|Home | About | Journals | Submit | Contact Us | Français|
The human immunodeficiency virus type 1 (HIV-1) nef gene is a crucial determinant in AIDS disease progression. Although several in vitro activities have been attributed to the Nef protein, identifying the one critical for in vivo pathogenicity remains elusive. In this study, we examined a large number of nef alleles derived at various time points from 13 perinatally infected children showing different progression rates: six nonprogressors (NPs), three slow progressors (SPs), and four rapid progressors (RPs). The patient-derived nef alleles were analyzed for their steady-state expression of a Nef protein, for their relative ability to downregulate cell surface expression of CD4 and major histocompatibility complex class I (MHC-I) and for their capacity to bind the clathrin adaptor AP-1 complex. We found that NP-derived nef alleles, compared to nef alleles isolated from SPs and RPs, had reduced CD4 and MHC-I downregulation activities. In contrast, SP- and RP-derived nef alleles did not differ and efficiently downregulated both CD4 and MHC-I. AP-1 binding was a conserved function of primary nef alleles not correlated with clinical progression. Defective Nef proteins from NPs, rather than sharing common specific changes in their sequences, accumulated various amino acid substitutions, mainly located outside the conserved domains previously associated with Nef biological properties. Our data indicate that Nef-mediated downregulation of cell surface CD4 and MHC-I significantly contributes to the expression of the pathogenic potential of HIV-1.
Following human immunodeficiency virus type 1 (HIV-1) infection, few individuals remain healthy with normal CD4+ cell counts for at least a decade and, therefore, are classified as long-term nonprogressors (6, 19, 28, 38). Some of these long-term nonprogressors were shown to carry viruses with gross deletions in their nef genes (13, 26, 42), strongly supporting the crucial role of the Nef protein in viral replication and progression toward AIDS, as clearly demonstrated in animal models (21, 24).
Several Nef in vitro activities that may contribute to AIDS pathogenesis have been described. The best characterized Nef property is the ability to decrease cell surface expression of the HIV receptor CD4. Nef binds to both CD4 and the components of the intracellular protein sorting machinery, such as the clathrin adaptor protein complexes AP-1 and AP-2, thus promoting receptor internalization and inhibiting receptor recycling to the membrane (reviewed in reference 37). Nef-mediated CD4 downregulation in infected cells may impair CD4+ helper T-cell immune functions (45), prevent superinfection (2) and allow new viral particle release (27, 41). By acting on an apparently different endocytic pathway, which involves the PACS-1/AP-1 complex and ARF6 (3, 39), Nef decreases the cell surface expression of MHC-I molecules, thus preventing lysis of HIV-1 infected cells by cytotoxic T lymphocytes (CTL) (11, 39, 44). By its ability to interact with signaling molecules, the Nef protein can support viral replication and spread as well as protect HIV-1-infected cells from apoptosis (reviewed in reference 14). Nevertheless, the relative contribution of the reported Nef functions for viral pathogenesis in vivo is unclear.
At present, functional studies performed with representative nef alleles derived from patients with different stages of disease are, in part, controversial. In some reports, nef alleles derived from individuals with nonprogressive infection were shown to be defective in CD4 downregulation (34), in enhancement of viral infectivity and replication (15), or in both activities (7, 47). However, in other studies nef alleles from patients with no or slow progression were not found to differ from progressor patients in their ability to downregulate CD4 (35, 40) or to increase viral infectivity and replication (22). As to MHC-I downregulation, nef alleles from nonprogressors, when compared to those from rapid and slow progressor patients, were found to be more efficient in one study (7), but equivalent in another (47). Since Nef in vitro activities may depend on the cellular system in which they are analyzed, some conflicting results may be explained by differences in experimental conditions (discussed in reference 15). On the other hand, due to the large variability of the nef gene, functional studies on a single allelic variant, although containing representative amino acid variations, may greatly constrain the analysis of possible Nef phenotypes in a given patient.
This is why in a previous study we isolated and sequenced a large number of nef alleles sequentially derived from perinatally infected children with different progression rates: six nonprogressors (NPs), three slow progressors (SPs), and four rapid progressors (RPs) (8). We found that a higher proportion of disrupted sequences were derived from NPs compared to the other patients. Nef proteins contained several mutations at various amino acid residues, although none of these changes was consistently associated with a specific progression rate. However, some residues localized in the folded core domain at CTL epitopes were frequently changed in NP-derived Nef proteins, while other residues located at the N- and C-terminal loops varied frequently in proteins derived from RP and SP patients (8). The role of these residues in the structure and function of the Nef protein is unknown, although it is conceivable that epitope variations in NP-derived proteins can induce a broad range CTL response thus limiting viral replication in NP patients.
In this study, we analyzed nef alleles from NP, SP, and RP patients for their ability to downregulate CD4 and MHC-I molecules and found that nef alleles from NP individuals were less efficient in both activities. These findings suggest that Nef-mediated CD4 and MHC-I downregulation is critical for in vivo viral replication and disease progression.
