Predicting an individual’s response to a diabetes drug

Scientists demonstrated that genetic variation predicts individual responsiveness to the antidiabetic drug rosiglitazone. Rosiglitazone reverses insulin resistance in type 2 diabetes, but its use is limited due to its significant side effects, which can include an increased risk of heart attack and stroke. Additionally, rosiglitazone’s beneficial insulin-sensitizing effect is seen in most but not all people, and scientists have been seeking to understand the reasons for these differential responses. With such knowledge, it might one day be possible to develop diagnostic tests to identify who will respond best to the drug and/or have the fewest side effects. In this study, a group of researchers developed a strategy--using stem-cell generated cells to identify human genetic variation--to study the differential response to rosiglitazone and, in doing so, revealed one way genetic variation alters its effects.

The researchers generated adipocyte (fat) cell lines from tissue samples from five women who had obesity, treated the cell lines with rosiglitazone, and identified genes that were “turned on” in each of the cell lines in response to the treatment. Interestingly, each cell line showed a unique signature of genes that were turned on. For example, 87 genes that were activated in four of the lines remained inactive in the fifth; on the other hand, 399 of the genes that were activated by rosiglitazone in one of the cell lines remained inactive in the other four. Because rosiglitazone is known to activate a protein called PPARγ, which binds to DNA to turn genes on, the scientists examined whether genetic differences between the participants affected PPARγ binding in these cell lines. They found specific genetic differences that not only altered PPARγ binding sites, but also accounted for corresponding differences in the responsiveness to rosiglitazone.

One genetic variation was of particular interest to the scientists because it was near a PPARγ-regulated gene whose protein is known to affect cholesterol levels. Treatment with rosiglitazone commonly leads to higher total cholesterol and low-density lipoprotein cholesterol, which may contribute to heart attacks and strokes among people taking the drug, but it is not well understood how rosiglitazone affects cholesterol levels. The scientists found that having a specific genetic variant (designated “C”) correlated with PPARγ responsiveness to rosiglitazone in that the nearby cholesterol-affecting gene was turned on, while it was not turned on in individuals with a different variant, “A.” The researchers went on to show that these genetic variants also affected cholesterol levels. People with the C variant had higher levels of total and low-density lipoprotein cholesterol in response to rosiglitazone while people who had the A variant near both copies of the cholesterol-affecting gene (cells have two copies of this gene) received the benefits of rosiglitazone treatment (lower blood glucose levels), but much less of the drug’s cholesterol-elevating side effect. This result suggests it may one day be possible for clinicians to identify people who may benefit from rosiglitazone without the adverse effects on their cholesterol levels. The study also presents an approach to identify how human genetic variation determines response to a drug and may be an important tool for understanding drug responses in other diseases.