|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 278, Issue 29, 26497-26504, July 18, 2003
Polymerization of Mycobacterial Arabinogalactan and Ligation to Peptidoglycan*![]() ová ![]() From the Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, Colorado 80523-1682
Received for publication, March 4, 2003 , and in revised form, April 28, 2003.
The cell wall of Mycobacterium spp. consists predominately of arabinogalactan chains linked at the reducing ends to peptidoglycan via a P-GlcNAc-( 13)-Rha linkage unit (LU) and esterified to a variety
of mycolic acids at the nonreducing ends. Several aspects of the biosynthesis
of this complex have been defined, including the initial formation of the LU
on a polyprenyl phosphate (Pol-P) molecule followed by the sequential addition
of galactofuranosyl (Galf) units to generate
Pol-P-P-LU-(Galf)1,2,3, etc. and Pol-P-P-LU-galactan, catalyzed by
a bifunctional galactosyltransferase (Rv3808c) capable of adding alternating
5- and 6-linked Galf units. By applying cell-free extracts of
Mycobacterium smegmatis, containing cell wall and membrane fragments,
and differential labeling with UDP-[14C]Galp and
recombinant UDP-Galp mutase as the source of
[14C]Galf for galactan biosynthesis and
5-P-[14C]ribosyl-P-P as a donor of [14C]Araf
for arabinan synthesis, we now demonstrate sequential synthesis of the simpler
Pol-P-P-LU-(Galf)n glycolipid intermediates followed by
the Pol-P-P-LU-arabinogalactan and, finally, ligation of the
P-LU-arabinogalactan to peptidoglycan. This first time demonstration of in
vitro ligation of newly synthesized P-LU-arabinogalactan to newly
synthesized peptidoglycan is a necessary forerunner to defining the genetics
and enzymology of cell wall polymer-peptidoglycan ligation in
Mycobacterium spp. and examining this step as a target for new
antibacterial drugs.
The cell envelope of Mycobacterium tuberculosis is composed of a conventional plasma membrane and a cell wall proper unique to some genera within the Actinomycetales order, consisting of a core of arabinogalactan (AG),1 peptidoglycan (PG), and mycolic acids interspersed with a variety of free lipids, lipoglycans, and proteins (1); there is also evidence for polysaccharides on the outer face of the cell wall (2). The mycolic acids are attached to the nonreducing ends of the arabinogalactan, whereas the reducing ends are covalently attached to the cross-linked peptidoglycan via phosphoryl-N-acetylglucosaminosyl-rhamnosyl linkage units (P-GlcNAc-Rha). This massive structure, the mycolate-arabinogalactan-peptidoglycan-complex (MAPc), is the basis of many of the physiological and pathogenic features of M. tuberculosis and the site of susceptibility and resistance to many of the anti-tuberculosis drugs (3).
Biosynthesis of this complex commences with attachment of the residues of the linkage unit, GlcNAc-1-P and Rha, donated by UDP-GlcNAc and dTDP-Rha, respectively, to a polyprenyl phosphate (Pol-P) carrier lipid (4). Formation of the linkage unit is followed by the sequential addition of galactofuranosyl (Galf) units donated by UDP-Galf, to provide simple Pol-P-P-linked AG intermediates (4). The bulk, if not all, of galactan biosynthesis is catalyzed by a membrane-associated bifunctional galactosyltransferase capable of adding the alternating 5- and 6-linked Galf units (5, 6). The demonstration that the direct donor of the arabinofuranosyl (Araf) units of the cell wall core is decaprenyl-P-Araf (7) and that 5-P-ribosyl-PP (PRPP) is a precursor of decaprenyl-P-Araf (8) now provides us with the means to characterize the subsequent polymerization steps in AG biosynthesis and the final ligation of the AG lipid-linked intermediates to PG to generate the fully formed cell wall core.
