Xanthomonas oryzae pv. oryzae recA is transcribed and regulated from multiple promoters

Abstract

Transcription regulation of Xanthomonas oryzae pv. oryzae recA was characterized. Primer extension experiments showed that recA is transcribed from three promoters designated P1, P2 and P3. The sequences of −10 and −35 regions of these promoters have moderate homology to the proposed consensus sequence for a Xanthomonas promoter. Putative SOS boxes were identified in the vicinity of P1 and P2 promoters. Deletion analysis and in vivo monitoring of promoter activity of these promoters revealed that the three promoters have different characteristics. P1 and P2 show stress-inducible high and low promoter strengths respectively. P3 is a non-inducible moderate promoter strength. These promoters are regulated by two SOS boxes. The multiplicity of promoters and SOS boxes provides back-up systems to ensure proper regulation of recA.

Transcription, Multiple promoter, SOS box, Inducible expression

1 Introduction

RecA is a highly conserved multi-functional protein that performs crucial roles in many processes. It can act as a recombinase, an ATPase, and a protease [1]. Its contribution to regulation of the SOS response to DNA damage and its roles in DNA repair and homologous recombination have been extensively studied [2,3]. Understanding the regulation of recA expression has been a major interest and current evidence suggests that the mechanisms of gene regulation are highly conserved among diverse bacteria [3]. Similar to other genes involved in the SOS response, recA is controlled by a repressor protein, LexA [4]. In uninduced cells, LexA binds to a region upstream of the transcription start site known as an SOS box or a LexA binding site [4]. The sequence of an SOS box is highly conserved in Gram-negative bacteria [4–6]. SOS boxes are normally located close to or between the −35 and −10 regions of the promoter, so that binding of LexA to the box prevents RNA polymerase from binding to the promoter and blocks transcription initiation. Upon exposure to DNA damaging agents, the damaged DNA acts as the signal to activate the co-protease function of RecA that cleaves the LexA repressor [4]. This reduces the concentration of the repressor protein and the number of LexA–SOS box complexes and in turn, allows more RNA polymerase to bind at LexA-regulated promoters and initiate transcription.

The highly repeated sequences in the Xanthomonas genome especially among avirulence genes are good substrates for recombination [7,8]. This, with high levels of RecA, could lead to intragenomic recombination. The altered gene structure that would result could effect the gene function and result in changes in pathogen host range and bacterial pathogenicity. A recent report indicates that a mutation in recA affects the pathogenicity of Xanthomonas [9]. Thus, it is important to understand how expression of recA is controlled in Xanthomonas. We have isolated and characterized recA from the rice bacterial phytopathogen Xanthomonas oryzae pv. oryzae [10]. A recA mutant shows increased sensitivity to mutagens but not to H₂O₂[11]. Furthermore, recA expression is strongly induced by H₂O₂ and mutagens [12]. Here, we showed that recA of X. oryzae pv. oryzae is transcribed from multiple promoters; two of these possess SOS boxes.

2 Materials and methods

2.1 Primer extension

X. oryzae pv. oryzae was grown in SB (Silva Buddenhagen, 0.5% sucrose, 0.5% yeast extract, 0.5% peptone, 0.1% glutamic acid, pH 7) at 28°C. SOS genes were induced by exposing the cells to 500 μM H₂O₂ for 10 min. Total RNA was extracted from uninduced and induced cells using a modified hot phenol method [13]. Primer BT147 (5′-TTCGATCTGGCTCAGTGC-3′) located at 51 bp downstream of the recA translation initiation codon was radioactively labelled with γ-P³²-ATP and T4 kinase. Five microgram of RNA were mixed with labelled primer and heat-denatured at 70°C for 3 min. Primer annealing was allowed to proceed at 55°C for 60 min. The reverse transcription reaction was initiated by adding 200 Units of MMLV reverse transcriptase and the reaction mix was incubated at 42°C for 60 min [13]. The reverse transcriptase products were analyzed on a sequencing gel. The sequence ladder was obtained by mixing labelled primer with pSMA1 [10] and a recombinant plasmid carrying the X. oryzae pv. oryzae recA gene using a fmol^® DNA cycle sequencing kit from Promega.

