Degradation of a Nonpolyadenylated Messenger: Histone mRNA Decay

The Response of Renal Tubular Epithelial Cells to Physiologically and Chemically Induced Growth Arrest

“…Because the H3/H4 histone tetramer forms the central core of the nucleosome, with H2A/H2B heterodimers capping each end of the core, H2A/H2B dimers are more readily displaced from the nucleosomal core than H3 and H4 (35). Histone mRNAs are rapidly degraded via the activation of a specific mRNA destabilization process (36) that is autoregulated by histone proteins and initiated by increases in cytoplasmic histones (37). Both core and linker histones induce the destabilization of histone H4 mRNA (38).…”

Specificities of Caenorhabditis elegans and human hairpin binding proteins for the first nucleotide in the histone mRNA hairpin loop

“…We therefore concentrated our analysis on the hairpin itself+ The hairpin RNA sequence is highly elegans RNA hairpins+ The indicated in vitro-synthesized RBDs (4+4 nM) were incubated with increasing amounts of 32 P-labeled mmHPs or ceHPs RNAs as described in Materials and Methods+ The protein-RNA complexes were separated from free RNA by nondenaturing gel electrophoresis (EMSA) and the amount of complexed RNA was determined by PhosphorImager+ Plotted is the amount of RNA bound at each RNA concentration versus the RNA concentration, allowing for a direct reading of the dissociation constant K d (whose values are listed in Table 1)+ Experiments were done at least in duplicate, and shown are the means, flanked by the highest and lowest values at each RNA concentration+ Symbols used for the two RNAs are defined in the insert of A+ Figure 6 and from two further identical experiments were quantitated by PhosphorImager and the average ratios of mmHPs complexes to ceHPs complexes are listed+ b Dissociation constants and maximal binding activities were read from Figure 4 (hs and ce RBD) and from similar graphs (mut3, mut 7, and mut9)+ Where no saturation was reached within the range of RNA concentrations tested, a minimal value for K d is listed in brackets and the maximal binding is given as nd+ conserved+ One difference between nematode and vertebrate histone hairpins is the nucleotide at position 1 of the U-rich 4-nt loop+ In vertebrates, it is invariably a U; in contrast, C. elegans is the only exception among all metazoans where a C is found at this first position (Marzluff, 1992;Marzluff & Hanson, 1993;Wittop Koning & Schümperli, 1994)+ To test whether a U at this position inhibits ce RBD binding, we replaced it by a U in the ceHP sequence, to produce ceHP-Uloop RNA (Fig+ 2)+ Whereas the hs RBD was able to bind to this RNA, complex formation with the ce RBD was greatly reduced to ;5-10% of complexes formed with ceHP RNA (Fig+ 5A, lanes 14 and 15; Fig+ 5B)+ This indicated that a C at the 59 position of the histone hairpin loop is crucial for the binding of the C. elegans RBD, whereas the human RBD can bind to the hairpin irrespective of the kind of pyrimidine present at this position+ A second difference between the vertebrate and C. elegans hairpins is the nucleotide at the fourth position of the hairpin loop+ All C. elegans histone genes have a U, whereas vertebrate histone genes can have any nucleotide, but most often have a C at this position+ To test whether this nucleotide is important for the sequence discrimination of the C. elegans RBD, we replaced it by a C in ceHP RNA, to produce ceHPCUUCloop RNA (Fig+ 2)+ Both hs and ce RBDs were able to form complexes with this RNA (Fig+ 5A, lanes 17 and 18; Fig+ 5B), indicating that this residue was not important for sequence discrimination+ For the human RBD, ceHP-CUUCloop RNA (Fig+ 5A, lane 17) was the least efficient substrate (;75% binding compared to ceHP RNA; Fig+ 5B), indicating that replacing the U at position 1 in the loop by a C also slightly affected RNA binding+ Identification of amino acids in the human and C. elegans RBD contributing to RNA binding specificity To identify which features of the RBD are involved in the recognition of the critical U or C nucleotides in the RNA hairpin loop, we tested whether replacing specific amino acids with the corresponding C. elegans ones would render the human RBD more selective for C. elegans hairpin RNA+ Based on the alignment of the five known HBP sequences shown in Figure 1, a consensus sequence of the RBD was deriv...…”

