Adenosine mediates the amelioration of social novelty deficits during rhythmic light treatment of 16p11.2 deletion female mice

Autism spectrum disorder (ASD) is a developmental disorder with social interaction and communication challenges, restricted/repetitive behavior, and intellectual disability. The 16p11.2 chromosomal deletion confers strong genetic risk for ASD, and 16p11.2 deletion mice recapitulate relevant behavioral abnormalities. Network-level dysfunction, including impaired power modulation and reduced phase synchronization across frequency bands, has been implicated in ASD, with reduced gamma oscillations reported during sensory processing. In mouse models (e.g., Neuroligin 3 R451C knock-in), gamma oscillation dysfunction in medial prefrontal cortex (PFC) leads to social deficits, and optogenetic stimulation of parvalbumin interneurons at 40 Hz and 8 Hz rescues social novelty deficits. It remains unknown whether 16p11.2 deletion mice display PFC gamma deficits and whether non-invasive rhythmic 40 Hz visual stimulation can ameliorate ASD-related behaviors in this model. Adenosine, a neuromodulator acting via A1, A2A, A2B, and A3 receptors, can provide feedback inhibition to limit excitability, and adenosine signaling has been linked to ASD symptomatology. Here, the study tests whether 40 Hz rhythmic light flicker can improve behavioral deficits in female 16p11.2 deletion mice and investigates mechanisms, focusing on adenosine signaling in PFC.

• ASD shows reduced gamma oscillations during sensory processing; PFC gamma dysfunction causes social deficits in mouse models (e.g., Neuroligin 3 R451C), rescued by PV interneuron stimulation at gamma frequencies.
• Rhythmic 40 Hz sensory stimulation has ameliorated deficits in Alzheimer's disease, cerebral ischemia, and Parkinson's disease models by entraining cortical oscillations.
• Adenosine modulates excitatory/inhibitory transmission via A1/A3 (inhibitory) and A2A/A2B (excitatory) receptors; elevated adenosine can attenuate excessive neuronal activity. Genetic variants in adenosine pathways (e.g., ADORA2A) and altered adenosine levels have been associated with ASD-related phenotypes.
• Sex differences exist in 16p11.2 deletion mouse behaviors; prior work used mixed sexes, underscoring a need to analyze females specifically.

Animals: Female 2–3-month-old 16p11.2 deletion mice (C57BL/6N × 129Sv background; Jackson Laboratory) bred in SPF facility (23 ± 1 °C; 50 ± 10% humidity; 12:12 h light/dark cycle). Group-housed, food/water ad lib. Randomization and blinding used. Sample size: ≥6 mice/group for behavior; ≥3 mice/group for electrophysiology, immunostaining, Golgi, photometry, microdialysis, and LFP.

Ethics: Approved by Animal Care Committee, Southern University of Science and Technology; ARRIVE guidelines followed.

Drugs: SCH58261 (A2A antagonist; 2 mg/kg, p.o. in 5% DMSO/95% corn oil) and DPCPX (A1 antagonist; 4 mg/kg, p.o.) daily for 14 d before 1 h 40 Hz stimulation. Adenosine (100 mg/kg, i.p. in 2.5% DMSO/saline) for 7 d before behavior. In vitro: adenosine 10 µM; DPCPX 0.3 µM; SCH58261 0.1 µM in aCSF.

Visual stimulation: Custom Tang-Guang system with six 36 V LEDs around transparent cages; control and experimental groups stimulated simultaneously once daily 6–7 pm for up to 14 d. Frequencies: 10 Hz, 40 Hz, 70 Hz; 1 h/day.

Behavioral assays: Open field (40×40×40 cm; 10 min), Novel Object Recognition (two sessions separated by 2 h; 10 min each), Three-chamber sociability and social novelty (10 min habituation; 10 min sociability with empty vs stranger mouse; 10 min novelty with familiar S1 vs novel S2). Sociability Index and Social Novelty Preference Index computed. Marble burying: 20 marbles for 20 min; marbles >75% covered counted.

Stereotaxic AAV and shRNA: Bilateral PFC injections (AP +1.98 mm, ML ±0.3 mm, DV −2.0 mm) of PAAV-U6-shRNA Adora1-CMV-EGFP or control virus (250 nL at 0.05 µL/min). Post-op analgesia/antibiotics for 3 d. Expression for 3 weeks. A1R shRNA sequence: 5'-CTCCTTGGGTGTGAATATTGA-3'; control: 5'-CCTAAGGTTAAGTCGCCCTCG-3'.

