C .J. fix \\x ‘3 __§‘3 ABSTRACT THE SYNTHESIS AND METHYLATION OF MESSENGER RNA IN NOVIKOFF HEPATOMA CELLS BY Ronald Charles Desrosiers The poly(A) tract found in eucaryotic mRNA has been utilized to investigate the possibility of methylation in mRNA obtained from Novikoff hepatoma cells. Methyl labeling of RNA was achieved with L-[methyl-3H1-methionine under conditions that suppress radioactive incorporation into the purine ring. Poly(A)+RNA was obtained from polysomal RNA by chromatography on oligo(dT)-cellulose. Sucrose density gradient centrifugation of the poly(A)+ mRNA revealed a pattern expected for mRNA. The poly(A)+ mRNA, obtained through standard (dT)-cellulose chroma- tography procedures, was used for methyl-labeled nucleo- side composition analyses following complete enzymatic degradation to nucleosides. Using DEAE-cellulose (borate) chromatography, which separates 2'-gémethy1nucleosides from normal and base-methylated nucleosides. approximately 50% of the radioactivity was recovered in the 2'-Qfmethyl- nucleoside fraction and 50% in the base-methylnucleoside Ronald Charles Desrosiers fraction. High speed liquid chromatography (Aminex A-S) of the 2'-Qfmethylnucleoside fraction produced four peaks coincident with the four 2'-Qfmethylnucleoside standards. Analysis of the base-methylnucleoside fraction revealed a unique pattern. While rRNA and tRNA possessed complex base methylnucleoside patterns, the distribution in mRNA was quite simple, consisting predominantly of N6 -methyl- adenosine. These results demonstrate a unique distribution of methylated nucleosides in mRNA. Standard (dT)-cellulose chromatography procedures were found to be unsatisfactory for the isolation of pure methyl-labeled mRNA. Heat denaturation of the RNA before application to (dT)-cellulose was required to release contaminants (mostly 18$ rRNA) that persisted even after repeated binding to (dT)-cellulose at room temperature. KOH digestion of methyl-labeled poly(A)+mRNA purified by (dT)-cellulose chromatography produced mononucleotide and multiple peaks of a large oligonucleotide (-6 to -8 charge) when separated on the basis of charge by Pellio- nex-WAX high Speed liquid chromatography in 7 M urea. Analysis of the purified poly(A)+mRNA by enzyme digestion, acid hydrolysis and a variety of chromato- graphic techniques has shown that the mononucleotide (53%) is due entirely to N6-methyladenosine. The large oligonucleotides (47%) were found to contain 7-methyl— guanosine and the 2'-Qfmethyl derivatives of all four nucleosides. No radioactivity was found associated with Ronald Charles Desrosiers the poly(A) segment. Periodate oxidation of the mRNA followed by B-elimination released only labeled 7-methyl- guanosine consistent with a blocked 5'-terminus containing an unusual 5'-5' bond. Alkaline phOSphatase treatment of intact mRNA had no effect on the migration of the KOH produced oligonucleotides on Pellionex-WAX. When RNA from which 7-methylguanosine was removed by B-elimination was used for the phosphatase treatment, distinct dinucleo- tides (Nmpr) and trinucleotides (Nmmepr) occurred after KOH hydrolysis and Pellionex-WAX chromatography. Thus Novikoff hepatoma poly(A)+mRNA molecules can contain either one or two 2'-Qfmethylnucleotides linked by a 5'-5' bond to a terminal 7-methylguanosine and the 2'-Qfmethyl- ation can occur with any of the four nucleotides. The 5' terminus may be represented by m7G5'ppp5' (Nmp)l or 2Np, a general structure proposed earlier as a possible 5'-terminus for all eucaryotic mRNA molecules (Rottman, F., Shatkin, A. & Perry, R. [1974], Cell 3, 197). The composition analyses indicate that there are 3.0 N6-methyl- adenosine residues, 1.0 7-methylguanosine residue and 1.7 2'-Qfmethylnucleoside residues per average mRNA molecule. Several characteristics of the transcriptional process in the Novikoff hepatoma (NlSl) cells have also been examined. The synthesis of ribosomal precursor RNA _ is very sensitive to cordycepin (3'-dA). The synthesis of hnRNA, however, is resistant to inhibition by concen- trations of 3'-dA that completely block the synthesis of Ronald Charles Desrosiers 45$ ribosomal RNA precursor. I have examined the RNA polymerases present in these cultured cells with regard to their sensitivity to cordycepin 5'-triphosphate (3'- dATP) in an effort to explain the differential inhibition of RNA synthesis observed in viva. RNA polymerases I and II were characterized on the basis of their chroma- tographic behavior on DEAE—Sephadex, as well as the response of their enzymatic activities to ionic strength, the divalent metal ions Mn2+ and Mgz+, and the toxin a-amanitin. For both enzymes the inhibition of in_yi2£g RNA synthesis by 3'-dATP was competitive for ATP. The Km values for ATP and the Ki values for 3'-dATP for the two enzymes were quite similar. RNA polymerase II, the enzyme presumed responsible for hnRNA synthesis, was actually slightly more sensitive to 3'-dATP than RNA polymerase I, the enzyme presumed responsible for ribosomal precursor RNA synthesis. Similar data were obtained when the RNA polymerases were assayed in isolated nuclei. These results indicate that the differential inhibition of RNA synthesis caused by 3'-dA ig_vigg cannot be simply explained by differential sensitivity of RNA polymerases I and II to 3'*dATP. THE SYNTHESIS AND METHYLATION OF MESSENGER RNA IN NOVIKOFF HEPATOMA CELLS BY Ronald Charles Desrosiers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1975 ACKN OWLEDGMEN T S I would like to thank all those friends and associates who made my stay at Michigan State University an enjoyable and satisfying experience. The help and understanding of my guidance com- mittee, Drs. Steve Aust, John Boezi, James Fairley and Jay Goodman and Alan Morris have been greatly appreciated. Special thanks go to the collaborators who shared the ups and downs of this work with me. These include Drs. Howard C. Towle and John Boezi who participated in the work described in part III and Ms. Karen Friderici who provided expert assistance and insight for the work in parts I and II. Dr. Fritz Rottman's encouragement, insight and assistance in all aspects of my graduate career provided an invaluable contribution to my work. PART I reprinted with permission from the Pro- ceedings of the National Academy of Science USA ll, 3971-3975 (1974). PART II reprinted with permission from Biochemistry l1, 4367-4374 (1975): Copyright by the American Chemical Society. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . vi LIST OF FIGURES. . . . . . . . . . . . . vii LIST OF ABBREVIATIONS. . . . . . . . . . . ix LITERATURE SURVEY . . . . . . . . . . . . l The Synthesis of Transfer RNA . . . . . . . 3 The Synthesis of Ribosomal RNA. . . . . . . 9 The Synthesis of Messenger RNA. . . . . . . 21 REFERENCES . . . . . . . . . . . . . . 41 PART I IDENTIFICATION OF METHYLATED NUCLEOSIDES IN MESSENGER RNA FROM NOVIKOFF HEPATOMA CELLS Summary . . . . . . . . . . . . . . 52 Introduction. . . . . . . . . . . . . 53 Methods . . . . . . . . . . . . . . 55 Cell Culture and Labeling . . . . . . . 55 Cell Fractionation and RNA Preparation . . . 55 Oligo(dT) -Cellulose Chromatography .. . . . 56 Poly(A)(- )RNA Fractionation. . . 57 Enzymatic Degradation of RNA and Resolution of Nucleosides . . . . . . . . . . 57 Results 0 O O O O O O O O O O O O O 58 Preparation of RNA Fractions . . . . . . 58 Separation of Base-Methylnucleosides from 2'-0-Methylnucleosides. . . . . . . . 62 2'-0-Methylnucleoside Distribution . . . . 64 Base-methylnucleoside Distribution . . . . 68 Discussion . . . . . . . . . . . . . 72 iii Page ACKNOWLEDGMENTS . . . . . . . . . . . . . 75 REFERENCES. 0 C O C O O I O O O O O O O 75 PART II CHARACTERIZATION OF NOVIKOFF HEPATOMA mRNA METHYLATION AND HETEROGENEITY IN THE METHYLATED 5'-TERMINUS Summary. . . . . . . . . . . . . . . 78 Introduction . . . . . . . . . . . . 79 Materials and Methods . . . . . . . . . . 83 Materials . . . . . . . . . 83 Cell Culture and Preparation of Methyl- Labeled RNA. . . . . . . . . . . 83 Oligo(dT)-Cellulose Chromatography. . . . . 84 KOH Hydrolysis and Pellionex-WAX Chroma- tography. . . . . . . . . 85 Enzyme and Acid Hydrolysis and Analysis of the Products . . . . . . . 86 Isolation of the Poly(A) Segment . . . . . 87 Periodate Oxidation and B-elimination. . . . 88 Results. . . . . . . . . . . . . . . 89 Discussion. . . . . . . . . . . . . . 113 REFERENCES. . . . . . . . . . . . . . . 119 PART III THE SENSITIVITY OF RNA POLYMERASES I AND II FROM NOVIKOFF HEPATOMA (NlSl) CELLS TO 3'-DEOXYADENOSINE S'-TRIPHOSPHATE Summary. . . . . . . . . . . . . . . 121 Introduction . . . . . . . . . . . . . 122 Experimental Procedures . . . . . . . . . 125 Synthesis of 3'-dATP . . . . . . 125 Effect of 3'-dA on in vivo hnRNA and rRNA Precursor Synthesis . . . . . . 126 Solubilization and Separation of RNA Poly- merase Activities. . . . . . . . . . 127 Assay of Solubilized RNA Polymerase Activity . 130 Assay of RNA Polymerase Activity Using Isolated Nuclei . . . . . . . . . . 131 Materials . . . . . . . . . . . . . 132 iv Page Results . . . . . . . . . . . . . . 132 Effect of 3' -dA on rRNA Precursor and hnRNA Synthesis. . . . . . . . . 132 RNA Polymerase Activities in Novikoff Hepatoma (N181) Cells. . . . . . 135 Inhibition of RNA Polymerases I and II by 3'-dATP O O O O O O O O O O O O 144 Discussion . . . . . . . . . . . . . 153 “FEWCES I O O O O O O O O O O O O O 160 Table LIST OF TABLES PART I Recovery of Labeled Methylnucleosides from DEAR-cellulose . . . . . . . . . Radioactive 2'-97Methylnuc1eoside Compo- Sition O I O O O C O I O O 0 PART II Summary of Methylnucleoside Composition Analyses . . . . . . . . . . . Isolation of Poly(A) Segment from Methyl- labeled Poly(A)+mRNA . . . . . . . Distribution of Methylated Species in an Average Poly(A)+mRNA Molecule . . . . PART III Summary of Kinetic Constants . . . . . vi Page 63 67 100 107 115 152 Figure LIST OF FIGURES PART I Sucrose density gradient sedimentation of methyl-labeled polysomal RNA . . . . . . High speed liquid chromatography of 2'—Qf methylnucleoside fraction . . . . . . . High speed liquid chromatography of base- methylnucleoside fraction . . . . . . . PART II The effect of heat denaturation on the elution of poly(A)+RNA from (dT)- cellulose . . . . . . . . . . . . DMSO-sucrose gradient centrifugation of poly(A)+ and poly(A)-RNA obtained from heat denaturation and (dT)-ce11ulose of poly(A)+RNA . . . . . . . . . . . Pellionex-WAX high speed liquid chromatography of KOH hydrolyzates . . . . . . . . . Aminex A-S high speed liquid chromatography of poly(A)+mRNA acid hydrolyzate. . . . . Aminex A-S high Speed liquid chromatography of the 2'—Qfmethylnucleoside fraction of the large oligonucleotides. . . . . . . Aminex A-S high speed liquid chromatography of the material removed by B-elimination . . Pellionex-WAX chromatography of KOH hydroly- zates of poly(A)+mRNA following B-elimi- nation and/or alkaline phosphatase treat- ment 0 O O O O O O O O O O O O 0 vii Page 60 66 70 91 94 97 102 106 110 112 Figure Page PART III 1. DMSO-sucrose gradient centrifugation of rRNA precursor and hnRNA . . . . . . . . . 134 2. DEAE-sephadex chromatography of Novikoff hepatoma nuclear RNA polymerases. . . . . 137 3. The effect of varying the concentration of (NH )2804 on in_vitro RNA synthesis by the NOVIkoff hepatoma RNA polymerases I and II . 140 4. The effect of varying the concentration of the divalent metal ion on in vitro RNA synthesis by the Novikoff hepatoma RNA polymerases I and II . . . . . . . . 142 5. The effect of 3'-deoxyadenosine 5'-triphos- phate on in_vitro RNA synthesis by the Novikoff hepatoma RNA polymerases I and II . 146 6. The effect of varying the concentration of ATP in the absence and presence of 3'- deoxyadenosine 5'-triphosphate on ig_vitro RNA synthesis by Novikoff hepatoma RNA polymerase I . . . . . . . . . . . 149 7. The effect of varying the concentration of ATP in the absence and presence of 3'- deoxyadenosine 5'-triphosphate on in vitro RNA synthesis by Novikoff hepatoma RNA polymerase II . . . . . . . . . . . 151 8. The effect of 3'-deoxyadenosine 5'-triphos- phate on RNA synthesis by Novikoff hepatoma nuclei . . . . . . . . . . . . . 155 viii 3'-dA 3'-dATP C Cm DEAE DMSO DNAase DTT EDTA G hnRNA mRNA oligo(dT)-cellulose poly(A) + RNA poly(A) - RNA LIST OF ABBREVIATIONS adenosine 2'-Qfmethyladenosine 3'-deoxyadenosine 3'-deoxyadenosine 5'-triphosphate cytidine 2'-Qfmethylcytidine diethylaminoethyl dimethylsulfoxide deoxyribonuclease dithiothreitol ethylenediaminetetraacetic acid guanosine 2'-Qfmethylguanosine heterogeneous nuclear RNA messenger RNA cellulose to which Oligo(dT) is covalently affixed RNA that contains a poly(A) segment RNA that does not contain a poly(A) segment ix rRNA RNAase SDS tRNA Tris Um wm ribosomal RNA ribonuclease lOmM Tris-HCl pH 7.4, 10 mM NaCl, 1.5 mM MgCl2 svedberg unit sodium dodecyl sulfate transfer RNA tris (hyroxymethyl) aminomethane uridine 2'-Qfmethyluridine pseudouridine 2'-Qfmethylpseudouridine LITERATURE SURVEY In bacteria, all types of RNA syntheses are carried out by the same basic core enzyme. Various factors have been described which affect the transcrip- tion by this core RNA polymerase (Burgess, 1971). Eucaryotes, on the other hand, have distinct trans- criptional machinery for the different classes of RNA synthesized in the nucleus (Roeder & Rutter, 1969; for review, Jacob, 1973). When RNA polymerase enzymes are solubilized from nuclei, three major peaks of activity are generally observed upon elution from DEAE-Sephadex. These RNA polymerases have been designated I, II and III based on their order of elution. RNA polymerase III is usually a small percentage of the total activity (5-10%) and the amount present in each of the distinct polymerases varies considerably with the cell type and physiological state (Schwartz et al., 1975). In addition to the order of elution from DEAE-Sephadex, the enzymes may be characterized by the salt concentration required for optimal activity and the response of each of the enzymes to the divalent metal ions Mg2+ and Mn2+. Probably the most useful property discovered to date has been the inhibition of enzyme activity caused by a-amanitin, a toxin obtained from the poisonous mushroom Amanita phalloides (Wieland, 1968; Blatti et al., 1970). RNA polymerase II is very sensitive to low concentrations of a-amanitin (lug/m1) while RNA polymerase III is only inhibited at very high concentrations (200 ug/ml). RNA polymerase I has been found to be resistant to inhibition at all concentrations of a-amanitin tested (Weinmann & Roeder, 1974). AS will be seen in subsequent sections, this drug has been effectively used to determine which RNA polymerase is responsible for the synthesis of each class of RNA. Another interesting feature of eucaryotic RNA synthesis is the wide range of post-transcriptional modification and maturation events that occur. In fact, it seems that most RNAS are originally synthesized as a precursor that is considerably longer than the mature functional RNA. The processing of the RNA precursor involves not only trimming the longer molecule to the desired size by ribonuclease action, but also enzymatic modification of the RNA chain. The modification may be a change in individual nucleotides, such as methylation or conversion of uridine to pseudouridine, or it may be a post-transcriptional addition of nucleotides to a pre-existing molecule, as will be described later in the addition of adenosine residues to mRNA precursor. The intense research currently involved with the pro- cessing of RNA precursors is directed at ultimately understanding the usefulness of such an energetically expensive process. In the sections that follow, I discuss the syn- thesis of each class of RNA individually (tRNA, rRNA and mRNA). In each section the synthesis of the RNA pre- cursor, the maturation steps that occur to produce the mature, functional RNA and the role that post-transcrip- tional modifications play in the function of that RNA is reviewed. The emphasis in this survey is on eucaryotes with reference to bacteria where applicable. The Synthesis of Transfer RNA Transfer RNA (cytOplasmic 4S RNA) comprises approximately 10% of the cytoplasmic RNA and in any cell type it is actually a mixture of many individual tRNA Species. Mature tRNA molecules are, on the average, about 85 nucleotides long. Unfractioned mammalian tRNA con- tains approximately 7% methylated nucleotides (Brown & Attardi, 1965; Perry & Kelley, 1974); approximately 20% represent 2'-Qfmethylation and approximately 80% base- methylation (Munns et al., 1974; Lane & Tamaoki, 1969). Among the most prominent mammalian tRNA base-methylations are l-methyladenine, l-methylguanine, 7-methylguanine, NZ-methylguanine, NzNz-dimethylguanine, 5-methylcytosine and S-methyluracil. g, 221; tRNA lacks NZ-methylguanine, NzNz-dimethylguanine, l-methyladenine and S-methylcytosine and this probably accounts for why g. 321; tRNA is an excellent substrate for mammalian tRNA methylases (Munns et al., 1974). A difference in the methyl—nucleoside composition has been observed with tRNA from different eucaryotic sources (Mittelman et al., 1969; Baguley & Staehelin, 1969). In addition to methylation, numerous other nucleo- tide modifications are associated with tRNA. These include nucleotides modified by the following derivatives: thiol groups (Lipsett, 1972), isopentene (Hall, 1970), 2—thio- methyl-6-isopentene (Gefter, 1969), pseudouridine (Johnson & 8011, 1970) and threonine (Chheda et al., 1972). That the transcription of the tRNA genes occurs via RNA polymerase III was originally suggested by the nucleoplasmic location of the enzyme (Blatti et al., 1970). The synthesis of pre-tRNA by isolated nuclei was subsequently found to be inhibited by a-amanitin, but only at high concentrations (Weinmann & Roeder, 1974). The a-amanitin inhibition curve for pre-tRNA synthesis was identical to the inhibition curve of solubilized RNA polymerase III activity. Thus the synthesis of tRNA can be assigned to RNA polymerase III. Research from many different laboratories has indicated that mature tRNA is not the product of direct transcription but arises through the processing of longer precursor tRNA molecules. When cells are exposed to radioactive nucleosides for periods as short as five minutes, the radioactive low molecular weight cytoplasmic RNA is distinctly different from the unlabeled 4S tRNA and the SS rRNA. The bulk of the labeled RNA under these rapid label conditions elutes from Sephadex G-100 (Vesco & Penman, 1968) or migrates on gel electrophoresis (Bern- hardt & Darnell, 1969) in a position between the SS RNA and the tRNA. This rapidly labeled 4.58 RNA has been observed in many mammalian cells (for review, Burdon, 1971). Kinetic studies have indicated that this 4.58 RNA is the precursor to mature tRNA (Bernhardt & Darnell, 1969; Kay & Cooper, 1969). These studies involved the use of longer labeling times and also Actinomycin D chase experiments where the label in 4.58 RNA could be followed into 48 RNA at later times. In addition, the conversion of this 4.58 precursor into 48 tRNA by cell extracts has been achieved (Mowshowitz, 1970; Smillie & Burdon, 1970). The elution characteristics of the 4.58 pre-tRNA Species can be ascribed to an increased length of approximately 20 nucleotides over mature tRNA (on the average) and not to any conformational change (Boedtker, 1967; Boedtker, 1968). Cellular fractionation after brief labeling periods has revealed that the 4.58 precursor RNA is found almost exclusively in the cytoplasmic fraction (Weinberg & Penman, 1968). This does not positively Show, however, that the size processing of the precursor takes place in the cytoplasm since the pre-tRNA could have easily leaked out of the nuclei during the isolation procedure. Similar cellular fractionation has also been used to localize the methylation event in the cytoplasm but possible leakage of pre-tRNA and methylases makes conclusions difficult (Burdon et al., 1967; Murumatsu & Fujisawa, 1968). Experiments designed to Show if methylation of pre-tRNA is necessary for conversion to mature tRNA have revealed that the precursor is deficient in modified nucleotides (Bernhardt & Darnell, 1969; Kay & Cooper, 1969; Lal & Burton, 1967; Mowshowtiz, 1970). Pulse chase experiments performed under conditions of methionine starvation indicate that size maturation can occur with- out methylation (Lal & Burdon, 1967; Bernhardt & Darnell, 1969). Further evidence for the lack of a methylation requirement for Size maturation will be discussed below. Investigation of the role of the methyl groups .in tRNA function has led to few answers. Methyl-deficient tRNA can be isolated from methionine requiring mutants of E. coli during methionine starvation (Borek & Srinivasan, 1966). Such methyl deficient tRNA can be analyzed for its ability to be aminoacylated, for bind- ing capacity to mRNA-ribosome complexes and for partici- pation in peptide bond formation. Most reports indicate that there is no significant difference in the rate or extent of amino acid acceptance between normal and methyl- deficient tRNA (Borek & Srinivasan, 1966; Biezunski et al., 1970). Shugart et a1. (1968), however, found that amino- acylation of undermethylated E, 92;; phenylalanine tRNA 12.21EEQ is greatly inhibited and can be restored to the normal level if the tRNA is