JOURNAL OF VIROLOGY, Dec. 1992, p. 7159-7167
Vol. 66, No. 12
0022-538X/92/127159-09$02.00/0 Copyright X 1992, American Society for Microbiology
Effects of the Human Immunodeficiency Virus Type 1 Tat Protein on the Expression of Inflammatory Cytokines LUIGI BUONAGURO,1 GIOVANNI BARILLARI,1 HSIAO K. CHANG,' CINDY A. BOHAN,2 VIVIAN KAO,1 RICHARD MORGAN,3 ROBERT C. GALLO,' AND BARBARA ENSOLI1* Laboratory of Tumor Cell Biology, National Cancer Institute, 1 and Molecular Hematology Branch, National Heart, Lung and Blood Institute,3 Bethesda, Maryland 20892, and Department of Biology, Georgia Institute of Technology, Atlanta, Georgia 303322 Received 18 June 1992/Accepted 20 August 1992
Increased levels of inflammatory cytokines, including tumor necrosis factor (TNF), interleukin-1 (IL-1), and IL-6, have been detected in specimens from human immunodeficiency virus type 1 (HIV-1)-infected individuals. Here we demonstrate that HIV-1 activates the expression of TNF but not of IL-1 and IL-6 in acutely and chronically infected T cells. The increase in TNF gene expression is due to activation of the TNF promoter by the viral gene product Tat. Transactivation of TNF gene expression requires the product of the first exon of the tat gene and is cell type independent. T cells chronically infected with pol-defective HIV-1 provirus constitutively express both Tat and TNF at levels significantly higher (fivefold) than those seen in control cells, and treatment with phorbol myristate acetate greatly enhances Tat expression and TNF production. As TNF can increase the production of IL-1 and IL-6 and these inflammatory cytokines all enhance HIV-1 gene expression and affect the immune, vascular, and central nervous systems, the activation of TNF by Tat may be part of a complex pathway in which HIV-1 uses viral products and host factors to increase its own expression and infectivity and to induce disease.
Tumor necrosis factors a and ,B (TNFa and TNFP), interleukin-1 alpha (IL-la) and IL-1I, and IL-6 have been found to be elevated in the serum or produced at high levels by cultured leukocytes of human immunodeficiency virus type 1 (HIV-1)-infected individuals (9, 42, 57). Previous studies have indicated that these inflammatory cytokines activate HIV-1 gene expression and replication, interrupting virus latency and increasing virus spread (1, 4, 30, 55). These cytokines also modulate the immune, vascular, and central nervous systems (15, 39, 54). These observations have suggested that these cytokines play a role in the pathogenesis of AIDS and associated diseases, such as Kaposi's sarcoma (KS), B-cell lymphomas, and neurological disorders (4, 10, 16, 34, 49). In particular, TNF, IL-1, and, to a lesser extent, IL-6 stimulate the proliferation of spindle cells derived from KS lesions of AIDS patients (AIDS-KS cells) (4, 49) and induce KS cell phenotypic changes in normal vascular cells, the potential progenitors of the AIDS-KS cells (4). The increased level of inflammatory cytokines in HIV-1infected individuals may be due, at least in part, to antigenic stimulation, which is a frequent event in most of the groups at risk for AIDS (56, 69). In addition, HIV-1 infection itself induces the production of inflammatory cytokines from monocytes, through the binding of the envelope protein of HIV-1 (gpl20) to the CD4 molecule, and from B cells through other pathways (47, 60, 76). An alternative mechanism is that cytokine gene expression is transactivated by HIV-1 regulatory proteins in infected T cells. In support of this hypothesis is the observation that a B-cell line constitutively expressing the HIV-1 tat gene product (Tat) produces TNFI (67). The tax gene product (Tax) of human T-cell leukemia/ *
