unc-89 mutants display reduced locomotion, and disorganization of muscle thick filaments, and usually lack M-lines.
unc-89 encodes 6 major isoforms, ranging in size from 156,000 to 900,000 Da. The largest of these isoforms, UNC-89-B, consists of 53 immunoglobulin (Ig) domains, 2 fibronectin type III (Fn3) domains, a triplet of SH3, DH and PH domains near its N-terminus, and two protein kinase domains called PK1 and PK2 near its C-terminus. The human homolog is called "obscurin". UNC-89 is localized to sarcomeric M-lines. To clarify how UNC-89 is localized and how it performs its functions, we are using a yeast 2-hybrid approach to identify binding partners. UNC-89 region Ig8-13 was used to screen a yeast 2-hybrid library and this yielded the protein CUL-1. CUL-1 is one type of cullin, a component of SCF complexes that direct the action of E3 ubiquitin ligases. A crucial protein for vertebrate muscle protein turnover is Atrogin-1, which is another component of an SCF complex that contains
cul1 (CUL-1 of vertebrates). When CUL-1 was used to screen 15 additional segments that cover all of UNC-89-B (8,081 amino acids), two other segments of UNC-89 were found to interact. The following regions of UNC-89 were found to be necessary and sufficient for interaction with CUL-1: Ig9-11, Ig18-23, and Ig50-51. A binding assay using purified recombinant proteins verified the interactions. With reference to the crystal structure of the human Cul1-Rbx1-Skp1 complex (Zheng et al. 2002), Ig8-13 requires structural Repeat 2, Ig18-23 requires Repeats 1, 2 and 3, and Ig48-52 requires Repeats 1 and 2 for binding. The interaction of UNC-89 with this N-terminal half of CUL-1 is compatible with the idea that UNC-89 influences interaction of CUL-1 with its adaptor protein SKR-1 (Skp1 in worms). RNAi knock down of
cul-1 results in disorganization of myosin thick filaments, a phenotype that is very similar to that of
unc-89(
su75), which lacks all the large UNC-89 isoforms and therefore lacks all CUL-1 binding sites. In addition,
cul-1(RNAi) results in nearly normal localization of UNC-112 to M-lines and dense bodies, but with some abnormal accumulations of UNC-112 protein. Interestingly,
unc-89 mutants display a thinner sarcomeric region, and we hypothesize that this might result from increased degradation of sarcomeric proteins. Furthermore, we hypothesize that normally UNC-89 sequesters CUL-1 and prevents it from promoting ubiquitin mediated degradation of sarcomeric proteins. Thus, potentially, an obscurin/cullin interaction may have an important role in preventing human muscle atrophy, a condition found in normal aging (sarcopenia), and in many chronic diseases. We are now conducting experiments to test our hypothesis.