methylated £2.X£E£2.by Spe- cific E. 32;; tRNA methylases. One possibility is that specific methylation helps assure that the right amino acid will be joined to the proper tRNA. Studies on the coding properties of methyl-deficient tRNA have also been inconclusive (Fleissner, 1967; Capra & Peterkofsky, 1968; Stern et al., 1970). Another approach to the problem is to study the properties of methyl-deficient tRNA isolated from mutant cells lacking a particular methylase enzyme. This has been done using a mutant unable to methylate uracil (forming thymine) in tRNA (Svensson et al., 1971; Bjork & Neidhardt, 1971). The isolated thymine-free tRNA behaved identically to normal tRNA in 12.!iEEQ amino- acylation and £2 XEEEQ polypeptide synthesis. Neverthe- less, wild type E. coli cells had a growth advantage over the mutant, suggesting some beneficial function of the methylation. Another interesting post-transcriptional event with tRNA is the synthesis and turnover of the -CCA terminus (for review, Deutscher, 1973). For many years it was unclear whether the -CCA nucleotides of the 3'-terminus were transcribed from DNA but it is known that these nucleotides are removed in the cytoplasm of cells and the -CCA rebuilt by tRNA nucleotidyltrans- ferases. Since the -CCA terminus can be formed in the cytoplasm, it was not known whether the original RNA transcripts contained the -CCA nucleotides. The finding that the terminal -CCA nucleotides in tRNA-DNA hybrids were more susceptible to pancreatic RNAse digestion sug- gested that there may not be complementary sequences on the DNA (Daniel at al., 1970). The elegant work of Altman (1971) and Altman and Smith (1971), however, on the composition and sequence of a tyrosine precursor tRNA molecule clearly shows that, at least for this molecule, the terminal -CCA sequence is transcribed from DNA . This tyrosine tRNA precursor is 129 nucleotides long, 44 residues longer than the mature tRNA species. It migrates just behind 58 RNA in polyacrylmide gels and has been completely sequenced (Altman, 1971; Altman & Smith, 1971). Forty-one of the extra nucleotides are at the 5'-terminus, with the 5'-termina1 residue being pppG. This shows that the 5'-terminus of the precursor originated from a primary transcription product. The 3'-end contains three additional nucleotides following the -CCA nucleotides that form the 3'-terminus of the mature tRNA. This shows that the -CCA in addition to the three extra nucleotides were transcribed from the DNA . The isolated tyrosine precursor tRNA was highly deficient in modified bases. A ribosome-bound ribo- nuclease was still able to specifically remove 12.21EEQ the 41 nucleotide fragment, exposing the 5'-end of the mature tRNA (Altman & Smith, 1971; Robertson et al., 1972). This indicates that the modification of nucleo- tides is not a prerequisite for the size maturation of the tRNA precursor. The Synthesis of Ribosomal RNA Ribosomes from all organisms contain two species of high molecular weight RNA. One (238 to 298) is found associated with the large ribosomal subunit and the other (168 to 188) with the small subunit. There is one species of low molecular weight RNA termed 58 RNA. In addition, eucaryotes contain an additional component termed 5.88 RNA (formerly 78 RNA) that is hydrogen bonded to the larger high molecular weight species and is only released by treatments which disrupt this interaction 10 (e.g., heat, DMSO, 7M urea). In bacteria, the high molecular weight species are usually 16S and 238. Some variation exists in the size of the high molecular weight rRNA among various eucaryotes but in mammals it is generally 188 and 288. The base composition of rRNA is usually very different from that of total DNA and in higher eucaryotes it is high in G + C (for review, Attardi & Amaldi, 1970). The high molecular weight ribosomal RNA components con- tain considerable modification. In bacteria, approxi- mately 0.7% of the nucleotides are methylated and approximately 80% of these methylations are on the bases (Fellner & Sanger, 1968; Hayashi et al., 1966; Isaakson & Phillips, 1968). In animal cells, approxi- mately 1.5% of the nucleotides are methylated (Brown & Attardi, 1965; Hudson et al., 1965) and 90% of these are 2'-Qfmethylations (Vaughan et al., 1967; Lane et al., 1969). l-methylguanine, NZ-methylguanine and NZNZ- dimethylguanine are conspicuously absent in animal high molecular weight rRNA but not in bacterial rRNA (Klags- brun, 1973). The methylation of the smaller component of the high molecular weight RNAS is greater (as per- centage of total nucleotides) than the larger component (Brown & Attardi, 1965; Zimmerman & Holler, 1967). Pseudouridine is also found in the high molecular weight rRNAS (Nichols & Lane, 1967) and one 11 2'-Qfmethy1pseudouridine has been found in HeLa 288 rRNA (Maden & Salim, 1974). A bit of confusion has arisen in the determination of percentage base vs. ribose methylation caused by the methods used for analysis (Attardi & Amaldi, 1970; Burdon, 1971). RNA can be degraded by alkali to produce mostly mononucleotides from base-methyl and normal nucleotides. A dinucleotide is also produced for each 2'-Qfmethylnucleotide due to the alkaline stability of a phOSphodiester bond adjacent to a 2'-Qfmethyl group. Separation of radioactive methyl-labeled mononucleotide from alkaline-stable dinucleotide thus provides a measure of base vs. ribose methylation. It has been reported that the instability of certain methylated bases causes an underestimate of base methylation when this technique is used (Iwanami & Brown, 1968). These authors have favored acid hydrolysis techniques to release free bases from the polynucleotide chain and using this technique have reported that 50-60% of rRNA methylations are on the bases. Acid treatment, however, can release 2'-Qf methyl groups as methanol (Baskin & Dekker, 1967) and perchloric acid digestion has actually been used in an assay for levels of 2'-Qfmethylation (Abbate & Rottman, 1972). Thus, acid hydrolysis may be of value in pre- serving the structure of unstable bases for analysis 12 but it is not worthwhile for determining percentage base vs. ribose methylation unless corrections are made for methanol production. The analysis of the methylated constituents of rRNA has recently culminated in the fingerprinting and sequencing of the methylated, enzymatically produced oligonucleotides from HeLa cell rRNA (Maden & Salim, 1974). Approximately 46 methyl groups were found associated with the 18S rRNA species and 70 methyl groups with the 288 rRNA species. Of these 116 methyl- ations, 11 have been found to be base-methylations. A sequence of two adjacent N6N6-dimethyladenosine residues is present in the 188 Species accounting for 4 of the 11 base-methylations. The alkaline stability of phosphodiester bonds adjacent to 2'-Qfmethy1 groups has permitted the sequenc- ing of oligonucleotides in rRNA containing two and three 2'-Qfmethylnucleotides in succession. The eucaryotic cells that have been examined to date possess two sequences of two 2'-Qfmethylnucleotides in succession and one sequence of three 2'-Qfmethylnucleotides in succession. In Novikoff hepatoma cells, the sequences that have been reported are UmmepUp, UmmepUp and AmmepCmpAp (Choi & Busch, 1974). Maden and Salim (1974) have found similar sequences in HeLa cells (UmmepUp, Ummepr and AmmepCmpAp). Slack and 13 Loenig (1974) have reported that Xenopus laevis contains two sequences of two 2'-Qfmethylnucleotides in succession and one sequence with three adjacent 2'-Qfmethy1nucleo- tides which is GmpAmpAmpAp. This alkaline stable tetra- nucleotide of Xenopus laevis is different from the mam- malian types listed above. All the methylated alkaline- stable tri- and tetranucleotides are present in the 288 species. Much evidence has accumulated demonstrating that the nucleolus is the site of rRNA production. UV-micro- beam irradiation of nucleoli results in the inhibition of synthesis of rRNA (Perry et al., 1961). Synthesis of ribosomes or ribosomal RNA does not occur in mutants that lack nucleolar organizers (Brown & Gurdon, 1964) and the amount of rRNA present in the cytOplasm is in direct proportion to the number of nucleolar organizers (Ritossa & Spiegelman, 1965). The high molecular weight rRNA also selectively hybridizes to the nucleolar DNA (Steele, 1968). All subsequent work using radioactive labeling, cellular fractionation and RNA composition analysis have confirmed the nucleolar location of rRNA synthesis. High molecular weight rRNA is made by RNA poly- merase I. Not only is RNA polymerase I found selectively in the nucleolus where rRNA synthesis occurs (Roeder & Rutter, 1970), but the synthesis of rRNA in isolated 14 nuclei displays the same sensitivity to d-amanitin as the isolated RNA polymerase I (Zylber & Penman, 1971; Reader & Roeder, 1972). It is now clear that the synthesis of both high molecular weight rRNA components in eucaryotes does not occur individually but as a large precursor molecule (~458) in the nucleolus that contains the sequences of both high molecular weight rRNA components (for review, Burdon, 1971). For HeLa and L cells, the 458 nucleolar component emerges very early (3-5 min.) over a background of heterogeneous RNA after a pulse label (Darnell, 1968; Greenberg & Penman, 1966). With longer times, label appears in 328 and 188 RNA and then in mature 288 RNA. Actinomycin D may be used to "chase" label out of 458 and 328 precursors into the mature 288 and 188 ribosomal species. The general processing scheme that has emerged from these kinetics of labeling and cellular fraction- ation studies is that ribosomal RNA is initially made as a 458 precursor that contains both the 288 and 188 sequences. The 458 is then Split in the nucleolus to form molecules which are precursors to the mature rRNA species (328 precursor to 288 rRNA and 208 precursor to 188 rRNA). Other components (418, 368 and 248) are also sometimes observed (Weinberg & Penman, 1970) and their origin will be discussed below. 15 Oligonucleotide mapping after RNAase digestion has shown that a 458 RNA precursor molecule contains the sequence of one 288 molecule and one 188 molecule (Jeanteur et al., 1968; Roberts & D'Ari, 1968). In addition, about 50% of the 458 molecule is comprised of sequences not present in the mature 288 + 188 rRNA. The sequences present in 28S rRNA are also present in the 328 precursor. Molecular weight determinations by equilibrium sedimentation indicate that the 458 molecule is about twice the size as the sum of the 288 + 188 components (McConkey & Hopkins, 1968). RNA-DNA hybridi- zation experiments have indicated that the 458 nucleolar RNA contains regions that correspond to the 288 and 188 rRNAs and also transcribed spacer nucleotides (Jeanteur et al., 1969; Quagliarotti et al., 1970). Thus the pro- cessing of the 458 rRNA precursor is nonconservative in that approximately 50% of the sequence is degraded in the nucleus. The processing scheme for 458 RNA discussed above has been confirmed using an ingenious technique termed secondary structure mapping (Wellauer & Dawid, 1973; Wellauer & Dawid, 1974; Wellauer et al., 1974). This technique is based upon the highly reproducible structure of hairpin loops (corresponding to regions of ordered secondary structure) produced after rRNA or one of its precursors are spread and examined in 16 an electron microscope. The 288 region has been located at the 5'-end of the 458 precursor by partial digestion with a 3'-exonuclease. The 458 precursor is cleaved E3 zigg to produce a 418 RNA (containing 288 + 188 RNA) and a 248 RNA from the 3'-terminus. In L cells, the 418 RNA is then cleaved to 368 and 188 components. The 368 RNA is processed via a 328 intermediate to 288 RNA. In HeLa cells, the 418 RNA is cleaved to produce the 328 component and a 208 component. The 208 RNA is the imme- diate precursor to the 188 RNA. The major pathways of processing of these two cell lines are thus very similar but apparently not identical. The order of cleavage at two Specific Sites on the 418 chain seems to be dif- ferent. In three cell types compared, L cell, HeLa cell and Xenopus laevis, the arrangement of regions on the large precursor is the same, as is the Shapes and location of specific loops in the 28S and 188 regions. Although the secondary structure of the ribosomal 288 and 188 regions of the precursor are highly conserved in evo- lution, the shapes of the loops in the transcribed spacer regions vary greatly and are not conserved. During the synthesis of the 458 precursor, or immediately thereafter, the RNA is specifically methylated (Greenberg & Penman, 1966). The methylation pattern of the 458 nucleolar RNA corresponds to that of an equimolar mixture of 288 and 188 rRNA (Wagner et al., 1967). The 17 methylation pattern of the 328 nucleolar RNA is very Similar to that of 288 rRNA but very different from the 188 rRNA. Almost all of the methylation of rRNA occurs at the 458 precursor stage and processing apparently involves no Significant gain or loss of 2'-Qfmethy1 groups. In the elegant work of Maden and Salim (1974), the fingerprints of methyl-labeled 188 and 288 RNA after nuclease digestion were compared to those of the 458 precursor. Their work confirms that the 458 is the main site of methylation of mammalian rRNA. They also con- firm the work of Zimmerman (1968) who described a late base-methylation event in 188 RNA maturation. Maden and Salim (1974) have found all the 2'-Qfmethylations present in 458 RNA to also be present in 288 + 188 rRNA. 0f the 11 base methylations found in mature rRNA, 5 of them occur in the 288 molecule and 6 in the 188 molecule. The 5 base-methylnucleotides of mature 288 rRNA are also found in the 28S portion of 458 precursor RNA indi- cating that these methylations occur in the nucleolus at the 458 stage. The 6 base methylations of the mature 188 rRNA, however, are not found in the 458 RNA and are thus termed late methylation events. These 6 methylations include the 4 methyl groups donated by the sequence of 6 6 2 N N -dimethyladenosine residues. It is not known whether the late methylations occur in the cytoplasm 18 or in the nucleus. All the methylations that occur on the 458 precursor are conserved in processing and the transcribed Spacer regions that are not conserved are also totally unmethylated. In 1970, Choi and Busch reported the sequence pCmpUp to be the 5'-terminus of 288 rRNA and also 458 RNA in Novikoff hepatoma cells. Maden and Salim (1974) have been unable to confirm this in HeLa cells and could find no evidence for a methylated sequence bearing a 5'-terminal phosphate. Slack and Loenig (1974) have also been unable to find such a sequence in Xenopus laevis rRNA; they found pUp at the 5'-terminus of 288 rRNA and the nucleolar precursor contains pGp at the 5'-terminus. Recently, Choi and Busch (1974) have reported that the alkaline stable tetranucleotide AmmepCmpAp is the 5'-terminus of 288 and 458 RNA of Novikoff hepatoma. In an attempt to see if methylation is required for processing, Vaughan et a1. (1967) have performed labeling and chase studies under conditions of methio— nine deprivation. Cells deprived of methionine incor- porate uridine into 458 and 328 precursor, although somewhat more Slowly, but this RNA is not processed into mature, cytoplasmic ribosomal RNA. Control studies using valine starvation Showed that processing of the 458 precursor to cytoplasmic 288 + 188 RNA did 19 occur. The 458 and 328 RNA made under the conditions of methionine deprivation appeared to be undermethylated. Addition of methionine to the media allowed the under- methylated RNA to be methylated and processed normally into cytoplasmic 288 and 188 rRNA. Until recently, researchers were unable to detect any larger precursors to bacterial 238 and 168 rRNA; many, in fact, believed that such precursors did not exist and that the 238 and 16S RNAS were synthesized separately. Use of a ribonuclease III mutant of Escher- ichia coli has permitted the observation of a 308 pre- cursor that contains the sequences of both 238 and 168 RNAS (Nikolaev et al., 1973; Dunn & Studier, 1973). The 308 RNA can be cleaved to 258 and 17.58 RNAS, the immediate precursors to 238 and 168 RNAS. The 308 pre- cursor is apparently about 27% larger than the sum of 23S and 168 RNAS (Nikolaev et al., 1973). 5.88 RNA (formerly 78 RNA) is attached by non- covalent linkage to a sequence in 288 rRNA and the HeLa 5.88 RNA has previously been reported to be unmethylated (Pene et al., 1968; Weinberg & Penman, 1968). The yeast, Saccharomyces cerevisiae, 5.88 RNA has been sequenced with no evidence for methylated nucleotides (Rubin, 1973). The 5.88 RNA is probably located some distance away from the 288 sequence within the precursor molecule and probably remains associated with a specific 288 20 sequence after the nucleotides joining them have been removed by nuclease action (Speirs & Birnstiel, 1974). In contrast, Maden and Salim (1974) have found the sequence GmpC in 5.88 RNA and reported that this is the only such sequence for the entire 288 + 5.88 complex. This same species was found in 458 RNA and the authors concluded that the 5.88 methylation occurs at the 458 precursor level. The genes for ribosomal RNA are among the few eucaryotic genes that have been isolated. Brown (1973) has described the arrangement of ribosomal genes in two related amphibians, the African clawed toads XenoBuS laevis and Xenopus mullieri. The sequence of the ribo- somal RNA precursor comprises only about 65% of the repeating unit of the ribosomal genes; the remainder is composed of untranscribed spacer DNA. In these organisms each cell contains approximately 450 ribosomal RNA genes. The nucleotide sequence of the rRNAs of the two Species are virtually identical. The untranscribed Spacer segments, however, have diverged considerably (at least 10%) during evolution. This is remarkable when one considers that the untranscribed spacer segments in the rRNA genes of any one Species, although markedly different from those of the other species, are very Similar to one another, perhaps identical. The 58 RNA molecules that have been examined to date are high in GC content and contain no methylated 21 residues (Brownlee et al., 1967; Brownlee et al., 1972; Labrie & Sanger, 1969; Averner & Pace, 1972). The sequence of the 58 RNA is highly conserved through evolution. A high proportion of the 58 RNA molecules are found with ppGp and pppGp 5'-termini arising from primary transcription (Hatlen et al., 1969). 