A Novel Mechanism for Resistance to the Antimetabolite N-Phosphonoacetyl-l-Aspartate by Helicobacter pylori

Abstract

The mechanism of resistance to N-phosphonoacetyl-l-aspartate (PALA), a potent inhibitor of aspartate carbamoyltransferase (which catalyzes the first committed step of de novo pyrimidine biosynthesis), in Helicobacter pylori was investigated. At a 1 mM concentration, PALA had no effects on the growth and viability of H. pylori. The inhibitor was taken up by H. pylori cells and the transport was saturable, with a K_m of 14.8 mM and a V_max of 19.1 nmol min⁻¹ μl of cell water⁻¹. By ³¹P nuclear magnetic resonance (NMR) spectroscopy, both PALA and phosphonoacetate were shown to have been metabolized in all isolates of H. pylori studied. A main metabolic end product was identified as inorganic phosphate, suggesting the presence of an enzyme activity which cleaved the carbon-phosphorus (C-P) bonds. The kinetics of phosphonate group cleavage was saturable, and there was no evidence for substrate inhibition at higher concentrations of either compound. C-P bond cleavage activity was temperature dependent, and the activity was lost in the presence of the metal chelator EDTA. Other cleavages of PALA were observed by ¹H NMR spectroscopy, with succinate and malate released as main products. These metabolic products were also formed when N-acetyl-l-aspartate was incubated with H. pylori lysates, suggesting the action of an aspartase. Studies of the cellular location of these enzymes revealed that the C-P bond cleavage activity was localized in the soluble fraction and that the aspartase activity appeared in the membrane-associated fraction. The results suggested that the two H. pylori enzymes transformed the inhibitor into noncytotoxic products, thus providing the bacterium with a mechanism of resistance to PALA toxicity which appears to be unique.

Helicobacter pylori has been established as the causative agent of chronic gastritis and a significant proportion of duodenal and gastric ulcers (14). Recently, the World Health Organization classified H. pylori as a group 1 carcinogen, owing to its role in the development of gastric cancer (10). The failure of some regimens in the treatment of H. pylori infection has motivated work in our laboratory directed at characterizing the physiology of the bacterium, with the aim of discovering potential sites for therapeutic intervention, including nucleotide biosynthetic pathways (24, 25).

Earlier studies on the uptake of nucleotide precursors by H. pylori showed that there was relatively little acquisition of pyrimidine nucleotide precursors by the salvage of preformed bases and nucleosides (24). Uracil, a commonly salvaged pyrimidine base, is also not required for the growth of this bacterium (34), suggesting that the majority of its pyrimidine nucleotides are synthesized through the de novo pathway. In contrast, humans can utilize the de novo or salvage pathway for the synthesis of pyrimidine nucleotides. Inhibitors of H. pylori de novo pyrimidine biosynthesis may therefore be potentially effective therapeutic drugs, as the host could still efficiently acquire its nucleotide requirements by salvage. This potential was demonstrated earlier by the finding that the inhibition of de novo pyrimidine biosynthesis at the second enzyme of this pathway, dihydroorotase, resulted in the killing of H. pylori cells (35).

Aspartate carbamoyltransferase (ACTase) (EC 2.1.3.2) catalyzes the first committed step in the de novo formation of pyrimidine nucleotides and is a key regulatory enzyme in bacteria (8). N-Phosphonoacetyl-l-aspartate (PALA) is a synthetic, transition state bisubstrate analogue of the intermediate of the ACTase-catalyzed reaction (5). PALA belongs to a group of organophosphorus compounds known as phosphonates, characterized by their extremely stable carbon-phosphorus (C-P) bond in place of the more common carbon-oxygen-phosphorus ester bond (39), which confers on them the advantage of inherent stability. Natural phosphonates are found in phosphonolipids, glycolipids, glycoproteins, and polysaccharides of many different organisms. PALA and other synthetic phosphonates have been produced for use as herbicides, antibacterial agents (1, 28), and even as agents of chemical warfare (38).