2.2 Construction of the recA promoter cat fusion

In vivo promoter activity was monitored by cloning various promoter fragments into a low copy number promoter probe vector containing a promoterless chloramphenicol acetyl transferase gene (cat) [14] as the reporter. Briefly, the reporter-selectable marker cat-Km^R gene cassette was cloned into pUC18sfi [15] to give p18cat-Km^R. The 498-bp SacI–SalI fragment containing all three putative recA promoters and a part of the recA coding region of X. oryzae pv. oryzae (Fig. 3A) was cloned into p18cat-Km^R in front of cat. The fragment containing the recA promoter-cat-Km^R sequence was excised from p18cat-Km^R using SacI–XbaI. The terminal single-strand overhang of XbaI site was gap-filled and the fragment was cloned into similarly digested and treated pUFR027, a low copy number broad host range vector [16]. This gave a new recombinant plasmid, pA-P₁₂₃ (Fig. 3B).

3

Map of recA and structure of vectors used in recA promoter analysis. A: Restriction map of recA. Only relevant restriction enzyme sites are shown. B: The structure of various vectors used in recA promoter analysis. The vectors were constructed as described in Section 2.

2.3 Deletion analysis of the recA promoters

The 285-bp PstI fragment containing only the P1 promoter and part of the recA coding region (Fig. 3A) was cloned into p18cat-Km^R. The resultant recombinant plasmid was then digested with SacI–XbaI and the XbaI end was gap-filled. The fragment was cloned into similarly digested and treated pUFR027. This generated a new recombinant plasmid pA-P₁ (Fig. 3B). A second deletion of the promoter was created by cloning the 101-bp SacI–PstI fragment containing only the P3 promoter (Fig. 3A) into p18cat-Km^R. Again, using a similar strategy for the construction of pA-P₁, the SacI–XbaI fragment was cloned into pUFR027. This gave a new recombinant plasmid pA-P₃ (Fig. 3B). These vectors were then transferred into X. oryzae pv. oryzae by electroporation, performed as previously described [11].

2.4 Western analysis of Cat

Growth conditions for H₂O₂ and a mutagen induction of recA promoter were as described by Rabibhadana et al. [12]. Cat levels were determined by Western immunoblot analysis. Cell lysates were prepared from cultures of X. oryzae pv. oryzae harboring various plasmids as previously described [16]. Thirty microgram of total protein were loaded into each lane. After SDS-PAGE electrophoresis, separated proteins were transferred to a nitrocellulose membrane. Western immunoblot analysis was performed as previously described [17]. Briefly, the membrane was reacted with a polyclonal anti-CAT antibody (5′–3′) and the immune complex was detected using alkaline phosphatase-conjugated goat anti-rabbit antibody.

3 Results and discussion

3.1 Determination of transcription start sites and analysis of recA promoters

Currently little is known about promoters in Xanthomonas, although, a consensus sequence for Xanthomonas promoters that has conserved regions equivalent to the −10 and −35 regions has been reported [18]. However, the space between these regions varies from 17 to 43 bp [18]. This makes assignment of a promoter, based on sequence homology, difficult and inaccurate. Thus, experiments were performed to locate the recA promoter and to investigate mechanisms of recA-induction. recA transcription start sites were determined using the primer extension technique on total RNA isolated from H₂O₂-induced and uninduced X. oryzae pv. oryzae cultures. The results showed multiple primer extension products (Fig. 1A). The major transcription start site, as judged by the amount of primer extension product was designated S1. It is located at the A residue 30 bp upstream from the translation initiation codon of recA. Examination of the sequence upstream of S1 showed a putative P1 promoter. P1 has the sequences CTGTGC and CATTAG at the −35 and −10 positions which have five out of six and four out of six bases respectively, that match the consensus sequence for Xanthomonas promoters [18]. Two other minor transcription start sites, designated S2 and S3 are located 42 and 69 bp respectively from the translation initiation codon of recA. Examination of sequences upstream of S2 and S3 showed two further promoters designated P2 and P3. The sequences of the −35 region of P2 (TCGCCT) and P3 (ATGGCG) have four out of six bases matched to the consensus sequence. The sequences of the −10 region of P2 (GATACA) and P3 (GTGTGT) have five out of six and three out of six bases respectively, that match the consensus sequenced [18]. The elements of the promoter boxes of P1, P2 and P3 are separated by 15–17 bp, placing the consensus elements at −35 and −10 in similar positions to other bacterial promoters. The sequence similarities of recA promoters to the consensus sequence are illustrated in Fig. 2. At present, there is insufficient information on Xanthomonas promoters to allow promoter strength to be deducted simply by examination of promoter sequences. Analysis of Xanthomonas campestris pv. citri recA promoter [19] and predicted sequences for Xanthomonas promoters [18] suggest that the two conserved elements for RNA polymerase binding could be separated by 17–43 bases. We think the long distances between conserved regions are excessive and it is unlikely that Gram-negative bacterial RNA polymerases could recognize them. We favor a more conventional separation of the −10 and the −35 regions by 15–21 bp as found for X. oryzae pv. oryzae recA promoters and for other well characterized bacterial promoters such as those from Escherichia coli and Bacillus subtilis.

