Oct 1, 1990 - recM mutant strains produces a wide variety of effects: i) markedly reduced cell .... killing action of 1 mM 4-nitroquinoline-l-oxide (4NQO). The ...
k.) 1990 Oxford University Press
Nucleic Acids Research, Vol. 18, No. 23 6771
Molecular cloning, genetic characterization and DNA sequence analysis of the recM region of Bacillus subtilis Juan C.Alonso*, Katsuhiko Shirahige1 and Naotake Ogasawara1 Max-Planck-Institut fOr molekulare Genetik, Ihnestrasse 73, D-1000 Berlin 33, FRG and 1Department of Genetics, Osaka University, Medical School, 3-57, Nakanoshima 4 Chome, Kita-Ku, Japan Received October 1, 1990; Revised and Accepted November 6, 1990
ABSTRACT In Bacillus subtilis the recM gene, whose product is associated with DNA repair and recombination, has been located between the dnaX and rrnA genes. The recM gene has been cloned and analyzed. Analysis of the nucleotide sequence (3.741-kilobase) around recM revealed five open reading frames (orl). We have assigned recM and dnaX to two of this orf, given the gene order dnaX-orfl 07-recM-orf74-orf87. The organization of genes of the dnaX-orf107-recM region resembles the organization of genes in the dnaXorf1 2-recR region of the Escherichia coil chromosome. Proteins of 24.2 and 17.0 kDa would result from translation of the wild type and in vitro truncated recM genes, and radioactive bands of proteins of molecular weights of 24.5 and 17.0 kDa were detected by the use of the T7promoter-expression system. The RecM protein contains a potential zinc finger domain for nucleic acid binding and a putative nucleotide binding sequence that is present in many proteins that bind and hydrolyze ATP. Strains, in which the recM gene has been insertionally inactivated, were generated and show a phenotype essentially the same as previously described recM mutants.
INTRODUCTION The bacterial replication origin and two clusters of genes involved in DNA metabolism followed by a rrn gene have been genetically identified within a 35 Kb segment of the Bacillus subtilis chromosome. The gene order in the first cluster is dnaA dnaN recF gyrB gyrA rrnO, with dnaA-dnaN and recF-gyrB as parts of two independent transcriptional units (1). The dnaA gene encodes the replisome assembly protein for initiation of chromosomal replication at oriC (2, 3). The dnaN gene encodes the ,3 subunit of DNA polymerase HI holoenzyme (4, 2), whereas recF codes for a product involved in DNA repair and recombination (5). Both gyrB and gyrA have been shown to code for the different subunits of DNA gyrase (6, 1). Except for gyrA, the gene order resembles the organization of the analogous genes in the Escherichia coli chromosome (1). In this bacterium, however, these genes seem to be both coordinately and separately regulated (7, 8, 9). *
To whom correspondence should be addressed
EMBL accession no. X17014
Within the second cluster, the gene order is dna-8132 dnaH recM recD rrnA (see 10). It has been postulated that the dnaH gene codes for a DNA polymerase IH auxiliary protein (11). The recM and recD gene products are associated with the AddABCindependent mechanism of DNA repair (5). DNA damage in recM mutant strains produces a wide variety of effects: i) markedly reduced cell viability, ii) reduced and delayed induction of the SOS response, and iii) cells impaired in genetic exchange in certain mutant backgrounds (5). Thus, the phenotype of recM mutants indicates that the recM gene is also involved in DNA metabolism. In this paper we report on the molecular analysis of the recM gene and its relationship to the flanking loci.
MATERIALS AND METHODS Bacterial strains and plasmids B. subtilis strains used were YB886 (trpC2 metB5 amyE sigB xin-1 attSPO) and its isogenic derivatives BG127 (recM13), BG103 (recM27), BG121 (recD41), BG201 [dnaH51(Ts)], BG8132 [dna-8132(Ts)] (5, 11), and BG128 (recMl::cat) (this report). On the basis of map position, phenotype and DNA complementation analysis the dna-8132 and dnaH51 mutations were renamed to dnaX8132 and dnaX51 (12). For the construction of the B. subtilis recM null mutant strain, we have inserted into an unique NruI site the blunt ended ClaIHpaII DNA fragment containing the chloramphenicol acetyltransferase (cat) gene of plasmid pC194 (see Fig. IC) in the two possible orientations. Since in B. subtilis transformation of competent cell with linear chromosomal DNA leads to marker replacement we used this technique to generate a recM null allele. To transfer the recM: : cat insertion allele (termed recM 1) to the B. subtilis chromosome, while avoiding duplication of the recombinant gene, we linearized the plasmids at one of their ends homologous to the recM region in the chromosome and used this DNA to transform rec+ cells (YB886) selecting for chloramphenicol resistance (CmR). All CmR transformants obtained with linearized plasmids were unable to form colonies on plates with 100 jg/ml of methyl methanesulfonate (MMS). This suggested that such a phenotype is independent of the direction of the transcription of the cat gene. The genetic configuration of the newly constructed strains was verified as follows: i) the insertional inactivation of the recM
6772 Nucleic Acids Research, Vol. 18, No. 23 gene was fully suppressed by a plasmid-borne recM+ allele (pBT52); ii) rec+ recipients transformed with DNA from BG128 and selected for the CmR phenotype became very sensitive to DNA damaging agents indicating a tight linkage between CmR and recMl; iii) the double crossover integration was checked by Southern blot hybridization. E. coli strains used were BL21DE3 hsdS20 gal [X-T7RNApol] (13) and JM103 [A(lac pro) thi strA supE endA sbcB hsdR F'traD36 proAB lacIq AM15 (14)]. The B. subtilis plasmids used were pUBIIO, its low copy number derivative pUBIlOcopI (15) and pC194 (16). The E. coli plasmids used were pUC18 (17), pLysE (18) and pT712 (from BRL,USA). The cloning of EcoRI chromosomal fragments from the region of the B. subtilis replication origin has been described previously (19, 20). A plasmid-borne 20 kb region of B. subtilis chromosomal DNA (p516-20) between the rmA-abrB genes (not shown in Fig. IA) was kindly provided to us by M. Marahiel (21). pNO2003 and pBR328-E21 (19) were used as the source of the 2.2 kb (E14) and 1.5 kb (E21) EcoRI DNA fragments (see Fig. lA). For the construction of plasmids pBT56, the 1.55 kb ClaI-PvuLl DNA fragment (coordinates 1767-3317) containing the recM region was cloned into the AccI-EcoRlcleaved plasmid pT712. For the truncation of the recM gene an in vitro deletion of 188 bp (NruI-NcoI) within the recM open reading frame (o4) was generated, and the plasmid termed
