Gene expression from the c-erbA alpha/Rev-ErbA alpha genomic locus. Potential regulation of alternative splicing by opposite strand transcription.

Alternative splicing of the c-erbA alpha gene transcript generates mRNAs encoding a thyroid hormone (T3) receptor (alpha 1) and a non-T3-binding variant (alpha 2). Also encoded at this genomic locus is Rev-ErbA alpha (Rev), a non-T3-binding member of the T3/steroid hormone receptor family. The DNA strand coding for Rev is opposite of that encoding the c-erbA alpha proteins, such that bidirectional transcription is required to generate the Rev and alpha 2 mRNAs. We have used cycloheximide as a tool for studying the regulation of gene expression from this complex genomic locus. In 235-1 cells cycloheximide treatment increased Rev mRNA levels by 50-100-fold. The effect was detectable after a 1-2-h incubation with 20 micrograms/ml cycloheximide and maximal at 24 h. The cycloheximide-induced increase in Rev gene expression was due to inhibition of protein synthesis since anisomycin caused a similar induction of Rev mRNA. Nuclear run-on assays revealed an approximately 10-fold increase in the transcriptional rate of the Rev gene during cycloheximide treatment. In addition, cycloheximide stabilized the Rev mRNA, as evidenced by half-life determinations in the presence of actinomycin D (5 ng/ml). Thus, labile proteins repress basal Rev gene expression both transcriptionally and post-transcriptionally. Cycloheximide treatment also increased alpha 1 and decreased alpha 2 mRNA levels, with the alpha 1/alpha 2 ratio increasing approximately 10-fold. Interestingly, however, these effects appeared not to be due to changes in rates of transcription or degradation of either mRNA. A likely alternative, given the structure of the c-erbA alpha gene, is that inhibition of protein synthesis resulted in a change in the c-erbA alpha splicing pattern. We hypothesize that this may be causally related to the accumulation of Rev mRNA (or the nascent Rev gene transcript), which could inhibit splicing to alpha 2 by hybridizing to complementary sequences, favoring alpha 1 mRNA production. Since alpha 1 is a T3 receptor whereas alpha 2 is an inhibitor of T3 action, such changes could influence target cell responsiveness to T3.