58 RNA is not transcribed in the nucleolus with the high molecular weight rRNA components since the 58 genes are not located there (Pardue et al., 1973). The a-amanitin sensitivity of 58 RNA synthesis in isolated nuclei has shown that RNA polymerase III is responsible for its synthesis (Weinmann & Roeder, 1974). The Synthesis of Messenger RNA A small portion (about 5%) of cytoplasmic RNA is mRNA. It is generally associated with ribosomes form- ing polysomes and is responsible for the synthesis of cellular proteins. Since most cells are engaged in the synthesis of many different proteins, mRNA populations are generally quite heterogeneous. The isolation of mRNA has posed considerable problems due to its heterogeneity and the low percentage of the total RNA pOpulation that is represented by mRNA. In the past, two techniques have generally been employed for its detection in eucaryotic cells. The radioactively labeled RNA present in polysomes after a short incorpor- ation consists primarily of a heterogeneous population 22 of DNA-like RNA molecules with sedimentation values of approximately 108 to 308 that is mRNA (Darnell, 1968; Perry & Kelley, 1968). After longer labeling periods, 188 rRNA followed by 288 rRNA appear. Messenger RNA is known to turn over more rapidly than other cytOplasmic RNA species (Singer & Penman, 1973; Perry & Kelley, 1973) and the rapid appearance of mRNA during a pulse label may be due to a more rapid rate of synthesis as well as a Shorter processing time. Low levels of Actinomycin D (0.04 to 0.08 ug/ml) can be used to selectively inhibit rRNA synthesis (Perry & Kelley, 1969; Penman et al., 1968). Use of this technique allows continued labeling of mRNA without appreciable interference by rRNA labeling. The distributions in HeLa cells of protein and messenger Sizes are closely parallel (Davidson & Britten, 1973) and the mRNA is generally not long enough to code for more than one protein. All analyses of eucaryotic, non- viral messengers to date have confirmed their monocis- tronic nature. Each messenger is Slightly longer than needed to code for its protein. The discovery of a sequence of adenylic acid residues at the 3'-end of mRNA has not only stimulated considerable interest in its role in mRNA function but has also been of great technical importance for the isolation of pure mRNA. Among the early reports on the existence of poly(A) were those of Edmonds and 23 Caramela (1969) using HeLa cells and Kates and Beeson (1970) with vaccinia virus cores. These early obser- vations were followed by a large number of reports detailing methods of isolation, the Site and mode of synthesis, location and the function of the poly(A). The poly(A) segments are found associated with pulse labeled polysomal RNA and RNA labeled in the presence of a low level of Actinomycin D (Kates, 1970; Darnell et al., 1971; Lee et al., 1971; Edmonds et al., 1971). These reports also Show that the poly(A) is in fact covalently bound to the mRNA. Most evidence has indicated that mRNA is gen- erally made as a longer precursor in the nucleus, a part of the hnRNA pOpulation (for review, Weinberg, 1973; Lewin, 1975a; Lewin, 1975b). When low levels of Actino- mycin D are used to inhibit rRNA synthesis, radioactive RNA first appears in the nucleus, in what is termed hnRNA, before radioactive mRNA appears in the cytoplasm (Penman et al., 1968; Perry & Kelley, 1970). The hnRNA has many of the properties one would expect for a pre- cursor to mRNA. Like mRNA, it has a base composition that resembles that of the DNA. It is heterogeneous in size, as one would expect for a precursor to the heter- ogeneous population of mRNAS in the cytoplasm. HnRNA has a marked tendency to aggregate and many estimates of its Size based on nondenaturing measurements are too 24 large. Rigorous denaturation conditions, however, indi- cate that hnRNA is on the average about five times larger than mRNA (Derman & Darnell, 1974). HnRNA is believed to be made in the nucleoplasm by RNA polymerase II (Roeder & Rutter, 1970; Zylber & Penman, 1971). Some hnRNA mole- cules also contain 3'-terminal poly(A) similar to mRNA. The poly(A) containing RNA may be separated from the rest of the RNA by the use of the affinity of poly(A) for complementary polynucleotides. Oligo(dt) may be covalently affixed to cellulose and under the proper ionic conditions, the Oligo(dt)-cellulose selectively retains poly(A)+RNA (Aviv & Leder, 1972). Other affinity chromatography techniques that have been used include poly(U)-Sepharose (Wagner et al., 1971; Adesnik et al., 1972) and poly(U) glass fiber filters (Sheldon et al., 1972). The tendency of poly(A) to bind to cellu- lose nitrate membrane filters (Millipore filters) has also been utilized in mRNA purification (Brawerman et al., 1972; Rosenfeld et al., 1972). Poly(A) may be complexed with poly(U) and bound to hydroxyapatite (Greenberg & Perry, 1972a). These procedures provide rather convenient means for the isolation of mRNA. The most commonly employed techniques are Oligo(dt) cellulose and poly(U)- Sepharose. .The poly(A) segment itself may be isolated by RNAase treatments that degrade heteropolymers but not poly(A). Poly(A) is resistant to digestion by 25 pancreatic RNAase and T1 RNAase at high ionic strength and this may be used to isolate the poly(A) segment intact (Darnell et al., 1971; Lee et al., 1971). Various techniques have been used to Show that the poly(A) is attached to the 3'-terminus of the mRNA. Alkaline hydrolysis of the poly(A) derived from mRNA produces approximately one residue of adenosine per 200 residues of AMP (Mendecki et al., 1972). The adenosine residue must be derived from the 3'-end of the RNA. Digestion of mRNA with a Specific 3'-exoribonuclease releases adenylic acid residues from poly(A) before the rest of the molecule (Molloy et al., 1972). Periodate oxidation and labeling of the 3'-termina1 nucleoside by reduction with 3H-NaBH4 labels the poly(A) segment, con- sistent with its 3'-terminal location (Yogo et al., 1972). The poly(A) appears to be synthesized in the nucleus where it is transported with mRNA to the cyto- plasm. In labelings as short as 45 sec., 90% of the cellular poly(A) is located in the nucleus (Jelinek et al., 1973). With longer times, the bulk of the labeled poly(A) is found in the cyt0plasm. No poly(A) is found unattached to higher molecular weight RNA (Jelinek et al., 1973) and cellular and viral genomes do not contain tracts of poly(dAsz) that could code for poly(A) (Wall et al., 1972; Bishop et al., 1974). It thus appears that poly(A) is not made as part of the original RNA 26 transcript but is added to the RNA chain post- transcriptionally. Nuclear and cytOplasmic enzymes that add adenylic acid residues to the 3'-end of RNA chains have been described (Winters & Edmonds, 1973; Tsiapalis et al., 1973). The many experiments that have been performed to elucidate the function of the poly(A) segment have not done much to define its role. The experimental rationale has been guided by three hypotheses: poly(A) is involved or is necessary in the transport of mRNA precursors from the nucleus to the cytOplasm; poly(A) is necessary for the proper translation of mRNA; and poly(A) is involved in the turnover of mRNA in the cyto- plasm. The nucleotide precursors 3'-deoxyadenosine and 3'-deoxycytidine have been used as RNA synthesis inhibi- tors to probe the role of poly(A) in transport of mRNA precursors from nucleus to cytoplasm. 3'-deoxyadenosine inhibits the synthesis of rRNA precursor (Siev et al., 1969; Penman et al., 1970) and poly(A) (Adesnik et al., 1972; Darnell et al., 1973). A considerably shorter polynucleotide results from the inhibition of synthesis of these Species (Siev et al., 1969; Mendecki et al., 1972). While neither the synthesis of hnRNA nor the size of the hnRNA product are altered by 3'-deoxyadeno- sine treatment, the synthesis of mRNA is drastically Cal.‘ ‘5".— 27 inhibited (Penman et al., 1970; Darnell et al., 1973). It has been proposed that proper addition of poly(A) in the nucleus is necessary for the transport of mRNA pre- cursors to the cytoplasm and that the inhibition of poly(A) addition is the cause of 3'-deoxyadenosine inhibition of mRNA appearance in the cytoplasm (Darnell et al., 1973; Adesnik et al., 1972). 3'-deoxycytidine Should not inhibit poly(A) synthesis. Consistent with the above hypothesis, Abelson and Penman (1972) have found that 3'-deoxycytidine can inhibit nucleolar rRNA precursor synthesis, but both hnRNA and mRNA synthesis go unaffected. Any scheme involving poly(A) in the transport of mRNA from nucleus to cytoplasm is burdened with the explanation of several other observations. The mRNA for histones contains no poly(A) segment (Adesnik & Darnell, 1972; Greenberg & Perry, 1972a). Contrary to an earlier report with L cells (Greenberg & Perry, 1972a), Milcarek et al. (1974) have recently reported that as much as 30% of HeLa mRNA contains no poly(A) segment and that this is not due to poly(A)+RNA that has lost poly(A) via nuclease or some other means. Thus, mRNA that lacks poly(A) is transported adequately from the nucleus. In fact, histone mRNA has been found to escape the usual 15 min. nuclear processing time and to appear almost immediately in the cytoplasm after its synthesis (Adesnik 28 & Darnell, 1972; Schochetman & Perry, 1972). Messenger RNAS of viruses that replicate in the cytoplasm still contain poly(A) even though they do not go through a nuclear cycle (Yogo et al., 1972; Eaton et al., 1972; Armstrong et al., 1972). A series of reports has appeared Showing that mRNA from which the poly(A) segment had been removed functions as well as polyadenylated mRNA in Eg'ziggg protein synthesis (Bard et al., 1974; Munoz & Darnell,., 1974; Williamson et al., 1974). Hemoglobin mRNA from which the poly(A) segment had been removed also is adequately translated when injected into Xenopus oocytes (Huez et al., 1974). Thus poly(A) does not seem to be required for translation. The newly made poly(A) of the nucleus (approxi- mately 200 nucleotides) has been found to be longer than the poly(A) in the cytoplasm and the length of poly(A) in the mRNA decreases with age (Greenberg & Perry, 1972b; Sheiness & Darnell, 1973; BishOp et al., 1974). The shortening of poly(A) may then be implicated in the turn- over of mRNA. The labeling kinetics of L cell mRNA, however, have suggested to Perry and Kelley (1973) that the turnover of any particular mRNA is independent of its age. Huez et a1. (1974) have made the interesting observation that the efficiency of translation in Xenopus oocytes of globin message, from which poly(A) 29 had been removed, declines very rapidly compared with the high level of continued translation of globin message containing poly(A). Whether this is due to destruction of the mRNA is not known. Isolation of mRNA by virtue of its poly(A) seg- ment has paved the way for fractionation techniques that allow the isolation of specific messengers. Most of these isolations utilize either a highly differentiated cell synthesizing a specific protein product or unusual properties (e.g., size) of the message. Hemoglobin mRNA (a and B) was the first and among the easiest to isolate (Chantrenne et al., 1967; Lockard & Lingrel, 1969; Lanyon et al., 1972) since both these properties were met. Over 90% of the protein synthesis of the reticulocyte is devoted to hemoglobin and the mRNA (98) is considerably smaller than the bulk of the mRNA observed in other systems. Separation of a mRNA and 8 mRNA has recently been achieved (Morrison et al., 1974). Other isolated mRNAS include histone mRNA (Levy et al., 1975), immuno- globin mRNA (Swan et al., 1972; Schecter, 1973), oval- bumin mRNA (Palacios et al., 1973), lens crystallin mRNA (Bloemendal et al., 1973) and silk fibroin mRNA (Suzuki % Brown, 1972). While most mRNA molecules contain a poly(A) segment, only about 20% of hnRNA molecules contain poly(A) (Greenberg & Perry, 1972a; Jelinek et al., 1973). 30 Studies on the kinetics of disappearance of nuclear poly(A) and the appearance of cytoplasmic poly(A) have led to conflicting reports on whether poly(A) is con- served in a hnRNA-mRNA transition (Jelinek et al., 1973; Perry et al., 1974; Puckett et al., 1975). It can be concluded that the rate of cytOplasmic accumulation of poly(A) is too rapid to be explained by conservation of poly(A) synthesized in the nucleus if mRNA is not very rapidly turned over in the cytoplasm. Nevertheless, the presence of poly(A) in both species is evidence for at least a portion of the hnRNA being the immediate pre- cursor to mRNA. HnRNA molecules with no poly(A) tail could arise from a number of different sources: newly synthesized molecules which have not yet received poly(A), molecules never destined to receive poly(A) or fragments arising from processing of primary transcripts destined to be degraded. The presence of poly(A), the base composition, the sensitivity of synthesis to Actinomycin D and the heterogeneous size provide good evidence for the precursor-product relationship of hnRNA and mRNA. Rigorous proof, however, requires that under conditions where transcription of hnRNA is inhibited, mRNA accumu- lation is derived from hnRNA breakdown. This has been very difficult to Show for hnRNA-mRNA for several reasons. One cannot do pulse-chase experiments with 31 unlabeled precursor due to the large pool sizes of eucaryotic cells. Actinomycin D can be used to inhibit transcription of hnRNA, but it also interferes with proper processing. HnRNA turns over very rapidly (half- life approximately 23 minutes) and only about 2% of the nucleotides of hnRNA have been estimated to enter mRNA (Brandhorst & McConkey, 1974). Highly radioactive DNA complementary to mRNA (cDNA) may be synthesized £2.21EES using reverse tran- scriptase and Oligo(dt) as a primer. The cDNA may then be used as a probe for sequences in high molecular weight nuclear RNA containing the mRNA sequence. Using such a technique, Macnaughton et al. (1974) have been able to isolate a single peak of nuclear RNA (148) that is a presumed precursor to duck globin mRNA in erythroblasts. The aggregation of hnRNA is an established phenomenon but the workers were careful to use rigorous denaturing techniques in the isolation and Size characterization of the presumed globin mRNA precursor. The 148 RNA cor- responds to molecules about three times the Size of globin mRNA. It is not known whether this 148 RNA is a primary transcript or a processing intermediate. Using Similar techniques for oviduct nuclear RNA, McKnight and Schimke (1974) were unable to find any precursor to ovalbumin mRNA (18S) larger than 288. Their technique could have detected as little as one molecule per cell. 32 RNA-DNA hybridization has been an effective tool for study of the hnRNA-mRNA relationship (for review, Lewin, 1975b). The repetitive content of hnRNA is never less than that of mRNA and is usually greater. The messengers of mammalian cells are derived mostly from nonrepetitive components of the genome. Five to 15% of different mRNAs are generally found to anneal to repetitive DNA sequences while corresponding hnRNA anneals 10 to 25% to repetitive DNA. Klein et a1. (1974) have Shown that, at least in HeLa cells, the minor fraction of mRNA that hybridizes at low CoT values (repetitive) represents a population of molecules separate from the nonrepetitive sequence transcripts. The mRNA does not contain repetitive sequence elements covalently linked to nonrepetitive sequence transcripts. Several reports have demonstrated that hnRNA molecules do contain repetitive sequences interspersed with non- repetitive sequences (Molloy et al., 1974; Smith et al., 1974; Holmes & Bonner, 1973). Hames and Perry have observed that the total sequence complexity of hnRNA is approximately 4x greater than that of mRNA, indicating some selection process must occur in the nucleus for transport of mRNA sequences to the cytoplasm (Lewin, 1975b). Hames and Perry were also able to isolate non- repetitive mRNA sequences in hnRNA (Lewin, 1975b). 33 Certain unusual regions of unknown function have been described for hnRNA. Nakazato et a1. (1974) have Shown that HeLa hnRNA contains short internal sequences of oligo(A) that are transcribed from the DNA. HnRNA is also known to contain RNAase resistant regions that are double-stranded in nature (Jelinek & Darnell, 1972). It has been reported that these double-stranded regions behave as repetitive sequences (Harel & Mon- taignier, 1971; Kimball & Duesberg, 1971). The occur- rence of oligo(U) approximately 30 residues long has also been reported in hnRNA (Burdon & Shenkin, 1972; Brawerman et al., 1972). Early work with bacteria produced strong evidence for the essential absence of methylation in the mRNA, being no higher than one methylation per 3500 nucleotides (Moore, 1966). T4 and R17 nRNA were Similarly found to be unmethylated. Analyses in eucaryotic cells have been more difficult in the past due to the difficulty in obtaining labeled mRNA uncontaminated with rRNA, which has a rather high level of methylation. Early studies with mammalian hnRNA indicated that methylation was either nonexistent or very low (Perry et al., 1970; Muramatsu & Fujisawa, 1968; Johnson, 1970). The avail- ability of techniques utilizing the poly(A) segment to obtain pure mRNA makes it possible to carefully study if eucaryotic mRNA is methylated. In 1974, Perry and 34 Kelley reported that mouse L cell mRNA contains approxi- mately 2.2 methyl groups per 1000 nucleotides (about 6 methyl groups per average mRNA molecule). This is one- sixth the level found in rRNA. The methylation of hnRNA was found to be even less than the 0.22% but still sig- nificant. This very low level probably explains the apparent absence observed in previous reports. A con- siderable portion of the methyl radioactivity was present as an alkaline stable oligonucleotide more negatively charged than -3 (larger than dinucleotide). During the course of the work described in this dissertation, numerous articles have appeared describing the methylation of cellular and viral mRNAs. A number of unpublished and unexplained observations prompted Rottman et al. (1974) to postulate that an unusual structure is present at the 5'-terminus of mRNAs. In this Structure, 7-methylguanosine is bonded to a 2'-Qf methylnucleoside via pyrophosphate in a 5'-5' bond rather than the usual 3'-5' phosphodiester bond. A second 2'-gf methylnucleoside was postulated to sometimes be present following the first 2'-Qfmethylnucleoside. The structure 7 5' 5 may be represented m G pp 2,2,7G5' '(Nmp)l or 2Np. An analogous structure, m3 pp 'AmpUmp has been reported for the 5'-terminus of certain low molecular weight nuclear RNAS (Ro-Choi et al., 1974; Ro-Choi et al., 1975). The ability to synthesize high Specific activity methylated animal virus mRNA E2_vitro using virion 35 associated RNA polymerase has facilitated the charac- terization of the viral mRNA methylated nucleotides. In these 12.21252 systems, the methylation occurs exclusively as a large alkaline-stable oligonucleotide with a charge estimated to be -5 to -6 by DEAE-cellulose chromatography in 7M urea (Urushibara et al., 1975; Wei & Moss, 1975; Furuichi et al., 1975a; Rhodes et al., 1974; Furuichi & Miura, 1975). Furuichi et al. (1975a) have analyzed this large oligonucleotide of reovirus mRNA synthesized 12.!iEEE and found that it contains 7-methylguanosine and 2'-Qfmethylguanosine in approxi- mately equal proportions. After labeling with (8,7-32P)- GTP, the large oligonucleotide was found to be blocked with respect to removal of 32 P by alkaline phosphatase. Periodate oxidation followed by reduction with 3H boro- hydride labeled the 7-methylguanosine of the large oligonucleotide, consistent with the structure origi- nally postulated by Rottman et a1. (1974) to be present on the 5'-end of eucaryotic mRNA. Furuichi et al. (1975a) I proposed the structure m7G5 ppp 'Gmpr for the 5'-terminus of reovirus mRNA. The same structure has been found in the mRNA Strand of the virion double-stranded RNA genome. The (-)strand does not contain this blocked structure (cap) and has been identified as ppGp (Furuichi et al., 1975b; Chow & Shatkin, 1975). 36 Wei and Moss (1975) have analyzed the vaccinia virus mRNA made $2.!lEEQ and have concluded the 5'- terminus consists of a similar structure, only the 2'-ge methylnucleoside is a mixture of Gm and Am. The mRNA of vesicular stomatitis virus (VSV) syn- thesized $3.2Eyg and iE.YiE£2.has been analyzed for methylation. The mRNA synthesized ig_gi£52_by the virion 5| Amp at the associated RNA polymerase contains m7G 'ppp 5'-terminus (Abraham et al., 1975a). VSV specific mRNAs isolated from infected cells contain a Similar methylated structure except the methylated adenosine has one more methyl in addition to the 2'-Qfmethyl group (Moyer et al., 1975). The genome RNA (-strand) is not blocked but con- tains ppAp at the 5'-terminus. When VSV mRNA is synthe- sized $2.21EEQ without S-adenosylmethionine, the 5'- terminal structure is Gs'pppS'Ap (Abraham et al., 1975b). CytOplasmic polyhedrosis virus has been reported I I 7 5 ppp5 AmpGp at the 5'-terminus of its to contain m G mRNA (Furuichi & Muira, 1975). Analysis of the methylated components of cellular mRNA has progressed less rapidly. This is probably due at least in part to the greater complexity of methylation. About 50% has been found in NG-methyladenosine that runs as a mononucleotide after KOH hydrolysis (Perry et al., 1975). This mononucleotide has not yet been observed with viral mRNAs synthesized E3 vitro. Its presence 37 has recently been observed in SV40 specific mRNA of infected cells (Lavi & Shatkin, 1975) but Moyer et a1. (1975) have reported its total absence in VSV specific mRNA of infected cells. A dinucleotide peak has also been observed in the poly(A)+mRNA fractions of L cells (Perry et al., 1975) and HeLa cells (Wei et al., 1975) but Perry et a1. (1975) suspected this may have been due to rRNA contamination, Since the dinucleotide peak was not observed in cells that were labeled in the presence of a dose of Actinomycin D that suppresses rRNA synthesis. As with viral mRNAs, a large alkaline stable oligonucleotide was also observed in these cellular mRNAs. Perry et al. (1975) have presented evidence that this oligonucleotide is derived from a blocked 5'-terminus. It was resistant to phosphorylation by polynucleotide kinase even after phosphatase treatment and phosphatase treatment only reduced its charge by two units. Wei et a1. (1975) have found that periodate oxidation and B- elimination removes the 7-methy1guanosine from the large oligonucleotide and have concluded that a 5'-5' bond similar to the viral 5'-terminus is also present. Furuichi et al. (1975c) have also analyzed HeLa cell mRNA and found blocked 5'-terminal structures of I I I I two general types: m7G5 ppp5 Nmpr and m7G5 ppp5 Nmp- Nmpr. In addition, about one-third of the 3 H-methyl label were present in NG-methyladenosine. All four 2'- Qfmethylnucleosides were identified. 38 Adams and Cory (1975) have recently fingerprinted and sequenced methyl-labeled oligonucleotides of mouse myeloma cell mRNA. In addition to N6—methyladenosine, the authors identified 10 of the common 5'-termina1 structures. m7GS'ppp5'CmpUp comprised 21% of the 5' structures. 