Repressor of Phage 16-3 with Altered Binding Specificity Indicates Spatial Differences in Repressor-Operator Complexes
http://www.100md.com
《细菌学杂志》
Institute of Genetics, Agricultural Biotechnology Center, Gdll, Szent-Gyrgyi A. 4., H-2100, Hungary,Department of Genetics, Etvs Lorand University and Research Group for Molecular Genetics of the Hungarian Academy of Sciences, Pazmany P. Setany 1/C, H-1117, Budapest, Hungary
ABSTRACT
The C repressor protein of phage 16-3, which is required for establishing and maintaining lysogeny, recognizes structurally different operators which differ by 2 bp in the length of the spacer between the conserved palindromic sequences. A "rotationally flexible protein homodimers" model has been proposed in order to explain the conformational adaptivity of the 16-3 repressor. In this paper, we report on the isolation of a repressor mutant with altered binding specificity which was used to identify a residue-base pair contact and to monitor the spatial relationship of the recognition helix of C repressor to the contacting major groove of DNA within the two kinds of repressor-operator complexes. Our results indicate spatial differences at the interface which may reflect different docking arrangements in recognition of the structurally different operators by the 16-3 repressor.
TEXT
The regulation of gene expression occurs mostly at the level of transcription through sequence-specific DNA-protein interactions. In several cases, even a single change in the binding sequences results in loss of function due to loss or weakened binding of the protein. Alternatively, examples are also known of when the DNA binding protein has relaxed binding specificity, that is, when several changes at given positions still allow proper binding (12, 23). Only a few naturally existing systems are known where the protein has the ability to bind specifically to sequences with different lengths. The Escherichia coli cyclic AMP receptor protein (CRP) recognizes 16- and 18-bp-long binding sites, in which 6- and 8-bp central spacers, respectively, are bracketed by the recognition sequences. According to Adhya's "geometric homeostasis" method of resolution in CRP-DNA complexes, a conformational shift from B- to A-form DNA over one helical turn covering the longer spacer allows sequence-specific binding of CRP (13). The CytR repressor recognizes binding sites consisting of two octamer repeats, in direct or inverted orientation, separated by 2 bp. However, in the presence of cyclic AMP-CRP, CytR instead recognizes inverted repeats separated by 10 to 13 bp or direct repeats separated by 1 bp. It was shown that the bases for recognizing the structurally different sites were conformational changes within the CytR protein induced by protein-protein interactions between CRP and CytR (14, 21). The CI repressor of E. coli phage 186 was found to recognize two distinct DNA sequences, termed A-type and B-type sites. The A-type binding sites were different in length since half-sites were separated by either 4-bp or 5-bp spacers (26). The binding of 16-3 C repressor to its operators is another example for recognition of structurally different DNA sites by the same protein. In this system, the possibility that either conformational shifts from B- to A-form DNA within the binding sites or interactions of the repressor with another protein is involved in formation of 16-3 repressor-operator complexes was ruled out (20).
Genes, proteins, and chromosomal sites for several functions of the temperate phage 16-3 of Rhizobium meliloti 41 have been studied in detail (1, 2, 5, 8, 9, 11, 19, 22, 24, 25). The most thoroughly analyzed region of the 16-3 genome has been the immC regulatory region which encodes a lambda type CI repressor (16-3 C repressor protein), and the cognate c cistron is flanked by two operator regions, OL and OR (3, 4, 6, 16-18, 20). An interesting feature of the immC region is its partial cross-compatibility with the phage 434 immC region: although the overall sequence homology is not significant (below 15%), the helix-turn-helix motifs of the two repressors (55% identity) and the OR operators are highly similar (4). In both of the 16-3 OL and OR operator regions, two repressor binding sites were found: the OL-type operators were 12 bp long (5'-ACAA-4 bp-TTGT-3'), while the OR-type operators were 14 bp long (5'-ACAA-6 bp-TTGT-3'). To explain all previous in vivo and in vitro binding data, we proposed the "rotationally flexible protein homodimers" model. Our model hypothesizes alternative interactions for the dimerization domains as the basis for structural flexibility in recognizing the two types of binding sites (20). The model supposes, based on analyses of repressor binding to mutant and truncated versions of both types of operators (for a review, see Fig. 2 in reference 20), that the spatial relationships of the appropriate side chains of the residues in the recognition helix and the contacting bases of the DNA at the interface are the same in both types of repressor-operator complexes. We thought that an altered binding specificity repressor mutant can be used suitably to test this assumption.
Determination of repression levels in vivo. To evaluate the strength of interaction between the wild-type and mutant repressors and the different operators, we used a single-copy reporter system (9). All measurements were done in R. meliloti at 28°C. Derivatives of pGSB1, containing the different operator/promoter units, were introduced by conjugation into R. meliloti 41 (E. coli with pCU101 [27] served as a helper). The site-specific integrative system of phage 16-3 (attP and the int gene), located in pGSB1 (9) and in its derivatives, resulted in the integration of the entire plasmid into the chromosome in each case. Another plasmid, either pSEM91 (25) (used as a no-repressor control) or one of its derivatives containing an allele of the c repressor gene, was introduced (E. coli harboring pRK2013 [10] served as a helper), and activities of the operator/promoter units were characterized.