pBT57. Media and transformation B. subtilis cells were grown on TY (22), and competent cells were prepared as described by Rottliinder and Trautner (23). E. coli cells were grown in either NB or M9 and transformed according to Maniatis et al. (14). Transformants were selected on agar medium containing phleomycin at 100 ng/ml, neomycin or chloramphenicol at 5 ,ug/ml or ampicillin at 50 ,tg/ml. DNA manpulations and nucleotide sequence determination Plasmids DNA was prepared on preparative and analytic scales as previously described (24, 25). Restriction endonucleases and DNA modifying enzymes were purchased from Boehringer Mannheim or Tanaka Shozo Corp. The enzymes were used as specified by their suppliers. The DNA templates for sequencing were produced by cloning in both orientations the 1.79 kb EcoRl-Pvul DNA fragment from plasmid pNO2003 into SmaI-cleaved pUC18, followed by the generation of nonspecific deletions with exonuclease Im (26). The DNA fragments containing the junction between the EcoRI DNA fragments E21 and E14 (see Fig. 1A) were amplified from the B. subtilis chromosome by the PCR method using two synthetic primers, one corresponding to the E21 sequence and the other to E14. The fragments were then cloned into pUC18 and sequenced. The DNA sequence was determined by the dideoxynucleotide chain-termination method of Sanger et al. (27). Both strands of the DNA fragments were sequenced and the adjoining points were confirmed by reading overlapping clones. [a32P]dATP (3000Ci/mmol) were obtained from Amersham Corp. Modified T7 DNA polymerase (version 2.0) and all deoxyand dideoxynucleotides were purchased as a Sequenase kit from United-States Biochemical Corp. Expression of plasmid-encoded proteins The host-vector cloning system BL21DE3[pT712] of Studier and Moffatt (13) was used to express the cloned genes, and their
protocol followed. Since a small residual synthesis of the T7 RNA polymerase made the plasmids highly unstable, a titration of the T7 RNA polymerase was performed with plasmid pLsyE as described by Moffatt and Studier (18). Protein labeling and SDS-polyacrylamide gel electrophoresis were essentially as published (28). [14C]labeled amino acids were obtained from Amersham Corp. The fluorography reagent (EN3HANCE) was from DuPont-New England Nuclear.
Computer analysis Amino acid sequences, RNA secondary structures and transcription terminators were predicted by using the computer software package of the University of Wisconsin Genetic Computer Group in a VAX computer (29, 30).
RESULTS Cloning and genetic characterization of the recM region In order to clone and identify the recM gene of Bacillus subtilis a genomic library from the B. subtilis origin region was used (see Fig. 1A). Upon chromosomal transformation of MMS sensitive (MMSS) recM13 and recM27 strains with the purified B. subtilis chromosomal EcoRI DNA fragments shown in Figure 1A, only the 2.2 kb EcoRI fragment 14 (E14) and the 1.79 kb EcoRI-PvuH fragment conferred methyl methanesulfanate resistance (MMSR). Furthermore, the EcoRI-PvuII DNA fragment also complemented the inability of the dnaH(Ts) strain to form colonies at 46°C, but failed to complement the recD41 and dna-8132(Ts) defects (see Fig. iB). To ascertain the limits of the recM gene, in vitro deletions of the 1.79 EcoRI-Pvull chromosomal DNA insert were undertaken. The deleted DNA fragments were then used to transform the recM13, recM27 and the dnaH51(Ts) competent cells, with selection for MMSR in the case of the recM strains and the ability to form colonies at 46°C in the case of the daH(Ts) strain. Chromosomal transformation with purified PstI-PvuII (coordinates 2089 to 3317) or Hindll-XhoII (coodinates 2233 to 3167) DNA fragments conferred MMSR (300 gg/ml) to the
MMSS recM13 and recM27 recipient strains (both are sensitive to 40 yg/ml). The HindI-NruI (coordinates 2233 to 2737) and
NruI-EcoRI (coordinates 2737 to 3741) DNA fragments complemented only the recM27 and recM13 deficiencies, respectively. Transformation with the EcoRI-HindlI (1522 to 2233) or XhoH-EcoRI (3167 to 3741) DNA fragments failed to confer MMSR (see Fig. iB). The former, however complemented the ability to form colonies at 46°C to the dnaH(Ts) strain. To define which of the two putative orfs present in the HindUXholI DNA fragment (see below) actually codes for the recM activity, the cat gene was cloned (in both orientations) into the unique NruI site (within the putative recM gene, see Fig. 1C). The linear plasmid DNA was then introduced into rec+ cells (YB886) with selection for CmR. All CmR transformants were unable to form colonies on 50 ,ug/ml of MMS, independent of the orientation of the cat gene. The resulting recM::cat allele (the one in which direction of transcription of the cat gene is shown in Fig. IC), and the mutant strain carrying it are designated recMl and BG128, respectively. Furthermore, using the same technique we could insertionally inactivate the o4107 gene, but we failed to do so with the dnaH gene (data not shown). A recM lesion renders B. subtilis cells very sensitive to the killing action of 1 mM 4-nitroquinoline-l-oxide (4NQO). The
Nucleic Acids Research, Vol. 18, No. 23 6773 orl C
A E5
E6
E20
B
EX
C
E19
[E27
[E13[E17l E21 E14 E26
E;4
E22------
5Kb P
H
arf ,107
dJr X
B
N
Nc
X Pv
recM
H
E
recMl3
orf 87
recM27 recD41
dna HS51
L~~~~~~~~~~~~~~~~~~~r X4
+
+
_
_
Nd
-
+
+
Nd
-
_
+
+
+
-
Nd Nd
+
-
C I
1522
2000
3000
lil 3741
Nucleotide Number
Fig. 1. Map of EcoRI generated DNA fragments of the B. subtilis replication origin region (oriC) and the location of genes (part A). The position of recD is uncertain. The restriction map and the functional analysis of the EcoRI fragment 14 (E14) is shown in (part B). The arrows show the direction and extent of the open reading frames determined from the DNA sequence present in this paper. The ability or inability of the DNA segment to complement the rec- and dnats defect is indicated on the right side of the figure by + and -, respectively (Nd = not determined). Abbreviations B, Bcll; C, ClaI; E, EcoRI; H, HindII; N, NruI, Nc, NcoI; P, PstI; Pv, PvuII; X, XhoII. (part C) shows the construction of the recM: cat. The arrows indicate the direction of transcription. The ability or inability of constructed strains to form colonies in methyl methanosulphonate containing plates (300isg/ml) is indicated by + and -, respectively. The cat fusion was tested in the both possible orientations.
