Science in China Series C: Life Sciences © 2009

www.scichina.com life.scichina.com www.springer.com/scp www.springerlink.com

SCIENCE IN CHINA PRESS

Springer

Article

Characterization and function analysis of a Halo-alkaline-adaptable Trk K+ uptake system in Alkalimonas amylolytica strain N10 GUO YongHao1, XUE YanFen1, LIU Jun2, WANG QuanHui1 & MA YanHe1† 1 2

State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China; Department of Pharmacology and Systems Therapeutics, Mount Sinai School of Medicine, New York 10029, USA

By functional complementation of Escherichia coli mutants defective in potassium (K+) uptake, two genes that are required for K+ uptake in halo-alkaliphilic Alkalimonas amylolytica strain N10 were cloned. These two genes, Aa-trkA (1337 bp) and Aa-trkH (1452 bp), were adjacent on the A. amylolytica N10 chromosome and transcribed in opposite directions. Complementation experiments revealed that Aa-TrkA and Aa-TrkH from A. amylolytica strain N10 restored the ability to grow at low K+ concentration in E. coli ΔtrkA and ΔtrkG ΔtrkH strains, respectively. In addition, Aa-TrkAH supported the growth of an E. coli ΔsapD strain, indicating that the ATP-binding protein TrkE was dispensable for the Trk system of A. amylolytica strain N10. The net K+ uptake was detected at different pH levels and the critical NaCl concentration indicated that Aa-TrkAH is an alkaline-adaptable and partially halo-adaptable K+ transporter. Kinetics determined by heterogeneous K+ transport experiments with an E. coli ΔtrkA strain revealed that Aa-TrkAH has an alkaline pH optimum close to 8.5 or higher. Site-directed mutagenesis of Aa-TrkH showed that Phe103 and Ser229 play certain key roles in K+ selection and transportation. The molecular chaperones groES-groEL and tig promoted Aa-TrkH and Aa-TrkA overexpression in vitro. +

K uptake system, haloalkaliphile, TrkH and TrkA, alkaline-adaptable, site directed mutagenesis

Potassium (K+) is the major monovalent intracellular cation in prokaryotes, and has physiological roles in osmotic homeostasis[1], enzyme activation[2] and signaling[3]. In addition, K+ is important for the regulation of intracellular pH[1]. These roles lead naturally to the inference that K+ accumulation is crucial for haloalkaliphilic bacteria, which have to survive in an environment of high salt and high pH. The accumulation of K+ is mediated by specific transport systems, which have been intensively studied at the genetic and physiological levels in nonhalophilic and halotolerant bacteria, such as Escherichia coli[4], Vibrio alginolyticus[5,6], Bacillus subtilis[7] and halophilic Halomona elongata. The major transport systems for K+ accumulation in these organisms are the transporter Kdp (E. coli), the Ktr system (V. alginolyticus and B. subtilis) and the Trk transporter (E. coli, V. alginolyticus and H.

elongata). Kdp is a high-affinity K+-translocation P-type ATPase, and homologues have been found in many other bacteria. The Ktr transporters, however, allow for medium- to low-affinity K+ uptake, and have also been identified in many bacteria[8,9]. The Trk system is evolutionarily related to the Ktr system, and is widespread in both bacteria and archaea[10]. It has a medium to low affinity for K+. The Trk system consists of a transmembrane protein, named TrkH or TrkG[11], which is the actual K+-translocation subunit, and the cytoplasmic membrane surface protein, Received June 8, 2009; accepted June 25, 2009 doi: 10.1007/s11427-009-0132-2 † Corresponding author (email: [email protected]) Supported by the National Key Basic Research and Development Program of China (Grant Nos. 2007CB714301 and 2007CB707801), the National High Technology Research and Development Program of China (Grant No. 2007AA021306) and the National Natural Science Foundation of China (Grant No. 30621005)

Citation: GUO Y H, XUE Y F, LIU J, et al. Characterization and function analysis of a Halo-alkaline-adaptable Trk K+ Uptake system in Alkalimonas amylolytica strain N10. Sci China Ser C-Life Sci, 2009, 52(10): 949-957, doi: 10.1007/s11427-009-0132-2