Electrophysiology: Acute PFC slices (350 µm). Whole-cell recordings at room temp. sEPSC recorded at −60 mV in bicuculline (20 µM); sIPSC at +10 mV in AP5 (50 µM) + CNQX (10 µM). PPR measured with paired pulses (20–500 ms intervals) under bicuculline. Intrinsic excitability assessed with current steps (0–380 pA). Analysis with Mini-analysis and Fitmaster.

Immunofluorescence: Fixed brains; 30 µm sections; antibodies: PSD95, VGLUT1, NeuN, Iba1, CD68, PV, GFAP, MAG; confocal imaging (Zeiss LSM980).

Synaptic puncta and colocalization: High-magnification confocal images; perisomatic VGLUT1 puncta density quantified; VGLUT1/PSD95 colocalization via ImageJ plugins.

Microglial morphology and engulfment: Skeleton analysis of Iba1 z-stacks; 3D Imaris analysis for VGLUT1 engulfment percentage.

Golgi staining: FD Rapid GolgiStain; 200 µm sections; dendritic spine density per 10 µm.

Fiber photometry: rAAV-hSyn-Ado1.0 adenosine sensor injected unilaterally in PFC (AP +1.98, ML 0.3, DV −2.0 mm); recordings during 40 Hz flicker; ΔF/F0 peaks defined as >3× baseline SD.

Microdialysis/HPLC: Guide cannula in PFC; aCSF perfusion 2 µL/min; dialysates collected every 20 min before and after 1 h 40 Hz; adenosine quantified by HPLC.

LFP recordings: 4-channel microwire array in PFC (AP +1.98, ML ±0.3, DV −2.0 mm). Freely moving mice; sampling 1000 Hz; 0.5–120 Hz band-pass; analysis via multitaper PSD and STFT.

Statistics: Data as mean ± SEM. Normality via Shapiro–Wilk. Unpaired t-test or Mann–Whitney U; One-way ANOVA with Tukey or Kruskal–Wallis with Dunn; Two-way RM ANOVA for firing rate/PPR; significance P < 0.05.

• Female 16p11.2 deletion mice showed normal locomotion/anxiety and sociability but significant social novelty deficit (three-chamber SNI; unpaired t-test P < 0.001 vs WT). NOR and marble burying were unaltered.
• 40 Hz light flicker (1 h/day) for 7 or 14 days did not affect locomotion/anxiety or sociability but significantly improved social novelty (One-way ANOVA SNI F(2,31)=5.299, P=0.01; Tukey P16p vs 16p+40Hz-7d=0.04; P16p vs 16p+40Hz-14d=0.02). Benefit dissipated 2 weeks after stopping stimulation. 10 Hz and 70 Hz did not improve SNI; 70 Hz increased anxiety.
• LFP power in PFC was elevated in 16p females across 4–13 Hz, 30–50 Hz, and 50–80 Hz bands versus WT (Mann–Whitney U, P < 0.001). 14 d of 40 Hz reduced LFP power to near/below WT levels (P < 0.001).
• Electrophysiology: sEPSC frequency increased in 16p vs WT (P=0.005) with unchanged amplitude; 40 Hz (7 d and 14 d) reduced sEPSC frequency (ANOVA F(2,34)=6.138, P=0.005; Tukey P=0.02 and 0.01). sIPSC largely unchanged; modest increase in amplitude only after 14 d (ANOVA F(3,53)=4.759, P=0.0052; Tukey P=0.011). PPR unchanged across groups, suggesting postsynaptic/synapse number effects rather than presynaptic release probability changes. Neuron counts (NeuN, PV) and intrinsic excitability (firing rate, rheobase, RMP) were not restored by 40 Hz.
• Synapses: VGLUT1+ excitatory boutons and dendritic spines increased in 16p; 40 Hz-14 d reduced both (VGLUT1 boutons: PWT vs 16p < 0.001; P16p vs 16p+40Hz-14d=0.013. Spines: PWT vs 16p=0.0042; P16p vs 16p+40Hz-14d=0.0025). VGLUT1/PSD95 colocalized puncta increased in 16p and reduced by 40 Hz.
• Microglia/astrocytes/myelination: No changes in microglial number; 16p showed increased branching metrics, but 40 Hz did not alter microglial morphology, CD68, or VGLUT1 engulfment. GFAP and MAG levels unchanged; 40 Hz had no effect, suggesting microglia-independent mechanism.
• Adenosine: Fiber photometry showed increased adenosine sensor peak amplitude (WT: P=0.036) and trend for frequency during 40 Hz; similar increases in 16p. Microdialysis confirmed PFC adenosine elevation after 1 h 40 Hz in WT. In slices from 16p, adenosine (10 µM) reduced sEPSC frequency (PACSF vs Ade=0.009), blocked by A1R antagonist DPCPX (PAde vs Ade+DPCPX < 0.001); A2AR antagonist SCH58261 had no effect on sEPSC frequency.
• In vivo adenosine (100 mg/kg i.p., 7 d) improved social novelty in 16p (P=0.024) without affecting sociability or locomotion.
• Receptor specificity: Systemic DPCPX, but not SCH58261, blocked 40 Hz improvement in SNI (ANOVA F(2,20)=6.776, P=0.006; Tukey P16p+VEH+40Hz-14d vs 16p+DPCPX+40Hz-14d=0.039) and reversed 40 Hz reductions in sEPSC frequency and amplitude; neither drug altered PPR. DPCPX also prevented 40 Hz reductions in VGLUT1 boutons.
• Genetic validation: PFC A1R knockdown via AAV-shRNA reduced A1R expression (P < 0.001), abolished 40 Hz improvement in SNI, increased sEPSC frequency (P=0.019) without affecting amplitude or PPR, prevented reduction of VGLUT1 boutons, and increased dendritic spine density.

