Individual histone messenger RNAs: Identification by template activity

Cell, Vol . 4, 2 3 9 -248, March 1975, Copyright © 1975 by MIT

Individual Histone Messenger RNAs : Identification by Template Activity

Shoshana Levy*, Periann Wood, Michael Grunstein, and Laurence Kedes Department of Medicine Stanford University School of Medicine and the Veterans Administration Hospital Palo Alto, California 94305

Summary Newly synthesized polysomal messenger RNAs from cleavage stage embryos of the sea urchin Arbacia punctulata and Lytechinus pictus that contain putative histone mRNAs have been fractionated on 6% polyacrylamide slab gels . At least 8 RNA species with unique electrophoretic mobilities have been recognized . The complex of RNAs has been eluted from the gels In three groups, A, B, and C, in Increasing order of mobility . The template activity of the three fractions and the unfractionated starting material was examined In the mouse Krebs 11 ascites tumor cell-free protein synthesizing system . The unfractlonated messenger complex programs the synthesis of proteins that coelectrophorese exclusively with sea urchin histories in both sodium dodecyl sulfate and acid urea gel systems . The products of In vitro protein synthesis stimulated by the Individual polyacrylamide gel RNA fractions were similarly examined . Each stimulated protein synthesis and was enriched for specific hlstone templates . We conclude that RNA fraction A Is template for histone 11, C is template for histone f2al, and B serves as template for f2b, f2a2, and f3 histones . A minor degree of contamination of the A and B RNA fractions was obvious from the production of other histories by each template . The co-electrophoresis of specific template activity with specific radiolabeled RNAs supports the concept that most or all of the labeled RNAs are Indeed themselves the histone mRNAs . Introduction The cleaving sea urchin embryo provides an excellent model for the study of the molecular biology of eucaryotic messenger RNA (mRNA) . Polysomeassociated RNA is easily isolated and radiolabeled, represents a major fraction of labeled cellular RNA, and is free, in early cleavage stages, of detectable amounts of newly synthesized ribosomal RNA species (Kedes and Gross, 1969a) . Study of the synthesis of chromosome-associated proteins in cleaving embryos led to the observation that a heterogeneous 9S RNA associated with light polyribosomes *Present address: Biochemistry Department, Weizmann Institute of Science, Rehovat, Israel .

engaged in histone synthesis (Kedes et al ., 1969 ; Nemer and Lindsay, 1969 ; Moav and Nemer, 1971) may be histone messenger RNAs (Kedes and Gross, 1969b) . When the 9S mRNAs are subjected to analytical electrophoretic techniques, several RNA species becomes evident (Kedes and Gross, 1969a) . These 9S RNA subfractions are more homogeneous than total 9S RNA as analyzed by RNADNA hybridization (Weinberg et al ., 1972 ; Grunstein, Schedl, and Kedes, 1973a) or by slab gel electrophoresis and oligonucleotide fingerprint examination (Grunstein et al ., 1973b ; Grunstein and Schedl, manuscript in preparation) . A more rigorous identification of 9S RNAs as histone messengers is important since these RNAs are different in several respects from any other eucaryotic mRNAs yet studied . These presumptive mRNAs for histones are represented some 4001000 times (species variation) in the urchin genome (Kedes and Birnstiel, 1971 ; Weinberg et al ., 1972 ; Grunstein et al ., 1973a), and are probably complementary to the DNA of a single locus or a restricted set of adjacent loci of Drosophila melanogaster (Pardue et al ., 1972) . In addition they contain no 3' OH polyadenylylate (Adesnik et al ., 1972 ; Grunstein et al ., 1973b), and their rate of appearance in the cytoplasm after transcription is quite rapid in tissue culture cells (Schochetman and Perry, 1972 ; Adesnik and Darnell, 1972) and is less than 2 min in sea urchins (Grunstein, unpublished results) . This characterization is based upon the assumption that the radioactively labeled 9S RNAs are the templates responsible for histone synthesis, but the identical nature of the labeled molecules and template activity has not yet been rigorously established . First steps toward chemical identification of 9S RNA as histone mRNA were recently taken when the unfractionated RNA was shown to be capable of directing the synthesis of the several histones in a cell free system derived from mouse ascites cells (Gross et al ., 1973) . Parallel observations regarding the restricted template specificity of a similar RNA preparation from HeLa cells have also been reported (Jacobs-Lorena, Baglioni, and Borun, 1972 ; Gallwitz and Breindel, 1972) . We have recently been able to fractionate labeled total 9S RNA on polyacrylamide slab gels into a series of reproducible size classes (Grunstein et al ., 1973a) . The base composition of these RNAs, the absence of polyadenylic acid, and preliminary sequence analysis (Grunstein et al ., 1973b) suggested that each RNA was the template for an individual histone, but direct evidence was lacking . We report here that the individual fractions of 9S RNA also stimulate the synthesis of histones in a cell free system . We therefore have been able to identify, among the several