Preparation of UDP-Galp Mutase, dTDP-Rha, P[14C]RPP, and UDP-MurNAc-L-Ala-D-Glu-meso-DAP-D-Ala-D-Ala (UDP-MurNAc-pentapeptide)Escherichia coli BL21 (DE3) (Stratagene, Cedar Creek, TX) was transformed with plasmid pORF6 containing Rv3809c as described (5). The recombinant UDP-Galp mutase was prepared and assayed as described (5); the concentration of protein in the final preparation was 2.0 mg per 100 µl. dTDP-Rha and P[14C]RPP were prepared from dTDP-Glc (9) and D-[U-14C]glucose (8), respectively, and were generous gifts from Dr. M. R. McNeil (Colorado State University). For the synthesis of UDP-MurNAc-pentapeptide, UDP-MurNAc was first prepared by a two-step coupled enzymatic conversion of UDP-GlcNAc to UDP-MurNAc (10) and identified through negative ion fast atom bombardment mass spectrometry as follows. The recombinant E. coli MurC, MurD, MurE, and MurF were overexpressed and purified from E. coli ER2566 as described (11) but using the Impact CN system (New England Biolabs) following the manufacturer's instructions. The purified enzymes were dialyzed extensively against 50 mM Tris-HCl (pH 8.0) containing 10 mM MgCl2 and 10% glycerol (v/v), and aliquots were stored at 80 °C. Reaction mixtures (30 ml) containing UDP-MurNAc (250 µM), L-Ala, D-Glu, DAP, D-Ala-D-Ala (1 mM each), TAPS (50 mM, pH 8), MgCl2 (5 mM), ATP (2.5 mM), and MurC, MurD, MurE, and MurF (75 µg/ml each) were incubated at 30 °C overnight and deproteinated by ultrafiltration, and the filtrate was loaded on a 10-ml Q-Sepharose (Amersham Biosciences) column equilibrated with 20 mM ammonium acetate. The bound material was eluted with a 201000 mM gradient of ammonium acetate. Fractions were monitored at A262 for the presence of UDP-containing compounds. Fractions containing UDP-MurNAc-pentapeptide were identified by TLC on silica gel plates in 2-butyric acid/1 M NH4OH (5:3), utilizing UV absorption and ninhydrin for detection. These fractions were pooled and lyophilized to remove buffer, and the final product, UDP-MurNAc-pentapeptide, was analyzed by mass spectrometry as described (11, 12). In most syntheses, the rate of conversion of UDP-MurNAc to UDP-MurNAc-pentapeptide was about 80%. Preparation of Chalaropsis MuramidaseChalaropsis sp. ATCC 16003 (American Type Culture Collection, Manassas, VA) was grown at 25 °C in medium consisting of glucose at 40 g/liter and peptone at 10 g/liter for 5 days (13, 14). The secreted muramidase was adsorbed from crude culture filtrates with Amberlite CG-50-H+ (Sigma) buffered at pH 5.0; protein was eluted from the matrix with 0.5 M ammonium acetate and the muramidase was precipitated with ammonium sulfate at 70% saturation (13, 14). The precipitate was redissolved in 10 mM ammonium acetate (pH 6.5). After dialysis to remove residual ammonium sulfate, the sample was passed over a Sephadex G-75 column (Amersham Biosciences), and fractions containing muramidase activity (measured by the reduction in A610 of Staphylococcus aureus whole cell suspension (13, 14)) were pooled. Purity was checked by SDS-PAGE, and the enzyme preparation showed a single band in SDS-PAGE gels stained with Coomassie Brilliant Blue R250. Yield from 10 liters of culture was about 200 mg of enzyme. One unit of enzyme was defined as the amount of enzyme that decreased the A610 of a S. aureus cell suspension at a rate of 0.008 OD/min. Preparation of an Enzymatically Active Cell Envelope Fraction from M. smegmatisM. smegmatis mc2155 cells were grown in nutrient broth to midlog phase (4), harvested, and stored at 70 °C until required. Approximately 8 g of bacteria (wet weight) were washed with a buffer containing 50 mM MOPS (pH 8.0), 5 mM 2-mercaptoethanol, and 10 mM MgCl2 (buffer A), resuspended in 24 ml of buffer A at 4 °C, and subjected to probe sonication as described (4). The