In this study, we demonstrate that RID-α is a potent regulator of intracellular cholesterol in gene transfer experiments as well as during acute Ad infections. Our data indicate that RID-α facilitates two distinct steps in cholesterol trafficking: egress from endosomes and transport to the ER necessary for homeostatic gene regulation (). RID-α rescues the cholesterol storage phenotype in NPC fibroblasts, suggesting that the viral protein regulates a molecular mechanism that operates independent of NPC1/NPC2. This supposition is further supported by evidence obtained with a palmitoylation-defective RID-α (C67S) mutant, which alters LE morphology to produce NPC-like LSOs and deregulates ER homeostatic mechanisms in CHO cells with functional NPC proteins (Cruz et al., 2000
; Naureckiene et al., 2000
). Even though the viral protein binds GTP-Rab7 effectors (Shah et al., 2007
), RID-α is not present in LAMP1-positive LEs but influences endosome function at a distance from a membrane compartment with characteristics of autophagic vesicles. Furthermore, RID-α is excluded from cholesterol-rich DRMs, suggesting that its function is not restricted by NPC mutations, in contrast to Rab7 which is inhibited by high levels of cholesterol (Kobayashi et al., 1998
). The RID-α (C67S) mutant also abrogates the cellular phenotype associated with a dominant-negative form of Rab7, providing further evidence that RID-α and GTP-Rab7 are not necessarily interchangeable. A previous study indicated that newly synthesized RID-α is sorted at the level of TGN via AP-1 clathrin adapters (Cianciola et al., 2007
), and the working model in suggests that RID-α regulates cholesterol trafficking in endosomes after TGN exit. A similar mechanism has been reported in T cells of the adaptive immune system, where antigen-loading compartments for major histocompatibility complex class II molecules receive continuous input from autophagosomes (Schmid et al., 2007
). However, we cannot exclude the possibility that RID-α facilitates retrograde cholesterol transport to the ER via a TGN intermediate (Urano et al., 2008
). The presence of β-COP in RID-α compartments is consistent with either role because COP1 coatomer regulates endocytosis and autophagy as well as Golgi to ER retrograde transport (McMahon and Mills, 2004
). It is also conceivable that RID-α regulates transport in both directions depending on sterol load and cell physiology.
Figure 8. Model of the role of RID-α as a coordinator of endosome trafficking. LDL-cholesterol (Chol) esters are internalized via the PM LDLR (1a), unesterified, and trafficked to MVBs/LEs (1b). Cholesterol is egressed out of these organelles by the coordinated (more ...)
Our results indicate the RID-α–containing membranes do not correspond to any well-defined intracellular organelle. Although the origin of these compartments remains unclear, it is possible that RID-α arrests maturation of a physiologically short-lived early autophagic compartment. Alternatively RID-α could create a virus-specific compartment by remodeling intracellular membranes. Our data favor the first possibility because RID-α vesicles are enriched in LC3 recruited to nascent autophagic membranes but devoid of LAMP1 found in mature autophagosomes. This is not the first example of a pathogen-hijacking host autophagic machinery to carry out a novel function, as poliovirus triggers formation of unique LC3-positive compartments (Taylor and Kirkegaard, 2007
). However, in contrast to poliovirus, which utilizes virally induced autophagosomal membranes to facilitate viral RNA replication, RID-α regulates innate immune responses (Ginsberg and Prince, 1994
; Stewart et al., 2007
). We have also demonstrated that RID-α promotes endosome to lysosome transport of select membrane cargo at the expense of enhanced autophagic flux in acutely infected cells. It is conceivable that RID-α compartments sequester key proteins required for autophagosome–lysosome fusion or alternatively provide a surplus of rate-limiting molecules required for efficient endocytosis. Evidence that RID-α regulates endosome sterol balance independent of NPC1/NPC2 favors the latter hypothesis. However, it is conceivable that RID-α integrates endocytosis and autophagy by a combination of these two mechanisms. The finding that RID-α function requires class III PI3K activity necessary for trafficking in both pathways (Backer, 2008
) also supports a coordinating role for RID-α.
