Use of histone antibodies for studying chromatin topography ... - NCBI

1 downloads 51 Views 1MB Size Report
Jun 22, 1984 - was supported by a grant to M.H.V.Van Regenmortel from the Institut .... Saccone,G.T.P., Skinner,J.D. and Burgoyne,L.A. (1983) FEBSLett., 157,.
The EMBO Journal vol.3 no. 10 pp.2431 - 2436, 1984

Use of histone antibodies for studying chromatin topography and the phosphorylation of chromatin subunits

Sylviane MuHler, Alice Mazen', Arlette Martinage2 and Marc H.V.Van Regenmortel Laboratoire de Virologie and 'Laboratoire de Biophysique, Institut de Biologie Mol&ulaire et Cellulaire du CNRS, 15 rue Descartes, 67000 Strasbourg, and 2Unite 409 CNRS, Institut de Recherche sur le Cancer, BP 311, 59045 Lille Cedex, France Communicated by L.Hirth

Polyclonal and monoclonal antibodies specific for histones as well as sera directed against synthetic peptides of histones were used to probe the topography of chromatin subunits. In native chromatin, the regions corresponding to residues 130 -135 of H3 and 6-18 of H2B were found to be exposed and able to interact with antibodies whereas the regions 26-35 and 36-43 of H2B and 80-89 and 85-102 of H4 were not. In vitro phosphorylation of H3 and H5 in native chromatin or of H3 in H1/H5-depleted chromatin led to a marked drop in the binding of antibodies specific for residues 130-135 of H3 and 6-18 of H2B. Phosphorylation of H1/H5-depleted chromatin also altered the degree of exposure of certain H2A epitopes but it did not affect the surface accessibility of residues 1-11 of H2B. Key words: synthetic peptide/histone monoclonal antibodies/ histone H5/histone phosphorylation/chromatin structure Introduction Antibodies to histones represent one of the most specific probes for studying the surface topography of chromatin. The usefulness of antibodies as analytical tools is enhanced when the antigenic determinants that they recognize have been identified. In recent years, about a dozen antigenic determinants (or epitopes) of the core histones have been located in their primary structures, within linear segments of 6-20 residues (Absolom and Van Regenmortel, 1977; Muller et al., 1982a, 1983; and in preparation). The availability of several monoclonal antibodies and antisera specific for some of these histone epitopes makes it possible to determine whether these regions of the histone molecules are exposed at the surface of core particles and nucleosomes. These antibodies are also very sensitive probes for monitoring conformational changes taking place in chromatin as a result of post-synthetic histone modifications such as acetylation (Muller et al., 1982b) and phosphorylation. Phosphorylation of chromosomal proteins has been reported to play a key role in the regulation of the eukaryotic genome (Gurley et al., 1978); Bradbury and Matthews, 1981). The exact function of these phosphorylations is not known but histone phosphorylation appears to correlate very closely with mitotic chromosome condensation. In Chinese hamster cells, for instance, essentially all the HI and H3 molecules are phosphorylated during mitosis but the phosphates are removed when the cells enter the GI phase (Lake, 1973; Gurley et al., 1978). Thus histone phosphorylation may be directly involved in stabilizing condensed mitotic chromatin (Lake, IRL Press Limited, Oxford, England.

1973; Gurley et al., 1978) or in initiating chromosome condensation (Matsumoto et al., 1980; Bradbury et al., 1974; Inglis et al., 1976). Phosphorylation leads to an alteration of the charge of modified serine and threonine residues and may thus modulate histone-DNA interactions. Sites of phosphorylation in individual histones or histone complexes have been investigated in vivo (Sung and Freedlender, 1978) and in vitro using the catalytic subunit of cyclic AMP-dependent protein kinase from pig brain (Kurochkin et al., 1977) and rat pancreas (Martinage et al., 1980, 1981). In core particles only one core histone was found to be phosphorylated by this enzyme, namely H3 at serine 10 (Taylor, 1982), whereas in long chains of chromatin, in addition to H3, HI was also phosphorylated at serine 38. The same sites have been shown to be modified in vitro in isolated HeLa metaphase chromosomes (Paulson and Taylor, 1982). In chicken erythrocyte chromatin, the number of phosphorylated sites in H5 has not been determined. We have compared the accessibility of a number of histone epitopes at the surface of chicken erythrocyte core particles, native chromatin and chromatin depleted of HI and H5. The removal of these two histones from chromatin was found to lead to a greater exposure of some of the antigenic regions of H2A, H2B and H3. The in vitro phosphorylation of histone H3 and H5 in native chromatin led to a decrease in accessibility of some epitopes of H3 and H2B. The C-terminal hexapeptide of H3 (residues 130-135) and the region 6-18 of H2B were particularly affected by phosphorylation. In addition, in phosphorylated Hl/H5-depleted chromatin a marked drop in the binding of antibodies specific for H2A was also observed.

