A KINETIC AND STRUCTURAL CHARACTERIZATION or; " - ADENOSINE 5' — TRIPHOSPHATE: RIBCIIIICLEIC ACID 1 “ ADENYLYLTRANSFERASE FROM ’ ' PSEUDOMONAS PUTIDA- Dissertation for the Degree~of Eh. D. I V MICHIGAN STATEUIIIVERSITII * ROBERT WILLIAM .BLAKESLEY_ ‘ ' 1974 ' This is to certify that the thesis entitled A KINETIC AND STRUCTURAL CHARACTERIZATION OF ADENOSINE 5'-TRIPHOSPHATE: RIBONUCLEIC ACID ADENYLYLTRANSFERASE FROM PSEUDOMONAS PUTIDA presented by Robert William Blakesley has been accepted towards fulfillment of the requirements for Ph . D B iochemis try degree in 24944844 Major professor Date September 13, 1974 0-7 639 ABSTRACT A KINETIC AND STRUCTURAL CHARACTERIZATION OF ADENOSINB S'-TRIPHOSPHATE: RIBONUCLEIC ACID ADENYLYLTRANSFERASE FROM PSEUDOMONAS PUTIDA By Robert William Blakesley This dissertation describes the research conducted on ATP:RNA adenylyltransferase isolated from two types of cells. The major portion of this work represents a kinetic and structural study of ATP:RNA adenylyltransferase from the bacterium Pseudomonas putida. This is followed by a brief investigation of the effect of 3'-deoxyadenosine 5'-triphos- phate on the AMP incorporation by ATPzRNA adenylyltransferase from Novikoff hepatoma cells. As an appendix to this dis- sertation there is presented a simple, rapid staining tech- nique for proteins separated by polyacrylamide gel electro- phoresis. ATP:RNA adenylyltransferase (EC 2.7.7.19) catalyzes the polymerization of ATP into the homopolymer, polyadenylic acid. The activity of this enzyme requires a divalent metal iOn and a polyribonucleotide as primer. The enzymatic activity has been detected in a wide variety of sourCes, Robert William Blakesley such as mammals, plants, avia, yeast and bacteria. The ubiquity of this enzymatic activity in nature suggests an importance for the activity in a cell. However, no direct evidence for the in viva function of ATP:RNA adenylyltrans- ferase has been presented. Although there were many reports of the isolation of ATPzRNA adenylyltransferase from various sources, few kinetic or structural characteristics of an enzyme of this type were known at the time this research began. Thus, a kinetic and structural study of ATP:RNA adenylyltransferase from the particulate fraction of the bacterium Pseudomonas putida was initiated. During the large-scale purification of this enzyme, designated adenylyltransferase B, a previously undetected ATP-incorporating activity, designated adenylyl- transferase A, was observed. Adenylyltransferases A and B were indistinguishable catalytically. Both of these ATP- specific, primer-dependent activities were optimally active at pH 9.5, were inhibited by inorganic pyrophosphate but not by orthophosphate, and were found to utilize effectively the same polynucleotides as primers. The divalent metal ion requirement for both activities was satisfied by Mg2+, with an optimum at 20 mM. Within experimental error, both adenylyltransferases had identical values for the Km for ATP (1-2 x 10‘4 M) and the Km for ribosomal RNA (2-4 x 10'4 M; expressed, by convention, in terms of nucleotide phosphate). Robert William Blakesley Adenylyltransferases A and B were distinguishable by their chromatographic and sedimentation properties. Adenylyl- transferases A and B were resolved by chromatography on phosphocellulose or poly(U)-Sepharose. When chromatographed on Bio-Gel P-100, adenylyltransferase A was excluded, while adenylyltransferase B penetrated the gel. By sedimentation through glycerol gradients, adenylyltransferases A and B were determined to have sedimentation coefficients (820,w) of 9.3 S and 4.3 S, respectively. The molecular weight of adenylyltransferase A was estimated to be 185,000 and that of adenylyltransferase B to be 50,000 to 60,000. Apparently, adenylyltransferase A was generated from adenylyltransferase B during the purification. Two derivatives of the antibiotic rifamycin, AF/013 and AF/DNFI, inhibited the incorporation of ATP into acid- insoluble form catalyzed by adenylyltransferases A and B. The polymerization rate by either adenylyltransferase was inhibited 50% at 5 pg per m1 AF/013 or 10 pg per ml AF/DNFI. Poly(U) and poly(dT) were competitive inhibitors of the ribosomal RNA-primed polymerization reaction. The Ki for 6 M (in terms of nucleotide poly(U) or poly(dT) was 4-10 x 10' phosphate) for adenylyltransferases A and B. The inhibition was a result of the competition between the non-priming poly(U), or poly(dT), and ribosomal RNA for the primer-binding site. The S'-triphosphate derivative (3'-dATP) of the drug cordycepin (3'-deoxyadenosine) was a competitive inhibitor with TP for both adenylyltransferases. The Ki for 3'-dATP was 6-10 x Robert William Blakesley 4 M. The activity of adenylyltransferase A or B was not 10- affected by several other analogs of ATP; namely, 2'-dATP, 2'-0-methylATP, or 3-B-D-ribofuranosylimidazo [2, l-il purine S'-triphosphate. The second research problem was undertaken to deter- mine the in vitro effect of 3'-deoxyadenosine S'-triphosphate on the AMP incorporation by a eukaryotic ATP:RNA adenylyl- transferase. The in vivo synthesis of the polyadenylate segment found covalently linked to eukaryotic messenger RNA or to its precursor, heterogeneous DNA-like RNA of the nucleus, is sensitive to the drug cordycepin (3'-deoxyadeno~ sine). If the eukaryotic adenylyltransferase is responsible for the jposttranscriptional. synthesis of the polyadenylic acid of heterogeneous DNA-like RNA as has been suggested, the adenylyltransferase should also be sensitive to the drug. The AMP incorporation by the adenylyltransferase purified from calf thymus nuclei or from rat liver nuclei, however, was not affected by the presence of cordycepin. The triphosphate derivative of cordycepin (3'-dATP) is probably the form in which cordycepin would inhibit adenylyl- transferase. To test this hypothesis ATP:RNA adenylyl- transferase was partially purified from the cytoplasm of Novikoff hepatoma cells. This enzyme was similar to the eukaryotic Mn2+-dependent adenylyltransferases previously reported. The activity was primer-dependent, ATP-specific (Km = Z x 10'5 M) and was sensitive to the sulfhydryl reagent, p-hydroxymercuribenzoate. The sedimentation Robert William Blakesley O 20,w enzyme was at pH 8. The preferred primers were polyadenylic coefficient (3 ) was 4.1 S. Optimal activity for the acid and yeast transfer RNA. Rifamycin AF/013 inhibited this adenylyltransferase; 50% of the activity remained at 20 pg per ml AF/013. When 3'-dATP was examined, it was shown to be a competitive inhibitor of the activity of adenylyltransferase. The Ki for 3'-dATP was 2 x 10.5 M. Also, in agreement with the results previously reported, cordycepin itself was ineffective in inhibiting the syn- thesis of polyadenylate by adenylyltransferase. In the appendix there is presented a new staining procedure. It was developed in order to provide a simpler, quicker staining technique for the identification of proteins separated by polyacrylamide gel electrophoresis. The staining solution was composed of Coomassie Brilliant Blue 6250 suspended in a mixture of sulfuric acid and trichloro- acetic acid. The staining technique was rapid (results as quick as 30 minutes), required little or no destaining, and had a sensitivity of at least 1 ug of protein per band. The clear, uniform background obtained was well-suited for densimetric scanning. In addition, a linear relationship was observed between absorbance and protein concentration in the concentration range of 1 ug to at least 10 ug per band. The staining technique can be used with sodium dodecyl sulfate as well as conventional standard polyacryl- amide gel electrOphoresis. A KINETIC AND STRUCTURAL CHARACTERIZATION OF ADENOSINE S'-TRIPHOSPHATE: RIBONUCLEIC ACID ADENYLYLTRANSFERASE FROM PSEUDOMONAS PUTIDA By Robert William Blakesley A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1974 DEDICATION To my entire family who provided confidence and moral support, and through whose sacrifice made possible the opportunity for this work. ii ACKNOWLEDGMENTS My sincere appreciation is expressed to Dr. John Boezi for his assistance, guidance and moral support given throughout my graduate training. My thanks are also given to the other members of my guidance committee; Dr. William C. Deal, Jr., Dr. Fritz Rottman, Dr. Loren Snyder, and Dr. John E. Wilson. I would also like to extend my thanks to the many members of the Department of Biochemistry, past and present, for their help throughout my graduate career. Special appreciation is expressed to Ron Desrosiers and Dr. Howard C. Towle for their advice and helpful discussions. Thanks are also given to Ron Desrosiers and Dr. Fritz Rottman for providing the Novikoff hepatoma cells and several valuable compounds for my research. I would like to thank the'Department of Biochemistry and the National Institutes of Health for financial support. iii CURRICULUM VITAE 0F ROBERT WILLIAM BLAKESLEY September 13, 1974 Biographical Sketch Name: Robert William Blakesley Born: January 15, 1947 in Fullerton, California Citizenship: United States of America Marital Status: Married Education (1952-1965) Elementary and Secondary Education--public schools in Fullerton, California. participant in Science Fair (1960) participant in California Math Contest (1965) (1965-1969) Undergraduate Education Fullerton Junior College, Fullerton, California A.A. degree earned in 1967 in Chemistry major University of California, Berkeley, California A.B. degree earned in 1969 in Biochemistry major (19(39-1974) Graduate Education Michigan State University, East Lansing, Michigan, Department of Biochemistry iv Thesis: A Kinetic and Structural Characterization of ATP:RNA Adenylyltransferase from Pseudomonas putida Thesis Advisor: Dr. John A. Boezi Stipend: National Institutes of Health Training Grant Publications l. Boezi, J. A.; Lee, L. F.; Blakesley, R. W.; Koenig, M.; and Towle, H. C. "Marek's Disease Herpesvirus- Induced DNA Polymerase." Journal of Virology (in press) 1974. 