CLEAVAGE AND EXCHANGE OF MAJOR HISTOCOMPATIBILITY COMPLEX LIGANDS EMPLOYING AZOBENZENE-CONTAINING PEPTIDES

Abstract

In one aspect, the disclosure relates to major histocompatibility complex (MHC) molecules comprising a ligand in the peptide binding groove of the MHC molecule, whereby the ligand comprises an azobenzene (Abc), and at least two amino acid residues separated by the azo-group of the Abc, and wherein the amino acid residues are positioned to interact with the peptide binding groove of the MHC molecule. The disclosure also relates, among others, to means and methods for producing and using such MHC molecules, and the ligands therefor.

Claims

1. A major histocompatibility complex (MHC) molecule comprising a ligand in the peptide binding groove of the MHC molecule, whereby said ligand comprises an azobenzene (Abc) wherein at least one of the aromatic rings comprises an electron-donor group and wherein the azobenzene comprises at least two amino acid residues separated by the azo-group of the Abc, and wherein the amino acid residues are positioned to interact with the peptide binding groove of the MHC molecule.

2. The MHC molecule according to claim 1, wherein said ligand is an MHC peptide antigen of which amino acid residues that are located between the amino acid residues have been replaced by an Abc.

3. The MHC molecule according to claim 1, wherein the Abc comprises the general formula I ##STR00011## wherein at least one of the aromatic rings comprise an electron-donating group; M is independently C, S, N or O; Z1 and Z4 each comprise an amino acid residue positioned to interact with the peptide binding groove of the MHC molecule; Z2 and Z3 are independently H, hydroxyl, carboxy, keto, or a linear or branched C.sub.1-C.sub.10 alkyl, optionally substituted with an oxy, hydroxyl, nitrogen, nitroxy, sulhydryl or sulfide group.

4. The MHC molecule according to claim 1, wherein the Abc comprises the general formula II ##STR00012##

5. The MHC molecule according to claim 1, wherein the ligand comprises the general formula III ##STR00013## wherein, A, B, C, D, X and Y are each independently an amino acid residue; n.sub.1, n.sub.2, n.sub.3 and n.sub.4 are each independently 0-11; and n.sub.1+n.sub.2+n.sub.3+n.sub.4 equals 2-18.

6. The MHC molecule according to claim 1, wherein said Abc is a trans-Abc.

7. The MHC molecule according to claim 5, wherein n.sub.2 or n.sub.3 or both are 1.

8. A complex comprising at least two MHC molecules according to claim 1.

9. A composition comprising an WIC molecule according to claim 1, and an MHC peptide antigen.

10. A method of producing an MHC molecule comprising: producing an MHC molecule according to claim 1; contacting the produced MHC molecule with a reducing agent; and contacting said MHC molecule with an MHC peptide antigen.

11. A method of detecting an MHC molecule comprising producing an MHC molecule according to the method of claim 10, and detecting the WIC molecule.

12. A method according to claim 11, wherein the MHC molecule, a peptide in the peptide binding groove of the MHC molecule, or both, comprise a label.

13. A solid surface comprising an MHC molecule according to claim 1.

14. An azobenzene of formula I ##STR00014## wherein at least one of the aromatic rings comprise an electron-donating group; M is independently C, S, N or O; Z1 and Z4 each comprise an amino acid residue positioned to interact with the peptide binding groove of an MHC molecule; Z2 and Z3 are independently H, hydroxyl, carboxy, keto, or a linear or branched C.sub.1-C.sub.10 alkyl, optionally substituted with an oxy, hydroxyl, nitrogen, nitroxy, sulhydryl or sulfide group.

15. An azobenzene according to claim 14, for use in the production of an MHC molecule comprising a peptide in the peptide binding groove of the MHC molecule.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0094] FIG. 1. Design of HLA-A*11:01-restricted Abc ligands. a) Replacement of four amino acids residues (11 bond lengths) with an azobenzene-containing (Abc) tetrapeptide isostere (12 bond lengths). b) MHC stability ELISA of UV-sensitive A*11:01 molecules peptide-exchanged with A*11:01-restricted epitope (1), A*02:01-restricted epitope (2) and A*11:01-restricted Abc ligands (3 to 6) upon UV irradiation. MHC molecules before (UV) and after (+UV) UV irradiation in the absence of rescue peptides were included as controls. c) Sequences of the epitopes (1 and 2) and newly synthesized Abc ligands (3 to 6) that were used in (b). The position of the Abc moiety in the parent peptides is indicated in red. Anchor residues of the peptides are underlined. d) Overlay of crystal structures of 4 (cyan) (PDB reference ID: 4BEO, this work) and its parent peptide (yellow) (PDB reference ID: 2HN7) in an A*11:01 molecule (grey).

[0095] FIG. 2. Cleavage kinetics and conditions of HLA-A*11:01-restricted Abc ligand. a) Cleavage of 4 resulted in two aniline products (7 and 8) upon addition of sodium dithionite. Reaction was confirmed using LC-MS. b) 4 is incubated in the presence of 9 at 1:1 ratio with 1, 2.5 or 5 mM of sodium dithionite and the reactions were quenched after 1, 2 or 5 minutes. The reaction mixtures were analyzed for the presence of intact ABC ligands using LC/MS. c) Refolded A*11:01 molecules bearing 4 were peptide-exchanged were with A*11:01-restricted peptides (1 and 10) and A*02:01-restricted peptide (2) in the presence of 5 to 20 mM sodium dithionite. Controls with no peptides added () to the MHC molecules were included. Stable MHC molecules were quantified using MHC stability ELISA. d) Freshly isolated PBMCs were incubated with 10 mM (middle column) and 100 mM (right column) sodium dithionite for 1 hour (top row) or 16 hours (bottom row) to assess cellular toxicity of sodium dithionite. Cells were stained with anti-CD8 antibodies, Annexin V and LIVE/DEAD viability dye, and analyzed on flow cytometry. Plots shown were gated on CD8.sup.+ cells. Numbers in each plot are cells expressed as a percentage of total CD8.sup.+ population.

[0096] FIG. 3. Detection of antigen-specific CD8.sup.+ T cells using A*11:01 MHC tetramers generated from UV-mediated peptide exchange or sodium dithionite-mediated peptide exchange. a) Schematic diagram of experimental workflow. Freshly isolated PBMCs from an A*11:01-positive volunteer were stimulated with A*11:01-restricted epitopes EBV BRLF1.sub.134-142 (1, upper row) and Influenza A MP.sub.13-21 (10, lower row) and clonally expanded for 14 days. Antigen-specific CD8.sup.+ T cells were then labeled with cognate peptide-bound A*11:01 MHC tetramers (red) and detected via flow cytometry. b) A*11:01-EBV BRLF1.sub.134-142 (columns 1 and 4), A*11:01-Influenza A MP.sub.13-21 (columns 2 and 5) and A*11:01-Abc (column 3) tetramers were incubated with the PBMCs to detect EBV BRLF1.sub.134-142 (1, upper row) and Influenza A MP.sub.13-21 (10, lower row) -specific CD8.sup.+ T cells 14 days post-stimulation. Number in each plot represents tetramer-positive cells as a percentage of total CD8.sup.+ cell population. Plots in columns 1 to 2 and columns 3 to 5 are tetramer staining performed using UV-derived and Abc-derived tetramers respectively.

[0097] FIG. 4. Effects of reductive and oxidative cleavage conditions on epitopes of interest. Mass spectrometry analysis of peptides Influenza A MP.sub.13-21 (10, left column) and EBV BMLF-1.sub.259-267 (11, right column) after incubation with PBS, 10 mM sodium dithionite or 0.3 mM sodium periodate. 10 and 11 remained unmodified when incubated in PBS or 10 mM sodium dithionite. The N-terminal serine of 10 was cleaved when incubated in 0.3 mM sodium periodate. Incubation with 0.3 mM sodium periodate resulted in oxidation of cysteine and methionine residues of 11. Unmodified and modified residues are shown in blue and red respectively.