Results Accessibility of various histone regions in native or HJ/H5depleted chromatin In an earlier report (Muller et al., in preparation) we located seven antigenic regions in the H2B molecule by means of antiH2B sera and monoclonal antibodies specific for H2B. A series of 23 natural and synthetic overlapping peptides covering the entire H2B molecule allowed the identification of several epitopes that were recognized by monoclonal antibodies. The antibodies 73 C3, 41 A5, C 13 and 69 B4 were shown to bind to regions of H2B situated in the N-terminal part of the histone, namely in residues 1 -11, 6-18, 10-35 and 26-35, respectively (see Table 1). The availability of these monoclonal antibodies as well as of an antiserum specific for the region 36-43 of H2B (Muller et al., in preparation) made it possible to monitor structural changes in a region of histone H2B that is considered to be involved in the functional regulation of chromatin. In a preliminary study, these antibodies were used to determine whether the corresponding epitope regions of H2B were exposed at the surface of chromatin and core particles. At the same time, the surface accessibility of different epitopes 2431

S.MuIer et al.

Table I. Binding of polyclonal and monoclonal antibodies to H2B with chicken erythrocyte chromatin, HI/H5-depleted chromatin and core particles Antibodies.

Region of H2B recognized~'

H2B antiserum 1-11 73 C3 6-18 41 A5 26-35 69 B4 10-35 C 13 Antiserum to 36-43 peptide (H2B)

Antigensc Chromatin

Hl/H5depleted chromatin

Core particles

1.82d

1.96 1.54 1.23 0.14 nd 0.12

1.88 >2 0.62 0.18 1.71 0.19

0.22 0.96 0.14 nde 0.10

aAntiserum to H2B or peptide 36-43 were diluted 1:100, ascitic fluid 73 C3, 41 A5 and 69 B4 were diluted 1:1000 and antibody C 13 was used at 4 tg Ig/mI. bFrom Muller et al., in preparation. 'The binding of antibodies to chromatin or chromatin subunits from chicken erythrocytes was measured in ELISA by coating wells with dilutions of material corresponding to 200 ng DNA/ml. dValues are expressed in optical density units measured at 405 nm. Background reading obtained in absence of antigen or with non-related serum or ascitis was < 0.20. end, not determined. 2.0

A A

A

1.0

c2.0

./A

-.

.

