During a search for cDNA clones from the P gene, several clones homologous to the direct repeats were recovered. An oligo dT-primed cDNA library was constructed using mRNA purified from fully pigmented pericarps (P-rr), and 1.8 million independent clones were screened using a 664 bp genomic DNA fragment which codes for the 3' end of the P-specific messenger RNAs. This genomic fragment is included within the repeat located at the 3' end of the P-gene. In addition to several P cDNA clones, we rescued six clones up to 850 bp long coded by the opposite strand from which P is transcribed. The 3' end of the transcripts represented by these clones would be 10 bp from the 3' end of the most extended P cDNA clones found. The six clones could be divided into two families depending upon the length of their homology with the direct repeat genomic sequence: clones in family I contain 343 bp identical to the direct repeat, while family II contains 393 bp identical to the direct repeat. Beyond the region of identity with the direct repeat sequence, the sequences of the six clones were completely unrelated to the P genomic sequence.
From where were these clones transcribed? To answer this, we made a 650 bp probe from the 3' end of one of these clones and used it to hybridize Southern blots of the 34 kbp of the P locus previously cloned; hybridization was found only over the 343 bp region homologous to both flanking repeats. Hybridizations of the probe to genomic Southern blots showed homology to at least five other sequences in the maize genome. The same hybridizing fragments were detected using the 664 bp P genomic sequence outside the P locus. A Genbank search showed no significant homology of the cDNA clones to any known sequence.
When used as a probe on RNA blots, the 650 bp probe from the clone failed to show any signal under conditions in which the P transcripts are readily seen. Only very weak signals were seen when a high specific activity single strand probe was used. Therefore, the abundance of these transcripts must be much lower than the P transcripts.
We don't yet know the significance, if any, of these transcripts for expression of P. However, our results resemble in some ways those obtained for the Bz2 locus by Schmitz, Theres and Starlinger (MNL 63:60, 1989). In the Bz2 case, the "anti-sense" transcripts are about 100 times less abundant than the Bz2 message; however, the "anti-sense" transcripts partially overlap with the Bz2 transcript.
We thank Susan Allan for technical assistance.
to the MNL 64 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page