1

Primer extension analysis of recA transcription. A: Primer BT147 was used in primer extension reactions. The experiment was performed on total RNA isolated from uninduced (U) and H₂O₂-induced (H) X. oryzae pv. oryzae cultures. Primer extension products were analyzed on a sequencing gel. S1, S2 and S3 indicate three primer extension products corresponding to transcription initiation sites of the three promoters. B: The sequence of the region upstream from recA showing various regulatory elements. Transcription start sites S1, S2 and S3 are shown. —−10 and −35 regions of P1; —−10 and −35 regions of P2; —−10 and −35 regions of P3. Location of putative SOS boxes ▲?▲ and *** are shown. RBS and ATG are the ribosome binding site and translation initiation codon of recA respectively.

2

Comparison of recA promoter sequences. The sequences of conserved regions of three recA promoters were compared to the consensus sequence for Xanthomonas promoter [18]. A capital letter in recA promoters indicated a matched base between the recA promoter and the consensus sequence.

Primer extension data suggested that X. oryzae pv. oryzae recA is transcribed from multiple promoters. This raised a question as to how these promoters were regulated. Comparison of primer extension products between uninduced and H₂O₂-induced samples clearly showed that the H₂O₂ treatment induced increases in transcription initiation at all three promoters (P1, P2 and P3)(Fig. 1A). The observation confirmed previous Northern analysis results that showed induction of recA expression by H₂O₂[12]. Moreover, the induction mechanism is consistent with the current model of recA-induction in other bacteria, namely, that increased transcription initiation is responsible for increased gene expression [4].

3.2 Identification of putative SOS boxes

Increased expression of many genes in the SOS regulon, including recA, has been observed in response to DNA damage [3]. In E. coli, a consensus SOS box sequence, 5′-TACTG-(TA)5-CAGTA-3′ has been identified as the LexA binding site [4]. Six of the outer bases (CTG-N10-CAG) and the CTG at position 1–3 are thought to be crucial for LexA binding [20]. The SOS box is conserved among different Gram-negative bacteria [5]. Thus, we searched for putative SOS boxes in vicinity of the recA promoters of Xanthomonas and found no perfect match to the E. coli SOS box consensus sequence. Nonetheless, several putative SOS boxes with minor mismatches were found close to the recA promoters. Two putative SOS boxes with the sequences 5′-CTG-N9-CCG-3′ and 5′-CTG-N9-CCT-3′ are located between the −10 and −35 regions of P1 and P2 respectively (Fig. 1B). These putative SOS boxes of X. oryzae pv. oryzae each contain the important CTG sequence at the 5′-end and the trinucleotide CCG and CCT at the 3′-end. However, in Xanthomonas these elements are separated by only 9 bp compared to 10 bp for the E. coli SOS box suggesting that for the X. oryzae pv. oryzae LexA geometry of binding to an SOS box might differ from that of E. coli LexA. Additional evidence supporting the conclusion comes from the observation that the sequence of X. campestris pv. citri LexA at the amino-terminus which is responsible for the DNA binding is differed from other Gram-negative bacteria [21]. This difference could be responsible for the minor changes in the recognition sequence and spacing of putative Xanthomonas SOS boxes. Similar sequence motifs and spacing for a putative SOS box has been observed in front of X. campestris pv. citri recA [19].

3.3 Deletion analysis of recA promoters

Primer extension data indicated that expression of recA is regulated at multiple promoters. A deletion analysis of the recA promoters was undertaken to determine the contribution of each of these promoters to recA expression. First, the 498-bp SacI–SalI fragment (Fig. 3A) containing all three putative recA promoters and a part of recA coding region was cloned in front of a promoterless cat on pUFR027 vector [16] (Fig. 3B). The resultant recombinant vector pA-P₁₂₃ was then transferred into X. oryzae pv. oryzae and Cat levels determined by Western immunoblot analysis. The ability of pA-P₁₂₃ to be induced by H₂O₂ and a mutagen methyl methane sulfonate (MMS) were tested. The results show that both compounds were potent inducers of the recA promoter (Fig. 4A). The observations are consistent with our previous reports [12]. Next, we investigated the effects of varying doses of H₂O₂ on the recA promoter activity. Clearly, increased concentration of H₂O₂ treatments led to higher induction of recA promoter. The recA promoter activity increased in parallel with increased H₂O₂ concentrations up to 400 μM H₂O₂. Treatment with higher concentrations of H₂O₂ did not produce increased promoter activity from the recA promoter (Fig. 4B).

