|Home | About | Journals | Submit | Contact Us | Français|
Modulation of T-cell receptor expression and signaling is essential to the survival of many viruses. The U24 protein expressed by human herpesvirus 6A, a ubiquitous human pathogen, has been previously shown to downregulate the T-cell receptor. Here, we show that U24 also mediates cell surface downregulation of a canonical early endosomal recycling receptor, the transferrin receptor, indicating that this viral protein acts by blocking early endosomal recycling. We present evidence that U24 is a C-tail-anchored protein that is dependent for its function on TRC40/Asna-1, a component of a posttranslational membrane insertion pathway. Finally, we find that U24 proteins from other roseoloviruses have a similar genetic organization and a conserved function that is dependent on a proline-rich motif. Inhibition of a basic cellular process by U24 has interesting implications not only for the pathogenicity of roseoloviruses but also for our understanding of the biology of endosomal transport.
Human herpesvirus 6 (HHV-6) and HHV-7 are closely related viruses and the sole human members of the Roseolovirus genus of the betaherpesviruses (12). Two subtypes of HHV-6 (HHV-6A and HHV-6B) have been identified (2, 53). These variants share many biological properties (2) and a high level of sequence homology (16, 24). Both HHV-6A and HHV-6B are acquired early in childhood and are ubiquitous within the human population (22). Primary infection is usually self-limiting but can be associated with exanthem subitum (mainly with HHV-6B  and HHV-7 ), a benign pathology characterized by a transient rash followed by fever. In rare occasions, severe complications can occur (40). Once primary infection is resolved, HHV-6 persists in peripheral blood mononuclear cells (PBMC) (48) and may be latent in other tissues as well (5, 15). HHV-6 reactivation has been proposed to play a role in the pathogenesis of AIDS (32) and several other diseases such as mesial temporal lobe epilepsy (19), multiple sclerosis (55, 60), and encephalitis (38, 61).
Although, HHV-6A and HHV-6B infect many cell types within the PBMC population, CD4+ T cells are believed to be the main targets (34, 58). The biological activity of T cells is tightly regulated by the T-cell receptor (TCR) complex expressed at the cell surface. Engagement of the TCR complex results in CD4+ T-cell activation through a series of biochemical events mediated by the CD3 molecules (25). Since CD4+ T cells play a central role in regulating the antiviral response, it is not surprising that many viruses have evolved mechanisms to modulate T-cell function. For example, Nef, a protein encoded by the human immunodeficiency virus type 2 (HIV-2) and simian immunodeficiency virus (SIV), downmodulates TCR/CD3 from infected T cells, blocking their responsiveness to activation (51). This property is thought to have evolved to maintain viral persistence in the context of an intact host immune system.
Similar to HIV-infected cells, HHV-6-infected CD4+ T cells express low levels of TCR-CD3 molecules (34). This phenotype results from a block in CD3 transcription (33) as well as from an inhibition of the constitutive recycling of the CD3 and TCR molecules back to the plasma membrane (57). Inhibition of TCR/CD3 recycling is mediated by U24, an HHV-6 gene product. Transfection of U24 in T-cell lines mediates a rapid relocalization of the CD3 and TCR molecules from the plasma membrane to early endosomes. As a consequence, U24-expressing cells are impaired in their ability to become activated (57).
Here, we provide evidence that U24 is a C-tail-anchored (TA) protein whose function can be inhibited by expression of a dominant negative (DN) version of TRC40/Asna-1, a protein involved in a posttranslational membrane insertion pathway (44, 56). We show that U24 mediates the downregulation of the TCR complex and also the transferrin receptor (TfR) through a PPXY motif near the amino terminus of U24, suggesting a general block in early endosomal recycling. Finally, we found that U24 is specific to roseoloviruses, with U24 sequence and function being conserved in the two subtypes of HHV-6 and in HHV-7.