TrkA, which is an NAD+ binding protein[12,13]. Dosch et al.[14] reported that, in E. coli, the TrkH system requires an ATP-binding protein named TrkE (SapD), which is thought to activate K+ transportation. The two Trk transporters recently isolated from halophilic H. elongata have been shown to play a critical role for growth in highly saline medium. Recently, proteomic analysis revealed that the expression level of membrane proteins participating in the ion transport in the haloalkaliphilic bacterium Alkalimonas amylolytica strain N10, which grows at NaCl concentrations up to 7% and with an optimum pH of 10 to 10.5[15], varies according to changes in pH and salt concentration[16]. This indicated the ion transporters are important for adaptation to high pH and salt in this haloalkaliphile. Indeed, several Na+ and K+ antiporters from this strain have been identified and their roles in promoting E. coli growth under high pH and salt concentration have been verified[17,18]. In spite of the potential importance of K+ in haloalkaliphiles, however, none of the K+ transport systems have been identified. Therefore, it would be very interesting to reveal which kinds of K+ transporters are employed by Alkalimonas amylolytica strain N10 and whether they have any specific adaptations to assist in the extreme environment experienced by haloalkaliphiles. In this study, a Trk system was successfully isolated from A. amylolytica strain N10 by growth complementation. The characterization of this system as well as its role and function in saline and pH adaptation was analyzed. Site-direction mutagenesis and protein expression of the Trk protein in vitro were also investigated, which may prove significant in revealing the K+ transport mechanism.

1 Materials and methods 1.1 Bacterial strains, growth media and growth conditions A. amylolytica strain N10 was grown under the conditions described previously[18]. The transporter-deficient E. coli strains LB650 (ΔtrkH, ΔtrkG) and transformants thereof were grown in LBK100 medium[19], LB2003 (ΔtrkA) and transformants thereof in LBK30 medium[19,20], and LB690 (ΔtrkH, ΔtrkG, ΔsapABCDF) and transformants thereof in K115 minimal mineral medium[21]. The minimal mineral K3Na111.5 medium[5], which contained 3 mmol/L K+, was used for genomic 950

library screening and K+-uptake complementation assays. When appropriate, ampicillin was added at a concentration of 100 μg/mL. All bacterial cultures were incubated at pH 7.5 and 37℃ unless otherwise noted. 1.2 Gene cloning and plasmid construction The gene library of A. amylolytica strain N10 was constructed as described previously[18] and transformed into E. coli strain LB650 (ΔtrkH, ΔtrkG). The transformants were screened for K+ uptake on K3Na111.5-ampicillin plates. Growth on this selection medium indicates the presence of K+ transport via a plasmid-encoded transporter from A. amylolytica strain N10. Clones were selected for further analysis by their ability to support the growth of E. coli strain LB650 on the K3Na111.5-ampicilin plate. Putative open-reading frames (ORFs) were identified using the ORF finder program and homologue searches were conducted using BLAST[22] and the network services of the National Center for Biotechnology Information (NCBI). Inverse PCR was used to obtain the complete trkA gene using a pair of primers (IVS-F and IVS-R) (Table 1). The chromosomal DNA fragments of A. amylolytica strain N10, partially digested with HindIII, were self-ligated and used as templates. Inverse PCR was performed in a 50 μL volume using LA-Taq polymerase (TaKaRa, Dalian, China). PCR was carried out at 94℃ for 5 min, followed by 30 cycles of 30 s at 94℃, 60 s at 65℃, 60 s at 72℃, and a final extension step of 10 min at 72℃. The inverse PCR products were purified and directly sequenced by primer walking. The gene of interest, which contained the upstream and downstream sequence, was cloned into pUC18 by the standard method[23] (Table 1). 1.3 Complementation assays in E. coli mutant strains The recombinant plasmids, pUC-Aa-trkH, pUC-Aa-trkA and pUC-Aa-trkAH, were transformed into K+ transporter-deficient E. coli strains LB650, LB2003 and LB690, respectively, for the complementation tests. The recombinant plasmids pUC-Ec-trkH containing the trkH gene from E. coli TK1001 and pUC-Ec-trkA containing the trkA gene from E. coli TK1001 could restore the K+ uptake activity at low K+ concentration of the corresponding transporter-deficient E. coli strain and were used as a positive control[24]. The pUC18 vector was used as a negative control. The test and control transformants were pre-cultured overnight in the correspond-

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

ing medium. The overnight cultures were adjusted to A600=0.5, and then spread on K3Na111.5-ampicillin plates and incubated for 18 h at 37℃. Cell growth was monitored visually. 1.4 Transport measurements K+ transport was measured as described by Tholema et al.[25]. The test transformants were grown to A600=1.5— 1.6 in LBK30 media. E. coli transformants LB2003 containing pUC-Aa-trkAH (KB2003AH) and LB650 containing pUC-Aa-trkH (KB650H) and derivatives were harvested from 100 mL cultures by centrifugation, and washed three times in DEA buffer containing 120 mmol/L diethanol amine HCl, 0.5% NaCl (W/V), pH 9.3[26], then re-suspended in the same buffer. To measure K+ transport at pH 7.5, the cells were washed 3 times in HEPES buffer containing 120 mmol/L HEPES (NaOH), 0.5% NaCl (W/V), pH 7.5, and re-suspended in the same buffer[9]. A 100 µL KCl solution of the suitable concentration was added to adjust the final K+ concentrations. The K+-containing contents of the cell pellets were resuspended in 1 mL trichloroacetic acid (5%) and frozen at −25℃. After thawing, 3 mL of CsCl solution (0.1%) was added to the trichloroacetic acid solution and heated at 90℃ for 10 min, and the denatured protein was removed by centrifugation (10 min, 12000×g). The supernatants were diluted with CsCl solution (0.1%), and analyzed by atomic absorption spectroscopy. Similar strategies were used to determine the net K+ uptake

