Mutations of a conserved residue within HIV-1 ribonuclease H affect its exo- and endonuclease activities

J. Mol. Hid. (1991) 220, 801-818

Mutations of a Conserved Residue within HIV-1 Ribonuclease H Affect Its Exo- and Endonuclease Activities Birgitta M. Wijhrl, Silke Volkmann and Karin Moel1ing-f MUX Plan& lnstitut fiir Molekulare Genetik, Abt. Schuster Ihnestrasse 73, D-1000 Berlin 33. Germane f Received 5 November

1990: a,ccapte:d 19 April

1991)

The human immunodeficiency virus 1 (HIV-l) reverse transcriptase (RT) is a protein of 66 kJ)a. ~66. which contains t,wo domains. an amino-terminal DNA polymeraso and an RNase PI at t,he carboxy terminus of’ the molecule. In order to characterize the mode of action of the RNase H, two previousfy, described mutant enzymes were used, with substitutions in the highly conserved histldme 539; which was mutated to the neutral amint, acid asparagine and to the negatively charged aspartate. The purified wild-t)ype (wt) and mutant (mt) enzyme activities are analyzed here using RNA-DNA hybrids consisting of it/ &ro t,ranscribed R,NA that harbors the polypurine tract (PPT) from HIV-I and DN’A oligonucleotides complementary to t,he PPT or to other regions of the RNA. Analysis of tlrc~ radioactively labeled RNA of these model hybrids after RNase H treatment) indicates that both. wt and mt enzymes, are capable of cleaving the R’NA in an endonucleolytic manner. The mt enzymes exhibit a severely reduced exonuclease activity. They are more sensitive t)owards salt and competition with excess of unlabeled hybrid. suggest)ing a reduced sub&rate binding affinity. DNA elongation by the RT is coupled with R,NA hydrolpsis by the 3%’ exonuclease of the wt) RNase H. The RNase Hmt of the mt, enzymes lJr,.,yrTyv ,Gry.not exhibit such processive 3-5’ exonuclease activity during DN -? .:,s;xi hi~,-~:~ I/:,: _j 1 “. ; sporadic endonucleolytic cuts, whereas the RT is not affect,ed. 1 1,. .~*.~ic~:ir: I;‘;i\, c’, : ;’ I- : I t,he RNase H mt enzymes exhibit cleavage preferences in the absdi::~*. r>: : :‘.;.+‘~.‘+~ : >‘b,I’i .I synthesis different from those of the wt enzyme. They cannot recognize speciric htiyut*ili:i ..: required to generate a PPT-primer and therefore cannot initiate plus-strand I>?;.~ synthesis in aitro at the 3’ end of the I’PT, which is essential for viral replication. h’eywords: HIV-l

RNase H mutants; reverse transcriptase: sequence specificit;\

1. Introduction

exo- and endonuclrasr.

tion of minus-strand DNA (Omer & Faras, 1982) and in creating a polypurine-rich RNA oligonucleotide at the polypurine tract (PPT) as a primer for the synthesis of plus-st,rand DXA (Champoux et al., 1984; Resnick et al., 1984: Rattray & Champoux, 1987, 1989: Huber &: Richardson. 1990: Luo et al.. 1990; Pullen & Champoux. 1990: W;iihrl & Moelling. 1990). Recent data have shown that during degradation of RNA-DNA hybrids the RNase H of retroviruses exhibits both endonuclease and exonuclease activities, and that during DNA polymerization along an RNA template the RT and RNase H activities can function in a concerted action (Krug & Berg, 1989; Oyama et aE., 1989; Dudding et al., 1990; Huber & Richardson, 1990; Luo et aZ., 1990; Schatz et al.: 1990; Wijhrl & Moelling, 1990). The RNase H domain of HIV-l resides in the carboxy-terminal part of the RT protein (Hansen et al., 1988) and sl?r~ws homology to other retroviral as well as to

Human immunodeficiency virus (HIVS) like all retroviruses possesses a reverse transcriptase (RT) that during replication of the retrovirus serves a variety of functions. The DNA polymerase activity of the RT can extend 3’ ends of a primer using either RNA or DNA as a template (Gilboa et al., 1979). The RNase H activity is an integral part of the RT enzyme. It degrades RNA only in RNA-DNA hybrids (Moelling et al., 1971; Keller & Crouch, 1972; Baltimore & Smoler, 1972; Moelling, 1974). The RNase H has furthermore been implicated in releasing the t-RNA primer necessary for the initiat Author to whom ail correspondence should be addressed. 1 Abbreviations used: HIV, human immunodeficiency virus; RT. reverse transcriptase: PPT, polypurine tract; mt, mutant: wt, wild-type.

801 ~~22--283S/~lj~~O~~ol-1x

$03.00/O

(in 1991 A~wirmic~ Press Limited

B. M. Wiihrl et al.

802

Escherichia coli RNase H (Johnson et al., 1986). RT from HIV-l virions is present as a 66 kDa/51 kDa heterodimer (p66/51) with a common N terminus (Di Marzo Veronese et al., 1986; Lowe et al., 1988). The p66 protein expresses both RT and RNase H activities (Hansen et al., 1988), whereas the ~51 protein lacks RNase H activity and has a strongly reduced RT function (Lori et al., 1988; Hansen et al., 1988; Tisdale et aZ., 1991; Schulze et al., 1991). To characterize further the endonucleolytic and exonucleolytic mode of action of the RNase H we investigated two mutants, each harboring an amino acid change in the carboxy-terminal region of the RT (Tisdale et aZ., 1991). By site-directed mutagenesis His539, which is highly conserved among retrovirus RTs and E. coli RNase H, was either mutated to the neutral amino acid Asn (H-+N) or to the negatively charged Asp (H+D). Previous thermal inactivation kinetics showed that the RNase H function of the H+N mutant (mt) enzyme is ninefold more thermolabile, compared to a twofold difference in the RT activity. In the H+D mt enzyme both activities exhibit about four- to fivefold more thermolability. The H+N mutation has also been shown to abolish virus infectivity (Tisdale et al., 1991). A similar finding was shown by Schatz et al. (19893), who mutated His539 to Phe. In this study the mode of action of the mt proteins was compared to the activities of the corresponding homodimeric p66 wild-type (wt) protein (Larder et al., 1987) and to a p66/51 heterodimeric enzyme (Hansen et al., 1987, 1988) using synthetic RNA-DNA model hybrids as substrates and analyzing their cleavage products in vitro. The model hybrids consisted of an in vitro transcribed RNA and DNA oligonucleotides either complementary to the PPT-region and flanking sequences or to other regions of the RNA with no known specificity. Our results indicate that mutation of His539 has strong effects on the endo- as well as exonuclease functions of the RNase H and on recognition of specific sequences. They also show independence of the RT and RNase H under these conditions, since the RT is unaffected. Analysis of the mt enzymes helps to elucidate the mode of action of the wt.

