One-shot design of functional protein binders with BindCraft

BindCraft design protocol

The input and design settings for running the BindCraft pipeline are organized into user-friendly JSON files. To initiate design trajectories, a target PDB format structure needs to be specified, along with the desired minimum and maximum length of the binders, and the desired number of final filtered designs. A target hotspot can be specified as either individual residues or entire chains, or can be omitted completely in which case a binding site is selected according to the combined design loss.

The binder hallucination process is performed using the ColabDesign implementation of AF2. The design process is initialized with a random sequence for the binder, which is predicted in single sequence mode, and a structural input template for the target. This is passed through the AF2 network to obtain a structure prediction and calculate the design loss. The design loss function is composed of several terms, with default weight values indicated in parentheses:

1. (1)

   binder confidence pLDDT (weight 0.1)

2. (2)

   interface confidence i_pTM (weight 0.05)

3. (3)

   normalized predicted alignment error (pAE) within the binder (weight 0.4)

4. (4)

   normalized predicted alignment error (pAE) between binder and target (weight 0.1)

5. (5)

   residue contact loss within binder (weight 1.0)

6. (6)

   residue contact loss between the target and binder: if hotspots are specified, the rest of the target is masked from this loss (weight 1.0)

7. (7)

   radius of gyration of binder (weight 0.3)

8. (8)

   ‘helicity loss’: penalize or promote backbone contacts every one in a three-residue offset to promote the hallucination of helical or non-helical designs (weight −0.3)

9. (9)

   optional ‘N&C termini loss’ increases the proximity of the N and C termini of the binder to allow splicing into protein loops (weight 0.1).

The loss function is used to calculate position specific errors, which are then backpropagated through the AF2 network to produce a L × 20 error gradient, where L is the sequence length. Using multiple iterations and stochastic gradient descent optimization, this error gradient is recomputed and used to optimize the input binder sequence for the next iteration to minimize the resulting loss. We backpropagate through the AF2 multimer model weights¹¹ and swap randomly between the five trained models at each iteration to ensure robust sequence generation and reduce the risk of overfitting to a single model.

As our goal is to arrive at a real discrete sequence for the binding interface, the sequence optimization is performed in four stages. The first sequence optimization stage is performed in a continuous sequence space using logit inputs. At each step, the sequence representation is based on linear combination of (1 − λ) × logits + λ × softmax(logits/T), where λ = (step + 1)/iterations and temperature (T) of 1.0. Here, many amino acids are considered per each binder position, which allows the exploration of a larger and less constrained sequence-structure space. After 50 iterations, we terminate trajectories showing poor AF2 confidence scores, as we found that such trajectories rarely converge to high confidence designs. Furthermore, if a beta-sheeted trajectory is detected, we increase the number of recycles during design from one to three to ensure accurate prediction. The continuous sequence space optimization is then continued for a further 25 iterations. During the second optimization stage, the sequence logits are normalized to sequence probabilities using the softmax function for 45 iterations to funnel the design space towards a more realistic sequence representation defined as softmax(logits/T) At each step, the temperature is lowered, where temperature is equal to (1 × 10⁻² + (1 − 1 × 10⁻²) × (1 − (step + 1)/iterations)²). The temperature is also used to scale the learning rate for rate decay. For the third stage, we implement the straight-through estimator, allowing the model to see the one-hot representation, but backpropagate through the softmax representation. This procedure is performed for five iterations. For the final fourth stage, the sequence inputs are converted to a one-hot discrete encoding. At each step, X random mutations are independently sampled and tested from the probability distribution of the softmax representation from the previous stage, and mutations with best loss are fixed. X is defined on the basis of the length of the binder sequence (0.05× binder length). This procedure is performed for 15 iterations. At the end, trajectories with pLDDT below 0.7, fewer than 7 interface contacts or significant backbone clashes are rejected.

Successful binder design trajectories are subjected to MPNN_sol sequence optimization to improve stability and solubility¹². To this end, we preserve binder residues in a 4 Å radius around the target interface, and design 20 new sequences for the remaining binder core and surface residues using the soluble weights of ProteinMPNN⁶, with a temperature of 0.1 and 0.0 backbone noise. These optimized sequences are then repredicted using the AF2 monomer model, with three recycles and two template-based models⁴⁹ in single sequence mode, to ensure robust and unbiased complex assessment. Each of the two resulting models is then energy minimized using Rosetta’s FastRelax protocol⁵² with 200 iterations, and interface scores are computed using the InterfaceAnalyzer mover⁵³ with side-chain and backbone movement enabled.

