Recognition by Viral and Cellular DNA Polymerases of Nucleosides

Proc. Natl. Acad. Sci. USA Vol. 92, pp. 6329-6333, July 1995 Biochemistry

Recognition by viral and cellular DNA polymerases of nucleosides bearing bases with nonstandard hydrogen bonding patterns [human immunodeficiency virus type 1 reverse transcriptase/DNA polymerases a, 13, and e/2'-deoxy-N1-methyloxoformycin B/ 2'-deoxyxanthosine/2,4-diamino-5-(3-D-2'-deoxyribofuranosyl)pyrimidineI JENNIFER HORLACHER*, MICHAEL HOThIGERt, VLADIMIR N. PODUSTt, ULRICH HOBSCHERt, AND STEVEN A. BENNER* *Bio-Organische Chemie, Eidgenossiche Technische Hochschule Zurich, Universitatstrasse 16, CH-8092 Zurich, Switzerland; and tDepartment of Veterinary Biochemistry, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland Communicated by Peter B. Dervan, California Institute of Technology, Pasadena, CA, April 3, 1995 (received for review September 22, 1994)

ABSTRACT The ability of DNA polymerases (pols) to N=\ N=\ N D H§ N Acceptor Hs ,H Acceptor catalyze the template-directed synthesis of duplex oligonucle- Donor Donor H, N l Donor otides a Watson-Crick base pair HN> N Donor N NH containing nonstandard Acceptor Acceptor N H NAN H between a nucleotide bearing a 5-(2,4-diaminopyrimidine) l[ , N s Donor Acceptor H ~~Donor Donor heterocycle (dKc) and a nucleotide bearing either deoxyxan- Acceptor N O guanineguanH thosine (dX) or N'-methyloxoformycin B (w) has been inves- cytosine K 'H H N-N x tigated. The K-X and IC-T base pairs are joined by a hydrogen H ' NyNw _ Acceptor TT Donor bonding pattern different from and exclusive of those joining H NS\ N_ Donor H N II Acceptor the AT and GC base pairs. Reverse transcriptase from human H a immunodeficiency virus type 1 (HIV-1) incorporates dXTP Acceptor 0 Acceptor it Donor t H into an oligonucleotide opposite dc in a template with good N y H ~~~~Donor ,N N N- Donor fidelity. With lower efficiency and fidelity, HIV-1 reverse Acceptor NllHo amino Acceptor O H X¶ *y thymine I adenine N N Acceptor transcriptase also incorporates d#cTP opposite dX in the Donor N OH y template. With dI in the template, no incorporation of dicTP 0 Acceptor H N=\ Donor N N H was with HIV reverse transcriptase. The Klenow MN-_ observed Donor H fragment of DNA pol I from Escherichia coli does not incor- Acceptor 0 Donor H N N-... N porate d#cTP opposite dX in a template but does incorporate Acceptor KN Y Donor H N.H Acceptor dXTP opposite dKc. Bovine DNA pols a, 1, and c accept neither Acceptor N N Acceptor Donor kNAN H Donor ,N H dXTP opposite dK< nor dKTP opposite dI. DNA pols a and E iso-guanine -,N Donor (but not .8) incorporate d#cTP opposite dX in a template but iso-cytosine Hi Acceptor IJ H'N'H discontinue elongation after incorporating a single additional base. These results are discussed in light of the crystal FIG. 1. Six base pairs that meet the constraints imposed by the structure for pol 13 and general considerations of how poly- Watson-Crick base pairing geometry. merases must interact with an incoming base pair to faithfully copy genetic information. replicatable bases joined in 6 base pairs by mutually independent hydrogen bonding patterns, provided that some bases Recognition between base pairs in complementary oligonu- were joined to the sugar via a carbon-carbon bond (a "C cleotide strands is mediated by two rules of complementarity: glycoside") (Fig. 1). We then asked whether natural DNA and size complementarity (large purines pair with small pyrimi- RNA polymerases (pols) could catalyze the template-directed dines) and hydrogen bonding complementarity (hydrogen incorporation of these pairs into duplex DNA. If so, this could bond donors are matched with acceptors). Hydrophobicity and increase the number of amino acids that might be encoded by planarity in the bases are also believed to be important for the a mRNA (4), provide nonstandard sequence "tags" on oligo- stability of the double helical structure. Hydrophobicity allows nucleotides used in diagnostic applications, expand the versa- transfer of the bases from water to the core of the double helix tility of experiments for the in vitro evolution of nucleic acids to make an energetic contribution to duplex formation. Pla- (5-7), and assist efforts to obtain self-replicating oligonucle- narity allows sequence-independent