lymphotropic virus types I and II (HTLV-I and HTLV-II) activates both viral and cellular gene expression, including cytokines and cytokine receptors (13, 22, 32, 38, 77). The HIV-1 tat gene product (Tat) has functions analogous to those of Tax (2, 14, 24, 71). Tat transactivates HIV-1 gene expression by interacting with sequences in the HIV long terminal repeat (LTR) (5, 23, 28, 61) to promote transcription initiation of the integrated proviral genome and to stimulate the elongation of newly initiated viral transcripts (6, 33, 36, 59). Like Tax, Tat can activate heterologous promoters (58, 67, 72) and can mediate additional activities on both cellular and viral functions (4, 17, 18). Furthermore, extracellular Tat acts as a growth factor for cells derived from AIDS-KS lesions (17) and for normal vascular cells after treatment with inflammatory cytokines (4). Collectively, these observations suggest that Tat may activate the expression of cellular genes. In this study, we have analyzed the effects of Tat on the expression of TNFa and TNFI, IL-la and IL-1p, and IL-6 during acute HIV-1 infection in T cells chronically infected with a pol-defective HIV-1 provirus, in transient-transfection experiments in monocytic and lymphocytic cell lines constitutively expressing Tat, and after exposure of AIDS-KS cells to concentrations of extracellular Tat protein that induce cell growth. MATERIALS AND METHODS Cell lines. The epithelial cell line COS-1, the CD4+ T-cell lines H9, Jurkat, and A301, the T-cell line 8E51, chronically infected with a pol-defective HIV-1 provirus (26), and the promonocytic cell line U937 were maintained in RPMI-1640 medium supplemented with 10% fetal calf serum at 37°C in a humidified atmosphere containing 5% carbon dioxide in air. AIDS-KS cells were established and cultured as previously described (16, 17, 20).
Corresponding author. 7159
7160
J. VIROL.
BUONAGURO ET AL.
Stimulation of 8E51 and A301 cells with phorbol 12myristate 13-acetate (PMLA) was performed as follows. For DNA transfection experiments, cells were cultured for 16 h in growth medium containing PMA (1.0 p,M), washed with phosphate-buffered saline, and transfected as described below. For Northern (RNA) blot analysis and determination of the TNF content by enzyme-linked immunosorbent assay (ELISA), cells were cultured in the presence of PMA (0.1 p,M) for 12 to 48 h, as shown in the legends to the figures. Control cells were cultured under the same conditions without PMA. Tat-expressing U937 cell line. The tat gene (40) was introduced into the retroviral vector pXT1, which is driven by the Moloney murine leukemia virus LTR and contains the neomycin resistance gene (neo). This construct was transfected into a packaging cell line and selected with G418, as described previously (8, 50). The resulting virus was used for infection of U937 cells, which were selected again with G418. Plasmids. The CD7-CAT plasmid, containing the HIV-1 LTR; the ABS-CAT plasmid, containing the HIV-1 LTR deleted of 4 bp at the top of the stem-loop structure (+32 to +36) of the Tat trans-activation response (TAR) sequence; the pCV-Tat plasmid and its control vector pCVO; and the pTax, pTAT, and pSVO-CAT plasmids have all been previously described (13, 17, 19, 29, 40). The TNFP-CAT plasmids were constructed as follows. The sequence from -305 to + 115, containing the TNFP promoter region and a portion of exon I for the gene (52), was amplified by the polymerase chain reaction technique, as described previously (64). The primers used for the polymerase chain reaction amplification were 5'-CAGCTGCTCA ACCACCTCCT-3' (-305 to -286) and 5'-CCAAGGCCCA GGCCCAGGCA-3' (+144 to + 163). The adaptor sequence 5'-ATTAGAAGC TT-3', cleaved by the HindIII restriction enzyme, was added to the 5' end of each primer. The amplified fragment was then digested with HindIII and cloned upstream of the chloramphenicol acetyltransferase (CAT) gene in the pSVO-CAT plasmid (29), in both the sense (pTNFP) and antisense (pASTNFP) orientation. The nucleotide sequence and the orientation of the two constructs were confirmed by sequence analysis. Transfection experiments and CAT assay. Transfections of the plasmid DNA were performed by the electroporation procedure, which was adapted to suspended and adherent cells. For the lymphocytic and monocytic cell lines, 3 x 107 cells were resuspended in 0.25 ml of the growth medium containing EGTA (ethylene glycol tetraacetic acid, 1 mM), transferred to Eppendorf tubes, and gently mixed with 10 jig of total plasmid DNA. After 10 min of incubation in ice, the cells were electroporated at 960 ,uF and 0.25 kV with a Bio-Rad Gene Pulser. After an additional 10 min in ice, the cells were resuspended in 25 ml of warm (37°C) growth medium and harvested 72 h later. For adherent cells, 80 