Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Latent cytomegalovirus (CMV) infection is an important cause of disease in immunocompromised hosts, including transplant recipients (1). The mechanisms of reactivation of CMV are unclear, but as activation of the major immediate early (IE) gene of CMV is necessary for productive infection (6), factors that stimulate this region could be involved in this process. Adenovirus coinfection with CMV has been detected in the lung in a number of cases of pneumonitis in patients with acquired immunodeficiency syndrome (AIDS) and transplant recipients (9, 10). We have previously shown that a gene product of the adenovirus E1A 13S region stimulates the CMV major IE promoter in inflammatory cells (11). As both adenovirus and CMV can persist in inflammatory cells (12) and adenovirus early genes are active during both latent and active infection (15, 16), this observation provides one potential mechanism of stimulation of CMV.

Products of the early genes of adenovirus, including E1A 13S 289R and E1A 12S 246R, stimulate downstream promoter elements of adenovirus, including adenovirus E2, E3, and E4 (17). We previously identified a specific conserved region of the 13S gene product called conserved region (CR)3 as responsible for transactivation of CMV major IE. The E1A 12S gene product, which is identical to the 13S gene product except for the absence of CR3, failed to stimulate CMV promoter activity (11). The zinc finger and carboxyl subdomains of the E1A 13S CR3 appear to be important for transactivation of adenovirus E3 and E4. One purpose of this study was to determine if these subdomains are also important for transactivation of the CMV IE promoter by adenovirus. To test this hypothesis, we performed transient transfections of the T lymphocyte-like Jurkat cell line and the monocyte/macrophage-like THP-1 cell line with plasmids expressing CR3 mutant and wild-type (WT) adenovirus E1A 13S.

To determine whether a similar mechanism involving general and specific transcription factors as proposed for E1A transactivation of adenovirus E3 was occurring and to confirm the importance of the specific domains of CR3 on transactivation of the CMV major IE promoter, we performed a series of mixing studies using the various E1A mutants. If the mutant protein contains a functional transactivating element, increasing ratios of mutant to WT DNA should result in enhanced transactivation. If no functioning element is present in the mutant protein, there should be a minimal effect of increasing ratios on transactivation. If the mutant proteins sequester limiting transcription factors, increasing the ratio of mutant to WT DNA should result in a net loss of transactivation (22).

We also sought to determine the important region of the CMV IE promoter involved in this phenomena. This promoter contains several positive and negative regulatory elements. These are contained in multiple repeat sequences. The positive regulatory elements include binding sites for transcription factors nuclear factor B (NF-B), cyclic AMP response element binding protein/activating transcription factor (CREB/ATF), and activator protein 1 (AP-1) (23). The NF-B and CREB/ATF sites are contained in the 18 and 19 bp repeats, respectively. The CREB/ATF element appears to be important in transactivation of the CMV IE promoter in epithelial cells (24). There are also negative regulatory elements present, including one for YY1 contained with SP-1 in the 21 bp repeat. To determine the important site for transactivation in our system, we performed transient transfections in THP-1 cells using 5' promoter deletions of the CMV IE promoter and adenovirus E1A expression vectors. Putative promoter elements identified were confirmed by repeating the initial studies using CMV IE single promoter element constructs. Finally, to determine if activation of the putative promoter element occurred through the specific subdomains of E1A described previously, the putative element was tested for transactivation activity in the presence of CR3 subdomain mutants.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture

The Jurkat (ATCC TIB 152) and THP-1 (ATCC TIB 202) cell lines were obtained from the American Type Culture Collection (Rockville, MD) and used for these studies as both adenovirus and CMV infect lymphocytes and monocyte/macrophages (25- 28). The Jurkat line was derived from a cell line established from a patient with T-cell leukemia. This derived line produces interleukin-2 and -interferon on stimulation and is CD2⁺, CD3⁺, CD4, and CD8. The THP-1 line was derived from a patient with acute monocytic leukemia, has Fc and C3b receptors, produces lysozymes, and is phagocytic. Both cell types were maintained in suspension cell culture in RPMI 1640 medium containing 10% fetal bovine serum (Hyclone Laboratories, Logan, UT), 2 mM L-glutamine, and 80 µg/ml gentamicin.

