Studies of the behavior and anatomy of C. elegans have determined that C. elegans has three major types of sensory receptors: chemoreceptors, thermoreceptors and mechanosensory receptors [3]. C. elegans behavior and its ability to adapt to environmental cues are mediated by these three modes of sensory input, and allows for interesting investigations into the neural circuitry and genetic pathways that underlie these systems.

Studies investigating chemotactic behavior have used behavioral analysis, genetic manipulation and laser ablation (killing) of specific neurons to lead to the identification of the 22 specific chemosensory neurons necessary for C. elegans to detect and respond to different chemical cues [4]. The 11 pairs of chemosensory neurons located in the anterior amphid sensillum (nose; see Fig. 1) have been divided into two separate classes responsible for either the sensation of taste or smell.C. elegans Neuroethology. Figure 1

Drawing of C. elegans showing some of the elements of the neural circuitry involved in chemotactic, thermotactic, and mechanosensory behaviors; sensory neurons represented by triangles; interneurons represented by circles; motor neurons represented by squares. The neurons shown are representative of the neurons used for these behaviors, not all neurons in the circuits described in the text are drawn.

As early as 1975, C. elegans was observed to show a behavioral response to thermal gradients [7]. When C. elegans are cultivated in a well-fed state at a given temperature and then transferred to an environment containing a thermal gradient they will migrate to 10.1007/978-3-540-29678-2_9 equivalent to the temperature in which they were raised (Fig. 2).C. elegans Neuroethology. Figure 2

Path of a C. elegans on an agar filled Petri dish with a radial thermal gradient of 15–25˚C after being well fed at 20˚C. Wild-type (N2) strain shows thermotactic behavior (a) Mutant strains, or laser-ablated wild-types can be abnormal showing cryophilic (b) thermophilic (c) or athermotactic behavior (d).

C. elegans responds to mechanosensory stimuli, in response to a touch to the head worms will swim backwards, in response to a touch to the tail worms will swim forwards. In addition, worms will swim backwards in response to a mechanical tap to their surroundings, such as within a laboratory Petri dish [8]. This tap-withdrawal response is a perfect example of how a seemingly simple behavior has more complexity than it originally appears; this response is plastic and shows short-term 10.1007/978-3-540-29678-2_8 and dishabituation, context conditioning and long-term memory for habituation [8]. Short-term habituation occurs with a series of taps separated by an inter-stimulus interval (ISI). Short ISIs lead to fast decreases in the withdrawal response, but this response is recovered quickly. Worms exposed to long ISIs, on the other hand, take longer to fully habituate, but have a much longer recovery period [8]. The tap-withdrawal response can also show long-term memory of habituation lasting at least 24 h [8]. The neural circuit for the tap response consists of the mechanosensory neurons ALM, AVM and PLM and the interneurons AVA, AVB, AVD and PVC [8]. Behavioral experiments led to the hypothesis that plasticity was occurring at the synapse between the sensory neurons and interneurons [8]. This synapse uses glutamate as its neurotransmitter, thus mutations that affect glutamate transmission affect both short-term and long-term memory. Mutations in the gene eat-4, which is responsible for expressing a pre-synaptic vesicular glutamate transporter, causes more rapid and complete short-term habituation of the tap-withdrawal response and no dishabituation [8]. When C. elegans have mutations in the gene, (glr-1) a post-synaptic glutamate receptor, no long-term memory of habituation is observed [9]. This indicates that glutamate transmission between mechanosensory neurons and interneurons is necessary for short-term and long-term memory for habituation of the tap-withdrawal response in C. elegans. This is of special interest because glutamate transmission has been implicated in long-term memory in many different animals [9], and may be part of a primitive, highly conserved mechanism of memory.

Natural variations in feeding behavior have been identified within C. elegans populations [10]. Wild type (N2) strains feed in a solitary fashion, they slow down when they contact food, and disperse randomly as they feed. C. elegans variants such as the RC301 strain and npr-1 mutants show a feeding pattern whereby they forage for food in groups, they speed up when they contact food and they congregate at the edges of food patches where levels of oxygen are low [10]. The differences between the solitary and social strains are the result of a single amino acid substitution in the gene npr-1 (a predicted G protein-coupled receptor of the neuropeptide Y receptor family). The npr-1 gene is expressed in several neurons (AQR, PQR and URX) that are exposed to C. elegans body fluid, and are involved in monitoring oxygen levels [10]. This research demonstrates how a “social” behavior can be dissected genetically, allowing a deeper understanding of the ways that neurons can produce and regulate behavior.

Detailed behavioral studies of the worm have shown that this tiny creature has a rich and complex behavioral repertoire, making it an ideal system in which to unravel the mysteries of the cellular control of behavior. Findings such as those described here contribute to the ever-growing database of knowledge within the C. elegans community. This database of knowledge (much of it available via the internet), coupled with the development of new investigative techniques, allows for research to be conducted at the behavioral, neuronal and genetic level making C. elegans an ideal model for the study of how genes regulate behavior.

Each of the examples that we have described show some of the ways that having all of the neurons in the nervous system identified and knowing their connectivity, as well as having a fully mapped and sequenced genome, has greatly aided the investigations of the roles of identified genes in behavior and extended our understanding of the neuroethology of C. elegans.

In a single vertebrate species such as humans, over 20 different subtypes of classical cadherins have been found so far, whereas the total number of non-classical cadherins is 80 at a moderate estimate.

Each vertebrate classical cadherin has five cadherin repeats, whereas the number of repeats and the presence or absence of other extracellular motifs are variable among invertebrate cadherins [1,2], even in a single species such as Drosophila (Fig. 1).

Compared to other cell-surface molecules, such as calcium-independent cell adhesion molecules, cadherins are unique in their ability to induce tight cell aggregates in which cells adhere to neighbors with maximum areas of contact. This ability appears to be dependent on the interaction of the cadherin–catenin multiprotein complex with the actin cytoskeleton. The classification of non-classical cadherins is based on sequence similarities of their intracellular regions and the overall structure of all domains. It is reasonable to assume that distinct subfamilies of non-classical cadherins are responsible for different functions and that they are coupled to different signaling pathways that have been under investigation.

Electrical signals conveyed along an axon are converted into chemical signals at the synapse. The basic mechanism for synaptic transmission was established at the frog neuromuscular junction, about 50 years ago, by Katz and his collaborators [4]. Ca²⁺ plays the central role in this story. The electrical signal at the presynaptic terminal results in a transient influx of Ca²⁺ through voltage-gated Ca²⁺ channels, which is detected by a specialized Ca²⁺ sensor. Through a series of molecular interactions, this signal causes rapid fusion of synaptic vesicles and transmitter release. Among Ca²⁺ -binding proteins in the presynaptic terminal, Syt is the most promising candidate for the Ca²⁺ sensor.

Synaptic transmission is malleable and plastic, which is believed to be the basis for memory and learning. Ca²⁺ again plays an important role in this process. CaMKII is the major player among Ca²⁺ binding proteins in the postsynaptic side of the synapse. This kinase has been studied extensively in the mammalian central nervous system as it constitutes the postsynaptic density, and is intimately involved in 10.1007/978-3-540-29678-2_12 of synaptic transmission in the brain.

Syt; Neurons, Endocrine and Exocrine cells, Presynaptic terminals, Synaptic vesicles, CaMKII; Mammalian brain, Hippocampus, Neurons, Synapses.

Syt: A major protein in the membrane of synaptic vesicles and dense core vesicles, which has an amino (N)-terminal transmembrane region followed by two C₂ domains (C2A and C2B). The C₂ domains are in the cytoplasm and uniquely situated to sense Ca²⁺ influx at the transmitter release site.

CaMKII: A major component in the postsynaptic density in the glutamatergic synapse.

Syt: Syt includes 13 isoforms in humans, and by the database search, an additional six potential isoforms have been suggested. Although most Syt’s are localized on transport vesicles, some (Syt III, VI, and VII) are present at the plasma membrane.

CaMKII, CaMKIIα, and CaMKIIβ are major isoforms in the brain.

Syt: Inter-neuronal communication, regulation of hormone secretion.

CaMKII: Postsynaptic modulation of synaptic transmission, synaptic plasticity, memory and learning.

Syt: 10.1007/978-3-540-29678-2_22, transmitter release.

CaMKII: Regulation of channel permeability and insertion of AMPA-type glutamate receptors in the postsynaptic density.

CaMKII: Extensive synaptic activities translocate CaMKII from cytosol in the postsynaptic 10.1007/978-3-540-29678-2_19 of dendrites to the postsynaptic density.

The CaMKII family is encoded by four different genes α, β, δ and γ, from which at least 28 isoforms can be derived. The gene encodes for three principal domains in the protein: catalytic, autoinhibitory and association domains. In the brain, the α and β genes are predominantly expressed. In the cell, the enzyme is not expressed as a single protein, but rather as a homo- or heteromeric complex (holoenzyme) of 12 subunits. The 12 catalytic and regulatory domains form 2 hexameric rings on each side of a central core comprising 12 association domains (Fig. 1).Calcium/Calmodulin-dependent Protein Kinase ll in Neurons. Figure 1

Structure of CaMKII. (a) Primary structure of CaMKII domains. (b) Model structure of CaMKII dodecamer. (c) Model of CaMKII activation by Ca²⁺ spikes [1,3].

CaMKII is implicated in various cellular functions, including the regulation of carbohydrate metabolism, membrane current, 10.1007/978-3-540-29678-2_14 synthesis and release, cytoskeletal organization, intracellular Ca²⁺ homeostasis, 10.1007/978-3-540-29678-2_20, synaptic plasticity and some forms of memory [1]. Here, we will focus on CaMKII implications in gene regulation and synaptic plasticity.

CaMKII is a multifunctional kinase highly abundant in the brain and critically involved in the regulation of several functions in neurons. Its unusual structural and regulatory properties support its multiple functions by enabling the enzyme to decode specific patterns of Ca²⁺ signals in time and in space, for brief and long-lasting effects. Since the discovery of CaMKII by Howard Schulman and colleagues three decades ago, many studies have unraveled some of its functions and mechanisms of action, but many aspects are still unresolved. Nevertheless, what we have learned so far about this enzyme has contributed to a better understanding of how signaling in cells can be achieved with specificity, multiplicity, efficiency, and persistence.

The principal pore-forming subunit of both LVA and HVA channels is the α₁-subunit, a high- molecular weight protein (190–250 kDa), which is structurally similar to the 10.1007/978-3-540-29678-2_14 α-subunit [1]. It is formed by four domains (I–IV) linked together in a single polypeptide chain and each domain contains six putative transmembrane segments (S1–S6), plus a loop (P) that dips partially into the pore to form the pore-lining region (Fig. 1). The cytoplasmic loops linking the four domains are structurally important for their interactions with β-subunits, second messengers, membrane binding proteins and channel 10.1007/978-3-540-29678-2_7.Calcium Channels — An Overview. Figure 1

Subunit structure of voltage-gated Ca²⁺ channels: transmembrane topology of the α₁ and associated auxiliary subunits (β, α₂-δ, γ). Predicted α-helices are depicted as cylinders. For the α₁-subunit the transmembrane spanning α-helices (1 to 6) are repeated in the four domains (I to IV). Red cylinders indicate the positively charged S4 segments (voltage sensors), and the thick green lines denote the pore loops (P) that line the permeation pathway for Ca²⁺. The lengths of lines are not intended to represent the exact lengths of the polypeptide segments indicated. The same is for the size of various subunits. Adapted and redrawn from ref [2].

As for other voltage-gated ion channels, the probability of Ca²⁺ channels opening is strictly voltage-dependent, i.e., the switch from a closed (non-conductive) to an open (conductive) configuration is strictly dependent on voltage, usually requiring 4–8 mV to change e-fold the probability of channel opening [6]. Structure-function studies have associated this property to the presence of positively charged lysine and arginine residues distributed in the S4 transmembrane segment of each domain, which form the “ 10.1007/978-3-540-29678-2_22 ” (Fig. 1). However, the exact location of the activation gates that regulate the open and closed states of the pore remains unknown.

During maintained depolarization, Ca²⁺ channels tend to inactivate, but the speed and extent of channel inactivation may vary dramatically. In general, Ca²⁺ channels inactivate by either Ca²⁺- or voltage-dependent mechanisms. Ca²⁺ -dependent inactivation is dominant for the cardiac Ca_v1.2 L-type channel, but may also occur for other HVA channels. Ca²⁺-dependent inactivation is a calmodulin-dependent process, involving sites on the C-terminal domain of the α₁-subunit. Voltage-dependent inactivation is a general term for inactivation that does not clearly depend on Ca²⁺. T-type Ca²⁺ channels tend to inactivate rapidly and almost completely, while inactivation of HVA channels is usually slow and incomplete. Spontaneous switch between inactivating and non-inactivating modes have been observed in single N-type channels, perhaps reflecting modulation by some intracellular signaling process. However, how Ca²⁺-binding or membrane voltage affect the inactivation gate is still widely unknown [6].

The present view of Ca²⁺ channel permeability is based on the existence of an intrapore binding site controlling both ion selectivity and channel block: the selectivity filter. This highly specialized region of Ca²⁺ channels consists of a ring of four negative charged groups inside the pore. In HVA channels, each of the four P loops contains a glutamate forming the EEEE locus, in T-type channels two glutamates are substituted by two aspartates in the corresponding position (EEDD locus). The spatial arrangement of the four negative charges in the P loops is postulated to closely coordinate two Ca²⁺ ions whose sequential entrance and subsequent interaction induce high Ca²⁺ fluxes while preserving high affinity for the pore-site [7]. Although this simple ion-ion interaction would explain the dual nature of Ca²⁺ as permeant ions at millimolar concentrations and as blockers of Na⁺ currents at micromolar concentrations, the complete understanding of Ca²⁺ channel permeability is likely to require a further hypothesis of ion-pore structure interactions. The recent availability of T-type channel clones [5] has allowed closer comparisons between LVA and HVA channel permeability properties, highlighting the role that the EEEE or EEDD locus play in the regulation of ion selectivity in the two channel groups. Ca_v3.1 T-type channels have apparently a narrower pore size (5.1 Å diameter) compared to the Ca_v1.2 L-type pore size (6.2 Å diameter) [7]. This structural difference may explain the different Ca²⁺/Ba²⁺ selectivity and blocking action of Cd²⁺ and Ni²⁺ of the two channel families (LVA and HVA): the L-, N- and P/Q-type channels being more permeable to Ba²⁺ than Ca²⁺ and more sensitive to the block by Cd²⁺ and the T-type channels being equally permeable to Ca²⁺ and Ba²⁺ and more sensitive to the block by Ni²⁺ (Table 1).Calcium Channels — An Overview. Table 1

Selective and often used blockers, distribution and established channelopathies [10] of the various voltage-gated Ca²⁺ channels

In the following paragraphs are described the tissue and cellular location, physiological role, pharmacology and related Channelopathies associated to each Ca²⁺ channel type as summarized in Table 1.

L-type channels (Ca_v1). L-type channels are widely expressed in many tissues and control a number of Ca²⁺-dependent responses in excitable cells. In central neurons they are preferentially located on proximal dendrites and cell bodies and are involved in postsynaptic integration, neuronal plasticity, gene transcription and mood behavior. In sensory neurons (10.1007/978-3-540-29678-2_8), they directly control neurotransmitter release and sensory perception. The L-type family includes four α₁-subunits (Ca_v1.1 to Ca_v1.4) with different structure-function characteristics but common blockers: 1,4-dihydropyridines (DHPs), phenylalkylamines and benzothiazepines. With the exception of Ca_v1.3 and Ca_v1.4, which activate at relatively low voltages, the other Ca_v1 channels activate at voltages much more positive than resting potential. Activation is fast and sharply voltage-dependent while inactivation is relatively slow in the presence of Ba²⁺ but speeds-up in the presence of Ca²⁺ (Ca^2±-dependent inactivation). Deactivation (giving rise to 10.1007/978-3-540-29678-2_20) is also fast, ensuring rapid closing of the channels on membrane repolarization to resting levels. L-type channels are distinguished from the other Ca_v channels by their high sensitivity to DHPs. DHP antagonists (nifedipine, nitrendipine) reversibly block the channels and help quantifying the amount of L-type channels expressed in a cell, while DHP agonists (Bay K 8644) prolong the open state of the channel, producing slow tail currents near resting potential. As such, DHP agonists allow measuring the 10.1007/978-3-540-29678-2_19 of high conductance L-type channels in membrane patches, otherwise hardly detectable [8]. L-type channels can be often distinguished from the other Ca²⁺ channels for their cAMP/PKA-mediated up-regulation which causes increased mean open times and probability of openings at the single channel level and increased Ca²⁺ current amplitude in whole-cell recordings [9]. Neuronal and neuroendocrine L-type channels can be also inhibited by a fast 10.1007/978-3-540-29678-2_7 mechanism activated by neurotransmitters which could be at the basis of an 10.1007/978-3-540-29678-2_1 of hormone release.

P/Q-type channels (Ca_v2.1). The Ca_v2.1 family includes two Ca²⁺ channels which are nearly indistinguishable except for their different affinity to a common blocker: the spider venom ω-agatoxin IVA. P-type channels are more sensitive to ω-agatoxin IVA (K_d 1–3 nM) than Q-type channels (K_d 100–200 nM). Ca_v2.1 channels are widely expressed in neurons but are also available in pancreatic, pituitary and chromaffin cells. Their main physiological function is to control neurotransmitter release in central neurons and mammalian 10.1007/978-3-540-29678-2_14 where they are highly expressed at the presynaptic sites. P/Q-type channels control also the excitation-secretion coupling in pancreatic and chromaffin cells. Ca_v2.1 channels play a key role in neurotransmitter release due to their high-density of expression at the release sites of central synapses and share most of the modulatory properties of N-type channels described below. Similarly to the N-type channels they bind tightly to the 10.1007/978-3-540-29678-2_19 proteins at the 10.1007/978-3-540-29678-2_19 of the II-III linker of the channel. P/Q-type channels are also effectively inhibited by the neurotransmitter activated GPCRs mechanism described below. Missense mutations of P/Q-type channels cause 10.1007/978-3-540-29678-2_6 associated to an apparent gain of function as a result of an increased probability of channel openings and alteration of synaptic transmission (Table 1).

N-type channels (Ca_v2.2). Ca_v2.2 channels are widely expressed in the central and peripheral nervous system and chromaffin cells. They are highly expressed at the nerve terminals, where they control neurotransmitter release, and to a minor degree at the dendritic sites, where they are involved in Ca²⁺ signaling. Ca_v2.2 channels control also hormone release in neuroendocrine cells. The channels activate at relatively high voltages. Maximal activation at positive potentials and deactivation on return to resting levels are both fast. Inactivation is variable but significantly faster than L-type channels and slower than T-type channels. Ca_v2.2 channels are insensitive to DHPs but are selectively blocked by ω-conotoxin GVIA and related cone snail toxins (Table 1).

N-type channels play a key role in neurotransmitter release due to their high-density of expression at the release sites and their tight binding to the SNARE complex of the vesicle release machinery (syntaxin 1A, SNAP-25, VAMP2/synaptobrevin and synaptotagmin). These proteins bind at the synprint motif of the II-III linker of the channel and the tight interaction affects the availability and gating of the channel. The SNARE complex in fact either steadily inactivates the channel or inhibits its activation through a G_βγ subunit. Interestingly, N-type channels are effectively modulated by GPCRs activated by neurotransmitters and the mechanism is thought to be at the basis of 10.1007/978-3-540-29678-2_16. Briefly, an activated G_βγ subunit binds directly to the pore-forming α₁-subunit of the N-type channel and shifts the gating mode from “willing” (from which the channel readily opens) to “reluctant” (from which the channel opens less frequently). The modulatory mechanism is 10.1007/978-3-540-29678-2_13, voltage-dependent and causes an increased delay of the channel opening which produces an overall slow activation of the “reluctant” channel. Strong depolarizations opening the channels reduce the affinity of G_βγ for the α₁-subunit and the channel recovers its normal gating mode.

R-type channels (Ca_v2.3). Ca_v2.3 channels are widely expressed in the central nervous system at the cell bodies, dendrites and presynaptic terminals. They are also expressed in 10.1007/978-3-540-29678-2_13, heart, pituitary and chromaffin cells. The Ca_v2.3 channel has been originally reported to encode a Ca²⁺ channel type with biophysical properties between LVA and HVA channels, or usually as an HVA channel resistant to DHPs, ω-toxins and thus called R-type (for “residual”). Ca_v3 channels are likely to form a family of several channels with fast activation but variable inactivation that could be fast and comparable to the Ca_v3 types or slow like the Ca_v1 channels. They are involved in neurotransmitter and hormone release, repetitive firing (→ 10.1007/978-3-540-29678-2_1) and 10.1007/978-3-540-29678-2_12. The tarantula toxin SNX-482 blocks exogenously expressed Ca_v2.3 currents but is only partially or not effective on native R-type currents, suggesting that Ca_v2.3 does not always conduct a significant portion of the R-type current which remains after blocking all the other voltage-gated Ca²⁺ channels. Ca_v2.3 channels are also sensitive to small doses of Ni²⁺. In some case the Ni²⁺ block has K_d comparable to that of the Ca_v3.2 T-type channel described below.

T-type channels (Ca_v3). Ca_v3 channels (Ca_v3.1 to Ca_v3.3) are ubiquitously expressed and sustain key physiological functions which derive from their unique properties [4,5]: (i) they activate and inactivate at unusually negative voltages and are responsible for a window-current near resting potentials, (ii) they exhibit fast and complete inactivation during sustained depolarization and deactivate slowly on repolarization, (iii) they are equally or slightly more permeable to Ca²⁺ than Ba²⁺ and have small single channel conductance, (iv) they outlast 10.1007/978-3-540-29678-2_13 and (v) they are preferentially blocked by low doses of mibefradil and Ni²⁺ (particularly the Ca_v3.2) (Table 1). At present, Ca_v3 channels are recognized to play a critical role in many physiological functions in which a low-threshold Ca²⁺ entry is required to trigger, or sustain, specific cell activities. This is particularly true for the generation of low-threshold spikes, pacemaking activity, hormone secretion, cell growth and differentiation. T-type channels play also a critical role in several pathologies in which either their recruitment, overexpression, or altered gating cause cardiac hypertrophy, hypertension, heart failure, 10.1007/978-3-540-29678-2_12, 10.1007/978-3-540-29678-2_14, and 10.1007/978-3-540-29678-2_16. Most recently, T-type channels are shown to control the vesicular release of neurotransmitters in neurons, and very recent data indicate that they are involved in fast catecholamine release in adrenal chromaffin cells.

The nervous system relies on charged molecules to relay information. Whereas the relatively inert Na⁺ and K⁺ ions serve mostly to establish membrane potential and carry action potentials along excitable membranes, Ca²⁺ has properties that allow it to accomplish additional functions. First, it has a unique affinity for binding ligands containing oxygen-donating groups such as carboxyls, carbonyls, ethers, and alcohols. Second, Ca²⁺ is kept at low levels within the cytosol since it precipitates organic anions and is toxic to cells at high concentrations. These features likely lead to the exploitation of the calcium ion not only as a charge carrier but also as a signaling molecule.

Nerve activity promotes subcellular calcium “hotspots” or microdomains around individual channels, and collectively these active channels contribute to global cellular processes including synaptic plasticity and neuronal survival. An example is long-term memory which involves synaptic connectivity changes activated via 10.1007/978-3-540-29678-2_12. Strengthening of synapses induced by LTP requires gene transcription mediated by calcium-sensitive signaling pathways [1]. Furthermore, activity-dependent neuronal survival, which is important during cognitive development, also requires calcium-dependent gene transcription [2].

“First responder” genes are termed 10.1007/978-3-540-29678-2_9, whose transcription is upregulated without a requirement for newly synthesized proteins. The expression of IEGs is tightly regulated in space and time in active neurons by the summation of nerve inputs and local Ca²⁺ flux. 10.1007/978-3-540-29678-2_2 is a classical calcium-activated IEG that mediates activity-dependent neuronal survival and is required for establishment of LTP (reviewed in [3]). BDNF is a neurotrophin that activates downstream signaling pathways through binding to 10.1007/978-3-540-29678-2_18 and low affinity receptor p75 (10.1007/978-3-540-29678-2_16) to exert immediate and long term effects on synaptic activity. Not surprisingly, mutations in the BDNF gene have been correlated with neuronal dysfunction including impaired episodic memory [3]. Numerous other genes encoding transcription factors, growth factors, cytoskeletal proteins, and proteins involved in signaling and metabolism have also been classified as IEGs including c-fos, fos B, c-jun, jun B, zif268, Egr-3, Cox-2, Rheb, Arc, Narp, β-activin, Homer, and nur77; many have been implicated in aspects of neuronal development and survival, and synaptic plasticity changes [2,4].

Calcium signaling depends on a cascade of electrochemical and biochemical events. These include stimulus sensation and opening of a channel or receptor and influx of Ca²⁺ into the cytosol, binding of Ca²⁺ ions to intracellular ligands, and signaling to terminal effectors. Ca²⁺ influx occurs along steep electrochemical gradients across the plasma membrane, through highly-selective channels or receptors including VGCCs and NMDA receptors, or from intracellular stores through ryanodine and inositol 1, 4, 5-triphosphate receptors (IP₃). Cytosolic Ca²⁺ is quickly buffered and limited to transient microdomains (Calcium Microdomains) surrounding the pores of open channels (Fig. 1c).Calcium Channels: Regulation of Gene Transcription. Figure 1

(a) Illustration of the subunit complex of high voltage-activated calcium channels. Calmodulin or CaM (shown in red) containing EF-hand motifs, bind the C-terminus of the α₁ subunits of calcium channels. (b) Cleaved C-termini of Ca_v1.2 channels translocate to the nucleus to activate gene transcription. A β₄ subunit isoform also translocates to the nucleus and interacts with nuclear factors involved in gene expression. (c) L-type channels mediate Ca²⁺ signaling via CaM bound to their C-termini. CaM is activated by Ca²⁺ entry through L-type (Ca_v1.2) channels and mediates both calcium-dependent inactivation and gene transcription.

Calcium/cAMP Response Element Binding protein (CREB) is a transcription factor whose activity is implicated in calcium-related signaling events and is activated through phosphorylation of a serine residue (Ser-133). Cytosolic Ca²⁺ microdomains trigger CREB phosphorylation via several pathways with distinct temporal kinetics, the most significant of which are the 10.1007/978-3-540-29678-2_13 and the calcium/calmodulin-dependent kinasePathway (CaMK) (reviewed in [3], Fig. 1c). Whereas the CaMK pathway is involved in early phosphorylation of CREB, the Ras/MAPK pathway is required for sustained CREB phosphorylation [5]. Both pathways are critical since a sustained CREB phosphorylation is required for transcriptional activation [5]. Phosphorylated CREB promotes transcription of a variety of IEGs including BDNF and mediates both activity-dependent neuronal survival and synaptic plasticity [3].

Besides CREB, additional calcium-dependent transcription factors have been identified. Like CREB, these are typically regulated by CaM or downstream members of the CaM signaling pathway, and translocate to the nucleus upon activation. These include:

10.1007/978-3-540-29678-2_14: NF-AT has been implicated in synaptic plasticity [reviewed in 6]. NF-AT translocates to the nucleus upon dephosphorylation by calcineurin, a protein phosphatase controlled by the CaMK pathway.

Of the known VGCCs in the nervous system (i.e., Ca_v1 (L-type), Ca_v2 (N-, P/Q-, R-types), and Ca_v3 (T-type)), L-types are the only channel type established to couple excitation with transcription. Selectively blocking L-type channels with dihydropyridines inhibits the expression of many IEGs. Intracellular microdomains of Ca²⁺ generated by L-type channels lead to both nuclear signaling and negative feedback regulation of channel gating itself by means of calcium-dependent inactivation. It has been reported that both processes are dependent on the tethering of calcium-sensitive molecules to the channel where they can sense the local increase in calcium concentration. CaM, a cytosolic protein able to directly bind calcium via its EF-hand motifs, remains tethered to L-type (Ca_v1.2) channels via an isoleucine-glutamine (IQ) motif in the channel’s C-terminus. Disrupting either the ability of CaM to interact with L-type channels or CaM’s ability to bind Ca²⁺ leads to loss of calcium-dependent inactivation [9] (Fig. 1c). Subsequent research from Dolmetsch and colleagues illustrates that CaM also mediates nuclear signaling by L-type channels [5] (Fig. 1c). This research group assayed the phosphorylation of CREB and activation of the Ras/MAPK signaling pathway in neurons where endogenous L-type channels had been blocked by dihydropyridines. By transfecting functional drug-resistant L-type channels into cultured primary neurons, Dolmetsch et al. showed that mutating the CaM-binding IQ motif or co-expressing calcium-insensitive CaM abrogated calcium-dependent activation of the Ras/MAPK pathway, CREB phosphorylation, and activation of MEF2. As illustrated in Fig. 1c, this research establishes CaM as a pivotal regulatory center for L-type channels, in channel self-regulation and the feed-forward control of gene expression.

Some intriguing twists in the story of how L-type channels modulate gene transcription are now emerging. The C-terminus of Ca_v1.2 translocates to the nucleus as a ~75 kDa peptide (termed CCAT) after proteolytic cleavage at the cell membrane. In the nucleus, CCAT peptide binds transcriptional regulator p54(nrb)/NonO, associates with endogenous promoters, and promotes a significant up-regulation of 16 genes (e.g., axon guidance factor 10.1007/978-3-540-29678-2_14 and gap junction protein Cx31.1) and down-regulation of 31 genes (e.g., 10.1007/978-3-540-29678-2_14and the 10.1007/978-3-540-29678-2_14Scl8A1) [10] (Fig. 1b). Nuclear translocation of CCAT is observed in several central neuron types, especially in adults and heterologous expression of CCAT promotes increased neurite outgrowth in cultured primary neurons. Interestingly, CCAT translocation requires low cytosolic Ca²⁺ levels suggesting a mode of transcriptional regulation unlike the CaM dependent pathway which requires Ca²⁺ and neuronal activity. It has been suggested that a CCAT-like cleavage mechanism may be widely applicable since the C-terminus of the P/Q-type channel (Ca_v2.1) is also cleaved and translocates to the nucleus; yet, it remains to be determined whether truncated Ca_v2.1 peptides affect gene expression.

With the exception of T-type channels, the pore-forming α₁ subunits of VGCCs associate with accessory subunits (termed α₂, δ, β, and γ) which significantly modulate the electrophysiological properties and surface expression of the α₁ subunits (Fig. 1a). Recent evidence suggests that a short β₄ isoform translocates to the nucleus and interacts with heterochromatin protein 1γ to promote gene silencing, however the underlying mechanisms and significance remains to be elucidated (Fig. 1b).

Calcium flux through post-synaptic NMDA receptors shares many of the downstream targets as L-type calcium channels to promote gene expression and neuronal survival. The similarities with L-type channels include activating CREB by phosphorylation, promoting the expression of some IEGs [8], and nuclear signaling via the CaMK and the Ras/MAPK signaling pathways. It has been reported that different calcium sources may be required for different phases of gene transcriptional activation. In particular, the NMDA receptors mediate an earlier phase of CREB phosphorylation (i.e., 10 min after depolarization), while the L-type channels are required in later stages. NMDA receptors also operate through downstream effectors associated with LTP induction, that do not overlap with L-type calcium channels. For example, EphB receptor tyrosine kinases localize at excitatory synapses and enhance the expression of NMDA receptor-mediated genes, but not those mediated by L-type channels. Interestingly, variations in the composition or location of the NMDA receptors create highly divergent signaling. Whereas synaptic NMDA receptors mediate neuronal survival, calcium entry via extra-synaptic receptors inhibits CREB phosphorylation and promotes cell death. Furthermore, changes in the composition of NMDA receptor subunit alters MAPK pathway signaling leading to either LTP or LTD (10.1007/978-3-540-29678-2_12).