stacking of the bases to otides as models for the origin of life (3). form an aperiodic crystalline structure necessary for duplex Previous work has shown that pols can accept the isocy- stability. tosine-isoguanine base pair (8-10), where both the purine and Even with these constraints, considerable structural flexi- the pyrimidine analogs are joined to the sugar ring by a bility remains within the Watson-Crick base pair. For example, carbon-nitrogen bond. The Klenow fragment of DNA pol I Rich (1) pointed out in 1962 that isocytosine, which presents also incorporates deoxyxanthosine (dX) (11), a purine pre- a hydrogen bond acceptor-acceptor-donor pattern to the senting a hydrogen bond acceptor-donor-acceptor pattern to complementary strand, might fit the Watson-Crick geometry a complementary strand, opposite 2,4-diamino-5-(13-D-2'- when paired with isoguanine (Fig. 1).t Zubay (2) proposed a deoxyribofuranosyl)pyrimidine (dK), a pyrimidine presenting base pair joined by a nonstandard hydrogen bonding pattern, although in a ring system that was not planar and therefore Abbreviations: pol, DNA or RNA polymerase; HIV-I, human immu- presumably not able to stack. nodeficiency virus type 1; BSA, bovine serum albumin; DTT, dithio- Some time ago, we noted (3) that the Watson-Crick for- threitol; M, marker; P, primer. malism could readily be extended to include 12 independently tIn dK, the heterocycle is joined to the sugar via a C-C bond. A crystal structure of ribo-K as a hydrochloride (not shown) (49) shows that the sugar ring of ribo-K exists in a pronounced C2'-endo conformation, The publication costs of this article were defrayed in part by page charge as in DNA. This suggests that the C-C bond need not by itself payment. This article must therefore be hereby marked "advertisement" in exclude dK as a substrate; other studies suggest that C-glycoside accordance with 18 U.S.C. §1734 solely to indicate this fact. linkages need not be intrinsically problematic for a pol (12). 6329 Downloaded by guest on October 1, 2021 6330 Biochemistry: Horlacher et al Po.Nt.Aa.SiProc. Natl. Acad. Sci. USAS 922(95(1995) the complementary donor-acceptor-donor hydrogen bonding amide gel (16 cm X 42 cm X 0.6 mm) containing 7 M urea. pattern. However, pots that accept triphosphates of C glyco- Following electrophoresis (50'C at 1500-1800 V), the gets sides have proven difficult to find. Recent work in these were fixed, dried,. and autoradiographed. (11-13) and other (14) laboratories has suggested that C nucleosides might generally form weaker base pairs and might be generally poor substrates for pots. We therefore undertook RESULTS a more systematic study of the enzymology of base pairing Klenow Fragment of DNA Pol I from E. coli. When the between 2,4-diamino-5-(13-D-2'-deoxyribofuranosyl)pyrimidine Klenow fragment of pot I was incubated at 230C with templates (dK, a C glycoside) and either deoxyxanthosine (dX) or 2'-deoxy- containing the nonstandard bases using enough pot to elongate 1-methyloxoformycin B (dir) (Fig. 1), focusing on viral and alt of the in 20 no of either dKTP or cellular DNA primer min, incorporation pots (15). dXTP opposite to its counterpart in a template could be detected (Fig. 2a). Elongation of the primer stops just before EXPERIMENTAL incorporation opposite the nonstandard base. With a 40-fold Synthesis of Nucleosides Bearing Nonstandard Bases. N'- a Methyloxoformycin B (i7r) was prepared by the procedure of Template sequence: Primer-GC'GGXCCGA-5' Primner-GCGGinCCG;A-5' P1rimez, GCGGicCCGA-5' Ugakar et at (16). 2,4-Diamino-5-(13-D-ribofuranosyl)pyrimi- Product i f i dine (K) was synthesized by the route of Chu et at (17). These were converted to 2'-deoxygenated nucleoside analogs as 26 described by Piccirilli (18) and to 5'-dimethoxytritylated f3-cya- 25 G noethyl phosphoramidites by standard methods (19). 5'- 24 G with both 23 rs/X Dimethoxytrityl-2'-deoxyxanthosine heterocyclic ~~~~~~~~~~~~~~~~~~22 C ~~~~~~~4i 4~~~~~~.....*...... ~ ring oxygens protected as p-nitrophenylethyl ethers was pre- ...... 