Engraftment of gene-edited hematopoietic stem cells after antibody-drug conjugate conditioning in nonhuman primates

Key Points

• •Antibody-based conditioning for HSC transplantation was directly compared with busulfan chemotherapy in a nonhuman primate model.
• •Targeting CD117 or CD45 with ADC enabled autologous transplant of CCR5-edited HSCs.

Visual Abstract

Abstract

Hematopoietic stem cell (HSC) gene therapies provide lifelong benefit in numerous hematological diseases and disorders, but safety and toxicity remain a critical barrier for routine application. In the setting of immunodeficiency syndromes and infectious diseases such as HIV infection, conditioning regimens may exacerbate immune dysfunction, blunting, or impairing overall efficacy. Here, we conduct a head-to-head comparison of 2 novel antibody-drug conjugates (ADCs) with a pyrrolobenzodiazepine payload for autologous transplant in rhesus macaques. These ADCs target either CD117 or CD45 and are benchmarked against the clinical standard busulfan. We quantified the extent of myeloablation and immunosuppression, time to hematopoietic recovery, long-term engraftment of CCR5 CRISPR–edited autologous HSCs, and resistance to infection when challenged with increasing concentrations of an HIV-like virus. Both ADCs enabled engraftment of CRISPR-edited HSCs, although with lower levels of long-term editing compared with busulfan. We observed myeloablation with similar times to hematopoietic recovery and preserved lymphocyte counts with all 3 conditioning regimens, but neither ADC conditioning nor busulfan enabled sufficient CCR5 editing for viral immunity. Although these results only apply to the specific ADC-conditioning protocols tested here, they are a step toward developing targeted strategies to engraft cells with therapeutic edits and highlight the need for further refinement of antibody-based selection.

Introduction

Safety and toxicity remain concerns for hematopoietic stem cell (HSC) therapies. In the setting of immunodeficiency syndromes and infectious diseases such as HIV infection, conditioning regimens may exacerbate immune dysfunction, undercutting measures of efficacy. We previously demonstrated this in a nonhuman primate (NHP) model of antiretroviral therapy suppressed HIV persistence. In these experiments, myeloablative conditioning with total body irradiation (TBI) potentiated viral reactivation following antiretroviral therapy cessation due to a conditioning-induced loss of immune control.^1,2 Minimizing the side effects of conditioning is necessary to maximize the therapeutic potential of HSCs to improve outcomes and enable broader application of cell therapy.

Our NHP model is well-suited to identify regimens that replace or augment hematopoietic subsets with minimal impact on host immunity. Our group and others have previously shown that myeloablative busulfan is a safe and effective alternative to TBI that supports engraftment of gene-modified HSCs derived from mobilized leukapheresis products without additional immunosuppression.^3,4 Antibody-drug conjugates (ADCs) allow targeted elimination of specific cell types and minimize off-target toxicity. ADCs targeting CD117 (c-kit, stem cell factor receptor) have been shown to preferentially deplete HSCs over other cells,⁵ but potential concerns remain as CD117 is also expressed in nonhematopoietic tissues, such as skin, lungs, and reproductive organs. CD45 is expressed on all leukocytes in addition to HSCs and has also been identified in tissue precursor cells.⁶ ADCs targeting both markers have been shown to ablate HSCs in vivo and support engraftment of gene-modified human HSCs in murine^5,7, 8, 9, 10 and autologous NHP models,^11,12 with minimal toxicity and preservation of fertility and immune function. Here, we characterized the engraftment of CRISPR-edited HSCs in rhesus macaques after conditioning with 1 of 2 ADCs using a pyrrolobenzodiazepine (PBD). The goal of this study was to identify ADC alternatives to busulfan conditioning that maximize engraftment of CRISPR-edited HSCs and redirect cytoreductive approaches while maintaining virus-specific adaptive immunity in our NHP model of HIV infection.

