The Lysis-Lysogeny Decision of Bacteriophage 933W: a 933W Repressor-Mediated Long-Distance Loop Has No Role in Regulating 933W P_RM Activity

Abstract

Our data show that unlike bacteriophage λ, repressor bound at O_L of bacteriophage 933W has no role in regulation of 933W repressor occupancy of 933W O_R3 or the transcriptional activity of 933W P_RM. This finding suggests that a cooperative long-range loop between repressor tetramers bound at O_R and O_L does not form in bacteriophage 933W. Nonetheless, 933W forms lysogens, and 933W prophage display a threshold response to UV induction similar to related lambdoid phages. Hence, the long-range loop thought to be important for constructing a threshold response in lambdoid bacteriophages is dispensable. The lack of a loop requires bacteriophage 933W to use a novel strategy in regulating its lysis-lysogeny decisions. As part of this strategy, the difference between the repressor concentrations needed to bind O_R2 and activate 933W P_RM transcription or bind O_R3 and repress transcription from P_RM is <2-fold. Consequently, P_RM is never fully activated, reaching only ∼25% of the maximum possible level of repressor-dependent activation before repressor-mediated repression occurs. The 933W repressor also apparently does not bind cooperatively to the individual sites in O_R and O_L. This scenario explains how, in the absence of DNA looping, bacteriophage 933W displays a threshold effect in response to DNA damage and suggests how 933W lysogens behave as “hair triggers” with spontaneous induction occurring to a greater extent in this phage than in other lambdoid phages.

INTRODUCTION

Studies of how the lysis-lysogeny decisions of lambdoid bacteriophages are regulated have illustrated the importance of short- and long-range cooperative DNA binding by proteins in controlling a complex gene regulatory network. In all lambdoid phages, the repressor protein directs the establishment and maintenance of the lysogenic state by simultaneously repressing transcription of the genes needed for lytic phage growth and activating transcription of a gene needed for lysogen formation (30). Based primarily on studies of bacteriophage λ, two different cooperative repressor-DNA binding events are thought to be required for regulation of lambdoid phage lysogen development. The first involves formation of a repressor tetramer between two repressor dimers, one bound to each of the adjacent O_R1 and O_R2 sites. A similar tetramer is also formed at the O_L1 and O_L2 sites. The “side-by-side” cooperative binding by a repressor tetramer is prerequisite to forming a stable λ lysogens (2, 20, 30).

Additional cooperative interactions between two tetramers, each bound to a pair of adjacent operators that are separated by >2.5 kb, occur in bacteriophages λ and P22 (13–15, 31). In λ, a repressor octamer-mediated O_L-cI₈-O_R complex forms with the intervening DNA looping out (3, 13, 15). This long-range cooperative interaction helps modulate the prophage's compensatory response to low doses of DNA damage (3) and thereby regulates lysogen stability. In all well-studied lambdoid phages, the amount of phage produced is not linearly related to the amount of DNA damage until a particular threshold amount of DNA damage is absorbed by the lysogen. This threshold is due to the compensatory effect of removing the repressor from a partially occupied O_R3, an occupancy that is facilitated by the long-range loop mediated by a repressor octamer bound to O_L1, O_L2, O_R1, and O_R2 (3). The discovery of the long-range interaction between repressor tetramers bound at O_R and O_L in phage λ seemingly solved two long-standing puzzles: (i) the function of cI binding at O_L3 and (ii) given that the λ repressor's affinity for O_R3 is too weak to significantly repress its own promoter, how does autogenous negative control work?

Despite the apparent importance of long-range cooperativity in stabilizing lambdoid phage lysogens, several recent observations indicate that this interaction may not be required. We and others have reported that unlike all other well-studied bacteriophages, bacteriophage 933W contains only two, not three, repressor binding sites in O_L (16, 23, 36) (Fig. 1). Bacteriophage 933W is derived from the Shiga toxin-producing Escherichia coli (STEC) strain EDL933. The gene encoding Shiga toxin is present on bacteriophage 933W and is part of an operon directly controlled by the bacteriophage P_R′ promoter. Shiga toxin is not produced when the bacteriophage is in its lysogenic state, but toxin production increases substantially upon phage lysogen induction. As a result of its downstream position in the lytic cascade, the activity of P_R′ and thereby Shiga toxin production by 933W are ultimately controlled by factors that influence 933W repressor DNA binding. Hence, in addition to providing information on bacteriophage gene control mechanisms, these studies will inform studies aimed at controlling Shiga toxin production in infected individuals.

Fig. 1.

Structure of the 933W immunity region. The positions of O_R1, O_R2, and O_R3, of O_L1 and O_L2, and of the repressor (cI) and cro genes (Cro) are indicated by boxes; the transcription start sites of P_L, P_RM, and P_R are indicated by bent arrows.

The absence of an O_L3 site suggests that 933W repressor binding to O_R3 may not be assisted by formation of a λ-like O_L-cI₈-O_R looped complex. Also, we reported earlier that in intact 933W O_R, the 933W repressor does not bind its O_R1 and O_R2 sites at an identical concentration (23), suggesting that the 933W repressor is incapable of binding cooperatively to these two adjacent sites. Cooperativity is not essential to the construction a of lysis-lysogeny switch in a lambdoid phage. For example, Babić and Little found that cis-acting mutations can allow a λ bacteriophage encoding a repressor that is incapable of forming DNA-bound tetramers (and, presumably, octamers) to form stable lysogens (2). Taken together, these results imply that the lysis-lysogeny regulatory circuitry of bacteriophage 933W does not involve cooperative binding of repressor tetramers or octamers.

Here we characterize the DNA binding and gene regulatory strategies that allow 933W bacteriophage to function as an outwardly “normal” temperate phage. Our findings confirm that the basic transcriptional behavior of 933W's lysis-lysogeny circuitry resembles that found in other well-characterized lambdoid phages. However, these data show that 933W repressors bound at O_L regulate neither O_R3 occupancy nor P_RM activity. Since the activity of 933W P_RM, the maintenance of lysogeny, is crucial, the stability of 933W lysogens is not regulated by a long-range DNA loop. Instead, our data show that O_R3 occupancy and thus P_RM activity depend on a small difference in the intrinsic affinities of the 933W repressor for O_R2 and O_R3. It is clear that 933W bacteriophage has evolved alternative strategies to regulate the lysogen stability decision. Remarkably, some of these strategies overlap those of mutant λ phage previously identified by Babić and Little (2).

MATERIALS AND METHODS

Bacterial strains, bacteriophages, and DNA.

