Introduction

The haloalkyl phosphorus compound tris(2-chloroethyl) phosphate (TCEP) has been widely used as a flame retardant or a plasticizer in plastics and forms throughout the world. This extensive use has resulted in its leakage into the environment, and thus it is now detected in indoor and outdoor air, rivers, lakes, disposal sites, drinking water, and even wild animals (Takahashi et al. 2013). TCEP has high physicochemical and microbiological stability in the environment and may exert neurotoxicity, reproductive toxicity, and carcinogenicity in rat and mouse (Takahashi et al. 2013). It has also been shown to exhibit toxic effects in aquatic fishes. This compound thus poses a serious threat to the health of humans and ecosystems. Therefore, methods for the removal of this environmental contaminant are needed.

To develop microbial techniques for the removal of TCEP from the environment, we have isolated two bacterial strains, Sphingobium sp. strain TCM1 and Sphingomonas sp. strain TDK1, that can utilize TCEP as a phosphorus source for cell growth (Takahashi et al. 2010). Strain TCM1 has a much higher TCEP degradation ability than strain TDK1. These bacterial strains are the only TCEP-degrading microorganisms reported so far, and they can also assimilate a broad range of other organophosphorus compounds for use as phosphorous sources, including tri(haloalkyl), triaryl, and trialkyl phosphorus compounds (Takahashi et al. 2010).

The utilization of organophosphorus compounds for cell growth in microorganism has been proposed to be mediated by three types of phosphoesterases: phosphotriesterase (PTE), phosphodiesterase (PDE), and phosphomonoesterase (PME) (Fig. 1) (Singh and Walker 2006). In sphingomonads, the involvement of PDE and PME is shown in the degradation pathway of tributyl phosphate (TBP) in Sphingobium sp. strain RSMS, although PTE has not identified in this strain (Rangu et al. 2014, 2016). From these findings, the metabolism of TCEP in strain TCM1 is considered to include these three types of enzymes (Fig. 1). At each step, one of the phosphate ester bonds is hydrolyzed to produce 2-chloroethanol (2-CE), and the final step also gives Pi for cell growth. Strain TCM1 cells cannot degrade 2-CE, which also exhibits some toxicity to humans and animals (National Toxicology Program 1985). Therefore, we have developed a complete microbial detoxification technique for TCEP using strain TCM1 combined with a 2-CE-degrading bacterium, Xanthobacter autotrophicus strain GJ10 (Takahashi et al. 2012). We have already identified a PTE of strain TCM1, named haloalkylphosphorus hydrolase (HAD), that can hydrolyze TCEP to produce bis(2-chloroethyl) phosphate (BCE) and 2-CE but not further degrade BCE (Fig. 1) (Abe et al. 2014). HAD can also hydrolyze a variety of other organophosphorus compounds, including tri(haloalkyl), trialkyl, and triaryl phosphates and some organophosphate pesticides (Abe et al. 2014). Conversely, the PDE and PME of the TCEP degradation pathway remain to be identified. The high TCEP elimination rate and low 2-CE generation rate of strain TCM1 suggest that the steps catalyzed by PDE, PME, or both might be rate limiting in the pathway (Takahashi et al. 2010). Identification of the PDE and PME involved in TCEP degradation in strain TCM1 is, therefore, important to develop an efficient and effective method for complete TCEP degradation using this microorganism.

Fig. 1
figure 1

Possible degradation pathway of TCEP in Sphingobium sp. strain TCM1. Enzymes: PTE phosphotriesterase, PDE phosphodiesterase, PME phosphomonoesterase, HAD haloalkylphosphorus hydrolase. Compounds: TCEP tris(2-chloroethyl) phosphate, BCEP bis(2-chloroethyl) phosphate, CEP 2-chloroethyl phosphate, 2-CE 2-chloroethanol, Pi inorganic phosphate

