Studies of Genetic and Proteomic Risk Factors of Amyotrophic Lateral Sclerosis Inspire Biomarker Development and Gene Therapy

Amyotrophic lateral sclerosis (ALS) is an incurable neurodegenerative disease affecting the upper and lower motor neurons, leading to muscle weakness, motor impairments, disabilities and death. Approximately 5–10% of ALS cases are associated with positive family history (familial ALS or fALS), whilst the remainder are sporadic (sporadic ALS, sALS). At least 50 genes have been identified as causative or risk factors for ALS. Established pathogenic variants include superoxide dismutase type 1 (SOD1), chromosome 9 open reading frame 72 (c9orf72), TAR DNA Binding Protein (TARDBP), and Fused In Sarcoma (FUS); additional ALS-related genes including Charged Multivesicular Body Protein 2B (CHMP2B), Senataxin (SETX), Sequestosome 1 (SQSTM1), TANK Binding Kinase 1 (TBK1) and NIMA Related Kinase 1 (NEK1), have been identified. Mutations in these genes could impair different mechanisms, including vesicle transport, autophagy, and cytoskeletal or mitochondrial functions. So far, there is no effective therapy against ALS. Thus, early diagnosis and disease risk predictions remain one of the best options against ALS symptomologies. Proteomic biomarkers, microRNAs, and extracellular vehicles (EVs) serve as promising tools for disease diagnosis or progression assessment. These markers are relatively easy to obtain from blood or cerebrospinal fluids and can be used to identify potential genetic causative and risk factors even in the preclinical stage before symptoms appear. In addition, antisense oligonucleotides and RNA gene therapies have successfully been employed against other diseases, such as childhood-onset spinal muscular atrophy (SMA), which could also give hope to ALS patients. Therefore, an effective gene and biomarker panel should be generated for potentially “at risk” individuals to provide timely interventions and better treatment outcomes for ALS patients as soon as possible.


Introduction
Amyotrophic lateral sclerosis (ALS), or "Lou Gehrig's disease", is a neurodegenerative disorder resulting in the degeneration of motor neurons inside the brain and spinal cord. The disease affects lower motor neurons (LMNs) in the brainstem and spinal cord and upper motor neurons (UMNs) in the motor cortex [1,2]. ALS patients also present additional symptoms, including cognitive decline and personality changes [3,4]. Disease phenotypes usually appear in mid-or later life (50s-70s) but occasionally occur in earlier adulthood (under 30) and to a lesser degree in elderly patients (over 80) [5]. The majority of patients die 3-5 years after the first disease symptoms, whilst a juvenile minority can exhibit prolonged survival (more than 10 years) [6]. Most ALS cases are sporadic, but approximately 5-10% of disease cases are associated with a positive family history of the disease. Disease inheritance can be autosomal dominant, recessive, de novo and X-linked (XL). ALS is a genetically complex disease with >50 genes associated with disease onset. The disease (XL). ALS is a genetically complex disease with >50 genes associated with disease onset. The disease usually starts with muscle issues such as tightness, weakness, and cramps, resulting in patients exhibiting reduced grip strength and poor balance. Initial symptoms can manifest in one of several body parts, such as bulbar-onset, arm-onset, or leg-onset. As the disease progresses, the facial muscles lose coordination, making speaking, breathing, and swallowing increasingly difficult. In the later stages, paralysis usually ensues, with patients losing the ability to walk unaided, rendering them entirely bedridden. Since the disease also affects the respiratory muscles, respiratory failure is often the leading cause of death [3,7]. In this regard, ALS and frontotemporal dementia (FTD) share several clinical, genetic, and pathological features. Approximately half of ALS patients develop FTD-like phenotypes (cognitive decline, personality changes), and 30% of FTD patients go on to develop motor neuron impairments [8][9][10].

Primary Causative Genes and Function
To date, there are four mendelian genes that cause ALS: The Cu-Zn superoxide dismutase (SOD1), TAR DNA Binding Protein (TARDBP), Fused in Sarcoma (FUS) gene and Chromosome 9 Open Reading Frame 72 (c9orf72) ( Table 1). Since its identification 30 years

