Ž .Biochimica et Biophysica Acta 1351 1997 261–266

Short sequence-paper

Isolation and nucleotide sequence of the gene encoding the XamI DNAmethyltransferase of Xanthomonas campestris pv. amaranthicola

Pilar Gomez, Rosa M. Ribas-Aparicio 1, Ana I. Pelaez, Ana Gomez, M. Rosario Rodicio )´ ´( ) ( )Departamento de Biologıa Funcional Area de Microbiologıa and Instituto UniÕersitario de Biotecnologıa IUBA , UniÕersidad de´ ´ ´

OÕiedo, 33006 OÕiedo, Spain

Received 29 November 1996; revised 4 February 1997; accepted 7 February 1997

Abstract

Ž .The gene xamIM encoding the DNA methyltransferase of the XamI restriction-modification system from Xan-Ž .thomonas campestris pv. amaranthicola M. XamI has been cloned in Escherichia coli and its nucleotide sequence

determined. The sequence predicts a protein of 527 amino acids that contains nine conserved motifs characteristic of DNAamino methyltransferases. In fact, M. XamI shows significant similarity with N6-adenine methyltransferases of the g group

Ž .of amino methyltransferases, including M.Sal I from the isoschizomeric Sal I restriction-modification system and M.TaqIŽ .the only N6-adenine methyltransferase for which a three-dimensional structure is available . M. XamI and M.SalI sharetwo highly conserved regions within the C-terminal domain, one of which aligns with one of the DNA recognition loopsproposed for M.TaqI. Analysis of the chromosomal DNA adjacent to xamIM led to the identification of an additional ORFŽ .275 codons , downstream, in the same transcriptional orientation. Although some limited similarities between the Sal Irestriction enzyme and the product deduced from this ORF were found, the clone carrying xamIM did not express theexpected endonuclease function. q 1997 Elsevier Science B.V. All rights reserved.

Keywords: DNA methyltransferase; Modification; Sal I; XamI

Type II restriction endonucleases are essential toolsin molecular biology. In nature, they are part ofrestriction-modification systems whose main functionis thought to be protection of bacterial cells againstinfection by foreign DNA. To date, more than 2650type II restriction endonucleases have been identified,

w xonly representing 207 specificities 1 . This implies

) Corresponding author. Fax: q34 8 5103148; E-mail: rrodi-cio@dwarf1.quimica.uniovi.es

1 Present address: Departamento de Microbiologıa, Escuela´Nacional de Ciencias Biologicas del Instituto Politecnico Na-´ ´cional, Mexico.´

that enzymes with the same specificity, i.e.,isoschizomers, are present in different bacteria.

Streptomyces albus G produces the restriction en-zyme Sal I that recognizes double-stranded DNA at

X X w xthe palindromic sequence 5 -GTCGAC-3 2 . Theendonuclease is part of a restriction-modification sys-

w xtem that has been characterized in detail 3–12 .Isoschizomers of Sal I have been detected in manydifferent bacteria, including strains of the genera

Ž w x.Xanthomonas R. Xam I; 2 , HerpetosiphonŽ w x. Ž w x.R. HgiDII; 13 , Rhodococcus R. Rrh4273I; 14

Ž w x.and Ruminococcus R. RflFI; 15 . None of the genesfor these endonucleases have been isolated. However,the sequence of the gene encoding the HgiDII C5-cy-

0167-4781r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0167-4781 97 00030-4

( )P. Gomez et al.rBiochimica et Biophysica Acta 1351 1997 261–266´262

tosine methyltransferase from Herpetosiphon gigan-w xteus Hpa2 has been reported 16 . In addition the

genes of the HincII and AccI systems, both recog-nizing degenerate sequences that include the Sal Itarget sequence as a subset, have been cloned and

Žw xsequenced 17–19 ; G. Wilson, personal communica-.tion .