Full size image

Alkaline phosphatase (APase) is a metal-dependent PME that can hydrolyze a wide variety of monophosphate esters most efficiently under alkaline conditions. APases can be mainly grouped into three families, namely PhoA, PhoD, and PhoX (Ragot et al. 2015; Sebastian and Ammerman 2009). PhoA APases have been identified in various organisms ranging from bacteria to plants and animals and are usually referred to as typical APases (Lin et al. 2013). Escherichia coli APase, a representative of the PhoA family, can hydrolyze various phosphomonoesters and requires Mg2+ and Zn2+ for activity (Kim and Wyckoff 1991). The bacterial APases of the PhoA family have a classic signal peptide and are exported to the extracellular space via the Sec pathway (Pugsley 1993). PhoD gene homologs are ubiquitous in both aquatic and terrestrial ecosystems, and are more abundant in marine bacteria than genes of the other APase families (Kageyama et al. 2011; Ragot et al. 2015). The APase of Bacillus subtilis, an archetype of the PhoD family, has significant PDE activity, as well as PME activity, and is secreted into the extracellular medium through a twin-arginine translocation (Tat) pathway (Pop et al. 2002). This PhoD requires Ca2+ for activity and contains Fe3+ and Ca2+ in the active site of the crystal structure (Rodriguez et al. 2014). The enzymes of the PhoX family are further divided into two subfamilies, PhoX-I and PhoX-II, although there is likely to be no close phylogenetic linkage between them (Zaheer et al. 2009; Lee et al. 2015). The homologous genes of PhoX-I are found in various members of Alphaproteobacteria and Gammaproteobacteria, while those of PhoX-II are often found in sphingomonads of Alphaproteobacteria (Zaheer et al. 2009; Lee et al. 2015). The Pseudomonas fluorescens PhoX-I APase is activated by Ca2+, and its structure shows the presence of Fe3+ and Ca2+ in the active site (Yong et al. 2014). An enzyme of the PhoX-II family (PhoK) was first characterized in Sphingomonas sp. strain BSAR-1, and another (mAP) was recently isolated from a metagenomic library constructed with tidal flat sediments (Nilgiriwala et al. 2008; Lee et al. 2015). PhoK requires Ca2+ and Zn2+ for catalysis but not Mg2+ (Bihani et al. 2011). Although PhoK displays very low PDE activity, PhoX-II APases are more similar to nucleotide phosphodiesterases than to members of the other APase families. PhoX-I enzymes have been shown to be secreted via the Tat pathway (Wu et al. 2007; Zaheer et al. 2009), while the secretion pathway of PhoX-II APases remains unclear.

It has been reported that E. coli APase can hydrolyze 2-fluoroethyl phosphate, a compound structurally similar to 2-chloroethyl phosphate (CEP, Fig. 1) (O’Brien and Herschlag 2002), suggesting that, in strain TCM1, APase may be involved in the utilization of TCEP as a phosphorus source. In this study, we identified four putative APase genes that belong to the PhoA, PhoD, and PhoX families in the draft genome sequence of strain TCM1. Expression of these genes in E. coli and characterization of gene deletion mutants of strain TCM1 showed that two APases, named SbPhoA and SbPhoX-II, are essential for utilization of TCEP as the sole phosphorus source. In particular, SbPhoX-II is the main APase involved in TCEP utilization for cell growth.

Materials and methods

Bacterial strains, plasmids, and culture conditions

Sphingobium sp. strain TCM, which has been deposited in the NITE Biological Resource Center (NBRC, Japan) under code NBRC 112374, was used as the wild-type strain (Takahashi et al. 2010). The wild-type strain and its mutants were grown at 30 °C in lysogeny broth (LB) or A-Cl minimal medium (Takahashi et al. 2010) containing 20 μM KH2PO4 or TCEP as a sole phosphorus source. SOC medium (20 g/l tryptone, 5 g/l yeast extract, 0.5 g/l NaCl, 2.5 mM KCl, 10 mM MgSO4, 10 mM MgCl2, 20 mM glucose, pH 7.0) was used for the recovery of TCM1 transformants. When necessary, 25 μg/ml kanamycin was added to the cultivation medium. E. coli strain DH5α was used as a host for DNA manipulation, and strain BL21(DE3) (New England Biolabs, Beverly, MA, USA) was used as a host for protein expression. E. coli cells were cultivated at 30 °C or 37 °C in LB or 2× YT medium (16 g/l tryptone, 10 g/l yeast extract, 5 g/l NaCl, pH 7.0). When necessary, 100 μg/ml ampicillin was added to the cultivation medium. For solid media, 1.5% (w/v) agar was added.

Identification of APase gene homologs

The draft genome sequence of strain TCM1 (LXVX00000000) was submitted to the Rapid Annotation using Subsystem Technology (RAST) server for annotation (http://rast.nmpdr.org/rast.cgi). The predicted amino acid sequences of APase genes were further analyzed using the blastp or tblastn program against a protein or bacterial genome database (DNA Data Bank of Japan; http://ddbj.nig.ac.jp/blast/blastn?lang=ja).

