SISTREBUFEDN 3F Z’HQMETE‘EYLNWLEGSiEES EN RNA Dissertation fer the Degree of Ph. D. MECHEGAR STATE fiNEVERSETY LEE M. ME 1974 itwuw‘“ i "b L :3 R A RY $1 Michigan .Statc Universxty 3'." @tev‘ _fl.’ This is to certify that the thesis entitled Distribution of 2'—Q-Methy1nucleosides in RNA presented by i Lee M. Pike has been accepted towards fulfillment of the requirements for Ph. D . degree in Biochemistry i ix: “(late a Major professor ABSTRACT DISTRIBUTION OF 2'-0-METHYLNUCLEOSIDES IN RNA BY Lee M. Pike In several classes of RNA there is a correlation of 2'-9: methylation with metabolic stability. The role of ribose-methylation in RNA has yet to be defined, but it has been shown to stabilize secondary structureznuipossibly decrease the nuclease susceptibility of certain synthetic polynucleotides. The degree of these effects may be influenced by the particular nucleotides which are modified as well as by the total abundance of methylation. For this reason, a study of the correlation between the occurrence of 2'-Q:methylation in natural RNA and its effect on structure or function should include a description of the extent to which each of the four nucleosides is modified. Although several procedures are available for the quantita- tive analysis of 2'-Q:methylation in RNA, none of them satisfy the requirements for multiple analytical determinations of the 2'-Q: methylnucleoside compositions of RNA samples, especially when using unlabeled RNA. In the course of this work, a rapid and sensitive procedure for determining the 2'-9:methylnucleoside composition of an RNA sample has been established. <7 ”o 7 ”13 t2???” Lee M. Pike The RNA to be analyzed is hydrolyzed to nucleosides through the combined action of bovine pancreatic ribonuclease, snake venom phOSphodiesterase, and bacterial alkaline phosphatase. The crude hydrolyzate is fractionated by DEAE-cellulose (borate) column chromatography. The 2'-9:methylnucleosides are collected at the void volume when the column is washed with boric acid while the ribo- nucleosides form a complex with borate and are retained. Nucleoside compositions are determined by resolving the 2'-9:methylnucleosides from each other on a high pressure liquid chromatography column. The method permits compositional analysis of a 5 mg sample of RNA provided 1-2% of the nucleosides are methylated on the ribose moiety. In this example, the 2'-9:methylnucleoside fraction isolated from the sample would be 0.1-0.3 umoles, a quantity sufficient for several independent determinations of its composition. The sensi- tivity of the determination may be significantly increased if the RNA samples are labeled with methyl-labeled methionine prior to analysis. A ribonucleoside composition may be obtained from the same sample if it is desired, since the liquid chromatographic compositional determination requires less than 1% of the original hydrolyzate. The sensitivity of the method is sufficient to distinguish two RNA samples which differ in their 2'-9:methylnucleoside composition by 2-3%. The accuracy of the procedure was verified by analysis of wheat germ ribosomal RNA which has been characterized for both ribo- nucleoside and 2'—9:methylnucleoside composition by Lane (1965). Since the values obtained with this method are in close agreement Lee M. Pike with those determined by Lane, it represents a reliable alternative to the more time consuming determinations used previously. The method has been used as the basis for structural comparisons between several groups of ribosomal and nuclear RNA's. DISTRIBUTION OF 2'-0-METHYLNUCLEOSIDES IN RNA BY Lee M. Pike A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1974 Dedicated to Mary, Jay, and Polly ii ACKNOWLEDGMENTS I thank Dr. Fritz Rottman for his guidance in my research efforts. His attention to the development of the scientist as a person as well as a researcher was instrumental in the successful completion of my graduate work. I appreciate the opportunity to have collaborated with Karen Friderici on certain aspects of this work. Our combined efforts have considerably broadened the scope of its application. I thank Marc Beversluis, an Undergraduate NSF Trainee during the summer, 1973, for his analysis of nuclear RNA from ethionine-treated Novikoff hepatoma ascites cells. I acknowledge the contributions from the members of my guidance committee, Dr. William Wells, Dr. Allan Morris, Dr. John Shaver, and Dr. Paul Kindel. In addition, Dr. Loran Bieber is thanked for his interest in me and my work. I am indebted to my colleagues, at both the graduate and post~graduate level, for helpful suggestions and discussions in the course of this work. I especially thank Mary for her love, faith and encouragement during my graduate training. The financial assistance from the National Institutes of Health and the Department of Biochemistry, Michigan State University is appreciated. iii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES LIST OF ABBREVIATIONS PART I: LITERATURE REVIEW . Natural occurrence of 2'-O-methylnucleosides Synthesis of 2'-0-methyl compounds . . . Biological effects of ribose-methylation . Physical effects of ribose-methylation Nucleolar ribosomal RNA precursors . Ribosomal RNA sequence variation . Low molecular weight nuclear RNA Quantitative asSays of 2'-Q:methylation Statement of the problem EXPERIMENTAL . Material . Wheat germ ribosomal RNA Wheat germ trinucleotides . . Preparation of rat tissues . . Novikoff hepatoma ascites cells . Rat liver ribosome fractionation . Rat brain, kidney, and testes cytoplasmic plus membrane-bound ribosomes . . . . . Novikoff cell ribosomes Rat liver nuclei . Novikoff cell nuclei Ribosomal RNA extraction . . . . . . Nuclear RNA extraction . . . . . . . . Salt fractionation of nuclear RNA Glycerol density gradient fractionation of nuclear RNA Cetyltrimethylammonium bromide precipitation of RNA Polyacrylamide gel electrophoresis . Ultraviolet Spectral analysis iv Page vi vii viii 22 22 22 23 24 24 25 26 27 27 29 30 31 32 33 34 34 36 Page Enzymatic hydrolysis conditions . . . . . . . . . . 36 Isolation of 2'-0-methylnucleosides . . . . . . . . 37 Gas chromatography of nucleosides . . . . . . . . . 38 Mass spectral analysis . . . . . . . . . . . . . 39 Liquid chromatography of nucleosides . . . . . . . . 39 Ethionine treated Novikoff cells . . . . . . . . . 41 RESULTS . . . . . . . . . . . . . . . . . . . 44 Wheat germ ribosomal RNA characterization . . . . . . 44 Tissue ribosomal RNA characterization . . . . . . . . 44 Nuclear RNA characterization . . . . . . . . . . . 46 Salt fractionation of nuclear RNA . . . . . . 57 Glycerol density gradient fractionation of nuclear RNA . . 57 Enzymatic hydrolysis of RNA . . . . . . . . . 58 Enzymatic hydrolysis of alkaline stable trinucleotides from wheat germ ribosomal RNA . . . . . . . 64 Isolation of the 2'-0-methy1nuc1eoside fraction of RNA . . 69 Gas chromatography of'nucleosides . . . . . . 71 Gas chromatOgraphy- -mass spectrometry of nucleosides . . . 75 Direct probe mass spectrometry of nucleosides . . . . . 75 Liquid chromatography of nucleosides . . . . . . . 79 Nucleoside composition of wheat germ ribosomal RNA . . . 85 Nucleoside composition of Novikoff cell and rat liver ribosomal RNA . . . . . . . . . . . . . 90 Nucleoside composition of rat tissue ribosomal RNA . . . 92 Nucleoside composition of Novikoff cell and rat liver high molecular weight nuclear RNA . . . . . . 92 Nucleoside composition of nuclear RNA from ethionine treated Novikoff cells . . . . . . . . . . . . 96 DISCUSSION . . . . . . . . . . . . . . . . . . 98 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 108 PART II: ALKYLATION OF ADENOSINE BY DIAZOETHANE IN AQUEOUS 1,2-DIMETHOXYETHANE: SOLVENT PARTICIPATION IMPLICATED BY MASS SPECTRAL ANALYSIS . . . . . . 116 Table 10. LIST OF TABLES Yield of ribosomal RNA from various tissues . Recovery of 2'-9:methylnucleosides from a DEAE— cellulose (borate) column . . . Gas chromatography of TMS-nucleosides . Unique ions occurring in the mass spectra of 2'—Q: methylnucleosides . . . . . Mass spectral response ratios for a mixture of 2'-9:methylnucleosides . . . . . . . . . Relative elution volumes of nucleic acid components during liquid chromatography with Buffer A ' Relative elution volumes of nucleic acid components during liquid chromatography with Buffer B Nucleoside composition of wheat germ ribosomal RNA Nucleoside compositions of Novikoff cell and rat tissue ribosomal RNA . . . . . . . . Nucleoside compositions of Novikoff cell and rat liver nuclear RNA . vi Page 45 7O 72 76 78 81 84 91 93 9S LIST OF FIGURES Figure Page 1. Processing of 458 nucleolar RNA (Weinberg and Penman, 1970) . . . . . . . . . . . . . . 10 2. UV scan of Novikoff cell ribosomal RNA . . . . . . 48 3. Polyacrylamide-agarose gel electrOphoresis scan of Novikoff cell ribosomal RNA . . . . . . . . . 50 4. Polyacrylamide-agarose gel electrophoresis scans of Novikoff cell and rat liver total nuclear RNA . . . 53 S. Polyacrylamide gel electrophoresis scans of Novikoff cell and rat liver low molecular weight nuclear RNA . 56 6. Glycerol density gradient centrifugation of Novikoff cell nuclear RNA . . . . . . . . . . 60 7. Polyacrylamide-agarose gel electrOphoresis scans of fractionated Novikoff cell nuclear RNA . . . . . 62 8. Paper electrophoresis of the enzymatic hydrolysis products of wheat germ ribosomal RNA . . . . . . 66 9. Enzymatic hydrolysis of alkaline stable trinucleotides from wheat germ ribosomal RNA . . . . . . . . 68 10. Effect of pH on the relative elution volumes of 2'-Q7methylnucleosides in liquid chromatography . . 83 11. Liquid chromatographic elution profile of ribo— nucleosides from hydrolysis of wheat germ ribosomal RNA . . . . . . . . . . . . . . 87 12. Liquid chromatographic elution profile of 2'-0- methylnucleosides from hydrolysis of wheat germ ribosomal RNA . . . . . . . . . . . . . . 89 vii LIST OF ABBREVIATIONS A adenosine Am 2'—9:methyladenosine aM molar extinction coefficient AmDP 2'-Q:methy1adenosine-5'-diphosphate AMU atomic mass unit A260 unit the quantity of material contained in 1 ml of a solution which has an absorbance of 1 at 260 nm, when measured in a 1 cm light path. BSA bis (trimethylsilyl) acetamide Buffer A 0.4 M_ammonium formate, pH 4.55 Buffer B 0.4 M_ammonium formate, pH 4.45, 40% ethylene glycol C cytidine Cm 2'—9;methylcytidine CTAB cetyltrimethylammonium bromide dA 2'-deoxyadenosine dC 2'-deoxycytidine DEAE diethylaminoethyl dG 2'-deoxyguanosine DNA deoxyribonucleic acid EDTA ethylenediaminetetraacetic acid viii G guanosine GC gas chromatography Gm 2'—9:methylguanosine HMW high molecular weight I inosine Im 2'—9:methylinosine LC liquid chromatography LMW low molecular weight M0 methoxyamine MS mass spectroscopy nRNA nuclear RNA oligo (dG) oligonucleotide of 2'-deoxyguanylic acid poly (A) homopolymer of adenylic acid (homopolymers of other nucleotides are represented in an analogous manner) RNA ribonucleic acid rRNA ribosomal RNA RSB Buffer 10 mM_NaC1, 15 mM_MgC12, 10 mM_Tris-HC1, pH 7.4 S Svedberg unit SDS sodium dodecyl sulfate T thymidine TEAB triethylammonium bicarbonate TKM Buffer 10 mM_Tris-HC1, pH 7.4, 10 mM_KCl, 2 mM MgCl2 TMS trimethylsilyl Tris tris (hydroxymethyl) aminomethane U uridine ix Um UmDP ultraviolet 2'-0-methyluridine 2'—9:methyluridine-5'-diphosphate PART 1: LITERATURE REVIEW Natural occurrence of 2'-0-methylnucleosides. The presence of an alkaline stable fraction in RNA samples was noted as early as 1951 by Magasanik and Chargaff. Reports of a similar fraction amount- ing to approximately 3% of yeast RNA were made by Smith and Allen (1953) and Crestfield gt_al: (1955). The nature of this material remained undetermined until Smith and Dunn (1959) reported the presence of 2'-g:methylribose in alkaline stable fractions of the RNA isolated from wheat germ, rat liver, and tobacco and beet leaves. The occurrence of a methoxy group on the 2'-carbon of some nucleosides resulted in the resistance of the adjacent internucleotide linkage to alkali due to interference with the formation of 2',3'-cyclic phosphate intermediates through which hydrolysis normally proceeds (Brown and Todd, 1955). Within a short period of time the 2'-9:methyl modification of RNA was shown to occur on all four of the major nucleosides. Biswas and Meyers (1960) reported the isolation of Cm from a preparation of cytidine obtained by acid hydrolysis of total RNA from the blue—green algae Anacystis nidulans. Cm and Gm were recovered from yeast transfer RNA and rat liver RNA by Morisawa and Chargaff (1963), and all four 2'-Q7methylnucleosides were isolated from yeast transfer RNA by Hall (1963). The widespread occurrence of ribose methylation became obvious over the next few years when several workers began more intensive studies on its distribution. Hall (1964) demonstrated the presence of significant levels of 