Nucleic Acids Research, Vol. 18, No. 3

1990 Oxford University Press 429

Inhibition of the p66/p51 form of human immunodeficiency virus reverse transcriptase by tRNALYs Bruno Bordier, Laura Tarrago-Litvak, Marie-Line Sallafranque-Andreola, Dominique Robert, Daniele Tharaud, Michel Fournier, Philip J.Barr1, Simon Litvak* and Leila Sarih-Cottin* Institut de Biochimie Cellulaire et Neurochimie du CNRS, 1 rue Camille Saint Saens, 33077 Bordeaux cedex, France and 'Chiron Co., Horton Street, Emeryville, CA, USA Received November 27, 1989; Revised and Accepted January 10, 1990

ABSTRACT Human immunodeficiency virus (HIV) reverse transcriptase (RT) uses host tRNALYS partially annealed to the primer binding site (PBS) as primer for the initiation of cDNA synthesis. When assaying cDNA synthesis with a template-primer complex formed by an RNA fragment carrying the PBS site and bovine tRNALYS we noticed that an excess of primer tRNA inhibited strongly the DNA polymerase activity of a recombinant HIV RT (p66-p51 heterodimeric form) produced in transformed yeast cells. The same inhibitory effect was observed with animal DNA polymerase a, while avian retrovirus RT was neither affected by tRNALYs nor by its specific primer tRNATrP. Although the strongest inhibition was observed with tRNALYS, other tRNas like tRNAPhe and tRNATrp inhibited also the HIV RT, whereas tRNAs specific for valine, proline and glycine had no effect on enzyme activity. Digestion of tRNALYS with pancreatic RNase abolished the inhibition ; on the other hand T1 RNase digestion had no effect on the inhibition suggesting a role of the anticodon region in this effect. The 12- and 14-mers corresponding to the anticodon regions of the three bovine tRNALYS isoacceptors inhibited RT activity, indicating that at least an important part of the inhibitory effect could be ascribed to this tRNA region. A strong stimulation of DNA polymerase activity was observed when the effect of tRNALYS was assayed on a recombinant HIV reverse transcriptase produced in a protease defficient yeast strain, which leads to the production of an active p66 enzyme. The same tRNAs that inhibited strongly the heterodimeric form stimulated the p66 form of HIV reverse transcriptase. The results suggest that although both enzymatic forms are able to interact with tRNALYS the topography, as well as the functional implications of the interaction between the precursor and the mature form of HIV reverse transcriptase with the tRNALYS primer, are different.

*

To whom

correspondence should be addressed

INTRODUCTION Retroviruses, the etiological agents of several forms of cancer and AIDS, encode an RNA-dependent DNA polymerase (reverse transcriptase) that carries several enzymatic functions: RNAdirected DNA synthesis of viral DNA minus strand, DNAdirected DNA synthesis of the plus strand of the proviral DNA and an RNase H activity [1-5]. This activity is responsible for viral RNA degradation once the minus strand DNA synthesis has occured. After the RNA moiety removal, the minus strand becomes available as a template for the synthesis of the DNA plus strand. This RNase H activity has been also implicated in releasing tRNA primer and in nicking viral RNA to generate the polypurine-rich oligonucleotide that primes plus-strand DNA synthesis [6,7]. As with all DNA polymerases, reverse transcriptase needs a primer carrying a 3' OH free group to initiate cDNA synthesis. Concerning the viral first strand DNA synthesis from an RNA template, the in vivo primer has been shown to be tRNA [8-10]. A region, near the 5' end of the retroviral RNA genome ('primer binding site' or PBS), is complementary to the last 18 nucleotides of the 3' CCA end of the specific primer. Avian myeloblastosis virus (AMV) reverse transcriptase uses host tRNAThP as primer while the Moloney murine leukemia virus (M-MuLV) enzyme uses tRNAPrO. In the case of human immunodeficiency virus (HIV), tRNALYs,3 is most probably the primer as deduced from the retroviral genome nucleotide sequence [11]. Previous work has shown that avian reverse transcriptase is involved in the selection of tRNATrP and its annealing to the avian retrovirus PBS [12,13]. More recently we have shown that a retrovirus-encoded small molecular weight nucleic acid binding protein can stimulate the in vitro tRNA-PBS annealing [14]. HIV reverse transcriptase is isolated as a 66-51 kDa (p66 and p51) dimer from the viral particle. The pSi subunit is derived from p66 form. The maturation occurs by partial proteolysis by the virus encoded acid protease which also cleaves the gag-pol precursor [15-18]. A recombinant HIV reverse transcriptase