2. Materials and Methods (a) Materials

Restriction enzymes, the vanadyl ribonucleoside complex and bacterial alkaline phosphatase were purchased from BRL, Berlin, Germany. HaeIII-restricted pBR322 DNA, primer (dT),,, phage T4 polynucleotide kinase RNase-free DNase I, 2’,3’-dideoxynucleoside triphosphates (ddNl?Ps), and 2’-deoxynucleoside triphosphates (dNTPs) were from Boehringer-Mannheim, Mannheim, Germany. RNasin was purchased from Promega Biotech, Heidelberg, Germany. Sequencing reactions were performed with the Sequenase kit from USB, Cleveland, OH. The vector pTZ19R harboring the phege T7 promoter was from Pharmacia, Uppsala, Sweden. Radiolabeled compounds [Y-~*P]ATP (3000 Ci/mmol), [a-‘*P]UTP (>400 Ci/mmol), [c~-~~S]UTPGIS (800 Ci/mmol) and [a-“S]dATPcrS (> 1000 Ci/mmol) were from Amersham

Buchler, Braunschweig, Germany. The RNA transcription kit and T7 RNA polymerase were purchased from Stratagene, Heidelberg, Germany. DNA oligonucleotides were synthesized on an Applied Biosystems oligonucleotide synthesizer and were purified by high-pressure liquid chromatography prior to use. (b) DNA manipulations Restriction endonuclease cleavage, DNA endlabeling of DNA, and transformations were as described by Maniatis et al. (1982). In vitro tion reactions were performed as described by facturer (Stratagene).

isolation, performed transcripthe manu-

(c) Plasmid construction Construction of the plasmid pTZP8 used for in vitro transcription reactions was as described (W6hrl & Moelling, 1990). (d) Expression and pur$cution

of HI V-l RTfRNase

H

The p66 wt protein was isolated as a recombinant protein from E. coli strain TGl harboring plasmid pRT2, a kanamycin-resistant derivative of pRT1, which expresses the p66 polypeptide after induction with isopropyl-/?-thiogalactoside. The 2 mt proteins were prepared from E. coli strain TGI after infection with fresh Ml3 phage stocks containing the mt derivatives of the Ml3 clone mpRT4. After induction with isopropyl-fl-thiogalactoside the 2 mt proteins are synthesized (Larder et al., 1987; Tisdale et al., 1991). The p66/51 wt protein was isolated from E. coli strain HBlOl/ptrpS-pol after induction with indole-3-acrylic-acid (Hansen et al., 1987). The from bacterial lysates by enzymes were purified DEAE-cellulose, phosphocellulose and subsequent poly(U)-Sepharose chromatography as described (Hansen et al., 1987, 1988; Tisdale et al., 1991). The RT/RNase H recovered from the poly(U)-Sepharose column has been shown to be free of E. coli RNase H (Hansen et al., 1987. 1988). The peak enzyme fractions, eluted with 250 mM-NaCl, were stored in 50% (v/v) glycerol at, -20°C. The protein concentrations were determined by scanning microdensitometry using various concentrations of bovine serum albumin as standard and amounted to 6 ng prot,ein/pl. (e) Sequencing reactions Sequencing reactions were carried out with a 32P-endlabeled oligonucleotide primer using the dideoxy sequencing kit from USB according to the manufacturer’s instructions. Portions of the reaction mixtures were analyzed on 8% polyacrylamide TBE-urea gels (see section (f), below). (f) RNase H cleavage of RNA-DNA oligonucleotide hybrids The reactions were carried out in RNase H buffer in a volume of 20 ~1 with @2 pmol of the uniformly labeled model substrates and 6 ng of the corresponding wt or mt RT/RNase H enzymes as described previously, with a substrate to enzyme molar ratio of about 4 to I (W6hrl$ Moelling, 1990). At, 30 min after the onset of the reactions, glycogen was added to a final.concentration of 2 pg/pl and LiCl to @4 M. The reaction products were precipitated with ethanol, the precipitates resuspended in

total

RNase H Mutants with Impaired urea loading buffer (7 M-urea in 1 xTBE, @l% each xylene cyan01 and bromphenol blue; TBE (pH @O) consists of @l M-Tris, 83 mw-boric acid, 1 mm-EDTA) and analyzed on polyacrylamide TBE-urea gels. In the time-course experiment the reaction volume of the assay was scaled up to 100 ~1 and the reaction started by the addition of l .O pmol of RNA-DNA-oligonucleotide hybrid. After the times indicated, 10-p] portions were taken from each sample, heat inactivated (3 min, 95°C) and then precipit.ated and analyzed as described above.

(g) KNOW H reaction coupled with cDNA elongation Reactions were carried out as described (W6hrl & Moelling, 1990) using 12 ng of enzyme and 1.5 pmol of hybrid X labeled to low specific activity, in 10 ~1 of RNase H buffer containing NaCl at a final concentration of 105 mM. The substrate to enzyme molar ratio was about 15 to 1. The reaction products were precipitated as described above and separated on 4% and 10% polyacrylamide TBE-urea gels to allow resolution of the 3’ and 5’ RNA. To check for eDNA elongation the same reaction was carried out, but. with the DNA oligonucleotide ‘*P-endlabeled and unlabeled RNA in the hybrid (W6hrl & Moelling, 1990). Portions were analyzed by electrophoresis on lo?/, polyacrylamide TBE-urea gels. In the time-course experiments coupled with cDNA elongation, the assay was scaled up to 100 ~1 and the enzyme was preincubated with hybrid (30 min, 37°C) where indicated to allow RNase H cleavage first without DNA synthesis. At the end of the preincubation time again 60 ng of enzyme was added. cDNA synthesis was started by the addition of 4 dNTPs to 125 PM. Portions (10 ~1) were taken from the sample after the onset of the reaction at the times indicated, heat inactivated (3 min, 95°C) and then precipitated and analyzed as described above.

(h) UTjRNase

H reactions using an RNA-M13DNA hybrid as aubatrate

Reactions were carried out as described, using the “P-endlabeled oligonucleotide N for the oligonucleotide extension assay and the corresponding sequence reaction with plasmid pTZP8 (WBhrl & Moelling, 1990). Each assay was extracted with phenol, precipitated with ethanol and the pellet resuspended in urea loading buffer.

(i) RT and RNaae H activity teats RT activity tests were performed as described (Hansen et aZ., 1987. 1988) in a total volume of 100 ~1 with 12 ng of t,hr corresponding enzyme, 025 pg of globin mRNA hybridized to oligo (dT),,, and 5 &Ji of [cr-35S]dATPcrS as substrates in RT buffer containing various salt concentrations. RNase H tests were also performed as described (Hansen et al., 1987, 1988) in a total volume of 100 ~1 with an [35S]RNA-M13DNA hybrid (Tisdale et al., 1991) and tinal NaCl concentrations of 65, 105, 145, 185 and 225 mM (values include the salt content of the enzyme). Reactions were started by the addition of 120 ng of the corresponding enzyme. At the end of the reaction the remaining radioactivity of the undigested hybrid was measured by scintillation counting. The initial values of each enzyme found at 65 mM-NaCl were similar and showed the highest RNaae H activity in each case. They were set to 10004 activity for each enzyme.