Construction of expression vectors

Total DNA of strain TCM1 was prepared from overnight cultures in LB medium as described by Bickley and Owen (1995). The open reading frame of each putative APase gene (locus_tags—A7Q26_23575, A7Q26_18390, A7Q26_15460, and A7Q26_04985) was amplified by PCR using the primers listed in Table S1 and total DNA of strain TCM1 as a template with Tks Gflex DNA Polymerase or TaKaRa Ex Taq Hot Start Version DNA polymerase (Takara Bio, Shiga, Japan) under the following conditions: initial denaturation at 94 °C for 1 min, followed by 30 cycles at 98 °C for 10 s, 60 °C for 15 s, and 68 °C for 1 min, with a final extension at 68 °C for 10 min. The resulting PCR products were digested with NdeI and BamHI and ligated into the same sites of pET25b to obtain their expression vectors. After sequencing analysis of the inserts, each plasmid was introduced into E. coli BL21(DE3).

E. coli BL21(DE3) cells harboring each APase expression vector were grown in 5 ml of LB medium containing 100 μg/ml ampicillin at 30 °C for 14 h. One milliliter of the culture was transferred into 100 ml of the same medium. When the absorbance at 600 nm reached 0.5, isopropyl-1-thio-β-d-galactopyranoside (IPTG) was added to the culture medium to a final concentration of 1 mM. After cultivation for 6 h at 30 °C, the cells were harvested by centrifugation at 5000 ×g for 10 min at 4 °C, resuspended in 10 ml of 50 mM Tris-HCl (pH 9.0), and disrupted by sonication using an Ultrasonic Disruptor UD-201 (Tomy Seiko, Tokyo, Japan). The homogenate was centrifuged at 20,000×g for 30 min at 4 °C, and the supernatant and the pellet were used as a crude extract (soluble fraction) and an insoluble fraction, respectively. The proteins in these fractions were separated using SDS–12% polyacrylamide gel electrophoresis (SDS-PAGE) and then stained with Coomassie brilliant blue R-250. Protein concentration was determined by the Bradford method using a protein assay kit (Bio-Rad, Hercules, CA, USA) with bovine serum albumin as a standard.

Enzyme assays

A plate assay for APase activity of E. coli cells was carried out by incubating cells overnight at 37 °C on LB agar containing 100 μg/ml ampicillin, 40 μg/ml 5-bromo-4-chloro-3-indolyl phosphate (BCIP), and 1 mM IPTG.

APase and PDE activities of crude extracts of E. coli cells were assayed by monitoring the release of p-nitrophenol (p-NP) from p-nitrophenyl phosphate (PNPP) and bis(p-nitrophenyl) phosphate (BNPP). The reaction mixture, containing 50 mM Tris-HCl (pH 9.0) and 5 mM PNPP or BNPP, was added to wells of a 96-well plate. After incubation of the mixture at 37 °C for 5 min, crude extract was added to the wells, and the release of p-NP was monitored spectrophotometrically at 405 nm (ε = 18,500 M/cm) at 37 °C on a microplate reader (SpectraMax; Molecular Devices, Sunnyvale, CA, USA). For analysis of the effect of metal ions, activity was measured in the presence of 5 mM Zn2+, Mg2+, or Ca2+. Mg2+ and Ca2+ were added as chloride salts, while Zn2+ was added as the sulfate salt. One unit of enzyme is defined as the amount of enzyme that liberates 1 μmol of p-NP per minute.

APase activity of intact TCM1 cells was determined according to the method of Manoil (1991). Briefly, cells were collected from 1 ml of culture medium by centrifugation at 10,000×g for 10 min at 4 °C, washed once with 1 ml of ice-cold 10 mM Tris-HCl (pH 9.0), and resuspended in 1 ml of the same buffer. The OD600 of the cell suspension was measured, and then 100 μl of the suspension was combined with 0.9 ml of 1 M Tris-HCl (pH 9.0) and 100 μl of a permeabilization solution (50 μl of 0.1% SDS and 50 μl of chloroform) and kept at 37 °C for 5 min. After the addition of 100 μl of 0.4% PNPP, the suspension was kept at 37 °C for 5 min. The reaction was stopped by adding 120 μl of a stop solution (a 1:5 mix of 0.5 M EDTA (pH 9.0):1 M KH2PO4), and then the OD550 and OD420 were measured. The activity per cell (OD600) was expressed in Miller units (Miller 1972), calculated as follows: 1000 × [OD420 − (1.75 × OD550)]/T (min) × V (ml) × OD600, where T represents the length of reaction time and V represents the cell suspension volume.

