We report here that the steady-state levels of Gag proteins in HIV-1-infected primary CD4+ lymphocytes and Hela cells are increased in the presence of Nef. Using imaging and biochemical approaches, we show that Nef changes the total amount and localization of both immature and mature forms of Gag proteins. In WT infected cells, Gag proteins are highly expressed, partially distributed in the cytoplasm and mostly localized at the plasma membrane. In the absence of Nef, expression of Gag proteins is lower, primarily cytoplasmic, and partially localized to the plasma membrane. Furthermore, the amount of HIV-1 p24 present in the supernatant of
Nef-infected cells is lower than that of cells infected with the WT virus (this work and [51
]). However, viral release, calculated as the ratio between the extracellular levels and the total amount (released plus cell-associated) of HIV-1 p24 is unaffected. The absence of Nef in infected lymphocytes does not affect their capacity to form virological synapses, confirming previous results [47
]. Once the virological synapse is formed, WT-infected cells transfer mostly mature viral particles to uninfected cells, whereas in the absence of Nef, the amount of transferred viral material is reduced and mostly immature. Together, these results strongly suggest a previously unappreciated effect of lentiviral Nef on the expression, intracellular localization and processing of Gag proteins, providing novel insights into how Nef optimizes viral replication.
How does Nef increase the amount of Gag proteins in infected cells? One possibility is that Nef, by affecting the HIV-1 long terminal repeat (LTR) activity, modulates Gag and GagPol expression. However, discrepant results have been reported regarding the effects of Nef on the transcriptional activity of the LTR [52
] . This hypothesis will deserve further investigation. Later in the viral life cycle, Nef may prevent the degradation of Gag proteins or of other proteins involved in the trafficking or assembly of Gag and/or GagPol. In support of this, it has been reported that proteasome inhibitors partly rescue the infectivity defect of
Nef viruses [60
]. Interestingly, Nef also increases levels of the cellular protein SOCS1 [61
], a protein that is important for Gag trafficking and stability in infected cells [62
The molecular and cellular mechanisms underlying the effects of Nef on the amount, trafficking and processing of Gag proteins remains to be further characterized. We hypothesize that Nef might positively impact the Gag biosynthetic pathway by acting at various levels. In the cytoplasm, Gag is mostly monomeric or dimeric, whereas higher ordered Gag multimers are found only at the plasma membrane [17
]. Gag assembly and viral budding is a cooperative process that depends on the amount of intracellular Gag [18
]. Thus, the reduced amount of Gag proteins observed in the absence of Nef could lead to an inefficient localization at the plasma membrane.
Nef is known to interact with the clathrin-dependent endocytic pathway and to modulate the surface expression of various cellular proteins [63
]. Gag contains a dileucine-like sorting motif that regulates association with multivesicular bodies [65
]. Gag interacts directly with the AP-3 complex, a component of the clathrin pathway [66
], and trafficking of Gag to late endosomes is part of a productive particle assembly pathway prior to budding from the plasma membrane [11
]. It is thus tempting to speculate that Nef may affect Gag trafficking through its effect on the clathrin-mediated cell sorting machinery. Nef also modulates actin dynamics by inactivating cofilin [68
], while the microtubule network is dispensable for proper targeting of Gag at the plasma membrane [11
]. It will be worth determining whether the effects of Nef on the actin cytoskeleton impact the overall levels of Gag proteins in infected cells.
Further work is thus warranted to determine which of these potential activities mediate the effects of Nef on Gag proteins. For instance, it may be of interest to determine which Nef mutants, known to be selectively defective in different activities of the viral protein, impact Gag localization and processing. Additionally, it has been shown that murine leukemia virus (MLV) glycosylated gag (Glycogag) proteins rescue the infectivity of Nef-defective virions [69
]. It will be worth examining whether Glycogag also impacts the trafficking and processing of HIV-1 Gag.