Materials and methods

NHP studies

Healthy juvenile rhesus macaques were housed at the University of Washington National Primate Research Center (WaNPRC) under conditions approved by the American Association for the Accreditation of Laboratory Animal Care. All experimental procedures performed were reviewed and approved by the Institutional Animal Care and Use Committee of the Fred Hutchinson Cancer Center (Fred Hutch) and University of Washington (protocol no. 3235-01). This study was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (The Guide), including at least twice-daily observation by animal technicians for basic husbandry parameters (eg, food intake, activity, stool consistency, and overall appearance), as well as daily observation by a veterinary technician and/or veterinarian. Animals were housed in cages approved by the Guide and in accordance with Animal Welfare Act regulations. Animals were fed twice daily and were fasted for up to 14 hours before sedation. Environmental enrichment included grouping in compound, large-activity, or run-through connected cages; perches; toys; food treats; and foraging activities. If a clinical abnormality was noted by WaNPRC personnel, standard WaNPRC procedures were followed to notify the veterinary staff for evaluation and determination for admission as a clinical case. Animals were sedated with ketamine HCl and/or telazol (tiletamine and zolazepam for injection) and supportive agents for balanced anesthesia (such as diazepam and midazolam) before all procedures. After sedation, animals were monitored according to WaNPRC standard protocols.¹³ WaNPRC surgical support staff are trained and experienced in the administration of anesthetics and have monitoring equipment available for electronic monitoring of heart rate, respiration, and blood oxygenation; audible alarms and digital readouts; monitoring of blood pressure, temperature, etc. For minor procedures, the presence or absence of deep pain was tested by the toe-pinch reflex, and the absence of response (leg flexion) indicated adequate anesthesia. In cases of general anesthesia, similar monitoring parameters were used, and anesthesia was tested by loss of palpebral reflexes (eye blink). Analgesics (generally buprenorphine with meloxicam or buprenorphine slow release) were provided as prescribed by clinical veterinary staff for ≥48 hours after the procedures and could be extended at the discretion of the clinical veterinarian based on clinical signs.

CD34⁺ HSC mobilization and enrichment

NHPs receiving autologous transplant, priming or mobilization, collection of cells, and CRISPR editing were treated to our previously published protocols. Granulocyte colony-stimulating factor (G-CSF) was administered once daily at a dose of 50 μg/kg per day for 5 days to rhesus macaques. On the fourth day, a single dose of 1 mg/kg AMD3100 was administered. A final dose of G-CSF was administered, and apheresis began the following morning, 10 hours after AMD3100.

Before enrichment for CD34⁺ cells, red cells were lysed twice in ammonium chloride lysis buffer. White blood cells were incubated for 25 minutes with a custom-produced 12.8 immunoglobulin M anti-CD34 antibody (Fred Hutch, Seattle, WA), and then washed and incubated for another 25 minutes with magnetic-activated cell-sorting anti-immunoglobulin M microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). The cell suspension was passed through magnetic LS columns (Miltenyi), enriching for CD34⁺ cell fractions with a maximum of 2 × 10⁹ cells per column. Enriched CD34⁺ cells were cultured in StemSpan (STEMCELL Technologies, Vancouver, BC) supplemented with penicillin-streptomycin (100 U/mL; Gibco by Life Technologies, Waltham, MA) and 100 ng/mL each of stem cell factor (PeproTech, Rocky Hill, NJ), thrombopoietin (PeproTech), and FMS-related tyrosine kinase 3 ligand (Miltenyi Biotec) overnight before electroporation. CCR5-edited and nonelectroporated control cells were maintained in vitro and assayed for up to 6 days after editing.

Flow cytometry

Flow cytometric analysis was performed on a FACSymphony (BD Biosciences, Franklin Lakes, NJ). Antibodies used for analysis of NHP HSCs included CD34 PE-CF594 and CD34 APC clone 563, CD90 PE-Cy7 clone 5E10, CD45 V450 clone D058-1283, and CD45RA APC-H7 clone 5H9, all from BD Biosciences. Dead cells and debris were excluded by forward/side scatter gating.

The following antibodies were used to assess posttransplantation hematopoietic recovery by flow cytometry at least weekly: CD20 APC clone 2H7 (BD), CD4 AF700 clone L200 (BD), CD16 APC-H7 clone 3G8 (BD), CD11b PE clone ICRF44 (BioLegend), CD14 PE-Cy7 (BD), CD45 V450 clone D058-1283 (BD), CD8 BV605 clone RPA-T8 (BioLegend), and CD3 BV786 clone SP34-2 (BD).

CRISPR-Cas9 editing of CD34⁺ cells

Purified Cas9 protein was acquired from Life Technologies (catalog no. A36499), and guide RNAs (gRNAs) were custom-ordered from Synthego (Redwood City, CA; sequence: UCAUCCUCCUGACAAUCGAU). Lyophilized gRNAs were resuspended in nuclease-free water at a concentration of 100 pmol/μL and stored as frozen aliquots at −80°C. Enriched CD34⁺ cells were cultured overnight after enrichment. CRISPR-Cas9 ribonucleoproteins were formed by combining 180 pmol Cas9 protein with 900 pmol gRNA for 10 minutes and used for the electroporation of 9- to 10- million cells per cuvette in a Lonza 4D-Nucleofector X Unit following the manufacturer’s instructions.

