Sensory processing dysfunction is a prevalent feature of autism spectrum disorder (ASD), affecting over 94% of individuals across age groups. Tactile sensitivity difficulties are particularly common, often appearing before ASD diagnosis. Early tactile processing alterations may predict the severity of ASD traits later in life. The high variability in ASD phenotypic presentation underscores the need to define neurophysiological signatures that may identify subgroups. Genetic mouse models provide a powerful tool for studying these mechanisms. Previous research showed that ASD mouse models with mutations in genes like *Mecp2*, *Gabrb3*, *Shank3*, and *Fmr1* exhibit tactile overreactivity. Developmental deletion of these genes in peripheral somatosensory neurons leads to abnormal tactile behaviors, social deficits, and increased anxiety in adulthood, highlighting the importance of intact somatosensory function during development. This study aimed to determine how loci of molecular and circuit disruptions in various ASD mouse models predict altered touch reactivity and other ASD-related behaviors.

Existing literature extensively documents sensory processing dysfunction in ASD, with tactile sensitivity issues being particularly prevalent and often preceding a formal diagnosis. Studies in both humans and animal models have shown a correlation between early sensory abnormalities and the severity of ASD symptoms later in development. However, the underlying neurobiological mechanisms remain largely unclear. Research using genetic mouse models has begun to reveal the role of specific genes and neuronal circuits in tactile processing. Studies focusing on genes such as *Mecp2*, *Gabrb3*, *Shank3*, and *Fmr1* have demonstrated a link between their dysfunction and altered tactile reactivity in adulthood. The developmental timing of these disruptions, however, has not been systematically investigated, and this study aims to address this gap in knowledge.

The study used several genetic mouse models of ASD, including mutations in *Mecp2*, *Gabrb3*, *Nlgn2*, and *Rorb*. Adult behavioral tests were conducted to assess tactile sensitivity (tactile prepulse inhibition, air puff response), anxiety-like behaviors (open field test, elevated plus maze), and social interaction (three-chamber social interaction test). To investigate the developmental timing of tactile overreactivity, a novel neonatal air puff responsivity assay was developed, measuring back hairy skin displacement using optical flow analysis. Optogenetic activation of low-threshold mechanoreceptors (LTMRs) at postnatal day 0 (P0) and embryonic day 18.5 (E18.5) was used to investigate the role of specific mechanoreceptor subtypes in tactile reactivity. Conditional knockout mouse lines were generated to determine the cell-autonomous roles of *Gabrb3* and *Nlgn2* in peripheral sensory neurons and spinal cord neurons. Immunohistochemistry was used to examine the expression of key proteins in the dorsal root ganglia (DRG) and spinal cord at different developmental stages. Finally, electrophysiological recordings (whole-cell patch clamp) from DRG neurons and spinal cord neurons were performed to assess the function of GABAergic and glycinergic synapses at neonatal and adult ages.

The study found considerable heterogeneity in adult behavioral phenotypes across different ASD mouse models. While all models displayed tactile overreactivity, only *Gabrb3⁺⁻* and *Mecp2⁺⁻* mice also exhibited increased anxiety-like behaviors and reduced sociability. The timing of tactile overreactivity varied: *Gabrb3⁺⁻* and *Mecp2⁺⁻* mice showed perinatal tactile overreactivity, while *Nlgn2⁻⁺* and *Rorbʰˡ⁺* mice displayed normal neonatal reactivity but enhanced reactivity later in life. Neonatal tactile overreactivity predicted adult anxiety-like behaviors and social deficits. Optogenetic activation of Aβ-LTMRs in *Gabrb3⁻⁻* mice showed enhanced reactivity at both P0 and E18.5, indicating early dysfunction. Conditional knockouts showed that *Gabrb3* is required cell-autonomously in peripheral somatosensory neurons for normal tactile reactivity during early postnatal development, while *Nlgn2* is needed in dorsal horn neurons for normal tactile reactivity in adulthood. Electrophysiological recordings revealed that presynaptic GABAergic inhibition is present in DRG neurons by P4, requiring *Gabrb3* for normal function, while feedforward inhibition in the spinal cord is immature at P4 but mature by P19. *Nlgn2* is required for the maturation of feedforward inhibition in the adult spinal cord but not at early postnatal stages. The lack of early tactile overreactivity in *Nlgn2⁻⁻* and *Rorb⁻⁻* mutants correlated with a lack of anxiety and social behavioral abnormalities.

These findings support a model where ASD-associated gene mutations affecting presynaptic inhibition of LTMRs manifest early in development, leading to profound effects on tactile reactivity and associated behavioral changes. Mutations affecting dorsal horn feedforward inhibition manifest later, without impacting early postnatal tactile reactivity. The developmental timing of aberrant sensory reactivity is crucial, with early overreactivity strongly linked to other ASD-related behavioral issues. The study highlights the importance of considering developmental trajectories in understanding ASD, suggesting that the heterogeneity of ASD may be partly due to the diverse timing of circuit alterations in sensory processing.

This research demonstrates that distinct cellular and synaptic mechanisms in peripheral and central nervous systems contribute to tactile overreactivity in ASD mouse models. The developmental timing of these disruptions significantly influences the presence of comorbid behavioral phenotypes. Early overreactivity stemming from peripheral presynaptic inhibition predicts adult anxiety and social deficits. Further studies should explore the precise timing of circuit maturation and its relation to behavioral outcomes, and these findings highlight the potential use of developmental sensory markers for early ASD diagnosis and intervention.

The study used mouse models, which may not perfectly reflect the complexities of human ASD. The behavioral assays used, while informative, may not capture the full spectrum of ASD-related behaviors. The sample sizes for some experiments were relatively small, warranting further studies with larger cohorts. While the study provides a comprehensive investigation of specific genes and circuits, other genetic and environmental factors contributing to ASD-related sensory processing dysfunction were not addressed.