Patients' characteristics as well as isolation and subcloning of patients' nef genes are described elsewhere (8). Briefly, we studied 13 perinatally HIV-1-infected children of Italian origin showing different modality of disease progression: six NPs, three SPs, and four RPs. NPs were children that reached at least 11 years of age with no or mild HIV-1-associated signs or symptoms (classified in stage N1 or A1 according to the Centers for Disease Control [CDC] classification for children . NPs did not receive therapy, with the exception of NP1 and NP6, who showed a decline in CD4+ cell number and increased viral loads around 12 and 11 years of age, respectively, and immediately after started a highly active antiretroviral therapy. The SP patients were comparable by age to NPs but had exhibited moderate signs and symptoms between 7 and 9 years of age (SP1, SP2, and SP3 were classified B2, A2, and B1 by CDC, respectively) and had received highly active antiretroviral therapy. RP patients had an onset of severe clinical manifestations (CDC category C) and/or profound immune suppression (CDC category 3) within the first 2 years of life and had received reverse transcriptase inhibitors. RP1 and RP4 also started highly active antiretroviral therapy at 39 and 34 months, respectively. Patients RP2, RP3, and RP4 died of AIDS-defining illness at 48, 36, and 38 months of age, respectively. Informed consent was obtained from all study participants or their guardians.
nef sequences from patients were amplified by nested PCR on peripheral blood mononuclear cell (PBMC)-derived genomic DNA obtained from two to five time points for each patient and subcloned into a retroviral vector, Pinco (20) (kindly provided by P. G. Pelicci, Milan, Italy). Based on length, at least five representative clones for each DNA sample were selected for subsequent analysis, with the exception of the earliest samples of three nonprogressors from whom either no nef sequences were amplified (NP4 and NP6 at 10 years) or only two recombinant clones were obtained (NP5 at 16 years). The recombinant clones contained a nef gene under the transcriptional control of the full-length long terminal repeat from the Moloney murine leukemia virus and are identified by their name: the first two letters specify the progression grouping, the first number specifies the individual patient and the last number identifies the clone. The retroviral vector contained also the ψ packaging sequences from Moloney murine leukemia virus, the gene encoding the enhanced green fluorescence protein under the control of the cytomegalovirus promoter, the EBNA-1 gene, and the origin of replication from the Epstein-Barr virus, which confer stable episomal maintenance in mammalian cells (20).
As controls for our assays, we subcloned (by standard insertional overlapping PCR) into the BamHI/EcoRI restriction sites of the Pinco vector the nef gene derived from the HIV-1 NL4-3 clone (NIH Reagent Program), generating Pinco-nefwt. To the NL4-3-derived nef gene and to some nef genes derived from patients the hemagglutinin (HA) epitope was added in frame to the 3′ end by PCR with the following primers: 5′-CGGGATCCTTCTCCACATACCTAGAAGA-3′ (forward) and 5′-GGAATTCATCACGGGAGGCTAGCGTAATCTGGAACGTCGTAGCAGTTCTTCAAGTACTCCGGATG-3′ (reverse). The sequences in bold and italic correspond to the HA epitope and to BamHI and EcoRI restriction sites, respectively. The resulting nef-HA clones were subcloned into the BamHI/EcoRI restriction sites of the Pinco vector and sequenced on both strands.
To create RP4-5R124W and NP2-11L136P, the specific mutations were introduced into RP4-5 and NP2-11, respectively, by standard site-directed mutagenesis based on recombinant overlapping PCR. Both mutants have been sequenced on both strands. The pGEX-6P-1 vector (Amersham Pharmacia Biotech) was used to express glutathione S-transferase (GST) fusion proteins with NL4-3-derived Nef protein and several Nef variants derived from patients. The coding region of nef genes was amplified from Pinco constructs by PCR with the following primers: 5′-CGGGATCCATGGGTGGCAAGTGGTCAAAA-3′ (forward) and 5′-GGAATTCAGCGGAAAGTCCCTTGTA-3′ (reverse). The italic sequences indicate BamHI and EcoRI restriction sites. The PCR products were subcloned in the BamHI/EcoRI sites of pGEX-6P-1 and sequenced on both strands.
HeLa-CD4+ cells (ARP019; MRC AIDS Reagent Project) were maintained in Dulbecco's modified Eagle's medium (D-MEM; Euro Clone) supplemented with 10% fetal bovine serum (FBS; Gibco-BRL, Rockville, Md.), 2 mM l-glutamine, 100 U of penicillin-streptomycin/ml, and 600 μg of G-418/ml (70% microbiological potency). Phoenix-ampho cells (kindly provided by G. Nolan, Stanford, Calif.) were maintained in D-MEM supplemented with 10% FBS, 2 mM l-glutamine, and 100 U of penicillin-streptomycin/ml. RMA-S-A2 (kindly provided by V. Barnaba, Rome, Italy) and Jurkat E6-1 cells (ARP027; MRC AIDS Reagent Project) were maintained in RPMI 1640 medium (Gibco-BRL) supplemented with 10% FBS, 2 mM l-glutamine, 100 U of kanamycin/ml, 1 mM sodium pyruvate, and 1× nonessential amino acids.
