1.The observable physical or biochemical characteristics of an organism, as determined by both genetic makeup and environmental influences.
2. An individual or group of organisms exhibiting a particular phenotype.
Phenotypes also refer to disease states such as risk of disease, response to therapy, a quantitative biomarker of medical relevance, or a physical trait such as height (as in the figure above).
For drug discovery, I have put forth the premise that human genetics is a useful tool to uncover novel drug targets that are likely to have unambiguous promotable advantage (see here). The starting point in a genetic study is to pick the right phenotype, one that is an appropriate surrogate for drug efficacy.
And phenotype matters!
Two illustrative examples are the autoimmune diseases type 1 diabetes and rheumatoid arthritis. In type 1 diabetes the immune system destroys the pancreas, thereby preventing insulin secretion and the control of blood glucose levels.
Human genetics has identified many alleles associated with the risk of type 1 diabetes, nearly all of which act on the immune system (see here). Drugs that are developed based on the genetics of type 1 diabetes might be expected to alter the immune system and prevent development of disease. However, once diagnosed, the primary treatment for type 1 diabetes is the administration of insulin to maintain glucose homeostasis. Thus, the genetic pathways that lead to the development of type 1 diabetes (immune dysregulation) are different from the biological pathways that are relevant for treating type 1 diabetes once diagnosed (glucose homeostasis).
By contrast, in patients with rheumatoid arthritis the immunological pathways that lead to the disease also seem to be related to the immunological pathways that contribute to symptoms in patients with established disease. As direct proof of concept, several genes that are implicated in the pathogenesis of rheumatoid arthritis are the targets of drugs that are effective therapies for this disease; for example, cytotoxic T lymphocyte antigen 4 (CTLA4) is targeted by abatacept (Orencia; Bristol-Myers Squibb) and interleukin-6 receptor (IL6R) is targeted by tocilizumab (Actemra; Roche). That is, the genetic pathways that lead to RA are similar to the biological pathways altered in patients with disease (in both, immune dysregulation). Our 2014 Nature paper provides direct, quantifiable evidence for connections between genetics of RA and drugs to treat RA (see here).
Thus, there must be a connection between the disease state used in the genetic study and the disease state for the drug indication.
Now think through several other examples, and ask yourself: Are the genetic pathways that lead to disease the same as the pathways that should be modulated to treat the disease? Here are some phenotypes to consider: asthma, heart failure, atherosclerosis, and type 2 diabetes. I would argue that some are more relevant genetic phenotypes for drug discovery than others.
More relevant phenotypes. If prior genetic data link a clinical phenotype and an approved therapy, then this provides strong evidence that the clinical phenotype is relevant for drug discovery. Two additional examples are psoriasis and osteoporosis / bone mineral density.
GWAS and related approaches have identified >40 psoriasis risk loci. Genes from these loci are involved in the differentiation of a subset of helper T cells, Th17, that preferentially produce interleukin-17 and play a major role in orchestrating inflammation. Such genes include IL23R, IL12B, IL23A, TRAF3IP2, and KLF4. Several studies have shown that anti-IL17-receptor therapy is remarkably effective in treating psoriasis (e.g., brodalumab, ixekizumab, secukinumab), as are drugs that block cytokines upstream of IL-17 (e.g., ustekinumab, which blocks two cytokines that share a co-receptor, IL-12 and IL-23).
More than 50 loci have been identified that modulate bone mineral density. Genes from several of these loci lie within the RANK-RANKL-OPG signaling pathway (encoded by TNFRSF11A, TNFSF11 and TNFRSF11B, respectively) and Wnt signaling pathway (e.g., LRP5, WNT3, WNT4, WNT5B, WNT16). While a formal comparison has not been performed with approved drugs that alter BMD, it is obvious that many GWAS hits overlap with approved BMD drugs (e.g., denosumab).
Less relevant phenotypes. Often, there is not a clear link between human genetics, clinical phenotypes and approved therapies, or there is no obvious biological connection between the pathways that lead to the phenotypes used in a genetic study and the pathways relevant to treatment. Examples of less relevant phenotypes include developmental diseases (e.g., autism, congenital heart defects) or degenerative diseases (e.g., late-stage Alzheimer’s disease, end-stage kidney disease). In these examples, genetics may uncover targets that, if pharmacologically modulated prior to disease onset, may prevent development of disease. However, it seems less likely that genetics will uncover targets that can be modulated once the disease is established.
Biological knowledge evolves. Importantly, phenotypes that are considered less relevant today may become highly relevant in the future, as additional knowledge about biological pathways is learned. One of the best examples is with Marfan’s syndrome, a genetic disease with defects in connective tissue. Prior to the discovery of mutations in the fibrillin-1 gene, the underlying defects observed in Marfan’s syndrome were considered an end-stage process that were not reversible. However, with a better scientific understanding of the dynamic nature between fibrillin-1 mutations and the TGF-beta pathway, new therapeutic hypotheses emerged about potential treatments. There are promising results of therapeutic modulation with angiotensin receptor blocks and risk of aortic aneurysms in patients with Marfan’s syndrome.
In summary, if one is to use human genetics to identify drug targets, the phenotype used in the genetic study must be relevant to the biological pathways important for treatment indication.
In the next two blogs, I will describe how to execute on a TIDVAL strategy anchored in human genetics.
Part 3: One gene, one target, one drug – human genetics can find genes with an allelic series to estimate dose-response curves at the time of target ID and validation.
Part 4: Many genes to pathways to targets – a systems biology approach, anchored in human genetic, is a powerful complement to the one gene “allelic series” TIDVAL approach.