One of the most striking discoveries by Chekanova
et al. [
2] was the role of the exosome in controlling the levels of antisense transcripts from rRNA loci. After RNAi-mediated suppression of RRP41 or RRP4, increases and decreases in the level of antisense transcripts occurred in chromosomal regions that form boundaries to the mature rRNAs, suggesting the possibility of a new layer of regulation for these highly transcribed genes. It seems likely that the antisense transcripts play a direct role in controlling rRNA levels, as their abundance correlates inversely with the levels of the adjacent mature rRNA. Perhaps antisense transcription affects the local chromatin environment and thus the accessibility of the respective rRNA promoter. Indeed, a recent report suggests that exosome control of the level of cryptic transcripts from the rDNA locus plays an important role in heterochromatic gene silencing [
8]. Chekanova
et al. [
2] also detected polyadenylated antisense transcripts that accumulated near mRNA promoter regions. Recently, Camblong and colleagues [
9] showed that antisense RNA production near the
PHO84 promoter in
S. cerevisiae results in the recruitment of the histone deacetylase Hda1 and subsequent repression of
PHO84 transcription. Interestingly, the stabilization of an antisense transcript near a gene from the same family,
PHO5, results in increased production of the mature mRNA [
10]. The increased transcription correlates with changes in the chromatin environment; in this case, however, the antisense RNA probably enhances chromatin plasticity, leading to activation of the gene traversed by the antisense sequence. The regulatory potential of antisense transcripts is highlighted by recent work showing that transcription of murine Xist RNA, which triggers X-chromosome inactivation, is negatively regulated
in cis by an antisense gene [
11].
Not all transcription from protein-coding genes results in full-length mRNA. Chekanova
et al. [
2] observed collinear transcription from the 5' ends of annotated genes, producing what they refer to as upstream noncoding transcripts (UNTs). This transcription appears distinct from that of the 'main' transcription units by RNA polymerase II, and the UNTs accumulate to higher levels than the mature mRNA. Similar transcripts, called promoter-associated short RNAs (PASRs) accumulate in human cells, where their expression correlates strongly with gene transcription [
12]. Human PASR expression also correlates with that from syntenic regions in mouse, suggesting a conserved function for these small RNAs [
12]. Although no experimental evidence for a role for PASRs exists, a clue comes from the study of small RNAs in budding yeast termed cryptic untranslated transcripts (CUTs) [
13,
14]. A CUT sequence called
SRG1 overlaps the promoter of the
SER3 gene and negatively regulates its expression by promoter occlusion [
15]. This glimpse of function and the widespread conservation of small RNAs overlapping the 5' end of genes suggest that further work on these transcripts will uncover novel roles in gene regulation.
The widespread, and apparently regulated, production of UNTs and CUTs suggests that these transcripts have physiological roles in the cell. In this view, the cell may only require appreciable levels of these transcripts at specific times or upon receiving specific stimuli. The cell could then upregulate production of these 'poised' transcripts by increasing recruitment of the transcriptional machinery or possibly by local inhibition of the exosome surveillance pathway. Indeed, in
Schizosaccharomyces pombe, meiosis-specific mRNAs accumulate upon loss of Rrp44 and of Cid14, a homolog of Trf4p, suggesting that the exosome constantly degrades those RNAs to prevent ectopic meiosis [
16]. The absence of accumulation of UNTs, PASRs and CUTs in wild-type cells implies a very limited or tightly regulated role
in vivo. Alternatively, some of this transcription may not result in functional RNA, but instead represents a vestige of nonspecific transcription that provides an organism with the ability to evolve new, functional transcription units [
17].
The work of Chekanova
et al. [
2] also points to an apparent functional specialization among subunits of the exosome and suggests the existence of unique exosome subcomplexes performing specific functions throughout the cell. Is it possible that a 'degradation exosome' or a 'processing exo-some' exists, or that location in the cell dictates an exosomes's function? Evidence for the latter case exists in
Drosophila, where individual exosome subunits have distinct localizations
in vivo [
18]. The inability of the
Arabidopsis RRP44 protein to co-purify with the plant exosome also points to the existence of exosome subcomplexes. The unique transcriptome profile of the
Arabidopsis csl4 mutant and the sub-stoichiometric amounts of the protein in affinity-purified exosomes also suggest the existence of functional exosome subcomplexes. Indeed, in humans a trimeric complex containing Rrp6, the RNA helicase Mtr4 and the RNA-binding protein MPP6 seems to participate in the correct processing of 5.8S rRNA [
19]. The plasticity may not stop there, as the poly(A) polymerases of the TRAMP complex also exhibit differences in specificity [
20]. It is conceivable that such dynamic behavior of the surveillance components evolved to regulate the large number of small RNAs produced by the transcriptome.
The importance of the exosome-dependent surveillance pathway is highlighted by the fact that cells deficient in this pathway accumulate chromosomal abnormalities similar to those observed in cancer patients [
16]. In addition, exosome-deficient yeast cells exhibit a growth defect in the presence of the chemotherapeutic drug 5-fluorouracil (5FU), and exposure of wild-type cells to 5FU results in exosome-enhanced accumulation of polyadenylated noncoding RNAs [
21]. The ubiquitous nature of the exosome makes it an ideal tool for identifying and understanding new classes of RNA. Importantly, the depletion of the individual exosome subunits in
A. thaliana leads to unique RNA profiles and developmental phenotypes [
2]. These new findings suggest a previously unrecognized role for the exosome in regulating the levels of noncoding RNAs that may play critical roles in gene regulation and organismal development.
1,2
