Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schertzer, J. W. Articles by Brown, E. D. Search for Related Content PubMed PubMed Citation Articles by Schertzer, J. W. Articles by Brown, E. D.

 Previous Article  |  Next Article 

Journal of Bacteriology, November 2008, p. 6940-6947, Vol. 190, No. 21
0021-9193/08/$08.00+0     doi:10.1128/JB.00851-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Use of CDP-Glycerol as an Alternate Acceptor for the Teichoic Acid Polymerase Reveals that Membrane Association Regulates Polymer Length

Jeffrey W. Schertzer and Eric D. Brown^*

Antimicrobial Research Centre and Department of Biochemistry and Biomedical Sciences, McMaster University, 1200 Main St. W., Hamilton, Ontario, Canada L8N 3Z5

Received 21 June 2008/ Accepted 14 August 2008

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The study of bacterial extracellular polysaccharide biosynthesis is hampered by the fact that these molecules are synthesized on membrane-resident carrier lipids. To get around this problem, a practical solution has been to synthesize soluble lipid analogs and study the biosynthetic enzymes using a soluble system. This has been done for the Bacillus subtilis teichoic acid polymerase, TagF, although several aspects of catalysis were inconsistent with the results obtained with reconstituted membrane systems or physiological observations. In this work we explored the acceptor substrate promiscuity and polymer length disregulation that appear to be characteristic of TagF activity away from biological membranes. Using isotope labeling, steady-state kinetics, and chemical lability studies, we demonstrated that the enzyme can synthesize poly(glycerol phosphate) teichoic acid using the elongation substrate CDP-glycerol as an acceptor. This suggests that substrate specificity is relaxed in the region distal to the glycerol phosphate moiety in the acceptor molecule under these conditions. Polymer synthesis proceeded at a rate (27 min^–1) comparable to that in the reconstituted membrane system after a distinct lag period which likely represented slower initiation on the unnatural CDP-glycerol acceptor. We confirmed that polymer length became disregulated in the soluble system as the polymers synthesized on CDP-glycerol acceptors were much larger than the polymers synthesized on the membrane or previously found attached to bacterial cell walls. Finally, polymer synthesis on protease-treated membranes suggested that proper length regulation is retained in the absence of accessory proteins and provided evidence that such regulation is conferred through proper association of the polymerase with the membrane.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Teichoic acids comprise a diverse group of phosphate-containing polymers that are covalently attached to peptidoglycan in the cell walls of gram-positive bacteria. In Bacillus subtilis 168, the major wall teichoic acid is a linear 1,3-linked poly(glycerol phosphate) polymer that is anchored to peptidoglycan through an N-acetylmannosamine-β(1-4)-N-acetylglucosamine disaccharide linkage unit (Fig. 1A). B. subtilis 168 is the best model for the study of teichoic acid biosynthesis as the roles of the enzymes involved in the pathway have been demonstrated biochemically (5, 12, 27, 34) or have been predicted based on strong similarity to archetypal enzymes (21, 25, 35). This work has given rise to a model for teichoic acid biosynthesis that involves successive elongation of undecaprenyl-linked intermediates, followed by export and attachment to nascent peptidoglycan (Fig. 1B). Despite this, we still understand little about the mechanisms used by teichoic acid polymerases and primases to carry out their functions and even less about how the many steps of biosynthesis are coordinated and regulated at the level of individual proteins or protein complexes. Recent evidence has suggested that a key point of regulation may be reached upon completion of the disaccharide portion of the linkage unit. Mutants having genetic knockouts in the tagO and tagA genes (and the orthologous staphylococcal tarO and tarA genes) have been shown to be viable and, in fact, to suppress the lethality of downstream knockouts (9, 10, 38; unpublished observations). Consequently, it is critical to our understanding of teichoic acid biosynthesis to learn what factors are necessary for the priming and polymerization machinery to pass through this possible checkpoint. Recently, we showed that the TagF polymerase and TagB primase likely carry out catalysis using the same mechanism and that enzymes of this type can be grouped into a TagF-like family (33). We began our study of the importance of membrane association to the main-chain teichoic acid biosynthetic enzymes with TagF, the enzyme for which we have the most robust tools.

View larger version (38K): [in this window] [in a new window]

FIG. 1. Structure and synthesis of teichoic acid in B. subtilis 168. (A) Chemical structure of poly(glycerol phosphate) teichoic acid attached to peptidoglycan. Gray, peptidoglycan; blue, GlcNAc; red, ManNAc; green, glycerol phosphate. (B) Teichoic acid biosynthetic pathway. The solid arrows indicate steps in the pathway that are catalyzed by the identified enzymes. The dashed arrows indicate the steps at which various nucleotide activated precursors are involved. For an explanation of the colors see above.

In an effort to study teichoic acid polymerization away from the potentially confounding presence of biological membranes, our group and the group of Walker (12, 28) synthesized soluble linkage unit mimics that proved to be competent for catalysis. These studies showed that specific recognition of the isopreniod portion of the linkage unit acceptor molecule was not strictly required for catalysis and that some aspects of catalysis under these conditions were noticeably different from aspects of catalysis in experiments performed using reconstituted systems, possibly because of the absence of membrane or membrane-associated factors. In the synthetic systems, purified enzymes synthesized teichoic acid at a higher rate and produced polymers with a distribution of lengths that was inconsistent with physiological observations and in vitro experiments using B. subtilis membranes as acceptors (6, 12, 28). Understanding these differences, particularly the loss of polymer length regulation, may shed light on important aspects of teichoic acid biosynthesis as a regulated system that no one has yet been able to address.

