The Pill That Could · BCR-ABL

BCR-ABL, part 4

The answer to the second question was also yes — in the bad direction. The fusion protein was a more active kinase than normal ABL was. The BCR portion at the front of the fusion was somehow disrupting the regulation that normally kept ABL quiet. (We now know, in some detail, how: the BCR sequences cause the fusion protein to dimerize — to pair up with another copy of itself — and the dimerization triggers the ABL kinase domain to activate the way it would normally activate only in response to the right cellular signal. In 1985 the molecular details weren't yet that clear, but the functional answer was.) BCR-ABL was a tyrosine kinase whose activity was no longer being properly turned off. It was, in the language we now use without thinking, constitutively active — stuck in the on position, phosphorylating its substrates whether or not the cell wanted it to.

This was, as a finding, exactly what the field had been hoping it would be.

* * *

There was one more question, and it was the question that mattered most for whether any of this would translate into anything useful for patients. Was BCR-ABL, by itself, sufficient to cause CML? Or was it just one piece of a more complicated puzzle — a contributor to the disease, but not the whole story?

The question is harder to answer than it sounds. CML, in a person, develops over years, in the context of all the other things that happen in a complex organism. You can show that a fusion gene exists in CML cells. You can show that it makes a fused protein with abnormal activity. But to show that the fusion is the cause — that it's enough to make CML happen, with nothing else required — you need to put just the fusion into a cell that doesn't already have it, and watch what happens.

In 1990, a paper appeared in Science that did this. The first author was George Daley, a graduate student in David Baltimore's lab at MIT (Baltimore by this point being one of the central figures in molecular biology, having shared the Nobel Prize fifteen years earlier for the discovery of reverse transcriptase). The other lead author was Richard Van Etten, a postdoc in the same lab. Their experiment was, in concept, straightforward.

They built a retrovirus carrying the BCR-ABL gene. (A retrovirus is a virus that, when it infects a cell, inserts its genetic material into the cell's DNA; the Abelson virus from the start of this chapter is a retrovirus, and so is HIV. Researchers had learned to take advantage of this property by removing the disease-causing parts of certain retroviruses and replacing them with whatever gene they wanted to deliver.) They took bone marrow from healthy mice. They infected the bone marrow cells with the BCR-ABL retrovirus, so that some fraction of the marrow cells now had the BCR-ABL gene inserted in their genome. They transplanted the infected marrow into mice whose own marrow had been wiped out with radiation — mice that, without the transplant, would have died of marrow failure within days.

The transplanted mice survived the marrow failure, because the donor cells took over and started making blood. And then most of them got sick. The disease they developed looked, in clinical and pathological detail, like CML: a high white blood cell count, a swollen spleen, infiltration of the bone marrow with too many immature myeloid cells, and — in mice that lived long enough — progression to something resembling blast crisis. The mice were dying of a disease that BCR-ABL alone, expressed in normal mouse marrow cells, was producing.

The case was as close to closed as biology gets. CML, the disease, was caused by BCR-ABL, the fusion gene. One translocation, one fusion, one constitutively active kinase. Everything that went wrong in a CML patient — the slow rise in white cell counts, the enlarged spleen, the eventual transformation into something acutely fatal — was downstream of one stuck-on enzyme.

* * *

This is the picture as the 1980s ended.

A reciprocal translocation between chromosomes 9 and 22, occurring in a single cell at a single moment in a single hematopoietic stem cell in a person's bone marrow, fuses the 5' part of the BCR gene to the 3' part of the ABL gene. The fused gene makes a fused protein, p210 BCR-ABL, in which the BCR portion forces the ABL portion into a permanently active state. The protein is a tyrosine kinase — an enzyme that attaches phosphate groups to tyrosine residues on its substrates. The substrates are proteins involved in cellular pathways that drive proliferation, suppress programmed cell death, and modify how the cell interacts with the bone marrow environment. Because the kinase is always on, those pathways are always active. The cell carrying the fusion grows when it shouldn't grow, divides when it shouldn't divide, fails to die when it should die. Its descendants — a whole lineage of myeloid cells, all carrying the same fusion — do the same. Over months and years, they accumulate. They expand into the bone marrow. They spill into the blood. They eventually transform into something more aggressive and kill the patient.

All of this is downstream of one enzyme.

If you could turn the enzyme off, you might be able to turn the disease off.

* * *

Brian Druker is, in 1990, finishing his postdoctoral fellowship in Tom Roberts's laboratory at the Dana-Farber Cancer Institute in Boston. He has been there since 1984. He was trained, originally, as a medical oncologist; he came to Dana-Farber because he wanted to understand cancer at the molecular level rather than just treat it, and Roberts's lab was a place where you could do that. He has spent his fellowship developing antibodies that recognize phosphotyrosine — antibodies that, in essence, light up tyrosine-phosphorylated proteins so you can see them and measure them. The antibodies are good. Other labs have started asking him for them.

He has also been reading the BCR-ABL literature.

The picture, by 1990, is the one this chapter has just laid out. Druker can see it as clearly as anyone in the field. He can also see, more clearly than most, what it implies. The disease is one enzyme. The enzyme is in a class — tyrosine kinases — that does its work through a particular chemical reaction. Chemical reactions can, in principle, be inhibited by small molecules that fit into the active site of the enzyme. If you could find a small molecule that fit into the active site of BCR-ABL, and didn't fit into the active sites of the kinases that the body needs in order to function, you might have a drug that turned off CML without killing the patient.

This is the obvious idea. It is also, in 1990, an idea that essentially nobody at any major pharmaceutical company is willing to take seriously, for reasons we'll come to in the next chapter.

There is, however, one chemist Druker has heard of, working at a Swiss company called Ciba-Geigy, who is making compounds that look as if they might be selective inhibitors of particular kinases. The compounds are intended as research tools, not as drugs. The chemist's name is Nick Lydon. The compounds are sitting on a shelf in Basel.

Druker writes him a letter.

All 13 chapters of The Pill That Could · Library · All courses