Recombinant AAV2 (rAAV2) and AAV9 (rAAV9) empty particles (cat. nos. P230508-1022fcd and P230508-1021rnj) and recombinant AAV2 and AAV9 vectors containing a luciferase reporter gene (cat. nos. P230802-1009dka and P230802-1013nav) were from VectorBuilder GmbH, Neu-Isenburg (Germany). Biotinylated rAAV2 and rAAV9 capsids were prepared in-house. Biotinylated mouse and unlabeled humanized recombinant anti-AAV2 (A20/H) and anti-AAV9 (4.5.136/1.30.34) antibodies were produced in-house. Horseradish peroxidase (HRP)-conjugated goat anti-human IgG (cat. No. 109–035–098) were from Jackson ImmunoResearch Europe Ltd., Ely (UK). Streptavidin (SA)-beads (cat. no. 28985799) were from Cytiva Europe GmbH, Freiburg (Germany), Low Cross Buffer (cat. no. 100500) was from Candor Biosciences GmbH, Wangen (Germany), and ONE-Glo™ EX Luciferase Assay System (cat. no. E8130) was from Promega, Walldorf (Germany). Poly-D-lysine plates (cat. no. 655944) were from Greiner Bio-One GmbH, Frickenhausen (Germany) and SA-coated microtiter plates were from Microcoat Biotechnologie GmbH, Bernried (Germany). Phosphate buffered saline (PBS), Dulbecco’s modified Eagle’s medium (DMEM), fetal calf serum (FCS) and other cell culture reagents as well as 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) were obtained from commercial suppliers. Serum for quality control (QC) samples preparation was obtained from Trina Bioreactives AG, Nänikon (Switzerland). Human embryonic kidney cells (HEK293T) were from Roche cell bank.

The sets of serum samples were obtained in two nonclinical development studies from gene therapy-naïve cynomolgus monkeys. The studies adhered to current ethical standards and applicable local and international regulations on animal experimentation.

Suspension of SA-Beads (200 μL) was washed 3 times with 500 μL PBS, mixed with 50 μL of biotinylated rAAV2 capsids (10 μg/mL [experiment 1 with 30 samples] or 20 μg/mL [experiment 2 with 70 samples]) and incubated at 800 rpm for 1 h. The prepared AAV2-capsid–beads were washed 3 times with 500 μL PBS and immediately used for sample depletion.

A serum sample was diluted with cell culture medium (DMEM/L-glutamine) to a final serum concentration of 2% (v/v), 250 μL of the diluted sample were mixed with the AAV2-capsid–beads prepared in the previous step and incubated for 2 h. Thereafter, the beads were magnetically separated and the depleted serum was recovered and stored frozen until analysis.

All procedures were performed at room temperature using a KingFisher Flex automated extraction instrument (Thermo Fisher Scientific). Incubation and washing steps were performed with shaking at > 700 rpm (or as specified) to keep the beads in suspension.

HEK293T cells were seeded at a density of 2 × 104 cells/well on 96-well poly-D-lysin plates and incubated for 24 h at 37°C and 5% CO2 in the cell culture medium supplemented with 10% FCS. Serum samples (diluted with the FCS-free cell culture medium to 2% v/v serum) were spiked with the AAV2 vector carrying a luciferase reporter gene (1:1, leading to a final serum concentration of 1% v/v), incubated at 300 rpm and room temperature for 60 min and then added to the cells (5 × 107 vg/well). The cells were incubated for another 24 h in the FCS-free cell culture medium and the luminescence signal generated by the produced luciferase (relative light units [RLU]) was measured using ONE-Glo™ EX Luciferase Assay System and a luminescence reader.

Samples depleted of AAV2 capsid-binding components were prepared individually for each analyzed sample (individual maximal transduction control; IMC) and measured in parallel. All measurements were conducted in duplicates and reported as mean normalized ratios to the IMC signal or % IMC signal. A sample was considered positive if its signal was ≤ 50% of the corresponding IMC signal (i.e., mean normalized ratio ≤ 0.5, Supplementary Fig.  1A). Humanized anti-AAV2 antibody (A20H) was used as a positive control. In assay qualification experiments, pooled serum from selected seronegative individuals was used as a maximal transduction control and the measurements were reported as mean RLU and mean % inhibition relative to the maximal transduction control.

The assay was performed as described previously (14). Briefly, serum samples were diluted with Low Cross Buffer (assay buffer) up to a final serum concentration 1% v/v, spiked with rAAV2 (2.5 × 1010 vp/mL) or buffer, incubated at room temperature for 30 min and then mixed with a capture antibody (biotinylated mouse anti-AAV2 antibody [A20], 0.5 µg/mL) and incubated again for 30 min at room temperature. The incubated samples were transferred onto a SA-coated plate, the plate was incubated at room temperature for 15 min, washed with assay buffer and the captured anti-AAV2 IgG–rAAV2 immune complexes were detected using HRP-labeled goat anti-human IgG (0.5 µg/mL incubated for 1 h) and ABTS substrate. Humanized anti-AAV2 antibody (A20H) was used as a positive control.

For each sample, a difference between measurements with and without capsid spike was calculated and reported. Samples were measured in duplicates (experiment 1) or singlicates (experiment 2). A sample was considered positive if its signal difference was ≥ 1.5 times the mean blank signal (assay buffer measured with spiked capsids; averaged across all plates) (Supplementary Fig. 1B).

Assays for anti-AAV9 detection were performed as described above with the following modifications. In the TI assay, cells were seeded at a density of 1 × 104 cells/well, serum samples were diluted 1:10 (resulting in a final serum concentration 5% v/v after AAV9 vector spike), the AAV9 vector was spiked at 5 × 108 vg/well and the incubation time with the vector was 48 h. AAV9-capsid–beads for IMCs preparation were generated using biotinylated rAAV9-capsids at a concentration of 40 μg/mL. In the TAb assay, samples were diluted to a final serum concentration of 2% v/v and rAAV9 capsids were spiked at 2 × 1011 vp/mL. Biotinylated mouse anti-AAV9 IgG (4.5.136) was used as a capture antibody. Humanized anti-AAV9 IgG (1.30.34) and mouse anti-AAV9 IgG (4.5.136) were used as positive control antibodies in the TAb and TI assays, respectively. Measurements were performed in singlicates.