Internal calcium sources may also participate in regulating gene transcription, but there is not much evidence for this mechanism occurring in neurons. IP₃ receptor activity can mediate the phosphorylation of CREB and activate NF-AT and NF-κB transcription factors in both muscle and neuronal cells (reviewed in [6]). Ryanodine receptor-mediated Ca²⁺ influx has been linked to activation of the transcription factor NF-AT and expression of the slow myosin heavy chain I gene in muscle (reviewed in [6]).

Gene transcriptional changes are the capacity of neurons to adapt to changing conditions, such as to modify their structure, their survivability and enhancement of connectivity. Regulation of gene expression is necessary in neural development, for the process of learning and memory, and for apoptosis. Importantly, regulatory mechanisms may be harnessed for possible treatment of neurodegenerative disease.

Specificity in gene regulation is governed by nerve activity and calcium flux. Different calcium sources mediated through L-type calcium channels as well as NMDA receptors and intracellular sources converge through multiple, and often overlapping, signaling pathways (e.g., CREB, CaMK, Ras/MAPK and cAMP dependent protein kinase A (reviewed in [3,6])). These likely contribute collectively to global cellular changes in nuclear Ca²⁺ to enhance the transcription of IEGs. At a local level, calcium is sensed by tethered CaM at the C-terminus of L-type calcium channels. CaM serves as a regulatory control point for tuning up or down calcium channel activity and its downstream regulators of gene transcription. Interestingly, low calcium also appears to be a signal for gene expression, as both the transcriptional modulator DREAM as well as the C-terminal cleavage product CCAT of L-type channels, require low nuclear and cytosolic Ca²⁺ levels respectively for activation [8,10]. Future experimentation will further clarify how the varied signaling mechanisms activated by calcium, coming through calcium channels or other sources (eg. NMDA receptors), contribute to the regulation of gene transcription.

The majority of the biological effects of cannabinoids are mediated through two class A (10.1007/978-3-540-29678-2_18-like) 10.1007/978-3-540-29678-2_7 (GPCRs) termed cannabinoid receptor 1 (CB1) and 2 (CB2). Although THC and a number of other cannabinoids have similar affinities for these receptors, their amino acid sequences share only 44% identity and their distributions in the body and 10.1007/978-3-540-29678-2_2 are very different. Not surprisingly therefore, the receptors are responsible for different functions.

CB1 is expressed in peripheral tissues including the lungs, liver, kidneys and reproductive organs, and is extremely abundant in the mammalian CNS. Distribution is widespread with particularly high levels present in the 10.1007/978-3-540-29678-2_2, 10.1007/978-3-540-29678-2_8, cerebellum and cortex (Fig. 1); areas which correlate well with the effects of cannabinoids on memory, perception and movement control. Consistent with the lack of lethality of cannabis administration, CB1 is essentially absent from regions of the brainstem controlling cardiovascular and respiratory functions [1]. Splice variants of CB1 have been reported, however their physiological relevance is not yet established.Cannabinoids. Figure 1

Immunohistochemical localisation of the CBI receptor in the mouse brain.

Several endogenous cannabinoid ligands (endocannabinoids) have been isolated and demonstrated to bind to cannabinoid receptors. These compounds are all derived from arachidonic acid. To date the most studied of these are 10.1007/978-3-540-29678-2_1 and 10.1007/978-3-540-29678-2_1 (Fig. 2). 2-arachidonoyl glycerol exhibits higher selectivity and efficacy for CB1 and CB2 receptors than anandamide, which also interacts with non-cannabinoid receptor targets (e.g., 10.1007/978-3-540-29678-2_20). The endocannabinoids are synthesized “on demand” in response to increased intracellular calcium, and released from cells immediately following their production. Recent studies have identified synthesizing and degradative enzymes for both of these compounds, and these enzymes provide new drug targets for the manipulation of endocannabinoid levels in vivo [3].Cannabinoids. Figure 2

Structures of the endocannabinoids.

Cannabinoids have been implicated in many neurological diseases and conditions that originate in or are mediated by the CNS. Those where aberrations in the endocannabinoid system have been linked to disease mechanisms and pathology include neuropsychiatric diseases such as Tourette’s syndrome, 10.1007/978-3-540-29678-2_2 and 10.1007/978-3-540-29678-2_19, and neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS; 10.1007/978-3-540-29678-2_12), 10.1007/978-3-540-29678-2_13 (MS) and 10.1007/978-3-540-29678-2_1, 10.1007/978-3-540-29678-2_16, and 10.1007/978-3-540-29678-2_8 diseases. Additionally, cannabinoids are postulated as symptomatic therapeutics for conditions including 10.1007/978-3-540-29678-2_19, 10.1007/978-3-540-29678-2_5, glaucoma, obesity and 10.1007/978-3-540-29678-2_1.

Few cannabinoid compounds are currently approved for pharmaceutical use (Fig. 3). Synthetic THC is available for prescription in capsule form as dronabinol (marketed as Marinol^®) and an analog nabilone (marketed as Cesamet^®). These products were originally approved for the suppression of nausea in cancer sufferers and were later approved as appetite stimulants for use by patients with HIV and cancer, who commonly experience a wasting syndrome. Subsequent approval has been granted for treatment of MS 10.1007/978-3-540-29678-2_14. An aerosol form of dronabinol is under investigation for migraine treatment in addition to the existing indications (http://www.marinol.com, http://www.cesamet.net).Cannabinoids. Figure 3

Cannabinoid compounds currently approved for therapeutic use.

Molecules that mediate cell-to-cell communication, such as hormones and neurotransmitters, are packaged and stored within cells in membrane-delimited vesicles. In neurons and other excitable cells, the trigger for release of these signaling molecules is influx of Ca²⁺ through voltage-gated Ca²⁺ channels upon membrane depolarization. The consequent elevation of the intracellular Ca²⁺ concentration ([Ca²⁺]_i) triggers the fusion of vesicles with the cell plasma membrane, and release of the vesicle contents to the extracellular space by the process of exocytosis. The vesicle membrane is taken up again into the cell by the process of endocytosis. Application of membrane capacitance measurements, in conjunction with other techniques, has greatly increased our understanding how Ca²⁺triggers exocytosis and how specific proteins and second-messenger cascades regulate exocytosis and endocytosis.

The capacitance of the cell membrane is proportional to its area, thus the insertion of vesicle membrane during exocytosis results in an increase in capacitance, whereas endocytosis decreases membrane capacitance. Just as patch-clamp techniques can resolve the currents due to the opening and closing of individual ion channel molecules, capacitance measurements from small membrane patches can resolve the fusion (exocytosis) and fission (endocytosis) of individual synaptic vesicles with millisecond time resolution [1].

The modern era of membrane capacitance measurements dates to the early days of the patch-clamp recordings. In a seminal report in 1982, Neher and Marty described changes in membrane capacitance associated with exocytosis in chromaffin cells [2]. Unlike single-channel recording, however, capacitance measurement was slow to catch on as a mainstream approach and only a handful of laboratories used the technique in the 1980s. Capacitance measurements blossomed in the 1990s, partly due to the emergence of powerful software that greatly simplified application of the method. The capacitance technique continues to make a large contribution towards a mechanistic understanding of exocytosis and endocytosis.

Present here is only a very brief survey of techniques for estimating changes in membrane capacitance related to exocytosis and endocytosis, see [11] for a detailed treatment. Most capacitance techniques are based on the three component equivalent circuit of the recording configuration depicted in Fig. 1. In this circuit, we neglect the pipette capacitance because the current passing through this pathway is electronically subtracted upon proper adjustment of the pipette capacitance compensation circuitry of the patch-clamp amplifier. During whole-cell recording of a neuroendocrine cell, typical parameters are C_m ~ 6 pF, R_A ~ 8 MΩ, and R_m > 2 GΩ. For on-cell recording, R_A is somewhat smaller and C_m is a fraction of a pF.

The most common approach to estimating capacitance changes is to apply a sinusoidal voltage stimulus and analyze the resulting sinusoidal current. The amplitude of the stimulus sinusoid is usually less than 50 mV, whereas the frequency ranges from ~ 1 kHz for whole-cell recordings to 50 kHz or higher for on-cell measurements. A significant limitation is that a single sinusoid only provides two pieces of information (magnitude and phase), whereas there are three unknown components of the equivalent circuit. Next are described two approaches used to obtain the additional information needed.

In the “sine + dc” approach, the dc (average) current is measured and used, together with an estimate of the extrapolated zero-current potential, to estimate the dc conductance (R_A + R_m) [12]. This approach is incorporated in “Pulse” and “PatchMaster” (HEKA Inc., Lambrecht, Germany) software, and is probably the most widely used approach currently for capacitance measurements in the whole-cell configuration. The simplest version of the sine + dc approach assumes the reversal potential does not change during the recording. The sensitivity of capacitance estimates to errors in the value of the assumed reversal potential is small if R_m is high (GΩ range). More complicated multi-sinusoid or square-wave approaches are necessary if the R_m is both small and changing.

The piecewise-linear approach is the original implementation of the patch-clamp capacitance technique by Neher and Marty [2]. A single-frequency sinusoid voltage stimulus is applied and a phase-sensitive detector (lock-in amplifier) is connected to the patch-clamp amplifier, to extract the component of the sinusoidal current that is proportional to changes in membrane capacitance. Thus, the output signal of the lock-in amplifier, when set to an appropriate phase, is directly proportional to changes in membrane capacitance but has little sensitivity to changes in R_A or R_m. Dithering of the membrane capacitance compensation knob of the patch-clamp amplifier is used both to find the appropriate phase setting of the lock-in amplifier and to calibrate the capacitance signal. A subsequent variant of the technique (“phase tracking”) uses computer-controlled dithering of series resistance introduced between the bath and ground to find the appropriate phase setting [13]. The piecewise-linear approach was initially quite popular because it can be implemented entirely in hardware (e.g. in the Cairn Optopatch patch-clamp amplifier), however, it has largely been replaced by more powerful (and less ad hoc) computational approaches for whole-cell capacitance measurements.

A number of multi-sinusoid approaches are also used. In the case of two sinusoids there are 4 pieces of information to determine the values of the three unknown parameters. Optimal use of the information to “fit,” i.e. estimate the parameters with the minimal variance “noise,” is a complex problem that is thoroughly addressed in an elegant study by Barnett and Misler [14]. A number of sub-optimal, but computationally simpler, ad hoc approaches to estimate the equivalent circuit parameters from dual sinusoid excitation have also been implemented. A challenging problem with all multi-sinusoid approaches is optimizing the choice of amplitudes and frequencies of the stimuli, and therefore these approaches tend to give noisier estimates of membrane capacitance than single-sinusoid techniques. In addition, the non-ideal frequency dependence of C_m results in an underestimation of R_m (unpublished observations).

Approaches using square-wave stimuli are also used. In response to a square step in pipette potential, the equivalent circuit depicted in Fig. 1 will respond with a transient current that decays with an exponential time course.

Fit of the current transient to an exponential function can be used to produce estimates of the three circuit parameters. An excellent implementation of this approach is described in [15]. This approach can be quite robust in that estimates of C_m can be generated that are quite insensitive to changes in R_A and R_m, and C_m estimates can be produced with nearly as low a noise as sinusoidal techniques [15]. This method requires a high bandwidth setting of the patch-clamp amplifier, a high sampling rate and is computationally intense. Nevertheless, an efficient algorithm running on a modern computer can generate estimates at ~100 Hz [15].

Any of the above techniques, when carefully applied, can produce valid estimates of changes in membrane capacitance related to exocytosis and endocytosis, so the choice of technique often depends on finding an attractive software package. In general, for capacitance techniques to be truly useful, they must be embedded within a powerful, flexible software package that is capable of executing complex stimulus protocols while recording and displaying multiple data streams in real time.

The equations governing the mechanics of both solid and fluid components are derived from the physical principles of conservation of mass and conservation of both linear and angular momentum. These equations are formulated in terms of stress tensors and either strain tensors for solid mechanics or strain-rate tensors for fluid mechanics. For the solid soft tissue component the displacements and strains are large (unlike those occurring in bone and in fact most engineering materials) and therefore the theoretical framework requires the use of large deformation elasticity theory. The relationship between stress and strain (or strain rate) is determined by the material properties of the tissue and has the following characteristics:

Stress-strain relations are highly nonlinear and typically exhibit “strain hardening” or stiffening with increased strain. This is illustrated in Fig. 1 where components of axial stress are plotted against the corresponding axial strain in experiments on a small block of myocardial tissue [1]. As the strain increases the material becomes stiffer (the slope of the stress-strain curve increases).

Both soft tissues and blood are usually considered to be incompressible (a consequence of the high water content) and for most soft tissues viscous behavior plays a small but important damping role. For the muscular components of the cardiovascular system (e.g. heart muscle and arterial smooth muscle) the mechanical function is also greatly influenced by the myofilament-generated forces that produce active muscle contraction.

Computational analysis of large deformation soft tissue mechanics is usually performed with finite element techniques, which divide the material into small coupled blocks or “elements” (see Fig. 3) and then apply integral stress balance equations to these elements. Continuous fields representing the geometric and material properties of the tissue are defined via nodal parameters, which are defined on the element boundaries and shared between adjacent elements in order to ensure continuity of the fields both within and across the elements. The equations governing blood flow are usually solved with finite element techniques, finite volume techniques or finite difference techniques. In this last case a Taylor series approximation of the governing Navier-Stokes partial differential equations produces a system of discrete equations for computation. Computational analysis of coupled solid-fluid structures is difficult and usually requires the two sets of equations to be solved simultaneously to achieve a converged solution.Cardiovascular Mechanics. Figure 3

A finite element mesh used for solving the large deformation mechanics of the left and right ventricles. The surfaces shaded red are the internal surfaces – endocardium – surrounding the left and right ventricles, and the outer translucent surface is the epicardium. The holes at the top are the orifices regulated by the four cardiac valves – the mitral and tricuspid valves link to the left and right atria respectively and the aortic and pulmonary valves link the ventricles to the aorta and pulmonary artery, respectively.

The cardiovascular reflexes are those reflexes which impact cardiovascular structures (and so, functions) either through direct innervation or secondarily, for example by influencing the release of substances such as ADH.

In humans, reflex sympathetic output to the cardiovascular system originates most immediately in the rostral ventrolateral medulla (VLM), and incorporates input from the hypothalamus and brain stem nuclei [1,3] (Central regulation of autonomic function). Specifically, parvocellular neurons of the paraventricular nucleus of the hypothalamus project both to the rostroventrolateral medulla and directly to spinal sympathetic preganglionic neurons (SPNs). Neurons of the rostral ventrolateral medulla are arranged in clusters which project in a topographic, and therefore viscerotropic, fashion to sympathetic preganglionic neurons. Hence, stimulation of more cephalad cell clusters may result in constriction of the renal vascular bed, whereas stimulation of more caudal cells results in constriction of mesenteric and hind limb muscle vascular beds. The VLM neurons project primarily to the intermediolateral cells columns of the thoracic cord, and release catecholamines and L-glutamate to excite spinal SPNs [7]. Within the spinal cord, the activity of some SPNs is inhibited by nitric oxide and glycine, although the effects of nitric oxide are complex and involve multiple pathways in the spinal cord.

Parasympathetic innervation to the heart originates in the dorsal vagal motor nucleus and the nucleus ambiguus. Neurons in the dorsal vagal motor nucleus, stimulated directly by the nucleus of the solitary tract (NTS), are probably of less importance to cardiac function than neurons of the nucleus ambiguus, which are indirectly stimulated by the NTS via the caudal VLM. Increased input to the NTS from baroreceptors brings about reflex attenuation of heart rate via vagal stimulation. The dorsal vagal motor nucleus and nucleus ambiguus are also the source of relatively sparse parasympathetic innervation to vascular beds in a variety of organs.

As somatic and visceral afferents are dealt with in depth elsewhere in this text, below we will particularly consider the lower level efferent components of cardiovascular reflexes, dealing first with sympathetic motor neurons and then parasympathetic motor neurons.

Spinal sympathetic preganglionic neurons are located principally within the intermediolateral and, to a lesser extent, intermediomedial columns of the thoracic and upper lumbar spinal cord [7]. In humans, some SPNs are also located within the central autonomic nuclei and the lateral funiculus. The cells of the spinal sympathetic nuclei are not uniformly distributed, but rather are arranged in nests along the length of the thoracic and upper lumbar cord. Axons from each cluster of SPNs exit the spinal cord mainly through the immediately adjacent nerve roots and so project to particular peripheral targets. Upper thoracic SPNs, the overwhelming majority arising from the first to fifth thoracic levels, project to the heart, particularly the ventricles, via the cervical sympathetic ganglia (10.1007/978-3-540-29678-2_1). Additionally, SPNs along the length of the thoracolumbar spinal cord also project in a segmentally-organized fashion to adjacent vascular beds.

The cell bodies of postganglionic neurons within the cervical sympathetic ganglion are arranged in a topographical fashion and form morphologically distinct clusters [7]. Hence, by way of example, stellate ganglion cells projecting to the heart may be distinguished from SPNs to, for example, the vascular bed of the sternocleidomastoid muscle.

Cardiac sympathetic nerves arising from the left and right cervical sympathetic chains and ganglia cross (and sometimes cross back again) from left to right on their journey to the heart [8]. Furthermore, it is apparent, and consistent with their anatomical targets, that left and right autonomic nerves serving the heart are functionally distinct. In particular, sympathetic fibers arising from the left stellate ganglion appear to have a greater influence on heart rate and heart rate variability, such that excessive activity of fibers from the left stellate ganglion or loss of modulating influences from the right stellate ganglion favor arrhythmogenesis.

The cardiac vagus nerve consists of parasympathetic preganglionic fibers which project through mixed sympathetic and parasympathetic plexi to parasympathetic ganglia close to or in contact with the heart. From these ganglia, postganglionic fibers course to the electrical conduction system of the heart and, to a lesser extent, to the myocardium [9]. In histological sections of human heart, the greatest concentration of acetylcholinesterase-positive cells is seen in the region of the sinoatrial node, with concentrations diminishing through the atrioventricular node, the atrioventricular bundle and into the bundle branches. In general, parasympathetic innervation of the sinoatrial node is predominantly from the right vagus nerve, while innervation of the atrioventricular node is predominantly from the left vagus nerve.

Autonomic output to the cardiovascular system originates from the interaction of a number of components of the central nervous system (Central regulation of autonomic function). In particular, the nucleus tractus solitarius (NTS) integrates sensory input from the viscera with information from higher brain centers in order to synthesize output to other brainstem and spinal reflex centers [1]. A portion of visceral afferent input to the NTS projects in a viscerotropic fashion such that information from particular organs projects to specific nuclei within the NTS. The anatomical basis therefore exists for reflex responses which are influenced by the current afferent input from particular visceral organs.

Baroreceptor afferents (10.1007/978-3-540-29678-2_2), which provide beat to beat information on cardiovascular function, project particularly to the intermediate (general visceral) region of the NTS. From the NTS afferent information of different types is relayed to various centers including the insular cortex (the primary viscerosensory cortex), the periaqueductal gray (PAG) of the midbrain, the parabrachial region of the pons, and the ventrolateral medulla. Baroreceptor information is especially relayed to the caudal then rostral ventrolateral medulla and cardiovagal neurons. Receptors similar to those of the carotid sinus, but perhaps less studied, have also been identified in the aortic arch and the myocardium, and pulmonary and thoracic stretch receptors also have a small role in modulating the rhythmicity of central autonomic neurons. There may also be a very short feedback loop by which intrinsic cardiac neurons increase their activity in response to local mechanical stimulation. Some vascular and pulmonary baroreceptors convey afferent information via the vagus nerve giving this nerve an afferent and efferent role in the reflex regulation of the cardiac cycle. Visceral pain and perhaps some baroreceptor information may also ascend via the spinothalamic, spinoreticular and spinomesencephalic pathways to influence baroreceptor reflex behavior. Noxious input, and so activation of the PAG, tends to dampen baroreflexes by augmenting the activity of cardiac sympathetic efferents. Conversely, baroreceptor excitation also depresses somato-cardiac sympathetic reflex responses to A- and C-fiber excitation [4].

Sensory input from the viscera may also be influenced at the level of the primary afferent by convergence with information from somatic tissues, including convergence of thoracic somatic and cardiac afferents onto single spinal neurons. Histological and physiological data also indicate projections of somatic afferents to visceromotor centers. The implied interaction is self-evident in, for example, cardiovascular responses to pain, and involves, at least in part, projections via the periaqueductal gray to the rostral ventrolateral medulla.

In preparation for increased demand on the cardiovascular system, anticipatory “central command” may increase heart rate and blood pressure prior to signaling of work from receptors within the muscle (10.1007/978-3-540-29678-2_1). Additionally, chemoreceptors and mechanoreceptors within muscle signal ongoing exercise. However, input from large diameter group 1a and 1b afferents generally has little influence on cardiovascular function. Central projections and effects of afferent input from neck muscles may represent a special case. Some low-threshold somatic afferents from the neck region project directly to the vestibular nuclei and thereby modulate vestibular influences on cardiovascular function. There is also physiological evidence of more direct connections between cervical muscle afferents and autonomic motor neurons; specifically, projection of muscle afferents directly to spinal sympathetic preganglionic neurons. These results suggest a special role for afferent input from axial muscles, especially those of the neck [6]. The adaptive rationale for these interconnections would include maintenance of appropriate blood pressure and blood volume distribution during postural changes.

Within the heart, the sympathetic efferents innervate many targets. Human heart contains subtypes of α₁, α₂, β₁ and β₂-adrenergic receptors. β-adrenergic receptor subtypes are the more abundant and their stimulation generally results in increased heart rate and increased force of contraction. Although these receptors are fairly evenly distributed throughout the atria and ventricles, the sinoatrial node is especially richly invested [10]. Both of α and β-adrenergic receptors are also found on vascular smooth muscle. Stimulation of α-adrenergic receptors leads to vasoconstriction, whereas the stimulation of β-adrenergic receptors on skeletal muscle and liver vasculature leads to vasodilation.

The effects of vagal stimulation on the heart have been attributed primarily to the release of acetylcholine and somatostatin. A number of subtypes of muscarinic acetylcholine receptors are found in the mammalian heart. The M2 subtype is the most abundant muscarinic receptor in human myocardium, and is more abundant in atrial than ventricular tissue.

While parasympathetic innervation of blood vessels is sparse when compared with sympathetic innervation, M3 muscarinic receptors are found in vascular endothelium. Activation of endothelial muscarinic receptors leads to the production of NO which diffuses into adjacent smooth cells initiating vasodilation.

It has been argued that as a result of integration of information from diverse receptors, cardiovascular reflexes are fundamentally adaptive and adaptable to the unique circumstances in which they arise in each individual. Interestingly, however, at the level of the spinal nuclei, virtually all stimuli result in what could be regarded as excitatory reflexes. Hence, both noxious and innocuous stimulation at any given site will initiate responses characteristic of “fight or flight.” It is the influence of descending excitatory and inhibitory pathways which then modulates (facilitates or depresses) the spinal reflexes in order to produce the final response most appropriate to the subject’s current circumstances [4]. The coordinated regulation of cardiovascular reflexes is most easily appreciated when the harmonious interaction of supraspinal and spinal components is disrupted by spinal injury, as described briefly below.

Collectively, the cardiovascular reflexes are designed to partition a limited blood supply among tissues with changing and often competing demands. Furthermore, this task must be accomplished in a dynamic creature whose movements may impose sudden gravitational challenges on the partitioning of the blood (not to mention other fluids and tissues of the body). The remarkable success of reflex regulation of cardiovascular function owes much to the effectiveness of the central nervous system in integrating afferent input from many sources: the cardiovascular system itself, other viscera, the somatic tissues of the musculoskeletal system and skin, and of course input from higher centers of the nervous system itself (10.1007/978-3-540-29678-2_8).

Cardiovascular reflexes come to clinical attention most often when pathology prevents their appropriate expression, for example when autonomic failure leads to phenomena such as orthostatic hypotension. However, the reflexes may themselves become pathological when dysfunction occurs in the afferent or efferent arms of the reflex pathway, or when there is an error in central processing. This occurs most dramatically in autonomic dysreflexia (10.1007/978-3-540-29678-2_1) following high spinal cord injury. Often following high spinal cord injury, more caudal spinal autonomic reflex centers are liberated from descending inhibition. Additionally, new and inappropriate synapses may form following injury and this may facilitate over-exuberant cardiovascular responses to somatic or visceral stimulation. Clinically, this may manifest as paroxysmal hypertension, which is an important source of morbidity and mortality in spinal patients.

Cardiovascular reflexes see little application in conventional western medicine. Clinicians, and patients themselves, may occasionally take advantage of the oculo-cardiac reflex or the baroreceptor reflex (by applying pressure to the eyes or the carotid bifurcation) to transiently attenuate hypertension or palpitations. In various forms of traditional medicine, however, innocuous or noxious stimulation may be applied to the skin (for example, via acupuncture) or muscles (for example, via mobilization or manipulation) with the intent of effecting long-term changes in cardiac function, systemic blood pressure or perfusion of particular vascular beds. The clinical effectiveness of these approaches is largely untested.

Carotid bodies (CBs) are peripheral arterial chemoreceptors located in the neck that detect changes in PO₂, PCO₂, pH, temperature and glucose in arterial blood. The importance of the CB to the control of breathing was first revealed by the Belgium physiologist, Corneille Jean Francois Heymans. Heymans discovered that cutting the neuronal connections between the carotid body and brainstem caused breathing to slow (hypoventilation) in the resting state and eliminated fast breathing (hyperventilation) caused by creating a local hypoxia-like situation by injecting cyanide into the CB circulation. These experiments demonstrated an essential role for the CBs in maintaining normal arterial carbon dioxide levels and initiating the primary respiratory and cardiovascular physiological responses to hypoxia. In 1938, Heymans was awarded the Nobel Prize for his discovery, arguably one of the most significant in respiratory research to date.

As demonstrated by Heymans, the CB is also a principal pH/Pco₂ chemoreceptor involved in ventilation. They compliment additional sets of CO₂/pH-sensitive cells located in the brainstem, cerebellum and hypothalamus, known as the central respiratory chemoreceptors. However, the relative contribution of the CBs and central respiratory chemoreceptors remains hotly debated [8]. A simple view is that the CBs provide the rapid response, but the central chemoreceptors provide most of the steady-state response. However, as Heymans demonstrated, transecting the CSN leads to hypoventilation, an increase in arterial PCO₂ and a resulting respiratory acidosis. Thus, while the CB’s are vital for maintaining normal PCO₂ the central chemoreceptors alone are insufficient.

There are two opposing hypothesis as to how hypercapnia (increase in partial pressure of CO₂)/fall in intracellular pH might elevate intracellular calcium and trigger CSN activity (Fig. 2): (i) Intracellular acidosis or hypercapnia inhibits membrane K⁺ channels, causing cell membrane depolarization and leading to Ca²⁺ entry through voltage-gated channels. (ii) Intracellular acidosis activates the Na⁺ – H⁺ exchanger system, extruding H⁺ and increasing Na⁺ influx. This results in a rise in intracellular Na⁺ and subsequent Ca²⁺ influx through the reversal of the Na⁺–Ca²⁺ exchanger.

Relative contribution of different O₂-sensing molecules to glomus cell excitability.

A majority of inputs to the basal ganglia is destined to the striatum (CD and putamen); the striatum acts as the input station of the basal ganglia. After being processed in the striatum, signals are sent to other nuclei in the basal ganglia, substantia nigra (SN) and 10.1007/978-3-540-29678-2_7 (GP). The final outputs of the basal ganglia are issued from part of the SN, which is pars reticulata (SNr), and part of the GP, which is the internal segment. The neural circuit in the basal ganglia related to eye movements originates in the CD and converges on the SNr, which then projects to the 10.1007/978-3-540-29678-2_19 (SC) [1] (Fig. 1).Caudate: Role in Eye Movements. Figure 1

Information processing in the caudate nucleus (CD) for the control of saccadic eye movement. CD, caudate nucleus; SNr, substantia nigra pars reticulata; SC, superior colliculus; SNc, substantia nigra pars compacta. Excitatory and inhibitory neurons and synapses are indicated by open and filled symbols, respectively. Gray symbol indicates a dopaminergic neuron which exerts modulatory effects on CD neurons. The thickness of the line (axon) roughly indicates the level of spontaneous activity. CD neurons receive spatial signals from the saccade-related areas in the cerebral cortex and reward-related signals from dopaminergic neurons in the SNc.

The saccade-related region in the CD roughly corresponds to the area that receives inputs from the 10.1007/978-3-540-29678-2_6 (FEF) and 10.1007/978-3-540-29678-2_19 (SEF) [2]. It is therefore likely that the CD receives saccade-related signals from these cortical eye fields. However, inputs to the CD are only loosely segregated. That is, the saccade-related region in the CD receives inputs from other cortical areas including the 10.1007/978-3-540-29678-2_4. In addition to these converging inputs, the entire CD (together with the putamen and the ventral striatum) receives diffuse inputs from dopaminergic neurons in the substantia nigra pars compacta (SNc) and its surrounding regions [3]. It is likely that particular combinations of these inputs create signals unique to CD neurons.

A majority of neurons comprising the CD are called medium-spiny neurons: neurons with medium-sized cell bodies and many dendrites thickly covered with spines. They are the projection neurons: neurons that project axons to the outside of the CD. They are GABAergic and inhibitory. The projection neurons are highly hyperpolarized in the resting state and emit action potentials only occasionally. A minority of neurons (less than 5%) in the CD consist of several types of interneurons. One conspicuous type is the cholinergic interneuron, which is characterized anatomically as a large-aspiny neuron. They fire tonically and irregularly and are often called “tonically active neurons” or “TANs” [4].

Single unit studies using monkeys trained on saccade tasks have revealed that many CD projection neurons are clearly related to 10.1007/978-3-540-29678-2_19. Some of them respond to visual stimuli that potentially induce saccades to them. Other neurons become active before saccades. These visual-saccadic neurons have response fields which are usually centered in the contralateral field. The responses are often highly dependent on the context. Visual responses may be enhanced if the animal attends to or memorizes the stimulus. Saccadic activity may be present only when the saccade is guided by memory, or only when it is guided by visual stimuli. The neurons usually do not fire in relation to 10.1007/978-3-540-29678-2_19. Intermingled with such visual-saccadic neurons are found more complex neurons, such as those related to expectation of task-specific events or 10.1007/978-3-540-29678-2_18. Such a complex nature of CD projection neurons appears to reflect the convergent inputs from the cortical eye fields (FEF and SEF) and from the dorsolateral prefrontal cortex.

Studies suggest that these saccade-related neurons in the CD neurons control saccadic motor outputs by modifying neuronal activity in the SNr and the SC (Fig. 1) (see the section 10.1007/978-3-540-29678-2_19). Electrical stimulation in the CD saccade-related region may elicit saccadic movements of the eye and the head to contralateral directions. If the electrical stimulation is short and weak, no movements are evoked, but it induces inhibitions and sometimes facilitations in SNr neurons. The former is likely to be mediated by the direct CD-SNr inhibitory connection, while the latter is likely to be mediated by the indirect pathway through the GP. Unlike CD projection neurons, SNr neurons are very active spontaneously, usually firing at more than 50 Hz. During the saccade tasks, many SNr neurons exhibit visual, saccadic, and memory-related activities, similarly to CD projection neurons. However, these activities usually show up as decreases in firing rates, unlike CD projection neurons. Since SNr neurons, especially those exhibiting saccade-related activity, have inhibitory connections to neurons in the SC, the decrease in SNr neuronal firing should lead to a disinhibition of SC neurons which control saccades to contralateral directions. In short, saccade-related activities in CD projection neurons would usually lead to facilitation of saccades to the contralateral directions. Note, however, the role of the CD on saccades may sometimes be suppressive, since some SNr neurons increase firing rates in relation to saccades presumably through the indirect pathways.

