m- qr. w—’-‘*'-qu"-”‘q’-WC-v A‘W‘WW "1 0' 0‘. ~~fia~v®0~ ~OQQ—FQQ-O..«Aa«- Q.- W o O o u 1 n o. ‘ o . - . ; _. - _‘.,. ,~-~"\- apcv 1* P ‘ 1.4.. .o. v-. u”. l""° ‘. . .5 r - mu‘otmu-ol'- ..." " I. l _v. Q~n_“,. o;~~ ’1 . ‘ A""J :.f_- a “0" 'c'I .‘~ O .,.a J . .— 02}. g‘yiyqnn-c'i‘fu', ""‘" "Uffvfi‘ofl-g '91:: :v.-’; mm L: ;V!~mh '8‘3911 a: ‘9‘5' - 5' ’4 H w -- .--¢' 5". ’ . «- br - 0‘ 7‘“ "’"" mt. A... v — ,,........._. _k ___4 'I- ' owl. ' u .. ‘- u r... . . . . v .'. ‘ - . . ’ . ‘ - ‘ _ I . 0 5er 'nw )(9 LIBRARY emoens 9mm". menu] \\‘ ~-~, ' -_. _/,. -/J ——I———_— ABSTRACT DETERMINATION OF 2'-Q-METHYL LEVELS AND DISTRIBUTION IN RIBOSOMAL RNA FROM RAT TISSUES By Karen Heinlein Friderici rRNA was prepared from purified ribosomes from normal rat tissues and neoplastic rat livers. The purity of the rRNA was established by gel electrophoresis and UV spectral analysis. 2'-Q-Methyl levels were determined by gas chromatography of methanol released upon perchloric acid hydrolysis of the RNA. The gas chromatography method used for these studies was optimized to permit detection of small variations in meth- ylation. For distributional analysis the rRNA was hydrolyzed to nucleosides by the combined action of pancreatic ribonuclease, snake venom phosphodiesterase, and bacterial alkaline phos- phatase. The 2'-Q-methyl-nucleosides were fractionated from the nucleoside mixture by DE-22 borate chromatography and analyzed by high speed liquid chromatography. The resulting levels and distributions were subjected to statistical analysis using a nested analysis of variance, to determine differences between tissues and confidence in- tervals for the values obtained. No differences in 2'-Q- methyl levels or distribution were found between the differ- entiated tissues or between neoplastic and normal rat liver. DETERMINATION OF 2'-Q-METHYL LEVELS AND DISTRIBUTION IN RIBOSOMAL RNA FROM RAT TISSUES By Karen Heinlein Friderici A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biochemistry 1974 dx (3%) Dedicated to my parents and my husband. 11 ACKNOWLEDGMENTS I would like to express my deep and sincere appreciation to Dr. Fritz Rottman for his support and guidance in this study. I also especially thank Lee Pike for his encourage- ment and many hours of helpful discussion. I thank William Ten Haaf for his help in the preparation of the ethionine treated rat rRNA. I thank Dr. John Boezi and Dr. Steven Aust for serving on my guidance committee. I gratefully acknowledge the assist- ance of Dr. John Gill in formulating the statistical tests used in this study. Last, but certainly not least, I thank my husband, Dan, for his love and patience during the course of this research. iii TABLE OF CONTENTS LIST OF TABLES ..... OOOOOOOOOOOOOIO LIST OF FIGURES...0.00.00.00.00... LIST OF ABBREVIATIONS............. LITERATURE REVIEW................. INTRODUCTION...OOOOOOOOOOOOOOOOIOO MATERIAL AND METHODSOOOOOOOOOOICOO Carcinogen treatment of rats. Collection of tissues........ Preparation of ribosomes..... Preparation of RNA........... Characterization of RNA...... Gas Chromatography........... Determination of nucleoside ratios RESULTSOOOOOOOO0.00000000000000IOOOOOOOO Purity and yield of rRNA........... RNA characterization by GC assay... Determination of nucleoside ratios. DISCUSSIONODOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIBLIOGRAPHY 0.000......00.00.000.000... iv Page vi vii 10 ll 11 11 12 14 15 17 19 23 23 28 45 57 62 LIST OF TABLES Yield of ribosomal RNA from rat tissues.. Determination of polymer phosphate....... Accuracy of CC assay..................... 2'-Q-Methyl levels for rat tissues....... Nested Analysis of Variance: 2'-Q- methyl levels. Ribonucleoside ratios for rat tissue rRNA........ 2'-Q-Methyl distribution for rat tissues......... Nested Analysis of Variance; 2'-Q-methyl distributions.IOD...0......IOOOOOOOOOOOOOOOIOOOOO Number and distribution of 2'-g-methyl nucleosides per ribosome for rat tissues..................... Page 25 31 42 43 44 52 53 54 58 LIST OF FIGURES UV Scan of rat brain ribosomes and rRNA........... Polyacrylamide-agarose gel electrophoresis of RNA from rat tissues.0.00......OOOOOOOOOOOOOOOOOOOOOOO Gas chromatography elution profile................ Time course of hydrolysis: methanol production.... Time course of hydrolysis: production of reaction pIOductSooooooocoooooooooooooooooooooooooooooooooo Dependence of 2'-Q-methyl determination on amount of sample hYdrOIyzedoso.ooooooooooooooooooooooooo. DE-22 borate chromatography of enzymatic hYdrOlySis.I0.0.0000000000000000000000000000...... HSLC elution of ribonucleosides from enzymatic hYdrOIYS18...O...I0.00.00....OOOOOOOOOOOOOOOOOOCOO HSLC elution of 2'-Q-methylnucleosides from enzymatic hydrOIYSiS...0......OOOOCUOOOOOOOOOOOOOO vi Page 27 30 34 36 39 40 49 51 56 DOC E[M.S.] OD Poly (Am,U) SDS Tl TKM UV LIST OF ABBREVIATIONS Cetyltrimethylammonium bromide Degrees of freedom Sodium deoxycholate Estimated mean squares Gas chromatography High speed liquid chromatography Mean squares Optical density unit Heteropolymer of 2'-Q-methyl adenosine and uridine Sodium laurel sulfate Takadiastase Buffer; 0.01 M Tris HCl pH 7.6, 0.01 M KC1, 0.002 M MgClz Ultraviolet vii LITERATURE REVIEW Ribosomes contain three size classes of RNA; 23-288, 16-188 and SS. The two high molecular weight components are transcribed from their DNA as a single 45$ precursor molecule (Scherrer and Darnell, 1962). This precursor is modified and cleaved during processing to yield, in mammals, 18 and 288 cytoplasmic rRNA (Weinberg and Penman, 1970; Perry, 1962). In Hela cells, for example, only about half of the transcrip— tional unit codes for stable rRNA. The other, nonconserved, fractions of the ribosomal RNA precursor molecules are tran- scribed from DNA segments lying between and to the side of the 283 and 188 cistrons (Weinberg and Penman, 1970; Wellanauer and Dawid, 1973). The genes for the 458 RNA are found in the nucleolus of the cell (Busch and Smetana, 1970), and are highly redundant. The extent of gene redundancy of the ribosomal cistrons has been determined by saturating DNA with homologous rRNA. The redundancy of Hela cells was found to be about 1000/ce11 for each of the two high molecular weight components; 28 and 188 (Jeanteur and Attardi, 1969). The genes are located serially on the nucleolar DNA with non-transcribed "spacer" regions separating them (Ford, 1972). SS RNA is very highly redundant with about 7600 copies in Hela cells (Hatlen and Attardi, 1971). This DNA is not found in the nucleolus but is apparently distributed on most 1 of the chromosomes (Brown, 1974). It is synthesized as a larger RNA molecule and trimmed to SS. This DNA also occurs as a serially repeating sequence with nontranscribed spacer regions separating the cistrons. On an evolutionary scale the extent of redundanty of the ribosomal genes roughly parallels the increase in overall genomic content (Birnstiel et a1, 1971). Since the cistrons show such a high degree of multiplicity, especially in higher organisms, the question arises as to whether this large number is constant throughout the life cycle of the cell, in differ- entiated tissues of an organism. Alternatively, they could be synthesized QEDEQXQ either by amplification of a few "master genes" or by some similar mechanism that is sensitive to cer- tain metabolic demands. Amplification of rDNA has been shown in oocytes of amphibians and insects. However, this amplifi- cation appears to be a response to demand for more rRNA. There is no evidence of rDNA amplification in any somatic tissue from vertebrates (Birnsteil, 1971). From partial digests with T1 RNase it is evident that nucleotide sequences of rRNA have changed considerably over the course of evolution. Therefore rDNA is subject to evolu- tionary changes. This raises the question whether traces of past evolutionary development can still be seen in the form of heterogeneity within the highly redundant ribosomal cistrons within one genome. RNA-DNA hybridization experiments show