Genomic Markers for Essential Tremor

There are many reports suggesting an important role of genetic factors in the etiopathogenesis of essential tremor (ET), encouraging continuing the research for possible genetic markers. Linkage studies in families with ET have identified 4 genes/loci for familial ET, although the responsible gene(s) have not been identified. Genome-wide association studies (GWAS) described several variants in LINGO1, SLC1A2, STK32B, PPARGC1A, and CTNNA3, related with ET, but none of them have been confirmed in replication studies. In addition, the case-control association studies performed for candidate variants have not convincingly linked any gene with the risk for ET. Exome studies described the association of several genes with familial ET (FUS, HTRA2, TENM4, SORT1, SCN11A, NOTCH2NLC, NOS3, KCNS2, HAPLN4, USP46, CACNA1G, SLIT3, CCDC183, MMP10, and GPR151), but they were found only in singular families and, again, not found in other families or other populations, suggesting that some can be private polymorphisms. The search for responsible genes for ET is still ongoing.

The etiopathogenesis of ET is not clearly established. Despite many reports in the literature suggesting an important role of genetic factors [23][24][25], these do not explain all cases, and a possible role of environmental factors has been suggested, especially to explain sporadic forms of ET [26][27][28]. The role of genetic factors in the etiology of ET is supported by the high frequency of positive family history of tremor in patients with ET, the description of genetic anticipation, that is, the onset of tremor at an earlier age in the next generation, and the higher concordance rates of ET for monozygotic than for dizygotic twins found in twin studies [23]. The most usual inheritance pattern of ET is an autosomal dominant mode (likely one or more autosomal dominant genes with low penetrance),

Linkage Studies
Linkage analysis is a genetic tool that searches for physical segments of the genome that co-segregate with certain phenotypes or traits through families. These type of studies identified 4 susceptibility loci for familial ET, which have been located at chromosomes 3q13 [29], 2p25-p22 [30], 6p23 [31], and 5q35 [32]. The results of linkage studies in families with ET are summarized in Table 1.
The first locus linked to ET, named FET1 (familial ET1) or ETM1, was reported in Icelandic families through a genome-wide scan study [29]. Linkage of ET to the ETM1 gene was found in only 4 of 30 ET families of Slavonic and Tajik origin [34], but was not confirmed in studies of ET families of other geographical origins [33,35,36]. However, further studies described an association between the rs6280 SNP in the dopamine receptor D3 (DRD3) gene (MIM/gene ID 126451/18149), which is responsible of the Gly9Ser amino acid change, and the risk for ET [43]. This gene is located in the ETM1 locus and is currently designated as ETM1 in the Gene Database. This variant was found in 23 of 30 French families with ET, and the presence of the DRD3Gly/Gly genotype was associated with an earlier ET onset [43]. Further association between the DRD3Gly allele with the risk for ET was found in two case-control association studies in North American [44] and Spanish populations [45]. However, the results of replication studies in other populations [46][47][48][49][50] and the lack of segregation in other families with ET [51,52] did not confirm such an association. The analysis of pooled results of case-control association studies showed a non-significant trend towards a slight overrepresentation of the DRD3Gly allele in ET patients compared with controls [23,53,54]. Gene mapped at chromosome 3q13 with a genome-wide significance assuming an autosomal-dominant model (parametrically LOD score = 3. 71 and non-parametrically LOD score = 4.70). The highest single-family LOD score was 1.29 [29] United States of America ETM1 (FET1) 3q13 190300/2111 Linkage analysis with microsatellite markers for ETM1, ETM2, and chromosome 4p in 38 members of a six-generation family with ET Lack of association with the analyzed loci, including ETM1 and with chromosome 4p [33] Russia ETM1 (FET1) 3q13 190300/2111 Linkage analysis for ETM1 and ETM2 loci, and for locus DYT1 on chromosome 9q32-34 in a group of Slavonic (11 patients) and Tajik (19 patients) families with ET.