Plasmids

The plasmid pSKE1AWT codes for the WT E1A 13S 289R protein. The plasmids pSKC171S and pSKS185N contain mutations resulting in single amino-acid substitutions in the carboxyl (S185N) and zinc finger (C171S) subdomains of CR3. Plasmid pSKC171S/ S185N encodes a double mutant of CR3. Plasmid pSKT178S contains a mutation causing a single amino-acid substitution in the region between the zinc finger and carboxyl subdomains. All of these plasmids were kind gifts of Dr. Robert Ricciardi (Wistar Institute, Philadelphia, PA) (22). The parent control plasmid, pBIISK, was obtained from Stratagene (La Jolla, CA). The CMV IE promoter- chloramphenicol acetyltransferase (CAT) construct pCAT-760WT plasmid contains the CMV promoter from 753 to +7 upstream of the protein coding region of bacterial CAT. Plasmids pCAT-dl36, pCAT-dl14, pCAT-dlA231, and pCAT-dl4 contain deletions of the CMV IE promoter, resulting in a loss of 16, 18, 19, and 21 bp repeats. Plasmid pCAT-dl760 contains a minimal WT promoter and initiation site (68 to +7) but lacks any of the upstream elements. Plasmids pIE1-161, pIE1-181, pIE1-191, and pIE1-211 contain the 16, 18, 19, and 21-bp single CMV IE promoter element repeats upstream of the minimal CMV IE promoter and initiation site linked to CAT. All are kind gifts of Dr. M. Stinski (University of Iowa, Iowa City, IA) (29). The human immunodeficiency virus 1 (HIV-1) long terminal repeat (LTR) promoter-CAT construct pHIV-1, which contains the HIV-1 LTR 450 to +180 upstream of CAT, was a kind gift of Dr. B. M. Peterlin (University of California, San Francisco, San Francisco, CA) (30). This construct was used to determine if transactivation by adenovirus E1A was virus-specific in the cells tested. The pTK--GAL plasmid contains the Herpes simplex virus thymidine kinase promoter linked to the -galactosidase protein coding gene and was constructed from the pSV--gal plasmid (Promega, Madison, WI) and the pMC-Neo plasmid (Promega). It was used as a control for transfection efficiency between separate transfections (22). The sequence of all plasmids was confirmed by automated DNA sequencing (ABI Systems, Foster City, CA).

Transient Transfections

Transfection of Jurkat cells was performed using electroporation. Cells were resuspended at a concentration of 2.5 × 10⁷ cells/ml in the same medium used for culture (described previously). A total of 0.4 ml of this suspension was added to 0.4-mm electroporation cuvettes. After adding plasmid DNA and gentle mixing, the cells were electroporated at settings of 960 µF and 0.29 V. Transfection of THP-1 cells was performed using the diethylaminoethyl (DEAE) dextran method as described previously (11). Plasmids were suspended in 10 ml of a solution containing 100 µg/ml DEAE dextran, Tris pH 7.0, and MgCL₂. The cultures were exposed to the plasmids for 20 min and washed once with RPMI 1640 containing 1.5 U/ml heparin and once with the same medium without heparin. After transfection, both cell types were then transferred to 100-cm culture dishes with 10 ml of culture medium. For each transfection, additional cells were transfected in triplicate using 10 µg of the pTKB plasmid for later -galactosidase assays. After 48 h of culture, cells were collected for CAT and -galactosidase assays.

In additional mixing experiments, increasing ratios of mutant to WT E1A 13S DNA were used in the transient transfections. The total amounts of DNA and ratios of mutant plasmid to E1A 13S WT were chosen to correspond to the linear range of transactivation in this system (11).

CAT Assays

CAT assays were performed as described by Gorman and coworkers (31). Cell extract protein was measured by the method of Bradford (32); equal amounts of protein were used for assays. The ¹⁴C-labeled chloramphenicol was separated into unacetylated and acetylated derivatives by ascending thin layer chromatography using chloroform/methanol (95:5). Chloramphenicol and its acetylated derivatives were quantitated by phosphorimaging (Molecular Dynamics, Sunnyvale, CA). Results were adjusted for -galactosidase activity in cells transfected with the pTKB plasmid to control for variations in transfection efficiency.