5'-terminal sequences with two adjacent 2'-Qfmethylnucleotides were also identified. Recent results from Shatkin's laboratory indicate that the methylated 5'-termina1 structure is required for translation (Both et al., 1975; Muthukrishnan et al., 1975). Capped and methylated reovirus, VSV, and hemo- globin mRNA stimulated protein synthesis i£.X$E£2 greater than RNA that is not blocked and methylated. S-adenosyl- methionine stimulated the translation of the RNA lacking methyls and caps and it was found that the RNA could be capped and methylated Ea XEEEQ. Inhibition of methylation ;g_giggg did not affect translation of the normally methylated RNA but reduced the translation of the unmethylated RNAS to background levels. In the early phases of my graduate research, it was generally assumed that mammalian mRNA was not methylated but I felt that the question had not been answered satisfactorily. My initial laboratory efforts involved the establishment of a continuous cell culture line (Novikoff hepatoma N181) as well as mastering techniques for radioactive labeling of RNA, isolation 39 of nuclei, cytoplasm and polyribosomes and the purifi- cation of mRNA. These techniques were necessary to per- form experiments to determine if mRNA contained methylated nucleotides. As these preliminaries were completed and as I prepared to attempt my first methyl-labeling of mRNA, the report of Perry and Kelley (1974) appeared in the literature stating that L cell mRNA was methylated at one-Sixth the level of rRNA. At this point, I felt it was important to confirm that mammalian mRNA was methylated and more importantly, to identify and char- acterize the methylated nucleosides. Techniques for the determination of normal and of 2'-Qfmethylnucleoside distributions using unlabeled RNA pioneered by Lee Pike in Dr. Rottman's laboratory (Pike & Rottman, 1974) were applied to the analysis of methyl-labeled mRNA. Following complete enzymatic digestion to nucleosides, 2'-Qfmethylnucleosides were separated from normal and base-methylnucleosides by chromatography on DEAR-cellulose (borate). High Speed liquid chromatography on Aminex A-5 was effectively used for 2'-gfmethylnucleoside distributions as well as for base-methylnucleoside distributions. Additional infor- mation on the unusual nature of the methylation in mRNA was obtained using acid hydrolysis, KOH hydrolysis followed by separation of the oligonucleotides on the basis of charge by pellionex-WAX chromatography, 40 periodate oxidation and B-elimination. The results of these analyses are presented in parts I and II of this dissertation. Part III deals with a particular aspect of the synthesis of mRNA precursor in Novikoff hepatoma cells, viz. its resistance to inhibition by cordycepin (3'dA). REFERENCES Abbate, J. and Rottman, F. (1972), Anal. Biochem. 31, 378. Abelson, H. T. and Penman, S. (1972), Biochim. Biophys. Acta 277, 129. Abraham, 6., Rhodes, D. P. and Banerjee, A. K. (1975a), Cell 2, 51. Abraham, 6., Rhodes, D. P. and Banerjee, A. K. (1975b), Nature 255, 37. Adams, J. M. and Cory, S. (1975), Nature 255, 28. Adesnik, M. and Darnell, J. E. (1972), J. Mol. Biol. g1, 397. Adesnik, M., Salditt, M., Thomas, W. and Darnell, J. E. (1972), J. Mol. Biol. 1;, 21. Altman, S. (1971), Nature New Biology 229, 19. Altman, 8. and Smith, J. D. (1971), Nature New Biology 233, 35. Armstrong, J. A., Edmonds, M., Nakazato, H., Phillips, B. A. and Vaughan, M. H. (1972), Science 176, 526. Attardi, G. and Amaldi, F. (1970), Ann. Rev. Biochem. E2, 183. Averner, M. J. and Pace, N. R. (1972), J. Biol. Chem. 247, 4491. Aviv, H. and Leder, P. (1972), Proc. Nat. Acad. Sci. USA g2, 1408. Baguley, B. C. and Staehelin, M. (1969), Biochemistry E, 257. Bard, E., Efron, D., Marcus, A. and Perry, R. P. (1974), Cell E, 101. 41 42 Baskin, F. and Dekker, C. A. (1967), J. Biol. Chem. 242, 5447. Bernhardt, D. and Darnell, J. E. (1969), J. Mol. Biol. 13, 43. Biezunski, N., Giveon, D. and Littauer, U. F. (1970), Biochim. Biophys. Acta 199, 382. Bishop, J. O., Rosbash, M. and Evans, D. (1974), J. Mol. Biol. EE, 75. Bjork, G. R. and Niedhardt, F. C. (1971), Cancer Res. 3;, 706. Bloemendal, H., Berns, A., Strous, G., Mathews, M. and Lane, C. D. (1973), in RNA Viruses--Ribosomes, 237, Amsterdam-North Holland. Boedtker, H. (1967), Biochemistry Q, 2718. Boedtker, H. (1968), J. Mol. Biol. E2, 61. Borek, E. and Srinivasan, P. R. (1966), Ann. Rev. Biochem. 35, 275. Both, G. W., Banerjee, A. K. and Shatkin, A. J. (1975), Proc. Nat. Acad. Sci. USA 1;, 1189. Brandhorst, B. P. and McConkey, E. H. (1974), J. Mol. Biol. fig, 451. Brawerman, G., Mendecki, J. and Lee, S. Y. (1972), Biochemistry El, 637. Brown, D. D. (1973), Scientific American 229 no. 2, 20. Brown, G. M. and Attardi, G. (1965), Biochem. Bi0phys. Res. Commun. 32, 298. Brown, D. D. and Gurdon, J. B. (1964), Proc. Nat. Acad. Sci. USA EE, 139. Brownlee, G. G., Cartwright, E. McShane, T. and Williamson, R. (1972), Fed. Eur. Biochem. Soc. Lett. EE, 8. Brownlee, G. G., Sanger, E. and Barrell, B. G. (1967), Nature 215, 735. Burdon, R. H. (1971), Prog. Nucl. Acid Res. Mol. Biol. ii, 33. 43 Burdon, R. H., Martin, B. T. and Lal, B. M. (1967), J. Mol. Biol. EE, 357. Burdon, R. H. and Shenkin, A. (1972), FEBS Lett. E4, 11. Blatti, 8., Ingles, C. J., Lindell, T. J., Morris, P. W., Weaver, R. F., Weinberg, F. and Rutter, W. J. (1970), Cold Spring Harbor Symp. Quant. Biol. EE, 649. Burgess, R. R. (1973), Ann. Rev. Biochem. 19, 711. Capra, J. D. and Peterkofsky, A. (1968), J. Mol. Biol. 3;, 591. Chantrenne, H., Burny, S. and Marbaix, G. (1967), Progr. Nucl. Acid Res. Mol. Biol. 1, 173. Chheda, G. B., Hong, C. I., Piskorz, C. F. and Harmon, Go A. (1972) ' BiOChemo J. 127' 5150 Choi, Y. C. and Busch, H. (1970), J. Biol. Chem. 245, 1954. Choi, Y. C. and Busch, H. (1974), Biochem. Biophys. Res. Commun. EE, 674. Chow, N. and Shatkin, A. J. (1975), J. Virol. EE, 1057. Daniel, V., Sarid, S. and Littauer, U. Z. (1970), Science 167, 1682. Darnell, J. E. (1968), Bacteriol. Rev. EE, 262. Darnell, J. E., Jelinek, W. R. and Molloy, G. R. (1973), Science 181, 1215. Darnell, J. E., Wall, R. and Tushinski, R. J. (1971), Proc. Nat. Acad. Sci. USA EE, 1321. Davidson, E. H. and Britten, R. J. (1973), Quart. Rev. Biol. fig, 565. Derman, E. and Darnell, J. E. (1974), Cell 2, 255. Deutscher, M. P. (1973), Prog. Nucl. Acid Res. Mol. Biol. E3, 51. Dunn, J. J. and Studier, F. W. (1973), Proc. Nat. Acad. Sci. USA 1Q, 3296. 44 Eaton, B. T., Donaghue, T. P. and Faulkner, P. (1972), Nature New Biol. 238, 109. Edmonds, M. and Caramela, M. G. (1969), J. Biol. Chem. 244, 1314. Edmonds, M., Vaughan, M. H. and Nakazato, H. (1971), Proc. Nat. Acad. Sci. USA EE, 1336. Fellner, P. and Sanger, F. (1968), Nature 219, 236. Fleissner, E. (1967), Biochemistry E, 621. Furuichi, Y. and Muira, K. (1975), Nature 253, 374. Furuichi, Y., Morgan, M., Muthukrishnan, 8. and Shatkin, ’ A. (1975a), Proc. Nat. Acad. Sci. USA 13, 362. Furuichi, Y., Muthukrishnan, S. and Shatkin, A. J. (1975b), Proc. Nat. Acad. Sci. USA 1;, 742. Furuichi, Y., Morgan, M., Shatkin, A. J., Jelinek, W., Salditt-Georgieff, M., and Darnell, J. E. (1975c), Proc. Nat. Acad. Sci. USA (in press). Gefter, M. L. (1969), Biochem. Biophys. Res. Commun. E9, 435. Greenberg, H. and Penman, S. (1966), J. Mol. Biol. EE, 527. Greenberg, J. R. and Perry, R. P. (1972a), J. Mol. Biol. 1;, 91. Greenberg, J. R. and Perry, R. P. (1972b), Biochim. et Biophys. Acta 287, 361. Hall, R. H. (1970), Progr. Nucl. Acid Res. Mol. Biol. EQ, 57. Harel, L. and Montaignier, L. (1971), Nature New Biol. 229, 106. Hatlen, L. E., Amaldi, F. and Attardi, G. (1969), Biochemistry E, 4989. Hayashi, H., Osawa, S. and Miura, K. (1966), Biochim. Biophys. Acta 129, 519. Holmes, D. 8. and Bonner, J. (1973), Biochemistry lg, 2330. Hudson, L., Gray, M. and Lane, B. G. (1965), Biochemistry i, 2009. 45 Huez, G., Marbaix, G., Hubert, E., Leclercq, M., Nudel, U., Soreq, H., Solomon, R., Lebleu, B., Revel, M. and Littauer, U. Z. (1974), Proc. Nat. Acad. Sci. USA El, 3143. Isaksson, L. A. and Phillips, 0. H. (1968), Biochim. Biophys. Acta 155, 63. Iwanami, Y. and Brown, G. M. (1968), Arch. Biochem. Biophys. 126, 8. Jacob, S. T. (1973), Prog. Nucl. Acid Res. Mol. Biol. EE, 93. Jeanteur, P., Amaldi, F. and Attardi, G. (1968), J. Mol. Biol. EE, 757. Jeanteur, P. and Attardi, G. (1969), J. Mol. Biol. fig, 305. Jelinek, W. and Darnell, J. E. (1974), Proc. Nat. Acad. Sci. USA EE, 2537. Jelinek, W., Adesnik, M., Salditt, M., Sheiness, D., Wall, R., Molloy, G., Phillipson, L. and Darnell, J. E. (1973), J. Mol. Biol. 1E, 515. Johnson, R. (1970), Biochem. J. 119, 699. Johnson, L. and 8611, D. (1970), Proc. Nat. Acad. Sci. USA 6_7, 943. Kates, J. (1970), Cold Spring Harbor Symp. Quant. Biol. EE, 743. Kay, J. E. and Cooper, H. L. (1969), Biochim. Biophys. Acta 186, 62. Kimball, P. C. and Duesberg, P. (1971), J. Virol. Z, 697. Klagsbrun, M. (1973), J. Biol. Chem. 248, 2612. Klein, W. H., Murphy, W., Attardi, G., Britten, R. J. and Davidson, E. H. (1974), Proc. Nat. Acad. Sci. USA 1;, 1785. Labrie, F. and Sanger, F. (1969), Biochem. J. 114, 29P. Lal, B. M. and Burdon, R. H. (1967), Nature 213, 1134. Lane, B. G. and Tamaoki, T. (1969), Biochim. Biophys. Acta 179, 332. 46 Lanyon, W. G., Paul, J. and Williamson, R. (1972), Eur. J. Biochem. EE, 38. Lavi, 8. and Shatkin, A. J. (1975), Proc. Nat. Acad. Sci. USA (in press). Lee, S. Y., Mendecki, J. and Brawerman, G. (1971), Proc. Nat. Acad. Sci. USA EE, 1331. Levy, 8., Wood, P., Grunstein, M. and Kedes, L. (1975), Cell E, 239. Lewin, B. (1975a), Cell E, 11. Lewin, B. (1975b), Cell E, 77. Lipsett, M. N. (1972), J. Biol. Chem. 247, 1458. Lockard, R. E. and Lingrel, J. B. (1969), Biochem. . Biophys. Res. Commun. El, 204. Macnaughton, M., Freeman, K. B. and Bishop, J. O. (1974), Cell E, 117. Maden, B. E. H. and Salim, M. (1974), J. Mol. Biol. EE, 133. McConkey, E. H. and Hopkins, J. W. (1969), J. Mol. Biol. EE, 545. McKnight, G. 8. and Schimke, R. T. (1974), Proc. Nat. Acad. Sci. USA EE, 4327. Mendecki, J., Lee, S. Y. and Brawerman, G. (1972), Bio- chemistry EE, 792. Milcarek, C., Price, R. P. and Penman, S. (1974), Cell E, l. Mittelman, A., Booney, R., Hall, R. H. and Grace, J. T. (1969), Biochim. Biophys. Acta 179, 242. Molloy, G. R., Jelinek, W., Salditt, M. and Darnell, J. E. (1974), Cell E, 43. Molloy, G., Sporn, M., Kelley, D. E. and Perry, R. P. (1972), Biochemistry EE, 3256. Moore, P. B. (1966), J. Mol. Biol. EE, 38. Morrison, M. R., Brinkley, 8. A., Gorski, J. and Lingrel, J. B. (1974), J. Biol. Chem. 249, 5290. 47 Mowshowitz, D. B. (1970), J. Mol. Biol. E9, 143. Moyer, S. A., Abraham, G., Adler, R. and Banerjee, A. K. (1975), Cell E, 59. Munns, T., Podratz, K. and Katzman, P. (1974), Bio- chemistry EE, 4409. Munoz, R. F. and Darnell, J. E. (1974), Cell E, 247. Muramatsu, M. and Fujisawa, T. (1968), Biochim. BiOphys. Acta 157, 476. Muthukrishnan, 8., Both, G. W., Furuichi, Y. and Shatkin, A. J. (1975), Nature 255, 33. Nakazato, H., Edmonds, M. and Kopp, D. W. (1974), Proc. Nat. Acad. Sci. USA EE, 200. Nichols, J. L. and Lane, B. G. (1967), J. Mol. Biol. EE, 477. Nikolaev, N., Silengo, L. and Schlesinger, D. (1973), Proc. Nat. Acad. Sci. USA EE, 3361. Palacios, R., Sullivan, D., Summers, N. M., Kiely, M. L. and Schimke, R. T. (1973), J. Biol. Chem. 248, 540. Pardue, M. L., Brown, D. D. and Birnstiel, M. L. (1973), Chromosoma EE, 191. Pene, J. J., Knight, E. and Darnell, J. E. (1968), J. Mol. Biol. EE, 609. Penman, 8., Rosbash, M. and Penman, M. (1970), Proc. Nat. Acad. Sci. USA El, 1878. Penman, 8., Vesco, C. and Penman, M. (1968), J. Mol. Biol. EE, 49. Perry, R. P., Cheng, T. Y., Freed, J. J., Greenberg, J. R., Kelley, D. E. and Tartof, K. D. (1970), Proc. Nat. Acad. Sci. USA EE, 609. Perry, R. P., Hell, A. and Errera, M. (1961), Biochim. Biophys. Acta EE, 47. Perry, R. P. and Kelley, D. E. (1968), J. Mol. Biol. EE, 37. 48 Perry, R. P. and Kelley, D. E. (1969), J. Cell. Physiol. EE, 235. Perry, R. P. and Kelley, D. E. (1970), J. Cell. Physiol. 1g, 127. Perry, R. P. and Kelley, D. E. (1973), J. Mol. Biol. EE, 681. Perry, R. P. and Kelley, D. E. (1974), Cell E, 37. Perry, R. P., Kelley, D. E., Friderici, K. and Rottman, F. (1975), Cell E, 387. Perry, R. P., Kelley, D. E. and LaTorre, J. (1974), J. Mol. Biol. EE, 315. Pike, L. M. and Rottman, F. (1974), Anal. Biochem. EE, 367. Puckett, L., Chambers, 8. and Darnell, J. E. (1975), Proc. Nat. Acad. Sci. USA EE, 389. Quagliarotti, G., Hidvegi, E., Wikman, J. and Busch, H. (1970), J. Biol. Chem. 245, 1962. Reeder, R. H. and Roeder, R. G. (1972), J. Mol. Biol. El, 433. Rhodes, D., Moyer, S. and Bannerjee, A. (1974), Cell E, 327. Ritossa, F. M. and Spiegelman, 8. (1965), Proc. Nat. Acad. Sci. USA EE, 737. Robertson, H. D., Altman, S. and Smith, J. D. (1972), J. Biol. Chem. 247, 5243. Ro-Choi, T., Reddy, R., Choi, T. C., Rej, N. B. and Henning, D. (1974), Fed. Proc. EE, 1832. Ro-Choi, T. 8., Choi, Y. C., Henning, D., McCloskey, J. and Busch, H. (1975), J. Biol. Chem. 250, 3921. Roeder, R. G. and Rutter, W. J. (1969), Nature 224, 234. Roeder, R. G. and Rutter, W. J. (1970), Proc. Nat. Acad. Sci. USA EE, 675. Rosenfeld, G. C., Comstock, J. P., Means, A. R. and O'Malley, B. W. (1972), Biochem. Biophys. Res. Commun. 21, 387. 49 Rottman, F. M., Shatkin, A. J. and Perry, R. P. (1974), Cell E, 197. Rubin, G. M. (1973), J. Biol. Chem. 248, 3860. Schechter, I. (1973), Proc. Nat. Acad. Sci. USA EE, 2256. Schochetman, G. and Perry, R. P. (1972), J. Mol. Biol. .EE, 591. Schwartz, L. B., Sklar, V. E. F., Jaehning, J. A., Wein- mann, R. and Roeder, R. G. (in press). Sheiness, D. and Darnell, J. E. (1973), Nature New Biology 241, 265. Sheldon, R., Jurale, C. and Kates, J. (1972), Proc. Nat. Acad. Sci. USA EE, 417. Siev, M., Weinberg, R. and Penman, S. (1969), J. Cell Biol. EE, 510. Singer, R. H. and Penman, S. (1973), J. Mol. Biol. EE, 321. Slack, J. M. and Loenig, V. E. (1974), Eur. J. Biochem. EE, 59. Slack, J. M. W. and Loenig, V. E. (1974), Eur. J. Biochem. EE, 69. Smillie, E. J. and Burdon, R. H. (1970), Biochim. Biophys. Acta 213, 248. Smith, M. J., Hough, B. R., Chamberlin, M. E. and David- son, E. H. (1974), J. Mol. Biol. EE, 103. Speirs, J. and Birnstiel, M. L. (1974), J. Mol. Biol. El, 2370 Steele, W. J. (1968), J. Biol. Chem. 243, 3333. Stern, R., Gonano, F., Fleissner, E. and Littauer, U. Z. (1970), Biochemistry E, 10. Suzuki, Y. and Brown, D. D. (1972), J. Mol. Biol. EE, 409. Svensson, I., Isaksson, L. and Henningsson, A. (1971), Biochim. Biophys. Acta 238, 331. Swan, D., Aviv, H. and Leder, P. (1972), Proc. Nat. Acad. Sci. USA 19, 173. 50 Tsiapalis, C. M., Dorson, J. W., DeSante, D. M. and Bollum, F. J. (1973), Biochem. Biophys. Res. Commun. E2, 737. Urushibara, T., Furuichi, Y., Nishimura, C. and Miura, K. (1975), FEBS Lett. EE, 385. Vesco, C. and Penman, 8. (1968), Biochim. Biophys. Acta 169, 188. Vaughan, M. H., Soeiro, R., Warner, J. R. and Darnell, J. E. (1967), Proc. Nat. Acad. Sci. USA EE, 1527. Wagner, A. F., Bugianesi, R. L. and Shen, T. Y. (1971), Biochem. Biophys. Res. Commun. EE, 184. Wagner, E. K., Penman, S. and Ingram, V. M. (1967), J. Mol. Biol. EE, 371. Wall, R., Phillipson, L. and Darnell, J. E. (1972), Virology E9, 27. Wei, C. and Moss, E. (1975), Proc. Nat. Acad. Sci. USA 1E, 318. Wei, C. M., Gershowitz, A. and Moss, B. (1975), Cell E, 379. Weinberg, R. A. (1973), Ann. Rev. Biochem. EE, 329. weinberg, R. A. and Penman, S. (1970), J. Mol. Biol. EE, 289. Weinberg, R. A. and Penman, S. (1970), J. Mol. Biol. El, 169. Weinmann, R. and Roeder, R. G. (1974), Proc. Nat. Acad. Sci. USA EE, 1790. Wellauer, P. K. and Dawid, I. B. (1973), Proc. Nat. Acad. Sci. USA 19, 2827. Wellauer, P. K., Dawid, I. B., Kelley, D. E. and Perry, R. P. (1974), J. Mol. Biol. EE, 397. Wellauer, P. K. and Dawid, I. B. (1974), J. Mol. Biol. EE, 379. Wieland, T. (1968), Science 159, 946. 51 Williamson, R., Crosley, J. and Humphries, S. (1974), Biochem. EE, 703. Winters, M. A. and Edmonds, M. (1973), J. Biol. Chem. 248, 4756. Yogo, Y. and Wimmer, E. (1972), Proc. Nat. Acad. Sci. USA EE, 1877. Zimmerman, E. F. and Holler, B. W. (1967), J. Mol. Biol. EE, 149. Zimmerman, E. F. (1968), Biochemistry E, 3156. Zylber, E. A. and Penman, S. (1971), Proc. Nat. Acad. Sci. USA EE, 2861. PART I IDENTIFICATION OF METHYLATED NUCLEOSIDES IN MESSENGER RNA FROM NOVIKOFF HEPATOMA CELLS 52 Summary The poly(A) tract found in eucaryotic mRNA has been utilized to study methylation in mRNA obtained from Novikoff hepatoma cells. Methyl labeling of RNA was achieved with L-[methyl-BH]methionine under conditions that suppress radioactive incorporation into the purine ring. Poly(A)(+)RNA was obtained from polysomal RNA by chromatography on Oligo(dT)-cellulose. Sucrose density gradient centrifugation of the poly(A)(+)RNA revealed a pattern expected for mRNA. The composition of the methyl-labeled nucleosides in the poly(A)(+)RNA was analyzed following complete enzymatic degradation to nucleosides. Using DEAE-cellulose (borate) chroma- tography, which separates 2'-Qfmethylnucleosides from normal and base-methylated nucleosides, approximately 50% of the radioactivity was recovered in the 2'-Ef methylnucleoside fraction and 50% in the base-methyl- nucleoside fraction. High speed liquid chromatography (Aminex A-5) of the 2'-Ehmethylnuc1eoside fraction pro- duced four peaks coincident with the four 2'-Qfmethyl- nucleoside standards. Analysis of the base-methyl- nucleoside fraction revealed a unique pattern. While rRNA and tRNA possessed complex base methyl-nucleoside patterns, the distribution in mRNA was quite simple, consisting predominantly of N6-methyladenosine. These results demonstrate a unique distribution of methylated 53 nucleosides in mRNA. By analogy to rRNA synthesis, the presence of methylnucleosides in mRNA may reflect a cellular mechanism for the selective processing of certain mRNA sequences. Introduction The important role of post-transcriptional RNA modification in the synthesis of eucaryotic RNA has become increasingly apparent. One of the most striking features of rRNA synthesis is the specific methylation that occurs on the 45 S precursor RNA molecule (1). Although the maturation process is nonconservative (about 50% of the original molecule is lost through degradation), the conserved regions present in the final 28 S and 18 8 products retain all of the methylated sequences (2). Methylnucleosides, containing both base- methyl and 2'-Qfmethylnuc1eosides, account for 1.5% of the total nucleosides in the mature RNA (3). Methylation of 45 8 RNA does not occur in the absence of methionine and mature ribosomal 28 S and 18 8 RNA is not formed (4). Methylation is thus required for proper rRNA processing. More recently, mRNA of eucaryotic cells has been found to undergo an unusual form of post-transcriptional modification involving the addition of a sequence of about 200 adenosine residues to the 3'-end (5). Studies on its function are still inconclusive, but several findings can be cited outlining its importance (5, 6, 7). 54 Like methylation in rRNA, adenylation seems to play an essential role in the metabolism of mRNA molecules. Since post-transcriptional events may represent an important control mechanism in the regulation of genetic expression, we have investigated the possibility of methylation as an additional post-transcriptional modification of mRNA. Earlier work with bacterial and phage mRNA produced strong evidence for the essential absence of methylation in these systems, being no higher than one per 3500 nucleotides (8). Other early studies with mammalian HnRNA indicated that methylation was either nonexistent or very low (9, 10). The discovery of poly(A) has now made it possible to obtain pure mRNA fractions through affinity chromatography and therefore to search for low levels of methylation in mRNA without interference from rRNA contamination. Recently Perry and Kelley reported the existence of methylation in mouse L cell mRNA at about one-sixth the level found in rRNA, and both base and ribose methylations were found (11). This paper reports the existence of methylated nucleosides in the mRNA of Novikoff hepatoma cells and identifies the unique distribution of methylated moieties. A preliminary report of these results has appeared elsewhere (12). 