Pyridoxal kinase knockout of Dictyostelium complemented by the human homologue

Transcription

1 FEMS Microbiology Letters 189 (2000) 195^200 Pyridoxal kinase knockout of Dictyostelium complemented by the human homologue Kunde Guo, Peter C. Newell * Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK Received 23 May 2000; received in revised form 5 June 2000; accepted 10 June 2000 Abstract The gene (pyka) encoding pyridoxal kinase which converts pyridoxal (vitamin B 6 ) to pyridoxal phosphate was isolated from Dictyostelium discoideum using insertional mutagenesis. Cells of a pyka gene knockout grew poorly in axenic medium with low yield but growth was restored by the addition of pyridoxal phosphate. Sequencing indicated a gene, with one intron, encoding a predicted protein of 301 amino acids that was 42% identical in amino acid sequence to human pyridoxal kinase. After expression of the wild-type gene in Escherichia coli, the purified PykA protein product was shown to have pyridoxal kinase enzymatic activity with a K m of 8.7 WM for pyridoxal. Transformation of the Dictyostelium knockout mutant with the human pyridoxal kinase gene gave almost the same level of complementation as that seen using transformation with the wild-type Dictyostelium gene. Phylogenetic analysis indicated that the Dictyostelium amino acid sequence was closer to human pyridoxal kinase than to pyridoxal kinases of lower eukaryotes. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords: Pyridoxal kinase; Vitamin B 6 ; Dictyostelium 1. Introduction Pyridoxal 5P-phosphate is an essential co-factor for amino acid transamination, a reaction that is essential in an organism such as Dictyostelium that has no nutrient input during its developmental stage and relies on protein degradation for energy [1^3]. It is formed by pyridoxal kinase (EC ) which phosphorylates the various forms of vitamin B 6 (pyridoxal, pyridoxine and pyridoxamine) to pyridoxal 5P-phosphate, pyridoxine 5P-phosphate and pyridoxamine 5P-phosphate (respectively). Pyridoxine 5P-phosphate and pyridoxamine 5P-phosphate are oxidised to pyridoxal 5P-phosphate by pyridoxine 5P-phosphate (pyridoxamine 5P-phosphate) oxidase [4]. Using the insertional mutagenesis technique of restriction enzyme-mediated integration (REMI) [5], we have isolated the pyridoxal kinase gene from Dictyostelium discoideum and designated it pyka. We nd that pyka knockout strains grew very poorly without exogenous pyridoxal phosphate and failed to develop, remaining as uniform lawns of cells. Both the growth and developmental defects in the pyka null cells were complemented by the human pyridoxal kinase gene (PKH). 2. Materials and methods 2.1. Cell growth AX2 is an axenic derivative of the wild-type strain NC4 [6]. The DdPYR5-6 mutant strain DH1 (lacking uridine monophosphate synthase activity: URA 3 ) was grown in axenic medium supplemented with 20 Wg ml 31 uracil. URA transformants were selected in FM medium (Gibco BRL) lacking uracil as described previously [7,8]. NEO R strains and transformants were selected and grown in medium supplemented with 20 Wg ml 31 of geneticin (G418), and blastocidin-resistant (bsr R ) transformants were selected and grown with 5 Wg ml 31 of blasticidin S [9]. The pyka 3 knockout strain was also grown in axenic media supplemented with 1035^10 M pyridoxal 5P-phos- 37 phate (Sigma). * Corresponding author. Tel.: +44 (1865) ; Fax: +44 (1865) ; newell@bioch.ox.ac.uk 2.2. REMI mutagenesis and molecular cloning pdiv5 was used as mutagenic plasmid [5]. The plasmid / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S (00)

2 196 K. Guo, P.C. Newell / FEMS Microbiology Letters 189 (2000) 195^200 was linearised with BglII and electroporated into DH1 cells (auxotrophic for uracil) along with BglII. URA transformants were selected in FM medium lacking uracil, cloned on SM agar with Klebsiella aerogenes OXF1 and screened by eye for developmental defects. A 3.5-kb sequence anking the integrated vector in a mutant was isolated by digesting the mutant DNA with EcoRI, religating and transforming into an Escherichia coli strain DH5a. The original REMI insertion was found to be in the 5P-non-coding region of a gene identi ed as the pyridoxal kinase (pyka) gene. To test the e ect of