Pleiotropy
Pleiotropy is a condition in which a single gene or genetic variant influences multiple phenotypic traits. A gene that has such multiple effects is referred to as a pleiotropic gene. Mutations in pleiotropic genes can affect several traits simultaneously, often because the gene product is used in various cells and affects different biological targets through shared signaling pathways.
Pleiotropy can result from several distinct but potentially overlapping mechanisms, including gene pleiotropy, developmental pleiotropy, and selectional pleiotropy. Gene pleiotropy occurs when a gene product interacts with multiple proteins or catalyzes different reactions. Developmental pleiotropy refers to mutations that produce several phenotypic effects during development. Selectional pleiotropy occurs when a single phenotype influences evolutionary fitness in multiple ways.
There are also three main types of genetic pleiotropic effects when a variant or gene is associated with more than one trait:
- Biological pleiotropy, where a genetic variant directly affects multiple traits through biological pathways.
- Mediated pleiotropy, where a variant influences one trait, which in turn causes changes in a second trait, and
- Spurious pleiotropy, where statistical or methodological biases make it falsely appear as though a variant is associated with multiple traits.
Pleiotropic gene action can limit the rate of multivariate evolution when natural selection, sexual selection or artificial selection on one trait favors one allele, while selection on other traits favors a different allele. Pleiotropic mutations can sometimes be deleterious, especially when they negatively affect essential traits. Genetic correlations and responses to selection most often exemplify pleiotropy.
Pleiotropy is widespread in the genome, with many genes influencing biological traits and pathways. Understanding pleiotropy is crucial in genome-wide association studies, where variants are often linked to multiple traits or diseases.
History
Pleiotropic traits had been previously recognized in the scientific community but had not been experimented on until Gregor Mendel's 1866 pea plant experiment. Mendel recognized that certain pea plant traits seemed to be inherited together; however, their correlation to a single gene has never been proven.The term "pleiotropie" was first coined by Ludwig Plate in his Festschrift, which was published in 1910. He originally defined pleiotropy as occurring when "several characteristics are dependent upon ... ; these characteristics will then always appear together and may thus appear correlated". This definition is still used today.
After Plate's definition, Hans Gruneberg was the first to study the mechanisms of pleiotropy. In 1938 Gruneberg published an article dividing pleiotropy into two distinct types: "genuine" and "spurious" pleiotropy. "Genuine" pleiotropy is when two distinct primary products arise from one locus. "Spurious" pleiotropy, on the other hand, is either when one primary product is utilized in different ways or when one primary product initiates a cascade of events with different phenotypic consequences. Gruneberg came to these distinctions after experimenting on rats with skeletal mutations. He recognized that "spurious" pleiotropy was present in the mutation, while "genuine" pleiotropy was not, thus partially invalidating his own original theory. Through subsequent research, it has been established that Gruneberg's definition of "spurious" pleiotropy is what we now identify simply as "pleiotropy".
In 1941 American geneticists George Beadle and Edward Tatum further invalidated Gruneberg's definition of "genuine" pleiotropy, advocating instead for the "one gene–one enzyme" hypothesis that was originally introduced by French biologist Lucien Cuénot in 1903. This hypothesis shifted future research regarding pleiotropy towards how a single gene can produce various phenotypes.
In the mid-1950s Richard Goldschmidt and Ernst Hadorn, through separate individual research, reinforced the faultiness of "genuine" pleiotropy. A few years later, Hadorn partitioned pleiotropy into a "mosaic" model and a "relational" model. These terms are no longer in use but have contributed to the current understanding of pleiotropy.
By accepting the one gene–one enzyme hypothesis, scientists instead focused on how uncoupled phenotypic traits can be affected by genetic recombination and mutations, applying it to populations and evolution. This view of pleiotropy, "universal pleiotropy", defined as locus mutations being capable of affecting essentially all traits, was first implied by Ronald Fisher's Geometric Model in 1930. This mathematical model illustrates how evolutionary fitness depends on the independence of phenotypic variation from random changes. It theorizes that an increasing phenotypic independence corresponds to a decrease in the likelihood that a given mutation will result in an increase in fitness.
Expanding on Fisher's work, Sewall Wright provided more evidence in his 1968 book Evolution and the Genetics of Populations: Genetic and Biometric Foundations by using molecular genetics to support the idea of "universal pleiotropy". The concepts of these various studies on evolution have seeded numerous other research projects relating to individual fitness.