In order to gain some insight into the evolutionaryrelationship among enzymes interacting with 5X-GTCGAC-3X, and as a step to establish the mecha-nism by which the target sequence is recognized, weextended these studies to the analysis of the geneŽ .xamIM encoding the XamI methyltransferase fromXanthomonas campestris pÕ. amaranthicola.

xamIM was isolated from a library of X.campestris pv. amaranthicola ATCC 11645 DNA,constructed in the Escherichia coli vector pBR328w x20 . Screening was based on expression of the genewhich renders the recombinant plasmid resistant to invitro cleavage by the isoschizomeric Sal I endonucle-

w xase 21,22 . To generate the library, X. campestrispv. amaranthicola DNA was digested by EcoRI andligated to pBR328, previously linearized with thesame enzyme. After transformation of E. coli RR1w x23 with the ligation mixture, plasmid DNA wasisolated from the pooled transformants. Then, 1 mg

Ž .was digested with SalI 5 units, 2 h, 378C , andre-transformed into E. coli. The enrichment for clonesresistant to in vitro cleavage by Sal I led to theisolation of a recombinant plasmid, pUO1050, carry-

Ž .ing an insert of 10 kb Fig. 1 . pUO1050, as well aschromosomal DNA of the E. coli strain containingthe plasmid were fully protected against SalI cleav-age. The approximate location of the modificationgene within the cloned DNA was established by

Ž .subcloning experiments Fig. 1 . These demonstratedthat xamIM is located within the 3 kb ClaI-EcoRVfragment present in pUO1055. Again, this plasmid aswell as the chromosomal DNA of the strain carryingit were resistant to in vitro cleavage with Sal I. The

Fig. 1. Restriction map of the cloned DNA and location ofxamIM demonstrated by subcloning experiments. Black arrowsabove the map show the position and orientation of xamIM andORF2. Protection against Sal I cleavage indicates an Mq pheno-type.

5X-GTCGAC-3X site present in pUO1055 was alsoprotected against digestion by HincII and AccI.

The nucleotide sequence of the ClaI-EcoRV seg-ment containing the functional xamIM gene was

Ž .determined Fig. 2 . Examination of the sequence inall six reading frames for potential initiation andtermination codons revealed the presence of two ORFsŽ .ORF1 and ORF2 of significant length, on the sameDNA strand.

The first possible initiation codon for ORF1 is anATG at nt 422-424, which is preceded by a potential

Ž . Ž X X.Shine and Dalgarno SD sequence 5 -AGAAAG-3 ,5 nt upstream. Since ORF1 extends to a TGA codonat positions 2003-2005, it would encode a protein of527 amino acids, with an estimated M of 57535.r

The GqC content of ORF1, 61%, is close to thelower limit reported for the genome of X. campestrisŽ w x.63.5–69.2%; 24 . Upstream from ORF1, sequencessimilar to y35 and y10 elements characteristic ofmajor eubacterial promoters were present.

Ž .Fig. 2. Nucleotide sequence of the xamIM region. Translation of ORF1 xamIM and ORF2 is presented in single-letter code. Potentialy10 and y35 promoter elements and SD sequences are shown in bold. The left inverted repeat of an insertion sequence detected in the

Ž .genome of X. campestris pv. amaranthicola, downstream from ORF2 P. Gomez, unpublished results is shown in bold and underlined.w xBoth strands were sequenced by the chain-termination method 32 , using a combination of sub-cloning, generation of unidirectional

w xdeletions by treatment with ExoIII 33 and the aid of synthetic primers. The GenBank accession number of the sequence is U77781.

( )P. Gomez et al.rBiochimica et Biophysica Acta 1351 1997 261–266´ 263

( )P. Gomez et al.rBiochimica et Biophysica Acta 1351 1997 261–266´264

( )P. Gomez et al.rBiochimica et Biophysica Acta 1351 1997 261–266´ 265

Ž .Fig. 4. Similarity between the product P of ORF2 and the Sal IŽ . w x Ž .restriction enzyme R.Sal I as detected by BLAST 34 . q

denote functionally similar amino acids.

