Ethics statement
The study was approved by the Ethics Committees of Kumamoto University (RINRI-1340 and GENOME-342) and the National Center for Global Health and Medicine (NCGM-A-000172-01). Written informed consent was obtained from all individuals according to the Declaration of Helsinki.
Study participants
One hundred and seven treatment-naive HLA-C*12:02+ individuals chronically infected with HIV-1 subtype B and nine HIV-1 seronegative individuals were recruited. Clinical data for these individuals are shown in Supplementary Table 6. PBMCs were separated from whole blood. HLA genotypes of the HLA-A, -B, and -C alleles were identified at the NPO HLA Laboratory (Kyoto, Japan).
Cell lines
CD4.221-C*12:02 cells and RMA-S-C*12:02 cells were previously generated28,55. CD4.221-B*15:01 cells were generated by transfecting the HLA-B*15:01 gene into 721.221 cells expressing human CD4 as previously described56. These cell lines were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and 0.15 mg/mL hygromycin B.
HIV-1 clones and pseudotyping
NL4-3 PolV472A mutant virus was previously generated by introducing the PolV472A mutation into the NL4-3 infectious molecular clone (IMC) by site-directed mutagenesis of plasmid pNL4-328.
To produce stocks of VSV-G pseudotyped NL4-3 and NL4-3 PolV472A virus, 12 μg plasmid DNA encoding the relevant IMC and 1.8 μg of the lentiviral packaging plasmid pMD2.G encoding the VSV-G envelope protein (Addgene plasmid #12259) were co-transfected into HEK293T cells using 1 μg/μl polyethylenimine (PEI, Polysciences) in OptiMEM media. Viral supernatants were harvested 72 h later, clarified by centrifugation at 1500 × g for 10 min, filtered, and stored at −80 °C. The infectivity of viral stocks was determined using a colorimetric reverse transcriptase assay (Roche Life Sciences, Cat #11468120910).
HLA stabilization assay
The binding of peptides to HLA-C*12:02 was measured as previously described25,35. Briefly, RMA-S-C*12:02 cells were pre-cultured at 26 °C for 16 h to fully express and present empty HLA class I molecules on the cell surface and then incubated with different concentrations of PolIY11 and PolIY11-9A peptides at the same temperature for 1 h. Cells were then incubated at 37 °C for another 3 h. After incubation, cells were stained with anti-HLA-C mAb DT-912 (culture supernatant of hybridoma cell line DT-9) and, subsequently, with FITC-conjugated sheep immunoglobulin G (IgG) (Jackson ImmunoResearch Laboratories, #515-095-003, 1:100). Surface expression of HLA-C was measured by flow cytometry, FACSCanto II (BD Biosciences). The HLA expression index was calculated as follows: (MFI [mean fluorescence intensity] of RMA-S-C*12:02 cells pre-pulsed with peptide minus MFI of RMA-S-C*12:02 cells without peptide pulsing)/(MFI of RMA-S cells kept at 26 °C indefinitely minus MFI of RMA-S cells without peptide pulsing).
Ex vivo IFN-γ ELISPOT assay
1 × 105 PBMCs from HIV-1-infected HLA-C*12:02+ individuals and 100 nM of PolIY10, PolIY11, or Pol IY11-9A peptide were added to 96-well polyvinylidene plates (Millipore) that had been coated overnight with 5 μg/ml anti-IFN-γ mAb 1-D1K (Mabtech, #3420-3-1000, clone 1-D1K). The plates were incubated for 16 h at 37 °C in 5% CO2, and then plate-bound IFN-γ spots were detected as previously described in detail15. The spots were counted with an Eliphoto-Counter (Minerva Teck). The number of CD8+ T cells in PBMCs was counted by flow cytometry. The number of spots was calculated per 106 CD8+ T cells. 200 spots/106 CD8+ T cells was defined as the cutoff for a positive response as described previously15.
