Photoaffinity Labeling of DNA Template-Primer Binding Site in ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol . 269, No. 34, Issue of August . 21828-21834, 1994 Printed in U.S.A.

Photoaffinity Labeling of DNA Template-Primer Binding Site in Escherichia coli DNA Polymerase I IDENTIFICATION OF INVOLVED AMINO ACIDS*

(Received for publication, December 21, 1993, and in revised form, May 1, 1994)

Virendra N. Pandey, Neerja Kaushik, and Mukund J. ModakS From the Department of Biochemistry and Molecular Biology, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey 07103

We have used two self-annealing template-primers (TPs) to covalently cross-link the Klenow fragment of Escherichia coli DNA polymerase I in its polymerase mode. The specificity of cross-linking is demonstrated by the observation that other template-primers, but not the template or primer alone, readily compete with self- annealing TPs. The enzyme-TP covalent complex is catalytically active and can incorporate one nucleotide on the primer terminus of the immobilized template- primer. Using a peptide mapping approach, we have identified a 17-amino acid tryptic peptide spanning residues 759-775 as a major constituent of the TP binding domain. Amino acid sequence analysis further revealed that Ile-765, Tyr-766 in the 0-helix and Ser-769, Phe-771 in the 0,-helix of the three-dimensional crystal structure of the Klenow fragment constitute the attachment site for TP.

The photochemically induced covalent linkage of proteins to DNA stabilizes macromolecular interactions, provides a specific means of labeling the domains involved in the binding, and allows one to identify the specific contact points between rele- vant amino acid residues and the polynucleotide substrate. Using this approach many different amino acids, which include serine, isoleucine, threonine, and cysteine, have been shown to be involved in the formation of covalent protein-nucleic acid linkage (reviewed by Williams and Konigsberg (1991)). Most importantly, in those instances where three-dimensional protein structures are available, the amino acids involved in the cross-linking to nucleotides are indeed found to be in the proximity of the polynucleotide substrates. Using this technology, His-881 (Pandey et al., 1987) and Tyr-766 (Rush and Konigs- berg, 1990) were shown to be involved in dNTP binding, whereas Leu 289-Thr-290 and Leu-295-Thr-296 in human im- munodeficiency virus reverse transcriptase (Basu et al., 1992) and Pro-76 in murine leukemia virus reverse transcriptase (Tirumalai and Modak, 1991) were identified as the contact points for the binding of primer DNA. The availability of the three-dimensional crystal structure of the large fragment (me- now fragment) of Escherichia coli DNA polymerase I (Ollis et al., 1985) and its recently described molecular model (Yadav et al., 1992, 1994a, 1994b), complete with side chain orientation, have provided an ideal model system for understanding the

* This work was supported in part by Grant GM-36307 from NIGMS, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must

U.S.C. Section 1734 solely to indicate this fact. therefore be hereby marked “advertisement” in accordance with 18

$ To whom correspondence should be addressed: Dept. of Biochemis- try and Molecular Biology, University of Medicine and Dentistry, New Jersey Medical School, 185 S. Orange Ave., Newark, N J 07103. Tel. 201-982-5515; Fax: 201-982-5594.

structure-function relationship in the polymerase class of enzymes. Although chemical modifications (Pandey and Modak, 1988a; Pandey et al., 1990; Basu et al., 1987, 1988; Basu and Modak, 1987; Mohan et al., 1988), affinity labeling (Pandey et al., 1987; Joyce et al., 1986; Rush and Konigsberg, 1990), and site-directed mutagenesis (Pandey et al., 1993, 1994; Polesky et al., 1990; Derbyshire et al., 1988; Polesky et al., 1992) have identified several important amino acid residues involved in substrate binding and catalytic functions, the template primer (TP)l binding domain of this enzyme has not yet been identified. Recently a structure of Klenow fragment co-crystallized with somewhat unusual double-stranded DNA containing a short 3”OH overhang has been described (Beese et al., 1993). Because of the 3’ overhang, the binding of the enzyme to TP has been proposed to be in the editing mode, whereas the position of the 5’ terminal nucleotide of the complementary strand is postulated to be in the template binding site (Beese et al., 1993). The double-stranded DNA binding region appears to be well defined in this structure. However, no indication for the binding of the template strand is given by these studies. In this paper, we demonstrate that a self-annealing TP can be covalently cross-linked to the Klenow fragment and that the binding is in the polymerase mode of the enzyme. From the amino acid sequence analysis of the appropriate peptide, we have identified the amino acid residues that are involved in the binding of TP. These results together with their biological im- plications are the subject of this paper.

EXPERIMENTAL PROCEDURES Materials-E. coli DNApolymerase I (Klenow fragment) was purified

from an overproducing exonuclease deficient strain (Derbyshire et al., 1991) generously provided by Catherine Joyce ofYale University. DNA- modifying enzymes were from Promega or Boehringer Mannheim. Ho- mopolymeric TPs and HPLC-purified dNTPs were obtained from Phar- macia Biotech Inc., while tritiated and 32P-labeled dNTPs and ATP were the products of DuPont NEN. Trypsin (TPCK-treated) was purchased from Worthington Corp. All other reagents were of the highest available purity grade and were purchased from Fisher, Millipore Corp., Boeh- ringer Mannheim, and Bio-Rad. Two synthetic self-annealing template- primers were obtained from Midland Certified Reagent Company. These were 25-mer and 37-mer species with the following sequences.

25-mer self-annealing TP CACGAGTGGG-3’ TCACTCACCCCAGAG-5’

37-mer self-annealing TP CACGCAGTCTTCTCC-3‘ TCACGTCAGAAGAGGATCCCTC-5’

template primer; Dm, dithiothreitol; PTH, phenylthiohydantoin; pol I, The abbreviations used are: TP, template-primer; E-TP, enzyme-

E. coli DNA polymerase I; PAGE, polyacrylamide gel electrophoresis; TPCK, L-1-tosylamido-2-phenylethyl chloromethyl ketone; WT, wild type; TEAA, triethylammonium acetate.

