Cell biology of retroviral RNA packaging

Abstract

Generation of infectious retroviral particles rely on the targeting of all structural components to the correct cellular sites at the correct time. Gag, the main structural protein, orchestrates the assembly process and the mechanisms that trigger its targeting to assembly sites are well described. Gag is also responsible for the packaging of the viral genome and the molecular details of the Gag/RNA interaction are well characterized. Until recently, much less was understood about the cell biology of retrovirus RNA packaging. However, novel biochemical and live-cell microscopic approaches have identified where in the cell the initial events of genome recognition by Gag occur. These recent developments have shed light on the role played by the viral genome during virion assembly. Other central issues of the cell biology of RNA packaging, such as how the Gag-RNA complex traffics through the cytoplasm toward assembly sites, await characterization.

Introduction

Early after infection, retrovirus RNA genomes are reverse transcribed into DNA via viral reverse transcriptase and integrated in the host genome by the viral integrase. Then, the cellular transcription machinery produces a single full-length precursor RNA (gRNA) that ensures three functions: it serves as pre-RNA for the production of spliced viral mRNAs, it serves as mRNA for the synthesis of Gag and Gag/Pol polyproteins, and it also acts as the viral genome, which will be specifically packaged into assembling viral particles.

Assembly of retroviruses is a highly regulated process that requires the viral structural components, as well as some key cellular factors, to converge on the correct cellular site at the correct time. Among retroviruses, two major pathways of assembly can be distinguished: B/D-type retroviruses, such as Mason-Pfizer monkey virus (MPMV), assemble immature procapsids at a pericentriolar location.¹ These procapsids are subsequently transported to the plasma membrane where budding occurs. C-type retroviruses, such as the lentivirus human immunodeficiency virus (HIV-1) and the gammaretroviruses murine leukemia virus (MLV), assemble and bud from an identical site on cellular membranes. However, over the past few years, the nature of these membranes has been a matter of debate, which we discuss in this review.

The retroviral genome contains three major structural genes: gag, pol and env. Fully assembled retroviral virions are composed of few thousand Gag proteins,² 1/20th of which are Gag-Pol pro-teins,³ 8 to 10 Envelope trimers^4,5 and two copies of the viral genome. The polyprotein Gag orchestrates the assembly process; its expression alone is sufficient to generate defective viral-like-particles (VLPs) that are morphologically indistinguishable from the ∼100 nm diameter immature virions produced by infected cells.⁶ Gag contains four domains that become cleaved from each other upon the action of the viral protease; a process that happens during or shortly after assembly and leads to the formation of mature particles.⁷ The N-terminal matrix (MA) domain targets Gag to the inner leaflet of the plasma membrane, the capsid (CA) domain drives Gag multimerization, the NC domain packages viral genomes and the C-terminal p6 region recruits members of the ESCRT complexes for particle release.⁷ In addition to acting as the major structural protein, Gag mediates the selective packaging of the dimeric RNA genome. Our understanding of when and where Gag encounters and recruits the genomic RNA have recently improved, due in large part to microscopic approaches in live cells. These central questions are the focus of this review.

Genome Recognition by Gag

Selection of the genomic RNA from a cytoplasmic pool that contains a substantial excess of non-viral and spliced viral RNAs is mediated by direct interaction between the NC domain of Gag (GagNC) and highly structured sequences called Psi, located in the 5′ untranslated region (5′UTR) of the gRNA (reviewed in ref. 8–10).

Conformations of the 5′UTR of HIV-1 and MLV have been well characterized and RNA mapping studies have identified the determinants for RNA packaging. The MLV Psi region forms an independent and highly structured domain including four stem-loop structures (called A–D) (Fig. 1A). The A and B stem-loops promote gRNA dimerization while C and D motifs form the core encapsidation signal.^11–17 Similarly, the Psi region of HIV folds into four stem-loops, called SL1 to SL4^18–20 (Fig. 1B). The major packaging signal includes SL1, which contains the dimerization initiation site (DIS), and SL3 (reviewed in ref. 8–10). For HIV-1 and MLV the major packaging determinants are located downstream of the 5′ splice donor site that generates all the viral spliced mRNAs²¹ (Fig. 1). This could explain, at least in part, the preferential packaging of the unspliced gRNA. Nevertheless, viral spliced RNAs are also selectively packaged in infectious particles and subsequently reverse-transcribed in infected cells.^22–27 Although the molecular mechanisms that govern RNA packaging selectivity remain unclear, initial evidence has shown that HIV-1 incorporates the gRNA and the spliced mRNA species in a competitive manner.²⁴

Figure 1.

