Two independent teams have identified activating mutations in the promoter of telomerase reverse transcriptase that occur in more than 70% of melanoma cases.1,2 The findings highlight the still-untapped potential of analyzing noncoding regions of cancer genomes and offer a hint about how some malignancies develop.

The functional role of these mutations still needs to be further explored before their therapeutic relevance can be assessed.

The chromosomes of all eukaryotic organisms are bookended by regions of repetitive DNA sequences known as telomeres, which are shortened during the replication process. Telomerase is an enzyme complex that synthesizes telomeric DNA and maintains telomere length, which is essential for chromosomal stability and continued cell division. Telomerase reverse transcriptase (TERT) is the core catalytic protein subunit of the enzyme.

In most mature tissues with limited cellular division, telomerase is tightly regulated and is either not expressed or expressed at a very low level. In contrast, highly proliferative cells, including almost all cancer cells, require increased telomerase expression to maintain telomere length and continue growth.

Because of this, the enzyme has been pursued as a therapeutic target for more than a decade, and at least one company, Geron Corp., has advanced a telomerase inhibitor into Phase II trials.3

Despite intensive study of the regulation of telomerase function, the mechanisms underlying the enzyme's activation in cancer have remained unclear.

Now, independent teams from the Broad Institute of MIT and Harvard and the German Cancer Research Center have identified highly prevalent mutations that could explain how telomerase is activated in melanoma and potentially many other cancers.

Neither team set out to study telomerase. Instead, both sought to use unbiased DNA sequencing analyses to identify new mutations associated with melanoma. The groups converged on the same result from completely different angles.

The German researchers studied a single family to understand why its members had an unusually high incidence of melanoma despite lacking known disease-associated mutations. Linkage mapping and high throughput sequencing identified a germline point mutation upstream of TERT in all affected family members and in one younger, thus far unaffected family member.

The mutation occurred in no healthy controls. In reporter assays, the mutation increased TERT promoter activity compared with that in wild-type controls.

Corresponding author Rajiv Kumar, professor of molecular genetic epidemiology at the German Cancer Research Center, told SciBX that the findings prompted his team to home in on the TERT promoter.

"This represents the first familial melanoma mutation since the discovery of mutations in CDKN2A and CDK4 in the 1990s. Following the logic that familial germline mutations have also been found as somatic mutations in spontaneous melanoma cases, we subsequently set out to sequence the region in a large number of melanomas," he said.

The results were striking. Mutations were found in the TERT promoter in 125 of 168 melanoma cell lines (74%). The mutations were not found in germline DNA samples from matched controls, indicating that although familial inheritance of TERT mutations is rare, TERT mutations can occur spontaneously with high frequency.

Indeed, the TERT mutations were found more frequently than mutations in the most common genes associated with melanoma, including BRAF at 53%, neuroblastoma Ras viral oncogene (NRAS) at 23% or cyclin dependent kinase inhibitor 2A (CDKN2A; INK4a; ARF; p16INK4a) at 50%.

"The TERT promoter mutations now become the most frequent lesions in melanoma," said Kumar. "These are more frequent than mutations in BRAF, which has been a very successful example of targeted therapy in melanoma."

The small molecule BRAF inhibitor Zelboraf vemurafenib is marketed by Roche and Daiichi Sankyo Co. Ltd. to treat metastatic melanoma in patients expressing the V600E BRAF mutation.

The Broad Institute team used a brute-force sequencing approach to arrive at the same conclusion. Whole-genome sequencing of 19 melanoma samples identified 1 of 2 mutually exclusive activating mutations in the TERT promoter in 17 cases (89%). In an independent set of 51 samples, 33 (65%) contained 1 of the mutations.

Both teams noted that the mutations generate binding sites for the Ets and ternary complex factor (TCF) class of DNA-binding transcription factors, which could explain the increased TERT transcription seen in reporter assays.

The Broad team also provided early evidence that TERT mutations could be found in additional cancers. The group analyzed data from 150 cell lines included in the Cancer Cell Line Encyclopedia.4 Of those, 24 (16%) contained the mutations, including 3 of 3 bladder cancer cell lines and 4 of 6 liver cancer cell lines.

Kumar said his team did not detect mutations in 22 esophageal squamous cell carcinomas but added that further studies in additional cancer types are warranted.

Results from both studies were published in Science.

Unraveling function

The results suggest activating mutations in telomerase could drive melanoma malignancy, but key functional studies are needed to firm up the case.

Jerry Shay, professor of cell biology at The University of Texas Southwestern Medical Center, said the high frequency of mutations strongly suggests they could be cancer-driving events. He now wants to see studies that test the cause-and-effect relationship of these mutations. Those findings, he said, will be essential to understanding the relative importance of telomerase upregulation in these samples.

"The bottom line is that the only evidence they have that TERT transcription is changing is an artificial promoter reporter system. That's a deficiency. If this is creating a functional binding site, you might expect samples with the mutations to have higher TERT levels than samples without the mutations," said Shay.

He added that telomerase regulation is complex, and studies looking at the role of the mutation in the context of the native promoter are absolutely essential to establish its functional importance.

Nevertheless, Shay said it was tempting to speculate that these findings could explain why benign naevi (moles) escape senescence and progress into malignant melanoma cells.5 "Most benign naevi are quiescent and do not express telomerase, while melanomas have robust telomerase expression. One could imagine this happening to drive the transition from benign naevi to melanoma," he told SciBX.