2. Blakesley, R. W.; and Boezi, J. A. "A Kinetic and Structural Characterization of ATP:RNA Adenylyltransferase from Pseudomonas putida," to be submitted for publication. 3. Blakesley, R. W., and Boezi, J. A. "The Effect of 3'-Deoxyademosine S'-Triphosphate on the Reaction Ca- talyzed by ATP:RNA Adenylyltransferase from Novikoff Hepatoma," to be submitted for publication. Teaching Experience (1970-1971) Department of Biochemistry, Michigan State University undergraduate laboratory supervised by Dr. J. Fairley (BCH 200) undergraduate lecture Course (reader) under Dr. J. Boezi (BCH 401) graduate laboratory supervised by Dr. J. Boezi (BCH 805) TABLE OF CONTENTS Page DEDICATION . . . . . . . . . . . . . . . ii ACKNOWLEDGMENTS . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . . ix LIST OF FIGURES . . . . . . . . . . . . . x GENERAL INTRODUCTION . . . . . . . . . . . . 1 SECTION I: THE ATP:RNA ADENYLYLTRANSFERASES OF ‘. 3 , PROKARYOTES, EUKARYOTES, AND VIRAL CORES . . . . 6 LITERATURE SURVEY . . . . . . . . . . . 7 Introduction . . . . . . . . 7 The Enzymes from Prokaryotes . . . . . . 10 The Enzymes from Eukaryotes . . . . . . . 19 The Enzymes from Viral Cores . . 24 A Summary of the ATP: RNA Adenylyltransferases . 25 Polyadenylic Acid . 28 REFERENCES . . . . . . . . . . . . . . 35 SECTION II: A KINETIC AND STRUCTURAL CHARACTERIZATION OF ATP:RNA ADENYLYLTRANSFERASE FROM PSEUDOMONAS PUTIDA . . . . . . . . . . . . . . . 40 ABSTRACT . . . . . . . . . . . . . . 41 INTRODUCTION . . . . . . . . . . . . 44 EXPERIMENTAL PROCEDURE . . . . . . . . . . 47 Materials . . . . . . . . . . . 47 Analytical Methods . . . . . . . . . . 48 Growth of Pseudomonas putzda . . . . . 49 Assay of ATP: RNA Adenylyltransferase . . . . 49 Preparation of Poly(U)- Sepharose . . . . . 50 .RESULTS . . . . . . . . . . . . . . . 51 Purification of ATP: RNA Adenylyltransferases A and B . . . . . . 51 vi Preparation of Initial Extract . . . Sedimentation . . . . . . Solubilization . . . . . . pH Fractionation . . . . . . pH Concentration . . . Bio- Gel P- 200 Gel Filtration . . DEAE-Sephadex Chromatography . . Phosphocellulose Chromatography . . Summary and Comments on Purification Through Phosphocellulose Chromatography Bio Gel P 100 Gel Filtration . . . . Properties of the Reaction Catalyzed by Adenylyltransferases A and B . . Characteristics of the Reaction Catalyzed by Adenylyltransferase A . The Effect of ATP Structural Analogs on Adenylyltransferases A and B . The Effect of Rifamycin Derivatives on Adenylyltransferases A and B . . . . The Effect of Synthetic Polynucleotides on AMP Incorporation by Adenylyltrans— ferases A and B with Ribosomal RNA as the Primer . . . . . Structural Properties of Adenylyltransferases A and B . . . Determination of the Sedimentation Coeffi- cient of Adenylyltransferases A and B . Determination of the Stoke's Radius for Adenylyltransferase B Using Gel Filtration on 810- Gel P 100 Determination of the Polypeptide Composition of Adenylyltransferase B Using SDS Polyacrylamide Gel Electrophoresis DISCUSSION . REFERENCES SECTION III: THE EFFECT OF 3'-DEOXYADENOSINE 5'- TRIPHOSPHATE ON THE REACTION CATALYZED BY ATP:RNA ADENYLYLTRANSFERASE FROM NOVIKOFF HEPATOMA ABSTRACT . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . EXPERIMENTAL PROCEDURE . . . . . . . Materials Analytical Methods . . Assay of ATP: RNA Adenylyltransferase Partial Purification of ATP: RNA Adenylyltransferase Vii 69 76 76 76 79 83 95 98 99 100 102 102 103 103 104 Page RESULTS AND DISCUSSION . . . . . . . . . . 107 Properties . . . . . . . . . . . . 107 Inhibitors . . . . . . . . . . . . lll Substrate Analogs . . . . . . . . . . 113 REFERENCES . . . . . . . . . . . . . . 118 APPENDIX: A RAPID, SIMPLE STAINING TECHNIQUE FOR PROTEINS SEPARATED IN POLYACRYLAMIDE GEL ELECTROPHORESIS . . . . . . . . . . . . 121 Introduction . . . . . . . . . . . . 121 Stain Preparation . . . . . . . . . . 123 Comments . . . . . . . . . . . . . 123 REFERENCES . . . . . . . . . . . . . . 127 viii Table II. III. IV. II. III. IV. LIST OF TABLES Section I A General Characterization of ATP:RNA Adenylyltransferase . . . Section II Purification of ATP: RNA Adenylyltransferases A and B . . . . Characteristics of the Reaction Catalyzed by Adenylyltransferase A . . . Primer Specificity for Adenylyltransferase A The Effect of Rifamycin Derivatives on Adenylyltransferase B Section III Characteristics of the Reaction Catalyzed by Adenylyltransferase . . . . Nucleotide Specificity for Adenylyltransferase Primer Specificity for Adenylyltransferase The Effect of Inhibitor Compounds on Adenylyltransferase ix Page 26 63 64 70 108 109 110 112 LIST OF FIGURES Figure Section II l. Bio-Gel P-200 gel filtration of fraction pH 3.2 precipitate . . . . . . . . 2. Phosphocellulose chromatography of fraction DEAEeSephadex . . . . 3. Bio—Gel P-100 gel filtration of fractions phosphocellulose A and phosphocellulose B 4. Inhibition of the ATP-incorporating activity of adenylyltransferase B in the presence of 3'-dATP 5. Inhibition of the ATP-incorporating activity of adenylyltransferase B in the presence of poly(U) or poly(dT) 6. Affinity chromatography of fractions concentrated phosphocellulose A and concentrated phos- phocellulose B on poly(U)-Sepharose 7. Glycerol gradient centrifugation of adenylyl- transferases A and B . . . . . . . 8. SDS-polyacrylamide gel electrophoresis of fraction phosphocellulose B . . . Section III l. The effect of ATP structural analogs on ATP incorporation by adenylyltransferase . Page 55 55 60 67 72 74 77 80 114 GENERAL INTRODUCTION The past two decades could be called the era of molecular biology. It was ushered in by the famous Watson and Crick model for DNA. The presentation of this model demonstrated that the chemical structure of a molecule could explain biological phenomena. In this case, the base se- quence of the DNA represented genetic information. Since then there has been a great deal of research interest in how the genetic information of DNA is replicated, and how the transmission of this information to various parts of the cell is controlled. The study of the enzymes that polymerize nucleic acids is an integral part to the under- standing of these processes. This dissertation covers research conducted on a nucleic acid polymerizing enzyme from two types of cells. The major portion of this work represents a structural and kenetic study of ATP:RNA adenylyltransferase from the bacterium Pseudomonas putida. This is followed by a brief investigation of the effect of 3'-deoxyadenosine S'-triphos- phate on the activity of ATP:RNA adenylyltransferase from the eukaryotic cell, Novikoff hepatoma. In addition, a literature review of the adenylyltransferases isolated from a wide variety of cells in nature, with emphasis on those from prokaryotes, is also presented. The in vitra synthesis of homopolynucleotides can be catalyzed by several types of enzymes, including DNA-dependent RNA polymerase, terminal nucleotidyl transferase and poly- nucleotide phosphorylase. If these enzymes are presented nucleotides of more than one base-type, then they will catalyze the formation of heteropolynucleotides. On the other hand, there are some nucleic acid polymerizing enzymes which only catalyze the synthesis of homopolyribonucleotides in vitra. These homopolyribonucleotide polymerases are found throughout nature and examples preferentially utilizing CTP, UTP, GTP, or ATP as substrate have been described. Their ubiquity in nature would seem to indicate their necessity to the biology of the cell. However, none of the investigations have provided direct evidence for the in viva role of these enzymes. From a separate line of investigation, messenger RNA, the transcription product of DNA-dependent RNA poly- merase, isolated from a wide variety of eukaryotic sources was recently found to contain a terminal homopolyribo- nucleotide sequence, polyadenylic acid. This homopolymer was not coded for by the template, but, in fact, synthesized on the end of the messenger RNA molecule after transcription had taken place. The presence of the polyadenylic acid sequence is presently thought to be necessary for the proper processing of messenger RNA from its precursor and for the functionality of messenger RNA in translation. It has been suggested that a homopolyribonucleotide polymerase is responsible for ths synthesis of the terminal polyadenylate sequence. Until the presentation of this hypothesis, the homopolyribonucleotide polymerases were little-known curiosities. Now some of these enzymes are being considered for an integral part in the genetic transfer of information within a cell. Of the four types of homopolyribonucleotide poly~ merases, that which synthesizes polyadenylic acid from ATP has been studied in the greatest detail. The many reports of this activity in nature are discussed, with emphasis on those from prokaryotes, in the literature survey which comprises section one. The accepted systematic name for this enzyme is ATP:RNA adenylyltransferase (EC 2.7.7.19); however, more often the trivial name poly A (Adenylate) polymerase is used. When the research for this dissertation began, the enzymatic activity of adenylyltransferase had been known for about ten years. The enzyme had been partially purified, the basic requirements of the reaction had been determined, but there was neither information on the structure of the protein nor detailed kinetics of the enzymatic reaction. The objectives of this research then were to elucidate many of the structural and kinetic II‘ [I ‘(ll‘ll‘l‘lll Ill] 1.