From DNA, cancer arises. Or so the thinking goes. The idea that DNA mutations drive cancer is not just deeply rooted, it is dogma. But what if we’ve gotten it all wrong?
Scientists at Fred Hutch Cancer Center and Seattle Children’s argue that we have. Instead of looking to mutations in DNA as the source of cancer, they contend that we should instead focus one level up, at RNA and proteins that package DNA.
They contend that cancer cells need more key packaging proteins to support break-neck proliferation, and hypertranscription ― or the wholesale ramping up of RNA production ― gets cancer cells the histones they need without creating a dangerous imbalance.
You will receive a free monthly newsletter from us about our innovative research and compassionate care as well as updates on events and other helpful tools and tips.
You may unsubscribe at any time.
Listen to the audio interview
Libsyn | Spotify | Apple Podcasts | Amazon Podcasts | YouTube
“We keep getting evidence that hypertranscription is what’s driving cancer, that cancer has an epigenetic origin,” said Fred Hutch molecular biologist Steve Henikoff, PhD, who studies the structure, function and evolution of our DNA molecules. Epigenetic changes alter gene expression, in which genes are transcribed into RNA, without any change to DNA sequence.
Henikoff believes by the time we can detect tumors, much of the DNA disarray they’ve accumulated can be traced to the biological processes encouraging hypertranscription and rapid proliferation. Henikoff and collaborator Jay Sarthy, MD, PhD, a pediatric hematologist at Seattle Children’s who trained in Henikoff’s lab, argue that their new framework helps answer some nagging questions about cancer and why some chemotherapies work even though they don’t damage DNA.
“We thought so many chemotherapies were targeting the central dogma of molecular biology, by hitting cancer and destroying it at the DNA level,” said Sarthy. Instead, he and Henikoff now think that chemotherapies have been targeting transcription all along: “We think the transcription-translation relationship is the critical one in cancer, and we’ve been tossing most of the data out.”
Now, he and Henikoff are revisiting — and expanding on — that overlooked data. Their teams are putting the finishing touches on two projects, each focused on a different RNA-making enzyme, which they contend reveal how hypertranscription drives cancer (and why tumors need it) and also explain how anthracycline chemotherapies (and others) actually work.
Stem cells and cancer cells share some noteworthy features, including the ability to divide and continue dividing. They are also less differentiated, or specialized, than most cells and this allows stem cells to turn into a range of different cell types within a tissue. De-differentiation is a hallmark of cancer.
Cancer cells rarely invent anything new. Mostly they co-opt biological protocols and processes used by normal cells, but apply them without the usual safeguards. This allows them to continue growing under conditions that would hobble normal cells and blow past checkpoints designed to keep DNA safe from harm. Co-opting hypertranscription enables cancer cells to take on stem-like characteristics, Henikoff argues.
By the time we can detect cancer, the original tumor cell has given rise to hundreds of thousands of cells. Minor changes picked up along the way get magnified and compounded into a dizzying mix of cells with unique characteristics. At this point, it’s hard to strip away changes that are side effects from those that caused the cancer, Henikoff said. Tumors’ similarities are obscured by their differences.
Like viewers seeing the hidden image in an optical illusion, he and Sarthy are reorienting their viewpoint: They’re bringing RNA and epigenetics to the foreground and letting DNA mutations fade into the background.
Henikoff says that it boils down to hypertranscription and histones, critical DNA packaging proteins (and important targets of epigenetic modifications that influence gene transcripiton), which enable tumor cells to divide rapidly.
Histones are proteins that interlock into a wagon wheel-shaped spindle called the nucleosome. The nucleosome, which DNA wraps around, is one of the most basic units of DNA organization and histones are essential players in gene regulation. Only 72 out of our 20-30,000 genes are histone genes, but they are so important that about 5% of our proteins are histones.
In many cancers, histone gene transcription goes way up — along with transcription of nearly every other gene. Histone gene upregulation was long dismissed as a side effect of a general cancer-associated uptick in transcription, but Henikoff believes it’s the opposite.
To divide at break-neck speed, cancer cells must fast-track DNA replication. During this process, DNA becomes vulnerable to damage, but histones offer shelter. In Nature Structural & Molecular Biology, Henikoff Lab staff scientist Kami Ahmad, PhD, showed that even in healthy cells, DNA replication (and therefore proliferation) depends on the rate of histone synthesis. If a cell can boost its histone production, it can turbo-charge its cell cycle.