disruption in the gene coding region, a pyka knockout strain was constructed in the developmentally competent strain AX2 by homologous recombination. A blastocidin resistance gene was inserted at the EcoRV site of the pyka coding region and, using this construct, transformants of AX2 were selected for their growth in the presence of blastocidin Plasmid constructs and transformation The pyka cdna fragment was generated by reverse transcription (RT)-PCR with the primers: PK1 (sense 5P-GAAGATCTATGGAACCAAAACTATTATCAATTC) and PK2 (antisense 5P-CCGCTCGAGTTATAATTTTT- CAGACTTAAATCTAATCTAATTTCAGA). Primer sequences corresponding to the coding sequence are shown in bold and the added restriction sites (BglII in PK1 and XhoI in PK2) in italic. Total RNA was extracted from the wild-type strain AX2 at the vegetative stage and RT-PCR carried out with a Titan1 One Tube RT-PCR System (Roche) following the manufacturer's instruction. The RT-PCR fragment, which contained the full-length amino acid sequence of PykA, was cut with BglII and XhoI and cloned into pact15gal (a gift from J. Williams, Department of Anatomy and Physiology, University of Dundee, UK) after removal of the L-galactosidase gene using digestion with BglII and XhoI. The pyka gene was driven by a Dictyostelium actin 15 promoter (actin 15: :pyka). It was linearised with ScaI and transformed into the pyka knockout strain by electroporation, with selection for growth in the presence of G418. A full-length human pyridoxal kinase (PKH) gene was ampli ed from PKH cdna (a gift from E. Kirkness, Institute for Genomic Research, Rockville, MD, USA) by PCR with the primers: PKH1 (sense 5P-GAAGATCTA- TGGAGGAGGAGTGCCGGGTGCTCTCC) and PKH2 (antisense 5P-CCCTCGAGTCACAGCACCGTGGCCTG- GACGACGAT). Primer sequences corresponding to the coding sequence are shown in bold and the added restriction sites (BglII in PKH1 and XhoI in PKH2) in italic. The PCR reaction was performed at 94³C for 30 s, 55³C for 1 min and 72³C for 1 min 30 s with 30 cycles. Similarly, the PCR product was digested with BglII and XhoI and cloned into the pact15gal with the L-galactosidase gene deleted. The human pyridoxal kinase (PKH) gene was fused downstream of the Dictyostelium actin 15 promoter (actin 15: :PKH). The construct was linearised with ScaI and transformed into the Dictyostelium pyka knockout strain, with selection for growth in the presence of G418. Both actin 15: :pyka and actin 15: :PKH constructs were con rmed by sequencing and shown to contain no mutations Bacterial expression of pyka protein The cdna fragment encoding amino acids 1^301 of pyka was ampli ed by RT-PCR as described above except with the primers PK3 and PK4. PK3 has the same coding sequence as PK1 but with a BamHI restriction site. PK4 has the same coding sequence as PK2 except for the omission of the stop code and the presence of a HindIII site. The fragment was digested with BamHI and HindIII and subcloned into the vector pet21d (Novagen) digested with the same enzymes. Sequencing con rmed that there were no mutations introduced during the ampli cation and that the His 6 -tag was fused at the C-terminal of the pyka gene. Expression of the protein (PykA) in the BL21 (DE3) host strain was induced by 1 mm IPTG when growth reached OD ^0.9. Incubation was continued for 5 h at 37³C. Bacteria were pelleted by centrifugation, resuspended in lysis bu er (50 mm Tris, ph 8.0; 100 mm KCl; 100 mm NaCl; 1 mm EDTA; 0.2 mg ml 31 lysozyme; 0.1 mg ml 31 ribonuclease A; 0.1% Tween-20 and 5% glycerol) and incubated on ice for 30 min. The lysate was then sonicated and centrifuged at Ug for 20 min. The supernatant was puri ed by Ni 2 -agarose (Qiagen) column chromatography, then aliquoted and stored at 370³C with 50% glycerol (v/v). Protein concentration was measured with a Bio-Rad kit by the method of Bradford [10]. Fig. 1. Schematic of REMI insertion and pyka knockout (KO) insertion. The dark boxes indicate the coding regions of the pyridoxal kinase gene. The REMI vector consisted of a pgem plasmid with the ura selection marker. The anking sequence was rescued using EcoRI. For the knockout experiment, a pyka construct bearing the blastocidin resistance gene was inserted into the EcoRV site of the pyka coding region by homologous recombination.

