De novo vs. inherited copy number variations in multiple sclerosis susceptibility

The Human Genome Project has identified, along with single nucleotide polymorphisms (SNPs), a range of other DNA sequence variations, including insertions and deletions of nucleotides and translocations of various segments of a chromosome.¹ These variations have, collectively, been named copy number variants (CNVs).¹ They represent a major source of genomic variation, with nearly 1500 variable regions covering approximately 12% of the human genome.² A CNV could act directly by affecting gene dosage and gene expression through complex mechanisms. It is thought, therefore, that CNVs in gene-dosage sensitive genes may have considerable influence on disease susceptibility.² Among genes overlapped by CNVs, significant enrichments in certain gene ontology categories have been identified, including those related to immune responses and interaction with the environment.³ CNVs have already been associated with several monogenic, syndromic and complex diseases.

Candidate gene association studies in multiple sclerosis (MS) have identified multiple common allelic variants. Given the important role of vitamin D in MS, polymorphisms in the vitamin D receptor (VDR) gene have been extensively studied in relation to MS susceptibility. A meta-analysis by Tizaoui et al.⁴ showed that SNPs in the VDR gene were significantly associated with MS risk. The genetic effects were modest, implicating other genetic variations such as CNVs, which may have larger effects than SNPs. Interestingly, the meta-analysis reported the interaction of VDR polymorphisms with exogenous factors. In fact, as a receptor for vitamin D, VDR expression is, directly and indirectly, dependent on environmental factors. Vitamin D deficiency may increase the incidence of CNVs. Genes involved in vitamin D pathways may be particularly prone to genetic variations. Gene ontology analyses have revealed that CNVs are frequent in genes implicated in immune responses and responses to external biotic stimuli.⁵ Although investigations of SNPs could provide important information in MS genetics, the results remain partial. In the future, more interest should be given to CNVs in genes related to vitamin D pathways.

Recently, CNV associations have been identified for various neurological and developmental diseases.⁶ As a complex disease, MS is related to any variation in the human genome that alters the expression of MS candidate genes. Taking advantage of studies indicating that DNA rearrangements are frequent, and may play an important role in complex disease susceptibility, Sato and collaborators ⁷ explored CNVs in MS. The most identified CNVs were 5 to 50 kb deletions at particular T cell receptor (TCR) gamma and alpha loci. These CNVs were observed in peripheral blood T cell subsets only, suggesting that the CNVs were somatically acquired. Furthermore, they found that certain genomic regions within the TRG and/or TRA loci were deleted. These CNVs were specifically present in CD31, CD41 and CD81 T cell subsets, indicating somatic mutations.⁷ Because the genomic region of the TCR usually undergoes somatic rearrangements in the embryonic thymus, it is important to assess CNVs within the TCR loci. Sample lymphocyte counts or the DNA source may influence rearrangements in the TCR loci.⁸ Sato et al.⁷ showed that the deletion-type CNV at the TRG locus involves the entire J gene segment and part of the V gene segment. Such alteration of the genome seems unlikely to occur with physiological rearrangements. These results are consistent with previous data showing that several inherited and de novo CNVs may cause pediatric MS.⁹ Researchers have reported a single rare CNV harboring the SACS gene. Patients with this CNV shared some features with both MS and the spastic ataxia of Charlevoix-Saguenay (ARSACS). It has been reported that SACS mutations cause autosomal-recessive ARSACS disease. This observation suggests that some autoimmune diseases may share CNVs as common risk factors.¹⁰De novo CNVs in the SACS gene region and other genes have been identified in MS.¹¹ Many of these regions contain paralogous genes, which tend overlap with segmental duplications.¹¹ In addition, McElroy et al.⁹ observed a de novo deletion in the HLA region, which included HCP5P10, HLA-H, HCG2P7, HCG4P6 and HLA-A genes. Within the MHC region, several CNVs implicated in autoimmunity have been described.¹² Retroviral insertions are often located either in the HLA-DR region or near the classical HLA class I genes. Their presence may promote molecular evolution by facilitating non-homologous recombination or gene conversion events.¹² In addition to common CNVs, de novo CNVs have also been associated with neurological disorders. Variants involving three brain-expressed genes involved in glutamate signaling were found only in patients with schizophrenia.¹³ It is speculated that these genes are differentially expressed during early human embryonic development.¹³ Rare de novo variants with high penetrance may underlie schizophrenia in some cases.¹⁴ Similar findings have been identified in parallel investigations into the genetics of autism. Sebat et al.¹⁵ found that the frequency of spontaneous mutations is 10% in sporadic cases of autism, compared with only 1% in unaffected controls. The interpretation of such data is complicated by the difficulty of proving that a particular CNV is de novo and establishing the normal rate of generation of de novo CNVs. Most current methodologies suffer from high false negative and false positive results.

