- Email: [email protected]
Cloning, sequencing and analysis of a gene encoding Escherichia coli proline dehydrogenase
ELSEVIER
FEMS Microbiology
Letters 127 (1995) 235-242
Cloning, sequencing and analysis of a gene encoding Escherichia cob proline dehydrogenase Mian Xia, Yuxian Zhu *, Xiaofeng Cao, Lingtao You, Zhangliang
Chen
The NationalLaboratoryof ProteinEngineering and Plant GeneticEngineering, Peking Unil,ersity, Beijing 100871, People’s Republicof China Received 8 December
1994; revised 31 January
1995; accepted
14 February
1995
Abstract Using a genomic subtraction technique, we cloned a DNA sequence that is present in wild-type Escherichia coli strain CSH4 but is missing in a presumptive proline dehydrogenase deletion mutant RM2. Experimental evidence indicated that the cloned fragment codes for proline dehydrogenase (EC 1.5.99.8) since RM2 cells transformed with a plasmid containing this sequence was able to survive on minimal medium supplemented with proline as the sole nitrogen and carbon sources. The
cloned DNA fragment has an open reading frame of 3942 bp and encodes a protein of 1313 amino acids with a calculated M, of 143808. The deduced amino acid sequence of the E. coli proline dehydrogenase has an 84.9% homology to the previously reported Salmonella typhimurium putA gene but it is 111 amino acids longer at the C-terminal than the latter. Keywords: Escherichia coli; Proline dehydrogenase activity
1. Introduction Proline dehydrogenase (ProDH) is a flavoprotein associated with Escherichia coli plasma membrane [1,2]. It is both the put repressor and an FAD-dependent proline dehydrogenase that catalyses the oxidation of L-proline to A’-pyrroline carboxylic acid (P5C) as well as an NAD+-dependent P5C dehydrogenase that converts P5C to glutamic acid [3]. It was reported that ProDH activity interacts with the membrane-associated respiratory chain and this activity is required for a bacteria to utilize L-proline [4]. Studies of the S. typhimurium putA gene, which encodes for
* Corresponding author. Tel.: + 86 (1) 250 1847; Fax: +86 (1) 250 1843; e-mail: [email protected] 0378-1097/95/$09.50 0 1995 Federation SSDI 0378- 1097(95)00067-4
of European
Microbiological
a functional ProDH, has shown that its expression is autogenously regulated and redox-modulated as well [5-71. In Drosophila melanogaster, researchers found that ProDH activity is related to motor activity since mutant flies with lowered mitochondrial ProDH activities behaved sluggishly [S]. However, most studies have focused on characterization of the enzyme and regulation of its function [9,10]. E. coli strain RM2 cannot grow on the minimal medium utilizing proline as the sole source of carbon and nitrogen, while CSH4 grows normally on this medium. It is believed that at least putA gene is missing in RM2. Since genomic subtraction technique provides a useful tool for cloning genes corresponding to a deletion mutant [ 111, we applied it to clone the putA gene based on genetic analysis of the two E. coli strains. By using sheared biotinylated Societies. All rights reserved
236
DNA gested CHS4, tional
M. Xia et al. / FEMS Microbiology Letters 127 (1995) 235-242
prepared from mutant RM2 and Suu3AI dinon-biotinylated DNA extracted from strain we were able to identify and clone a funcE. coli putA gene.
2. Materials and methods
2.1. DNA preparation and subtraction Genomic DNA from both E. coli strain CSH4 and the RM2 were isolated and purified by CsCl gradient centrifugation [ 121, Genomic subtraction were carried out as previously reported ill]. An excess (50 pg) of sonicated, biotinylated, deletion mutant RM2 DNA was denatured in the presence of 0.5 pg Sau3ALdigested wild-type CSH4 DNA and allowed to reassociate for 48 h at 68” C. Since DNA molecules from CSH4 that corresponded to the deleted region in RM2 had no biotinylated complementary strands with which to hybridize, these fragments cannot form hybrids. In the next step, the biotinylated DNA and any DNA that had reassociated with it were removed from the sample by incubating with streptavidin (Vector labs). The unbound DNA, which was enriched for sequences that are missing in the deletion mutant, is collected and rehybridized with another batch of freshly biotinylated mutant DNA.
amplified in a thermal cycler (Perkin-Elmer/Cetus) using the shorter adaptor strand as primers. The parameters for each of the 35 cycles were as follows: 94” C, 30 s; 55” C, 1 min; 72” C, 3 min. PCR not only produced enough DNA molecules for further experiment, but also eliminated contaminations from subtracting reactions since only wild-type DNA fragment which possessed Sau3AI-specific ends was able to be amplified. 2.3. Southern hybridization The major PCR product was purified and denatured in the presence of one adaptor molecule and labelled with [32P]dCTP using the Klenow fragment of DNA polymerase I. The labelled probe was then used to hybridize genomic DNA isolated from either CSH4 or RM2, digested to completion with HindIII. The same probe was also used to screen a wild-type CSH4 genomic library constructed in ZAP11 (Zhu et al., unpublished results). Nylon membranes (Gene screen, NEN) were used for both blots. Single positive plaques were obtained after two rounds of hybridization. E. coli cell hosting recombinant plasmid was isolated by in vivo excision upon superinfection with a helper phage R408.
