|Home | About | Journals | Submit | Contact Us | Français|
Tetherin has been characterized as a key factor that restricts viral particles such as HIV and hepatitis C virus on plasma membranes, acts as a ligand of the immunoglobulin-like transcript 7 (ILT7) receptor in tumor cells, and suppresses antiviral innate immune responses mediated by human plasmacytoid dendritic cells. However, the normal cellular function of Tetherin without viral infection is unknown. Here we show that Tetherin not only serves as a substrate of autophagy but itself regulates the initiation of autophagy. Tetherin interacts with the autophagy/mitophagy suppressor LRPPRC and prevents LRPPRC from forming a ternary complex with Beclin 1 and Bcl-2 so that Beclin 1 is released to bind with PI3KCIII (class III PI3K) to activate the initiation of autophagy. Suppression of Tetherin leads to impairment of autophagy, whereas overexpression of Tetherin causes activation of autophagy. Under mitophagic stress, Tetherin is concentrated on mitochondria engulfed in autophagosomes. Tetherin plays a general role in the degradation of autophagosomes containing not only the symbiotic mitochondria but also, possibly, the infected virus. Therefore, Tetherin may enhance autophagy and mitophagy to suppress tumorigenesis, enhance innate immune responses, or prevent T cell apoptosis or pyroptosis.
Mammalian cells primarily use the autophagy system to degrade dysfunctional organelles, misfolded/aggregated proteins, and other macromolecules (1). The initiation of autophagy is regulated by the mammalian target of rapamycin (mTOR)2 pathway. The antiapoptotic protein Bcl-2 or B cell lymphoma extra-large (Bcl-xL) may exhibit opposite functions in the initiation of autophagy. Both Bcl-2 and Bcl-xL inhibit the initiation of autophagy through the PI3K-AKT-mTOR pathway by sequestering the Bcl-2 interactive Beclin 1 or activate the initiation of autophagy through the liver kinase B1 (LKB1)-5′-adenosine monophosphate-activated protein kinase-mTOR pathway by increasing the levels of the cyclin-dependent kinase inhibitor 1B (P27) (2, 3). After autophagy is initiated, the autophagy marker microtubule-associated light chain 3 (LC3) is converted to LC3-II, which is inserted into the isolation membrane (phagophore) (4). LC3-interactive microtubule-associated protein family 1 small form (MAP1S) has been identified as a positive regulator of autophagy, and its depletion leads to autophagy defects under nutritive stress and accumulation of dysfunctional mitochondria (5). The substrates of autophagy, including protein aggregates and dysfunctional organelles, are brought into autophagosomes by binding with the LC3-II-interactive substrate receptor P62 (6, 7). LRPPRC is a mitochondrion-associated protein. It binds with autophagy regulators Beclin 1 and Bcl-2 and the mitophagy regulator Parkin (8, 9). It prevents the formation of the Beclin 1-PI3KCIII complex of autophagy initiation (8). Under mitophagic stress, LRPPRC sequesters the Parkin that translocates to the mitochondrion to suppress mitophagy (9).
Tetherin, also called HM1.24 antigen, bone marrow stromal cell antigen 2 (BST-2), or cluster of differentiation 317 (CD317), is a type II transmembrane protein consisting of 180 amino acids. It expresses in certain human bone marrow stromal cell lines, terminally differentiated B cells, multiple myeloma cell lines, and various normal tissues (10,–12). It has been reported that Tetherin distributes on the cell surface or in the intracellular trans-Golgi network (12). Tetherin has been found to directly tether and retain human immunodeficiency virus, type 1 (HIV-1) virions on the cell surface, facilitate their endocytosis, and block their release (13, 14). Thereafter, Tetherin has become an extensively studied host restriction factor of retroviruses (15). In addition, the endogenously expressed Tetherin in tumor cells acts as a ligand of the human plasmacytoid dendritic cell-specific receptor ILT7 and suppresses the antiviral innate immune responses mediated by plasmacytoid dendritic cells (16). A systematical RNAi study has suggested that some autophagy regulatory proteins are required for HIV infection (17). More recently, extensive studies have indicated that Tetherin is degraded in lysosomes (18,–20). Although it has been proposed that Tetherin has arisen in mammals primarily, and likely exclusively, as a directly acting innate protection against viral disease (21), its wide distribution suggests that it may have roles in some basic cellular processes in addition to its viral restriction function. Involvement of autophagy in the host defense to viral infection through xenophagy (22,–24) and identification of Tetherin as an interactive partner of LRPPRC in our early yeast two-hybrid screening prompted us to investigate the roles of Tetherin in autophagy.
