Immune selection in vitro reveals human immunodeficiency virus type 1 Nef sequence motifs important for its immune evasion function in vivo

Abstract

Human immunodeficiency virus type 1 (HIV-1) Nef downregulates major histocompatibility complex class I (MHC-I), impairing the clearance of infected cells by CD8(+) cytotoxic T lymphocytes (CTLs). While sequence motifs mediating this function have been determined by in vitro mutagenesis studies of laboratory-adapted HIV-1 molecular clones, it is unclear whether the highly variable Nef sequences of primary isolates in vivo rely on the same sequence motifs. To address this issue, nef quasispecies from nine chronically HIV-1-infected persons were examined for sequence evolution and altered MHC-I downregulatory function under Gag-specific CTL immune pressure in vitro. This selection resulted in decreased nef diversity and strong purifying selection. Site-by-site analysis identified 13 codons undergoing purifying selection and 1 undergoing positive selection. Of the former, only 6 have been reported to have roles in Nef function, including 4 associated with MHC-I downregulation. Functional testing of naturally occurring in vivo polymorphisms at the 7 sites with no previously known functional role revealed 3 mutations (A84D, Y135F, and G140R) that ablated MHC-I downregulation and 3 (N52A, S169I, and V180E) that partially impaired MHC-I downregulation. Globally, the CTL pressure in vitro selected functional Nef from the in vivo quasispecies mixtures that predominately lacked MHC-I downregulatory function at the baseline. Overall, these data demonstrate that CTL pressure exerts a strong purifying selective pressure for MHC-I downregulation and identifies novel functional motifs present in Nef sequences in vivo.

Fig 1

Phylogeny of nef quasispecies in the absence and presence of immune selection by Gag-specific CTLs. Plasma nef sequences from nine subjects (input sequences, n = 94, blue circles), nef sequences passaged in the absence of CTL selection (control sequences, n = 71, green squares), and nef sequences passaged under Gag-specific CTL selection (selected sequences, n = 67, red triangles) were aligned with NL4-3 nef to create a neighbor-joining phylogenetic tree. Independent clusters for each subject had >99% bootstrap support, with the exception of subjects 00035 and 00039, whose sequences were previously found to be related (40). Significant clustering of CTL-selected sequences (bootstrap values of >70%) is indicated by asterisks.

Fig 2

CTL selection exerts evolutionary pressure on the nef quasispecies. The nef quasispecies sequences were examined for changes resulting from selection by the Gag-specific CTLs, comparing the input, control, and selected sequences. (A) The percentages of sequences with nonsense mutations (frameshift and/or early stop mutations) are plotted for each group across all subjects. (B) For nef sequences from each subject, pairwise diversity (calculated for each group using 500 bootstrap replicates to give the standard error of the mean) is plotted for each group. (C) For each subject, the change in nef diversity due to CTL selection (comparing the control and selected groups) is plotted, and the median across all subjects is indicated. An asterisk indicates a P value <0.05 for the difference between control and selected sequences.

Fig 3

Passaging of HIV-1 in the presence of Gag-specific CTLs results in purifying selection of nef. The input, control, and selected sequences were evaluated for evidence of selective pressure as reflected by the dN/dS ratios. (A) Maximum-likelihood estimates of the global dN/dS ratios with 95% CIs are plotted for the three groups of sequences. An asterisk indicates nonoverlapping CIs. (B) Site-by-site analysis for CTL selection was performed by multiple methods; shown here are results from the SLAC method. The plot shows the estimated dN/dS ratios for codons (numbered according to the HXB2 numbering system) that demonstrated significant selection (P < 0.05) by both the SLAC and FEL methods.

Fig 4

Structural locations and conservation of amino acids associated with MHC-I downregulation. (A) The 13 codons determined to be under purifying selection are indicated in the predicted three-dimensional structure of the Nef protein (composite crystal structure kindly provided by Art F. Y. Poon). The probability of each amino acid, based on an alignment of all complete, nonrecombinant Nef sequences, including genotypes A to K, submitted to the LANL HIV-1 sequence database through 2010 (n = >2,100 sequences), was calculated for the sites under purifying selection shown in Table 1 (B) and sites previously identified as important for Nef MHC-I downregulation (C). term, terminus.

Fig 5

Downregulation of MHC-I by mutant forms of Nef identified by CTL immune selection. Eight mutant forms with changes at the seven selected sites with no previously reported role in Nef function were individually introduced into NL4-3 recombinant reporter viruses. Specific amino acid changes were selected on the basis of their presence in primary plasma isolates before selection, except H171A. Their abilities to downregulate HLA-A*0201 were measured by flow cytometry and compared to those of NL4-3 Nef, Delta Nef, and M20A Nef, a mutant form specifically deficient in MHC-I downregulation. (A) Summary of the average HLA-A*0201 downregulation of each mutant form relative to NL4-3 Nef based on at least three separate infections. A single asterisk indicates a significant difference from NL4-3 with a P value of <0.05; double asterisks, P < 0.001. (B) Histogram plots of the levels of HLA-A*0201 on cells infected with a virus that produces either mutant Nef (filled histograms, mutation shown in the upper right corner) or NL4-3 Nef (open histograms).

Fig 6

CTL-selected Nef sequences have preserved MHC-I downregulatory function. Levels of A*0201 on the surface of cells infected with reporter viruses carrying input versus CTL-selected Nef quasispecies were measured by flow cytometry. (A) Histogram plots of A*0201 on cells infected with wild-type NL4-3 Nef and M20A Nef, deficient in MHC-I downregulation (top panels) and for subjects 00021 and 00022 before and after immune selection (middle and bottom panels, respectively). Open histograms are cells without Nef; and filled histograms are cells infected with the Nef allele shown in the upper right corner. (B) Summary plots of Nef quasispecies from subjects 00021, 00022, 00030, 00034, and 00037 for input and selected viruses, as well as a wild-type NL4-3 virus control that underwent CTL selection. The error bars indicate the standard deviation for three independent experiments with each input virus group. Note that only one sample of each selected virus was available for testing.