Enhanced Utilization of Phosphonate and Phosphite by Klebsiella aerogenes

Abstract

Klebsiella aerogenes ATCC 9621 was able to utilize phosphonates (P_n), including aminoethylphosphonate, ethylphosphonate, methylphosphonate (MP_n), and phosphonoacetate, and inorganic phosphite (P_t) as sole sources of phosphorus (P). The products of the phn gene cluster were absolutely required for P_n breakdown and P_t oxidation to inorganic phosphate (P_i) in this organism. To determine if K. aerogenes ATCC 9621 could be engineered to enhance the utilization of P_n and P_t, a multicopy plasmid, pBI05, which carried the entire phn gene cluster, was introduced into this strain. Despite the increased dosage of the phn genes, K. aerogenes ATCC 9621(pBI05) could utilize only up to 1.1-fold more P_n and P_t than did the control strain with the parent vector alone. These results suggested that P_i, which was generated from P_n and P_t, might limit further utilization of these P compounds. Consequently, to convert the resulting P_i to polyphosphate (polyP), the plasmid pKP28, which carried the K. aerogenes ppk gene (which encodes polyP kinase), was introduced into K. aerogenes ATCC 9621(pBI05). Overexpression of the ppk gene in K. aerogenes ATCC 9621(pBI05, pKP28) resulted in a 2.5-fold increase in P_t utilization over that of the control strain. This recombinant strain also accumulated approximately sixfold more P than did the control strain when the cells were grown with MP_n as a sole source of P.

Phosphorus (P) compounds are major contaminants in industrial and municipal effluents (13). Inorganic phosphate (P_i) is recognized as one of the major nutrients contributing to the eutrophication of lakes, bays, and other natural bodies of water (13). Treatment of inorganic phosphite (P_t) associated with high-tech industries is also becoming a difficult problem (10, 15). For example, large amounts of hypophosphite (H₂PO₂⁻) have been used to reduce metal ions in chemical plating processes such as those employed in compact-disk manufacturing plants (15). After metal plating, wastewaters contain high concentrations of P_t (HPO₃²⁻) as well as organic acids which were used to adjust the pH and stabilize metal ions. Because of the very narrow market for P_t, it is necessary to oxidize P_t to P_i because it has a wider range of applications. As of now, however, there is no effective means of P_t oxidation in the presence of high concentrations of organic acids (15).

Phosphonates (P_n), which are P compounds with a direct carbon-phosphorus (C—P) bond in place of the more familiar carbon-oxygen-phosphorus ester (C—O—P) bond, are widely used as pesticides, detergent additives, antibiotics, and flame retardants (4, 17). Since P_n are relatively rare in living organisms, they are generally recalcitrant to biodegradation (17). The C—P bonds are also resistant to chemical hydrolysis and thermal decomposition, and the fate of P_n in the environment is not well understood (17).

Some bacteria are capable of utilizing P_n and P_t as sole sources of P (2, 4, 17). For example, Escherichia coli utilizes aminoethylphosphonate (AEP_n), ethylphosphonate (EP_n), methylphosphonate (MP_n), and P_t as a sole source of P when P_i is not available (2, 12). In E. coli P_n utilization requires cleavage of the C—P bond by C-P lyases (18), and P_t seems to be enzymatically oxidized to P_i before being used as a P source. The E. coli phn locus, which comprises 14 genes (named alphabetically from phnC to phnP) in an operon with a single promoter, is required for utilizing P_n and P_t (2, 12). PhnC, PhnD, and PhnE probably comprise a periplasmic binding protein-dependent P_n transporter which also transports P_t, P_i esters, and P_i (12). PhnG, PhnH, PhnI, PhnJ, PhnK, PhnL, and PhnM are likely to be components of a C-P lyase pathway or C-P lyase enzyme complex which catalyzes C—P bond cleavage and P_t oxidation (2). Two additional proteins, PhnF and PhnO, appear to be regulatory proteins. The expression of the E. coli phn genes is known to be activated under conditions of P_i limitation (2, 12).

