• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. Jun 2004; 24(12): 5391–5403.
PMCID: PMC419869

Histone mRNAs Do Not Accumulate during S Phase of either Mitotic or Endoreduplicative Cycles in the Chordate Oikopleura dioica


Metazoan histones are generally classified as replication-dependent or replacement variants. Replication-dependent histone genes contain cell cycle-responsive promoter elements, their transcripts terminate in an unpolyadenylated conserved stem-loop, and their mRNAs accumulate sharply during S phase. Replacement variant genes lack cell cycle-responsive promoter elements, their polyadenylated transcripts lack the stem-loop, and they are expressed at low levels throughout the cell cycle. During early development of some organisms with rapid cleavage cycles, replication-dependent mRNAs are not fully S phase restricted until complete cell cycle regulation is achieved. The accumulation of polyadenylated transcripts during this period has been considered incompatible with metazoan development. We show here that histone metabolism in the urochordate Oikopleura dioica does not accord with some key tenets of the replication-dependent/replacement variant paradigm. During the premetamorphic mitotic phase of development, expressed variants shared characteristics of replication-dependent histones, including the 3′ stem-loop, but, in contrast, were extensively polyadenylated. After metamorphosis, when cells in many tissues enter endocycles, there was a global downregulation of histone transcript levels, with most variant transcripts processed at the stem-loop. Contrary to the 30-fold S-phase upregulation of histone transcripts described in common metazoan model organisms, we observed essentially constant histone transcript levels throughout both mitotic and endoreduplicative cell cycles.

In eukaryotic cells, DNA replication requires the synthesis of histone proteins to package newly replicated DNA into nucleosomes. When these two events are not coupled, the cell faces severe consequences, such as defects in chromatin assembly, chromosome loss, or the arrest of S-phase progression (17, 18, 20). Histones are characterized as replication-dependent or replacement variants according to their expression during the cell cycle. Replacement variants are expressed throughout the cell cycle, their mRNAs are usually polyadenylated, and they may contain introns. Expression of replication-dependent variants is restricted to S phase, with upregulation at the G1-to-S transition, and a subsequent reduction in transcript levels as S-phase is completed (20). In metazoans, transcripts for these variants are the only mRNAs that do not end in a poly(A) tail. Instead, the 3′ of these mRNAs contains a stem-loop with a highly conserved sequence (7). The 30-fold increase in replication-dependent histone mRNAs at S-phase onset is the result of a 3- to 5-fold stimulation of transcription, an 8- to 10-fold increase in pre-mRNA processing induced by association of the stem-loop binding protein (SLBP), and an extension of the half-life of histone mRNA to 45 to 60 min (10 to 15 min outside S phase) (20). In mammals, the major basis for histone pre-mRNA regulation is cell cycle regulation of the SLBP (32). In plants and yeast, where replication-dependent histone transcripts are polyadenylated, coupling of DNA replication and histone synthesis is achieved mainly through transcriptional regulation (11, 27, 28).

There are occasions when DNA is replicated outside of the canonical G1-S-G2-M cell cycle. An example is the assembly of chromatin during rapid embryonic divisions that occur in the absence of cell growth. During these phases, organisms such as Drosophila melanogaster and Xenopus laevis, maintain the link between chromatin assembly and DNA replication through maternal stocking of histone transcripts and proteins in the developing oocyte (16, 33). In this instance, histone mRNA synthesis is not coupled to DNA replication, but nucleosome assembly remains coupled to the replicative process through posttranscriptional control. In early Drosophila development, the SLBP is still required for proper histone mRNA processing (26). Embryos in which the SLBP is mutated fail to produce histone mRNA with normal 3′ stem-loop ends. Instead, polyadenylated histone transcripts accumulate in the cytoplasm. The phenotypes of SLBP mutants range from female sterility due to failure to accumulate histone mRNA in the oocyte to zygotic lethality late in development. Another important deviation from the standard cell cycle is polyploidy through replication of DNA without cytokinesis. Such endocycles occur in most organisms from protists to humans but, thus far, there is an absence of information on regulation of histone metabolism in this common variation of the cell cycle.

Changes in cell cycles during development are not only characterized by altered kinetics in the synthesis of histones but also by the use of specific histone variants. Distinct sets of transcripts coding for stage-specific core and linker histone variants are associated both with early embryonic cleavage cycles and terminally differentiated tissues (6, 19, 21, 33). In some cases, the use of a specific variant at a specific time is critical, and other members of the same histone subtype are unable to substitute. In Drosophila the H2A variant, H2A.vD, is absolutely required in developing embryos with mutations in this protein being lethal (30, 31). The progressive replacement of linker histone variant H1 M by somatic H1 during Xenopus development is correlated with changes in the expression of specific genes (22) and can play an active role during commitment to the formation of some tissues (25).

The transparent marine urochordate, Oikopleura dioica, allows investigation of the use of histone variants and the regulation of histone metabolism in an organism that undergoes an extensive shift to endoreduplicative cycles during development and growth (Fig. (Fig.1).1). The short life cycle (5 days) begins with rapid mitotic cleavage cycles (5 min) of the 100-μm-diameter fertilized egg. During organogenesis, increasing numbers of cells enter endocycles and after metamorphosis, the organism grows essentially by increasing cell volume via polyploidization. Previously, we showed that the histone gene complement of O. dioica was distinguished from its phylogenetic neighbors by the extent of divergence among its variants (5). As in other organisms, the variants could be divided into groups on the basis of the presence or absence of S-phase-responsive elements in the promoters. Most of the genes featured a strong stem-loop consensus sequence in the 3′-untranslated region (3′UTR) but also contained a polyadenylation signal downstream of the stem-loop. The histone downstream element, implicated in processing of the 3′UTR at the stem-loop in other organisms (7), was absent in O. dioica, and we observed transcripts containing both the stem-loop and a poly(A) tail. Here we show quantitative differences in the developmental stage-specific expression of O. dioica histone variants. During the mitotic phase of development, polyadenylated histone transcripts were abundant. After metamorphosis there was a global downregulation in histone mRNA levels and a loss of polyadenylation for most variants. There was no apparent change in histone density on chromatin throughout the life cycle. Contrary to what has been described for other organisms, there was no extensive S-phase upregulation of histone messages in either the mitotic or endoreduplicative phases of development.