All plasmids were propagated in either Escherichia coli strain K802 (40) or strain JM101 (27). 933W repressor was purified from the E. coli strain BL21(DE3)::pLysS (Novagen, Madison, WI) bearing a plasmid that directs its overexpression (p933WR) as described previously (23). Construction of the plasmid containing the 933W O_R operator was described previously (23). Templates for DNase I footprinting and transcription were generated by PCR from these plasmids as described below.

The λimm⁴³⁴ bacteriophage used in lysogen induction experiments was from our collection. λimm^933W (36) was the generous gift of D. Friedman (University of Michigan). Lysogens were formed in MG1655.

Deletion of the O_L region of λimm^933W was effected by using the lambda Red/Gam homologous recombination system expressed on helper plasmid pKD46 (12). For this process, O_L, as well as part of the N gene found on p933WO_L (23), was replaced with the spectinomycin resistance gene. Deletion of the 933W O_L region within the lysogen was confirmed via PCR. As a consequence of the deletion of part of the N gene, the λimm^933WΔOL lysogen is not inducible with mitomycin.

Primers for site-directed mutagenesis, quantitative reverse transcription-PCR (RT-PCR), and the primer for detecting 933W repressor occupancy of O_R (see below) were obtained from Integrated DNA Technologies (Coralville, IA).

In vitro transcription assays.

Template DNA was PCR amplified from the appropriate plasmid by using the standard M13 forward and reverse primers and gel purified. Increasing amounts of 933W repressor were incubated with 8 nM template DNA for 10 min at 25°C in transcription buffer (0.2 M Tris-HCl [pH 7.5], 75 mM KCl, 50 mM MgCl₂, 0.05% Triton X-100, 1 mM dithiothreitol [DTT]). E. coli RNA polymerase (2 mg/ml) was added and allowed to form open complexes at 37°C for 10 min. Single-round transcription was then initiated by adding a nucleoside triphosphate mix (2.5 mM each for ATP, CTP, and GTP, 0.5 mM UTP and [α-³²P]UTP, and 1 mM heparin), and the reaction mixture was incubated at 37°C for 15 min. Transcription was stopped with formamide dye and heated to 95°C for 4 min, and products were separated on a 6% acrylamide–7 M urea gel in Tris-borate-EDTA (TBE; 89 mM Tris, 89 mM borate, 1 mM EDTA). Radiolabeled products were visualized and quantified by phosphorimaging.

UV induction of phage lysogens.

Measurements of the UV induction threshold response were performed using methods similar to those described previously (25). Briefly, lysogens were grown in LB to mid-log phase and then chilled, centrifuged, and washed with TMG buffer (10 mM Tris-HCl [pH 8.0], 10 mM MgSO₄, 10 μg/ml gelatin). Resuspended cells were irradiated at a distance of 20 cm (∼1 μW/cm²) by using a 254-nm UV light (Mineralight UVG-54). At various exposure times, aliquots of the cultures were plated on LB to ascertain viable bacterial cell counts. The UV-exposed culture was then diluted 1:10 in 5 ml of LB and incubated at 37°C shaking for 2 h. The induced culture was treated with CHCl₃, and debris was removed by centrifugation at 18,000 × g for 1 min. Phage titers were measured by plating as described previously (1).

EMSA.

Electrophoresis mobility shift assays (EMSA) were performed as essentially described (23). DNA containing the individual naturally occurring 933W repressor binding sites was obtained by annealing 60-base oligonucleotides containing the 15-bp 933W binding site sequence and two additional naturally occurring bases on either end of the identified site. The DNA fragments were radioactively labeled at their 5′ ends by incubating the DNA with [γ-³²P]ATP (6,000 Ci/mmol; Perkin-Elmer, Boston, MA) in the presence of T4 polynucleotide kinase (Epicentre, Inc., Madison, WI). O_R1 DNA was also radiolabeled using [α-³²P]dATP (3,000 Ci/mmol; Perkin-Elmer, Boston, MA) in place of dATP in a PCR mixture with an O_R1-containing plasmid as a template to increase the radioactive signal and detect DNA to concentrations below 0.1 nM. Labeled DNA was incubated with the specified concentrations of 933W repressor protein in binding buffer (10 mM Tris [pH 8.0], 50 mM NaCl, 1 mM MgCl₂, 10% glycerol, 100 μg/ml bovine serum albumin [BSA], 1 mM isopropyl-β-d-thiogalactopyranoside [IPTG], 1 mM DTT) for 10 min at 25°C. The protein-DNA complexes were resolved on 5% polyacrylamide gels at 25°C. The electrophoresis buffer was 1× TBE. The amounts of protein-DNA complexes present on the dried gels were quantified using a Storm imager (GE Lifesciences, Piscataway, NJ).

Values of the dissociation constant (K_D) were determined by nonlinear squares fitting of the EMSA data using Prism 4.0 software (GraphPad Software Inc.). Each dissociation constant was determined from at least eight replicate measurements. Two-tailed t tests were employed to determine the significance of observed differences in dissociation constants of the various 933W repressor-DNA complexes.

DNase I footprinting.

Template DNA was PCR amplified from the desired plasmid using 5′-end-labeled standard M13 forward and unlabeled M13 reverse primers and gel purified. Following phenol-chloroform-isoamyl alcohol extraction and ethanol precipitation, this DNA (≤0.05 nM) was incubated without or with 933W repressor in binding buffer (10 mM Tris [pH 8.0], 50 mM NaCl, 1 mM MgCl₂,100 μg/ml BSA, 1 mM DTT) for 10 min at 25°C prior to addition of sufficient DNase I to generate, on average, one cleavage per DNA molecule with 5 min of additional incubation. The cleavage reactions were terminated by precipitation with ethanol and sec-butanol dehydration, and the DNA was dissolved in 90% formamide solution containing tracking dyes. The products along with chemical sequencing reactions (Maxam Gilbert 1980) derived from the same templates were resolved on 6% acrylamide gels containing 7 M urea in 1× TBE. Cleavage products were visualized using a Storm imager (GE Lifesciences, Piscataway, NJ).

Values of the dissociation constant (K_D) were determined by nonlinear squares fitting of the measured intensities of repressor-protected bands for each of the three individual binding sites. Each dissociation constant was determined from measurements made in three independent experiments. The reported dissociation constants varied by ≤20%.

In vivo dimethylsulfate (DMS) footprinting.