The Jurkat CD4+ T-cell line was generously provided by Art Weiss (University of California, San Francisco, CA), and the SLK endothelial cell line (23) and 293T cell line were provided by Don Ganem (University of California, San Francisco, CA). HSB-2 cells and HHV-6A (strain GS)-infected HSB-2 cell lines were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH, from Electro-Nucleonics, Inc., and Robert Gallo, respectively (3, 4, 31, 46). Jurkat and HSB-2 cell lines were grown in RPMI (Gibco) medium containing 5% fetal bovine serum (FBS) and 100 μg/ml penicillin-streptomycin. 293T and SLK cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 10% FBS and 100 μg/ml penicillin-streptomycin. Jurkat transfections were performed as described previously (56). Briefly, 16 × 106 Jurkat cells were transfected with 30 μg of DNA via electroporation. Cotransfections were performed in the same manner except that the enhanced green fluorescent protein (EGFP)-encoding plasmid was transfected at a ratio of 1:6 with the non-EGFP-encoding plasmid for a total of 30 μg. 293T and SLK transfections were performed using Fugene HD (Roche) according to the manufacturer's specifications. Phycoerythrin (PE)-conjugated CD3 (UCHT1), αβ TCR (IP26), and HLA-A, -B, and -C antibodies were purchased from BD Biosciences while purified polyclonal anti-hemagglutinin (HA) (clone Y11) antibody and rhodamine-conjugated donkey anti-mouse IgG were purchased from Santa Cruz Biotechnology. Monoclonal anti-HA antibody (clone 3F10) was purchased from Roche Applied Science. Horseradish peroxidase (HRP)-conjugated goat anti-rat IgG was purchased from GE Healthcare. Anti-CD3 (OKT3) antibodies were a gift from Nilahb Shastri (University of California, Berkeley). Cy5-conjugated donkey anti-rabbit IgG was purchased from Jackson.
U24 and EGFP-Rab constructs (57) as well as Mir1 constructs (11, 13) have been described previously. EGFP-Rab7a was amplified from Jurkat cDNA and cloned as described previously (57). HA-U24 and HA-EGPF were constructed by adding the sequence encoding the hemagglutinin epitope (MYPYDVPDYA) to the 5′ end of the U24 and EGFP open reading frames (ORFs), respectively. TRC40/Asna-1 and Hsp40 (DNAJB1) were amplified from Jurkat cDNA and cloned into pCDNA 3.1(+) (Invitrogen). TRC40/Asna-1 DN (G46R) was created by changing the guanine at position 136 to a cytosine using standard overlap PCR methods and cloned into pCDNA3.1(+). Hsp40J, containing the J domain of Hsp40, was created by a carboxy-terminal truncation of Hsp40 after amino acid N86. U24 mutants were created in a similar fashion at the positions indicated in the text. U24 from HHV-6B (strain Z29) and HHV-7 (strain RK) were cloned into pIRES2-EGFP by PCR amplification using virus-infected cells kindly provided by Dharam Ablashi (HHV-6 Foundation).
A total of 5 × 105 cells were used for each condition and incubated with PE-conjugated primary antibodies for 30 min and subsequently washed twice with 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS). Fluorescence was determined using a Beckman-Coulter EPICS XL flow cytometer, and cells were analyzed and gated on appropriate forward and side scatter profiles using FlowJo (Tree Star, Inc.) software.
Jurkat or 293T cells were transiently transfected with relevant plasmids. After 36 h of expression, cells were washed four times in ice-cold PBS, resuspended in 100 mM sodium carbonate, pH 11.0 (extraction solution), lysed with 10 strokes in a 2-ml dounce homogenizer, and centrifuged at 68,000 × g. The soluble fraction was removed, and the pellet was rinsed once with extraction solution before resuspension in extraction solution to remove residual soluble fraction. Genomic DNA present in the membrane fraction was sheared by multiple passages through a 30-gauge needle. Soluble and membrane fractions were run on a 12.5% acrylamide denaturing gel, transferred to a polyvinylidene difluoride (PVDF) membrane, and blotted with monoclonal anti-HA antibody and HRP-conjugated goat anti-rat IgG antibody. Samples were neutralized by adding an equivalent volume of gel loading buffer (31 mM Tris, pH 6.8, 1% SDS, 0.002% bromophenol blue, 5% glycerol, and 5% β-mercaptoethanol).
Protein concentration was quantified using a bicinchoninic acid (BCA) protein assay kit (Pierce). Equivalent amounts of protein were run on a 12.5% acrylamide denaturing gel, transferred to a PVDF membrane, and blotted with monoclonal HA antibody followed by HRP-conjugated goat anti-rat IgG antibody.