To generate infectious retroviral particles, the recombinant Pinco-derived plasmids were transfected into the amphotropic packaging Phoenix cell line by the calcium-phosphate/chloroquine method (25). For the infections with recombinant retroviruses, 1.2 × 104 HeLa-CD4+ cells were plated on a 12-well cell culture cluster in complete medium without G-418 24 h before the infection. Infections were performed by incubating cells for 8 h at 37°C with 500 μl of freshly prepared Phoenix cell supernatants/well filtered through a 0.45-μm-pore-size filter (Millex-GP; Millipore) and supplemented with 8 μg of Polybrene/ml. Cells were then cultured in complete medium without G-418 for 40 h before analysis. RMA-S-A2 cells were resuspended at 1.4 × 106 cells/ml in the viral supernatants, infected by four spin-infection cycles (20), placed back in growth medium and cultivated 40 h before analysis. To obtain suboptimal viral infections, RMA-S-A2 cells were infected by two spin-infection cycles. For each recombinant retrovirus, three independent infections were carried out for each cell line.
Forty hours after transfection with a Pinco-nef clone, Phoenix cells were washed twice with PBS and were then lysed in JS buffer (50 mM Tris-HCl, pH 8.0, 1% Triton X-100, 1.5 mM MgCl2, 150 mM NaCl, 5 mM EGTA, 10% glycerol, 0.2 mM phenylmethylsulfonyl fluoride, 1 μg of pepstatin/ml, 5 μg of aprotinin/ml, and 1 μg of leupeptin/ml) for 20 min on ice. After centrifugation at 12,000 × g for 10 min, equal amounts of clarified supernatant containing cell lysates were resolved on 14% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted onto nitrocellulose membrane (Schleicher & Schuell, Keene, N.H.). Immunoblot analysis of Nef proteins was performed by standard methods with a sheep polyclonal anti-HIV Nef serum (ARP444; MRC AIDS Reagent Project) diluted 1:350 in blocking solution. After washing, the membranes were incubated with horseradish peroxidase-conjugated G-protein (Bio-Rad, Hercules, Calif.) diluted 1:2,000 in blocking solution. To analyze HA-tagged Nef proteins, an anti-HA monoclonal antibody (sc-7392; Santa Cruz Biotechnology, Santa Cruz, Calif.) diluted 1:1,000 was used followed by horseradish peroxidase-conjugated sheep anti-mouse immunoglobulin antibodies (Amersham Pharmacia Biotech) diluted 1:10,000. The proteins were detected with the ECL system (Amersham Pharmacia Biotech). In order to evaluate the levels of transfection, the membranes were stripped and reprobed with an anti-GFP monoclonal antibody (Clontech, Palo Alto, Calif.) diluted 1:10,000. The amount of proteins was quantified by densitometric analysis.
Cells infected with Pinco-nef recombinant retroviruses, as well as mock-infected cells, were collected from culture dishes, washed three times with phosphate-buffered saline (PBS) containing 1% FBS and resuspended in the same solution. HeLa-CD4+ cells were reacted with saturating amounts of phycoerythrin (PE)-conjugated monoclonal antibody MT310 specific for CD4 (Dako), for 30 min on ice. After staining, cells were washed three times with PBS-1% FBS and resuspended in FACSFlow (Becton Dickinson) before analysis. RMA-S-A2 cells, were stained with saturating amounts of biotinylated monoclonal antibody anti-human HLA-A2 (One Lambda, Inc.) for 30 min on ice. Cells were then washed three times with PBS-1% FBS and incubated with indodicarbocyanine-conjugated streptavidin (Southern Biotechnology Associates, Inc.) for 30 min on ice. Finally, cells were washed, resuspended in FACSFlow and analyzed.
Fluorescence intensities for GFP and either for CD4 or HLA-A2 were analyzed by two-color flow cytometry on a FACSCalibur with CellQuest software (Becton Dickinson). Parameters were set in order to acquire 104 living cells/sample. The percentage of infected cells was evaluated as the percentage of GFP-expressing cells (90 to 100% in a typical experiment). The amounts of cell surface CD4 or HLA-A2 molecules were determined as the geometric means of the specific fluorescence in cells gated (R1 region) for medium GFP fluorescence (102 to 103 green fluorescence). When indicated, cells were infected with smaller amounts of viral supernatant and gated in a wider region (R2) comprising also cells with low GFP expression (101 to 103 green fluorescence units). The relative downregulation efficiency was calculated for each nef variant considering as 100% of activity the downregulation of CD4- and HLA-A2-specific fluorescence observed with the NL4-3 nef allele.
Total bacterial extracts expressing NL4-3 Nef and Nef variants derived from patients fused to GST were generated and purified on glutathione-Sepharose beads according to the manufacturer's instructions (Amersham Pharmacia Biotech). For the binding studies, growing Jurkat E6-1 cells were washed twice in PBS and lysed in JS buffer as described above for Phoenix cells. Equal amounts (100 μg) of GST or GST fusion proteins immobilized on beads were incubated with gentle rotation for 4 h at room temperature with 2 mg of Jurkat cell lysates in a final 1,500-μl volume of PBS-T buffer (1× PBS, 0.1% Triton X-100 plus protease inhibitors). The beads were then collected by centrifugation at 500 × g for 5 min followed by supernatant removal, and washed six times at room temperature by addition of 1 ml of PBS-T buffer, inversion three times, and centrifugation. The bound proteins were eluted by boiling in reducing sample loading buffer, separated by 10% SDS-PAGE and analyzed by immunoblotting as decribed above with an anti-γ-adaptin (AP1) (clone 100/3, Sigma) antibody diluted 1:100 and with an anti-GST (sc-459; Santa Cruz Biotechnology, Santa Cruz, Calif.) antibody diluted 1:10,000.
