Every biochemical reaction speeds up as temperatures rise, which should make any molecular clock run fast on warm days. Yet circadian clocks, the internal timers that govern daily rhythms of sleep, metabolism and behavior, hold their roughly 24-hour period remarkably steady across a wide range of physiological temperatures. This property, known as temperature compensation, has puzzled chronobiologists for decades. A new genetic study in the fruit fly Drosophila melanogaster now identifies two conserved enzymes, the kinases CK2 and DBT, as central players in this thermal buffering, and shows that they act through separate phosphorylation pathways on the clock proteins PER and TIM.

The research, conducted in Drosophila, builds on the well-established architecture of the fly circadian oscillator. The transcription factors CLOCK and CYCLE activate the genes period and timeless, whose protein products, PER and TIM, accumulate in the cytoplasm, pair up, and enter the nucleus to shut down their own transcription. This negative feedback loop takes about a day to complete and forms the core of the clock. But the timing of the loop depends heavily on post-translational modifications, especially phosphorylation, which controls the stability, localization and activity of clock proteins. The DBT kinase, the fly counterpart of mammalian casein kinase 1 delta/epsilon, is famous for setting the pace of PER degradation, while CK2 has long been implicated in circadian timing across fungi, plants, insects and mammals.

To probe the role of these kinases in temperature compensation, the researchers measured free-running circadian periods, the rhythms that persist in constant darkness, in flies carrying loss-of-function mutations at three temperatures: 18, 25 and 29 degrees Celsius. Wild-type flies kept nearly constant periods across this range. In contrast, heterozygous flies carrying the Ck2alphaTik or Ck2alphaTikR mutations, which impair the catalytic subunit of CK2, showed clear undercompensation: their periods shortened by roughly 1.2 hours between 18 and 29 degrees, and the Q10, a measure of how much a process changes per 10-degree shift, increased significantly. Notably, Ck2alphaTikR largely suppresses the long-period phenotype of Ck2alphaTik yet failed to restore normal temperature compensation, showing that compensation is not simply a byproduct of baseline clock speed.

DBT told a similar story. Flies heterozygous for the catalytically impaired dbtAR allele had periods of about 27.8 hours at 18 degrees and 26.0 hours at 29 degrees, again with an elevated Q10. Two hypomorphic homozygous alleles, dbtS and dbtL, which produce short and long baseline periods respectively, both showed undercompensation, with periods shortening by roughly 1.5 to 2 hours across the temperature range. Because these alleles have opposite effects on period length at standard temperature yet share the same compensation defect, the authors conclude that temperature compensation is mechanistically separable from the determination of absolute circadian period.

The team then traced which phosphorylation sites on PER mediate the DBT effect. Previous work had identified two PER regions implicated in thermal robustness: the N-terminal phosphodegron spanning residues Ser44 to Ser47, and the so-called perShort cluster at amino acids 583 to 596. Within the perShort cluster, DBT directly phosphorylates Thr583, Ser585 and Ser589, while the kinase NEMO phosphorylates Ser596, which primes subsequent DBT action in a hierarchical cascade. Combining dbtAR with the classic perS allele, which carries a Ser589Asn mutation, lengthened the period but did not change the magnitude of temperature-dependent period shortening, indicating that DBT and the perShort cluster act in the same pathway.

Ser47, the terminal site of that phosphorylation cascade and a validated direct DBT target, proved equally important. Phosphonull S47A mutations lengthen the period, while phosphomimetic S47D mutations shorten it, yet both variants show similar overcompensation, with periods lengthening by about 2 hours between 18 and 29 degrees. When the researchers combined S47D with dbtAR, the double mutant displayed an overcompensation phenotype of roughly 2 hours, essentially the same as S47D alone. This epistatic relationship indicates that the S47D mutation overrides the effect of reduced DBT activity, supporting the idea that DBT regulates temperature compensation through phosphorylation of Ser47. By contrast, mutations at the neighboring Ser45 residue showed additive effects with dbtAR, arguing that Ser45 does not participate in the DBT pathway.

