Me vs. “MuTEs”: Enhancing Cancer’s Hallmarks via Mutation

Share this post:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Reddit


A small insertion mutation creates a landing site for transcription factor proteins, which drives overproduction of TAL1 protein and leukemia cell traits.

Cancer scientists say all the time that “cancer is a disease of the genome,” but I still have trouble understanding exactly what that means. I think they think they mean that cancer cells grow better/faster/more toxically than healthy cells because the DNA in these cells differs from the DNA of healthy cells. These differences include mutations that a cell’s DNA acquires over a person’s life. Understanding which mutations matter and how they impact the processes of cancer cells is the goal of hundreds of labs and millions of dollars of research. I am but one of the researchers seeking this knowledge.

In 2014, I worked to understand what one particular mutations does, and wound up shining a light on an under-appreciated subtype of mutation and how it seems to impact the processes happening in cancer cells.

The sequence of DNA letters in a cancer cell can differ from that of the healthy cell it originated from at thousands of locations; these differences are called “somatic mutations.” (There’s also mutations that you inherit from your parents, but that’s a whole other post.) The vast majority of these somatic mutations in tumor DNA don’t matter or contribute to a cell being cancerous and are just passengers along for the ride. A tiny minority of the thousands of mutations in cancer cells’ genomes allow these cells to adopt “hallmark” cancer cell traits. Hallmark traits of cancer cells include dividing more frequently, avoiding detection by the immune system, staving off cell death, etc. For a given tumor, we cancer geeks see only the end, meaning we see some traits of the tumor cells and have the list of mutations in the cells’ genomes. What about all the processes and steps that connect the mutations to the traits of that particular cancer? That part is still pretty much a black box.

It is this connection between acquired mutations and acquired traits that we wanted to probe by testing how one particular mutation in one particular type of cancer contributed to the hallmarks of cancer.

From millions of research dollars, thousands of mutations, and hundreds of labs, a principle has emerged: the mutations that matter to cancer cells do something to change which proteins are made in those cells. Different types of healthy cells produce different complements of thousands of proteins at different levels, and the amounts of proteins produced in tumor cells differ from those amounts in the healthy cell it once was. Despite this principle, the research into changing levels of proteins has rarely run up against the research into where the mutations in the genome are located. The mutations we know the most about alter the structures of proteins, and some of these proteins directly control which genes are transcribed; this is one way to change protein levels. But these protein-structure-altering mutations are only a tiny fraction of all the mutations in a cancer genome, and we know comparatively little about other kinds of mutations that alter protein levels.

The mutation we investigated doesn’t break a protein, but we thought it could explain why too much of a signature protein gets made by this particular type of leukemia.

The cells in T cell leukemia often produce troublesome amounts of a protein called TAL1— that’s how the gene was named: “T cell acute leukemia 1.” TAL1 is one of these proteins that controls which proteins get made, so when it gets overproduced, the levels of many proteins are altered. The trouble is that we don’t know how TAL1 gets overproduced. So my collaborators, Tom and Marc, looked at the enhancers around the TAL1-encoding gene. Enhancers are bits of DNA that serve as decision-making “boardrooms” that control whether or not a gene gets transcribed into an mRNA and translated into a protein. They found that some cases of T cell leukemia had an unusually large enhancer right next door to the instructions for making TAL1 protein.

Tom and Marc discovered a mutation in an enhancer in genomic proximity to the TAL1-encoding gene in this sample, and they asked us to help suss out what effects it had on cancer cells.

Proteins stick to specific DNA letters at enhancers when they’re signaling that it’s time to start transcribing a gene. Transcription is the first step on the way to a protein being made from the instructions in that gene. These DNA-binding, transcription-regulating proteins are called “transcription factors,” and they usually like to stick to specific sequences of DNA letters. A mutation that changes as little as one DNA letter can create or destroy the ability of a transcription factor protein to bind the DNA at enhancers. Altering which proteins bind to which enhancers is another way tumor cells alter the levels of proteins made from genes.

Our team discovered that this mutation near the TAL1-encoding gene created a sequence that a transcription factor protein called MYB likes to stick to.

We checked if MYB bound the DNA there, and it does, but MYB has some traits that make this binding event more important than just that protein binds DNA. MYB doesn’t like to work alone. MYB likes to stick to other transcription-regulating proteins and form a protein complex. Oddly enough, the sequences these transcription factors like to bind were already in the area where the mutation happened. So this one mutation happened in a perfect spot to cause not only MYB to bind but other transcription factors as well. Allowing MYB and its buddy proteins to bind seems to have nucleated a really powerful enhancer near the TAL1-encoding gene.

To make sure TAL1 production depended on this enhancer and this enhancer depended on this mutation, we deleted it from the genomes of leukemia cells and followed the cells’ progress.

Specifically deleting the mutation from the leukemia cells’ DNA with that fancy CRISPR technique everyone’s talking about led to a 60% decrease in transcription from the TAL1-encoding gene. A similar percent loss in another study led to a significant decrease in tumor cell multiplication. We found these “MuTEs” (Mutations of the TAL1 Enhancer) in a handful of real patient samples, so this event is happening in the cells of actual people and might help drive their leukemias.

Altogether, this means that some event inserts a few DNA letters into a region of DNA with several pre-existing landing sites for transcription factor proteins. This insertion creates a landing site for another transcription factor, leads to an enhancer where there isn’t usually one, drives production from the instructions in the TAL1-encoding gene, and alters the levels of thousands of proteins in the leukemia cell. These are links between one mutation and the hallmark cancer traits of survival and proliferation, and I got to help uncover them.

