ovalbumin and lysyme genes (oviduct). - BioMedSearch

Volume 14

Volume 14 Number Number 88 1986 1986

Nucleic Acids Research Research Nucleic Acids

Isolation of oligonucleosomes from active chromatin using HMG17-specific monoclonal antibodies

Tomislav Dorbic and Burghardt Wittig*

Institut fur Molekularbiologie und Biochemie, Freie Universitat Berlin, Amimallee 22, D-1000 Berlin 33, FRG Received 9 December 1985; Revised and Accepted 17 March 1986

ABSTRXT We report the prepaation of HMOGI 7-containing oligonucleosomes from chicken embryos and from liver and oviduct of laying hens. Monoclonal antibodies apinst HMO1 7 were used for their isolation. An unusual size distribution with respect to their repeat number was observed. The oligonucleosomes of repeat number up to N6 were highly enriched for DNA of the vitellogenin 11 ge (liver) and for DNA of the ovalbumin and lysyme genes (oviduct).

INTRODUCTION For the correct unWerstanding of the mechanisms of gene expression in eukaryotic cells it is indispensable to have at hand those limited fractions of chromatin which, in a certain situation, are actually transcribed In recent years, a number of characristics of active chromatin have been published (histone modification, nuclease sensitivity, non-histne proteins [especially the high mobility group, HMO proteins], etc.; for erence se [ 1]). On the basis of the geral nucleas sensitivity and the HM017/14 content of active chromatin, among others, Weintraub and coworkers [2,3] suceded in Isolating mononuclosomes (Ni) and nucilosomal core particles (NC) from active globin domains of chicken erythroyte chromatin. In addition to other criticisms brought forward ainst these results [4,5], their significac is certainly limited by two facts: only Ni and NC particles are obtained; eryhrocytes are terminally differentiated cells containing histone H5 in their chromatin. In our present Investigation we also make use of the fact that active chromatin contains HMO1 7. But insted of utilizing the recognition by HM01 7 of those nucleosomes which previous to their dissociation had already contained HMiOl 7 (of which the spificity was questioned in literature) [4,5] we used the high spificity of a monoclonal HMOI 7-antibody for the isolation. This enabled us to finally obtain oligonucleo es of r t number up to N6 from active chromatin of two specialized tissues. In hybridization experiments with probes for lysoyme, ovalbumin, and vitellogenin we could prove the presence of the three genes in the respective HMGI 7-containing active chromatin.

MlETHODS Immunization and monoclonal antibodies Balb/C mice (6 weeks old) were injected in 2 week intervals with a) 50 jgg of a preparation of HMO17/14 [61 from chicken embryo nuclei, b] 100 Ig of a crude HMO preparation from chicken red © I R L Press Limited, Oxford, England.

