THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1984 by The American Society of Biological Chemists, Ine.

Vol. 259, No. 2, Issue of January 25, pp. 805-814,1984 Printed in U.S.A.

Characterization of the Escherichia coli SSB- 113 Mutant Single- stranded DNA-binding Protein CLONING OF THE GENE, DNA AND PROTEIN SEQUENCE ANALYSIS, HIGH PRESSURE LIQUID CHROMATOGRAPHY PEPTIDE MAPPING, AND DNA-BINDING STUDIES*

(Received for publication, August 1, 1983)

John W. Chase$§, James J. L’ItalienTIl/, Janet B. Murphy$, Eleanor K. Spicery, and Kenneth R. WilliamslI** From the $Department of Molecular Biology, Albert Einstein College of Medicine, Bronx, New York 10461 and the TDepartment of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut 06510

The ssb-113 (formerly lexC113) gene encoding a mutant single-stranded DNA binding protein (SSB) has been cloned into plasmid pSClOl resulting in 5- to 10- fold more mutant protein than strains carrying only one (chromosomal) copy of the gene. Analysis of tryptic and chymotryptic peptides of the mutant protein by high pressure liquid chromatography and solid phase protein sequencing has shown that the ssb-113 muta- tion results in the substitution of serine for proline at residue 176 of SSB. This change could only occur in one step by a C + T transition in the DNA sequence. Physicochemical studies of the homogeneous mutant protein have shown that it binds as well as wild type SSB to single-stranded DNA and that it is a slightly better helix-destabilizing protein than wild type SSB as measured by its ability to lower the thermal melting transition of poly[d(A-T)]. In vivo studies of ssb-113 strains carrying the cloned ssb-113 gene in pSClOl have shown that overproduction of the mutant protein does not complement the temperature-sensitive condi- tional lethality caused by the ssb-113 mutation when present in single gene copy in contrast to effects re- cently observed in ssb-1 strains overproducing the ssb- 1 encoded protein (Chase, J. W., Murphy, J. B., Whit- tier, R. F., Lorensen, E., and Sninsky, J. J. (1983) J. Mol. Biol. 164, 193-211).

Also noted in this report are two corrections to the DNA sequence of wild type SSB, one of which places glycine (codon GGC) at residue 133 rather than serine as previously reported (Sancar, A., Williams, K. R., Chase, J. W., and Rupp, W. D. (1981) Proc. Natl. Acad. Sci. U. S. A. 78,4274-4278). The second correction to the DNA sequence is in the serine 39 codon, previously reported to be TCA and now correctly shown to be TCC.

* This investigation was supported in part by Public Health Service Grant GM12607 to William H. Konigsberg. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

f Supported by Public Health Service Grants GM11301, GM23451, CA13330. An established investigator of the American Heart Asso- ciation, grant 78-129.

11 Supported by Dr. Frank F. Richards (United States Public Health Service Grants AI 09614-14 and AI-15530-03 and by American Heart Association Grant 81-661). Present address, Molecular Genetics, Inc., 10320 Bren Road East, Minnetonka, MN 55343.

** Recipient of National Science Foundation Grant PCM-81- 04118.

It is now well established that Escherichia coli ssDNA’ binding protein (SSB’) serves a necessary role in DNA repli- cation, repair, and recombination (Meyer et al., 1979; Whittier and Chase, 1983; Lieberman and Witkin, 1983; Golub and Low, 1983; for reviews see Coleman and Oakley, 1979; Ko- walcykowski et al., 1981; Williams and Konigsberg, 1981). This has become apparent through in. vivo analysis of ssb- strains. At present, two clearly defined mutations in SSB have been identified and analyzed-ssb-1 and ssb-113 (for- merly lexCll3). Both mutations confer temperature-sensitive conditional lethality (Vales et al., 1980) and effect the ampli- fied synthesis of RecA protein following DNA damage (Baluch et al., 1980) and the UV-induced activation of RecA protease (Lieberman and Witkin, 1983). In addition, recent measure- ments of recombination in ssb- strains (Golub and Low, 1983) have detected more significant defects than first reported (Glassberg et al., 1979). Deficiencies in SSB therefore result in pleiotropic effects and, in general, the two ssb mutations so far studied cause similar deficiencies. However, subtle differences exist between ssb-1 and ssb-113 mutant strains particularly with regard to the effect of temperature. All defects studied to date in ssb-I mutant strains are tempera- ture-dependent; that is, ssb-1 skrains appear essentially nor- mal at 30 “C and exhibit all effects of SSB deficiency at 43 “C while ssb-113 mutant strains exhibit these defects even at the temperature permissive for growth. In addition, overproduc- tion of the ssb-l-encoded protein reverses most of the defects caused by the ssb-1 mutation when present in single gene copy (Chase et al., 198313) whereas our preliminary character- ization reported here suggests that increased quar.tities of the ssb-113-encodedprotein does not complement the effects of a chromosomal ssb-113 mutation. On the basis of these and other arguments, we have previously suggested that the mech- anism of action of SSB may be more complex than earlier thought (Chase et al., 198313; Whittier and Chase, 1983). Although there is as yet no direct evidence that SSB functions other than through binding to ssDNA, we have speculated that the multiple defects exhibited by ssb- strains in general and the differences we have observed between ssb-1 and ssb-

The abbreviations used are: ssDNA, single-stranded DNA; SSB, the Escherichia coli single-stranded DNA-binding protein, which has also been referred to as a DNA helix-destabilizing protein (EcoHD- protein I); poly[d(A-T)], alternating copolymer of deoxyadenylate and deoxythymidylate; kb, kilobases; UV, ultraviolet light; HPLC, high pressure liquid chromatography; PTH, phenylthiohydantoin.

We will use SSB to refer to the wild type protein purified from ssb+ strains and will also use the appropriate allele number as a suffix when referring to a particular mutant protein (e.g. SSB-113 will refer to ssb-113 encoded ssDNA-binding protein).

805

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant ssDNA Binding Protein

113 strains could more easily be explained if SSB also directly interacts with other proteins. We have therefore undertaken in vitro analysis of mutant SSB proteins in order to better understand a t a biochemical level how SSB participates in DNA replication, repair, and recombination. We report here the cloning of the ssb-113 gene and the amino acid sequence and preliminary physicochemical studies of the ssb-113 en- coded protein.

MATERIALS AND METHODS

Bacterial Strains and Plasmids The E. coli K12 strains used were KLC788 (F-, melA7, rha-8,

thyA36, amp-50, deoC2), KLC792 (as KLC788 but ssb-lid), KLC818 (as KLC788 but uorA6). HMS91 (W3llOrha-,thy-), and KLC221 (xonA, his-, gnd- derivative of KS439b (Konrad and Lehman, 1974)). The E. coli B strains used were WP2 (trpE65, lonAl, sulAI, malB15) and PAM33 (ssb-113, malB') (Johnson, 1977). The plasmids used were pSClOl (Cohen and Chang, 1977), pPM24 (Meacock and Cohen, 1980), and pKAC4 (Chase et al., 1983b).

Media Cells were routinely grown in T-broth (1% tryptone, 0.5% NaCI).

Efficiency of plating was determined on L-broth agar plates without NaCl (1% tryptone, 0.5% yeast extract, 1.5% agar). Mitomycin C (Sigma) was added to 0.5 pg/ml where indicated. Tetracycline (Sigma) was added to 5 pg/ml.