recMI strain (recM::cat null allele) resembles the previously described recM13 and recM27 mutant carrying strains in viability and DNA repair (see Fig. 2). Based on the above results, the 1,79 Kb EcoRI-PvuH fragment was cloned into the unique EcoRI site of the pUBi 10 low copy plasmid derivative (pUBI lOcopi), generating plasmid pBT52. An external promoter reading into the cloned DNA is present in both orientations. The plasmid was then transformed into the different genetic backgrounds and the transformant cells challenged with 1 mM of 4NQO. The presence of plasmid pBT52 fully complemented the sensitivity of all recM strains (see Fig. 2). From these results we can infer that: i) the recM complementing activity can be placed in the 934 bp HindI-Xhol DNA fragment (coordinates 2233 to 3167), ii) the recM gene is expressed in pBT52, while an insertional inactivation at the NruI site renders cells sensitive to DNA damaging agents, iii) recM and recD are two independent genes, iv) the dnaH complementing activity can be placed within the 711 bp EcoRIHindll DNA fragment and, v) the recM and orjfo7 genes are dispensable, but dnaH is strictly required for cell viability. DNA sequence analysis The nucleotide sequence of the 1.79 Kb EcoRI-Pvull DNA fragment (see Fig. iB), which has been shown to contain a functional recM gene, is presented in Fig. 3. Since genes involved in DNA metabolism have been mapped in the vicinity of recM, we extended our sequencing analysis towards the 5' and 3' ends of the EcoRI-PvuH DNA segment. Hence, in Fig. 3 the nucleotide sequence of the 1.5 kb EcoRI (E21) and the 0.42 kb PvuH-EcoRI DNA fragments are also included. From the DNA sequence
1-1
-o2
-3
CJM
.' 10 -4
10
10
20
40
Exposure time( in min) Fig. 2. Effect of the recM impairment and the plasmid-borne recM gene on the survival of cells in 1 mM 4-nitroquinoline-l-oxide. Strains tested were YB886 (x), BG127 (0 and 0), BG128 (U and Ol) and BG103 (A and A). Filled symbols denote the presence of the plasmid-bome recM gene (pBT52), open symbols plasmid-free cells.
analysis we learned that the EcoRl fragments E21 and E14 are contiguous DNA segments and that the 1.5 kb EcoRI segment (corresponding to the EcoRI fragment E21, Fig. IA) contains the 3'-moiety of the scr gene and an orf homologous to dnaX
6774 Nucleic Acids Research, Vol. 18, No. 23 GAATTCATGAACCATGTCAGGTCCGGAAGGAAGCAGCATTAAGTGAAACCICTCATGTGCCGCAGGGTTGCCTGGCCGAGCTAACTGCT TAAGTAACG
1TTA404TACLLAAICGACAG -10
-35
121 241
AAGGTGCACGGTAAATCAATCATCAAATTTTCAGACTCACCTIATAJAGGTGAGTTT T TTGTTATGTAAAAAGAGCTIGA CGAAACAAGGT TCAT
MAST
GGAATGATGAA
TAACGGAGGAGGGCAAACCCGT GAGT TACCAAGCT TTATATCGAGTAT TCAGGCCT CAGCGCT T TGAAGAT GT GGT CGGACAAGAACACA T TACAAAAAC GC T uCAAAAT GCCC TT T TGC
sS l
T K T L 0 N A L L O V S Y 0 A L Y R V F R P S R F E D V V G S E H RBS AAAAAAAGTT T TCT CACGCCTAT CTGT T TTCCGGGCCTAGAGGAACCGGAAAAACCAGT GCAGCCAAAAT AT T TGCTAAGGC TGT CAAC TGT GAACA TGC TCC TG T TGA TGAGC CAT GCA F A K A V N C E H A P V D E P C N K K F S A A Y L F S G P R G T G K T S A A K ACGAATGT GCGGCCTGTAAAGGGATAACAAACGGGTCAATAT CCGAT GT CATAGAAAT TGACGCCGC TTC TAAT AACGGT GT TGAT GAGAT TCG TGACA T TCGCGA TAAAG TGAAGT TT G
bO
CCCCATCGGCCGTCACATATAAGGTATATATCATAGATGAGGTGCATATGCT TTCTATCGGCGCT T TTAATGCAT TACTAAAAACAT TAGAAGAGCCGCCT GAGCA T TGT ATT T TCAT AT
721
TAGCAACAACTGAACCGCACAAAATTCCTTTAACCATTATCTCTAGATGTCAGCGTTTTGACTTTAAACGAATTACCTCCCAAGCGATTGTAGGCCGCATGAATAAAATTGTTGATGCTG
361
E P
C S
A
A
A
V
C
T
K
Y
T
G
K
V
Y
N
G
S
S
D
E
V
D H
V M
E L
S
D G
AA S A
F