Lack of association with ETM1
[35] Linkage analysis in a large American-Czech family with "pure" autosomal-dominant ET (138 members, with 18 affected with ET; genetic anticipation over generations) This gene was mapped close to D2S272 at chromosome 2p25-p22 (maximum LOD score = 5.92). Affected relatives showed a CAG repeat expansion not clearly located in the ETM2 locus. [30] United States of America ETM2 2p25-p22 602134/2112 Linkage analysis in the previous large American-Czech family with "pure" autosomal-dominant ET and in 3 additional, unrelated American families using fine mapping results in an "only-affected" model Positive combined pairwise LOD scores (Z) at the ETM2 locus with a Z(max) = 5.94 at a recombination fraction (theta) = 0.00 for locus D2S220. Haplotype reconstruction places the ETM2 gene in a 9.10 cM interval (D2S224-D2S405) Multipoint linkage analysis suggested that the ETM2 gene was a the 2.18 cM interval (D2S2150 and D2S220; Z(max) = 8.12). [37] United One family showed genome-wide significant linkage to ET in chromosomes 5 and 18, but shared segment analysis reduced the 5q35 region by 1 Mb, and excluded the 18p11 candidate region. No causative variants in the 5q35 region were identified after exome sequencing. [32] Regarding the ETM2 gene, the initial linkage found in a large American-Czech family with "pure" autosomal dominant ET [30] was confirmed by the same research group in other independent American families [37] and a case-control association study [38]. Additionally, a decrease in short tandem repeats in the ETM2 gene, designated as ETM1234 microsatellite, was found to be associated with the risk for ET in the Korean population [41]. In contrast, studies in other populations failed to find any linkage between ET and the ETM2 locus in different populations [33][34][35][36]42]. Among the possible candidate genes included in locus ETM2, the rs11680700 variant within the HS1BP3 gene (HCLS1 binding protein 3, MIM 609359, gene ID 64342) was described in two families with ET [55], and in 12 of 73 (16.4%) patients with familial ET unrelated among them, while this variant was absent in 304 healthy controls [55,56]. This variant was infrequent in 2 families with 27 members affected by ET (it was only present in 3 subjects of the same family) [57] and was not associated with the risk for ET in a case-control association study [58]. Linkage of ET to the ETM3 locus found in a study of large American families [31] was not confirmed in Italian families [35,36]. Finally, linkage of ET was reported in 1 of 5 families at a locus on chromosome region 5q35, but to our knowledge, there are no replication studies [32].

Genome-Wide Association Studies (GWAS)
GWAS consist in analyzing many common spaced genetic variants in cases and controls trying to look for genetic markers that are associated with a disease. Identification of genetic markers can be useful to understand the contribution of genes to the risk of disease and to develop strategies for its prevention and therapy.
The first two GWAS reported, respectively, an association of the risk for ET with two SNPs in the Leucine rich repeat and Ig domain containing Nogo receptor interacting protein-1 gene (LINGO1) [59,60], and with an intronic variant in the solute carrier family 1-glial affinity glutamate transporter-, member 2 (SLC1A2) gene with ET [60]. However, further replication studies on these variants showed controversial results, that will be discussed in the next section.

