杏吧原创

THOUSANDS have been granted over the past decade, not just to industry, but to universities and research institutes as well. Nevertheless, is it right?

To date, the debate over the patenting of human DNA has largely consisted of stating and restating entrenched positions: industry, keen to protect investment, is on one side; on the other a mix of NGOs, scientists and other people, who see human DNA as a common heritage that should be protected from commercial exploitation. But there is a tenable middle-ground position, and last week my colleagues and I attempted to outline it in a new report from the Nuffield Council on Bioethics.

Our report does not say that patenting genes is wrong in principle. What it concludes is that far too many gene patents are being granted by a system that is failing to apply the rules strictly enough. The purpose of patents is to stimulate innovation for the public good and reward people for new inventions. But many new patents for human DNA are likely to impede innovation and create powerful monopolies capable of charging over the odds for tests and drugs based on human genes. Patents involving human DNA should be granted only in rare cases. They should be the exception rather than the norm.

How did we reach these conclusions? The most common objection to this type of patent is that human genes occur naturally: they are there to be discovered and not invented. This is not persuasive. Isolated DNA does not occur naturally, and without isolating and cloning a gene, you cannot decipher its sequence. Moreover, the patent system has long recognised useful applications of discoveries as inventions.

So patent offices are right to conclude that DNA can be considered part of an invention. But it does not follow from this that all patents on all human genes should be allowed in all circumstances. The key question is where to draw the line. Inventions must pass three tests to qualify for a patent. They must be useful, novel and not obvious to someone familiar with the field. To look at how these criteria should apply to genes, we distinguished between four different uses of DNA sequences: in genetic tests, as research tools, in gene therapy and for producing therapeutic proteins.

First, genetic testing. Heated controversy continues over an American company called Myriad Genetics and its patent for the breast cancer genes BRCA1 and BRCA2. The company is charging $2,400 a time to test patients and claiming royalties from others who offer tests based on the same gene. But finding a link between a gene and a disease is not itself an invention: it is a discovery. And once such a link comes to light, it is obvious that the gene might form the basis of a genetic test. There is little inventive about it. That is why we concluded that gene patents based on claims about diagnosis should seldom be granted.

The same reasoning should apply to patents that are based on claims about gene therapy. Once a gene has been linked to a disease, the notion of using it in a treatment is obvious and shouldn鈥檛 merit the reward of a patent. Patents should be reserved for those who invent safe and effective methods for getting genes into tissues.

Also to be discouraged are patents on genes of unknown medical value. These days, scientists can identify human genes by trawling through databases of human DNA and making 鈥渂est guesses鈥� about the biological functions and potential uses of the genes they find there. But without further work, such genes are research tools, not inventions. And past experience suggests that allowing speculation to pass for actual evidence of usefulness is a recipe for hindering research.

The salutary lesson involves the human gene for a receptor known as CCR5. In 1995, a US company called Human Genome Sciences applied to patent the gene for its usefulness in combating viral infections, solely on the basis of its similarity to known DNA sequences. Shortly after, researchers elsewhere discovered that the receptor was HIV鈥檚 passport into cells. So because of the speculative patent it holds, Human Genome Sciences is entitled to levy royalties from companies that use the receptor to look for potential HIV drugs.

One type of gene patent is acceptable, however. Companies and labs should be allowed to own the rights to genes whose protein products are used directly as medicines, such as human insulin or erythropoetin. The information encoded in such genes is being used to make something of value, and it is clearly in the public interest to create incentives to encourage the costly process of developing a medicine. But since genes often carry the instructions for more than one protein, there is a caveat. The rights to the gene should only extend to one protein.

Are these recommendations realistic? Most do not require new laws. They could be achieved by applying the existing criteria for patents more stringently. Without this, we face the prospect of more attempts to monopolise genes, more high prices for gene tests, and the tying up in legal red tape of ever more human DNA sequences.