DnaK Therapeutics
01 / The scienceDnaK, p53 and chemoresistance

How one bacterial protein blunts chemotherapy.

Everything below is drawn from ten peer-reviewed papers published between 2018 and 2026. Each claim carries the citation it rests on.

02 / Where we are focusedACS 2026 · IARC GLOBOCAN 2024

The program targets the aerodigestive cancers: head and neck, esophagus, stomach, colon and rectum. It also targets prostate. In each of these the bacteria and the chemotherapy rationale meet. Why these cancers →

The aerodigestive and prostate cancers the program targets
Cancer type Bacteria detected or reported US, 2026 Worldwide
Aerodigestive
Head and neckOral cavity, pharynx and larynxF. nucleatum72,770965,243
EsophagusF. nucleatum22,530494,320
StomachH. pylori is an IARC Group 1 carcinogenH. pylori, M. fermentans, F. nucleatum31,510980,286
Colon and rectumF. nucleatum, M. fermentans158,8501,984,605
Other
ProstateReported association, no tissue data of our ownMycoplasma spp.333,8301,546,112
Total619,4905,970,566
Share of all cancers diagnosed29%29%

US figures are American Cancer Society estimates of new cases for 2026 (all cancers, 2,114,850). Worldwide figures are IARC GLOBOCAN 2024 estimates of new cases (all cancers, 20,636,441). Head and neck is the sum of the oral cavity, pharynx and larynx sites; colon and rectum is the sum of the two separate sites. Naming a bacterium here means it has been detected in, or reported in association with, that tumor type. It does not mean bacterial DnaK has been shown to cause the cancer, and it does not mean every patient with that diagnosis carries the bacterium.

03 / How it worksCited to the published record

Chaperones

A newly made protein is a limp chain of amino acids. It only works once it folds into the right shape. Chaperone proteins do that folding, and refold proteins that have come apart under stress. DnaK is the chaperone of the Hsp70 family found in bacteria, and nearly every bacterium carries one.

Bacterial DnaK does not stay inside the bacterium. It is secreted into the surrounding tissue, gets taken up by neighboring human cells, and once inside it binds human proteins it was never meant to meet.[6] A 2026 structural analysis found that DnaK and its human binding partners are unusually disordered proteins, which is what lets one bacterial chaperone engage so many different human pathways.[10]

The three bacteria

Three genera of tumor-associated bacteria matter here: Fusobacterium nucleatum, Helicobacter pylori, and several Mycoplasma species. Between Mycoplasma and Fusobacterium the DnaK amino-acid sequence is nearly identical, and both are clearly distinct from the DnaK of bacteria not associated with cancer.[1] H. pylori DnaK is similar to those two, though not near-identical. A comparable case can be made for it, and we hold it at a lower level of confidence. That shared structure is why a single inhibitor is a plausible goal instead of three separate programs.

The three differ in how firmly they are tied to cancer. H. pylori is the clearest case: the International Agency for Research on Cancer classifies it as a Group 1 carcinogen, meaning there is sufficient evidence that it causes cancer in humans, specifically gastric cancer. F. nucleatum and Mycoplasma species are strongly associated with several tumor types in the published literature, and the mechanisms by which Mycoplasma species drive inflammation and cellular transformation have been reviewed in detail.[3] Association is still not the same as causation. It is the word the evidence supports, so it is the word we use.

What DnaK does

The founding paper showed that Mycoplasma promotes malignant transformation in living animals, and that its DnaK alone has broad oncogenic properties.[1] Follow-up work identified DnaK as the specific bacterial component responsible for that transformation.[2] Three activities have been characterized:

  • It reduces p53. p53 is the protein that halts a damaged cell before it divides. Bacterial DnaK binds proteins in the p53 pathway and lowers p53 levels and activity.[1]
  • It impairs DNA repair. Interactome mapping in cancer cells showed DnaK binding human proteins that control DNA repair and chromatin organization, including XRCC1 and USP10, and interfering with those pathways.[6]
  • It causes genetic instability on its own. In mice, expressing Mycoplasma DnaK increased DNA copy number variation.[7] In a further in-vivo study, DnaK expression accelerated cancer development after DNA damage.[8]

DnaK introduced from outside a cell also changes protein-kinase signaling inside human cancer cell lines,[4] and behaves as expected once inside infected human colon carcinoma cells.[5]

Why chemotherapy breaks

Several of the most widely used chemotherapy drugs work by damaging DNA or blocking its synthesis, then relying on the cell's own machinery, p53 prominently, to recognize the damage and trigger cell death. Cisplatin and 5-fluorouracil are examples of drugs that depend on that signaling, and they are the two our published work has tested.

