![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Microbiology and Immunology, Loyola University Medical Center, Maywood, Illinois 60153,1 Bacteriology Division, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland 21702-5011,2 Department of Biological, Chemical, and Physical Sciences, Illinois Institute of Technology, Chicago, Illinois 60616,3 Department of Molecular, Cellular, and Developmental Biology, University of Michigan, Ann Arbor, Michigan 48109,4 Headquarters, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland 21702-50115
Received 26 June 2006/ Accepted 1 November 2006
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
The significant differences between Bacillus spp. in coat and exosporium morphology and the high degree of conservation of morphogenetic coat protein genes raise the question of how spore assembly programs among species have diverged to generate variations in spore ultrastructure and function. A critical aspect of this question is the role of the conserved morphogenetic coat proteins. Functional analysis of these genes in a variety of species should provide significant insight into how programs of cellular assembly evolve.
Here, we characterize B. anthracis homologues of B. subtilis morphogenetic coat proteins. B. anthracis coat assembly is of particular interest in anthrax pathogenesis. Studies of B. subtilis strongly argue that the coat (but not necessarily any individual coat protein) will play an essential role in B. anthracis spore resistance and in germination, both of which are needed for disease (16, 18, 19, 32, 75). Coat proteins are also likely to be useful as vaccine candidates (9, 13, 49, 90) and ligands for biological weapon detectors (10, 84, 92). Proteins to be used for these purposes should be required for virulence; otherwise, a determined enemy might prepare spores that are missing these proteins. In light of this, an especially critical open question is the possible role, if any, of the exosporium in virulence. This has not yet been tested using a virulent B. anthracis strain. However, an important study by Sylvestre et al. (80) showed the dispensability of a major exosporium protein, BclA, in an infection model that uses the attenuated Sterne strain (which lacks the capsule, a critical virulence factor) (22) and an immunocompromised mouse. BclA is the major structural component of the fringe of hair-like projections present on the exosporium basal layer (78, 80, 81). Therefore, in this model, BclA is not needed for disease. The role of any given coat protein in virulence is also unknown; while some degree of coat formation is undoubtedly required for anthrax, at least some individual proteins are likely to be dispensable.
We find that in spite of the large percentage of conserved coat proteins, coat assembly differs significantly between B. subtilis and B. anthracis. To a large degree, this is because important conserved morphogenetic proteins function in strikingly different ways in the two species. Mutations in morphogenetic protein genes may therefore be particularly important in the evolution of species-specific features of spores and their adaptation to novel niches. In addition, we find that the exosporium, the outermost structure of the B. anthracis spore, is dispensable for virulence in both intramuscular and intranasal challenge models. Therefore, an individual who might prepare B. anthracis as a biological weapon has the option of generating spores with or without an exosporium. As a result, we argue that proteins in both the coat and the exosporium should be considered for the use of vaccines or detector ligands.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ames strain vegetative cultures were grown in LB broth, and spores were prepared as previously described (36) by using Leighton-Doi medium (47) or Difco sporulation agar (73). Spores were then purified by centrifugation through a Hypaque-76 gradient (Nycomed, Inc., Princeton, NJ). Prior to use, spores were heated at 65°C for 30 min and then kept on ice.
Strains and plasmids used in this study are described in Table 1. Primer sequences are available upon request. Recombinant DNA procedures and chromosomal DNA isolation were performed according to methods described previously by Sambrook et al. (71) and Cutting and Vander Horn (14), respectively. Coat protein extraction and sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) analysis were performed as described previously (48). Sporangial lysates were prepared from 1 ml of a sporulating culture according to a method described previously (11). Western blot analysis was performed as described previously (4) using B. subtilis SpoIVA antiserum and B. anthracis CotE antiserum (see below). Electron and atomic force microscopy (AFM) were performed according to methods described previously by Catalano et al. (11) and Chada et al. (12), respectively. The calcium release assay was carried out according to a method described previously (87), except that spores were cultured on solid modified G medium (15 g/liter agar) for 3 to 5 days and germinated on AAC medium (89). Fluorescent dye uptake, heat sensitivity, and lysozyme sensitivity assays were carried out according to methods described previously by Welkos et al. (89), Bozue et al. (8), and Cutting and Vander Horn (14), respectively. The tetrazolium overlay assay was performed as described previously (14), except that the medium used to induce germination was supplemented with L-alanine (1.7 g/liter instead of 1 g/liter for B. subtilis), Casamino Acids (0.4 g/liter), and adenosine (1.7 g/liter).
|
Phase-contrast microscopic examination of germination. Water-washed, heat-activated spores were suspended in water to a final optical density at 600 nm of 5.0. Twenty microliters of the spore suspension was spotted onto a slide that was first treated with 0.1% (wt/vol) poly-L-lysine. After 10 min of incubation, the remaining liquid was aspirated, nonadherent spores were removed by two water washes with 20 µl of water, and the sample was allowed to dry. To initiate germination, 5 µl of AAC medium was spotted onto the slide, followed by the placement of an 18-mm by 18-mm coverslip over the drop. Phase-contrast microscopy was performed using a Nikon Labophot microscope. Spores from three independent sporulations were scored as phase bright or phase dark for each strain at each time point. One hundred to 300 spores were counted for each time point. To prevent drying, slides were placed in a humid chamber between observations. The experiment was carried out at room temperature. The percentage of germination was calculated as follows: (total number of spores – number of phase-bright spores)/total number of spores.