“ characteristics of this enzyme, which, in turn might provide for an understanding of its in viva role. A short time before this work began, the enzyme from the bacterium Pseudamanas putida had been partially purified on a small scale in this laboratory. Scaling up and im- proving this already established purification procedure was necessary in order to carry out the proposed characterization of adenylyltransferase. During the course of this investi- gation a previously undetected adenylyltransferase activity was observed. It was identical catalytically, but was different in size and chromatographic behavior from the enzyme previously studied. The second section of this dis- sertation describes the kinetic and structural characteri- zation of both of these adenylyltransferases. This section is to be submitted for publication in a scientific journal. The third section of this dissertation presents a brief investigation of a eukaryotic ATP:RNA adenylyltrans- ferase. As previously described, eukaryotic messenger RNA contains polyadenylic acid, proposed.to be posttranscrip- tionally synthesized by an adenylyltransferase. The in vitra polyadenylation of messenger RNA can be inhibited by the drug cordycepin. If adenylyltransferase is responsible for the polyadenylation of messenger RNA, this drug should inhibit the activity of the enzyme in vitra. Adenylyl- transferase was purified from Novikoff hepatoma, then briefly characterized. To test the hypothesis, the effects of the drug and its triphosphate derivative on the reaction catalyzed by adenylyltransferase were examined. The results of this study are presented in the format of a brief scientific journal article, which also will be submitted for publication. SECTION I THE ATP:RNA ADENYLYLTRANSFERASES OF PROKARYOTES, EUKARYOTES, AND VIRAL CORES LITERATURE SURVEY Introduction The in vitra formation of homopolymers of nucleic acid can be catalyzed by a variety of enzymes. DNA-dependent RNA polymerase can, by the reiterative copying of short sequences of thymidylic acid residues in the DNA template, produce a homopolymer (1). Without templates, terminal deoxynucleotidyl transferase can polymerize deoxynucleoside triphosphates into homopolydeoxynucleotides (2); and poly— nucleotide phosphorylase, which proved instrumental inIthe early deciphering of the genetic code, can catalyze the formation of homopolyribonucleotides from ribbnucleOSide diphosphates (3). However, when presented nucleotides of more than one base-type, these enzymes will produce hetero- polymers. Unlike these enzymes, there exists one group of polymerases which synthesize only homopolymers, regardless of the type of nucleotide bases available for reaction. These unusual enzymes can be distinguished by their pro— duction of polymers, up to 200 residues in length, consisting of only one of the four common ribonucleoside 5'-triphos- phates. The enzymatic activity requires a divalent metal ion and a primer. The primer is a polyribonucleotide with a free 3'-hydroxyl end to act as an initiation point for the polymerization. Homopolyribonucleotide polymerizing 7 I l‘l‘Il I'l‘ll‘ll ‘lll‘ll .(l‘l‘.1‘l|l‘| (I. ([l {l‘ enzymes which are specific for each of the ribonucleoside triphosphates, UTP (4,5), CTP (6,7), GTP (8,9), and ATP (9-62) have been isolated. Examples from a wide variety of sources, such as, bacteria (10-24), mammals (5-7, 25-45), plants (9, 46-50), avia (4, 51, 52), and viruses (57-62) have been reported. Of these enzymes those which polymerize ATP into polyadenylic acid have received the greatest attention. Nearly fifteen years ago Edmonds and Abrams (25) reported the characteristics of an enzyme partially purified from the particulate fraction of calf thymus nuclei. This activity polymerized ATP into long chains of AMP residues. The activity required Mg2+, but did not require exogenous polynucleotide as primer. The specificity for ATP as substrate, the inhibition of AMP incorporation into acid- insoluble form by the presence of the other three common ribonucleoside 5'-triphosphates or inorganic pyrophosphate, and the product length of 25 to 100 AMP residues distin- guished this activity from the activities of several other enzymes that could synthesize homopolymers. Later, by eliminating contaminating RNA from their preparation, these investigators found that this enzymatic activity required a polynucleotide to prime the reaction. The primer require- ment was satisfied by polyadenylic acid, but not by yeast RNA or by calf thymus DNA (26). A second group of investigators, August et a1. (10), also made a major early contribution to the characterization of a polyadenylate synthesizing activity. In 1962 (10), they reported the purification of the enzyme from the particulate fraction of Escherichia caZi. Through a detailed analysis, the enzyme was found to utilize only ATP as sub— strate (Km = 6 x 10.5 M) and to be Mg2+-dependent. The enzyme, optimally active at pH 10.5, could polymerize up to 200 AMP residues in chain length. Specific AMP incorpo- ration by the enzyme was inhibited 20 to 30% in the presence of the other three ribonucleoside 5'-triphosphates. In contrast to the calf thymus enzyme, the E. caZi enzyme could be stimulated by many different natural RNAs, for example, ribosomal RNA, transfer RNA, and the in vitra transcription product of E. caZi DNA-dependent RNA polymerase. However, the synthetic homopolyribonucleotides, polyuridylic acid, polycytidylic acid or polyadenylic acid would not function as primers. The general reaction catalyzed by an ATP-specific polyadenylate synthesizing enzyme was: Mn2+ or Mg2+ n ATP + (nucleotide)m: A (nucleotide)m+ n + n pyrophosphate. The accepted systematic name, ATP:RNA adenylyltransferase (EC 2.7.7.19), is descriptive of this reaction. More 10 frequently, the trivial name, poly A (Adenylate) polymerase, is given to an enzyme that catalyzes this type of reaction and displays properties similar to those already mentioned. The reports from calf thymus and E aaZi stimulated many workers to search for poly A polymerase activity in other organisms. The ubiquity of the activity was thereby well established; however, little new information was obtained. A majority of the subsequent reports were limited characterizations of partially purified enzymes. They confirmed the distinctive characteristics of the enzyme, but failed to provide a functional role for the activity in the biology of the cell. The Enzymes from Prokaryotes The investigators of the prokaryotic poly A poly- merase made several attempts to determine its physiological role. Those that utilized the purification procedure of August et a1. (10) found that the enzymatic activity was tightly associated with the particulate fraction of E. caZi. This suggested that the enzyme might be an integral part of the ribosomes and, thereby, be providing a role in the translational process. Smith and August found that the activity was bound to the ribosomes of E. caZi upon puri- fication and speculated that the enzyme might be involved in protein synthesis. However, the activity could be released by suspension of the ribosomes in 20 mM glycylglycine, 11 pH 9.5 (12). Because this treatment was known to release electrostatically bound proteins from ribosomes (82), they stressed that the observed binding need not restrict the enzyme to a similar location in viva. Hardy and Kurland elaborated on this last point by noting the failure of the enzyme-ribosome relationship to meet four criteria. The enzyme did not affect protein synthesis; it was not specific for either the 30 S or 50 S ribosomal subunit; the number of binding sites on the ribosome was not well-defined; and, as before, the enzyme was not associated with the ribosome when subjected to conditions which removed absorbed contami- nants (13,14). Thus, the observed association of the enzyme was probably an artifact generated by an electrostatic attraction to the ribosomes upon cell rupture. Another method used to determine the physiological significance of poly A polymerase was analysis of the enzyme's activity following bacteriOphage infection. Investigators of prokaryotic systems were accustomed to this technique for discovering many aspects in the control of DNA and RNA metabolism in a cell. Thus, if poly A polymerase had a function essential to this metabolism, its activity might be altered upon infection. In fact, upon infection by a T-even phage, the E. aaZi particulate poly A polymerase was irreversibly inhibited. The inhi- bition was specific only for this enzyme, as other nucleic acid enzymes seemed to be unaffected. Appearance of the 12 inhibition was chloramphenicol-sensitive, indicating that a protein might be responsible (15). Later, the inhibitor was localized in the non-particulate fraction of T4-infected E. 001i and Shigella dysenteriae. The inhibitor from either source could inhibit the particulate poly A polymerase from E. caZi or from S. dysenteriae (16). The concomitant inhi- bition of the particulate enzyme and the host RNA metabolism suggested the possibility that this enzyme had some important physiological role in the RNA metabolism of the host cell (15). However, the lack of further evidence left this hypothesis undeveloped. The last approach was a familiar method used by microbiologists for deducing the function of an enzyme. The technique was to isolate a temperature-sensitive mutant of the enzyme, then determine what function was affected at the non-permissive temperature. Terzi at al. (18) attempted to demonstrate that an observed thermo-sensitivity in the RNA metabolism of an E. aaZi mutant was the result of a temperature-sensitive poly A polymerase activity. In crude extracts, DNA, RNA and protein synthesis, in addition to poly A polymerase activity were arrested at the non- permissive temperature. Upon purification of the poly A polymerase, the supposed temperature sensitivity of the activity was lost. Thus, no conclusive results were obtained from their experiments. As part of the investigation, though, these workers compared the poly A polymerase and E. aaZi 13 DNA-dependent RNA polymerase by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis. They concluded that the two enzymes "have some common subunits," and that "poly A polymerase seems to be just a