-Galactosidase assays and calculations of promoter activities (expressed in Miller units) were carried out as described previously (15). Repression (R) values were calculated using the following equation (9, 20): R = 1 – (promoter activity in the cell when repressor was added from plasmid/promoter activity in the cell without repressor).
Role of residues of the recognition helix in repressor binding. Plasmids pGSB42 and pGSB62, which contain the sequences of an OL2 operator and an OR2 operator, respectively, between the –35 and –10 regions of a promoter, were constructed as previously described (9). Plasmid pSEM91 was used to express the 16-3 C repressor protein and its mutant derivatives. pPM232 (20) expressed the wild-type repressor (Cwt). Plasmid pPM238 was created. pPM238 is identical to pPM232, except that MscI and EagI restriction sites, in front of and following the recognition helix, respectively, were introduced into the coding sequence of the c repressor gene without altering the amino acid sequence of the C protein. Expression of functional repressor (CME) by pPM238 was verified by an in vivo experiment. Plasmid pPM238 was used to construct different mutant repressor genes by replacing the MscI-EagI region with synthetic oligonucleotides. To determine the importance of the residues in the recognition helix, the first seven residues were replaced with alanine, except at position 39, where the wild-type sequence originally contained an alanine residue. The resulting repressors were CQ37A, CQ38A, CI40A, CN41A, CN42A and CL43A. Binding of the wild-type (Cwt and CME) and mutant repressors to OL-type and OR-type operators was measured in vivo, and the calculated repression values are summarized in Table 1. Repression values of Cwt and CME are basically the same for each operator, indicating that the changes (silent at the protein level) which were introduced to create restriction sites in the c gene had no effect on the repression level. Replacement of the residues that were most likely to make contact with the operators (positions 37, 38, 41, and 42) by alanine resulted in significant loss in repressor binding activity irrespective of the type of operator used. Alanine at positions 40 and 43 also resulted in reduced binding, indicating the importance of the side chains of the original residues, which may be involved in maintaining the overall structure of the repressor in the DNA-protein complex.
Isolation of mutant repressors with altered binding specificities. Plasmid pGSB1 was used to construct different promoter/operator units in front of a promoterless lacZ gene. To isolate altered binding specificity mutant repressors, synthetic oligonucleotides were used to introduce all possible symmetrical changes into an OR2 operator at positions 1 and –1 (OR21–1), 2 and –2 (OR22–2), and 3 and –3 (OR23–3), as well as 4 and –4 (OR24–4) (Table 2). Due to our experimental setup, the R values directly correlate with the in vivo binding efficiencies of the repressors to the operators. Repression values for the binding of the Cwt repressor to mutant OR2-type operators are listed in Table 3. To correspond the results to the complex functional level of lysogenization, we can consider the phenotypic expression of two phage 16-3 mutants: (i) the immunity insensitivity of virulent mutant 16-3 Orc–1 , which carries the OR2c operator, and (ii) the heat inducibility of the 16-3 cti3 mutant, which carries a ts allele of the repressor. The phenotypes of the two mutants provide options to relate the R values to functional thresholds. Binding of Cwt to the OR2c operator (contains Orc–1 mutation in OR2) resulted in about 70% of the wild-type repression level (i.e., Cwt binding to the wild-type OR2 operator) (20). This result means that when R is 0.56, the level of repression would not be enough to complete its function toward lysogenization. The repression values of CL15M (encoded by the cti3 allele of the c gene) were 0.66 and 0.69 at 28°C for the OL2 and OR2 operators, respectively, and 0.31 at 37°C for both. This result means that when R is 0.66, the level of repression is enough to reach the stable lysogenic stage. In the light of these data, Cwt recognizes OR23–3/gc and OR24–4/cg mutant operators with near wild-type affinity, while in the rest of the cases, the strength of binding should be below the functional threshold needed for lysogen formation.
Repressor residues at positions 37 and 38 (the first and second positions in the recognition helix) were replaced with different amino acids in an attempt to identify mutations that restore the lost binding of the repressor to any of the mutant OR2-type operators. Repressors with different substitutions at position 37 were tested for binding to OR21–1 operators, and mutant repressors with changes at position 38 were tested for binding to OR22–2 and OR23–3 operators. Although we have not done a full-scale analysis (data not shown), one mutant repressor has been found to suppress an operator mutation. Repressor CQ37A binds well to OR21–1/cg (Table 4). Since binding of the CQ37A repressor to the OR2wt operator is very poor, like binding of Cwt to the OR21–1/cg operator (Table 4), we concluded that the CQ37A repressor had altered binding specificity for the OR-type operators.