Designs are finally filtered using a set of predefined filters to ensure the selection of high quality designs for experimental testing. Filters were initially defined based on experimental observations from previous binder design studies^2,3,4,5 and refined over the course of this work. These include:

1. (1)

   AF2 confidence pLDDT score of the predicted complex (>0.8)

2. (2)

   AF2 interface predicted confidence score (i_pTM) (>0.5)

3. (3)

   AF2 interface predicted alignment error (i_pAE) (<0.35)

4. (4)

   Rosetta interface shape complementarity (>0.60)

5. (5)

   number of hydrogen bonds at the interface (>3)

6. (6)

   number of unsaturated hydrogen bonds at the interface (<4)

7. (7)

   hydrophobicity of binder surface (<35%)

8. (8)

   r.m.s.d. of binder predicted in bound and unbound form (<3.5 Å)

9. (9)

   fewer than three lysines and methionines at the binder interface.

We allow only two MPNN_sol generated sequences per individual AF2 trajectory to pass filters to promote interface diversity amongst selected binders. This design procedure is set up to loop until a defined number of final desired designs is reached. For optimal results, we recommend running the design pipeline until at least 100 designs pass computational filters. This generally requires the sampling of about 300–3,000 trajectories. We then usually pick 10 designs from the top 20 (ranked by i_pTM) for experimental testing.

To generate designs against targets described in the section ‘Accurate design of de novo binders’, we used the input structures, binder specifications and hotspot designations described in Supplementary Table 1. For AF2 predictions, we used full-length input sequences from UniProt. In all cases, the amino acid cysteine was excluded from sequence design. For AAV targets, the N-termini and C-termini loss is activated with default weight.

Computational benchmarks of BindCraft

To evaluate the flexibility of the target structure post-design, the input PDB structure of the target was aligned to the target chain A of the design trajectory, and r.m.s.d._Cα was calculated using PyRosetta. For increasing target flexibility, the sequence of the input target template was masked by enabling the flag ‘rm_target_seq’ in ColabDesign for trajectory hallucination⁵⁴, and 200 trajectories were generated.

For the impact of the helicity loss on binder secondary structure composition, the ‘weights_helicity’ flag in BindCraft was set to 1, 0, −0.3, −1, −2 and −3, and 200 trajectories were generated for each instance using otherwise default settings.

To compare the design capabilities of AF2 monomer and multimer weights, we generated 200 trajectories each. For AF2 multimer trajectories, we used the default settings in which AF2 multimer models 1–5 are used for design and AF2 monomer models 1–2 trained with templates are used for reprediction. For AF2 monomer this is inverted, we use AF2 monomer models 1 and 2 for design and AF2 multimer models 1–5 for reprediction.

For benchmarks involving design and trajectory success rates, we run the design pipeline either for 200 trajectories or until 100 designs passing in silico filters are accumulated (where indicated). We then designate trajectories with pLDDT above 0.7 as ‘passing’, whereas trajectories that have a pLDDT below 0.7, more than 1 Cα backbone clash between chains or fewer than 3 contacts between the binder and target are designated ‘low confidence’.

RFdiffusion benchmarks were performed as described in the original publication⁵, with the exception of running the pipeline in deterministic mode for tracking purposes. Briefly, backbones of designated lengths were sampled using RFdiffusion against selected targets and sequences were designed using original ProteinMPNN weights with a temperature of 0.0001 and 8 sequences per backbone. Each complex was predicted using AF2 monomer model 1 and two MPNN designed sequences for each backbone were allowed to pass filters as defined in the original publication (pLDDT > 0.8, i_pAE < 0.32, binder r.m.s.d. < 1.0 Å). The pipeline was run until 100 designs passed filters. The computational time was calculated as backbone generation time + ProteinMPNN sequence generation + AF2 complex prediction for each design. Notably, although single model prediction was used in the case of RFdiffusion, we used prediction using two template-based AF2 models in the case of BindCraft.

Pairwise structural similarities and sequence identities across targets and binders in Supplementary Data 2 were extracted using Foldseek⁵⁵ exhaustive search and TMalign alignment type.

To determine fold and interface novelty of designed binder complexes, we searched the binder chain against the PDB using Foldseek in TMalign mode. Hits with the highest template modelling score (qtmscore) and their sequence identities (fident) for each binder were plotted. Owing to the low resolution structural representations in Foldseek, an alternative strategy was used to assess interface novelty. Residues were extracted using PPIRef in a 6 Å radius around the designed interface, then searched against the precomputed PDB interaction pairs using the iDist method, with a default threshold of 0.04 (ref. ¹⁴). The closest interface hit is then aligned using USalign to calculate the template modelling score and sequence identity¹⁵.

Benchmarking of designs from other design pipelines was performed using the BindCraft prediction method of either AF2 monomer or multimer in single sequence, with templates provided for the target according to the specifications in their respective publications.