Cloning of the ssb-I13 Mutant Gene Chromosomal DNA was isolated from strain KLC792 as previously

described (Chase et aL, 198313) and dissolved in 10 mM Tris-HCI buffer (pH 8.1), 0.1 mM EDTA (TE buffer). This DNA (100 pg) was digested in two independent reactions with 90 and 225 units of EcoRI (New England Biolabs) in 200-4 reactions for 4 h at 37 "C. Under these conditions, the DNA incubated with 90 units of EcoRI, appeared to be partially digested. The EcoRI-digested DNA was phenol ex- tracted, ethanol precipitated, and dissolved in TE buffer. Purification of plasmid DNA (pSC101) was performed as described by Kupersz- toch and Helinski (1973). Plasmid DNA (50 pg) was digested with 45 units of EcoRI in a 10O-pl reaction for 2 h a t 37 "C, phenol extracted, ethanol precipitated, and dissolved in TE buffer. The ligation reaction (400 pl) contained 100 pg of chromosomal DNA (50 pg from each EcoRI digestion), 50 pg of plasmid DNA, 2000 units of T4 DNA ligase (New England Biolabs) and was incubated for 18 h at 12 "C (Cohen et al., 1973). Following ligation, strain KLC818 was transformed with this DNA in four independent transformations (Cohen et al., 1972). After transformation, cells were incubated at 30 "C for 4 h in T-broth and then diluted 10-fold into T-broth containing 5 pg of tetracycline/ ml. Incubation was continued until all cultures reached an AF,W of approximately 1.5. T-broth agar plates containing 5 pg/ml of tetra- cycline and 0.5 pg of mitomycin C/ml were spread with 0.2 ml of cells each and irradiated with 17 J/m2 of UV. Plates were incubated at 30 "C. Surviving colonies were grown in T-broth containing 5 pg of tetracycline/ml and tested for sensitivity to UV.

Measurement of SSB in Cell Extracts Cells growing exponentially a t 30 "C in T-broth were harvested by

centrifugation and resuspended in 10 mM Tris-HCI buffer (pH 8.1), 0.1 mM EDTA to a concentration of 2.5 X lo9 bacteria/ml. This was equivalent to a protein concentration of approximately 1 mg/ml. Cells were lysed by sonic irradiation with a Branson W185 sonicator.

Two methods were utilized for the determination of SSB in cell extracts. One method measures SSB-binding activity and can be used to quantitatively estimate wild type as well as the ssb-113-encoded protein in pure fractions and in crude extracts. We have previously described this assay in detail and have shown it to be at least 99% specific for E. coli SSB (Whittier and Chase, 1980). The specific activity of the DNA employed in the assay reported here was 39,000 cpm/nmol.

A solid phase radioimmunoassay was also used to measure SSB (Askenase and Leonard, 1970). Anti-SSB serum was collected from rabbits and the y-globulin fraction was prepared as previously de- scribed (Whittier and Chase, 1980). SSB was labeled with "'1 by the procedure of Syvanen et al. (1973).

Purification of SSB-113 Mutant Protein

Since the ssb-113-encoded protein binds a t least as well as wild type SSB to single-stranded DNA (see below), the procedure em- ployed to purify the wild type protein utilizing ssDNA cellulose chromatography (Chase et al., 1980) was used to purify the SSB-113 mutant protein with one alteration (PBE-94 chromatography in place of DEAE-Sephadex) which is now also employed in purification of the wild type protein. The purification described was from 160 g of cell paste of strain KLC792/pKAC20 and is summarized in Table I. In Fig. 1 is shown sodium dodecyl sulfate-polyacrylamide gel analysis of the SSB-113 protein and wild type SSB. Cells were grown and the SSB-113 protein was purified through the ssDNA cellulose step as described (Chase et al., 1980). Following elution from ssDNA cellu- lose, the 2 M NaCl eluate was dialyzed to equilibrium against 25 mM imidazole buffer (pH 7.01, 1 mM j3-mercaptoethano1, 10% glycerol (Buffer B) containing 0.15 M NaCI.

PBE-94 Chromatography-A column of PBE-94 resin (Pharmacia) (0.64 cm2 X 8 cm) was equilibrated with Buffer B containing 0.15 M NaCl and the sample was applied. The resin was washed with 10 ml of this same buffer. SSB-113 was eluted with a 100-ml linear gradient from 0.15 to 0.7 M NaCl in Buffer B. SSB-113 eluted a t about 0.35 M NaCI. Fractions of 2 ml were collected and several fractions in the region of the gradient between 0.2 to 0.5 M NaCl were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis in order to visualize contaminants and more efficiently pool SSB-113 containing fractions.

Concentration and Storage of SSB-I 13"Pooled fractions from PBE-94 chromatography were diluted with Buffer B to a conductivity

TABLE I Purification of ssb-113 encoded SSB

Fraction Activity Protein activit4 Spcifz Yield

units X IO" mglrnl unitslmg %

I Extract 10.0 26.0 8.4 100 I1 2 M DNA-cellulose 1.1 0.026b 2080 11

111 PBE-94 (concen- 0.82 2.53b 1560 8.2 wash (dialyzed)

trated) a One unit is 1 nmol of DNA bound determined by filter binding

assay (see "Materials and Methods"). Determined by amino acid analysis.

r-" ~165,000 7155.000 - 90,boo

-40,000

I 2 3 FIG. 1. Analysis of SSB-113 and SSB by sodium dodecyl

sulfate-polyacrylamide gel electrophoresis. Lane 1,3 pg of SSB- 113; Lane 2, 3 pg of SSB, Lune 3, protein standards, E. coli RNA polymerase (165,000-, 155,000-, 90,000-, and 40,000-dalton subunits).

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant ssDNA Binding Protein 807

corresponding to 0.15 M NaCl in Buffer B. SSB-113 was absorbed to a 1-ml column of PBE-94 equilibrated with 0.15 M NaCl in Buffer B and eluted with 0.7 M NaCl in Buffer B. Fractions (0.5 ml) were collected and the concentration of SSB-113 was determined using a molar extinction coefficient of 2.7 X 10' at 280 nm (Chase et al., 1980) or amino acid analysis. The final concentration of the preparation described here was 2.53 mg/ml (by amino acid analysis) and the total yield was 5.3 mg. Fractions were stored at -45 "C in small aliquots to avoid refreezing as much as possible. Although we have no direct evidence that refreezing solutions of SSB-113 is damaging, as a matter of practice we avoid it.

Purification of Single-stranded DNA Binding Protein from Various Strains

Wild type SSB and SSB-113 from E. coli B used for comparative peptide mapping were purified from strains WP2 and PAM33 respec- tively, by the original published procedure (Chase et al., 1980). Single- stranded DNA-binding protein from KLC792/pKAC23 and wild type SSB from E. coli K12 were purified from KLC221/pKAC4 by the procedure described here for the SSB-113 protein from KLC792/ pKACPO.