[0098] FIG. 5. (figure S1)

[0099] Mass spectrometry (MS) analysis of the synthesized Abc ligands by LC-MS IT-TOF. The peptides were first separated by liquid chromatography (LC) on a C18 column prior to the measurement of its mass to charge ratios (m/z). Details of the Abc ligands are shown in Table S1. The reference to Z-indicates Z the presence of an Abc of formula II

A) Conditional ligand: AIF-Z-TK; Empirical formula: C43H59N9O9;

Exact Mass: 845.44; Molecular Weight: 845.98

[0100] Peak #1; Retention Time: 11.520 min; Base Peak m/z: 846.3542;

Base Peak Intensity: 937599; Polarity: Pos

[0101] B) Conditional ligand: AIM-Z-YPK; Empirical formula: C49H68N10O10S;

Exact Mass: 988.48; Molecular Weight: 989.19

[0102] Peak #2; Retention Time: 9.033 min; Base Peak m/z: 495.2475;

Base Peak Intensity: 1413471; Polarity: Pos

[0103] C) Conditional ligand: QVPL-Z-YK; Empirical formula: C51H71N11O11;

Exact Mass: 1013.53; Molecular Weight: 1014.19

[0104] +ESI Scan; 15 scans: 6.088-6323 min Frag=180.0V; Polarity: Pos
D) Conditional ligand: KTF-Z-PK; Empirical formula: C45H62N10O9;
Exact Mass: 886.47; Mol. Wt.: 887.04
Peak #: 1; Retention Time: 10.733 min; Base Peak m/z: 887.3648;

Base Peak Intensity: 658640; Polarity: Pos

[0105] E) Conditional ligand: FLPS-Z-SV; Empirical formula: C46H61N9O11;

Exact Mass: 915.45; Molecular Weight: 916.05

[0106] Peak #: 2; Retention Time: 12.513 min; Base Peak m/z: 916.4301;

Base Peak Intensity: 6800116; Polarity: Pos

[0107] F) Conditional ligand: LLF-Z-YV; Empirical formula: C50H64N8O9;

Exact Mass: 920.48; Molecular Weight: 921.11

[0108] Peak #: 2; Retention Time: 14.293 min; Base Peak m/z: 921.4783;

Base Peak Intensity: 2512246; Polarity: Pos, Event

[0109] G) Conditional ligand: NLVP-Z-TV; Empirical formula: C44H64N10O11;

Exact Mass: 908.48; Molecular Weight: 909.05

[0110] Peak #: 2; Retention Time: 12.793 min; Base Peak m/z: 909.4549;

Base Peak Intensity: 1161661; Polarity: Pos

[0111] H) Conditional ligand: NLVP-Z-VATV; Empirical formula: C52H78N12O13;

Exact Mass: 1078.58; Molecular Weight: 1079.27

[0112] Peak #: 2; Retention Time: 13.300 min; Base Peak m/z: 1079.5823;

Base Peak Intensity: 1070704; Polarity: Pos

[0113] I) Conditional ligand: GLS-Z-RL; Empirical formula: C38H57N11O9;

Exact Mass: 811.4341; Molecular Weight: 811.9275

[0114] Peak #: 1; Retention Time: 9.613 min; Base Peak m/z: 406.7002;

Base Peak Intensity: 3287573; Polarity: Pos

[0115] J) Conditional ligand: FAP-Z-AL; Empirical formula: C41H52N8O8;

Exact Mass: 784.39; Molecular Weight: 784.91

[0116] Peak #: 1; Retention Time: 9.693 min; Base Peak m/z: 785.3885;

Base Peak Intensity: 582117; Polarity: Pos

[0117] K) Conditional ligand: FAP-Z-KL; Empirical formula: C44H59N9O8;

Exact Mass: 841.45; Molecular Weight: 842.01

[0118] Peak #: 2; Retention Time: 9.253 min; Base Peak m/z: 421.7008;

Base Peak Intensity: 2788083; Polarity: Pos

[0119] FIG. 6. (figure S2)

[0120] Binding of Abc ligands to HLA-A*02:01 and H2-K.sup.b. Photocleavable ligands on a) A*02:01 and b) K.sup.b molecules were peptide-exchanged with either previously identified peptide antigens (1, 9, 12 and 18) or Abc ligands (13 to 17, 19 and 20) following UV irradiation. Peptide ligands that can bind to the MHC will stabilize the MHC complex. ELISA was used to detect intact MHC molecules before UV irradiation (UV), after UV irradiation in the absence (+UV) or in presence of binding peptides.

[0121] FIG. 7. (figure S3)

[0122] In vitro refolding and biotinylation of HLA-A*11:01, A*02:01 and H2-K.sup.b containing Abc ligands (4, 17 and 20 respectively). a) The refolded MHC complexes were purified using S200 size exclusion chromatography. Fractions corresponding to approximately 45 kDa were collected as indicated in red. b) Gel shift SDS-PAGE was performed to assess the proportion of biotinylated MHC molecules. Purified MHC molecules yielded two distinct bands corresponding to the heavy chain (33 kDa) and beta2m (12 kDa). In the presence of soluble streptavidin, the biotinylated heavy chains bind to streptavidin forming complexes of high molecular size. IB:HC and IB: beta 2m refers to the heavy chain and beta2m extracted from E. coli inclusion bodies respectively.

[0123] FIG. 8. (figure S4)

[0124] Crystal structure of HLA-A*11:01 (grey) in complex with the Abc ligand, AIM-Z-YPK (cyan). a) Side view of the complex in cartoon format showing that the azobenzene moiety protrudes from the peptide binding cleft of the MHC. b) Top-down view in cartoon format showing the orientation that the Abc peptide resides in the binding cleft. c) Representation of the complex in its side surface view to show the depth of the ligand binding in the cleft. d) Top-down surface view shows that the ligand fits into the pockets of the MHC binding cleft. Z is the Abc of formula II.

[0125] FIG. 9. (figure S5)

[0126] Interactions between AIM-Z-YPK and residues in the binding groove of HLA-A*11:01. a) Top-down zoomed-in view of the HLA-A*11:01 peptide binding cleft. HLA-A*11:01 residues that contact the Abc ligand are highlighted as grey sticks. Electron density omit map (dark grey mesh) of the AIM-Z-YPK ligand (cyan). b) Interaction map depicting the contacts (represented by black dotted lines) made between the Abc ligand (amino acid in blue azobenzene moiety in red) and HLA-A*11:01 residues (black). Numbers representing bond distances are in . Z is the Abc of formula II.

[0127] FIG. 10. (figure S6)

[0128] Alternate conformations of the AIM-Z-YPK ligand. The normal cis binding confirmation of the Abc ligand (left) and the trans cross-linking conformation (right), of the two complexes in the asymmetric unit are shown. Z is the Abc of formula II.

[0129] FIG. 11. (figure S7)

[0130] HLA-A*11:01 molecules with AIM-Z-YPK ligand bound in alternate conformations give rise to two different molecular species. The canonical cis binding conformation of the ligand results in HLA-A*11:01 monomeric complexes and the noncanonical trans binding conformation results in HLA-A*11:01 dimeric complexes. a) The two species yielded two fractions in size exclusion chromatography. b) Repeated size exclusion chromatography with the separated fractions (fractions 1, left and 2, right) shows that the species did not interconvert in solution. c) Particle size of the HLA-A*11:01:AIM-Z-YPK species were determined using dynamic light scattering to be 171 and 98 in diameter for fraction 1 (left) and 2 (right) respectively. Z is the Abc of formula II.

[0131] FIG. 12. (figure S8)

[0132] Crystal structure of HLA-A*02:01 (grey) in complex with the Abc peptide, GLS-Z-RL (orange). a) The side view of the structure reveals that the Abc moiety of GLS-Z-RL ligand sits lower in the MHC peptide-binding cleft and is not as exposed as in the HLA-A*11:01 complex. b) 90 degree flip around the x-axis to show the top-down view in cartoon format of the peptide binding in the cleft. c) View of (a) in surface format. d) View of (b) in surface format. Z is the Abc of formula II.

[0133] FIG. 13. (figure S9)

[0134] Interactions between GLS-Z-RL and residues in the binding groove of HLA-A*02:01. a) Top-down zoomed-in view of the HLA-A*02:01 peptide binding cleft. Abc ligand-interacting residues of HLA-A*02:01 are highlighted as grey sticks. Electron density omit map (dark grey mesh) of the AIM-Z-YPK ligand (orange). The aromatic rings of the azobenzene moiety are not in the same plane due to a slight twist around the NN bond. b) Interactions (represented by black dotted lines) made between the Abc ligand (amino acid in orange, azobenzene moiety in red) and HLA-A*02:01 residues (black) are shown in an interaction map. Z is the Abc of formula II.

