Oncology | Tumor Microbiome | Published in Signal Transduction and Targeted Therapy, April 7, 2026
University of Illinois Chicago
For most of the history of cancer medicine, the tumor was the enemy. A mass of mutant cells, proliferating out of control, evading the immune system, consuming resources, spreading. The goal of treatment was always the same: find the tumor and destroy it.
What scientists have discovered over the past two decades is that the tumor is not alone. It is an ecosystem. Inside and around it live bacteria, fungi, viruses, and immune cells. Some of these organisms make things worse. Some appear to be neutral. And a small but growing body of evidence suggests that some, very counterintuitively, make things better.
The big idea
Tumors are not just cancer cells. They are microbial ecosystems, and some bacteria inside them may carry molecules that can be turned against cancer.
Tohru Yamada has spent over two decades of his scientific life at the University of Illinois Chicago asking a question that most of his colleagues once thought was absurd: what if the bacteria living inside tumors are not passengers, but potential allies, carrying molecular tools that cancer cells cannot resist?
A paper published in Signal Transduction and Targeted Therapy on April 7, 2026 represents the furthest he has yet gone in answering that question. The result is a new compound called aurB, a synthetic peptide designed from a bacterial protein found in photosynthetic bacteria living inside tumors. And unlike most of what came before it in this unusual field of research, aurB does not need a functioning copy of the tumor suppressor gene p53 to work.
That distinction, which may sound like a technical detail, is the difference between a therapy that works in some cancers and one that could work in many.
Coley’s Toxins and a Century-Old Intuition
The idea that bacteria and cancer interact in ways that might be therapeutically useful is not new. It is, in fact, over a hundred years old, which makes its persistent neglect by mainstream oncology one of the more interesting episodes in the history of medicine.
In the late 1880s, William Coley, a New York surgeon, noticed something that did not fit the expected pattern. Some of his patients with inoperable bone tumors who had contracted severe bacterial infections, particularly erysipelas, appeared to have their tumors shrink during or after the infection. It was an observation that every reasonable physician of the time would have been expected to dismiss. Instead, Coley pursued it.
He began deliberately infecting cancer patients with bacteria, and then, when that proved too dangerous, with heat-killed bacterial preparations that came to be known as Coley’s Toxins. The results were inconsistent and controversial, partly because Coley’s methods were not standardized and partly because the immune mechanisms behind his observations were not understood at the time. His work was sidelined as radiation and chemotherapy rose to dominance in the mid-twentieth century.
But Coley was, in some fundamental sense, right. His clinical observations were real. The mechanisms he could not explain, we now understand in outline: bacterial infection activates the innate immune system, and that activation can spill over into the tumor microenvironment, triggering immune responses that attack cancer cells alongside the invading bacteria. Modern cancer immunotherapy, particularly BCG treatment for bladder cancer, in which a weakened strain of Mycobacterium tuberculosis is instilled directly into the bladder, is a direct descendant of Coley’s intuition. So is the entire field of cancer immunotherapy, which Coley’s contemporary, more mechanistically literate, admirers acknowledge when they call him the father of immunotherapy.
Old idea, new precision
Coley used bacteria to trigger anti-tumor effects. Yamada’s work asks whether a single bacterial protein can deliver the benefit without the infection risk.
Using live bacteria against cancer is an old idea, Yamada said. More than a century ago, surgeon W. Coley treated patients who had inoperable sarcoma with live bacteria and found anti-tumor effects. Coley’s finding still influences the field of cancer immunotherapy today.
The problem with live bacteria, as a therapeutic strategy, is obvious: you are introducing a potentially dangerous pathogen into an already sick patient. What Yamada’s research has spent two decades working toward is something more precise. Not live bacteria. Not even killed bacteria. A single protein, purified and synthesized in the laboratory, that carries the cancer-fighting activity without the infection risk.
Azurin: The First Bacterial Anti-Cancer Protein
Yamada’s story with bacterial cancer proteins begins not with the study published in 2026, but with a paper published in the Proceedings of the National Academy of Sciences in 2002. In that paper, a team led by Yamada and his longtime collaborator Tapas K. Das Gupta at UIC reported something remarkable.
