Logo of prosciprotein sciencecshl presssubscriptionsetoc alertsthe protein societyjournal home
Protein Sci. 2008 Mar; 17(3): 518–526.
PMCID: PMC2248322
PMID: 18287284

Intersubunit linker length as a modifier of protein stability: Crystal structures and thermostability of mutant TRAP

Masahiro Watanabe,1 Yumiko Mishima,2,3 Ichiro Yamashita,2,3,4 Sam-Yong Park,1 Jeremy R.H. Tame,1 and Jonathan G. Heddle1,2,3,5
Author information Article notes Copyright and License information Disclaimer

Abstract

The ability of proteins to self-assemble into complex, functional nanoscale structures is expected to become of significant use in the manufacture of artificial nanodevices with a wide range of novel applications. The bacterial protein TRAP has potential uses as a nanoscale component as it is ring-shaped, with a central, modifiable cavity. Furthermore, it can be engineered to make a ring of 12-fold symmetry, which is advantageous for packing into two-dimensional arrays. The 12mer form of TRAP is made by linking multiple subunits together on the same polypeptide, but the usefulness of the 12mers described to date is limited by their poor stability. Here we show that, by altering the length of the peptide linker between subunits, the thermostability can be significantly improved. Since the subunit interfaces of the different 12mers are essentially identical, stabilization arises from the reduction of strain in the linkers. Such a simple method of controlling the stability of modular proteins may have wide applications, and demonstrates the lack of absolute correlation between interactions observable by crystallography and the internal energy of a complex.

Keywords: self-assembly, nanoengineering, ring protein, RNA-binding, strain

From the beginning of nanotechnology, proteins have been heralded as ideal candidates for nanodevices (Astier et al. 2005), not least because they already carry out just such a role in nature. However, designing proteins to fulfill specific functions that do not occur naturally is a major challenge in synthetic biology. Obstacles include biological complexity, the time and work required to construct and characterize synthetic systems, and spontaneous variations in the behavior of biological systems (Endy 2005). Advances have been made in the use of massively parallel computing to simulate and predict protein folding for small proteins (Shirts and Pande 2001; Zagrovic et al. 2001), but proteins of normal size remain far beyond current computational limits.

Until it becomes possible to predict tertiary structure from amino acid sequence alone, one of the most promising approaches to engineering new, artificial protein structures is a “building block” technique, where protein domains with desired structural or functional properties are chosen from existing, characterized proteins and connected together. This approach has been used to make a number of artificial proteins including a novel chimaeric DNA binding protein (Chevalier et al. 2002), a biotin-modified aldolase linked together by streptavidin to form a two-dimensional grid (Ringler and Schulz 2003), a backward-moving myosin made from three preexisting protein domains (Tsiavaliaris et al. 2004), a charge-trapping chimaeric protein made by a fusion of cytochrome b562 and green fluorescent protein (Choi et al. 2006), and a ball-and-stick supramolecule (Sugimoto et al. 2006).

Similarly, linking tandem copies of a protein on the same polypeptide chain is another potentially powerful technique that can be used to alter the stability and functional properties of a protein. The native 11mer TRAP has been engineered in this way to form an artificial 12-membered ring (Heddle et al. 2006, 2007a). TRAP is an RNA-binding protein involved in the control of tryptophan synthesis in various species of Bacillus (Gollnick 1994; Babitzke 2004). The TRAP monomer is ∼8.2 kDa in size and associates to form a ring-shaped homo-11mer that is highly thermostable (Baumann et al. 1997; Heddle et al. 2006). The ring is ∼8.4 nm in diameter and 2.5 nm in height, with the central hole being ∼2.4 nm in diameter. TRAP is able to bind cognate single-stranded RNA (ssRNA) once it has bound tryptophan. In vivo and in vitro studies show that TRAP will bind ssRNA carrying 11 copies of the sequence NAG (Babitzke et al. 1994, 1995, 1996; Antson et al. 1999) with an optimal spacing of two nucleotides between triplet repeats.

Ring-shaped proteins and peptides are of particular interest as potential components of nanodevices. They can be used to create arrays (McMillan et al. 2002) or tubes (Hartgerink et al. 1998) of metals or semiconductors. TRAP also offers the potential to arrange materials such as gold nanodots. The protein surrounding the nanodot can be utilized to organize the dots in a desired pattern such as a tightly packed array. Recently, a modified TRAP protein was used to place gold nanodots on a titanium oxide surface (Heddle et al. 2007a).