[0135] FIG. 14. (figure S10)

[0136] MHC stability ELISA of sodium dithionite-mediated peptide-exchanged HLA-A*02:01 and H2-K.sup.b molecules. a) A*02:01 refolded in vitro with 17 were peptide-exchanged with two A*02:01-restricted epitopes (12 and 21) and an A*11:01-restricted epitope (1) in the presence of 5 to 20 mM sodium dithionite. b) Similar to (a), refolded K.sup.b molecules bearing 20 were peptide-exchange with two K.sup.b-restricted epitopes (18 and 22) and an L.sup.d-restricted epitope (9). Negative controls with no peptides added () to the MHC molecules were included.

[0137] FIG. 15. (figure S11)

[0138] Preferential cleavage of Abc ligands by sodium dithionite. a) A mixture containing 9 and 17 at 1:1 molar ratio (0.123 mM each) and varying concentrations of L-Glutathione oxidized (0 to 125 mM) were incubated in the absence (empty bars) or presence of 2.5 mM sodium dithionite (filled bars) for 5 minutes. b) Similar to (a), peptide mixture of 9 and 17 (0.123 mM each) were added to varying concentrations of L-Cystine (0 to 125 mM) prior to treatment with 2.5 mM sodium dithionite. Data are represented as the ratio of the intact Abc ligand 17 to dithionite-resistant 9 detected in LC-MS after sodium dithionite treatment.

[0139] FIG. 16. (figure S12)

[0140] Flow cytometric analysis on the viability of sodium dithionite-treated CD8.sup.+ T cells. Freshly isolated human PBMCs were incubated with sodium dithionite ranging from 1 mM to 100 mM to assess cellular toxicity of sodium dithionite. Cells were treated for 1 hour (top row), 1 hour followed by rested overnight in fresh culture media (middle row) or 16 hours (bottom row) prior to staining with anti-CD8 antibodies, Annexin V and LIVE/DEAD viability dye. Data represented above are based on CD8.sup.+ cells. Numbers in each plot are expressed as a percentage of total CD8.sup.+ population.

[0141] FIG. 17. (figure S13)

[0142] Detection of antigen-specific CD8.sup.+ T cells using A*02:01 tetramers generated from UV-mediated peptide exchange or sodium dithionite-mediated peptide exchange. Freshly isolated peripheral blood mononuclear cells from an A*02:01-positive volunteer were stimulated with A*02:01 epitopes EBV LMP2426-434 (upper row) and CMV pp65495-503 (lower row). A*02:01-EBV LMP2426-434 (columns 1 and 4), A*02:01-CMV pp65495-503 (columns 2 and 5) and A*02:01-Abc (column 3) tetramers were used to perform tetramer staining 14 days post-in vitro stimulation. Number in each plot represents tetramer-positive cells expressed as a percentage of total CD8.sup.+ T cells. Plots in columns 1 to 2 and columns 3 to 5 are tetramer staining performed using UV-derived tetramers and Abc-derived tetramers respectively.

[0143] FIG. 18. (figure S14)

[0144] Detection of antigen-specific CD8.sup.+ T cells using H2-K.sup.b MHC tetramers generated from UV-mediated peptide exchange or sodium dithionite-mediated peptide exchange. Freshly isolated splenocytes from naive and OTI-TCR transgenic C57/BL6 mice were mixed in 1:1 ratio. K.sup.b-OVA.sub.257-264 (columns 1 and 4), K.sup.b-Tgd057.sub.57-64 (columns 2 and 5) and K.sup.b-Abc (column 3) tetramers were used to detect OT1 cells from the splenocyte mixtures. Numbers in each plot represent tetramer-negative (left) and tetramer-positive (right) CD8.sup.+ splenocytes expressed as a percentage of total splenocyte mix. Plots in columns 1 to 2 and columns 3 to 5 are tetramer staining performed using UV-derived tetramers and Abc-derived tetramers respectively.

[0145] FIG. 19.

[0146] Peptide binding of MBP.sub.85-99 (yellow) to HLA-DR2 (blue) molecule. 15 residues (P-4 Glu to P11 Arg) of the MBP peptide were shown with peptide side chains of the P1 Val, P4 Phe, P6 Asn and P9 Thr occupying pockets within the peptide binding groove of the MHC molecule. (Smith et al. J. Exp. Med. 1998, 188, 1511-1520.) This figure is not part of the manuscript.

[0147] FIG. 20.

[0148] Crystal structure (left) versus model (right) of the HLA-DR2-MBP.sub.85-99 complex. MBP.sub.85-99 (yellow) binds to HLA-DR2 (blue) with the Anp residue at P4 position occupying the large hydrophobic P4 pocket of the MHC molecule. (Grotenbreg et al. J. Biol. Chem. 2007, 282, 21425-21436.) This figure is not part of the manuscript.

[0149] FIG. 21.

[0150] Abc-ligands for HLA-A*02:01.

[0151] FIG. 22.

[0152] Peptide exchange of known non-(CTELKLSDY and IVTDFSVIK); intermediate-(SLENFRAYV; ALQLLLEV and VMLRWGVLA) and high affinity binders (NLVPMVATV; GILGFVFTL, SLYNTVATL and NMLSTVLGV) to HLA-A*02:01 using HLA-A*02:01-ILKZGV (A) or HLA-A*02:01-ILKZKV (B) and Na2S2O4-induced peptide exchange (values are the meanSD of two independent experiments) or HLA-A*02:01-KILGFVFJV and UV-induced peptide exchange (C). The presence of intact HLA complex was determined by MHC stability ELISA. The measured absorbances at 414 nm were evaluated relative to that of the high affinity binder NLV which was put to 100%

[0153] FIG. 23

[0154] Human peripheral blood cells (PBMC) were stained for the presence of antigen-specific T cell responses using PE-labeled tetramers. The flow cytometric results are depicted in the figure. The Abc and UV tetramers render similar results.

Staining of antigen-specific T cell responses against 4 different CD8 epitopes restricted to HLA-A*02:01 in a PBMC sample.
Abc: PE-labeled tetramers generated using Abc ligand peptide exchange technology; HLA-A*02:01-ILKZGV.
Abc*: PE-labeled tetramers generated using Abc ligand peptide exchange technology; HLA-A*02:01-ILKZKV.
UV: PE-labeled tetramers generated using UV-induced peptide exchange technology; HLA-A*02:01-KILGFVFJV.

SI: Stain Index.

EXAMPLES

Materials and Methods

Abc Ligand and Antigenic Peptide Synthesis

[0155] The azobenzene-containing (Abc) MHC ligands were manually constructed by standard Fmoc-based solid-phase peptide synthesis. Fmoc-protected amino acids and Wang-based resins were purchased from Advanced ChemTech. The azobenzene linker was constructed as described (Verhelst et al., 2007). All other chemicals were purchased from Sigma-Aldrich. Deprotection and coupling of amino acids was carried out manually in a rotating glass reactor vessel at 0.2 mmol scale. For each peptide, the MBHA Resin HS, 100-200 mesh, 1% DVB (105 mg, 0.2 mmol, 1 equiv) was allowed to swell for 12 min in N-methyl-2-pyrolidinone (NMP). Installation of HMPB linker (120 mg, 0.5 mmol, 2.5 equiv) was accomplished using hydroxybenzotriazole (HOBT) (68 mg, 0.5 mmol, 2.5 equiv), benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate (PyBOP) (260 mg, 0.5 mmol, 2.5 equiv) and N,N-diisopropylethylamine (DIPEA) (246 ul, 1.5 mmol, 7.5 equiv) in 4 ml NMP. The HMPB-linked resin was washed for 12 min in NMP, followed by 12 min in dichloromethane (DCM). The first amino acid (0.8 mmol, 4 equiv) was coupled using N,N-diisopropylcarbodiimide (DIC) (124 ul, 0.8 mmol, 4 equiv), 4-dimethylaminopyridine (DMAP) (4 mg, 0.033 mmol, 0.165 equiv) in 4 ml DCM. The resin was then washed in DCM for 12 min, followed by 12 min in NMP. The amino acid/azobenzene linker was Fmoc-deprotected for 15 min using a solution of 20% piperidine in NMP. Following amino acid couplings were carried out using HOBT (108 mg, 0.8 mmol, 4 equiv), PyBOP (416 mg, 0.8 mmol, 4 equiv) and DIPEA (392 ul, 2.4 mmol, 12 equiv) in 4 ml NMP. Azobenzene linker (204 mg, 0.4 mmol, 2 equiv) coupling was carried out twice using PyBOP (208 mg, 0.4 mmol, 2 equiv) and DIPEA (196 ul, 1.2 mmol, 6 equiv) in 2 ml NMP. A Kaiser test (Kaiser et al., 1970) was used to monitor reaction completeness. Stepwise deprotection and coupling of the appropriate amino acids or azobenzene linker furnished the desired peptide on-resin. The peptides were cleaved, and simultaneously deprotected from dried resin using 5 ml trifluororoacetic acid (TFA) solution containing 2.5% distilled water and 2.5% triisopropyl silane (TIS) over 24 hrs. The peptide solution was precipitated in cold diethyl ether, and dried under vacuum. The peptide identities were confirmed by IT-TOF LC/MS analysis (Shimadzu).

