Nucleic Acids Research, Vol. 20, No. 10 2591-2596
Transcriptional activity of the human immunodeficiency virus-1 LTR promoter in fission yeast Schizosaccharomyces pombe Reiko Toyama, Steven M.Bende+ and Ravi Dhar* Laboratory of Molecular Virology, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA Received September 6, 1991; Revised and Accepted April 6, 1992
ABSTRACT We have analyzed the transcriptional activity of the human immunodeficiency virus type I (HIV-1) LTR promoter in the fission yeast Schizosaccharomyces pombe (S.pombe). The ability of a series of 5'-deleted forms of the HIV-1 LTR promoter to direct transcription of the chloramphenicol acetyltransferase reporter gene was studied. We found that the HIV-1 promoter is functional in S.pombe and that deletion of sequences upstream of the NF-kB binding site previously identified to contain the negative regulatory element (NRE) in mammalian cells, resulted in an about thirty-fold increase in transcriptional activity. Sequences in the HIV-1 promoter that bind NF-kB were found to be essential for transcriptional activation in S.pombe. In mammalian cells, transactivation of the HIV-1 LTR requires TAR sequences and the viral Tat protein. In fission yeast, Tat failed to transactivate the HIV-1 LTR, suggesting that S.pombe may lack a cellular factor(s) required for the Tat transactivation process. INTRODUCTION The human immunodeficiency virus type 1 (HIV-1) encodes several regulatory genes, which are essential for virus production (1-4). The HIV-1 Tat protein is essential for transactivation of the HIV-1 LTR promoter (2, 5-8). The Tat protein requires a cis-acting RNA sequence called TAR which is located downstream of the transcription start site (+ 19 to +42) (9-11). The TAR region is capable of forming a stable RNA stem-loop secondary structure (6). In vitro, the Tat protein binds to the bulge region of the TAR RNA, and mutations within this region abolish transactivation (12, 13). Mutations within the loop region abolish Tat transactivation even though the Tat protein is still capable of binding to the TAR RNA (12). Since a 68 kD cellular protein has been shown to bind to the loop region of TAR RNA (14), the data would imply that both Tat and a cellular factor(s) are required for transactivation. Regulation of the HIV-1 promoter *
activity by Tat has been proposed to occur by several mechanisms. Tat has been found to increase the rate of initiation of HIV-1 transcription and also to stimulate elongation of short transcripts within the HIV-1 LTR leader sequences, possibly by antitermination mechanisms (7, 8, 10, 11, 14-16). Binding sites for the transcriptional factor Spi have been shown to be required for Tat-dependent transactivation (17) leading to the suggestion that the interaction of Tat with Spl may be one mechanism responsible for the increased rate of transcription initiation. Studies of HIV-1 transcription in mammalian cells have identified a number of cis-acting regulatory elements such as TATA box and NF-kB and Spl binding sites. The HIV-1 LTR contains a negative regulatory element (NRE) upstream of the NF-kB binding site (position -420 to -157 as defined in ref. 18; Figure 1), deletion of which results in an increased rate of Tat-dependent transcription (18). Specific sequences within the NRE which bind the transcription factors NFAT-1 and USF have been shown to contribute to the inhibitory effect of NRE on HIV-1 promoter activity (18). Recently, the transcriptional machinery of yeast and mammalian cells has been shown to be highly conserved (19, 20). For example, the TATA box binding factor TF-IID of yeast will complement for the mammalian TF-IID factor in in vitro transcription assays (21). Also, mammalian enhancer sequences and cellular transcription factors will activate transcription in yeast (19). A number of mammalian promoters, including the SV40 early, human cytomegalovirus (CMV), human chorionic gonadotropin c-subunit (CGca) and Adenovirus region 3 promoters, have been shown to be functionally active in the fission yeast Schizosaccharomyces pombe (S.pombe) (20). The transcription start sites of these promoters in S.pombe correspond precisely to the authentic 5' RNA ends observed in mammalian cells (20). In this study, we have characterized the transcriptional activity of the HIV-1 LTR promoter in S.pombe. We found that the HIV-1 LTR promoter has a weak activity in this system. Deletion of NRE sequences within the HIV-1 LTR resulted in a
To whom correspondence should be addressed
'Present address: Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892, USA
2592 Nucleic Acids Research, Vol. 20, No. 10 considerable increase in promoter activity which was dependent on the presence of the NF-kB biding sites located downsteam of the NRE. Unlike in mammalian cells, expression of the Tat protein in S.pombe did not lead to an increase in HIV-1 LTR promoter, activity.
METHODS Strains and media The S.pombe genotypes ura4-294 h- (ATCC38436) and ura4-294 leul-32 h- (gift from K.Okazaki) were used as hosts for transformation. A modified minimal medium (22) was used for growing S.pombe.
Plasmid construction JB12 (kindly provided by Julie Brown), a modified yeast replicating vector, was derived from S.pombe replicating vector pIRTl containing the URA3 gene of Saccharomyces cerevisiae (S. cerevisiae) as a selectable marker for transformation (kindly provided by A.Klar). A 146 bp DNA fragment containing a SacI restriction site at its 5' end followed by two poly A signals in opposite orientation (underlined), restriction sites for Ball, BglII, XhoI, KpnI, EcoNI, NurI, Notl, and EagI, a second stretch of two poly A signals in opposite orientation (underlined), and a BamHI site at its 3' end (5'-AGCTCGCGCAATAAAAG ATATm7ATTTCATTAGATATGTGTGTTGG TTTTTTGTGTGGCCAAGATCTCTCG AGGTACCTCGCGAAGGCGGCCGCAAAATAAAAATCAATGCAATTGTTGTTGTTA ACTTGTTTATTGTAACGGATC-3') was inserted at the SacI and BamHI sites of the pIRTI multiple cloning site to create JB12. A series of HIV-1 LTR promoter constructs, containing different 5' promoter deletions, the CAT reporter gene, and SV40 early splice and poly A signals (23) were subcloned into plasmid pBR322 using XbaI (5' end) and BamHI (3' end) linkers (kindly provided by J.Rappaport). The Xbal-BamHI fragments were cut out from the different HIV-1 LTR promoter deletion constru, blunt-ended, and cloned into the Ball site present in the polylinker region of JB12. The clones were designated pCHIV452 through pCHIV65, the numbers corresponding to the number of nucleotides upstream of the RNA start site (+ 1) present in these constructs (Figure 1). Clone pCHIV80 was generated by PCR using clone PCHIV452 as a template. The upstream primer contained a Stul restriction site at the 5' end followed by a sequence corresponding to nucleotides -80 to -66 of the HIV-1 LTR. The downstream primer was 5'-GCTTCCTGG GGATCCAGACATGATAAG-3', and contained a BamHI site at the 5' end and the complementary sequence of the 3' end of the SV40 poly A region which exists downstream of CAT coding region. The resulting 1.75 kb PCR product was inserted at the BalI/BglII sites into the polylinker region of JB12. The expression vector pCMVL (R.Toyama et al., unpublished data) was derived from pTR (20). The S.cerevisiae URA43 coding region in pTR was replaced with the entire coding region of S. cerevisiae LEU2 gene. The enhancer/promoter region of the major IE gene of CMV (-671 to +71) (24) was inserted at the BamHI/ClaI sites of the pTR multiple cloning site. The resultinig plasmid pCMVL contains a polylinker region with unique HindIl, BglI, Pvull, Sacdl and XhoI restriction sites downstream of the CMV promoter. The HIV-1 tat gene was tagged with a short sequence encoding
a 11 amino acid epitope (HA) of the Haemophilus influenzae
haemagglutimin protein (25) at its 3' end. A DNA fragment
containing the entire tatHA coding region was ed by PCR using the following primers: upstream (5'-1CG GCCOGCGGCCGCAAGCTTGACAGAGGAGAGCAAOAA, Hin site at the 5' end), downstream (5'-TGGCCATOGCCACTCGAGT-
CACAAGCTAGCGTAATCTGGAACAi-CGTATGGGTA TTCCTTCGGGCCTGTCG-3', HA sequen is, underlined, XhoI site at the 5' end). The resulting 325 bp PCR product was cloned'into pCMVL at the HindlI (5') and XhoI (3') sites of the multiple cloning site. pCEH was derived from pTR containing a 1.6 kb CGa promoter followed by the CAT coding secpe(20) (pTRCOa). The CGa promoter is extremely active in S.pombe (20). A 5'-deletion of the CGca promoter sequence extending to position -100 upstream of the transcription inittion site abolished promoter activity. However, the resulting core promoter could be activated by insertion of upstream enhancer sequences (Toyama et al., unpublished data). pCEH contains the CGa promoter and has XhoI and Bglll sites for the insertion of enhancer sequences upsntm of this core promoter which directs the transcription of a downstream CAT reporter gene. Transformation A high efficiency transformation protocol was used as previously described (22). HIV-1 LTR promoter deliconstructs were transformed into S.pombe ura4-294 h- (Figures 2 and 5) or ura4-294 leul-32 h- (Figure 3) either aloe or together with pCMVL-tatHA or pCMVL. S.pombe taini ng pCMVL-tatHA showed no detectable changes in growth rate. Extract preparation and CAT assay Transformants (approximately 1.5 x 106 cells) were plated on mininal medium plates and cultured for 5 days at 300C. Transformants were harvested, washed twice wi4h water, and suspended in 1 ml of a solution containing I M sorbitol, 0.1 M EDTA, and 14 mM 2-mercaptoethanol. The cells were then treated with lyticase (Sigma, 0.1 units per Dtl) at 300C for 1 hr.