ORF1 was assigned to xamIM, since the deducedproduct shares considerable similarity with aminomethyltransferases and contains the nine amino acid

Ž .conserved motifs motifs I to VIII and X characteris-tic of the catalytic domain of this type of enzymew x25 . The highest scores were obtained with M.Sal I,M. Acc I, M. HincII, M. PaeR71, M. Eco57I,

Ž .M.TthHB8I and M.TaqI Fig. 3 , all belonging to thew xg group of amino methyltransferases 25 . This group

is composed of N6-adenine methyltransferases inwhich the order of the motifs is similar to that seen in

w xC5-cytosine methyltransferases 26 , differing only inposition of motif X. Assignment of M. XamI to the g

group strongly suggests that it would methylate thesingle adenine present within the 5X-GTCGAC-3X tar-get sequence. In addition, the enzyme may have atertiary structure similar to that of M.TaqI, for whicha crystal structure complexed with S-adenosyl-L-

w xmethionine has been provided 27 : the N-terminalcatalytic domain, that contains motifs I to VIII and XŽ .see above , and a C-terminal DNA binding domain.By modelling a complex between M.TaqI and DNAcontaining the 5X-TCGA-3X recognition sequence, twoloops were identified within the C-terminal domain,that can make base specific contacts with the DNAw x28 . Comparison of the C-terminal domains of theXamI and Sal I methyltransferases revealed the pres-

Ž .ence of two highly conserved regions see Fig. 3 .Currently, we are investigating a possible involve-

ment of these regions, one of which aligns with oneof the recognition loops proposed for M.TaqI, in theinteraction of M. XamI and M.Sal I with the 5X-GTCGAC-3X recognition sequence.

Since modification and restriction genes are usu-ally linked, the DNA carrying xamIM was searchedfor the cognate endonuclease function. For this, ex-tracts were prepared from E. coli strains carrying

ŽpUO1050 and each of the different subclones see.Fig. 1 . None of them contained detectable endonu-

clease activity, as shown by restriction assays per-w xformed according to 29 , using Pst I linearized

w xpBR322 30 as substrate. In agreement with this, inŽvivo restriction tests using phage l EMBL4 which

w x.has 4 SalI sites in its DNA; 31 failed to reveal theexpected restriction phenotype. Therefore, although

ŽORF2 that follows xamIM in the same transcrip-.tional orientation, Fig. 2 showed some limited simi-

larities with the endonuclease of the Sal I systemŽ .Fig. 4 , conclusive identification of the restrictiongene of the XamI system awaits further research.

P.G and A.G were recipients of grants from FYCIT,Asturias, Spain. R.M.R.-A was the recipient of afellowship from COFAA-IPN and Banco de Mexico,´Mexico. The work was supported by grant BIO91-´0563, CICYT, Spain.

References

w x Ž .1 R.J. Roberts, D. Macelis, Nucleic Acids Res. 24 1996223–235.

w x2 J.R. Arrand, P.A. Myers, R.J. Roberts, J. Mol. Biol. 118Ž .1978 127–135.

w x Ž .3 K.F. Chater, L.C. Wilde, J. Bacteriol. 128 1976 644–650.w x Ž .4 K.F. Chater, L.C. Wilde, J. Gen. Microbiol. 116 1980

323–334.w x Ž .5 A. Maxwell, S.E. Halford, Biochem. J. 203 1982 77–84.w x Ž .6 A. Maxwell, S.E. Halford, Biochem. J. 203 1982 85–92.w x Ž .7 A. Maxwell, S.E. Halford, Biochem. J. 203 1982 93–98.w x Ž .8 M.R. Rodicio, K.F. Chater, Mol. Gen. Genet. 213 1988

346–353.

Fig. 3. Comparison of M. XamI with other DNA methyltransferases. Conserved motifs present within the N-terminal putative catalyticŽ . w xdomain of N6-adenine methyltransferases I-VIII and X are indicated. The DNA recognition loops proposed for M.TaqI 28 are

underlined on the C-terminal domain of the enzyme. Highly conserved regions on the C-terminal domains of M. XamI and M.Sal I areŽ .overlined. Accession numbers GenBank or Swissprot are as follows: M.Sal I, UO1232; M. AccI, P25201; M. HincII, P17744;

M. PaeR7I, P05103; M. Eco57I, P25240; M.TthHB8I, P29749; M.TaqI; P14385.

( )P. Gomez et al.rBiochimica et Biophysica Acta 1351 1997 261–266´266

w x Ž .9 M.R. Rodicio, K.F. Chater, Gene 74 1988 39–42.w x10 M.A. Alvarez, K.F. Chater, M.R. Rodicio, Mol. Microbiol.