ICS assay
CD4.221-C*12:02 cells pre-pulsed with PolIY10, PolIY11, or PolIY11-9A peptide, CD4.221-B*15:01 cells pre-pulsed with PolLY9, PolIY10, or PolIY11 peptide, CD4.221-C*12:02 cells infected with NL4-3 or PolV472A mutant virus, and CD4+ T cells infected with NL4-3 or NL4-3 PolV472A virus were used as stimulator cells. These cells were added to a 96-well plate together with cultured T cells, and the cells were incubated for 4 h at 37 °C with brefeldin A (10 μg/ml). The cells were then stained with Pacific blue (PB)-labeled anti-CD3 monoclonal mAb (Biolegend, #300330, clone HIT3a, 1:50), Allophycocyanin (APC)-labeled anti-CD8 mAb (Biolegend, #300912, clone HIT8a, 1:50), and LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit (Invitrogen, Cat #L34975) and subsequently fixed with 4% paraformaldehyde and incubated in permeabilization buffer (0.1% saponin–5% FBS–phosphate-buffered saline (PBS)). Thereafter, the cells were stained with Phycoerythrin (PE)-labeled anti-gamma interferon (anti-IFN-γ) mAb (Biolegend, #502509, clone 4S.B3, 1:50). Staining data were acquired on a FACSCanto II (BD Biosciences) and analyzed using FlowJo 10.7.1 software.
Detection of HIV-1-specific T cells using HLA-C*12:02 tetramers
HLA-C*12:02-PolIY11 or HLA-C*12:02-PolIY11-9A-tetramers were generated as previously described55. PBMCs from HIV-1-infected HLA-C*12:02+ individuals or HIV-1-specific T cell clones were stained with PE-conjugated tetramers at 100 nM and/or APC-conjugated tetramers at 100 nM at 37 °C for 30 min. The cells were then washed twice with RPMI 1640 medium containing 5% fetal calf serum (R5), followed by staining with FITC-labeled anti-CD8 mAb (BioLegend, #300906, clone HIT8a, 1:50) and 7AAD (BD Pharmingen, #559925, 1:25), or FITC-labeled anti-CD3 mAb (Biolegend, #300440, clone UCHT1, 1:50), PerCP-Cyanine5.5-labeled anti-CD8 mAb (Biolegend, #301032, clone RPA-T8, 1:200), and LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit (Invitrogen, Cat #L34975) at 4 °C for 30 min. Finally, the cells were washed twice with R5 and then analyzed using a FACS Canto II (BD Biosciences).
Isolation of KIR2DL2+2DS2+2DL3−NK cells
KIR2DL2+2DS2+2DL3−, KIR2DL2−2DS2−2DL3+, and KIR2DL2−2DS2−2DL3−NK cells were purified by cell sorting after staining PBMCs from KIR2DL2+2DS2+2DL3+ individuals and KIR2DL2−2DS2−2DL3+ individuals, with Pacific blue-labeled anti-CD3 mAb (BD Pharmingen, #558117, clone UCHT1, 1:50), PE-labeled anti-CD56 mAb (Biolegend, #318306, clone HCD56, 1:100), APC-labeled anti-KIR2DL3/CD158b2 mAb (R&D Systems, #FAB2014A-100, clone 180701, 1:20), and FITC-labeled CD158b mAb cross-recognizing 2DL2, 2DS2, and 2DL3 (BD Pharmingen, #559784, clone CH-L, 1:25).
Binding of HLA-C*12:02-tetramers to KIR2DL2+2DS2+2DL3−NK cells
Purified KIR2DL2+2DS2+2DL3− and KIR2DL2−2DS2−2DL3+NK cells were stained with PE-conjugated HLA-C*12:02-PolIY10 tetramer, HLA-C*12:02-PolIY10-9A-tetramer, HLA-C*12:02-PolIY11-tetramer, or HLA-C*12:02-PolIY11-9A-tetramer (100 nM) at 37 °C for 30 min. The cells were then washed twice with R5, followed by staining with FITC-conjugated anti-CD56 mAb (BD Pharmingen, #340410, clone NCAM16.2, 1:50), Pacific blue-labeled anti-CD3 mAb (BD Pharmingen), and 7AAD (BD Pharmingen, #559925, 1:25) at 4 °C for 30 min. Finally, the cells were washed twice with R5 and then analyzed using a FACS Canto II (BD Biosciences).
Sequencing of HIV-1 Pol IY11
Bulk sequencing of the HIV-1 PolIY11 encoding region of plasma viral RNA from 27 HLA-C*12:02+ individuals infected with HIV-1 subtype B viruses was performed26. Plasma was separated from whole blood and stored at −80 °C. HIV-1 RNA was extracted from plasma samples using a QIAamp UltraSens virus kit (Qiagen, Cat #53704) and then used for two rounds of PCR. PCR and Sequences were analyzed as previously described26.