PALA is a potent inhibitor of the ACTase-catalyzed reaction in a range of prokaryotic and eukaryotic organisms, including Escherichia coli (5), Pyrococcus abyssi (33), and Leishmania donovani (29), and in mammalian cells (36). Owing to its stability and toxic effects on a key regulatory enzyme, PALA has been employed as an antitumor agent to inhibit the growth of rapidly proliferating cancer cells (9, 36). The inhibitor was also suggested as a possible antimetabolite for the protozoan pathogen L. donovani due to its cytotoxic effects on this organism (29). However, we have not found any detailed studies investigating the effects of PALA on the viability of bacterial cells. Recent results indicated that PALA is a potent inhibitor of ACTase activity in H. pylori, with 50% inhibition of enzyme activity observed at 0.1 μM PALA, and that PALA binds to the enzyme over 2,500 times more tightly than carbamoyl phosphate (3). This finding suggested that ACTase in H. pylori was a potential target for therapeutic intervention. However, initial results in our laboratory showed that PALA did not have inhibitory effects on the growth and viability of the bacterium.

The aim of this work was to elucidate the mechanism(s) for H. pylori resistance to the potentially toxic effects of PALA. The effects on growth and viability, the transport of the inhibitor into whole cells, and the metabolic fate of this compound inside the cell were investigated by radiotracer analyses and nuclear magnetic resonance (NMR) spectroscopy.

MATERIALS AND METHODS

Substrates and reagents.

Carbamoyl phosphate, carbamoyl aspartate, dipyridamole, phosphonoacetate, glyphosate, methylphosphonate, and N-acetyl-l-aspartate were from Sigma (St. Louis, Mo.). Blood agar base no. 2 and Iso-Sensitest media were from Oxoid (Basingstoke, United Kingdom). PALA was kindly provided by Jill Johnson (Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment, National Cancer Institute, Bethesda, Md.). Phosphonoacetaldehyde was a kind gift from H. B. F. Dixon (Department of Biochemistry, University of Cambridge, Cambridge, United Kingdom). All other reagents were of analytical grade.

Strains and growth conditions.

H. pylori NCTC 11639 and SS1 and a recent clinical isolate, UNSW 92100/5, were grown on blood agar base no. 2 plates, supplemented with 5 to 7% (vol/vol) horse blood. The medium was supplemented with amphotericin B (Fungizone) (2 μg ml⁻¹), trimethoprim (5 μg ml⁻¹), polymyxin B (2.5 μg ml⁻¹), and vancomycin (10 μg ml⁻¹). H. pylori cultures were incubated in a Stericult incubator (Forma Scientific, Sydney, New South Wales, Australia) in an atmosphere of 10% CO₂ in air and 95% relative humidity at 37°C. Cells were passaged every 30 h, checked for purity by phase-contrast microscopy, and tested for urease and catalase activity. For cell viability studies, cells were grown in a liquid culture of Iso-Sensitest medium supplemented with 1% bovine serum albumin and 0.5% catalase and the above-mentioned antibiotics. Cultures were grown in vented tissue culture flasks in a microaerophilic environment generated in a jar that contained an anaerobic gas pack without a catalyst. Jars were incubated at 37°C with shaking.

Preparation of cell extracts.

After approximately 30 h, cells were harvested in 150 mM NaCl and centrifuged at 17,000 × g for 8 min at 4°C. Each resulting pellet was washed three times and resuspended in 150 mM NaCl (ca. 4.5 mg ml⁻¹). To prepare lysates, cells were disrupted by being frozen in liquid nitrogen and thawed twice. To study the cellular location of enzyme activity, lysates were centrifuged at 27,000 × g for 45 min at 4°C, and the supernatant was carefully separated from each pellet. Pellets were washed another three times and resuspended in 150 mM NaCl (5:1). After the same results regarding resistance to PALA and breakdown of this compound were observed in all strains, NCTC 11639 was used for the detailed analysis of the mechanism of resistance.

Measurements of cell viability.

Cell proliferation was measured in the presence of various concentrations of PALA. Cultures were grown for 48 h and sampled at 4-h intervals to measure growth and viability. Cell optical density was measured at 600 nm, and viability was determined by the viable-plate technique (27). Cultures grown under identical conditions in the absence of any added PALA were used as positive controls, and uninoculated flasks were used as negative controls.