poly HlVLTR A
TAR
CAT
upstream regulatory element/NRE
Splicing poly A
NF TB s Sp II TATA
ghlO5h82
-79
-46
-27
.
pCHI V65 nrw 0vAn
-rCHI V 176
Flgure 1. Dagrmnmtic representation of the HIV-1 LTR promter showing upstream promoter deletion mutants. The positions of the rgultory domains including the negative regulatory element (NRE), NFPk, SptI TATA and TAR sequences are indicated. The '-end of the RNA co tsondt position + I. The numbering of the pCHIV deletion constructs rfers to the 5'-end points of the LTR sequences. pCHIV452 contains the entire HIV-I LTR, with 452 nucleotides upstream of the 5' end of the RNA. All contnc were cloned into the Ball restriction site within the polylinker region of JB12, an pIRTI modified yeast replicating vector containing the URM3 gene of Saccsaromyces cerevisiae as a selectable marker. The polylinker is flanked by mammalian poly A sites at both ends and in both orientations.
Nucleic Acids Research, Vol. 20, No. 10 2593 The spheroplasts were harvested by centrifugation at 3000 rpm (Beckman table top centrifuge) for 10 min at 4°C, and resuspended in 1.2 ml of 92 mM Tris-HCl (pH 7.4) containing 42 mM EDTA. The cells were sonicated on ice by a Branson sonifier 450 (output control 2.5) for 1 min and cell debris was removed by centrifugation. Protein concentrations of the extracts were measured according to Bradford (26) using the BioRad protein assay kit. CAT assays were performed according to published protocols using 1 ytg of protein extract (20).
35S labelling of proteins and immunoprecipitation Exponentially growing cultures (100 ml in minimal medium) of S.pombe transformed with pCMVL or pCMVL-tatHA were harvested by centrifugation at 3000 rpm (Beckman table top centrifuge) for 20 min at 4°C. The cells were resuspended in 10 ml of minimal medium and labelled with 35S-Cysteine (100 AtCi/ml) for 1 hr. Cells were collected by centrifugation then washed 2 times with 1 ml of the cell breakage buffer (0.1 M Tris-HCl [pH 7.5], 1 mM DTT, 20% glycerol and a mixture of the protease inhibitors (100 Agg/ml each) N-Tosyl-L-lysine chloromethyl ketone (TLCK), N-Tosyl-L-phenylalanine chloromethyl ketone (TPCK) and Phenylmethanosulfonyl fluoride (PMSF)). Cells were resuspended in 0.3 ml of cell breakage buffer, and acid washed glass beads (Sigma, 425-600 Am) were added to just below the meniscus. Cells were vortexed vigorously in 1.5 ml microfuge tubes 6 times for 1 min, keeping the cells on ice during the intervals. Cell extracts were separated from glass beads, and about 0.3 ml of extract was obtained. The extract was diluted with modified RIPA buffer (20 mM MOPS, 150 mM NaCl, 1 mM EDTA, 1 % NP-40, 1 % deoxycholate, 0.1 % SDS;
pH 7.0) containing 1 % apoprotein. Cell debris was removed by centrifugation. 1 ml of extract was incubated with 2 IA of monoclonal antibody 12CA5 (obtained from Berkeley Antibody Inc.) directed against the 11 amino acid HA epitope (see above) for 4 hr at 4°C. 50 1l of protein A Sepharose CL4B beads (Pharmacia) were added and extracts were allowed to incubate for an additional 90 min. Following five times washes in modified RIPA buffer, Sepharose beads were resuspended in 30 l1 of sample buffer for electrophoresis, heated at 100°C for 5 min, and separated on 1.5% SDS-polyacrylamide gels. Gels were then fixed with glacial acetic acid/methanol, treated with Enlightning (NEN), dried and autoradiographed.
Northern blot analysis and primer extension Total RNA was extracted from S.pombe transformants as previously described (20). 10 tsg of RNA was denatured and fractionated on 1 % agarose gels containing 6.7% formaldehyde. RNA was transferred to nitrocellulose filters and hybridized to a probe (32P-labelled by nick translation) containing either the whole CAT coding sequence, the entire tatHA coding sequence or a fragment of the S.pombe actin gene. The DNA fragments
A
B
C
2.3 71.35-
-TAT
woo
46I-
0.24-
30_n
_,
A
7.462 1.5-
4,
B
-TAT 14.3-
2.37-
kb
6.5-
7.461.35-
"
S
- CAT
3.4-
4.402.37-
i.
1.35-
CAT
1
2
3
2 37-
4
5
1
6
2
Actin
1.35-
1
2 3 4 5 6
2.37-
*..* 1
2
0*
3
4
5
1.35-.
6
Actin
1
2
3 4
5 6
Figure 2. Transcriptional activity of HIV-1 LTR promoter deletion mutants. S.pombe cells were transformed with the different HIV-1 LTR promoter deletion constructs shown in Figure 1. The results of CAT assays (A) and Northern blot analysis (B) are shown. Protein (A) or total RNA (B) was extracted from transformants carrying the following plasmids: pCHIV452 (lane 1, full length LTR); pCHIV278 (lane 2); pCHIV 176 (lane 3); pCHIV1 17 (lane 4); pCHIV80 (lane 5); pCHIV65 (lane 6). 1 isg of protein extract was used for the CAT assays. Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values 4 S.E.M; n = 3) were the following: pCHIV452 (0.78 + 0.03), pCHIV278 (1.3 + 0.2), pCHIV176 (1.2 0.1), pCHIV117 (29 + 7), pCHIV80 (1.4 + 0.1), pCHIV65 (0.12 + 0.02). In panel B (upper section), a 0.9kb BamHI restriction fragment containing the entire CAT coding sequence (cut out from pcDCAT, H.Okayama, unpublished data) was used as a probe. The lower section of panel B shows the same blot reprobed with a fragment of the Spombe actin gene (see Methods) as a control. The molecular size markers (kb) are shown on the left of the blot.
Figure 3. Transcriptional activity of the HIV-1 LTR promoter in the presence and absence of Tat. S.pombe cells were transformed with different HIV-1 LTR promoter deletion constructs together with pCMVL-tatHA or PCMVL (control). S.pombe extracts were assayed for CAT activity (A) and the presence of CAT and Tat mRNA (B). pCHIV452 (lane 1), pCHIV1 17 (lane 2), or pCHIV65 (lane 3) were cotransformed with pCMVL-tatHA. In parallel, pCHIV452 (lane 4), pCHIV1 17 (lane 5), or pCHIV65 (lane 6) were cotransformed with PCMVL which does not contain the tat gene. 1 og of protein extract was used for the CAT assays (A). Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values t S.E.M; n = 3) were: pCHIV452 + pCMVLtatHA (0.59 4 0.2), pCHIV 117 + pCMVL-tatHA (10.2 + 1.2), pCHIV65 + pCMVL-tatHA (0.15 4 0.03), pCHIV452 + pCMVL (0.36 i 0.10), pCHIV117 + pCMVL (17.4 3.5), pCHIV65 + pCMVL (0.21 X 0.01). (B) The probes used for the Northern analysis are described under Methods. A fragment of the S.pombe actin gene was used as a control probe (panel B, lower section). The molecular size markers (kb) are shown on the left of the blot. (C) Immunoprecipitation of TatHA fusion protein. Spombe cells transformed with pCMVL-tatHA (lane 1) or pCMVL (lane 2, control) were labelled with 35Scysteine as described under Methods. TatHA fusion protein was immunoprecipitated with monoclonal antibody 12CA5 directed against the HA epitope (see Methods). The molecular size markers (kD) are shown on the left of the panel.
