Abstract
Polyomavirus JC (JCV) is the aetiological agent of progressive multifocal leukoencephalopathy (PML), a frequently fatal infection of the brain afflicting nearly 4 % of AIDS patients in the USA. Human immunodeficiency virus type 1 (HIV-1) Tat, acting together with cellular proteins at the JCV non-coding control region (NCCR), can stimulate JCV DNA transcription and replication. Tat in the brain is secreted by HIV-1-infected cells and incorporated by oligodendroglia, cells capable of infection by JCV. Thus far the effects of Tat on JCV have been studied primarily with protein encoded by the HIV-1 B clade most common in North America. Here, we determine the abilities of Tat from different HIV-1 clades to alter JCV early and late gene transcription and DNA replication initiated at the JCV origin. Tat from all clades tested stimulates both JCV early and late gene promoters, with clade B Tat being significantly most effective. Tat proteins from the HIV-1 clades display parallel patterns of differences in their effects on HIV-1 and JCV transcription, suggesting that Tat effects in both cases are mediated by the same cellular proteins. Clade B Tat is most effective at directing Smad mediators of tumour growth factor beta and cellular partner Purα to the NCCR. Tat proteins from all non-B clades inhibit initiation of JCV DNA replication. The effectiveness of HIV-1 clade B Tat at promoting JCV transcriptional and replicative processes highlights a need for further investigation to determine which molecular aspects of Tat from distinct HIV-1 substrains can contribute to the course of PML development in neuroAIDS.
Introduction
Nearly 50 % of individuals with AIDS in the USA experience neurological complications (Clifford, 1999; Price, 1996), many caused directly by human immunodeficiency virus type 1 (HIV-1). AIDS patients also suffer opportunistic infections of the central nervous system (CNS) caused by a variety of viral agents, including human polyomavirus JC (JCV). Mortality due to HIV-associated neurological disease has been reduced by combination treatments, highly active antiretroviral therapy (HAART) and combination antiretroviral therapy (CART). Although none of the antiretroviral therapies is effective against DNA viruses such as JCV, JCV activation in the brain is due in large part to immunosuppression, which is ameliorated by CART. HIV-1 is, however, associated with significant neurological deficits even with treatment (Heaton et al., 2010, 2011).
JCV is the aetiological agent of progressive multifocal leukoencephalopathy (PML), a brain-demyelinating disease afflicting nearly 4 % of people with AIDS in the USA and Western Europe (Berger & Major, 1999; Frisque et al., 1984; Khalili et al., 2006; Major et al., 1992; Padgett et al., 1971; Cinque et al., 2003). There is little information on neuroAIDS, including PML, in many areas of the world. Although there are several potential reasons for this, such as under-reporting, the possibility that the distinct clades of HIV-1 affect JCV differently has yet to be incisively addressed (Robertson et al., 2010; Robertson & Hall, 2007; Shankar et al., 2003). HIV-1 clade B is prevalent in the USA and Europe, but clade C is currently the most prominent HIV-1 subtype worldwide (Liner et al., 2007). A recent review suggests that clade C may have reduced neuroinvasive capacity due to the different primary structure of its Tat protein (Rotta & Almeida, 2011). Other HIV-1 subtypes may also differ in pathology or incidence (Liner et al., 2007). The present study aims, through molecular means, to address an important aspect of the problem of HIV interaction with JCV: the ability of Tat of different clades to activate JCV DNA transcription and replication.