N
N
A
L
G
L
V
K
D T
R
E
L
E
E
P
9tl
1081 1201 321
14'41 1561
1681
b0 @21
2041l 2161 2281
P
E
D
H
K
C
V
K
F
F
A L
V C A E T S O A M V G RA F D F K R S R C P L T A T T E P H K AACAGC TGCAAGT GGAGGAAGGAT CGCT GGAAA TCAT CGCGAGT GC TGCAGACGGAGGGA TGAGGGAT GCCC TGAGCCT TC T TGAT CAGGCCAT A TCA T TCAGCGGCGA TA TCC TGAAAG L K V S F S G D A S A A D G G M R D A L S L L DO A G L O V E E G S L E T CGAAGAT GCGCT T TTAAT TACGGGT GCT GT TT CTCAAT TATATATCGGGAAGCT TGCAAAAT CCCT GCA TGAT AAAAACGT T TCT GACGCACT GGAAACAT TAAAT GAA T TGC T TCAGC R
841
R
D
N
K
L A E L L OG G K L A K S L H D K N V S D A L E l T G A V S Q L Y E D A L L AAGGAAAAGAT CCAGCTAAGCTGATAGAAGATAT GAT T TTCTAT TTCAGGGACATGC TGCT GTACAAAACAGCCCCTGGCT TAGAGGGAGT GC T TGAGAAAGT AAAAG TCGAT GAAACGT F Y F R D M L L Y K T A P G L E G V L E K V K V D E T F E D M G K D P A K L T CCGGGAACTAAGT GAACAAAT TCCGGCTCAGGCCC TATAT GAAAT GAT TGATAT TCT GAACAAGAGCCAT CAGGAAATGAAATGGACAAAT CAT CCGCGT AT C TTT T TCGAAGT GGCCG F F E V A V D L N K S H O E M K W T N H P R R E L S E OI P A S A L Y E M T TGT GAAGAT T TGCCAAACCTCACATCAAT C TGC TGC TGATCT CCCGGAAGT GGATAT GC TGAT GAAAAAAAT CCAGCAGCT CGAACAGGAAGT AGAGCGGC TCAAAACAAC GGGCAT AA K E V E R L K T T G L EK K S S V K C S T S H O S A A D L P E V D M L M AAGCAGC TGCGGAAAGCCCGAAAAAAGAAGCGCCGCGT GT GCCAAAGGGCGGGAAAT CCAAT TACAAAGCGCCAGT AGGCAGAA T TCA TGAGA TT T TGAAGGAAGC CACAAGACCGGAT C L K E A T R P D L H E A A A E S P K K E A P R V P K G G K S N Y K A P V G R
TTGATCTGCTCAGAAACAGTTGGGGCAAGCTTCTTGCTCATCTCAAACAGCAAAACAAAGTATCACATGCCGCTTTGCTGAATGACAGTGAACCTGTGGCTGCCGGATCGGCGGCATTTG N K V S H A A L L N D S E P V AA G SA A AF V D L L R N S W G K L L A H L KE 0 T TCT GAAAT TCAAATATGAAAT TCAT TGTAAAAT GGTCGCTGAGGATAACAACGGAGTCCGGACTAATCT TGAGCAGAT T TTAGAAT CGAT GCT CGGAAAAAGAAT GGA T TTGAT TGGCG G V H C K M V A E D N N G V R T N L EO I L E S M L G K R M D L E L K F KE T TCCAGAAGCACAAT GGGGT AAAAT AAGAGAAGAAT T TT TAGAAGAT CAT CAGCAGGCAAAT GAAGGAT CAAA TGAACCGGC TGAAGAAGACCCGCT TA T TGC CGAAGCGAAAAAGC TT G A E A K L V A OO A N E G S N E P A E E D P L R E E F L E D H P E A W G K T TGGAGCGGAT T TAAT TGAAATAAAAGAC TAACAAAAT GAAAGAGAGT GAATGC TA TGCGT GGCGGAAT GGG TAA TA TGCAAAAAA TGAT GAAACAAA TGCAAAAAA TGCAAAAGGACA T M R G G M G N M S K M M K S M S K M S K D M RBS K D * E G A D L GGCGAAGGC TCAAGAAGAGC TTGCAGAAAAGGT TGT TGAAGGAACT GCAGGCGGCGGTAT GGTAACAGTAAAAGCAAACGGCCAAAAAGAAAT CC TGGA TGT TAT CAT CAAAGAAGAAGT K E E V L D V A K A S E E L A E K V V E G T A G G G M V T V K A A G O K E CGTTGATCCTGAAGATATTGATATGCTTCAAGAT T TAGTGCT TGCTGCAACGAATGAAGCT T TGAAAAAAGTTGACGAAATCACAAACGAAACAATGGGTCAAT T TACAAACGCAATGAA D M F T K G M N G T N E T D M L O D L V L A A T N E A L K K V D E V D P E D
CAT GCCAGGTTTATTCTAGGGGGATAAAGAACATGCAATATCCTGAACCAATATCAAAGCTGATTGACAGCTTTATGAAATTGCCAGGGATCGGACCGAAAACAGCGGTTCGTCTGGCT M S Y P E P G P K T A V R L A S K L D S f M K L P G M P G L F * TT TT T TGT TCTAGGTATGAAAGAAGATGTAGTAT TAGAT T TTGCGAAAGCAT TAGTAAATGCGAAACGCAACCT GACATAT TGT TCAGT T TGCGGGCAT A TTACAGAT CAGGACCCT TGC T D O D P C EK R N L T Y C S V C G H F F V L G M K E D V V L D F A K A L V N
RBS
240 1
2521
TATATATCGTGAGATACGCGCAGGGATAGTACTGTTATCTGTGTTGTGCAAGACCCTAAAGACGTTATCGCTATGGAGAAAATGAAGGAATACAACGGACAGTATCACGTTCTTCACGGC
2641
Y H V L H G A M E K M K E Y N G S l C V V O D P K D V C E D T R R D K G V GC TAT T TCTCCAAT GGACGGCATCGGACCGGAGGATAT TAAAAT ACCAGAAT TGT TAAAACGAT TACAGGAT GAT CAAGT GACAGAAGTGAT CC TCGCGACAAACCC TAAT A TAGAAGGG E G P F L L K R L O D D O V T E V I L A T N P N K G P E D A S P M D G