2594 Nucleic Acids Research, Vol. 20, No. 10 used as probes were prepared as follows: The entire CAT coding region (0.9 kb) was excised from pcDCAT (gift from H.Okayama) with BamHI. For the tat probe, the whole tatHA coding region was generated by PCR using pCMVL-tatHA as a template and the upstream primer 5'-CTCGAGTTACTCGACAGAGGAGAGCAAGAA-3' and the downstream primer 5 '-ATCGCGATCACAAGCTAGCGTAATCTGGAACAT-
CGTATGGGTATTCCTTCGGGCCTGTCG-3' which covered with whole HA epitope. A 400 bp S.pombe actin gene fiagment corresponding to nucleotides 1083 to 1484 (as defined in ref. 27) was generated by PCR using 5'-CCAACCCTCAGCITI-3' as an upstream primer, 5'-TCAGAAGAATCGATGT-3' as a downstream primer, and a S.pombe genomic DNA library (28) as a template. The actin gene probe was used as a control probe to determine the amount and quality of yeast RNA loaded onto the gels. This fragment hybridizes with the three transripts of the actin gene which are 1.8, 1.6 and 1.2 kb long. For primer extension studies, an antisense oligonucleotide (5 '-CAACGGTGGTATATCCAGT-3') corresponding to nucleotides + 12 to +31 of the CAT coding sequence was labeled at its 5' end using 32P-ATP and T4 polynucleotide kinase and annealed to total yeast RNA. Primer extension was perforned as described (20). The reaction mixture was fractionated on a 6% polyacrylamide gel containing 8 M urea. A dideoxy sequencing reaction was run in parallel as a size marker.
RESULTS AND DISCUSSION Transcription The transcriptional activity of various HIV-1 LTR promoter deletion constructs (Figure 1) was measured by assaying the chloramphenicol acetyltransferase (CAT) activity in protein extracts prepared from S.pombe transformants. In addition, gene expression was monitored by Northern blot analysis using the CAT gene as a radiolabelled probe. The plasmid containing the whole HIV-l LTR (pCHIV452) and constructs with 5' pr r deletions extending downstream to -176 (pCHIV278 and pCHIV176) showed about 1% conversion of radiolabelled chloramphenicol (Figure 2A, lanes 1-3). pCHIV1 17, in which sequences upstream of the two NF-kB binding sites of the HIV-l LTR promoter have been deleted, showed an about 30-fold increase in CAT activity compared to pCHIV452 (Figure 2A, lane 4). pCHIV80 was at least 15 times less active than pCIHIV1 17 (Figure 2A, lanes 4 and 5). This plasmid (pCHIV80) does no longer contain the two NF-kB binding sites, but still has three copies of the Spl binding site (Figure 1). pCHIV65, which contains the core promoter and one and a half binding sites for the transcription factor Spl, did not give any detectable CAT activity (Figure 2A, lane 6). The results obtained by Northem blot analysis (Figure 2B) were in good agreement with the data obtained in the CAT assays (Figure 2A). Transformants containing pCHIV1 17 (lane 4) showed significantly higher levels of CAT mRNA (- 1.35 kb) than pCHIV452, pCHIV278, or pCHIV176 (Figure 2B, lanes 1-4). The amount of CAT mRNA seen with PCHIV8O (lane 5) was low, similar to that obtained with pCHIV452, pCHIV278, and pCHIV176. No CAT mRNA transcript was observed with pCHIV65. In this case only a high molecular weight band appeared on the Northern blot which is usually seen when the promoter activity of the CAT reporter construct is extremely low (R.Toyama, unpublished data). As a control, the same blot was
hybridized to a S.pombe actin gene probe, verifying that equal amounts of RNA were loaded in each lan of te gel. These data suggest that sequences between nucleoides -176 and -117 of the HIV-1 LTR promoter repress gene expression, and dut the region containing the NF'kR aepunces (-117 through -81) are required for strong HIV-l LTR promoter activity in Spombe. It is unlikely that the construct pCHIV1 17 results in an artifactual positive cis-elenrn at the junction of vector and -117 site, since an oligonucleotide -112 to -79 is able to efficiently activate a heterologous promoter (see below for details in the activadon of hete s ptonr). I Hmammlian cells, a number of tranriptin as incuding APl (-349 to -330), NFAT-1 (-254 to -216) and USF (-173 to -159) recognize sequences within the NRE (18). Deletion of NFAT-1 and USF sequences witiin the HIV-l LTR results in an increase in gene expression in ma lian cells (18). In this study, deletion of the segment betwee -176 to - 117 of the HIV-1 LTR allowed NF-kB sequences tlactivMe tanscription from the HIV-1 LTR promoter in S.pobe. tetngly, this sequence element (-176 to -117) con sa binding site for USF (-173 to -159) which has been shn toconributeto the repression of HIV-l LTR promoter activity in mamnalian cells (18). A more detailed analysis is required to elucidate whether S. pombe contains an USF-like factor and whedter the recognition sequence is identical to that found in mammalianl cells. The HIV-1 Tat protein is known to bind to the TAR RNA element (Figure 1) and to transactivate tnscription from the HIV-1 LTR promoter (12, 13). Control of transactivation can occur at different levels including increased rate of initation of transcription or elongation of the short RNAs which truncate within the HIV-1 leader sequence (7, 8, 10, 11, 14-16). We have tested the possibility that Tat may transactivate the HIV-1 LTR pro tr in the fission yeast. Initially, the plasmnid pCMVLtatHA encoding the Tat protein tagged with a short sequence encoding a 11 amino acid epitope (HA) of the Haemnphilus influnzae haemagglutinin protein (25) to allow its unolocal
t 50 142
12 FI1gua 4. 5' end analysis using primer extension. An adnwies oligomuciotide + 12 to + 31 of the CAT coding asoquan was labelled at its 5' end with using polynucleotide kinase, amealed to S.pombe RNA, and extended by reverse transcriptase. Total RNA, prepomd from cells cotransformed with pCHIV 1 17 and pCMVL -(lane, 1, contro; 16 ,sg RNA) or with pCHIV117 and pCMVLatHA (lane 2; 22 og RNA) was usd as a .tenbIt. In mamalian cells, transcription initiation of the HIV-I LTR occn at position
