Eur. J. Biochem. 128. 209-213 (1982) Q FEBS 1982

The Participation of Poly(ADP-Ribosy1)ated Histone H1 in Oligonucleosomal Condensation

May WONG. Najma MALIK, and Mark SMULSON

Department of Biochemistry, School of Medicine and Dentistry, Georgetown University, Washington DC

(Received March lh / Ju ly 26, 1982)

The chromatin-associated enzyme poly(ADP-Rib) polymerase causes an NAD-dependent crosslinking of modified oligonucleosomes, as demonstrated by electrophoretic and sedimentation analysis [Butt, T. R. and Smulson, M. (1980) Biochemistry, 19, 5235- 52421. It was speculated that poly(ADP-ribosy1)ation of histone H1 and subsequent formation through crosslinking to an H1 dimer may be an important component of this phenomenon. To study this process, a method of complexing histone HI to chromatin was required that promoted the restoration of accurate poly(ADP-ribosy1)ation of this histone. Previously we have established that two histone H1 molecules are crosslinked by a chain of poly(ADP-Rib) 15 or 16 units in length. In the current study, we made use of the ability of oligonucleosomes, reconstituted with H1, to carry out the synthesis of the poly(ADP-Rib)-HI complex in order to monitor the accuracy of reconstitution. It appears that a specific distance and juxtaposition of adjacent H1 molecules along the polynucleosome fiber is required for the enzymatic synthesis of this modified histone complex.

We established that a controlled trypsin digestion of oligonucleosomes removed H1 histone with minimal perturbation of other nuclear proteins associated with chromatin. In addition, poly(ADP-Rib) polymerase was partially removed from chromatin by this procedure. Subsequently, methods utilizing gradient salt dialysis have been employed to reconstitute both the polymerase and histone H1 to the depleted oligonucleosomes.

The reassociation of H1 (and polymerase) to specific binding sites within oligonucleosomes was accomplished by the above procedures. Poly(ADP-Rib) - H1-dimer synthesis was not observed in depleted oligonucleosomes, but this capacity was found to be partially restored in the reconstituted chromatin. Similiarly, the ability of NAD to promote crosslinking of nucleosomes was restored in the reconstituted samples. These results provide a basis for further studies on how the poly(ADP-ribosy1)ation of histones alters the structure of chromatin.

The chromatin-associated enzyme, poly(ADP-Rib) poly- merase, catalyzes the successive transfer of the ADP-ribose moiety of NAD to various nuclear acceptor, including core nucleosomal histones, histone H1 and a 112-kDa non-histone protein, shown to be poly(ADP-Rib) polymerase [I]. With increasing concentrations of NAD, the polymer reaches chain lengths exceeding 100 units long [2]. It is established that a chain of 15 or 16 ADP-Rib units crosslinks two histone H1 molecules [3]. This reaction could thus represent a consider- able physical alteration of the structure of those nucleosomes undergoing this modification, since recent data suggest that core histones are also crosslinked upon extensive poly(ADP- ribosy1)ation [2,4]. These reactions could account for the long chains of poly(ADP-Rib) noted in vitro and in vivo, especially if nucleosomes distal from each other were con- nected by poly(ADP-Rib). However, it should be pointed out that this has not been clearly established.

Recently, we have demonstrated, in v i m , that when oligo- nucleosomes are incubated with concentrations of [32P]NAD 10 pM and higher, the labeled poly(ADP-ribosy1)ated nucleo- somes migrate in native polyacrylamide gels to positions of far greater size than the bulk of the nucleosomes [2]. This apparent condensation of nucleoprotein complexes was also demonstrated by an increase in the sedimentation of these ADP-ribosylated chromatin components in velocity sucrose gradients, where the sedimentation coefficient of complexed nucleosomes was noted to be approximately 65 S while that

Ahhreviurion. ADP-Rib, adenosine(5’)diphospho(5)-~-~-ribose.

of unmodified nucleosomes was 40 S. A direct correlation between NAD concentration, the level of chromatin aggrega- tion and the length of poly(ADP-Rib) chains in the complex was detected.