Azurin, a purified bacterial redox protein from Pseudomonas aeruginosa, enters human cancer cells and induces apoptosis. The induction of apoptosis occurs readily in melanoma cells harboring a functional tumor suppressor protein p53, but much less efficiently in p53-null mutant melanoma cells.
Azurin is a cupredoxin, a member of a family of small, water-soluble, copper-containing proteins whose normal job in bacteria is to shuttle electrons between other proteins in the bacterial energy-transfer chain. In Pseudomonas aeruginosa, an opportunistic pathogen best known as a cause of hospital-acquired infections in immunocompromised patients, azurin helps the bacterium generate energy. It has been evolved for bacteria. It has nothing, in theory, to do with human cancer biology.
Except that it does. Azurin is a multi-target anticancer agent that preferentially enters human cancer cells and induces apoptosis or growth inhibition, interfering in the p53 signaling pathway and the non-receptor tyrosine kinases signaling pathway.
How Yamada and his colleagues discovered this, and why azurin preferentially enters cancer cells rather than healthy ones, is a story about the peculiar molecular architecture of cancer cell membranes, which differ from normal cell membranes in ways that make them more permeable to amphipathic peptides. But the core finding was startling enough on its own: a bacterial protein from a lung pathogen could enter cancer cells, bind to the tumor suppressor protein p53, stabilize it against degradation, and push the cancer cell toward programmed self-destruction.
The first breakthrough
Azurin showed that a bacterial protein could enter cancer cells and push them toward self-destruction. But it depended heavily on functional p53.
Over the following years, the team developed a 28-amino acid peptide derived from the azurin sequence, called p28, that retained the anticancer activity in a smaller, more drug-like form. P28, an azurin-derived peptide designated NSC745104, induced p53-dependent tumor growth inhibition in animal models of a broad range of cancer types and showed preliminary efficacy in two Phase I clinical trials without apparent toxicity in patients with advanced solid tumors and pediatric patients with recurrent and refractory central nervous system tumors.
Phase I clinical trials. Real patients. No apparent toxicity. Preliminary efficacy signals. This was, by the standards of first-in-human studies for an entirely novel class of compound, a promising result.
But it had a fundamental limitation built into its mechanism. P28 worked by stabilizing p53. That means it required p53 to be present and capable of doing its job. The induction of apoptosis occurs readily in cancer cells harboring a functional tumor suppressor protein p53, but much less efficiently in p53-null mutant cells.
The limitation
More than 50% of human tumors carry p53 mutations. A therapy that needs working p53 may fail in many of the hardest cancers.
And p53 is the most commonly mutated gene in human cancer. More than 50% of all human tumors carry p53 mutations. In many cancers, including the castration-resistant prostate cancer that would become the focus of the 2026 study, p53 dysfunction is a defining feature of the most aggressive, treatment-resistant disease. A therapy that requires working p53 is a therapy that will fail in exactly the patients who need it most.
Yamada knew this. And it set the direction for everything that followed.
The Tumor Microbiome: A New Frontier
While Yamada continued work on p28, a parallel scientific revolution was unfolding that would eventually converge with his research.
For decades, tumors were assumed to be sterile. The received wisdom was that a healthy immune system would destroy any bacteria that managed to enter the tumor environment, and that whatever microbial contamination appeared in tumor samples was the result of sampling artifacts. A tumor, by definition, was a place bacteria could not survive.
That assumption was methodologically convenient and entirely wrong.
Beginning in the 2010s, advanced DNA sequencing technologies, particularly 16S ribosomal RNA sequencing that can identify bacterial species from tiny quantities of DNA without requiring the bacteria to be cultured, began revealing that tumors across multiple cancer types were colonized by diverse bacterial communities. Pancreatic tumors contained Fusobacterium. Colon tumors had specific microbial signatures that predicted prognosis. Breast tumors had their own distinct bacterial inhabitants. The pattern was consistent enough that researchers began referring to the tumor microbiome as a distinct biological phenomenon.
This raised a new set of questions. Where did these bacteria come from? What were they doing there? Were they affecting tumor growth, drug resistance, immune evasion, or patient survival? And, as Yamada’s work had suggested, were any of them producing proteins that might be exploitable therapeutically?