2...... pared following Van Aerschot et aL (20) and converted to the phosphoramidite. The triphosphates dKTP and dXTP were prepared following literature procedures (21, 22). Standard AAA K...... K dNTPs were from Pharmacia. Oligonucleotides were prepared .X ... by solid-phase synthesis (Applied Biosystems) from 13-cyano- 2b Templatesequence: Primer-GCGGXCCGA-5'Primer-G.G.....A...P...... ethyl-protected phosphoramidites, purified by the "trityl on" Product~~~~~~~~~~~~~~~~~...... ' ":~P procedure, deprotected, and purified again by HPLC (23). :~::~~~~~:::::fff~~;:::::~:~~~~~:::~:~ :~~;::::~:~~~~: .;: :::::::::: ~ -: -:::::::. ~ Primer (18 bases, 5'-TAATACGACTCACTATAG-3') was labeled at the 5' end with [-y_33P]ATP (Amersham) using T4 polynucle*itide kinase (24). Enzymes, Proteins, and Chemicals. Human immunodeficiency type 1 (HIV-1) reverse transcriptase overexpressed the in Escherichia coli was a 26 C using plasmid pJS3.7 purified by 25 G published procedure (25). Calf thymus DNA pots a, /3, and s 24 G were purified according to the methods of Podust et al (26), 23 sIX and Weiser et at The Klenow 22 C Chang (27), (28), respectively. 21 C fragment of pot I (sequencing grade) was from Boehringer 20 G Mannheim. DNase-free bovine serum albumin (BSA) was 19C from New England Biolabs. Other reagents were of analytical Primer grade from Fluka, Merck, or Aldrich. U U U U C* U U U C. UU U U Nucleic Acids. The G G G G G* G G G G G000 G template (5'-AGCCQGGCGCTAT- T T T' T T T T AGTGAGTCGTATTA, 27 bases, Q is a nonstandard nudle- A A* A A otide) was annealed with primer in 750 jil of total buffer [50 x XI pmol of template and 15 pmot of labeled primer, in 1.8 mM 0.5 MM and 23 mM c Tris-HCI (pH 7), containing MgCl2 NaCI] Template sequence: Primcr-GCGGXCCGA-5' Priner-GCGGerCCGA-5' Primner-GCGGKCCGA-5' by cooling from 750C to the assay temperature over a period Product tI I of 1 hr. 27 T DNA Pol Assays. Pot assays were performed in a final 26 C volume of 20 Al the for a, 20 mM 25 G containing following: pot 24 G potassium phosphate (pH 7.2), 0.1 mM EDTA, 4 mM dithio- 23 Kt/X threitol (DTT), 0.25 mg of BSA per ml, and 10 MM MgCt2; for 22 C 3, 40 mM Tris-HCI 1 mM DTT, 0.2 mg of BSA 21 C pot (pH 7.5), 20 G per ml, and 10 MM MgCl2; for pot E, 20 mM Tris (pH 7.2), 0.5 19C mM DTT, 0.1 mg of BSA per ml, 10 MM MgCl2, and 10 mM Primer MgCt2; for HIV-1 reverse transcriptase, 50 mM Tris-HCI (pH Mc c c c c- c c c c c- c c c c c- 7.2), 5 MM MgCI2, 80 mM KCI, 1 mM DTT, and 0.5 mM 0 G G G G* G G 0 G G* G G G G G* for Klenow of 50 mM Tris-HCI TT T T T T T T T EDTA; fragment pot I, (pH A A* A A A A* 7.5), 10 MM MgCt2, 0.1 mg of BSA per ml, and 1 mM DTT. (K ~~K K X X* The amount of enzyme used and the concentrations of templates, primers, and nucleoside triphosphates are specified in FIG. 2. Primer extension by the Klenow fragment of DNA pol I the figure legends. One unit of pot is defined as the amount of from E. coli. Triphosphates present are indicated below. M indicates enzyme that incorporates 1 nmot of total dNTP per hour at markers (primer, full-length product). (a) Deoxynucleoside triphosphates (2 .tLM; 50 ttM when indicated by *) incubated at 230C for 20 370C. min with primer/template (as 3'-OH primer end, 0.12 pmol) and 0.15 The reactions were quenched by adding a formamide (3-5 unit of the Klenow fragment of pollI in a final volume of 20 pil. (b) As gl) containing xylene cyanot (0.3%), bromphenol blue (0.3%), in a but with 0.8 unit of the Klenow fragment of pot I and incubation and EDTA (0.37%), the samples were heated for 10 min at for 30 min (c) As in a but with 6.0 units of the Klenow fragment of 750C, and aliquots (3-5 gl) were loaded onto a 13% polyacryl- pollI and incubation for 30 min. Downloaded by guest on October 1, 2021 Biochemistry: Horlacher et at Proc. Natl. Acad. Sci. USA 92 (1995) 6331 increase in pol concentration, no incorporation of dKTP was good yields and high selectivity. However, the pol still clearly observed opposite it in the template (Fig. 2c); nor could recognizes dXTP as foreign, as considerable chain termination convincing evidence be found for the incorporation of dKTP occurred immediately before incorporation of dXTP and, at opposite dX. In both cases, incorporation of dKTP and further concentrations of dXTP where pausing is eliminated, after elongation were observed, but control experiments suggested incorporation of the dX-dK base pair into the duplex product. that these might reflect misincorporation of a standard base With templates containing dX, a small amount of fully opposite dX. An additional band (Fig. 