Alternative Splicing Role in New Therapies of Spinal Muscular Atrophy

It has been estimated that 80% of the pre-mRNA undergoes alternative splicing, which exponentially increases the flow of biological information in cellular processes and can be an attractive therapeutic target. It is a crucial mechanism to increase genetic diversity. Disturbed alternative splicing is observed in many disorders, including neuromuscular diseases and carcinomas. Spinal Muscular Atrophy (SMA) is an autosomal recessive neurodegenerative disease. Homozygous deletion in 5q13 (the region coding for the motor neuron survival gene (SMN1)) is responsible for 95% of SMA cases. The nearly identical SMN2 gene does not compensate for SMN loss caused by SMN1 gene mutation due to different splicing of exon 7. A pathologically low level of survival motor neuron protein (SMN) causes degeneration of the anterior horn cells in the spinal cord with associated destruction of α-motor cells and manifested by muscle weakness and loss. Understanding the regulation of the SMN2 pre-mRNA splicing process has allowed for innovative treatment and the introduction of new medicines for SMA. After describing the concept of splicing modulation, this review will cover the progress achieved in this field, by highlighting the breakthrough accomplished recently for the treatment of SMA using the mechanism of alternative splicing.


Introduction
Splicing is an essential part of pre-mRNA maturation in a eukaryotic cell. That process consists of excising noncoding intronic sequences from the initial product of gene transcription and ligating remaining exons before translation to protein [1]. Splicing reaction is controlled by the spliceosome, the macromolecular ribonucleoprotein complex. Determination of the beginning and end of the intron, which is marked by the 5 and 3 splice sites (5 ss and 3 ss), plays a key role in the splicing mechanism. Specific sequences are recognized by the spliceosome, the macromolecular ribonucleoprotein structure that catalyses splicing [2]. An alternatively spliced gene is a source of multiple mRNA isoforms, which increases coding potential of the eukaryotic genome. Alternative splicing is regulated by the cis-acting splicing regulatory elements (SREs) that recruit trans-acting factors in a sequence-unique manner [3]. Trans-acting RNA binding proteins (RBPs) bound to an intronic or exonic splice enhancer (ISE or ESE) stabilize spliceosome formation and lead to exon recognition and retention. Analogically RBPs bound to intronic or exonic splice silencing motifs (ISS or ESS) preclude the formation of the spliceosome and promote exon removing [4]. The most common RBPs are proteins rich in serine/arginine (SR) rests and the heterogeneous ribonucleoprotein (hRNP). Due to tissue-specific RBPs bounding, final products from the same gene primary transcript can be different depending on the tissue. the tissue-specific RBPs bounding, final products from the same gene primary transcript can be different depending on the tissue. The binding of these proteins is a variable cells [37]. According to hypotheses, the SMN protein in the cytoplasm was shown to have an important role in the transport of mRNA through axons and transport of containing b-actin ribonucleoprotein complexes. Another hypothesis states that the SMN protein takes part in synthesis of small nuclear RNA (snRNA) and therefore plays a key role in the formation of a spliceosome that removes introns from pre-mRNA into functional mRNA [38,39]. As motor neurons are sensitive to malfunctioning of the spliceosome directly or indirectly through misspliced mRNAs, any damage to motor neurons results in the development of dysfunctions in proteins essential for neuronal function [21].

Mechanisms of SMN2 Splicing Regulation Targeted by Therapeutics
In human cells, there are two nearly identical genes responsible for SMN protein production. The presence of two SMN genes is attributed to large tandem chromosomal duplication [40]. In the region of this duplication, on the long arm of chromosome 5 (5q13.2) lie four protein-encoding genes: SMN, NAIP, GTF2H2A, and SERF1A. The duplicated genes are identical to their partner gene (SERF1B), differ in a low number of nucleotides (SMN2) or are pseudogenes (ΨGTF2H2B and ΨNAIP∆5) [41]. Both SMN genes consist of 10 exons (1,2A,2B,3-6,6B, 7,8). It is worth noting that Exon 6b is a new discovery and is generated by exonification of the Alu element in intron 6 [42]. Under certain conditions such as starvation, hypoxia, or oxidative stress, transcription of these genes may proceed differently [8,43]. Factors that regulate SMN levels and modify transcription are tissue-specific [44].