Yamada is also researching other bacteria-cancer interactions, including the makeup of the tumor microbiome, the bacterial species that live in the tumor microenvironment. The National Cancer Institute recently funded that project through an R01 grant. Targeting the bacteria unique to tumors, as compared to those in other parts of the body, may generate new approaches to treating cancer.
With NCI funding and the accumulated knowledge of two decades of cupredoxin research, Yamada’s team turned their attention to the tumor microbiome as a source of new drug leads. Specifically, they were looking for cupredoxin proteins in tumor-associated bacteria that might have anticancer activity through a pathway that did not depend on p53.
Auracyanin: The Protein From Photosynthetic Bacteria Inside Tumors
The discovery that led to aurB began with a sequencing experiment.
The team analyzed tumor samples from breast cancer patients and used DNA sequencing to identify the bacteria present. Photosynthetic bacteria from the phylum Chloroflexota, including a member of the class Chloroflexia, identified in tumors, carry the cupredoxin auracyanin gene.
The finding was immediately unusual for two reasons.
First, Chloroflexota are photosynthetic bacteria. They are found in hot springs, shallow marine sediments, and other sunlit environments. They are not organisms you would expect to find inside a human tumor, which is both hypoxic and definitionally dark. How they got there, whether through some route involving the gut microbiome or through other mechanisms, is not yet understood. But their genetic signature, and specifically the signature of their cupredoxin gene, was present in tumor tissue.
Second, auracyanin is structurally and functionally similar to azurin in important ways. Both are cupredoxins. Both carry copper and participate in electron transfer. Both have the amphipathic molecular characteristics that allow cupredoxins to insert into and cross biological membranes. But auracyanin, as a protein from a photosynthetic rather than a respiratory bacterium, has a different active site chemistry. And crucially, when the team analyzed how it interacts with cancer cell biology, they found that its mechanism of action pointed not toward p53, but toward the mitochondria.
The key pivot
Auracyanin did not point toward p53. It pointed toward mitochondria, the energy infrastructure of cancer cells.
The mitochondrial connection would prove to be everything.
Why the Mitochondria Are Both a Problem and an Opportunity in Cancer
To understand why targeting mitochondria differently from everything else, you need to understand something about cancer metabolism that took oncologists decades to fully appreciate.
In 1924, the German biochemist Otto Warburg made an observation that became one of the most cited and debated findings in cancer biology: cancer cells consume glucose at a dramatically higher rate than normal cells and convert most of it to lactate rather than fully oxidizing it in the mitochondria, even when oxygen is plentiful. This phenomenon, known as aerobic glycolysis or the Warburg effect, was initially interpreted to mean that cancer cell mitochondria were damaged or inactive.
That interpretation was wrong, or at least incomplete.
Cancer cells often exhibit enhanced mitochondrial ATP production, metabolic flexibility, and the ability to switch between energy sources such as glucose, glutamine, and pyruvate. In cancer, mitochondrial function is often profoundly altered, reflecting the metabolic reprogramming that supports rapid cell proliferation, survival under stress, and resistance to therapy.
Cancer cells are not using dysfunctional mitochondria. They are using profoundly adapted mitochondria, mitochondria that have been reprogrammed to serve the needs of a cell committed to uncontrolled growth. The metabolic reprogramming experienced by cancer is generally known as the Warburg effect, characterized by the sharp increase in the rates of glycolysis and of the fermentation of pyruvate into lactate, which allows the regeneration of NAD+ for glycolysis to proceed at high rates. Concurrent with this, the oxidation of pyruvate coupled to oxidative phosphorylation in mitochondria is marginally increased and partially inhibited in cancer cells, to spare the backbone of carbon skeletons for the synthesis of precursors.
What this means is that cancer cells are running both pathways simultaneously: aerobic glycolysis to feed their voracious demand for biosynthetic precursors, and enough mitochondrial activity to maintain the energy levels needed for survival. The mitochondria are not bystanders. They are essential.
Why mitochondria matter
Cancer cells are metabolically reprogrammed, but they still depend on mitochondrial energy production to survive and proliferate.
