The team started with an optogenetic trick. They engineered non-excitable HeLa cells to express melanopsin, a light-sensitive, G-protein-coupled receptor that humans normally use in intrinsically photosensitive retinal ganglion cells to help regulate circadian rhythms. When blue light at 470 nanometers strikes melanopsin, the receptor triggers a cascade that releases calcium from the endoplasmic reticulum through inositol trisphosphate receptors. By wiring computer-controlled light-emitting diodes into the cell incubator, the researchers could force the cells to experience calcium surges on demand, once every minute or once every two minutes, for minutes or hours at a stretch. Cells transduced with an empty vector ignored the light entirely, confirming that the responses depended on the introduced receptor.

The system proved remarkably reliable. In a stable cell line selected for high melanopsin expression, roughly 40 percent of cells responded to every single light pulse delivered over twenty minutes of repeated stimulation, and the light-evoked calcium peaks, lasting around twenty to forty seconds, rose synchronously across the population. Twelve hours of pulsing every minute did not harm the cells: their morphology was unchanged and the early apoptotic marker Annexin V did not increase. Because melanopsin needs its vitamin A-derived cofactor retinal to function, some experiments added all-trans retinal to the culture medium, and the strength of the calcium response tracked how much melanopsin each cell expressed.

With the hardware in place, the researchers first tested a known calcium-frequency decoder: the transcription factor NF-κB, which shuttles in and out of the nucleus and is thought to respond preferentially to low-frequency calcium oscillations in the millihertz range. A luciferase reporter assay showed that NF-κB-associated transcriptional activity rose with stimulation frequency, with pulses every minute producing stronger reporter output than pulses every other minute over a one-hour window. The light-driven activation reached a similar order of magnitude to that provoked by acute lipopolysaccharide, a potent inflammatory stimulus, suggesting the optogenetic route engages the pathway substantially.

Zooming in on individual genes revealed something subtler. Using quantitative PCR, the team measured a panel of NF-κB-dependent genes and found that TNF, encoding tumor necrosis factor, was significantly increased only under the fast, once-per-minute regime, while IL8, encoding interleukin-8, responded to both fast and slow stimulation. Pharmacology pointed to partially divergent wiring: the IKKβ inhibitor SC-514 preferentially blunted TNF induction, whereas the IκBα ubiquitination inhibitor Ro 106-9920 more clearly suppressed IL8, and the InsP3 receptor inhibitor 2-APB reduced both, consistent with calcium-dependent signaling upstream. Intriguingly, when stimulation stretched to twelve hours, neither gene showed significant induction, hinting that the response is transient or adapts over longer horizons.

Timing mattered in a second, unexpected way. When the researchers scrambled the schedule, delivering 55 light pulses irregularly across one hour instead of at fixed intervals, gene induction dropped significantly for both TNF and IL8 compared with regular stimulation, even though the total number of pulses was similar. In a matched-pulse comparison, thirty pulses delivered every two minutes for an hour produced significantly lower TNF induction than thirty pulses delivered every minute for half an hour followed by half an hour of darkness, while IL8 again showed no significant difference between the two schedules. The authors are careful about interpretation: because intracellular calcium waveforms were not directly quantified under these paradigms, the experiments do not isolate periodicity itself as the causal variable, but they do demonstrate that downstream gene responses differ between distinct temporal patterns of melanopsin activation.

To see the whole transcriptomic landscape rather than a handful of genes, the team ran bulk RNA sequencing on cells stimulated for one or twelve hours at the fast or slow frequency, controlling for both light exposure and the transgene itself. After one hour, the fast regime produced 215 differentially expressed genes relative to the slow regime, 131 upregulated and 84 downregulated. After twelve hours, only 27 such genes emerged, and none overlapped with the one-hour set, marking clearly distinct transcriptional endpoints at the two time points. An interaction-model analysis identified 78 genes whose fast-versus-slow response depended on melanopsin expression, 23 of which overlapped the broader set, flagging them as higher-confidence candidates for melanopsin-dependent temporal decoding. Enrichment analysis of the one-hour genes pointed toward translation and ribosome biology, including cytoplasmic translation, ribosome biogenesis, and RNA binding.