Lane
A
B
2.2. PCR amplification After three rounds of subtraction, the remaining DNA was amplified through PCR. A Sau3AI adaptor was prepared by annealing the synthetic oligonucleotides GACACTCTCGAGACATCACCGTCC and GATCGGACGGTGATGTCTCGAGAGTG. The latter oligonucleotides were phosphorylated at the 5’ end using T4 polynucleotide kinase (New England biolabs). An equimolar amount of the two strands was mixed and heated to 65” C for 5 min in a heating block and then allowed to cool to room temperature within the block. One end of the resulting adaptor is Suu3AI compatible with a 5’ phosphate while the other is not. This is to ensure that only one adaptor molecule can be ligated onto each end of a Suz&U fragment. DNA molecules capped with adaptors (l/10 of the ligation reaction) were
Fig. 1. Southern blot analysis of RM2 and CSH4 genomic DNA using 32P-Iabelled PCR product which was generated after three rounds of subtractions as a probe. Lane 1, 10 pg of HindIII-digested RM2 DNA, Lane 2, 10 pg of HindIII-digested CHS4 DNA. Arrows indicate standard migration of 7.5 kb (upper) and 2 kb (lower).
237
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
2.4. Assay of ProDH activity Competent RM2 cells were transformed with the plasmid containing a putative putA coding sequence and selected for their ability to grow on minimal medium which contained 50 pg/ml of ampicillin and 1 mM of IPTG and utilized proline as the sole sources of carbon and nitrogen. Proline dehydrogenase activities of the transformant as well as CSH4 and RM2 cells were then determined following previously described methods [l]. 2.5. DNA sequencing
and computer analysis
DNA sequencing was carried out using CsClpurified, double-stranded plasmid DNA by the dideoxynucleotide chain termination method using
Fig. 2. Nucleotide and predicted amino acid sequence EMBL data bank under accession number X78340.
Sequenase II (USB) following the manufacturer’s instructions. Deletion of the plasmid template was generated by a combination of Exonuclease III and Sl nuclease. DNA sequence was compiled and analysed using the GCG package [13].
3. Results and discussion 3.1. Cloning the E. coli putA gene using a probe that was generated from subtraction When genomic DNA isolated from CSH4 cells were subjected to three rounds of subtraction by large quantities of RM2 DNA molecules, the pattern of PCR products simplified down to only a few bands (data not shown). The main product (a 1.3-kb
of E. coli putA. The nucleotide
sequence
reported
here has been submitted
to the
238
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
Fig. 2 (continued).
band on agarose gel) was used to probe DNA isolated from the two different bacterial strains to make sure that such a generated probe had no representation in RM2 cells. As shown in Fig. 1, a 6.0-kb HindIII-digested DNA fragment from CSH4 hybridized to the probe specifically (lane B) while there is no signal from RM2 DNA (lane A). This was taken as a good indication that the probe was pure enough to be used for cloning of the putA gene via plaque hybridization against a CSH4 partial genomic library. Although genetic complementation in a bacterial cell is quite common, and the DNA sequences of putC as well as part of p&P and putA from E. coli K12 were known when we began our work [14-161, we failed to clone putA gene using this method. Several independent reasons may account for the
failure. First, the size of putA gene makes it a bad candidate to be cloned as a functional unit; second, the growth rate of RM2 cells on minimal medium is too low to permit reliable screen; third, for some reason, the rate of RM2 cell transformation was much lower comparing to those standard cell lines. In our view, the genomic subtraction technique is a good way to clone those genes that do not have a ready probe. It may not be as convenient as a screen library, but the probe we generated was pure enough to permit cloning of the puti gene. 3.2. Verification of the cloned sequence by enzyme activity assay To verify that some of the positive plaques may contain the entire put4 coding sequence, we trans-
M. Xia et al. /FEMS
Microbiology Letters 127 (1995) 235-242
formed competent RM2 cells with a plasmid excised from the vector and then selected for cells that grew on minimal medium. ProDH activities of the transformants as well as CSH4 and RM2 cells were
239
determined. While the ProDH activity of RM2 cells was indeed very low (too low to be detected by our assay), the transformed RM2 cells (ZX16) possessed higher activity than that of wild-type CSH4 cells
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
660 PVINPAEPKDIVGWGREATESEVEQALQNAVNQAPVWFATVLMEDQM . . . . .. .. . . . .. .. .. . .. . . .. . .. . .. . . .. . . .. .. . . . . . .. . . .. . . . . . . . . . . . .. . .-TEEVICHVE.GDKAD.DK.VRQ.FQIGSP.RTMDAS..GCL.NKL.D...RDR ET....NG-NVLATVGA.GRED.DR.VKS.QQGQKI.ASMTAM..SR..R..VDILRERN ..S. .LDNSTLAEIACASA.-Q....VAS.RETFAS.KE..VS...RVML.YQA.LKEHH
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
720 QQLIGLLVREAGKTFSNAIAEVREAVDFLHYYAGQVRDDFDNETHRPLGPWCISPWNFP .. . .. . . . . . .. . . .. . . .. . . ... . . .. . . .. . . . .. . . . . .. . .. . .. . . . . .. . . . . VL.ATMESMN...I.TH.LLDTEVSIKA.K.F..KIHGQ.TYTRRE.I.VCGQ.I...G. DE.AK.ETLDT..AY.ESTVDIVTGA.V.E....PL.ET.VYTRRE.,.V.AG.GA..Y. DE.AKIVSS.L... .ED.KGD.WRGIEWEHLM.ETVENVARNIDQ...VC.G.T.F...