Stone et al. [3] have shown that ADP-ribosylation of histone H1 leads to the dimerization of two HI molecules connected by a 15-unit-long poly(ADP-Rib) chain. More recently, we have studied in detail the synthesis of this H1 dimer in purified oligonucleosomes [5 ] . Larger polynucleo- somes were found to be more effective in H1 dimer con- struction than shorter chromatin fragments. It was of interest that optimal poly(ADP-Rib) - H1-dimer synthesis in vitro occurred at those concentrations of NAD favoring chromatin aggregation [6].

Attempts were subsequently made to assess whether ADP- ribosylated histone HI plays a key role in the oligonucleo- some condensation reaction described above. While depletion of HI from oligonucleosomes abolished the NAD-promoted aggregation, attempts to restore the complex formation by the reconstitution of H1 to the depleted chromatin were un- successful [6]. A stringent criterion for the faithful reconsti- tution of H1 to depleted chromatin was required. We have principally addressed this problem in the work presented below. Poly(ADP-Rib)- H1-dimer formation in the HI - reconstituted oligonucleosomes has been investigated. The primary aim has been to establish whether H1 crosslinking by poly(ADP-Rib) plays a major role in chromatin aggrega- tion. This will be of significance because of the strong evidence implicating this histone with the maintenance of higher- ordered chromatin structures [7].

210

MATERIALS AND METHODS

Materials

[32P]NAD was purchased from New England Nuclear (specific activity 32-56 Ci/mmol); ‘251-protein A was ob- tained from Amersham. Micrococcal nuclease was purchased from Worthington; bovine serum albumin, trypsin and soy- bean trypsin inhibitor were products of Sigma Chemical Co.; Dowex 50WX2 was obtained from Bio-Rad Inc. Histone H1 antibody was a gift from Michael Bustin. Kodak SB5 X-ray film was used for autoradiography. HeLa S3 cells were main- tained in suspension cultures at 37°C in spinner flasks. All the cells used in the present studies were grown asynchronously to a midlog density (8- 10 x lo5 cells/ml of culture medium).

Preparation of Chromatin .from HeLa Nuclei

HeLa cell nuclei were prepared by the method of Sporn et al. [8] and washed with 0.25 M sucrose. 5 mM Tris/HCl, pH 7.5, 1 mM CaC12 and 80 mM NaCl containing Triton

The nuclei were incubated with nuclease digestion buffer at lo8 nuclei/ml and treated with micrococcal nuclease (30 units/108 nuclei) for 3 - 5 min at 37 “C. The reaction was terminated and the nuclei were lysed with 1 mM EDTA as described previously [9]. Under these digestion conditions, 10- 12 ”/, of the chromatin DNA was rendered acid-soluble. Chromatin was separated on a 10-30”/, linear sucrose gradient as described previously [9,10].

X-100 (0.3 %).

Analysis ?f Acid-Soluble Proteins on Acetic AcidlUrea Gels

Chromatin fragments were labeled with [32P]NAD and extracted for histone H1 and H1 complex as described by Nolan et al. [5]. Acetic acid/urea gels (1 5 % polyacrylamide) were performed according to Panyim and Chalkley [ll]. The gels were stained with amido black, destained with 40% methanol, 7% acetic acid, dried on a Bio-Rad model 224 slab gel drier and exposed to X-ray film an -80°C for appropriate time periods.

Native Chromatin Gels

[ 32P] Poly (ADP - Rib) - containing chromatin fragments were loaded onto a 2.5% polyacrylamide gel (acrylamide/ N,N‘-methylenebisacrylamide, 20 : 1, w/w) containing 0.5 ”/, agarose which was pre-electrophoresed for I h at 150 V. When very low concentration [32P]NAD (i.e. picomolar) is utilized in these experiments, the labeled nucleosomes migrate identically with bulk nucleosomes since only short chains of poly(ADP-Rib) are generated. When 10 pM [32P]NAD is used, long chains of poly(ADP-Rib) are synthesized, cross- linking these nucleosomes, and they can be distinguished from bulk chromatin by their slower electrophoretic mobility [2]. The gel buffer contained 89 mM Tris, 89 mM boric acid and 2.5 mM EDTA at pH 8.3. The gel was stained with ethidium bromide, fixed with 7% acetic acid and processed for autoradiography.