that the multiple cistrons are very similar (DiGirolamo et al., 1969; Retel and Planta, 1968) but there is no evidence that they are identical (Birnstiel, 1971). If variability of the genes does exist, heterogeneity of the population of each of the rRNA components within the same cell could result. If all the rRNA cistrons are not equally expressed in different cell types of the same organ— ism, differences between rRNA preparations from different tissues or developmental stages of the same organism could exist. This would have far reaching implications in determin— ing variations in function between ribosomes. Of especial interest is the possibility of differences between neoplastic and normal tissues. Another possibility is that the various rRNA genes are expressed in a coordinate fashion, though at a different rate under all conditions. In such a case all rRNA samples prepared from the same organism will be equally heterogeneous. The problem of heterogeneity of rRNA in a population of ribosomes has been approached on several levels. Loening I (1968) examined rRNA from several organisms. He found a change in size of the rRNA components as shown on acrylamide gels in different species. He did not find heterogeneity in size be- tween tissues from the same organism. However, this method would not detect changes in chain length of less than 4%. Gould (1966, 1967) looked for differences in the large rRNA component between organisms and tissues from the same organism on 5% gels after limited digestion with T1 RNase. His method produced large fragments with a molecular weight of 10,000-40,000. While he showed large differences between organisms, especially between eucaryotes and procaryotes, he could detect little difference between different tissues from the mouse. Analysis of base composition of the rRNA has produced varying results. Hirsh (1966) found no significant base dif- ferences in rRNA from rabbit reticulocytes and liver. With this method a 40-60 nucleotide change would be required to produce a detectable difference. Amaldi and Attardi (1971) likewise reported no difference between HeLa cells and several normal human tissues. Higashi et al. (1971) have not found base compositional differences in mature ribosomes from rat tumor and normal tissues. Nevertheless, differences in the base composition of 458 precursors in these tissues have been shown (Yadzi, et al., 1969). A more sensitive technique for determining differences between base sequences involves finger printing of oligo- nucleotide fragments. In this procedure radioactively labeled RNA is degraded with Pancreatic RNase A in combination with T1 RNase. The resulting oligonucleotide mixture is then chromatographed by two dimensional high voltage ionophoresis (Sanger, et a1. 1965). Most workers using this technique do not recover the oligonucleotides of rRNA in molar yields. Depending on the bias of the investigator, this can be explain- ed by a heterogeneity of the base sequence expressed, or could be the result of inaccuracies in the method. Using this technique, Fellner, et a1. (1972) found hetero- geneity of the 168 RNA primary sequence in E. coli. One of the fragments which was recovered in less than molar yield was methylated while another fragment with the same base sequence was unmethylated. Attardi and Amaldi (1971) using labeled HeLa cell rRNA compared the oligonucleotides produced by T1 digestion with unlabeled rRNA from normal human tissues. They observed no difference in relative specific activity of the fragments be- tween HeLa cell rRNA and any normal tissues. Other investigators, however, have reported differences based on the oligonucleotide distributions of rRNA. Busch's group (Seeber and Busch, 1971; Wikman, et al., 1970; Nazar and Busch, 1972) has found a significant difference in 28S rRNA from Novikoff ascites cells and rat liver. They have shown a difference in distribution in GpU and GpC dinucleo- tides produced by pancreatic RNase digestion. Hashimoto and Muranatsu (1973) studied rRNA from mouse liver and mouse MB 134 hepatoma. They identified two nucleo- tide sequences (GpGBCp and ApApApApUp) in 18S RNA of liver and kidney which were always absent from the hepatoma. Higashi, et a1. (1971) reported different distributions of the various size classes of oligonucleotides produced by pancreatic and T1 RNase digestion of rat liver and AH-l30 tumor cells. In addition he studied kidney and embryonic liver tissues. It was shown that the oligonucleotide distributions of the adult liver and kidney were similar, but not necessarily identical. A similarity in the frequencies of individual components of dinucleotides was also found on comparison of tumor and fetal liver cells, but the rRNAs of these less differentiated tis- sues differed from those of the less actively growing adult liver and kidney (Higashi, et al., 1972). Small degrees of variation of base sequence in the SS rRNA has also been noted. In each of two strains of E. 321;, two forms have been found (with one in common), in about equal amounts which differ from each other in sequence in one posi- tion. (Brownlee, et al., 1968). Ford (1972) has indicated that in Xenopus there are at least three and possibly five different SS RNA sequences. There is tissue-specific expres- sion of the different sequences, in that only one appears to be expressed in kidney cells but all are found in total RNA from ovary. On the other hand, the SS species from man (Forget and Weismann, 1969) and mouse were shown to be identi- cal (Williamson and Brownlee, 1969). Another possible source of variation in ribosome popula- tion lies in modification of the basic nucleotide sequence by methylation of the rRNA. The role of methylation of rRNA is not completely clear. Addition of methyl groups to the base or ribose moieties of RNA is known to be a very specific, non-random process (Salim and Maden, 1973; Klagsbrun, 1973). Methylation appears to be determined by specificites of the methylating enzymes for sequence and structure (Salim and Maden, 1973). It has also been shown that methylation occurs in all the major RNA species of the eucaroytic cell: rRNA, tRNA and mRNA (Perry, 1974; Desrosiers, et a1, 1974). Methylation of the high molecular weight components of eucaryotic ribosomal RNA is about 902 2'-Q-methyl (Brown and Attardi, 1968; Lane and Tamaoki, 1969; Vaughn, et al., 1967; Salim and Maden, 1973). This methylation occurs while the rRNA 4SS precursor is being synthesized in the nucleolus (Greenberg and Penman, 1966; Zimmerman and Holler, 1967). All or most of the methyl groups which are incorporated into the 458 precursor are conserved during its processing to 18 and 28S rRNA, despite the fact that about one-half of the pre- cursor is lost (Wagner, et al., 1967; Choi and Busch, 1970; Wellanauer and Dawid, 1973). 