correspwnding to nulode 32P
147. A dideoxy sequencing reaction is shown on the left as a size standard.
Nucleic Acids Research, Vol. 20, No. 10 2595 detection, was tested for its ability to induce transactivation of the HIV-1 LTR promoter in Hela cells. Transactivation of the HIV-1 LTR by the TatHA fusion protein was compared to that by the Tat protein without the HA epitope. We found that the TatHA fusion protein was able to transactivate the HIV-1 promoter as efficiently as the Tat protein alone (data not shown). Subsequently, the ability of pCMVL-tatHA to induce synthesis of TatHA protein in S.pombe was tested by labelling exponentially growing S.pombe transformants with 35S-cysteine. In parallel, cells transformed with the vector (pCMVL) alone were processed as a control. Figure 3C shows that antibody 12CA5 was able to immunoprecipitate TatHA protein in cells transformed with pCMVL-tatHA but not in those transformed with pCMVL alone. Our effort to map the subcellular localization of the TatHA fusion protein in S.pombe was not successful because of the high background observed with the 12CA5 monoclonal antibody in the absence of Tat. The yeast strain [genotype: ura4-294 leul-32 h-] was cotransformed with pCMVL-tatHA and various HIV-1 LTR promoter deletion constructs described above. The plasmid pCMVL which does not carry the tatHA gene was used as a control. Transformation of S.pombe with pCHIV452, pCH1V1 17 or pCHIV65 together with pCMVL-tatHA resulted in CAT activities (Figure 3A, lanes 1-3) and CAT mRNA levels (Figure 3B, lanes 1-3) similar to those observed after cotransformation with pCMVL (Figure 3A, B, lanes 4-6). Tat mRNA was found in cells cotransformed with pCMVL-tatHa but not in those cotransformed with pCMVL (Figure 3B). Our data therefore suggest that Tat does not transactivate the HIV-1 LTR promoter in the fission yeast. Primer extension The 5' end of the TAR RNA sequence is important for Tat to transactivate the HIV-1 LTR promoter in mammalian cells (10). In order to examine the possibility that the demonstrated lack of Tat transactivation of the HIV-I LTR in S.pombe may be due to a different site of transcription initiation in S.pombe (resulting
A
in a loss of TAR sequences), primer extension analysis was performed. An antisense oligonucleotide (corresponding to nucleotides + 12 to +31 of the CAT coding sequence) was 5' end labelled, annealed to total RNA prepared from yeast cells cotransformed with pCHIV1 17 and pCMVL (Figure 4, lane 1) or cotransformed with pCHIV1 17 and pCMVL-tatHA (Figure 4, lane 2), and extended with reverse transcriptase. The 5' end analysis showed two major clusters of 5' ends, each cluster containing 3 to 4 bands (Figure 4). The major 5' ends were located 142 and 150 nucleotides away from the 5' end of the primer. The transcription initiation start site(s) of the HIV-1 LTR in S.pombe is therefore very similar to that found in mammalian cells (147 bases away from the 5' end of the primer). In fact, there are minor transcripts in S.pombe whose syndtesis is initiated exactly at the same site as in mammalian cells (Figure 4). Since the transcripts found in S.pombe would allow transactivation of the HIV-1 LTR by Tat in mammalian cells, the observed lack of transactivation by Tat in S.pombe appears not to be due to different points of transcription initiation. Recent reports suggest that transactivation of the HIV-1 promoter by Tat is partly dependent on the presence of Spi binding sites within the core promoter (17). One possible explanation for the inability of Tat to transactivate transcription from the HIV-1 LTR promoter in S.pombe may therefore be the lack of the Spl transcription factor in the fission yeast. Alternatively, S.pombe may lack a cellular factor(s) (other than Spi) necessary for the Tat transactivation process which has been proposed to exist in mammalian cells (17). Activation of a heterologous promoter by NF-kB sequences As outlined above, a short segment in the HIV-I LTR (-117 to -81) containing two NF-kB recognition sequences is responsible for the increase in promoter activity seen after removal of upstream promoter sequences. To further characterize this sequence element we inserted a 34 bp oligonucleotide
(5'-GCTACAAGGGACTTTCCGCTGGGGACTTTCCAGG-3') (-112 to -79; NF-kB sequence shown underlined) upstream of
oligo cloning
B
M,
ura 3
m NF kll3 GC TACAAGGAw
T TCTGG G C C
INF-KBm :GCTACAACTCACTTTCCGCTGC TCAC TTCCAGG
1
2
3
Figure 5. Transcriptional activity of pCEH core vector constructs containing NF-kB binding sequences. (A) An oligonucleotide corresponding to the NF-kB protein recognition sequence -83 to -105 (Figure 1) or the mutant sequence, NF-kBm, was inserted into the polylinker site upstream of the pCEH core promoter vector. pCEH contains the CGa core promoter directing transcription of the CAT reporter gene. (B) CAT analysis of protein extracts prepared from S.pombe cells transformed with pCEH (lane 1, control), pCEH-NF-kB (lane 2) and pCEH-NF-kBm (lane 3). Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values L S.E.M; n = 3) were: pCEH (1.67 fi- 0.24), pCEH + NFkB (17.6 0.4), pCEH + NFkBm (6.17 -0.55).
2596 Nucleic Acids Research, Vol. 20, No. 10 the CGa core promoter into the yeast replicating vector pCEH. This URA3 based vector contains the CGa core promoter upstream of the CAT gene flanked by a 5' polylinker region which allows the insertion of enhancer sequences (Figure 5). For comparison, an additional oligonucleotide (5'-GCTACAACTCACTTTCCGCTGCTCACTTTCCAGG-3') containing mutations (shown in bold lefters) in the NF-kB protein recognition sequence which abolishes NF-kB binding in mammalian cells (29) was also cloned into pCEH (Figure 5). Yeast cells were transformed with both plasmids and CAT activities were determined in cell extracts as described. Transformants containing wild type NF-kB sequences displayed >9-fold activation of the CG a core promoter, whereas the mutant sequence (NF-kBm) resulted in only 2-fold activation. Together with the results presented in Figures 2 and 3, hes data suggest the existence of a cis-acting element containing two NFkB binding site (between -117 to -81) in the HIV-1 LTR which is required for promoter activation in the fission yeast. However, our data do not show that the tranription factor which recognizes this sequence element in S.pombe has a recognition sequence identical to that of NF-kB. It is unclear whether S.pombe will serve as a useful system for the study of the different functional domains of the HIV-1 LTR promoter, because the cis-acting sequences identified here may interact with yeast factors not relevant to the process in mammalian cells. On the other hand, the inability of Tat to activate the HIV-l LTR promoter in S.pombe may allow the isolation of genes which are involved in mediating this process in mammalian cells. To test this, we are presently developing a genetic selection system based on cotransformation of S.pombe with mammalian cDNA libraries.
ACKNOWLEDGMENTS We are grateful to Drs Susan Marriott and Jurgen Wess for critical review of the manuscript and to Dr Julie Brown for providing yeast replicating vectors before publication. We are also thankful to Drs Jay Rappaport and Cindy Bohan for providing us with plasmids carrying different HIV-1 LTR CAT constructs.
REFERENCES 1. 2. 3. 4. 5.
6. 7. 8. 9. 10. 11.
12.
13. 14. 15.
16.
Cullen,B.R. and Greene,W.C. (1989) Cell 58, 423 -426. Cullen,B.R. (1991) J. Virol. 65, 1053-1056. Pavlakis,G.N. and Felber,B.K. (1990) New Biologist 2, 20-31. Rosen,C.A. (1991) Trends Genet. 7, 9-14. Wright,C.M., Felber,B.K., Paskalis,H. and Pavlakis,G.N. (1986) Science 234, 988-992. Muesing,M.A., Smith,D.H. and Capon,D.J. (1987) Cell 48, 691-701. Laspia,M.F., Rice,A.P. and Mathews,M.B. (1989) Cell 59, 283-292. Feinberg,M.B., Baltimore,D. and Franker,A.D. (1991) Proc. Natl. Acad. Sci. USA- 88, 4045-4049. Hauber,J. and Cullen,B.R. (1988) J. Virol. 62, 673-679. Jakobovits,A., Smith,D.H., Jakobovits,E.B. and Capon,D.J. (1988) Mol. Cell. Biol. 8, 2555-2561. Selby,M.J., Bain,E.S., Luciw,P.A. and Peterlin,B.M. (1989) Genes Dev. 3, 547-558. Roy,S., Delling,U., Chen,C.-H., Rosen,C.A. and Sonenberg,N. (1990) Genes Dev. 4, 1365-1373. Weeks,K.M., Ampe,C., Schultz,S.C., Steitz,T.A. and Crohers,D.M. (1990) Science 249, 1281-1285. Marciniak,R.A., Garcia-Blanco,M.A. and Sharp,P.A. (1990) Proc. Natl. Acad. Sci. USA 87, 3624-3628. Kao,S.-Y., Calman,A.F., Luciw,P.A. and Peterlin,B.M. (1987) Nature 330, 489-493. Laspia,M.F., Rice,A.P. and Mathews,M.B. (1990) Genes Dev. 4, 2397-2408.
17. Kamine,J., Subramanian,T. and Chinnadurai,G. (1991) Proc. Natl. Acad. Sci. US 88, 8510-8514. 18. Lu,Y., Touzjian,N., Stenzel,M., Dorfman,T., Sodroski,J.G. and Haseltine,W.A. (1990) J. Virol. 64, 5226-5229. 19. Jones,R.H., Morene,S., Nurse,P. and Jones,N.C. (1988) Cell 53, 659-667. 20. Toyama, R., and Okayama, H. (1990) FEBS Let. 268, 217-221. 21. Hoffmann,A., Horikoshi,M., Wang,K.C., Schlr,S., Weil,P.A. and Roeder,R.G. (1990) Genes Dev. 4, 1141-1148. 22. Okazaki,K., Okazaki,N., Kume,K., Jinno,S., Tanaka,K. and Okayama,H. (1990) Nucleic Acids Res. 18, 6485-6489. 23. Bohan,C.A., Robinson,R. A., Luciw,P.A. and Srinivasan,A. (1989) V7rology 172, 573-583. 24. Boshart,M., Weber,F., Jahn,G., Dorsch-Haser,K., Flcckenstein,B. and Schaffner,W. (1985) Cell 41, 521-530. 25. Field,l., Nikawa,J., Brock,D., MacDonald,B., Rodgers,L., Wilson,I.A., Lrnier,R.A. and Wigler,M. (1988) Mol. Cell. Biol. 8, 2159-2165. 26. Bradford,M. (1976) Anal. Biochem. 72, 248-254. 27. Mertins,P. and Gailwitz,D. (1987) Nucleic Acids Res. 15, 7369-7379. 28. Wright,A., Maundrell,K., Heyer,W.-D., Beach,D. and Nurse,P. (1986) Plasmid 15, 156-158. 29. Duh,E.J., Maury,W.J., Folks,T.M., Fauci,A.S. and Rabson,A.B. (1989) Proc. Natl. Acad. Sci. USA 86, 5974-5978.