Recent studies have revealed another striking feature of CD neurons: Relation to reward-oriented behavior. Here, the amount of reward is biased depending on the direction of saccade: for example, rightward saccades are followed by a big reward and leftward saccades are followed by a small reward. This task has a strong behavioral impact: the saccade to the position associated with big reward is faster and earlier than that associated with small reward. The visual response of CD projection neurons is greatly enhanced and diminished if the saccade to the visual stimulus is expected to be followed by a bigger and smaller reward, respectively [5]. A minority of neurons exhibit the opposite reward modulation. The modulation is very common among visually responsive CD neurons (about 80%), and is often very strong such that the original directional tuning can be completed reversed. Similar reward-dependent modulation occurs for saccadic activity.

The relation of the CD to reward-oriented behavior is highlighted by a conspicuous group of CD projection neurons which cannot be classified as simply related to visual-saccadic processes [6]. They exhibit growing activity while the animal is waiting for the go-signal for a saccade. It appears to be related to saccade preparation. However, the activity occurs before instruction is given to which position the saccade should be made. The activity is nonetheless spatially selective in that it is present only when saccades to a contralateral, rather than ipsilateral, position are followed by a big reward. If there is no positional bias in reward, the anticipatory activity is much weaker.

Further studies suggest that the reward-position-sensitive anticipatory activity is transmitted to the SC through the SNr. Suppose a bigger reward is associated with saccades to a right target than a left target, CD neurons on the left side would exhibit stronger anticipatory activity than those on the right side (according to the findings described above). Since a major effect of CD neurons on SNr neurons is inhibitory, SNr neurons on the left side would exhibit a stronger decrease in firing rates than those on the right side. This is actually observed experimentally [7]. Since SNr neurons in turn inhibit SC neurons, the excitability of SC neurons on the left side would be elevated compared with those on the right side. This has also been confirmed experimentally [8]. There is now a clear imbalance in excitability between the two sides of the SC: Neurons in the left SC are more excitable than neurons in the right SC. This occurs before any instruction is given and far before a saccade is executed. This is an internal process based on the knowledge of the positional difference in reward amount. Under such a biased condition, a target that appears on the right side (i.e., associated with a big reward), which activates neurons on the left SC, would trigger a saccade more easily and more quickly than a target on the left side. In other words, the animal would make saccades more quickly to a more highly rewarded position. Such a behavioral bias is consistently observed experimentally. These results suggest that the CD is a critical brain area for reward-oriented motivational behavior.

Such strong reward-dependent modulation of CD neurons may be caused by dopaminergic inputs (Fig. 1). As mentioned above, CD projection neurons have dendrites with many spines on which both axons from cortical neurons and axons from midbrain dopaminergic neurons make synapses. A majority of cortical axons originate from the cortical eye fields and are therefore likely to carry spatial signals. In contrast, midbrain dopaminergic neurons are known to carry reward-related signals, but not spatial signals [9]. They respond to reward which is given unexpectedly or, if reward is expected, to a sensory stimulus that predicts the reward. In the reward-biased saccade task, dopaminergic neurons respond by excitation and inhibition to sensory stimuli that predict a reward that is larger and smaller than what is expected, respectively.

How could dopamine influence activity of CD projection neurons? Dopamine does not exert fast excitatory or inhibitory actions, but is thought to modulate other synaptic inputs, especially glutamatergic inputs from the cerebral cortex. These findings led to the following hypothesis: the spatial signals from the cortical eye field are enhanced if dopaminergic inputs are increased (i.e., a bigger reward is expected) and depressed if dopaminergic inputs are decreased (i.e., a smaller reward is expected). The interaction may occur within individual 10.1007/978-3-540-29678-2_4. This interaction may be due to the interactions among different ionic conductances. Or, it may be due to 10.1007/978-3-540-29678-2_12 or depression (LTD). Recent studies have indicated that LTP indeed occurs if cortical inputs come in simultaneously with dopaminergic inputs and if the CD neuron fires [10]. In support of this hypothesis, the reward-dependent bias in saccades is reduced if dopaminergic transmission in the CD is blocked by injecting 10.1007/978-3-540-29678-2_4 [11].

The role of interneuronal processing in the CD is less clear. TANs, which are thought to be cholinergic, respond to reward or reward predictor, similarly to dopaminergic neurons, but also respond to a sensory stimulus that predicts the absence of reward or punishment. It has been suggested that cholinergic interneurons indicate that the reward is not equal among actions to choose, but do not indicate which action is the best or the worst. The latter function would be carried out by dopaminergic neurons. Cholinergic interneurons might detect the condition in which rewards are unequal and guide dopaminergic neurons to fully operate. However, this hypothesis needs to be examined in future experiments.

The role of the CD in eye movements is usually not emphasized in clinical literature. However, patients with degenerative diseases involving the CD, such as Parkinson’s disease and Huntington’s disease, may exhibit severe difficulty in making eye movements [1]. The deficits may be more evident when the patients are asked to make eye movements voluntarily or based on memory; the deficits are less clear or absent when eye movements are made to visible targets. Local deprivation of dopaminergic innervation in the CD in monkeys leads to the severe paucity of spontaneous saccades and deficits in 10.1007/978-3-540-29678-2_13 to the side contralateral to the denervation. However, eye movement deficit after a lesion in the CD is not a universal finding. This may partly be due to the anatomical configuration of the eye movement-related region in the CD. In monkeys trained on saccade tasks, many saccade-related neurons are found distributed in an anterior-posteriorly elongated zone in the CD excluding the most anterior part. A small lesion in the CD may not seriously disrupt information processing for saccadic eye movement.

A single cell division cycle has two major phases, namely the DNA synthesis (S) phase and the cell division or mitosis (M) phase. Between these two phases, there are two “gap” or “inter-” phases called gap 1 (G1) and gap 2 (G2) phases. There is another cell cycle state called the G0 phase, when cells remain resting but capable of reentering into a proliferative cell cycle. It is believed that in the developing brain most, if not all, newly developed neurons are not in the G0 state but in the terminally differentiated state. The G1 phase is initiated as M phase is completed and completed as S phase begins; the G2 phase is initiated as S phase is completed and completed as M phase begins. Dividing cells proceed through these four phases repeatedly. Thus, this continuing process is called the cell “cycle.”

Among the four phases of the cell cycle, the G1 phase is considered to be a critical period when proliferative cells receive extracellular “cues” that may lead these cells to either proceed to S phase or to exit from the cycle. These extracellular cues include extrinsic molecules such as neurotrophic factors and mitogens/anti-mitogens of various kinds and environmental substances such as drugs and pollutants.

Generally, the total cell cycle length of undifferentiated progenitor cells varies greatly, not only among different types of tissues/organs to which these cells give rise but also among different time points during ontogeny of the tissue/organ. For example, while neural progenitor cells in the pseudostratified ventricular epitherium (PVE, Fig. 1a) lining the surface of the lateral ventricles of the mouse embryonic forebrain take ~8 h to complete a single cell cycle on embryonic day 11 (E11), those on E16 take as long as 18 h (Fig. 2) [1].Cell Cycle. Figure 1

Overview of Critical Events of Neocortical Neuronogenesis. (a) Schematic representation of interkinetic nuclear migration. The position of cell nuclei of neuronal progenitor cells changes systematically as they proceed through cell cycle phases (M, G1, S, G2 along the abscissa). A fraction of cells exits the cycle during G1 phase (upward arrow). PVE pseudostratified ventricular neuroepitherium; LV lateral ventricle. (b) The founder population of neural progenitors and their progenies execute 11 cell cycles (CC₁-CC₁₁) over the neuronogenetic interval. The early-formed neurons are destined for the deepest cortical layers (dotted curved lines connecting the PVE and layers VI), whereas the later formed neurons are destined for the more superficial layers (solid curved lines connecting the PVE and layers IV and II/III). (c) The fraction of daughter cells that exits the cell cycle (Q fraction) increases slowly during the initial phase of neuronogenesis and then accelerates to reach the final value of 1.0 at the completion of neuronogenesis when 100% of progenitor cells leave the cycle.

It is of note that G1 phase is the phase of the cell cycle when a given proliferative progenitor cell chooses whether to proceed to S phase and remain in the proliferative cell cycle or to leave the cycle and become terminally differentiated (Fig. 1a). In other words, for a single G1 phase progenitor, there are only two fates to follow (all-or-none kind of decision making) either to stay in or to exit the cycle [2]. Obviously, the population of neuronal progenitor cells is composed of numerous proliferative cells, which makes the population probability of cell cycle exit anywhere between 0% and 100%. In the course of neocortical histogenesis, the probability of cell cycle exit or quiescent (Q) fraction increases from 0 (i.e. 0%) at the outset to reach 1.0 (i.e. 100%) at the completion of the interval of neuron generation (i.e. neuronogenesis, Fig. 1c) [2].

In the developing central nervous system, multipotent neural stem cells progressively become mature neurons or glial cells. This process involves three steps, (i) fate determination, (ii) subtype selection and (iii) maturation. In the fate determination step, the cell fate is determined; cells acquire a neuronal or glial 10.1007/978-3-540-29678-2_9. During or subsequent to the fate determination step, neuronal and glial subtypes are selected. In the case of neurons, a subtype selection is made from many options, such as motor versus sensory subtypes and excitatory (glutamatergic) versus inhibitory (GABAergic) subtypes. For glial cells, selection for oligodendrocytes versus astrocytes is made. In most cases, the fate determination and subtype selection steps proceed at the same time. Each subtype of cells then becomes morphologically and functionally mature.

The developing central nervous system initially consists of neuroepithelial cells, which have epithelial cell characteristics such as tight junctions and adherens junctions at the apical side and a basal lamina on the basal side (Fig. 1) [1,2].Cell Differentiation. Figure 1

The cell fate determination step. Neuroepithelial cells, the first form of neural stem cells, have epithelial cell characteristics such as the junctional complex at the apical side and the basal lamina on the basal side. These cells undergo symmetrical cell divisions and do not usually give rise to neurons. Neuroepithelial cells gradually change into the second type of neural stem cells, called radial glial cells. These cells undergo asymmetrical cell divisions and give rise first to neurons. After the production of neurons, radial glial cells give rise to oligodendrocytes, astrocytes and ependymal cells. Radial glial cells disappear after birth, but some astrocytes or astrocyte-like cells function as neural stem cells in the adult brain.

Until relatively recently the nerve cell membrane appeared to be a somewhat simple structure with a few simple, internal stereotyped tasks, whereas the accomplishment of complex neuronal tasks was thought to be the exclusive domain of networks of neurons. What we knew about membrane structure through the 1960’s was comparatively modest. The membrane was unmistakably very thin, on the order of 50 nm. It was made up largely of proteins and lipids, the latter organized into a bilayer. It was electrically charged (polarized) at rest (10.1007/978-3-540-29678-2_13). It had aqueous channels through which various ions passed that allowed the cells to be excitable, i.e., to generate 10.1007/978-3-540-29678-2_1 for communication among cells. It was endowed with 10.1007/978-3-540-29678-2_18, defined mainly by pharmacological testing, that enabled cells to communicate chemically through release and receptor binding of 10.1007/978-3-540-29678-2_14. The membrane was also thought to somehow facilitate growth and development of 10.1007/978-3-540-29678-2_14 (dendrites and axons) for local and distant cell-cell communication.

The introduction of new methodologies beginning in the 1970’s, including x-ray diffraction, freeze fracture electron microscopy, advances in crystallography, computerized methods for analysis and modeling, and an avalanche of molecular biological methodologies and discoveries, brought a new appreciation of cell membrane structural complexity; and with it, the discovery of heretofore unknown, built-in mechanisms of synaptic control and 10.1007/978-3-540-29678-2_14.

This article provides a contemporary survey of membrane structural components, how they are assembled and how they contribute to nerve cell function.

The unit membrane of the nerve cell is depicted in Fig. 1. It is, on average, about 50 nm thick and comprised of various types of 10.1007/978-3-540-29678-2_16, proteins and carbohydrates. Proteins, being the largest molecules, make up the greatest membrane mass but the smaller phospholipids are the largest in number and carbohydrate molecules are the fewest. The molecules making up the membrane proper, or attached to it, are mobile, interactive and in many cases functionally interdependent. They are replaced by intracellular biosynthesis, and turned over by a process called 10.1007/978-3-540-29678-2_20.Cell Membrane Components and Functions. Figure 1

Diagram of the nerve cell membrane. Shown are phospholipids, cholesterol, and various proteins (I–XIII) that make up membrane structure. See text for a description of their chemical properties and functions. Revised composite assembled from [4–6,10].

Membrane lipids are esters of glycerol phosphate attached to two long-chain fatty acids, each generally 14–20 carbon atoms long, and arranged in a bilayer with the glycerol phosphates facing the extracellular and intracellular fluids, and the fatty acid chains arranged in rows side by side in the membrane.

The phospholipid molecules are synthesized in the endoplasmic reticulum (ER), mainly in the cytoplasmic monolayer. Four different phospholipids are the major constituents of the bilayer: sphingomyelin, phosphatidylcholine, phosphatidylethanolamine and phosphatidyl-serine. Smaller amounts of other phospholipids such as inositol phosphates are also found in the membrane. Phospholipids are differentially distributed in the cell membrane. More sphingomyelin and phosphatidylcholine are found in the outer leaflet of the bilayer, while more phosphatidylethanolamine and phosphatidyl-serine are found in the inner leaflet.

Phospholipid molecules have a high degree of lateral mobility in the bilayer, which facilitates movement of small nonpolar molecules across the cell membrane. Fluidity of cell membrane phospholipids also facilitates 10.1007/978-3-540-29678-2_20 processes and enzyme activities. In fact, some membrane proteins require the presence of phospholipids for proper function. Less frequently, phospholipid molecules will “flip-flop,” i.e., migrate from a monolayer on one side to that on the other.

Insulation and barrier properties. The lipid bilayer acts as a barrier to passage of polar substances including water and electrolytes, although gases such as oxygen and carbon dioxide pass through it, along with various lipid–soluble substances including alcohol and local anesthetics. The barrier property protects the cell from loss of vital polar cytoplasmic constituents and entry of many potentially harmful extracellular substances.

The extremely thin, expansive lipid bilayer of the nerve cell membrane has a 10.1007/978-3-540-29678-2_13 on the order of 1 μF/cm² that produces a charge of about 8 × 10⁻⁹ coulombs/cm² at a 10.1007/978-3-540-29678-2_18 (membrane potential – basics) of –80 mV, or approximately 5 × 10¹¹ monovalent ions/cm². Even under steady state, or resting conditions, membrane channels, including some K⁺ and Na⁺ channels, stay open and generate a “10.1007/978-3-540-29678-2_12” [1]. In effect, what results is a leaky RC circuit, with a specific membrane resistance of about 1,000 Ωcm² [2].

The RC circuit properties have functional consequences (Fig. 2b). In the resting state, the net movement of K⁺ ions down its concentration gradient through leak channels will leave behind impermeant cytoplasmic organic anions that accumulate on the inner side of the cell membrane and an accumulation of cations on the extracellular side [3], which accounts for the potential difference (Vm) of about −80 mV across the cell membrane (membrane potential – basics).

Figure 2b also shows subthreshold membrane potential responses to square wave pulses of current applied through a microelectrode inserted through the cell membrane (also membrane potential – basics). Each depolarizing current pulse will first mainly charge the membrane capacitor, and then the voltage difference will promote ion movement through the leak channels. The consequence is that the voltage transients caused by the applied current steps from the microelectrode develop more gradually than the current changes. Synaptically evoked 10.1007/978-3-540-29678-2_16 (PSPs) will also build up more slowly than their corresponding synaptic currents. The RC properties of the membrane act as a low-pass filter, and an integrator of PSPs if frequency of occurrence leads to temporal summation. They can also affect the time required for an action potential to reach threshold, and consequently the interspike interval during a burst of action potentials.

The nerve cell membrane contains large amounts of cholesterol, which is synthesized mainly in the endoplasmic reticulum (ER). Cholesterol enhances the permeability-barrier property of the lipid bilayer. The hydroxyl groups of cholesterol are in proximity to the polar heads of the phospholipid molecules (Fig. 1) and partially immobilize the hydrocarbons close to the polar heads. This renders the lipid bilayer less permeable to small water-soluble molecules.

These lipids contain carbohydrate groups, usually galactose but also glucose, inositol or others, and are found only on the extracellular side of the cell membrane. 10.1007/978-3-540-29678-2_7 associate into micro-aggregates and are believed to be involved in cell-cell interactions. Five to ten percent of the total lipid mass consists of a particular type of glycolipid called a 10.1007/978-3-540-29678-2_7.

Gangliosides are thought to alter the electrical field across the cell membrane, as well as the concentration of Ca²⁺ ions along the external surface of the cell membrane. They may also be involved in cell-cell recognition at the extracellular matrix that promotes cell aggregation.

Figure 1 illustrates only 13 of a much larger group of currently identified and characterized membrane proteins. Proteins are subdivided according to position into integral (types I-V in Fig. 1) and peripheral (VI–XIII). 10.1007/978-3-540-29678-2_9 completely traverse the cell membrane, whereas 10.1007/978-3-540-29678-2_16 are anchored to either the cytoplasmic or extracellular side. Singer [4] subdivided integral proteins into four general types, I–IV. Types I and II have just one transmembrane segment, and terminal amino and carboxyl groups on opposite sides of the cell membrane. Some transport proteins are part of this group. Members of the Type III group, which include the β−10.1007/978-3-540-29678-2_1 and adenylyl cyclase, an intracellular signaling component, have polypeptide chains that traverse the cell membrane several times. Members of type IV, which includes 10.1007/978-3-540-29678-2_9, have several domains that are arranged around an aqueous pore that serves as the channel. Hydrophobic parts of integral proteins are positioned within the cell wall, in parallel with the lipid bilayer. Hydrophilic parts of integral proteins face the cytoplasm and extracellular fluid. The proteins have some degree of lateral mobility, less so than phospholipids.

Membrane proteins exhibit function-dependent polarity. For example, transporting enzymes have ATP-binding sites on the cytoplasmic side and glycoproteins have sugar residues on the outer surfaces.

Synthesis of cell membrane proteins takes place largely in the soma endoplasmic reticulum (ER), under the direction of nuclear DNA in ribosomal RNA-protein complexes. Selective axonal and dendritic transport processes deliver proteins to all regions of the neuron. Rough ER bears the ribosomes during protein synthesis. Newly synthesized protein is stored in cisternae, transported in vesicles through the Golgi apparatus and inserted into the cell membrane.

Cell membrane proteins are also synthesized on polyribosomes and stored in membranous cisterns in dendrites and axons, where they play important roles in 10.1007/978-3-540-29678-2_19 and 10.1007/978-3-540-29678-2_1.

The locations of different types of proteins in the cell membrane serve as general predictors of how they function in the nerve cell.

Integral membrane proteins serve as:

10.1007/978-3-540-29678-2_9, moving ions against a concentration gradient, using energy derived form ATP

Glycoproteins have three main subgroups, each involved one way or another with cell-cell recognition: the immunoglobulin super-family, the cadherin family and the integrins. The immunoglobulin family imparts Immunoreactivity, homophilic cell-cell interactions and outgrowth of neurites and fiber bundling during development. The cadherin family promotes Ca²⁺-dependent neurite growth and axon bundling. The integrins promote cellular interactions with the extracellular matrix and also promote neurite growth and extension of axons to their targets.

10.1007/978-3-540-29678-2_16, another group of integral proteins thought to be involved in cell-cell recognition, have long sugar chains that form a structure around the cell called the glycocalix that is important for structural support.

Peripheral proteins function as:

Receptors for neurotransmitters, 10.1007/978-3-540-29678-2_14, 10.1007/978-3-540-29678-2_8 and other chemical messengers that trigger membrane ion permeability changes.

The fidelity of chemically mediated 10.1007/978-3-540-29678-2_19 is affected by the affinity of an agonist for a receptor and by the number of available receptors. Each of the two factors plays a key role in neural network responsiveness to endogenous 10.1007/978-3-540-29678-2_14 as well as drugs such as opiates.

Desensitization of response to a 10.1007/978-3-540-29678-2_18 most often occurs after prolonged or repeated receptor binding, particularly if the agonist has a high affinity for the receptor.

As shown in Fig. 3, binding of the agonist to its receptor triggers activation of several cytoplasmic membrane proteins, including G-Protein-coupled receptor kinases and a family of proteins known as arrestins. The consequence is internalization of the receptor, which is either degraded or reincorporated into the cell membrane.Cell Membrane Components and Functions. Figure 3

Trafficking (cycling) of a G protein-coupled receptor (GPCR) under the influence of GPCR kinase (GRK) and Arrestin. After agonist binding and G-protein-mediated activation (or suppression) of a signal pathway, GRKs phosphorylate GPCR and Arrestin forms a complex that terminates signaling and translocates the complex to a clathrin-coated pit, followed by internalization and either degradation or recycling. Adapted from [6].

The nerve cell membrane consists of a functionally efficient organization of phospholipids, proteins and carbohydrates that orchestrate static functions such as insulation and membrane electrical charge, and dynamic functions related to cell excitability, cell-cell communication and cell responsiveness to receptor agonists.

One of the key features of many cellular clocks is that they can function autonomously, i.e., they do not require intercellular communication. A seminal experiment demonstrating this revealed that mammalian circadian pacemaker neurons cultured under conditions where synaptic communication is abolished continue to exhibit circadian rhythms of action potential firing rate [1]. This leads to the conclusion that circadian oscillation is an intrinsic property of the cell itself, and is not solely a tissue-level phenomenon. This constrains the search for underlying mechanisms of circadian oscillation to processes that occur at the cellular level. Nevertheless, intercellular communication at the tissue level and communication between different tissues at the organismic level play important roles in the integrative physiology of circadian timekeeping.

A key common feature of the cellular clock in fungi, plants, and animals is negative transcriptional feedback, a mechanism in which a gene product negatively regulates the gene that encodes it [for review, see 2]. Many “clock proteins” are transcription factors that, upon nuclear entry from the cytoplasm where they are synthesized, inhibit transcription of the “clock genes” that encode them. This mechanism ensures that when clock protein levels increase, there is a decrease in clock gene transcription, and consequent decrease in clock protein synthesis. In conjunction with clock protein degradation, this leads to a decrease in clock protein levels. This results in release from inhibition of clock gene transcription, and a consequent increase in clock protein levels, thus completing one cycle of an oscillation in the abundance of both clock gene transcripts and, with a delay, clock proteins.

While this simple model of the cellular clock explains how negative transcriptional feedback can underlie a 10.1007/978-3-540-29678-2_19 of clock gene transcript and clock protein abundance, it does not account for the fact that circadian 10.1007/978-3-540-29678-2_15 cycle with a 10.1007/978-3-540-29678-2_16 very close to 24 h. The period of oscillation is determined by the time occupied by each of the steps of the cycle outlined above: transcription of clock gene, translation of clock gene transcript into clock protein, nuclear import of clock protein, degradation of clock protein. The most important mechanisms for setting the period of oscillation of the cellular clock appear to be regulation of the rates of nuclear import and degradation of clock proteins, implemented via post-translational modification of clock proteins [for review, see 3]. The two main post-translational modifications of clock proteins are protein phosphorylation – the covalent attachment of phosphate groups – and ubiquitination – the covalent attachment of the small polypeptide ubiquitin. Phosphorylation of clock proteins catalyzed by protein kinase enzymes regulates both nuclear import and ubiquitination. Ubiquitination of clock proteins catalyzed by ubiquitin ligase enzymes regulates degradation. Clock proteins can also be dephosphorylated and/or deubiquitinated by protein phosphatases and ubiqutin-specific proteases, respectively. Thus, the balances of phosphorylation and dephosphorylation, and of ubiquitination and deubiquitination, ultimately determine the period of oscillation of the cellular clock, with an appropriate balance resulting in a period of oscillation close to 24 h. Point mutations either in clock-protein transcription factors or the kinases that phosphorylate them that affect phosphorylation lead to aberrant periods substantially shorter or longer than 24 h. Some of these point mutations have been implicated in human disorders of circadian regulation of the 10.1007/978-3-540-29678-2_19 [for review, see 4].

While most cellular clocks appear to require negative transcriptional feedback, circadian oscillation in cyanobacteria – photosynthetic prokaryotes – can occur under some circumstances in the complete absence of transcription altogether. When a completely purified cyanobacterial clock protein that has protein kinase activity is incubated in a test tube with ATP, the clock protein itself exhibits a circadian rhythm in its level of phosphorylation [5]. Since this rhythm occurs in a reconstituted cell-free system without any gene transcription or protein translation, it establishes the existence of cellular clock mechanisms that do not rely on negative transcriptional feedback.

In addition to negative transcriptional feedback and post-translational modifications, some circadian pacemaker neurons require membrane depolarization for continued oscillation (for review, see [6]). When the plasma membrane of fruit fly or mammalian pacemaker neurons is chronically hyperpolarized, cellular oscillation is severely impaired. Interference with intracellular calcium signaling also severely impairs cellular oscillation. These kinds of studies have led to a model of cellular oscillation in which circadian rhythms of membrane potential and/or intracellular calcium also participate – along with negative transcriptional feedback – in circadian timekeeping.

One of the most fascinating features of cellular clocks is that they are temperature compensated, meaning that they run with the same period at a relatively wide range of temperatures. Since cellular clocks are based on a complicated set of interlocking biochemical reactions, and since the rates of biochemical reactions have a temperature dependence determined by principles of thermodynamics, then the simplest prediction would be that the period of cellular oscillation would also be temperature dependent. The fact that cellular clocks are temperature compensated thus implies the existence of specific compensatory mechanisms that counteract the effect of temperature on the biochemical reactions that underlie cellular timekeeping. While the nature of these compensatory mechanisms remains obscure, it is noteworthy that the reconstituted cell-free cyanobacterial oscillator is temperature compensated [5].

An enormous amount of information has been gained from in vitro mammalian brain preparations. Examples range from the discovery of long term potentiation to mechanisms of synaptic transmission and oscillatory neuronal behavior, as well as imaging of somatic and dendritic integration. The in vitro brain slice preparations evolved from the need to circumnavigate the technical obstacles and limitations encountered when working on the brain in vivo. Indeed, in vitro brain slice preparations are advantaged by the ability to control precisely multiple physiological variables (e.g., temperature; osmolarity) as well as the extracellular milieu thereby enabling the administration of pharmacological agents that include those that would be toxic if administered in vivo. Of major benefit is the mechanically stable environment of in vitro brain slice preparations. For example, maintaining intracellular recordings in vivo is plagued by the constant movement of the brain caused by the cardiac pulse and/or breathing cycle. With the significant advances in live imaging at the cellular level, brain slices, particularly those from neonates (which are more transparent as myelination is incomplete), allow visualization of cells (neurons, glia or vessels) and measurement of intracellular events such as calcium fluxes and translocation of fluorescently tagged proteins. Importantly, the brain slice is insentient and data are not compounded by the unphysiological effects of anesthesia.

The viability of the brain slice is determined by its thickness. Thus, the neuronal circuitry and connectivity is restricted. Without a circulation oxygen delivery is dependent on diffusion. This is limited as demand for oxygen by brain tissue is relatively high. To assist in delivery, high concentrations of oxygen are used (95% with 5% carbon dioxide, or carbogen) to elevate the diffusion gradient. Measurements in slices indicate that the tissue oxygen levels at the surface of the slice are hyperoxic (10.1007/978-3-540-29678-2_8) but levels decline rapidly such that anoxia occurs by 150–175 μm below the surface [1]. En-bloc brainstem and brainstem-spinal cord preparations of neonatal rats have been used in cardiovascular and respiratory research but these are known to have an anoxic core and viability is limited to the early neonatal period only. To improve the viability of thicker in vitro brain preparations and to allow studies to be performed on adult tissues, researchers developed arterially perfused in vitro preparations which, for example, included those of the brainstem [2] and cerebellum.

Despite these technological advances, there was a requirement to study the brain in situ. In situ means studying the brain within the body of the animal. This had the distinct advantage over isolated in vitro brain preparations (slice, en bloc, arterially perfused) of not only preserving both significant regions of the brain but also maintaining the peripheral afferent pathways and their peripheral receptors intact. It was apparent that the motor pathways were also preserved allowing kinesiological (Kinesiology) studies as target organs were functional. Motor outflows (autonomic and somatic) could be shown to respond appropriately to stimulation of classical reflex pathways such as those mediating nociception, baroreceptor and peripheral chemoreceptor information. With such integrity of the in situ preparation, the question of how it was different to in vivo preparations and what added benefits there were arose. The in situ approach is distinct to in vivo in that anesthesia could be avoided by decortication or decerebration, the pulse pressure that caused mechanical instability in vivo was either minimized or abolished meaning the brainstem was more receptive to intracellular recording and imaging (see below). If forebrain structures were required then anesthetic agent could always be added to the perfusate. Finally, there was good pharmacological access as drugs could be applied topically or to the perfusate. A number of in situ preparations from multiple researchers using a variety of species have been utilized previously (e.g., [3]) and all demonstrated superior mechanical stability of the brainstem relative to in vivo rats and the ability to antagonize receptors with drugs given systemically.

Since 1995 three new in situ preparations have been developed (i) the working heart-brainstem preparation (WHBP; [4]); (ii) the perfused hind limb and trunk preparation and, (iii) the decerebrate arterially perfused whole rat preparation [5]. In all cases, the preparations were perfused arterially with a cell-free perfusate consisting of a Ringer’s solution containing an oncotic agent (to prevent edema) with an osmolarity of 290 mosm.kg⁻¹.H₂O and gassed with carbogen (pH 7.35) at 31–33°C. All these variables can be “clamped” or manipulated as the experiment demands. No oxygen carriers were required. This was a major benefit as they are both expensive and difficult to dissolve in aqueous solution. Adequate oxygenation is achieved because the perfusate is less viscous than blood allowing higher flow rates to be used for a given arterial pressure. Additions of vasopressin to the perfusate were effective in increasing vascular tone allowing long term manipulation of perfusion pressure to within physiological or hypertensive levels (Fig.1).Central Integration of Cardiovascular and Respiratory Activity Studied In Situ. Figure 1

Integrating across systems in situ. Simultaneous recording of arterial pressure (AP), heart rate (HR; beats per minute, bpm), integrated thoracic sympathetic activity (SNA_th), integrated phrenic nerve activity (∫PNA) and two expiratory neurons from the Bötzinger complex. Note the Hering-Traub waves in the arterial pressure trace (solid arrow) and the sinus arrhythmia (open arrow); the former reflect the respiratory-related increase in sympathetic discharge whereas the latter indicate heightened excitability of cardiac vagal motoneurons in early expiration. Stimulation of the baroreceptor reflex by raising systemic pressure (achieved by increasing perfusion pump rate transiently) reduces both heart rate and SNA(*) but prolongs expiratory time which is coincident with an activation of the post-inspiratory neuron (Post-Insp) and an inhibition of the expiratory augmenting (Exp-Aug) neuron. Note the prolonged post-inspiratory neuron firing and enhanced sinus arrhythmia (two open arrows) after the baroreceptor stimulus. Unpublished (D. Baekey, T. E. Dick & J. F. R. Paton).