21828

Template-Primer Binding Domain of Pol I 21829

Enzyme Assay-The polymerase activity of the Klenow fragment was assayed as described before using activated calf thymus DNA or synthetic homopolymeric template-primers and c3H1dNTP or [cu-32PldNTPs as the source of radiolabeled substrate (Pandey et al., 1987; Pandey and Modak, 1988a).

End Labeling of Self-annealing TPs-The self-annealing TPs were kinased at the 5’ end using [y-[32P]ATP and T4 polynucleotide kinase or labeled at the 3’ end using the respective [CX-~~PI~NTP and Klenow fragment essentially as described by Sambrook et al., (1989). The labeled TPs were purified by electrophoresis on 20% polyacrylamide, 7 M

urea gel as described by Maxam and Gilbert (1980). W-mediated Cross-linking of Enzyme to TP-Photochemical cross-

linking of 5’ 32P-labeled 25-mer or 37-mer TP to the Klenow fragment was performed on ice in a buffer containing 50 mM Tris-HC1, pH 7.5, 1 mM MgCl,, 1 mM DTT, and 2% glycerol. Optimum cross-linking conditions were established by varying the concentration of Klenow fragment and TP from 0.25 to 100 p for Klenow fragment and from 0.5 to 200 p for the TP. For the cross-linking reaction, the enzyme and radiolabeled TP were incubated on ice for 10 min, followed by exposure to UV irradiation in a Spectrolinker (Spectronic Corp.) at a dose rate of 250 mJ/ em2. Measurement of covalent attachment of labeled TP to enzyme protein was assessed by SDS-PAGE, followed by autoradiography and quantitation by either densitometric scanning of the autoradiogram or by excision of the radioactive bands from the gel and determination of Cerenkov counts associated with the band (Pandey and Modak, 1988~).

Preparative Scale Cross-linking of Labeled TP to KZenow Fragment and Isolation of E-TP Covalent Complex-100 nmol of Klenow fragment (exo-) was mixed with 200 m o l of labeled 25-mer or 37-mer TP in 1 ml of incubation buffer, incubated on ice for 10 min, and dispensed in 50-1.11 aliquots on parafilm, which was placed on an ice block. Samples were irradiated as described above. The irradiated samples were pooled, adjusted to 250 mM NaCl, and applied to a DEAE-Sephadex A-25 column (3 ml), which was preequilibrated with 10 mM Tris-HC1, pH 8.0, containing 1 mM DTT and 250 m~ NaCl. The un-cross-linked enzyme was recovered in the column flow-through and the buffer wash, while enzyme cross-linked to TP and free TP were both retained on the column. The column was further washed extensively with the same buffer to remove traces of free enzyme. E-TP covalent complex, as well as bound TP, was eluted with 1 M NaCl in the same buffer. The column eluate was desalted using a Centriprep-30 ultrafiltration device with a molecular mass cut-off of 30 kDa (Amicon, W. R. Grace & Co.). During this procedure greater than 98% of NaCl, as well as free radiolabeled TP, was removed. SDS-PAGE analysis of the E-TP covalent complex showed the presence of a complex with the expected mobility.

Assay of Nucleotidyltransferase Activity of the E-TP Covalent Complex-10 nmol of enzyme was UV-irradiated with 20 nmol of unlabeled TP, and the E-TP covalent complex formed was purified by DEAE- Sephadex chromatography as described above. Traces of un-cross- linked TP present in the isolated complex were further removed by using a small Bio-Rex 70 (Bio-Rad) column (3 ml) preequilibrated with 10 mM Tris-HC1, pH 8.0. The column was washed with the equilibration buffer until the A260 was zero, followed by elution of the E-TP covalent complex with 0.25 M NaCl in the same buffer. The column eluate containing the E-TP covalent complex, free from un-cross-linked enzyme and TP, was concentrated and desalted as described above. The nucleotidyltransferase activity of the purified E-TP covalent complex was then measured in a reaction mixture consisting of 50 mM Tris-HC1, pH 7.8, 1 mM Dm, 5 IILM MgCl,, 1 p [CX-~~PI~NTP (5000 cpdpmol) corresponding to the first template base and 5-50 n~ purified E-TP covalent complex in a final volume of 50 p1. The reaction mixture was incubated for 30 min at room temperature or 20 min at 37 “C, and reactions were terminated by the addition of 1% SDS and 20 mM EDTA. An aliquot of the reaction mixture was subjected to 10% SDS-PAGE followed by autoradiography. The radioactivity associated with the E-TP covalent complex was determined by Cerenkov counting after excising the radioactive band from the gel.

Isolation of Klenow Zkyptic Peptide Cross-linked to Labeled TP-The labeled E-TP covalent complex isolated as described above was lyophilized and dissolved in 200 pl of 8 M urea in 50 mM Tris-HC1, pH 8.0, and then diluted to 800 pl with 0.1 M NH,HCO,, pH 8.0. The E-TP covalent complex was then trypsinized with TPCK-treated trypsin at a protein: trypsin ratio of 503. After 2 h of incubation at 37 “C, a second aliquot of trypsin (50:l ratio) was added and incubation was continued for an additional 12-16 h. The tryptic digests were directly loaded on a 2-ml DEAE-Sephadex A-25 column preequilibrated with 250 m~ triethylammonium acetate (TEAA) buffer, pH 8.0. After extensive washing of the column with 250 mM TEAA to remove the un-cross-linked tryptic peptides, the TP-cross-linked peptides were eluted with 2 ml of 1 M TEAA

buffer. The eluates were lyophilized, dissolved in 1 ml of H,O, and subjected to gel filtration through a NAP-10 column (Pharmacia) to remove traces of TEAA buffer. An aliquot of the sample was analyzed by C4 reverse phase HPLC equilibrated with 10 mM sodium phosphate, pH 6.8 (solvent A). Peptides were eluted by increasing the acetonitrile concentration (see Fig. 41, and 1-ml fractions were collected at a flow rate of 1 mumin. Each fraction was monitored for absorbance at both 220 and 254 nm. The radioactivity in individual fractions was determined by Cerenkov counting.