2761 2881
3 301 3121
3241
GAAGCAACAGCGATGTATATATCAAGGCTCCTCAAGCCGTCTGGTATT AAGCTCTCCCGTATTGCCCACGGACTGCCCGTCGGCGGTGATTTGGATATGCTGATGAGGTCACTCTTTCTA
A H G L P V G G D L D M L M R S L F L K L S R S R L L K P S G E A T A M Y AAGCAC T TGAAGGAAGACGT GAAT TGTAAGGAGGAAAAAGCGA TCCAT GGGT T TTCT TCGCAAGAAAACAT TAAGAAGAGAGT T TGAT GAAAAAC TAACCGAGCAGCT TT T TAAGCAAAA K K R V * H G F S S O E N K H L K E D V N C K E E K A F K S K M G F L R K K T L R R E F D E K L T E S L RBS GGAAGAG TGGAACAGGCAAAtAAAAGCT GGT TGAAAAAAGC T TAGAACCGT CT GC TGAAGT GTTAT ACGAAC TGAAAGTAGCT GAAGCGAAGTA TT TT TT T TAC T TAAGGGAAGCGAAGCA A L Y E L K V A E A K Y F F Y L R E A EK K F O E E W N R O K K LAKE GCGAAAT TTAAAAATCAGCCGGT GGAAGTAAGT GGT CTAAACT CCT GGATCT TCT CATAAGCT TGTACTAGAACAAGCGAAGGAGATGAGAAGAT TCAT GGAGCC TA T T TTTA TTAT TGG R E P I F G RBS R N L K S R W K ' GCTT GCTGGGTTTGTGTAAA N C L V A F V A A L L V G V K T G L L F L S G S A A K P L K W L G L V
GATTATTTTAGGACTGGTTATTCTTCTTTTTTTATCAGGTTCAGCAGCGAAGCCTTTAAAGTGGATTGGCATCACAGCTGTITAAATGTCTGTGGCAGGT
3481
TATGT TTGGCGGCAGTCTTTGCATTCATGTGCCGATTAATCTGGTTACAACAGCTATCAGCGGAATTTTAGGAATACCCGGAATAGCTGCGTTAGTCGTCATTAAGCAATTTATCATTTA K S F A A L V V MV P P G L G S G N L V T T A M F G G S L C 35 - 10 -35 TAAATG TTTTTGAAAAGAAAATGCTAAAAAGT ATT
3601 3721
_ -10 TFATGA+ATAAAGTCGCT TAAACGAGCGGT AAACAAAGT TCT T TGAAAACTAAACAAGACAAAACGTACCT GT TAAT TCATT T TTATAAAT CGCACAGCAAT GT GCGT AGT CAGTCAAA CTACTTTATCGGAGAGTTTGA
3361
GATATTATGTAIE3AGACAACTGAAGGTG93___ATATACGTCGCTGATGACGAACAGCT
TGTGACTCAGCCGG
Fig. 3. Nucleotide sequence of the B. subtilis recM and adjacent genes. The nucleotides are numbered from the EcoRI at the 5'-end of the B. subtilis dnaX gene. The deduced amino acid sequences are shown below the nucleotide sequences. The boxed nucleotide sequences denote the RNA polymerase consensus sequences (-35) and (-10). The underlined and bolded letters indicate the potential ribosomal binding sites (RBS). A potential terminator sequence in the 5'-noncoding region of the dnaX gene is indicated by converging arrows. Broken arrows mark a potential stem and loop structure that could enhance frameshifting (see 42; 43; 44). Wavy arrows indicate the mnRNA initiation sites (see 19).
from E. coli, as reported by Struck et al., (31, 32). We have corroborated that the nucleotide sequence reported previously is identical to the one determined in this work (see 20, 31, Fig. 3). The analysis of the sequence between the 3'-moiety of the scr gene and the 5'-moiety of the rrnA gene reveals five major orfs (see Table I). The first open reading frame (dnaX) is preceded by typical RNA polymerase consensus regions at nucleotides 200-205 (-35) and 223-228 (-10) (31). The utilization of this promoter region was identified in vivo (by measuring its activity using a promoterless cat reporter gene) and in vitro (by a RNA polymerase binding assay and by run-off transcription)
(32). Between the scr and dnaX genes a putative rho-independent terminator with a hairpin-forming region beginning at nucleotide 153 can be detected (see also 31). The second orf is preceded by the sequence AGAGAG (see Table I). Flower and McHenry (33) have detected an identical sequence for its E. coli counterpart (see below) and suggested that it could correspond to the RBS. A putative RNA polymerase consensus region at nucleotides 2073-2078 (-35) and 2098 -2103 (-10) within the orfl07 was predicted, however, the window between the two consensus regions is larger (19 nucleotides) than expected.