Antibodies against human LC3 (catalog no. NB 100-2331) and Tetherin (catalog no. H00000684-D01) were purchased from Novus Biologicals. The IgG control antibodies from mouse (catalog no. SC-2025) and rabbit (catalog no. SC-2027); primary antibodies against β-tubulin (catalog no. SC-9104), β-actin (catalog no. SC-47778), LRPPRC (mouse, catalog no. SC166178; rabbit, catalog no. SC-66845), P27 (catalog no. SC-776), Beclin 1 (catalog no. SC-11427), and GFP (catalog no. SC-8334); and random sequence control siRNA (catalog no. sc-44234) and siRNA specific to Tetherin (catalog no. sc-60294) were from Santa Cruz Biotechnology, Inc. Antibodies against Bcl-2 (catalog no. 2870) and PI3KCIII (catalog no. 4263) were from Cell Signaling Technology. Antibodies against lysosome-associated membrane glycoprotein 2 (LAMP2, catalog no. ab18528) and Tetherin (catalog no. ab88523) were from Abcam. The antibody against P62 (SQSTN1, catalog no. BWL-PW9860) was from Enzo Life Sciences International Inc. The antibody against the mitochondrial 20-kD outer membrane protein (Tom20, catalog no. 612278) was from BD Transduction Laboratories. HRP-conjugated secondary antibodies against mouse (catalog no. 172-1011) and rabbit (catalog no. 172-1019) were from Bio-Rad. Rhodamine Red-X goat anti-mouse IgG and Alexa Fluor 633 goat anti-rabbit IgG (catalog nos. R6393 and A-21070, respectively), Lipofectamine 2000 (catalog no. 11668-027), and Oligofectamine (catalog no. 12252-011) were from Invitrogen. RFP-LC3 was supplied by Dr. Mizushima (25). The plasmid pCDNA3 vector control and pCDNA3 with the HA-Tetherin insert were constructed as described in our previous report (26). The plasmid carrying GFP-Bcl-2 (plasmid 17999) was purchased from Addgene. MG-132, Bafilomycin A1 (BAF), and carbonyl cyanide m-chlorophenylhydrazone (CCCP) were from Sigma. The protein G beads were from Amersham Biosciences.
HeLa cells, HeLa cells stably expressing RFP-LC3, or 293T cells were established and cultured as described previously (8, 9). Cells were transfected with random or Tetherin-specific siRNA molecules packed with Oligofectamine or plasmids carrying HA-Tetherin and/or GFP-Bcl-2 packed in Lipofectamine 2000 in a similar way as we described previously (8, 9). Cell lysates were prepared from attached cells, and coimmunoprecipitation was performed as described previously (8, 9). Cell lysates with the same amounts of total proteins were subjected to immunoprecipitation with equal amounts of specific antibodies and control antibodies (IgG) from the same species.
Immunofluorescence staining and mitochondrial tracking were performed, and images were captured with a laser-scanning microscope in a manner similar to the one we described previously (8, 9). HeLa cells or HeLa cells stably expressing RFP-LC3 were grown in 6-well culture plates. They were transfected with Tetherin-specific siRNA or a plasmid carrying HA-Tetherin. Cells were then left untreated or treated with CCCP in the absence or presence of the lysosomal inhibitor BAF, fixed, and processed for confocal fluorescence microscopy. The number of RFP-LC3 punctate foci was analyzed using the ImageJ program. The standard protocol to stain intracellular proteins included a step to permeabilize the cell membranes with 0.1% Triton X-100 before fixation (8, 9). This step was skipped to label cell surface-distributed Tetherin only.