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
780 LAIFTGQIAAALAAGNSVLAKPAEQTSLIAAQGIAILLEARTAGRRATVAGTGRNRRRPA . .. .. . . .. . .. . . .. .. . . .. . . ... .. .. . . .. . . . . .GVPPGWQLLPGRGETVGAQ .IL.IWK.G...SC..T.IV...... P.T.LYMASLIK..GPP.WNV.P.Y.STAGAAI IQ.ALWKS.P..... .AMIF..S.V.P.T.LKLAE.YS..LPD.VFNVLP.V.AETGQYL AM.PLWMFPL.I.C..AFIL..S..VP.TSVRLAELF . ..GAPKGVLQ.VHG.KEQVDQL
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
840 YRRCACTRVMFTGSTEVATLLQRNIATRLDAQGRPIPLIAETGGMNRMIVDSSALTEQW LTAD.RV..................................................... SSHMD1DK.S.. . . ...GK.IKEAAG---KSNLKRVT.--.L..KSPC..FAD.DLDSA. TEHPGIAK.S...GVASG---KKVM.NSAASSLKEVTM--.L..KSPL...DAD.DLAAD LKHPQVKA.S.V..VA.GQYVYHTGT----.HNKRVQSF.--.AK.H.VIMPD.DKA..I
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
900 VDVLASAFDSAGQRCSALRVLCLQDDIAEHTL~LRRPHGGVSVGESGRLTTDIGPVIDS . . .. .. ..RQRRTTLFR.... . .. . .. . .. . . .. ..GAMAECRM.NP............. EFAHQGV.FHQ..I.V.ASR.FVEES.YDE---FV.. ---S.ERAKKYV.GNPLDSG.SQ IAMM.NF.-.S..V.NGT..FVPAKCK.AFEQ.I.A.VE-RIRA.DVFDPQ.NF..LVSF SNLVGASVGA.....M.IS.AV.VGAAR.W-IPEI.DALAK.RP.PWDDSGASY....NP
ECPUTA STP'JTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
960 EAKANIERHIQTMRAKGRPVSQ~RENSDDAQEWQTGTFVMPTLIELENFAELEKEVFGP . . . . . .. .. . . .. . . . .. ..F....................................... GPQIDK.Q.AKILDLIESGKKEG.KLECGGGR-.NK.F..Q..VFNVTDEMRIA..I... PHRD.VL.Y.AKGKEE.AR.LCGG--DVLKGDGFDN.AW.A..VFTSDDMTIVRE.I... Q...R...L.GQGVEE.AQLLLDG.GV--KVEGYPD.NW.G...FARPDM.IYRE.....
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
1020 VLHWRYNRNQLAELIEQINASGYGVTLGVHTRIDE . . . . . .. . . . . ..DV..........L.......... .QQIMKFKS--ID.V.KRA.NTP..LAA..F.KL.R .MSILT.ESED--.V.RRA.DTD..LAA.IV.ADLN ..CLA--EVDS.EQA.RL..E.P..NGTSIF.SSGA
Fig. 3. Comparison of deduced E. coli putA protein amino acid sequences 660-1060 with various regions of the Salmonella typhimurium putA, rat cytosolic aldehyde dehydrogenase, E. coli betaine aldehyde dehydrogenase and Pseudomonas aeruginosa methylmalonate-semialdehyde dehydrogenase. Multiple alignment was done with the GCG programs. The sources of the sequences were taken from refs. [5,17-191.
240 Table 1 Determination
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
of the enzyme activity of the cloned puti
Strains
Activity of ProDH (nmol min -1 (mg protein)- ’ 1
CSH4 RM2 ZX16 a
12.6 f 1.0
a ZX16 is the RM2 cell line that contains putative puti gene.
gene
the plasmid
of a
(Table 1). This encouraged us to think that we actually cloned sequences which encode for a functional ProDH.
cofactor binding and may be needed for dehydrogenase activity 131.It shares a 84.9% identity with the Salmonella typhimurium putA protein although it is 111 amino acids longer at the carboxyl terminal end. The significance of these 333 extra nucleotides present at the 3’ end of the E. coli put.4 gene is not known. Since E. coli ProDH is such a large protein with multiple functions and binds a number of substrates, such as proline, P5C, DNA, FAD and NAD, the relationships between its structure and function could be complicated. Further analysis on functional domains are warranted for a better understanding of proline utilization.