Genes and Molecular Mechanisms Contributing to ALS Primary Causative Genes and Function
To date, there are four mendelian genes that cause ALS: The Cu-Zn superoxide dismutase (SOD1), TAR DNA Binding Protein (TARDBP), Fused in Sarcoma (FUS) gene and Chromosome 9 Open Reading Frame 72 (c9orf72) ( Table 1). Since its identification 30 years ago, almost 200 SOD1 mutations have been attributed to familial and sporadic forms of ALS; most of the disease-causing variants are missense mutations and comprise 15-30% of inherited ALS cases and less than 2% of sporadic ALS cases [17]. The TARDBP gene constitutes 4.2% and 1.5% of European and Asian fALS cases, respectively, whereas FUS is more common in Asian than European fALS cases (6.4% and 2.8%, respectively). Finally, the c9orf72 gene is common in European familial ALS cases (more than 30%), while it may be a rare causative factor among Asian fALS cases (less than 3%). However, these genes are rarely encountered in either European or Asian (less than 8% and 3%, respectively) sporadic ALS (sALS) cases [18]. Table 1. ALS involvement and significant mutations in SOD1, C9orf72, TARDBP and FUS genes [18]. The SOD1 gene is located on chromosome 21, producing a 153 conserved-amino-acidlong protein in the cytoplasm, nucleus and mitochondria. In the active configuration, the protein exists as a homodimer enzyme containing eight antiparallel beta strands housing copper and zinc ions, which play a role in the dimerization and disulfide bond stabilization (between Cys57 and Cys146). SOD1 protein functions as an antioxidant enzyme, which protects the cells from oxidative stress by the conversion of catalyzing toxic superoxide anions into hydrogen peroxide and oxygen molecules, lowering the steady state of radical ions (nitric oxide, superoxide, and peroxynitrite), regulating apoptosis [10,[19][20][21]. SOD1 dysfunctions may also impact another motor disease, spastic tetraplegia and axial hypotonia (STAHP) [22]. Most SOD1 mutations are heterozygous with autosomal dominant inheritance, but homozygous variants with autosomal recessive inheritance have been reported in rare cases [23]. SOD1 mutations increase protein instability and the capacity to misfold due to metal ion and disulfide bond deficiencies. Collectively, such deficits can result in oligomerization, abnormal protein aggregation, and the formation of amyloid-like aggregates [24][25][26]. In addition, the potential prion-like properties of pathogenic SOD1 can lead to nuclear and mitochondrial DNA damage [19], with resultant amyloid-like fibrils shown to be intracellularly transmissible [24]. SOD1 variants disrupt cytosolic and mitochondrial calcium homeostasis, leading to abnormal cell functions and alterations of glial cell-motor and skeletal muscle-motor neuronal crosstalk, triggering apoptosis [27].
Several geographical mutations are associated with SOD1; for example, A4V is a common mutation in American populations, which is thought to accelerate progressive familial ALS, reducing disease duration to 2 years or less. The disease initially presents with limb and bulbar muscle weakness, primarily affecting the lower motor neurons [28,29]. Common among Japanese patients, H46R is a variant ALS-causing mutation associated with slower disease progression (up to 15 years). The H46R mutation is thought to impair the copper-binding site of SOD1 [30][31][32]. Another common ALS mutation associated with rapid disease progression is SOD1 G93A. Transgenic mice carrying this variant are frequently used in ALS drug screening [33,34]. Figure 2 summarizes the possible impact of SOD1 mutations (including A4V or H46R) in ALS.
with limb and bulbar muscle weakness, primarily affecting the lower motor neurons [28,29]. Common among Japanese patients, H46R is a variant ALS-causing mutation associated with slower disease progression (up to 15 years). The H46R mutation is thought to impair the copper-binding site of SOD1 [30][31][32]. Another common ALS mutation associated with rapid disease progression is SOD1 G93A. Transgenic mice carrying this variant are frequently used in ALS drug screening [33,34]. Figure 2 summarizes the possible impact of SOD1 mutations (including A4V or H46R) in ALS. In 2006, prion-like mutations were identified in the C-terminus-encoding regions of the TARDBP gene and deemed causal for autosomal dominant ALS and associated TDP43 proteinopathy. The TARDBP gene contains six exons on chromosome 1, expressed ubiquitously, residing mainly in the nucleus [35]. The TDP43 protein encoded by this gene is essential for RNA transcriptional regulation, splicing and stabilization, potentially impacting metabolism [36,37]. In addition, TARDBP protein could bind to different proteins (such as hnRNP proteins) and influence other protein-protein interactions [38], some of which are integral to mitochondrial function. Its overexpression and mislocation can lead to TDP43 proteinopathy, impairing mitochondrial functions [39]. The protein is 414 amino acids long, possessing several domains, including the N-terminal domain (between aa1 and 76), domain for nuclear localization signal (NLS; aa82-98), RNA recognition motif-1 (RRM1, aa106-176) and RNA recognition motif-2 (RRM2, aa191-259), and the C-terminal domain (aa 274-414). The RNA recognition motif-2 contains the specific sequence of nuclear export signal (aa239-250). The N-terminal domain is involved in the self-regulation of TARDBP transcription and RNA splicing through homodimer formation [40]. The RNA recognition motifs participate in numerous processes, such as the direct binding of DNA or RNA, splicing and translation, limiting oligomer formation and potential aggregation [40,41]. Finally, the C-terminal domain contains a prion-like domain between residues 274 and 414. This domain comprises polar amino acids essential in TDP43 aggregation [41][42][43]. The majority of ALS-related TARDBP mutations are associated with the misfolding aggregation of TDP43 (for example, Q331K, M337V, Q343R, N345K, R361S, G376D, or N390D) protein [44] through different mechanisms. For example, G376D is thought to increase the mislocalization of protein into the cytoplasm [45]. Furthermore, additional In 2006, prion-like mutations were identified in the C-terminus-encoding regions of the TARDBP gene and deemed causal for autosomal dominant ALS and associated TDP43 proteinopathy. The TARDBP gene contains six exons on chromosome 1, expressed ubiquitously, residing mainly in the nucleus [35]. The TDP43 protein encoded by this gene is essential for RNA transcriptional regulation, splicing and stabilization, potentially impacting metabolism [36,37]. In addition, TARDBP protein could bind to different proteins (such as hnRNP proteins) and influence other protein-protein interactions [38], some of which are integral to mitochondrial function. Its overexpression and mislocation can lead to TDP43 proteinopathy, impairing mitochondrial functions [39]. The protein is 414 amino acids long, possessing several domains, including the N-terminal domain (between aa1 and 76), domain for nuclear localization signal (NLS; aa82-98), RNA recognition motif-1 (RRM1, aa106-176) and RNA recognition motif-2 (RRM2, aa191-259), and the C-terminal domain (aa 274-414). The RNA recognition motif-2 contains the specific sequence of nuclear export signal (aa239-250). The N-terminal domain is involved in the self-regulation of TARDBP transcription and RNA splicing through homodimer formation [40]. The RNA recognition motifs participate in numerous processes, such as the direct binding of DNA or RNA, splicing and translation, limiting oligomer formation and potential aggregation [40,41]. Finally, the C-terminal domain contains a prion-like domain between residues 274 and 414. This domain comprises polar amino acids essential in TDP43 aggregation [41][42][43]. The majority of ALS-related TARDBP mutations are associated with the misfolding aggregation of TDP43 (for example, Q331K, M337V, Q343R, N345K, R361S, G376D, or N390D) protein [44] through different mechanisms. For example, G376D is thought to increase the mislocalization of protein into the cytoplasm [45]. Furthermore, additional mutations, such as D169G or R361S, are associated with larger stress granular production and disease acceleration. Mutant TARDBP may result in impaired protein-protein interactions between TARDBP protein and its binding partners, such as FUS or ataxin. The TDP-43 aggregations may disturb the autophagic mechanisms and proteosome formation. Furthermore, mutations in TARDBP could impact the processing of miRNAs and lncRNAs ( Figure 3) [46][47][48].
mutations, such as D169G or R361S, are associated with larger stress granular production and disease acceleration. Mutant TARDBP may result in impaired protein-protein interactions between TARDBP protein and its binding partners, such as FUS or ataxin. The TDP-43 aggregations may disturb the autophagic mechanisms and proteosome formation. Furthermore, mutations in TARDBP could impact the processing of miRNAs and lncRNAs ( Figure 3) [46][47][48]. The FUS gene could also be a significant causative gene for ALS. The initially suspected oncogene is located on chromosome 16 and encodes a nuclear RNA-binding protein. The respective protein has been implicated in RNA metabolism, transcription regulation, RNA transport, micro-RNA processing, and genome maintenance [49] and is considered a key player in neuronal homeostasis (neuronal plasticity and dendrite integrity) [50]. The protein is 526 amino acids long, containing several domains, including an Nterminal QGSY-rich region (1-165aa), two glycine-rich or prion-like domains (aa165-222 and aa391-407), with a nuclear export signal motif (NES, 280-377aa), three arginine-glycine rich (RGG) domain boxes (RGG1-aa222-280, RGG2-aa377-418, RGG3-aa454-501), zinc finger (aa418-454) and a C-terminal area, called proline-tyrosine nuclear localization signal (aa501-526). Inside RGG2, another prion-like domain was identified, which is located between residues 391 and 407 ( Figure 4). Along with The RNA recognition domains zinc finger (Znf), motifs could be involved in FUS interaction with RNA. In addition, RGG domains are also implicated in nucleotide binding and other cellular pathways, such as cell division or the formation of stress granules. The N-terminal region and the glycinerich (GC) domain may impact transcription regulation. Furthermore, the GC domain may also influence DNA repair mechanisms, nuclear localization signals, and importation [48,49,51,52]. The FUS gene could also be a significant causative gene for ALS. The initially suspected oncogene is located on chromosome 16 and encodes a nuclear RNA-binding protein. The respective protein has been implicated in RNA metabolism, transcription regulation, RNA transport, micro-RNA processing, and genome maintenance [49] and is considered a key player in neuronal homeostasis (neuronal plasticity and dendrite integrity) [50]. The protein is 526 amino acids long, containing several domains, including an N-terminal QGSY-rich region (1-165aa), two glycine-rich or prion-like domains (aa165-222 and aa391-407), with a nuclear export signal motif (NES, 280-377aa), three arginine-glycine rich (RGG) domain boxes (RGG1-aa222-280, RGG2-aa377-418, RGG3-aa454-501), zinc finger (aa418-454) and a C-terminal area, called proline-tyrosine nuclear localization signal (aa501-526). Inside RGG2, another prion-like domain was identified, which is located between residues 391 and 407 ( Figure 4). Along with The RNA recognition domains zinc finger (Znf), motifs could be involved in FUS interaction with RNA. In addition, RGG domains are also implicated in nucleotide binding and other cellular pathways, such as cell division or the formation of stress granules. The N-terminal region and the glycine-rich (GC) domain may impact transcription regulation. Furthermore, the GC domain may also influence DNA repair mechanisms, nuclear localization signals, and importation [48,49,51,52].  A significant ratio of FUS mutant ALS cases is associated with an early-onset disease (under 40 years of age) [51,52], many of which are either gain-of-function or loss-of-function variants, influencing multiple pathways. FTD and ALS with FUS mutations may present TDP43 inclusions in their brain, but TDP43 proteinopathy can also be absent in ALS patients with FUS, particularly when the abnormal transcript utilizes TDP43-independent  A significant ratio of FUS mutant ALS cases is associated with an early-onset disease (under 40 years of age) [51,52], many of which are either gain-of-function or loss-of-function variants, influencing multiple pathways. FTD and ALS with FUS mutations may present TDP43 inclusions in their brain, but TDP43 proteinopathy can also be absent in ALS patients with FUS, particularly when the abnormal transcript utilizes TDP43-independent pathways [53]. For example, the postmortem brains of ALS patients can exhibit active p38 MAPK elevations, suggesting that aberrant expression may perturb the p38 MAPK pathway, leading to stress-induced neuronal loss [54]. Moreover, other FUS mutations are thought to exercise their toxicity by reducing ion channel expression and synaptic activity, inhibiting intra-axonal translation [55]. Furthermore, additional mutations such as R521G, R522G, or ∆Exon15 may result in the abnormal splicing of FUS transcript and autoregulation of FUS levels [56]. As a result, the abnormal protein could induce nonsense-mediated decay (NMD) and promote the up-regulation of several pro-NMD factors (Up-Frameshift Suppressor 3 Homolog B, UPF3b; and UPF1 RNA Helicase and ATPase, UPF1) and the down-regulation of NMD inhibitors, such as UPF3a, leading to NMD hyperactivity and neurotoxicity [57]. Additional investigations have identified FUS immunoreactive cytoplasmatic inclusions inside neurons and glial cells. The inclusions usually occur in the nucleus [58], but can induce conformational changes in cytoplasmic proteins due to their prion-like domain, leading to further aggregation [59]. Other mutations include R521L, R521G, R521H, R521C, P525L, and R495Qfs527X or R495X, primarily associated with FUS protein mislocalization and mitochondrial impairment. Impairment can manifest as mitochondrial shortening and fragmentation, whilst the mutant R495X is thought to reduce the organelle's size [60]. In addition, other FUS mutations may down-regulate genes (such as cyclooxygenases, COX) essential to oxidative phosphorylation, impacting skeletal muscle functions [61].
The next causative pathogenic variant is the hexanucleotide repeat expansion mutation (G4C2; (g.26724GGGGCC (3_23)) located between the non-coding exons 1a and 1b of the c9orf72 gene. DeJesus Hernandez et al. (2011) revealed that c9orf72 was present in 11.7% of familial FTD and 23.5% of familial ALS patients. In healthy individuals, repeat expansions typically range from 3 to 30. G4C2 is the abbreviation of hexanucleotide repeat expansion mutation (HREMs) and ranges from 30 to 100 in ALS patients, although G4C2 expansions in the hundreds or thousands have been reported in sALS patients [62]. In a normal brain, the standard splicing of the c9orf72 transcript results in three different isoforms: (1) variant 1, which contains exon 1b and coding exons 2-11, but not exon 1a or the G4C2 repeats; (2) variant 2, which contains the exon 1a and repeat expansion, but only contains coding exons 2-5; and (3) variant 3, containing exon 1a, the repeat expansion region, and all coding exons. Variants 1 and 3 can manifest as a 481 amino-acid-long protein, while variant 2 was predicted to encode a 222 amino-acid-long protein. However, in the brains of FTD or ALS patients with repeat expansions, the c9orf72 variant 1 is often undetectable. An initial study suggested that the repeat expansion could limit the expression of c9orf72 variant 1, significantly impacting brain development and function [63][64][65]. Moreover, longer G4C2 expansions were suggested to alter the methylation pattern of the c9orf72 gene and its promoter activity. Repeat expansions were associated with a more significant degree of methylation, leading to reduced transcriptional activity and haploinsufficiency, resulting in dipeptide repeat protein (DPR) and glutamate receptor elevations. Furthermore, intermediate repeats, ranging between 7 and 24 G4C2 units, showed a slight increase in methylation compared to a low number of repeats (2-6), suggesting that "intermediate alleles "may also act as risk alleles [65,66]. Even though c9orf72 participates in loss-of-function (LOF) mechanisms and has the potential for gain-of-function toxicity (GOF) (Figure 5), the variants alone are not considered a primary cause of motor degeneration [67]. Another factor potentially contributing to c9orf72-related disorders is the translation of expanded intronic G4C2 through an alternative START codon (such as Kozak-sequence-or repeat-associated non-ATG-initiated translation). The abnormal translation of G4C2 repeats could result in DPR, such as poly-(Gly-Ala or polyGA), poly-(Gly-Pro or polyGP) and poly-(Gly-Arg or polyGR) proteins, which have a strong tendency to aggregate due to their high degree of