Tat, secreted from HIV-1-infected cells (Ensoli et al., 1993), including CNS microglial cells, astrocytes and monocytic blood cells, can be incorporated by other accessible cells (Ensoli et al., 1993; Frankel & Pabo, 1988). Tat is avidly incorporated by oligodendrocytes (Daniel et al., 2004), where it can exert positive control over both transcription (Krachmarov et al., 1996) and replication (Daniel et al., 2001) of JCV DNA. Tat does not bind directly to JCV DNA, but it interacts with cellular proteins and protein complexes that bind to sequences in the JCV non-coding control region (NCCR) (Chang et al., 1994, 1996; Chen et al., 1995; Daniel et al., 2001, 2004; Gallia et al., 1998, 1999). The NCCR contains the origin of JCV DNA replication (ori) as well as promoter sequences for both JCV early (E) and late (L) gene transcription. Tat interacts with known cellular transcription factors, including Purα (Bergemann et al., 1992; Gallia et al., 2000; Johnson, 2003) and Smad proteins (Sawaya et al., 1998; Stettner et al., 2009) that have demonstrated effects on the JCV-NCCR. Tat is the primary viral facilitator of HIV-1 gene transcription (Chepenik et al., 1998; Jones, 1997; Karn, 1991). Two partner proteins that have been particularly well characterized with regard to gene transcription and interaction with resulting transcripts are Cyclin T1/Cdk9 (Garber et al., 1998; Mancebo et al., 1997) and Purα (Chepenik et al., 1998; Gallia et al., 1999; Johnson, 2003; Johnson et al., 2006; Krachmarov et al., 1996; White et al., 2009). These cellular proteins both bind to a critical Tat cysteine-rich structural domain, involved in Zn coordination (Campbell et al., 2011), which plays an essential role in RNA binding (Tahirov et al., 2010). Sequences of Tat proteins from HIV-1 strains of clade C differ in this C-rich domain.
In the present study, we found that Tat from HIV-1 clade B has a significantly more stimulatory effect on transcription of both JCV early and late gene transcription than do Tat proteins from each of the other tested clades. The ability of Tat from different HIV-1 clades to influence initiation of JCV DNA replication correlates well with Tat presence on the JCV origin of replication in live oligodendroglial cells, as measured by chromatin immunoprecipitation (ChIP). Results of this study help to provide a molecular basis for comparison of the activation of JCV in PML by the different substrains of HIV-1 affecting people worldwide.
Results
Tat proteins from distinct HIV-1 clades have differential effects on transactivation of the 5′- long-terminal repeat (LTR) promoter
Fig. 1 shows an alignment of the Tat amino acid sequence encoded by the first exon from seven different variants of four different HIV-1 clades, all functionally analysed for this study. Several structural features are notable. First, most amino acid mutations in the first 68 aa occur from aa 19–39, in the region classified as C-rich, Zn-binding. These Cs in clade B Tat are essential for coordinating two Zn atoms (Frankel et al., 1988; Tahirov et al., 2010), and most of the Cs are essential for Tat functionality. For example, mutation of C22 abrogates Tat transactivation of HIV-1 (Kashanchi et al., 1996; Ruben et al., 1989; Wortman et al., 2000), Tat binding to cellular partner Purα and resulting Tat activation of JCV DNA replication (Daniel et al., 2001). Notably, all tested strains of clade C possess S31 instead of C31. Yet clade C Tat can clearly transactivate HIV-1. Cm Tat is a naturally occurring variant that has a C27 to Y27 mutation. In the present study, we compare the efficacies of clade C Tat strains to those of Tat from other clades in several functional tests. The clade E variants are also structurally interesting. E-2 Tat contains a 2 aa insert in the transporter sequence responsible for Tat intercellular mobility. E-9 Tat72 contains a 17 aa C-terminal sequence apparently transduced from human dual-use phosphatase 10. In transactivation experiments using Tat variants from clades B and C, no differences were noted between Tat72 from a given clade, representing exon 1, and longer Tat variants with sequences from exon 2 (e.g. Tat86 or Tat101). Therefore, this report presents comparisons of Tat72 and counterparts from each HIV-1 clade.
Amino acid coding sequences for exon 1 of Tat from HIV-1 clades examined in this study. The Cm mutation of clade C Tat has a base change in the transactivation C-rich domain, resulting in a C27 to Y27 transition. Domains for HIV-1 TAR-dependent transactivation, C-rich, Zn-binding and transporter activities are indicated at the bottom.
Oligodendroglial cell line KG-1C was co-transfected with plasmids expressing Tat from each different clade, with luciferase-expressing pLTR-luc, and with control plasmid, Renilla. Fig. 2 shows the effect of each clade construct on the transactivation of the HIV-1 5′-LTR promoter. Empty vector, pcDNA3.1 is a control, as is pTat, a plasmid expressing Tat, clade B, from its LTR. A parallel control at the bottom of Fig. 2 shows that the pcDNA3.1 vector is equally effective at expressing each of the different clade Tat proteins. A striking finding in Fig. 2 is that proteins generated by both clade B constructs, pTat and B, are significantly more effective at activating HIV-1 transcription than are Tat proteins from all other tested clades. All clade Tat proteins in Fig. 2 are compared with Tat of lane B for P-values because these are all under control of the same CMV MIE promoter. Tat E-9, containing an unusual transduced sequence, is next most effective after Tat B.