special form of RNA polymerase, special in having the same polymerase core but with a different head." This statement, based upon a partially purified enzyme, raised the question of the uniqueness of poly A polymerase. To determine whether or not the structures of poly A polymerase and RNA polymerase were related, some labo- ratories carried their characterization of poly A polymerase further. Based upon a previous report that RNA polymerase could synthesize a polyadenylic sequence (1) and their own SDS polyacrylamide electrophoresis data, Ohasa at al. con- cluded that the E. aaZi poly A polymerase and the a-subunit of the RNA polymerase complex (aZBB'o) were the same poly- peptide (19,20). However, this argument was severely weakened by their poor purification of the enzyme, as evidenced by the lack of conformity of the characteristics of their activity to these known for poly A polymerase. For instance, the activity was inhibited by an+ or Mg2+, and by inorganic orthophosphate. In addition, the activity lacked substrate specificity as CTP, UTP and GTP were incorporated nearly as well as ATP. The reports in support of equating poly A polymerase and a subunit of RNA polymerase (18-20) suffered from a 14 failure to purify the enzyme. The first report was based on finding RNA polymerase subunits in poly A polymerase prepa- rations (18), while the second report rested on locating poly A synthesizing activity in RNA polymerase preparations (19,20). In order to best answer the question, it was necessary to purify poly A polymerase to homogeneity. In 1973, Sippel reported the extensive purification of a poly A polymerase from E. aaZi. SDS polyacrylamide gel electro- phoresis revealed the enzyme to be a single polypeptide of approximately 50,000 molecular weight, well-differentiated from any of the subunits of RNA polymerase. By Sephadex gel chromatography, the native molecular weight was esti- mated to be 58,000. Two inhibitors of RNA polymerase, rifampicin and streptolydigin, were ineffective against the ATP-incorporating activity of this enzyme (22). This evidence together with the distinctive kinetic character- istics already established contradicted the claim that this poly A polymerase and the o-subunit of RNA polymerase were the same. However, as will be demonstrated, there are possibly several distinct poly A polymerases in prokaryotes. Each will have to be examined for its possible relationship to RNA polymerase in order to definitively prove or disprove the hypothesis. Poly A polymerase has been isolated from the parti- culate fraction of sources other than E. caZi. The enzyme was partially purified from S. dysenteriae by the procedure 15 of August et al. (10). From a brief characterization, this enzyme was found to be slightly different than that from E. caZi. Optimal activity occurred at pH 9 and at Mg2+ concentrations of less than 1 mM (16). On the other hand, pH 10.5 and 20 to 30 mM Mg2+ were optimal for the E. aaZi enzyme (10). A poly A polymerase activity has also been detected in the particulate fraction of Pseudamanas putida. Optimal activity occurred at pH 9.5 and at concentrations 2+ of 20 to 25 mM Mg The Km for ATP as substate was deter- mined to be 3 x 10'4 M. The activity was entirely dependent upon exogenous RNA. This primer requirement could be satisfied by ribosomal RNA, transfer RNA or Ap(Ap)2A, but not by DNA, polyadenylic acid or (Ap)4. The product was in excess of 100 AMP residues in length covalently attached to the 3'-end of the primer molecule (24). A comparison of these reports with those of E. caZi demonstrate the close correspondence between prokaryotes of the characteristics of the particulate poly A polymerase. However, there are several other reports of polyadenylate synthesizing activities in E. aaZi whose characteristics are sufficiently different to suggest that there are several distinct enzymes. An early report (11) of a partially purified enzyme from E. aaZi had two features that were- taken to be significant. First, the in vitra reaction 2+ .. . + . . . . mixture contained an in addition to Mg No eV1dence was presented to determine whether either or both divalent 16 metal ions were required for activity. Second, the activity was detected in the non-particulate (100,000 x g supernatant) fraction of E. caZi. The localization upon purification, though, was not a sufficient criterion for establishing the activity as a unique enzyme. In addition, the investigators did not examine their particulate fraction for polyadenylate synthesizing activity. Thus, it is not possible to determine that this activity was different from that detected in the particulate fraction. The effect of bacteriophage infection upon poly- adenylate synthesizing activity stimulated Gross and Alberty (17) to investigate the effect on the ”non-particulate enzyme." They concluded that T4 infection resulted in a 2.5-fold increase in the non-particulate activity and in a physical modification of that enzyme. The particulate poly A polymerase was unchanged by T4 infection. While observing the changes in poly A polymerase activity upon infection, the activity of another enzyme, ATP(CTP)-transfer RNA nucleotidyltransferase, was to serve as a control. The assay conditions for these two activities were similar so the workers used only those of the latter enzyme in their analysis of the modification on poly A polymerase. The failure to assay specifically for poly A polymerase activity generates considerable doubt in the validity of their con- clusion. This work also suffered from the false assumption that the determination of the cellular fraction of this 17 activity upon purification was sufficient evidence for a different polyadenylate synthesizing enzyme. Modak and Srinivasan (23) have reported the purifi- cation of a primer-independent ATP-incorporating enzyme from the non-particulate fraction of E. caZi. The activity was either Mg2+-v or Mn2+-dependent and optimally active at pH 7.9 to 8.9. The Km for ATP as substrate was determined to be 1.4 x 10-3 M. The incorporation of AMP could be inhibited by streptolydigin, but not by rifampicin or a-amanitin. A sedimentation coefficient (320,w) of 8-9 8 was determined for the enzyme. These characteristics are different from those already described for the particulate poly A polymerase, but the fact these workers emphasized most was that the enzyme did not require exogenous RNA for AMP incorporation. However, the activity was stimulated by the addition of ribosomal RNA and transfer RNA, an indication that their preparation might not be free of nucleic acid. Their best evidence that the enzyme did not require an exogenous primer was the sucrose gradient centrifugation of a reaction mixture 3 incubated with exogenous primer. The H-labeled AMP incorpo- 14C-labeled exogenous ration product was separated from primer by the centrifugation, indicating no covalent linkage between the two and, therefore, the activity's primer— independence. However, if contaminating nucleic acid were present in their enzyme preparation, it could preferentially prime the reaction and be undetected by this procedure. Thus, 18 their claim that exogenous primer was not required for this enzyme's activity is questionable. But the other character- istics which were established do distinguish this ATP- incorporating enzyme from the particulate ATP:RNA adenylyl- transferase. Recently, Sippel (22) has characterized a highly purified "adenyltransferase" from the particulate fraction of E. caZi. This enzyme is similar in respect to ATP- specificity and primer-dependency to the particulate poly A polymerases already described. The Km for ATP as substrate was 5 x 10'5 M. QB RNA and E. caZi transfer RNA satisfied the requirement for polynucleotide as primer. The activity of this enzyme is distinguished from that of other prokaryotic 2+ over Mg2+, and poly A polymerases by its preference for Mn by a pH optimum of 8. This adenylyltransferase was also inhibited by inorganic orthophosphate as well as by pyro- phosphate, but not by rifampicin or streptolydigin. In addition, enzymatic activity was stimulated approximately 2.5-fold in the presence of 0.35 M KCl. From SDS poly- acrylamide gel electrophoresis and from Sephadex G-100 gel filtration, the enzyme appeared to be a single polypeptide of approximately 58,000 molecular weight. Finally, one report (21) of an ATP—incorporating activity from E. aaZi, designated RNA-primed polynucleotide- pyrophosphorylase, clearly pointed out the distinction of this activity from poly A polymerase. The activity 19 incorporated CTP and UTP as well as ATP, had pH optima at 7 and 10, was RNA-primed, and was inhibited by actinomycin. Structurally, the enzyme was about 40,000 molecular weight, composed of two 20,000 molecular weight subunits. The enzyme had a sedimentation coefficient (320,w) of 3.1 8. Even though polyadenylate can be synthesized, this enzyme is distinct from ATP:RNA adenylyltransferase. In summary, there are several ATP-incorporating acti- vities in prokaryotes. How many of these reported activities represent truly distinct enzymes is difficult to determine from the limited amount of data available. In general, ATP: RNA adenylyltransferase utilizes almost exclusively ATP as substrate. The divalent metal ion requirement with one exception (22) is preferentially satisfied by Mg2+. Several synthetic and natural RNAs with a free 3'-hydroxyl end can satisy the requirement for exogenous polynucleotide (primer). The optimum pH is alkaline, varying from pH 8.0 to 10.5. The product is composed almost entirely of adenylic acid, generally greater than 100 residues in length. The activity is not inhibited by rifampicin, actinomycin or streptolydigin. And, in one case (22), the enzyme is probably a single polypeptide of approximately 60,000 molecu- lar weight. The Enzymes from Eukaryotes Only a brief review of the more important facts about eukaryotic enzymes will be presented. Polyadenylate 