55 Methods Cell Culture and Labeling The N181 strain of Novikoff hepatoma cells was grown in culture in Swimm's 8-77 media supplemented with 4 mM glutamine and 10% dialyzed calf serum (13). The cells were grown in an atmosphere of 5% CO2 in air in sealed screw cap Erlenmeyer flasks with a doubling time of approximately 12 hr at 37°. For labeling with L-[methyl-BHJmethionine, cells in mid-log were pelleted aseptically and resuspended in fresh warm media containing 0.02 mM methionine (one- fifth the normal concentration) at a cell concentration of approximately 1.5 x 106/ml. Labeling was performed for three h in the presence of 20 mM sodium formate and 20 uM each of adenosine and guanosine; these conditions have been shown to effectively suppress nonmethyl purine ring labeling via the H4-folate pathway (14). Cell Fractionation and RNA Preparation Cells were poured over frozen crushed saline solution and harvested by centrifugation. The cells were washed once and disrupted by Dounce homogenization in RSB (0.01 M Tris-HCl, pH 7.4, 0.01 M NaCl, 0.0015 M MgClZ). Nuclei were removed by centrifugation at 800g for 2 min and mitochondria at 10,000g for 7 min. The cytoplasmic supernatant was centrifuged through a 56 10.5 ml 15-45% sucrose gradient in RSB for the isolation of polysomes. The polysome fraction was made 0.1 M NaCl, 0.01 M EDTA and 0.5% SDS, 0.25 mg carrier rRNA was added per ml of solution and the RNA collected by ethanol pre- cipitation. After centrifugation, the RNA pellet was taken up in SDS buffer (0.01 M TriS°HC1, pH 7.4, 0.1 M NaCl, 0.01 M EDTA and 0.5% SDS) and incubated at room temper- ature for 10 minutes following the addition of a small crystal of proteinase K (m 0.3 mg per 108 cells) (15). From this point, the RNA extraction procedure followed the technique described by Singer and Penman (16). The final aqueous phase was removed and the RNA precipitated with 2 volumes absolute ethanol. Oligo(dT)-Cellulose Chromatography The isolation of poly(A)(-) and poly(A)(+)RNA was achieved using Oligo(dT)-cellulose chromatography. The procedure used was essentially that described by Aviv and Leder (17) except that nonadsorbed material was eluted by continued washing with 0.12 M NaCl, 0.01 M Tris-HCl, pH 7.4, 0.001 M EDTA and 0.2% SDS. The material retained was then eluted by the same buffer lacking NaCl. Carrier rRNA was used for ethanol pre- cipitation and the poly(A)(+)RNA was further purified by a second passage through oligo(dT)-ce11ulose. Over 57 90% of the radioactivity in the second application was retained by the column. PoEy(A)(-)RNA Fractionation The poly(A)(-)RNA was fractionated into a 4 8 fraction and a 28 S + 18 8 fraction by sedimentation through sucrose gradients. Further analysis of the 4 8 RNA fraction required deacylation of 3H-peptidyl and 3H-methionyl tRNA by incubation in 1.8 M Tris-HCl, pH 8.1 at 37° for 100 min. Enzymatic Degradation of RNA and Resolution of Nucleosides The detailed method for degradation of RNA to nucleosides and the subsequent quantitative analysis of methylnucleosides is described in another publication (18). In brief, RNA samples were completely hydrolyzed to the nucleoside level by simultaneous treatment with alkaline phosphatase, pancreatic ribonuclease A and phospho- diesterase I. Completeness of the reaction was monitored by paper electrophoresis. DEAE-cellulose in the borate form was used to chromatographically separate those nucleosides blocked at the 2'-prosition with a methyl group, from the remainder of the normal and base-substituted nucleosides. The 2'-Qfmethylnucleosides are not retained and elute with 0.15 M boric acid just after the void volume. The base-methyl radioactivity is recovered with the normal 58 ribonucleosides by elution with 0.7 M boric acid. Each nucleoside fraction was taken to dryness by flash evaporation and the boric acid removed as its methyl ester by successive flash evaporations with absolute methanol (18). The 2'-Qfmethylnucleosides present in the first DEAE-cellulose (borate) fraction were separated by high speed liquid chromatography (18). The base-methyl- nucleoside distribution was determined using the same resin but elution was with the buffer described for normal nucleoside resolution (18). In this case, the UV absorbance peaks corresponded to the four normal ribonucleosides plus any base-methyl standards added for reference. Results Prgparation of RNA Fractions Prior to the characterization of methyl-labeled components in mRNA, it is necessary to carefully document that the methylated RNA being examined is mRNA and is not contaminated with other cellular RNA species. The sucrose density gradient sedimentation profiles of the methyl- labeled poly(A)(-) and poly(A)(+)RNA are Shown in Figure 1. The purification of the poly(A)(+)RNA (mRNA) involved repeated binding of the RNA sample to Oligo(dT)-cellulose. Figure 1B shows the characteristic heterogeneous sedi- mentation profile of mRNA, with the peak of radioactivity 59 Figure l. Sucrose density gradient sedimentation of methyl-labeled polysomal RNA. (A) An aliquot of poly(A)(-)RNA obtained from Oligo(dT)-cellulose was layered over a 4.8 ml 5-20% sucrose gradient (0.01 M Tris, pH 7.4 and 0.005 M EDTA) and centrifuged at 4° for 160 min at 45,000 RPM in a SW 50.1 rotor. The fractions containing 28 S + 18 8 RNA indicated by the horizontal bar were pooled for ethanol precipitation. The fractions containing 4 8 RNA were also pooled. (B) An aliquot of poly(A)(+)RNA was sedimented as described above. (C) DMSO-sucrose gradient sedimentation of poly(A)(+)RNA. 5 ul of poly(A)(+)RNA was mixed with 100 pl of 99% DMSO (10 mM LiCl, 1 mM EDTA) and heated to 60° for two minutes. All of this material was layered over 4.8 ml of a 5-20% sucrose gradient in 99% DMSO (10 mM LiCl, 1mM EDTA). Centrifugation was for 15 h at 27° and 45,000 RPM in a SW 50.1 rotor. couu'rsmm no") 60 A POLY A P) % IIS’IIS | I 43 ) ”l 0 “l O ' r I .0770“ IO 20 30 TOP 8 POLY A (4) 205 :03 as I 4 4 Id 05* O 1 ' j ' nono- '° 20 so ,0, C POLY Mo) (OM80) 4( us no: 43 I 1 I :1 O r - narrow IO 20 IOP F RACTION NUMBER 61 sedimenting slightly faster than 18 S. The mRNA pro- file obtained using L-[methyl-BHlmethionine label is virtually identical to that obtained when 3H-uridine is used for the label (data not shown) and is very similar to uridine labeled mRNA profiles obtained for HeLa and L cells by others (11, 16). The 3H-methyl radioactivity recovered in mRNA generally represented 3.5-4.0% of the radioactivity recovered in 28 S + 18 8 RNA in the three h labeling used here. This level is considerably lower than that obtained when a three h uridine label is employed. This suggests mRNA methyl- ation is considerably less than rRNA methylation (as % of total nucleotides), in agreement with the more quantitative determinations of Perry and Kelley (11). To insure that the radioactivity)seen in mRNA is not due to adventitious binding of smaller, methionine- labeled RNA fragments, the labeled mRNA was analyzed on denaturing DMSO-sucrose gradients (19). As shown in Figure 1C, most of the RNA remained larger than 18 S. The region of 28 S + 18 S rRNA and of 4 8 RNA indicated by the brackets in Figure 1A were pooled for further methylnucleoside analysis to compare with the mRNA. After pooling the 4 8 region, the tRNA was deacylated as described in Materials and Methods prior to further analysis. 62 Separation of Base-Methylnucleosides from 2'-0-Methy1nucleosides To determine the proportion of methyl groups attached to the base moiety as opposed to those attached to the ribose moiety, each RNA fraction was completely degraded enzymatically to the nucleoside level; the nucleosides were then separated by DEAF-cellulose (borate) chromatography into a 2'-Qfmethylnucleoside fraction and a base-methylnucleoside fraction. Table 1_ shows the recoveries in the 2'-Qfmethylnucleoside and the base~methylnucleoside fractions as a percentage of the radioactivity present in the original RNA sample. While rRNA methylation occurs predominantly on the ribose moiety, and 4 8 RNA methylation occurs pre- dominantly on the bases, the mRNA is intermediate with approximately 50% being 2'-Qfmethy1ation. Note that essentially complete recovery of radioactivity was obtained for the labeled mRNA and rRNA. Only about 80% of the 4 8 RNA radioactivity could be recovered, how- ever, from the DEAE-cellulose (borate) column. This incomplete recovery, was observed only with 4 8 RNA, and does not interfere with the further characterization of the mRNA since complete recovery was obtained with this fraction. 63 Table 1 Recovery of Labeled Methylnucleosides from DEAR-cellulose (borate)a % of Total Radioactivity RNA Species Base- 2'-E- Methylnucleosides Methylnucleosides 28 S + 18 S 11 89 4 S 55 25 mRNA 50 50 aEach RNA fraction containing 1.2 mg carrier RNA was hydrolyzed at 37° and pH 9.0 for 34 h with 1 unit alkaline phosphatase, 10 pg ribonuclease A and 0.22 units phosphodiesterase I per mg RNA. The nucleosides produced were then resolved into a 2'-Qfmethylnucleoside fraction and a base-methylnucleoside fraction on a 6 mm x 120 mm DEAF-cellulose (borate) column. 64 2'-0-Methylnucleoside Distribution The ribose methylated nucleoside fraction for each of the RNA classes studied was further analyzed by high Speed liquid chromatography (Fig. 2). The separation of 2'-Ermethylnuc1eosides derived from mRNA (Fig. 2c) revealed the presence of all four 2'-thethylnucleosides also found in rRNA (Fig. 2a). The tRNA 2'-E—methy1- nucleoside distribution depicted in Figure 2b contains two additional peaks. These additional peaks represent 2'-Efmethylnucleosides that contain some other modifi- cation. This radioactivity present in each particular 2'-Qfmethylnucleoside peak may be used to determine the percentage of each labeled 2'-Qemethylnucleoside in each class of RNA. As Shown in Table 2, the 2'-Qfmethyl- nucleoside distribution of mRNA differed considerably from that found for 4 8 RNA. The 2'-Qfmethylnucleoside distribution of mRNA was quite similar, however, to that found for rRNA. This similarity in distribution contrasts with the different sizes of oligonucleotides produced by alkaline digestion. Alkaline digestion of rRNA produces primarily mononucleotides and small amounts of dinucleo- tides, due to the alkaline stability of phosphodiester bonds adjacent to 2'-Qfmethyl groups. In studies not Shown, reversed phase chromatography was used to compare 65 Figure 2. High Speed liquid chromatography of 2'-Qfmethylnucleoside fraction. 20 pl of each 2'-E- methylnucleoside fraction dissolved in 0.4 ammonium formate in 40% ethylene glycol adjusted to pH 4.25 with formic acid was injected onto a high speed liquid chromatography column. The column was developed at 31.5° and 2500 psi. The flow rate was 0.4 ml per min and 0.5 ml fractions were collected. (a) 28 S + 18 8 RNA, (b) 4 S am. (e) mRNA. -2 COUNTS/MIN (I IO ) 20') 66 4+ __—:.¢-f M ‘3'- 5 (o) neso 203 mu ‘ H U. 4 n c- ) ‘ t T -N 26 4'0 so so (0) 48 RNA (c) mRNA vv v v'v " I T so so FRACTION NUMBER 67 Table 2 Radioactive 2'-E-Methylnucleoside Compositiona % of Total Radioactivity 28 S + 18 8 RNA mRNA 4 S RNA Um 18.0 21.9 40.0 Am 33.4 35.0 14.0 Gm 26.8 23.2 19.2 Cm 21.5 19.9 11.9 Y 0 0 8.9 Z 0 0 6.1 aThe radioactivity present in each of the peaks shown in Figure 2 is presented as percentage of the total radioactivity recovered from high speed liquid chroma- tography. 68 the Sizes of alkaline stable oligonucleotides produced by NaOH digestion of methyl-labeled mRNA and rRNA. While almost all the alkaline-stable radioactivity of rRNA eluted in the dinucleotide region, a large portion of the mRNA alkaline-stable radioactivity eluted con- siderably past the dinucleotide region. This indicates that some of the methyl groups occur as adjacent 2'-Er methylnucleotides or in some Species containing multiple phOSphates. Base-methylnucleoside Distribution The base-methylnucleoside distribution in mRNA is distinctly different from either 4 8 RNA or rRNA. The separation of 28 S + 18 8 RNA base-methylnucleosides by Aminex A-S high speed liquid chromatography is Shown in Figure 3a. It is significant that essentially no radioactivity appears coincident with the unmodified nucleosides guanosine and adenosine (Fig. 3a, markers F and G, respectively). This indicates that nonmethyl purine labeling has indeed been effectively suppressed and all the radioactivity seen in RNA is in fact due to methylation. The primary peaks of 28 S and 18 8 base- methylnucleosides correspond to N6-methyladenosine, N6-dimethy1adenosine, 5-methy1cytidine and material that migrates near the solvent front (primarily modified uridine). This corresponds well with observations made by others with HeLa cell RNA using acid hydrolysis 69 Figure 3. High speed liquid chromatography of base-methylnucleoside fraction. 250 pl of each base- methylnucleoside fraction dissolved in H20 was injected onto a high Speed liquid chromatography column. The column was developed at 2500 psi at 31.5°. The flow rate was 0.8 ml per min and 0.67 ml fractions were collected. At fraction 35, the volume collected per fraction was changed to 1.35 ml. The letters correspond to the location of the following standards: A, 3-methyl- uridine, thymine riboside and uridine; B, l-methylinosine; C, l-methylguanosine; D, NZ-dimethylguanosine; E, N2- methylguanosine; F, guanosine; G, adenosine and N4- methylcytidine; H, N6-methy1adenosine; I, S-methylcyti- dine; J, NG-dimethyladenosine; K, 1-methyladenosine; L, 7-methylguanosine. (a) 28 S + 18 8 RNA, (b) 4 8 RNA, (c) mRNA. 70 (O) .8 0 205 RNA LI. ' .w KI. L .w J II. "I. To 6|. 2 Fl. Al. 43 RNA (0) II. "I. El. DI. all. A4 fl 1 m m. Ave. : gmh‘aou z'o (c) mRNA ‘1 8% 0—4 FRACTION "UNDER 71 techniques except for the absence of l-methyladenosine and 7-methylguanosine (20, 21). l-Methyladenosine is relatively unstable at neutral and alkaline pH (22) and is readily converted, under the incubation conditions used here to generate nucleosides, to N6-methyladenosine. Thus the peak of N6-methyladenosine radioactivity contains an unknown contribution from 1-methyladenosine. 7-Methy1- guanosine is also relatively unstable (22) and is con- verted under our hydrolysis conditions to a product which migrates near the solvent front on high speed liquid chromatography. Also in agreement with studies using HeLa cells (20, 21), the 4 8 RNA does not contain much NG-dimethyl- adenosine but does contain the unique nucleosides 1-methylguanosine, NZ-methylguanosine and Nz-dimethyl- guanosine (Fig. 3b). Peaks of radioactivity also occur coincident with S-methylcytidine, N6-methyladenosine and with the methylated uridines. Klagsbrun has reported no N6-methyladenosine for 4 8 RNA in HeLa cells (20), so in this case the peak at NG-methyladenosine may be due entirely to l-methyladenosine. The base methyl distribution for mRNA is strik- ingly simple (Fig. 3c). Approximately 80% of the radioactivity e1utes.with N6-methyladenosine indicating tfllat base methylation of mRNA is primarily N6-methyl- adenosine and/or 1-methyladenosine. Small peaks of 72 radioactivity also occur near the solvent front (W 15%) and coincident with N6-dimethyladenosine (m 5%). Discussion Several lines of evidence presented in this paper demonstrate that the radioactivity recovered in the mRNA fraction is indeed due to methylation of mRNA. The RNA has been passed twice through Oligo(dT)-cellulose columns and is apparently free of rRNA and tRNA contamination. The incorporated radioactivity is only in methylated Species and the amount of radioactivity found in normal adenosine and guanosine is below detectable levels. The distribution of radioactive methyl groups between base and ribose moieties and the base-methylnucleoside dis- tribution are distinctly different for mRNA when compared to rRNA or tRNA labeled under identical conditions. Furthermore, the pattern of alkaline stable oligonucleo- tides is different in mRNA. Thus, not only does mRNA contain methylated Species, but the distribution of the methylated species is unique. The distribution of radioactivity between base and ribose groups for rRNA determined here agrees well with values obtained for HeLa cell RNA by others using alkaline digestion (23, 24). Using acid hydrolysis techniques, the percentage of ribose methylation observed by others is considerably less (40-50%) (20, 21). It is most likely that acid treatment resulted in 73 this lower estimate of 2'-Efmethylation due to release of 2'-Efmethyl groups as methanol (25, 26). Complete recovery of the rRNA radioactivity applied to the DEAE- cellulose (borate) column strengthens the value for ribose methylation reported here (89%). Since 100% of the mRNA radioactivity was recovered, we probably have observed all the methylated mRNA nucleosides labeled under these conditions. Although the role of 2'-Qfmethylation in rRNA processing is not clearly understood, its presence does seem necessary for the proper processing of the large 45 8 precursor (4). 2'-E:Methy1 groups are known to alter the secondary structure of RNA (27) and thus may be involved in conferring the proper structure for the precise cleavages that occur. The presence of 2'-Qf methylnucleotides stabilizes synthetic RNA molecules against hydrolysis by a 3'-OH-exoribonuclease thought to be involved with rRNA processing (28). As suggested by these £2.21EEQ studies, a 2'-Efmethylnucleotide con- ceivably could function as a stop signal for a pro- cessing exonuclease. A 2'-Qfmethylnucleotide conceivably could also function as a recognition site for an endo- nuclease. Similar mechanisms may be involved in mRNA precursor processing, if indeed methylation occurs at the precursor stage. The processing of mRNA found in eucaryotes has the additional interesting feature of being a potentially 74 important control point in gene expression. If such control does in fact occur, as some hybridization studies indicate (29), we are faced with the problem of identify- ing the means by which certain messenger sequences are selected. The addition of poly(A) may not be sufficient to define which mRNA sequences are to be transported Since all poly(A) synthesized in the nucleus does not exit to the cytoplasm as part of functional messenger RNA (30). Methylation could function in such a selection process. The labeled mRNA observed here most probably represents many different species of messenger. It is interesting, then, that the base methylation Should occur primarily with adenine. Calculations based on the level of methylation determined by Perry and Kelley (11) and the distribution of methylation described in these studies, as well as the average size of mRNA, indicate that there is sufficient methylated adenosine for each messenger molecule to contain this modified nucleoside at least once. The presence of this base-methylnucleoside in all mRNA molecules would suggest some essential function for the modification. The existence of a poly(A) tract in mRNA molecules raises the possibility that the base modification occurs in this segment. This is presently being investigated. ACKNOWLEDGMENTS The authors thank Dr. Jay Goodman for his assistance in setting up the cell culture and Dr. J. J. Fox of the Sloan Kettering Institute for a generous gift of N4-methylcytidine. REFERENCES Greenberg, H. and Penman, S. (1966) J. Mol. Biol. EE, 527—535. Wagner, E. K., Penman, S. and Ingram, V. (1967) Jo M010 BiOlo 22" 371-387. Brown, G. M. and Attardi, G. (1965) Biochem. Biophys. Vaughan, M. H., Soerio, R., Warner, J. R. and Darnell, J. E. (1967) Proc. Nat. Acad. Sci. USA EE, 1527- 1534. Jelinek, W., Adesnik, M., Salditt, M., Sheiness, D., Wall, R., Molloy, G., Phillipson, L. and Darnell, J. E. (1973) J. Mol. Biol. EE, 515-532. Greenberg, J. R. and Perry, R. P. (1972) J. Mol. BiOlo 12., 91-980 Nakazato, H., Edmonds, M. and Kopp, D. W. (1974) Proc. Nat. Acad. Sci. USA 1E, 200-204. Moore, P. B. (1966) J. Mol. Biol. EE, 38-47. Perry, R. P., Cheng, T. Y., Freed, J. J., Greenberg, J. R., Kelley, D. E. and Tantof, K. D. (1970) Proc. Nat. Acad. Sci. USA EE, 609-616. 75 lo. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 76 Murumatsu, M. and Fujisawa, T. (1968) Biochim. Biophys. Acta 157, 476-492. Perry, R. P. and Kelley, D. E. (1974) Cell E, 37-42. Desrosiers, R., Friderici, K. and Rottman, F. (1974) Federation Proceedings Abstracts EE, 1432. Morse, P. A., Jr. and Potter, V. R. (1965) Cancer Research EE, 499-508. Maden, B. E. H., Salim, M. and Summers, D. F. (1972) Nature New Biology 237, 5-9. Wiegers, V. and Hilz, H. (1972) FEBS Letters EE, 77-82. Singer, R. H. and Penman, S. (1972) Nature 240, 100-102. Aviv, H. and Leder, P. (1972) Proc. Nat. Acad. Sci. USA EE, 1408-1412. Pike, L. M. and Rottman, F. (1974) Anal. Biochem. EE, 367. Lindberg, U. and Persson, T. (1972) Eur. J. Biochem. EE, 246-254. Klagsbrun, M. (1973) J. Biol. Chem. 248, 2612-2620. Iwanami, Y. and Brown, G. M. (1968) Arch. Biochem. Biophys. 126, 8-15. Hall, Ross H. (1971) The Modified Nucleosides in Nucleic Acids, Columbia University Press. Lane, B. G., and Tamaoki, T. (1969) Biochim. Biophys. Salim, M. and Maden, B. E. H. (1973) Nature 244, 334-336. Baskin, F. and Dekker, C. A. (1967) J. Biol. Chem. 242, 5447-5449. Abbate, J. and Rottman, F. (1972) Anal. Biochem. 51, 378-388. Rottman, F., Friderici, K., Comstock, P. and Khan, K. (1974) Biochemistry EE, 2762—2771. 28. 29. 