Exo- and Endonuclease Activities

803

3. Results (a) The exonuclease of the two mt RNase H a&&es

is

reduced The mode of action of the RNase H was investigated with a homodimeric p66 wt RT/RNase H enzyme (Larder et al., 1987) and two p66 mt RT/RNase H enzymes that contained mutations of His539 to Asn or Asp (Tisdale et al., 1991). For a p66/51 heterodimeric wt RT/ comparison, RNase H was used (Hansen et al., 1987, 1988). The RNase H activity of all enzymes was inhibited by the vanadyl ribonucleoside complex proving their viral origin (Krug & Berger, 1989; data not shown). Two RNA-DNA oligonucleotide hybrids (designated V and Y) were used to analyze the effect of wt and mt RT/RNase H activities in vitro. One hybrid represented a region of the viral RNA with no known specificity (V) whereas the other one contained the PPT-RNA and flanking viral sequences (Y) (Scheme 1). Decreasing concentrations of RT/RNase H of the two wt -and two mt enzymes were incubated with these hybrids in the absence of dNTPs so that no DNA synthesis occurred. The RNA cleavage products were then analyzed by electrophoresis of polyacrylamide TBE-urea gels and subsequent autoradiography. The results obtained with hybrid V and hybrid Y are shown in Figure l(a) and (b), respectively. Due to the different lengths of the cleavage products obtained with the two hybrids, the 5’ and 3’ RNA fragments are shown either on one 6% polyacrylamide gel or on two gels with 10% and 4% polyacrylamide, respectively. Determination of the exact cleavage site in case of hybrid Y is based on sequencing analysis described by Wijhrl & AMoelIing (1990) (Fig. l(b)). The resulting 3’ and 5’ RNA fragments indicate discrete endonucleolytic cleavage sites for both hybrids with the wt and mt enzymes. However, the two mt proteins preferentially cleave both hybrids closer to the 3’ end of the RNA (Fig. 1(a) and (b)). The 3’ RNA fragments produced by the mt enzymes are complementary to the main 5’ cleavage sites (Fig. l(a) and (b), top). In contrast, for the wt the 5’ and 3’ RNA cleavage products are not complementary. Both, the shortest 5’ and 3’ RNA fragments appear at the highest enzyme concentrations. This may be interpreted by the action of an endonuclease activity of the wt enzymes followed by an additional 5’-3’ and 3’-5’ exonuclease function, respectively, which gives rise to the shorter 5’ and 3’ RNA fragments. For the mt proteins the exonuclease activities appear to be reduced, thus the larger 3’ and 5’ RNA fragments prevail (see lanes 1 and 2 of wt and mt in Fig. I (a) and (b)). To understand better the details of the differences between wt and mt RT/RNase H activities, a timecourse experiment was performed with incubation periods ranging from 15 seconds to 30 minutes using the PPT-containing hybrid y (Fig. l(c)). As can be seen, the cleavage patterns between wt and mt RT/RNase H again exhibit differences. After 15

B. M. W6hrE et al.

804

--m-----

5’ RNA

x (63-82)

----

‘1 PIT 7=1 ~~b,j*~~*2’ *M $Br?*:,,: ; ,r‘;\x RNA

5’-..-

---m

----w-wv (137-176)

T (207-226)

yr

85

534

3’ RNA

91 1

UAGCCACUUUUUAAMaAAMaOOOOQA CUGGAAGGGCUAAUUCACUAAA.. * - 3’ 3’-TGAAAAATTTTCTTTTCCCCCCTGACCTTCCCGATTAAGT-5’ 3’-AATTTTCTTTTCCCCCCTGA-5’

VrRNA DNA

5’ -“‘AUCUACCACACACAAGGCUACUUCCCUGAUUGGCAGAACUACACACCAGG...-33’ 3’-GGTGTGTGTTCCGATGAAGGGACTAACCGTCTTGATGTGT-5’

T[RNA DNA

5’ -**.UGACCUUUGGAUGGUGCUACAAGCUAGUAC...-3 3’-AAACCTACCACGATGTTCGA - 5’

Scheme 1. The RPu’A-DNA-oligonucleotide substrates X, Y, V and T used in this study are shown. The major RKase H cleavage sites in hybrid Y are indicated by arrows. In all Figures the PPT-region is indicated by a black bar. the cleavage pattern of the 5’ RNA produced by the wt enzymes consists of a series of cleavages, at sites other than nucleotides 84185, but without preference. With the increase of incubation time, the bands representing the shortest 5’ RNA fragment cut at nucleotide 72 and also the shorter 3’ RNA fragments become more intense. This result suggests that the wt enzymes exhibit endo- as well as exonuclease activities within 15 seconds and therefore the endonuclease cut is masked by the action of the exonuclease. The cleavage patterns of both wt enzymes shown are very similar. The two mt enzymes apparently perform a prefer;.!:fial endonucleolytic cut at nucieotides $4/85. The predominant 3’ fiNA is initially complementary. With increase of incubation time the two mt enzymes also give rise to some shorter 5’ and 3 RNA fragments (compare Fig. 1(c), lanes 7 and 8 of mt enzymes to lane 1 of wt enzymes). The time required by the mt enzymes for these additional cleavages is at least 80 to 12O-fold longer than required by the wt enzymes. Apparently the exotluclease activities of the mt enzymes are strongly whereas their endonuclease activities impaired, appear to be only weakly affected in this assay. seconds,

(b) The two mt RNase H enzymes exhibit reduced substrate binding affinities In order to characterize further the two mt RNase H enzymes, they were tested in the presence of increasing salt (NaCl) concentrations using the PPT-containing hybrid Y (Fig. 2(a)). The 5’ (bottom) and 3’ RNA fragments (top) are shown. As can be seen, the cleavage patterns resemble those received with decreasing enzyme concentrations (Fig. l(b)). At low NaCl concentrations, or high enzyme concentrations, the faster migrating shorter 3’ and 5’ RNA fragments prevail. The severely impaired exonuclease activities of the mt enzymes are further reduced at higher NaCl concentrations

(compare 3’ and 5’ RNA fragments of wt and mt in lanes 4 and 5; Fig. 2(a)), whereas the endonucleolytic cut of the mt enzymes is clearly detectable at nucleotides 84/85. The 3’ RNA fragments of the mt enzymes consist of the band complementary to the main 5’ RNA cleavage site (at nucleotides 84/85) and of additional smaller ones. These results again point to the existence of 3’-5’ and 5’-3’ exonuclease activities for both wt enzymes (see lanes 1 to 4 of both wt enzymes), whereas, in the case of the mt RNase H enzymes, the exonucleases are impaired. To quantify the sensitivities of the ~66 wt and mt enzymes t,owards salt. an alternative test was chosen by measuring the overall hydrolysis of radioactively labeled RNA within an RNAMlSDNA hybrid by the RT/RNase H without specifying endo-or exonuclease activity. After the reaction, the undigested hybrid is filtered onto glass fibre discs and its remaining radioactivity measured by scintillation counting (Hansen et al., 1987, 1988). This RNase H activity test allows a quantitative determination of the salt concentration thal leads to 500/o inhibition of the RNase H activity. Such a reaction is shown in Figure 2(b). Inhibition of the RNase H activities to 500/, is reachrd at about 150 mM-Nac!l for the two mt enzymes and at 250 mM for the p66 wt RNase H, respectively. Therefore, the two mt RNase H activities can apparently be characterized by reduced substrate binding affinities. A similar assay was performed to measure the dependence of the RT activity on salt concentration, whereby both, wt and mt RT activities are inhibited to 50% at 120 mM-salt (data not shown). reduced substrate binding To substantiate further, an additional approach was taken. It has been described earlier that the RNase H of p66/51 cannot be as easily competed for by additional hybrid substrate as the ~15 RNase H (Hansen et al., 1988). This property was tested here for one wt and mt RT/RNase H enzymes. The RNase H assay was

RNwse H Mutants with Impaired

p66

RNase H 2I2

3456’1

j

805

Exo- and Endonuclease Activities

p66 H+N WiA’

j !