Transport measurements.

Cells were harvested in log phase (ca. 30 h) in 0.9% (wt/vol) NaCl, and the preparations were centrifuged at 17,000 × g for 8 min at 4°C. The transport of [³H]PALA (18.01 μCi μmol⁻¹) was measured by a centrifugation-through-oil method previously described for H. pylori (19). Radioactivity was counted with a Packard Tri-Carb 2100TR scintillation system.

NMR spectroscopy.

The fates of specific metabolites were monitored over time by ¹H NMR or ³¹P NMR spectroscopy. Free-induction decays were collected with a Bruker DMX-500 spectrometer operating in the Fourier transformation mode. Measurements were carried out at 37°C. Sequential spectra were acquired automatically at 500.13 MHz with a presaturation of the water resonance. The instrumental parameters were a spectral width of 5,340.7 Hz, a memory size of 8 kilobytes, a recycling time of 3.5 s, a number of transients of 144, and a pulse angle of 50° (8 μs). Exponential filtering of 1 Hz was applied prior to Fourier transformation. Substrates from 0.3 M stock solutions were dispensed into 5-mm-diameter NMR tubes (Wilmad, Buena, N.J.) containing NaCl (0.15 M) or HEPES (0.25 M) buffer. The reactions were started by adding 200 μl of cell extract. The total sample volume was 600 μl, and H₂O-²H₂O (11:1, vol/vol) buffer mixtures were employed to provide a deuterium frequency lock for the spectrometer.

The evolution over time of the utilization of substrates and the appearance of products was monitored with sequential spectra. Assignments of the spectra arising from the products were made by adding the appropriate compounds to cell preparations and observing the overlap of resonances of the unknown product and the added compound. Progress curves were obtained by measuring the integrals of substrate and product resonances at each point in time. Maximal rates were calculated from good fits (correlation coefficients ≥0.99) of the data to straight lines for the first 30 min of the reactions. Calibrations of the peaks arising from substrates were performed by extrapolating the resonance intensity data to zero time and assigning the appropriate concentration to this intensity. The intensities of resonances corresponding to products were calibrated by adding the appropriate metabolite to cell suspensions and constructing standard concentration curves.

Estimation of the molecular size of the enzyme catalyzing C-P bond cleavage.

The approximate size of the protein catalyzing C-P bond cleavage was determined by membrane filtration. A volume of 1 ml of H. pylori cell extract was filtered through a 100-kDa-cutoff membrane of a Centricon concentrator (Amicon, Beverly, Mass.) by centrifugation at 1,000 × g. The retentate and filtrate were collected, and the filtrate was concentrated further through a 50-kDa-cutoff Centricon membrane. The retentate and filtrate were kept, and the volumes of each suspension were brought to 300 μl by concentration through a 10-kDa-cutoff Centricon membrane. Each fraction was tested for C-P bond cleavage activity as described above.

Enzyme activity and protein concentration determination.

ACTase activity was measured in H. pylori cell extracts by a microtiter plate protocol previously described (7). Protein concentrations were estimated by the bicinchoninic acid method (with a kit from Pierce Chemical Co., Rockford, Ill.).

Kinetic analyses.

The kinetic parameters K_m and V_max were determined from measurements of initial rates of 10 time courses for each substrate. The values for the kinetic parameters were calculated by nonlinear regression analysis with the program Enzyme Kinetics (Trinity Software, Campton, N.H.).

RESULTS

Effect of PALA on H. pylori growth and viability.

At concentrations of 10 μM, 100 μM, and 1 mM, PALA had no effect on the growth and viability of H. pylori NCTC 11639, SS1, and UNSW 92100/5. There was no significant difference between the growth of cultures containing the enzyme inhibitor and the growth of those without PALA for 48 h, according to both turbidity measurements and viable counts. There was also no effect on cell growth in cultures grown in the presence of PALA and a 1 mM concentration of the uridine uptake inhibitor dipyridamole.