Studies on LINGO Gene Family
The first GWAS described a strong statistical association between the intronic rs9652490 and rs11856808 SNPs in the LINGO1 gene (chromosome 15q24.3, MIM 609791, gene ID 84894) and the risk for ET in the Icelandic population, but, after adjusting for the rs9652490 genetic effect, the association with rs11856808 disappeared, and the only confirmed in diverse studies was that of rs9652490 with ET [59]. However, LINGO1 should be an interesting candidate gene to modify risk of ET by many reasons, which are outside of the scope of the present review [23,61], and even could be potentially related to the therapy of this disease [61].
A meta-analysis found no association whatsoever for the rs9652490G allele, and identified a weak association with the risk of developing ET and the presence of the rs11856808T allele [72]. Nevertheless, both rs9652490G and rs11856808T alleles had a weak association with ET limited to patients with a positive family history of ET [72], whereas two other meta-analyses described association [54] and lack of association [73] between rs9652490 SNP and ET risk. Interestingly, some authors reported a significant association of the LINGO1 rs9652490AA genotype and the rs9652490A allele with ET risk, under a recessive model, in a North American series [63,65], while the association described by other authors was with the minor allele rs9652490G [59,60,62,66]. This flipflop phenomenon [74] could have affected the results of the meta-analyses. Moreover, a further two-stage GWAS involving 2807 ET patients and 6441 controls of European descent did not find an association with rs9652490 [75].
Other variants in the LINGO1 gene have been reported to be associated with the risk for ET or to an earlier onset of ET in single studies. An association between the rs8030859T allele and the risk for ET was reported in the German population [60], and a weak association between the rs7177008, 13313467, and rs8028808 and early-onset ET was reported in North Americans [66]. A case-control association study in Chinese Han did not find any association of rs2271398, rs2271397, rs3743481, and a novel G → C transition ss491228439 SNPs variants and the risk for ET, although rs2271397 and ss491228439 variants could contribute to the risk for ET among females [76].
A study in North Americans described an association of 5 tagging SNPs within, or close to, the LINGO1 and LINGO2 (MIM 609793, Gene ID 158038) genes (rs4886887, rs3144, rs8028808, rs12905478, and rs1412229) with the risk of developing ET [63,65], and another study involving Asian populations found an association of the LINGO2 rs7033345CC genotype and the LINGO2 rs10812774C allele with the risk for ET under a recessive model [77]. Finally, another study in Chinese Han found a lack of association between the rs61746299 and rs1521179 SNPs in the LINGO4 gene (MIM 609794, Gene ID 339398) and ET risk [78].

Studies on SLC1A2 Gene
The SLC1A2 gene (solute carrier family 1-glial affinity glutamate transporter-member 2, chromosome 11p13-p12, MIM 600300, Gene ID 6506) encodes a member of a family of solute transporter proteins, one of them being the main transporter of the excitatory neurotransmitter glutamate. This neurotransmitter plays an important role in the pathogenesis of ET [79], although a detailed description of this issue is outside of the scope of this review.
The second GWAS described a strong association (odds ratio (95% CI) = 1.59 (1.36-1.84)) between the SNP rs3794087 in the SLC1A2 gene and the risk for definite ET in a GWAS of 990 patients with ET (658 with definite ET) and 1490 healthy controls [60]. This variant could be an interesting genetic marker for ET, since one study has replicated this association in the Chinese [80] and Taiwanese populations [81]. However, this association was not confirmed in additional replication studies [82][83][84], two meta-analyses [54,85], and was not found in several families with ET [52], nor in two-stage GWAS involving 2807 ET patients and 6441 controls of European descent [75].

Results of Other GWAS Studies
A two-stage GWAS involving 2807 ET patients and 6441 controls of European descent reported association of ET with two markers: the intronic SNP rs10937625 in the serine/threonine kinase 32B gene (STK32B, chromosome 4p16.2, Gene ID 55351; the protein encoded by this gene participates in the transfer of phosphate molecules to the oxygen atoms of serine and threonine), and rs17590046 in the PPARG coactivator 1 alpha gene (PPARGC1A, chromosome 4p15.2; MIM 604517, gen ID 10891; PPARGC1 protein acts as a transcriptional coactivator that regulates the genes involved in energy metabolism) [75]. In addition, in a combined analysis, this study found a significant association of the markers rs12764057, rs10822974, and rs7903491 in the catenin alpha 3 gene (CTNNA3, chromosome 10q21.3, MIM 607667, Gene ID 29119; CTNNA3 plays a role in cell-cell adhesion in muscle cells) [75]. Interestingly, this study also described increased expression of STK32B in the cerebellar cortex of ET patients and association between the minor allele of rs10937625 and reduced expression of STK32B in the cerebellar cortex [75].
A replication study in Asian patients involving 469 ET patients and 470 controls confirmed the association of ET with PPARGC1A rs17590046, but not with the STK32B rs10937625 variant [86]. A Canadian study did not find any rare exonic variant in STK32B, PPARGC1A, and in CTNNA3 genes in a whole-exome and whole-genome sequencing study involving 14 autosomal-dominant multiplex ET families and in a targeted massive parallel sequencing study of these 3 genes in 269 ET patients and 287 controls [87].
An estimation of narrow-sense heritability by using the genomic-relationship matrix restricted maximum likelihood (GREML-LDMS) to measure the phenotypic variance ex-plained by genetics in a study involving 1751 ET cases and 5311 controls showed that ET is a highly heritable condition with an important role of common variability, with chromosomes 6 and 21 being those that contained potential causative risk variants influencing genetic susceptibility to ET [88].