Cleavage of Abc Ligands with Sodium Dithionite

[0156] Abc ligand and IPAAAGRFF were mixed at 1:1 molar ratio (0.123 mM each) and incubated in the presence of 1 mM, 2.5 mM or 5 mM freshly prepared sodium dithionite (in 200 mM phosphate buffer, pH 7.4). The reactions were allowed to proceed for 1 to 5 min until quenched using ZipTipC18 (Milipore) to extract the peptides from the sodium dithionite solution. The peptides were then eluted in 0.1% trifluoroacetic acid containing 5% acetonitrile and analyzed on IT-TOF LC/MS (Shimadzu).

In Vitro Folding and Purification of MHC Complexes

[0157] MHC molecules were generated as described previously (Garboczi et al., 1992). Genes encoding human 2-microglobulin and luminal portion of HLA-A*11:01, A*02:01 and H2-K.sup.b engineered with a C-terminal BirA recognition sequence were cloned into pET-28a (+) vector (GenScript). The plasmids were transformed and overexpressed in E. coli BL21 induced by 1 mM isopropyl -D-thiogalactopyranoside. The expressed proteins were extracted and purified from the inclusion bodies under reducing conditions and solubilized in 8 M urea. In vitro refolding of the MHC molecules was carried out with at least 10-fold molar excess of either UV-cleavable or Abc ligands for 24 to 36 h. The proteins were dialyzed into 20 mM Tris (pH8.0), biotinylated in vitro by recombinant BirA and purified using S200 size exclusion chromatography. Biotinylated MHC molecules were conjugated with Streptavidin-PE (Invitrogen) at 4:1 molar ratio to form MHC tetramers. For MHC molecules used in crystallography, refolding and purification were carried out in a similar fashion with the exception that unbiotinylated constructs were used. Also, the proteins purified from size exclusion chromatography were further subjected to ion exchange chromatography on a Mono Q column in 20 mM Tris (pH 8.0) and eluted over a gradient of increasing salt concentration with 20 mM Tris (pH 8.0), 1 M NaCl. For both HLA-A*11:01 and HLA-A*02:01, the proteins eluted at approximately 100-150 mM NaCl.

Peptide Exchange Conditions on MHC Monomers and Tetramers

[0158] MHC monomers used for MHC stability ELISA were peptide-exchanged in the presence of 100 fold molar excess of peptide ligands. For photocleavable MHC monomers, preparations of 500 nM MHC monomers in PBS were subjected to 365 nm longwave UV irradiation on ice for 15 minutes using UVP CL-1000 L Ultraviolet crosslinker (UVP), followed by the addition of 50 uM peptide ligands and 1-hour incubation on ice. For Abc MHC monomers, preparations containing 500 nM MHC monomers, 50 uM peptide ligands and 5 to 20 mM sodium dithionite in 50 mM HEPES (pH 7.4) were incubated for 30 minutes on ice. To stain antigen-specific CD8.sup.+ T cells, photocleavable MHC tetramers were diluted to 40 ug/ml with cold PBS containing 200 M peptides, subjected to 365 nm longwave UV irradiation on ice for 15 minutes and followed by 1-hour incubation on ice. 40 g/ml Abc MHC tetramers were incubated with 10 mM sodium dithionite in 50 mM HEPES (pH 7.4) containing 200 M peptides and followed by 30-minute incubation on ice. After incubation, all MHC monomers and tetramers were further incubated for 1 hour at 37 C. with shaking at 850 rpm and were centrifuged at 16 000g, 4 C. for 10 minutes prior to use.

MHC Stability ELISA

[0159] Assessment of ABC ligand binding to MHC molecules and optimization of ABC peptide exchange conditions were performed using an established protocol (Rodenko et al., 2006). Briefly, wells of a 384-well microplate (Corning) coated overnight at room temperature (RT) with 50 l of 2 g/ml streptavidin in PBS were washed and treated with 100 l of 2% BSA in PBS for 30 min at RT. The 2% BSA was discarded and 25 l of 20nM peptide-exchanged MHC was added to each wells and incubated on ice for 1 h. Wells were then washed and incubated with 25 l 1 g/ml HRP-conjugated anti-2m antibodies (Clone D2E9, Abcam) on ice for 1 h. Subsequently, wells were washed and developed with 25 l of ABTS solution (Invitrogen) for 10 to 15 min at RT. The development is quenched by the addition of 12.5 l of 0.01% sodium azide in 0.1M citric acid. Absorbance was measured at 415 nm using Spectramax M2 microplate reader (Molecular Devices). Each washing procedure involves rinsing the wells four times with 100 l of 0.05% Tween 20 in PBS. Samples were measured in quadruplicates.

Cells and MHC Tetramer Staining

[0160] Fresh whole blood was obtained from A*11:01 and A*02:01-positive volunteers. Isolation of PBMCs from these samples was performed via Ficoll-Paque density-gradient centrifugation. The isolated PBMCs were frozen for later staining without stimulation or were cultured in RPMI 1640 containing 2.05 mM L-glutamine (Invitrogen) supplemented with 40 M 2-mercaptoethanol (Gibco), 100 IU/ml penicillin/streptomycin (Invitrogen) and 5% pooled human AB serum (Invitrogen) at 37 C., 5% CO.sub.2. Briefly, PBMCs were stimulated with peptides at 10 g/ml. 25 U/ml interleukin-2 (IL-2) (R&D systems) was added to the culture 2 days post peptide stimulation. Half medium change was carried out and 25 U/ml IL-2 was supplemented every 2 to 3 days from 5 to 14 days post stimulation.

[0161] Mouse splenocytes were extracted from spleens of nave and OTI-TCR transgenic C57/BL6 mice using conventional splenocyte extraction protocol. Briefly, spleen meshed and homogenized in cold PBS was passed through a cell strainer. The resultant cells were washed with cold PBS and treated with 3 ml of RBC lysis buffer (pH 7.4) containing 155 mM NH.sub.4Cl, 10 mM KHCO.sub.3 and 0.1 mM EDTA for 2 min. Finally, the cells were washed twice with 10 ml of cold PBS and resuspended in 5 ml of cold PBS.

[0162] Cells were first stained with cell viability LIVE/DEAD fixable near-IR stain (Molecular Probes) prior to tetramer staining Subsequently, cells were washed with PBS and incubated with 80 nM peptide exchanged PE-conjugated MHC tetramers on ice for 20 min. Cells from A*11:01-positive donor were stained with 200 nM MHC tetramers instead. All PBMCs and murine splenocytes were stained with anti-human CD8 (Clone RPA-T8, BD biosciences) or anti-murine CD8 (Clone 53-6.7, BD biosciences) Pacific Blue antibodies for 15 min respectively. Cells were then washed again with PBS and fixed with 1% paraformaldehyde in PBS. Flow cytometry data were acquired on BD LSRII flow cytometer and analyzed using FlowJo (Tree Star).