Comparative HPLC Peptide Mapping of SSB and SSB-13 The SSB and SSB-13 proteins were denatured by trichloroacetic

acid precipitation (final concentration 10% (v/v)) and then concen- trated by centrifugation. The precipitate was washed three times with cold acetone and resuspended in 0.5 ml of 25 mM NH,HCO, by sonication for 5 min in a Precision water bath sonicator. Proteolytic digestion was then performed by the addition of either trypsin (EC 3.4.21.4) a t a substrate-to-enzyme ratio of 25 (w/w) or chymotrypsin (EC 3.4.21.1) at a substrate-to-enzyme ratio of 30 (w/w) and allowing the reaction to proceed a t 37 "C. After 6 h (for trypsin) or 8 h (for chymotrypsin) the digestions were terminated by quick freezing and lyophilization. The resulting peptides were dissolved in 0.5 ml of 5% acetic acid and applied to a Waters Associates pBondapak C-18 column (0.39 X 30 cm) equilibrated in 10 mM potassium phosphate (pH 2.5) as described in the figure legends. Peptides were eluted from the column with linear gradients of acetonitrile into the aqueous phase. Column eluents were monitored by direct UV detection and peptide peaks were collected as discrete fractions.

Aliquots of all major peaks were mixed with 2 nmol of norleucine (to serve as an internal standard for calculating yields), dried under a stream of nitrogen a t 55 "C and hydrolyzed at 115 "C for 16 h in 6 N HCl, 0.2% phenol. Amino acid compositions were determined using a Beckman 121M amino acid analyzer. The HPLC purified, COOH- terminal tryptic peptide (T-14,15; residues 116-177) from SSB and SSB-113 was neutralized with triethylamine and dried under nitrogen prior to further digestion with chymotrypsin (protein/protease (w/w) ratio of 30 in 50 mM NH,HCO, a t 37 "C for 4 h) or pepsin (protein/ protease (w/w) ratio of 50 in 5% (v/v) formic acid a t 23 "C for 2 h). Proteolytic digestions on peptides were terminated and HPLC was carried out as described above.

Amino Acid Sequencing of SSB and SSB-113 Peptides The peptides from SSB and SSB-113 which were selected for

sequence analysis following HPLC purification and amino acid anal- ysis were neutralized with triethylamine and dried under a stream of nitrogen (L'Italien and Laursen, 1981). The resulting peptides plus their buffer salts were immobilized to aminopolystyrene (Sequemat) using the water-soluble carbodiimide procedure which we have re- cently described (L'Italien and Strickler, 1982). The peptides which were immobilized by this procedure appear to have been selectively coupled to the COOH-terminal carboxyl group of the peptide because both aspartic and glutamic acid residues were readily discernible and there was no detectable loss of sequencable peptide in the cycles following either of these residues, as previously noted (L'Italien and Laursen, 1981; L'Italien and Strickler, 1982).

Peptides immobilized to aminopolystyrene were sequenced using a Sequemat Mini-15 solid phase sequencer employing a 65-min sequen- cer program which has been previously described (L'Italien and Laursen, 1982). Coinversion of anilinothiazolinones to PTH amino acids was performed automatically using a Sequemat P-6 auto con- verter as described (L'Italien and Strickler, 1982). P T H amino acids, resulting from the sequencing process, were identified by HPLC using the conditions previously described (L'Italien and Lawsen, 1981; L'Italien and Strickler, 1982).

Fluorescence-quenching Studies Fluorescence measurements were carried out on an SLM spectro-

fluorimeter (Model SOOO), equipped with a magnetic stirrer and temperature-regulated cuvette holders. All fluorescence-quenching measurements were made at an excitation wavelength of 285 nm and emission wavelengths of 340-350 nm. Changes in protein fluorescence were corrected for quenching due to oligonucleotide absorption of incident radiation by subtracting the normalized change in fluores- cence of N-acetyltryptophanamide due to nucleotide addition. All experiments were performed in 50 mM Na,HPO,, 1 mM Na2EDTA, 1 mM 0-mercaptoethanol, in a volume of 2.5 ml, with initial monomer protein concentrations of 0.15-2.5 pM.

Association constants for the noncooperative binding of protein to 01igo[d(pT)~] were computed from Scatchard plots, assuming that each monomer can bind one d(pT)s molecule. Apparent association constants for the binding of SSB and SSB-113 to poly(dT) were determined from the fluorescence titration curves, assuming Scat- chard-type binding of tetramers to independent sites, as previously described (Williams et al., 1983). Titrations done at high protein concentrations (>2 p ~ ) with single-stranded fdDNA indicated that there was no significant difference in the apparent binding site size of SSB as compared to SSB-113. In calculating binding constants the occluded site sizes for both these proteins were taken to be 45.2 nucleotides per tetramer (Williams et al., 1983).

Other Methods Isoelectric focusing, thermal melting, and sedimentation studies

and measurements of the rate of dissociation of ssDNA.protein complexes were performed as previously described (Williams et al., 1983). Protein concentrations were determined by the method of Lowry et al. (1951) or by amino acid analysis as noted. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis in 12% gels was performed using the buffer system of Laemmli (1970) as previously described (Chase et al., 1980). Concentrations of DNA are expressed as equivalents of nucleotide phosphorus. All pH measurements were made a t a buffer concentration of 0.05 M at room temperature.

RESULTS

Cloning of the ssb-113 Mutant Gene The ssb and uvrA genes are contiguous and the entire ssb

gene and the portion of the uurA genes encoding the NH,- terminal portion of the protein are carried on a 7-kb EcoRI DNA fragment (Sancar and Rupp, 1979; Sancar et al. 1981a). We have devised a selection which allows cloning of the uvrA gene or the portion of the gene carried on this 7-kb DNA fragment and hence cloning of the adjacent ssb gene (Chase et al., 1983b). The entire uvrA gene could only be cloned from a partial EcoRI DNA digest. However, since the uurA6 mu- tation occurs within the portion of the uurA gene carried on the 7-kb EcoRI DNA fragment, a wild type uvrA gene can be produced on the chromosome by homogenotization or recom- bination between the cloned 7-kb EcoRI DNA fragment from a uwA+ strain and a uurA6 chromosomal gene. Although ssb- 113 mutant strains are UV-sensitive, uurA6 strains are ap- proximately 3 orders of magnitude more UV-sensitive and uur- ssb-113 double mutant strains are still further reduced in survival following UV irradiation (Whitier and Chase, 1981). Therefore, survival of a uurA+ derivative is strongly favored. The procedure employed to clone the ssb-113 mutant gene utilized a double selection based on the sensitivity of uvrA- mutant cells to mitomycin C as well as UV and has already been described in detail (Chase et al., 198313). DNA isolated from an ssb-113 mutant strain (KLC792) was digested with EcoRI and ligated with EcoRI-cleaved DNA of plasmid pSC101. Transformed cells were spread on mitomycin C- containing plates after an expression period and irradiated with 17 J/m' of UV (see "Materials and Methods"). Surviving cells were then individually tested for sensitivity to UV by streak tests. In this way, five pSC101 derivative clones were isolated which showed increased UV resistance compared to

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant SSDNA Binding Protein

the uurA6 parental strain (KLC818). These clones were iso- lated from three independent transformations, however only one clone (KLC955) was studied in detail. All were shown to produce approximately 5-10-fold more single-stranded DNA- binding protein than a control strain by both a radioimmu- noassay using antibody to wild type SSB and a filter-binding assay (see “Materials and Methods”). This is consistent with the quantity of protein anticipated to be produced by the gene cloned into pSC101. The isolated clones each contained only a 7-kb DNA insert. Therefore, the entire uurA gene could not have been cloned and the complementation of UV sensitivity observed was probably due to conversion of the mutant uvrA6 gene to wild type either by recombination or homogenotiza- tion as was previously observed during the cloning of the ssb- 1 gene (Chase et al., 1983b).

