Supplementary Data

Patient ID 108

This patient had a maternal deletion of 17 bp in exon 33 of the COL4A3 gene, but the paternal disease allele was missing. Step 2 analysis revealed exon 46 skipping (Figure 2-I-a). This finding prompted genomic DNA sequencing of introns 45 and 46, which led to identification of a point mutation upstream of the 3′ splice site of intron 46, c.4028-27A>T (IVS46-27A>T) (Figure 2-I-b). Adenosine mutations at this location that diminish exon inclusion may affect branch sites, key splice-site recognition sequences that are usually located 18–40 bp upstream of the 3′ splice sites.[1] To investigate this hypothesis, we used the Human Splicing Finder program (www.umd.be/HSF) and found c.4028-27A as a motif for branch point. As intron 46 is very short and bridging interactions across the intron during spliceosome assembly are probably critical, the branch site remains to be confirmed experimentally.

Patient ID 155

The parents of this patient were first cousins, suggesting homozygosity; however, they did not show any mutations in Step 1 analysis. Step 2 analysis identified a 139 bp insertion between COL4A3 exon 48 and exon 49 (Figure 2-II-a), corresponding to a 139 bp sequence of intron 48. Genomic DNA sequencing detected a point mutation, c.4463-523C>G (IVS49-523C>G), activating a cryptic exon in intron 49 (Figure 2-II-b). Analysis of repetitive sequences by RepeatMasker (http://www.repeatmasker.org/) showed that the new exon was derived entirely from an antisense AluJ sequence (Figure 2-II-c). The length of the new exon was identical to that of the average human exon (139 nt) [2], suggesting that activation was facilitated by the optimal exon size. Interestingly, the mutation was at the same position of the Alu consensus in an intron of the OAT gene as reported previously [3]; this study was the first reported example of a hereditary disease resulting from Alu exonization. The size of the new OAT exon was also very similar (142 nt) to the cryptic exon in this patient.

As mutagenesis is often linked to transcription and recurring human gene mutations can be facilitated by structural changes of nascent transcripts [4], we carried out predictions of local secondary structures of overlapping segments across the new exon. This analysis suggested that the mutation could destabilize a relatively stable stem-loop structure encompassing the new 5′ splice site (Figure 2-II-d), which may have contributed to efficient recognition of the new exon.

Patient ID 165

Using genomic DNA analysis, we detected heterozygous c.1576-20_-6del15bp (IVS25-20_-6del15bp) on the paternal allele and heterozygous c.689G>A (p.G230D) in exon 13 on the maternal allele of COL4A3 (Figure 2-III-a). Analysis of mRNA extracted from leukocytes (Figure 2-III-b) and urinary sediments (data not shown) revealed exon 25 skipping in the patient’s mRNA, consistent with the notion that shortening of exons may interfere with their recognition.

Patient ID 169

With Step 1 analysis, we detected only a heterozygous point mutation of c.4793T>G (p.L1598R) on the paternal allele of the COL4A3 gene; a mutation on the maternal allele was absent. Step 2 analysis did not reveal any RNA processing abnormalities. Step 3 analysis revealed that the heterozygous deletion involving exon 43 and 44 was on the maternal allele (Figure 2-IV-a). The break point of this large deletion in genomic DNA found by long PCR and direct sequencing was c.3752-511_3955+576del (IVS43-511 and IVS44+576del) (Figure 2-IV-b).

Patient ID 218

With Step 1 analysis, we detected only a heterozygous point mutation of c.2878G>A (p.G960R) on the paternal allele of the COL4A4 gene; a maternal disease allele was absent. Step 2 analysis failed to identify obvious RNA processing defects. Step 3 analysis revealed a heterozygous deletion encompassing exons 8–25 on the maternal allele (Figure 2-V-a). We detected the break point of this large deletion in genomic DNA using long PCR and direct sequencing method as c.559-491_1460-808delinsPolyT (IVS8-491_IVS26-808delinsPolyT) (Figure 2-V-b), again confirming compound COL4A4 heterozygosity.


References

1. Kol G, Lev-Maor G, Ast G (2005) Human-mouse comparative analysis reveals that branch-site plasticity contributes to splicing regulation. Hum Mol Genet 14:1559-1568.

2. Kralovicova J, Vorechovsky I (2007) Global control of aberrant splice-site activation by auxiliary splicing sequences: evidence for a gradient in exon and intron definition. Nucleic Acids Res 35:6399-6413.

3. Mitchell GA, Labuda D, Fontaine G, Saudubray JM, Bonnefont JP, Lyonnet S, Brody LC, Steel G, Obie C, Valle D (1991) Splice-mediated insertion of an Alu sequence inactivates ornithine delta-aminotransferase: a role for Alu elements in human mutation. Proc Natl Acad Sci U S A 88:815-819.

4. Vorechovsky I, Jones AV, Cross NC (2013) Why do we see JAK2 exon 12 mutations in myeloproliferative neoplasms? Leukemia Epub ahead of printing.


Figure legends

Supplementary Figures.

1. Genetic analysis of Patient ID108.

a. COL4A3 cDNA analysis showed an extra band in the patient’s sample. Direct sequencing showed exon 46 skipping. b. Sequencing of genomic DNA (gDNA) encompassing intron 46 revealed a putative branch point mutation, c.4028-27A>T (IVS46-27A>T).

2. Genetic analysis of Patient ID155.

a. Amplification of COL4A3 cDNA showed a cryptic exon activation between exon 48 and 49. b. Sequencing of intron 48 showed a point mutation, c.4463-523C>G (IVS49-523C>G). This point mutation activated a cryptic 5′ splice site. c. Alignment of the mutation-induced COL4A3 cryptic exon with the AluJ retroposon. Alignment was carried out using the cross_match tool from RepeatMasker. The cryptic exon is highlighted gray; the mutated residue creating the 5′ splice site (CT>GT) is shown with an arrow. i: transitions; v: transversions. d. Secondary structure predictions across the cryptic exon. The mutated residue is denoted by the arrow. The C>G mutation may destabilize the loop structure by reducing predicted free energy from −10.2 to −9.3 kcal/mol.

3. Genetic analysis of Patient ID165.

a. COL4A3 genomic DNA analysis showed c.1576-20_16del15bp (IVS25-20_16del15bp). A 15 bp segment (highlighted red) was deleted on the mutant allele. b. cDNA analysis showed exon 25 skipping.

4. Genetic analysis of Patient ID169.

a. Semiquantitative PCR for COL4A3 showed two bands encompassing exon 43 and 44 in the patient’s and his mother’s samples. Primer pairs amplifying the MARK gene were used as internal controls. COL4A3 exon 42, 45, 46 and MARK genes showed equal expression. b. Detection of large deletion break point by genomic DNA (gDNA) analysis. Sequencing chromatogram revealed a large deletion, c.3752-511_3955+576del spanning introns 42–44.

5. Genetic analysis of Patient ID218.

a. Semiquantitative PCR for COL4A4 showed half-density bands for exons 8–25 in the patient’s and his mother’s samples. Gene primer pairs for the FN1 gene were used as internal controls. COL4A4 exon 7, 26 and MARK genes showed equal expression, consistent with exons 8–25 being deleted. b. Detection of a large deletion break point by genomic DNA (gDNA) analysis. Sequence chromatogram revealed a large deletion, c.559-491 1460-808delinsPolyT, spanning introns 7–25.