Another molecule enriched in RID-α compartments is ORP1L, a member of the family of oxysterol-binding proteins implicated in a variety of cellular functions (Olkkonen et al., 2006
). ORP1L is a GTP-Rab7 effector linked to LE MT-dependent motility and lipid trafficking that also binds RID-α (Johansson et al., 2005
; Shah et al., 2007
). ORP1L is comprised of multiple amino-terminal ankyrin repeats, a pleckstrin homology domain, and a carboxyl-terminal oxysterol regulatory domain capable of binding cholesterol and 25-HC in vitro (Suchanek et al., 2007
). Contrary to GTP-Rab7, which recognizes ORP1L ankyrin repeats, RID-α interacts with oxysterol regulatory domain sequences (Shah et al., 2007
), suggesting that it may directly regulate sterol binding. Although palmitoylation does not alter RID-α membrane partitioning or intracellular compartmentalization, this modification does influence RID-α function presumably by regulating conformation of the RID-α carboxyl tail and its ability to interact with protein-binding partners (Charollais and Van Der Goot, 2009
). Thus, failure of RID-α to undergo palmitoylation may sequester ORP1L and/or inhibit its sterol-binding properties, leading to deregulated cholesterol homeostasis. RID-α also binds a second GTP-Rab7 effector, RILP, and this interaction is required for RID-α biological activity in Ad2-infected cells (Shah et al., 2007
). It will be interesting in the future to determine whether RILP is also recruited to RID-α–induced compartments because RILP binds ESCRT components involved in the dual regulation of endocytosis and autophagy independent of GTP-Rab7 (Progida et al., 2006
; Wang and Hong, 2006
; Fader et al., 2008
In contrast to CHO cells and normal fibroblasts, RID-α compartments in NPC fibroblasts are also enriched for LBPA, which is known to regulate cholesterol clearance from endosomes (Chevallier et al., 2008
). The NPC phenotype is partially reversed by the addition of exogenous LBPA, suggesting that LBPA is a limiting factor contributing to disease pathology (Chevallier et al., 2008
). Therefore, it is possible that RID-α restores NPC cholesterol trafficking by up-regulating a basal mechanism involved in LBPA transport to the endocytic system. The finding that LBPA is only detectable in RID-α compartments in cells with inherent cholesterol imbalance reveals the plasticity of RID-α–based mechanisms.
Although its role in controlling innate immune responses is well known, our experiments have identified at least three new RID-α–associated functions during acute Ad infections. First, RID-α subverts virus-induced autophagy. Although the molecular basis for Ad2-induced autophagy is currently unknown, it is conceivable that this pathway is activated to remove endosomes damaged by membrane lysis immediately after Ad internalization (). Ad-induced endosome membrane lysis may also deplete a critical rate-limiting factor sufficient to disrupt trafficking throughout the endocytic pathway that is reintroduced to the system via a RID-α–dependent mechanism (). Second, in contrast to receptors involved in innate immune responses, RID-α may divert MPRs to the PM, where they fulfill distinct functions compared with their normal role in shuttling acid hydrolases to lysosomes (Ghosh et al., 2003
). Third, RID-α regulates endosome cholesterol egress in acutely infected cells. Cholesterol is required for Ad2 internalization from the PM and endosome escape, suggesting that Ads may be taken up to specialized cholesterol-enriched domains that execute membrane lysis (Imelli et al., 2004
). Therefore, it is conceivable that RID-α restores the cholesterol balance that is perturbed during the early stages of an acute Ad infection. RID-α–induced cholesterol trafficking may also be important in latent infections in which E3 proteins are thought to have a prominent role (McNees et al., 2002
). However, in contrast to NPC1/NPC2, RID-α selectively regulates cholesterol trafficking to the ER and consequently could trigger cholesterol imbalance in other intracellular compartments during an acute infection.
This study highlights several new areas for future investigation. First, given its capacity to reconstitute cholesterol trafficking in cells with defective NPC proteins, it will now be of interest to fully characterize the newly identified RID-α–induced pathway. Second, this study raises the possibility that the NPC1/NPC2 machinery is impaired during acute Ad infections. Third, RID-α–containing vesicles may shed new light on the intracellular membrane origin of autophagic vesicles.