Cloning, mutagenesis, and purification of the heat shock protein (Hsp) cage from an E. coli heterologous expression system was previously described (Flenniken et al 2003). The HspG41C cages, with interior exposed thiols, were characterized by size exclusion chromatography (Superose-6, Amersham-Pharmacia), dynamic light scattering (DLS) 90 plus (Brookhaven Instruments), and by transmission electron microscopy (TEM) (LEO 912 AB) (Supplemental Figure 1).

Two genetic variants of CCMV were utilized for these experiments, CCMV S102C for epifluorescence studies and CCMV K42R for biodistribution studies. Purification of CCMV S102C from the Pichia pastoris (yeast) heterologous expression system was performed as previously described (Speir et al 1993; Brumfield et al 2004). The primer sequences utilized for PCR mediated mutagenesis of CCMV S102C were: S102C for 5′ gttgcttcccagtgtttgtggtaccgtgaaatcctgtg 3′ and S102C rev 5′ cacaggatttcacggtaccacaaacactgggaagcaac 3′. The integrity of the entire CCMV S102C coding sequence was confirmed by sequencing on an ABI 310 automated capillary sequencer using Big Dye chain termination sequence technology (Applied Bio Systems, Foster City, CA, USA). The CCMV S102C capsids were purified and characterized by size exclusion chromatography (Superose 6, Amersham-Pharmacia), dynamic light scattering (DLS) 90 plus (Brookhaven Instruments), and transmission electron microscopy (TEM) (LEO 912 AB) (Supplemental Figure 1). CCMV K42R capsids were characterized (Fox et al 1996) and purified from plants as previously described (Bancroft and Hiebert 1967).

Cysteine containing genetic variants, CCMV S102C and HspG41C, were chemically derivatized with Texas Red^® C₂-maleimide (Invitrogen) using methods described in (Gillitzer et al 2002; Flenniken et al 2003). The cages (100 mM HEPES 50 mM NaCl pH 6.5) were reacted with Texas Red^® C₂-maleimide (Invitrogen) in a concentration of 3-fold molar excess per subunit. The reactions proceeded for 30 minutes at room temperature followed by overnight incubation at 4 °C. Texas-Red^® labeled cages were purified from free dye by size exclusion chromatography (Superose 6, Amersham-Pharmacia) in Dulbecco’s phosphate buffered saline pH 7.4.

Protein cages were iodinated utilizing IODO-BEADS Iodination Reagent (Pierce) according to manufacturer’s instructions. Unincorporated ¹²⁵I were removed with desalting columns (Zeba Desalt Spin Columns, Pierce): two rounds of desalting were performed. The iodinated protein cages was characterized by absorbance spectroscopy, size exclusion chromatography (Superose 6, Amersham Pharmacia), dynamic light scattering (Brookhaven 90Plus), and transmission electron microscopy (TEM) (Leo 912 AB) (Supplemental Figure 1). The specific activity of ¹²⁵I labeled protein cages was determined from counts associated with the sample compared to a linear regression fit to the counts from ¹²⁵I standards (ranging from 0–1000 nCi) (x-axis) plotted with their corresponding counts per minute (y-axis) (Beckman Gamma 4000 Counting System). Typically, specific activity for Hsp cage samples was ~20 nCi/μg and ~250 nCi/μg for CCMV samples. The incorporation of ¹²⁵I, in terms of specific activity (nCi/μg), was about 10-fold greater in CCMV K42R as compared to HspG41C cages. This discrepancy is most likely because CCMV has more surface accessible tyrosines (5 per subunit; 900 per cage; exterior Tyr 50, 157, 159, 190, interior Tyr 138) than Hsp (2 per subunit; 48 per cage; interior Tyr 96, subunit interface Tyr 106).

Mice received an intravenous dose, via the tail vein, of either 50 μg of ¹²⁵I-HspG41C or ¹²⁵I-CCMV cage in a 200 μL volume. The dose of radiation ranged from approximately 0.3 to 0.5 μCi of cage. The nanocages were allowed to circulate for 1 to 24 hours before animals were sacrificed with a lethal dose of sodium pentobarbital. When a lack of pedal response was apparent mice were opened and the abdominal aorta clipped. A 70 μL blood sample was obtained and all organs excised and weighed separately. The blood, tissue and injected dose samples were then transferred into vials and counted on a Beckman Gamma 4000 Counting System.