Since teichoic acid is a major surface-exposed anionic polysaccharide of the gram-positive cell envelope, we may be able to make predictions about the coordination and regulation of its biosynthetic machinery by comparison to the synthesis of lipopolysaccharide (LPS) in gram-negative bacteria, about which much more is known. The distal O antigen of LPS is most analogous to the main chain of teichoic acid as it is more chemically diverse than the core or lipid A regions and it is a polymer composed of repeating (often anionic) carbohydrate units that, when extracted from cells, displays a consistent modal distribution of lengths. A number of O-antigen biosynthetic pathways have been described, which can be differentiated based on the involvement of a dedicated synthase enzyme, an ABC transporter or the Wzy polymerase (for a review, see reference 29). In each of the latter two pathways, the observed modal distribution of polymer lengths is achieved through the use of a multiprotein regulatory mechanism. Chain extension catalyzed by the Wzy polymerase is regulated by interaction with the Wzz protein (4), whereas polymerization in the ABC transporter-dependent pathway is terminated by chemical modification of terminal residues by independent enzymes (8). It is not known how polymer length is controlled in the synthase-dependent pathway. Teichoic acid biosynthesis involves the same undecaprenyl-phosphate acceptor that is used in peptidoglycan and LPS synthesis (2), is dependent on an ABC-2 family transporter (21), and appears to proceed through processive addition of monomeric substituents to the membrane-distal end of the polymer (17). Because of these common features, it is reasonable to suspect that the teichoic acid biosynthetic pathway might employ regulatory mechanisms analogous to those that control O-antigen biosynthesis.

In this paper, we report the surprising discovery that the TagF polymerase is capable of using the elongation substrate CDP-glycerol as an initiator substrate. We also provide evidence that association of the enzyme with the membrane itself may provide the basis for polymer length regulation and that this likely occurs without the involvement of accessory proteins.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
General methods. The TagF enzyme was purified as previously described (33). CDP-glycerol, CDP-[U-¹⁴C]glycerol, and [-³²P]CDP-glycerol were synthesized using glycerol-3-phosphate cytidylyltransferase from Staphylococcus aureus and previously described methods (3). [U-¹⁴C]glycerol-3-phosphate, Ni²⁺ chelating columns, and Superdex 200 columns were obtained from Amersham Biosciences (Baie d'Urfé, Quebec, Canada). -[³²P]CTP and scintillation fluid were obtained from PerkinElmer Life Sciences (Woodbridge, Ontario, Canada). ProteinPak 300SW, Ultrahydrogel 120, and Nova-Pak C₁₈ high-performance liquid chromatography (HPLC) columns were obtained from Waters (Mississauga, Ontario, Canada). Chromatography was performed using either an Amersham Biosciences Äkta fast protein liquid chromatography system or a Waters HPLC system. Polyclonal rabbit anti-TagF antibodies were raised for us by Cocalico Biologicals (Reamstown, PA). Rabbit anti-FtsY antibodies were kindly provided by David Andrews. Donkey anti-rabbit horseradish peroxidase-conjugated antibodies were obtained from Bio/Can Scientific (Mississauga, Ontario, Canada). Filters were obtained from Millipore (Nepean, Ontario, Canada). Dithiothreitol, isopropyl-β-D-thiogalactopyranoside (IPTG), imidazole, and ampicillin were obtained from Bioshop (Burlington, Ontario, Canada). Potassium phosphate was obtained from EM Science (Darmstadt, Germany). Complete EDTA-free protease inhibitor cocktail tablets were obtained from Roche (Laval, Quebec, Canada). All other reagents were obtained from Sigma (Mississauga, Ontario, Canada).

Poly(glycerol phosphate) polymerase assay. Experiments investigating TagF activity in the presence of B. subtilis membranes were carried out as previously described (34). The enzyme and substrate concentrations, as well as the incubation times, are described in the text. TagF activity was also assayed in the absence of membranes under similar conditions. The products of the reactions were analyzed by gel filtration chromatography as described below.

Chromatography of polymeric products. Alkaline-extracted membrane polymers (see below) and the products of membrane-free experiments were analyzed by gel filtration chromatography using a Waters ProteinPak 300SW column eluted with buffer GF2 (50 mM Tris, 100 mM NaCl, 10 mM EDTA; pH 7.5) as the mobile phase. The HPLC system was equipped with an in-line scintillation counting apparatus for the detection of radiolabeled molecules. The same chromatography procedure was used with a Waters Ultrahydrogel 120 gel filtration column for analysis of acid-hydrolyzed reaction products.

CMP release assay. Reactions using mixtures (200 µl) containing various concentrations of purified TagF protein (50 to 125 nM) in buffer CR {25 mM EPPS [4-(2-hydroxyethyl)piperazine-1-propanesulfonic acid], 35 mM MgCl₂; pH 8.0} were initiated by addition of CDP-glycerol to a concentration of 6 mM. The reactions were allowed to proceed at room temperature for appropriate time intervals before they were quenched by addition of 4 M urea. Products were then separated by paired ion chromatography (1) with a Waters Nova-Pak C₁₈ column using a linear gradient of buffer PIC-A (15 mM potassium phosphate, 10 mM TBAHS [tetrabutylammonium hydrogen sulfate]; pH 7.2) and buffer PIC-B (35 mM potassium phosphate, 10 mM TBAHS, 30% acetonitrile; pH 7.2). Products were detected by using absorbance at 271 nm, and reaction progress was monitored by examining the movement of absorbance intensity from the CDP-glycerol peak to the CMP peak. The identities of peaks were confirmed by injection of appropriate standard compounds.

Denaturing proteolysis. TagF protein (10 µg) or membrane-free reaction products were incubated in proteolysis buffer (50 mM HEPES, 40 mM KCl, 10 mM MgCl₂, 4 M urea; pH 7.3) containing 20 µg/ml trypsin for 0, 10, 25, or 60 min. Reactions were quenched by addition of 1 mM phenylmethylsulfonyl fluoride, followed by addition of boiling Laemmli buffer (20). Samples were then analyzed by discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 3% stacking and 12% separating gels and visualized with Coomassie blue stain. Identical experiments were also performed in the presence of CDP-glycerol (2 mM).