3 K. Guo, P.C. Newell / FEMS Microbiology Letters 189 (2000) 195^ Pyridoxal kinase assay The pyridoxal kinase activity of puri ed enzyme was measured by a uorimetric assay [11]. Kinase activity is de ned as the amount of pyridoxal phosphate formed (nmol min 31 ) at 37³C per mg protein. Fig. 3. Growth of parental strain AX2 and the pyka knockout mutant in axenic medium with or without added pyridoxal phosphate (PLP). 3. Results 3.1. Isolation of the pyka gene A mutant later shown to be in the pyridoxal kinase gene was isolated by the insertional mutagenesis technique of REMI [5] using a vector transformed by electroporation into the strain DH1. The mutant was found to grow on lawns of Klebsiella on SM nutrient agar but failed completely to aggregate and form fruiting bodies (agg 3 ). A 3.5-kb sequence anking the integrated vector was isolated by digesting the mutant DNA with EcoRI (an enzyme that does not cut the inserted vector) followed by ligation and transformation into E. coli. Sequencing revealed that this fragment consisted of 0.2 kb upstream of the REMI insertion site and 3.3 kb downstream. The insertion site was found to be in the 5P-non-coding region of a gene identi- ed by its sequence as pyridoxal kinase (pyka) (Fig. 1) A pyka knockout strain Fig. 2. Puri cation of recombinant D. discoideum PykA protein expressed in E. coli. (A) Proteins were separated by SDS^PAGE and stained with Coomassie blue. Lane 1, crude supernatant from E. coli expressing D. discoideum PykA. Lane 2, eluate from Ni 2 -agarose a nity column chromatography showing PykA protein (M r 35 kda). (B) Enzyme activity of recombinant pyridoxal kinase. Puri ed PykA protein was assayed uorimetrically. Results show activity as a double reciprocal plot of initial reaction rate (V) versus pyridoxal concentration. Data points are the mean values from the duplicate experiments. A knockout in the pyka coding region in strain AX2 was isolated by inserting a blastocidin resistance gene into the EcoRV site of the coding region of the gene (Fig. 1) using homologous recombination and selection for growth in the presence of blasticidin S. The insertion was con- rmed by Southern blotting (not shown). The knockout strain had the same agg 3 phenotype as the original REMI mutant Sequencing the gene The sequence of the rescued genomic DNA (GenBank

4 198 K. Guo, P.C. Newell / FEMS Microbiology Letters 189 (2000) 195^200 Fig. 4. Amino acid sequence alignment of D. discoideum pyridoxal kinase (pyka) with homologous genes from human (U89606) [13], S. scrofa (AF041255) [14], T. brucei (U96712) [15] and E. coli (U53700) [16]. Identities are shaded. accession number AF136753) indicated the presence of one intron of 88 bases consisting of greater than 96% A/T. The sequence of the open reading frame was con- rmed by isolation of a full-length cdna by RT-PCR from total RNA extracted from AX2 cells using primers designed according to the coding sequence of the genomic DNA of pyka. The pyka sequence predicts a 34.2-kDa protein with 301 amino acids and enriched in charged residues. It has a calculated pi of 5.91 and is hydrophilic [12] Expression of the protein in E. coli The pyka gene was inserted into a bacterial vector with a C-terminal His 6 -tag (see Section 2) and the recombinant PykA protein was expressed in E. coli and puri ed using Ni 2 -agarose. A strong single band of approximately 35 kda was found by SDS^PAGE analysis (Fig. 2A). Using a uorimetric assay, the puri ed recombinant PykA protein was found to have pyridoxal kinase activity and displayed a K m value of 8.7 WM for pyridoxal (Fig. 2B) PykA is required for growth and development The pyka knockout strain and AX2 parental cells were cultured in axenic medium at 22³C with shaking at 220 rpm. The mutant cells were found to grow poorly compared with the parental cells with low (50%) yield. However, exogenous pyridoxal phosphate restored the growth yield of the cells in proportion to the amount added between and M, the nal cell density with M pyridoxal phosphate being similar to that of the parental cells (Fig. 3). Cells grown on bacteria or in axenic medium subsequently failed to aggregate or develop. In contrast to the e ect on growth, added pyridoxal phosphate had no obvious e ect and even at M it could not restore development Homology to other pyridoxal kinase proteins The deduced amino acid sequence of PykA was compared with the pyridoxal kinases from other organisms, as