Tissue radioactivity measures were in counts-per-minute (cpm) and background adjusted. The percent of recovered dose per gram of tissue or per organ was calculated as follows: % of injected dose/g of tissue           = Activity recovered/g of tissueAdministered dose radioactivity × 100% of injected dose/organ           = Activity recovered/organAdministered dose radioactivity × 100

In order to compare the burden of deposition in various organs the percent of injected dose per gram of tissue was used. However, some tissues weigh less than 1 gram and result in percentages greater than 100. The percent injected dose/organs, urine and feces samples were totaled for mice at 1 and 24 hours so as to determine the total percent of dose recovered in each experiment.

The experimental animal protocol used for all studies including the ¹²⁵I trials was approved by the Institutional Animal Care and Use Committee and the Safety and Risk Management Office at Montana State University. Female BALB/c mice, 10–14 weeks old, weighing on average 18–22 gm each, were purchased from Charles River Laboratories (Wilmington, MA). The 24 hour mice were housed in Nalgene^® Metabolic Cage for Single Mouse (Rochester, NY) under normal conditions with ad libitum access to finely ground sterilized rodent chow and water.

Female BALB/c mice were immunized with either a 50 μg HspG41C or CCMV protein dose intravenously injected into the tail vein. Mice received the dose in a 200 μL volume at week 1 and two weeks later (week 3). At week 4 a blood sample was obtained to confirm immunization and then week 5 mice received an IV challenge of 50 μg of ¹²⁵I-labeled cage to determine particle biodistribution as stated below.

Samples were run using SDS-Polyacrylamide Gel Electrophoresis (PAGE) under denaturing conditions with TR-CCMV S102C and TR-HspG41C controls. Gels were visualized first by fluorescence (inverted image; Molecular Imager FX, BioRad) and then by Coomassie Brilliant Blue staining and compared to molecular weight markers.

Mouse lungs were instilled with 3% low melt agarose (kept at 37 °C) and allowed to set-up in cold PBS for later sectioning. One-hundred micron thick agarose lung sections were cut on a Leica VT1000S vibrating-blade microtome (Leica Microsystems Nussloch GmbH, Germany). Fresh kidney and liver cuts (100 microns thick) were also made using the vibrating microtome; these tissues samples were not agarose filled. Tissue sections were wet mounted on slides, cover slipped and viewed with the Nikon Eclipse E800 microscope (Nikon, Inc., Melville, NY) and imaged using the Nikon Digital Camera DXM1200 and MetaVue software (Molecular Devices Corporation, Downingtown, PA).

A morphological examination of lung, liver, spleen and kidney was conducted with tissues obtained from mice sensitized and challenged with either of HspG41C or CCMV cage protein. The organs were excised from mice at 1 and 24 hrs after the challenge and fixed in 10% phosphate-buffered formalin (Fisher Scientific), dehydrated in a graded series of ethanol, embedded in paraffin, and cut into 5 micron thick serial sections. Sections were then stained using hematoxylin and eosin (H and E), and the morphology of the various tissues examined using the Nikon Eclipse E800 microscope and images taken with the Nikon Digital Camera DXM1200 and MetaVue software as stated above.

Total IgG specific HspG41C and CCMV in sera was measured by direct enzyme-linked immunosorbent assay (ELISA). The sera were kept frozen at −80 °C before analysis. Each well of a 96-well polystyrene plate was coated with 50 μL of either a CCMV or Hsp solution at a protein concentration of 0.5–1.0 μg/mL and incubated overnight at 4 °C. The plate was then washed four times with a 2% horse serum solution prior to each well being blocked with 100 μL of a non-specific binding site blocking buffer (1% bovine serum albumin in PBS) and again incubated for two hours at room temperature. The plate was washed once more with the 2% horse serum solution and 100 μL of serum diluted in blocking buffer was added and the plate incubated for 2 hours at room temperature. The plate received four additional washes and 100 μL per well of goat anti-mouse peroxidase conjugated IgG, diluted in blocking buffer to a 1:500 concentration, was added then incubated for one hour at room temperature. The samples were again washed four times and ABTS substrate added, allowed to react for 15 minutes and visualized at 415 nm on a microplate reader (VERSA_MAX Microplate Reader, Molecular Devices, Sunnyvale, CA, USA).