Jurkat cells were transiently cotransfected with plasmids encoding EGFP and either TRC40/Asna-1 wild type, TRC40/Asna-1 DN, dynamin wild type, or dynamin K49E. After 36 h of expression, live cells were enriched by Histopaque 1083 separation. Cells were serum starved in RPMI medium for 30 min, after which Alexa Fluor 647-conjugated transferrin was added to cells at a concentration of 500 ng/ml. Aliquots were taken at 0, 2, 5, 15, 30, and 60 min and placed in an ice bath to stop internalization of transferrin. Cells were washed in ice-cold PBS, and extracellularly bound transferrin was removed by two washes with PBS, pH 5. Cells were resuspended in PBS and mean Alexa Fluor 647 fluorescence was determined by flow cytometry (LSR-II; BD Biosciences) gating on the live-cell population and GFP-expressing cells. The percentage of internalization at each time point was calculated as follows: (Stx − St0)/(St30 − St0) × 100, where Stx is the mean fluorescence of cells at each time point, St0 is the mean fluorescence of cells at time zero, and St30 is the mean fluorescence of cells collected at 30 min after internalization when the maximum internalization of transferrin was observed in control cells.
Jurkat cells were transiently transfected with plasmids encoding a bicistronic transcript of HA-U24 and GFP-tagged Rab proteins. After 24 h of expression, cells were sorted for moderate expression of GFP corresponding to cells expressing moderate to low CD3 at the cell surface. Cells were then fixed with 3% paraformaldehyde (PFA)-0.03 M sucrose, permeabilized with 0.02% Triton X-100, and blocked with 10% goat serum-3% BSA in PBS. Cells were stained with anti-CD3 (OKT3) antibodies and polyclonal HA (Y11) antibodies, followed by rhodamine-conjugated anti-mouse and Cy5-conjugated anti-rabbit antibodies, respectively. Cells were visualized with a Zeiss LSM 510/NLO META confocal microscope. Micrographs of Jurkat cells were obtained using a Plan-Apochromat 63×/1.3 oil objective scanned at a magnification of ×6 with a line average value of 16. Micrographs of SLK cells were obtained without prior cell sorting, using the same objective, but were scanned first without magnification and subsequently at a magnification of ×4 with a line average of four. Images were analyzed with Zeiss LSM software.
We have previously reported that the U24 ORF from HHV-6A encodes a protein that downregulates the TCR complex from the surface of CD4+ T cells (57). Similarity searches performed using U24's protein sequence identified only positional homologues in HHV-6 variant B and HHV-7, both CD4+ T-cell-tropic viruses of the Roseolaviridae family (12). While U24 from HHV-6 variant B (U24-6B) is 84% identical to its counterpart in variant A (U24-6A), the positional homologue in HHV-7 (U24-7) shares only a 28% amino acid identity with U24-6A (Fig. (Fig.1A).1A). In order to test whether these positional homologues also function to downregulate the TCR complex from the cell surface, U24 ORFs from both variants of HHV-6 and HHV-7 were cloned into the multiple cloning site of pIRES2-EGFP. This vector contains an internal ribosome entry site between the multiple cloning site and EGFP open reading frame. As a consequence, cells expressing EGFP should also express our gene of interest. U24-expressing vectors were transfected into the CD4+ T-cell line Jurkat, and GFP-expressing cells were analyzed for CD3 and αβ TCR surface expression. While there have been conflicting reports as to the extent of downregulation of the TCR complex in in vitro infected cells (20, 21), we found that the ability to downregulate this complex is encoded in the genomes across this viral family (Fig. (Fig.1B1B).
In addition to its sequence and functional similarities, U24 proteins from these three viruses also share a basic architecture. All three contain a hydrophobic region at the carboxy terminus preceded by an unstructured amino-terminal tail lacking a signal sequence. Analysis with 3′ and 5′ rapid amplification of cDNA ends (RACE) using RNA from infected HHV-6A cultures confirmed that U24 is not an exon of a larger coding region (data not shown), suggesting that U24 may be a tail-anchored (TA) protein, a class of proteins characterized by a carboxy-terminal transmembrane region that is posttranslationally inserted into membranes (8). Localization of TA proteins into specific membranes, such as the endoplasmic reticulum, peroxisomes, and mitochondria, is, in part, dictated by charged amino acids flanking either side of the transmembrane domain (8, 9, 30). Similar to TA proteins, U24 encodes triplets of positively charged amino acids flanking both ends of its putative transmembrane domain.