Map of HIV-1 and MLV genomes. Schematic presentation of the RNA secondary structures of the Psi region are given for HIV-1 (A) and MLV (B). The stem-loops forming the packaging Psi region are depicted as well as the major splice-donor site (SD), the gag initiation codon, the 5′ cap, the unstranslated regions (UTR) and the polyadenylation site (Aⁿ).

The selective recruitment of the gRNA is at least in part the result of the high affinity binding of Psi sequences for GagNC protein. Abundant studies have focused on analysis of the GagNC binding affinity to the 5′UTR of gRNA. GagNC contains 2 highly conserved ‘CCHC’ zinc fingers (ZFs), with the exception of MLV, which has only a single ZF domain (Fig. 2). These ZFs confer to the NC potent nucleic acid binding and annealing activities. Moreover, the basic residues flanking the ZFs, also contribute to the nucleic acid chaperone activity of NC.^28,29 Deletion of the NC domain of Gag results in the production of viruses defective in Gag-Gag multimerization, virus budding and particle morphogenesis (reviewed in ref. 30). Mutation of the functional domains of NC has shown that the ZFs and basic residues of the HIV and MLV GagNC are involved in gRNA selection and dimerization through tight interactions with the Psi packaging signal.^12,31–38

Figure 2.

HIV-1 and MLV nucleocapsid proteins. The NC domain is located at the C-terminus of the Gag polyprotein. The amino acid sequences of are shown in one-letter code. The zinc fingers (ZFs) with the C-C-H-C consensus sequence are depicted with the central zinc atom (Zn). The basic regions that also contribute to the RNA packaging, are shown in gray boxes.

Strikingly, loosening the interactions between GagNC and the gRNA by mutating either the NC ZFs or its 5′ basic residues, not only reduces the gRNA content of mutant virus but turns HIV-1 into a DNA-containing virus.^38–40 These mutant HIVs activate viral DNA synthesis by the viral reverse transcriptase (called “late RT”) in HIV-1 producer cells before virus release leading to the packaging of viral DNA (up to 100-fold increase of DNA level as compared with wild-type virions).^38,39,41 These findings raise new questions about the control of the interactions between viral components at assembly sites; a process that appears to be more complex than previously thought.

C-Type Retroviruses Assemble at the Plasma Membrane

Until recently, much confusion and controversy existed as to where within the cell retroviral assembly was initiated and completed. Based on electron microscopy or immunofluorescent detection of Gag and budding virions at the surface of infected cells,⁴² it was historically thought that retroviral assembly took place at the plasma membrane. However, later studies have demonstrated that considerable amounts of HIV-1 Gag proteins and virions could be detected in cellular compartments containing late endosomal markers, such as CD63 and lysosome-associated membrane protein-1 (Lamp1). This was true both in standard tissue culture cell lines, as well as in T-lymphocyte cell lines and macrophages, which are physiological targets of HIV-1 infection.^43–50 Consistent with these microscopic studies, extracellular macrophage-derived virions could be immunoprecipitated with CD63 antibodies.^45,51 Moreover, these observations were extended to several other retroviruses, such as MLV and human T-cell leukemia virus type-1 (HTLV),^52–54 suggesting that assembly in late endosomes is a general feature of retroviruses. Altogether, these data were supporting a model implying that retroviral Gag initiates assembly on endosomal membranes in all cell types and that assembled virions are subsequently transported to the cell surface within late endosomes.

