There are two Y-27632 ROCK inhibitor possible causes for this result: First, the necessary full set of core circadian clock genes are not yet expressed at this stage of differentiation. However, expression of the major core clock genes in mouse neurospheres has been reported. Second, the clock genes are expressed, but the oscillator cannot operate because necessary non-rhythmic positive inputs are missing or an inhibitory factor is present in the spheres during early differentiation. It is also possible that the growth factors in SCM suppress functioning of the clock mechanism. It seems unlikely that either of the added growth factors can completely suppress circadian activity because circadian rhythms were detected in SCM, although these were rare during the late component of imaging sessions. Stem cell state and circadian rhythmicity were negatively correlated, but the rhythmicity of spheres undergoing differentiation in vitro from the most stem-like state in SCM did take various paths, such as changing from a non-rhythmic or ultradian state to circadian. When examining all of the possible paths, the neurospheres that were ultimately circadian during the last 3�C4 days of imaging were ones that had been given forskolin and then maintained in either SM or B27 medium. Spheres under these medium conditions attached and began propagating into neuroblast and glial-like cells that, by day 6 or 7, stained for Dcx and GFAP, respectively, further indicating that more differentiated spheres are more likely to be circadian. It is likely that neurospheres are composed of many individual circadian oscillator cells as well as non-clock cells that are unable to sustain a circadian rhythm without input of timing information from other cells. Similarly, some brain areas when isolated as explant cultures produce circadian activity, whereas others do not. Several major brain structures have been grouped into three categories: endogenous circadian clocks, rapidly damping slave oscillators, and non-circadian. One reason why circadian rhythms were not common in SCM spheres could be because individual circadian clock cells are present but they are not adequately synchronized to a common phase to be detected in the whole-sphere recordings. To test for this possibility, spheres in SCM were given a pulse of forskolin before imaging but the percentage of circadian spheres did not increase. It is possible, but seemingly unlikely, that the less differentiated cells present in SCM are not responsive to forskolin but might contain a circadian clock. The circadian rhythms in spheres imaged in SM did respond to forskolin by showing a significantly clustered phase that was near the phase expected for this Rapamycin treatment, about 24 hours after the pulse. By the third cycle, the forskolin-treated SM spheres had drifted out of phase and were no longer clustered significantly, according to circular statistics. Spheres in B27 medium given a forskolin pulse were significantly clustered, but this occurred at a phase 12 hours away from the expected phase. It is possible that the transition into B27 medium had its own phase-shifting effect that acted in combination with forskolin. B27 medium has been shown to elevate mPer1 expression in cortical astrocyte cultures, suggesting that it could cause a phase-shift by altering the level of this core clock component. Although the forskolin-treated SM spheres were in SCM during the forskolin treatment, and so were mostly undifferentiated, some circadian clock cells must have been present for the forskolin to produce synchronization.