An artificial 12-membered version of the TRAP ring was produced by fusing together three or four copies of the TRAP gene (to make fusions called “TRAP3” and “TRAP4,” respectively). In these constructs the copies of the monomers are connected by three alanine residues (Heddle et al. 2006). Three molecules of TRAP4 and four molecules of TRAP3 self-assemble to form ring proteins with subunit contacts essentially identical to the wild-type protein in the crystal structures, but with 12 instead of 11 subunits in the ring (Heddle et al. 2006). However, the linker peptide was not visible in the resulting electron density maps, and it was unclear if they were located inside or outside the ring (Fig. 1). Simplistic linear measurements of the two possible paths suggest a loop length of ∼47 Å for the inside route and 66 Å for the outside route (including not just the linker peptide but residues disordered in previous TRAP crystal structures). This might suggest that the inside path would be preferable but does not take into account the possible effects of crowding eight copies of the linker within the cavity. It is also possible that a linker taking the outside path might pack more closely with the existing structures, reducing its path length. Both of these cases would significantly promote the outside path, and only by solving the crystal structure, revealing part of the linker, can the actual path be verified. Although the hydrogen bonds and apolar interactions between subunits are almost identical to those of the wild-type protein, the 12mer TRAP was found to be significantly less stable, with differential scanning calorimetry suggesting that initial peaks in heat capacity occur at more than 10°C lower than for the wild-type protein (Heddle et al. 2006). The 12-membered ring was apparently destabilized significantly by the short length of the linker peptide, which introduces mechanical strain into the system. In this paper we have investigated the position of the linker peptides in the ring structure and the effect of the linker length on the stability of the engineered 12mer ring. The crystal structures of two 12mer TRAP proteins with increased linker length are described, showing for the first time part of the linker and N-terminal region of the protein pointing toward the center of the ring. Increasing the length of the linker increases the stability of the protein, although the crystal structures of the different 12mers are essentially identical.

An external file that holds a picture, illustration, etc. Object name is 518fig1.jpg

Schematic figure comparing wild-type 11mer TRAP and artificial 12mer TRAPs showing possible positions of linker peptides. (A) A wild-type TRAP monomer. Folding and assembly (indicated by the arrow) gives the 11mer wild-type TRAP ring. The RNA-binding region around the outside of the ring is indicated by the blue stripe. (B) The artificial monomer TRAP, T3A7, consisting of three wild-type monomers that are covalently linked by seven-alanine linker peptides (shown in red). Four of these linked proteins assemble into a “12mer” ring. The linker connecting the C and N termini of the linked subunits could do so via a path around the outside of the ring (external linker) or through the inside of the ring (internal linker). In the former case, the linker would obstruct the RNA-binding site.

Results

In the original 12mer protein (T3A3), two linker sequences (L) of three alanine residues join three copies of TRAP (T) in the pattern “T-L-T-L-T.” The newly produced 12mers in this work have linkers of five or seven alanines (T3A5 and T3A7, respectively). Transmission electron microscopy (TEM) was used to confirm that the mutant proteins form rings, and all of the mutant proteins showed characteristic doughnut-like rings, indistinguishable from TEM images of the wild-type protein (see Fig. 4).

An external file that holds a picture, illustration, etc. Object name is 518fig4.jpg

Transmission electron micrographs of mutant TRAPs deposited on a carbon-coated copper mesh. (A) T3A5. (B) T3A7. We applied 0.05mg/mL of each protein in the presence of tryptophan. The white scale bar is 50 nm.

RNA-binding assays

The previously determined crystal structures of 12mer forms of TRAP gave no indication of the position of the linker peptides connecting the subunits. To see if the linker peptide runs around the outside of the protein or through the central cavity, we tested the RNA binding of wild-type 11mer and mutant 12mer forms of TRAP. If the linker peptide runs around the outside of the 12mer ring, it would block RNA binding to the outer surface of the ring (Yang et al. 1997). If the linkers run through the center of the ring, then RNA binding affinity is expected to be close to that of wild-type TRAP. Gel retardation assays (Fig. 2) showed all of the mutant TRAP proteins were able to bind RNA, with affinity close to that of wild-type protein, strongly suggesting that the linker peptides run through the center of the ring. Interestingly, while wild-type TRAP protein gave a sharp band of the protein–RNA complex, the mutant TRAPs all gave a more smeared band at higher molecular weights. This may be because the 12mer TRAP is not completely enclosed by the 55mer RNA, and so excess RNA may lead to some higher order complexes forming.

An external file that holds a picture, illustration, etc. Object name is 518fig2.jpg

RNA binding to wild-type and mutant TRAPs. (A) Mixtures of 2.5 μM single-stranded RNA with wild-type or mutant TRAPs were applied to an agarose gel that was stained with ethidium bromide for the presence of protein (top) and RNA (bottom). Protein concentrations were 0 μM, 2.5 μM, and 5 μM. All of the 12mer TRAPs show very high apparent molecular weight in the presence of RNA, suggesting the 55-base-long RNA binds more than one 12mer ring. (B) Mixtures of 100 pM single-stranded RNA and increasing amounts of wild-type (left-hand gel) or T3A7 (right-hand gel) were run on a 5% native-PAGE TBE gel. Lanes are marked with the picomolar concentration of protein used. Lane C is a control containing RNA only. 3A, 5A, and 7A denote TRAP proteins T3A3, T3A5, and T3A7, respectively.