AlphaFold3 predictions of designed BindCraft complexes were performed using the AlphaFold3 server⁴⁹ with multiple-sequence alignments and templates enabled.

Pairwise Pearson correlation coefficients (r) among experimental binding (yes, 1, no, 0), Affinity (nanomolar, length and all AF2 and Rosetta-derived features were computed and visualized as a heatmap to assess linear relationships and correlation across all pairs of values. Coefficient values outlined in the cells are considered significant at |r| ≥ 0.7.

Protein expression, purification and characterization

DNA sequences of designed proteins, as well as BBF-14, Der f7, Der f21 and Bet v1 targets were ordered from Twist Biosciences with Gibson cloning adaptors for cloning into bacterial expression vectors pET21b or pET11. Proteins were expressed in Escherichia coli BL21 Codon Plus (DE3) cells (Novagen) by inducing with 0.5 mM isopropyl-β-d-thiogalactoside for 6 h at 18 °C. Pellets were resuspended and lysed in lysis buffer (50 mM Tris-HCl pH 7.5, 500 mM NaCl, 5% glycerol, 1 mg ml⁻¹ lysozyme, 1 mg ml⁻¹ phenylmethylsulfonyl fluoride and 1 µg ml⁻¹ DNase) using sonication. Cell lysates were clarified using ultracentrifugation, loaded on a 1 ml Ni-NTA Superflow column (Qiagen) and washed with 7 column volumes of 50 mM Tris-HCl pH 7.5, 500 mM NaCl and 10 mM imidazole. Proteins were eluted with 10 column volumes of 50 mM Tris-HCl pH 7.5, 500 mM NaCl, 500 mM imidazole. Claudin binders were dialysed against 20 mM HEPES pH 8.0, 150 mM NaCl, 4% glycerol and directly frozen.

The Fc-fused PD-L1 target³, IFNAR2 target, IFNA2 cytokine and antibodies were expressed using a mammalian Expi293 secreted expression system (Thermo Fisher Scientific, A14635). Six days posttransfection, the supernatants were collected, cleared and purified either using a 1 ml Ni-NTA Superflow column (Qiagen) or protein A affinity column (Qiagen). SAS-6 (ref. ²²), SpCas9 (ref. ⁵⁶), CbAgo and the catalytic mutant of CbAgo (D541A, D611A)⁴⁰ have been purified as described previously.

Remaining bacterial and mammalian expressed proteins were then concentrated and injected onto a Superdex 75 16/600 or Superdex 75 10/300 gel filtration column (GE Healthcare) in 50 mM Tris-HCl pH 7.5, 250 mM KCl or PBS. Proteins after size exclusion were concentrated, frozen in liquid nitrogen and stored at −80 °C. Molar mass, sample homogeneity and multimeric state were confirmed using SEC–MALS (miniDAWN TREOS, Wyatt) by injecting 100 µg of protein in PBS (Column, Superdex 75 10/300 or Superdex 200 10/300, GE Healthcare). Folding, secondary structure content and melting temperatures were assessed using circular dichroism in a Chirascan V100 instrument from Applied Photophysics in PBS at a concentration of 0.1–0.3 mg ml⁻¹.

Expression and purification of PD-1 target and binders

DNA sequences were synthesized in the pcDNA3.4 vector with an osteonectin secretion signal at the N terminus (Twist Biosciences). De novo designs were fused to the N terminus of human IgG1 Fc. The extracellular domain (25–167) of human PD-1 (UniProtKB Q15116) was fused to a C-terminal AviTag and His tag. Plasmid DNA was prepared from glycerol stocks (Twist Biosciences) using Cowin Biosciences GoldVac EndoFree plasmid maxi kit. Plasmids were transfected into 3 ml or 50 ml cultures of Expi293F (Gibco) cells as per the manufacturer’s recommendations. Cells incubated at 37 °C for 4–5 days before collection. Following protein expression, the cell culture supernatant was filtered through a 0.22-µM filter and purified using MabSelect protein A affinity chromatography resin (Cytiva). The column was washed with PBS and the protein was eluted in Tris glycine buffer pH 2.5. Following elution, proteins were dialysed into PBS using a 10-kDa molecular weight cut-off dialysis cassette. For production of biotinylated PD-1 protein, the PD-1 plasmid was cotransfected with BirA plasmid (2:1 ratio). The BirA plasmid contains the BirA sequence (UniProtKB P06709) with a C-terminal Flag tag in the pcDNA3.4 vector.