Cryopreservation

Enriched CD34⁺ cells were cultured overnight after electroporation and cryopreserved at a density of 10 million cells per mL in CryoStor CS10 (STEMCELL). Cells were cooled to −80°C overnight in a Mr Frosty Freezing Container (Thermo Fisher Scientific, Waltham, MA) and stored at −80°C for up to 2 weeks until infusion.

Colony forming cell assays

A total of 1000 to 1200 enriched cells were seeded into 3.5 mL ColonyGEL 1402 (ReachBio, Seattle, WA) in 35-mm petri dishes. Hematopoietic colonies were scored after 12- to 14- days. Arising colonies were identified as colony-forming unit (CFU) granulocyte, CFU macrophage, CFU granulocyte-macrophage, and burst-forming unit-erythrocyte. Colonies consisting of erythroid and myeloid cells were scored as CFU-GEMM (granulocyte, erythrocyte, monocyte, megakaryocyte).¹³

Quantifying indels with MiSeq

Editing at the CCR5 locus was measured by next-generation sequencing using Illumina-barcoded 2 × 150 base paired-end MiSeq primers for complete sequencing; forward: TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGTGGCCATCTCTGACCTGCTT; reverse: GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGACCCCAAAGGTGACTGTCCT. Library primers used were Nextera XT Index 15055293 FC-131-1001, and bioinformatic processing of sequencing data was conducted using a custom-built pipeline.¹⁴

Busulfan dosing and cell infusions

Busulfan was administered in 4 IV doses of 5.5 mg/kg per day (22 mg/kg total) commencing 3- to 10- days after mobilized apheresis. Cryopreserved cells were warmed to 37°C and infused into conditioned animals no sooner than 24 hours after the final dose of busulfan.

ADC dosing, pharmacokinetics, and cell infusions

CD117-ADC and CD45-ADC were both acquired from Magenta Therapeutics (Cambridge, MA). Dosing and washout periods were chosen based on previous pharmacokinetic studies using ADC conditioning with PBD payloads for HSC transplantation (HCT).^11,12 Animals were given a single IV injection of CD117-ADC (0.3 or 0.4 mg/kg) or CD45-ADC (0.2 mg/kg), commencing 3- to 20- days after mobilized apheresis. Plasma concentrations of ADC were measured by enzyme-linked immunosorbent assay for A21043 (CD117-ADC, 0.3 mg/kg) and A22062 (CD45-ADC, 0.2 mg/kg) as previously described.^11,12 Cryopreserved cells were warmed to 37°C and infused into conditioned animals 6 to 10 days after treatment with ADC. Details for individual animals are provided in Table 1.

Table 1.

CRISPR-Cas9–edited HSC infusions in rhesus macaques

Animal ID	Conditioning	Apheresis date	Conditioning start date	Cell infusion date	Weight (kg)	CCR5 editing ex vivo (%)	Infused CD34+ cells per kg	Infused CD90+ cells per kg	Reference
A21032	Busulfan 5.5 mg/kg per day ×4	13 December 2021	17 December 2021	21 December 2021	7.8	93	1.2 × 107	3.7 × 106	3
A21042	Busulfan 5.5 mg/kg per day ×4	7 February 2022	12 February 2022	16 February 2022	6.7	94	5.9 × 106	2.6 × 106
A21039∗	Busulfan 5.5 mg/kg per day ×4	28 February 2022	3 March 2022	7 March 2022	5.5	90	5.4 × 106	3.1 × 106
A21040	Busulfan 5.5 mg/kg per day ×4	11 July 2022	21 July 2022	25 July 2022	6.0	90	3.6 × 106	2.2 × 106
Average					6.5	92	6.5 × 106	3.3 × 106
A21041	CD117-ADC 0.3 mg/kg	24 January 2022	26 January 2022	1 February 2022	6.8	92	3.3 × 106	1.9 × 106
A21037	CD117-ADC 0.4 mg/kg	22 February 2022	25 February 2022	7 March 2022	6.7	90	4.3 × 106	1.7 × 106
A21038∗	CD117-ADC 0.4 mg/kg	11 April 2022	15 April 2022	25 April 2022	6.5	76†	1.1 × 106	8.3 × 105
A21043	CD117-ADC 0.3 mg/kg	13 September 2022	4 October 2022	10 October 2022	6.0	93	5.3 × 106	3.0 × 106
Average					6.5	85	3.5 × 106	1.9 × 106
A22062	CD45-ADC 0.2 mg/kg	10 October 2022	1 November 2022	7 November 2022	6.6	88	2.4 × 106	1.2 × 106
A22086	CD45-ADC 0.2 mg/kg	30 January 2023	21 February 2023	27 February 2023	5.4	91	5.7 × 106	2.3 × 106
A22053	CD45-ADC 0.2 mg/kg	21 February 2023	14 March 2023	20 March 2023	6.1	90	2.4 × 106	3.0 × 105
Average					6.0	90	3.5 × 106	1.6 × 106
A20131	No transplant	N/A	N/A	N/A	5.2	N/A	N/A	N/A
A22061	No transplant	N/A	N/A	N/A	8.3	N/A	N/A	N/A
A21044	No transplant	N/A	N/A	N/A	4.9	N/A	N/A	N/A