It’s been calculated that to make the same amount of histone proteins in one-third of the time, cells need to load histone genes with 1,000 times the typical amount of RNA polymerase II.
“If histones were to go up out of proportion [to other genes], that would cause an imbalance that would likely make the cells unfit,” Henikoff said, meaning the cells wouldn’t be able to out-divide their neighbors. “But if transcription goes up proportionally, this could drive cancer from the very first cell.”
So, to get those much-needed histones safely, cancers shift all gene transcription into overdrive, he said.
The hints that epigenetics, not DNA mutations, lies at cancer’s root have been lurking in the data for years, Henikoff said, but they were obscured by technological limitations and a focus on how tumors differ.
Previous work on the Myc family of transcription factors, which regulate gene expression and are known cancer drivers, had suggested that they foster tumor development by amplifying the output of a given cell’s entire gene expression program. But there’s more to the story, said Henikoff.
The enzymes that “read” a gene’s sequence of DNA bases and produce an RNA transcript are called RNA polymerases. RNA polymerase II (pol II) transcribes genes into messenger RNAs (mRNA) that will then be translated into proteins. The RNA molecules transcribed by RNA polymerases I and III (pol I and pol III) are incorporated into the ribosome, the protein-translating apparatus. The three polymerases also produce other types of RNA, like tiny microRNAs, that have various gene regulatory functions. Myc propels hypertranscription by influencing all three polymerases, like a maestro leading a complex orchestra to a united crescendo.
“Myc is the driver, but the RNA polymerases are what’s under the hood,” he said, likening the transcriptional apparatus to a car driving along a DNA highway. “Everything else is the gas and brake pedals, but the RNA polymerases are where the rubber meets the road. And we’d missed it.”
This was because standard techniques to assess a cell’s RNA output focus on the mRNA produced by pol II. (Most biological processes run on proteins, so they get a lot of scientific attention.) These pol II-focused strategies miss the surges in ribosomal RNAs from pol I and pol III that occur with hypertranscription.
“The cancer field doesn’t love this message, because they want RNA polymerases to be doing these magical things of turning on and off lineage-specific factors,” Sarthy said. “But I think this is just about making enough histone and ribosomal protein to keep the cell going.”
Despite the work linking Myc to hypertranscription, cancer geneticists had ignored RNA polymerases, Henikoff said. He and Henikoff Lab data analyst Jorja Henikoff used publicly available datasets to discover a microRNA cluster upstream of ribosomal RNA genes. They revealed that two highly cancer-associated microRNAs, miR-1275 and miR-6724, arise from the same 50-base pair stretch of DNA upstream of ribosomal RNA genes.
One, miR-1275, “is the best cancer biomarker anyone knows about,” Henikoff said, noting that most of the thousands of studies on miR-1275 focus on cancer. “Between 30-60% of mRNAs have an microRNA that targets them.”
This includes transcription factors, which regulate gene expression programs, and cancer takes full advantage of a hypertranscription-related rise in microRNAs and a “level of gene regulation we may never understand,” he said.
So far, it seems their functions don’t change, but the sheer abundance of the microRNAs pushes them to the top of the list of cancer biomarkers and can have big effects on tumor cells and their environment, Henikoff said.
Henikoff and his team were inspired to look for evidence of hypertranscription in their own data, revisiting epigenetic studies they had previously done of archived tumor samples.
Using a technique he and Fred Hutch collaborator and brain cancer expert Eric Holland, MD, PhD, had developed, Henikoff and his team surveyed historical tumor samples from patients to profile RNA pol II’s location on cancer genomes and directly measure hypertranscription.
With Holland, who holds the Pigott Family Endowed Chair, Ahmad and Henikoff examined meningiomas (a type of brain tumor) and breast cancers. In work published in the journal Science, they found evidence of hypertranscription via RNA pol II location, but also that RNA pol II at histone genes in particular correlated with poor prognosis and rapid recurrence.
