3.1.2 Variants of c.1437C>G (p.Cys479Trp) and c.1564G>A (p.Glu522Lys) in exon 13 did not alter pre-mRNA splicing
Variant c.1437C>G (p.Cys479Trp), located at nucleotide position +6 from the 5’ end of exon 13, was predicted to lead to the disruption of two ESEs and the generation of four new ESS sites using bioinformatic tool HSF. Presumed missense variant c.1564G>A (p.Glu522Lys) is located 133 bp upstream of exon 13 (Table 1). The results of HSF analysis indicated that this variant inactivated four potential ESE sites and generated two potential ESSs. However, analysis of the minigenes containing variant c.1437C>G and c.1564G>A in exon 13 of SLC4A1 resulted in RT-PCR products that matched in size with those generated by the respective WT constructs (Figure 2A). This was further confirmed by sequencing analysis. Therefore, these variants did not affect pre-mRNA splicing.
3.2 Splicing outcome of sequences variations of ATP6V1B1
3.2.1 Variants c.368G>T (p.Gly123Val) andc.370C>T (p.Arg124Trp) led to exon 5 skipping compared with the WT plasmids.
Presumed missense variant c.368G>T (p.Gly123Val) is caused by the first nucleotide substitution in exon 5 ofATP6V1B1 . Bioinformatic analysis with BDGP demonstrated that this variant drastically reduces the score of the WT 3’ splice site from 0.96 to 0.76 (Table 1). Variant c.370C>T (p.Arg124Trp) result from substitutions at the +3 nucleotide of exon 5 and was demonstrated that this variant marginally reduces the score of the WT 3’ splice site from 0.96 to 0.94 with BDGP. Additionally, analysis of HSF to two variants predicted that they both generated three ESSs. To determine the splicing effect of variants c.368G>T and c.370C>T, a minigene containing exon 5 and flanking intronic sequences was used. The result of RT-PCR analysis showed that the WT lane demonstrated 2 different fragments of 263 bp and 368 bp, respectively (Figure 2B). Direct sequencing showed that the larger fragment contained exon 5 flanked by two exons of the pSPL3 vector, while the smaller one included only the 3’ and 5’ pSPL3 exons. In contrast, the mutant lane of c.368G>T revealed one unique fragment of 263 bp corresponding to skipping of exon 9 in the mRNA; whereas, the result of variant c.370C>T revealed two different electrophoresis bands whose sizes correspond to the WT products. Analysis of cDNA prepared from HEK293T showed that there was a significant increase of exon 5-skipping transcript of c.370C>T compared with the WT plasmids (28.19% versus 4.78%, P<0.05, Figure 2D). Deletion of exon 5 in the mutant transcript would result in the abnormal connection between exons 3 and 4, leading to the loss of 26 amino acids at the protein level without changing the open reading frame. Taken together, both of variants c.368G>T and c.370C>T broken proper recognition of the acceptor splicing site, resulting in exon 6 skipping of ATP6V1B1 .
3.2.2 Nonsense variant c.484G>T ( p.Glu162*) led to the skipping of exon 6 .
Variant c.484G>T of the internal position of exon 6 inATP6V1B1 alters a GAG codon for Glu to a premature TAG stop codon ((p.Glu162*), which is predicted to produce a truncated and nonfunctional protein (Table 1). However, the in silico analysis by HSF software indicated that c.484G>T not only disrupts five ESEs (CCCGAGG A, CCGAGG , CCGAGG AG, CGAGG A, GAGG AG), but also creates two ESSs (GT AGAT, T AGATG). Besides, HSF also predicted that this variant would cause the activation of a cryptic donor site (CGAGT AGAT). To further clarify the impact of variant c.484G>T on the splicing process, we inserted either the WT or c.484G>T-mutated exon 6, along with the nearby intron sequences, into the vector pSPL3, and analyzed the spicing pattern in HEK293T. The WT minigene resulted in two different transcripts corresponding to a mature mRNA and a truncated mRNA (1.22%) by sequencing analysis. The mutated minigene, also, produced two mature transcripts: a larger one corresponded to the single band produced by the WT, and a shorter one with the splicing of exon 6 (Figure 2B). Because exon 6 has 140 nucleotides, loss of exon 6 would disrupt the reading frame of the ATP6V1B1 transcript. Quantitative analysis of the RT-PCR products revealed that the rate of transcript lacking exon 6 to the full transcripts was 60.35% (P<0.05, Figure 2D). Taking these findings together, the splicing pattern of the mutant minigene indicated that nonsense variant c.484G>T induces exon 6 skipping.
