The Pill That Could · BCR-ABL

BCR-ABL, part 3

This was the question de Klein was asking. She was working in the laboratory of Gerard Grosveld, a molecular biologist at Erasmus University, with collaborators including Ad Geurts van Kessel and others, and in close communication with a group at the National Cancer Institute in Frederick, Maryland — John Groffen and Nora Heisterkamp, working in Stephen O'Brien's lab — who were doing complementary experiments.

The technique de Klein used was somatic cell hybridization. You take a human cell and a mouse cell and you fuse them together. The resulting hybrid cell, in the early days of growth, has all the chromosomes from both parents. As it divides, it tends, for reasons that are a little arcane, to lose human chromosomes more or less at random while keeping the mouse ones. After many divisions you have a population of hybrid cells, each of which retains only some of the original human chromosomes. You can sort the cells, find ones that have, say, human chromosome 22 but not human chromosome 9, or human chromosome 9 but not 22, and ask: does the human ABL gene track with chromosome 9, or with chromosome 22?

In normal human cells, you would expect ABL to track with chromosome 9, because that is where it lives. In cells from a CML patient, with the Philadelphia translocation, the answer might be different. If the translocation broke the chromosome at a point distal to the ABL gene — farther out toward the tip than ABL — then ABL would still be on chromosome 9, where it had always been. If the translocation broke the chromosome at a point closer to the centromere than ABL, then the piece carrying ABL would have moved with the rest of the distal long arm to chromosome 22.

De Klein's blot showed ABL on chromosome 22.

In the hybrid cells from CML patients, the ABL gene tracked with the small Philadelphia chromosome — the one that should not have it — not with the normal-looking chromosome 9. The translocation had broken chromosome 9 on the centromere side of ABL, picking up the ABL gene along with the rest of the distal long arm and depositing all of it onto chromosome 22.

This was the gene at the breakpoint. The Philadelphia translocation, in 1982, finally had a known piece of cellular machinery on at least one of its sides.

* * *

The paper appeared in Nature in November 1982, with seven authors and a careful, narrow title: “A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukaemia.” It was a clean result, and the field absorbed it quickly. The cellular cousin of v-Abl was on the Philadelphia chromosome. ABL, a tyrosine kinase that was already on the field's list of suspects in cancer biology, had been moved by this specific translocation in this specific disease.

The next question was the partner. ABL was now sitting on chromosome 22, but where exactly? The translocation breakpoint on chromosome 22, in band q11, must have happened in or near some other gene. What was that gene?

Heisterkamp and Groffen, with collaborators that included the same Rotterdam group, took up this question over the next two years. They cloned the breakpoint region from chromosome 22, working outward from where ABL had landed. They found something striking: the breakpoints from different CML patients all clustered in a relatively small stretch of DNA on chromosome 22, no more than a few thousand letters long. Different patients had breaks in slightly different places, but always within this one tight region.

They named the region the breakpoint cluster region, abbreviated bcr. The gene that contained this region was named, with the same letters, BCR. By 1984 the picture had filled in: the t(9;22) translocation broke the ABL gene on chromosome 9 in one of two introns (an intron being a non-coding stretch within a gene that gets spliced out before the gene's protein is made), and broke the BCR gene on chromosome 22, also typically in an intron, and joined them. The resulting fused gene, which sat on the small Philadelphia chromosome, was BCR on its 5' end (the upstream end, the part that gets read first) and ABL on its 3' end (the downstream end). When the cell read this fused gene and made a protein from it, the protein was a hybrid: BCR up front, ABL behind.

It was a fusion gene. Just as the new questions, in the late 1970s, had suggested it might be.

* * *

Now what did it do?

The first answers came from Owen Witte's lab. Witte had trained as a postdoctoral fellow with David Baltimore at MIT in the late 1970s, working on the Abelson virus protein, and had been part of the work showing that v-Abl was a tyrosine kinase. By the early 1980s he had moved to UCLA and set up his own lab, where he and a graduate student named John Konopka turned their attention to the BCR-ABL fusion protein.

They wanted to know two things. First, did the fusion protein have kinase activity, the way both v-Abl and normal ABL did? Second, if it did, was the activity regulated normally, or was it stuck in the on position, the way v-Abl's activity was?

The answer to the first question was yes. The fusion protein, which they showed was a single polypeptide of about 210 kilodaltons (a kilodalton being a unit of protein mass; 210 kilodaltons makes BCR-ABL a fairly large protein, but well within the normal range), retained the kinase domain of ABL and could phosphorylate substrates in the test tube. The ABL part of the fusion was still doing its enzymatic job.

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