-

©

10 Q

0

~~~~~~~0

01.0

0

/

100

200

CHROMATIN

100

200

(ng DNA/ ml)

Flg. 1. ELISA binding curves of histone antibodies to increasing quantities of chicken erythrocyte chromatin (A and C) and Hl/H5-depeleted chromatin (B and D). Antisera were diluted 1:100 anti-H5 (V-V), anti-H3 ( A ), anti-C-terminal hexapeptide (IRGERA) of H3 (A), anti-H4 (-U), anti-H2A ([O E) and anti-H2B (0 *); ascitic fluids 73 C3 (-), 41 AS (0-0 ) and 69 B4 ( O> 0) were diluted 1:10 000. Binding to normal rabbit serum and non-related ascitic fluids was isgnificant.

recognized by antisera to histones H2A, H3, H4 and H5 was also determined. All four histone antisera reacted strongly with the corresponding free histones in solution (Muller et al., 1982a, 1983). The interaction between the chromatin and specific antibodies was measured in ELISA by coating the wells of microtiter plates with chromatin or core particles diluted in 2432

phosphate buffer pH 7.4 of 220 mM ionic strength, as described in previous studies (Muller et al., 1982b, 1983). As shown in Figure IA and C, H2B, H3 and H2A were found to be the most accessible histones in chromatin while H4 was the least accessible. The results confirm earlier findings of several authors (Goldblatt and Bustin, 1975; Bustin et al., 1977; Absolom and Van Regenmortel, 1978; Romac et al., 1981; Muller et al., 1982b, 1983). The accessibility of the epitopes of H2B situated in the N-terminal part of the molecule was then studied with monoclonal antibodies and an anti-peptide serum (Table I, Figure IC). Monoclonal antibody 41 A5 was able to bind to chromatin whereas antibodies 73 C3 and 69 B4 and the anti-36-43 peptide serum did not (Table I). The accessibility of the four core histones was also studied in Hl/H5-depleted chromatin from chicken erythrocytes. The results shown in Table I and Figure 1 were obtained with the same initial chromatin preparation after HI and H5 had been removed (see Materials and methods). A comparison of Figure IA and B shows that the removal of HI and H5 led to an increase in the immunoreactivity of H2A and H3 in chromatin whereas that of H4 remained unchanged. This was confirmed in several additional experiments using intermediary substrate incubation times in ELISA. Antibodies directed against the synthetic C-terminal hexapeptide 130-135 of H3 (IRGERA) also reacted better with HI and H5-depleted chromatin, whereas antibodies directed against the synthetic fragments 85-102 and 80-89 of H4 did not (results not shown). Polyclonal antiserum to H2B and monoclonal antibody 41 A5 reacted about equally well with depleted chromatin as with native chromatin (Table I), whereas the interaction between antibody 73 C3 and chromatin greatly increased when chromatin was depleted in HI and H5 (Figure IC and D, Table I). In contrast, no significant reaction was observed between HI/H5-depleted chromatin and antibody 69 B4 or antiserum to peptide 36-43 of H2B. Anti-H5 serum was also used as a control and was found not to bind to Hl/H5-depleted chromatin (data not shown). These results indicate that the region 6-18 of H2B (recognized by antibody 41 A5) is readily accessible at the surface of chromatin, and that the region 1-11 of H2B (recognized by antibody 73 C3) is masked in native chromatin but exposed when HI and H5 have been removed from chromatin. Since antibodies specific for the regions 26-35 (antibody 69 B4) and 36-43 (anti-peptide serum) do not bind to chromatin, we assume that the region 26-43 is buried in the nucleoprotein. The accessibility of H2B epitopes at the surface of core particles was also investigated. As shown in Table I, anti-H2B serum, and monoclonal antibodies 73 C3, C 13 and 41 A5 were able to bind to core particles whereas antibodies 69 B4 and anti-36 -43 peptide antibodies did not. These results show that the same regions are exposed in core particles as in Hl/H5-depleted chromatin except for the region 6-18 which appears to be less exposed in core particles than in chromatin. Study ofphosphorylated chromatin Native chicken erythrocyte chromatin and Hl/H5-depleted chromatin were phosphorylated in vitro by cyclic AMPdependent protein-kinase purified from porcine heart according to the method described by Mangeat et al. (1978). Only the histones H3 and H5 were phosphorylated as shown by electrophoresis on acid urea or Triton gels (A.Mazen, in preparation). It was determined that the totality of the H5