In order to establish that U24 is an integral membrane protein, we performed an alkaline-carbonate extraction of cells expressing an HA-tagged U24. This construct was made by adding the HA epitope tag to the amino terminus of U24. When transiently transfected into Jurkat cells, HA-U24 is able to downregulate the TCR complex to identical levels to its untagged counterpart (Fig. (Fig.1C,1C, left histogram). Furthermore, HA-U24 is expressed proportionally to EGFP when a bicistronic vector encoding these two proteins is transfected into cells (Fig. (Fig.1C,1C, right histogram). Jurkat cells transiently expressing either HA-U24 or HA-tagged controls were fractionated into membrane and soluble components. HA-U24, as well as the type III transmembrane protein HA-tagged Mir1, was present in the membrane fraction while HA-tagged EGFP was present mostly in the soluble fraction (Fig. (Fig.1D,1D, first six lanes). Mutating the three positively charged amino acids on the amino-terminal side of the transmembrane domain to neutral glutamines (U24 RRK) almost completely abolishes the ability of U24 to downregulate CD3 while a truncation that eliminates the charged amino acids carboxy-terminal to the transmembrane domain does not affect U24 function (Fig. (Fig.1E,1E, left histogram). Interestingly, when these two mutations occur in the same protein, there is a slight recovery of the ability of U24 to downregulate CD3 (Fig. (Fig.1E,1E, right histogram), highlighting the positional importance of charged residues and their context within a TA protein. The functional defect of U24 RRK is not due to lack of expression or membrane insertion (Fig. (Fig.1D,1D, last two lanes). Altogether, these data indicate that U24 has the basic architecture of TA proteins.
Two pathways have been identified in the posttranslational insertion of TA proteins into membranes. The first involves TRC40/Asna-1, a homologue of a bacterial ATPase, ArsA, that was shown to be essential to TA protein membrane insertion in biochemical reconstitution assays (18, 56). A dominant negative mutation in the ATPase domain of TRC40/Asna-1 blocked membrane insertion of TA proteins in these assays (54). More recently, the Saccharomyces cerevisiae homologue of TRC40/Asna-1, Get3, was shown to be a component of a Get complex that is essential for membrane insertion and localization of certain TA proteins (54). The second pathway involves a chaperone pair, Hsc70 and Hsp40, that has been shown as important in TA protein membrane insertion for a separate subset of TA proteins that are not inserted by the TRC40/Asna-1 pathway (1, 44). In order to determine if TRC40/Asna-1 is involved in the membrane insertion of U24, plasmids encoding wild-type or dominant negative (DN) versions of TRC40/Asna-1 were transfected into Jurkat cells together with plasmids encoding either a bicistronic transcript of U24 and EGFP (U24/EGFP) or EGFP alone. After 24 h, levels of surface CD3 were determined by flow cytometry. While TRC40/Asna-1 DN had no effect on CD3 levels in the absence of U24, the ability of U24 to downregulate CD3 was inhibited in the presence of TRC40/Asna-1 DN (Fig. (Fig.2A).2A). Furthermore, cotransfection of U24 with the J domain of Hsp40, which has been shown to block the ability of Hsc70-related proteins (39), did not affect the ability of U24 to downregulate CD3. Interestingly, however, there was a slight effect on CD3 surface levels in cells that were not transfected with U24.
