Durable contraception in the female domestic cat using viral-vectored delivery of a feline anti-Müllerian hormone transgene

The study addresses the need for safe, efficient, and permanent non-surgical contraception for domestic cats, a species with a large free-roaming population impacting welfare and wildlife. Anti-Müllerian hormone (AMH) is a TGFβ family ligand critical in sexual differentiation and is expressed in granulosa cells in the feline ovary. Prior work in mice showed that supraphysiological AMH delivered via adeno-associated virus (AAV) can suppress folliculogenesis and induce permanent contraception by inhibiting primordial follicle activation and maturation of pre-antral follicles. Existing non-surgical strategies (contraceptive vaccines, sex steroid analogs, GnRH modulators, barrier devices) have not provided effective long-term contraception in cats and none are FDA/EMA-approved. With advances in AAV gene therapy and favorable AAV9 tropism and low pre-existing anti-AAV9 antibodies in cats, the authors hypothesize that a single-dose AAV9 vector delivering a feline AMH transgene can produce durable, safe contraception in female domestic cats while maintaining estrous cycling and sex steroid production.

The paper reviews prior contraceptive strategies for wildlife and companion animals, including vaccines, sex steroid analogs, GnRH agonists/antagonists, and barrier methods, noting limited long-term efficacy and lack of regulatory approval for cats. It references advances in AAV vector engineering for human gene therapy that open opportunities for animal contraception, and prior mouse studies demonstrating that AAV-mediated AMH overexpression suppresses folliculogenesis and induces permanent contraception through early, gonadotropin-independent blockade of follicle development. The authors also note AAV9’s muscle tropism and low neutralizing antibody prevalence in cats, which supports vector selection.

Transgene design and mouse validation: A first-generation feline AMH transgene (fcMISv1) was created using cat genome v8.0 with carnivora consensus sequence filling gaps; a second-generation transgene (fcMISv2) was later constructed using the complete cat genome v9.0, correcting 31 amino acid differences. Both were codon-optimized and cloned for expression and AAV packaging. Recombinant FLAG-tagged proteins were purified and bioactivity confirmed via fetal rat urogenital ridge bioassay (Grade 4 regression at 5 µg/ml). In nude mice, AAV9-fcMISv2 (5e12 or 1e13 vg/kg, IP) produced AMH expression in multiple tissues, elevated serum AMH, ovarian size reduction, and significant decreases in primary, secondary, and antral follicles after 1 month.

Pilot study in cats (fcMISv1): Three adult females received AAV9-fcMISv1 (5e12 vg/kg, IM). Safety monitoring included physical exams, bloodwork, CRP, and long-term follow-up (≈4 years). AMH and anti-fcMISv1 antibodies were measured by ELISA. At 42 months post-treatment, ovariohysterectomy enabled histology of reproductive tracts; in situ RNA hybridization assessed transgene expression in muscle and liver post-mortem in one cat.

Main cat study (fcMISv2): Nine sexually mature, nulliparous female cats were randomized into three groups (n=3 each): control (AAV9-empty, 5e12 vp/kg), low-dose (AAV9-fcMISv2, 5e12 vg/kg), and high-dose (AAV9-fcMISv2, 1e13 vg/kg). A single IM injection into the right caudal thigh was administered on Day 0. Health monitoring included daily welfare checks post-injection, injection-site exams (daily for 14 days, weekly for 2 weeks, monthly thereafter), physical exams and CBC/chemistry (pre-treatment, Day 0, every 3 months in Year 1, every 6 months thereafter). Safety endpoints included renal (BUN, creatinine, SDMA), hepatic (ALT, AST), and skeletal muscle (CK) markers.

Viral shedding: Quantitative PCR measured AAV genomes in blood (Days 0, 2, 7, 14, 21, 28, monthly to Month 6), urine and feces (daily during BSL-2 housing), and oral swabs (Days 0, 2, 7, 14).

Immunogenicity and AMH levels: A direct antigen-antibody capture ELISA was developed using purified FLAG-fcMISv2 to quantify anti-fcMISv2 IgG. Serum AMH concentrations were measured monthly for 2 years using AMH Gen II ELISA with appropriate dilutions; newborn kitten serum (if any) within 24 h of birth was to be sampled for AMH.

Reproductive hormones and cycling: Fecal samples were collected three non-consecutive days per week from 6 months pre-treatment through 24 months post-treatment. Fecal estradiol (E2) and progesterone (P4) metabolites were quantified by EIAs. Serum measurements included inhibin A, inhibin B, follistatin, testosterone, C-type natriuretic peptide (CNP), and luteinizing hormone (LH; validated EIA). Estrous phases were inferred from fecal E2 (≥2 consecutive samples >1.5 SD over baseline mean). Luteal phases were defined by P4 (≥6 consecutive samples >1.5× baseline mean). Analyses compared a 2-month transition period post-injection (0–2 months) to a post-treatment period (2–24 months), excluding mating/pregnancy/lactation.

Mating trials: Two 4-month group-housed mating trials began at ~8 and ~20 months post-treatment, each with a different proven breeder male. Females and male were co-housed 8 h/day, 5 days/week, with continuous video monitoring. Breeding interactions were scored as breeding attempts or successful breedings (confirmed intromission with appropriate female responses). Weekly transabdominal ultrasonography assessed pregnancy. Pregnant controls were removed for maternity care. Outcomes included breeding activity, luteal phase occurrence following breeding, pregnancies, and number of kittens.

Histology: In the pilot fcMISv1 cats, reproductive tracts removed at 42 months were processed for H&E histology to assess uterine and ovarian pathology. In situ RNA hybridization probed for fcMISv1 expression in skeletal muscle and liver.

