The Biochemistry of Survival Motor Neuron Protein Is Paving the Way to Novel Therapies for Spinal Muscle Atrophy.

Spinal muscle atrophy (SMA) is the leading genetic cause of infant mortality. SMA originates from the loss of functional survival motor neuron (SMN) protein. In most SMA cases, the SMN1 gene is deleted. However, in some cases, SMN is mutated, impairing its biological functions. SMN mutants could provide clues about the biological functions of SMN and the specific impact on SMA, potentially leading to the identification of new pathways and thus providing novel treatment alternatives, and even personalized care. Here, we discuss the biochemistry of SMN and the most recent SMA treatment strategies.

SMA-linked SMN mutants prevent phase separation properties and SMN interactions with FMRP family members.

Although recent advances in gene therapy provide hope for spinal muscular atrophy (SMA) patients, the pathology remains the leading genetic cause of infant mortality. SMA is a monogenic pathology that originates from the loss of the SMN1 gene in most cases or mutations in rare cases. Interestingly, several SMN1 mutations occur within the TUDOR methylarginine reader domain of SMN. We hypothesized that in SMN1 mutant cases, SMA may emerge from aberrant protein-protein interactions between SMN and key neuronal factors. Using a BioID proteomic approach, we have identified and validated a number of SMN-interacting proteins, including fragile X mental retardation protein (FMRP) family members (FMRFM). Importantly, SMA-linked SMNTUDOR mutant forms (SMNST) failed to interact with FMRFM In agreement with the recent work, we define biochemically that SMN forms droplets in vitro and these droplets are stabilized by RNA, suggesting that SMN could be involved in the formation of membraneless organelles, such as Cajal nuclear bodies. Finally, we found that SMN and FMRP co-fractionate with polysomes, in an RNA-dependent manner, suggesting a potential role in localized translation in motor neurons.

Xmn I polymorphism associated with concomitant activation of Gγ and Aγ globin gene transcription on a ß0-thalassemia chromosome.

The -158 (C→T) nucleotide change, known as Xmn I polymorphism, occurs in (G)γ-globin gene promoter, and results in elevated fetal hemoglobin (HbF). We found this mutation in cis of a ß(0)-thalassemia splicing mutation. Despite the complete absence of adult HbA, the phenotype was only moderately severe with no detectable alteration of α-globin gene expression. Interestingly, the ß-globin locus haplotype has not been described to bear the (G)γ promoter mutation. Using a gene-specific real-time RT-PCR approach, we found a dramatic increase of both (G)γ and (A)γ mRNA accumulated in the reticulocytes, suggesting that the (G)γ-promoter mutation, alone or in association with another genetic modification, alters in concert the transcription of both (G)γ and (A)γ. This observation is discussed in light of recent regulatory model for ß-globin locus.

Homozygous deletion of EPB41 genuine AUG-containing exons results in mRNA splicing defects, NMD activation and protein 4.1R complete deficiency in hereditary elliptocytosis.

Complete loss of protein 4.1R in red blood cell membrane is a very rare condition in humans. We here explore the third case. The morphological and biochemical observations suggested that the proband suffers from homozygous hereditary elliptocytosis. Both parents, who are consanguineous, have an elliptocytosis with no cell fragmentation, typical of a heterozygous 4.1R deficiency with a silent allele. A genomic deletion was found; it encompasses about 50 kb of genomic DNA, and suppresses the two key exons 2 and 4, which contain the two functional AUG translation initiation sites in erythroid and nonerythroid cells. The alternative first exons are intact, hence preserving the transcription potential of the altered gene. Extensive analysis of 4.1R transcripts revealed multiple splicing defects upstream of the deleted sequences. Importantly, we found that most of the transcripts generated from the altered gene are intercepted by the nonsense-mediated mRNA decay mechanism, suggesting that the massive degradation of the mRNA species jeopardizes the production of shortened but functional protein 4.1R from an alternative translation initiation site downstream of the deletion.

Nonsense-mediated mRNA decay (NMD) blockage promotes nonsense mRNA stabilization in protein 4.1R deficient cells carrying the 4.1R Coimbra variant of hereditary elliptocytosis.