Chromatin topography and phosphorylation studied by histone antibodies

E

A.

c

E

.-,

,

_._r__, _,

,__,_

0

200

n10

©

©

°2.0

200

Fig. 3. ELISA binding curves of native and phosphorylated histones H5 (A) and H3 (B). Closed symbols correspond to native histones, open symbols to phosphorylated histones. Sera were diuluted 1:100:anti-H5 (O and 0), anti-H3 (- and El) and anti-IRGERA (A-and A). The concentration of histones in this expeximent was four time higher than in the other expeiments presented; thins does not influence the differentiation

/00

1.0

loo

H3 (ng /ml)

Hs (ng ml)

between the different curves. 2K 0

0-0. 100

9 200

;/=t-9

=-

100

200

C HROMATI N (ng DNA / ml) Fig. 2. ELISA binding curves of histone antibodies to increasing quantities of native chicken erythrocyte chromatin (A and C) and phosphorylated chromatin (B and D). For symbols of histone antibodies, see Figure 1. The curves in A and C correspond to the same experiment shown in Figure IA and C.

and H3 populations was modified. The amount of radioactivity incorporated into H3 and H5 were compatible with one and three phosphates per molecule respectively. In HU/H5-depleted chromatin, only H3 was phosphorylated; the same level of one phosphate per molecule was found. Free histones H3 and H5 were also phosphorylated and were found to incorporate one phosphate per H3 molecule and three phosphates per H5 molecule (A.Mazen, in preparation). The influence of H3 and H5 phosphorylation on the immunochemical reactivity of chromatin was studied with the antisera specific for individual histones and histone fragments and with the monoclonal antibodies described above. Comparative immunoassay binding curves with native chromatin and phosphorylated chromatin are presented in Figure 2. The major effect observed in these assays was the decrease in binding between phosphorylated chromatin and antibodies specific for certain determinants of H3 and H2B (Figure 2). The binding of antibodies specific for the C-terminal hexapeptide of H3 was reduced whereas the binding of antisera specific for whole H3 and H2A was only slightly affected by the phosphorylation (Figure 2A and B). The decrease in the anti-IRGERA curve in Figure 2B represents the 'hook' effect which is often observed in ELISA; the nature of this phenomenon is not well understood (Pesce et al., 1983). No difference in binding was observed with antibodies specific for H4 and H5. The interaction between H2B antiserum and H2B was much reduced when chromatin was modified by H3 and H5 phosphorylation (Figures 2C and D); the binding of monoclonal antibody 41 A5 specific for the region 6-18 of H2B was also greatly decreased (70- 90%7o) whereas the binding of antibody 73 C3 specific for the region 1-11 of H2B was unchanged or sometimes only slightly increased. The ability of antibody 69 B4 to react with chromatin remained unchanged after chromatin phosphorylation. Antibody C 13 (specific for

the region 10-35 of H2B) had the same binding properties as antibody 41 A5 (data not shown). The changes of antigenic reactivity of chromatin observed after phosphorylation could result from the fact that phosphorylation modified residues situated within antigenic determinants of histone H3 and H5. This possibility was tested by labelling isolated H5 and H3 using the same method as described for chromatin. H5 antiserum was found to bind in ELISA equally well to native and phosphorylated H5 (Figure 3A) while antisera to H3 and to the IRGERA peptide of H3 reacted only weakly with H3 after phosphorylation of the histone (Figure 3B). The decrease of antigenic reactivity after aferphosphorylation of the H3 molecule was not due to a destruction of the epitope located in the C-terminal end of this histone. Recurrent degradation of phosphorylated H3 with carboxypeptidases A and B showed that the same C-terminal amino acid residues (alanine and arginine) were cleaved as in the case of the unmodified histone. The decrease in binding of antibodies specierc for the IRGERA peptide observed after H3 phosphorylation (Figure 3B) cannot be ascribed, therefore, to an artefactual destruction of the C-tenrninal region of H3 during the phosphorylation reaction. The effect of phosphorylation on the antigenic properties of H t/H5-depleted chromatin was also determined in the same manner. As shown in Figure 4A and B, chromatin submitted to in vitro phosphorylation by