in 120 mmol/L Tricine buffer at pH 8.0, TAPS buffer at pH 8.5 and Ampso buffer at pH 9.0, respectively. The same strategies were used to detect the net Rb+-uptake at different pH, except that 5% (V/V) HNO3 was substituted with 0.1% CsCl (W/V). 1.5 Site-directed mutagenesis Site-directed mutations of Phe103, Phe223, Ser107 and Ser229 of Aa-trkH were introduced using PCR and a site-directed mutation kit (Stratagene). Using this method the Ser codon AGT and Phe codon TTT were individually replaced by GCT (alanine), GAT (aspartate), AAA (lysine), CTT (leucine) and CAT (histidine). The variants growth was tested on K3Na111.5 medium containing 100 μg/mL ampicilin. 1.6 Overexpression of Aa-TrkA and Aa-TrkH in vitro The E. coli protein expression system pET28a/E. coli BL21 (DE3) was used (Novagen, Darmstadt, Germany). The entire Aa-trkA and trkH genes were individually cloned into the pET28a vector. After the transformants were cultured in LB medium to A600=0.6, 0.5 mmol/L IPTG was added into the culture to induce protein expression. TaKaRa’s chaperone plasmid set, which contained pG-KJE8, pGro7, pKJE7, pG-Tf2 and pTf16 (TaKaRa), was used to assist protein overexpression. Molecular chaperone expression was induced by 10 mg/L arabinose or 5 ng/L tetracycline.

Table 1 Bacterial strains, plasmids and primers Strains or plasmid Relevant phenotype or genotype A. amylolytica N10 Extreme alkaliphiles, gram negative LB650 TK1001 ΔtrkH: :CamrΔtrkG: :Kanr LB2003 TK1001 ΔtrkA LB690 TK1001 ΔtrkG::KmR ΔtrkH::CmR ΔsapABCDF pUC18 carrying A. amylolyticus N10 trkH gene which contained 436 bp of upstream sequence pUC-Aa-trkH and 350 bp of downstream sequence. pUC18 carrying A. amylolyticus N10 trkA gene which contained 214 bp of upstream sequence pUC-Aa-trkA and 165 bp of downstream sequence. pUC18 carrying A. amylolyticus N10 trkAH gene which contained the upstream sequence and pUC-Aa-trkAH the downstream sequence. pUC-Ec-trkH pUC18 carrying a 2340 bp trkH fragment gene from E. coli strain pUC-Ec-trkA pUC18 carrying a 1941 bp trkA fragment gene from E.coli strain Primers Primer IVS-F IVS-R TRKH-F TRKH-R TRKA-F TRKA-R

Primer sequence TCGATTTGCTGCAAGGAGGTGAAGTCGA AAGCACCACGTTTGATAAGCACGATGG ACCGAATTCTGCCGAAAGTGCTGGCTGA (EcoRI) AAATCTAGAAGGTGAAGTCGATATCGC (HindIII) CCATCTAGAGACCGCAGCCTGCTT (HindIII) CAAAAGCTTCAGCTCTGGCCTCTGCTC(XbaI) GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

Source or reference [15] [5] [5] [21] This study This study This study This study This study Production Aa-trkA 3′-end Aa-trkH Aa-trkA

951

1.7 Data analysis Transport parameters for different types of transformants were determined at least three times. The given data shown are representative examples. Kinetic parameters for the uptake of K+ were determined by plotting the data using the non-linear regression mode of the Graphpad software (http://www.graphpad.com). To calculate the K+ content of the cells (in nmol per mg of dry cell weight), we assumed that a cell suspension with A600= 1.0, measured using a Unico 7200 spectrophotometer, contained 0.33 mg of dry cell weight per mL. The sequence data has been submitted to the GenBank database under accession No. FJ589782.

2 Results and discussion

TrkH from E. coli. The incomplete putative trkA gene was found at the 3′ end of the 3.5 kb insert. We used inverse PCR to obtain the complete trkA gene. An 8-kb inverse PCR product was obtained and by primer-walk sequencing the full trkA ORF was revealed. This ORF is 1337 bp in length, encoding a 459-amino acid protein with a calculated molecular mass of 50.5 kD. The deduced protein sequence showed 67.8% identity to the NAD+/NADH binding protein TrkA from E. coli (24). It was designated as Aa-TrkA as shown in Figure 1. The K+ uptake genes Aa-trkA and Aa-trkH in A. amylolytica strain N10 are adjacent but reversely transcribed. This gene arrangement is different from the Trk systems characterized in other strains, where the genes are either dispersed in the genome or tandemly arranged and co-transcribed.

