Linker histone partial phosphorylation: effects on secondary structure and chromatin condensation

Linker histones are involved in chromatin higher-order structure and gene regulation. We have successfully achieved partial phosphorylation of linker histones in chicken erythrocyte soluble chromatin with CDK2, as indicated by HPCE, MALDI-TOF and Tandem MS. We have studied the effects of linker histone partial phosphorylation on secondary structure and chromatin condensation. Infrared spectroscopy analysis showed a gradual increase of β-structure in the phosphorylated samples, concomitant to a decrease in α-helix/turns, with increasing linker histone phosphorylation. This conformational change could act as the first step in the phosphorylation-induced effects on chromatin condensation. A decrease of the sedimentation rate through sucrose gradients of the phosphorylated samples was observed, indicating a global relaxation of the 30-nm fiber following linker histone phosphorylation. Analysis of specific genes, combining nuclease digestion and qPCR, showed that phosphorylated samples were more accessible than unphosphorylated samples, suggesting local chromatin relaxation. Chromatin aggregation was induced by MgCl₂ and analyzed by dynamic light scattering (DLS). Phosphorylated chromatin had lower percentages in volume of aggregated molecules and the aggregates had smaller hydrodynamic diameter than unphosphorylated chromatin, indicating that linker histone phosphorylation impaired chromatin aggregation. These findings provide new insights into the effects of linker histone phosphorylation in chromatin condensation.     

Presence of histone H1o-related fraction in chicken liver

Abstract. The lysine-rich histones of chicken liver were studied in order to see whether a protein similar to mammalian histone H1o was present in this lower vertebrate. The following biochemical methods were used: sodium dodecylsulphate and acid-urea electrophoresis, gel exclusion chromatography on BioGel P100, and ion-exchange chromatography on BioRex 70. Specific polyclonal antibodies were elicited against purified mouse liver Hlo and chicken erythrocyte H5, and applied for the further characterization of the chicken H1 subfractions obtained chromatographically. The results from microcomplement fixation and enzymelinked immunosorbent assays showed that the presumptive chicken liver Hlo shared common antigenic determinants with the mammalian H1o and the chicken liver H5. Based on the combined biochemical and immunological evidence, we conclude that an H1o-like protein is present in quiescent differentiated avian cells. The data of Smith et al. [34], who did not find this specific lysine-rich histone in resting chicken cells, are discussed.     

Presence of histone H1o-related fraction in chicken liver

The lysine-rich histones of chicken liver were studied in order to see whether a protein similar to mammalian histone H1o was present in this lower vertebrate. The following biochemical methods were used: sodium dodecylsulphate and acid-urea electrophoresis, gel exclusion chromatography on BioGel P100, and ion-exchange chromatography on BioRex 70. Specific polyclonal antibodies were elicited against purified mouse liver H1o and chicken erythrocyte H5, and applied for the further characterization of the chicken H1 subfractions obtained chromatographically. The results from microcomplement fixation and enzyme-linked immunosorbent assays showed that the presumptive chicken liver H1o shared common antigenic determinants with the mammalian H1o and the chicken liver H5. Based on the combined biochemical and immunological evidence, we conclude that an H1o-like protein is present in quiescent differentiated avian cells. The data of Smith et al. [34], who did not find this specific lysine-rich histone in resting chicken cells, are discussed.     

Thyroxine stimulation of tadpole liver histone phosphorylation in vivo

The in vivo phosphorylation of histones in the livers of Rana catesbeiana tadpoles was followed during the course of thyroxine-induced metamorphosis. Phosphorylation of histones H1 and H2a, and possibly of histone H4 at a low level, was observed in all animals. After correction for specific radioactivity of liver inorganic phosphate pools, an apparent wave of phosphorylation of histones was found to occur between 2 and 8 days of thyroxine treatment, with a peak increase of approximately 2- to 5-fold for histones H2a and H1. The increases in liver histone phosphorylation are approximately coincident with well-documented increases in the levels of various liver enzymes and occur in the absence of any change in the low basal rate of histone or DNA synthesis in this organ. This is apparently the first instance of increased phosphorylation of both H1 and H2a which is not coincident with or precedent to increases in cellular proliferation rates.     

