Comment on Xie et al, page 1980
Xie and colleagues examine the contribution to viral pathogenesis of an important protein-protein interaction motif in the transforming protein of HTLV-1.
Human T-cell leukemia viruses encode a protein called Tax that is the consummate adaptor protein. Interactions between Tax and a variety of protein partners activate transcription of virus and cellular genes, redirect signal transduction pathways, and disable cell cycle regulators. A transforming protein in vitro and in vivo, Tax sets the stage for multistep carcinogenesis by increasing cell proliferation and diminishing DNA repair processes.1 Several years ago, a new protein interaction motif was identified in HTLV-1 Tax (Tax1) that is shared with oncoproteins from DNA tumor viruses.2-4 In this issue of Blood, Xie and colleagues begin to unravel the actions of this PDZ-binding motif (PBM) in HTLV-1-infected T cells.
There are compelling reasons to believe that the PBM is important. HTLV-1 and HTLV-2 are close relatives, with about 70% sequence similarity, but only HTLV-1 is firmly associated with adult T-cell leukemia or inflammatory neurologic diseases. The Tax1 protein—but not Tax2—contains a PBM at its C-terminus that binds to proteins that contain PDZ domains.2-4 Of interest, the E6 oncoproteins from high-risk human papilloma virus (HPV) strains contain a PBM, whereas those from low-risk isolates do not. The adenovirus 9 (Ad9) ORF1 oncoprotein also contains a PBM, and HTLV-1 Tax, HPV18 E6, and Ad9 ORF1 proteins all associate with a PDZ domain in the human Dlg protein. Yeast 2-hybrid screens with Tax1 as bait have also recovered other PDZ-containing proteins. The hDlg protein is of interest here because it forms a complex with adenomatous polyposis coli protein (APC), which negatively regulates cell cycle progression. The APC gene is mutated in colon cancers and the mutated gene products generally lack a PBM. Moreover, the PBMs of Tax1, HPV E6, and Ad9 ORF1 interact with the same PDZ domain on Dlg as APC, suggesting that they may disrupt the APC-Dlg complex and deregulate cell cycle progression.
The paper by Xie and colleagues is the first to examine how the Tax1 PBM affects the course of virus infection in primary human T cells, the natural target for HTLV-1 replication. Previous studies on viral PBMs and Dlg have been done in nonlymphoid cells, including work from Fujii's laboratory (Hirata et al5 ), which showed that Tax1 PBM was necessary for efficient transformation of Rat1 fibroblasts in vitro. Xie et al have now done experiments with wild-type and mutant HTLV-1 and HTLV-2 viruses in primary T cells and in a rabbit model system. Deletion of the Tax PBM in HTLV-1 rendered the mutant virus unable to establish persistent infections in rabbits. Although important, this result does not provide mechanistic insights because we do not know whether the mutation affects cell proliferation, cell survival, immune response, or virus spread. The really interesting studies here addressed Tax1 PBM function by infecting primary lymphocytes and monitoring cell growth and survival. These experiments indicated that the Tax1 PBM contributes to an early boost in cell proliferation, reminiscent of the increased cell proliferation early after HPV31 infection of human keratinocytes. The HTLV-1 experiments are challenging on many levels, because of limitations in the cell culture and animal models and because Tax is a multifunctional protein and the effects of the PBM are subtle.
The convergence of HTLV-1 with DNA tumor viruses at the level of oncoproteins that target the same PDZ-containing cellular factors is intriguing. There are many compelling reasons to believe that the Tax1 PBM contributes to HTLV-1 pathogenesis and the preliminary data are consistent with this notion, but there is much work to do. The studies described here illustrate how future work should proceed—that is, by examining primary T cells infected with recombinant HTLV. There is a need to characterize the functions of Dlg and other PDZ domain-containing proteins in T lymphocytes so that we can better understand the consequences of Tax1 association. In these ways, we may soon discover which of the many PDZ-containing proteins is the real target of Tax1 and why it is necessary for the virus to interact with it. Furthermore, we would like to know how HTLV-2 establishes a life-long persistent infection without this function; after all, why does HTLV-2 not cause leukemia? ▪