Construction of gene-disruption vectors

For constructing markerless gene deletion vectors for each APase gene, approximately 1.0-kbp regions just upstream and downstream of each putative APase gene coding region were amplified by PCR using the primers listed in Table S1 and total DNA of strain TCM1 as a template with Tks Gflex DNA Polymerase (Takara Bio) under the following conditions: initial denaturation at 94 °C for 1 min, followed by 30 cycles at 98 °C for 10 s, 60 °C for 15 s, and 68 °C for 1.5 min, with a final extension at 68 °C for 10 min. The resulting PCR products were ligated into SmaI-digested pK18mobsacB using the In-Fusion HD Cloning Kit (Clontech, Palo Alto, CA, USA) to obtain the respective APase gene-disruption vectors.

Construction of deletion mutants

Strain TCM1 cells were grown in 10 ml of LB medium at 30 °C for 12 h with shaking at 169 rpm. One milliliter of the culture was transferred into 50 ml of LB medium and incubated at 30 °C for 8 h with shaking at 200 rpm. Cells were harvested by centrifugation at 5000×g for 10 min at 4 °C, washed twice in ice-cold 10% (w/v) glycerol, and resuspended in 100 μl of ice-cold 10% glycerol. The cell suspension was kept at −80 °C until use.

Five micrograms of an APase gene-disruption vector was mixed with the cell suspension and kept on ice for 1 min. The cell suspension was transferred into a precooled 2-mm-gap electroporation cuvette and subjected to a pulse of 2.5 kV in a MicroPulser electroporation apparatus (Bio-Rad). The cell suspension was then combined with 1 ml of ice-cold SOC medium and incubated at 30 °C for 6 h with shaking at 169 rpm. An aliquot of the cell suspension was plated onto LB agar medium containing 25 μg/ml kanamycin and incubated for 2 days. The cells grown on the medium were inoculated into 10 ml of LB medium and incubated for 1 day, and then 500 μl of the culture was spread onto LB agar medium containing 10% (w/v) sucrose to select gene deletion strains. The gene deletions were confirmed by PCR using the primers listed in Table S1 and total DNA of the mutants prepared using a PowerSoil DNA Isolation Kit (MO BIO Laboratories, Carlsbad, CA, USA). The PCR conditions were the same as those used for construction of the gene-disruption vectors except that the extension time was 2 min.

Characterization of mutants

Wild-type and mutant cells were grown in 100 ml of A-Cl medium containing 20 μM NaH2PO4 as the sole phosphorus source at 30 °C for 96 h with shaking at 162 rpm and collected by centrifugation at 5000×g for 10 min at 4 °C. Cells were washed with 100 ml of A-Cl medium lacking a phosphorus source, suspended in 100 ml of the same medium, and incubated at 30 °C for 96 h with shaking at 162 rpm to allow consumption of available phosphorus sources in the cells. The cells were then collected by centrifugation, washed once with the same medium, and resuspended in the same medium to an OD600 of 10. An aliquot of the cell suspension was transferred into A-Cl medium containing 20 μM NaH2PO4 or TCEP as the sole phosphorus source, and cell growth was monitored spectrophotometrically at 600 nm. APase activity of the cells was determined as described above.

Nucleotide sequence accession numbers

The nucleotide sequences of SbphoA, SbphoD1, SbphoD2, and SbphoX-II genes have been deposited at DDBJ/EMBL/GenBank under the accession numbers LC157910, LC157909, LC157912, and LC157911, respectively.

Sequence analysis

Multiple alignments were performed using Clustal Omega at EMBL-EBI (http://www.ebi.ac.uk/Tools/msa/clustalo/). Phylogenetic trees were generated using MEGA6.0 (http://www.megasoftware.net/) by the maximum-likelihood method with 1000 bootstrap replicates to verify tree topology. The amino-terminal signal sequence was predicted using SignalP 4.1 (http://www.cbs.dtu.dk/services/SignalP/) or TatP 1.0 servers (http://www.cbs.dtu.dk/services/TatP/).