Previously, we examined a variety of bacterial species for their ability to utilize P_n and P_t as sole sources of P (14). After examining the bacterial growth on P_n and P_t, we found that Klebsiella aerogenes ATCC 9621 was able to grow well in minimal medium containing AEP_n, EP_n, MP_n, phosphonoacetate (P_nAc), or P_t as a sole source of P (14). The growth rate of K. aerogenes ATCC 9621 was threefold higher than that of E. coli IFO13168 (B strain) in the medium with P_t. Therefore, for further study, we have also cloned the entire phn gene cluster from K. aerogenes ATCC 9621 by using E. coli phnCDE as a DNA probe (2).

The present work was undertaken to determine if K. aerogenes ATCC 9621 could be engineered to enhance the utilization of P_n and P_t as sole sources of P by increasing the dosage of the phn genes. In the course of this study, we found that P_i, which was released from P_n and P_t, apparently limited further utilization of these P compounds. Therefore, to enhance the conversion of the resulting P_i to polyP, we employed the K. aerogenes ppk gene, which encodes polyphosphate (polyP) kinase (PPK) (9); this enzyme catalyzes the formation of polyP from ATP.

MATERIALS AND METHODS

Bacterial strains and plasmids.

E. coli MV1184 (16) and K. aerogenes ATCC 9621 (14) were grown on 2× YT medium (16). Plasmid pQF50, which carried the promoterless lacZ gene, was described previously (9). pKP01 is a derivative of pUC119 (Pharmacia) containing the K. aerogenes ppk and its own promoter (9). pBR322 (16) and pSTV28 (Takara Shuzo Co., Ltd., Kyoto, Japan) were used as cloning and expression vectors. To construct pBR322MC, a 56-bp EcoRI-HindIII fragment containing a multiple cloning site of pUC118 (Takara Shuzo Co., Ltd.) was ligated to pBR322 which had been cut with EcoRI and HindIII. Plasmid pUC4K (Pharmacia) carrying a kanamycin resistance (Km^r) gene cassette was used for disrupting the chromosomal phn gene cluster. Antibiotics used for the selection of transformants were 50 mg of ampicillin, 50 mg of kanamycin, and 20 mg of chloramphenicol per liter. Standard procedures were used for plasmid preparations, restriction enzyme digestion, ligation, and agarose electrophoresis (16).

Disruption of the chromosomal phn gene cluster.

Plasmid pPI04, which carried a 14-kb BglII-SmaI fragment containing the entire K. aerogenes phn gene cluster, was described previously (14). pPI04 was digested with EcoRI and ligated with a 1.2-kb EcoRI site-flanked Km^r gene cassette from pUC4K to construct pPI02.2. K. aerogenes ATCC 9621 was then transformed with pPI02.2 by electroporation, and Km^r transformants were selected on 2× YT medium containing kanamycin. The disruption of the chromosomal phn gene cluster was confirmed by Southern hybridization (16).

P_n and P_t uptake experiments.

K. aerogenes cells were grown in 2× YT medium with shaking at 37°C for 7 h, inoculated (1% inoculum) into T₀ minimal medium (7) containing either P_n or P_t as a sole source of P. The cultures were then incubated for 24 h under the same conditions, and samples were taken at intervals for the determination of growth (measured as optical density at 600 nm [OD₆₀₀]) and P concentrations. P_n were measured by ammonium peroxidisulfate digestion (120°C, 30 min) followed by P_i measurement (14) by an ascorbic acid method (9). P_t was analyzed with a high-performance liquid chromatography system with an IC-Anion-SW column (4.6 by 50 mm; Tosoh Co., Ltd., Tokyo, Japan). The total P content of K. aerogenes cells was determined as described previously (9).

PolyP analysis.