Amino Acid Sequence Analysis of TP-cross-linked Zkyptic Peptide- Since the tryptic digest of TP-cross-linked enzyme protein yielded only a single peptide peak upon its resolution on a C4 HPLC reverse phase column, sequence analysis was performed using the DEAE-Sephadex- purified peptide fraction. Amino acid sequence analysis of peptides was performed using an Applied Biosystems 470A gas phase sequenator a t the protein chemistly facility of Yale University under the supervision of Dr. K. R. Williams. The 03RPTH program supplied with this instru- ment was used without modification, and the resulting phenylthiohydantoin-derivatives were analyzed using an on-line Applied Biosystems model 470A microbore HPLC.

RESULTS

Properties of UV-mediated Cross-linking of TP to Klenow Fragment-In this study, we used two self-annealing TPs of 25 and 37 nucleotides in length. Both the TPs contain an intramo- lecular double-stranded region with a 5’ overhang consisting of 5-7 nucleotides, which constitutes the template strand. The advantage of using such TPs is that the slippage or breathing between template and primer strands during bindinghross- linking to the enzyme is effectively avoided. The self-annealing TPs were prelabeled with 32P either at the 5’ or at the 3”OH position before cross-linking to the Klenow fragment. Under optimal conditions of photolabeling, nearly 7-9% of the enzyme was found to be covalently cross-linked to the TP (Fig. L4). A typical dose-response pattern of the cross-linking of 37-mer TP to the Klenow fragment as a function of ultraviolet energy is shown in Fig. L4. The covalent cross-linking of TP to enzyme increased linearly up to 250 mJ/cm2 and reached a maximum of about 0.09 moVmol of enzyme. We therefore used 250 mJ/cm2 UV energy for covalent cross-linking in subsequent studies. Under these conditions, (a) no detectable UV-induced proteoly- sis or interprotein cross-linking was noticeable as judged by SDS-PAGE analysis and Coomassie Blue staining of the gel, ( b ) HPLC profile of tryptic digests of both irradiated and unirradiated Klenow fragments were identical (data not shown), and ( c ) nearly 90% of the catalytic activity of the irradiated enzyme was consistently recovered. A short incubation of enzyme-DNA complex was sufficient to reach equilibrium, since preincuba- tion times ranging from 30 s to 30 min showed nearly identical extent of cross-linking. Autoradiography of the TP-Klenow covalent complex resolved by SDS-polyacrylamide gel electrophoresis showed a shift in molecular mass of labeled enzyme species to 75- and 78-kDa positions with 25-mer and 37-mer TP, respectively (Fig. 1B). The shift in the migration pattern of cross-linked enzyme is consistent with the expected increase in the molecular mass of the complex due to the presence of TP. These observations also suggest that only a single molecule of TP is cross-linked to a single enzyme subunit. Some character- istics of the cross-linking reaction are shown in Table I. The ability of enzyme to cross-link to TP is affected by inclusion of other TPs, such as poly(dA).(dT),, and poly(dC).(dG),,, but not by template or primer alone. Similarly, substrate dNTPs have no effect on the extent of cross-linking of enzyme to TP. Ionic strength of the medium, which influences binding of TP to enzyme, does have a negative impact on the cross-linking reaction as judged by the decrease or near complete loss of cross- linking with increasing salt concentrations (Table I). The covalent labeling of the enzyme to TPs required the presence of the active enzyme species since heat-denatured enzyme did not show any Cross-linking. The extent of cross-linking increased

21830 Template-Primer Binding Domain of Pol I

L W - 0.10 , W

0.06 .“. n 2 3 0.04 1 0 m i p, I- 0.02 j / W A

0.00 Y . I I 1 . I I

0 100 2 00 3 00 400 5 00

68 --: kDa !

UV DOSE (mJlcmxcm) Flc. 1. A, covalent cross-linking of :iZP-labeled self-annealing TP to Klenow fragment as a function of UV irradiation. The standard reaction

mixture (100 pl) contained 3 nmol of 5’ “‘P-labeled 37-mer TP (300 cpdpmol) and 1 nmol of Klenow fragment. The mixture was incuhated on ice for 10 min and subjected to W irradiation as described under “Experimental Procedures.” The extent of covalent attachment of labeled TP was determined by SDS-PAGE. Labeled protein was localized by Coomassie Blue staining and autoradiography ofthe gel. The labeled protein band was excised from the gel, and the radioactivity was quantitated by Cerenkov counting. The figure shown in the inset is an autoradiogram of a cross-linking experiment with 30 pmol of Klenow fragment and 90 pmol of labeled 37-mer TP, irradiated at the indicated UV energy. B, gel analysis of the migration pattern of Klenow fragment cross-linked with ‘”P-labeled 25-mer TP, 37-mer TP, and TTP. The cross-linking of Klenow fragment with labeled 37-mer and 25-mer TP followed by SDS-PAGE and autoradiography of the gel were carried out as described for panel A. Simulta- neously, 30 pmol of Klenow fragment was also cross-linked with [ cr-:”PITTP and processed on the polyacrylamide gel. Note the shift in the migration patterns of enzyme labeled with 25-mer TP, 37-mer TP, and “P.