Plasmid DNA was isolated from a culture of KLC955 grown from the colony from the original selection plate and was used to transform KLC792 ssb-113. These transformants were individually tested for UV- and temperature-sensitivity. Two classes of transformants were apparent from this analysis. One class (64% of those tested) appeared to be nearly as UV- and temperature-sensitive as KLC792 alone while the other class (36% of those tested) was nearly as UV- and tempera- ture-resistant as wild type. Analysis of the SSB isolated from each cell type (including comparative peptide mapping and amino acid sequencing, see below, Fig. 2) showed that the UV- and temperature-sensitive cell type produced exclusively ssb-I13 mutant protein (the plasmid isolated from this strain contained the ssb-113 mutation and was designated pKAC20) while the UV- and temperature-resistant cell type produced a mixture of wild type and ssb-113 mutant protein. The plasmid isolated from the latter strain contained the ssb+ gene and was designated pKAC23. The observation that the ssb+ gene was transferred to the chromosome can be accounted for by the same recombination or homogenotization process that resulted in the production of uurA+ cells from the original uurA6 mutant strain used in the selection. Observations of a similar nature made in the cloning of the ssb-1 gene utilizing the same cloning strategy have already been described (Chase et al., 1983b).

Preliminary Characterization of Strains Producing Increased Quantities of ssb-113 Mutant Protein

The ssb-113 (pKAC2O) and ssb+ (pKAC23) plasmids iso- lated in this study were used to transform ssb-113 (KLC792) and ssb+ (HMS91) strains in order to begin an analysis of the cellular effects of increased quantities of ssb-113 mutant pro- tein. As shown in Table 11, both of the plasmids result in the production of a 5-lo-fold excess of SSB as measured by either a radioimmunoassay or measurement of binding to ssDNA directly in extracts. It is interesting to note that the radioim- munoassay using antibody to wild type SSB consistently detected less ssb-113 mutant protein compared to the wild type. This may suggest that antibody to wild type SSB reacts less efficiently to the mutant protein; however, this point has not been investigated further. In addition, consistently more binding activity was detected in extracts of ssb-113 mutant cells than in wild type cells. This may be significant consid- ering the fact that ssb-113 mutant protein is a slightly better helix-destabilizing protein than is the wild type protein (see below, Fig. 7). Finally, it should be noted that overproduction ofssb-113 mutant protein, a t least at the levels observed here, does not complement and in fact may further reduce the temperature survival of an ssb-113 mutant strain. This is in contrast to the effect observed in overproduction of the ssb-1 mutant protein in an ssb-1 mutant strain where complemen-

tation of the temperature defect was observed (Chase et al., 1983b). Although these results are difficult to directly compare since different expression vectors were employed in each study and the overproduction of the mutant proteins varied in each case, they do suggest different effects of SSB-113 on essential cellular reactions compared to SSB-1.

Comparative HPLC Peptide Mapping of SSB and SSB-113

Previous comparative tryptic peptide mapping studies of SSB and the ssb-113 mutant (Williams et al., 1982) demon- strated that the ssb-1 mutation results in a change in the elution time of one of the tryptic peptides from an HPLC reverse phase column. This peptide, T-14,15 (which corre- sponds to residues 116-177), exhibits a significant decrease in its retention time on the column. A portion of this tryptic fragment, T-14 (residues 116-154), was shown to co-migrate with its corresponding wild type peptide indicating that the mutation must be located in T-15 (residues 155-177). This conclusion was substantiated by comparative chymotryptic peptide-mapping experiments which demonstrated that the SSB-113 mutation decreased the elution time for C-4 (resi- dues 157-177) from 93 to 89 min (Fig. 2 A ) . The amino acid composition of this peptide indicated a possible proline to serine substitution (Table 111). These results were consistent with a previous study involving a site of proline-serine micro- heterogeneity (Teeter et al., 1981). HPLC peptide mapping of the microheterogenous peptides using a similar column and chromatographic conditions to those employed here resulted in the serine peptide eluting earlier than the proline peptide.

The ssb-113 mutation did not alter the elution position or amino acid composition (data not shown) of any other chy- motryptic or tryptic peptide. In addition, based on both of these later criteria, SSB from E. coli K12 strains is identical to that from B strains. Fig. 2A also clearly demonstrates that bacteria (KLC792) containing a chromosomal ssb-113 muta- tion and the pKAC20 plasmid produce only the SSB-113 protein.

As described in a previous section, based on the phenotypic properties (UV sensitivity) of a strain (KLC792) containing a chromosomal ssb-113 mutation and carrying the pKAC23 plasmid, we surmised that this plasmid, which was also iso- lated during the cloning of the ssb-113 gene, probably carried the ssb+ gene. We therefore isolated ssDNA-binding protein from this strain and analyzed chymotryptic peptides by HPLC compared to those produced from both pure SSB and SSB- 113 protein. As shown in Fig. 2B, the ssDNA-binding protein produced by KLC792/pKAC23 is clearly a mixture of SSB and SSB-113 protein approximately in a 5 to 1 ratio as would be expected from an ssb+ gene cloned into pSClOl carried in a strain with a chromosomal ssb-113 mutation.

Identification of the ssb-113 Mutation by Solid Phase Protein Sequencing

The amino acid compositions of all of the tryptic and chymotryptic peptides which we isolated from the wild type SSB were in complete agreement with the published sequence for this protein with one exception (Sancar et al., 1981b). The amino acid analysis of the tryptic peptide T-14 (residues 116- 154) indicated that it contained only two serines rather than three as predicted from the SSB sequence (data not shown). This discrepancy was resolved following HPLC purification (Fig. 3) of the peptide C-1 (residues 116-135) which resulted from chymotryptic digestion of the SSB tryptic peptide T- 14,15 (residues 116-177). Amino acid analysis of C-1 indicated that this peptide did not contain serine (Table 111). Direct solid phase protein sequencing of this peptide, as well as

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant ssDNA Binding Protein 809 - -

008 - -

0 0 6 -

- 0 0 4 -

0 0 2 - -

0 -

- 0 0 8 -

- 006 -

-

004 - 2 - (' 0.02 -

E + 008

f

0 - O -

0 c -

- 0 v) a Q 006

- - -

004 -

- 002 -

- 0 -

-

0 0 4 1 ,

::: r

0 0 2 1

SSB (B)

SSB-113 (B)

SSB-113 pKAC20)

20 40

I Minutes

0 08 ISSB I 0 O o 6 I 04 1 " 1

0

SSB/SSB-113 0 08

0 06

0 04

0 02

0 0 20 40 60 80 100

Mlnutes

FIG. 2. HPLC analysis of chymotryptic peptides from SSB and SSB-113. A, HPLC separation of chymotryptic peptides from 2.0 nmol of SSB from E. coli K12 and B strains and SSB-113 from an E. coli B strain containing a chromosomal ssb-113 mutation and a K12 strain containing a chromosomal ssb-113 mutation as well as a plasmid carrying the ssb-I13 mutant gene. The Waters C-18 column was eluted at a flow rate of 0.7 ml. min" with linear gradients of acetonitrile (solvent B) into 10 mM KH~POI, pH 2.5. 0-86 min (0-30% B), 86-109 min (30-60% B). B, HPLC separation of chymotryptic peptides from 2.2 nmol of SSB or SSB-113 and from ssDNA- binding protein purified from a strain containing a chromosomal ssb-113 mutation and a plasmid thought to carry the ssb' gene (see text). The column was eluted at a flow rate of 0.7 m1.min-I with linear gradients of acetonitrile (solvent B) into 10 mM KH2P04, pH 2.5. 0-72 min (0-25% B), 72-78 min (25-26.5% B), 78-118 min (26.5-30% B), 118-123 min (30-80% B).

