Interactive Organization of the Circadian Core Regulators PER2, BMAL1, CLOCK and PML.

The evolutionarily conserved circadian clock is a fundamental mechanism in most life forms, a biological response to Earth’s diurnal cycles. The transcription/translation feedback loops of the mammalian circadian clock sustain the oscillating expression pattern of the key clock genes and clock output genes. In the core loop, the BMAL1/CLOCK heterodimer binds to the E-boxes in the promoters of Per and Cry genes and drives their expression. The prevailing view is that the protein products of the Per and Cry genes form a PER/CRY dimer that enters the nucleus and represses BMAL1/CLOCK-mediated expressions of clock target genes1. The circadian regulator PERIOD 2 (PER2) was identified almost 2 decades ago2,3. We have reported that Per2 mutant and null mice have disrupted circadian rhythm behaviors2,4. At the molecular level, PER2 deficient mice displayed a reduced expression of core clock genes in the suprachiasmatic nucleus (SCN) and in peripheral tissues suggesting that PER2 acts as a positive regulator of the clock transcriptional mechanism in vivo5. Indeed, several independent studies have now supported PER2’s role as a positive regulator of BMAL1/CLOCK mediated transcription. A recent study demonstrated that PER2 acts as a positive regulator via PER2 inhibition of CRY1 mediated repression of BMAL1/CLOCK transcriptional activity6. These observations are consistent with biochemical reconstitution studies using purified recombinant proteins that demonstrated that PER2, in contrast to CRY1 and CRY2, apparently does not interact with the BMAL1/CLOCK:DNA complex7. Rather the study reported that PER2 serves to remove CRY1 from binding the BMAL1/CLOCK:DNA complex. These observations raised the possibility that PER2’s interaction with BMAL1 or CLOCK is not direct but rather through intermediary partner(s). Consistent with the notion that PER2 requires an intermediary partner to interact with the BMAL1/CLOCK transcriptional complex, we have previously reported that in the absence of promyelocytic leukemia (PML) protein, immunoprecipitation (IP) of PER2 did not bring down BMAL1 or CLOCK8. The PML protein has been implicated in many important biological processes such as the DNA damage response, cell division control, chromosome instability and circadian rhythm8,9. In addition, Pml gene translocation with RARα is the major underlying cause of acute promyelocytic leukemia (APL)10. Our studies revealed that PML is an important circadian clock regulator8. We observed that Pml−/− mice have an abnormal phase shift response to a light pulse and have a period length that displays reduced precision and stability. At the molecular level, the loss of PML disrupts PER2 nuclear localization and dampens BMAL1/CLOCK mediated transcription. That the absence of PML leads to alteration of PER2 nuclear localization presented an opportunity to determine whether PER2 interaction with the BMAL1/CLOCK heterodimer complex is direct or occurs via an intermediary partner(s). PML is known to be able to form PML nuclear bodies (PML-NB)11. It has long been suggested that PML-NB’s are macromolecular scaffolds that modulate post-translational modifications and availability of many nuclear proteins with various cellular functions12. In our earlier study, we observed that PML enhances BMAL1/CLOCK transcriptional activity only in the presence of PER28, suggesting that PML may be involved in PER2/BMAL1/CLOCK complex formation. Here we investigate how PML organizes PER2/BMAL1/CLOCK complex formation. We undertook immunofluorescence (IF) and co-immunoprecipitation (co-IP) studies using Pml+/+ and Pml−/− cells to investigate the cellular distribution and protein-protein interactions of PER2 with BMAL1 and CLOCK. Our studies revealed that PML mediates the binding of PER2 to BMAL1 in the BMAL1/CLOCK heterodimer. Our studies also revealed that BMAL1 but not CLOCK is pivotal for BMAL1/CLOCK heterodimer interaction with PML. We observed that the cellular distribution of PER2 and CLOCK but not BMAL1 is influenced by PML. Using Bmal1−/− MEF, we observed that nuclear distribution of CLOCK but not PML is influenced by BMAL1.