The key difference between these genes lies in the splicing of exon 7. The amino acids encoded in exon 7 are responsible for SMN stability as they determine the crucial C-terminus of the protein. In the SMN2 gene due to alternative splicing, exon 7 is more often skipped, resulting in more of the truncated, partially functional and unstable SMN∆7 protein than full-length SMN [45,46]. The primary reason exon 7 is excluded is C-to-T substitution at position 6 of exon 7 (C6U) (Figure 1). Mutation or deletion of the SMN1 gene is a major cause of spinal muscular atrophy, through deficiency of SMN [39]. Restoring exon 7 inclusion has therapeutic benefits proven in mouse models [47].
Genes 2021, 12, 1346 3 of 17 a pivotal role in homeostatic cellular pathways in all cells [37]. According to hypotheses, the SMN protein in the cytoplasm was shown to have an important role in the transport of mRNA through axons and transport of containing b-actin ribonucleoprotein complexes. Another hypothesis states that the SMN protein takes part in synthesis of small nuclear RNA (snRNA) and therefore plays a key role in the formation of a spliceosome that removes introns from pre-mRNA into functional mRNA [38,39]. As motor neurons are sensitive to malfunctioning of the spliceosome directly or indirectly through misspliced mRNAs, any damage to motor neurons results in the development of dysfunctions in proteins essential for neuronal function [21].

Mechanisms of SMN2 Splicing Regulation Targeted by Therapeutics
In human cells, there are two nearly identical genes responsible for SMN protein production. The presence of two SMN genes is attributed to large tandem chromosomal duplication [40]. In the region of this duplication, on the long arm of chromosome 5 (5q13.2) lie four protein-encoding genes: SMN, NAIP, GTF2H2A, and SERF1A. The duplicated genes are identical to their partner gene (SERF1B), differ in a low number of nucleotides (SMN2) or are pseudogenes (ΨGTF2H2B and ΨNAIPΔ5) [41]. Both SMN genes consist of 10 exons (1,2A,2B, 3-6,6B,7,8). It is worth noting that Exon 6b is a new discovery and is generated by exonification of the Alu element in intron 6 [42]. Under certain conditions such as starvation, hypoxia, or oxidative stress, transcription of these genes may proceed differently [8,43]. Factors that regulate SMN levels and modify transcription are tissuespecific [44].
The key difference between these genes lies in the splicing of exon 7. The amino acids encoded in exon 7 are responsible for SMN stability as they determine the crucial C-terminus of the protein. In the SMN2 gene due to alternative splicing, exon 7 is more often skipped, resulting in more of the truncated, partially functional and unstable SMN 7 protein than full-length SMN [45,46]. The primary reason exon 7 is excluded is C-to-T substitution at position 6 of exon 7 (C6U) (Figure 1). Mutation or deletion of the SMN1 gene is a major cause of spinal muscular atrophy, through deficiency of SMN [39]. Restoring exon 7 inclusion has therapeutic benefits proven in mouse models [47]. Mechanisms regulating exon 7 splicing are good potential therapeutic targets. The best described to date splicing factors, binding directly to exon 7 splicing enhancer regions SE1 and SE2 are serine/arginine-rich splicing factor 1 (SRSF1) and transformer 2 protein Mechanisms regulating exon 7 splicing are good potential therapeutic targets. The best described to date splicing factors, binding directly to exon 7 splicing enhancer regions SE1 and SE2 are serine/arginine-rich splicing factor 1 (SRSF1) and transformer 2 protein homolog β (Tra2B) [48]. The best-known negative regulators of exon 7 splicing are heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) [49] and src-associated substrate in mitosis 68 (Sam68) [50]. C6U substitution results in hnRNA A1 or Sam68 being bound to Genes 2021, 12, 1346 4 of 17 SE1 in place of the positive regulator. Exon 7 exclusion can also occur through binding of hnRNP A1 to SE2 sequences or the intronic silencer sequence N1 (ISS-N1) [51] (Figure 2). Other factors showing altering splicing activity include SRp30c [52], TDP-43 [53], TIA1 [54], hnRNP Q [55], hnRNP G [56].
