The Pill That Could · The Translocation

The Translocation, part 3

The second thing was on chromosome 9.

Chromosome 9 is, in a normal cell, recognizable by its banding pattern. The long arm has its own characteristic sequence of light and dark stripes, and the very end of the long arm is, by Q-banding, a region that fluoresces brightly. In the CML cells, chromosome 9 looked almost normal — but with something extra at the end of the long arm. A pale band, beyond what should have been the terminal bright band. A piece of something attached.

Rowley measured. The extra material on chromosome 9 was about the same size as the piece that was missing from chromosome 22. The banding pattern of the extra material — once she had it isolated cleanly enough to read — matched the banding pattern of the missing tip of 22.

The piece that was supposed to be on chromosome 22 was on chromosome 9.

It had not been deleted. It had been moved.

* * *

This kind of move is called a translocation. A piece of one chromosome detaches and reattaches somewhere else.

Translocations come in two varieties. In a non-reciprocal translocation, a piece of one chromosome moves to another, and the donor chromosome simply loses what it gave away. In a reciprocal translocation, two chromosomes exchange pieces — a piece of A goes to B, and a piece of B goes to A — and both chromosomes are altered. The cell now has the same total amount of DNA it had before; it has just shuffled some of it.

Rowley's reading of the Philadelphia situation was that it was reciprocal. The piece missing from chromosome 22 was on chromosome 9, but chromosome 9, looking carefully, also had something extra at the tip of its long arm — and the small chromosome 22 had received, in exchange, only a tiny piece of chromosome 9, which is why the chromosome ended up shorter overall. The two chromosomes had swapped pieces. 9 and 22 had each broken on their long arms; the broken-off pieces had stuck to the wrong partner; the cell had carried on with the reshuffled set.

In the language that would become standard, the Philadelphia chromosome was a t(9;22) — a translocation between chromosomes 9 and 22 — with breakpoints in specific bands (q34 on chromosome 9, q11 on chromosome 22, in the eventual nomenclature).

This was a reinterpretation of a well-known abnormality, not a discovery of a new one. But the reinterpretation mattered. It changed the question from “what does it mean for a piece of chromosome 22 to be lost?” — a question with no obvious experimental approach in 1973 — to “what does it mean for a piece of chromosome 22 and a piece of chromosome 9 to be joined?” — a question with a much sharper edge, because joining two pieces of DNA can create something specific, namely a junction. At the junction, two sequences that should never have been near each other are now adjacent. If a gene happens to span the junction, you get a fusion gene. If the gene downstream of the junction is normally regulated by the sequences it now finds itself next to, it will be regulated differently. Either way, the junction is the kind of thing you can, in principle, investigate.

You couldn't investigate it in 1973. The tools weren't there. But you could now state, clearly and correctly, what kind of investigation would be needed.

* * *

Rowley's paper appeared in Nature in June 1973. Like the Nowell and Hungerford paper before it, it was short — a one-page letter, with a small figure showing the banded chromosomes from a CML patient and an arrow pointing to the extra material on chromosome 9. The paper's title was careful: “A New Consistent Chromosomal Abnormality in Chronic Myelogenous Leukaemia identified by Quinacrine Fluorescence and Giemsa Staining.” The word “new” referred to the previously unrecognized abnormality on chromosome 9, not to the Philadelphia chromosome itself.

The paper's reception was mixed. Some readers — the ones already thinking about cancer in chromosomal terms — saw immediately what it meant. Others were skeptical that translocations could really be reproducible features of cancer; the standard view was that cancer cells were chromosomally unstable in a generic way, and that a specific translocation appearing in patient after patient was probably a coincidence or an artifact. Rowley herself encountered some of that skepticism firsthand. An earlier paper of hers, on a different translocation in a different leukemia, had had a difficult time finding a home before being published.

What changed the field's mind, over the next several years, was that Rowley kept finding translocations. So did other people. The list grew.