430 Nucleic Acids Research expressed in transformed yeast cells [16] was biochemically characterized. Furthermore we showed that the p66-pSi enzyme can form a stable complex with bovine tRNALYS [19]. During the synthesis of a natural primer-template complex using bovine tRNALYS and a viral RNA fragment carrying the PBS site, we observed a strong inhibition of cDNA synthesis by free tRNALYS. The same effect has been observed with activated DNA or a natural RNA template, while no inhibition was observed with the synthetic poly(rA)-oligo(dT) template, routinely used for assaying reverse transcriptase [19]. In this article we have pursued a detailed study of the effect of several purified tRNAs on the mature (p66-pS 1) and precursor (p66) forms of HIV reverse transcriptase, as well as on other DNA polymerases.

MATERIALS AND METHODS Materials Unlabelled nucleotides, oligonucleotides or polynucleotides were obtained from Sigma or Pharmacia. Radioisotopes were purchased from Commissariat 'a l'Energie Atomique-Saclay (CEA-France). AMV reverse transcriptase was purchased from Genofit. T7 RNA polymerase, PstI and SphI restriction enzymes and T, ribonuclease were purchased from BRL, proteinase K from Boehringer and pancreatic ribonuclease A from Sigma. Xenopus laevis oocytes DNA polymerases a and y were purified as described previously [20 and 19 respectively]. Rat liver DNA polymerase (3 was a kind gift of Dr. J.M. Rossignol (VillejuifFrance). tRNAs purification tRNAs from beef liver were purified by classical chromatogrphic methods (benzoylated DEAE-cellulose, DEAE-sephadex): tRNATrP, tRNAvld and tRNAPh were prepared according to Fournier et al. [21] ; tRNALYS, tRNAGIY and tRNAPr were obtained as described by Fournier [22]. The purity of each tRNA was determined by its aminoacylation level using a pool of amynoacyl-tRNA synthetases prepared from wheat germ as described by Nishimura and Weinstein [23]. The purified tRNAs showed an aminoacylation level between 1200 and 1600 picomoles per 260 nm absorbancy unit. HIV reverse transcriptase purification HIV reverse transcriptase p66/pS1 form was obtained as before [19]. HIV reverse transcriptase p66 form was purified as described previously for the p66/p5i form, but using a proteasedeficient yeast strain (JSC 302, derivative of AB 116: MATa, leu2-, trp-1, ura3-52, prBl-1122, pep4-3, prCI-407 [cir°]) transformed with the expression vector pBS24RT5. The parent vector, pBS24, has been described previously [24]. Reverse transcriptase assays Incubation was carried out at 37°C for various times: a) In the presence of poly(rA)-oligo(dT), the reaction mixture contained in a final volume of 0.05 ml, 50 mM Tris-HCl pH 8.0, 10 mM MgCl2, 10 mM dithiothreitol, 80 mM KCl, 20 jg/ml of poly(rA)-oligo(dT), 0.5-1 ytCi [3H]-dTTP (56 Ci/mmol), 50 ,uM dTTP and enzyme as indicated. b) In the presence of activated DNA : the reaction mixture contained in a final volume of 0.05 ml, 50 mM Tris-HCl pH 8.0, 10 mM MgCl2, 10 mM dithiothreitol, 15 ltg/ml of activated DNA, 1 yCi [3H]-dTTP (56 Ci/mmol), 5 ztM of dTTP, 50 AM each of dATP, dCTP and dGTP and enzyme as indicated.

c) When the retroviral RNA fragment, carrying the PBS region, was used as template with tRNALYS as primer, the incubation mixture (final volume : 50 1l) contained the following reagents: 50 mM Tris-HCl pH 8.0, 10 mM DTT, 10 mM MgCl2, 50 IM each of dTTP, dCTP and dGTP, 5 ytM [a-32P] dATP (3000 Ci/mmol), 10 nM of 'PBS' HIV RNA and different amounts of bovine tRNALYS. Reactions were stopped by the addition of 1 ml of cold 10 % trichloroacetic acid plus 0.1 M sodium pyrophosphate. The precipitates were filtered through nitrocellulose membranes, washed with 2 % trichloroacetic acid, dried and counted in a PPO/POPOP/toluene scintillation mixture.

Animal DNA polymerase assays Incubations were performed with calf thymus activated DNA under the same conditions as for reverse transcriptase.