PolyP was extracted from bacterial cells with a 4 M guanidine thiocyanate solution and bound to silicate glass powders (Gene Clean kit II; Funakoshi Co., Ltd., Tokyo, Japan). The glass powders binding polyP were washed with an ethanol-NaCl solution, and then polyP was recovered with distilled water from the glass powder solution (1). PolyP was determined as described by Crooke et al. (3). The ability to serve as a substrate for the PPK-catalyzed conversion of [¹⁴C]ADP to [¹⁴C]ATP was used to quantitate levels of polyP. The amount of [¹⁴C]ATP was determined by thin-layer chromatography followed by visualization by an image analyzer (BAS1000; Fuji Co., Ltd., Tokyo, Japan).

Chemicals.

AEP_n, MP_n, and P_nAc were purchased from Sigma Chemical Co. (St. Louis, Mo.). P_t was obtained from Aldrich Chemical Co. (Milwaukee, Wis.). EP_n and dimethylphosphite (DMP_t) were from Nacalai Tesque, Inc. (Kyoto, Japan), and [¹⁴C]ADP was from Dai-ichi Chemical Co. (Tokyo, Japan).

RESULTS AND DISCUSSION

P_n and P_t utilization.

There are two P_n degradation pathways which are commonly referred to as the phosphonatase and the C-P lyase pathways (11). They differ in regard to their substrate specificities and mechanism of C—P bond fission. Interestingly, E. coli degrades P_n solely by the C-P lyase pathway, whereas Salmonella typhimurium appears to contain genes only for the phosphonatase pathway (5). The phn genes, whose products are involved in the C-P lyase pathway, have been cloned from K. aerogenes ATCC 9621 (14).

To determine if K. aerogenes ATCC 9621 possesses solely the C-P lyase pathway for utilizing P_n and P_t, we constructed the chromosomal phn mutant strain PHN1 using a Km^r gene cassette from pUC4K (Fig. 1). K. aerogenes PHN1 could not grow in T₀ medium containing AEP_n, EP_n, MP_n, P_nAc, DMP_t, or P_t as a sole source of P (Fig. 2). To further confirm the essentiality of the phn gene cluster, a 14-kb XbaI-KpnI fragment which carried the entire K. aerogenes phn gene cluster (14) was excised from pPI04 and cloned into pBR322MC to construct pBI05 (Fig. 1). Plasmid pBI05 could restore the ability of K. aerogenes PHN1 to grow on T₀ medium containing either P_n or P_t as a sole source of P (Fig. 2). These results indicate that K. aerogenes ATCC 9621 possesses solely the C-P lyase pathway for utilizing P_n and P_t as a sole source of P.

FIG. 1.

Schematic representation of plasmids pKP28 and pBI05 (A) and the phn region of K. aerogenes PHN1 chromosomal DNA (B).

FIG. 2.

Growth of K. aerogenes strains ATCC 9621 (hatched bars), PHN1 (open bars), and PHN1(pBI05) (solid bars) in T₀ medium containing either P_n or P_t as a sole source of P. Cultures were grown for 24 h at 37°C with shaking, and cell growth was determined by OD₆₀₀ measurements. The initial concentrations of P compounds in T₀ medium were 0.5 mM. Error bars represent standard errors.

Regulation of the K. aerogenes phn locus.

To determine if K. aerogenes ATCC 9621 could be engineered to enhance the utilization of P_n and P_t as sole sources of P, P_n and P_t uptake experiments were performed with K. aerogenes ATCC 9621(pBR322MC), a control strain, and ATCC 9621(pBI05). However, despite the increased dosage of the phn gene cluster, K. aerogenes ATCC 9621(pBI05) removed only up to 1.1-fold more P_n and P_t from T₀ medium than did ATCC 9621(pBR322MC) (data not shown). Since both P_n breakdown and P_t oxidation could result in an elevated accumulation of cellular P_i, it is highly possible that P_i, which was released from P_n and P_t, limited further utilization of these P compounds. We also observed that the presence of P_i in the growth medium strongly inhibited the utilization of P_n and P_t by K. aerogenes ATCC 9621 (data not shown).