3.3. Sequence analysis of the E. coli put4 gene
Acknowledgements
The complete nucleotide (nt) sequences of our E. coli putA gene was shown together with the deduced amino acid sequence in Fig. 2. The largest ORF extends from the ATG start codon, which is preceded by a potential RBS at nt 295 to the stop codon at nt 4234, which is immediately followed by multiple stop codons in all reading frames. The major ORF codes for a polypeptide of 1313 amino acid with a calculated M, of 143 808. These values are in accordance with biochemical data obtained earlier [2]. The E. coli putA gene is rich in G and C (about 60%). Codon usage analysis revealed a high frequency (65%) of codon ending with G or C which correlates well with the reported frequency of E. coli codon usage. Examination of the deduced E. coli proline dehydrogenase amino acid sequence shows that some part of it is homologous to various previously characterized dehydrogenases including that of bacteria, fungi and eukaryotic organisms. Computer comparisons of deduced amino acid sequence of the E. coli put4 with EMBL database revealed sequence similarities to a number of proteins. The N-terminus of the E. coii putA protein has 100% identity to deduced amino acid sequence of the S. typhimurium put4 protein [5], indicating that this sequence may have an important function. The region from amino acid 660-1060 has also some similarities (25-44%) to a number of NAD-dependent dehydrogenases from a wide variety of organisms (Fig. 3). Previous work suggests that this domain may be important for
We thank Miss Xiaowen Liang for her assistance in database search and computer sequence analysis. We are grateful to Miss Qun Jing and Xiaohua Li for typing the manuscript. This work was supported by a grant from the national high-technology ‘863’ program of China to Y.-X.Z.
References [l] Dendinger, S. and Brill, W.J. (1970) Regulation of proline degradation in Salmonella typhimurium. J. Bacterial. 130, 144-1.52. [2] Scarpulla, R.C. and Soffer, R.L. (1978) Membrane-bound proline dehydrogenase from Escherichia coli. J. Biol. Chem. 17, 5997-6001. [3] Brown, E.D. and Wood, J.R. (1993) Conformational change and membrane association of the putA protein are coincident with reduction of its FAD cofactor by proline. J. Biol. Chem. 268,8972-8979. [4] Wood, J.M. (1987) Membrane association of proline dehydrogenase in Escherichia coli is redox dependent. Proc. Natl. Acad. Sci. USA 84, 373-377. [5] Allen, S.W., Senti-Willis, A. and Maloy, S.R. (1993) DNA sequence of the putA gene from Salmonella typhimurium: a functional membrane-associated dehydrogenase that binds DNA. Nucleic Acids Res. 21, 1676. [6] De Spicer, P.O. and Maloy, S. (1993) puti protein, a membrane-associated flavin dehydrogenase, acts as a redoxdependent transcriptional regulator. Proc. Natl. Acad. Sci. USA 90, 4295-4298. [7] Maloy, S. and Stewart, V. (1993) Autogenous regulation of gene expression. J. Bacterial. 175, 307-316. [8] Hayward, D.C., Delaney, S.J., Campbell, H.D., Ghysen, A., Benzer, S., Kasprzak, A.B., Cotsell, J.N., Young. LG. and
M. Xia et al. /FEMS
[9]
[lo]
[ll]
[12]
[13]
Microbiology
Miklos, G.L. (1993) The sluggish-A gene of Drosophila melanogaster is expressed in the nerve system and encodes proline oxidase, a mitochondrial enzyme involved in glutamate biosynthesis. Proc. Natl. Acad. Sci. USA 90, 29792983. Brown, E.D. and Wood, J.R. (1992) Redesigned purification yields a fully functional puti protein dimer from Escherichia coli. J. Biol. Chem. 267, 13086-13092. Zhu, Y.-X., Shearer, G. and Kohl, D.H. (1992) Proline fed to intact soybean plants influences acetylene reducing activity, and proline content and metabolism in bacteroids. Plant Physiol. 98, 1020-1028. Li, W.-B, Gruber, C.E., Lin, J.J., Lim, R., D’AJdessio, J.M. and Jessee, J.A. (1994) The isolation of differentially expressed genes in fibroblast growth factor stimulated BC3Hl cells by subtractive hybridization. BioTechniques 16, 722729. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Springs Harbor Press, Cold Springs Harbor, NY. Devereaux, J., Haeberli, P. and Smithies, 0. (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387-395.
Letters 127 (1995) 235-242
241
[14] Mogi, T., Yamamoto, H., Nakao, T., Yamato, I. and Anraku, Y. (1986) Genetic and physical characterization of putP, the proline carrier gene of Escherichia coli K12. Mol. Gen. Genet. 202, 35-41. [15] Nakao, T., Yamato, I. and Anraku, Y. (1988) Mapping of the multiple regulatory sites for putP and putA expression in the pufC region of Escherichia coli. Mol. Gen. Genet. 214, 379-388. [16] Chen, L.-M. and Maloy, S. (1991) Regulation of proline utilization in enteric bacteria: cloning and characterization of the Klebsiella put control region. J. Bacterial. 173, 783-790. [17] Hsu, L.C., Chang, W.C. and Yoshida, A. (1989) Genomic structure of the human cytosolic aldehyde dehydrogenase gene. Genomics 5, 857-865. [18] Dunn, T.J., Koleske, A.J., Lindahl, R. and Pitot, H.C. (1989) Phenobarbital-inducible aldehyde dehydrogenase in the rat: cDNA sequence and regulation of the mRNA by pentobarbital in responsive rats. J. Biol. Chem. 264, 13057-13065. [19] Hirji, R. and Selvaraj, G. (1991) Characterization of an E. coli gene encoding betaine aldehyde dehydrogenase (BADH): structural similarity to mammalian ALDHs and plant BADH. Gene 103, 45-52.
