A comprehensive splicing characterization of COL4A5 mutations and prognostic significance in a single cohort with X-linked alport syndrome

Abstract

Introduction: X-linked Alport syndrome (XLAS), caused by mutations in the COL4A5 gene, is an X-linked hereditary disease typically characterized by renal failure, hearing loss, and ocular abnormalities. It is a leading hereditary cause of end-stage renal disease (ESRD) worldwide. Studies on the genotype-phenotype correlation in Alport syndrome suggest that splicing mutations result in more severe clinical phenotypes than missense mutations. Determining whether COL4A5 mutations lead to aberrant mRNA splicing is critical for diagnosis and prognosis.

Methods: This study retrospectively reviewed pediatric XLAS patients with COL4A5 gene mutations from a single-center cohort, summarizing and analyzing their clinical features. Minigene assay was employed to evaluate the mRNA splicing functionality of 26 single-nucleotide variants (SNVs), both intronic and exonic, identified in XLAS patients. Bioinformatics tools were used to evaluate the accuracy and sensitivity of splicing mutation prediction. Additionally, linear mixed models were applied to analyze the relationship between mutation types and prognosis in patients' estimated glomerular filtration rate (eGFR), exploring genotype-phenotype correlations.

Results: In this cohort, we screened 41 XLAS pediatric patients, including 32 with confirmed XLAS and nine suspected XLAS. The cohort included 21 males (51.2%) and 20 females (48.8%), with a median age at onset of 4.42 years. Among the patients, 22 presented with both hematuria and proteinuria, while 18 exhibited hematuria alone. Notably, only one patient had isolated proteinuria. Regarding mRNA splicing, among the 26 intronic and exonic SNVs, 10 mutations (38.5%) were found to cause aberrant mRNA splicing, as demonstrated by the minigene assay. Sensitivity and specificity assessments of bioinformatics tools revealed that ESE Finder demonstrated higher sensitivity, while RNA Splicer exhibited greater specificity. Furthermore, These splicing abnormalities were closely associated with a faster decline in eGFR.

Conclusion: This study demonstrates that 38.5% of SNVs in the COL4A5 gene result in aberrant mRNA splicing, which is closely linked to renal function decline in XLAS. Splicing mutations are correlated with more rapid renal progression, highlighting the importance of determining the splicing effects of SNVs during genetic screening for XLAS.

Keywords: COL4A5; alport syndrome; mRNA; minigene assay; splicing.

FIGURE 1

The schematic diagram of the minigene splicing assay constructed by pSPL3 exon trapping vector.

FIGURE 2

The flowchart of this study.

FIGURE 3

Agarose gel electrophoresis and statistical analysis of RT-PCR products of Intronic SNVs in the COL4A5 gene. Note (A) Lane 1: Marker DL 2000; Lane 2: pSPL3 (263 bp); Lane 3: Exon 13 (356 bp); Lane 4: c.780 + 243 T>C (263 bp); Lane 5: Exon 13 (356 bp); Lane 6: c.780 + 5G>A (263 bp); Lane 7: Exon 31 (431 bp); Lane 8: c.2510–2A>G (431 bp); Lane 9: Exon 40 + 41 (263 bp and 500 bp); Lane 10: c.3604 + 65A>G (263 bp and 500 bp); Lane 11: Exon 50 (263 bp and 436 bp); Lane 12: c.4976 + 1G>A (263 bp). (B) Quantification is expressed as the percentage (%) of exon skipping; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 4

Agarose gel electrophoresis and statistical analysis of RT-PCR products of exonic SNVs in the COL4A5 gene. Note (A) Lane 1: Marker DL 2000; Lane 2: pSPL3 (263 bp); Lane 3: Exon 3 (353 bp); Lane 4: c.220C>T (353 bp); Lane 5: Exon 4 (308 bp); Lane 6: c.262C>T (308 bp); Lanes 7, 9, 11: Exon 11 + 12 (341 bp); Lane 8: c.638G>A (263 bp and 341 bp); Lane 10: c.641C>T (341 bp); Lane 12: c.679G>A (341 bp); Lane 13: Exon 17 (317 bp); Lane 14: c.973G>A (317 bp); Lane 15: Exon 20 (437 bp); Lane 16: c.1276G>A (437 bp) (C) Lane 1: Marker DL 2000; Lane 2: pSPL3 (263 bp); Lane 3: Exon 22 (356 bp); Lane 4: c.1499G>A (356 bp); Lane 5: Exon 24 (263 bp); Lane 6: c.1634G>A (263 bp and 455 bp); Lane 7: Exon 25 (432 bp); Lane 8: c.1781G>T (432 bp); Lane 9: Exon 28 (263 bp); Lane 10: c.2228G>A (361 bp); Lane 11: Exon 29 (263 bp); Lane 12: c.2314G>C (263 bp and 414 bp); Lane 13: Exon 29 (263 bp and 414 bp); Lane 14: c.2329C>A (263 bp); Lane 15: Exon 31 (431 bp); Lane 16: c.2605G>A (431 bp). (E) Lane 1: Marker DL 2000; Lane 2: pSPL3 (263 bp); Lane 3: Exon 31 (263 bp and 431 bp); Lane 4: c.2624G>A (263 bp); Lane 5: Exon 32 (263 bp and 353 bp); Lane 6: c.2678G>A (263 bp); Lane 7: Exon 35 (263 bp); Lane 8: c.3059T>C (353 bp); Lane 9: Exon 37 (390 bp); Lane 10: c.3319G>A (390 bp); Lane 11: Exon 41 (449 bp); Lane 12: c.3667G>A (263 bp and 449 bp); Lanes 13, 15: Exon 42 (397 bp); Lane 14: c.3808G>T (397 bp); Lane 16: c3809G>A (397 bp) (B, D and F) Quantification is expressed as the percentage (%) of exon skipping; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 5

Constructing a linear mixed model based on mutation types. Note (A) Linear mixed model of three mutation types: missense mutation group (blue, n = 14), splicing mutation group (red, n = 10), and truncating mutation group (green, n = 11) (B) Linear mixed model of SNVs and splicing mutations after minigene analysis: SNV group (blue, n = 26) and splicing mutation group (red, n = 10). Shaded areas represent the 95% confidence interval (95% CI). Dots indicate eGFR values at different ages for individual patients. Straight lines connect data points from the same individual.