3o. 77 Stuart, 8. E. and Rottman, F. (1973) Biochem. Biophys. Res. Commun. EE, 1001-1008. Shearer, R. W. and Smuckler, R. W. (1972) Cancer Research EE, 339-342. Perry, R. P., Kelley, D. E. and LaTorre, J. (1974) J. Mol. Biol. EE, 315-330. PART II CHARACTERIZATION OF NOVIKOFF HEPATOMA mRNA METHYLATION AND HETEROGENEITY IN THE METHYLATED 5'-TERMINUS 78 Summary KOH digestion of methyl-labeled poly(A)+mRNA purified by (dT)-ce11ulose chromatography produced mono- nucleotide and multiple peaks of large oligonucleotides (-6 to -8 charge) when separated on the basis of charge by Pellionex-WAX high speed liquid chromatography in 7 M urea. Heat denaturation of the RNA before application to (dT)-ce11ulose was required to release contaminants (mostly 18 S rRNA) that persisted even after repeated binding to (dT)-cellulose at room temperature. Analysis of the purified poly(A)+mRNA by enzyme digestion, acid hydrolysis and a variety of chromatographic techniques has Shown that the mononucleotide (53%) is due entirely to N6-methy1adenosine. The large oligonucleotides (47%) were found to contain 7-methylguanosine and the 2'-Ef methyl derivatives of all 4 nucleosides. No radioactivity was found associated with the poly(A) segment. Periodate oxidation of the mRNA followed by B-elimination released only labeled 7-methylguanosine consistent with a blocked 5'-terminus containing an unusual 5'-5' bond. Alkaline phosphatase treatment of intact mRNA had no effect on the migration of the KOH produced oligonucleotides on Pel- lionex-WAX. When RNA from which 7-methylguanosine was removed by B-elimination was used for the phosphatase treatment, distinct dinucleotides (Nmpr) and trinucleo- tides (Nmmepr) occurred after KOH hydrolysis and 79 Pellionex-WAX chromatography. Thus Novikoff hepatoma poly(A) + mRNA molecules can contain either one or two 2'-Qfmethylnucleotides linked by a 5'-5' bond to a terminal 7-methylguanosine and the 2'-Efmethylation can occur with any of the four nucleotides. The 5' terminus may be represented by m7G5'ppp5'(Nmp)l or 2Np, a general structure proposed earlier as a possible 5'- terminus for all eucaryotic mRNA molecules (Rottman, F., Shatkin, A. & Perry, R. [1974], Cell E, 197). The compo- sition analyses indicate that there are 3.0 N6 -methyl- adenosine residues, 1.0 7-methylguanosine residue and 1.7 2'-Qfmethylnucleoside residues per average mRNA molecule. Introduction Only in recent years has considerable understand- ing of the structure and composition of eucaryotic mRNA been obtained. The existence of a poly(A) tract not only accounts for a portion of the large untranslated region but also has greatly facilitated the isolation of pure mRNA from cells. More recently, poly(A)+mRNA has been found to contain a low level of methylated nucleotides (Perry & Kelley, 1974) and the distribution of methylation is distinct from rRNA and tRNA (Desrosiers et al., 1974). The finding of a Simple distribution of base-methylnucleosides in cellular mRNA is interesting since many different species of messenger coding for 80 many proteins were actually present in the analyzed material (Desrosiers et al., 1974). The ability to synthesize high specific activity methylated viral mRNA i2.YlE£2.uSin9 virion associated RNA polymerase has facilitated the characterization of the viral mRNA methylated nucleotides. In these EE gEggg systems, the methylation occurs exclusively as a large alkaline stable oligonucleotide with a charge estimated to be -5 to -6 by DEAE chromatography in 7 M urea (Urushibara et al., 1975; Wei & Moss, 1975; Furuichi et al., 1975; Rhodes et al., 1974; Furuichi, 1974). Furuichi et a1. (1975) have analyzed this large oligo- nucleotide of reovirus mRNA synthesized $2.!iEEQ and found that it contains 7-methylguanosine and 2'-Qf methylguanosine in approximately equal proportions. After labeling with [8,7-32PJGTP, the large oligo- nucleotide was found to be blocked with respect to removal of 32 P by alkaline phosphatase. Periodate oxi- dation followed by reduction with [3H]borohydride labeled the 7-methylguanosine of the large oligonucleotide, con- sistent with the structure originally postulated by Rottman et a1. (1974) to be present on the 5' end of eukaryotic mRNA. Furuichi et a1. (1975) proposed the structure m7G5'ppp 'Gmpr for the 5'-terminus of reovirus mRNA. Wei and Moss (1975) have analyzed the vaccinia virus mRNA made EE_vitro and have concluded 81 the 5'-terminus consists of a similar structure, only the 2'-Efmethylnucleoside is a mixture of Gm and Am. I I 2'2'7 5 pp5 AmpUmp has been An analogous structure m3 G reported for the 5'-terminus of certain low molecular weight nuclear RNAS (Ro-Choi et al., 1974; Ro-Choi et al., 1975). Quantitative analysis of the methylated compo- nents of cellular mRNA has progressed less rapidly. This is probably due at least in part to the greater complexity of methylation. About 50% has been found in NG-methyladenosine that runs as a mononucleotide after KOH hydrolysis (Desrosiers et al., 1974; Perry et al., 1975). This mononucleotide has not yet been observed with viral mRNAs synthesized $2.!iEEQR Its presence has recently been reported in SV40 Specific mRNA of infected cells (Lavi & Shatkin, 1975) but Abraham et a1. (1975) have reported its total absence in VSV specific mRNA of infected cells. A dinucleotide peak has also been observed in the Poly(A)+mRNA fractions of L cells (Perry et al., 1975) and HeLa cells (Wei et al., 1975) but Perry et a1. (1975) sus- pected this may have been due to rRNA contamination, since the dinucleotide peak was not observed in cells that were labeled with a dose of Actinomycin D that suppresses rRNA synthesis. AS with viral mRNA made E3 vitro, a large oligonucleotide was also observed 82 in these cellular mRNAs. Perry et al. (1975) have presented evidence that this oligonucleotide is derived from a blocked 5'-terminus, Since it was resistant to phosphorylation by polynucleotide kinase even after phOSphatase treatment and phosphatase treatment only reduced its charge by two units. Wei et a1. (1975) have found that periodate oxidation and B-elimination removes the 7-methylguanosine from the large oligonucleo- tide and have concluded that a 5'-5' bond similar to the viral 5'-terminus is also present. In the characterization of methyl-labeled Novikoff hepatoma poly(A) + mRNA, we have found radioactive con- tamination with other species that cannot be eliminated by standard room temperature (dT)-ce11ulose procedures. We present evidence that the only methylated constituents of the poly(A) + mRNA are the base-modified mononucleotide and the heterogeneous large oligonucleotide produced by alkaline digestion. Extensive compositional and structural analyses are shown including depurination by acid hydroly- sis under conditions that preserve 7-methylguanine structure and do not allow conversion of l-methyladenine to NG-methyladenine, and periodate oxidation followed by B-elimination of 7-methylguanosine from the blocked 5'-terminus. 83 Materials and Methods Materials Cell culture materials were purchased from Grand Island Biologicals Co. L-[methyl-3H1methionine (5-8 Ci/mmole) was obtained from Amersham Searle. Chroma- tographically pure proteinase K was purchased from EM laboratories (Elmsford, N.Y.). Methylated bases and nucleosides used as standards in chromatography were obtained from Sigma Chemical Co. and Calbiochem. Oligo(dT)-cellulose was prepared as described by Gilham (1964). Sigma urea was passed over a mixed bed ion exchange resin before use. AL-Pellionex-WAX was pur- 14C poly(A) was prepared chased from Reeve Angel. using polynucleotide phoSphorylase as described (Rottman & Heinlein, 1968). Cell Culture and Preparation of MetherLabeled RNA Novikoff hepatoma cells (N181 strain) were main- tained in Swimm's 8-77 medium with 10% calf serum essen- tially as described (Desrosiers et al., 1974). Cells in late log phase were harvested and resuspended at approximately 106 cells per ml in fresh warm medium containing 10% calf serum and the normal methionine concentration (0.1 mM). Labeling (generally 5 mCi per 200 ml of cells) was performed with L-[methy1-3H1methio- nine for 13 hours in the presence of 20 mM sodium formate 84 and 40 pM of adenosine and guanosine to suppress purine ring labeling. No radioactivity was detected in adeno- sine or guanosine. At the completion of labeling, cells were pelleted and washed once with a balanced salt solution. After swelling for 8 minutes in hypotonic buffer (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 1.5 mM MgClz), the cells were disrupted by Dounce homogenization. Nuclei were removed by centrifugation at 800 g for 2 min and mito- chondria at 10,000 g for 7 minutes. The supernatant was made 0.1 M NaCl, 0.01 M EDTA, and 0.5% SDS and the RNA was isolated with a proteinase K digestion and phenol- chloroform extraction as previously described (Desrosiers et al., 1974; Singer & Penman, 1973). Oligo(dT)-cellulose Chromatogrgphy RNA containing a poly(A) segment was isolated by two passes through (dT)-cellulose essentially as described (Desrosiers et al., 1974). RNA was applied to the column in 0.12 M NaCl, 0.01 M Tris-HCl (pH 7.4), 1 mM EDTA and 0.2% SDS at room temperature and the RNA lacking a poly(A) segment was eluted by continued washing with this buffer. The material retained by the column was then collected by elution with the same buffer lacking NaCl. The material eluted by buffer without NaCl in the first pass was chromatographed a second time through 85 the column. Approximately 0.5% of the original radio- activity present in total cytOplasmic RNA was recovered in the final poly(A)+mRNA. For studies on the effect of heat denaturation on retention of RNA on (dT)-ce11ulose, the poly(A)+RNA isolated by two passes through (dT)-ce11ulose was dissolved in 0.2% SDS, heated to 60° for 2 minutes and chilled rapidly in an ice bath. An equal volume of 0.24 M NaCl, 0.02 M Tris-HC1 (pH 7.4) and 2 mM EDTA was added and the RNA was rechromatographed on (dT)- cellulose at room temperature as above. KOHEydrolySis and Pellionex- WAX Chromatograpgy Alkaline hydrolysis was performed in a volume of 0.5 ml 0.4 N KOH overnight at 37°. The solution was neutralized with perchloric acid in the cold and the insoluble KClO4 removed by centrifugation. The super- natant was lyophilized prior to chromatographic analyses. The nucleotides produced by alkaline hydrolysis were separated on the basis of net negative charge by high Speed liquid chromatography in 7 M urea on Pellionex- WAX, a weak anion exchanger. This system was found to separate principally on the basis of charge since purine and pyrimidine oligonucleotide markers eluted identically and the 16 dinucleotides produced by KOH digestion of methyl-labeled rRNA eluted in a narrow band with -3 charge. 86 The alkaline hydrolyzate from above was dissolved in 0.005 M sodium phosphate, 7 M urea at pH 7.8 along with oligonucleotide markers and injected onto the Pellionex- WAX column. A 100 ml linear gradient of 0 to 0.2 M (NH4)2804 in 0.005 M sodium phosphate, 7 M urea at pH 7.8 was used to develop the column. The eluate was passed through a Gilford UV monitor for location of the markers and collected either in test tubes or directly in scintillation vials. The mononucleotide (-2) and dinucleotide (-3) were located from the UV absorbance of the hydrolyzed RNA while oligouridylate markers were used for the location of more highly charged Species. When particular fractions were to be recovered from Pellionex-WAX and further analyzed, the nucleotide was bound to charcoal to remove salt and urea and the nucleotide eluted with 10% pyridine in water (Dlugajczyk & Eiler, 1966). The mononucleotide was dephosphorylated by treatment with alkaline phosphatase for 16 hours at 37°. The dinucleotide and large oligonucleotide were degraded to nucleosides by treatment with pancreatic ribonuclease, snake venom phosphodiesterase, and alkaline phosphatase (Pike & Rottman, 1974). Enzyme and Acid Hydrolysis and Analysis of the Products Techniques for the enzymatic degradation of RNA to nucleosides as well as separation into 2'-Efmethyl- nucleoside and base-methylnucleoside fractions by 87 DEAE-borate chromatography and analysis by Aminex high Speed liquid chromatography have been completely described (Pike & Rottman, 1974; Desrosiers et al., 1974). The inclusion of pancreatic ribonuclease in the digestion mixture has been found to be unnecessary and was frequently left out in these studies. Acid hydrolysis was performed with 88% formic acid at 100° for 2 hours in sealed Pyrex Tubes as described (Munns et al., 1974). These conditions were chosen since complete depurination without ring Opening of 7-methy1guanine or conversion of l-methyladenine to NG-methyladenine has been reported to occur (Munns et al., 1974). The hydrolyzates were dried under a stream of N2 and used for analysis by Aminex A-5 high Speed liquid chromatography. The column was developed with 0.4 M ammonium formate pH 5.6. Unhydrolyzed material (from 2'-Qfmethylnucleotides) eluted in the solvent front and excellent resolution was obtained among the purine bases studied. The radioactivity present as 2'-Qf methylnucleotides was determined by adding the amount present in the solvent front in Aminex chromatography to the radioactivity lost by conversion of the 2'-Qf methyl to methanol (Baskin & Dekker, 1967). Isolation of the Poly(A) Segment 3H RNA was mixed with a known amount of 14C poly(A) and digested with 5 pg/ml pancreatic ribonuclease ,88 and 6 units/ml T1 ribonuclease in 0.3 M NaCl, 0.01 M Tris-HCl pH 7.4 and 0.01 M EDTA for 60 minutes at 37°. Upon completion, SDS was added to 0.5% and a brief proteinase K digestion was performed to remove nuclease activity. The poly(A) segment was isolated by phenol- chloroform extraction, ethanol precipitation and chroma- tography on (dT)-cellulose. The yield of nuclease 3H radioactivity was corrected to obtain the 14 resistant amount in poly(A) based on the recovery of C poly(A). l4C poly(A) was usually 50-60%. Poly(A) The recovery of isolated in this manner from 12 hour adenosine-labeled mRNA has a size of 140 to 200 nucleotides estimated by 10% polyacrylamide gel electrophoresis. Periodate oxidation and B-elimination Periodate oxidation and B-elimination were per- formed as described by Fraenkel-Conrat and Steinschneider (1967). After cleavage of terminal residues that con- tained free 2' and 3' hydroxyls, the RNA was precipi- tated with ethanol. The alcohol supernatant was dried and the residue acid hydrolyzed as described above to identify the Species released by B-elimination. Half of the RNA precipitate was used for direct KOH hydrolysis and the other half was treated with alkaline phosphatase (0.15 mg/ml) for 2 hr. at 45° C in 0.05 M ammonium for- mate, 0.002 M MgClz, pH 9 prior to KOH hydrolysis and Pellionex-WAX chromatography. 89 Results The methyl-labeled Novikoff mRNA purified by repeated binding to (dT)-cellulose appeared pure by aqueous sucrose gradient centrifugation and further (dT)-cellulose chromatography. Other findings suggested to us that in fact it was contaminated with other RNA species. Aqueous sucrose gradient centrifugation in 10 mM TriS°HCl pH 7.4 and 5 mM EDTA revealed no hint of rRNA or tRNA contamination in the broad mRNA profile, but sedimentation in DMSO-sucrose under denaturing con- ditions produced a rather sharp peak at 18 S superimposed on a heterogeneous profile (Desrosiers et al., 1974). Furthermore, we discovered that when poly(A)+RNA was recovered from DMSO-sucrose gradients, the expected 100% binding to (dT)-cellulose was no longer achieved. Figure 1A Shows that approximately 95% of the labeled mRNA elutes as poly(A)+RNA when passed a third time through (dT)-cellulose at room temperature. When the mRNA isolated by two passes was first heat denatured as described in Materials and Methods before room tem- perature application to (dT)-cellulose, only about 70% was retained as poly(A)+ material (Fig. 1B). To characterize the RNA released by heat denatur- ation as nonmessenger contaminants, the poly(A)+ and the poly(A)-RNA isolated by heat denaturation and (dT)-ce11ulose chromatography (heat denatured poly(A)+ and heat denatured poly(A)-) were analyzed by 90 Figure 1. The effect of heat denaturation on the elution of poly(A)+RNA from (dT)-cellulose. Poly(A)+RNA that had been isolated by two passes through (dT)-cellulose as described in Materials and Methods was re-chromatographed on (dT)-cellulose’ after: (A) no treatment (b) heating to 60 for 2 minutes in 0.2% SDS. 91 400 "' O.I2 M NoCl I zooI— T POLY (AHRNA no heat treatment 0.0 M NoCI I + . .001 3H CPM POLY (AH RNA with heat troaimont TI O.I2 M NOCI 0.0 M NOCI ZOO '- 00 IO 20 30 FRACTION NUMBER 92 DMSO-sucrose gradient centrifugation. The heat denatured poly(A)-RNA produced three peaks that correspond to the cytoplasmic 4 S, 18 S, and 28 S RNAS (Fig. 2B). The 18 8 peak always predominated and in some preparations made up over 80% of the heat released radioactivity. We assume this peak is due to 18 S rRNA contamination. The heat denatured poly(A)+RNA sedi- mentation profile in DMSO-sucrose (Fig. 2A) reveals no Sharp peak at 18 S and retains the size distribution expected of mRNA. Alkaline hydrolysis of RNA produces mainly mononucleotide, but a ribose 2'-Qfmethyl group confers alkaline stability on the adjacent phosphodiester bond. A dinucleotide (-3 charge) is thus produced by alkaline hydrolysis for each 2'-Efmethylnucleotide while a mono- nucleotide (-2 charge) is produced for each base- methylnucleotide. Alkaline hydrolysis followed by identification of the charged species produced is an efficient method of characterization of methyl-labeled RNA Since most known RNAS have distinctive patterns of methylation. While rRNA methylation is approximately 90% 2'-Qfmethylation, tRNA is approximately 80% base- methylation (Desrosiers et al., 1974, Munns et al., 1974). Alkaline hydrolysis of 18 S + 28 8 rRNA from the poly(A)-RNA of the first (dT)-ce11ulose pass produced 82.6% dinucleotide and 9.9% mononucleotide 93 Figure 2. DMSO-sucrose gradient centrifugation of poly(A)+ and poly(A)-RNA obtained from heat denatur- ation and (dT)-cellulose of poly(A)+RNA. The poly(A)- and poly(A)+ fractions from Figure 1B were pooled and precipitated in ethanol. An aliquot of each RNA fraction was made 90-95% DMSO in 10 mM LiCl, 1mM EDTA and heated to 60° for 2 min. All of this material was layered over 4.8 ml of a linear 5-20% sucrose gradient in 99% DMSO (lOmM LiCl, 1mM EDTA). Centrifugation was for 14 h. at 26° and 45,000 rpm in a SW 50.1 rotor. Marker RNA was run in parallel tubes. (A) heat denatured poly(A)+; (B) heat denatured poly(A)-. 94 r I POLYIAI+ RNA 29s ms 43 no, I Ioo- aor- 2 a. o I I o I 300- POLY 00- RNA to 288 :83 4f zoo— tool- 1 00 IO 20 FRACTION NUMBER 95 (Fig. 3A). Mammalian rRNA is thought to contain two sequences of two adjacent 2'-Qfmethylnucleotides and one sequence of three adjacent 2'-Qfmethylnuc1eotides (Choi & Busch, 1974; Slack & Loening, 1974; Maden & Salim, 1974). The resolving power of the Pellionex-WAX column can be seen by the elution of radioactivity in the trinucleotide and tetranucleotide region of Figure 3A. Figure 3B shows that KOH hydrolysis of poly(A)+mRNA isolated without heat denaturation produced mononucleotide, dinucleotide and large oligonucleotides similar to what has been observed in L cells (Perry et al., 1975) and HeLa cells (Wei et al., 1975). The amount observed as dinucleotide in Novikoff hepatoma cells has varied with the preparation (18%, cf. 3B) but appears Slightly higher than that reported for the other two cell lines mentioned above. Note that the large oligonucleotides elute as a series of three peaks of charge -6, -7 and -8, with the predominant species being the -7 charged oligonucleo- tide. This is, on the average, at least one charge unit larger than has been observed by others for the large oligonucleotides of methylated viral mRNA (Urushibara et al., 1975; Wei & Moss, 1975; Furuichi et al., 1975; Rhodes et al., 1974) and cellular mRNA (Perry et al., 1975; Wei et al., 1975) using DEAF-Sephadex chromatography in 7 M urea. The reason for the observed differences between Pellionex-WAX and DEAE-Sephadex is presently 96 .omoHSHHoouaepv nmsounu odouo when» m opomoo cousumcoc too: we mommom omoHsHHoouApr 03» mo couoaomw <2m+A¢v>Hom cons +IacsHom mm mmusflm Haaum pan» azm "azm+lavsHom omnnumcmc umma lac .mmoHsHHmouleov monounu oaomo cuss“ m ouomon wousuocop boon ma mommom omOHsHHoOIABUV 03» an UoDMHOmw mzm+advmaom con3 Iamvhaom mo moDSHo pony dzm ”oc mos sadaoo one .mswodu Hoouo mooacamum :m\H Eoum ocme SESHoo x¢3|xocowaaom .80 mm o ouco nonconsw can «.5 mm as sons 2 h .ouonmmonm Edflpom z moo.o ca po>a0mmwc ouoz coauooum dzm sumo mo mononmaonchn mom one .mououmaouoan mom mo mammumovofiouno pwsvfla poomm amen xdzlxoSORHHom .m ousmwm 97 b m u m 2 3 z o... 0.4 8 so . I r :1 8. u b u o La La LII be ”as Erna. J. “ u 0 j I. I I I 18. I I : 3a 0 ”4H: 4H: ILII {as new. .zrnLiJmI - Po zo_._.o<¢u W0 out on 00 I - p... I I I I I III... m {a .3 {a {a ”as .222 .2 o ,. .. +8. I < p :EI, is .3. .3: is {a P b PL .2 1800 HdO 98 uncertain. The Pellionex-WAX profile obtained after KOH hydrolysis of heat denatured poly(A)+mRNA (Fig. 3D) reveals that almost all the dinucleotide peak is removed by the heat denaturation procedure. The ratio of counts in mononucleotide to large oligonucleotides is 53/47. In the profile obtained from heat denatured poly(A)-RNA (Fig. 3C), approximately 66% of the radioactivity elutes as dinucleotide and 34% as mononucleotide. 18 S rRNA is known to contain a higher percentage of base- methylnucleotides than 28 8 rRNA (Maden & Salim, 1974) and any contribution from tRNA would be mostly as mono- nucleotides. The elution profile of the heat denatured poly(A)-RNA is thus consistent with it being rRNA and tRNA contamination. Previously our analyses of 3 hour labeled mRNA using enzyme digestion yielded 50% base and 50% ribose methylation. The base-methylnucleosides identified 6,N6-dimethyladenosine were NG-methyladenosine (40%), N (~2.5%), and an unidentified component (7.5%). The unidentified component was probably 7-methylguanosine since the digestion conditions cause ring-opening of 7flmethylguanosine and the unidentified component migrated with ring opened 7-methylguanosine. The 2'- Qfmethylnucleosides were distributed among all four Species but heat denaturation was not used in the isolation procedure. The methylnucleoside composition 99 of the 13 hour labeled poly(A)+mRNA obtained by heat denaturation was determined using enzyme digestion and acid hydrolysis techniques. The results are summarized in Table l. N6,N6-dimethyladenosine was no longer detected in the mRNA, consistent with the removal of 18 8 rRNA contamination. 18 S rRNA is known to contain two such moieties (Maden & Salim, 1974). The methylated Species present in 13 hr. labeled mRNA are the same as found in 3 hr. labeled mRNA. The lower value for per- centage Qfmethylation reported here is due to the removal of residual contamination by the heat treatment- (dT)-cellulose procedure. Acid hydrolysis proved useful in further analyses since the conditions did not ring-open 7-methy1guanosine and did not convert 1-methyladenosine to the N6-deriva- tive. Control digestions with methyl-labeled tRNA revealed the expected percentage of l-methyladenine and almost no NG-methyladenine as has been previously reported (Munns et al., 1974). Figure 4 shows the separation of the bases released upon acid hydrolysis of the heat denatured poly(A)+mRNA by Aminex A-S high Speed liquid chromatography. The identity of N6- methyladenine and 7-methylguanine were further confirmed by descending paper chromatography in isopropanol, conc. HCl, H20 (680, 170, 144). These results identify 7-methylguanine and N6-methyladenine as the base- methylated species of poly(A)+mRNA. Determinations 100 Table 1 Summary of Methylnucleoside Composition Analyses Species Method of Hydrolysis m7G m6A Nm poly(A)+mRNAa enzymes 9.5 48 42 poly (A) +mRNAa acid 19 51 30 poly (A) mama B-elimination 17 mononucleotideb enzymes 53 oligonucleotidesb enzymes 14.5 32.5 aAnalyses were performed with poly(A)+mRNA obtained with heat denaturation and (dT)-cellulose. The RNA contained approximately 4% dinucleotide contamination. bAnalyses were performed with fractions isolated from Pellionex chromatography. 