FIG. 1.
Cell cycle transitions during the life cycle of O. dioica. During embryonic cleavage and organogenesis prior to metamorphosis, proliferative mitotic cell cycles predominate. Over the subsequent 4 to 5 days, as the animal grows 10-fold, an increase in ...


Appendicularia culture and sequence accession.

Animals were collected from fjords around Bergen, Norway, and cultured in 6-liter beakers with constant stirring at 14 to 15°C (6-day life cycle) as described previously (5). To sample animals at different developmental stages, they were transferred to watch glasses, rinsed two to three times in sterile-filtered seawater (SFSW), and houses and house rudiments were removed. To collect unfertilized oocytes, mature females were transferred to SFSW and the gonad was burst by gentle pipetting to liberate the oocytes. The free oocytes were then rinsed three to four times in SFSW. For in vitro fertilizations, a 5-ml suspension of sperm was obtained from one to two mature male ejaculates in SFSW. An aliquot of the sperm suspension was then added at a final dilution of 1:100 to oocytes collected from mature females in a watch glass. When 90% of the fertilized oocytes had emitted polar bodies, the embryos were rinsed three times in SFSW. Embryos were then left to develop at room temperature (19 to 21°C), and specific stages were sampled as indicated in Fig. Fig.1.1. Samples were immediately frozen in liquid nitrogen and stored at −80°C. EMBL accession numbers for new sequences reported in the present study are AJ626745 (OdHira), AJ626746 (OdH3.3), AJ626747 (OdH1.3), and AJ626748 to AJ626752 (OdH4.2 to OdH4.6).

Semiquantitative reverse transcription-PCR (RT-PCR).

Total RNA from selected developmental stages was isolated by the guanidium thiocyanate-acid phenol method. First-strand cDNA synthesis was performed by incubating 2 μg of DNase I-treated (PCR grade; Gibco-BRL) total RNA with 100 pmol of random hexamers or 0.5 μg of oligo(dT12-18) primer (Gibco-BRL), 10 mM dithiothreitol, 1 U of RNasin (Promega)/μl, and 0.5 mM deoxynucleoside triphosphates in 50 mM Tris-HCl-75 mM KCl-3 mM MgCl2 (pH 8.3) for 1 h at 37°C in the presence (RT+) or absence (RT) of 400 U of Moloney murine leukemia virus reverse transcriptase (Gibco-BRL). Real-time PCR (LightCycler; Roche) reactions contained cDNA synthesized from an equivalent of 50 ng of total RNA, 2 μl of LightCycler-FastStart DNA Master SYBR Green I (Roche), 0.2 μM concentrations of primers (Table (Table1),1), and 3 to 5 mM Mg2+ (primer dependent [see Table Table1])1]) in a total volume of 20 μl. After initial denaturation for 10 min at 95°C, 35 cycles of 95°C for 5 s, 50 to 70°C (primer dependent [see Table Table1])1]) for 10 s, and 72°C for 5 s were conducted, with a final extension for 5 min at 72°C. RT negative controls were run to 40 cycles of amplification.

Primers specific for O. dioica genes used in this study

Whole-mount immunofluorescence.

Animals were fixed and permeabilized with 4% paraformaldehyde in 0.5 M NaCl-0.1 M morpholinepropanesulfonic acid (pH 7.5)-0.1% Triton X-100 for 1 h at room temperature. Unless otherwise indicated, all washing and equilibration steps in this and the next section were for 20 min. Animals were washed once with phosphate-buffered saline containing 0.1% Tween 20 (PBS-T), twice with PBS-T containing 0.1 M glycine, and twice with PBS-T prior to blocking with 3% acetylated bovine serum albumin (Sigma) in PBS-T at 4°C from 4 h to overnight. Primary antibodies were added at the appropriate dilution in blocking buffer and incubated for 6 to 7 days at 4°C. Samples were then washed repeatedly with PBS-T and postfixed overnight at 4°C with 4% paraformaldehyde in PBS-T. The fixative was washed out and quenched as for the primary fixation. Fab fragment secondary antibodies (Jackson Immunoresearch Laboratories) were applied in blocking buffer, followed by incubation as for primary antibodies. Anti-mouse Fab fragments conjugated with either fluorescein isothiocyanate or Rhodamine Red-X were used for detection of the anti-pan histone antibody. Secondary antibodies were washed extensively and samples were counterstained with 1 μM TO-PRO-3 (Molecular Probes) in PBS-T for 15 min at room temperature. After two washes with PBS-T, samples were mounted on slides with Vectashield mounting medium (Vector Labs). Images were acquired with a Leica TCS-SP confocal microscope by using Leica confocal software.

Whole-mount in situ hybridization combined with BrdU labeling of replicating DNA.