Creation of B. anthracis mutants.
Sterne and Ames strain mutants were generated using approaches based on methods described previously by Koehler et al. (41) and Mendelson et al. (53), respectively. To generate insertion-deletion mutations in the Sterne strain, we first constructed plasmid pMR1. To do this, a DNA fragment bearing the 
Km-2 kanamycin resistance cassette generated by digesting plasmid pUC4Km-2 (62) with BamHI was isolated and ligated into plasmid pSL1180 (Strategene) digested with BglII, resulting in plasmid pES1. The SacI-SalI fragment of pES1 (bearing Km-2) was ligated into plasmid pUTE29 (41) digested with the same enzymes, resulting in pMR1. To construct each coat protein gene mutation, we used PCR to amplify approximately 1-kb stretches of DNA on either side of each gene using primers. Each resulting pair of DNA fragments was then inserted into pMR1 such that they flanked the kanamycin resistance marker in the plasmid. These recombinant DNA manipulations were carried out using the restriction enzyme sites built into the primers. The plasmids were amplified in E. coli GM1684, purified using the QIAGEN Plasmid Mini kit (to increase transformation efficiency) (41), and introduced into the Sterne strain by electroporation. Transformants were selected on the basis of tetracycline resistance as described previously (41).
We isolated recombinants by a two-step procedure. First, we grew transformants in LB medium for 3 days, diluting them by a factor of 1:10 twice a day. On the third day, the culture was plated onto LB agar supplemented with kanamycin. We screened colonies by PCR to confirm that the expected single integration (Campbell-like) event had occurred. To inactivate the target gene, we needed a further reciprocal recombination to excise the plasmid backbone and, in effect, replace the target gene open reading frame with the kanamycin resistance cassette. Therefore, we continued the culture regimen to allow this event to take place. After several days, samples of the liquid culture were plated onto LB agar supplemented with kanamycin. We isolated kanamycin-resistant and tetracycline-sensitive colonies by replica plating and used PCR to confirm that the expected recombination had occurred.
We inactivated spoIVA in the Sterne strain by an insertional mutation (53). To do this, we first used PCR and appropriate primers to generate an internal fragment of spoIVA. We digested the PCR product and plasmid pEO-3 (53) with BamHI and HindIII and ligated the resulting DNA fragments to build pRG23. We passaged pRG23 through E. coli GM1684, introduced it into the Sterne strain by electroporation, and selected for cells in which the plasmid had inserted into the genome by single reciprocal recombination, as described above. This operation inserted the erythromycin-bearing plasmid backbone into spoIVA and should generate a null allele, based on extensive analysis of the highly related B. subtilis orthologue (11). We then backcrossed this mutation into the Sterne strain by CP51ts-mediated generalized transduction (30). We sporulated the resulting strain (RG154) in the presence of erythromycin, as this was needed to maintain the phenotype (data not shown).
We insertionally inactivated cotE, cotH, and cotB in the Ames strain as previously described (53). To do this, we used PCR (and appropriate primers) to generate DNA fragments corresponding to internal regions of each coat protein gene, digested each fragment with KpnII and NotI, and ligated each fragment with similarly digested pEO-3 (53). We passaged the resulting plasmid through E. coli GM2163 cells, introduced it into the Ames strain by electroporation, and selected for integrants as described previously (53).
To insertionally inactivate spoIVA in the Ames strain, we first used PCR to generate a DNA fragment bearing the spoIVA gene. The fragment was digested with NotI and KpnI and ligated into similarly digested pEO-3 (53). The resulting plasmid was digested with EcoRV (which cuts once within spoIVA) and was ligated into a DNA fragment bearing the Km-2 cassette (62), which was generated by SmaI digestion of plasmid pJRS100.2 (M. Caparon, personal communication), a derivative of pJRS102.0 bearing Km-2 at the EcoRI site. This plasmid was then used to transform the Ames strain as previously described (53), and a strain bearing the plasmid as a single reciprocal integration was identified. Next, we cultured this strain in Leighton-Doi medium for several days, as described above for mutagenesis of the Sterne strain, and screened for erythromycin sensitivity and kanamycin resistance to identify a strain (Ames-JAB-11) in which the plasmid backbone had been excised by a reverse reciprocal excision event. We confirmed that this occurred by PCR analysis.
We note that it is highly unlikely that our mutagenesis procedures resulted in significant alterations of the genome beyond what was intended, as in each case, mutations were generated in two backgrounds that produced essentially the same phenotypes (including the case of the cotB mutant) (data not shown).