-Galactosidase Assays

-galactosidase assays were performed using a chemiluminescent assay (Tropix, Bedford, MA). The assay is sensitive to 2 femtogram of -galactosidase and linear over approximately five orders of magnitude.

Western Blot Analysis

Jurkat and THP-1 cells transfected with 10 µg of the control, mutant, or WT E1A plasmid DNA and 10 µg of either the CMV IE CAT or HIV-1 LTR reporter gene constructs were cultured 48 h and collected as described for CAT assays. Equal amounts of cell extract protein were fractionated on a 10% polyacrylamide gel and transferred onto a nitrocellulose membrane (MSI, Westborough, MA). The membrane was blocked with 5% non-fat dry milk overnight at 4°C, and the E1A protein was detected by chemiluminescence (NEN, Boston, MA) after reacting with anti-E1A 13S primary antibody (PharMingen, San Diego, CA) and secondary antibody, horseradish peroxidase-conjugated goat antimouse (PharMingen).

Statistical Analysis

Statistical significance was determined using the two-tailed t test with Bonferroni correction for multiple comparisons. Significance was considered as P < 0.05 (33).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of E1A WT and Mutant Plasmids in Jurkat Cells

To evaluate expression of the E1A mutant proteins in our cell system, we transfected both Jurkat and THP-1 cells with various E1A 13S CR3 mutant plasmids and the CMV IE and HIV-1 LTR constructs using electroporation (Jurkat cells) and DEAE dextran (THP-1 cells) followed by Western blot analysis. This analysis showed that Jurkat and THP-1 cells transfected with E1A and mutant plasmids accumulated equivalent levels of the E1A WT and mutant proteins (Figure 1B, Jurkat cells; Figure 1C, THP-1 cells). Immunoreactive protein was resolved into two major bands with apparent molecular masses of 47 and 45 kD, as expected (22). Equivalent accumulation of the E1A WT and CR3 mutant proteins in the presence of both the CMV IE and HIV-1 LTR reporter gene constructs makes it likely that any difference in transactivation of the CMV IE and HIV-1 LTR promoters in these cells can be attributed to a difference in activity of the proteins.

View larger version (20K): [in this window] [in a new window]   View larger version (52K): [in this window] [in a new window]   View larger version (52K): [in this window] [in a new window]   Figure 1.   (A) Schematic representation of CR3 of adenovirus E1A 13S 289R. The location of the E1A mutations expressed by the plasmids used is depicted together with the WT and mutant amino-acid residues. (B) Western blot analysis of adenovirus E1A CR3 mutant proteins in Jurkat cells or (C) THP-1 cells. Cells were transfected with 10 µg of either pBIISK (control), E1A WT, or E1A 13S plasmids containing mutations in the zinc finger (C171S), interdomain (T178S), carboxyl (S185N), or both the zinc finger and carboxyl regions (C171S/S185N), and 10 µg of either the pCAT760WT(+CMV CAT) or pHIV-1 (+HIV-1 CAT) plasmids. A total of 5 µg of cell extract protein per lane was fractionated by polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and probed with anti-E1A 13S antibody and horseradish peroxidase-conjugated secondary antibody.

Effect of CR3 Mutation of Adenovirus E1A 13S on Transactivation of the CMV IE Promoter

To evaluate the effect of mutation of specific regions of the adenovirus E1A 13S CR3 on transactivation of the CMV IE promoter, Jurkat cells and THP-1 cells were cotransfected with E1A CR3 mutant plasmids and the CMV IE promoter-CAT construct pCAT760WT plasmid.