P66 H+D

23456!123

45611

p66/51

I

’ 274

3’

587 RNA 504 uncut 434 3’ RNA fragments

267 234 213 192 184 5’ RNA fragments ....

124

-

(a) Fig. 1.

performed with radioactively labeled model substrate Y and was competed for with a tenfold excess of unlabeled hybrid before and 60 seconds after the initiation of the reaction by the addition of enzyme. Incubation was allowed to continue for 20 minutes (Fig. 2(c)). Analysis of the reaction products on a 10% polyacrylamide TBE-urea gel shows that both wt and mt RNase H activities, can be competed after the onset of the reaction. Especially the mt enzymes exhibit after competition a decrease of smaller 5’ RNA fragments, suggesting further reduction of the 3’-5’ exonuclease activity due to their lower substrate binding affinities. These results resemble those shown in Figures l(a) and (b) and 2(a) and (b) with lower enzyme or higher salt concentrations.

(c) Concerted action

of

RT/JLVasr JJ

A concerted a&ion of RT and RKase H activities has been demonstrated recently for the wt HTV-I RT/RNase H (Schatz et al., 1990; Wijhrl & Moelling, 1990). Therefore, we wanted to determine whether the RNase H activity of the mt enzymes behaved in the same way. Both wt and mt enzymes were allowed to react with model hybrid T, which contains random sequences (see Scheme 1), under conditions where primer extension is possible. For comparison some of the reactions with the p66/51 wt are included. Either all four dNTPs or three dNTPs plus one ddNTP which leads to chain termination (Sanger et al., 1977) were added to the assay. The resulting 3’ and 5’ RNA cleavage products were analyzed on a 10% polyacrylamide

806

B. M. Wiihrl et al.

p66

p66

: p66

I p66/51 i

-

124

91 84/M

89

’ RNA t30; -agments

79 72

-

-, 1

(b) Fig. 1.

TBE-urea gel (Fig. 3(a)). Depending on the ddNTP added, extension of the DNA oligonucleotide by one, three, four or five deoxynucleotides was achieved. The DNA polymerization by the RT is accompanied by a simultaneous enzymatic hydrolysis of the RNA by one, three, four or five nucleotides with the wt proteins (compare Fig. 3(a) lanes 1 to 4 and lane 19). After the addition of all four dNTPs, all cleaved 5’ RNA is degraded by the wt proteins (lanes 5 and 20), whereas the 3’ RNA fragments remain intact (Fig. 3(a), top). Analysis of

the 5’ RNA fragments created by the mt enzymes after addition of ddNTP gives rise to cleavage patterns similar to those of the wt (Fig. 3(a), bottom, lanes 7 to 10, and 13 to 16) except that RNA hydrolysis by the mt RNase H activities is reduced, leaving a larger proportion of the input RNA uncut (Fig. 3(a), top, lanes 7 to 10, 13 to 16). When all four dNTPs are added the mt enzymes perform only a few endonucleolytic cuts, which give rise to distinct 5’ RNA fragments; the most prominent band is indicated by an arrow (Fig. 3(a),

RNase H Mutants with Impaired

~66 6-2

P66 H-+N a:______1

Exo- and Endonuclease Activities

j

~12345678;12345678~12345678,12345678,M

P66 H+D

-

807

~66151 wt

-13 time ,540 q-

3’

RNA uncut

- 504 T 458

3’ RNA -i 434 fragmer

5

- 104 91 84/85

89

f

*’

! ...I

72

-

4

5’RNA fragments

-64 - 57

5’

w

Figure 1. Parameters affecting the cleavage patterns of model substrates V and Y with wt and mt enzymes. (a, b) Hependence on enzyme concentrations. (a) Hybrid V (0.2 pmol) was treated with 24, 6, 3, 1.5, 075 and 0375 ng of the caorresponding wt or mt RT/RNase H enzyme (lanes 1 to 6) at 37°C for 30 min. The resulting 3’ and 5’ RNA cleavage products were separated on a 6% polyacrylamide TBE-urea gel. The schematic drawing on the left shows the hybrid used. The arrows in (a) on the left point to the cleavage sites in the hybrid. The first and the last nucleotide of the hybrid region are indicated by numbers. (b) Hybrid Y was treated with 24, 12, 6, 3 and 1.5 ng of the corresponding wt or mt RT/RNase H (lanes I to 5). Portions of the resulting 3’ and 5’ RNA cleavage products were separated on a 40/O (top) or IOo/, (bottom) polyacrylamide TBE-urea gel, respectively. The numbers on the left indicate the size of the 6 RNA fragments, the arrows point to the cleavage sites. Nucleotide 79 corresponds to the next to last nucleotide at the 3’ end of the PPT-RNA. -RT indicates incubation of hybrid without enzyme. M represents HueHI fragments of pBR322 as molecular weight markers. Fragment sizes are indicated on the right. The stars in the RNA symbolize uniform radioactive labeling. (c) Time-course analysis. Hybrid Y (1 pmol) was incubated with 30 ng of wt, or mt enzyme in RNase H buffer at a final BaCl concentration of 65 mM in a total volume of 100 ~1 of RT/RNase H. After 15 and 30 s, 1, 2, 5. 10. 20 and 30 min IO-n1 portions were taken from the reaction mixture (lanes 1 to 8) and analyzed on a 49;) (top) or loo+, (bottom) polyacrylamide TBE-urea gel, respectively. For explanation of numbers, see above.

bottom, lanes 11 and 17). Thus the mt enzymes appear to lack the processive 3’-5’ exonuclease activities. To prove that DNA primer elongation has really taken place, a control experiment was carried out, in which the DNA oligonucleotide of model substrate T was labeled radioactively instead of the RNA strand and extended under the same condi-

tions as described above in the presence of all four dNTPs, or three dNTPs and one ddNTP, which leads to an elongation by one, three, four or five deoxynucleotides (Wijhrl & Moelling, 1999). Analysis of the DNA reaction products proves that the mt enzymes are able to carry out DNA polymerization just like the wt enzymes under the chosen assay conditions (Fig. 3(b)).

808

B. M. W6hrl et al.

2 540 . RNA uncut - 504

IqT458

3’ RNA .___, fragments - 434

- 124

-104

.-.: 4 89

-91 = 84/85 --

79

--

72

5’ RNA fragments

(a) Fig. 2.