2c) obtained with very extended product is observed at low concentrations (2 ,uM) of high concentrations of dKTP and enzyme suggested that dKTP dKTP. Substantial amounts of full-length product are observed might be very inefficiently incorporated. at 50 ,uM dKTP when the standard dNTPs remained at low The Klenow fragment did, however, incorporate dXTP concentration (2 ,uM). Misincorporation of the standard opposite dK in a template with increasing enzyme concentra- dNTPs (50 ,uM) also yielded full-length products (Fig. 3). tions (Fig. 2 a-c), and increasing dXTP concentrations (to 50 Thus, the discrimination between dK and the standard dNTPs ,tM). Incorporation of dXTP is selective, as similar elongation opposite dX in a template is not particularly good. However, was not observed in the absence of dXTP with the four it is interesting to note that HIV-1 reverse transcriptase does standard bases. A substantial amount of product missing the not pause significantly after incorporation of dKTP opposite final base was also observed. This is frequently seen with fill-in dX in the template, in contrast to the pause observed after the experiments using the Klenow fragment. These results confirm incorporation of dXTP opposite dK. Thus, although dKTP is and extend results obtained by Piccirilli et at (11). incorporated with lower selectivity than dXTP by HIV-1 HIV-1 Reverse Transcriptase. HIV-1 reverse transcriptase reverse transcriptase, it is incorporated with higher overall incorporated dXTP opposite dK in a template at low concen- efficiency. trations of enzyme and low concentrations of dXTP (Fig. 3a). Experiments at higher concentrations of enzyme showed the A product with a single base incorporated opposite dK and a selectivity of incorporation of dXTP opposite dK in the smaller amount of fully extendible product were seen. No template. With dGTP and dCTP alone, primer extension stops misincorporation of the standard dNTPs could be detected just before the dK in the template without substantial misin- under similar conditions. At 25-fold higher concentrations of corporation. Further, neither dT nor dA was misincorporated dXTP, primer elongated by addition of dXTP was the major in the absence of dX at 2 ,uM. Even at high (25 ,uM) product, with substantial amounts of full-length product also concentrations of each of the four standard dNTPs, only a observed. Higher concentrations of standard dNTPs yielded trace of full-length product was observed. only a trace of misincorporation. Thus, HIV-1 reverse tran- CalfThymus DNA Pols a, fB, and c. These three pols all did not scriptase incorporates dXTP opposite dK in a template with incorporate dXTP opposite dK in a template (Fig. 4 b and d). Furthermore, no misincorporation was observed. Similarly, nei- a ther incorporation of dKTP nor misincorporation was observed Temiiplate sequence: P'rimiecr-(;C(GGXCCGA-5' Prircrir-c;(;;n'C(;A-5' Primer-GCG(;KCCGA-5' opposite iT (Fig. 4a). With templates containing dX, the pols Proluct -- 1 F I sequlence: 27 T- behaved differently. No incorporation of dKTP was detectable se Len 27 2626CC ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.* with pol ,B, even at 50 ,uM concentrations. Pol a incorporated 24 G.... dKTP opposite dX in a template (Fig. 4c). Primer extension stops, ...... however, after incorporation of a single additional base following ...... 22 ....IC AA ... 22IC21 C ***r dK, with substantial amounts of pausing. In the presence of dCTP 20 G and dGTP alone with pol a, primer extension was terminated at amount of misin- 19 ...... the nonstandard base. A small but significant ... corporation opposite dX was observed only at high concentra- c c c c c* c c c c C* c c c c C* GG G G G' G G G G G* G G G G G* tions of standard dNTPs (50 ,uM each). With pol s (29-32), a T T T* T T T* T T T* similar but not identical behavior was observed (Fig. 4d). Incu- A A* AA A A* bation in the presence of dKTP, dCFP, and dGTP showed K* K IC' x xI incorporation of a base opposite dX in the template and contin- ued elongation by only a single additional base. Unlike with pol b Templale seqeCnlC'C: Pritiei- -(((;GGCC'CGA-5' a, however, less pausing was observed after dKTP was incorpo- Pr-oduct rated. The amount of dK incorporated increased at higher con- seqttcnce: 27 T centrations of dKTP.