Results

Identification of putative APase genes of strain TCM1

To identify APase-coding genes of Sphingobium sp. strain TCM1, we searched its genome sequence for APase gene homologs using the RAST server (Kera et al. 2016). The search revealed four APase gene homologs (locus_tags—A7Q26_23575, A7Q26_18390, A7Q26_15460, and A7Q26_04985) with open reading frames of 1548, 1125, 1665, and 1584 bp, respectively, which code for proteins composed of 516, 375, 555, and 528 amino acids, respectively (Table 1). BLAST analysis of these amino acid sequences showed that the proteins coded by genes A7Q26_23575, A7Q26_18390, A7Q26_15460, and A7Q26_04985 have the highest identity to putative APases of Sphingobium sp. strain C100, Sphingobium yanoikuyae (Pho1, UniProt accession number A0A0J9CTQ7; and Pho2, A0A085K328), and an uncharacterized protein of Sphingomonas sanxanigenens DSM 19645 that is highly homologous to a putative APase of Sphingobium sp. strain SYK-6 (UniProtKB accession no. G2ILT4), respectively (Table 1). Phylogenetic analysis of the gene products classified A7Q26_18390 into the PhoA family, A7Q26_23575 and A7Q26_04985 into the PhoD family, and A7Q26_15460 into the PhoX-II family (Fig. 2). The amino acid sequences of these products include most of the functional amino acid residues involved in the binding of catalytic metal ions and Pi (Figs. S1–S3). Therefore, we named these genes SbphoA (A7Q26_18390), SbphoD1 (A7Q26_23575), SbphoD2 (A7Q26_04985), and SbphoX-II (A7Q26_15460). SbPhoA has 26.9% amino acid identity to E. coli PhoA APase. SbPhoD1 and SbPhoD2 have 37.9 and 39% identity to B. subtilis PhoD APase. SbPhoX-II has 85.4% identity to Sphingomonas sp. BSAR-1 PhoK. Analyses performed using the SignalP and TatP servers showed that SbPhoA and SbPhoX-II contain a signal sequence for the Sec pathway and that SbPhoD2 possesses a Tat signal (Figs. S1–S3). While SbPhoD1 does not have an obvious Tat signal, it does have a Tat signal-like sequence (Fig. S3).

Table 1 Putative APase genes of Sphingobium sp. strain TCM1
Full size table
Fig. 2
figure 2

Phylogenetic tree of putative APases of Sphingobium sp. strain TCM1 and other known and putative APases. The phylogenetic analysis was performed using the maximum-likelihood method in MEGA6.0 with 1000 bootstrap trials. The numbers at nodes indicate bootstrap value percentages. Amino acid sequences from the bacteria listed below were used for the analysis (accession numbers are shown in parentheses). PhoA cluster: Bacillus subtilis subsp. subtilis str. 168 (NCBI protein accession no. NP_388822.2), Staphylococcus aureus subsp. aureus str. JKD6008 (NCBI no. WP_000953739.1), Oceanobacillus iheyensis (NCBI no. WP_011065431), Shewanella sp. AP1 (UniProtKB no. Q8RQU7), Desulfobacterium autotrophicum HRM2 (UniProtKB no. C0QFB4), Antarctic bacterium TAB5 (UniProtKB no. Q9KWY4), Cobetia amphilecti (UniProtKB no. Q1W622), Vibrio sp. G15-21 (UniProtKB no. Q93P54), Halobacterium salinarum (NCBI no. WP_010904149), Erwinia pyrifoliae DSM 12163 (UniProtKB no. D2T9N9), Escherichia coli str. K12 substr. W3110 (UniProtKB no. P00634), Pseudomonas aeruginosa PAO1 long type (NCBI no. NP_251986), Yersinia pestis biovar Microtus str. 91001 (NCBI no. NP_993417), Agrobacterium tumefaciens CCNWGS0286 (UniParc no. UPI000233382C), Ralstonia eutropha H16 (NCBI no. YP_726648), Hyphomonas neptunium ATCC15444 (UniProtKB no. Q0BWI9), and Sphingobium yanoikuyae Pho2 (NCBI no. WP_048939515). PhoD cluster: Pseudomonas aeruginosa PAO1 short type (UniProtKB no. P35482), Pseudomonas syringae (NCBI no. WP_011266682), Sphingomonas sanxanigenens DSM 19645 (UniProtKB no. W0A6N3), Bacillus subtilis PhoD (NCBI no. WP_009966461), Rhodopirellula baltica SH1 (NCBI no. NP_868602), Sphingomonas wittichii (NCBI no. WP_011951389), Sphingobium sp. C100 (NCBI no. WP_024019602), and Aphanothece halophytica (UniProtKB no. F5HRA4). PhoX-II cluster: Roseobacter sp. MED193 (NCBI no. WP_009807681), Sphingopyxis sp. MC1 (NCBI no. WP_003042625), Sphingopyxis alaskensis RB2256 (NCBI no. YP_615103), Sphingobium sp. SYK-6 (NCBI no. YP_004833865), Sphingobium yanoikuyae Pho1 (NCBI no. WP_037510231), Citrobacter rodentium ICC168 (NCBI no. YP_003364064), Sphingomonas sp. BSAR-1 (UniProtKB no. A1YYW7), and mAP (UniProtKB no. K9JA07). PhoX-I cluster: Parvularcula bermudensis (NCBI no. WP_013300011), Hahella chejuensis KCTC2396 (NCBI no. YP_437582), Pasteurella multocida (NCBI no. A1C3J6), Pseudomonas fluorescens Pf0-1 (UniProtKB no. G8Q1D5), Vibrio cholerae (UniProtKB no. Q6JJM1), Campylobacter jejuni (NCBI no. WP_021137608), Nitrosococcus oceani (NCBI no. WP_002811528), Sulfitobacter sp. NAS-14.1 (UniProtKB no. A3SXY0), Fulvimarina pelagi HTCC2506 (UniProtKB no. Q0G5E0), and Sinorhizobium meliloti 1021 (NCBI no. NP_385195)

