Natural C-independent expression of restriction endonuclease in a C protein-associated restriction-modification system

Abstract

Restriction-modification (R-M) systems are highly prevalent among bacteria and archaea, and appear to play crucial roles in modulating horizontal gene transfer and protection against phage. There is much to learn about these diverse enzymes systems, especially their regulation. Type II R-M systems specify two independent enzymes: a restriction endonuclease (REase) and protective DNA methyltransferase (MTase). Their activities need to be finely balanced in vivo Some R-M systems rely on specialized transcription factors called C (controller) proteins. These proteins play a vital role in the temporal regulation of R-M gene expression, and function to indirectly modulate the horizontal transfer of their genes across the species. We report novel regulation of a C-responsive R-M system that involves a C protein of a poorly-studied structural class - C.Csp231I. Here, the C and REase genes share a bicistronic transcript, and some of the transcriptional auto-control features seen in other C-regulated R-M systems are conserved. However, separate tandem promoters drive most transcription of the REase gene, a distinctive property not seen in other tested C-linked R-M systems. Further, C protein only partially controls REase expression, yet plays a role in system stability and propagation. Consequently, high REase activity was observed after deletion of the entire C gene, and cells bearing the ΔC R-M system were outcompeted in mixed culture assays by those with the WT R-M system. Overall, our data reveal unexpected regulatory variation among R-M systems.

Figure 1.

Csp231I R-M system and its transcription initiations. (A) Genetic organization of the Csp231I R-M system comprising of its regulator (C gene and its promoter P_C), REase (R gene and its two promoters: major P_R1 and minor P_R2) and MTase (M gene and its P_M) (not to scale). The identified promoters are designated by arrows. The operator for P_C promoter, C-box, consists of two inverted repeats CTAAG-n₅-CTTAG, left and right, marked as O_L and O_R respectively. A presence of Rho-independent transcription terminator in the 152-nt intergenic region separating R and M genes is predicted and depicted here as hairpin. (B) Identified promoters activity was measured as a transcriptional fusion of appropriate DNA fragments with indicated promoter to the reporter lacZ gene. In case of P_C, deletion of left operator (ΔO_L) also has been tested. The transcription activity in context of C protein absence (vector control, white bars) or C protein presence (pBAD-CWT, dark bars) is presented in Miller units. The results are the averages (±SD) of at least three independent experiments. (C) Mapping the transcription start sites for the promoters with confirmed activity. For each reaction, total RNA from E. coli harboring the p18 plasmid with entire, functional R-M system was used as template for primer extension method using radioactively labeled primers and performed as indicated in Materials and Methods. The primer extension products (marked as +1) were resolved on a denaturing 6% polyacrylamide gel along with the nucleotide sequencing reactions (G, A, T, C) performed with the same labeled primer and appropriate DNA template. At the bottom, sequence of relevant DNA containing the indicated promoters is also shown. The -10 and -35 promoter motifs are underlined and the start codons (ATG) and ribosome binding sites are in bold. In case of P_C promoter, the sequence of two inverted repeats CTAAG-n₅-CTTAG is boxed. For REase promoters, the start of transcription for major promoter P_R1 is indicated by thicker arrow and bold -10 and -35 boxes, in contrast to the minor P_R2 promoter marked by thin arrow. (D) Promoter activity for REase gene was tested by ONPG hydrolysis in plasmid constructs of pLEX3B, where reporter lacZ gene was fused in-frame to REase gene (pLEX-P_R1WT). The major P_R1 promoter (thicker bent arrow) was knocked-out by mutation of -10 box of TTAAAT→CCCGGG (pLEX-P_R1mut). The transcription activity in context of C gene presence or absence was measured as in panel B.

Figure 2.

C.Csp231I protein and its action in vitro. (A) LOGO of C proteins sequence for: C.Csp231I family (21 sequences, upper part); C.PvuII family (65 sequences, bottom part). Helical structure, designated as seven white tubes, is derived from C.Csp231I crystal analysis (45). The regions of proteins responsible for activation of transcription and DNA recognition are inferred based on structural analysis of C.AhdI, C.Csp231I or mutational analysis other C.PvuII-like members. Grey boxes represent well conserved residues with the highest similarity regions between the two C protein families: C.PvuII and C.Csp231I. Logo was generated from software at http://weblogo.berkeley.edu. (B) Preparations of C-terminal His-tag fusion of WT C protein (3μg) and its variants: C-ARQmut (2.8μg) and C-SQEmut (2μg), resolved on a 10% acrylamide Tricine SDS gel and Coomasie Blue stained. Overproduction and purification were carried out as described in Materials and Methods section. (C) C.Csp231I and its variants binding to the P_C promoter/operator region (C-box) containing the two inverted repeats CTAAG-n₅-CTTAG. A 576-bp target DNA fragment was prepared by PCR amplification, as well as its control with no C-box sequence (515-bp) containing a DNA fragment of comparable size but lucking the C-box (Supplementary Table S2). Each binding reaction was carried out with the same amount of DNA (100 nM) and increasing concentration of proteins: 0, 200, 500, 1000 and 2000 nM. Reactions were processed further as outlined in Materials and Methods and finally resolved on 6% native polyacrylamide gels. DNA was visualized by ethidium bromide staining. Open and filled arrows denote positions of unbound DNA and protein–DNA complexes, respectively. The comparable data were obtained with modified protocol (Supplementary Figure S3).

Figure 3.

