Phosphorylation of non-histone chromosomal proteins

Phosphorylation of non-histone chromosomal proteins

One of most unsual properties of non-histone proteins in the extensive degree to which they are phosphorylated. The protein-bound phosphate content of this fraction as a whole is enough to account for several phosphorylated amino acids for every hundred residues present, giving the non-histone protein fraction the highest concentration of phosphorylated polypeptides of any protein fraction in the cell. All the properties discussed thus far as characterizing the non-histone protein fraction as a whole (i.e. heterogeneity, tissue/species specificity, association with active chromatin and changes in association with alteration in gene activity) apply to the phosphorylated non-histone proteins as well^{44, 45, 63}. In regard to these properties, especially striking is the vast number of different experimental system in which changes in non-histone protein phosphorylation have been observed in association with changing nuclear metabolism, chromatin structure, and gene activity (Table 3.1).

            Such observation have stimulated interest in isolating and identifying individual phosphorylated non-histone proteins so that their spesific roles in nuclear functions can be investigated. In the process, it has become gradually apparent that many different kinds of non-histone proteins are phosphorylated, and that the physiological roles of these various proteins may be quite diverse. Thus, among the non-histone nuclear protein now known to be phosphorylated one can distinguish at least a dozen different catagories: chromatin associated proteins which influence transcription, nucleolar proteins, the S-100 (brain-specific) protein, RNA polymerase and associated factors, DNA-binding protein, poy (A) polymerase, histone deacetylase, glucocorticoid receptors, nucleosomal proteins (including HMG proteins), protein associated with heterodisperse nuclear RNA, nuclear envelope proteins, and nuclear matrix protein. A detailed discussion of the properties of each of these groups of the phosphorylated proteins is beyond the scope of this chapter, especially since the subject has been recently reviewed in depth elsewhere⁶³. In the present context, only a few general issues raised by this diverse group of phosphorylated proteins will be touched upon.

            The most critical issue regrads the significance of the fact that these proteins are all phosphorylated. The functional significance of protein phosphorylation has been most clearly estabilished for cytoplasmic enzymes such as glycogen phosphorylas, phosphorylase kinase, and glycogen synthetase, where such modifications serve to either increase or decrease catalytic activity depending on the enzyme in question^{27, 49, 51, 78, 88}. By analogy it has generally been assumed that phosphorylation serve to regulate the activities of other proteins known to be phosphorylated, though in most instances such a regulatory role is yet to be clearly demonstrated. In the case of the phosphorylated non-histone proteins listed above,

TABLE 3.1 Summary of experimental systems in which changes in non-histone protein phosphorylation have been observed (for detailed reference, see reference 63)

Growth and Development Lymphocytes activity Avian erytrocytes Physarium polycephalum HeLa cells Reuben hepatoma cells Yoshida sarcoma Sea urchin embryos Trout embryos Muscle differentiation Neuronal and glial cells Kidney regeneration Barley germination Chinese hamster cells (K12) WI-38 fibroblast Chick embryo fibroblasts Hamster kidney cells (BHK21C13) Testis and epididymis Uterus ChO cells (heat shock) Liver (aging) Liver (regeneration) Liver (cell culture) Landschutz tumor cells (amino acid starvation) Chemical Agents Isoproterenol (salivary gland) Isoproterenol (pineal gland) Acetylcoline (lymphocytes) Norephinephrine (glioma) Carbachol (lymphocytes) Phenobarbitone (liver) α-1, 2, 3, 4, 5, 6-Hexachlorocyclohexane (liver) Morphine (brain) Hemin (liver) Vitamin D (kidney, liver) Prostaglandin (lymphocytes) 2, 4-Dichlorophenoxyacetic acid (soybeans) Helanalin (Ehrlich ascites) Cytarabine (salivary gland) Bleomycin (salivary gland) Chlorambucil (Yoshida sarkoma) Melphalan (Yoshida sarkoma) Cyclophosphamide (Yoshida sarkoma) Eupahyssopin (Ehrlich ascites)

