The highly conserved HIV accessory protein Viral Protein R (Vpr) is vital for HIV infection and replication in vivo, particularly within non-dividing cells, including terminally differentiated T-cells and macrophages where nuclear envelope integrity is permanently maintained 12345. On its own and within the context of HIV infection 67, Vpr has been shown to localize to the nucleus and induce G₂ cell-cycle arrest through hyperphosphorylation of p34-cdc2, which provides the most optimal environment for viral replication 8910111213, followed closely by apoptosis 1415161718192021. During HIV infection, Vpr associates with the viral cDNA containing Pre-integration Complex (PIC) increasing its affinity for components of the cellular nuclear import machinery 22232425 through the activity of 2 distinct nuclear localization signals (NLSs) within its N- 26 and C-termini 27, thus driving productive HIV infection.

We previously reported 28 on a cohort spanning 1992 to 2000, comprising an HIV⁺ long-term non-progressor (LTNP) (donor A) and two recipients (B and C) who upon transmission (autumn and summer 1989 respectively) from donor A progressed to AIDS. Vpr sequences derived from HIV pro-viral DNA isolated from PBMCs as well as circulating virus, revealed that sequences from the progressing recipients differed markedly to those of the LTNP founder virus host, providing the first evidence for Vpr positive selection during disease progression in an epidemiologically-linked cohort (see also 29).

In this study, cohort-derived GFP-Vpr mammalian cell expression vectors were used to investigate Vpr subcellular localisation and nuclear import. In total, 4 HIV-1 vpr clones were isolated from donor A (A1, A3, A5 and A6 corresponding to years 1996, 1998, 1999 and 2000), 4 from recipient B (B3, B4, B5 and B6 corresponding to years 1997, 1998, 1998 and 1999) and 7 from recipient C (C2, C3, C5, C6, C8, C10 and C11 from 1992, 1992, 1993, 1994, 1994, 1998 and 1999 respectively) and used to generate the GFP-Vpr expression constructs. Lipofectamine™ 2000 (Invitrogen) was used to transfect constructs into HeLa and COS-7 (not shown) cells, followed by imaging 14 hours post transfection using an Olympus FV-1000 confocal laser scanning microscope (CLSM), with similar results. Initial analysis of late-stage vpr gene products from all 3 donors revealed a reduction in nuclear fluorescence accompanied by perinuclear accumulation for LTNP sample A6-2 (patient A, time-point 6, clone #2) (Fig. 1A(iii)) compared to pUC18-NL4.3 derived 30 wild-type GFP-Vpr^1–96 (Fig. 1A(ii)): a phenotype not observed in samples from patients B (B6-4) nor C (C11-1) (Fig 1A(iv, v)). Quantification of nuclear import was determined using ImageJ analysis software 31 where the ratio of nuclear (Fn) to cytoplasmic (Fc) fluorescence, subtracting background (B) was calculated, Fn/c = (Fn-B)/(Fc-B). A highly significant (p < 0.0001) reduction in Fn/c was observed between GFP-Vpr^1–96 (Fn/c ~2.4) and GFP-Vpr^A6-2 (Fn/c ~0.9) indicating a reduction in nuclear accumulation. No significant difference in Fn/c was observed between GFP-Vpr^1–96 and GFP-Vpr^B6-4 or GFP-Vpr^C11-1 (Fig. 1B).

To establish if decreased Vpr nuclear accumulation was a specifically conserved viral feature of the LTNP or an acquired trait, GFP-Vpr constructs from early-cohort timepoints were analysed. Samples from 1996 (GFP-Vpr^A1-2) and 1998 (GFP-Vpr^A3-1) (Fig. 1A(vi, vii)) both displayed levels of nuclear accumulation comparable to wildtype GFP-Vpr^1–96 (Fig. 1B), implying that loss of nuclear import occurred post 1998.

With the loss of efficient nuclear accumulation identified as a late-stage LTNP trait, further analysis of late-stage clones A6-3 and A6-5 (clone A6-5 data not shown as sequence is homologous to A6-3) was performed, revealing Fn/c values significantly lower than wildtype GFP-Vpr^1–96 (Fig. 1B). Of interest, a single late-stage clone (GFP-Vpr^A6-4) displayed a small degree of nuclear accumulation (Fig. 1A(ix)), although at levels significantly reduced compared to wildtype (Fig. 1B). Sequence analysis revealed that all clones with reduced nuclear accumulation (Vpr^A6-2 and Vpr^A6-3/5) contained a phenylalanine to leucine mutation at amino acid position 72 (F72L) (Fig. 1C). This mutation is absent in the nuclear localizing Vpr^A6-4, which is otherwise identical in sequence to late-stage samples Vpr^A6-3/5. Analysis of our database containing over 1200 published and unpublished Vpr sequences revealed that the F72L mutation has only been reported once previously 32.