Cell Viability Assay

[0163] 10.sup.6 freshly isolated PBMCs from healthy volunteers were incubated in 1 ml RPMI 1640 culture media containing HEPES-buffered 1 mM to 100 mM sodium dithionite (pH 7.4) at 37 C., 5% CO.sub.2. After 1-hour or 16-hour incubation, the cells were immediately assessed for cell viability or rested overnight in fresh media (for 1-hour treatment only). The cells were harvested, washed with PBS twice and stained with cell viability LIVE/DEAD fixable near-IR stain (Molecular Probes). The cells were then washed again with PBS and stained with anti-human CD8 (Clone RPA-T8; BioLegend) Brilliant Violet 421 antibodies for 15 minutes. Thereafter, the cells were washed once with PBS and once with 1 Annexin V binding buffer (10 mM HEPES, pH 7.4; 140 mM NaCl; 2.5 mM CaCl.sub.2) prior to incubation with Annexin V FITC (eBioscience) for 10 minutes. The stained cells were immediately analyzed on BD LSRII flow cytometer and data were processed using FlowJo (Tree Star).

Mass Spectrometry Analysis of Epitope Modification

[0164] 50 uM of Influenza A MP.sub.13-21 and EBV BMLF-1.sub.259-267 peptides were incubated with 10 mM Na.sub.2S.sub.2O.sub.4 in 50 mM HEPES (pH 7.4) or 0.3 mM NaIO.sub.4 in PBS at RT for 2 h. After which, the peptides were extracted from the buffer using ZipTipC18 (Milipore) and loaded on LC/MS IT-TOF (Shimadzu) for analysis. 50 uM peptides in PBS were used as a control.

Competition Assay for Cleavage of Abc Ligands and Disulfide Bonds

[0165] GLS-Z-RL and IPAAAGRFF (0.123 mM each) were mixed with 2.5 mM, 25 mM or 125 mM L-Glutathione oxidized (Sigma-Aldrich) or L-Cystine (Sigma-Aldrich) and incubated with 2.5 mM freshly prepared sodium dithionite (in 200 mM phosphate buffer, pH7.4). After 5 minutes, the peptides were extracted from the L-Glutathione oxidized or L-Cystine, and sodium dithionite mixture using ZipTipC18 (Milipore). Elution of the peptides was carried out in 0.1% trifluoroacetic acid containing 5% acetonitrile prior to analysis on IT-TOF LC/MS (Shimadzu).

X-Ray Structures of HLA-A*11:01:AIM-Z-YPK and HLA-A*02:01:GLS-Z-RL Complexes

[0166] We performed X-ray crystallographic studies to determine the molecular details in which class I MHC molecules bind to the azobenzene-containing peptide.

Crystallization Conditions for HLA-A*11:01:AIM-Z-YPK and HLA-A*02:01:GLS-Z-RL Complexes

[0167] Crystals for HLA-A*11:01:AIM-Z-YPK were grown at room temperature using the sitting drop, vapor-diffusion method with a well solution of 15% (w/v) PEG4000, 0.2 M ammonium sulfate, 0.1 M tri-sodium citrate (pH 5.6). Crystals for HLA-A*02:01:GLS-Z-RL were grown at room temperature using the sitting drop method with a well solution of 20% (w/v) PEG4000, 10% (w/v) isopropanol, 0.1 M HEPES pH 7.5. Crystals were harvested and frozen rapidly in liquid nitrogen for data collection.

X-Ray Data Collection and Structure Refinement of HLA-A*11:01:AIM-Z-YPK and HLA-A*02:01:GLS-Z-RL Complexes

[0168] X-ray diffracted intensities for HLA-A*11:01:AIM-Z-YPK were collected at 100 K using a FRE generator at the Biopolis Shared Facilities, Singapore with a R-AXIS IV++ imaging plate detector from Rigaku. The data was collected at X-ray wavelength of 1.54 . X-ray data for HLA-A*02:01:GLS-Z-RL were collected at 100 K using the X06DA beamline (X-ray wavelength of 1.0 ) at the Swiss Light Source with a Pilatus detector. Diffraction data (Table S3 for A*11:01 and Table S5 for A*02:01) for both HLA complexes were integrated with Mosflm and intensities were scaled with SCALA (Evans, 2006; Leslie, 1992). The structures were solved by molecular replacement in the program MOLREP (Vagin and Teplyakov, 2000), using the HLA-A*11:01 structure with PDB code 2HN7 (Blicher et al., 2006) or the HLA-A*02:01 structure with PDB code 3V5H, as search probe for HLA-A*11:01 and HLA-A*02:01 respectively. For HLA-A*11:01, refinement was carried out with REFMAC and BUSTER (Murshudov et al., 1997; Smart et al., 2012), with a final refinement was carried out on REFMAC. For HLA-A*02:01, the structure was refined initially with REFMAC, followed by final refinement rounds with Buster. Validation of the models and the x-ray data were checked with MOLPROBITY (Davis et al., 2007), and figures were generated using PyMOL (Delano, 2002). The coordinates and structure factors (code 4BEO for the HLA*A11:01 complex and 4BLH for the HLA*A02:01 complex) have been deposited in the Protein Data Bank.

The Crystal Structure of HLA-A*11:01:AIM-Z-YPK and HLA-A*02:01:GLS-Z-RL Complexes

[0169] We performed X-ray crystallographic studies to determine the molecular details of the interaction between class I MHC molecules and the azobenzene-containing peptide.

Overall Description

[0170] The X-ray structure of the HLA-A*11:01 molecule in complex with the azobenzene containing peptide was determined to 2.43 resolution (Table S3 and Fig S4). The model contains residues 1-274 of the heavy chain of HLA-A*11:01, residues 1-99 of 2-microglobulin and the azobenzene containing peptide, AIM-Z-YPK. There are two molecules in the asymmetric unit. The overall structure of the HLA-A*11:01/2m/peptide complex is similar to the native peptide complex [PDB code 2HN7], and the RMSD for all C atoms of the alpha chain of the MHC molecules is 0.607 . The structure of the HLA-A*02:01 complex, which consists of residues 1-275 of the heavy chain, residues 1-100 of 32-microglobulin, and the azobenzene-containing peptide, GLS-Z-RL, was determined to 2.1 resolution (Table S5 and Fig S7). There are two molecules in the asymmetric unit. Superimposition of the HLA-A*11:01/2m/peptide complex with the previously solved structure of HLA-A*02:01 structure [PDB code 3V5H] is similar overall, and the RMSD for all C atoms of the alpha chain of the MHC molecules is 1.11 .

REFERENCES CITED IN THE MATERIALS AND METHODS SECTION

[0171] Blicher, T., Kastrup, J. S., Pedersen, L. ., Buus, S., and Gajhede, M. (2006). Structure of HLA-A*1101 in complex with a hepatitis B peptide homologue. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 62, 1179-1184. [0172] Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., Snoeyink, J., Richardson, J. S., et al. (2007). MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 35, W375-W383. [0173] Delano, W. L. (2002). The PyMOL Molecular Graphics System, DeLano Scientific LLC. Palo Alto, Calif., USA. [0174] Evans, P. (2006). Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72-82. [0175] Garboczi, D. N., Hung, D. T., and Wiley, D. C. (1992). HLA-A2-peptide complexes: refolding and crystallization of molecules expressed in Escherichia coli and complexed with single antigenic peptides. Proc. Natl. Acad. Sci. U.S.A. 89, 3429-3433. [0176] Kaiser, E., Colescott, R. L., Bossinger, C. D., and Cook, P. I. (1970). Color test for detection of free terminal amino groups in the solid-phase synthesis of peptides. Anal. Biochem. 34, 595-598. [0177] Leslie, W. A. G. (1992). Recent changes to the MOSFLM package for processing film and image plate data. Joint CCP4+ESF-EAMCB News-Letter on Protein Crystallography. [0178] Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53, 240-255. [0179] Rodenko, B., Toebes, M., Hadrup, S. R., van Esch, W. J. E., Molenaar, A. M., Schumacher, T. N. M., and Ovaa, H. (2006). Generation of peptide-MHC class I complexes through UV-mediated ligand exchange. Nat Protoc 1, 1120-1132. [0180] Smart, O. S., Womack, T. O., Flensburg, C., Keller, P., Paciorek, W., Sharff, A., Vonrhein, C., and Bricogne, G. (2012). Exploiting structure similarity in refinement: automated NCS and target-structure restraints in BUSTER. Acta Crystallogr. D Biol. Crystallogr. 68, 368-380. [0181] Vagin, A., and Teplyakov, A. (2000). An approach to multi-copy search in molecular replacement. Acta Crystallogr. D Biol. Crystallogr. 56, 1622-1624. [0182] Verhelst, S. H. L., Fonovi, M., and Bogyo, M. (2007). A mild chemically cleavable linker system for functional proteomic applications. Angew. Chem. Int. Ed. Engl. 46, 1284-1286.