As Yamada put it directly: “The mitochondria are very important for a cell to survive; they are the energy factories. Many cancer cells exhibit altered mitochondrial number and activity, because a cancer cell has to grow aggressively and rapidly. Therefore, the mitochondria would be an ideal target for cancer therapy.”
The specific target inside the mitochondria that aurB attacks is ATP synthase.
ATP synthase is one of the most elegant molecular machines in biology. It sits embedded in the inner membrane of the mitochondrion and uses the electrochemical gradient generated by the electron transport chain to synthesize ATP, adenosine triphosphate, the universal energy currency of the cell. As protons flow through the membrane channel in ATP synthase from regions of high concentration to low, the energy of that flow is captured and used to add a phosphate group to ADP, producing ATP. Without ATP, virtually no cellular process can proceed. Muscle contraction stops. Ion pumps fail. Protein synthesis halts. The cell dies.
ATP synthase is the cornerstone molecular machine of cellular energy production. Targeting it cuts off the fundamental energetic supply of the tumor.
Cancer cells, with their elevated proliferative demands, are particularly dependent on sustained ATP production. When you target ATP synthase in cancer cells, you are not merely inconveniencing them. You are attacking the molecular infrastructure without which uncontrolled growth is impossible.
Designing aurB: The Synthetic Peptide
With auracyanin identified as the target protein and ATP synthase identified as its mechanism of action, the team at UIC designed a synthetic peptide called aurB that captures the active portion of the auracyanin protein in a small, drug-like molecule.
The design process was guided by structural analysis of auracyanin and by the earlier work on p28, which had established that a relatively short peptide could retain and concentrate the key biological activities of a full cupredoxin protein while being more manufacturable, more stable, and more amenable to pharmaceutical development.
Based on the structural and chemical characteristics of auracyanin, the team designed a novel cell-penetrating peptide, aurB. Unlike p28, which induced p53-dependent tumor growth inhibition, aurB was designed specifically to work through a p53-independent pathway.
Cell-penetrating peptides are a class of short protein sequences, typically 5 to 30 amino acids long, that can cross the cell membrane and deliver cargo into the interior of cells. They exploit specific properties of membrane lipid composition to insert and translocate, and many have been shown to preferentially enter cancer cells over normal cells because of the distinctive lipid asymmetry of cancer cell membranes. This selectivity is not absolute, but it is real and it is therapeutically relevant.
How aurB attacks cancer cells
aurB enters cancer cells, targets mitochondria, binds ATP synthase, and disrupts the energy supply needed for rapid tumor growth.
Once inside a cancer cell, aurB is designed to locate the mitochondria and bind to ATP synthase, specifically to a region of the enzyme that impairs its ability to produce ATP. With ATP production disrupted, the cancer cell’s energy budget is thrown into deficit. The processes that depend on sustained ATP availability, particularly the energetically expensive processes of rapid cell division and DNA replication that define cancer, begin to fail.
The team tested aurB in cell lines lacking active p53 and in mouse models of prostate cancer that no longer respond to hormone therapy. When combined with radiation, a standard treatment for prostate cancer, aurB significantly reduced tumor growth without clear signs of toxicity.
The Prostate Cancer Model: Why This Specific Disease
The choice of castration-resistant prostate cancer as the primary preclinical model for aurB was not arbitrary.
Prostate cancer is the second most common cancer in men worldwide. Early-stage disease is typically managed with surgery or radiation. When it recurs or spreads, hormone therapy, which reduces the production of testosterone that fuels most prostate tumors, is the standard approach.
But in a significant proportion of men, the cancer eventually stops responding to hormone therapy. This state, called castration-resistant prostate cancer or CRPC, is one of the most difficult situations in oncology. The cancer has evolved mechanisms to grow without the hormonal signal it originally depended on. By this stage, p53 dysfunction is extremely common, which means that the earlier azurin-p28 approach would have limited effectiveness.
CRPC was therefore an ideal test case for aurB, specifically because it is p53-deficient, specifically because it represents an unmet clinical need, and specifically because it responds to radiation, allowing the combination to be tested.
“The combination significantly enhanced the activity of the peptide and the tumor became much smaller. This approach is promising. Using a well-established tibial bone metastatic model, we demonstrated significant inhibition of tumor growth, preclinically.”
