Prolonged viral suppression with anti-HIV-1 antibody therapy – Nature

Study design

We conducted a phase 1b, open-label, randomized clinical trial of two treatment groups of chronically infected people living with HIV-1 (http://www.clinicaltrials. gov; NCT03526848). Study participants were enrolled sequentially according to eligibility criteria and randomized to either group 1 or group 2 in a 3:1 ratio. The planned sample size was 40 participants; however, owing to the COVID-19 pandemic, enrolment was stopped early with 26 participants, with 18 enrolled in group 1 and 8 in group 2. Participants received seven infusions of 3BNC117 and 10-1074 intravenously at a dose of 30 mg per kg body weight of each antibody, at weeks 0, 2 and 4, 8, 12, 16 and 20 unless viral rebound occurred. Viral rebound was defined as plasma HIV-1 RNA level > 200 copies per ml in two consecutive measurements. Group 1 discontinued ART 2 days after the first antibody infusions, whereas group 2 remained on ART before discontinuing the medication 26 weeks after the first antibody infusions. During the treatment interruption periods of the study, ART was resumed if there was a confirmed >30% decline in CD4⁺ T cell count or confirmed CD4⁺ T cell count decrease to < 350 cells per μl from baseline (day 0). Plasma HIV-1 viral RNA levels were monitored weekly through study week 38 and virologic criteria for ART re-initiation accommodated different phases of the follow-up period: (part A) week 0 to week 26, 2 consecutive plasma HIV-1 RNA levels > 200 copies per ml measured weekly, (part B) week 26 to week 38, sustained (>4 weeks) HIV-1 RNA levels > 1,000 copies per ml measured weekly; and (part C) week 38 to week 48, 2 consecutive plasma HIV-1 RNA levels > 1,000 copies per ml measured every other week. Other ART restart criteria included pregnancy and symptoms consistent with acute retroviral syndrome. Study participants were followed for 48 weeks after the first infusions. The study hypotheses were that the administration of repeated infusions of 3BNC117 and 10-1074 in the absence of ART would be safe and well tolerated, maintain viral suppression in people living with HIV during ATI and interfere with the HIV-1 reservoir. The primary endpoints of the study were: (1) rate of viral rebound 12 weeks after ART interruption (week 12 for group 1, and week 38 for group 2); (2) time to return of viraemia (first plasma HIV-1 RNA level > 200 copies per ml of 2 consecutive measurements) after ART interruption; (3) size of the proviral HIV-1 reservoir before and after 7 antibody infusions; and (4) rate and severity of adverse events and serious adverse events. All participants provided written informed consent before participation in the study and the study was conducted in accordance with good clinical practice. The protocol was approved by the US Food and Drug Administration and the Rockefeller University and Mass General Brigham Human Research Institutional Review Boards (IRBs).

Study participants

Study participants were recruited at the Rockefeller University Hospital, New York, USA, and the Massachusetts General Hospital, Boston, USA. Eligible participants were adults aged 18–65 years, HIV-1-infected, on ART with plasma HIV-1 RNA levels of <50 copies per ml for at least 12 months (one viral blip of >50 but <500 copies per ml during this period was allowed), plasma HIV-1 RNA levels < 20 copies per ml at screening visit, reported or confirmed CD4⁺ T cell count nadir of > 200 cells per μl and a current CD4⁺ T cell counts > 500 cells per μl. Participants on an ART regimen that included a NNRTI were switched to an integrase inhibitor-based regimen (dolutegravir plus tenofovir disoproxil fumarate and emtricitabine) at least four weeks before treatment interruption due to the prolonged half-life of NNRTIs. Exclusion criteria included medical history of resistance to two or more classes of antiretroviral medication, prior anti-HIV-1 monoclonal antibody therapy, concomitant hepatitis B or C infection, clinically relevant physical findings, medical conditions or laboratory abnormalities and pregnancy or lactation. Time to viral rebound was defined as the first of two consecutive viral loads of >200 copies per ml with the exception of participant 5104 who experienced two viral blips (two viral loads of <500 copies per ml) which were subsequently followed by suppressed viraemia before rebounding with sustained viraemia at week 33.