Nucleic Acids Research, Vol. 18, No. 23 6775 Table I. Open reading frames predicted from the DNA sequence
orf dnaX
odf07 recM
orfl4 orfl7
RBSa AGGAGG AAAGAGAGf AGGGOG AAGGAGG AAGGAGATGA
window b
startC
8 8 10 12 8
GTG ATG ATG ATG ATG
stopc (261) (1976) (2314) (2927) (3218)
TAA TAG TGA TAA TAA
(1950) (2297) (2965 (3149) (3479)
extentd
molecular weigthe
1689 321 651 222 261
62.5 11.7 24.2 9.2 9.0
a nucleotide sequence with extensive complementarity to the 3' terminal region of the B. subtilis 16S rRNA. b the number of nucleotides between the proposed ribosomal binding site (RBS) and the initiator codon define the window. c sequence position (between parenthesis) corresponding to the first and the last codon. d extent in nucleotides. e predicted molecular weight in kDa of polypeptide after translation of the open reading frame. f see text and ref. 33
From our in vitro deletion analysis and complementation experiments, we conclude that the third orf, which is potentially translatable, is the recM gene. Downstream from orJ87, two RNA polymerase consensus regions have been identified and the 5'-ends of the rrnA transcripts reported (see 19, Fig. 3). The analyzed DNA region is characterized by short distances between the end of one orf and the beginning of the following one. Since by computer-assisted analysis we have failed to detect promoter(s) and rho-independent terminator(s) other than the ones described in the text (see 19, 31, Fig. 3) we assume that these genes (dnaX orflO7 recM orf74 orJ87) are part of an operon. Alternatively, two putative mRNAs, one starting upstream from dnaX and a second one beginning within the orJ107 could be synthesized. Identification of the recM product To visualize and possibly overexpress the recM gene product, the 1.5 kb ClaI-PvuIl fragment that contains this gene (and orflO7 and orf74), was cloned into AccI-SmaI-cleaved pT712, generating plasmid pBT56. An in vitro deletion of the 188 bp NruI-NcoI fragment (coordinates 2737 to 2925) within the recM gene was similarly constructed (plasmid pBT57). Here the last 67 residues of the RecM protein are replaced by 14 new residues and the RBS of the orf74 is not present. Thus, it is expected that the molecular mass of the truncated RecM protein should be 17.0 kDa and no synthesis of the orf74 should be observed. Plasmids pT712, pBT56 or pBT57 were introduced into the E. coli T7 RNA polymerase expression system (BL21DE3 strain), bearing plasmid pLysE (13, 18). The pattern of labelled proteins is shown in Fig. 4. With plasmid pBT56, the main proteins were 24.5 and 11.2 kDa in size. When plasmid pBT57, which contained an in vitro truncated recM gene was analyzed, a product (17.0 kDa) of expected mobility predicted from the use of the ATG initiator codon at coordinate 2314 and the 11.2 kDa product were observed (see Fig. 3). We failed to detect, in E. coli, a product that could be encoded by the putative orf74. From these results, we conclude that, under denaturing conditions, the molecular weights of the proteins encoded in the ClaI-PvuII DNA fragment are 24.5 kDa (recM gene) and 11.2 kDa (orfl07), which are consistent with the sizes predicted from the DNA sequence (24.3 and 11.7 kDa). At present it remains unknown whether the putative orf74 is expressed in the homologous host. Under our experimental conditions the expression of the orf07 gene, which has an atypical RBS, seems to be 2- to 3-fold higher than the recM gene. Similar results were reported for its E. coli counterpart (see 34).
.I3 _ --
'U
1
-
+
3
Fig. 4. Autoradiogram of 14C-protein hydrolysate-labeled extracts from E. coli BL21DE3 strain. The labeled extracts from cells contaiiing plasmids pT712 (lane 2), pBT56 (lane 3), or pBT57 (lane 4) were subjected to sodium dodecyl sulfate/20% polyacrylamide gel electrophoresis. Lanes 1 and 5 contain the molecular weight markers. dno X
orf 12
recR
109a
646ao
52bp
201aa
E coli
-lbp
'
23bpl
4bp
|
dnaX
orf 107
rec M
562cia
107ac
217cc
B.subtilis
Fig. 5. Comparison of gene organization. The E. coli dnaX-recR region (33, 34, 36) was compared to the B. subtilis dnaX-recM region (31, this report). Arrows indicate open reading frames. Shaded areas show regions homologous in amino acid sequences. The distance between the open reading frames is indicated.
DISCUSSION Our studies indicate that the B. subtilis recM gene, which resides between dnaX and rrnA genes, could be replaced by a segment of DNA coding for resistance to chloramphenicol (cat gene) without a major effect on cell viability. All the defects caused
6776 Nucleic Acids Research, Vol. 18, No. 23 1
! _AA ./
400
200
I
600
50
100
150
200
.
200
/
/
400
'.0
LrI
150
C-
(N.
0
0 C..
C.
aU
0
Ca1
'2FE
Jt
/
-200
100
C-i
OD
C-,
C-)
50
/
Eco
dna X
to 643
Eco recR lto201
Eco orf 12 1 tolO9
Fig. 6. Dot matrix analysis of homology between homologous pairs of gene products from E. coli (Eco) and B. subtilis (Bsu) origin. Segments of 10 residues from the vertical axis were compared with segments from the horizontal axis, and a dot was placed at the appropriate distance when the total score for comparisons was 6 or more.