N/A, not applicable.

^∗Euthanized after transplant.

^†Measured at day 2 after electroporation. Other samples were collected on day 6.

Cell counts

Complete blood counts were performed to determine postconditioning numbers of total white blood cells, platelets, neutrophils, and lymphocytes. Cell counts for CD34⁺ enrichment and infusion products were determined using the Countess II Automated Cell Counter (Thermo Fisher) with trypan blue staining. CD34⁺ cell purity was confirmed by flow cytometry.

SHIV challenge and plasma viral load assays

After HCT and recovery, animals underwent mucosal challenge to quantify resistance to a chimeric Simian-HIV (SHIV) chimera encoding a transmitted/founder clade C HIV-1 Env adapted for increased replication fitness in rhesus macaques (SHIV.C.CH505.v2^15,16). Increasing quantities of titered SHIV.C.CH505.v2 inoculum stock (149 ng of p27 per mL) were administered intrarectally every other week, and viral load was measured in cell-free plasma by reverse transcription-polymerase chain reaction as described previously.² Challenges 1 to 3 administered 1 μL inoculum, challenges 4 and 5 administered 5 μL, and challenge 6 administered 20 μL. SHIV plasma viral loads remained negative in 1 animal (ID A22086) through 1 week after challenge 6. The following week (2 weeks after challenge 6), A22086 received challenge 7, consisting of 1 mL inoculum administered IV. However, plasma viral load data showed the first SHIV-positive reading on the day of challenge 7 (2 weeks after challenge 6). Our data therefore indicate that this animal was infected after 6 intrarectal SHIV challenges.

Statistics

Statistical analyses were performed in GraphPad Prism 10. Assays of CCR5 editing in varying hematopoietic lineages used a 2-way analysis of variance with Dunnett correction for multiple comparisons. Comparisons of editing after SHIV infection used a 1-way analysis of variance with Dunnett multiple comparisons.

Results

Collection, editing, and transplant of autologous HSCs

Fourteen rhesus macaques were enrolled in this study (Table 1). Animals were mobilized with a regimen of G-CSF and plerixafor for 5 days.^3,17, 18, 19 CD34⁺ HSCs were manually enriched from leukapheresis products, electroporated with a CRISPR ribonucleoprotein targeting CCR5, and cryopreserved as previously reported.³ After manufacturing and cryopreservation of the autologous HSC products, animals were conditioned for transplantation. They received either a myeloablative busulfan dose of 5.5 mg/kg per day for 4 days (Figure 1A) or a single IV injection of ADC targeting either CD117 or CD45 carrying a PBD payload.^8,10, 11, 12 ADC conditioning included a washout period before cell infusion to ablate resident CD34⁺ HSCs in bone marrow without targeting the subsequently infused CCR5-edited HSCs (Figure 1B).

Figure 1.

Hematopoietic nadirs and recovery after HCT with busulfan or ADC conditioning. (A) Timeline of mobilization and conditioning for transplants of hematopoietic stem and progenitor cells (HSPCs) with busulfan or (B) ADC conditioning. (C-E) Cell count nadirs in peripheral blood after conditioning and HSPC infusion. (F-G) Times to neutrophil and platelet recovery after transplant. ∗P < .05. †One busulfan and 1 CD45-ADC animal did not reach neutropenia. Two busulfan and 1 CD45-ADC animal did not reach thrombocytopenia. ‡One busulfan and 1 CD117-ADC animal did not recover from thrombocytopenia.

