ESCRT-independent budding of HIV-1 gag virus-like particles from Saccharomyces cerevisiae spheroplasts

Abstract

Heterologous expression of HIV-1 Gag in a variety of host cells results in its packaging into virus-like particles (VLPs) that are subsequently released into the extracellular milieu. This phenomenon represents a useful tool for probing cellular factors required for viral budding and has contributed to the discovery of roles for ubiquitin ligases and the endosomal sorting complexes required for transport (ESCRTs) in viral budding. These factors are highly conserved throughout eukaryotes and have been studied extensively in the yeast Saccharomyces cerevisiae, a model eukaryote previously utilized as a host for the production of VLPs. We used heterologous expression of HIV Gag in yeast spheroplasts to examine the role of ESCRTs and associated factors (Rsp5, a HECT ubiquitin ligase of the Nedd4 family; Bro1, a homolog of Alix; and Vps4, the AAA-ATPase required for ESCRT function in all contexts/organisms investigated) in the generation of VLPs. Our data reveal: 1) characterized Gag-ESCRT interaction motifs (late domains) are not required for VLP budding, 2) loss of function alleles of the essential HECT ubiquitin ligase Rsp5 do not display defects in VLP formation, and 3) ESCRT function is not required for VLP formation from spheroplasts. These results suggest that the egress of HIV Gag from yeast cells is distinct from the most commonly described mode of exit from mammalian cells, instead mimicking ESCRT-independent VLP formation observed in a subset of mammalian cells. As such, budding of Gag from yeast cells appears to represent ESCRT-independent budding relevant to viral replication in at least some situations. Thus the myriad of genetic and biochemical tools available in the yeast system may be of utility in the study of this aspect of viral budding.

Figure 1. Expression and localization of HIV Gag-GFP fusion proteins.

(A) Live cells (SEY6210 or JPY69 for ) expressing the indicated form of Gag-GFP were visualized by fluorescence microscopy. Scale bar for each image represents 4 . (B) Yeast cells expressing either WT Gag-GFP or were analyzed for their ability to incorporate these chimeras into VLPs. Whole cell extracts (W.C.) and the VLP fractions were subjected to SDS-PAGE and western blotting with anti-GFP antibody to allow visualization. (C) Representative images of fluorescence recovery after photobleaching performed on wild type cells expressing Gag-GFP (pre-bleach, first image captured post-bleach, and 25 minute recovery). The red box highlights a punctum that was subjected to bleaching while the red arrows highlight Gag-GFP puncta that were not. (D) Quantification of fluorescence recovery after photobleaching (at one frame per minute) for wild type and cells expressing wild type Gag-GFP.

Figure 2. Characterization of VLPs.

(A) VLPs were collected from wild type cells expressing wild type Gag-GFP and subjected to centrifugation across a sucrose gradient (20–70%, weight to volume). 1 ml fractions were subsequently collected, the density of each fraction was measured, and each fraction was subjected to SDS-PAGE and anti-GFP immunoblotting. (B) Wild type cells expressing either Gag-GFP (WT) or (G2A) were equivalently processed to isolate the whole cell and VLP fractions. The VLP fraction was subsequently incubated with either PBS alone (PBS), PBS+Trypsin (T), or PBS+Trypsin+Triton X-100 (T+TX100), and post reaction material was visualized by immunoblotting using anti-GFP antibody. Whole cell lysates (W.C.) from cells expressing either WT Gag-GFP or used to generate the VLP fraction were loaded onto the gel to indicate the relative amount of expression.

Figure 3. HIV Gag-GFP VLP budding from yeast spheroplasts does not require Gag L-domains.

(A) and (B) Wild type yeast were transformed with , , , or . VLP budding of the Gag-GFP constructs was assayed as described in the Methods section. Gag-GFP was visualized by Western blotting of whole cell extracts and VLP fractions with anti-GFP antibody.

Figure 4. HIV Gag-GFP VLP budding from yeast spheroplasts is not diminished by loss-of-function mutations in Rsp5.

was transformed into an background complemented with , , or . (A) Localization of by fluorescence microscopy on the indicated live cells. (B) VLP budding assays were conducted with cells of the indicated genetic backgrounds, and whole cell extracts (W.C.) and VLP fractions were visualized by anti-GFP immunoblotting.

Figure 5. HIV Gag-GFP VLP budding from yeast spheroplasts is independent of ESCRT function.

(A) or GFP-Cps1 were transformed into wild type cells or strains deleted for a component of each ESCRT complex, ESCRT-0 (), ESCRT-I (), ESCRT-II (), ESCRT-III () or VPS4 (), and GFP localization was assessed by fluorescence microscopy of live cells. In the wild type background GFP-Cps1 was sorted into the lumen of the yeast vacuole, while loss of ESCRT function, Bro1 function, or Vps4 function resulted in missorting to the vacuolar limiting membrane and accumulation within the aberrant class E compartment. (B) or were expressed in the indicated strains strains and whole cell extracts (W.C.) and VLP fractions were analyzed by Western blotting with anti-GFP antibody. (C) VLPs were collected from cells expressing wild type Gag-GFP and subjected to centrifugation across a sucrose gradient (20–70%, weight to volume). 1 ml fractions were subsequently collected, the density of each fraction was measured, and each fraction was subjected to SDS-PAGE and anti-GFP immunoblotting. (D) cells expressing wild type Gag-GFP were processed to isolate the whole cell and VLP fractions. The VLP fraction was subsequently subjected to incubation with either PBS alone (PBS), PBS+Trypsin (T), or PBS+Trypsin+Triton X-100 (T+TX100) and the reactions were visualized by immunoblotting using anti-GFP antibody. Whole cell lysate (W.C.) from the cells used to generate the VLP fraction was loaded to indicate expression.