FEMS Microbiology
Letters 127 (1995) 235-242
Cloning, sequencing and analysis of a gene encoding Escherichia cob proline dehydrogenase Mian Xia, Yuxian Zhu *, Xiaofeng Cao, Lingtao You, Zhangliang
Chen
The NationalLaboratoryof ProteinEngineering and Plant GeneticEngineering, Peking Unil,ersity, Beijing 100871, People’s Republicof China Received 8 December
1994; revised 31 January
1995; accepted
14 February
1995
Abstract Using a genomic subtraction technique, we cloned a DNA sequence that is present in wild-type Escherichia coli strain CSH4 but is missing in a presumptive proline dehydrogenase deletion mutant RM2. Experimental evidence indicated that the cloned fragment codes for proline dehydrogenase (EC 1.5.99.8) since RM2 cells transformed with a plasmid containing this sequence was able to survive on minimal medium supplemented with proline as the sole nitrogen and carbon sources. The
cloned DNA fragment has an open reading frame of 3942 bp and encodes a protein of 1313 amino acids with a calculated M, of 143808. The deduced amino acid sequence of the E. coli proline dehydrogenase has an 84.9% homology to the previously reported Salmonella typhimurium putA gene but it is 111 amino acids longer at the C-terminal than the latter. Keywords: Escherichia coli; Proline dehydrogenase activity
1. Introduction Proline dehydrogenase (ProDH) is a flavoprotein associated with Escherichia coli plasma membrane [1,2]. It is both the put repressor and an FAD-dependent proline dehydrogenase that catalyses the oxidation of L-proline to A’-pyrroline carboxylic acid (P5C) as well as an NAD+-dependent P5C dehydrogenase that converts P5C to glutamic acid [3]. It was reported that ProDH activity interacts with the membrane-associated respiratory chain and this activity is required for a bacteria to utilize L-proline [4]. Studies of the S. typhimurium putA gene, which encodes for
* Corresponding author. Tel.: + 86 (1) 250 1847; Fax: +86 (1) 250 1843; e-mail: [email protected] 0378-1097/95/$09.50 0 1995 Federation SSDI 0378- 1097(95)00067-4
of European
Microbiological
a functional ProDH, has shown that its expression is autogenously regulated and redox-modulated as well [5-71. In Drosophila melanogaster, researchers found that ProDH activity is related to motor activity since mutant flies with lowered mitochondrial ProDH activities behaved sluggishly [S]. However, most studies have focused on characterization of the enzyme and regulation of its function [9,10]. E. coli strain RM2 cannot grow on the minimal medium utilizing proline as the sole source of carbon and nitrogen, while CSH4 grows normally on this medium. It is believed that at least putA gene is missing in RM2. Since genomic subtraction technique provides a useful tool for cloning genes corresponding to a deletion mutant [ 111, we applied it to clone the putA gene based on genetic analysis of the two E. coli strains. By using sheared biotinylated Societies. All rights reserved
236
DNA gested CHS4, tional
M. Xia et al. / FEMS Microbiology Letters 127 (1995) 235-242
prepared from mutant RM2 and Suu3AI dinon-biotinylated DNA extracted from strain we were able to identify and clone a funcE. coli putA gene.
2. Materials and methods
2.1. DNA preparation and subtraction Genomic DNA from both E. coli strain CSH4 and the RM2 were isolated and purified by CsCl gradient centrifugation [ 121, Genomic subtraction were carried out as previously reported ill]. An excess (50 pg) of sonicated, biotinylated, deletion mutant RM2 DNA was denatured in the presence of 0.5 pg Sau3ALdigested wild-type CSH4 DNA and allowed to reassociate for 48 h at 68” C. Since DNA molecules from CSH4 that corresponded to the deleted region in RM2 had no biotinylated complementary strands with which to hybridize, these fragments cannot form hybrids. In the next step, the biotinylated DNA and any DNA that had reassociated with it were removed from the sample by incubating with streptavidin (Vector labs). The unbound DNA, which was enriched for sequences that are missing in the deletion mutant, is collected and rehybridized with another batch of freshly biotinylated mutant DNA.
amplified in a thermal cycler (Perkin-Elmer/Cetus) using the shorter adaptor strand as primers. The parameters for each of the 35 cycles were as follows: 94” C, 30 s; 55” C, 1 min; 72” C, 3 min. PCR not only produced enough DNA molecules for further experiment, but also eliminated contaminations from subtracting reactions since only wild-type DNA fragment which possessed Sau3AI-specific ends was able to be amplified. 2.3. Southern hybridization The major PCR product was purified and denatured in the presence of one adaptor molecule and labelled with [32P]dCTP using the Klenow fragment of DNA polymerase I. The labelled probe was then used to hybridize genomic DNA isolated from either CSH4 or RM2, digested to completion with HindIII. The same probe was also used to screen a wild-type CSH4 genomic library constructed in ZAP11 (Zhu et al., unpublished results). Nylon membranes (Gene screen, NEN) were used for both blots. Single positive plaques were obtained after two rounds of hybridization. E. coli cell hosting recombinant plasmid was isolated by in vivo excision upon superinfection with a helper phage R408.