In 1957 evolutionary biologist George C. Williams theorized that antagonistic effects will be exhibited during an organism's life cycle if it is closely linked and pleiotropic. Natural selection favors genes that are more beneficial prior to reproduction than after. Knowing this, Williams argued that if only close linkage was present, then beneficial traits will occur both before and after reproduction due to natural selection. This, however, is not observed in nature, and thus antagonistic pleiotropy contributes to the slow deterioration with age.
Mechanism
Pleiotropy describes the genetic effect of a single gene on multiple phenotypic traits. Recent genetic research distinguishes between three forms of pleiotropy:Biological pleiotropy
Biological pleiotropy also known as horizontal pleiotropy is when a causal variant or gene has direct and independent effects on more than one phenotypes. There are two sub-types of biological pleiotropy, single-gene pleiotropy and multigene regulatory pleiotropy.Single-gene pleiotropy
Causal risk variants can affect several traits by acting on a single gene that has many different effects. There are several ways that this kind of gene pleiotropy can happen, and these possibilities can overlap. For example, a gene might have more than one molecular function, be involved in several separate biological pathways or cellular processes, or be active in different organs, tissues, or times and places in the body, each influencing different traits. Also, one gene can produce different versions of a protein, called isoforms, which vary in structure and function and contribute to the gene's wide range of effects.Multigene regulatory pleiotropy
Pleiotropy also occurs when a causal variant changes the expression of many genes. Every one of these genes may play a role in shaping different phenotypic outcomes. Regulatory pleiotropy can also arise from genetic influences on the 3D structure of chromosomes.Mediated pleiotropy
Also known as vertical pleiotropy and happens when a causal variant effect on one trait which itself causes effect on a different trait. An example of mediated pleiotropy is that gene variants that affect low-density lipoprotein also affect coronary artery disease.Spurious pleiotropy
Sometimes, what looks like pleiotropy can be caused by problems in how the study is designed or how risk genes and traits are defined, leading to incorrect conclusions about pleiotropy. Spurious pleiotropy occurs when there is a misclassification either at the genomic level or the phenotypic level. At the genomic level, this might happen when a region of the genome linked to a special trait includes causal variants that are related. When this is the case, variants that influence different phenotypes through separate biological mechanisms may wrongly appear as a single pleiotropic locus.Other
Polygenicity-induced horizontal pleiotropy
There has been introduced a fourth type, polygenicity-induced horizontal pleiotropy, where several genetic loci with causal effects could be linked to multiple phenotypic traits.Network pleiotropy
Another model that has been proposed is network pleiotropy. In this model, a single causal variant influences several traits through one or more intermediate cell types within the same network. It may be especially useful for explaining multi-dimensional psychiatric disorders such as schizophrenia and bipolar disorder.Polygenic risk scores and pleiotropy in complex traits
One of the key challenges is to figure out if a gene actually influences more than one trait. One reason is that it's not always clear how traits should be grouped or named when studying them. Another challenge is that many of the methods used to test for pleiotropy, do it in an indirect way. Usually, these methods start by assuming that a gene doesn't affect any traits, and then look for evidence to prove otherwise. To solve this, researchers have developed better ways to test if a gene affects several traits at the same time, using methods that don't rely on these indirect assumptions.Early genome-wide association studies that revealed links between many genetic loci and multiple traits were often described in terms of cross-phenotype associations. When such associations can be traced back to a shared biological mechanism at the causal locus, they can be more precisely defined as pleiotropic effects.
Genome-wide association studies and machine learning analysis of large-scale genomic data have made it possible to develop SNP-based polygenic predictors for complex human traits. The goal of GWAS was to identify how strongly a specific genetic variant, typically a single-nucleotide polymorphism, is associated with a particular human trait.
One way to quantify pleiotropy is by measuring the proportion of shared genetic variance between two complex traits. Analyses of hundreds of trait pairs have shown that often, the genomic regions involved are largely distinct, with only modest overlap. This suggests that, for the complex traits studied so far, pleiotropy is generally limited. Still, identifying genetic variants through GWAS and linking them to biological pathways offers valuable opportunities to improve diagnosis, develop new therapies, and better prevent diseases.
Polygenic risk scores, built from these findings, holds promise for predicting individual risk for various conditions. However, while PRS has many strengths, their predictive power remains probabalistic. The accuracy and reliability of these scores are currently under scrutiny, emphasizing the need for cautious interpretation when applying them to clinical or public health contexts.