We describe a new approach to stabilize nonsense mRNA, based on the inhibition of the NMD mechanism, by combining cycloheximide-mediated inhibition of translation, and caffeine-mediated inhibition of UPF1 phosphorylation. This approach aimed to identify the impact of a 4.1R splicing mutation. This mutation is involved in a partial deficiency of 4.1R in the homozygous state in a patient with hereditary elliptocytosis and a moderated hemolytic anemia. We show that, in addition to two known minor shortened and stable spliceoforms, the mutation activates an intronic cryptic splice site, which results in a nonsense mRNA major isoform, targeted to degradation in intact cells by NMD. This accounts for the main cause of 4.1R partial deficiency. In a general perspective, blocking the NMD mechanism would help to identify a missing isoform, and pave the path for a molecular targeting strategy to circumvent a deleterious splicing pathway in favor of a therapeutic splicing pathway.

Reinforcement of a minor alternative splicing event in MYO7A due to a missense mutation results in a mild form of retinopathy and deafness.

PURPOSE: Recessive mutations of the myosin VIIA (MYO7A) gene are reported to be responsible for both a deaf-blindness syndrome (Usher type 1B [USH1B] and atypical Usher syndrome) and nonsyndromic hearing loss (HL; Deafness, Neurosensory, Autosomal Recessive 2 [DFNB2]). The existence of DFNB2 is controversial, and often there is no relationship between the type and location of the MYO7A mutations corresponding to the USH1B and DFNB2 phenotype. We investigated the molecular determinant of a mild form of retinopathy in association with a subtle splicing modulation of MYO7A mRNA. METHODS: Affected members underwent detailed audiologic and ocular characterization. DNA samples from family members were genotyped with polymorphic microsatellite markers. Sequencing of MYO7A was performed. Endogenous lymphoid RNA analysis and a splicing minigene assay were used to study the effect of the c.1935G>A mutation. RESULTS: Funduscopy showed mild retinitis pigmentosa in adults with HL. Microsatellite analysis showed linkage to markers in the region on chromosome 11q13.5. Sequencing of MYO7A revealed a mutation in the last nucleotide of exon 16 (c.1935G>A), which corresponds to a substitution of a methionine to an isoleucine residue at amino acid 645 of the myosin VIIA. However, structural prediction of the molecular model of myosin VIIA shows that this amino acid replacement induces only minor structural changes in the immediate environment of the mutation and thus does not alter the overall native structure. We found that, although predominantly included in mature mRNA, exon 16 is in fact alternatively spliced in control cells and that the mutation at the very last position is associated with a switch toward a predominant exclusion of that exon. This observation was further supported using a splicing minigene transfection assay; the c.1935G>A mutation was found to trigger a partial impairment of the adjacent donor splice site, suggesting that the unique change at the last position of the exon is responsible for the enhanced exon exclusion in this family. CONCLUSIONS: This study shows how an exonic mutation that weakens the 5' splice site enhances a minor alternative splicing without abolishing a complete exclusion of the exon and therefore causes a less severe retinitis pigmentosa than the USH1B-associated alleles. It would be interesting to examine a possible correlation between intrafamilial phenotypic variability and the subtle variation in exon 16 inclusion, probably related to genetic background specificities.

Cryptic splicing sites are differentially utilized in vivo.

It has long been considered that cryptic splice sites are ignored by the splicing machinery in the context of intact genuine splice sites. In the present study, it is shown that cryptic splice sites are utilized in all circumstances, when the authentic site is intact, partially functional or completely abolished. Their use would therefore contribute to a background lack of fidelity in the context of the wild-type sequence. We also found that a mutation at the 5' splice site of beta-globin intron 1 accommodates multiple cryptic splicing pathways, including three previously reported pathways. Focusing on the two major cryptic 5' splice sites within beta-globin exon 1, we show that cryptic splice site selection ex vivo varies depending upon: (a) the cell stage of development during terminal erythroid differentiation; (b) the nature of the mutation at the authentic 5' splice site; and (c) the nature of the promoter. Finally, we found that the two major cryptic 5' splice sites are utilized with differential efficiencies in two siblings sharing the same beta-globin chromosome haplotype in the homozygous state. Collectively, these data suggest that intrinsic, sequence specific factors and cell genetic background factors both contribute to promote a subtle differential use of cryptic splice sites in vivo.