Lane
A
B
2.2. PCR amplification After three rounds of subtraction, the remaining DNA was amplified through PCR. A Sau3AI adaptor was prepared by annealing the synthetic oligonucleotides GACACTCTCGAGACATCACCGTCC and GATCGGACGGTGATGTCTCGAGAGTG. The latter oligonucleotides were phosphorylated at the 5’ end using T4 polynucleotide kinase (New England biolabs). An equimolar amount of the two strands was mixed and heated to 65” C for 5 min in a heating block and then allowed to cool to room temperature within the block. One end of the resulting adaptor is Suu3AI compatible with a 5’ phosphate while the other is not. This is to ensure that only one adaptor molecule can be ligated onto each end of a Suz&U fragment. DNA molecules capped with adaptors (l/10 of the ligation reaction) were
Fig. 1. Southern blot analysis of RM2 and CSH4 genomic DNA using 32P-Iabelled PCR product which was generated after three rounds of subtractions as a probe. Lane 1, 10 pg of HindIII-digested RM2 DNA, Lane 2, 10 pg of HindIII-digested CHS4 DNA. Arrows indicate standard migration of 7.5 kb (upper) and 2 kb (lower).
237
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
2.4. Assay of ProDH activity Competent RM2 cells were transformed with the plasmid containing a putative putA coding sequence and selected for their ability to grow on minimal medium which contained 50 pg/ml of ampicillin and 1 mM of IPTG and utilized proline as the sole sources of carbon and nitrogen. Proline dehydrogenase activities of the transformant as well as CSH4 and RM2 cells were then determined following previously described methods [l]. 2.5. DNA sequencing
and computer analysis
DNA sequencing was carried out using CsClpurified, double-stranded plasmid DNA by the dideoxynucleotide chain termination method using
Fig. 2. Nucleotide and predicted amino acid sequence EMBL data bank under accession number X78340.
Sequenase II (USB) following the manufacturer’s instructions. Deletion of the plasmid template was generated by a combination of Exonuclease III and Sl nuclease. DNA sequence was compiled and analysed using the GCG package [13].
3. Results and discussion 3.1. Cloning the E. coli putA gene using a probe that was generated from subtraction When genomic DNA isolated from CSH4 cells were subjected to three rounds of subtraction by large quantities of RM2 DNA molecules, the pattern of PCR products simplified down to only a few bands (data not shown). The main product (a 1.3-kb
of E. coli putA. The nucleotide
sequence
reported
here has been submitted
to the
238
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
Fig. 2 (continued).
band on agarose gel) was used to probe DNA isolated from the two different bacterial strains to make sure that such a generated probe had no representation in RM2 cells. As shown in Fig. 1, a 6.0-kb HindIII-digested DNA fragment from CSH4 hybridized to the probe specifically (lane B) while there is no signal from RM2 DNA (lane A). This was taken as a good indication that the probe was pure enough to be used for cloning of the putA gene via plaque hybridization against a CSH4 partial genomic library. Although genetic complementation in a bacterial cell is quite common, and the DNA sequences of putC as well as part of p&P and putA from E. coli K12 were known when we began our work [14-161, we failed to clone putA gene using this method. Several independent reasons may account for the
failure. First, the size of putA gene makes it a bad candidate to be cloned as a functional unit; second, the growth rate of RM2 cells on minimal medium is too low to permit reliable screen; third, for some reason, the rate of RM2 cell transformation was much lower comparing to those standard cell lines. In our view, the genomic subtraction technique is a good way to clone those genes that do not have a ready probe. It may not be as convenient as a screen library, but the probe we generated was pure enough to permit cloning of the puti gene. 3.2. Verification of the cloned sequence by enzyme activity assay To verify that some of the positive plaques may contain the entire put4 coding sequence, we trans-
M. Xia et al. /FEMS
Microbiology Letters 127 (1995) 235-242
formed competent RM2 cells with a plasmid excised from the vector and then selected for cells that grew on minimal medium. ProDH activities of the transformants as well as CSH4 and RM2 cells were
239
determined. While the ProDH activity of RM2 cells was indeed very low (too low to be detected by our assay), the transformed RM2 cells (ZX16) possessed higher activity than that of wild-type CSH4 cells
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
660 PVINPAEPKDIVGWGREATESEVEQALQNAVNQAPVWFATVLMEDQM . . . . .. .. . . . .. .. .. . .. . . .. . .. . .. . . .. . . .. .. . . . . . .. . . .. . . . . . . . . . . . .. . .-TEEVICHVE.GDKAD.DK.VRQ.FQIGSP.RTMDAS..GCL.NKL.D...RDR ET....NG-NVLATVGA.GRED.DR.VKS.QQGQKI.ASMTAM..SR..R..VDILRERN ..S. .LDNSTLAEIACASA.-Q....VAS.RETFAS.KE..VS...RVML.YQA.LKEHH
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
720 QQLIGLLVREAGKTFSNAIAEVREAVDFLHYYAGQVRDDFDNETHRPLGPWCISPWNFP .. . .. . . . . . .. . . .. . . .. . . ... . . .. . . .. . . . .. . . . . .. . .. . .. . . . . .. . . . . VL.ATMESMN...I.TH.LLDTEVSIKA.K.F..KIHGQ.TYTRRE.I.VCGQ.I...G. DE.AK.ETLDT..AY.ESTVDIVTGA.V.E....PL.ET.VYTRRE.,.V.AG.GA..Y. DE.AKIVSS.L... .ED.KGD.WRGIEWEHLM.ETVENVARNIDQ...VC.G.T.F...