Cell 240

species of RNAs in the 9S group, the individual templates for histones f1, f2al, and a cluster of RNAs containing the templates for the histones f2b, f3, f2a2 . [Histone nomenclature continues to be confusing . We use here the alpha-numerical designations of Phillips and Johns (1965) . We feel these are synonymous with the nomenclature of Panyim and Chalkly (1969) as f1 = 1 ; f3 = 2 ; f2b = 3 ; f2a2 = 4 ; and f2al = 5 . We recognize at least 6 separable histones in stained gels including the developmental-stage specific double fl proteins (f1 M and f1 G) in confirmation of the finding of Seale and Aaronson (1973) and Ruderman and Gross (1974) .] Additional evidence obtained by nucleotide sequence analysis demonstrates that the f2al template RNA contains the expected codons and will be presented elsewhere (Grunstein and Schedl, manuscript in preparation) . Results The proteins synthesized in the Krebs ascites cell free system by reactions containing 9S and 20S RNA (Figure 1) and 26S rRNA or endogenous RNA were compared by polyacrylamide gel electrophoresis (data not shown) . No discrete radioactive bands were found in the products of reactions containing 26S or endogenous RNA . The 9S and 20S products, on the other hand, consist of several discrete proteins with coincident electrophoretic mobilities . Since the leucine used as radioactive tracer in this experiment is an amino acid commonly found in all proteins, it would not be ex-

pected to preferentially label one protein over another. The programming of discrete polypeptides by only the 9S and 20S RNAs thus suggests that they contain a restricted number of functioning templates, whereas the heterogeneity of the products of the 26S and endogenous incubations probably reflects heterogeneity of the functional size of any templates they might contain . That 20S RNA produces proteins similar in size to those made by the smaller 9S RNA reflects the fact that 20S RNA is a hydrogen-bonded complex (Kedes and Gross, 1969a) containing nucleotide sequences and template specificities identical to those of the 9S RNAs (Kedes, Wood, Grunstein, Levy, unpublished data) . Since Gross et al . (1973) have convincingly demonstrated the presence of histone templates among 9S-sized sea urchin polysomal RNA, we compared the products of in vitro incorporation directly with histones in two widely discriminating electrophoretic systems . First, the in vitro product directed by total 9S RNA and labeled with 3H-lysine was compared on an SDS-polyacrylamide gel (SDS, sodium dodecyl sulfate) with 14C-lysine-labeled histones made in vivo (Figure 2) . Second, the lysine-labeled histones were compared with the products of in vitro synthesis with 9S RNA using 35S-methionine as amino acid label (Figure 3) by electrophoresis on acid-urea slab gels at pH 2 .7 . The products of in vitro synthesis contain proteins which comigrate with the f1 and f2al histone in both the SDS and

c') 10 5 x 2a. 4 U

10

2

5

Top

5

10

15

20

Bottom

Fraction Number Figure 1 . Analytical SDS-Sucrose Density Gradient of Labeled RNA from Polyribosomal Pellets 10,000 cpm 3H-uridine-labeled polyribosomal RNA from Lytechinus pictus (0-0) and 7500 cpm 14C-uridine marker RNA from myeloma cells (0----0) were loaded on a 5 ml 15-30% sucrose-SDS gradient . The gradients were centrifuged in a Spinco SW 50 .1 rotor for 3 .5 hr at 20°C . 0 .2 ml fractions were collected and counted. The myeloma RNA was a gift of Dr . R . Shutt.

1 10

i 20

1 30

1 40

50

Slice Number (mm) Figure 2 . Co-electrophoresis of Cell-Free Products and Histories of Arbacia punctulata on SDS Polyacrylamide Gels 0) 14C-labeled (0----0) in vitro product with 9S RNA added, (0 histones .



Identification of Individual Histone mRNAs 24 1

acid-urea systems. The less well separated f2b, f3, and f2a2 fractions also co-electrophorese with discrete products of the in vitro incorporation system . When 3 H-leucine is incorporated, the f1 region is much less radioactive than the other histones (data not shown), whereas when 3 H-lysine is used there is relatively more label incorporated into this lysinerich histone (fraction 20 of Figure 2) as might be predicted . It should also be noted that when 35 Smethionine was used as label, no in vitro product electrophoretically coincident with f2a2 is detectable . This result supports our belief that the in vitrosynthesized proteins are histones, since f2a2 histone contains no methionine residue (Yeoman et al ., 1972) . The major proteins encoded by sea urchin 9S polysomal mRNA thus are histones in

agreement with the findings of Gross et al . (1973) who could find no evidence for in vitro synthesis of any proteins except histones . Isolation of Individual Histone Template RNAs