2.1 Cloning of trk transporter genes from A. amylolytica strain N10 The genomic library of A. amylolytica strain N10 was screened for colony forming cells at pH 7.5 on K3Na111.5-ampicilin plates. One plasmid, designated as pK1, survived and was picked for sequence analysis by PCR. The results revealed that the DNA insert was 3524 bp in length, and included two complete ORFs and one incomplete ORF. As shown in Figure 1, the first complete ORF (38 to 1516 bp from the 5′ end) was 1479 bp in length, encoding a 492-amino acid peptide with a calculated molecular mass of 56.2 kD. By BLAST search, this deduced protein sequence displays 67% identity to an X-pro dipeptidase (prolidase) gene pepQ. The second complete ORF (1558 to 3009 bp from the 5′ end), Aa-trkH, was 1452 bp in length, encoding a 483-amino acid protein with a calculated molecular mass of 52.9 kD. This deduced that the amino acid sequence has 60.5% identity to the transmembrane protein

Figure 1 Gene organization of Aa-trkAH at the insert fragment of pK1 from A. amylolytica N10 genomic DNA. The end of the 3524 bp insert is the BamHI locus and the downstream 1110 bp was amplified by inverse-PCR.

2.2 Functional assays using the cloned Aa-Trk genes The functions of the Aa-trkH and Aa-trkA genes in the growth of E. coli chromosomal K+-uptake mutant strains were examined by complementation assays. trkH and trkG deficient E. coli LB650, trkA deficient E. coli LB2003 and trkH, trkG and sapD deficient E. coli LB690 did not grow on the K3Na111.5 plate. Figure 2 shows that pUC-Aa-trkH and pUC-Aa-trkA, as well as the positive controls pUC-Ec-trkH and pUC-Ec-trkA, all

Figure 2 Complementation capacities of A. amylolytica N10 genes in low concentration K+ sensitive E. coli LB650 (A), LB2003 (B) and LB690 (C). Plate assays were conducted as described in Section 1. The test plasmids were pUC-Aa-trkA (Aa-trkA), pUC-Aa-trkH (Aa-trkH) and pUC-Aa-trkAH (Aa-trkAH). The positive control was pUC-Ec-trkA and pUC-Ec-trkH. The negative control was pUC18. 952

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

complemented the corresponding E. coli mutants growing on the low K+ plate. As an E. coli Trk system consists of several components, this result suggests that functional hybrids were formed between E. coli and A. amylolytica N10 trk gene products. Figure 2A shows that LB650 supported by Aa-trkH had a lower growth rate than Ec-trkH and Aa-trkAH, probably because the hybrid between Aa-TrkH and Ec-TrkA had relatively low activity than that between Aa-TrkAH and Ec-TrkAH. Unexpectedly, the product encoded by pUC-Aa-trkAH supported the growth of strain LB690 at 3 mmol/L K+ (Figure 2C). The K+ TrkH transporter system needs the ATP binding protein TrkE encoded by the sapD gene (25). TrkE independence was also observed for the Va-TrkH system in V. alginolyticus[6]. Nakamura et al. suggested that TrkH could borrow other ATP-binding subunits from ABC transporters to explain this phenomenon. It seems that TrkH has weak “borrowing” ability. When trkH was located on the chromosome, the lower copy of TrkH showed full dependency on TrkE. On the other hand, the multi-copy TrkH encoded by the cloned plasmid gene showed the quantity accumulation effect.

transport activity was evident in the pH range 7.5—8.5, increasing linearly over the range 7.5—8.0. Higher pH values could not be tested due to the limitations of the

2.3 Wild-type Aa-TrkAH has high K+ transport activity with a high pH optimum and high osmotic pressure Assays were initiated to characterize the K+ transport properties of chromosomally encoded A. amylolytica TrkAH using atomic absorption spectrophotometry of E. coli LB2003 expressing Aa-TrkAH from plasmid pUC-Aa-trkAH. The transformant containing the empty pUC18 vector and the pUC-Ec-trkAH vector was used as a control. The cells were depleted of more than 85% cytoplasmic K+ by washing 3 times with DEA buffer. The washing procedure had no deleterious effect on K+ transport, and all cells were able to accumulate about 350 nmol of K+ per mg dry weight, which corresponded to the K+ content found in exponentially growing cells of E. coli. Although not shown, wild-type A. amylolytica Aa-TrkAH, compared to positive control Ec-TrkAH, supported the rapid growth of E. coli LB2003 in the K3 medium at high pH and high osmotic pressure (created using NaCl). It was speculated that Aa-TrkAH was adaptable to the haloalkaliphilc environment because it came from one haloalkaliphile. The K+ transport activity of Aa-TrkAH was evaluated by detecting the net K+ uptake. Figure 3 shows that wild-type Aa-TrkaAH K+