Purification of 'PBS' HIV RNA A 1.1 Kbp DNA fragment, containing R, U5 and PBS HIV sequences (gift from Dr. J.L. Darlix ; CNRS-Toulouse), was cloned in the PstI site of a Bluescribe plasmid. After SphI vector linearization, the insert was transcribed by T7 RNA polymerase. About 1 fig of linearized DNA was incubated with T7 RNA polymerase buffer, 50 mM DTT, 150 jtM of each rXTP, 35 units/4d of RNasin and 5 units/4l of T7 RNA polymerase in a final volume of 50 Id for 1 hour at 37°C. The reaction was stopped by adding 100 yg/ml of proteinase K for 15 min at 37°C. Transcription products were extracted twice with phenol, twice with phenol/chloroform/isoamyl alcohol (v/v/v : 24/24/1) and precipitated in 2.5 volumes of ethanol. RNA/DNA pellet was then washed with 70 % ethanol and resuspended in sterile water. After heating for 2 min at 85°C the solution was cooled rapidly, loaded on a cesium chloride solution (5.7 M) and centrifuged for 12 hours at 170,000 xg. RNA was recovered from the bottom of the tube, resuspended and precipitated in 2.5 volumes of ethanol. Ribonucleases digestions RNase A: 0.1 unit A260 of denatured tRNA (2 mg/ml tRNA in 0.1 M NaCl, heated for 5 min at 60°C and cooled in ice) was incubated with 6 units of RNase A for 30 min at 37°C.

T1 RNase: 0.1 unit A260 of denatured tRNA was incubated with 600 units of T1 RNase for 30 min at 47°C. Reactions were stopped by phenol extraction. Urea-PAGE analysis 8 M urea-polyacrylamide gels were composed of two parts: a 8 % polyacrylamide stacking gel and a 20 % polyacrylamide separating gel. Samples were dried and resuspended in sample buffer (50 % sucrose, 0.2 % bromophenol blue, 0.2 % xylene cyanol, 8 M urea) and migration buffer TBE (450 mM Tris, 450 mM borate, 10 mM EDTA, pH 8.0). Migration was performed for 16 hours at 250 V. Silver nitrate staining was performed according to Igloi [25], except that development was with a 0.28 % formaldehyde-0.75 M KOH solution. The reaction was stopped by 2.3 M citrate. Oligonucleotide purification When the purification of the 12 and 14-mers (TI RNase tRNALYS digestion products) was needed, gels were stained with methylene blue (this tpe of staining allows tRNA aminoacylation after its elution from the gel). Acrylamide sections containing

Nucleic Acids Research 431

oligonucleotide bands were eluted in elution buffer (0.5 M ammonium acetate, 1 mM EDTA, 0.1 % SDS). Oligonucleotides were then precipitated in 2.5 volumes of ethanol and 0.3 M of sodium acetate, pH 5.2. Production of anti-RT antibodies Polyclonal antibodies against highly purified HIV RT (FPLC fractions [19]) were prepared according to Vaitukaitis method [26]. 100 jig of RT were injected twice in the back of rabbits, with a period of three weeks between the two injections. The rabbits were then immunized with 50 jig of RT every two weeks. Antibody levels were checked a week after each injection using the ELISA technique. When anti-RT antibodies level reached a maximum, animals were sacrified.

Immunoblotting Proteins or peptides were separated by electrophoresis on a 12 % polyacrylamide-SDS gel [27]. Proteins were then transferred to Immobilon membranes (0.45 izM Millipore) for three hours in a buffer containing 10 mM Caps, pH 11.0, 10% methanol. Blots were saturated in buffer A (10 mM phosphate buffer pH 7.0, 140 mM NaCl) with 1 % bovine serum albumin for three hours at room temperature and incubated with anti-RT antibodies (dilution 1/5000 in buffer A plus 0.3% bovine serum albumin). Blots were then rinsed in buffer A and incubated overnight in peroxidase linked goat anti-rabbit IgG (purchased from Pasteur Institute) for 1 hour at 37°C. After rinsing in buffer A, development was performed according to Hawkes et al. [28] using chloronaphtol (Sigma) as substrate for peroxidase.

RESULTS AND DISCUSSION Inhibition of HIV reverse transcriptase activity by tRNALYS

fragment and cDNA synthesis was measured. As seen in Figure 1, only when the primer/template ratio was 1 (about 10 nM) or less, an expected DNA synthesis increase was observed. The high incorporation obtained in the absence of primer tRNA is probably due to the presence of a high degree of secondary structure in the RNA fragment [29]. Using the same template, AMV reverse transcriptase also gave very high incorporation values without primer tRNALYs indicating that the high background was related to the RNA and not to HIV transcriptase. A strong DNA synthesis inhibition was observed as soon as

an excess of tRNA was present in the incubation mixture ;