Heat stability tests

The thermostability of different forms of TRAP was tested by incubating the protein, with or without added tryptophan, in a heating block and raising the temperature in steps of 10°C. Wild-type TRAP and the three mutant TRAPs showed no visible precipitation after a 5-min incubation at 70°C in the presence of saturating amounts of tryptophan. T3A3 however showed precipitation in absence of tryptophan (Fig. 3A). At 80°C T3A5 showed significant precipitation in the absence of tryptophan (Fig. 3B), and at 100°C both T3A3 and T3A5 showed precipitation both in the presence and absence of tryptophan (Fig. 3C). At 110°C T3A7 started to show very slight precipitation in the presence but not absence of tryptophan (Fig. 3D). Even at 110°C, wild-type TRAP remained soluble whether or not tryptophan was added. In summary, wild-type TRAP appears the most stable. Generally, TRAP stability is increased by the addition of tryptophan (with the exception of T3A7 at 110°C), and mutant TRAPs increase in thermostability as the linker length increases from three to seven alanine residues. Both T3A5 and T3A7 also retain the expected ring shape (Fig. 4).

An external file that holds a picture, illustration, etc. Object name is 518fig3.jpg

Heat stability of the mutants. Wild-type and mutant proteins (1 mg/mL) were heated to 70°C (A), 80°C (B), 100°C (C), and 110°C (D) for 5 min before cooling to room temperature. Precipitation at 110°C in the case of T3A7 was too slight to be readily visible. Denatured protein formed a white precipitate. (+) Added tryptophan, (−) no added tryptophan. 3A, 5A, and 7A denote TRAP proteins T3A3, T3A5, and T3A7, respectively.

Differential scanning calorimetry

As previously found with T3A3 (Heddle et al. 2006), the DSC results showed broad, noisy peaks indicative of complex interactions rather than a two-state process and a reduced thermostability compared to wild-type protein. However, each DSC trace contains a sharp trough that corresponds to the irreversible denaturation of the protein (see Supplemental Fig. S1). The temperature at which this trough occurs is significantly higher in T3A5 and T3A7 than in T3A3 and corresponds closely to the denaturation temperature observed in the heat stability tests (Table 1). This confirms that the mutant proteins with longer linkers are significantly more stable than the protein with the short linker, but less stable than the wild-type protein.

Table 1.

Melting temperature of different TRAP proteins

An external file that holds a picture, illustration, etc. Object name is 518tbl1.jpg

X-ray crystallography

T3A5 and T3A7 were crystallized in space groups P4212 and P4, respectively (Table 2). The crystal packing of the T3A5 rings is identical to that previously found for T3A3 with one TRAP3 unit (equivalent to three wild-type monomers) per asymmetric unit. In the case of T3A7 there are two. In both cases the new structures closely resemble the previous model in that four such units combine to form a single ring facing the crystallographic fourfold axis. These rings form continuous tubes running through the crystal, unlike the previously solved wild-type TRAP structures (Antson et al. 1994, 1999; Chen et al. 1999; Hopcroft et al. 2002). The models of T3A5 and T3A7 are essentially identical in tertiary structure to T3A3 and wild-type TRAP and subunit interface hydrogen bonds are also essentially unchanged but in the case of T3A7 additional residues of the linker peptide are visible (Figs. 5–7; Table 3). Overlaying the Cα backbones of T3A3 and T3A5 (Fig. 7) gives an RMS deviation of 0.16 Å. The one significant difference with T3A7 is that several additional linker residues are now visible, protruding into the central cavity. These regions are not ordered in T3A3 and T3A5, which more closely resemble the wild-type TRAP structures also missing the first six N-terminal and two C-terminal residues. In T3A7 the C-terminal residues form a hairpin turn between Glu73 and Gly74 and the chain forming a β-strand running up the central cavity of the ring adjacent to residues 67–71. The visible region of the linker is stabilized by a number of hydrogen bonds with surrounding residues (Supplemental Table S1). The region of the hole close to the C-terminal face of T3A7 is significantly narrowed due to the presence of the linker and at the N-terminal face due to the presence of N-terminal residues that were not visible in previous structures (Figs. 5, ​,77).

Table 2.

Data collection and refinement statistics

An external file that holds a picture, illustration, etc. Object name is 518tbl2.jpg

Table 3.

