Enhanced fear limits behavioral flexibility in Shank2-deficient mice

Abstract

Background: A core symptom of autism spectrum disorder (ASD) is repetitive and restrictive patterns of behavior. Cognitive inflexibility has been proposed as a potential basis for these symptoms of ASD. More generally, behavioral inflexibility has been proposed to underlie repetitive and restrictive behavior in ASD. Here, we investigated whether and how behavioral flexibility is compromised in a widely used animal model of ASD.

Methods: We compared the behavioral performance of Shank2-knockout mice and wild-type littermates in reversal learning employing a probabilistic classical trace conditioning paradigm. A conditioned stimulus (odor) was paired with an unconditioned appetitive (water, 6 µl) or aversive (air puff) stimulus in a probabilistic manner. We also compared air puff-induced eye closure responses of Shank2-knockout and wild-type mice.

Results: Male, but not female, Shank2-knockout mice showed impaired reversal learning when the expected outcomes consisted of a water reward and a strong air puff. Moreover, male, but not female, Shank2-knockout mice showed stronger anticipatory eye closure responses to the air puff compared to wild-type littermates, raising the possibility that the impairment might reflect enhanced fear. In support of this contention, male Shank2-knockout mice showed intact reversal learning when the strong air puff was replaced with a mild air puff and when the expected outcomes consisted of only rewards.

Limitations: We examined behavioral flexibility in one behavioral task (reversal learning in a probabilistic classical trace conditioning paradigm) using one ASD mouse model (Shank2-knockout mice). Thus, future work is needed to clarify the extent to which our findings (that enhanced fear limits behavioral flexibility in ASD) can explain the behavioral inflexibility associated with ASD. Also, we examined only the relationship between fear and behavioral flexibility, leaving open the question of whether abnormalities in processes other than fear contribute to behavioral inflexibility in ASD. Finally, the neurobiological mechanisms linking Shank2-knockout and enhanced fear remain to be elucidated.

Conclusions: Our results indicate that enhanced fear suppresses reversal learning in the presence of an intact capability to learn cue-outcome contingency changes in Shank2-knockout mice. Our findings suggest that behavioral flexibility might be seriously limited by abnormal emotional responses in ASD.

Keywords: Classical conditioning; Fear; Reversal learning; Shank2.

Fig. 1

Behavioral task. A The experimental setting. Head-fixed mice performed a probabilistic classical conditioning task. B Task schematic. A conditioned stimulus (CS; odor) was delivered for 1 s, there was a 1-s delay, and then an appetitive (water, 6 µl) or aversive (air puff) US was delivered in a probabilistic manner. Each CS was paired with 75% water (CS_Rw) or 75% strong air puff (CS_Pn(strong)) in Task 1, with 75% water or 75% mild air puff (CS_Pn(mild)) in Task 2, and with 80% water (CS_80%) or 20% water (CS_20%) in Task 3

Fig. 2

Measurement of eye closure. The eye closure response to air puff was estimated by measuring pupil diameter with an infrared camera. A Schematic for the experimental setting. B A sample eye closure response to an air puff. Video images and detected eye regions before and after air puff delivery are shown at the bottom

Fig. 3

Reversal learning is impaired in adult male Shank2-KO mice. A Left, two odor cues, CS_Rw and CS_Pn, were paired with 75% reward (water, 6 µl; blue) and 75% strong air puff (100 ms, 3 psi; red), respectively (Task 1). Right, mean licking responses (lick density functions, σ = 100 ms) of WT (left; n = 10) and Shank2-KO (right; n = 10) mice during the last acquisition session. Trials were grouped according to CS and outcome. B Mean delay period anticipatory lick rates in response to CS_Rw (blue) or CS_Pn (red) during initial acquisition (three sessions). C The LDI of WT (black) and Shank2-KO (red) mice in each stage (100 trials) during initial acquisition. D Sample reversal learning sessions. Blue and red lines indicate anticipatory licking responses to CS_Rw→Pn and CS_Pn→Rw cues, respectively, in a moving average of 25 trials. Gray and black asterisks at the top indicate significantly higher anticipatory licking response to the CS_Rw→Pn (gray) or CS_Pn→Rw (black) cue compared to the other cue, in a moving window of 25 trials (p < 0.05, t test). Open triangles indicate the first trial since meeting the cue-outcome contingency reversal criterion. E The number of trials required to exceed the reversal criterion. F The LDI during the first reversal session (moving average of 25 trials). G The LDI during the last 10 trials in F. H, I Relationships between reversal learning performance (ordinate; H, the number of trials needed to reach the reversal criterion; I, LDI during the first reversal session [last 10 trials in F]) and the mean anticipatory lick rate in CS_Rw trials, that in CS_Pn trials, their difference, and the LDI during the last acquisition session (abscissa). Gray dashed lines represent least-squares linear fit. Circles indicate individual animal data E, F, H and I. Squares and bar graphs B, C, E and H present the mean across 10 mice. Shading and error bars indicate the SEM across 10 mice A–C, E, G and H. *p < 0.05, t test

Fig. 4

Reversal learning is impaired in juvenile male Shank2-KO mice. Results of reversal training with a strong air puff (Task 1) in juvenile male WT (n = 10) and Shank2-KO (n = 11) mice. The results are presented as described for Fig. 3

Fig. 5

Intact reversal learning in female Shank2-KO mice. Results of reversal training with a strong air puff (Task 1) in adult female WT (n = 12) and Shank2-KO (n = 10) mice. The results are presented as described for Fig. 3A–G

Fig. 6

Eye closure response to air puff differs between male, but not female, WT and Shank2-KO mice. A Eye closure responses (average of 5 trials) of adult male WT (black; n = 12) and Shank2-KO (red; n = 10) mice (shading, SEM across mice) to strong (100 ms, 3 psi; left), mild (5 ms, 3 psi; middle), and very strong (100 ms, 30 psi; right) air puffs. Shaded rectangles indicate time periods before (green, 1.5 s before air puff onset), during (orange, 1 s after air puff onset), and after air puff delivery (purple, between 2.5 and 4 s after air puff onset). B The difference in eye closure response between Shank2-KO and WT mice to 16 different combinations of air puff duration (abscissa) and pressure (ordinate) before (left), during (middle), and after (right) air puff delivery. C Eye closure responses of adult of female WT (black; n = 11) and Shank2-KO (red; n = 10) mice. The results are presented as described for panels A–C. *p < 0.05, **p < 0.01, using two-way mixed ANOVA followed by post hoc Bonferroni test

Fig. 7

Adult Shank2-KO mice exhibit intact reversal learning with the use of a mild air puff. Results of reversal training with a mild air puff (Task 2) in adult male WT (n = 10) and Shank2-KO (n = 10) mice. The results are presented as described for Fig. 3A–G, except that pink lines indicate anticipatory lick rates in response to the CS_Pn predicting a mild air puff (5 ms, 3 psi)

Fig. 8

Shank2-KO mice exhibit intact learning in reward probability reversal. Results of reversal training using only appetitive outcomes (Task 3) in adult male WT (n = 10) and Shank2-KO (n = 10) mice. The results are presented as described for Fig. 3A–G, except that the purple and green lines denote anticipatory lick rates in response to CS_80% and CS_20%, respectively

Fig. 9

DCS does not rescue the abnormal eye closure response of male Shank2-KO mice. Eye closure responses (average of 5 trials) of adult male WT (black; n = 11) and Shank2-KO (red; n = 10) mice (shading, SEM across mice) treated with DCS (dashed line) or saline (solid line) injection and exposed to the strong, mild, and very strong air puffs