Protein phosphorylation in normal and neoplastic hepatocytes]

Partially purified preparations of protein kinase were isolated from the cytoplasm and nuclei of rat liver and hepatoma 27 and characterized in terms of their substrate specificity. The protein kinases from normal liver and hepatoma revealed some differences in phosphorylating protein substrates. Hepatoma protein kinases were found to phosphorylate arginine-rich histones (H3, H4); differences in phosphorylation of histone H1 were revealed. Hepatoma protein kinases phosphorylated the C-terminal fragment of histone H1, whereas normal liver protein kinases produced no such effect. It was assumed that the phosphorylation of histone H1 in rapidly dividing and resting cells is operated through different channels.     

Declining histone phosphorylation during myogenesis revealed by in vivo and in vitro labeling

To investigate histone phosphate levels during myogenesis, proliferation (d 1), pre-fusion postmitotic (d 2) and myotube (d 3) stage cultured chicken myoblasts were phosphorylated in vivo with [32P]orthophosphate or in vitro by incubating isolated nuclei with 32P-gamma-ATP. Incorporation of radioactive phosphate into histone was assessed by SDS and acid/urea/Triton-X-100 (AUT) gel electrophoresis and radioautography. During proliferation, in vivo labeling with [32P]orthophosphate revealed that all histones except H2b were phosphorylated in the following order of decreasing modification: H1 a greater than H2a greater than H1 b greater than H3 greater than H4. In pre-fusion post-mitotic cells phosphorylation of histones H1 a, H3 and H4 declined, whereas all histones exhibited significantly decreased modification at the myotube stage. It is unlikely that these changes resulted from decreased specific radioactivity of intracellular inorganic phosphate pools, since uptake of [32P]orthophosphate by myotubes increased six-fold, compared with proliferating cells. Isolated nuclei incubated with 32P-gamma-ATP displayed similar decreases during myogenesis; however, 1 a, H1 b and H3 were the only histones modified by in vitro phosphorylation.     

Reactivation of DNA replication in erythrocyte nuclei by Xenopus egg extract involves energy-dependent chromatin decondensation and changes in histone phosphorylation.

Reactivation of chicken erythrocyte nuclei for DNA replication in Xenopus egg extracts involves two phases of chromatin remodelling: a fast decondensation leading to a small volume increase and chromatin dispersion occurring within a few minutes (termed stage I decondensation), followed by a slower membrane-dependent decondensation and enlargement of up to 40-fold from the initial volume (stage II decondensation). Chromatin decondensation as measured by nuclear swelling and micrococcal nuclease digestion required ATP. We observed a characteristic change in the phosphorylation pattern of erythrocyte proteins upon incubation in egg extract. While histones H5, H2A, and H4 became selectively phosphorylated during decondensation, the phosphorylation of histone H3 and of several nonhistone proteins was prevented. Furthermore, histone H5 was selectively released from erythrocyte nuclei in an energy-dependent reaction. These molecular changes already occurred during stage I decondensation and they persisted during stage II decondensation. DNA replication was confined to nuclei of stage II decondensation which incorporated lamin LIII from the egg extract. These results show that initiation of DNA replication in chicken erythrocytes requires in addition to ATP-dependent chromatin remodelling (stage I), further changes in chromatin structure that correlates with lamin LIII incorporation, and stage II decondensation.     