Later, the ‘late endosomal assembly’ model was challenged by several studies conducted with HIV-1, both in tissue culture cell lines and in macrophages. First, ultra-structural electron-microscopy studies, performed in HIV-1 infected macrophages, have shown that CD63-positive invaginations of the plasma membrane can be easily mistaken for intracellular virus-containing compartments.⁵⁵ Second, following Gag localization over time, either by immunofluoresence approaches⁵⁶ or by subcellular fractionation coupled to pulse-chase labeling,⁵⁷ revealed that Gag is targeted to the plasma membrane early after expression and is found in late endosomes only at later time points. Third, abolishing late endosome motility within cells using 2 pharmacological inhibitors with distinct mechanisms of action does not affect the amount of released particles.⁵⁶ Finally, targeting of Gag proteins to late endosomes by fusing them to a specific late endosome membrane binding domain failed to produce extracellular particles, in contrast to Gag proteins that were fused to a plasma membrane targeting domain.⁵⁶ Taken together, these studies constitute solid evidence that HIV-1 assembly occurs at the plasma membrane (Fig. 3) and that a subpopulation of virions can be subsequently internalized by endocytosis. Endosomal MLV virions can also arise via endocytosis from the cell surface,⁵⁸ extending the plasma membrane assembly notion to other retroviruses. This model is entirely consistent with previous observations of virions in late endosomes.

Figure 3.

Cellular sites of initial recognition of viral genome by Gag. Transcription of the integrated retroviral genome generates the grNA that includes the Psi site . After export into the cytoplasm, gRNA is used either for Gag protein translation or as genome for packaging into progeny virions. The latter fate requires gRNA dimerization, occurring inside the nucleus (e.g., RSV, MLV) or within the cytosol (e.g., HIV, FIV). Cytoplasmic gRNA dimers then interact with Gag via the NC domain. The site of Gag-RNA interaction varies among the different retroviruses.

More recently, this model was reinforced by imaging experiments that have allowed the visualization of retroviral assembly in real time, at the scale of individual virions.^59–63 These experiments were performed using spinning disc confocal microscopy or total internal reflection fluorescent (TIRF) microscopy, a technique well-suited to study events that occur at the plasma membrane because it results in a selective excitation of fluorophores within 100 nm of the coverslip.⁶⁴ Stringent criteria such as the measurement of Gag proteins proximity by fluorescence resonance energy transfer (FRET) and recruitment of Gag proteins by fluorescence recovery after photobleaching (FRAP) were applied to ensure that de novo particle genesis was observed.⁶⁰ Observation and characterization of the assembly of HIV-1, MLV and equine infectious anemia virus (EIAV) virions in real time have confirmed that retroviral assembly is nucleated and completed at the plasma membrane.^59–63 These studies have also established that retroviral assembly is completed within approximately 10 min. The next challenging issue was then to resolve how the genome is recruited to the plasma membrane, and when this recruitment occurs relative to virion assembly. Gag could recruit the RNA in the cytoplasm and the Gag/RNA complex would then traffic together to the plasma membrane. Alternatively, RNA and Gag could reach independently assembly sites and the capture of the genome by Gag could occur during or after completion of assembly.

Cellular Site(s) of Initial Recognition of RNA by Gag

The mechanisms that trigger the targeting of HIV-1 and MLV Gag to the plasma membrane are well described. The MA domain of Gag contains a bipartite signal that promotes stable interaction with the plasma membrane: a N-terminal myristyl group and a patch of basic residues within MA that interacts specifically with PI(4,5)P₂, a class of phospholipids that resides exclusively in the plasma membrane.^65–67 What was less clear until recently is how the viral genome is recruited to assembly sites. Characterization of the localization of two lentiviral genomes in live cells, in the presence or absence of Gag, revealed that Gag is responsible for the recruitment of the genome to the plasma membrane^61,68 (Fig. 3). HIV-1 and Feline Immunodeficiency Virus (FIV) RNAs were fluorescently tagged using the MS2 system, which involves the coat protein of the bacteriophage MS2. This protein recognizes with high affinity and high specificity a stem loop within the phage RNA.⁶⁹ Twenty-four copies of these stem loops were inserted into HIV-1 and FIV RNA reporters, which were then recognized by MS2 fused to GFP. Detection of MS2-GFP-bound RNA in VLPs collected from cell culture supernatant, either by microscopy^61,70,71 or by protein gel blotting analysis,⁶⁸ indicated that the modified genomes were efficiently packaged into virions. Importantly, this encapsidation was dependent on the RNA packaging signal.^61,68 When expressed without Gag, FIV and HIV-1 MS2 bound genomes, represented by green puncta under epifluorescent microscopes, were detected in the nucleus, at the nuclear membrane, and in the cytoplasm.^61,68 Live cell TIRF microscopy imaging of the HIV-1 RNA molecules revealed that they were moving in and out of the proximity of the plasma membrane, remaining there for no more than few seconds.⁶¹ By contrast, when Gag was expressed with the viral RNA reporters, both HIV-1 and FIV genomes accumulated and remained at the plasma membrane for the entire imaging period (Fig. 3). Moreover, viral genomes could no longer be detected at plasma membrane when Gag mutants incapable of binding the plasma membrane were coexpressed with RNA reporters. Altogether, these data strongly suggest that Gag is responsible for anchoring the genome at assembly sites.^61,68