TABLE I Effect of competitors and ionic strength on the cross-linking of

37-mer TP to Klenow fragment 50 pmol of Klenow fragment cross-linked in the presence of 50 pmol of

“2P-labeled 37-mer in a standard reaction mixture consisting of 50 mM Tris-HCI, pH 7.5, 2% glycerol, and 1 mM MgCI,, was taken as 100%. Indicated concentration of each competitor was mixed with the labeled 37-mer TP before the addition of enzyme.

Additions ”‘P-TP Percent of cross-linked cross-linking

pmol 70

Enzyme + “‘P-labeled 37-mer TP 3.08 100 + Poly(&) (0.5 p ~ ) 2.77 90 + Oligo(dT) (0.5 p ~ ) 3.01 98 + Poly(dA)4dT),, (0.5 p ~ 1 ) 0.64 21 + Poly(dC) (0.5 phf) 2.92 95 + Oligo(dG) (0.5 p ~ ) 3.06 100 + Poly(dC)4dG),, (0.5 p ~ ) 0.36 12 + TTP (200 phr) 3.09 100 + dATP (200 p ~ ) 2.86 93 + NaCl (100 mht) 2.46 80 + NaCl (500 mb1) 0.104 3 + NaCl (1 11) 0.01 0.01

Heat-denatured enzyme + “P-TP 0.01 0.01

with increasing TP concentration. The optimum molar ratio of TP to enzyme was found to be 3:l (Table 11).

Specificity of TP Cross-linking to Klenow Fragment-The specificity of binding of the 37-mer to the Klenow fragment was suggested by reduction in its cross-linking to enzyme in the presence of other TPs but not by the template poly(dA) or primer oligo(dT) (Table I). A typical dose-response pattern of TP, template, and primer concentrations versus the observed cross-linking of 32P 37-mer DNA to enzyme is shown in Fig. 2. Results confirm the competitive nature of binding between 37- meric DNA and poly(dA).(dT),,. We estimate that the extent of labeling with TP is reduced to approximately 50% when the molar concentration of 3“OH termini contributed by poly(dA).(dT),, or poly(dC).(dG),, (data not shown) is close to their K,, values for the polymerase reaction.

TABLE I1 Extent of template-primer cross-linking to Klenow fragment as a

function of TP concentrations The standard irradiation mixture contained 50 pmol of the Klenow

fragment and increasing concentrations of ””P-labeled 37-mer DNA. The E-TP complex was exposed to ultraviolet light, and the extent of cross- linking was estimated as described under “Experimental Procedures.”

Addition TP cross-linked cross-linked

Protein

pmol %

100 pmol of TP (no W exposure) 0 0 25 pmol of TP 1.5 3.0 50 pmol of TP 2.5 5.0 100 pmol of TP 3.75 7.5 150 pmol of TP 4.5 9.0 200 pmol of TP 4.55 9.1 250 pmol of TP 4.57 9.14

Cross-linking of TP to Enzyme Represents Productive Binding-The most compelling evidence indicating that the cross-linking of enzyme to TP is a specific binding (i.e. in the polymerase mode) was obtained by the ability of the cross- linked enzyme to catalyze a nucleotide addition reaction. To demonstrate the catalytic competence of the TP-cross-linked enzyme, the non-radioactive 37-mer TP-Klenow and the 25-mer TP-Klenow complexes were separated from the free enzyme on a DEAE-Sephadex column as described under “Experimental Procedures.” The purified complexes from the two TPs were then mixed with :j2P-labeled substrate dNTP in a standard reaction mixture lacking external TP. The incorporation of the labeled nucleotide was monitored by SDS-PAGE of the reaction products followed by autoradiography. As shown in Fig. 3A, the covalently linked E-TP complex is catalytically active and can effectively incorporate dNTP at the 3”OH terminus of the cross-linked TP. This indicates that the covalently cross-linked TP is oriented in the polymerase mode and that the catalytic center of the enzyme in this complex is functional. It may be argued that the observed nucleotidyltransferase reaction may be due to traces of free enzyme, which may be present in the


I

poly dA

80 oligo dT

Competitor DNA (nM)

FIG. 2. Effect of DNA competitors on the cross-linking of labeled self-annealing TP to Klenow fragment. A standard irradiation mixture in a final volume of 50 pl contained 1 p~ Klenow fragment and 3 pu '"P-labeled 37-mer TP (500 cpm/pmol) and indicated concentrations of DNA competitor. The reaction conditions and the procedures for UV-induced cross-linking, SDS-PAGE, and autoradiography were the same as described in Fig. 1. The figure shown in the inset is an autoradiogram of the Klenow fragment cross-linked with labeled 37- mer TP, in the presence of 400 nM DNA competitor.

c--78 kC

A R FIG. 3. A, nucleotidyltransferase activity of E-TP covalent complex

and addition of nucleotides corresponding to 1st and 2nd template base. The Klenow fragment was cross-linked with unlabeled 37-mer or 25- mer TP and the E-TP covalent complexes were selectively purified by DEAE-Sephadex chromatography as described under "Experimental Procedures." The E-TP covalent complex equivalent to 5 nM protein was incubated in 50 1.11 of reaction mixture containing 50 mM Tris-HCI, pH 7.8,l mM DTT, 5 mM MgCl,, and 1 VM [a-"2PldNTPcomplementary to 1st template base (lane 1 ). In another set of experiments, the E-TP covalent complex was preincubated at room temperature for 20 min with 20 PM unlabeled dNTP complementary to 1st template base. 1 pal [cr-:'?PldNTP corresponding to the second template base (lane 2 ) was then added and incubation continued for another 30 min. Additions of 1st (lane 1 ) and 2nd nucleotides (lane 2 1 were analyzed by SDS-PAGE and autoradiography as described under "Experimental Procedures." B, effect of salt concentrations on the nucleotidyltransferase activity of E-TP covalent complex. The nucleotidyltransferase reaction of the purified E-TP covalent complex was carried out in a reaction mixture containing 5 nhi E-TP complex, 50 r n h i Tris-HCI, pH 7.8, 5 mM MgCI,, 1 mhl DTT, 1 PM la-"'PlTTP, and indicated concentrations of NaCl in a final volume of 50 pl. The reaction mixture was incubated at 37 "C for 20 min and was subjected to SDS-PAGE and autoradiography as described under "Experimental Procedures."