Exosporium isolation and removal. To isolate exosporium material from the Sterne strain, 1 x 109 water-washed spores/ml were passed three times through a French pressure cell at 20,000 lb/in2. Fragments of the exosporium were then separated from spores by centrifugation at 4,000 x g for 15 min. The supernatant was filtered through a 0.45-µm Millex HA filter (Millipore) and was then filtered through a 0.22-µm polyvinylidene difluoride filter (Fisher Scientific) to remove any residual spores. The filtrate was centrifuged at 100,000 x g for 1 h, and the resulting pellet was resuspended in 50 mM Tris-HCl and 0.5 M EDTA (pH 8). Pressed spores were washed once with water prior to Western blot analysis. The anti-BclA EG4-4-1 antibody (1.3 mg/ml) was used at a dilution of 1:100,000 to detect exosporium material (78).
Residual exosporium was removed from the Ames strain cotE mutant by sonication (5) in a biological safety cabinet, with the spores in a sealed atmosphere chamber and on ice, using a Branson 450 sonifier (Danbury, CT). Spores were then purified by centrifugation through a Hypaque-76 gradient (Nycomed, Inc.) to remove debris. The removal of the exosporium was confirmed by electron microscopy (data not shown).
Antibody production. To produce a polyhistidine-tagged version of CotE in E. coli, we used PCR with chromosomal DNA as the template and appropriate primers to generate a DNA fragment encoding CotE. We then digested this DNA fragment with NdeI and XhoI, ligated it with the similarly digested plasmid pET24b (Novagen), and used the resulting plasmid (pMR10) to transform E. coli BL21(DE3). We verified that the cotE sequence was free of mutations by DNA sequence analysis. We overproduced CotE and purified it from E. coli by nickel chromatography according to the manufacturer's directions (Novagen) and contracted with a commercial facility (Sigma Genosys) to produce antibody in rabbits. We used the resulting anti-CotE serum at a dilution of 1:50,000 in Western blot experiments.
Interestingly, neither the B. anthracis CotE protein in a spore extract nor B. anthracis CotE overproduced in E. coli was recognized by a previously generated anti-B. subtilis CotE antiserum (data not shown) (4), in spite of the high degree of identity (58%) between these homologues. Likewise, the anti-B. anthracis CotE antiserum does not recognize B. subtilis CotE (data not shown).
Immunofluorescence microscopy.
Ten microliters of culture was placed into each well of a multiwell microscope slide (MP Biomedicals) that was first treated with 0.01% (wt/vol) poly-L-lysine and allowed to dry. The slide was then treated with methanol for 5 min at –20°C and allowed to dry. Ten microliters of phosphate-buffered saline was then placed into each well and was replaced with 2% (wt/vol) bovine serum albumin in phosphate-buffered saline prior to the addition of primary antibody. As primary antibodies, we used either mouse immunoglobulin G1
1 control antibody (0.5 mg/ml; Pharmingen) at a 1:2,000 dilution or mouse monoclonal anti-BclA EG4-4-1 antibody (1.3 mg/ml) (78) at a 1:5,000 dilution. We used Alexa Fluor 568-conjugated goat anti-mouse antibody (Molecular Probes) as a secondary antibody at a 1:300 dilution. Chromosomes were visualized using the DNA-specific dye Hoechst 33342 (Sigma) (11). Images were collected using a Leica DM IRB fluorescence microscope equipped with a MagnaFire cryocooled charge-coupled-device camera and processed with Adobe Photoshop 7.0 software.
Spore survival in macrophages. RAW264.7 macrophage murine cells (American Type Culture Collection) were infected at a multiplicity of infection of approximately 5 to 10 spores/macrophage and were cultured in Dulbecco's modified Eagle's medium with high glucose, 10% heat-inactivated fetal bovine serum, and 1% sodium pyruvate. Spore survival was measured by in vitro assays, and spores were observed within macrophages by staining as described previously by Welkos et al. (88).
Animal virulence models. (i) Intramuscular challenge experiments. Ten female Hartley Guinea pigs (350 to 400 g each, obtained from Charles River Laboratories) were injected intramuscularly with approximately 2,000 spores of either wild-type or cotE or cotH mutant strains, representing the equivalent of 20 Ames 50% lethal doses (LD50s) (35). Guinea pigs were monitored several times each day, and mortality rates were recorded.
(ii) Intranasal challenge experiments. Female BALB/c mice (approximately 6 to 8 weeks of age) were anesthetized with 100 µl of ketamine, acepromazine, and xylazine injected intramuscularly. The mice were then challenged with approximately 7.2 x 105 spores of either the wild-type or cotE mutant strain, representing the equivalent of 9 Ames LD50s, via intranasal instillation. The total volume of inoculum instilled was 50 µl. Mice were monitored several times each day, and mortality rates were recorded for 14 days.
Statistics. Survival rates were compared between each treatment group and control group by Fisher exact tests with permutation adjustment for multiple comparisons. Mean times to death were compared between each treatment group and control group by t tests with permutation adjustment for multiple comparisons. All analyses were conducted using SAS version 8.2 (SAS Institute Inc.).
