Biospecific labeling with undecagold: visualization of the biotin-binding site on avidin.

The biotin-binding site on avidin has been labeled with biotin conjugated to undecagold, an organometallic cluster compound containing 11 gold atoms in a core angestroms in diameter. Examination of unstained specimens by scanning transmission electron microscopy reveals the labeled sites directly, without computational averaging or filtering of the images. This approach should be widely applicable for determining the locations of subunits and functional site in biological macromolecules at a resolution at a resolution in range of 15 angstroms.

2-Amino-6-methoxypurine arabinoside: an agent for T-cell malignancies.

Earlier studies have shown guanine arabinoside (ara-G) is an effective agent against growth of T-cell lines and freshly isolated human T-leukemic cells. However, poor water solubility of ara-G limits clinical use. 2-Amino-6-methoxypurine arabinoside (506U) is a water-soluble prodrug converted to ara-G by adenosine deaminase. 506U is not a substrate for deoxycytidine kinase, adenosine kinase, or purine nucleoside phosphorylase and is phosphorylated by mitochondrial deoxyguanosine kinase at a rate 4% that of ara-G phosphorylation. Mitochondrial DNA polymerase was the least sensitive to ara-GTP inhibition of the five human DNA polymerases tested. [3H]506U was anabolized to ara-G 5'-phosphates in CEM cells but not to phosphorylated metabolites of 506U. 506U was selective for transformed T over B cells and also inhibited growth in two of three monocytic lines tested. 506U given i.v. to cynomolgus monkeys was rapidly converted to ara-G; the ara-G had a half-life of approximately 2 h. 506U had in vivo dose-dependent efficacy against human T-cell tumors in immunodeficient mice. A Phase 1 trial of 506U against refractory hematological malignancies is now in progress at two study sites.

Field Evaluation of a New Sequential Sampling Technique for Determining Apple Scab "Risk".

Most fungicide sprays applied to apple orchards in the New England states are targeted at the management of apple scab. Researchers have developed action thresholds that aid in decision-making on whether early spring fungicide applications could be eliminated without a significant increase in the incidence of fruit scab at harvest. To facilitate grower adoption of these thresholds, a simplified, sequential sampling technique in autumn to determine the "scab risk" of an orchard for the following spring was proposed in the scientific literature. However, this technique had not been evaluated in the field. In autumn 1999, 2000, and 2001, orchards were evaluated using the new sequential sampling technique to determine scab risk. Risk ratings were compared with those obtained by the original, nonsequential procedure in each orchard. Data also were examined using a simulation sequential sampling computer program to determine whether or not risk ratings would change if different trees or shoots were used. In two of the assessed orchards, "delayed-spray" experiments involving two treatments (a delayed-spray and full-spray treatment) were conducted in 2000 and 2001. Delayed-spray replicates were to receive no fungicide sprays until after the third primary infection period (but before the fourth) or until the pink stage of bud development, whichever came first; full-spray replicates received fungicide sprays starting at the green-tip stage of bud development. The sequential sampling technique provided scab-risk ratings consistent with the original, nonsequential procedure, at potentially significant time savings. Also, following the delayed-spray strategy in low-risk orchards did not result in significant differences in fruit scab at harvest compared with initiating spraying at the green-tip phenological bud stage.

3'-Azido-3',5'-dideoxythymidine-5'-methylphosphonic acid diphosphate: synthesis and HIV-1 reverse transcriptase inhibition.

3'-Azido-3'-deoxythymidine-5'-phosphonate was synthesized by a five-step reaction sequence. The 5'-phosphonate was inactive against HIV-1 in MT4 cells. The absence of activity against HIV-1 was at least partially explained by demonstrating that the Km value for the 5'-deoxy-5'-methylphosphonic acid diphosphate analog with HIV-1 reverse transcriptase (RT) was 320-fold greater than the Km value for 3'-azido-3'- deoxythymidine-5'-triphosphate (AZTTP), and the kcat value for the 5'-deoxy-5'-methylphosphonic acid diphosphate analog was one-seventh the value for AZTTP. These differences in kinetic constants were due to a change in the rate-determining step from dissociation of the RT chain-terminated template-primer complex to the catalytic step. Thus, substitution of a methylene group for the 5'-oxygen atom of AZTTP resulted in an 1800-fold reduction in the rate constant for RT-catalyzed phosphodiester bond formation.

Phosphorylation of ganciclovir phosphonate by cellular GMP kinase determines the stereoselectivity of anti-human cytomegalovirus activity.