Binding characterization of PD-1

Designs were initially screened for binding to biotinylated human PD-1 or a random protein using BLI (Sartorius OctetRED384). Biotinylated human PD-1 protein and biotinylated lysozyme (GeneTex) were prepared at 500 nM in PBS containing 0.1% bovine serum albumin (BSA) (PBSA). The designs were diluted to 5 µM in PBSA. Streptavidin-labelled biosensors were saturated with either biotinylated human PD-1 or biotinylated chicken lysozyme. The designs were then allowed to associate with the immobilized ligand for 60 s, followed by a dissociation step in PBSA. The baseline subtracted signal (nanometres) was calculated and used to prioritize human PD-1 specific binders for further characterization.

To determine the affinity of selected designs, 100 nM biotinylated human PD-1 prepared in PBSA was immobilized onto a streptavidin-labelled biosensor for 15 s. Serial dilutions of the designs (from 2.5 µM to 5 nM) were then allowed to associate with the immobilized ligand for 180 s, followed by a dissociation step in PBSA for 300 s. Following background subtraction of the BLI binding curves using the buffer only (PBSA) curve, the K_d was determined using the 1:1 model in the Data Analysis HT v.11.1 curve fitting module.

To determine whether the designed protein competed with pembrolizumab for binding to PD-1, 100 nM biotinylated human PD-1 in PBSA was immobilized onto streptavidin coated biosensors for 15 s. An initial association with 200 nM pembrolizumab prepared in PBSA was performed for 180 s, followed by a second association with 200 nM design prepared in PBSA for 180 s.

SPR binding and competition assays

SPR measurements were performed using the Biacore 8 K system (Cytiva) in HBS-EP + buffer (10 mM HEPES pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.005% (v/v) Surfactant P20 GE Healthcare). Target proteins were immobilized on a CM5 chip (GE Healthcare) through amide coupling in 10 mM NaOAc pH 4.5 for 130–250 s at a flow rate of 10 µl min⁻¹ aiming for 100 relative response units. Designed binders or control proteins were injected as analytes in either a single 10 µM concentration during binder prescreening or in serial dilutions to assess binding kinetics. These were injected at a flow rate of 30 µl min⁻¹ for a varying contact time, followed by dissociation. If necessary, the chip surface was regenerated after each injection using 10 mM Glycine-HCl pH 2.5 for 30 s at a flow rate of 30 µl min⁻¹. Binding curves were fitted with a 1:1 Langmuir binding model in the Biacore 8K analysis software. Steady-state response units were plotted against analyte concentration and a sigmoid function was fitted to the experimental data in Python v.3.9 to derive the K_d.

Competition assays were performed as follows. For PD-L1 and IFNAR2, target receptors were immobilized, and binders and competitors were injected as analytes. Two subsequent injections were performed with only competitor (A,1 µM), only design (B,1 µM) or first competitor (1 µM, A) and then design + competitor (both 1 µM, A + B). For Bet v1, REGN5713 (Antibody format) was immobilized on the SPR chip and in a first injection (1) loaded with Bet v1 allergen (1 µM), before either REGN5714 (Fab format) or Birch–binder2 were injected (both 1 µM) (2).

Cell-surface specificity measurements

For specificity measurements, PD-1–b4 was expressed and purified as a His-tagged protein. PD-1-Fc was produced with mutations at glycosylation sites (N → D) and free cysteine residues (C → S). All other proteins were purified as previously described.

BLI experiments were performed using a Gator BLI system and GatorOne software (Gator Bio, v.2.7.3.0728). Assays were conducted in a running buffer containing 10 mM HEPES (pH 7.4), 150 mM NaCl, 3 mM EDTA and 0.005% (v/v) Surfactant P20 (GE Healthcare).

For immobilization, Fc-tagged target proteins (PD-L1, PD-1 and IFNAR2) were diluted to 5 µg ml⁻¹ and captured onto protein A biosensor tips (Gator Bio). After immobilization, the biosensor tips were dipped into 1 µM solutions of purified binder.