Transcriptional activity of the human immunodeficiency virus-1 LTR promoter in fission yeast Schizosaccharomyces pombe Reiko Toyama, Steven M.Bende+ and Ravi Dhar* Laboratory of Molecular Virology, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA Received September 6, 1991; Revised and Accepted April 6, 1992
ABSTRACT We have analyzed the transcriptional activity of the human immunodeficiency virus type I (HIV-1) LTR promoter in the fission yeast Schizosaccharomyces pombe (S.pombe). The ability of a series of 5'-deleted forms of the HIV-1 LTR promoter to direct transcription of the chloramphenicol acetyltransferase reporter gene was studied. We found that the HIV-1 promoter is functional in S.pombe and that deletion of sequences upstream of the NF-kB binding site previously identified to contain the negative regulatory element (NRE) in mammalian cells, resulted in an about thirty-fold increase in transcriptional activity. Sequences in the HIV-1 promoter that bind NF-kB were found to be essential for transcriptional activation in S.pombe. In mammalian cells, transactivation of the HIV-1 LTR requires TAR sequences and the viral Tat protein. In fission yeast, Tat failed to transactivate the HIV-1 LTR, suggesting that S.pombe may lack a cellular factor(s) required for the Tat transactivation process. INTRODUCTION The human immunodeficiency virus type 1 (HIV-1) encodes several regulatory genes, which are essential for virus production (1-4). The HIV-1 Tat protein is essential for transactivation of the HIV-1 LTR promoter (2, 5-8). The Tat protein requires a cis-acting RNA sequence called TAR which is located downstream of the transcription start site (+ 19 to +42) (9-11). The TAR region is capable of forming a stable RNA stem-loop secondary structure (6). In vitro, the Tat protein binds to the bulge region of the TAR RNA, and mutations within this region abolish transactivation (12, 13). Mutations within the loop region abolish Tat transactivation even though the Tat protein is still capable of binding to the TAR RNA (12). Since a 68 kD cellular protein has been shown to bind to the loop region of TAR RNA (14), the data would imply that both Tat and a cellular factor(s) are required for transactivation. Regulation of the HIV-1 promoter *
activity by Tat has been proposed to occur by several mechanisms. Tat has been found to increase the rate of initiation of HIV-1 transcription and also to stimulate elongation of short transcripts within the HIV-1 LTR leader sequences, possibly by antitermination mechanisms (7, 8, 10, 11, 14-16). Binding sites for the transcriptional factor Spi have been shown to be required for Tat-dependent transactivation (17) leading to the suggestion that the interaction of Tat with Spl may be one mechanism responsible for the increased rate of transcription initiation. Studies of HIV-1 transcription in mammalian cells have identified a number of cis-acting regulatory elements such as TATA box and NF-kB and Spl binding sites. The HIV-1 LTR contains a negative regulatory element (NRE) upstream of the NF-kB binding site (position -420 to -157 as defined in ref. 18; Figure 1), deletion of which results in an increased rate of Tat-dependent transcription (18). Specific sequences within the NRE which bind the transcription factors NFAT-1 and USF have been shown to contribute to the inhibitory effect of NRE on HIV-1 promoter activity (18). Recently, the transcriptional machinery of yeast and mammalian cells has been shown to be highly conserved (19, 20). For example, the TATA box binding factor TF-IID of yeast will complement for the mammalian TF-IID factor in in vitro transcription assays (21). Also, mammalian enhancer sequences and cellular transcription factors will activate transcription in yeast (19). A number of mammalian promoters, including the SV40 early, human cytomegalovirus (CMV), human chorionic gonadotropin c-subunit (CGca) and Adenovirus region 3 promoters, have been shown to be functionally active in the fission yeast Schizosaccharomyces pombe (S.pombe) (20). The transcription start sites of these promoters in S.pombe correspond precisely to the authentic 5' RNA ends observed in mammalian cells (20). In this study, we have characterized the transcriptional activity of the HIV-1 LTR promoter in S.pombe. We found that the HIV-1 LTR promoter has a weak activity in this system. Deletion of NRE sequences within the HIV-1 LTR resulted in a
To whom correspondence should be addressed
'Present address: Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892, USA
2592 Nucleic Acids Research, Vol. 20, No. 10 considerable increase in promoter activity which was dependent on the presence of the NF-kB biding sites located downsteam of the NRE. Unlike in mammalian cells, expression of the Tat protein in S.pombe did not lead to an increase in HIV-1 LTR promoter, activity.
METHODS Strains and media The S.pombe genotypes ura4-294 h- (ATCC38436) and ura4-294 leul-32 h- (gift from K.Okazaki) were used as hosts for transformation. A modified minimal medium (22) was used for growing S.pombe.
Plasmid construction JB12 (kindly provided by Julie Brown), a modified yeast replicating vector, was derived from S.pombe replicating vector pIRTl containing the URA3 gene of Saccharomyces cerevisiae (S. cerevisiae) as a selectable marker for transformation (kindly provided by A.Klar). A 146 bp DNA fragment containing a SacI restriction site at its 5' end followed by two poly A signals in opposite orientation (underlined), restriction sites for Ball, BglII, XhoI, KpnI, EcoNI, NurI, Notl, and EagI, a second stretch of two poly A signals in opposite orientation (underlined), and a BamHI site at its 3' end (5'-AGCTCGCGCAATAAAAG ATATm7ATTTCATTAGATATGTGTGTTGG TTTTTTGTGTGGCCAAGATCTCTCG AGGTACCTCGCGAAGGCGGCCGCAAAATAAAAATCAATGCAATTGTTGTTGTTA ACTTGTTTATTGTAACGGATC-3') was inserted at the SacI and BamHI sites of the pIRTI multiple cloning site to create JB12. A series of HIV-1 LTR promoter constructs, containing different 5' promoter deletions, the CAT reporter gene, and SV40 early splice and poly A signals (23) were subcloned into plasmid pBR322 using XbaI (5' end) and BamHI (3' end) linkers (kindly provided by J.Rappaport). The Xbal-BamHI fragments were cut out from the different HIV-1 LTR promoter deletion constru, blunt-ended, and cloned into the Ball site present in the polylinker region of JB12. The clones were designated pCHIV452 through pCHIV65, the numbers corresponding to the number of nucleotides upstream of the RNA start site (+ 1) present in these constructs (Figure 1). Clone pCHIV80 was generated by PCR using clone PCHIV452 as a template. The upstream primer contained a Stul restriction site at the 5' end followed by a sequence corresponding to nucleotides -80 to -66 of the HIV-1 LTR. The downstream primer was 5'-GCTTCCTGG GGATCCAGACATGATAAG-3', and contained a BamHI site at the 5' end and the complementary sequence of the 3' end of the SV40 poly A region which exists downstream of CAT coding region. The resulting 1.75 kb PCR product was inserted at the BalI/BglII sites into the polylinker region of JB12. The expression vector pCMVL (R.Toyama et al., unpublished data) was derived from pTR (20). The S.cerevisiae URA43 coding region in pTR was replaced with the entire coding region of S. cerevisiae LEU2 gene. The enhancer/promoter region of the major IE gene of CMV (-671 to +71) (24) was inserted at the BamHI/ClaI sites of the pTR multiple cloning site. The resultinig plasmid pCMVL contains a polylinker region with unique HindIl, BglI, Pvull, Sacdl and XhoI restriction sites downstream of the CMV promoter. The HIV-1 tat gene was tagged with a short sequence encoding
a 11 amino acid epitope (HA) of the Haemophilus influenzae
haemagglutimin protein (25) at its 3' end. A DNA fragment
containing the entire tatHA coding region was ed by PCR using the following primers: upstream (5'-1CG GCCOGCGGCCGCAAGCTTGACAGAGGAGAGCAAOAA, Hin site at the 5' end), downstream (5'-TGGCCATOGCCACTCGAGT-
CACAAGCTAGCGTAATCTGGAACAi-CGTATGGGTA TTCCTTCGGGCCTGTCG-3', HA sequen is, underlined, XhoI site at the 5' end). The resulting 325 bp PCR product was cloned'into pCMVL at the HindlI (5') and XhoI (3') sites of the multiple cloning site. pCEH was derived from pTR containing a 1.6 kb CGa promoter followed by the CAT coding secpe(20) (pTRCOa). The CGa promoter is extremely active in S.pombe (20). A 5'-deletion of the CGca promoter sequence extending to position -100 upstream of the transcription inittion site abolished promoter activity. However, the resulting core promoter could be activated by insertion of upstream enhancer sequences (Toyama et al., unpublished data). pCEH contains the CGa promoter and has XhoI and Bglll sites for the insertion of enhancer sequences upsntm of this core promoter which directs the transcription of a downstream CAT reporter gene. Transformation A high efficiency transformation protocol was used as previously described (22). HIV-1 LTR promoter deliconstructs were transformed into S.pombe ura4-294 h- (Figures 2 and 5) or ura4-294 leul-32 h- (Figure 3) either aloe or together with pCMVL-tatHA or pCMVL. S.pombe taini ng pCMVL-tatHA showed no detectable changes in growth rate. Extract preparation and CAT assay Transformants (approximately 1.5 x 106 cells) were plated on mininal medium plates and cultured for 5 days at 300C. Transformants were harvested, washed twice wi4h water, and suspended in 1 ml of a solution containing I M sorbitol, 0.1 M EDTA, and 14 mM 2-mercaptoethanol. The cells were then treated with lyticase (Sigma, 0.1 units per Dtl) at 300C for 1 hr.