“…The binding target for HBP, the hairpin structure at the 39 end of animal histone mRNAs, has been highly conserved in evolution (Marzluff, 1992;Marzluff & Hanson, 1993;Wittop Koning & Schümperli, 1994;)+ However, the sequences at the 39 end of C. elegans histone mRNAs show a number of important deviations from those of other metazoans+ First, not only the hairpin and the following 4 nt, but also 12 nt preceding the hairpin are virtually invariant among C. elegans histone genes (Wittop Koning & Schümperli, 1994)+ In comparison, the conserved region preceding the hairpin is shorter and more variable in vertebrates (the main characteristic of this region being the absence of Gs)+ Second, the sequence flanking the hairpin on the 39 side is ACA A / U in C. elegans, ACCA in sea urchins and ACCCA in vertebrates+ Moreover, the C. elegans histone hairpins all have a C at the first position of the loop, whereas an invariant U is found at this position in all other metazoans+ Finally, C. elegans seems to lack a wellconserved downstream or spacer element+ The best candidate so far is the sequence A / U AAUCC (Wittop Koning & Schümperli, 1994), which has, however not been confirmed experimentally, since the C. elegans U7 snRNA, supposed to interact with this sequence, has not yet been identified+ Figure 2+ The RNA-protein complexes were quantitated from three independent experiments by PhosphorImager+ The experiment indicates that the new selectivity of both mutant RBDs is mostly for the pyrimidine at the first position in the hairpin loop+ Here we have exploited the availability of cloned cDNAs for human and C. elegans HBP to analyze if the two proteins display similar or different RNA binding specificities+ Interestingly, the C. elegans HBP bound very selectively to its own, but not to a vertebrate hairpin, and this high selectivity was preserved when only the C. elegans RBD was analyzed (Fig+ 3)+ This selectivity was not the result of misfolding of the C. elegans RBD, as it bound to ceHPs RNA with similar affinity as the human RBD (Fig+ 4) and because the binding selectivity was observable over a wide range of salt concentrations (Fig+ 7)+ We further showed that the nucleotide at the first position was mostly responsible for the exclusive binding of the C. elegans HBP and its RBD to its cognate RNA hairpin and that this nucleotide had to be a C (Fig+ 5)+ In contrast, exchanging the sequences flanking the hairpin only slightly reduced the binding of the C. elegans RBD+ This highly selective binding behavior of the C. elegans HBP and its RBD contrasted with that of human HBP+ Both human HBP and its RBD bound about equally well to all the mammalian, C. elegans, and chimeric hairpins analyzed (Figs+ 3 and 5)+ Previous studies had shown that the sequence of the stem and the sequences immediately flanking the hairpin were critical determinants for mammalian HBP binding but that the conformation of the loop was not critical (Pandey et al+, 1991(Pandey et al+, , 1994Williams & Marzluff, 1995)+ However, when the conserved uridine at the first position of the loop was mutated to guanosine, the resulting RNA showed only 15% binding to HBP present in mouse myeloma nuclear extract (Williams & Marzluff, 1995)+ We recently found a similar but slightly weaker reduction in binding of recombinant human HBP to the same hairpin mutation …”

“…The sequence of the hairpin element in vertebrate and C. elegans histone genes varies both in the loop and in the sequences flanking the hairpin+ One difference in the loop is the nucleotide at position 1 of the U-rich 4-nt loop+ In all other species, including humans, there is a U at this position, whereas in C. elegans it is a C (Marzluff, 1992;Marzluff & Hanson, 1993;Wittop Koning & Schümperli, 1994)+ The situation is reversed at the 39 end of the loop where the C. elegans hairpin has a U, whereas in other species, this nucleotide is less well conserved but most often a C+ Human, mouse, Xenopus, and C. elegans HBPs have been identified that bind specifically to this histone RNA hairpin element (Wang et al+, 1996(Wang et al+, , 1999Martin et al+, 1997)+ An ;73-amino-acid minimal RNA-binding domain (RBD) of HBP has been characterized by deletion analysis (Wang et al+, 1996)+ This RBD has no obvious primary sequence homology with other RNA-binding motifs characterized so far (Burd & Dreyfuss, 1994;Draper, 1995Draper, , 1999)+ The RBD also encompasses a short 67-amino-acid region (residues 9-75 in Fig+ 1) that is the most conserved part among the known animal HBPs (Wang et al+, 1996(Wang et al+, , 1999Martin et al+, 1997)+ In this region, C. elegans and human HBP are 55% identical, whereas the overall sequence identity between these two proteins is 38%+ This is lower than the ;63% identity in the RBD between the human HBP and Xenopus SLBP2, the most distantly related vertebrate HBPs+ Here we have determined experimentally whether these differences between C. elegans and human HBPs are reflected by and contribute to a selective RNA binding+ These experiments revealed that the C. elegans HBP binds with great preference to C. elegans hairpin RNA, and that the C. elegans RBD is responsible for this preference+ Human HBP, on the other hand, as well as the human RBD, did not show a strong preference for either vertebrate or C. elegans hairpin RNA+ Binding reactions with mutant hairpin RNAs demonstrated that the first nucleotide in the loop is an important specificity determinant for the binding of the C. elegans RBD to its cognate RNA+ To identify amino acid sequence elements involved in this binding specificity, we have changed single or multiple residues in the human RBD to the corresponding C. elegans sequence and tested these chimeric proteins for RNA binding+ This approach enabled us to identify two amino acid segments of HBP that participate in RNA binding through interaction with specific nucleotides in the hairpin loop+…”

“…Replication-dependent histone mRNAs are not polyadenylated, but are formed at the 39 end by an endonucleolytic cleavage of the primary transcripts+ This nuclear cleavage event requires two cis-acting elements: a purine-rich sequence (or spacer element) downstream of the cleavage site that is recognized by the U7 small nuclear ribonucleoprotein (snRNP) through RNA-RNA base pairing (Mowry & Steitz, 1987a;Cotten et al+, 1988;Soldati & Schümperli, 1988;Bond et al+, 1991) and a conserved 26-nt sequence containing a hairpin that interacts with a factor called hairpin-binding protein (HBP) or stem-loop-binding protein (SLBP) (Mowry & Steitz, 1987b;Vasserot et al+, 1989;Melin et al+, 1992;Wang et al+, 1996;Martin et al+, 1997)+ The cleavage of histone pre-mRNA requires a third factor, termed heat-labile factor (HLF) that has not been well characterized (Gick et al+, 1987)+ Besides histone RNA 39 processing, the hairpin is also essential for nuclear export, translation, and stability regulation of histone mRNAs (reviewed in Marzluff, 1992;Marzluff & Hanson, 1993;Wittop Koning & Schümperli, 1994;)+ As HBP is also present on polysomeassociated histone mRNA (Pandey et al+, 1991;Dominski et al+, 1995), it is thought to control most, if not all, posttranscriptional aspects of histone gene expression and regulation+…”