The integrative nature of the WHBP is portrayed by the central nervous coupling between multiple systems. Fig. 1 shows that the preparation exhibits respiratory sinus arrhythmia that occurs naturally to assist the matching of cardiac output with minute ventilation. This has a central nervous correlate with the firing of neurons located in the ventrolateral medulla. These are expiratory neurons exhibiting a decrementing discharge that fire maximally coincident with the onset of the sinus arrhythmia bradycardia (Fig. 1). The firing of these neurons occurs at an identical time as well as exhibiting a similar pattern to the decrementing expiratory motor activity recorded from the central vagus nerve. This decrementing (post-inspiratory) motor activity targets cardiac vagal post-ganglionic neurons but also laryngeal adductor muscle. Interestingly, it is possible to measure the respiratory phase-dependent changes in laryngeal resistance in the preparation, which is a good example of how the preparation offers a kinesiological approach. Respiratory modulation of the airway acts to facilitate inhalation (upper airway 10.1007/978-3-540-29678-2_15) as well as stalling exhalation (laryngeal 10.1007/978-3-540-29678-2_1) to prolong time for gas exchange at the level of the alveoli. In addition, sympathetic nerve activity is also respiratory phase modulated peaking at the transition between the end of inspiration and start of expiration (Fig. 1). Again, expiratory neurons (Bötzinger augmenting type) show a similar firing pattern and temporal relationship with sympathetic nerve activity (Fig. 1).

The WHBP has contributed to the understanding of central integration of cardiovascular and respiratory reflexes. These include reflexes originating from nociceptors (somatic and visceral), peripheral chemoreceptors, cardiac, pulmonary, nasal, pharyngeal and oesophageal receptors. Additionally, the baroreceptor reflex is functional evoking the classical response of bradycardia, sympathoinhibition and prolongation of expiratory time (Fig. 1). Incidentally, the gain of the cardiac baroreceptor reflex is comparable to that measured in the conscious unrestrained rat (i.e., ∼1.8 bts/min/mmHg). With the ability to precisely control arterial pressure, including its resting level, we were able to demonstrate a difference in the pressure threshold for baroreceptor reflex evoked vagal bradycardia versus sympathoinhibition. This lead to the new idea of separate reflex sympathetic and parasympathetic arcs existing at the level of the nucleus tractus solitarii, the site of termination of baroreceptor afferents [8]. In addition, for the first time baroresponsive cells were recorded intracellularly from the nucleus tractus solitarii using whole cell patch recording, while stimulating the baroreceptors using the physiological stimulus of pressure. This has allowed novel insight into their sub-threshold activity, intrinsic membrane properties and morphology, for example. The WHBP has also played an important role in the introduction of viral gene transfer as a method to unravel central mechanisms involved in regulating the sensitivity of the baroreceptor reflex at the level of the nucleus tractus solitarii [9]. In the absence of a pharmacological antagonist for endothelial nitric oxide synthase (eNOS), adenoviral gene delivery of an eNOS dominant negative was used to demonstrate that the well established depressant effect of angiotensin II acting at the level of the nucleus tractus solitarii on baroreceptor reflex gain was via production of nitric oxide generated by stimulation of eNOS. This led to the novel idea of vascular-neuronal signaling in which paracrine signaling by chemical messengers released from the endothelium, such as nitric oxide, cross the blood brain barrier to affect neural processing of baroreceptor reflex circuitry. Subsequently, it was functionally shown that this process plays an essential role in the homeostatic reflex regulation of arterial pressure in both health and disease.

With the robust and eupneic respiratory motor pattern generated by the WHBP, the preparation has been adopted by multiple laboratories to understand neural mechanisms governing respiratory rhythm and pattern generation. In a recent study, a voltage sensitive dye was used to image spontaneous respiratory activity from the pre-Bötzinger complex (Fig. 2) in the WHBP [10]. This was the first time that the adult mammalian central respiratory rhythm generator had been visualized during eupneic-like activity. The study unearthed the temporal and spatial organization of neural circuitry employed during eupneic- and gasp- like respiration which led to the conclusion that there was a significant difference in the spatial organization of the gasp relative to the region exhibiting eupnea.Central Integration of Cardiovascular and Respiratory Activity Studied In Situ. Figure 2

Imaging brainstem respiratory network activity during eupnea and gasping in situ. Using the WHBP, we exposed the lateral edge of the medulla oblongata and injected a voltage-sensitive dye (Di-2 ANEPEQ) into the Pre-Bötzinger complex (a). Using a fast CCD camera, we performed phrenic nerve discharge (PND) triggered imaging of respiratory activity during eupnea (b) and hypoxic-induced gasping (c). Temporal and spatial patterns of activity were compared (d) and indicated distinct sites for the genesis of eupneic- and gasp- like respiratory patterns. The vertical colored scale indicates degree of depolarization with red being greatest. Abbreviations: D, dorsal; R, rostra; VRG, ventral respiratory group; V, VII, VIII, IX, X and XII are all cranial roots; ∫, integrated. From [10], with permission.

For the future, a new direction for the WHBP will be to image single functionally identified cardiorespiratory neurons using two-photon imaging (Fig. 3). This may be possible because the brainstem is mechanically stable and overlying superficial structures of the brainstem can be trimmed off to expose underlying cardiovascular and respiratory brainstem regions. Such an approach would allow unprecedented analysis of somatic and dendritic intracellular calcium fluxes and synaptic integration under different physiological and pathophysiological conditions (i.e., hypoxic (10.1007/978-3-540-29678-2_8) driven gasping) on physiologically characterized neurons.Central Integration of Cardiovascular and Respiratory Activity Studied In Situ. Figure 3

The future in situ. A future challenge would be to image neurons within an intact functional brainstem using multi-photon microscopy. Such a system equipped with a physiological recording rig and perfusion circuit for a WHBP is shown in (a). Using a custom designed stage it is possible to mount a WHBP beneath the objective turret of a two-photon microscope giving the potential to image brainstem cardiovascular and respiratory neurons (b).

Three electrophysiological techniques have provided a better understanding of the organization of the cMRF and its role in gaze control: (i) electrical microstimulation; (ii) single unit recording; and (iii) reversible inactivation of the MRF in awake, behaving monkeys (Macaca mulatta). (For details of preparation and localization see [3]. All experiments have been carried out with the approval of the Animal Care and Use Committee of the University of Connecticut Health Center.

Much of the human brain is devoted to acquiring visual information and then reorienting gaze (i.e. the combined movements of the head and eyes) to view targets of interest. The visual regions (e.g. V1, lateral geniculate nucleus, etc.) are organized topographically, such that neurons in a particular potion of the brain are only activated when a particular portion of the contralateral visual field is illuminated (see also Visual Cortex, connectivity and Visual Field Defects). This constitutes a spatial map: each individual neuron has a “visual receptive field”. Once a target of interest activates such a receptive field, a series of steps ensue that they activate excitatory burst neurons in the paramedian pontine reticular formation (the PPRF) of the brain stem whose temporal pattern of discharge is closely associated with the force and speed of contraction required to move the muscles of the eyes (a temporal code) [6]. A similar area for control of head movement is thought to reside in the nucleus reticularis gigantocellularis (NRG) located at the ponto-medullary junction. The neuronal mechanisms necessary to convert topographically organized sensory information into the temporal pattern of activity required to carry out motor actions has been termed a spatial to temporal transformation (10.1007/978-3-540-29678-2_19 STT).

One locus in the brainstem where this STT may begin is in the superior colliculus (SC) of the midbrain (see also 10.1007/978-3-540-29678-2_19). In the primate SC, neurons in the intermediate and deep layers are topographically organized and are activated before a specific sub-set of contraversive head and eye movements (i.e. gaze) called a “movement field” (see also 10.1007/978-3-540-29678-2_19). As a result, the discharge of neurons in the intermediate and deep layers of the SC encodes the gaze displacement signal in retinotopic coordinates required to shift the fovea using a combination of head and eye movement to fixate the new visual target. Critical to understanding the current discussion is the observation that the temporal discharge pattern of an individual SC neuron is poorly correlated with the velocity of the upcoming saccade, and the duration of discharge is only moderately related to gaze duration [7]. The movement fields of SC neurons are either “closed” (i.e. activity abates for gaze shifts larger than an optimal amplitude) or “open” (i.e. spike number increases to a maximum and then begins to decline, but does not disappear for movements beyond the optimal amplitude) (see also 10.1007/978-3-540-29678-2_19). In other words, these open movement fields are “non-monotonic” with respect to amplitude. Thus, while neurons in the SC, especially those in the caudal portion of the SC, discharge for a wide range of gaze movements, their activity could not encode the temporal pattern required to precisely activate the muscles moving the eyes or head. This suggests that the tectal output must undergo further processing before being incorporated into the activity of the short-lead, excitatory burst neurons of the PPRF whose discharge profile is closely correlated with both saccade duration and velocity.

A variety of techniques including electrical microstimulation, single neuron recording, and reversible inactivation, have demonstrated that neurons in the central Mesencephalic Reticular Formation (cMRF) participate in the control of saccadic eye movements. Four neuronal sub-types have been identified. Evidence that the burst duration and peak discharge of pre-saccadic, monotonically increasing open movement field cMRF neurons were closely associated with saccade duration and peak velocity, respectively, suggested that these neurons were similar to the directional long-lead burst neurons found in the PPRF [6]. Since cMRF neurons have direct projections to the PPRF, they could be a critical component of an indirect tecto-reticular-pontine pathway that parallels the direct tecto-pontine pathways [6]. The close association of the directional long-lead burst neurons of the pons, and the monotonically increasing long-lead burst neurons of the MRF with saccade dynamics, supports a role for these neurons in the conversion of gaze signals coded spatially in the SC (i.e. movement fields organized topographically) into the temporal pattern of activity (i.e. rate of discharge) found in the excitatory burst neurons of the pons [8].

The precise role of the other neuronal sub-types in the cMRF remains unclear. For example, the physiological characteristics of cMRF neurons with non-monotonically open movement fields and neurons with closed movement fields were most similar to neurons located in the superior colliculus. The evidence of hypermetric saccades following the reversible inactivation of the caudal portion of the cMRF suggested that these two neuronal sub-types could participate in a feedback pathway that provided the SC with a signal of the current change in horizontal eye position [9]. However, a distinct feature of the non-monotonic open movement field neurons was the close association of their burst duration with saccade duration, particularly the horizontal component. This suggests an alterative idea that these neurons could feedback a saccade duration signal from the omnipause neurons to the SC [7]. By extension, MRF neurons located rostral to the posterior commissure could provide a feedback signal corresponding with either the current change or duration of the vertical component of eye movement to the SC. Finally, the physiological properties of the post-saccadic neurons are unique and may reflect aspects of either a feed-forward or feedback mechanism for controlling movements of the head. Future work will be directed to better understanding the different roles played by these other neurons in the control of gaze movements.

Respiration serves to exchange O₂ and CO₂ between body and atmosphere. The initiation and maintenance of respiration occurs in the brainstem and involves a network of 10.1007/978-3-540-29678-2_18 (see Respiratory Network Analysis). The amount of respiration depends on the response of this neuronal network to inhibitory and excitatory afferent input from peripheral and central sensors (see Respiratory Reflexes; Carotid Body Chemoreception and Respiratory Drive). Here we discuss central chemoreception.

Central chemoreception was initially identified by the presence of respiratory responses to the perfusion of acidic fluids within the brain ventricles [1]. It was then localized to cells of unknown type at or beneath the ventral surface of the medulla oblongata by the direct application of acidic fluids [2]. But studies in vitro showed that neurons from many other locations within the hindbrain were responsive to pH changes see [3].

More recent experiments in vivo have provided support for the hypothesis that central chemoreception is a distributed property, i.e., that central chemoreceptor sites are widespread within the hindbrain [3–7]. In conscious animals, respiration is stimulated by the focal application of an acidic stimulus (increased CO₂) by reverse microdialysis at the (i Retrotrapezoid nucleus [3], (ii) caudal aspect of the nucleus of the tractus solitarius [3], (iii) medullary raphe [3] (see Medullary Raphe Nuclei and Respiratory Control), and (iv) the fastigial nucleus of the cerebellum [4]. Similar studies in anesthetized animals also showed chemosensitivity at the locus ceruleus and the 10.1007/978-3-540-29678-2_22 see [3]. Others have indicated the presence of chemoreception near the surface of the caudal ventral medulla as well [5].

In the retrotrapezoid nucleus, the chemosensitive neurons are glutamatergic and are identified by the presence of the vesicular glutamate transporter 2 (VGLUT2) see [3,7]. Their location is ventral to the caudal aspect of the facial nucleus just below one ventral medullary surface site at which earlier studies identified central chemoreception by application of acidic fluids [2]. The retrotrapezoid nucleus chemosensitive neurons express Phox2b, a gene associated with the development of the autonomic nervous system, and they also receive information arising in the peripheral chemoreceptors, the carotid bodies, via the nucleus of the tractus solitarius [7]. Inhibition or lesion of retrotrapezoid nucleus neurons diminishes the respiratory response to exogenously elevated CO₂ [3].

In the medullary raphe, the chemosensitive neurons are serotonergic [6]. They are located in the ventral medulla with some being quite close to the ventral medullary surface. The blood supply to the medulla originates at the ventral surface and many chemosensitive serotonergic neurons are anatomically situated close to penetrating arteries. Inhibition or lesions of medullary raphe serotonergic neurons diminishes the respiratory response to exogenously elevated CO₂ [3]. (see Medullary Raphe Nuclei and Respiratory Control).

The chemosensitive neurons of the caudal part of the nucleus of the tractus solitarius, the fastigial nucleus of the cerebellum and the caudal ventral medulla have yet to be identified in terms of their cell type.

The chemosensitive neurons of the locus ceruleus are catecholaminergic see [3]. Their location is in the dorsal pons and they too are situated close to blood vessels. Lesions of locus ceruleus and other catecholaminergic neurons diminishe the respiratory response to exogenously elevated CO₂.

ATP release from cells of unknown type at the ventral medulla has also been postulated as involved in central chemoreception see [3].

Central chemoreception is a widely distributed property with many types of neurons involved. Each of the central chemoreceptor locations has known neuronal projections to the major groups of brainstem respiratory neurons.

The control of respiration, designed to maintain normal levels of arterial PO₂ and PCO₂, depends on constant feed-back from peripheral and central chemoreceptors. The peripheral chemoreceptors in the carotid body located at the bifurcation of the carotid artery detect and produce rapid respiratory responses to small changes in arterial PCO₂. They also respond to lowered arterial PO₂ levels but the magnitude of the response is small until arterial PO₂ is about 70 mm Hg (normal arterial PO₂ = 90 mm Hg). The central chemoreceptors are not directly affected by changes in PO₂. Their response to changes in PCO₂ is robust (∼60% of the steady-state response to elevated CO₂ levels) but occurs more slowly than that of the carotid bodies [8,9]. These differences between central and peripheral chemoreceptors in response magnitude and dynamics indicate the presence of complex chemical-feedback system for CO₂ that governs respiration [8]. In order to maintain normal respiration, input from both receptors is necessary as lesions of either result in decreased respiration (and elevated arterial PCO₂ levels). Thus both contribute a drive to normal respiration. In contrast, sudden decreases in arterial PCO₂ can result in the cessation of breathing (apnea) if they occur in non-waking conditions. This hypocapnic apnea arises in the more rapidly responding peripheral chemoreceptors and is likely tempered by the slower central chemoreceptor response [8,9].

Central chemoreceptors are located within the parenchyma of the hindbrain. While some are situated quite close to arteries and vessels, in terms of overall function they are able to detect the pH of brain interstitial fluid [1]. In body fluids, CO₂ is in rapid equilibrium with water forming H₂CO₃ and then H⁺ and HCO₃ ions. Thus a pH sensing mechanism can be responsive to both primary changes in PCO₂ and in pH. Brain interstitial fluid pH is determined by three interacting processes: (i the level of alveolar ventilation, which determines the arterial PCO₂, (ii) cerebral metabolic rate, which determines the rate of CO₂ production, and (iii) cerebral blood flow. For example, brain interstitial fluid can become more acidic and respiration stimulated if the alveolar ventilation is decreased, if metabolic CO₂ production is increased, or if cerebral blood flow is decreased, which would slow the clearance of tissue CO₂. Central chemoreceptors can be viewed as detecting a product, pH, of these three vital processes. Further, dysfunction of alveolar ventilation or cerebral blood flow such as to impair O₂ delivery to the brain would be detected as a change in interstitial fluid pH thus allowing central chemoreceptors to act as indirect sensors of cerebral oxygen delivery.

Central chemoreceptors, activated by brain interstitial fluid pH, provide a drive to breathe. They may also act as a “buffer” to modulate rapid responses that might arise from the peripheral chemoreceptors. There are many locations and cell types that are chemosensitive and the specific functions of each within the overall system design are not well understood. For example, different chemoreceptor sites can interact dramatically. Simultaneous inhibition of the retrotrapezoid nucleus and the medullary raphe produces a much greater inhibitory effect on the respiratory response to exogenously elevated CO₂ than does inhibition of either site alone [10].

Although we do not know the exact mechanism by which the central nervous system (CNS) is damaged in multiple sclerosis (MS), an example of a CNS demyelinating disease, 10.1007/978-3-540-29678-2_22 and immune responses have been suggested to play important roles in its pathogenesis. Historically, among the various effector mechanisms of the immune system, the antibody was first suggested as an effector molecule; this has been supported by findings, such as oligoclonal immunoglobulin G (IgG) bands in the cerebrospinal fluid (CSF) and demyelinating antibodies in organotypic cultures. Later, T cells were regarded as another candidate effector. The delay was partly because analyses of cellular immune responses were established after analyses of humoral immune responses [1]. Currently, many in the field consider MS to be a major histocompatibility complex (MHC) class II-restricted CD4⁺ T helper 1 (Th1)-mediated disease, despite the observation that CD8⁺ T cells have been found more frequently than CD4⁺ T cells in demyelinating lesions of MS patients. This could reflect the technical feasibility of analyzing MHC class II-restricted CD4⁺ Th cells in vitro, compared with that of MHC class I-restricted CD8⁺ cytotoxic T lymphocytes (CTLs).

Endogenous antigens (usually made within the cells) are presented by MHC class I molecules, while exogenous antigens are presented by MHC class II molecules with a few exceptions [1]. Thus, if investigators add a protein of interest into cultures in vitro or inject protein into animals in vivo, protein in the extracellular space will be taken up by antigen presenting cells (APCs) and presented with MHC class II molecules, which enable detection of sensitized or stimulated CD4⁺ T cells specific for the protein of interest. On the other hand, to stimulate CD8⁺ T cells, a protein of interest needs to be expressed in the cytoplasm (in general) of APCs or target cells [1]. For this purpose, researchers usually either transfect APCs and target cells with cDNA encoding the particular protein or infect APCs and target cells with virus encoding the protein. In addition to the above technical difficulty stimulating CD8⁺ T cells, there are storage and handling problems in CD8⁺ T cell analyses. For instance, standard CTL assays have low throughput and require handling of chromium-51 (⁵¹Cr) that decays by electron capture and gamma (γ) emission with a short half-life of 27.7 days, while standard helper T cell assays, to detect lymphoproliferative responses, have high throughput using 96-well microtiter plates and require tritiated (³H)-thymidine that emits a weak beta (β) ray with a half life of 12.3 years. The ⁵¹Cr release assay is still a standard assay to detect CD8⁺ CTL responses, although several alternative detection methods have been introduced to detect CD8⁺ T cell responses [1], such as flow cytometry with intracellular cytokine staining and lactate dehydrogenase (LDH) release assays.

Various viruses have been found to induced demyelination in laboratory animals. One of the most studied experimental models is infection of mice with Theiler’s murine encephalomyelitis virus (TMEV) [2]. TMEV belongs to the family Picornaviridae. Although the precise mechanism of demyelination is not known, several effector mechanisms have been demonstrated to play important roles, including direct oligodendrocyte infection, TMEV-specific antibody (reviewed in [2]), TMEV-specific CD4⁺ Th1 cells, macrophages, apoptosis of oligodendrocytes and axonal damage [3].

We have demonstrated the generation of autoimmune CD8⁺ CTLs that cross-react with both virus and autoantigen in TMEV infection. Here, we will describe the history and direction of our investigations into the autoimmune CTLs generated in TMEV infection: (i) discovery and characterization of autoimmune T cells in bulk lymphocyte culture derived from mice infected with TMEV; (ii) establishment and characterization of T cell clones; (iii) generation of T cell hybridomas; and (iv) T cell receptor (TCR) and CTL epitope analyses, which will provide useful information to elucidate its pathogenesis and to develop tailor-made immunotherapies such as DNA immunization and tolerance induction (Fig. 1).Central Nervous System Degeneration Caused by Autoimmune Cytotoxic CD8+ T Cell Clones and Hybridomas. Figure 1

From mice infected with TMEV, we detected autoimmune CD8⁺ T cells that kill both uninfected and TMEV-infected syngeneic target cells. To characterize the autoimmune CTLs induced following TMEV infection, it was necessary to establish long-term T cell lines and clones. To maintain cytotoxicity, supplementation with interleukin (IL)-2 was necessary except during the first week of in vitro stimulation with TMEV-infected APCs. From the T cell clones, we established autoimmune CD8⁺ T cell hybridomas. Unlike the T cell clones, the T cell hybridomas can be grown and expanded without the addition of APCs or exogenous IL-2. We found that CD8⁺ T cell clones and hybridomas can recognize both a self antigen and a viral protein. Determination of TCR Vβ and CDR3 spectratyping of TMEV-specific T cell hybridomas, clones and uncloned bulk autoimmune T cell cultures will allow us to attempt modulation of TMEV infection by treating mice with Vβ antibodies or by vaccination with cDNA encoding TCR Vβ. Our novel experimental findings and approach can be applicable to elucidation of involvement of CD8⁺ CTLs in immune-mediated diseases, including MS, where CD8⁺ T cells have been demonstrated in demyelinating lesions. Analyses of TCR usage and CTL epitopes of viral and self antigen will provide useful information to elucidate its pathogenesis and to develop tailor-made immunotherapy such as DNA immunization and tolerance induction. This will extend the studies of several other groups where autoreactive CD4⁺ T cells have been described as containing degenerate TCRs that can recognize both autoantigen and microbial peptides. Since CD4⁺ T cells have been reported to recognize both self and microbial antigens from MS patients, similar autoreactive CD8⁺ CTLs could also contribute to the pathogenesis in MS and other virus infections.

CD8⁺ CTLs have been suggested to play an important role in not only eradication of virus but also demyelination in TMEV infection. To explore the role of CTLs, we monitored CTL activity using a 5h ⁵¹Cr release assay [4]. We utilized splenic mononuclear cells (MNCs) from SJL/J mice (H-2 ^s) infected with the Daniels (DA) strain of TMEV as effector cells and a syngeneic fibroblast cell line, PSJLSV (PSJL, H-2 ^s), as target cells. We stimulated the MNCs with TMEV-infected APCs for 1 week in vitro, and used these stimulated cells as effector cells in CTL assays. To our surprise, the MNCs killed not only TMEV-infected PSJL, but also uninfected PSJL (syngeneic or autoimmune killing). The autoimmune CTLs showed the highest killing against syngeneic target cells, intermediate killing against F1 target cells (H-2 ^b/s), and low killing against allogeneic target cells (H-2 ^b). The phenotype of the CTLs was CD3⁺/CD4⁻/CD8⁺. The autoimmune killing required cell-to-cell contact and was mediated by the Fas-FasL pathway, not by the perforin pathway. Killing was associated with interferon (IFN)-γ production in enzyme linked immunospot (ELISPOT) assays [5]. The CTLs were efficiently induced by vaccinia virus (VV) encoding the DA virus capsid proteins, but not by APCs infected with GDVII virus, a non-demyelinating strain of TMEV. Injection of the CTLs into the brain of naïve mice caused meningitis and perivascular cuffing, not only in the brain parenchyma, but also in the spinal cord distant from the injection site.

By limiting dilution, we established CD3⁺/CD4⁻/CD8⁺ CTL clones [6]. The CTL clones showed MHC class I-restricted killing of both TMEV-infected and uninfected syngeneic target cells, although infected target cells were killed more efficiently. Intracerebral (i.c.) injection of the clones into naïve mice induced large CNS degenerative lesions with loss of myelin and oligodendrocyte apoptosis. In contrast, we did not see degenerative lesions in mice injected with control CD8⁺ T cells activated with concanavalin A (ConA) or CD8⁺ enriched lymphokine-activated killer (LAK) cells.

In most CTL assays, T cells mediate killing through direct lysis. For example, in Fas-mediated killing, APCs present antigen to and thereby activate CTLs. This leads to up-regulation of FasL on CTLs, enabling the CTLs to kill the APCs positive for Fas. But in rare instances, CTLs have been shown to mediate killing through bystander lysis, where the APCs present antigen to and activate CTLs, inducing expression of FasL, but the target is a third Fas-positive cell that lacks the appropriate MHC restriction or antigen presentation. Here, the FasL-positive CTLs recognize the antigen presented by MHC molecules on the APCs, and the CTLs kill Fas-positive target cells without the trimolecular interaction of TCR with antigen and MHC molecules on the target cells.

In our system, uncloned splenic populations and T cell clones contain APCs [6]. After the 1 week in vitro culture of bulk T cells with TMEV-infected APCs, live cell populations were separated by a density gradient and used as effector cells in CTL assays. Here, the effector cells consisted mainly of TMEV-specific T cells, since the majority of other cell types were not able to survive after 1 week in culture. However, we cannot rule out the possibility of the existence of small numbers of live TMEV-infected APCs mixed in among the effector cell populations. Thus, this raised the question that the killing by splenic bulk culture or CTL clones might be due to bystander killing of the target, not by direct lysis of the target. To rule out this possibility, we developed T cell hybridomas having similar specificities and properties to those of the T cell clones.

All T cell hybridomas were CD4^– (Fig. 2). The majority of clones were also TCR⁺, although three hybridomas were TCR^–. Some hybridomas express low levels of CD8 antigen, while others were CD8^–. Hybridomas produced IFN-γ when incubated with infected or uninfected PSJL cells and TMEV-infected astrocytes but not uninfected astrocytes (Fig. 3a). The IFN-γ production was inhibited by the addition of CD8 and MHC class I antibodies to the cultures, but not by addition of CD4 antibody (Fig. 3b). I.c. injection of some hybridomas resulted in CNS pathology characterized by parenchymal and perivascular cell infiltration and gliosis (Fig. 4).Central Nervous System Degeneration Caused by Autoimmune Cytotoxic CD8+ T Cell Clones and Hybridomas. Figure 2

We generated the autoimmune T cell hybridomas, using TMEV autoimmune CD8⁺ CTL clone cells, 8a-1A or 8b-1C, with the BWα-β- (BW-1100.129.237) cell line that lacks the α and β chains of the TCR. Using flow cytometry, we characterized the T cell hybridomas according to surface antigens. T cell hybridomas were tested for surface markers by using monoclonal antibodies directed against CD4, CD8 and TCR. The surface phenotype of TMEV-induced autoimmune T cell hybridomas was CD3⁺ CD4⁻ CD8^dim TCR⁺ (8a-1A1, top and 8b-1Cs, middle) or CD3^dim CD4⁻ CD8⁻ TCR⁻ (8a-1Ae, bottom). CD8 may be preferentially down-regulated during prolonged in vitro culture or fusion, because the parent T cell clones were originally highly positive for CD8. FITC, fluorescein isothiocyanate. PE, phycoerythrin.

One of the characteristics of organ-specific 10.1007/978-3-540-29678-2_1 is that the development of the disease is closely associated with, or induced by, a particular type of T cell reactive to organ-specific antigens [7]. Thus, it has been postulated that autoantigen-reactive T cells bearing particular types of TCRs are expanded clonally during the course of the disease. Based on limited usage of the TCR repertoire in some types of experimental autoimmune encephalomyelitis, a particular single TCR has been suggested to be involved in encephalitogenicity (V-region disease hypothesis). More recently, to identify TCRs expressed on clonally expanded T cells, there are at least three types of analysis: complementarity determining region (CDR) 3 spectratyping, single-stranded conformational polymorphism (SSCP) and heteroduplex analysis [7]. CDR3 is encoded by the V(D)J junctional sequences and may directly contact the MHC-bound peptide, thus conferring T cell specificity for a particular peptide-MHC complex (CDR1 and 2 are encoded by germline and allow the TCR to interact with the α helices of the MHC molecule). In SSCP and heteroduplex analyses, each band on the gel represents expansion of a particular single TCR clone. In contrast, each band demonstrated in CDR3 spectratyping simply represents those TCRs with a CDR3 region of the same size. Therefore, the nucleotide sequence of the CDR3 region of the TCR clones derived from expanded bands should be determined to confirm the presence of clonal expansion in the bands [8].

TCR usages and CDR3 spectratyping are the methods that show TCRs of oligoclonally expanded T cells, compared with control samples. The methods do not require knowledge of putative autoantigens that are not known in most autoimmune diseases, including MS. Nucleotide and amino acid sequences of the CDR3 region of the TCR clones derived from the spectratypes of interest can be used to determine whether there is a clonal expansion and whether specific CDR3 motifs are used or not. For example, Matsumoto and colleagues [9] found that the Vβ5.2 spectratype is expanded more frequently than other Vβs in MS patients, suggesting that the finding provides useful information for designing TCR-based immunotherapy in MS. Thus, information on the identified pathogenic TCRs can be used in the prognosis of the disease or future treatment using antibodies and DNA vaccination against TCRs.

It is controversial whether specific TCR Vβ or CDR3 motifs are used in TMEV infection; some reports support specific TCR usage and others do not. We attempted to determine TCR usage by TMEV-induced CD8⁺ T cell hybridomas, using flow cytometry with a mouse Vβ TCR screening panel, containing monoclonal antibodies against Vβ 2, 3, 4, 5.1, 5.2, 6, 7, 8.1, 8.2, 8.3, 9, 10^b, 11, 12, 13, 14, and 17^a TCR. We found that hybridoma, 8a-1A1, was positive for TCR Vβ3 and 8b-1Cs was negative for all antibodies included in the panel. TCR Vβ region gene sequence analyses also showed the presence of Vβ3 in 8a-1A1 (Table 1).Central Nervous System Degeneration Caused by Autoimmune Cytotoxic CD8+ T Cell Clones and Hybridomas. Table 1

TCR Vβ chain gene segment usage in TMEV specific hybridomas^a

There are several methods currently employed for identifying antigens recognized by CTLs: (i) molecular biology approaches using cDNA libraries from a microbe or tissue of interest [10]; (ii) direct acid elution and amino acid sequence analysis of MHC-associated peptides; (iii) epitope mapping conducted with a series of overlapping peptides in cases where the antigenic protein is known; (iv) the use of algorithms that employ known MHC binding motifs for epitope predictions within a protein of interest; and (v) synthetic combinatorial libraries that do not require knowledge about T cell specificity or MHC restriction. Of these methods, expression cloning of cDNA libraries and peptide elution have been the preferred methods to identify antigens recognized by T cells for which no molecular information is available. ELISPOT assays have improved the sensitivity and efficiency of T cell antigen cloning from cDNA expression libraries [10].

We have developed T cell hybridomas that are easy to grow and produce IFN-γ in ELISPOT assays in response to TMEV-infected or uninfected PSJL, or infected astrocytes but not uninfected astrocytes. These T cell hybridomas can be used to define the viral epitope as well as the self epitope that is recognized by the cross-reactive TCR. We are using astrocytes transfected with a pCMV vector encoding each of the viral capsid proteins, VP1, 2, 3, and 4 [2] as well as the P2 or P3 regions of the TMEV genome encoding nonstructural proteins [2]. We will use a similar approach to identify what cellular antigen(s) is recognized by the TMEV-induced autoimmune CTL by transfecting astrocytes with an expression library constructed from the CNS of SJL/J mice. Once the protein is identified, either deletion mapping or overlapping peptides can be used to identify the self epitope. If it is a cell-specific protein found only in oligodendrocytes, then one could predict that direct killing of the oligodendrocyte would be a plausible mechanism for the demyelination. In vivo roles of TMEV-induced autoimmune CTLs can be clarified by sensitization of TMEV-infected or uninfected mice with (i) the CTL epitope peptide (or its altered ligand) of TMEV and autoantigen, emulsified in complete Freund’s adjuvant (disease induction and exacerbation) or in incomplete Freund’s adjuvant (treatment or tolerance induction); (ii) recombinant VV; or (iii) cDNA encoding the epitope. Induction of the CTLs that recognize both virus and autoantigen can result in viral clearance in the CNS of mice infected with TMEV, while the CTL induction in uninfected mice can lead to CNS degeneration.