Transport of [³H]PALA.

The transmembrane transport of [³H]PALA was investigated to establish whether the inhibitor was able to enter intact, metabolically competent cells. Uptake of PALA into H. pylori at 20°C was linear for 10 min at a fixed permeant concentration of 1 mM (Fig. 1), and under these conditions the rate of influx was calculated as 1.55 nmol min⁻¹ μl of cell water⁻¹. The kinetic parameters of PALA entry into cells were determined over the concentration range of 0.1 to 40 mM. At the 2-min point the transport was linear at both the lower and upper inhibitor concentrations. Initial rates as a function of PALA concentration were linear up to 12 mM, with transport showing saturation at PALA concentrations over 20 mM. Nonlinear regression analysis of the data revealed a K_m of 14.8 mM and a V_max of 19.11 nmol min⁻¹ μl of cell water⁻¹. To study the specificity of the transport process, the influx of [³H]PALA was measured in the presence of 10 mM phosphonoacetate; the uptake was inhibited by 54% when this moiety was added (Fig. 1).

FIG. 1.

Transport of [³H]PALA by H. pylori whole cells. Measurements were carried out at 20°C in suspensions of cells in 150 mM NaCl, with a permeant concentration of 1 mM. (Inset) Effect of 10 mM phosphonoacetate on the transport of 1 mM [³H]PALA. Values are expressed relative to those for a control without added phosphonoacetate.

Metabolism of PALA.

NMR spectroscopy was employed to investigate the metabolic fate of PALA in H. pylori cells. The catabolism of PALA with the stoichiometric release of inorganic phosphate (P_i) was observed in lysates and cells by ³¹P NMR spectroscopy (Fig. 2), and the rate of P_i formation at a 10 mM PALA concentration was 2.57 μmol min⁻¹ mg⁻¹. By ¹H NMR spectroscopy, it was established that other enzyme activities also cleaved PALA, with the formation of succinate and malate as the major products (Fig. 3).

FIG. 2.

³¹P NMR spectra illustrating C-P bond cleavage in H. pylori lysates. The PALA concentration was 10 mM. The resonances corresponding to the substrate, PALA, and the product, P_i, are shown. The dashed line indicates the evolution over time of the phosphonate resonance of PALA. The time at which each spectrum was acquired is shown on the right.

FIG. 3.

¹H NMR spectra showing the metabolism of 10 mM PALA by H. pylori lysates. The resonances corresponding to the substrate, PALA, and the major products, succinate and malate, are shown. The time at which each spectrum was acquired is shown on the right.

Metabolism of PALA moieties.

To determine the enzyme activities involved in the formation of these metabolic products, the fates of phosphonoacetate, N-acetyl-l-aspartate, and l-aspartate, compounds which are also moieties of the PALA molecule, were studied by ³¹P and ¹H NMR spectroscopy. The catabolism of phosphonoacetate by H. pylori lysates yielded P_i at a rate of 2.80 μmol min⁻¹ mg⁻¹. ¹H NMR experiments also showed the production of acetate from phosphonoacetate in bacterial lysates. N-Acetyl-l-aspartate would be a product from cleavage of the PALA C-P bond, and succinate and malate were the major products formed in incubations of N-acetyl-l-aspartate with H. pylori lysates. At a 10 mM substrate concentration, a catabolic rate of 0.95 μmol min⁻¹ mg⁻¹ for N-acetyl-l-aspartate was measured in lysates. Succinate and malate were also formed from l-aspartate metabolism, and the rates of l-aspartate breakdown decreased in the presence of N-acetyl-l-aspartate, while the presence of PALA did not affect the rates of l-aspartate catabolism. N-Acetyl-l-aspartate was not metabolized by intact, metabolically competent cells, suggesting that this compound was not transported into the bacterium. To test whether this compound also inhibited ACTase once inside the cell, ACTase activity was measured in lysates in the presence of 10 mM N-acetyl-l-aspartate. Under these conditions, N-acetyl-l-aspartate inhibited ACTase activity in a dose-dependent manner, with 40% inhibition at 10 mM (Fig. 4).Control experiments employing ¹H and ³¹P NMR spectroscopy in which substrates at 10 mM concentrations were incubated without bacterial preparations indicated that there was no chemical cleavage of PALA, phosphonoacetate, N-acetyl-l-aspartate, or l-aspartate under the experimental conditions used. Controls containing only bacterial preparations without substrates showed no significant formation of P_i under the same experimental conditions.