Why the bone metastasis model matters
Prostate cancer commonly spreads to bone. Showing activity in a tibial bone metastasis model suggests possible relevance to one of the hardest forms of late-stage prostate cancer.
The tibial bone metastatic model deserves particular attention. When prostate cancer spreads beyond the prostate, bone is one of its most common destinations. Bone metastases from prostate cancer are notoriously painful and difficult to treat, and they significantly worsen the prognosis for patients with advanced disease. Demonstrating activity in a tibial bone metastasis model is not just a preclinical milestone. It is evidence that the compound might eventually address one of the most clinically challenging aspects of late-stage prostate cancer.
A Therapy With No p53 Requirement: Why This Matters
The absence of a p53 requirement is worth dwelling on because it shapes the entire potential clinical scope of aurB.
P53 is often called the guardian of the genome. When a cell sustains DNA damage, p53 is activated and orchestrates either DNA repair or apoptosis, programmed cell death, depending on the severity of the damage. It is the cell’s most important internal checkpoint against becoming cancerous.
Because p53 is so important to cancer suppression, cancer cells are under strong evolutionary pressure to disable it. Mutations in TP53, the gene encoding p53, are found in more than 50% of all human cancers. In certain tumor types, including ovarian cancer, colorectal cancer, and the late-stage prostate cancer used in this study, p53 dysfunction rates are even higher. In some cancers, particularly those that have progressed through multiple prior treatments and have evolved resistance to conventional therapy, functional p53 is the exception rather than the rule.
Why p53 independence is huge
Many aggressive cancers disable p53. aurB targets mitochondrial energy production instead, meaning it may still work when p53 is broken.
A therapy that requires p53 to work cannot help these patients. A therapy that targets mitochondria independently of p53 can.
Unlike the previous azurin-derived cupredoxin work which induced p53-dependent tumor growth inhibition, aurB was designed specifically to work in cancers that have lost or mutated p53.
This is the conceptual pivot at the heart of the 2026 paper: taking a scientific lineage built on a p53-dependent mechanism, recognizing its limitation, and finding a new bacterial protein within the tumor microbiome itself that circumvents that limitation. The tumor’s own bacterial inhabitants pointed the way to a therapy the tumor cannot defend against with its most common resistance mechanism.
What Comes Next: Patents, Trials, and a Broader Vision
The researchers have secured a patent for aurB with support from UIC’s Office of Technology Management and are now exploring how to move the therapy into human clinical trials.
The path from a promising preclinical result to a clinical trial is measured in years and requires significant funding, regulatory engagement, and pharmaceutical manufacturing development. The first step would typically be an Investigational New Drug application to the FDA, allowing Phase I human trials to begin. Phase I trials would focus on safety, dosing, and pharmacokinetics: how the body absorbs, distributes, metabolizes, and excretes aurB.
The earlier Phase I work on p28 provides meaningful precedent. P28 went through Phase I in adults with advanced solid tumors and in children with recurrent brain tumors without apparent toxicity, which suggests that the cupredoxin peptide class has a reasonable safety profile at the doses tested. This precedent does not guarantee aurB’s safety, but it provides a scientific basis for optimism about the class.
The combination with radiation, which produced the most dramatic results in the preclinical models, would likely be incorporated into the early clinical design, given the logical rationale: radiation damages DNA and induces cell death through mechanisms that aurB’s mitochondrial disruption should complement.
Yamada sees aurB not as a destination but as a proof of concept for a broader approach. “There are many other bacterial proteins that could be a source of cancer drugs. We simply haven’t tried them yet.”
The broader vision
The tumor microbiome may be a pharmaceutical library hiding inside cancer itself.
This statement, almost casual in its delivery, contains a genuinely large idea. If the tumor microbiome is a diverse ecosystem of bacteria, each carrying their own protein repertoire, and if those proteins have been evolving in direct contact with cancer cells for however long the tumor has been developing, then some fraction of them may have acquired activities that interfere with cancer cell biology. Auracyanin appears to be one. Azurin was another. The tumor microbiome of a pancreatic cancer patient, or a lung cancer patient, or a glioblastoma patient, may contain cupredoxins or other bacterial proteins that no laboratory has yet isolated and tested.
