Disentangling the molecular mechanisms of multiple sclerosis: The contribution of twin studies

Highlights

• Identical twin pairs discordant for MS are ideal for the co-twin control design.

• This approach allows studies, controlling for genetic and non-genetic influences.

• The co-twin control design provides clues on non-genetic factors of MS etiology.

• MS-discordant monozygotic twins inform on early, possibly reversible, endophenotypes.

Abstract

Twin studies of disease concordance are useful to weight the relative contribution of genetic and environmental factors to the cause of common complex disorders. In multiple sclerosis (MS) different twinning rates from geographic areas at different prevalence suggested that heritable and non-heritable factors contribute in different proportions and ways to MS risk in diverse populations. This concept prompted genome-wide association studies, and the implementation of the co-twin control design, that allows stringent experimental approaches in MS-discordant identical pairs, controlling for genetic influences and many other known and unknown factors. The co-twin control design provided important clues on MS molecular model. These studies will be reviewed, focusing on those showing significant differences between affected and healthy co-twins. In some cases, differences that emerged in non-twin patients compared to matched controls were not confirmed in identical MS-discordant pairs, suggesting an ‘MS subclinical trait’. Early patterns of magnetic resonance imaging and predictive biomarkers that characterize ‘healthy’ co-twins may be useful for the identification of a prodromal reversible phase of the disease.

Introduction

Multiple sclerosis (MS) is a demyelinating disease of the central nervous system (CNS) beginning with a relapsing-remitting phase in most cases, for which many disease-modifying treatments are now available. Nonetheless, an etiologic therapeutic approach is still out of reach since molecular mechanisms underlying disease, as well as accurate etiopathogenic models remain largely unknown. Among others, the increased of MS incidence and female to male ratio in recent decades, as well as many works in progress on disease etiology suggest complex interactions between heritable, non-heritable and stochastic events for illness development (Dobson and Giovannoni, 2019; Sadovnick, 2019). As for other common complex disorders (where there is an interplay between heritable and non-heritable etiological factors), also in MS the availability of twin registries is of invaluable value. Twin studies of disease concordance have historically been used to evaluate the relative contribution of genetic and environmental factors to the cause of common complex disorders (Burbridge, 2001): greater concordance in monozygotic(MZ) than in dizygotic (DZ) twins suggests a role of genetic factors, while similar concordance rates between MZ and DZ twins, or discordance in MZ twins, suggest causative non-heritable factors. This approach may also consider the recurrence risk to sibs of twin probands, to determine whether DZ twins are at more risk than their non-twin sibs (Willer et al., 2003).

In Nordic countries, where MS prevalence is high and population-based registries are available, twin studies in MS consistently demonstrated a higher disease concordance in MZ twins, highlighting the importance of genetic factors (Mumford et al., 1994; Willer et al., 2003; Hansen et al., 2005a,b; Islam et al., 2006; Kuusisto et al., 2008; Westerlind et al., 2014). In Italy, a country of medium MS prevalence, the potentially largest twin registry in the world was made available for medical research thanks to the existence of the fiscal code: all Italian citizens receive, for fiscal purposes, an alphanumeric code that identifies the individual’s surname, first name, and date and place of birth. Individuals with the same family name and place and date of birth are potential twins, and this made it possible to create a nation wide list of more than 1 million potential twins (Salvetti et al., 1997). This list evolved and led to the Italian Twin Registry that was regularly updated using municipalities’ records (Stazi et al., 2002). The availability of the registry of potential twins allowed a population-based study that showed lower twin concordance rate in MZ twins, compared to Nordic countries (Ristori et al., 2006). This work confirmed, at least in part, the results of a previous French study (French Research Group on Multiple Sclerosis, 1992), pointing to a relation between relative weight of genetic influences and disease prevalence. Overall, different twinning rates in MS patients from geographic areas at different prevalence suggest that in diverse populations both heritable and non-heritable factors contribute in different proportions and ways to MS risk, extending the concept of causative heterogeneity from individual to population level.

Twin studies also offer the possibility of deriving quantitative estimates of the relative contribution of heritable and non-heritable factors to the cause of complex diseases (Boomsma et al., 2012).Our group performed a meta-analysis of eight twin studies in MS (Table 1, reorganized from Fagnani et al., 2015), selected through various steps. After a MEDLINE search (‘multiple sclerosis [Title/Abstract] AND twins [Title/Abstract]’) that identified 166 studies, we manually refined the search to eliminate the works that were not focused on the estimation of MS concordance rate in MZ and DZ pairs. Of the resulting 18 candidates (Bobowick et al., 1978; Cendrowski, 1968; Currier and Eldridge, 1982; Ebers et al., 1986; French Research Group on Multiple Sclerosis, 1992; Hansen et al., 2005a,b; Heltberg and Holm, 1982; Islam et al., 2006; Kinnunen et al., 1987; Kuusisto et al., 2008; Mackay and Myrianthopoulos, 1966; Mumford et al., 1994; Ristori et al., 2006; Sadovnick et al., 1993; Westerlind et al., 2014; Willer et al., 2003; Williams et al., 1980), we selected eight studies that met the criteria of systematic case ascertainment and adequate mean age of the population studied (age at onset may affect the estimate of the concordance since it may not be necessarily the same for the co-twins). In case of multiple reports from the same population, we considered the updated study. We therefore obtained quantitative summary estimates of heritability (i.e. the proportion of total variance in disease liability that is explained by genetic variance) and of shared and unique environmental components of MS susceptibility, disentangling their relative contributions with a reasonable degree of confidence. Using a method that took into account the study-specific design and the population-specific etiological background, we found the following meta-analytic estimates: 0.50 (95 %CI:0.39–0.61)for heritability,0.21(95 %CI: 0.11–0.30) for shared environmental component and 0.29 (95 % CI: 0.26–0.33) for unique environmental component. These results support the efforts in progress to identify unknown genetic mechanisms that may fill the gap of ‘missing heritability’; on the other hand, a ‘missing environmentality’ emerges in the context of non-heritable components, which contributes to MS development as both shared and individual-specific exposures.

The ‘missing heritability’ is the focus of intense research in MS and other complex traits through genome-wide association studies (GWAS) on large populations aimed at identifying genetic susceptibility variants and at deciphering their biological plausibility in the etiopathogenic cascade. The ‘missing environmentality’ may be studied by diverse approaches; exploiting MS-discordant identical twin pairs, through the ‘co-twin control design’, may be an invaluable option to explore environmental components (Goldberg and Fischer, 2005).