The study addresses whether non-invasive 40 Hz visual flicker can ameliorate ASD-like social deficits in 16p11.2 deletion females and elucidates mechanisms. Long-term 40 Hz stimulation selectively improved social novelty without altering general activity or sociability, and normalized elevated PFC LFP power. At the synaptic level, 16p mice showed increased excitatory transmission and synapse density; 40 Hz reduced sEPSC frequency and excitatory synapse markers without changing presynaptic release probability, neuron numbers, or intrinsic firing rates, suggesting modulation of excitatory synapse function/number. Microglial activation and phagocytosis were not enhanced by 40 Hz, indicating a microglia-independent mechanism. Instead, 40 Hz induced adenosine release in PFC; adenosine suppressed excitatory transmission via A1R, and both pharmacological blockade (DPCPX) and genetic knockdown of A1R abolished the behavioral and synaptic benefits, while A2AR antagonism was ineffective. These findings support a model in which 40 Hz entrainment elevates adenosine signaling to engage postsynaptic A1R, reducing excessive excitatory transmission and synapse density, thereby improving social novelty processing. The heterogeneous magnitude of response across individuals reflects variability in the ASD model and suggests differential sensitivity to gamma entrainment. Potential ancillary contributions from adenosine-mediated neurovascular effects could further support network function in 16p11.2 deletion.

This work demonstrates that rhythmic 40 Hz visual stimulation alleviates social novelty deficits in female 16p11.2 deletion mice, normalizes elevated PFC LFP power, and reduces excessive excitatory transmission and synapse density. Mechanistically, 40 Hz evokes adenosine release in PFC, which acts through A1 receptors to suppress excitatory neurotransmission; blocking or knocking down A1R abrogates these benefits. Adenosine administration alone partially phenocopies the behavioral rescue. These results identify adenosine–A1R signaling as a key mediator linking gamma-frequency sensory entrainment to circuit-level and behavioral improvements in an ASD genetic model and support 40 Hz light flicker as a promising non-invasive therapeutic strategy. Future work should assess durability and dosing paradigms (given benefits waned after 2 weeks without stimulation), extend testing to additional ASD models (e.g., SETD5), explore sex differences and individual variability in responsiveness, and delineate downstream A1R signaling mechanisms and potential neurovascular contributions.

• Model specificity: The 16p11.2 deletion represents ~5.5% of ASD genetic cases; generalizability to other etiologies is uncertain.
• Sex and genotype scope: Experiments focused on female 16p11.2 deletion mice; effects in males or other ASD models were not tested.
• Durability: Behavioral benefits dissipated 2 weeks after cessation of 40 Hz stimulation, indicating need for continued exposure or optimized schedules.
• Response heterogeneity: Variable rescue magnitudes across individuals suggest differential sensitivity to 40 Hz entrainment.
• Microglia/astrocyte measures were negative; while suggesting independence, they do not exclude subtle glial or neurovascular contributions not captured by assays.