Interestingly, the viral genome was stabilized at the plasma membrane few minutes before Gag accumulation was detectable at the same site.⁶¹ These observations indicate that a few Gag molecules are likely to be responsible for anchoring the HIV-1 RNA to the plasma membrane, yet these events were under the detection limit of the TIRF microscope. As such, these microscopic analyses have demonstrated that the viral genome is anchored at the plasma membrane by Gag, but they were not sensitive enough to determine where in the cell Gag proteins first recruit the viral genome. However, some hints about the initial site of genome recognition by FIV Gag can be deduced from such microscopic observations. When expressed alone, FIV Gag was accumulating both at the plasma membrane and at the nuclear membrane. In the presence of the genome, Gag and RNA co-localized at the nuclear membrane.⁶⁸ This suggests that FIV Gag is likely to recruit its genome at the nuclear envelope, and then traffics to the cell surface as a protein-RNA complex (Fig. 3).

It was previously reported that initial HIV-1 Gag-RNA interactions could occur at a perinuclear/centrosomal site.⁷² However, recent live cell co-imaging of the HIV-1 RNA and Gag proteins did not support this localization, since Gag did not accumulate to perinuclear sites.⁶⁸ Recent biochemical evidence strongly supports the notion that HIV-1 Gag and RNA interact in the cytoplasm before membrane association⁷³ (Fig. 3). RNA-immunoprecipitation, coupled to membrane flotation assays, performed either with Gag and RNA reporter in transfected cells or in infected cells, revealed that Gag-RNA complexes are present both in membrane and cytoplasmic fractions. Viral RNA could also be immunoprecipitated from the cytoplasmic fraction with great efficiency by Gag mutants that are unable to bind the plasma membrane. By coupling protein cross-linking experiments with membrane flotation approaches, it was shown that Gag forms low-order multimers in the cytoplasm and, as expected, high-order multimers at the plasma membrane.⁷³ Sequential RNAimmunoprecipitation experiments using two populations of Gag fused to different tags further demonstrated that Gag proteins interact with RNA as low-order multimers. Therefore, HIV-1 genome is recruited by few Gag molecules in the cytosol (Fig. 3). As for FIV, this implies that Gag and the viral genome traffic to assembly sites as a sub-complex.

Complex retroviruses have evolved so called accessory proteins, such as the HIV-1 Rev protein, that are able to export the unspliced gRNA from the nucleus.⁷⁴ Simple retroviruses, such as Rous sarcoma virus (RSV), do not have such proteins and interestingly, the Gag protein itself mediates the nuclear export of the unspliced genome. The regulation of Gag nuclear trafficking is finely tuned by different signals. To enter the nucleus, the RSV Gag protein utilizes two nuclear localization signals (NLS), one in NC and the other one in MA. Each one of these NLS is independent and interacts with different importins. The NC NLS involves the classical importin-α/β pathway, whereas the MA NLS interacts with Kap120p and Mtr10p (also called transportin SR and importin-11 in higher eukaryotes) import receptors.⁷⁵ Why Gag contains two independent NLS is still unclear. Based on some in vitro studies it seems that Gag interacts with the importin receptors as a monomer.⁷⁶ Once in the nucleus, Gag recruits the gRNA⁷⁶ via an interaction between NC and the Psi site of the unspliced gRNA (Fig. 3). The formation of this complex seems to promote Gag dimerization or oligomerization.⁷⁷ It was proposed that Gag multimerization induces a change of conformation of the complex exposing the nuclear export signal (NES),⁷⁶ which resides in a leucine-rich region of 18 residues within the p10 sequence of Gag. The NES interacts with the CRM1 nuclear export receptor which promotes the export of the Gag/RNA cargo from the nucleus to the cytoplasm by the CRM1 export pathway^78,79 (Fig. 3). It has been speculated that once in the cytoplasm, the complex does not re-associate with importins because MA and NC domains are engaged in higher affinity with the gRNA. Thus, the complex can be directed to the plasma membrane to complete particle assembly.⁷⁶

Altogether, these data suggest that very few Gag molecules initiate the recognition of the genome,^61,73 at a variety of cellular sites, such as the nucleus (RSV), the nuclear membrane (FIV) or the cytosol (HIV-1) (Fig. 3). In all these cases, the RNA therefore traffic to assembly sites bound to Gag.