101 .muoxuoe omon onwusm mo cowumooa on» mouooecne msouuo on» no cowuwmom one .conooaaoo ouo3 AHE m.oev mnowuooum mono wuno3e .m.m mm ouoEHom EanoEEo z v.o nuHS ouommoum ESEHRME nuance oo~v .omv um comoao>oc mo: ssdaoo one .SEdHoo mud xonfiEd Hooum mmoacwmuo .80 cm o 0» poaammo can Hommsn SESHoo cw oo>a0mmep .m2 m0 Eoouum o Hoods cownc mo3 omoHSHHoOIAecv can coaumpsumcoc noon an ponaouno «sz+A¢VmHom mo ouonmaouvmn pwoo one .ououhaoupmn neon 42m8+nmvhaom mo enmoumoumEouno Uflsvea poomm nmen mud xonHE¢ .v ousmem 102 cm mmmzaz 202.012: 0? ON J! : 4—9 a do... <08 05:. @—-> n n a... n scream CON 00v WdO HS 103 of the percentage 7-methylguanosine by enzyme hydrolysis consistently yielded lower values than when measured by acid hydrolysis or by B-elimination which will be described below. A decreased recovery of the ring- opened form of 7-methylguanosine has been previouSly reported (Hall, 1971) and a similar phenomenon is pro- bably responsible for the low values obtained here using the enzyme digestion procedure. Thus the value of 17-19% 7-methylguanosine determined by acid hydrolysis and by B-elimination is most probably correct. Analysis of the 2'-Qfmethylnucleosides of heat denatured poly(A)+ mRNA produced by enzyme digestion revealed the presence of all four ribose-methylated species (Um, Gm, Am, Cm) (data not shown). The mononucleotide and all large oligonucleotides were recovered from Pellionex-WAX chromatography (Fig. 3) and enzymatically digested to nucleosides. The dephos— phorylated mononucleotide was composed entirely of N6-methyladenosine and this accounted for all of the NG-methyladenosine present in the poly(A)+mRNA. This agrees well with the findings of Perry et al. (1975) for the mononucleotide of L cell mRNA. Analysis of the large oligonucleotides after digestion revealed that they were approximately one-third 7-methylguanosine and two-thirds 2'-Efmethylnucleosides. The distribution of 2'-Qfmethylnucleosides obtained from the large 104 oligonucleotide (Fig. 5) agrees well with the 2'-Ef methylnucleoside distribution obtained when entire heat denatured poly(A)+mRNA was analyzed. This hetero- geneity in 2'-Efmethylnucleosides contrasts with the distinct Species observed with viral mRNAs (Furuichi et al., 1975; Wei & Moss, 1975; Urushibara et al., 1975) and probably reflects the heterogeneous pOpulation of messengers being analyzed. The high percentage of N6-methyladenosine in mRNA suggested to us that at least part of these moieties could be present in the poly(A) segment (Desrosiers et al., 1974). A nuclease digestion was thus performed on methyl- 1abe1ed mRNA using conditions that are known to leave the poly(A) segment intact. Less than 2% of the total methyl radioactivity was found associated with the poly A seg- ment and we conclude that the poly A is devoid of any methylated species (Table 2). This conclusion has also been reached by Perry et a1. (1975) for L cell mRNA. Periodate oxidation of RNA and B-elimination of the oxidized nucleoside require that both 2' and 3' hydroxyls be free. Since the poly(A) tract occupies the 3' end of each of these RNA molecules, B-elimination of radioactive nucleoside from methyl-labeled poly(A)+ mRNA would occur if a methylated 5'-5' structure were present at the 5' terminus. Furthermore, B-elimination of such a blocking group should free the phosphate in 105 .oouooaaoo ouoz AHE m.oev mnoeuooum mono noouufine .oASmmon ESEonE monsom oomm can own an mm.v mm doomam onoamnuo mow .oumEuom SSHcoEEo z «.0 nufls pomoao>on mos nEdHoo one .SEdHoo mud xocae< Eu cm 8 on pofiammo mos Hommsn :EdHoo Se noeuomuw opwmooaosnaenuoELmu.N one .mnmmnmouofiounv ououonum¢mo en sowuoouw opwmooaoscaenuoelomon m can noeuoouw opamooaoscaenpoELdl.~ m once oououomom oHo3 mocemooaosc one .Ho>oa ocemooaosc on» on poumomep madmoeuoeanno can mm onomem SH mo mnemumouoaouno xonoaaaom some oouo>ooou ouos o€mu Hmanwume man no hammnmoumaouno canvaa Ummmm sown mud xmcaa< .m musmflm 110 mum! :2 zo_hoflnm© mm3 h.m mo msam> one .Amhmav .Hm um muumm can a .H manna mam hammumoumeouno xmcoHHme Eoum Umcwmuno mosam> muwmomfioo mum cEdHoo was» cw coma mumnadc ones >.H h.m om.o EZ.~ o.m h.m Nm.o £05 o.H h.m mH.o one masomaoz amazomaoz Hmmcmmmmz mmmum>¢ mom mmfiommm comm mo Hmnfidz MHMUOB ummcmmmmz mmmu0>¢ umm x no cowuomum mamgumz mo umnssz annoy mmflommm manomaoz ezms+x4 cm cw mowommm omumamnumz mo coausnfluumfio m manna 116 have such a methylnucleoside composition, the numbers we have obtained are consistent with it. Perry et al. (1975) have observed a greater frequency of methylation per 1000 nucleotides in the smaller mRNA molecules, as would be the case if all mRNA molecules had either 5 or 6 methyl groups. Two factors are involved in the heterogeneity we have observed in the 5'-terminal structure: the 2'-Qfmethylation occurs with the four common nucleosides (Um, Gm, Am, Cm) and the 5'-structure may contain either one or two 2'-Qfmethylnucleotides. The exact nature of each of the large oligonucleotide peaks obtained from direct KOH hydrolysis of the Poly(A)+mRNA and Pellionex- WAX chromatography is not entirely certain. Three peaks were consistently observed and there is no simple way to correlate their percentages with the percentages observed as dinucleotide (Nmpr) and trinucleotide (Nmmepr) after B-elimination, phosphatase treatment and KOH hydrolysis. It seems possible that the unusual structure, the variable base composition, and the variable size (one vs. two 2'-Qfmethylnucleotides) can completely account for the heterogeneous oligo- nucleotide pattern. If the heterogeneity is indeed due to the fact that a mixed population of messengers is being examined, methyl-labeling and isolation of a specific mRNA should simplify the analysis. 117 Both et al. (1975) have recently found that the initiation and subsequent in gitrg translation of viral mRNA is dependent on the presence of 7-methylguanosine in a blocked 5' structure. It is thus possible that the methylation serves some important function in_yi!g in the translation of viral and cellular mRNA or its control. In addition, 2'-Qfmethylation may serve some role in the stabilization of certain RNA sequences to nuclease degradation. Although the maturation process of rRNA conserves only about 50% of the original molecule, the conserved regions retain all of the methylated sequences (Wagner et al., 1967). The presence of 2'-Qf methylnucleotides stabilized synthetic RNA molecules against hydrolysis by a processing 3'-0H exoribonuclease (Stuart & Rottman, 1973) and Wei et al. (1975) have observed that the blocked S'-terminus is resistant to 5'-nuclease action. Any increased stabilization caused by methylation may be important not only for nuclear processing but also for cytoplasmic turnover of the mRNA. In this latter regard, Singer and Penman (1973) have observed two populations of poly(A)+mRNA that were dis- tinguished on the basis of half-life (7 h. and 24 h.) and it is interesting to speculate whether these p0pu- lations are differentiated by the number of 2'-Qf methylnucleotides at the S'-terminus. Kinetic studies on the rate of appearance of the structures with one 118 vs. two 2'-Qfmethylnucleotides should reveal whether termini containing one 2'-Qfmethylnucleotide are derived from mRNA that turns over more rapidly. REFERENCES Abraham, G., Moyer, S. A., Adler, R. and Bannerjee, f A. K. (1975), Fed. Proc. 34, 2779. 1 Baskin, F. and Dekker, C. A. (1967), g. Biol. Chem. 242, 5 5447. § Both, G. W., Bannerjee, A. K. and Shatkin, A. J. (1975), Proc. Nat. Acad. Sci. USA (in press). Choi, Y. C. and Busch, H. (1974), Biochem. Bi0phys. Res. 3 Commun. 58, 674. Desrosiers, R., Friderici, K. and Rottman, F. (1974), Proc. Nat. Acad. Sci. USA ll, 3971. Dlugajczyk, A. and Eiler, J. (1966), Proc. Soc. Exp. Biol. Med. 123, 453. Fraenkel-Conrat, H. and Steinschneider, A. (1967), Methods in_Enzymology12B, 243. Furuichi, Y. (1974), Nucleic Acids Research I, 809. Furuichi, Y., Morgan, M., Muthukrishnan, S. and Shatkin, A. (1975), Proc. Nat. Acad. Sci. USA 23, 362. Gilham, P. T. (1964), g, Amer. Chem. Soc. §§, 4982. Hall, R. H. (1971), The Modified Nucleosides in Nucleic Acids, p. 147. Lavi, S. and Shatkin, A. (1975), Fed. Proc. 34, 1725. Maden, B. and Salim, M. (1974), g. Mol. Biol. 88, 133. Munns, T., Podratz, K. and Katzman, P. (1974), Bio- chemistry lg, 4409. Perry, R. and Kelley, D. (1974), Cell 1, 37. 119 120 Perry, R., Kelley, D., Friderici, K. and Rottman, F. (1975), Cell 3, 387. Pike, L. and Rottman, F. (1974), Anal. Biochem. 33, 367. Rhodes, D., Moyer, S. and Banerjee, A. (1974), Cell 3, 327. Ro-choi, T. S., Choi, Y. C., Henning, D., McCloskey, J. and Busch, H. (1975), 3. Biol. Chem. 250, 3921. Ro-choi, T., Reddy, R., Choi, T. C., Rej, N. B. and Hennings, D. (1974), Fed. Proc. 33, 1832. Rottman, F., Shatkin, A. and Perry, R. (1974), Cell 3, 197. Rottman, F. and Heinlein, K. (1968), Biochemistry 1, 2634. Singer, R. H. and Penman, S. (1973), 3. Mol. Biol. 13, 321. Slack, J. and Loening, V. (1974), Eur. 3. Biochem. 33, 69. Stuart, S. E. and Rottman, F. M. (1973), Biochem. Biophys. Res. Commun. 33, 1001. Urushibara, T., Furuichi, Y., Nishimura, C. and Miura, K. (1975), FEBS Letters 33, 385. Wagner, E. K., Penman, S. and Ingram, V. (1967), 3. Mol. Biol. 33. 371. Wei, C., Gershowitz, A. and Moss, B. (1975), Cell 3, 379. Wei, C. and Moss, B. (1975), Proc. Nat. Acad. Sci. USA 13, 318. “" PART III THE SENSITIVITY OF RNA POLYMERASES I AND II FROM NOVIKOFF HEPATOMA (NISI) CELLS TO 3'-DEOXYADENOSINE 5'-TRIPHOSPHATE _ ‘ Cum-.1! 121 Summary The synthesis of ribosomal precursor RNA in Novikoff hepatoma (NlSl) cells is very sensitive to cor- dycepin (3'-dA). The synthesis of hnRNA, however, is resistant to inhibition by concentrations of 3'-dA that completely block the synthesis of 455 ribosomal RNA pre- cursor. I have examined the RNA polymerases present in these cultured cells with regard to their sensitivity to cordycepin 5'-triphosphate (3'-dATP) in an effort to explain the differential inhibition of RNA synthesis observed in 3332. RNA polymerases I and II were char- acterized on the basis of their chromatographic behavior on DEAF-Sephadex, as well as the response of their enzymatic activities to ionic strength, the divalent metal ions Mn2+ and Mgz+, and the toxin a-amanitin. For both enzymes the inhibition of ifl.!i££2.RNA synthesis by 3'-dATP was competitive for ATP. The Km values for ATP and the Ki values for 3'-dATP for the two enzymes were quite similar. RNA polymerase II, the enzyme pre- sumed responsible for hnRNA synthesis, was actually slightly more sensitive to 3'-dATP than RNA polymerase I, the enzyme presumed responsible for ribosomal precursor RNA synthesis. Similar data were obtained when the RNA polymerases were assayed in isolated nuclei. These results indicate that the differential inhibition of 122 RNA synthesis caused by 3'-dA 33 vivo cannot be simply explained by differential sensitivity of RNA polymerases I and II to 3'-dATP. Introduction Concentrations of 3'-dA (cordycepin) that do not inhibit nucleoplasmic hnRNA synthesis in HeLa cells cause drastic inhibition and apparent termination of ribosomal RNA precursor synthesis in the nucleolus (Siev, Weinberg & q , Penman, 1969; Penman, Rosbash & Penman, 1970). The most IF", .1"... 3’. ‘ commonly cited explanation for this phenomenon is that there are distinct RNA polymerases in nuclei with dif- ferent sensitivities to the triphosphate derivative of the drug (Siev et al., 1969; Plagemann, 1971; Sarkar, Goldman & Moscona, 1973; Grahn & Lovtrup-Rein, 1971). The inhibition by 3'-dA of mRNA synthesis but not of hnRNA synthesis was originally used as evidence for the separate synthesis of these two RNA species (Penman et al., 1970). Since the discovery of poly(A), evidence has been presented by Adesnik et al. (1972), Darnell et al. (1971) and Abelson and Penman (1972) that the inhibition of mRNA appearance in the cytoplasm caused by 3'-dA is a result of the inhibition of post-transcriptional poly(A) addition. Eucaryotic cells from a variety of sources have been shown to contain at least three physically and bio- chemically distinct DNA-dependent RNA polymerases (Roeder & Rutter, 1969; for review, Chambon, 1974). 123 The activities which most consistently appear have been classified I, II and III on the basis of their order of elution from DEAR-Sephadex. RNA polymerase I is optimally active at low salt concentrations (ammonium sulfate less than 0.05 M), utilizes Mn2+ and Mg2+ almost equally well to satisfy its divalent metal ion requirement and is insensitive to inhibition by the toxin a-amanitin at concentrations up to 200 ug/ml. This polymerase is of nucleolar origin and has been demonstrated to be responsible for the synthesis of ribosomal precursor RNA (Roeder & Rut- ter, 1970; Zylber & Penman, 1971; Reeder & Roeder, 1972). RNA polymerase II is optimally active at ammonium sulfate concentrations between 0.1 and 0.15 M, prefers Mn2+ over Mg2+ and can be completely inhibited by very low concen- trations of a-amanitin. This RNA polymerase activity is found in the nucleoplasm and is probably responsible for most, if not all, hnRNA synthesis (Roeder & Rutter, 1970; Zylber & Penman, 1971; Blatti et al., 1970). The nucleo- plasmic RNA pol-merase III is found only in small quanti- ties, is sensitive to a-amanitin only at high concentra- tions (N 200 ug/ml) and is responsible for SS ribosomal and transfer RNA synthesis (Weinmann & Roeder, 1974; Weil & Blatti, 1975). The existence of distinct RNA polymerases for the synthesis of ribosomal precursor RNA and hnRNA makes it reasonable to test the hypothesis that the selective 124 inhibition of ribosomal RNA transcription by 3'-dA is due to a greater sensitivity of RNA polymerase I than RNA polymerase II to 3'-dATP (3'-deoxyadenosine 5'- triphosphate; cordycepin triphosphate). In fact, there is precedent for two distinct RNA polymerases in a single cell varying markedly in their response to 3'-dATP. 5* The bacteriophage gh-l induced RNA polymerase isolated g from infected Pseudomonas putida cells displayed a 100- fold greater sensitivity to 3'-dATP than the host 3. putida RNA polymerase (Towle, Jolly, & Boezi, 1975). an. Vesicular Stomatitis viral mRNA and poly(A) syntheses are very insensitive 39 1329 to 3'-dA and 33_git£g to 3'-dATP (Ehrenfeld, 1974; Bannerjee, Moyer & Rhodes, 1974). In contrast, the host RNA synthesis is very sen- sitive to the 3'-dA (Ehrenfeld, 1974). Unlike host RNA synthesis, influenza viral mRNA synthesis is insensitive to 3'-dA (Etkind & Krug, 1974). The purpose of this research was to determine whether 33 gitgg RNA syntheses by the isolated forms of RNA polymerase I and II display sensitivities to 3'-dATP which would explain the action of 3'-dA in whole cells. Blatti et al. (1970) have reported inhibition of calf thymus polymerase I to a similar extent as polymerase II by 3'-dATP but I believe it is important to use enzymes from cells that can be shown to exhibit the differential inhibition of RNA synthesis 33 vivo and to do a more complete kinetic 125 analysis. This report also describes the isolation and characterization of RNA polymerases I and II from Novikoff hepatoma (NlSl) cells since this has not been previously reported. Experimental Procedures Synthesis of 3'-dATP 3'-dA was phosphorylated by the method of Yoshik- kawa (1967) using POCl3 in triethylphosphate. Pure 3'- deoxyadenosine 5'-monophosphate was isolated by chroma- tography on Bio-Rad AG l-XZ (formate) in approximately 60% yield. The 5'-triphosphate derivative was prepared by the displacement of the morpholidate derivative of 3'-deoxyadenosine 5'-monophosphate with the tri-n- butylamine salt of perphosphate (Moffatt, 1967). Pure 3'-deoxyadenosine 5'-triphosphate was isolated by Bio- Rad Ag l-X2 (Cl-) chromatography using a LiCl gradient in 0.01 M HCl. The triphosphate was desalted by DEAE- cellulose chromatography and converted to the tetrasodium salt prior to analysis. Approximately 70% yield was obtained from the monophosphate. The 3'-deoxyadenosine 5'-triphosphate was judged pure by chromatographic analyses and released 3.0 moles of phosphate per mole of adenosine upon complete acid hydrolysis. “J, .. 126 Effect of 3'-dA on in vivo hnRNA and rRNA Precursor Synthesis Novikoff hepatoma (NlSl) cells were grown in culture at 37° as described by Desrosiers, Friderici and Rottman (1974). For labeling of ribosomal precursor RNA, cells (4-6 X 105 per ml) were preincubated 20 min with 100 uM 3'-dA (treated) or 100 uM adenosine (control) prior to the start of labeling with 3H-uridine (0.33 pM; 0.8 uCi per ml). After 20 min of incorporation at 37°, the cells were poured over frozen crushed balanced salt solution. The nuclei were prepared by swelling in hypotonic buffer and dounce homogenization as described 14C-cytOplasmic RNA was (Desrosiers et al., 1974). added to the pelleted nuclei to monitor recovery and as a sedimentation marker for sucrose gradient centrifu- gation. The nuclei were lysed with 1.0 ml 0.5 M NaCl, 0.01 M Tris pH 7.4 and 0.05 M MgCl2 and the DNA digested with a small crystal of Worthington electrophoretically pure DNAase. To the DNAase digested solution was added 1.0 ml 0.1 M EDTA and 0.12 ml 10% SDS. Protein was digested with proteinase K and the RNA isolated by phenol chloroform extraction (Singer & Penman, 1973; Desrosiers et al., 1974). For labeling of hnRNA, cells (4-6 X 105 per ml) were preincubated with 0.06 ug/ml Actinomycin D for 40 min, the last 20 min of which additionally included 100 uM 3'-dA (treated) or 100 uM adenosine (control). The cells were then labeled and 127 the nuclear RNA isolated exactly as described above. DMSO-sucrose gradient centrifugation under denaturing conditions was performed as described (Desrosiers et al., 1974). The RNA samples in 90-95% DMSO (10 mM LiCl, 1 mM EDTA) were heated to 65° for 3 min prior to layering on gradients and centrifugation. Solubilization and Separation of RNA Polymerase Activities When the culture had reached a density of 6 to 8 X 105 cells/ml, cells were harvested by centrifugation. All subsequent procedures were performed at 0 to 4°C. From 2 to 2.5 X 109 cells were used for the typical purification. Cells were washed and concentrated with a balanced salt solution. The cells were then suspended in 40 ml of hypotonic lysis buffer containing 10 mM Tris°HCl, pH 7.6, 6 mM KCl, 5 mM MgCl2 and allowed to swell for 10 minutes. After this time, a 10% (v/v) solution of NP-40 detergent was added to the cell suspension to a final concentration of 0.5%. The sus- pension was mixed briefly and the nuclei collected by centrifugation at 1,000 x g for 2 minutes. The nuclear pellet was washed twice with hypotonic lysis buffer. The final nuclear pellet was suspended in about 15 ml of a buffer containing 10 mM Tris-RC1, pH 7.9, 50 mM MgC12, l M sucrose, 1 mM dithiothreitol. 128 The solubilization and chromatography on DEAE- Sephadex of Novikoff hepatoma RNA polymerases were carried out by a modification of the procedure of Roeder and Rutter (1970). A solution of 4 M (NH4)ZSO4, pH 8.0, was added to the nuclear suspension to give a final con- centration of 0.3 M (NH4)ZSO4. The viscous suspension was sonicated on a Biosonik sonicator at full voltage for 1 1/2 minutes (in lO-second bursts with intermittent cooling). The sonicated solution was diluted with two volumes of a buffer containing 50 mM Tris-HCl, pH 8.0, 12.5% (v/v) glycerol, 5 mM MgCl 0.1 mM EDTA, 0.5 mM 2: dithiothreitol and then centrifuged at 40,000 RPM in a Spinco 40 rotor for 1.25 hours. To the supernatant fluid from the centrifugation, 0.28 grams of (NH4)ZSO4 per m1 of solution were slowly added. The solution was stirred for 30 minutes and the precipitated material collected by centrifugation at 15,000 x g for 20 minutes. This (NH4)ZSO4-precipitated material was resuspended in a small volume of buffer containing 50 mM Tris-HC1, pH 8.0, 25% (v/v) glycerol, 5 mM MgC12, 0.1 mM EDTA, 0.5 mM dithiothreitol (TGMED Buffer). The resuspended material was dialyzed for 8 hours against 1 liter of TGMED Buffer containing 0.03 M (NH4)ZSO4. After dialysis, the insoluble material present was removed by centrifu- gation at 15,000 x g for 20 minutes and the supernatant solution used for further purification. 