Comparison of subunit interface hydrogen bonds in T3A5, T3A7, and wild-type B. stearothermophilus TRAP

An external file that holds a picture, illustration, etc. Object name is 518tbl3.jpg
An external file that holds a picture, illustration, etc. Object name is 518fig5.jpg

Comparison of the crystal structures of T3A3, T3A5, and T3A7. Each protein ring is shown in colored cartoon form with the molecular surface of the protein shown in semi-transparent gray. Images in the top row show the rings from “above.” The middle row is after a rotation of 90° about an axis parallel to the plane of the ring, and the bottom row after an additional 180° rotation about the same axis. In T3A7, the extra residues visible in the center of the ring are Thr5 and Asn6.

An external file that holds a picture, illustration, etc. Object name is 518fig7.jpg

(A) Peptide backbone overlay of T3A3 (red), T3A5 (green), and T3A7 (blue). (B) Peptide backbone overlay of TRAP3A3 (red), TRAP3A5 (green), and TRAP3A7 (blue), showing the N-terminal region of one wild-type monomer equivalent and the C-terminal of the neighboring monomer equivalent. The residue shown as cyan sticks is residue 79, the third alanine of the T3A7 linker, which points into the central hole of TRAP. (C) Close-up view of part of the peptide backbone overlay of T3A3 (red), T3A5 (green), and T3A7 (blue) with the same orientation as in B. The dashed line connects the last visible C-terminal residue (A79, the fourth linker residue) with the first visible N-terminal residue of the next monomer (T5). The dashed line indicates the connection of these two residues via the remaining four alanine residues, which are not visible in the crystal structure.

An external file that holds a picture, illustration, etc. Object name is 518fig6.jpg

T3A7 linker residues, with a final, weighted 2Fo-Fc electron density map, calculated to 1.8 Å resolution, displayed with CCP4mg (Potterton et al. 2002). The map is contoured at 1.0 σ. The first three linker residues (Ala77–Ala79) and Lys76 as well as residues Thr5, Asn6 are not visible in the electron density maps of earlier 11mer and 12mer TRAP crystal structures. The model shown here unambiguously demonstrates that the intersubunit linkers pass through the central hole of the ring.

As expected, the hydrophobic tryptophan-binding pocket of the mutants is identical to that seen in the wild-type protein. The buried surface area between protein domains as calculated by AREAIMOL (Collaborative Computational Project, Number 4 1994) is 1700, 1780, and 1860 Å2, for T3A3, T3A5, and T3A7, respectively, compared to 1630 Å2 for the wild-type protein. The slightly larger area for T3A7 is mainly due to the presence of extra C-terminal residues in the crystal structure.

Previous examination of the contacts in the wild-type and T3A3 mutants showed that hydrogen bonds and hydrophobic interactions across the monomer interface were unchanged (Heddle et al. 2006). Similar examination of contacts between monomer interfaces in T3A5 and T3A7 using CONTACT, part of the CCP4 package (Collaborative Computational Project, Number 4 1994), again showed no significant deviations (Table 3). The crystal structures alone therefore give no immediately obvious explanation for the change in thermostability such as lost hydrogen bonds or other interactions.

Discussion

As a thermostable ring protein, TRAP has considerable potential as a component of nanoscale devices and/or as a biomineralization agent. Biomineralization using cage proteins such as ferritin and viral capsids has been shown to be possible if suitable mutations are introduced into the protein cavity (Douglas et al. 2002; Miura et al. 2006). We have shown it is possible to alter the inner cavity of the TRAP ring in the same way (Heddle et al. 2007a). To form ordered arrays on surfaces, however, 12-fold symmetry is much more suitable than 11-fold symmetry. We have previously constructed artificial 12mer TRAP rings by linking together TRAP monomers with three alanine residues (Heddle et al. 2006), but these proteins were found to have low thermostability. The crystal structures showed that the tri-alanine linker peptides in T3A3 must be tightly stretched between the N and C termini of the subunits, but offered no other obvious structural cause for the reduced stability. Here we have shown that increasing linker length does indeed significantly stabilize 12mer TRAP.

Since the N- and C-terminal residues were not visible in the electron density maps (Heddle et al. 2006), it could not be determined previously whether the linkers passed through the cavity or around the outside of the ring. The crystal structure of T3A7 shows clearly the linker peptides pass through the central cavity. This result is confirmed by RNA-binding experiments that show RNA is capable of binding to the 12mer forms of TRAP. If the linkers lie on the outside of the protein they would block access to the RNA-binding site, which consists primarily of residues Glu36, Lys37, Asp39, Lys56, and Arg58 located around the circumference of the ring (Yang et al. 1997; Hopcroft et al. 2002).