Effects of Tat from different HIV-1 clades on transcription controlled by the HIV-1 5′ LTR promoter. KG-1C cells were transfected and cell lysates harvested at 48 h for luciferase assay as detailed in Methods. Plasmid pTat, under the HIV-1 5′-LTR promoter, was used as a positive control. A no-vector mock transfection was used as a negative control and pcDNA3.1 as a negative empty vector control. The control immunoblot at bottom shows HA-Tat protein generated under the CMV MIE promoter. **P<0.01, ***P<0.001 vs B Tat.
Tat Cm (clade C Tat, C27 to Y27 mutation) is an effective inhibitor of Tat transactivation of HIV-1 transcription by unmutated clade C Tat
Clade C Tat is an effective transactivator of clade B LTR-trans-activation response element (TAR)-driven HIV-1 transcription, whereas Cm is not (Fig. 2). We sought to determine whether Cm can inhibit the ability of clade C Tat to transactivate, and the results are presented in Fig. 3. Cm is a very effective inhibitor. When plasmid expressing Cm is transfected together with plasmid expressing C, approximately 50 % inhibition is seen with Cm plasmid at 50 % the amount of C plasmid. It has previously been reported that C27 is essential for clade B Tat transactivation (Ruben et al., 1989), but to obtain this level of inhibition of clade C Tat activity by adding a 72 aa polypeptide with a single aa change is remarkable.
Inhibition of clade C Tat-mediated HIV-1 5′ LTR transcription by clade C Tat mutant, Tat Cm. The plasmid LTR-luc was transfected into KG-1C cells with clade C Tat plasmid and the indicated amounts of Cm plasmid. Tat Cm is clade C Tat72, C27 to Y27. The LTR-luc plasmid contains the clade B LTR with the 5′-UTR, including the TAR element. **P<0.01, ***P<0.001 vs C Tat.
The Tat proteins of different HIV-1 clades vary in effects on transactivation of both JCV-E and -L promoters
PML occurs in the brains of people infected with different HIV-1 clades. For example, PML lesions have been visualized in cases of clade C brain infection (Shankar et al., 2003). The question thus arises: do Tat proteins from the different HIV-1 clades influence JCV gene transcription differently? Co-transfection experiments were done as described using all Tat clades and the JCV early (E) and late (L) promoters individually. Fig. 4(a) shows the results from luciferase assays using the JCV-E promoter. Our data indicate that the L promoter has both higher basal transactivation and more significant stimulation than does the E promoter in response to Tat of all clades, as seen in Fig. 4(b). While it is likely that the E and L promoters respond differently to different concentrations of Tat, it is likely that there are differences in how the Tat clades stimulate both the early and late promoters of JCV.
Effects of Tat from different HIV-1 clades on transcription controlled by the JCV early (JCV-E) and late (JCV-L) promoters. The luciferase assay was done in triplicate as described in Methods and the legend to Fig. 2. KG-1C cells were transfected for 48 h with the JCV-E (a) or L (b) plasmid together with plasmids expressing individual Tat proteins. As in Fig. 2, controls for Tat protein expression showed no significant difference among the clades. *P<0.05, **P<0.01, ***P<0.001 vs B Tat.
Differences in KG-1C cell Smad4 mRNA levels in response to Tat from different HIV-1 clades
It is not known whether Tat effects on JCV E- and L-gene promoters are indirect, involving Tat effects on other cellular processes. Previous publications indicate that the DNA/RNA-binding protein Purα is able to co-localize with Tat and bind DNA of the JCV-NCCR in KG-1C cells (Daniel & Johnson, 1989; Daniel et al., 2004). It has been reported that Tat presence in cells stimulates release of the tumour growth factor beta 1 (TGF-β1) cytokine that triggers the cascading Smad gene regulatory pathway (Sawaya et al., 1998) as well as triggering other TGF-β-related pathways (Rahimi & Leof, 2007). It has previously been reported that Smads -2, -3 and -4 stimulate JCV gene transcription (Stettner et al., 2009). We thus sought to determine whether Tat proteins from different clades induce different levels of expression of the TGF-β-mediating Smad4 transcription factor. We transfected KG-1C glial cells to express Tat from each clade and subjected the cells to mRNA analyses. RNA was analysed for purity prior to being subjected to PCR as described in Methods. Fig. 5(a) presents the results of RT-PCR using primers specific for the Smad4 transcription factor. It is not yet known whether the observed effects of Tat on Smad4 mRNA levels are due to effects on gene regulation or on post-transcriptional events. Note that in Fig. 5(b) there are no changes in expression of the control β-actin gene. We conclude that there are differences between the Tat clades in effects on the transcriptional expression of the Smad4 transcription factor in oligodendrocytes. Notably, Tat from clade B is again most effective.