Additional Rare ALS-Related Genes
Even though SOD1, TARDBP, c9orf72, and FUS mutations are relatively common in familial ALS, several additional genes were identified which could play a role in ALS onset. In addition, approximately 44.5% of European and~60% of Asian fALS cases are associated with rare genetic factors. More than 50 genes (Table 2, Figure 6) have been associated with ALS, and this number is set to increase [18]. Table 2. Examples of rare genetic factors and their potential roles in ALS ("fs" represents frameshift mutation) [18].   . Potential biological processes/functions affected by common and rare ALS-related genes during ALS progression [18,74]. The differently colored edges may represent the the proterin-protein associations. The light blue means, association was verified from curated databases. Pink means, interactions were proven experimentally. The darker green means "gene neighbourhood". Red means possible gene fusion. Darker blue means gene co-occurrence. Light green means "text mining" or possible association from literature. The black means that the genes may interact through co-expression. The purple edge means that these genes may share homology. Network analyses were performed by STRING software version 11.5 (https://string-db.org/, accessed on 20 July 2023).  Figure 6. Potential biological processes/functions affected by common and rare ALS-related genes during ALS progression [18,74]. The differently colored edges may represent the proterin-protein associations. The light blue means, association was verified from curated databases. Pink means, interactions were proven experimentally. The darker green means "gene neighbourhood". Red means possible gene fusion. Darker blue means gene co-occurrence. Light green means "text mining" or possible association from literature. The black means that the genes may interact through coexpression. The purple edge means that these genes may share homology. Network analyses were performed by STRING software version 11.5 (https://string-db.org/, accessed on 20 July 2023).