FIG. 4.

Effects of N-acetyl-l-aspartate on ACTase activity. Aspartate and carbamoyl phosphate, each at a concentration of 5 mM, were used as substrates. Assays were performed at 37°C for 10 min, which is within the linear region of the evolution over time of the reaction. Samples were analyzed with the microtiter assay, and 100% activity was defined as the rate in the absence of N-acetyl-l-aspartate.

Substrate specificity of C-P bond cleavage activity.

To identify the type of C-P bond cleavage operative in H. pylori, several substrates characteristic of known C-P bond cleavage pathways were studied. The organophosphonates phosphonoacetaldehyde, glyphosate, and methylphosphonate were not metabolized by H. pylori lysates under the ³¹P NMR assay conditions described for PALA catabolism.

Effects of temperature and metal ions on C-P bond cleavage and approximate size of the enzyme.

The temperature dependence of phosphonoacetate C-P bond cleavage by H. pylori lysates was measured over the range of 10 to 45°C at a substrate concentration of 10 mM. The optimum temperature was 37°C, and the activity decreased at 45°C (Fig. 5). The energy of activation was calculated at 3.19 kJ mol⁻¹ from an Arrhenius plot.

FIG. 5.

Effect of temperature on C-P bond cleavage in H. pylori lysates. ³¹P NMR was employed to measure the rates of C-P bond cleavage for phosphonoacetate over the temperature range of 10 to 45°C. Rates were measured by the decrease in the resonance intensity of a 10 mM concentration of the phosphonate moiety.

The effects of various metal ions on PALA C-P bond cleavage were examined in H. pylori lysates. Enzyme activity was measured under the same conditions with metal ions added to a final concentration of 1 mM. Addition of Mg²⁺, Ca²⁺, Fe²⁺, or Zn²⁺ had no effect on activity. Owing to the paramagnetic effects of Mn²⁺ and Co²⁺, which broaden the ³¹P NMR signals, the concentration of these ions was reduced to 0.1 mM, and there was no effect on activity at this concentration. However, C-P bond cleavage activity was almost completely lost with the addition of 5 mM EDTA.

The results of membrane filtration experiments indicated the existence of C-P bond cleavage activity in the 50- to 100-kDa fraction. No significant activity was observed in fractions containing proteins less than 50 kDa or greater than 100 kDa in size.

Cellular location of enzyme activities.

The rates of enzyme activities measured in the pellet and supernatant fractions were compared to ascertain whether enzymes were located in the soluble fraction or were associated with the cell envelope. A summary of the findings is shown in Table 1. The C-P bond cleavage activity was associated with the soluble fraction, while the aspartase activity was localized in the pellet fraction.

TABLE 1.

Relative metabolism of PALA and associated moieties in supernatant and pellet fractions determined by ³¹P NMR

Substratea	% of metabolism in:
	Supernatant	Pellet
PALA	95	5
Phosphonoacetate	96	4
N-Acetyl-l-aspartate	3	97

^a The concentration of each substrate was 10 mM. 

Kinetic properties of PALA metabolism.

The substrate saturation curves for PALA, phosphonoacetate, and N-acetyl-l-aspartate were all hyperbolic and did not exhibit substrate inhibition at higher concentrations. The kinetic parameters of PALA metabolism were an apparent K_m of 10.4 mM and a V_max of 4.6 μmol min⁻¹ mg⁻¹. Apparent K_m values of 7.3 and 12.1 mM were determined for phosphonoacetate and N-acetyl-l-aspartate, respectively. The V_max was 5.2 μmol min⁻¹ mg⁻¹ for phosphonoacetate and 2.1 μmol min⁻¹ mg⁻¹ for N-acetyl-l-aspartate.