The PCR product of the gene encoding histone OdH4.1 was subcloned into the pCR2.1-TOPO vector (Invitrogen), sequenced, and used as a template for in vitro transcription of sense and antisense RNA probes by using T7 polymerase (Promega) in the presence of digoxigenin-labeled UTP (Roche). Bromodeoxyuridine (BrdU) was diluted in seawater at a final concentration of 1 mM. Animals were exposed to pulses of this solution for 10 min (tadpoles at 4 h), 45 min (tadpoles at 5 to 7 h), or 20 min (animals at day 3). At the end of the pulse animals were fixed and permeabilized as described above. After one wash with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% Tween 20 (2×SSCT), two washes with 2×SSC-T-0.1 M glycine, and two washes with 2×SSCT, the samples were equilibrated once in 2×SSCT-20% formamide, once in 2×SSCT-40% formamide, and twice in 2×SSCT-50% formamide. Samples were then prehybridized (50% formamide, 15% dextran sulfate, 50 μg of carrier RNA/ml, 2×SSCT) for 30 min at 37°C. RNA probes were added to a final concentration of 5 ng/μl and denatured together with the sample for 2 min at 90°C. Hybridizations were carried out at 42°C (16 to 60 h). Samples were then washed six times in 50% formamide-2×SSCT for 15 min at 42°C and once in 25% formamide-2×SSCT for 5 min at room temperature, followed by five washes in 2×SSCT. Samples were blocked with 3% acetylated BSA in 2×SSCT overnight at 4°C. Immunohistochemistry was performed as described above. Digoxigenin was detected with a sheep anti-digoxigenin-FITC-conjugated Fab fragment (1:100 in blocking buffer). The mouse monoclonal anti-BrdU antibody (clone IU-4, 1:100; Accurate Chemicals) was detected by using donkey anti-mouse immunoglobulin G-Rhodamine Red-X (1:200; Jackson Immunoresearch Laboratories). Images were acquired on the Leica TCS-SP confocal microscope.


Developmental expression profiles of histone variants.

To examine the developmental regulation of expression of histone variants in O. dioica, particularly with respect to the shift from mitotic proliferation to endoreduplicative growth, real-time RT-PCR was conducted by using primer pairs specific for histone variants (5) (Fig. (Fig.22 to to6).6). Absence of cross-reactivity of primer pairs among the H2A, H2B, H3, and H1 variants (except H1.1 and H1.2; coamplified because of high nucleotide sequence similarity) was verified by using phage or plasmid DNA containing the respective variants. RT was primed with either oligo(dT) or random hexamers to distinguish polyadenylated histone transcripts from total histone transcripts. The developmental expression profile of the ribosomal protein L23 (RbL23), used as an internal control for all reactions, is shown in Fig. Fig.22.

FIG. 2.
Developmental expression profile of the O. dioica homologue of ribosomal protein RbL23. Total RNA was isolated from the stages described in Fig. Fig.1,1, and cDNAs were synthesized by using either random hexamers (Rhex, [filled square]) or oligo(dT) ...
FIG. 6.
Developmental expression patterns of histone H4 in O. dioica. (A) Total RNA was isolated from the stages described in Fig. Fig.11 and cDNAs synthesized by using either random hexamers ([filled square]) or oligo(dT) (□). The unit value of the ...

Specific developmental expression of different histone variants was observed and could be grouped into three general patterns. The first pattern consisted of genes expressed mainly during early development, starting around 2 to 3 h postfertilization (p.f.), increasing to a peak level at 4 to 6 h, decreasing 5- to 10-fold at day 1 and remaining at low levels during later stages. Expression levels for genes in this group were moderate to abundant and included OdH1.1/2 (Fig. (Fig.3),3), OdH2B1 (Fig. (Fig.4),4), OdH3.1 (Fig. (Fig.5),5), and OdH4.1 (Fig. (Fig.6).6). OdH2A1 was also abundantly expressed during early development with a decrease at day 1 but differed from the pattern of the first group in not showing the same extent of downregulation from days 2 to 5. The second pattern consisted of genes with expression restricted principally to juvenile (days 1 to 3) and/or adult stages (days 4 to 5). The expression levels for these genes could be low (OdH1.3, Fig. Fig.3),3), moderate (OdH2A2 and OdH2A4, Fig. Fig.4),4), or abundant (OdH3.3, Fig. Fig.5).5). Of these genes, OdH3.3 showed the broadest expression range, with significant levels extending prior to metamorphosis during the early tadpole stage. The third group of genes shared the characteristic of very low levels of expression and included OdH2B2 and OdH2B4 (Fig. (Fig.4)4) and OdH3.2 (Fig. (Fig.5).5). Based on their expression profiles, it was possible to correlate the first two groups of histone variants with the two main phases of O. dioica development: the first group with the premetamorphic phase characterized by mitotic divisions and the second group with the postmetamorphic endoreduplicative phase. The weak expression profiles of the third group of histone genes was reminiscent of replacement or tissue-specific variants in other organisms.

FIG. 3.
Expression profiles of O. dioica linker histone variants. (A) Amino acid sequence alignment of the histone H1 variants. Identities are represented by dots, and gaps are indicated by dashes. Motifs for putative phosphorylation sites are underlined with ...
FIG. 4.
Developmental expression profiles for O. dioica H2A and H2B variants. Total RNA was isolated from stages described in Fig. Fig.1,1, and cDNAs were synthesized by using either random hexamers ([filled square]) or oligo(dT) (□). The unit value ...
FIG. 5.
Expression profiles of O. dioica histone H3 variants. (A) Amino acid sequence alignment of the histone H3 variants. Dots represent identities. The four residues in the histone fold distinguishing replacement histone H3.3 from somatic histone H3.1 in all ...

Polyadenylation of O. dioica histone mRNAs is developmentally regulated.