Smad4 mRNA levels in the presence of Tat from different HIV-1 clades. KG-1C cells, plated in triplicate, were lysed 48 h after transfection with Tat-expressing plasmids and total RNA purified. One-step RT-PCR was performed using primers specific for Smad4. Primers for β-actin were used as a positive loading control. PCR cDNA products were subjected to electrophoresis and resulting stained bands were quantified using Photoshop 5.5. Quantification of the Smad4 mRNA is presented in Fig. 5(a), and stained PCR products for Smad4 and β-actin in Fig. 5(b).
Tat proteins from different HIV-1 clades cause Purα and Smad4 to bind the JCV-NCCR to varying degrees as determined by ChIP
Because Tat and the cellular protein, Purα, interact (Chepenik et al., 1998; Daniel et al., 2001; Gallia et al., 1999; Krachmarov et al., 1996) to stimulate JCV DNA replication (Daniel et al., 2001) and to synergistically enhance late-gene transcription (Krachmarov et al., 1996), we sought to determine whether the levels of Purα bound to JCV-NCCR DNA vary based on the presence of Tat from the different HIV-1 clades. KG-1C glial cells were co-transfected with Tat clades along with the JCV-NCCR plasmid, pMad1, which expresses a 260 bp segment of the JCV-NCCR. It should be noted that this segment contains overlapping elements for E- and L-gene transcription as well as for initiation of DNA replication. Cells were incubated 72 h before cross-linking proteins to DNA, and ChIP was performed as described in Methods. Quantification was done using primers specific for the JCV-NCCR with normalization to a standard curve generated from real-time PCR using pMad1 DNA. Results in Fig. 6(a) indicate that Tat from each clade has a different effect on levels of Purα binding to the JCV-NCCR. Consistent with all Tat effects examined thus far, pTat and B are most effective at enhancing levels of Purα bound.
ChIP analysis of Purα (a) and Smad4 (b) binding to the JCV-NCCR. KG-1C cells were co-transfected with pMad1, containing the NCCR, and plasmids expressing Tat from the different clades. After ChIP, a 260 bp segment of the NCCR containing the JCV early and late-gene promoters and the origin of DNA replication was amplified. Data were analysed using GraphPad Prism software and normalized to the empty pcDNA3.1 vector. **P 0.01, ***P<0.001 vs B Tat.
Because Tat interacts with Smad4 to influence JCV gene transcription (Stettner et al., 2009), the ability of Tat from different clades to alter Smad4 binding to the JCV-NCCR was also examined, as shown in Fig. 6(b). Smad4 binding is affected differently from that of Purα in that Tat of clade D is more effective than Tat of E-2 or E-9 at enhancing Smad4 binding. In other respects, however, the Smad4 pattern of results mimics that for Purα. B and pTat are most effective at enhancing both Purα and Smad4 binding.
The results of Fig. 6(a) reveal a new aspect of the functional interaction of Tat with Purα. It has previously been published that Tat and Purα together stimulate JCV early and late-gene transcription (Krachmarov et al., 1996). Because Tat does not itself bind DNA, and Purα does, a likely explanation would be that Purα recruits Tat to the JCV control region. The present results indicate a more reciprocal functional interaction of the two proteins. Fig. 6(a) shows that Tat proteins from different HIV-1 clades act differently to stimulate Purα binding to the promoter region. Thus, Tat is not just being recruited but it is changing the ability of Purα to bind to the JCV DNA.