Both MG1655::λ^imm933W and MG16655::λimm^933WΔOL lysogens were transformed with pGP1-2 (35), which encodes T₇ RNA polymerase and whose synthesis is under the control of a temperature-sensitive lambda repressor. Each of these two strains were subsequently separately transformed with p933WR or pET17b (EMB Biosciences).

Cultures of the two plasmid-containing lysogens were grown for 16 h at 37°C in LB with 100 μg/ml ampicillin and 50 μg/ml kanamycin. DMS was added to the cultures at a final concentration of 7.9 μM and incubated with shaking for 5 min at 37°C. The reactions were stopped by addition of an equal volume of ice-cold LB and immediately placed on ice. Genomic DNA from 1.5 ml of treated cells was isolated using lysozyme-phenol-RNase A essentially as described previously (32) and resuspended in PCR buffer (0.2 mM deoxynucleoside triphosphates, 1.5 mM MgCl₂, 50 mM KCl, 0.001% gelatin). Resuspended genomic DNA (20 μg) was used as template for each primer extension reaction.

A primer complementary to a sequence just outside the 933W O_R region, in the cro gene, 5′-CCTTTAATCGGCTCATCAAGATTTTGCAT-3′, was 5′ end labeled with T₄ polynucleotide kinase (New England BioLabs, Beverly, MA) and [γ-³²P]ATP (6,000 Ci/mmol; Perkin-Elmer, Boston, MA). Labeled primer was added to the dissolved genomic DNA to a final concentration of 1.4 μM, and the extension reaction was initiated with the addition Taq polymerase followed by thermocycling. After 35 rounds of extension, the products of extension reactions (one-half the reaction volume) were loaded onto 6% acrylamide–7 M urea gel (89 mM Tris, 89 mM borate, 1 mM EDTA) after addition of formamide dye and boiling of samples for 4 min. A sequencing ladder was also loaded next to the primer extension reaction mixtures (Sequenase DNA sequencing kit; USB, Cleveland, OH). Gels were dried and imaged with a Storm imager.

Quantitative RT-PCR.

Cultures of the plasmid-containing MG1655::λ^imm933W and MG166::λimm^933WΔOL lysogens were grown to early log phase at 30°C in LB with 100 μg/ml ampicillin and 50 μg/ml kanamycin. Cultures were shifted to 42°C and grown until late log phase. RNA was extracted from 0.5 ml of cells by using the QuickExtract RNA extraction kit (Epicentre, Madison, WI). RNA was further purified by acid phenol extraction followed by chloroform-isoamyl alcohol extraction and precipitation. RNA was resuspended in 22 μl of DNase I buffer and RiboGuard, and residual genomic DNA was removed by treatment with RNase-free DNase I (Epicentre, Madison, WI) for 1 h at 37°C. cDNA synthesis reaction mixtures containing purified RNA, Affinityscript buffer, forward primer, and Affinityscript reverse transcriptase/RNase block enzyme mix (Agilent Technologies, Cedarville, TX) were incubated at 25°C for 5 min to allow primer annealing, 45°C for 45 min for cDNA synthesis, and finally at 95°C for 5 min to heat kill the reverse transcriptase. Quantitation of RNA was performed by real-time PCR. DNA products were detected using Sybr Green I (Invitrogen/Molecular Probes Inc., Carlsbad, CA) in a MiniOpticon real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA). Primers used for amplification of P_RM transcript cDNA were complementary to a portion of the 933W open reading frames 23 and 24. (933W P_RM trans FWD, 5′-GCCGCTCTAACACCTAGTATTCTC-3′, and 933W P_RM trans RVS, 5′-TAAGGCCGCCTGAACATATC-3′). As an internal control for RNA preparation, separate real-time PCR analyses were performed on the cDNA preparations of the E. coli UmuC gene transcript from the same RNA preparations as the P_RM transcripts. These real-time PCR assays used primers complementary to a portion of the UmuC gene (FWD UmuC, 5′-GATTTATGGGGTAAACCGGTGG-3′; RVS UmuC, 5′-CAGTCAGATCGAGACAATTACG-3′). Standard curves for the real-time PCR analyses were made using known template amounts of a plasmid containing the region of interest.

Numerical simulation of transcription data.

Occupancies of O_R1, O_R2, and O_R3 at various 933W repressor concentrations were calculated using the dissociation constants given in Fig. 5, below, and the following equation: fr_{Ox bound} = −{([O_{x total}] + [933W R_total] + K_D) − {([O_{x total}] + [933W R_total] + K_D)² − 4[O_{x total}][933W R_total]}^1/2}/(2[O_{x total}]), where O_x is either O_R1, O_R2, or O_R3 and R_total is the total repressor concentration. For the simulations, the DNA concentration was fixed at the value used in the transcription experiments. Relative activities of P_RM transcription were calculated using the experimentally verified assumptions that 933W repressor bound at O_R2 activates P_RM and that 933W repressor bound at O_R3 represses the activity of this promoter. Maximal promoter activity was set to 60, corresponding to the observed maximal activity of P_RM relative to P_R, as measured on a template bearing a mutation in O_R3 that prevented repressor binding.

Fig. 5.

Sequences and intrinsic affinities of 933W for separate naturally occurring 933W operators. (A) Binding of 933W repressor to individual naturally occurring binding sites. Various concentrations of 933W repressor were mixed with radiolabeled binding site-containing DNA, and protein-DNA complex formation was detected as described in Materials and Methods. Points represent averages of ≥8 replicate experiments. Lines represent best nonlinear least-squares fits to the data based on a hyperbolic equation. (B) The sequences and dissociation constants (K_D [lswb]± standard deviation]) of 933W repressor binding to O_R1, O_R2, O_R3, O_L1, and O_L2. Dissociation constants were derived from data shown in panel A and determined as described in Materials and Methods. Standard deviations were calculated from the averages of ≥8 replicate experiments.

Statistical methods.

Where employed, tests for statistical significance of differences between paired data sets were performed using two-tailed t tests.

RESULTS

Examination of the threshold response to UV light.