Genes Involved in Protein Trafficking
Charged multivesicular body protein 2 (CHMP2B) is on chromosome 3 and encodes the 213 amino-acid-long protein. This essential protein plays multiple roles in the Endosomal Sorting Complex Required For Transport III (ESCRT-III), from the biogenesis and sorting of endosomes to the trafficking of multivesicular bodies (MVBs), proteostasis, and neuronal homeostasis. CHMP2B mutations are not unique to ALS and have been associated with numerous diseases, including FTD and progressive muscular atrophy (PMA) [75,76]. However, the mutation in CHMP2B intron5 was recently implicated in the hyperphosphorylation of TDP-43 via the ubiquitin-proteasome system (UPS), decreasing ubiquitination and turnover of casein kinase 1 (CK1), highlighting a genetic link between TARDBP and CHMP2B and the associated ALS pathology [77].
Another gene implicated in ALS is the Valosin-containing protein (VCP) which exhibits chaperon-like properties that regulate the ubiquitin-proteasome system (UPS) and autophagic processes. VCP mutations are rare causative factors for ALS [78] and are associated with other diseases, such as inclusion body myopathy (IBM), Paget's disease, and FTD. VCP plays a crucial role in UPS-mediated protein degradation processes, and its impairment leads to the accumulation of ubiquitinated proteins. However, the specific role of VCP in autophagy remains unclear, as the protein has been implicated in several processes, including autophagy activation, autophagic complex maturation, autophagosome, and lysosome fusion and homeostasis [79]. In addition to VCP, the vesicle-associated membrane-protein-associated protein B (VAPB) gene plays a role [80,81] in vesicle trafficking and calcium homeostasis, modulating neuromuscular junction activity. The observable impact of VAPB mutants (P56S) includes disorganized endoplasmic reticulum (ER) structures and an enhanced propensity of a cell to form aggregates [82,83]. Another rare gene implicated with ALS is phosphoinositide 5-phosphatase (FIG4), which regulates endosomal/lysosomal vesicle transport [84]. FIG4 has a potential role in Charcot-Marie-Tooth disease, ALS, primary lateral sclerosis (PLS), recessively inherited hereditary motor and sensory neuropathy, or Yunis-Varón syndrome [84]. Furthermore, additional FIG4 mutations in ALS include I41T, F254Sfs*8 or Y647C, which can exhibit autosomal dominant inheritance patterns [85][86][87][88]. Spastacin (SPG11) is a primary causative factor for autosomal recessive hereditary spastic paraplegia (HSP). Still, it may also play a role in other neurodegenerative diseases, including autosomal recessive juvenile ALS and CMT. SPG11 could play a role in multiple processes, including lysosome trafficking, cholesterol transport, autophagy, and calcium homeostasis. ALS's SPG11 mutations (such as c.704_705delAT and c.5199delA) may be associated with a lack of thinned corpus callosum and cognitive and ocular abnormalities [89,90]. NIPA Magnesium Transporter 1 (NIPA1) was initially associated with Prader-Willi/Angelman syndrome 1 or Type 6 HSP, but it may also impact ALS [91][92][93]. NIPA1 impacts magnesium metabolism inside the cells. It may also affect endosome regulation inside neurons and epithelial cells [92]. The GCG repeat (or polyalanine repeat) expansion in the 5 end of NIPA1 was suggested to impact ALS risk. ALS patients with NIPA1 CGG repeat expansions may be associated with increased ALS risk, earlier disease onset, and reduced survival [91,93]. The impact of NIPA1 in ALS was verified by a large-scale replication analysis performed by Tazelaar et al. in 2019. The pathogenic mechanism of CGG repeat expansion remained unclear. The polyalanine repeats may result in abnormal protein conformation, protein-protein interactions, and aggregation [94].
Ubiquilin-2 (UBQLN2) is a member of the ubiquitin proteosome system and plays a role in the homeostasis of different proteins and the quality control of mitochondrial proteins. Also, UBQLN2 participates in the degradation of ubiquitinated proteins through proteosome and autophagic pathways. Different mutations have been linked to FTD and ALS [99,100]. Specifically, ALS-related UBQLN2 mutations (such as R497H, P506S, and P509S) could restrict proteosome-mediated degradation due to their reduced binding ability, leading to aggregation and neurotoxicity [95]. Furthermore, mutations in UBQLN2 may decrease the expression of ATPase H+ Transporting V1 Subunit G1 (ATP6v1g1), leading to impaired autophagy due to failed acidification and maturation [100,101].
The chaperon small heat-shock protein family B member 1 (HSPB1) mutations are thought to impact CMT and distal hereditary motor neuropathy and, therefore, considered a potential risk factor for ALS. HSPB1 mutations are suspected to enhance protein misfolding and aggregation. For example, HSPB1 mutations reported in sporadic ALS cases, such as Q190H and A204Gfs* 6, were assumed to reduce the protein stability and chaperon activity, rendering motor neurons more sensitive to stress. HSPB1 mutations may also be associated with impaired autophagy by reducing the ability of Sequestosome 1 (SQSTM1)/p62 body formation [102][103][104][105].

Variants with the Potential to Impact Autophagic Pathways
Sequestosome 1 (SQSTM1) gene encodes the sequestosome 1/p62 protein, with suspected roles in FTD, Paget bone disease, or familial and sporadic forms of ALS. The SQSTM1 mutations could impact the turnover of p62, promoting aggregation or autophagy, and impair its interaction with the Kelch-like ECH-associated protein 1 (KEAP1) [106]. For example, ALS-associated mutation L341V in SQSTM1, may impair autophagy by reducing SQSTM1 affinity for the lipid-anchored form of LC3 protein, impairing phagophore development and reducing cell survival [107]. Another mutation, G427R, may abolish an important phosphorylation site in serine 351, limiting interactions between KEAP1 and SQSTM1, reducing the expression of erythroid-derived 2 and Nuclear Factor Erythroid 2-Related Factor 2 (NFE2L2/Nrf2)-targeted genes, elevating the TDP-43 associated stress granule formation under oxidative stress [108].
TANK-binding kinase 1 (TBK1) is an NF-κB activator kinase protein that could impact ALS and FTD. TBK1 protein consists of a kinase, a ubiquitin-like, and two scaffold dimerization domains, permitting its participation in several pathways, including mitophagy (mitochondrial clearance) and autophagy by phosphorylating several adaptor proteins, such as optineurin (OPTN) or Parkin (PRKN). In addition, TBK1 may be involved in T-cell activation and autoimmunity [109,110]. TBK1 mutations could potentially generate a variety of disease phenotypes via several mechanisms. For example, R47H, D135N, G217R, or R228H expression could result in lower autophosphorylation. Specifically, G217R could decrease mitochondrial fragmentation and the interaction between TBK1 and PRKN. Other mutations, such as G217R, R357Q, and M559R, could disrupt TBK1 assembly and mitophagy events. In addition, D135N or D559N mutations could impair mitophagy by inhibiting LC3 recruitment, increasing the likelihood of accumulation of damaged mitochondria. Moreover, these mutations could also elevate mitochondrial reactive oxygen species (ROS), promoting aggregate formations and neuronal cell death and neuroinflammation [110,111].
Optineurin (OPTN) was initially reported as a causative gene for glaucoma and was later found to contribute to ALS pathology. The gene is expressed in several human organs, such as the brain, skeletal muscles, liver, kidney, or placenta muscle [112]. The OPTN functions as a receptor for autophagy after being phosphorylated by TBK1 on serine 177, 473 and 513 [113]. The autophagic adaptor plays a role in several forms of autophagy, including mitophagy, xerophagy, receptor recycling, or aggrephagy, as well as inhibiting the NF-kappa B pathway reducing neuroinflammation [114][115][116]. The inheritance pattern of OPTN mutation-associated ALS could be either autosomal dominant or autosomal recessive. However, the disease-related mechanisms can vary between these two traits.

Genes Impacting Cytoskeletal Dynamics or Microtubule Functions
Dynactin 1 (DCTN1) encodes the p150Glued protein, a member of the dynactin protein complex. DCTN1 protein directly binds microtubules via the highly conserved N-terminal glycine-rich and cytoskeleton-associated protein (CAP-Gly) domain and plays a role in retrograde vesicle transport. DCTN1 can interact with TDP43 forming co-aggregates in the cytoplasm. Most disease-related mutations of DCTN1 (such as G71A) are located in the glycine-rich domain [117][118][119]. NMA-related kinase 1 (NEK1) is a serine/threonine kinase and a mitotic checkpoint protein, which could control cell cycle progression and mitosis. It could also be involved in DNA repair. NEK1 was also suggested to regulate the cilia formation, microtubule stability, and dynamics [120]. NEK1 dysfunctions are associated with brain diseases such as hydrocephalus and mental retardation. In 2016, NEK1 was initially associated with ALS since a loss-of-function mutation, R261H, was observed as a risk factor for the disease [121]. Additional loss-of-function mutations (such as N181S, G399A, and M545T) were also reported, reducing NEK1 expression and haploinsufficiency [122].
Profilin-1 (PFN1) is a 140 amino acid actin-binding protein that promotes nucleotide exchange on actin, resulting in the conversion of monomeric ADP-actin to ATP-actin [123,124] and the formation of PFN1-ATP-actin complexes. Additionally, PFN1 binds to phosphoinositides and an extensive network of proteins with poly-L-proline stretches. Some mutations, such as G118V or M114T, could result in disturbances in the polyproline-binding region of PFN1, leading to abnormal actin assembly. Another variant, PFN1 C71G, potentially reduces PFN1 expression [125], whilst others, such as M114T or E117G, could also result in abnormal autophagic processes resulting in mitochondrial dysfunctions [126].
Peripherin (PRPH) is a type III intermediate filament, a multifunctional protein implicated in neuronal processes, cytoskeletal dynamics, axonal progressions, transport and myelination, neurite stability, and growth, with possible roles in processing DNA/RNA, mitochondrial functions, and vesicle transport. In addition, PRPH is considered a risk factor for diseases such as ALS, Charcot-Marie-Tooth Type 2B, diabetes, or enterovirus-A71 infection [127,128]. PRPH is overexpressed in injured neurons and promotes axon regeneration. Mutations in PRPH, such as R133P and D141Y [129] or splice site variants [130], may form intracellular aggregates promoting neurotoxicity. In addition, the suppression of Neurofilament Heavy Chain (NEFH) or Neurofilament Modest Chain (NEFM) expression via a small group of ALS-linked miRNAs is shown to perturb cytoskeleton composition and disrupt axon structure and transport exhibiting impaired N-methyl-D-aspartate (NMDA)-mediated calcium influx [131].
The Tubulin Alpha 4a (TUBA4A) gene is a rare risk gene for FTD and ALS. It encodes a subunit for alpha-tubulin, which forms a heterodimer with beta-tubulin and is involved in microtubule formation [132,133]. Mutations in the TUBA4A gene, such as W407X, could disrupt the assembly of alpha and beta tubulins, leading to reduced assembly of microtubules. ALS cases with TUBA4A were spinal onset, but upper motor neurons can also be involved. Cognitive dysfunctions and frontal involvement may also be involved [133,134]. Alsin (ALS2) is a Rho Guanine Nucleotide Exchange Factor, which could impact young onset autosomal recessive ALS or hereditary spastic paraplegia [135,136]. Mutations in ALS2 can be either missense, such as I94V, E159K, M368V or affecting RNA splicing, such as IVS7 + 3A>G [137]. The exact pathological role of mutant ALS2 remains unclear, but ALS2 may impact cytoskeletal dynamics [135]. It was also suggested that the knockout of ALS2 could significantly damage postsynaptic subsynaptic reticulum (SSR) functions. ALS2 defects may also impair late endosome trafficking, synaptic development, and neuroprotection [137].

