American Scientist

Tumor cells, like many lost souls, have vices that push them toward misbehavior. However, the vices tumor cells need are hard to shut down completely and systemically without harming innocent bystander cells. Being able to specifically target such processes depends on what exactly that vice is. Surprisingly, researchers are finding that gene transcription—an essential step to making all proteins in all cells—can be abused by tumor cells that transcribe the wrong or broken genes. Experimental drugs targeting this process have already shown promising results in treating tumors.

Most anti-cancer drugs work by inhibiting the tasks that tumor cells do differently than normal cells. Tumor cells can abuse normal activities like sending and receiving protein signals, growing, or recruiting blood vessels. Many labs focus on developing treatments that target and inhibit these processes with small chemical compounds. This protocol has given us hormone substitute drugs like tamoxifen that counteract estrogen signals in estrogen receptor–positive breast cancer; it’s given us erlotinib, which halts the growth signals and prevents cell proliferation in lung cancers; it’s given us drugs like bevacizumab that halt construction of blood vessels sustaining tumors in the colon. But these vital processes—sustaining signaling, relentless growing, and inducing blood flow—are all processes essential to tumor cells, which they can easily evolve to misuse.

Not all targetable processes differ so obviously between healthy cells and cancer cells, and it’s become surprisingly hip to target a process that all cells do—transcribe DNA genes into mRNA transcripts. Transcription of certain genes or the process of transcription itself seems to be especially important for certain tumor cells. In a way, transcription is involved in all processes in a cell. Without it, no new genes make mRNA copies, so the pool of resulting proteins becomes depleted. These proteins receive hormone signals or convince the body to form new blood vessels. So targeting transcription could affect all of the vices of a tumor cell. So far, inhibiting the processes involved in transcription has a significant impact on survival of certain cancers but doesn’t seem to adversely affect normal cells in culture or in mice treated with these chemical compounds.

All cells have the same genes in their nuclei, but to make the different kinds of cells in a human body, only certain genes are transcribed into mRNA and translated into protein. Cells can differ from each other in at least two ways:

1. if they’re two different kinds of healthy cells—for example, white blood cell versus skin cell;
2. or they can be healthy or cancerous—white blood cell versus leukemia.

Because cancer arises from preexisting cells with preexisting identities, they retain aspects of this identity and acquire some additional traits through mutation that can either alter protein quality or protein amount.

The traits a cell has and the jobs it does come from the sum of all gene transcription. A complex of several proteins forms a molecular machine that transcribes DNA genes into mRNA transcripts. This machine, called RNA polymerase, takes the information permanently stored in DNA and encodes it in a slightly different, less long-lived, chemical format used as a scratch blueprint to produce a protein.

Different individual tasks in transcription are handled by different subunit proteins of the RNA polymerase II holoenzyme (RNA Pol II), which is the most commonly used polymerase. For instance, one protein subunit can recognize where to start transcribing, and another can help unwind DNA. A chain of amino acids pokes out from the back end of one of the subunits and acts as a kind of status reporter or gauge for what the polymerase machine has clearance to do. Other proteins called kinases add phosphate chemical groups to certain amino acids on this tail to indicate that the polymerase is getting ready to bind, or fully licensed to transcribe a gene.

One such kinase, cyclin-dependent kinase 7 (CDK7), is part of a vast web of transcriptional regulators that eventually feed into whether or not RNA Pol II can transcribe a gene. It adds phosphate groups to many proteins including regulators of the cell cycle and the tail of RNA Pol II.

A group at the Dana-Farber Cancer Institute led by Dr. Nathanael Gray developed a chemical compound that binds CDK7 and prevents it from adding phosphate groups to other proteins. To do this, they had to solve some common problems with targeting members of the CDK family, because each of these proteins has very similar structures. The compound, THZ1, does a great job of either killing or halting growth of cancer cells in a dish.

The mechanism by which THZ1 has this effect is a little more complicated, a matter I’m all too familiar with as one of the team members researching what THZ1 is doing to cells. We want to know what vice it is inhibiting, and why it isn’t doing so in healthy cells. Tests in three different types of cancer suggest variants on a similar hypothesis: THZ1 inhibits the genes controlling the cancer cell identity.

Not all cancers are equally destructive. So, with our collaborators at Dana-Farber, we have tested THZ1 in T cell leukemia, small cell lung cancer, and neuroblastoma. On the surface, these kinds of cancer are not that similar. They share few common mutations, and they arise from different cells of origin. They are, however, generally difficult to treat and lack the kinds of targets used to treat other cancers. These projects show that they are also particularly sensitive to THZ1.

Cells start dying from THZ1 at so low a dose that transcription of only a few genes is really affected. If the cells die when we are affecting transcription of only certain genes, something about these genes must be playing a role in the cells’ survival. In this way, not only are we finding out that affecting transcription of certain genes is an essential vice of these cells, but we can see what exactly these cells depend on.

Across the tested cancers, the genes whose transcription is shut down aren’t the same, so THZ1 doesn’t just hit one “master” gene that controls its effect. It hits genes that are important for that particular kind of cancer. In T cell leukemia, it hits the oncogene network that keeps cells too immature. In small cell lung cancer, it hits the proteins that regulate transcription. In neuroblastoma, it hits the most problematic and frequently duplicated gene, and its carcinogenic effect disintegrates.

Some of the vices tumor cells abuse are still beyond our reach. Identifying them has already shown common themes—most importantly that they can be stopped by cancer-fighting drugs. Like all living things, cancer cells will take advantage of whatever they can to survive by evolving defenses as their environment selects for advantageous mutations. These defensive processes or proteins paint a target on these cells’ proverbial backs. Studies with THZ1 and other transcriptional therapeutics are showing that, not only can tumor cells become addicted to transcription of certain genes, but also that treating this vice can help the body come clean.