Tat proteins from different HIV-1 clades vary in effects on Smad4 localization to the nucleus
KG-1C glial cells were co-transfected with plasmids expressing Tat and Smad4 in order to visualize changes in Smad4 localization within the cells. Results after 48 h are visualized in Fig. 7. Alexa Fluor secondary antibodies were used to stain Tat (red) and Smad4 (green), and TO-PRO-3 was used to visualize nuclei (blue). Immunofluorescent co-localization of Tat and Smad4 (yellow) is visualized throughout both the cytoplasm and nucleus with Tat of clades C and E. In contrast, co-localization of Tat and Smad4 using Tat from clades B and D is primarily limited to highly specific nuclear structures (Fig. 7). Results with Tat of clades C and E-2 show that Smad4 is distributed approximately evenly in the nucleus and cytoplasm. In the presence of pTat and B, however, Smad4 is localized nearly exclusively in the nucleus. This is visualized clearly in the Fig. 7 inset displaying effects of transfection with pTat on two similar cells. Compare this inset with the No Tat panels directly above showing primarily cytoplasmic distribution of Smad4 in the absence of Tat. Although in the pTat inset Smad4 (green) has clearly been directed to the nucleus, there is only moderate co-localization with Tat (red). Whereas Tat is present in the cytoplasm and in well-defined round nuclear bodies, Smad4 is located in a dense, aggregated chromatin structure. In the majority of cells transfected with pTat or B, Smad4 is highly directed to nuclear localization although there is minimal co-localization with Tat. In contrast, with Tat of clade D Smad4 is co-localized with Tat throughout the nucleus, and cytoplasmic Tat is minimal. These results indicate a clear difference in the abilities of Tat from different clades to direct an important transcription factor to the nucleus. Clades B and D Tat proteins are distinct in their ability to localize Smad4 to nuclei, and Tat of clade B is unique in its ability to direct Smad4 to nuclear structures without actually co-localizing with it.
Different effects of HIV-1 Tat clades on Smad4 localization to the nucleus. Confocal imaging was performed 48 h after co-transfection of KG-1C cells with Tat clades and Smad4. Antibodies identified Tat (red) and Smad4 (green). Co-localization of Tat and Smad4 is visualized as yellow. TO-PRO-3 iodide was used for nuclear staining (blue). Negative controls using empty vector, no Tat plasmid and no primary antibodies are shown to the right. The blue channel is only presented for control micrographs. In the pTat inset arrows indicate the localization of both Tat and Smad4.
Effects of Tat from different HIV-1 clades on replication initiated at the JCV origin of DNA replication in KG-1C oligodendroglial cells
Tat interacts with Purα to stimulate replication initiated by large T-antigen at the JCV origin (ori). Replication is stimulated by Tat when both proteins are moderately exogenously expressed (Daniel et al., 2001). High levels of Purα in the absence of Tat inhibit initiation (Chang et al., 1996). It has previously been observed that Tat of clade B has little effect on the JCV ori in the absence of added Purα in KG-1C cells. These cells have low levels of endogenous Purα, similar to an embryonic phenotype in which levels of another Pur family member, Purγ, are relatively high (Khalili et al., 2003). To determine the effects of different clade Tat proteins on replication initiated by T-antigen, KG-1C cells were transfected with pJCV-T, pMad1, containing the JCV ori, and plasmids expressing Tat of different clades (Fig. 8). The DpnI assay for replication was employed, as shown in Fig. 8(a). Briefly, DpnI only cleaves when adenosine is methylated on both DNA strands of its cleavage site, as in a plasmid grown in bacteria. Replication in mammalian cells renders the pMad1 plasmid resistant to DpnI. As is consistent with the previous results, clade B Tat, expressed from either pTat or B, had little effect on replication initiated by T-antigen, as shown in Fig. 8(a). A control in Fig. 8(b) indicates that Tat effects on replication are not due to any effect of the different Tat proteins on T-antigen expression. Intriguingly, Tat from each of the other clades in Fig. 8(a) inhibited replication to varying degrees. Essentially no replication was seen using Tat of clade D, a striking departure from clade B Tat. Because Tat does not itself bind to DNA, Tat from other clades interacts with cellular partner protein(s) differently than does Tat from clade B. This result could have implications concerning abilities of different HIV-1 clades to activate JCV in oligodendroglia.