Protein crystallization and structure determination

The BBF-14–binder4 complex was crystallized at a concentration of 5 mg ml⁻¹ using sitting drop vapour diffusion at 16 °C in 0.1 M MES pH 6.0, 0.2 M sodium acetate trihydrate, 20% w/v polyethylene glycol (PEG) 8000 buffer (SG1-Eco Screen, Molecular Dimensions). The Der f7–binder2 complex in P21 crystal form was crystallized at a concentration of 15 mg ml⁻¹ using sitting drop vapour diffusion at 16 °C in 0.1 M MES pH 6.5, 0.2 M KSCN, 25% w/v PEG 2000 MME buffer (Clear Strategy Screen I, Molecular Dimensions). The Der f7–binder2 complex in C121 crystal form was crystallized at a concentration of 15 mg ml⁻¹ using sitting drop vapour diffusion at 16 °C in 0.1 M MES pH 6.5 and 20% v/v PEG smear high BCS (BCS Screen, Molecular Dimensions). The Der f21–binder10 complex was crystallized at a concentration of 30 mg ml⁻¹ using sitting drop vapour diffusion at 16 °C in 0.1 M sodium citrate pH 5.6, 1.0 M LiSO₄, 0.5 M NH₄SO₄ buffer (SG1-Eco Screen, Molecular Dimensions). Crystals were cryoprotected in 25% glycerol and flash-cooled in liquid nitrogen. Diffraction data were collected at the European Synchrotron Radiation Facility MASSIF-3 and ID30B beamlines, Grenoble, France at a temperature of 100 K. Crystallographic data were processed using the autoPROC package⁵⁷. Phases were obtained by molecular replacement using Phaser⁵⁸. Atomic model refinement was completed using COOT⁵⁹ and Phenix.refine⁵⁸. The quality of refined models was assessed using MolProbity⁶⁰. Structural figures were generated using ChimeraX⁶¹.

Cryo-EM structure determination

SpCas9 was mixed with a threefold excess of either binder3 or binder10, and the complex was purified using S200 10/300 gel filtration column (GE Healthcare) in 20 mM Tris-HC pH 7.5, 250 mM KCl. The purified complex was applied to a glow discharged 300-mesh holey carbon grid 300-mesh holey carbon grid (Au 1.2/1.3 QuantifoilMicro Tools), blotted for 4 s at 95% humidity, 10 °C, plunge frozen in liquid ethane (Vitrobot Mark IV, FEI) and stored in liquid nitrogen. Data collection was performed on a 300 kV Titan Krios G4 microscope equipped with a FEI Falcon IV detector and SelectrisX energy filter. Micrographs were recorded at a magnification of ×165,000, pixel size of 0.726 Å and a nominal defocus ranging from −0.8 mm to −2.2 mm.

Acquired cryo-EM data were processed using cryoSPARC v.4.6.0 (ref. ⁶²). Micrographs were patch motion corrected, and micrographs with a resolution estimation worse than 5 Å were discarded after patch contrast transfer function estimation. Initial particles were picked using blob picker at 90–135 Å. Particles were extracted with a box size of 360 × 360 pixels, downsampled to 220 × 220 pixels. After two-dimensional classification, clean particles were used for ab initio three-dimensional (3D) reconstruction and initial non-uniform 3D reconstruction⁶³. This model was used for extra template-based picking of particles. Following several rounds of 3D classification, in which classes containing unbound Cas9 were excluded, the class with the most detailed binder features was re-extracted using full box size and subjected to non-uniform and local refinement to generate final reconstructions. The local resolution was calculated and visualized using ChimeraX⁶¹. The in silico models were docked into density using ChimeraX⁶¹.

Birch allergen blocking assay

Anti-Bet v1 binder blocking capacity was assessed by first coating NuncSorp (Thermo Fisher) plates with 2 μg ml⁻¹ of anti-human IgE monoclonal antibody (NBS-C BioScience; clone Le27; 0908-1-010) in coating buffer (15 mM Na₂CO₃, 34.87 mM NaHCO₃) and incubating overnight at 4 °C. The plates were washed with PBS + 0.05% Tween and blocked using PBS + 1% BSA for 2 h at room temperature. Then, sera from patients allergic to birch were added at a concentration of 4 ng ml⁻¹ of anti-Bet v1 IgE. Biotinylated Bet v1 allergen at 1 nM concentration was preincubated for 2 h at room temperature with fourfold serial dilutions of the Bet v1–binder2 starting at 2 μM or with fivefold serial dilutions of the cocktail of REGN5713, REGN5714 and REGN5715 (starting at 50 nM each) and then added to the IgE coated plate. After 2 h of incubation at room temperature, the plates were washed with PBS + 0.05% Tween and streptavidin horseradish peroxidase (BD Pharmigen; 554066; 1:1,000 dilution) was added and incubated for 1 h. Plates were washed and tetramethylbenzidine substrate (BD Biosciences; 555214) was added and incubated for another 20 min. The reaction was stopped with 2 M sulfuric acid. Absorbance was measured on a spectrophotometer at 450 nm with a 630-nm reference, and blocking percentage was measured by subtracting the absorbance of the sample in the absence of the binder.

MST

CLDN1 WT was labelled with Cy5 by adding a 1:5 molar excess of dye and incubating for 2 h on ice. The excess dye was removed by passing through a PD-10 column. The labelled protein was collected and stored in small aliquots at −80 °C after flash freezing in liquid nitrogen.