Dosing and timing of ADC conditioning were based on prior pharmacokinetic studies.^11,12 CD117-ADC animals received a dose of either 0.3 mg/kg with a 6-day washout before cell infusion or 0.4 mg/kg with a 10-day washout. CD45-ADC animals received 0.2 mg/kg ADC with a 6-day washout period in the first animal (A22062). The washout period was extended to 10 days for A22086 and A22053. Consistent with pharmacokinetic analyses, plasma ADC concentration was below the lower limit of quantification within 48 hours for CD117-ADC and within 6 hours for CD45-ADC (supplemental Figure 1). Busulfan-conditioned animals received a dose of 5.5 mg/kg per day for 4 days. Details on cell counts, conditioning, and transplantation times are listed in Table 1.

The mean cell infusion was 4.6 × 10⁶ cryopreserved CD34⁺ cells per kg, with 10 of 11 animals exceeding 2.0 × 10⁶ cells per kg. The mean number of cryopreserved CD34⁺CD90⁺CD45RA⁻ long-term engrafting HSCs cells was 2.3 × 10⁶ cells per kg, with all transplantations exceeding the 1.1 × 10⁵ cells per kg threshold as previously observed by our laboratory²⁰ (Table 1).

CD117-ADC enables myeloablation without lymphodepletion, but with off-target effects

Three of the 4 busulfan-conditioned animals achieved myeloablation, defined as a neutrophil nadir <500 cells per μL in peripheral blood. Both doses of CD117-ADC proved myeloablative, with near-complete depletion of neutrophils (Figure 1C; supplemental Figure 2) and significant decreases in postconditioning platelet nadirs compared with busulfan (Figure 1D; supplemental Figure 3). After conditioning, animals transplanted with all 3 regimens showed minimal lymphocyte toxicity and no lymphopenia (defined as lymphocyte counts <500/μL in peripheral blood), with no significant differences in lymphocyte count nadirs even after CD45-targeted conditioning (Figure 1E; supplemental Figure 4). As previously reported,³ we euthanized 1 busulfan-conditioned animal (A21039) 90 days after transplant due to sustained thrombocytopenia despite transfusions of platelet-rich plasma. One CD117-ADC–conditioned animal (A21038) was euthanized 11 days after transplantation with severe thrombocytopenia and cerebral hemorrhage. Neutrophil recovery did not differ significantly between conditioning regimens (Figure 1F). Time to platelet recovery was similarly variable, with 2 busulfan-conditioned animals and 1 CD45-ADC–conditioned animal experiencing no thrombocytopenia without transfusion support (Figure 1G; supplemental Figure 3). We observed no significant signs of liver or kidney toxicity by serum chemistry (supplemental Figure 5). CD4⁺ and CD8⁺ T cells were detected in peripheral blood in the first week after transplant in all CD117-ADC–conditioned animals, which are likely endogenous and not transplant derived²⁰ (Figure 2).

Figure 2.

Multilineage preservation and recovery after conditioning with both CD117-ADC and CD45-ADC for autologous HCT. Numbers are presented as percentages of CD45⁺ white blood cells after conditioning and cell infusion. Myeloid cells: CD11b⁺; granulocytes: CD11b⁺CD14⁻SSA^hi; monocytes: CD11b⁺CD14⁺; natural killer cells. cells: CD11b⁻CD16⁺; B cells: CD11b⁻CD20⁺; CD4⁺ T cells: CD11b⁻CD3⁺CD4⁺CD8⁻; CD8⁺ T cells: CD11b⁻CD3⁺CD4⁻CD8⁺.

Unexpectedly, we observed a dose-dependent loss of pigmentation in the fur of all 3 CD117-ADC–conditioned animals (Figure 3). Depigmentation persisted >15 months after CD117-ADC administration in animals that received either the high dose (A21037) or the low dose (A21041), suggesting long-term toxicity to melanocytes not seen with other conditioning regimens.

Figure 3.

Off-target effects of CD117-ADC on melanocytes. Depigmentation in animals >1 year after conditioning with CD117-ADC. The greatest loss of pigmentation was observed in the animal that received the highest dose of ADC.

CD45-ADC enables engraftment with minimal myeloablation and preserved T cells

Likewise, CD45-ADC dosed at 0.2 mg/kg enabled engraftment of HSCs with similar hematopoietic recovery as busulfan (Figure 1). As CD45 is expressed by all white blood cells, it is notable that the dose used here did not fully ablate lymphocytes, with both CD4⁺ and CD8⁺ T cells detectable in peripheral blood samples taken from all animals during the first 2 weeks after transplantation (Figure 2). Of note, 1 animal (A22086) successfully engrafted with no neutropenia (Figure 1E; supplemental Figure 2) or thrombocytopenia (Figure 1F; supplemental Figure 3).

