A surprising number of functional activities have been observed for Nef. Lymphoid cell activation and downregulation of CD4, CD3, CD28, and MHC class I have been assayed in our current experiments. In addition, infectivity enhancement and induction of Fas ligand have also been described as functional activities (40
). Nef has also been reported to induce lymphocyte chemotaxis toward infected macrophages (57
) and to inhibit proapoptotic signaling in infected cells (19
). Infectivity enhancement may relate in whole or in part to CD4 downregulation (34
) and increased cellular activation (3
). Induction of Fas ligand has been suggested to increase apoptotic death of reactive cells trying to respond to infection (60
The goal of our current studies was to obtain evidence for the contribution of one activity of this multitude, MHC class I downregulation, to the ability of SIV to replicate in rhesus monkeys. Although the MHC-selective NefΔ239-240 and Nef238stop/fs/fs mutations had no effect on lymphoid cell activation in the 221 cell assay or on downregulation of CD4, CD3, or CD28, we do not know whether these mutations affected other known or unknown activities. Nonetheless, the stronger CD8 responses in mutant-infected animals and the strong selection for unusual compensatory sequence changes that restored MHC-downregulating activity provide convincing evidence for a contribution of this activity to the persistent replication of SIV in rhesus monkeys. Although a number of other viruses encode gene products that interfere with MHC expression (7
), we are aware of only one other analogous report in which a functional role in immune evasion has been validated in an animal system as we have done here. Stevenson et al. (51a
) knocked out the class I MHC-downregulating K3 gene from murine gammaherpesvirus 68 and found stronger virus-specific CD8+
T-cell responses and lower viral loads in mice.
Mutations became fixed in the population of Nef sequences in monkeys infected with the MHC-selective mutants at an astonishing rate, resulting in about 7% amino acid changes per year. Even the highly variable gp120 envelope protein accumulates sequence changes at a rate of only about 2% per year (12
), more than three times less than the rate seen in Nef in the monkeys infected with the MHC-selective mutants in the current study. This suggests strong selective pressure for the appearance of sequence variants and strong selective advantage of the mutant forms that became fixed in the population. In a previous study (41
), a simple point mutation that eliminated MHC-downregulating activity reverted within a few weeks of monkey infection. In all clones analyzed at 16 and 32 weeks after infection in our current study, the difficult-to-revert NefΔ239-240 and Nef 238stop/fs/fs mutations were retained, forcing the virus to try to compensate by changing residues elsewhere in the Nef protein.
The changes in Nef progressively increased the ability to downregulate MHC from weeks 0 to 16 to 32 such that by week 32 the ability to downregulate MHC class I was similar to that of the parental SIV239 Nef. Based on the large number of sequence changes and the time required to restore full MHC-downregulating activity, the virus was clearly under strong selective pressure over a prolonged period to restore the activity; this provides unambiguous evidence for the contribution of MHC downregulation to the ability of SIV to replicate in monkeys. The ability to restore the activity by sequence changes quite distant in the linear sequence is testimony to the remarkable flexibility and multiplicity built into the Nef protein structure.
Surprisingly, despite similar functional activities and about 30% sequence identity at the amino acid level, HIV-1 nef
differs from SIVmac nef
in the location of mutations that will selectively knock out the ability to downregulate MHC class I (54
). Several groups have shown that mutation of HIV-1 nef
in the vicinity of the coding sequences for PXXPXXPXXP, corresponding to residues 101 to 110 in Fig. , will selectively knock out the ability to downregulate MHC class I (21
). In contrast, Swigut et al. (56) and our data here show that MHC-selective mutations map to the C-terminal region of SIV239 Nef. Minimal amino acid changes that restored MHC-downregulating activity to NefΔ239-240 and Nef238stop/fs/fs are interestingly located in the same general vicinity in the Nef linear sequence as the MHC-selective mutations in HIV-1 Nef. In fact, the S101P change seen in SIV Nef in all four animals in our current study creates a PXXPXXP sequence that in HIV-1 is essential for HIV-1 Nef's ability to downregulate MHC class I. Amazingly, more than 50% of the amino acid changes in SIV Nef in the monkeys infected with SIV239 NefΔ239-240 and SIV239 Nef238stop/fs/fs resulted in the consensus HIV-1 amino acid at that location. This could be viewed as an accelerated, partial recapitulation of the evolution that occurred naturally over millennia in the emergence of HIV-1. A converse evolution of HIV-1 to SIV-like nef
sequences was observed previously in monkeys infected with a recombinant SIV containing an HIV-1 nef
gene (SHIVnef) (2
CD8+ T-cell responses were significantly higher in monkeys that received the MHC-selective mutant SIVs than in monkeys that received parental SIV239, from 4 to 14 weeks postinfection. It seems likely that the evolution of sequence variants with restored or partially restored MHC-downregulating activity was responsible for decreased ability to discern effects on the number of virus-specific CD8 cells at later time points. Stronger CD8+ antiviral responses would be expected to translate into lower viral loads, but we were unable to demonstrate any statistically significant differences in viral loads among the groups, excluding of course the SIVΔnef animals. It seems likely that the evolution of restored MHC-downregulating activity minimized the magnitude and duration of any effects of the mutations on viral load. Other factors may also have contributed. The number of monkeys employed was necessarily limited, and significant differences may have been observed with much larger numbers of monkeys.
A controlling immune response is likely to consist of multifactorial components; the effects of the MHC activity of Nef influenced the levels of only one of these multifactorial components, virus-specific CD8+ cellular responses. SIV destruction of virus-specific CD4+ helper cells may have limited the effectiveness of increased numbers of virus-specific CD8+ cells that resulted from the loss of MHC-downregulating activity. Nonetheless, although we were not able to demonstrate significant differences in viral load, the selective pressure for sequence change to restore MHC-downregulating activity clearly indicates that the absence of this activity did serve to limit the replication of the virus at least somewhat.
A reasonable portion of the nef clones at 16 weeks lost the ability to downregulate CD28, but by 32 weeks the vast majority of clones behaved like the wild type for both CD28 and MHC downregulation. This result suggests that some of the sequence evolution between 16 and 32 weeks resulted in restoration of CD28-downregulating activity and further suggests that downregulation of CD28 is also a biologically relevant functional activity of Nef.
The large number of functional activities assigned to Nef has raised doubts about whether all are relevant. Our results validate MHC downregulation as a biologically relevant functional activity of Nef. It should be possible to use the SIV/rhesus monkey system as we have done here to validate the relative importance of other functional activities of Nef. Also, vaccine strategies that employ Nef expression as one component of the immunogen should modify the nef gene in such a way as to eliminate the ability to downregulate MHC class I.