Effects of Tat from different HIV-1 clades on DNA replication initiated at the JCV origin (ori). (a) Replication was assayed as described in Methods. KG-1C cells were co-transfected with plasmids pMad1 (4.3 kb), containing a 260 bp segment with the JCV ori, and pJCT, expressing JCV large T-antigen (T-Ag). Intensities of the DpnI-resistant HindIII 4.3 kb bands, representing fully replicated plasmid DNA, were analysed using Imagequant 1.2. (b) Cell lysate proteins were assayed by immunoblot to determine the levels of large T-antigen expressed for each Tat clade. β-Actin was used as an internal and loading control. Intensities of the T-Ag bands were also quantified. Intensities for the band position of each lane, relative to the lane, +T-Ag (no Tat), assigned a value of 100, are as follows: -T-Ag, 30.6; -JCV ori, 98.5; pTat, 98.5; B, 98.0; C, 99.0; E2, 96.9; E9, 99.5; Czeo, 98.5; D, 98.0; Cm, 100.
Discussion
The HIV-1 protein, Tat, is a major contributor to neuronal and glial pathology. Tat is secreted by HIV-1-infected cells (Ensoli et al., 1993), can be taken up by neighbouring cells (Frankel & Pabo, 1988), can incite neuropathology (Bachis et al., 2009; Gurwell et al., 2001) and is avidly incorporated by oligodendroglial cells (Daniel et al., 2004), which myelinate and insulate neuronal axons. It is well established that Tat contributes to activation of JCV (Gallia et al., 2000; Johnson, 2003; Krachmarov et al., 1996; Tada et al., 1990). It is less well established whether different strains, or clades, of HIV-1 can affect JCV differently. In this regard, it is notable that Tat proteins representing clade B, i.e. pTat and B, are more effective than Tat of any other clade tested at stimulating processes involved in activation of JCV in the CNS. These results may be considered in light of the fact that different HIV-1 clades are prevalent in different geographical areas of the world. The different Tat proteins display the same variation among clades in activation of transcription of three genes of completely distinct character. It should be noted that HIV-1 and JCV infect different cells types. HIV-1 infects primarily microglial cells of the brain, and JCV infects primarily oligodendrocytes. It would be informative to examine the effect of Tat on the HIV-1 LTR in microglial cells. Nonetheless, because Tat is active in both types of cells, and because JCV is responsive to Tat in oligodendrocytes, the KG-1C oligodendroglial cell line is used in the present study to determine effects of Tat proteins on all three genes belonging to the two different viruses. Those genes are controlled by the HIV-1 LTR promoter and TAR element (Fig. 2) and by the JCV early and late gene promoters (Fig. 4). In each of these three cases Tat from clade B is significantly more stimulative than Tat from any of the other clades. Also in each case clade E-9 Tat is second best, with Tat from clades E-2, C and D being approximately equal. In the interest of brevity and clarity, we used the clade B LTR and TAR element for all transactivation assays. The relative luciferase levels among the different Tat proteins could be compared for each gene. Note that pTat in these experiments is Tat from clade B. Because it activates its own transcription, it produces more Tat protein than the other Tat clones, which are all under control of the hCMV MIE promoter. There is, however, no statistical significance between pTat and Tat clade B in effects on gene transcription. There is also no difference in the amounts of Tat mRNA or the amounts of Tat protein generated by the plasmids utilizing the CMV MIE promoter (Fig. 4). Therefore, the differences observed between Tat proteins of different clades on transcription of each distinct gene are due solely to amino acid sequences of the Tat proteins.
Two tests were performed to assay Tat effects on binding of demonstrated mediators of HIV-1 pathology, Purα and Smad4 (Krachmarov et al., 1996; Stettner et al., 2009) to the JCV-NCCR. The results obtained with Tat proteins from the different clades (Figs 5, 6) mimic closely, although not exactly, those obtained in our assays of Tat transactivation of HIV-1 or JCV. There is, however, one outstanding similarity: clade B Tat is far more effective than Tat of other clades at stimulating binding of both proteins to the JCV-NCCR chromatin. Because there are no perfect Smad consensus elements in the JCV-NCCR, the general similarity of the two panels of Fig. 6 may reflect Tat interaction with Purα, which has several binding elements in the control region DNA and which can also bind Smad4. Further experiments would be necessary to affirm this. Another aspect of Fig. 6 is important. That is, Tat from all the clade C clones is worse at stimulating binding of either Purα or Smad4 to the control region than Tat of any other clade. This indicates a functional difference in the amino acid sequence of clade C Tat from those of all the other clades. One clear sequence difference is the lack of C31 in Tat C. It remains to be determined whether C31 is critical for Tat binding to Purα or to the JCV-NCCR.