CCR5-edited progeny persists long-term in multiple hematopoietic lineages

HSCs were enriched from a single-mobilized apheresis procedure, with a final CD34⁺ cell dose ranging from 1.2 × 10⁷ (A21032) to 1.1 × 10⁶ (A21038; cells per kg (Figure 4A). We observed similar degrees of CCR5 editing in each HSC product at the time of cryopreservation, 1 day after nucleofection (Figure 4B), and similar colony-forming potential of edited HSCs (supplemental Figure 6). As previously observed,^3,21, 22, 23, 24 measured editing continued to increase in a subset of cells cultured in vitro for up to 6 days (Figure 4C). Cryopreserved infusion products were thawed and administered after busulfan or ADC conditioning. We monitored CCR5 editing in the peripheral blood of each animal for a minimum of 4 months (A22053) to a maximum of 18 months (A21032) after transplantation. All 3 conditioning regimens supported long-term engraftment of edited HSCs, measured by CCR5 editing in peripheral white blood cells (Figure 4D-F). Busulfan supported the highest levels of CCR5-edited HSC engraftment, with CCR5 editing frequencies averaging 33% in total leukocytes 3 months after transplantation (Figure 4G). Both CD117-ADC and CD45-ADC conditioning regimens supported engraftment of edited cells but to a significantly lower degree than busulfan. To quantify editing of distinct hematopoietic lineages, we purified peripheral blood for markers of lymphocytes (T cells: CD3, CD4, and CD8; B cells: CD20) and myeloid subsets (monocytes: CD14). We observed CCR5 editing in all subsets, with significantly lower levels of editing in T-cell lineages after busulfan conditioning (Figure 4H). Despite the broad range of cell doses administered, we saw no correlation between the number of CD34⁺ or CD34⁺CD90⁺ cells infused and CCR5 editing in vivo, either within or across conditioning groups (supplemental Figure 7).

Figure 4.

Gene editing in peripheral blood after busulfan- or ADC-conditioned HCT. (A) Total counts of CD34⁺ and CD34⁺CD90⁺ cells cryopreserved and transplanted. (B) CCR5 editing of HSCs collected at time of cryopreservation. (C) CCR5 editing of infusion product HSCs cultured in vitro up to 6 days. (D-F) CRISPR-mediated CCR5 editing in total white blood cells following autologous transplant. (G) Comparison of gene editing in white blood cells 3 months after infusion. (H) CCR5 editing frequencies in multiple hematopoietic lineages 4 to 18 months after HSC infusion. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001. ns, not significant.

Overall, we found that busulfan, CD117-ADC, and CD45-ADC all induced myeloablation and engraftment of autologous CCR5–edited HSCs in rhesus macaques without lymphopenia and with persistence of CD4⁺ and CD8⁺ T cells. However, busulfan conditioning proved most effective at enabling long-term engraftment of edited cells and their progeny.

No difference in SHIV infection rates between ADC and busulfan conditioning

We are interested in exploring alternative HSC conditioning regimens that maximize HSC engraftment while minimizing immunosuppression relative to TBI.² To compare the impact of busulfan, CD117-ADC, and CD45-ADC conditioning on susceptibility to an HIV-like virus in our NHP model, we next challenged animals that received transplants and 3 control animals that did not receive transplants with SHIV.C.CH505.v2. We performed ≤6 dose-escalating mucosal challenges via the intrarectal route, followed by a seventh IV challenge, if necessary (Figure 5A). In addition to comparing the impact of each conditioning regimen, this experiment allowed us to further explore threshold CCR5 editing levels necessary for protection against de novo SHIV infection.^1,25 Consistent with our previous results,³ we found that SHIV acquisition was comparable between animals that did not receive transplants and animals that were conditioned with busulfan, whereas CD117-ADC and CD45-ADC conditioning showed trends of earlier and later SHIV acquisition, respectively (Figure 5B). CCR5 editing levels of ≤40% in peripheral blood granulocytes and 25% in CD3⁺ T cells (Figure 4H) were insufficient to impact SHIV acquisition relative to controls that did not receive transplants (Figure 5B). We found no correlation between conditioning regimen or level of CCR5 editing and control of plasma viral load after infection (supplemental Figure 8). To test if ongoing SHIV replication selects for CCR5-edited cells as observed in TBI-conditioned animals,¹ we repeated the analysis of CCR5 editing in lymphoid and myeloid lineages at a time point 4 months after SHIV infection in each animal (Figure 5C). Consistent with the prediction that SHIV specifically depletes CD4⁺ cells and that CCR5-edited cells should be protected, the statistically significant difference in editing difference between CD4⁺ cells and granulocytes vanished after infection in animals that were conditioned with busulfan, with no change in editing of other lineages including CD8⁺ cells. Although our model of autologous CCR5-edited HCT failed to achieve a protective threshold against SHIV infection, this observation suggests a protective effect of transplant-derived CCR5-edited progeny and indicates that higher levels of bulk CCR5 editing could inhibit mucosal SHIV acquisition.

