The abundance of Alu RNA is transiently increased by heat shock in human cell lines. This effect is specific to Alu repeats among Pol III transcribed genes, since the abundance of 7SL, 7SK, 5S and U6 RNAs is essentially unaffected by heat shock. The rapid induction of Alu expression precedes the heat shock induction of mRNAs for the ubiquitin and HSP 70 heat shock genes. Heat shock mimetics also transiently induce Alu expression indicating that increased Alu expression is a general cell-stress response. Cycloheximide treatment rapidly and transiently increases the abundance of Alu RNA. Again, compared with other genes transcribed by Pol III, this increase is specific to Alu. However, as distinguished from the cell stress response, cycloheximide does not induce expression of HSP 70 and ubiquitin mRNAs. Puromycin also increases Alu expression, suggesting that this response is generally caused by translational inhibition. The response of mammalian SINEs to cell stress and translational inhibition is not limited to SINEs which are Alu homologues. Heat shock and cycloheximide each transiently induce Pol III directed expression of B1 and B2 RNAs in mouse cells and C-element RNA in rabbit cells. Together, these three species exemplify the known SINE composition of placental mammals, suggesting that mammalian SINEs are similarly regulated and may serve a common function.

Promoter elements derived from the 7SL RNA gene stimulate RNA polymerase III (Pol III) directed Alu transcription in vitro. These elements also stimulate expression of Alus transfected into 293 cells, but transcripts from these same constructs are undetectable in HeLa cells. A terminator resembling the terminator for the 7SL RNA gene has no effect on in vitro Alu template activity, but increases expression in vivo in a position independent manner. Alu transcripts generated from templates with and without this terminator have identical half-lives, indicating that this terminator stimulates expression by increasing template activity. Together, these results show that Alu expression may be regulated at multiple levels and can respond to cis-acting elements. This new found ability to express Alu transcripts by transient transfection provides an opportunity to monitor their post-transcriptional fate. Primary Alu transcripts are not extensively adenylated or deadenylated following transcription, but are short-lived compared to 118 nt scAlu RNA. In addition to Alu RNA, transfected templates encode scAlu RNA, but very high levels of Alu RNA expression does not increase the abundance of scAluRNA. ScAluRNA is not merely a transient RNA degradation product, but is instead tightly regulated by factors other than the abundance of primary transcripts.