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
780 LAIFTGQIAAALAAGNSVLAKPAEQTSLIAAQGIAILLEARTAGRRATVAGTGRNRRRPA . .. .. . . .. . .. . . .. .. . . .. . . ... .. .. . . .. . . . . .GVPPGWQLLPGRGETVGAQ .IL.IWK.G...SC..T.IV...... P.T.LYMASLIK..GPP.WNV.P.Y.STAGAAI IQ.ALWKS.P..... .AMIF..S.V.P.T.LKLAE.YS..LPD.VFNVLP.V.AETGQYL AM.PLWMFPL.I.C..AFIL..S..VP.TSVRLAELF . ..GAPKGVLQ.VHG.KEQVDQL
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
840 YRRCACTRVMFTGSTEVATLLQRNIATRLDAQGRPIPLIAETGGMNRMIVDSSALTEQW LTAD.RV..................................................... SSHMD1DK.S.. . . ...GK.IKEAAG---KSNLKRVT.--.L..KSPC..FAD.DLDSA. TEHPGIAK.S...GVASG---KKVM.NSAASSLKEVTM--.L..KSPL...DAD.DLAAD LKHPQVKA.S.V..VA.GQYVYHTGT----.HNKRVQSF.--.AK.H.VIMPD.DKA..I
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
900 VDVLASAFDSAGQRCSALRVLCLQDDIAEHTL~LRRPHGGVSVGESGRLTTDIGPVIDS . . .. .. ..RQRRTTLFR.... . .. . .. . .. . . .. ..GAMAECRM.NP............. EFAHQGV.FHQ..I.V.ASR.FVEES.YDE---FV.. ---S.ERAKKYV.GNPLDSG.SQ IAMM.NF.-.S..V.NGT..FVPAKCK.AFEQ.I.A.VE-RIRA.DVFDPQ.NF..LVSF SNLVGASVGA.....M.IS.AV.VGAAR.W-IPEI.DALAK.RP.PWDDSGASY....NP
ECPUTA STP'JTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
960 EAKANIERHIQTMRAKGRPVSQ~RENSDDAQEWQTGTFVMPTLIELENFAELEKEVFGP . . . . . .. .. . . .. . . . .. ..F....................................... GPQIDK.Q.AKILDLIESGKKEG.KLECGGGR-.NK.F..Q..VFNVTDEMRIA..I... PHRD.VL.Y.AKGKEE.AR.LCGG--DVLKGDGFDN.AW.A..VFTSDDMTIVRE.I... Q...R...L.GQGVEE.AQLLLDG.GV--KVEGYPD.NW.G...FARPDM.IYRE.....
ECPUTA STPUTA DHAC-RAT DHAB-ECOLI MMSA-PSEAE
1020 VLHWRYNRNQLAELIEQINASGYGVTLGVHTRIDE . . . . . .. . . . . ..DV..........L.......... .QQIMKFKS--ID.V.KRA.NTP..LAA..F.KL.R .MSILT.ESED--.V.RRA.DTD..LAA.IV.ADLN ..CLA--EVDS.EQA.RL..E.P..NGTSIF.SSGA
Fig. 3. Comparison of deduced E. coli putA protein amino acid sequences 660-1060 with various regions of the Salmonella typhimurium putA, rat cytosolic aldehyde dehydrogenase, E. coli betaine aldehyde dehydrogenase and Pseudomonas aeruginosa methylmalonate-semialdehyde dehydrogenase. Multiple alignment was done with the GCG programs. The sources of the sequences were taken from refs. [5,17-191.
240 Table 1 Determination
M. Xia et al. /FEMS Microbiology Letters 127 (1995) 235-242
of the enzyme activity of the cloned puti
Strains
Activity of ProDH (nmol min -1 (mg protein)- ’ 1
CSH4 RM2 ZX16 a
12.6 f 1.0
a ZX16 is the RM2 cell line that contains putative puti gene.
gene
the plasmid
of a
(Table 1). This encouraged us to think that we actually cloned sequences which encode for a functional ProDH.
cofactor binding and may be needed for dehydrogenase activity 131.It shares a 84.9% identity with the Salmonella typhimurium putA protein although it is 111 amino acids longer at the carboxyl terminal end. The significance of these 333 extra nucleotides present at the 3’ end of the E. coli put.4 gene is not known. Since E. coli ProDH is such a large protein with multiple functions and binds a number of substrates, such as proline, P5C, DNA, FAD and NAD, the relationships between its structure and function could be complicated. Further analysis on functional domains are warranted for a better understanding of proline utilization.