Identification of gene BAS2377.
After SDS-PAGE, a band migrating as a 13-kDa species in a 15% SDS-polyacrylamide gel was excised and digested with trypsin using a MassPrep robot (MicoMass, United Kingdom). The peptides were extracted from the gel plug with 2% acetonitrile-0.1% trifluoroacetic acid and purified and concentrated using C18 Zip tips (Millipore). The peptides were eluted from the Zip tips with 5 µl 60% acetonitrile-0.1% trifluoroacetic acid. Two microliters of matrix-assisted laser desorption ionization (MALDI) matrix was added, and the sample was evaporated to dryness in a Speed-Vac, dissolved in 
1 µl of solvent, and spotted in its entirety onto the MALDI target. Mass spectra were acquired using an Applied Biosystems 4700 proteomics analyzer (tandem time of flight). The data from each tandem mass spectrometry spectrum were searched against the NCBI database (accessed in November 2005) using the Mascot search engine (v1.9; MatrixScience, London, United Kingdom).
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have shown previously that B. subtilis and B. anthracis have similar numbers of coat proteins, although the electrophoretic patterns differ significantly (44, 46, 68) (Fig. 3, compare lanes 1 and 8). To learn whether this level of polypeptide complexity is a general property of spores of Bacillus spp. and therefore to better understand the significance of the differences in the electrophoretic patterns of B. subtilis and B. anthracis, we examined several other species as well as multiple strains of B. subtilis and Bacillus cereus (Fig. 3). All the strains that we examined, except for Bacillus pumilis, possessed similarly large numbers of coat proteins. Given the morphological differences between the coats of B. subtilis and B. anthracis as well as the other species (1, 34), we suggest that the degree of structural complexity observed by electron microscopy is not, in general, correlated with the numbers of coat polypeptide species. For the most part, protein bands from the various isolates did not comigrate, unless they were extracted from spores of strains of the same or closely related species. Therefore, the biochemical composition of coats from different species is likely to vary significantly.
|
Role of SpoIVA in B. anthracis. Given the importance of the morphogenetic coat proteins SpoIVA, CotE, and CotH in B. subtilis (19), we analyzed the phenotypes of B. anthracis strains (both Sterne and Ames) bearing mutations in the corresponding homologues. The resulting strains grew normally and entered sporulation, as judged by light microscopy (data not shown). However, the spoIVA mutant strains in the Sterne and Ames backgrounds (RG154 and JAB11, respectively) both failed to progress beyond very early stages of sporulation. Light microscopy revealed that neither strain developed refractile forespores during sporulation or released a refractile body after extended culturing under sporulation conditions (data not shown).
To precisely determine the nature of the sporulation defect in these cells, we carried out thin-section electron microscopy. We found that after a period of sporulation that was sufficient to allow the formation of mature forespores in sporulating cells of a wild-type strain, spoIVA mutant sporulating cells were clearly defective (Fig. 1C and D and data not shown). The forespores lacked the germ cell wall or cortex, and the mother cells were filled with swirls of electron-dense material (Fig. 1D and data not shown). This is extremely close to the B. subtilis spoIVA mutant phenotype, where we know that the swirls are unattached coat material (20, 70). Based on the lack of forespore-associated structures, the swirls of material in the mother cell, and the B. subtilis spoIVA mutant phenotype, we consider it very likely that these swirls contain both coat and exosporium material. We infer, therefore, that as in the case of B. subtilis, SpoIVA is responsible for the attachment of the coat (and, indirectly, the exosporium) to the forespore surface (20, 64, 65).
The view that SpoIVA in B. anthracis directs early sporulation events, such as the initial assembly of the coat and cortex, requires that the protein be synthesized at an appropriately early time. To test this, we carried out Western blot analysis using anti-B. subtilis SpoIVA antibodies (11). We detected 47-kDa and 45-kDa species in sporulating cells of the Sterne strain (but not the spoIVA mutant) beginning at 1 h after the onset of sporulation (Fig. 4A), consistent with the timing of spoIVA gene expression from microarray analysis (for B. anthracis) (50) and Northern blot analysis (for B. subtilis) (70). The 47-kDa species was not detected after hour 3. Neither species was detectable after hour 4, likely as a result of the difficulty of extracting the protein when the maturation of the coat proceeded past an intermediate stage. Taken as a whole, we interpret these experiments to indicate that in B. anthracis, as in B. subtilis, SpoIVA is required for germ cell wall and cortex formation and for attachment of the coat.
|
To more carefully characterize the cotE spore surface, we imaged it in the dry state by AFM in the tapping mode, a method that generates a topographical relief map (25, 76). Wild-type Sterne strain spores have ridges on the coat surface that extend in a relatively straight line from pole to pole (12) (Fig. 1I). In contrast, the cotE Sterne strain mutant had either a series of disorganized ridge-like structures that usually did not span the long axis of the spore (Fig. 1J) or no ridges at all (Fig. 1K). In the latter case, the mutant spore coat surfaces resembled those of the corresponding B. subtilis mutant (12).