20 synthesizing activities have been detected in a wide variety of eukaryotic sources, including mammals (25-45), plants (9, 46-50), avia (51-52), yeast (53,54), and echinoderms (55,56). Poly A polymerase can be isolated from many parts of the cell, for example, the nucleus (25-28, 46, 47, 53, 55, 56), ribosomes (54), mitochondria (43-46), chloroplasts (9,46), or the soluble fraction (34-42, 48, 49, 51). Unlike prokaryotes, the characteristics of these activities permit a subclassification into three groups; the Mg2+-dependent activities, predominantly located in the nucleus, the Mn2+- dependent activities, isolated outside the nucleus, and the activities from organelles, such as mitochondria and chloro- plasts. Generally, they are all ATP-specific; incorporation yields products significantly less than 100 residues in length; the activity is dependent on a RNA primer and either Mn2+ or Mg2+; and optimal activity occurs near pH 8. Unfortunately, most of the work on the eukaryotic enzymes have been confined to brief reports of these basic characteristics of reaction from crude preparations. The trend has been away from purification to homogeneity or extensive characterization of these enzymes, but rather toward identification of new sources. In addition, little research into the function of this activity has been made compared to the attempts with the prokaryotic enzyme. The Mg2+-specific poly A polymerase isolated from calf thymus nuclei in 1960 (25) was further purified and 21 characterized by Winters and Edmonds (27). They found that the primer requirement could now be satisfied by RNAs other than polyadenylic acid (28). Transfer RNA proved to be especially effective as a primer. The native molecular weight determined from gel chromatography was estimated to be 140,000 to 160,000. Incorporation of ATP into acid- insoluble form was not affected by e-amanitin, actinomycin D or cordycepin (27). These drugs also failed to inhibit the enzyme isolated from rat liver nuclei (31, 32). How- ever, one derivative of rifamycin, AF/013, which was shown to be singularly effective against a variety of nucleic acid polymerizing enzymes (81), inhibited the rat liver Mg2+- specific poly A polymerase (32). The Mn2+-specific poly A polymerase has been best purified and characterized by a group of workers headed by Bollum (34). This enzyme from the cytoplasm of calf thymus electrophoresed as a single polypeptide of 60,000 molecular weight in polyacrylamide gels containing SDS. Upon electro- phoresis in the absence of SDS, the enzyme appeared as a single, coincident band of protein and enzymatic activity. The enzyme was determined to have a sedimentation coefficient (S20,w) of 3.5 S from sucrose gradient centrifugation analy- sis. Bollum's group also has isolated this enzyme from the cytoplasm of human lymphocytes (35). The activity was similar to that of the calf thymus enzyme, including the preference for oligoadenylic acid as primer. The molecular 22 weight of the enzyme was 60,000. Interestingly, upon stimu- lation of these cells by phytohemaglutinin, the poly A polymerase activity increased three- to six-fold. This was the first demonstration of the enzyme from a "mammalian, non-viral system in which increases in . . . activity were observed with changes in the physiological state of the cell" (35). The Mn2+-specific enzyme is not restricted to the cytOplasm, as this type of activity has also been detected in the nuclei of rat brain (29), yeast (53), sea urchin embryo (55, 56) and guinea pig, and ex cerebral gray and white matter (30). The characterizations.of these activities, although limited, have shown that they are catalytically similar to the enzyme from the cytoplasm of calf thymus. The enzyme from the mitochondria of rat liver was not purified sufficiently to allow many significant charac- + . 2 , but no primer terizations. The activity used Mg2+ and Mn requirement was established. In contrast to the other poly A polymerases, the activity was not inhibited by the presence of the other three ribonucleoside 5'-triphosphates (43). The activity from wheat chloroplasts has been better charac- terized. The divalent metal ion requirement was satisfied equally well by Mn2+ or Mg2+. Many synthetic and natural RNAs primed the reaction. AMP incorporation was inhibited by the other three ribonucleoside triphosphates, but not by actinomycin D or a—amanitin (9). No characterization of the structure of either of these two enzymes has been made. 23 In addition to normal cells, there have been several reports of polyadenylate synthesizing activities in trans- formed cells. The activity from HeLa cytoplasm was Mn2+- dependent and used polyadenylic acid or transfer RNA as primer. The drugs rifamycin AF/AP and AF/ABDMP could inhibit AMP incorporation when added at very high concentrations. It was not determined, however, whether the nucleotide incorporating activity was specific for only ATP (38). An ATP-specific polymerizing activity was also detected in the mitochondria of rat liver tumors. This activity used Mn2+ and Mg2+ and was primer-dependent. Polyadenylic acid satisfied the requirement for primer. AMP incorporation into acid-insoluble form was not inhibited when the other three ribonucleoside 5'-triphosphates were present. Compared to the mitochondrial activity of normal tissue, the specific activity of the mitochondrial enzyme from rat liver tumors was nearly the same (44, 45). Probably one of the best purified and characterized poly A polymerases of transformed cells was reported from the cytoplasm of mouse ascites cells. This Mn2+-dependent enzyme had an activity optimum at pH 8.1 to 8.5. The primer requirement was fulfilled by transfer RNA, ribosomal RNA or polyadenylic acid. The enzyme had a sedimentation coefficient (320 w) of 4.8 S and an isoelectric ’ point of pH 6 (36, 37). 24 The Enzymes from Viral Cores The investigation into viruses as a source of poly A polymerase activity is a relatively new and, therefore, limited field. Polyadenylic acid synthesizing activity has been detected in cells infected with vesicular stomatitis virus (57,58). Samples from a crude cytoplasmic extract of virus-infected cells were assayed. However, proper controls were not included, so the observed synthesis of polyadenylic acid cannot be taken as proof of a viral poly A polymerase. On the other hand, there are several reports of poly A polymerase activity in purified virions. For example, in reovirus there is an ATP-specific incorporating activity. The activity utilized Mn2+ and had a pH optimum of 7.5 to 8.0. Due to a partial purification, the activity was not primer-dependent (59). The best evidence for a viral poly A polymerase has been the study of the activity from purified cores of vaccinia virus (60-62). The incorporation was ATP-specific, Mn2+-dependent, and inhibited in the presence of the other three ribonucleoside 5'-triphosphates. Enzymatic activity was stimulated upon addition of several synthetic homopolymers, for example, oligoadenylic and polyadenylic acid, however, primer-dependency was not established (60,62). SDS poly- acrylamide gel electrophoresis of the partially purified vaccinia enzyme revealed polypeptides of 35,000 and 50,000 molecular weights (62). 25 A Summary of the ATP:RNA Adenylyltransferases Poly A polymerase activity has been detected in a large number of biological sources. For a majority of the cases the enzyme was partially purified and its activity was minimally characterized. Because of these qualities and the repetitiveness of the data, it was not practical for many of the literature reports to be represented heretofore. A table to summarize all of the reported poly A polymerase activities is now presented (Table I). It is readily apparent; from the table that poly A polymerase activity is ubiquitous in nature. Enzymatic activity has been reported in eukaryotes: nuclei, cytoplasm, mitochondria, chloroplasts, and ribosomes; in prokaryotes: particulate, and non-particulate; and eukaryotic viruses. It is also evident that the enzyme is ATP-specific; utilizes Mn2+ or Mg2+ to satisfy the divalent metal ion requirement; has optimal activity at alkaline pH; and is not inhibited by a-amanitin, actinomycin, strip- tolydigin or rifampicin. Many of the activities are found not to be primer-dependent, but primer-stimulated. This is a feature of characterizing impure enzyme preparations. Although the variation between reports for some characteristics can be explained by enzyme impurity or different experimental technique, there are several trends that may be of significance. For instance, with one exception (22), the prokaryotic enzymes utilize only Mg2+, while the eukaryotes have numerous examples using Mn2+, Mg2+ or both. _ - 1.0. if}. mWWLliP. c .I‘At IF . .6 .. ”in... link. R. ALI, ._...l... ...2..,..1. .. .- 26 II Ii -. I: - I , It. 1.2.1.1. -Alwia no... I . - I I - .I.I...u.w....~mu.I.- .L.» 3 21: am: .LE 3“ .. + . 28> om-mv om +mcz e.m + - + .coom owazz mm .#m +Nuz m.m-m - o>ELEo xu_;u p A+VH mac ”A-Vu pom .aem oflm H.wy +ch m.n + + + «sewage: CcOx_>oz mm .om ufim m.v~ oe-om +ch m.w + + mou_oma omzoz mm 3“ .25 x2 as: + Sex Na .3. 3.8. 4:: + + + c3: umz ow +N£a + + + ORCLEo omsoz mm m-v” LAM .uom .me oco.co +ch - mouxuoczex_ snazz cm ooo.oc +ch m.m + + m:Ex;u *_mmJI Em:H&0u>u Mm A-v” Hum .mem ooo.oo~ +ch + + + umao> S 3” a: ”I” use .ch .teaz 9m + . Bee. sees; a :1 8e .aa 2 .ch + . Lees :22 ca a-vn pom +Nmz - + mmofi ooomnoh om .mm A-Vu Hum v-m +ch + + + oxcnso :_;oc: aom mm +Nwz - + mouwumm Nozzumwcmq mm 3” 8... 3A As: 3 . - Ens 51 cm A-Vn uom +ch o.w + - + :_:LL x0 on A-Vu now + :2 o.m + - + swaps wan mocflso N... .3 TV” So ”:1 a: :8 .3... as AS. :2 .. - + 82.: 51 RAN 3” 8e :8 . 2a 2562 87mm -wza N.” + + .. waif tea a +e _ _o_u:z I muec»x~uwmcom uzwfioz space; sea _1_ a maez m m vocfisvoz w ufluquoam mouczom 030335 336202 Unwed 13¢: :35 EC .850 .3 SEC; E.