'The' gene? For what?

There is yet another story in the news these days, about a diabetes drug that apparently raises the risk of heart disease. In this case, the chemical known as rosiglitazone binds to a receptor called PPAR-gamma in fat cells. Variation in a PPAR-gama gene has been found by several studies to be associated with higher risk of type 2 (adult-onset) diabetes, the form of the disease in which cells do not respond to circulating insulin, thus leading to elevated glucose in the blood, and to pathological consequences.

However, there is some evidence accumulating, according to the news stories, that rosiglitazone may increase the risk of heart attacks.

Whether or not this turns out to be true, it is but one of many stories about side effects. Indeed, as a follow-up story suggest, there are some boys-will-be-boys elements of questionable ethics--where the company was hesitant to accept or acknowledge at least one study that showed the elevated heart-attack risk--resembling many other stories of yielding to heavily vested interests, an aspect of the ethics of science that could use some serious scrutiny. Again, we see the news stories but can't claim to know what the preponderance of the facts are in this case.

Separate from any ethical questions, most drugs, if you read the small print, can have many side effects. The small print in drug inserts may contain several anti-lawyer rather than true biological effects of the drug, since if an effect is rare in a drug trial it will be difficult to prove a true casual connection.

Above is one image of one section across one stage (mid-gestation) of a mouse embryo, stained (purple) to show cells expressing PPAR-gamma. This is from the great gene-usage site GenePaint.org that shows the mouse mid-gestation expression of around 20 thousand genes. Go there and you can look at all sections in two different mouse embryos available for this gene (here's one).

The point is that even at one stage in one strain of animal, the gene is expressed in many tissues. Ironically, but not necessary atypically, the gene is not expressed in the heart at this stage. However this is part of a system of energy-related cellular functions that are used by most if not all cells, and the gene is in fact expressed in most cells at some level (see Wikipedia for PPARg).

Regardless of how the drug-reaction story turns out, it and many other like it show a point that we tried to make in The Mermaid's Tale, and that we were by no means at all the first to observe: many genes--probably the vast majority of genes--are expressed in several different tissues at different times or under various circumstances. Signaling and related cell-communication and gene-expression regulating systems (of which PPAR is an example), are often used in many different types of cells. It is not the gene, but the combination of genes, that has functional effects.

Thus, it is totally to be expected that a drug that targets a gene product can, or probably will, have multiple effects. The effect may be positive relative to disease in one type of cell, but harmful in other cells. Usually, doses are adjusted to minimize such effects, but humans and our experiences and cellular contexts are so variable that it is difficult to avoid some cost for a given benefit.

This is part of the way life is organized. It's how life evolved. It's how organisms with many different tissues and organs evolved--because an organism is the result of exquisitely complex and highly stereotyped cell-to-cell interactions.

Attempts to target just one gene in one type of cell are a real challenge as a result. That's why even targeting a mutant gene in a given type of cancer has proved to be so challenging, as another story in the NY Times has discussed.