Ž .8 1993 243–252.w x11 M.R. Rodicio, T. Quinton-Jager, L.S. Moran, B.E. Slatko,

Ž .G.G. Wilson, Gene 151 1994 167–172.w x12 M.A. Alvarez, A. Gomez, P. Gomez, J.E. Brooks, M.R.

Ž .Rodicio, Mol. Gen. Genet. 253 1996 74–80.w x13 M. Kroger, G. Hobom, H. Schutte, H. Mayer, Nucleic Acids¨ ¨

Ž .Res. 12 1984 3127–3141.w x14 M.J. Yebra, I.S. Novella, C. Barbes, J.F. Aparicio, C.G.´

Martin, C. Hardisson, J. Sanchez, J. Gen. Microbiol. 132Ž .1991 1279–1284.

w x15 M. Morrison, R.I. Mackie, B.A. White, Appl. Env. Micro-Ž .biol. 58 1992 66–69.

w x Ž .16 A. Dusterhoft, M. Kroger, Gene 106 1991 87–92.¨ ¨ ¨w x17 H. Ito, A. Sadaoka, H. Kotani, N. Hiraoka, T. Nakamura,

Ž .Nucleic Acids Res. 18 1990 3903–3911.w x18 B. Kawakami, C. Hilzheber, M. Nagatomo, M. Oka, Agric.

Ž .Biol. Chem. 55 1991 1553–1559.w x Ž .19 S. Brassard, H. Paquet, P.H. Roy, Gene 157 1995 69–72.w x Ž .20 X. Soberon, L. Covarrubias, F. Bolivar, Gene 9 1980

287–305.w x Ž .21 E. Szomolanyi, A. Kiss, P. Venetianer, Gene 10 1980

219–225.w x22 K.D. Lunnen, J.M. Barsomian, R.R. Camp, C.O. Card, S.-Z.

Chen, R. Croft, M.C. Looney, M.M. Meda, L.S. Moran,D.O. Nwankwo, B.E. Slatko, E.M. Van Cott, G.G. Wilson,

Ž .Gene 74 1988 25–32.

w x23 E.A. Raleigh, N.E. Murray, H. Revel, R.M. Blumenthal, D.Westaway, A.D. Reith, P.W.J. Rigby, J. Elhai, D. Hanahan,

Ž .Nucleic Acids Res. 16 1988 1563–1575.w x Ž .24 J.F. Bradbury, in: N.R. Krieg and J.G. Holt Eds. , Bergey’s

Manual of Systematic Bacteriology, vol. 1, Williams andWilkins, Baltimore, MD, 1984, pp. 199-210.

w x25 T. Malone, R.M. Blumenthal, X. Cheng, J. Mol. Biol. 253Ž .1995 618–632.

w x26 S. Kumar, X. Cheng, S. Klimasauskas, S. Mi, J. Posfai, R.J.Ž .Roberts, G.G. Wilson, Nucleic Acids Res. 22 1994 1–10.

w x27 J. Labahn, J. Granzin, G. Schluckebier, D.P. Robinson,W.E. Jack, I. Schildkraut, W. Saenger, Proc. Natl. Acad.

Ž .Sci. USA 91 1994 10957–10961.w x28 G. Schluckebier, J. Labahn, J. Granzin, I. Schildkraut, W.

Ž .Saenger, Gene 157 1995 131–134.w x29 G. Theriault, P.H. Roy, K.A. Howard, J.S. Benner, J.E.

Brooks, A.F. Waters, T.R. Gingeras, Nucleic Acids Res. 13Ž .1985 8441–8461.

w x30 R. Bolivar, R.L. Rodriguez, P.J. Greene, M.C. Betlach, H.L.Ž .Heynecker, H.W. Boyer, Gene 2 1977 95–113.

w x31 A.M. Frischauf, H. Lehrach, A. Poustka, Y.N. Murray, J.Ž .Mol. Biol. 170 1983 827–842.

w x32 F. Sanger, S. Nicklen, A.R. Coulson, Proc. Natl. Acad. Sci.Ž .USA 74 1977 5463–5467.

w x Ž .33 S. Henikoff, Gene 28 1984 351–359.w x34 S.F. Altschul, W. Gish, W. Miller, E.W. Myers, D.J. Lip-

Ž .man, J. Mol. Biol. 215 1990 403–410.