While wild-type TRAP clearly prefers the 11mer form, there has been limited experimental evidence suggesting it can also make a 12mer ring (McCammon et al. 2004). Our results show that all forms of 12mer TRAP tested can bind single-stranded RNA with an appropriate sequence (Fig. 2A), suggesting that the artificial proteins could replace 11mer TRAP in vivo. Further testing showed that 12mer TRAP has a similar affinity for RNA as the wild-type protein (Fig. 2B), consistent with the linker peptide being present in the central cavity rather than crossing the RNA-binding site. The binding results do, however, reveal one obvious difference in RNA binding between the wild-type and 12mer TRAPs: The 12mer proteins give higher molecular weight complexes that appear to precipitate. In fact, sensitive mass spectrometry studies have shown that wild-type protein also forms small amounts of higher order complexes of two or three TRAP rings bound to RNA (McCammon et al. 2004). This wild-type complex, however, is not visible in the gel-shift studies, suggesting that the 12mer complex is able to form complexes of even higher molecular weight that precipitate. This is confirmed by our own mass spectrometry studies (results not shown). This may be due to the fact that the 12mer ring has 12 RNA binding sites but the RNA used in these binding experiments has only 11 copies of the NAG binding sequence. Each 12mer therefore has one “free” binding site to which a second RNA strand could bind while simultaneously binding a second TRAP. In this case, some portion of the RNA would not be bound to the protein and therefore it would be easier to stain than protein-bound nucleotides. Consistent with this interpretation, we find that the RNA-12mer complexes could be stained with either ethidium bromide or SYBR-Gold, whereas the RNA complexed with the wild-type protein could only be stained successfully with SYBR-Gold, which has a stronger affinity for single-stranded nucleic acid.

Linking different domains by incorporating them into the same polypeptide is an obvious method for producing bifunctional proteins, whether for nanoscale protein-based devices or other purposes. Our results show that the linker region can play an important role in the properties of the resulting protein, and linker length should be considered carefully when designing such proteins. Given the essential identity between the intersubunit contacts of the TRAP subunits in the different mutant structures, these proteins are also an example of the difficulty of predicting protein stability from crystallographic models. The differences in stability of the different proteins described here largely arise from parts of the structure invisible in the electron density map, and the strain induced in the visible parts has no measurable effect on the atomic coordinates of the refined models.

Materials and Methods

Construction of genes encoding mutant proteins

The plasmid encoding modified TRAP 12mer protein consisting of a fusion of three TRAP monomer genes was previously described (Heddle et al. 2006). The three wild-type monomers in this protein are linked via a linker peptide consisting of three alanines. In this report, this protein is referred to as T3A3. The T3A3 coding sequence was further modified by the stepwise addition of the six bases, 5′-GCGGCG-3′, encoding two additional alanine residues, to both linker regions to produce a new gene encoding T3A5. A second mutant (called T3A7) was constructed by further addition of the same six-base sequence to T3A5, lengthening the linker regions by a further two alanine residues. Insertions were made using the Quikchange mutagenesis kit (Stratagene).

Protein expression and purification

Plasmids encoding the mutant proteins were transformed into BL21 DE3 cells (Stratagene), which were grown in liquid LB medium at 37°C and induced with 500 μM IPTG at an optical density at 695 nm of ∼0.5. Protein purification was carried out using a method similar to that described previously (Heddle et al. 2006). In brief, cell pellets of induced cells were resuspended in 50 mM Tris HCl (pH 8.5), 100 mM NaCl. Resuspended cells were lysed by sonication. Supernatant was applied to a Q-sepharose column (Amersham) equilibrated in 50 mM Tris HCl (pH 8.5) and eluted with an ascending salt gradient. Mutant TRAP proteins eluted from the column at ∼450 mM NaCl. Fractions containing TRAP protein were pooled and exchanged into 20 mM MES (pH 6.5), 100 mM NaCl, and applied to a heparin sepharose column (Amersham) equilibrated in the same buffer. Protein was eluted with an ascending salt gradient. Protein was found to elute at ∼1 M NaCl. As a final step, eluted protein was added directly to a Sephadex 200 16–60 gel filtration column (Amersham) equilibrated in 20 mM Tris HCl (pH 8.5), 100 mM NaCl. Protein collected from this final step was more than 95% pure.

Protein crystallization

Mutant TRAP proteins were crystallized using the hanging-drop vapor diffusion method using 2.5 mg/mL protein stocks in a total drop volume of 2 μL. T3A5 was crystallized in buffer containing 0.09 M CAPS (pH 10.5), 30% (w/v) PEG 300, 0.15 M ammonium sulfate, and 10 mM L-tryptophan. T3A7 was crystallized in buffer containing 0.1 M sodium citrate (pH 5.5), 30% (w/v) MPD, 0.2 M ammonium acetate, and 10 mM L-tryptophan. Data for T3A5 and T3A7 were collected at beamlines NW12A and BL17A, respectively, at the Photon Factory, Tsukuba, Japan, and processed with HKL2000 (Otwinowski and Minor 1997). T3A5 and T3A7 crystallized in space groups P4212 and P4, respectively.