DNA-sequencing studies, established that residue 133 is gly- cine (codon GGC) and not serine as was previously reported (Sancar et al., 1981b).3

The ssb-113 mutation was identified by solid-phase se- quencing of the chymotryptic peptide C-3,4 which was isolated as described in Fig. 3. The peptide was coupled to amino-

'' We have also found that the codon previously reported (Sancar et al., 1981b) for serine-39 (TCA) is incorrect. Although the correct codon (TCC) does not change the amino acid residue, it. does introduce a BstNl recognition site.

polystyrene in a 30% yield to give 4.1 nmol of immobilized peptide. One-half of the peptide-resin (2 nmol) was sufficient to allow the complete sequence analysis of this 30-residue peptide (Fig. 4). The only difference between the sequence of C-3,4 shown in Fig. 4 and that expected for SSB (Sancar et al., 1981b) is at cycle 29 where there is a serine residue (Figs. 4 and 5 ) in place of proline 176. Although the yield of PTH serine at cycle 29 was only about 60 pmol, it was nonetheless readily identified (Fig. 5 ) and there was no detectable increase in the yield of PTH proline at this cycle. Additional confir- mation that SSB-113 contains a serine at residue 176 was

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

810 SSB-113 Mutant ssDNA Binding Protein

TABLE I1 Properties of strains containing a cloned ssb-l I3 gene

Genotype Plasmid’ Ratio of hg SSB‘

Chromosome” Plasmid to mg protein Binding activityd EOP (43 “CY

(nrnoks DNA bound)/ (mg protein)

ssb- I13 0.14 1.37 1 x 10-3 ssb-113 PSClOl 0.18 1.36 9 x 10-4 ssb- I13 ssb- 113 pKAC2O 1.05 5.55 7 x 10-6 s.+113 ssb’ pKAC23 1.43 5.58 ssb’ 0.20 0.79 0.98

1.05

ssb’ pSCl0l 0.24 0.84 0.94 ssb’ ssb-I I 3 pKAC2O 2.20 4.72 0.80 ss b’ ssb’ pKAC23 2.10 4.64 1.06 ~-

“The ssb-113 host strain is KLC792; the ssb’ host strain is HMS91. All plasmids are derived from pSC101. Determined by solid phase radioimmunoassay of crude extracts using antibody to wild type SSB (see “Materials

Binding activity determined in crude extracts by the method of Whittier and Chase (1980). and Methods”).

“EOP (efficiency of plating) a t 43 “C is the ratio of viable cells at 43-30 “C determined on LB agar plates without NaCl (see “Materials and Methods”).

TABLE I11 Amino acid composition of SSB and SSB-113 peptides

Peptides were obtained by chymotryptic (C) or pepsin (P) digestion of the COOH-terminal tryptic peptide (T- 14, 15; residues 116-177) of SSB and SSB-113 as described under “Materials and Methods.” The numbers in parentheses are derived from the previously reported sequence of SSB assuming that both SSB and SBB-113 contain a glycine a t position 133 and that SSB-113 contains a serine in place of proline 176 in SSB. ND, not determined.

Amino acid

Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Tryptophan Residue

numbers Yield ( 7 6 )

c- 1 c-2 c-3,4 c - 4 P- 1

SSB SSB-113 SSB SSB-113 SSB SSB-113 SSB SSB-113 SSB SSB-113

1.2 (1) 1.3 (1) 1.1 (1) 1.1 (1) 5.1 (5) 5.2 (5) 5.0 (5) 4.7 ( 5 ) 3.0 (3) 3.1 (3)

3.2 (4) 4.1 (5) 1.8 (2) 2.5 (3) 1.0 (1) 3.3 (3) 3.3 (3) 5.0 (5) 5.0 (5) 4.2 (4) 4.3 (4) 2.3 (2) 2.4 (2) 2.0 (2) 2.0 (2) 2.1 (2) 2.1 (2) 6.4 (6) 5.7 (5) 5.2 (5) 4.8 (4) 1.2 (1) 9.2 (10) 9.2 (10) 2.9 (3) 2.9 (3) 2.2 (2 ) 2.3 (2) 2.2 (2 ) 2.2 (2) 4.2 (4) 4.3 (4) 3.1 (3) 3.0 (3)

1.0 (1) 1.1 (1) 0.8 (1) 0.9 (1) 0.7 (1) 0.8 (1) 1.0 (1) 1.0 (1) 1.0 (1) 0.9 (1) 0.9 (1) 0.9 (1)

1.0 (1) 1.1 (1) ND (1) ND (1)

116-135 136-147 148-177 157-177 171-177

46 59 49 83 32 31 9 26 65 40

obtained by pepsin digestion of T-14,15 from SSB and SSB- 113. Using the digestion conditions given under “Materials and Methods,” a rapid and preferential cleavage occurred between the aspartic acid-phenylalanine sequence at position 170-171. The resulting COOH-terminal peptide (P-1, residue 171-177) from both SSB and SSB-113 was isolated by reverse phase HPLC (data not shown) and subjected to amino acid analysis. As shown in Table 111, the amino acid composition of the P-1 peptide from SSB is consistent with the published sequence for SSB (Sancar et al., 1981b) while that from SSB- 113 is consistent with this protein having a serine in place of proline 176. A similar analysis was also done on SSB isolated from the led’-114 strain recently reported by Johnson (1982). HPLC tryptic peptide maps of SSB from lexC-114 were iden- tical to that for SSB-113 (data not shown). In addition, when the T-14,15 peptide from both of these proteins was digested with pepsin and then subjected to comparative HPLC peptide

mapping, the two resulting chromatograms were identical (data not shown). Amino acid analyses of the P-1 peptide (residue 171-177) from the lexC-114 SSB were also identical to that for the P-1 peptide from SSB-113. These results indicate that SSB from le&-114 strain is identical to SSB- 113 and that both proteins contain serine in place of proline 176.

Properties of SSB-113 Mutant Protein

The PI of the SSB-113 mutant protein was 6.0, identical to that of the wild type protein (Weiner et al., 1975). This is consistent with the amino acid replacement in the mutant protein (proline 176 to serine) which would not be expected to cause a significant change in the charge of the protein. The susceptibility of the SSB-113 protein to limited thermolysin digestion was identical to that previously observed for the wild type SSB (Williams et al., 1983), suggesting that the

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant ssDNA Binding Protein 81 1

I I I I I I I I I

t n S S B i

I I I I I I

1 o z 0 3 o 4 o 5 0 6 o m

Minutes FIG. 3. HPLC separation of 1.5 nmol of chymotryptic pep-

tides from the COOH-terminal tryptic peptide, T-14,15 of SSB and SSB-113. Preparative HPLC separations on 20 nmol of these digests looked identical to the analytical maps shown above. In hoth instances, the column was eluted a t a flow rate of 1.0 ml . mi& with linear gradients of acet.onitrile (solvent B) into 20 mM KH,PO,, pH 2.5. 0-80 min (0-40% B).