We scanned the nucleotide sequence upstream from the K. aerogenes phn gene cluster for the presence of a Pho box, the consensus sequence shared by the Pho regulon promoters, by using the consensus sequence published by Kasahara et al. [CTGTCATA(A,T)A(A,T)CTGTCA(C,T)] (6). Inspection of the promoter sequence revealed that the K. aerogenes phn gene cluster was preceded by a Pho box sequence (CTGTCATCAAACTGCGCT). There was a 14-of-18-bp match with the consensus Pho box sequence. We further confirmed that the promoter of the K. aerogenes phn gene cluster is activated under conditions of P_i limitation. A 2.1-kb XbaI-PstI fragment which carried the phn promoter was excised from pBI05 (Fig. 1) and inserted upstream from the promoterless lacZ gene (encoding β-galactosidase) in the vector pQF50 to construct pQF01. K. aerogenes ATCC 9621 was transformed with either pQF50 or pQF01, and the β-galactosidase levels were measured in K. aerogenes cells grown in T₀ medium with or without P_i. High enzyme levels were observed only with K. aerogenes ATCC 9621(pQF01) cells grown in T₀ medium without P_i (data not shown).

Conversion of P_n and P_t to polyP.

We attempted to place the phn gene cluster under the control of the tetracycline resistance (tet) promoter of pBR322 to express it constitutively. However, because of the absence of appropriate restriction sites for the phn gene cluster, this attempt has not yet been successful. As we have previously demonstrated (9), K. aerogenes is able to accumulate high levels of polyP. Therefore, it may be possible that the conversion of P_i to polyP, if enhanced, could improve the utilization of P_n and P_t in this organism.

Previously, we have cloned ppk from K. aerogenes ATCC 9621 (9). To enhance the conversion of P_i to polyP, a 2.2-kb EcoRI-HindIII fragment, which contained the K. aerogenes ppk gene and its own promoter, was excised from pKP01 (9) and cloned into a vector pSTV28 to construct pKP28 (Fig. 1). This recombinant plasmid was then introduced into K. aerogenes ATCC 9621(pBI05) by electroporation. K. aerogenes ATCC 9621(pBI05, pKP28) could remove 2.5-fold more P_t from T₀ medium containing 0.5 mM P_t than did the control strain ATCC 9621(pBR322, pSTV28) (Fig. 3). Since the growth of K. aerogenes ATCC 9621(pBI05, pKP28) was almost equivalent to that of the control strain, this recombinant strain accumulated approximately 2.1-fold more P (23 mg of P per g [dry weight] of cells) than did the control strain. The levels of polyP in K. aerogenes ATCC 9621(pBI05, pKP28) were approximately 3.5 mg of P per g [dry weight] of cells after the 24-h incubation in T₀ medium containing 0.5 mM P_t (Fig. 4). However, the time course analysis of polyP accumulation revealed that the levels reached a maximum of 12 mg of P per g (dry weight) of cells around 6 h after the start of incubation (Fig. 5). No significant amount of polyP was observed with K. aerogenes ATCC 9621(pBR322, pSTV28) and ATCC 9621(pBI05, pSTV28). K. aerogenes ATCC 9621(pBR322, pKP28) could accumulate polyP only when P_i was available (Fig. 4). K. aerogenes ATCC 9621(pBI05, pKP28) also accumulated a maximum of 10% of its dry weight as polyP (33 mg of P per g [dry weight] of cells) when the cells were grown in T₀ medium containing 0.5 mM MP_n as a sole source of P (Fig. 5). In this experiment, the total P content of K. aerogenes ATCC 9621(pBI05, pKP28) reached a maximum of 62 mg of P per g (dry weight) of cells, or approximately sixfold more than that of the control strain with the parent vectors alone (data not shown).