Reconstitution of’ Histone HI and Polymerase with Oligonucleosomes

One molecule of HI per nucleosome was estimated as equivalent to 0.19 mg Hl/mg DNA [12]. This assumes an average molecular weight of histone HI of 24000 and an

average of 190 base pairs of DNA per nucleosome [12]. Histone-H1-depleted chromatin. HeLa HI and semipurified poly(ADP-Rib) polymerase, obtained as described earlier [5,13], were combined and dialyzed against a solution con- taining 0.25 M KC1, 1 mM EDTA, 1 mM dithiothreitol, 50 mM NaHSO3 and 10 mM Tris/HCl pH 7.0 for 4 h. The buffer was then changed to a solution containing 80mM KCI, 1 mM EDTA, 1 mM dithiothreitol, 50 mM NaHS03 and 10 mM Tris/HCl pH 7.0 for an additional 4 h and to another solution containing 1 mM EDTA, 1 mM dithio- threitol and 10 mM Tris/HCI pH 7.0 overnight. Chromatin samples were subsequently passed through a Sephadex G-50 column to remove that fraction of histone H1 which did not bind to chromatin.

RESULTS Selection of Conditions for H I Depletion

Van Lente and Weintraub [14] described the use of con- trolled trypsin cleavage to remove selectively various loosely bound proteins from chromatin. One major complication of this method might be secondary proteolytic cleavage of non- histone proteins including possibly poly(ADP-Rib) poly- merase. Oligonucleosomes were treated with three concen- trations of trypsin (0.5, 1 and 5 Fg/ml) for various time periods. The reactions were terminated with soybean trypsin inhibitor and the chromatin analyzed for histone content. Almost total H1 depletion was achieved with trypsin at 1 pg/ml for 5 min (see Fig. lB, lane 2; stain). H1 was removed from all classes of oligonucleosomes as determined by a highly sensitive immunological procedure [15]. No significant loss of core histones was noted using this procedure. Gel analysis indicated some cleavage of amino-terminal peptides of core histones ; however, under the limited digestion conditions employed, the majority of the core histones remained intact. In addition, identical results for all experiments provided here have also been obtained utilizing chromatin depleted of HI by the use of DNA-agarose.

The activity of poly(ADP-Rib) polymerase was followed during the time course required for H1 depletion of oligo- nucleosomes. All concentrations of trypsin studied were found to cause varying degrees of reduction in enzymatic activity. It was consequently decided to use trypsin at 1 pg/ml ( 5 min) for future studies, since under these conditions sig- nificant removal of H1 was achieved with the retention of approximately 20 ”/, of the polymcrase activity. Howevcr, be- cause of the loss in enzymatic activity, it was clear that it would be necessary to reconstitute not only H l , but also purified poly(ADP-Rib) polymerase to stripped polynucleo- somes in order to perform the experiments anticipated.

Reconstitution of H 1 and Poly (ADP-Rib) Pol-vmerase with HI-Depleted Chromatin

The following experiment was designed to determine whether the poly(ADP-Rib)-induced crosslinking of histone H1 play a role in the aggregation of chromatin described earlier [2]. Such an experimental approach would require (a) polynucleosomes depleted of H1 but reconstituted with poly(ADP-Rib) polymerase and (b) stripped nucleosomes to which both histone HI and polymerase have been reconsti- tuted. Specific criteria were required to establish that HI and polymerase had been reconstituted to polynucleosomes in a similar manner to their association with native chromatin.