to 90% confluent COS-1 cells from a 175-cm3 flask were treated with cold trypsin, washed, and gently resuspended in 1.6 ml of electroporation buffer (272 mM sucrose, 7 mM potassium phosphate [pH 7.4], 1 mM MgCl2). Eight hundred microliters of the cell suspension was mixed gently with 30 ,g of total plasmid DNA and incubated in ice for 10 min. Cells were electroporated at 25 ,uF and 0.28 kV. After an additional 10 min in ice, the cells were very gently resuspended in 25 ml of warm (37°C) growth medium and harvested 48 h later. CAT assays were performed with cell lysates (normalized for total protein content) as described previously (19, 29). Lysates from cells transfected with the CD7-CAT plasmid were incubated for 0.5 to 1.5 h, while those from cells transfected
with TNF,B-CAT plasmids were incubated for 6 h or overnight. HIV-1 infection of H9 cell line. H9 cells (2 x 107, 106/ml) were infected with 6 x 105 cpm of reverse transcriptase from the HIV-1 IIIB isolate (17). H9 cells (106/ml) were maintained in growth medium, and viral infection was monitored by evaluating syncytium formation, reverse transcriptase activity, and HIV-1 p24 expression. Total RNA was extracted from infected and uninfected control cells at 3-day intervals postinfection for 12 days (RNAzol; CINNA/
BIOTECX). Northern blot analysis. Northern blot analyses were performed with 10 ,ug of total RNA, as described previously (46). The filters were incubated at 37°C for 2 h in prehybridization solution (3x SSC [lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 5 x Denhardt's reagent, 0.7% sodium dodecyl sulfate [SDS], salmon sperm DNA [100 ,ug/ml], yeast tRNA [200 jig/ml], 50% formamide, 10% dextran sulfate) and subsequently hybridized, under the same conditions, with 2 x 106 cpm of radiolabeled genomic DNA probe per ml and antisense oligonucleotides for TNFa, TNFP, IL-la and IL-1,, IL-6, and tat. Hybridizations with the antisense oligonucleotide probes were performed at 37°C with the same solution as above except that 5 x SSC and 30% formamide were also used. The filters were washed extensively with 2x SSC-1% SDS at 65°C and exposed on X-ray film (Kodak) at -70°C for autoradiography. Cytokine ELISA. The intracellular and extracellular TNFi content was determined for unstimulated and PMA-stimulated (0.1 ,uM) 8E51 and A301 cells. Eight milliliters of cell supernatant was harvested, centrifuged for 30 min (4°C) at 3,000 x g, and concentrated by ultrafiltration (Centricon 10; Amicon). The values obtained for the concentrated cell supernatants were normalized to the original volume. The cellular pellets were extracted as described before (19, 29) and normalized to the total protein content. The TNF3 ELISA was performed as described before (3) with monoclonal and polyclonal anti-TNFI antibodies (BoehringerMannheim and Genzyme, respectively), human recombinant TNFI (R&D System), and revelatory enzymatic systems (Kirkegaard & Perry Laboratories, Inc.). The results were confirmed with a commercially available ELISA kit (R&D
System). RESULTS Cytokine gene expression in acutely HIV-1-infected H9 cells. To investigate whether HIV-1 infection was directly involved in inducing the expression of inflammatory cytokines, CD4+ H9 T cells were infected acutely with HIV-1, and the expression of TNFa and TNF,B, IL-la and IL-1, and IL-6 was analyzed at 3, 6, 9, and 12 days postinfection (p.i.) (Fig. 1A and data not shown). The amount of TNFa and TNFP mRNAs increased during acute HIV-1 infection (2.6and 8.5-fold, respectively, at 9 days p.i.) compared with the amounts in uninfected control cells, whereas IL-1- and IL-6-specific transcripts were undetectable in both uninfected and HIV-1-infected cells (data not shown). These results were corroborated by the results from a radioimmunoprecipitation assay and ELISA of these cytokines in cell extracts and supernatants from infected and uninfected H9 cells (data not shown). Hybridization with the tat cDNA probe indicated that the kinetics of TNF and Tat-containing viral transcripts were temporally associated, as the levels of Tat and TNF,B transcripts peaked at 9 days p.i. (237- and 8.5-fold, respectively) (Fig. 1A and B). At this time, both
VOL. 66, 1992
A
DAYS POST INFECTION
H9
H9-HIV-1
3
6
M
9 12
TNF .'