Genes Involved in the Regulation of DNA/RNA Processing
In addition to TARDBP and FUS, other genes, such as sentaxin or angiogenin, can potentially influence RNA-related processes. Sentaxin (SETX) is a helicase for DNAs and RNAs and is involved in different processes, including DNA repair, replication or recombination, RNA processing, and translation [138]. Several SETX mutations (such as N264S, M386T, or T1118I) have been implicated in impaired DNA repair and gene expression [139] associated with autosomal recessive cerebellar ataxia [138] and rare autosomal dominant juvenile ALS.
The TATA box-binding protein (TBP)-associated factor 15 (TAF15) protein is closely related to FUS, since they have similar structures and both belong to heterogeneous nuclear ribonucleoparticle (hnRNP) proteins. TAF15 was verified to regulate RNA synthesis and also protein splicing. Mutations in TAF15 (A31T, D386N, R388H, or R395Q) may be related to familial and sporadic ALS [140,141].
Angiogenin (ANG) was suggested as a rare risk factor for Alzheimer's disease (AD), Parkinson's disease (PD), and ALS. ANG is a ribonuclease A protein which activates ribosomal RNA regulating its transcription. Moreover, its involvement could also extend to angiogenesis, cell survival, and neuronal growth [142]. For example, multiple ANG mutants (such as H13R, K40R, K54R V103I, H114R, or R121C) have been associated with impaired RNase activity, leading to a lower rate of cell proliferation and angiogenesis [143].
Heterogeneous Nuclear Ribonucleoprotein A1 (HNRNPA1) is a member of a heterogeneous nuclear ribonucleoprotein, which could impact mRNA processing, including splicing and RNA assembly, regulating the stability of RNA and gene expression. Mutations in the prion-like domain of HNRNPA1, such as P288A, G264R, or D262V, were identified in ALS cases but could also impact other diseases, such as FTD, multisystem proteinopathy (MSP), or inclusion body myopathy (IBM) [146]. In ALS patients, the immunoreactivity of HNRNPA1 was reduced in the motor neurons of the spinal cord. This process may be associated with abnormal RNA processing [147]. TDP43 could regulate the functions of HNRNPA1, leading to alternative splicing and aggregation [148,149].
Matrin3 (MATR3) is a conserved DNA-RNA binding protein, which may play a role in several mechanisms, such as mRNA stabilization, DNA repair and transport or splicing. Mutations in MATR3 could impact ALS, FTD, and distal myopathy [150]. Different MATR3 mutations were identified in ALS, such as S85C, F115C, P154S, or T622A, which were associated with impaired RNA processing, such as abnormal mRNA export, especially the mRNA for TDP43 or FUS [151,152].
The Sigma Non-Opioid Intracellular Receptor 1 (SIGMAR1) gene plays a role in neuroprotection, neuroplasticity, and mitochondrial calcium transport through the inositol trisphosphate receptor (IP3) receptor, which impacts both FTD and ALS [161,162]. Thus, associated mutations (E102Q) have the potential to perturb calcium-dependent mitochondrial ATP production, leading to suppressed proteosome activity under high ER stress and lower mitochondrial respiration. Moreover, SIGMAR1 mutants may contribute to TDP43 mislocalization and ubiquitinoylation, leading to additional mitochondrial and autophagosome impairment in motor neurons [162]. Also, mutant SIGMAR1 could result in abnormal autophagy and stress granule aggregation. Furthermore, mutant SIGMAR1 expression may result in abnormal cytoplasmic RNA binding protein assemblies involving TDP43, FUS or MATR3 proteins [163].

Genes, Which Could Impact Other Pathways
Erb-B2 Receptor Tyrosine Kinase 4 (ERBB4 or HER4) belongs to the tyrosine kinase family and is member of the epidermal growth factor receptor subfamily. ERBB4 could bind to neuregulin proteins (NRGs), leading to ERBB4 activation and elevated tyrosine kinase activity. Activated ERBB4 may be associated with multiple cellular pathways, including neurodevelopment. ERBB4 gene was established to impact cancers, but it may also play a role in ALS. Mutations, such as R927Q or R1275W, may disrupt the tyrosine kinase activity of ERBB4, resulting in reduced neuregulin downstream signaling. These mutations were associated with slow progressive ALS, where both upper and lower motor neurons were involved, but cognitive function was spared [164,165]. Annexin 11 (ANXA11) is a calcium-dependent phospholipid-binding protein that could impact ALS and FTD onset. Mutations in ANXA11, such as D40G, G228Lfs*29, H390P, or R456H, could be associated with both gain-and loss-of-function mechanisms, including abnormal protein aggregation and loss of intracellular calcium responses, respectively. Also, the mutations may result in abnormal stress granule dynamics. ANXA11 protein could co-aggregate with other ALS-related RNA-binding proteins, such as FUS or hnRNPA1, inside lower motor neurons or the brain [166][167][168]. Calcium-Responsive Transactivator (CREST or SS18L1) is a calcium-dependent transcription activator involved in dendrite growth. It was established to impact cancer onset, but its role in ALS is also investigated. The exact mechanisms of how CREST mutations may be related to ALS remained unclear. Mouse studies revealed that an ALS-related CREST mutation, Q388X, could repress the cytokine expression (C-C Motif Chemokine Ligand 2; Ccl2 or C-X-C motif chemokine ligand 10; Cxcl10) inside neurons [169]. D-Amino Acid Oxidase (DAO) could cause ALS and bipolar diseases. DAO protein regulates brain D-serine, which could act as a neuromodulating factor. Also, DAO may have a neuroprotective role in the brain by protecting against D-amino acid aggregation by modulating their degradation. Furthermore, DAO may also impact dopamine synthesis. ALS-related DAO mutations like R199W may be associated with elevated D-serine levels. The D-serine could accumulate in the spinal cord, leading to motor neuron degradation [170,171].