Trace amounts of 32 P-labeled 9S RNA were added to unlabeled polysomal RNA and electrophoresed on 6% polyacrylamide slabs as outlined in Experimental Procedures. Bands of radioactivity detected by autoradiography (Figure 4) were excised from the gel as well as interband regions . The RNA fractions were eluted from the gel pieces with 80-90% recovery of 32P marker . In order to compare the stimulatory activity of individual fractions with each other and with the total 9S preparation from which they were derived, fixed proportions of the total yield of each template were added to the in vitro system . For example, starting with 3 mg of total polysomal RNA, we compared aliquots representing 10% of the 9S RNA fraction with aliquots representing 10% of the eluates from individual acrylamide gels bands . RNA from individual gel bands stimulated the reaction from 2-5 fold above endogenous incorporation using 3 H-lysine as label (compared to 30 fold for total 9S RNA) . Individual RNAs from 3 separate preparations showed no consistent rank order of their stimulatory activity . Interband material had no detectable stimulatory effect . The sum of counts in-

0 .6-

0 .5-

J

E C

°~

0 .4-

a>

U C

n

0 .3-

0N Q 0.2-

0.1 2

3

4

5

Centimeters migrated Figure 3 . Acid-Urea Gel Electrophoresis of Cell-Free Product and 14C-Labeled Histories of Arbacia punctulata

35 S-methionine was uct and 14C marker

used as label in the cell-free system . The prodhistones were electrophoresed in separate wells of a 1 .5 mm thick slab gel . Autoradiography was carried out for 6 weeks . The film was scanned (0 .05 mm slit width) in a Gilford 2400-S spectrophotometer at 550 nM . FlM and f1G mark the f1 proteins synthesized during morula and gastrula stages, respectively . The in vitro template, isolated from pregastrular embryos, seems to synthesize only f1 M in confirmation of the finding of Ruderman and Gross (1974) .

Mobility (cm) Figure 4 . Fractionation of 9S RNA on SDS-Polyacrylamide Slab Gel 100,000 cpm 32P polyribosomal RNA of Arbacia punctulata was electrophoresed and autoradiographed as described in Experimental Procedures . The X-ray film was scanned as described in the legend to Figure 3 . The original X-ray film was used as a template for excision of gel fragments corresponding to regions A, B, and C .

Cell 242

corporated above background from the individual fractions was lower than the total stimulated incorporation programmed by total 9S RNA . When larger aliquots of gel eluates were added to the in vitro cell-free system, incorporation was progressively inhibited . We believe this is consistent with an inhibition of incorporation by materials eluted from the gel, such as ammonium persulfate or acrylamide monomers, rather than a degradative effect of electrophoresis on the templates themselves . Assessment of Purity of Individual RNA Fractions Total 9S RNA seems to be template for histones and for no other major class of identifiable protein(s) . We reasoned that since 9S RNA can be separated into individual molecules with unique nucleotide sequences, each should act as template for a single histone . When we subjected fractionated Lytechinus pictus RNAs to reelectrophoresis, marked enrichment of individual moieties was detected (Figure 5), but serious cross contamination of some species was also evident . Most marked was the contamination of bands within each group (for example, contamination of band B1 with other B group bands) . Contamination of A group RNA fractions with RNA from the dominant B group was also evident . C group RNA on the other hand shows little cross contamination on reelectrophoresis . This has been independently verified by the purity of fingerprint patterns obtained from ribonuclease digests of C band RNAs (Grunstein et al ., 1973b ; Grunstein and Schedl, manuscript in preparation) and from the heterogeneity of such patterns with A and B groups RNAs . The contaminants are usually less than 10% the activity of the enriched fractions (but not always, for example, A2 band in Figure 5) and are easily detected by autoradiography . Thus at present the method allows a separation of C RNA but only a marked enrichment of most of the other labeled RNA species . Analysis of Cell-Free Products of Enriched RNA Fractions When tracer amounts of RNA are reelectrophoresed, cross-contamination is not a problem (data not shown) . But in order to isolate the near microgram quantities of individual templates needed for in vitro protein synthesis, preparative gels had to be loaded with several milligrams of cellular RNA . Thus trailing of RNA boundaries into slower migrating species is most likely due to overloading of the gel . For the most rapidly migrating RNAs this is not a problem, but RNAs with slower mobilities will tend to be contaminated by them . If C RNA is a reasonably pure template, its in vitro product thus should be almost exclusively a single protein . Because of the cross-contamination of the other indi-