homolog β (Tra2B) [48]. The best-known negative regulators of exon 7 splicing are heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) [49] and src-associated substrate in mitosis 68 (Sam68) [50]. C6U substitution results in hnRNA A1 or Sam68 being bound to SE1 in place of the positive regulator. Exon 7 exclusion can also occur through binding of hnRNP A1 to SE2 sequences or the intronic silencer sequence N1 (ISS-N1) [51] (Figure 2). Other factors showing altering splicing activity include SRp30c [52], TDP-43 [53], TIA1 [54], hnRNP Q [55], hnRNP G [56]. Exon 7 skipping may be caused by increased activity of a regulatory sequence located at the 3′ end of exon 7, called terminal stem loop 2 (TSL2). It exhibits inhibitory activity, probably by competing with U1 snRNP for a binding site [57]. The inhibitory effect of TSL2 was confirmed by observing the effect of U40G or A54C substitution on exon 7 splicing. Separately, they disrupted TSL2 by promoting exon 7 incorporation, but combined they reproduced the structure of TSL2 and thus inhibition of exon 7 splicing [58]. Modification of splicing through TSL2 requires further study. It can be a good target for screening small molecules [59].
In 2006, Singh et al. discovered an ISS SMN2 intron-7 in the human SMN1/2 gene, named ISS-N1 [60]. ISS-N1 is a sequence located immediately downstream of the 5′ss of exon 7. It is 15-nt long and binds positions 10 to 24 of intron 7, producing a strongly inhibitory effect on exon 7 inclusion [61]. Interestingly, ISS-N1 deletion reduced the requirement for positive cis-elements in exon 7 inclusions, an effect similar to the A54G mutation [51]. Blocking ISS-N1 with low concentrations of antisense oligonucleotide (ASO) effectively increased SMN protein levels in studies in mouse models or fibroblasts collected from SMA patients. This demonstrates the high availability and binding efficiency of ISS-N1 for this group of compounds [47]. Importantly, modifying alternative splicing of SMN2 by targeting ISS-N1 is already used by the ASO drug nusinersen (Spinraza) in the first approved therapy to treat SMA [62].
A more specific target for ASO than ISS-N1 appeared to be the GC-rich sequence (GCRS), which spans from the 7th to 14th position of intron 7 overlapping ISS-N1. Studies on SMA type I patient cells and severe SMA mouse models demonstrated the efficacy of 8-mer ASO binding to GCRS, which not only elevated SMN levels but also increased the levels of Gemin 2 and Gemin 8 factors involved in snRNP biogenesis and Tra2-β1 and hnRNP Q, responsible for proper RNA splicing [63,64].
GCRS participates in the formation of a 5′ strand of a unique RNA structure called the internal stem formed by long-distance interactions (ISTL-1). 279-nts divides the two 8bp ISTL1 strands. The last position of ISTL-1 and the first position of ISS-N1 is the C resi- Exon 7 skipping may be caused by increased activity of a regulatory sequence located at the 3 end of exon 7, called terminal stem loop 2 (TSL2). It exhibits inhibitory activity, probably by competing with U1 snRNP for a binding site [57]. The inhibitory effect of TSL2 was confirmed by observing the effect of U40G or A54C substitution on exon 7 splicing. Separately, they disrupted TSL2 by promoting exon 7 incorporation, but combined they reproduced the structure of TSL2 and thus inhibition of exon 7 splicing [58]. Modification of splicing through TSL2 requires further study. It can be a good target for screening small molecules [59].
