Characterization of Biological Endophenotypes Associated With AIS
To address the phenotypic and genotypic heterogeneity of AIS, we identified a common disease mechanism among all patients and then subdivided these patients based on their biochemistry. These biological endophenotypes, defined by Moreau et al., can then partition biological variation and thus increase the power to detect genetic associations.[8–11] By definition, an endophenotype is a term from genetic epidemiology used to parse behavioral symptoms into more stable phenotypes with a clear genetic connection. This notion was first suggested by Bernard John and Kenneth R Lewis in an article called "Chromosome variability and geographic distribution in insects", where a specific geographic distribution of insects is more likely due to internal endophenotypes instead of obvious external phenotypes. The concept of endophenotypes has been widely applied to complex psychiatric disorders such as schizophrenia. Endophenotypes are believed to have a polygenic background with multifactorial origins. They are associated with illness in the population, are heritable and can manifest in an individual whether disease is active or not, and endophenotypes segregate with illness within families.[13,14] Moreover, endophenotypes are usually present in affected and nonaffected family members at a higher rate than in the general population.
In complex genetic disorders such as AIS, endophenotypes have potential utility both in identifying risk genes and in illuminating the pathophysiology. We have identified three distinct heritable endophenotypes for AIS, based on the previously reported G inhibitory (Gi) signaling dysfunction in AIS patients. Using multiple cell types, a whole-cell assay was developed using cellular dielectric spectroscopy to measure the activity of G-protein-coupled receptors among AIS patients, controls and asymptomatic children with a family history of AIS. These observations led us to classify patients into three distinct groups (referred to as FG1, FG2 and FG3), depending on the maximal Gi response observed after stimulation. This disparity is very specific, as such patterns are not observable among healthy controls. The use of these endophenotypes has allowed for in-house patient stratification that has been correlated to explicit molecular expression profiles and differential prognoses [Yuan Q et al. Genome-wide expression profiling of idiopathic scoliosis endophenotypes (2012), Manuscript in preparation]. The use of endophenotypes has allowed us to highlight some specific clinical features of the disease. Importantly, the occurrence of severe scoliosis cases (Cobb angle ≥45°) is strikingly different among each biological endophenotype. In our French–Canadian pediatric cohort, the percentages of severe cases are 13% for FG1, 60% for FG2 and 27% for FG3 endophenotype. These findings are currently being validated in two independent cohorts. Of note, our blood tests should be viewed first as prognosis tools in affected patients to establish at an early stage the risk of disease progression. However, the same tests applied to asymptomatic children at risk (e.g., offspring of scoliotic parents) could be used to predict the risk of developing scoliosis before disease onset.
Personalized Medicine. 2013;10(1):97-103. © 2013 Future Medicine Ltd.