For MST-based interaction studies, the Monolith (Nanotemper) instrument was used. Serial dilutions of the ligand (CLDN1–b12/CpE–Nd33) were made in buffer B (25 mM HEPES pH 8.0, 200 mM NaCl, 5% glycerol 0.03% DDM) and mixed with 10 nM labelled CLDN1 WT. After 10 min of incubation, samples were transferred to capillaries (Monolith standard capillary) and readings were initiated. The spectral shift data were plotted and fitted into a K_d model, and estimated K_ds were obtained. When data were not fitted using the K_d model, the Hill model was used to fit data. For studying the competitive binding of CpE–Nd33 and CLDN1–b12 to the target CLDN1 WT, a second set of experiments was performed. CLDN1 WT was incubated with CLDN1–b12 (2 × K_d) and subsequently challenged with CpE–Nd33.

Cytotoxicity assay

To study whether claudin binders were able to inhibit pore formation in Sf9 cells expressing claudins, adherent Sf9 cells in a 24-well plate were infected with baculovirus containing either CLDN1 or CLDN4. The assay was performed as shown previously²⁰. Briefly, for each claudin, a 12-well experiment was performed. Six wells were used to test the effect of binders on the pore-forming capacity of CpE–Nd33 and the other six wells were used as controls. After 36 h of infection, 4 µM of each binder were added into six different wells and the plate was then gently mixed by swirling and incubated for 5 min. After that, 300 nM of CpE–Nd33 was added to each of the six wells. The following controls were used in experiment 1. Sf9 without baculovirus infection, 2. Sf9 infected with claudin but not treated with CpE–Nd33, 3. Sf9 infected with Claudin and treated with CpE–Nd33 4. Sf9 infected with Claudin and treated with COP4 Fab (referred to as CpE inhibitor) 5. Sf9 infected with Claudin and treated by COP4 followed by addition of CpE–Nd33 after incubation for 5 min. The number of cells dead or alive were then measured after 18 h of incubation by staining the cells with trypan blue and measuring the number of cells using an automated cell counter (Invitrogen Countess).

SpCas9 gene editing

For SpCas9-single-guide RNA (sgRNA) plasmid cloning, lentiCRISPR v2 (Addgene no. 52961, a gift from F. Zhang) was digested with BsmBI (NEB). Oligonucleotides encoding for the sgRNA targeting the NSD2 gene were annealed and ligated into the digested lentiCRISPR v2 plasmid. All binders were human-codon optimized using the GenSmart Codon Optimization tool and ordered as inserts with homology overhangs for cloning from Twist bioscience. Final binder plasmids were generated by isothermal assembly (NEBuilder HiFi DNA Assembly Cloning Kit, NEB).

HEK293T (ATCC CRL-3216) cells were maintained in DMEM plus GlutaMax (Thermo Fisher Scientific), supplemented with 10% (vol/vol) fetal bovine serum (Sigma-Aldrich) and 1 × penicillin-streptomycin (Thermo Fisher Scientific) at 37 °C and 5% CO₂. Cells were maintained at confluency below 90% and passaged every 2–3 days. For testing inhibitor efficiency, HEK293T cells were seeded in 48-well cell culture plates (Greiner) and transfected at 70% confluency using 300 ng Cas9 + sgRNA plasmid, 500 ng of inhibitor plasmid and 5 µl of Lipofectamine 2000 according to the manufacturer’s instructions (Thermo Fisher Scientific). The next day, cells were split and selected with Puromycin, Blasticidin or both. Three days posttransfection, cells were gathered and genomic DNA was isolated by direct lysis.

The DNA from the cell lysate was prepped for next-generation sequencing as previously described⁶⁴. In the first PCR round, genomic regions of interest were amplified using GoTaq Green Master Mix (Promega) and primers that included Illumina forward and reverse adaptor sequences. A second PCR round, also using GoTaq Green Master Mix (Promega), introduced p5–p7 barcodes into the products from the first round. The resulting amplified amplicons were pooled and quantified using a Qubit 3.0 fluorometer (Invitrogen). The libraries were then sequenced using a MiSeq platform (Illumina, 150 bp, paired-end). Sequencing data and resulting gene editing insertion-deletion rates were analysed using CRISPResso2 (ref. ⁶⁵).