In addition, although variant c.481G>A (p.Glu161Lys) close to c.484G>T was also predicted by the software HSF to destroy five ESEs and produce two ESSs, it did not affect splicing through the minigene analysis (P>0.05).
3.2.3 Variant c.1102G>A ( p.Glu368Lys) resulted in partial skipping of exon 11.
Variant c.1102G>A identified at nucleotide position 43 of exon 11 in ATP6V1B1 were predicted by HSF to broke seven ESEs (CACAG A, CACAG AG, ACAG AGGG, CAG AGG, CAG AGGG, G AGGGA) and create one ESS site (AA AGGG) corresponding to a new hnRNPA1 binding site (Table 1). In order to verify whether this variant affected mRNA splicing, minigene splicing experiments in vitro were also carried. As a result, two products of 346 bp and 236 bp were both found from the RT-PCR products of the WT minigene and mutant (Figure 2B). Direct sequencing of two products showed that the larger amplicon was the transcript containing exon 11, and the smaller was the transcript excluding exon 11. Quantitative analysis of two products showed that c.1102G>A altered weakly splicing resulting in a 9.51% of the aberrant transcript that broke the open reading frame compared with WT of 2.60% (P<0.05, Figure 2D). In addition, the skipping of exon 11 in the mutant transcript would result in the loss of 83 nucleotides that disrupt the reading frame.
3.3 Splicing outcome of sequences variations ofATP6V0A4.
3.3.1 Variant c.322C>T (p.Gln108*) resulted in skipping of exon 6.
Variant c.322C>T, as a nonsense variant (p.Gln108*) located at nucleotide position +31 from the 5’ end of exon 6, is predicted to generate a truncated and nonfunctional protein (Table 1). The software HSF demonstrated that c.322C>T not only destroyed six ESEs (C AGGAA, TAC AGGA, TAC AGGAA, TTAC AG, GTTAC A, AGTTAC ), but also engenders three ESSs (TAT AGG, TTAT AG, AGTTAT A). Consequently, two fragments of 389 bp and 263 bp were detected from that of the mutant vector by minigene assay, respectively (Figure 2C). Direct sequencing results showed that the larger fragment containedATP6V0A4 exon 6 flanked by two exons of the pSPL3 vector, while the smaller one included only the 3’ and 5’ pSPL3 exons (24.22%). However, the WT vector also revealed a smaller fragment (1.78%) of 263 bp corresponding to incorrect skipping of exon 6 in ATP6V0A4expect for a mature transcript (Figure 2D). The skipping of exon 6 in the mutant transcript would result in the loss of 42 amino acids (residues 98-139) at the protein level without altering the open reading frame. Therefore, variant c.322C>T (p.Gln108*) causes partial exon 9 skipping because of ESEs destruction and ESSs generation.
3.3.2Missense variant c.1571C>T (p.Pro524Leu) and synonymous variant c.1572G>A (p.Pro524Pro) resulted in skipping of exon 15.
Missense variant c.1571C>T (p.Pro524Leu) caused by substitutions at the -2 nucleotide of exon 15 downstream of the 3’ splice site, was demonstrated that this variant reduces the score of the WT donor splice site of intron 15 from 0.8 to 0.68 with BDGP (Table 1). Synonymous variant c.1572G>A (p.Pro524Pro) affected the G the last nucleotide substitution in exon 15. Bioinformatic analysis with BDGP demonstrated that the score of the donor site of intron 15 is 0.8, whereas it could not be analyzed after variation. Additionally, analysis of this variant with HSF predicted to break the WT donor sites (CCG GTAATA), most probably affecting splicing. Taken together, to examine the splicing effect of two variants, we also used a minigene containing exon 15 and surrounding intronic sequences. RT-PCR analysis results showed the splicing products produced by the mutant and WT minigenes were different. The WT lane demonstrated one fragment of 357 bp that contains ATP6V0A4 exon 15, whereas both of mutant c.1571C>T and c.1572G>A generated two different fragments of 263 bp and 357 bp, respectively (Figure 2C). Direct sequencing of all products showed that the larger amplicons were the exons-included transcripts and the smaller amplicons are the exons-excluded transcripts. Analysis of cDNA prepared from HEK293T revealed that there was a significant increase of exon 15-skipping in c.1571C>T and c.1572G>A with the control plasmid (3.74% versus 0 and 84.48% versus 0, respectively, Figure 2D). Besides, the skipping of exon 15 in the mutant transcript would result in the alteration of the open reading frame due to the loss of 94 nucleotides.