Kb -2.4
.
--1.4
C-, z
0
c)
U)
0~LU
U B
H9-HIV-1
3
6
9
7161
1.2r
08
0.6
x 0.4 UZ 0.2
DAYS POST-INFECTION H-9
A
Tat ACTIVATION OF TNFs DURING HIV INFECTION
M
12 Kb
0
3 H9
6
9 2 H9-H IV-1 DAYS POST-INFECTION
f
L)
z 0
_9.5 -7.5 -4.4
Ln
LXI
CL
-2.4
1III 3
-1.4
H9
6
9 H9-HIV-1 DAYS POST- INFECTION
12
FIG. 1. Steady-state levels of TNFI mRNA and tat-containing viral transcripts during acute HIV-1 infection of H9 cells. (A) Total cellular RNA was isolated at 0, 3, 6, 9, and 12 days p.i. and analyzed by the Northern blot technique (10 ,ug per lane), as described in Materials and Methods. M, size markers. The ethidium bromide-stained gel is shown at the bottom left. The filters were hybridized with a genomic TNF3 probe (left panel) and subsequently with the TNFa, IL-1, and IL-6 probes (data not shown). The right panel shows the quantitation of TNF, hybridization results by integral analysis on the PhosphorImager (Molecular Dynamics). RNA levels were normalized for amount, transfer, and hybridization by comparison with hybridization results obtained with an actin cDNA probe on the same filters (data not shown). The intensity of the radioactive signal in the uninfected control cells was assigned a value of 1. (B) The filters in panel A were stripped and hybridized with a tat cDNA probe (left panel). The right panel shows the quantitation of tat hybridization radioactive signal of the three viral transcripts with the Phosphorlmager. The intensity of the radioactive signal detected at day 3 p.i. was assigned a value of 1. At day 9 p.i., the increase in viral mRNA expression was 237-fold.
Tat-expressing cells and TNF-positive infected cells were detected by immunohistochemistry (data not shown). These results indicated that HIV-1 infection activated the expression of TNFa and TNF,B but not that of other inflammatory cytokines and suggested that Tat may be responsible for this induction. TNF gene expression and promoter activation in Tat-expressing U937 cells. To determine whether the tat gene was able to activate TNF gene expression, a monocytic cell line (U937) which expresses no TNF constitutively was infected with a Moloney murine leukemia virus LTR-driven retrovirus vector carrying the tat gene and the neo gene, which confers resistance to the selective agent G418 (U937-Tat). The control cell line was infected with the same retrovirus vector that lacked the tat sequence (U937-CR). After 3 weeks of selection, cells were analyzed for the expression of tat and cytokine genes (Fig. 2). The U937-Tat cell line but not the U937-CR cells expressed high levels of both Tat and TNF mRNAs (Fig. 2), while transcripts of the other cytok-
ines were not detectable (data not shown). As expected, HTLV-I infection, known to activate the expression of TNF and of other cytokines (32, 38, 75, 77), resulted in higher levels of TNFI mRNA, which served as a positive TNFI control (Fig. 2). These results confirmed previous data (67) and indicated that HIV-1 Tat was sufficient to activate TNF gene expression. To investigate the mechanism of Tat induction of TNF gene expression, studies were performed with the TNFI gene promoter. The promoter region and part of the first exon of the TNFI3 gene (-305 to +115) (52) was cloned in both the sense (pTNF3) and antisense (pASTNF3) orientation into pSVO-CAT (29) (Fig. 3A). TNFP promoter-CAT plasmids were transfected into U937-Tat and U937-CR cells. TNF,B-CAT expression was detected only in U937-Tat cells (Fig. 3B, right panel; Table 1). Maximal activation of the TNF3 promoter correlated well with high levels of tat mRNA and Tat activity, as measured from transfections involving Tat and the HIV-1 LTR (CD7-CAT) (17). The
7162
J. VIROL.
BUONAGURO ET AL.