In 2006, Singh et al. discovered an ISS SMN2 intron-7 in the human SMN1/2 gene, named ISS-N1 [60]. ISS-N1 is a sequence located immediately downstream of the 5 ss of exon 7. It is 15-nt long and binds positions 10 to 24 of intron 7, producing a strongly inhibitory effect on exon 7 inclusion [61]. Interestingly, ISS-N1 deletion reduced the requirement for positive cis-elements in exon 7 inclusions, an effect similar to the A54G mutation [51]. Blocking ISS-N1 with low concentrations of antisense oligonucleotide (ASO) effectively increased SMN protein levels in studies in mouse models or fibroblasts collected from SMA patients. This demonstrates the high availability and binding efficiency of ISS-N1 for this group of compounds [47]. Importantly, modifying alternative splicing of SMN2 by targeting ISS-N1 is already used by the ASO drug nusinersen (Spinraza) in the first approved therapy to treat SMA [62].
A more specific target for ASO than ISS-N1 appeared to be the GC-rich sequence (GCRS), which spans from the 7th to 14th position of intron 7 overlapping ISS-N1. Studies on SMA type I patient cells and severe SMA mouse models demonstrated the efficacy of 8-mer ASO binding to GCRS, which not only elevated SMN levels but also increased the levels of Gemin 2 and Gemin 8 factors involved in snRNP biogenesis and Tra2-β1 and hnRNP Q, responsible for proper RNA splicing [63,64].
GCRS participates in the formation of a 5 strand of a unique RNA structure called the internal stem formed by long-distance interactions (ISTL-1). 279-nts divides the two 8-bp ISTL1 strands. The last position of ISTL-1 and the first position of ISS-N1 is the C residue located at the 10th intronic position (10C) [65]. Interestingly, an experiment was conducted by targeting ISS-N1 with ASOs of equal length (14-mer). F-14 binding to the first nucleotides of ISS-N1, including 10C, promoted exon 7 inclusion, while L-14 targeting the terminal positions of ISS-N1 without 10C had the opposite effect, promoting exon 7 exclusion [66]. ISTL-1 has been shown to negatively regulate exon 7 splicing independently of snRNP A1. Destabilization of ISTL-1 induces abolition of long-distance interaction (LDI) mediated by C10. ASO-mediated sequestration of the 3 strand of ISTL-1 and upstream sequences form ISS-N2, which results in correction of alternative splicing and restoration of full-length SMN secretion [66,67].
Sequences flanking exon 7 such as Element 1 located in intron 6 also appear to be promising therapeutic targets. Element 1 is a cis-element that is an extended inhibitory sequence located upstream of the 3 ss exon 7 [68]. The effect of promoting full-length SMN expression was demonstrated by targeting Element 1 with morpholino ASOs in mouse models [69].

Antisense Oligonucleotides
A novel approach to the therapy of SMA and other genetically determined diseases is represented by the use of ASO [70]. ASOs are short (about 15-30 nucleotides in length), single-stranded molecules of chemically modified nucleic acids or nucleotide analogs that, on the basis of complementarity, recognize and bind target sequences in RNA through Watson-Crick base pairing [71,72]. Depending on the binding site, ASOs affect transcript inactivation or splicing, leading to changes in exon content [73]. ASOs are designed to pair bases and form a steric block for binding splicing factors to pre-mRNAs. RNA alters the recognition of splicing sites by the spliceosome, leading to a change in the normal splicing of the target transcript [72]. Modified sequence-dependent ASOs can appropriately lead to the exclusion or inclusion of an exon that would have been excised, as is the case in SMA [72].
This relies on exon 7 appearing in the mature SMN2 transcript. It is necessary to block the action of the intron folding silencer. Antisense oligonucleotides recognize, on the basis of nitrogenous base complementarity, precisely this SMN2 pre-mRNA fragment and sterically block its recognition by appropriate proteins. This prevents the formation of a complex that would inhibit detection of the exon/intron boundary. Subsequently, the split between exon 7 and intron 7 is detected, resulting in the incorporation of exon 7 into the mature transcript [74]. The use of appropriate ASOs to treat spinal muscular atrophy allows exon 7 to be incorporated into the transcript of the SMN2 gene.