2'-Q-Methylation is necessary for the proper processing of the precursor to its constituent 28 and 188 rRNA products. In HeLa cells, Vaughn, et al., 0967) have shown 458 precursor synthesized in the absence of methionine is not methylated and is not processed to cytoplas- mic rRNA, but is turned over in the nucleus. While all the methyl groups found in the ribose moeities are added before cleavage of the 458 precursor this is not true for all the base modifications (Salim and Maden, 1973). A late base methylation in ribosomal RNA which has been studied recently, occurs on adenosine in the nucleotide sequence AmSAmSAACUG. This particular sequence has been found in the smaller high molecular weight component (16S - 188) of E, 231; (Ehresman, et al., 1971), yeast (Klootwijk, et al., 1972) and HeLa cells (Salim and Maden, 1973). It has been suggested that this methylation plays some role in the final maturation of the small subunit (Ehresman, et al., 1972) or functions during protein synthesis itself (Salim and Maden, 1973). It is of interest that a mutant of g. 321; which is resistant to the antibiotic kasugamycin is not methylated on this sequence (Hesler, et al., 1971). Heterogeneity of methylation of rRNA has been investigated. Lau and Lane (1971) found upon alkaline hydrolysis of yeast rRNA that certain dinucleotides (CmA, CmU, UmC and UmU) could not be present in every 178 polynucleotide chain. Lane and Tamaoki (1969) reported that not all the sugar and base meth— ylations could occur on every chain in 16S and 28S RNA from HeLa cells. On the other hand Higashi, et al., (1972) using similar methods could find no evidence for differences in total 2'-g-methylation between mouse liver and AH-l30 tumor. Salim and Maden (1973) showed that the majority of methylated products from T1 digestion of HeLa cell rRNA occurred in molar yields. Most of the conclusions concerning heterogeneity or lack of heterogeneity of methylation are open to question since similar results are interpreted differently, depending on the investigator. In view of the fact that methylation of rRNA is of im- portance in its processing, and possibly in its function, it is of interest to look for differences between the normal and cancerous tissues. Abnormalities in the morphology of the nucleolus (the site of rRNA synthesis) generally accompany the neoplastic state. This raises the question of involvement be- tween carcinogenesis and rRNA synthesis and processing. Pos— sible differences in expressed sequences have already been discussed. Certain carcinogens are known to affect the base and ribose modifications of rRNA. Ethionine induction of liver tumors in rats has been studied by Farber's group (Axel, et al., 1967). Administration of L-ethionine to rats leads to the ethylation of proteins and nucleic acids in the liver. The ethylation of nucleic acids after ethionine administration could be produced by S-adenosyl-L-ethionine substituting for S-adenosyl-L-methionine in the reactions catalyzed by nucleic acid methylases. Pegg (1972) found incorporation of H3-L- ethionine into tRNA of rats after one dose of 100 mg. to male rats. Of the ethylation all occurred on bases where methyla- tion usually occurs. He also reported that 23% of the ethyla- tion was on the sugar ribose. Diethylnitrosamine is an acute hepatotoxin and is carcino- genetic in many animals. Prolonged feeding to rats at a high dosage in the diet produces liver tumors. Dimethyl and diethyl- nitrosamines alkylate nucleic acids of the target organ (Swann and Magee, 1971). Presumably, the mode of action is direct alkylation rather than S-adenosyl mediated alkylation. Swann and Magee (1971) studied alkylation of the N-7 position of guanine in animals injected with Gl4-diethylnitrosoamine. Ethylation of microsomal, soluble, and nuclear RNA was found. INTRODUCTION Little is known about the role of methylation of rRNA. It is known to be a very specific, non-random process and important in the processing of rRNA. The nucleolus is the intranuclear site of sugar methylation of rRNA. Since many carcinogens cause significant changes in nucleolar structure or function, it is of importance to determine whether these agents specifically affect ribose methylation of RNA. It has been suggested that certain chemical carcinogens may inter- fere with the normal mechanisms for processing newly synthe- sized nuclear RNA (Al Arif and Sporn, 1972). There is some evidence for variability in basic nucleo- tide sequence of rRNA in different cell types. This hetero- geneity could extend to variability in methylation of the cytoplasmic rRNA of these cells. In this study 2'-Q-methy1ation levels and distributions were studied for rRNA prepared from a variety of rat tissues. Several normal differentiated tissues were investigated. Rats were also treated with carcinogens which were known to affect methylation of tRNA. Liver tissue from these rats was studied and compared to normal rat liver. As ascites hepatoma which is very deviated from its parent liver tissue was also investi- gated. 10 MATERIALS AND METHODS Carcinogen treatment of rats Diethylnitroamine was administered to eight week male Fisher rats in their drinking water at a concentration of 40 mg/l. The rats were maintained on this water for two months and then transferred to tap water for one to two months before they were sacrificed. Ethionine was administered to 150 g female Sprague- Dawley rats by intraperitoneal injection. Four to five ml of a solution of 25 mg/ml L-ethionine in saline was injected once daily for four days. The rats were starved overnight and sacrificed on the fifth day. Novikoff ascites cells were maintained in 160 to 200 3 male Sprague-Dawley rats by intraperitoneal injection of 1.5 to 2.0 ml of fluid. Cells were transferred every seven days. Collection of tissues Animals used for liver preparations were fasted overnight before sacrificing by decapitation. The abdominal cavity and the thorax were opened and the liver was perfused with cold saline through the hepatic vein. The kidneys and testes were removed from the abdominal cavity. The brain was obtained by cutting open the skull with heavy scissors. The tissues were trimmed of connective tissue, rinsed in cold saline, blotted dry and weighed. Livers from diethylnitrosamine treated rats 11 12 were trimmed of normal tissue before tumorous tissue was weighed. The Novikoff ascites cells were drained from the ab- dominal cavity of the rat after six to seven days of growth. The cells were centrifuged from the ascites fluid at 10,000 x g for ten minutes. The red cells contaminating the fluid form a layer below the ascites cells in the centrifuge tube. The ascites cells were scraped from above the red cells and resuspended in 0.25 M sucrose in TKM buffer (0.01 M Tris—HCl pH 7.6, 0.01 M KC§, 0.002 M MgClz) in a volume equal to the original fluid volume. The cells were recentrifuged at, 10,000 x g for ten minutes. Preparation of ribosomes The rat tissues were minced with a scissors and suspend- ed in 0.25 M sucrose in TKM buffer at a concentration of 0.2 g tissue per ml buffer. Homogenization was accomplished by five to six passes in a motor driven glass-teflon homogenizer (0.013" clearance). The suspension was rehomogenized in an- other glass-teflon homogenizer with 0.007" clearance. The homogenate was filtered through two layers of cheesecloth, and centrifuged at 8,000 x g for ten minutes to remove cell debris, mitochondria and nuclei. One tenth volume of 10% DOC was added to the supernatant and this was centrifuged at 28,000 rpm for ninety minutes in a Spinco thirty rotor, to pellet ribosomes. The ribosomal pellet was resuspended in a hand held glass-teflon homogenizer in 0.25 M sucrose in TKM buffer in a volume equal to 102 that of the 8,000 x g 13 supernatant and one tenth volume of 10% DOC was added. Three ml was then layered over six ml 1 M sucrose in TKM buffer and centrifuged at 38,000 rpm for two hours in a Spinco forty rotor. Alternatively, if the volume was large, five ml of the ribosome suspension was layered over ten ml of l M sucrose in TKM buf- fer and centrifuged at 28,000 rpm for three hours in a Spinco thirty rotor. Purified ribosomes were suspended in 0.1 M Na acetate pH 5.1 in a hand held glass-teflon homogenizer and 1/20 volume of 10% SDS was added. The concentration of ribosomes was ap- proximately 50 A260 units of ribosomes per ml buffer. At this point the preparation was usually frozen at -20°C overnight. For preparation of ribosomes from Novikoff cells, the washed cells were suspended in six to seven ml TKM buffer per g cells and allowed to stand in an ice bath for ten minutes. The swollen cells were centrifuged at 2,000 x g for ten min- utes. The supernatant was carefully decanted and the volume of the cells recorded. TKM buffer was added to make the con- centration 0.35 g of cells per ml buffer. The cells were broken by eight passes in a Dounce homogenizer with the tight fitting pestle. Breakage was monitored by a phase-contrast microscope. When cells were well broken, the homogenate was added to an equal volume of 0.5 M sucrose in TKM buffer. This was centrifuged at 30,000 x g for thirty minutes to remove cell debris, mitochondria and nuclei. The rest of the ribo- some purification was then performed as for the rat tissues. 