The extent of polyadenylation of histone mRNAs was determined by comparing real-time RT-PCR results obtained on first-strand cDNA primed with random hexamers versus oligo(dT). A first unusual feature was the extensive use of polyadenylation for histone transcripts in O. dioica despite all genes, except OdH3.3, containing a stem-loop consensus sequence in the 3′UTR (5). Among variants of the same histone subtype, those expressed early in development generally showed a preponderant fraction of polyadenylated transcripts. In contrast, variants that expressed postmetamorphosis exhibited limited or no polyadenylation. These trends were observed in comparing the fraction of poly(A)+ mRNA for OdH1.1/2 to OdH1.3 (Fig. (Fig.3)3) and for OdH2A1 transcripts at early stages compared to OdH2A1, OdH2A2, and OdH2A4 transcripts at later stages (Fig. (Fig.4).4). This same developmental profile in polyadenylation of transcripts was also observed when total transcripts for all H4 genes (Fig. (Fig.6)6) were amplified. There were two exceptions to this trend in the use of polyadenylation: the variant OdH2B2 (Fig. (Fig.4)4) and the OdH3 variants (Fig. (Fig.5).5). The trend was reversed for OdH3.1 and OdH3.3: a minor fraction of transcripts for OdH3.1 was polyadenylated early in development, whereas a major proportion of OdH3.3 transcripts were polyadenylated at juvenile and adult stages.

Cytoplasmic adenylation and deadenylation can contribute to the developmental regulation of gene expression. A reduction in poly(A) tail length during development could result in inefficient priming by oligo(dT), leading to a false impression of a shift to processing at the stem-loop. We have previously shown by RNase protection that processing of O. dioica histone H4 transcripts can indeed occur at the stem-loop (5). Here we show by using an additional RT-PCR approach a developmental shift to processing at the stem-loop (Fig. (Fig.7).7). O. dioica H2A1 variants showed a clear reduction in the amount of polyadenylation in juvenile and adult stages (Fig. (Fig.4).4). To test the possibility that this could result from a reduction in poly(A) tail length, we designed primers specific for the H2A1c 3′UTR sequence both prior to and after the stem-loop. The RT-PCR results clearly showed that, in contrast to 3′UTR sequences upstream of the stem-loop, the portion of the 3′UTR downstream of the stem-loop was no longer detected in transcripts of this variant at later developmental stages, indicating that a shift to processing at the stem-loop rather than a simple reduction in poly(A) tail length accounts for the developmental shift in polyadenylation profile.

FIG. 7.
Developmentally regulated polyadenylation of an OdH2A1 variant. (A) Strategy for RT-PCR to distinguish total transcripts (primers a and b) and those containing 3′UTR sequences beyond the stem-loop (primers a to c). (B) The left-hand panel shows ...

Global histone mRNA levels drop as the proportion of endocycling cells increases.

To investigate cell cycle regulation of histone mRNA transcripts during the two different phases of O. dioica development, we used mRNAs coding for OdH4 genes as a marker of the global content of histone transcripts. All nucleosomes contain two H4 molecules, and conserved regions in the coding sequence of all OdH4 genes allowed design of a primer pair that could amplify transcripts from any of them (Fig. (Fig.6B).6B). A general 5- to 10-fold decrease of global OdH4 mRNA levels was detected over the transition between early tadpole (ET) and day 1 animals, with this lower level then maintained until day 5 (Fig. (Fig.6A,6A, lower panel). This result was surprising since, after metamorphosis, the animal grows very rapidly, increasing 10-fold in volume from days 1 to 5. This is mainly achieved by cell growth rather than cell proliferation but involves a concomitant increase in DNA content to respect nucleocytoplasmic ratios. For example, in the epithelium, covering the entire trunk of the animal, endoreduplication of DNA becomes increasingly rapid from days 1 to 5 (9).

One key factor linking histone gene expression to DNA replication is the product of the HIRA gene. Hira acts in part as a transcriptional repressor of replication-dependent histone gene expression (18, 23). To determine whether the downregulation of histone transcripts correlated with any upregulation of HIRA transcription, the 5′ region of the coding sequence for the O. dioica Hira homolog (OdHira) was cloned by PCR (Fig. (Fig.8A)8A) and was used to design primers to carry out semiquantitative RT-PCR throughout development. No correlation was found between the developmental profile of histone H4 mRNA levels (Fig. (Fig.6A)6A) and that of OdHira (Fig. (Fig.8B).8B). OdHira mRNA levels were essentially constant throughout development, with the exception of a brief mild upregulation during early embryonic cleavage cycles.

FIG. 8.
Developmental expression profile of the O. dioica Hira homologue (OdHira). (A) Alignment of the N-terminal regions of the Hira homologues from yeast (Sc, S. cerevisiae), human (Hum, Homo sapiens), chicken (Ch, Gallus gallus), frog (X, X. laevis), O. dioica ...