An O_L3-bound λ repressor cooperatively assists binding of repressor to O_R3 (15). This interaction requires the formation of an octameric λ repressor-DNA complex containing two tetramers of λ repressor, one bound at O_R1 and O_R2 and another at O_L1 and O_L2 (15). Deletion of O_L eliminates the increased P_RM activity in response to low doses of UV (no derepression effect), suggesting the DNA damage “threshold response” depends on formation of the O_L-cI₈-O_R loop in phage λ (3). The observation that bacteriophage 933W lacks an O_L3 site (16, 23, 36) (Fig. 1) indicates that unlike other lambdoid phages (3), 933W either does not display a threshold response or that the threshold response of this phage is constructed differently from that of other lambdoid phages. To distinguish between these alternatives, we compared the ability of UV light to induce lysogens of two different hybrid bacteriophages: (i) λimm^933W, a bacteriophage that contains the entire immunity region of 933W and whose lysis-lysogeny decision is controlled by 933W repressor-DNA interactions, and (ii) λimm⁴³⁴, a bacteriophage that bears the immunity region of bacteriophage 434 and whose repressor is capable of cooperatively binding its DNA sites. Other than the immunity regions (i.e., DNA between the N gene [immediately upstream of O_L] and the end of the cro gene [immediately downstream of O_R]), the sequences of these two phages are identical. Hence, any differences in the ability of UV light to induce these two phages can be directly attributed to differences in gene regulation mechanisms encoded within the immunity region.

At low UV dosages, neither the λimm⁴³⁴ nor λimm^933W lysogen produces a significant number of bacteriophage (Fig. 2). In contrast, once the UV exposure reached a threshold dose of ∼3.5 J/m², the amount of phage produced by both of these strains rose sharply and increased linearly with UV dosage (Fig. 2). These findings show that the 933W immunity region codes for a threshold response to DNA damage. The UV threshold set point value for 933W is nearly identical to that observed with bacteriophage λ lysogens (3, 25), but since 933W lacks an O_L3 site, its response must be enabled by a mechanism different from phage λ and its related bacteriophages.

Fig. 2.

Demonstration of the λimm^933W threshold response to UV induction, similar to that of λimm⁴³⁴. Lysogens containing either 933W (λimm^933W) or 434 (λimm⁴³⁴) immunity regions were subjected to either no UV light or increasing doses of UV light. Lysogens were then diluted and incubated with aeration at 37°C for 2 h to allow phage induction. Phage titers were determined by plating on MG1655. Cell numbers were determined by plating aliquots of cells on LB plates prior to the 2-h incubation. Phage numbers are expressed as the fraction of the maximal amount of phage-forming units (PFU) per cell-forming units (CFU). Error bars represent standard deviations calculated from the averages of 5 replicate experiments.

Repressor occupancy of O_R3 is independent of the presence of O_L.

Despite the absence of an O_L3 site (16, 23), 933W repressor bound at O_L could have a role in regulating repressor occupancy of O_R3, the key element of the threshold response (Fig. 2). We wished to directly test whether the O_L region of 933W has any role in regulating repressor occupancy of O_R3. To do this we first used DMS footprinting to examine 933W repressor occupancy in vivo in both a λimm^933W lysogen (which bears an intact 933W O_R and O_L) and a λimm^933WΔOL lysogen, in which the O_L region of λimm^933W has been deleted. These strains were transformed with either a control plasmid or one that directs the synthesis of 933W repressor (see Materials and Methods). These strains were exposed to DMS, and the DMS modification/protection pattern was detected by primer extension and using a radioactively labeled primer.

To assist with identification of the 933W repressor-protected bands in the samples isolated from the cells, genomic DNA was isolated from untreated cells and treated in vitro with DMS in the absence or presence of purified 933W repressor. As expected, 933W repressor occupancy of O_R1, O_R2, and O_R3 on isolated DNA in vitro was detectable by a repressor-dependent decrease in the intensity of several DMS-reactive guanine residues in these sites (Fig. 3A, lanes 7 and 8). For example, the DMS reactivity at positions 5′G, 7′G, and 2G in O_R1 and 5′G in O_R2 in wild-type DNA decreased in the presence of 933W repressor. The ability of the repressor to protect these bases from modification in vitro was unaffected by deletion of the O_L region (Fig. 3A, compare lanes 7 and 8 with lanes 8 and 9).

Fig. 3.

Deletion of 933W O_L does not affect repressor occupancy at O_R3. (A) Representative DMS footprinting gel. DMS methylation of 933W genomic DNA template was used to detect repressor protection of O_R sites. DMS was either not added (lanes 1 and 2) to lysogenic cells (lanes 3 to 6) or isolated genomic DNA from lysogens was added (lanes 7 to 10) and allowed to react at 37°C for 5 min. Isolated genomic DNA in all lanes was subjected to primer extension using a radiolabeled DNA primer complementary to regions just outside of the 933W O_R and Taq polymerase (see Materials and Methods). Locations of guanine bases in O_R3, O_R2, and O_R1 sensitive to methylation by DMS are indicated. For the in vivo DMS treatment (lanes 3 to 6), DMS was added to overnight cultures of λimm^933W (wt) or λimm^933WΔOL (ΔO_L) grown at 37°C. The level of 933W repressor was determined by either endogenous lysogen levels (lanes 3 and 5) or endogenous levels plus additional repressor expressed from p933WR (lanes 4 and 6). No DMS was used in lanes 1 and 2. For the in vitro DMS treatment (lanes 7 to 10), DMS was added to purified genomic DNA isolated from λimm^933W (lanes 7 and 8) or λimm^933WΔOL (lanes 9 and 10) lysogens. Lanes 7 and 9 had no 933W repressor present upon DMS addition. In lanes 8 and 10, saturating amounts of purified 933W repressor were added to genomic DNA prior to the DMS methylation reaction. (B) Quantification of in vitro DMS methylation intensities of guanines in O_R3 (see panel A, lanes 7 to 10). (C) Quantification of in vivo DMS methylation intensities of guanines in O_R3 (see panel A, lanes 3 to 6). In panels B and C, intensities were normalized to the reactivity at position 1′. Error bars in panels B and C represent standard deviations derived from 4 replicate experiments.

Our major focus with respect to the effect of O_L deletion on 933W repressor occupancy of its binding sites in O_R concerned O_R3. Repressor binding to this site was therefore considered in detail. In the absence of any 933W repressor, the DMS reactivities of the four guanine residues on the top strand in O_R3 (positions 2, 7′, 5′, and 1′) were unaffected by the presence or absence of the O_L region (Fig. 3A, compare lanes 7 and 9). Also, for both these DNAs, added 933W repressor protected the guanines at positions 2 and 5′ of O_R3 from DMS modification (Fig. 3A, compare lanes 8 and 10). The extent of repressor-mediated protection of each of the guanines at positions 2 and 5′ was essentially identical in both the λimm^933W and λimm^933WΔOL DNAs (Fig. 3B). These observations establish the protection DMS modification pattern of the repressor-bound O_R3 site and provide a basis for understanding the in vivo DMS modification patterns.