129 A DEAE-Sephadex column (2 x 12 cm) was packed and equilibrated with TGMED Buffer containing 0.03 M (NH4)2SO4. The enzyme solution was loaded onto the column at a rate of 0.5 ml per minute or less. The column was then washed with two column volumes of TGMED Buffer containing 0.03 M (NH4)ZSO All detectable RNA 4. polymerase activity adsorbed to the DEAE-Sephadex under these conditions. The column was eluted with a linear gradient from 0.03 M to 0.6 M (NH4)ZSO in TGMED Buffer 4 (total volume of 300 ml) and fractions of about 3.5 ml were collected. Samples of the column fractions were assayed in the Limiting UTP Reaction Mixture described below. The pooled RNA polymerase fractions were dialyzed against 1 liter of TGED Buffer (same as TGMED Buffer except no MgClz) containing 0.03 M (NH4)ZSO4. Small phosphocellulose columns with total bed volumes of 1.5 ml were packed and equilibrated with the same buffer. The pooled dialyzed fractions were loaded on the phospho- cellulose columns and the columns were washed with five column volumes of TGED Buffer with 0.03 M (NH4)ZSO4. RNA polymerase was then eluted with TGED Buffer contain- ing 0.5 M.(NH4)ZSO4 and 0.5 ml fractions were collected. RNA polymerase activity was generally eluted in 1 to 1.5 m1 of buffer with a 70 to 80% recovery of activity. At this stage, MgCl2 was added to a final concentration of 5 mM and, on occasion, bovine serum albumin was 130 added to a final concentration of 0.5 mg/ml. Final enzyme samples were dialyzed against TGMED Buffer con- taining 0.03 M (NH4)ZSO4 for 6 hours, divided into small aliquots, and stored at -80°C. Assay of Solubilized RNA Polymerase Activity The standard reaction mixture for RNA polymerase I and II contained in a volume of 0.125 ml: 50 mM Tris'HCl, pH 7.9, 1.6 mM MnCl 1 mM dithiothreitol, 2: 160 ug/ml calf thymus DNA, 0.6 mM each of ATP, GTP, and GTP, 0.1 mM [3H]UTP (1.25 no per assay), either so mM (NH4)ZSO4 (RNA polymerase I) or 100 mM (NH4)ZSO4 (RNA polymerase II) and enzyme samples. The Limiting UTP Reaction Mixture for assaying column fractions contained in 0.125 ml: 50 mM TriS'HCl, pH 8.0, 1.6 mM MnClZ, 1 mM dithiothreitol, 160 ug/ml calf thymus DNA, 0.6 mM each 4 mM [BHJUTP (1.25 uc of ATP, GTP, and GTP, 6.7 x 10' per assay) and enzyme samples. All reactions were ini- tiated by the addition of enzyme and incubated for 10 minutes at 37°C. Reactions were terminated by the addi- tion of 10 pl of 0.2 M EDTA and placed in an ice bath. Samples of 0.1 ml of each reaction mixture were pipetted onto Whatman DE8l filter discs, which were then washed by the procedure described by Blatti et a1. (1970). Filters were dried at 100°C for 5 minutes and counted 131 in a scintillation fluid containing 4 grams of 2,5-bis-2- (5-tert-Buty1benzoxazoly)-Thiophene per liter of toluene. Assay of RNA Polymerase Activity Using IsoIated Nuclei Nuclei were prepared as described for the solu- bilization of RNA polymerase activities, only dounce homogenization was used instead of NP 40 to achieve cell breakage. The pelleted nuclei were washed once with the hypotonic lysis buffer, resuspended in 25% glycerol, 5 mM Mg Acetate, 50 mM Tris-H01, pH 7.9, 5 mM DTT, 10 mM KCl and frozen in aliquots immediately at -80°. The reaction mixture contained in a volume of 0.125 ml: 50 mM TriS'HC1, pH 7.9, 1.0 mM MnCl 5 mM Mg Acetate, 2: 3 mM DTT, 1.0 mM each of ATP, GTP and CTP, 0.05 mM [3H]-UTP (5 HO per assay) and 100 mM KCl. Reactions were initiated by the addition of nuclei and incubated for 10 minutes at 30°. Reactions were terminated by the addition of 0.1 ml 0.1% SDS and 2 ml of a cold solution of 10% trichloroacetic acid - 1% sodium pyrophosphate. Measurement of acid insoluble radioactivity was performed as described (Boezi et al., 1974). RNA polymerase I was measured by the inclusion of 3 ug/ml a-amanitin in the assay. RNA polymerase II was measured by isolating nuclei from cells that had received a 40 min pretreatment with 0.08 ug/ml Actinomycin D to inhibit RNA polymerase I. Assays using nuclei from Actinomycin D pretreated cells 132 measured mostly RNA polymerase II since 3H-UTP incor- poration was inhibited approximately 70% by 3 ug/ml a-amanitin. Materials DEAE-Sephadex (A-25) was purchased from Pharmacia Fine Chemicals. NP-40 detergent was obtained from Shell. Phosphocellulose (P-11) and Whatman DE-8l filter discs were from Reeve-Angel. Calf thymus DNA was purchased from P-L Biochemicals and further purified by two suc- cessive SDS-phenol extractions, followed by extensive dialysis. Cordycepin was isolated as described by Kredich and Guarino (1960). Results Effect of 3'-dA on rRNA Precursor and hnRNA Synthesis The effect of 3'-dA on cellular incorporation of 3H-uridine into total nuclear RNA is shown in Figure 1A. At 100 uM 3'-dA, the incorporation of 3H uridine into RNA is probably a true measure of RNA synthesis since the uptake of radioactive uridine into total cell material is not inhibited at this concentration of 3'-dA (Plage- mann, 1970). Without 3'-dA pretreatment a considerable portion of the radioactivity after a 20 min labeling period is present in 45 S rRNA precursor. 3H-Uridine incorporation in the presence of 100 uM 3'-dA, however, 133 .omummuu dol.m .QIIIO “Houucoo OIIIO «szmcn .m «nomusomum «2mm .d .mHMH> coaumaaflucfiom oucfi hauomuflo ommmflun mum3 mc0fluomum .mucmwomum may on Umflammm mcflmn muowmn sHE m How omo on vmumms mumB Adaam SEA .Hqu SE oav cmzo wma-om an mmamemm .uouou H.om 3m 6 an smm ooo.m¢ um 0mm pm as m.ma now Adana zed .Hoflq zsoav Oman mam an mucmflemum mmouosm womum nmsounu mm: coaummSMHHucwo .mQOHuHocoo mcwnsumcmo Hoods coaummsmwuusmo ou nowum mousomooum Hmucmeflummxm cw confluommo mm omumaomfl new umamnma was 42m HmmHosz .fluosvcoo ha vocfleuouoo mos mcowuoouw m50fiuo> cw wommawmzv mo coauouucoocoo one .ousuxfiz sofiuooom m9: mcwufiEHq on» ad muw>wu0o omouosmaom 42m How oohommo ouos macauooum msowuo> mo AH: omv moameom .mousooooum Houcosauomxm ca oonfluomoo mo UoEHowuom mm3 cadaoo xooonmomum¢mn o co maaoo osouomon «moxfl>oz mo uoouuxo Hooaosc o no unmoumouoeouno .momouosmaom dz“ Hooaosc mEouomon mmoxfl>oz mo mammumouofiouno xooonmomlm¢mo .m ousmflm 137 (-‘W) [’oszmm] 0 l0 "' 0.45 d I l J I 0 V' N ("9.0I X N60) NOIIVUOdNOONI dflfl 50 7O FRACTION NUMBER 30 IO 138 0.15 M (NH4)ZSO4 and constituted from 45 to 60% of the total RNA polymerase activity; RNA polymerase II, which was found at 0.28 M (NH4IZSO4 and contained 30 to 45% of the total activity; and RNA polymerase III, which eluted at 0.33 M (NH4)ZSO4 and constituted from 5 to 10% of the activity. The third RNA polymerase activity was designated as RNA polymerase III on the basis of its relative insensitivity to inhibition by a-amanitin but it was not further studied. Fractions with the majority of RNA polymerase I and II activity were pooled as shown in Figure 2 and concentrated by phosphocellulose chroma- tography for further characterizations. At this stage, the RNA polymerases were completely dependent on the presence of DNA, the four nucleoside triphosphates, and a divalent metal ion for enzymatic activity. The effect of various concentrations of (NH4)ZSO4 on the activity of RNA polymerases I and II is shown in Figure 3. RNA polymerase I was found to display optimal enzyme activity at relatively low concentrations of (NH4)ZSO4, from 0.04 to 0.06 M. RNA polymerase II, on the other hand, required higher concentrations of (NH4)ZSO4, from 0.1 to 0.15 M for optimal enzyme activity. The effect of varying the concentrations of the divalent metal ions, Mn2+ and Mgz+, on the activity of the two RNA polymerases is shown in Figure 4. Both RNA polymerases I and II show maximal enzyme 139 Figure 3. The effect of varying the concen- tration of (NH4)ZSO4 on in ygtgg RNA synthesis by the Novikoff hepatoma RNA polymerases I and II. Standard reaction mixtures were prepared as described in Experimental Procedures except that (NH4)ZSO4 was added to the reaction mixtures at the concentrations indicated. Reactions were initiated by either the addition of RNA polymerase I ( O ) or RNA Polymerase II ( I ), incubated for 10 min, and analyzed for RNA synthesis as described in Experimental Procedures. (W) WOSW’HMJ 140 UMP INCORPORATION (0PM x IO'3,o) 1. 9' A m o in I I .0 0— Ul . q I .0 5' n .2- u - .0 NP- .- O .0 N— .— OI JR I I 9 - .- m N 00 UMP INCORPORATION (CPM x IO‘3,.) .1 nI-nw 1‘ m 9 I' >VIV' ('4' f 141 Figure 4. The effect of varying the concentration of the divalent metal ion on in yitgg RNA synthesis by the Novikoff hepatoma RNA polymerases I and II. Standard reaction mixtures were prepared as described in Experimental Procedures except that the concentration of Mn2+ was varied as indicated ( 0 ) or Mn2+ was replaced by Mg2+ at the concentrations indi- cated ( I ) . Reactions were initiated by either the addition of RNA polymerase I (upper diagram) or RNA polymerase II (lower diagram), incubated for 10 min, and analyzed for RNA synthesis as described in Experi- mental Procedures. 142 — _ q 1 _ _ I [R n I ll IMU M II I 18.”. w I T 16“». .l E II 14.-"l. ll 12 _ _ _ p 3 2 I 0 5 w n. 0 3:0. x 2&3 zo_._.._._>_._.o< hzmomwn— _ O 3 IO- 147 the absence and presence of 3'-dATP was determined. Within experimental error, 3'-dATP appeared to act as a competitive inhibitor of ATP for both RNA polymerases I and II (Figs. 6 and 7). The apparent Km values for ATP for both RNA polymerases were similar: 3.5 x 10.5 M for RNA polymerase I and 4 x 10-5 M for RNA polymerase II. F“ The apparent K value for 3'-dATP for RNA polymerase I I was calculated to be 1.4 x 10-5 M, while that for RNA polymerase II was 7 x 10.6 M. A summary of the kinetic constants obtained is shown in Table 1. As was seen in Figure 5, RNA polymerase II appeared to be more sensitive to inhibition by 3'-dATP than RNA polymerase I. Vir- tually identical results were obtained when a preparation of Novikoff hepatoma DNA was used as template instead of calf thymus DNA (data not shown). Furthermore, the RNAase activity in both enzyme preparations was negli- gible. No significant inhibition of RNA polymerase activity was observed with 3'-dA, 3'-dADP and 2'-dATP. Thus, the isolated forms of RNA polymerase I and II do not possess the relative difference in sensitivity to 3'-dATP in in yitgg RNA synthesis that would explain the selective inhibition of ribosomal RNA synthesis by 3'-dA observed with whole cells. The relative sensitivities of the RNA polymerases to 3'-dATP were also assayed in isolated nuclei. RNA polymerase I was measured by including 3 ug/ml a-amanitin 148 .mousooooum Hounoefluomxm cH confiuomoo mo memonunhm 42m How ooumaono oco touocfleuou ouoz mcowuooom .me¢©|.m S OH x min . I $931.... 2 3 x o.m 3 Semen; on .o $36-3 «o 303328 Ml Inoo mcflzoaaom onu coneounoo Omao mousuxwe noeuooom .Uouoowonfl mo oowuo> mos me< mo coeuouunoonoo onu ponu umooxo monsooooum Hounoaeuomxm ca confluomoo on can coo couomoum oHoB H omoHoE%Hom «2m How mousuxwe coeuooou ouoocoum .H omoHthHom «zm mEouomon muonfl>oz an memonucwm dzm ouuw> mm no ouonmmonmeuuu.m onemonoooexoonl.m mo oonomoum ono oonomno onu ca med «0 soauouunoonoo on» mnewno> mo uoommo one .m ousmflm 149 '20 I I I . I. O I I I I I I I I I O O O '0 N - 'I._NdO) ,0: x All IO 20 3O 4O 50 l/[ATPJ (MM-II 'IO -"'-. Aha-v- _...- .1- 150 .mousooooum HousoEeHomxm cw wonwuomou mo memonunmm «2m How commaoco Ugo oouonHEHou oHo3 ms0fluooom .medol.m 2 OH x m 0.... . I £56-; 2 -3 x NA .4 3946-8 on .u "3%)... no 203388 m Icoo mnfisoHHOM on» conflounoo Omao mousuer coauooom .oouoowonw mo oowuo> mo3 me< mo conuouucoonoo on» pony umooxo monotonoum Hounoefluomxm ca tonfiuomoo mo can cam ooummoum ouoz HH omouoEMHOQ «2% you mousuxwa noeuooou ouoonoum .HH omouoeeaom dzm oEouomon muonw>oz an mflmonucmm nzm ouuw> mm no ouonmmonmwuuu.m ocflmononomxooon.m mo oonomoum tam oonomnm onu ca med mo nowumuunoonoo on» mnwmuo> mo uoomuo one .5 ousmfim I J3. rill... . . 1v. ..o 1. ' .F 151 I 40 3O I 20 I/ [ATP] (MM-'I L IO 1- 4... -I_ 1 -IO d’ -20 I l O O N '- (._m0) ,0: x All 152 Table 1 I” Summary of Kinetic Constants a b K I Enzyme Km KI KE' : I ": 1 RNA Polymerase I 35 14 2.5 RNA Polymerase II 40 7 5.7 auM for ATP bum for 3'-dATP 153 in the reaction mixture. To measure RNA polymerase II activity, nuclei were isolated from cells that received a pretreatment with a low level of Actinomycin D to inhibit RNA polymerase I (Zylber & Penman, 1974). Since only 70% of the activity was inhibited by 3 ng/ml a-amanitin, approximately 30% of the nuclear activity from these Actinomycin D treated cells was from poly- merases other than RNA polymerase II. Nevertheless, the results shown in Figure 8 using isolated nuclei clearly show that, as was the case with the isolated enzymes, RNA polymerase II is slightly more sensitive than RNA polymerase I to inhibition by 3'-dATP. Discussion 3'-Deoxyadenosine has been shown to inhibit rRNA precursor production when present at concentrations which do not affect the synthesis of heterogenous nuclear RNA. A reasonable explanation for the differential effect of 3'-dA is that the enzyme responsible for the synthesis of hnRNA (RNA polymerase II) is much more resistant to inhibition by 3'-dATP than the enzyme responsible for the synthesis of rRNA precursor (RNA polymerase I). In this paper, it has been shown that in 23552 RNA synthesis by RNA polymerase II is actually slightly more sensitive to 3'-dATP than is in yitgg RNA synthesis by RNA polymerase I. Thus, the isolated forms of RNA polymerases I and II do not exhibit the specificity W-._.____—._‘e-J§.'- aux ' I “L... . . -Iak 154 Figure 8. The effect of 3'-deoxyadenosine 5'-triphosphate on RNA Synthesis by Novikoff hepatoma nuclei. RNA polymerase I ( 0 ) or RNA polymerase II (I ) was selected in assays using isolated nuclei as described in Experimental Procedures. The concentration of ATP in the reaction mixture was 1.0mM. RNA synthesis in reactions containing 3'-dATP was expressed as the per- centage of RNA synthesis in control reactions containing no 3'-dATP. In reaction mixtures containing no 3'-dATP, 100% activity was equivalent to 920 CPM of UMP incor- poration in 10 min for RNA polymerase I and 1850 CPM for RNA polymerase II. 155 — 0 5 >H_>_._.o< Hzmomma IOO- 75- 25- PI 3 -LOG [3'-dAT 4 156 towards 3'-dATP necessary to explain the effects of 3'-dA in whole cells. The studies reported here for Novikoff hepatoma (NlSl) cells is to my knowledge the first time that the effects of 3'-dA on whole cells have been coupled with accurately determined K values I for 3'-dATP for a single cell type. '6‘ 3. One possible explanation for the discrepancy 1 between the effects of 3'-dA in whole cells and the relative sensitivities of the RNA polymerases towards 3'-dATP is that the isolated forms of the RNA polymerase J differ from the forms of the enzyme functioning in the cell. In this regard, it has been shown that isolated RNA polymerases do not faithfully transcribe defined templates that are available (Mandel & Chambon, 1974; Roeder, Reeder & Brown, 1970). Several protein factors which affect in yitgg RNA synthesis by eukaryotic RNA polymerases have also been described (for example; Seifert, Juhasz & Benecke, 1973; Benson, Spindler & Blatti, 1974). It is conceivable that such protein factors may be bound to the RNA polymerases in the nucleus as part of the transcriptional complex, but are lost during purification. The binding of such factors could possibly alter the RNA polymerase to make it more or less sensitive to 3'-dATP. One might expect factors which affect the transcriptional activity of RNA polymerases in the cell, to still be present in 157 isolated nuclei. When isolated nuclei were used to assay the RNA polymerases, however, RNA polymerase II still appeared slightly more sensitive than RNA poly- merase I to 3'-dATP. Thus, either such factors are not functional in isolated nuclei or they play no role in RNA synthesis inhibition caused by 3'-dATP. It must also be considered a distinct possibility from these studies that the selective inhibition of ribo- somal RNA transcription by 3'-dA in whole cells is not due to the differential sensitivities of RNA polymerases I and II to 3'-dATP. A conceivable but unlikely alter- native is that the bulk of hnRNA synthesis is not per- formed by RNA polymerase II in the cell. Numerous hypo- theses concerning the mechanism of 3'-dA inhibition of ribosomal RNA production can be envisioned, with little, if any, supportive data favoring any particular one. For instance, it is conceivable that different nucleo- tide pools for ribosomal and heterogenous nuclear RNA synthesis might exist. If this is so, 3'-dATP might be selectively concentrated in the nucleolar pool causing a higher molar ratio of 3'-dATP to ATP in this pool. Another possibility is that the block in nucleolar RNA synthesis is not due to direct inhibition by 3'-dATP but rather it is somehow related to the stringent response originally observed in bacteria but more recently described for eucaryotic cells (Gross & Pogo, 1974). 158 Many bacteria respond to a stress situation, such as the loss of an essential amino acid, by almost imme- diately restricting the synthesis of a variety of cri- tical metabolic products including stable RNA, i.e. rRNA and tRNA (Edlin & Broda, 1968). Bacteria synthesize no more ribosomes than can be efficiently engaged in my 4.”;- . I I protein synthesis (Maaloe, 1969). Usually stable species of RNA are also found to be extensively degraded (Kaplan a & Apirion, 1973). The bacterial response is related to the levels of guanosine polyphosphates (ppGpp and pppGpp) ‘ M9? fila‘u ‘. 4U ,,. 4F ‘ ‘ “If: in the cell (Cashel & Gallant, 1969) and these compounds have recently been discovered in mammalian cells (Irr, Kaulenas & Unsworth, 1974). The stringency phenomenon has not been extensively studied in eucaryotic cells but remarkably the response drastically affects nucleolar rRNA precursor synthesis with little or no effect on the synthesis of hnRNA (Franze-Fernandez & Pogo, 1971; Gross & Pogo, 1974). Willis, Baseman and Amos (1974) and Yu and Feigelson (1972) have also observed a greater effect of protein synthesis inhibition on nucleolar RNA synthe‘ sis than on other types of RNA synthesis. It seems possible then that 3'-dA or one of its metabolites could cause some stress situation in the cell, such as a decrease in available mRNA, of protein synthesis or of the level of some amino acid. A resultant stringent response could cause a block in rRNA synthesis without 159 any direct effect of 3'-dATP on polymerase activity. This is of course highly speculative but amenable to experimental analysis. I‘flm-l’ux S-. “A. .J “n' " . REFERENCES Abelson, H. T. and Penman, S. (1972), Biochim. Biophys. Acta 277, 129. T" Adesnik, M., Salditt, M., Thomas, W. and Darnell, J. E. I (1972), J. Mol. Biol. 11, 21. Banerjee, A. K., Moyer, S. A. and Rhodes, D. P. (1974), Virology 31. 547. Benson, R. H., Spindler, S. R. and Blatti, S. (1974), J. Cell Biol. 33, 21a. m. Blatti, S. P., Ingles, C. J., Lindell, T. J., merris, P. W., Weaver, R. F., Weinberg, F. and Rutter, W. J. (1970), Cold Spring Harbor Symp. Quant. Biol. 33, 649. Boezi, J. A., Lee, L. F., Blakesley, R. W., Koenig, M. and Towle, H. C. (1974), J. Virol. 13, 1209. Cashel, M. and Gallant, J. (1969), Nature 221, 838. Chambon, P. (1974), In The Enzymes, vol. 10, P. Boyer ed. (New York: Academic Press), p. 261. Darnell, J. E., Philipson, L., Wall, R. and Adesnik, M. (1971), Science 174, 507. Desrosiers, R., Friderici, K. and Rottman, F. (1974), Proc. Nat. Acad. Sci. USA 11. 3971. Derman, E. and Darnell, J. E. (1974), Cell 3, 255. Edlin, G. and Broda, P. (1968), Bacteriol. Rev. 31, 206. Ehrenfeld, E. (1974), J. Virol. 13, 1055. Etkind, P. R. and Krug, R. M. (1974), Virology 31, 38. Franze-Fernandez, M. T. and Pogo, A. O. (1971), Proc. Nat. Acad. Sci. USA 33, 3040. 160 161 Grahn, B. and Lovtrup-Rein, H. (1971), Acta Physiol. Scad. 1. Gross, K. J. and Pogo, A. O. (1974), J. Biol. Chem. 249, 568. Irr, J. D., Kaulenas, M. S. and Unsworth, B. R. (1974), Cell 3, 249. Kaplan, R. and Apirion, D. (1974), J. Biol. Chem. 249, 149. -1 Kredich, N. M. and Guarino, A. J. (1960), Biochim. Biophys. Acta 5;. 363. Maaloe, O. Twenty-eighth Symposium of Society for Developmental Biology: Communications in Development, University of Colorado (1969). F5] Mandel, J. and Chambon, P. (1974), Eur. J. Biochem. 31, '3 379. Moffatt, J. G. (1967), In Methods in Enzymoloby 12A, L. Grossman and K. Moldave, eds. (New York: Academic Press), p. 182. Penman, S., Rosbash, M. and Penman, M. (1970), Proc. Nat. Acad. Sci. USA 31, 1878. Penman, S., Vesco, C. and Penman, M. (1968), J. Mol. Biol. 33, 49. Perry, R. P. and Kelley, D. E. (1969), J. Cell Physiol. 11. 235. Plagemann, P. G. (1971), Arch. Biochem. and Biophys. 144, 401. Reeder, R. H. and Roeder, R. G. (1972), J. Mol. Biol. 31, 433. Roeder, R. G., Reeder, R. H. and Brown, D. D. (1970), Cold Spring Harbor Symp. Quant. Biol. 33, 722. Roeder, R. G. and Rutter, W. J. (1969), Nature 224, 234. Roeder, R. G. and Rutter, W. J. (1970), Proc. Nat. Acad. Sci. USA 33, 675. Sarkar, P. K., Goldman, B. and Moscona, A. A. (1973), Biochem. Biophys. Res. Commun. 33, 308. 162 Seifart, K. H., Juhasz, P. P. and Benecke, B. J. (1973), Eur. J. Biochem. 33, 181. Siev, M., Weinberg, R. and Penman, S. (1969), J. Cell Biology 31, 510. Singer, R. H. and Penman, S. (1973), J. Mol. Biol. 8, 321. Towle, H. C., Jolly, J. F. and Boezi, J. A. (1975), J. Biol. Chem. 250, 1723. Weil, P. A. and Blatti, S. P. (1975), Biochemistry 13, 1636. Weinmann, R. and Roeder, R. G. (1974), Proc. Nat. Acad. Sci. USA 11, 1790. Willis, M. V., Baseman, J. B. and Amos, H. (1974), Cell 3, 179. Yoshikawa, M. (1967), Tetrahedron Letters 33, 5065. Yu, Fu-Li, and Feigelson, P. (1972), Proc. Nat. Acad. Sci. Zylber, E. A. and Penman, S. (1971), Proc. Nat. Acad. Sci. USA 33, 2861. er PEA“ I J'. I .III 1 III al- "’11111111111111“