Disorders of the central nervous system (CNS) cover a wide range of diseases from depression, Alzheimer’s disease (AD), multiple sclerosis (MS), Parkinson’s disease (PD), 10.1007/978-3-540-29678-2_3 and viral encephalopathies. However, 10.1007/978-3-540-29678-2_14 currently available on the market are not as effective as originally anticipated. Furthermore, side effects associated with long term use of these medicines discourage their use in chronic CNS pathologies. In order to develop effective treatment strategies, insight into the molecular basis of these disorders is essential. The complexity of the brain architecture, its components and neurocircuitry has made this task immensely challenging. The advent of rapidly evolving technology has enabled us to understand the common hallmarks of the pathophysiology of most of these disorders. The root cause in each one of them may differ from the other, but all of them show the involvement of inflammatory responses initiated by cells of the immune system in the brain. In the current essay, the primary factors involved in the etiologies of AD and HAD are described, and both primary and secondary factors contributing towards 10.1007/978-3-540-29678-2_14 (Autoimmune, Chronic) in major CNS disorders are summarized in Fig. 1.Central Nervous System Disease – Natural Neuroprotective Agents as Therapeutics. Figure 1

The primary and secondary factors involved in the etiology of major neurodegenerative disorders of the brain (AD, PD, HAD and MS) are outlined in this figure. The primary factors ( Lewy body, Aβ peptide, diffuse plaques, well defined amyloid plaques, neurofibrillary tangles, demyelinated neurons, HIV with gp41, and Nef are shown in the inner dotted eclipse. The secondary pro-inflammatory mediators, 10.1007/978-3-540-29678-2_3, activated microglia. T cells, macrophages, COX enzymes, astrocytes, free radicals, and NF-κB are shown in the outer eclipse. The neurodegeneration due to these factors is symbolised by a cartoon of damaged neuron in the centre.

Neurodegenerative disorders are marked by the complexity of their pathogenesis. The primary etiologies of most of these disorders differ from each other, but most of them show some common pathological hallmarks. Ingredients of natural origin offer significant potential for the development of effective treatment strategies, by their specific actions on the root cause or by targeting common pro-inflammatory mediators. However, proper study of route of administration, bioavailability, side effects, adverse interactions and standardization of dose is necessary for the development of efficient neuroprotective agents.

Primary Sjögren’s syndrome is generally diagnosed between the ages of 30 and 60 years and affects women nine times as often as men. The main features are dry eyes and mouth, often involving caries, but the disease can become systemic and affect other organs, leading to symptoms such as muscle and joint pain, itchy skin, vaginal dryness, gastroesophageal reflux and dry cough. Among neurological findings, peripheral neuropathy is most common, affecting between 10 and 20% of patients. It may cause progressive and initially distal tingling/numbness and weakness of the upper and lower extremities (polyneuropathy), asymmetrical sensory and/or motor deficits (e.g., radiculopathy, multiple mononeuropathy) or cranial nerve dysfunction.

The symptoms and course of CNS involvement are also heterogeneous. Patients may experience a sudden onset of focal deficits, such as hemiparesis or speech disturbance, suggesting stroke. They may also develop symptoms subacutely (e.g., transverse myelitis, optic neuritis) or over months and years (e.g., chronic myelopathy), and the course of disease can be progressive or relapsing-remitting. Rarely, CNS disease in pSS may manifest as seizures. Symptoms are sometimes discrete and not easily detected in routine neurological exams. This applies especially to mild cognitive disturbances, which may reflect brain damage, but can also be associated with fatigue and/or depression, other common symptoms in pSS.

It is important to note that CNS disease may occur before typical sicca symptoms (10.1007/978-3-540-29678-2_19) [5].

The prevalence of reported CNS disease is controversial and ranges from 0 to 62%. Reasons for the differences in prevalence in previous studies are (a) the use of diverging diagnostic criteria [1–3], (b) varying definitions of CNS involvement and (c) the selection of patients from varying populations [4]. To standardize the diagnostic criteria, the revised international (American European Consensus Group) classification criteria [1] were formulated in 2002. Definitions of CNS involvement differ regarding the inclusion or exclusion of nonfocal symptoms, specifically cognitive and neuropsychiatric disturbances. Studies disregarding patients with cognitive and neuropsychiatric impairment tend to underestimate the prevalence of pSS-related CNS symptoms. Moreover, a selection bias towards more severe CNS disease is likely to occur in tertiary referral centers, which may help explain the high prevalences reported in some hospital-based studies [6]. Another reason for discrepancies in the reported prevalence of CNS-involvement has recently been suggested: CNS disease may be more rarely associated with immunological markers of pSS than, for example, PNS involvement [5].

Sjögren’s syndrome is an autoimmune disorder in which immune cells destroy the exocrine glands that produce tears and saliva. At the beginning of the disease process, plasma cells and lymphocytes infiltrate the periductal salivary tissue. The original glandular structure is replaced by dense infiltrates of lymphocytes. Sjogren’s syndrome may result from T-cell abnormalities or may be caused by a deficiency of T-lymphocytes and subsequent hyperactivity of B-lymphocytes and the production of autoantibodies. Both environmental and genetic factors are likely to contribute to the immunological dysregulation that occurs in pSS [7].

The mechanisms of neurological disease in pSS are still unclear. Regarding peripheral nervous system involvement, different mechanisms seem to be associated with specific clinical features. Thus, sensory ataxic, painful and trigeminal neuropathy may be related to more immediate neuropathic processes than multiple mononeuropathy and multiple cranial neuropathy, which appear to result from vasculitis [8].

Evidence on the pathology of CNS damage is diverse. Some studies have suggested ischemic mechanisms. Other possible mechanisms are mononuclear cell infiltration in CNS tissue, immunologically mediated CNS vascular damage and the action of antibodies (antineuronal and/or anti-Ro/SSA antibodies [5]). The types of tissue damage observed in pSS-associated CNS disease vary greatly, in accord with the heterogeneous mechanisms mentioned above. Thus patients may develop severe conditions resulting from vascular damage or obstruction, such as subarrachnoid hemorrhage and ischemia; other types of tissue injury, such as meningoencephalitis and transverse myelitis indicate demyelination, and may be hard to distinguish from multiple sclerosis (MS).

Among animal viruses, arboviruses are unique in that they are transmitted by blood-sucking arthropods (vectors) to vertebrates, a mode of transmission commonly known as biological transmission involving the three essential components: virus, vector and vertebrate. Arboviruses generally require horizontal transmission by arthropod vectors among vertebrate hosts for their natural maintenance. On the basis of their arthopod vector, arboviruses that cause neurological infections can be classified as mosquito borne, tick borne, and viruses transmitted by different species of sandflies and by other vectors.

Among mosquito-borne viruses some members of Alphavirus genus in the family Togaviridae represent an important group of neurological disease agents. Alphaviruses of neurological interest include Western Equine Encephalitis virus and Eastern Equine Encephalitis virus and Venezuelan equine encephalitis virus, which can cause severe disease in horses and encephalitis in humans. Human outbreaks of all three of these viral diseases occur shortly after outbreaks are observed in horses.

Alphavirus encephalitis results in either localized or diffuse signs of cerebral dysfunction. Signs of meningeal irritation (meningoencephalitis) are nearly always present but may not be evident in the very young, the very old or the comatose patients. Inflammation of the leptomeninges may occur in some patients without evidence of brain dysfunction (10.1007/978-3-540-29678-2_1). Onset of neurological disease is preceded by a period in which the patient has an influenza-like illness. Encephalitis may follow quite soon after the onset or may follow days or weeks later [1].

While in the New World the arboviruses of neurological importance belong to the Togaviridae family (with the exception of West Nile virus), in the Old World (Europe, Asia and Africa) the viruses involved belong to the Flaviviridae family and are transmitted by mosquitoes and by ticks, as illustrated in Table 1.

Flaviviruses that cause neurological disease can be classified on the basis of their mode of transmission as tick-borne or mosquito-borne viruses. In Europe the most important agent of human disease is the Tick-borne encephalitis virus (TBE), transmitted by ticks; in Southern and eastern Asia the Japanese encephalitis virus (JEV) is the most important pathogen transmitted by mosquitoes. Recently the Flavivirus West Nile spread from the Old World to America and since 1999 it represents an important cause of neurological disease in the United States.

The Bunyaviridae are a large group of viruses that infect a diversity of arthropod vectors and animal hosts. They have a worldwide distribution and can be the cause of human illness. The most important viruses in the family Bunyaviridae that produce neurological disease in humans belong to genus Bunyavirus, transmitted by mosquitoes and Phlebovirus, transmitted by different species of sandflies.

It is well known that alcohol abuse results in structural and functional damage to the brain. Damage to the CNS is clearly evidenced in alcoholic individuals by a significant reduction in both brain weight and brain volume compared to control subjects. Interestingly, the neuropathology caused by alcohol appears to be region and cell type specific. For instance, the cerebral cortex, hypothalamus and cerebellum are quite vulnerable to the adverse effects of alcohol. Additionally, while both neurons and astroglia are affected by alcohol, astroglia are particularly susceptible to the detrimental effects of alcohol. Ethanol alters astroglial cell function and proliferation and the reduction in brain size is likely due in part to the cytotoxic effects of ethanol on astroglia. These ethanol-mediated effects on astroglia are potentially very important given that astroglia are essential for neuronal survival and function and are instrumental in response to infectious and traumatic insults to the CNS. The mechanism responsible for ethanol-induced brain damage is not fully understood but a major contributor appears to be inflammation. That is, ethanol induces inflammation in the brain and alters inflammatory pathways in the brain. These changes in inflammatory pathways likely contribute to brain damage but may also alter CNS immunocompetence and response to injury. For instance, risk and fatality of bacterial meningitis is greater in patients with alcoholic liver cirrhosis than in patients with non-alcoholic cirrhosis [1]. The poor outcome of alcoholic individuals with bacterial meningitis is likely a consequence of altered BBB function and CNS immunocompetence, given the involvement of BBB breakdown, leukocyte infiltration and neuronal injury in the pathogenesis of bacterial meningitis. Another instance in which alcohol abuse may be particularly detrimental to infection related neuropathogenesis is 10.1007/978-3-540-29678-2_8 (HIV) infection. More specifically, neuroimaging analysis indicates common loci of neuropathology in the human brain between HIV infected and alcoholic individuals suggesting that co-occurrence of these diseases may compound neuropathology [2]. Experimental findings suggest that ethanol and the HIV protein, 10.1007/978-3-540-29678-2_20, synergistically increase oxidative stress and proinflammatory gene expression in the brain [3]. These clinical insights, as well as other key experimental findings have led researchers in the field to target several inflammatory molecules as likely molecules involved in ethanol modulation of CNS immunocompetence and response to injury. The remainder of this essay will discuss key insights regarding ethanol effects on these inflammatory molecules.

There are four members of the JAK family (JAK1, JAK2, JAK3 and TYK2) and seven members of the STAT family (STAT 1, 2, 3, 4, 5A, 5B and 6). JAKs have a molecular weight of 120–140 kDa and contain seven JAK homology (JH) domains. The C-terminal JH1 domain has tyrosine kinase activity while JH2 has a pseudokinase structure but no catalytic function and regulates JAK activity. Regions JH3 to JH7 are required for interaction with the receptor. The molecular weight of the seven STATs ranges from 80 to 115 kDa. They also show a related structure with an N-terminal dimerisation domain and a central SH2-domain that are required for STAT dimerisation. The SH2 domain also contains a conserved tyrosine residue that serves as a substrate for the JAKs. Phosphorylation of this tyrosine is essential for STAT activity. Adjacent to the dimerisation domain are a coiled-coil domain that is involved in protein-protein interactions and the DNA binding domain. The transcriptional activation domain (TAD) is close to the C-terminus.

The SOCS family currently contains eight members, SOCS1-SOCS7 and CIS that range in molecular weight from 22 to 63 kDa. Members of this family share two common motifs – a central SH2 domain interacts with phosphorylated tyrosine residues and the C-terminal SOCS-box mediates ubiquitination and degradation of the SOCS protein.

A variety of diseases may give rise to CP (Table 1). The most common and well described central pains are central post-stroke pain and CP in spinal cord injury and 10.1007/978-3-540-29678-2_13, but any lesion along the sensory neuraxis from the dorsal horn to the brainstem, thalamus, subcortical white matter and probably cortical areas may cause CP [1]. Patients with 10.1007/978-3-540-29678-2_16, which is dominated by rigidity, bradykinesia and tremor, may also experience pain and sensory disturbances, but some of these pains may be related to dystonia and fluctuations in anti-parkinsonian medication [1]. Patients with 10.1007/978-3-540-29678-2_5 may have pain as part of a seizure or aura [1].Central Pain. Table 1

Etiology of central neuropathic pain

Central pains are characterized by a specific lesion or disease of the CNS and

Pain located in a neuroanatomical area with partial or complete sensory loss.

These symptoms/signs may occur in various combinations, but do not necessarily have to be present altogether. The underlying disease itself may also influence the pain and sensory pattern and contribute to heterogeneity of the core phenomena of CP.

Stimulus-independent pains are spontaneous pains and can be either continuous or paroxysmal. The character differs, but shooting, shock-like, aching, cramping, crushing, burning types of pain are descriptors that have been used. The pain may be described as superficial or deep or both. Other sensations such as 10.1007/978-3-540-29678-2_16 and 10.1007/978-3-540-29678-2_4 may be present spontaneously or evoked (e.g., ongoing tight or tingling sensations evoked by touching the area). The onset of CP varies, but in most cases patients develop CP within three to six months after their central nervous system lesion. Any delayed onset of neuropathic pain should prompt an examination for other causes (e.g., syringomyelia in cases of spinal cord injury). After a 10.1007/978-3-540-29678-2_19, CP may be distributed in a hemi body fashion or it may affect a smaller part of one side with sensory disturbance (e.g., part of a limb). In lateral medullary 10.1007/978-3-540-29678-2_9, the symptoms may be crossed with pain in one side of the face and the other side of the rest of the body (Fig. 1).Central Pain. Figure 1

Examples of distribution of central pain in a patient with central post-stroke pain following a lateral medullary infarction (a), a patient with at- and below level neuropathic pain in spinal cord injury (b), and a patient with central pain following multiple sclerosis (c).

An essential part of neuropathic pain is loss of sensory function. In some cases, sensory changes are subtle and a thorough sensory examination is needed, including perhaps the use of quantitative methods. Abnormal temperature and pain sensibility is the most consistent abnormality in post-stroke pain and it is suggested that a spino-thalamo-cortical sensory deficit is a necessary, albeit not a sufficient condition for the occurrence of CP [2–4]. The sensory deficit may be dissociated from a decrease in thermal and pinprick sensations and a relative preservation of vibration and other somatosensory functions. In addition to the sensory deficits, some patients may have a paradoxical sensitivity to cold and heat such that cold is perceived as hot and vice versa.

There is limited information on the frequency of CP. In a prospective study that included 207 consecutive stroke patients, 8% developed CP within the first year after their stroke [2]. Lesions of the thalamus and lateral medullary infarction are associated with higher risks of developing CP. Multiple sclerosis pain is likewise frequent [6], and Österberg et al. reported that 28% of patients with multiple sclerosis have CP [7]. In spinal cord injury, CP occurs in about 30–45% [8]. Pain at the level of injury seems to have an early onset, while CP below the level of injury may develop later after the spinal injury. Both types of pain tend to persist despite attempts at management [8]. Although some studies have indicated a higher incidence of CP in patients with incomplete lesions, other studies suggest that there is no relationship between the extent or site of spinal lesion and the presence of pain. Older age at time of injury has been found to be related to spinal core injury neuropathic pain.

The mechanisms responsible for CP are still unclear, but various theories have been advanced to explain these pains. The frequent incidence of evoked pain and decreases in mechanical thresholds in painful areas suggest the presence of hyperexcitability, and clinical and experimental studies indicate the presence of sensitization of 2nd or 3rd order neurons in the CNS that have lost their normal patterned input [3]. After a central nervous system lesion, several changes, including release of glutamate, up-regulation of sodium channels, activation of glia and loss of inhibition, are thought to increase the excitability of central neurons from which abnormal input may arise. In addition, the thalamus is thought to play a key role in CP [3] and bursting activity and reorganization has been demonstrated in the thalamus following central lesions. Disinhibtion due to partial lesions and imbalance between pathways has also been suggested to contribute to the development of CP. Among the more interesting recent theories, Craig has suggested that CP is due to loss of a normal inhibitory effect exerted by cool-signaling pathways from lamina I projecting to the thalamus and insula [9]. According to this hypothesis, a lesion of the lateral cool projection system disinhibits the medial system of heat-pinch-cold neurons passing from Lamina I to the medial part of the thalamus. This disinhibition results in a release of cold allodynia, burning and ongoing pain [9]. Disruption of thermosensory integrations leads to a disinhibition of thalamocortical neurons that respond to noxious inputs and a sensation of burning pain [9].

Like other chronic pain conditions, CP is a complex psychological experience which may have consequences for daily activities, sleep, cognition, emotion, behavioral and social relations and a broad approach to the treatment is essential. There is limited data on the pharmacological treatment of CP. Gabapentin, pregabalin, tricyclic antidepressants, lamotrigine and cannabinoids are treatments that have been shown to relieve CP, but other drugs like serotonin-noradrenaline reuptake inhibitors and opioids have not yet been studied in CP conditions [10]. Gabapentinoids and antidepressants are often considered first drugs of choice, but as in other neuropathic pain conditions these drugs, even when given in effective and tolerable doses, only reduce pain to a variable extent and other drugs or drug combinations may be considered. Patients with CP often have concurrent medical problems and impairment, are treated with multiple drugs with unwanted side-effects, and they may be elderly, which should be considered when treating CP.

The concept of central pattern generation was introduced in the early part of the twentieth century to account for experiments which demonstrated that deafferented hind limbs in anaesthetized cats were still able to produce rhythmic movements/muscle contractions [1]. This observation suggested that the rhythmic pattern of alternating contractions of flexor and extensor muscles underlying limb movements during locomotion are generated centrally within the spinal cord without the requirement of sensory feedback from the contracting muscles.

This “central control hypothesis” contradicted the “peripheral control hypothesis” of locomotion that was prevalent at the time. The “peripheral control hypothesis” considered the reflex as the basic functional unit in the nervous system and proposed that rhythmic movements (e.g. walking, swimming) are caused by the activation of alternating reflexes, i.e. contraction of a flexor muscle causes the activation of a reflex that triggers contraction of the antagonistic extensor muscle, which in turn activates the reflex that causes contraction of the flexor muscle leading to rhythmic movements. Furthermore, it was thought that the sequential activation of individual reflexes, where the action of one reflex causes a sensory response that triggers a second reflex and so on (10.1007/978-3-540-29678-2_18), also underlies the control of complex behavioral sequences.

Based on a wide range of studies in both invertebrates and vertebrates this debate has been settled in favor of the central control hypothesis and the CPG has emerged as a general principle of neuronal organization. However, it has also been recognized that phasic sensory feedback has an important role to play in shaping CPG output.

It should be noted that CPGs do not only generate rhythmic activity that directly controls motor behaviors, but that they also play a role in CNS activity patterns that are believed to be important for cognitive functions (e.g. hippocampal gamma and theta rhythms [2]). However, this essay concentrates on motor pattern generation as it has proved particularly useful to study the organization and function of CPGs.

Mechanisms for the generation of rhythmic activity in CPGs have frequently been divided into two broad categories – pacemaker-driven CPGs and 10.1007/978-3-540-29678-2_14 oscillators. Pacemaker-driven CPGs rely on neurons with intrinsic bursting properties (10.1007/978-3-540-29678-2_9), so called 10.1007/978-3-540-29678-2_5, for their rhythm generating ability (Fig. 1a). The inherent ability of endogenous bursters (pacemaker) to generate rhythmic membrane potential oscillations that drive bursts of activity is due to the specific interplay between various ion channels. Most commonly, the sustained depolarization of the membrane potential during a burst, a so called 10.1007/978-3-540-29678-2_16, is caused by voltage-activated persistent Na⁺ currents, voltage-activated Ca⁺⁺ currents or the activation of NMDA receptors. Calcium-dependent K⁺ currents or slow I_A currents are the most common channels responsible for the termination of the plateau potential and the repolarization of the membrane potential. The activation of a hyperpolarization-activated slow depolarizing current I_h by the repolarization at the end of the burst is frequently responsible for the initiation of the next burst. In some pacemaker neurons, so called conditional bursters, the bursting property is dependent on the action of modulatory neurotransmitters (e.g. serotonin, dopamine).Central Pattern Generator. Figure 1

Diagrams of various CPG rhythm generating mechanisms. (a) Pacemaker CPG. The upper diagram shows various ion channels that commonly underlie the endogenous bursting property in pacemaker neurons (I_Nap: persistent Na⁺ current, LVA I_Ca: low-voltage activated Ca⁺⁺ current, I_K(Ca): Ca⁺⁺-dependent K⁺ currents, I_A:slow activating K⁺ currents, I_h:hyperpolarisation-activated inward currents). The lower trace shows a schematic representation of the electrical activity pattern in a pacemaker neuron. The grey arrows indicate which ion channels are responsible for the different phases of the bursting pattern. (b-d) Schematic representations of the network configurations and activity patterns of three types of network oscillators. The diagrams at the top of each panel shows the connectivity between the network elements. Circles denote inhibitory synapses, whilst bars denote excitatory synapses. The traces below the diagrams show the activity pattern inthe correspondingly labelled network elements. See text for more details.

Most insights into the rhythm generating mechanisms of CPGs have been derived from in vitro experiments using isolated nervous system preparations that were completely deprived of sensory information. Whilst these experiments have proved extremely useful to demonstrate that the basic rhythm generating properties do not require sensory information, they ignore the role of sensory feedback in rhythm generation. However, in the intact animal CPGs do not operate in isolation but receive constant inputs from peripheral sense organs and proprioreceptors (Fig. 2). The influence of sensory feedback on pattern generation is obvious in locusts where removal of phasic sensory feedback from wing proprioreceptors significantly reduces wingbeat frequency. The natural burst frequency of the flight CPG can be restored by electrical stimulation of afferent sensory fibers in phase with the CPG activity pattern [6]. Similarly, sensory feedback significantly influences the rhythmic activity in the lamprey swim CPG as can be demonstrated by alternate bending of the caudal segments to simulate natural swimming movements. This activates stretch receptors along the spinal cord that can entrain the internal locomotor rhythm generated by the swim CPG to the frequency of the external movements [7]. Extensive evidence also exists for an important role of sensory feedback in mammalian locomotion. Overall sensory feedback appears to fulfill three main functions. Firstly, it can provide a corrective signal to adapt CPG activity to changes in an unpredictable environment as is necessary for the accurate step placement in a rough terrain or the adjustment of a wingbeat in a turbulent air stream. Secondly, sensory feedback can provide timing cues about ongoing activity resulting in the stabilization of the pattern. For example, stepping movements in cats become more variable in the absence of sensory feedback [7]. Similarly, mechanosensory feedback from the lips reduces the variability of bite intervals in the pond snail Lymnaea [8]. Thirdly, sensory feedback provides information about the position of the body or a limb within space which is important for adapting and planning further movements. Thus, in most systems central pattern generating mechanisms and sensory feedback mechanisms are closely integrated and interact to produce a robust rhythmic activity pattern that can be rapidly adjusted to cope with unpredictable environmental disturbances.Central Pattern Generator. Figure 2

Diagram of interactions between command systems, CPGs, motor systems and the environment. CPGs need to be considered in the context of the entire organism and its interaction with the environment to fully understand their function. Whilst CPGs can generate a basic motor pattern, this pattern is influenced by command and modulatory signals from higher order command systems as well as feedback from the motor system and the environment. Black arrows: command signals, white arrows: feedback signals.

CPG driven activity patterns can be active continuously throughout the life of an organism (e.g. mammalian respiration) or can be short episodic events triggered by a specific stimulus (e.g. fish escape response). The study of command systems that drive CPG activity has concentrated particularly on well-defined, robust episodic behaviors in relatively simple preparations that are reliably triggered by a specific stimulus. These preparations promised the possibility of identifying specific neurons, so called command neurons that can trigger a specific CPG activity/behavior. The definition of what actually constitutes a command neuron has been intensely debated and it has been proposed that only neurons that are both sufficient and necessary for the initiation of a specific behavior should be considered command neurons [9]. However, very few neurons actually fulfill these stringent criteria. Whilst there are a range of neurons, in particular in invertebrates, that fulfill the sufficiency criteria (e.g. slow oscillator and some cerebral buccal interneurons for activating the feeding CPG in molluscs such as Lymnaea and Aplysia, the Mauthner cell for the escape behavior in fish, etc.), very few also fulfill the necessity criteria (e.g. the dorsal ramp interneuron for activating escape swimming in the marine mollusc Tritonia).

This observation is consistent with the recognition that CPG activity can be driven by different stimuli and that there are usually parallel pathways that all contribute to the activation of a CPG. Thus, it is not surprising that most neurons that can drive a CPG do not appear to be absolutely necessary to trigger activity in a specific CPG. Furthermore, CPGs are flexible and can generate different activity patterns depending on the precise nature of the stimulus and an organism’s requirements. Whilst the different patterns utilize the same muscle groups, motoneurons and CPG interneurons, the sequence and phase relationship of activation of these elements can differ. For example, the Aplysia feeding CPG can produce ingestive and egestive motor patterns. These different patterns can both be driven by activity in the cerebral-buccal interneuron 2 (CBI-2). However, if CBI-2 is activated on its own the elicited feeding pattern is more ingestive-like, whilst co-activation of CBI-2 together with CBI-3, a second cerebral-buccal interneuron produces a more egestive-like activity pattern [10]. There are now many examples of modulatory interneurons that can affect the pattern of CPG activity. Some of the most striking and best characterized examples of network reconfigurations by the action of neuromodulators and higher order interneurons have been provided by studies of the pyloric and gastric mill CPGs in the crustacean stomatogastric nervous system. Here, it has been analyzed in great detail how different neuromodulators can alter CPG activity, how individual CPG interneurons can switch between different CPGs, and how higher-order interneurons can cause the complete reconfiguration of CPG networks to produce different activity patterns [4]. Thus, CPGs are not hard-wired, but dynamic 10.1007/978-3-540-29678-2_16 that can be reconfigured by the action of higher order interneurons, which are part of a general command system. This is clearly considerably more efficient than individual hard-wired CPGs for different behaviors (e.g. walking, running and jumping in mammals). Furthermore, the inherent flexibility in the CPGs and command systems that drive CPG activity provides the basis for behavioral choice as it enables motor patterns to be chosen for and adapted to specific requirements.

The insular cortex is the site of cortical representation of visceral, pain, and temperature sensation [2]. Nuclei in the ventromedial portion of the thalamus relay these sensory modalities to the insula. This cortical area is connected to the amygdala, lateral hypothalamus and brainstem autonomic nuclei, and is also interconnected with the anterior cingulate and ventromedial prefrontal cortices. The anterior cingulate cortex has a major role in regulation of affective behavior, and modulation of bodily arousal via the autonomic nervous system [4]. It has extensive connections with the prefrontal cortex, amygdala, hypothalamus, and brain stem and receives thalamic inputs involved in arousal and relay of nociceptive information.

The amygdala nuclear complex attaches emotional significance to sensory stimuli, including pain, and initiates the autonomic responses associated with emotion, including fear [5]. The amygdala receives inputs from all sensory modalities via projections from the brainstem, thalamus, and association areas of the cerebral cortex. The central nucleus of the amygdala is the effector structure of the amygdala complex and projects to the hypothalamus and brain stem areas involved in autonomic, endocrine, and motor expression of emotional responses [5].

The hypothalamus has a central role in the integration of autonomic and endocrine responses required for homeostasis and adaptation to internal or external stimuli. It is subdivided functionally into a periventricular zone, involved in circadian and neuroendocrine control, a medial zone involved in control of foraging behavior, and a lateral zone controlling arousal and motivated behavior. Several hypothalamic nuclei innervate brain stem and spinal targets controlling sympathetic and parasympathetic neurons. These include the 10.1007/978-3-540-29678-2_16 (PVN), the dorsomedial nucleus, the arcuate (infundibular) nucleus, and the posterior lateral hypothalamus (perifornical region) [1–2].These hypothalamic regions contain separate populations of neurons that project to different subsets of preganglionic neurons to generate distinct patterns of autonomic response according to specific stimuli.

The PAG consists of different longitudinal columns that receive specific inputs from sensory pathways, hypothalamus, and cerebral cortex and initiate stimulus-specific autonomic, somatic, and antinociceptive responses to external stressors [6]. The parabrachial nucleus, located in the dorsolateral pontine tegmentum, is a major relay center for converging visceral, nociceptive, and thermoreceptive information to the forebrain and contains separate subnuclei involved in taste, salivation, gastrointestinal activity, cardiovascular activity, and respiration [2,3]. The NTS is the first relay station for taste and general visceral afferents in the brainstem and conveys this information to all central autonomic regions, both directly and via the parabrachial nucleus. The NTS is also critically involved in all medullary reflexes controlling cardiovascular, respiratory, and gastrointestinal functions [3,7,8].

The VLM contains neurons that control sympathetic vasomotor tone, cardiac function, respiration, and endocrine function [3,7,8]. The rostral VLM contains glutamatergic neurons that provide the major tonic excitatory input to the sympathetic preganglionic vasomotor neurons and mediate most descending and reflex influences controlling arterial pressure. Epinephrine synthesizing C1 neurons of the rostral VLM contribute to this glutamatergic input and are required for sympathoexcitatory reflexes. The caudal VLM contains GABAergic neurons that, via their projections to the rostral VLM, mediate the baroreflexes and other sympathoinhibitory reflexes. Norepinephrine synthesizing A1 neurons of the caudal VLM project to the hypothalamus and participate in control of endocrine function, including secretion of arginine vasopressin. The medullary raphe contains serotonergic neurons that project to the spinal cord and control nociceptive, sympathetic, and respiratory functions. Medullary raphe-spinal pathways are involved in thermoregulatory responses, including skin vasoconstriction [2,3].

The insular and anterior cingulate cortices, amygdala, hypothalamus, and PAG form a functional unit that has a critical role in integrated responses to stress, emotional responses, and motivated behavior [1,4,5,6]. These areas receive and integrate inputs from several sources. Inputs from visceral receptors, nociceptors, and thermoreceptors reach these areas via both the dorsal horn and the NTS. The dorsal horn receives inputs from the dorsal root ganglia and projects via spinothalamic and spinobulbar pathways [2]. The NTS relays inputs from taste receptors, baroreceptors, chemoreceptors, pulmonary, and gastrointestinal receptors, carried via the facial, glossopharyngeal, and particularly the vagus nerves [2,3]. Both the dorsal horn and the NTS project to the parabrachial nucleus, hypothalamus, and amygdala, as well as to the thalamus, which then relays visceral inputs to the insular and anterior cingulate cortices [2]. At all these levels, there is integration of visceral with pain and temperature sensations. Humoral information, including levels of circulating peptides such as angiotensin II or cytokines, reaches the central autonomic structures in part via the circumventricular organs, which lack a blood brain barrier. These include the area postrema at the level of the fourth ventricle, and the subfornical organ and vascular organ of the lamina terminalis at the level of the anterior wall of the third ventricle [1]. The central autonomic structures, either directly or via the hypothalamus, receive influences from the suprachiasmatic nucleus (circadian pacemaker), the limbic cortical areas, and the cholinergic and monoaminergic cell groups involved in behavioral arousal and regulation of the sleep-wake cycle.