purified E-TP covalent complex. This possibility was ruled out by the observation that only a single nucleotide could be added on the primer terminus of the cross-linked TP. Addition of the 2nd nucleotide corresponding to the second template base in the cross-linked TP was not observed (Fig. 3A, lane 2). Thus, the cross-linked enzyme is unable to translocate across the template strand. In contrast, the free enzyme efficiently incor-

TAIIIX 111 Effect of salt (ionic strength) and dioalent cation on the incorporation

of substrate by cross-linked and free enzyme The Klenow fragment cross-linked with unlabeled 37-mer was puri-

fied by DEAE-Sephadex and Bio-Rex 70 columns chromatography as described under "Experimental Procedures." Five pmol of the covalently cross-linked purified E-TP complex was incubated at 37 "C for 20 min with 3 pCi of Ia-:"PldTTP (5000 cpm/pmol) in a standard reaction mixture consisting of 50 my Tris-HC1, pH 7.8, 1 mM DTT, 5 mhl MgCI,, and the indicated salt concentrations. The acid-insoluble radioactivity was counted to determine the extent of TTP incorporation. To compare the extent of incorporation by the free enzyme under similar reaction conditions, 0.5 pmol of free enzyme was preincubated with 25 pmol of 37-mer and incorporation of la-:"PldTTP in the presence of indicated salt concentrations was determined.

Reaction condition dNTP incorporated

rpm x IO-' E-TP covalent complex + la-RZPITTP

Control 20.0 - Mgy' 0.01 + 100 mM NaCl 20.0 + 500 m>l NaCl 19.0 + 1 N NaCl 7.0

Control 70.0 - M$' 0.2 + 100 mM NaCl 40.0 + 500 mv NaCl 0.5 + 1 >I NaCl 0.45

Free enzyme + TP + I n-:'2PITTP

porates both the 1st and 2nd (and even 3rd, 4th, and 5th) nucleotide under similar assay conditions with the same TP (data not shown). The salt-resistant nature of the nucleotidyltransferase reaction catalyzed by the TP-cross-linked enzyme also suggests that the observed polymerization of dNTP is carried out by the enzyme species in the complex. As seen from Fig. 3B, the observed nucleotidyltransferase activity of the E-TP covalent complex was only slightly affected by increasing the salt concentration and was detected even at 1 M salt concentration (Fig. 3B). In contrast, the incorporation of nucleotide on the primer terminus of the 37-mer TP by free enzyme is ex- tremely sensitive to high salt concentration due to the salt induced dissociation of TP from the enzyme (Table 111). For example, the polymerase activity of free Klenow fragment is reduced to background levels in the presence of a salt concentration above 500 mM. As expected, the nucleotidyltransferase reaction by E-TP covalent complex is strictly Me-dependent (Table 111) and appears to be quantitative. A simple calculation suggests that nearly 80% of the cross-linked TP has been ex- tended by the resident enzyme (Table 111).

cJ2PJdNMP Incorporated a t the Primer Terminus of E-TP Co- valent Complex Is Not Accessible to 3' -> 5'-Exonuclease--Other evidence in support of bindingkross-linking of TP in the polymerase mode of enzyme came from the following experiment. Following in situ incorporation of a single nucleotide catalyzed by the E-TP complex on the primer terminus of the cross-linked TP molecule, its accessibility to 3' - 5'-exonuclease activity was examined. Presumably, if cross-linked TP represents its binding in the polymerase mode of the enzyme then the nucleotide incorporated on the covalently linked TP should not be accessible to 3' 4 5'-exonucleolytic cleavage activity of the externally added WT Klenow fragment. The results shown in Table IV clearly indicate that the primer terminus of the E-TP covalent complex is not available for 3' + 5'-exonucleolytic cleavage by the WT Klenow fragment. In contrast, when identical product synthesized by free enzyme was exposed to WT Klenow fragment, near complete removal of 3"labeled nucleotide was evident (Table IV). These results clearly show that the enzyme in the E-TP covalent complex is catalytically active.

Isolation of Klenow Tryptic Peptide Cross-linked to ."'P-La- heled TP-In order to identify the amino acid residues involved

21832 Template-Primer Binding Domain of Pol I TABLE IV

Sensitivity of the product synthesized by E-TP covalent complex to 3‘ - 5‘-exonuclease

Fifty pmol of purified E-TP covalent complex was incubated with [U-~~PITTP as described in Table 111. Following the incorporation of labeled TTP onto the primer terminus of the E-TP complex, unincorpo- rated nucleotide was removed by gel filtration through NAP-10 column pre-equilibrated with Klenow reaction buffer. Accessibility of the incorporated label to 3’ --f 5’-exonucleolytic cleavage was examined by incu- bating an aliquot of the labeled E-TP complex (18,000 cpm) with indicated concentrations of WT Klenow fragment at 37 “C for 30 min. The acid precipitable radioactivity remaining after the cleavage reaction was determined as described in the text. For comparison the labeled product (30,000 cpm) synthesized by the free enzyme under identical conditions was used as a positive control for exonucleolytic cleavage by the externally added WT Klenow fragment.