Similar to the pattern of DMS reactivity seen in vitro, once inside cells DMS modifies guanines at positions 2, 7′, 5′, and 1′ of O_R3. However, compared to the results obtained in vitro in the absence of repressor, the relative DMS reactivities of the guanine residues at positions 2 and 5′ in O_R3 are lower than seen at positions 7′ and 1′ (Fig. 3A, compare lane 3 to lanes 7 and 8 and lane 5 to lanes 9 and 10; see also B and C). In the absence of repressor in vitro, the intensities of these bands are essentially equal, indicating near-identical accessibility to DMS (Fig. 3A, lanes 7 and 9, and B). Quantification of the results in lane 3 showed that the DMS reactivities of the guanines at positions 2 and 5′ were 2-fold lower than that at position 1′ (Fig. 3C). This finding suggests that at the level of 933W repressor found in the λimm^933W lysogen, O_R3 is partially occupied by the 933W repressor. Consistent with this suggestion, the reactivities of the bases at positions 2 and 5′ were reduced by an additional 2-fold when cells were transformed with a plasmid that directs synthesis of an additional amount of 933W repressor (Fig. 3A, compare lanes 3 and 4 and lanes 5 and 6). Moreover, in the presence of the excess repressor, the levels of DMS reactivity of position 2 and 5′ guanines were reduced to the same level as that seen when this DNA reacted with DMS in the presence of saturating levels of 933W repressor in vitro (see Fig. 5B and C, below).

Importantly, we found that both the pattern and degree of protection of the bases in O_R3 in the λimm^933WΔOL lysogen are identical to that seen in the λ^imm933W lysogen (Fig. 3A, compare lanes 3 and 5, and B and C). This result indicated that the absence or presence of the O_L region has no effect on the ability of the 933W repressor to occupy O_R3 in vivo at the levels of repressor found in a lysogen. We also found that the degree to which excess (plasmid-derived) repressor enhanced the protection of the guanines at positions 2 and 5′ inside λimm^933WΔOL lysogens was identical to that found with the λimm^933W lysogen (Fig. 3C). Taken together, these results show that 933W repressor binding to sites in O_L does not influence repressor occupancy of O_R3. These observations are consistent with the suggestion that a repressor-mediated DNA loop between O_L and O_R does not form in bacteriophage 933W. The DMS modification patterns in the region around O_R1 and O_R2 obtained in vitro differ slightly from those found in vivo. We suspect these differences do not stem from differences in repressor occupancy under these two conditions, but rather arise from the partial occupancy of this region by RNA polymerase bound at P_R and/or P_RM. Regardless, inspection of the results shown in Fig. 3A, lanes 4 and 6, shows that the DMS modification pattern of these sites was unaffected by the absence or presence of O_L, indicating deletion of O_L does not affect 933W repressor binding to these sites. This finding is consistent with our observation that deletion of O_L does not affect repressor occupancy of O_R3.

P_RM transcriptional activity is not dependent on O_L.

To further examine the potential role of O_L-bound 933W repressor on O_R3 occupancy, we determined if the presence or absence of the O_L region affected the transcriptional activity of P_RM in vivo. For these experiments, we compared the amount of P_RM transcript produced in both the λimm^933W and λimm^933WΔOL lysogens in the absence and presence of additional 933W repressor.

Quantification of P_RM transcripts found in λimm^933W and λimm^933WΔOL showed that the differences in the steady-state amounts of P_RM-derived mRNA in these two lysogens were statistically insignificant (P = 0.34) (Fig. 4). Production of additional 933W repressor inside each of these lysogens repressed P_RM transcript levels by an identical amount (Fig. 4). Since repressor occupancy of O_R2 and O_R3 positively and negatively regulates P_RM activity, respectively (see Fig. 7, below), these results showed that O_L has no effect on repressor occupancy of O_R2 and O_R3. These findings are completely consistent with the results of our in vivo DMS footprinting study (Fig. 3), which showed that deletion of the repressor binding sites in O_L did not have any effect on 933W repressor occupancy of O_R3. These observations do differ dramatically from what was found with bacteriophage λ. In that case, deletion of λ repressor binding sites in λO_L increased transcription from λP_RM as a consequence of the decreased occupancy of λO_R3 by 2- to 3-fold (3, 13). Thus, our results show that occupancy of the binding sites in 933W O_R by the 933W repressor is by interactions between repressors bound at O_R and O_L.

Fig. 4.

Deletion of 933W O_L does not affect in vivo levels of P_RM transcripts. Quantitative real-time PCR was used to determine P_RM cDNA levels produced from reverse transcription of total RNA extracted from λimm^933W and λimm^933WΔOL lysogens producing wild-type levels of 933W repressor (containing pET17b and pGP1-2) or excess 933W repressor (+++-containing p933WR and pGP1-2). Amounts of P_RM transcripts were normalized to the amount of UmuC transcripts, determined in parallel reverse transcription reactions (see Materials and Methods). Differences between the amounts of transcript without or with excess repressor were significant (P < 0.0001). The amount of transcripts from O_L⁺ and ΔO_L lysogens were not significantly different (P = 0.34), regardless of the absence or presence of the repressor-producing plasmid. Data are derived from eight replicate experiments.

Fig. 7.

O_R2 and O_R3 are required for regulation of P_RM transcription by repressor. (A) Representative transcription gels. 933W DNA templates containing wild-type (wt) O_R or O_R regions bearing mutations in either O_R2 (O_R2⁻) or O_R3 (O_R3⁻) were transcribed in vitro in the absence of repressor (lane 1) and at repressor concentrations increased in 2-fold steps (lanes 2 to 7), starting with 4.5 nM protein. Positions of transcripts initiated from P_R and P_RM are indicated. The 933W repressor was incubated with DNA template at 25°C for 10 min, followed by addition of E. coli RNA polymerase. The reaction mixture was transferred to 37°C for 10 min before the transcription reaction was initiated by the addition of nucleotides and heparin. (B) Graphical representation of the amount of P_RM transcript synthesized as a function of 933W repressor concentration from the template bearing wild-type O_R or templates bearing a mutation in O_R3 or O_R2. The transcript amounts are quantified as a percentage of maximal P_R transcription as a function of 933W repressor concentration. Error bars represent standard deviations calculated from the averages of at least three replicate experiments. (C) Numerical simulation of transcription data using O_R occupancy data calculated from dissociation constants given in Fig. 5 (see also Materials and Methods). The lines represent simulated data. Points are the measured P_RM activity, as described for panel B.

Affinities of the 933W repressor for naturally occurring 933W O_R binding sites in separate sites and intact O_R.