As a result of the COVID-19 pandemic, one participant in group 1 (5122M) re-initiated ART despite not having experienced viral rebound and one participant in group 2 (5216) decided against analytical treatment interruption after completion of all seven antibody infusions. Participant 5126M was withdrawn after the first antibody infusion because his NNRTI-based ART regimen was inadvertently not switched according to protocol and he was found to have a viral load of 615 copies per ml at baseline (day 0). In addition, two participants (5207 and 5219) opted to withdraw from the study for COVID-19-related and other personal reasons.

Peripheral blood mononuclear cell (PBMC) samples from four out of seven study participants in a parallel NIH/NIAID-sponsored clinical trial who received 3BNC117 in combination with 10-1074 during ART interruption (http://www.clinicaltrials.gov; NCT03571204) were included for reservoir assessments (Sneller et al., submitted manuscript). This trial included participants that initiated ART within 90 days of being diagnosed with acute or early HIV infection. Enrolled participants discontinued ART 3 days after the first antibody infusions and received a total of 8 3BNC117 and 10-1074 infusions over 24 weeks. Reservoir assessments were performed at baseline and week 24. The 4 participants were selected for analysis because they had measurable reservoirs and maintained viral suppression while receiving all 8 infusions of 3BNC117 and 10-1074 (30 mg per kg body weight of each antibody) over a period of 24 weeks. One additional participant who maintained viral suppression over 24 weeks did not have a measurable reservoir by Q4PCR and was excluded from the analysis. The protocol was approved by the Institutional Review Board of the National Institute of Allergy and Infectious Diseases, National Institutes of Health (Sneller et al., submitted manuscript).

Ten additional individuals on suppressive ART and with similar baseline clinical characteristics were followed over time under a parallel observational protocol for repeated blood donations and reservoir assessments while remaining on suppressive ART (ART-alone group; Extended Data Fig. 1). Clinical data collection and management were carried out using the software iRIS by iMedRIS Version 11.02.

Study procedures

3BNC117 and 10-1074 were prepared according to the Rockefeller University Pharmacy and MGH Pharmacy Standard Operating Procedures and administered intravenously at a dose of 30 mg kg⁻¹. Monoclonal antibody intravenous infusions were administered sequentially, each antibody over 60 min. Study participants were observed at the Rockefeller University Hospital or the Massachusetts General Hospital for 1 h after the last antibody infusion.

Participants returned for weekly (part A and part B) and bi-weekly (part C) follow-up visits during the ATI period for safety assessments, which included physical examination and measurements of clinical laboratory parameters such as haematology, chemistries, urinalysis and pregnancy tests (for women). Viral load was monitored on a weekly (weeks 1 to 38) or bi-weekly (weeks 38–48) basis and CD4⁺ T cell counts were monitored every 1–2 weeks during analytical treatment interruption and 2-4 weeks after ART re-initiation. During every study visit, participants were counselled of the importance to use barrier protection while off ART and until viral load became undetectable after ART re-start to decrease the risk of HIV transmission to a sexual partner. Two participants acquired HCV during study follow-up (one probable sexual transmission after ART had been resumed but prior to viral suppression and one linked to drug use during ATI) (Supplementary Table 3). The protocol was modified to include monthly testing for bacterial sexually transmitted infections while participants remained off ART and until viral suppression after ART restart. Study investigators evaluated and graded adverse events according to the Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events version 2.1 (July 2017) and determined causality.

Leukapheresis was performed at the Rockefeller University Hospital or at the Massachusetts General Hospital at week −2 and week 26. Blood samples were collected before and at multiple times after 3BNC117 and 10-1074 infusions. Samples were processed within 4 h of collection, and serum and plasma samples were stored at −80 °C. PBMCs were isolated by density gradient centrifugation. The absolute number of PBMCs was determined using an automated cell counter (Vi-Cell XR; Beckman Coulter) or manually, and cells were cryopreserved in fetal bovine serum plus 10% DMSO before storage in liquid nitrogen.

Plasma HIV-1 RNA levels

HIV-1 RNA levels in plasma were measured at the time of screening, at week −2, day 0 (before infusion), weekly during ATI, and every two weeks to every eight weeks after viral rebound had occurred. HIV-1 RNA levels were determined using the Roche COBAS AmpliPrep/COBAS TaqMan HIV-1 Assay (version 2.0) or the Roche cobas HIV-1 quantitative nucleic acid test (cobas 6800), which quantitate HIV-1 RNA over a range of 2 × 10¹ to 1 × 10⁷ copies per ml. These assays were performed at LabCorp or at the local laboratory at Massachusetts General Hospital.