Full size image

Expression of putative APase genes in E. coli

To determine whether the APase gene homologs code for active APases, we expressed them in E. coli BL21(DE3) and determined APase activity on an agar plate containing BCIP as a substrate (Fig. 3). APase activity was clearly observed in E. coli cells expressing SbphoA or SbphoX-II but not in those expressing SbphoD1 or SbphoD2. The activity of SbPhoX-II appeared to be much higher than that of SbPhoA. We analyzed the activity of crude extracts of the E. coli cells using PNPP as the substrate (Table 2). Similar to the plate assay, analysis of crude extracts showed significant APase activity for SbPhoA and SbPhoX-II, but not for SbPhoD1 and SbPhoD2. The activity of SbPhoX-II was approximately 6500 times higher than that of SbPhoA. These results show that at least two APase gene homologs, SbphoA and SbphoX-II, encode active APases. We also determined PDE activity of the crude extracts using BNPP as the substrate (Table 2). Compared with the control extract, significant PDE activity was observed in that of SbPhoA and SbPhoX-II. The PDE activity of SbPhoA was comparable to the APase activity, while the activity of SbPhoX-II was considerably lower, suggesting much higher specificity of SbPhoX-II toward phosphomonoesters, similar to PhoK of Sphingomonas sp. strain BSAR-1 (Bihani et al. 2011). We confirmed expression of the gene products in E. coli by SDS-PAGE (Fig. 3b). In the crude extract (soluble fraction), a likely candidate gene product was observed only for the SbphoX-II gene, while other likely candidate gene products were found in the insoluble fractions, suggesting their accumulation in inclusion bodies.

Fig. 3
figure 3

Expression of putative APase genes of Sphingobium sp. strain TCM1 in E. coli. a Plate assay of APase activity of E. coli BL21(DE3) cells expressing each putative APase gene or harboring an empty vector (pET-25b). E. coli cells were plated on LB agar medium containing 100 μg/ml ampicillin, 40 μg/ml BCIP, and 1 mM IPTG and incubated overnight at 37 °C. b SDS-PAGE analysis of the expression of putative APase genes in E. coli. Soluble (S) and insoluble (IS) proteins of E. coli BL21(DE3) cells expressing each putative APase gene or harboring an empty vector (pET-25b) were separated on SDS–12% polyacrylamide gels and stained with Coomassie brilliant blue. M molecular size markers. Arrowheads indicate the positions of the likely candidate products of putative APase genes

Full size image
Table 2 PME and PDE activities of crude extracts of E. coli cells expressing APase gene homologs from Sphingobium sp. strain TCM1
Full size table