The chi-square test was used to compare the range of CD4 and MHC-I downregulation activity (<25%, 25 to 50%, and >75% activity, considering 100% activity that of NL4-3 nef) of nef alleles for the NP, SP, and RP groups of patients. Comparisons of mean CD4 and MHC-I downregulation activities of nef alleles for the three progression groups were performed by variance analysis. All statistical calculations were done with the SPSS/PC+ software (version 5.0; SPSS Inc., Chicago, Ill.).
The nef sequences can be retrieved from GenBank with accession numbers AY116676 to AY116854 and AY254895 to AY254897.
We have previously characterized 182 nef alleles isolated at various time points from NP, SP, and RP children and found that the frequency of defective nef alleles containing large deletions, premature stop codons, and mutations in the initiation codon was significantly higher in NPs (26%) compared to RPs (14%) and SPs (5%) (8). Beside gross defects, small in-frame deletions/insertions or alterations in conserved as well as in variable regions occurred in several patient-derived nef genes. Since such changes could interfere with Nef protein stability, we analyzed by immunoblotting the steady-state expression of Nef proteins in transfected cells expressing nef genes isolated from patients. The level of the GFP protein encoded by the vector used (20) was analyzed in order to evaluate transfection efficiencies (Fig. (Fig.1).1). This analysis was performed on all nef alleles isolated from patients at the indicated time points (Fig. (Fig.1),1), including those that were grossly defective, and with the nef gene derived from the NL4-3 viral strain as a positive control (Pinco-nef).
As expected, the Nef protein could not be detected in extracts of cells expressing nef alleles with disrupted open reading frames due to the presence of defects such as mutations in the initiator codon (NP2-18), large deletions (NP1-13, NP2-7, NP2-8, NP2-9, NP2-10, NP4-1, NP4-7, NP4-8, NP4-9, NP6-2, NP6-3, NP6-4, NP6-5, NP6-6, NP6-8, NP6-9, NP6-10, and RP1-18), premature stop codons (NP6-11, SP3-12, RP1-13, and RP3-1) and both large deletions and premature stop codons (NP6-1, NP6-12, RP1-1, and RP4-6). However, two nef alleles, NP3-9 and SP2-3, that contained stop codons very close to the 5′ end (at positions 47 and 39, respectively, with regard to the NL4-3 nef sequence) expressed Nef proteins with an apparent molecular mass of 21 kDa that was consistent with the beginning of translation at the second methionine codon (position 58). Besides, a peptide of 14 kDa or smaller reacting with the anti-Nef antibody was detected in extracts of cells expressing some nef alleles with premature stop codons derived from patient RP1 (RP1-2, RP1-7, RP1-8, and RP1-9). All other prematurely terminated nef alleles did not express Nef-derived peptides that could be detected even with long immunoblot exposure (data not shown).
The remaining 149 nef alleles encoded for Nef proteins with an apparent molecular mass ranging between 25 and 35 kDa. Twelve of these alleles expressed proteins that were barely detectable by immunoblotting, albeit GFP levels were comparable to those of other samples (Fig. (Fig.1).1). In order to test whether such proteins were unstable or reacted weakly with the anti-Nef antiserum, we expressed them as HA-tagged proteins and analyzed their expression levels by immunoblotting with an anti-HA monoclonal antibody. The majority of the HA-tagged Nef proteins gave strong signals in the anti-HA immunoblotting analysis comparable to that of the NL4-3-derived HA-Nef protein, indicating that they contained amino acid substitutions that impaired efficient recognition by the anti-Nef serum but did not alter the protein expression levels (data not shown). Conversely, HA-RP4-5 and HA-NP2-11 proteins were weakly detected by both the anti-Nef and the anti-HA antibody giving signals which corresponded to less than 5% of that specific for the NL4-3-derived HA-Nef protein (Fig. (Fig.2A).2A). These results suggested that RP4-5 and NP2-11 proteins contained mutations that destabilized the protein structure, as confirmed by pulse-chase analysis of protein stability (data not shown).
By sequence alignment analysis, indeed, both proteins contained one substitution at a highly conserved residue (W124→R in RP4-5 and P136→L in NP2-11, numbered with regards to the NL4-3 Nef protein) that was never mutated in other variants derived from the same patients and that is buried in the Nef protein core domain (29). The reverse mutations at such residues (R124→W and L136→P in the RP4-5R124W and in NP2-11L136P mutants, respectively) restored normal levels of Nef protein expression (Fig. (Fig.2B).2B). Therefore, we concluded that the two mutations selected in vivo in the RP4-5 and NP2-11 variants abolished the structural integrity of the Nef protein.