2'-9:methylnucleosides in yeast, Esherichia coli, and several mammalian RNA's. Lane and his co-workers studied the occurrence of alkaline stable oligonucleotides in wheat germ (Singh and Lane, 1964), yeast (Gray and Lane, 1967), Esherichia ggli_(Nichols and Lane, 1967), and mouse L-cells (Lane and Tamaoki, 1967). Until recently methylation of specific nucleosides in RNA had been demonstrated only in ribosomal and transfer RNA and in a group of low molecular weight RNA molecules which are usually confined to the nucleus (Starr and Sells, 1969; Hall, 1971; Abbate, 1971). Messenger RNA from both eukaryotes and prokaryotes had been assumed to be free of methylated nucleosides since Moore (1966) demonstrated that bacterial and viral messenger RNA contain less than one methyl group per 3500 nucleotides. This assumption was apparently supported by the observation that heterogeneous nuclear RNA, the presumed pre- cursor to messenger RNA in eukaryotic cells, was not detectable labeled by methyl-labeled methionine (Perry g£_al:, 1970; Johnson, 1970). However, direct analysis of messenger RNA following incubation of mouse L-cells with methyl-labeled methionine has recently led to the conclusion that messenger RNA is methylated on both the base and ribose moieties to the extent of about 2 to 3 methyl groups per 1000 nucleotides (Perry and Kelley, 1974). In contrast to the earlier reports, the authors also detected a very low level of methylation in heterogeneous nuclear RNA. There are some RNA molecules which do not appear to contain any methylated nucleosides. The relatively short chain RNA molecules associated with the structural ribosome are unmodified. The SS ribosomal RNA molecules from bacterial (Brownlee et_al:, 1967), yeast (Hindley and Page, 1972), Xenopus (Brownlee et_al:, 1972; Ford and Southern, 1973), and several mammalian sources (Forget and Weissman, 1967; Williamson and Brownlee, 1969; Labrie and Sanger, 1969; Averner and Pace, 1972) have been sequenced and do not contain any modified nucleosides. In addition, neither the 7S ribosomal RNA from HeLa cell (Weinberg, 1973) nor the analogous 5.68 yeast ribosomal RNA (Rubin, 1973) is methylated. Synthesis of 2'-0-methyl compounds. The chemical synthesis of 2'-9:methylnucleosides served three purposes: a) structural confirmation of the naturally occurring modified nucleosides, b) a convenient source of standards for identification of methylated nucleosides isolated in small quantities from RNA samples, and c) a source of substrate for the synthesis of synthetic polynucleotides containing 2'-Q:methylnucleosides. The first direct synthesis of a 2'-Q:methy1nucleoside was performed by Brown and Robins (1965). They treated adenosine with diazomethane in an aqueous solution of 1,2-dimethoxyethane and discovered that the 2'-hydroxyl reacted preferentially to give 2'-9:methy1adenosine in 40% yield. Gin and .Dekker (1968) further purified and characterized the products of this reaction with adenosine and also reported a yield of about 140% for 2'-Q:methyladenosine. Furukawa gt_al: (1965) prepared 2'-Q:methyluridine and 2'-9:methylcytidine using the 3',S'—ditritylated ribonucleosides in the presence of silver oxide with methyl iodide as the methyl donor. The procedure is somewhat laborious in that it involved separation of N?- from Qfi-ring-methylated products and subsequent chemical transformations. Later, Martin et_al: (1968) treated cytidine directly with diazomethane in aqueous 1,2-dimethoxyethane to obtain 2'-Q:methylcytidine in 14% yield. In addition, 2'-9:methyluridine was prepared by deaminating the 2'-9:methylcytidine with potassium nitrite in acetic acid. Treatment of guanosine with diazomethane leads to preferential methylation of the heterocyclic rings of the base moiety. Khwaja and Robins (1966) were able to circumvent this difficulty by methylating 2-amino-6-chloropurine riboside with diazomethane. The 2'-hydroxyl of this compound reacted preferentially; the major product was isolated and converted to 2'-Q:methylguanosine by a series of chemical modifications. A modified version of the Broom and Robins method using diazo- methane and catalytic amounts of stannous chloride dihydrate had been used to prepare the 2'507methy1 derivatives of cytidine and adenosine (Robins and Naik, 1971). This procedure resulted in greatly improved yields of 2'-0-methylcytidine (74% of the starting material). The yield of 2'-0-methyladenosine remained about 40%, but the balance of the adenosine was quantitatively converted to 3'-Q:methyladenosine, a significant increase in the yield of this desirable side product. Kusmierek, Giziewicz, and Shugar (1973) reported the prepara- tion of 2'-_Q_-alkylcytidine (-methyl or -ethyl) in 12% yield by treatment of cytidine with dialkylsulfate in a strongly alkaline medium. Although the yield of Cm was not increased relative to the method of Martin e£_al, (1968), the yield of potentially useful side products was enhanced with this procedure. Subsequent bisulfite catalyzed deaminations of the products gave the corresponding deriva- tives for uridine. However, the method is limited to alkylation of cytosine nucleosides or nucleotides (Kusmierek, Kielanowska, and Shugar, 1973) at the present time. Direct synthesis of 2'-0:alkyl- nucleotides (—methyl or -ethyl) was reported by Tazawa et_31: (1972) using an alkyl iodide at alkaline pH in the presence of dimethyl- formamide. The 3':S'-cyclic nucleotides of adenosine and cytidine were successfully alkylated in 40 to 50% yield. The unexpected discovery of Rottman and Heinlein (1968) that polynucleotide phOSphorylase would incorporate 2'-9:methy1adenosine diphosphate (AmDP) into poly Am made possible the synthesis of several polymers containing 2'-9:methylnucleosides. The substrate for this reaction was prepared by chemically coverting Am to the 5'-mononucleo- tide and raising it to the diphOSphate level with rabbit muscle myokinase. The synthesis of poly Cm (Zmudzka g£_gl;, 1969) utilized similar conditions except that manganese was substituted for magnesium in the reaction mixture. Poly Um was prepared by deamination of poly Cm or by incorporation of UmDP using polynucleotide phosphorylase (Zmudzka and Shugar, 1970). Heteropolymers containing ribonucleotides and 2'-Q:methylnucleotides in varying proportions have also been ;prepared with similar techniques (Rottman and Johnson, 1969). Biological effects of ribose-methylation. Considerable effort has been made to elucidate the functional role of 2'-Q:methylation in RNA. The availability of homOpolymers or hetero- polymers containing 2'-9:methylnucleotides has made it possible to examine the effect of this modification on the biological and physical properties of RNA. For example, sugar-methylation has been shown to have little effect on the binding of lysyl-tRNA to ribosomes mediated by oligoadenylates (Price and Rottman, 1970). In addition, it was found that polymers only partially methylated on the 2'-hydroxyl directed the incorporation of amino acids into protein in a bacterial cell-free amino acid incorporating system (Dunlap e£_al:, 1971). Homopolymers of 2'-9:methylnucleotides were found to have no template activity under similar conditions (Dunlap et_al:, 1971), but the increased secondary structure of these polymers may have been respon- sible for this effect (see discussion below). Sugar methylation does not interfere with the in_vi££g_trans- cription of polypyrimidines by Pseudomonas putida RNA polymerase in either the single-stranded or duplex form, but it does interfere with transcription of poly(A) in either form (Gerard e£_gl:, 1972). The inability of poly (Am) to serve as a template was not explained, but the authors ruled out an anomaly in the secondary structure of the molecule as the cause. Interestingly, in the DNA—directed synthesis of RNA, the polymerase can utilize 2'-9:methyladenosine-5'- triphosphate as a substrate to a limited extent (Gerard §£_al:, 1971). Several potentially useful observations have been made as the result of specific attributes of 2'-9:methylnucleoside containing polymers. One is the use of poly (Cm)°oligo (dG) as a template for RNA directed DNA polymerase activity (Gerard s£_§l:, 1974). This template was found to be capable of distinguishing reverse transcriptase activity from several different DNA-dependent DNA polymerase activi- ties. Degradation studies using 2'-Q:methylnucleoside containing heteropolymers have indicated that methylated nucleotides may confer nuclease resistance to a polymer (Dunlap et_al:, 1971; Gerard et_al., 1974). Nuclease resistance is a desirable factor in designing a synthetic polynucleotide capable of antiviral activity through induc- tion of interferon (DeClercq g£_al:, 1971). First reports indicated that homopolymers of 2'—0:methylnucleotides had no such activity (DeClercq st 21., 1972), but subsequent investigations with partially methylated heteropolymers have provided more positive results (Merigan and Rottman, 1973). High levels of both poly (Am) and poly (A) inhibit the uptake of oncornavirus in a manner similar to other polyanions (Tennant g£_gl:, 1973). Low levels of poly (Am) don't inhibit uptake of oncor- navirus, but do inhibit replication and transformation of tissue cultured cells by oncornavirus. Poly (A) was ineffective at compar- able concentrations. Since low levels of poly (Am) did not have this effect on other RNA viruses which do not contain reverse transcriptase, an essential role for this enzyme is suggested in replication of oncornavirus as well as transformation of cells. Physical effects of ribose-methylation. The influence of 2'-0:methylation of polynucleotides on their structure has been investigated by UV and circular dichromism Spectroscopy. Early work had suggested that the 2'-hydroxyl group may participate in stabil- ization of ordered structures of polyribonucleotides by intramolecular hydrogen bonding with nearby groups in the polymer chain (Brahms g£_gl:, 1967; Ts'o §£_al:, 1966). Surprisingly, the physical charac— teristics of poly (Am) indicated that it had more secondary structure than the analogous poly (A) (Bobst, Cerutti, and Rottman, 1969). It was further shown that an unsubstituted 2'-hydroxy1 group in poly (A) was not essential for the stabilization of the double-stranded conformations of poly (A) in acidic solutions or the double-stranded duplex of poly (A)'poly (U) (Bobst, Rottman, and Cerutti, 1969a). Increased secondary structure was also reported for poly (Cm) (Zmudzka g£_al:, 1969) and poly (Um) (Zmudzka and Shugar, 1970) rela- tive to their polyribonucleotide counterparts. No such effect was detected when the 2'-hydroxyl was replaced by hydrOgen (Zmudzka and Shugar, 1970), fluorine (Janik £3 31:, 1972), chlorine (Hobbs 33 51:, 1971), or amine (Hobbs e£_al:, 1972). These results seemed to support the pr0posa1 that the important requirement for increased stability of polyribonucleotides relative to polydeoxynucleotides was the presence of an oxygen atom in the 2'-position of the sugar (Bobst, Rottman, and Cerutti, 1969b). Upon finding that 2'-Q:ethylation further increased secondary structure for a given polynucleotide (Khan and Rottman, 1972; Kusmierek, Kielanowska, and Shugar, 1973) the hypothesis was advanced that steric factors may account for some of the observed differences. The steric role seems to be supported by the report of increased secondary structure in poly (2'-azido-2'-deoxyuridylic acid) which lacks a 2'-oxygen (Torrence g£_gl:, 1973). Using both methylated and ethylated polymers, Rottman g£_al. (1974) and Khan gt_al, (1974) observed that the degree of secondary . - .11 '6‘ he» a... v'u I 1'. 