Changes in subfraction composition of chicken lysine-rich histone during ontogenesis]

Lysine-rich histone isolated from different chicken tissues was separated electrophoretically into 4-5 subfractions. The subfrations reffered to as 1, 2, 3, and 4, occur in each the tissue studied, erythrocyte lysine-rich histone containing an additional subfraction 1a. F1 histone from mitotically active tissues (intestinal mucosa, thymus, testes) has a higher content of subfraction 2, while the same histones from mototically inactive tissues (liver, heart, brain) contain an elevated amount of subfraction 3. F1 histone isolated from liver, brain and heart of 21-day embryo has much more of subfraction 2, than the same histone of adult animal. During the chicken development from hatching till maturation the content of subfraction 2 in these organs decreases, and the content of subfraction 3 increases. The rate of this change in liver corresponds to the rate of DNA synthesis. In F1 histone of erythrocytes the content of subfraction 4 falls down during the post hatching ontogenesis.     

Tissue-specific histones in the erythrocytes of chicken and turtle

Mature erythrocytes from Leghorn chickens contain lysine-rich histone F1 and a tissue-specific histone F2c. The composition of the F1 fraction was found to be similar to the F1 histones in higher vertebrates. In the erythrocytes of a sea turtle (Chelonia mydas), only lysine-rich histones F1 could be detected. One of these fractions (F1b) differed in amino acid composition from the typical F1 histones described in the literature. The F1b histone fraction was not found in turtle liver. Chromatographic analysis of tryptic peptides of the chicken erythrocyte F1 and F2c histones and of the turtle erythrocyte F1a and F1b histones revealed considerable similarities between these four fractions, thus indicating their possible phylogenetic relationships.     

In vivo phosphorylation of histones H1 and H5 in calf thymus and chicken erythrocyte as studied by 31P nuclear magnetic resonance spectroscopy

The 31P NMR method was first applied to characterize in vivo phosphorylation of H1 and H5 in calf thymus and chicken erythrocytes as well as in vitro phosphorylation of H1 and H5 by cAMP-dependent protein kinase. The amino acid residues phosphorylated in vivo in the histones were exclusively serine residues, and the mole fraction of phosphoserine was estimated to be 0.34 and 0.27 per molecule of calf thymus H1 and chicken erythrocyte H5, respectively. Interestingly, chicken erythrocyte H1 was not phosphorylated in vivo. Three H1 subtypes from calf thymus H1 varied in the 31P NMR spectra, and the bisected fragments of calf thymus H1 and chicken erythrocyte H5 exhibited characteristic spectral patterns, indicating that there are considerable diversities of the degree of phosphorylation and phosphorylation sites in very-lysine-rich histones. Furthermore, it was found that the microenvironment of phosphoserine residues phosphorylated in vivo in calf thymus H1 and chicken erythrocyte H5 is quite distinct from that of phosphoserine residues phosphorylated in vitro by bovine heart cAMP-dependent protein kinase.     

Relationship between histone phosphorylation and premature chromosome condensation

The relationship between histone phosphorylation and chromosome condensation was investigated by determining changes in phosphorylation levels of histones H1 and H3 following fusion between mitotic and interphase cells and subsequent premature chromosome condensation. We detected significant increases in the levels of phosphorylation of H1 and H3 from interphase chromatin in which a majority of nuclei had undergone premature chromosome condensation. In addition, we noted significant decreases in the phosphate content of the highly phosphorylated mitotic H1 and H3, presumably resulting from phosphatase activity contributed by the interphase component of mitotic/interphase fused cells. These observations further strengthen the correlation between histone phosphorylation and the changes in chromosome condensation associated with the initiation of mitosis. This study also suggests that maintenance of the mitotic chromosomes in a highly condensed state does not require the continued presence of histones in a highly phosphorylated form.     

Phosphorylation of rat thymus histones, its control and the effects thereon of gamma-irradiation.