WT Klenow Cleavage of product synthesized by E-TP covalent

complex fra ent

a g d Free enzyme

nM

None 50

200 100

300 400

90

0 40 70 85 90 98

~ ~ ~~

in DNA binding, Klenow fragment labeled with 32P-labeled TP was subjected to trypsin digestion. Due to the relatively large charge difference between the TP-cross-linked peptides and the free peptides, we used anion exchange chromatography to selectively isolate the 5‘ 32P-labeled TP tryptic peptides (Pandey and Modak, 1988b; Williams and Konigsberg, 1991). Tryptic digestion of the 68-kDa Klenow fragment would be expected to generate 75 tryptic peptides. At or above pH 7, all of the Klenow tryptic peptides except tryptic peptide 4 should have a net charge ranging from -3 to +l. Tryptic peptide 4 has a net negative charge of -7. In contrast, the peptides cross-linked with 25-mer or 37-mer TP should have a net negative charge of at least -24 or -36, respectively. Therefore all of the free peptides may be expected to elute at a lower salt concentration from the anion-exchange column, well before the TP-linked peptides.

The TP-cross-linked peptide was found to be selectively retained on a DEAE-Sephadex column as described under “Experimental Procedures.” All the neutral tryptic peptides, as well as those having a net positive charge, were excluded in the flow-through, while the acidic tryptic peptides, including tryptic peptide 4, were effectively removed from the column by extensive washing with 250 mM TEAA buffer. The tryptic peptide covalently linked with the labeled TP, as well as free TP present in the tryptic digest, was quantitatively eluted from the column at high TEAA concentration (1 M). A portion of the DEAE-Sephadex-purified fraction containing the TP-cross- linked tryptic peptide was analyzed by C4 reverse phase HPLC (Fig. 4). Greater than 70% of the radioactivity applied to the column was recovered from the column. A typical profile of the resolution of DEAE-Sephadex-purified labeled tryptic digest is shown in Fig. 4. Of the recovered radioactivity, approximately 60% of the 32P label was contained in the fraction eluting at 26 min and the remainder was in the fraction eluting at 42 min. Both radioactive peaks exhibited high absorbance at 254 nm indicating the presence of TP. The radioactive material eluting at 26 min showed no associated peptide material upon amino acid composition analysis and was subsequently determined to be free TP. TP alone (irradiated or unirradiated), analyzed on the same C4 reverse phase column, yielded an identical chro- matogram, as shown in Fig. 4, except for the absence of the 42-min peak (data not shown). Furthermore, the tryptic digests ofunlabeled Klenow fragment (irradiated or unirradiated) processed through DEAE-Sephadex, followed by C4 reverse phase

12 36 60

TIME ( MIN)

FIG. 4. C4 HPLC profile of DM-Sephadex-purified tryptic peptide cross-linked with SaP-labeled template-primer. DEAE- Sephadex-purified tryptic peptides covalently linked with 32P-labeled 37-mer TP were applied to a Vydac C4 reverse phase column and eluted at a flow rate of 1 mumin. The profile was developed with a linear gradient of solvent B (70% CH,CN) into solvent A (20 m~ sodium phosphate buffer, pH 6.8) at 0-5 min (0% B), 5-15 min (04% B), 15-130 min (5-30% B). 1-ml fractions were collected, and radioactivity in the individual fraction was determined by Cerenkov counts. Each fraction was monitored for absorbance at 220 and 254 nm on a dual pen recorder attached to a Polychrome 9060 diode array detector. The recording of 254 nm absorbance was slightly shifted to avoid overlaps between 220 and 254 nm absorbance.

HPLC, yielded no peptide peaks (data not shown). Therefore the DEAE-Sephadex-purified peptide material, as well as the fraction from the C4 column eluting at 42 min of the gradient, were considered to represent the peptideb) cross-linked to the 37-mer DNA. Identical results were obtained when 25-mer DNA was substituted for 37-mer DNA in the above experiment (data not shown).

Amino Acid Sequence Analysis of TP-cross-linked W p t i c Peptide-Amino acid sequence analysis of the purified 37-mer, as well as 25-mer, TP-linked peptide revealed it to be a 17- amino acid peptide spanning residues 759-775 in the primary amino acid sequence of pol I (Table V). Examination of the yields of PTH amino acid in each cycle of sequencing revealed that yields of Ile, Tyr, Ser, and Phe corresponding to Ile-765, Tyr-766, Ser-769, and Phe-771 were significantly reduced as compared to the other PTH amino acids (Table IV). Since the amino acid residues involved in covalent cross-linking with the base moiety of the nucleic acid may be expected to be structur- ally altered, their identification in sequencing analysis is based on negative results, i.e. the absence of their identifiable phenylthiohydantoin-derivative at the appropriate cycle in the sequence analysis. The absence of or significant reduction in the yields of PTH-derivatives of Ile-765, Tyr-766, Ser-769, and Phe- 771 during their respective sequencing cycles indicates that these residues are the probable cross-linking points between the TP and the enzyme. Identical amino acid sequencing results were obtained with labeled 25-mer TP (Table v).

DISCUSSION

Klenow fragment of E. coli DNA pol I serves as a model enzyme for the DNA-dependent DNA polymerase class of enzymes. Its crystal structure has been resolved, and local domain structures responsible for catalytic activity, substrate dNTP binding, and template-primer binding have been suggested (Yadav et. al., l992,1994b, Beese et al., 1993). Anumber of chemical modifications and site-directed mutagenesis studies have been carried out with this enzyme, and the functional contribution of various amino acids residues of the Klenow fragment to the overall process of catalysis has been suggested (Pandey and Modak, 1988a; Pandey et al., 1987, 1990, 1993;


TABLE V Amino acid sequencing of tryptic peptide-TP complex isolated by

DEAE-Sephadex column chromatography The tryptic peptide covalently linked with 25-mer TP or 37-mer TP

was selectively purified by DEAE-Sephadex chromatography as described under “Experimental Procedures.” An aliquot of the purified peptide was subjected to gas phase sequencing and the yield of PTH- derivatives of amino acids in each sequencing cycle was recorded. Pico- mole values of each PTH-derivative given above are normalized to 10 pmol of norvaline and have been corrected for percent injection. The numbers shown in the parentheses indicate the position of the amino acid residues in the primary amino acid sequence of pol I.