poly HlVLTR A
TAR
CAT
upstream regulatory element/NRE
Splicing poly A
NF TB s Sp II TATA
ghlO5h82
-79
-46
-27
.
pCHI V65 nrw 0vAn
-rCHI V 176
Flgure 1. Dagrmnmtic representation of the HIV-1 LTR promter showing upstream promoter deletion mutants. The positions of the rgultory domains including the negative regulatory element (NRE), NFPk, SptI TATA and TAR sequences are indicated. The '-end of the RNA co tsondt position + I. The numbering of the pCHIV deletion constructs rfers to the 5'-end points of the LTR sequences. pCHIV452 contains the entire HIV-I LTR, with 452 nucleotides upstream of the 5' end of the RNA. All contnc were cloned into the Ball restriction site within the polylinker region of JB12, an pIRTI modified yeast replicating vector containing the URM3 gene of Saccsaromyces cerevisiae as a selectable marker. The polylinker is flanked by mammalian poly A sites at both ends and in both orientations.
Nucleic Acids Research, Vol. 20, No. 10 2593 The spheroplasts were harvested by centrifugation at 3000 rpm (Beckman table top centrifuge) for 10 min at 4°C, and resuspended in 1.2 ml of 92 mM Tris-HCl (pH 7.4) containing 42 mM EDTA. The cells were sonicated on ice by a Branson sonifier 450 (output control 2.5) for 1 min and cell debris was removed by centrifugation. Protein concentrations of the extracts were measured according to Bradford (26) using the BioRad protein assay kit. CAT assays were performed according to published protocols using 1 ytg of protein extract (20).
35S labelling of proteins and immunoprecipitation Exponentially growing cultures (100 ml in minimal medium) of S.pombe transformed with pCMVL or pCMVL-tatHA were harvested by centrifugation at 3000 rpm (Beckman table top centrifuge) for 20 min at 4°C. The cells were resuspended in 10 ml of minimal medium and labelled with 35S-Cysteine (100 AtCi/ml) for 1 hr. Cells were collected by centrifugation then washed 2 times with 1 ml of the cell breakage buffer (0.1 M Tris-HCl [pH 7.5], 1 mM DTT, 20% glycerol and a mixture of the protease inhibitors (100 Agg/ml each) N-Tosyl-L-lysine chloromethyl ketone (TLCK), N-Tosyl-L-phenylalanine chloromethyl ketone (TPCK) and Phenylmethanosulfonyl fluoride (PMSF)). Cells were resuspended in 0.3 ml of cell breakage buffer, and acid washed glass beads (Sigma, 425-600 Am) were added to just below the meniscus. Cells were vortexed vigorously in 1.5 ml microfuge tubes 6 times for 1 min, keeping the cells on ice during the intervals. Cell extracts were separated from glass beads, and about 0.3 ml of extract was obtained. The extract was diluted with modified RIPA buffer (20 mM MOPS, 150 mM NaCl, 1 mM EDTA, 1 % NP-40, 1 % deoxycholate, 0.1 % SDS;
pH 7.0) containing 1 % apoprotein. Cell debris was removed by centrifugation. 1 ml of extract was incubated with 2 IA of monoclonal antibody 12CA5 (obtained from Berkeley Antibody Inc.) directed against the 11 amino acid HA epitope (see above) for 4 hr at 4°C. 50 1l of protein A Sepharose CL4B beads (Pharmacia) were added and extracts were allowed to incubate for an additional 90 min. Following five times washes in modified RIPA buffer, Sepharose beads were resuspended in 30 l1 of sample buffer for electrophoresis, heated at 100°C for 5 min, and separated on 1.5% SDS-polyacrylamide gels. Gels were then fixed with glacial acetic acid/methanol, treated with Enlightning (NEN), dried and autoradiographed.
Northern blot analysis and primer extension Total RNA was extracted from S.pombe transformants as previously described (20). 10 tsg of RNA was denatured and fractionated on 1 % agarose gels containing 6.7% formaldehyde. RNA was transferred to nitrocellulose filters and hybridized to a probe (32P-labelled by nick translation) containing either the whole CAT coding sequence, the entire tatHA coding sequence or a fragment of the S.pombe actin gene. The DNA fragments
A
B
C
2.3 71.35-
-TAT
woo
46I-
0.24-
30_n
_,
A
7.462 1.5-
4,
B
-TAT 14.3-
2.37-
kb
6.5-
7.461.35-
"
S
- CAT
3.4-
4.402.37-
i.
1.35-
CAT
1
2
3
2 37-
4
5
1
6
2
Actin
1.35-
1
2 3 4 5 6
2.37-
*..* 1
2
0*
3
4
5
1.35-.
6
Actin
1
2
3 4
5 6
Figure 2. Transcriptional activity of HIV-1 LTR promoter deletion mutants. S.pombe cells were transformed with the different HIV-1 LTR promoter deletion constructs shown in Figure 1. The results of CAT assays (A) and Northern blot analysis (B) are shown. Protein (A) or total RNA (B) was extracted from transformants carrying the following plasmids: pCHIV452 (lane 1, full length LTR); pCHIV278 (lane 2); pCHIV 176 (lane 3); pCHIV1 17 (lane 4); pCHIV80 (lane 5); pCHIV65 (lane 6). 1 isg of protein extract was used for the CAT assays. Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values 4 S.E.M; n = 3) were the following: pCHIV452 (0.78 + 0.03), pCHIV278 (1.3 + 0.2), pCHIV176 (1.2 0.1), pCHIV117 (29 + 7), pCHIV80 (1.4 + 0.1), pCHIV65 (0.12 + 0.02). In panel B (upper section), a 0.9kb BamHI restriction fragment containing the entire CAT coding sequence (cut out from pcDCAT, H.Okayama, unpublished data) was used as a probe. The lower section of panel B shows the same blot reprobed with a fragment of the Spombe actin gene (see Methods) as a control. The molecular size markers (kb) are shown on the left of the blot.
Figure 3. Transcriptional activity of the HIV-1 LTR promoter in the presence and absence of Tat. S.pombe cells were transformed with different HIV-1 LTR promoter deletion constructs together with pCMVL-tatHA or PCMVL (control). S.pombe extracts were assayed for CAT activity (A) and the presence of CAT and Tat mRNA (B). pCHIV452 (lane 1), pCHIV1 17 (lane 2), or pCHIV65 (lane 3) were cotransformed with pCMVL-tatHA. In parallel, pCHIV452 (lane 4), pCHIV1 17 (lane 5), or pCHIV65 (lane 6) were cotransformed with PCMVL which does not contain the tat gene. 1 og of protein extract was used for the CAT assays (A). Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values t S.E.M; n = 3) were: pCHIV452 + pCMVLtatHA (0.59 4 0.2), pCHIV 117 + pCMVL-tatHA (10.2 + 1.2), pCHIV65 + pCMVL-tatHA (0.15 4 0.03), pCHIV452 + pCMVL (0.36 i 0.10), pCHIV117 + pCMVL (17.4 3.5), pCHIV65 + pCMVL (0.21 X 0.01). (B) The probes used for the Northern analysis are described under Methods. A fragment of the S.pombe actin gene was used as a control probe (panel B, lower section). The molecular size markers (kb) are shown on the left of the blot. (C) Immunoprecipitation of TatHA fusion protein. Spombe cells transformed with pCMVL-tatHA (lane 1) or pCMVL (lane 2, control) were labelled with 35Scysteine as described under Methods. TatHA fusion protein was immunoprecipitated with monoclonal antibody 12CA5 directed against the HA epitope (see Methods). The molecular size markers (kD) are shown on the left of the panel.