When the analogous reaction was performed using hybrid X, which contains the PPT-RNA sequence (see Scheme l), the result obtained with the wt enzymes was similar, extension of the DNA by one, four, five and eight deoxynucleotides leads to the equivalent shortening of the RNA (Fig. 3(c)). The mt enzymes, however, exhibit some surprising differences that may be attributed to this particular hybrid containing the PPT. An endonucleolytic cut can be found when the DNA oligonucleotide is

elongated by only one deoxynucleotide (Fig. 3(c) lanes 8 and 14); however, extension of the DNA by four, five and eight deoxynucleotides did not lead to significant equivalent cuts in the RNA (Fig. 3(c), lanes 9 to 11 and 15 to 17). Here the endonuclease appears to be inhibited. Again, in the presence of all four dNTPs the mt RNase H proteins induce some sporadic endonucleolytic cuts (Fig. 3(c), bottom part, lane 7 or 13), one of them apparently circumventing the PPT region (see arrow above the marker

RNase H Mutants with Impaired

Exo- and Endonuclease Activities

809

this case as well as in Figure 3(a). When a preincubation period is allowed, both the wt and the mt RNase H perform a strong endonucleolytic cut at the 3’ end of the PPT at nucleotide 79 (Fig. 3(e). right). The RNA fragment originating from this endonucleolytic cut by t’hr wt rnzyme disappears with time due to its RKase H exonnciease activit\,. The mt, RNase H shows with increasing t,lme ad& tional distinct 3’ and 5’ RNA fragments somtb of which are the same as in the experiment wltjhout preincubation (arrows and dotted lines). in caw of the wt the bottom part, of the gei allows the tietect,ion of some degradation product’s (Fig. 3(e) right. bot,tom bracket,). These RNA fragments may ht. produced when thts R,T reaches the end of t,htx RNA template. We have previously c~akulatetl t,hat thrs active centers of the RT and RNasr H are ahout 1% nucleotides apart. corresponding roughly to thr size of the reaction products (Wiihrl K! iIlorl11ng. l!NOi

O> 50

100

150 200 NaCl [mM]

250

( bi

Fig. 2.

nucleotide 57). The cause of the other sporadic cleavage sites is not obvious. The equivalent 3’ RNA fragments are detectable and longer than in the case of the wt (Fig. 3(c) top, lanes 7 and 13). Again, DNA elongation is shown as control and indicates that DNA synthesis of wt and mt enzymes is indistinguishable (Fig. 3(d)). In order to elucidate the mechanism of the reaction and to investigate whether the RNase H of the mt has also an impaired endonuclease, as might be suggested by the data, a time-course analysis allowing DN4 synthesis was performed. Two approaches were taken. one in which four dNTPs were added at the beginning of the reaction and the other one in which the four dNTPs were added after a preincubation time of 30 minutes (Fig. 3(e) left and right,, respectively). The reactions were performed with the p66 wt and one mt enzyme. The result indicates that the mt RNase H leads to a few distinct 5’ RNA cleavage products that become shorter with increasing time (Fig. 3(e), lanes 9 to 16). This suggests that the mt RNase H apparently cleaves the RNA endonucleolytically at certain sites during RT-directed DNA synthesis and not afterwards. Thereby the complementary 3’ RNA fragments are created (Fig. 3(e) top, lanes 9 to 16). The endonucleolytic cut at nucleotide 79 performed by the wt cannot be detected in case of the mt enzyme. The mt enzyme excludes the PPT sequence and sets the first endonucleolytic cut around the 5’ end of the PPT (Fig. 3(e), arrow below the 64 marker). The specificity of the mt enzyme for the other cuts that lead to the shorter 5’ RNA fragments is unknown in

Previous in v&o analysis with a pretormed DNA-RNA hybrid covering the T’PT-RNA has demonstrated that the HTV-1 wt RTjRKase H ih capable of initiating plus-strand J)NA synthesis in the presence of’ deoxyribonucleot~idrs. This sssaj was repeated with the p66 wt’. one of’ the t,wo mt enzymes (H -+N) and the p66/5f het~erotiinrer us14 previously (W6hrl $ Moelling, 1990: Fib!. 1). As can be seen, the mt enzyme does not s!ktic*aliy c*lravt~ at the 3’ end of the PPT in cant rast t)o t hr wt. indicating that, t,he mutation affects recognition of the PPT-specific hybrid structure. This may &her be a consequence of the reduced suhstratr binding affinity or of the reduced catalytic a,c*tlvity of the mt RNase H. The predominant cleava,gcA sitr at the 3’ end of the PPT is located above the (‘.(i pair, (Fig. 4, upper arrow). Previouslv. we ant1 others showed that cleavage below the (“. (; pair appeared to be preferred (Huber & Richardson. t 990; Putlen & Champoux, 1990: Wiihrl $ Jloellir~g. 1990; Fig. 4, a difference which suggests some lower arrow), imprecision of the RNase H. also oi)srrvetl her Hu ber & Richardson ( 1990).

4. Discussion The results presented suggest that mutations in His539 of the HTV-1 RNase H impair t’he RNase H without detectable effect on the RT activity. The mutations lead to a significant reduction of the RNase H exonuclease cleavage rate. Kven after 30 minutes the mt RNase H activities have not .vet generated a cleavage pattern produced by the wt in 15 seconds (Fig. l(c)). We interpret the differences in the cleavage patterns of the mt RNasr H activities by a reduced cleavage rate of the rxonuclease activity. An alternative interpretation that not an exonuclease but a series of endonucleolytic cuts generates the distinct RNA cleavage products seen for the wt (Fig. l(a) and (b)) can he ruled out by analysis of the mt enzymes. which lead to fewer cuts

B. M. W6hrl et al.

810

competition

.124 104

80 5’ RNA fragments

64

5’

Figure 2. Parameters affecting the substrate binding affinity. (a) Dependence on NaCl concentrations. The p66 wt and mt proteins (6 ng) were allowed to react for 30 min at 37°C with 62 pmol of model hybrid Y at final concentrations of NaCl of 25, 65, 105, 145 and 225 mM (lanes 1 to 5). For comparison, the p66/51 wt was also included in the test as described (lanes 1 to 5, right) (Wiihrl & Moelling, 1990). The resulting 3’ and 5’ RNA fragments were analyzed on 4% (top) or 10% (bottom) polyacrylamide TBE-urea gels. For explanation of other details, see Fig. I. (b) Quantification of NaCl sensitivity. Overall RNA hydrolysis in an [3’S]RNA-M13DNA hybrid by the p66 wt and mt enzymes was measured at final NaCl concentrations of 65, 105, 145, 185 and 225 mM-NaCl. The RNase H activity was quantified by acid precipitation and collection of the undigested hybrid on filters and subsequent determination of the radioactivity in a scintillation counter. The initial RNase H activities at 65 mM-NaCl of all enzymes were similar and set to 100% for each enzyme. (c) Competition with unlabeled substrate. Radioactively labeled hybrid Y (62 pmol) was incubated for 30 min at 37°C with 6 ng of the wt or mt enzymes in standard RNase H reaction mixtures containing NaCl at a final concentration of 65 mM (lane 1). A lo-fold excess of unlabeled hybrid (2 pmol) was added either before or 60 s after starting the reaction by the addition of the corresponding RT/RNase H (lanes 2 and 3, respectively). The resulting reaction products were analyzed on a 10% polyaerylamide TBE-urea gel. For explanation of other details, see Fig. 1.

RNase H Mutants

I

with Impaired

X11

Act6uitie.s

Exo- and Endonuclease

2 3 4 5 6 7 8 9 IO II 121I3 I4 IS I6 I7 IS’ I9 20 21 22 23

3’

/ 540 MA uncut - 434 3’ RNA ..i/ fragments

“7

- 267

1I : :I : : :I ; 3 207 : +I+ 226

- 234

.: $z

+ 192 j- 184

[ 5’RNA I fragments

I

+3; +4 +S $

5’

p66 wt

p66 H+N

(a) Fig. 3.

:

P66 H+D

~p66/;i .,,,, wt

‘,

t

89

$

80

812

B. M.

Wiihrl

et al.

