In contrast with the failure of visceral endoderm formation resulting in embryonic day (E)6.5 lethality of A-/A- mice, replacement with NM II-B or chimeric NM IIs restores a normal visceral endoderm. This finding is consistent with NM II's role in cell adhesion and also confirms an essential, isoform-independent requirement for NM II in visceral endoderm function. The knock-in mice die between E9.5 and 12.5 because of defects in placenta formation associated with abnormal angiogenesis and cell migration, revealing a unique function for NM II-A in placenta development. In vitro results further support a requirement for NM II-A in directed cell migration and focal adhesion formation. These findings demonstrate an isoform-specific role for NM II-A during these processes, making replacement by another isoform, or chimeric NM II isoforms, less successful. The failure of these substitutions is not only related to the different kinetic properties of NM II-A and II-B, but also to their subcellular localization determined by the C-terminal domain. These results highlight the functions of the N-terminal motor and C-terminal rod domains of NM II and their different roles in cell-cell and cell-matrix adhesion. We have successfully expressed full length wild type and mutated NM II proteins using the Sf9-baculovirus system. We also expressed two chimeric NM II proteins Wang A et al. PNAS 2010, 107(33):14645-50 and GFP-NM II fusion proteins. We find that: I) although full length NM II-A, II-B and II-C exhibit biochemical differences, the morphology of the filaments determined by negative-staining electron microscopy (EM) is essentially indistinguishable among the three paralogs. In the presence of ATP all three paralogs display a similar ability to adopt the 10S compact conformation. II) EM images of chimeric molecules show that the tail domains of the paralogs are interchangeable in terms of filament formation and formation of the 10S compact conformation;III) In contrast to a previous report, the presence of point mutations in full length NM IIA proteins (N93K, D1424N, E1841K) causing human diseases has little or no obvious effects on filament formation;IV) GFP fused to NM II allows us to directly analyze in vitro motility by TIRF microscopy. We found that the gene targeting locus, a 6 kb range around Myh9 gene exon2, displays an extremely high and repeatable frequency of homologous recombination in mouse embryonic stem (ES) cells (95% in this case vs. 1-10% in most cases). To our knowledge, this is the highest rate that has been reported. Interestingly, there is lack of unique residues responsible for high targeting efficiency since reducing the length of the targeting arms results in a corresponding decrease of targeting frequency. Importantly, this finding has important applications for stem cell biology because the mouse Myh9 gene locus represents a particularly good alternative to site-directed transgeneis based on the unique properties of Myh9 gene and high HR frequency, as substantiated by our works. These applications include: I) generating genetic replacement models to study the isoform and domain specificity of nonmuscle myosin IIs (Wang A, et al. PNAS 2010;Zhang Y et al. Unpublished data);II) generating Myh9 related disease (Myh9-RD) mouse model, which mimics Myh9-RD in human (Zhang Y, Blood 2012);III) obtaining ES cell-derived cardiomyocytes with the selection marker being site-specifically integrated into myh9 gene locus. In each case above, the desired ES cell clones were easily obtained mainly facilitated by high HR frequency at this locus. Currently, we are further expanding the application of this useful targeting site.