sonicate was centrifuged at 27,000 x g for 15 min at 4 °C, and the pellet, containing the cell envelope, was resuspended in buffer A to a final volume of 16 ml. Percoll was added to achieve a 60% suspension, and the mixture was centrifuged at 27,000 x g for 60 min at 4 °C. The particulate, upper band was collected, washed twice with buffer A, resuspended in buffer A to a protein concentration of 1520 mg/ml, and used as the enzyme source in all experiments. Reaction Mixtures for [14C]Gal Labeling and Fractionation of Reaction ProductsThe basic reaction mixtures for assessing [14C]Gal incorporation into lipid-linked AG precursors were prepared as follows. UDP-[U-14C]Galp (1 µCi; 3.5 nmol; 289 mCi/mmol; PerkinElmer Life Sciences) was dried under a stream of N2, dissolved in 38 µl of buffer A, and incubated with 2 µl of the UDP-Galp mutase preparation (0.13 mg of protein) at 37 °C for 15 min. Other reagents and buffer A were added to yield a final volume of 320 µl containing a 10.8 µM mixture of UDP-[U-14C]Galp and UDP-[U-14C]Galf, 60 µM UDP-GlcNAc, 20 µM dTDP-Rha, 100 µM ATP, and the envelope enzyme fraction (2 mg of protein). The reaction mixtures were incubated at 37 °C for the indicated period of time. In some cases, 60 µM PRPP and/or 200 µM UDP-MurNAc-pentapeptide were also included in the reaction mixtures. In the case of the ligation assays, these [14C]Gal labeling reaction mixtures containing both PRPP and UDP-MurNAc-pentapeptide were incubated at 28 °C for appropriate periods. After incubation, reaction mixtures were extracted with CHCl3/CH3OH (2:1), the resultant pellet was washed thoroughly with 0.9% NaCl and extracted with CHCl3/CH3OH/H2O (10:10:3), followed by "E-soak" (water/ethanol/diethyl ether/pyridine/ammonium hydroxide; 15:15:5:1:0.017) (15) as described (16). The CHCl3/CH3OH (2:1) extract was partitioned with water (17). The backwashed lower (organic) phase was dried under a stream of N2, and the residue was dissolved in 200 µl of CHCl3/CH3OH/H2O/NH4OH (65:25:3.6:0.5) prior to liquid scintillation counting and analysis by TLC. In order to obtain a completely insoluble residue, rich in MAPc, the E-soak insoluble pellet was extracted three times with boiling 60% methanol containing 0.1% ammonium hydroxide.
To examine product-precursor relationships between lipid-linked
intermediates and the insoluble residue, [14C]Gal-labeled
CHCl3/CH3OH/ H2O (10:10:3)-soluble
lipid-linked polymers were synthesized using the basic reaction conditions for
[14C]Gal labeling described above. These enzymatically synthesized
[14C]Gal-labeled compounds (
Labeling of the Arabinan Component of AG with
P[14C]RPPThe basic reaction mixture contained 3.3
µM P[14C]RPP ( Analysis of the Insoluble ResidueThe insoluble residue, enriched in MAPc, was subjected to base treatment with 2 ml of 0.5% KOH in methanol for 4 days at 37 °C with gentle stirring. After washing three times with methanol, the methyl esters of the mycolic acids were removed with two diethyl ether extractions. The residual pellet was dried under N2 and digested with 100 µg/ml of Proteinase K (Roche Applied Science) in 250 µl of 10 mM sodium acetate (pH 7.5) at 37 °C overnight. Radioactivity released into the supernatant from the insoluble pellet by Proteinase K treatment was quantitated by liquid scintillation counting and subjected to sugar analysis as described below. After washing with 10 mM sodium acetate buffer, the residual pellet was treated with 2.5 units of purified Chalaropsis muramidase in 250 µl of 10 mM sodium acetate (pH 5.0), 500 units/ml of lysozyme in 10 mM Tris-HCl buffer (pH 7.5), or Proteinase K at 37 °C overnight. Aliquots of radiolabeled materials solubilized by these treatments were subjected to liquid scintillation counting and sugar analysis. AnalysisIn order to facilitate