Previous Article | Next Article ![]()
Journal of Bacteriology, September 2008, p. 6271-6275, Vol. 190, No. 18
0021-9193/08/$08.00+0 doi:10.1128/JB.00860-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Center for Structural Biology,1 Department of Microbiology and Immunology, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157,2 Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Michigan 481093
Received 24 June 2008/ Accepted 3 July 2008
|
|
|---|
|
|
|---|
In their transcriptional profiling of the B. anthracis life cycle, in which five distinct temporal waves of gene expression were identified from germination through sporulation, Bergman et al. (4) showed that the coaX, coaBC (BA4007), and coaD (BA4139) genes encoding the first three enzymes in the Pan
CoASH pathway (20) are upregulated in waves I and II. The coaX gene was also upregulated more than twofold between 1 and 2 h postinfection within host macrophages (3); in this respect (4), PanKBa may be particularly significant as a potential target for the design of therapeutically useful molecules. CoASH biosynthesis has very recently been reviewed as an antimicrobial drug target (31); crystal structures are now available for PanKBa (25) and the type III PanKs from Pseudomonas aeruginosa (15) and Thermotoga maritima (35, 36), and they include complexes with Pan and ADP and with the 4'-phosphopantothenate product. These have led to the identification of new motifs for the Pan-binding pocket and suggest, based on differences in the binding modes for both Pan and ATP substrates (relative to the type II human PanK) (14), potential modes of design for new inhibitors specifically targeting the type III enzymes.
As suggested by the absence of glutathione, CoASH also functions in the thiol-disulfide redox biology of B. anthracis (25); this scheme includes an NAD(P)H-dependent CoA-disulfide reductase (BACoADR) that maintains the reduced intracellular state of CoASH in vegetative cells via an enzymatic Cys42-SSCoA intermediate (34). This is quite distinct from the system described for B. subtilis, which has both type I and type III PanKs (6, 37) but lacks CoA-disulfide reductase (34). While the recent report of Hochgräfe et al. (13) supports the conclusion that S-thiolation by Cys represents a general, reversible mechanism for the protection of protein-SH groups during disulfide stress in B. subtilis, earlier studies with Bacillus megaterium linked CoA-disulfide reductase (32) with the reduction of spore protein-SSCoA (soluble proteins S-thiolated with CoASH) early in germination. These protein-SSCoA mixed disulfide forms account for
45% of the total CoASH in B. megaterium spores (30).
Bioinformatics analyses of the PanKBa locus (26) indicated that coaX might be linked with the hslO and cysK-1 genes (BA0066 and BA0067, respectively) in a transcriptional unit (Fig. 1A). The same coaX-hslO-cysK-1 clusters are conserved in some other Bacillus species, in Geobacillus kaustophilus, and in several strains of Listeria monocytogenes. In B. subtilis, the yacD gene (encoding a protein with limited sequence identity to the B. subtilis PrsA peptidyl-prolyl isomerase) (11) is inserted within the coaX cluster (Fig. 1A), between the hslO and cysK loci. The B. anthracis hslO and cysK-1 genes are expressed during waves II and III of the life cycle (4), and cysK-1 is also upregulated more than twofold between 1 and 2 h postinfection within host macrophages (Table 1) (3).
![]() View larger version (27K): [in a new window] |
FIG. 1. coaX is part of a tricistronic operon during the exponential growth of B. anthracis Sterne. (A) Genomic contexts (26) for coaX in B. anthracis Sterne (top) and B. subtilis (bottom). Primer sets 1 to 5 for endpoint RT-PCR (arrows) are indicated above the B. anthracis open reading frames and correspond as follows: 1, ftsH-coaX; 2, coaX-hslO; 3, hslO-cysK-1; 4, cysK-1 inner; and 5, cysK-1-pabB. (B) Endpoint RT-PCR showing that coaX is transcribed as part of a contiguous transcript during exponential growth. The gel shows contiguous products (at an A600 of 0.4 to 0.6) from primer sets 2 to 4 but not from pairs 1 or 5. Controls with genomic DNA and without RT were done for each experiment. (C) Partial sequence of the intergenic region between B. anthracis Sterne ftsH and coaX. Elements of the A-dependent promoter sequence (–35, TG, and –10) are underlined, as is the ribosome binding site. The coaX transcription start site, as determined by 5' RACE analysis, is indicated by a bent arrow; the start codon (ATG) is boxed.
|
|
View this table: [in a new window] |
TABLE 1. Bioinformatics of the B. anthracis coaX gene cluster
|
In order to test the suggestion that coaX, hslO, and cysK-1 are transcribed as a single mRNA unit in B. anthracis, we performed reverse transcriptase PCR (RT-PCR) (Fig. 1B) with RNA isolated from exponentially growing brain heart infusion (BHI) cultures of B. anthracis Sterne. These results revealed, as suggested by the bioinformatics analyses, that coaX is the first gene in a tricistronic operon. This operon assignment is also consistent with the results obtained with the operon prediction algorithm developed by Bergman et al. (5).
