The organisation and structure of C. elegans
histone genes is similar to histone genes in other metazoans. The conserved hairpin element in the 3' UTR, the core element of the binding site for HBP/SLBP, is clearly present, and is followed by a conserved element that may serve as the U7 snRNP binding site HDE (Fig. and additional file 1
). However all C. elegans
histone genes have a conserved potential polyadenylation signal immediately downstream of the HDE, a feature not shared with histone genes from other organisms. We have used three independent approaches to analyse the structure of histone mRNA 3' ends. They reveal that histone mRNAs end after the hairpin structure, indicating that 3' end processing of C. elegans
histone mRNAs occurs via cleavage immediately after the hairpin, and that cleavage following the putative polyadenylation signal occurs rarely, if at all. First, we found that C. elegans
histone mRNAs were lost during selection of poly(A) mRNA using oligo-dT cellulose, similar to HeLa cell histone mRNA known to end after the hairpin and to lack a poly(A) tail (Fig. ). Next, we investigated the structure of histone mRNA ends using an RNase protection assay designed to distinguish between RNAs ending after the hairpin, and RNAs ending after the polyadenylation signal (Fig. ). The protection patterns obtained indicate that the vast majority of transcripts of the histone genes analysed (his-9, his-13
) end after the hairpin signal. Finally, cloning and sequencing of synthetically poly(I/A) tailed histone mRNAs showed that none of the mRNAs included the putative HDE element or polyadenylation signal (Table ). Instead this revealed that histone mRNAs end three to six nucleotides after the hairpin structure and, as they are a substrate for poly(A) polymerase, have a 3'OH group. Thus our data indicate that the structure of C. elegans
histone mRNA 3' ends is similar to the structure of histone mRNA 3' ends in vertebrates and Drosophila
], and is typical for histone-specific mRNA 3' end processing mediated by the U7snRNP and associated factors. It will be interesting to analyse the function of the conserved AATCC element. We believe that this is the core motif of a U7 snRNP binding site. We will exploit this to search for U7 snRNA, using bioinformatics and experimental approaches.
It is known that polyadenylation signals contribute to transcription termination by changing the properties of the passing RNA polymerase [33
]. Perhaps, in C. elegans
replication-dependent histone genes the polyadenylation signal functions as transcription terminator rather then as an RNA processing site. We also cannot currently exclude the possibility that, instead of histone-specific processing, the mRNA 3' ends are formed by cleavage after the polyadenylation signal, followed by 3'-5' exonucleolytic trimming of the pre-mRNA up to the hairpin element.
We previously described that cdl-1(RNAi)
results in the inhibition of histone synthesis [13
]. Here we analyse this in more detail and show that cdl-1(RNAi)
does not significantly affect histone mRNA levels, but, as shown previously, severely inhibits histone protein expression (Fig. ). This indicates that CDL-1 functions similar to HBP/SLBP in mammals and in Drosophila
. In these animals, HBP/SLBP controls post-transcriptional regulation of histone gene expression including mRNA 3' end processing and translation. It will be interesting to determine whether similar to the role of HBP/SLBP in other organisms, CDL-1 is required for histone mRNA 3' end processing, translation and perhaps also mRNA stability control in C. elegans
We have shown that during C. elegans development, histone mRNA levels positively correlate with mitotically active tissues (Fig. ). Thus, as in other organisms, C. elegans histone gene expression is cell cycle regulated. In wild-type embryos we detected a two-fold reduction of histone mRNA levels, relative to 18S rRNA, in L3-stage animals compared to L1 larvae, presumably reflecting the fact that proportionally more cells are proliferating in L1 larvae compared to L3 larvae. Relative histone mRNA levels were again increased in adult N2 animals, which was expected since the adult germ line is mitotically active, and most adults would contain a number of highly proliferating early embryos. This pattern was independent of temperature and was similar at 16°C and 25°C. The fact that no histone mRNAs could be detected from glp4(bn2) adult animals grown at 25 °C, which essentially lack germ line proliferation when grown at the restrictive temperature, shows that in adults most, if not all histone gene expression is confined to the germ line. Thus our findings indicate that the link between DNA synthesis and histone gene expression observed in mammals and Drosophila is also the case for C. elegans.