size exclusion chromatography of the polyprenyl-P-linked polymers, the CHCl3/CH3OH/H2O (10:10:3)-soluble, E-soak-soluble, and Chalaropsis muramidase-solubilized materials were hydrolyzed in mild acid as follows to selectively cleave the prenyl phosphate. Samples were suspended in 50 µl of 1-propanol by bath sonication, followed by 100 µl of 0.02 N HCl, and the resulting mixture was incubated for 30 min at 60 °C (18, 19). After neutralization with 10 µl of 0.2 N NaOH, the released water-soluble products were applied to a Biogel P-100 column (1 x 118 cm), equilibrated, and eluted with 100 mM ammonium acetate (pH 7.0). SDS-PAGE analysis of enzymatically radiolabeled products was done using Novex® 1020% Tricine gels (Invitrogen) under conditions recommended by the manufacturer. After electrophoresis, samples were blotted to nitrocellulose membranes, which were dried at room temperature, and subjected to autoradiography. CHCl3/CH3OH (2:1)-soluble materials were analyzed on silica gel TLC plates developed in CHCl3/CH3OH/NH4OH/1 M ammonium acetate/H2O (180:140:9:9:23), which were then subjected to autoradiography. For [14C]sugar analysis, samples were subjected to acid hydrolysis in 2 M CF3COOH for1hat120 °C. Hydrolysates were analyzed on silica gel TLC plates (silica gel G60, aluminum-backed; EM Science, Gibbstown, NJ), developed in pyridine/ethyl acetate/glacial acetic acid/water (5:5:1:3), and autoradiography. Radioactive spots were identified by comparative chromatography with standard sugars. Protein concentrations were estimated using the BCA protein assay reagent (Pierce).
Synthesis of Polyprenyl-P-linked IntermediatesWe previously established a cell-free assay system using membranes from M. smegmatis for the synthesis of the simple Pol-P-P-GlcNAc, Pol-P-P-GlcNAc-Rha, and Pol-P-P-GlcNAc-Rha-(Galf)14 intermediates in AG biosynthesis (4). The evidence for the nature of these products was based on solubility in organic solvents, susceptibility to mild acid hydrolysis, the presence of the appropriate radiolabeled sugar, and pulse-chase experiments (5). These experiments led to the identification of one of the galactosyltransferases involved in galactan synthesis, a bifunctional enzyme capable of adding the majority of the alternating 5- and 6-linked Galf units (5, 6). In the present study, we modified this cell-free system in order to demonstrate the sequential synthesis of the simpler glycolipid intermediates, followed by polyprenyl-P-P-linked galactan and arabinan intermediates, and in vitro ligation of these lipid-linked arabinogalactan intermediates to PG. The cell wall-membrane fraction of M. smegmatis, the source of endogenous glycosyltransferases and polyprenyl-P, was supplemented with UDP-GlcNAc, dTDP-Rha, the precursors of LU, and UDP-[14C]Galp and UDP-Galp mutase as the source of the Galf units of galactan. Reaction products were extracted with the organic solvents, CHCl3/CH3OH (2:1), CHCl3/CH3OH/H2O (10:10:3), and "E-soak" (15), and finally with boiling 60% methanol containing 0.1% ammonium hydroxide to remove residual soluble material, providing the insoluble MAPc-containing cell wall core. The incorporation of [14C]Galf from UDP-[14C]Galp into these four fractions is shown in Table I. Analysis of the CHCl3/CH3OH (2:1)-, CHCl3/CH3OH/H2O (10:10:3)-, and E-soak-soluble materials for 14C-labeled sugars showed that [14C]Gal was the sole radioactive sugar component (data not shown). TLC analysis of CHCl3/CH3OH (2:1)-soluble materials revealed a hierarchical array of glycolipids previously identified (5) as polyprenyl-P-P-GlcNAc-Rha-Galf, polyprenyl-P-P-GlcNAc-Rha-(Galf)2, and polyprenyl-P-P-GlcNAc-Rha-(Galf)3,4 (a mixture of tri-Galf- and tetra-Galf-containing glycolipid intermediates) (Fig. 1). No further simple glycolipid intermediates