CRISPR–Cas9-edited SHANK3-mutant cynomolgus macaques recapitulate sleep, motor, social, repetitive-behaviour and functional-connectivity phenotypes of Phelan–McDermid syndrome (PMS) and autism spectrum disorder (ASD), and a 2026 F1-generation study has now replicated and extended these phenotypes in a larger, non-mosaic heterozygous cohort with systematic cognitive testing and EEG biomarkers. What remains unresolved is the causal chain: how reduced SHANK3 protein produces the observed local and long-range connectivity alterations, which circuits drive which behaviours, whether the behavioural phenotypes are primary or anxiety/motor-mediated, whether the phenotypes are attributable to SHANK3 loss rather than uncharacterized off-target edits, and whether candidate translational biomarkers (altered pupil responses, diminished auditory-evoked EEG responses) generalize to PMS patients. The F1 study resolves the cohort-size and cognitive-phenotyping questions; the mechanism, anxiety/motor-confound and edit-specificity questions remain open; patient validation of pupillometry remains open, while auditory-evoked patient event-related-potential evidence is conflicting.

Background

SHANK3 encodes a postsynaptic scaffolding protein that organizes excitatory synapses, and its haploinsufficiency (loss of one copy) is the major cause of Phelan–McDermid syndrome (PMS), a 22q13-deletion/SHANK3-mutation neurodevelopmental disorder featuring intellectual disability, hypotonia and, frequently, autism spectrum disorder (ASD). Because rodents lack various primate-specific cortical circuits and higher social cognition, several groups engineered SHANK3-mutant non-human primates: an independent cynomolgus model generated with the CRISPR–Cas9 genome-editing system (targeting SHANK3 exons 6 and 12) showed impaired social interaction, repetitive behaviours and reduced brain-network activity on positron-emission tomography, reversible with fluoxetine (an antidepressant that increases serotonergic signalling) (Tu et al. 2019), and the consortium of Zhou et al. (2019, Nature 570:326–331) generated germline-transmissible CRISPR–Cas9 SHANK3 mutations in cynomolgus macaques with confirmed reduced SHANK3 protein in brain biopsies. In the founder mutants, Zhou et al. reported sleep disturbances, motor deficits, increased repetitive behaviours, reduced vocalization, social and learning impairments, an altered pupillary light reflex (PLR — the reflexive constriction/dilation of the pupil to light, previously reported atypical in autistic children, e.g. Daluwatte et al. 2013), and — from resting-state functional MRI (rsfMRI, which measures correlated spontaneous blood-oxygenation signals between brain regions as an index of functional connectivity) — dysregulated local and long-range (global) connectivity indicative of circuit abnormalities. The authors explicitly flagged what remained unresolved: the founder cohort was small, mosaic (each founder carried a different mixture of edited alleles), and genetically heterogeneous; off-target assessment was limited to the top predicted sites (Supplementary Table 1 of that paper); behavioural differences could be secondary to anxiety or motor impairment; the generality of the pupil findings to PMS patients was unknown; and cognitive findings were task- and subject-dependent. In 2026 the same consortium (Jiang et al., Neuron, advance online publication) generated a larger F1 generation (the first-generation offspring of the founder mutants, here non-mosaic heterozygous SHANK3+/- animals carrying one mutated and one intact allele) precisely to overcome the cohort and mosaicism limits, replicating sleep disturbances, diminished exploration, atypical social interactions, stereotypies and altered functional connectivity, adding electroencephalogram (EEG) evidence of a markedly diminished auditory-evoked response, finding working memory largely intact but paired-associate learning impaired, and building a multi-task phenotyping array intended to establish biomarkers for testing therapeutics. Reviews of genetically modified non-human-primate models (e.g. Feng et al. 2020; Zhang and Wang 2025) concur that mapping circuit-level mechanisms — and standardizing behavioural validation — remains the central gap these models have not yet closed. Adjacent large-animal work (Shank3-mutant beagle dogs, Tian et al. 2023; altered pupil responses in those dogs, Ren et al. 2023) supports cross-species conservation of sensory/pupil phenotypes but does not address primate circuit mechanism or patient generality, and general edited-macaque genomics (e.g. Schmidt et al. 2023, who found large on-target deletions and off-target edits in CRISPR-edited cynomolgus blastocysts) keeps edit specificity an open accountability issue for this line. On the patient side, event-related-potential (ERP — the averaged EEG waveform time-locked to a stimulus) studies of genetically characterized PMS cohorts report intact N1/P2/N2 responses and habituation to simple tones (Isenstein et al. 2022) but delayed, low-amplitude P50/N1/P2 responses and weak inter-trial phase-locking in an auditory gating task (Smith et al. 2025), so whether the macaques' markedly diminished auditory-evoked EEG response — and the altered pupillary light reflex, for which no targeted patient pupillometry study exists — generalizes to patients remains unsettled.