We further report a possible consequence of the effects of Nef on Gag proteins. Using two short-term co-culture systems (co-culture of infected HeLa cells with Jurkat target cells, and co-culture of primary CD4+ T cells), we demonstrate that the quantity and quality of the viral material transferred is different in the presence or in the absence of Nef. Without Nef, the percentage of targets having received viral material is significantly reduced. Moreover, by western blot analysis, we demonstrate that the p24/p55 ratio on targets, reflecting the amount of mature viral material passing from donor to target cells is also significantly reduced in the absence of Nef. In donor cells, without Nef, we observed a slight reduction of the p24/ p55 ratio as compared to cells infected with WT virus. This raises questions about how and which viral material is actually transferred from donors to targets. Recent fluorescence microscopy methods demonstrated that viral assembly, budding and release in the supernatant are rapid processes [70
]. It will be interesting to understand whether the dynamics of HIV-1 assembly are similar in the presence or in the absence of virological synapses [72
] and if Nef plays a role in Gag assembly. Hubner and colleagues [63
] showed that Gag proteins moving across the synapse toward the target cells originate from regions of the donor cells close to the cell-contact site. Thus, in the absence of Nef, the decreased quantity of viral material transferred to targets could be a direct consequence of the reduced amount of Gag proteins at the plasma membrane of donor cells. Moreover, Nef modulates the lipid content and the nature of the cellular proteins present at the cell membrane, a process that may enhance viral infectivity [74
]. Since rafts and other membrane microdomains polarize at the site of viral transfer and are considered as privileged Gag assembly sites [9
], a modulation of the composition of the cellular membranes could affect Gag/GagPol assembly and subsequent processing. At the membrane, virions bud mostly as immature particles. Maturation starts during the late phases of or immediately after budding, when the autocatalytic cleavage of the PR activates this enzyme to produce the mature viral core [77
]. Nef directly binds the GagPol-p6* transframe protein, but not Gag-p6, and redirecting Nef to the endoplasmic reticulum inhibits the activity of Nef on Gag processing and virion production [78
]. Gag processing occurs during Gag/GagPol assembly at the plasma membrane, but not during membrane trafficking [79
] and the ratio between Gag and GagPol significantly impacts the intracellular distributions of mature Gag and the infectivity of the viral particles produced [79
]. Thus Nef binding to p6* may modulate the trafficking of the viral structural proteins and affect their processing.
The increase of HIV-1 p24 in membrane fractions induced by Nef is visible in HIV-1 infected cells (HeLa or primary CD4+ lymphocytes), as well as in HeLa cells expressing only GagPol. In contrast, Nef does not seem to promote p55 accumulation at the membrane when this viral protein is expressed alone. This suggests that the increase of p55 and p24 at the plasma membrane observed in infected cells might depend on the interaction of Nef with the GagPol proteins synthesized during viral replication. Of note, a part of the p24 signal detected in the membranes fraction could also result from virus being released and re-internalized [9
At later time points, this altered transfer of viral material is associated with a reduction of productive infection of target cells. This is demonstrated here after 24h of viral replication in Jurkat cells that have been separated after 2h of contact with Hela donor cells. Of note, a positive effect of Nef on viral replication was observed in primary CD4+ T cells when the sources of infection were either cell-free viruses (Figure a) or infected lymphocytes co-cultivated with autologous targets (not shown). The effect of Nef on Gag proteins amount and trafficking described help explain the slower kinetics of replication observed for
Nef viruses. Portillo and colleagues [32
] showed that cell-to-cell transmission significantly increases the number of copies of viral DNA integrated in the host genome. The differences in the quantity and quality of transferred viral material between WT and
Nef could thus affect the number of integration events and the amount of viral proteins produced per infected cells. Moreover, there are less mature virions transferred in the absence of Nef, so the fusion events at the virological synapse could also be affected [81
]. Additionally, we show that in the absence of Nef, the amount of transferred immature viral material is significantly increased. It has been recently proposed [40
] that immature viral particles are first endocytosed and then undergo maturation inside the target cell. Without Nef, this “excess” of immature viral material may necessitate a longer maturation time, which may further delay viral replication.
The profile of maturation of Gag proteins in extracellular virions is considered to be similar with or without Nef ([82
], and not shown). However, these analyses were generally performed on viral particles harvested after long periods of times (i.e. after a few hours to a few days). This may have masked short-term effects of Nef on the kinetics of viral maturation after extracellular release.