To determine the specific role of F72L in abrogation of Vpr nuclear import, an F72L substitution was engineered into wildtype Vpr and an expression construct (GFP-Vpr^F72L) produced. Subsequent transfection (Fig. 1A(x)) revealed a reduction in GFP-Vpr^F72L nuclear accumulation when compared with wildtype GFP-Vpr^1–96 (Fig. 1A(ii)), in a similar fashion to clinically derived GFP-Vpr^A6-2 (Fig. 1B), thus confirming F72L as a probable mechanism for the observed reduction in nuclear accumulation within late-stage LTNP samples. This reduction in nuclear import was accompanied by an accumulation of GFP-Vpr protein within the perinuclear region of the cell, concordant with an area encompassing the proposed location of the Golgi. Colocalisation studies with the Golgi-specific marker protein γ-adaptin revealed that Vpr proteins harbouring the F72L point mutation, but not those with wildtype phenylalanine, were localized to the Golgi apparatus, implying that reduced nuclear accumulation could in part be attributed to this mislocalisation of GFP-Vpr (see Additional file 1). The lack of colocalisation between F72L-containing Vpr proteins and the endoplasmic reticulum (ER) marker Calnexin (Sigma, C-4731), confirmed that Vpr mislocalisation to the Golgi is not a result of general mistargeting (see Additional file 2).

Although the precise pathways involved are still debated, efficient nuclear export 33 and/or cytoplasmic retention 34 of Vpr during viral infection has been identified as a key requirement for virion incorporation. With this in mind, we assessed the ability of nuclear import-inhibited cytosolic GFP-Vpr^F72L to incorporate into forming virions. Briefly, 293T cells were cotransfected with a panel of GFP-Vpr expression vectors as well as the proviral plasmid pUC18-NL4.3, with the resultant virions analysed by Western blotting for GFP-Vpr incorporation. Interestingly we found that all F72L-containing GFP-Vpr proteins that failed to significantly localize within the nucleus, were absent from purified viral lysates (Fig. 2A), implying a possible link between the efficient nuclear transport of Vpr and virion incorporation. To examine the effect of the F72L mutation on HIV infectivity, we infected γ-irradiated, non-dividing CD4⁺ MAGI (

M

A

G

I

Nuclear accumulation of Vpr is governed by 2 distinct NLSs; a leucine-rich N-terminal α-helix (²⁰LELLEEL²⁶) which has been linked to virion incorporation 3738, and a C-terminal arginine-rich bipartite (

underlined

RI

C

RQRR

R

It does not seem unreasonable to speculate that in vivo selective pressures have driven the evolution of functionally-attenuated vpr to ensure long-term viral survival and replication in vivo 25404142. Our study shows that F72L-containing Vpr does not accumulate within host cell nuclei or incorporate into nascent virions; subsequent infections with such a Vpr-deficient virus are less efficient, leading to reduced virus production 414243 and ultimately a lack of deterioration in T-cell numbers, as in the case of the LTNP. Many of our observations are concordant with analysis of LTNP infection status in 2000, which revealed normal CD4⁺ T-cell levels (>550 cells per μl) suggesting the absence of significant HIV-mediated T-cell targeting and depletion. Since our data here suggests that c. 75% of late-stage proviral strains in the LTNP would produce virions lacking Vpr due to the F72L mutation, the levels of observed CD4⁺ T-cells within the LTNP may be due to the inability of HIV lacking Vpr to import the viral genome into the non-dividing host cell nucleus and carry out subsequent steps of infection 1242544. Presumably one advantage for the virus is that the preservation of patient health status leads to enhanced probability of transmission to new hosts, thus leading to increased virus spread. Low levels of virus production (<425 HIV copies/ml plasma) 29 are presumably maintained through the activity of a small subset of Vpr quasi-species that do not contain the F72L mutation (A6-4) or perhaps viruses that we were unable to characterize due to their low prevalence. Additional to all of the above considerations, it is important to remember that the effects of decreased Vpr nuclear import through the F72L mutation are likely to be influenced by many other factors, including host genetic make-up and immune response, which in themselves can influence disease progression 284546474849.

In summary this study provides evidence for a naturally occurring Vpr mutation within an epidemiologically-linked cohort and its possible contribution to non-progressive HIV disease status through disruption of efficient nuclear transport and apparent lack of viral incorporation, leading to reduced infectivity of non-dividing cells (Fig. 2B). To demonstrate formally the link between Vpr nuclear localisation and disease progression, further analysis of Vpr^F72L in a HIV infectious system is required, eg. to dismiss the possibility that the effects in terms of lack of assembly of Vpr^F72L may in part be due to impaired p6gag binding, stemming from conformational or other effects. With this proviso in mind, however, the work here and elsewhere (see also 50) implies that functional selection of vpr viral quasispecies in concert with host selection pressure over time, as evident in the LTNP, may be a factor in determining the rate of disease progression.