2594 Nucleic Acids Research, Vol. 20, No. 10 used as probes were prepared as follows: The entire CAT coding region (0.9 kb) was excised from pcDCAT (gift from H.Okayama) with BamHI. For the tat probe, the whole tatHA coding region was generated by PCR using pCMVL-tatHA as a template and the upstream primer 5'-CTCGAGTTACTCGACAGAGGAGAGCAAGAA-3' and the downstream primer 5 '-ATCGCGATCACAAGCTAGCGTAATCTGGAACAT-
CGTATGGGTATTCCTTCGGGCCTGTCG-3' which covered with whole HA epitope. A 400 bp S.pombe actin gene fiagment corresponding to nucleotides 1083 to 1484 (as defined in ref. 27) was generated by PCR using 5'-CCAACCCTCAGCITI-3' as an upstream primer, 5'-TCAGAAGAATCGATGT-3' as a downstream primer, and a S.pombe genomic DNA library (28) as a template. The actin gene probe was used as a control probe to determine the amount and quality of yeast RNA loaded onto the gels. This fragment hybridizes with the three transripts of the actin gene which are 1.8, 1.6 and 1.2 kb long. For primer extension studies, an antisense oligonucleotide (5 '-CAACGGTGGTATATCCAGT-3') corresponding to nucleotides + 12 to +31 of the CAT coding sequence was labeled at its 5' end using 32P-ATP and T4 polynucleotide kinase and annealed to total yeast RNA. Primer extension was perforned as described (20). The reaction mixture was fractionated on a 6% polyacrylamide gel containing 8 M urea. A dideoxy sequencing reaction was run in parallel as a size marker.
RESULTS AND DISCUSSION Transcription The transcriptional activity of various HIV-1 LTR promoter deletion constructs (Figure 1) was measured by assaying the chloramphenicol acetyltransferase (CAT) activity in protein extracts prepared from S.pombe transformants. In addition, gene expression was monitored by Northern blot analysis using the CAT gene as a radiolabelled probe. The plasmid containing the whole HIV-l LTR (pCHIV452) and constructs with 5' pr r deletions extending downstream to -176 (pCHIV278 and pCHIV176) showed about 1% conversion of radiolabelled chloramphenicol (Figure 2A, lanes 1-3). pCHIV1 17, in which sequences upstream of the two NF-kB binding sites of the HIV-l LTR promoter have been deleted, showed an about 30-fold increase in CAT activity compared to pCHIV452 (Figure 2A, lane 4). pCHIV80 was at least 15 times less active than pCIHIV1 17 (Figure 2A, lanes 4 and 5). This plasmid (pCHIV80) does no longer contain the two NF-kB binding sites, but still has three copies of the Spl binding site (Figure 1). pCHIV65, which contains the core promoter and one and a half binding sites for the transcription factor Spl, did not give any detectable CAT activity (Figure 2A, lane 6). The results obtained by Northem blot analysis (Figure 2B) were in good agreement with the data obtained in the CAT assays (Figure 2A). Transformants containing pCHIV1 17 (lane 4) showed significantly higher levels of CAT mRNA (- 1.35 kb) than pCHIV452, pCHIV278, or pCHIV176 (Figure 2B, lanes 1-4). The amount of CAT mRNA seen with PCHIV8O (lane 5) was low, similar to that obtained with pCHIV452, pCHIV278, and pCHIV176. No CAT mRNA transcript was observed with pCHIV65. In this case only a high molecular weight band appeared on the Northern blot which is usually seen when the promoter activity of the CAT reporter construct is extremely low (R.Toyama, unpublished data). As a control, the same blot was
hybridized to a S.pombe actin gene probe, verifying that equal amounts of RNA were loaded in each lan of te gel. These data suggest that sequences between nucleoides -176 and -117 of the HIV-1 LTR promoter repress gene expression, and dut the region containing the NF'kR aepunces (-117 through -81) are required for strong HIV-l LTR promoter activity in Spombe. It is unlikely that the construct pCHIV1 17 results in an artifactual positive cis-elenrn at the junction of vector and -117 site, since an oligonucleotide -112 to -79 is able to efficiently activate a heterologous promoter (see below for details in the activadon of hete s ptonr). I Hmammlian cells, a number of tranriptin as incuding APl (-349 to -330), NFAT-1 (-254 to -216) and USF (-173 to -159) recognize sequences within the NRE (18). Deletion of NFAT-1 and USF sequences witiin the HIV-l LTR results in an increase in gene expression in ma lian cells (18). In this study, deletion of the segment betwee -176 to - 117 of the HIV-1 LTR allowed NF-kB sequences tlactivMe tanscription from the HIV-1 LTR promoter in S.pobe. tetngly, this sequence element (-176 to -117) con sa binding site for USF (-173 to -159) which has been shn toconributeto the repression of HIV-l LTR promoter activity in mamnalian cells (18). A more detailed analysis is required to elucidate whether S. pombe contains an USF-like factor and whedter the recognition sequence is identical to that found in mammalianl cells. The HIV-1 Tat protein is known to bind to the TAR RNA element (Figure 1) and to transactivate tnscription from the HIV-1 LTR promoter (12, 13). Control of transactivation can occur at different levels including increased rate of initation of transcription or elongation of the short RNAs which truncate within the HIV-1 leader sequence (7, 8, 10, 11, 14-16). We have tested the possibility that Tat may transactivate the HIV-1 LTR pro tr in the fission yeast. Initially, the plasmnid pCMVLtatHA encoding the Tat protein tagged with a short sequence encoding a 11 amino acid epitope (HA) of the Haemnphilus influnzae haemagglutinin protein (25) to allow its unolocal
t 50 142
12 FI1gua 4. 5' end analysis using primer extension. An adnwies oligomuciotide + 12 to + 31 of the CAT coding asoquan was labelled at its 5' end with using polynucleotide kinase, amealed to S.pombe RNA, and extended by reverse transcriptase. Total RNA, prepomd from cells cotransformed with pCHIV 1 17 and pCMVL -(lane, 1, contro; 16 ,sg RNA) or with pCHIV117 and pCMVLatHA (lane 2; 22 og RNA) was usd as a .tenbIt. In mamalian cells, transcription initiation of the HIV-I LTR occn at position
