Selective Isolation of Retroviruses from Extracellular Vesicles by Intact Virion Immunoprecipitation

Background

Widely used approaches for isolating retroviruses, such as the Human Immunodeficiency Virus (HIV) and Murine Leukemia Virus (MLV), include precipitation, chromatography, ultrafiltration, ultracentrifugation, as well as various other means of particle separation (Reviewed in Nestola et al., 2015). While each technique has its specific advantages, drawbacks and limitations, a common concern for all methods is the co-purification of cell secreted extracellular vesicles (EVs).

EVs constitute a heterogeneous population of membrane-derived vesicles secreted by all cell types (Yanez-Mo et al., 2015). There are strikingly similar biophysical and biochemical characteristics between retroviruses and EVs (Table 1), especially with the small 50-150 nm vesicles secreted through the endosomal pathway, better known as exosomes (Reviewed by Nolte-'t Hoen et al., 2016). Some retroviruses, such as HIV and MLVs, can also share with exosomes the pathways of biogenesis and egress through the endocytic system (Orenstein et al., 1988; Raposo et al., 2002; Houzet et al., 2006; Sandrin and Cosset, 2006; Akers et al., 2013; Madison and Okeoma, 2015; Martin et al., 2016; Nolte-'t Hoen et al., 2016). This imparts inherent biochemical composition similarities between the two types of particles, which extend to their cargo (e.g., proteins, mRNAs, miRNAs) and host-derived surface membrane proteins and antigens (e.g., CD9, CD63, CD81), which inevitably increases difficulties in telling them apart (Eckwahl et al., 2016; Nolte-'t Hoen et al., 2016; Telesnitsky and Wolin, 2016). Given such similarities, selective isolation of retroviruses requires a unique identifying marker to confidently discriminate them from exosomes and EVs in general.

Table 1. Retroviruses and EVs are nearly indistinguishable by their biochemical and biophysical characteristics


Nanoscale flow cytometry (NFC), also called flow virometry or NanoFACS, is an optimization of flow cytometry techniques, sample preparations and hardware for the analysis of particles smaller than 200 nm, which is the average detection limit of most commercial flow cytometers (Tang et al., 2016 and 2017; Lippe, 2018). This technology is especially useful for the immune phenotypic profiling of markers on the surface of viruses and EVs. By using this approach, we previously determined that a fluorescently tagged viral envelope glycoprotein (Env-eGFP) of the Moloney MLV (M-MLV) was almost exclusively expressed on the surface of these virus particles, and thereby constituted a very reliable selection marker (Figure 1) (Tang et al., 2017).


Figure 1. Env-eGFP represents a selection marker for identifying retroviruses by nanoscale flow cytometry. This figure has been adapted from Tang et al. (2017). 293T cells were mock transfected with an empty plasmid (A), with Env-eGFP (B) or eGFP (D) expression plasmids, or with the Env-eGFP expressing M-MLV viral plasmid (C). The M-MLV used in this study contained an eGFP reporter inserted into the proline-rich region of the extracellular domain of the envelope glycoprotein (Sliva et al., 2004). Supernatants were 450 nm-syringe filtered prior to NFC analysis. Transfection efficiency was monitored by eGFP expression in the transfected cells and was similar in each relevant condition (data not shown). For NFC analysis, particles were detected by triggering off of side-scattered light (SSC). A square gate was set above background in the Mock sample where eGFP+ events are expected (A). Side and top boundaries of this gate were determined by the limits of the eGFP+ events in the MLVeGFP sample (C). Numbers in green represent eGFP+ particles detected and enumerated in the gate during a fixed acquisition time window, which was the same for all samples analyzed. For NFC analysis, SSC is more sensitive than forward scattered (FSC) light to detect particles smaller than 200nm on our instrument (Tang et al., 2016). The results show that, in our system, the membrane-expressed Env-eGFP does not substantially associate with EVs (B). However, cytosolic eGFP is incorporated as cargo inside EVs (D). Env-eGFP is highly enriched only on the surface of viruses (C).

The protocol described here was specifically developed to study an enigmatic virus-encoded integral membrane protein called Glycogag (or gPr80) inserted in the envelope of M-MLV (Pillemer et al., 1986; Fujisawa et al., 1997 and 2001; Rosales Gerpe et al., 2015; Renner et al., 2018). Our goal was to assess the incorporation and orientation of full-length gPr80 in the envelope of M-MLV. A major caveat to this particular study was the release of EVs by the infected cells that contaminated our virus samples. This proved to be especially problematic, as we found that the gPr80 protein associated with both EVs and virions (Renner et al., 2018). But given that Env-eGFP was highly enriched on the surface of M-MLV virions and poorly incorporated on EVs (Figure 1) (Tang et al., 2017), we thus developed an intact virion immunoprecipitation (IVIP) assay designed to specifically isolate structurally intact viral particles expressing Env-eGFP on their surface. Using this approach, we successfully identified the orientation of gPr80 as a Type-I integral membrane protein on virions but as a Type-II integral membrane protein on EVs that are devoid of Env-eGFP (Renner et al., 2018). In conclusion, IVIP has the ability to selectively isolate and discriminate retroviruses from EVs with minimal physical manipulation and without compromising the structural integrity of either particle type.