a 21%, not significantly different than the 10% reduction from IFN-b treatment alone. At 96 h, myotube area was reduced 45% with AG1024 site siISG15 treatment compared to 53% reduced from IFN-b alone. Indeed, at the latest time point, 120 h, siISG15 and IFN-b treated cells were significantly more impaired than IFN-b treatment alone.These findings suggest that ISG15 does not mediate early C2C12 IFN-b myotoxicity. antiproliferative effects of type 1 IFNs in other cell lineages . Regarding the mechanism of type 1 IFN myocyte toxicity, we focused on the potential role of ISG15, a gene whose transcript and protein are highly elevated in human DM muscle. ISG15 is a member of the ubiquitin-like proteins family, with significant sequence homology to ubiquitin. Like ubiquitin, ISG15 is conjugated to many cellular proteins; the conjugation process requires the conjugating enzymes Ube1L, Ube2L6, and HERC5 as well as a deconjugating enzyme, USP18. Previous studies indeed demonstrated both the marked increase in free ISG15 protein as well as numerous ISG15 conjugated proteins in DM muscle and human skeletal muscle culture, suggesting a potential role as a mediator of IFN-mediated toxicity. We therefore examined the effect of IFN-a, IFN-b, and IFN-c in the C2C12 model we developed and demonstrated that IFN-b, but not IFN-a, or IFN-c, resulted in marked and sustained upregulation of ISG15. The lack of IFN-c upregulation of ISG15 is explained by its known mechanism of action, as it binds to the gamma activation sequence in promoter regions, and ISG15 does not contain this 
GAS element. In contrast, the known mechanisms of IFN-a and IFN-b involve binding to the interferonstimulated regulatory element, multiple copies of which are present in ISG15. However, as both IFN-a and IFN-b bind to a common receptor, the more potent and sustained upregulation of ISG15 by IFN-b is an empirical observation of unknown mechanism. Regardless of mechanism, we found that ISG15 silencing did not reverse IFN-b mediated toxicity, providing evidence against ISG15 being a major factor mediating the myotoxic effects of IFN-b. Limitations of our studies include the use of an in vitro immortalized 17984313 cell line model. Nevertheless, the molecular responses of this model to IFN-b are consistent with those found in numerous previously published studies of DM human tissue . It is likely that other type 1 IFN induced genes are involved in myotoxicity, but the detailed mechanisms remain to be established. An approach similar to the one taken here, inhibiting other genes, might yield candidate mechanisms. However, currently no animal model of dermatomyositis exists, so that in vivo exploration of these mechanisms will likely await future development of such a model. Materials and Methods Cell Culture and Treatment with type 1 IFNs Mouse skeletal muscle cells were cultured in 6-well plates, with growth medium consisting of DMEM supplemented with 20% fetal bovine serum. Differentiation was initiated 72 h after seeding, by replacing the growth medium with D-MEM supplemented with 2% horse serum. At approximately 95% confluence, cells were treated daily with IFN-a or -b, 11404282 at two different final concentrations. Human skeletal muscle cells were cultured in 6well plates and treated with human IFN-b as previously described. Discussion Recent studies have identified type1 IFNs or their inducible gene products as biomarkers of DM, suggesting a mechanism of myofiber and capillary injury driven by type 1 IFNs. We th