vidual RNAs, however, we anticipated that the in vitro product of A or B RNA would be a mixture of proteins rather than a single species, but that the product of each template might be greatly enriched for one or a few proteins . Identification of the individual RNAs therefore would depend on a partial separation and purification of template activity from the mixture of RNAs . Accordingly, the RNAs from gel group A, B, and C were used individually as templates for protein synthesis . The in vitro products of each reaction were analyzed by electrophoresis with marker histones in both the SDS and acid-urea gel systems (Figures 6 and 7) . As expected, only when C RNA was used as template was the product predominantly a single protein (Table 1) . The product co-migrates with f2al histone in both gel systems above the background of heterogeneous endogenous incorporation . Though a small number of counts also co-migrate with the other histories in the acid urea gel (Figure 7), the preparation examined in Figure 6 seems free of proteins other than f2al . This is in agreement with the purity of C RNA as determined from nucleotide sequence analysis (Grunstein and Schedl, manuscript in preparation) . When we examined the in vitro products synthesized with A and B RNAs, the situation was less clearcut . By comparing the electrophoretic mobilities of the products of the fractionated RNAs with those encoded by total 9S RNA (Figure 2 and 3), however, several features do emerge . While fl histone is a minor product (<20%) of in vitro synthesis stimulated by 9S RNA (Figure 2 and 3 and Table 1), the fraction of product which co-electrophoreses with f1 rises to 33-54% when A RNA is separated and used as template (Figure 6 and 7 and Table 1) . Similarly, when B RNA is used as template, a mixture of proteins is synthesized which predominantly co-migrates in both gel systems with f2a2, f2b, and f3 histones (Figures 6 and 7) . Since these three histones already represent the major products of synthesis engendered by total 9S RNA, the identification of B RNA as their template must rest on the fact that the B RNA group codes for little or no f1 (especially see Figure 6) or f2al (compare Figure 6 and 7, and Table 1) . Discussion The mouse Krebs ascites tumor cells have proven to be a valuable tool in studying the template activity and specificity of a variety of eucaryotic (Mathews et al ., 1972 ; Housman, Pemberton, and Taber, 1971) and viral RNAs (Mathews and Korner, 1970 ; McDowell et al ., 1972) . We have confirmed here the finding of Gross et al . (1973) that 9S RNA from cleavage sea urchin embryos contains template ac-

Identification 243

Figure

of Individual

5. Reelectrophoresis

Histone

mRNAs

of Individual

a*P-Labeled

RNA

Fractions

on 6% Acrylamide

RNA bands eluted from a preparative gel were layered on a new gel and electrophoresed as described in Experimental Procedures. The 10 wells of the gel were loaded from left to right with unfractionated 9s RNA and fractions C3, C2, Cl, 84, 83, 82, Bl , A2, and Al, respectively. The total 9s and C region bands are overexposed. On shorter exposures C3 RNA demonstrated only one C region band.

Cell 244

tivity for histories when analyzed in the Krebs system . In addition we have separated the 9S RNAs into discrete size classes, each of which is highly enriched with templates for specific histones . Identification of the products of in vitro incorporation as histories rests on several experimental findings . First, the products co-electrophorese with histones

labeled in vivo in both neutral SDS gels and in the discriminating 4 M urea gel system at pH 2 .7 . Second, the in vitro labeling characteristics of the fl histone with lysine and leucine as substrates and the failure of the methonine-free f2a2 protein to label with 35 S-methionine as substrate (see Results) strongly supports the identification of the in vitro products . Third, using sea urchin 9S RNA as template, Gross et al . (1973) rigorously demonstrated by tryptic peptide analysis that the in vitro protein products are indeed histones . Finally, in the work reported here and the experiments of Gross et al . (1973), there is no evidence for any specific incorporation into any proteins other than histones . 10

N O x

CL

0

Slice Number (mm) Figure 6 . SDS-Polyacrylamide Gel Analysis of Cell-Free Products Stimulated by Fractionated RNAs of Arbacia punctulata 2 yCi of 3H-lysine was added to each reaction mixture containing RNA from fractions A, B, or C (Figure 4) . The incubation conditions and analysis of products are described in Experimental Procedures . Histones labeled in vivo with 1 4C-lysine were added to the incubation mixture after in vitro protein synthesis was terminated but before dialysis. The products were electrophoresed in separate gels and aligned by the 14C marker patterns . The position of the 14C marker histones is indicated in Panel A . F3+ refers to the mobility of the poorly resolved f3, f2b, and f2a2 histones . ( •- --- 0) product of A RNA ; (0---- 0) product of B RNA ; (A ---- A) product of C RNA.

Slice

Number (mm)

Figure 7 . Acid-Urea Polyacrylamide Gel Analysis of Cell-Free Products Stimulated by Fractionated RNAs Symbols and details as in Figure 9 .