correspwnding to nulode 32P
147. A dideoxy sequencing reaction is shown on the left as a size standard.
Nucleic Acids Research, Vol. 20, No. 10 2595 detection, was tested for its ability to induce transactivation of the HIV-1 LTR promoter in Hela cells. Transactivation of the HIV-1 LTR by the TatHA fusion protein was compared to that by the Tat protein without the HA epitope. We found that the TatHA fusion protein was able to transactivate the HIV-1 promoter as efficiently as the Tat protein alone (data not shown). Subsequently, the ability of pCMVL-tatHA to induce synthesis of TatHA protein in S.pombe was tested by labelling exponentially growing S.pombe transformants with 35S-cysteine. In parallel, cells transformed with the vector (pCMVL) alone were processed as a control. Figure 3C shows that antibody 12CA5 was able to immunoprecipitate TatHA protein in cells transformed with pCMVL-tatHA but not in those transformed with pCMVL alone. Our effort to map the subcellular localization of the TatHA fusion protein in S.pombe was not successful because of the high background observed with the 12CA5 monoclonal antibody in the absence of Tat. The yeast strain [genotype: ura4-294 leul-32 h-] was cotransformed with pCMVL-tatHA and various HIV-1 LTR promoter deletion constructs described above. The plasmid pCMVL which does not carry the tatHA gene was used as a control. Transformation of S.pombe with pCHIV452, pCH1V1 17 or pCHIV65 together with pCMVL-tatHA resulted in CAT activities (Figure 3A, lanes 1-3) and CAT mRNA levels (Figure 3B, lanes 1-3) similar to those observed after cotransformation with pCMVL (Figure 3A, B, lanes 4-6). Tat mRNA was found in cells cotransformed with pCMVL-tatHa but not in those cotransformed with pCMVL (Figure 3B). Our data therefore suggest that Tat does not transactivate the HIV-1 LTR promoter in the fission yeast. Primer extension The 5' end of the TAR RNA sequence is important for Tat to transactivate the HIV-1 LTR promoter in mammalian cells (10). In order to examine the possibility that the demonstrated lack of Tat transactivation of the HIV-I LTR in S.pombe may be due to a different site of transcription initiation in S.pombe (resulting
A
in a loss of TAR sequences), primer extension analysis was performed. An antisense oligonucleotide (corresponding to nucleotides + 12 to +31 of the CAT coding sequence) was 5' end labelled, annealed to total RNA prepared from yeast cells cotransformed with pCHIV1 17 and pCMVL (Figure 4, lane 1) or cotransformed with pCHIV1 17 and pCMVL-tatHA (Figure 4, lane 2), and extended with reverse transcriptase. The 5' end analysis showed two major clusters of 5' ends, each cluster containing 3 to 4 bands (Figure 4). The major 5' ends were located 142 and 150 nucleotides away from the 5' end of the primer. The transcription initiation start site(s) of the HIV-1 LTR in S.pombe is therefore very similar to that found in mammalian cells (147 bases away from the 5' end of the primer). In fact, there are minor transcripts in S.pombe whose syndtesis is initiated exactly at the same site as in mammalian cells (Figure 4). Since the transcripts found in S.pombe would allow transactivation of the HIV-1 LTR by Tat in mammalian cells, the observed lack of transactivation by Tat in S.pombe appears not to be due to different points of transcription initiation. Recent reports suggest that transactivation of the HIV-1 promoter by Tat is partly dependent on the presence of Spi binding sites within the core promoter (17). One possible explanation for the inability of Tat to transactivate transcription from the HIV-1 LTR promoter in S.pombe may therefore be the lack of the Spl transcription factor in the fission yeast. Alternatively, S.pombe may lack a cellular factor(s) (other than Spi) necessary for the Tat transactivation process which has been proposed to exist in mammalian cells (17). Activation of a heterologous promoter by NF-kB sequences As outlined above, a short segment in the HIV-I LTR (-117 to -81) containing two NF-kB recognition sequences is responsible for the increase in promoter activity seen after removal of upstream promoter sequences. To further characterize this sequence element we inserted a 34 bp oligonucleotide
(5'-GCTACAAGGGACTTTCCGCTGGGGACTTTCCAGG-3') (-112 to -79; NF-kB sequence shown underlined) upstream of
oligo cloning
B
M,
ura 3
m NF kll3 GC TACAAGGAw
T TCTGG G C C
INF-KBm :GCTACAACTCACTTTCCGCTGC TCAC TTCCAGG
1
2
3
Figure 5. Transcriptional activity of pCEH core vector constructs containing NF-kB binding sequences. (A) An oligonucleotide corresponding to the NF-kB protein recognition sequence -83 to -105 (Figure 1) or the mutant sequence, NF-kBm, was inserted into the polylinker site upstream of the pCEH core promoter vector. pCEH contains the CGa core promoter directing transcription of the CAT reporter gene. (B) CAT analysis of protein extracts prepared from S.pombe cells transformed with pCEH (lane 1, control), pCEH-NF-kB (lane 2) and pCEH-NF-kBm (lane 3). Conversion rates of 14C chloramphenicol product (given as% acetylation/30 min; mean values L S.E.M; n = 3) were: pCEH (1.67 fi- 0.24), pCEH + NFkB (17.6 0.4), pCEH + NFkBm (6.17 -0.55).
2596 Nucleic Acids Research, Vol. 20, No. 10 the CGa core promoter into the yeast replicating vector pCEH. This URA3 based vector contains the CGa core promoter upstream of the CAT gene flanked by a 5' polylinker region which allows the insertion of enhancer sequences (Figure 5). For comparison, an additional oligonucleotide (5'-GCTACAACTCACTTTCCGCTGCTCACTTTCCAGG-3') containing mutations (shown in bold lefters) in the NF-kB protein recognition sequence which abolishes NF-kB binding in mammalian cells (29) was also cloned into pCEH (Figure 5). Yeast cells were transformed with both plasmids and CAT activities were determined in cell extracts as described. Transformants containing wild type NF-kB sequences displayed >9-fold activation of the CG a core promoter, whereas the mutant sequence (NF-kBm) resulted in only 2-fold activation. Together with the results presented in Figures 2 and 3, hes data suggest the existence of a cis-acting element containing two NFkB binding site (between -117 to -81) in the HIV-1 LTR which is required for promoter activation in the fission yeast. However, our data do not show that the tranription factor which recognizes this sequence element in S.pombe has a recognition sequence identical to that of NF-kB. It is unclear whether S.pombe will serve as a useful system for the study of the different functional domains of the HIV-1 LTR promoter, because the cis-acting sequences identified here may interact with yeast factors not relevant to the process in mammalian cells. On the other hand, the inability of Tat to activate the HIV-l LTR promoter in S.pombe may allow the isolation of genes which are involved in mediating this process in mammalian cells. To test this, we are presently developing a genetic selection system based on cotransformation of S.pombe with mammalian cDNA libraries.
ACKNOWLEDGMENTS We are grateful to Drs Susan Marriott and Jurgen Wess for critical review of the manuscript and to Dr Julie Brown for providing yeast replicating vectors before publication. We are also thankful to Drs Jay Rappaport and Cindy Bohan for providing us with plasmids carrying different HIV-1 LTR CAT constructs.
REFERENCES 1. 2. 3. 4. 5.
6. 7. 8. 9. 10. 11.
12.
13. 14. 15.
16.
Cullen,B.R. and Greene,W.C. (1989) Cell 58, 423 -426. Cullen,B.R. (1991) J. Virol. 65, 1053-1056. Pavlakis,G.N. and Felber,B.K. (1990) New Biologist 2, 20-31. Rosen,C.A. (1991) Trends Genet. 7, 9-14. Wright,C.M., Felber,B.K., Paskalis,H. and Pavlakis,G.N. (1986) Science 234, 988-992. Muesing,M.A., Smith,D.H. and Capon,D.J. (1987) Cell 48, 691-701. Laspia,M.F., Rice,A.P. and Mathews,M.B. (1989) Cell 59, 283-292. Feinberg,M.B., Baltimore,D. and Franker,A.D. (1991) Proc. Natl. Acad. Sci. USA- 88, 4045-4049. Hauber,J. and Cullen,B.R. (1988) J. Virol. 62, 673-679. Jakobovits,A., Smith,D.H., Jakobovits,E.B. and Capon,D.J. (1988) Mol. Cell. Biol. 8, 2555-2561. Selby,M.J., Bain,E.S., Luciw,P.A. and Peterlin,B.M. (1989) Genes Dev. 3, 547-558. Roy,S., Delling,U., Chen,C.-H., Rosen,C.A. and Sonenberg,N. (1990) Genes Dev. 4, 1365-1373. Weeks,K.M., Ampe,C., Schultz,S.C., Steitz,T.A. and Crohers,D.M. (1990) Science 249, 1281-1285. Marciniak,R.A., Garcia-Blanco,M.A. and Sharp,P.A. (1990) Proc. Natl. Acad. Sci. USA 87, 3624-3628. Kao,S.-Y., Calman,A.F., Luciw,P.A. and Peterlin,B.M. (1987) Nature 330, 489-493. Laspia,M.F., Rice,A.P. and Mathews,M.B. (1990) Genes Dev. 4, 2397-2408.
17. Kamine,J., Subramanian,T. and Chinnadurai,G. (1991) Proc. Natl. Acad. Sci. US 88, 8510-8514. 18. Lu,Y., Touzjian,N., Stenzel,M., Dorfman,T., Sodroski,J.G. and Haseltine,W.A. (1990) J. Virol. 64, 5226-5229. 19. Jones,R.H., Morene,S., Nurse,P. and Jones,N.C. (1988) Cell 53, 659-667. 20. Toyama, R., and Okayama, H. (1990) FEBS Let. 268, 217-221. 21. Hoffmann,A., Horikoshi,M., Wang,K.C., Schlr,S., Weil,P.A. and Roeder,R.G. (1990) Genes Dev. 4, 1141-1148. 22. Okazaki,K., Okazaki,N., Kume,K., Jinno,S., Tanaka,K. and Okayama,H. (1990) Nucleic Acids Res. 18, 6485-6489. 23. Bohan,C.A., Robinson,R. A., Luciw,P.A. and Srinivasan,A. (1989) V7rology 172, 573-583. 24. Boshart,M., Weber,F., Jahn,G., Dorsch-Haser,K., Flcckenstein,B. and Schaffner,W. (1985) Cell 41, 521-530. 25. Field,l., Nikawa,J., Brock,D., MacDonald,B., Rodgers,L., Wilson,I.A., Lrnier,R.A. and Wigler,M. (1988) Mol. Cell. Biol. 8, 2159-2165. 26. Bradford,M. (1976) Anal. Biochem. 72, 248-254. 27. Mertins,P. and Gailwitz,D. (1987) Nucleic Acids Res. 15, 7369-7379. 28. Wright,A., Maundrell,K., Heyer,W.-D., Beach,D. and Nurse,P. (1986) Plasmid 15, 156-158. 29. Duh,E.J., Maury,W.J., Folks,T.M., Fauci,A.S. and Rabson,A.B. (1989) Proc. Natl. Acad. Sci. USA 86, 5974-5978.