Vol. 66, No. 12
0022-538X/92/127159-09$02.00/0 Copyright X 1992, American Society for Microbiology
Effects of the Human Immunodeficiency Virus Type 1 Tat Protein on the Expression of Inflammatory Cytokines LUIGI BUONAGURO,1 GIOVANNI BARILLARI,1 HSIAO K. CHANG,' CINDY A. BOHAN,2 VIVIAN KAO,1 RICHARD MORGAN,3 ROBERT C. GALLO,' AND BARBARA ENSOLI1* Laboratory of Tumor Cell Biology, National Cancer Institute, 1 and Molecular Hematology Branch, National Heart, Lung and Blood Institute,3 Bethesda, Maryland 20892, and Department of Biology, Georgia Institute of Technology, Atlanta, Georgia 303322 Received 18 June 1992/Accepted 20 August 1992
Increased levels of inflammatory cytokines, including tumor necrosis factor (TNF), interleukin-1 (IL-1), and IL-6, have been detected in specimens from human immunodeficiency virus type 1 (HIV-1)-infected individuals. Here we demonstrate that HIV-1 activates the expression of TNF but not of IL-1 and IL-6 in acutely and chronically infected T cells. The increase in TNF gene expression is due to activation of the TNF promoter by the viral gene product Tat. Transactivation of TNF gene expression requires the product of the first exon of the tat gene and is cell type independent. T cells chronically infected with pol-defective HIV-1 provirus constitutively express both Tat and TNF at levels significantly higher (fivefold) than those seen in control cells, and treatment with phorbol myristate acetate greatly enhances Tat expression and TNF production. As TNF can increase the production of IL-1 and IL-6 and these inflammatory cytokines all enhance HIV-1 gene expression and affect the immune, vascular, and central nervous systems, the activation of TNF by Tat may be part of a complex pathway in which HIV-1 uses viral products and host factors to increase its own expression and infectivity and to induce disease.
Tumor necrosis factors a and ,B (TNFa and TNFP), interleukin-1 alpha (IL-la) and IL-1I, and IL-6 have been found to be elevated in the serum or produced at high levels by cultured leukocytes of human immunodeficiency virus type 1 (HIV-1)-infected individuals (9, 42, 57). Previous studies have indicated that these inflammatory cytokines activate HIV-1 gene expression and replication, interrupting virus latency and increasing virus spread (1, 4, 30, 55). These cytokines also modulate the immune, vascular, and central nervous systems (15, 39, 54). These observations have suggested that these cytokines play a role in the pathogenesis of AIDS and associated diseases, such as Kaposi's sarcoma (KS), B-cell lymphomas, and neurological disorders (4, 10, 16, 34, 49). In particular, TNF, IL-1, and, to a lesser extent, IL-6 stimulate the proliferation of spindle cells derived from KS lesions of AIDS patients (AIDS-KS cells) (4, 49) and induce KS cell phenotypic changes in normal vascular cells, the potential progenitors of the AIDS-KS cells (4). The increased level of inflammatory cytokines in HIV-1infected individuals may be due, at least in part, to antigenic stimulation, which is a frequent event in most of the groups at risk for AIDS (56, 69). In addition, HIV-1 infection itself induces the production of inflammatory cytokines from monocytes, through the binding of the envelope protein of HIV-1 (gpl20) to the CD4 molecule, and from B cells through other pathways (47, 60, 76). An alternative mechanism is that cytokine gene expression is transactivated by HIV-1 regulatory proteins in infected T cells. In support of this hypothesis is the observation that a B-cell line constitutively expressing the HIV-1 tat gene product (Tat) produces TNFI (67). The tax gene product (Tax) of human T-cell leukemia/ *