were observed in the more polar CHCl3/CH3OH/H2O (10:10:3) extract; the bulk of its radioactivity remained at the origin of the TLC plate (results not shown), supporting the evidence that this fraction contained the polyprenyl-P-P-GlcNAc-Rha-AG intermediates (5) (see below). The addition of exogenous PRPP and UDP-MurNAc-pentapeptide, precursors of arabinan and peptidoglycan synthesis, respectively, stimulated incorporation of [14C]Gal into the MAPc-containing residue in an additive manner with a concomitant reduction of radioactivity in the CHCl3/CH3OH/H2O (10:10:3) and E-soak extracts (Table I), suggesting that the polyprenyl-P-P-GlcNAc-Rha-(Gal)14 and the polyprenyl-P-P-GlcNAc-Rha-AG intermediates in these extracts were precursors of the mature PG-bound AG. However, the increase in insoluble material is not fully matched by a concomitant decrease in the other fractions. There is a substantial loss of radioactivity from the CHCl3/CH3OH/H2O (10:10:3) and E-soak extracts in incubations containing the additional precursors UDP-MurNAc-pentapeptide and PRPP. This is presumably due to a shortage of endogenous lipid carrier, required for de novo PG and decaprenyl-P-Araf synthesis when the UDP-MurNAc-pentapeptide and PRPP precursors are added.
Arabinan Polymerization Steps in AG BiosynthesisTo define the steps leading to the synthesis of the arabinan component of AG, cell-free reactions containing P[14C]RPP as the ultimate precursor of Araf were prepared in bulk and extracted with CHCl3/CH3OH (2:1), CHCl3/CH3OH/H2O (10:10:3), and E-soak. Parallel reactions containing UDP-[14C]Galp were run, and similar extracts were prepared. Complete acid hydrolysis and TLC analysis for radioactive sugar showed that all of the [14C]Gal remained as such, and the majority of the P[14C]RPP radiolabel was converted into [14C]Ara-containing material; a minority appeared as [14C]ribose, apparently from intermediates of an unidentified riban (8). TLC of the CHCl3/CH3OH (2:1)-soluble products showed a preponderance of polyprenyl (C50)-P-Araf in this fraction (7) (Fig. 1). Mild acid hydrolysis of the [14C]Ara-labeled CHCl3/CH3OH/H2O (10:10:3)-soluble and E-soak-soluble lipid polymers to remove the presumed polyprenyl-P and subsequent gel filtration showed considerable overlap but not complete coincidence in the profiles of these two sets of [14C]Ara-containing polymers (Fig. 2). Profiles were similar to those of the [14C]Gal-labeled polymers, labeled, released, and extracted under similar conditions (Fig. 2). In both cases, the E-soak-extractable material appeared to be slightly but reproducibly larger than that extractable with CHCl3/CH3OH/H2O (10:10:3), pointing to the presence of a population of polyprenyl-P-linked AG intermediates, partially resolvable by the two extractants. These lipid-linked intermediates were also analyzed by Tricine SDS-PAGE (Fig. 3). Overnight exposure of autoradiograms revealed a population of [14C]Gal-labeled CHCl3/CH3OH/H2O (10:10:3)- and E-soak-soluble lipid-linked polymers, whereas no images were seen in lanes containing [14C]Ara-labeled lipid-linked polymers, presumably due to lower labeling efficiency when using P[14C]RPP as the precursor. However, after 14 days of exposure, the [14C]Ara-containing CHCl3/CH3OH/H2O (10:10:3)- and E-soak-soluble lipid-linked polymers were also visible on the autoradiograms and showed a degree of heterogeneity similar to that of the [14C]Gal-labeled material. The E-soak soluble [14C]Ara- and [14C]Galf-containing lipid polymers again appeared to be larger than the CHCl3/CH3OH/H2O (10:10:3)-soluble material, supporting the trend seen in the size exclusion analysis. Analysis for radioactive sugar content in these extracts confirmed the sole presence of [14C]Ara and [14C]Gal in the respectively labeled polymers.