Refinement

General data handling was carried out with the CCP4 package (Collaborative Computational Project, Number 4 1994), and the structures were solved by molecular replacement using Phaser (McCoy et al. 2005). Manual adjustment of the models was carried out with TURBO-FRODO (Roussel and Cambillau 1989), and refinement using CNS (Brunger et al. 1998) and REFMAC5 (Murshudov et al. 1999). Water molecules were added to the models by manual inspection of the 2Fo-Fc and Fo-Fc maps. Ramachandran plots for the T3A5 and T3A7 models showed 97.7% and 93.5% of the residues, respectively, to be in the most favored regions, with the remaining 2.3% and 6.5% of the residues in additional allowed regions. Data collection and refinement statistics are shown in Table 2.

Differential scanning calorimetry

Wild-type TRAP, T3A3, T3A5, and T3A7 in 50 mM MOPS (pH 7.8), 250 mM NaCl, 500 μM L-tryptophan were used at a final concentration of 1.0 mg/mL. A VPDSC instrument (Microcal) was used, and the scanning speed was 60°C per hour. Before each sample run at least 10 control runs in the absence of protein were carried out. DSC scans in the presence of protein were performed two or three times for each protein used. Results were analyzed using the manufacturer's software based on the Origin graphing program.

Heat stability tests

The heat stability of wild-type and mutant TRAPs was assessed by heating 1.0 mg/mL protein in 100 μL of 50 mM Tris HCl (pH 8.5), 100 mM NaCl, in the presence or absence of 0.5 mM L-tryptophan. The samples were placed in a heat block, and the temperature raised from 50°C in steps of 10°C. Incubation was continued for 5 min at each temperature, and the temperature at which visible precipitation occurred was recorded.

RNA-binding assays

RNA binding was assessed by gel-shift assays. To see if the mutant proteins were able to bind RNA, varying amounts of TRAP, in the presence of excess tryptophan, were added to 2.5 μM of the 55-base single stranded RNA consisting of 11 GAGAA repeats (Dharmacon), known to bind strongly to TRAP (Babitzke et al. 1996). Reaction mixtures were incubated at 37°C for 10 min. The presence or absence of RNA binding was assessed by running 15 μL of the reaction mixture on a 0.5% Tris-borate agarose gel run at a constant current of 20 mA for 75 min. The position of RNA and protein on the gels were visualized using SYBR-Gold stain (Invitrogen) and CBB stain, respectively, and visualized under UV and visible light. The migration of RNA with TRAP bound through the gel was significantly retarded. A more sensitive binding assay with T3A7 was to the same RNA sequence except that an extra three bases (GCC) were added to the 5′ end and the terminal 5′ base was labeled with FAM (Hokkaido System Science). TRAP-binding to RNA was measured quantatively by titrating increasing amounts of TRAP into a solution of 100 pM RNA. Reaction mixtures were incubated at 37°C for 15 min. The presence or absence of RNA binding was assessed by running 20 μL of the reaction mixture on a 5% native-PAGE_TBE gel in 1× TBE buffer. Gels were run at 150 V for 30 min and imaged with a Typhoon 9400 multiformat imager. The approximate Kd was calculated as the amount of protein required to cause 50% of the RNA to be shifted.

Transmission electron microscopy

Transmission electron microscopy was carried out using a 200-keV JEM-2200 (JEOL). TRAP proteins were used at ∼0.05 mg/mL in 50 mM Tris-HCl (pH 8.5), 100 mM NaCl buffer, 0.5 mM L-tryptophan. Protein was added to a carbon-coated copper grid and stained with 3% potassium tungsten acetate.

Accession codes

Coordinates for T3A5 and T3A7 have been deposited in the Protein Data Bank with the accession codes 2ZD0 and 2ZCZ, respectively.

Electronic supplemental material

The electronic supplemental material contains one figure and one table. The figure shows thermostability of T3A3 in the absence of tryptophans, measured via DSC. The table shows hydrogen bonds to the residues of the linker visible in TRAP3A7.

Acknowledgment

We thank the beamline staff at the Photon Factory for help with data collection.

Footnotes

Supplemental material: see www.proteinscience.org

Reprint requests to: Jeremy R.H. Tame, Protein Design Laboratory, Yokohama City University, Tsurumi, Suehiro 1-7-29 Yokohama 230-0045, Japan; e-mail: pj.ca.uc-amahokoy.imurust@ematj; fax: 81-45-08-7366; or Jonathan G. Heddle, Protein Design Laboratory, Yokohama City University, Tsurumi, Suehiro 1-7-29 Yokohama 230-0045, Japan; e-mail: pj.ca.hcetit.m@aa.j.elddeh; fax: 81-45-08-7366.