Chemical analysis of reaction products. Acid and alkaline hydrolysis of the products of both the membrane-free and on-membrane reactions was performed as previously described (34). Reaction products were incubated either in 1 M HCl for 3 h at 100°C or in 0.5 N NaOH for 25 min at 37°C.

Estimation of polymer length by using the isotope ratio. Two parallel membrane-free experiments were performed. In one experiment TagF (100 nM) was incubated in the presence of 150 µM CDP-[¹⁴C]glycerol, while in the other experiment [-³²P]CDP-glycerol was used. Reactions were allowed to proceed for 48 h, and then they were quenched by addition of 4 M urea. Samples were analyzed by gel filtration chromatography using a Waters ProteinPak 300SW column. By determining the ratio of the radioactive signal converted into product to the total radioactivity for each sample, we were able to calculate the amount of each radiolabeled species that was incorporated into the teichoic acid polymer. The values determined using the two different substrates were then compared to obtain the ratio of the number of ¹⁴C-labeled glycerol phosphate units (polymer) to the number of ³²P-labeled units (initiator) and thus the average length of the polymers.

Treatment of membranes with proteinase K. Inverted membranes were pelleted by centrifugation at 257,000 x g for 40 min and resuspended in 300 µl PK buffer (50 mM Tris-HCl [pH 8.0], 1 mM CaCl₂) with or without 3% Triton X-100. Proteinase K was added to a final concentration of 250, 500, or 1,000 µg/ml, and the mixture was incubated at 37°C for 90 min. The reaction was quenched by addition of 4 mM phenylmethylsulfonyl fluoride. The mixture was then diluted into 6 ml of 500 mM potassium acetate, and the membranes were pelleted by centrifugation at 104,000 x g for 45 min. The resulting pellet was washed once in membrane resuspension buffer (50 mM Tris-HCl, 10 mM MgCl, 1 mM EDTA; pH 7.2) and pelleted, and then the volume was adjusted so that the volume was equal to that of the starting material. The protease-treated membrane was then used as an acceptor in the polymerase assay.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
TagF is active in the presence of CDP-glycerol alone. We showed previously that TagF activity was enhanced by addition of B. subtilis membranes (34) or a soluble linkage unit mimic (28), consistent with the hypothesis that the natural acceptor of glycerol phosphate units is a membrane-associated undecaprenyl-pyrophosphate-linked disaccharide (19, 31). In the work presented here, we tested the ability of TagF to synthesize polymer in the presence of CDP-[¹⁴C]glycerol alone. To our surprise, TagF catalyzed the production of a large radioactive product when a reaction mixture was incubated under these conditions (Fig. 2).

View larger version (8K): [in this window] [in a new window]

FIG. 2. Synthesis of a large glycerol phosphate-containing product in the absence of membrane. TagF (100 nM) was incubated in the presence of 500 µM CDP-[14C]glycerol. The reaction was quenched after 23 h by addition of urea to 4 M. Reaction products were analyzed by gel filtration chromatography using an HPLC equipped for in-line scintillation counting. The results for zero time (solid line) and 23 h (dotted line) are shown. Formation of a high-molecular-weight product in the absence of membrane was observed (asterisk).

Poly(glycerol phosphate) was synthesized using CDP-glycerol alone. A series of experiments were conducted to confirm the identity of the large reaction product. First, the membrane-free experiment was repeated using the TagF-H474A variant, which we have previously shown to be catalytically inactive (33). No large radioactive product was produced when the H474A variant was used, suggesting that TagF activity is required for the production of this product (data not shown). To ensure that the large size of the product was not due to aggregation of the substrate with protein or the result of covalent modification of the enzyme, we subjected the reaction product to denaturing proteolysis in the presence of 20 µg/ml trypsin and analyzed the products by gel filtration. The results of the control reactions shown in the inset in Fig. 3A indicate that the proteolysis conditions were sufficient to completely destroy TagF both in the presence and in the absence of CDP-glycerol. Despite this, proteolysis did not result in a reduction in the size of the radioactive product as judged by gel filtration chromatography (Fig. 3A). Finally, we employed acid hydrolysis to confirm that the radioactive product was poly(glycerol phosphate). Hydrolysis of the cell wall extracted poly(glycerol phosphate) with hot acid hydrolyzes the polymer to its monomeric constituents (2, 34). Acid hydrolysis of the large radioactive product yielded two species that coeluted with glycerol phosphate and glycerol when they were analyzed by gel filtration chromatography (Fig. 3B). These results were consistent with the hypothesis that TagF synthesized an unusually large poly(glycerol phosphate) polymer in the presence of CDP-glycerol alone.

View larger version (22K): [in this window] [in a new window]

FIG. 3. Analysis of the membrane-free reaction product. (A) Membrane-free reaction products were subjected to denaturing proteolysis in the presence of 20 µg/ml trypsin for 60 min and were analyzed by gel filtration chromatography (Waters ProteinPak 300SW). (Inset) Purified TagF (10 µg) was subjected to identical proteolysis for 0, 10, 25, and 60 min in the presence or absence of CDP-glycerol (2 mM). A sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel of the products of proteolysis is shown. The positions of molecular mass standards (in kDa) are indicated on the left. The molecular mass of TagF is 90 kDa. (B) Purified membrane-free reaction products were treated in the presence of 1 M HCl for 3 h at 100°C and analyzed by gel filtration chromatography (Waters Ultrahydrogel 120). The peaks at 15 and 20 min coeluted with glycerol phosphate and glycerol standards, respectively.

