From Genetics to Genomics of Epilepsy

The introduction of DNA microarrays and DNA sequencing technologies in medical genetics and diagnostics has been a challenge that has significantly transformed medical practice and patient management. Because of the great advancements in molecular genetics and the development of simple laboratory technology to identify the mutations in the causative genes, also the diagnostic approach to epilepsy has significantly changed. However, the clinical use of molecular cytogenetics and high-throughput DNA sequencing technologies, which are able to test an entire genome for genetic variants that are associated with the disease, is preparing a further revolution in the near future. Molecular Karyotype and Next-Generation Sequencing have the potential to identify causative genes or loci also in sporadic or non-familial epilepsy cases and may well represent the transition from a genetic to a genomic approach to epilepsy.


Introduction
In the last decades a large number of gene discoveries have changed our views of idiopathic and symptomatic epilepsy [1]. Indeed, idiopathic epilepsy has the considerable genetic advantage to be found very often in informative autosomal dominant families that have been of great relevance to map and to positional clone the causative gene, opening insight into the biology and molecular pathology of this condition [2,3].
The search of epilepsy genes has allowed the identification of several genes in idiopathic generalized epilepsy (Table 1), the vast majority of which are channelopathies [4,5] or affect the activity of excitatory or inhibitory neurotransmitters in central nervous system [6]. It is possible that the dominant nature of these genes due to the multisubunit composition of the molecules have greatly overestimated the role of their mutations in the disease.
Other important insights came from the discoveries of causative genes of syndromic epilepsy (Table 2) [7] and other disorders where epilepsy is associated with encephalopathies (Table 3) [8], mental retardation with brain malformation (Table 4) [9,10], other neurologic conditions including neuronal migration disorders (Table 5) [11], and inborn errors of metabolism (Tables 6 and 7) [12,13]. Without any doubt, these discoveries have been great advances in the field; however, their impact on the management of epileptic patients was limited because of the failure to collect significant genetic information from each patient to distinguish the large number of genetic defects that can lead to the disease. Therefore, genetic testing was possible only for few or selected family cases.
Technical improvements in human chromosomes recognition and better definition of chromosome regions realized by increasing the number of detectable chromosome bands have provided higher resolution of normal and pathological karyotype. It is today well established an association between epileptic seizures and chromosome abnormalities recognized by high-resolution chromosome banding [14,15]. However, the type and the size of the chromosome defects are not always easy to detect even by the highest-resolution cytogenetic techniques available for light microscopes.
The identification of the specific genetic defect in a patient with epilepsy may clarify the diagnosis (diagnostic testing), suggest the prognosis, assist with treatment and management (e.g., the use of a ketogenic diet in glucose transporter type 1 deficiency syndrome or the avoidance of lamotrigine, phenytoin, and carbamazepine in Dravet syndrome), elucidate the risk of a disease in family members and future children, and save the patient from further diagnostic evaluation and potentially invasive testing.
In asymptomatic subjects with increased risk of seizures because of a family history, genetic test may predict onset of epilepsy (predictive testing) [16,17]. Despite such potential benefits, genetic testing has also potential harms, such as its ethical, legal, and social implications, and the potential for stigma, distress, adverse labeling, and nonconfidentiality that exists in the setting of inadequate safeguards against discrimination [18]. Considering that our understanding of the epidemiology and clinical utility of genetic testing in the epilepsies is incomplete, the assessment of these potential benefits and harms is particularly complex and is closely linked to the clinical scenario.
