Clock genes and the molecular machinery of circadian rhythms

That paper was published in the Proceedings of the National Academy of Sciences with modest fanfare. The field it launched led, 46 years later, to a Nobel Prize.

Jeffrey Hall and Michael Rosbash at Brandeis University, and Michael Young at Rockefeller University, spent the 1980s and 1990s working out what the period gene actually does — and discovering the molecular feedback loop that makes it a clock. The 2017 Nobel Prize in Physiology or Medicine recognized this work, which by then had expanded far beyond fruit flies to encompass the molecular clockwork present in virtually every living thing on Earth. The discovery wasn't that organisms have circadian rhythms — that had been known since the 18th century. The discovery was that the clock is written in DNA, transcribed into proteins, and run by a self-sustaining molecular feedback loop that you can find in a skin cell, a liver cell, a neuron, a heart muscle cell, and in each of them it keeps approximately the same time.

Here is how the core loop works. Two proteins, BMAL1 and CLOCK, form a heterodimer — they join together as a functional pair — and bind to specific regulatory sequences in DNA called E-boxes. E-boxes are found in the promoter regions of the Period genes (PER1, PER2, PER3) and the Cryptochrome genes (CRY1, CRY2). When BMAL1 and CLOCK bind these E-boxes, they drive the transcription of the PER and CRY genes, producing PER and CRY messenger RNA and ultimately PER and CRY proteins. As the day goes on, PER and CRY proteins accumulate in the cytoplasm of the cell. They eventually form heterodimeric complexes — PER/CRY pairs — that are phosphorylated by casein kinase enzymes, which targets some for degradation but allows stable complexes to translocate back into the nucleus. Once in the nucleus, the PER/CRY complex inhibits the activity of the BMAL1/CLOCK heterodimer, suppressing transcription of the PER and CRY genes. The proteins that the genes produce circle back and shut the genes down. This is the negative feedback loop. It takes approximately 24 hours to complete, from E-box activation through protein accumulation, nuclear translocation, repression, and then degradation of the repressor proteins to allow the next cycle to begin.

Secondary loops add precision and connect the clock to metabolism. REV-ERBα and REV-ERBβ are transcription factors whose own expression is driven by BMAL1/CLOCK; they then repress BMAL1 gene transcription, adding a second negative loop that reinforces the periodicity and helps maintain robustness against perturbation. ROR proteins compete with REV-ERBα for binding and support BMAL1 expression, creating a balanced positive/negative regulation of the master clock activator. These secondary loops are not peripheral ornaments on the clock — they are the molecular bridges between the core timekeeping function and the metabolic and immune processes that the clock regulates. REV-ERBα, in particular, is a regulator of lipid metabolism, bile acid synthesis, and inflammatory gene expression. When you alter the clock, you alter metabolism and immunity through these connections.

The precision of the cellular clock is maintained partly by post-translational modifications — phosphorylation, ubiquitination, sumoylation of clock proteins — that control their stability, location, and activity. Casein kinase 1 delta and epsilon phosphorylate PER proteins; the phosphorylation pattern determines whether a PER protein is degraded quickly or accumulates to feed back into the nucleus. Mutations in these kinases are the molecular basis of familial advanced sleep phase syndrome and familial delayed sleep phase syndrome — rare but real inherited conditions where the human circadian period is substantially shorter or longer than 24 hours, causing people to be compelled toward extremely early or extremely late sleep timing regardless of social schedules. These clinical conditions provided some of the earliest human genetic evidence that clock gene function is directly relevant to human sleep physiology.

The scale of circadian gene regulation across the body is difficult to overstate. Large-scale transcriptomic studies — measuring all the transcripts produced in a tissue across the 24-hour cycle — have found that between 10% and 80% of expressed genes in different tissues show significant circadian oscillation in abundance. The percentage varies by tissue: the liver, which is a metabolically active organ with strong circadian demands, shows among the highest percentages of rhythmic transcripts. The brain, heart, muscle, gut, skin, and immune cells all show substantial fractions of circadian transcription. Nearly all of the most important drug targets in pharmacology — the enzymes and receptors that most medications act on — are clock-controlled in at least some tissues. This is the molecular substrate for chronopharmacology: the same drug, given at different times of day, encounters different levels of its target, different metabolic enzyme activity, different transport protein expression. Time is a pharmacological variable.

The cancer connection is one of the most striking implications of this biological reality. Cell division is clock-controlled — the machinery that governs DNA replication and mitosis interacts directly with clock proteins, and cells are more likely to replicate at certain times of day than others. Disruption of circadian rhythms is associated with increased cancer risk in epidemiological studies of shift workers; the International Agency for Research on Cancer classifies shift work involving circadian disruption as a Group 2A probable carcinogen. Mechanistically, clock disruption may weaken the temporal gating of cell division that limits the proliferative window in which errors can accumulate. Chronotherapy — timing cancer drug administration to the phase of the circadian cycle when tumor cells are most vulnerable and normal cells are most resistant — is an active research area with some striking preclinical results and more modest but real human clinical data. Several chemotherapy agents, including oxaliplatin and irinotecan, have documented differential toxicity and efficacy depending on administration time in clinical trials.