lymphotropic virus types I and II (HTLV-I and HTLV-II) activates both viral and cellular gene expression, including cytokines and cytokine receptors (13, 22, 32, 38, 77). The HIV-1 tat gene product (Tat) has functions analogous to those of Tax (2, 14, 24, 71). Tat transactivates HIV-1 gene expression by interacting with sequences in the HIV long terminal repeat (LTR) (5, 23, 28, 61) to promote transcription initiation of the integrated proviral genome and to stimulate the elongation of newly initiated viral transcripts (6, 33, 36, 59). Like Tax, Tat can activate heterologous promoters (58, 67, 72) and can mediate additional activities on both cellular and viral functions (4, 17, 18). Furthermore, extracellular Tat acts as a growth factor for cells derived from AIDS-KS lesions (17) and for normal vascular cells after treatment with inflammatory cytokines (4). Collectively, these observations suggest that Tat may activate the expression of cellular genes. In this study, we have analyzed the effects of Tat on the expression of TNFa and TNFI, IL-la and IL-1p, and IL-6 during acute HIV-1 infection in T cells chronically infected with a pol-defective HIV-1 provirus, in transient-transfection experiments in monocytic and lymphocytic cell lines constitutively expressing Tat, and after exposure of AIDS-KS cells to concentrations of extracellular Tat protein that induce cell growth. MATERIALS AND METHODS Cell lines. The epithelial cell line COS-1, the CD4+ T-cell lines H9, Jurkat, and A301, the T-cell line 8E51, chronically infected with a pol-defective HIV-1 provirus (26), and the promonocytic cell line U937 were maintained in RPMI-1640 medium supplemented with 10% fetal calf serum at 37°C in a humidified atmosphere containing 5% carbon dioxide in air. AIDS-KS cells were established and cultured as previously described (16, 17, 20).
Corresponding author. 7159
7160
J. VIROL.
BUONAGURO ET AL.
Stimulation of 8E51 and A301 cells with phorbol 12myristate 13-acetate (PMLA) was performed as follows. For DNA transfection experiments, cells were cultured for 16 h in growth medium containing PMA (1.0 p,M), washed with phosphate-buffered saline, and transfected as described below. For Northern (RNA) blot analysis and determination of the TNF content by enzyme-linked immunosorbent assay (ELISA), cells were cultured in the presence of PMA (0.1 p,M) for 12 to 48 h, as shown in the legends to the figures. Control cells were cultured under the same conditions without PMA. Tat-expressing U937 cell line. The tat gene (40) was introduced into the retroviral vector pXT1, which is driven by the Moloney murine leukemia virus LTR and contains the neomycin resistance gene (neo). This construct was transfected into a packaging cell line and selected with G418, as described previously (8, 50). The resulting virus was used for infection of U937 cells, which were selected again with G418. Plasmids. The CD7-CAT plasmid, containing the HIV-1 LTR; the ABS-CAT plasmid, containing the HIV-1 LTR deleted of 4 bp at the top of the stem-loop structure (+32 to +36) of the Tat trans-activation response (TAR) sequence; the pCV-Tat plasmid and its control vector pCVO; and the pTax, pTAT, and pSVO-CAT plasmids have all been previously described (13, 17, 19, 29, 40). The TNFP-CAT plasmids were constructed as follows. The sequence from -305 to + 115, containing the TNFP promoter region and a portion of exon I for the gene (52), was amplified by the polymerase chain reaction technique, as described previously (64). The primers used for the polymerase chain reaction amplification were 5'-CAGCTGCTCA ACCACCTCCT-3' (-305 to -286) and 5'-CCAAGGCCCA GGCCCAGGCA-3' (+144 to + 163). The adaptor sequence 5'-ATTAGAAGC TT-3', cleaved by the HindIII restriction enzyme, was added to the 5' end of each primer. The amplified fragment was then digested with HindIII and cloned upstream of the chloramphenicol acetyltransferase (CAT) gene in the pSVO-CAT plasmid (29), in both the sense (pTNFP) and antisense (pASTNFP) orientation. The nucleotide sequence and the orientation of the two constructs were confirmed by sequence analysis. Transfection experiments and CAT assay. Transfections of the plasmid DNA were performed by the electroporation procedure, which was adapted to suspended and adherent cells. For the lymphocytic and monocytic cell lines, 3 x 107 cells were resuspended in 0.25 ml of the growth medium containing EGTA (ethylene glycol tetraacetic acid, 1 mM), transferred to Eppendorf tubes, and gently mixed with 10 jig of total plasmid DNA. After 10 min of incubation in ice, the cells were electroporated at 960 ,uF and 0.25 kV with a Bio-Rad Gene Pulser. After an additional 10 min in ice, the cells were resuspended in 25 ml of warm (37°C) growth medium and harvested 72 h later. For adherent cells, 80 to 90% confluent