A racemic mixture of ganciclovir phosphonate was resolved by stereoselective phosphorylation using GMP kinase. The R-enantiomer of ganciclovir phosphonate was active against human cytomegalovirus but the S-enantiomer was less active. We show that enantiomeric selectivity of antiviral for ganciclovir phosphonate was conferred by stereoselective phosphorylations by mammalian enzymes, not by stereoselective inhibition of DNA polymerase from human cytomegalovirus.

Herpes simplex virus type 1 and human DNA polymerase interactions with 2'-deoxyguanosine 5'-triphosphate analogues. Kinetics of incorporation into DNA and induction of inhibition.

The ability of herpes simplex virus type 1 (HSV-1) DNA polymerase, HeLa polymerase alpha, and HeLa polymerase beta to utilize several dGTP analogues has been investigated using a defined synthetic template primer. The relative efficiencies of the triphosphates of 9-[(2-hydroxyethoxy)methyl]guanine (acyclovir triphosphate, ACVTP), 9-[(1,3-dihydroxy-2-propoxy)methyl] guanine (ganciclovir triphosphate, DHPGTP), and 2',3'-dideoxyguanosine (ddGTP) as substrates for the three polymerases were: HSV-1 polymerase, dGTP greater than ACVTP approximately equal to DHPGTP greater than ddGTP; polymerase alpha, dGTP greater than ACVTP approximately equal to DHPGTP much greater than ddGTP; polymerase beta, ddGTP greater than dGTP much greater than ACVTP approximately equal to DHPGTP. The potent inhibition of HSV-1 polymerase by ACVTP has been shown previously to be due to the formation of a dead-end complex upon binding of the next 2'-deoxynucleoside 5'-triphosphate encoded by the template after incorporation of acyclovir monophosphate into the 3' end of the primer (Reardon, J. E., and Spector, T. (1989) J. Biol. Chem. 264, 7405-7411). This mechanism was shown here to be a general mechanism for inhibition of polymerases by the obligate chain terminators, ACVTP and ddGTP. The ACVTP-induced inhibition was 30-fold more potent with HSV-1 polymerase than with polymerase alpha. This difference may contribute to the antiviral selectivity of this nucleotide analogue. The effect of ganciclovir monophosphate incorporation (a nonobligate chain terminator) on subsequent primer extension was also evaluated. With HSV-1 polymerase and polymerase alpha, although there was a considerable reduction in the efficiency of utilization of the 3'-DHPGMP-terminal primer, contrasting kinetic behavior was observed. With HSV-1 polymerase, insertion of DHPGTP resulted in a significant reduction in Vmax for subsequent nucleotide incorporations. In contrast, with polymerase alpha, a relatively small decrease in Vmax was accompanied by increased Km values for subsequent nucleotide incorporations.

Herpes simplex virus type 1 DNA polymerase. Mechanism-based affinity chromatography.

The potent inhibition of herpes simplex type 1 (HSV-1) DNA polymerase by acyclovir triphosphate has previously been shown to be due to the formation of a dead-end complex upon binding of the next 2'-deoxynucleoside 5'-triphosphate encoded by the template after incorporation of acyclovir monophosphate into the 3'-end of the primer (Reardon, J. E., and Spector, T. (1989) J. Biol. Chem. 264, 7405-7411). This mechanism of inhibition of HSV-1 DNA polymerase has been used here to design an affinity column for the enzyme. A DNA hook template-primer containing an acyclovir monophosphate residue on the 3'-primer terminus has been synthesized and attached to a resin support. In the absence of added nucleotides, the column behaves as a simple DNA-agarose column, and HSV-1 DNA polymerase can be chromatographed using a salt gradient. The presence of the next required nucleotide encoded by the template (dGTP) increases the affinity of HSV-1 DNA polymerase for the acyclovir monophosphate terminal primer-template attached to the resin, and the enzyme is retained even in the presence of 1 M salt. The enzyme can be eluted from the column with a salt gradient after removal of the nucleotide from the buffer. Traditionally, the affinity purification of an enzyme relies on elution by a salt gradient, pH gradient, or more selectively by addition of a competing ligand (substrate/inhibitor) to the elution buffer. In the present example, elution of HSV-1 polymerase is facilitated by removal of the substrate from the buffer. This represents an example of mechanism-based affinity chromatography.

Human immunodeficiency virus reverse transcriptase. A kinetic analysis of RNA-dependent and DNA-dependent DNA polymerization.