Yield of PTH-derivatives in each

sequencing Cycle Sequence cycle of tryptic peptide linked with

25-mer TP 37-mer TP

1 2

Ala (759) 88.1 Ile (760)

58.2 82.4

3 Asn (761) 34.3

4 Phe (762) 24.0 14.1

5 47.2 24.3

Gly (763) 53.6 16.7 6 7

Leu (764) 55.9 36.9 Ile (765) 6.4

8 Tyr (766) 3.8

7.1 1.2

10 Met (768) 39.8 22.9 48.3

11 Ser (769) X 12 Ala (770) 36.1

2.6

13 Phe (771) 8.6 8.5 21.6

14 Gly (772) 20.7 29.7 15 Leu (773) 16

20.4 19.8 Ala (774) 16.7 15.5

17 Arg (775) 1.1 3.3

9 Gly (767) 40.3

Basu and Modak, 1987; Basu et al., 1987, 1988; Mohan et al., 1988; Polesky et al., 1990, 1992). In spite of this progress, very little is known about the binding site for the TP. In our effort to identify the overall TP binding site or domain in this enzyme we have used two self-annealing TPs as affinity probes. We find that the use of self-annealing TP is advantageous over the use of two separate complementary oligomeric strands in the cross- linking reaction since the integrity of the TP is maintained and the possibility of a single-stranded oligomeric DNA serving as a ligand in the cross-linking reaction is eliminated. We have demonstrated that pol I protein can be covalently modified by UV- mediated cross-linking of 32P-labeled TP within its binding domain. Under optimal conditions ofphotolabeling, approximately 7-9% of the protein is converted into covalent complex. The specificity of cross-linking of TP to the TP binding site of the Klenow fragment is demonstrated by the observations that the E-TP cross-linking is effectively competed out by poly(dA).(dT),, and poly(dC).(dG),, TPs but is unaffected by single-stranded template or primer DNA (Table I, Fig. 2). Furthermore, in the presence of high salt concentration, no template-primer cross-linking to enzyme is apparent (Table I), which is consistent with the biochemical observation that TP binding to enzyme is severely affected in high salt medium. The most compelling observation, however, is the fact that E-TP covalent complex shows catalytic activity with bound TP. The activity is restricted to the addition of one nucleotide onto the cross-linked TP. The possibility that nucleotidyltransferase activity observed with E-TP complex may be due to free enzyme, which may catalyze the polymerase reaction on TP of the E-TP-cross-linked complex was ruled out based on the following results. 1) Only a single nucleotide could be incorporated quantitatively on the primer terminus of cross- linked TP, suggesting that the enzyme covalently immobilized on the TP molecule could not translocate to add the next nucleotide. 2) The incorporated nucleotide at the primer terminus of the E-TP covalent complex is not accessible to 3’ + 5”exonucleolytic cleavage activity of externally added WT Klenow fragment (Table IV). 3) Nucleotidyltransferase reaction ofE-TP complex can occur in high salt medium (up to 1 M) suggesting that

covalently linked TP remains bound to the enzyme and that dNTP substrate can bind to enzyme and be transferred in high ionic strength media. Under similar conditions, the polymerase activity of free enzyme is undetectable due to the inability of enzyme to bind to TP. 4) The E-TP covalent complex purified by DEAE-Sephadex column was found to be devoid of free enzyme as judged by SDS-PAGE. Furthermore, addition of severalfold excess free enzyme to the purified complex did not increase the extent of nucleotide incorporation on the bound TP compared to E-TP alone (data not shown). 5 ) Upon peptide mapping of the enzyme cross-linked to two different TPs, only a single labeled peptide containing the cross-linked TP was obtained. These results suggest that both 37-mer and 25-mer TPs bind to enzyme in the polymerase mode, resulting in a catalytically active covalent E-TP binary complex.

Anion exchange chromatography of the tryptic digest of TP- cross-linked enzyme was the most effective purification step in the isolation of labeled peptides. The peptide was identified as a 17-amino acid tryptic peptide spanning residues 759-775 in the primary amino acid sequence of pol I. The site of cross- linking is expected to be limited to those amino acids that are at the interface of the Klenow-TP complex. The identification of such amino acids in the TP-linked peptide is made on the basis of amino acid sequencing, where the site of cross-linking is inferred from the missing amino acids in a particular cycle of sequencing. The absence of or reduction of PTH-derivatives of Ile-765, Tyr-766, Ser-769, and Phe-771 in the sequencing of labeled tryptic peptide indicates that these amino acids are the possible contact points in the binding of enzyme to TP. These results also suggest that TP makes relatively few contacts within the binding domain.

Catalan0 et al. (1990) have shown that the terminal nucleotide of primer containing the photoaffinity azido group cross- links specifically with Tyr-766. In the present studies, we find that 4 residues, namely Ile-765, Tyr-766, Ser-769, and Phe-771 are involved in TP binding. Ile-765 and Tyr-766 in the 0-helix may be in close proximity to the primer terminus, and Ser-769 and Phe-771 in the 0,-helix may form the contact points for the template strand.