Figure 3 Net K+ uptake by K+-depleted cells of strain LB2003 containing plasmid pUC-Aa-trkAH (●), pUC-Ec-trkAH (□) and pUC18 (△). Cells were grown at the LBK30 medium described in Materials and methods. K+-depleted, energized cells were prepared as described in Materials and methods. At t=0,1 mmol/L KCl was added to the cell suspension. At the time points indicated on the abscissa, cells from a 1-mL sample were centrifuged through silicone oil. The K+ content of the pellet was analyzed by flame photometry. A: the detection buffer is pH 7.5 (HEPES 120 mmol/L, 0.5% NaCl, NaOH), B: the detection buffer is pH 8.0 (Tricine 120 mmol/L, 0.5% NaCl, NaOH), C: the detection buffer is pH 8.5 (TAPS 120 mmol/L, 0.5% NaCl, NaOH).

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

953

heterologous E. coli system. Based on these results, Aa-TrkAH obviously has a wider pH profile than the positive control Ec-TrkAH. At pH 8.0, Aa-TrkAH also accumulated an abundance of intracellular K+, but the Ec-TrkAH activity declined very quickly. At pH 8.5, the detected K+ uptake value of Aa-TrkAH declined slightly. At both pH 8.0 and 8.5, the K+ accumulation value of Aa-TrkAH was about twice that of Ec-TrkAH. It was concluded that the difference in pH value between Aa-TrkAH and Ec-TrkAH was 0.5 to 1. The optimum pH of Aa-TrkAH may be 8.0 to 8.5. Based on growth experiments, it was known that E. coli LB2003 cannot survive in the media containing 3.5% NaCl, even if adequate K+ was added into the medium (data not shown). The critical NaCl concentration that inhibited E. coli growth was used to evaluate Aa-TrkAH activity under hyper-osmotic pressure. Figure 4 shows that the critical NaCl concentration severely weakened the K+ transport activity of Aa-TrkAH and Ec-TrkAH at any pH. The detected value of intracellular K+ concentration decreased nearly one order of magnitude. Aa-TrkAH exhibited relatively high activity for K+ uptake, especially, at pH 8.0 and pH 8.5, compared to Ec-TrkAH. The K+ uptake activity of Aa-TrkAH was not significantly different from that of Ec-TrkAH at high NaCl concentration and it had obvious good alkali-adaptability and slight halo-adaptability. 2.4 Assay of the transport kinetics of Aa-TrkAH The activity of the Aa-TrkAH protein was studied as a function of [K+] at pH 7.5 in E. coli LB2003. The results exhibit Michaels-Menten kinetics with an apparent Km of 0.31 mmol/L and Vmax of 129.3 nmol of K+/min·(mg of dry weight) (Figure 5). At pH 8.0, transport kinetics analysis of Aa-TrkAH revealed a Km value of 0.17 mmol/L and a Vmax of 125.6 nmol of K+/min·(mg of dry weight). Even at pH 8.5 (the limit for the assay), Aa-TrkAH exhibited a Km of 0.13 mmol/L and a Vmax of 75 nmol of K+/min·(mg of dry weight). The detected Vmax of Aa-TrkAH was lower than that of E. coli TrkH which was detected by Dosch DC using the same method. Kinetic analysis showed that Aa-TrkAH had the lowest Km value at pH 8.5, and the Vmax of K+ transport was half that at pH 8.0. It is known that the intracellular cations of E. coli efflux quickly if the environmental pH reaches 8.5. Therefore, the intracellular K+ concentration detected was in fact the K+ uptake minus the K+ efflux caused by the high pH buffer. As a consequence, the 954

Figure 4 Net K+ uptake by K+-depleted cells of strain LB2003 containing plasmid pUC-Aa-trkAH (●), pUC-Ec-trkAH (□) and pUC18 (△) under hyper osmotic pressure. K+ uptake by plasmid containing E. coli LB2003 cells was determined as described in the legend to Figure 3 except that, at t=−1.5 min, cells were treated with3.5% NaCl (W/V). A: pH 7.5, B: pH 8.0, C: pH 8.5.

transport velocity measured by the intracellular K+ concentration at pH 8.5 was significantly lower than the actual velocity. Our results indicated that about 50% K+ were extruded under the test conditions, Kraegeloh et al.[26] reported 70%, so the actual Aa-TrkAH K+ transport velocity at pH 8.5 may be double the detected value. The Aa-TrkAH optimum pH may be reached at 8.5.