Identification of Individual Histone mRNAs 245

Since only several hundred cpm of radioactive amino acids were incorporated into any one histone protein by A, B, or C RNAs ultimately isolated from large embryo preparations, our attempts at analysis of individual products by trypsin digestion or cyanogen bromide cleavage were not conclusive . Our analyses of the in vitro products of unfractionated 9S RNA have repeatedly shown that fewer counts are incorporated into the f1 histone relative to the other histories than are incorporated into histones labeled in vivo . This is independent of whether lysine or leucine are used as the labeled precursors . It is possible that at the stage of development from which 9S RNA is extracted (cleavage), f1 templates and f1 production may not be as great as at the time the in vivo marker proteins are isolated (early pluteus stage) . This possibility is consistent with the report that the fl protein changes during development and a second species appears after cleavage (Seale and Aaronson, 1973 ; Ruderman and Gross, 1974 ; and Figure 3) . Thus the content of f1 messenger may be low at cleavage relative to the other templates but increase at later stages . Alternatively, fi template may be a less efficient in vitro template than the other messengers . In this regard, the optimum ionic conditions for in vitro translation may vary from one histone messenger to another, and we have made no systematic evaluation of ideal conditions for the translation of the fractionated templates . Histories isolated from the chromatin of the two species of sea urchin used in these experiments have identical electrophoretic mobilities when compared in both SDS and acid-urea gels (Levy, unpublished data) . It seems reasonable to assume that because of the high degree of evolutionary conservation of several of the histones, few if any amino acid sequence differences exist between the histones of Arbacia punctulata and Lytechnius pictus . However, their respective mRNAs are neither identi-

cal in size as measured by electrophoretic mobility (compare Figure 4 and Figure 5) nor in nucleotide sequence (Grunstein et al ., 1973a ; Grunstein and Schedl, manuscript in preparation) . Accordingly, the molecular differences between the templates must reflect a combination of differences due to degeneracy of the amino acid code and differences due to the length (if not sequence) of nontranslated portions of the messengers . Caution thus must be exercised in attempts to isolate or identify mRNAs for homologous proteins from different species merely by physical separation on the basis of electrophoretic mobility . Some additional criterion, such as template specificity or nucleotide sequence analysis must be sought . The use of slab gel electrophoresis coupled with autoradiography of 32P tracer RNA has proven to be a powerful tool for separation of RNAs of similar molecular weight . The method has allowed us to isolate fractions greatly enriched in single template specificities . Though we are unable to measure directly the specific activity of the individual mRNAs (that is, the stimulatory activity per ug), we believe that we recover individual templates quantitatively from gels . We are able to estimate the template content of 32 P or 3H total 9S RNA preparations as 1-2 x 10 5 cpm/pg, and can follow the yields of these markers through gel fractionation . Accordingly, we recover 3-6 pg of 9S RNA from 3 ml of eggs . Approximately 10% of the total preparation (0 .3-0 .6 pg of 9S RNA) was used to stimulate the in vitro protein synthesizing system to produce the products of Figures 2 and 3, while the reaction mixtures of Figures 6 and 7 contained approximately 0 .1-0 .2 pg or less of the individual A, B, or C RNAs . The unique features ascribed to unfractionated histone messenger RNAs (absence of Poly A, rapid cytoplasmic appearance) and to histone DNA (reiteration, clustering, linkage) by our previous studies and those of others rely on the assumption that the

Table 1 . Relative Stimulation of In Vitro Histone Synthesis by 9S RNA and its Subfractionso Proportion of Protein Product (%) f1

f3 + b

f2al

Template RNA

SDS<

pH 2 .7d

SDS

pH 2 .7

SDS

pH 2 .7

Total 9S

19

20

46

46

22

22

RNA "A"

33

54

53

38

7

7

RNA "B"

8

31

70

63

18

4

RNA "C"

11

16

20

22

59

62

oThe data represent the fraction of total radioactivity above background recovered from the gel and co-migrating in the slices with 14C marker histone . The fractional amounts of histone proteins stimulated by the various templates are less than 100% because of the presence of nonspecific incorporation . Data are derived from several sets of electrophoretic analysis, some of which are presented in Figures 2, 3, 6, and 7 . bRefers to the poorly resolved f2a2, f2b, and f3 histones . Data listed are derived from SDS-acrylamide gels . dData listed are derived from acid-urea acrylamide gels .