Copy Number Variants (CNVs) in ALS
Besides the repeat expansions in c9orf72, ATXN2, and NIPA1, the copy number variants (CNVs) are associated with the gain (multiplication) or loss (deletion) of a certain sequence into the genome, which could result in abnormal protein expression. The CNVs were suggested to impact different disorders, including CMT, autism, schizophrenia, and epilepsy, but they could also play a role in ALS [172]. For example, the survival motor neuron (SMNs, SMN1 and SMN2). SMN proteins may impact several cellular functions, including RNA splicing, recycling, and axonal transport. SMN gene mutations could impact motor diseases, including spinal muscular atrophy (SMA). SMN1 duplications were suggested to play a role in sporadic ALS, but they failed to find an association between SMN2 and ALS.
However, further studies are needed on the mechanism of SMN1 duplication in ALS onset [173,174]. A study by Moisse et al. (2021) failed to find an association between ALS risk and SMN1 or SMN2. Also, this study suggested that SMN genes may not impact the ALS disease course or disease severity [175]. Blauw et al. (2008) investigated studies on CNVs, which may be a risk for ALS, but they did not find any common CNVs in ALS patients. However, they found gene deletions (including Crystallin Beta B3-CRYBB3, Crystallin Beta B2-CRYBB2, LDL Receptor Related Protein 5 Like-LRP5L, Crystallin Beta B2 Pseudogene 1-CRYBB2P1), which were more common in ALS patients, compared to controls, suggesting CNVs as possibly ALS susceptibility factors [176]. An additional study by the same group performed a large genome study on CNVs and suggested that dipeptidyl-peptidase 6 (DPP6) and NIPA1 CNVs may be possibly promising candidates for ALS association [177].

Potential ALS Biomarkers
ALS is a complex disease that shares similar phenotypes with other diseases, such as FTD or PD. To limit misdiagnosis, a battery of comprehensive tests, including brain-and spine imaging, biofluid analysis, and neuropsychological tests, are routinely employed [74]. Current ALS diagnosis utilizes the simplified Gold Coast criteria (evolved from the revised El Escorial and the Awaji criteria), which employs three stages. The first stage involves progressive motor impairments documented from repeated clinical assessments preceded by normal motor function. The second stage involves the identification of upper (UMN) and lower motor neuron (LMN) dysfunctions in at least one of the four body regions (bulbar, cervical, thoracic, and lumbar) or LMN dysfunction in at least two regions via electromyography (EMG). The final stage is finding evidence to exclude other disorders. Electromyography (EMG) remains an unrivalled diagnostic tool, but it is far from 100% sensitive, resulting in significant diagnostic delays regarding motor neuron disease (MND). As a supplement or alternative to EMG, several molecular and imaging biomarkers can be employed to monitor MND progression and assess the efficacy of potential therapeutic candidates [178,179]. Proteomic markers may be the easiest to apply, as they are compatible with current methods such as traditional mass spectrometry, immunoassays, or aptamer technologies. Moreover, protein profiling is compatible with various biofluids such as plasma, serum or cerebrospinal fluid (CSF) [179,180].
Given the diversity of pathological mechanisms related to ALS (excitotoxicity, oxidative stress, neuroinflammation, metabolism dysfunctions) [178], a biomarker panel must rule out the possibility of other diseases that closely mimic ALS (spinal and bulbar muscular atrophy (SBMA) and multifocal motor neuropathy (MMN)), allow the disease risk prediction, identify patients genetically suited for trials, reduce the time of diagnosis, and provide the earliest opportunity to start therapy [181].
A novel set of ALS-specific biomarkers might include a combination of multiple biomarkers, including neurophysiological and neuroimaging and markers from biological fluids and tissues, which could effectively diagnose and monitor the disease progression [182]. Altered gene expression analysis after muscle biopsy may help distinguish the earlier and later forms of disease pathology. However, a biopsy is an invasive method that is not always required for disease diagnosis [183], whereas fluid biomarkers (neurofilaments (NFs) sourced from Table 3, CSF, serum, or blood ) associated with ALS-related changes in the brain [182] can provide a less invasive option.   [202] p-Tau/ T-Tau NA Metaanalysis Increased t-Tau and reduced pTau/tTau ratio in ALS may be promising candidates.

Biomarkers of Disease Diagnosis and Differential Diagnosis
Neurofilaments are part of the cytoskeleton; their level could be higher in biological fluids in case of axonal damage. NFs have three subunits: light (NFLs), medium (NFMs), and heavy (NFHs) chains. It was suggested that their fluid levels potentially reflect the degree of axonal injury in the case of neurodegenerative diseases [183]. Multiple studies revealed that NFs in blood or CSF were elevated in different neurodegenerative diseases, such as ALS, parkinsonism, AD, or multiple sclerosis [184][185][186][187]. NFLs have been extensively studied in ALS. For example, Verde et al. (2019) revealed that serum NFLs in ALS patients were higher than in controls and non-ALS patients (AD, FTD, or PD), aiding differential diagnosis. A recent study (n = 429) further highlighted the potential of serum NFL to distinguish ALS from other diseases, implying that the routine examination of serum NFL can improve survival prediction and diagnostic accuracy [188]. In addition, Sun et al. (2020) measured NFL in CSF, and sporadic ALS patients revealed higher levels of NFLs compared to controls. Unfortunately, this study failed to find significant differences in CSF NFL levels between ALS and other CNS or peripheral neuropathy diseases. These findings suggested that CSF-NFL levels cannot significantly distinguish ALS from other diseases. However, a negative correlation was found between ALS function scores and CSF-NFL levels, implying CSF-NFL might be a helpful marker in monitoring disease severity and progression [184,189]. Besides NFL, phosphorylated NFH (pNFH), present in (CSF) and serum, offers similar diagnostic utility comparable to NFL. In a recent study, including ALS patients (n = 234), ALS mimics (n = 44), as well as non-neurological disorder controls (n = 9), SF NFL, CSF pNFH, and plasma NFL levels were significantly increased in ALS patients, with authors concluding pNFH had diagnostic value capable of differentiating ALS from clinically relevant mimics [190]. Higher levels of pNFH in serum and CSF may also correlate with the degree of motor neuron impairment [191]. In addition, NFHs in plasma may also reflect the degree and pace of neurodegeneration, disease progression, and survival [192]. Behzadi et al. (2021) suggested that neurofilaments may be higher in ALS patients than in patients with other motor neuron diseases (including spinobulbar muscular atrophy, progressive muscular atrophy), myopathies, or neuropathies. Furthermore, neurofilaments may be able to distinguish the different ALS subgroups (spinal-bulbar or truncal onset). Bulbar-onset ALS patients presented higher levels of plasma NFLs than spinal-onset patients, but there was no difference in patients with truncal-onset. No difference was seen in the CSF-NFL levels among the different subtypes [190].
Multiple studies have confirmed that inflammatory cytokines and T-cell-related immune response are elevated in body fluids of ALS patients, compared to unaffected age-and gender-matched controls [193][194][195][196], suggesting the disease might have a unique inflammatory pattern. Cytokines in plasma may also be useful in ALS diagnosis. For example, Tortelli et al. (2020) analyzed five inflammatory markers (interleukins, such as IL-2, IL-6, IL-10, interferon; IFN-gamma, and tumor necrosis factor; TNF-alpha) in plasma; all were higher in patients compared to the unaffected controls, with IL6 being the most potent marker [197]. Whether inflammation is causal, exacerbates disease pathogenesis, or is the consequence of neurodegeneration is debatable. Also, further research is needed on whether the inflammatory response could be damaging or protective [193,197]. In 2017, a serum analysis of ALS patients showed elevations for malondialdehyde (MDA), 8-hydroxy-2 -deoxyguanosine (8-OHdG), the ratio of glutathione disulfide (GSSG)/glutathione (GSH), and IL6 and IL8, indicating a reduction in the total antioxidant status (TAS) of ALS patients [198]. Additional serum analysis involving 106 serum cytokines, growth factors, and markers for blood-brain barrier breakdown, sourced from 15 patients, was conducted by Cao et al. (2022). No significant differences were detected, with authors attributing the findings to small sample size; however, 20 markers involved in angiogenesis and growth were reportedly enriched (such as fractalkine, Brain-derived neurotrophic factor BDNF, Epidermal Growth Factor EGF, Platelet-derived growth factor PDGF, Dickkopf WNT signaling pathway inhibitor 1, macrophage migration inhibitory factor, angiopoietin-2, or S100 calcium-binding protein B) and segregated with ALS [199].