5 K. Guo, P.C. Newell / FEMS Microbiology Letters 189 (2000) 195^ shown in Fig. 4 [13^16]. It was found to share 28% amino acid identity with the pyridoxal kinase of E. coli (U53700) [16], and 33% with Trypanosoma brucei (U96712) [15]. Compared with the putative pyridoxal kinases of yeast and the Nematode worm (from the GenBank database; sequences not shown), pyka showed 31% amino acid identity with Schizosaccharomyces pombe (Z98981), 33% with Saccharomyces cerevisiae (U18530) and 37% with Caenorhabditis elegans (O01824). Perhaps surprisingly, it was found to share greater homology with mammals than with yeasts, sharing 40% identity with pig (Sus scrofa: AF041255) [14] and 42% identity with human pyridoxal kinase (U89606) [13] Transformation of the pyka knockout mutant with Dictyostelium pyka The coding region of the pyka cdna generated by RT- PCR was inserted into a Dictyostelium expression vector under the control of the actin 15 promoter. This construct was transformed into the pyka knockout cells and selected by G418. The transformants were found to have lost their agg 3 phenotype and developed normally to produce mature fruiting bodies of wild-type appearance on agar plates, demonstrating that a functional pyka gene was essential for aggregation. Moreover, the poor growth yields found with the pyka mutant in axenic medium were also fully complemented by transformation with the Dictyostelium gene (Fig. 5) Transformation with the human pyridoxal kinase gene When cells of the pyka mutant strain were transformed with the human pyridoxal kinase (PKH) driven by the Fig. 5. Growth of pyka 3 mutant transformed with pyka or PKH (human pyridoxal kinase gene). Both pyka and PKH were driven by the D. discoideum actin 15 promoter. Data points are the mean values from the duplicate experiments. Fig. 6. Phylogenetic relationships between pyridoxal kinases. The tree was constructed using the `neighbour-joining' method [18] based on the amino acid sequences. Branch lengths are drawn to scale with the number of amino acid substitutions per site indicated on the scale bar. The sequences were obtained from the GenBank database: E. coli (U532700), S. cerevisiae (U18530), S. pombe (Z98981), T. brucei (U96712), C. elegans (O01824), D. discoideum (this study, AF137653), human (Homo sapiens, U89606), pig (S. scrofa, AF041255). Dictyostelium actin 15 promoter (actin 15: :PKH), the agg 3 phenotype on agar plates was eliminated and normal fruiting bodies were formed in a similar manner to that seen after transformation with Dictyostelium pyka. The growth yields in axenic medium were also restored as effectively as with transformation with the Dictyostelium pyka gene. 4. Discussion The data presented demonstrate that an active pyridoxal kinase is required for growth and development of Dictyostelium. In axenic culture, growth yields of the pyka knockout mutant were dependent on the amount of exogenous pyridoxal phosphate present from to M and the mutant was unable to initiate subsequent aggregation or development, even in the presence of exogenous pyridoxal phosphate. We do not know the reason for the inability of added pyridoxal phosphate to restore development. We speculate, however, that it may be required in developing cells at a higher concentration than during growth because of its essential co-factor role in transamination, which is required for energy production from protein during development [1^3], coupled with the relatively ine cient uptake of exogenous compounds into cells during the development phase (when the process of pinocytosis used for nutrient uptake during growth does not occur).

