SCIENCE JUNE 13, 2013
The Supreme Court decided today that human gene sequences cannot be patented, invalidating some of Myriad Genetics's patent claims on the breast cancer genes BRCA1 and BRCA2. But it was only a partial victory for scientists and patients' rights advocates—the Court upheld the company's patents on the cDNA versions of those genes. What's cDNA? Clearly the Court doesn't understand, or else wouldn't have made such a ruling.
For a long time, biologists have known that the DNA all coiled up in your cells is mostly useless. DNA's job is to serve as a blueprint for proteins—but about 98 percent of your genetic material doesn't code for anything. When a cell's machinery goes to read the DNA, it just ignores those gibberish bits, which are called introns. (Lots of research has gone into learning about what, if any role these mysterious introns play. There are some theories, but the evidence is mixed. Humans have a ton of introns, but some species have very few and seem to function just fine.) The clean version of the blueprint, with all the junk taken out, is called complementary DNA, or cDNA.
The BRCA1 and BRCA2 genes are regions that code for for proteins that help repair DNA. If these genes are mutated, then the proteins don't come out right, and the body has a harder time fixing damaged DNA—which dramatically increases a person's risk for cancer. Myriad's patents would have prevented anyone from handling the genes, or even testing for mutations in them. The Supreme Court rightly dismissed those claims. Myriad only discovered the location of those genes. It didn't create them, it found them. As Justice Sotomayor pointed out in the oral argument, the law "doesn't patent ideas."
But like most genes, BRCA1 and BRCA2 contain a ton of introns, and the Court said that its patents on the clean versions of the genes—the cDNA—are fine, even though they're basically the same thing. The reasoning that the Court used is bizarre, but to understand it, we first have to understand how cDNA is made.
The journey from DNA to protein, which is called gene expression, involves a complicated copying process. First the DNA is unwound (remember, it's coiled up in big X-shaped chromosomes) and unzipped. A working copy of the data is created using RNA—think of RNA as scratch paper—and the body then goes to work taking out all the introns and putting together all the useful bits. The final copy is called the mature mRNA transcript and it's sent off to get turned into a protein.
Scientists who want a clean copy of a gene, then, can intercept the mRNA transcript and figure out what it says. They can even use the mRNA to create complementary DNA, which has the same sequence. The cDNA is what our DNA would look like if it were purged of all introns. But it doesn't exist naturally because our bodies like to keep our DNA safely locked away. Only RNA copies of our DNA go through the process of having the junk bits cut out. So cDNA is an artificial creation, even though it contains the same data as the mRNA, which is a natural creation.
That's a fine distinction, but it was good enough for the Supreme Court, which notes that "the lab technician unquestionably creates something new when cDNA is made." Because cDNA is something manmade, unnatural, it is therefore patentable. Of course, the Court was ruling on a very narrow part of patent law, which says that "products of nature" can't be patented. Someone could still argue that cDNA fails the "obviousness" test, as the body does all the work to remove introns—the researcher just copied it back into DNA form.
Still, saying that cDNA is patentable but natural DNA isn't misunderstands the central complaint about gene patents, which is that genes are basically information, and information can't be patented. The body has its own code, it's a natural code, and the body naturally manipuates that code, making copies, edits and deletions all on its own. Genetics is data. The provenance of the molecules that carry that data, whether they are DNA or RNA, whether the DNA version exists naturally or the RNA version exists naturally, is completely irrelevant.