Exome and Whole-Genome Sequencing Studies
Exome sequencing of whole-exome sequencing is an efficient strategy to selectively sequence the exome (that is, all the protein-coding regions of the genome). Whole-genome sequencing is the process of determining the complete DNA sequence (both nuclear and mitochondrial DNA) of all of the genome with a single run. During the last decade, there have been an important number of contributions reported looking for genetic markers for ET through exome sequencing and increasing studies even using whole-genome sequencing.

Fused in Sarcoma/Translated in Liposarcoma (FUS/TLS or FUS) Gene
The first whole-exome sequencing published in familial ET found the p.Q290X mutation (rs387907274) in the fused in sarcoma/translated in liposarcoma gene (FUS/TLS, FUS or FUS RNA protein, currently designated as ETM4, chromosome 16p11.2, MIM 137070, gene ID 2521; this gene encodes a protein which is component of a heterogeneous nuclear riboprotein, which is involved in pre-mRNA splicing and the export of fully processed mRNA to the cytoplasm). Mutations of this gene were previously described in families with amyotrophic lateral sclerosis and frontotemporal dementia. The FUS p.Q290X mutation segregated with ET in a large Canadian family, and two rare missense variants (p.R216C-rs267606832 and p.P431L-rs186547381) were found in a further screening of 270 ET cases [89]. Interestingly, some authors generated a transgenic model in Drosophila expressing hFUS-WT and hFUS-Q290X and found that expression of hFUS-Q290X caused a motor dysfunction linked to the impairment in the GABAergic neurotransmission which was partially rescued with gabapentin [90].
The presence of a gene variant causing the amino acid exchange R471C substitution in the gene EWSR1 (EWSR1 binding protein, chromosome 22q12.2, MIM 133450, Gene ID 2130) has been reported, related with FUS (as genes encoding RNA-binding proteins), in a single subject with familial ET from two subsets of ET patients (n = 661) and controls (n = 886) [100].

Mitochondrial Serine Peptidase 2 (HTRA2) Gene
A study in a six-generation consanguineous Turkish kindred identified the p.G399S mutation (rs72470545) in the HtrA serine peptidase 2 (HTRA2, chromosome 2p13.1, MIM 606441, Gene ID 27429; this gene encodes a serine protease localized in the endoplasmic reticulum and in the mitochondria that is released and has a role in apoptosis, and it is suggested to be involved in familial Parkinson's disease (PD), designated as the PARK13 gene as well), as it is proposed to be responsible for both ET and PD [101]. ET was present both in heterozygous or homozygous for this allele, while only homozygotes developed PD, and homozygosity was related with earlier disease onset and higher severity of tremor [101]. However, further studies in different populations showed that HTRA2 mutations were very infrequent or absent [99,[102][103][104][105][106].

Teneurin Transmembrane Protein 4 (TENM4) Gene
A whole-exome sequencing followed by targeted resequencing found missense mutations in the teneurin transmembrane protein 4 gene (TENM4, ETM5, chromosome 11q14.1, MIM 610084, Gene ID 26011, TENM4 protein plays a role in establishing proper neuronal connectivity during development, and is a regulator of axon guidance and central myelination), and showed that TENM4 variants segregated in an autosomal-dominant fashion in three Spanish families with ET [107]. However, studies in 3 cohorts of ET patients and controls detected several missense variants in both groups, but the allele frequencies did not differ significantly among ET patients and control groups [99,108,109]. Interestingly, an ET phenotype has been reported in Tenm4 knockout mice [110].