In Jurkat cells, WT E1A resulted in a 13-fold increase in CMV IE promoter activity as assessed by CAT activity. Substitution of serine for cysteine in the E1A CR3 zinc finger subdomain resulted in a significant 96% loss of CMV IE transactivation relative to E1A WT (C171S; Figure 2A, P < 0.05); substitution of asparagine for serine in the carboxyl region resulted in a significant 96% loss of activity also (S185N; Figure 2A, P < 0.05). Importantly, substitution at both sites (C171S/S185N; Figure 2A) resulted in a similar significant loss of activity (99%, P < 0.05). Transactivation using this double mutant was similar to that seen with the single mutants (P = 0.18), suggesting that mutations in one subdomain did not unmask a negative regulatory site in the other.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Effect of CR3 mutations of adenovirus E1A 13S on transactivation of the CMV IE promoter. Jurkat and THP-1 cells were transfected with either the CMV IE CAT (solid bars) or HIV-1 LTR (open bars) plasmid (1 µg/ml) and 1 µg/ml of either the pBIISK (control), plasmids expressing E1A 13S WT, E1A 13S with a mutation in the zinc finger region (C171S), carboxyl region (S185N), both regions (C171S/S185N), or a mutation between these regions (T178S). CAT assays were performed as described. (A) Jurkat cells transfected with the mentioned plasmids using electroporation. CAT activity is expressed on the ordinate as percent of CAT activity in the presence of the plasmid expressing E1A 13S WT together with the CMV IE CAT plasmid. (B) THP-1 cells transfected with the mentioned plasmids using DEAE dextran. CAT activity is expressed on the ordinate as percent of CAT activity in the presence of the plasmid expressing E1A 13S WT together with the CMV IE CAT plasmid (mean ± standard error of the mean [SEM] of three separate studies for both cell types).

In contrast to the previous findings, substitution of serine for threonine in the interdomain region of CR3 resulted in only a 24% loss of activity (T178S; Figure 2A). Transactivation using this mutant was significantly different from that seen with all of the subdomain mutants (P < 0.05). These results suggest that transactivation of the CMV IE promoter by adenovirus E1A 13S involves specific subdomains in CR3.

In THP-1 cells, WT E1A resulted in an 18-fold increase in CMV IE promoter activity as assessed by CAT activity. Mutation of the zinc finger subdomain resulted in a 98% loss of transactivation (C171S; Figure 2B, P < 0.05). Mutation of the carboxyl region also resulted in a significant 99% reduction in activity (S185N; Figure 2B, P < 0.05), as did mutation of both regions combined (99%, C171S/ S185N; Figure 2B, P < 0.05.)

As in the Jurkat cells, transactivation of the CMV IE promoter was less affected (23%) by mutation in the interdomain region of CR3 (T178S; Figure 2B).

Transactivation by adenovirus E1A 13S appears to be viral-specific in unstimulated Jurkat and THP-1 cells as no transactivation of the HIV-1 LTR was seen by either WT or mutant adenovirus E1A 13S (Figures 2A and 2B).

Effect of Increasing Concentrations of CR3 Mutants on Transactivation of the CMV IE Promoter by E1A 13S WT

Increasing ratios of the interdomain mutant plasmid T178S to E1A 13S WT increased CMV transactivation (Figure 3A). This suggests that there is a functional transactivational element in the T178S plasmid.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Effect of CR3 zinc finger region mutations of adenovirus E1A 13S on transactivation of the CMV IE promoter. Jurkat cells were transfected using the electroporation method with the CMV IE CAT plasmid (1 µg/ml) and increasing ratios of the interdomain mutant T178S (A), zinc finger mutant C171S (B), or the carboxyl domain mutant S185N (C) to E1A 13S WT. The amount of WT and mutant plasmids used at the various ratios was 2.5 µg/1 µg (2.5:1), 5 µg/1 µg (5:1), and 10 µg/1 µg (10:1). CAT assays were performed as previously described. CAT activity is expressed on the ordinate as percent conversion of 14C- labeled chloramphenicol to its acetylated derivatives. The ratio of mutant to E1A WT plasmids used is expressed on the abscissa. Percent conversion for T178S/E1A 13S WT (A), C171S/E1A 13S WT (B), or S185N/E1A 13S WT mixtures (C) is plotted as the mean ± SEM of five, four, and four experiments, respectively.

In contrast, increasing ratios of the zinc finger mutant C171S to E1A WT plasmid did not appear to alter the absolute level of transactivation (Figure 3B). Therefore, we cannot exclude the possibility that the C171S mutant protein sequesters a transcription factor, albeit weakly. In any case, the absence of an increase in promoter activation confirms the previous results that this region is necessary for transactivation of the CMV IE promoter. Similarly, the effect of the carboxyl region mutant (S185N) confirmed the importance of the carboxyl subdomain in transactivation (Figure 3C).