Posted in Science 101, Work I've Done | Leave a comment

Snarking the Science of Scorpion

Share this post:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Reddit

I don’t normally watch the TV show Scorpion, the pitch meeting for which I imagine included someone saying “Imagine Big Bang Theory was a cop show.” But my girlfriend’s parents were visiting, so I didn’t have control of the remote. Without commenting on the quality of the series as a whole or its likeability, I am pleased to say that the episode we watched gave me the most unintentionally hilarious moment of TV in weeks.

The Fast and the Nerdiest” episode featured a plotline entwining car races, bioweapons, and something about The Price is Right, because reasons. I would be overselling the attention I paid if I said I could describe the plot in detail, but one moment caused my ears to prick up and earned a hearty bad-science-laugh.

Sly has just stumbled onto an improvised lab. While no one’s home, he finds a red bag with a biohazard label and DNA letters printed on it and gives the camera his best worried face. He’s on the phone with Walt, who is drag racing some gorgeous classic cars– not particularly important, since, hey, classic cars!

Sly: I am looking at biohazard labels with a genome string on it

Walt: Sly, read it aloud to Toby.


Walt: Hold on, back to the CC part


Walt: Toby, you hear what I hear?

Toby: Yes, that whole string is a common cold virus but that CCT runner in the middle, that’s a mitochondrial gene, unique to people of mesoamerican descent

Sly: Like Aztecs found in Balio?

Toby: Yes, exactly like the Aztecs in Balio. This weapon is designed to spread via the common cold, while attaching a specific gene that’s only found in that particular race of people.

This is the spot. This is what got me calculating if there was enough room betwixt my guests and myself to roll on the floor laughing. Please, dear reader, allow me to dissect the… ahem… science in this exchange.

First, the positive: The writers used DNA letters!

Second, the negative: The rhinovirus (common cold virus) has an RNA genome, so the letters aren’t the same. DNA uses ATCG. RNA uses AUCG.

Next, the really negative: A 14-letter stretch of DNA sequence is hugely unlikely to be only found in one strain of one virus that specifically targets one human subpopulation.

In the world of DNA words, size really does matter.

Scorpion_DNA_hits_taxaBeing a bit of a DNA geek, I took this sequence and did a BLAST search. BLAST is a freely-available web tool that tries to guess what species a bit of DNA came from and specifically what part of that species’ genome. TGCTCCTATCCCGA could come from at least 100 different spots in the genomes of different species. And not one of these hits is from the common cold (rhinovirus).

While any two humans have >90% of the same sequence in the same spot in their genomes, we can differ at millions of places because our genomes are large– 3,400,000,000 letters large. The likelihood of four DNA letters sufficiently distinguishing one POPULATION from another is absurd. We researchers have to keep sequencing thousands of persons’ genomes to figure out the baseline level of difference between populations, and we already have millions of places people differ from each other.

For a show about hackers, they sure missed a great opportunity to show a screen with gobs of information on it. Why not have the biohazard bag with a long, complex string of letters, have Sly take a photo, send it to the team, opening the door to a quick gag about texting-while-driving, and transmitting the data? This source sequence could come from the publicly available databases of cold virus genome sequence.

This choice the writers made seems small, but it could very well affect how people perceive bioweapons and their feasibility. The concept of targeting a specific genetic marker of a population with a virus or other drug is the dream of many a comic book despot, but it’s not there yet. We’re not good at distinguishing populations based on their genes. We need millions of datapoints to classify people even roughly on their backgrounds. Even the commercial genealogy tests based on these millions of datapoints are only good enough to give a score for what percentage of your genes can likely be traced back to one general geographic area.

So this “virus,” with its teeny string of wrong DNA letters would not be able to do much of anything, let alone target a specific population for a nasty cold.

Posted in Snarkings | Leave a comment

Beating Cancers’ Unexpected Vice: Transcription – at

Share this post:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Reddit
Many thanks to Tom Dunne from Sigma Xi for this cartoon of vice-ridden cancer cells.

My second post over at American Scientist’s renovated blog network tackles some of the latest thinking about a new process to target that is especially important for cancer cells. Targeting gene transcription– making mRNA from DNA genes– does a number on some of the nastiest types of cancer. Head over to my post to find out how and why.

Posted in Work I've Done | Leave a comment

Mutations that don’t break proteins – at

Share this post:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Reddit


I’ve been given the opportunity to write for’s renovated blog network. Head over and check out my first post on how not every disease-causing DNA variant breaks a protein-coding gene.

Posted in Uncategorized | Leave a comment

Snarking Stem Cells: The Webinars

Share this post:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Reddit

How does the same genome make different cell identities?

Got half an hour for stem cells? I’ve been working with the International Society of Stem Cell Research (ISSCR) to develop a four-part webinar series on what everyone needs to know about stem cells and stem cell biology.

The first session is half an hour, and it explains how your cells, which received the same genes, can adopt many different jobs and identities with the end goal of making you. The session is up on their “Stem Cells in Focus” website. Yes, to view, you have to register, but registration is free, and they don’t spam your email address.

Give it a watch, and send me your comments/questions, which I’ll try to append to this post with responses.

FYI, the next one is coming up on March 24, so join us then to hear about the basics of reprogramming cells.

Posted in Uncategorized | Leave a comment