Problem Statement

In the germline-transmissible SHANK3-mutant cynomolgus macaque model (now embodied in the F1 heterozygous SHANK3+/- cohort), determine the causal mechanisms by which reduced SHANK3 protein gives rise to the observed neural-circuit and behavioural phenotypes. Concretely: (a) map the developmental trajectories and causal links whereby SHANK3 haploinsufficiency alters local versus long-range functional connectivity, using approaches suited to establishing causality, and identify which circuit alterations drive which specific behaviours (such as sleep disturbance, diminished exploration and vocalization, stereotypies, atypical social interaction, paired-association learning impairment, diminished auditory-evoked EEG response); (b) disentangle primary social and cognitive deficits from secondary effects of anxiety or motor limitations, using dedicated behavioural paradigms and physiological measures; (c) establish edit specificity through comprehensive genome-wide off-target analysis, so that the phenotypes can be attributed to SHANK3 haploinsufficiency itself; and (d) determine, through targeted clinical investigation of genetically confirmed Phelan–McDermid syndrome patients, whether the candidate translational biomarkers generalize to patients: for the altered pupillary light reflex no patient pupillometry study exists at all, and for the diminished auditory-evoked EEG response the existing patient event-related-potential evidence is conflicting — intact responses and habituation to simple tones versus delayed, low-amplitude early responses with weak phase-locking in auditory gating (Isenstein et al. 2022; Smith et al. 2025) — leaving generalization unsettled.