Sortilin 1 (SORT1) Gene
A whole-exome sequencing and subsequent approaches including functional analysis, in a Spanish family with an autosomal-dominant form of early-onset ET, described a disease-segregating mutation p.Gly171Ala (rs750957839), that was absent in the normal population, in the sortiline 1 gene (SORT1, chromosome 1p21.3-p13.1, MIM 602458, Gene ID 6272; this gene encodes a member of the VPS10-related sortilin family of proteins which are proteolytically processed by furin to generate the mature receptor, that plays a role in the protein trafficking to either the cell surface or subcellular compartments such as lysosomes and endosomes) [111]. The p.Gly171Ala variant impaired the expression of sortilin and decreased mRNA levels of its binding partner p75 neurotrophin receptor implicated in neurotransmission, neuronal apoptosis, and brain injury [111].

Sodium Voltage-Gated Channel Alpha Subunit (SCN11A) Gene
A whole-exome sequencing in a four-generation Chinese family with early-onset familial episodic pain and adult-onset familial ET showed the missense mutation p.Arg225Cys (rs138607170) in the sodium voltage-gated channel alpha subunit gene (SCN11A, chromosome 3p22.2, MIM 604385, Gene ID 11280; SCN11A is a transmembrane glycoprotein complex composed of a large alpha subunit with 24 transmembrane domains and one or more regulatory beta subunits that are responsible for the generation and propagation of action potentials in neurons and muscle. SNC11a proteins are highly expressed in the nociceptive neurons of dorsal root ganglia and trigeminal ganglia and participate in peripheral inflammatory pain hypersensitivity) [112]. The authors suggested that, according to these findings, ET should be considered as a channellopathy.

Notch 2 N-Terminal-Like (NOTCH2NLC) Gene
A study using a research strategy that combined linkage analysis, whole-exome sequencing, long-read whole-genome sequencing, repeat-primed polymerase chain reaction, and GC-rich polymerase chain reaction, in 197 Chinese pedigrees with ET, identified in 11 of them (co-segregating with the disease) an abnormal CGG repeat expansion in the 5 region of the Notch 2 N-terminal-like gene (NOTCH2NLC or ETM6, chromosome 1q21.2, MIM 618025, Gene ID 100996717, mutations in this gene are associated with neuronal intranuclear inclusion disease, or NIID) [113]. Subjects carrying this mutation had higher severity of tremor, and these 11 families showed genetic anticipation [113]. This gene has been designated as ETM6.
Other authors described abnormal CGG repeat expansions (>60) in the NOTCH2NLC in Asiatic patients with ET [114][115][116]. A whole-exome sequencing in 30 members from 15 Chinese families with ET (10 of them diagnosed with ET) found abnormal CGG repeat expansions in 16 subjects, 4 of them developed cognitive impairment, and 3 were finally diagnosed with NIID [114]. Another study identified pathogenic NOTCH2NLC CGG expansions in 4 of 285 Singaporean individuals with sporadic ET (one of them developed motor and cognitive impairment 8-10 years later) and in none of 125 ET patients with a family history of ET, in 52 probands from ET pedigrees, and in 200 controls (although 4 patients with a family history of ET showed 47 to 53 "intermediate" repeats) [115]. Finally, there have been abnormal NOTCH2NLC CGG expansions found in 3 of 28 probands of families with ET [116].
In contrast, abnormal NOTCH2NLC CGG expansions are very rare in European ET patients. One group did not find any abnormal expansion in a series of 111 European patients with ET (74 with "pure" ET and 37 with ET-plus) [117], and only one mutation in another cohort of 203 ET patients [118], and another study did not find any abnormal expansion in 204 ET patients and 408 controls of European ancestry [119].