cyclic AMP-dependent protein kinase presented a marked drop in the binding of antibodies specific for H2A, H3 and the IRGERA peptide. Such a decrease in the binding of H2A antibodies following h3sphosphorylation was not observed in th e of chromatin containing H3 and H5. The interaction betweenH./H5depleted chromatin and H4 antiserum was not modified by phosphorylation. However, in spite of the fact that in Hl/H5-depleted chromatin there is only one site of phosphorylation, namely on histone H3, the binding of H2B antiserum and of antibody 41 A5 was also greatly diminished (75 -85 o) when this chromatin was phosphorylated (Figure 4C and D); in contrast, interaction of antibody 73 C3 remained almost constant after phosphorylation. Antibody 69 B4 did not react at aH with phosphorylated He/H5-depleted chromatin.

Discussion Studies by neutron diffraction at

a

resolution of 16Ai 2433

S.MuIer et al.

located at the entrance and exit points of the nucleosomal DNA (Allan et al., 1980). The removal of HI leads to a more open structure of chromatin. Although all the domains of HI (and H5) are involved in the organization of chromatin (Thoma et al., 1983), the precise mechanisms used for bringing about the superstructure of chromatin are still controversial (Russo et al., 1983). Our results are in agreement with earlier data describing a relaxation of chromatin depleted of HI and H5. Some of the antigenic determinants that became unmasked may also correspond to regions in these histones E' © © that interact with HI; cross-linking experiments have shown 0 2.0 that HI is interacting with various core histones (Ring and Cole, 1979; Boulikas et al., 1980; Bakayev et al., 1981). When native chromatin was phosphorylated in vitro by cyclic AMP-dependent protein kinase, the major immuno0. chemical change was a decrease in the ability to bind H3 and H2B antibodies. It seemed that the regions 130-135 of H3 and 6 - 18 of H2B, in particular, were considerably less ac$ 0~~~~~~~~~~_O~ 00 cessible to antibodies (Figure 2). In the case of HI/H5depleted chromatin, phosphorylation had an even greater ef'F~~~~~~~~~~~~ 0 fect on the antigenic reactivity of H3 and of the region 6-18 0 200 200 100 100 of H2B (Figure 4). In addition, the antigenicity of H2A was also considerably reduced. As shown in Figure 3B, phosCHROMATIN (ng DNA/ ml) phorylation of the native H3 molecule on its own also Flg. 4. ELISA binding curves of histone antibodies to increasing quantities decreased the binding of anti-IRGERA antibodies. Since the of Hl/H5-depleted chromatin (A and C) and phosphorylated H1/H5site in H3 is present in the N-terminal region phosphorylated depleted chromatin (B and D). For symbols of histone antibodies, see of the it seems that phosphorylation leads to a conmolecule, 1. Figure The curves in A and C correspond to the same experiment formational change in H3 which modifies the ability of the shown in Figure lB and D. C-terminal region to bind IRGERA antibodies. The same (Bentley et al., 1984) and by X-ray diffraction at a resolution antibodies also react less well with phosphorylated chromatin of 25 A (Finch et al., 1981) have not yet made it possible to (Figure 2A and C). These two observations are clearly due to visualize the relative positions and orientations of the difa change in conformation of the C-terminal end of H3 as a ferent histones inside core particles. Little information is result of a residue modification in the N-terminal part of the therefore available concerning the orientation of discrete molecule. In monomeric proteins, it is commonly observed histone domains and of the mobile histone tails which are that the C and N termini are in close opposition at the surface known to play a key role in the stability of the nucleoprotein of the molecule (Thornton and Sibanda, 1983). It is, of (Allan et al., 1982; Grigoryev and Krasheninnikov, 1982). course, unclear whether the H3 termini are also juxtaposed in The N-terminal region of histone H2B, in particular, seems to chromatin and whether the modification of residue 10 alters interact preferentially with the linker DNA, and this could be the C-terminal region in exactly the same fashion in the responsible for the characteristic length of different linkers in monomeric histone as in the histone octamer. chromatin extracted from various species (Pospelov et al., When HI and H5 have been removed from chromatin, 1979; Zalebskaya