To test whether Nef proteins derived from our study population differed in their ability to downregulate cell surface CD4 and MHC-I, a retrovirus-based transduction system and two-color flow cytometry analysis were employed. HeLa-CD4+ and a mouse T-cell line expressing human HLA-A2 molecule, RMAS-A2, were chosen to study Nef activity on CD4 and MHC-I, respectively. Both cell lines were efficiently infected with the Pinco retrovirus as they became 90 to 100% GFP+ by flow cytometry analysis (Fig. (Fig.3).3). Upon infection with Pinco-nef, the red fluorescence intensity values specific for CD4 or MHC-I of GFP+ cells were decreased 7- to 10-fold, depending on the experiment, compared to Pinco-infected cells, thus showing that CD4 and MHC-I downregulation activities of a wild-type Nef protein could be readily measured in this assay.
For the quantitation of Nef-mediated CD4 or class I MHC downregulation, the geometric mean of red fluorescence was determined in cells expressing medium GFP levels (gated in the R1 region) and that comprise the vast majority of infected cells in a typical experiment (Fig. (Fig.3).3). The activity of each nef allele derived from NP, SP, and RP patients was tested and calculated as the percentage of NL4-3 Nef activity, as described in Materials and Methods. The immunoblotting analysis of the relative expression of Nef protein variants in infected cells (data not shown) confirmed results shown in Fig. Fig.1.1. The flow cytometric analysis of six nef alleles, two for each progression group, with abnormal CD4 and/or MHC-I downmodulation activity, is shown in Fig. Fig.3.3. Quantitative results obtained from the functional analysis of all nef alleles derived from patients are shown in Fig. Fig.4.4. The data obtained with disrupted nef alleles that, as expected, were completely impaired in both CD4 and MHC-I downregulation activities were omitted from Fig. Fig.44.
Several of the NP-derived nef alleles encoded Nef proteins capable of downregulating cell surface CD4 and MHC-I expression as efficiently as the nef gene of NL4-3 (Fig. (Fig.4A).4A). However, among nef alleles derived from NPs, one allele (NP2-17) had lost the ability to downregulate CD4, 2 alleles (NP2-13 and NP2-15) had no effect on MHC-I expression and three alleles (NP2-11, NP3-9, and NP4-4) had no effect on both CD4 and MHC-I (Fig. (Fig.33 and and4A).4A). Besides, three nef alleles (NP1-5, NP2-13, and NP2-16) had reduced CD4 downregulation activity (about 60 to 70% the activity of the NL4-3-derived nef gene), five alleles (NP1-6, NP1-10, NP2-5, NP2-14, and NP2-17) had reduced MHC-I downregulation activity (50 to 70%) and 3 (NP1-18, NP5-7, and NP5-8) were less active on the expression of both CD4 (40 to 70%) and MHC-I (14 to 70%) (Fig. (Fig.33 and and4A).4A). On the other hand, none of the SP-derived nef alleles had lost activity, although one allele (SP1-1) downregulated both CD4 and MHC-I less efficiently (70%) than the nef gene of NL4-3 and one allele (SP3-7) was less efficient (73%) in MHC-I downregulation (Fig. (Fig.33 and and4B).4B). Among RP-derived nef alleles, two (RP2-7, RP4-11) had no effect on CD4 expression and one (RP4-5) was inactive on both CD4 and MHC-I expression (Fig. (Fig.33 and and4C).4C). In the latter group of alleles, six (RP2-4, RP4-1, RP4-2, RP4-3, RP4-4, and RP4-9) had a CD4 downregulation activity reduced to 60 to 70% of that of the NL4-3-derived nef.
For each progression group, the occurrence of nef genes expressing Nef proteins with functional defects and of grossly defective nef genes was evaluated, and the mean CD4 and MHC-I downregulation activities were calculated (Table (Table1).1). The nef alleles from NPs downregulated both CD4 and MHC-I expression (64.4% and 63% mean activity, respectively) less efficiently than nef alleles from SPs (91.4% and 92.5% mean activity) and RPs (80.8% and 90.3% mean activity) (Table (Table1).1). The differences between groups were statistically significant (P values ranging from 0.047 to 0.005). On the other hand, the mean activities of CD4 and MHC-I downregulation did not significantly differ between SP- and RP-derived nef alleles. Since patients NP1 and NP6 cannot be considered bona fide long-term nonprogressors (see Materials and Methods), the average nef activities were also calculated by excluding these two patients from the NP group. Even in this case, the NP-derived nef genes downregulated both CD4 and MHC-I to a lesser extent (70.8% and 68% mean activity, respectively) if compared to nef alleles derived from SPs and RPs, although the difference in CD4 downregulation between NPs and RPs did not reach statistical significance.
