ABOVE: An artist’s rendition of transcription inside a nucleus © ISTOCK.COM, SELVANEGRA

The rate of transcription across all genes could serve as a biomarker of cancer with clinical applications. A newly developed algorithm allowed researchers to measure transcription levels more accurately than ever before, leading to the discovery that the degree of hypertranscription in a patient’s tumor cells is fairly predictive of their survival chances, according to a November 23 study in Science Advances

“Hypertranscription is the global elevation of gene activity that can occur normally during stages of growth in development or in stem cells,” explains Miguel Ramalho-Santos, a cell biologist at the Lunenfeld-Tanenbaum Research Institute at Mount Sinai Hospital, who was not involved in the study, in an email to The Scientist. “This study finds that hypertranscription is pervasively co-opted by cancer cells across different types of cancer, and is correlated with poorer patient survival,” he says, adding that the work has several important implications, including raising awareness to the pervasiveness of hypertranscription in cancer and pointing to the phenomenon as a potential target for cancer therapies. 

Quantifying transcription in hypermutated cells

Highly mutated DNA and transcription gone awry are defining features of cancer. However, whether transcription levels are altered across all genes in tumors had not been studied closely—the phenomenon, known as hypertranscription, has mostly been described in cell lines. In large part, that’s due to technical challenges: When RNA and DNA are extracted from tumor samples, information as to how many cells they were extracted from is lost, making it difficult to calculate the per-cell transcription amount.  

See “Cancers Ramp Up Overall mRNA Expression as They Progress

To get around this issue, the team developed a new computational tool, which they dubbed RNAmp. It uses the number of tumor-specific mutations as well as tumor-specific copy number changes—changes in chromosome structure that lead to gain or loss of DNA sections—to normalize RNA measures, allowing for accurate estimation of the transcription levels of tumor cells. 

“When you use this algorithm and matched DNA sequencing, you can estimate the proportion of the transcriptome . . . that derives from the tumors as opposed to the surrounding [nontumor] cells”, says study coauthor Adam Shlien, a cancer genomics researcher with The Hospital for Sick Children (SickKids) in Toronto.

Using RNAmp, Shlien and his team examined transcription levels in 7,494 tumors from 31 different cancer types, finding that transcription levels were universally higher in tumor cells than in surrounding nontumor cells. “All of the genes are elevated above baseline level. It’s as if there’s a person with a megaphone who is amplifying this really corrupted message across the entire genome,” says Shlien. 

The researchers identified several transcription factors involved, but Shlien says “the way that [transcription factors] drive hypertranscription is really through the absence of their suppression. . . . So instead of accelerating . . . you’re taking your foot off the brakes.”

Poor prognosis is associated with hypertranscription

As the transcriptomic and genomic data came from The Cancer Genome Atlas, the researchers were able to correlate the degree of hypertranscription with health outcomes, particularly 5-year survival data. Any hypertranscription was associated with poorer survival prognoses, says Shlien,  “above and beyond stage, grade, and tumor subtype.” Furthermore, patients with tumors that had higher levels of hypertranscription did not survive as long as patients whose tumors had lower levels of hypertranscription.

Why hypertranscription is correlated with worse outcomes is “still a mystery,” Shlien says. Global upregulation of transcription may influence signaling pathways involved in the cancer’s aggressiveness, he suggests, but that’s just a hypothesis. What did emerge from the data is that for some tumor types, hypertranscription levels vary widely among patients, while in others, there is little “wiggle room,” as Shlien puts it.

Even without the underlying mechanism to explain it, the connection between hypertranscription and poor prognosis might present a new avenue for improving treatments, additional analyses in melanoma patients suggest. In these patients, the number of mutations in a tumor’s cells, known as the tumor mutational burden (TMB), is typically used to determine whether to proceed with immunotherapy, as patients with a high TMB usually respond better to the treatment. But Shlien and colleagues found that tumors can exhibit higher levels of hypertranscription despite having a low TMB, and that patients with such tumors also benefit from immunotherapy. “We think these patients should become eligible for immunotherapy based on the hypertranscription,” says Shlien, who has filed a patent application related to using tumor-specific transcription to predict patient prognosis. He says that based on these results, he’s seeking to expand the eligibility for immunotherapy in the precision medicine arm of the SickKids Cancer Sequencing program.