21 1

Table 1. The activit! of poly(ADP-Rib) polymerase in H1-depleted oligo- nucleosome after reconstitution with histone H I and semi-purified poly- merase Oligonucleosomes were depleted of histone H1 using trypsin (1 pg/ml). H1-depleted chromatin was then reconstituted with both purified HeLa histone H 1 and semi-purified poly(ADP-Rib) polymerase as described in Materials and Methods

Oligonucleosomes p2P]NAD incorpored

Untreated HI-depleted Reconstituted

counts min-' A260 unit-'

1200 000 70000

720 000

Jump et al. [13] recently provided conditions for the reconsti- tution of purified polymerase with nucleosomes. Isolated oligonucleosomes, stripped of endogenous poly(ADP-Rib) polymerase activity, were reconstituted by step salt-dialysis with purified HeLa polymerase. The enzyme was shown to bind the internucleosomal DNA of polynucleosomes by both velocity sedimentation analysis and by micro- coccal nuclease digestion studies [13]. Accordingly, in the experiment shown in Table 1, a modification of the con- ditions established earlier was employed to reconstitute both histone H 1 and poly(ADP-Rib) polymerase to trypsin-treated oligonucleosomes. A stoichiometric amount of purified HeLa histone HI was used for reassociation. The sample of HeLa cell histone H 1 utilized in these experiments yielded essentially only bands in the H1 region when a highly overloaded sample was subjected to electrophoresis (data not shown). After dialysis. excess H1 and polymerase were partitioned from oligonucleosomes by gel filtration through Sephadex G-50. The reconstitution was considered successful since 60 % of the specific activity for poly(ADP-Rib) polymerase was regained after reconstitution (Table 1).

Fidelitj, of HI Reassociation with Nucleosomes I t was critical for the interpretation of later experiments

that during the reconstitution procedure, H1 was shown to bind specifically and stoichiometrically to various classes of nucleosomes. Random binding to contaminating proteins, DNA or chromatin aggregates might have arisen during dialysis. Direct quantitative electrophoretic analysis of the H 1 content of reconstituted dinucleosomes, trinucleosomes and oligonucleosomes was not feasible due to the limitation of material and low level of sensitivity of histone analysis. However, gel analysis indicated that the amount of H1 in reconstituted chromatin was approximately the same as in native chromatin (Fig. lB, lanes 1 and 3; stain).

After reconstitution, nucleosomes were electrophoresed and transferred to nitrocellulose by the method of Towbin et al. [16]. The blots were subsequently tested qualitatively and quantitatively for the presence of polymerase and histone HI by the use of antibodies to these proteins (data not shown). Despite the complications associated with such analysis, the data indicated that both polymerase and H1 reassociated with H1-depleted oligonucleosomes to about the same extent, per oligonucleosome chain length, as found in native chro- matin.

Effect of HI Reassociation on Modijication Profiles

A most exacting criterion for assessing the proper alig- ment of H1 after reconstitution would be an efficient restora- tion of poly(ADP-Rib) - H1-dimer synthesis in vitro. This H1 adduct has been shown to be connected by 15 or 16 ADP-Rib units [3]. This fact suggests that a specific distance and juxta- position of adjacent histone H1 molecules along the poly- nucleosome fiber is required for the enzymatic synthesis of this complex. We have recently established the conditions required for H1 dimer synthesis in isolated nucleosomes [5,6].

To establish the proper conditions for HI dimer synthesis for the current study, the experiment of Fig. 1 A was per- formed using native chromatin. Oligonucleosomes incubated with 0.1 pM [32P]NAD synthesized predominantly lower multimers of ADP-ribosylated H I (lane 1). These inter- mediates were chased into larger HI derivatives by the addi- tion of 100 pM nonradioactive NAD by 1 min (lane 2 ) and into mainly H I dimer, by 5 min (lane 3). Since only ["'PINAD was present in the substrate, the autoradiogram will detect only poly(ADP4bosyl)ated histone HI.