DISCUSSION

Although PALA has previously undergone trials as an antimetabolite for mammalian tumor cell lines (9), its efficacy was limited because of high-level drug resistance in different cell lines. The main mechanism for resistance of rapidly proliferating mammalian cells to PALA has been thought to be the amplification of the multifunctional CAD gene, which codes for carbamoylphosphate synthetase (EC 6.3.5.5), ACTase, and dihydroorotase (EC 3.5.2.3), causing an increase in the amounts of the corresponding mRNA and ACTase protein (32). Another mechanism of resistance is an increase in the activity of the salvage pathway of pyrimidine biosynthesis to circumvent the block in de novo pyrimidine production (40). Furthermore, PALA may not be able to enter the target cells. H. pylori ACTase activity was completely inhibited in situ at nanomolar concentrations of PALA (3), yet this compound did not have an effect on cell growth and viability; this suggested that PALA may not have been able to enter H. pylori cells. However, transport experiments with [³H]PALA indicated that it is taken up by the bacterium via a saturable system and thus was available in vivo to act on the ACTase. It was then hypothesized that H. pylori, too, possessed some mechanism for resistance to PALA. The finding that the uridine uptake inhibitor dipyridamole did not affect cell viability in the presence of PALA suggested that the pyrimidine salvage pathway did not play a major role in this resistance process in H. pylori.

The identification of an in situ saturable and temperature-dependent C-P bond cleavage activity acting on PALA suggested that once inside the cell, the inhibitor was catabolized by an enzyme-mediated process, preventing its action on ACTase, and thus was detoxified. This activity also cleaved the C-P bond of phosphonoacetate, a moiety of PALA, but not the organophosphonates phosphonoacetaldehyde, glyphosate, and methylphosphonate, demonstrating some specificity for the phosphonoacetate moiety. Another mechanism of PALA detoxification in H. pylori was found by observing the cleavage of the compound in situ by an aspartase activity also capable of catabolizing N-acetyl-l-aspartate. The catabolic rates and the cellular locations of these enzyme activities in H. pylori (Table 1) suggested an ordered mechanism for their actions on PALA. The C-P bond cleavage activity was localized in the soluble fraction, while the aspartase activity appeared to be membrane associated. Significantly, when PALA was incubated with the pellet fraction very little catabolism was measured (Table 1), but the breakdown of N-acetyl-l-aspartate and aspartate in this fraction suggested the presence of an active aspartase activity. These findings suggested that in H. pylori, the C-P bond of PALA was cleaved by one enzyme activity and that the N-acetyl-l-aspartate product was then cleaved at the C-N bond by an aspartase activity. This conclusion was supported by the observations that PALA had no effect on aspartase activity, while N-acetyl-l-aspartate competed with aspartate for this activity.

The importance of this second enzyme activity for H. pylori survival was established by the observation that N-acetyl-l-aspartate also inhibits ACTase in the bacterium (Fig. 4). Overall, then, it appeared that both the C-P bond cleavage activity and aspartase activity may have been necessary to prevent pyrimidine biosynthesis inhibition by PALA in H. pylori. Aspartase activity in H. pylori has been previously reported (20), and a gene with 73.2% similarity to aspA, the gene coding for aspartase in E. coli, has been identified in the H. pylori genome (37). The results of this study added a new putative role for aspartase in inhibitor detoxification, in addition to its function in fumarate production and energy metabolism.

Although H. pylori appears to possess an efficient mechanism for the detoxification of PALA, the bacterium would not be expected to have a PALA-specific mechanism because PALA is a synthetic compound. It is more likely that this enzyme system is a normal component of the cells and that the activity on PALA is a side activity of enzymes currently having alternate physiological roles. Several benefits may accrue to the organism as a result of its having pathways for phosphonate breakdown. These compounds may serve as a source of carbon and/or phosphorus for H. pylori. Because phosphorus assimilation is a fundamental process in bacterial physiology, the development of enzyme systems to cleave the C-P bond in phosphonates and release P_i would be beneficial to cell survival. Systems for P_i utilization in H. pylori include the Entner-Doudoroff pathway (21), the tricarboxylic acid cycle via pyruvate (23), and the ATP produced by the fumarate reductase complex (22). The identification of energy and phosphorus stores in the form of polyphosphate granules in H. pylori (2) further illustrates the importance of phosphorus assimilation in this microorganism. Another possible role for phosphonate breakdown was suggested to be in pathogenesis. As some phosphonates are primarily found in phosphonolipids of eukaryotic cells (15), it was proposed that enzymes of the phosphonate degradative pathway may aid in membrane destruction by invasive bacteria (11).

