Not a Deficit, Just Different: Prepulse Inhibition Disruptions in Autism Depend on Startle Stimulus Intensities

Hypothetical model of the acoustic startle response and PPI. The primary acoustic startle pathway (yellow) consists of the cochlear nucleus, caudal pontine reticular nucleus (PnC), and spinal cord motor neurons. The primary circuit responsible for PPI (purple) involves the cochlear nucleus and inferior colliculus of the ascending auditory pathway (gray dotted frame) and pedunculopontine tegmental nucleus (PPTg). PPTg inhibitory projections to the PnC are assumed to mediate the attenuation of the startle response. Additional modulatory input to the PPTg includes projections from the medial geniculate body, auditory cortex, substania nigra pars reticulata, and ventral pallidum. Figure modified from Möhrle et al. (2021) and El-Cheikh Mohamad et al. (2023). Based on data and information from Koch (1999), Fulcher et al. (2020), Gómez-Nieto et al. (2020), and Weible et al. (2020).

Hypothetical model of how the baseline startle response curve is changed by startle and sound scaling. A, Hypothetical startle response curve (black) is scaled by a prepulse resulting in a downward and rightward shift (purple), indicative of startle and sound scaling, respectively. Classical analyses of PPI calculate the %PPI using one startle stimulus intensity (gray dotted frame) and consequently, cannot be informative of scaling components. B, Startle scaling results from a reduction in response amplitude and is shown by a downward shift in the startle response curve (orange). Startle scaling is determined by the change in the maximum response magnitude (Top). C, Sound scaling results from a reduction in sound sensitivity and is shown by a rightward shift in the startle response curve (green). Sound scaling is determined by the change in threshold, ES50, and saturation point. Figure modified from El-Cheikh Mohamad et al. (2023). Based on data and information from Martin-Iverson and Stevenson (2005), Miller et al. (2021), and El-Cheikh Mohamad et al. (2023).

Cntnap2 KO rats have increased startle reactivity that is associated with a leftward shift of the startle I/O curve in males but not females. Cntnap2 WT rats are represented in blue and Cntnap2 KO rats in orange. All graphs show group medians from the 40 repetitions per startle stimulus SPL with error bars representing IQR. Scatterplots show individual values. A, Baseline startle response curves. The colored arrows point to the maximum startle response value (Top) for each genotype. Cntnap2 WT IQR is visualized as the blue-shaded area. Goodness of fit Sy.x: male Cntnap2 WT = 449.7, male Cntnap2 KO = 2,336, female Cntnap2 WT = 181.1, female Cntnap2 KO = 881.9. B, Scaled startle response curve. The colored arrows point to the threshold, ES50, and saturation point for each genotype. Cntnap2 WT IQR is visualized as the blue-shaded area. Goodness of fit Sy.x: male Cntnap2 WT = 0.2048, male Cntnap2 KO = 0.1257, female Cntnap2 WT = 0.1956, female Cntnap2 KO = 0.1937. C, Maximum startle response (Top). The gray arrow indicates that there are values outside the limits of the y-axis, but graphs were zoomed in to visualize the data more clearly. Sex was collapsed to examine the effect of genotype. Cntnap2 KO rats showed a greater maximum startle response magnitude than Cntnap2 WT rats. D, Threshold. Male but not female Cntnap2 KO rats showed a lower response threshold than Cntnap2 WT rats. E, ES50. Male but not female Cntnap2 KO rats showed a lower ES50 than Cntnap2 WT rats. F, Saturation point. Cntnap2 KO males showed a lower saturation point than Cntnap2 WT males. Cntnap2 KO females showed a higher saturation point than Cntnap2 WT females. *p < 0.05, **p < 0.01, ***p < 0.0001, no asterisk or ns indicates nonsignificance of the comparison.