et al., 1981). only a single residue remains modified by phosphorylation, In this study we examined whether certain antigenic deternamely the serine at position 10 in H3 (Taylor, 1982; Paulson minants of histone molecules are located at the surface of and Taylor, 1982). This allows the influence of H3 chromatin subunits, and we used for this purpose not only phosphorylation on chromatin antigenicity to be perceived antisera to histones but also monoclonal antibodies specific more clearly. The ability of a single modified residue of H3 to for well-defined regions of histone H2B. This made it possiaffect so markedly the surface exposure of different parts of ble to show that the regions 1-11, 26-35 and 36-43 of histones H2B, H2A and H3 would seem to corroborate histone H2B are not accessible at the surface of native previous observations indicating that the N-terminal region of chromatin, in contrast to the region 6-18 which is exposed H3 plays a decisive role in maintaining the superstructure of and able to bind to antibodies (Table I; Figure lA and C). chromatin (Chatterjee and Walker, 1973; Saccone et al., The surface location of the first N-terminal residues of 1983; Marion et al., 1983a, 1983b) and in bringing about its H2B in chromatin has previously been demonstrated by condensation (Ajiro et al., 1983). various approaches such as limited proteolysis (Rill and Although it has been reported that the condensation of Oosterhof, 1982; Diaz and Walker, 1983), cross-linking chromosomes during mitosis is linked to the phosphorylation (DeLange et al., 1979; Suda and Iwai, 1979) and immunoof histones HI and H3 (Gurley et al., 1978; Bradbury and chemical studies (Di Padua Mathieu et al., 1981). It is wellMathews, 1981), the exact connection between these two established that histone HI plays a particular role in the conevents remains uncertain (Krystal and Poccia, 1981; Hanks et densation of chromatin and that it stabilizes the core particle al., 1983; Hohmann, 1983). It is mentioned, however, that (Thoma and Koller, 1977, 1981; Worcel and Benyajati, 1977; histone phosphorylation influences histone-DNA and Thoma et al., 1979; Stratling, 1979; Butler and Thomas, histone-histone interactions and that it plays a role in dif1980; Harborne and Allan, 1983). Histone HI is situated on ferential gene activity (Allfrey, 1980; Bradbury and Matthe outside of the core particle and its central globular part is thews, 1981). 2434

Chromatin topography and phosphorylation studied by histone andbodies

Our results demonstrate that the alterations of chromatin structure induced by phosphorylation can be monitored by immunochemical means. They show that phosphorylation not only alters the superstructure of chromatin but also modifies the topography of core histones in chromatin subunits.

Materials and methods Preparation of chromnatin and core particles Nuclei and adult chicken erythrocytes and chromatin were prepared as described previously (Mazen et al., 1982). Micrococcal nuclease digestion was carried out until 1% of DNA was rendered acid-soluble. Long chains of chromatin (20-50 nucleosomes) were fractionated at 4°C by centrifugation on a 5-28Oo sucrose gradient in TEP buffer (I mM Tris-HCl, I mM EDTA, 0.1 mM phenyhnethylsulfonylfluoride (PMSF) pH 7.4. Hl/H5-depleted chromatin was obtained by centrifugation on a 5 - 281o sucrose gradient containing 650 mM NaCl in TEP buffer, and dialysed against TEP buffer. If necessary, the resulting chromatin fractions were concentrated on Diaflo XM 50 membranes in a stirred AMICON cell. The quality of chromatin preparations was routinely controlled by electron microscopy. The DNA content was analysed in 1 %o agarose gel electrophoresis and the proteins were characterized in SDS-18Wo polyacrylamide gels. The ratio Hl/H5 was about 1:3 as reported previously (Mazen and Champagne, 1972; Wright and Olins, 1975). Chromatin samples were stored at 4°C in the presence of 0.1 mM PMSF and 0.2 mM EDTA buffer pH 7.4 and used within 4 days. Chicken erythrocyte core particles were prepared and characterized as previously described (Muller et al., 1982a). These particles contained the four core histones H3, H4, H2B and H2A in equal proportions but no Hl nor H5; the size of DNA was 145 : 3 bp. Polyacrylamide gel electrophoresis, circular dichroism and electron microscopy were used to monitor the various samples. Cyclic AMP-dependent protein kinase from porcine