DISCUSSION 23 The ability of pols to accept the dK-dX nonstandard base pair 22 C that all pols 21 is remarkably variable, especially considering

20G appear to be homologous (judging by the common three- dimensional fold in those studied). In some cases, dKTP was 19 C not incorporated. In others, difficulties in polymerization were Primner ,- encountered only after the incorporation of dKTP, as if the pol C C C C C C C G G G G G G* G rejected dK in the primer. dXTP was easily incorporated by T T T T* some polymerases and firmly rejected by others. During the A* A x x X- divergent evolution of the polymerase superfamily, features in the active site that make contact with the bases in the template, FIG. 3. Primer extension by HIV-1 reverse transcriptase. Triphos- primer, and triphosphate must have diverged considerably, and phates present are indicated below. (a) Deoxynucleoside triphosphates these interactions (rather than an intrinsic property of the (2 ,uM; 50 ,uM when indicated by *) incubated at 37°C for 30 min with nonstandard nucleosides themselves) must be the primary primer/template (as 3'-OH primer end, 0.13 pmol) and 0.12 unit of determinant of the outcome of a replication experiment with HIV-1 reverse transcriptase in a final volume of 20 ,ul. (b) De- oxynucleoside triphosphates (2 ,uM; 25 ,uM when indicated by *) a nonstandard base. incubated at 37°C for 45 min with primer/template (as 3'-OH primer The ability of pols to accept a nonstandard base does not end, 0.08 pmol) and 0.37 unit of HIV-1 reverse transcriptase in a total correlate cleanly with overall specificity. Pol ,B is the least volume of 20 ,lI. selective of the pols examined here. Indeed, pol 13 is so efficient Downloaded by guest on October 1, 2021 6332 Biochemistry: Horlacher et at Proc. Natl. Acad. Sci. USA 92 (1995) b Template sequence: Prinier-GCGGKCCGA-5' c a Polymerase: teemplate sequLenlce: Primer-GCGGItCCGA-5' ProduLct a D Template sequence: Primer-GCGGXCCGA-5' Polyvicrase: sequence: Polymerase: ae ProdLuCt ProduLct __ 27 T F sequence: I t-I sc(tience: 26C 27 T 27TT 25 G 26 C 26 C ...... 24 25 G ...... 25 G 24 G 23 X ...... 24 ...... I .. 24 23KK ...... J ... :: :. ,J22 .N..M...... _ .. 22 Cr ...... _.*.. .~ 21 C ...... 21 C 20 G ...... 2.I. 20 G 20 G ...... :.. 19 C 1 9 C ...... ' ..... Primer _., ...... r...... zaPrimer ._~~~~~~~~~~~~~~~...... ~~~~~~~~...... NM Pr C C C C C C CCC M C C* N. l1 Pr c c c C* c C C C* C C C* C* GG G G GG G* G G G G* G G G G* G G G G* G G G G* T T T* T T* T T* T T* T T* A* A* A* A' A*

x X.. x X.. K K* K- K* d Ftc. 4. Primer extension by calf thymus DNA pols aY, 3, and Triphos- hates presenit are indicated below. (fi) Deoxynucleoside (2 FM: Template sequcnce: 1riiiic'-tiC(iX'XC('t(iA-5 Priilmea(,( ((;CrtCCGA-5' Primrier-GCG(;KCC(XiA-5' triphosphatcs 5 jM whcn indicated by I()()M wheni indicated by when pol a was Product 100; 2 T F I sted; I (,uM; 25 ,uM when indicatcd by I00 AM whctn indicated by whiell SC(ILIC(lCe: _ 2-)7t TI ~~~~~~~~~~~~~~~~~~~~~.p 26C ol wcas tested) incubated at 370C for 30 min with prinicr/template 25 C) tntain ing r (as 3'-OH primer cnd, (). 17 pmol when polrw stested; 0.15 pmol 24 G hen pol was tested). The amounts of pol oa and pol B tcsted in a final volume 23 K/X 20 Al were 0.24 unit and 6.6 uniits, respectively. (b) As in ta but with a 22CC ....* . 