9: ‘C u:._‘ “A! l structure was influenced by the particular nucleotide which was modified. For this reason, a correlation between the occurrence of 2'-Q:methylation and its effect on structure or function may depend on the extent to WhiCh each 0f the four nucleotides are modified in addition to the total number of modifications. Nucleolar ribosomal RNA precursors. Research over the last decade has made clear that the nucleolus is the site of the bio- synthesis of ribosomal RNA. Scherrer and Darnell (1962) found that in HeLa cells, nucleotides are first incorporated into 45S RNA, which is subsequently processed into cytoplasmic ribosomal RNA through a series of intermediates in the nucleolus (Perry, 1962). The pro- cessing scheme for nucleolar ribosomal RNA precursors in HeLa cells (Weinberg and Penman, 1970) is presented in Figure 1. Similar schemes have been derived for yeast (Retel and Planta, 1967) and Novikoff hepatoma ascites cells (Quagliarotti g£_al:, 1970) with some varia- tion in the sizes of the precursors or intermediates. The general pattern seems to be the same for other eukaryotic organisms as well (Rubenstein and Clever, 1971). One point of Special interest has been the role of methylation in the processing. Kinetic studies with [14C-methyl]-methionine labeling of HeLa cells (Greenberg and Penman, 1966; Zimmerman and Holler, 1967) showed that 458 RNA was methylated simultaneously or immediately following its synthesis in the nucleolus. Further, the methyl label could be chased through the intermediates and into cytoplasmic ribosomal RNA. The amount of label incorporated indicated 10 A A o v a .flonmfl .cmecom ecu whencfiozv oom um unmwu on umoa Eoem we: :oflumemfle one ooflHmmm mm: oz mo ceom mamouonmonuooao How omonmwmuoofiemflxuomxfiom .m chewed 50 m ounwflm mm to map was 51 agreement with the reported value of 73 pg per g of tissue (Hodnett and Busch, 1968). Recovery of Novikoff cell nuclear RNA averaged 420 pg per g of cells (wet weight), but this represented only about 65% of the nuclear RNA which the Novikoff cell contained (Moriyama gt_al,, 1969). This discrepancy may reflect loss of some Novikoff cell nuclei due to mechanical breakage during the extended homogeni- zation required to completely disrupt the Novikoff cells (Busch et_al., 1972). The UV spectral data obtained from the nuclear RNA's were similar to the data reported above for ribosomal RNA's. Samples of Novikoff cell and rat liver nuclear RNA were examined by 2% poly- acrylamide, 0.5% agarose gel electrophoresis as shown in Figure 4A and 4B, resPectively. The high molecular weight ribosomal RNA precursor bands were well resolved and discrete bands were observed for 458, 358, 288, and 18S (nomenclature after Choi and Busch, 1970). The proportions of these precursor molecules were relatively constant in all preparations from either tissue, but there were differences between the two tissues. Novikoff cell nuclear RNA consistently contained higher levels of 458 and 35S RNA than did rat liver, perhaps reflecting differences in the metabolic state of the tissues with respect to ribosomal RNA synthesis. The amount of 188 RNA found in rat liver nuclear RNA was higher than expected since this species is normally transported out of the nucleus soon after the 45S precursor molecule is cleaved (Weinberg gt_al:, 1967); Grierson et_al:, 1970). Possible cytoplasmic contamination was probably minimal, however, since the ratio of 52 Figure 4. Polyacrylamide-agarose gel electrophoresis scans of Novikoff cell and rat liver total nuclear RNA. The gel consisted of 2% polyacrylamide, 0.5% agarose buffered with Peacock's Buffer, pH 8.3; 0.30 A260 unit RNA was applied for both samples and migration was from left to right at 200v, 0°C for 2 hr 15 min; A) Novikoff cell and B) rat liver. 54 358 RNA to 28$ RNA was relatively high. Conclusive information concerning the relative amounts of individual Species of RNA could not be obtained from the relative sizes of gel bands since the bands did not always stain quantitatively. Samples of nuclear RNA were further examined by 10% poly- acrylamide gel electrophoresis to further resolve the low molecular weight species as shown for Novikoff cell in Figure 5A and for rat liver in Figure 5B. The nomenclature used for the low molecular weight nuclear RNA was established by Busch and his colleagues (Prestaykogt £11., 1970; Busch 3E. _a_1_., 1971). The gel patterns were qualitatively similar for the two tissues and with one exception the same bands occurred on both gel scans. The exception is in the 88 region, where two bands were seen for rat liver nuclear RNA and only one was present for Novikoff cell nuclear RNA. It should be noted that whereas all other bands were closely correlated in location on the two gels; neither rat liver 88 band migrated with the Novikoff cell 88 RNA band. The 88 bands occurred to the same relative extent in all nuclear RNA gel scans. Consistent quantitative differences between Novikoff cell and rat liver were seen in the proportions of Ulb, 58, and 4S RNA relative to the other low molecular weight nuclear RNA bands. These differences may reflect the difficulty of obtaining quantitative staining of the bands in a gel. Further studies are in progress concerning the distribution of low molecular weight nuclear RNA in the two tissues. Figure 5. Polyacrylamide gel electrOphoresiS scans of Novikoff cell and rat liver low molecular weight nuclear RNA. The gel consisted of 10% polyacrylamide buffered with Peacock's Buffer, pH 8.3; 0.30 A260 unit RNA was applied for both samples and migration was from left to right at 200v, 0°C for 7 hr; A) Novikoff cell and B) rat liver. 57 Salt fractionation of nuclear RNA. Fractionation of nuclear RNA preparations was undertaken in order to obtain the high molecular weight ribosomal RNA precursors as a group for subsequent nucleoside composition analysis. Sodium acetate fractionation of the RNA was attempted initially at 2 M salt concentration as described in the Experimental section. The salt fractions were assayed by 2% poly- acrylamide, 0.5% agarose gel electrophoresis to check for separation of high molecular weight RNA from low molecular weight RNA. The supernatant from the initial precipitation (LMW I) contained 17% of the RNA and included most of the 4S RNA, some 4.58 RNA, and no larger Species of RNA. The pellet remaining after two precipitations from 2 M_sodium acetate (HMW II) contained the ribosomal precursor bands, but all low molecular weight RNA species were present as well. Further fractionation of HMW II was then per- formed with 0.5 M_sodium acetate. The supernatant (LMW III) contained an additional 6% of the original sample. All Species of low molecular weight RNA were present while no larger RNA Species were represented. However, all size classes of the low molecular weight RNA were repre- sented in the precipitate (HMW III) as well. Since the low molecular weight RNA did not quantitatively separate from the high molecular weight RNA under these conditions, salt fractionation could not be used to prepare either fraction. Glycerol density gradient fractionation of nuclear RNA. Glycerol density gradient centrifugation was successfully employed fer fractionation of nuclear RNA samples. Linear gradients of 58 10-30% glycerol were used on an analytical or a preparative scale. The sedimentation profile of a preparative gradient of Novikoff cell nuclear RNA is Shown in Figure 6. The fractions were pooled as described in the legend and RNA was recovered as the cetyltrimethyl- ammonium salt. The high molecular weight fraction comprised about 70% of Novikoff cell nuclear RNA preparations and about 50% of rat liver nuclear RNA preparations. Separation of the low molecular weight RNA from high molecular weight RNA was complete as shown by 2% polyacrylamide, 0.5% agarose gel electrophoresis. Gel patterns for the high molecular weight and low molecular weight RNA fractions prepared from Novikoff cell nuclear RNA are presented in Figure 7A and 78, respectively. The scans indicate that all bands present on the total nuclear RNA gel (Figure 3A) were present in the fractionated RNA, and there was little cross contamination of either fraction. Further analysis of the low molecular weight fraction was performed by 10% polyacrylamide gel electrOphoresis to insure that uniform recovery of low molecular weight RNA had been achieved, since discrete species of low molecular weight nuclear RNA were difficult to identify on low percentage gels. All low molecular weight bands were present in the same proportions as in total nuclear RNA (Figure 5A) (data not shown). Enzymatic hydrolysis of RNA. Hydrolysis of RNA to the nucleo- side level was achieved with the combination of enzymes described by Hall (1964). Snake venom phosphodiesterase was used for production of 5'-mononucleotides since it attacks all nucleic acids, regardless S9 .noz mo cowummsmflnucoo ucofiomuw xpfimcoo Honooxao .o chewed 6O okuop c ohswwm .02 20:U(¢u ON econ was man“ flv nfiv 0; W909! iDNVIIOSIV 61 Figure 7. Polyacrylamide-agarose gel electrophoresis scans of fractionated Novikoff cell nuclear RNA. The gel consisted of 2% polyacrylamide, 0.5% agarose buffered with Peacock's Buffer, pH 8.3; 0.20 A260 unit was applied for both samples and migration was from left to right at 200v, 0°C for 2 hr 15 min; A) high molecular weight RNA and B) low molecular weight RNA. 62 288 355 4S vFigure 7 63 of the sugar moiety (Laskowski, 1971). The lack of sugar specificity was Significant since it was important that all phosphodiester link- ages involving 2'-0:methy1nucleosides were cleaved. Pancreatic ribonuclease was added to produce oligonucleotides since they are a better substrate for phosphodiesterase than intact RNA (Laskowski, 1966). Nucleotides were dephosphorylated with alkaline phosphatase. Precipitation of magnesium from the hydrolysis mixture pre- sented a problem in establishing reaction conditions since phospho- diesterase has a magnesium requirement (Butler, 1955). The upper limit of magnesium phosphate solubility is 3.6 mM and 10 mM_phOSphate was generated by hydrolysis of nucleotides. The magnesium level was lowered to 2 mM_to minimize precipitation. Ammonium formate was used to buffer the reaction when it was discovered that ammonium carbonate removed virtually all the magnesium through precipitation of a complex involving magnesium, carbonate, and phOSphate. Optimum concentrations of alkaline phOSphatase, snake venom phosphodiesterase, and ammonium formate were arrived at empirically by following the course of several reaction mixtures by paper electrophoresis. Reaction mixtures were routinely assayed prior to further analysis to insure that hydrolysis was complete. Both paper electro- phoresis and inorganic phosphate release were used to follow the progress of hydrolysis; of the two, paper electrOphoresis proved the more useful. When inorganic phOSphate release was assayed, precipi- tation of phosphate from the reaction mixture caused a decrease in assayable phOSphate after 4.5 hr of incubation. Paper electro- phoresis of the reaction products indicated incomplete hydrolysis 64 by the occurrence of UV absorbing material in the oligonucleotide region of an electrOphoreogram as shown diagrammatically in Figure 8A. Within the limits of detection, the hydrolysis was judged complete by the presence of all detectable UV absorbing material (greater than 97%) near the origin (Figure 8B). Enzymatic hydrolysis of alkaline stable trinucleotides from wheat germ ribosomal RNA. Since Gray and Lane (1967) reported that the dinucleoside monOphosphate AmpA is more stable to the action of snake venom phosphodiesterase than the ribonucleoside analogue ApA, it was necessary to demonstrate that the reaction conditions employed for hydrolysis of our RNA samples were adequate to completely hydrolyze any phosphodiester linkages involving 2'-9:methylnucleo- sides. Accordingly, hydrolysis of a sample of alkaline Stable 2'-0:methylated trinucleotides isolated from wheat germ ribosomal RNA was used to test our reaction conditions. Duplicate samples of the trinucleotide were subjected to enzymatic hydrolysis as shown in Figure 9. Hydrolysis to the nucleoside level was greater than 92% complete in 32 hr at 37°C by the criterion described in the Experimental section. There was no visible UV absorbing material left in the oligonucleotide region of the electrophoreogram after 32 hr, but a small amount of fluorescent material was present. Analysis of the UV spectrum of the fluorescent material indicated that it was not characteristic of nucleic acid components. Therefore, the hydrolysis of essentially all phospho- diester linkages involving 2'-9:methylnucleosides was assumed to be 65 .mwmxfionnx: ouonEoo mm ocm mwmxaonw»: Hmfiupmm n< “cofluaHOmnm >3 so oopoouoe ones muomm one .cfie me How >oov um .m.n :a .Humumfieb.m mo.o ca woeuomeom mm: mfimonogdonuooam .moefiu ooumoflocfi one we comma H .02 cmEumgz co coupomm one: come pea: oom< H xfioumefixoemmm mcflcflmucoo muosvflam ampflflopofle cop “cofluoom Hensosfinodxm on» :H confinomon mm mm: enouxfle :ofiuomoa och .m douuumm .< ovuuoofiuoz-.m mouuoofioscowwao ovfiooofiooz 67 .cofluoom Hmucoefiuodxm can an voowuomoo m:oauwocoo ecu hoods meowpmcfienouoo opmowddsv sow 03Hm> ommuo>m ecu mucomoumen venom poem .< .onma ..He no fiuuoumflmmso ecu .Hnma .nomsm ecu .fiogu .wzwmm so mongoose moSHw> mo ommho>oU 69¢?»me enamocw modam>d H.H~ o.ms m.o~ o.m~ mufl use 4.0 n m.n~ H.o n s.o~ s.o n s.m~ H.o n H.om emoEooonuo some - oz so a: so e< 0.0m A.RH m.em v.5H m nanosecond - nfi use m.o n m.Hm A.o n N.RH m.o n 0.5m s.o n n.5H emosomosnn wagon - uu use ~.H n H.Hm s.o n s.mH o.H n o.