In the CD4 and MHC-I downregulation assays, some nef alleles isolated from patients were slightly more active than the NL4-3-derived nef gene (see Fig. Fig.4).4). To further analyze these alleles, cells expressing suboptimal levels of the corresponding Nef proteins or NL4-3-Nef were analyzed for their cell surface expression levels of CD4 and MHC-I. By this analysis, two nef alleles derived from the latest time point of patient RP4 (RP4-9 and RP4-11) showed twofold-higher activity in MHC-I downregulation than did the NL4-3-derived nef (Fig. (Fig.5),5), while all other alleles were not consistently more efficient than the control for both activities (data not shown). Interestingly, the average MHC-I downregulation activities of RP-derived Nef proteins (103% that of NL4-3-Nef) and of SP-derived Nef proteins (98%) were similar, and both were significantly higher (P < 0.005 and 0.036, respectively) than that of NP-derived Nef proteins (85%). Conversely, differences in the average CD4 downregulation activity among Nef proteins derived from the three progression groups (87% for NPs, 97% for SPs, and 92% for RPs) were not significant.
During this work, several Nef protein variants with functional defects have been isolated from our study population. As expected, NP3-9 and SP2-3 proteins that had lost the N-terminal myristoylation motif required for Nef association with cellular membranes and for virtually all its biological activities (reviewed in reference 16) were not able to downregulate cell surface expression of CD4 and MHC-I. The RP4-5 and NP2-11 proteins that were found to be unstable due to the W124→R and P136→L mutations, respectively, had indeed no effect on the expression of both CD4 and MHC-I. The other defective Nef proteins carried mutations at various residues, some of which are highly conserved and probably important for Nef protein structure and function. The NP5-7 and NP5-8 proteins, which were inefficient for both CD4 (40%) and MHC-I (15%) downregulation activities, differed at a poorly conserved C-terminal residue (position 194 was a leucine in NP5-7 rather than histidine, as in all other NP5-derived variants), but both carried a P78→L substitution in the highly conserved proline-rich domain. Since the fully active NP5-1 variant differed from NP5-8 only at position 78, the P78→L substitution was by itself sufficient to reduce both activities with a stronger effect on MHC-I downregulation. In a recent study, proline 78 was indeed shown to be crucial for Nef activity in MHC-I downregulation (49). As for CD4 downregulation, all four prolines of the PxxP domain were shown to be irrelevant (49) or, in agreement with our data, to contribute to this Nef activity (1). Discrepancies could possibly be based on differences in background of selected nef alleles and/or in the experimental systems used. In the NP4-4 protein, the L165→S substitution, by destroying the dileucine motif required for binding to adaptor proteins, explains the loss of activity on CD4. However, the inability of the NP4-4 protein to downregulate MHC-I is not due to mutations in the motifs which are known to be important for this function [N-terminal α-helix, EEEE(62) and (PxxP)3(72)], but rather to one or more of those substitutions scattered along the protein that are not present in functional proteins derived from the same as well as other patients.
The two Nef variants with enhanced MHC-I downregulation activity, RP4-9 and RP4-11, contained an unusual feature consisting of two additional PxxP motifs, one located in the N-terminal loop and one located in the folded core domain, that could amplify the function of the (PxxP)3(72) domain (8). However, since all RP4-derived Nef variants contained the additional PxxP motifs and differed from each other at various amino acid positions, an extensive mutagenic analysis should be performed in order to identify those features that can increase the MHC-I downregulation activity of Nef.
In summary, by comparing the protein sequences of functionally characterized Nef variants, several residues of Nef that could contribute to a variable extent to CD4 and/or MHC-I downregulation activities were identified. Most of these residues are located at various positions in the Nef protein that are not yet credited with biological function and await further characterization.
The CD4 and MHC-I downregulation activities of the Nef protein can be genetically and functionally separated and seem to involve different endocytic pathways (reviewed in reference 16). However, both Nef activities require the association of Nef with the clathrin adaptor AP-1 complex which may serve to redirect sorting and exit of CD4 and MHC-I molecules from the Golgi apparatus or from endosomes (4, 17, 30). Binding of AP-1 to a Nef dileucine motif (LL165) is required for CD4 downregulation, but is dispensable for the activity on MHC-I (4, 12, 17, 33). Conversely, the Nef acidic domain (EEEE65) mediates the indirect binding to AP-1 through the PACS-1 sorting molecule and is crucial for Nef activity on MHC-I (3, 39) but superfluous for Nef-mediated CD4 downregulation (1, 18).
In order to analyze the ability of functional and defective Nef proteins derived from patients to associate with AP-1, we generated a series of GST-Nef fusion proteins and tested them for AP-1 binding in an in vitro assay. With this system, the AP-1 adaptor bound efficiently to the NL4-3-Nef protein but not to the GST alone (Fig. (Fig.6).6). As for Nef proteins derived from patients, we found that AP-1 binding was a conserved function of Nef, with the exception of a variant isolated from a nonprogressor patient, NP4-4, that contained few amino acid variations including a mutated dileucine motif (Fig. (Fig.6).6). In general, we observed that the AP-1 binding efficiency of Nef proteins derived from patients was extremely variable, ranging from 8% to 90% of that measured for the NL4-3-Nef protein. Most likely, several residues that are not strictly conserved in Nef proteins derived from patients contribute to the overall accessibility of Nef to the AP-1 complex. However, the amount of captured AP-1 did not correlate with the relative ability of downregulating either CD4 or MHC-I. As shown in Fig. Fig.6,6, Nef proteins that were as efficient as the NL4-3-Nef protein for both activities captured an amount of AP-1 that was less than that bound by NL4-3-Nef (down to 8% for the RP3-7 protein) or by Nef proteins that were defective for one or both functions. On the other hand, if comparing functional Nef proteins with defective or hyperactive variants derived from the same patient (for instance, NP5 and RP4), the AP-1 binding efficiency did not vary by a relative decrease or increase. Thus, with the exclusion of the NP4-4 variant, the functional variations here described for patient-derived Nef proteins cannot be attributed to the efficiency of forming a complex with the AP-1 adaptor. Binding to AP-1 was a conserved feature of Nef proteins derived from NPs, RPs, and SPs (Fig. (Fig.66 and data not shown) and did not correlate with the clinical progression of the patients.