Identification of a residue-base pair contact in a repressor-OR2wt operator complex. Sequence-specific binding of proteins to DNA depends on proper docking arrangements at the interface. It requires complementarities of the surfaces of the protein and DNA concerning size, shape, and charge. Altered binding specificity mutants have been isolated for several DNA binding proteins. They were used to identify contacts between a residue of the protein and a base pair in DNA even in the absence of the corresponding crystal structures (7, 29).
Cross-reactions between the 16-3 repressor and the 434 operator as well as between the 434 repressor and the OR-type 16-3 operator (4) suggested that the glutamine residue at position 37 of the 16-3 repressor (the first residue of the recognition helix) interacts with A at the first position of the operator half-site. Isolation of the CQ37A repressor as an altered specificity mutant confirms this prediction, since its binding to the OR21–1/cg mutant operator is as strong as Cwt binding to the OR2wt wild-type operator, while binding levels of Cwt to OR21–1/cg and CQ37A to OR2wt were reduced significantly. Substitution of the glutamine by an alanine at the first position in the recognition helix of the 434 repressor (Gln 28 Ala 28) resulted in altered binding specificity. The corresponding mutant 434 operator had an A · T T · A base pair change at the first position of both half-sites (28). It was rather surprising that altered binding specificity of the mutant 16-3 repressor is based on an A · T C · G change at the first position of the 16-3 operator half-sites, since the hydrophobic and short side chain of alanine would be unlikely to find a suitable atomic group to make specific contact with the C · G base pair. One possibility is that the new contact is made with an atomic group of the nucleotide other than in the base itself. This could explain a strong binding of CQ37A repressor to the OR21–1/cg mutant operator. However, CQ37A repressor has the ability to distinguish specifically between potential binding sites (R values of binding to OR21–1/gc and OR21–1/ta are 0.34 and 0.38, respectively).
Binding of the altered specificity mutant repressor to OL-type operators. In addition to identifying a residue-base pair contact, CQ37A was used to test the spatial arrangements when the 16-3 repressor binds to the structurally different OL-type and OR-type operators. The basic idea is that if the spatial arrangements are the same in both types of repressor-operator complexes, an altered binding specificity mutant isolated using OR-type operators should affect binding to OL-type operators in the same way, since the atomic groups required for the new contact are at the same place in both repressor-operator complexes. To assay binding of CQ37A repressor to OL-type operators, its binding to OL2wt and OL21–1/cg (5'-CCAATTGATTGG-3', where underlined type indicates changes relative to OL2wt ) was tested. Unexpectedly, the CQ37A repressor, contrary to its binding to the mutant OR21–1/cg operator, did not bind well to its parallel OL version, the OL21–1/cg mutant (Table 4).
Although it is not clear what the exact molecular basis of the suppressor phenotype is when CQ37A repressor binds to the OR21–1/cg mutant operator, the atomic groups ensuring contacts between the alanine residue and the C · G base pair in the OR-type operator-repressor complex could be out of reach to make contact in the OL-type complex. This means that the spatial relationship between the major grooves and the recognition helixes at the interface are different in the two kinds of complexes, implying that the 16-3 repressor may use different docking arrangements to recognize OL-type and OR-type operators.
In the light of our new results, we can extend our previously suggested "rotationally flexible protein homodimers" model by drawing conclusions concerning the binding domains. Now we can say that the spatial arrangements and the conformation of the binding domains of the repressor dimer are somewhat different when a 16-3 repressor forms a complex with an OL-type or an OR-type operator. However, this finding does not influence the basic characteristic of the model, which is the "rotational flexibility" of the dimers at the interface of the dimerization.
ACKNOWLEDGMENTS
We thank Kornelia Szorath Gal, Magdolna Toth Peli, and Csilla Santa Trk for excellent technical assistance and Dhruba Chattoraj for discussion and helpful comments on the manuscript.
REFERENCES
Blaha, B., S. Semsey, S. Ferenczi, Z. Csiszovszki, P. P. Papp, and L. Orosz. 2004. A proline tRNA(CGG) gene encompassing the attachment site of temperate phage 16-3 is functional and convertible to suppressor tRNA. Mol. Microbiol. 54:742-754.
Csiszovszki, Z., Z. Buzas, S. Semsey, T. Ponyi, P. P. Papp, and L. Orosz. 2003. immX immunity region of Rhizobium phage 16-3: two overlapping cistrons of repressor function. J. Bacteriol. 185:4382-4392.
Dallmann, G., F. Marincs, P. Papp, M. Gaszner, and L. Orosz. 1991. The isolated N-terminal DNA binding domain of the c repressor of bacteriophage 16-3 is functional in DNA binding in vivo and in vitro. Mol. Gen. Genet. 227:106-112.
Dallmann, G., P. Papp, and L. Orosz. 1987. Related repressor specificity of unrelated phages. Nature 330:398-401.
Dorgai, L., I. Papp, P. Papp, M. Kalman, and L. Orosz. 1993. Nucleotide sequences of the sites involved in the integration of phage 16-3 of Rhizobium meliloti 41. Nucleic Acids Res. 21:1671.