14 Preparation of RNA Ribosomes suspended in 0.1 M Na acetate and SDS were ex- tracted two times with redistilled phenol (H20 saturated) at room temperature. The RNA was precipitated overnight from the aqueous phase with two volumes of absolute ethanol. RNA which was to be used for nucleoside ratio determinations was resuspended in 0.1 M Na acetate pH 5.1 and reprecipitated with ethanol. RNA to be used to determine 2'-Q-methyl levels was dissolved in water and extracted 2x with ether to remove phenol. The RNA solution was made 0.1 M in Na acetate pH 5.1 at a con- centration of about thirty A260 units/ml. The RNA was then precipitated with CTAB (Bellemy and Ralph, 1968) followed by two washes with 70% ethanol/0.1 M Na acetate pH 5.1. (CTAB does not precipitate neutral or negatively charged poly- saccharides). The precipitate was dissolved in 0.1 M Na ace— tate pH 5.1 and precipitated with two volumes of ethanol. The RNA was then dissolved in H20 with a drOp of NH4OH added to facilitate solubilization of the RNA. The RNA solution was stored at -20°C until further analyses were performed. tRNA was prepared from the supernatant of the first high speed ribosomal sedimentation. The supernatant was extracted three times with an equal volume of phenol at room temperature. The aqueous phase was extracted with ether one time and ap- plied to a water washed, 1.5 x 10 cm DEAE Cl‘ column at the rate of one ml/min. The column was washed with about 500 ml of water until no UV material eluted from the column. Any small molecular weight material was eluted with 250 m1 of 0.2 15 M NaCl. tRNA was removed from the column with approximately 75 ml of 1.0 M NaCl. The tRNA was dialyzed against water to remove the NaCl, and lyophilized. Synthetic polynucleotides used as standards for CC analy- sis were polymerized from nucleoside diphosphates with poly— nucleotide phosphorylase (Rottman and Heinlein, 1968). Poly- nucleotides were checked for contaminating nucleoside diphos- phates by chromatography of one OD on Whatman No. 1 paper in Heppel's solvent (n-propanol:NH4OH:H20, 55:10:35). The base ratio of Poly (Am,U) was determined by HSLC. Characterization of RNA The purity of the rRNA and tRNA preparations was checked by 2% acrylamide/0.5% agarose gel electrophoresis prepared by the method of Dingman and Peacock,(l968). A sample of RNA (0.2 A260 units) was applied to the gel slot and electrophoresis was carried out at 200 v for 2 to 2.5 hours at 4°C. The gel was stained in 'Stains all' (0.005% in 50% formamide) over- night. After destaining, the gel was cut into 0.5 cm x 10 cm strips and scanned in a Gilford linear transport gel scanner at 570 nm. For ultraviolet spectral analysis the RNA solution was diluted to about 50ug/ml in TKM buffer. The absorbances at 260 and 280 nm were recorded and ultraviolet spectra were ob- tained from 210 to 310 nm with a Gilford Scanning spectro- photometer. To determine the nmoles of phosphate per A260 unit of RNA a total phosphate assay was performed (Ames and Dubin, 1960). 16 Pyrex or Kimax test tubes (10 cm x 1.2 cm) were soaked over- night in aqua regia to remove any residual phosphate from de- tergent used to wash glassware. RNA was diluted to a concen- tration of 0.2 mg/ml and 50 ul aliquots were placed in the sample tubes. Using the same 50 ul pipette, aliquots were added to 950 ul TKM and the absorbance at 260 nm was recorded. Standard and blank tubes were also prepared using a phosphorous standard solution purchased from Sigma. RNA was ashed by adding 25 ul of 10% MgNOj in ethanol and heating in the flame of a bunsen burner until brown fumes were no longer produced. After cooling the tubes, 0.6 ml of 0.5 N HCl was added, and the tubes were mixed on the vortex mixer until the white ash was completely dissolved. The tubes were topped with marbles and heated for twenty minutes in a boiling water bath. The samples were cooled and briefly centrifuged in the clinical centrifuge to remove condensed liquid from the sides of the tubes. Inorganic phosphate was assayed using the method of Chen et a1. (1956). To each tube 1.4 m1 of Reagent C (10% ascorbic acid: 0.42% NH3molybdate in 1 N H2804, 1:6) was added. The samples were incubated at 45°C for twenty minutes. Absorbance was measured at 820 nm. The standard curve was linear from 0 - 90 nmoles of phosphate. Adequacy of the ashing procedure and accuracy of the phos- phate standard was verified using AMP. The concentration of an AMP solution was determined by measuring the absorbance of an aliquot at 257 nm in 0.01 N HCl. An extinction coefficient 17 of 15.1 x 10'3 was used to calculate the concentration. Vary- ing amounts of this solution were carried through the ashing procedure and the inorganic phosphate was measured as above. Gas Chromatography The determination of 2'-Q-methy1 levels by GC was adapted from the method of Abbate and Rottman, (1972). Hydrolysis tubes were prepared by cutting 0.5 cm pyrex tubing into ten cm. lengths. The tubes were soaked in aqua regia overnight, rinsed, boiled in detergent for fifteen minutes, then in dis- tilled water for fifteen minutes, rinsed well with distilled water and sealed at one end. Six foot glass columns for the gas chromatograph were cleaned in chromic acid cleaning solution overnight. The col- umn was rinsed well with distilled water and washed with nano- grade methanol. The column was then washed with 5% dimethyl- dichlorosilane in hexane and rinsed with methanol and dried. This silanization was necessary to reduce trailing of water due to interaction with active sites on the glass. The column was packed with Porapak 0 100-120 mesh (Waters assoc.). Packing was accomplished by attaching a vacuum hose from a water aSpirator to one end of the column and pouring the packing in the other while vibrating the column with an electric vibrator. Careful packing was necessary to produce a column with adequate resolution and reasonably symmetrical peaks. If this was not accomplished the column was discarded and a new one made. Occasionally a particular lot of Porapak Q was found to be inadequate for the resolution of methanol and 18 acetaldehyde. The t0p three to six inches of the column was replaced every few runs. This was necessary to remove the accumulation of tar and degradation products from the hydro- lysis reaction. The column was conditioned at a flow rate of 0.3 cc/sec at 240°C for five to six hours, then at 200°C at least twenty-four hours prior to use. For hydrolysis, a 40 to 100 u1 aliquot of the RNA solu- tion was introduced into the hydrolysis tube. The solution was centrifuged to the bottom of the tube, frozen in dry ice and lyophilized. Thirty ul of 70% HC104 was added to each tube, the tube was sealed and centrifuged in a clinical cen- trifuge. Hydrolysis was carried out at 100°C, for sixty minutes, in a sand bath kept in an oven. After hydrolysis the tubes were cooled slightly and placed in powdered dry ice. While the contents remained frozen, the tubes were opened and 40 ul of 10 N KOH was added to the upper portion of the tube and resealed. The KOH solution was carefully titrated against the HC104 so that the final solution was slightly basic (pH 7 to 9), since acidic solutions rapidly destroyed the packing. After sealing, the KOH and H0104 were mixed by vortexing. Care was taken not to allow the mixture to become too hot and the tubes were replaced in dry ice to facilitate precipitation of KC104. The tubes were centrifuged at 20,000 x g for twenty minutes. Thirty-five ul of the aqueous supernatant was care- fully withdrawn with a Hamilton syringe and transferred to a cold 6 x 50 mm pyrex tube. To this, 10 ul of a 0.4 ug/ul solution of n—propanol was added, as an internal standard. 