The effects of Tat on replication initiated at the JCV DNA origin are complex, but they again reveal the great differences in functional abilities of Tat from the different clades. In the absence of transfected, expressed Purα, clade B Tat has little effect on replication initiated by large T-antigen (Daniel et al., 2001). This is confirmed in Fig. 8, in which clade B Tat and T-antigen alone are approximately equal in effect on initiation. Tat proteins from all of the other clades are inhibitory to T-antigen. This is conceivably due to the abilities of different Tat proteins to bind a cellular factor that cooperates with T-antigen. Purα could be one such factor, but initiation involves several cellular proteins that could also play a role. Previous studies have shown that endogenous clade B Tat can enhance initiation of JCV replication at molar concentrations below 10−12 M, whereas effects of Tat are reduced at higher concentrations (Daniel et al., 2004). There is no reliable means to determine the molar concentrations of Tat in the transfected cells at sites of JCV DNA replication activity. It is therefore possible that inhibitory effects of Tat from certain clades could be due to excessive concentrations of Tat. This remains to be further explored. Note, however, that Tat from the different clades is expressed in the transfected KG-1C cells at similar levels, as exemplified in Fig. 2. Therefore, different effects of the Tat proteins on replication in Fig. 8 are most probably due to differences in the protein structures.
Tat proteins of different clades vary considerably in their ability to influence nuclear localization of Smad4 in oligodendroglia. For example, Tat from clades B and D stimulate an intense localization of Smad4 to nuclear structures, whereas Tat from clades C and E-2 promote a more diffuse co-localization with Smad4 throughout the cell (Fig. 7). The interaction of Tat and Smad4 with the JCV-NCCR is complex because neither protein can bind directly to the NCCR, but both can bind to mutual partner, Purα, which binds many NCCR DNA sites (Fig. 6, Stettner et al., 2009). Whereas Purα alone suppresses initiation of JCV DNA replication (Chang et al., 1996), the clade B Tat–Purα complex at appropriate concentrations stimulates initiation (Daniel et al., 2001, 2004). It is not known what effect Smad4 would have on JCV replication in a potential complex of the three proteins. It is conceivable that Smad4 plays a role in the differing effects of the various Tat proteins on initiation of JCV replication seen in Fig. 8.
Conclusions drawn from these results may be useful in assessing the course of PML in the context of people infected by different substrains of HIV-1 worldwide. In every test performed there are highly significant differences in functional efficacy among the Tat proteins from different HIV-1 clades. Results indicate that Tat of clade B is more effective than Tat of other clades at promoting transcription and initiating replication of JCV in oligodendroglia. An important corollary of these results is that, although they do so to different degrees, Tat from each of the HIV-1 clades is capable of stimulating JCV gene expression. The present focus on Tat does not presume that other HIV-1 proteins, which also undergo structural changes among different clades, are less important regarding either HIV-1 or JCV infection. Because Tat is known to interact with cellular proteins that play key roles in JCV replication (Gallia et al., 2000; Johnson, 2003), it is timely to understand how Tat structural changes alter the response of JCV to the different HIV-1 clades. The extent and results of such differences among clades represent gaps in our knowledge that need to be filled.
Methods
Plasmids, cell lines and cell culture.
Clones of HIV-1 clades C (GenBank accession #AF286227) and D (GenBank accession #U88822) were obtained from the NIH AIDS Research and Reference Reagent Program. The PCR products for exon 1 of each Tat gene (Tat72) were purified and ligated into pcDNA3.1/Zeo-FLAG. Clones encoding Tat 72 from clades B, another C, and two different forms of E were obtained as acknowledged. HIV-1 clade B LTR-luciferase construct, pLTR-luc, and pJCV-T, expressing JCV large T-antigen were obtained as acknowledged. Smad2, -3 and -4 cDNAs were in vector pRK-5 (Genentech), FLAG-tagged and under the control of the CMV-MIE promoter. JCV-E and -L, bearing JCV early (E) and late (L) gene promoters, in luciferase expression vector pGL3 (Promega), were obtained as acknowledged. Plasmid pTAT, expressed clade B Tat86 (86 aa) under control of the HIV-1 LTR promoter. Plasmid pMad1 contains a 260 bp segment of the JCV control region (JCV-NCCR). KG-1C cells are an oligodendroglioma line that supports replication initiated at the JCV ori (Daniel et al., 2001; Tanaka et al., 1983).