3363

Nucleic Acids Research bloodoells [7] containing also HMGJ/2 and histones H1/HS, c) 50 ig of the preparation b, d)10 1ig of a highly purified HM017 preprtion from chicken embryo nuclei (5]. All antigen preparations were mixed with complete Freund's adjuvant and Injected intraperitoneally. After the third injection the antibody titer was determined by ELISA using preparation b) as the antigen. Four days after the fourth injection the spleens of two mice were mechanically dissociated and the cells fused with myeloma cells (strain P3-X63-Ag8). Cell fusion, seWlction and propagation of hybridoma clones as well as ascites production were performed by standrd methods [8,9]. A first screening of hybridoma clones was by ELISA as above. Positive hybridoma supernatants were then selectd for indirect immunoprecipitation of HMO 17 using formalin-fixed Staph. A cells (Immuno-Precipitin, BRL/Oibco). Precipitates were analyzed by SDS-PAGE (see below). The antibody preparations employed for immunoprecipitation of nucleosomes were either (NH4)2S04 concentraed cell culture supernatants or ascites fluids purified by chromatogaphyon Affi-Oel Blue (BIO-RAD) [ 10]. Isolation of nuclei and rordation of chromatin Nuclei were prepared from 13 d old chicken embryos essentially as described earlier [ I 1]. This procedure was slightly modified for nuclei from liver of laying hens (centrifugation through 52 2 (w/v) sucrose instead of 58 2) and for nuclei from the oviduct of laVing hens (4 times washing in 0.25 M sucrose solution Instead of twice). Digestion of chromatin from the three tissues investigeW here with Staphylococcus nucleae (MNase) was identical to our publisd procedure [ I 1 ,12]. Following removal of nuclar debris the suspension was edjusted to 0.15 M salt with NaCl. 30 ml a]iquots were layered on a dDuble-l ercof sucrose (5 ml of 10 % and 5 ml of 50 X (w/v) sucrose in 130 mM NeCl, 20 mM KC1, 1 mM EDTA, 5 mM Tris-HCl, pH 7.4. After centrifugation at 85,000 x g and 2 °C for 18 h in a SW28 rotor (Beckman Instruments) the bottom 9 ml were recovered and dialyzed twice against 2 1 of IP-buffer (50 mM Nil, 5 mM EDTA, 50 mM Tris-HCl, pH 7.4). Insoluble material was pelleted ( 2,000 x g, 4 °C, 10') and the nucleosome preparation stored in aliquots at -70 OC. Immunoorectietation of nucleosomes end gAl electroghoresIs 1. Immunoprecipitation for the analyss of nucloosomal proteins: 0.2 A260U of chromatin and 10 PI ( 0.5 mg/ml) of a HMO1 7-specific monoclonal antibody prepaation (either cloms 3-C-2, 4-C-5, 9-B-2, 9-C-2 or a mixture of the four) were incubated in IP-buffer containing 0.25 2 (w/v) bovine serum albumin (IP/BSA-buffwr); incubation was for 3 h at room teperature in a volume of 0.1 ml. Following incubation the 0.1 ml were d to a pellet of formalin-fixed Staph. A cells (Immuno-Precipitin, BRL-Olbco; 0,1 ml of the 10 2 (w/v) suspension of Immuno-Precipitin were pelleted for 1', 15,000 x g, Microfuge), the pellet sued by pipetting, and the incubation continued for 1 h at 4 *C (shortly vortexing every 10'). The suspension was then pelleted as above, the supernant discarded, and the pellet resded in 0.2 ml of IP/BSA-buffer by pipetting. This washing produre was repet twice, the final pellet suspended in 50 lil of cracking/loading solution (2 mM EDTA, 1 s (w/v) sodium ylsulfate (SDS), 1 2 (v/v) 2-mcaptoethaol, 30 2 (v/v) glycerol, 0.05 2 (w/v) bromphenol blue, 50 mM Tris-HCl, pH 8.0), and incubatd for 3' at 95 "C. Immuno-Prscipitin was pelleted as above and the supernatnt containing the disociated proteins loed on a 3364