The NTS, VLM, and medullary raphe are involved in autonomic reflexes and mediate the effects of rostral areas, including the amygdala, hypothalamus, and PAG, on sympathetic and parasympathetic outflow [3,7,8]. The medullary cardiovascular and respiratory reflexes have several features in common. Baroreceptor, cardiac receptor, chemoreceptor and pulmonary mechanoreceptor afferents provide an excitatory input to the NTS that via direct and indirect propriobulbar connections, activate or inhibit the sympathoexcitatory neurons of the rostral VLM, vagal neurons of the nucleus ambiguus or dorsal vagal nucleus, and the neurons of the ventral respiratory group [3,7,8].

The preganglionic sympathetic or parasympathetic neurons are the final central effectors of the forebrain and brainstem structures controlling autonomic output [2]. The sympathetic preganglionic neurons are located primarily in the intermediolateral cell column at T1–L2 levels of the spinal cord and are organized into different functional units that specific targets via the paravertebral ganglia, prevertebral ganglia, or adrenal medulla. The sympathetic outflow is critical for responses to stress, such as hypoglycemia or hemorrhage, control of arterial blood pressure, and thermoregulation. Descending pathways from the hypothalamus and brain stem exert a differential influence on the different populations of sympathetic preganglionic neurons so that there is patterned activation of preganglionic outflow according to the physiological needs. For example, the rostral VLM activates muscle and splanchnic vasoconstrictor preganglionic neurons for maintenance of arterial pressure, whereas the medullary raphe controls skin vasomotor preganglionic neurons related to thermoregulation [2,3].

The vagus nerve provides the most widespread cranial parasympathetic output. Vagal preganglionic neurons are located in the dorsal motor nucleus, which controls respiratory and abdominal viscera, and in the ventrolateral region of the nucleus ambiguus, which innervates the heart. The vagus has a critical role in beat-to-beat control of the heart rate and regulation of gastrointestinal motility and secretion [3]. The sacral parasympathetic outflow arises from the sacral parasympathetic nucleus, located at the S2-S4 segments of the spinal cord, and is critical for micturition, defecation, and penile erection.

The anterior cingulate cortex and the amygdala control autonomic responses associated with motivated behavior and emotion. The human anterior cingulate cortex is activated during goal-directed behaviors associated with sympathetic activation [4]. Stimulation of the anterior cingulate cortex elicits increases or decreases in blood pressure, heart rate or respiration; mydriasis; piloerection and facial flushing; salivation; nausea or vomiting; and bowel or bladder evacuation. The amygdala receives inputs from all sensory modalities, both directly via the thalamus or parabrachial nucleus, and indirectly after cortical processing in association areas, particularly the insula and anterior temporal cortex [5]. In humans, the amygdala is activated by exposure to emotionally arousing stimuli, passive viewing of facial expressions (particularly fear), and\conditioned aversive stimuli. Together with the orbitomedial prefrontal cortex, the amygdala is critical for emotional and decision making on the basis of previously experienced sensations. The central nucleus of the amygdala is the effector structure for emotional responses. Both directly or via the bed nucleus of the stria terminalis, it innervates the hypothalamus, PAG, NTS and VLM, which initiate sympathoexcitation, release of stress hormones, and motor responses, including startle and vocalization [5].

The hypothalamus is critical for integration of autonomic with endocrine and behavioral responses required for homeostasis and adaptation [1,2,9,10]. The hypothalamic autonomic nuclei receive direct input from the anterior cingulate cortex, insula, and hippocampal formation, amygdala, and basal forebrain, as well as ascending inputs from the NTS, parabrachial nucleus, and A1/C1 catecholaminergic neurons of the VLM. The PVN provides the most widespread autonomic output of the hypothalamus and is crucial for coordinated endocrine and autonomic responses to stress [9]. Different neuronal subpopulations of the PVN, including the magnocellular neurons that secrete AVP to the general circulation, the parvocellular neurons that synthesize corticotrophin releasing hormone and activate the pituitary-adrenocortical axis, and the neurons projecting to autonomic nuclei of the brain stem and spinal cord, are activated, in a stimulus-specific fashion, by hypoglycemia, hypovolemia, cytokines, pain, and environmental stressors [2,9]. An important group of neurons in the posterior lateral hypothalamus synthesize hypocretin (also called orexin) and provide widespread projections to the hypothalamus, brain stem, and spinal cord. Via these projections, the hypocretin/orexin neurons prevent abrupt transitions between wakefulness and sleep, promote food intake and regulate sympathetic function [10].

The PAG is a critical component of the circuits involved in emotion and stress responses, including those triggered by pain [6]. The different columns of the PAG receive specific inputs and generate stimulus-specific responses. The lateral column of the PAG, which receives well-localized cutaneous nociceptive inputs, initiates flight-or-flight responses characterized by sympathetic activation with hypertension and tachycardia and blood flow redistribution to the face (fight) or lower limbs (flight) responses; these responses are associated with opioid-independent analgesia. In contrast, the ventrolateral PAG, which receives poorly localized somatic, visceral, and muscle inputs, elicits hypotension, bradycardia, immobility, and hyporeactivity to the environment; this is associated with opioid-dependent analgesia. The lateral and ventrolateral columns of the PAG provide descending inputs to different targets in the VLM and ventromedial medulla, which mediate both the cardiovascular and pain-modulatory responses [6].

The rostral VLM has a critical role in tonic maintenance of arterial blood pressure [7]. Medullary reflexes are critical for control of the blood pressure, heart rate, respiration, and gastrointestinal function [3,7,8]. A typical example is the baroreceptor reflex (baroreflex), which provides a powerful moment-to-moment negative feedback regulation of arterial pressure that minimizes the fluctuations of arterial pressure during standing, exercise, emotion, and other conditions. An increase in arterial pressure activates mechanosensitive baroreceptor terminals in the carotid sinus and aortic arch. Baroreceptor afferents excite neurons in the NTS that (i) directly activate the cardiovagal neurons in the nucleus ambiguus (leading to a decrease in the heart rate; (ii) via GABAergic neurons in the caudal VLM, inhibit the sympathoexcitatory neurons of the rostral VLM controlling vasomotor tone in muscle and visceral blood vessels (resulting an a decrease in total peripheral resistance); and (iii) via polysynaptic pathways, inhibit AVP release from the hypothalamus. Unloading of the baroreceptors, as occurs during standing, elicits opposite responses vasoconstriction and tachycardia) [3,8]. Vasoconstriction of muscle and splanchnic blood vessels is critical to prevent orthostatic hypotension.

The central control of autonomic functions can be affected by focal or degenerative disorders. Ischemic stroke involving the insular cortex can produce cardiac arrhythmias, which are a potential cause of sudden death Limbic seizures arising from the amygdala or anterior cingulate cortex may produce cardiac arrhythmias, cutaneous vasomotor and sudomotor changes, mydriasis, vomiting, or respiratory manifestations. Hypothalamic disorders are commonly associated with disturbances in thermoregulation, which may be paroxysmal or chronic and are commonly associated with disturbances in the sleep-wake cycle and food intake. Neurologic catastrophes, such as head trauma and subarachnoid hemorrhage, may manifest with paroxysmal sympathetic hyperactivity (hypertension, tachycardia, pallor, excessive sweating, hypothermia or hyperthermia) due to activation or disinhibition of hypothalamic and medullary sympathoexcitatory regions, including the PVN and rostral VLM. Medullary lesions, such as tumors, strokes, or syringobulbia, may manifest with paroxysmal hypertension, orthostatic hypotension, cardiovagal failure, or sleep apnea. High spinal cord lesions interrupting descending inputs to the preganglionic neurons may manifest with orthostatic hypotension and thermoregulatory failure as well as with paroxysmal unpatterned reflex sympathetic activity triggered by bladder distension and other stimuli (autonomic dysreflexia). Neurodegenerative disorders, such as multiple system atrophy, produce sympathetic and parasympathetic failure due to loss of preganglionic sympathetic and parasympathetic neurons, as well as neuronal loss in the VLM, medullary raphe, and other central autonomic nuclei.

The vestibular system is divided into a peripheral portion, housed in the labyrinth of the inner ear, the vestibular portion of the 10.1007/978-3-540-29678-2_22 (= acoustic-vestibular) cranial nerve, and the central connections of the vestibular nerve. Clinically, disorders of the vestibular system are divided into peripheral (labyrinth) and central (CNS) [1]. Ear specialists (usually ear, nose and throat surgeons) may consider disorders of the VIII nerve as central but 10.1007/978-3-540-29678-2_14 regard them either as peripheral or extra-axial (outside the neural axis).

Central vestibular disorders can be classified in different ways, according to (i) the underlying pathological process (e.g., inflammatory, demyelinating, tumoural, vascular, 10.1007/978-3-540-29678-2_4, traumatic), (ii) topography (e.g., medullary, cerebellar, cortical) and (iii) system involved (e.g., vestibulo-spinal, vestibulo-autonomic, vestibulo-ocular, vestibulo-cortical). Whenever possible, a physician would apply all these classifications simultaneously to his/her patient; for instance a patient can have a degenerative disorder of the cerebellum, predominantly involving the vestibulo-ocular system, as in the down-beat nystagmus syndrome (see below).

In clinical practice it is customary to divide patients’ problems into symptoms, what the patient reports (e.g., dizziness) and signs, what the examining doctor finds (e.g., nystagmus). The main symptoms in patients with vestibular disorders are vertigo, dizziness, oscillopsia and unsteadiness.

Vertigo is an illusion of body movement. The more common form is rotational or “true” vertigo in which patients feel that they are spinning round. Patients can also report that they see the world spin around them. Rotational vertigo is a useful symptom for diagnosis as it indicates involvement of the semicircular canals or their central projections. Apart from the canals and will nerve (as in peripheral vestibular disorders) the more frequent lesion site inducing vertigo is the area of the vestibular nuclei in the floor of the IV ventricle in the pontomedullary junction. More rarely patients describe sensations of linear bodily motion, called linear vertigo and thought to reflect involvement of the otolith organs or their central pathways.

Dizziness is more difficult to define. Patients with central vestibular disorders often use terms such as light-headedness, giddiness, dizziness or rocking sensations to describe their symptoms. Although common, these symptoms are less specific than vertigo for indicating vestibular system disease as many general medical conditions (anemia, hypoglycaemia, arterial hypotension, psychological disorders) also provoke dizziness [1].

Oscillopsia is the illusion of movement or oscillation of the visual world. It is due to loss of stability of the visual image on the retina. Oscillopsia results from basically two situations, either the vestibulo-ocular reflex (VOR) is significantly reduced and as a result ocular stability during head movements is lost or the eye has involuntary movements such as nystagmus [2]. Bilateral peripheral disorders are usually responsible for the former whereas central disorders are usually the cause in the latter.

Unsteadiness. This is the result of direct disturbance in the vestibulo-spinal projection or functionally related structures, such as the vestibulo-cerebellum (the flocculo-nodular lobe). If the lesion is unilateral and acute, patients tend to fall to one side (lateropulsion), usually ipsilesionally as in lateral medullary stroke involving the vestibular nuclei (Wallenberg syndrome). If the lesion is bilateral or diffuse, as in progressive cerebellar degenerative disease, the patient is globally unsteady. It is important to realize that most central vestibular disorders are not confined to vestibular pathways and thus the unsteadiness observed in any individual patient is likely to result from involvement of other balance mechanisms, e.g., motor, cerebellar and proprioceptive pathways.

The clinical findings that allow a distinction between a peripheral and a central vestibular disorder to be made are emphasized here. Apart from the presence of abnormal findings on the general neurological examination, such as limb weakness, anesthesia or ataxia, most signs indicative of a central vestibular disorder concern eye movement abnormalities. These include various forms of nystagmus, briefly mentioned below, as well as abnormalities in smooth pursuit, vestibulo-ocular reflex suppression and saccades, not discussed here.

Glutamatergic afferences are coming from many brain areas including several cortical regions and some hippocampal structures. The feedback projections coming from olfactory cortex are the main centrifugal fibers innervating the olfactory bulb (Fig. 1). They project mainly onto somata or basal dendrites of granule cells, the main GABAergic interneurons of the olfactory bulb [1].Centrifugal Fibers in Olfactory System. Figure 1

Main glutamatergic projections onto olfactory bulb. The olfactory bulb receives feedback projections from every part of the olfactory cortex. It also receives glutamatergic projections from other cortical areas (including frontal cortex) and from hippocampal structures.

In addition to the massive innervation by glutamatergic terminals, the olfactory bulb receives inputs from neuromodulatory regions (Fig. 2). Cholinergic, noradrenegic, and serotoninergic fibers reach the olfactory bulb circuit to modulate the network activity at several synaptic and extrasynaptic levels.Centrifugal Fibers in Olfactory System. Figure 2

Neuromodulatory projections onto olfactory bulb. Neurons from the horizontal limb of the diagonal band of broca (HDB) are releasing acetylcholine (Ach). Neurons from the dorsal raphe nucleus (DRN) are releasing serotonine (5-HT). Neurons from the locus coeruleus (LC) are releasing norepinephrine (NE).

Centrifugal projections extending from olfactory and neuromodulatory structures act together to regulate activity of the main olfactory bulb. This concerted modulation of olfactory information processing illustrates the intensive crosstalk between these areas of the brain.

The large variety of centrifugal projections and cell types innervated by these fibers give this system a high level and various sources of plasticity. There are many situations in which the olfactory bulb needs to be highly plastic. Behavioral studies clearly demonstrate the major role of centrifugal projections in enhancement of spontaneous odor discrimination, olfactory learning, and recall of specific olfactory memories. Furthermore, computational modeling has suggested that these inputs may increase in contrast in mitral cell responses to various odorants.

Physiological studies have shown the possibilities of long lasting changes in the strength of centrifugal inputs and the excitability of olfactory bulb neurons. Centrifugal fiber’s stimulation in fish can induce long-term potentiation (LTP) at the mitral cell to granule cell synapse [8]. Additionally, in vivo recordings in anesthetized rats have shown that high-frequency stimulation in the GCL can induce LTP of centrifugal inputs to granule cells [9].

Another major source of plasticity in the olfactory bulb network is continuous neurogenesis in the adult, consisting of production of granular and periglomerular interneurons throughout the life of the animal. As interneurons are the main targets of centrifugal projections and as adult neurogenesis is regulated by olfactory experience and sensory activity [10], this extreme form of network plasticity may be controlled by the concerted action of neuromodulators and feedback excitatory projections.

Feedback projections from cortical structures play a major role in attentional processes in other sensory pathways, including the visual system. The existence of attentional mechanisms in the olfactory system is under debate; presence of the classic type of attentional processes, as in other sensory systems, is excluded by the fact that the olfactory centrifugal projections do not pass through the thalamus. However, some of the defects in olfactory performances reported in behavioral studies of animals with altered centrifugal innervation may be interpreted as an impairment of olfactory attention.

Structurally, the cerebellum consists of a foliated cortex and the deep cerebellar nuclei. The output neurons of the cerebellar cortex, the Purkinje cells, project to the cerebellar nuclei as the corticonuclear projection (cerebellar corticonuclear projection), and the output neurons of the deep nuclei convey the information processed in the cerebellum to the brainstem and thalamus. The cortex and corresponding regions of the cerebellar nuclei are organized into sagittal zones. Traditionally, the cerebellum is divided into three primary sagittal zones: the midline (vermal), the intermediate (paravermal), and the lateral (hemispheric) zones. However, detailed neuroanatomical studies indicate that there are at least eight such zones in higher vertebrates, depending on the species and the region of the cerebellum. In general, these more restrictive zones specify the location of Purkinje cells within the cerebellar cortex projecting to a specific medio-lateral location in the deep cerebellar nuclei. In addition these zones define the topographic distribution of the climbing fiber projection (the olivocerebellar projection) originating from specific locations in the inferior olive. The relation of these zones to the olivocerebellar system will be discussed below. These relationships are characterized in Fig. 1.Cerebellar Functions. Figure 1

Diagrammatic illustration of the cerebellar-olivary loop, a set of interconnections relating the nuclear projection of Purkinje cells (PC) to the projection of climbing fibers from the inferior olive to the cerebellar cortex. Note the corresponding projections from Zones A and B to the related regions of the deep nuclei. Inhibitory nuclear neurons project in turn to olivary neurons which ultimately terminate on Purkinje cell dendrites in the same sagittal zones. Brachium conjunctivum: ascending output projection from the cerebellar nuclei. In the cerebellar nuclei and inferior olive, clear cells are excitatory and filled cells are inhibitory.

One of the primary functions of the cerebellum is the real-time control and coordination of a wide variety of movements. Characteristically, the more precise and complex the movement and the greater the integration required for its execution, the more the cerebellum is involved in its control. In general, the cerebellum is particularly important for the coordination of discrete, goal-directed smooth pursuit movements of the eyes (10.1007/978-3-540-29678-2_19), control of 10.1007/978-3-540-29678-2_7, and the regulation of multijoint movements of the extremities, particularly those requiring the integration of postural changes with phasic limb movements. This structure is also important for the coordination of combined eye and hand movements.

Cerebellar circuits utilize both feedforward and feedback control mechanisms in regulating these movements. However, much of the literature emphasizes the importance of the cerebellum in predictive or feedforward control mechanisms (See [3] for Review). In general, feedforward mechanisms are utilized to generate preparatory modifications in motor output that enhance the performance of previously rehearsed or experienced movements. Feedforward control involving the cerebellum is particularly important in establishing 10.1007/978-3-540-29678-2_13 and scaling of responses when novel tasks require modification of movements to attain a target accurately.

An example of this is shown in Fig. 2. In these experiments [4], cerebellar patients and normal controls were asked to move a manipulandum from a start position to a specific target through a force field, a velocity-dependent directional load imposed on the movement.Cerebellar Functions. Figure 2

Trajectories of a control subject and two cerebellar patients (WFN: 28 and WFN: 51) from a start position (circle, lower left) to a target (circle, upper right). A–C: control trials. Catch trials, trials in which no force field was applied during the trials in D–I, are shown as dashed lines. D–F: trials during adaptation. G–I: trials during the test for retention. Note that the force field consistently displaced the trajectory of the cerebellar patients. However, the control subject effectively compensated and also retained the task. Catch trials illustrate that the control subject had learned a new strategy to compensate for the load, since in the absence of the load, the movements demonstrated an after effect, a movement approximately opposite to the one learned in order to compensate for the load. SCA, spinocerebellar ataxia. Figure is from [4].

Considerable evidence has implicated the cerebellum in the learning of a wide variety of motor behaviors. These range from classically conditioned reflexes to complex, operantly conditioned tasks. The specific contributions of the cerebellum to this function are reviewed in other entries in the Encyclopedia. Consequently, this overview will focus on the cerebellum’s involvement in higher functions other than motor learning.

Studies implicating the cerebellum in other nonmotor functions have utilized imaging techniques such as fMRI and PET to illustrate changes in the activity of cerebellar regions during the execution of certain complex tasks, or they have examined the deficits manifested by cerebellar patients in related behaviors. Acknowledging that studies of this type provide strong inferences that the cerebellum is involved in these behaviors, they do not implicate this structure causally nor do they indicate how the cerebellum might be involved. That said, there is substantial evidence for the involvement of the cerebellum in the following higher order functions: solving tasks requiring the manipulation and perception of objects during the solving of puzzles such as the Tower of Hanoi, certain word association and word selection tasks, perception of tone duration, the characteristics of imagined movements, and the perception of object shapes during ongoing movement. In addition, the occurrence of autism and schizophrenia has been associated with structural abnormalities in the cerebellum (See [6] for review).

The above sections emphasize that the cerebellum receives information from virtually every sensory system as well as from projections originating from structures important in motor control. In addition, the cerebellum plays at least some role in most if not all aspects of motor behavior. This heterogeneity of involvement has made it very difficult to assign a single function to this interesting structure. In attempting to integrate this information, general hypotheses have been proposed suggesting that the cerebellum acts as a “mediator” or “metasystem” for integrating information from multiple sensory systems on-line with information characterizing the task and the state of the organism in order to generate an optimized, well-coordinated movement ([7–9], see also the “context linkage” proposal of Thach [6]). The sensory systems convey information about the external environment or the influence of the environment on the body, each with its own unique reference frame (See [10] for overview of reference frames). These data must be effectively integrated with data represented in internal reference frames reflecting features of the intended movement and body scheme as well as with reference frames describing muscle space and execution space. Although the cerebellum and its afferent and efferent projections appear to be appropriately organized to contribute to this complex function, little is known regarding precisely how this integration is performed.

In addition to its role in regulating on-line motor behavior, the cerebellum also contributes to functions related to motor learning as well as other complex behaviors which are not movement related. The specific role the cerebellum plays in these higher order functions is still a matter of discussion. Similar to its role in regulating movements, its role in higher order functions may also be task- and condition-dependent. Consequently, its specific contribution to storing motor engrams and in regulating the storage-related processes at other sites remains uncertain and may be dependent on the type of behavior being learned and the conditions under which the task is being performed. Its role in other non-motor functions has been inferred largely from testing cerebellar patients and from imaging studies. The current evidence clearly shows that activity in certain cerebellar regions is modulated during the performance of certain higher order tasks, and that some complex higher order functions are impaired in cerebellar patients [6].

Many questions remain regarding the precise mechanisms by which the cerebellar cortex and nuclei contribute to the multiple types of neuronal and system interactions required for the execution of the wide variety of behaviors in which this structure is involved. Attaining the answers to these questions is confounded not only by the complexity of the integration required but also by the fact that the extent of the cerebellum’s involvement in both the on-line control of movement and higher functions associated with motor control appear to be task- and/or condition-dependent. Thus, the extent and the nature of the cerebellum’s involvement in any given movement is likely dependent on factors such as: the type or class of movement (reflexive, volitional, postural, etc.), the characteristics of the motor sequence, the novelty of the movement, the extent to which learning is required, the association of posture and movement, the occurrence of a perturbation during execution, and the requirement of feedforward and/or feedback control.

FTNs are usually only one or two synapses distant from the motor output. Thus, while Purkinje cells receive mixed sensory and motor signals, FTNs tend to be more related to the motor system. FTNs receive the terminals of cerebellar cortical Purkinje neurons, usually upon their somata and proximal dendrites. This results in powerful mono-synaptic inhibition that is also tonic, as the average firing rate of a Purkinje cell, at least in the cerebellar 10.1007/978-3-540-29678-2_6, is about 100 impulses/s. FTNs, in turn usually fire at high rates, ca. 120 impulses/s because they are bombarded by excitatory inputs (10.1007/978-3-540-29678-2_6). This balance between FTN afferent excitation and Purkinje cell inhibition determines the moment-to-moment firing rate of FTNs.

The head can be envisioned as a sphere mounted on a ball joint, the neck, and is thus free to rotate in pitch, roll, and yaw. The head and body can also translate linearly. (An example of linear translation is walking.) The brain needs to be kept informed of the linear motion and position of the head and body and of the angular motion of the head. Relevant information is carried by the primary vestibular afferents that originate within the vestibular labyrinth. The labyrinth, located within the inner ear, is composed of linear accelerometers that sense the impulsive and gravitational components of linear acceleration and angular accelerometers that sense angular head motion. The angular sensors, the semicircular canals, are three in number bilaterally, and anatomically situated in the pitch, roll, and yaw axes of head rotation. Primary vestibular afferents terminate within the vestibular nuclei in the brainstem, while a subset projects directly to the cerebellum. Fig. 1 illustrates these neural connections.Cerebellum – Flocculus Target Neurons. Figure 1

A schematic of the connections between the flocculus of the cerebellum and brainstem that control the VOR. FTN is a flocculus target neuron, FPN is a flocculus projecting neuron, PVP is a position-vestibular-pause neuron, Y Cell is a cell in the Y group of the vestibular nuclei, Int. is an interneuron, SVN is a superior vestibular nucleus, MVN is a medial vestibular nucleus, IIIrd and IVth nuc. are the oculomotor and trochlear nuclei, respectively.

The somadendritic morphology of vestibular neurons is correlated with their axonal projection targets. Vestibular VOR neurons project rostrally to the oculomotor nuclei and to the cerebellum. Many superior vestibular nucleus (SVN)-VOR neurons are FTNs. Mitsacos et al. [2,3] injected SVNs neurons with horseradish peroxidase for morphological study. Cells were identified as projecting to the oculomotor nuclear complex (VOR-SVN) or to the cerebellum. VOR-SVN neurons vary in shape from pyramidal or multipolar to ovoidal or elongated. Most neurons exhibited their longest dendritic extent along the rostro-caudal axis while their shortest extent was in the coronal plane. Most of the dendrites remained within the cellular boundaries of the nucleus. The branching pattern of VOR neurons is isodendritic [4] i.e. most dendrites follow a straight course and branch in such a manner that the primary dendritic segments are shorter than the secondary ones and these, in turn, are shorter than the tertiary ones. Fig. 2 illustrates examples of FPN and VOR neurons. Note the differences in dendritic trees.Cerebellum – Flocculus Target Neurons. Figure 2

Reconstructions of two superior vestibular nucleus neurons. The upper neuron is the soma and dendrites of an SVN-VOR neuron. The curved arrow points to a terminal dendritic formation, s is a dendritic spine, w is a wavy dendrite and a is the axon. The bottom neuron is an SVN-cerebellar projecting neuron. Arrowheads point to dendritic segments displaying an allodendritic branching pattern, p is a dendritic process, and w is a wavy dendrite. Calibration is 100 μm; arrow in the calibration bar points to the midline.

The 10.1007/978-3-540-29678-2_15 (cerebellum – oculomotor vermis) consists of lobules VIc and VII of the midline cerebellar cortex. It was first defined in the monkey as the region where low amplitude (<10 μA) electrical microstimulation elicits saccadic eye movements, but slower, smooth-pursuit like eye movements are also elicited. Prominent saccade, smooth-pursuit, and head-movement related signals are present in the discharges of neurons in this region [2]. Nearby regions that are not in the circumscribed oculomotor vermis (e.g. vermal lobules VIa,b and VIII) may nonetheless participate somewhat in the control of eye movements, since eye movements can be elicited in this area by microstimulation at moderate current strengths, and eye- and head-movement neuronal discharges can also been recorded.

The 10.1007/978-3-540-29678-2_6 (Cerebellum – Fastigial Oculomotor Region (FOR)) is a circumscribed region in the caudal fastigial nucleus, the most medial of the deep cerebellar nuclei, where saccades are elicited by microstimulation and where neurons exhibiting saccade-related and smooth-pursuit related discharges are found. This region corresponds closely with the part of the fastigial nucleus that receives Purkinje-cell input from the oculomotor vermis.

The 10.1007/978-3-540-29678-2_6 and ventral paraflocculus are contiguous structures adjacent to the cerebellar hemispheres and overlying the eighth cranial nerve. Due to past inconsistencies in the naming of the ventral paraflocculus and to similarities of its connections and neuronal discharges with those of the flocculus (see below), the two areas are sometimes lumped together as the 10.1007/978-3-540-29678-2_6. Collectively, they participate in the generation of smooth pursuit and the regulation of the VOR. The flocculus receives direct input from the vestibular portion of the eighth nerve, and so is one part of the vestibulocerebellum.

The nodulus and uvula are vermal regions on the underside of the cerebellum that corresponds to midline lobules X and IX, respectively. The nodulus and rostral uvula also receive direct input from the vestibular nerve and have heavy reciprocal connections with the vestibular nuclei. It is the second component of the vestibulocerebellum. The nodulus/uvula is integral to the velocity storage mechanism, and so, participates in controlling the time course and direction of prolonged vestibularly and optokinetically induced eye movements (see 10.1007/978-3-540-29678-2_22).

The ventral portion of the monkey posterior interpositus nucleus and adjacent portions of the caudal dentate nucleus have been implicated in the control of saccades. Inputs to this area derive from saccadic and/or smooth-pursuit regions of parietal cortex by way of the dorsal and dorsolateral pontine nuclei. The same inputs innervate the dorsal and ventral paraflocculus, which project back to the interpositus/dentate. This region also projects directly to the superior colliculus and interstitial nucleus of Cajal, and indirectly to the frontal eye fields; all structures known to participate in the generation of saccades. In addition, the interpositus contains neurons that exhibit saccade-related discharges, and its transient inactivation using the GABA agonist, muscimol, results in an upward bias in the endpoints of saccades (dysmetria). This data is suggestive but preliminary, and a better understanding of the role of the ventral posterior interpositus and dentate area will require additional data.

A second more rostral oculomotor part of the dentate nucleus, which may overlap the part of the first area, is an extension of the y-group of the vestibular complex. This region contains neurons that are excited during upward eye velocity during smooth pursuit and during upward head rotation with the VOR suppressed. In macaques, these eye and head signals are roughly equal, and approximately cancel during VOR in the dark (i.e. they encode 10.1007/978-3-540-29678-2_7). Some neurons have eye position sensitivity and most have saccadic eye-movement sensitivity. As the dentate/y-group area receives inputs from the paraflocculus and projects to the oculomotor nucleus, this region is thought to participate in generating vertical smooth pursuit.

Finally, evolution in humans produced a huge expansion of the lateral cerebellum along with its target nucleus, the dentate. There are strong indications that this expanded region participates in human cognitive functions. Accordingly, a portion of the lateral cerebellum has increased activity in functional-MRI studies of humans generating memory-guided saccades (10.1007/978-3-540-29678-2_13) and 10.1007/978-3-540-29678-2_1.

Oculomotor portions of the cerebellum receive direct input from the eighth nerve, the vestibular nuclei, the pontine reticular formation (paramedian pontine reticular formation (PPRF)), raphe nuclei in the pons and medulla, and indirectly from the superior colliculus and specific regions of the cerebral cortex. The latter are relayed through the 10.1007/978-3-540-29678-2_14 and the pontine nuclei (see below). The cortical inputs to these relay nuclei include the 10.1007/978-3-540-29678-2_6, the 10.1007/978-3-540-29678-2_19, parietal areas middle temporal (MT), medial superior temporal (MST), and 10.1007/978-3-540-29678-2_12. (Each of these areas is discussed more fully elsewhere in this Encyclopedia). A previous concept that each cortical region served one particular type of eye movement, (e.g. the FEF subserved saccades and area MST subserved smooth pursuit) has been replaced after the demonstration that the FEF, SEF, and LIP each have adjoining or partially overlapping saccade-related, smooth-pursuit related, and sometimes vergence-related areas, and that visual motion processing areas MT and MST have connections with the saccade-related as well as pursuit-related cortical and subcortical structures. Accordingly, both saccade-related, smooth-pursuit related and vergence-related signals have been recorded from neurons in NRTP and the dorsolateral pontine nuclei.

The inferior olive provides climbing fiber inputs to all portions of the contralateral cerebellum, but these inputs are not thought to be important in short-term signal processing and will not be considered further.

There are several targets of the cerebellar output that are best examined in the context of the pathways mediating each type of eye-movement. For the saccadic system, cerebellar efferents arise from the FOR and project to the saccadic burst generators in the contralateral pontine and midbrain reticular formations, the contralateral superior colliculus, and thalamus. The burst generators excite agonist and inhibit antagonist motoneurons to generate the saccade (see 10.1007/978-3-540-29678-2_2). For the smooth-pursuit system, different fastigial-nucleus efferents project to the vestibular nuclei and the pontine and midbrain reticular formations near the burst generators. An additional smooth pursuit pathway traverses the floccular lobe, which in turn, also projects to the vestibular nuclear complex. Specific output targets, which are better known for the floccular pathway, include the superior vestibular nucleus, the medial vestibular nucleus, the ventrolateral vestibular nucleus, and the y group. Signals are then conveyed to the ocular motoneurons to produce smooth-pursuit eye movements. There is also a projection of the floccular lobe to the basal interstitial nucleus of the cerebellum, whose function is not known. Details regarding these pathways are elaborated for each eye-movement system below.