By using thin-section electron microscopy, we detected inner and outer layers in Sterne cotE mutant spores (Fig. 1F, inset), in contrast to the corresponding B. subtilis mutant, which lacks an outer coat (94). To characterize the composition of the cotE mutant spore coat, we visualized the extractible coat proteins after fractionation by SDS-PAGE. In B. subtilis, a cotE mutation results in the loss of about half of the bands that are normally present (4, 48, 94). In contrast, the electrophoretic patterns of extracts of wild-type and cotE Sterne spores were very similar (Fig. 5, lanes 1 and 2, respectively). The major consistent difference between wild-type and mutant spore extracts was the absence of a band migrating as an approximately 13-kDa species. In B. subtilis, especially severe changes in coat composition, including those due to the cotE mutation, can impair the ability of the coat to function as a sieve that excludes large molecules, as measured by the ability to exclude lysozyme (1, 4, 14). To determine whether this is the case for the B. anthracis cotE mutant, we incubated spores form strain RG56 with lysozyme. We found that these spores were not detectably more sensitive to lysozyme than the wild type, consistent with the view that the coat is not severely disrupted by the mutation. Therefore, in B. anthracis, CotE has a major role in exosporium assembly but a more limited, although significant, role in coat formation.
|
In B. subtilis, CotE controls the assembly of at least 24 coat proteins (39, 48, 94). Sixteen of these proteins have orthologues in B. anthracis, and for eight of them, paralogues are present as well. Given the modest role of CotE in B. anthracis coat protein assembly (Fig. 5, lane 2), it appears that CotE is required for the assembly of very few, if any, of these orthologues and paralogues. To investigate this finding further, we tested whether CotE has a role in the assembly of the two B. anthracis orthologues of CotB (46), as the assembly of B. subtilis CotB is known to be CotE controlled (39, 52, 94). First, we generated a version of the Sterne strain (RG35) in which the cotB orthologues were deleted. As anticipated, a band previously identified as a CotB orthologue was undetectable in spore extracts from the cotB mutant (Fig. 5, lane 4) but was present in cotE mutant spores (Fig. 5, lane 5), suggesting that B. anthracis CotB is CotE independent.
In B. subtilis, CotE is synthesized early in sporulation, consistent with its role at an early stage in coat assembly (94). The mutant phenotypes for strain RG56 left unclear whether CotE acts early or late in the sporulation in B. anthracis. To help clarify this, we monitored the steady-state levels of CotE over the course of sporulation by using Western blot analysis with an anti-CotE antiserum. We first detected CotE 2 h after the start of sporulation (Fig. 4B). At all time points, we detected a prominent band migrating as a 26-kDa species, which is likely to be the monomer. At later times, we detected slower-migrating species, including bands at 55 kDa and a band that migrates as larger than 170 kDa, which are likely the result of cross-linking. In addition, we detected smaller species between 17 and 26 kDa, which are probably due to proteolysis. Similar low-molecular-weight bands were detected previously in similar experiments with B. subtilis (4). The effect of CotE on both coat and exosporium assembly (Fig. 1) is consistent with a possible location in either or both structures. To address this, we separated the exosporium from Sterne strain spores using a French press and differential centrifugation and analyzed both fractions by Western blot analysis using anti-B. anthracis CotE and anti-BclA antibodies to identify coat and exosporium material, respectively (Fig. 6A and B). The CotE antiserum specifically recognized multiple species at less than 26 kDa (Fig. 6A). Differences between the bands of sporangial or spore extracts detected in Western blot analysis raise the possibility of modifications of CotE during coat assembly. The anti-BclA antibody recognized a species larger than 170 kDa in the Western blot analysis of spore extracts, consistent with previous reports (Fig. 6B, lane 1) (78, 80, 81). Even after French press treatment, at least some exosporium was still spore associated, as judged by Western blot analysis (Fig. 6B, lane 2). More importantly, CotE was undetectable in the exosporium fraction (Fig. 6A, lane 3). As expected, we did not detect bands with either antibody when we analyzed extracts of the cotE mutant (Fig. 6A and B, lane 4). The exosporium fraction was free of spores, as judged by light and electron microscopy (Fig. 6C and data not shown). Therefore, CotE is largely, if not entirely, restricted to the coat or the interspace. Given that CotE is a coat protein in B. subtilis (20, 94), our working model is that CotE is also a coat protein in B. anthracis.