Effect of Deletion of CMV IE Promoter Elements on Transactivation by E1A 13S WT

To determine the importance of specific elements of the CMV IE promoter on transactivation of the CMV IE promoter by adenovirus, THP-1 cells were cotransfected with E1A 13S WT or control plasmid and CMV IE promoter deletion CAT constructs. The results demonstrate that progressive deletion of the 19-bp CREB/ATF element resulted in a progressive loss of activity (Figure 4). Interestingly, loss of the 21-bp element appears to result in enhancement of activity also (pCAT-760WT, pCAT-dl36, and pCAT-dl14; Figure 4).

View larger version (22K): [in this window] [in a new window]   Figure 4.   Effect of deletion of CMV IE promoter elements on transactivation by E1A 13S WT. THP-1 cells were transfected using DEAE dextran with either the pBIISK (control; open bars) or E1A 13S WT plasmids (1 µg/ml; solid bars) and 1 µg/ml of either WT (pCAT-760WT) or 5' promoter deletions of the CMV IE promoter (pCAT-dl36, pCAT-dl14, pCAT-dlA231, pCAT-dl4, and pCAT-dl760). CAT assays were performed as described. CAT activity is expressed on the ordinate as percent of CAT activity in the presence of the plasmid expressing E1A 13S WT together with the WT CMV IE CAT plasmid. The number of 16-, 18-, 19-, and 21-bp repeats present in the plasmids used is represented on the abscissa. The data is expressed as the mean ± SEM of four experiments.

Transactivation of Single CMV IE Promoter Elements by E1A 13S WT

To confirm the importance of the 19-bp CREB/ATF repeat in transactivation of the CMV IE promoter, cotransfection was performed using the adenovirus E1A 13S WT expression vector or control plasmid and CMV single promoter elements upstream of CAT.

The results demonstrated that the 19-bp single promoter element containing the CREB/ATF binding site conferred a fourfold increase in CAT activity on the minimal promoter element pCAT-dl760 (Figure 5, P < 0.05). No negative effect of the 21-bp element was seen in this experiment, suggesting that this function of the 21-bp element can only occur in the context of the entire promoter.

View larger version (14K): [in this window] [in a new window]   Figure 5.   Transactivation of single CMV IE promoter elements by E1A 13S WT. THP-1 cells were transfected using DEAE dextran with either the pBIISK (control; solid bars) or E1A 13S WT plasmids (1 µg/ml; open bars) and 1 µg/ml of plasmids containing either the 16- (pIE1-161), 18- (pIE1-181), 19- (pIE1-191), or 21-bp (pIE1-211) CMV IE repeats linked to CAT. CAT assays were performed as described. CAT activity is expressed on the ordinate as fold increase in CAT activity over that seen in the presence of the control plasmid together with the minimal CMV IE promoter element (pCAT-dl760). Data are expressed as the mean ± SEM of four separate studies.

Effect of CR3 Mutation of Adenovirus E1A 13S on Transactivation of the CMV IE Promoter 19-bp CREB/ATF Repeat

To determine if transactivation of the critical CMV IE promoter 19-bp element occurred through similar subdomains of adenovirus E1A 13S CR3 as for transactivation of the entire promoter, THP-1 cells were cotransfected with the CMV IE 19-bp single promoter element CAT construct and the WT or CR3 mutant adenovirus E1A 13S expression vectors. WT E1A enhanced CAT activity from this construct by fourfold as compared with control vector. Mutation of the zinc finger (C171S) or carboxyl region (S185N) of CR3 or both regions resulted in a significant loss of activity (83, 94, and 85%, respectively; Figure 6, P < 0.05). In contrast mutation of the interdomain region (T178S) resulted, if anything, in an enhancement of transactivation (+28%). These results suggest that transactivation of this promoter element requires the zinc finger and carboxyl element of adenovirus E1A CR3.