SSB-113 mutation does not substantially alter the normal folding of the COOH-terminal domain of SSB. In addition, raising the temperature of the digest to 55 "C did not specifi- cally enhance cleavage of the SSB-113 protein so, unlike the P7 temperature-sensitive mutation in the bacteriophage T4 ssDNA-binding protein (Williams and Konigsberg, 1983), the SSB-113 mutation does not seem to result in a thermolabile protein that unfolds at restrictive temperature. The s ~ ( , , ~ and K,,,, of the mutant protein were also found to be nearly identical to those previously reported for the wild type protein (Weiner et al., 1975; Williams et al., 1983) (data not shown). Taken together, these results suggest that the overall tetra- meric structures of both proteins are similar and that their complexes with ssDNA are equally stable.

SSB-I 13 Mutant Protein Binds As Well As Wild Type S S B to ssDNA-To assess the effect of the ssb-113 mutation on the DNA binding properties of SSB, we have determined the association constants for the binding of SSB-113 to oligonu- cleotide d(pT), and to poly(dT). As reported previously (Mol- ineux et al., 1975; Williams et al., 1983), SSB exhibits an intrinsic tryptophan fluorescence which is partially quenched upon binding to nucleic acids. As expected, SSB-113 exhibits an intrinsic fluorescence which is also quenched by nucleic acid binding. This fluorescence quenching has been used to measure the binding constants of oligonucleotides and poly- nucleotides to SSB and, in this report, to SSB-113. The fluorescence intensities of equimolar solutions of SSB and SSB-113 are essentially equal. Similarly, the excitation and emission spectra of SSB and SSB-113 are indistinguishable; both proteins exhibit maximum fluorescence at XeXCitatllln = 280 nm and X,,,,,,,, = 348 nm (data not shown).

Table IV summarizes the fluorescence-quenching measure- ments made with SSB and SSB-113. The fluorescence of both proteins is quenched to a small degree by binding to ~ l igo[d(pT)~] and to a much larger degree by binding to

1.5

I .o

0.5

I .o

Q) 0.5

E

z 1.0

u)

0

0 c 0

-

0.5

I .o

0.5

Phe 1 Pro c Ser 1

IO 20 30 IO 20 30 IO 20 30

Cycle Number I 10

Ser-Gly-GLy-Alo-Gln-Ser-Arg-Pro-Gln-Gln-Ser-Alo-Pro-Ala-Aio -Pro-Ser-Asn-Glu-Pro-Pro-Met-Asp-Phe-Asp-Asp-Asp-Ile-Ser-Phe

FIG. 4. Yields of PTH amino acids from the sequence anal- ysis of the SSB-113 peptide C-3,4. Identification of the residues present in each cycle was made by comparison with standard PTH amino acid retention times following high performance liquid chro- matographic separation. Integrated peak areas were converted to picomoles based on recovery of the 200 pmol of norleucine standard added to each cycle. The presence of each amino acid in the sequence is represented in the above figure by a large spot in the sequencer cycle in which it appears.

20 30

poly(dT), suggesting that both proteins undergo a conforma- tional change upon binding polynucleotides. Fig. 6 shows the fluorescence titration curves obtained for binding of SSB-113 and SSB to poly(dT). Titration curves were generated for ~ l igo [d (pT)~] binding to SSB and SSB-113, in a similar man- ner. The association constants for the noncooperative binding of SSB and SSB-113 to d(pT)B were computed by Scatchard plot analysis of the binding data and were found to be ap- proximately equal as indicated in Table IV. Raising the tem- perature of the binding reaction from 23 to 45 "C had only a slight effect on the affinity of these proteins for dfpT),. As shown in Table IV, the affinity of SSB-113 for poly(dT) is the same as that of SSB, at 23 "C. In addition, at 45 "C, both proteins' affinities for poly(dT) are undiminished and are essentially equal. Thus, within the limits of the fluorescence- quenching measurements, the binding parameters for SSB- 113 binding noncooperatively or cooperatively to ssDNA are not significantly different from that of wild type SSB.

The ssb-113 Mutation Results i n SSB Protein with Zn- creased Helix-destabilizing Ability-Previous studies have demonstrated that the helix-destabilizing "activity" of both the T4 and E. coli ssDNA-binding proteins are greatly in- creased by proteolytic removal of their acidic COOH termini (Hosoda and Moise, 1978; Williams et al., 1983). Although these effects result from extreme alterations of the proteins (i.e. deletion of the COOH terminus), we wondered if a more subtle change within the COOH terminus might have a cor-

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

812 SSB-113 Mutant SSDNA Binding Protein

I I Nor

0 I O ,

V c 0 n

I

0 5 IO I 5 20 25 - .

M l n u t e s FIG. 5. Actual HPLC traces of PTH-amino acids from cycles

26-30 of the SSB-113 peptide C-3,4 coupled to aminopoly- styrene by the carbodiimide method as described under "Ma- terials and Methods. " An internal standard (200 pmol of PTH norleucine) was added to each sequencer cycle which was dried under N,, dissolved in methanol, and then directly applied to a 5 - ~ m Altex Ultrasphere RP-18 column for identification. The sequence is shown hy the one-letter amino acid code placed over the peak corresponding to that, residue. The SSB-113 mutant shows a serine in cycle 29 of this peptide while the wild type SSB has a proline present in this position.

responding, if less drastic, effect. We therefore examined the effect of the SSB-113 mutant protein on the melting transi- tion of poly[d(A-T)]. As shown in Fig. 7, the SSB-113 mutant protein decreased the melting temperature of poly[d(A-T)] approximately 1.5 "C below that of SSB isolated from ssb+ E. coli K12 and B strains. Although this effect is small compared to the reduction in T, observed for the proteolytic products of SSB (Le. 15-20 "C, see Williams et al., 1983) it does suggest that the interaction of the mutant protein with DNA is altered as a result of the ssb-113 mutation.

TABLE IV DNA bindirza properties of SSB and SSB-113

Protein Ligand Temperature 76 AF" L-

"C "

23 SSB-113 d(pT), 23 10 1.0 x lo6 SSB 8.0 1.3 X lo6 SSB d(pT)s 45 13 5.3 X 105

SSB 23 74 1.1 x lo8

poly(dT) 23 73 1.0 x lo8 SSB-113 poly(dT) SSB

45 69 1.2 x 108 poly(dT) 45 68 1.2 x loR

SSB-113 d(pT)8 45 13 6.5 X 105

SSB-113 poly(dT)

The % AF is the maximum per cent fluorescence quenching that was observed upon the addition of a saturating concentration of ligand.

80 8 7

i 2 4 6 8 IO 12 14 16 18 20 22 24 26

NUCLEOTIDE / PROTEIN MOLE RATIO

FIG. 6. Fluorescence quenching of SSB-113 and SSB by poly(dT) binding. Measurements were made a t 45 "C in 50 mM Na,HPO,, pH 7.0, 1 mM Na2EDTA, and 1 mM 0-mercaptoethanol. The initial monomer protein concentration was 0.6 p M for both SSB- 113 (0) and SSB (0). Excitation wavelength was 285 nm; emission wavelength was 345 n n ~ .