1932112223

I 23 4

+206 4 k f I

234 213 184

226 5’

124 104 89 80

64 57 51

25 21 18 11

21 20

P66 wt

3’

P66 H+N

1

(b) Fig. 3.

p66 HjD

b66/5i, i wt “.;

RNase H Mutants with Impaired

Exo- and Endonuclease Activities

E

E-

X

813

89

: I 3 I i 3 b : I I : I I +I +4 +Si +8+ 5’ RNA fragments

P66 WI

P66

H+N (cl Fig. 3.

P66

p66/5 I

H+D

WI

B. M. Wiihrl et al.

814

M’I

23

UQ

I

23

-

I

2

”

3

-

I

82

-28

21 18 -

I

Figure 3. Coupling of RT and RNase H activities. (a) Concerted action of RT and RNase H using hybrid T. RNase H reactions were carried out in RNase H buffer in a total volume of 10 ~1 containing NaCl at a final concentration of 105 mM, 15 pmol of RICA-DNA hybrid T (see Scheme 1) labeled to low specific activity (Wiihrl & Moelling, 1990) and 12 ng of wt or mt enzymes. The assay was performed in the absence or presence of 4 dNTPs or 3 dNTPs plus one ddNTP as described (WGhrl & Moelling, 1990). Incubation was for 30 min at 37°C. The 3’ and 5’ RNA cleavage products were separated on a 10% polyacrylamide TBE-urea gel. Lanes 1 to 4, 7 to 10, 13 to 16 and 19 show incubations of wt and mt RT/RNase H enzymes in the presence of 1 ddNTP as indicated and the other 3 dNTPs (ddA, ddG, ddC and ddT allow respectively). Lanes 5, 11, 17 and 20 show the elongation of the DNA primer by 5, 1, 4 and 3 deoxynucleotides, fragments obtained in the presence of all 4 dNTPs and lanes 6, 12, 18 and 21 in the absence of all 4 dNTPs. Lane 22 shows hybrid without enzyme (-RT), lane 23 contains Hue111 fragments of pBR322 as molecular weight markers (M). The start point of RNA hydrolysis and the nucleotides incorporated during DNA elongation and concomitant RNA hydrolysis are indicated on the left. The 3’ RNA fragments generated during the reaction are indicated by brackets, the 5’ RNA fragments by brackets and arrows (right). (b) Evidence for cDNA elongation using hybrid T. RT/RNase H reactions were carried out as described in (a) with hybrid T and wt and mt enzymes as indicated, but with unlabeled RNA and 32P~endlabeled DNA oligonucleotide in the hybrid. The DNA products were separated on a 10% polyacrylamide

RNase H Mutants with Impaired

H+N

__“-‘--12 3 4 5 67

~(56

~(36

P66 I

Exo- and Endonuclease Activities

R 91011

1213141S16

____? 1718

wt

x15

p66

-time

H-+N

1920212223242526172029M31323334353637’38

79

89

-80

19

-64 -

’

51

-

21

-

IX

-

II

:5’

THE-urea gel. Numbering of the lanes corresponds to those of (a). The number of nucleotides incorporated and the total length of the DNA copies obtained are indicated on the left, nucleotide markers are indicated on the right (M). (c) Concerted action of RT and R&se H using hybrid X. RNase H reactions were carried out as described in (a). Lanes 1. 7, 13 and 21 show 3’ and 5’ RNA fragments obtained in the presence of all 4 dNTPs, and lanes 6, 12. 18, 19 in the absence of dNTPs with wt and mt RT/RNase H enzymes as indicated. Lanes 2 to 5, 8 to 11, 14 to 17 and lane 20 show incubations of enzymes in the presence of 1 ddNTP as indicated and the other 3 dNTPs (ddA, ddG, ddT and ddC allow elongation of the DNA primer by 1,4,5 and 8 deoxynucleotides, respectively). For lanes 22, 23 and description of scheme see (a). (Determination of the cleavage site at nucleotide 80 refers to sequencing analysis described by Wiihrl & Moelling (1990)). (d) Evidence for cDNA elongation using hybrid X. RT/RNase H reactions were carried out as described in (b). with the p6S wt and mt enzymes. Lane 1 shows hybrids incubated with wt and mt enzymes in the absence of dNTPs and lane 2 in the presence of 4 dNTPs. Lane 3 shows DNA products obtained after incubation of wt and mt enzymes with 3 dNTPs and ddCTP, which leads to a primer elongation by 8 deoxynucleotides from 20 to 28. The number of nucleotides incorporated and the total length of the DNA copies obtained are indicated on the right. The size of the Hue111 fragments of pBR322 used as molecular weight markers as indicated on the left (M). (e) T’ime-course analysis. Reactions were carried out as described in (a) with the ~66 wt and the H -+K mt enzyme. The assay was scaled up to 100 ~1. cDNA synthesis was allowed by the addition of dNTPs to 125 PM, either without preincubation (left) or with a preincubation period of hybrid and enzyme of 30 min (right). To the latter assay an additional 120 ng of enzyme was added before the start of cDNA synthesis. Portions (10 ~1) were taken from the reaction mixtures before the addition of dNTPs (right, lanes 20 and 29, wt and mt enzyme, respectively) or 15 and 30 s, 1, 2, 5, 10, 20 and 30 min afterwards (left, lanes 1 to 8 and 9 to 16 for wt and mt without preincubation; right, lanes 21 to 28 and 30 to 37 for wt and mt after preincubation, respectively). Portions were treated as described in (a) and applied to a 4% or 10% polyacrylamide TBE-urea gel to analyze the 3’ and 5’ RNA cleavage products, respectively (see top and bottom of panels, respectively). Lanes 17 and 19 contain hybrid without enzyme. Lanes 18 and 38 contain pBR322 Hue111 fragments as molecular weight markers, Fragment sizes are shown on the right of each panel. Arrows on the right point to the 3’ (top) and 5’ RNA fragments (bottom). The bracket at the bottom of the right panel indicates end products of RNA hydrolysis. The scheme on the left symbolizes RNA-directed cDNA synthesis, the arrows point to the preferential RNA cleavage sites of the mt enzyme.

B. M. Wiihrl et al.

816

dr (A

U)5

(T

A)4

CG (T

A)4

(C G)5 CGZ TA GC AU (C G)2 (T A)2 CC G)3 12345678 Figure 4. Site-specific initiation of plus-strand DPU’A in an M13DIGA-RNA hybrid. An M13DNA-RKA hybrid harboring the PPT region was incubated with 4 dPu’TPs without RT/RKase H (lane 2) or with 120 ng of the p66 wt (lane 4) or the H + ;hu’mt enzyme (lane 3), respectively, to allow synthesis of plus-strand DNA. To determine the initiation site of the plus-strand DPU’A, a primer extension assay was carried out after the remaining RPr’A had been removed by treatment with alkali. A 32P-endlabeled oligonucleotide K complementary to a, sequence 66 to 86 bases downstream from the PPT was annealed to the plus-strand DKA and the primer extended up to the start point of plus-strand DKA synthesis by means of the DKA polymerase. Sequenase was as described (W8hrl & Moelling, 1990). The reaction products were applied to an 8% polyacrylamide sequencing gel together with the products of dideoxy sequencing reactions as indicated carried out plasmid pTZP8 and the same 32P-endlabeled primer oligonucleotide N (lanes 5 to 8). The reaction performed with the pSS/ 51 wt published recently (WBhrl & Moelling, 1990) is shown for comparison (lane 1). The bold letters C and G indicate the end-points of the primer extension products. Arrows mark the start-points of the plus-strand DNA

synthesis.