Rowley had already identified, slightly earlier, a t(8;21) translocation in a subset of patients with acute myeloid leukemia — a different leukemia from CML, with a different translocation, but also reproducible. By the mid-1970s, the picture was filling in: Burkitt lymphoma was associated with a t(8;14); follicular lymphoma with a t(14;18); a particular subtype of acute promyelocytic leukemia, identified later, with a t(15;17). Each cancer had its own characteristic chromosomal swap. The swaps were not random. They appeared in the same patients, the same diseases, the same breakpoints, over and over.

It became increasingly difficult to argue that the translocations were incidental. The simpler explanation — the only explanation that fit the accumulating data — was that the translocations were causal. Something specific happened at the breakpoint that turned a normal cell into a cancer cell. Different breakpoints in different cancers meant different specific events. The genetics of cancer, which had been an abstract question in 1960 and an open question in 1970, was starting to look, by the late 1970s, like a tractable problem with specific cases.

* * *

What this meant, by the end of the 1970s, was that the field had a new question.

The old question, the one that had been frozen in place since 1960, had been: what does the Philadelphia chromosome do? The new question, made possible by Rowley's work and the cascade of similar findings in other cancers, was: what gene is at the breakpoint?

The two questions are not the same. The old question treated the chromosome as the unit of analysis — a chunk of stuff missing or rearranged, in some general way. The new question treated the gene as the unit of analysis — a specific piece of DNA, with a specific function, broken or fused or moved in a specific way. The new question pointed at experiments. If you knew where the breakpoint was on chromosome 22, you could try to find the gene at that location. If you knew where it was on chromosome 9, you could try to find the gene there. If both were genes, and the translocation was joining them, you might be able to identify the fusion gene and learn what protein it made.

This was a research program. It was not a research program that anyone could pursue in 1973 with any realistic hope of success — the tools were too crude — but it was a research program that the field could see, even from where it was standing. The technology to investigate genes at the molecular level was being developed in parallel, in other labs, for other reasons. Recombinant DNA techniques — cutting DNA at specific sequences, copying it, moving it from one organism to another — were being worked out in the early 1970s. DNA sequencing, the technique of reading the letters of DNA one by one, was still primitive but was getting better. The tools that would eventually let you go from “the breakpoint is in band q11 of chromosome 22” to “the breakpoint is in the middle of this specific gene, and here is the sequence of the resulting fusion protein” were, in the early 1970s, not yet in working order. They would be, soon.

Rowley herself would continue to work on cancer cytogenetics for the rest of her life. She would identify or co-identify a number of additional cancer-associated translocations over the next two decades. She would receive, eventually, the National Medal of Science, the Lasker Award, and the Presidential Medal of Freedom — most of the major scientific honors short of the Nobel, which she did not get. She would die in 2013, at eighty-eight, having lived to see the small chromosome from her photographs become the molecular target of a drug that worked.

The drug was a long way off in 1973.

* * *

Here, then, is where the story stands at the end of the 1970s.

Cancer is genetic, in a real and specific sense. The genetics can be read, with the right tools, in the chromosomes of cancer cells. Specific cancers have specific chromosomal rearrangements, reproducible from patient to patient. In CML, the rearrangement is a swap between chromosomes 9 and 22 — the swap that creates the Philadelphia chromosome, and that has, somewhere in the swap, joined together two pieces of DNA that should not be joined.

What is at the joint?

What protein, if any, does it make?

What does the protein do, that it would turn a stem cell in a patient's bone marrow into a stem cell that produces a slow-rising flood of abnormal white blood cells, and eventually kills the patient?

The next chapter is about people figuring out the answers to those questions. It takes place in the 1980s, in laboratories on three continents, using techniques that did not exist in 1973. By the end of it the field will know, in the precise language of biochemistry, what the Philadelphia chromosome does.

Knowing what it does, of course, is not the same as being able to fix it. But it is the necessary first step.

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