Clock gene variants in human populations create the biological basis for chronotype — the morning-versus-evening orientation that most people regard as personality but is substantially genetic. PER3 has a variable-number tandem repeat polymorphism (the gene comes in 4-repeat and 5-repeat versions) associated with chronotype differences and sleep architecture differences: the longer 5-repeat version is associated with earlier sleep timing and more slow-wave sleep. CLOCK has a SNP in its 3' untranslated region associated with evening preference and, in some studies, susceptibility to mood disorders. These are probabilistic associations, not deterministic ones — the genetic variants shift distributions, not individual outcomes — but they establish that chronotype is a trait with molecular underpinnings, not simply a behavioral choice.

The aging of the molecular clock happens at multiple levels. Clock gene expression amplitude declines with age in peripheral tissues, as measured in post-mortem brain tissue studies and in accessible blood cells and skin biopsies. The BMAL1/CLOCK transcriptional activation of PER and CRY becomes less robust — the peaks are lower, the troughs are higher, the oscillation is dampened. In the SCN specifically, the electrical coupling between clock neurons that normally amplifies and stabilizes the master clock output may weaken, reducing the robustness of the central timing signal. Peripheral clocks that should be tightly synchronized drift gradually out of phase with each other and with the SCN. The whole temporal architecture of the body becomes less well-organized.

The metabolic consequences of clock decline with aging reinforce each other. Clock gene disruption impairs glucose metabolism — BMAL1 knockout mice develop severe metabolic syndrome, and circadian disruption in humans is associated with insulin resistance and type 2 diabetes risk. Because clock-regulated metabolic function normally improves glucose handling during the biologically active period, its decline means that age-related metabolic deterioration and clock deterioration are not independent processes; they are partially the same process. Inflammatory regulation loses its circadian structure, contributing to inflammaging. DNA repair pathways are clock-controlled, and their temporal organization may matter for the accumulation of the DNA damage that underlies cellular senescence and cancer risk with aging.

The interventions that engage directly with clock gene biology are, once you understand the molecular mechanism, obvious in their logic. Time-restricted feeding is the most powerful feeding-based synchronizer of peripheral clocks, particularly hepatic and metabolic clocks, because the liver clock responds strongly to feeding cues. Consistent meal timing reinforces peripheral clock gene expression rhythms in ways that caloric restriction without timing consistency does not. Animal studies have shown that time-restricted feeding restores clock gene amplitude in aged mice, and preliminary human data suggest improvements in metabolic parameters that may partly reflect clock resynchronization.

Light exposure timing works at the level of the SCN through melanopsin-containing retinal ganglion cells. The molecular mechanism involves light-induced transcription of the PER genes in the SCN — the same negative feedback loop, but with light providing an acute stimulus that overrides the current phase of the clock and resets it. Morning light advances the clock (shifts it earlier); late-night light delays the clock (shifts it later). The directional effect depends on where in the circadian cycle the light hits, which is why the timing of light exposure is as important as the intensity.

Rapamycin — the mTOR inhibitor with well-documented lifespan extension in animal models — has been found to interact with clock function in some studies, with mTOR signaling affecting PER translation and clock period. This is a tentative mechanistic connection between one of the most promising longevity pharmacology approaches and the circadian system; it does not tell us that rapamycin's longevity effects are circadian in mechanism, but it suggests the two systems are not independent.

Where the Khavinson-tradition peptides intersect with circadian biology: epitalon, a tetrapeptide derived from the pineal gland and researched extensively in Russian literature, has been studied for effects on melatonin secretion and circadian regulation in aging animals and humans. Some of the Russian clinical data suggests that epitalon may support pineal function and circadian amplitude in older populations. The evidence base exists primarily outside Western peer-reviewed journals, limiting independent replication and scrutiny, and the compound is not FDA-approved. The biological hypothesis — that restoring pineal signaling strength might support SCN entrainment and downstream circadian amplitude — is mechanistically coherent, even if the evidence is insufficient by standard clinical criteria.

What the molecular biology of clock genes ultimately teaches is something counterintuitive about complexity. The most intricate regulatory architecture in human biology — the coordination of metabolism, immunity, cell division, repair, and behavior across the 24-hour cycle — rests on a feedback loop that, at its core, is simple. A gene makes a protein. The protein comes back and shuts the gene off. Repeat. That loop, plus the secondary loops that connect it to metabolism and immunity, plus the hierarchy that synchronizes trillions of cells to a common master clock, plus the entrainment by light that keeps it aligned with the external world — this architecture generates every circadian phenomenon in the human body.

When it runs well, it is invisible. The liver is ready for breakfast before you sit down. The immune system shifts from surveillance to repair at night without being asked. The brain clears its metabolic waste in the slow-wave window allocated for it. The timing is built in. When it runs poorly — through aging, through light disruption, through feeding dysregulation, through shift work — the invisibility becomes visible as fatigue, inflammation, metabolic dysfunction, cognitive decline, and accumulated cellular damage. Time-of-day biology is not a detail. It is the scaffolding on which everything else runs, and the scaffolding is among the first things to weaken as the decades pass.