21 C mplite/pr-inmrer(ntaininigidK. (c ) De)oxynucleoside triphosphaltes (2 uM 5t() 20 G .M when indicated by when pol (x was tested: I ,M; 5t0 MM when indicated 19C) was incuibated at 30 with c A. y* when pol tested) 37C {or min primcr/tcmplate Pm9 .,-f.. ...::::::...-i"; ...... A*k _-I..._.::"6;~~!_. L ^... Z Primler ^ih I'll._ a)nttaining X (as 3-OH primcr cnd, 0.15 pnmol) and either- pol cy (0.16 unit) C c C C- C C C C* C C C C* M r pol [3 (4.4 units) in a final volumc of 20 M1l. (cl) Deoxynucleoside G G G G* G G G G* GG G G* *iphosphates (2 AM: 50 when indicated by ) incubated with pol aE (0.007 T T* T T* T T* AM A' A' A* nit) at 37TC for 30 min withi primer/template (as 3-Ol-1 primcr end, t).16 K K.: X X* nmol) in a final volume of 20 M'.

at incorporating mismatches into duplex DNA that early information in a product (DNA or RNA). The pol must choose speculation suggested that it required no template (33), a the appropriate sugar, enforce a Watson-Crick geometry (as speculation reinforced by its homology with terminal de- opposed to a wobble geometry, for example) (47), and provide oxynucleotidyltransferase. Yet pol is extremely discriminat- most of the selective binding to the pentacoordinated phosphate ing with the nonstandard dK-dX base pair. This discrimination in the transition state to catalyze the reaction. However, the cannot be explained by proofreading (34-36), as pol X3 does not template, not the pol, should choose the pattern of hydrogen bond have a proofreading 3' -> 5' exonuclease activity. The dis- donor and acceptor groups in the incoming base once the crimination must occur during the elongation reaction itself. Watson-Crick geometry is imposed. Any preference that a pol Direct contacts between the bases and the side chains of itself might have for one of the four natural base pairs would residues in pol ,B may be invoked to explain the extreme decrease its overall efficiency and fidelity. specificity of the enzyme. Pelletier et at (37) report that In this view, the pol should make contacts with the bases at Lys-234 in pol ,B forms a hydrogen bond to the lone pair of only those positions where the standard bases have similar electrons on oxygen-2 of a cytosine in the template, Tyr-271 structural features, for example, with the lone pairs of elec- a a in forms hydrogen bond to the 0-2 lone pair of cytosine the trons on 0-2 of pyrimidines and N-3 of purines. The crystal forms a bond to a lone on N-4 primer, and Arg-283 hydrogen pair structure of pol , suggests that these contacts, and no others, of a in the These lone of electrons are guanine template. pairs are made between the enzyme and the bases (37). significant, as they are the only structural feature presented in This view cannot, however, be general. The Klenow frag- common to the grooves of duplex DNA by all four base pairs (AT, ment accepts X as a triphosphate and K as a template. Both TA, GC, and CG). This lone pair is replaced by an N-H bond the common lone of N-3 of in both dK and dX. While Tyr-271 might be ambivalent (accepting replace pair (0-2 pyrimidines, an NH It is hard to see what Klenow be both hydrogen bond donor and acceptor groups) at this position purines) by group. might identifying that allows it to discriminate against dK (but not dX) in the oligonucleotide, neither Lys-234 nor Arg-283 should tol- erate this replacement. Accordingly, site-directed mutagenesis as a triphosphate, or against dX (but not dK) in the template, experiments targeted against these two residues might yield