In the absence of O_L-mediated regulation of O_R3 occupancy, how then might the UV threshold response of bacteriophage 933W (Fig. 2) be determined? To begin to answer this question, we measured the affinity of the 933W repressor for its naturally occurring sites in O_R and O_L, both on the individual sites and in the context of the intact operator. The intrinsic affinity of the 933W repressor for the individual sites was measured in an electrophoresis mobility shift assay. Figure 5 shows that the 933W repressor displayed a hierarchy of affinities for the sites in O_R, as it bound with highest affinity to O_R1 and with ∼100- and 15-fold-lower affinities to O_R2 and O_R3, respectively. The differences in affinities of the 933W repressor for O_R1, O_R2, and O_R3 were significant (P ≤ 0.005), and the relative affinities of the 933W repressor for its sites in O_R were qualitatively and quantitatively different than what was seen with other well-studied lambdoid bacteriophage repressors, e.g., λ (O_R1 of 1, O_R2 of 75, and O_R3 of 30), 434 (O_R1 of 1, O_R2 of 14, and O_R3of 5.5), and P22 (O_R1 of 1, O_R2 of 14, and O_R3 of 6) (5, 6, 19, 21, 33, 34, 37).

Our previously reported DNase I footprinting results also suggested that the 933W repressor does not bind cooperatively to O_R1 and O_R2 when these sites are present within intact O_R (23). However, the conditions under which those experiments were performed are not directly comparable to those used in the EMSA determinations reported here. Also, those experiments were performed at concentrations of DNA that, based on the data in Fig. 5, were either within or near the stoichiometric range for repressor binding to O_R1 and O_R2, respectively. Therefore, we repeated the footprinting experiments under conditions identical to those used in the EMSA determinations (i.e., low DNA concentration and absence of competitor DNA) to provide a comparable measure of the affinities of 933W repressor for its individual sites in intact O_R. Unlike the case with other lambdoid bacteriophage repressors, the 933W repressor binds the three sites in intact O_R at distinctly different concentrations (Fig. 6). The dissociation constants deduced from a series of footprinting experiments were as follows: O_R1, 0.22 nM; O_R2, 3.0 nM; O_R3, 5.8 nM (± ∼20%). Hence, the affinities of the 933W repressor for O_R1, O_R2, and O_R3 in intact O_R are essentially identical to the affinities of the 933W repressor for the individual, separated sites shown in Fig. 5. The similarities between 933W repressor affinities for the separated sites and the sites present in intact O_R led us to modify our earlier suggestion that the 933W repressor may bind with weak cooperativity to adjacent sites within the intact operators (23). Instead, our data indicate that the 933W repressor does not cooperatively bind to adjacent sites in O_R.

Fig. 6.

DNase I footprinting of complexes between the 933W repressor and 933W O_R. Shown is a representative phosphorimage of a representative gel. DNA templates containing 933W O_R radioactively labeled were partially digested with DNase I in the presence of increasing amounts of the 933W repressor. Lane 1 shows the DNase I cleavage pattern of the DNA in the absence of added repressor. In lanes 2 to 13, repressor concentrations were increased in 1.5-fold steps starting at 0.17 nM protein. The arrows identify positions of protected bands used in measuring site occupancies of O_R1, O_R2, and O_R3.

Our measurements revealed that the 933W repressor binds to O_L1 and O_L2 with similar affinities (Fig. 5). We showed previously that in intact 933W O_L, the 933W repressor binds its respective O_L1 and O_L2 with nearly equal affinity (23). Similar measurements under comparable conditions to our EMSA studies confirmed this finding (data not shown). Thus, the equal affinities of 933W repressor for O_L1 and O_L2 seen in intact O_L are apparently not a result of cooperative binding by this protein.

Regardless of whether 933W repressor binds DNA cooperatively, our data confirm the earlier finding that its affinities for O_R2 and O_R3 differ by 1.5- to 2-fold. This observation provides a potential explanation as to how the UV threshold in bacteriophage 933W is generated. Instead of being regulated by a long-range loop, 933W repressor occupancy of the sites in O_R is instead governed solely by the intrinsic affinities of the repressor for the individual sites in O_R.

Positive and negative regulation of P_RM transcription by the 933W repressor.

To further explore how the small differences in the 933W repressor's intrinsic affinities for sites in O_R might contribute to the UV threshold, we determined how the bacteriophage 933W repressor regulates transcription from P_RM. In vitro transcription analysis of a 933W wild-type O_R template, containing the 933W O_R region and its P_R and P_RM promoters, showed that in the absence of 933W repressor, no P_RM transcript was observed, while P_R was robustly transcribed. As the 933W repressor concentration was increased to up to ∼30 nM, an increasing amount of P_RM transcript was synthesized and transcription of P_R was repressed (Fig. 7). Transcription from P_RM was inhibited as the concentration of 933W repressor was increased from 30 nM to 120 nM (Fig. 7). Consistent with previous observations (23), these findings indicate that, similar to other bacteriophage repressors, the 933W repressor both activates and represses P_RM transcription and functions as a repressor of P_R transcriptional activity.

In the other well-studied bacteriophages, the repressor-O_R2 complex is needed to activate transcription of P_RM. To investigate the role of the 933 repressor-O_R2 complex in stimulating 933W P_RM transcription, we created a transcription template bearing two base substitutions in O_R2 at positions 5′ and 4′ (G_5′T_4′ → T_5′C_4′) (see Fig. 1 for wild-type sequences). Control experiments established that these sequence changes prevented 933W repressor binding to O_R2, without detectably altering binding of this protein to O_R1 or O_R3 or the concentration of repressor needed to repress P_R transcription (data not shown). As opposed to the case with wild-type template, 933W repressor was incapable of activating P_RM transcription from the O_R2⁻ template in vitro (Fig. 7A). This finding showed that, similar to other lambdoid phages, a 933W repressor bound to O_R2 is needed to activate transcription from P_RM.

In all other bacteriophages, the repressor-O_R3 complex inhibits transcription initiated at P_RM. We investigated the mechanism by which the 933W repressor negatively regulates P_RM, and thus its own synthesis, by examining the effects of mutations that prevent 933W repressor binding to O_R3 on P_RM transcription in vitro. Mutations that blocked 933W repressor binding to O_R3 prevented the 933W repressor from repressing transcription from P_RM (Fig. 7A). These results show that 933W repressor binding to O_R3 mediates repression of transcription from P_RM.