During the juvenile and adult phases of the O. dioica life cycle, the 5- to 10-fold reduction in the global level of histone transcripts must be reconciled with an increasing DNA content of the organism. Possible explanations include: stocking of histone proteins earlier in development, alteration in the rates of production of histone proteins from a given unit of histone transcripts, or progressive dilution of nucleosomes on the DNA of endocycling cells. The fact that histone synthesis is required for progression of DNA replication and correct chromatin assembly (18) suggests that the latter possibility would be as improbable as it would be novel. With regard to the first possibility, immunostaining with a pan-histone antibody throughout the life cycle revealed no subcellular cytoplasmic stocking of histone proteins at any developmental stage (not shown). To examine the second possibility of an altered histone metabolism, we first attempted quantitative Western blotting with antibodies targeted to the conserved C-terminal regions of histones H3 and H4. By this method, H3 and H4 content could be quantified during all premetamorphic developmental stages (not shown), but no histone bands were detected on immunoblots of protein extracts from days 1 to 5. This was despite the fact that protein bands in the histone molecular weight range were clearly visible by silver staining of electrophoretic gel separations of protein extracts from these stages. Indeed, a panel of antibodies directed against diverse histone modifications worked well for immunofluorescence studies throughout development and also on Western blots of premetamorphic stages but were also completely negative on Western blots for the period from days 1 to 5. Furthermore, higher-molecular-weight proteins, such as the β-actin (50 kDa) internal control, were revealed on immunoblots at all stages. Likely candidates for the problem with specific detection of low-molecular-weight, basic histones on Western blots of postmetamorphic stages were the acidic mucopolysaccharides (24) that are massively secreted from the oikoplastic epithelial cells for repetitive building of the house structure from days 1 to 5. Different attempts at partial purification to relieve blocking of histone detection while still maintaining quantitative recovery of the internal β-actin control were unsuccessful. Although not as quantitative, an alternative approach capable of revealing a 5- to 10-fold reduction in histone density on chromatin was whole-mount immunofluorescence with the anti-pan-histone antibody (Fig. (Fig.9).9). Comparison of the intensity of DNA counterstaining to that of the pan-histone staining yielded no evidence of significantly reduced nucleosomal density on chromatin throughout the O. dioica life cycle. This finding indicates that although global histone transcripts decreased in concentration over the period from days 1 to 5 compared to that observed in pre-tail bud to ET stages, the ratio of histones to DNA was maintained at proportional levels throughout development. This suggests a change in the stability and/or translation rates of histone mRNAs during the endoreduplicative phase.

FIG. 9.
Histone density on chromatin throughout development. Representative immunofluorescence images showing histones detected by an anti-pan-histone monoclonal antibody (Boehringer; green, A to D) with TO-PRO-3 counterstaining of DNA (red; E to H). (A and E) ...

No S-phase increase in histone mRNA was observed in either mitotic or endocycling cells.

In mitotically proliferating cells, replication-dependent histone mRNAs accumulate during S phase with a 30-fold increase at the G1-S transition, whereas replacement histones are expressed at low levels throughout the cell cycle (20). An important basis for cell cycle regulation is the stability and interaction of the SLBP with the stem-loop structure in the 3′UTR of histone transcripts. Since the majority of histone transcripts terminate in the stem-loop during the endoreduplicative phase of the O. dioica life cycle, we investigated whether histone mRNA accumulation occurred during the S phase of endocycles. In parallel, we also studied cycling of histone mRNA levels during developmental phases characterized by the presence of a substantial fraction of poly(A)+ histone transcripts and rapid mitotic divisions.

An RNA probe complementary to the highly conserved coding sequence of histone H4 was used for in situ hybridization studies in concert with BrdU incorporation to identify replicating cells (Fig. (Fig.10).10). The experiment was performed with 4-h tadpoles (H), in which most cells are proliferating mitotically; with 5.5- and 7-h tadpoles (ET), in which mitotic and endocycling cells are both represented; and on day 3 animals; in which the majority of the tissues have entered endocycles. Surprisingly, at 4 h (Fig. 10A), 5.5 h (Fig. 10B), and 7 h (Fig. 10C), histone mRNA was never enriched in cells in S phase. Instead, histone H4 mRNA levels were quite homogeneous, regardless of the point in the cell cycle at the moment of the BrdU pulse. Even cells in which mitotic figures were visible (Fig. 10A) had a histone H4 mRNA content similar to cells replicating DNA. In Fig. 10C (7-h tadpole), only two endocycling cells of the field of Eisen were in S phase, but no difference in histone mRNA content was observed compared to the other five nonreplicating cells in the field. The lack of modulation of histone H4 mRNA levels with respect to phase of the cell cycle was also observed in day 3 animals in both mitotically proliferating (e.g., gonad) and endocycling (e.g., epithelium) cells. Thus, independently of whether polyadenylated transcripts predominated in early proliferating cells or stem-loop transcripts were preponderant in later endocycling cells, we observed nothing similar to the 30-fold upregulation of histone transcripts associated with the S phase in other organisms (20). We also observed that the in situ signal for histone H4 mRNA was reduced in day 3 animals compared to developing tadpoles. This may in part be explained by dilution of mRNA in the larger cytoplasmic volume of polyploid cells, but it also fits with the 5- to 10-fold reduction in mRNA levels observed in the RT-PCR profiles described above. The length of the S phase is considerably shorter during the cleavage stage and ET development (minutes to tens of minutes) than it is during endocycles (hours), and therefore differential rates of DNA replication may also impact on the respective cellular levels of histone mRNA.

FIG. 10.
Histone H4 mRNA levels in replicating versus nonreplicating cells during development. Whole-mount in situ RNA hybridization was combined with detection of BrdU incorporation. Transcripts for histone H4 are in green, DNA counterstaining is in blue, and ...


In higher eukaryotes, the classification of histones as replication-dependent or replacement variants is related to two fundamental processes in the life of a eukaryotic cell: (i) the need for rapid bulk packaging of newly replicated DNA into chromatin and (ii) the requirement to modulate nucleosome structure for gene expression, DNA or nucleosome repair, and chromosome movement and compartmentalization. The kinetics of these two types of processes can be quite different and, as outlined in the introduction, this is reflected in both the structural features of the corresponding genes and in their metabolic processing. The data presented here make it difficult to apply several important features of this convenient classification to the chordate O. dioica.