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

2.5 Aa-TrkH-Phe103 and Ser229 variants had altered activity

ring), Thr (-OH) and Leu (aliphatic), at these two sites. The growth results showed that Phe103Asp, Phe103Lys, Phe103His and all of the Ser229 variants did not support growth of LB650 in the K3Na111.5 media (Figure 6). The Phe103Thr variant supported the growth of LB650 in K3Na111.5 media as well as the wild type, and the Phe103Leu variant had a slightly lower growth rate. These results indicated that Ser229 is particularly critical for Aa-TrkH K+ transport. Because Ser229 is located at the junction between the helix loop and the transmembrane helix, alteration of the helix direction may not be possible in the Ser229 variants. We thought that Ser229 may be one of the K+ binding center amino acids and constructed the hydrogen bond with other amino acids by its hydroxide radical to select and transport K+ because of its conservation and indispensability. In contrast to Ser229, the conservation of the Phe103 residue was less important for Aa-TrkH activity, and its role can be taken over by Thr and Leu. However, when this site was changed into a small molecular Ala or charged

To identify the conserved amino acids, 6 TrkH amino acid sequences, including 3 TrkH sequences that have been experimentally verified to have K+ transport activity, i.e., E. coli, V. alginolyticus, H. elongata and one KtrB from V. alginolyticus[5,6,14,26,27], were aligned with Aa-TrkH. The results showed that 20 amino acids were strictly conserved, consisting of 11 Gly residues, two Phe residues, two Ser residues and another 5 residues. Tholema et al.[25] investigated the roles of four Gly residues (namely Gly70, Gly185, Gly290 and Gly402, corresponding to Gly113, Gly222, Gly320 and Gly448 in Aa-TrkH, respectively) in KtrB, which is the closest homologue of TrkH, and verified these Gly residues play key roles in substrate selection. Considering the strict conservation of the two Phe residues and two Ser residues, i.e., Phe103, Phe223, Ser107 and Ser229, we wanted to know whether these residues also play a key role in K+ transport. Firstly, the 4 amino acids were substituted by Ala. The Ser107Ala and Phe223Ala variants supported the growth of LB650 in the K3Na111.5 media as well as the wild type, whereas the Phe103Ala variant had a decreased growth rate and the Ser229Ala variant did not support the growth of LB650 in K3Na111.5 media. This indicates that K+ transport ability was at least partially damaged in Phe103Ala and Ser229Ala variants. To further investigate the role of Phe103 and Ser229, a half-saturation mutation was used to introduce Asp (negative charge), Lys (positive charge), His (imidazole

Figure 6 Growth of transformed E. coli as a measure of K+ uptake activity of Aa-TrkH variants. A, Aa-TrkH position 103 Phe variants; B, Aa-TrkH position 229 Ser variants. Aa-trkH plasmid-containing E. coli LB650 cells were grown in the minimal medium with glucose as the carbon source at the 3 mmol/L K+ concentrations. (○) Aa-trkH (positive control); (●) Ala; (■) His; (▲) Leu; (◆) Thr; (▼) Lys; (△) Asp.

Figure 5 Kinetic analysis of K+ transport via Aa-TrkAH in E. coli mutation strains. Transport buffer was HEPES (pH 7.5, NaOH ○), Tricine (pH 8.0, NaOH □), TAPS (pH 8.5, NaOH △). Transport was started by adding potassium to the assay. The substrate KCl concentration was 0.1, 0.5, 1, 2, 5 and 10 mmol/L. After KCl was added into the supernate, a 1-mL sample was centrifuged at different time points and then analyzed by flame photometry.

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

955

residue, either positively- or negatively-charged, the K+ transport of Aa-TrkH was completely abolished, indicating that this site is also critical for Aa-TrkH function. To further investigate the activity of the Phe103 variant, we analyzed the net K+ and Rb+ uptake of the Phe103Ala, Phe103Asp and Phe103Thr variants (Figure 7). The results showed that Phe103Ala Rb+ transport activity was not damaged compared to the positive control Aa-TrkAH (Figure 7B). While the other variants, including Phe103Thr, which had full K+ transport activity, lost Rb+ transport activity. This suggested that Phe at this position is responsible for anion selection rather than anion transport. It is challenging to determine the complex K+ uptake mechanism of the TrkH system without the crystal structure. To date, no crystal structure for TrkH or TrkA has been reported. One major reason for this is that it