Nucleic Acids Research

SDS-polyacrylamide gel (SDS-PAGE) as decribed by Laemmli [ 13] (4.75 2 (w/v) acrylamide stacking gel, 15 % (w/v) acrylamide running gel, 0.1 2 (w/v) SDS; gel dimensions: 500 mm x 250 mm x 0.6 mm). The gels were silver-stained employing a published procedure [ 14] with our own modification of the flxing procedure (incubation in: 1. 1 l of 50 (v/v) methanol, 12 (v/v) acetic acid for 3 h, 2. 1 of I0 (v/v)ethanol, 52 (v/v)acticacidfor20', 3. 1 1 of H20for 20', 4. 0.5 l of 1 (v/v) glutaraldehyde, 50mM Na2B407 for 20', 5. 3 times in 11 of H20 for 20' each ). 2. Immunoprecipitation for the preparation and analysis of nucleosomal nucleic acids (fractions HMe-N, SUP, and PEL, definition see below): 10 assas, each containing 0.4 - 0.5 A260U of chromatin and 10 ml of HMOG 7-specific monoclonal antibody preparation were incubated as described above. Then, the Immuno-Precipitin was pelleted ( 1', Microfuge) and the supernatant recovered The pellet was resuspended in 0.2 ml IP/BSA-buffer, centrifuged, and the supernatant recovered and combined with the first supernatant; we will refer to this fraction as SUP. The pellet was resuspended in 0.2 ml of IP/BSA-buffer containing 0.2 mM NaCl (instead of 0.05 mM), centrifuged, and the supernatant discarded. The pellet was resuspended in IP/BSA-buffer containing 0.4 M NaCl, incubated for 20' at room temperature, centrifuged, and the supernatant recovered. The pellet was resuspended in 0.2 ml of the same buffer, centrifuged, the supernatant recovered and combined with the previous supernatant; we will refer to this fraction as HMO-N. The remaining pellet fraction will be referred to as PEL. The H1M-N and SUP fraction were deproteinized by proteinase K digestion (0.2 mg/ml, 0.17 2 (w/v) SDS) and twice by phenol extraction. The PEL fraction was suspedd in 1.7 2 (w/v) SDS, 0.3 M NaCl, 40 mM Tris-HCI, pH 7.5, and then deproteinized as both other fractions (the proteinase K concentration here being 2 mg/ml). Nucleic acids in HMO-N, SUP, and PEL were precipitated with ethanol, dissolved and aliquts dd on a 1.5 2 (w/v) horizontal agarose gel (300 mm x 250 mm x 5 mm, "submarine" in the Tris/acetate-buffer system [ 15]). Hae III digsted (P X1 74 RF DNA and the "ladder" of nucleosomal DNA of the starting chromatin material served as length standards. Gels were stained with ethidium bromide ( I pg/ml). Hybridization exoeriments and uauntification. From the agerose gels from corresponding positions equal bands were cut out according to the starting

chromatin "ladder", electro-eluted with the help of NACS 52 (BRL-Gibco) minicolumns, precipitated with ethanol, the precipitate dissolved in 100 j.l H20, and the rovery of nucleic acids measured by UV-spectrophotometry. Aliquots of these samples sufficient for dot-blot hybridization were precipitated with ethanol and processed for dot-blotting [15] (except of quantitatively scintillation counting of hybridized samples instead of scanning x-ray photographs) on GeneScreenPlus membranes (New Englano Nuclear). From each sample of nucleosomal DNA we applied to the membrane: 10 gg of strting material DNA, 20 jig of HMO-N DNA when probing for an inactive gene, or 5 jig of HMG-N DNA when probing for an active gene.The membranes were covered with "Frischhaltefolie" Frappan (an equivalent to Saran Wrap) and irradiated with UV-light for 3' using a disassembled transilluminator at a distance of 100 mm [ 161. Hybridization (65 *C, overnight) and washings were performed in the buffer systems used for "genomic sequencing" [ 16]. Hybridization buffer was 1 2 BSA (crystalline grade, Sigma Chemie), 7 2 (w/v) SDS 3365


A

2 A *

,,*

.,

..

I.-

i;

k-aI

.._ iF

IVW,

.:....

AI -i,11111~*.~_...:. ~~~.P-~ v ~~~~

'O.

-ow-ft

_ 4bw,___ 0 In "O0_6. .-o,* ~ _ a __ wo

3

l

C-, -.1

h-S

b -J

~-

cli

iII

---HMG14 HMG17 --H3 --HMG 17I -H2B

--H2At -H4

3366

14 M. C, -

-.!Al4

_'4t

2

B

I4 %S

.41

_

3

*

I:-

H MG, -.' IH N4 C,

''