LAMA2 mRNA processing alterations generate a complete deficiency of laminin-alpha2 protein and a severe congenital muscular dystrophy.

An increasing number of genomic variations are no more regarded as harmless changes in protein coding sequences or as genetic polymorphisms. Studying the impact of these variations on mRNA metabolism became a central issue to better understand the biological significance of disease. We describe here a severe congenital muscular dystrophy (CMD) with lumbar scoliosis and respiratory complications in a patient, who died at the age of 10. Despite a poor linkage to any form of CMD, total deficiency of laminin-alpha2 rather suggested the occurrence of an MDC1A form. Extensive analysis of LAMA2 gene revealed two novel mutations: a (8007delT) frameshift deletion in exon 57, and a de novo 7nt deletion in intron 17. Using an ex vivo approach, we provided strong evidence that the intron mutation is responsible for complete exon 17 skipping. The mutations are in trans and they each generate a nonsense mRNA potentially elicited to degradation by NMD. We further discuss the impact of mRNA alterations on the subtle phenotypic discrepancies.

Severe MDC1A congenital muscular dystrophy due to a splicing mutation in the LAMA2 gene resulting in exon skipping and significant decrease of mRNA level.

Congenital muscular dystrophies (CMDs) are a clinically and genetically heterogeneous group of neuromuscular disorders, with autosomal recessive inheritance. We report a patient with severe congenital muscular dystrophy and total deficiency in the laminin alpha2 chain. Genetic analyses showed a linkage to the MDC1A locus for the patient's family, and DNA sequencing revealed in the propositus of a new homozygous mutation in the donor splice site of intron 58 of the LAMA2 gene. RT-PCR experiments performed on total RNA from a patient's muscle biopsy showed a complete skipping of exon 58 in LAMA2 cDNA and a significant decrease in the LAMA2 mRNA level. This exon skipping altered the open reading frame of the mutant transcript and generated a premature termination codon (PTC) within exon 59, which potentially elicits the nonsense mRNA to degradation by NMD (nonsense-mediated mRNA decay). However, the residual exon 58-lacking mRNA could potentially be translated, and the resulting truncated alpha2 chain would lack its LG4 and LG5 domains that are involved in binding with alpha-dystroglycan. These results demonstrate the utility of mRNA analysis to understand the mutation primary impact and the disease phenotype in the patients.

Spi-1/PU.1 but not Fli-1 inhibits erythroid-specific alternative splicing of 4.1R pre-mRNA in murine erythroleukemia cells.

The inclusion of exon 16 in mature protein 4.1R mRNA arises from a stage-specific splicing event that occurs during late erythroid development. We have shown that mouse erythroleukemia (MEL) cells reproduce this erythroid-specific splicing event upon induction of differentiation. We here found that this splicing event is regulated specifically in erythroleukemic cells that have the potential to differentiate and produce hemoglobin, regardless of the nature of the differentiation inducer. Knowing that dysregulated expression of spi-1/pu.1 and fli-1 oncogenes is involved in MEL cell differentiation arrest, we looked at their effect on exon 16 erythroid splicing. We found that exon 16 inclusion requires Spi-1/PU.1 shutdown in MEL cells, and that enforced expression of Spi-1/PU.1 inhibits exon selection, regardless of the presence or absence of a chemical inducer. By contrast, endogenous overexpression or enforced expression of Fli-1 has no effect on exon selection. We further showed that Spi-1/PU.1 acts similarly on the endogenous and on a transfected exon 16, suggesting a promoter-independent effect of Spi-1/PU.1 on splicing regulation. This study provides the first evidence that Spi-1/PU.1 displays the unique property, not shared with Fli-1, to inhibit erythroid-specific pre-mRNA splicing in erythroleukemia cell context.

Protein 4.1R expression in normal and dystrophic skeletal muscle.