View larger version (11K): [in this window] [in a new window]   Figure 6.   Effect of CR3 mutation of adenovirus E1A 13S on transactivation of the CMV IE promoter 19-bp CREB/ATF repeat. THP-1 cells were transfected with 1 µg/ml of either the pBIISK (control), plasmids expressing E1A 13S WT, E1A 13S with a mutation in the zinc finger region (C171S), carboxyl region (S185N), both regions (C171S/S185N), or a mutation between these regions (T178S), and 1 µg/ml of the plasmid containing the 19-bp CREB/ATF binding site upstream of the minimal CMV IE promoter element linked to CAT. CAT assays were performed as described. CAT activity is expressed on the ordinate as percent of CAT activity in the presence of the plasmid expressing E1A 13S WT together with the CMV IE 19-bp CREB/ATF CAT plasmid. Data are expressed as the mean ± SEM of three separate studies.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results confirm and extend previous work showing that the adenovirus E1A 13S gene product transactivates the CMV IE gene in inflammatory cells (11). The findings also indicate that the mechanism involves specific subdomains in CR3. Localization of the sites of E1A 13S 289R responsible for transactivation is important in helping to determine which of several possible mechanisms of transactivation is likely to be important in the particular activity presented herein. In particular, it is important to know whether CMV transactivation involves interaction of one of these subdomains with limiting transcription factors as proposed for transactivation of downstream adenovirus promoter elements by E1A (34).

With regards to transcriptional activation by the various adenovirus E1A proteins, there are three conserved regions in the 12S 246R and 13S 289R proteins (35, 36). CR1 and CR2 are present in both proteins, whereas CR3 is present only in the 289R protein. We showed that transactivation of the CMV IE promoter does not occur with the CR1 and CR2 containing 12S gene product alone (11) but instead requires the presence of CR3 contained exclusively in the 13S gene product. Our findings presented in this present report confirms the importance of this region as we saw a loss of transactivation with mutation in either of the two putative subdomains of CR3.

CR3-dependent transactivation of adenovirus E1A 13S appears to involve the binding of CR3 subdomains to general and specific transcription factors. Specifically, the CR3 zinc finger domain (aa 147-177) appears to interact with the general transcription factor TFIID by binding to TATA-box binding protein (TBP) and TBP-associated factors, whereas the carboxyl region (aa 180-188) appears to bind to the cellular or specific transcription factors (37- 39). Thus, E1A may act as a bridge between cellular transcription factors and the basal transcriptional complex. This may be how E1A plays a role in altering transcription rates in cooperation with specific DNA binding transcription factors. Mixing experiments using the adenovirus E1A E3 promoter suggested that primarily carboxyl region mutants inhibited transactivation by E1A WT protein presumably by sequestration of less abundant general transcription factors by the zinc finger subdomain. If the mechanism of transactivation of CMV IE is identical to adenovirus E3, inhibition of transactivation of CMV IE in the mixing studies would have occurred primarily by the carboxyl region mutant S185N due to sequestration of a limiting general transcription factor. In fact, if any transdominant suppression occurred, it was in the presence of the zinc finger mutant C171S, which contains an intact carboxyl region (Figure 3B). It is possible that our results differ from those with E3 transactivation because of a difference of the relative amounts of general and cellular transcription factors in Jurkat cells versus HeLa cells used in the E3 studies. It is more likely that the specific cellular transcription factor required for CMV IE transactivation may be the limiting factor in Jurkat cells. Alternately, if the mechanism of transactivation of CMV IE is similar to that of adenovirus E2A and involves cooperation of non-CR3 regions with CR3, transdominance may be obscured.

Phosphorylation may also be important in CR3-dependent transcriptional regulation as it is in CR3-independent regulation. Phosphorylation of specific residues of E1A 13S CR3 is catalyzed by mitogen-activated protein kinase (MAPK). This kinase was shown to directly phosphorylate E1A 13S at the serine 185 in the carboxyl subdomain of CR3. The functional result of phosphorylation was to augment CR3-mediated transactivation of the adenovirus E4, but not adenovirus E3, promoter (40). Presumably, augmentation is due to enhanced binding of phosphorylated 289R to cellular transcription factors. The MAPK kinase/ Elk-1 pathway may also be involved in activation of the CMV IE promoter (34). Although the protein kinase C- activated pathway involving NF-B was a potential candidate as it is activated by E1A and may also be important in CMV IE promoter activation by other stimuli (41, 42), we did not see involvement of the NF-B containing the 18-bp repeat in transactivation by E1A. On the other hand, both pathways can be activated by phorbol myristate acetate and lipopolysaccharide, and we have previously demonstrated that treatment of Jurkat and THP-1 cells with these agents enhances transactivation of the CMV IE gene by adenovirus E1A 13S (11). We are currently investigating which of the potential signaling pathways are involved in CR3 transactivation of the CMV major IE promoter.