CD4⁺ T cells

CD4⁺ T cell counts were determined by a clinical flow cytometry assay, performed at LabCorp or at the local laboratory at Massachusetts General Hospital.

Measurement of 3BNC117 and 10-1074 serum levels

Serum concentrations of active 3BNC117 and 10-1074 were determined before and at the end of each 10-1074 infusion, every 4 weeks during the ATI period, every 4 to 8 weeks in the follow-up period after ART re-initiation, and at the time of viral rebound by a validated luciferase-based neutralization assay in TZM-bl cells as previously described¹⁸. In brief, serum samples were tested using a primary 1:20 dilution with a fivefold titration series against HIV-1 Env pseudoviruses Q842.d12 and X2088_c9, which are highly sensitive to neutralization by 3BNC117 and 10-1074, respectively, while fully resistant against the other administered antibody. In the case of the post-infusion time points of 10-1074, instances where the serum inhibitory dose (ID₅₀) titres against X2088_c9 were >100,000, serum samples were also tested against a less sensitive strain, Du422. Similarly, in the case of the post-infusion time points of 3BNC117, instances where serum ID₅₀ titres against Q842.d12 were >100,000, serum samples were also tested against a less sensitive strain, Q461.e2 (Supplementary Tables 4 and 5). To generate standard curves, 3BNC117 and 10-1074 clinical drug products were included in every assay set-up using a primary concentration of 10 μg ml⁻¹ with a fivefold titration series. Serum concentrations of 3BNC117 and 10-1074 for each sample were calculated as follows: serum ID₅₀ titre (dilution) × 3BNC117 IC₅₀ or 10-1074 IC₅₀ titre (μg ml⁻¹) = serum concentration of 3BNC117 or 10-1074 (μg ml⁻¹). Env pseudoviruses were produced using an ART-resistant backbone vector that reduces background inhibitory activity of antiretroviral drugs if present in the serum sample (SG3ΔEnv/K101P.Q148H.Y181C; M. S. Seaman, unpublished data). Virus pseudotyped with the envelope protein of murine leukemia virus (MLV) was used as a negative control. Antibody concentrations were calculated using the serum ID₈₀ titre and monoclonal antibody IC₈₀ if non-specific activity against MLV was detected. The lower limits of detection were 0.007 µg ml⁻¹ and 0.005 µg ml⁻¹ for 3BNC117 and 10-1074, respectively. All assays were performed in a laboratory meeting GCLP standards.

Antiretroviral drug screening

Plasma samples were screened for a panel of antiretrovirals (tenofovir (TFV), emtricitabine (FTC), lamivudine (3TC), abacavir (ABC), zidovudine (ZDV), nevirapine (NVP), amprenavir (APV), atazanavir (ATV), darunavir (DRV), etravirine (ETR), elvitegravir (EVG), efavirenz (EFV), lopinavir (LPV), ritonavir (RTV), raltegravir (RTG), dolutegravir (DTG), maraviroc (MVC) and rilpivirine (RPV)) by the University of North Carolina Center for AIDS Research Clinical Pharmacology and Analytical Chemistry Core. Results were reported as ‘peak’ or ‘no peak’. No peak results correspond to a non-detectable antiretroviral concentration in the sample of >1 ng ml⁻¹ for TFV, FTC, 3TC, ABC, ZDV and NVP, or >10 ng ml⁻¹ for APV, ATV, DRV, ETR, EVG, EFV, LPV, RTV, RTG, DTG, MVC and RPV. Samples from participants who continued to maintain viral suppression for >12 weeks (group 1) or >12 weeks from last dose (group 2) were tested at approximately 12-week intervals until viral rebound. An additional sample from participant 5120 collected approximately 24 months after ART interruption was also tested.