19 The tubes were sealed immediately and stored at ~20°C until further use. Gas chromatography was carried out on a Hewlett-Packard F and M 402 gas chromatograph with a flame ionization detector. The column was run at 140°C with a nitrogen flow of 75 cc/min. The optimum hydrogen flow rate varied considerably and had to be determined for each column. The pH of each sample was checked by spotting a small amount on pH paper prior to injection. If the pH was below 6, a small amount of 10 N NaOH was added to the sample before injecting into the column. Sample size was generally 4 ul. The methanol/propanol ratio was obtained by cutting out the area under each peak and weighing the paper on an analyti- cal balance. In each run 2'-Q-methyladenosine was carried through the hydrolysis as a standard. The amount of 2'-Q- methyladenosine was determined by measuring the absorbance at 257 nm and using an extinction coefficient of 15.1 x 10'3. The methanol/propanol ratio produced by this known standard was set at 100% methylation and the values for the other samples were calculated accordingly. The mole percent of methanol was determined by dividing the moles of methanol pro- duced by the moles of polymer phosphate in the sample. Determination of nucleoside ratios in RNA The determination of ribonucleoside and 2'-g-methylnu— cleoside ratios was primarily that described by Pike and Rottman, (1974). RNA was hydrolyzed to its component nucleo- sides by a combination of enzymes. The reaction mixture 20 contained 5 mg of RNA in 1.2 ml 50 mM NH4formate, pH 9.0, and 2 mM MgAc. Bacterial Alkaline phosphatase (Worthington Bio- chemical Corp., BAPF) was dialyzed six hours against 0.05 M NH4HCO3 and 5 units was used for the reaction. Bovine pan- creatic ribonuclease A (Sigma Chemical Co. Type III A) was added at 10 ug per mg RNA. Phosphodiesterase I from Crotalus adamanteus venom (Worthington Biochemical Corp., VPH) was dis— solved in water with a resulting concentration of 5 mg/ml and 0.3 mg was used for the reaction. Hydrolysis was carried out at 37°C for thirty-six hours. Hydrolysis of RNA to nucleosides was assayed by paper electrophoresis of a 10 ul aliquot of the reaction mixture on Whatman No. 1 paper. Electrophoresis was performed in 0.05 M Tris C1" pH 7.8, at 400 v for forty-five minutes in a Gelman electrophoresis chamber. The reaction was considered complete if all visible UV absorbing material (estimated to be greater than 97%) remained near the origin. 2'—Q-Methy1 nucleosides were separated from the normal nucleosides in the reaction by DE-22 borate chromatography. Fibrous DE-22 (Whatman) was recycled with 0.5 M HCl and 0.5 M NaOH, then converted to the borate form with 0.7 M boric acid. A 1.1 x 13 cm column was poured, washed with 250 ml 0.7 boric acid and equilibrated with 150 ml 0.15 M boric acid, prior to use. One ml of the reaction mixture was applied and the col- umn was eluted with 0.15 M boric acid. Three to four ml frac- tions were collected at the rate of 1 ml/min. 2'-QrMethyl nucleosides were eluted in the first twenty m1 following the 21 void volume. Normal nucleosides were eluted with 0.7 M boric acid. The 2'-g-methyl nucleoside fraction was taken to dry- ness by flash evaporation and the boric acid was removed as its methylester by two flash evaporations with methanol. The nucleosides were dissolved in water, transferred to a small test tube and lyophilized. The column could be regenerated and reused about ten times before loss of resolution occurred. Ribonucleoside compositions were determined by High Speed Liquid Chromatography (HSLC) on a Chromatronix SS-2-500 column. The column was packed at 3000 psi with Bio-Rad Aminex A-S cat- ion exchange resin which had been defined and washed with 50% acetone, 50% ethanol and 3 M NH4formate. The buffer used was 0.4 M NH4formate adjusted to pH 4.55 with concentrated formic acid. The ribonucleosides were eluted at 26°C with a flow rate of 33 mllhr and a column pressure of 2300 to 2500 psi. 2'-Q-Methylnucleoside compositions were determined on a ninety cm column of stainless steel tubing, 1/8 in. 0.D. by 0.02 in. I.D. packed with Aminex A-S treated as above. The buffer used was 0.4 M NH4formate in 40% ethylene glycol adjust- ed to pH 4.15. The 2'-Q-methyl nucleosides were eluted at 40°C with a flow rate of fifteen mllhr and the column pressure was 3000 psi. The buffers for both columns were filtered through a millipore filter (0.25 u) and degassed. The buffers were pumped through the columns with a Milton Roy Mini-pump capable of 5000 psi and elution was monitored with a Gilford 22 spectrophotometer using a 40 ul flow cell with a 1 cm light path. Full scale on the chart was generally 0.10 OD. Most nucleosides were monitored at 260 nm, but 2'-Q-methylcytidine was usually monitored at 271 nm. Areas under the peaks were calculated by measuring heights x width at half—height. The areas were converted to relative molar amounts using extinction coefficients determined by measuring a known concentration of each nucleoside at pH 4.55 at the appropriate wavelength. (A and Am, 3M x 10"3 = 15.2 at 260 nm; G and Gm, an x 10..3 = 11.7 at 260 nm, U and Um, an x 10'3 = 10-0 at 260 nm; C and Cm, aM x 10"3 = 6.9 at 260 nm or 3M x 10'3 = 10.0 at 271 nm). These values agreed with those calculated from Circular OR-lO, P. L. Biochemicals Inc. Compositions were determined as the portion of each nucleoside relative to the sum of all four nucleosides. RESULTS Purity and yield of rRNA Purified ribosomes were used to prepare the rRNA for these studies to insure that a well defined population of RNA, free of DNA, was obtained. The procedure used was optimized to pro- duce relatively large yields of ribosomes with high purity. The preparation of ribosomes was carried out in the cold as quickly as possible to reduce degradation of the rRNA. Generally, tissues used for ribosome purification were fresh. Occasionally testes or brain were frozen at -80 C for use later. Freezing reduced the yield of ribosomes consider- ably. Since the rRNA extracted from frozen tissues showed little degradation the reduction in yield was presumably due to difficulty in breaking frozen cells.. The yield of rRNA varied between tissues (Table 1). This is probably a reflection of the efficiency of cell breakage as well as differences in actual ribosome number. Rat tissue homogenates were centrifuged at 8,000 x g to remove nuclei and cell debris, instead of 10,000 or 15,000 x g as is generally done. It was found that microsomes tended to be lost in the higher speed centrifugation. Due to the lower speed spin the ribosomes may be slightly contaminated with mitochondria. UV absorbance spectra were taken at various stages during the rRNA preparation. As can be seen in Figure l, centrifugation 23 24 through 1M sucrose greatly reduced the amount of contaminating proteins in the ribosome preparation. Phenol extraction to remove ribosomal proteins, and ethanol precipitation, results in a 260 nm/ 280 nm absorbance ratio of 2.03 to 2.09. Purity of the rRNA was also analyzed by 2% polyacrylamide, 0.5% agarose gel electrOphoresis. Figure 2 shows the profiles obtained for various tissues. Only kidney rRNA showed con- siderable degradation bands. These could be reduced or in- creased depending on the rapidity with which the ribosomes were purified. To calculate mole percent of 2'-Q-methyl nucleosides for the RNA it was necessary to quantitate the number of nucleo- tides in a given RNA sample. This was done using the total phosphate assay to determine polymer phosphate per A260 unit of RNA. Typically, rRNA from rat tissue was about 130 nmoles/ A260 (Table 2). CTAB precipitation of RNA used for the GC analysis is important in that it elimates neutral and basic polysaccharides. Polysaccharides frequently contain methoxy residues which would contribute extraneous methanol during the methyl analysis of RNA. Plant tissues present a real problem in this respect, since they contain high levels of polysaccharides (Jakob and Tal, 1973). Polysaccharides with a negative charge will co- purify with RNA in the CTAB precipitation. Table 1. Yield of ribosomal RNA from rat tissues 25 — f Grams tissue A260 units / g tissue Rat Ribosomes rRNA 'Rat liver 15.0 25 12.0 Rat brain 1.7 10 8.0 Rat testes 5.2 21 12.8 Rat kidney 2.0 24 14.5 Rat liver Ethionine treated 8.0a 16 15.0 Rat liver b Nitrosodiethylamine 12.0 25 15.2 Novikoff ascites cells 6.5 35 18.0 aWhile rats generally used in these studies were young adult males (250-300 g ), ethionine treated rats were smaller females (150- 200 g ). eight represents areas of liver which were tumerous. tissue was discarded. Healthy 26 Figure 1. UV scan of rat brain ribosomes and rRNA. All measurements were at pH 7.8 in 0.01 M Tris Cl", 0.01 M KCl, and 0.003 M Mg C12. A) Ribosomes sedimented from 0.25 M sucrose in TKM; A260/A280 = 1.42. B) Ribo- somes purified by centrifugation through 1.0 M sucrose; A260/A280 = 1.77. C) Purified ribosomal RNA after ethanol precipitation; A260/A230 = 2.04. 