Ipdab, a Virulence Factor in Mycobacterium Tuberculosis
Total Page:16
File Type:pdf, Size:1020Kb
IpdAB, a virulence factor in Mycobacterium PNAS PLUS tuberculosis, is a cholesterol ring-cleaving hydrolase Adam M. Crowea, Sean D. Workmana,b, Nobuhiko Watanabea,b, Liam J. Worralla,b, Natalie C. J. Strynadkaa,b, and Lindsay D. Eltisa,c,1 aDepartment of Biochemistry and Molecular Biology, Life Sciences Institute, The University of British Columbia, Vancouver, BC, Canada, V6T 1Z3; bThe Center for Blood Research, Life Sciences Institute, The University of British Columbia, Vancouver, BC, Canada, V6T 1Z3; and cDepartment of Microbiology and Immunology, Life Sciences Institute, The University of British Columbia, Vancouver, BC, Canada, V6T 1Z3 Edited by William R. Jacobs Jr., HHMI and Albert Einstein College of Medicine, Bronx, NY, and approved February 23, 2018 (received for review October 2, 2017) Mycobacterium tuberculosis (Mtb) grows on host-derived choles- degradation (10) beginning with 3aα-H-4α(3′-propanoate)-7aβ- terol during infection. IpdAB, found in all steroid-degrading bacteria methylhexahydro-1,5-indanedione (HIP). Briefly, HIP is thioesterified and a determinant of pathogenicity, has been implicated in the to HIP-CoA, then subjected to a series of β-oxidative reactions to hydrolysis of the last steroid ring. Phylogenetic analyses revealed yield HIEC-CoA, in which rings C and D are still intact. It was that IpdAB orthologs form a clade of CoA transferases (CoTs). In a proposed that ring D is then hydrolyzed by EchA20, a member coupled assay with a thiolase, IpdAB transformed the cholesterol of the crotonase superfamily, to yield (R)-2-(2-carboxyethyl)-3- catabolite (R)-2-(2-carboxyethyl)-3-methyl-6-oxocyclohex-1-ene-1- methyl-6-oxocyclohex-1-ene-1-carboxyl-CoA (COCHEA-CoA) carboxyl-CoA (COCHEA-CoA) and CoASH to 4-methyl-5-oxo-octanedioyl- (Fig. 1). Incubation of COCHEA-CoA with a CoA thiolase, CoA (MOODA-CoA) and acetyl-CoA with high specificity (kcat/Km = FadA6, and IpdAB yielded a product presumed to be 4-methyl- 5.8 ± 0.8 × 104 M−1·s−1). The structure of MOODA-CoA was con- 5-oxo-octanedioyl-CoA (MOODA-CoA; Fig. 1) based on mass sistent with IpdAB hydrolyzing COCHEA-CoA to a β-keto-thioester, spectral data and the observation of MOODA accumulation in a thiolase substrate. Contrary to characterized CoTs, IpdAB exhibited a ΔfadE32 mutant (10). This last step indicates that either no activity toward small CoA thioesters. Further, IpdAB lacks the IpdAB or FadA6 is responsible for cleavage of steroid ring C. β Δ catalytic glutamate residue that is conserved in the -subunit of char- Further, ipdAB mutants in each of Mtb and two other mycolic BIOCHEMISTRY acterized CoTs and a glutamyl-CoA intermediate was not trapped acid-producing Actinobacteria, Mycobacterium smegmatis and during turnover. By contrast, Glu105A, conserved in the α-subunit Rhodococcus jostii RHA1 (RHA1), accumulated COCHEA-CoA of IpdAB, was essential for catalysis. A crystal structure of the when incubated with cholesterol (10), suggesting that COCHEA- IpdAB·COCHEA-CoA complex, solved to 1.4 Å, revealed that Glu105A CoA is the physiological substrate of IpdAB. However, IpdAB is positioned to act as a catalytic base. Upon titration with COCHEA- has been annotated as a CoA transferase (CoT), sharing 27% CoA, the E105AA variant accumulated a yellow-colored species amino acid sequence identity with glutaconate CoT (GCT) from (λmax = 310 nm; Kd = 0.4 ± 0.2 μM) typical of β-keto enolates. In Acidaminococcus fermentans, a heterotetrameric class I β-keto-CoT the presence of D2O, IpdAB catalyzed the deuteration of COCHEA- (11). It is unclear how a thiolase or a CoT might catalyze fission of CoA adjacent to the hydroxylation site at rates consistent with kcat. a cyclohexene ring. Based on these data and additional IpdAB variants, we propose a IpdAB has also been implicated in Mtb pathogenesis. Based on retro-Claisen condensation-like mechanism for the IpdAB-mediated transposon mapping studies, ipdAB was predicted to be essential hydrolysis of COCHEA-CoA. This study expands the range of known reactions catalyzed by the CoT superfamily and provides mechanistic Significance insight into an important determinant of Mtb pathogenesis. All steroid-degrading bacteria utilize IpdAB, a predicted CoA steroid catabolism | coenzyme A transferase | carbon bond cleavage | transferase (CoT) that has been implicated in the hydrolysis of a tuberculosis | enzymology carbon–carbon bond, an unprecedented reaction in CoTs. In Mycobacterium tuberculosis, IpdAB is required for degrading he causative agent of tuberculosis, Mycobacterium tuberculo- host cholesterol and virulence. We used a combination of X-ray Tsis (Mtb), was responsible for 1.8 million deaths in 2015, more crystallographic and biochemical studies to elucidate the mech- than any