E 0 1 0 - 1 I - CI Q N

558-113

a. b 2 004 Q

z

-

- 0 0 2 - Y v) Q

U 0 -

z I L I

4 0 50 60

TEMPERATURE I T ) FIG. 7. Thermal melting of poly[d(A-T)] in the presence of

SSB purified from wild type and ssb-113 mutant strains. Poly[d(A-T)] a t a concentration of 33.5 p~ (phosphate) was melted in the presence of 2.82 and 2.88 pM (monomer) SSB purified from E. coli ssb+ B and K12 strains, respectively, or 2.84 PM SSB-113 mutant protein. DNA:protein concentrations were adjusted to give approxi- mately stoichiometric ratios of nucleotides per SSB monomer using binding site sizes determined by fluorescence quenching.

DISCUSSION residues 1-115 at the NH2 terminus. The portion of the protein in the region near residues 115-135 is important in

Our previous studies of E. coli SSB have allowed us to subunit interactions which maintain the native tetrameric define at least three functional domains of the protein (Wil- structure and are probably necessary for cooperative interac- hams et aA, 1983). The DNA binding site is contained within tions in ssDNA binding. The COOH-terminal portion of the

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

SSB-113 Mutant SSDNA Binding Protein 813

protein (residues 136-176) is suggested to regulate or modu- late DNA binding inasmuch as its removal increases the helix- destabilizing ability of the remaining protein core. We have begun an effort to define the existing mutant SSB proteins within this framework.

Although a deficiency in SSB results in generally predict- able defects (i .e. , deficiency in DNA replication, postreplica- tion repair defects, and impaired recombination ability), i n vivo characterization of the ssb-1 and ssb-113 mutations in SSB has demonstrated a number of subtle differences which can now be studied at the biochemical level (Whittier and Chase, 1981, 1983). The in vitro analysis of the mutant proteins has been facilitated as a result of our previous cloning of the ssb-1 gene (Chase et al., 1983b) and the cloning of the ssb-113 gene reported here. Although these are both condi- tional lethal mutations and DNA replication ceases rapidly upon elevation to the restrictive temperature, ssb-113 mutant strains are deficient in DNA repair and recombination proc- esses at the temperature permissible for growth as well as the restrictive temperature, in contrast to ssb-1 strains which are essentially normal at the temperature permissible for growth and only show these other deficiencies after elevation to the restrictive temperature. We have determined that the amino acid change which results from the ssb-1 mutation occurs in the region of the protein we have shown to contain the DNA binding site (Williams et al., 1982, 1983). Although the ssb-l- encoded protein has not been studied in detail (one study has been published, see Meyer et al., 1980), it appears that its binding to ssDNA is effected more by ionic strength than is the wild type p r ~ t e i n . ~

The detailed ssDNA-binding studies reported here of the ssb-113-encoded protein clearly show that this protein binds at least as well to ssDNA as wild type SSB and may be a slightly better helix-destabilizing protein. Although a temper- ature-dependent ssDNA-binding effect might have been an- ticipated since the ssb-213 mutation is conditional lethal, none was detected (Table IV). In addition, the similar stabilities of SSB or SSB-113.ssDNA complexes and the nature of their thermal melting transitions does not suggest any obvious deficiency in cooperative DNA binding. The defects caused by the ssb-213 mutation cannot therefore result from a gross deficiency in the binding of SSB-113 to ssDNA. Whether these defects are due to increased helix-destabilizing ability of the ssb-113 encoded protein is at the moment unclear. Although this latter effect is small (T , reduced by 1.5-2 "C compared to wild type) it does qualitatively correlate with a recent electron microscopic observation. Evidence by electron microscopy in low salt shows no qualitative difference between wild type SSB and SSB-113 in the appearance of the nucleo- protein filament; however, as NaCl concentration is increased, mutant protein remains able to melt out secondary structure under conditions where wild type SSB cannot (Chrysogelos and Griffith, 1982).s If the defects caused by SSB-113 protein are directly the result of alterations of ssDNA binding, then the effects are subtle and escape easy detection. Of more basic significance, if such subtle differences cause the pleiotropic effects of the ssb-I13 mutation, then our current understand- ing of interactions of this type may be too limited to permit analysis at the present time. However, excluding the possibil- ity of a basic lack of understanding of this physical interaction and its role in biological processes, we favor the simpler interpretation that while the small effect in helix destabili- zation we have observed may be a contributing factor, the

J. W. Chase, J. B. Murphy, and K. R. Williams, unpublished

J. Griffith, unpublished observations. results.

major deficiency probably lies elsewhere. We have suggested previously that the pleiotropic effects exhibited by strains deficient in SSB could more simply be explained if SSB directly interacts with other DNA replication, recombination, and repair proteins (Chase et al., 1983a, 1983b). Circumstan- tial evidence pointing to this possibility has been accumulat- ing for some time. Tessman has recently reported genetic evidence suggesting an interaction of the ssb and rep gene products (Tessman and Peterson, 1982) and we have identi- fied a mutation mapping very close to rep which partially complements several effects of an SSB deficiency.6 McEntee et al. (1980) have shown an anomalous effect of SSB-113 on RecA protein-catalyzed strand assimilation compared to wild type SSB. Finally, it has recently been shown that approxi- mately one to three molecules of SSB copurify with a gene 5 ssDNA complex isolated from M13- or fl-infected bacteria, although it has not yet been proven that this interaction is specific.'

We have recently sequenced a ssDNA-binding protein from F sex factor which has extensive homology to E. coli SSB (Chase et al., 1983a). While the NH,-terminal portions of these proteins which contain the ssDNA-binding region are remarkably homologous, most of their COOH termini have extensively diverged with the exception of one region. It is striking that 6 of the last 7 COOH-terminal amino acid residues of both proteins are identical and the proline residue which is altered in SSB-113 is also present as proline in F plasmid SSB. It is suggestive therefore that the integrity of this region of the protein is essential for the functioning of both proteins. We have speculated that, if SSB does interact directly with other DNA metabolic proteins, then the COOH terminus of the protein may be a site of interaction (Chase et al., 1983a, 1983b).

We suggest that SSB may participate in various DNA metabolic processes by at least three mechanisms. The first and most obvious results from its direct binding to ssDNA and the roles it could play in protection, extending regions of ssDNA thus making them more accessible to various enzymes, and in catalyzing renaturation of ssDNA. The second requires specific interactions between SSB and other DNA metabolic enzymes. At least one such interaction is already documented between protein n and SSB (Low et al., 1982) and, as noted above, others are suggested. Finally the possibility for a very subtle, delicately balanced mechanism exists, considering the specific ssDNA conformation that appears to result as a consequence of SSB binding (Chrysogelos and Griffith, 1982). This could incorporate both direct interactions with SSB, which might only be seen as a result of the conformational change which SSB undergoes when bound to ssDNA (Wil- liams et al., 1983); and specific interactions of various DNA replication, recombination, and repair proteins (perhaps re- quiring a conformational change of these proteins as well) which might only occur efficiently in the presence of the appropriate SSB -ssDNA complex. Subtle changes in SSB structure could therefore easily be imagined to have far reach- ing and varied effects. Although we still find SSB-113 some- what of an enigma, we believe it is very likely that the eventual understanding of this and other SSB mutant proteins at the molecular level will yield much greater insight into the partic- ipation of SSB in DNA metabolic processes than could earlier have been imagined.

The sequencing of the SSB-113 mutant protein reported here also represents a technical achievement worth emphasiz-

e J. Chase, unpublished observations. ' R. Webster, personal communication. " J. W. Chase and K. R. Williams, unpublished results.