than with the wt enzyme and generate smaller 5’ RNA fragments with time (Fig. l(c)). If only endonuclease activities were involved one would have expected a distribution of the various fragments for the initial time points of the experiment also for the mt protein. The slowly reacting mt enzymes therefore help to resolve the mode of action of the fast WI enzyme. They allow a retarded view of the catalyti!. activity of the wt, enzyme. All the hybrids analyzed here require at least one endonucleolytic cut before the exonuclease can proceed. At first sight the endonuclease of the mf RNase H enzymes does not appear to be altered in its activity (Fig. I(a) and (b)). However, when different concentrations of monovalent ions were applied, the mt RNase H appeared to be morca sensitive (Fig. 2(a) and (b)). Thus, the substrate binding affinity of the mt RNase H required for th(L endonuclease function is reduced. The mt RNase H is. in agreement with this property. also more sensi tive towards competition with unlabeled hybrid after the onset of the reaction (Fig. 2(c)). The reduced substrate binding affinity of the mt enzymes is probabiy also responsible for the prominent difference between wt and mt, RNase H that hecomes apparent when limited 1)NA synthesis is allowed by the addition of ddNTPs. The mt RNase H enzymes are not capable of performing endonucleolytic cuts inside the PPT region under these conditions (Fig. 3(c)), only with .a random sequence containing hybrid can endonucleolyti!* cleavage take place (Fig. 3(a)), A time-course analysis of this reaction and an analysis of the sequence of events by performing a preincubation before initiation of DNA synthesis aliowed us to understand the properties of the ml, cndonucleases (Fig. 3(e) without and with preincubation). The mt endonuclease cleaves at the 3’ end of the PPT with a strongly reduced cleavage rate. With DNA synthesis allowed immediately. the mt RT/RNase H molecule moves too rapidly to give the endonuclease a chance for cleaving at nucleotide 79, like the wt enzymes. During DNA synthesis the mt enzyme cleaves the RNA sporadically, once around the 5’ end of the PPT (Fig. 3(e)) and at a few other sites, the specificity of which is unclear. Thus the mt enzymes differ from the wt in three respects: by a reduced exonuclease cleavage rate, by a reduced substrate binding affinity and by a reduced endonuclease activity at specific sequences. This latter property may be a consequence of the second one, the reduced substrate binding affinity. The specificity for the PPT-RNA-DNA hybrid could be due either to its nucleotide sequence or to conformational properties. Analyses by Rattray & Champoux (1989) lend support to the second possibility, since minor sequence changes within the PPT had no influence

on cleavage

site selection.

The

sporadic

endonucleolytic cuts that give rise to the distinct 5’ RNA fragments with the mt enzymes (Fig. 3(a), (c), (e)) may arise either at specific sequences or may reflect some pausing of the RT during DNA elongation that allows the mt RNase H to set a cut. Since

&Vase H Mutants with Impaired t,he active center of the RT is 18 nucleotides apart from that of the RNase H (Wiihrl & Moelling, 1990), we analyzed the sequence of the RNA 18 nucleotides away but were not able to detect any specific properties. Secondary RNA structures may play a role in retarding the RT. Only recently, some sequence dependence of the RNase H has been suggested (Furfine & Richardson, 1991), in agreement with our observations. A consequence of reduced substrate binding affinity is also reflected by t,he inability to initiate plus-strand DNA synthesis at the PPT. which we demonstrate for one of the mt enzymes (Fig. 4). It is worth noting, that the PPT-containing hybrid is cleaved by the wt RNase H within the I’PT (see Fig. 2(b) or 3(c)), when limited DNA synthesis or no DNA synthesis is allowed, whereas initiation of plus-strand DNA synthesis starts at the 3’ end of the PPT without apparently destroying it (Fig. 4). The two experiments cannot be directly caornpared with each other because primer extension of PPT-containing hybrids may force the RNase H to perform cleavages that are imposed on it for steric reasons such as the distance between RT and Without these catalytic centers. itNase H constraints initiation of DNA synthesis at the PPT occurs correctly. Results deduced from crystal structure analysis of t,hc: RNase H from E. coli suggest that the highly conserved His residue mutated in the two mt enzymes is not part, of the putative active site (Katayanagi et al.. 1990; Yang et al., 1990). It is apparentlv located there at the rim of the surface t:avity that harbors the conserved triad (Asp, Glu, Xsp) that probably represents the catalytic center of t)he 6. coli RNase H (Kanaya et al., 1990; Kat,ayanayi et ai.. 1990). Mutational analyses of ? ht,s:e r,hrer amino acid residues from H and oi’ the analogous three Chase :&tines irom the HTV-I RNase H

the E. coli amino acid support the

!;otion that, the KNase H from E. coli and HIV-l :!re cioselv related (Schatz et al., 1989a; Kanaya et :ci.. 1990;” >lizrahi of nl.. 1990). Therefore, the effect, t’ His.539 from HIV-1 KNase H on the catalytit .i(‘I rvit,v i:: prot)ahly Indirect. and an involvement c(,r:le klncl of subst,rat,r interaction appears to l:‘
in be

Exo- and Endonuclease Acti,vities

x17

Champoux, J. J., Gilboa. E. & Baltimore. D. (1984). Mechanism of RPU‘A primer removal by the RPjase H activity of avian myeloblastosis virus reverse transcriptase. J. Viral. 49. 686-691. Di Marzo Veronese, F., Copeland, T. I).. De Vito. -4. I,.. It., Oroszlan. S., (:allo. R. (‘. 8: Rahman, Sarngadharan, M. G. (1986). (Characterization of highly immunogenic p66/p51 as t,he reverse tranHTLV IIT/LAV. Scirncr?. 231. scriptase of 1289-1291. Dudding, L. R., Harington. A. & Mizrahi. \r. ( 1990). Endoribonucleolytic cleavage of RNA: oligodeoxynucleotide hybrids by the ribonunlease H activity of Biochem. Biophys. fies. HIV- 1 reverse transcriptase.

Commun. 167, 244-250. Purfine. E. S. Dz Reardon. J fi:. ( 1991). Reverse transcriptase. RNase H from the human immunodeficiency virus, Relationship of t,he i)RiA polymerase and RR;A hydrolysis activit,ies ./. H1:ol. (‘hem.

266. 406G412. Gilboa, E., Mitra, 8. W.. Gaff. S. & Baltimore, I). ( 1979). 9 detailed model of reverse transcription and tests of crucial aspects. CleEI. 18. 93- 100 Hansen: ,J.. Schulze, T. & Moelling, K. (1987). RXase i( activity associated with bacterially expressed reverse transcriptase of human T-cell Iymphotropic virus virus. .I. Biol. III/lymphadenopathy-associated Chem. 262, 12393-12396. Hansen, J., Schulze. T.. Mellert, W. bt Moelling, K. (1988). Identification and characterization of HIV-sperific RXase H by monoclonal antibody. EMBO .I. 7.

239-243. Huber, H. E. & Richardson. C. C. (1990). Processing of the primer for plus strand DNA synthesis by human immunodeficiency virus I reverse transcriptase.