We used the mitochondria isolation kit for cultured cells (Pierce, catalog no. 89874) from Thermo Fisher Scientific. Mitochondria were prepared from HeLa cells as guided by the associated instructions. Nuclei and cell debris were removed after low-speed centrifugation at 700 × g for 10 min, and the mitochondria and cytosol were separated from the supernatant after centrifugation at 12,000 × g for 15 min. The mitochondrial pellets were resuspended in buffer to the same volumes of cytosolic fraction. The mitochondrial purity was monitored by immunoblot analysis with antibodies against LRPPRC and Tom20.
The yeast two-hybrid screening system has been described in detail in our previous publication (27). Bait was subcloned into the yeast plasmid pGBKT7 (Clontech). A plasmid carrying different bait was transformed into yeast strain AH109 separately. After the expression of bait was confirmed and the transcriptional activation and cell toxicity caused by bait were eliminated, transformed yeast AH109 cells were mated with yeast strain Y187 pretransformed with the human liver cDNA library to screen transformants. The library plasmid DNA was isolated from positive yeast colonies, recovered from Escherichia coli, and sequenced. Positive clones expressing proteins with the right reading frame were cotransformed with their respective baits into AH109 cells. The strength of the interaction was quantified by β-galactosidase activity as directed by the protocols in the associated Clontech manual.
When HeLa cells were treated with either the proteasomal inhibitor MG-132 and/or the lysosomal inhibitor BAF and stained with a Tetherin-specific antibody, we observed a dramatic increase of Tetherin signals in the presence of either inhibitor and a synergistic increase in the presence of both inhibitors (Fig. 1A). The same trends were confirmed by immunoblotting (Fig. 1, B and C). The increased Tetherin signals were distributed mainly inside cells, and the strength of cell surface-distributed Tetherin signals was not changed in the presence of either MG-132 or BAF but enhanced dramatically in the presence of both MG-132 and BAF (Fig. 1A).
To investigate the stability of Tetherin, we treated cells with the protein synthesis inhibitor cycloheximide to stop protein synthesis and then observed the rate of degradation of Tetherin. Because the inhibitors MG-132 and BAF had not been included in the culture medium to exert any impact, the initial levels of Tetherin at 0 h for different treatments were similar, as expected (Fig. 1D). The half-life of Tetherin was about 12 h. Inhibition of proteasomal activity with MG-132 increased the half-life of Tetherin to about 20 h, whereas inhibition of lysosomal activity with BAF or inhibition of both proteasomal and lysosomal activities with a combination of BAF and MG-132 increased the half-life of Tetherin to more than 30 h. Therefore, the stability of Tetherin was enhanced dramatically when either proteasomal or lysosomal activity was blocked (Fig. 1, D and E). Tetherin was colocalized with RFP-LC3 punctate foci and LAMP2-labeled lysosomes in the presence of BAF (Fig. 1, F and G). Under mitophagic stress, the majority of Tetherin is degraded through the autophagy machinery, which possibly includes macroautophagy and the recently characterized ESCRT (endosomal sorting complex required for transport)-mediated endosomal microautophagy (28).
To test whether Tetherin plays any role in the regulation of autophagy, we examined the relationship of levels of Tetherin with autophagy markers. We suppressed the expression of Tetherin with a specific siRNA in HeLa cells and found that the levels of LC3-II were reduced significantly but that levels of P62 were not changed in either the absence or presence of BAF (Fig. 2, A–C). We repeated the experiment in HeLa cells stably expressing RFP-LC3 and found that the numbers of RFP-LC3 punctate foci were similarly reduced (Fig. 2, D and E). Immunoblot analysis of the cells stably expressing RFP-LC3 revealed that not only the levels of both exogenous RFP-LC3-II and endogenous LC3-II but also the levels of P62 were reduced significantly in Tetherin-suppressed cells (Fig. 2, F and G). The reduction in the levels of endogenous LC3-II and P62 in HeLa cells stably expressing RFP-LC3 caused by Tetherin silencing was repeated with 293T cells (Fig. 2, H–J).