Evidence for in Vitro Ligation of AG to PGThe basic cell-free systems capable of catalyzing the synthesis of lipid-linked polymer intermediates of AG do not allow appreciable transfer of intermediates from the Pol-P carrier to PG. However, the addition of UDP-MurNAc-pentapeptide as a precursor of PG synthesis and longer incubation times resulted in linear incorporation of radioactivity into the insoluble residue over 16 h. Interestingly, incorporation of radioactivity into the E-soak-soluble fraction reached a plateau after 1 h (Fig. 4). Size exclusion, Tricine SDS-PAGE, and sugar analysis of the CHCl3/CH3OH/H2O (10:10:3)- and E-soak-soluble products from 16-h incubations revealed that lipid-linked [14C]Gal-labeled polymers similar to those generated in the shorter incubation periods described above had been synthesized. Thus, incubation times up to 16 h were used for further characterization of this first time demonstration of in vitro ligation of cell wall polysaccharide to PG.
Previously, we had demonstrated that Pol-P-P-[14C]GlcNAc and Pol-P-P-[14C]GlcNAc-Rha are precursors of more glycosylated versions of lipid-linked polymer intermediates that are soluble in CHCl3/CH3OH/H2O (10:10:3) and E-soak (5). Experiments in which [14C]Gal-prelabeled CHCl3/CH3OH/H2O (10:10:3)-soluble lipid-linked polymers were incubated with fresh enzyme and cold precursors for AG and PG synthesis were conducted. A comparison of radioactivity distributed among the three fractions showed that a linear decrease in the amount of [14C]Gal-labeled CHCl3/CH3OH/H2O (10:10:3)-soluble compounds (i.e. the lipid-linked polymers) was accompanied by an increase in the radioactivity found in both the E-soak- and the MAPc-containing residue (Fig. 5). The amount of radioactivity lost from the CHCl3/CH3OH/H2O (10:10:3) fraction was approximately equivalent to that appearing in the other fractions, indicating that the [14C]Gal-labeled CHCl3/CH3OH/H2O (10:10:3)-soluble precursors were converted into larger, more heavily glycosylated and insoluble products. Tricine SDS-PAGE analysis showed a size shift from CHCl3/CH3OH/H2O (10:10:3)-soluble precursor to E-soak-soluble products similar to that seen in Fig. 3. A small amount of radioactivity (0.4%) was also found in the CHCl3/CH3OH (2:1) fraction, suggesting that some degradation of the CHCl3/CH3OH/H2O (10:10:3)-soluble lipid-linked polymers had occurred.