thus,

less than 10% of residual activity was detected at 50 nM tRNA. As previously shown, a similar inhibition by tRNALYS was obtained when calf thymus activated DNA or a plant virus RNA (turnip yellow mosaic virus, TYMV) were used as templates with HIV reverse transcriptase, the inhibition is clearly competitive regarding activated DNA template [19]. It seemed then important to study if the tRNALYs inhibitory effect was specific towards the HIV encoded DNA polymerase or if other DNA polymerases were also affected. As seen in Figure 2, the inhibition of mammalian DNA polymerase a by tRNALYs was equal to that seen with HIV RT while, under the same conditions, AMV reverse transcriptase was not inhibited. Preliminary experiments with animal DNA polymerases -y and ,B showed that these enzymes were much less inhibited than DNA polymerase ca, while E. coli DNA polymerase I was not affected at all (results not shown).

Specificity of the HIV RT inhibition by different tRNAs. Search for the tRNALYS region involved in the inhibition In order to assess the ability of other tRNAs, besides tRNALYS, to inhibit HIV RT, we studied the effect of several highly purified

In order to develop an in vitro assay using a natural primertemplate, we have synthesized a short fragment of 1100 nucleotides carrying the PBS site using the T7 RNA polymerase system. Beef liver purified tRNALYS was annealed to the RNA

100 80

200 0-%

60 P.

150

5U

-

2'

; _;

100

40

wo

a 0q' ot4C

_

0 o

U

3K a

P-

f-0

20

50

0 0 tRNA Lys (nM)

Figure 1: Inhibition of HIV reverse transcriptase DNA polymerase activity by tRNALYS with 'PBS' HIV RNA as template. HIV reverse transcriptase p66/pSl (40nM) was incubated at 37°C for 15 min as described in Materials and Methods (Reverse transcriptase assays: § c). Template-primer annealing was performed by heating the viral RNA with different quantities of tRNALYS at 850C for 1-2 min and then incubated 15 min at 37°C.

t RNA Lys (pM) Figure 2: Effect of tRNALYS on various DNA polymerases activities. HIV reverse transcriptase p66/pSl (A), X. laevis oocytes DNA polymerase a (o) or AMV reverse transcriptase (O) were incubated for 15 min at 37°C with different concentrations of tRNALYS and 50 mM Tris-HCI, pH 8.0, 10 mM DTT, 10 mM MgCI2, 50 jsM dXTP (dATP, dCTP and dGTP), 1 !Ci [3H] dTTP (56 Ci/mmol), 5 AM dTTP and 15 itg/ml of calf thymus activated DNA. Reactions were carried out in a final volume of 50 1l.

432 Nucleic Acids Research 1

-

:t

S) I.4

t:

a

_

0

a

*a

u

a _

L)

0

ito

-;1 %-

0

1 tRNA

2

1

0

tRNATrp (hiM)

(PM)

Figure 3: Effect of various beef liver tRNAs on HIV reverse transcriptase DNA polymerase activity. HIV RT p66/p5I DNA polymerase activity was assayed with different concentrations of tRNAval (U), tRNAGIY ( * ), tRNAPro (O), tRNATrp (A), tRNAPhe (A) and tRNALYS ( o) in the presence of calf thymus activated DNA as primer-template system. Incubation conditions are described in figure 2.

2

Figure 4: Effect of tRNATrP on HIV and AMV reverse transcriptase activities. Reverse transcriptases were incubated for 15 min in the same conditions as in figure 2. HIV (A) or AMV (O) reverse transcriptase activities were assayed in the presence of different concentrations of tRNATrP.

A

tRNAs on HIV reverse transcriptase activity. As seen in Figure 3, two groups of tRNAs were found. One group comprising tRNAs specific for valine, glycine and proline did not affect HIV reverse transcriptase activity, while tRNAs specific for phenylalanine, lysine and tryptophan were able to inhibit the enzyme at different degrees, the stronger inhibition beeing obtained with tRNALys. A similar study performed with DNA polymerase using the same tRNAs gave a similar pattern of inhibition (not shown). In order to assess the ability of other reverse transcriptases, besides HIV RT, to be inhibited by their specific tRNA primers, we performed the experiment described in Figure 4. As it can be seen in this Figure, tRNATrp which is the primer of avian reverse transcriptase, does not inhibit the AMV DNA polymerase activity while markedly decreasing the HIV reverse transcriptase activity. It is important to point out that the experiments described above were performed with purified tRNALys which comprise the three main isoacceptors found in beef liver. By using bidimensionnal electrophoresis we were able to separate and purify the 3 isoacceptors. When assayed separately, the three tRNALYS forms inhibited very strongly HIV reverse transcriptase: no significant differences in the inhibition degree was obtained for the three seperate isoacceptors (results not shown). In parallel experiments we have found that the complex formation between tRNALYS and HIV reverse transcriptase, determined by gel retardation assay, was equally found with the three purified isoacceptors, showing that the minor nucleotide changes which characterize the three isoacceptors [30] do not affect their capacity to bind reverse transcriptase (D.R. et al., submitted for publication). Sequence comparison of phenylalanine, lysine and tryptophan tRNAs from animal sources gave no clues on a common motif that could be involved in the inhibition process though there is a relatively high degree of sequence homology [31].