The International League Against Epilepsy (ILAE) Genetic Commission presented a tool in the approach to specific tests for epilepsy [16]. According to ILAE report, the diagnostic genetic testing is "very useful" in individual affected by early-onset spasms, X-linked infantile spasms, Dravet and related syndromes, Ohtahara syndrome, epilepsy and mental retardation limited to females, early-onset absence epilepsy, autosomal dominant nocturnal frontal lobe epilepsy, and epilepsy with paroxysmal exercise-induced dyskinesia; the predictive testing is "very useful" in unaffected relatives of individuals affected by Dravet syndrome and epilepsy and mental retardation limited to females [16]. Considering   Zinc finger E-box binding homeobox 2 Mowat-Wilson syndrome the potential harms, genetic testing should always be performed with the patient's consent or parental consent in the case of minors. A team approach, including a genetic counselor, a psychologist, and a social worker, is recommended throughout the process of evaluation.
In the last years a number of new molecular genetic technologies became available and they promise to change genetic testing for epilepsy, allowing to extend genetic analysis also to sporadic or nonfamilial cases. Two are the major new technologies that can affect the management of epileptic patients: Oligonucleotide Arrays Comparative Genomic Hybridization (Array-CGH) and Next-Generation Sequencing (NGS).

Molecular Karyotype
During the last 50 years cytogenetics has evolved from simple chromosome counting or banded chromosome morphological identification under light microscope to a molecular approach where chromosomes are analyzed through sophisticated computer system for their ability to hybridize to specific oligonucleotides spanning the entire genome [19]. Array-CGH is nowadays a basic diagnostic tool for clinical diagnosis of several types of developmental delays [20], intellectual disabilities [21,22], and congenital abnormalities [23]. Epilepsy is also enjoying several advantages from the use of this technology that significantly improves diagnostic resolution of classic cytogenetics [24,25].
Chromosomes did not become individually identifiable before the discovery that several procedures could create reproducible, permanent, and specific banding patterns [26,27]. This was fundamental for gene mapping and positional cloning of disease genes and also revealed a large number of rare and subtle pathological conditions that disturbed the normal band patterning of chromosomes. The improvements of high-resolution banding techniques allowed the identification of several subtle chromosomal abnormalities associated with epilepsy [14,15]. The possibility to study   these chromosomes regions with specific hybridization DNA probes through fluorescence in situ hybridization (FISH) greatly improved sensitivity to detect small chromosomal aberrations in specific regions [28]. Today molecular karyotyping is rapidly replacing conventional cytogenetics and FISH. This name refers to the analysis of all chromosomes using hybridization to standard DNA sequences arranged on a "chip" rather than microscope observation. The technological development of this approach allows now clinicians to evaluate the entire genome for copy number variants (CNVs, duplications, deletions) in a single test. The high resolution of this approach is however limited by the difficulty to identify balanced chromosome translocations or inversions, even if this powerful technique recognizes in many of them microdeletions or cryptic anomalies at the chromosomal breakpoints. Detection of deletions or duplications is based on the comparison of two genomes ( Figure 1). Labelled patient DNA is cohybridized with control DNA to an array spotted with oligonucleotide DNA probes spanning the entire genome at critical intervals. The distance between these oligonucleotide sequences in the genome marks the resolution of the technique and can be as low as 1000 bp. The intensity of the signal from patient and control are then read and normalized by an electronic scanning device coupled with a software that generates a graphic plot of intensities for each probe.
