The Human Knockout Project
Share |

The Human Knockout Project

Team Representative:
Danish Saleheen, MBBS, PhD
Assistant Professor of Epidemiology in Biostatistics and Epidemiology, Perelman School of Medicine at the University of Pennsylvania
E-mail: saleheen@pennmedicine.upenn.edu
Phone: 215-573-6323

Media Contact: 
Holly Auer
E-mail: holly.auer@uphs.upenn.edu
Phone: 215-349-5659

Summary of findings in layman's terms: 

The biological function of the vast majority of the >18,000 human genes remains unknown. Naturally occurring inactivating mutations in genes in humans represents a powerful entry point into this uncharacterized biology. These deleterious mutations can also mimic the therapeutic modulation of a target, hence providing reliable insights for the development of medicines. Such deleterious mutations are however very rare in the general population, and finding individuals who have deleterious mutations in both copies of a given gene (“human knockouts”) is exceedingly rare.

Consanguineous marriages are much more likely than unions between unrelated people to result in human knockouts. This study leveraged a highly consanguineous cohort from Pakistan in whom 40% of the participants were born of first cousin unions. The genetic code of all protein-coding genes in ~10,500 adults was sequenced. For 1,317 distinct genes, both copies were found to have deleterious mutations in at least one individual, resulting in ‘human knockouts’ for a remarkable number of new genes for whom human knockouts have never been reported. About 1 in 5 sequenced participants had at least one gene completely inactivated.

This study confirmed the biochemical deficiency in these knockouts as well. One gene in particular, APOC3, which regulates triglyceride metabolism, was inactivated in several dozen individuals, resulting in the world’s first APOC3 human knockouts. These individuals were challenged with a high-fat meal. Compared with family members who were not knockouts, the APOC3 knockouts did not have the usual post-fatty meal rise in plasma triglycerides. Their genetic makeup has provided unique insights about the APOC3 biology, which may help in validating APOC3 inhibition as a therapeutic target for cardiometabolic diseases. Several other human knockouts for other genes were also characterized. This study population should continue to provide an incredible source of new knowledge about how genes influence human health and disease. 

Specific biological innovation of study: 

Despite many decades of progress in human biology, three-quarters of human genes have no known associated phenotype, and a third of them are of completely unknown function. Studying humans knocked out for a gene due to loss of function (LoF) mutations is an innovative way to understand biology; however such individuals are very rare.

To identify human knockouts, Dr. Saleheen has enrolled a unique cohort of 100,000 Pakistanis – 40% of whom are born of consanguineous unions. In this study, whole-exome sequencing was conducted in 10,500 of these participants. The study participants were observed to have a six times higher genome-wide median homozygosity compared to other populations. 

This study further led to the identification of 1843 human knockouts for 1317 genes - many of these genes were observed completely missing for the first time in humans. This study additionally confirmed the biochemical deficiency in participants who were identified to be human knockouts. For instance, APOC3 knockouts had absent plasma apolipoprotein C-III. Furthermore, homozygosity for null mutations at CYP2F1 was found associated with higher plasma interleukin-8 concentrations; and at either A3GALT2 or NRG4, with markedly reduced plasma insulin C-peptide concentrations; hence identifying novel phenotypic associations at these genes.

Additionally, knockouts for 96 genes were identified that were thought to be essential in humans thereby challenging prior methods to catalogue “essentiality” in humans. This study has also helped prioritize therapeutic pathways that are safe and beneficial for inhibition in humans (e.g., APOC3) compared to those that are not (e.g., PLA2G7). Finally, this study demonstrated that provoked hypothesis driven physiological studies in human knockouts should discover new biology; for instance, after challenge with oral fat, APOC3 knockouts displayed marked blunting of the usual post-prandial rise in plasma triglycerides compared to wild-type family members. This study therefore provides a roadmap to the “Human Knockout” project. 

Potential impact on human care and/or how the findings contributed to an improved understanding:

Gene disruption in cells and model organisms followed by phenotypic analysis has been a traditional approach to understand gene function. Studies conducted on these preclinical disease models, however often have limited ability to predict effects in patients. This has been suggested by unsustainably high rate of failures of new compounds in clinical trials of coronary heart disease (CHD) alone, exemplified by failures of varespladib (Anthera), dalcetrapib (Roche), and niacin / laropiprant (Merck) in the past few years.

Naturally occurring loss of function (LoF) mutations that disrupt the gene’s function or expression provide an opportunity to understand the long term consequences of genetic deficiency in humans. These LoF variants represent “experiments of nature” that can mimic the effect of therapeutic modulation of a target. A striking example of the use of LoF alleles to guide therapeutic development for CHD is PCSK9, for which LoF mutations were associated with low levels of LDL cholesterol and protection from CHD; this led to the development of PCSK9 inhibitors.

Identification of human knockouts also has the potential to inform human physiology and the biological function. This study identified 1,317 distinct genes found to be knocked out in humans; many of these genes (e.g., APOC3, PLA2G7, LPA) are being pursued as therapeutic targets, whereas many others are novel genes of unknown function. Observing viability for complete deficiency of these genes in humans demonstrate that achieving a >50% therapeutic inhibition should be theoretically safe. Moreover, deep phenotyping studies conducted in these human knockouts, as illustrated in the case of APOC3 and PLA2G7, can help characterize several important features of a relationship between therapeutic modulation of a target and disease risk, specifically providing insights on: (i) underlying biological mechanisms of a gene; (ii) dose-response associations with disease risk; (iii) pleiotropic effects on other pathways; and (iv) safety implications. 

Journal citation:

Saleheen D, Natarajan P, Armean IM, Zhao W, Rasheed A, Khetarpal SA, Won HH, Karczewski KJ,O'Donnell-Luria AH, Samocha KE, Weisburd B, Gupta N, Zaidi M, Samuel M, Imran A, Abbas S, Majeed F, Ishaq M, Akhtar 2, Trindade K, Mucksavage M, Qamar N1, Zaman KS, Yaqoob Z, Saghir T, Rizvi SNH, Memon A, Hayyat Mallick N, Ishaq M, Rasheed SZ, Memon FU, Mahmood K, Ahmed N, Do R, Krauss RM, MacArthur DG, Gabriel S, Lander ES, Daly MJ, Frossard P, Danesh J, Rader DJ, Kathiresan S.

Human knockouts and phenotypic analysis in a cohort with a high rate of consanguinity. Nature. 2017 Apr 12;544(7649):235- 239.

https://www.ncbi.nlm.nih.gov/pubmed/28406212