4.1R pre-mRNA alternative splicing results in multiple mRNA and protein isoforms that are expressed in virtually all tissues. More specifically, isoforms containing the alternative exon 17a, are exclusively expressed in muscle tissues. In this report, we show that these isoforms are preferentially present in the myoplasm of fast myofibres. 4.1R epitopes are also found at the sarcolemma of both slow and fast myofibres in normal muscle. Interestingly, they are absent from dystrophin-deficient sarcolemma of DMD muscle, and colocalize with partially expressed dystrophin in BMD muscle. We also show that alternative splicing of exons 16 and 17a is regulated during muscle differentiation in an asynchronous fashion, with an early inclusion of exon 16 in forming myotubes, and a late inclusion of exon 17a. Consistently, Western blot analysis led to characterize mainly an approximately 96/98-kDa doublet bearing exons 16-17a-encoding peptide, exclusively occurring in the differentiated muscle.

HnRNP A1 tethers KSRP to an exon splicing silencer that inhibits an erythroid-specific splicing event in PU.1-induced erythroleukemia.

Exon 16 inclusion is a critical splicing event that triggers the production of a functional protein 4.1R in mature normal erythroblasts, and is obviated in PU.1-induced erythroleukemia cells. Exon 16 contains an exonic splicing silencer (ESS16) that interacts with hnRNP A/B in heterologous cell context. We here show that ESS16 promotes the recruitment of a protein complex containing hnRNP A1 and a 79-kDa protein in nuclear extracts from either proliferative erythroleukemia cells or cells induced to terminal differentiation. By using 2D gel fractionation and mass spectrometry, we unambiguously identified KSRP as the 79-kDa component interacting with ESS16. Furthermore, we show that KSRP slightly decreases in erythroleukemia cells induced to terminal erythroid differentiation. Yet, KSRP inducible knockdown, through stable transfection of small hairpin KSRP RNA, did not alter exon 16 splicing, suggesting that KSRP alone does not modulate the splicing event. Interestingly, absence of hnRNP A1 prevented KSRP from binding to ESS16. Reciprocally, KSRP interaction with ESS16 was recovered when hnRNP A1 expression is restored in hnRNP A1-null cells. Collectively, this study establishes that hnRNPA1 is part of a KSRP-containing RNP complex, and emphasizes that, aside from its function in AU-rich element-mediated mRNA decay and its role in microRNA biogenesis, KSRP associates with hnRNP A1 to bind an ESS. These findings further support the role of members of the KH-domain protein family in organizing large RNA-protein complex formation, rather than primarily in modulating specific splicing events.

[Oncogenic transcription factors as splicing regulators]. / Des oncogènes à l'interface entre transcription et épissage.

Oncogene activity ranges from transduction signals to transcription factors. Altered expression of oncogenes, either by chromosomal translocation, proviral insertion or point mutations, can lead to tumor formation. More specifically, data accumulated through the last two decades have shown that disregulation of oncogenic transcription factors can interfere with regulatory cascades that control the growth, differentiation, and survival of normal cells. There is also evidence that alterations of oncogene activity are associated with pre-mRNA splicing defects. The insights gained from the pivotal role of RNA polymerase II in coupling transcription and splicing have instigated a new line of research regarding the possible role of oncogenic transcription factors in pre-mRNA splicing regulation. This review focuses on recent advances addressing this question. Understanding the impact of alterations in the expression and/or function of oncogenes have important prognostic implications that can guide the design of new therapeutic drugs to promote differentiation and/or apoptosis over cell proliferation.

Proteasome-mediated proteolysis of SRSF5 splicing factor intriguingly co-occurs with SRSF5 mRNA upregulation during late erythroid differentiation.