Figure 5.

Rates of acquisition and CCR5 editing after SHIV infection. (A) Schematic of intrarectal challenge with SHIV.C.CH505.v2 following autologous HCT. (B) Kaplan-Meier survival plot of time to infection after SHIV challenge. (C) CCR5 editing frequencies in multiple hematopoietic lineages 6 months after initiation of SHIV challenge. ∗P < .05, ∗∗P < .01. IR, intrarectal.

Discussion

We comprehensively compared 2 ADCs with a PBD payload as alternative conditioning regimens to busulfan, the current standard in autologous HCT and gene therapy. Busulfan is commonly used in combination with the immunosuppressant fludarabine, especially for allogeneic HCT. Although lymphodepletion may have a role in autologous HCT, especially in situations involving surface-expressing transgenes such as a T-cell receptor or chimeric antigen receptor, we did not use additional lymphodepletion for this study. This was to minimize the effect of conditioning on the immune response to SHIV and was chosen because fludarabine has been shown to have no immunosuppressive effect on rhesus macaques.²⁶ Both CD117-ADC and CD45-ADC led to myeloablation without lymphodepletion, with similar times to neutrophil and platelet recovery after transplantion as seen with busulfan conditioning. We observed long-term persistence of CCR5-edited HSCs and progeny in vivo following all 3 conditioning regimens, with busulfan enabling the highest levels of CCR5 indels after engraftment. With a maximum of 40% CCR5 editing in peripheral white blood cells, none of the assessed transplantation regimens was protective against a dose-escalating mucosal challenge with an HIV-like virus.

The NHP is an immunocompetent, clinical-scale model of HCT with the ability to conduct long-term follow-up. The kinetics of hematopoietic recovery after HSC engraftment are similar to those in humans and are well characterized after TBI conditioning^20,27,28 and after busulfan, which enables myeloablation and autologous HCT in NHPs without immunosuppression or allogeneic HCT with additional immunosuppressive agents.^3,4,11,28 In recent years, the field has advanced from editing HSCs derived from bone marrow^29,30 toward a more translational model using HSCs enriched from mobilized peripheral blood.^31,32 Our range of 28% to 43% CCR5 editing in peripheral blood leukocytes after busulfan conditioning compares favorably with published results of gene-edited HCT using Cas9 or an adenine base editor (Abe8) editing the CD33 and HBG loci with more intense TBI conditioning.29, 30, 31, 32, 33, 34 Demirci et al^12,35 have shown that even higher editing efficiencies can be achieved at the BC11A locus.

This NHP model can provide critical insights on the safety and specificity of antibody-based conditioning in hematopoietic and nonhematopoietic compartments. For example, CD117 is expressed on numerous nonhematopoietic subsets, including germ cells.^36,37 Notably, CD117-ADC–based conditioning has been reported to support engraftment of lentivirus-transduced HSCs without inducing infertility, as seen with busulfan,¹¹ and CD117-targeting conditioning regimens for HCT are currently in clinical trials.^38,39 Here, we observed that CD117-ADC conditioning in rhesus macaques transiently suppressed peripheral neutrophil and platelet counts yet spared the lymphocyte compartment. Neutrophil and platelet nadirs were comparable between conditioning regimens, but engraftment of edited cells was reduced, with CCR5 editing in peripheral blood ranging from 4% to 9% of alleles at the end point vs 29% to 37% after busulfan. This is similar to the observation by Uchida et al¹¹ of lentiviral vector copy numbers of 0.01 to 0.1 after transplantation, which were also reduced compared with busulfan controls. Although these results are unique to autologous HCT, we postulate that in the allogeneic setting CD117-ADCs may be able to replace myeloablative busulfan when combined with additional lymphodepletion. Unexpectedly, we observed an effect on melanocytes, with multiple animals losing pigmentation (Figure 3) that persisted to the end of the study. Similar effects have been seen in mouse models using naked monoclonal antibodies against CD117,⁴⁰ but as this was not reported in previous NHP studies of this same ADC,¹¹ we did not quantify the extent or kinetics of the change. Although a mild side effect, this finding shows that on-target but non-HSC effects of CD117-based conditioning remain a concern.

