In response to an original research article published in Nature by Sekar Kathiresan and colleagues (see here), I penned a News & Views piece for Nature (here), a blog for the Timmerman Report (here, here), and a podcast for BBC Inside Science (here). An important theme for drug discovery & development is that human knockouts can rule-in and to rule-out drug targets. For human knock-out data, the key concept is to understand the effect of maximum genetic perturbation on human physiology.
- Rule-in drug targets: As has been described by Matt Nelson and colleagues from GlaxoSmithKline (see 2015 Nature Genetics), and David Cook and colleagues from AstraZeneca (see 2014 Nature Reviews Drug Discovery), therapeutic molecules developed against targets with human genetic data are more likely to lead to regulatory approval than those without. PCSK9 represents the poster child for human genetic knockouts in drug discovery & development (see my plengegen.com blog here). But there are many other examples, too.
- Rule-out drug targets: But human genetics can also rule-out drug targets or mechanisms that are nominated through animal models, human epidemiology or other approaches. A prominent example is related to raising HDL cholesterol, the so-called “good cholesterol”. For decades, human epidemiology has suggested that therapeutically raising HDL through pharmacological perturbations of targets such as cholesterylester transfer protein (CETP), will protect from cardiovascular disease. Human genetics, however, now clearly demonstrates that genetic variants that solely raise HDL do not protect people from cardiovascular disease. Multiple failed clinical trials of CETP-inhibitors support human genetics over human epidemiology. Similarly, the Nature study by Sekar Kathiresan and colleagues provide support that pharmacologically inhibiting the enzyme Lp-PLA2, which is an inflammatory enzyme expressed in atherosclerotic plaques, will not have cardioprotective effects. The Nature study identifies human knockouts of the gene, PLA2G7, that codes for Lp-PLA2 and demonstrates no protection from cardiovascular disease. This observation, which conflicts with the putative beneficial effects predicted from human epidemiological studies (see Lancet 2010), has also been born out in clinical trials, where an Lp-PLA2 inhibitor (darapladib) has not had a beneficial impact on cardiovascular disease (see NEJM 2014, JAMA 2014).
But to accomplish this goal, the industry needs a more complete database of human knockouts. For more on this topic, I encourage plengegen.com readers to review my Timmerman Report blog (here, here), as well as the references cited at the beginning of this short blog.