SR proteins exhibit diverse functions ranging from their role in constitutive and alternative splicing, to virtually all aspects of mRNA metabolism. These findings have attracted growing interest in deciphering the regulatory mechanisms that control the tissue-specific expression of these SR proteins. In this study, we show that SRSF5 protein decreases drastically during erythroid cell differentiation, contrasting with a concomitant upregulation of SRSF5 mRNA level. Proteasome chemical inhibition provided strong evidence that endogenous SRSF5 protein, as well as protein deriving from stably transfected SRSF5 cDNA, are both targeted to proteolysis as the cells undergo terminal differentiation. Consistently, functional experiments show that overexpression of SRSF5 enhances a specific endogenous pre-mRNA splicing event in proliferating cells, but not in differentiating cells, due to proteasome-mediated targeting of both endogenous and transfection-derived SRSF5. Further investigation of the relationship between SRSF5 structure and its post-translation regulation and function, suggested that the RNA recognition motifs of SRSF5 are sufficient to activate pre-mRNA splicing, whereas proteasome-mediated proteolysis of SRSF5 requires the presence of the C-terminal RS domain of the protein. Phosphorylation of SR proteins is a key post-translation regulation that promotes their activity and subcellular availability. We here show that inhibition of the CDC2-like kinase (CLK) family and mutation of the AKT phosphorylation site Ser86 on SRSF5, have no effect on SRSF5 stability. We reasoned that at least AKT and CLK signaling pathways are not involved in proteasome-induced turnover of SRSF5 during late erythroid development.

Combined inhibition of PI3K and activation of MAPK p38 signaling pathways trigger erythroid alternative splicing switch of 4.1R pre-mRNA in DMSO-induced erythroleukemia cells.

There is increasing evidence showing that many extracellular cues modulate pre-mRNA alternative splicing, through different signaling pathways. We here show that 4.1R exon 16 splicing is altered in response to specific signals. The switch from erythroblastic isoform lacking exon 16 to mature erythrocytic isoform containing this exon is tightly regulated during late erythroid differentiation, and blocage of this splicing switch in erythroleukemia cells is seen as a consequence of the deregulation of important regulatory pathways. We support that combined inhibition of PI3K and activation of p38 signaling pathways impinge on erythroid 4.1R pre-mRNA alternative splicing switch, and on cell differentiation as witnessed by hemoglobin production. By contrast, MEK/ERK signaling appeared not to affect neither cell hemoglobin production nor erythroid 4.1R pre-mRNA splicing. We also found that the signal-induced alternative splicing is not typically distinctive of EPO-non-responsive cells, but operates in EPO-responsive cells as well. Pre-mRNA splicing is a major regulatory mechanism at the crossroad between transcription and translation. We here provide evidence that inhibition of PI3K activates the splicing switch in a promoter-dependent manner, whereas p38 activation induces this event in a promoter-independent fashion. Our data further support that constitutive activation of EPO-R by the viral protein gp55 and the short form of the tyrosine kinase receptor Stk, transduces PI3K proliferation signal, but not MAPK p38 differentiation signal. Concurrently, this work lend credence to the concept that DMSO triggers transient activation of p38 signaling and irreversible inhibition of PI3K/AKT signaling pathway, hence uncovering an old conundrum regarding the mechanism by which DMSO induces erythroleukemia cell differentiation.

Subtle discrepancies of SF2/ASF ESE sequence motif among human tissues: A computational approach.

The intron removal during the pre-mRNA splicing in higher eukaryotes requires the accurate identification of the two splice sites at the ends of the exons, or exon definition. However, the consensus sequences at the splice sites provide insufficient information to distinguish true splice sites from the large number of the false ones that populate the primary transcripts. Additional information is provided by cis-acting regulatory sequences that serve to enhance or repress splicing, and that may be exonic or intronic in nature: the splicing enhancers and the splicing silencers, respectively. In this study, we tested by computational and statistical approaches if the exonic splicing enhancer motif binding to the SF2/ASF SR protein is conserved among several groups of human genes. The results showed that the SF2/ASF ESE consensus was conserved between genes within the same chromosome, within different chromosomes and between different levels of muscular cells differentiation. However, this motif displays subtle variations within the consensus sequence between genes expressed in different tissues. These results can emphasize the presence of different translational isoforms of the SFRS1 gene encoding for the SF2/ASF, or different post-translational protein maturations in different tissues. This tissular discrepancy can also account for the alternative splicing of several genes between tissues.

Downregulation of the Spi-1/PU.1 oncogene induces the expression of TRIM10/HERF1, a key factor required for terminal erythroid cell differentiation and survival.