|
|
|
In B. subtilis, coat proteins with significant roles in coat assembly usually also affect germination (16). In B. anthracis, conversion of the spore to the vegetative state is essential for infection (54). Therefore, we studied the possible involvement of B. anthracis CotE in germination (55, 74). To measure early events in germination, we used a calcium release assay (87) for the Sterne strain and a quantitative fluorescent dye uptake assay for both the Sterne and Ames strains (89) (since safety considerations make the use of radioactivity, needed for the calcium release assay, inappropriate for use with Ames). cotE mutant spores from the Sterne and Ames strains had significant delays in early germination events based on the dye uptake assay and significant, although much more modest, delays based on the calcium release assay (Fig. 9A and C). Differences in the sporulation media used in the two assays prevent direct comparisons of the data generated using different methods. To measure later germination events, we monitored the appearance of heat-sensitive (i.e., germinated) spores after resuspension in BHI medium. Even though it begins to appear early in germination, heat sensitivity can be an appropriate measure for later events, as the acquisition of complete heat sensitivity occurs relatively late (56, 74). cotE spores were partially germination defective, as measured by the loss of heat resistance during incubation in BHI medium (Fig. 10A and data not shown).
|
|
Germination in phagocytic cells is very likely to be important to anthrax pathophysiology (15, 31). We have argued previously that when spores germinate, they become sensitive to macrophage killing (88). Therefore, we addressed the question of whether the in vitro germination defect of cotE mutant spores (discussed above) affected germination in macrophages. After macrophage infection, we recovered more cotE mutant cells than wild type Ames cells, as would be anticipated for germination-defective spores (Fig. 10B). To further address whether the increase in the number of cotE mutant cells recovered from macrophages was due to a germination defect, we harvested bacteria from spore-infected macrophage cultures and measured germination by testing heat sensitivity. Surprisingly, we found that cotE spores survive in the macrophage even after losing heat resistance (i.e., after germination), as the total number of Ames strain cells recovered from the macrophage was greater than the number of heat-resistant (ungerminated) cells (Fig. 10B). This was true for the Sterne strain cotE mutant as well (data not shown). In contrast, and as expected, after infection by the wild-type Ames strain, the number of heat-resistant bacteria (i.e., ungerminated spores) was similar to the total number of cells recovered (Fig. 10B). Therefore, the germination defect in cotE mutant spores cannot fully explain their ability to survive in the macrophage. One possible explanation for the survival of these spores would be an inherent heat resistance defect of dormant cotE mutant spores (as opposed to a defect in germination per se). If this were true, heat resistance would not accurately reflect germination, since the dormant spores would be heat sensitive. We ruled out this possibility, however, as dormant cotE mutant spores do not have a heat resistance defect (data not shown). Given this, it is not immediately obvious how to account for our data, since we would expect heat-sensitive cells to have also shed the coat, making them vulnerable to macrophage lysozyme. We favor the possibility that coat shedding is delayed in cotE mutant spores, thereby allowing them to endure macrophage stress even after a loss of heat resistance.
We also detected a defect in cotE mutant spore germination in macrophages by using a stain that distinguishes dormant and germinating spores (91; data not shown). In the case of infection with wild-type Ames, after 1 h of incubation, most spores within macrophages were committed to germination, as indicated by a purple counterstain. In contrast, when macrophages were infected with cotE mutant spores, the majority of spores in macrophages stained with malachite green, indicating that germination had not initiated (data not shown). After 2 h of additional incubation, 62% of the intracellular cotE mutant spores were counterstained purple, compared to 99% of the intracellular wild-type (Ames) spores, consistent with the results described above (data not shown). Taken as a whole, these data indicate that both Sterne and Ames strain cotE mutants have a partial defect at an early stage in germination and, possibly, at a later stage. This could be due to changes in the coat, exosporium, or both.
Role of CotE in virulence. Previously, it was shown that the major exosporium protein BclA is dispensable for infection by the attenuated Sterne strain in an immunocompromised mouse model (80). The exosporium defect of cotE mutant spores allowed us to evaluate the role of all the exosporium components in disease by using a virulent strain. We sonicated cotE mutant Ames strain spores to completely remove any residual exosporium and tested their ability to cause infection in a Guinea pig intramuscular challenge model (35). Spores that were missing their exosporia were fully virulent in this model (Fig. 11A). Furthermore, taken together with the data described above, we conclude that modest defects in germination do not necessarily prevent disease.
|
Role of CotH in B. anthracis. In B. subtilis, the CotE-controlled protein CotH directs the assembly of at least seven coat proteins (39, 48, 57; M. Hahn and A. Driks, unpublished observations). The B. anthracis CotH homologue is 56% identical and 72% similar to the B. subtilis protein. We tested the role of CotH in B. anthracis spore function and assembly by inactivating its gene in the Sterne and Ames strains (resulting in strains RG73 and JAB3, respectively) and analyzing the resulting cells as we did for the cotE mutant. cotH mutant cells grew and sporulated normally and were indistinguishable from the wild type by light and thin-section transmission electron microscopy (data not shown). However, AFM analysis of the Sterne cotH mutant strain revealed that the coat surface ridges were largely absent (Fig. 1L). SDS-PAGE analysis revealed differences in the electrophoretic patterns of Coomassie-stainable bands from extracts of wild-type and cotH mutant spores in both the Sterne and Ames backgrounds (Fig. 5, lanes 6 to 9). In particular, bands of 43, 29, 24, and 13 kDa were reduced in intensity in mutant strain extracts. The 13-kDa band may be BAS2377, which is controlled by CotE (see above). In most analyses, we also detected increased amounts of the 116-, 92-, and 70-kDa bands in the Ames mutant. Because the changes in the cotH electropherogram were more severe than those seen for the cotE mutant, it is unlikely that CotH is a CotE-controlled protein in B. anthracis, as it is in B. subtilis (95).