Cell 246

radioactively labeled RNAs being examined are the histone messenger RNAs . But one must also consider the possibility that while the labeled material is an interesting set of RNAs, the template activity might be unlabeled and merely electrophoretically coincident with the label when examined by sucrose density gradient fractionation (Skoultchi and Gross, 1973) . We believe that the experiments presented in this paper make this alternative hypothesis unlikely, since the radioactively labeled RNAs and template activities can be resolved into a number of coincident fractions . In addition, template activity is found coincident with labeled RNA and not between radioactive RNA species as separated in 6% polyacrylamide gels (interband regions) . C3 RNA contains approximately 400 nucleotides (Grunstein et al ., 1973b ; Grunstein and Schedl, manuscript in preparation) and, as template for the 102 amino acid protein, histone f2al, it must also contain some untranslated sequences . From the relative electrophoretic mobilities of the templates for the 212 amino acid-long f1 protein and the f2b, f2a2, f3 histories (125, 129, and 135 amino acids long respectively), we conclude that the sequence lengths of the messengers are proportional to the lengths of the proteins they program, and that the amount of untranslated sequence in the template does not vary enough to disturb that proportionality . Experimental Procedures Development and Labeling of Embryos Gametes from specimens of Lytechinus pictus and Arbacia punctulata were obtained by injection of 0 .5 M KCI or by dissection . Eggs were washed three times in artificial sea water and fertilized with dilute fresh sperm. Only cultures that exceeded 95% fertilization were used. Cultures were grown at a density of 1 ml of embryos in 100 ml of Millipore-filtered artificial sea water containing 50 units/ml of penicillin and 50 gg/ml of streptomycin . The developing embryos were stirred gently at 18°C . Radioisotopic Labeling RNA of embryos was labeled by adding 3H-5-uridine (specific activity > 25 Ci/mmol, New England Nuclear) at a final concentration of 50 pCi/ml from the 16 cell stage until harvesting . In some cases RNA was prepared from embryos grown in phosphate-free sea water with carrier-free 32P-orthophosphate (New England Nuclear) at a concentration of 40-100 pCi/ml . Proteins were labeled from the 8 cell stage for 20 hr with 14C-lysine (specific activity 0 .3 Cl/ mmol (New England Nuclear) at a final concentration of 0 .1 pCi/mI. Cell Fractionation Polyribosomes were prepared from embryos at the early blastula stage . Embryos were harvested by low speed centrifugation, washed twice with calcium and magnesium-free sea water and once with iced medium M (0 .4 M KCI, 0 .01 M MgCI2 . 0 .05 M TrisHCI, pH 7 .7). The washed embryos were homogenized at 4°C with at least 4 vol of medium M in a Dounce homogenizer with a B pestle (Kontes Glass Co .). Homogenization was monitored by phase microscopy and averaged 5-10 strokes . The homogenate was spun at 10,000 x g for 10 min . 8 ml of the supernatant were layered over 2 ml of 50% sucrose in medium M and spun for 3

hr at 60,000 rpm in a Spinco 65 rotor . The supernatant was removed by aspiration, the polyribosomal pellet was rinsed with a few drops of medium M and processed or stored at -70°C . Extraction of RNA Polyribosomal pellets were dissolved in a small volume (1-2 ml) of SDS buffer (0 .1 M NaCl ; 1 mM EDTA, 0 .5% SDS and 10 mM Tris-HCI, pH 7 .4) at room temperature. Equal volumes of redistilled water-saturated phenol and chloroform -4% isoamyl alcohol were added simultaneously, and the RNA was extracted by vigorous agitation on a vortex mixer for 5 min in the cold . Centrifugation for 5 min at 5000 x g separated the phases . If a significant interface was present, the organic solvents were replaced and the aqueous layer reextracted . A 0.1 vol of 1 N NaCl and 2 .5 vol of ethanol at -20°C were added to the clear supernatant . The RNA was precipitated at-20°C overnight, deposited by centrifugation, washed with 70% ethanol, and either drained dry or lyophilized . Preparation of Specific RNA Size Classes Sucrose density gradient centrifugation on 15-50% sucrose-SDS gradients was performed as described by Kedes and Gross (1969a) . Figure 1 illustrates fractionation of a typical preparation . The fractions marked as 9S and 20S were pooled and used as templates for protein synthesis . RNA was also fractionated by SDS-polyacrylamide slab gel electrophoresis (Grunstein and Schedl, manuscript in preparation) . 32Plabeled polysomal RNA from the homologous species was added routinely as a marker for nonlabeled or 3 H-uridine-labeled RNA . The gels were exposed to Kodak SB/54 X-ray film, and the autoradiographic image was used to locate the fractions . RNA electrophoresed in a 3-6% polyacrylamide gel separated as illustrated in Figure 4 . The scan is of an autoradiograph of Arbacia punctulata RNA labeled with 32 P-orthophosphate from the 16 cell stage up to early blastula stages . Gel slices of regions A, B, and C were removed and eluted with 0 .15 M NaCl in the cold, precipitated in 70% ethanol overnight at -20°C, deposited by centrifugation, suspended in 0 .1 M NaCl, and reprecipitated in ethanol at least 2 more times. The final RNA precipitate was dried and resuspended in H 2O for use as template in the in vitro protein synthesis system . Nuclei Nuclei were prepared by a modification of the method of Hogan and Gross (1972) from swimming embryos at late blastula stages . The embryos were washed twice in sea water and homogenized in 5 vol of STC buffer (0 .5 M sucrose, 3 mM CaCI ; 50 mM Tris-HCI, pH 8 .0) using the tight fitting Dounce homogenizer. An equal volume of sea water was added, and the cells were centrifuged for 10 min at 700 x g . The pellet was suspended in 25 ml of 0 .25% Triton X-100 in STC and homogenized ; the homogenization was monitored by phase microscopy and averaged 5 strokes to give clean nuclei . An equal volume of sea water was added, and nuclei were spun for 10 min at 1000 x g . The nuclei were washed in STC twice more and were used for chromatin extraction directly or stored at -20°C . Preparation of Histories Chromatin was extracted from nuclei as previously described (Levy et al., 1973). Histories were prepared by suspension of the chromatin in 0 .4 N H2SO4 for 1 hr at 0°C . The acid insoluble material was removed by centrifugation for 10 min at 10,000 x g, and the supernatant liquid was dialyzed against H 2 O and lyophilized to dryness. in Vitro Protein Synthesis Mice bearing Krebs II ascites tumors were generously provided by Dr. R . Shields. Transplanted ascites fluid has been subsequently maintained in BALB/c mice . Ascites cells were collected 7 days post inoculation, washed, homogenized, and processed according to Mathews and Korner (1970), except that the preincubation medi-