Experimental lesions of the midline cerebellum (oculomotor vermis and caudal fastigial nucleus) produce substantial deficits in saccade generation and in smooth pursuit. Saccade deficits are more pronounced with lesions or chemical inactivation of one side. Inactivation of the caudal fastigial nucleus causes ipsiversive saccades to overshoot the target by up to a factor of two, and contraversive saccades to undershoot the target by as much as half. Unilateral lesions or inactivation of the oculomotor vermis cause reversed effects. Vertical saccades are misdirected for both types of lesions. Bilateral lesions cause smaller and more balanced 10.1007/978-3-540-29678-2_19 [1]. Smooth pursuit is similarly affected. Unilateral fastigial inactivation increases ipsiversive eye acceleration for ramp targets up to two times normal, and decreases contraversive acceleration as much as 30%. Bilateral inactivation of the FOR or lesions of the vermis decreased smooth-pursuit gains by 30–50% when subjects track periodic stimuli [1]. The residual smooth pursuit reflects the fact that the FOR/oculomotor vermis is only one of at least two smooth-pursuit pathways, and that pursuit is a closed-loop system.

Experimental lesions of the flocculus and ventral paraflocculus together produce a 50–60% deficit in smooth pursuit and in suppression of the VOR. The gain of the VOR in the dark is affected very little. The ability to hold eccentric gaze is severely affected, as the eye returns towards a neutral point with a 10.1007/978-3-540-29678-2_20 around 2 s. This has been interpreted as disruption of the velocity-to-position integrator (see 10.1007/978-3-540-29678-2_14) by loss of a high-gain feedback loop through the flocculus.

Localized cerebellar lesions in humans caused by infarcts, tumors, surgery, head trauma, or degenerative disease produce the same symptoms. However, such lesions are rarely as well confined as experimental lesions, so humans exhibit additional eye-movement disorders. These include several different forms of gaze deviations or inability to hold eccentric gaze and nystagmus [10].

The folding of nascent peptides into a specific three-dimensional mature conformation is essential to allow proper function of all proteins. Chaperones are a large family of factors that promote this process, preventing aggregation of misfolded intermediates and by targeting terminally misfolded proteins for degradation. Chaperone mediated protein folding is required in all cells and is particularly important in the nervous system, as aberrant protein aggregation of misfolded proteins is a hallmark of many of the major neurodegenerative diseases. Although the majority of proteins fold properly and are quite stable in their functional conformation, different environmental stresses that are placed on the cell can result in unfolding or misfolding of proteins. The neuronal synapse is a particularly challenging area in maintaining proper protein folding conformations due to a dense protein filled cytosol and rapidly changing ion and pH levels. Proteins with abnormal conformations can accumulate over time due to exposure to extrinsic or naturally produced oxidizing agents or environmental stresses and this can result in the protein quality control system being taxed beyond its ability to deal with these misfolded proteins. Therefore, chaperones are thought to play an integral role in the manifestation of neurodegenerative diseases whose common pathological feature is protein aggregates and neuronal cell death.

The neuron has long cellular extensions that require transport from the cell body where proteins are made to the dendritic and axonal termini. Synaptic proteins are translated at the cell body and must be transported to the site of action via vesicles that move along the cytoskeleton. A loss in efficiency of vesicle and cellular traffic increases the stress on the protein quality control network at synapses, thus increasing the chance that a toxic aggregation is undetected by the chaperone system. Components of the protein degradation pathway are common at the synapse and associate with endocytic vesicles, suggesting a link between cellular traffic and cytosolic chaperones. Although aggregates are the universal indicator of many neurodegenerative diseases the underlying cellular cause of neuronal loss has many different effectors, one of which includes the role of protein trafficking in maintaining protein quality.

The functional specificity of a protein is dictated by its conformation and folding is driven by the interaction between the aqueous cytosol and hydrophobic amino acids incorporated into the protein. Normally, the hydrophobic amino acids are driven into the centre of the protein to avoid the association with water and this drives the folding of the hydrophilic domains around them. A form of protein secondary structure called the β-pleated sheet is prone to aggregation due to exposed hydrophobic domains. Self-propagating aggregation is a fundamental property of 10.1007/978-3-540-29678-2_1, or aggregates of peptides that polymerize into a cross beta structure. Chaperones can prevent aggregate formation by recognizing stretches of hydrophobic amino acids that are exposed upon translation or misfolding and can initiate one of three pathways: promoting proper folding by an ATP dependent mechanism, preventing aggregation by shielding the hydrophobic domains that would normally aggregate with hydrophobic domains of other proteins or targeting misfolded proteins for degradation by recruiting ubiquitinating enzymes and subsequent degradation by the 26S proteasome (Fig. 1). The proteasome is a major component of the protein quality control system that recognizes ubiquitinated proteins that have been targeted for degradation by chaperones.Chaperones: Protein Trafficking. Figure 1

Generalized chaperone function. Chaperones, particularly the Hsp70 complex, associate with hydrophobic domains of newly translated proteins off the ribosome. The Hsp70 complex can aid in the proper folding of the protein into its functional conformation. Environmental stress and exposure to toxic agents can cause denaturation of the folded protein. Depending upon the degree of damage, and the type of chaperone associated with the misfolded peptide, chaperones can prevent aggregation of misfolded intermediates, promote refolding, or target the peptide for degradation by ubiquitination and subsequent degradation by the proteasome.

Protein aggregates are a unifying feature of a large number of neurodegenerative diseases that affect the human population (Table 1).Chaperones: Protein Trafficking. Table 1

Types of aggregates in common neurodegenerative diseases

Alzheimer’s disease (AD) is the most prevalent neurodegenerative disease where 10% of individuals over the age of 65 will eventually develop AD and 1–2% of the same demographic will develop Parkinson’s disease (PD). The diseases show diverse clinical manifestations as a result of cell death occurring in different neural subtypes in the brain where specific toxic peptides cause eventual death of those neurons and the formation of brain lesions. Alzheimer’s disease, Parkinson’s disease and the polyglutamine diseases represent a major class of neurodegenerative disorders that result from defects in protein folding and aberrant aggregation.

Alzheimer’s Disease: AD is characterized by the formation of both intra- and extracellular protein aggregates. The extracellular plaques are composed mainly of β-amyloid whereas the intracellular neurofibrillary tangles (NFT) are composed mainly of a microtubule associate protein, tau. It has been known for some time that amyloid precursor protein (APP) is associated with AD; APP is a single pass membrane protein that is normally trafficked through the endocytic and secretory pathways (Fig. 2). Cleavage of APP produces the disease causing peptide (β-amyloid) and this cleavage occurs during normal cell protein turnover by enzymes located in the endosomes releasing the extracellular domain. It is the location of the cleavage site that causes the formation of neutral or toxic β-amyloid. β-amyloid can be produced intracellularly during receptor recycling of cellular vesicles, and it can also occur in the ER/Golgi network prior to APP secretion, suggesting that the formation of the toxic peptide is a result of normal cell protein turnover. The reduced efficiency of vesicle movement with age or due to mutation may result in a traffic jam of cellular components in the axon; this decreased movement of vesicles may increase the susceptibility of cleaving APP into β-amyloid.Chaperones: Protein Trafficking. Figure 2

Production of β-amyloid in the neuron. Amyloid precursor protein (APP) is a single pass transmembrane protein that is recycled via the endocytic pathway. Once in the endosome an enzyme called β-secretase can cleave APP at amino acid 671 leaving a membrane bound fragment. As the endosome enters the lysosome another secretase (γ-secretase) can cleave the sAPPβ fragment into β-amyloid. β-amyloid can be degraded by the proteasome through chaperone-mediated degradation or, if the UPS system is overloaded, may aggregate into disease causing protofibrils.

Protofibril formation and protein misfolding are likely occurring throughout a lifetime and it may be that the loss of chaperone activity with age contributes to the late age of onset for the majority of non-inherited neurodegenerative diseases. Many unknowns still exist regarding the initial toxic agent in different neurodegenerative diseases but all potential therapeutic interventions must address the initial cause of disease progression. Chaperones provide a potential therapeutic target for dealing with the early stages of neurodegeneration because upregulation of chaperones has been shown in multiple different disease models to suppress the progression of the disease. How chaperones suppress disease progression is unknown but may act at multiple different stages including protofibril formation, prevention of fibrillar structures, promotion of protofibril refolding or degradation or promotion of amorphous aggregates. The complexity of the cellular effects of aggregation diseases indicates that simply interfering with one of the downstream effects may not prevent disease progression and neurodegeneration.

Chemical stimuli are molecules and ions and chemical senses are molecular senses. Molecules are transported to the contact of chemosensory cells either by air or by water. Yet, knowing the nature of the diffusion medium does not offer a sufficient criterion to classify sense organs. Even though the olfactory system detects airborne stimuli at low concentration in mammals and insects while the gustatory system detects waterborne, sapid molecules with a lower sensitivity, this is not the case for fish and aquatic crustaceans that have a true olfactory system detecting waterborne odorants with a high sensitivity [1].

Another system bringing chemosensory information is represented by the taste system specialized in the detection of chemical properties of potential foods and evaluation of their nutritional value. In contrast, the olfactory system has a much wider involvement in the detection of environmental and social odors. Yet, odors are also strongly involved in food intake and cooperate with taste as essential components of flavors.

Odorants emitted and sensed by individuals of the same species, for example sex attractants, are called pheromones. There are several categories of pheromones. Reproductive pheromones trigger sex recognition, courtship displays and sexual activity; in addition to these behavior-oriented functions, they also initiate long-term physiological – mainly hormonal – changes in the recipient animal. Maternal pheromones guide newborn towards mother nipples; recognition pheromones are involved in labeling the identity and social status of individuals; aggregation and dispersion pheromones maintain individual spacing. Odorants and pheromones (at least sex pheromones) were thought to be detected by two distinct sensory organs, the olfactory epithelium (OE) and the vomeronasal organ (VNO), respectively. In fact, the OE can also detect pheromones, whereas the VNO can also detect ordinary odorants.

All animals detect odorants using seven-transmembrane domain receptors that activate G protein-based signalling cascades. In 1991, Buck and Axel [2] discovered a wide family of these G protein-coupled receptors (GPCRs) expressed in the rat OE and proposed that they were chemosensory receptors called odorant receptors (ORs). Subsequent molecular and bioinformatic studies confirmed the pioneer discovery and indicated that mammals, birds, fish, and amphibians have large numbers of genes for olfactory GPCRs expressed in olfactory organs. The OR repertoire contains around 1,000 genes in mouse and rat, 500–750 genes in human, and 100 genes in zebrafish and catfish. Independently expanded families of chemosensory GPCRs have a chemosensory function in invertebrate species [1].

The VNO of the accessory olfactory system expresses two families of GPCRs called V1R and V2R [3]. Some empirical evidence suggests that V1Rs bind to small airborne chemicals whereas V2Rs recognize water-soluble molecules.

In the gustatory system, two GPCR families, T1R and T2R, are involved in the detection of sapid molecules inducing sweet, bitter and umami sensations [4]. Identified proteins, T1R2 and T1R3, might compose a heterodimeric receptor to sweet tasting molecules whereas the T2R family contains around 30 different bitter receptors. Another family of GPCRs, the heterodimere T1R1/T1R3 that responds to many aminoacids in mice is especially sensible to L-glutamate and L-aspartate in humans. This receptor is therefore a good but non exclusive candidate to the role of umami taste detection.

Gustation use channel-receptors differing from GPCRs to sense salt and sour tastes [5]. For example, one of the most studied Na + taste receptor is a Na + highly selective channel known as ENaC, the amiloride sensitive epithelial sodium channel. This receptor is an oligomeric complex comprising three homologous subunits. Receptors to sour taste can be ordered in two groups: one group includes a channel-like EnaC that conducts an inward proton current when protons are available in the oral cavity. The second group comprises several H + -gated channels.

Knowledge of trigeminal chemoreception that operates in both oral and nasal cavity has benefited from recent studies on primary somatosensory neurons [6]. Studies led to discovering receptors to irritant substances like capsaicin, the active principle of chilli pepper and isothiocyanates found in mustard oil and horseradish. These receptors are members of the TRP ion channel family. Interestingly, capsaicin-sensitive nerve fibers are also activated by noxious heat. TRPV1 (also known as vanilloid receptor 1 or VR1), can be activated by both capsaicin and high temperature. Alternatively, low temperature activates a cold receptor called TRPM8 that is also sensible to menthol, a chemical eliciting a sensation of cold.

The size of the OR gene superfamily varies considerably among species. Duplications have greatly increased the number of genes but deletions and inactivating mutations resulting in a large number of pseudogenes contributed to reduce the size of the functional repertoire [7]. The proportion of intact OR genes and pseudogenes is also quite variable. For example, the human genome contains less than 400 functional OR genes and over 400 OR pseudogenes. This decline of the OR gene repertoire is already visible, even though less marked, in apes and Old World monkeys, coinciding with the acquisition of a full trichromatic vision, which may indicate that improving visual abilities made olfaction redundant in some its adaptive functions.

Considerable variations in gene repertoire are also observed for VNO receptors, V1R and V2R. The differences in number of V1R and V2R functional genes among vertebrates seem to point to an asymmetric evolution followed by the two gene families [8]. Only very few intact VR1 genes are found in fishes whereas the mouse and rat have over 100 genes. In contrast, the V2R repertoire is virtually lost or severely reduced in many terrestrial vertebrates (in the cow, dog and primates, including humans), but not in the opossum and rodents that have a wide V2R gene repertoire. The existence of a functional vomeronasal organ in adult humans is highly questionable.

Olfactory neurons use two main intracellular signalling pathways utilizing cyclic nucleotides and phosphoinositide-derived signals. Cyclic nucleotide signalling is common in vertebrates and is thought to operate in nematodes (Caenorhabditis elegans) and arthropods [1]. Cyclic nucleotides target the olfactory cyclic nucleotide-gated ion channel through which calcium enters the cell and secondarily activates a chloride current. When activated, ion channels generate a graded voltage response that triggers action potentials. In crustaceans, olfactory receptors seem to use phosphoinositide signalling, whose target is a presumptive homolog of the TRP family of ion channels. Phosphoinositide signalling has been implicated in some way in olfactory transduction in other, phylogenetically diverse species, including nematodes, insects, fish and mammals.

Taste transduction is complex [5]. Bitter taste detection involves the taste- specific signalling G protein, gustducin. Two transduction pathways are simultaneously activated by receptors of the T2R family: receptor activation triggers a transient decrease in cAMP and cGMP, along with a transient increase in IP3. Like bitter- taste transduction, sweet-taste transduction has cyclic nucleotides and IP3 as intracellular messengers. Sugars and artificial sweeteners are supposed to use distinct pathways.

The peripheral organization of sensory pathways differs notably between the main and accessory olfactory systems, on the one hand, and the taste system, on the other hand. Several common features can be observed in olfactory systems of vertebrates and arthropods. In most animals, the primary olfactory afferents that are axons of receptor neurons, project to the CNS without intermediate synapsing. The first synaptic relay, that is the olfactory bulb in mammals, antennal lobe in insects and olfactory lobe in crustaceans, is similarly organized in arthropods and mammals. The olfactory afferents converge into the dense neuropile of glomeruli where they terminate on both projection neurons and local interneurons. In mammals, this projection is narrowly selective: all receptor neurons expressing the same type of OR converge onto one or two glomeruli. In turn, each projection neuron connects one or a few glomeruli to the primary olfactory cortex in mammals and the lateral protocerebron and corpora pedunculata in arthropods.

A similar organization pattern is shown by the accessory olfactory system. VNO receptor cells that are neurons project to glomeruli in the accessory olfactory bulb (AOB) located in the caudal part of the main OB. In those mammals that have both V1R- and V2R-expressing sensory neurons (rodents, opossum), the two populations separately project their axons to segregated (anterior and posterior) regions of the AOB. In all other examined mammals that have VNO, the projection system is uniform. Then, relay neurons directly project to the hypothalamus.

Gustatory receptor cells are grouped in taste buds inside three types of taste papillae. The spatial distribution of the different taste receptors in the tongue and the mouth is not homogenous but expression zones of different receptors overlap to some extent. Differing from olfactory sensory cells, taste cells have an epithelial origin, they are not neurons. The apical portion of a taste cell possesses fine expansions, called microvilli, equipped with taste receptors. Receptor activation by sapid molecules triggers ionic currents generating action potentials that are synaptically transmitted to fibers of the gustatory nerves (cranial nerves VII, IX, X). A single afferent fiber makes synapses with several taste cells. Taste afferents project to the rostral part of the Nucleus Tractus Solitarius (NTS) in the brain stem.

Chemical senses have the function of allowing animals to identify substances, objects, places or living beings on the basis of their molecular properties that induce sensations endowed with specific qualities. Understanding how the molecular identity of an odor or a taste is coded in corresponding sensory organs and pathways is a fundamental question. Some common principles and notable differences can be found between olfaction and taste. In the olfactory system it is generally agreed that each cell expresses a single type of OR, individual receptor cells can be activated by different odorants and individual odorants activate multiple receptor cells [9]. These properties give support to the concept of combinatorial coding [9] that is valid in both vertebrates and insects: an odor is represented as a specific combination of excited neurons that reproduces the specific combination of activated ORs. The other possible coding strategy consisting in dedicating neurons to the detection of a particular odorant (“labeled lines” coding), seems to be exceptional (lobster). Recent studies suggest that the discriminative capacity of the static combinatorial coding could be improved in projection neurons by including a temporal dimension.

There is less agreement regarding the coding of taste qualities [10]. Like in the olfactory system, in the taste system individual afferent fibers respond to multiple stimuli and even to stimuli representing different “basic” qualities: i.e., NaCl, HCl, sucrose, quinine. This can be explained because individual afferent fibers innervate several taste cells and taste cells individually display multiple chemical sensitivities. A combinatorial coding, originally called “across-fiber pattern” [10] coding could therefore operate in the taste system. However, the taste system differs from the olfactory system in that a limited number of taste qualities can be recognized in the former, whereas qualities cannot be consensually classified in the latter. A solution was proposed once to conciliate apparently opposite views: among several stimuli inducing responses from a taste afferent, one would be more efficient than every other; afferents could thus be grouped according to their “best” stimuli. This proposal failed to close the debate. Whether neural activity data support the view that taste qualities form a continuum or are more in favor of the traditional “four basic taste” theory is still a matter of individual belief. Falsification of either model seems to be difficult or impossible. Hopefully, unrevealing mechanisms of taste reception does not require previous solving of the conceptual puzzle.

Chloride is a unique ion in the cellular physiology of the nervous system. Whereas the electrically important inorganic cations (Na⁺, K⁺, Ca²⁺) are maintained with strong, predictable gradients across neuronal membranes (10.1007/978-3-540-29678-2_13), Cl⁻ anion gradients are variable, tailored to the physiological role of this anion in a given cell type at a given developmental stage. In some cells, Cl⁻ is passively distributed, adjusting its gradient according to the 10.1007/978-3-540-29678-2_18 of the cell. Activation of Cl⁻ conductances in those cells cannot therefore change the membrane potential; it can, however, attenuate 10.1007/978-3-540-29678-2_5 by reducing the 10.1007/978-3-540-29678-2_13. In other neurons, Cl⁻ concentrations are more tightly controlled, being either actively extruded or accumulated. The resulting shift in the Cl⁻ 10.1007/978-3-540-29678-2_5, by controlling whether activation of Cl⁻ channels is inhibitory or excitatory, has a major effect on excitability in the central nervous system (CNS). This effect can be dynamic, as in the 10.1007/978-3-540-29678-2_19, where there is a daily oscillation in the neuronal response to 10.1007/978-3-540-29678-2_7 activation, or it can be developmentally regulated, as in 10.1007/978-3-540-29678-2_19 containing GABA_A receptor Cl⁻ channels which switch during development from excitatory to inhibitory [1]. (For further information see essays on Chloride homeostasis and development and 10.1007/978-3-540-29678-2_9).

Chloride conductances serve a wide range of physiological roles in the nervous system. These functions result not only from channel activity but also from the working of a series of transporters that move Cl⁻, both passively and actively. Perhaps the most thoroughly studied neuronal Cl⁻ channel is the GABA_A receptor, a member of the cys-loop family of 10.1007/978-3-540-29678-2_14-activated (10.1007/978-3-540-29678-2_12) channels. As this protein is covered thoroughly in other articles, we will focus here on other Cl⁻-transport mechanisms. These other pathways fall into two broad categories: some have been identified at a molecular level, offering the possibility of genetic (and other) manipulations; others have been described only in terms of their functional behavior, observed in 10.1007/978-3-540-29678-2_16 recordings of neuronal cells but not yet cloned.

Chloride (Cl⁻) is the most abundant permeant anion in cells. In many non-neuronal cells, active transport does not maintain a Cl⁻ gradient across the neuronal membrane. The resulting passive distribution of this ion establishes an equilibrium potential for Cl⁻ (E_Cl) which is equal to the resting membrane potential (Vr). E_Cl is the membrane voltage at which there is no net flow of Cl⁻ across the membrane. Unlike non-neuronal cells, neurons express cation-chloride cotransporters (CCC) which precisely regulate Cl⁻ homeostasis throughout development and in the mature CNS [1–3]. Electroneutral cotransport of Cl⁻ and K⁺ and/or Na⁺ is mediated by members of the CCC gene family SLC12 (solute carrier family 12) [4]. The CCCs are all secondarily active cotransporters that utilize energy from extrusion of ion(s) down their electrochemical gradient(s) in order to fuel the transport of other ions against their electrochemical gradient(s).

The Na⁺ -K⁺ -2Cl⁻ (NKCC1) is a ~1,280 amino acid protein SLC12 gene family member, which is widely expressed in both epithelial and nonepithelial cells, including neurons and glia [4]. Early in development there is a high neuronal expression of NKCC1, which produces an influx of Na⁺, K⁺, and Cl⁻ (Fig. 1).Chloride Homeostasis and Development. Figure 1

The balance of CCCs determines [Cl⁻]_i during development. (a) NKCC1 is the dominantly expressed CCC in immature neurons. This inward transport of Cl⁻ results in a relatively high [Cl⁻]_i. (b) In mature neurons, the developmental up-regulation of KCC2, coupled with decreased NKCC1 expression, produces a net Cl⁻ extrusion which maintains a low [Cl⁻]_i.

The neurotransmitters GABA and glycine both bind to ionotropic receptors (GABAARs and glycineRs, respectively) which are permeable to Cl⁻. Early in development the dominant expression of NKCC1 maintains a high [Cl⁻]_i which maintains E_Cl more depolarized than the action potential threshold. Under these circumstances GABAergic and glycinergic synaptic transmission can produce action potential firing, and thus their actions are excitatory [2] (Fig. 2).Chloride Homeostasis and Development. Figure 2

E_Cl Determines the polarity of GABAergic and glycinergic synaptic transmission. (ai) In immature neurons E_Cl is more depolarized than the action potential threshold rendering GABAergic and glycinergic synaptic transmission excitatory. (aii) Under such conditions inward GABAergic currents are recorded electrophysiologically. (bi) In mature neurons, when E_Cl is more hyperpolarized than V_r, GABAergic and glycinergic synaptic transmission is inhibitory, producing outward currents (bii).

GABA is present and functional prior to the development of synaptic, where it has been shown in vivo to exert trophic functions during neuronal development. Premature lowering of [Cl⁻]_i in a subpopulation of ventricular neural progenitors, by in utero electroporation of KCC2, eliminated the tonic excitatory actions of GABA [3]. This resulted in a severe impairment of the morphological maturation of those newly born neurons. In addition to the role of excitatory GABA on neurite morphology, excitatory GABA also plays a critical role in the establishment of excitatory and inhibitory inputs within neuronal circuits. When KCC2 was prematurely expressed in immature tectal cells of the Xenopus tadpole tectum, there was increased GABAergic synaptic input to tectal neurons, while the normal developmental increase in AMPA receptor-mediated transmission was blocked [6].

Early excitatory glycinergic signaling also exerts trophic functions which are required for proper nervous system development. Early excitatory glycinergic signaling was perturbed in the zebrafish using morpholino oligonucleotides (morpholinos) [7]. The resulting knockdown of the glycine receptor α₂-subunit produced morphants with disrupted interneuron differentiation. This decrease in excitatory glycinergic signaling resulted in a reduction in the frequency of spontaneous glycinergic and glutamatergic synaptic transmission.

The excitatory trophic actions of GABAergic and glycinergic signaling are not restricted to development. In the dentate gyrus of the hippocampus, where neurogenesis continues into adulthood, excitatory GABA regulates neurogenesis, morphological maturation, and synapse formation. In particular, when NKCC1 expression is reduced and GABA’s actions are converted from excitatory to inhibitory, the dendritic development of newly generated granule cells is impaired [8].

Many of the excitatory actions of GABA and glycine are likely mediated by receptor-induced membrane depolarization, which regulates Ca²⁺ influx through both voltage-gated channels and neurotransmitter receptors. Membrane depolarization which triggers action potentials, will in turn activate voltage-gated calcium channels (VGCCs) allowing significant Ca²⁺ influx. In addition, membrane depolarization may be sufficient to remove the Mg²⁺ block from the Ca²⁺ -permeable NMDA receptor. The VGCC- and/or NMDA-dependent rise in [Ca²⁺]_i may then trigger Ca²⁺ -dependent signaling cascades responsible for developmental processes such as progenitor proliferation, neuronal migration, neurite growth, and synapse formation.

Peripheral nerve injury can lead to neuropathic pain, which results from hyperexcitability of dorsal horn neurons in the spinal cord. Nerve injury leads to an increased synthesis and activation of the ATP receptor on microglia. Recently, it was shown that ATP stimulates brain-derived neurotrophic factor (BDNF) release from microglia, which acts via the TrkB receptor to depolarize E_Cl in spinal lamina I neurons [9]. The depolarization of E_Cl was significant enough to convert GABAergic and glycinergic synaptic transmission from inhibitory to excitatory. A link between neurotrophic factors and Cl⁻ homeostasis has also been demonstrated in the hippocampus. Following epileptic seizure activity there is a BDNF-induced decrease in KCC2 expression [10].

In the mature CNS, when GABAergic synaptic transmission is inhibitory, physiological patterns of neuronal activity can regulate the strength of inhibition [3]. When hippocampal GABAergic synapses are repetitively stimulated with spike-timing induction protocols or when the postsynaptic neuron is stimulated alone, there is a Ca²⁺ -dependent decrease in KCC2 function. This activity-induced regulation of KCC2 leads to an increase in neuronal [Cl⁻]_i which depolarizes E_CI Because E_Cl determines the strength of inhibition, the activity-dependent increase in [Cl⁻]_i decreases the effectiveness of inhibition. Inhibitors of Ca²⁺ -dependent protein kinase C (PKC) abolished the postsynaptic spiking induced E_Cl shift, whereas activation of protein tyrosine kinases or phosphatases were not required, suggesting that Ca²⁺ may act via a PKC-dependent pathway to regulate KCC2. The postsynaptic spiking-induced and spike-timing induced down-regulation of KCC2 activity occurred within minutes, suggesting an alteration in membrane trafficking, and/or posttranslational modification of KCC2, as opposed to changes in gene transcription or protein synthesis.

NKCC1- and KCC2-mediated Cl⁻ homeostasis determine the polarity and strength of GABAergic and glycinergic synaptic. Early in development when [Cl⁻]_i is high, excitatory GABAergic and glycinergic synaptic transmission play important roles in progenitor proliferation, neuronal migration, neurite growth, and synapse formation. In the mature nervous systems when [Cl⁻]_i is low, both physiologically- and pathologically-induced neuronal activity can alter Cl⁻ homeostasis via a regulation of KCC2, which effectively weakens the strength of inhibition.

In 1936, the English physiologist Sir Henry Dale and the German-Austrian-American pharmacologist Otto Loewy shared the Nobel Prize in physiology and medicine for their discoveries of 10.1007/978-3-540-29678-2_1 (10.1007/978-3-540-29678-2_1) as a neurotransmitter. In fact, ACh was the first chemical transmitter to be recognized as a neurotransmitter. In cholinergic neurons, the enzymes choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) are synthesized at the cell body and then move to the nerve terminals [1]. Terminals of the cholinergic neurons contain acetyl coenzyme A (acetyl-CoA), produced by an intermediary metabolic step, and choline, taken up from the synaptic cleft and other parts of the extracellular space by an active, sodium dependent, high-affinity transport system. In the nerve terminals, acetyl-CoA and choline are combined by the enzymatic activity of ChAT to synthesize ACh molecules [1]. These Ach molecules are stored within synaptic vesicles and are released in response to action potentials. Released ACh activates postsynaptic as well as presynaptic receptors. Since cholinergic neurons synthesize both ACh synthesizing enzyme, ChAT, and a degradation enzyme, AChE, the 1960s studies first aimed at the identification of brainstem cholinergic neurons attempted to do so by localizing cells containing the enzyme AChE. Later it was discovered that AChE is also synthesized by non-cholinergic neurons containing cholinergic receptors. It turns out that the presence of the ACh synthesizing enzyme ChAT is a more specific marker for the identification of cholinergic neurons. In the 1980s, for the first time, the cholinergic neurons in the brainstem were identified definitively with immunohistochemical staining of ChAT [2].

There are two categories of cholinergic neurons in the 10.1007/978-3-540-29678-2_2. Neurons in the first category have their cell bodies in the brainstem whereas their axons terminate in the periphery. This category of cholinergic cells is located in the hypoglossal nucleus, the nucleus ambiguus, the dorsal motor nucleus of the vagus nerve, the facial nucleus, the salivatory and lacrimatory complexes, the motor nucleus of the trigeminal nerve, the trochlear nucleus, the oculomotor complex, and the Edinger-Westphal nucleus. Another category of brainstem cholinergic neurons is completely contained within the central nervous system. This category of brainstem cholinergic cells is mainly located in the pontomesencephalic junction in two aggregates [2]. One aggregate of these cholinergic neurons is located in the pedunculopontine tegmentum (PPT, also called Ch5 sector) and another aggregate of neurons is located in the laterodorsal tegmentum (LDT, also called Ch6 sector). Cholinergic neurons within the PPT and LDT are medium to large in size (diameter ranging from 15 to 25 μm) [3]. These cholinergic neurons are variously shaped but are mainly fusiform, polygonal, and oval. Cholinergic neurons in the PPT are slightly more irregularly shaped than those in the LDT. The cytoplasm of these cholinergic neurons is highly developed and remarkably rich in organelles, including free ribosomes, rough and smooth endoplasmic reticulum, Golgi apparatus, and mitochondria [3]. The nucleus of these neurons typically contains large, dense, eccentric nucleoli with invaginated nuclear membranes.

Physiological characteristics of brainstem cholinergic cells have been studied by recording extracellular-single-cell-unit (single-unit) activity of PPT and LDT neurons [4]. Single-unit activity recording is a technique that records extracellularly the occurrence and duration of action potentials of a neuron. The frequency of extracellularly recorded action potentials (number of action potentials per second) is expressed as firing rate (also called discharge rate). Increased firing rate of a neuron means increased physiological activity of that neuron and similarly decreased firing rate of a neuron means decreased physiological activity. Based on the patterns of sleep-wake-state-dependent single-unit activity rates, five different types of cholinergic neurons have been identified in the PPT and LDT of cats and rats [4,5]. The firing rate of the first type of neurons, called 10.1007/978-3-540-29678-2_18, increases as the animal transitions from wakefulness to non-REM (NREM) sleep (10.1007/978-3-540-29678-2_14) and then to 10.1007/978-3-540-29678-2_18. The firing rate of the second group of neurons, called 10.1007/978-3-540-29678-2_23, increases during both wakefulness and REM sleep. The third group of neurons, 10.1007/978-3-540-29678-2_23, begins to discharge seconds before sleep is terminated and then maintains a high discharge rate until the end of the waking period. This type of cell remains silent throughout the sleep period. The fourth group of neurons, called 10.1007/978-3-540-29678-2_18, stops firing during REM sleep. The firing rates of the neurons in the fifth group, called state-independent cells, do not change as a function of sleep-wake behavioral states. The majority of the PPT and LDT cells are of REM-on and Wake-REM-on types. The durations of action potentials of these cholinergic cells are between 0.8 and 2.0 ms [5]. Another important physiological characteristic of a neuron is the mode of firing. Normally, a neuron fires in two different modes: tonic and bursting. In the tonic mode the neuron discharges as a single spike and in bursting mode the neuron discharges as a cluster of three to five action potentials within a very short period of time. In behaving rats, PPT and LDT cholinergic cells fire only in a tonic mode. In the cat, PPT and LDT cholinergic cells fire mostly in a tonic mode, but occasionally during REM sleep, about 5% of those cells also fire in a bursting mode.