Mechanisms of Transport of RNA-Gag Complex to Assembly Sites

The mechanisms by which retroviral RNAs are transported through the cytoplasm are poorly understood. Cellular mRNAs have been reported to traffic through the cytoplasm either by passive diffusion followed by local trapping or by directional movement at speeds that are too fast to be explained by passive diffusion, suggesting an active transport along the cytoskeleton.^69,80 In fact, some mRNA movements have been shown to be actin-dependent, whereas others seem to depend on the microtubule network.⁸⁰ These two scenarios can be envisaged for retroviral gRNAs. Observations of MS2-GFP-bound HIV-1 RNA in live cells with a TIRF microscope have revealed that viral RNA molecules are highly dynamic and that their rapid movements are rarely directed or confined, but rather ‘random’.⁶¹ This suggests that HIV-1 RNAs are not actively transported through the cytoplasm. More studies are definitely warranted to investigate this issue.

The mechanisms of transport of the RNA-Gag complex, from the initial binding site to the plasma membrane, have not been thoroughly studied. However, one study, based of live cell imaging of cells stably expressing MLV Gag and Env, reports that MLV Gag and MS2-bound MLV genomes are associated with endosomes that exhibit directed movements through the cytoplasm.⁵² MLV RNA and Gag were therefore proposed to traffic together to the plasma membrane within endosomes in a microtubule-dependent fashion. Moreover, assembled Gag-RNA particles inside endosomal vesicles were visualized by electron microscopy in situ hybridization (EM-ISH) and immunolabeling (EM-IF) analyses.⁵⁴ However, this does not mean that packaging nor assembly occur there. The RNA-Gag complexes detected here may well represent assembled virions that have been endocytosed subsequent to assembly. In fact, after completion of assembly at the plasma membrane, MLV and HIV-1 particles containing viral genomes are internalized into CD63-positive endosomes that exhibit highly dynamic movements reminiscent of microtubule-based transport.^58,60,61,68 Therefore, the mechanism of retroviral RNA-Gag complexes transport to the plasma membrane is still an open question.

Gag-RNA Complexes Nucleate Retroviral Assembly

It is long known that retroviral RNA is not only the support of the genetic information but also plays a critical role in viral assembly. First evidence to support this notion came from in vitro studies performed with purified HIV-1, MLV and RSV Gag. It was shown that nucleic acid greatly increases the efficiency of particles assembly, that the size of these particles was dependent on the size of the RNA, and that treatment of particles with ribonuclease triggered disassembly.^81,82 Later studies performed on MLV and HIV-1 particles generated in mammalian cells have demonstrated that, in the absence of viral gRNA, VLPs specifically package cellular mRNAs randomly, as mRNA species detected in VLPs are in proportion to their cellular levels,^24,83,84 at the exception of some RNA species such as 7SL, and U6 RNAs, that are specifically enriched in wild-type HIV-1 and MLV particles and whose function remains unclear.^24,25,83,84 Interestingly, retroviral cores containing either viral or cellular RNAs were disassembled by RNase treatment,⁸³ confirming the importance of the RNA-Gag interaction in the formation and structure of retroviral particles.