6 200 K. Guo, P.C. Newell / FEMS Microbiology Letters 189 (2000) 195^200 The nding that the human pyridoxal kinase gene transformed into the Dictyostelium pyka knockout mutant allowed complementation of the missing pyridoxal kinase as e ciently as did the Dictyostelium wild-type pyka gene has the important consequence that this system can be exploited for structure/function studies of the human pyridoxal kinase protein. For example, transformation of the Dictyostelium cells with mutated forms of the human pyka/pkh gene allows study of the e ects of these mutations on the processes of cell growth and development. A knockout of the pyridoxal kinase gene has not been reported for mammalian cells and, being a haploid organism, Dictyostelium allows much simpler and more direct interpretation of results in such studies than would be possible in diploid mammalian systems. The sequencing results also show that the Dictyostelium PykA protein has greater identity to mammalian homologues than to other lower eukaryotes (42% to human compared to 31% to S. pombe). A phylogenetic tree constructed for pyridoxal kinases using the `neighbour-joining' method of Sartou and Nei [17] indicates the close evolutionary relationship of the Dictyostelium PykA protein to higher mammalian systems (Fig. 6). This result may seem surprising in view of the popular conception of slime moulds being primitive organisms. However, the ndings are consistent with the work of Loomis and Smith [17] who reported evidence based on molecular phylogenetic analyses that Dictyostelium had a more recent common ancestor with metazoans than yeast. Acknowledgements We thank Dr E.F. Kirkness (Institute for Genomic Research, Rockville, MD, USA) for the gift of human pyridoxal kinase cdna, Dr Bela Tiwari (Oxford University Bioinformatics Centre) for the guidance on the phylogenetic analysis of pyridoxal kinases, Dr C. Pears, Dr G.A. Lundberg and Prof. J.G. Gross for very helpful discussions, and the BBSRC for nancial support. References [1] Gregg, J.H., Hackney, A.L. and Krivanek, J.O. (1954) Nitrogen metabolism of the slime mold Dictyostelium discoideum during growth and morphogenesis. Biol. Bull. 107, 226^235. [2] Firtel, R.A. and Brackenbury, R.W. (1972) Partial characterization of several protein and amino acid metabolising enzymes in the cellular slime mold Dictyostelium discoideum. Dev. Biol. 27, 307^321. [3] Kelleher, J.K., Kelly, P.J. and Wright, B.E. (1979) Amino acid catabolism and malic enzyme in di erentiating Dictyostelium discoideum. J. Bacteriol. 138, 467^474. [4] McCormick, D.B. and Chen, H. (1999) Update on interconversions of vitamin B-6 with its coenzyme. J. Nutr. 129, 325^327. [5] Kuspa, A. and Looms, W.F. (1992) Tagging developmental genes in Dictyostelium by restriction enzyme mediated integration of plasmid DNA. Proc. Natl. Acad. Sci. USA 89, 8803^8807. [6] Watts, D.J. and Ashworth, J.M. (1970) Metabolism of the cellular slime mould Dictyostelium discoideum grown in axenic culture. Biochem. J. 119, 171^174. [7] Guo, K., Anjard, C., Harwood, A., Kim, H.-J., Newell, P.C. and Gross, J.D. (1999) A myb-related protein required for culmination in Dictyostelium. Development 126, 2813^2822. [8] Guo, K., Chang, W.-T. and Newell, P.C. (1999) Isolation of spermidine synthase gene (spsa) of Dictyostelium discoideum. Biochim. Biophys. Acta 1449, 211^216. [9] Sutoh, K. (1993) A transformation vector for Dictyostelium discoideum with a new selectable marker bsr. Plasmid 30, 150^154. [10] Bradford, M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein^dye binding. Anal. Biochem. 72, 248^254. [11] Sussmane, S. and Kootz, J. (1995) A uorometric assay for pyridoxal kinase applicable to crude extracts. Anal. Biochem. 225, 109^112. [12] Kyte, J. and Doolittle, R.F. (1982) A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157, 105^132. [13] Hanna, M.C., Turner, A.J. and Kirkness, E.F. (1997) Human pyridoxal kinase. cdna cloning, expression and modulation by ligands of the benzodiazepine receptor. J. Biol. Chem. 272, 10756^ [14] Gao, Z.G., Lau, C.K., Lo, S.C.L., Choi, S.Y., Churchich, J.E. and Kwok, F. (1998) Porcine pyridoxal kinase cdna cloning, expression and con rmation of its primary sequence. Int. J. Biochem. Cell Biol. 30, 1388^1397. [15] Scot, T.C. and Phillips, M.A. (1997) Characterization of Trypanosoma brucei pyridoxal kinase: puri cation, gene isolation and expression in Escherichia coli. Mol. Biochem. Parasitol. 88, 1^11. [16] Yang, Y., Zhao, G. and Winkler, M.E. (1996) Identi cation of the pdxk gene that encodes pyridoxine (vitamin B6) kinase in Escherichia coli K-12. FEMS Microbiol. Lett. 141, 89^95. [17] Loomis, W.F. and Smith, D.W. (1995) Consensus phylogeny of Dictyostelium. Experientia 51, 1110^1115. [18] Saitou, N. and Nei, M. (1987) The neighbour-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406^425.