Sustained expression of the Spi-1/PU.1 and Fli-1 oncoproteins blocks globin gene activation in mouse erythroleukemia cells; however, only Spi-1/PU.1 expression inhibits the inclusion of exon 16 in the mature 4.1R mRNA. This splicing event is crucial for a functional 4.1R protein and, therefore, for red blood cell membrane integrity. This report demonstrates that Spi-1/PU.1 downregulation induces the activation of TRIM10/hematopoietic RING finger 1 (HERF1), a member of the tripartite motif (TRIM)/RBCC protein family needed for globin gene transcription. Additionally, we demonstrate that TRIM10/HERF1 is required for the regulated splicing of exon 16 during late erythroid differentiation. Using inducible overexpression and silencing approaches, we found that: (1) TRIM10/HERF1 knockdown inhibits hemoglobin production and exon splicing and triggers cell apoptosis in dimethylsulfoxide (DMSO)-induced cells; (2) TRIM10/HERF1 upregulation is required but is insufficient on its own to activate exon retention; (3) Fli-1 has no effect on TRIM10/HERF1 expression, whereas either DMSO-induced downregulation or shRNA-knockdown of Spi-1/PU.1 expression is sufficient to activate TRIM10/HERF1 expression; and (4) Spi-1/PU.1 knockdown triggers both the transcription and the splicing events independently of the chemical induction. Altogether, these data indicate that primary Spi-1/PU.1 downregulation acts on late erythroid differentiation through at least two pathways, one of which requires TRIM10/HERF1 upregulation and parallels the Spi-1/PU.1-induced Fli-1 shutoff regulatory cascade.

Cytoplasmic nonsense-mediated mRNA decay for a nonsense (W262X) transcript of the gene responsible for hereditary tyrosinemia, fumarylacetoacetate hydrolase.

Messenger RNAs containing premature stop codons are generally targeted for degradation through the nonsense-mediated mRNA decay (NMD) pathway. The subcellular localization of the NMD process in higher eukaryotes remains controversial. While many mRNAs are subjected to NMD prior to their release from the nucleus, a few display cytoplasmic NMD. To understand the possible impact of NMD on the pathogenesis of hereditary tyrosinemia type I, a severe metabolic disease caused by fumarylacetoacetate hydrolase (FAH) deficiency, we examined the metabolism of FAH mRNA harboring a nonsense mutation, W262X, in lymphoblastoid cell lines derived from patients and their parents. W262X-FAH transcripts show a approximately 20-fold reduction in abundance in mutant cells, which is translation-dependent. Cellular fractionation shows that this down-regulation of the W262X transcript occurs in the cytoplasm. Thus, the W262X FAH is another example of nonsense mRNAs subjected to the NMD pathway in the cytoplasm.

A splicing alteration of 4.1R pre-mRNA generates 2 protein isoforms with distinct assembly to spindle poles in mitotic cells.

The C-terminal region of erythroid cytoskeletal protein 4.1R, encoded by exons 20 and 21, contains a binding site for nuclear mitotic apparatus protein (NuMA), a protein needed for the formation and stabilization of the mitotic spindle. We have previously described a splicing mutation of 4.1R that yields 2 isoforms: One, CO.1, lacks most of exon 20-encoded peptide and carries a missense C-terminal sequence. The other, CO.2, lacks exon 20-encoded C-terminal sequence, but retains the normal exon 21-encoded C-terminal sequence. Knowing that both shortened proteins are expressed in red cells and assemble to the membrane skeleton, we asked whether they would ensure 4.1R mitotic function in dividing cells. We show here that CO.2, but not CO.1, assembles to spindle poles, and colocalizes with NuMA in erythroid and lymphoid mutated cells, but none of these isoforms interact with NuMA in vitro. In microtubule-destabilizing conditions, again only CO.2 localizes to the centrosomes. These data suggest that the stability of 4.1R association with centrosomes requires an intact C-terminal end, either for a proper conformation of the protein, for a direct binding to an unknown centrosome-cytoskeletal network, or for both. We also found that 4.1G, a ubiquitous homolog of 4.1R, is present in mutated as well as control cells and that its C-terminal region binds efficiently to NuMA, suggesting that in fact mitotic spindles host a mixture of the two 4.1 family members. These findings led to the postulate that the coexpression at the spindle poles of 2 related proteins, 4.1R and 4.1G, might reflect a functional redundancy in mitotic cells.