A minimal kinetic mechanism for HIV reverse transcriptase (RT)-catalyzed RNA-dependent and DNA-dependent DNA polymerization was determined by pre-steady-state kinetic methods to be: [formula: see text] where E, TP, dNTP, and PPi are RT, template-primer, 2'-deoxynucleoside 5'-triphosphate, and inorganic pyrophosphate, respectively. Defined sequence template-primers that encode for incorporation of dTTP were prepared by annealing either a 44-mer RNA template or a 44-mer DNA template (of the same sequence) to a 21-mer DNA primer (r44:d21-mer and d44:d21-mer, respectively). The values of the above kinetic constants were determined for dTMP and 3'-azido-3'-deoxythymidine 5'-monophosphate (AZTMP) incorporation into both template primers. The kcat and Km values calculated from these kinetic constants were similar to the values directly determined from steady-state experiments. Further, the net rate constants for processive incorporation of three successive nucleotides into the r44:d21-mer were similar indicating that a rate-determining step did not follow catalysis. A 20-fold difference in the rate constants (kp) for incorporation of dTMP into the r44:d21-mer versus the d44:d21-mer was largely responsible for the difference in the calculated processivity numbers of 340 and 5, respectively. Finally, the rate constant for pyrophosphorolysis of the 3'-AZTMP-terminated r44:d21-mer (kpyro) was similar to the rate constant for dissociation of the chain-terminated template primer from the enzyme (koff) indicating that millimolar concentrations of intracellular inorganic pyrophosphate would be required for pyrophosphorolysis of AZTMP-terminated retroviral genomes.

Human immunodeficiency virus reverse transcriptase: steady-state and pre-steady-state kinetics of nucleotide incorporation.

Steady-state and pre-steady-state kinetic constants were determined for reverse transcriptase catalyzed incorporation of nucleotides and nucleotide analogues into defined-sequence DNA primed-RNA templates. 3'-Azido-3'-deoxythymidine 5'-triphosphate (AZTTP) was almost as efficient a substrate (kcat/Km) as dTTP for the enzyme. In contrast, the four 2',3'-dideoxynucleoside 5'-triphosphates and 3'-deoxy-2',3'-didehydrothymidine 5'-triphosphate (d4TTP) were 6-30-fold less efficient substrates of the enzyme. The kcat values for all nucleotide analogues were similar, consistent with a kinetic model in which the steady-state rate-limiting step was dissociation of the template-primer from the enzyme [Reardon, J. E., & Miller, W. H. (1990) J. Biol. Chem. 265, 20302-20307]. The pre-steady-state kinetics of single-nucleotide incorporation were consistent with the kinetic model: [formula: see text] where E, TP, and dNTP represent reverse transcriptase, a defined-sequence DNA primed-RNA template, and 2'-deoxynucleoside 5'-triphosphate (or analogue), respectively. The dissociation constant (Kd1) for template-primer binding was 10 nM, and the estimated rate constants for association and dissociation of the enzyme.template-primer complex were 4 x 10(6) M-1 s-1 and 0.04 s-1, respectively. The dissociation constants (Kd2) for dTTP, AZTTP, and 3'-deoxythymidine 5'-triphosphate (ddTTP) were 9, 11, and 4.6 microM, respectively. Thus, the differences in steady-state Km values were not due to differences in binding of the nucleotide analogues to the enzyme. In contrast, the rate-limiting step during single-nucleotide incorporation (kp) was sensitive to the structure of the nucleotide substrate.(ABSTRACT TRUNCATED AT 250 WORDS)

Inhibition of cholesterol biosynthesis by fluorinated mevalonate analogues.

The conversion of mevalonate to cholesterol in rat liver homogenates (IC50 = 0.01-1.0 mM) is inhibited by 6- (I), 6,6-di- (II), and 6,6,6-trifluoromevalonate (III), as well as 4,4-difluoromevalonate (IV). Addition of compound I, III, or IV to rat liver homogenates results in the accumulation of 5-phospho- and 5-pyrophosphomevalonate. The conversion of isopentenyl pyrophosphate to cholesterol is not inhibited by the fluorinated analogues. It thus appears likely that the decarboxylation of mevalonate 5-pyrophosphate is inhibited. Rat liver homogenates catalyze the phosphorylation of I and III. The inhibition of the decarboxylation of mevalonate 5-pyrophosphate by I and III was demonstrated directly with partially purified decarboxylase. Compound I is a remarkably effective inhibitor of the decarboxylation (Ki = 10 nM). Similar results were reported by Nave et al. [Nave, J. F., d'Orchymont, H., Ducep, J. B., Piriou F., & Jung, M. J. (1985) Biochem. J. 227, 247]. It is likely that the phosphorylated or pyrophosphorylated forms of all inhibitors tested are responsible for inhibition. We also describe a chemical method for the synthesis of mevalonate 5-pyrophosphate.

Herpes simplex virus type 1 DNA polymerase. Mechanism of inhibition by acyclovir triphosphate.

Acyclovir triphosphate (ACVTP) was a substrate for herpes simplex virus type 1 (HSV-1) DNA polymerase and was rapidly incorporated into a synthetic template-primer designed to accept either dGTP or ACVTP followed by dCTP. HSV-1 DNA polymerase was not inactivated by ACVTP, nor was the template-primer with a 3'-terminal acyclovir monophosphate moiety a potent inhibitor. Potent inhibition of HSV-1 DNA polymerase was observed upon binding of the next deoxynucleoside 5'-triphosphate coded by the template subsequent to the incorporation of acyclovir monophosphate into the 3'-end of the primer. The Ki for the dissociation of dCTP (the "next nucleotide") from this dead-end complex was 76 nM. In contrast, the Km for dCTP as a substrate for incorporation into a template-primer containing dGMP in place of acyclovir monophosphate at the 3'-primer terminus was 2.6 microM. The structural requirements for effective binding of the next nucleotide revealed that the order of potency of inhibition of a series of analogs was: dCTP much greater than arabinosyl-CTP greater than 2'-3'-dideoxy-CTP much greater than CTP, dCMP, dCMP + PPi. In the presence of the next required deoxynucleotide (dCTP), high concentrations of dGTP compete with ACVTP for binding and thus retard the formation of the dead-end complex. This results in a first-order loss of enzyme activity indistinguishable from that expected for a mechanism-based inactivator. The reversibility of the dead-end complex was demonstrated by steady-state kinetic analysis, analytical gel filtration, and by rapid gel filtration through Sephadex G-25. Studies indicated that potent, reversible inhibition by ACVTP and the next required deoxynucleoside 5'-triphosphate also occurred when poly(dC)-oligo(dG) or activated calf thymus DNA were used as the template-primer.

Synthesis of undecagold cluster molecules as biochemical labeling reagents. 1. Monoacyl and mono[N-(succinimidooxy)succinyl] undecagold clusters.

This paper describes the synthesis and characterization of succinyl, phthalyl, and N-(succinimidooxy)-succinyl derivatives of the undecagold cluster complex tricyanoheptakis[4,4',4"-phosphinidynetris(benzenemethana mine)]undecagold, 1, molecular formula Au11(CN)3[P(C6-H4CH2NH2)3]7. These are useful as electron-dense reagents for labeling biological structures in preparation for electron microscopic analysis. Limited reaction of 1 with succinic or phthalic anhydrides produces a mixture of mono-, bis-, etc. (N-succinyl)-1 or (N-phthalyl)-1, which can be separated by anion-exchange chromatography at pH 11.5. Yields of monoacylated derivatives can be maximized by controlling the ratio of succinic or phthalic anhydride to 1. The remaining 20 primary amino groups can be dialkylated or acetylated, blocking their participation in further chemical modifications of the carboxylic functional group introduced in the succinylation or phthalylation of 1. These carboxyl groups can be activated as N-hydroxysuccinimido esters, which are acylating derivatives of 1. An example is mono[N-(succinimidooxy)-succinyl]icosa(N,N-dimethyl)-1 whose synthesis is described. Bis- and tris(N-succinyl) and -(N-phthalyl) derivatives of 1 are also produced and isolated in usable quantities.

Human immunodeficiency virus reverse transcriptase ribonuclease H: specificity of tRNA(Lys3)-primer excision.

Two model substrates were prepared to examine the mechanism of tRNA-primer excision catalyzed by reverse transcriptase associated ribonuclease H (RT-RNase H). The first model substrate contained sequences from the HIV genome and was designed to be structurally similar to the DNA-extended tRNA created by initiation of minus-strand DNA synthesis during retroviral replication. The DNA-extended RNA was a template and was annealed to a DNA oligonucleotide that primed reverse transcription of the RNA in the template. The second model substrate was structurally similar the first substrate but contained sequences unrelated to the HIV viral genome. The RT-RNase H catalyzed excision of the RNA from the template of the two model substrates was examined. Human immunodeficiency virus (HIV) and Moloney murine leukemia virus RT-RNase H hydrolyzed the substrates to leave a single ribonucleotide 5'-phosphate at the 5'-terminus of the model DNA genome. In contrast, avian myeloblastosis virus RT-RNase H hydrolyzed the phosphodiester bond at the DNA-RNA junction. These hydrolytic specificities were not highly dependent on substrate sequence. The importance of these specificities to retroviral integration is discussed. Additional data indicated that the HIV polymerase and RNase H active sites are separated by a distance equivalent to the length of a 15-nucleotide RNA-DNA heteroduplex.

Mechanism of action of isopentenyl pyrophosphate isomerase: evidence for a carbonium ion intermediate.

Isopentenyl pyrophosphate isomerase catalyzes the interconversion of isopentenyl pyrophosphate and dimethylallyl pyrophosphate. The isomerase from yeast has been purified to near homogeneity (purity greater than 90%). The substrate analogue (Z)-3-(trifluoromethyl)-2-butenyl pyrophosphate reacts at less than 1.8 X 10(-6) times the rate of dimethylallyl pyrophosphate. The enzyme is irreversibly inactivated by 2-(dimethyl-amino)ethyl pyrophosphate (I). These observations are consistent with a carbonium ion mechanism for the isomerization. Compound I is an analogue of the intermediate carbonium ion and probably acts as a transition state analogue. For I, kon' = 2.1 X 10(6) M-1 min-1. No off-rate was detected and, therefore, Ki less than 1.4 X 10(-11) M. Upon denaturation of the inactivated enzyme, I is released unchanged. 2-(Trimethylammonio)ethyl pyrophosphate also inhibits with Ki' = 7 X 10(-7) M, kon' = 4.4 X 10(4) M-1 min-1, and koff = 0.03 min-1. Substrate analogues without a positively charged nitrogen were relatively poor inhibitors. The best inhibitor of these is ethyl pyrophosphate, Ki = 10(-4) M. The enzyme is inactivated by sulfhydryl-selective reagents. These reagents also prevent binding of I to the enzyme. The inactivation by iodoacetamide is dependent upon one ionizable group (pK = 9.3). The pH dependence of V and V/K for the isomerase-catalyzed reaction also depends upon a group with pK = 9.3.

Reverse transcriptase.RNase H from the human immunodeficiency virus. Relationship of the DNA polymerase and RNA hydrolysis activities.

Human immunodeficiency virus reverse transcriptase.RNase H (RT.RNase H) has an RNA hydrolysis specificity which was influenced both by the sequence of the DNA primer-RNA template and by binding of the polymerase active site to the primer 3' terminus. RT.RNase H selectively hydrolyzed the phosphodiester bond between the 15th and 16th ribonucleotide back from the ribonucleotide that is complementary to the primer 3'-terminal deoxynucleotide. The cleavage site for RT.RNase H remained a fixed distance behind the 3'-primer terminus as the polymerase extended the primer. This cleavage was not strongly affected by the position of the DNA primer on the template nor was it affected by reducing the primer length from 40 to 25 nucleotides. These results suggest that the distance between the RNase H and polymerase active sites corresponds to the length of a 15-16-nucleotide DNA-RNA heteroduplex. Since one helical turn is approximately 10 nucleotides, the distance between the active sites is 1.5 helical turns of heteroduplex. Therefore, the two active sites (catalyzing reactions on opposite strands) bind the same side of the RNA-DNA double helix. RT.RNase H also showed some sequence dependence for the site of hydrolysis. This sequence dependence has not been fully characterized. The rate of RT.RNase H cleavage was weakly inhibited by the next coded deoxynucleoside triphosphate following the incorporation of a dideoxynucleotide.

Human immunodeficiency virus reverse transcriptase. Substrate and inhibitor kinetics with thymidine 5'-triphosphate and 3'-azido-3'-deoxythymidine 5'-triphosphate.

3'-Azido-3'-deoxythymidine 5'-triphosphate (AZTTP) was an efficient substrate for the human immunodeficiency virus 1 reverse transcriptase. It was incorporated into both homopolymer and defined sequence DNA-primed RNA templates and DNA-primed DNA templates. The substrate and inhibitor kinetics of both AZTTP and dTTP were dependent on the template-primer and reaction conditions used. dTMP was incorporated into poly(rA).oligo(dT) and into a defined sequence DNA-primed RNA template (when the other three 2'-deoxynucleoside 5'-triphosphates were present) as a conventional substrate, with steady-state Km values of 5-10 microM. The results suggest that the reverse transcriptase was capable of processive DNA polymerization on these DNA-primed RNA templates. In contrast, in the absence of the other three 2'-deoxynucleoside 5'-triphosphates, the time course for incorporation of dTMP into the same defined sequence DNA-primed RNA template was biphasic. A burst of product formation was observed followed by a slow steady-state rate with a Km value of 0.082 microM. AZTMP incorporation into poly(rA).oligo(dT) and into the defined sequence DNA-primed RNA template produced similar biphasic time courses and steady-state Km values. These results were consistent with rate-limiting dissociation of the polymerase.template-primer complex after "forced" termination of polymerization. AZTMP and dTMP were both incorporated into the homopolymer DNA-primed DNA template, poly(dA).oligo(dT), and a defined sequence DNA-primed DNA template as conventional substrates. Their Km values were similar (2-10 microM). The absence of biphasic time courses suggested that dissociation of the DNA-primed DNA templates from the enzyme, after forced termination, was not rate-limiting. This was consistent with a more distributive mode of DNA polymerization. With the defined sequence template-primers and poly(dA).oligo(dT), Ki values for both dTTP and AZTTP were comparable to their Km values. Thus, AZTTP appeared to be a simple competitive substrate-inhibitor with respect to dTTP. AZTTP inhibition of dTMP incorporation into poly(rA).oligo(dT) was linear competitive at low concentrations (0-100 nM) of AZTTP (Ki = 35 nM) but became hyperbolic (decreasing potency) at concentrations of AZTTP above this range. A mechanism for this nonlinear inhibition is discussed.

Synthesis of undecagold cluster molecules as biochemical labeling reagents. 2. Bromoacetyl and maleimido undecagold clusters.

Derivatives of heptakis[4,4',4"-phosphinidynetris(benzenemethanamine) ]undecagold, 1, molecular formula Au11(CN)3[P(C6H4CH2NH3)]7, are described. These include undecagold complexes with a single free primary amino group, a single bromoacetyl group, and a single maleimido group per molecule. Hydrolysis of mono(N-phthalyl)icosa(N-acetyl)-1 at pH 3.2 and 46 degrees C under anaerobic conditions and in the presence of NaBH3CN produces icosa(N-acetyl)-1. Partial acylation of 1 with 1.3 equiv of 2,3-dimethylmaleic anhydride followed by complete acetylation with acetic anhydride produces a mixture consisting largely of mono- and bis(dimethylmaleyl)peracetyl-1. Hydrolysis of 2,3-dimethylmaleimides at pH 3.2 for 1 at 25 degrees C produces a mixture of icosa(N-acetyl)-1, with a single free amino group, and nondea(N-acetyl)-1. This mixture can be quantitatively separated by cation-exchange chromatography at pH 7, giving homogeneous icosa(N-acetyl)-1 in an overall yield of 55%. Icosa(N-acetyl)-1 serves as the starting material for the synthesis of the alkylating derivatives mono(N-bromoacetyl)icosa(N-acetyl)-1 and mono[N-(p-maleimidobenzoyl)]icosa(N-acetyl)-1. These derivatives can be used for alkylating proteins in preparation for electron microscopy.

Human immunodeficiency virus reverse transcriptase. Effect of primer length on template-primer binding.

Poly(rA).oligo(dT)n binding to human immunodeficiency virus type-1 reverse transcriptase heterodimer (p66-p51) was primer length-dependent. The estimated Kd for (n = 10-14) was 20-30 nM and for (n = 16-20) was 0.11-0.14 nM. Gel electrophoretic analysis of the patterns of primer extension was consistent with an abrupt change in the Kd between a primer length of 14 and 16 nucleotides. Further, the rate constant for dissociation of the reverse transcriptase-template-primer complex was determined from steady state kinetics and enzyme-template-primer trapping experiments to be independent of primer length. Thus, the abrupt change in Kd was most likely due to a change in the rate constant for formation of the reverse transcriptase-template-primer complex. A similar shift in the Kd for template-primer binding was observed with poly(dA).oligo(dT)n. Reverse transcriptase homodimer (p66) catalyzed the incorporation of dTMP into poly(rA).oligo(dT)n with the same primer length dependence observed for the heterodimer. In contrast, binding of the p51 homodimer to poly(rA).oligo(dT)n was independent of primer length. Thus, the RNase H domain may contribute to reverse transcriptase heterodimer or p66 homodimer binding to template-primers in which the primer length is greater than 14 nucleotides.