CDP-glycerol was used as an acceptor in the absence of membrane. To test whether the enzyme used CDP-glycerol as an acceptor in the experiments described above, we first synthesized [-³²P]CDP-glycerol to determine whether the nucleotide portion of the molecule was incorporated into the polymer. TagF incorporated ³²P radioactivity into poly(glycerol phosphate) (Fig. 4A), while no radioactivity was incorporated in the absence of enzyme (data not shown). It was found that the amount of radioactivity incorporated into the polymer was much smaller when [-³²P]CDP-glycerol was used as substrate than when CDP-[U-¹⁴C]glycerol phosphate was used (data not shown). This was consistent with the apparently large size of the membrane-free polymer. The presence of CDP-glycerol at the initiation terminus of the polymer was further confirmed by performing alkaline lability studies. Treatment in dilute alkali is known to cleave poly(glycerol phosphate) from the linkage unit but leave the polymer largely intact (19, 26, 32). We have also previously confirmed this observation for teichoic acids synthesized on purified membranes (34). We reasoned that the pyrophosphate linkage within the hypothesized CDP-glycerol acceptor molecule (Fig. 4A, inset) should be more alkali labile than the glycerol phosphodiester linkages in the body of the polymer and, therefore, that treatment of the reaction product in dilute NaOH should selectively liberate ³²P radioactivity from the polymer if the hypothesis were true. Treatment with 0.5 N NaOH for 25 min at 37°C resulted in the removal of ³²P radioactivity from the reaction product (Fig. 4B). This cleavage coincided with the production of a peak that cochromatographed with CMP. Control experiments using poly([U-¹⁴C]glycerol phosphate) showed that treatment under these conditions left more than 80% of the polymers intact (data not shown). Together, these results demonstrated that TagF can use CDP-glycerol both as a source of activated glycerol phosphate and as an acceptor for teichoic acid polymerization when assays are performed in the absence of membrane.

View larger version (10K): [in this window] [in a new window]

FIG. 4. CDP-glycerol functions as the acceptor in the absence of membrane. (A) TagF (200 nM) was incubated with 150 µM [-32P]CDP-glycerol for 48 h. The reaction product was purified and analyzed by gel filtration chromatography using an HPLC system equipped for in-line scintillation counting. (Inset) Predicted structure of the reaction product, showing the position of 32P. (B) Purified reaction product was treated in the presence of 0.5 M NaOH for 25 min at 37°C. Neutralized samples were analyzed by gel filtration chromatography as described above for panel A. The presence of base-labile [-32P]CMP in the membrane-free polymers suggested that CDP-glycerol was used as the acceptor for polymerization.

Steady-state turnover in the presence of membrane and steady-state turnover in the absence of membrane were comparable. To study membrane-free activity, we measured the production of CMP over time by absorption spectroscopy analysis of products separated using paired ion chromatography. The rates of production of CMP were measured at four different TagF concentrations (Fig. 5). The fact that the response was linearly related to both time and enzyme concentration showed that the assay was robust and therefore suitable for investigation of membrane-free activity. The slope of the line describing the relationship between velocity and enzyme concentration (expressed in min^–1) (Fig. 5, inset) is an estimate of the turnover of the enzyme at steady state. In this system, TagF liberated CMP from CDP-glycerol at a rate of 27 min^–1.

View larger version (12K): [in this window] [in a new window]

FIG. 5. Dependence of membrane-free activity on time and enzyme concentration. TagF protein at concentrations of 50 (•), 75 (), 100 (), and 125 nM () was incubated in the presence of 6 mM CDP-glycerol for 120, 180, and 240 min. The reactions were quenched by addition of urea to 4 M, and the products were analyzed by paired ion chromatography. The production of CMP was assessed as described in Materials and Methods and plotted against time. (Inset) Steady-state velocity plotted against the concentration of TagF. The slope of the line represents the turnover of the enzyme under experimental conditions (27 min–1) and is very similar to that previously obtained for membrane-associated TagF (34).

Association with the membrane-regulated polymer length. The polymers synthesized away from the membrane eluted in the void volume of a gel filtration column with a manufacturer's stated exclusion limit of 150,000 Da (for a random coil peptide). Compared to single-stranded DNA standards, elution in the void translates to a polymer that is well over 100 glycerol phosphate units long, whether calibrated to molecular weight or the number of intramolecular bonds (28, 34). By analyzing the products from matched isotope labeling reactions allowed to go to completion, we were able to estimate the average ratio of initiator units to polymeric units in the membrane-free polymer. We observed that the incorporation of [U-¹⁴C]glycerol phosphate into the polymer was 140-fold greater than the incorporation of [-³²P]CDP-glycerol, suggesting that membrane-free polymers were, on average, 140 glycerol phosphate units long. To ensure that this result was not simply an artifact of long incubation times, we extracted [U-¹⁴C]glycerol phosphate polymers synthesized in the presence of membrane after 24 h of incubation. While the membrane-free polymers synthesized over a similar time period migrated during gel filtration with a retention time consistent with the large polymers described above (Fig. 2), Fig. 6 shows that polymers synthesized on the membrane had a physiological distribution of lengths indistinguishable from that observed previously (34). Thus, in the absence of cell membranes TagF catalyzed the synthesis of polymers that were several times longer than the polymers synthesized in the presence of cell membranes.

View larger version (22K): [in this window] [in a new window]

FIG. 6. Analysis of polymers synthesized on native and protease-treated membranes. (A) TagF (50 nM) was incubated in the presence of 750 µM CDP-[14C]glycerol and 20 mg protein/ml native B. subtilis membranes. (B) Inverted B. subtilis membranes were treated with proteinase K as described in Materials and Methods and subsequently used as an acceptor for TagF. The results obtained for membranes treated with 1,000 µg/ml proteinase K are shown. After 24 h, each reaction was quenched by addition of urea to 4 M, membranes were separated by ultracentrifugation, and the synthesized polymers were extracted by treatment in 0.5 M NaOH for 25 min at 37°C. Extracted polymers were analyzed by gel filtration chromatography. (Inset) Western blot analysis of the membrane preparations treated with 0, 250, 500, or 1,000 µg/ml proteinase K. Lane TX contained membranes treated with 1,000 µg/ml proteinase K in the presence of 3% Triton X-100.