Results from Other Exome and Whole-Genome Sequencing Studies in ET Patients
A study using SNP arrays followed by whole-exome sequencing in a family with highly penetrant autosomal-dominant tremor (17 members, 5 of them affected with ET) did not identify any copy number variation or mutation related to the ET phenotype [120].
A whole-genome sequencing study involving 40 individuals from 8 ET families identified the deleterious and damaging variant p.Arg456Gln (rs116920450) in the calcium voltage-gated channel subunit alpha1 G gene (CACNA1G, chromosome 17q21.33, MIM 604065, gene ID 8913; the T-type low-voltage activated calcium channel encoded by this gene generates transient currents, owing to fast inactivation, and tiny currents, owing to small conductance, and is thought to be involved in pacemaker activity, low-threshold calcium spikes, neuronal oscillations, resonance, and rebound burst firing) in one family, and a variant in the slit guidance ligand 3 gene (SLIT3, chromosome 5q34-q35.1, MIM 603745, Gene ID 6586; the protein encoded by this gene acts as an axon guidance molecule) in another [122].
Finally, a study in 40 individuals from 8 families with autosomal-dominant ET by using whole-exome sequencing followed by a case-control association study comprising a total of 1310 ET patients and 1366 controls from two cohorts, looking for the association of rare variants with ET risk, found co-segregation with the disease in at least one family with the variants rs749875462, located in the coiled-coil domain containing 183 (CCDC183, chromosome 9q34.3, MIM 615955, Gene ID 849609), rs535864157, located in the matrix metalopeptidase 10 (MMP10, chromosome 11q22.2, MIM 185260, Gene ID 4319), and rs114285050, located in the G protein-coupled receptor 151 genes (GPR151, chromosome 5q.32, MIM 618487, Gene ID 134391) [123]. MMP10 protein belongs to the peptidase M10 family of matrix metalloproteinases, which are involved in the breakdown of extracellular matrix in normal physiological processes, such as embryonic development, reproduction, and tissue remodeling. GPR151 protein belongs to a class of rhodopsin-like family of G-protein-coupled receptors, which also includes somatostatin, opioid, galanin, and kisspeptin receptors. However, the frequency of these variants was very low both in ET patients and in controls in the replicatory case-control association study [123].

Transcriptomic Studies
Transcriptomics technologies are used to study an organism's transcriptome, that is, the sum of all of its RNA transcripts. For the first transcriptomic analysis by direct sequencing of RNA from frozen cerebellar cortex tissue in 33 ET patients compared to 21 normal controls, differential expression analysis between ET patients and controls identified 231 differentially expressed gene transcripts, that contributed to the regulation of axon guidance, microtubule motor activity, endoplasmic reticulum to Golgi transport, and calcium signaling/synaptic transmission [124].
A case-control RNA-sequencing study analyzing cerebellar cortex and dentate nuclei from 16 ET patients and 16 controls, using a multi-omics approach (phenome-wide association study, or pheWAS, genome-wide gene association study, or GWGAS, and transcriptome-wide association study, or TWAS) reported differences in the expression of several genes in ET patients compared with controls (PRKG1-kinase function, SAC3D1mitotic function, SHF-apoptotic function, TRAPPC11-protein trafficking function, NELL2neuronal survival function, and CACNA1A-calcium channel function in the cerebellar cortex, and PLCG2-phospholipase and ALDH3A2-dehydrogenase in the dentate nucleus) [125].
A study with human cerebellar DAOY cells with overexpression of STK32B RNA using an RNA-Seq approach to identify differentially expressed genes (DEGs), by comparing the transcriptome profile of these cells to one of the control DAOY cells, identified dysregulation in several potentially relevant ET genes, including FUS, CACNA1C, and CACNA1A, and differentially expressed genes including olfactory transduction, axon guidance, and calcium ion transmembrane transport genes [126].