Because TA proteins have been found to be involved in membrane vesicle fusion and other aspects of the secretory pathway (47), it remains possible that TRC40/Asna-1 may not be affecting U24 membrane insertion but, instead, may affect other TA proteins in endosomal pathways necessary for U24 function. We first tested whether TRC40/Asna-1 affects the function of another viral protein, Mir1, that mediates major histocompatibility complex class I (MHC-I) downregulation from the cell surface and its trafficking to the lysosome for degradation. The ability of the type III transmembrane protein Mir1 to downregulate MHC-I was not affected by TRC40/Asna-1 DN expression in the same cell type (Fig. (Fig.2B).2B). These data suggest that there is no defect in the ability of Mir1 and MHC-I to traffic to the membrane or the ability of MHC-I to become endocytosed. In order to further examine the effect of TRC40/Asna-1 on endosomal trafficking, we transiently transfected TRC40/Asna-1 as well as TRC40/Asna-1 DN into Jurkat cells and measured the internalization of an Alexa Fluor 647-conjugated transferrin by flow cytometry. Transferrin and transferrin receptors, important factors in iron uptake by cells, have been shown to traffic in a similar manner to the TCR complex. Jurkat cells transiently transfected with plasmids encoding TRC40/Asna-1 wild type and DN show no defect in transferrin internalization in contrast to a dominant negative version of dynamin, a protein required for most types of endocytosis (Fig. (Fig.2C).2C). Furthermore, transferrin receptor levels at the cell surface remain nearly unchanged upon expression of TRC40/Asna-1 DN (Fig. (Fig.2C,2C, histogram). Together, these data suggest that there is no inherent defect in early endosomal trafficking and surface receptor expression in TRC40/Asna-1 DN-expressing cells.
We next sought to characterize the effect of TRC40/Asna-1 on the insertion of U24 into membranes. 293T cells were transiently cotransfected with plasmids encoding TRC40/Asna-1 wild type or DN and HA-U24. Integral membrane proteins were separated from soluble proteins by carbonate extraction, separated by SDS-PAGE, transferred to PVDF membranes, and blotted with anti-HA antibodies (Fig. (Fig.2D).2D). Similar transfections with HA-Sec61β were performed in order to compare the effects of TRC40/Asna-1 on another TA protein previously shown to be inserted into membranes through this pathway (44, 54, 56). HA-Mir1 was used as a negative control as it is a type III membrane protein and should therefore be cotranslationally inserted into membranes. EGFP was also cotransfected in each case as a marker of transfection efficiency and as a marker for the soluble fraction of cell lysates. While a small fraction of HA-U24 is consistently soluble when coexpressed with TRC40 DN, the vast majority of U24 is inserted into membranes (Fig. (Fig.2D,2D, first six lanes) and thus cannot account for the defect in U24-mediated downregulation of the TCR complex from the cell surface. These results agree with a recent paper suggesting that the yeast homologue of TRC40/Asna-1, Get3, is essential for the correct subcellular localization of the target TA protein. Schuldiner et al. show that in the absence of Get3, TA proteins normally inserted into endoplasmic reticulum membranes were inserted into mitochondrial membranes instead (54). Insertion into membranes other than those from which U24 may act could explain the functional defect seen in U24-expressing cells. Although we have not seen colocalization between U24 and mitochondria upon expression of TRC40/Asna-1 DN (data not shown), we have not confirmed whether U24 could be mislocalized to other membranes.
Charged residues in the transmembrane domains of the TCR complex have been shown to be important in complex formation (14). While the placement of positively charged residues proximal to U24's transmembrane region is consistent with the known architecture of TA proteins, it remains possible that these residues are important for association with the TCR complex and that mutations at these positions disrupt the complex. Since numerous attempts to immunoprecipitate the TCR complex with U24 have not been successful (data not shown), we examined CD3 and U24 subcellular localization using immunofluorescence and confocal analysis.
Previously, our group has shown that upon expression of U24, CD3 becomes excluded from Rab11 recycling endosomes and is sequestered mostly in Rab4- and Rab5-containing early endosomes as well as some increased localization to Rab9-containing late endosomes (57). To determine U24 localization, these experiments were repeated using cells expressing HA-U24. Cells undergoing CD3 downregulation but not having been fully downregulated were selected from this pool and stained for confocal microscopic analysis. This was done in order to visualize the location of CD3 and HA-U24 as the downregulation was occurring.
As previously reported, CD3 colocalized with Rab4 sorting and Rab5 early endosomes in U24-expressing Jurkat cells (Fig. (Fig.3,3, top two rows) (57). Interestingly, colocalization of these endosomal markers with HA-U24 was also observed in over 80% of cells; however, we rarely saw colocalization of CD3 and HA-U24 (Fig. (Fig.3,3, last column). HA-U24 also colocalized with late endosomes (Fig. (Fig.3,3, Rab7 and Rab9). Furthermore, while we observed colocalization of HA-U24 with several of our Rab constructs, HA-U24 did not fully colocalize with any one Rab construct, suggesting that HA-U24 exists in several endosomal compartments.