If p53 is suppressed, the damage is still done but the response is muted. A 2024 paper tested this directly and found that bacterial DnaK reduces the activity of cisplatin and 5-FU.[9] The company is built on that finding: chemoresistance with a cause the tumor's genome does not encode, which can therefore be addressed without changing the chemotherapy itself.

Reversing it

If bacterial DnaK is what suppresses p53, then inhibiting DnaK should restore it. Our therapeutic program is a set of selective small molecules designed to do exactly that, given alongside the chemotherapy a patient is already receiving, not in place of it.

The published proof of concept used ARV-1502, a DnaK-binding peptide. In cancer cell lines exposed to M. fermentans or F. nucleatum DnaK, and in primary tumor cells from a DnaK knock-in mouse model, ARV-1502 restored the anti-cancer activity of cisplatin and 5-FU.[9] Those are cell-culture and animal results. ARV-1502 is a research tool, not our clinical candidate. Selecting one is described on the pipeline page.

Two consequences follow from the combination design. The claim is narrower, which makes for a cleaner regulatory path than a replacement therapy. And it fits how oncologists already work, since nothing about the existing regimen has to change.

The target is a bacterial protein, not a human one, so the therapeutic only helps patients who actually carry it. The companion diagnostic is not an afterthought. It is what makes the therapeutic targetable.

Why these cancers

The bacteria are not hypothetical here. Mining human cancer sequencing data from The Cancer Genome Atlas, F. nucleatum was found to be particularly present in cancers of the gastrointestinal tract, specifically head and neck, esophageal, colon and rectal tumors.[9] Our team then detected both Mycoplasma and Fusobacterium DnaK by quantitative PCR in primary stomach and colon adenocarcinoma samples, at varying copy number.[9] Prostate is included on the Mycoplasma association reported in the literature. We have no tissue data of our own for it, and we would rather say so than blur the point.

These bacteria have been reported in a wider set of tumor types still, and the DnaK mechanism is not in principle limited to this list. Later programs can take that up. Today the work is the list above.

References

All ten papers, with abstracts and DOI links, are listed in the newsroom. Numbered references used above:

  1. Zella D, et al. Mycoplasma promotes malignant transformation in vivo, and its DnaK, a bacterial chaperone protein, has broad oncogenic properties. PNAS, December 2018. 10.1073/pnas.1815660115
  2. Benedetti F, et al. Role of Mycoplasma Chaperone DnaK in Cellular Transformation. Int J Mol Sci, February 2020. 10.3390/ijms21041311
  3. Benedetti F, et al. Mycoplasmas–Host Interaction: Mechanisms of Inflammation and Association with Cellular Transformation. Microorganisms, September 2020. 10.3390/microorganisms8091351
  4. Benedetti F, et al. Exogenous bacterial DnaK increases protein kinases activity in human cancer cell lines. J Transl Med, February 2021. 10.1186/s12967-021-02734-4
  5. Curreli S, et al. Analysis of DnaK Expression from a Strain of Mycoplasma fermentans in Infected HCT116 Human Colon Carcinoma Cells. Int J Mol Sci, April 2021. 10.3390/ijms22083885
  6. Curreli S, et al. Characterization of the interactome profiling of Mycoplasma fermentans DnaK in cancer cells reveals interference with key cellular pathways. Front Microbiol, October 2022. 10.3389/fmicb.2022.1022704
  7. Benedetti F, et al. Mycoplasma DnaK increases DNA copy number variants in vivo. PNAS, July 2023. 10.1073/pnas.2219897120
  8. Benedetti F, et al. Mycoplasma DnaK expression increases cancer development in vivo upon DNA damage. PNAS, February 2024. 10.1073/pnas.2320859121
  9. Benedetti F, et al. Bacterial DnaK reduces the activity of anti-cancer drugs cisplatin and 5FU. J Transl Med, March 2024. 10.1186/s12967-024-05078-x
  10. Benedetti F, Rahman T, Uversky VN, et al. DnaK unmasked: potential contributions of intrinsic disorder to the hijacking of human proteostasis by a bacterial chaperone. Int J Biol Macromol, June 2026. 10.1016/j.ijbiomac.2026.153019

Every claim above traces to a published paper.