Studies on Candidate Genes
Postural and intention tremors are frequently observed in many neurological diseases, such as PD, dystonia, and spinocerebellar ataxias (SCAs), among others. For this reason, in an attempt to search for common etiological factors, many researchers have analyzed the possible role of genes related with some of these diseases in the risk for ET. In fact, the relationship between ET and PD is supported by many epidemiological, genetic, clinical, neuropathological, and neuroimaging data [127,128]. For this reason, many studies looking for possible genomic markers for ET analyzed both genes showing an association with PD in hypothesis-driven case-control association studies or genes related with monogenic familial PD. Table 2 summarizes the results of studies on candidate genes for ET that had been previously related to PD. Although a number of studies showed weak associations between several variants in certain genes with the risk for ET, replication studies did not confirm these associations. Replication studies on the possible contribution of several allelic variants in the genes CYP2C19, CYP2C9/CYP2C8, RIT2, and IL1B, which have shown an association with the risk for ET in single studies, are lacking. Table 3 summarizes the results of studies on other candidate genes, not related to PD, in the risk for ET. In summary, a lack of association of genes have been described related with idiopathic torsion dystonia [34,165,166], spinocerebellar ataxias [167][168][169][170], and fragile X-associated tremor/ataxia syndrome (FXTAS) [171][172][173][174][175][176][177], with genes related with potassium and sodium channels [167,168], GABAergic pathways, calcium and glutamate signaling pathways, and with mitochondrial genes.
Interestingly, despite that the rs1800435 variant in the amino-levulinic acid dehydratase (ALAD) gene has not shown a direct association with ET risk in two studies [178,179], this variant showed association with this risk in interaction with serum lead levels [178] and with the heme-oxygenase 2 (HMOX2) rs1051308G variant [179].
Firstly, a knockout gamma-aminobutyric acid type A receptor subunit alpha1 GABR Alpha1 (GABRA1) mouse showed postural and kinetic tremor and motor un-coordination [188], and, secondly, a knockout mouse for the GABA transporter (GAT)-1 gene showed a motor disorder, including a 25 to 32 Hz frequency tremor [189]. However, three case-control association studies failed to find any association between GABR or GAT genes and the risk for ET [182][183][184].
Since ET is associated with restless leg syndrome [190], two case-control association studies have addressed the possible role of genes previously related to restless leg syndrome in the risk for ET. The initial description of an association of the rs6494696 SNP and a haplotype (rs4489954, rs3784709, rs2241420, rs1026732, and rs6494696) with the risk for ET [185] was not replicated by other study [186].    Lack of association with ET, with the exception of a significantly higher frequency of mutated allelic variants in ET patients exposed to pesticides than in non-exposed. [148] Spain   OR: odds ratio (OR); 95% CI: 95% confidence intervals (CI); NAD: non-available data.          The tremulous dominant mutant Kyoto rat is a rat model that shows a spontaneous tremor resembling human ET. Recent studies have identified a missense mutation (c.1061 C>T, p. A354V) in the hyperpolarization-activated cyclic nucleotide-gated potassium 1 channel (Hcn1) gene [191], and a missense substitution (c. 866T>A, p. I289N) in the potassium calcium-activated channel subfamily N member 2 (Kcnn2) gene in this model [192]. To our knowledge, associations of variants in the equivalent human genes (HCN1, chromosome 5p12, MIM 602780, Gene ID 348980, and KCNN2, chromosome 5q22.3, MIM 605879, Gene ID 3781) with the risk for ET in humans have not been analyzed to date. A study reported the lack of association between CAG repeat expansions in the KCCN3 gene and the risk for ET [167].