A

4C

C

C4-

C

G

pGC

_CG

-CG C G -G

G

pGGC C G c

4-

C G G C

a

(A) G

tRNAI (2)

tRNA 3

Figure 5 : tRNA LYs isoacceptors: Digestion by ribonuclease A. The cleavage sites of RNase A are shown by arrows. The 12-mer and 14-mer fragments resulting from the TI digestion are boxed in the clover leaf structure.

Having shown that tRNALYS produced the strongest inhibition, we studied the effect of pancreatic ribonuclease (RNase A) a ribonuclease from Aspergillus oryzae RNase) on

and the inhibitory effect of tRNALYS to ascertain whether the observed effect was due to some structural features of tRNALYS. In Figure 5, the RNase A cleavage sites on the three tRNALYs isoacceptors are shown. It can be seen that while RNAse A leads to very short degradation products, T, RNase which cleaves only at G residues, will lead to a 14-mer in the case of tRNALYs, I and 2 and to a 12-mer in the case of tRNALYS,3. The digestion of tRNALYS

(TI

Nucleic Acids Research 433

[Al1 tRNA

2

3

4

5

-

6

-d

._ "._

100

14-merU

.Z*. ux

I.

12- mer.-OC

50 I

1

2

3

Figure 6A: T1 ribonuclease hydrolysis of tRNALYs'Analysis of digestion products on a 20 % polyacrylamide gel -8 M urea stained with silver nitrate. 0.25 /M of tRNALYs were incubated with different amounts of T1 RNase for 30 min at 47°C in a final volume of 20 A1l. tRNALYS fragments were extracted twice with phenol and twice with chloroform. Aqueous phases were then dried and resuspended in a sample buffer (50 % sucrose, 0.2 % bromophenol blue, 0.2 % xylene cyanol, 8 M urea and migration buffer TBE). Migration was performed for 16 hours at 250 V. Lanes 1, 2 and 3: tRNALYS digestion by 150 (lane 1), 300 (lane 2) and 600 (lane 3) units of T, RNase. Lane 4 : tRNAIYS. Lane 5: Size marker: oligo(rA)1O. Figure 6B: Effect of 12-mer and 14-mers on HIV reverse transcriptase DNA polymerase activity. 40 nM HIV RT p66/p51 was assayed in the presence of a gel eluate containing no oligonucleotides (column 1; a control without gel eluate gave a similar activity level) or in the presence of 12-mer or 14-mers (columns 2 and 3 respectively). Incubation conditions and the reaction mixture are described in fig. 2. Oligonucleotides concentration was 6 /AM.

by RNase A, which degrades extensively the anticodon region of the three isoacceptors, abolished the inhibitory effect of this tRNA, whereas incubation with TI RNase, under conditions of complete tRNA digestion, did not affect the inhibitory effect of tRNALYs (Table 1). This result suggests a possible role of the anticodon region in the inhibitory effect. This suggestion was confirmed by the experiment shown in Figure 6A: increasing concentrations of TI RNase lead to the total digestion of tRNALYS and the emergence of the 12 and 14-mers fragments corresponding to the anticodon region of this tRNA. When these fragments were separately eluted and their effect assayed on the DNA polymerase activity of HIV reverse transcriptase, a strong inhibiton was obtained as shown in Figure 6B. These results argue in favour of the involvement of the anticodon region of tRNALYS in the inhibition on HIV reverse transcriptase. The sum of these results points out to a strong inhibition of the DNA polymerase activity of recombinant HIV reverse transcriptase by bovine tRNALYs. The enzyme used throughout this work is the recombinant reverse transcriptase over-expressed in transformed yeast cells and isolated as a p66-p51 dimer thus very similar to the viral reverse transcriptase isolated from HIV

Table 1. Effect of digestion products of tRNALYS by ribonuclease A or by ribonuclease on HIV reverse transcriptase activity

T,

Activity Assay system

Addition

(%)