COS-1 cells from a 175-cm3 flask were treated with cold trypsin, washed, and gently resuspended in 1.6 ml of electroporation buffer (272 mM sucrose, 7 mM potassium phosphate [pH 7.4], 1 mM MgCl2). Eight hundred microliters of the cell suspension was mixed gently with 30 ,g of total plasmid DNA and incubated in ice for 10 min. Cells were electroporated at 25 ,uF and 0.28 kV. After an additional 10 min in ice, the cells were very gently resuspended in 25 ml of warm (37°C) growth medium and harvested 48 h later. CAT assays were performed with cell lysates (normalized for total protein content) as described previously (19, 29). Lysates from cells transfected with the CD7-CAT plasmid were incubated for 0.5 to 1.5 h, while those from cells transfected
with TNF,B-CAT plasmids were incubated for 6 h or overnight. HIV-1 infection of H9 cell line. H9 cells (2 x 107, 106/ml) were infected with 6 x 105 cpm of reverse transcriptase from the HIV-1 IIIB isolate (17). H9 cells (106/ml) were maintained in growth medium, and viral infection was monitored by evaluating syncytium formation, reverse transcriptase activity, and HIV-1 p24 expression. Total RNA was extracted from infected and uninfected control cells at 3-day intervals postinfection for 12 days (RNAzol; CINNA/
BIOTECX). Northern blot analysis. Northern blot analyses were performed with 10 ,ug of total RNA, as described previously (46). The filters were incubated at 37°C for 2 h in prehybridization solution (3x SSC [lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 5 x Denhardt's reagent, 0.7% sodium dodecyl sulfate [SDS], salmon sperm DNA [100 ,ug/ml], yeast tRNA [200 jig/ml], 50% formamide, 10% dextran sulfate) and subsequently hybridized, under the same conditions, with 2 x 106 cpm of radiolabeled genomic DNA probe per ml and antisense oligonucleotides for TNFa, TNFP, IL-la and IL-1,, IL-6, and tat. Hybridizations with the antisense oligonucleotide probes were performed at 37°C with the same solution as above except that 5 x SSC and 30% formamide were also used. The filters were washed extensively with 2x SSC-1% SDS at 65°C and exposed on X-ray film (Kodak) at -70°C for autoradiography. Cytokine ELISA. The intracellular and extracellular TNFi content was determined for unstimulated and PMA-stimulated (0.1 ,uM) 8E51 and A301 cells. Eight milliliters of cell supernatant was harvested, centrifuged for 30 min (4°C) at 3,000 x g, and concentrated by ultrafiltration (Centricon 10; Amicon). The values obtained for the concentrated cell supernatants were normalized to the original volume. The cellular pellets were extracted as described before (19, 29) and normalized to the total protein content. The TNF3 ELISA was performed as described before (3) with monoclonal and polyclonal anti-TNFI antibodies (BoehringerMannheim and Genzyme, respectively), human recombinant TNFI (R&D System), and revelatory enzymatic systems (Kirkegaard & Perry Laboratories, Inc.). The results were confirmed with a commercially available ELISA kit (R&D
System). RESULTS Cytokine gene expression in acutely HIV-1-infected H9 cells. To investigate whether HIV-1 infection was directly involved in inducing the expression of inflammatory cytokines, CD4+ H9 T cells were infected acutely with HIV-1, and the expression of TNFa and TNF,B, IL-la and IL-1, and IL-6 was analyzed at 3, 6, 9, and 12 days postinfection (p.i.) (Fig. 1A and data not shown). The amount of TNFa and TNFP mRNAs increased during acute HIV-1 infection (2.6and 8.5-fold, respectively, at 9 days p.i.) compared with the amounts in uninfected control cells, whereas IL-1- and IL-6-specific transcripts were undetectable in both uninfected and HIV-1-infected cells (data not shown). These results were corroborated by the results from a radioimmunoprecipitation assay and ELISA of these cytokines in cell extracts and supernatants from infected and uninfected H9 cells (data not shown). Hybridization with the tat cDNA probe indicated that the kinetics of TNF and Tat-containing viral transcripts were temporally associated, as the levels of Tat and TNF,B transcripts peaked at 9 days p.i. (237- and 8.5-fold, respectively) (Fig. 1A and B). At this time, both
VOL. 66, 1992
A
DAYS POST INFECTION
H9
H9-HIV-1
3
6
M
9 12
TNF .'