Several other microorganisms metabolize phosphonate compounds, and three different pathways for phosphonate degradation have been identified thus far: (i) a C-P lyase pathway which has a broad substrate range (12, 15, 41), (ii) a phosphonotase activity which appears to be specific for phosphonoacetaldehyde (6, 15), and (iii) a phosphonoacetate hydrolase pathway recently discovered in Pseudomonas fluorescens (17, 18). The results of the present study suggested that H. pylori may possess the last pathway, because phosphonoacetate was metabolized and substrates characteristic of the other pathways were not. Another similarity was the complete inhibition of C-P bond cleavage activity by the chelating agent EDTA, suggesting a metal dependence of this enzyme that was also seen in P. fluorescens (17), though unlike with the enzyme of this organism, metals such as Zn²⁺ did not enhance C-P bond cleavage in H. pylori lysates under the experimental conditions used. Further support for this putative assignment of the enzyme activity in H. pylori was the observation of C-P bond cleavage only in the fractions containing 50- to 100-kDa proteins, which agreed with the 80-kDa size of the phosphonoacetate hydrolase of P. fluorescens. The enzymes involved in C-P bond cleavage in the lyase pathway were found to be greater than 100 kDa in size (30), and although the 62-kDa size of phosphonoacetaldehyde hydrolase is within the 50- to 100-kDa range (6), that enzyme was shown to be unable to cleave phosphonoacetate (31). A comparison of the nucleotide sequence for phosphonoacetate hydrolase (13) with the open reading frames of the H. pylori complete genome (37) revealed no significant similarity, but this may indicate that the genes encoding such a system in P. fluorescens simply are not homologous to those in H. pylori. However, the exact nature and mechanism of C-P bond cleavage in H. pylori need to be investigated further to characterize the pathways operational in this bacterium.

The saturable nature of PALA transport also suggested that the uptake of this compound is controlled by a specific carrier or carriers. As with the metabolism of PALA, it is highly unlikely that this would be a specific uptake system for PALA but rather a broad phosphonate-substrate uptake system. This conclusion was supported by the finding that phosphonoacetate inhibits PALA transport in H. pylori. A comparison of several putative phosphonate uptake genes of the phn operon in E. coli (4) with the complete genome sequence of H. pylori (37) revealed an open reading frame in H. pylori with 72.2% identity to the phnA gene of E. coli. An open reading frame with homology to the E. coli phnA gene has also been recently identified in our laboratory (16). It was proposed that the phnA gene in E. coli was involved in phosphonate uptake (4), but recent results suggest that it does not have a role in phosphonate metabolism in E. coli (26). Although this finding does not completely preclude a role for phnA in phosphonate uptake in H. pylori, the understanding of the metabolism of C-P compounds in this bacterium would be further enhanced by molecular and genetic analyses of the genes involved in H. pylori.

In conclusion, this study has identified a physiological mechanism for the detoxification of PALA in H. pylori. Although this does not discount the possibility that resistance may also be gained through gene amplification or other processes, it revealed a novel system of tolerance to this compound thus far not demonstrated in other systems. This process may be necessary but not sufficient to prevent complete de novo pyrimidine biosynthesis inhibition, and a further study of other possible mechanisms involved would enhance the understanding of PALA resistance. The importance of studying enzyme activities in crude extracts is also demonstrated by the finding of two distinct activities acting on PALA in H. pylori. The presence of a C-P bond cleavage enzyme activity in H. pylori allows the further characterization of the physiology of this bacterium and provides a more cogent understanding of the importance of the organism’s metabolism in the overall disease process.