One of the principles underlying the use of any compound or factor as a therapeutic agent is its capacity to selectively affect the target with little or no off-target effects. With this concept in mind, recent reports regarding the ability of a variant of the host factor Sam68 to selectively regulate the expression of several key components of HIV-1 take on particular interest. HIV-1 replication is critically dependent on the expression of its structural proteins, Gag, Gagpol and Env 1. As a result, any factor able to inhibit expression of these proteins would force the virus into a state akin to latency. In addition, HIV-1 Nef has been implicated as a major player in the pathogenesis of this virus 23, expression of Nef alone in transgenic mice reproducing many aspects of the pathology seen by the intact virus in humans 4. The recent reports that a mutant of Sam68, Sam68ΔC (lacking the C-terminal nuclear localization signal), is able to interfere at both the level of HIV structural protein and Nef synthesis makes it of particular interest 56.

Initial experiments 7 identified Sam68ΔC as a dominant inhibitor of HIV-1 replication. While subsequent work determined that inhibition was dependent upon the cytoplasmic localization of Sam68ΔC and associated with formation of cytoplasmic granules around the outside of the nuclear envelope 8, the underlying mechanism remained unclear. However, the recent work of Marsh et al. 6 provided some detail as to the mechanism. Using various expression vectors, they showed that Sam68ΔC selectively inhibited mRNA expressing Gag exported via the exportin-1 pathway, with little to no effect on the same Gag coding sequence delivered to the cytoplasm by Nxf1. Co-expression of Sam68ΔC and an unspliced Env expressor resulted in translation inhibition of the latter and disruption of the cytoplasmic bundles failed to restore expression of the encoded protein. Rather, despite a normal polyA tail, inhibition by Sam68ΔC was attributed to a block in translation of the affected RNA due to reduced binding of PABP-1 (Fig. 1A). The ability of Sam68ΔC to selectively affect only those RNAs exported in a Rev- and exportin-1-dependent fashion suggested that it recognizes some features unique to the mRNPs exported by this pathway. In parallel work by Henao-Mejia et al. 5 and consistent with Marsh et al., it was shown that constructs functionally similar to Sam68ΔC had the capacity to repress Rev-dependent protein expression. Surprisingly, inhibition of Rev-independent Nef synthesis was also observed with little or no alteration in Tat or Rev levels. Given that these three proteins are expressed from multiply spliced HIV-1 RNAs that all use the Nxf1 export pathway (Fig. 1A), selective repression of Nef expression may require a different mechanism than that outlined by Marsh et al. Inhibition of Nef expression was reported to be associated with the accumulation of nef mRNA in cytoplasmic granules that co-stained with markers of stress granules (SGs); these observations led Henao-Mejia et al. to suggest that reduced Nef synthesis was due to sequestration in these bodies. At present, it is unclear whether the granules characterized by Henao-Mejia et al. are similar or distinct from those formed by Sam68ΔC and incompletely spliced HIV-1 RNAs and whether Sam68ΔC inhibition of Nef synthesis is dependent upon their integrity. The two studies suggest that, while the route different RNAs take to repressive sites can differ (the Exportin1 pathway for Rev-dependent RNAs versus the Nxf1 pathway for nef mRNA), a similar mechanism may underlie repression of HIV-1 structural protein and nef mRNAs by Sam68ΔC. However, whether the mechanism is simple sequestration in SGs or something more complex remains to be determined. This is based on the observation of Marsh et al. that RRE-containing RNAs are still repressed upon dispersal of Sam68ΔC granules, although dispersion into functional "nano" granules cannot be dismissed and should be investigated. In addition, ongoing studies (Marsh and Cochrane, unpublished) showing that Sam68ΔC-induced granules contain mRNAs whose expression is not repressed suggest that sequestration to such granules alone is insufficient to explain translational repression. Consequently, additional experiments are needed to assess whether common or distinct mechanisms underlie repression of HIV-1 structural protein and nef mRNAs by Sam68ΔC.

The suggestion that Sam68ΔC can discriminate nef mRNA from that of tat and rev is of particular interest given that these RNAs not only share a common export pathway but are almost identical except for differences in their 5' untranslated regions (Fig. 1B). The determination by Henao-Mejia et al. that sensitivity to Sam68ΔC is due to sequences in the 3'UTR of nef mRNA that are also present in tat/rev mRNAs raises questions about how repression is restricted to nef mRNAs. One hypothesis is based on the position of the different reading frames and the influence of translation on 3' UTR structure/RNP composition. Both tat and rev mRNAs contain reading frames encoding the respective proteins (Tat or Rev) and that of Nef, while nef mRNA has only one reading frame (Fig. 1B). Since translation requires the unfolding of RNA secondary structure as well as disruption of protein-RNA interactions, it is possible that the sequence spanning the Nef reading frame within tat and rev mRNAs could have very different secondary structure and/or RNP composition than nef mRNA. Consequently, repression specificity could be achieved by Sam68ΔC binding to RNPs containing alternative structure/composition in the region common to the three mRNAs. Such a hypothesis is readily testable and will provide important insights into the determinants that specify susceptibility to regulation by Sam68ΔC. Defining the mechanism by which Sam68ΔC selectively inhibits the expression of several key HIV-1 mRNAs will provide important insights into their regulation and potentially lead to new approaches to controlling the pathogenesis of this virus.