27 u q _ _ _ s. 0 1| Ru \_ \\ \ \ \\\ O \\ .II“\ [9 \\\ \\\\ \\ \11 \ \‘ \\ \X\\. \\\H \\\\\ m \ x :2 \ w [I / I, .2 v I. \‘III‘ \ A1. llllllll O ““““““ 3 Bl 1J2 CIIII .m L — 2 O. 5. I 0 mUZwa own as c was mamamm .N coaumnamuum .OH mwamm ou .m aowumoomau« .H swamm Baum sowumooouum as mwomno moumofivaw AHm Ou musofia use some moo aoum woman uumno ow mwomno mmumuavofi noxomun ownuws mou< .oawmoum oofiuoao maamwwOumaowno mew .m muowam 34 m ouowfim Al m2: 35 .oz oaoas H do vmauowuom aofiumowaumumv oumuwamso .ooauusvowm Hoomzuoz “mamhaoumh: mo mmusoo mafia .q madman 36 e muomfim m _m>Jomo>I no NE: on ON 0. _ 11 - ‘\. _ CF |.|.\ 0.. ON 'IAHLBW'O". 2 124308 3:! 37 Figure 5. Time course of hydrolysis; produc- tion of reaction products. A) Furfural, B) Acetone, C) Acetaldehyde. PEAK HEIGHT/PROPANOL PEAK AREA 38 0.0M " 0.010 " 0.2 " B 0.. 1 \ .\._l . 02 u / \°\. \. OJ - / \_ J l l I l IO 20 30 40 60 60 TIME OF HYDROLYSIS (min) Figure 5 39 Figure 6. Dependence of 2'-Q-methyl determina- tion on amount of sample hydrolyzed. Determinations were made in duplicate. A) E. coli rRNA, B) Novikoff cell rRNA. PERCENT 2"0" METHYL 40 0 Z - \ \. \ 0““. ‘~_____‘__ 0 0 OJ *- fi 1 l l 2.0 4.0 6.0 I T V 3.0 *- 2.o - ’ \. \. ‘.—. . LoL I.O , ' 2.0 3.0 uMOLES POLYMER PHOSPHATE Figure 6 41 easily lost during later manipulations. Poly (A) was assay- ed by itself to make certain that it contained no 2'sg-methyl components. The accuracy of the CG assay was investigated using RNA with known 2'-Q-methyl contents. The results of this study are shown in Table 3. As can be seen, the method is accurate over a range of 0-100% 2'-Qfmethylation. The 2'-g-methy1 levels for the various rat tissues is shown in Table 4. A nested analysis of variance was used to look for differences between the tissues (Table 5). An over- all tissue difference was found at the 0.001 confidence level (F-9.18; critical value F0.001,5,19-6.18). Using Scheffe's procedure each of the tissues was then compared to liver. Kidney rRNA was the only tissue with a significant difference from liver rRNA at the 0.05 confidence level. However, this tissue was also the only one in which degradation was a prob- lem. Therefore Cochran's test was used to examine variability in the preparation of this tissue. It was found that kidney was the only tissue in which there was significant variability between preparations (C - 0.66; critical value €0.05,6,2 . 0.62) It is probable that kidney rRNA lost some of its non- methylated segments during isolation and thus showed an arti- factual increase in 2'-grmethy1 levels. Some of the kidney preparations showed less rRNA degradation and these gave lower 2'-Qfmethyl levels (Table 4). Therefore the variation in 2'fiQ-methyl levels between preparation of kidney rRNA was probably due to greater or lesser degrees of degradation. 42 Table 3. Accuracy of CC Assay. ====:;l .14; Percent 2'1Q-methyl Experimental Expected E, coli rRNA 0.13 o.la,o.12b, 0.17c Novikoff rRNA 1.44 1.47d Poly (Am,U) 8.43 8.6e Poly (Am) 99.5 100 Poly (Um) 106.0 100 Poly (A) N.D. O a Nichols and Lane (1966). bIsaksson and Phillips (1968). c Hayashi, et a1. (1966). d Egawa, et a1. (1971). eDetermined by HSLC Method. N.D. none detectable. 43 Table 4. 2'7Q-Methyl levels for rat tissues. t M Percent 2'7Q-Methyl Average for each Average for each preparation tissue Rat liver rRNA Prep. 1 1.57 (5)8 Prep. 2 1.58 (5) Prep. 5 1.47 (6) 1.51 + 0.055b Prep. 4 1.44 (9) ‘— Prep. 5 1055 (6) Prep. 6 1.55 (14) Rat testes rRNA Prep. 1 1.44 (10) Prep. 2 1.50 (6) l.47-+ 0.062 Prep. 5 1.47 (6) —— Prep. 4 1.47 (6) Rat brain rRNA Prep. 1 1.50 (6) 1.49 + 0.086 Prep. 2 1.49 (6) ‘- Rat kidney rRNA Prep. 1 1.65 (9) Prep. 2 1.76 (8) 1.67 :_0.062 Prep. 5 1.76 (6) Prep. 4 1.49 (5) Novikoff ascites rRNA Prep. 1 1.46 (8) Prep. 2 1.49 (9) Prep. 5 1.47 (10) 1.44 :_0.049 Prep. 4 1.59 (7) Prep. 5 1.41 (17) Prep. 6 1.45 (8) Diethylnitrosamine treated rat liver rRNA Prep. 1 1.40 (8) 1.58 + 0.084 Prep. 2 1.57 (15) .- L-ethionine treated rat liver rRNA Prep. 1 1.60 (2) Prep. 2 1.55 (5) 1.57 + 0.081 Prep. 5 1.58 (3) .7 8Number in parenthesis indicates the number of determinations. 95% confidence interval calculated from nested analysis of variance. Table 5. Nested Analysis of Variance; 44 2'1Q-methyl levels. s.s. d.f. M.S. E[M.SJ f Ratio (df) Tissues 1.45797 6 0.259662 0.025774 9.299 (19) Preparations in each tissue 0.50189 20 0.025094 0.025094 GC runs in each preparation 0.47167 65 0.007256 0.007256 Samples in 00 runs 0.56452 111 0.005282 0.005282 45 Rats treated with ethionine could possibly incorporate 2'fiQ-ethyl groups in their rRNA. It has been shown that 2'—Q-ethylation does occur in tRNA from animals given a single dose of H3-Léethionine. Perchloric acid hydrolysis of 2'-Q- ethyl ribose releases ethanol and can be detected by the gas chromatography method. RNA from ethionine treated rats was lyophilized 3-4 times to remove residual ethanol remaining from ethanol precipitation of the rRNA. No ethanol was de- tected in rRNA from ethionine treated rats. tRNA from these rats was also studied. There was no decrease in methylation of this tRNA and no ethanol was produced. This could indicate that the treatment with ethionine was not sufficiently ex- tensive to detect incorporation into the 2' position over the large background of non-altered tRNA. Determination of nucleoside ratios Enzymatic hydrolysis was performed as described in ma- terials and methods. The action of Snake Venom phosphodiester- ase was not always consistant. After repeated freezing and thawing, the enzyme became less potent and a contaminating deaminase seemed to become more active. When this occurred an Im peak was found in the Z'fig-methyl distribution analysis with a corresponding decrease in Am. Phosphodiesterase pur- chased from Sigma Biochemicals showed an even greater deamin- ase activity than that from Worthington. To overcome decreased activity of the phosphodiesterase and increased deaminase activity several precautions were taken. Every reaction was monitored by paper electrOphoresis after 46 twenty-four hours of incubation. If UV absorbing material was found in the dinucleoside, indicating incomplete hydrolysis, a new enzyme solution was used for subsequent degradations. To reduce deamination, the reaction was not stored after com- pletion of hydrolysis but applied immediately to the DE-22 borate column. 2'-Q-Methy1 nucleosides were isolated from the normal nucleosides in RNA since they represent only one to two percent of the total mixture. This was accomplished by taking advantage of the ability of ribonucleosides to form a complex with the borate ion through which they could be re- tained on an anion exchange column. The 2'-Qf methyl nucleo- sides could not form this complex with their blocked 2'- hydroxyl group and were not retained. Figure 7 shows a typical elution profile from the DE-22 column. After repeated use of the column the volume in which the 2'-Q-methy1 nucleosides eluted became larger. Ribonucleoside ratios were determined by direct injection of small amounts of the reaction mixture, prior to borate chromatography, into the Chromtronix SS-2-500 HSLC column and elution with 0.4 M NH4formate at pH 4.55. This had to be done soon after the reaction was complete as deamination could occur if the reaction was allowed to remain at room temperature. Freezing the reaction mixture caused precipitation of guanosine and phosphate salts. Alternatively, the non-2'-Q-methyl nucleo- sides (ribonucleosides) could be collected from the DE-22 borate column with 0.7 M boric acid elution. An elution profile for the ribonucleotides compositional analysis is shown in Figure 8. 