The histone genes expressed during the early mitotic phase of O. dioica development resemble the replication-dependent variants described in other organisms in that they contain S-phase-responsive promoter elements such as E2F, they lack introns, and they contain a stem-loop consensus sequence in their 3′UTR. These genes were principally expressed between 2 and 7 h p.f., peaking in early tadpoles, before declining to reduced or very limited expression during growth after metamorphosis. In view of their similarities to replication-dependent variants, as described above, an unexpected characteristic was that a significant fraction of their transcripts were polyadenylated. Extensive use of polyadenylation of histone transcripts has been described in mitotic cycles of lower eukaryotes and plants (4, 11, 20), where histone genes may contain introns and lack the stem-loop consensus, but in insects and higher chordates, replication-dependent variants are generally not polyadenylated and terminate in the stem-loop. In the latter groups of organisms, polyadenylated transcripts that also contain the stem-loop sequence have been noted, but they comprise a very minor fraction of the histone mRNA pool (15). Therefore, the early histone genes of O. dioica do not fit the schema in that they are expressed in actively dividing cells as abundant polyadenylated transcripts containing the 3′UTR stem-loop sequence.

There was a second group of histone variants whose major expression was restricted to the endoreduplicative phase of O. dioica development. This included genes coding for OdH1.3, OdH2A2, OdH2A4, OdH3.3, and OdH2B3. None of these genes had known cell cycle-responsive promoter elements. With the exception of OdH3.3, they did not have introns and did contain a 3′ stem-loop consensus sequence. In O. dioica all endoreduplicating cells are terminally differentiated. In metazoans, only replacement variants are expressed as polyadenylated transcripts in differentiated cells. However, in O. dioica, again with the exception of H3.3, the transcripts for this group of late genes were not significantly polyadenylated and were processed through the stem-loop, a characteristic of replication-dependent variants in other animals. In general then, O. dioica histone mRNAs are processed preferentially through polyadenylation during the early mitotic phase and through the stem-loop sequence during the later endoreduplicative phase of O. dioica development. A clear and intriguing exception to this is the processing of genes encoding the H3 variants (Fig. (Fig.55).

Although expressed early in development, transcripts for OdH3.1 were not polyadenylated and, during the endoreduplicative phase, the abundantly expressed OdH3.3 variant was polyadenylated, in contrast to mRNAs of the other histone subtypes. Recently, it has been shown in D. melanogaster that incorporation of different variants of histone H3 into nucleosomes depends on distinct pathways of chromatin assembly (1, 2). The replication-dependent variant of H3, expressed as transcripts terminating in the 3′ stem-loop, is assembled into chromatin only during S phase by a replication-coupled assembly pathway. It does not deposit onto DNA if expressed at other cell cycle phases. The other two Drosophila H3 variants, H3.3 and Cid, are expressed as polyadenylated mRNAs (3, 10). H3.3 is expressed at constant levels in differentiating cells that have exited the cell cycle (14) and is incorporated into chromatin by a replication-independent assembly pathway (2). Cid, the centromere-specific variant, is incorporated by a replication-independent pathway distinct from that of H3.3 and can be assembled both during the replication of centromeres and for centromere repair after mitosis. It is the protein sequence of Drosophila H3 that specifies its replication-coupled assembly, and both H3 homodimers and H3-H3.3 heterodimers can be assembled into chromatin during DNA replication. The lack of polyadenylation of the somatic H3.1 variant of O. dioica may then make sense. Since the protein produced from H3.1 transcripts can only be assembled during DNA replication by a replication-coupled mechanism, translation of poly(A)-stabilized transcripts outside of S phase would appear to be of little utility. Once O. dioica cells enter endocycles, the major function of the centromere in chromosome segregation at mitosis is no longer required. On the other hand, polyploid cells are generally highly metabolically active, as the 10-fold growth of O. dioica within 4 days attests. High rates of metabolism imply high rates of transcription and, since H3.3 is known to be preferentially incorporated at transcriptionally active loci, such as the ribosomal gene repeats in Drosophila, the abundant expression of the H3.3 variant during the endocycling phase of O. dioica development would seem logical.

To examine more globally the metabolism of histone transcripts throughout the life cycle, we focused on H4 transcripts. This is because all nucleosomes contain H4, it is the most conserved histone subtype, and it was possible therefore to design common reagents to evaluate transcripts for all H4 encoding genes. The developmental profile for general H4 mirrored that for many individual variants of the diverse histone subtypes described earlier. This included significant processing by polyadenylation during the early mitotic phase of development, a 5- to 10-fold downregulation in global H4 transcript levels during endocycles after metamorphosis, and a shift to a preponderance of nonpolyadenylated transcripts during these latter stages. Among the most striking features of global H4 transcript levels was the lack of detectable S-phase accumulation of transcripts in either the early proliferative phase or later endocycles (Fig. (Fig.1010).