A search of Online Mendelian Inheritance in Man (OMIM) clinical synopsis with the term "seizure" reveals that there are at least 754 mendelian disorders in which epilepsy is or can be part of the clinical condition, but not the main feature. Many of these disorders can be associated with DNA sequence mutations or subtle chromosomal anomalies that can be conveniently detected by array-CGH. Some will be private or sporadic cases and others will be familial. With the widespread use of CGH in both circumstances, many more genetic events will be reported in patients and the genetic aetiology will be recognized making possible over the time to saturate the genome with all possible loci and events that have an epileptogenic role.  knowledge of human genome has considerably changed [29] and the technologies to sequence DNA are today of common use in diagnostic practice and much cheaper. Chain termination or Sanger's method [30] was largely used for the Human Genome Project and has dominated the past decades. The logic of this technology was to create by synthesis a population of DNA fragments of different size each one terminated at all possible positions by one of the four labelled dideoxynucleotides (ddNTPs) terminators. Separation of these fragments by polyacrylamide gel or capillary electrophoresis allowed the reading of the sequence through the first developed sequencing machines that could distinguish the fluorescence emitted by the blocking ddNTP [31]. Therefore, these are considered the "first generation" of DNA sequencing technologies. The need to reduce the cost of large sequencing projects has stimulated the development of a variety of cheaper sequencing technologies that are generally called "Next-Generation Sequencing" (NGS) [32,33]. The final goal of this new field is to reduce the cost of human genome sequencing till or lower than $1,000 per genome to make it available  for common medical practice and diagnostic use [34]. The development of further third-generation sequencing technologies should make possible to sequence single DNA molecules in real time with a cost that it is projected to be very close to the goal [35]. The development of NGS platforms was a major progress in the technology because, differently from Sanger method, rather than producing about one thousand nucleotides for run, they are able to produce orders of magnitude more sequence data using massive parallel process, resulting in substantial increase of data at a lower cost per nucleotide [36,37].

Next-Generation Sequencing
Several commercial platforms are today available, including Roche/454 [38], Illumina/Solexa and Life Technologies/SOLiD (Table 8(a)). In very general terms these platforms follow similar process that includes: (a) template preparation by breaking large DNA macromolecule to generate short fragment libraries with platform-specific synthetic DNA adapters at the fragment ends, (b) massive and parallel clonal amplification of individual DNA fragment molecules on glass slide or microbeads by PCR [39] to generate a sufficient copy number of the labelled fragment to be detected by the machine optical system, and (c) sequencing by several cycles of extensions that are repeated and detected  automatically to create short reads [40]. The data of these reads are then collected by the device, and the alignment of the short reads with specific software allows to rebuild the initial template sequence. Helicos and Pacific Biosystem platforms (Table 8(b)) are substantially different because they use a more advanced laser-based detection system that does not require massive parallel amplification with the considerable advantages to simplify preparation process, to eliminate PCR-induced bias and errors, and to make easy data collection. Ion Torrent developed an entirely new approach to sequencing based on hydrogen ion release when a nucleotide is incorporated into a DNA strand by polymerase (Table 8) [41]. An ion sensor can detect hydrogen ions and convert this ion chemical signal to digital sequence information eliminating the need of optical reading at each dNTP incorporation.
Other third-generation platforms under development make use of nanophotonic visualization chamber, ion semiconductor, electron microscopy, a variety of nanotechnologies like nanopores (Oxford Nanopore Technologies), nanochannels (BioNanomatrix/nanoAnalyzer), nanoparticles (GE Global Research), nanoballs (Complete Genomics), nanowells (CrackerBio), nanoknifes (Renveo), and specially engineered sensor DNA polymerase (VisiGen Biotechnologies) [42]. They promise even larger and faster data production although they are still under development and a few years away from commercial use. In principle also DNA microarrays could allow sequencing by hybridization using ultrafast nonenzymatic methods (Genizon BioSciences) and somebody even suggests that mass spectrometry might be used to determine mass differences between DNA fragments produced by chain termination [43].
The beginning of several individual genome projects has gradually decreased the cost of sequencing an individual genome, and it is likely that the $1,000 cost per person will be reached in few years. In medicine, the "personal genome" age made possible by NGS will be an important milestone for the entire genomic field and will mark a transition from single gene testing to whole genome evaluation [44].
It is impossible to predict today which NGS will eventually dominate genomic research, but it is sure that cost reductions, sequencing speed, and better accuracy will make NGS an essential molecular tool in several areas of biology and medicine.
Although the cost of whole genome sequencing has dropped significantly, it remains a major obstacle since it can reach $100,000 for a single individual. However, targeting sequencing of specific regions of interest can decrease the overall cost and improve efficiency of NGS making this technology ready for diagnostic use [45].