C.Csp231I is a transcriptional positive and negative autoregulator. Cells were grown in minimal media with 0.2% glucose (glu) or arabinose (ara) at indicated concentration from 0–0.2%. Cells contained two plasmids: one with WT promoter or its variant fused to reporter gene, second, compatible plasmid delivered WT C gene under inducible P_araBAD promoter, its C variant or C-absent vector. (A) Transcriptional profiles were measured as LacZ specific activity. Circles represent values for WT C gene promoter/ operator including two inverted repeats (O_L and O_R) fused to lacZ reporter gene in p20 plasmid. Diamonds represent values for C gene promoter with O_L operator deleted and fused to lacZ reporter gene in p12 plasmid. As a source of C inducible expression following plasmids were used: pBAD-CWT; pBAD-sqe; pBAD-arq or pBAD33 vector as a C-negative control. For plot clarity, C.Csp231I effect on its own transcription is represented by connected points: black circles in C presence and open circles in C absence. (B) Crude extracts from cells showing WT C-dependent transcription profile indicated by black circle in panel A, were resolved by 10% acrylamide Tricine-SDS PAGE and analyzed by western blotting using the rabbit anti-C.Csp231I polyclonal antibodies to detect C protein level (visualized by chemiluminescence). Bottom part serves as a protein loading control stained by coomassie brilliant blue. (C) Effect of altered C-box on C-dependent transcription profile. Mutation within O_L or O_R were introduced (CTTAG→GTATC; pLex15-O_Rmut or pLex15-O_Lmut) and measured as translational LacZ activity in WT C protein presence (circles; pBAD-CWT) or C absence (diamond, pBAD33). Transcriptional (panel A) and translational (panel C) activity was measured as LacZ specific activity determined by linear regression of the slopes for the lines generated by plotting LacZ activity versus optical density of the culture (modified Miller units) (53). Error of each point was measured with R-squared value not less than 0.94. Note, LacZ activity is from a transcriptional fusion in panel A, and from a translational fusion in panel C.

Figure 4.

Transcripts initiated from P_C are bicistronic. (A) Schematic diagram of the C and R::lacZ genes in tested plasmid (not to scale). Triangles depict the primers (a–f) used in PCR reaction, where cDNA reversely transcribed from E. coli harboring pLexWT total RNA (right panel) or control pLexWT DNA (left panel) was used. The expected PCR products were: 1 (150bp); 2 (230bp); 3 (500bp); 4 (399 bp) and 5 (300 bp). (B) Transcription activity measured downstream of P_R promoter. The reporter lacZ gene was fused in-frame to REase gene generating a series of plasmids, which important features are indicated on the diagrams (red in left panel; not to scale). Briefly: Cmut indicates a substitution: A33G; R34E; Q37A in C protein; P_R1 promoter was knocked-out as in Figure 1D. The transcription activity in context of a second plasmid carrying no C gene (pLex3B vector, white bars) or C gene (pBAD-CWT, dark bars) is presented in modified Miller units, as in Figure 3. Error of each point was measured with R-squared value not less than 0.95.

Figure 5.

C protein is not essential for REase expression, but affects the R-M system transfer efficiency. Schematic diagram of the Csp231I R-M system WT and its variants (not in scale) are presented and data for particular panel A-D are adjusted to be read horizontally. The unchanged MTase gene is not shown. Changed elements are depicted in red color. Briefly: p18 (WT); p19 (deletion of O_L from C-box sequence); p23 (substitutions: A33G; R34E; Q37A in C protein); p28 (substitutions: S16A; Q17A; E18A in C protein); p30 (deletion of C gene and its upstream region including C-box and P_C); p32 (mutation of -10 box of P_R1, TTAAAT→CCCGGG); p24 (knock-out of REase gene, XhoI cut and Klenow filling) (A) Efficiency of plaque formation was assayed as plaque forming units on tested plasmid divided by plaque forming units on pBR322. The standard deviation from at least three experiments is shown. (B) Relative restriction is indicated as 1 for WT R-M system. (C) Efficiency of transformation is a relative number of transformants obtained from the same amount of plasmid DNAs carrying the indicated R-M system variants referred to vector pBR322 (D) Level of C protein in bacterial crude extracts was tested by western blot (horizontal bands) using the rabbit anti-C.Csp231I polyclonal antibodies and visualized by nitroblue tetrazolium (NBT) as the color development reagent. Culture samples were normalized by OD₆₀₀. Loading control was prepared as one from Figure 3B; it is not shown for clarity. C.Csp231I protein is marked by arrow.

Figure 6.

Cells with the WT R-M system have a fitness advantage over cells carrying the C-deleted variant. Mixed cultures were prepared by adding equal numbers of the two competing E. coli strains into medium without any antibiotic (Materials and Methods). Each type carried plasmid with a specific R-M system variant and its distinct antibiotic marker (tetracycline or ampicillin), as indicated below the diagram. One flask co-cultures were diluted every 24 generations into fresh medium and CFUs of competing cells were measured. Relative competitive fitness (W) was estimated individually for each mixed culture represented on diagram as a single symbol, calculated as W = log(CFU_amp/CFU_tet) and normalized to vector control (pBRamp versus pBRtet) (Materials and Methods). Black diamonds represents seven separate co-cultures, where WT R-M system in ampicillin resistant cells (p18amp) competed with C-deleted R-M system (p30tet) in tetracycline resistant cells. In control, parallel co-cultures, cells with plasmids with restriction-negative and modification-positive variants (p17amp vs. p31tet; white diamonds) were used.