~n 5.0 n o.w~ emosooosun nose - nu use ~.H n m.o~ H.o n ”.ms ~.H n w.mm H.o n m.aH oossa> onsoanoonu - oz N.o n m.om H.o n o.wH s.o n s.nm H.o n o.es seesaw osooono - oz o a o < e.oz mo mcofiufimomaou commooaonz .m canoe 94 centrifugation into high and low molecular weight fractions as described earlier. The high molecular weight fractions (greater than 188) consisted of primarily, but not exclusively, nucleolar pre- cursors to cytOplasmic ribosomal RNA as seen from the agarose gel electrophoresis scan of this fraction (Figure 7A). These fractions were shown to differ in both ribonucleoside and 2'-0-methylnucleoside composition when rat liver was compared to Novikoff cell (Table 10). Two observations are helpful in explaining these results which may have been unexpected a_priori in light of the similarity in nucleoside composition observed between the respective cytoplasmic ribosomal RNA's. First, in Novikoff cells approximately one-half of the nucleolar 458 ribosomal RNA precursor is lost during its pro- cessing into 188 and 288 RNA (Quagliarotti g£_§l:, 1970). Second, the distribution of Novikoff high molecular weight nuclear RNA in various size classes differs from that of rat liver (Figure 3). Quagliarotti g£_al: (1970) have shown that the ribonucleoside composition of the conserved sequences differs considerably from the composition of the sequences which are lost. Further, Choi and Busch (1970) found that approximately 16 2'-Q:methylnucleosides are lost during the processing events and indicated that the losses alter the 2'-Q:methylnuc1eoside content of the products in relation to the precursor. Similar losses presumably occur during production of liver ribosomal RNA, thereby affecting its precursor-product relationship. Since the precursors differ in nucleoside composi- tion, the compositional differences observed between Novikoff cell 95 .cofiumcweuouoo oncfim .onmfl .gomsm can mono so poppedon mosfim> n v .mcoflumcwauouoe N we oweoo>oe eumvcmum menace“ monam>m m.mH m.aH m.mm n.mm eozflm> oasumnouflq mm AH. mm om omHHou common» ocficoflnum em ma mm on nmafioo Honucou oz N.mN w.o~ w.om «.mm nwa pom H.H h m.HN N.o u w.oa m.H h m.om m.o a N.Hm ooz EU 5: Eu a< m.m~ w.mH m.~m m.wH nfia Hem m.~m 0.5H H.em w.m~ poz m.wH won one Haoo mmoxfi>oz mo mcowumeQEoo commooaosz .oH oHneH 96 and rat liver high molecular weight nuclear RNA may be explained by the distributional differences of the precursors in the two fractions. Nucleoside composition of nuclear RNA from ethionine treated Novikoff cells. Beaud and Hayes (1971a and b) observed a slow growth rate for Escherichia coli EAl (met-, bio-, RGStr) grown in the absence of methionine but in the presence of ethionine. The ethionine treated bacteria contained ribosomes whose RNA was sub- methylated, indicating an abnormality in the post-transcriptional modification processes. Villa-Trevino g£_al: (1966) found that ethionine induces a marked inhibition of rat liver RNA synthesis. Steward and Farber (1968) later showed that this inhibition was not due to a rapid acceleration of nuclear RNA breakdown. The effect of ethionine on rat liver RNA synthesis may be Similar to that observed in E: spli_with respect to post transcriptional modification (i.e. reduced methylation), but since eukaryotic RNA apparently must be modified before it can be processed (Vaughan SE 31., 1967), there might be a decrease in the overall rate of production of mature RNA. The extent of methylation in nuclear RNA from Novikoff cells was examined in the presence or absence of ethionine using [3H-methyl]- L-methionine as a methyl donor. In the presence of ethionine the incorporation of methyl label into nuclear RNA was reduced to about 42% of the level of control cells suggesting that ethionine was somehow altering methylation of nuclear RNA. On the other hand, 3H-uridine was incorporated into Nbvikoff cell nuclear RNA under the same conditions at 47% of control cell levels. 97 The methionine labeled nuclear RNA from ethionine treated cells was further examined by 2'-0-methylnucleoside composition analysis to determine if evidence for abnormal modification distribu- tions could be detected. The RNA from both ethionine treated and control cells was enzymatically hydrolyzed to nucleosides and applied to a DEAE-cellulose (borate) column. For both samples over 85% of the label was recovered in the 2'-9:methylnucleoside fraction, estab- lishing that methylation was predominantly occurring on the ribose moiety. The 2'-0:methylnucleoside compositions for both samples are presented in Table 10. The composition of nuclear RNA from ethionine treated cells was identical to the composition of nuclear RNA from control cells, even though the total methylation was only about 40% of the control level. The composition similarity is consistent with complete methylation of a ribosomal precursor molecule once modifica- tion has been initiated, even though the overall rate of modification might be reduced. The 2'-0:methylnucleoside composition of Novikoff cell high molecular weight nuclear RNA (greater than 188) as deter- mined by Choi and Busch (1970) is also presented in Table 10. The composition was determined for RNA samples isolated from Novikoff cells labeled with [14C-methyl]-L-methionine for 12 hr ip_!iyg: The results differ considerably from those determined above for a 30 min labeling period in tissue culture. DISCUSSION In preparation of RNA samples for compositional analysis, it is important that they be free of DNA. If the RNA sample being analyzed was initially contaminated with trace amounts of DNA, the deoxynucleosides resulting from complete hydrolysis of the sample would have been isolated with the 2'-0:methylnucleosides on DEAE- cellulose (borate) columns as discussed earlier. Ribosomal RNA was prepared from isolated ribosomes to minimize this difficulty. Nuclear RNA was prepared at pH 5.1 as described by Quagliarotti g£_al: (1970); the reduced solubility of DNA at this pH resulted in preparation of RNA which was free of detectable amounts of DNA. DNA contamination would be detected on the LC column with Buffer B if it exceeded 0.1% of the RNA since dC separates from Cm (see Table 7). The presence of deoxynucleosides could be verified by LC of the 2'0:methylnucleoside fraction with Buffer A Since both dG and dC would be detected as discrete peaks (see Table 6). Complete hydrolysis of RNA samples to nucleosides is also essential for an accurate determination of 2'-9:methy1nucleoside composition. This is especially important in light of the increased stability towards snake venom phosphodiesterase of phosphodiester bonds involving 2'-0:methylnucleosides (Gray and Lane, 1967). The 98 99 mixture of enzymes used for hydrolysis was shown to be adequate under the described conditions since it completely hydrolyzed a sample of wheat germ alkaline stable trinucleotides in 30 hr under the described conditions. Deamination of a significant portion of A or Am during enzymatic hydrolysis of RNA by impurities in the enzyme preparations could be monitored by the appearance of inosine or Im in the LC elution profiles (see Tables 6 and 7). Deamination was not generally a problem, but it occasionally occurred when Sigma snake venom phos- phodiesterase was used in place of Worthington phosphodiesterase I. Hydrolysis of nucleosides would also be detected since several of the free bases would form discrete LC peaks (see Tables 6 and 7), but hydrolysis was never observed under these conditions. Several approaches were investigated for nucleoside composition determinations before a satisfactory procedure was devised. The pre- cision of the GC techniques was unsatisfactory due to inconsistent formation of TMS derivatives for uracil and cytosine nucleosides. It is potentially useful for the direct analysis of nucleic acid hydroly- zates in spite of these difficulties. However, an additional diffi- culty was encountered in forming TMS derivatives of 2'-Q:methyl- nucleoside fractions isolated by DEAE-cellulose (borate) column chromatography. Incomplete removal of boric acid from this fraction with methanol resulted in incomplete formation of TMS derivatives for all 2'-9:methy1nucleosides; in some cases no derivatives were formed at all. Since this difficulty was not readily overcome, other procedures were investigated. 100 Direct probe mass Spectrometry of a mixture of nucleosides could potentially be a satisfactory approach to compositional analysis. The major difficulties encountered in these studies were unsatisfactory resolution of the unique ions for each nucleoside at low resolution and irregular thermal decomposition of the nucleosides under the slightly erratic conditions of manual temperature pro- gramming. High resolution mass spectrometry would circumvent the first problem by permitting the unique ions listed in Table 4 to be measured with low cross contamination from other nucleosides. The latter difficulty could be eliminated by using a programmed temperature increase. One potential advantage of determining nucleoside composi- tions with MS is the prospect of analyzing relatively impure samples. High resolution MS would permit the unique ions to be identified and quantitated in the presence of contaminating ions produced by other classes of compounds. This procedure has obvious clinical implica- tions, since pure nucleosides would be difficult to prepare from some clinical samples (e.g., blood or urine). Liquid chromatography proved to be the most satisfactory approach attempted for determining nucleoside compositions. Resolu- tion of ribonucleosides was achieved with a 0.4 M ammonium formate buffer at pH 4.55 (Buffer A). Use of an organic solvent (ethylene glycol) in the elution buffer (Buffer B) resulted in satisfactory resolution of 2'-0:methylnucleosides. Several potential sources of error in nucleoside composition values could be detected from the elution profiles as discussed above. The resolving potential of the 101 LC column is further demonstrated by the resolution of 2'-0:methyl- nucleosides from analogous 3'-Q:methy1nucleosides in several cases. Uracil nucleosides were not resolved from one another since they are not retained well under these conditions. The accuracy of the procedure was verified by analysis of wheat germ ribosomal RNA (188 plus 288), which has been characterized for both ribonucleoside composition and 2'-0:methylnucleoside composition by Lane (1965). The values obtained with this method are in close agreement with those determined by Lane (see Table 8). The method is more direct and less time consuming than previously reported procedures involving ion-exchange (Lane, 1965; Al-Arif and Sporn, 1972a) or liquid partition (Hall, 1964) column methods which were coupled with paper chromatographic separations. The sensitivity of the method is sufficient to distinguish two RNA samples which differ in their 2'-Q:methylnucleoside composition by 2-3%. In the case of mammalian ribosomal RNA, such as HeLa cell with 104 2'-Q:methylnucleosides per 188 plus 288 (Salim and Maden, 1973) or Novikoff hepatoma ascites cell with about 94 2'-0:methylnucleosides per 188 plus 288 (Choi and Busch, 1970), it would be possible to distinguish RNA samples which differ by only 2 or 3 2'-Q:methylnucleosides in their total complement. Consequently, the method is potentially useful in comparative studies of the ribo- somal RNA from various tissues such as normal tissue, hepatomas, or drug induced carcinomas. Several separate studies of nucleoside composition were performed along these lines. The first study compared the nucleoside 102 compositions of rat liver ribosomal RNA and ribosomal RNA from the liver-derived Novikoff hepatoma ascites cell. Both free and membrane bound ribosomal RNA from rat liver were Similar to Nevikoff cell ribosomal RNA in ribonucleoside and 2'-Q-methylnucleoside composition. The results are consistent with the occurrence of identical primary sequences for the ribosomal RNA from the three sources, but variation could still exist. Similarity in ribonucleoside composition cannot be taken as evidence for identical sequences when comparing ribosomal RNA's due to the large number of nucleosides involved in the molecules (approxi- mately 6400 nucleosides in 188 plus 28S ribosomal RNA). Minor varia- tions may be masked or sequence variations in different regions of the RNA chains may balance each other. Indeed, evidence for minor differences between the primary sequences of rat liver and NOvikoff cell 288 cytoplasmic ribosomal RNA have been reported (Wikman, Quagliarotti, g£_§l:, 1970; Seeber and Busch, 1971). The 2'-0:methy1nucleoside composition data is more conclusive than the ribonucleoside composition data since it is based on a total of approximately only 95 nucleosides per ribosome (Choi and Busch, 1970). The sensitivity of the assay permits detection of a difference in 2'-0;methylnuc1eoside composition on the order of 2-3 nucleosides as discussed above. Therefore, any variation in 2'-9:methylation of Novikoff cell ribosomal RNA relative to liver ribosomal RNA would of necessity require that a 2'-0:methylnucleoside of one type at one site in the chain be replaced by a 2‘-9:methylnucleoside of the same type at another site in the chain. Any