CbAgo in vitro cleavage assay

For in vitro cleavage assays, binders, CbAgo, 5′-phosphorylated 16-nt single-stranded DNA (ssDNA) guide (oDS423) and Cy5-labelled 45-nt ssDNA target (oDS401) were mixed to final concentrations of 2:0.4:0.4:0.2 μM in 10 mM HEPES pH 7.5, 125 mM KCl and 2 mM MgCl₂. To this end, first the binder protein and CbAgo were mixed and incubated at 37 °C for 15 min, after which the mixture was incubated on ice and guide ssDNA and target ssDNA were added. Subsequently, reaction mixtures were incubated at 37 °C, and samples were taken at 0-min, 4-min, 10-min, 30-min and 60-min timepoints. Samples taken at each timepoint were directly quenched by adding 2× RNA loading dye (25 mM EDTA, 5% v/v glycerol, 90% v/v formamide) and heating for 5 min at 95 °C. Cleavage products were resolved using denaturing (7 M urea) 20% polyacrylamide gel electrophoresis, and gels were imaged using a Amersham Typhoon gel scanner (Cytiva Life Sciences). Cleavage reactions were performed in triplicates for each binder protein. CbAgo target cleavage was quantified using ImageQuant TL 1D v.8.2.0 (Cytiva Life Sciences), and fitted with nonlinear least squares fit (nlsLM from R package minpack.lm) to a double-exponential decay model to model initial (fast) and turnover (slow) cleavage:

If fitting to a double-exponential decay model yielded no fit after 1,024 iterations with residuals and gradient convergence tolerance of 1 × 10⁻⁹, the turnover cleavage (slow) was considered negligible and a single-exponential decay model (that is, B = 0) was used.

For all samples K_cat was calculated from the fit constants for the initial rate (A and K₁):

The mean and standard deviation of K_cat was calculated using the three experimental replicates.

CbAgo BLI binding kinetics

BLI measurements were conducted using the Gator BLI system and GatorOne software (Gator Bio, v.2.7.3.0728). The running buffer consisted of 150 mM KCl, 20 mM HEPES (pH 7.5) and 0.5% BSA. His-tagged CbAgo binders were immobilized on the sensor tips at a concentration of 10 µg ml⁻¹. After immobilization, the tips were transferred into serial dilutions of CbAgo. Binding curves were globally fitted using a 1:1 interaction model in the Gator software.

CbAgo SEC binding verification

Purified CbAgo was diluted to 0.8 mg ml⁻¹ (9.3 µM) and mixed with 0.2 mg ml⁻¹ binder protein and 9.3 µM 5′-phosphorylated 16-nt ssDNA guide (oDS423) in SEC buffer (20 mM HEPES pH 7.5, 250 mM KCl and 2 mM MgCl₂). The mixture was incubated for 15 min at room temperature. After incubation samples were resolved at room temperature on a Superdex 200 Increase 10/300 GL column (Cytiva Life Sciences) connected to a 1260 Infinity II high-performance liquid chromatography system (Agilent) using SEC buffer with a flow rate of 0.75 ml min⁻¹. The elution was measured using a Agilent 1260 Infinity II Multiple Wavelength Detector at 280 nm. The data were analysed using Astra v.8.1 (Wyatt Technology).

AAV engineering

HEK293 cells adapted to culture in orbitally shaken bioreactors (HEKExpress, ExcellGene SA) were maintained in Serum-free BalanCD HEK293 Medium (Irvine Scientific) supplemented with l-alanyl-l-glutamine (Gibco GlutaMax) at 37 °C, 80% humidity, 5% CO₂, under constant shaking at 180 rpm (shaking diameter 5 cm). Cells were passaged every 3–4 days to a concentration of 0.2 × 10⁶ cells per ml. For the generation of cell lines stably expressing the target receptors used in the AAV transduction experiments, the receptor complementary DNAs (cDNAs) were obtained from an open reading frame collection (HER2, ORFeome Collaboration cDNA Clone, PD-L1, Addgene no. 121142) and cloned into a pRRLSIN lentiviral shuttle construct (Addgene no. 12252) with expression under the control of the human phosphoglycerate kinase (hPGK) promoter. Lentiviral particles were generated using standard procedures for calcium phosphate transfection of HEK293T cells with the pRRLSIN-hPGK-WPRE, p8.92, pMD2G and pAdVAntage plasmids. At 48 h, the vector-containing supernatant was harvested, filtered and concentrated by ultracentrifugation. The number of lentiviral particles present in the obtained vector suspension was quantified using a p24 antigen ELISA kit (ZeptoMetrix). HEKExpress cells were transduced in a six-well plate at a density of 3.0 × 10⁶ cells per well using a multiplicity of infection (MOI) of 100 vg per cell (conversion factor 1 pg p24 = 1 × 10⁴ vg). After 5 days, the cells were stained for the presence of the respective target receptor using an APC-conjugated antibody (0.8 µg ml⁻¹, BioLegend, 329707 (PD-L1), 324407 (HER2)) in staining buffer (PBS containing 0.5% BSA (Merck)) and sorted by flow cytometry using a Sony SH800 cell sorter. After expansion, the cells were aliquoted and frozen at −80 °C until further use.