In this study we analyzed distinct functions of a large number of nef alleles isolated at various time points from HIV-1-infected children with different rates of disease progression. Our results demonstrate that nef alleles derived from NP patients are less efficient than nef alleles derived from SP and RP children in the downregulation of cell surface expression of both CD4 and MHC-I. Moreover, SP- and RP-derived nef alleles did not differ in these Nef activities.
The functional defects that we observed in NP-derived nef alleles are in part due to structural defects, since mutations incompatible with protein expression were present in 26% of the sequences (8). In particular, several grossly deleted nef genes were isolated from patient NP6, in agreement with previous reports of long-term nonprogressors carrying HIV-1 with a nef deletion (13, 15, 26, 42). Interestingly, when NP6 reached 12 years of age, concomitant with an increase of the viral load and a decline in the clinical status, we were able to isolate a full-length nef gene that turned out to be functional for both CD4 and MHC-I downregulation activities. This result strongly supports the crucial role of Nef in HIV-1 replication in vivo and in disease progression. Besides, as discussed previously (8), the data indicate that patient NP6 was originally infected with a virus carrying a wild-type nef gene and that viral forms with a nef deletion have been positively selected in this patient, possibly by a strong antiviral immune response persisting for several years until infection has progressed. Therefore, emergence of full-length nef genes in long-term nonprogressors carrying nef-deleted viral forms should be monitored carefully.
Several nef alleles with disrupted open reading frames, coexisting with full-length nef alleles functional for both CD4 and MHC-I downregulation activities, were also isolated from patient RP1. However, most of the disrupted nef alleles derived from RP1 contained point mutations that could be more easily repaired than large deletions mainly found in NP patients. Moreover, four RP1-derived nef alleles expressed stable Nef proteins of reduced molecular weight that could have maintained some Nef properties, such as enhancement of viral replication, as already demonstrated for truncated Nef proteins found in vivo (10, 43). Indeed, as RP1 progressed toward the disease, functional alleles progressively outgrew the disrupted ones.
In addition to structural defects in nef alleles from NPs, the NP-derived Nef proteins were less efficient than proteins derived from SP and RP patients in both activities, although only differences in MHC-I downregulation reached statistical significance. At the protein level, there were many discernible differences between defective and functional Nef proteins as, more in general, there was a great intrapatient and intragroup variability in Nef protein sequences. This is why for only two variants, NP5-7 and NP5-8, has it been possible to clearly correlate a specific change (P78→L) to functional Nef defects. Mutations in conserved Nef motifs that are known to be important for function, such as the N-terminal myristoylation motif, the PxxP domain, and the dileucine motif, were found rarely in patient-derived Nef proteins. However, by comparing the protein sequences of the functionally characterized Nef variants, several residues located outside the conserved domains, potentially contributing to CD4 and/or MHC-I downmodulation activities, could be identified.
Apparently, rather than sharing common specific changes in their sequences, NP-derived Nef proteins accumulate various amino acid substitutions, including those described previously (8), that result in inefficient CD4 and MHC-I downregulation activities. Such NP-associated changes might interfere with Nef functions through several possible mechanisms, including the alteration of Nef protein structure and accessibility to cellular interacting proteins. Future studies on defective NP-derived Nef proteins, based on mutational analysis assisted by computer-based protein structure predictions, will help the understanding of those features in the Nef protein that regulate its right structural conformation and functional properties.
Apparently, binding of NP-derived Nef proteins to the clathrin adaptor AP-1 complex was conserved, with the exclusion of the NP4-4 variant that was indeed not functional. Besides, the amount of captured AP-1 in our in vitro binding assay did not correlate with the relative ability of Nef proteins derived from any progression group in CD4 or MHC-I downregulation. As previously shown for PAK (15) and Hck (47) kinases, in vitro association of Nef with AP-1 does not correlate with clinical progression.
Our data are only partly in agreement with previous studies in HIV-1-infected patients. Three groups reported that CD4 downregulation activity was reduced or abolished in long-term nonprogressor-derived nef alleles (7, 34, 47), and two groups reported that nef alleles isolated from slow progressor patients were as efficient as wild-type nef (35, 40). However, two groups reported that the ability to downregulate MHC-I expression was retained in long-term nonprogressor-derived nef alleles (7, 47). Carl and coworkers also showed that nef alleles obtained from RP patients at late stages of infection did not efficiently downregulate MHC-I and suggested that, in the immunodeficient host unable to mount an efficient antiviral CTL response, Nef activity on MHC-I has lost a selective advantage (7). On the contrary, we found that Nef proteins derived from RPs as well as SPs maintain an efficient MHC-I downregulation activity, as opposed to NP-derived Nef proteins. Moreover, two out of six Nef variants isolated from rapid progressor RP4 shortly before his death, RP4-9 and RP4-11, showed an enhanced MHC-I downregulation activity.