by this mutation are complemented by a plasmid-borne recM gene. On the basis of map position and phenotype it has been suggested that the recD and recM mutations affect the same gene (35). We show here that recD and recM must be two independent genes because the cloned recM gene does not restore MMSR to the recD41 mutant; however, it does restore resistance to the recM mutants. The 3.7 kb recM region has been cloned and sequenced. DNA sequence analysis revealed five orfs, which are transcribed from the same strand of DNA and encode polypeptides of 62.5 kDa, 11.7 kDa, 24.2 kDa, 9.2 kDa and 9.0 kDa, respectively. When the recM open reading frame was expressed using a T7 expression system, a 24.5 kDa polypeptide was observed. The organization of genes in the dnaX-orflO7-recM interval of B. subtilis and the dnaX-orfl2-recR region of E. coli and the relative location of each homologous gene pair is remarkably conserved (see 33, 34, 36, Fig. 5). In B. subtilis, downstream from the recM gene, two small orfs (74 and 87 amino acids long) and the rrnA gene are present, whereas in E. coli downstream from recR were mapped the htpG and adk genes (37). The direction of transcription of the homologous region of the chromosomes is also similar in both bacteria. The DNA sequence upstream of the E. coli dnaX gene revealed two potential promoters flanked by putative dnaA boxes (33). Hence, a DnaAdependent transcription termination could modulate dnaX expression (see 38). In B. subtilis the dnaX gene is preceded by a typical promoter, but here no DnaA consensus regions (see 2) were identified. The utilization of the B. subtilis dnaX promoter could be regulated by two putative antisense RNA molecules (32). In both bacteria the genes downstream from the dnaX gene seem to be part of an operon (34, this report). Amino acid sequence comparison between homologous pairs of gene products from E. coli and B. subtilis is shown in Fig. 6. The overall homology between the dnaX gene product of E. coli and B. subtilis is 29.2%, but increases to 48.2% when conserved amino acids are also scored. The N-terminal domains of the proteins (first 250 amino acids) have strong homology, whereas little or no homology is observed at the carboxyl-termini of both proteins. Walker et al., (39) identified two amino acid sequence motifs (motif A and B) which are present in most nucleotide binding proteins. A putative ATP binding motif GXXGXGKT (motif A) was identified between residues 44 to
RecM RecR
RecM
1 MQYPEPISKLIDSFMKLPGIGPKTAVRLAFFVLGMKEDVVLDFAKALVNA 50 *11 ** I* ** * ***I*** * *I** 1* 1 MQTSPLLTOLMEALRCLPGVGPKSAORMlAFTLLQRDRSGGMRLAQALTRA 50 51 KRNLTYCSVCGHITDQDPCYICEDTRR.DKSVTCVVODPKDVIAMEKMKE 99 ** * -***I' * *l *I* 1*1*1 * ** 51 MSEIGHCADCRTFTEQEVCNI CSNPRRQENGQT CVVESPADIYAI EQTGQ 10 0
-1
RecR
RecM 100 YNGQYHVLHGAISPMDGIGPEDIKIPELLKRLODDQVTEVILATNPNIEG
*1**** *I**
*
**
149
*****I*
RecR 101 FSGRYFVLMGHLSPLDGIGPDDIGLDRL.EQRLAEEKITEVILATNPTVEG 150 RecM 150 EATAMYISRLLKPSGIKLSRIAHGLPVGGDLDrILMRSLFLKHLKEDVNCK
***
*
***|**
199
*
RecR 151 EATANYIAELCAQYDVEASRIAHGVPVGGELErIVDGTTLSHSLAGRHKIR
200
FSg. 7. Comparison of amino acid sequences between RecM and RecR. The amino acid sequences, shown by a single letter symbol, are deduced from nucleotide sequences of cloned genes from B. subtilis and E. coli, respectively. Conserved amino acids are marked by asterisks, and conserved replacements by a vertical line. The putative zinc finger and ATP-binging motifs are bolded and underlined.
51 (see Fig. 3) of the deduced dnaX product (35, 31). The B. subtilis dnaX is 81 amino acids shorter than its counterpart in E. coli. In E. coli, the dnaX (formerly termed dnaZ and dnaX) gene has one orf for a 71 kDa polypeptide from which two distinct DNA polymerase Ill subunits [gamma ('y, 52 kDa) and tau (-r, 71 kDa)] are produced (33, 36, 40, 41). Recently. it has been shown that translational frameshifting generates the 7y subunit (42-44). A similar situation may exist in B. subtilis, since there is the enigma that the two mutations dna-8132 an dnaH are located within this region, but they are believe to be associated with two distinct genes (10). We are interested to pursue biochemical evidences for the possible existance of translational frameshifting within the B. subtilis dnaX gene. The homology between the B. subtilis orflO7 and the or112 of E. coli (34) is 39.8%, and increases to 58% when conserved amino acids are aligned. The E. coli polypeptide is two residues longer than that of B. subtilis. The B. subtilis recM and the E. coli recR genes are associated with AddABC- and the RecBCD-independent mechanism of DNA repair (45, 46, 5). Conservation of 43.2% that increases to 61.3 % when conserved amino acids are allowed suggests that they are very similar in structure and function. An amino acid sequence motif CXXC-(X)n-CXXC which has been found to be
Nucleic Acids Research, Vol. 18, No. 23 6777 associated with DNA-binding and metal-binding (zinc finger, 47-49) is located between residues 57 to 72 of the deduced RecM product (see Fig. 7). A putative ATP binding motif (motif B, 39) was identified. Here, an aspartate residue (position 180, Fig. 7) preceded by hydrophobic amino acids correspond to the nucleotide binding pocket. Furthermore, the amino acids in this region (residues 178 to 180) conform to a consensus sequence for ATP-binding derived from phosphofructokinase (50). Both the zinc finger and the ATP-binding motifs are conserved between recM and recR products (see Fig. 7). The CXXC-(X)n-CXXC motif is also present in proteins which bind damaged sites in DNA as the E. coli UvrA (51) or the Saccharomyces cerevisiae RAD18 (52) proteins. RecM mutants are defective in postreplication repair and showed a reduced and delayed induction of the SOS response (5, 53). One possibility for the action of RecM is that it binds the damaged site in the template DNA. The RecM protein could bind a nucleotide such as ATP, and the bound form of RecM might have higher affinity for damaged DNA. Thus, the binding could amplify the SOS inducing signal and allow activation of the RecE protein (54) to a form that mediates LexA-like repressor cleavage. Another possibility, not excluded, for the action of RecM is that it is a transcriptional regulator, via its zinc finger, of other genes involved in DNA repair (48, 49).
ACKNOWLEDGEMENTS We are very grateful to the following people: T.A.Trautner and H. Yoshikawa for critical reading of the manuscript and their continuous interest in this project; to B. Lloyd and J. Struck for the communication of results prior to publication; to M. Marahiel for providing as plasmid p516-20 and to A.C. Stiege and G. Luder for excellent technical assistance. This work was supported in part by Deutsche Forschungsgemeinschaft (SFB 344) and a Grant-in-aid for Special Project Research from the Ministry of Education, Science and Culture, Japan.