Hormonal Stimulation Testosterone (Prostate) Antiandrogens (prostate) Glucocorticoids (Insect salivary glands) Glucocorticoids (liver) Glucocorticoids (lymphosarcoma) Aldosterone (kidney) Oestradiol (uterus) Oestradiol (mammary carcinoma) Oestradiol (oviduct) Oestradiol (brain) Prolactin (mammary gland) Chrionic gonadotropin (ovary) Calcitonin (bone cells) Triiodothyronine (liver) Triiodothyronine (heart) Thyrotropin (thyroid) Abscisic acid (Lemna) Cyclic AMP (rat liver) Cyclic AMP (rat heart) Cyclic AMP (lymphocytes) Cyclic AMP (mammary carcinoma) Cyclic AMP (adrenal medulla) Cyclic AMP (salivary gland) Cyclic AMP (Neurospora) Cyclic AMP (pineal gland) Cyclic AMP (lymphocytes) Cyclic AMP (salivary gland) Maglinant and Transformed Cells Mammary carcinoma Azo-dye carcinogenesis Dimethylbenzanthracene carcinogenesis Methylcholanthrene carcinogenesis SV-40 transformed fibroblast Morris hepatoma Novikoff hepatoma Friend erythroleukimia Ehrlich ascites Walker tumor Adenovirus transformed cells Murine sarcoma virus transformed cells Human leukimia

Phosphorylation has been shown to affect activation of gene transcription^{14, 15, 47, 50} non-histone protein binding to DNA^{35, 50, 61, 67}, poly (A) polymerase activity^{75, 76}, RNA polymerase activity⁴⁸ and glucocorticoid binding activity⁶⁵. For other phosphorylated non-histone proteins, such as S-100 histone deacetylase, nucleosomal phosphoprotein, hnRNA-associated phosphoprotein, and nuclear envelope and nuclear matrix phosphoproteins, no direct evidence concerning the functional significance of phosphorylation is avaliable. Though there is no shortage of attractive ideas (e.g. phosphorylation of hnRNA-associated proteins may be involved in regulating RNA processing and/or transport, phosphorylation of histone deacetylase may regulate its catalytic activity, phosphorylation of nucleosomal proteins may influence chromatin structure, etc), one should exercise extreme caution in such instance where there are no definitive data avaliable.

            The fact that the phosphorylation state of non-histone proteins influences their ability to stimulate RNA synthesis and bind to DNA in vitro raises the question of the possible role of non-histone phosphorylation in the control of spesific gene expression. This question is difficult to answer unequivocally because of the severe limitations of the cell-free systems in which such investigations have been carried out. However, after discussing the remaining properties of non-histone protein protein fraction, we will return to this general issue in more detail, discussing both the limitation of the current data implicating non-histone proteins in specific gene regulation and the prospects for obtaining more definitive information in the future.

DNA binding of non-histone chromosomal proteins

Selective gene expression implies the restriction of some genetic loci and the specific activation of others. If the non-histone proteins play a role in these processes, it is reasonable to postulate that the mechanism through which transcriptional control is achieved will depend, in part, on associations between these proteins and DNA. In prokaryotes these are several examples of DNA binding proteins which regulate RNA synthesis, for example lac represormediated inhibitor of transcription of the lac operon^{2, 26} and the cAMP receptor-mediated facilitation of initation of lac and gal transcription^{4, 104}.

            Spesific DNA binding non-histone proteins have been isolated by several groups of workers. In the section we will discuss some of the properties of these DNA binding proteins which may have a bearning on their postulated role in the control of gene expression.

Sequence-specific DNA binding non-histone chromosomal proteins

DNA binding non-histone proteins have been most commonly isolated by affinity chromatography on coloums containing immobilized homologous or heterologous DNAs. Only a small proportion of the proteins in non-histone preparations bind specifically to homologous DNA^{46, 92} and many of these are phosphoproteins⁴³. Considerable confusion exists as to heterogeneity of these proteins, some groups observing a highly complex population of non-histone protein^{43, 89}, whereas others do not^{68, 92, 94}. Similarly, most groups have found that non-histones bind selectively to homologous DNA, though others have found little species-specific binding and argue that their interaction with DNA has severaly restricted the evolutionary diversity of these proteins. Such data must generally be considered with some coution because binding conditions very from group to group, different non-histone protein preparations are used, in most cases considerable overlap of proteins in DNA binding fractions has been observed, and rechromatography of ‘specific’ DNA binding proteins fractions has yielded contradictory results.