Rapid amplification of cDNA ends (RACE) analysis was performed with B. anthracis RNA in order to analyze the 5' region preceding the coaX coding sequence (Fig. 1C). The coaX transcription start site is located 54 bp upstream of the ATG initiation codon, and the –35 and –10 elements of the
A-dependent promoter sequence conserve three (CTGATT) and five (TATGAT) bases, respectively, of the consensus hexanucleotide sequences identified for B. subtilis (12). The two promoter elements are separated by a 16-base spacer region, which includes the TG dinucleotide motif at positions –16 and –15. The predicted ribosome binding site (AGTGG), deduced by comparison with the highly conserved pentanucleotide sequence described for B. subtilis genes (22), is also indicated in Fig. 1C. For the 12-bp region containing this motif (the spacing from GTGG to the ATG initiation codon is seven bases), eight bases are complementary to the 3' end of 16S rRNA, including one stretch of five consecutive bases (GGTGA).
Given that CysK-1 was identified within the B. anthracis cytoplasmic proteome (2), and in consideration of the longer 105-bp intergenic region separating hslO and cysK-1 (5), we examined the possibility that cysK-1 could be transcribed independently in B. anthracis. For the corresponding 5' RACE analysis, we prepared total RNA from the hslO::pBKJ236 mutant strain constructed by temperature-dependent gene disruption (see below). The chromosomal integration of the 6.3-kb plasmid (17) is expected to have a polar effect on the downstream transcription of cysK-1 from the coaX promoter, but it would not be expected to interfere with the monocistronic transcription of cysK-1 from its own promoter. RACE analysis with cysK-1-specific primers should yield only product from the predicted monocistronic transcript; there should be no tricistronic transcript from the mutant strain. However, analyses of RNA preparations from cultures harvested either at an A600 of 0.4 in BHI broth (known to give the tricistronic coaX-hslO-cysK-1 transcript) (Fig. 1B) or at an A600 of 1.0 in LB broth (known to give CysK-1 protein expression in B. anthracis UM23C1-2) (2) failed to provide evidence for any cysK-1 transcript (data not shown).
We were unable to construct a markerless deletion mutant of coaX in B. anthracis, suggesting that PanKBais essential. We therefore decided to construct a conditional mutant by placing coaX under the control of Pspac, following the temperature-dependent pNFd13 insertion mutagenesis method (9). PCR was used to assess the genomic structure of the coaX::pNFd13 mutant (COAXd) and the absence of replicating plasmid at 39°C; these results are shown in Fig. 2A. Unlike the markerless gene replacement protocol (17), the pNFd13 insertion mutagenesis method is expected to introduce polar effects on the downstream hslO and cysK-1 genes. We tested the possibility that the hslO and/or cysK-1 gene products might be essential for vegetative growth, using the temperature-dependent gene disruption protocol described above. Both hslO::pBKJ236 and cysK-1::pBKJ236 mutant strains grew on BHI-erythromycin plates; Fig. 3 shows the PCR results confirming the genomic structures of the respective chromosomal integrants. Neither hslO nor cysK-1 gene products are essential for B. anthracis growth under these conditions; the results presented for COAXd (see below) are solely attributable to the conditional disruption of the coaX gene and/or the stable suppressor mutation.
![]() View larger version (33K): [in a new window] |
FIG. 2. Construction of the COAXd mutant strain and the effect of IPTG addition on growth. (A) At the permissive temperature of 30°C, PCR analyses confirm the presence of the independently replicating pNFd13::coaX' plasmid (lane 2), the intact chromosomal copy of coaX (lane 3), and the coaX::pNFd13 integrant (lanes 4 and 5). At the restrictive temperature of 39°C, PCR analyses confirm the absence of both the independently replicating plasmid (lane 6) and the intact coaX locus (lane 7), as well as the presence of the chromosomal integrant (lanes 8 and 9). Lane 1 is a 1-kb ladder. The four primer combinations used, both for lanes 2 to 5 and for lanes 6 to 9, were NFd-FOR/NFd-REV, US-FOR/DS-REV (coaX), US-FOR/NFd-REV, and NFd-FOR/DS-REV. Sequences are available on request. (B) Effects of IPTG addition on the B. anthracis Sterne COAXd mutant strain. The COAXd strain, in which the expression of coaX is controlled by Pspac, was grown in the presence of 50 µM IPTG and diluted 40-fold into fresh prewarmed BHI broth with ( ) or without 50 µM IPTG. The wild-type strain ( ) is included as a control. At 4 h, different concentrations of IPTG were added (indicated by the arrow) to those cultures inoculated without IPTG. , COAXd without IPTG; , COAXd with 5 µM IPTG; , COAXd with 10 µM IPTG; and , COAXd with 50 µM IPTG. The COAXd(Su) suppressor mutant is responsible for growth observed in the absence of IPTG.