CD45 targeting remains a promising avenue for treating HIV with stem cell modification. CD45 is broadly and exclusively expressed by hematopoietic cells, effectively trading specificity to HSCs for the elimination of on-target effects beyond the hematopoietic compartment. This would additionally enable simultaneous depletion of HSCs and lymphocytes that comprise the HIV reservoir. Both CD117-ADC and CD45-ADC conditioning supported similar levels of engraftment in this study. Unexpectedly, some CD45-targeted animals engrafted with reduced or no neutropenia, and none of the 3 animals in this study displayed postconditioning lymphopenia. It is important to reiterate that these findings are specific to the dose and timing tested here. Demirci et al¹² have also recently shown that higher doses of this same CD45-ADC enable long-term engraftment of CRISPR-edited HSCs at even higher levels of editing, with subsequent elevation of fetal hemoglobin and preserving CD3⁺ T cells during conditioning. Thus, a carefully titrated CD45 ablation may allow therapeutic levels of HSC editing with similar or fewer complications than myeloablative busulfan. CD45’s ubiquity on leukocytes also makes it an appealing candidate for hematological malignancies, and we hypothesize that the same agent at higher doses may target malignant cells and enable HCT with or without additional lymphodepletion.

It should be acknowledged that post-engraftment CRISPR editing is dependent at least partially on the number of cells transplanted.²⁰ In this study, animals were assigned to conditioning cohorts prior to mobilization and HSC enrichment, resulting in the busulfan cohort receiving a larger average cell dose. However, editing in engrafted cells did not match the lowest busulfan-conditioned animal (A21040) in either the CD117-conditioned animal (A21043) or the CD45-conditioned animal (A22053), which received the highest cell doses. Additionally, the differences in cell doses were not statistically significant. Optimal times for HCT were predicted to be 6- to 10- days after administration of ADC,¹¹ but it is possible that this difference in efficacy between busulfan and ADC conditioning could be due to toxicity to the transplanted cells from residual ADC in the blood and bone marrow. Lee et al³² found that HSCs that have undergone electroporation and CRISPR editing, especially via the homology-directed repair pathway, engraft at a lower rate after TBI conditioning compared with HSCs that had been lentivirally transduced. Although this study is not a direct comparison of methods for gene modification, it is possible that unedited and residual endogenous HSCs remain healthier after ADC conditioning when compared with busulfan or other more intense conditioning regimens, allowing them to more readily outcompete the gene-edited and heavily manipulated transplant product.

Neither ADC significantly impacted mucosal virus acquisition in our HIV model. These results, paired with the persistence of peripheral blood T cells after transplantation, suggest that targeted conditioning regimens can support HSC engraftment without complete lymphodepletion. Nevertheless, even the highest level of CCR5 editing in this study failed to reach the threshold necessary to protect against SHIV infection, which is expected to exceed 90%.^25,41 Although CCR5 editing alone has been shown to be protective in humanized mouse models, the high degree of editing required makes it likely that further mechanisms for selecting edited cells in vivo will be necessary for therapeutic autologous CCR5 knockout, along with approaches for depleting the viral reservoir. Recent successes with the “Geneva patient” (HIV cure without CCR5Δ32 transplant) and the “next Berlin patient” (HIV cure with a CCR5Δ32 heterozygous donor) support this view.⁴² We are currently investigating the ability of this CD45-ADC approach to clear latent viral reservoirs in our model of persistent HIV infection, and the ability of CCR5-edited HCT to synergize with treatment strategies using broadly neutralizing antibodies. These results are a promising step toward developing targeted strategies to engraft cells with therapeutic edits and antibody-based selection without the need for alkylating chemotherapies, these approaches are also relevant for malignancies and hemoglobinopathies.

Conflict-of-interest disclosure: H.-P.K. has received support as the inaugural recipient of the José Carreras/E. Donnall Thomas Endowed Chair for Cancer Research and the Stephanus Family Endowed Chair for Cell and Gene Therapy and is or was a consultant to and has or had ownership interests with Rocket Pharmaceuticals, Homology Medicines, Vor Biopharma, and Ensoma Inc. S.R. is a consultant to Forty Seven Inc (Gilead Sciences) and Ensoma Inc. K.J.B. has served on Data and Safety Monitoring Boards or Scientific Advisory Boards for ViiV and AbbVie. The remaining authors declare no competing financial interests.