< :flououm UQHL;_;:_ msocouozm .omceoum:msu~»~»:op< oenn am am -2 tie: e as. .e + - . a: _. seam mm .3 02-2. Au: - Access: 2:28.) maxi; mm .2 .2 :1 be J: .25 3?. .13 92-2., A»: .Acz if: . - + II- Us..s....wI.W| - . Mo~c-3o_ Lc.x_-rzsz v.3. T: m3 1:“ be. .CC :8 25.3 2: La: 31 . . + sass; .a CH $.42 0.9. e..,...a.~.m.§som..we new NN . :u be a: 89.3 3% as: saw + + + 3% a 2-2 2 :1 Be a? 3 STE. Au: 3:. + + + IIIIIC . eun_:udueme mfibbpgoma em 81: Am: 3st m .m + . . Emma...” moEomosz a :1 8a .28 2 Am: .Aez . + as: :25 ow A-vn uum +mw2 - + chm Ouuagowli mummwmmwcflzu me .wq +nw2 .+m:2 - + + assuage: mqecfla me A-Vn mac +Nuz .+m:2 - - + eo>dfl Dex as 3 u 3... tea: - + .1 e2 metamfi-.. mwcvcozu0uwm 28 The enzymes of prokaryotes generally are Optimally active at pHs greater than pH 8, the optimum for the enzymes of eukaryotes. And, comparing chain lengths of polyadenylate synthesized, the poly A polymerases from prokaryotes generally polymerize greater than 100 AMP residues, while those from eukaryotes polymerize significantly less than 100 AMP residues. The enzyme from viruses seems to be similar to the eukaryotic poly A polymerases in pH optimum and in the divalent metal ion utilized. Finally, the poly A polymerase from mitochondria could be unique in the failure of UTP, CTP and GTP together to inhibit AMP incorporation. Polyadenylic Acid Poly A polymerase has been found in many different types of cells throughout nature. Despite the importance of this enzyme as suggested by its ubiquity in nature, investi- gators have failed to deduce the role the enzyme plays in the physiology of a cell. Recently, a separate line of investigation has led to a hypothesis for the function of the eukaryotic poly A polymerase. This investigation has involved the identification of polyadenylic acid in viva and the processing of functional messenger RNA from a transcriptional precursor. The purpose of this section of the survey is only to briefly cover the facts from this investigation that are relevant to an understanding of the hypothesis. For a more thorough review of the proposed 29 biogenesis of eukaryotic messenger RNA the reader is directed to articles by Darnell et a1. (63), and Brawerman (80). Polyadenylic acid was discovered in viva many years ago as ribonuclease-resistant segments in RNA digests. An adenine-rich polymer was found associated with the messenger RNA of rabbit reticulocytes (64); the cytoplasm of rat liver microsomes contained polyadenylic acid sequences (65); and Edmonds and Abrams discovered that in viva polyadenylic acid was responsible for priming the calf thymus nuclear poly A polymerase (26). But no significance was attached to these discoveries. The rediscovery several years later of naturally occurring polyadenylic acid in the messenger RNA isolated from vaccinia virus-infected cells (66), sparked a wide spread interest in the isolation of polyadenylate-rich RNA from a variety of sources. Of the RNA species of eukaryotic cells, it was found that only messenger RNA (mRNA) and the large, non-descrete RNA of the nucleus, called heterogeneous nuclear RNA (HnRNA), contained long polyadenylic acid sequences. These sequence were nearly 200 residues in length and located at the 3'-hydroxyl end of the molecule. These discoveries were made with RNA isolated by new techniques that made use of the unusual properties of polyadenylic acid. Polyadenylate containing polynucleotides can be separated from other nucleic acids by extraction in aqueous phenol at pH 9 (67), or by the selective binding to "Millipore" filters at high salt (68), or by binding 30 to polyuridylate- (69) or polydeoxythymidylate-cellulose (70). The significance of this work was two-fold. First, a possibility existed that mRNA had a transcriptional precursor, HnRNA, that now could be identified and whose in viva processing could be studied, as had been done with ribosomal RNA. The polyadenylic acid would provide a "handle" for the selective isolation and identification of these species. Second, the methods of selective isolation of polyadenylate-containing RNA could be taken advantage of for the isolation of "pure" eukaryotic message, a heretofore difficult task. From the studies on the precursor-product relation- ship of HnRNA and mRNA, it was postulated that the polyadenylic acid sequence might be necessary for the proper processing, transport and functionality of mRNA (for a review, see LaTorre and Perry [71]). This hypothesis was suggested from the results of in viva studies on the effect of two anti- biotics on HnRNA and polyadenylic synthesis. Actinomycin blocked the synthesis of HnRNA, but had almost no effect on polyadenylic acid synthesis. Cordycepin, 3'-deoxyadenosine, preferentially inhibited polyadenylate synthesis. In either case, little functional mRNA on polyribosomes was found. The selective inhibition by cordycepin and the knowledge that the transcription was from non-polydeoxythymidylate containing DNA suggested that the polyadenylic acid sequence was a posttranscriptional product. That is, the polyadenylate 31 was not coded for in the template read by DNA-dependent RNA polymerase, but synthesized in the nucleus after the trans- cription had taken place. These qualities suggested an enzyme, separate from DNA-dependent RNA polymerase, that could synthesize long polyadenylic acid sequences on the 3'-hydroxyl end of nuclear transcription products was re- sponsible. Of course, this description fit poly A polymerase. Thus, the enzymatic activity that had earlier been a mere curiosity was implicated as having an important physiological function in the biogenesis of mRNA. Even though this is an attractive hypothesis, there are several additional facts about polyadenylate-containing RNA which need to be accounted for. For instance, the eukaryotic poly A polymerases are found not only in the nucleus, but also in the cytoplasm. A trivial explanation for the enzyme in the cytoplasm could be that some of the enzyme leaked out of the nucleus. However, as has already been demonstrated there are some catalytic differences for distinguishing between the enzymes isolated from the nucleus and the cytoplasm. A different explanation for the activity in the cytoplasm involves some additional findings about polyadenylic acid. The mRNA of certain viruses that replicate only in the cytoplasm contain polyadenylic acid (72). There is also evidence for the polyadenylate sequence of mRNA becoming shorter with time when in the cytoplasm (73). It is possible that the addition of ribonuclease-resistant 32 polyadenylic acid to mRNA provides a mechanism by which a cytoplasmic poly A polymerase could regulate certain classes of mRNA. The extent of polyadenylation would then be cor- related to the effective lifetime of that mRNA. In fact, the lifetime of histone mRNA, the only mRNA found not to contain polyadenylic acid, is very short in comparison to other mRNA (69,74). The ubiquity of poly A polymerase in nature suggests an importance for this activity in the biology of a cell. If the hypothesized in viva function for the enzyme is cor- rect, one might expect to find polyadenylated-mRNA also throughout nature. This certainly seems to be true for the nuclear and cytoplasmic fractions from the many types of eukaryotic cells studied (63-65,69,7l,73). Viral (66,75) and tumor (68,70) mRNA also contain these polyadenylic acid sequences. Even mitochondrial DNA transcripts have a polyadenylic acid tract about 50 residues in length at their 3'-hydr0xyl end (76). Not all of the characteristics of the polyadenylation of nuclear mRNA, though, can be expected to apply to mito- chondrial "mRNA." For example, in mitochondria there is no analogous nuclear to cytoplasmic transport for which polyadenylation could be associated. Also, cordycepin was found to inhibit the transcription of HeLa mitochondrial DNA, but did not effect the posttranscriptional synthesis of polyadenylic acid on the messenger-like RNA of mitochondria 33 (77). This is just opposite the effect that was observed in the nucleus (69,71). Thus, the enzymes responsible for transcription and posttranscriptional polyadenylation in mitochondria could be expected to be catalytically different than those from the nucleus. Indeed, it has already been shown that the poly A polymerases are different. Since polyadenylate synthesis probably occurs subsequent to the processing of the mitochondrial DNA transcript (77), only the imparting of functionality for translation or regulating the turnover of mRNA from the proposed model remains appli- cable to mitochondria. Based upon the frequently used analogy between prokaryotes and mitochondria, one would expect to find in bacteria polyadenylated-mRNA and probably the same limi- tations on the model as has been described for mitochondria. Two groups of investigators attempted to demonstrate the existance of polyadenylated-mRNA in E. aaZi. Edmonds and Kopp (78) reported the isolation of polyadenylic acid seg- ments, similar in size to those from mammalian cells, from the steady state-labeled total RNA of E. caZi. The relative content of polyadenylic acid from those cells, though, was one-tenth of that found in mammalian cells. Perry at al. (79) found fault with their analysis on the basis that rather inaccurate results for estimating the polyadenylate content were obtained by using steady state-labeled RNA. In addition, Edmonds and Kopp had failed to demonstrate that the 34 polyadenylate was actually part of polyribosomal RNA. In order to obtain a more satisfactory answer, Perry and his coworkers isolated pulse-labeled polyribosomal mRNA. Based upon a variety of the usual techniques used for the analysis of mammalian mRNA, the polyribosomal mRNA of E. aaZi was found to contain no appreciable polyadenylic acid. Even though there are limitations to the resolution of these techniques, Perry and coworkers demonstrated that prokaryotic mRNA does not contain polyadenylate tracts even to the extent found in mitochondria. The lack of polyadenylated—mRNA in bacteria suggests that the prokaryotic poly A polymerase might not mediate the same process in viva as has been suggested for its eukaryotic counterpart. This presents a paradox; the many in vitra catalytically similar polyadenylate synthesizing enzymes do not have the same unique function in all levels of evolu- tionary development. Of course, it is also very probable ‘that the true in viva function for any poly A polymerase 1135 not yet been discovered. Gaining a better understanding c>f these unusual and important enzymes will provide a basis fc)r an understanding of their role in viva. L10. ill. 112. 