Recent real-time microscopic studies have allowed the visualization of MS2-GFP-bound viral genome during Gag driven viral assembly⁶¹ and have confirmed the role played by the genome in particle formation. When express alone, HIV-1 genomes were not detected at the plasma membrane. By contrast, in the presence of Gag, a population of viral gRNA that remains at the plasma membrane for long period of time and until the end of the imaging period was detected. This population was not detected when the packaging-impaired viral gRNA (Psi-minus) was used and is therefore likely to represent Gag bound genomes. Interestingly, these genomes first appeared to be Gag-free and approximately few minutes later, Gag becomes detectable at the same location as the anchored RNA molecules. The Gag signal increased progressively during approximately 5–10 min on the RNA puncta, until it reaches a plateau; a pattern which represent an assembly events.^59,60 Interestingly, the increase in Gag fluorescence coincided with reduced lateral velocities of the RNA, suggesting that Gag accumulation around the genome leads to a more and more efficient binding of the RNA-Gag complex to the membrane (Fig. 3). Gag mutants lacking the C-terminal domain of CA (Gag-ΔCTD) are unable to assemble into virions due to the absence of Gag-Gag interaction but retain RNA- and membrane-binding activity. The genomes associated with these Gag mutants were also docking at the plasma membrane but had a tendency to dissociate from it after few minutes. This was never observed with genome bound to WT Gag. These findings are in perfect agreement with biochemical data performed with the same Gag mutants showing that Gag-ΔCTD was capable of recruiting the RNA but that the Gag-ΔCTD/RNA complexes were less stable than the WT Gag complexes.⁷³ Therefore, Gag molecules that are yet to assemble into VLPs are capable of recruiting viral RNA molecules to the plasma membrane.

It is also worth noting that the appearance of RNA at assembly sites was always represented by a single step in fluorescent signal increases which level remained constant as assembly proceeds. Since complete virions are known to contain dimeric genomes,^70,85–88 this suggests that the genomes are recruited as a preformed dimer, rather than as individual molecules that subsequently dimerize during or after packaging. Consistent with this notion, it was reported that MLV gRNA dimerization is coupled to splicing and transcription processes, suggesting that MLV gRNA exits the nucleus as dimeric forms^89–92 (Fig. 3).

The visualization of genome during HIV-1 assembly therefore supports a model in which a small number of Gag molecules recruits dimeric genomes in the cytoplasm. Then, Gag-RNA complexes are targeted to the plasma membrane where they nucleate virion assembly (Fig. 3).

Concluding Remarks

Several recent studies, based on live cell microscopy of single RNA molecule, and on original biochemical approaches, have begun to reveal the mechanisms that accompany the initial events of RNA packaging. Such studies have been limited to very few retroviruses and have revealed that a small number of Gag molecules can recruit the viral genome at a variety of cellular sites, including the nucleus, the nuclear membrane and the cytosol. Extending such studies to a broad spectrum of retroviruses should revealed if any of these sites are preferred for initial RNA recognition by retroviral Gag. Other key questions of RNA packaging await characterization. For instance, the sorting mechanisms that drive the gRNA either to the translational machinery for production of Gag and GagPol polyproteins, or to specific cellular sites, where it will be recognized by Gag, remain elusive. One plausible hypothesis is that a conformational switch controls the fate of gRNA: the structural change induced by RNA dimerization could act as a positive signal for packaging and as a negative signal for translation. Such RNA conformational switches have been elucidated upon dimerization of HIV-1 and MLV RNAs.^{11,12,19,93–99} However, direct insights into RNA structure-function relationship are needed to validate this hypothesis. Another possibility is that the interaction between the gRNA and Gag, or between the gRNA and a-yet-to-be-identified cellular factor, could modulate the dual targeting of the genome.

Similarly, the spatiotemporal control of the gRNA dimerization in infected cell remains poorly characterized for most retroviruses. MLV RNA dimerization is likely to occur in the nucleus and is coupled to transcription and splicing events.^89–92 For HIV-1 RNA, dimerization is likely to happen later in the cytoplasm.⁷¹ However, where exactly in the cytoplasm HIV-1 dimerization occurs and when it occurs relative to Gag binding remained to be elucidated.

Finally, the present review did not report on the cellular proteins packaged into virus particles (e.g., Staufen, nucleolin, APOBEC3G). The elucidation of the contribution of these cellular proteins should help to more precisely define the main events of virus assembly. A better understanding of these processes in which viral RNA plays a key role, may provide attractive antiviral drug targets.^100–104

Abbreviations

HIV

human immunodeficiency virus

MLV

murine leukemia virus

MPMV

Mason-Pfizer monkey virus

FIV

feline immunodeficiency virus

HTLV

human T-cell leukemia virus

EIAV

equine infectious anemia virus

RSV

Rous sarcoma virus

VLPs

viral-like-particles

NC

nucleocapsid

GagNC

NC domain of Gag

ZF

zinc fingers

gRNA

genomic RNA

UTR

untranslated region

ESCRT

endosomal sorting complex required for transport

TIRF

total internal reflection fluorescent microscopy

FRET

fluorescence resonance energy transfer

FRAP

fluorescence recovery after photobleaching