FIG. 3.

Time course of OD₆₀₀ (circles) and P_t concentration (squares) during growth of K. aerogenes ATCC 9621(pBR322, pSTV28) (A) and ATCC 9621(pBI05, pKP28) (B).

FIG. 4.

PolyP accumulation in K. aerogenes strains. (A) Thin-layer chromatography of [¹⁴C]ATP generated by the PPK-catalyzed conversion of [¹⁴C]ADP. [¹⁴C]ATP and [¹⁴C]ADP were resolved with 1 M formic acid and 0.4 M LiCl as a solvent system and visualized with an image analyzer (BAS1000; Fuji Co. Ltd.). Lanes 1 and 5, K. aerogenes ATCC 9621(pBR322, pSTV28); lanes 2 and 6, K. aerogenes ATCC 9621(pBI05, pSTV28); lanes 3 and 7, K. aerogenes ATCC 9621(pBR322, pKP28); lanes 4 and 8, K. aerogenes ATCC 9621(pBI05, pKP28). Cultures were grown in T₀ medium containing either P_i (lanes 1 to 4) or P_t (lanes 5 to 8) as a sole source of P. PolyP levels were determined with K. aerogenes cells sampled around 24 h after the start of incubation. (B) PolyP contents of K. aerogenes strains quantitated by the image analyzer. Error bars represent standard errors.

FIG. 5.

Time course of OD₆₀₀ (circles) and polyP content (squares) of K. aerogenes ATCC 9621(pBI05, pKP28) during growth in T₀ medium containing either 0.5 mM MP_n (A) or 0.5 mM P_t (B).

We have examined, for the first time, the possibility of genetic improvement of P_n and P_t utilization in bacteria. The phn genes were essential for P_n and P_t utilization in K. aerogenes ATCC 9621, although their roles in P_n breakdown and P_t oxidation are specifically uncharacterized. In addition, it was suggested that P_i, which was released from P_n and P_t, appeared to limit further utilization of P_n and P_t. In fact, we found that enhanced conversion of P_i to polyP could improve the utilization of P_n and P_t as sole sources of P in K. aerogenes ATCC 9621. It has been known that exopolyphosphatase plays an important role in the degradation of polyP to P_i (3, 8). Previously, we showed that the K. aerogenes ppx gene, encoding exopolyphosphatase, existed immediately downstream of the ppk gene without transcriptional termination (9). The chromosomal ppk mutant, which was constructed by inserting a Km^r gene cassette into the wild-type gene, showed a reduced polyPase activity in K. aerogenes ATCC 9621 (data not shown). We also constructed a ppk ppx double mutant of K. aerogenes ATCC 9621 and introduced pKP28 into this mutant by electroporation. When this strain was grown in T₀ medium with 0.5 mM P_t, it accumulated 12-fold more polyP than did ATCC 9621(pKP28) (data not shown). Consequently, we attempted to further improve the utilization of P_n and P_t in K. aerogenes by introducing both pBI05 and pKP28 into the ppk ppx double mutant. However, simultaneous introduction of pBI05 and pKP28 appeared to be detrimental to the ppk ppx double mutant, and this attempt has not yet been successful.

Our previous work demonstrated that high levels of P_i accumulation in E. coli were achieved by modifying the genetic regulation and increasing the dosage of the E. coli genes encoding PPK and the P_i-specific transport system (8). When P_i is available, the E. coli recombinants accumulated as much as 16% of their dry weight as P (48% as P_i), or approximately 10-fold more P than did the control strain. This level of P content was surprisingly high and even surpassed those of natural phosphorite deposits (typically 14% as P). K. aerogenes ATCC 9621(pBI05, pKP28) could accumulate up to sixfold more P than did the control strain when MP_n was available as a sole source of P. Further strain improvement would be possible if the functions of the phn gene products in K. aerogenes were specifically understood.