without creating an interaction that will disrupt the neutrality of mutant forms of pol ,B that accept dK and dX. the pol with respect to the four standard base pairs. This becomes that Crystal structures for other pols are available (38-40), but doubly problematic in light of recent results that show without a full complement of bound substrates. Thus, specific Klenow accepts isocytosine in the template and as a triphosphate, residue-substrate interactions that might explain specificity in although with some bias against isocytosine (8). Klenow cannot these enzymes are difficult to assign, especially as conforma- be searching rigorously for the 0-2 lone pair (for pyrimidines) or tional changes evidently occur upon formation of the active the N-3 lone pair (for purines). complex (41-44). For these pols, site-directed mutagenesis A final point must be made concerning HIV-1 reverse tran- (45, 46) should, in combination with studies using nonstandard scriptase, which accepts a wide range of substrates, including bases, prove valuable in interpreting crystallographic informa- nucleoside triphosphates modified on the sugar (48). There is no tion regarding these interactions. reason a priori why tolerance of modified sugars should fore- The idiosyncracies displayed by pols are also surprising in light shadow tolerance of nonstandard base pairs. That HIV tolerates of the physiological role that a pol must play. Pols must faithfully variation in both is consistent, however, with the common view convert information in a template strand (DNA or RNA) into that the active site in HIV-1 reverse transcriptase is "looser" than Downloaded by guest on October 1, 2021 Biochemistry: Horlacher et aL Proc. Natl. Acad. Sci. USA 92 (1995) 6333 in other pols. The implications of this for the design of inhibitors 23. Vogel, J. (1994) Ph.D. dissertation (Eidgenossiche Technische for HIV reverse transcriptase are evident. Hochschule, Zurich). 24. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular We thank A. Muller for the help with nucleoside synthesis and M. Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Haugg for discussion. This work was supported by the Swiss National Plainview, NY). Science Foundation (31-28592.90 to U.H.; 31-36713.92 to M.H.). M.H. 25. Hafkemeyer, P., Ferrari, E., Brecher, J. & Hubscher, U. (1991) is a recipient of a Swiss National Science Foundation grant for the Proc. Natl. Acad. Sci. USA 88, 5262-5266. M.D.-Ph.D. program. 26. Podust, V. N., Mikhailov, V., Georgaki, A. & Hubscher, U. (1992) Chromosoma 102, 133-141. 1. Rich, A. (1962) in Horizons in Biochemistry, eds. Kasha, M. & 27. Chang, L. M. S. (1973) J. Biol. Chem. 248, 3789-3795. Pullmann, B. (Academic, New York), pp. 103-126. 28. Weiser, T., Gassmann, M., Thommes, P., Ferrari, E., Hafkem- 2. Zubay, G. (1988) in The Roots of Modern Biochemistry, eds. eyer, P. & Hubscher, U. (1991)J. Biol. Chem. 266, 10420-10428. Kleinkauf, H., von Doehren, H. & Jaenicke, L. (de Gruyer, New 29. Hubscher, U. & Thommes, P. (1992) Trends Biochem. Sci. 17, York), pp. 911-916. 55-58. 3. Benner, S. A., Allemann, R. K., Ellington, A. D., Ge, L., Glas- 30. Campbell, J. L. (1993) J. Biol. Chem. 268, 25261-25264. feld, A., Leanz, G. F., Krauch, T., MacPherson, L. J., Moroney, 31. Podust, V. N. & Hiubscher, U. (1993) Nucleic Acids Res. 21, S., Piccirilli, J. A. & Weinhold, E. (1987) Cold Spring Harbor 841-846. Symp. Quant. Biol. 52, 53-63. 