The phosphate content of rat thymus histones was determined. As expected for a replicating tissue, histones 1 and 2B were more phosphorylated and had higher 32P uptakes than did histones from resting liver nuclei; the other histones all showed 32P uptake, but the phosphate content and uptake of histone 2A was about half that for liver histone 2A. When thymus nuclei were incubated in a slightly hypo-osmotic medium, non-histone proteins and phosphorylated histones were released into solution; this was enhanced if ATP was present in the medium. [gamma-32P]ATP was incorporated into non-histone proteins, including protein P1, and into the ADP-ribosylated form of histone 1; negligible 32P was incprporated into the other, bound, histones. Histones 1 and 2B added to the incubation medium were extensively, and histones 2A and 4 slightly, phosphorylated. Histones released by increasing the ionic strength of the medium were phosphorylated. Added lysozyme and cytochrome c were neither bound nor phosphorylated, but added non-histone protein P1 was phosphorylated, causing other histones to be released from the nuclei, especially histones 2A and 3. The released histones were phosphorylated. gamma-Irradiation decreased 32P uptake into the non-ADP-ribosylated histones 1 and 4; phosphorylation of histone 1 in vitro was unaffected. The importance of non-histone proteins, ATP availability and nuclear protein kinases to the control of histone phosphorylation in vivo is discussed.     

Regulation of ultraviolet B-induced phosphorylation of histone H3 at serine 10 by Fyn kinase

Ultraviolet B (UVB) induces phosphorylation of histone H3 at serine 10, and mitogen-activated protein kinases are involved in this signal transduction pathway. Here we provide evidence that Fyn kinase, a member of the Src kinase family, is involved in the UVB-induced phosphorylation of histone H3 at serine 10. UVB distinctly increased Fyn kinase activity and phosphorylation. Fyn kinase inhibitors 4-amino-5-(4-chlorophenyl)-7(t-butyl)pyrazol(3,4-d)pyramide and leflunomide, an Src kinase inhibitor, suppressed both UVB-induced phosphorylation of histone H3 at serine 10 and Fyn kinase activity and phosphorylation. UVB-induced phosphorylation of histone H3 at serine 10 was blocked by either a dominant-negative mutant of Fyn (DNM-Fyn) kinase or small interfering RNA of Fyn kinase. UVB-induced phosphorylation and activities of ERKs and protein kinase B/Akt were markedly inhibited by DNM-Fyn kinase. However, DNM-Fyn kinase did not inhibit UVB-induced phosphorylation of p38 MAPK or c-Jun N-terminal kinases. Active Fyn kinase phosphorylated histone H3 at serine 10 in vitro, and the phosphorylated Fyn kinase could translocate into the nucleus of HaCaT cells. These results indicate that Fyn kinase plays a key role in the UVB-induced phosphorylation of histone H3 at serine 10.     

C-terminal phosphorylation of murine testis-specific histone H1t in elongating spermatids

Previous studies gave differing results as to whether the testis-specific histone H1t was phosphorylated during rodent spermatogenesis. We show here that histones extracted from germ cell populations enriched with spermatids at different stages of development in rat testes reveal an electrophoretic shift in the position of H1t to slower mobilities in elongating spermatids as compared to that from preceding stages. Alkaline phosphatase treatment and radioactive labeling with (32)P demonstrated that the electrophoretic shift is due to phosphorylation. Mass spectrometric analysis of histone H1t purified from sexually mature mice and rat testes confirmed the occurrence of singly, doubly, and triply phosphorylated species, with phosphorylation sites predominantly found at the C-terminal end of the molecule. Furthermore, using collision-activated dissociation (CAD) and electron transfer dissociation (ETD), we have been able to identify the major phosphorylation sites. These include a new, previously unidentified putative H1t-specific cdc2 phosphorylation site in linker histones. The presence of phosphorylation at the C-terminal end of H1t and the timing of its appearance suggest that this post-translational modification is involved in the reduction of H1t binding strength to DNA. It is proposed that this could participate in the opening of the chromatin fiber in preparation for histone displacement by transition proteins in the next phase of spermiogenesis.     