The pRepCap plasmids for the AAV production by transient transfection of HEKExpress cells, encoding the rep (AAV2) and cap (varying) genes, were chosen according to the different variants as indicated. For serotype 6 wild-type AAV, an AAV6 plasmid was ordered from the manufacturer (Aldevron, pALD-AAV6). For the variant carrying the knockout mutations to deplete the primary interactions with heparin (K459S and K531E) and sialic acid (V473D, N500E and T502S), a corresponding gene fragment was ordered as an insert with homology overhangs (Twist Biosciences) and cloned into pALD-AAV6 by BspEI/MscI yielding the pRepCap knockout. To introduce the sequence encoding the designed miniprotein binders, an intermediate plasmid was created that, in addition to the five mutations to deplete the primary interactions, carries two silent mutations yielding MluI/NheI restriction sites in proximity to the chosen site of binder insertion between amino acid positions 497 and 498 of the serotype 6 VP3. The DNA sequences for the designed miniprotein binders, flanked by a single –(GSG)₁– linker at both termini, were human-codon optimized using the GenSmart Codon Optimization tool and ordered as inserts with homology overhangs (Twist Biosciences) for subcloning.

For the AAV production for screening at small scale, the cells were seeded in 24-well cell culture plates at a density of 0.4 × 10⁶ cells per ml in a volume of 500 µl and transfected with 520 ng of pHelper (Aldevron, pALD-HELP), 250 ng of shuttle plasmid (Aldevron, pALD-ITR-GFP), 270 ng of pRepCap (varying) and 1.5 µg of polyethyleneimine (Polysciences). If applicable, the variants’ pRepCap plasmids were respectively mixed in a ratio of 1:2 with the pRepCap knockout plasmid. Then 12 h after transfection, the cells’ media was exchanged and supplemented with 4 mM valproic acid (Sigma). The cell culture was incubated at the standard conditions described above but without shaking, and the AAV containing medium was harvested on day 5 by collecting the supernatant using centrifugation at 400g for 5 min at room temperature to remove cells.

For the AAV production for validation at a normalized MOI, the cells were seeded at a density of 1.0 × 10⁶ cells per ml in a volume of 300 ml in a TubeSpin 600 bioreactor tube (TPP) and transfected with 231 µg pHelper (Aldevron, pALD-HELP), 105 µg of shuttle plasmid (Aldevron, pALD-ITR-GFP), 105 ng pRepCap (varying) and 900 µg polyethyleneimine (Polysciences). If applicable, the pRepCap plasmids were obtained as described above, and mixed in a ratio of 1:2 with the pRepCap knockout plasmid. Then 6 h after transfection, the cell culture medium was supplemented with 4 mM valproic acid (Sigma). The cell culture was incubated at the standard conditions described above for 7 days, and the vector was harvested on day 3–4 and on day 7 by collecting the supernatant after centrifugation of the bioreactor tube at 800g for 10 min at room temperature. The supernatant was filtered (Stericup Quick Release, Millipore Express PLUS 0.22 μm PES, 1,000 ml, Merck Millipore). The particles were concentrated from the cell culture supernatant to a concentration of at least 3.0 × 10¹⁰ vg ml⁻¹ by using Amicon Ultra-15 centrifugal filter units at a molecular weight cut off of 100 kDa (Merck). The wild-type and knockout variant particles were processed alternatively according to Gaudry et al.⁶⁶. In short, the particles were purified from the cell culture supernatant using the POROS CaptureSelect AAVX resin (Thermo Fisher Scientific) on an ÄKTA Pure chromatography system followed by buffer exchange to PBS, 0.001% Pluronic F-68 (10% stock solution, Gibco) through Amicon Ultra-15 centrifugal filter units, at a molecular weight cut off of 100 kDa (Merck). The number of genome-containing AAV particles was determined after treatment with DNase I (Thermo Fisher) by digital PCR using the QIAcuity system and PCR kit (Qiagen).

For transduction, the target cells were seeded in 96-well cell culture plates at a density of 3 × 10⁵ cells per ml in a volume of 100 µl. After 6 h, the cells’ media was replaced with 100 µl of AAV containing medium from the production in 24-well cell culture plates, or a 100-µl dilution to 3 × 10¹⁰ vg ml⁻¹ of the material from the production in 300 ml of culture to yield a MOI of 1 × 10⁵ vg per cell. If applicable, 0.8 µg ml⁻¹ target receptor blocking antibody (BioLegend, 329707) was added. After 48 h, the cells were washed twice with 100 µl of PBS containing 0.5% BSA (Merck), and the transduction signal (GFP) was measured by flow cytometry on an Attune NxT analyser (Thermo Fisher) equipped with an automated plate reader. The results were analysed using FlowJo v.10.8 Software (BD Life Sciences).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.