Conclusions and Future Directions
Despite that genetic factors have an important role in ET, the search for the responsible gene(s) is still ongoing. The identification of 4 genes/loci in several families through linkage studies (the 3 first reported designated as ETM1, ETM2, and ETM3 genes) has not been confirmed in other family studies, and, moreover, they should only explain a small percentage of familial ET, and the responsible genes remain to be identified. As it was previously mentioned, genetic factors do not explain all cases of ET, and recent studies have suggested a possible role of several environmental factors such as β-carboline alkaloids and ethanol, agricultural work, pesticide, lead, and harmanes, with antioxidant intake and smoking being possible protective agents [27,28,[194][195][196].
Regarding treatment of ET, the drugs that have shown higher efficacy are the betablocker propranolol and the antiepileptic drug primidone, but other drugs such as 1-octanol and octanoid acid, and drugs acting on the glutamatergic system, the extra-synaptic GABAA receptors, or LINGO-1 could be interesting therapeutic options, and injections of botulinum toxin A have shown to be useful in the treatment of refractory ET [197].
The results of GWAS studies reported to date have not been conclusive. Despite the results of the first GWAS that pointed to 2 LINGO1 variants, further case-control association studies showed a weak association of these variants with the risk for familial ET, and a further GWAS failed to replicate the findings [75]. The results of the second GWAS suggested the association of the rs3794087 variant in the SLC1A2 gene with ET but, again, these were not replicated [54,72,73,75]. The role of several variants in the STK32, PPARG1A, and CTNNA3 genes suggested by the third GWAS [75] has not been confirmed by other groups [87], except for PPARG1A in the Chinese population [86].
Exome and whole-genome sequencing studies have found several candidate variants possibly responsible for ET in a small number of families, in split genes such as FUS (designated as ETM4), HTRA2, TENM4 (designated as ETM5), SORT1, SCN11A, NOTCH2NLC (designated as ETM6), NOS3, KCNS2, HAPLN4, USP46, CACNA1G, SLIT3, CCDC183, MMP10, and GPR151. However, replication studies on FUS, HTRA2, TENM4, and NOTCH2NLC genes have found that these mutations are infrequent in other families and populations, while results on mutations of other genes remain to be replicated.
Finally, candidate gene studies have not identified an association with ET risk for genes previously related with other degenerative diseases such as PD, idiopathic torsion dystonia, hereditary ataxias, or others, except for the findings on several variants of CYP2C19 [137], CYP2C9/CYP2C8 [145], RIT2 [158], and IL1B [147], and the increased risk for carriers of the ALAD rs1800435 variant in interaction with serum lead levels [178] or with a variant in the HMOX2 gene [179]. However, the results of these studies have not been replicated so far.
Several factors should be taken into account as limitations in the investigation of genomic markers for ET, such as the lack of disease-specific non-genetic markers for ET (the diagnosis is done on clinical grounds), its frequent overlap with other disorders such as dystonia and PD, and the possible inclusion of phenocopies in genetic studies [23]. Table 4 summarizes the minimal conditions that should be fulfilled by studies trying to address genomic markers for ET. Table 4. Design recommendations for studies focused on genetic research of essential tremor (adapted from text of Reference [23]).

Selection of Index Patients and Controls
Index patients should have a positive family history of ET and be diagnosed with definite and "pure" or "monosymptomatic" ET according to standardized criteria.
Index patients could participate both in family studies and in case-control association studies or family studies.
Inclusion of controls in case-control association studies as "healthy" should imply the absence of a family history of tremor and other movement disorders and the neurological interview and examination to exclude the presence of tremor or other movement disorders.

Selection of Relatives
All available first-degree relatives of the index patient should undergo a clinical examination, including rating scales for tremor.
ET families should be divided into several subtypes, that should be sub-analyzed separately, according to the coexistence or not of other neurological diseases such as dystonia and PD ("pure ET", "ET-dystonia", "ET-PD", "ET-dystonia").

Study Design
Multicenter, multiethnic, and prospective design.
Long-term follow-up to assess further development of PD or other associated disorders in the index patients, in their relatives, or both, and development of ET during the follow-up period by relatives of ET patients who had no tremor in the initial assessment

Blood Collection
Obtention of blood for DNA extraction both from patients, their relatives, and healthy controls. The samples obtained will be used for future genetic studies attempting to establish the role of genetic factors in the different clinical subtypes of ET. Funding: This work was supported in part by Grants RETICS RD16/0006/0004 (ARADyAL), PI15/00303, and PI18/00540 from Fondo de Investigación Sanitaria, Instituto de Salud Carlos III, Madrid, Spain, and GR18145 and IB16170 from Junta de Extremadura, Mérida, Spain. Partially funded with FEDER funds.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest:
The authors declare no conflict of interest.