Accumulation of histone transcripts during S phase in a number of organisms suggests that there has been selective pressure in evolution to restrict the bulk of histone synthesis to S phase (16). This may also be reflected in a number of organisms by the selective maintenance of large clusters of histone genes facilitating coordinate transcription and pre-mRNA processing. Studies in the yeast, Saccharomyces cerevisiae (23), have demonstrated that uncoupling histone transcription from the cell cycle is not necessarily lethal, but it has been suggested that continuous histone expression throughout the cell cycle, particularly with accumulation of polyadenylated transcripts, is not compatible with metazoan development (16). A common exception to this is the uncoupling of histone mRNA levels and DNA replication during embryonic development, particularly in species, such as Xenopus and Drosophila, with very rapid cleavage cycles. In these instances, tight coupling seems to be established when a full cell cycle with G1-S regulation appears (13). The structure of the histone mRNAs during these phases is the same as that for the replication-dependent histones with transcripts terminating in a stem-loop. Despite very rapid early development, O. dioica does not contain large clusters of histone repeats (5). The histone genes expressed during this period have promoter elements and gene structures that share features with replication-dependent variants in other organisms but, in contrast, they were significantly polyadenylated. Even as cell cycles slowed and became fully regulated during larval tadpole development, polyadenyated transcripts persisted and transcript levels were not modulated throughout the cell cycle. Nonetheless, this continuous level of histone mRNA throughout the cell cycle, with accumulation of polyadenylated transcripts, is clearly compatible with the development of this complex metazoan. Even after entry into endocycles, there was no apparent cell cycle modulation of histone transcript levels, although most variant transcripts were then processed principally through the DNA replication-associated stem-loop. Recently, it has been demonstrated that the SLBP is not necessary for the degradation of stem-loop transcripts, since histone mRNA degradation occurs independently of levels of SLBP. The observation that histone mRNAs are degraded as a consequence of induced arrest of DNA replication, whereas SLBP is not, has led to the hypothesis that although SLBP is regulated by cell cycle signals, histone mRNA levels respond to changes in DNA replication rate (34). It has been proposed that the mechanism linking rates of DNA synthesis to histone mRNA levels involves monitoring of histone protein occupancy on chromatin assembly factors (12). In endocycling cells in O. dioica, the rate of DNA replication appears to be insufficient to control histone mRNA degradation, since these cells have as much histone mRNA in the gap phase as in S phase. This raises the possibility that histone mRNA degradation may also be regulated by cell cycle signals that are absent in O. dioica endocycles.

It has been suggested that the biological rationale for linking bulk histone biosynthesis to S phase might be related to the need for orderly propagation of epigenetic information through serial cell cycles in development (16). In this regard, endocycles are a substantial simplification of mitotic cell cycles. The combination of epigenetic information required to transmit the memory of transcriptional states and structural information through chromosome condensation and segregation at mitosis would no longer be required in endocycles. This speculative idea would then be consistent with a more relaxed cell cycle regulation of histone biosynthesis in endocycles. What is strikingly evident, however, is that several important facets of the replication-dependent/replacement histone variant paradigm established in common model metazoans do not apply to the metabolism of histones in the chordate O. dioica.


This study was supported by grants from the Norwegian Research Council and Ministry of Education and by NFR Biotechnology grant 146653/431.


