In C2C12 cells, S6K1 activity increased almost 3-fold from subconfluence until myotubes were completely formed. Concomitant with the increase in S6K1 activity was a decrease in IRS-1 signaling and protein levels, consistent with the negative feedback loop from S6K1 to IRS-1. However, knocking down S6K1 had no effect on IRS-1 protein levels, suggesting that physiological activation of S6K1 during differentiation did not lead to the reduction of IRS-1 protein. Even though IRS-1/2 function decreases during differentiation, phosphotyrosine associated p85 increases at confluence and remains high throughout differentiation. These data suggest that during myogenesis the predominant source of PI3K activity is phosphotyrosine complexes other than IRS-1/2.
S6K1 activity increases substantially from subconfluent C2C12 myoblasts to fully differentiated myotubes. Recently, S6K1 has been identified as a key component of a negative feedback loop in insulin/growth factor signaling . This feedback is mediated through serine phosphorylation and direct regulation of IRS-1 protein and mRNA. In differentiated L6 myotubes, activation of S6K1 for as little as 1hr, inhibits insulin stimulated IRS-1 associated PI3K activity . Therefore, we hypothesized that the prolonged activation of S6K1 during myogenesis would lead to a reduction in IRS-1 protein levels. Consistent with this hypothesis, IRS-1 protein declined by more than 50% between Day 1 and Day 3 of differentiation. Since growth factor mediated differentiation is thought to involve IGF-II signaling through IRS-1 to PI3K and PKB, we further hypothesized that decreasing S6K1 would increase IRS-1 protein and result in greater IGF-II signaling and myoblast differentiation. Contrary to this hypothesis, knockdown of S6K1, its structural homologue S6K2, or both enzymes together had no effect on IRS-1 protein level or muscle cell differentiation. These data suggest that during differentiation S6K1 does not regulate the level of IRS-1 protein or muscle cell differentiation. In agreement with the current work, Park et al [16, 20] have elegantly shown that S6K1 is not required for muscle cell differentiation but plays an important role in myocyte hypertrophy once differentiation is complete.
Measuring the amount of p85 associated with a particular scaffolding protein provides an estimate of the PI3K activity of that complex. Due to the requirement on IRS-1 for human myoblast differentiation  we had expected that IRS-1 associated p85 would be substantially elevated from the point of plating. However, both IRS-1 and IRS-2 associated p85 were significantly suppressed during differentiation. Instead of associating with IRS-1 or 2, PI3K associates with other phosphotyrosine complexes during differentiation in C2C12 cells. These data do not preclude a role for IRS in myogenesis. Indeed, cross-talk between cell surface receptors and the insulin/IGF-I pathways has recently been described in mice with skeletal muscle lacking β1-integrins . In mice without β1-integrins insulin-induced phosphorylation of PKB at Ser473 was impaired even though IRS-1 signaling appeared normal. Since integrins and cadherins interact with integrin-linked kinase (ILK;[23, 24], and ILK binds to both PKB and the Ser473 kinase mTORC2 , this suggests that cell surface receptors work together with IRS-1 to activate PI3K and PKB. The current data suggest that in differentiating muscle, the interaction between the cell surface receptors and IRS-1 is reversed. At high confluence, cell-cell contact results in the activation of cadherins and integrins , which function as high confluence-activated receptors. PI3K can be recruited to cell-cell contacts and activated by cadherin , but this has yet to be demonstrated in muscle cells. During differentiation, cadherins are upregulated in C2C12 cells , suggesting that they may play an important role in confluence-dependent differentiation. Some researchers have suggested that cadherins can activate PKB during differentiation, but this has previously been attributed to Rho kinase  or Cdo/JLP  dependent activation of p38. The current data suggest a more direct mode of activation is possible: the direct recruitment of PI3K, possibly in association with a permissive level of activated IRS-1, to the membrane and the local production of PIP3. While the cadherins have yet to be definitively identified as the phospho-tyrosine containing protein that increases its interaction with p85 during differentiation, the location, function, and dynamics of cadherins during differentiation are all suggestive of a role for these proteins.
Our profile of PKB phosphorylation differs from that previously published [29, 30]. We observed a biphasic response, whereas others observe a linear increase in PKB phosphorylation at Ser473 and Thr308 [29, 30]. This discrepancy could be due to the differentiation protocol used. We include data from cells at subconfluence and withdraw serum at 90-100% confluence whilst Deldicque et al  withdrew serum at 70% confluence. Gonzalez et al.  include data from cells at subconfluence however they collected their cells at a higher confluence and they plated their cells on gelatin coated plates which could affect the cell surface receptors that may be involved in PKB activation. These differences may explain why Gonzalez et al. did not observe an initial reduction in PKB Thr308 phosphorylation upon reaching confluence. Using a similar protocol and measuring PKB activity directly, Cuenda and Cohen observed the same increase in PKB activity 1.5 days following serum withdrawal followed by a decrease at day 3 . These results highlight the importance of maintaining a consistent protocol for differentiation studies.