            Specific binding of non-histone proteins to DNA probably occurs under a limited set of conditions, and optimization of specific binding in vitro requires the determination of a number of binding parameters. The nitrocellulose filter binding of nucleoprotein complexes has provided a rapid and flexible DNA binding assay. Several and co-workers^{39, 40, 81}, for example, used tandem heterologous and homologous DNA affinity chromatography to fractionate a rat liver non-histone protein preparation, and then employed a filter binding assay to analyze the binding specificities of the protein fractions obtained. At saturation, the heterologous DNA binding proteins bound 60-70% of whole rat liver DNA whereas the homologous DNA binding proteins boundonly 35% implying the existance of some sequence specificity. Competition experiments estabilished that the homologous DNA binding protein fraction bind homologous DNA  preferentially, with no such preference being observed for the heterologous DNA binding protein fractions. Analysis of the DNA retaind by filters showed that it is enriched four-fold unique DNA sequences and depleted in repetitive sequence relative total DNA.

            Cloned DNA sequences have also been used to identify sequence-spesific binding. Hsieh and Brutlag³⁶ isolated a protein from Drosophila embryos which specifically binds to a cloned repeating unit of satellite DNA when the plasmid and its insert are superhelical conformation. Retriction endonuclease mapping of this DNA-protein complex revealed that the protein protects one specific restriction site on the inserted DNA fragment. Similiarly, Weideli et al.⁹⁶ have isolated a DNA binding protein (DB-2) from Drosophila and bound it to random or retriction fragments of whole cell DNA. Bound DNA fragments were recovered from nitrocellulose filters, inserted into plasmids, and cloned in bacteria by standard recombinant DNA procedures. The cloned DNA fragments were reacted with the DNA binding protein again, and bound DNA fragments were cloned a second time. This process yielded a cloned DNA fragment which selectively binds to the DB-2 protein at high ionic strength (2.0 mol/ℓ NaCl). Immunocytochemical localization of the DB-2 protein revealed that is concentrated specifically on chromosome 3 (section 95a/b) in intact cells, supporting the postulate that the protein is associated with a specific region of the Drosophila genome in vivo.

            One obvious disandventage to experiments such as those described above, in which one monitors the binding of proteins to purified DNA sequence, is that in vivo the DNA is part of chromatin complex which may behave quite differently from naked DNA. A model system which attempts to overcome this difficulty has been devised by Chao et. al.¹³ who used 1 203 base pair DNA fragment containing the E.coli lac operators sequence to reconstitute nucleosomal structures with call thymus histone. When such nucleosomal were bound to the lac repressor protein, sequence-spesific recognition could be detected without with the diplacement of histone. Since the binding was quantitative and spesific, the results suggest that the reconstitution protocol yields structure where the bound DNA strand faces the outer surface of the nucleo-histone complex. Chemical crosslinking of the histone inhibits binding of the lac repressor only occurs when the operator sequence is positioned in the region of nucleosome stabilized by H2A-H2B interaction (but not H3-H4 interactions). This implies that squence-spesific binding in eukaryotic chromatin may be modulated by DNA interactions with histone proteins which define the accessibility of sequence to intermoleculer interactions.

            Histone may not be the only chromatin which influences the binding of non-histone proteins to DNA. Dastugue and Crepin¹⁹ have reported that the binding of non-histone proteins to isolated DNA exhibits less species specifity than the binding of the same proteins to intact chromatin. This observation suggests that the spesific binding of non-histone proteins to DNA may be influenced by chromatin proteins other than histone histones, since the histones do not vary much from species to species. Although the physicological significance, if any, of the in vitro binding of non-histone proteins to chromatin is not certain, it is interesting to note that such binding is accompanied by stimulation of transcription in homologous, but not heterologous, protein/chromatin structure mixtures¹⁸.

            An alternative approach to the identification of the spesific DNA : non-histone protein complexes formed in vivo has been provided by Bekhor and coworkers^{6, 7, 23, 66}. Essentially, chromatin is extensively washed with 2.0 mol/ℓ NaClto remove loosely bound proteins, and the protein-depleted DNA is then dialysed against 10 mmol/ℓ Tris-HCl. Low speed centrifugation yields as nucleoprotein complex containing some 1 % of the total nuclear DNA complexed to a spesific non-histone protein fraction. The DNA in this complex is enriched twenty-fold in transcribed sequences, and is depleted of non-transcribed sequences, inactive gene sequences, and highly repentitive DNA sequences. Since the proteins bind spesific DNA sequences with such high affinity, rearragement during isolation is an unlikely explanation for this nucleoprotein complex.