Dudas, B., and L. Orosz. 1980. Correlation between map position and phenotype of Cti mutants in the C cistron of Rhizobium meliloti phage 16-3. Genetics 96:321-329.
Ebright, R. H. 1991. Identification of amino acid-base pair contacts by genetic methods. Methods Enzymol. 208:620-640.
Elo, P., S. Semsey, A. Kereszt, T. Nagy, P. Papp, and L. Orosz. 1998. Integrative promoter cloning plasmid vectors for Rhizobium meliloti. FEMS Microbiol. Lett. 159:7-13.
Ferenczi, S., A. Ganyu, B. Blaha, S. Semsey, T. Nagy, Z. Csiszovszki, L. Orosz, and P. P. Papp. 2004. Integrative plasmid vector for constructing single-copy reporter systems to study gene regulation in Rhizobium meliloti and related species. Plasmid 52:57-62.
Figurski, D. H., and D. R. Helinski. 1979. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad. Sci. USA 76:1648-1652.
Ganyu, A., Z. Csiszovszki, T. Ponyi, A. Kern, Z. Buzas, L. Orosz, and P. P. Papp. 2005. Identification of cohesive ends and genes encoding the terminase of phage 16-3. J. Bacteriol. 187:2526-2531.
Hengen, P. N., S. L. Bartram, L. E. Stewart, and T. D. Schneider. 1997. Information analysis of Fis binding sites. Nucleic Acids Res. 25:4994-5002.
Ivanov, V. I., L. E. Minchenkova, B. K. Chernov, P. McPhie, S. Ryu, S. Garges, A. M. Barber, V. B. Zhurkin, and S. Adhya. 1995. CRP-DNA complexes: inducing the A-like form in the binding sites with an extended central spacer. J. Mol. Biol. 245:228-240.
Kallipolitis, B. H., and P. Valentin-Hansen. 2004. A role for the interdomain linker region of the Escherichia coli CytR regulator in repression complex formation. J. Mol. Biol. 342:1-7.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
Orosz, L. 1980. Methods for analysis of the C cistron of temperate phage 16-3 of Rhizobium meliloti. Genetics 94:265-276.
Orosz, L., A. Pay, and G. Dallmann. 1980. Heterozygosis of phage 16-3 of Rhizobium meliloti: moderate level of mismatch repair or gene conversion. Mol. Gen. Genet. 179:163-167.
Orosz, L., K. Rostas, and R. Hotchkiss. 1980. A comparison of two-point, three-point and deletion mapping in the C cistron of rhizobiophage 16-3, with an explanation for the recombination pattern. Genetics 94:249-263.
Papp, I., L. Dorgai, P. Papp, E. Jonas, F. Olasz, and L. Orosz. 1993. The bacterial attachment site of the temperate Rhizobium phage 16-3 overlaps the 3' end of a putative proline tRNA gene. Mol. Gen. Genet. 240:258-264.
Papp, P. P., T. Nagy, S. Ferenczi, P. Elo, Z. Csiszovszki, Z. Buzas, A. Patthy, and L. Orosz. 2002. Binding sites of different geometries for the 16-3 phage repressor. Proc. Natl. Acad. Sci. USA 99:8790-8795.
Pedersen, H., and P. Valentin-Hansen. 1997. Protein-induced fit: the CRP activator protein changes sequence-specific DNA recognition by the CytR repressor, a highly flexible LacI member. EMBO J. 16:2108-2118.
Putnoky, P., V. Deak, K. Bekasi, A. Palvolgyi, A. Maasz, Z. Palagyi, G. Hoffmann, and I. Kerepesi. 2004. H protein of bacteriophage 16-3 and RkpM protein of Sinorhizobium meliloti 41 are involved in phage adsorption. J. Bacteriol. 186:1591-1597.
Rice, P. A., S. Yang, K. Mizuuchi, and H. A. Nash. 1996. Crystal structure of an IHF-DNA complex: a protein-induced DNA U-turn. Cell 87:1295-1306.
Semsey, S., B. Blaha, K. Koles, L. Orosz, and P. P. Papp. 2002. Site-specific integrative elements of rhizobiophage 16-3 can integrate into proline tRNA (CGG) genes in different bacterial genera. J. Bacteriol. 184:177-182.
Semsey, S., I. Papp, Z. Buzas, A. Patthy, L. Orosz, and P. P. Papp. 1999. Identification of site-specific recombination genes int and xis of the Rhizobium temperate phage 16-3. J. Bacteriol. 181:4185-4192.
Shearwin, K. E., I. B. Dodd, and J. B. Egan. 2002. The helix-turn-helix motif of the coliphage 186 immunity repressor binds to two distinct recognition sequences. J. Biol. Chem. 277:3186-3194.
Thatte, V., D. E. Bradley, and V. N. Iyer. 1985. N conjugative transfer system of plasmid pCU1. J. Bacteriol. 163:1229-1236.
Wharton, R. P., and M. Ptashne. 1987. A new-specificity mutant of 434 repressor that defines an amino acid-base pair contact. Nature 326:888-891.