The absence of colocalization between CD3 and U24 suggests that U24 might alter the early endocytic recycling pathway rather than having a specific effect restricted to the TCR complex. To test this hypothesis, we tested the effect of U24 on transferrin receptor surface expression. The transferrin receptor (TfR) has been extensively studied as an early endosomal recycling receptor (17, 43) and has been observed to traffic in a manner similar to the TCR complex (7, 28). Cells that express (Jurkat) or lack (293T and SLK) the TCR complex were transiently transfected with U24/EGFP. TfR surface expression was measured by flow cytometry. As shown in Fig. Fig.4A,4A, the TfR is downregulated in cell lines of distinct lineages, regardless of TCR complex expression. These data support the idea that U24 mediates a general block of endosomal recycling. This block in recycling seems to be specific to early endosomal recycling as surface levels of receptors that recycle through different endosomes, MHC-I (6, 37) and CD1d (26), did not change upon transient transfection of U24 (Fig. (Fig.4B).4B). In addition, expression of TRC40/Asna-1 DN also affected the ability of U24 to downregulate the TfR in SLK cells (Fig. (Fig.4C).4C). Similar to what we observed in Jurkat cells, U24 expressed in SLK cells largely colocalized with Rab4 sorting endosomes and, to a lesser extent, with Rab5 early endosomes but not with Rab7- or Rab11-containing compartments (Fig. (Fig.4D4D).
Based on conserved amino acids in the three roseolovirus U24 ORFs, a limited alanine-scanning of U24-6A was performed to identify essential regions for U24 function. U24 wild type and mutants were transiently transfected into Jurkat cells and assayed for surface expression of CD3. This analysis revealed a proline rich region that is essential to U24's function (Fig. (Fig.5A,5A, asterisk) while alanine substitutions in neighboring amino acids or substitutions in a separate proline-rich region had no effect on CD3 downregulation (Fig. (Fig.5A,5A, circles with x's). The amino-terminal proline-rich region is highly conserved in U24-6B and U24-7. In order to determine whether these prolines are functionally conserved across the roseolovirus family, we performed analogous mutations in U24s encoded by HHV-6B and HHV-7. Mutations in these regions abolished the ability of U24-6B and U24-7 to downregulate CD3 (Fig. (Fig.5B),5B), indicating a conserved mechanism of TCR downregulation.
In order to identify specific motifs that may be important for U24-mediated CD3 downregulation, further mutational analysis was performed. Furthermore, levels of cell surface transferrin receptors were also assessed to determine whether CD3 and TfR downregulation occurred by separate mechanisms or whether the same motif was responsible for both phenotypes. Jurkat cells were transiently transfected with vectors encoding wild-type U24-6A and U24 mutants and assayed for both TfR and CD3 downregulation (Fig. (Fig.5C).5C). Both TfR and CD3 were downregulated in the presence of wild-type U24-6A while U24 PPXY mutants could not downregulate either receptor (Fig. (Fig.5D),5D), suggesting that CD3 and TfR are downregulated through a common mechanism. Western blot analysis of wild-type U24 and mutants revealed similar expression levels (Fig. (Fig.5D5D).
Surprisingly, while some mutants reveal a distinct banding pattern compared to the wild-type U24, this pattern does not correlate with the ability of U24 to block endosomal recycling. The significance or composition of the doublet seen in U24 Western blots is still unclear. The doublet, however, is not due to a truncation at the amino terminus (where the HA tag is located) or the carboxy terminus as Western blots of a functional U24 missing the six amino acids carboxy-terminal to the transmembrane region show the same doublet that runs only slightly faster than the wild-type protein (Fig. (Fig.5D,5D, lower blot). In addition, while the size of U24 is predicted to be 10 kDa, HA-tagged constructs of U24 consistently run at double this size, indicating extensive but still unidentified posttranslational modifications.