1. Ahmad, K., and S. Henikoff. 2002. Histone H3 variants specify modes of chromatin assembly. Proc. Natl. Acad. Sci. USA 99:16477-16484. [PMC free article] [PubMed]
2. Ahmad, K., and S. Henikoff. 2002. The histone variant H3.3 marks active chromatin by replication-independent nucleosome assembly. Mol. Cell 9:1191-1200. [PubMed]
3. Akhmanova, A. S., P. C. Bindels, J. Xu, K. Miedema, H. Kremer, and W. Hennig. 1995. Structure and expression of histone H3.3 genes in Drosophila melanogaster and Drosophila hydei. Genome 38:586-600. [PubMed]
4. Chaboute, M. E., N. Chaubet, B. Clement, C. Gigot, and G. Philipps. 1988. Polyadenylation of histone H3 and H4 mRNAs in dicotyledonous plants. Gene 71:217-223. [PubMed]
5. Chioda, M., R. Eskeland, and E. M. Thompson. 2002. Histone gene complement, variant expression, and mRNA processing in a urochordate Oikopleura dioica that undergoes extensive polyploidization. Mol. Biol. Evol. 19:2247-2260. [PubMed]
6. Dimitrov, S., G. Almouzni, M. Dasso, and A. P. Wolffe. 1993. Chromatin transitions during early Xenopus embryogenesis: changes in histone H4 acetylation and in linker histone type. Dev. Biol. 160:214-227. [PubMed]
7. Dominski, Z., and W. F. Marzluff. 1999. Formation of the 3′ end of histone mRNA. Gene 239:1-14. [PubMed]
8. Flood, P., and D. Deibel. 1998. The appendicularian house, p. 105-124. In Q. Bone (ed.), The biology of pelagic tunicates. Oxford University Press, New York, N.Y.
9. Ganot, P., and E. M. Thompson. 2002. Patterning through differential endoreduplication in epithelial organogenesis of the chordate, Oikopleura dioica. Dev. Biol. 252:59-71. [PubMed]
10. Henikoff, S., K. Ahmad, J. S. Platero, and B. van Steensel. 2000. Heterochromatic deposition of centromeric histone H3-like proteins. Proc. Natl. Acad. Sci. USA 97:716-721. [PMC free article] [PubMed]
11. Kapros, T., A. J. Robertson, and J. H. Waterborg. 1995. Histone H3 transcript stability in alfalfa. Plant Mol. Biol. 28:901-914. [PubMed]
12. Kaufman, P. D., R. Kobayashi, N. Kessler, and B. Stillman. 1995. The p150 and p60 subunits of chromatin assembly factor 1: a molecular link between newly synthesized histones and DNA replication. Cell 81:1105-1114. [PubMed]
13. Lanzotti, D. J., H. Kaygun, X. Yang, R. J. Duronio, and W. F. Marzluff. 2002. Developmental control of histone mRNA and dSLBP synthesis during Drosophila embryogenesis and the role of dSLBP in histone mRNA 3′ end processing in vivo. Mol. Cell. Biol. 22:2267-2282. [PMC free article] [PubMed]
14. Lennox, R. W., and L. H. Cohen. 1988. The production of tissue-specific histone complements during development. Biochem. Cell Biol. 66:636-649. [PubMed]
15. Levine, B. J., N. Chodchoy, W. F. Marzluff, and A. I. Skoultchi. 1987. Coupling of replication type histone mRNA levels to DNA synthesis requires the stem-loop sequence at the 3′ end of the mRNA. Proc. Natl. Acad. Sci. USA 84:6189-6193. [PMC free article] [PubMed]
16. Marzluff, W. F., and R. J. Duronio. 2002. Histone mRNA expression: multiple levels of cell cycle regulation and important developmental consequences. Curr. Opin. Cell Biol. 14:692-699. [PubMed]
17. Meeks-Wagner, D., and L. H. Hartwell. 1986. Normal stoichiometry of histone dimer sets is necessary for high fidelity of mitotic chromosome transmission. Cell 44:43-52. [PubMed]
18. Nelson, D. M., X. Ye, C. Hall, H. Santos, T. Ma, G. D. Kao, T. J. Yen, J. W. Harper, and P. D. Adams. 2002. Coupling of DNA synthesis and histone synthesis in S phase independent of cyclin/cdk2 activity. Mol. Cell. Biol. 22:7459-7472. [PMC free article] [PubMed]
19. Newrock, K. M., C. R. Alfageme, R. V. Nardi, and L. H. Cohen. 1978. Histone changes during chromatin remodeling in embryogenesis. Cold Spring Harbor Symp. Quant. Biol. 42:421-431. [PubMed]
20. Osley, M. A. 1991. The regulation of histone synthesis in the cell cycle. Annu. Rev. Biochem. 60:827-861. [PubMed]
21. Poccia, D., J. Salik, and G. Krystal. 1981. Transitions in histone variants of the male pronucleus following fertilization and evidence for a maternal store of cleavage-stage histones in the sera urchin egg. Dev. Biol. 82:287-296. [PubMed]
22. Sera, T., and A. P. Wolffe. 1998. Role of histone H1 as an architectural determinant of chromatin structure and as a specific repressor of transcription on Xenopus oocyte 5S rRNA genes. Mol. Cell. Biol. 18:3668-3680. [PMC free article] [PubMed]
23. Sherwood, P. W., S. V. Tsang, and M. A. Osley. 1993. Characterization of HIR1 and HIR2, two genes required for regulation of histone gene transcription in Saccharomyces cerevisiae. Mol. Cell. Biol. 13:28-38. [PMC free article] [PubMed]
24. Spada, F., H. Steen, C. Troedsson, T. Kallesoe, E. Spriet, M. Mann, and E. M. Thompson. 2001. Molecular patterning of the oikoplastic epithelium of the larvacean tunicate Oikopleura dioica. J. Biol. Chem. 276:20624-20632. [PubMed]
25. Steinbach, O. C., A. P. Wolffe, and R. A. Rupp. 1997. Somatic linker histones cause loss of mesodermal competence in Xenopus. Nature 389:395-399. [PubMed]
26. Sullivan, E., C. Santiago, E. D. Parker, Z. Dominski, X. Yang, D. J. Lanzotti, T. C. Ingledue, W. F. Marzluff, and R. J. Duronio. 2001. Drosophila stem-loop binding protein coordinates accumulation of mature histone mRNA with cell cycle progression. Genes Dev. 15:173-187. [PMC free article] [PubMed]
27. Tanimoto, E. Y., T. L. Rost, and L. Comai. 1993. DNA replication-dependent histone H2A mRNA expression in pea root tips. Plant Physiol. 103:1291-1297. [PMC free article] [PubMed]
28. Taoka, K., N. Ohtsubo, Y. Fujimoto, K. Mikami, T. Meshi, and M. Iwabuchi. 1998. The modular structure and function of the wheat H1 promoter with S phase-specific activity. Plant Cell Physiol. 39:294-306. [PubMed]
29. Thompson, E. M., T. Kallesoe, and F. Spada. 2001. Diverse genes expressed in distinct regions of the trunk epithelium define a monolayer cellular template for construction of the oikopleurid house. Dev. Biol. 238:260-273. [PubMed]
30. van Daal, A., and S. C. Elgin. 1992. A histone variant, H2AvD, is essential in Drosophila melanogaster. Mol. Biol. Cell 3:593-602. [PMC free article] [PubMed]
31. van Daal, A., E. M. White, M. A. Gorovsky, and S. C. Elgin. 1988. Drosophila has a single copy of the gene encoding a highly conserved histone H2A variant of the H2A. F/Z type. Nucleic Acids Res. 16:7487-7497. [PMC free article] [PubMed]
32. Whitfield, M. L., L. X. Zheng, A. Baldwin, T. Ohta, M. M. Hurt, and W. F. Marzluff. 2000. Stem-loop binding protein, the protein that binds the 3′ end of histone mRNA, is cell cycle regulated by both translational and posttranslational mechanisms. Mol. Cell. Biol. 20:4188-4198. [PMC free article] [PubMed]
33. Wolffe, A. 1998. Chromatin structure and function, 3rd ed. Academic Press, Inc., San Diego, Calif.
34. Zheng, L., Z. Dominski, X.-C. Yang, P. Elms, C. S. Raska, C. H. Borchers, and W. F. Marzluff. 2003. Phosphorylation of stem-loop binding protein (SLBP) on two threonines triggers degradation of SLBP, the sole cell cycle-regulated factor required for regulation of histone mRNA processing, at the end of S phase. Mol. Cell. Biol. 23:1590-1601. [PMC free article] [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...