REFERENCES 1. Ogasawara, N., Moriya, S., von Meyenburg, K., Hansen, F.G. and Yoshikawa, H. (1985) EMBO J. 4,3345-3350 2. Fukuoka, T., Moriya,S., Yoshikawa, H. and Ogasawara, N. (1990) J. Biochem. 107,732 -739 3. Ogasawara, N., Fujita, M.Q., Moriya, S., Fukuoka, T., Hirano, M., and Yoshikawa,H. (1989) In Bacterial Chromosome (Riley,M and Drlica,K. eds) pp287-295, Amarican Society for Microbiology, Washigton,D.C. 4. Ogasawara, N., Moriya, S., Mazza, P.G., and Yoshikawa, H. (1986) Gene
45,227-231 5. Alonso, J.C. Tailor, R.H. and Luder, G. (1988) J. Bacterial. 170,3001-3007 6. Lampe, M.F. and Bott, K. (1985) J. Bacterial. 162,78-84 7. Armengod, M.E., Garcia-Sogo, M. and Lambies, E. (1988) J. Biol. Chem.
263,12109-12114 8. Quinones, A. Kaasch, J., Kaasch, M., and Messer, W. (1989) EMBO J.
8,587-593 9. Menzel, R. and Gellert, M. (1987) J. Bacteriol. 169,1272-1278 10. Dean, D.H. and Zeigler, D.R. (1989) Catalog of the Bacillus Genetic Stock Center, Fourth edition. Ohio. U.S.A. 11. Alonso, J.C., Stiege, A.C. Tailor, R.H. and Viret, J.-F. (1988) Mol. Gen. Genet. 214,482 -489 12. Struck, J.C.R., Alonso,J.C., Toschka,H.Y. and Erdmann,V.A. (1990) Mol. Gen. Genet. 222, 441-445 13. Studier, F.W. and Moffatt, B.A. (1986) J. Mol. Biol. 189,113-130 14. Maniatis, T., Fritch, .F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York 15. Leonhardt, H. (1990) Gene 94,121-124
16. Iordanescu,S. (1975) J: Bacteriol. 124,597-601 17. Yanisch-Perron, C., Vieira, J. and Messing, J. (1985) Gene 33,103-119 18. Moffatt, B.A. and Studier, F.W. (1987) Cell 49,221-227 19. Ogasawara, N., Moriya, S., and Yoshikawa, H. (1983) Nucleic Acids Res. 11,6301-6318 20. Ogasawara, N., Moriya, S., and Yoshikawa, H. (1985) Nucleic Acids Res. 13,2267-2279 21. Robertson, J.B., Gocht,M. Marahiel, M.A. and Zuber, P. (1989) Proc. Nati. Acad. Sci. U.S.A. 86,8457-8461 22. Biswal, N., Kleinschmidt, H.C., Spatz, H.C., and Trautner, T.A. (1967) Mol. Gen. Genet. 100,39-55 23. Rottlander, E., and Trautner, T.A. (1970) Mol. Gen. Genet. 108,47-60 24. Alonso, J.C. and Trautner, T.A. (1985) Mol Gen. Genet. 198,427-431 25. Birnboim, H.C. and Doly, J. (1979) Nucleic Acids Res 7,1513-1523 26. Henikoff, S. (1984) Gene 28,351-358 27. Sanger, F., Nicklen, S., Coulson, A.R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,5463-5467 28. Laemnmli, U.K. (1970) Nature 227,680-685 29. Brendel, V. and Trifanov, E.N. (1984) Nucleic Acids Res. 12,4411-4427 30. Devereux, J., Haeberli, P., and Smithies, 0. (1984) Nucleic Acids Res. 12,387-395 31. Struck, J.C.R., Vogel, D.W., Ulbrich, N. and Erdmann, V.A. (1988) Nucleic Acids Res. 16,2720 32. Struck, J.C.R. (1988b) Dissertation Thesis, Freie Universitat Berlin. 33. Flower, A. and McHenry, C.S. (1986) Nucleic Acids Res. 14,8091-8101 34. Mahdi, A.A, and Lloyd, R.G. (1989) Nucleic Acids Res.17,6781-6794 35. Mazza, G. and Galizzi, A (1978) Microbiologica 1,111- 135 36. Yim, K.-C., Blinkowa, A. and Walker, J.R. (1986) Nucleic Acids Res. 14,6541-6549 37. Brune, M., Schumann, R. and Wittinghofer, F. (1985) Nucleic Acids Res. 13,7139-7151 38. Schaefer,C., and Messer,W. (1989) EMBO J. 8,1609-1613 39. Walker,J.E., Saraste,M., Runswich,M.J. and Gay,N.J. (1982) EMBO J. 1,945 -951 40. Lee, S.-H., Kanda, P., Kennedy, C., Walker, J.R. (1987) Nucleic Acids
Res. 15,7663-7675 41. Tsuchihashi, Z. and Komberg, A. (1989) J. Biol. Chem. 264,17790- 17795 42. Tsuchihashi, Z. and Komberg, A. (1990) Proc. Natl. Acad. Sci. U.S.A.
87,2516-2520 43. Blinkowa, A.L. and Walker, J.R. (1990) Nucleic Acids Res. 18,1725-1729 44. Flower, A.M. and McHenry, C.S. (1990) Proc. Natl. Acad. Sci. U.S.A.
87,3713-3717 45. Telander-Muskavitch, K.M. and Linn, S. (1981) In The Enzymes (Boyer, P.D. ed.) Vol XIV part A, pp 234-250, Academic Press, New York 46. Mahdi, A.A, and Lloyd, R.G. (1989) Mol. Gen. Genet.216,503-510 47. Miller, J., McLachlan, A.D. and Klug, A. (1985) EMBO J.4,1609- 1614. 48. Klug, A. and Rhodes, D. (1987) Trends Biochem. Sci. 12,464-469 49. Berg, J.M. (1990) J. Biol. Chem. 265,6513-6516 50. Serrano, R. (1988) Biochem. Biophys. Acta 947,1-28. 51. Doolittle, R.F., Johnson, M.S., Husain, I., Van Houten, B., Thomas, D.C. and Sancar, A. (1986) Nature 323,451-453 52. Jones, J.S., Weber, S., and Prakash, L. (1988) Nucleic Acids Res. 16,7119-7131 53. Gassel, M. and Alonso, J.C. (1989) Mol. Microbiol. 3,1269-1276 54. Lovett, C.M. and Roberts, J.W. (1985) J. Biol. Chem.260,3305-331