Results

[0183] We explored the application of azobenzene-containing (Abc, Z) linkers that are sensitive to sodium dithionite (Na.sub.2S.sub.2O.sub.4). The recently developed stereocenter-free building block is accessible from readily available starting materials by a straightforward and cost-effective synthesis route. Furthermore, the Abc moiety is unaffected by reducing agents common to biological protocols (e.g. TCEP, DTT) and the correct fragmentation conditions have been demonstrated to be compatible with biomolecules and living systems.[5]

[0184] The Abc-linker, with its 12 bond lengths separating the amino- and carboxylic acid functionalities, cannot formally be regarded as tetrapeptide isostere (11 bond lengths, FIG. 1a), but was envisaged to act as a surrogate for four amino acid residues, making allowances for the double bond and aromatic systems counting towards the peptidomimetic backbone. Conditional ligands were designed such that the Abc building block strategically replaced non-essential residues within a parent epitope of high affinity (Table S1 and S2), which improves the likelihood of the resulting Abc ligand to bind to and stabilize the recombinant MHC sufficiently during in vitro refolding and purification. For example, in the HLA-A*11:01-restricted epitope from hepatitis B virus DNA polymerase 110-118 (Table S1), the residues at positions P4-P7 are solvent-exposed, identifying them as candidates for Abc replacement.[6] Moreover, the key N- and C-terminal anchor residues Ile (P2) and Lys (P10) were conserved to ultimately furnish the Abc-homologue AIM-Z-YPK (4), which was obtained through standard Fmoc-based solid-phase peptide synthesis (SPPS) (Fig S1). Applying the same strategy, we obtained a panel of Abc ligands for HLA-A*11:01, HLA-A*02:01 and H2-Kb (Table S1, Fig S1); covering allelic variants of MHC predominantly found in Asian, and Caucasian populations, as well as in common murine disease models.

[0185] To determine whether the Abc-ligands' binding to the MHC product they were designed for was unperturbed by the tetrapeptide isostere, we started with a UV-sensitive complex, discharged its peptide cargo by traditional irradiation, and subsequently measured the capability of the Abc-ligand (which is inert to photocleavage) to prevent disintegration of the emptied complex by MHC stability ELISA (FIG. 1b for HLA-A*11:01, Fig S2 for HLA-A*02:01 and H2-Kb). As the protein requires all subunits to maintain a stable conformation, peptides that rescued the complex were deemed appropriate to produce purified Abc-ligand:MHC molecules (Fig S3). Definitive proof of peptide association, and the molecular details in which the Abc ligand binds the MHC, was furnished by X-ray crystallographic studies. The structure of HLA-A*11:01 in complex with AIM-Z-YPK (4) was determined to 2.43 resolution (Table S3, Fig S4). The conditional ligand (FIG. 1d, cyan) engages the HLA in a way very similar to the parent peptide (FIG. 1d, yellow) and occupies the peptide-binding groove by preserving crucial hydrogen bonds and salt bridges formed by the parent peptide via its N- and C-terminal anchor residues (Table S4, Fig S5). The central azobenzene moiety protrudes straight from the groove, is solvent exposed, and sufficiently straddles the four amino acids it was designed to replace. We furthermore observed 4 to occupy two alternative confirmations in the crystal (Fig S6). Optimization of the size exclusion protocol demonstrated that the two refolded A*11:01 complexes could be separately obtained (Fig S7a and b). These molecular species did not interconvert when left in solution, and had hydrodynamic volumes of 98 and 171 as judged by dynamic light scattering, that likely corresponding to the monomeric and dimeric MHCs respectively (Fig S7c). A second crystal structure of HLA-A*02:01 binding to GLS-Z-RL (17) at 2.1 resolution essentially displays the same features (but in this case no alternative conformation for the ligand is observed, Tables S5 and S6, Fig S8 and S9).

[0186] We next examined how to facilitate rapid and complete Abc-peptide exchange. Exposure of 4 to dithionite indeed resulted in fragmentation towards the expected two aniline products 7 and 8 as confirmed by LC/MS (FIG. 2a). By mixing in a stable internal standard 9 at 1:1 ratio, and interrupting the reaction (ranging from of 1 to 5 mM of sodium dithionite) by solid phase extraction, the kinetics could be tracked by LC/MS analysis (FIG. 2b). An incubation period of 5 min with 2.5 mM Na2S2O4 (aq) was sufficient for the original Abc ligand to fall below the limit of detection, indicative of (near) quantitative peptide cleavage in solution.

[0187] The peptide exchange efficiency (spanning 5 to 20 mM dithionite) was analyzed by ELISA on purified Abc-ligand:MHC complexes with established T cell epitopes (Table S2). Reduction-promoted peptide exchange could be observed at all tested dithionite concentrations (FIG. 2c for A*11:01, Fig S10 for A*02:01, and Kb). Disulfide bonds remain intact under these conditions (Fig S11) and the method therefore appears to have limited effect on the overall stability of the protein complexes. For HLA-A*11:01, the highest signal-to-background ratio was obtained at 20 mM Na2S2O4, yet this trend was the reverse for HLA-A*02:01, highlighting that every allelic variant carrying a tailored Abc-ligand will have unique stability characteristics.

[0188] A further impetus for moderating the amount of employed reducing agent is to prevent toxicity towards cells. It would be preferable that the MHC tetramers of novel specificity can be directly deployed, which involves them being shortly (<1 h) incubated with CD8+ T cells, without requiring the removal of any component (i.e. employed reagents or side-product) that could unnecessarily lengthen or complicate the peptide exchange and/or staining protocol. Both primary and cultured cells of various origins, fortunately, were very tolerant to buffered dithionite, showing little sign of apoptosis or cell death at high (10 mM) concentration and prolonged (16 h) exposure (FIG. 2d, Fig S12). Balancing the above constraints, we employed 10 mM Na2S2O4 (aq) in the ensuing experiments.

[0189] To confirm that our strategy enables detection of antigen-specific cells from peripheral blood, a short-term expanded T cell line from an A*11:01-carrying donor responsive to Epstein Barr Virus (EBV) antigen (BRLF1134-142, 1, FIG. 3 top row) was labeled with MHC tetramers before, and after replacement with the canonical epitope (1), and an irrelevant peptide (10). This established that MHC tetramers generated through chemical- or UV-mediated exchange were equally capable in detecting frequencies of CD8+ T cells (i.e. 4.16% and 4.03%, respectively) only of the correct specificity and with minimal background. This could be replicated in an alternative CD8+ T cell line reactive towards Influenza A M113-21 peptide (10) when presented by A*11:01 (FIG. 3, bottom row), and reductive exchange of Abc-ligands was comparably successful for human HLA-A*02:01 and murine H2-Kb tetramers (FIGS. S13 and S14 respectively).

[0190] Next to preserving protein integrity, it is vital that cleavage conditions do not alter any functionality on the replacement epitope either. Such modifications could pose problems when they occur on critical residues that anchor the peptide to the MHC or are important for T cell receptor engagement, possibly resulting in failure to identify a given T cell population. A major limitation, for example, of vicinal diol- or alkanolamine-containing amino acids that can be cleaved by periodate, is that e concomitant oxidation of the Cys-, Met-, N-terminal Ser- or Thr-residues can be oxidised.[7] We therefore compared reductive (i.e. 10 mM dithionite) with oxidative (i.e. 0.3 mM periodate) cleavage conditions on well-established T cell epitopes containing said residues. Incubation with periodate predictably cleaved the N-terminal Ser of A*11:01-restricted Influenza A MP13-21 epitope (10), and (partially) oxidized the Cys and Met of EBV BMLF-1259-267 epitope (11), whereas dithionite treatment left the epitopes unaffected (FIG. 4).

[0191] Collectively, we have established a truly bio-orthogonal and robust strategy for conditional peptide exchange based on a unique panel of chemolabile Abc-ligands that can provide functional libraries of T cell labeling reagents both for human MHC molecules frequently found in both Asian and Caucasian populations, as well as for murine MHC. The true value of our method lies in the facile epitope replacement without the need for dedicated UV-irradiation equipment under conditions that are neither detrimental to the protein, the epitope, nor to the cells. Broad population coverage, through the inclusion of diverse MHC allelic variants, is currently under development, as we believe this will allow widespread application of this high-throughput method with which we can tackle the sprawling diversity of biologically relevant T cell populations in both basic research and clinical settings.