Youderian, P., A. Vershon, S. Bouvier, R. Sauer, and M. Susskind. 1983. Changing the DNA-binding specificity of a repressor. Cell 35:777-783.(Szilamer Ferenczi, Laszlo)
ABSTRACT
The C repressor protein of phage 16-3, which is required for establishing and maintaining lysogeny, recognizes structurally different operators which differ by 2 bp in the length of the spacer between the conserved palindromic sequences. A "rotationally flexible protein homodimers" model has been proposed in order to explain the conformational adaptivity of the 16-3 repressor. In this paper, we report on the isolation of a repressor mutant with altered binding specificity which was used to identify a residue-base pair contact and to monitor the spatial relationship of the recognition helix of C repressor to the contacting major groove of DNA within the two kinds of repressor-operator complexes. Our results indicate spatial differences at the interface which may reflect different docking arrangements in recognition of the structurally different operators by the 16-3 repressor.
TEXT
The regulation of gene expression occurs mostly at the level of transcription through sequence-specific DNA-protein interactions. In several cases, even a single change in the binding sequences results in loss of function due to loss or weakened binding of the protein. Alternatively, examples are also known of when the DNA binding protein has relaxed binding specificity, that is, when several changes at given positions still allow proper binding (12, 23). Only a few naturally existing systems are known where the protein has the ability to bind specifically to sequences with different lengths. The Escherichia coli cyclic AMP receptor protein (CRP) recognizes 16- and 18-bp-long binding sites, in which 6- and 8-bp central spacers, respectively, are bracketed by the recognition sequences. According to Adhya's "geometric homeostasis" method of resolution in CRP-DNA complexes, a conformational shift from B- to A-form DNA over one helical turn covering the longer spacer allows sequence-specific binding of CRP (13). The CytR repressor recognizes binding sites consisting of two octamer repeats, in direct or inverted orientation, separated by 2 bp. However, in the presence of cyclic AMP-CRP, CytR instead recognizes inverted repeats separated by 10 to 13 bp or direct repeats separated by 1 bp. It was shown that the bases for recognizing the structurally different sites were conformational changes within the CytR protein induced by protein-protein interactions between CRP and CytR (14, 21). The CI repressor of E. coli phage 186 was found to recognize two distinct DNA sequences, termed A-type and B-type sites. The A-type binding sites were different in length since half-sites were separated by either 4-bp or 5-bp spacers (26). The binding of 16-3 C repressor to its operators is another example for recognition of structurally different DNA sites by the same protein. In this system, the possibility that either conformational shifts from B- to A-form DNA within the binding sites or interactions of the repressor with another protein is involved in formation of 16-3 repressor-operator complexes was ruled out (20).
Genes, proteins, and chromosomal sites for several functions of the temperate phage 16-3 of Rhizobium meliloti 41 have been studied in detail (1, 2, 5, 8, 9, 11, 19, 22, 24, 25). The most thoroughly analyzed region of the 16-3 genome has been the immC regulatory region which encodes a lambda type CI repressor (16-3 C repressor protein), and the cognate c cistron is flanked by two operator regions, OL and OR (3, 4, 6, 16-18, 20). An interesting feature of the immC region is its partial cross-compatibility with the phage 434 immC region: although the overall sequence homology is not significant (below 15%), the helix-turn-helix motifs of the two repressors (55% identity) and the OR operators are highly similar (4). In both of the 16-3 OL and OR operator regions, two repressor binding sites were found: the OL-type operators were 12 bp long (5'-ACAA-4 bp-TTGT-3'), while the OR-type operators were 14 bp long (5'-ACAA-6 bp-TTGT-3'). To explain all previous in vivo and in vitro binding data, we proposed the "rotationally flexible protein homodimers" model. Our model hypothesizes alternative interactions for the dimerization domains as the basis for structural flexibility in recognizing the two types of binding sites (20). The model supposes, based on analyses of repressor binding to mutant and truncated versions of both types of operators (for a review, see Fig. 2 in reference 20), that the spatial relationships of the appropriate side chains of the residues in the recognition helix and the contacting bases of the DNA at the interface are the same in both types of repressor-operator complexes. We thought that an altered binding specificity repressor mutant can be used suitably to test this assumption.
Determination of repression levels in vivo. To evaluate the strength of interaction between the wild-type and mutant repressors and the different operators, we used a single-copy reporter system (9). All measurements were done in R. meliloti at 28°C. Derivatives of pGSB1, containing the different operator/promoter units, were introduced by conjugation into R. meliloti 41 (E. coli with pCU101 [27] served as a helper). The site-specific integrative system of phage 16-3 (attP and the int gene), located in pGSB1 (9) and in its derivatives, resulted in the integration of the entire plasmid into the chromosome in each case. Another plasmid, either pSEM91 (25) (used as a no-repressor control) or one of its derivatives containing an allele of the c repressor gene, was introduced (E. coli harboring pRK2013 [10] served as a helper), and activities of the operator/promoter units were characterized.
