Prenatal VPA exposure impairs pair-pulse induced LTD at the synaptic level without altering spiking output in the dentate gyrus

Objectives

Prenatal valproic acid (VPA) exposure is a well‑established environmental model of autism spectrum disorder (ASD). Long‑term depression (LTD) is a form of synaptic plasticity critical for cognitive flexibility and is known to be disrupted in ASD. However, most studies have evaluated only synaptic responses (fEPSP) without examining how these changes translate to neuronal spiking output (population spike, PS). We investigated whether pair‑pulse stimulation (PPS)‑induced LTD is altered at both fEPSP and PS levels in the dentate gyrus of VPA‑exposed rats in vivo, and explored underlying molecular mechanisms.

Methods

Pregnant Wistar rats received a single injection of VPA (500 mg/kg, i.p.) on embryonic day 12.5 to generate the VPA group (n=9); control animals (n=10) received saline. At postnatal days 55-60, in vivo electrophysiology was performed under urethane anesthesia. The perforant pathway was stimulated using a pair‑pulse protocol (30 ms interstimulus interval, 1 Hz, 900 pulses, total duration 15 minutes) to induce LTD. Field excitatory postsynaptic potentials and population spikes were simultaneously recorded from the dentate gyrus. For molecular analyses, western blotting was performed to measure total synapsin, phospho‑synapsin, GluA1, and GluA2 levels. Immunohistochemistry was used to assess PSD‑95 expression.

Results

In control rats, PPS reliably induced LTD of fEPSP slopes. In contrast, VPA‑exposed rats showed marked impairment of fEPSP‑LTD. The overall difference in fEPSP LTD between groups was statistically significant (p = 0.0085), and the difference during the last 5 minutes of recording was also significant (p = 0.0419). Strikingly, despite this clear impairment at the synaptic level, population spike amplitude did not differ between control and VPA groups at any time point throughout the recording period (p > 0.05), indicating that spiking output remained preserved even though synaptic depression was deficient. Molecular analyses revealed no significant differences between groups in total synapsin or GluA1 levels. However, phospho‑synapsin was significantly decreased in the VPA group compared to controls (p = 0.0186). GluA2 expression was also significantly reduced in VPA animals (p = 0.01). Immunohistochemistry further demonstrated a significant reduction in PSD‑95 expression in the VPA group (p = 0.04).

Conclusions

Prenatal VPA exposure impairs PPS‑induced LTD at the synaptic level in the dentate gyrus in vivo, but spiking output remains unchanged, revealing a dissociation between synaptic plasticity and neuronal firing. This deficit is associated with decreased phospho‑synapsin, reduced GluA2, and lower PSD‑95. Importantly, our findings suggest that the functional impairment in ASD may arise not merely from synaptic dysfunction, but rather from a mismatch at the neuronal circuit level, specifically an alteration in EPSP‑spike coupling. These results provide a multi‑level framework linking molecular alterations to synaptic dysfunction in the VPA model of ASD while revealing an unexpected resilience of network output and highlight that evaluating both fEPSP and PS simultaneously is critical for understanding circuit‑level pathophysiology in autism.

This work was supported by the Erciyes University Scientific Research Projects (SRP) Department under grant numbers TSAÜ-2024-13381.

References

No references