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

814 SSB-113 Mutant ssDNA Binding Protein

ing. Using solid phase amino acid sequencing, it was possible to sequence a 30-residue peptide in its entirety. It would have been extremely difficult to perform such an analysis using any other current methodology. Also worth noting is the comparative peptide mapping of tryptic and chymotryptic peptides which not only allowed the identification of the peptide containing the SSB-113 mutation despite the fact that it does not result in an alteration of charge, but also permitted the analysis of the ssDNA-binding protein pro- duced in a strain containing both the ssb+ and ssb-113 genes. These techniques are clearly more powerful than more con- ventional approaches and can obviously be applied to any system for similar types of analyses.

Acknowledgments-We are grateful to Dr. William Konigsberg for helpful discussions and to Elke Lorenson and Mary LoPresti for expert technical assistance.

REFERENCES

Askenase, P. W., and Leonard, E. J. (1970) Immunochemistry 7, 29-

Baluch, J., Chase, J. W., and Sussman, R. (1980) J. Bacterid. 144,

Chase J. W., Whittier, R. F., Auerbach, J., Sancar, A., and Rupp, W.

Chase, J . W., Merrill, B. M., and Williams, K. R. (1983a) Proc. Natl.

Chase, J. W., Murphy, J. B., Whittier, R. F., Lorensen, E., and

Chrysogelos, S., and Griffith, J. (1982) Proc. Natl. Acad. Sci. U. S. A.

Cohen, S. N., and Chang, A. C. Y. (1977) J. Bacteriol. 132, 734-737 Cohen, S. N., Chang, A. C. Y., and Hsu, L. (1972) Proc. Natl. Acad.

Cohen, S. N., Chang, A. C. Y ., Boyer, H. W., and Helling, R. B. (1973)

Coleman, J., and Oakley, J . (1979) Crit. Reu. Biochem. 3, 247-289 Glassberg, J., Meyer, R. R., and Kornberg, A. (1979) J. Bacteriol.

Golub, E. I., and Low, K. B. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,

Hosoda, J., and Moise, H. (1978) J. Biol. Chem. 253, 7547-7555 Johnson, B. F. (1977) Mol. Gen. Genet. 157 , 91-97 Johnson, B. F. (1982) Mol. Gen. Genet. 186 , 122-126 Konrad, E. B., and Lehman, I. R. (1974) Proc. Natl. Acad. Sci. U. S. A.

71

489-498

D. (1980) Nucleic Acids Res. 8, 3215-3227

Acad. Sci. U. S. A. 80, 5480-5484

Sninsky, J . J. (198313) J. Mol. Biol. 164 , 193-211

79,5803-5807

Sci. U. S. A. 69,2110-2114

Proc. Natl. Acad. Sci. U. 5'. A. 70 , 3240-3244

140 , 14-19

1401-1405

71,2048-2051 Kowalcvkowski. S.. Bear. D., and von Hippel, P. (1981) in The

E n z y k s (Boyer,'P., ed.) Vol. 14a, pp. 373-444, Academic Press, New York

Kupersztoch, Y. M., and Helinski, D. R. (1973) Biochem. Biophys. Res. Commun. 54 , 1451-1459

Laemmli, U. K. (1970) Nature (Load.) 227,680-685 Lieberman, H. B., and Witkin, E. M. (1983) Mol. Gen. Genet. 190 ,

92-100 L'Italien, J. J., and Laursen, R. A. (1981) J. Biol. Chem. 256,8092-

8101 L'Italien, J. J., and Laursen, R. A. (1982) in Methods in Protein

Seouence Analysis (Elzinga. M.. ed.) PP. 383-399, Humana Press, Clifton, NJ -

- . . "

L'Italien, J . J., and Strickler, J . E. (1982) Anal. Biochem. 127, 198- 212

6242-6250 Low, R. L., Shlomai, J., and Kornberg, A. (1982) J. Biol. Chem. 257,

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951)

McEntee, K., Weinstock, G. M., and Lehman, I. R. (1980) Proc. Natl.

Meacock, P. A., and Cohen, S. N. (1980) Cell 20,529-542 Meyer, R. R., Glassberg, J., and Kornberg, A. (1979) Proc. Natl. Acad.

Meyer, R. R., Glassberg, J., Scott, J. V., and Kornberg, A. (1980) J.

Molineux, I. J., Pauli, A., and Gefter, M. L. (1975) Nucleic Acids Res.

Sancar, A., and Rupp, W. D. (1979) Biochem. Biophys. Res. Commun.

Sancar, A., Wharton, R. P., Seltzer, S., Kacinski, B. M., Clarke, N.

Sancar, A., Williams, K. R., Chase, J . W., and Rupp, W. D. (1981b)

Syvanen, J . M., Yang, Y. R., and Kirschner, M. W. (1973) J. Biol.

Teeter, M. M., Mazur, J. A., and L'Italien, J. J. (1981) Biochemistry

Tessman, E. S., and Peterson, P. K. (1982) J. Bacteriol. 152, 572-

Vales, L. D., Chase, J . W., and Murphy, J. B. (1980) J. Bacteriol.

Weiner, J. H., Bertsch, L. L., and Kornberg, A. (1975) J. Biol. Chem.

Whittier, R. F., and Chase, J. W. (1980) Anal. Biochem. 106,99-108 Whittier, R. F., and Chase, J. W. (1981) Mol. Gen. Genet. 183 , 341-

347 Whittier, R. F., and Chase, J. W. (1983) Mol. Gen. Genet. 190, 101-

111 Williams, K. R., and Konigsberg, W. (1981) in Gene Amplification

and Analvsis (Chirikiian. J.. and Papas. T., eds.) Vol. 2, PP. 475-

J. Biol. Chem. 193,265-275

Acad. Sci. U. S. A. 77,857-861

Sci. U. S. A. 76, 1702-1705

Biol. Chem. 255,2897-2901

2,1821-1837

90,123-129

D., and Rupp, W. D. (1981a) J. Mol. Biol. 148,45-62

Proc. Natl. Acad. Sci. U. S. A. 78,4274-4278

Chem. 248,3762-3768

20,5437-5443

583

143,887-896

250,1972-1980

508, ElseGier/North Holland, New York ..

Williams, K. R., and Konigsberg. W. H. (1983) in The Bacteriophage 2'4 (Matthews, C., Kutter, B.:'Mosig, G., and Berget, P., eds.) pp. 82-89, American Society for Microbiology, Washington, D. C.

Williams, K. R., L'Italien, J. J., Guggenheimer, R. A., Sillerud, L., Spicer, E., Chase, J. W., and Konigsberg, W. (1982) in Methods in Protein Seouence Analysis (Elzinza. M.. ed.) pp. 499-507, Humana Press, Clifion, NJ

I . - . "

Williams, K. R., Spicer, E. K., LoPresti, M. B., Guggenheimer, R. A., and Chase, J . W. (1983) J. Biol. Chem. 258,3346-3355

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from

J W Chase, J J L'Italien, J B Murphy, E K Spicer and K R Williamshigh pressure liquid chromatography peptide mapping, and DNA-binding studies.

DNA-binding protein. Cloning of the gene, DNA and protein sequence analysis, Characterization of the Escherichia coli SSB-113 mutant single-stranded

1984, 259:805-814.J. Biol. Chem. 

  http://www.jbc.org/content/259/2/805Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/259/2/805.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on June 14, 2018http://w

ww

.jbc.org/D

ownloaded from