Natural human knockouts and the era of genotype to phenotype

Cohorts sequenced to date in search of human knockouts have given different estimates
of the number of knockout events per person depending on whether they were enriched
for complete (biallelic) knockout events. The study by Stefansson and colleagues 9] is particularly noteworthy not only because of the size of the sequenced cohort (2,636)
or its enrichment for biallelic knockout events (due to the founder effect of the
Icelandic population) but also because the LOF alleles were imputed by high-density
genotyping such that the cohort size screened for these events was effectively increased
to 100,000 9]. This is by far the largest study to date on human knockouts and provides a valuable
resource in the discovery of the role of complete human knockouts in general populations.
Although the study was performed on a genetic isolate, some of its results seem to
be generalizable; for example, the predilection of knockout events to affect olfactory
genes was similarly observed by MacArthur et al.8] and Alsalem et al.4]. As the study cohort was mixed, their impressive list of about 5,000 genes with biallelic
LOF 9] should not be interpreted as a list of genes that do not cause severe Mendelian diseases
in humans. For instance, they proposed that their knockout subjects for LRIG3 and OTOP1 are candidates for auditory evaluation to assess for hearing loss given the established
role for these genes in ear development 9]. On the other hand, the cohorts studied by MacArthur et al.8] and Alsalem et al.4] were selected such that severe Mendelian diseases (or their casual variants) were
excluded. The resulting partially overlapping lists of 221 and 169 genes, respectively
(the former excluding splicing variants), were therefore significantly depleted for
severe Mendelian phenotypes. Consequently, these knockout events represent a rich
resource to generate hypotheses about the apparent tolerance of the human genome to
these knockout events.

One of the most intriguing, testable hypotheses is that this tolerance is context
dependent. The context could be genetic or environmental. For example, there are numerous
single knockouts in model organisms that appear completely normal but a phenotype
is observed in double knockouts 5]. A second knockout event is not necessary, however, because milder genetic modifiers
(referred to as genetic background in mouse models) can also be important in determining
the phenotypic expression of the human knockout events as they do in knockout animal
models. Environmental factors may also be important determinants; for example, FUT2 knockout may lead to clinically consequential B12 deficiency only in nutrition-deficiency
states 4]. We have also observed knockout events in several immune-related genes and it is
possible that these will only manifest phenotypically under certain microbial exposures
4]. Knockout events may have a role in human phenotypic diversity in less subtle ways
5]. For example, knockout of olfactory and keratin genes may influence flavor preferences
and hair texture, respectively. These are just two of the many phenotypes that are
not typically assessed upon enrolling ‘healthy’ individuals in these sequencing studies,
so it is inevitable that reverse phenotyping will become commonplace as we move from
the discovery of knockout events in genes to evaluating their potential phenotypic
consequences, taking a lead from their known or predicted biological roles 5].