Figure 7 One Phe103 residue Aa-TrkH half-saturation variants mediate K+ and Rb+ uptake. (■) Phe103Thr variants; (▲) Phe103Ala variants; (▼) Phe103Asp variants; (□) positive control Aa-TrkH; (▽) negative control empty plasmid pUC18. A: K+ uptake; B: Rb+ uptake. The K+-depleted cells of strain LB650 were used. Cells were grown at the LBK100 medium as described in Materials and methods. K+ or Rb+ uptake by energized Aa-trkH plasmid-containing E. coli LB650 cells was determined as described under “Experimental Procedures.” At t=0, 1 mmol/L KCl or 2 mmol/L RbCl was added to the cell suspension. The detection buffer pH was 8.0 as described in Figure 3 legend. 956

has proven difficult to overexpress the two membrane proteins in vitro. 2.6 Over-expression of Aa-TrkH and Aa-TrkA in vitro The difficulty in overexpression of membrane proteins is the bottleneck for further characterization of the Trk systems. We have tried to overexpress Aa-TrkH and Aa-TrkA in E. coli. Although no Aa-TrkA was secreted into the supernatant, appreciable quantities of Aa-TrkA were expressed in inclusion bodies. Aa-TrkH failed to be overexpressed. Molecular chaperones have been reported to be able to improve expression and solubility of exogenous proteins. In this study, we tried to use this method in Aa-TrkH and Aa-TrkA expression. Five plasmids containing different molecular chaperones, namely pG-KJE8, pGro7, pKJE7, pG-Tf2 and pTf16, were transformed into E. coli BL21 (DE3) containing Aa-trkH and Aa-trkA. The coexpression results showed that pGr07, encoding the molecular chaperones groESgroEL, was able to assist Aa-TrkA overexpression as a soluble protein (Figure 8), and pGr07 and pTf16, encoding the molecular chaperone tig, were able to assist Aa-TrkH overexpression. Overexpressed Aa-TrkH did

Figure 8 The high expression of Aa-TrkH and Aa-TrkA in the E. coli BL21 assisted by molecular chaperone. The expression of Aa-trkH and Aa-trkA located on the pET28a was induced by 0.5 mmol/L IPTG at 16℃ for 12 h. The plasmids pGr07 contained molecular chaperone groES-groEL and ptf16 contained molecular chaperone tig were preceding transformed into BL21. The expression of the two chaperones was induced by 0.5 mg/L arabinose. 1: marker; 2: empty plasmid pET28a; 3: The expression of Aa-TrkA without molecular chaperone; 4: The co-expression of Aa-TrkA and molecular chaperone groES-groEL; 5: The expression of Aa-TrkH without molecular chaperone; 6: The co-expression of Aa-TrkH and molecular chaperone groES-groEL; 7: The co-expression of Aa-TrkH and molecular chaperone tig. The cell was broken by ultrasonics. 10 μL supernate was used for PAGE.

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

not form inclusion bodies. These results are important because no information on protein expression of Trk system proteins has been previously reported. Furthermore, we tried to obtain the correctly folded Aa-TrkA protein by urea denaturation and renaturation, but the efficiency was very low. The overexpression of Aa-TrkH 1 2

Epstein W. The roles and regulation of potassium in bacteria. Prog

and Aa-TrkA in vitro was an important first step towards crystallizing the two proteins and the findings from this study open up possibilities for future work. We thank Dr. Evert P. Bakker for kind providing us with E. coli strains LB650, LB690 and LB2003. 15

nov., and Alkalimonas delamerensis gen. nov., sp. nov., novel alka-

Suelter C H. Enzymes activated by monovalent cations. Science, 1970,

liphilic bacteria from soda lakes in China and East Africa. Extremo-

168: 789—795 3

4

philes, 2004, 8: 193—200

Prince W S, Villarejo M R, Osmotic control of proU transcription is

16

proteomic responses of Alkalimonas amylolytica N10 to different

scription complex. Biol Chem, 1990, 265: 17673—17679

external pHs with combination strategy of de novo peptide sequencing.

Horton R M, Hunt H D, Ho S N, et al. Engineering hybrid genes

Proteomics, 2009, 9: 1254—1273 17

from the soda lake alkaliphile Alkalimonas amylolytica N10 com-

Nakamura T, Yuda R, Unemoto T, et al. KtrAB, a New Type of Bac-

plement an alkali-sensitive Escherichia coli mutant. Microbiology

180: 3491—3494

(Reading, England), 2007, 153: 2168—2179 18

gene cluster and characterization of the Trk K(+)-uptake system of

monas amylolytica is adaptive for the extreme environment. J Bacte-

Vibrio alginolyticus. Microbiology (Reading, England), 1998, 144:

riol, 2005, 187: 7589—7595 19

pression and primary characterization of Vibrio parahaemolyticus

uptake systems in Bacillus subtilis and their role in adaptation to

K+/H+ antiporter genes in Escherichia coli. Mol Microbiol, 2006, 59: 651—663

Roosild T P, Miller S, Booth I R, et al. A mechanism of regulating

20

cherichia coli. Arch Microbiol, 1997, 167: 126—136

tional switch. Cell, 2002, 109: 781—791 Tholema N, Bakker E P, Suzuki A, et al. Change to alanine of one out