NK cell response analysis
KIR2DL2+2DS2+2DL3−, KIR2DL2−2DS2−2DL3+, and KIR2DL2−2DS2−2DL3−NK cells were purified by cell sorting after staining PBMCs from a KIR2DL2+2DS2+2DL3+ individuals and a KIR2DL2−2DS2−2DL3+ individuals, with Pacific blue-labeled anti-CD3 mAb (BD Pharmingen, #558117, clone UCHT1, 1:50), PE-labeled anti-CD56 mAb (Biolegend, #318306, clone HCD56, 1:100), APC-labeled anti-KIR2DL3/CD158b2 mAb (R&D Systems, #FAB2014A-100, clone 180701, 1:20), and FITC-labeled CD158b mAb cross-recognizing 2DL2, 2DS2, and 2DL3 (BD Pharmingen, #559784, clone CH-L, 1:25). Sorted cells were cultured in serum-free medium Cell Growth Medium (CellGenix) supplemented with 10% FBS, 22 ng/ml OKT-3 mAb, 20 ng/ml rIL-2 (ProSpec), and 5 × 104 irradiated (60 Gy) PBMC from a healthy donor for 3–4 weeks as previously shown57.
RMA-S cells expressing HLA alleles were used as stimulator cells in previous studies of NK cell function25,58,59,60,61,62. We used RMA-S-C*12:02 cells as stimulator cells. RMA-S-C*12:02 were pre-cultured at 26 °C for 18 h and then incubated with PolIY10 or PolIY10-9A peptide at 1 μM at same temperature for 1 h; cells were then co-cultured with purified KIR2DL2+2DS2+2DL3−, KIR2DL2−2DS2−2DL3+, or KIR2DL2−2DS2−2DL3−NK cells (E/T ratio = 1:4) for 1 h at 37 °C, after which Brefeldin A (10 μg/ml) and Alexa Fluor® 647 labeled anti-CD107a (LAMP-1) mAb (BioLegend, #328612, clone H4A3, 1:40) were added and the cells were cultured for another 5 hrs. NK cells were then stained with FITC-labeled anti-CD56 mAb (BD Pharmingen, #340410, clone NCAM16.2, 1:50) and 7AAD (BD Pharmingen, #559925, 1:25), and subsequently fixed with 4% paraformaldehyde and incubated in permeabilization buffer (0.1% saponin–5% FBS–PBS). Thereafter, the cells were stained with PE-labeled anti-IFN-γ mAb (BioLegend, #502509, clone 4S.B3, 1:50) and analyzed by FACSCanto II (BD Biosciences). 7AAD−CD56+ cells producing IFN-γ and/or expressing CD107a were all counted as the responding NK cell population.
TCR clonotype analysis
PBMCs from KI-890 were stained with both PE-conjugated HLA-C*12:02-PolIY11-tetramer and APC-conjugated HLA-C*12:02-PolIY11-9A-tetramer, followed by staining with FITC-labeled anti-CD3 mAb (DAKO, #F0818, clone UCHT1, 1:50), Pacific Blue-labeled anti-CD8 mAb (BD Pharmingen, # 558207, clone RPA-T8, 1:50), and 7AAD (BD Pharmingen, #559925, 1:25). Among CD3+CD8+7AAD− cells, HLA-C*12:02-PolIY11-tetramer+ cells were sorted into a 96-well plate by using a FACS Aria I. Unbiased analysis of TCR gene usage was performed as previously described40,63.
Cloning, expression, and purification of recombinant proteins
A fragment of the gene encoding the KIR2Ds extracellular domain was cloned into the pGMT7 expression vector64. KIR2Ds, TCR α and β chains, the HLA-C*12:02 heavy chain, and beta-2-microglobulin were expressed in Escherichia coli strain BL21(DE3)pLysS, and inclusion bodies were isolated. KIR2Ds and HLA-C*12:02 were refolded by a dilution method, and the refolding solution was concentrated with a VIVAFLOW50 system (Sartorius) and an Amicon Ultra (Millipore). The concentrated protein solution was loaded on a gel filtration column, Superdex 75 (GE Healthcare). For crystallization, further purification was performed by anion-exchange chromatography with Resource Q (GE Healthcare). TCR recombinant protein was prepared as a stable heterodimer with disulfide bonds by introducing cysteine residues at the Thr48 of α chain and Ser57 of the β chain. Cys75 of the β chain, a free cysteine, was substituted with alanine to avoid undesired disulfide bond formation. TCR heterodimer was refolded by a dilution method and was concentrated using a VIVAFLOW50 system with replacement of 20 mM Tris-HCl, pH 8.0, 90 mM NaCl, 5% Glycerol solution. The concentrated protein was diluted 2-fold with 20 mM Tris-HCl, pH 8.0, and was purified by anion-exchange chromatography with Resource Q (GE Healthcare) and gel filtration chromatography using a Superdex200 column (GE Healthcare).