further definitive conclusion 103 with reSpect to identifying methylation sites will be answered only when the complete sequences are obtained. When rat liver ribosomal RNA was compared to several other rat tissue ribosomal RNA'S, the ribonucleoside compositions were similar, but the 2'-0:methylnuc1eoside composition of liver RNA differed from those of the other tissues. An increase of 3-4 Cm residues and a decrease of 3-4 Gm residues per ribosome for liver relative to other tissues was indicated by these results. If real, the observed differences could be the result of a) methylation at the same nucleoside residue of the ribosomal RNA precursor chain with different primary sequences for various tissues, or b) a change in methylation patterns of ribosomal RNA precursors with the same primary sequence. It is not possible to distinguish between these alternatives at the present time. Further study is required to determine whether both the 188 and 28S ribosomal RNA molecules are altered, or whether only one of them contains all the differences. One chain may be more highly methylated while the other contains fewer of these modified nucleosides. The analysis of the nucleoside compositions of nuclear RNA in rat liver and Novikoff cell was complicated by the low amounts of nuclear RNA available. Since insufficient quantities were obtained for the isolation of homageneous ribosomal RNA precursor molecules, the entire high molecular weight fraction was analyzed. Both the ribonucleoside and the 2'-0:methylnucleoside compositions differed when rat liver high molecular weight nuclear RNA was compared to that of Novikoff cell. These differences are probably due to the 104 difference in distribution of ribosomal RNA precursors in these fractions as discussed earlier. Analysis of the low molecular weight nuclear RNA fractions was not attempted, since they also represent a mixture and generally even less material was available. This study would be more feasible if methyl-labeled methionine was used to label the RNA and homogeneous RNA samples were prepared prior to analysis. Treating Novikoff cells with ethionine had no effect on the 2'-9-methylnucleoside composition of nuclear RNA, although the incorporation of methyl label into this fraction was reduced to about 42% of the control level. Under the same conditions 3H-uridine incorporation was reduced to 47% of the control level. The decrease could result from a decreased rate of synthesis or an increased rate of degradation of the SH-uridine labeled nuclear RNA in the presence of ethionine. It accounts for most of the reduction in methyl label incorporation, but there could be a slight build-up of undermethylated RNA under these conditions similar to that observed by Vaughan g£_§l: (1967). The observation that ethionine treatment has no effect on the 2'-Q:methylnucleoside composition of Novikoff cell nuclear RNA is consistent with complete methylation of a ribosomal RNA precursor molecule prior to further processing once modification has been initiated. This possibility is attractive since methylation of the 458 precursor is required for processing in HeLa cell (Vaughan ggngl., 1967). The end result would be production of normal 188 and 28S ribosomal RNA molecules at a reduced rate in the presence of ethionine. 105 There is agreement between the 2'-0:methylnucleoside com- positions obtained for Novikoff cell nuclear RNA by either UV analysis or methyl label incorporation following the 30 min labeling period with methyl-labeled methionine (Table 10). The explanation for the difference between the results of Choi and Busch (1970) and those reported here may lie in the relative amount of incorporation of formyl groups into purine rings via the tetrahydrofolate pathway during the two different experiments. For example, Choi and Busch observed that less than 25% of their total methyl label eluted with the alkaline stable dinucleotide fraction upon DEAE-cellulose (chloride) column chromatography following alkaline hydrolysis of the RNA. This percentage is lower than predicted from the observa- tion that over 90% of eukaryotic ribosomal RNA methylation occurs on the ribose moiety (Wagner, Penman, and Ingram, 1967; Lane and Tamaoki; 1969). By contrast, over 85% of the methyl label was found in the 2'-Q:methylnucleoside fraction when ring incorporation was inhibited by adding sodium formate and purines in the short term labeling experiment described here. Consistent with high levels of purine ring labeling in the results reported by Choi and Busch is the observation that the relative amounts of Am and Gm are elevated while the percentage of both Um and Cm are decreased relative to the results reported here. The Um to Cm ratio would not be expected to be altered by purine ring labeling, and this ratio is constant at about 0.75 in both experiments. Further, the amount of label in Am and Gm is approxi- mately the same in both cases, as would be expected if label 106 was incorporated into both adenosine and guanosine rings to the same extent. In summary, a rapid and sensitive procedure for determining the 2'-0:methylnucleoside composition of an RNA sample has been established. It permits analysis of a 5 mg sample of RNA provided 1-2% of the nucleosides are ribose-methylated. In this case the 2'-g:methylnucleoside fraction will be 0.1-0.3 umoles, a quantity sufficient for several independent determinations of its composition. The sensitivity of the determination may be significantly increased if the RNA samples are labeled with methyl-labeled methionine prior to analysis and the LC column fractions are collected and counted. A ribonucleoside composition may be obtained from the same sample prior to DEAE chromatography if it is desired, Since this analysis requires less than 1% of the original hydrolyzate. The method has been used for the structural comparison of several groups of ribosomal and nuclear RNA's. The Novikoff hepatoma ascites cell, which is liver derived, was shown to contain ribosomal RNA with a Similar 2'-0-methylnucleoside composition to that of both free and membrane bound liver ribosomal RNA's. A possible difference in the 2'-0:methylnuc1eoside composition of liver ribosomal RNA with respect to several other rat tissue ribosomal RNA'S was found. Nuclear RNA nucleoside composition comparisons between cell types were found to be ambiguous since they were based on RNA preparations which differed in the distribution of various RNA'S in their complement. Satisfactory results were achieved when comparisons of nuclear RNA 107 were made within a cell type, probably since these RNA preparations would have more similar complements. When comparing two sets of RNA on a structural basis, it is essential to work with a well defined preparation. As new methods for purification of Specific RNA molecules become available, it becomes easier to prepare relatively homogeneous RNA samples. This method pro- vides a rapid, sensitive structural analysis and will be of continued interest in the characterization of such samples as messenger RNA, heterogeneous nuclear RNA, specific low molecular weight nuclear RNA molecules, or specific ribosomal RNA molecules and their nucleolar precursors. BIBLIOGRAPHY BIBLIOGRAPHY Abbate, J., 1971, ”The Determination of 2'-9:Methylation in RNA," Ph.D. Thesis, Michigan State University. Abbate, J. and F. Rottman, 1972, Anal. Biochem. 41, 378. Al-Arif, A. and M. B. Sporn, 1972a, Anal. Biochem. 483 386. Al-Arif, A. and M. B. Sporn, 1972b, Proc. Nat. Acad. Sci., U.S. 69, 1716. _ Amaldi, F. and G. Attardi, 1971, Biochemistry_193 1478. Aronson, A. I. and F. W. Wilt, 1969, Proc. Nat. Acad. Sci., U.S. 22, 186. Averner, M. J. and N. R. Pace, 1972, J. Biol. Chem. 247, 4491. Balle, G., P. A. Cerutti, and B. Witkop, 1966, J. Am. Chem. Soc. 88, 3946. Baskin, F. and C. A. Dekker, 1967, J. Biol. Chem. 242, 5447. Beaud, G. and D. H. Hayes, 1971a, Eur. J. Biochem. 19, 323. Beaud, G. and D. H. Hayes, 1971b, Eur. J. Biochem. 293 525. Biswas, B. B. and J. Myers, 1960, Nature 186, 238. Bobst, A. M., P. A. Cerutti, and F. Rottman, 1969, J. Am. Chem. Soc. 91, 1246. Bobst, A. M., F. Rottman, and P. A. Cerutti, 1969a, J. Mol. Biol. 46, 221. Bobst, A. M., F. Rottman, and P. A. Cerutti, 1969b, J. Am. Chem. Soc. 9__1_, 4603. Brahms, J., J. C. Maurizot, and A. M. Michelson, 1967, J. Mol. Biol. 25, 481. 108 109 Bray, G. A., 1960, Anal. Biochem. 13 279. Broom, A. D. and R. K. Robins, 1965, J. Am. Chem. Soc. 81, 1145. Brown, D. M. and A. R. Todd, 1955, in "The Nucleic Acids," (E. Chargaff and J. N. Davidson, Eds.), Vol. 1, 409, Academic Press, New York. Brownlee, G. G., E. Cartwright, T. McShane, and R. Williamson, 1972, FEBS (Fed. Eur. Biochem. Soc.) Lett. 25, 8. Brownlee, G. G., E. Sanger and B. G. Barrell, 1967, Nature 215, 735. Busch, H., 1967, in "Methods in Enzym010gy," (L. Grossman and K. Moldave, Eds.), Vol. 12A, 421, Academic Press, New York. Busch, H., Y. C. Choi, I. Daskal, A. Inagaki, M. O. J. Olson, R. Reddy, T. S. Ro-Choi, H. Shibata, and L. C. Yeoman, 1972, Karolinska Symposium on Research Methods in Reproductive Endocrinology 5, KarolinSka Institutet,78tockholm. Busch, H., T. S. Ro-Choi, A. W. Prestayko, H. Shibata, S. T. Crooke, S. M. El-Khatib, Y. C. Choi, and C. M. Mauritzen, 1971 (Autumn), Perspect. Biol. Med. 15, 117. Busch, H. and K. Smetana, 1970, "The Nucleolus," Academic Press, New York. Butler, C. C., 1955, in "Methods in Enzymology," (S. P. Colowick and N. 0. Kaplan, Eds.), Vol. 2, S61, Academic Press, New York. Butts, W. C., 1970, J. Chrom. Sci. 8, 474. Chen, P. 8., Jr., T. Y. Toribara, and H. Warner, 1956, Anal. Chem. 283 1756. Choi, Y. C. and H. Busch, 1970, J. Biol. Chem. 245, 1954. Craddock, V. M., 1969, Biochim. BioPhys. Acta 195, 351. Crestfield, A. M., K. C. Smith, and F. W. Allen, 1955, J. Biol. Chem. 216, 185. DeClercq, E., R. D. Wells, R. C. Grant, and T. C. Merigan, 1971, J. Mol. Biol. 593 83. DeClercq, E., B. Zmudzka, and D. Shugar, 1972, FEBS (Fed. Eur. Biochem. Soc.) Lett. 243 137. Dingman, C. W. and A. C. Peacock, 1968, Biochemistry Z! 659. 110 Dounce, A. L., 1955, in "The Nucleic Acids," (E. Chargaff and J. N. Davidson, Eds.), Vol. 2, 93, Academic Press, New York. Dounce, A. L., 1963, Exptl. Cell Res., Suppl. 9, 126. Dunlap, B. E., K. H. Friderici, and F. Rottman, 1971, Biochemistry 10, 2581. Egawa, K., Y. C. Choi, and H. Busch, 1971, J. Mol. Biol. 39, 565. El-Khatib, S. M., T. S. Ro-Choi, and H. Busch, 1970, J. Biol. Chem. 245, 3416. Ford, P. J. and E. M. Southern, 1973, Nature New Biology_24l, 7. Forget, B. G. and 8. M. Weissman, 1967, Science 158, 1695. Friderici, K., 1974, M. S. Thesis, Michigan State University. Furukawa, Y., K. Kobayashi, Y. Kanai, and M. Honjo, 1965, Chem. Pharm. Bull. 33, 1273. Gerard, G. F., F. Rottman and J. A. Boezi, 1971, Biochemistry 10, 1974. _- Gerard, G. F., F. Rottman, and J. A. Boezi, 1972, Biochem. BiOphys. Res. Comm. 59, 1095. Gerard, G. F., F. Rottman, and M. Green, 1974, Biochemistry 33, in press. Gin, J. B. and C. A. Dekker, 1968, Biochemistry_z, 1413. Gray, M. W. and B. G. Lane, 1967, Biochim. Biophys. Acta 134, 243. Greenberg, H. and S. Penman, 1966, J. Mol. Biol. 23, 527. Grierson, 0., M. E. Ragers, M. L. Sartirana, and U. E. Loening, 1970, Cold Spring Harbor Symp. Quant. Biol. 33, 589. Hall, R. H., 1963, Biochim. Biophys. Acta 33, 278. Hall, R. H., 1964, Biochemistry 3, 876. Hall, R. H., 1971, "The Modified Nucleosides in Nucleic Acids," Columbia University, New York. Hashimoto, S. and M. Muramatsu, 1973, Eur. J. Biochem. 33, 446. Hecht, S. M., A. S. Gupta, and N. J. Leonard, 1970, Anal. Biochem. 38, 230. 111 Higashi, K., T. Matsuhisa, S. Gotoh, K. Nishinaga, and Y. Sakamoto, 1971, Biochim. Biophys. Acta 232, 352. Hindley, J. and S. M. Page, 1972, FEBS (Fed. Eur. Biochem. Soc.) Lett. 33, 157. Hobbs, J., H. Sternbach, and F. Eckstein, 1971, FEBS (Fed. Eur. Biochem. Soc.) Lett. 33, 345. Hobbs, J., H. Sternbach, and F. Eckstein, 1972, Biochem. Biophys. Res. Comm. 19, 1509. Hodnett, J. L. and H. Busch, 1968, J. Biol. Chem. 243, 6334. Howlett, H. A., M. W. Johnson, A. R. Trim, J. Eagles, and R. Self, 1971, Anal. Biochem. 32, 429. Jacobson, M., J. F. O'Brien, and C. Hedgcoth, 1968, Anal. Biochem. 25, 363. Janick, B., M. P. Kotick, T. H. Krieser, L. F. Reverman, R. G. Sommer, and D. P. Wilson, 1972, Biochem. Biophys. Res. Comm. 46, 1153. -_' Johnson, R., 1970, Biochem. J. 119, 699. Khan, M. K. A., P. Comstock, and F. Rottman, 1974, Manuscript submitted. Khan, M. K. A., and F. Rottman, 1972, FEBS (Fed. Eur. Biochem. Soc.) Lett. 33, 25. Khwaja, T. A. and R. K. Robins, 1966, J. Am. Chem. Soc. 33, 3640. Klootwijk, J. and R. J. Planta, 1973, Eur. J. Biochem. 32, 325. Kusmierek, J. T., J. Giziewicz, and D. Shugar, 1973, Biochemistry 12, 194. Kusmierek, J. T., M. Kielanowska, and D. Shugar, 1973, Biochem. Biophys. Res. Comm. 33, 406. Labrie, F. and F. Sanger, 1969, Biochem. J. 114, 29P. Lane, B. C., 1965, Biochemistry g, 212. Lane, 8. G. and T. Tamaoki, 1967, J. Mol. Biol. 31, 33s. Lane, B. G. and T. Tamaoki, 1969, Biochim. Biophys. Acta 179, 332. 