47 If the hydrolysis reaction was not complete, a peak eluting before uridine was seen. This was due to negatively charged nucleotides appearing in the solvent front. The ratios of the ribonucleosides for the various rat tissue is shown in Table 6. All the tissues had very similar nucleoside ratios. This data was not subjected to statisti- cal analysis. The ratio of the 2'-Qfimethyl nucleosides was determined by HSLC on a 0.02 in. by 90 cm column eluted with 0.4 M NH4 formate in 40% ethylene glycol, pH 4.15 (Figure 9). At this pH all the 2'-Q-methyl components are well separated. The pH of this buffer must be checked each day, as the pH tends to increase as the buffer stands. At higher pH the Gm and Am peaks become less well resolved. The ratio of the 2'-Q-methyl nucleosides for the various tissues is shown in Table 7. Using a nested analysis of vari- ance, no difference between tissues was found for any of the 2'-g-methy1 nucleosides at the 0.01 confidence level (Table 8). 48 .o .< .D .0 ma mopfimomaosoonfiu mo oofiusao mo “mono .oeoo oeeoe z k.o rose movamooaosaonau mo sowusam Am .vwom canon z mH.o now? movwmooaos: HmcumaLdI.N no eoeooam Ad .mvonuoz vow mHmHuoumz as wonwuomov mm woumowwv mus mzmH uo>HH umm .mfimzaouumn oHumamnoo wo zzemuwoumaouno summon NNImo .m musmwm 49 30 W“ 092 SONVBEIOSBV I )- 8 .- /./ '- 8 1...... l..._ K.——f )- h‘ fi-Ok .- g ‘0\ ‘0 ____.__l C )- fin‘ d 9 \,_ i. .z. . _ . o 73 a: T r ‘d r- 2 L e - 0"T"’—.——8 -'¢) O\. I I ./ 0| '§ 0' O FRACTION NUMBER Figure 7 50 Figure 8. HSLC elution of ribonucleosides from enzymatic hydrolysis. Column temperature was 26°C; col- umn pressure was 2400 psi. Approximately three nmoles of nucleosides from rat liver rRNA were injected. Order of elution is U, G, A, and C. '1 j 2:. 33 822.283 1 5 1 _ 5 I ifll .llnh ”(Inglis 5.113104- 441E.Iirlllll Fl. I. (I .I . . u . its . . . I. | -I a. i.ll-...fim!i£.¥w in]. . . aria». .... .i 30 20 'IO 1.0-)- 5 TIME (min) Figure 8 52 m.mn m.mH m.mm w.rH m .eoum m.mn o.Hm m.mH m.mH w.mm :.Hm o.mH m.>H H .eoum oz w.om n.eH m.em s.eH m .eoum o.Hm m.Hn m.>H m.eH n.2m o.em m.>H H.eH m .ooem w.Hm m.>H m.em m.eH H .ooem azmu HoeeHx pom ®.Hm H.AH o.em e.sH n .oopm o.Hn m.Hn m.>H >.>H m.nn e.nm m.>H m.>H m .eoom m.Hm w.>H 6.2m m.eH H .eoum mzmH mmumou umm m.Hn m.mH :.Hm m.eH m .oopm o.Hm m.on n.eH e.eH H.em m.Hm m.eH o.>H H .ooem HH pom .o>m .aoue .o>w .noua .o>w .eoun .o>m .nown. ongvfluhp oawkua. mammoamow. oawmoaov< .dZMu oommwu umn wow moHumw ovwmoofiosaonwm .0 manna 53 .mocmwum> mo mammHmem pmumoc scum wouwasofimo amphmucw mucouwmcoo fimm n .mooHumumeoua oommwu mo nomads moumowvcw mamonuooume a“ possum“ Hmv «zen po>HH 65H H mHm SH H 5: 5.6 H 58 3.6 H Hmom Hop 3833:5338 RH H THm No.6 H 08 2.6 H Tom 9.0 H TR H8 42% 3:3... 39:32 RH H mom mwd H 6.9 $6 H Won id H Hon 3 same 85.3 03H SH HHHN SH 0112 HooHHom 88.3.8 Q as. so; .2 05H H mew No.6 H 92 E6 H Pom Hm.o H {on 3 $20 338 one 3H H mdm 8.6 H 4.3 8.6 H Son HES H Ton may «zap 02,: ”am So a: so e< .8533 so... How ooHoothoHo Harooz.d-.m .2. Some 54 Table 8. Nested Analysis of Variance; 2'-Qfmethyl distributions. 8.8. .f. M.S. E[M.S.] f ratio (df) 2'19-Methyladenosine Tissues 11.79711 5 2.59425 0.55579 4.49 (5,11) Preparations per tissue 5.45585 11 0.49580 Runs per preparation 10.56749 44 0.25562 2'1Q-Methylguanosine Tissues 8.59806 5 1.67961 1.52400 1.10 (5,15) Preparations per tissue 12.25785 11 1.11255 Runs per preparation 11.86085 44 0.26956 2'jQ-Methyluridine Tissues 47.75257 5 9.5508 1.85795 5.14 (5,10) Preparations per tissue 18.21796 11 1.65618 Runs per preparation 12.05985 44 0.27409 2'fiQfiMethylcytidine Tissues 24.16472 5 4.85294 1.57810 5.06 (5,10) Preparations per tissue 15.75585 11 1 45217 Runs per preparation 19.00700 44 0.45198 Critical f ratio Um, f0.01,5,1o=5-643 Am, fo.01,s,11=5-323 Cm, f0.01,5,io-5-6’+; Gm, f0.01,s,1s=4-61+3 55 umvuo In< .au .a< .80 .6: ma soauoao mo .vmuomneH mos HH umn aoum mmvfimomause mo moHoae manna maoumsaxoue .Hme comm mas muommmue easaoo moan.am mos ououmuoeamu aasnoo .mfimhaouvH: oaumshuao Bonn movaooHosaHmnuoaldl.N mo coausao UAmm .m musth 56 m ouawfim AEEV mi: 00 on o . & Jud is. 33NV880$8V DISCUSSION The use of the gas chromatography technique for determina- tion of 2'-g-methyl levels in RNA has several advantages. Re- latively small amounts of RNA can be analyzed without the ne- cessity of radioactive labeling procedures. This means that tissues which are difficult to label or require high levels of precursor radioactivity, can be studied. Also, the problem of pool sizes and kinetics of labeling do not arise. The G.C. procedure lends itself well to studies involv- ing analysis of numerous RNA samples. However, certain as- pects of the method require attention. It is very important that methanol and acetaldehyde are adequately resolved by the Porapak column. The methanol peak becomes difficult to quanti- tate if it trails into the acetaldehyde. This resolution is difficult to attain and the precautions described in Materials and Methods must be observed. When the procedure is care- fully done, reasonably precise data can be obtained. If one assumes 6520 nucleotides in rRNA (28 S - 4500, Seeber and Busch, 1971; 18 S a 1900, Qualiarotti et al., 1970; 5 S 3 120, Brown, 1974), the number of ribose methylated nucleotides in liver rRNA is 98.51’3.5 at a 95% confidence interval. As is shown in Table 9 a difference of six to ten nucleotides per rRNA could be detected. 57 58 By combining the results from 2'-Q-methyl level analysis with those from the distribution studies, the absolute amount of each 2'-g-methy1ated nucleoside can be determined (Table 9). A change of only two or three nucleosides could be detected at the 90% confidence limits. The use of proper statistics is important when studying suspected quantitative differences. It is not adequate to look at one or two experiments and conclude that two similar samples are quantitatively the same or different. Confidence limits for the technique should be established since only in this way is it possible to determine how much variation could be detected and if differences are significant or merely due to imprecision in the method. rRNA was prepared from purified ribosomes rather than from a high salt precipitation of cytOplasmic RNA, which would, admittedly, be simpler and quicker. However, RNA from puri- fied ribosomes eliminated the possibility of contamination by DNA which would interfere with both 00 analysis of 2'-Qfmethy1 levels and the HSLC determination of 2'-Q-methyl distribution. It is known that endonucleolytic "nicks" in rRNA can occur in intact ribosomes. These "nicks" produce discrete RNA molecules with lengths smaller than eighteen or twenty- eight S. Presumably this cleavage occurs in intact ribosomes in portions of the RNA which are not protected by proteins or 2'-g-methyl groups. It was thought that if degradation did occur during purification, these shorter pieces would be re- tained in the RNA extracted from whole ribosomes. Apparently 59 .aom mg Ha>uoucw ooeuvamaoa mew suHsmou any .muHEHH ooeopwweoo Hosea on» mean: mp vocwsuouov swoon noses any .ovaoooaoso some you uwBHH oocopHmaoo *mm Home: ecu an mopHaomHose Hanuosimr.m ueoowoe may mo uaswa moeovwmeoo fimm Home: 0:» weHHHe uHuHsB Hp woeHsumuov mes swamp Home: use .ueoaoue novHuooH05: Hhsuosrmr.m Hquu mo Hones: owuuo>m onu mp voHHnHuHos was wwwuoofiosc nous mo unmouoe one .QEOQOQHH use ovuoooaosc nous no access wan Mom n .szu m m + 2 + m8 2.32? use 6me ow nonmawumo was mopHmooHooe mo unease HmuOu 0:9 .wammwu nous mo mHo>oH HHfiuoauOu.m map you Hu>uou:H ooeopfimeoo.mmm osu wean: woumfisofiwo was osmmfiu