Classical analysis confirms PPI deficits in Cntnap2 KO rats. Cntnap2 WT rats are represented in blue and Cntnap2 KO rats in orange. Scatterplots represent individual values and black lines represent the median with error bars as IQR. The gray arrows indicate that there are values outside the limits of the y-axis, but graphs were zoomed in to visualize the data more clearly. A, Classical %PPI analysis at 100 dB startle stimulus. Male but not female Cntnap2 KO rats showed a PPI deficit with the 75 and 85 dB prepulse. B, Classical %PPI analysis at 110 dB startle stimulus. Sex was collapsed to examine the effect of genotype. Cntnap2 KO rats showed a PPI deficit with a 75 and 85 dB prepulse. *p < 0.05, **p < 0.01, ***p < 0.0001, no asterisk or ns indicates nonsignificance of the comparison.

Cntnap2 KO rats show intact startle scaling through prepulses. Cntnap2 WT rats are represented in blue and Cntnap2 KO rats in orange. All graphs show group medians from the 40 repetitions per startle stimulus SPL with error bars representing IQR. Scatterplots show individual values. A, Startle response curves for baseline and prepulse conditions (75 and 85 dB) for Cntnap2 WT and KO rats. Note: Y-axis scales for Cntnap2 WT and KO rats are different as Cntnap2 KO rats have a greater baseline startle response magnitude than WT rats. Baseline IQR is visualized as the shaded area. Goodness of fit Sy.x: Cntnap2 WT males baseline = 449.7, 75 dB = 249.6, 85 dB = 54.27, Cntnap2 KO males baseline = 2,336, 75 dB = 1,739, 85 dB = 641.1, Cntnap2 WT females baseline = 181.1, 75 dB = 44.16, 85 dB = 27.03, Cntnap2 KO females baseline = 881.9, 75 dB = 513.5, 85 dB = 242.9. B, Maximum startle response (Top). The gray arrows indicate that there are values outside the limits of the y-axis, but graphs were zoomed in to visualize the data more clearly. For the 75 dB prepulse, data was collapsed across sex to examine the effect of genotype. The change in the maximum startle response (Top) due to a prepulse did not differ between Cntnap2 KO and WT rats. *p < 0.05, **p < 0.01, ***p < 0.0001, no asterisk or ns indicates nonsignificance of the comparison.

Cntnap2 KO rats have greater sound scaling by a prepulse at lower startle stimulus intensities and mildly impaired sound scaling at higher startle stimulus intensities. Cntnap2 WT rats are represented in blue and Cntnap2 KO rats in orange. All graphs show group medians from the 40 repetitions per startle stimulus SPL with error bars representing IQR. Scatterplots show individual values. A, Scaled startle response curves for baseline and prepulse conditions (75 and 85 dB) for Cntnap2 WT and KO rats. Baseline IQR is visualized as the shaded area. Goodness of fit Sy.x: Cntnap2 WT males baseline = 0.2048, 75 dB = 0.09425, 85 dB = 0.2019, Cntnap2 KO males baseline = 0.1257, 75 dB = 0.1458, 85 dB = 0.3584, Cntnap2 WT females baseline = 0.1956, 75 dB = 0.2305, 85 dB = 0.9432, Cntnap2 KO females baseline = 0.1937, 75 dB = 0.1971, 85 dB = 0.1531. B, Change in threshold with a prepulse from baseline. For the 75 dB prepulse, sex was collapsed to examine the effect of genotype. With a 75 dB prepulse, Cntnap2 KO rats showed a greater change in threshold than WT rats. With an 85 dB prepulse, the change in threshold did not differ between Cntnap2 KO and WT females. However, the change in threshold for Cntnap2 KO males was trending toward being significantly greater than WT males. C, Change in ES50 with a prepulse from baseline. With a 75 and 85 dB prepulse, Cntnap2 KO males showed a greater change in ES50 than WT males. Cntnap2 KO and WT females did not differ in the change in ES50 with a 75 dB prepulse, but with an 85 dB prepulse, Cntnap2 KO females had a reduced change in ES50 than WT females. D, Change in saturation with a prepulse from baseline. The change in saturation did not differ between Cntnap2 KO and WT males with a 75 or 85 dB prepulse or between Cntnap2 KO and WT females with a 75 dB prepulse. However, with an 85 dB prepulse, the change in saturation for Cntnap2 KO females was trending toward being significantly lower than WT females. *p < 0.05, **p < 0.01, ***p < 0.0001, no asterisk or ns indicates nonsignificance of the comparison.