key: cord-0823114-5g1ute03
authors: Chatzileontiadou, Demetra S.M.; Szeto, Christopher; Jayasinghe, Dhilshan; Gras, Stephanie
title: Protein purification and crystallization of HLA-A*02:01 in complex with SARS-CoV-2 peptides
date: 2021-06-09
journal: STAR Protoc
DOI: 10.1016/j.xpro.2021.100635
sha: 2decc3bafd3a10d2fac7fecf03d4e184102ee322
doc_id: 823114
cord_uid: 5g1ute03

Understanding T cell responses requires identifying viral peptides presented by Human Leukocyte Antigens (HLA). X-ray crystallography can be used to visualize their presentation. This protocol describes the expression, purification, and crystallization of HLA-A*02:01, one of the most frequent HLA in the global population in complex with peptides derived from the SARS-CoV-2 Nucleocapsid protein. This protocol can be applied to different HLA-class I molecules bound to other peptides.

This protocol can be applied with a variety of different peptides. It has been optimized for HLA-A*02:01 molecule but it has been used for different HLA-class I molecules by optimizing the refolding time (see below) depending on the HLA protein of interest. Here, we describe the complex formation with a SARS-CoV-2 Nucleocapsid protein-derived peptide N 222-230 LLLDRLNQL. Other peptides can be substituted but the conditions may vary as described in Szeto C, et al., 2021 (iScience, doi: 10.1016 /j.isci.2021 where five different SARS-CoV-2 peptides have been used.

Timing: 6 hours Note: A DNA plasmid (here, pET-30a(+) vector, but any vector for protein expression in bacterial system can be used) encoding the heavy chain of the HLA-class I molecule of interest, here HLA-A*02:01, and a plasmid encoding human β2-microglobulin need to be available and be kept at -20 o C.

Note: The peptide of interest, here peptides derived from SARS-CoV-2 Nucleocapsid protein, needs to be at least 95% pure, usually as powder kept at -20 o C.

Note: When working with bacterial cultures make sure everything is sterile. Clean everything (bench, pipettes, etc.) with 70% EtOH, and work in aseptic conditions (under flame) to avoid any contamination.

Note: For the protein expression, different bacterial transformation methods can be also used, such as electroporation, with the use of electrocompetent E. coli protein expression strains instead. Step-by-Step Method Details

Expression of HLA-A*02:01 and human β2-microglobulin & extraction of insoluble proteins (Inclusion Bodies)

Timing: 5 days This step describes the expression of HLA-A*02:01 heavy chain and human β2-microglobulin proteins as well as their recovery from inclusion bodies from Escherichia coli expression using a mild solubilization process. This protocol can be applied to a variety of different HLA-class I molecules.

1. HLA-A*02:01 heavy chain and β2-microglobulin expression. (The protocol can be applied for both  proteins) DAY 1, Timing: 2 hours a. Chemical transformation of competent BL21 (DE3) E. coli cells with the pET-30a (+) vector encoding HLA-A*02:01 heavy chain, or β2-microglobulin.

Note: The gene sequence of β2-microglobulin (sequence below) was inserted in pET-30a (+) that was cut with NdeI/HindIII restriction enzymes IQRTPKIQVYSRHPAENGKSNFLNCYVSGFHPSDIEVDLLKNGERIEKVEHSDLSFSKDWSFYLLYYTEFTPTEKDEYAC RVNHVTLSQPKIVKWDRDM The gene sequence of HLA-A*02:01 heavy chain (sequence below) was inserted in the pET-30a (+) vector that was cut with NdeI/HindIII restriction enzymes. GSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAASQRMEPRAPWIEQEGPEYWDGETRKVKAHSQTHR VDLGTLRGYYNQSEAGSHTVQRMYGCDVGSDWRFLRGYHQYAYDGKDYIALKEDLRSWTAADMAAQTTKHKWE AAHVAEQLRAYLEGTCVEWLRRYLENGKETLQRTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLTWQRDGED QTQDTELVETRPAGDGTFQKWAAVVVPSGQEQRYTCHVQHEGLPKPLTLRWEPSS Both sequences and successful sub-cloning into pET30a vector have been confirmed by sequencing by Genscript.

J o u r n a l P r e -p r o o f i.