-Galactosidase assays and calculations of promoter activities (expressed in Miller units) were carried out as described previously (15). Repression (R) values were calculated using the following equation (9, 20): R = 1 – (promoter activity in the cell when repressor was added from plasmid/promoter activity in the cell without repressor).
Role of residues of the recognition helix in repressor binding. Plasmids pGSB42 and pGSB62, which contain the sequences of an OL2 operator and an OR2 operator, respectively, between the –35 and –10 regions of a promoter, were constructed as previously described (9). Plasmid pSEM91 was used to express the 16-3 C repressor protein and its mutant derivatives. pPM232 (20) expressed the wild-type repressor (Cwt). Plasmid pPM238 was created. pPM238 is identical to pPM232, except that MscI and EagI restriction sites, in front of and following the recognition helix, respectively, were introduced into the coding sequence of the c repressor gene without altering the amino acid sequence of the C protein. Expression of functional repressor (CME) by pPM238 was verified by an in vivo experiment. Plasmid pPM238 was used to construct different mutant repressor genes by replacing the MscI-EagI region with synthetic oligonucleotides. To determine the importance of the residues in the recognition helix, the first seven residues were replaced with alanine, except at position 39, where the wild-type sequence originally contained an alanine residue. The resulting repressors were CQ37A, CQ38A, CI40A, CN41A, CN42A and CL43A. Binding of the wild-type (Cwt and CME) and mutant repressors to OL-type and OR-type operators was measured in vivo, and the calculated repression values are summarized in Table 1. Repression values of Cwt and CME are basically the same for each operator, indicating that the changes (silent at the protein level) which were introduced to create restriction sites in the c gene had no effect on the repression level. Replacement of the residues that were most likely to make contact with the operators (positions 37, 38, 41, and 42) by alanine resulted in significant loss in repressor binding activity irrespective of the type of operator used. Alanine at positions 40 and 43 also resulted in reduced binding, indicating the importance of the side chains of the original residues, which may be involved in maintaining the overall structure of the repressor in the DNA-protein complex.
Isolation of mutant repressors with altered binding specificities. Plasmid pGSB1 was used to construct different promoter/operator units in front of a promoterless lacZ gene. To isolate altered binding specificity mutant repressors, synthetic oligonucleotides were used to introduce all possible symmetrical changes into an OR2 operator at positions 1 and –1 (OR21–1), 2 and –2 (OR22–2), and 3 and –3 (OR23–3), as well as 4 and –4 (OR24–4) (Table 2). Due to our experimental setup, the R values directly correlate with the in vivo binding efficiencies of the repressors to the operators. Repression values for the binding of the Cwt repressor to mutant OR2-type operators are listed in Table 3. To correspond the results to the complex functional level of lysogenization, we can consider the phenotypic expression of two phage 16-3 mutants: (i) the immunity insensitivity of virulent mutant 16-3 Orc–1 , which carries the OR2c operator, and (ii) the heat inducibility of the 16-3 cti3 mutant, which carries a ts allele of the repressor. The phenotypes of the two mutants provide options to relate the R values to functional thresholds. Binding of Cwt to the OR2c operator (contains Orc–1 mutation in OR2) resulted in about 70% of the wild-type repression level (i.e., Cwt binding to the wild-type OR2 operator) (20). This result means that when R is 0.56, the level of repression would not be enough to complete its function toward lysogenization. The repression values of CL15M (encoded by the cti3 allele of the c gene) were 0.66 and 0.69 at 28°C for the OL2 and OR2 operators, respectively, and 0.31 at 37°C for both. This result means that when R is 0.66, the level of repression is enough to reach the stable lysogenic stage. In the light of these data, Cwt recognizes OR23–3/gc and OR24–4/cg mutant operators with near wild-type affinity, while in the rest of the cases, the strength of binding should be below the functional threshold needed for lysogen formation.
Repressor residues at positions 37 and 38 (the first and second positions in the recognition helix) were replaced with different amino acids in an attempt to identify mutations that restore the lost binding of the repressor to any of the mutant OR2-type operators. Repressors with different substitutions at position 37 were tested for binding to OR21–1 operators, and mutant repressors with changes at position 38 were tested for binding to OR22–2 and OR23–3 operators. Although we have not done a full-scale analysis (data not shown), one mutant repressor has been found to suppress an operator mutation. Repressor CQ37A binds well to OR21–1/cg (Table 4). Since binding of the CQ37A repressor to the OR2wt operator is very poor, like binding of Cwt to the OR21–1/cg operator (Table 4), we concluded that the CQ37A repressor had altered binding specificity for the OR-type operators.
Identification of a residue-base pair contact in a repressor-OR2wt operator complex. Sequence-specific binding of proteins to DNA depends on proper docking arrangements at the interface. It requires complementarities of the surfaces of the protein and DNA concerning size, shape, and charge. Altered binding specificity mutants have been isolated for several DNA binding proteins. They were used to identify contacts between a residue of the protein and a base pair in DNA even in the absence of the corresponding crystal structures (7, 29).
