Functional analysis of missense mutations G36A and G51A in PAX6, and PAX6(5a) causing ocular anomalies

The PAX6 has been described a "master regulator of eye development". A specific ratio of PAX6, and its alternatively spliced isoform, PAX6(5a), has also been observed essential for optimal function. Mutations into PAX6 lead to a number of ocular, and neuronal defects of variable penetrance and expressivity but the mechanism is either poorly understood or underrepresented. This report describes analysis of functions of two missense mutations, G36A, and G51A, causing optic-nerve hypoplasia and optic-disc coloboma in humans, respectively. Mutations were created by site-directed mutagenesis. Products were detected by in-vitro translation and transient transfection to the cultured NIH-3T3 cells. Their DNA-binding, and transcriptional activation properties were analysed through electrophoretic mobility shift assay and luciferase reporter assay, respectively. Mutations induced changes in conformation and secondary structure of PAX6, and PAX6(5a) not only restrict to specific site of mutation in the paired-domain but extend to homeodomain, and transactivation domain. The PAX6-G36A showed reduced binding to PAX6-consensus binding sequence and PAX6(5a)-consensus binding sequence but its binding affinity to homeodomain binding sequence was unaffected. It showed significantly higher transactivation potential through PAX6-consensus binding sequence but reduced activity with PAX6(5a)-consensus binding sequence and homeodomain binding sequence containing luciferase reporters. The PAX6(5a)-G36A showed enhanced transactivation potential with PAX6-consensus binding sequence, PAX6(5a)-consensus binding sequence, and homeodomain binding sequence containing luciferase reporters. The binding affinity of PAX6(5a)-G36A was significantly higher to PAX6-consensus binding sequence, and PAX6(5a)-consensus binding sequence as compared to PAX6(5a) but remains unaffected to homeodomain binding sequence. The enhanced binding affinity was observed by PAX6-G51A to PAX6-consensus binding sequence, PAX6(5a)-consensus binding sequence, and homeodomain binding sequence. The transactivation potential was observed higher with PAX6-consensus binding sequence but significant reduction was evident with PAX6(5a)-consensus binding sequence, and homeodomain binding sequence containing luciferase reporters. The lower binding affinity to PAX6-consensus binding sequence and PAX6(5a)-consensus binding sequence was observed by PAX6(5a)-G51A but loss of binding affinity was detected to homeodomain binding sequence. However, PAX6(5a)-G51A showed significantly higher transactivation with PAX6-consensus binding sequence, PAX6(5a)-consensus binding sequence, and homeodomain binding sequence containing luciferase reporters. With the eye-specific α-A-crystallin promoter, PAX6-G36A and PAX6-G51A mutants were found to have higher ability to transactivate whereas PAX6(5a)-G36A and PAX6(5a)-G51A have lower transactivation potential compared to their respective wild type forms. Thus, variable DNA-binding and transactivation properties of the mutants with different PAX6-binding sequences provide an insight towards their variable penetrance and expressivity.