Page 319 - Haematologica March 2020

NGS of F7 and minigene studies identify molecular bases of FVII deficiency
exon 6, from usage of the weak cryptic 5’ss at position +79 and, most importantly, from a pseudo-exonization event (transcripts 1, 3G, and 4, respectively). More pre- cisely, this new pseudo-exon 5b originates from the usage of a cryptic 5’ss and of a 282 bp upstream cryptic 3’ss. This finding was further strengthened by a PCR using a primer in the pseudo-exon 5b (F7PsExR), which gave rise to amplified fragments only in cells expressing the pIVS6- 392G construct (Figure 2B, lower panel). The inclusion of the pseudo-exon leads to a frame-shifted mRNA with a premature nonsense codon, predicted to encode a dys- functional FVII protein. Semi-quantitative evaluation of transcripts by fluorescent labeling of amplicons and dena- turing capillary electrophoresis revealed that the correct and the aberrant forms are present in the relative propor- tion of 74% and 26%, respectively (Figure 3C), compared to 90% and 10% in the pIVS6-wt context (Figure 3A).
Overall, these data demonstrate that both mutations exert their detrimental effect by impairing FVII splicing, strengthening the usage of cryptic 5’ss.
Investigation by antisense U7snRNA
Since the observed aberrant splicing is caused by the usage of new 5’ss, we hypothesized that masking them
would weaken or abolish their detrimental role. To this purpose, we exploited variants of the U7 small nuclear RNA (U7smOPT)41 as potent antisense molecules to tar- get the alternative splice sites (Figure 4A).
Co-expression of the c.571+78G>A change with anti- sense U7smOPT variants resulted in an appreciable res- cue of splicing, as evaluated by densitometric analysis of bands upon semi-quantitative PCR. In particular, the pro- portion of correct transcripts, barely appreciable in untreated conditions, remarkably increased to approxi- mately 9% or to approximately 20% of total transcripts upon co-expression of the pU7+78Ash or pU7+78A, respectively (Figure 4B, bottom).
Concerning the c.572-392G variant, co-expression of pU7-392G, designed on the cryptic 5’ss, resulted in a 3- fold reduction in aberrantly spliced mRNA containing the pseudo-exon 5b and conversely favored (1.3-fold increased) the synthesis of correctly spliced transcripts that rose from approximately 70% to approximately 90% of all forms (Figure 4B, bottom), resembling the propor- tion observed in the wild-type context.
Overall, these data further demonstrate the causative role of the mutations that create/strengthen cryptic intronic 5’ss.
AB
C
Figure 2. Alternative splicing patterns associated with the c.571+78G>A and c.572-392C>G mutations. (A) Schematic representation of the pIVS6 minigene. Mutations (blue) are reported on top. The presence of cryptic splice sites (5’ss in light blue, 3’ss in green), with related scores are indicated by arrows. Polymerase chain reaction (PCR) oligonucleotides are shown in red. (B) The schematic representation of splicing patterns is reported together with relative sequencing chro- matograms of the amplicons, obtained by T7bisF-F7ex7R PCR (see panel A). PCR fragments were cloned before sequencing. (C) Splicing pattern analysis in HEK293T cells transiently transfected with pIVS6 wild type (wt) or with pIVS6 variants c.571+78A (+78A) and c.572-392G (-392G); the PCR with T7bisF-F7ex7R oligonu- cleotides is reported in the upper panel; PCR with T7bisF-F7psExR oligonucleotides specifically designed to amplify transcripts with the pseudo-exon5b, is reported in the lower panel. M:100 bp ladder; hd: heteroduplex.
haematologica | 2020; 105(3)
833