This data confirms that NAD concentrations in the range of 100 pM favor H1 dimer synthesis. When HI-depleted oligonucleosomes were incubated with 100 pM [32P]NAD (Fig. 1 B, lane 2) no ADP-ribosylation of HI was detected, in contrast to HI dimer synthesis noted under the same conditions with native chromatin (lane 3). The data with reconstituted chromatin was supportive of the Fact that a large percentage of H1 was bound in a native fashion. Lane 1 shows that reconstituted nucleosomes carried out significant levels of HI dimer synthesis at 100 pM NAD. The reconsti- tuted nucleosomal sample was not quite as efficient as the native chromatin in complete synthesis of H1 dimer, i.e. more H1- poly(ADP-ribosy1)ated intermediates were noted in the reconstituted sample. This may be due to the various steps involved in the sample preparation and to some non-specific binding of H1 during reconstitution. In an earlier study, it was shown that the above intermediates in the reaction migrate close, to unmodified H1 on sodium dodecyl sulfate gel electrophoresis, while the dimer migrates at a position indicating a molecular weight approximately twice that of HI [S]. We have routinely encountered 32P-labeled material not entering gels under these conditions [S]. This material might represent free poly(ADP-Rib) or aggregated histone HI.

Assuming that the synthesis of poly(ADP-Rib) - H1 dimer is a highly specific reaction in isolated nucleosomes, these data suggest that the conditions used to reconstitute both polymerase and H1 placed most of these molecules in a proper alignment in the reconstituted samples. Accordingly, the importance of HI histone and its modification by poly- (ADP-Rib) polymerase in chromatin condensation could be studied further.

NAD-Promoted Nucleosome Aggregation As studied in detail previously [6], NAD causes only those

oligonucleosomes undergoing poly(ADP-ribosy1)ation to ag- gregate with respect to both their sedimentation properties and electrophoretic mobilities. This is shown in Fig. 2 by the coincidence of [32P]poly(ADP-Rib) incorporation with nu- cleosome repeat sizes when incubation was performed with 0.1 pM NAD (lane 1) and, alternatively, the aggregation of these labeled nucleosomes upon subsequent incubation with 100 pM NAD (lane 2) . The more rapidly migrating radio- active bands in lane 1 (i.e. low NAD concentration) represent

212

Fig.

A 1 2 3

Restoration qf poly(AD

B Stain Autorodiogrom

1 2 3 1 2 3

Rib) - H I complex synthesis in reconstituted oligonucleosornes. Nucleosomes were deplete of I by the trypsin method and reconstituted with H I and semi-purified polymerase as previously described [ l ] and in Materials and Methods. H1 dimer synthesis was studied in ritro with nucleosomes by labeling with r2P]NAD. The incubation mixture (700 pl final volume) contained 50 mM Tris/HCI. pH 8, 10 mM MgC12. 1 mM dithiothreitol, [3ZP]NAD and oligonucleosomes (1 A260 unit). Samples were 'pulsed' with [32P]NAD (0.1 pM, 7 ptci) for 30 s and 'chased' by the addition of non-radioactive NAD to 100 pM for 5 min [2]. Selective H1 extraction, clectrophoresis on acetic acid/urea/polyacrylamide gels and autoradiography were performed as described [2,5] in Materials and Methods. (A) Untreated oligonucleosomes (non-stripped): lane 1, pulse; lanes 2 and 3, chased for 1 min and 5 min. respectively. (B) Lane 1 , reconstituted oligonucleosomes, chased for 15 min; lane 2, H1-depleted oligonucleosomes chased for 15 min; lane 3. untreated oligonucleosomes chased for 15 min

Fig. 2 . Recon.ytilution of histone H1 restores poly(ADP-Rib)-induced nucleosome aggregation. Oligonucleosomes were treated with 0.1 pM r*P]NAD for 30 s (pulse) and subsequently chased with 500 pM non- radioactive NAD for 5 min. The chromatin samples were analyzed by native 2.5 polyacrylamide electrophoresis as described previously [2]. Lanes 1 and 2, autoradiogram of untreated oligonucleosomes, pulse and chase-ADP-ribosylation, respectively; lanes 3 and 4, autoradiogram of H 1 and poly (ADP-Rib) polymerase reconstituted oligonucleosomes; pulse and chase-ADP-ribosylation, respectively

mononucleosomes and the slower band represents dinucleo- somes. This has been well characterized in an earlier study (see Fig. 1A of [ 9 ] ) ; confirmation by base pair size analysis on a similarly prepared preparation has also been provided previously [9].