Kb -2.4
.
--1.4
C-, z
0
c)
U)
0~LU
U B
H9-HIV-1
3
6
9
7161
1.2r
08
0.6
x 0.4 UZ 0.2
DAYS POST-INFECTION H-9
A
Tat ACTIVATION OF TNFs DURING HIV INFECTION
M
12 Kb
0
3 H9
6
9 2 H9-H IV-1 DAYS POST-INFECTION
f
L)
z 0
_9.5 -7.5 -4.4
Ln
LXI
CL
-2.4
1III 3
-1.4
H9
6
9 H9-HIV-1 DAYS POST- INFECTION
12
FIG. 1. Steady-state levels of TNFI mRNA and tat-containing viral transcripts during acute HIV-1 infection of H9 cells. (A) Total cellular RNA was isolated at 0, 3, 6, 9, and 12 days p.i. and analyzed by the Northern blot technique (10 ,ug per lane), as described in Materials and Methods. M, size markers. The ethidium bromide-stained gel is shown at the bottom left. The filters were hybridized with a genomic TNF3 probe (left panel) and subsequently with the TNFa, IL-1, and IL-6 probes (data not shown). The right panel shows the quantitation of TNF, hybridization results by integral analysis on the PhosphorImager (Molecular Dynamics). RNA levels were normalized for amount, transfer, and hybridization by comparison with hybridization results obtained with an actin cDNA probe on the same filters (data not shown). The intensity of the radioactive signal in the uninfected control cells was assigned a value of 1. (B) The filters in panel A were stripped and hybridized with a tat cDNA probe (left panel). The right panel shows the quantitation of tat hybridization radioactive signal of the three viral transcripts with the Phosphorlmager. The intensity of the radioactive signal detected at day 3 p.i. was assigned a value of 1. At day 9 p.i., the increase in viral mRNA expression was 237-fold.
Tat-expressing cells and TNF-positive infected cells were detected by immunohistochemistry (data not shown). These results indicated that HIV-1 infection activated the expression of TNFa and TNF,B but not that of other inflammatory cytokines and suggested that Tat may be responsible for this induction. TNF gene expression and promoter activation in Tat-expressing U937 cells. To determine whether the tat gene was able to activate TNF gene expression, a monocytic cell line (U937) which expresses no TNF constitutively was infected with a Moloney murine leukemia virus LTR-driven retrovirus vector carrying the tat gene and the neo gene, which confers resistance to the selective agent G418 (U937-Tat). The control cell line was infected with the same retrovirus vector that lacked the tat sequence (U937-CR). After 3 weeks of selection, cells were analyzed for the expression of tat and cytokine genes (Fig. 2). The U937-Tat cell line but not the U937-CR cells expressed high levels of both Tat and TNF mRNAs (Fig. 2), while transcripts of the other cytok-
ines were not detectable (data not shown). As expected, HTLV-I infection, known to activate the expression of TNF and of other cytokines (32, 38, 75, 77), resulted in higher levels of TNFI mRNA, which served as a positive TNFI control (Fig. 2). These results confirmed previous data (67) and indicated that HIV-1 Tat was sufficient to activate TNF gene expression. To investigate the mechanism of Tat induction of TNF gene expression, studies were performed with the TNFI gene promoter. The promoter region and part of the first exon of the TNFI3 gene (-305 to +115) (52) was cloned in both the sense (pTNF3) and antisense (pASTNF3) orientation into pSVO-CAT (29) (Fig. 3A). TNFP promoter-CAT plasmids were transfected into U937-Tat and U937-CR cells. TNF,B-CAT expression was detected only in U937-Tat cells (Fig. 3B, right panel; Table 1). Maximal activation of the TNF3 promoter correlated well with high levels of tat mRNA and Tat activity, as measured from transfections involving Tat and the HIV-1 LTR (CD7-CAT) (17). The
7162
J. VIROL.
BUONAGURO ET AL.