112 Larsen, C. J., P. Lebowitz, S. M. Weissman, and B. Dubuy, 1970, Cold Spring Harbor Symp. Quant. Biol. 33, 35. Laskowski, M., Sr., 1966, in "Procedures in Nucleic Acid Research," (G. L. Cantoni and D. R. Davies, Eds.), 154, Harper and Row, New York. Laskowski, M., Sr., 1971, in "The Enzymes," Third Edition (P. D. Boyer, Ed.), Vol. 4, 313, Academic Press, New York. I. Lazarus, H. M. and M. B. Sporn, 1967, Proc. Nat. Acad. Sci., U.S. 31, 1386. . Levin, S. and N. Fausto, 1973, Biochemistry_33, 1282. Maden, B. E. H., C. D. Lees, and M. Salim, 1972, FEBS (Fed. Eur. Biochem. Soc.) Lett. 33, 293. Magasanik, B. and E. Chargaff, 1951, Biochim. Biophys. Acta Z, 396. Martin, D. M. G., C. B. Reese, and G. F. Stephenson, 1968, Biochemistry 7, 1406. Merigan, T. C. and F. Rottman, 1973, Personal communication. Mirault, M. E. and K. Scherrer, 1972a, FEBS (Fed. Eur. Biochem. Soc.) Lett. 33, 233. Mirault, M. E. and K. Scherrer, 1972b, FEBS (Fed. Eur. Biochem. Soc.) Proc. 31, 197. Moore, P. B., 1966, J. M01. Biol. 13, 38. Morisawa, S. and E. Chargaff, 1963, Biochim. Biophys. Acta 33, 147. Moriyama, Y., J. L. Hodnett, A. W. Prestayko, and H. Busch, 1969, J. Mol. Biol. 33, 335. Morse, P. A., Jr., and V. R. Potter, 1965, Cancer Res. 33, 499. Nakamura, T., A. W. Prestayko, and H. Busch, 1968, J. Biol. Chem. 243, 1368. Nazar, R. N. and H. Busch, 1972, Cancer Res. 33, 2322. Nichols, J. L. and B. G. Lane, 1967, J. Mol. Biol. 33, 477. Penman, s. J., 1966, J. Mol. Biol. 11, 117. Perry, R. P., 1962, Proc. Nat. Acad. Sci., U.S. 33, 2179. 113 Perry, R. P., T. Y. Cheng, J. J. Freed, J. R. Greenberg, D. E. Kelley, and K. D. Tartof, 1970, Proc. Nat. Acad. Sci., U.S. 33, 609. Perry, R. P. and D. E. Kelley, 1972, J. Mol. Biol. 29, 265. Perry, R. P. and D. E. Kelley, 1974, Cell 1, SEE Pike, L. M., 1971, "A Rapid and Sensitive Technique for the Determination of the 2'-Q:Methy1nucleoside Ratio of an RNA Molecule," M.S. Thesis, Michigan State University. Prestayko, A. W., B. G. Lewis, and H. Busch, 1972, Biochim. Biophys. Acta 269, 90. Prestayko, A. W., B. G. Lewis, and H. Busch, 1973, Biochim. Bi0phys. - Acta 319, 323. Prestayko, A. W., M. Tonato, and H. Busch, 1970, J. Mol. Biol. 31, 505. Price, A. R. and F. Rottman, 1970, Biochemistry 3, 4524. Quagliarotti, G., E. Hidvegi, J. Wikman, and H. Busch, 1970, J. Biol. Chem. 245, 1962. Reddy, R., T. S. Ro-Choi, D. Henning, H. Shibata, Y. C. Choi, and H. Busch, 1972, J. Biol. Chem. 247, 7245. Rein, A. and S. Penman, 1969, Biochim. Biophys. Acta 190, 1. Retel, J. and R. J. Planta, 1967, Eur. J. Biochem. 3, 248. Retel, J., R. C. van den 805, and R. J. Planta, 1969, Biochim. Biophys. Acta 195, 370. Robins, M. J. and S. R. Naik, 1971, Biochim. Biophys. Acta 246, 341. Ro-Choi, T. S., Y. Moriyama, Y. C. Choi, and H. Busch, 1970. J. Biol. Chem. 245, 1970. Ro-Choi, T. S., R. Reddy, D. Henning, T. Takano, C. W. Taylor, and H. Busch, 1972, J. Biol. Chem. 247, 3205. Rottman, F., K. Friderici, and M. K. A. Khan, 1974, Manuscript submitted. Rottman, F. and K. Heinlein, 1968, Biochemistry Z, 2634. Rottman, F. and K. L. Johnson, 1969, Biochemistry_3, 4354. Rubenstein, L. and U. Clever, 1971, Biochim. Biophys. Acta 246, 517. 114 Rubin, c. M., 1973, J. Biol. Chem. 243, 3860. Salim, M. and B. E. H. Maden, 1973, Nature 244, 334. Scherrer, K. and J. E. Darnell, 1962, Biochem. Biophys. Res. Comm. 7, 486. Seeber, S. and H. Busch, 1971, Cancer Res. 33, 1888. Shibata, H., R. Reddy, D. Henning, T. S. Ro-Choi, and H. Busch, 1973, Fed. Proc. 33, 664 Abst. Singh, H. and B. G. Lane, 1964, Can. J. Biochem. 33, 1011. Singhal, R. P. and W. E. Cohn, 1972, Anal. Biochem. 45, 585. Was-1:. - Singhal, R. P. and W. E. Cohn, 1973, Biochemistry 33, 1532. Smith, J. D. and D. B. Dunn, 1959, Biochim. Biophys. Acta 33, 573. Smith, K. C. and F. W. Allen, 1953, J. Am. Chem. Soc. Z3, 2131. Soeiro, R., M. H. Vaughan, J. R. Warner, and J. E. Darnell, Jr., 1968, J. Cell Biol. 33, 112. Starr, J. L. and B. H. Sells, 1969, Physiol. Rev. 53, 623. Stewart, G. A. and E. Farber, 1968, J. Biol. Chem. 243, 4479. Stuart, S. E. and F. Rottman, 1973, Biochem. Biophys. Res. Comm. 55, 1001. Sweeley, C. C., B. 0. Ray, W. I. Wood, and J. F. Holland, 1970, Anal. Chem. 33, 1505. Tazawa, I., S. Tazawa, J. L. Alderfer, and P. 0. P. Ts'o, 1972, Biochemistry_£l, 4931. Tennant, R. W., J. G. Farrelly, J. N. Ihle, B. C. Pal, F. T. Kenney and A. Brown, 1973, J. Virol. 33, 1216. Tomlinson, R. V. and G. M. Tener, 1963, Biochemistry 3, 697. Torrence, P. F., A. M. Bobst, J. A. Waters, and B. Witkop, 1973, Biochemistry 33, 3962. Ts'o, P. 0. P., S. A. Rapaport, and F. J. Bollum, 1966, Biochemistry 5, 4153. 115 Uziel, M., C. K. Koh, and W. E. Cohn, 1968, Anal. Biochem. 33, 77. van den 805, R. C. and R. J. Planta, 1973, Biochim. Biophys. Acta 294, 464. Vaughan, M. H., Jr., R. Soeiro, J. R. Warner, and J. E. Darnell, Jr., 1967, Proc. Nat. Acad. Sci., U.S. 33, 1527. Villa—Trevino, S., K. Shull, and E. Farber, 1966, J. Biol. Chem. 241, 4670. Wagner, E. K., L. Katz, and S. Penman, 1967, Biochem. Biophys. Res. Comm. 33, 152. Wagner, E. K., S. Penman, and V. M. Ingram, 1967, J. Mol. Biol. 29, 371. Weinberg, R. A., 1973, Ann. Rev. Biochem., E. E. Snell, Ed., Vol. 42, 329. Weinberg, R. A., U. Loening, M. Willems, and S. Penman, 1967, Proc. Nat. Acad. Sci., U.S. 33, 1088. Weinberg, R. A. anui S. Penman, 1968, J. Mol. Biol. 33, 289. Weinberg, R. A. and S. Penman, 1969, Biochim. Biophys. Acta 190, 10. Weinberg, R. A. and S. Penman, 1970, J. Mol. Biol. _4_7_, 169. Wikman, J., E. Howard, and H. Busch, 1970, Cancer Res. 33, 773. Wikman, J., G. Quagliarotti, E. Howard, Y. C. Choi, and H. Busch, 1970, Cancer Res. 33, 2749. Williamson, R. and G. G. Brownlee, 1969, FEBS (Fed. Eur. Biochem. Soc.) Lett. 3, 306. Yazdi, E., T. S. Ro-Choi, J. Wikman, Y. C. Choi, and H. Busch, 1969, Cancer Res. 33, 1755. Zapisek, W. F., A. G. Saponara, and M. D. Enger, 1969, Biochemistry_ 8, 1170. Zimmerman, E. F., 1968, Biochemistgy Z, 3156. Zimmerman, E. F. and B. W. Holler, 1967, J. Mol. Biol. 33, 149. Zmudzka, 8., C. Janion, and D. Shugar, 1969, Biochem. Biophys. Res. Comm. 33, 895. Zmudzka, B. and D. Shugar, 1970, FEBS (Fed. Eur. Biochem. Soc.) Lett. 8, 52. PART II ALKYLATION OF ADENOSINE BY DIAZOETHANE IN AQUEOUS 1,2-DIMETHOXYETHANE: SOLVENT PARTICIPATION IMPLICATED BY MASS SPECTRAL ANALYSIS Lee M. Pike, M. Khurshid A. Khan, and Fritz Rottman Contribution from the Department of Biochemistry Michigan State University East Lansing, Michigan 48823 116 ABSTRACT Synthesis of 2'-3:ethy1adenosine by treatment of adenosine with diazoethane in aqueous 1,2-dimethoxyethane produced several unexpected alkylation products. Characterization of the products by several methods, including mass spectrometry of their trimethylsilyl derivatives, indicated that both 3rmethyl and 3fethyl compounds were being produced. Analysis of the reaction conditions employing palmitic acid as an alkyl acceptor implicated the solvent (1,2-dime- thoxyethane) as a source of extraneous alkyl groups. In addition, the mass Spectral analysis of 3:methyl nucleosides as their trimethyl- silyl derivatives provided insight into the origin of several ions derived from the ribose moiety. In a few cases, the mechanism of fragment formation was found to be different for TMS-nucleosides than for TMS-nucleotides or TMS-sugar phosphates. These ions were used in elucidating structures for the alkylation products of adenosine. 117 118 The natural occurrence of 2-3fmethylribose in RNA was first reported by Smith and Dunn.2 Subsequent work has indicated that 2'-3:methy1nuc1eotides are ubiquitous; they occur in the RNA from all 3 Considerable effort has been made to 4-7 organisms examined to date. elucidate the functional role of this modification in RNA. Several homopolymers of 2'-3:methy1 substituted nucleotides were prepared to examine the effects these modifications might have on RNA structure.8'10 Surprisingly, the physical characteristics of these polymers indicated that they had more self-structure than 10-12 This result was somewhat analogous polymers of ribonucleotides. unexpected since intramolecular hydrOgen bonding between the 2'- hydroxyl moiety and nearby groups in the polymer chain had been postulated to contribute to the ordered conformation of RNA.13'14 It appears that substitution on the 2'-carbon of poly- nucleotides by a bulky group, perhaps requiring an oxygen atom, represents a more important contribution to polynucleotide stability than the hydrOgen bond forming potential of unsubstituted 2'-hydroxy1 groups. The role of oxygen at the 2'-position is not clear, but the possible steric effect of a hydroxyl or methoxyl group is among several factors to be considered. Therefore it was of interest to attempt the synthesis of an RNA molecule containing a 2'-3:ethy1 group (Poly Ae),15’16 using procedures similar to those employed earlier in the synthesis of poly Am,8 to further study this steric effect. During the synthesis of the 2'-3:ethy1adenosine monomer using diazoethane in aqueous 1,2-dimethoxyethane, we noted several 119 unexpected alkylation products. This paper describes the charac- terization of the major products by combined gas chromatOgraphy-mass Spectrometry of their trimethylsilyl derivatives and indicates that the reaction probably involves the solvent (1,2-dimethoxyethane) as a source of extraneous alkyl groups. The origins of several specific ions produced by mass Spectral analysis of the TMS deriva- tives of ribonucleosides have been defined by utilizing the unique fragmentation patterns of nucleosides alkylated on various hydroxyls of the ribose moiety. EXPERIMENTAL Material. Adenosine was purchased from Raylo Chemicals Limited (Canada). Authentic 2'-3:methyladenosine, 3'-3:methyl- adenosine, 2',3'-di-3:methy1adenosine, and 3?, 2'-3:dimethy1adenosine were obtained from adenosine treated with diazomethane in an aqueous solution of 1,2-dimethoxyethane.8 Bis (trimethylsilyl) acetamide was purchased from Regis Chemical Co. Methyl palmitate, ethyl palmitate and palmitic acid were purchased from Applied Science Laboratories, Inc. The gas chromatography column packing [3% SE-30 on silanized Supelcoport (100/120 mesh)] was purchased from Supelco, Inc. Alkylation of adenosine and fractionation of the products. The preparation of diazoethane from nitrosoethylurea and the alkyla- tion of adenosine were performed by a modification of the method 17,18 used for preparation of 2'-3:methyladenosine. Diazoethane was generated by addition of 40 g Ernitrosoethylurea to 200 ml of 120 1,2-dimethoxyethane in the presence of 100 ml 50% KOH at -18° C. After drying for 3 hr over KOH pellets at -18° C, 160 ml of the diazo- ethane solution was added to two grams of adenosine in 80 m1 of water at 80° C. The crude reaction mixture was stirred overnight at room temperature, evaporated to dryness, and redissolved in 40% ethanol. The products were separated by ion exchange chromatography on Bio-Rad AG 1-X2 (OH'), 200/400 mesh. Nucleosides were eluted with 40% ethanol and they were characterized by descending paper chroma- t0graphy in the following solvent systems: (A) 2-propanol-NH40H - 0.1 ! boric acid (7:1:2), (B) ethyl acetate-l-propanol-HZO (4:1:2, upper phase), (C) l-butanol-NH40H - H20 (86:5:14), and (D) ethanol-1 !