We analyzed the ability of the cotH mutants to germinate, as described above. In contrast to the case of cotE, cotH mutants of either the Sterne or Ames background showed no significant defect by calcium release, dye uptake, or heat resistance (when germinated in BHI medium) (Fig. 9B and C and 10A and data not shown). However, we were able to detect differences in the cotH mutant strains in the following assays. First, we measured the ability of spores to resume metabolism after germination by using a modified tetrazolium assay (described above). cotH mutant strains (RG73 and JAB3) produced a color that was greater in intensity than that produced by the wild type (data not shown). This could occur if the percentage of cotH mutant spores that successfully germinate is higher than that of wild-type spores. Therefore, we germinated wild-type and cotH mutant cells (RG1 and RG73) (in triplicate) and monitored the refractility of spores by phase-contrast light microscopy (Fig. 9D). At the initiation of germination, all of the spores were phase bright, indicating a dehydrated, and therefore dormant, spore. After 15 min of germination, 7% of wild-type spores had become phase dark, and 94% of cotH mutant spores were phase dark, consistent with the results of the tetrazolium overlay assay and indicating that more cotH mutants spores germinated than wild-type spores. After 60 min, 47% of wild-type spores were phase dark, while 98% of the cotH mutant spores had become phase dark. We also examined the role of CotH in germination by assessing spore survival after uptake by macrophages. We found that fewer cotH mutant spores were recovered after macrophage uptake, consistent with an increase in cotH mutant spore germination (Fig. 10C). Overall, our measurements of early and intermediate events in germination by calcium release, dye uptake, or heat resistance suggest that these occur largely normally in cotH mutants, whereas the other assays, phase-dark transition and macrophage survival, indicate that more cotH mutant spores complete germination than do wild-type spores. Because we used the same germinant to detect the consequence of the cotH mutation when calcium release was analyzed, dye uptake and phase transition, our data suggest a defect in a relatively late stage in germination that results in an increase in the number of germinating spores. Regardless, the cotH mutation does not appear to have a dramatic effect on disease, as the cotH mutation had no effect on virulence in the Guinea pig intramuscular challenge virulence model (Fig. 11A).
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A model for assembly in B. anthracis. The results presented here allow us to describe the outlines of coat assembly in B. anthracis (Fig. 2E to H). In this view, the first critical event in B. anthracis coat assembly is the deposition of SpoIVA at the forespore surface (as is also the case in B. subtilis) (Fig. 2A and E). We expect this to be true in Clostridium spp. as well, since extremely close homologues of spoIVA are present in the genomes of the sequenced strains from this genus. Next, a shell of CotE forms (or begins to form) around the spore (Fig. 2F) in a SpoIVA-dependent manner, as in B. subtilis. CotE, in turn, directs the assembly of an electron-translucent structure that assembles at the mother cell pole of the forespore. The surface of this structure could serve as the site of exosporium assembly initiation. This early coat structure is reminiscent of the precoat in B. subtilis (20). Once coat and exosporium assembly begin, their formation proceeds essentially concomitantly, with the layers of the coat becoming evident, the exosporium forming a contiguous shell, and the eventual appearance of the hair-like projections (Fig. 2G). Soon after forespore engulfment, the close connection between the "precoat" and the exosporium apparently changes, resulting in variability in the size of the interspace around the circumference of the spore. We speculate that the spring-like quality of the interspace mentioned above appears at this time. The exosporium has previously been observed to be flexible in Bacillus mycoides (6), perhaps due to the spring-like nature of the interspace. We also infer that an as-yet-unidentified material spans the interspace to connect the coat and exosporium and that CotE attaches this material to the coat (Fig. 2H). Although no connecting material is evident by thin-section electron microscopy, we assume that it exists, as otherwise, the exosporium would not close around the spore during its assembly.
How CotE guides the exosporium around the forespore is unclear. Possible clues are provided by recent results showing that the spore proteins ExsY and CotY play important roles in exosporium assembly in B. anthracis and in B. anthracis and B. cereus, respectively (7, 37). Interestingly, ExsY and CotY are orthologues of the B. subtilis CotE-controlled protein CotZ (39). This raises the speculative possibility that CotE directs exosporium assembly at least partially through the control of ExsY and CotY. Notably, in B. anthracis exsY mutant sporangia, a fragment of exosporium is detected at the mother-cell-facing pole of the forespore (7). This suggests that the cotE mutant phenotype is more severe than that of exsY, consistent with the speculations mentioned above.