[0192] Alternate conformations of the Abc ligand in the HLA-A*11:01 complex We also observed electron density suggesting that the Abc ligand has an alternate conformation that is non-canonical to peptide-HLA binding; the C-terminal portion of the Abc ligand proceeding from the azobenzene group flips out and binds the adjacent molecule in the asymmetric unit, forming what appears to be a cross-link that would allow the two MHC molecules to dimerize (Fig S6). The occupancies of the canonical and non-canonical conformation were estimated to be 36% and 64% respectively. This was calculated based on their expected average B-factor values.

TABLE-US-00001 TABLE S1 Sequences of synthesized Abc ligands Peptide Parent epitope no. Abc ligand Restriction Sequence Organism Protein Location IEDB ID 3 AIF-Z-TK A*11:01 AIFQSSMTK Human Reverse 158 to 166 1913 immunodeficiency transcriptase virus 1 4 AIM-Z-YPK A*11:01 IMPARFYPK Hepatitis B virus DNA polymerase 110 to 118 27530 (2024) (peptide homologue) 5 QVPL-Z-YK A*11:01 QVPLRPMTYK Human Nef protein 73 to 82 52760 immunodeficiency virus 1 6 KTF-Z-PK A*11:01 KTFPPTEPK SARS coronavirus Nucleoprotein 362 to 370 33667 13 FLPS-Z-SV A*02:01 FLPSDFFPSV Hepatitis B virus Core protein 18 to 27 16833 14 LLF-Z-YV A*02:01 LLFGYPVYV Human Transcriptional 11 to 19 37257 T-lymphotropic activator Tax virus 1 15 NLVP-Z-TV A*02:01 NLVPMVATV Human herpesvirus 5 65 kDa lower 485 to 493 44920 matrix phosphoprotein 16 NLVP-Z-VATV A*02:01 NLVPMVATV Human herpesvirus 5 65 kDa lower 485 to 493 44920 matrix phosphoprotein 17 GLS-Z-RL A*02:01 GLSRYVARL Hepatitis B virus Polymerase 412 to 420 21145 19 FAP-Z-AL K.sup.b and D.sup.b FAPGNYPAL Sendai virus Nucleoprotein 324 to 332 15248 20 FAP-Z-KL K.sup.b and D.sup.b FAPGNYPAL Sendai virus Nucleoprotein 324 to 332 15248

[0193] Table S1: Alternate conformations of the Abc (Z) in the HLA-A*11:01 complex. We also observed electron density suggesting that the Abc ligand has an alternate conformation that is non-canonical to peptide-HLA binding; the C-terminal portion of the Abc ligand proceeding from the azobenzene group flips out and binds the adjacent molecule in the asymmetric unit, forming what appears to be a cross-link that would allow the two MHC molecules to dimerize (Fig S6). The occupancies of the canonical and non-canonical conformation were estimated to be 36% and 64% respectively. This was calculated based on their expected average B-factor values. The design of Abc ligands is based on the following parent epitopes. The restriction element, sequence, organism and protein source of the parent epitopes are listed. Residues in these epitopes that are replaced by the Abc moiety are underlined. IEDB ID refers to the epitope identification number in the immune epitope database and analysis resource (URL: http://www.immuneepitope.org/).

TABLE-US-00002 TABLE S2 Sequences of antigenic peptides Peptide no. Sequence Restriction Organism Protein Location IEDB ID 1 ATIGTAMYK A*11:01 Human Transcription activator 134 to 142 5002 herpesvirus 4 BRLF1 2 FLPSDFFPSV A*02:01 Hepatitis B virus Core protein 18 to 27 16833 10 SIIPSGPLK A*11:01 Influenza A virus Matrix protein 1 13 to 21 58567 11 GLCTLVAML A*02:01 Human BMLF1 protein 259 to 267 20788 herpesvirus 4 12 CLGGLLTMV A*02:01 Human Latent membrane 426 to 434 6568 herpesvirus 4 protein 2 21 NLVPMVATV A*02:01 Human 65 kDa lower matrix 485 to 493 44920 herpesvirus 5 phosphoprotein 18 SIINFEKL K.sup.b Gallus gallus Ovalbumin 258 to 265 58560 22 SVLAFRRL K.sup.b Toxoplasma Tgd057 57 to 64 146017 gondii 9 IPAAAGRFF L.sup.d Toxoplasma Rhoptry protein ROP7 435 to 443 103992 gondii

[0194] Table S2: Previously identified antigenic peptides that were used in this study for MHC stability ELISA and generation of peptide-specific MHC tetramers are listed. IEDB ID refers to the epitope identification number in the immune epitope database and analysis resource (URL: http://www.immuneepitope.org/).

TABLE-US-00003 TABLE S3 Data collection and refinement statistics of HLA-A*11:01:AIM-Z-YPK Data collection Name HLA-A*11:01:AIM-Z-YPK Beamline Rigaku Detector R-AXIS IV++ Space group P1 Cell dimensions a, b, c () 52.14, 71.46, 75.43 , , () 106.74, 96.74, 105.28 Resolution () 29.93-2.43 (2.494-2.431)* R.sub.merge (%) 5.2 (17) I/(I) 17.4 (6.8) Completeness (%) 94.7 (90.7) Redundancy 4 (4) Refinement Resolution () 29.93-2.43 (2.49-2.43)* Number of reflections 33278 (2260) R.sub.work/R.sub.free 0.18/0.25 Number of atoms Protein 6252 Ligand 79 Water 249 B-factors (.sup.2) Protein 30.30 Ligand 23.73 Water 27.67 RMSD values Bond lengths () 0.014 Bond angles () 1.693 Ramachandran values Most favoured (%) 96.3 Additional allowed (%) 3.7 Disallowed (%) 0.0 *Values for the highest resolution shell are shown

TABLE-US-00004 TABLE S4 Interactions between AIM-Z-YPK and HLA-A*11:01 Hydrogen-bond partner Abc-Peptide HLA Distance Van der Waals Residue Atom Residue Atom () interactions Ala1 N Tyr7 OH 3.2 Met5, Tyr7, Glu63, N Tyr171 OH 2.9 Tyr195, Arg163, Tip167, O Tyr159 OH 2.6 Tyr171 Ile2 N Glu63 O1 2.8 Tyr7, Tyr9, Met45, Glu63, Asn66, Val67, Tyr99, Tyr159, Arg163 Met3 N Tyr99 OH 3.2 Asn66, Tyr99, Arg114, Tyr159 Tyr8 O Trp147 N1 2.9 Ala152 Pro9 Asp77 Lys10 N Asp77 O1 3.0 Asp77, Thr80, Tyr84, Asp116, Thr143, Lys146 OXT Tyr84 OH 2.7 OXT Thr143 O1 3.0 O Lys146 N 3.0 N Asp116 O2 2.8 H-bond cut off <3.5 , Van der Waals: 3.6-4.0

TABLE-US-00005 TABLE S5 Data collection and refinement statistics of HLA-A*02:01:GLS-Z-RL Data collection Name HLA-A*02:01:GLS-Z-RL Beamline Swiss Light Source X06DA Detector Pilatus Space group P2.sub.1 Cell dimensions a, b, c () 57.79, 79.58, 83.97 , , () 90, 89.96, 90 Resolution () 28.89-2.10 (2.21-2.10)* R.sub.merge (%) 4.6 (9.4) I/(I) 14.0 (7.9) Completeness (%) 92.3 (81.3) Redundancy 2 (1.7) Refinement Resolution () 28.89-2.10 (2.21-2.10) Number of reflections 43603 (2861) R.sub.work/R.sub.free 0.18/0.22 Number of atoms Protein 6217 Ligand 681 Water 537 B-factors (.sup.2) Protein 16.73 Ligand 16.59 Water 23.58 RMSD values Bond lengths () 0.010 Bond angles () 1.04 Ramachandran values Most favoured (%) 98.4 Additional allowed (%) 1.6 Disallowed (%) 0.00 *Values for the highest resolution shell are shown