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.073059308.

References

  • Antson, A.A., Brzozowski, A.M., Dodson, E.J., Dauter, Z., Wilson, K.S., Kurecki, T., Otridge, J., Gollnick, P. 11-fold symmetry of the trp RNA-binding attenuation protein (TRAP) from Bacillus subtilis determined by X-ray analysis. J. Mol. Biol. 1994;244:1–5. [PubMed] [Google Scholar]
  • Antson, A.A., Dodson, E.J., Dodson, G., Greaves, R.B., Chen, X., Gollnick, P. Structure of the trp RNA-binding attenuation protein, TRAP, bound to RNA. Nature. 1999;401:235–242. [PubMed] [Google Scholar]
  • Astier, Y., Bayley, H., Howorka, S. Protein components for nanodevices. Curr. Opin. Chem. Biol. 2005;9:576–584. [PubMed] [Google Scholar]
  • Babitzke, P. Regulation of transcription attenuation and translation initiation by allosteric control of an RNA-binding protein: The Bacillus subtilis TRAP protein. Curr. Opin. Microbiol. 2004;7:132–139. [PubMed] [Google Scholar]
  • Babitzke, P., Stults, J.T., Shire, S.J., Yanofsky, C. TRAP, the trp RNA-binding attenuation protein of Bacillus subtilis, is a multi-subunit complex that appears to recognize G/UAG repeats in the trpEDCFBA and trpG transcripts. J. Biol. Chem. 1994;269:16597–16604. [PubMed] [Google Scholar]
  • Babitzke, P., Bear, D.G., Yanofsky, C. TRAP, the trp RNA-binding attenuation protein of Bacillus subtilis, is a toroid-shaped molecule that binds transcripts containing GAG or UAG repeats separated by two nucleotides. Proc. Natl. Acad. Sci. 1995;92:7916–7920. [PMC free article] [PubMed] [Google Scholar]
  • Babitzke, P., Yealy, J., Campanelli, D. Interaction of the trp RNA-binding attenuation protein (TRAP) of Bacillus subtilis with RNA: Effects of the number of GAG repeats, the nucleotides separating adjacent repeats, and RNA secondary structure. J. Bacteriol. 1996;178:5159–5163. [PMC free article] [PubMed] [Google Scholar]
  • Baumann, C., Xirasagar, S., Gollnick, P. The trp RNA-binding attenuation protein (TRAP) from Bacillus subtilis binds to unstacked trp leader RNA. J. Biol. Chem. 1997;272:19863–19869. [PubMed] [Google Scholar]
  • Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., et al. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta. Cryst. D. 1998;54:905–921. [PubMed] [Google Scholar]
  • Chen, X., Antson, A.A., Yang, M., Li, P., Baumann, C., Dodson, E.J., Dodson, G.G., Gollnick, P. Regulatory features of the trp operon and the crystal structure of the trp RNA-binding attenuation protein from Bacillus stearothermophilus . J. Mol. Biol. 1999;289:1003–1016. [PubMed] [Google Scholar]
  • Chevalier, B.S., Kortemme, T., Chadsey, M.S., Baker, D., Monnat, R.J., Stoddard, B.L. Design, activity, and structure of a highly specific artificial endonuclease. Mol. Cell. 2002;10:895–905. [PubMed] [Google Scholar]
  • Choi, J.W., Nam, Y.S., Lee, B.H., Ahn, D.J., Nagamune, T. Charge trap in self-assembled monolayer of cytochrome b562-green fluorescent protein chimera. Curr. App. Phys. 2006;6:760–765. [Google Scholar]
  • Collaborative Computational Project, Number 4. The CCP4 Suite: Programs for protein crystallography. Acta Cryst. D. 1994;50:760–763. [PubMed] [Google Scholar]
  • Douglas, T., Strable, E., Willits, D., Aitouchen, A., Libera, M., Young, M. Protein engineering of a viral cage for constrained nano-materials synthesis. Adv. Mater. 2002;14:415–418. [Google Scholar]
  • Endy, D. Foundations for engineering biology. Nature. 2005;438:449–453. [PubMed] [Google Scholar]
  • Gollnick, P. Regulation of the Bacillus subtilis trp operon by an RNA-binding protein. Mol. Microbiol. 1994;11:991–997. [PubMed] [Google Scholar]
  • Hartgerink, J.D., Clark, T.D., Ghadiri, M.R. Peptide nanotubes and beyond. Chem. Eur. J. 1998;4:1367–1372. [Google Scholar]
