In this report, the mechanisms by which the Ad RID complex downregulates the levels of cell surface TNFR1 are described. It is known that the tyrosine sorting motif of receptors can interact with the μ2 subunit of AP2 directly to mediate receptor endocytosis (52
). In addition, previous reports suggested that the tyrosine sorting motif in RIDβ was important in downregulating FAS, TRAIL-R, and EGFR (29
). We therefore produced RIDβ mutants in which 122
Y was mutated to either alanine (A) or phenylalanine (F) to determine if the tyrosine sorting motif in RIDβ is also critical for TNFR1 downregulation. The abrogation of TNFR1 downregulation by RIDβ mutants suggested that the tyrosine sorting motif in RIDβ played an important role in TNFR1 downregulation and 122
F could not substitute for 122
Y as a functional transport motif in RIDβ. The results obtained with the RIDβ(YA) mutant also support early reports that the tyrosine sorting motif in RIDβ was critical for downregulation of FAS (29
). Interestingly, the observation that 122
F in RIDβ also abolished FAS downregulation differs from a previous study in which the authors found that the RIDβ(YF) mutant was able to downregulate FAS (41
). The discrepancy between the two studies could be explained by the different cell types or assays that were used. In our study, RIDβ(YF) mutants of the Ad2 subtype were expressed in an adenoviral vector; in this case, there was a potential interaction between RID and other adenoviral proteins. In addition, we used FACS analysis to observe the endogenous surface level of FAS in 293 and HeLa cells. In contrast, in the previous study, the RIDβ(YF) mutant was from an Ad5 subtype expressed from plasmids and immunofluorescence microscopy was used to study the surface level of overexpressed FAS in COS7 and A549 cells.
Our observations that RIDβ mutants did not affect surface levels of FAS or affect the WT RID-mediated FAS downregulation in mixing experiments corroborated the conclusions of a previous study (29
). In that study, the authors showed that there was a three- to fourfold increase in surface expression of the RIDβ(YA) mutant compared to WT RID. Thus, it is likely that RID affects the FAS pathway intracellularly, such as in the early endosomes. Other data, however, favored the model in which RID downregulated FAS by acting on the plasma membrane, since the RIDα mutants which did not localize to the cell surface were defective for FAS downregulation (68
). Surprisingly, in contrast to FAS, we showed that the surface TNFR1 levels increased by 170 to 220% in the presence of RIDβ mutants. In addition, WT RID-mediated downregulation of TNFR1 was partially inhibited by the RIDβ mutants. These differential effects not only suggest that there are mechanistic differences in the downregulation of FAS and that of TNFR1 by RID but also may indicate that the interaction of TNFR1 and RID in the same complex, as shown by our data, probably occurs on the cell surface. Whereas WT RID downregulates surface TNFR1, the RIDβ mutants may inhibit the normal internalization of TNFR1, thereby increasing its surface expression.
Our present data cannot distinguish between a direct protein-protein interaction between TNFR1 and RIDβ, in contrast to both existing in a larger complex of proteins held together by an unidentified bridging molecule. Previously there had been no experimental evidence that RID physically associates with any of its targeted receptors on the plasma membrane. Indeed, the only protein interaction data were from Crooks et al. (10
), who showed an association of RIDα with EGFR in the early endosomes. Together with the evidence that the endocytosis rate of EGFR was not accelerated by RID (31
), it was proposed that RID downregulates EGFR by rerouting the constitutively recycling receptors to the lysosomes, where they are degraded. Therefore, it is unlikely that RID interacts with EGFR on the cell surface. Here we demonstrated for the first time an association between RID and a member of the death receptor family, TNFR1, by co-IP. Interestingly, the RIDβ mutants which cannot interact with μ2 of AP2 (29
) still associate with the TNFR1, suggesting that the tyrosine sorting motif in RIDβ does not mediate the complex formation of TNFR1 and RID. Together with the effects of RIDβ mutants on increasing cell surface levels of TNFR1, we propose a model in which RID associates with TNFR1 on the cell surface (Fig. ), whereas RID and FAS are likely to associate in an intracellular organelle such as an endosome, in which the targeted receptors proceed to the lysosomes and are degraded, while RID recycles to the cell surface with the help of the dileucine motif in RIDα (29
). It is noteworthy that surface levels of TRAIL-R1 were shown to be increased by the RIDβ(YA) mutant; therefore, in addition to TNFR1, RID may also associate with TRAIL-R1 on the cell surface (29
). The concept of endosomal/lysosomal degradation of FAS and TNFR1 agrees with observations by others and data from our study that degradation of FAS and TNFR1 by RID was inhibited in the presence of Bal A1 or NH4
). Moreover, RID-mediated TNFR1 degradation was not inhibited by the proteasomal inhibitor MG115 (data not shown), further suggesting that TNFR1 degradation by RID is via an endosomal/lysosomal pathway.