Identification of Individual Histone mRNAs 247

um was 75 mM KCI, 25 mM Mg-Acetate, 4 .5 mM /3-Mercaptoethanol, 20 mM Tris-HCI (pH 7.6) . In vitro protein synthesis was performed in a total volume of 50 µl, which contained 15 µl preincubated S30 fraction and various RNA concentrations as indicated in the individual experiments in a final concentration of 75 mM KCI ; 3 .5 mM Mg-Acetate; 20 mM Tris-HCI ; 2 .4 mM $-Mercaptoethanol ; 4 mM phosphoenolpyruvic acid (Sigma) ; 0 .6 units pyruvate kinase (Sigma) 1 mM ATP (adenosine triphosphate), (Sigma) ; 0 .1 mM GTP (guanidine triphosphate), (Sigma) ; 40 µM of each of the 19 nonlabeled amino acids ; and a labeled amino acid as indicated in each of the experiments . The reaction was incubated at 37°C for 1 hr and terminated by rapid cooling . 5 µl aliquots were added to 0 .2 ml of 0 .1 M KOH and incubated at 37°C for 30 min, then cooled at 0°C and precipitated with 1 ml of 20% cold trichloracetic acid . The precipitates were collected on Whatman GF/C glass fiber filters, dried, and radioactivity was measured in a liquid scintillation spectrometer . The in vitro translation of total 9S RNA in the cell-free system derived from mouse Krebs II ascites tumor was dependent on the ionic concentration of magnesium with an optimum of 3 .5 mM which we chose for all the cell-free synthesis experiments with isolated RNA fractions . KCI concentration above 75 mM reduced incorporation into protein . Increased )3-Mercaptoethanol concentrations had a stimulatory effect only at low Mg++, but had no stimulatory effect when 3 .5 mM Mg++ was used . Amino acid incorporation increased with time for 1 hr but declined thereafter. Three different S30 preparations from the Krebs 11 ascites cells were used in the experiments reported in this article . The endogenous incorporation rates differed for each preparation . Maximal stimulation of incorporation with 9S mRNA was 29 fold greater than endogenous activity when the labeled amino acid was lysine . In contrast, the maximal stimulation of incorporation with radioisotopic leucine was only 15 fold greater than the endogenous level . The leucine incorporation ratio at 3 .5 mM Mg++ was comparable to the ratio we obtained when 9-12S polysomal RNA from rabbit reticulocytes (kindly provided by R . Rhodes) was used as template at 3 .0 mM Mg++ (the optimum for reticulocyte RNA). Sea urchin 9S RNA thus is a highly efficient template for protein synthesis in vitro . The 2 fold greater stimulation using lysine rather than leucine is most likely related to the fact that lysine-rich histone polypeptides are preferentially synthesized in the presence of 9S RNA (see Results) . Increasing amounts of 9S RNA increased incorporation into protein over a 3 fold range of concentrations from 50-150 µg/ml . The 20S RNA fraction also stimulated labeled amino acid incorporation into protein, but this stimulation was already maximal at the lowest concentration tested . The 26S RNA fraction had moderate stimulatory activity when 2 .5 µg and 5 µg were added to the 50 µl reaction mixture, but higher concentrations inhibited protein synthesis . The stimulation at lower concentrations of 26S RNA is probably due to the presence of authentic templates, since the addition of either purified sea urchin ribosomal RNA or E . Coli tRNA with the 9S RNA fraction inhibited the reaction at concentrations similar to those reported by Jacobs-Lorena and Baglioni (1972) .

Radioactive Determinations Radioactivity was measured on a Nuclear Chicago Mark II, a Nuclear Chicago Isocap 300, or a Beckman LS350 liquid scintillation spectrophotometer with appropriate corrections made for isotope crossover in different channels . Quenching was monitored by the external standard ratios method .

Analysis of the Products of the Cell Free System 10% polyacrylamide-SDS gels were prepared according to Weber and Osborn (1969). Incubation mixtures were dialyzed against 0.1 % SDS ; 0 .1 % $-Mercaptoethanol, 1 mM phosphate buffer (pH 7 .2). Gel buffer and running buffer were 0 .1% SDS, 30 mM phosphate buffer (pH 7 .2) . Gels (6 mm x 100 mm) were run at 6 mA/gel for 4 hr . 15% polyacrylamide gels in 4 M Urea at pH 2 .7 were prepared as described by Panyim and Chalkley (1969) . Gels (6 mm x 100 mm) were prerun overnight at 2 mA/gel . Samples were electrophoresed for 3 .5 hr at 2 mA/gel . The gels were frozen and cut into 1 mm sections on a Joyce-Loebel slicer . The gel slices were incubated at 60°C overnight in 3% Protosol in LiquifluorToluene (New England Nuclear).