heart Purification of the catalytic subunit was performed in the laboratory of Dr G.Marchis-Mouren according to a simplified procedure adopted from Mangeat et al. (1978). The 80 000 g supernatant was applied to a DEAESepharose column and, after extensive washing, the catalytic subunit was then eluted with 10 juM cyclic AMP and directly adsorbed on a CM-Sepharose column connected in tandem. The subunit was eluted with 150 mM NaCl. The yield was 8000 units per kg of porcine heart. The enzymatic activity was assays as described by Taylor (1982). In vitro phosphorylation of histones and chromatin The phosphorylation procedure was adapted from the work of Nelson and Taylor (1981) and will be described in detail elsewhere (Mazen, in preparation). Briefly, in vitro phosphorylation of individual histones and of Hl/H5depleted chromatin was carried out at a protein concentration of 500 Ag/ml in a reaction mixture containing 20 mM Tris-HCl pH 7.0, 5 mM MgCl2 and 0.5 mM [32PJATP (sp. act. 20-60 c.p.m./pmol) (Amersham, UK). Phosphorylation of long chains of native chromatin was found to be more effective in 2 mM MgCl2. Twenty units of the catalytic subunit were used per ml of reaction mixture. The incubations were performed at 35°C for 2 h. The solubility of the chromatin was restored by complexing the magnesium ions with EDTA. Preparation of histones Chicken erythrocyte and calf thymus histones H2B, H2A, H3 and H4 were extracted and purified according to the acid method of Johns (1964). Histones HI and H5 were prepared as previously described (Champagne et al., 1968). Preparation of antisera and monoclonal antibodies Antibodies to histones H2A, H2B, H3, H4 and H5 were raised in rabbits by immunizing with histone-RNA complexes as described by Stollar and Ward (1970). Antisera specific for C-terminal hexapeptide 130-135 of H3, fragments 85-102 and 80-89 of H4 and 36-43 of H2B were obtained as previously described (Muller et al., 1982a, 1983, and in preparation). Monoclonal antibodies C 13, 73 C3, 69 B4 and 41 A5 were prepared and characterzed as described by Laskov et al. (1984) and Muller et al. (in

preparation). Enzyme-linked immunosorbent assay (ELISA) ELISA was carried out with rabbit antisera as described by Muller et al. (1983) and with monoclonal antibodies as described by Laskov et al. (1984). Antigen dilutions used for coating microtiter plates were prepared in phosphate buffered saline (pH 7.4) containing 0.05% Tween 20 (PBS-T) in the case of chromatin and core particles, and in 0.5 M sodium carbonate buffer (pH 9.6) in the case of histones.

Acknowledgements We are grateful to Dr J.Gordon (Basel) for gifts of monoclonal antibodies 73 C3, 41 A5 and 69 B4 and to Dr D.Eilat (Jerusalem) for monoclonal antibody C 13. We thank Drs G.Marchis-Mouren and A.Tirard (Marseille) for providing the cyclic AMP-dependent protein kinase. We also thank Dr Y.Boulanger (Strasbourg) for amino acid analysis and Drs P.Sautiere (Lille) and M.Champagne (Strasbourg) for helpful discussions. The technical assistance of J.Dunand and D.Buhr is gratefully acknowledged. This work was supported by a grant to M.H.V.Van Regenmortel from the Institut National de la Sante et de la Recherche Medicale.

References Absolom,D. and Van Regenmortel,M.H.V. (1977) FEBS Lett., 81, 419-422. Absolom,D. and Van Regenmortel,M.H.V. (1978) FEBS Lett., 85, 61-64. Ajiro,K., Nishimoto,T. and Takahashi,T. (1983) J. Biol. Chem., 258, 45344538. Allan,J., Hartman,P.G., Crane-Robinson,C. and Aviles,R.X. (1980) Nature, 288, 675-679. Allan,J., Harborne,N., Rau,D.C. and Gould,H. (1982) J. Cell Biol., 93, 285-297.

Allfrey,V.G. (1980) in Goldstein,L. and Prescott,D.M. (eds.) Gene Expression: The Production of RNA's, Academic Press, pp. 347-437. Bakayev,V.V., Bakayeva,T.G. and Domansky,N.N. (1981) Mol. Biol. Rep., 7, 209-216. Bentley,G.A., Lewit-Bentley,A., Finch,J.T., Podjarny,A.D. and Roth,M. (1984) J. Mol. Biol., in press. Boulikas,T., Wiseman,J.M. and Garrard,W.T. (1980) Proc. Natl. Acad. Sci. USA, 77, 127-131. Bradbury,E.M. and Matthews,H.R. (1981) in Sarma,R.H. (ed.), Proceedings of the Second SUNYA Conversation in the Discipline Biomolecular Stereodynamics, Vol. II, Adenine Press, NY, pp. 125-144. Bradbury,E.M., Inglis,R.J., Matthews,H.R. and Langan,T.A. (1974) Nature, 249, 553-556. Bustin,M., Simpson,R.T., Sperling,R. and Goldblatt,D. (1977) Biochemistry (Wash.), 16, 5381-5385. Butler,P.J.G. and Thomas,J.O. (1980) J. Mol. Biol., 