Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science


Anatomy and Cell Biology


Willmore, Katherine E.


The purpose of this study was to determine whether connexin 43 (Cx43) contributes to craniofacial phenotypic variability. Skull shape and variation were compared within and among two heterozygous mutant mouse models (G60S/+ and I130T/+) that exhibit different levels of Cx43 channel function when compared to their wildtype counterparts (~80% and ~50% reduction in function, respectively). Results indicated mutants have significant differences in skull shape compared to wildtype littermates. Similar patterns of shape difference were found in both mutants. Increased skull shape variation and a disruption in the covariation of skull structures were observed in G60S/+ mutants only. These results show that while a 50% reduction in Cx43 function causes a shift in mean skull shape, there is a lower threshold at which Cx43 function disrupts craniofacial phenotypic robustness. This study demonstrates that Cx43 can contribute to phenotypic variability of the skull through a nonlinear relationship between Cx43 function and phenotypic outcomes.

Summary for Lay Audience

The skull is complex in its development and function. This complexity makes it vulnerable to disruptive forces, and upwards of one third of inherited abnormalities in humans involve the skull. Abnormalities in the skull can be viewed as extreme examples of variation, and it is generally assumed that both normal and extreme variation is structured through development. If this assumption is correct, then we would expect that patterns of variation will be similar if the developmental mechanisms are similar, and that shape variation can then be used to help uncover potential areas of developmental disruption in the case of unusual development. This study tests the hypothesis that the function of a protein important for communication between cells, particularly those in bone (connexin 43), has a nonlinear relationship with how susceptible the skull is to shape change. This hypothesis was tested using mutant mice with normal and abnormal skull shapes. Two mouse models with mutations to the gene encoding for connexin 43 (Cx43) were used to test the relationship between Cx43 function and shape variation in the skull. The I130T/+ mutant mouse has a fifty percent reduction in channel function, whereas the G60S/+ mutant has an eighty percent reduction. Therefore, it was predicted that the G60S/+ mutants would have a greater difference in shape than I130T/+ but that the localization of shape differences would be similar in both mutants. Shape analyses indicated that the two mutants have significantly different skull shapes compared to non-mutant littermates. As predicted, shape changes and shape variation are more severe in G60S/+mice than the I130T/+, but the areas and patterns of change are similar in both mutants. While the fifty percent reduction in Cx43 in I130T/+ mice was enough to change the mean shape, the range of variation (or variability) was only increased in the G60S/+ mutants, meaning that there is a nonlinear relationship between phenotype and function in the skull. This study supports the suggestion that any gene that contributes to trait development can likewise modulate trait variability, extending to genes that play important though not crucial roles in development.