Our results show that the effects of 933W repressor binding to O_R2 and O_R3 on P_RM transcriptional activity are similar to those seen in other bacteriophages. However, inspection of the results shown in Fig. 7 reveals that the activity of 933W P_RM, in response to varied 933W repressor concentrations, differs considerably from that seen in other lambdoid bacteriophages. In 933W, the maximal repressor-stimulated P_RM transcription from the O_R3⁻ template was nearly 4-fold higher than that observed from the wild-type template. In contrast, in bacteriophage λ, the maximal amount of repressor-stimulated P_RM transcript obtained on wild-type and O_R3⁻ templates is the same (17, 28). Similar results were obtained with bacteriophage 434 (10, 42). Based on the observation that the 933W repressor affinity for O_R2 is less than 2-fold higher than its affinity for O_R3 (Fig. 5), the results in Fig. 7A indicate that at 933W repressor concentrations needed to completely occupy O_R2, the 933W repressor partially occupies O_R3. This conclusion was verified by accurately simulating these transcription results, using operator occupancy data derived from the dissociation constants shown above (Fig. 7C).

On the wild-type O_R template, complete repression of P_RM transcription occurred with the addition only 4-fold more 933W repressor than is needed maximally for this promoter (Fig. 7). The difference in 933W repressor concentration needed to activate versus repress P_RM transcription was much lower than that seen in other lambdoid phages. In bacteriophage λ, >20-fold more λ repressor is required to repress P_RM transcription from an O_R template than is needed to maximally stimulate this promoter's activity (17, 29). Similar results were obtained with bacteriophage 434 (10, 43).

DISCUSSION

Our data clearly show that in vivo, deletion of the 933W O_L region does not influence repressor occupancy of O_R3 or the activity of the P_RM promoter. Hence, bacteriophage 933W regulates P_RM activity, and therefore repressor levels, by a mechanism that apparently does not involve long-range cooperative interactions between 933W repressors. Had a long-range DNA loop formed in 933W and had such a loop been required for modulating 933W repressor occupancy of sites in O_R, we would also have anticipated that the affinities of 933W repressor for the sites in O_L should be equal to or greater than its affinities for O_R1 and O_R2. However, the intrinsic affinities of 933W repressor for O_L1 and O_L2 were ≥7-fold lower than its affinities for O_R1 (Fig. 5). It is possible that 933W repressor binding at O_L1 and O_L2 stabilizes 933W repressor binding at O_R. However, this view is inconsistent with the finding that deletion of O_L apparently does not impact 933W repressor occupancy of any sites in O_R (Fig. 3).

How then does bacteriophage 933W regulate repressor levels so that it can form a stable and yet inducible lysogen? In general, 933W repressor-mediated regulation of P_RM activity in bacteriophage 933W employs a strategy similar to that of other lambdoid phages. Like those phages, 933W repressor bound to O_R2 is responsible for activation of P_RM, and repressor bound to O_R3 is required for repression. However, unlike other lambdoid phages, in bacteriophage 933W negative regulation of P_RM transcriptional activity by the repressor is governed solely by differences in the 933W repressor's intrinsic affinities O_R2 and O_R3. The ∼2-fold difference between the affinities of 933W repressor for 933W O_R2 and 933W O_R3 (as opposed to the 15- to 30-fold differences in affinities of the λ repressor for λO_R2 and λO_R3 in intact λ O_R) results in tight repressor autoregulation of P_RM activity. The small differences in relative affinities of the 933W repressor for 933W O_R2 and 933W O_R3 are sufficient to allow repressor occupancy at O_R3 in vivo for negative regulation of P_RM in 933W. That is, the closely matched affinities of O_R2 and O_R3 allow for the counterbalancing of the autogenous positive- and negative-control activities of the 933W repressor-DNA complexes formed at 933W O_R2 and 933W O_R3, respectively, on the transcriptional activity of P_RM.

Interestingly, repression of 933W P_RM transcription in vitro, which is mediated by the 933W repressor binding to O_R3, occurs at 933W repressor concentrations where full occupancy of O_R2, and thus maximal activation of P_RM transcriptional activity, has not yet been reached. Furthermore, complete repression of P_RM occurs at levels of repressor that are only 4-fold higher than is needed for activation, further arguing for the partial occupancy of O_R3 at concentrations where O_R2 is not yet completely filled. The lack of repressor-mediated interactions between O_L and O_R suggests that the in vitro transcription findings accurately represent the situation in vivo.

The narrow range of 933W repressor concentrations needed to activate and repress 933W P_RM transcription is strikingly different from that seen in vitro with other lambdoid phages. In bacteriophages λ and 434, repression of P_RM (via occupancy of O_R3) does not occur until a repressor concentration at which P_RM is fully activated (or O_R2 is maximally filled) is reached. This is because in these bacteriophages, at least 20-fold more repressor is required to fully repress P_RM than that needed to maximally stimulate transcription. Therefore, in both λ and 434 bacteriophages, the activity of maximal repressor-stimulated P_RM activity on wild-type O_R and O_R3 mutant templates, where repressor cannot repress P_RM transcription, is essentially identical (10, 29).

It is important to note that to enable repression of P_RM in vivo, the weak intrinsic affinity of λ repressor for O_R3 must be overcome by the long-range DNA loop mediated by repressor bound at O_L and O_R sites (15). As a consequence, repression of P_RM in bacteriophage λ in vivo requires only 4-fold more λ repressor than is needed for activation, as opposed to the ≥20-fold needed in vitro (15, 29). Therefore, the range of λ repressor concentrations needed to activate and repress λ P_RM in vivo transcription matches that seen for 933W. However, as a consequence of the cooperative binding of λ repressor to O_R1 and O_R2, repressor-stimulated transcriptional activity of P_RM in vivo reaches >70% of the maximum amount seen in an O_L3⁻ or O_R3⁻ λ bacteriophage. Hence, in λ, repression of P_RM only occurs at repressor concentrations higher than where P_RM is essentially fully activated. This observation contrasts with our findings for intact 933W O_R, where P_RM reached only ∼25% of the maximum activation possible.