Van Ommen ¹⁶ estimated that over 99% of CNVs are inherited, whereas others are generated de novo during meiosis. Similarly, population genetic analysis of SNPs and CNVs in HapMap samples showed that approximately 80% of the observed copy number differences between pairs of individuals were due to common CNVs and over 99% were inherited rather than new mutations.¹⁷ Although family studies have suggested that de novo CNVs may occur during meiosis, copy number differences have been reported between monozygotic twins. In addition, CNVs occurring between different tissues in the same individual have been reported.¹⁸ Growing evidence suggests that CNVs occur either in meiotic or somatic tissues. The evolutionary drive to generate and maintain de novo CNVs or somatic CNV mosaicism is not yet completely explored. However, some suggest that CNVs may be constantly generated de novo.¹⁹

The development of complex human diseases may increase the incidence of CNVs in the genome. Abnormal immune function, which includes increased cellular proliferation, excessive secretion of cytokines and the presence of free radicals may result in DNA damage. Additional endogenous and environmental agents may cause constant DNA damage. DNA repair pathways may be linked to numerous rearrangements that predispose individuals to severe disorders. Some CNVs initially appear benign, but in combination with other genetic and environmental exposures, may predispose an individual to disease. For example, homozygous deletions in the GSTM1 gene are thought to increase susceptibility to some cancers, possibly due to a reduction in the metabolism of certain carcinogenic compounds in the environment.²⁰ Diet and tobacco smoking are suspected to be risk factors for MS. It has been shown that some components of cigarette smoke can directly mutate cellular DNA.²¹ Additionally, vitamin D deficiency is associated with increased CNVs. It has been shown that CNVs are enriched within genes important in molecular-environmental interactions and may influence the immune defense and disease resistance or susceptibility of humans.¹

The incidence of de novo CNVs is favored by infections, chronic inflammation, cellular stress, free radicals generated by the inflammatory response and diverse physiologic factors. Somatic variations may cause genomic fragility, increasing the likelihood of further genomic rearrangements. Inherited variants can decrease or increase the expression of risk genes, which, in combination with somatic variation, may result in a predisposition for severe diseases. The large proportion of CNVs in the genome indicates that a significant number of SNPs may fall in these regions (Figure 1). SNPs are significantly enriched in known human CNVs, which may act in conjunction with SNPs to produce an adverse phenotype. Thus, future integrated approaches should consider genetic variants at both the copy number and nucleotide levels.³ The interplay between somatic variation and environment, diet or medicine, has opened some new fields of research. Personalized medicine and pharmacogenomics represent the development of preventative, diagnostic and therapeutic approaches specific to the genotypic profile of an individual patient.²² Knowledge about CNVs is important to the understanding of the pathogenesis of autoimmune diseases; however, many findings need to be replicated in independent populations and different ethnic groups. In fact, the association of CNVs with gene expression appears to be more highly population-specific than the identified SNP associations.²³ Genome-wide association studies have characterized the effects of common SNPs on complex human traits but have done a far less comprehensive investigation of the effects of CNVs. One fascinating question is whether there is any relationship between the inherited CNVs in MS-relevant genomic regions and the incidence of the various genomic losses and gains that occur during MS progression.

Figure 1.

Proposed schema for the mutual relationship between DNA damage and the inflammatory process during multiple sclerosis pathogenesis. Various triggers, including but not limited to viral infection, sunlight, nutrients and smoking, may lead to the onset of multiple sclerosis in genetically susceptible individuals. Genetic variations such as SNPs and CNVs interact with each other and with environmental factors, leading to the development of the disease. Early immune events may cause somatic mutations and DNA damage. These mutations may disturb normal cellular machinery and aggravate the condition. CNVs harboring exons and regulatory elements have extensive effects on gene expression. Duplications in regions harboring a gene of interest may amplify the expression of that gene. However, deletions harboring exons and/or regulatory elements may abolish genetic function. De novo CNVs may result in rearrangements in the genome. Therefore, regulation of the immune response may be perturbed, resulting in the persistence of inflammation in the central nervous system. Mutually, outputs of chronic inflammation may induce DNA damage, producing new rearrangements. CNV, copy number variation; SNP, single nucleotide polymorphism; CD, cluster differentiation; E1, exon 1; E2, exon 2; E3, exon 3.

Conflict of interest

The author declares no conflict of interest.