4

In vivo analysis of recA promoters. Cat levels were determined by Western immunoblot analysis performed as described in Section 2. Thirty microgram of total protein were loaded into each lane. A: X. campestris pv. phaseoli harboring pA-P₁₂₃ were treated with 200 μM H₂O₂ or 0.01% MMS (U, uninduced and I, induced). B: A dose response of the same strain as in A treated with various concentrations of H₂O₂, (U, uninduced). C: In vivo analysis of various recA promoter constructs. The designation of various plasmids are the same as in Fig. 3B, (U, uninduced and H, induced with 200 μM H₂O₂).

The recA promoter deletion analysis experiments were done. Clearly, pA-P₁₂₃ that contained the recA promoter fragment directs high H₂O₂-inducible expression of cat (Fig. 4C). The first deletion of the promoter was made by cloning the 285-bp PstI fragment, containing only the P1 promoter and part of recA coding region, in front of the cat in pUFR027 vector, as described in Section 2. The resultant recombinant vector, pA-P₁ was transferred into X. oryzae pv. oryzae and Cat levels determined with and without H₂O₂-induction (Fig. 3A,B). The P1 promoter alone directed high H₂O₂-inducible expression of cat (Fig. 4). These data are consistent with the primer extension findings that showed P1 as a strong inducible promoter. Analysis of the sequences around the P1 promoter showed a possible SOS box positioned between the −10 and −35 regions. The expression from P1 is H₂O₂-inducible, consistent with this arrangement. The second deletion was created by cloning the 101-bp SacI–PstI fragment containing only the P3 promoter in front of the promoterless cat on pUFR027 vector (Fig. 3B). The resultant recombinant vector pA-P₃ was then transferred into X. oryzae pv. oryzae. Analysis of Cat levels in cells harboring the vector showed that the P3 promoter directs moderate levels of constitutive expression of cat (Fig. 4). This is in agreement with the lack of SOS box near P3. However, these data contradicted the primer extension data which showed that P3 directed inducible expression of recA. A likely reason for this discrepancy is that in uninduced cells LexA binds to the SOS boxes downstream of the transcription start site for P3 (S3) in the vicinities of P1 and P2 (Fig. 1B). This blocks RNA polymerase binding to P1 and P2; in addition transcription elongation from P3 is also blocked. The net effect would be a reduction in recA expression. In induced cells, LexA is cleaved by activated RecA. This reduces both the concentration of the repressor and the number of SOS boxes occupied by LexA. Consequently, this allows RNA polymerase to initiate transcription at all three promoters and to read-through points of repressor blockade into recA. In summary, P1 and P2 respectively are strong and weak inducible promoters which are regulated by the LexA repressor. While, P3 is a moderate strength non-inducible promoter.

A question remains as to why there are three differentially regulated promoters to control recA expression? The answer might be in the important roles recA plays in survival strategy during stressful conditions. High RecA levels are required when cells have been exposed to DNA damaging conditions. Such conditions could arise when Xanthomonas are exposed to plant-generated H₂O₂, a part of the active plant defense response [22]. Such conditions could induce mutations in recA promoters. In the event that P1, the major LexA-regulated promoter of recA, is inactivated by mutation, regulated recA expression could be maintained by transcription from P2 and P3. Similarly, if P2 and/or P3 are lost by mutation, controlled expression of recA can still be achieved from P1. After DNA damage has been repaired, recA expression is repressed by LexA repressor binding once again to SOS boxes. Thus, by having two SOS boxes, all three promoters can be repressed by LexA. The arrangement of multiple promoters and SOS boxes controlling X. oryzae pv. oryzae recA allows for back up for both induction and repression of recA expression.

Acknowledgements

We thank P. Bennett for reviewing the manuscript and Supa Utamapongchai for performing several experiments. The research was supported by grants from Chulabhorn Research Institute to the Laboratory of Biotechnology and a NSTDA career development award RCF-01-40-005 to S.M.