Phenotypic characterization and surface staining

Frozen PBMCs were thawed and washed with PBS. Cells were then surface-stained for 30 min in the dark at 4 °C with viability reagent (BD Horizon Fixable Viability Stain; BD Biosciences) and a 13-colour cocktail of monoclonal antibodies containing surface antibodies against CD19 (clone SJ25C1, dilution 1/250), CD20 (clone 2H7, dilution 1/100), CD66b (clone G10F5, dilution 1/250), CD14 (clone M5E2, dilution 1/150), CD3 (clone SK7, dilution 1/100), CD4 (clone OKT4, dilution 1/200), CD8 (clone RPA-T8, dilution 1/500), CD38 (clone HIT2, dilution 1/250), CD69 (clone FN50, dilution 1/500), CD71 (clone M-A712, dilution 1/100), CD25 (clone 2A3, dilution 1/300), CD152 (clone BNI3, dilution 1/250), CD279 (clone EH12.1, dilution 1/500), TIGIT (clone 741182, dilution 1/500), CD223 (clone T47-530, dilution 1/250) and CD336 (clone 7D3, dilution 1/500). After labelling, cells were washed and fixed in PBS containing 2% paraformaldehyde and stored at 4 °C before flow cytometry acquisition within 24 h.

Flow cytometry analysis

All events (∼1,200,000–1,800,000 events per sample) were collected on a BD FACSymphony A5 Cell Analyzer (BD Biosciences). The lymphocytes were gated for further analysis, as described in Extended Data Fig. 2 using FlowJo version 9.9.6 (BD Biosciences).

PhenoSense Monoclonal Antibody Assay

The PhenoSense Monoclonal Antibody Assay (Labcorp-Monogram Biosciences) is a phenotypic cell-based infectivity assay capable of assessing the susceptibility of pseudovirions bearing plasma or PMBC-derived HIV-1 envelope proteins to anti-envelope monoclonal antibodies. Populations of full-length envelope sequences, amplified from either plasma derived HIV RNA from viremic individuals or cell-associated HIV DNA from aviremic individuals, were cloned into an envelope expression vector. Envelope expression vector populations and a HIV genomic reporter vector, with env replaced by luciferase, were cotransfected into producer cells to generate luciferase reporter pseudovirions. Pseudovirions were then were tested for neutralization sensitivity to 3BNC117 or 10-1074 in an in vitro cell-based assay.

Clinical cut-off values for neutralization sensitivity have not been established for the PhenoSense Monoclonal Antibody Assay. In this study, the following exploratory values were used to define monoclonal antibody sensitivity: IC₉₀ < 1 µg ml⁻¹ and a maximum per cent inhibition equal to or > 98%.

Single-genome amplification of reservoir env genes

Genomic DNA (gDNA) was isolated using Puregene Cell and Tissue Kit (Qiagen, catalogue no. 158388) according to the manufacturer’s instructions. DNA concentration was measured using Nanodrop. HIV-1 envelope was amplified from gDNA by single-genome amplification (SGA). To achieve a Poisson distribution, in which ≤30% of wells yield a PCR product, gDNA was endpoint diluted in 384-well plates. The first-round PCR was performed at 94 °C for 2 min; 94 °C for 15 s, 58.5 °C for 30 s, and 68 °C for 3 min × 35; and 68 °C for 15 min. One microlitre of the PCR product from first-round PCR was used as a template for second-round nested PCR at 94 °C for 2 min; 94 °C for 15 s, 61 °C for 30 s, and 68 °C for 3 min × 45; and 68 °C for 15 min. PCR2 products were checked using 1% 96-well E-Gels (Invitrogen). Bands with the expected size of the HIV-1 envelope were subjected to library preparation and sequencing using the Illumina Miseq platform (Control Software v3.1.0.13). All PCRs were performed using Platinum Taq Polymerase. Primers used in the first round were envB5out 5′–TAGAGCCCTGGAAGCATCCAGGAAG–3′ and envB3out 5′–TTGCTACTTGTGATTGCTCCATGT–3′. In cases where there was no amplification in the first round, 3′ half-genome primers were used instead (B3F3 5′-TGGAAAGGTGAAGGGGCAGTAGTAATAC-3′ and R3B6R 5′-TGAAGCACTCAAGGCAAGCTTTATTGAGGC-3′). The second-round nested primers used were envB5in 5′–TTAGGCATCTCCzTATGGCAGGAAGAAG–3′ and envB3in 5′–GTCTCGAGATACTGCTCCCACCC–3′.