Two independent computational lenses converged on MYC, the infamous growth-regulating transcription factor. Gene set enrichment analysis flagged the Hallmark MYC Targets V1 and V2 pathways among the one-hour responses, alongside epithelial-mesenchymal transition, inflammatory response, and mTORC1 signaling, while transcription factor enrichment analysis ranked MYC among the top predicted regulators, associated with 116 of the 215 differentially expressed genes. Notably, MYC messenger RNA itself did not change between fast and slow stimulation, raising the possibility that oscillatory signaling modulates MYC-associated programs through post-transcriptional mechanisms rather than by altering MYC expression. At twelve hours, the picture shifted: cholesterol homeostasis and apical junction pathways dominated, and HIF1A replaced MYC as the most connected predicted regulator.

Phosphoproteomics added a third layer. Mass spectrometry after one hour of stimulation detected 641 differentially abundant proteins between the fast and slow regimes, 193 up and 448 down, and 211 differentially abundant phosphopeptides, 109 up and 102 down, corresponding to roughly 90 unique proteins on each side. Kinase enrichment analysis of the phosphoproteins surfaced a striking cluster: CSNK2A1, CHEK2, ATR, ATM, PRKDC, and other masters of the DNA damage response and cell-cycle checkpoint control. When the researchers independently predicted kinases from the transcription-factor interactor network, five of the top ten overlapped with the phosphoproteomic list: PRKDC, CHEK2, ATM, ATR, and AURKA. The authors stress that these networks represent candidate regulatory relationships inferred computationally, not direct evidence of kinase activation, and that limited separation in principal-component analysis and missing-value imputation make the phosphoproteomic findings hypothesis-generating rather than definitive.

The convergence is biologically suggestive. ATM is canonically activated by DNA double-strand breaks and phosphorylates targets including CHEK2 to restrain cell-cycle progression or trigger apoptosis, while PRKDC participates in repairing such breaks, and both have documented ties to MYC biology: ATM counteracts genomic instability in MYC-overexpressing cancers, and PRKDC inhibition is lethal to MYC-overexpressing human lung fibroblasts. Although these kinases are not classical calcium-dependent enzymes, emerging evidence links ATM activity to calcium handling, including mitochondrial calcium uptake. The study’s authors frame their results as temporal decoding of melanopsin-evoked signaling within an intact Gq-coupled network, not as proof that calcium oscillation frequency is decoded in isolation, since melanopsin also engages phospholipase C, protein kinase C, and potentially MEK/ERK branches. HeLa cells, a carcinoma-derived line, may also deploy calcium machinery differently from neurons or developing tissue, leaving generalizability open. Even with those caveats, the demonstration that the same light-driven receptor, ticking at different rhythms, reshapes inflammatory genes, MYC-linked programs, and stress-response phosphorylation gives researchers a powerful new handle on one of biology’s oldest questions: how cells read the rhythm, not just the message.

Subject of Research: Optogenetic control of melanopsin-evoked calcium signaling patterns and their downstream transcriptional and phosphoproteomic responses in HeLa cells

Article Title: Temporal patterns of melanopsin-evoked calcium signaling engage NF-κB-associated transcriptional and phosphoproteomic responses

Article References: Nikpour, P., Varas-Godoy, M., Tenghamn, J., Uhlén, P., & Smedler, E. (2026). Temporal patterns of melanopsin-evoked calcium signaling engage NF-κB-associated transcriptional and phosphoproteomic responses. iScience, 29(11), Article 118031. https://doi.org/10.1016/j.isci.2026.118031

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.118031

Keywords: calcium signaling, optogenetics, melanopsin, NF-κB, TNF, IL8, MYC, phosphoproteomics, RNA sequencing, HeLa cells, DNA damage response, frequency decoding