  • Heddle, J.G., Yokoyama, T., Yamashita, I., Park, S.-Y., Tame, J.R.H. Rounding up: Engineering 12-membered rings from the cyclic 11-Mer TRAP. Structure. 2006;14:925–933. [PubMed] [Google Scholar]
  • Heddle, J.G., Fujiwara, F., Yoshii, S., Nishio, K., Addy, C., Yamashita, I., Tame, J.R.H. Using the ring-shaped pProtein TRAP to capture and confine gold nanodots on a surface. Small. 2007a;3:1950–1956. [PubMed] [Google Scholar]
  • Heddle, J.G., Okajima, T., Scott, D.J., Akashi, S., Park, S.Y., Tame, J.R. Dynamic allostery in the ring protein TRAP. J. Mol. Biol. 2007b;371:154–167. [PubMed] [Google Scholar]
  • Hopcroft, N.H., Wendt, A.L., Gollnick, P., Antson, A.A. Specificity of TRAP-RNA interactions: Crystal structures of two complexes with different RNA sequences. Acta. Cryst. D. 2002;58:615–621. [PubMed] [Google Scholar]
  • McCammon, M.G., Hernandez, H., Sobott, F., Robinson, C.V. Tandem mass spectrometry defines the stoichiometry and quaternary structural arrangement of tryptophan molecules in the multiprotein complex TRAP. J. Am. Chem. Soc. 2004;126:5950–5951. [PubMed] [Google Scholar]
  • McCoy, A.J., Grosse-Kunstleve, R.W., Storoni, L.C., Read, R.J. Likelihood-enhanced fast translation functions. Acta Crystallogr. D Biol. Crystal. 2005;61:458–464. [PubMed] [Google Scholar]
  • McMillan, R.A., Paavola, C.D., Howard, J., Chan, S.L., Zaluzec, N.J., Trent, J.D. Ordered nanoparticle arrays formed on engineered chaperonin protein templates. Nat. Mater. 2002;1:247–252. [PubMed] [Google Scholar]
  • Miura, A., Hikono, T., Matsumura, T., Yano, H., Hatayama, T., Uraoka, Y., Fuyuki, T., Yoshii, S., Yamashita, I. Floating nanodot gate memory devices based on biominerlized inorganic nanodot array as a storage node. Jpn. J. Appl. Phys. 2006;45:L1–L3. doi: 10.1143/JJAP.45.L1. [CrossRef] [Google Scholar]
  • Murshudov, G.N., Vagin, A.A., Lebedev, A., Wilson, K.S., Dodson, E.J. Efficient anisotropic refinement of macromolecular structures using FFT. Acta Crystallogr. D Biol. Crystallogr. 1999;55:247–255. [PubMed] [Google Scholar]
  • Otwinowski, Z., Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymol. 1997;276:307–326. [PubMed] [Google Scholar]
  • Potterton, E., McNicholas, S., Krissinel, E., Cowtan, K., Noble, M. The CCP4 molecular-graphics project. Acta Cryst. D. 2002;58:1955–1957. [PubMed] [Google Scholar]
  • Ringler, P., Schulz, G.E. Self-assembly of proteins into designed networks. Science. 2003;302:106–109. [PubMed] [Google Scholar]
  • Roussel, A., Cambillau, C. Silicon Graphics Geometry Partners Directory. Silicon Graphics; Mountain View, CA: 1989. pp. 77–78. [Google Scholar]
  • Shirts, M.R., Pande, V.S. Mathematical analysis of coupled parallel simulations. Phys. Rev. Lett. 2001;86:4983–4987. [PubMed] [Google Scholar]
  • Sugimoto, K., Kanamaru, S., Iwasaki, K., Arisaka, F., Yamashita, I. Construction of a ball-and-spike protein supramolecule. Angew. Chem. Int. Ed. Engl. 2006;45:2725–2728. [PubMed] [Google Scholar]
  • Tsiavaliaris, G., Fujita-Becker, S., Manstein, D.J. Molecular engineering of a backwards-moving myosin motor. Nature. 2004;427:558–561. [PubMed] [Google Scholar]
  • Yang, M., Chen, X., Militello, K., Hoffman, R., Fernandez, B., Baumann, C., Gollnick, P. Alanine-scanning mutagenesis of Bacillus subtilis trp RNA-binding attenuation protein (TRAP) reveals residues involved in tryptophan binding and RNA binding. J. Mol. Biol. 1997;270:696–710. [PubMed] [Google Scholar]
  • Zagrovic, B., Sorin, E.J., Pande, V. β-Hairpin folding simulations in atomistic detail using an implicit solvent model. J. Mol. Biol. 2001;313:151–169. [PubMed] [Google Scholar]
Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society