Effects of the hMLH 1 and hMSH2 mutations on exon inclusion. Results of the splicing assay with the different hMLH1(A) and hMSH2(B) mutated constructs. Numbering is relative to the nucleotide position in the ORF. Cos-7 cells were transfected with 1 μg of the indicated mutant minigene variants or the corresponding wild-type exon, RNA was extracted, reverse transcribed, and amplified with primers SD6 and SA2. The RT-PCR products were resolved on GeneGel Excel, stained with ethidium bromide and quantitated with an image analyzer (see methods). V = vector only; Mw = size standard. The black arrowhead represents the exon skipped product. The percentage of exon inclusion is indicated above each lane. The white asterisks show the splicing product deriving from use of an internal cryptic donor site. Below the gel are reported the predictions for the three algorithms: = no change; + the mutation creates an ESE or abrogates an ESS sequence; - the mutation creates an ESS or abrogates an ESE sequence; 0 the mutations is not localised in, and does not create or disrupt any regulatory sequence. (C, D) Graphic representation of the splicing assay results. The average of percent exon inclusion is reported in the y-axis and represents the mean of two independent transfections done in triplicate for each construct (x-axis). White bars are used for normal alleles, patterned for mutated constructs. Mutations within the same exons are grouped together and with their corresponding normal exon. Error bars represent standard deviation. The mutated constructs causing significant differences when data were analysed using Student's t test are underlined (* = P < 0.05, ** = P < 0.01,*** = P < 0.001).