Sera cage specific IgE was also measured by an ELISA assay. An affinity purified anti-mouse IgE capture mAb (2 μg/mL; clone R35-72, BD Bioscience, San Diego, CA, USA) was used to coat a 96-well ELISA plate (flat-bottom, high-binding, polystyrene; Costar, Corning, NY, USA) at 4 °C overnight. Next, the plates were washed three times with PBS/Tween 20 and incubated with a non-specific binding site blocking buffer (1% bovine serum albumin in PBS) for 1 h at room temperature. After incubation, the plates were again washed three times with PBS/Tween 20 and serum samples, controls, and standards applied; 1 hr incubation at 37 °C followed. Again the plates were washed three times and 100 μL per well of biotinylated anti-mouse IgE (2 μg/mL; clone R35-118, BD Biosciences) was added and the plate incubated for 1 h at room temperature. After washing, 100 μL per well of horseradish peroxidase avidin D (0.5 μg/mL; Vector Laboratories, Inc, Burlingame, CA) was added and the plate incubated at 37 °C for 30 minutes. The plates were washed with PBS/Tween 20 five last times. The presence of bound Ab was detected by the addition of 100 μL per well of o-phenylenediamine dihydrochloride (SIGMAFAST™ OPD, Sigma-Aldrich Co., St. Louis, MO) and visualized at 405 nm on a microplate reader (VERSA_MAX Microplate Reader, Molecular Devices, Sunnyvale, CA). The Ig contents of the serum samples were compared with standard curves generated from the 405 nm absorbance levels of known amounts of IgG or IgE.

The majority of the data were analyzed using GraphPad PRISM (version 4.00; 2003). Data are presented as mean values ± standard deviation (SD) and the unpaired Student t-test with calculations of a two-tailed p-value were utilized to determine significance; differences with a p-value < 0.05 were considered significant. Results from the ¹²⁵I-CCMV, ¹²⁵I-HspG41C, and free ¹²⁵I experiments were evaluated by GraphPad PRISM one-way ANOVA to determine statistical significance; differences between the free ¹²⁵I and ¹²⁵I-labeled cage were evaluated using the Bonferroni’s Multiple Comparison Test for multiple post-hoc comparisons. The comparisons of percent injected dose/organ for selected organs from the ¹²⁵I-CCMV and ¹²⁵I -HspG41C experiments were made using two-way ANOVA analysis with the GLM procedure of SAS (version 8.2; Statistical Analytical Systems Institute, Inc., Cary, NC, USA). The main effects of immune status (naïve vs. immunized) and time (1 hr vs 24 hr), and the interaction between immune status and time were evaluated. When the interaction was not significant the main effect means are reported. If the interaction was significant, then the interaction mean was reported. If the treatment F-test was significant (p < 0.05) the treatment least square means were separated by the LSD method.

Genetic variants of both CCMV (CCMV S102C) and Hsp (Hsp G41C) were chemically derivatized with Texas Red^®(TR) in order to visualize the cages by epi-flourescence microscopy. Texas Red-labeled CCMV and TR-HspG41C cages were intravenously (IV) injected into mice from which various tissues were subsequently harvested and evaluated. Fluorescent images from vibratome sections of tissues demonstrated TR-HspG41C cages in the lung, liver and kidney and TR-CCMV in the liver and kidney at 1 hour post IV injection (Figures 1 and 2). Neither the TR-HspG41C nor TR-CCMV particles appear to be associated with any one cell type in the lung; rather, they were visible on pulmonary endothelial surfaces, interstitial perivascular areas, and in alveolar lymphatic ducts (Figure 1). The TR-HspG41C introduced into the mouse intranasally (IN) appeared rapidly in the lymph nodes of the respiratory tract (tracheobronchial and cervical lymph nodes) within 1 hour post instillation (data not shown). The TR-labeled protein cages appeared to freely move in and out among the epithelial cells of the airways, the pulmonary endothelial cells, and pulmonary draining lymph nodes.

TR-labeled cages were also evident in the sinusoids of the mouse liver at 1 hour after IV injection (Figures 1 and 2). They appeared to be associated with either phagocytic Kupffer cells or endothelial cells of the sinusoid radiating from the terminal hepatic venule, but were otherwise evenly dispersed throughout the liver. This association of the cages with these cell types was not surprising, as both have the capacity to interact with particulate materials via phagocytosis, thus protecting the systemic circulation by clearing the plasma of foreign material (Steffan et al 1986; Dini et al 1995).

As seen in lung and liver, both cage types were present in the kidney. One hour after IV injection, cages could be seen in the convoluted tubules and interlobular connective tissue of the mouse kidney. By 24 hours post injection cages were still visible primarily in the convoluted tubules of the kidney (Figure 1), although with less intensity than seen at 1 hour, and in the cortical lymphatics. In summary, these results indicate that both non-cell targeted CCMV and Hsp protein cages can be rapidly found within the lungs, kidneys and liver, declining by 24 hours post IV injection.