Sugar analysis of the insoluble MAPc-containing residue revealed the
presence of both [14C]Gal and [14C]Glc, indicating that
some randomization of the radiolabel had occurred over the long incubation
period and that both sugars had been incorporated into that fraction
(Fig. 6). After base treatment
of the final residue to remove mycolic acids, the pellet was treated with
Proteinase K, which resulted in 6065% solubilization of radiolabel
without apparent loss of the volume of the pellet. Sugar analysis of the
solubilized radioactive compounds revealed [14C]Gal and
[14C]Glc in similar amounts. A second treatment with Proteinase K
did not solubilize more of the remaining radioactivity. However, when the
remaining insoluble residue was subjected to treatment with purified
Chalaropsis muramidase, an enzyme known to hydrolyze the
Synthesis of AG begins with the linkage unit, in a manner analogous to that of the cell wall teichoic acids of Gram-positive bacteria (21). GlcNAc-1-P is transferred to Pol-P from UDP-GlcNAc catalyzed by an as yet unidentified GlcNAc-1-P phosphotransferase; the closest homolog of wecA (formerly rfe) is Rv1302 (Fig. 8) (4, 5). This event is sequentially followed by the addition of Rha donated by dTDP-Rha, which is catalyzed by the rhamnosyl transferase Wbbl (Rv3265c) (22),2 and the addition of Galf residues donated by UDP-Galf. The only galactofuranosyl transferase recognized to date, Rv3808c, is reported to be a bifunctional enzyme capable of adding alternating 5- and 6-linked Galf residues (5, 6) and is probably responsible for the synthesis of bulk galactan; whether it is responsible for the synthesis of all of the galactan, especially the initial units, is not clear. The products of the emb operon were originally implicated in the transfer of the D-Araf units from the decaprenyl-P-Araf donor to the growing polymer (23); however, mutants in which embA and embB were inactivated by homologous recombination showed a selective deletion of the terminal -D-Araf-(1 2)- -D-Araf
extensions from the 3-position of the terminal, branching 3,5-linked
-D-Araf residues of AG
(24), suggesting a role beyond
simple arabinofuranosyl group transfer. Employing differential labeling with
UDP-[14C]Gal and P[14C]RPP, as the sources of galactan
and arabinan, respectively, we have now defined new aspects of the gross
polymerization steps leading to the synthesis of AG and MAPc. Specifically,
arabinan is added to galactan at the polyprenyl-linked stage, prior to
ligation to PG, and the polyprenyl-linked AG polymers are intermediates in the
synthesis of MAPc (Fig. 8).
However, the evidence that these
CHCl3/CH3OH/H2O (10:10:3)- and E-soak-soluble
intermediates are polyprenyl-P-linked is based on lipid solubility and
susceptibility to mild acid hydrolysis
(5), and is not unequivocal.
Thus, the possibility of other intermediary steps, including transfer of
intermediates to different carriers in the later stages of AG assembly, is
possible.
Most significantly, we demonstrated ligation of the polyprenyl-linked [14C]Gal-labeled AG to PG in a cell-free system. The ligation of AG to PG involves the transphosphorylation of the terminal GlcNAc-1-P of the LU of the AG polymer from its polyprenyl-P carrier to the 6-position of the N-glycolylmuramic acid residues of PG (25). The reaction closely resembles the ligation of teichoic acid and other anionic polymers to PG in a wide range of Gram-positive bacteria and is a prime candidate for antibiotic intervention in the case of Staphylococcus and Streptococcus infections (21). In Bacillus subtilis 168, the attachment of cell wall teichoic acid or teichuronic acid to PG has been demonstrated in vivo and required the simultaneous synthesis of both polymers (26). This ligation has also been achieved in toluenized cells of B. subtilis W23 under conditions that reduced cell wall autolytic activity and was independent of de novo PG synthesis (27). Using cell wall-membrane preparations, Ward et al. (2830) demonstrated the ligation of teichuronic acid to PG by formation of a phosphodiester bond between the reducing GalNAc terminus of the teichuronic acid and the 6-hydroxyl groups of muramic acid residues in the glycan chain of PG; a linkage unit was not involved, in contrast to that found for teichoic acids. Concomitant synthesis of both polymers was apparently necessary (30).
In general, the criterion for ligation of Gram-positive cell wall anionic
polymers to peptidoglycan was insolubility in 4% boiling SDS
(2730).