Henikoff and his team argue that cancer hallmarks, including extrachromosomal circles of straggling bits of orphaned DNA and aneuploidy, in which chromosomes gain or lose entire arms, can also be traced back to hypertranscription and their rising tide of histones.
In the Science paper, the team had shown that chromosome arm loss predominated over chromosome arm gain and correlated with RNA pol II at histone genes. Ahmad, Henikoff and then-postdoctoral fellow Ye Zheng, PhD (now a professor at the MD Anderson Cancer Center), proposed that chromosomal arm loss could be traced back to histones as well.
A specific type of histone, called CENP-A, helps form the centromere, the cinched-in waist of the chromosome where duplicated chromosomes attach to each other before being separated during cell division. The team hypothesized that as hypertranscription pushes histone levels up, standard (or “canonical”) histones (known as H3 and H4) can shoulder out CENP-A and muck with centromere integrity.
Weakened centromeres could then increase the chance that a chromosome arm will break off. In the Proceedings of the National Academy of Sciences, the collaborators showed that whole-arm loss could predict patient outcomes across different types of cancer. They also presented evidence that centromere breakage could be enough to explain loss or gain of chromosome arms.
Hypertranscription can also explain the cancer-associated rise of microRNAs, Henikoff said.
MicroRNAs are tiny RNAs that exert effects on gene expression, acting in between transcription and translation to hobble mRNAs before they can get translated into proteins.
And their new framework also remains consistent with cancer’s association with aging, Henikoff said. Cancer incidence rises with age, which most scientists have attributed to the fact that we accumulate DNA mutations over time. But Henikoff Lab postdoctoral fellow Sarah Leichter, PhD, showed that just as cancer appears to acquire stem-like programs, aging stem cells become more cancer-like, increasing their proliferation and showing similar patterns of RNA pol II at histone genes.
Henikoff and Sarthy contend that evidence that transcription, not DNA, is the true target of most chemotherapies has also been overlooked by most investigators.
For example, inhibitors of enzymes that facilitate transcription, such as transcriptional kinases, are effective against cancer, Henikoff said.
Ribosomes, which build proteins, are also a target of chemotherapies. Drugs called topo poisons, which introduce torsion into DNA by targeting the topoisomerases that untangle DNA, and histone deacetylase (HDAC) inhibitors (which affect how DNA wraps around nucleosomes) affect the nucleolus. This is where ribosomes, the protein-building machines, are created. Researchers had thought the effects on the nucleolus were a byproduct of the drugs’ actual activity, but Sarthy disagrees: “I think it’s the main effect, it’s how these drugs kill cancer cells,” he said.
He and Henikoff think most chemotherapies work by interfering in some way with transcription or translation, rather than by damaging DNA. They began appreciating how important transcription is to cancer cells while studying how a class of chemotherapies called anthracyclines work.
One of the most widely used chemotherapies, doxorubicin (known to patients as the “red devil” for its toxicity) is an anthracycline. Doxorubicin, like other anthracyclines, inserts itself in between the DNA bases. It pushes them apart, changing the torque and tension of the DNA helix.
To relieve this, topoisomerases snip the DNA to release the torsion. Normally, they also repair these breaks — but doxorubicin, a topo poison, blocks the DNA repair step. With this damage comes side effects like heart muscle damage and an increased risk of secondary cancers.
“The previous model was that DNA damage by anthracyclines is what’s killing cancer cells,” Sarthy said.
But in 2013, Henikoff and Fred Hutch colleague and functional genomics expert Chris Kemp, PhD, showed that aclarubicin, another anthracycline, can increase turnover of histone proteins without leading to DNA breakage. They found that when doxorubicin and anthracycline muscle into DNA, they shoulder out histones. Jacques Neefjes, PhD, then of the Netherlands Cancer Institute, independently discovered the same phenomenon.
Aclarubicin, still used widely in China, never caught on in the U.S and fell out of favor in Europe after its patent expired. But it is nearly as effective, though much less damaging long-term, as doxorubicin. Neefjes is working to reintroduce aclarubicin to Europe while Sarthy works to bring it to the U.S.
“The patients who need it most sit at opposite ends of the age range,” Sarthy said.
Anthracycline therapy can be curative, but many older cancer patients don’t receive it because they are too frail to withstand the heart damage it causes. Children can withstand anthracycline toxicity, but pay a price decades later: Anthracycline treatment raises the risk that survivors of childhood cancers will develop heart disease and frailty much earlier than average.