Thaw competent cells on ice. Note: competent cells are very "fragile", do not warm up the cells by holding the bottom of the tube in your fingers. Note: work under a flame, or in a fume hood, to maintain aseptic conditions. Alternatively, if not applicable, make sure that everything is sterile by cleaning bench, pipettes, etc. with 70 % EtOH to avoid any contamination.

ii.

Add 0.5-1 μL (<100 ng) of plasmid to the competent cells. Note: Use 50-100 ng of plasmid. The volume of plasmid added to the competent cells shouldn't exceed 10 % of the final volume as the H 2 O may cause lysis of the cells.

iii.

Incubate cells-DNA mixture on ice for 20 mins. iv.

Heat shock competent cells at 42°C in a water bath (or heat block) for 45 seconds. v.

Add 1 Inoculate 6x 800 mL of LB broth containing 34 μg/mL Kanamycin with 8 mL of starter culture (100-fold dilution) and grow at 37°C until the OD 600nm reaches 0.6 (approximately 3 hours).

Note: The volume of bacterial cultures determines the final protein yield and can be adjusted to your needs Note: When growing bacterial cultures, it is important not to let the OD get any higher than 0.6. The OD should be carefully monitored and checked often, especially when it gets above 0.2, as the cell growth is exponential.

ii. Add 800 μL of 0.5 M isopropyl-β-D-thiogalactopyranoside (IPTG) to each 800 mL culture to induce protein expression (HLA-A*02:01, or β2-microglobulin) and incubate for 4 hours. iii.

Harvest cells by centrifugation at 5000x g for 15 mins at 4 o C. Resuspend cell pellets in MilliQ H 2 O (20 mL in total for the entire 6x 800 mL culture). We recommend using 50 mL falcon tubes for easy storage and manipulation. iv.

Store cell pellet in the -80°C (long term storage) or -20°C (short term storage) freezer. Pause Point: You can either proceed to the preparation of insoluble proteins (Inclusion bodies) or store the cell pellets as described above. Alternatives: Other methods can also be used for the lysis of bacterial cells (e.g. French press, sonication). Homogenise and centrifuge at 10,000 g at 4˚C for 10 min. iv.

Repeat i-iii until pellet is clean of dark debris (minimum of 4 washes). v.

Discard supernatant and resuspend pellet in 150 mL of Wash Buffer 2. vi.

Add 1 mM DTT (150 μL of 1 M stock) and 0.2 mM PMSF (150 μL of 200 mM stock). vii.

Homogenise and centrifuge at 10,000 g at 4˚C for 10 min. viii.

Discard supernatant and resuspend pellet in 5 mL Guanidine·HCl Buffer for 6x 800mL culture (1 mL/L of culture), on a rotating wheel for 14-17 hours at 4˚C. Note: If the mixture turns into a jelly-like consistency, add in 10 mM DTT.

Note: For problems that may arise during this procedure please refer to Troubleshooting Problems 1 and 2 below. Note: The molecular mass of HLA α-chain is ~32 kDa and human β-2-microglobulin is ~10 kDa.

Aliquot them out to:  10 mg for human β2-microglobulin  30 mg for HLA-A*02:01 heavy chain ii. Store in -80˚C or proceed to the next step (Refolding).

Pause Point: You can proceed to the refolding steps below or store the inclusion bodies long term at -80 o C.

Refolding of the HLA-class I molecule with human β2-microglobulin and the peptide of interest (SARS-CoV-2 N 222-230 LLLDRLNQL) followed by dialysis Timing: 1.5 days

The refolding step produces soluble peptide-HLA complexes. During this step, the pHLA complex will be formed, which is composed of the HLA-A*02:01 heavy chain, human β2-microglobulin, and the peptide of interest, in this case the SARS-CoV-2 N 222-230 LLLDRLNQL peptide. a. Prepare, wash and equilibrate the DEAE-C resin as described in the "Before You Begin". b. Filter dialyzed protein sample using a 0.22 μm syringe filter and load onto the column at 4°C. c. Elute bound proteins with 5 CV (column volume) of 10 mM Tris-HCl pH 8.0, 150 mM NaCl. d. Visualise your protein purity on SDS-PAGE. e. DEAE resin can be washed with 1 M NaCl followed by 10 mM Tris-HCl pH 8.0 and stored in 20 % EtOH to be reused at 4°C.

