Vibrio cholerae phosphatases required for the utilization of nucleotides and extracellular DNA as phosphate sources

Abstract

Phosphate is essential for life, being used in many core processes such as signal transduction and synthesis of nucleic acids. The waterborne agent of cholera, Vibrio cholerae, encounters phosphate limitation in both the aquatic environment and human intestinal tract. This bacterium can utilize extracellular DNA (eDNA) as a phosphate source, a phenotype dependent on secreted endo- and exonucleases. However, no transporter of nucleotides has been identified in V. cholerae, suggesting that in order for the organism to utilize the DNA as a phosphate source, it must first separate the phosphate and nucleoside groups before transporting phosphate into the cell. In this study, we investigated the factors required for assimilation of phosphate from eDNA. We identified PhoX, and the previously unknown proteins UshA and CpdB as the major phosphatases that allow phosphate acquisition from eDNA and nucleotides. We demonstrated separable but partially overlapping roles for the three phosphatases and showed that the activity of PhoX and CpdB is induced by phosphate limitation. Thus, this study provides mechanistic insight into how V. cholerae can acquire phosphate from extracellular DNA, which is likely to be an important phosphate source in the environment and during infection.

Figure 1

Deletion of five putative phosphatases hinders growth on eDNA as a source of phosphate. Growth on eDNA was assessed using mid‐exponential phase bacteria, which were washed twice before putting in to the test conditions. The growth medium used was MOPS‐glucose supplemented with either (A) sheared salmon sperm DNA consisting of 0.5 mM phosphate or (B) no phosphate. The Δ5 phosphatases mutant is: ΔphoX ΔushA ΔcpdB ΔVCA0608 ΔVCA0545. Shown are two biological replicates assayed on the same day. The growth assay was performed two times, with a total of four biological replicates; each experiment exhibited the same results.

Figure 2

ushA is required for growth on all 5′ nucleotides when supplied as sources of phosphate. A–D. Bacteria were pre‐grown to mid‐exponential phase in MOPS‐glucose minimal medium, supplemented with 10 mM KH₂PO ₄. Cultures were washed two times in MOPS medium containing no phosphate and inoculated into 200 μl MOPS‐glucose medium with either (A) 0.1 mM 5′dAMP, (B) 0.1 mM 5′dGMP, (C) 0.1 mM 5′dCMP, or (D) 0.1 mM 5′TMP. Strains were grown at 37°C with aeration in a 96 well plate. At least four biological replicates, assayed on at least two different days, are shown for all growth assays. Doubling times are reported in Table 1. E. For complementation, ushA was expressed from the IPTG‐inducible P_tac promoter carried on the pMMB67EH plasmid. After pre‐growth and washing as described above, strains were inoculated into 200 μl MOPS‐glucose medium with 0.1 mM 5′dGMP + 0.5 mM IPTG. The strains harboring the expression vector were grown in the presence of Ap. Strains were grown at 37°C with aeration in a 96 well plate. Four biological replicates, assayed on at least two different days, are shown.

Figure 3

UshA is required for 5′ nucleotidase activity. Wild type and ΔushA strains were grown to an OD ₆₀₀ of ∼0.5 in 10 ml LB cultures. Cultures were washed once in 10 mM Tris pH 7.5 and lysed by sonication. Lysates were mixed with a final concentration of 1 mM (A) 5′dAMP, (B) 5′dGMP, (C) 5′dCMP or (D) 5′ TMP. At 0, 5, 10 and 15 min after addition of the substrate, aliquots of the reaction were removed and mixed with 0.1 N HCl to prevent further enzymatic activity. After all samples were collected, cellular debris was removed by centrifugation and the supernatants were incubated with the ammonium molybdate solution (1% ascorbic acid and 1 N H₂SO ₄) at 45°C for 20 min. Nanomoles of phosphate released by enzymatic activity was determined by measuring the OD at 820 nm and converting to nmole through use of a standard curve. The mean and standard error of at least three replicates, assayed on at least two different days, are shown for each assay.

Figure 4

ushA is not required for growth on eDNA as a source of phosphate. Growth on eDNA was assessed using mid‐exponential phase bacteria, which were washed twice before putting in to the test conditions. The growth medium used was MOPS‐glucose supplemented with either (A) 0.5 sheared salmon sperm DNA consisting of 0.5 mM phosphate or (B) no phosphate. Shown are two biological replicates assayed on the same day. The growth assay was performed four times, with a total of eight biological replicates; each experiment exhibited the same results.