_ammonium acetate, pH 7:5 (7:3), all solvent proportions by volume. The nature of the alkyl ether formed on the ribose moiety was determined by perchloric acid hydrolysis of the nucleosides combined with gas chromatographic analysis of the released alcohol.19 Hydrolysis of 10 ug of each nucleoside was achieved by addition of perchloric acid to the lyophilized sample and heating in sealed pyrex glass tubes at 110° C for 90 min. After neutralization with KOH in the resealed tube, the upper liquid layer was analyzed by gas chromatography. Alkylation of palmitic acid and analysis of the products. For control reactions designed to demonstrate the absence of methyl donors in diazoethane generated in diethylether, diazoethane was generated by addition of 4 g Ernitrosoethylurea to 20 ml diethylether in the presence of 10 ml 50% KOH at -18° C. The alkyl esters of palmitic acid were formed under the following two reaction conditions: 121 a) 23 mg palmitic acid was dissolved in 1 ml diethylether and 2 m1 ethereal diazoethane was added at room temperature. The reaction proceeded for 1 hr, at which time the sample was evaporated under a stream of N2. b) 23 mg palmitic acid was placed in 1 ml water, heated to 80° C, and 2 ml of 1,2-dimethoxyethane containing diazoethane at -l8° C was added. The reaction proceeded for 1 hr at room. temperature. The aqueous phase was extracted 2 times with 4 m1 ether, the ether layers were combined, washed with 4 ml water and the ether phase was evaporated with N2. Both samples were dissolved in hexane for gas chromatography. Methyl and ethyl palmitate were resolved by gas chromatography using a Hewlett-Packard F and M 402. The column of U-shaped glass tubing (6 ft x 1/8 in i.d.) was packed with 3% SE-30 on silanized SupelcOport (100/120 mesh). Nitrogen carrier gas was maintained at 50 ml per minute and the separations were carried out isothermally at 155° C. Trimethylsilylation procedure and combineg:g§s chromatography- mass spectrometry of the nucleosides. Derivatization of the purified nucleosides for GC-MS was performed in 1/2 dram vials sealed with a teflon-lined screw cap. Trimethylsilylation was achieved by adding BSA in 100 mole excess and heating the solution at 120° C for 120 min.20 Low resolution mass spectra were recorded on an LKB—9000 combined GC-MS using an ionizing energy of 70 eV. The ion source was maintained at 290° C and the molecular separator at 230° C with a trap current of 60 uamps. The GC column consisted of coiled glass tubing (4 ft x 3 mm i.d.) packed with 3% SE-30 on silanized Supelcoport 122 (100/120 mesh). Helium carrier gas was maintained at 35 ml per minute. The column was maintained isothermally at 230° C and the flash heater was at 250° C. The data was processed by an on-line digital computer system.21 RESULTS AND DISCUSSION Fractionation of the alkylation3products of adenosine and their preliminary characterization. The products of adenosine alkylation by diazoethane were resolved by ion exchange column chromatography on Bio-Rad AG-l as described in Methods. The results of a typical elution are presented in Table I. The resolution of six prominent ultra-violet absorbing fractions was unexpected, since in the analogous elution of methylated adenosine, only four fractions are observed.8 The yields of each component recovered are also indicated in Table I; the combined yield of these fractions represents about 34% of the original adenosine applied to the column. Unreacted adenosine and presumably some alkylated compounds were not recovered under the conditions employed. The relative amounts of fractions 3 and 4 varied in different preparations (10-20%) but they were always the largest fractions obtained. Under Similar reaction conditions diazomethane preferentially 17’18 Since the third alkylates the 2'-hydroxy1 group of adenosine. and fourth fractions are the major components eluted from the column, it was assumed that one of these fractions contained 2'-3:ethy1- adenosine. Free 2'-hydroxy1 groups have a greater tendency for ionization and subsequent column retention under the conditions 123 Table I. Fractionation of alkylated nucleosides on Bio-Rad AG 1 column.a Recovery . Fraction Presumed Tubesiiooled (% of ” Number Identity . Adenosine Each Fraction A . pplied) l 12 — 24 3.8 2 38 - 44 1.7 3 2'-3:ethyladenosine 45 - 58 11.9 4 2'537methy1adenosine 68 - 81 9.8 5 3'-3pthyladenosine 105 - 120 3.8 6 3'-3:methy1adenosine 179 - 210 2.8 8The column consisted of Bio-Rad AG 1 -x2 (OH’), 200/400 mesh, 4 x 40 cm, equilibrated with 40% ethanol prior to use. ‘95,100 A260 units of the crude reaction mixture was applied in 40% ethanol and eluted with 40% ethanol at a flow rate of 2 m1/min; tube volume was 20 m1. 124 employed in this separation;22 therefore 3'-alkyl ethers typically elute after 2'-alky1 ethers.8 Fractions 1 and 2 were assumed to contain dialkylated compounds by analogy to the elution of diazo- methane derivatives of adenosine. Components of all six AG-l fractions were characterized by descending paper chromatography in solvent systems A,B,C, and D. Methylation of either the 2'- or 3'-hydroxy1 groups of adenosine in- creases its migration in all four solvent systems as seen in Table II. This result is consistent with the reduction of polarity in the mole— cule through formation of an ether which renders it less soluble in the stationary, aqueous phase of the developing solvent and more soluble in the mobile, organic phase. Ethylation should further increase migration due to an even greater reduction in polarity. The compounds in fractions 3,4,5, and 6 each gave a single ultra-violet absorbing spot in all four chromatography systems, and were estimated to be greater than 97% pure. Conditions employed would have detected a single ultra-violet absorbing contaminant exceeding 2% of the total. Although the alkylation conditions were expected to yield only ethylated products, surprisingly the compound in fraction 4 migrated with 2'-3:methy1adenosine in all four systems and the fraction 6 component migrated with 3'-3:methy1adenosine in all four systems as shown in Table 11. These compounds were resolved from each other by solvent systems C and D. The compounds in fractions 3 and S migrated faster than 2'-3:methyladenosine and 3'-3fmethyladenosine and they were resolved from each other only by solvent system D. These results were consistent with the tentative identification of fraction 3 as 125 Table 11. Paper chromatography of alkylated adenosine derivatives. C°mp°und Soisgnt A Solsgnt B Solsgnt c Solsgnt D Fraction 3 0.70 0.53 0.57 0.82 Fraction 4 0.61 0.38 0.47 0.76 Authentic 2'-3f methyladenosine 0.63 0.37 0.47 0.77 Fraction 5 0.73 0.51 0.55 0.68 Fraction 6 0.60 0.35 0.40 0.64 Authentic 3'-37 methyladenosine 0.59 0.37 0.40 0.64 Adenosine 0.37 0.19 0.23 0.63 Authentic 2',3'-di-3: methyladenosine 0.71 0.61 0.62 0.82 Authentic 3?,2'-9:di- methyladenosine 0.77 0.55 0.68 0.77 126 2'-3:ethy1adenosine and fraction 5 as 3'j3fethy1adenosine Since the 2'-ether should have eluted from the AG-l column before the 3'-ether based on the behavior of the methyl derivatives of adenosine as discussed above. Fraction 1 contained several components, one of which migrated with 2',3'-di-3:methy1adenosine in all four chromatography systems and the rest of which migrated slightly faster, indicating the presence of two ethyl groups or mixtures of alkyl groups at the 2'- and 3'- positions. Fraction 2 also contained several components; the Rf's indicated that they were alkylated on the EF-amino group of the ring in addition to the 2'-hydroxyl group since they migrated slightly faster than 3?, 2'-3:dimethyladenosine. The mixtures in fraction 1 and fraction 2 were not further resolved since these compounds repre- sented only small percentages of the total product and were not useful for subsequent polynucleotide synthesis. Alkyl ethers on the ribose moiety are released as the free 23 and the 19 alcohol when nucleosides are treated with perchloric acid, alcohol can be identified by gas chromatOgraphy of the hydrolyzate. When the nucleosides in fraction 4 and 6 were subjected to perchloric acid hydrolysis, they released methanol while those in fractions 3 and 5 released ethanol. These results confirm the diversity in the nature of the alkyl substitution in the reaction products of adenosine treated with diazoethane in aqueous 1,2-dimethoxyethane. Mass spectral characterization of the alkylation products of adenosine. The four products in fractions 3-6 were characterized 127 by combined GC-MS of the trimethylsilyl derivatives of the nucleosides to identify the suspected 2'-3:ethyl- and 3'-3:ethy1adenosine and 2'-3;methy1- and 3'-3:methyladenosine. 'Trimethylsilylation has been shown to be an effective means of decreasing the polarity of nucleo- sides, thereby increasing their volatility.20 The more volatile nucleoside derivatives may be introduced to the mass spectrometer by gas chromatOgraphy rather than by direct inlet probe. This procedure provides clean separation from impurities with low amounts of material and reduces the effect of thermal decomposition of the relatively non-volatile, underivatized nucleoside. Moreover, several diagnostic ions discussed below appear to be stabilized by the conversion of free hydroxyl groups to TMS-ethers and they occur at a significant level for derivatized nucleosides. Mass spectra of the TMS derivatives of authentic adenosine (I), 2'-3:methy1adenosine (II), and 3'-3:methyladenosine (III) are presented in Figure 1. The molecular ion M+, at 212.555 for I, dis- plays a shift of 58 mass units to 912.497 for both 11 and III, con- sistent with the occurrence of an 3fmethyl group in the molecule. The ions at M-lS represent loss of a methyl group and are more prominent than the molecular ions since the TMS derivatives of carbohydrates readily lose a TMS-methyl group.24 It is interesting to note that the M-IS ion occurs at a higher relative intensity for 111 than for I or II. The mass spectrum of underivatized 3'-3_-methyladenosine25 contains a sizeable M—15 ion while the ion is much reduced for 2'-3:methyl- adenosine and absent for adenosine. Therefore, loss of an 3:alkyl 128 b F - DOM”. D D b8“. 5 D oyh IPL D Poms. P DOM“. 8“ on“ 8" on" 8" on ‘ 1 A! A 1- R b b I D qulbr. ”‘TP I Dr. I I nihibh b a. a 3 _ __ a. a a a mum . u o3 oo~ .8 + «:3 0.19 m N m\ . «a. i «9. .. .. \ one. on. .3 w 0. SN ._. N+x I: o # £32: 18 z .- SN 3 A A: J \ IOQ 3v 3: «at... nu a 2 09 L L! Lomn b VLsmn- b > home. F DOESD - V Lommb F F >3”. b F Don" S“ on" 8" on 4 J .4 4 i J— o i 1 ybilhdb. p 3L 4 2 0mm \ N0? m VON 1 9-: m :80 on... I z x wnw h .5: 3.5353258 N o 8N w: I n A 4 uxusah N+o B :8 6:? 9... o5 3: 3.5353158 . .. .m-.~-nmE A: 83353-43. . / J. \ ,8 Se 3: mo 1.300% and: 4 0.353 2:... 2. 2 02 on... can one 09 on» com onu com on. 02 on ”EUHl a ozmomq b D F K 5. 1P 1P b h P rfi‘bk PPb P h h I . I on... s. ova A 9-: w non 3: 29:... mp 02 mm BAILVWSH m AllSNSlNI 129 group from the molecular ion seems to be characteristic for 3'-alky1 ethers of nucleosides. Fragmentation of the molecule is consistent with the stability of the purine ring as reported by others for underivatized nucleo- 25,26 28 and for TMS derivatives of the nucleosides,27 nucleotides 29 sides, or purine bases. Simple cleavage of the glycosidic bond with charge retention on the ring leads to ion B for all three derivatives. This ion is accompanied by the more abundant B + 1 and B + 2 ions, indicat- ing the presence of one or two hydrogens abstracted from the ribose moiety. This triplet at m£3_206-8 is very characteristic for TMS derivatives of adenosine analogues with a free amino group. A methyl group is frequently lost from the TMS group of the B + 1 ion resulting in a large peak at g£g_192 representing ion C. A ribosyl TMS group may become associated with the B + 1 ion giving rise to a small, but characteristic peak at E£g_280. An ion (D) at 212.236 is produced in large relative abundance by cleavage of the molecular ion to include 3:1', 374', and two ribosyl hydrogens with the purine base fragment as with the underivatized nucleoside. When the ribose moiety retains the charge following cleavage of the glycosidic bond, a hydrogen is typically lost leading to ion S, reflected in a small ion at £12 348 for I and g£g_290 for 11 and III. Loss of a labile hydrogen from 3:3' or 3:5' was implicated in formation of the analogous ion for underivatized 2'-3:methy1- adenosine,25 while the ion was not formed for adenosine or 3'-3: methyladenosine. Apparently an alternate mechanism and structure is involved in formation of ion S for the TMS derivatives of these 130 compounds since formation of the ion occurred for I, II, and III and neither hydrogen is available in any of these compounds. Loss of 3:5' with its constituents from ion S frequently occurs resulting in a relatively abundant ion T at mig_245 for I and 213.187 for II and III. The ions discussed above demonstrated the occurrence of a T1] methyl group on either the 2'- or 3'-hydroxy1 of the ribose moiety of II and III by a shift of ions S and T to peaks 58 mass units from their location in the mass spectrum of I. Ions which include various ‘FJ‘U . " Van-Ina: portions of the sugar moiety distinguish between these two alternatives. Ion E (text figure i) was produced when the charge was retained with the purine base fragment which included both r _ .NHTMS H’ + CH 1' / HC 2. m/e 522, R = oms m/e 26h, R = OCH3 I R Text figure 1 131 3:1' and 3:2' as discussed by McCloskey for the underivatized 25 and TMS derivatives of nucleotides.28 Since ion E was nucleosides located at E£2.322 for both I and III and was shifted to 212.264 for 11, it is diagnostic for substitution on the 2'-hydroxy1. Three additional ions were diagnostic in locating the position of alkyl substitutions on the sugar hydroxyl groups of TMS nucleoside deriva- tives and analogous ions did not occur at a significant level with underivatized nucleosides. One is located at g£g_230 and two others are located at glg_217, in the mass spectrum of 1. Analysis of compounds which were methylated on specific, known hydroxyls of adenosine prior to forming TMS-ethers of the remaining hydroxyls enabled us to look more closely at the origin of the above ions. Ion W, at g£g_230, has been well documented in mass spectra of TMS-derivatives of sugar phosphates30 