140, 505-529. Champagne,M., Mazen,A. and Wilhelm,X. (1968) Bull. Soc. Chim. Biol., 50, 1261-1272. Chatterjee,S. and Walker,I.O. (1973) Eur. J. Biochem., 34, 519-526. DeLange,R.J., Williams,L.C. and Martinson,H.G. (1979) Biochemistry (Wash.), 18, 1942-1946. Diaz,B.M. and Walker,I.O. (1983) Biosci. Rep., 3, 283-292. Di Padua Mathieu,D., Mura,C.V., Frado,L.L.Y., Woodcock,C.L.F. and Stollar,B.D. (1981) J. Cell Biol., 91, 135-141. Finch,J.T., Brown,R.S., Rhodes,D., Richmond,T., Rushton,B., Lutter, L.C. and Klug,A. (1981) J. Mol. Biol., 145, 757-769. Goldblatt,D. and Bustin,M. (1975) Biochemistry (Wash.), 14, 1689-1695. Grigoryev,S. and Krasheninnikov,I.A. (1982) Eur. J. Biochem., 129, 119-125. Gurley,L.R., Tobey,R.A., Walters,R.A., Hildebrand,C.E., Hohmann,P.G., D'Anna,J.A., Barham,S.S. and Deaven,L.L. (1978) in Jeter,J.R., Cameron,I.L., Padilla,G.M. and Zimmerman,A.M. (eds.), Cell Regulation, Academic Press, NY, pp. 37-60. Hanks,S.K., Rodriguez,L.V. and Rao,P.N. (1983) Exp. Cell Res., 148, 293302.

Harborne,N. and Allan,J. (1983) FEBS Lett., 155, 88-92. Hohmann,P. (1983) Mol. Cell. Biochem., 57, 81-92. Inglis,R.J., Langan,T.A., Matthews,H.R., Hardie,D.G. and Bradbury,E.M. (1976) Exp. Cell Res., 97, 418-425. Johns,E.W. (1964) Biochem. J., 92, 55-59. Krystal,G.W. and Poccia,D.L. (1981) Exp. Cell Res., 134, 41-48. Kurochkin,S.N., Andreeva,N.B., Gasaryan,K.G., Severin,E.S. and Kurchatov,I.V. (1977) FEEBS Lett., 76, 112-114. Lake,R.S. (1973) Nature New Biol., 242, 145-146. Laskov,R., Muller,S., Hochberg,M., Giloh,H., Van Regenmortel,M.H.V. and Eilat,D. (1984) Eur. J. Immunol., 14, 75-81. Mangeat,P.H., Chaninian,H. and Marchis-Mouren,G.J. (1978) Biochimie, 60, 777-785. Marion,C., Roux,B., Pallotta,L. and Coulet,P.R. (1983a) Biochem. Biophys. Res. Commun., 114, 1169-1175. Marion,C., Roux,B. and Coulet,P.R. (1983b) FEBS Lett., 157, 317-321. Martinage,A., Mangeat,P., Laine,B., Couppez,M., Sautiere,P., MarchisMouren,G. and Biserte,G. (1980) FEBS Lett., 118, 323-329. Martinage,A., Mangeat,P., Sautiere,P., Couppez,M., Marchis-Mouren,G. and Biserte,G. (1981) FEBS Lett, 134, 107-110. Matsumoto,Y.-I., Yasuda,H., Mita,S., Marunouchi,T. and Yamada,M.-A. (1980) Nature, 284, 181-183.

2435

S.Muler et al. Mazen,A. and Champagne,M. (1972) Biochimie, 54, 1273-1279. Mazen,A., de Murcia,G., Bemard,S., Pouyet,J. and Champagne,M. (1982) Eur. J. Biochem., 127, 169-176. Muller,S., Himmelspach,K. and Van Regenmortel,M.H.V. (1982a) EMBO J., 1, 421-425. Muller,S., Erard,M., Burggraf,E., Couppez,M., Sautiere,P., Champagne,M. and Van Regenmortel,M.H.V. (1982b) EMBO J., 1, 939-944. Muller,S., Soussanieh,A., Bouley,J.P., Reinbolt,J. and Van Regenmortel, M.H.V. (1983) Biochim. Biophys. Acta, 747, 100-106. Nelson,N.C. and Taylor,S.S. (1981) J. Biol. Chem., 256, 3743-3750. Paulson,J.R. and Taylor,S.S. (1982) J. Biol. Chem., 257, 6064-6072. Pesce,A.J., Krieger,N.J. and Michael,J.G. (1983) in Avrameas,S., Druet,P., Masseyeff,R. and Feldmann,G. (eds.), Immunoenzymatic Techniques, Elsevier Scientific Publishers, The Netherlands, pp. 127-138. Pospelov,V.A., Svetlikova,S.B. and Vorob'ev,V.I. (1979) FEBS Lett., 99, 123-128. Rill,R.L. and Oosterhof,D.K. (1982) J. BioL Chem., 257, 14875-14880. Ring,D. and Cole,R.D. (1979) J. Biol. Chem., 254, 11688-11695. Romac,J., Bouley,J.P. and Van Regenmortel,M.H.V. (1981) Anal. Biochem., 113, 366-371. Russo,E., Giancotti,V., Crane-Robinson,C. and Geraci,G. (1983) Int. J. Biochem., 15, 487-493. Saccone,G.T.P., Skinner,J.D. and Burgoyne,L.A. (1983) FEBS Lett., 157, 111-114. Stollar,B.D. and Ward,M. (1970) J. Biol. Chem., 245, 1261-1266. Stratling,W.H. (1979) Biochemistry (Wash.), 18, 596-603. Suda,M. and Iwai,K. (1979) J. Biochem., 86, 1659-1670. Sung,M.T. and Freedlender,E.F. (1978) Biochemistry (Wash.), 17, 18841890. Taylor,S.S. (1982) J. Biol. Chem., 257, 6056-6063. Thoma,F. and Koller,T. (1977) Cell, 12, 101-107. Thoma,F. and Koller,T. (1981) J. Mol. Biol., 149, 709-733. Thoma,F., Koller,T. and Klug,A. (1979) J. Cell. Biol., 83, 403-427. Thoma,F., Losa,R. and Koller,T. (1983) J. Mol. Biol., 167, 619-640. Thornton,J.M. and Sibanda,B.L. (1983) J. Mol. Biol., 167, 443-460. Worcel,A. and Benyajati,C. (1977) Cell, 12, 83-100. Wright,E.B. and Olins,D.E. (1975) Biochem. Biophys. Res. Commun., 63,

642-650. Zalebskaya,I.A., Pospelov,V.A., Zalensky,A.O. and Vorob'ev,V.I. (1981) Nucleic Acids Res., 9, 473-487. Received on 16 May 1984; revised on 22 June 1984

2436