Brain-specific polyA- transcripts are detected in polyA+ RNA: do complex polyA- brain RNAs really exist?

Transcripts encoded by 2 different rat genomic clones, rg13 and rg100, appear to be typical brain-specific polyA- RNAs, as defined by previous criteria (rare, polysomal, and postnatally expressed from single-copy genes). However, we have found by using a sensitive nuclease protection assay that low levels of these transcripts (10% and 3%, respectively) are detected in polyA+ RNA. To determine if rg transcripts that fractionate as polyA- could have resulted from nicking of polyA+ RNA, we assessed the integrity of 2 known polyA+ RNAs, those of tyrosine hydroxylase, a 2-kilobase (kb) mRNA, and sodium channel, a 9.5-kb RNA. Using RNA prepared by several different procedures, including LiCl-urea and guanidine thiocyanate followed by CsCl centrifugation, the shorter message fractionated as polyA+ after 2 cycles over oligodeoxythymidine (oligo-dT) cellulose, whereas the majority of the longer sodium channel RNA fractionated as polyA, as assayed by nuclease protection using probes from the 5' end of the 2 genes. However, on Northern blots, the same RNA preparations showed an intact 9.5-kb sodium channel band only in polyA+ RNA, suggesting that the polyA- RNAs were randomly cleaved, resulting in a smear of sizes that could not be detected as a discrete band. These data imply that long messages may be nicked during standard isolation procedures and that this would not be detected by Northern blot analysis.(ABSTRACT TRUNCATED AT 250 WORDS)