Purified Tat, antibodies and immunoblotting.
Purified HIV-1 Tat protein (clade B, 86 aa) was obtained from ProSpec. Antibodies recognizing Smad2/3 (rabbit polyclonal), Smad4 (mouse monoclonal), HA epitope (mouse monoclonal F-7) and Fast-1 (rabbit polyclonal) were from Santa Cruz Biotechnology. Mouse monoclonal anti-Tat antibody was from Abcam. Mouse monoclonal anti-polyomavirus large T-antigen antibody was from Calbiochem. Secondary antibodies for gel visualization from Li-Cor were labelled with IRDye 680 or IRDye 800CW near-infrared fluorophores. Imaging of immunoblots was performed using a Li-Cor Odyssey infrared detection system. TO-PRO-3 iodide (642/641, Invitrogen) was used for nuclear staining.
Luciferase reporter assays.
KG-1C cells (2.5×104 per well) were transfected as previously described (Stettner et al., 2009) using FuGene-6 reagent (Roche). Luciferase and Renilla activities were detected at 48 or 72 h using the Dual-Luciferase Reporter Assay system (Promega). Data were processed, including standard errors and P-values, using GraphPad Instat (Version 3.0) and Prism (Version 4) software. Relative luciferase (RLU) values were calculated in triplicate as follows: (sample Firefly luciferase/sample Renilla luciferase) – (control vector Firefly luciferase/control Renilla luciferase).
Immunofluorescence.
Coverslips with KG-1C cells were treated with 4 % paraformaldehyde, washed with PBS and incubated with methanol. Primary antibody diluted in 3 % BSA was applied for 12 h at a 1 : 50 conc. Confocal imaging employed a Zeiss LSM-510 Laser Confocal Microscope.
ChIP and real-time PCR.
KG-1C cells were transiently transfected with pBLCAT3-Mad1 plasmid, containing a 388 bp segment from the JCV Mad-1 strain (gbJ02226.1) comprising the origin of replication, for 72 h (Daniel et al., 2001). ChIP was performed as described previously (Kinoshita & Johnson, 2004). After reversal of the cross-linking, DNA analysis was done by quantitative real-time PCR using a Bio-Rad iCycler and primers 5′-CTTCTGAGTAAGCTTGGAGGC-3′ and 5′-GTTCCCTTGGCTGCTTTCCAC-3′, representing nt 5103–5123 and 212–232 of JCV Mad1, which amplified a JCV-NCCR sequence of 260 bp. Standard errors and P-values were calculated using GraphPad Prizm software. One-step reverse transcriptase PCR (RT-PCR) to measure Smad4 mRNA transcription was done with the Smad4 primer set: 5′-AAAGGTGAAGGTGATGTTTGGGTC-3′ (forward) and 5′-CTGGAGCTATTCCACCTACTGATCC-3′ (reverse) to amplify a 268 bp segment.
DNA replication initiated at the JCV origin.
KG-1C oligodendroglial cells were co-transfected with plasmids pMad1 (4.3 kb), containing a 260 bp segment with the JCV ori, and pJCT, expressing JCV large T-antigen (T-Ag). Plasmid DNA recovered 48 h after transfection was subjected to HindIII linearization, DpnI digestion and Southern blot hybridization. DpnI-resistant bands were detected on a nylon membrane by hybridization to the 260 bp NCCR insert labelled with 32P-phosphate. Intensities of the DpnI-resistant HindIII 4.3 kb bands, representing fully replicated plasmid DNA, were analysed using Imagequant 1.2.
Acknowledgements
NIH-NINDS grant NS35000 was to E. M. J. We thank Dr J. Hiscott for clones expressing Tat72 from HIV-1 clades B, C and E. We thank Dr Kamel Khalili for plasmids pJCV-T, pLTR-luc, pTat and pJCV-E and -L.