• Vector expression and safety: AAV9-fcMISv2 IM delivery achieved robust, sustained supraphysiological serum AMH that declined over the first year and then plateaued above a target 0.25 µg/ml through year 2. No clinically relevant adverse effects were observed; one high-dose cat had transient mild edema at the injection site (Days 3–4). Serial CBC/chemistry revealed no sustained renal, hepatic, or muscle abnormalities.
• Viral shedding: AAV genomes were readily detected in blood by Day 2 and remained elevated for ~2 months before declining by 3–4 months. Urine shedding was detectable acutely by Day 1 and decreased over 7 days. Oral swab and fecal shedding were variable across individuals and groups.
• Immunogenicity: Unlike the fcMISv1 pilot (2/3 cats developed anti-transgene antibodies), no anti-fcMISv2 IgG above background was detected in any treated cat over 24 months.
• Hormones and cycling: Inhibin B did not significantly decline post-treatment compared to the transition period in treated cats. Fecal E2 metabolites increased significantly from pre- to post-treatment in controls and high-dose groups (P<0.0001 for both), but not in the low-dose group. Serum LH levels were higher in the post-treatment vs transition period in both low- and high-dose groups (P=0.0040 and 0.0076, respectively), with no change in controls. Mean fecal P4 concentrations decreased significantly post-treatment in both treated groups (P<0.0001), with no change in controls. Estrous phase frequency (based on E2) did not differ pre- vs post-treatment in any group. Luteal phase frequency was significantly reduced in the high-dose group post-treatment (P=0.0263), but not in low-dose or controls. CNP increased in controls post-treatment (P=0.045) but was unchanged in treated cats; serum follistatin decreased in controls post-treatment (P<0.0001). Testosterone did not decline in treated cats; inhibin A and follistatin were similar across transition vs post-treatment in high-dose cats.
• Contraceptive efficacy: In two 4-month mating trials at ~8 and ~20 months post-treatment, all control females conceived after their first breeding bout and produced healthy litters (Trial 1: 10 kittens; Trial 2: 11 kittens). No AAV9-fcMISv2-treated females became pregnant in either trial; ultrasound showed no gestational sacs/fetuses. Some treated females allowed breeding bouts, but no luteal phases followed any bouts. Chi-square tests showed significant differences between groups for number of luteal phases following breeding (P=0.0498), number of pregnant females (P=0.0498), and total kittens produced (P<0.0001) in both trials.
• Pilot fcMISv1 observations: One cat without anti-fcMISv1 antibodies maintained high AMH and had normal endometrium; two cats with anti-fcMISv1 antibodies had cystic endometrial hyperplasia, and one showed multiple corpora lutea indicating spontaneous ovulation. In situ RNA hybridization confirmed persistent muscle expression of fcMISv1 at 49.5 months post-treatment.
• Mouse validation: AAV9-fcMISv2 produced high serum AMH and significantly reduced primary, secondary, and antral follicles in mice, consistent with ovarian suppression.

The findings demonstrate that a single IM dose of AAV9-fcMISv2 induces sustained supraphysiological AMH, leading to durable contraception in female domestic cats for at least two years, without clinically significant adverse effects. Despite preserved estrous cycling and generally unchanged E2, inhibin A/B, follistatin, and testosterone, treated cats exhibited elevated LH, reduced progesterone, and fewer luteal phases (notably in high-dose), indicating suppression of ovulation and/or corpus luteum formation. The lack of luteal phases after natural breeding strongly supports impairment of breeding-induced ovulation. Mechanistically, contraception in cats may differ from mice: AMH may allow some antral-like follicle development sufficient for E2 production while blocking ovulation. Alternatively, AMH may modulate pituitary function and the LH surge, contributing to anovulation. Elevated LH alongside reduced P4 suggests a mild hypergonadotropic hypogonadism phenotype. Behaviorally, some treated females still displayed estrus and permitted mating, yet remained infertile, indicating that complete suppression of estrus is not required for contraceptive efficacy. The pilot study suggests that sustained high AMH may protect against cystic endometrial hyperplasia by preventing ovulations. Overall, the data indicate that AMH-based gene therapy can safely prevent pregnancy by inhibiting ovulation while maintaining sex steroid production and estrous cycling in many animals, supporting its relevance as a non-surgical alternative for population control.

This study shows that AAV9-mediated delivery of a feline AMH transgene (fcMISv2) provides safe, durable, single-dose contraception in female domestic cats by preventing ovulation and pregnancy for at least two years. The approach maintains estrous cycling and sex steroid profiles while reducing luteal phases, and it elicits minimal immunogenicity with favorable safety monitoring and limited, transient viral shedding. Mating trials confirmed complete infertility in treated cats versus consistent fertility in controls. These results position AMH-based, AAV9-delivered gene therapy as a viable non-surgical contraceptive alternative to spay surgeries for domestic cats. Future research should define the precise ovarian and neuroendocrine mechanisms (including LH surge dynamics and corpus luteum formation), assess long-term uterine histopathology in treated intact females, evaluate potential effects on seasonal or male-induced hormonal changes, explore behavioral outcomes, and investigate scalability and field deployment for community cat population management.

• The study could not determine whether AMH affects seasonal or male-induced changes in antral follicle hormones (e.g., CNP, follistatin), which was identified as a limitation.
• Uterine histopathology in the AAV9-fcMISv2-treated cats could not be evaluated because females remain intact; assessments relied on serial exams and ultrasounds.
• Maternal-fetal transmission of AMH could not be assessed in treated cats due to absence of pregnancies; only reference newborn AMH values from control litters were obtained.
• Pilot fcMISv1 cats showed variable anti-transgene immune responses that affected AMH levels, complicating interpretation for that vector generation (addressed with the optimized fcMISv2).