In order to evaluate the biodistribution of CCMV and Hsp cages they were radiolabeled with ¹²⁵I. Both the CCMV and Hsp protein cages were rapidly and broadly distributed to a majority of the mouse tissue and organ types. Within 1 hr post injection, both the CCMV and Hsp protein cages were detected in all examined tissues and organs, with 0.1% or less of the injected dose in the brain (Table 1). The biodistribution of ¹²⁵I-HspG41C or ¹²⁵I-CCMV cages in naïve mice at the 1 hour time point did not significantly differ from one another in terms of percent injected dose/gram of tissue (%ID/g), with the exception of the salivary gland (Figure 3). The salivary gland of mice injected with ¹²⁵I-HspG41C had 56% greater %ID/g of tissue than the ¹²⁵I-CCMV injected mice. The five sites of greatest deposition for the ¹²⁵I-HspG41C or ¹²⁵I-CCMV cages at 1 hour were the bladder, thyroid, liver, spleen, and salivary gland. Thus, independent of protein cage size or amino acid composition, a similar pattern of rapid biodistribution of both protein cage types to most tissue and organ types was observed.

The biodistribution of ¹²⁵I-labeled cage was evaluated in naïve and immunized mice. Overall, by 24 hours post injection, both the CCMV and Hsp protein cages had been significantly cleared to a similar extent from most organs in naïve mice except the thyroid where levels actually increased over 24 hours (Figures 3 and 4; Tables 1 and 2). In general, greater than 80% of the recovered labeled protein cages had been removed by 24 hrs as compared to 1 hr post injection. Even though there was an overall reduction of both protein cages, there were significant differences (p < 0.05) in distribution of the two cage types in the liver, spleen and kidney, such that a greater %ID/g of liver, spleen and kidney was found in ¹²⁵I-CCMV versus ¹²⁵I-HspG41C injected mice. It was evident the majority of the ¹²⁵I-HspG41C or ¹²⁵I-CCMV cages protein cages were secreted through the urine and feces (56%–66% of dose) (Table 2). These results demonstrated that independent of protein cage type, size or amino acid composition, a similar pattern of rapid removal of both protein cages from most tissue and organ types in naive mice was observed.

The CCMV and the Hsp protein cages demonstrated similar biodistribution patterns in both naïve and immunized mice. Similar to the biodistribution patterns observed in naive mice, both the CCMV and the Hsp protein cages were detected in all tissue and organ types except the brain in immunized mice. When ¹²⁵I-HspG41C or ¹²⁵I-CCMV protein cages were administered IV to mice previously immunized with either the Hsp or CCMV protein cages, the distribution pattern at 1 hour was similar for both types of protein cages expressed as either a percentage of dose per gram of tissue or per organ basis (Figures 3 and 4; Tables 1 and 2). Two exceptions to this were the %ID of ¹²⁵I-CCMV found per gram of both lung and kidney of immunized mice differed significantly from the %ID of ¹²⁵I-HspG41C injected in those tissues (p < 0.05). In immunized mice at 1 hr, the %ID/gm of lung found in mice receiving iodinated HspG41C cage was 4.5 times greater than that in the lungs of mice receiving iodinated CCMV (30.8% ID/g versus 6.8% ID/g) (Table 2 and Figure 3). Conversely, a comparison of the %ID/g of kidney indicated a greater amount (1.4 times) of iodinated cage present in CCMV-immunized mice than in HspG41C-immunized mice at 1 hour.

When the biodistribution of labeled cages was examined on a per total organ basis, the pattern of deposition of the two types of nanocages among mice organs at 1 and 24 hours was similar (p > 0.05) (Figure 4). With regards to naïve versus immunized mice, the distribution of nanocages in the whole organs was similar with the exception of the liver and GI tract. At 1 hour, the ¹²⁵I-HspG41C IV administered naïve mice had approximately 2 times more of the injected dose in the GI tract than the Hsp-immunized mice, while the CCMV-naïve mice had 2.5 times more cage associated counts in the GI tract than CCMV-immunized mice. The opposite was true for the %ID in the liver; at 1 hour 2.8 times more of the ¹²⁵I-HspG41C dose and 1.6 times more of the ¹²⁵I-CCMV dose was found in the liver of immunized versus naïve mice. The %ID/g of thyroid was appreciably less at the 1 hour time point in immunized versus naïve mice for ¹²⁵I-HspG41C or ¹²⁵I-CCMV cage (p < 0.05; 3.5 and 22.3 times less respectively) (Figure 3).

