The Transactivation Functions of HTLV-I Tax Are Suppressed by Foxp3
Previous studies by Bettelli and colleagues have demonstrated that Foxp3 can repress both the basal levels of NF-kappaB activation as well as tumor necrosis factor-alpha-stimulated NF-kappaB activation [15]. Our next step was to determine whether Foxp3 could also suppress the activation of NF-kappaB caused by a strong viral transactivator protein. HTLV-I encodes a multifunctional transactivator protein, Tax, capable of activating both the NF-kappaB and CREB pathways [24-29]. Since the HTLV-I Tax protein can function at multiple levels in both the cytoplasm and the nucleus to stimulate activation of NF-kappaB [28,29], we hypothesized that overexpression of Foxp3 may interfere with this process. However, since Tax-dependent HTLV-I gene expression is independent of NF-kappaB [18], we also hypothesized that Foxp3 would not affect Tax-dependent activation of the HTLV-I LTR. To test these hypotheses, we overexpressed HTLV-I Tax, full-length Foxp3, and/or DeltaFKH in HEK 293T cells cotransfected with an HTLV-I LTR or NF-kappaB reporter vector. As shown in Figure 4A, HTLV-I Tax strongly up-regulated NF-kappaB-dependent transcriptional activation (~60-fold). Interestingly, overexpression of Foxp3, but not DeltaFKH, suppressed Tax-mediated activation of NF-kappaB-dependent transcription. These observations further suggest that the carboxyl-terminal FKH domain is required for inhibiting activation of NF-kappaB in the presence of Tax in HEK 293T cells, strikingly similar to the requirements of Foxp3 inhibition of basal NF-kappaB activation shown in Figure 2B. Transactivation of the HTLV-I LTR was stimulated about 55-fold by overexpression of Tax (Figure 4B), while transfection of Foxp3 suppressed Tax-dependent HTLV-I LTR activation, although HTLV-I LTR activation in the presence or absence of Tax is independent of NF-kappaB or NF-AT (another transcriptional activator known to interact with Foxp3). Furthermore, overexpression of DeltaFKH also led to suppression of HTLV-I transactivation by Tax to a similar extent as full-length Foxp3 (Figure 4B). The suppressive effects shown in Figure 4A and 4B were not the result of Foxp3 down-regulating the expression of the transfected Tax plasmid as determined by real-time RT-PCR (Figure 4C). These results strongly suggest that Foxp3 interacts with transcriptional regulators in addition to NF-kappaB and NF-AT, and that the carboxyl-terminal FKH domain, and therefore localization to the nucleus, are not required for inhibition of Tax-mediated HTLV-I LTR activation (even in HEK 293T cells). 
To determine whether Foxp3 inhibited the transactivation functions of Tax by directly associating with this viral protein, we generated an expression vector in which HTLV-I Tax was fused in-frame to the carboxyl terminus of the Gal4 DNA-binding domain (Gal4-BD). This Gal4-BD-Tax fusion protein activated transcription of a synthetic promoter containing five Gal4 binding sites, while Gal4-BD was insufficient to stimulate transcription by itself (Figure 4D). Transactivation of the Gal4-resposive promoter by Gal4-BD-Tax remained relatively unaffected by overexpression of either EGFP (control), Foxp3, or DeltaFKH, suggesting that Foxp3 does not repress Tax transactivation by directly interfacing with the HTLV-I Tax protein. To confirm that Foxp3 had a direct effect on HTLV-I replication, we transfected HEK 293T cells with a well-characterized HTLV-I infectious molecular clone (termed ACH) [30] in the presence of full-length Foxp3, DeltaFKH, or control vector. ACH has been previously shown to direct the expression of viral antigens, produce infectious virus, and transform CD4+ T cells both in vitro and in vivo [31,32]. After 24 h, the amount of viral antigen expression, in this case Tax mRNA, was detected by a sensitive real-time RT-PCR assay. As illustrated in Figure 5, the level of Tax mRNA synthesized from ACH was down-regulated in the presence of Foxp3 compared to the level produced in the presence of the control vector. DeltaFKH did not have a discernable affect on Tax expression. These data indicate that Foxp3 is capable of repressing the expression of Tax from an infectious HTLV-I molecular clone. 