References

  1. Zhou Y, Sharma J, Ke Q, Landman R, Yuan J, Chen H, Hayden DS, Fisher JW, Jiang M, Menegas W, Aida T, Yan T, Zou Y, Xu D, Parmar S, Hyman JB, Fanucci-Kiss A, Meisner O, Wang D, Huang Y, Li Y, Bai Y, Ji W, Lai X, Li W, Huang L, Lu Z, Wang L, Anteraper SA, Sur M, Zhou H, Xiang AP, Desimone R, Feng G, Yang S. Atypical behaviour and connectivity in SHANK3-mutant macaques. Nature 570(7761):326–331 (2019). doi:10.1038/s41586-019-1278-0, https://doi.org/10.1038/s41586-019-1278-0
  2. Jiang M, Li R, Yang X, Ren Z, Yang Z, Azevedo F, Sharma J, Menegas W, Choi S, Yang L, Li M, Su Z, Zhu L, Guo P, He Z, Huang W, Anteraper S, Dai J, Wang L, Yang S, Desimone R, Feng G, Lu Z. Behavioral phenotypes and neuronal biomarkers in F1 mutant macaque model of SHANK3-associated autism spectrum disorders. Neuron (2026), advance online publication. doi:10.1016/j.neuron.2026.04.018, https://doi.org/10.1016/j.neuron.2026.04.018
  3. Tu Z, Zhao H, Li B, Yan S, Wang L, Tang Y, Li Z, Bai D, Li C, Lin Y, Li Y, Liu J, Xu H, Guo X, Jiang YH, Zhang YQ, Li XJ. CRISPR/Cas9-mediated disruption of SHANK3 in monkey leads to drug-treatable autism-like symptoms. Human Molecular Genetics 28(4):561–571 (2019). doi:10.1093/hmg/ddy367, https://doi.org/10.1093/hmg/ddy367
  4. Feng G, Jensen FE, Greely HT, Okano H, Treue S, Roberts AC, Fox JG, Caddick S, Poo MM, Newsome WT, Morrison JH. Opportunities and limitations of genetically modified nonhuman primate models for neuroscience research. PNAS 117(39):24022–24031 (2020). doi:10.1073/pnas.2006515117, https://doi.org/10.1073/pnas.2006515117
  5. Tian R, Li Y, Zhao H, Lyu W, Zhao J, Wang X, Lu H, Xu H, Ren W, Tan Q, Shi Q, Wang G, Zhang YP, Lai L, Mi J, Jiang YH, Zhang YQ. Modeling SHANK3-associated autism spectrum disorder in Beagle dogs via CRISPR/Cas9 gene editing. Molecular Psychiatry 28(9):3739–3750 (2023). doi:10.1038/s41380-023-02276-9, https://doi.org/10.1038/s41380-023-02276-9
  6. Ren W, Huang K, Li Y, Yang Q, Wang L, Guo K, Wei P, Zhang YQ. Altered pupil responses to social and non-social stimuli in Shank3 mutant dogs. Molecular Psychiatry 28(9):3751–3759 (2023). doi:10.1038/s41380-023-02277-8, https://doi.org/10.1038/s41380-023-02277-8
  7. Ren X, Wang H, Lv X, Zhou Y, Fan Y, Yu Y, Turck CW, Chen Y, Lv L, Hu Y, Li H, Wang W, Qin D, Feng X, Hu X. Establishment of a standardized daily behavior collection and analysis system for brain disease models of rhesus and cynomolgus monkeys and its application in autism spectrum disorder. Journal of Zhejiang University-Science B 25(11):972–995 (2024). doi:10.1631/jzus.B2400294, https://doi.org/10.1631/jzus.B2400294
  8. Matsumoto J, Kaneko T, Kimura K, Blanco Negrete S, Guo J, Suda-Hashimoto N, Kaneko A, Morimoto M, Nishimaru H, Setogawa T, Go Y, Shibata T, Nishijo H, Takada M, Inoue KI. Three-dimensional markerless motion capture of multiple freely behaving monkeys toward automated characterization of social behavior. Science Advances 11(26):eadn1355 (2025). doi:10.1126/sciadv.adn1355, https://doi.org/10.1126/sciadv.adn1355
  9. Zhang L, Wang JH. The utilization of non-human primates in the investigation of autism spectrum disorder. Brain Research 1866:149900 (2025). doi:10.1016/j.brainres.2025.149900, https://doi.org/10.1016/j.brainres.2025.149900
  10. Schmidt JK, Kim YH, Strelchenko N, Gierczic SR, Pavelec D, Golos TG, Slukvin II. Whole genome sequencing of CCR5 CRISPR-Cas9-edited Mauritian cynomolgus macaque blastomeres reveals large-scale deletions and off-target edits. Frontiers in Genome Editing 4:1031275 (2023). doi:10.3389/fgeed.2022.1031275, https://doi.org/10.3389/fgeed.2022.1031275
  11. Daluwatte C, Miles JH, Christ SE, Beversdorf DQ, Takahashi TN, Yao G. Atypical pupillary light reflex and heart rate variability in children with autism spectrum disorder. Journal of Autism and Developmental Disorders 43(8):1910–1925 (2013). doi:10.1007/s10803-012-1741-3, https://doi.org/10.1007/s10803-012-1741-3
  12. Ponson L, Gomot M, Blanc R, Barthelemy C, Roux S, Munnich A, Romana S, Aguillon-Hernandez N, Malan V, Bonnet-Brilhault F. 22q13 deletion syndrome: communication disorder or autism? Evidence from a specific clinical and neurophysiological phenotype. Translational Psychiatry 8:146 (2018). doi:10.1038/s41398-018-0212-9, https://doi.org/10.1038/s41398-018-0212-9
  13. Isenstein EL, Grosman HE, Guillory SB, Zhang Y, Barkley S, McLaughlin CS, Levy T, Halpern D, Siper PM, Buxbaum JD, Kolevzon A, Foss-Feig JH. Neural markers of auditory response and habituation in Phelan-McDermid syndrome. Frontiers in Neuroscience 16:815933 (2022). doi:10.3389/fnins.2022.815933, https://doi.org/10.3389/fnins.2022.815933
  14. Smith MR, Berry-Kravis E, Thaliath A, Isenstein EL, Durkin AR, Foss-Feig J, Siper PM, Nelson CA, Baczewski L, Levin AR, Powell CM, Pulver SL, Mosconi MW, Kolevzon A, Ethridge LE. Phenotypic variation in neural sensory processing by deletion size, age, and sex in Phelan-McDermid syndrome. Journal of Neurodevelopmental Disorders 17(1):51 (2025). doi:10.1186/s11689-025-09642-4, https://doi.org/10.1186/s11689-025-09642-4