Figure 5

cpdB is required for growth on 3′AMP and 3′dGMP when supplied as sources of phosphate. A–D. Bacteria were pre‐grown to mid‐exponential phase in MOPS‐glucose minimal medium, supplemented with 10 mM KH₂PO ₄. Cultures were washed twice in MOPS medium containing no phosphate and inoculated into 200 μl MOPS‐glucose medium with either (A) 0.1 mM 3′AMP, (B) 0.1 mM 3′dGMP, (C) 0.1 mM 3′CMP, or (D) 0.1 mM 3′TMP. Strains were grown at 37°C with aeration in a 96 well plate. At least two biological replicates are shown for each curve. Doubling times are reported in Table 1. E. For complementation, cpdB was expressed from the IPTG‐inducible P_tac promoter carried on the pMMB67EH plasmid. After pre‐growth and washing as described above, strains were inoculated into 200 μl MOPS‐glucose medium with 0.1 mM 3′dGMP + 0.5 mM IPTG. The strains harboring the expression vector were grown in the presence of Ap. Strains were grown at 37°C with aeration in a 96 well plate. Four biological replicates, assayed on at least two different days, are shown.

Figure 6

CpdB is required for 3′ nucleotidase activity on certain nucleotides. Wild type and ΔcpdB strains were grown to an OD ₆₀₀ of ∼0.5 in 10 ml LB cultures. Cultures were washed once in 10 mM Tris pH 7.5 and lysed by sonication. Lysates were mixed with a final concentration of 1 mM (A) 3′AMP, (B) 3′dGMP, (C) 3′CMP, or (D) 3′ TMP. At 0, 5, 10 and 15 min after addition of the substrate, aliquots of the reaction were removed and mixed with 0.1 N HCl to prevent further enzymatic activity. After all samples were collected, cellular debris was removed by centrifugation, and the supernatants were incubated with the ammonium molybdate solution (1% ascorbic acid and 1 N H₂SO ₄) at 45°C for 20 min. Nanomoles of phosphate released by enzymatic activity was determined by measuring the OD at 820 nm and converting to nmole through use of a standard curve. The mean and standard error of at least two replicates are shown for each assay.

Figure 7

Deletion of cpdB, together with ushA, does not abolish growth on eDNA as a sole source of phosphate. This growth assay was performed using mid‐exponential phase bacteria, which were washed twice before putting into the test conditions. The growth medium used was MOPS‐glucose supplemented with either (A) sheared salmon sperm DNA consisting of 0.5 mM phosphate or (B) no phosphate source. The experiment was performed three times with a total of five biological replicates. Two biological replicates assayed in the same experiment are shown.

Figure 8

Deletion of ush A, cpd B and pho X mimics the delta 5 phosphatase mutant. This growth assay was performed using mid‐exponential phase bacteria, which were washed twice before putting into the test conditions. The growth medium used was MOPS‐glucose supplemented with either (A) sheared salmon sperm DNA consisting of 0.5 mM phosphate or (B) no phosphate source. The experiment was performed twice with two biological replicates each time. Two replicates from the same experiment are shown.

Figure 9

3′nucleotidase activity is induced by phosphate limitation. Bacterial cultures were grown to mid‐exponential phase in MOPS‐glucose medium supplemented with 10 mM KH₂PO ₄, washed twice in no phosphate MOPS‐glucose, and resuspended into two test conditions: MOPS‐glucose medium supplemented with 10 mM KH₂PO ₄ or no phosphate. After 2 h of incubation at 37°C in the test conditions, the bacteria were washed once in 10 mM Tris pH 7.5 and resuspended in 100 μl of the same buffer. The cells were mixed with assay buffer and a final concentration of 1 mM (A) 5′dAMP, (B) 5′dGMP, (C) 3′AMP, or (D) 3′ dGMP. At 0, 5, 10 and 15 min after addition of the substrate, aliquots of the reaction were removed and mixed with 0.1 N HCl to prevent further enzymatic activity. After all samples were collected, cellular debris was removed by centrifugation, and the supernatants were incubated with the ammonium molybdate solution (1% ascorbic acid and 1 N H₂SO ₄) at 45°C for 20 min. Nanomoles of phosphate released by enzymatic activity was determined by measuring the OD at 820 nm and converting to nmole through use of a standard curve. The mean and standard error of four replicates (5′dGMP and 3′AMP) or two replicates (5′dAMP and 3′dGMP) are shown for each assay.

Figure 10

Model for the utilization of eDNA as a source of phosphate in V . cholerae. Extracellular DNA is broken down by Xds and Dns in the extracellular space. Dns is an endonuclease in the EndA family of nuclease and presumably cleaves at the 3′ carbon, leaving a 5′ phosphate attached to the DNA strand. Alternatively, Xds is an exonuclease and may cleave at the 5′ carbon, leaving a 3′ phosphate. Once produced, nucleotides can pass across the outer membrane into the periplasm through porins. UshA and CpdB, which we hypothesize as located in the periplasm, remove phosphate groups from 5′ and 3′nucleotides respectively. Additionally, we hypothesize that PhoX contributes to removal of phosphate from 3′dAMP and 3′dCMP. Released phosphate can traverse the inner membrane via the Pst/PhoU system, whereas nucleosides can pass through the nucleoside transporters (e.g. NupC). Low phosphate conditions induce transcription of phosphate starvation genes such as xds and pho X, in a PhoB‐dependent manner. Additionally, CpdB activity is induced under phosphate limiting conditions.