Also the field of epilepsy is potentially affected by NGS. Indeed too many genes and genetic conditions can be associated with epilepsy to make impossible for the clinicians a general use of specific monogene test for the vast majority of nonsyndromic or idiopathic epileptic patients. NGS is changing this situation by targeting several genome regions where known epilepsy genes are located and using enrichment techniques to significantly reduce the cost and improve efficiency. Targeted sequencing usually tests all protein-coding exons (functional exosome) which only requires roughly 5% as much sequencing than whole genome. This strategy will reduces the cost to about $3,000 or even less per single individual. Targeted selection technologies have been marketed and successfully used in different NGS projects and are becoming the tool of choice in several conditions, including epilepsy [46].

Targeting Sequencing and Epilepsy Gene Panels
A diagnostic panel is the contemporaneous targeted sequencing of a number of known genes that have already been identified as cause of a particular disease. A diagnostic panel is very different from whole genome or exome sequencing. Only genes clearly associated with a disease are examined. The genes included in the panel can be decided by the prescribing physician or by ad hoc committees of experts that can reach a consensus on the number and type of genes to test making commercially available diagnostic panel kits for specific diseases. This strategy should make easier to detect genetic variants that after validation by Sanger sequencing can be interpreted as the cause of the disease. Of course diagnostic panels and targeted sequencing make sense only if the condition is caused by several or very large number of genes. Many genetic disorders fall in this condition and are excellent candidate for the development of diagnostic panels. Epilepsy is an excellent example of such situation since it is a relative frequent disease affecting 1% of the population in a variety of forms, at different ages, with different progression. A genetic cause of epilepsy can be reasonably supposed in sporadic cases if trauma, tumor, or infection can be ruled out. In such circumstance all genetic information about epilepsy genes identified over the years in familial cases can   Peroxisomal biogenesis factor 26 ZWS-NADL-IRD be used to identify the causative gene through an epilepsy diagnostic panel. Indeed in the case of epilepsy the identified genes are so many that they can be classified in subpanels of genes that underline a common clinical entity (Table 9). Clinical considerations may suggest the clinician to include in a diagnostic panel other genes or genes from different subpanels. At the present the first diagnostic panels for epilepsy that can analyze up to 400 different genes (CeGaT GmbH) are commercially available [47].

Future Perspectives and Conclusions
The genetics of epilepsy has evolved from ion channel and neurotransmitter receptor subunits to newly discovered genes highlighting the importance of different pathways in the epileptogenesis. Furthermore, it has been demonstrated that copy number variations collectively explain a larger portion of idiopathic epilepsy than any single gene. These studies have identified structural genomic variations associated with idiopathic epilepsy, representing a change from the conventional knowledge that chromosome microarray analysis is useful only for patients with intellectual disability or dysmorphism [48]. Genetic testing techniques are rapidly evolving and whole exome or whole genome sequencing, performing at increasingly cheaper costs, will allow rapid discovery of other pathogenic mutations, variants in noncoding DNA, and copy number variations encompassing several genes. This rapidly accumulating genetic information will expand our understanding of epilepsy, and will allow more rational and effective treatment. However, along with the ability to identify genetic variants potentially associated with epilepsy it is imperative to validate genetic associations and analyze their clinical significance. Is it worth? The main objective of a diagnostic panel for epilepsy is to discover the molecular defect in all possible cases to create a specific and personalized treatment of the disease than can be pharmacologically different for different types of molecular defects. Personalized therapy will be possible only within a genomic medicine. But genomic medicine at the same time will raise other questions: what to do if more than one genetic variant is identified in the same epileptic patient? Can we understand how genetic interactions will modulate the disease severity and prognosis? Can interaction of specific genetic variants and environmental factors modulate the clinical spectrum of the disorder? What to do if the diagnostic panel is inconclusive? Are the costs affordable? These are only few questions that the genomics of epilepsy will raise. The answers will require time, a lot of sequencing, and probably the development of new and cheaper sequencing technologies.