13.. L4. 15. REFERENCES Chamberlin, M., and Berg, P. (1964) J. Mol. Biol. 8, .708-726. Kato, K.; Gonclaves, J. M.; Houts, G. E.; and Bollum, F. J. (1967) J. Biol. Chem. 242, 2780-2789. Klee, C. B. (1971) in Procedures in Nucleic Acid Re— search. Edited by S. L. Cantoni and D. R. Davies. Vol. II. New York: Harper and Row. pp. 896-911. Straus, D. B., and Goldwasser, E. (1961) J. Biol. Chem. 236, 849-856. Klemperer, H. G., and Kammen, H. O. (1961) Biochem. Biophys. Res. Commun. 6, 344-348. Cory, J. G.; Benson, A. W.; and Girgenti, A. J. (1971) Biochem. Biophys. Res. Commun. 42, 778-783. Edmonds, M. (1965) J. Biol. Chem. 240, 4621-4628. Federoff, N. V., and Zinder, N. D. (1971) Prac. Nat. Acad. Sci. U.S.A. 68, 1838-1843. Burkard, G., and Keller, E. B. (1974) Proc. Nat. Acad. Sci. U.S.A. 7l, 389-393. August, J. T.; Ortiz, P. J.; and Hurwitz, J. (1962) J. Biol. Chem. 237, 3786-3793. Gottesman, M. E.; Canellakis, Z. N.; and Canellakis, E. S. (1962) Biochim. Biophys. Acta 61, 34-42. Smith, I., and August, J. T. (1966) J. Biol. Chem. 241, 3525-3532. Hardy, S. J. S., and Kurland, C. G. (1966) Biochemistry 5, 3668-3676. Hardy, S. J. S., and Kurland, C. G. (1966) Biochemistry 5, 3676-3684. Ortiz, P. J.; August, J. T.; Watanabe, M.; Kaye, A. M.; and Hurwitz, J. (1965) J. Biol. Chem. 240, 423-431. ‘35 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. J28. 2S9. 30. 31 - 32. 33. 36 Colvill, A. J. E., and Terzi, M. (1968) Biachim. Biophys. Acta l55, 394-400. Gross, H. J., and Alberty, H. (1962) Biochem. Biophys. Res. Commun. 46, 1581-1587. Terzi, M.; Cascino, A.; and Urbani, C. (1970) Nature 226, 1052-1054. Ohasa, S., and Tsugita, A. (1972) Nature New Biol. 240, 35-38. Ohasa, S.; Tsugita, A.; and Mii, S. (1972) Nature New Biol. 240, 39-41. Schifer, R.; Zillig, W.; and Priess, H. (1972) Fed. Eur. Biochem. Soc. Lett. 25, 87-90. Sippel, A. E. (1973) Eur. J. Biochem. 37, 31-40. Modak, M. J., and Srinivasan, P. R. (1973) J. Biol. Chem. 248, 6904-6910. Payne, K. J., and Boezi, J. A. (1970) J. Biol. Chem. 245, 1378-1387. Edmonds, M., and Abrams, R. (1960) J. Biol. Chem. 235, 1142-1149. Edmonds, M., and Abrams, R. (1962) J. Bial. Chem. 237, 2636-2642. Winters, M. A., and Edmonds, M. (1973) J. Biol. Chem. 248, 4756-4762. Winters, M. A., and Edmonds, M. (1973) J. Biol. Chem. 248, 4763-4768. Dravid, A. R.; Pete, N.; and Mandel, P. (1971) J. Neurochem. 18, 299-306. Kato, T., and Kurokawa, M. (1970) Biochem. J. ll6, 599-609. Niessing, J., and Sekeris, C. E. (1972) Fed. Eur. Biochem. Soc. Lett. 22, 83-88. Niessing, J., and Sekeris, C. E. (1973) Nature New Biol. 243, 9-12. Burden, R. H. (1963) Biochem. Biophys. Res. Commun. ll, 472-476. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 445. 46. 417. 48- 49. so. 51. 37 Tsiapalis, C. M.; Dorson, J. W.; DeSante, D. M.; and Bollum, F. J. (1973) Biochem. Biophys. Res. Commun. 50, 737-743. Coleman, M. S.; Hutton, J. J.; and Bollum, F. J. (1974) Nature 248, 407-409. Giron, M. L., and Huppert, J. (1972) Biochim. Biophys. Acta 287, 438-447. Giron, M. L., and Huppert, J. (1972) Biachim. Biophys. Acta 287, 448-455. Mans, R. J. (1973) Fed. Eur. Biochem. Soc. Lett. 33, 245-246. Edmonds, M., and Abrams, R. (1957) Biachim. Biophys. Acta 26, 226-227. Hadidi, A., and Sagar Sethi, V. (1974) Fed. Prac. Fed. Amer. Soc. Exp. Biol. 33, 1485. Klemperer, H. G. (1963) Biachim. Biophys. Acta 72, 416-420. Klemperer, H. G. (1965) Biachim. Biophys. Acta 95, 251-261. Jacob, S. T., and Schindler, D. G. (1972) Biochem. Biophys. Res. Commun. 48, 126-134. Jacob, S. T.; Schindler, D. G.; and Morris, H. P. (1972) Science Z78, 639-640. Rose, K. M., and Jacob, S. T. (1974) Fed. Prac. Fed. Amer. Soc. Exp. Biol. 33, 1485. Sachar, R. c. (1968) Biochim. Biophys. Acta Z69, 58-66. Sasaki, K., and Tazawa, T. (1973) Biachem. Biophys. Res. Commun. 52, 1440-1449. Mans, R. J., and Walter, T. J. (1971) Biochim. Biophys. Acta 247, 113-121. ‘ Walter, T. J., and Mans, R. J. (1970) Biachim. Biaphys. Acta 217, 72-820‘ Mans, R. J. (1971) Biochem. Biophys. Res. Commun. 45, 980-983. Chung, C. W.; Mahler, H. R.; and Enrione, M. (1960) J. Biol. Chem. 235, 1448-1461. 52. S3. 54. 55. 56. S7. 58. 59. 60. 61. 62. 63. 64. (55. 6(5. 677. 68. 38 Venkataraman, P. R., and Mahler, H. R. (1963) J. Biol. Chem. 238, 1058-1067. Haff, L. A., and Keller, E. B. (1973) Biochem. Biophys. Res. Commun. 5l, 704-710. Twu, J. S., and Bretthauer, R. K. (1971) Biochemistry l0, 1576-1582. Hyatt, E. A. (1967) Biachim. Biophys. Acta Z42, 246- 253. Hyatt, E. A. (1967) Biochim. Biophys. Acta Z42, 254- 262. Galet, H., and Prevec, L. (1973) Nature New Bial. 243, 200-203. Reichman, M. E.; Villareal, L. P.; Kohne, D.; Lesnaw, J.; and Holland, J. J. (1974) Virology 58, 240-249. Stoltzfus, C. M.; Morgan, M.; Banerjee, A. K.; and Shatkin, A. J. (1974) J. Viral. 23, 1338-1345. Brown, M.; Dorson, J. W.; and Bollum, F. J. (1973) J. Virol. 12, 203-208. Kates, J., and Beeson, J. (1970) J. Mal. Biol. 50, 19-33. Moss, B.; Rosenblum, E. N.; and Paoletti, E. (1973) Nature New Biol. 245, 59-62. Darnell, J. E.; Jelinek, W. R.; and Molloy, G. R. (1973) Science l8l, 1215-1221. Lim, L., and Canellakis, E. S. (1970) Nature 227, 710-712. Hadjivassiliou, A., and Brawerman, G. (1966) J. Mal. 87:01. 20, 1'70 Kates, J. (1970) Cold Spring Harbor Symp. Quant. Biol. 35’ 743-752. . Brawerman, G.; Mendecki, J.; and Lee, S. Y. (1972) Biochemistry ll, 792-798. Lee, S. Y.; Mendecki, J.; and Brawerman, G. (1971) Proc. Nat. Acad. Sci. U.S.A. 68, 1331-1335. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. £32. 39 Adesnik, M.; Salditt, M.; Thomas, W.; and Darnell, J. E. (1972) J. Mal. Biol. 7l, 21-30. ' Edmonds, M., and Caramela, M. G. (1969) J. Biol. Chem. 244, 1314-1324. LaTorre, J., and Perry, R. P. (1973) Biachim. Biophys. Acta 335, 93-101. Yogo, Y., and Wimmer, E. (1972) Prac. Nat. Acad. Sci. U.S.A. 69, 1877-1882. Sheiness, D., and Darnell, J. E. (1973) Nature New Bial. 24l, 265-268. Adesnik, M., and Darnell, J. E. (1972) J. Mal. Bial. 67, 395-406. Armstrong, J. A.; Edmonds, M.; Phillips, B. A.; and Vaughan, M. H. (1972) Science Z76, 526-528. Hirsch, M., and Penman, S. (1973) J. Mal. Biol. 80, 379-391. Hirsch, M., and Penman, S. (1974) J. Mal. Biol. 83, 131-142. Edmonds, M., and Kopp, D. W. (1970) Biochem. Biophys. Res. Commun. 4l, 1531-1537. Perry, R. P.; Kelley, D. E.; and LaTorre, J. (1972) Biochem. Biophys. Res. Commun. 4l, 1593-1600. Brawerman, G. (1974) Annu. Rev. Biochem. 43, 621-642. Gerard, G. F.; Gurgo, C.; Grandgenett, D. P.; and Green, M. (1973) Biochem. Biophys. Res. Commun. 53, 194-201. Petermann, M. L. (1964) The Physical and Chemical Properties of Ribosomes. New York: Elsevier, p. 174. SECTION II A KINETIC AND STRUCTURAL CHARACTERIZATION OF ATP:RNA ADENYLYLTRANSFERASE FROM PSEUDOMONAS PUTIDA 40 ABSTRACT A catalytic and structural study of ATP:RNA adenylyl- transferase (EC 2.7.7.19) from the particulate fraction of Pseudamanas putida was made. During the large-scale purifi- cation of this enzyme, designated adenylyltransferase B, a previously undetected ATP-incorporating activity, designated adenylyltransferase A, was observed. Adenylyltransferases A and B were indistinguishable catalytically. The incorpo- ration of AMP from ATP into a polymeric product by the adenylyltransferases required Mg2+ and a polyribonucleotide as primer. The rate of incorporation of CTP was about 12% that of AMP, while GTP, UTP, 2'-dATP or ADP were not utilized by these adenylyltransferases. The Km for ATP was 1-2 x 10-4 M. The primer requirement was satisfied by transfer RNA, ribosomal RNA, poly (C), Ap(Ap)zA, and ApApA. DNA, QB RNA, poly(A), poly(U), poly(dT), or ApA were not effective primers. The Km for ribosomal RNA, in terms of nucleotide phosphate, was 2-4 x 10.4 M. The polymerization reaction catalyzed by the two adenylyltransferases gave the fastest rate at pH 9.5, and was inhibited by inorganic pyrophosphate but not orthophosphate. Adenylyltransferases A and B were distinguishable by their chromatographic and sedimentation properties. 