Biomarkers of ALS Prognosis and Disease Progression
Early studies showed CSF-Tau could be a reliable marker for several neurogenerative diseases, including ALS [200,201]. However, its role regarding ALS may be limited [200], as Scarafino et al. (2018) [201] and Kojima et al. (2021) [202] highlighted that its inclusion was best served as a disease prognosticator. More recently, Thapa et al. (2023), via a meta-analysis of ALS studies, demonstrated that the t-Tau to pTau/tTau ratio was a better indicator of ALS in patients than either alone [203]. In addition, Sun et al. (2020) monitored eight prognostic markers (including serum creatine, albumin, C-Reactive Protein CRP, and glucose, calcium, and potassium) in the blood of 399 ALS patients. The study highlighted an association between disease biomarkers and the possible risk of higher mortality. In conclusion, the authors suggested further studies were needed regarding the potential application of these markers in ALS prognosis [204]. Other studies have shown that ALS patients with c9orf72 repeat expansion tend to have higher pNFH levels in CSF, which can exhibit quicker disease progression and shorter survival times than those without c9orf72 expansion. Such findings suggest that pNFH may be helpful in the prognosis of disease progression with or without intervention in patients with c9orf72 expansion [205]. Gaiani et al. (2017) investigated a study on NFL levels in the blood and CSF of ALS patients. This study revealed that NFLs can potentially support disease diagnosis and prognosis in ALS patients. High early-stage NFL levels may be associated with poorer disease outcomes [206]. Vacchiano et al. (2021) measured NFL in plasma and CSF using the SIMOA platform to differentiate ALS patients from those who mimic ALS. NFL concentrations were higher in plasma and CSF than those with ALS mimic. ALS patents' quick disease progression and shorter survival showed higher NFL levels in both plasma and CSF [189].
TDP43 is a neurodegenerative marker frequently used to confirm AD and FTD [207,208] and, more recently, as a potential candidate for ALS progression. For example, Beyer et al. (2021) revealed that the beta-sheets of TDP43 sourced from the CSF of ALS patients were significantly more enriched compared to TDP43 samples taken from PD patients and controls. In support of those observations, faster progressive ALS patients have also exhibited higher beta-sheet enrichment than slower progressive patients [209].
Other inflammatory markers in a potential ALS panel are the metalloproteases (MMPs) MMP-2 and MMP-9, and their tissue inhibitors (TIMPs), essential in regulating enzyme activity. TIMPs, specifically TIMP-1, could inhibit the proteinase activity of several MMPs, impacting ALS progression [210][211][212]. For example, early studies associated higher MMP9 levels with peripheral nerve and muscle damage, a common feature in the disease's initial stage [213]. A membrane type-matrix metalloproteinase (MT-MMP1) and MMP9 in serum could be potential markers in distinguishing ALS patients from asymptomatic individuals [221].
In another study, using larger sample (CSF and serum) sizes (n = 105), researchers evaluated the potential of secreted macrophage and epithelial proteins, Chitotriosidase-Cells 2023, 12, 1948 20 of 39 1 (CHIT1), chitinase-3-like protein 1 (YKL-40), and monocyte chemoattractant protein-1 (MCP-1) as ALS prognostic biomarkers. CSF levels of these markers were elevated in ALS patients compared to the controls; however, they may not be effective as discriminatory markers between ALS and ALS mimics. A correlation was observed between CHIT1, YKL-40, and disease progression rate. The CSF CHIT1 levels were higher in the ALS patients, who presented more regions with motor neuron degeneration. The YKL-40 and MCP-1 in CSF could predict the disease survival of ALS patients. Furthermore, CHIT1 and YKL40 were weakly correlated with disease progression rate [210]. Other chitinases, such as chitinase-3-like-2 (CHI3L2) or chitinase-3-like-1 (CHI3L1), were also elevated in the CSF of ALS patients, compared to controls. However, they may not help distinguish ALS from primary lateral or multiple sclerosis. Chitinase levels in ALS patients may correlate with molecules involved in neuroinflammation, such as MCP-1, C-reactive proteins or soluble TREM2. Also, CHIT1 expression may correlate with interleukin expression (IL18 or IL16), and CHI3L1 and CHI3L2 may correlate with complement molecules, such as C1 components. These findings suggested that chitinases may be markers for inflammatory processes in ALS [219,220,223]. Perivascular fibroblast proteins, such as secreted phosphoprotein 1 (SPP1) and Collagen Type VI Alpha 1 Chain (COL6A1), were analyzed in the plasma of European (Swedish, Dutch, German, and Belgian) ALS patients and controls. High SPP1 and COL6A1 were associated with shorter survival [222].

Biomarkers of ALS Onset Prediction
Predicting the impairment of regulatory processes and the formation of abnormal DPR proteins [224] due to enlarged C9orf72 repeat expansions may be one of the risk factors for ALS. Thus, elevated c9orf72 poly-GP protein levels in clinical (CSF and peripheral blood mononuclear) samples could be prognostic for c9orf72 carriers. Elevated levels of poly-GR and poly-GPs can appear in patients and asymptomatic carriers, and are reduced in murine models following exposure to antisense oligonucleotides (ASO) [215], suggesting the protein has potential in assessing the rescue effects of therapeutic candidates. In addition to poly-GP, elevated polyGR, polyGA, and SOD1 have exhibited [216] similar potential following ASO treatments [217,218]. Abnormal TDP43 splicing (with cryptic exon inclusion) is thought to occur before disease onset, suggesting that alternatively spliced TDP43 can potentially predict ALS onset [214]. NFL may also be useful in identifying pre-symptomatic individuals. Benatar et al. (2020) revealed that serum and CSF NfL levels of asymptomatic carriers of fALS mutations were higher than the controls but lower than the ALS patients. Elevation of NFL levels may predict ALS onset approximately a year before the clinical symptoms appear [225]. The CHIT1 elevation appears to be a feature of the late pre-symptomatic to early symptomatic phases in patients carrying mutations in C9orf72 or SOD1 [219].
The potential outlook regarding the clinical application of miRNA signatures for SALS is somewhat limited, given the lack of multi-center studies and serum miRNAs.
In an attempt to address this issue, a recent study examined miR-181 levels in 252 ALS patients and found an association between higher levels in the plasma and a higher risk of mortality. Moreover, when miR-181 was paired with NFLs, the combination offered superior prognostication capacity for ALS, leading to improved patient stratification [235].

Plasma small and large EVs
Deregulated miRNAs in EVs of ALS patients (such as hsa-miR-206, hsa-miR-205-5p, miR-1-3p, hsa-miR-205-5p), EVs may be helpful in differential diagnosis [238] The contents of EVs can potentially be employed to diagnose different stages of ALS and other neurodegenerative diseases. The vehicles are involved in the transport and exchange of RNAs, lipids, and proteins between recipient cells. Since they can cross the BBB, they may accurately reflect the disease-specific changes inside the brain. Also, they may be sufficiently present in peripheral areas, making them easy to obtain from several sources, including plasma, whole blood, CSF, or saliva. Alterations in the protein/RNA contents were suggested as promising markers that may reflect the disease-related changes [237][238][239][240][241].
Elevated TDP43 and hyperphosphorylated TDP43 levels have been reported in plasma EVs [242,243]. TDP43 in EVs may also be a valuable biomarker to monitor ALS progression since their contents can vary with disease progression [243]. Pasetto et al. revealed reduced levels of Hsp90 in plasma EVs compared to controls [244]. Immune molecules, such as IL6 (systemic pro-inflammatory), were also elevated in plasma EVs, which may reflect the presence of neuroinflammation in the brain of ALS patients [246].