However, this treatment precludes the subsequent use of muramidase,
susceptibility to which is clearly diagnostic for ligation of a target polymer
to PG. Therefore, in the present work, extraction with refluxing 60% methanol
containing 0.1% ammonium hydroxide was introduced; this treatment appeared to
be more stringent than SDS extraction in that 60% more radioactivity was
solubilized. It should be noted that Proteinase K treatment liberated
compounds containing both [14C]Glc and [14C]Gal from the
final solvent-insoluble pellet. Obviously, the long incubation times required
for ligation resulted in epimerization of UDP-[14C]Gal to
UDP-[14C]Glc and dispersal of the [14C]Glc. This
observation suggests the association of glucosylated and galactosylated
proteins with the cell wall core. Alternatively, the source may be residual
glycans, which have been identified on the surface of mycobacteria
(28), or glycogen storage
granules (29), although the
mechanism by which Proteinase K mediates the release of this type of material
is not clear. Following Proteinase K digestion, the remaining radioactivity
associated with the insoluble residue was successfully solubilized by
Chalaropsis muramidase treatment. Chalaropsis muramidase is
a specific enzyme that hydrolyzes the The ligation reaction is clearly complex, requiring the admixture of enzymes from different cellular compartments present in wall-membrane preparations. Although there has been no detailed study of the organization of cell wall synthetic enzymes at the molecular level in Gram-positive bacteria or mycobacteria, diverse investigations of Gram-positive bacteria (21, 26, 3336) lead to the conclusion that the cytoplasmic membrane contains ordered assemblies of the enzymes of polymer and PG synthesis together with shared anchor lipids. Attachment of PG and anionic polymer almost certainly occurs at the outer surface of the membrane. However, in the case of mycobacteria, membranes alone will not accomplish the task. Despite the availability of the genome of B. subtilis, the gene(s) involved in teichoic acid ligation remain unidentified, although tagA and tagB were regarded as candidates at one time (37). It should be noted that mycolic acids were removed at the first step in analysis of the final residue from the M. smegmatis incubations in order to facilitate effective digestion with Chalaropsis muramidase. Thus, it remains to be seen whether they are attached to AG before or after ligation to PG. However, recent observations through whole cell labeling with different precursors suggest that mycolylation of arabinan termini follows ligation of AG to PG (38). The previous identification of a bifunctional membranous galactofuranosyl transferase, the fact that the embA-C operon may encode the capability of both Araf addition and polymerization, the present development of cell-free AG polymerization, and subsequent ligation assays now set the stage for the complete enzymatic definition of AG-PG formation and its genetic basis.
* This work was supported by NIAID, National Institutes of Health, Grants AI-18357 and AI-46393 (to P. J. B.) and AI-49151 (to D. C. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed. Tel.: 970-491-6700; Fax: 970-491-1815; E-mail: Patrick.Brennan{at}ColoState.edu.
1 The abbreviations used are: AG, arabinogalactan; Galf,
galactofuranose; Araf, arabinofuranose; Galp,
galactopyranose; GL, glycolipid; LU, linker unit; MOPS,
4-morpholinepropanesulfonic acid; PG, peptidoglycan; PRPP,
5-phosphoribose-pyrophosphate; Rha, rhamnose; UDP-MurNAc-pentapeptide, uridine
diphosphoryl-N-acetylmuramate-L-Ala-D-Glu-meso-DAP-D-Ala-D-Ala;
Tricine, N-[2-hydroxy-1,1-bis(hydoxymethyl)ethyl]glycine; MAPc,
mycolate-arabinogalactan-peptidoglycan-complex; Pol-P, polyprenyl phosphate;
MurNAc, N-acetylmuramic acid; TAPS,
3-{[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino}-1-propanesulfonic acid; DAP,
2,6-diaminopimelic acid.
2 The cell wall arabinogalactan linker formation enzyme, dTDP-Rha: GlcNAc
1-diphosphoryl polyprenol rhamnosyltransferase, is essential for mycobacterial
viability (J. A. Mills, K. Motichka, M. Jucker, H. P. Wu, B. C. Uhlic, R. J.
Stern, M. S. Scherman, V. D. Vissa, W. Yan, M. Kundu, and M. R. McNeil,
unpublished data).
We thank Dr. Michael R. McNeil, Michael S. Scherman, and their colleagues for preparation of materials and helpful discussions.
This article has been cited by other articles:
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||