We have previously shown that U24 expression in the Jurkat T-cell line causes the downregulation of the TCR complex from the cell surface by inhibiting the recycling of this complex back to the plasma membrane (57). In the present study, we expanded our findings to include the U24 ORF encoded by HHV-6B and HHV-7. There have been conflicting reports of CD3 downregulation in these viruses. While it has been established by several groups that HHV-6A can downregulate CD3 (20, 21, 33, 34), Furukawa et al. have tested the ability of one strain of HHV-7 to downregulate CD3 in in vitro infected human CD4+ primary T cells and found no effect (20). However, we found that U24 cloned from the same HHV-7 strain (RK) downregulates CD3. Similarly, while the same group also observed only a modest downregulation of CD3 in HHV-6B-infected cells (20), we found that U24 isolated from the same HHV-6B strain mediated a robust downregulation of the TCR complex from the surface of transfected T cells. Another study using a strain of HHV-6B (PL1) isolated from a healthy individual demonstrated a nearly total downregulation of CD3 from infected ex vivo lymphoid tissues (21). Our results indicate that both variants of HHV-6 as well as HHV-7 encode the ability to downregulate the TCR complex. Whether these differences are due to experimental procedures, virus strains, or viral regulation of U24 is unknown. Clinical work studying the reactivation of these viruses in healthy individuals and the viruses' effects on TCR cell surface downregulation may aid in reconciling these conflicting data.
U24 had originally been characterized as a protein that downregulates the TCR complex from the surface of T cells (57). Our present data indicate that the transferrin receptor is also downregulated by U24. The downregulation of TfR by U24, as well as the inability to detect any physical or spatial interaction between U24 and the TCR complex, suggests that U24 acts to block early endosomal recycling. Interestingly, U24 does not affect receptors that recycle through other means. Surface levels of CD1d, a molecule that presents lipid antigens to NK T cells (49, 65), are unaffected by U24 expression, indicating that a late endosomal/lysomal recycling pathway is unaffected. Furthermore, cell surface levels of MHC-I, which recycles though an ADP ribosylation factor-6 dependent, tubular endosome recycling pathway, are also unaffected (10, 37). Our data indicate that U24 may be functioning in one of two ways: either by affecting proteins that specifically target proteins to certain endosomal compartments or by affecting vesicular fusion of specific endosomes carrying TCR and TfR. While the data presented here do not rule out either possibility, we prefer the latter explanation as we have previously shown that internalization of TCRs is partially blocked in U24-expressing cells after 24 h of expression. This effect could be due to the partial sequestration of unknown factors important in TCR internalization that may also recycle back to the plasma membrane. Possible effects on vesicular fusion mediated by U24 provide a simple explanation of how non-target proteins involved in receptor endocytosis may be inadvertently sequestered, thereby affecting subsequent internalization of the target receptor. This hypothesis may also explain why the affected cargo and U24 do not colocalize. However, it remains possible that these internalization factors may recycle through the same mechanisms as the TCR and TfR. Elucidation of the molecular mechanism of U24 may lead to a greater understanding of how early endosomal recycling is regulated.
A PPXY motif has been identified as essential to U24 function, but its binding partners have yet to be identified. PPXY sequences have been known to bind type 1 WW domains, small protein interaction domains containing, in most cases, two conserved tryptophan residues (35). Preliminary experiments have demonstrated that HA-U24 can be coimmunoprecipitated with overexpressed WW domain-containing proteins and that this association is dependent on a complete PPXY motif (data not shown). While this suggests that U24 function is dependent on a WW domain-containing protein, the cognate ligand has not yet been identified due to the promiscuous nature of the in vitro association between U24 and WW domain-containing proteins.
In this report, U24 has been characterized as a membrane protein with distinct characteristics of TA proteins, namely, charged residues abutting the transmembrane domain as well as the absence of a signal peptide. By using a dominant negative TRC40/Asna-1 initially characterized by Stephanovic and Hedge (56), we confirmed that U24 is likely a TA protein. This is the first instance that a component of the TA protein insertion machinery has been shown to affect a TA protein function in a viable mammalian cell as opposed to observations made based on biochemical assays (1, 18, 44, 56) or assays using permeabilized cells (44). Our data correlate well with experiments performed in S. cerevisiae, where insertion of TA proteins into membranes is seen in the absence of Get3, the yeast homologue of TRC40/Asna-1 (54). Schuldiner et al. have demonstrated that, in the absence of Get3, TA proteins are mislocalized to mitochondrial membranes (54). While we have not seen any mitochondrial localization of U24 in the presence of TRC40/Asna-1 DN (unpublished observations), U24 may be mislocalized to other membranes within the cell, thereby explaining the decreased ability of U24 to downregulate both the TCR complex (Fig. (Fig.2A)2A) and TfR from the cell surface (Fig. (Fig.4C4C).