(recrystallized from ethanol),1 mM EDTA, 0.5 M "NaHP04", pH 7.2. [32P]-labelled RNA-probes with >5 x 1 08 cpm/gig at 5 x 107 cpm/ml were emplyed. Membranes were washed three times for 5' each at 65 C in WB I ( 0.5 % (w/v) BSA fractlon V grade, 5 % (w/v) SDS, 1 mM EDTA, 40 mM "NaHP04", pH 7.2, three times under the same conditions in WB2 ( 1 % (w/v) SDS, 1 mM EDTA, 40 mM "NaHP04", pH 7.2, twice under same the conditions in WB3 ( 100 "NaHP04", pH 7.2), and finally in WB3 at 65 *C for 20'. RNA probes were synthesized using pSP64/65 recombinants and SP6 RNA-Polymerese (New England Nuclear and Promega Blotec) as described in [ 17], employing [a32P]-UTP (3000 Ci/mmole, 10 mCi/mi, New England Nuclear; no unlabelled UTP added). The RNA-probes for hybridization experiments had the following extensions relative to the starting point of transcription ( + 1 ): Lysozyme base positim +290 to -+400; ovalbumin base position + 155 to -'+250; vitellogenin base position +58 to + 138.

RESULTS In order to obtain a specific tool for the isolation of HMO1 7 and/or HMO1 4 containing nucleosomes, a monoclonal antibody preparation against one of these two non-histone proteins was needed. Cell culture supernatants of trybridoma clones were screened for antibodies which show a high specificity for HMO1 7

and an effective binding to crosslinked cell wall: protein A-complexes of Staphylococcus aureus (Immuno-Precipitin). Fig. lA shows SDS-PAGE analysis of immunoprecipitates with aliquats of highly purified chicken embryo HMO1 7/HMO14. The electrophoresis of dissociated HMO: antlbody-complexes from a random collection of Ig-secreting hybridoma clones exhibits only HMO 17-specific antibodies. Among some 50 such samples investigated, only a few also showed crossreaction with HMO14 (not shown here). Chicken embryo HMG17 shows two bands in SDS-PAGE, one in a position corresponing to the regular species. The second one (HMO17' in the HMO control track in Fig. IA) migrates slightly faster and may be a subspecies or a modifled form (G. Ooodwin, personal communicatlon). In some cases the

FLU.1: A. Screening of hybridoma clones for HMOGI 7-specific antibodies. SDS-PAGE was performed on dissociated immunoprecipitates (complexes of HMOi 7-protein, monoclonal antibodies [AB], and Immuno-Precipitin). 10 Ill of hybridoma supernatant, 10 jgg of HM017/HMG14-preparatlons from chicken embryos, and Immuno-Precipitin were incubated and procsed as described in METHODS. a: Immuno-Precipitin remnants; b, c: Culture medium components; 1: Large subunits of HMO1 7-specific AB; s: Small subunits of HMG I7-specific AB; Track HMO : Control with 4 jig of HMOI 7/14 -preparation (used as antigen in screening) only. Antibody preparations from ascites forms of the *-marked clone were usd for further experimentation. B. SDS-PAGE of dissociated immunoprecipitates employing supernatants from non-AB-producing hybridoma cells and Immuno-Precipitin. 1, 2, 3 : (el tracks from increasing amounts of supernatant; a: Immuno-Precipitin remnants; b, c: Culture medium components. C. SDS-PAGE of dissociated immunoprecipitates of chicken embryo oligonucleosomes with HM017specific AB. 1 : Precipitation as described in METHODS (Immunoprecipitation for the analysis of nucleosomal

proteins.