The systematic exploration of the tumor microbiome as a pharmaceutical library is still in its infancy. The Yamada lab’s work represents one of the most advanced examples of what that exploration can produce.
The Bigger Picture: Bacteria as Drug Designers
There is something philosophically satisfying about this research that the dry language of scientific papers does not fully convey.
Cancer is often described as evolution gone wrong: cells that have accumulated mutations allowing them to escape the normal constraints on growth, to proliferate without limits, to spread. Cancer exploits the machinery of life to serve its own survival at the expense of the organism it inhabits.
What Yamada’s research suggests is that the microorganisms that have been colonizing tumors, living in that hostile, oxygen-poor, immune-suppressed environment for months or years, may have been doing something analogous on a microscopic scale: evolving their own proteins to survive in and interact with cancer cells. And some of those survival adaptations may have, as a byproduct, produced molecules that interfere with cancer cell biology in ways that a pharmaceutical chemist designing from scratch might never have imagined.
The takeaway
Bacteria inside tumors may not be passive passengers. Some may carry molecular tools that can be turned into cancer therapies.
Bacteria as inadvertent drug designers. The tumor’s own microbial inhabitants as the source of its undoing.
This is not metaphor. It is the mechanism of auracyanin, documented in a peer-reviewed paper in one of the world’s leading translational research journals, from a team that has been building toward this finding for over two decades.
The tumor thought the bacteria were passengers. They may have been witnesses, and one day, medicines.
References and Attributions
Primary Source:
Naffouje S.A., Tran D.B., Rademacher D.J., Botti V., Christov K., Green A., Li W., Phong N.H.T., Cannistraro S., Bizzarri A.R., Das Gupta T.K., Yamada T. “Suppression of mitochondrial energy production by a photosynthetic bacterial cupredoxin peptide inhibits tumor growth.” Signal Transduction and Targeted Therapy, April 7, 2026.
DOI: 10.1038/s41392-026-02703-7
URL: https://www.nature.com/articles/s41392-026-02703-7
Yamada Lab foundational work:
Yamada T., Goto M., Punj V., Zaborina O., Chen M.L., Kimbara K., Majumdar D., Cunningham E., Das Gupta T.K., Chakrabarty A.M. “Bacterial redox protein azurin, tumor suppressor protein p53, and regression of cancer.” PNAS, October 2002.
DOI: 10.1073/pnas.222539699
PMID: 12393814
Punj V., Bhattacharyya S., Saint-Dic D., Vasu C., Cunningham E.A., Graves J., Yamada T., et al. “Bacterial cupredoxin azurin as an inducer of apoptosis and regression in human breast cancer.” Oncogene, March 25, 2004.
DOI: 10.1038/sj.onc.1207376
PMID: 14981543
p28 clinical trial:
First-in-human Phase I trial of p28 (NSC745104) in patients with p53+ metastatic solid tumors. NCI/Pediatric Brain Tumor Consortium. Referenced in Signal Transduction and Targeted Therapy, 2026 paper, references 27-29.
Background science on cancer mitochondria and Warburg effect:
Sánchez-Aragó M., Formentini L., Cuezva J.M. “The Mitochondrial ATP Synthase/IF1 Axis in Cancer Progression: Targets for Therapeutic Intervention.” Cancers, 2023.
DOI: 10.3390/cancers15153775
Bioenergetics of cancer cells and Warburg effect regulation of ATP synthase. Molecular Medicine, October 2025.
DOI: 10.1186/s10020-025-01378-0
Institutional background:
UIC Today. “UIC’s Yamada elected National Academy of Inventors fellow for cancer work.” April 2, 2024.
URL: https://today.uic.edu/uics-yamada-elected-national-academy-of-inventors-fellow-for-cancer-work/
SciTechDaily. “Scientists Turn Cancer’s Own Bacteria Against It in Breakthrough Therapy.” May 8, 2026.
URL: https://scitechdaily.com/scientists-turn-cancers-own-bacteria-against-it-in-breakthrough-therapy/
Funding:
This research was supported by the National Cancer Institute through an R01 grant to Tohru Yamada and the University of Illinois Cancer Center.
0 Comments