“Maintaining chemotherapy’s effectiveness while making it safer is the most important problem in oncology, in my opinion,” Sarthy said. “This new lens gives us a way to get there.”
Sarthy and Henikoff have continued to dig into how aclarubicin works, seeking not just insights into how to improve chemotherapies, but also into cancer’s vulnerabilities.
With post-baccalaureate fellow Kevin Nguyen, postdoctoral fellow Matthew Wooten, PhD, and Ahmad, Henikoff showed that aclarubicin trips up RNA polymerase II at histone genes during DNA replication. In work posted on bioRxiv, Nguyen found that aclarubicin treatment causes stalled pol II to build up at histone genes at higher levels than other genes.
Sarthy’s team (including collaboration with Neefjes) examined other RNA polymerases and found similar results. Their findings also showed that anthracyclines ultimately interfere with creation of ribosomes (which translate mRNAs into proteins). A lack of ribosomes and protein translation would parch cancer cells thirsty for histones.
Anthracycline’s apparent affinity for histone genes is less about the genes themselves as the fact that they’re easy targets, Sarthy said.
“There’s no specificity to the drug, it’s just the landscape of DNA torsion,” he explained. “These drugs are just intercalating in the most accessible genes.”
The most accessible genes are the ones that are the most highly transcribed: histones.
Other work supports the idea that anthracyclines work through chromatin and transcription, not DNA damage. Scientists have modified various anthracyclines to retain one activity over the other, and anthracyclines that retain histone-ejecting activity also retain cancer-killing ability. Amrubicin, an anthracycline that can damage DNA but not bounce histones, works poorly.
There’s evidence from other cancer drugs that inhibiting transcription, histone genes or histone translation can have powerful effects on tumors.
PRMT5 inhibitors, a class of anti-cancer drugs under clinical development, modify histones and lead to rapid suppression of histone genes. Their anti-cancer effects have been attributed to defects in splicing mRNA, but Ahmad and Henikoff argue that they come primarily from their ability to turn off histone transcription.
Many chemotherapies that damage DNA also interfere with transcription as a byproduct. Sarthy has come to believe that the transcriptional and translational “byproducts” of most chemotherapies are actually their main mode of action.
“We’ve been targeting transcription horizontally,” he said. “We’ve be targeting the [biological] dogma vertically by, yes, targeting DNA — but we’re inhibiting transcription and translation as the main phenotypic consequence.”
He hopes that their insights can help improve chemotherapies by enabling scientists to focus on the developing therapies that target transcription and translation without accompanying DNA damage.
“Some anthracyclines essentially ‘bend’ DNA but don’t break it, and in doing so inhibit RNA polymerases,” Sarthy said. “Since DNA breakage is linked to a slew of chemotherapy side effects, ‘bend but not break’ may be enough to treat cancer patients effectively while reducing toxicity.”
This work was funded by the Howard Hughes Medical Institute, the National Cancer Institute, the National Human Genome Research Institute, the National Institute of General Medical Sciences, the National Institute of Environmental Health Sciences, an Invent at Seattle Children’s Postdoctoral Scholar Award, a Prostate Cancer Foundation Challenge Award, an American Cancer Society Discovery Boost Award, a Burroughs Wellcome CAMS Award, the Andy Hill CARE Fund, the Cush It to the Limit Foundation, Unravel Pediatric Cancer, Nancy and Dick Bernheimer, the Matthews Family, the Stinchcomb Family and the Tomas & Patricia Wright Memorial Funds.
Sabrina Richards, a senior editor and writer at Fred Hutch Cancer Center, has written about scientific research and the environment for The Scientist and OnEarth Magazine. She has a PhD in immunology from the University of Washington, an MA in journalism and an advanced certificate from the Science, Health and Environmental Reporting Program at New York University. Reach her at srichar2@fredhutch.org.
Are you interested in reprinting or republishing this story? Be our guest! We want to help connect people with the information they need. We just ask that you link back to the original article, preserve the author’s byline and refrain from making edits that alter the original context. Questions? Email us at communications@fredhutch.org
Every dollar counts. Please support lifesaving research today.