Abell et al. have shown in biochemical assays (1) and, more recently, in semipermeabilized cells that the chaperone pair Hsp40/Hsc70 is essential to TA membrane insertion (44). We used the J domain of Hsp40 as a dominant negative inhibitor of the Hsp40/Hsc70 complex and found that it had no effect on U24 function (Fig. (Fig.2A).2A). While the effectiveness of this dominant negative has not been confirmed in our own experiments, a report published by Rabu et al. suggests that the TRC40/Asna-1 and Hsp40/Hsc70 pathways are distinct and that the transmembrane regions of TA proteins determine which pathway will be utilized (44). Since expression of functional U24 can be easily assessed (downregulation of TCR complex and TfR), this protein may be useful in determining other components of the TA protein insertion complex. TA proteins have been increasingly identified as essential in many basic cellular processes including protein trafficking (47) and cell survival (52). The ability of U24 to disrupt endocytic recycling is consistent with the role of other TA proteins, such as the syntaxins (47), to facilitate vesicular traffic and fusion. While the molecular mechanism of U24 function has not yet been elucidated, the putative cytoplasmic domain of U24 may be acting to recruit factors that disrupt membrane fusion.
What is the physiological significance of U24 function? We have previously shown that U24 can block activation of T cells mediated by antigen-presenting cells (57). This function of U24 may play a role in controlling viral replication by blocking extracellular signals that might dramatically increase viral titers. Unregulated viral replication may be harmful to a virus whose most effective means of immune evasion is latency (27, 29). It is unclear whether downregulation of the transferrin receptor affects virus survival. Close inspection of the flow cytometry data shows that at the point of maximal TCR downregulation, the transferrin receptor has not yet reached full downregulation (unpublished observations), indicating that U24 is more adept at downregulating the TCR complex than the transferrin receptor. Whether this is due to the difference in the kinetics of internalization and recycling is unclear. Although the proline-rich motif shown to be essential to the downregulation of both the TCR and TfR suggests that the two receptors are downregulated through the same mechanism, TfR downregulation may be incidental to viral infection. Interestingly, Nef from HIV-1 has also been shown to alter endosomal recycling compartments through a leucine motif, thereby affecting transferrin receptor recycling back to the cell surface (36). Unlike U24, however, Nef from HIV-1 does not typically have the ability to downregulate the TCR, a phenotype more commonly associated with Nef from SIV (50). Clearly, these processes have been decoupled in HIV/SIV through the expression of differential Nef proteins, and this suggests that TCR expression and early endocytic recycling modulation may each play a separate role in the pathogenesis of these closely related viruses.
The possibility exists, however, that U24 may be acting in a completely separate mechanism such as virus assembly and egress. Recent work by Mori et al. has demonstrated that assembly and egress of HHV-6 occurs in multivesicular bodies that release small vesicles in addition to mature virions, a pathway that is distinct from that of the well-characterized alphaherpesviruses (42). Though we have tested several dominant negative inhibitors of the multivesicular body pathway and seen no effect on U24 function (unpublished observations), we cannot exclude the possibility that U24 plays a role in this process.
Reactivation of HHV-6 has often been seen in multiple sclerosis patients, but whether HHV-6 plays a role in the pathogenesis of this autoimmune disease is unknown (41). Interestingly, HHV-6 can reside in oligodendrocytes (45) that synthesize the myelin sheath (63). Recent work by Winterstein et al. has shown that some myelin sheath proteins may traffic in a manner similar to transferrin (62). Whether the block in endosomal recycling by U24 causes a disruption in myelin sheath formation is unknown, but the possible pleiotropic effects of U24 expression on the surface receptors of T cells as well as other HHV-6A target cells and its effects on disease warrant further study.
We acknowledge the CHPS Imaging Core for assistance with microscopy as well as the CHPS Mouse/Virus core for providing viral stocks. We thank Hector Nolla for his assistance in cell sorting. We also thank Nadine Jarousse for invaluable assistance with the preparation of this manuscript and Ken Cadwell for his cloning expertise.
Published ahead of print on 18 November 2009.