In mammals, a set of core clock genes form transcription–translation feedback loops to generate circadian oscillations. We and others recently identified a novel transcript at the Period2 (Per2) locus that is transcribed from the antisense strand of Per2. This transcript, Per2AS, is expressed rhythmically and antiphasic to Per2 mRNA, leading to our hypothesis that Per2AS and Per2 mutually inhibit each other's expression and form a double negative feedback loop. By perturbing the expression of Per2AS, we found that Per2AS transcription, but not transcript, represses Per2. However, Per2 does not repress Per2AS, as Per2 knockdown led to a decrease in the Per2AS level, indicating that Per2AS forms a single negative feedback loop with Per2 and maintains the level of Per2 within the oscillatory range. Per2AS also regulates the amplitude of the circadian clock, and this function cannot be solely explained through its interaction with Per2, as Per2 knockdown does not recapitulate the phenotypes of Per2AS perturbation. Overall, our data indicate that Per2AS is an important regulatory molecule in the mammalian circadian clock machinery. Our work also supports the idea that antisense transcripts of core clock genes constitute a common feature of circadian clocks, as they are found in other organisms.

Circadian transcriptome studies identified a novel transcript at the Period2 (Per2) locus, which we named Per2AS. Per2AS is a long non-coding RNA transcribed from the antisense strand of Per2, and is expressed rhythmically and anti-phasic to Per2 mRNA. Previously, we mathematically tested the hypothesis that Per2AS and Per2 mutually inhibit each other’s expression by forming a double negative feedback loop, and found that Per2AS expands the oscillatory domain. In this study, we have experimentally tested this prediction by perturbing the expression of Per2AS in mouse fibroblasts. We found that Per2AS represses Per2 pre-transcriptionally in cis and regulates the amplitude of the circadian clock, but not period or phase. Unexpectedly, we also found that Per2 positively regulates Per2AS post-transcriptionally, indicating that Per2AS and Per2 form a single negative feedback loop. Because knock-down of Per2 does not recapitulate the phenotypes of Per2AS perturbation and Per2AS also activates Bmal1 in trans, we propose that Per2AS regulates the amplitude of the circadian clock without producing a protein by rewiring the molecular clock circuit.