and nucleotides.28 The structure of this ion for TMS-sugar phosphates was proposed to include both the 2'- and 3'- TMS ether groups; for TMS-nucleotides, the origin of the two TMS groups in the ion could not be determined. Our results with partially methylated adenosine analogues indicate that ion W consists of 3:2', 3', 4', and 5' of the ribose moiety with loss of the 3'-ether group (text figure ii). Loss of only the + r—‘A—fi TMS-O-CHg-CHuaCHmCH-O-R mle 250, R=TMS 5' 4' 3' 2' mge 172, R = TMS w Text figure ii 132 3'-ether was indicated since both I andulll produce a peak at y: 230 while a shift of 58 mass units to 913.172 was seen for both II and TMSz-Z', 3'-di-3:methyladenosine (IV) (data not shown). If the 3'- ether was present in ion W, or if the peak at 233 230 for I was composed of two fragments, only one of which involved loss of the 3'-ether, we would expect to observe a peak at £13,172 on the mass spectrum of III. Extensive rearrangement is not required to produce ion W; it could be produced either from ion S or by retention of the charge with the ribosyl fragment in a cleavage similar to that which firm“; produces ion C. Distribution of the charge over 3:2', 3', and 4' in this four carbon fragment probably accounts for its stability, reflected in the high relative intensity of the ion from all these compounds. The peak at 212.217 for I results from extensive fragmenta- tion of the nucleoside and both 332', 3', and 4' (ion 2) and 3:3', 4', and 5' (ion X) (text figures iii and iv) fragments are proposed to + r—*““-—1 TMS-O-CH&“CH=:CH-O-R £13 217, R= TMS 5' 4' 3' X 213 159. R: C113 Text figure iii + CH2 = COR' -CH-0-R 313 217, R= R' = TMS 4' 3' 2' Z m 159, R: CH3, 1" = TMS or R= TMS, R' =-- CH3 Text figure iv 133 contribute to it. Separation of the two components of this ion occurs on the mass spectra of both 11 and IV reflecting the asymmetry of methylation in these compounds. The presence of a methyl group on only the 2'-hydroxyl as in 11 produces a shift of 58 mass units to 212.159 for ion Z, and a peak remains at g£g_217 representing ion X. When both the 2'- and 3'-hydroxyl groups are methylated prior to TMS derivatization (as in IV), ion X is shifted 58 mass units to 212. 159 while ion Z reflects the presence of two methyl groups by a shift of 116 mass units to 913.101° Since the 3' ether is present in both ions, a single peak at g£§_159 represents both X and Z for III. Both X and Z would be produced without extensive rearrangement of the carbon skeleton or its constituents. The greater stability of 2 due to distribution of the charge over three carbon atoms is reflected in the two- to four-fold greater relative abundance of ion 2 with reSpect to ion X in the mass spectra of both II and IV. Several ions are characteristically produced on mass Spectra of the TMS derivatives of sugar compounds due to anomolies of the trimethylsiloxyl function.24 The peak at 212.103 retains one carbon of the ribose ring with a TMS group. The ion at g£§_l47 is formed by loss of a methyl group from the di-TMS ether. Other examples are ions 212.45» 59, 73, 75, and 89. The 212.73 ion is produced in sufficient quantity to be the base peak in most TMS-nucleoside Spectra. Comparison of the fragmentation patterns of the TMS deriva- tive of the compound in fraction 4 from the AG-l column and the TMS derivative of authentic 2'-3:methy1adenosine (11) indicated that they were identical as shown in Table III. Similarly, fragmentation 134 o.mM\mmH o.e~\omH o.mm\mafi e.mH\AH~ H.o~\AH~ n.mH\AHN m.vm\AHN N o.mM\omH o.e~\mmfi o.mm\mafl ~.o\omH o.m\mmH o.o\mAH m.e~\nflm x o.am\om~ 4.Ne\om~ m.mm\om~ m.oM\NAH m.mm\~afl o.em\owH ~.Ae\om~ z H.om\nmfl m.H~\AmH m.vm\flo~ m.o~\awfi m.e~\smfl m.oN\Ho~ m.H~\m4N H m.oH\omN o.HH\om~ m.HH\eom A.4\oo~ m.m\om~ A.o\eon w.m\w4m m o.w\mmm o.o\mNn o.w\NNm m.m\eo~ N.m\vo~ v.m\m5~ m.e\-m m “.mm\omm o.m~\om~ m.a~\om~ m.mM\om~ m.mm\om~ o.om\on~ 4.A¢\omm a m.mN\NmH o.AH\NmH A.mH\~mH A.oH\NmH w.ow\~mH m.mH\NmH n.HH\~mH o w.mfi\wo~ o.NH\mom H.4H\wow o.m4\wow m.n4\mow N.me\mom H.oH\wom N + m w.e\now N.m\aom m.m\ao~ e.ma\no~ n.mH\Aom m.mm\aom H.m\ao~ H + m m.n\oom s.m\oom m.m\oo~ m.m\oom o.o\oo~ o.4m\oom o.4\oom m 4.34\Nm4 H.om\~we N.mz\om4 m.oH\~me m.mH\Nw4 m.mfl\om4 A.oH\oem mH-z H.N\am4 o.~\am4 o.H\HHm o.H\Amv 4.N\Ams m.N\HHm m.N\mmm 2 E<.m o cofiuomhm m cowpomum E<.N v cowuomum m :ofiuomum ocfimocov< :oH . - u u - - - - - u u u u u u - - xuwmaoucfi o>HpmHou\mmmmv u 1 u n u u u - u u u u u u u u u u u .mo>fium>wnov mzb ovfimoofiosc mo ouuoomm mode on» aoum meoH vouuoaow .HHH oHan 135 of the TMS derivative of the compound from fraction 6 is identical to that of the TMS derivative of authentic 3'-3:methyladenosine (111). These results confirm the earlier tentative identifications which were based on paper chromatographic migration behavior in solvent systems A-0 and the release of methanol by perchloric acid hydrolysis 'h of the nucleosides. The occurrence of 2'-3:methyladenosine and 3 3'-3:methyladenosine among the derivatives eluted from the AG-l column (fractions 4 and 6 respectively) demonstrate that methylation of adenosine is occurring to a significant extent when this nucleo- if side is treated with diazoethane in aqueous 1,2-dimethoxyethane. The TMS derivative of the fraction 3 component displays some similarities in its fragmentation behavior to II, but both the molecular ion and M-15 ion have been shifted 14 mass units to g!3_511 and 496 respectively, consistent with the presence of an ethyl group rather than a methyl group in the molecule. Ions S and T are also shifted 14 mass units to 212.304 and 201 reSpectively, indicating the presence of an ethyl ether in the ribose moiety. Ions E, W, and 2, which are diagnostic for the 2'-position, all support the presence of the ethyl ether at this position by the expected shift of 14 mass units. A similar analysis of the TMS derivative of the compound from fraction 5 indicates that it is related to 111 but that it contains a 3'-ethy1 ether rather than a 3'-methy1 ether. The presence of a 3'-ethy1 ether is further supported by the unique occurrence of a large M-29 ion for this compound, consistent with extensive loss of an ethyl group from the 3'-ether as discussed above. The components of fractions 3 and S 136 from the AG-l column have thus been conclusively identified as 2'-3:ethy1adenosine and 3'~3:ethyladenosine, respectively. Alkylation of palmitic acid and characterization of the products. The presence of significant levels of methylated nucleo- sides among the reaction products obtained with diazoethane raised the question as to the possible origin of the methyl groups. The most obvious possibility was the presence of a methyl donor in the alkylating reagent, e.g., the inadvertant generation of diazo- methane during the preparation of diazoethane. An alternate possi- bility was the participation of the solvent in the reaction, perhaps through exchange of a methyl group from 1,2-dimethoxyethane with the alkylating species prior to its reaction with adenosine. Both of these possibilities were examined by employing diethyl ether as an alternate solvent in the diazoethane reaction. Palmitic acid was chosen as an appropriate acceptor molecule since it is soluble in both diethyl ether and 1,2-dimethoxyethane. The esters which formed were characterized by their retention times relative to standards during gas chromatOgraphy and by combined GC-MS. As shown in Table IV, only the ethyl ester of palmitic acid was formed when the reaction was run in diethyl ether. The presence of as little as 0.01% of methyl ester relative to ethyl ester could have been detected under the conditions employed. However, when the reaction was performed with aqueous 1,2-dimethoxyethane as a solvent, both methyl and ethyl palmitate were produced. The above results indicate that the diazoethane preparation was not contaminated with a methyl donor such as diazomethane and 137 oo.H oo.H ~.mm oo.H o.ooH oonoflefioo Haeom ma.o ma.o w.o - Ho.ov opooflsflno Aseooz cowucouon coaucopon posooum noumo :owucouon nonvoum Houmo o>flumflom o>flumaom Hmuou mo « o>fiumaom Hauop mo w whoumo ocmsuoxxonuoeflv-m.a Honuo axgpofiv pudendum one ON: a“ ocwnpoonmfio ca ocozuoouwfia .nfiom owufieamm mo whoumo Hxxflm on» mo acmmnmoumEougo mmo .>H oHan 138 therefore the solvent is implicated as a source of extraneous alkyl groups. A clear precedent for solvent participation in diazoalkyla- tion reactions is not known, but it is interesting to note that methyl ether formation of glucopyranosyl derivatives with diazomethane is catalyzed by the presence of methanol.31 In addition, the use of P‘ methanol as a component of the solvent in esterification of p: hydroxybenzoic acid by diazomethane results in complete formation of the 3fmethyl ether in addition to the ester; in the absence of ' methanol only about 1-5% of the 3:methyl ether forms.32 The possible E. exchange of the ethyl donor with a methyl group from 1,2-dimethoxy- ethane mentioned above could involve formation of a trialkyl oxonium 33 ion intermediate since both the triethyloxonium ion and the tri- methyloxonium ion34 have been prepared as the fluoroborate salts. This intermediate could then either serve as a direct alkylating agent or dissociate to form a methyl donor. Acknowledgments. We are indebted to Dr. C. C. Sweeley for use of the mass Spectral facility, to M. Bieber, Dr. R. Hammond, and Dr. R. Laine for helpful discussions in interpretation of mass spectra, and to J. Harten for running the mass spectra. We also wish to thank Dr. W. H. Reusch and Dr. J. C. Speck for critically reading the paper prior to publication. 10. 11. 12. 13. 14. REFERENCES This work was supported by Grant GB-20764 from the National 3 on Science Foundation and by Public Health Service Research Grant No. CA-13175 from the National Cancer Institute. LMP was a trainee on NIH Training Grant GM-1091. Michigan Agriculture Experiment Station Article No. 6367. . D. Smith and D. B. Dunn, Biochim. Biophys. Acta, 3, 73 (1959). . L. Starr and B. H. Sells, Physiol. Rev., 33, 623 (1969). . 1L Vaughan,Jr., R. Soeiro, J. R. Warner, and J. E. Darnell, Jr., Proc. Nat. Acad. Sci., U.S., 33, 1527 (1967). . R. Price and F. Rottman, Biochemistgy, 3, 4524 (1970). . E. Dunlap, K. H. Friderici, and F. Rottman, Biochemistry, 19, 2581 (1971). . F. Gerard, F. Rottman, and J. A. Boezi, Biochem. Biophys. Res. Commun., 53, 1095 (1972). . Rottman and K. Heinlein, Biochemistry, Z, 2634 (1968). . Zmudzka, C. Janion, and D. Shugar, Biochem. Biophys. Res. Commun., 31, 895 (1969). . Zmudzka and D. Shugar, FEBS (Fed. Eur. Biochem. Soc.) Lett., g, 52 (1970). . M. Bobst, P. A. Cerutti, and F. Rottman, J. Am. Chem. Soc., g}, 1246 (1969). . M. Bobst, F. Rottman, and P. A. Cerutti, J. Mol. Biol., 33, 221 (1969). . O. P. Ts'o, S. A. Rapaport, and F. J. Bollum, Biochemistry, 3, 4153 (1966). . Brahms, J. C. Maurizot, and A. M. Michelson, J. Mol. Biol., 33, 481 (1967). 139 28. .9. ed w \J 1" W‘V 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 140 The abbreviations used in this paper are as follows: BSA, Bis(trimethylsilyl) acetamide; GC, gas chromatography; MS-GC, combined mass spectrometry-gas chromatography; 2'-Am, 2'-g:methy1adenosine; 3'Am, 3'-0-methyladenosine; poly Ae, poly-2'-0-ethy1adeny1ic acid;_poly Am, poly-2'-0- methyladenylic acid; and TMS, trimethylsilyl. _- M. K. A. Khan and F. Rottman, FEBS (Fed. Eur. Biochem. Soc.) Lett., 39, 25 (1972). ' A. D. Broom and R. K. Robins, J. Am. Chem. Soc., 91, 1145 (1965). J. B. Gin and C. A. Dekker, Biochemistry, Z, 1413 (1968). J. Abbate and F. Rottman, Anal. Biochem., 11, 378 (1972). M. Jacobson, J. F. O'Brien, and C. Hedgcoth, Anal. Biochem., 39, 363 (1968). C. C. Sweeley, B. D. Ray, W. 1. Wood, and J. F. Holland, Anal. Chem., 53, 1505 (1970). c. A. Dekker, J. Am. Chem. Soc., 32, 4027 (1965). F. Baskin and C. A. Dekker, J. Biol. Chem., 242, 5447 (1967). D. C. De Jongh, T. Radford, J. D. Hribar, S. Hanessian, M. Bieber, G. Dawson, and C. C. Sweeley, J. Am. Chem. Soc., 91, 1728 (1969). S. J. Shaw, D. M. Desiderio, K. Tsubozama, and J. A. McCloskey, J. Am. Chem. Soc., 93, 2510 (1970). S. M. Hecht, A. S. Gupta, and N. J. Leonard, Anal. Biochem., 99, 230 (1970). J. A. McCloskey, A. M. Lawson, K. Tsuboyama, P. M. Krueger, and R. N. Stillwell, J. Am. Chem. Soc., 99, 4182 (1968). A. M. Lawson, R. N. Stillwell, M. M. Tacker, K. Tsuboyama, and J. A. McCloskey, J. Am. Chem. Soc., 99, 1014 (1971). E. White, V. P. M. Krueger, and J. A. McCloskey, J. Org. Chem., 91, 430 (1972). M. Zinbo and W. R. Sherman, J. Am. Chem. Soc. 99, 2105 (1970). M. Aritoni and M. Kawasaki, Chem. Pharm. Bull., $9, 677 (1970). 141 32. C. M. Williams and C. C. Sweeley, in "Biomedical Applications of Gas Chromatography," Vol. 1, H. A. Szymanski, ed., Plenum Press, New York, 1964, p. 231. 33. H. Meerwein, Org. Syn., 49, 113 (1966). 34. H. Meerwein, Org. Syn., 39, 120 (1966).