DAY 7, Timing: 3 hours 6. Ion exchange chromatography -HiTrap Q HP, 1mL a. Concentrate the eluted protein using a centrifugal concentrator with a 10 kDa cut-off and buffer exchange over 10 mM Tris-HCl pH 8 four times.

Critical: This step is crucial as NaCl needs to be removed from the sample before loading it onto the HTQ column.

Note: Protein of interest must be filtered with 0.22 μm filter before loading onto the column to remove particulate material. The starting volume of the sample should be less than 1 mL for the HTQ column and diluted in equilibration buffer (see below) to less than 5 mL (5 fold dilution). 5 mL loop should be used on FPLC machine for injection onto the HTQ column. Check the pressure limit for the column and set this parameter on the FPLC. If sample volume is higher than the volume of the loop multiple injections can be done.

b. Connect the column on the FPLC with desired buffer (Pump A: 10 mM Tris-HCl pH 8.0). c. Wash the column with five column volumes of 10 mM Tris-HCl pH 8.0, 1 M NaCl (Pump B). d. Equilibrate the column with five column volumes of 10 mM Tris-HCl pH 8.0 (Pump A). e. Start fractionation before protein injection. Ensure the deep well or tubes are at the right position. Collect 1 mL fractions. f. Inject protein sample and run at a flow rate of 1 mL/min. g. After two loop volumes (10 mL) set a gradient from 0 to 20 % of 10 mM Tris-HCl pH 8.0, 1 M NaCl (Pump B) over 20 mins. h. Usually, peptide-HLA complex elute at ~15 % of 10 mM Tris-HCl pH 8.0, 1 M NaCl although this can vary between different proteins. i. Run SDS-PAGE gel of the fractions of interest. j. Pool the fractions needed based on purification profile and gel analysis. k. Concentrate the protein sample using a 10 kDa cut-off concentrator and measure the concentration of protein using a UV spectrophotometer at 280 nm absorbance. The protein sample needs to be concentrated to the desired concentration for crystallization. The HLA-A*02:01-SARS-CoV-2 N 222-230 LLLDRLNQL complex was concentrated at 5 mg/mL to set up crystal trays.

Note: Beer-Lambert law is used to calculate concentration based on absorbance at 280nm: A = εcl, where A = Absorbance; ε = Molar absorption coefficient, M -1 cm -1 ; c = Molar concentration, M; l= optical path length, cm.

Note: The protein sample can be stored at 4 o C short term or at -20 o C long term. However, it is highly recommended to be used straight away to set up trays.

Note: For problems that may arise during this procedure please refer to Troubleshooting Problems 3 and 4 below.

Crystallization of the HLA-A*02:01-SARS-CoV-2 N 222-230 LLLDRLNQL complex Timing: time varies for crystal formation (hours to months)

This step describes the crystallization process of the HLA-A*02:01-N 222-230 LLLDRLNQL (pHLA) complex. The condition used is: 20 % PEG3350 w/v, 0.2 M Potassium Formate, 1mM CaCl 2 , via sitting-drop, vapour diffusion at 20˚C with a protein: reservoir drop ratio of 1:1, at a protein concentration of 5 mg/mL in 10 mM Tris-HCl pH 8.0, 150 mM NaCl (Szeto et al., 2021) . This process can be applied for a variety of different proteins/ peptides/ protein complexes, however, many factors such as the crystallization condition, protein concentration, use of seeds and additives, can vary significantly.

Note: The purity of the protein sample should be evaluated prior to crystallization. For the initial crystallization screening, the protein sample should be at least 95 % pure on SDS-PAGE. It is recommended, if possible, to also evaluate the homogeneity and monodispersity of the protein sample as both factors can significantly influence the crystallization outcome.

Critical: Avoid freeze/thawing protein samples multiple times and try to use freshly purified protein for crystallisation. Try to set up the trays within a week after the purification or pass your protein sample through a HiTrap Q HP column again to remove any unfolded material prior to crystallisation trial set up.

Timing: 2 hours 7. Set up sitting drop 48-well-plates according to or using reagent formulations from Hampton Research commercial screening kits. The kit used here is the PEG/Ion HT Crystallization reagent kit. a. Prepare approximately 50 μL (per plate to be set up) of pHLA-A*02:01 peptide complex with concentration at approximately 5 mg/mL in the buffer used for protein purification (10 mM Tris-HCl pH 8.0, 150 mM NaCl). Note: The concentration of the protein needed to form crystals can vary.

b. Pipette 100 μL of crystallisation solution into the deep well c. Pipette 1 μL of crystallisation solution from each deep well onto the well, followed by 1 μL of the protein sample d. Seal the plate with Crystal Clear plate Sealing Tape and store the plate at 20C Note: Alternatively, hanging drop plates can be used with silicon glass cover slides. e. Wait until crystals appear Note: Crystallization trial is a multi-factor process, and not all factors are under your control. Please see below Troubleshooting Problem 5 for some alternative tips that may facilitate the process.

Timing: 2 hours J o u r n a l P r e -p r o o f

Depending on the crystal quality such as size, shape, and diffraction quality, optimization steps may be needed. Here we optimized the initial conditions that crystals were grown in by using the Additive Screen HT and micro seeds from the drop that gave crystals. a. Set up sitting drop 48-well-plates using the condition that produced crystals. b. Pipette 1 μL of the crystallisation solution from each deep well onto the well, followed by 1 μL of the protein.

Note: The concentration of the protein may need to be adjusted depending on the initial crystals. For example, if there are multiple clusters of crystals in the drop, or high precipitation, the protein concentration may need to be decreased.

c. Add 0.2 μL of additive included in the Additive Screen HT in your drop (protein and crystallization solution), in this case 1mM CaCl 2 was used. d. Add crystal micro seeds if initial crystals have been obtained. Crush up the seeds finely and add 0.2 μL of crushed seeds into each drop. e. Seal the plate with Crystal Clear plate Sealing Tape and store the plate at 20 o C. f. Wait until crystals appear.

The yield of Inclusion Bodies for HLA-A*02:01 heavy chain is typically 125 mg/ L of bacterial culture, and for human β2-microglobulin the typical yield is 48 mg/ L of bacterial culture. The purity of the Inclusion Bodies can be increased by increasing the number of washes with Inclusion Bodies Wash Buffer 1 (Figure 1) .

The final yield of pHLA is expected to be 0.5 -5 mg, depending on the peptide and the HLA molecule. After the first purification step, using the DEAE-C resin, the purity of the pHLA complex is expected to be > 80 % as shown in Figure 2 . During the second purification step, the pHLA complex elutes in approximately 15 mS/cm in the HiTrap Q HP column (Figure 3 ). The purity of the complex is expected to be > 90 % as shown in Figure 4 .

Crystals of the pHLA complex were grown in 20 % PEG3350 w/v, 0.2 M Potassium Formate, 1mM CaCl 2 ( Figure 5A ) via sitting-drop, vapour diffusion at 20˚C with a protein: reservoir drop ratio of 1:1, at a protein concentration of 5 mg/mL in 10 mM Tris-HCl pH 8.0, 150 mM NaCl (Szeto et al., 2021) . Crystals appeared in 48 hours. Optimization trials using micro-seeds from the drop shown in Figure 5A led to bigger crystals, as shown in Figure 5B , that appeared in 24 hours.

Limitations of the protocol are mentioned in the Troubleshooting section below. 

Firstly, check that the speed and time for centrifugation is adequate. We recommend using 250 mL centrifuge bottles to hold the 150 mL solutions of wash buffer. Secondly, the highly viscous consistency can be caused by excess nucleic acids so ensure that you have added the appropriate amount of DNase. Thirdly, check that the E. coli cells are resuspended in solution. If cells are not completely lysed, the wash buffer solutions can permeabilise cell membranes leading to spillage and contamination of inclusion bodies with nucleic acids and cellular debris. If enzymatic lysis is insufficient, resuspend the cell pellet in lysis buffer, homogenise the cell pellet and allow additional time (30 minutes -1 hour) for enzymatic lysis to occur. If this is still insufficient, consider using a French Press or sonication to fully lyse cells before proceeding to inclusion body washing.

Inclusion bodies resuspended in 6 M Guanidium HCl turned into a jelly-like consistency after the 14-17 hours incubation (Step 2 -Preparation and extraction of insoluble proteins).

This viscosity is caused by reformation of disulphide bonds from denatured proteins. Adding in small amounts of DTT (20 mM DTT) and allowing the mixture to continue rotating at ~25 o C will re-solubilize the protein.

Problem 3: 8. The protein elution profile from HTQ (anion exchange chromatography) has a "shoulder" peak (

Step 6 -Ion exchange chromatography -HiTrap Q HP).

A number of HLAs including HLA-A*02:01 and the mouse MHC H-2Db elute with shoulder peaks (before or after the main peak). This could be due to oligomerisation at high protein concentrations that reflect multiple isoelectric points and therefore elution of proteins at different salt concentrations. It is always important to check both peaks on an SDS-PAGE to confirm purity of the refolded pHLA complex. If the gel shows contamination with other proteins, size-exclusion chromatography should be performed to further purify the protein sample. Remember that if the pHLA sample is used for crystallisation trial, the purity of the protein is critical to enable crystal formation. Potential Solution:

Multiple factors affect the crystallization procedure including the solubility of the protein in certain buffers, the choice of the right precipitant, the pH of the solution, and the temperature (20 o C, 16 o C, or 4 o C). Instead of setting up trays manually, fast screening of crystallization conditions using sitting drop techniques in 96-well plates with sealing film can be used instead by using commercially available sparse matrix or grid screen solution kits (Jancarik et al., 1991; Cudney et al., 1994; Tran et al., 2004 ) (plates and kits available at: Emerald Biostructures Inc.; Hampton Research; Molecular Dimensions; Qiagen-Nextal Biotechnology, Canada; Jena Bioscience, Germany; Axygen Biosciences), following the order: (i) sparse matrix, (ii) matrix screen PEG/ions, (iii) grid screen ammonium sulfate, (iv) grid screen PEGs, (v) grid screen PEG/LiCl, (vi) grid screen alcohols, (vii) grid screen salts. Shake crystallization solutions (i.e., sparse matrix or grid screen kit solutions) thoroughly before use. Dispense 100 μL of crystallization solution in the deep well and drops consisting of 0.5-1 ul of protein solution + 0.5-1 μl of crystallization solution (equal amounts).

Note: Pipette the crystallization solution into the protein solution and let the two diffuse together without mixing. Once taken from the refrigerator, the screening solutions should be kept approximately 30 min at ~25 o C to allow equilibration.

This procedure facilitates the finding of the condition where the protein forms crystals. Using this condition, optimization steps may be needed in order to obtain good quality crystals, such as the use of seeds and/ or additives, as well as a range of pH, protein and precipitant concentrations.

Lead Contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Professor Stephanie Gras, s.gras@latrobe.edu.au

This study did not generate new unique reagents.

This study did not generate/analyze datasets/code. peptide. Elution peak was pooled together and an 8 μL aliquot was taken to run the SDS-PAGE. As the pHLA protein complex is non-covalently linked, on an SDS-PAGE gel, it is shown as two chains: a heavy α chain (~32 kDa) and a β-2-microgolubin chain (~10 kDa). First lane of the gel is the protein molecular weight marker in kDa (MW). Excess of human β2-microglobulin elutes at around 8 mS/cm while pHLA complex elutes at around 12-18 mS/cm. Protein that hasn't refolded is eluted at 100% buffer B (aggregation peaks). The top bar of the graph shows the numbers of the collected fractions (i.e. A/66-A/68). Figure 4 . SDS-PAGE performed for Hi Trap Q elution for the HLA-A2 in complex with the SARS-CoV-2 N 222-230 peptide. Elution peak was pooled together and an 8 μL aliquot was taken to run the SDS-PAGE.

As pHLA protein complexes are non-covalently linked, on an SDS-PAGE gel, it is shown as two chains: a heavy α chain (~32 kDa) and a β-2-microgolubin chain (~10 kDa). First lane of the gel is the protein molecular weight marker in kDa (MW). 

Screening and optimization strategies for macromolecular crystal growth

A systematic comparison of sitting and hanging-drop crystallization using traditional and crossdiffusion microbatch crystallization plates

Sparse matrix sampling: a screening method for crystallization of proteins

The presentation of SARS-CoV-2 peptides by the common HLA-A * 02:01 molecule

Statistical experimental design of protein crystallization screening revisited

The authors have no competing interests.