For SPR analysis, HLA-C*12:02 with a BirA tag (ASLHHILDAQKMVWNHR) at the C-terminus for site-specific biotinylation was prepared in the same way as non-tagged HLA-C*12:02, described above.
Surface plasmon resonance (SPR) analysis
Interaction analyses were performed using a Biacore 3000 (GE Healthcare). The binding of HLA-C*12:02 to TCR or to KIR2Ds was performed using a sensor chip CAP (GE Healthcare) with immobilized biotinylated HLA-C*12:02-PolIY10, -PolIY11, and -PolIY11-9A proteins. As a negative control, biotinylated HLA-C*04:01 or BSA was immobilized on the sensor chip. The KIR2Ds or TCR solutions were buffer-exchanged to HBS-EP (10 mM HEPES, pH 7.4, 150 mM NaCl, 3.4 mM EDTA, and 0.005% surfactant P20) and flowed over the immobilized proteins. For the kinetic analysis of TCR-HLA-C*12:02 binding, five different concentrations of TCR protein solution (0.09–1.4 M) were injected at a flow rate of 20 μL/min at 25 °C. For equilibrium binding analysis of KIR2Ds against HLA-C*12:02 proteins, the KIR2D proteins were injected over immobilized HLA-C*12:02 proteins at a rate of 10 μL/min.
Data analysis was performed using BIAevaluation software (version 4.1.1, Cytiva) and OriginPro 2017 (OriginLab).
Crystallization and data collection
Crystals of heterotrimeric HLA-C*12:02 heavy chain, β2-microglobulin, and PolIY10 peptide complexes (HLA-C*12:02-PolIY10) (8.6 mg/mL) were obtained after 3 days in 0.2 M Potassium thiocyanate, 0.1 M Tris propane, pH 7.5, 20% (w/v) PEG3350 at 293 K. Crystals of HLA-C*12:02-PolIY11 (6.1 mg/mL) were obtained in 0.1 M CH3COONa, pH 5.0, 1.8 M (NH4)2SO4 at 293 K by the sitting drop vapor diffusion method. Crystals of HLA-C*12:02-PolIY11-9A (8.2 mg/mL) were obtained in 0.2 M Sodium bromide, 20% (w/v) PEG 3350, respectively, at 293 K by the sitting drop vapor diffusion method.
Purified HLA-C*12:02-PolIY11 mixed with the TCR at a 3:1 molar ratio was incubated at 4 °C for 4 h and purified as a complex by gel filtration chromatography (Superdex200). The complexes in 20 mM Tris-HCl, pH 8.0 were concentrated to 7.7 mg/mL. Crystals were obtained in 0.1 M HEPES, pH 7.0, 1.6 M ammonium sulfate, 0.2 M sodium thiocyanate at 293 K by the hanging drop vapor diffusion method.
HLA-C*12:02-PolIY10 was mixed with KIR2DL2 at a molar ratio of 2:1, and then desalted to 20 mM Tris-HCl, pH 8.0, and concentrated by Amicon Ultra (MWCO: 10000 Da, Millipore). The crystals obtained from the initial screening (PEGs suite (QIAGEN)) were used to collect X-ray diffraction data.
The crystals obtained were analyzed using the synchrotron radiation beamline BL41XU (HLA-C*12:02-IY10) at Spring-8 in Harima Science Garden City, Hyogo Prefecture, and the synchrotron radiation beamline AR-NW12A (HLA-C*12:02-PolIY11), BL-1A (HLA-C*12:02-PolIY11-9A and HLA-C*12:02-PolIY10-KIR2DL2 complex), and BL-17A (HLA-C*12:02-PolIY11-TCR complex) at Photon Factory in the Tsukuba Campus of the High Energy Accelerator Research Organization (KEK).
Structure determination and refinement
The structure of HLA-C*12:02-PolIY10, tentatively solved in our laboratory at 2.5 Å, was used for analysis by the molecular replacement method. Coot ver. 0.9.8.165, Refmac566, and phoenix.refine67 were used for refinement. Detailed statistics are summarized in Supplementary Table 2.