|
![]() View larger version (74K): [in a new window] |
FIG. 3. Construction of hslO and cysK-1 fusion mutants. Diagnostic PCR confirming the correct integrations of pBKJ236::hslO' (lanes 2 and 3) and pBKJ236::cysK-1' (lanes 4 and 5) into hslO and cysK-1 loci, respectively. Lane 1 is a 1-kb ladder. The respective primer combinations used were as follows (sequences available upon request): lane 2, US-FOR/BKJ-REV (hslO); lane 3, BKJ-FOR/DS-REV (hslO); lane 4, US-FOR/BKJ-REV (cysK-1); and lane 5, BKJ-FOR/DS-REV (cysK-1).
|
7 h after inoculation. In order to test the possibility that the slow growth observed from COAXd spores on solid LB-kanamycin media in the absence of IPTG could be due to a stable suppressor mutant [COAXd(Su)], a single colony from an overnight growth was used to inoculate a fresh plate. The confluent growth observed after an approximately 12-h incubation is very similar to that observed originally in the presence of IPTG and supports the conclusion that a stable suppressor mutation allows growth in the absence of IPTG in both liquid (Fig. 2B) and solid media. We considered two plausible explanations for the stable suppressor mutation, the first of which involves a possible gain-of-function mutation in the inactive B. anthracis homolog ("hypothetical protein BA2901"; an E value of 4e–45) (25) of the functional Staphylococcus aureus type II PanK (15, 19). However, attempts to complement a temperature-sensitive E. coli coaA mutant with the BA2901 expression plasmid were unsuccessful. This is consistent with our earlier report (25) that the purified recombinant BA2901 protein is inactive in the in vitro PanK assay; the BA2901 protein is likely an ATP-dependent small-molecule kinase, but Pan is not the acceptor substrate.
The second scenario considered, however, involves an altered expression mutation within the Pspac promoter employed in pNFd13 mutagenesis (9). We evaluated this suggestion by isolating multiple clones of COAXd, grown in the presence of IPTG, and of COAXd(Su), grown in the absence of IPTG. Representative sequence results for the Pspac regions from pNFd13, COAXd, and COAXd(Su) are shown in Fig. 4A. The pNFd13 and COAXd Pspac sequences (four independent COAXd clones) are identical. However, the sequences of four independent COAXd(Su) clones reveal a single-base substitution of adenine for guanine at base pair 5 (the lac operator base numbering system of Sadler et al. [28]) of the lac operator. That transcription of the wild-type coaX operon is indeed occurring in COAXd(Su) is demonstrated in Fig. 4B, where RT-PCR results using the same primers described for Fig. 1B are shown for COAXd and for COAXd(Su). The same tricistronic coaX transcripts are present in both cases.
![]() View larger version (30K): [in a new window] |
FIG. 4. The COAXd(Su) suppressor mutant carries a single-base substitution within the lac operator and constitutively produces the tricistronic coaX transcript. (A) Multiple alignment of partial sequences for the Pspac promoter regions of pNFd13 (Pspac), the COAXd conditional mutant grown in the presence of IPTG, and the COAXd(Su) suppressor mutant isolated in the absence of IPTG. Four individual colonies were analyzed for both COAXd and COAXd(Su), with internally identical results. The A-dependent promoter and lac operator sequences are boxed. The COAXd(Su) lac operator carries a guanine adenine mutation in position 5; this position is boxed and indicated with an asterisk. (B) Total RNAs were isolated from COAXd and COAXd(Su) as described for wild-type B. anthracis Sterne, but cultures were grown in the presence and absence, respectively, of IPTG and at 39°C. Samples were analyzed by endpoint RT-PCR using the primer sets described in Fig. 1B. The gels show contiguous coaX transcripts for both samples.
|
CoASH biosynthetic enzymes, PanK in particular, continue to be recognized as possible new antimicrobial targets that may provide for the development of therapeutically useful molecules (31). As distinguished from the type II human PanK (14), the type III PanKs share very limited sequence identity and exhibit distinct kinetic properties and substrate preferences (6, 15); they also reveal drastic differences in binding modes for both ATP and Pan substrates (36). Yang et al. (36) have concluded that these differences should be exploited in the development of new inhibitors specifically targeting PanKBa and other type III enzymes. This work provides evidence supporting the conclusion that the type III PanK is an essential enzyme in B. anthracis, thus contributing to its validation as a target for such development.
The findings, opinions, and recommendations expressed in this paper are those of the authors and not necessarily those of NIAID, SERCEB, NIH, DHS, DOE, or ORISE.
Published ahead of print on 18 July 2008. ![]()
|
|
|---|
-aminoacrylate intermediate and an enzyme-inhibitor complex. J. Biol. Chem. 282:23473-23481.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»