The discrepancies between our results and others may be due to differences in the selection of nef alleles and patient cohorts. First, rather than testing the function of single representative nef alleles displaying amino acid changes that were present in most of the alleles isolated at a given time point (7, 47), we examined, with few exceptions, an average of six nef alleles at two or more time points for each patient. As shown previously (8), primary nef alleles isolated from a PBMC sample display indeed a degree of sequence diversity that makes the selection of few representative specific changes very restrictive in the analysis of all possible variants. Second, our study population consisted of perinatally infected children rather than adults. HIV-1-infected children have a shorter disease incubation period and a more severe immunosuppression than do adults, which have been attributed to the relative incompetence of the immune system at the time of infection. In infected adults, HIV-specific CTLs represent the first detectable antiviral immune response and can be found at all disease stages even though their reduction in number is associated with disease progression. On the contrary, in infected children HIV-specific CTLs are less frequent and can be detected only in children with no or mild symptoms (6, 48). It is conceivable that, during primary infection in children, an efficient Nef mediated downregulation of MHC-I results in the lack of an early development of HIV-specific CTL responses, thus allowing a rapid viral spread and a severe HIV-induced immunosuppression. In pediatric NPs, the inefficient Nef activity in MHC-I downregulation that we observed may result in enhanced presentation of viral antigens and enable the infected host to mount early CTL responses able to control viral replication.
As opposed to NPs, the SP patients analyzed in this study harbored nef alleles that were as efficient as those isolated from RPs in both CD4 and MHC-I downregulation activities. It is possible that factors not considered in this study, such as exact timing of viral transmission (in utero versus neonatal), mother's viral load, and host genetic background, could have determined the ability of SPs to effectively control viral replication for at least 7 years. It is also possible that defects in other Nef activities rather than in CD4 and MHC-I downregulation could have contributed to the slow disease progression in these patients. Of note, we observed that several SP-derived Nef proteins contained mutations at the G3 residue that is required for an efficient Nef stimulation of viral replication and infectivity (8). Further functional investigation of patient-derived nef alleles should help to better understand the relative contributions of the different Nef activities to AIDS pathogenesis.
Experiments performed in animal models are in agreement with our results. In infected monkeys, the CD4 downregulation activity of Nef is crucial for in vivo viral replication and disease progression (23). Moreover, Nef-mediated MHC-I downregulation provides a selective advantage for viral replication in monkeys (36). Of note, monkeys infected with mutated viruses carrying a nef gene defective for MHC-I downregulation activity remained asymptomatic for about 1 year, when the virus reverted to wild-type forms; by that time animals infected with wild-type virus had developed simian AIDS and died (36). Unfortunately, antiviral immune responses were not quantitated in these studies.
Our results indicate a link between Nef-mediated CD4 and MHC-I downregulation and the progression rate in infected children. As shown by previous studies, a decrease of cell surface CD4 molecules can facilitate progeny virion release (27, 41), prevent superinfection (2), and impair helper T-cell function (45). Thus, CD4 downregulation by Nef can have a profound effect on disease induction in vivo by both stimulating viral replication and repressing virus-specific immune responses. The Nef-mediated reduction of cell surface MHC-I expression apparently does not contribute directly to viral replication (31, 32, 46), but protects HIV-1-infected cells from killing by CTLs (11, 39, 44). We envision that inefficient CD4 and MHC-I downregulation may result in progressive development of HIV-specific immunity in infected children. It will be relevant to examine whether NP patients harboring nef alleles defective in CD4 and MHC-I downregulation display more efficient virus-specific CD4+ T-cell and CTL responses than do RPs and SPs. Such studies may bring insights into the mechanisms by which HIV-1 can be controlled by the host immune system and, possibly, contribute to the design of an effective vaccine.
Since the HIV-1 Nef protein has been credited with a series of in vitro functions, other Nef activities beside CD4 and MHC-I downregulation analyzed in the present study may correlate with clinical outcome. However, the data presented here should encourage efforts to discover drugs targeting these two conserved activities of the Nef protein.
This study was supported by grants from the Istituto Superiore di Sanità, Rome, Italy, II, III, and IV National Program of AIDS Research, “Patogenesi, Immunità e Vaccino” (40B40, 40C37, and 40D.38), by the Ricerche Correnti of Children's Hospital “Bambino Gesù,” Rome, Italy, and by EU Shared Cost Action (BMH4-97-2262). G.D.M. was supported by a fellowship from the Istituto Superiore di Sanità, Rome, Italy.
We thank P. G. Pelicci, G. Nolan, and V. Barnaba for kindly providing reagents. We are indebted to M. Harris and R. Axel for materials obtained from the NIBSC Centralised Facility for AIDS Reagents supported by the EU Programme EVA and the United Kingdom Medical Research Council. We also thank Silvia Di Cesare for help with FACS analysis.