In Fig. 2, reconstituted oligonucleosomes were incubated with 0.1 pM ["PINAD for 30 s (lane 3) and subsequently chased with 500 pM nonradioactive NAD (lane 4). During the pulse, most of the radioactivity was incorporated into small nucleosome chain sizes. The resolution of this material was not as clear as a similar pulse performed with native chromatin (lane 1) where distinct labeled nucleosomal repeat sizes were noted. This was probably due to the variety of treatments required to deplete and subsequently reconstitute this sample; moreover it may also be due to some false binding of polymerase and HI during reconstitution. Despite these complications, the data (lane 4) show that the ["PI- poly(ADP-Rib) incorporated into nucleosomes was chased, upon incubation with 500 pM non-radioactive NAD, into chromatin with greatly reduced electrophoretic mobility, in a manner quite similar to that observed with native chromatin (lane 2). It should be noted that 500 pM NAD was utilized in this assay to ensure aggregation potential. However, similar results were obtained with 100 pM NAD. The poly- merase utilized in these experiments represents a preparation from a mid-step in the enzyme purification procedurc as reported by Jump and Smulson [I] ; this preparation contains other non-histone proteins besides thc polymerase itself. The latter did not appear to participate in the aggregation reaction since the preparation alone (i.e. without HI ) did not cause complex formation. Additionally, some minimal polymerase activity and contaminating non-histone proteins are still

21 3

retained on stripped nucleosomes, yet no aggregation of nucleosomes was observed in the absence of the reconstitution of histone HI [2]. Based upon the above data, we tentatively conclude that the poly(ADP-ribosy1)ation of this histone may be involved in the process.

DISCUSSlON

We have been interested in how the modification of histone HI may be an important component in the condensation of oligonucleosomes in vitro promoted by poly(ADP-ribosyl)a- tion. This chromatin condensation was previously shown to occur in an NAD concentration-dependent manner, presum- ably via long-chain poly(ADP-Rib) crosslinking of chromatin nuclear proteins [2,6]. In the earlier studies, the poly(ADP- ribosyl)ation of histone HI appeared to be required for this reaction since selective removal of this histone from oligo- nucleosomes abolished the effect. However, attempts to restore the aggregation by the reconstitution of exogenous H1 were not successful. It was therefore obvious that better precision was required during reconstitution of HI with chromatin so as to maintain its natural profile for poly(ADP-ribosy1)ation.

The results of the above work shown that histone HI and poly(ADP-Rib) polymerase can be reassociated to H1 -de- pleted oligonucleosomes near their native sites. This con- clusion is supported by the following observations. (a) The specific activity of the polymerase in reconstituted oligo- nucleosomes was approximately 60 of that of native chro- matin (Table 1). (b) The capacity of poly(ADP-Rib)-HI- dimer synthesis (Fig. 2B) was restored upon reconstitution. (c) The ability of poly(ADP-ribosy1)ation to cause poly- nucleosome aggregation [2], lost upon H1 depletion, was partially regained upon reconstitution with HI and poly- merase (Fig. 2). (d) The reconstituted oligonucleosomes be- haved electrophoretically like native chromatin. (e) Antibodies directed against either H1 or polymerase [I71 were bound with some specificity and quantitativeness to native or re- constituted nucleosomes (data not shown).

There is considerable evidence that histone HI plays an important role in stabilizing the higher-order packing of chromatin through interactions with nucleosomes and also interaction with other HI molecules. It is of interest that Jorcano and coworkers [I81 have recently shown that isolated oligonucleosomes are able to interact with each other through the very lysine-rich histones (HI and H5) and form hetero- genous globular particles with a mean diameter of about 30 nm. Electron micrographic studies have earlier shown that, upon addition to H1-depleted chromatin, HI causes con- densation of internucleosomal linker DNA and the formation of complex structures (20-30 nm in diameter) [19,20].

