The Pill That Could

Chapter 3: BCR-ABL

In a laboratory at Erasmus University in Rotterdam, in the autumn of 1982, Annelies de Klein is holding an X-ray film up to a lightbox.

The film shows a column of dark bands floating on a gray background. Each band is the answer to a small question: does this fragment of DNA, taken from this kind of cell and cut in this particular way, contain the gene she is hunting for? The dark bands say yes. The blank spaces say no.

She has been doing this experiment, on and off, for months. The technique is called Southern blotting, after Edwin Southern, the British biologist who developed it in the mid-1970s. It is, in 1982, the standard way to ask whether a specific gene is present in a sample of DNA. You take the DNA out of the cells, you cut it into fragments using an enzyme that recognizes a specific stretch of letters, you separate the fragments by size in a slab of gel, you transfer them onto a sheet of nitrocellulose, you wash the sheet with a small radioactive piece of the gene you're looking for, and you put the whole thing against an X-ray film overnight. If the gene is there, the radioactive piece sticks to it. The film, when you develop it the next morning, shows a dark band.

The gene de Klein is hunting for is called ABL.

What she is finding is that ABL is in the wrong place.

* * *

To make sense of why this matters, we need to back up several years and follow three threads. They start in different places in the 1970s and come together, by 1982, into a question that one experiment can answer.

The first thread is a virus that causes leukemia in mice. The second is a discovery about how cancer-causing genes get to be cancer-causing. The third is a piece of unexpected biochemistry — a chemical reaction nobody knew cells were doing, found in the wrong place by accident.

By 1982 these threads have braided together enough that a careful scientist in Rotterdam can ask whether a particular known gene sits at the breakpoint of a particular known translocation. By 1985 she'll have the answer to a sharper version of the same question. By 1990 the field will have shown, in mice, that a single fused gene at that breakpoint is sufficient — by itself, with no other help — to cause a disease that looks like CML.

The threads are worth taking one at a time.

* * *

The virus first.

In 1969, a researcher named Herbert Abelson, working at the National Institutes of Health, was studying viruses that cause leukemia in mice. He had a strain of one such virus that, when injected into young mice, caused an unusually fast and aggressive disease — a lymphoma involving immature B-cell precursors, killing the animal within weeks. The virus was eventually named the Abelson murine leukemia virus, or A-MuLV, and the gene inside it that was responsible for the cancer-causing activity was given the short name v-abl — the v for viral, the abl for Abelson.

This was, in 1969, a particular kind of finding. The field had known for decades that some viruses could cause cancer — Peyton Rous's chicken sarcoma virus, the one that took the field fifty years to take seriously, was the original example. By the late 1960s, several such viruses were known: the Rous sarcoma virus in chickens, several murine leukemia viruses including A-MuLV, an avian leukosis virus complex, a feline leukemia virus. Each one had a small genome — just a few genes — and each one carried, somewhere on that small genome, a single gene that was responsible for the cancer-causing activity. These genes were called oncogenes, from the Greek onkos, a swelling or mass.

v-abl was the oncogene of A-MuLV. v-src, the cancer-causing gene of the Rous sarcoma virus, was the most-studied. There were, by the early 1970s, perhaps a dozen viral oncogenes catalogued. Each was named after the virus that carried it. None of them, at this point, had been connected to anything in the cancers of humans.

The question that hung over the field in 1970 was: where do these genes come from?

* * *

That is the second thread.

In 1976, two researchers at the University of California, San Francisco — a virologist named Harold Varmus and a biochemist named J. Michael Bishop — published a paper in Nature that answered the question, at least for one virus, in a way that nobody quite expected.

They had been working with v-src, the chicken-virus oncogene. They made a small radioactive piece of v-src — essentially the same kind of probe de Klein would later use, made a few years later — and used it to ask whether anything in normal chicken DNA bound to it. If v-src was a uniquely viral creation, nothing in chicken cells should match. If it had been picked up from a cell at some point in the virus's evolutionary past, something should.

Something did. Normal, healthy, non-cancerous chicken cells contained a gene that was almost identical to v-src. Then they tried it in other species. Quail had a similar gene. So did mice. So did fish. So did humans. The gene that the Rous sarcoma virus had been carrying around in its tiny viral genome was, in some only slightly altered form, present in essentially all vertebrates. It had a normal function in normal cells. The virus had picked it up at some point — perhaps tens of millions of years ago, perhaps more recently — and kept a slightly modified version, and that slightly modified version was what caused cancer.

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