Discussion and Future Promise of ALS Therapy
ALS is a motor neuron disease which could affect the neurons responsible for muscle movement control. The phenotypes of ALS may overlap with other disorders, including FTD or PD, which may make the diagnosis more difficult [247]. ALS is a genetically complex disease since multiple genes were suggested as causative or risk factors. Besides SOD1, c9orf72, TARDBP, and FUS, many other rare disease-causing genes were discovered, which may impact either familial or sporadic forms of ALS. With the availability of next-generation sequencing technologies, the cost and time of sequencing has been reduced. With novel sequencing technologies, more and more novel genes were discovered, not only in the coding but also in the non-coding regions. The discovery of new genetic variants revealed that multiple pathways might be involved in ALS progression, including regulation of gene expression, protein trafficking, or autophagy [10,18]. Recently, the discovery of ALS-related genetic factors has been accelerated, and in the future, more possible candidates may be discovered [248,249]. Understanding the involvement of different genetic factors in ALS progression is an excellent challenge in disease-related research. Determining the possible disease-causing, risk factor genes or other biomolecules is becoming increasingly important in disease diagnosis or therapy [249,250]. In the last few years, multiple biomarkers and possible candidates for ALS were identified in biological fluids, such as CSF or blood. The discovery of biomarkers should be essential in disease diagnosis, monitoring its progression, and screening the efficacy of drug candidates. Research on potential biomarker candidates is ongoing, and several promising proteomic markers were discovered, including NFLs, Tau protein, CHIT1, and other inflammatory markers [250].
ALS is clinically recognized as a heterogenic (upper or lower motor neurons, or both affected, bulbar or limb onset) disease with a complex molecular background. As mentioned previously, novel genetic factors continue to be identified, whose functional role in ALS involving endosome composition, autophagy, or cytoskeletal system is yet to be determined. In addition, several biomarkers (proteomic markers or miRNAs) in the biological fluids could play a role in ALS diagnosis, prognosis, or prediction (either the clinical course or the disease onset). Moreover, these markers' timely presence may indicate a patient's responses to therapies [253,254]. As such several novel methods (such as combined tissue-fluid proteomics) were developed to discover possible biomarker candidates, some with applications in phenotypic stratification [255]. ALS stratification is essential for optimized personal clinical care and treatment. For example, phenotypic studies involving relatively small cohorts suggest neurofilament markers could be used in ALS patient stratification [256]. With the aid of machine learning and a novel scheme, "The Stratification Challenge" (DREAM Prize4Life ALS), Kueffner et al. (2019) successfully performed clustered analysis and the classification of ALS patients and ALS subgroups [257].
Similar to most neurodegenerative diseases, no successful therapy is currently available for ALS. However, promising strategies are available through gene therapy. In other motor neuron diseases, spinal muscular atrophy (SMA), and several gene therapeutic approaches, such as Nusinersen, Zolgensma, or Risdiplam, have been approved, which could lead to improved life quality of patients. The gene therapy strategy may also be promising for ALS drug development [258,259]. Antisense molecule delivery, which targets genetic mutations in SOD1, FUS, TARDBP, or c9orf72, was suggested to be helpful in disease therapy. The gene therapy methods may be helpful in patients positive for ALS-causing mutations. For example, ALS patients who received ASO-Tofersen were associated with reduced SOD1 and NFL levels in CSF and plasma of ALS patients, respectively. However, the drug did not improve the clinical symptoms of patients [260]. Tofersen or BIIB067 drug may be helpful to SOD1 carriers in pre-symptomatic ALS individuals with higher levels of plasma NfL. The drug may delay the disease's progression and is currently in phase III clinical trials [261]. Another potentially available ASO for c9orf72 repeat expansion is C9orf72-631 or BIIB078. BIIB078 may protect the neurons from gain-of-function toxicity of repeat expansion and may prevent glutamate-related toxicity, and is in phase I of clinical trials [262][263][264]. For FUS mutations, an ASO-related drug called Jacifusen (NCT04768972 or ION363) is also in phase III of clinical trials [265].
Research is ongoing on viral-based strategies of gene therapy as well. For example, adeno-associated virus (AAV)-related gene therapies may also be promising approaches in ALS mutants. AAV-related gene silencing in SOD1 mutant mice showed reduced SOD1 expression and improved motor skills [266]. AAV-miR-SOD1 was tested in two ALS patients with SOD1 mutants. A slight improvement was observed in one of the patients in leg strength, while the other patient's condition remained stable. These findings suggested that AAV-related gene silencing may slow down the disease progression. However, AAV therapy may need immunosuppression to avoid side effects, such as meningoradiculitis [267,268]. AAV-related gene silencing was also attempted to target c9orf72 repeat expansion in ALS mouse models (AAV5-miC). These mice showed reduced RNA aggregation with repeat expansion in nerve cells, leading to reduced neurotoxicity [269]. The AAV5-related gene silencing was also screened in human cell lines, such as HEK293T-and iPSC-derived neurons, and it showed reduced levels of mutant c9orf72 mRNA in both plasma and cytoplasm of the cells. In addition, RNA foci were reduced in the nucleus [270]. Studies are going on to "reverse the mutation" by genome editing (such as the CRISPR-Cas9 method) for SOD1, C9orf72, FUS, ATXN2, and TARDP mutations. These studies showed promising data in iPSC cell lines derived from ALS patients, such as increased survival and reduced mutant protein-related toxicity [268].
Treatment of neurodegenerative diseases may be more effective in the early disease stage or even in the pre-symptomatic phase. Studies are needed regarding pre-symptomatic ALS to identify the individuals who may develop disease phenotypes. Genetic analysis may effectively identify at-risk individuals; however, the known genetic factors (such as SOD1, c9orf72) may not define all ALS cases. Several biomarkers, such as elevated NFL in the serum and CSF, may be present years before the disease symptoms appear. As well as the lack of definitive markers or ratios, pre-symptomatic ALS diagnoses face many other challenges. One of the most significant is a patient's psychological readiness before presymptomatic testing, comprehension of a positive result, their legal protections regarding information disclosure, and potential discrimination based on genetic results, particularly if a mutation or risk factor is present; counseling should be tailored. The other issue with pre-ALS testing is the logistical complexity. For example, potential pre-ALS individuals may live remotely and are inexperienced with telemedicine. Moreover, if trained personnel are unavailable, monitoring the initial stages of the disease may go undetected, particularly given the overlap of FTD with ALS, leading to further complications [271][272][273][274].
Even though studies have yielded promising results, research on gene therapies may have limitations. One difference between SMA and ALS is that most SMA cases (type I and II) occur in early childhood, but most ALS cases are adult onset. Therefore, the gene therapies for SMA may be given before the second year of age, but the treatment may not be efficient later, especially after the clinical symptoms appear [240]. The other issue with therapy is that ALS is a genetically complex disease; gene therapies may not be available yet for all ALS-related genetic factors. Nevertheless, the gene therapies may work in the pre-symptomatic disease stage, suggesting that obtaining genetic and biomarker data may be essential to identify the "at risk" individuals. Discovering the early markers in pre-symptomatic individuals may also improve ALS drug development [261,271].

Conclusions
ALS is a complex disease, since many diseases associated genes were identified, including SOD1, c9orf72, TARDBP, or FUS. Emerging studies are also discovering new genetic factors, which could cause disease or contribute to disease progression. Also, several proteomic (including NFs, Tau, TDP43, inflammatory) and microRNAs were discovered, which could play a role in disease diagnosis, prognosis, and prediction. Diagnosing ALS in early or even pre-clinical stages should be important. Currently, there are no therapies available for ALS. However, gene therapies were successfully adapted against other motor diseases, such as SMA. Also, ASO-based gene therapies (such as Tofersen or Jacifusen) are currently in clinical trial phases, but AAV-based therapies are also under development.
Finding the possible genetic and proteomic markers for ALS diagnosis, prognosis, and prediction should be essential for gene therapies against the disease.

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