The narrow window of repressor concentrations between the amount needed to activate versus repress 933W P_RM transcription explains why 933W bacteriophage displays a “threshold response” to DNA damage. The observation of a threshold response has been attributed to the response of P_RM activity to various levels of repressor. As discussed above, in both 933W and λ lysogens, the transcriptional activity of P_RM is partially repressed due to partial occupancy of O_R3 by repressor. At low UV doses, as a consequence of RecA-mediated repressor degradation, O_R3 occupancy decreases, leading to increased P_RM activity and a consequent increase in repressor levels. At high UV doses, the rate of RecA-mediated repressor degradation exceeds the rate of new repressor synthesis, and the repressor concentration falls below the level needed to occupy O_R1 and O_R2, leading to lysogen induction. Hence, the key feature of the threshold response is partial occupancy of O_R3 by the repressor. In bacteriophage λ, O_R3 binding is facilitated by a repressor-mediated loop between O_L and O_R that allows an O_L3-bound repressor to closely approach O_R3 and to help a repressor bind that site via cooperative interactions that are thought to mimic those used by λ repressor bound at adjacent sites (3). In bacteriophage 933W, O_R3 occupancy is facilitated by the closely matched affinities of 933W repressor for the binding site that activates (O_R2) and the one that represses (O_R3) P_RM transcription.

This proposed strategy for generating a UV threshold response in bacteriophage 933W not only eliminates the need for long-distance cooperative interactions between two DNA-bound repressor tetramers but also the subsequent role for cooperative binding between spatially adjacent O_R3- and O_L3-bound repressor dimers in the looped complex. Since the interaction between repressors bound at O_R3 and O_L3 is anticipated to use the “side-by-side” cooperativity interface, this conclusion is consistent with the indication that the 933W repressor is incapable of cooperatively binding to adjacent sites on O_R (Fig. 5 and 6). Together with the fact that bacteriophage 933W forms stable lysogens, this suggestion indicates that the ∼9-fold difference in relative affinity of the 933W repressor for O_R2 and O_R1 is sufficient to enable efficient occupancy of O_R2 and positive regulation of P_RM in the absence of side-by-side cooperative repressor interactions. In contrast, the λ phage repressor's intrinsic affinity for O_R1 is 30 times higher than its affinity for O_R2 (21, 22, 34). It is known that efficient occupancy of O_R2 in wild-type λ phage, and therefore stable λ lysogen formation, requires cooperative repressor binding to O_R1 and O_R2 (2, 11).

The different levels of maximal P_RM activity in vivo in λ and 933W bacteriophages and the apparent inability of the 933W repressor to cooperatively bind adjacent sites may help explain the observation that 933W lysogens behave as a “hair trigger,” i.e., the frequency of DNA damage-independent (spontaneous) induction is much higher in 933W than in other related bacteriophages (26). The increased induction frequency in these stx-encoding phages has been attributed to the requirement for a lower concentration of active RecA necessary for induction (26). The critical level of RecA needed for induction of 933W phage could be determined by an increased sensitivity of the 933W repressor to RecA, a decrease in the strength of the 933W repressor's binding interactions with its operators, and/or a decreased total amount of 933W repressor present in the lysogen.

Our findings show that the absolute affinities of the 933W repressor for its DNA sites are not dramatically different than the affinities of other lambdoid phage repressors for their cognate operators (Fig. 5). We have not directly assessed the sensitivity of 933W to RecA-mediated autocleavage. Results of in vitro transcription assays indicated that the maximal activity of 933W repressor-stimulated P_RM is 60% of the value of P_R (Fig. 7), whereas similar experiments have indicated that the maximal activity of λ repressor-stimulated P_RM is ∼3-fold lower, i.e., ∼20% of λ P_R activity (18, 39, 41). Our data also indicate that the in vivo activity of 933W P_RM is only 25% of its maximal level. Together these observations suggest that the amount of repressor is lower in 933W lysogens than in λ lysogens. Therefore, the increased sensitivity of 933W lysogens to spontaneous induction could be due, at least in part, to a smaller amount of repressor in the 933W lysogen than in λ lysogens.

Babić and Little (2) were able to isolate cis-acting mutants that allow a λ bacteriophage encoding a repressor that is incapable of cooperatively forming DNA-bound tetramers to form stable lysogens. The mutant phages identified by Babić and Little compensated for the repressor cooperativity defect, in part, by increasing the relative affinity of O_R2 for the λ repressor. We showed here that the affinity of the 933W repressor for its O_R2 is 4- to 6-fold higher, relative to 933W O_R1, than seen in wild-type λ. Therefore, the compensatory O_R2 mutation found in the mutant λ phages mimics a key native feature of the 933W bacteriophage lysis-lysogeny circuitry.

The increased affinity of repressor for O_R2 in the mutant λ phage creates a scenario where, akin to bacteriophage 933W, the λ repressor binds O_R3 and the mutant O_R2 with similar affinities. Consequently, the smaller difference between O_R2 and O_R3 affinities would mean that at λ repressor concentrations where full occupancy of O_R2 has not yet been reached, O_R3 would be partially occupied, and thus P_RM partially repressed. With P_RM partially repressed, the mutant λ phage requires a more active P_RM promoter to maintain required levels of λ repressor to form stable lysogens. This observation is consistent with the inference that the maximal stimulated activity of 933W P_RM is greater than that of wild-type λP_RM.

The 933W repressor is apparently unique among the lambdoid bacteriophage repressors in its native inability to bind cooperatively to either adjacent or widely separated sites. The C-terminal domain (CTD) of lambdoid phage repressors is responsible for mediating any cooperative binding (30). We hypothesize that the 933W repressor is incapable of binding DNA cooperatively due to critical sequence differences between its CTD and the CTDs of other lambdoid repressors. A sequence comparison of phage repressor CTDs revealed only a weak (10 to 11%) similarity between the sequence of the 933W CTD and those of λ, 434, and P22 phages (Fig. 8), In contrast, the sequences of the CTDs of λ, 434, and P22 were highly similar (36 to 60%) to each other. Strikingly, the sequence alignment showed many of the λ repressor residues found in the tetramer interface and those tested for mediating cooperative interactions, such as E188, K192, S198, G199, Q209, Y210, and M212, differ from those in analogous positions in 933W in terms of both identity and character (4, 7–9, 38). In contrast, most of the residues in these regions of the 434 and P22 CTDs, which are known to bind DNA cooperatively, are similar to those found in the λ repressor. Hence, we suggest that these sequence differences underlie the inability of the 933W repressor to cooperatively bind DNA.

Fig. 8.

Alignments of the CTDs of the 933W repressor with three other lambdoid phage repressors known to cooperatively bind DNA. Multiple sequence alignments were performed using ClustalW (24). Shaded boxes were used to demonstrate the output based on residue matches (black) and functional similarities (gray). This program was written by Kay Hofmann and Michael D. Baron. Numbers indicate the residue within the protein that defines the beginning of the CTD. Black dots indicate residues in the λ repressor that contribute to cooperative interactions between repressor dimers.