Q4PCR

Q4PCR was performed as previously described^21,39. In brief, genomic DNA from 1 million to 5 million total CD4⁺ T cells was isolated using the Gentra Puregene cell kit (Qiagen) or phenol-chloroform, and the DNA concentration was measured using a Qubit high-sensitivity kit (Thermo Fisher Scientific). Next, an outer PCR (NFL1) was performed on genomic DNA at a single-copy dilution using outer PCR primers BLOuterF (5′-AAATCTCTAGCAGTGGCGCCCGAACAG-3′) and BLOuterR³⁷ (5′-TGAGGGATCTCTAGTTACCAGAGTC-3′). Undiluted 1-µl aliquots of the NFL1 PCR product were subjected to a Q4PCR reaction using a combination of four primer–probe sets that target conserved regions in the HIV-1 genome. Each primer–probe set consists of a forward and reverse primer pair as well as a fluorescently labelled internal hydrolysis probe as previously described²¹ as follows: PS forward, 5′-TCTCTCGACGCAGGACTC-3′; reverse, 5′-TCTAGCCTCCGCTAGTCAAA-3′; probe, 5′-/Cy5/TTTGGCGTA/ TAO/CTCACCAGTCGCC-3′/IAbRQSp (Integrated DNA Technologies); env forward, 5′-AGTGGTGCAGAGAGAAAAAAGAGC-3′; reverse, 5′-GTCTGGCCTGTACCGTCAGC-3′; probe, 5′-/VIC/CCTTGGGTTCTTGGGA-3′/MGB (Thermo Fisher Scientific); gag forward, 5′-ATGTTTTCAGCATTATCAGAAGGA-3′; reverse, 5′- TGCTTGATGTCCCCCCACT-3′; probe, 5′-/6-FAM/CCACCCC- AC/ZEN/AAGATTTAAACACCATGCTAA-3′/ IABkFQ (Integrated DNA Technologies); and pol forward, 5′-GCA CTTTAAATTTTCCCATTAGTCCTA-3′; reverse, 5′-CAAATTTCTACTAATGCTTTTATTTTTTC-3′; probe, 5′-/NED/AAGCCAGGAATGGA-TGGCC-3′/ MGB (Thermo Fisher Scientific). Each Q4PCR reaction was performed in a 10-µl total reaction volume containing 5 µl TaqMan universal PCR master mix containing Rox (catalogue no. 4304437; Applied Biosystems), 1 µl diluted genomic DNA, nuclease-free water, and the following primer and probe concentrations: PS, 675 nM forward and reverse primers with 187.5 nM PS internal probe; env, 90 nM forward and reverse primers with 25 nM env internal probe; gag, 337.5 nM forward and reverse primers with 93.75 nM gag internal probe; and pol, 675 nM forward and reverse primers with 187.5 nM pol internal probe. Quantitative PCR (qPCR) conditions were 94 °C for 10 min, 40 cycles of 94 °C for 15 s, and 60 °C for 60 s. All qPCRs were performed in a 384-well plate format using the Applied Biosystem QuantStudio 6 or 7 Flex real-time PCR system. qPCR data analysis was performed as previously described²¹ using ThermoFisher Design and Analysis Software 2.4.3. Generally, samples showing reactivity with two or more of the four qPCR probes were selected for a nested PCR (NFL2) with the exception of participant 5108 for whom samples showing reactivity with one or more of the four qPCR probes were selected for NFL2. This focused approach detects the subset of defective proviruses that are positive for at least 2 of the 4 oligonucleotide probes which in turn leads to an underestimation of the absolute number of defective proviruses. The NFL2 reaction was performed on undiluted 1-µl aliquots of the NFL1 PCR product. Reactions were performed in a 20-µl reaction volume using Platinum Taq high-fidelity polymerase (Thermo Fisher Scientific) and PCR primers 275F³⁷ (5′-ACAGGGACCTGAAAGCGAAAG-3′) and 280R (5′-CTAGTTACCAGAGTCACACAACAGACG-3′) at a concentration of 800 nM. Library preparation and sequencing were performed as previously described²¹. Sampling time points without recovery of intact sequences were determined at lower limit of detection for statistical analyses. Limit of detection is calculated as half of the proviral frequency assuming one intact sequence out of the highest input of sampled cells.

Single-genome amplification of plasma rebound virus env genes

Sequencing of HIV-1 plasma rebound env genes was performed as previously described^12,51.