Regulation of polymer length did not require accessory proteins. Having shown that association with the membrane directs TagF to synthesize teichoic acid polymers that are the physiological length, we asked whether any protein apart from TagF is required for the regulation of polymer length. Native B. subtilis inverted membranes were treated with 1 mg/ml proteinase K for 90 min at 37°C. This treatment was sufficient to abolish endogenous TagF activity (data not shown). Western blot analysis was performed to probe for the presence of the TagF and FtsY proteins after protease treatment (Fig. 6B, inset). Each protein was destroyed by proteinase K treatment, suggesting that membrane-associated proteins were generally destroyed by this treatment. We noticed that a consistent minimum amount of FtsY appeared to remain even after treatment with the highest concentrations of proteinase K and that the amount could not be reduced further by addition of more protease. By repeating the proteolysis in the presence of 3% Triton X-100, we confirmed that this minimum level represented the amount of protein contained within the approximately 5% of vesicles in a noninverted orientation formed when bacterial cells are lysed by a French press (11). This phenomenon was not as apparent for TagF, presumably because TagF is a less abundant protein. Proteolyzed membranes were incubated with purified TagF as described above, and the polymers synthesized were extracted and analyzed by gel filtration chromatography. Polymers synthesized on proteinase K-treated membranes (Fig. 6B) behaved during gel filtration like the polymers synthesized on untreated membranes (Fig. 6A) (34) and not like the polymers synthesized in the absence of membranes (Fig. 2). This result suggests that TagF is capable of synthesizing polymers that are the physiological length in the absence of additional protein factors.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The study of biosynthesis of bacterial extracellular polysaccharides, including teichoic acid, has been hampered by the fact that these molecules are synthesized on membrane-resident carrier lipids, necessitating the use of cumbersome and ill-defined reconstituted assay systems (7, 13, 14, 16, 18, 23, 24). Recently, we described a completely soluble system for the study of individual teichoic acid synthetic enzymes (28). Observations made in this work suggested that the teichoic acid polymerase displayed relaxed acceptor specificity, no polymer length regulation, and a distributive mechanism of polymer synthesis in the absence of membranes. In the work presented here we sought to further explore the differences between TagF activity on the membrane and TagF activity off the membrane and to identify the factors that govern the synthesis of physiologically relevant polymers. We have shown previously that the TagF enzyme from B. subtilis strain 168 is responsible for poly(glycerol phosphate) polymerization (34) and that it is a good model for other enzymes in the teichoic acid primase/polymerase family (33). Therefore, we believe that insights gained in this work can be generalized to the synthesis of related teichoic acid polymers in other organisms, about which much more is rapidly becoming known.

It has been observed that some polysaccharide polymerase enzymes can function in the absence of their endogenous acceptors if activated precursors or oligomeric primers are provided in excess. For example, glycogen synthase can catalyze glycosyl transfer to glucose at sufficiently high concentrations (30), and the NeuS glycosyltransferase involved in Escherichia coli K1 antigen biosynthesis shows activity with endogenous acceptor or oligosaccharide primers (36, 37). It is in this context that we made the surprising discovery that the TagF enzyme could synthesize a large radiolabeled product in the presence of CDP-[¹⁴C]glycerol alone. The time-dependent production of a large ¹⁴C-containing molecule under these conditions suggested either that poly(glycerol phosphate) was produced in the absence of a known acceptor or that the enzyme and radioactive material aggregated together or became covalently associated. We performed several experiments to discriminate between these possibilities. Several properties of the large reaction product, in addition to the fact that active TagF was required to synthesize it, led to the conclusion that TagF did indeed synthesize poly(glycerol phosphate) in the absence of the physiological acceptor or a closely related acceptor mimic.

We have shown previously that TagF activity is dependent on the concentration of a suitable acceptor molecule (28, 34). Therefore, it was important to understand how polymerization was initiated in the current system. To examine this, we synthesized [-³²P]CDP-glycerol to track the nucleotide portion of the activated precursor. We hypothesized that TagF may have used CDP-glycerol itself as an acceptor since this molecule loosely resembles the disaccharide linkage unit in structure and could possibly bind in an alternate mode at an initiator site on the enzyme. The only way that a radioactive polymer could be synthesized using [-³²P]CDP-glycerol would be if the CMP portion of the molecule, usually released as a product, was instead incorporated into the polymer. The fact that a ³²P-labeled polymer was produced in these experiments supported our hypothesis and eliminated the possibility that polymers had been synthesized on some unknown E. coli component that may have copurified with the enzyme. We confirmed that CDP-glycerol was used as the acceptor through treatment of the reaction product in dilute NaOH. Under these conditions the predicted pyrophosphate linkage at the initiator terminus of the polymer would be more susceptible to hydrolysis than the phosphodiester linkages found throughout the body of the polymer. Consequently, if the proposed initiator structure was true, alkaline hydrolysis would preferentially remove ³²P radioactivity from the polymer while leaving the phosphdiester linkages largely intact. Our results showed that this was indeed the case, confirming that CDP-glycerol can be used as an initiator molecule and expanding the observation that acceptor specificity is relaxed when the enzyme is examined away from the membrane. Consistent with the elution of the membrane-free polymer at the void volume of the gel filtration column, we observed that 140 times more [U-¹⁴C]glycerol phosphate equivalents than [-³²P]CMP equivalents were incorporated into the large polymers in matched experiments. These results provided evidence that there was a loss of polymer length regulation in the absence of membrane that resulted in the synthesis of polymers several times larger than physiological polymers.