Cross-reactions between the 16-3 repressor and the 434 operator as well as between the 434 repressor and the OR-type 16-3 operator (4) suggested that the glutamine residue at position 37 of the 16-3 repressor (the first residue of the recognition helix) interacts with A at the first position of the operator half-site. Isolation of the CQ37A repressor as an altered specificity mutant confirms this prediction, since its binding to the OR21–1/cg mutant operator is as strong as Cwt binding to the OR2wt wild-type operator, while binding levels of Cwt to OR21–1/cg and CQ37A to OR2wt were reduced significantly. Substitution of the glutamine by an alanine at the first position in the recognition helix of the 434 repressor (Gln 28 Ala 28) resulted in altered binding specificity. The corresponding mutant 434 operator had an A · T T · A base pair change at the first position of both half-sites (28). It was rather surprising that altered binding specificity of the mutant 16-3 repressor is based on an A · T C · G change at the first position of the 16-3 operator half-sites, since the hydrophobic and short side chain of alanine would be unlikely to find a suitable atomic group to make specific contact with the C · G base pair. One possibility is that the new contact is made with an atomic group of the nucleotide other than in the base itself. This could explain a strong binding of CQ37A repressor to the OR21–1/cg mutant operator. However, CQ37A repressor has the ability to distinguish specifically between potential binding sites (R values of binding to OR21–1/gc and OR21–1/ta are 0.34 and 0.38, respectively).
Binding of the altered specificity mutant repressor to OL-type operators. In addition to identifying a residue-base pair contact, CQ37A was used to test the spatial arrangements when the 16-3 repressor binds to the structurally different OL-type and OR-type operators. The basic idea is that if the spatial arrangements are the same in both types of repressor-operator complexes, an altered binding specificity mutant isolated using OR-type operators should affect binding to OL-type operators in the same way, since the atomic groups required for the new contact are at the same place in both repressor-operator complexes. To assay binding of CQ37A repressor to OL-type operators, its binding to OL2wt and OL21–1/cg (5'-CCAATTGATTGG-3', where underlined type indicates changes relative to OL2wt ) was tested. Unexpectedly, the CQ37A repressor, contrary to its binding to the mutant OR21–1/cg operator, did not bind well to its parallel OL version, the OL21–1/cg mutant (Table 4).
Although it is not clear what the exact molecular basis of the suppressor phenotype is when CQ37A repressor binds to the OR21–1/cg mutant operator, the atomic groups ensuring contacts between the alanine residue and the C · G base pair in the OR-type operator-repressor complex could be out of reach to make contact in the OL-type complex. This means that the spatial relationship between the major grooves and the recognition helixes at the interface are different in the two kinds of complexes, implying that the 16-3 repressor may use different docking arrangements to recognize OL-type and OR-type operators.
In the light of our new results, we can extend our previously suggested "rotationally flexible protein homodimers" model by drawing conclusions concerning the binding domains. Now we can say that the spatial arrangements and the conformation of the binding domains of the repressor dimer are somewhat different when a 16-3 repressor forms a complex with an OL-type or an OR-type operator. However, this finding does not influence the basic characteristic of the model, which is the "rotational flexibility" of the dimers at the interface of the dimerization.
ACKNOWLEDGMENTS
We thank Kornelia Szorath Gal, Magdolna Toth Peli, and Csilla Santa Trk for excellent technical assistance and Dhruba Chattoraj for discussion and helpful comments on the manuscript.
REFERENCES
Blaha, B., S. Semsey, S. Ferenczi, Z. Csiszovszki, P. P. Papp, and L. Orosz. 2004. A proline tRNA(CGG) gene encompassing the attachment site of temperate phage 16-3 is functional and convertible to suppressor tRNA. Mol. Microbiol. 54:742-754.
Csiszovszki, Z., Z. Buzas, S. Semsey, T. Ponyi, P. P. Papp, and L. Orosz. 2003. immX immunity region of Rhizobium phage 16-3: two overlapping cistrons of repressor function. J. Bacteriol. 185:4382-4392.
Dallmann, G., F. Marincs, P. Papp, M. Gaszner, and L. Orosz. 1991. The isolated N-terminal DNA binding domain of the c repressor of bacteriophage 16-3 is functional in DNA binding in vivo and in vitro. Mol. Gen. Genet. 227:106-112.
Dallmann, G., P. Papp, and L. Orosz. 1987. Related repressor specificity of unrelated phages. Nature 330:398-401.
Dorgai, L., I. Papp, P. Papp, M. Kalman, and L. Orosz. 1993. Nucleotide sequences of the sites involved in the integration of phage 16-3 of Rhizobium meliloti 41. Nucleic Acids Res. 21:1671.
Dudas, B., and L. Orosz. 1980. Correlation between map position and phenotype of Cti mutants in the C cistron of Rhizobium meliloti phage 16-3. Genetics 96:321-329.
Ebright, R. H. 1991. Identification of amino acid-base pair contacts by genetic methods. Methods Enzymol. 208:620-640.