Mitotic phosphorylation of histone H3 threonine 80

The onset and regulation of mitosis is dependent on phosphorylation of a wide array of proteins. Among the proteins that are phosphorylated during mitosis is histone H3, which is heavily phosphorylated on its N-terminal tail. In addition, large-scale mass spectrometry screens have revealed that histone H3 phosphorylation can occur at multiple sites within its globular domain, yet detailed analyses of the functions of these phosphorylations are lacking. Here, we explore one such histone H3 phosphorylation site, threonine 80 (H3T80), which is located on the nucleosome surface. Phosphorylated H3T80 (H3T80ph) is enriched in metazoan cells undergoing mitosis. Unlike H3S10 and H3S28, H3T80 is not phosphorylated by the Aurora B kinase. Further, mutations of T80 to either glutamic acid, a phosphomimetic, or to alanine, an unmodifiable residue, result in an increase in cells in prophase and an increase in anaphase/telophase bridges, respectively. SILAC-coupled mass spectrometry shows that phosphorylated H3T80 (H3T80ph) preferentially interacts with histones H2A and H4 relative to non-phosphorylated H3T80, and this result is supported by increased binding of H3T80ph to histone octamers in vitro. These findings support a model where H3T80ph, protruding from the nucleosome surface, promotes interactions between adjacent nucleosomes to promote chromatin compaction during mitosis in metazoan cells.     

Only a small fraction of avian erythrocyte histone is involved in ongoing acetylation.

We have studied histone acetylation in chicken erythrocytes. We find that about 30% of the histone in these cells is acetylated, however the majority of these histones are not in a dynamic steady state typical of other chicken cells and of mammalian cells, but rather are frozen in this state of modification. A very small fraction of erythrocyte histones are being modified normally but cannot be detected as shifting to higher levels of acetylation upon treatment with butyrate because the amount of histone so modified is small. Nonetheless, chicken erythrocytes incorporate 3H-acetate into histones about 40% as well as seen in the dynamically active HTC cells. This is most likely due to the formation of very high specific activity Acetyl CoA pools in erythrocytes which have very low levels of coenzyme A. We conclude that these genetically inactive cells are involved in only a minor way with histone acetylation.     

NAP-2: histone chaperone function and phosphorylation state through the cell cycle

We have recently cloned the human nucleosome assembly protein 2 (NAP-2). Here, we demonstrate that casein kinase 2 (CKII) from HeLa cell nuclear extracts interacts with immobilized NAP-II, and phosphorylates both NAP-2 and nucleosome assembly protein 1 (NAP-1) in vitro. Furthermore, NAP-1 and NAP-2 phosphorylation in crude HeLa cell extracts is abolished by heparin, a specific inhibitor of CKII. Addition of core histones can stimulate phosphorylation of NAP-1 and NAP-2 by CKII. NAP-2 is also a phosphoprotein in vivo. The protein is phosphorylated at the G0/G1 boundary but it is not phosphorylated in S-phase. Here, we show that NAP-2 is a histone chaperone throughout the cell cycle and that its cell-cycle distribution might be governed by its phosphorylation status. Phosphorylated NAP-2 remains in the cytoplasm in a complex with histones during the G0/G1 transition, whereas its dephosphorylation triggers its transport into the nucleus, at the G1/S-boundary, with the histone cargo, suggesting that binding to histones does not depend on phosphorylation status. Finally, indirect immunofluorescence shows that NAP-2 is present during metaphase of HeLa and COS cells, and its localization is distinct from metaphase chromosomes.     

Phosphorylation of nuclear proteins

Many nuclear proteins are phosphorylated: they range from enzymes to several structural proteins such as histones, non-histone chromosomal proteins and the nuclear lamins. The pattern of phosphorylation varies through the cell cycle. Although histone H1 is phosphorylated during interphase its phosphorylation increases sharply during mitosis. Histone H3, chromosomal protein HMG 14 and lamins A, B and C all show reversible phosphorylation during mitosis. Several nuclear kinases have been characterized, including one that increases during mitosis and phosphorylates H1 in vitro. Factors have been demonstrated in maturing amphibian oocytes and mitotic mammalian cells that induce chromosome condensation and breakdown of the nuclear membrane. The possibility that they are autocatalytic protein kinases is considered. The location of histone phosphorylation sites within the nucleosome is consistent with a role for phosphorylation in modulating chromatin folding.