There are three different ways to identify cerebral sites that receive ACh from the brainstem cholinergic cells: (i) anatomically, by localizing the sites that receive axonal terminals of brainstem cholinergic cells, (ii) neurochemically, by measuring brainstem cholinergic cell activation-induced ACh release in different parts of the brain, and (iii) by combining electrophysiological and pharmacological techniques to identify brainstem cholinergic cell activation-induced postsynaptic cholinergic effects [6]. Brain areas that receive axonal terminals of the brainstem cholinergic cells have been identified by visualizing the movements of different tracers from the PPT and LDT cholinergic cells to other parts of the brain and vise versa. A limited number of studies have also measured brainstem cholinergic cell activation-induced ACh release in some areas of the brainstem and forebrain [6]. There are some studies that have used a combination of electrophysiological recordings and pharmacological techniques to identify the postsynaptic targets of brainstem cholinergic cells and the postsynaptic effects of ACh released from those brainstem cholinergic cells [7]. Collectively, those anatomical pathway tracing, ACh release, and electrophysiological-pharmacological studies have provided evidences that the PPT and LDT cholinergic cells are the main supplier of ACh for the entire brainstem and also the thalamus [6]. Those studies also provided evidence that the PPT and LDT cholinergic cells are part of the Ach source for several nuclei in the hypothalamus, basal forebrain, and limbic areas. Although the targets for Ach released from the cholinergic cells of the PPT and LDT are mostly common, some parts of the brain receive ACh from only the PPT or the LDT. For example, in the medial prefrontal cortex, the brainstem source of ACh is the LDT and in the suprachiasmatic nucleus the source of brainstem ACh is the PPT. Interestingly, the PPT and LDT cholinergic cells are not the direct source of ACh for the cerebral cortex. In summary, the brainstem cholinergic cells act as a source of ACh for many different parts of the brain but the brain areas and proportion of total ACh are area specific.

There is some conclusive evidence to suggest that the cholinergic cells in the PPT and LDT are directly involved in the regulation of REM sleep [6]. REM sleep is characterized by a constellation of events including the following: (i) activated cortical EEG; (ii) marked atonia of the postural muscles; (iii) rapid eye movements; (iv) a theta rhythm within the hippocampus; (v) spiky field potentials in the pons [10.1007/978-3-540-29678-2_16], lateral geniculate nucleus, and occipital cortex (also called 10.1007/978-3-540-29678-2_16); (vi) myoclonic twitches, most apparent in the facial and distal limb musculature; and (vii) pronounced cardio-respiratory and core body temperature fluctuations. Distinct cell groups located in the brainstem generate individual events of REM sleep. They are discrete components of a widely distributed network rather than a single REM sleep “center” [6]. For example, muscle atonia is executed by the activation of neurons in the 10.1007/978-3-540-29678-2_12 alpha, rapid eye movements result from the activation of neurons in the peri-abducens reticular formation, PGO waves emerge by the activation of neurons in the caudo-lateral peribrachial area of predator mammals and in the dorsal part of the nucleus subcoeruleus of prey mammals, hippocampal theta rhythm is produced via the activation of neurons in the pontis oralis, muscle twitches appear with the activation of neurons in the nucleus gigantocellularis (especially the caudal part), and increased brain temperature and cardio-respiratory fluctuations occur via the activation of neurons in the parabrachial nucleus.

The desynchronized cortical EEG signature of REM sleep, however, is executed jointly by the activation of neurons in the mesencephalic reticular formation and rostrally projecting bulbar reticular formation [6]. Each of these particular cell groups is simply a set of executive neurons for an individual sign. For the final expression of an individual sign, the relevant executive neurons employ a specific neuronal circuit unique to that REM sleep sign. In essence, each of these REM sleep signs has a separate, specialized network. Thus, each of these REM sleep signs could be modulated with multiple neurotransmitters at multiple sites within their circuits. Turn-on or turn-off conditions of REM sleep generating executive neurons are regulated by the ratios of available aminergic and cholinergic neurotransmitters within those cell groups [6]. The source of aminergic neurotransmitters is the locus coeruleus (LC) and raphe nuclei (RN), while cholinergic neurotransmitters, as stated above, originate from the LDT and PPT [6]. The activity of both aminergic and cholinergic cells is approximately equal during wakefulness and the onset of NREM sleep results in an equal reduction in activity. Therefore, during wakefulness and NREM sleep the ratio of aminergic to cholinergic neurotransmitters in REM sleep generators is proportionate. During REM sleep, however, aminergic cell activities are markedly reduced or absent and cholinergic cell activities are comparatively high. Thus, when a hypothetical ratio of aminergic and cholinergic neurotransmitters is low (1:1 ratio), the REM sleep sign-generators remain in the turned-off condition; however, when this ratio is high (0:0.65 ratio), the generators are turned-on to express REM sleep signs [5]. This is how the activity of PPT and LDT cholinergic cells regulate REM sleep. To initiate and maintain the activation of REM sleep sign generators, brainstem cholinergic cells in the PPT are activated by the stimulation of kainate-type glutamate receptors [6]. Activation of GABA-B receptors inhibits PPT cholinergic cell activity to terminate REM sleep [6].

Single-cell recordings, chemical stimulation, and anatomical pathway tracing studies also suggest that the activation of PPT cholinergic cells could promote wakefulness [6]. Single-cell recording studies examining sleep-wake state-dependent firing patterns in the PPT and LDT of behaving cats and rats identified the five aforementioned major groups of cholinergic cells: (i) REM-on, (ii) Wake-REM-on, (iii) Wake-on, (iv) REM-off, and (v) sleep-wake state-unrelated. Of those five categories of cells, Wake-REM-on cells are active during both wakefulness and REM sleep [5,6]. In contrast, the Wake-on cells were found to be active only during wakefulness. The presence of Wake-REM-on and Wake-on types of cells suggests that the PPT and LDT are involved in promoting wakefulness. A recent single cell recording study in freely moving rats has shown that the Wake-REM-on cell population within the cholinergic cell compartment of the PPT increases neuronal activity as a prelude to wakefulness and remains very active until 5–8 s before the end of wakefulness [5]. Consistent with this analysis, Ach release in the thalamus is highest during waking, slightly less during REM sleep, and at a minimum during SWS. The causal evidence that the PPT is involved in generating wakefulness came from local chemical stimulation studies [6]. Chemical stimulation studies have demonstrated that the activation of NMDA-type 10.1007/978-3-540-29678-2_7 receptors within the PPT cholinergic cell compartment induces wakefulness. PPT and LDT cholinergic cells project directly to multiple wake-promoting areas of the forebrain via thalamo-cortical, hypothalamo-cortical, basalo-cortical pathways [6]. Activation of the PPT NMDA receptors promotes wakefulness by activating multiple wake-promoting areas of the forebrain. On the contrary, inhibition of PPT cholinergic cells reduces the total amount of wakefulness [6].

The projections from the PPT and LDT to the thalamic nuclei constitute a major component of the ascending reticular activating system [4,7]. Increased activity of these cholinergic cells increases the levels of ACh release in the thalamus. ACh is an excitatory neurotransmitter for the thalamo-cortical cells. Since those PPT and LDT cells are active during both wakefulness and REM sleep, ACh release from these brainstem cholinergic cells may play an important role in cortical activation during both waking and REM sleep. Indeed, the NMDA receptor-mediated activation of PPT cholinergic cells increases cortical activation [4,6]. Cortical activation is a prerequisite for cognitive functions, and thus, the activation of PPT cholinergic cells may contribute positively to cognitive processing. In summary, the normal functioning of the brainstem cholinergic cells in the PPT and LDT is involved in the regulation of REM sleep, wakefulness, and cortical activation processes.

Although there is no known direct causal relationship between the patho-physiology of brainstem cholinergic cells and any disease condition, a number of degenerative neurological diseases involve brainstem cholinergic cells [8,9]. For example, both amyotrophic lateral sclerosis and Mobius syndrome involve degeneration of brainstem motoneurons. Parkinson’s disease (PD) and progressive supranuclear palsy (PSP), which are both characterized pathologically by the loss of PPT neurons, share varying degrees of insomnia and motor dysfunction as clinical manifestations. More recently, using a transgenic animal model, it has been shown that cholinergic cell loss in the PPT may be a causal factor for a number of symptoms in Alzheimer’s disease, especially the disturbances in REM sleep [6]. It is also suspected that developmental abnormalities of PPT cholinergic cells during prenatal and early postnatal periods could be a significant contributing factor for the development of endogenous depression and schizophrenia in adolescence and adulthood [10].

A major advance towards the identification of target genes of specific transcription factors was the development of ChIP [1]. ChIP involves a biochemical rather than genetic approach to the identification of direct target genes. Unlike cDNA microarray analysis, ChIP isolates direct targets from relevant tissues under physiologic conditions [2]. Gould et al. [1] used an antibody to the Ubx homeobox gene product to immunopurify Ubx bound to DNA sequences from soluble chromatin. After immunoprecipitation (IP), the complexes were disrupted and the released genomic DNA (gDNA) fragments were cloned. The DNA fragments isolated by ChIP were enriched for sequences matching defined 10.1007/978-3-540-29678-2_8 binding sites and used to screen genomic DNA libraries for transcription units. Modifications of the ChIP protocol include stabilization of DNA-protein adducts in intact nuclei by cross-linking using ultraviolet (UV) light, formaldehyde (cross-links protein to DNA and protein to protein) and cisplatin, a chemotherapy agent that cross-links protein to DNA [3]. Utilizing these methods, homeodomain targets have been isolated and these targets include: homeobox genes, other transcription factors, growth factors, adhesion molecules, and secreted proteins. Until recently, the search for homeobox targets in vertebrates has not been as productive as in Drosophila melanogaster. Although several candidate genes have been identified using in vitro and tissue culture methods, the significance of these interactions in vivo remains to be confirmed. Takahashi’s group also modified the ChIP approach to identify mgl-1, a 10.1007/978-3-540-29678-2_20 and putative adhesion molecule, as a downstream target of HOX-C8 from mouse spinal cord [4], showing that it is possible to isolate other direct vertebrate homeobox gene targets.

The most widely applied ChIP method is adapted [4,5] and modified by our group [6] (refer to Fig. 1).Chromatin Immunoprecipitation and the Study of Gene Regulation in the Nervous System. Figure 1

Schematic outline of the chromatin immunoprecipitation (ChIP) procedure.

Cisplatin or cis-diamminedichloroplatinum (II) will not cross-link protein to protein and may be a more selective cross-linking reagent than formaldehyde compounds [6]. The cell lysate is added to hydroxylapatite resin. RNA and proteins not cross-linked to DNA are washed away. The cross-linking is reversed by incubating in thiourea, releasing proteins from the hydroxylapatite, while the DNA will remain bound. ChIP using cisplatin cross-linking of protein to DNA may isolate different gene targets than by paraformaldehyde cross-linking, such as those associated with the 10.1007/978-3-540-29678-2_14. It is reported that transcriptionally active genes are nuclear matrix associated.

The utilization of biochemical approaches such as ChIP provides several advantages. Identified target genes are directly downstream and are derived from physiological transcription factor-DNA complexes obtained in vivo. The isolated gDNA fragments may be from transcriptional regulatory elements of previously identified or novel genes. ChIP may be applied to diverse species, including Drosophila melanogaster and vertebrates. The advantage of cross-linking is preservation of a naturally existing (in situ) protein-DNA interaction. Identification of a transcription factor consensus binding sequence using ChIP-chip arrays may be more efficient than by screening random oligonucleotide pools.

Chromatin immunoprecipitation does have several limitations. The choice of the cross-linking reagent may influence whether targets are indirectly or directly downstream. IP screens require specific antibodies and require the construction of separate DNA libraries for each protein for which targets are sought and these libraries may be hampered by low cloning efficiency. It might be difficult to identify targets that interact with the regulatory protein in only a few cells or during brief developmental periods. It may be necessary to perform ChIP at several developmental time points to obtain different functional classes of transcription factor targets. Another problem may be that there is promiscuous binding. One way to reduce the non-specific DNA obtained from the ChIP procedure is to subtract the ChIP-DNA with input DNA before sub-cloning. In addition, binding may be significantly distant from the coding region As well, since most homeodomain proteins, for example, bind to a consensus TAAT core motif, many of the immunopurified fragments may not be specifically regulated by the homeobox gene itself. Finally, multiple factors may be required for the regulated expression of the target gene.

The morphological reaction of chromatolysis has been extensively studied for more than a century in experimental animal models. It was Nissl in 1894 and Marinesco in 1898 who first described the reaction using light microscopy. The classical chromatolytic appearance of the neuronal cell body can easily be recognized using light microscopy and cresyl violet or toluidin stained tissues. It includes disintegration of the basophilic Nissl bodies to a dust like appearance, peripheral condensation of basophilic substances, eccentricity of the nucleus, a basophilic nuclear cap and crenation (folding) of the nucleolemma. Often the cell body is surrounded by activated small basophilic sattelite glial cells (sattelitosis). Swelling of the cell body is frequently reported as part of the chromatolytic reaction in early neurocytological studies. In a series of studies from our own laboratory using modern stereological methods we have shown an initial cellular shrinkage amounting to approximately 30% following nerve crush and nerve transection [3,4]. In surviving cell bodies shrinkage was followed by a return to normal cellular dimensions after 3–5 months [4]. The initial morphological changes appear during the first 2–5 days following nerve damage [2] and progress during the following weeks followed by a gradual recovery with return to normal cell morphology among surviving cells. Four days after axotomy no significant cell loss can be detected. After 15 days the cell loss amounts to 31% without further progression during the following weeks.

Chromatolysis is observed in neuronal cells in the peripheral and central nervous system. Furthermore, the reaction is not confined to the neurons but also involves the surrounding glial cells. Some authors refer to this as an activation of the sattelitic glial cells and in the light microscope it is recognizable as sattelitosis, the neuron being surrounded by basophilic glial cells. The role of the activated glial cells is controversal, but it is hypothesized that they play a key role in supplying the neuron with neurotrophic growth factors which it is denied because of damage to the peripheral axon.

Sattelitosis should not be mistaken as an inflammatoric response. There is no immune reaction surrounding the neurons and when cell death occurs it is by apoptosis and not by necrosis.

Chromatolysis leads to intracellular reorganisation of the cytoplasma and its organelles. Electron microscopic studies have only provided sparse information as to the actual changes of the organelles, and the details are beyond the scope of this essay. The nissl bodies, smooth endoplasmatic reticulum, lysosomes, cytosheleton, nucleus and nucleolus are all influenced by the shift to a state of regeneration with increased synthesis of 10.1007/978-3-540-29678-2_8. The characteristically morphological changes occurring in the cell (eccentricity of the nucleus and peripheral displacement of basophilic substances) has traditionally been hypothesised to be caused by osmotic swelling. Later studies, however, revealed an abundance of nissl body-free cytoskeletal components stockpiled in the cytoplasma of axotomised neuronal cells [5] believed to cause the displacement of the organelles, including that of the nucleus.

Chromatolysis is a temporary condition of regeneration in response to a harmful stimulus rather than a step in a chain of inevitable events leading to cell death. Axotomy, traumatic, pathological as well as toxicologic conditions can result in a condition leading to chromatolysis. With regard to axotomy, which by definition leads to loss of the axon terminal, the chromatolytic regenerative state is supposed to be caused by massive intracellular reorganization caused by the need to initiate a growth program for the formation of the axonal growth cone to replace the amputated axonal terminal [5,6]. Dependent on the harmful stimulus a fraction of the cells will die, by apoptosis, and another fraction will regenerate and resume normal morphology and function when possible. Neurons in the central nervous system seem very vulnerable to traumatic damage whereas neurons in the peripheral nervous system seem rather resilient to traumatic damage and in most cases regenerate dependent of the severity of the trauma. In a series of experiments performed in our laboratory, permanent central nerve damage of spinal nerves in rats lead to a reduction of L5 dorsal root ganglion (DRG) neuronal cells of 26% after 3 months whereas 46% of the cells were chromatolytic. In diabetic rats there was no increase in cell loss and 46% of the remaining neurons were also chromatolytic [7]. This indicates that in spite of a severe trauma of the spinal nerve, 1 cm distal to the DRG only one fourth of the neurons succumb to apoptosis, and roughly half of the remaining cells are in a state of chromatolysis with the potential to survive.

The amount of loss of neuronal DRG cells is dependent on the distance to the DRG. In sciatic nerve axotomy in rats the DRG cell loss is smaller than after spinal nerve axotomy and occurs later [8].

Structural damage as well as other pathological conditions can cause a neuron to enter a state of chromatolysis. The signal for this transformation has not yet been identified but strong evidence points toward loss or reduction of neurotrophic factors derived from the periphery [1]. 10.1007/978-3-540-29678-2_14 has been suggested to play a role because experiments have shown that axotomy-induced chromatolysis of sympathic ganglia cell bodies in vitro is blocked by application of NGF and that application of NGF antisera leads to axotomy like changes [1]. Other neurotrophic factors might play a role and up and down regulation of neurotrophic receptors in pathological conditions might well be involved.

The signal leading to chromatolysis has been debated for decades, and in comprehensive reviews during the 70’s [2,9] not less than ten possible candidates were presented. Strong evidence, however, suggests an important role of axonally retrogradely transported signaling proteins. In reviews, [1,6] the lack of retrogradely transported peripheral trophic factors important for the continuous upheaval of the cell body integrity as well as the presence of peripheral electrical stimuli via calcium influx en masse and retrogradely transported chromatolysis inducing proteins are discussed. It seems likely, however, that the signal for chromatolysis is in part the lack of retrogradely transported peripheral trophic factors usually achieved from the axon terminal as well as traumatic initiated electrical or biochemical stimuli of the cell body. In contrast to the traditionally held belief that all neuronal proteins are produced centrally and anterogradely transported to the axon and the axon terminal, new research indicates that protein synthesis capacity is present in the axon [6,9], and that damage to the axon might trigger translation of mRNAs. Peripherally produced peptides possibly play an important role in initiating the chromatolytic reaction in response to axonal damage by retrogradely transporting to the cell body, including the formation of a growth cone for regeneration of the axon terminal [6,10], the latter only being an ability characteristic of the peripheral nervous (Fig. 1). Chromatolysis. Figure 1

Non-chromatolytic dorsal root ganglion cells with preserved nissl substance and a centrally placed nucleus (a). Chromatolytic dorsal root ganglion cells showing disintegration of nissl substance, displacement of nucleus and sattelitosis of glial cells around the neuronal cell body (b).

Chromatolysis is considered to be a state of regeneration in damaged neuronal cells characterized by increased synthesis of cytoskeletal and other housekeeping proteins with down regulation of neurotransmitter-related enzymes and receptors. The state is a shift from external functioning to internal build up and the surface of the cell is covered by glial profiles from activated sattelite glial cells leading to a temporary loss of most presynaptic dendritic terminals.

The bulk of experimental work regarding chromatolysis has been performed in animal models inflicting physical (axotomy or chrush [3,4]), thermal or chemical (acyrylamide intoxication) damage to the peripheral (DRG or sympathetic ganglions) or central nervous system (visual cortex and retinal ganglia or olfactory bulb). Chromatolysis is, however, observed is part of many disease processes including multiple schlerosis, porphyria, pellagra, amyotrophic lateral sclerosis and poliomyelitis.

Chromatolysis is a regenerative cytologic response to harmful physical or metabolic exposure. In accordance with this statement there is no information about conditions in which chromatolysis is the primary pathology. In traumatic injury it seems attractive to support regrowth of damaged axonal or dendritic processes by supplying patients with neurotrophic growth factors, or by manipulation of axonal protein synthesis, growth cone formation and propagation. The knowledge of this field is still limited and on a strictly experimental basis. In the future, however, stimulation and manipulation of chromatolysis and regeneration of neuronal cells might prove to be a new approach in posttraumatic neurology.

10.1007/978-3-540-29678-2_14

By far the most studied biological rhythm is the circadian rhythm (Latin: circa – about – and dies – a day). It is an excellent example of how circa-rhythms affect living systems at all levels of biology – from gene expression, hormone secretion and physiology to complex behavior. The circadian clock represents internal time-of-day and ensures that the appropriate biological functions in cells, tissues and organs occur at the right time in relationship to other endogenous processes and to the external environment.

The signature of circadian rhythms is their persistence in constant conditions, (shielded from all zeitgebers) revealing their 10.1007/978-3-540-29678-2_6 period. Examples of circadian rhythms are the sleep-wake behavior in humans and other animals, leaf movement in plants, fungal spore formation, and virtually all of gene expression in cyanobacteria, to name only a few. Circadian rhythms are ubiquitous, i.e., they have been identified in organisms of all phyla, and, in each organism, they modulate all aspects of biology [1].

In spite of being built by different cellular and molecular components in different organisms (see below), circadian rhythms share basic properties. They are (i) rhythmic and (ii) self-sustained (i.e., non-dampened), (iii) with a circa 24-h period in constant conditions; (iv) circadian rhythms are both robust in their amplitude (sufficient to drive output rhythms) and precise in their period (though not exact, circadian rhythms have been shown to continue for years with deviations of only minutes [2]); (v) their period is compensated against spurious environmental changes (e.g., of temperature or nutrients); (vi) circadian rhythms can be synchronized by appropriate environmental signals (zeitgebers). This synchronization is a complex, active process called entrainment. Under natural conditions, circadian clocks are perfectly entrained to the 24-h rotation of the Earth by using the environmental changes that have shaped circadian clocks through evolution (predominantly light, but poikilotherms also use temperature) as zeitgeber signals.

Similar to the day, the year also shows distinct characteristics in its temporal structure. With growing distance from the equator towards the poles, seasonal changes in day length become increasingly obvious (even at the equator, seasonal progression is apparent, for example, by different amounts of rain). Two different chronobiological strategies allow organisms to adapt their physiology and behavior to the progression of seasons: the circannual clock and photoperiodism.

Similar to the circadian clock, a circannual clock represents internal time-of-year and ensures that the appropriate biological functions in cells, tissues and organs occur at the right time in relationship to both other endogenous functions and to external time-of-year. Similar to the case of the circadian clock, alterations in light and dark are the predominant zeitgeber that entrain circannual rhythms – in this case, alterations in changing photoperiod. Circannual clocks can run free when photoperiod is kept constant, slightly deviating from 365.25 days (some circannual clocks only show a free-run in a specific constant photoperiod of, for example, 10 h light and 14 h darkness, LD10:14 [3]).

While the entrained circannual program ensures continuous adjustment of immunological, metabolic and behavioral processes to seasonal environmental changes, photoperiodism opens a once-a-year window, which is called the critical photoperiod, triggering a (photoperiodic) response. In most plants and animals, this response is related to reproduction. The mechanisms that detect this critical photoperiod involve the circadian system as an internal reference (abnormal photoperiodic timing is typical for Circadian Clock mutants [4]). The sensitivity to certain critical photoperiods requires a previous sensitization by, for example, short days. Hamsters provide an impressive example for a photoperiodic response, as they rapidly enlarge testes and become reproductive following exposure to days with photoperiods over 12 h [5].

Compared to the circadian program, we know far less about the anatomical structures, genes and molecular mechanisms which form the basis of both circannual rhythmicity and photoperiodism.

Genetics has been broadly applied to describe the circadian clock mechanism, an approach pioneered in the lab of Seymour Benzer. Mutant screens have revealed a complex network of so-called clock genes – genes that, when mutated or deleted, change at least one of the six fundamental properties of circadian rhythms (see above). Clock genes, involved in generating the circadian rhythmicity at the cellular level, form a transcriptional-translational negative feedback loop. Activators control the production of gene transcripts leading to proteins which undergo a progressive modification (mainly phosphorylation) and eventually feed back to inhibit their own transcription. In the cyanobacterial system, circadian oscillations have been definitively shown to depend on rhythmic phosphorylation and de-phosphorylation of a set of proteins, a process which even persists in a test tube [6]. It is not clear how this may relate to eukaryotic molecular clocks.

Clock genes have been identified in model genetic organisms from all phyla. Interestingly, animals, plants, fungi and bacteria all feature distinct gene sets, which nonetheless function similarly on the molecular level. This suggests that these are species-specific adaptations to their environment, rather than evidence of a primordial, common clock.

One of the easiest ways to understand chronobiology is to recall common human daily behaviors. For example, the human 10.1007/978-3-540-29678-2_19 occurs once per 24 h when entrained but runs free (with a circa 24-h period) when shielded from zeitgebers. There is, however, a tremendous difference in when sleep occurs in different individuals. The temporal differences in these so-called chronotypes are not restricted to sleep, but extend to other clock-controlled processes, such as melatonin or cortisol production. Within a population, the frequencies of different chronotypes show an almost normal distribution, reflecting that chronotype is a multi-genic, highly complex trait. Chronotypes result from individual differences in entrainment characteristics due to a variety of reasons: within the population there are polymorphisms or mutations in clock genes [7]; the late-to-bed, late-to-wake teenager is well known to all of us, and it reflects a systematic effect of development on the circadian clock [8]. From childhood to adolescence, the clock entrains progressively later, a trend that reverses – at the population level – at around the age of 20; there are gender differences in chronotype also, at least until the age of menopause, with females typically being earlier chronotypes than males; exposure to strong or weak zeitgebers also determines when the clock is entrained within the daily cycle. People who work outside in broad daylight are generally earlier chronotypes than office workers [8]. Thus, genes, environment, age and gender all contribute to chronotype.

The implications of chronotype are manifold. If chronotype, for example, is not incorporated into medical practice, results of tests or the efficacy of treatments may differ merely due to the patient’s chronotype. Chronotype is also a quality of life issue. The more discrepancy between internal and external time (e.g., between an individual’s circadian timing and his or her work hours), the more sleep debt accumulates during the work-week, culminating in a chronic “social jetlag.” The larger this social jetlag, the more likely an individual is to be a smoker, indicating that a chronic jetlag acts as a stressor [8].

Because chronobiology has an impact on broad aspects of an organism’s biology, it represents a scientific specialty similar in scope to development or reproduction. Circadian rhythms are a fundamental property of all organisms (with few exceptions). The concept of selective advantage due to increased fitness is inherent to evolutionary theory. The adaptive advantage of biological clocks lies in the benefit of being able to anticipate environmental changes. The activity of animals is frequently restricted to certain times of day, and straying outside of these domains can increase the risk of predation, for instance [9]. Sessile organisms such as plants and fungi also benefit from prediction of temperature, nutrient, or humidity changes. The fitness concept was recently substantiated in vitro using cyanobacteria, showing that a circadian oscillation with a period similar to the environmental one is more successful [10]. Similar adaptive advantages hold for all circa clocks, whereby organisms prepare for seasons, tides or nocturnal light levels.

Traditionally regarded as part of the limbic system, the cingulate cortex is a large and heterogeneous part of the cerebral cortex that can be partitioned based on architecture, connectivity and functional properties [1,2]. First, cingulate cortex is divided into a posterior part (Brodmann’s area 23) and an anterior part. The anterior cingulate cortex can be divided into a ventral zone (occupying the surface of the cingulate gyrus, containing Brodmann’s areas 24a, 24b and the subcallosal area 25) and a dorsal zone (mainly in the cingulate sulcus, containing Brodmann’s areas 24c and 32). In humans, this functional area often extends into the surrounding paracingulate gyrus.

A putative cingulate eye field has been described in the caudal portion of anterior cingulate cortex (Fig. 1), and visual and saccade-related activity has been described in a portion of posterior cingulate cortex (not shown). Anterior cingulate cortex can contribute indirectly to ocular motor function through dense, reciprocal connectivity with the supplementary eye field and a weaker linkage with the frontal eye field, superior colliculus and ocular motor thalamic nuclei [3].Cingulate Cortex – Role in Eye Movements. Figure 1

Medial view of macaque monkey (top) and human (bottom) brains showing estimated location of zone in cingulated cortex related to movements of the eyes. Abbreviations: CG, cingulated gyrus; cing s, cingulate sulcus; PCG, paracingulate gyrus.

Neurons in posterior cingulate cortex of macaque monkeys discharge in response to visual stimuli and following saccadic eye movements [4], and functional imaging in humans has described activation in posterior cingulate cortex associated with visually guided saccades [5]. However, more evidence indicates a role in gaze control for a caudal zone in anterior cingulate cortex. Saccadic eye movements can be evoked by electrical microstimulation of a region in the upper bank of the cingulate sulcus directly ventral to the SEF, in area 24c [6]. Functional brain imaging studies have reported activation in anterior cingulate cortex during production of self-generated saccades guided by arbitrary cues [7].

In macaque monkeys performing a task that requires inhibition of a partially prepared movement in response to an imperative stop signal, neurons in anterior cingulate cortex were modulated following errors or when reinforcement was earned but not delivered (Fig. 2) [8]. This signal from single neurons corresponds to a scalp potential referred to as the error-related negativity, which may originate from a single dipole in anterior cingulate cortex. In addition, a diversity of neurons in anterior cingulate cortex signal when reinforcement is earned and received, earned and not received, or delivered but not earned. The activation of these neurons can guide adjustments of performance, probably derived from signals arriving from brainstem dopamine neurons, the ventral striatum or orbital frontal cortex. These results are consistent with a body of research indicating that anterior cingulate cortex monitors performance for executive control.Cingulate Cortex – Role in Eye Movements. Figure 2

Monitoring signals in anterior cingulate cortex. Activity of a single neuron is shown aligned on the time of a saccade (top) or time of reinforcement (bottom) on trials when the saccade was correct and earned reinforcement (black), when the saccade was an error and received no reinforcement (red), and when the saccade was correct but reinforcement was not delivered (blue). This representative neuron signaled the unexpected absence of reinforcement.

Three general perspectives have framed hypotheses about the function of cingulate cortex: motor control, performance monitoring and motivation. Cingulate cortex seems to contribute indirectly to gaze control through mediating the influence of motivation derived from the consequences of previous actions.

Damage to cingulate cortex in humans results in diverse disorders. Lesions focused in a limited part of anterior cingulate cortex result in impairments in producing memory-guided saccades and antisaccades [9]. These deficits involved impaired suppression of reflexive saccades as well as increased latency of visually guided saccades.

The ACC is involved in assessment of valenced information and the long-term storage of emotional objects and events and contributes to tonic mood states (Fig. 2). It appears to separately store happy events rostrally in area 32 and sad events caudally in areas 24 and 25. The ACC receives a massive input from the amygdala and a link between emotion and 10.1007/978-3-540-29678-2_1 occurs in area 25 in the subgenual ACC which projects to autonomic regulatory sites including the hypothalamus, periaqueductal gray and parabrachial nucleus. The highest concentration of cingulate glutamate receptors is in the ACC. There are proportionately fewer γ-aminobutyric acid (GABA) receptors, which are higher in the PCC [2].

The MCC coordinates decision-making about behavioral outcomes in a number of ways, i.e., anticipation of outcomes, comparing actual with expected outcomes and modifying behaviors as rewards are reduced [3]. In view of its prominent role in both rewarded and aversive (i.e., painful) outcomes, it is not surprising that an important feature of this region is the two cingulate premotor areas located along the cingulate sulcus that regulate skeletomotor outputs (Fig. 2). Both cingulate motor areas have extensive motor system projections that include the frontal cortical motor areas and subcortical structures. Important subcortical projection targets include the putamen, red nucleus, pontine nuclei, facial motor nucleus and spinal cord [4].