Analysis of the 3' end of the human pro-α2(I) collagen gene. Utilization of multiple polyadenylation sites in cultured fibroblasts

Three overlapping genomic clones covering 28 kilobases of the human pro-α2(I) collagen gene have been isolated from a λ phage library. The analysis of 12 introns and 12 exons in the 3' end region has shown that the human gene has a structure remarkably similar to that reported for the homologous chicken gene. One large intron, in the α-chain domain, contains an AluI sequence flanked by short direct repeats; a second AluI sequence is present 4 kilobases downstream from the termination codon. The analysis of the exon coding for the 3'-untranslated region has revealed that the pro-α2(I) collagen gene transcribes at least four different mRNAs in cultured fibroblasts. The colinearity and exact location of the termini of these transcripts was determined by Northern blots, R-looping analysis, S1 protection, and DNA sequencing. The ends of two transcripts are closely preceded by the canonical polyadenylation signal (AAUAAA), whereas two of its variations (AUUAAA and AUUAA) precede the ends of the other two transcripts.