Ser45 instead belongs to the CK2 pathway. Double mutants combining S45A with Ck2alphaTik or Ck2alphaTikR showed additive effects on baseline period length but no further worsening of temperature compensation, indicating that Ser45 and CK2 act within the same pathway. Other CK2-regulated PER sites, Ser151 and Ser153, which delay nuclear entry when mutated, had no effect on compensation. The researchers therefore turned to TIM, another CK2 substrate. A conserved serine/threonine-rich region of TIM containing three CK2-targeted residues, Thr305, Ser309 and Ser313, regulates nuclear accumulation of the PER/TIM complex in small ventral lateral neurons. Flies carrying phosphonull substitutions at all three sites showed long periods and pronounced undercompensation, with periods of about 28 hours at 18 degrees and 26 hours at 29 degrees, and no additive effect when combined with Ck2alphaTikR. The TIM site Ser1404, which controls nuclear export, did not affect compensation.

Combining the dbtAR mutation with Ck2alphaTik produced robust additive effects on temperature compensation, confirming that the two kinases operate through distinct, parallel pathways. Further genetic interactions reinforced this conclusion: Ck2alphaTik clearly suppressed the overcompensation of S47A mutants, and combining undercompensated Ck2alphaTik with overcompensated S47D restored completely normal compensation across the 18-to-29-degree range. The authors note that Ck2alphaTikR, a complete loss-of-function allele, produced weaker but still detectable suppressive effects, likely because the dominant-negative Ck2alphaTik more profoundly reduces total CK2 activity in heterozygous flies.

Molecular analyses of the severely undercompensated S45A;dbtAR double mutant, whose period shortens by about 4.3 hours across the temperature range, revealed what goes wrong biochemically. These flies accumulated elevated levels of PER, particularly hyperphosphorylated PER, and showed reduced amplitude in PER and TIM oscillations. Crucially, wild-type flies maintain nearly identical PER phosphorylation profiles at 18 and 29 degrees, meaning PER phosphorylation rhythms are themselves temperature compensated. In the double mutant, however, the ratio of hyperphosphorylated to hypophosphorylated PER was consistently higher at 18 degrees than at 29 degrees, showing that PER phosphorylation dynamics had lost their own thermal compensation, mirroring the behavioral defect.

Imaging of the s-LNv pacemaker neurons added a spatial dimension. In wild-type flies, PER shuttles between nucleus and cytoplasm on a schedule that is itself temperature compensated. In the double mutant, PER accumulated excessively in the nucleus at circadian time 16 and was inefficiently cleared during the subjective day, with the localization rhythm delayed at warm temperature. The authors propose that the critical determinant of temperature compensation is not simply how long PER stays in the nucleus, but its phosphorylation state there: nuclear hypophosphorylated PER appears to promote compensation, while nuclear hyperphosphorylated PER antagonizes it. Even the most severe defect in the study raised the Q10 only to about 1.13, underscoring how robust and partially redundant the underlying mechanisms are. The researchers caution that the relationships between kinases and phosphorylation sites are inferred from genetic interactions, that Ser45 may act structurally rather than as a CK2 substrate, and that their biochemical analyses of head extracts reflect multiple tissues rather than pacemaker neurons alone. Still, the work paints temperature compensation as an emergent property of a balanced phosphoregulatory network, with CK2 and DBT buffering the fly clock against the thermal physics that would otherwise derail it.

Subject of Research: CK2 and DBT kinase pathways regulating temperature compensation of the Drosophila circadian clock

Article Title: CK2 and DBT promote temperature compensation of the Drosophila circadian clock via distinct pathways

Article References: Xia, Y., Louis, V., North, L. E., & Emery, P. (2026). CK2 and DBT promote temperature compensation of the Drosophila circadian clock via distinct pathways. iScience, 29(11), Article 118004. https://doi.org/10.1016/j.isci.2026.118004

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.118004

Keywords: circadian clock, temperature compensation, Drosophila, CK2, DBT, PER, TIM, phosphorylation, kinases, chronobiology, genetics, promote