J. Viol. Chem. 265, 10565~10573. Johnson, M. S., McClure. M. A.. Feng, I). F.. Grav. ,I. & Doolittle, R. F. (1986). Computer analysis or” retroviral pot genes: assignment of enzymatic functions to specific sequences and homologies with nonviral .Vat. =1rctcl. SC/., I’.S.ll. 83. enzymes. Proc.

i648-765%. Kanaya, S.. Kohara. A.. Miura. Y.. Sekiquchi. .-\.. Iwal. S.. inour. H.. Ohtsuka. E. & Jktahara. M. I 1990). Identification of the atnino acs~d residues involved in an active site of t/:scheri&o co/i rihonuc+astA M 1)) site-directed mutagenesis. .i Niol. f ‘henl 265. 16 1Fi--462 I Katayanagi, K.. Miyagawa. .\I.. Ylatsushima. )I.. Tshikawa, 31.. Kanaya. S.. Tkehara. Al.. Mat,suxakr. ‘I’. cUL. Mnrikawa. K. (l!,YO). Three tllmensionai structure I)f ribonuc~lease H from ti. r&i .\‘mtlcrc { /,ondtwr/,

347. 306-309. ‘E’;!‘. :tutnors are grateful to Dr M. Tisdale for supply of r: t r!liItalll, c,onst,ructs and lt. Oeltjen for excellent t.echI. .:..I! CGxt,ance. We also thank E. Philippi for preparing ;’ I- drawings. This work was supported by a grant from 7 +:#a J?~:ndrnminister fiir Forschung und Technologie :!i3iFT) FKZII-019-86 and the Dr Mildred Schee] stittung PA6.

References Baltimore. D. & Smoler. D. F. (1972). Association mdoribonuolease with the avian myeloblastosis deoxyribonucleic acid polymerase. J. Biol.

247, 7282-7287.

of an virus

Chem.

Keller. W. $ DNA-RNA polymerases

Crouch. IC. (1972). I)rgradation of hvbrids h,v ribonucleasta H and l>;\jA oi cellular and viral 0rigzIn. i’roc. .L’&.

rlcud. Sri., 1:.S.d. 69. X60-333364. Krug.

MM. S. 62 Berger. d. 1,. (1989). Itihonuc+asr H activities associated with viral rtavyrse transcriptasrs are endonucleasrs. Proc. Sat. <~cYu~. SC+., i..S.,-l 86. 33539--3.543. Larder. I$.. Purifoy. I).. Powell. K. & Dart)>,. t :. (1987). .AlI)S virus reverse transcriptaar tfefinrd bv high level expression in E. coli. EMBO ,/. 6. 6133&~137. Lori, F.. Scovassi. A. I.. Zella. Lt., ;\chiili, C:., (latt,anro. E.. (lasoli. I’. & Uertazzoni. I’. (1988). Enzymatlcaily active forms of reverse transcariptase of the human immunodeficienc*v virus. .-I I IAS IZrs. Hum.

/Zetrovimser, 4. i93-398.

818

B. M.

Wiihrl

Lowe, D. M., Aitken, A., Bradley, C., Darby, G. K., Larder, B. A., Powell, K. L., Purifoy, D. J., Tisdale. M. & Stammers, D. K. (1988). HIV-l reverse transcriptase: crystallization and analysis of domain structure by limited proteolysis. Biochemistry, 27, 8884-8889. Lou, G. X., Sharmeen, L. & Taylor, J. (1990). Specificities involved in the initiation of retroviral plus-strand DNA. J. Viral. 64, 592-597. Miniatis, T., Fritsch, E. F. & Sambrook, ,J. (1982). Molecular Cloniny: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Mizrahi, V., Usdin, M. T., Harington, A. & Dudding. L. R. (1990). Site-directed mutagenesis of the conserved Asp-443 and Asp-498 carboxy-terminal residues of HIV-l reverse transcriptase. Nucl. Acids Res. 18, 535995363. Moelling, K. (1974). Characterization of reverse transcriptase and RNase H from Friend-murine leukemia virus. Virology, 62, 46-59. Moelling, K., Bolognesi, D. P., Bauer, W., Biisen, W., Plassmann, H. W. & Hausen, P. (1971). Association of viral reverse transcriptase with an enzyme degrading the RNA moiety of RNA-DNA hybrids. Nature New Biol. 234, 240-243. Omer, C. A. & Faras, A. J. (1982). Mechanism of release of the avian retrovirus tRNATrp primer molecule from viral DNA by ribonuclease H during reverse transcription. Cell, 30, 797-805. Oyama, F., Kikuchi, R., Crouch, R. ,J. & Uchida, T. (1989). Intrinsic properties of reverse transcriptase in reverse transcription. Associated RNase H is essentially regarded as an endonuclease. J. Biol. Chem. 264, 18808-18817. Pullen, K. A. & Champoux, J. J. (1990). Plus-strand origin for human immunodeficiency virus type 1: implications for integration. J. Viral. 64, 6274-6277. Rattray, A. J. t Champoux, J. J. (1987). The role of Moloney murine leukemia virus RNase H activity in the formation of plus-strand primers. J. Viral. 61, 2843-2851. Edited

et al.

Rattray, A. J. & Champoux, J. J. (1989). Plus-strand priming by Moloney murine leukemia virus. The sequence features important for cleavage by RNase H. J. Mol. Biol. 208, 445-456. Resnick, R., Omer, C. A. t Faras, A. ,J. (1984). Involvement of retrovirus reverse transcriptase-associated RNase H in the initiation of strong-stop (+) DNA synthesis and the generation of the long terminal repeat. J. Virol. 51, 813-821. Sanger, F., Nicklen, S. & Co&on, R. A. (1977). DNA sequencing with chain terminating inhibitors. Proc. Nat. Ad. Sci., U.S. A. 74, 5463-5467. Schatz, O., Cromme, F. V., Griininger-Leitch, F. & Le Grice, 6. F. J. (1989a). Point mutations in conserved amino acid residues within the C-terminal domain of HIV-l reverse transcriptase specifically repress RNase H function. FEBS Letters, 257, 311-314. Schatz, O., Cromme, F. V., Naas, T., Lindemann, D., Mous, J. & Le Grice, S. F. J. (19893). Inactivation of the RNase H domain of HIV-l reverse transcriptase blocks viral infectivity. In Gene Regulation and AIDS (Papas, T., ed.),pp. 293-303, Portfolio Publishing, Houston, TX. Schatz, O., Mous, J. & Le Grice, S. F. J. (1990). HIV-l RT-associated ribonuclease H displays both endonuclease and 3-r 5’ exonuclease activity. EMBO J. 9, 1171-1176. Schulze, T., Nawrath, M. & Moelling, K. (1991). Cleavage of the HIV-l p66 reverse transcriptase/RNase H by the p9 protease in vitro generates active p15 RNase H. Arch. Viral. in the press. Tisdale, M., Schulze, T., Larder, B. A. & Moelling, K. (1991). Mutations within the RNase H domain of HIV-l reverse transcriptase abolish virus infectivity. J. Gen. Virol. 72, 59-66. Wiihrl, B. M. & Moelling, K. (1990). Interactions of HIV-l RNase H polypurine tract containing with RNA-DNA hybrids. Biochemidry, 29, 10141-10147. Yang, W., Hendrickson, A., Crouch, R. J. t Satow, Y. (1990). Structure of ribonuclease H phased at 2 A resolution by MAD analysis of the selenomethionyl protein. Science, 249, 1398-1405. by J. Karn