Kumar agreed that was a possibility. "Our next step is to try to understand how the discovered mutations affect melanoma formation, and if those mutations relate to disease progression and outcome. We will try to screen mutations in melanoma lesions from different stages of progression in order to understand at what stage these mutations arise in the development of melanoma," he said.

Geron EVP, CMO and head of R&D Stephen Kelsey said what struck him was that the mutations are cytidine to thymidine transitions, which are indicative of damage caused by UV light and could explain how sun exposure causes skin cancer. "It is quite remarkable that this particular region of the TERT promoter is very sensitive to UV light-induced damage," he said.

He added that the findings were all the more striking because very few mutations have been described in the TERT coding region, and there have not been clear correlations identified between TERT mutations and cancer susceptibility.

Kelsey said functional studies are the clear next step to determine the importance of telomerase in melanoma samples carrying the mutations. He suggested that, counterintuitively, these mutations could actually make tumors less susceptible to telomerase inhibition.

"Here's the irony: if telomerase is upregulated early on in the evolution of the tumor, the telomeres will not have the chance to shorten. And while we haven't done an extensive screen of melanoma, it turns out the telomere length in melanoma is relatively normal, as you would expect if the enzyme were upregulated early on," he said. "The implication is that if you switch off telomerase in melanoma, they will have a lot of telomere length left, and it would take a long time for the telomeres to shorten to a critical length, causing cellular senescence or death. We have done some preliminary in vitro work with melanomas, and we have not seen that they are particularly sensitive to imetelstat."

Geron's imetelstat, a modified oligonucleotide inhibitor of telomerase, is in Phase II testing in multiple cancers. Kelsey said the company is refining a telomere length companion diagnostic for the product, which could be used to select patients whose tumor cells have short telomeres for imetelstat treatment. In a prespecified subgroup analysis of a Phase II study of imetelstat in non-small cell lung cancer (NSCLC), patients with tumor cells with short telomeres had a greater progression-free survival benefit with imetelstat than patients with tumor cells with medium to long telomeres.

Genome gazing

Both teams told SciBX that it was initially surprising that the TERT promoter mutations had not been uncovered so far, given the intense prior studies of telomerase regulation and the frequency with which they found the mutation.

"The level of recurrence suggested it was a technical artifact, and we really had to nail down that it was real before we put the result out there," said Eran Hodis, a graduate student at the Broad Institute who was co-first author of the Broad study with postdoc Franklin Huang.

He added that it made sense that labs have largely focused their sequencing studies on coding regions in the past. "The two factors are cost and interpretability, and you have to take into account where previous driver mutations have been found. We have known for many years of driver mutations in coding regions, so it made sense to look there, and it saves you time and money," said Hodis.

"We think in the near future there will be more such discoveries in other types of cancer as whole-genome sequencing becomes more prevalent," said Kumar.

He added that one reason these particular mutations may not have been detected before is because they are located in a highly GC-rich sequence that is not easy to amplify. He also said the proposed mechanism of action described by these studies-a point mutation driving gene expression by generating a new transcription factor binding site-has not been seen before.

Roman Thomas, professor and chair of the Department of Translational Genomics at the University of Cologne, said the findings provide a compelling argument for the importance of whole-genome vs. exome sequencing.

The exome is the exon-containing, protein-coding region of the genome, which makes up about 1% of total genetic material and thus requires fewer resources to sequence.

"It shows that looking just at exons represents a liability for not making important discoveries. It was only a matter of time before people would find something in genomes that would not be detectable by exome sequencing," said Thomas.

Thomas is also cofounder and scientific director of Blackfield AG, which is focused on cancer genome analysis.

Thomas warned that although the transition from exome sequencing to whole-genome sequencing is inevitable, the speed of the transition may be slowed due to funding and pricing considerations. Exome sequencing can cost less than $1,000, whereas whole-genome sequencing can cost about $5,000 on average.

"People would love to do more genomes, and we sequence them whenever we can. The problem is that recently the prices have stopped dropping, in contrast to the previously exponentially dropping costs in genome sequencing," said Thomas. "That is just not the case anymore, and it is a bit of a disaster for many people because we have written grants and proposed experiments expecting to factor in falling prices."

Results from both studies are unpatented.

Cain, C. SciBX 6(7); doi:10.1038/scibx.2013.157 Published online Feb. 21, 2013


1.   Huang, F.W. et al. Science; published online Jan. 24, 2013; doi:10.1126/science.1229259 Contact: Levi A. Garraway, Broad Institute of MIT and Harvard, Cambridge, Mass. e-mail:

2.   Horn, S. et al. Science; published online Jan. 24, 2013; doi:10.1126/science.1230062 Contact: Rajiv Kumar, German Cancer Research Center, Heidelberg, Germany e-mail: Contact: Susanne Horn, same affiliation as above e-mail: Contact: Dirk Schadendorf, Essen University Hospital, Essen, Germany e-mail:

3.   Harley, C.B. Nat. Rev. Cancer 8, 167-179 (2008)

4.   Barretina, J. et al. Nature 483, 603-607 (2012)

5.   Michaloglou, C. et al. Nature 436, 720-724 (2005)


      Blackfield AG, Cologne, Germany

      Broad Institute of MIT and Harvard, Cambridge, Mass.

      Daiichi Sankyo Co. Ltd. (Tokyo:4568; Osaka:4568), Tokyo, Japan

      German Cancer Research Center, Heidelberg, Germany

      Geron Corp. (NASDAQ:GERN), Menlo Park, Calif.

      Roche (SIX:ROG; OTCQX:RHHBY), Basel, Switzerland

      University of Cologne, Cologne, Germany

      The University of Texas Southwestern Medical Center, Dallas, Texas