The first drug approved for the treatment of SMA was nusinersen (Spinraza TM) [75]. Its discovery took place in 2010 [76]. It is an antisense oligonucleotide that binds to the splicing inhibitory sequence of intron 7. Nusinersen is an 18-mer oligonucleotide in which the sugar-phosphate backbone has been chemically modified [77]. Nusinersen complementary hybridizes to ISS-N1 to block hnRNP recruitment, resulting in the inclusion of exon 7 incorporation into the SMN2 transcript, resulting in higher levels of fully functional SMN protein [78] (Figure 3). This protein is associated with SMA. As the amount of SMN protein increases, the degeneration of motor neurons stops and the disease progresses [78]. This influences a patient's longer survival, better motor function and faster achievement of milestones. Nusinersen does not cross the blood-brain barrier and therefore requires intrathecal administration. The half-life of the drug is 163 days, and doses must be repeated throughout life [79]. Monitoring in patients of thrombocyte count, prothrombin time, partial thromboplastin time, and urinalysis results is necessary during nusinersen therapy because it can lead to thrombocytopenia and coagulation disorders, and is nephrotoxic [80]. It can be used to treat all types of SMA [81]. Another breakthrough in the treatment of SMA was the December 2016 approval by the American Food and Drug Administration of the medicine nusinersen (SpinrazaTM), also known as ISIS-SMNRx or ISIS [82]. If a patient does not reach an advanced stage of muscle atrophy, appropriate physiotherapy and multidisciplinary care in combination with nusinersen can produce a significant improvement in the condition of the treated patient [83,84] (Table 1).  [82]. If a patient does not reach an advanced stage of muscle atrophy, appropriate physiotherapy and multidisciplinary care in combination with nusinersen can produce a significant improvement in the condition of the treated patient [83,84] (Table 1).

Small Molecules
A project led by PTC-Roche (PTC Therapeutics, South Plainfield, New Jersey and Hoffmann-La Roche, Basel, Switzerland) to identify an orally available molecule to treat SMA began about a decade ago. Both groups identified small molecules and reported three orally delivered compounds, namely SMN-C1 (isocoumarin), SMN-C2 (coumarin), and SMN-C3 (pyridopyrimidinone derivative); each promoted exon 7 inclusion from SMN2 [92]. Small molecules can exhibit high selectivity, affecting the modulation of RNA folding of only one or a few genes, among the many thousands of genes expressed in cells [93]. Most drugs are inhibitors of enzyme proteins or receptors. It is worth noting that it is possible to obtain modulators of interactions in RNA-RNA and RNA-protein complexes [94]. Risdiplam is being developed by Roche, PTC Therapeutics Inc and the SMA Foundation for the treatment of SMA. In August 2020, the European Medicines Agency (EMA) approved the use of risdiplam to treat patients with the SMN1 gene mutation [95]. This experimental drug manifests high selectivity for modulation of RNA folding against the SMN2 transcript. It affects the alternative splicing of a small pool of other genes, such as FOXM1, MADD or STRN3 [94,95]. Risdiplam is not a substrate for the transport protein MDR1, and thus crosses the blood-brain barrier well. It is properly distributed in the CNS and peripheral tissues of mice, rats, and monkeys after single or repeated oral or intraperitoneal administration. Risdiplam also increased levels of functional SMN protein in the CNS and peripheral tissues of mouse models of SMA [96].
Risdiplam is a highly potent SMN2 splicing modifier that increases exon 7 inclusion in SMN2 mRNA transcripts in in vitro assays and in transgenic mouse models of SMA [95][96][97]. Risdiplam binds to the SMN2 transcript at two sites-the exonic splicing enhancer 2 (ESE2) in exon 7 and 5 ss of intron 7, thereby dislocating hnRNPG and enhancing 5 ss recognition and binding by U1snRNP. This results in exon 7 not being excised from the transcript and the full SMN protein being able to be synthesized [94,98]. Risdiplam can also increase the binding of far upstream element binding protein 1 (FUBP1) and KHtype splicing regulatory protein (KHSRP) splicing modulators to the SMN2 pre-mRNA complex, activating SMN2 splicing [99]. Some of the first preclinical studies have shown that risdiplam can reach the central nervous system and peripheral organs in vivo and can lead to significant increases in SMN protein levels in blood, brain, and muscle, with increased survival in various mouse models of SMA [100,101]. The advantage of this drug is the oral route of administration [102]. Preclinical studies allow for hypothesizing the possibility of a therapeutic effect also in tissues other than the nervous system [102]. This is particularly important because numerous studies in human and animal models indicate that SMA may be considered a multisystem disorder with involvement of the neuromuscular junction, gastrointestinal tract, cardiovascular system, and lung and liver tissues [103,104]. According to the Food and Drug Administration (FDA), on 7 August 2020 risdiplam was approved for the treatment of spinal muscular atrophy in adults and children 2 months of age and older [105]. A recent study analyzing the administration of risdiplam to infants from 1-7 months of age (type 1 SMA) has led to increased expression of functional SMN protein in the blood [100].