other infectious agent worldwide (1). Current treatments anism of IpdAB. Superposition of the IpdABMtb active site with are onerous and, with the emergence of drug-resistant strains such those of CoTs reveals distinct architectural features which, in as MDR- and XDR-TB, novel therapeutics are urgently needed. conjunction with the biochemical data, indicate that IpdAB cat- The virulence and persistence of Mtb, an intracellular pathogen, is alyzes a retro-Claisen-like ring-opening reaction. This reaction is predicated on its ability to catabolize host-derived lipids, including unique for a member of the CoT superfamily. This study provides cholesterol (2). Cholesterol catabolism is of burgeoning interest as insights into bacterial steroid catabolism and facilitates the devel- disruption of a number of cholesterol catabolic genes generates opment of potential antituberculosis therapeutics targeting IpdAB. attenuated or avirulent strains (3–5).Morerecently,asignificant Author contributions: A.M.C., S.D.W., N.W., L.J.W., N.C.J.S., and L.D.E. designed research; number of compounds selected for their ability to inhibit the in- A.M.C., S.D.W., and N.W. performed research; A.M.C., S.D.W., N.W., L.J.W., N.C.J.S., and tracellular growth of Mtb specifically inhibited cholesterol catab- L.D.E. analyzed data; and A.M.C., S.D.W., L.J.W., N.C.J.S., and L.D.E. wrote the paper. olism, suggesting that this process is a viable target for novel The authors declare no conflict of interest. antituberculosis therapeutics (6). However, the function of many This article is a PNAS Direct Submission. ∼ of the 80 cholesterol catabolic genes is poorly understood. Published under the PNAS license. In Mtb and other mycolic acid-producing Actinobacteria, choles- Data deposition: The atomic coordinates and structure factors have been deposited in the terol catabolism can be described as three processes: (i) β-oxidation Protein Data Bank, www.wwpdb.org (PDB ID codes 6CO6, 6CO9, 6COJ, and 6CON). of the alkyl side chain, (ii) oxygenolytic cleavage of rings A and 1To whom correspondence should be addressed. Email: [email protected]. β B, and (iii) -oxidation of rings C and D. For the first two pro- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. cesses, many of the enzymatic steps have been well described 1073/pnas.1717015115/-/DCSupplemental. (3, 7–9). We recently proposed a pathway for rings C and D www.pnas.org/cgi/doi/10.1073/pnas.1717015115 PNAS Latest Articles | 1of10 Downloaded by guest on October 2, 2021 CHOLESTEROL HIP 5OH-HIC-CoA HIEC-CoA O O O C D C D O A B O O HO O SCoA O SCoA O O- EchA20 O O * O ** CoAS 8 4 - IpdAB CoAS FadA6 8 1 O 10 8 4 - O 4 1 O- 1 O O O 10 CoASH AcCoA O O O O SCoA COCHEA-CoA MeDODA-CoA MOODA-CoA − Fig. 1. Cholesterol rings C and D degradation in Mtb. R and R′ represent CoAS or O . Carbon numbering for COCHEA-CoA, MeDODA-CoA, and MOODA-CoA used within the text is displayed. for the growth of Mtb in macrophages (12). Indeed, deletion revealed that IpdA homologs from steroid-degrading bacteria mutants of ipdAB in Mtb had highly reduced growth rates in formed a clade distinct from each of the three formed by the α macrophages (10). Interestingly, an ipdAB mutant in the related subunits of the class I CoTs, class I β-keto-CoTs, and class II horse pathogen Rhodococcus equi has been patented as a live CoTs, respectively (Fig. 2). IpdAs clustered most closely with the vaccine for use in foals (13). Finally, ΔipdAB mutants of Mtb, M. heterotetrameric class I β-keto-CoTs and, within the IpdA clade, smegmatis, and RHA1 did not grow on cholesterol and were un- orthologs from actinobacteria (blue) and proteobacteria (green) able to grow on other carbon sources in the presence of choles- formed subclades. Analysis of IpdBs with the CoT β subunits terol (10). This cholesterol-dependent toxicity correlated with revealed similar relationships except that PcaI, the β subunit of highly depleted CoASH levels suggests that disruption of ipdAB results in CoA sequestration. CoTs catalyze the reversible transfer of CoA from a CoA PCT thioester to a free carboxylate (14, 15). Three classes of CoTs YdiF have been identified, the first two of which belong to the same superfamily. In class I CoTs, CoA is transferred from the acyl CitC CftA Bsub group of the donor substrate to a free carboxylate of an acceptor ACT CitF Hpy ScoA substrate, typically a small organic acid (11, 16, 17). These en- ScoA SCT zymes utilize a ping-pong mechanism in which a conserved glu- 84 tamyl residue acts as a nucleophile to form a glutamyl-CoA 70 intermediate before CoA transfer to the acceptor (Fig. S1A) (15, 71 18). In all class I CoTs characterized to date, the conserved 49 0.4 glutamate occurs in the β subunit of the α2β2 enzymes or the 79 equivalent domain in the α4 enzymes, toward the back of the active site (14, 16, 18, 19).