To confirm that the surprising length of polymers synthesized off the membrane was not simply an artifact of the soluble system, we compared the polymerization rates on and off the membrane and also examined whether long polymers could be made on the membrane, even after long incubation times. We monitored polymerase activity by observing the production of the CMP product by absorbance spectroscopy. This response was linearly related to the enzyme concentration and time over several hours, and, by replotting the rates at various enzyme concentrations versus the enzyme concentration, we showed that the steady-state turnover in the absence of the membrane was very similar to that observed in the presence of the membrane (27 and 16 min^–1 [34], respectively). Therefore, the absence of the membrane in this system had very little effect on the rate at which TagF synthesized poly(glycerol phosphate). However, the lag in activity observed when CDP-glycerol was used as an acceptor was interesting. This lag may suggest that CDP-glycerol, although usable, is a poor substrate for initiation. According to this hypothesis, TagF would synthesize the first CMP-(glycerol phosphate)_n oligomers at a low rate until a threshold size and/or concentration of competent acceptors was reached, at which point polymerization would continue at the normal rate. Although this distinction makes little difference to the ultimate production of poly(glycerol phosphate) under these conditions, it is interesting because it suggests that alterations in TagF that affect only initiation might be indicated by modulation of the lag period. In this way, it is possible that we could use this system to separately study initiation and polymerization and thereby identify which residues or regions of TagF are important for each function. We were also interested in determining if it was possible to find long polymers that were synthesized on the membrane. To investigate this, we incubated TagF in the presence of membranes and CDP-[U-¹⁴C]glycerol for up to 24 h. The polymers extracted from membranes after this treatment were identical to those observed previously (34); namely, they were the physiological length. These results confirmed that the association of TagF with the membrane regulates teichoic acid length in some way.

The issue of teichoic acid length regulation has not been addressed previously. Extracellular polysaccharides from both gram-positive and gram-negative organisms most often display a fairly discreet modal distribution of lengths, suggesting that there is some form of regulation. In gram-negative bacteria, these systems involve accessory proteins that either modulate polymerase activity or "cap" the polysaccharide with modified substituents (29). The production of unregulated teichoic acid polymers away from the membrane could easily be explained by the absence of either type of factor. Indeed, the existence of a teichoic acid synthetic complex has been suggested (2, 15, 22). However, no terminally modified teichoic acids have been described, and neither bioinformatics nor biochemical analysis have led to the identification of any potential regulatory proteins. To address this question, we treated purified B. subtilis inverted membrane vesicles with proteinase K to destroy membrane-associated proteins. We then analyzed the teichoic acids that were synthesized on the membranes after incubation with only purified TagF and CDP-glycerol. The protease treatment used was more stringent than that used previously to destroy TagB (5) and was found in this work to completely destroy TagF. Thus, two proteins known to be involved in teichoic acid biosynthesis were eliminated by this procedure. Furthermore, the abundant membrane-associated protein FtsY (signal recognition particle receptor homologue) was also destroyed. When protease-treated membranes were used as a teichoic acid acceptor, the pattern of extracted polymer lengths was identical to that observed with untreated membranes. This interesting result suggests that no accessory proteins are required to regulate teichoic acid polymer length and that the observed length regulation can be attributed entirely to the interaction between the membrane and TagF. This could be explained if association of the polymerase with phospholipids or the isoprenoid portion of the linkage unit, which was also absent from the soluble system, caused a conformational change in the protein, thereby imparting the proper processivity. Such a phenomenon would not be unprecedented in the field of extracellular polysaccharide biosynthesis as many polyisoprenyl-phosphate glycosylation steps in both prokaryotes and eukaryotes involve the interaction of a conserved polyisoprenyl recognition sequence in the glycosyltransferase with the isoprenoid substrate. This interaction has been shown to alter the conformation of both the polyisoprenoid molecule and the enzyme (39), although it occurs within the bilayer and consequently is restricted to membrane-spanning glycosyltransferases. Alternatively, association of TagF with the membrane could impose a physical restriction on how far the enzyme is free to move during synthesis or could simply discourage reinitiation on completed polymers. The latter possibility is interesting in light of the observation that the polymerase appears to operate in a distributive mode away from the membrane (28), a polymerization mechanism that does not seem likely to produce the regular polymer length distribution observed in cells and is at odds with the prediction of Walker's group that the Staphylococcus aureus polymerase TarL operates via a processive mechanism (6). More work is needed to flesh out all of the details surrounding teichoic acid polymer length regulation. However, the work presented here demonstrates that, despite sharing some features of each of the pathways governing gram-negative O-antigen synthesis, teichoic acid biosynthesis appears to be mechanistically unique.

In this study, we further explored the properties necessary for a molecule to serve as an acceptor for teichoic acid polymerization and identified the factors that are important for the synthesis of polymers that are physiological length. We showed that the TagF enzyme could use CDP-glycerol as an acceptor and that the rate of polymer synthesis when this acceptor was used was comparable to the rate when the true acceptor molecule present in purified B. subtilis membranes was used. We also observed a lag in membrane-free catalysis that could possibly be exploited to investigate the steps involved in polymerase initiation. Although the membrane was not required for the polymerase to be active, we showed that proper association of the enzyme with the membrane was critical for regulating teichoic acid polymer length and that this was likely accomplished without the aid of additional protein factors. In summary, the work presented here provided great insight into teichoic acid polymerization and how it is controlled. It also raised important questions about how membrane association enables proper polymer length regulation and what features of the enzyme-membrane association mediate this control. We will continue to address these questions in our laboratory as we seek to fully describe the mechanisms of teichoic acid biosynthesis.

 
This work was supported by Canadian Institutes of Health Research grant MOP-15496, by a Canada Research Chair in Chemical Biology (to E.D.B.), and by a Canadian Institutes of Health Research Canada graduate scholarship (to J.W.S.).

We thank David Andrews and his group for providing antibodies for the detection of the B. subtilis FtsY protein.