2: Same as 1, but prior to immunoprocipitation oligonucleosomes were incubated with 0.4 M NtCl (to renove HMO proteins) and centrifuged through a 10 X (w/v) sucrose lwr. 3: Same as 2 using 0.6 M NCl (to remove additionally histone HI). b: Culture medium components; ch: Chromatin proteins; s: small subunits of HM01 7-specificAB. 3367


~~~~~~ ~~~~~~~~~~~.~~~7

Track 1 is starting material chromatin as a reference, in A 3 jig, in B 6 jig. Tracks 2 to 6, immunoprecipitates after exposure to 50 mM, 1 50 mM, 350 mM, 400 mM, 600 mM Nacl (A), to 15 pmole, 60 pmole, 150 pmole, 240 pmole, 360 pmole HMOGi 7 (B).

HMOGI 7-specific antibody, binds to both HMOi1 7 species. Control electrophoreses shown in Fig. 1 B allow fortheidentifeationoftheotherbandsasdecl&ardinthelegendtoFig. lA. BandcinFig. B as also in Fig. IA represents culture medium components (with non-specific binding to Immuno-Precipitin) not present in monoclonal antibody preparations from corresponding ascites fluids. All our main subsequent experiments were performed using the ascites form of clone 3C2, 4C5, 9B32, 9C2. In Fig. 1 B we show three tracks ( 1,2,3) employing increasing amounts of culture medium. The bands with increasing intensity through 1 ,2,3, therefore represent culture medium comnponents (designated as b in Fig. 1A,B,C). Remnants of Immuno-Precipitin are designated as a. In Fig. iC we show SDS-PAGE of chicken embryo oligonucleosomes precipitated with HMOI 7-specific monoclonal antibodies. The result of these experiments is that pretreatment of oligonucleosomes with 0.4 or 0.6 M NaCI leads to an almost complete loss of the core nucleosome histone bands. This is due to the fact that 0.4 M NaCl removes HMO 1 7 from the nucleosomes and thus renders themn unprecipitable by HMO 17-specific antibodies. 0.6 M NaCl would remove even histone HI1. These observations will be relevant to the discussion of unspecific binding of Hi1 containing nucleooes. Fig. 2A shows SOS-PAGE analyses of immunoprecipitated nuclesoe preparations which were treated with differet NaCl concentrations. We know (own experiments and from literature) that es do not dissociate from HM*i7 nuc.i at NaCl coentrations below 0.35 M. Therefore, the silve stained bands from the assay's with 0.35 M most of the nucleosmxal material Is relese into the supernatant; conseqetly, only trwe amounts of coe histones remnain visible in the precipitated fraction. In order to proyide more evidenc for -selective disengageent of the nucleooms from the HMOG1 7: antlbody-comnplex, we exposed the nucleoscxme: HMO-bonds within the Immunoprecipitates to an inreasing competitive excess of HMOGI 7-protein. Fig. 2B demonstrates that the comnpetition with HMGi1 7 relae nucleosomes from the precipitate when at a 5- to 10-fold molar excess. Prime facie, It 3368


I. Fi ii.

ii!

fjg. 3: Nucleic acid content of nucleosornal particles from chicken embryos (meon of 4 measurements) and oviduct and liver from laying hens (moan of 3 measurements). On the Nucleosome Repeat No.-categry axis, (NC means subnucleosomal particles, NC nucleosome core particles. The insert shows an agarose gel electrophoresis track of oviduct nucleosomes, the brackets indicating the material excisW for analysis.

msems surprising that neither in Fig. 2A nor in 2B bands for HMG1 7 itself can be seen. This is explained simply by two facts. First, HMOJ 7-protein is present, at the most, at half the amount compared to core histons [6,18]. Second, it is known that the silvwr staining method is, with respect to HMOi 7, much less seitive than to histones [ 14,19, own observations]. Regardless how strong these arguments msy be, the relevant result is that with incraing HMO1 7 concentration nucleosomes are I relsd from a complex containing HM017. In experiments not shown here we obtained quite similar results when anplcVing a polyclonal antl-HMO 17 serum (provided by M. Bustin) [20]. The persisting fraction of material containing core histones seen in Fig. 2A and B reflects unspecifically insolubilized subnucleosomal