Recently Beese et al. (1993) reported the x-ray structure of the Klenow fragment with a duplex DNAconsisting of a 12-mer template with an 8-mer complementary primer, which contained an epoxyribo(A) a t its 3’ terminus. However, during crystallization of the complex, rearrangement of TP had oc- curred such that a distorted duplex of B-DNA with a 3-nucleotide (single-stranded) primer overhang was formed. The 3’ overhang of the duplex appeared oriented toward the 3’ +

5’-exonuclease active site thus forming an editing complex of the enzyme rather than the expected polymerase complex. The viable duplex DNA portion in this structure was found to bind in the cleft, which runs at nearly a right angle to the originally described large cleft containing the polymerase active site (01- lis et al., 1985). A nearly identical position for DNA binding has also been inferred by Yadav et al., (1994) based on a revised structure generated by the computer assisted modeling and electrostatic potential studies. Steitz and colleagues have postulated that the binding position for the duplex part of the DNA remains constant for the editing as well as the polymerase complex but that the non-complementary extension of primer moves from the polymerase to the exonuclease site. Since contact points between the Klenow fragment and the duplex region of DNA are expected to occur through interaction with the DNA phosphate backbone, the photochemically induced covalent linkage of TP to a region containing Ile-765, Tyr-766, Ser-769, and Phe-771 may represent the binding domain for single- stranded template. This region constitutes part of 0- and 0,-

21834 Template-Primer Binding Domain of Pol I

helices and forms the wall of the cleft opposite the thumb sub- domain, where the template strand has been postulated to bind (Beese et al., 1993).

Site-directed mutagenesis of a t least 2 residues, Lys-758 and Tyr-766, present in the 0-helix region has been carried out, and the mutations seem to affect polymerase activity (Polesky et al., 1992; Pandey et al., 1994). The major defect in both mutant enzymes is found to be their inability to catalyze DNA synthesis in a processive manner. Furthermore, synthesis directed by poly(dA) but not the poly(dC) templates seems to be severely affected in both mutants (Desai et al., 1994). Since the 2 residues are located in the vicinity of each other and their side chains are oriented toward the cleft, it seems likely that it is not the individual residue but a region spanned by them that may constitute the TP interactive region. Investigation of site-directed mutagenesis ofresidues between Lys-758 and Tyr-766 and those between Tyr-766 and Phe-771 may be helpful in illustrating the putative role of the 0-helix in the binding of and translocation across the template strand. Further identification of contact points between the cross-linked residues and corresponding nucleotides of the template primer together with definition of the duplex binding region would provide real directionality of TP binding that has remained obscure and speculative in spite of the availability of three-dimensional structure of this enzyme.

REFERENCES

Basu, S., Basu, A., and Modak, M. J . (1988) Biochemistry 27, 67104716 Basu, A,, Basu, S., Ahluwalia, K., and Modak, M. J. (1992) Biochemistry 31,

616422

Beese, L. S., Derbyshire, V., and Steitz, T. A. (1993) Science 260, 352355 Catalano, C. E.,Allen, D. J., and Benkovic, S . J. (199O)Biochemistry 29,3612-3621 Derbyshire, V., Freemont, P. S., Sanderson, M. R., Beese, L., Friedman, J. M.,

Desai, S. D., Pandey, V N., and Modak, M. J. (1994) Biochemistry, in press Joyce, C. M., Ollis, D. L., Rush, J., Steitz, T. A,, Konigsberg, W. H., and Grindley,

Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A,, and Steitz, T. A. (1992)

Maxam, A. M., and Gilbert, W. (1980) Methods Enzymol. 65,499-560 Ollis, D. L., Brick, P., Hamlin, R., Xoung, N. G., and Steitz, T. A. (1985)Nature 313,

Pandey, V. N., and Modak, M. J. (1988a) J. Biol. Chem. 263,6068-6073

Pandey, V. N., and Modak, M. J. (1988~) J. Biol. Chem. 263,3744-3751 Pandey, V. N., and Modal, M. J. (1988b) J. Chromatogc 444, 133-139

Pandey, V. N., Williams, K. R., Stone, K. L., and Modak, M. J. (1987) Biochemistry

Pandey, V. N., Kausbik, N., Pradhan, D. S., and Modak, M. J . (1990) J. Biol. Chem.

Pandey, V. N., Kaushik, N., Sanzgiri, R. P., Patil, M. S., Modak, M. J., and Barik,

Pandey, V. N., Kaushik, N., and Modak, M. J. (1994) J. Biol. Chem. 269, 13259-

Polesky, A. H., Steitz, T. A,, Grindley, N. D. E, and Joyce, C. M. (1990) J. Biol.

Polesky,A. H., Dahlberg, M. E., Benkovic, S. J., Grindley, N. D. F., and Joyce, C. M.

Rush, J., and Konigsberg, W. H. (1990) J. Biol. Chem. 265,48214827 Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Labo-

ratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Joyce, C. M., and Steitz, T. A. (1988) Science 240, 199-201

N. D. F. (1986) UCLA Symp. Mol. Cell. Biol. 32, 197-205

Science 256, 1783-1790

762-766

26,7744-7748

265,36793684

S. (1993) Euc J. Biochem. 214,59-65

13265

Chem. 265,14579-14591

(1992) J. Biol. Chem. 267,84174428

Williams, K. R., and Konigsberg, W. H. (1991) Methods Enzymol. 208, 516-539 Tirumalai, R. S., and Modak, M. J. (1991) Biochemistry 30, 643M443

Yadav, P. N. S., Yadav, J. S., and Modak, M. J. (1992) Biochemistry 31,2879-2886 Yadav, P. N. S., Yadav, J. S., Arnold, E., and Modak, M. J. (1994a) J. Biol. Chem.

Yadav, P. N. S., Yadav, J. S., and Modak, M. J. (199413) J. Mol. Recognit., in press 269,716-720

Photoaffinity Labeling of DNA Template-Primer Binding Site in ...

Documents

Transcript of Photoaffinity Labeling of DNA Template-Primer Binding Site in ...