 
* Corresponding author. Mailing address: Department of Biochemistry and Biomedical Sciences, McMaster University, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada. Phone: (905) 525-9140, ext. 22392. Fax: (905) 522-9033. E-mail: ebrown{at}mcmaster.ca

Published ahead of print on 19 August 2008.

Present address: Department of Molecular Genetics and Microbiology, University of Texas at Austin, Austin, TX 78712.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Allinquant, B., C. Musenger, and E. Schuller. 1985. Reversed-phase high-performance liquid chromatography of nucleotides and oligonucleotides. J. Chromatogr. 326:281-291.[CrossRef]
2
2. Anderson, R. G., H. Hussey, and J. Baddiley. 1972. The mechanism of wall synthesis in bacteria. The organization of enzymes and isoprenoid phosphates in the membrane. Biochem. J. 127:11-25.[Medline]
3
3. Badurina, D. S. 2003. CTP:glycerol 3-phosphate cytidylyltransferase (TarD) from Staphylococcus aureus catalyzes the cytidylyl transfer via an ordered Bi-Bi reaction mechanism with micromolar K_m values. Biochim. Biophys. Acta 1646:196-206.[Medline]
4
4. Bastin, D. A., G. Stevenson, P. K. Brown, A. Haase, and P. R. Reeves. 1993. Repeat unit polysaccharides of bacteria: a model for polymerization resembling that of ribosomes and fatty acid synthetase, with a novel mechanism for determining chain length. Mol. Microbiol. 7:725-734.[Medline]
5
5. Bhavsar, A. P., R. Truant, and E. D. Brown. 2005. The TagB protein in Bacillus subtilis 168 is an intracellular peripheral membrane protein that can incorporate glycerol phosphate onto a membrane-bound acceptor in vitro. J. Biol. Chem. 280:36691-36700.[Abstract/Free Full Text]
6
6. Brown, S., Y. H. Zhang, and S. Walker. 2008. A revised pathway proposed for Staphylococcus aureus wall teichoic acid biosynthesis based on in vitro reconstitution of the intracellular steps. Chem. Biol. 15:12-21.[CrossRef][Medline]
7
7. Burger, M. M., and L. Glaser. 1964. The synthesis of teichoic acids. I. Polyglycerolphosphate. J. Biol. Chem. 239:3168.[Free Full Text]
8
8. Clarke, B. R., L. Cuthbertson, and C. Whitfield. 2004. Nonreducing terminal modifications determine the chain length of polymannose O antigens of Escherichia coli and couple chain termination to polymer export via an ATP-binding cassette transporter. J. Biol. Chem. 279:35709-35718.[Abstract/Free Full Text]
9
9. D'Elia, M. A., K. E. Millar, T. J. Beveridge, and E. D. Brown. 2006. Wall teichoic acid polymers are dispensable for cell viability in Bacillus subtilis. J. Bacteriol. 188:8313-8316.[Abstract/Free Full Text]
10
10. D'Elia, M. A., M. P. Pereira, Y. S. Chung, W. Zhao, A. Chau, T. J. Kenney, M. C. Sulavik, T. A. Black, and E. D. Brown. 2006. Lesions in teichoic acid biosynthesis in Staphylococcus aureus lead to a lethal gain of function in the otherwise dispensable pathway. J. Bacteriol. 188:4183-4189.[Abstract/Free Full Text]
11
11. Futai, M. 1974. Orientation of membrane vesicles from Escherichia coli prepared by different procedures. J. Membr. Biol. 15:15-28.[CrossRef][Medline]
12
12. Ginsberg, C., Y. H. Zhang, Y. Yuan, and S. Walker. 2006. In vitro reconstitution of two essential steps in wall teichoic acid biosynthesis. ACS Chem. Biol. 1:25-28.[CrossRef][Medline]
13
13. Glaser, L. 1964. The synthesis of teichoic acids. II. Polyribitolphosphate. J. Biol. Chem. 239:3178-3186.[Free Full Text]
14
14. Guan, S., A. J. Clarke, and C. Whitfield. 2001. Functional analysis of the galactosyltransferases required for biosynthesis of D-galactan I, a component of the lipopolysaccharide O1 antigen of Klebsiella pneumoniae. J. Bacteriol. 183:3318-3327.[Abstract/Free Full Text]
15
15. Hancock, I. C., and J. Baddiley. 1973. Solubilisation of a teichoic acid-synthesising system from the membrane of Bacillus licheniformis by freezing and thawing. FEBS Lett. 34:15-18.[CrossRef][Medline]
16
16. Harrington, C. R., and J. Baddiley. 1985. Biosynthesis of wall teichoic acids in Staphylococcus aureus H, Micrococcus varians and Bacillus subtilis W23. Involvement of lipid intermediates containing the disaccharide N-acetylmannosaminyl N-acetylglucosamine. Eur. J. Biochem. 153:639-645.[Medline]
17
17. Kennedy, L. D., and D. R. D. Shaw. 1968. Direction of polyglycerolphosphate chain growth in Bacillus subtilis. Biochem. Biophys. Res. Commun. 32:861-865.[CrossRef]
18
18. Kido, N., V. I. Torgov, T. Sugiyama, K. Uchiya, H. Sugihara, T. Komatsu, N. Kato, and K. Jann. 1995. Expression of the O9 polysaccharide of Escherichia coli: sequencing of the E. coli O9 rfb gene cluster, characterization of mannosyl transferases, and evidence for an ATP-binding cassette transport system. J. Bacteriol. 177:2178-2187.[Abstract/Free Full Text]
19
19. Kojima, N., Y. Araki, and E. Ito. 1983. Structure of linkage region between ribitol teichoic acid and peptidoglycan in cell walls of Staphylococcus aureus H. J. Biol. Chem. 258:9043-9045.[Abstract/Free Full Text]