The post-synthetic modifications of this histone may therefore be an important means for altering the higher

levels of chromatin organization. The poly(ADP-ribosy1)ation reactions of H1 provide an attractive means by which this histone might interact covalently with both nucleosomes and; or other histone H1 molecules along chromatin fibers. As suggested by Stone et al. [3] the H1 dimer may be a transient modification which causes localized regions of chromatin to condense or open depending upon the biological activities occurring in that domain. With the approach described in the current work, it should be possible ultimately to define the biological consequences of poly(ADP-ribosy1)ation of HI within chromatin in a very systematic fashion. Our current direction is to reconstitute HI with polynucleosomes con- taining widely differing lengths of internucleosomal chromatin to assess the structural restraints of this reaction.

We wish to thank Dr Michael Bustin for providing histone H1 anti- body. This work was supported by National Institutes of Health grants. CA 13195 and CA 25344.

REFERENCES 1 . Jump, D. B. & Smulson, M . (1980) Biochemisrr,: 19. 1024- 1030. 2. Butt, T. R . & Smulson, M. (1980) Biochemistrj, IY, 5235-5242. 3. Stone, P. R., Lorimer 111, W. S. & Kidwell, W. R. (1977) Eur. J .

4. Adamietz, P., Klapproth, K. & Hilz, H. (1979) Biochem. Biophw

5. Nolan, N . L., Butt, T. R., Wong, M., Lambrianidou, A. & Smulson.

6. Butt, T. R., DeCoste, B., Jump, D. B., Nolan, N. & Smulson, M.

7. Worcel, A. (1978) Cold Spring Harbor Symp. Quunt. Bid. 42.

8 . Sporn, M. B., Berkowitz, D. M., Glinki, R. P., Ash, A. B. & Steven,

9. Butt, T. R., Brothers, J. F., Giri, C. P. & Smulson, M. (1978) Nucleic

10. Butt, T. R., Jump, D. B. & Smulson, M. (1979) Proc. Nut1 Acud.

1 1 . Panyim, S . & Chalkley, R. (1969) Arch. Biochem. Biophys. 130,

12. Nelson, P. P.. Albright, S. C., Wiseman, J . M. & Garrard, W. T.

13. Jump, D. B., Butt, T. R. & Smulson, M. (1980) Biochemisfry. IY,

14. Weintraub, H. & Van Lente, F. (1974) Proc. Nut1 Acud. Sci. USA.

15. Goldblatt, D. & Bustin, M. (1975) Biochemi.~frj: 14, 1689- 1695. 16. Towbin, H., Shaehelin, T. & Gordon, J . (1979) Proc. Nut1 Acud.

17. Malik, N.. Bustin, M . & Smulson, M. (1982) Nucleic Acid Res. 10.

18. Jorcano, J. L., Meyer. G. , Day. L. A. & Renz, M. (1980) Pmc. Nut1

19. Finch, J . T. & Klug, A. (1976) Proc. Nafl Acurl. Sci. USA, 73.

20. NOH. M. & Kornberg, R. D. (1977) J. Mol. B i d . 109, 393-404.

Biochem. 81,Y - 18.

Res. Commun. 91, 1232-1238.

M . (1980) Eur. J . Biochem. 113, 15-25.

( 1 980) Biochemistry, 19, 5243 - 5249.

313-324.

C. L. (1969) Science (Wu.vh. DC) 164, 1408-1410.

Acids Rex 5.2775 -2778.

Sci. USA, 76,1628-1632.

337 - 346.

(1979) J. Biol. Chem. 254, 1 1 751 - 11 760.

1031 -1037.

71.4249-4253.

Sci. USA, 76,4350-4354.

2939 -2950.

Acud. Sci. USA. 77, 6443-6447.

1897-1901.

M. Wong, N. Malik, and M. Smulson, Department of Biochemistry. School of Medicine and Dentistry, Georgetown University, Washington, District of Columbia. USA 20007