TABLE-US-00006 TABLE S6 Interactions between GLS-Z-RL and HLA-A*02:01 Hydrogen-bond partner Abc-Peptide HLA Distance Van der Waals Residue Atom Residue Atom () interactions Gly1 N Tyr7 OH 2.6 Met5, Tyr7, Glu63, N Tyr171 OH 2.8 Tyr159, Trp167, O Tyr159 OH 2.7 Leu2 O Lys66 N 2.9 Tyr7, Glu63, Lys66, Val67, Tyr99, His70, Tyr159 Ser3 O Tyr99 OH 2.9 Lys66, His70, Tyr99, N Tyr99 OH 2.9 Tyr159 Abc O1 Tyr116 OH 3.5 Arg8 O Trp147 N1 2.8 Thr73, Val76 Leu9 N Asp77 O1 3.1 Asp77, Leu81, Tyr116, OXT Tyr84 OH 2.8 Tyr123, Thr143, OXT Thr143 O1 2.7 Trp147 H-bond cut off <3.5 , Van der Waals: 3.6-4.0

TABLE-US-00007 TABLE S7 Common human Crystal Reference MHC Class II Frequency structure Design of three (PubMed molecules of allele.sup.a) (PDB ID#).sup.b) Parent ligand.sup.c) Abc ligands.sup.e) ID).sup.e) HLA-DR1 (DRA, 5% 2IAM GELIGILNAAKVPAD GELI-Abc-AAKVPAD 17334368 DRB1*01:01) GELIGI-Abc-KVPAD GELIGIL-Abc-VPAD HLA-DR2 (DRA, 13% 1YMM ENPVVHFFKNIVTPR ENPVV-Abc-NIVTPR 15821740 DRB1*1501) ENPVVH-Abc-IVTPR ENPVVHFF-Abc-TPR HLA-DR4 (DRA, 17% 3O6F FSWGAEGQRPGFG FSW-Abc-QRPGFG 21297580 DRB1*04:01) FSWG-Abc-RPGFG FSWGAE-Abc-GFG HLA-DP2 4%, 6% 3LQZ RKFHYLPFLPST RKF-Abc-FLPST 20356827 (DPA1*01:03, RKFH-Abc-LPST DPB1*02:01) RKFHYL-Abc-ST HLA-DQ8 23%, 43% 4GG6 QQYPSGQGSFQPSQQNPQ QQYPSGQ-Abc-PSQQNPQ 23063329 (DQA1*03:01, QQYPSGQG-Abc-SQQNPQ DQB1*03:02) QQYPSGQGSF-Abc-QNPQ Common mouse Crystal Reference MHC Class II structure Design of three (PubMed molecule N/A (PDB ID#).sup.b) Parent ligand Abe ligands.sup.c) ID).sup.d) H-2-IAb 3C5Z FEAQKAKANKAVD FEA-Abc-ANKAVD 18308592 FEAQK-Abc-KAVD FEAQKA-Abc-AVD H-2-IAd 1IAO ISQAVHAAHAEINEAGR IS-Abc-AAHAEINEAGR 9529149 ISQA-Abc-HAEINEAGR ISQAV-Abc-AEINEAGR H-2-IEk 1KT2 ADLIAYLKQATK ADLI-Abc-QATK 11956295 ADLIA-Abc-ATK ADLIAYL-Abc-K .sup.a)http://www.ncbi.nlm.nih.gov/projects/gv/mhc/ihwg.cgi?cmd=PRJOV&ID=9 .sup.b)http://www.rcsb.org./pdb/home/home.do .sup.c)P1, P4, P6 and P9 anchor residues are indicated by bold and underlined format .sup.d)Structural design of Abc conditional ligands, MHC binding and fragmentation is achieved if the Abc moiety is incorporated betwe the critical P1 and P9 anchors, and replaces 4 amino acid residues. .sup.e)http://www.ncbi.nlm.nih.gov/pubmed indicates data missing or illegible when filed

CITED ART

[0195] [1] a) E. M. Sletten, C. R. Bertozzi, Angew. Chem. Int. Ed. Engl. 2009, 48, 6974-6998; b) M. Grammel, H. C. Hang, Nat. Chem. Biol. 2013, 9, 475-484. [0196] [2] a) G. Leriche, L. Chisholm, A. Wagner, Bioorg. Med. Chem. 2012, 20, 571-582; b) G. C. Rudolf, W. Heydenreuter, S. A. Sieber, Curr Opin Chem Biol 2013, 17, 110-117. [0197] [3] a) J. D. Altman, P. A. H. Moss, P. J. R. Goulder, D. H. Barouch, M. G. McHeyzer-Williams, J. I. Bell, A. J. McMichael, M. M. Davis, Science 1996, 274, 94-96; b) M. M. Davis, J. D. Altman, E. W. Newell, Nat. Rev. Immunol. 2011, 11, 551-558. [0198] [4] a) M. Toebes, M. Coccoris, A. Bins, B. Rodenko, R. Gomez, N. J. Nieuwkoop, W. van de Kasteele, G. F. Rimmelzwaan, J. B. A. G. Haanen, H. Ovaa, et al., Nat. Med 2006, 12, 246-251; b) G. M. Grotenbreg, M. J. Nicholson, K. D. Fowler, K. Wilbuer, L. Octavio, M. Yang, A. K. Chakraborty, H. L. Ploegh, K. W. Wucherpfennig, J. Biol. Chem. 2007, 282, 21425-21436; c) A. H. Bakker, R. Hoppes, C. Linnemann, M. Toebes, B. Rodenko, C. R. Berkers, S. R. Hadrup, W. J. E. van Esch, M. H. M. Heemskerk, H. Ovaa, et al., Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 3825-3830; d) G. M. Grotenbreg, N. R. Roan, E. Guillen, R. Meijers, J.-H. Wang, G. W. Bell, M. N. Starnbach, H. L. Ploegh, Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 3831-3836; e) E.-M. Frickel, N. Sahoo, J. Hopp, M.-J. Gubbels, M. P. J. Craver, L. J. Knoll, H. L. Ploegh, G. M. Grotenbreg, J. Infect. Dis. 2008, 198, 1625-1633; f) S. Gredmark-Russ, E. J. Cheung, M. K. Isaacson, H. L. Ploegh, G. M. Grotenbreg, Journal of Virology 2008, 82, 12205-12212; g) C. X. L. Chang, A. T. Tan, M. Y. Or, K. Y. Toh, P. Y. Lim, A. S. E. Chia, T. M. Froesig, K. D. Nadua, H.-L. J. Oh, H. N. Leong, et al., Eur. J. Immunol. 2013, 43, 1109-1120. [0199] [5] a) S. H. L. Verhelst, M. Fonovi, M. Bogyo, Angew. Chem. Int. Ed. Engl. 2007, 46, 1284-1286; b) Y.-Y. Yang, M. Grammel, A. S. Raghavan, G. Charron, H. C. Hang, Chemistry & Biology 2010, 17, 1212-1222; c) F. Landi, C. M. Johansson, D. J. Campopiano, A. N. Hulme, Org. Biomol. Chem. 2010, 8, 56-59; d) G. Budin, M. Moune-Dimala, G. Leriche, J.-M. Saliou, J. Papillon, S. Sanglier-Cianfrani, A. Van Dorsselaer, V. Lamour, L. Brino, A. Wagner, Chembiochem 2010, 11, 2359-2361; e) G. Leriche, G. Budin, L. Brino, A. Wagner, Eur J Org Chem 2010, 4360-4364. [0200] [6] a) J. Sidney, H. M. Grey, S. Southwood, E. Celis, P. A. Wentworth, M. F. del Guercio, R. T. Kubo, R. W. Chesnut, A. Sette, Hum. Immunol. 1996, 45, 79-93; b) T. Blicher, J. S. Kastrup, L. . Pedersen, S. Buus, M. Gajhede, Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 2006, 62, 1179-1184. [0201] [7] a) B. Rodenko, M. Toebes, P. H. N. Celie, A. Perrakis, T. N. M. Schumacher, H. Ovaa, J. Am. Chem. Soc. 2009, 131, 12305-12313; b) A. Amore, K. Wals, E. Koekoek, R. Hoppes, M. Toebes, T. N. M. Schumacher, B. Rodenko, H. Ovaa, Chembiochem 2013, 14, 123-131.