20
20. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T7. Nature 227:680-685.[CrossRef][Medline]
21
21. Lazarevic, V., and D. Karamata. 1995. The tagGH operon of Bacillus subtilis 168 encodes a two-component ABC transporter involved in the metabolism of two wall teichoic acids. Mol. Microbiol. 16:345-355.[Medline]
22
22. Leaver, J., I. C. Hancock, and J. Baddiley. 1981. Fractionation studies of the enzyme complex involved in teichoic acid synthesis. J. Bacteriol. 146:847-852.[Abstract/Free Full Text]
23
23. Liu, D., A. M. Haase, L. Lindqvist, A. A. Lindberg, and P. R. Reeves. 1993. Glycosyl transferases of O-antigen biosynthesis in Salmonella enterica: identification and characterization of transferase genes of groups B, C2, and E1. J. Bacteriol. 175:3408-3413.[Abstract/Free Full Text]
24
24. Mauck, J., and L. Glaser. 1972. An acceptor-dependent polyglycerolphosphate polymerase. Proc. Natl. Acad. Sci. USA 69:2386-2390.[Abstract/Free Full Text]
25
25. Mauel, C., M. Young, A. Monsutti-Grecescu, S. A. Marriott, and D. Karamata. 1994. Analysis of Bacillus subtilis tag gene expression using transcriptional fusions. Microbiology 140:2279-2288.[Abstract/Free Full Text]
26
26. McArthur, H. A., F. M. Roberts, I. C. Hancock, and J. Baddiley. 1978. Lipid intermediates in the biosynthesis of the linkage unit between teichoic acids and peptidoglycan. FEBS Lett. 86:193-200.[CrossRef][Medline]
27
27. Park, Y. S., T. D. Sweitzer, J. E. Dixon, and C. Kent. 1993. Expression, purification, and characterization of CTP:glycerol-3-phosphate cytidylyltransferase from Bacillus subtilis. J. Biol. Chem. 268:16648-16654.[Abstract/Free Full Text]
28
28. Pereira, M. P., J. W. Schertzer, M. A. D'Elia, K. P. Koteva, D. W. Hughes, G. D. Wright, and E. D. Brown. 2008. The wall teichoic acid polymerase TagF efficiently synthesizes poly(glycerol phosphate) on the TagB product lipid III. Chembiochem 9:1385-1390.[CrossRef][Medline]
29
29. Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71:635-700.[CrossRef][Medline]
30
30. Salsas, E., and J. Larner. 1975. Glycogen synthase can use glucose as an acceptor. J. Biol. Chem. 250:1833-1837.[Abstract/Free Full Text]
31
31. Sasaki, Y., Y. Araki, and E. Ito. 1980. Structure of linkage region between glycerol teichoic acid and peptidoglycan in Bacillus cereus AHU 1030 cell walls. Biochem. Biophys. Res. Commun. 96:529-534.[CrossRef][Medline]
32
32. Sasaki, Y., Y. Araki, and E. Ito. 1983. Structure of teichoic-acid-glycopeptide complexes from cell walls of Bacillus cereus AHU 1030. Eur. J. Biochem. 132:207-213.[Medline]
33
33. Schertzer, J. W., A. P. Bhavsar, and E. D. Brown. 2005. Two conserved histidine residues are critical to the function of the TagF-like family of enzymes. J. Biol. Chem. 280:36683-36690.[Abstract/Free Full Text]
34
34. Schertzer, J. W., and E. D. Brown. 2003. Purified, recombinant TagF protein from Bacillus subtilis 168 catalyzes the polymerization of glycerol phosphate onto a membrane acceptor in vitro. J. Biol. Chem. 278:18002-18007.[Abstract/Free Full Text]
35
35. Soldo, B., V. Lazarevic, and D. Karamata. 2002. tagO is involved in the synthesis of all anionic cell-wall polymers in Bacillus subtilis 168. Microbiology 148:2079-2087.[Abstract/Free Full Text]
36
36. Steenbergen, S. M., T. J. Wrona, and E. R. Vimr. 1992. Functional analysis of the sialyltransferase complexes in Escherichia coli K1 and K92. J. Bacteriol. 174:1099-1108.[Abstract/Free Full Text]
37
37. Vimr, E. R., W. Aaronson, and R. P. Silver. 1989. Genetic analysis of chromosomal mutations in the polysialic acid gene cluster of Escherichia coli K1. J. Bacteriol. 171:1106-1117.[Abstract/Free Full Text]
38
38. Weidenmaier, C., J. F. Kokai-Kun, S. A. Kristian, T. Chanturiya, H. Kalbacher, M. Gross, G. Nicholson, B. Neumeister, J. J. Mond, and A. Peschel. 2004. Role of teichoic acids in Staphylococcus aureus nasal colonization, a major risk factor in nosocomial infections. Nat. Med. 10:243-245.[CrossRef][Medline]
39
39. Zhou, G. P., and F. A. Troy II. 2003. Characterization by NMR and molecular modeling of the binding of polyisoprenols and polyisoprenyl recognition sequence peptides: 3D structure of the complexes reveals sites of specific interactions. Glycobiology 13:51-71.[Abstract/Free Full Text]

---

Journal of Bacteriology, November 2008, p. 6940-6947, Vol. 190, No. 21
0021-9193/08/$08.00+0     doi:10.1128/JB.00851-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Sewell, E. W. C., Pereira, M. P., Brown, E. D. (2009). The Wall Teichoic Acid Polymerase TagF Is Non-processive in Vitro and Amenable to Study Using Steady State Kinetic Analysis. J. Biol. Chem. 284: 21132-21138 [Abstract] [Full Text]  

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schertzer, J. W. Articles by Brown, E. D. Search for Related Content PubMed PubMed Citation Articles by Schertzer, J. W. Articles by Brown, E. D.