A team of scientists led by researchers at Baylor College of Medicine and the University of Texas at Austin have harnessed Escherichia coli bacteria to help discover human proteins that can initiate DNA damage, and which may play a role in triggering cells to become cancerous. The novel approach, which involved engineering bacteria to overexpress one of each of the 4,000 E. coli genes, also identified biological mechanisms that underpin the DNA damage resulting from protein overproduction. The researchers suggest that their findings could point to new strategies for human cancer therapy, and help to identify individuals at risk of cancer.

“One way proteins can cause DNA damage is by being overproduced, which is a relatively frequent cellular event,” commented Susan M. Rosenberg, PhD, Ben F. Love chair in cancer research and professor of molecular and human genetics, of molecular virology and microbiology and of biochemistry and molecular biology at Baylor, who is co-corresponding author of the researchers’ published paper in Cell. “In this study, we set out to uncover proteins that, when overproduced by the cell, cause damage to DNA in ways that can lead to cancer.”

The scientists report on their studies and results in a paper titled, “Bacteria-to-Human Protein Networks Reveal Origins of Endogenous DNA Damage.”

graphical abstract
Source: Cell Press

Damage to DNA can lead to spontaneous mutations that result in the development of cancer and genetic diseases, as well as underpinning evolution, the researchers explained. “Cancer is a disease of mutations,” Rosenberg commented. “A normal cell that has accumulated several mutations in particular genes becomes likely to turn into a cancer cell.”

Although DNA damage can result from environmental insults such as sun damage or tobacco smoke, most damage to DNA results from cellular processes that involve cell components and molecules, including proteins. However, the researchers noted, “the identities and functions of endogenous DNA damage-promoting proteins in any organism are poorly understood.”

One way of discovering which proteins can promote DNA damage is to investigate protein overproduction, which is a major cancer driver, the researchers continued. To do this Rosenberg and colleagues devised an unconventional approach. Rather than try to identify human proteins directly, they turned to E.coli, and looked for bacterial proteins that, when overproduced, caused DNA damage in the bacterial cells. “Given DNA biology conservation across life, proteins that promote spontaneous DNA damage may be conserved, and their identification could potentially inform strategies for prevention, diagnosis, and treatment of disease, including cancer, aging, and pathogen evolution,” the team stated.

“This was a wild idea,” Rosenberg acknowledged. However, she noted, “although bacteria and people are different, their basic biological processes are similar, so with this approach we thought we might find common mechanisms of DNA damage that could be relevant to cancer.”

To carry out their search the team genetically modified E. coli to overexpress the 4,000 bacterial genes individually, and to fluoresce red when protein overproduction was associated with endogenous DNA damage. They called these proteins DNA “damage-up proteins (DDPs). The findings threw up some unexpected results, including the indication that only 8% of the 208 DDPs identified were involved in DNA repair.

“We uncovered an extensive and varied network of proteins that, when overproduced, alter cells in ways that lead to DNA damage,” Rosenberg continued. “Some of these proteins are, as expected, involved in DNA processing or repair, but, surprisingly, most are not directly connected to DNA. For instance, some of the DNA damage-up proteins participate in the transport of molecules across the cell membrane.” Further tests with 32 of the E. coli DDPs confirmed that their overproduction was linked with endogenous DNA damage that increased mutation rates. “Thus, overproduction of diverse E. coli proteins causes DNA-damage and mutations of essentially all kinds,” the authors wrote.

The team subsequently identified 284 human proteins with amino acid sequences similar to 58 of the E. coli DDPs, and which represented candidate human DDPs (hDDPs). They determined that the human DDPs were linked to cancer more often than a random set of proteins, while evaluation of protein and RNA data from resources including The Cancer Genome Atlas (TCGA) suggested that many of the human DDPs were linked with reduced patient survival and higher total tumor mutation loads in different types of cancer. “These data highlight the network properties of the 284 DDP homologs, their frequent overexpression in cancers, and predictive power for poor survival and high tumor mutation loads,” the team noted. “The correlations of the 284 human homolog RNAs with tumor mutation loads and poor survival remain strong even when both known or predicted cancer drivers and human proteins validated as DNA damage-instigating here are removed.”

When the researchers overproduced these proteins in human cells in the lab, half of the proteins triggered DNA damage and mutation. These validated human DDPs included different classes of protein that had not previously been predicted to promote DNA damage, and had not been linked with cancer. “Many hDDPs span diverse protein functions, the cancer-driving roles of which may be obscure or misassigned,” they commented. “Some of the mechanisms of hDDP action may necessitate reevaluation of their cancer-driving mechanisms and also of the drugs designed to inhibit them.”

“We showed that E. coli can help to identify DNA damage-up proteins and mechanisms of action in human cells quickly and inexpensively,” stated co-corresponding author Christophe Herman, PhD, professor of molecular and human genetics and molecular virology and microbiology at Baylor College of Medicine and member of the Dan L. Duncan Comprehensive Cancer Center. “Some of the proteins and their mechanisms were known to be involved in cancer, but many others were not suspected of being in the cancer-causing list.”

The authors suggest that DNA damage could feasibly represent a biomarker for predicting cancer. “The ability to detect high DNA damage loads in cells could potentially make DNA damage screening attractive for early identification of at-risk individuals, useable before genome sequencing would identify disease-associated mutations,” they wrote. “Our data suggest that DNA damage or upregulation of DDPs may predict tumorigenic processes and susceptibilities in various cancers.”

Rosenberg maintains that the study results could also help to progress both basic research and clinical developments. “We provide a previously unknown understanding of the diverse mechanisms that can generate DNA damage leading to cancer,” she said. “In the future, this finding may lead to new ways to identify people who are likely to develop cancer so that strategies to prevent it, slow it down, or catch it early can be used.”

“I think it is extraordinary to identify so many ways DNA can be damaged,” added co-corresponding author Kyle M. Miller, PhD, associate professor of molecular biosciences at the University of Texas at Austin and member of the Dan L. Duncan Comprehensive Cancer Center at Baylor. “This study is opening up new avenues for discoveries of novel mechanisms that protect our genomes and how their dysfunction can alter the integrity of our DNA and cause cancer.”

Previous articleMachine Learning Provides New Pathway to Better Understand Brain Disorders
Next articleTime to Chew the Fat on Lipidomics