3.3. Sequence analysis of the E. coli put4 gene
Acknowledgements
The complete nucleotide (nt) sequences of our E. coli putA gene was shown together with the deduced amino acid sequence in Fig. 2. The largest ORF extends from the ATG start codon, which is preceded by a potential RBS at nt 295 to the stop codon at nt 4234, which is immediately followed by multiple stop codons in all reading frames. The major ORF codes for a polypeptide of 1313 amino acid with a calculated M, of 143 808. These values are in accordance with biochemical data obtained earlier [2]. The E. coli putA gene is rich in G and C (about 60%). Codon usage analysis revealed a high frequency (65%) of codon ending with G or C which correlates well with the reported frequency of E. coli codon usage. Examination of the deduced E. coli proline dehydrogenase amino acid sequence shows that some part of it is homologous to various previously characterized dehydrogenases including that of bacteria, fungi and eukaryotic organisms. Computer comparisons of deduced amino acid sequence of the E. coli put4 with EMBL database revealed sequence similarities to a number of proteins. The N-terminus of the E. coii putA protein has 100% identity to deduced amino acid sequence of the S. typhimurium put4 protein [5], indicating that this sequence may have an important function. The region from amino acid 660-1060 has also some similarities (25-44%) to a number of NAD-dependent dehydrogenases from a wide variety of organisms (Fig. 3). Previous work suggests that this domain may be important for
We thank Miss Xiaowen Liang for her assistance in database search and computer sequence analysis. We are grateful to Miss Qun Jing and Xiaohua Li for typing the manuscript. This work was supported by a grant from the national high-technology ‘863’ program of China to Y.-X.Z.
References [l] Dendinger, S. and Brill, W.J. (1970) Regulation of proline degradation in Salmonella typhimurium. J. Bacterial. 130, 144-1.52. [2] Scarpulla, R.C. and Soffer, R.L. (1978) Membrane-bound proline dehydrogenase from Escherichia coli. J. Biol. Chem. 17, 5997-6001. [3] Brown, E.D. and Wood, J.R. (1993) Conformational change and membrane association of the putA protein are coincident with reduction of its FAD cofactor by proline. J. Biol. Chem. 268,8972-8979. [4] Wood, J.M. (1987) Membrane association of proline dehydrogenase in Escherichia coli is redox dependent. Proc. Natl. Acad. Sci. USA 84, 373-377. [5] Allen, S.W., Senti-Willis, A. and Maloy, S.R. (1993) DNA sequence of the putA gene from Salmonella typhimurium: a functional membrane-associated dehydrogenase that binds DNA. Nucleic Acids Res. 21, 1676. [6] De Spicer, P.O. and Maloy, S. (1993) puti protein, a membrane-associated flavin dehydrogenase, acts as a redoxdependent transcriptional regulator. Proc. Natl. Acad. Sci. USA 90, 4295-4298. [7] Maloy, S. and Stewart, V. (1993) Autogenous regulation of gene expression. J. Bacterial. 175, 307-316. [8] Hayward, D.C., Delaney, S.J., Campbell, H.D., Ghysen, A., Benzer, S., Kasprzak, A.B., Cotsell, J.N., Young. LG. and
M. Xia et al. /FEMS
[9]
[lo]
[ll]
[12]
[13]
Microbiology
Miklos, G.L. (1993) The sluggish-A gene of Drosophila melanogaster is expressed in the nerve system and encodes proline oxidase, a mitochondrial enzyme involved in glutamate biosynthesis. Proc. Natl. Acad. Sci. USA 90, 29792983. Brown, E.D. and Wood, J.R. (1992) Redesigned purification yields a fully functional puti protein dimer from Escherichia coli. J. Biol. Chem. 267, 13086-13092. Zhu, Y.-X., Shearer, G. and Kohl, D.H. (1992) Proline fed to intact soybean plants influences acetylene reducing activity, and proline content and metabolism in bacteroids. Plant Physiol. 98, 1020-1028. Li, W.-B, Gruber, C.E., Lin, J.J., Lim, R., D’AJdessio, J.M. and Jessee, J.A. (1994) The isolation of differentially expressed genes in fibroblast growth factor stimulated BC3Hl cells by subtractive hybridization. BioTechniques 16, 722729. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Springs Harbor Press, Cold Springs Harbor, NY. Devereaux, J., Haeberli, P. and Smithies, 0. (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387-395.
Letters 127 (1995) 235-242
241
[14] Mogi, T., Yamamoto, H., Nakao, T., Yamato, I. and Anraku, Y. (1986) Genetic and physical characterization of putP, the proline carrier gene of Escherichia coli K12. Mol. Gen. Genet. 202, 35-41. [15] Nakao, T., Yamato, I. and Anraku, Y. (1988) Mapping of the multiple regulatory sites for putP and putA expression in the pufC region of Escherichia coli. Mol. Gen. Genet. 214, 379-388. [16] Chen, L.-M. and Maloy, S. (1991) Regulation of proline utilization in enteric bacteria: cloning and characterization of the Klebsiella put control region. J. Bacterial. 173, 783-790. [17] Hsu, L.C., Chang, W.C. and Yoshida, A. (1989) Genomic structure of the human cytosolic aldehyde dehydrogenase gene. Genomics 5, 857-865. [18] Dunn, T.J., Koleske, A.J., Lindahl, R. and Pitot, H.C. (1989) Phenobarbital-inducible aldehyde dehydrogenase in the rat: cDNA sequence and regulation of the mRNA by pentobarbital in responsive rats. J. Biol. Chem. 264, 13057-13065. [19] Hirji, R. and Selvaraj, G. (1991) Characterization of an E. coli gene encoding betaine aldehyde dehydrogenase (BADH): structural similarity to mammalian ALDHs and plant BADH. Gene 103, 45-52.