In order to test the possibility of free ¹²⁵I release from the radiolabeled protein cages skewing the distribution, an investigation into the biodistribution of free ¹²⁵I in mice using identical methods to those used in the protein cage biodistribution was conducted. The biodistribution of free ¹²⁵I was notably different than the biodistribution of the injected iodinated protein nanocages. The results of the free ¹²⁵I study indicated that at 1 hour post injection the majority of free ¹²⁵I was found in the tegument (24.6 % ID) and gastrointestinal tract (16.44 %ID). The organs associated with the highest %ID/g of tissue were the thyroid and bladder (221.3% and 208.9%/g tissue respectively). Gamma counts obtained from the 24 hour collection of urine were high as well (63.7%ID, 74.28% of dose recovered) (Figure 5). The presence of free ¹²⁵I- in the lungs of injected mice had less %ID/g of tissue than the ¹²⁵I-HspG41C-immunized injected mice at 1 hour (5.03% vs 30.78%; p < 0.05). The %ID/g of liver and spleen at 1 and 24 hours was significantly less (p < 0.05) in free ¹²⁵I-injected mice compared to mice injected with ¹²⁵I-labeled cage (Figure 5). One-way ANOVA analysis revealed a significant difference at the 1 and 24 hour time points (p < 0.05) when the distribution of ¹²⁵I-HspG41C or ¹²⁵I-CCMV in the thyroids of naïve and immunized mice were compared to free ¹²⁵I injected mice; immunized mice receiving iodinated HspG41C or CCMV had less %ID/g of thyroid than mice receiving free ¹²⁵I alone (Figure 5).

In general, the biodistribution of two different protein nanocages in the organs and tissues of naïve or immunized mice were similar, both at 1 and 24 hours post injection. As well, no real pattern of preferential accumulation in one tissue or organ type was readily apparent, with the possible exception of the thyroid and liver.

Both cage types (¹²⁵I-HspG41C and ¹²⁵I-CCMV) were nearly completely cleared from mouse tissues by 24 hours. The total percent of injected dose recovered in the urine alone ranged from 51% to 60%; from the feces and urine combined values were 56.6%–63.3%, while the total recovered dose ranged from 64% to 89% and none of these were significantly different.

Various analyses were conducted to determine the form of the excreted cage in the urine. In this regard, direct visualization of viral concentrates of urine samples (100,000 × g viral pellet fraction) by transmission electron microscopy failed to identify intact virus particles. In addition, enzyme-linked immunosorbent assay (ELISA) and western immunoblot analysis utilizing CCMV polyclonal antiserum failed to detect intact either 28 nm CCMV particles or intact coat protein subunits. No evidence was found that CCMV protein cages were secreted in the urine intact. Urine samples from mice IV injected with TR-HspG41C or TR-CCMV were analyzed using a SDS-polyacrylamide gel electrophoresis; first imaged with fluorescence and subsequently stained with Coomassie Blue. Results, when compared to control urine spiked with TR-HspG41C or TR-CCMV, indicated that the cages were not intact and coat protein subunits were not detected (data not shown). These observations suggest that in the ¹²⁵I-HspG41C or ¹²⁵I-CCMV injected mice the cages were degraded in vivo and excreted in components smaller than the cage protein subunits.

An ELISA was used to analyze serum from immunized mice for either HspG41C- or CCMV-specific IgG. Serum samples were obtained from mice seven days after receiving two 50 μg IV injections two weeks apart. Results indicated animals were highly positive for cage protein-specific IgG (data not shown). These mice were then used in the ¹²⁵I-labeled cage biodistri-bution experiments described above. The CCMV-immunized mice, and to a lesser extent HspG41C-immunized mice, exhibited signs of distress 5–7 minutes post injection of the ¹²⁵I-labeled cages, including hypopnea, ruffled fur, and lethargy. Signs of recovery began approximately 1.5 hrs post injection and a full recovery was apparent at 2.5 hrs post IV administration. Also, the spleens of the 1 hr CCMV injected mice were darker in color, the lungs appeared congested, and it was difficult to obtain the usual 70 μL volume of blood sample from the mice after clipping the abdominal aorta.

Immunized mice injected with either HspG41C or CCMV nanoparticles under the same conditions as the ¹²⁵I-injected mice did not have any remarkable histological pathology, based on observation of liver, kidney or spleen hematoxylin and eosin stained sections. However, observations of lung sections from CCMV-immunized mice revealed pulmonary vascular congestion, which may have accounted for some of the respiratory distress observed in these mice. Spleen sections from HspG41C- or CCMV-injected mice appeared different from each other; areas of white pulp in the HspG41C-injected mouse spleen were greater in size with well defined germinal centers. Serum samples taken from immunized mice prior to challenge with IV cage injections were analyzed using an ELISA technique for cage specific IgE levels; no IgE was detected (data not shown).