Effect of metal ions on APase activity

To characterize the metal ion dependency of each gene product, we analyzed the effect of several divalent metal ions on the APase activity of the crude extracts (Table 2). The activity of SbPhoA was increased 6.9-fold and 1.7-fold by adding 5 mM Zn2+ and 5 mM Ca2+, respectively, while the activity of SbPhoX-II was increased 1.5-fold by adding Mg2+. Interestingly, Zn2+ markedly decreased the activity of SbPhoX-II to 19% of the value obtained without Zn2+. In addition, significant activity (12.5 mU/mg) was observed for SbPhoD1 in the presence of Zn2+, suggesting that SbphoD1 might also encode an APase. Conversely, no significant activity was obtained for SbPhoD2 with all metal ions tested, although the activities with metal ions were slightly higher than the activity of the control extract. These results suggest that SbPhoA and SbPhoD1 might require Zn2+ and SbPhoX-II might utilize Mg2+ for catalysis.

Involvement of APase genes in TCEP utilization

To determine whether the APase genes are involved in the utilization of TCEP, we created gene deletion mutants and analyzed their growth on TCEP as the sole phosphorus source (Fig. 4a, b). In a minimal medium containing 20 μM NaH2PO4 as the sole phosphorus source, the APase gene deletion mutants grew well, similar to the wild-type strain, although the growth of the SbphoA and SbphoX-II deletion mutants was slightly lower (Fig. 4a). On medium with 20 μM TCEP as the sole phosphorus source, the SbphoD1 and SbphoD2 deletion mutants grew like the wild type, while the growth of the SbphoX-II and SbphoA deletion mutants was two thirds and one fifth lower, respectively, than that of the wild type even after 120 h of cultivation (Fig. 4b). We also analyzed the APase activity of the mutant cells during cultivation (Fig. 4c). The SbphoD1 and SbphoD2 mutant cells maintained a high activity, comparable to that of the wild-type cells, during cultivation, while the activities of the SbphoA and SbphoX-II mutant cells were about two thirds and one third lower, respectively, than the activity of the wild-type cells.

Fig. 4
figure 4

Growth and APase activity of APase gene deletion mutants of Sphingobium sp. strain TCM1. The growth of wild-type and APase mutant cells on 20 μM NaH2PO4 (a) or TCEP (b) as the sole phosphorus source was determined by measuring the OD600. c APase activity of wild-type and APase mutant cells during cultivation on 20 μM TCEP as the sole phosphorus source. Open circles, wild-type; closed circles, ΔsbphoA; closed triangles, ΔsbphoD1; closed squares, ΔsbphoD2; closed diamonds, ΔsbphoX-II. These data are the averages ± standard deviations (error bars) of at least three measurements

Full size image

To further analyze the involvement of the SbphoA and SbphoX-II genes in TCEP utilization, we created an SbphoA and SbphoX-II double deletion mutant and determined its growth on TCEP (Fig. 5). The double deletion mutant grew well on 20 μM NaH2PO4 as the sole phosphorus source, although the growth was a little slower compared with that of the wild type (Fig. 5a). In contrast, no growth of the mutant was observed on 20 μM TCEP (Fig. 5b). These results clearly show that SbPhoA and SbPhoX-II, but not SbPhoD1 and SbPhoD2, are PMEs responsible for the utilization of TCEP as a phosphorus source, and that SbPhoX-II is the main PME needed for TCEP utilization.

Fig. 5
figure 5

Growth and APase activity of an SbphoA and SbphoX-II double deletion mutant of Sphingobium sp. strain TCM1. The growth of wild-type and APase double deletion mutant cells on 20 μM NaH2PO4 (a) or TCEP (b) as the sole phosphorus source was determined by measuring the OD600. Open circles, wild type; closed circles, ΔsbphoAΔsbphoX-II. These data are the averages ± standard deviations (error bars) of at least three measurements

Full size image

Discussion

In this study, we sought to identify the PME responsible for utilization of TCEP as a phosphorus source in Sphingobium sp. strain TCM1. We identified four APase gene homologs belonging to three different families, PhoA, PhoD, and PhoX, and showed that two of them, SbphoA and SbphoX-II, are involved in TCEP utilization. SbPhoX-II was identified as the main PME. Because the hydrolysis of CEP or PME is suggested to be a rate-limiting step in the TCEP degradation pathway in strain TCM1, the identification of this enzyme may facilitate development of an efficient technique for detoxification of this environmental contaminant.

