Longitudinal Assessment of Fgf17-/- Mouse Social Behavior On the C57Bl/6 Background

Background: Fgf17 mutant mice on a mixed background (129X1/SvJ, C57Bl/6NCrl) have a reduction both in the size of the dorsomedial frontal cortex and in the extent of frontal cortex subcortical projections, as well as cerebellar defects.  In addition, Fgf17^-/- pups at post-natal day (PND) 8 emit fewer isolation-induced ultrasonic vocalizations (USVs) than their wild-type littermates.  Furthermore, Fgf17 ^-/- adults show reduced affiliative interactions and interaction time with a novel conspecific.  Because behaviors are strain-dependent it was uncertain whether the phenotype characterized in the original description of Fgf17 ^-/- mouse behavior resulted from the loss of Fgf17 function, or was related to the decreased social interaction typical of the 129X1/SvJ strain.

Objectives: We re-examined the behavioral repertoire of Fgf17 ^-/- mice at the N5 generation of crossing onto the C57Bl/6 background, both during early post-natal development and late in life, to confirm and extend the results of the original behavioral analyses. 

Methods: Pups were generated from Fgf17^-/- males mated with Fgf17^-/- females of the N5 backcross generation.  On PND 8, pups from 2 litters of each genotype were tested for their USVs in response to maternal isolation, according to the protocol used in the original behavior analysis.  After these pups reached late adulthood (11-12 months of age) their social behaviors were assessed according to previously published protocols.  Data were analyzed using linear regression and repeated measures ANOVAs.

Results: There was a statistically significant difference in vocalization patterns between N5 Fgf17^-/- pups and Fgf17^+/+ pups. Fgf17^-/- mutants showed a trend toward a statistically significant negative effect on total vocalizations as well as a statistically significant negative effect on potentiation (i.e., Fgf17^-/- pups vocalize less overall, and have less potentiation after maternal contact).  When these pups were tested as aged adults (11-12 months old), there were genotype-based differences in their social behaviors.   In a test of social exploration, Fgf17^-/- adult mice had significantly more brief interactions, and showed trends toward fewer extended interactions, as well as more aggressive and mounting behaviors (similar to the results from the original behavior analysis).  In a test of social approach, Fgf17^-/- adults were also significantly slower to approach the side of the chamber occupied by a novel stimulus mouse.  In another test of social interaction, transmission of food preference, there was a sex difference in the effect of genotype:  Fgf17^-/- females ate more cued social food, while Fgf17^-/-males ate less cued social food, than their wild-type counterparts.

Conclusions: In a more homogeneously C57Bl/6 background, Fgf17^-/- mice continue to exhibit social deficiencies that were generally comparable to the original phenotype.  These results suggest that the Fgf17^-/- mutants’ decreased USVs as pups and abnormal social interactions as adults are likely to result directly from the absence of Fgf17.  In these behavioral analyses, mice lacking Fgf17 manifested impairments in two of the three symptom domains (reciprocal social interaction and communication) characterizing autistic spectrum disorders.  Repetitive behavior, the third symptom domain, has yet to be tested.