To observe the impact of Tetherin overexpression, we transfected HeLa cells stably expressing RFP-LC3 with a plasmid encoding HA-tagged Tetherin. We observed a significant increase in the number of RFP-LC3 punctate foci in the absence or presence of BAF (Fig. 3, A and B). Further immunoblot analyses demonstrated that levels of exogenous RFP-LC3-II and endogenous LC3-II and P62 were increased similarly (Fig. 3, C and D). The same impact of Tetherin overexpression on the levels of endogenous LC3-II and P62 was observed in 293T cells overexpressing HA-Tetherin (Fig. 3, E and G).
Surprisingly, neither the level of P62 nor the level of LC3-II was altered when HA-Tetherin was overexpressed in native HeLa cells (Fig. 3, H and I). Compared with HeLa cells stably expressing RFP-LC3, native HeLa cells expressed lower levels of Bcl-2, higher levels of P62, and similar levels of Beclin 1, PI3KCIII, and P27 (Fig. 3J). We reasoned that the two types of cells had different levels of Bcl-2 and then different levels of autophagy flux. Native HeLa cells had higher levels of autophagy flux so that the inhibition imposed by the depletion of Tetherin was obvious, whereas the further enhancement imposed by the overexpression of Tetherin was not easily detected. P62 proteins exists in autophagosomes and aggresomes that have not been packaged in autophagosomes (29). The high levels of P62 in native HeLa cells may suggest an accumulation of a large amount of aggresomes. Therefore, the amount of P62-labeled aggregates influenced by Tetherin-triggered activation or inhibition of autophagy may only occupy a small fraction of the total P62 levels, and the resulting variation may be insignificant. We overexpressed GFP-Bcl-2 in native HeLa cells to suppress the initiation of autophagy and found that levels of LC3-II were reduced as expected. Although the exact mechanism is still under investigation, the levels of P62 were also decreased coincidentally (Fig. 3K). The enhancement effect of Tetherin on autophagy, as indicated by increasing levels of LC3-II and P62, was recovered upon overexpression of Bcl-2 (Fig. 3, L and M). Therefore, Tetherin is confirmed to be an activator of autophagy flux.
When mitophagy was induced with CCCP, a dramatic decrease in the levels of Tetherin was observed (Fig. 4A). We found that a significant amount of Tetherin was colocalized with autophagosome-contained mitochondria when lysosomal activity was blocked (Fig. 4, B and C). Induction of mitophagy with CCCP caused the majority of Tetherin to be associated with mitochondria (Fig. 4D). Therefore, Tetherin probably plays a role in mitophagy.
To understand the mechanism of how Tetherin enhances autophagy flux, we examined the role of Tetherin in the regulation of the initiation of autophagy. On the basis of analyses of the amino acid sequence of full-length human LRPPRC, we designed four fragments of LRPPRC protein as bait for yeast two-hybrid screening (Fig. 5A). After a series of stringent tests, no interactive protein was found when baits 1 and 4 were used to screen a human liver cDNA library (Fig. 5B). As we reported previously (27), five interactive proteins of LRPPRC were found when bait 3 was used to repeat the screening (Fig. 5B). An interaction between Tetherin and LRPPRC was revealed in a yeast two-hybrid complementation system in which a fragment from residue Leu-70 to Gln-180 of the 180 amino acid residue full-length Tetherin was captured by bait 2 that was made of a fragment from residues Leu-407 to Glu-674 of the human LRPPRC (Fig. 5, A–C).
Because LRPPRC has been reported to be exclusively colocalized with mitochondria (8, 9), the mitochondrial association of Tetherin (Fig. 4) made the interaction between Tetherin and LRPPRC physiologically meaningful. We further confirmed that LRPPRC interacted with Tetherin by coimmunoprecipitation in lysates from HeLa cells. Immunoprecipitation with the antibody to LRPPRC led to the coprecipitation of Tetherin, whereas immunoprecipitation of Tetherin with a polyclonal antibody led to precipitation of Tetherin isoforms with different mobility and coprecipitation of LRPPRC (Fig. 5D). We have reported previously that LRPPRC interacts with Bcl-2 and Beclin1 and prevents Beclin 1 from forming a complex with PI3KCIII to activate the initiation of autophagy (8). We examined the relationship among the related proteins and found the Tetherin-LRPPRC complex in addition to two other distinct protein complexes: the LRPPRC·Bcl-2·Beclin 1 and Beclin 1-PI3KCIII complexes (Fig. 5, E and F).