Complete

None + tRNALYs + RNase A

100 38 90 75 100 39

+ tRNALYS + RNase A

+ T1 RNase + tRNALYS + T1 RNase

RNase A digestion was carried out at 37°C during 30 min; the reaction mixture contained, in a final volume of 20 ,u, lOmM Tris-HCI pH 7.5, 15 mM NaCl, 0.25 1tM tRNALYS and 0.3 units/ll RNase A. T, RNase digestion was carried out at 47°C during 30 min; the reaction mixture contained, in a final volume of 20 1A, 60mM Tris-HCI pH 7.5, 0.25 1sM tRNALys and 30 units/4l T1 RNase. 40 nM reverse transcriptase p66-pSi form was assayed in the presence of activated DNA as described in Materials & Methods, before and after tRNALYS digestion.

particles [19]. In experiments performed in collaboration with Drs B. Masquelier and H. Fleury (Virology Department. University of Bordeaux II), HIV virions (LAV, Bru isolate) [32]

434 Nucleic Acids Research 20

0 a 0.0 0

15,

_k c

'0. 0

10

0

5

E 0-

°-1 0.0

.

,

0.1

.

,

0.2

.

.

0.3

,

, 0.4

. 0.5

tRNA Lys (gM X Figure 8 Stimulation of the p66 form by tRNALYS. The HIV reverse transcriptase p66 form (47 nM) was incubated in the presence of different amounts of tRNALYs and its activity was determined after 15 min of incubation at 37°C. The reaction mixture contained 50 mM Tris-HCI pH 8.0, 10 mM DTT, 5 mM MgCl2, 80 mM KCI, 50 tsM each of dATP, dCTP, dGTP, 2.5 jsM dTTP, I /Ci of [3H] dTTP (56 Ci/mmol), 340 ug/ml TYMV RNA and 0.03 units A260/ml oligo(dG). Inset: same conditions but in the presence of the reverse transcriptase p66-pSI form (36 nM).

The p66 form of the HIV reverse transcriptase is stimulated

by tRNALys

Figure 7A: Purification of HIV reverse transcriptase extracted from a proteasedeficient yeast. One unit of enzyme activity is defined as that amount catalyzing the incorporation of nmole of [3H] dTTP/h at 370C. Assays were performed as described in Materials and Methods, using poly(rA)-oligo(dT) as templateprimer. Figure 7B: Western blot analysis of both forms of HIV reverse transcriptase. Immunoblotting was performed as described in Materials and methods. Size scale (arrows) was given by prestained molecular weight proteins markers (Pharmacia): subunit molecular mass of phosphorylase b 94 kDa, bovine serum albumin = 67 kDa, ovalbumin = 43 kDa, carbonic anhydrase = 30 kDa, soybean trypsin inhibitor = 20 kDa. Lane 1: preimmune serum. Lane 2: RT (p66/p51) produced in yeast. Lane 3: RT (p66) produced in the protease-deficient yeast strain. =

were lysed or absence

and assayed for reverse transcriptase in the presence of tRNALYS. The human retroviral enzyme behaved identical to the recombinant polymerase indicating that tRNALYS inhibition cannot be attributed to a peculiarity of the yeastexpressed recombinant reverse transcriptase.

The HIV reverse transcriptase inhibition by its putative primer tRNA may seem paradoxical since reverse transcriptases are supossed to interact with their primer tRNAs as a compulsory step for the initiation of cDNA synthesis. Thus, the HIV RT inhibition by its primer would jeopardize the elongation process catalyzed by the retroviral DNA polymerase. This apparent paradox can be explained by the results obtained when we used a recombinant reverse transcriptase expressed in proteasedeficient yeast cells. The purification procedure used was very similar to that described before [19] but we omitted the FPLC step since the enzyme was rather unstable after the last chromatographic step. As it can be seen in Figure 7A, a decrease of HIV reverse transcriptase specific activity was observed between the phosphocellulose and the heparine-agarose steps. The low specific activity of the HIV RT p66 form can be compared to f2 precursor of the cf3 avian reverse transcriptase [33]. As seen in Figure 7B, the immunoblot with antibodies against the heterodimeric enzyme showed that the protease deficient yeast cells produced essentially the HIV reverse transcriptase p66 form. Preliminary experiments of our laboratory, as well as the size determination of the p66 form produced in E. coli by other groups [34-36], indicate that in the native state the p66 form is found as an homodimer. Interestingly when the HIV reverse transcriptase p66 form was assayed in the presence of tRNALYS, a marked stimulation of DNA synthesis was observed; under the same experimental conditions, the p66-p51 form was inhibited by tRNALYS (Figure 8). Different p66 preparations gave a variable level of DNA