Moav, B ., and Nemer, M . (1971) . Biochemistry 10, 881 .

Acknowledgments This work was supported by a grant from the National Institutes of Health . S . L . is a postdoctoral fellow of the Muscular Dystrophy Association of America, Inc . M . G . is a postdoctoral fellow of the Leukemia Society of America, Inc . L. K . is a scholar of the Leukemia Society of America, Inc ., and Investigator, Howard Hughes Medical Institute Reprint requests should be sent to Dr. Kedes. Received September 9, 1974 ; revised December 23, 1974 References Adesnik, M ., and Darnell, J . (1972) . J . Mol . Biol . 67, 397 . Adesnik, M ., Salditt, M ., Thomas, W., and Darnell, J . (1972). J . Mol . Biol . 71, 21 . Gallwitz, D ., and Breindl, M . (1972) . Biochem . Biop. Res . Comm . 47, 1106 . Gross, K ., Ruderman, J ., Jacobs-Lorena, M ., Baglioni, C ., and Gross, P. R . (1973) . Nature New Biol . 241, 272 . Grunstein, M ., Schedl, P ., and Kedes, L . H . (1973a) . In Molecular Cytogenetics (Gatlinburg, Tennessee : Oak Ridge National Laboratory) . Grundstein, M ., Levy, S ., Schedl, P ., and Kedes, L . (1973b) . Cold Spring Harbor Symp. Quant . Biol . 38, 717 . Hogan, B ., and Gross, P . R . (1972) . Exp. Cell . Res . 72, 101 . Housman, D ., Pemberton, R ., and Taber, R . (1971). Proc . Nat . Acad . Sci . USA 68, 2716 . Jacobs-Lorena, M ., and Baglioni, C . (1972) . Biochemistry 11, 4970 . Jacobs-Lorena, M ., Baglioni, C ., and Borun, T . W. (1972). Proc . Nat . Acad . Sci . USA 69, 2095 . Kedes, L . H ., and Birnstiel, M . (1971). Nature New Biol . 230, 165 . Kedes, L. H ., and Gross, P . R . (1969a) . J . Mol . Biol . 42, 559 . Kedes, L. H ., and Gross, P . R . (1969b) . Nature 223, 1335 . Kedes, L . H ., Gross, P. R ., Cognetti, G ., and Hunter, A . L. (1969) . J . Mol . Biol . 45, 337 . Levy, R ., Levy, S ., Rosenberg, S . A ., and Simpson, R . T. (1973) . Biochemistry 12, 224 . Loaning, U . E . (1969) . Biochem . J . 113, 131 . Mathews, M . B ., and Korner, A. (1970). Eur . J . Biochem . 17, 328 . Mathews, M . B ., Osborn, M ., Berns, A . J . M ., and Bloemendal, H . (1972) . Nature New Biol . 236, 5 . McDowell, M . J ., Joklik, W . K., Villa-Komaroff, L., and Lodish, H . (1972) . Proc . Nat . Acad . Sci . USA 69, 2649 . Nemer, M ., and Lindsay, D . T. (1969) . Biochem . Biophys . Res . Comm . 35, 156 . Panyim, S ., and Chalkley, R . (1969). Arch . Biochem . Biophys . 130, 337 . Pardue, M . L., Weinberg, E ., Kedes, L . H ., and Birnstiel, M . (1972) . J . Cell Biol . 55, 199a . Phillips, D . M . P ., and Johns, E . W. (1965). Biochem . J . 94, 127 . Reid, M . S ., and Bieleski, R. L. (1968) . Anal . Biochem . 22, 374 . Ruderman, J ., and Gross, P . R . (1974) . Devl . Biol . 36, 286 . Schochetman, G., and Perry, R . P . (1972) . J . Mol . Biol . 63, 591 .

Cell 248

Skoultchi, A ., and Gross, P . R . (1973) . Proc . Nat . Acad . Sci . USA 70, 2840 . Seale, R ., and Aronson, A . (1973) . J. Mol . Biol . 75, 647 . Weber, K ., and Osborn, M . (1969). J . Biol . Chem . 244, 4406. Weinberg, E . S ., Birnstiel, M . L., Purdom, I . F ., and Williamson, R. (1972) . Nature 240, 225 . Yeoman, L. C ., Olson, M. 0 ., Sugano, N ., Jordon, J . J ., Taylor, C. W., Starbuck, W . C ., and Busch, H . (1972) . J . Biol . Chem . 247, 6018 .