21

+

+

+

nitude decrease in the affinities for both K and Na of the Na de-

systems Trk(H) and Trk(G) from Escherichia coli K-12. Microbiol-

pendent K+ uptake system KtrAB from Vibrio alginolyticus. FEBS letters, 1999, 450: 217—220

ogy (Reading, England), 2001, 147: 2991—3003 22

Durell S R, Hao Y, Nakamura T, et al. Evolutionary relationship between K(+) channels and symporters. Biophys J, 1999, 77: 775—

Harms C, Domoto Y, Celik C, et al. Identification of the ABC protein SapD as the subunit that confers ATP dependence to the K+-uptake

of four selectivity filter glycines in KtrB causes a two orders of mag-

10

Stumpe S, Bakker E P. Requirement of a large K+-uptake capacity and of extracytoplasmic protease activity for protamine resistance of Es-

transmembrane potassium flux through a ligand-mediated conforma9

Radchenko M V, Waditee R, Oshimi S, et al. Cloning, functional ex-

Holtmann G, Bakker E P, Uozumi N, et al. KtrAB and KtrCD: Two K+ hypertonicity. J Bacteriol, 2003, 185: 1289—1298

8

Liu J, Xue Y, Wang Q, et al. The activity profile of the NhaD-type Na+(Li+)/H+ antiporter from the soda Lake Haloalkaliphile Alkali-

Nakamura T, Yamamuro N, Stumpe S, et al. Cloning of the trkAH

2281—2289 7

Wei Y, Liu J, Ma Y, et al. Three putative cation/proton antiporters

tension, Gene, 1989, 77: 61—68 terial K+-Uptake System from Vibrio alginolyticus. J Bacteriol, 1998, 6

Wang Q, Han H, Xue Y, et al. Exploring membrane and cytoplasm

mediated through direct action of potassium glutamate on the tran-

without the use of restriction enzymes: Gene splicing by overlap ex5

Ma Y, Xue Y, Grant W D, et al. Alkalimonas amylolytica gen. nov., sp.

Nucleic Acid Res Mol Biol, 2003, 75: 293—320

Altschul S F, Gish W, Miller W, et al., Basic local alignment search tool. J Mol Biol, 1990, 215: 403—410

23

Joseph S, Russell D W. Molecular cloning: A laboratory manual. 3rd

24

Stumpe S, Schmid R, Stephens D L, et al. Identification of OmpT as

788

ed. New York: Cold Spring Habor Laboratory Press, 2001

11

Rhoads D B, Epstein W. Energy coupling to net K+ transport in Escherichia coli K-12. J Biol Chem, 1977, 252: 1394—1401

the protease that hydrolyzes the antimicrobial peptide protamine be-

12

Schlosser A, Hamann A, Bossemeyer D, et al. NAD+ binding to the

fore it enters growing cells of Escherichia coli. J Bacteriol, 1998, 180:

Escherichia coli K(+)-uptake protein TrkA and sequence similarity between TrkA and domains of a family of dehydrogenases suggest a

13

Tholema N, Vor der Bruggen M, Maser P, et al. All four putative se-

role for NAD+ in bacterial transport. Mol microbiol, 1993, 9:

lectivity filter glycine residues in KtrB are essential for high affinity

533—543

and selective K+ uptake by the KtrAB system from Vibrio alginolyti+

Bossemeyer D, Borchard A, Dosch D C, et al. K -transport protein TrkA of Escherichia coli is a peripheral membrane protein that re-

14

4002—4006 25

cus. J Biol Chem, 2005, 280: 41146—41154 26

Kraegeloh A, Amendt B, Kunte H J. Potassium transport in a halo-

quires other trk gene products for attachment to the cytoplasmic

philic member of the bacteria domain: Identification and characteri-

membrane. J Biol Chem, 1989, 264: 16403—16410

zation of the K+ uptake systems TrkH and TrkI from Halomonas elongata DSM 2581T. J Bacteriol, 2005, 187: 1036—1043

Dosch D C, Helmer G L, Sutton S H, et al. Genetic analysis of potassium transport loci in Escherichia coli: Evidence for three consti-

27

Schlosser A, Meldorf M, Stumpe S, et al. TrkH and its homolog, TrkG,

tutive systems mediating uptake potassium. J Bacteriol, 1991, 173:

determine the specificity and kinetics of cation transport by the Trk

687—696

system of Escherichia coli. J Bacteriol, 1995, 177: 1908—1910

GUO Y H, et al. Sci China Ser C-Life Sci | Oct. 2009 | vol. 52 | no. 10 | 949-957

957