some you nonwoooaoso Hmnuosrmrem mo Henson HmuOu may: H.mH New m.mH H23 TH H 92 m.mH 6.8 m.mH 0.8 42% quH sou oewfismouuweaxnuowa m.mH {om HH H 02 TH H 9mm «H H 0mm m.mH 58 so: 8:83 33:82 o.n H mom H.m H 6.: PNH 0mm ONH 6.8 Pm. H mém «zen 53s use 11m H T? 6H H H.mH TH H 9mm TH H H.mm 6.: H 23 ezmu souoou use H.mH Tom HHH H.mH H..HHo.on H.HHm.mm m.nH 0me see 33: use 50 B: so Ed uoBOmonHH use momwmoofiuss assemonwu use wouoooHooc coma mo_uonssz woumfi>nuoaumu.m Hmuoa .mosmmwu umu How oEOmonHH use mopHmooHooc HASuoaimr.m mo :oHuonHHume was Honasz .m wanna 60 this is not true when extensive degradation occurs, as evi- denced by the preparations from kidney. In the case of kidney rRNA the 2'-Qrmethyl distribution was the same as for the other tissues. The 2'—Qrmethyl level, however, was significantly higher. This could result if there were selective loss of non-methylated portions of the rRNA. Methylation does protect RNA from the action of various nucle- ases (Dunlap, et al., 1971; Stuart and Rottman, 1973). There- fore preferential 1033 of unmethylated sequences could have occurred either by exonucleolytic attack or by endonucleo- lytic production of small pieces of RNA which would have been lost during ribosome purification or ethanol precipitation of rRNA. To study possible differences between normal and cancerous tissue, three different carcinogenic states were chosen. Novikoff ascites hepatoma has been maintained for several years and has deviated considerably from its parent liver cells. It, therefore, represents a tissue which is dividing very rapidly and has no morphological resemblance to normal liver. Diethyl— nitrosamine is a potent carcinogen which is known to alkylate RNA bases. The mode of alkylation is presumably direct with no enzymatic action required to insert the alkyl group in the RNA. L-Ethionine also alkylates RNA but is thought to act through an S-adenosyl-ethionine intermediate. With S-adenosyl- ethionine substituting for S-adenosyl-methionine, ethylation can occur at sites that would normally be methylated. In this case the possibility of 2'-Q-methylation could exist. These 61 three types of abnormal tissues have been shown to contain altered tRNA alkylation when compared to normal liver (Pegg, 1971; Nau, 1974; Swann and Magee, 1971). The question he- comes; could these alterations extend to rRNA also? In this study no differences in methylation of rRNA were found under any conditions. The studies with ethionine were not conclusive, however, since no ethylation of tRNA, from the treated rats, was found. Long term administration of ethionine to rats in their diet might still result in detect- able 2'-Q-ethylation of rRNA. This study would probably be better approached by use of radioactive ethionine. The 2'-Q- alkylated products could then be isolated with the DE-22 bor- ate column and the products characterized with the HSLC method. It is significant that no variation in level or distribu— tion of methylation was found either between normal tissues or between normal and carcinogenic material. Depending on the particular nucleoside, differences of methylation of two to five nucleosides would have been detected. This indicates that if variation in base sequences does exist it does not significantly alter the methylation pattern of the rRNA. The uniformity of methylation of rRNA, under conditions that affect tRNA alkylation, may indicate that prOper methylation of RNA is vital in the processing of precursor RNA into mature rRNA, and that a complete complement of 2'-Qfimethyl groups are re- quired for biological function. This constancy of structure suggests a major role of 2'-Q§methylation. BIBLIOGRAPHY BIBLIOGRAPHY Abbate, J. A. and F. Rottman, 1972, Anal. Biochem. 41, 378. Al-Arif, A. and M. Sporn, 1972, Proc. Nat. Acad. Sci. USA 62, 1716. Amaldi, F. and G. Attardi, 1971, Biochemistry, 10, 1478. Ames, B. N. and D. T. Dubin, 1960, J. Biol. Chem. 235, 769. Axel, R., I. B. Weinstein and E. Farber, 1967, Proc. Nat. Acad. Sci. USA 58, 1244. Bellamy, A. R. and R. K. Ralph, 1968, in "Methods in Enzymology" (L. Grossman and K. Moldave, eds.), Vol. XII, Part B, 156, Academic Press, New York. Birnsteil, M. L., M. Chipchase and J. Spiero, 1971, Frog; Nucl. Acid Res. Mol. Biol.‘1l, 351. Brown, D. D. 1974, Federation Proceedings, Symposium. Brown, G. M. and G. Attardi, 1968, Biochem. Biophys. Res. Comm. 22, 298. Brownlee, G. G., F. Sanger, and B. G. Barrell, 1968, J. Mol. Biol. 34, 379. Busch, H. and K. Smetana, 1970, "The Nucleolus," Academic Press, New York. Chen, P. S., T. Y. Toribara, and H. Warner, 1956, Anal. Chem. ‘gg, 1756. Choi, Y. C. and H. Busch, 1970, J. Biol. Chem. 245, 1954. Desrosiers, R., K. Friderici and F. Rottman, 1974, Proc. Nat. Acad. Sci. USA, in press. DiGriolamo K., E. Busiello and M. DiGirolamo, 1969, Biochim. Biophys. Acta 182, 169. Dingman, C. W. and A. C. Peacock, 1968, Biochemistry l, 659. 62 63 Dunlap, B. E., K. H. Friderici and F. Rottman, 1971, B10- chemistry 10, 2581. Egawa, K., Y. C. Choi, and H. Busch, 1971, J. Mol. Biol. _§J 565. Ehresmann, G., P. Fellner and J. P. Ebel, 1971, FEBS (Fed. Eur. Biochem. Soc.) Lett. 3, 309. Ford, P. J., 1972, Proc. of the Biochem. Soc. 528. Forget, B. G. and S. M. Weisman, 1969, J. Biol. Chem. 244, 3148. Greenberg, H. and S. Penman, 1966, J. Mol. Biol. 21, 527. Gould, H., S. Bonanou and K. Kanaligham, 1966, J. Mol. Biol. 22, 397. Gould, H. J., 1967, J. Mol. Biol. 29, 307. Hashimoto, S. and M. Muranatsu, 1973, Eur. J. Biochem. 33, 446. Hatlen, L. E. and G. Attardi, 1971, J. Mol. Biol. Hayashi, Y., S. Osawa and K. Miura, 1966, Biochim. Biophys. Acta 129, 519. Helser, T. L., J. E. Davies and J. E. Dahlberg, 1971, Nature, New Biol. 233, 12. Higashi, K., T. Matsuhisa, S. Gotoh, K. Nishinaga, and Y. Sakamoto, 1971, Biochim. Biophys. Acta 232, 352. Higashi, K., S. Gotoh, K. Nishinaga and Y. Sakamoto, 1972, Biochim. Biophys. Acta 262, 320. Hirsh, C. A., 1966, Biochim. Biophys. Acta 123, 246. Isaksson, L. A. and O. H. Phillips, 1968, Biochim. Biophy_. Acta 155, 63. Jakob, K. M. and I. Tal, 1973, Biochem. Biophys. Acta 308, 296. Jeanteur, Ph. and G. Attardi, 1969, J. Mol. Biol. 45, 305. Klagsbrun, M., 1973, J. Biol. Chem. 248, 2512. Klootwijk, J., R. C. van den Bos, and R. J. Planta, 1972, Abstr. Commum. Meet. Fed. Eur. Biochem. Soc. 8, No. 492. 64 Lane, B. G. and T. Tamaoki, 1969, Biochim. Biophys. Acta 179, 332. Lau, R. Y. and B. C. Lane, 1971, Can. J. Biochim. 49, 431. Loening, U. E., 1968, J. Mol. Biol. _8, 355. Nau, F., 1974, Biochemistry 13, 1105. Nazar, R. N. and H. Busch, 1972, Cancer Res. 52, 2322. Nichols, J. L. and B. G. Lane, 1966, Biochim. Biophys Acta 119, 649. Pegg, A. E., 1972, Biochem, J. 128, 59. Perry, R. F., 1962, Proc. Nat. Acad. Sci. USA 48, 2179. Perry, R. P. and D. E. Kelley, 1974, Cell 1, 37. Pike, L. M. and F. Rottman, 1974, Anal. Biochem, in press. Quagliarotti, G., E. Hidvegi, J. Wikman, and H. Busch, 1970 J. Biol. Chem. 245, 1962. Retel, J. and R. J. Planta, 1968, Biochim. Bigphys. Acta, 169, 416. Rottman, F. and K. Heinlein, 1968, Biochemistgy l, 2634. Salim, M. and B. E. H. Maden, 1973, Nature, 244, 334. Sanger, F., G. C. Brownlee, and B. G. Barrell, 1965, J. Mol. Biol. 13, 373. Scherrer, K. and J. E. Darnell, 1962, Biochem. Biophys. Res. Comm. 1, 486. Seeber, S. and H. Busch, 1971, Cancer Res. 31, 1888. Stuart, 8. E. and F. Rottman, 1973, Biochem. Biophys. Res. Comm. 55, 1001. Swann, P. E. and P. N. Magee, 1971, Biochem. J. 125, 841. Vaughn, M. H., Jr., R. Soeiro, J. R. Warner and J. E. Darnell, Jr., 1967, Proc. Nat. Acad. Sci., USA 58, 1527. Weinberg, R. A. and S. Penman, 1970, J. Mol. Biol. 7, 169. Wellanauer, P. K. and I. B. Dawid, 1973, Proc. Nat. Acad. Sci. USA. 12, 2827. Wikman, J., E. Howard, and H. Busch, 1970, Cancer Res. 0 773. 65 Williamson, R. and G. G. Brownlee, 1969, FEBS (Fed. Eur. Biochem. Soc.) Lett. 1. 309. Yadzi, E., T. S. Ro-Choi, J. Wikman, Y. C. Choi, and H. Busch, 1969, Cancer Res. 22, 1755. Zimmerman, E. F. and B. W. Holler, 1967, J. Mol. Biol. 2;, 149. I! l l ( (l l I( (I l. l