FIG. 8. Model of RID-mediated TNFR1 and FAS downregulation. The RIDβ(YA) and RIDβ(YF) mutants (as shown on the right) increased surface levels of TNFR1 but not FAS. In addition, the observations of others have shown that surface levels of RIDβ (more ...)
Although we have previously shown that sucrose inhibited RID-mediated TNFR1 downregulation (17
), there are reports suggesting that sucrose may also be involved in the clathrin-independent endocytotic process (53
). We therefore used additional methods to verify the role of clathrin in RID-mediated TNFR1 downregulation. By transfecting 293 cells with a dominant negative form of dynamin (K44A), we demonstrated that TNFR1 downregulation by RID is dynamin dependent. Although it is generally agreed that dynamin is involved in scission of clathrin-coated vesicles, it has been shown in certain cell lines dynamin also plays a role in caveolin-mediated receptor endocytosis (26
). Since clathrin-mediated receptor endocytosis utilized AP2 to recruit targeted receptors into the coated pit, by demonstrating siRNA against the μ2 subunit of AP2 blocked RID-mediated TNFR1 downregulation, our data provide strong evidence that TNFR1 downregulation by RID is via an AP2-, clathrin-dependent process. Although RID has been shown to interact with both AP1 and AP2 (29
), which mediate the trafficking of cargo proteins from the Golgi apparatus and the plasma membrane, respectively, to the endosome (36
), our data are the first to show that the μ2 subunit of AP2 is functionally important in the downregulation of cell surface receptors by RID. The role of AP2 in the RID-mediated downregulation of other receptors remains to be determined.
Thus far, four different types of TNFR1 turnover mechanism from the plasma membrane have been described. Two well-known processes include clathrin-mediated endocytosis and proteolytic shedding (51
). Recently, two new mechanisms, caveola-mediated endocytosis in endothelial cells (11
) and exosome-like vesicle release (25
), have been identified. In addition to studying the role of clathrin in RID-mediated TNFR1 downregulation, we also tested if RID downregulated surface TNFR1 by increasing exocytosis of TNFR1. However, our data demonstrated that RID reduced the amount of TNFR1 in exosome-like vesicles (data not shown), thus excluding this pathway in the downregulation of TNFR1 by RID.
In order to understand why RID specifically downregulates certain cell surface receptors but not others, and to further dissect the different interactions between RID and its target receptors, it is important to identify the subcellular organelles where RID and various receptors interact, as well as to distinguish if RID interacts with different receptors directly or through a common adaptor molecule. Since RID is able to downregulate receptors of different families with little homology, it appears more likely that there are common molecules bridging RID and its target receptors. Although the μ2 subunit of AP2 was shown to interact with RID (29
) and a functional tyrosine sorting motif was identified in TNFR1 (54
), it is unlikely that μ2 acts as the bridging molecule between RID and TNFR1, based on our finding that the RIDβ mutants, which cannot bind to μ2, still associate with TNFR1. Instead, we speculate that RID acts as a connector between TNFR1 and μ2 to provide an alternative or complementary mechanism to link TNFR1 to the components of clathrin-coated pits (CCP). In this regard, RID may function like the human immunodeficiency virus nef protein, which has been shown to downregulate CD4 from the cell surface by acting as a molecular connector of CD4 and μ2 (50
). In fact, RID may be a functional homolog of cellular molecules which act as linkers between cell surface receptors and CCP, such as Eps15 and Shc, which are bridging molecules between EGFR and CCP (46
). Nonetheless, experiments are under way to examine if TNFR1 and RID interact directly or if there is yet another molecule bridging this interaction.
In summary, our genetic and functional studies indicate that RID downregulates TNFR1 via an AP2-, clathrin-mediated pathway, followed by degradation of TNFR1 in the endosomes/lysosomes. Experiments with our RIDβ mutants suggest that RID downregulates TNFR1 and FAS through distinct mechanisms. To our knowledge, this is the first study providing evidence for a model in which RID downregulates surface TNFR1 by associating with it, most likely on the cell surface. The ability of RID to coordinately downregulate different death receptors, including TNFR1, FAS, and TRAIL-R, may result in enhancement of Ad replication and its persistence in the host. The potential of utilizing individual Ad E3 proteins as therapeutic agents to enhance transgene persistence with the advantage of organ-specific immunosuppression underscores the importance of dissecting the precise mechanism of downregulation of TNFR1, as well as other receptors, by RID.