particles. We shall refer to this effect later in the context of transcribed nucleotide sequenes within these nucleosomal particles(especially in PEL). Having establishd a method to selectively isolate HM31 7-containing nucleosomal particles, we proe to detrmine the number distribution of their repeat units. In order to have a ference sstem for the HM017-containing mono- and oligonucleosomes, we first analyzed (by agarose gel eletrophoreis) the DNA-fragments isolated from nucleosomal prepaations of chromatin from chicken embryos, from liver of laying hens and from oviduct of laying hens after digestion with Staphyl Ioccal nuclems (MNe). As Fig 3 clarly dmnstrates, the distribution of DNA repeat lengths of the respective mono- and oligonucleooms up to N8 is indistinguishable for the three tisu investigated Only in the case of liver the amount of nucleic acid fragments from subnucleosomal particles is significantly larger. It will be relevant for further discussion that this fraction consists in 30 % of RNA and 70 % of DNA For the two specialized tissues, liver and oviduct, the mme analysis was performed with respect to the HlMO 1 7-containing nucleosomes enplaying immunoprecipitation with a HMG 1 7-specific antibody 3369

Nucleic Acids Research Table 1. Recovery of nucleic acids In SUP, HMO-N, and PEL after immunoprecipitation. Chromatin rrom Liver

CompIeteAways

Input

30

atin

I I26U 15

CompleteAssays

Assays minusAB

No. of Sumof

Ass

Chromatin from Oviduct

No. of

As" 100

Sum of

No. of Sum of

A260U

%

A

A260U

Am

50

5.0 100

10

Assays minus AB Sumof T

SAm"A260U 9.0

20

22.5 100

3

100

After immunoprecipitation and fractionation into SUP, HMO-N, and PEL, 10 essaVs each were pooled SUP

3 1

7.1471

1

3.4 68

5

11.8 521

2

HMO-N

3f

2.8

19

1 j

0.2

4

5

5.9 261

2

0.25

3

PEL

3

4.2

28

1

1.0

20

5

3.7 16

2

1.1

12

d | N¢C*oiA low WIO- N Ouct

77

B

A

0SPLi

16.00 14.00

HMO-NLlvow

2.00

7.0

.,00 1,60

12.00

1.40

10.00

1.20

0.00 1,00

6.00

0,00

4.00 2.00

0,60 0,40

%P

7

0.00

0.20

'C

0,00

I

C

NC

13

NI

16

N4

IaN

NC.Ni

N7

NucbuumPNpu

2

144

I N5 N

47

NB

K~~~~~~~~~Iosmb p,Al& lb

I No,

D 10.00* [PePSoL l.00

1| PooL 0

6.00

I

UK

LW

Nh. Nil

*

N;

4.00

N;i N1~

2.0

.N :s

NH

D- N7 Ne :'- No -N4

_N3 _ N2 Ni

'C

NC

Ml

112

Nc

N4

147

wxu

Reps& lb.

NC C

FiA4: Nucleic Acid content of A HMO-N, B SUP, C PEL nucleosomal fractions (in all c men of 3 measurements), and D exanples of agerose gel electrophoresis tracks; 1, 2, 3: HMG-N, SUP, PEL from liver; 4, 5, 6: HMO-N, SUP, PEL from oviduct; c is a antibody-minus control of HMO-N.

3370

Nucleic Acids Research preparation. This differentiates the chromatin into three fractions: the supernatant from the precipitation step (SUP), the HMG17-containing nucleosomes after dissociation from the precipitate with 0.4 M NoCl (HMO-N), and the pellet remaining therefter (PEL) (as documented in Table 1). Here we show the quntitative recovery and relative distribution of DNA between the chromatin starting material and the three fractions. The corresponding values of minus antibody controls demonstrate clearly that almost all of the HMO-N DNA appears in SUP when HMOl 7-specific antibody is omitted in the precipitation st. On the other hand, the nucleic acid content of the PEL fraction is hardly affected by the presence of HMOlI 7-specific antibodies during precipitation. The agmse gel electrophoresis of the nucleic acids from the three fractions (Fig 4) reveals a striking difference in number distribution of repeat units in HMO1J 7 mono- and oligDnucleosomes. For the SUP fraction from liver as well as from oviduct the distribution is (with the exception of subnucleosomal particies) essentially the same as in the starting material (Fig. 48 and Fig. 3). As the electrophoresis of PEL shows (Fig. 4C) this fraction is not merely an undefined background (as might have been insinuated by the PEL recovery values In Table 1), but rather consists exclusively In subnucleosomal (