HIV-1 infection of CD4.721.221 cells
CD4.221-C*12:02 cells were infected with VSV-G pseudotyped NL4-3 (n = 1) or NL4-3 PolV472A (n = 2) as previously described47. Briefly, 9 × 107 cells were infected with virus stocks (reverse transcriptase (RT) value = 200 ng/mL) in a low volume of RPMI 1640 medium (Thermo Fisher) containing 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10 mM HEPES (R10). The cells were then incubated for 1.5 h at 37 °C, after which 15 ml of R10 was added and the cells were cultured overnight. At day 2 postinfection, a further 15 ml R10 was added. On day 3 postinfection, the proportion of cells infected with HIV-1 was determined by intracellular p24 staining using FITC-labeled anti-HIV-1 p24 mAb (Beckman Coulter, #6604665, clone KC57, 1:100), as previously described47. Cells were then harvested for immunoprecipitation of peptide-HLA complexes.
HLA-I immunoprecipitation and peptide purification
A total of 1.5 × 108 CD4.221-C*12:02 cells were harvested from infected cultures, washed in PBS, and lysed in 5 ml of lysis buffer (1% IGEPAL 630, 300 mM NaCl, and 100 mM Tris [pH 8.0] plus protease inhibitors) at 4 °C for 45 min. Two centrifugation steps (2000 × g for 10 min followed by 20,000 × g for 30 min at 4 °C) were then employed to clear the lysates of infected cells prior to overnight capture of HLA-peptide complexes on W6/32-coated protein A-Sepharose beads. W6/32 (culture supernatant of hybridoma cell line HB-95)-bound HLA-peptide complexes were sequentially washed with 20 ml of wash buffer 1 (0.005% IGEPAL, 50 mM Tris [pH 8.0], 150 mM NaCl, 5 mM EDTA), wash buffer 2 (50 mM Tris [pH 8.0], 150 mM NaCl), wash buffer 3 (50 mM Tris [pH 8.0], 400 mM NaCl), and, finally, wash buffer 4 (50 mM Tris [pH 8.0]) under gravity flow in Econo-Column glass chromatography columns (Bio-Rad). Peptide-HLA complexes were eluted from the beads with 5 ml of 10% acetic acid, and following drying under a vacuum, they were loaded onto a 4.6- by 50-mm ProSwift RP-1S column (Thermo Fisher Scientific) and eluted from an Ultimate 3000 HPLC system (Thermo Scientific) using a 1 ml/min flow rate over 5 min from 2 to 34% buffer B (0.1% trifluoroacetic acid [TFA] in acetonitrile) in buffer A (0.1% TFA and 1% acetonitrile in water). Alternate 0.5 ml fractions (odd and even) were divided into two separate pools and dried under vacuum prior to resuspension in loading buffer for LC-MS/MS analysis.
Liquid chromatography-tandem mass spectrometry
HPLC fractions were dissolved in loading solvent and analyzed by an Ultimate 3000 HPLC system coupled to a high-field Q-Exactive (HFX) Orbitrap mass spectrometer (Thermo Scientific). Peptides were initially trapped in loading solvent, before RP separation with a 60 min linear acetonitrile in water gradient of 2–25% across a 75 mm 350 cm PepMap RSLC C18 EasySpray column (Thermo Scientific) at a flow rate of 250 nl/min. Gradient solvents contained an additional 1%(v/v) DMSO and 0.1%(v/v) formic acid. An EasySpray source was used to ionize peptides at 2000 V, and peptide ions were introduced to the MS at a transfer tube temperature of 250 °C. Ions were analyzed by data-dependent acquisition. Initially, a full-MS1 scan (120,000 resolution, 60 ms accumulation time, AGC 3 × 106) was followed by 20 data-dependent MS2 scans (60,000 resolution, 120 ms accumulation time, AGC 5 × 105), with an isolation width of 1.6 m/z and normalized HCD energy of 25%. Charge states of 1–4 were selected for fragmentation. Dynamic exclusion was set for 30 s. Each sample was acquired once.
Analysis of LC-MS/MS datasets
The analysis of all LC-MS/MS data sets (.raw files) was performed using PEAKS v10 software (Bioinformatic Solutions). No enzyme was specified during the peptide spectral matching, and mass tolerance settings of 5 ppm (for precursor ions) and 0.03 Da (for fragment ions) were used. Spectral sequence annotation was performed against the annotated Homo sapiens Swiss-Prot database appended with a 6-frame translation of the HIV-1 NL4-3 or NL4-3 PolV472A genome, respectively. A false discovery rate of 1% was set using a parallel decoy database search. Further filtering was performed to exclude known HLA-II contaminants68.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