32. Jessberger, R., Podust, V. N., Hubscher, U. & Berg, P. (1993) J. 4. Bain, J. D., Chamberlin, A. R., Switzer, C. Y. & Benner, S. A. Biol. Chem. 268, 15070-15079. (1992) Nature (London) 356, 537-539. 33. Green, R. & Korn, D. (1970) J. Biol. Chem. 245, 254-260. 5. Irvine, D., Tuerk, C. & Gold, L. (1991)J. Mol. Biol. 222,739-761. 34. Muzyczka, N., Poland, R. L. & Bessman, M. J. (1972) J. Biol. 6. Szostak, J. W. (1992) Trends Biochem. Sci. 17, 89-93. Chem. 247, 7116-7122. 7. Giver, L., Bartel, D. P., Zapp, M. L., Green, M. R. & Ellington, 35. Fersht, A. R. & Knill-Jones, J. W. (1981) Proc. Natl. Acad. Sci. A. D. (1993) Gene 137, 19-24. USA 78, 4251-4255. 8. Switzer, C. Y., Moroney, S. E. & Benner, S. A. (1989) J. Am. 36. Kunkel, T., Schaaper, R. M., Beckman, R. A. & Loeb, L. A. Chem. Soc. 111, 8322-8323. (1981) J. Biol. Chem. 256, 9883-9889. 9. Switzer, C. Y., Moroney, S. E. & Benner, S. A. (1993) Biochem- 37. Pelletier, H., Sawaya, M. R., Kumar, A., Wilson, S. H. & Kraut, istry 32, 10489-10496. J. (1994) Science 264, 1891-1903. 10. Tor, Y. & Dervan, P. B. (1993)J. Am. Chem. Soc. 115,4461-4467. 38. Jacob-Molina, A., Ding, J. P., Nanni, R. G., Clark, A. D. & Lu, 11. Piccirilli, J. A., Krauch, T., Moroney, S. E. & Benner, S. A. (1990) X. D. (1993) Proc. Natl. Acad. Sci. USA 90, 6320-6324. Nature (London) 343, 33-37. 39. Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A. & Steitz, 12. Piccirilli, J. A., Moroney, S. E. & Benner, S. A. (1991) Biochem- T. A. (1992) Science 256, 1783-1790. istry 30, 10350-10356. 40. Davies, J. F., Almassy, R. J., Hostomska, Z., Ferre, R. A. & 13. Voegel, J. J. & Benner, S. A. (1994) J. Am. Chem. Soc. 116, Hostomska, Z. (1994) Cell 76, 1123-1133. 6929-6930. 41. Wong, I., Patel, S. S. & Johnson, K. A. (1991) Biochemistry 30, 14. Rosenberg, I., Soler, J. F., Tocik, Z., Ren, W.-Y., Ciszewski, 526-537. L. A., Kois, P., Pankiewicz, K. W., Spassova, M. & Watanabe, 42. Kati, W. M., Johnson, K. A., Jerva, L. F. & Anderson, K. S. K. A. (1993) Nucleosides Nucleotides 12, 381-401. (1992) J. Biol. Chem. 267, 25988-25997. 15. Hubscher, U. & Spadari, S. (1994) Physiol. Rev. 74, 259-304. 43. Tanabe, K., Bohn, E. W. & Wilson, S. (1979) Biochemistry 18, 16. Ugakar, B. G., Revankar, G. R. & Robins, R. K. (1984) J. 3401-3406. Heterocycl. Chem. 21, 1865-1870. 44. Mizrahi, V., Henrie, R. N., Marlier, J. F., Johnson, K. A. & 17. Chu, C. K., Reichman, U., Watanabe, K. A. & Fox, J. J. (1977).J. Benkovic, S. J. (1985) Biochemistry 24, 4010-4018. Org. Chem. 42, 711-714. 45. Polesky, A. H., Dahlberg, M. E., Benkovic, S. J., Grindley, 18. Piccirilli, J. A. (1989) Ph.D. dissertation (Harvard Univ., Cam- N. D. F. & Joyce, C. M. (1992) J. Biol. Chem. 267, 8417-8428. bridge, MA). 46. Polesky, A. H., Dahlberg, M. E., Benkovic, S. J., Grindley, 19. Sinha, N. D., Biernat, J. & Koster, H. (1983) Tetrahedron Lett. 24, N. D. F. & Joyce, C. M. (1992) J. Biol. Chem. 267, 8417-8428. 5843-5846. 47. Joyce, C. M., Sun, X. C. & Grindley, N. D. F. (1992) J. Biol. 20. Van Aerschot, A., Mag, M., Herdewijn, P. & Vanderhaeghe, H. Chem. 267, 24485-24500. (1989) Nucleosides Nucleotides 8, 159-178. 48. Schinazi, R. F. (1993) Perspect. Drug Discovery Des. 1, 151-180. 21. Ludwig, J. & Eckstein, F. (1989) J. Org. Chem. 54, 631-635. 49. Krauch, T. (1989) Ph.D. dissertation (Eidgenossiche Technische 22. Ludwig, W. & Follman, H. (1978) Eur. J. Biochem. 82, 393-403. Hochschule, Zurich). Downloaded by guest on October 1, 2021