Elo, P., S. Semsey, A. Kereszt, T. Nagy, P. Papp, and L. Orosz. 1998. Integrative promoter cloning plasmid vectors for Rhizobium meliloti. FEMS Microbiol. Lett. 159:7-13.
Ferenczi, S., A. Ganyu, B. Blaha, S. Semsey, T. Nagy, Z. Csiszovszki, L. Orosz, and P. P. Papp. 2004. Integrative plasmid vector for constructing single-copy reporter systems to study gene regulation in Rhizobium meliloti and related species. Plasmid 52:57-62.
Figurski, D. H., and D. R. Helinski. 1979. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad. Sci. USA 76:1648-1652.
Ganyu, A., Z. Csiszovszki, T. Ponyi, A. Kern, Z. Buzas, L. Orosz, and P. P. Papp. 2005. Identification of cohesive ends and genes encoding the terminase of phage 16-3. J. Bacteriol. 187:2526-2531.
Hengen, P. N., S. L. Bartram, L. E. Stewart, and T. D. Schneider. 1997. Information analysis of Fis binding sites. Nucleic Acids Res. 25:4994-5002.
Ivanov, V. I., L. E. Minchenkova, B. K. Chernov, P. McPhie, S. Ryu, S. Garges, A. M. Barber, V. B. Zhurkin, and S. Adhya. 1995. CRP-DNA complexes: inducing the A-like form in the binding sites with an extended central spacer. J. Mol. Biol. 245:228-240.
Kallipolitis, B. H., and P. Valentin-Hansen. 2004. A role for the interdomain linker region of the Escherichia coli CytR regulator in repression complex formation. J. Mol. Biol. 342:1-7.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
Orosz, L. 1980. Methods for analysis of the C cistron of temperate phage 16-3 of Rhizobium meliloti. Genetics 94:265-276.
Orosz, L., A. Pay, and G. Dallmann. 1980. Heterozygosis of phage 16-3 of Rhizobium meliloti: moderate level of mismatch repair or gene conversion. Mol. Gen. Genet. 179:163-167.
Orosz, L., K. Rostas, and R. Hotchkiss. 1980. A comparison of two-point, three-point and deletion mapping in the C cistron of rhizobiophage 16-3, with an explanation for the recombination pattern. Genetics 94:249-263.
Papp, I., L. Dorgai, P. Papp, E. Jonas, F. Olasz, and L. Orosz. 1993. The bacterial attachment site of the temperate Rhizobium phage 16-3 overlaps the 3' end of a putative proline tRNA gene. Mol. Gen. Genet. 240:258-264.
Papp, P. P., T. Nagy, S. Ferenczi, P. Elo, Z. Csiszovszki, Z. Buzas, A. Patthy, and L. Orosz. 2002. Binding sites of different geometries for the 16-3 phage repressor. Proc. Natl. Acad. Sci. USA 99:8790-8795.
Pedersen, H., and P. Valentin-Hansen. 1997. Protein-induced fit: the CRP activator protein changes sequence-specific DNA recognition by the CytR repressor, a highly flexible LacI member. EMBO J. 16:2108-2118.
Putnoky, P., V. Deak, K. Bekasi, A. Palvolgyi, A. Maasz, Z. Palagyi, G. Hoffmann, and I. Kerepesi. 2004. H protein of bacteriophage 16-3 and RkpM protein of Sinorhizobium meliloti 41 are involved in phage adsorption. J. Bacteriol. 186:1591-1597.
Rice, P. A., S. Yang, K. Mizuuchi, and H. A. Nash. 1996. Crystal structure of an IHF-DNA complex: a protein-induced DNA U-turn. Cell 87:1295-1306.
Semsey, S., B. Blaha, K. Koles, L. Orosz, and P. P. Papp. 2002. Site-specific integrative elements of rhizobiophage 16-3 can integrate into proline tRNA (CGG) genes in different bacterial genera. J. Bacteriol. 184:177-182.
Semsey, S., I. Papp, Z. Buzas, A. Patthy, L. Orosz, and P. P. Papp. 1999. Identification of site-specific recombination genes int and xis of the Rhizobium temperate phage 16-3. J. Bacteriol. 181:4185-4192.
Shearwin, K. E., I. B. Dodd, and J. B. Egan. 2002. The helix-turn-helix motif of the coliphage 186 immunity repressor binds to two distinct recognition sequences. J. Biol. Chem. 277:3186-3194.
Thatte, V., D. E. Bradley, and V. N. Iyer. 1985. N conjugative transfer system of plasmid pCU1. J. Bacteriol. 163:1229-1236.
Wharton, R. P., and M. Ptashne. 1987. A new-specificity mutant of 434 repressor that defines an amino acid-base pair contact. Nature 326:888-891.
Youderian, P., A. Vershon, S. Bouvier, R. Sauer, and M. Susskind. 1983. Changing the DNA-binding specificity of a repressor. Cell 35:777-783.(Szilamer Ferenczi, Laszlo)