To understand the significance of the Tetherin-LRPPRC interaction in the regulation of autophagy, we suppressed the expression of Tetherin to test the impact on the formation of protein complexes. Suppression of Tetherin did not lead to a change in the levels of LRPPRC, Beclin 1, and Bcl-2 but did cause a reduction in the amount of Tetherin bound to LRPPRC (Fig. 6, A and B). The reduction in the amount of Tetherin bound to LRPPRC led to the release of LRPPRC and formation of a ternary complex of LRPPRC·Beclin 1·Bcl-2 (Fig. 6, A and C–F). Suppression of Tetherin did not alter the levels of PI3KCIII (Fig. 6G) but limited the formation of Beclin 1-PI3KCIII complexes (Fig. 6, G–I).
Similarly, overexpression of Tetherin did not alter the levels of LRPPRC, Beclin 1, Bcl-2, and PI3KCIII (Fig. 6, J and K). However, the overexpression of Tetherin did enhance the binding of Tetherin on LRPPRC (Fig. 6, J and L) and reduced the binding of Beclin 1 and Bcl-2 with LRPPRC (Fig. 6, J, M, and N), of LRPPRC with Bcl-2 (Fig. 6, J and O), of Beclin 1 with Bcl-2 (Fig. 6, J and P), and of Bcl-2 with Beclin 1 (Fig. 6, K and Q). Overexpression of Tetherin did enhance the binding of PI3KCIII with Beclin 1 (Fig. 6, K and R) so that the initiation of autophagy was activated.
On the basis of a chemical carcinogen-induced mouse model of liver cancer, we have suggested previously that autophagy is activated to release the oxidative stresses caused by either carcinogen in the initiative stage or by a metabolic imbalance imposed by genome instability in tumor foci (30, 31). Reactive oxygen species cause telomere attrition and DNA double strand breakage (32, 33) and simultaneously subvert mitotic checkpoints (34, 35). Therefore, genome instability is amplified and eventually leads to tumorigenesis (31). Therefore, autophagy generally plays a tumor-suppressive role.
The endoplasm reticulum-distributed Bcl-2 sequesters Beclin 1, which is distributed in both the endoplasm reticulum and mitochondria, and prevents it from formatting the autophagy initiation complex with PI3KCIII (36). Autophagy-related 14 (ATG14) enhances the Beclin 1-PI3KCIII complex to induce the initiation of autophagy (37). On the basis of the distribution of ATG14, the source of the isolation membrane has been proposed to originate from the mitochondrion-associated endoplasm reticulum membrane (38). Tetherin residing within a “lipid raft” contains a C-terminal glycosylphosphatidylinositol anchor to link with the cytoskeleton or provide a platform for the assembly of a signaling complex (39). LRPPRC associates with Beclin 1, Bcl-2, and Parkin on mitochondria and serves as a checkpoint protein to suppress the initiation of autophagy/mitophagy to verify whether the events of mitophagy are ready to be executed (8, 9). Tetherin, being expressed at enhanced levels in tumor cells (16), interacts with LRPPRC and causes the release of more Beclin 1 to form the autophagy initiation complex nearby to induce autophagy/mitophagy.