HIV-1 genome reconstruction

HIV-1 genome reconstruction was performed using our in-house pipeline, Defective and Intact HIV genome Assembler (DIHIVA), a pipeline for the assembly of raw sequencing reads into annotated HIV genomes, capable of reconstructing thousands of genomes within hours. First, it removes PCR amplification and performs error correction using clumpify.sh from BBtools package v38.72 (http://sourceforge.net/projects/bbmap). After, a quality-control check is carried out by Trim Galore package v0.6.4 (https://github.com/FelixKrueger/TrimGalore) to trim Illumina adapters and low-quality bases. The pipeline also uses bbduk.sh from BBtools package to remove possible contaminant reads using HIV sequences from the Los Alamos HIV database (https://www.hiv.lanl.gov) as a positive control. A k-mer-based assembler, SPAdes v3.13.0⁵² is used to assemble the HIV-1 sequences. The longest assembled contig is aligned via BLAST to HXB2 HIV-1 sequence to set the correct orientation and transfer annotations. Sequences with double-peaks, that is, regions indicating the presence of two or more viruses in the sample (cut-off consensus identity for any residue < 70%), or samples with a limited number of reads (≤500 sequencing reads) are omitted from downstream analyses. Finally, reconstructed sequences are classified as intact or defective proviruses. We divided defective proviral sequences into five subcategories: (1) non-functional, that is, complete, full-length genome however with one or multiple early stop codons in at least one coding genes; (2) major splice donor mutation mutation, that is, complete, full-length, in-frame genome with a mutation in the 5′ donor splice site 1 (D1); (3) missing internal genes, that is, truncated provirus, with large internal deletions spanning one or more of the 9 HIV coding genes; (4) structural variation, that is, genomes containing indels, duplications or inversions; (5) LTR defect, that is, genomes missing either 3′ LTR or 5′ LTR and PSI region

Post hoc model for 10-1074 sensitivity testing

Reconstructed env sequences were analysed for the presence of computationally predicted 10-1074 escape mutations⁵³. The env sequences were classified as sensitive when obeying all of the following amino acid rules (using HXB2 numbering): 332: N; 333: not P; 334: S or T; 325: D, N, or T and 330: H or Y. The performance of the model was tested against a 10-1074 neutralization panel that included 557 strains (n = 371 10-1074-sensitive (IC₅₀ = 2.0 µg ml⁻¹), n = 186 10-1074-resistant (IC₅₀ ≥ 2.0 µg ml⁻¹)). Of 393 strains predicted to be sensitive, 366 were shown to be true sensitive and 27 were false sensitive. Of 164 strains predicted to be resistant 159 were shown to be true resistant and 5 were false resistant resulting in a sensitivity of 99% and specificity of 85%.

Phylogenetic analysis

Nucleotide alignments of env sequences were translation aligned using MAFFT v7.487 under the BLOSUM62 cost matrix. Maximum likelihood phylogenetic trees were then generated from these alignments with PhyML v3.1 using the general time-reversible (GTR) model with 1,000 bootstraps.

Statistical analysis

Rates of viral rebound 12 weeks after ART interruption in group 1 and group 2 were calculated. Differences in time to viral rebound after ART interruption and after last antibody infusions between group 1 and a shorter antibody combination therapy study¹² were determined using Kaplan–Meier curve comparison and log-rank (Mantel–Cox) tests. Differences in time to viral rebound between group 1 and group 2 were also determined. Differences in time since diagnosis and time on uninterrupted ART between bNAb therapy and ART-alone group were determined using two-tailed Mann–Whitney U-test. Differences in time to viral rebound between participants with phenotypically resistant, partially resistant or sensitive reservoir viruses was determined using one-way ANOVA with Tukey multiple comparison test. Differences in log normalized frequencies of intact and defective proviral genomes per 10⁶ CD4⁺ T cells were determined using paired Student’s t-test. Relative change in proviral frequencies was determined using two-tailed Wilcoxon matched-pairs signed-rank test for matched comparisons within bNAb therapy and ART-alone group. Pharmacokinetic parameters were estimated by performing a non-compartmental analysis using Phoenix WinNonlin Build 8.3 (Certara), using all available pharmacokinetic data starting with the time point after the last infusion of 10-1074 from the TZM-bl assay. These and all other statistical tests as indicated in the manuscript and figure legends were performed using GraphPad Prism 9.1.

Data presentation

Figures were arranged in Adobe Illustrator 2021.

Reporting summary

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