Nucleic Acids Research 435 Table 2. Effect of various beef liver tRNAs on HIV reverse transcriptase p66 DNA polymerase activity Assay system

Addition

R

Complete

None + tRNALYS + tRNATrP + tRNAPhe + tRNAVal

I 22 20 21 1

47 nM HIV RT p66 form was assayed with 50 nM of various beef liver tRNAs in the presence of poly(rA)-oligo(dT). Incubation conditions were the same as indicated in Materials and Methods. R=Activity in the presence of tRNA/activity in the absence of tRNA.

synthesis stimulation (values from 3 to 20 fold were obtained). On the other side, the stimulation was always observed, independent of wether the template used was poly(rA)-oligo(dT), activated DNA or TYMV RNA. When the effect of different tRNAs on the p66 form DNA polymerase activity was studied, we found that phenylalanine, tryptophan and lysine specific tRNAs stimulated the enzymatic activity, while tRNAval did not affect it (Table 2). From these experiments, we can conclude that the tRNAs, capable of stimulating the p66 form, were the same as those involved in the heterodimeric form inhibition. If we assume that the p66----> p66-p5l proteolysis maturation step occurs in the virion prior to the initiation of cDNA synthesis, it is possible to think that the p66 form may be involved in the selection, and eventually in the annealing of primer tRNALYs to the PBS. If this is the case, the mature heterodimer p66-p5l would play a role only in the elongation of the annealed tRNA primer. Important modification of the secondary and tertiary structure of tRNA during the annealing to the PBS site may explain that the hybridized primer does not inhibit cDNA synthesis. Further experimental work is necessary to prove this hypothesis.

CONCLUSION The inhibitory effect of some tRNAs, and more specifically of the anticodon region of tRNALYS, constitutes a new exciting source in the search of inhibitors of the key enzyme in the replicative cycle of HIV. While we were in the process of writing this article, a recent publication concerning the interactions of HIV RT and bovine tRNALYS came to our attention [37]. These authors confirm our previous results concerning the formation of a complex between the putative primer tRNA and HIV RT and show that the anticodon region of bovine tRNALYs interacts strongly with the enzyme. The strong interaction of the anticodon region with the human retroviral polymerase can be related to the inhibition of HIV RT by tRNALYs described in this work, since we have located the inhibitory activity of tRNA precisely in the anticodon region. The fact that animal cell DNA polymerase cx shows the same inhibition pattern by tRNAs as does HIV reverse transcriptase may be a drawback in designing RT inhibitors based on this observation. However this problem can potentially be overcome based on the different compartments where these two enzymes are active in vivo, since retrovirus replication takes place in the cytoplasm while the replicative DNA polymerase a acts in the nucleus in a highly condensed and poorly accessible chromatin complex. Experiments are in progress to delineate both the minimal region of tRNA able to inhibit the enzyme, and also,

to target these inhibitory oligonucleotides to the retrovirus

replicative complex.

ACKNOWLEDGMENTS We are greatly indebted to Drs A. Araya, M. Castroviejo and P.V. Graves for discussions and suggestions. This work was supported by the Agence National des Recherches sur le SIDA, l'Association des Recherches contre le Cancer, 1'INSERM, l'Universite de Bordeaux II, le CNRS et le Conseil Regionale d'Aquitaine. D.R. was supported by the Ligue Nationale Franqaise de Lutte contre le Cancer (Comit6 Dordogne) and M-L.S-A was supported by the Fondation pour la Recherche M6dicale.

REFERENCES 1. Verma, I.M. (1977) Biochim. Biophys. Acta 473, 1-38 2. Verma, I.M. (1981) In The Enzymes (Boyer, P.D., Ed.) vol. 14, part A, pp 87-103, Academic Press, New York. 3. Varmus, H. (1988) Science 240, 1427-1435 4. Sarih, L., Araya, A., Heyman, T. and Litvak, S. (1988) Research in Retroviruses and Oncogenes (INSERM Ed.), pp 55-93, John Libbey

Eurotext, London-Paris. 5. Sanchez-Pescador, R., Power, M.D., Barr, P.J., Steimer, K.S., Stempien, M.M., Brown-Shimer, S.L., Gee, W.W., Renard, A., Randolph, A., Levy, J.A., Dina, D. and Luciw, P.A. (1985) Science 227, 484-492 6. Kikuchi, Y., Ando, Y., Ichimura, N. and Noda, A. (1989) J. Biochem. 105, 974-978 7. Krug, M.S. and Berger, S.L. (1989) Proc. Natl. Acad. Sci. USA 86, 3539-3543 8. Harada, F., Sawyer, R.C. and Dahlberg, J.E. (1975) J. Biol. Chem. 250, 3487-3497 9. Harada, F., Peters, G. and Dahlberg, J.E. (1979) J. Biol. Chem. 254, 10979-10985 10. Litvak, S. and Araya, A. (1982) Trends Biochem. Sci. 7, 361-364 11. Wain-Hobson, S., Sonigo, P., Danos, 0. and Alison, M. (1985) Cell 40, 9-17 12. Panet, A., Haseltine, W., Baltimore, D., Peters, G., Harada, F. and Dahlberg, J.E. (1975) Proc. Natl. Acad. Sci. 72, 2535-2539 13. Sarih, L., Araya, A. and Litvak, S. (1988) FEBS Lett. 230, 61-66 14. Prats, A.C., Sarih, L., Gabus, C., Litvak, S., Keith, G. and Darlix, J.L. (1988) EMBO J. 7, 1777-1783 15. Lightfoote, M.M., Coligan, J.E., Folks, T.M., Fauci, A.S., Martin, M.A. and Venkatesan, S. (1986) J. Virol. 60, 771-775 16. Barr, P.J., Power, M.D., Lee-Ng, C. T. , Gibson, H. and Luciw, P. (1987) Biotechnology 5, 486-489 17. Hansen, J., Schulze, T., Mellert, W. and Moelling, K. (1988) EMBO J. 7, 239-243 18. Stanes, M.C. and Yung-Chi Cheng (1989) J. Biol. Chem. 264, 7073 -7077 19. Sallafranque-Andreola, M.L., Robert, D., Barr, P.J., Fournier, M., Litvak, S., Sarih-Cottin, L. and Tarrago-Litvak, L. (1989) Eur. J. Biochem. 184, 367-374 20. Zourgui, L., Tharaud, D., Solari, A., Litvak, S., and Tarrago-Litvak, L. (1986) Biochim. Biophys. Acta 846, 2312-2323 21. Fournier, M., Dorizzi, M., Sarger, C. and Labouesse, J. (1976) Biochimie 58, 1159-1165 22. Fournier, M., (1981), Ph. D. Thesis University of Bordeaux II 23. Nishimura, S. and Weinstein, I. B. (1969), Biochemistry, 8, 832-842 24. Sabin, E.A., Lee-Ng, C.T., , Shuster, J.R. and Barr, P.J. (1989) Biotechnology 7, 705-709 25. Igloi, G.L. (1983) Anal. Biochem. 134, 184-188 26. Vaitukaitis, J.L. (1981) Methods in Enzymology 73, 46-52 27. Laemli, U.K. (1970) Nature (London) 227, 680-685 28. Hawkes, R., Niday, E. and Gordon, J. (1982) Anal. Biochem. 119, 142-147 29. Le, S.Y., Chen, J.H., Braun, M.J., Gonda, M.A. and Maizel, J.V. (1988) Nucleic Acids Res. 16, 5153-5168 30. Raba, M., Limburg, K., Burghagen, M., Katz, J.R., Simsek, M., Heckman, J.E., Rajbhandary, U.L. and Gross, H. (1979) Eur. J. Biochem. 97, 305-318 31. Sprinzl, M., Hartmann, T., Weber, J. et Zeidler, R. (1989) Nucleic Acids Res. r17, rl-r67 32. Barre-Sinoussi, F., Cherman, J.C., Rey, F., Nugeyre, M.T., Chamaret, S., Gruest, J., Danguet, C., Axler-Blin, C., Vezinet-Brun, F., Rouzioux,

436 Nucleic Acids Research C., Rozenbaum, W. and Montagnier, L. (1983) Science 220, 868-870 33. Grandgenett, D.P., Vora, A.C. and Faras, A.J. (1976) Virology 75, 26-32 34. Tanese, N., Sodroski, J., Haseltine, W.A. and Goff, S.P. (1986) J. Virol. 59, 743-745 35. Farmerie, W.G., Loeb, D.D., Casavant, N.C., Hutchison, C.A., Edgell, M.H. and Swanstrom, R. (1987) Science 236, 305-308 36. Lowe, D.M., Aitken, A., Bradley, C., Darby, G.K., Larder, B., Powel, K, Purifoy, D., Tisdale, M., and Stammers, D.(1988) Biochemistry 27, 8884-8889 37. Barat, C., Lullien, V., Schatz, O., Keith, G., Nugeyre, M.T., GruningerLeitch, F., Barre-Sinoussi, F., Legrece, S.F.J. and Darlix, J.L. (1989) EMBO J. 8, 3279-3285