We have also reported previously that LRPPRC-suppressed autophagy and mitophagy are associated with a poor prognosis in prostate cancer patients and that prostate cancer with high levels of LRPPRC tends to be more malignant (8, 9, 40). The autophagy activator MAP1S is induced during the development of tumorigenesis, and prostate cancer patients with high levels of MAP1S survive for a longer period than those with low levels of MAP1S (5, 31, 41). Tetherin knockout mice develop their major organ systems normally, and Tetherin becomes a key effector of the antiretroviral activity of type I interferon (21). Similarly, activated autophagy induced by enhanced levels of Tetherin may represent an adaptive mechanism of cancer cells to counteract the tumorigenesis-trigged oxidative stresses and suppress the development of more malignant cancer. Therefore, Tetherin enhances autophagy and mitophagy and may suppress tumorigenesis.
Currently, most of studies of Tetherin focus on its role in viral restriction. Tetherin tethers virus particles on the cell surface and prevents the diffusion of virus particles after budding from infected cells (13, 14). Because cell surface-distributed Tetherin only occupies a small fraction of the total Tetherin molecules in cells, whether the tethering function of Tetherin is its major role or just one of several functions is still unknown. Whether Tetherin binds with HIV-1 particles and helps deliver the particles to autophagosome-like structures for degradation in lysosomes has not been reported, although both viral particles and Tetherin molecules distributed in lysosomes have been reported (42, 43). We know that LRPPRC suppresses autophagy/mitophagy (8, 9). Its depletion may lead to activation of autophagy to enhance the degradation of viral particles so that the HIV infection is attenuated (44). On the basis of endosymbiotic theory, mitochondria may originate from previously free-living bacteria when they were taken inside eukaryotic cells as an endosymbiont (45). We observed that mitochondria associate with Tetherin in autophagosomes upon exposure to an inducer of mitophagy. Therefore, it is logical to predict that Tetherin may also direct virus particles to autophagosomes for degradation in cells, including macrophages, dendritic cells, or some other primary cells. Furthermore, Tetherin depletion may enhance viral replication not only by increasing the release of virions but also by blocking autophagy.
One of the consequences of HIV infection is the depletion of cluster of differentiation 4 (CD4) T cells and further immunodeficiency. Apoptosis accounts for the death of the activated and productively infected CD4 T cells that account for only a small fraction of total CD4 T cells. The remaining 95% of quiescent lymphoid CD4 T cells die by caspase-1-mediated pyroptosis triggered by abortive viral infection (46). Both apoptosis and pyroptosis are triggered by dysfunctional mitochondria resulting from autophagy/mitophagy defects. It has been suggested that the toxicity of viral infection to host cells is caused by mitochondrial dysfunction and apoptosis mediated by virus-encoded proteins such as HIV-1 viral protein R (Vpr) and hepatitis B virus X protein (HBx) (47). We hypothesize that viral infection may also cause cell death, mainly by overloading the autophagy machinery and interfering with the normal turnover of dysfunctional mitochondria. Infection with HIV-1 causes the ubiquitination of Tetherin that is activated by HIV-1 viral protein U (Vpu) and its degradation in lysosomes (48). The degradation of Tetherin may lead to the stabilization of the LRPPRC·Beclin 1·Bcl-2 complex and the suppression of autophagy and mitophagy. If the autophagy process is blocked before autophagosomal formation, the fragmented mitochondria will release cytochrome c and other small molecules to induce apoptosis (49,–51). If autophagosomes are not degraded efficiently, the accumulated mitochondria may become damaged by their own production of superoxide, start to leak electrons, lose their membrane potentials, and induce even further robust oxidative stress (52). Oxidative stress, in turn, activates the NLRP3 inflammasome that leads to proinflammatory cytokine maturation and pyroptotic cell death (53, 54). The suppression of autophagy because of viral overloading may be the true etiology of human HIV infection and AIDS.
We thank Dr. Noboru Mizushima (Department of Physiology and Cell Biology, Tokyo Medical and Dental University Graduate School) for the gift of LC3 cDNA.
*This work was supported, in whole or in part, by NCI/National Institutes of Health Grant CA142862 (to L. L.) and NIAID/National Institutes of Health Grant AI099007 (to J. T. K.). This work was also supported by Department of Defense New Investigator Award W81XWH (to L. L.) and by Baylor College of Medicine-University of Texas Houston Medical School Center for AIDS Research Grant AI036211 (to J. T. K.).
2The abbreviations used are: