Structural basis of protocadherin-1 recognition by New World hantaviruses

Protein purification

Soluble spike proteins

To produce the soluble ANDV spikes, we inserted synthetic genes for Gn and Gc of ANDV (strain Chile-9717869, UniProt: Q9E006), codon-optimized for expression in Drosophila cells in a bicistronic plasmid to produce the p-sSpike plasmid. We designed the hetero-octamerization domain from HALC4_136, a synthetic homo-tetramerization domain formed by three antiparallel helices, which has the appropriate dimensions (PDB: 8D09)²⁰. In the first cistron of the plasmid, we cloned the ectodomain of Gn (amino acids 21 to 483) followed by the first two helices of HALC4_136, which we call HALC4_136.^N. The second cistron contains the complete ectodomain of Gc (amino acids 652 to 1107) followed by the third α-helix of HALC4_136, named HALC4_136.^C. To prevent aggregation, we introduced four mutations into Gc: F1089A, F1093A, L1100A, and L1104A. To further stabilize the hetero-oligomerization domain, we designed a disulfide bond connecting the C terminus of HALC4_136.^N with the N terminus of HALC4_136^C. For purification purposes, we introduced a double strep tag at the C terminus of HALC4_136.^C. For structural studies, we cloned the EC1–4 extracellular cadherin domains of human PCDH1 (amino acids 58–503, UniProt: Q08174) into a pMT/BiP plasmid in frame with a His and a double strep tag to produce plasmid p-sEC1–4. We used these plasmids to obtain stable transfectants of Drosophila S2 cells (Thermo Fisher Scientific, R69007, authenticated by Thermo Fisher Scientific by karyotyping and isozyme profiling), along with pCoPuro plasmid (1:20 ratio) for puromycin selection⁴⁶. We generated stable cell lines with p-sSpike (soluble spike) plasmid alone and cell lines co-transfecting p-sSpike and p-sEC1-4 (EC1-4). Stable cell lines were selected and maintained in LM-SFM4 medium (Cytiva) containing 8 μg ml⁻¹ puromycin. Cultures of 1 to 3 L were grown in spinner flasks in medium supplemented with 1% penicillin-streptomycin antibiotics to approximately 10 L.⁷ cells per ml and protein expression was induced with 4 μM CdCl₂. After 5 days, the supernatant from S2 medium was concentrated to 40 ml using a VivaFlow 10 kDa threshold concentrator (Sartorius) and supplemented with Biolock (IBA) and 0.1 M Tris-HCl (pH 8.0), centrifuged for 30 min at 20,000 °C.gand purified by Strep-Tactin affinity chromatography using a StrepTrap HP column (Cytiva). For co-expression of soluble ANDV spikes and PCDH1 EC1-4, purification of Strep-Tactin was followed by His affinity chromatography using the TALON column (Cytiva). Proteins were then purified by size exclusion chromatography (SEC) on a Superose6 10/300 augmentation column (Cytiva) in 10 mM Tris pH 8.0, 100 mM NaCl. Protein purity and the presence of (Gn – Gc)₄ spikes and PCDH1 in different SEC fractions after co-expression were assessed by SDS–PAGE under reducing and non-reducing conditions.

ANDV-VLP

To obtain the PCDH1–EC1 complex with ANDV-VLP, the nucleic acid sequence of the first extracellular domain (amino acids 58 to 165) of human protocadherin 1 plus a C-terminal 5×GGGS linker were codon optimized, synthesized by GenScript, and inserted into the plasmid pCAG-ANDV-DM-VLP.⁵ to make the plasmid pCAG-PCDH1-EC1-GS5-ANDV-DM-VLP. The insertion took place between the native signal peptide sequence of ANDV (1–18), predicted by the SignalP 5.0 server.⁴⁷and the N-terminal residue of Gn. For protein expression, plasmid pCAG-PCDH1-EC1-GS5-ANDV-DM-VLP was mixed with 25 kDa polyethyleneimine (PEI; Polysciences) and transfected into FreeStyle 293-F cells (Thermo Fisher Scientific) at a density of 10.⁶ cells per ml. Transfected cells were incubated in a shaking incubator at 37°C with 8% CO₂ and 80% humidity for 72 hours. After transfection, supernatants were collected by centrifugation, filtered with 0.22 μm Rapid-Flow filters (Nalgene), and concentrated to approximately 40 ml using a Centricon Plus-70 filtration unit (MilliporeSigma). The concentrated supernatant was then purified by discontinuous sucrose density gradient ultracentrifugation (20%, 50%, 70% sucrose in PBS) at 4 °C, 135,000g for 2 hours. Fractions corresponding to VLPs were collected, pooled, and concentrated to approximately 60 μl in PBS (pH 7.4) for cryo-EM grid preparation.

PCDH1 fragments

To obtain soluble His-tagged versions of the human PCDH1 domains EC1-4 (sEC1-4), EC1-2 (sEC1-2), and EC1 (sEC1) for ELISA and BLI analyses, codon-optimized human PCDH1 fragments (amino acids 58-503 for sEC1-4, 58-284 for sEC1-2, and 58-160 for sEC1) were cloned. in a pCAGGS (sEC1–4 and sEC1) or pcDNA3.1 (sEC1–2) plasmid in a frame with a histidine tag. sEC1–4 and sEC1 plasmids were used to transfect Expi293 (Thermo Fisher Scientific, A14527, unauthenticated) at a density of 3 × 10⁶ cells per ml using FectoPRO DNA transfection reagent (Polyplus). After 5 days of incubation at 37°C, the supernatant was collected and centrifuged for 30 min at 3100°C.g. Proteins were purified by TALON affinity chromatography followed by SEC on the HiLoad 16/600 Superdex 75 column (Cytiva) in 10 mM Tris pH 8.0 and 100 mM NaCl. Plasmid sEC1-2, which also included a FLAG tag, was used to transfect FreeStyle 293-FS cells (Thermo Fisher Scientific) at a density of 1 × 10 .⁶ cells per ml using linear PEI (Polysciences). After 6 days of incubation at 37°C, the supernatant was collected and clarified by centrifugation at 4000 rpm for 20 min. Proteins were purified by nickel chelation chromatography with Nickel-NTA resin (Qiagen). The protein was eluted in PBS containing 250 mM imidazole and the buffer was exchanged with PBS.

Antibody

To obtain full-length IgG versions of the monoclonal antibodies used in this study, synthetic variable regions of the light and heavy chains of each antibody were cloned into pCAGGS plasmids upstream of the corresponding human IgG1 constant regions. All antibodies were expressed in Expi293 cells as described above. Antibodies were purified by protein G affinity chromatography, eluted with 0.1 M glycine pH 2.7, neutralized with 1 M Tris-HCl pH 8.0, and further purified by SEC on the HiLoad 16/600 Superdex 200 column (Cytiva) in PBS. To obtain scFv versions of the monoclonal antibodies used in this study, we cloned the linker-linked heavy and light chain variable regions (GTGGSGGGGSGGGGSGGGAS) into an in-frame pMT/BiP plasmid with an enterokinase cleavage site and a double Strep tag. These plasmids were used to generate stable transfectants of Drosophila S2 cells, as described above. Antibody fragments were purified by Strep-Tactin affinity chromatography using the StrepTrap HP column (Cytiva), followed by SEC on a HiLoad 16/600 Superdex 75 column (Cytiva) in PBS.

Cryo-EM grid preparation and data collection

Soluble spike proteins

To perform the cryo-EM studies, we diluted the purified soluble tips alone or in complex with PCDH1 EC1-4 to a concentration of 0.34 mg ml⁻¹ in 10 mM Tris pH 8.0, 100 mM NaCl and deposited 3 µl onto 300 mesh UltrAufoil R1.2/1.3 copper grids which were discharged twice for 25 s in the PELCO easyglow cleaning system (15 mA, 0.3 mBar). After 5 s waiting time, excess protein was removed for 3 s using Whatman #1 filter paper with a transfer force of −1 at 15 °C in 100% humidity before being frozen by diving into liquid ethane using Vitrobot Mark IV (Thermo Fisher Scientific). The frozen grids were imaged on the Titan Krios POLARIS system (Synchrotron Soleil) equipped with a FALCON 4i detector (Thermo Fisher Scientific), operating in super-resolution mode at a magnification of ×165,000, corresponding to a calibrated pixel size of 0.7455 Å per pixel. A full description of the cryo-EM data collection parameters is provided in Supplementary Table 1.

ANDV-VLP

For VLPs, the 300 mesh ultrAuFoil R1.2/R1.3 grids were glow discharged with a setting of 20 mA, 1 min. Next, 3 μl of concentrated PCDH1-EC1–GS5–ANDV-DM-VLP sample was added to glow discharge grids, transferred with a transfer force of 4 and a transfer time of 20 s, and submerged frozen in liquid ethane using a Vitrobot Mark IV (Thermo Fisher Scientific). The prepared grids were examined to determine the optimal ice thickness on a 200 kV Glacios (Thermo Fisher Scientific) at the University of Texas at Austin. The filtered grids were transferred to the cryo-EM facilities at UT Southwestern (UTSW) Medical Center for high-resolution data collection. Data collection was performed on the Titan Krios G4 system (Thermo Fisher Scientific) equipped with an X-FEG camera, K3 Summit (Gatan) and a BioContinuum imaging filter (Gatan). A total of 18,507 videos were acquired with a pixel size of 0.413 Å, with a total dose of 56 e.⁻ A⁻²and an energy filter slot width of 15 eV.

EM data processing and model building

Dissolvable tips

Motion correction, CTF estimation, particle selection, 2D classification, 3D classification and non-uniform 3D refinement were performed using cryoSPARC (v.4.6.2)⁴⁸. For model creation, the map was refined using DeepEMhancer⁴⁹ . The initial model was generated using ChimeraX⁵⁰ by fitting the high-resolution ANDV tetramer structure (PDB: 9P3I)⁵ and the crystallographic structure of PCDH1 EC1 (PDB: 6BX7)²⁶ in the electron density map. The model was then built manually in Coot⁵¹ and iteratively refined using phenix.refine⁵²using a refinement strategy that includes minimization_global, local_grid_search and adp, with rotamer and Ramachandran restrictions. The complete cryo-EM data processing workflow is shown in Supplementary Figure 1 and model refinement statistics are provided in Supplementary Table 1.

ANDV-VLP

For VLPs, data processing follows a similar pipeline reported for the single tetramer of the ANDV glycoprotein.⁵ . In short, the dataset was imported into CryoSPARC (v.4.6.2)⁴⁸for patch-based motion correction, contrast transfer function (CTF)…