Branaplam is another small molecule, administered orally, that modulates SMN2 splicing with high specificity. It is currently in Phase 2 clinical trials [101,106]. It has been shown to modulate splicing, increase full-length SMN protein levels, and increase survival in a mouse model of severe SMA [107]. The mechanism of action is similar to risdiplam.
To the best of our knowledge, two more molecules PK4C9 and TEC-1, according to recent reports, increase exon 7 SMN2 inclusion with high specificity [59,101,107]. TEC-1 permeabilizes the central nervous system and confers therapeutic efficacy in a mouse model of SMA [59,108]. PCK4C9 targeting the TSL2 tri-loop appeared to cover the "3 -cluster," a negative element identified by in vivo selection of the entire exon 7 [109,110] (Table 2). SMN-C2-binding to the AGGAAG SMN2 pre-mRNA exon 7 SMN-C3-hypothetically targets a sequence of RNA on or close to exon 7 or a splicing regulatory protein or protein complex that is specific to exon 7.
Small molecules complementary to nucleic acids modulate pre-mRNA splicing and can have a therapeutic influence on SMA. Future studies should concern recognition sequence of FUBP1 and KHSRP and their contribution in splicing regulation.

Future Prospects
The advancement of SMA therapies has allowed many patients to survive and improve their lives. Current drugs focus on replacing the SMN1 gene (onasemnogene abeparvovec) or changing SMN2 splicing (nusinersen, risdiplam). Work is currently underway on a complementary treatment independent of SMN. This applies, for example, to neuroprotective drugs, nerve connection stabilizers, myostatin inhibitors, or activators of muscle function [106,[119][120][121] (Figure 4). Many studies also emphasize the importance of early diagnosis and treatment implementation, even presymptomatically. Efforts should be made to develop effective neonatal screening for SMA and to update treatment regimens due to the evolving phenotype of the disease [122][123][124]. Currently, the most important modifier of SMA is the SMN2 copy number; however, it has been noticed that patients with the same SMN2 copy number show a difference in the disease phenotype. Work is currently underway to find new biomarkers of disease evolution [125][126][127]. It is also important to provide multidisciplinary care for treated children [128,129]. 12,1346 Work is currently underway to find new biomarkers of disease evolution [125also important to provide multidisciplinary care for treated children [128,129].

Conclusions
The use of the molecular basis of SMA by drugs such as nusinersen, Zolgen risdiplam has brought significant benefits to patients with this fatal disease. Mod of exon 7 alternative splicing turns out to be a key mechanism and target for fu search. Current research proves that this therapeutic strategy can effectively inc level of the SMN protein and, as a result, reduce the course of the disease. Early d and initiation of treatment in the patient allow for the extension of lifespan achievement of milestones. Work is underway on the implementation of oth pounds in SMA therapy that bind to factors involved in the regulation of splicin

Conclusions
The use of the molecular basis of SMA by drugs such as nusinersen, Zolgensma, and risdiplam has brought significant benefits to patients with this fatal disease. Modification of exon 7 alternative splicing turns out to be a key mechanism and target for further research. Current research proves that this therapeutic strategy can effectively increase the level of the SMN protein and, as a result, reduce the course of the disease. Early diagnosis and initiation of treatment in the patient allow for the extension of lifespan and the achievement of milestones. Work is underway on the implementation of other compounds in SMA therapy that bind to factors involved in the regulation of splicing.