With regards to the target of E1A transactivation in the present study, the CMV IE promoter, it is known that this promoter contains several positive and negative regulatory elements. The positive regulatory elements include binding sites for transcription factors NF-B, CREB/ATF, and AP-1 (43). Cotransfections in HeLa cells suggest that the CREB/ATF site is of major importance in transactivation of the CMV IE promoter by adenovirus E1A (24). The current study confirms this initial observation as the 19-bp CREB/ATF site appeared to be important for transactivation of the CMV IE promoter in the monocyte/macrophage-like THP-1 cells (Figure 5). Furthermore, transactivation of this element appeared to involve both the zinc finger and carboxyl regions of adenovirus E1A 13S CR3 as mutation of these regions resulted in a loss of this activity (Figure 6). Although this likely occurs through modulation of the formation of a transcriptional complex binding to this site, we were unable to detect an E1A-specific mobility shift using labeled oligonucleotide complementary to the CREB/ATF consensus sequence (data not shown). This suggests that E1A does not cause transactivation in this case by directly interacting with DNA. E1A can act through NF-B (41, 44), but as mentioned, the studies described here did not show that this element contained in the 18-bp repeat was involved. AP-1 family factors can also play a role in E1A transactivation, but it is unlikely these factors are important in transactivation of CMV IE, seen here as the pCAT-dl36, pCAT-dl14, and pCAT-dlA231 constructs, which lack the CMV IE AP-1 site at 232 yet retain significant activity (Figure 4). The CMV IE promoter also contains negative regulatory elements that also may be important for regulation of activation of this promoter. Many of these negative regulatory elements contain binding sites for the transcription factor YY1 (47). This factor is an important repressor of CMV IE reporter activity in several cell types. In transient transfections in HeLa cells, adenovirus E1A 13S has been shown to overcome repression due to YY1 (48, 49). The current studies using CMV promoter deletions suggest that the YY1 sites contained in the 21-bp repeat may act as a negative regulatory element in the context of the entire promoter (Figure 4). The lack of a negative effect for this element in the single promoter element repeat studies also suggests that this element requires interaction with other positive regulatory elements to act as a repressor as has been demonstrated in activation of c-fos (50). In any case, based on the studies presented, CMV IE promoter transactivation by E1A 13S is likely due to interaction with, or effects on, positively and negatively acting transcription factors, one of which is likely CREB/ATF.

Reactivation of CMV by adenovirus in vivo would be hard to demonstrate; however, codetection of adenovirus and CMV has been demonstrated simultaneously in the lung in patients with AIDS and transplant recipients (9, 10). In addition, significant disease due to adenovirus and CMV coinfection of the same organ has been demonstrated microscopically. When AIDS patients with suspected viral colitis were biopsied, most were found to have CMV infection by immunohistochemistry; however, 12% of those patients had adenovirus infection. Most patients with adenovirus infection also had coinfection with CMV, but only 10% of those with CMV colitis had coinfection with adenovirus, thus suggesting that adenovirus may play a role in stimulating CMV infection (51).

Our results confirm that transactivation of the CMV major IE promoter by adenovirus involves CR3 of the E1A 289R protein in inflammatory cells in vitro. We also confirm that as in transactivation of downstream promoter elements of adenovirus by E1A 13S, transactivation of the CMV IE promoter involves the zinc finger and carboxyl subdomains of CR3 acting, at least in part, through the 19-bp CREB/ATF consensus sequence. Our data provide new insights into the mechanisms whereby reactivation of CMV infection may occur in susceptible patients by adenovirus.