Background

Retroviruses have evolved replication cycles that excel in circumventing host antiviral responses. One strategy that retroviruses have developed is to shield their replication intermediates from cytosolic nucleic acid sensors such as cGAS, TREX1, IFI203, and DDX41 (Yan et al., 2010; Gao et al., 2013; Lahaye et al., 2013; Stavrou et al., 2015). During replication, retroviruses produce RNA-DNA hybrids and unmethylated double-stranded proviral DNA in the cytosol that are common targets for innate immune sensors (Yan et al., 2010; Gao et al., 2013; Lahaye et al., 2013). The mature retroviral capsid core is constituted by about 1,500 units of the viral capsid protein (CA) that assemble to form a rigid structure that houses two copies of the retroviral RNA genome, the viral reverse-transcriptase, various other host-derived molecules (e.g., miRNAs, proteins, dNTPs), and in some cases, viral accessory proteins (Ganser et al., 1999; Briggs et al., 2004; Cantin et al., 2005; Campbell and Hope, 2015). This structure is permissive to the diffusion of dNTPs, yet it is impermeable to most host proteins (Jacques et al., 2016). In order to successfully and efficiently deliver the proviral DNA associated with the pre-integration complex (PIC) to the nucleus of the cell, the viral core must maintain a certain level of stability. Once it is physically situated proximal to a nuclear pore complex, the PIC may traverse into the nucleus and deliver the proviral DNA. Capsid cores that lack structural integrity or are destabilized by host restriction factors, like TRIM5α, will trigger innate responses and fail to deliver proviral DNA to the nucleus (Sayah et al., 2004; Stremlau et al., 2004 and 2006).

One of the most commonly used methods of analyzing retrovirus capsid core stability is known as the ‘fate of capsid assay’ (Stremlau et al., 2006; Perron et al., 2007; Yang et al., 2014). This assay involves infecting cells and then detecting the amount of intact pelletable capsid cores in the lysates of the infected cells. While this assay can compare relative levels of cores that persist in the cytosol over time, it is quite labor intensive, time-consuming and requires very large quantities of virus. In our hands, the ‘fate of capsid assay’ required the equivalent of 10⁸ Transducing Units (TU–as measured by productive infections) of Moloney Murine Leukemia Virus (M-MLV) which was barely over the limit of detection in our conditions (Renner et al., 2018a). Additionally, this method cannot differentiate endocytosed, non-infectious capsids, from those capable of a productive infection. Other alternatives to this approach include visual tracking of intact fluorescent cores, and the direct assessment of viral uncoating kinetics in a cyclosporine A washout assay (Fricke et al., 2013; Campbell and Hope, 2015).

The protocol presented here is an adaptation of a similar method (Forshey et al., 2002; Aiken, 2009; Shah and Aiken, 2011). It is designed to rapidly assess the relative stability of capsid cores from different viral mutants. In this modified approach, retroviruses are pre-treated with a detergent to strip away the viral envelope, and then the naked capsids are spun through a sucrose gradient with a detergent layer at the top (Renner et al., 2018a). Measuring the amount of intact cores recovered following ultracentrifugation provides an easy way of determining the comparative stability of the cores from different viruses. In the context of our published study, we also directly compared this method with a conventional ‘fate of capsid assay’, which yielded similar results but was much more involved to carry out and required a substantially larger viral input. The retroviral capsid core stability assay described here poses as an easy and technically reproducible alternative to other approaches evaluating capsid stability (Renner et al., 2018a).