Although E. coli has only one APase gene, most bacterial strains possess more than one. For example, P. fluorescens has two APase genes (phoX and phoD) (Monds et al. 2006) and B. subtilis has three (phoA, phoB, and phoD) (Antelmann et al. 2000). We also found multiple APase gene homologs in the genome sequences of several sphingomonads, such as Sphingobium sp. C100, Sphingobium japonicum UT26S, S. yanoikuyae, and S. sanxanigenens. S. yanoikuyae and S. sanxanigenens have the same set of APase gene homologs found in strain TCM1, while Sphingobium sp. C100 lacks PhoA APase and S. japonicum UT26S has only PhoX-II APase, suggesting that PhoX-II APase may be the main PME in sphingomonads. These different APase compositions among sphingomonads might be caused by differences in the availability of Pi and phosphorus compounds among habitats. The different substrate preference of SbPhoA and SbPhoX-II suggests that the multiple APases of strain TCM1 could make it possible for this bacterium to utilize various organophosphorus compounds.

We have shown that two of the four APases of strain TCM1, SbPhoA and SbPhoX-II, are involved in the utilization of TCEP as a phosphorus source, and SbPhoX-II is the main APase involved in this utilization (Fig. 4). We initially hypothesized that SbPhoA is the main PME for TCEP utilization because E. coli PhoA APase can hydrolyze haloalkyl phosphates like CEP (O’Brien and Herschlag 2002). Our study is the first to show that PhoX-II APases can also hydrolyze a haloalkyl phosphate. It should be noted that SbPhoX-II and PhoK of Sphingomonas sp. strain BSAR-1, which is also a member of the PhoX-II family (Fig. 2), have different metal requirements despite their high amino acid sequence identity (85.4%): SbPhoX-II is markedly activated by Mg2+, but not by Ca2+ or Zn2+ (Table 2), while PhoK of BSAR-1 requires Ca2+ and Zn2+ for its activity, but not Mg2+ (Bihani et al. 2011). Further studies on the enzymatic properties of SbPhoX-II and determination of its three-dimensional structure might provide insights into its substrate preference and the different metal requirement.

In the TBP-degrading bacterium Sphingobium sp. strain RSMS, two phosphoesterase proteins are identified as potential candidate enzymes involved in TBP degradation pathway (Rangu et al. 2016). These proteins are similar to the phosphohydrolase of Aliivibrio fischeri or exopolyphosphatase-like protein of Marinomonas mediterranea MMB-1 and show PME and PDE activities but not PTE activity in vitro assay. Strain TCM1 can also assimilate TBP as a phosphorus source (Takahashi et al. 2010), but we find no proteins with significant homology to the two phosphoesterases in the genome sequence, suggesting that some enzymes involved in the metabolism of TBP might be different between the strains. It is now unknown whether strain RSMS can degrade and assimilate TCEP. If possible, different types of enzymes could also be involved in their TCEP degradation pathways. Further studies are desired to reveal the degradation pathway of organophosphorus compounds in these Sphingobium strains.

PhoX APases are known to possess a Tat signal sequence (Wu et al. 2007; Lee et al. 2015). We predicted that PhoX-I APases have a Tat signal sequence, but we found that PhoX-II enzymes, including SbPhoX-II and the BSAR-1 PhoX-II APase, have a Sec signal instead of a Tat signal, except for the mAP enzyme derived from a metagenomic library (Fig. S5). This suggests that, in contrast to PhoX-I enzymes, most PhoX-II enzymes are likely transported to the extracellular space via the Sec pathway. This difference in transport pathways and the lack of significant homology between PhoX-I and PhoX-II APases (Fig. S 4) indicate that it may be appropriate to change the name of the PhoX-II family to a less confusable name, such as PhoK.

We have shown that the SbphoD1 and SbphoD2 genes are not involved in the utilization of TCEP as a phosphorus source (Figs. 4 and 5). Although it is still uncertain whether the phoD genes encode active APases due to their inactive expression in E. coli, their high homologies to known PhoD APases, the conservation of catalytic amino acid residues, and the presence of a Tat or Tat-like signal strongly indicate that they might code for active PhoD APases. In B. subtilis, PhoD APase is considered to be involved in the utilization of teichoic acid, a phosphodiester-linked glycopolymer found in the cell wall of gram-positive bacteria, as a phosphorus source for cell growth (Eder et al. 1996). Because Sphingobium sp. strain TCM1 is a gram-negative bacterium, SbPhoD1 and SbPhoD2 do not play the same role but might have other specific functions. Further analyses are needed to determine their enzymatic properties and physiological functions in strain TCM1.