CotH plays a significant role in the deposition of certain coat proteins in both B. anthracis and B. subtilis. The observation that germination appears to be increased in the B. anthracis cotH mutant spores raises the possibility that CotH controls the assembly of germination-inhibitory proteins. We do not know how CotH assembly is controlled. However, the phenotypes of cotE and cotH mutants indicate that CotE and CotH assemble largely independently. The previously identified B. anthracis morphogenetic coat protein, Cot
, is also likely to assemble independently of CotE or CotH, given that the cot mutant coat assembly phenotype is more severe than that of a cotE or cotH mutant (40). However, phenotypic comparisons do not exclude the possibility that Cot directs CotH assembly.
The phenotypic differences between cotE and cotH mutants in B. anthracis and B. subtilis are reminiscent of the differences in the morphological consequences in mutations in safA in B. subtilis and its orthologue, exsA, in B. cereus (3). ExsA, which controls coat and exosporium assembly in B. cereus, appears to have a more severe effect on the B. cereus coat than does safA on the B. subtilis coat (60, 82).
By approximately hour 8 of sporulation, the exosporium has divided the mother cell into two compartments. Given that the exosporium is probably impermeable to proteins (27, 29, 42), it is reasonable to suggest that after hour 8, any assembly at the coat or the underside of the exosporium depends on proteins in the interspace. It might be argued that the number of ribosomes in this region is insufficient to synthesize sufficient amounts of coat proteins. However, we have proposed that coat proteins are synthesized in excess in B. subtilis (11). Possibly, a similar excess of coat proteins in B. anthracis facilitates continued assembly after closure of the exosporium.
CotE has a poorly characterized common role in both B. subtilis and B. anthracis: enforcing close contact between the coat and the cortex (4, 16). Possibly, CotE imparts an elastic property to the coat. In this view, as the forespore volume decreases (during core dehydration), the coat maintains contact by constricting in a CotE-dependent manner. In the absence of CotE, this elastic property is lost and the coat retains the larger diameter that it possessed prior to forespore dehydration. This perspective implies that the contacts that mediate the initial deposition of CotE at the forespore are of low affinity or otherwise lost as development proceeds. The elasticity could be a direct result of CotE or CotE-controlled proteins, or it could be an emergent property of the entire coat. An additional common role for CotE in B. subtilis and B. anthracis is in germination (94). It seems likely that the mechanistic basis of the impact of CotE on germination will differ between these species due to the different roles in coat assembly. However, in B. subtilis, CotE controls the assembly of CwlJ, which participates in a redundant pathway in later stages of germination (2, 61) and has two homologues in B. anthracis, as identified by BLAST analysis. Possibly, in B. anthracis, CotE controls CwlJ assembly as well.
Our data are consistent with the view that various Bacillus species use a conserved core of morphogenetic coat proteins in related but different ways. This raises the possibility that relatively modest numbers of mutations in the morphogenetic coat protein genes would permit significant changes in function and enable rapid adaptation to novel niches. This would be particularly useful to bacilli, given the diversity of environments that they inhabit (58, 66). From analyses of coat protein gene sequences of a variety of species, it may be possible to address this issue by measuring the degree of evolutionary pressure on key amino acids of morphogenetic coat proteins.
The exosporium is probably not a virulence factor. A principal result from this work is the finding that, at least in the context of a cotE mutation, the exosporium is dispensable for virulence as measured in two animal models. As the outermost surface of the B. anthracis spore, it is reasonable to imagine that the exosporium plays a significant role in disease. However, Sylvestre et al. (80) showed that a deletion of the gene encoding the major exosporium protein BclA from the Sterne strain had no significant effect on the LD50 in a mouse subcutaneous model of anthrax infection. Additionally, exosporia are found on spores of nonpathogenic species including B. megaterium and Bacillus odysseyi (45, 85), suggesting that their primary role need not be in disease. Nonetheless, the exosporium may detectably affect spore function. We found that cotE mutant spores are more resistant to killing by macrophages, likely as a result of less efficient germination. However, Kang et al. (38) showed that when the exosporium was removed from the Sterne strain by sonication, macrophage killing occurred more readily. Clarification of the role of the exosporium in spore resistance, disease, and maintenance in the environment will be an important future goal. Regardless of its role in the wild, the fact that the exosporium can be removed (mechanically [68] or genetically, as we have done in this work) and that the resulting spores are likely to be suitable for use as a weapon has important implications for the design of future detection, decontamination, and vaccine strategies involving the exosporium (26). Our view is that we should anticipate the potential deployment of spores without exosporia. Therefore, any future antispore vaccine and detection strategies, for example, should include both exosporial as well as coat surface antigens.
|
ACKNOWLEDGMENTS |
|---|
This work was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, National Research Council, 1996. The facility where this research was conducted is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.
This work was supported by grants GM53989 and AI53365 from the National Institutes of Health (NIH) (A.D.), grant AI053360 from the NIH (J.R.M.), and the U.S. Army Medical Research and Materiel Command under project numbers 02-4-5C-018 (J.B.), 02-4-5C-023 (S.W.), and 04-0-IL-002 (S.W.).
Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Army.
|
FOOTNOTES |
|---|
Published ahead of print on 17 November 2006.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
