Founder's role text (READ): "Sense DNA damage and replication stress and enforce cell-cycle checkpoints to allow repair." Founder's key components (READ): ATM, ATR, CHEK1, CHEK2, WEE1, DNA-PK. In the recommender: No.
DRY
DRY: none located. Queries: the three dry-fast queries, plus "intermittent fasting" OR "time-restricted eating" ATM ATR CHK1 CHK2 checkpoint kinase human trial and fasting ATM activation AMPK mouse in vivo starvation DNA damage response liver.
THE SEPARATION THE FOUNDER ASKED FOR, STATED EXPLICITLY
ATM has a large and real metabolic and oxidative-stress literature that has nothing to do with fasting. That literature is held separate here and contributes no fasting row. Located, none opened, all retrieval debts:
- "ATM kinase: Much more than a DNA damage responsive protein." DNA Repair, record S1568786415301221.
- "Ataxia Telangiectasia Mutated Protein Kinase: A Potential Master Puppeteer of Oxidative Stress-Induced Metabolic Recycling." PMCID PMC8032526.
- "ATM directs DNA damage responses and proteostasis via genetically separable pathways." Science Signaling, DOI 10.1126/scisignal.aan5598.
- "Metabolic Stress and Mitochondrial Dysfunction in Ataxia-Telangiectasia." DOI 10.3390/antiox11040653.
- "ATM Protein Kinase: Old and New Implications in Neuronal Pathways and Brain Circuitry." DOI 10.3390/cells9091969.
- "The emerging relationship between metabolism and DNA repair." Cell Cycle, DOI 10.1080/15384101.2021.1912889.
None of these is a fasting intervention. None of them may be used to support a fasting row on this node. They establish that ATM is a redox and metabolic kinase, which is a reason to be more careful here, not less.
Disconfirming evidence, first
- 26.D1. In NORMAL human cells, nutrient withdrawal does not activate ATM, and it suppresses ATM activation by a genotoxin. This is the central disconfirming result for the node and it comes from the same paper as the positive. Shi Y, Felley-Bosco E, Marti TM, Orlowski K, Pruschy M, et al., Stahel RA (corresponding). BMC Cancer 2012;12:571. DOI 10.1186/1471-2407-12-571. Published 4 December 2012. (Full author list truncated on the opened page: retrieval debt.) MEASURED in SDM104, primary human mesothelial cells: starvation alone did not activate ATM. It suppressed cisplatin-induced ATM activation. It did increase p53 and p21 through AMPK, since compound C blocked that effect. MEASURED in A549 human lung adenocarcinoma: starvation induced Chk2 activation, p53 accumulation and gamma-H2AX, but ATM and AMPK activation were not detected (reported as data not shown, which is itself a reporting debt). So the ATM response to nutrient withdrawal is cell-type dependent, and in the one normal human primary cell line tested it was absent or inverted.
- 26.D2. The dedicated 2026 review does not discuss checkpoint kinases for CR, DR or fasting. van Galen et al. 2026 (Nutrients 18(2):246), READ: "Not discussed directly." The related readouts it does report are reduced gamma-H2AX foci and lower damage-inducible transcripts (Bax, Bad, Casp9, Casp11, Gadd45a). None of those is a checkpoint kinase. gamma-H2AX is an ATM and ATR substrate, so citing it as checkpoint-kinase evidence is a one-hop inference and must be tagged M.
- 26.D3. The in vivo fast in the only fasting paper on this node did not measure the node. From the opened Shi 2012: nude mice bearing ZL55 or A549 xenografts fasted with water ad libitum, 32 hours before and 16 hours after weekly cisplatin (3 mg/kg i.p.) for three weeks. The in vivo arm measured tumor growth and regression, not ATM, CHK2 or p53 signaling. So a whole-animal fast exists in this paper and the checkpoint readout does not.
Graded rows
| Row | Modality | Species / system | Tier | Grade | What was actually measured | Direction |
|---|---|---|---|---|---|---|
| 26.1 | serum withdrawal, 24 h | Human tumor cell lines ZL55, A549, HCT116 | C1 | C | ATM-pS1981, pATM, pChk2-T68, p53, gamma-H2AX | Up in tumor lines |
| 26.2 | serum withdrawal, 24 h | Normal primary human mesothelial SDM104 | C1 | C | ATM activation | No activation; ATM suppressed under cisplatin |
| 26.3 | PF, 32 h pre and 16 h post, water ad libitum | Mouse xenograft | A1 for the intervention, not for this node | C | Tumor growth only. No checkpoint kinase measured. | Node not measured |
| 26.4 | CR / DR, 30% | Mouse, Ercc1 delta/- liver | M | C | gamma-H2AX foci. One hop from ATM/ATR. | Damage down |
26.1 and 26.2 numbers MEASURED from the opened paper: in ZL55 human mesothelioma, serum starvation alone raised ATM-pS1981-positive nuclei about fivefold and pATM by about 40% on Western blot; it increased pChk2-T68 and p53 protein by about 50%, without changing p53-Ser15 phosphorylation. Combined with cisplatin, pATM rose further (P < 0.001), exceeding 90% positive nuclei. The ATM inhibitor Ku-55933 blocked starvation-induced p53 accumulation in ZL55 and HCT116. In HCT116, the cisplatin sensitization by starvation was lost in p53-/- cells. Starvation mode MEASURED: cells moved to M199/MCDB106 medium without supplements; flow cytometry used 24 hours of starvation; for cisplatin experiments, 8 micromolar cisplatin for 16 hours starting 8 hours after starvation began. Glucose and amino acid starvation were not tested in this study.
26.4 source: Vermeij et al. 2016 (Nature 537:427-431), cited in full under node 22. Tagged M, not A1, for this node, because gamma-H2AX is a substrate of these kinases and not a measurement of the kinases themselves. The hop is shown.
Open questions
1. Zero human fasting evidence. ATM, ATR, CHEK1, CHEK2, WEE1 and DNA-PK have never been measured in a human fasting study located by this sweep. (DNA-PK was measured in mouse skin under 4-week 40% CR: see row 23.1. That is CR, in mouse, and it belongs to node 23.) 2. Direction is cell-type dependent and unresolved: up in tumor lines, absent or inverted in normal human primary cells. Both sides come from the same paper, so there is no independent replication of either side. 3. Shi 2012 reports the A549 ATM and AMPK null as "data not shown." That is a reporting debt on the disconfirming side of the record, and it weakens the disconfirming row as well as the supporting one.
← 25. Direct reversal (MGMT) All 41 nodes 27. Telomere maintenance / telomerase →
dryfasting.com has been published since 2014. This evidence record was compiled on 22 November 2026. Every row names the fasting modality the study actually ran, the system it was measured in and the grade. Where no source carries a DOI or PMID the row says so rather than supplying one. Nothing is sold on this site.