41 42 Adenylyltransferases A and B were resolved by phosphocellulose and by poly(U)-Sepharose chromatographies. When chromato- graphed on Bio—Gel P-100, adenylyltransferase A was excluded, while B penetrated the gel. By sedimentation through glycerol gradients, adenylyltransferase A was determined to have a sedimentation coefficient (S20,w) of 9.3 S and B of 4.3 S. The molecular weight of adenylyltransferase A was estimated to be 185,000 and that of adenylyltransferase B to be 50,000 to 60,000. Apparently, adenylyltransferase A was generated from adenylyltransferase B during the purification. Two derivatives of the antibiotic rifamycin, AF/013 and AF/DNFI, inhibited the AMP incorporation catalyzed by adenylyltransferases A and B. The polymerization rate by either enzyme was inhibited 50% at 5 ug per ml of AF/Ol3 or 10 pg per ml AF/DNFI. The 5'-triphosphate derivative (3'-dATP) of the drug cordycepin (3'-deoxyadenosine) was a competitive inhibitor with ATP for both adenylyltransferases. The Ki for 3'-dATP was 6-10 x 10'4 M. Several other analogs of ATP, 2'-dATP, 2'-0-methylATP, or the fluorescent 3-8-0- ribofuranosylimidazo [2, 1-i] purine 5'-triphosphate did not affect the activity of adenylyltransferase A or B. Poly(U) and poly(dT) were competitive inhibitors of the ribosomal RNA-primed polymerization reaction. The Ki for poly(U) or poly(dT), in terms of nucleotide phosphate, was 4-10 x -6 10 M for adenylyltransferases A and B. The inhibition was a result of the competition between the non-priming 43 poly(U), or poly(dT), and ribosomal RNA for the primer- binding site on the enzyme. INTRODUCTION ATP:RNA adenylyltransferase (EC 2.7.7.19) catalyzes the polymerization of ATP into polyadenylic acid. This enzyme preferentially utilizes ATP as substrate. Its activity depends upon a divalent metal ion and a primer. The primer is a polyribonucleotide with a free 3'-hydroxy1 end to act as an initiation point for the polymerization. Because of the nature of the product it polymerizes, this enzyme is commOnly referred to by the trivial name, poly A (Adenylate) polymerase. Poly A polymerase is ubiquitous in nature, having been isolated from such a wide variety of sources as mammals (1-6), plants (7), avia (8), yeast (9,10), echinoderms (11,12), bacteria (13-16) and viruses (17,18). It has been widely observed that polyadenylic acid (polylAJ) sequences are covalently linked to the heterogeneous DNA-like RNA in the nucleus (HnRNA) and to messenger RNA (mRNA) in the cytOplasm of eukaryotic cells (for a review, see Darnell et a2. [19], and Brawerman [20]). Polyadeny- lation appears to be important for the proper processing of HnRNA to mRNA and for the transport of functional mRNA to the cytoplasm. The poly(A) sequences are believed to arise by a posttranscriptional addition to the 3'-hydroxy1 end 44 45 of HnRNA. This synthesis is speculated to be catalyzed not by a DNA-dependent RNA polymerase, but by a poly A polymerase- type enzyme. The mRNA of eukaryotic mitochondria also contain 3'-termina1 poly(A) sequences (21,22). An exami- nation of E. cali mRNA, however, failed to detect any similar poly(A) tracts (23,24). The ubiquity of poly A polymerase in nature certainly suggests an importance for this activity in the biology of the cell. In eukaryotes this function might be the poly(A) addition on DNA transcripts. However, none of the investi- gations on this enzyme have provided direct evidence of its in viva role. The reports on the enzymes from prokaryotes have presented many different structural and catalytic properties for poly A polymerase, making for a confusing situation. August et al. reported the first purification of a particulate poly A polymerase from E. cali (13). Recently, Sippel purified a particulate enzyme from E. cali, different from the August enzyme in its pH optimum and preference for metal ion (14). From the non-particulate fraction of E. cali there was isolated an apparent primer- independent enzyme, also with unique pH optimum, metal ion preference and molecular size (15). Another ATP-incorporating activity, RNA-primed polynucleotide pyrophosphorylase, was detected in E. cali by Schafer et al. (43). However, this enzyme was distinguishable from poly A polymerase, since other nucleotides were incorporated as efficiently as ATP. 46 It is difficult to determine from the data whether or not there are several distinct poly A polymerase activities in a prokaryotic cell. Such a decision must await the further characterization of the reported poly A polymerases. When the work for this report was initiated, no structural and few kinetic properties of poly A polymerase from prokaryotes was known. It was the objective of this project to further purify the particulate enzyme from the bacterium Pseudomonas putida first reported by Payne and Boezi (16), then obtain structural and further kinetic characteristics of it. During the course of the purifi- cation, a previously undetected ATP:RNA adenylyltransferase activity was observed. This new activity and that previously described (16) were indistinguishable catalytically, but they differed in molecular size and in chromatographic properties. That which follows is the structural and ' kinetic characterization of these adenylyltransferases. EXPERIMENTAL PROCEDURE Materials DEAE—Sephadex (A-ZS), Sepharose 4B and Blue Dextran 2000 were purchased from Pharmacia Fine Chemicals, Inc. and Whatman cellulose phosphate (P-l) was from H. Reeve Angel, Inc. Bio-Gel P-100 (100-200 mesh) and P-200 (50-100 mesh) were obtained from Bio-Rad Laboratories. Calf thymus DNA, herring sperm DNA, poly(A), poly(C), poly(G), poly(l), 1 hemoglobin (type I), cytochrome poly(U), ApA, ApApA, PMSF, c (type VI), ribonuclease A (type I-A), pepsin, and bovine serum albumin were from Sigma Chemical C0. Poly(dA), poly(dT), and Q8 viral RNA were obtained from Miles Labo- ratories, Inc. Pancreatic DNase I and E. cali alkaline phosphatase (both electrophoretically purified) were from Worthington Biochemical Corp. Dithiothreitol, Ap(Ap)2A, 2'-dATP, e-ATP, and all unlabeled 5’-triphosphate derivatives of the ribonucleotides were obtained from P-L Biochemicals, Inc. Nitrocellulose membrane filters (type B-6) were obtained from Schleicher and Schuell, Inc. I3H-labeled nucleatides.were purchased from Schwarz/Mann. Rifamycin derivatives were the lThe abbreviatiOns used that are not listed in the Journal of Biological Chemistry 249, l (1974) are: PMSF, phenylmethylsulfonyl flouride; e-ATP, 3-BeD-ribofuranosyl- imidazo [2, l-il purine 5'-triphosphate; 3'-dATP, 31-, deoxyadenosine 5'-triphosphate; poly(Am), poly(Z'-0-methyl adenylic acid); poly(Um), poly(2'-0-methyluridylic acid); AmTP, 2'-0-methyladenosine 5'-triphosphate; and SDS, sodium dodecyl sulfate. 47 48 gifts of Dr. Luigi Silvestri, Gruppo Lepetit, Inc., Milan, Italy. .SH-labeled R17 viral RNA was a gift from B. Hall, University of Washington, Seattle. Chromatographically pure 3'-dATP was a figt from R. Desrosiers of this department. B. subtilus ribosomal RNA, yeast transfer RNA, E. cali ribosomal and transfer RNA, poly(Am), poly(Um), and AmTP were all gifts from F. Rottman also of this department. P. putida ribosomal RNA was prepared according to a procedure previously described (16). Ribosomal RNA and transfer RNA were chromatographed electrophoretically on polyacrylamide gels and found to be homogeneous. Analytical Methods Protein concentration was determined by the method of Lowry et al. (25) with bovine serum albumin as the standard. Calf thymus DNA was denatured by heating to 100° for 10 min followed by quick cooling in ice. QB RNA was purified by sedimentation through a sucrose gradient before use. The concentrations of Q8 RNA, oligoadenylic acid and calf thymus DNA were determined spectrophotometrically based on the extinction coefficient B150 = 200. The molar extinctiOns, e(P), used to determine the concentration 3 of ribosomal and transfer RNA at 260 nm were 7.4 x 10 (26) and 7.2 x 103 (27), respectively. The molar extinctions, 3 e(P), for polyribonucleotides were: 10.5 x 10 at 257 nm, 3 3 9.2 x 10 at 260 nm and 6.5 x 10 at 267 nm for poly(A), 3 poly(U), and poly(C), respectively (28) and 10.2 x 10 and 3 9.4 x 10 at 259 nm for poly(Am) and poly(Um), respectively 49 (29). The molar extinctions, e(P), for polydeoxyribo- 3 nucleotides were 8.1 x 103 and 9.7 x 10 at 260 nm for poly(dT) and poly(dA), respectively (26). Growth of Pseudomonas putida P. putida (ATCC 12633) was grown in a medium which contained the following in grams per liter: 6, NazHPO4; 3, KH PO 8, NaCl; 5, bacto yeast extract (Difco); 5, 2 4’ bacto-tryptone (Difco); and 5, glucose. P. putida was grown in lOO-liter volumes in a New Brunswick Fermacell, model F-l30, at 32° with aeration at 140 liters per min. After entering early stationary phase, the cells were harvested in a Sharples centrifuge. The 1000 to 1300 g (wet weight) of packed cells were stored at -20°. Assaonf ATP:RNA Adenylyltransferase The standard assay measured the conversion of [3H] ATP into acid-insoluble product. Unless otherwise indicated, the reaction mixture (0.25 ml) contained 30 mM glycine-NaOH 3 3 (pH 9.5), 20 mM MgCl 0.8 mM [ HJATP (about 2 x 10 cpm 2, per nmol), saturating amounts of P. putida ribosomal RNA, 100 mM KCl and an appropriate amount of enzyme. After incubation at 37° for 10 min, the reaction mixture was mixed with 0.1 ml of 0.1% SDS. Cold 10% trichloroacetic acid -1% sodium pyrophosphate (Sml) and 250 ug of herring Sperm DNA were added. After 10 min at 0-4°, insoluble material was collected on a nitrocellulose filter. The 50 filter was washed with three 5 ml portions of cold 10% trichloroacetic acid -1% sodium pyrophosphate, dried and then monitored for radioactivity using liquid scintillation spectrometry. The scintillation fluid (5 ml) contained 4 g of 2,5-bis[2-(5-tert-butylbenzoxazolyl)] thiophene per liter of toluene. One unit of enzyme activity is defined as that amount of enzyme which catalyzes the incorporation of one nmol of AMP into an acid-insoluble product in 10 min at 37°. Specific activity is expressed as units per mg of protein. The amounts of enzyme activity which were added to the reaction mixture were within the activity range that gave direct proportionality to the rate of AMP incorporation. Preparation of Poly(U)-Sepharose Poly(U)-Sepharose was prepared by a procedure similar to that of Poonian et al. (30). After activation by cyanogen bromide, the Sepharose 4B was thoroughly washed by filtration with 25 volumes of 100 mM potassium phosphate (pH 7.5). A slurry of activated Sepharose 4B was obtained by mixing it with an equal volume of the above buffer. To this slurry was added an equal volume of poly(U), 2.4 x 10.4 M nucleotide phosphate, and the entire volume was stirred slowly for 18 hours at 4°. The final product was 2 x 10-7 mol nucleotide phosphate of poly(U) bound per ml of packed Sepharose as determined by the method of Failla and Santi (31). RESULTS Purification of ATP:RNA Adenylyltransferases A and B The entire purification was performed at 0-4°. Preparation of Initial Extract.--A block of frozen Pseudomonas putida cells (500 g wet weight) was cut into small pieces, then 500 ml of buffer (10 mM Tris-HCl [pH 2 8.0], 10 mM MgCl 0.1 mM EDTA, 5 mM 2-mercaptoethanol 2, and 0.1 mM PMSF) was added. The mixture was stirred until a smooth suspension was obtained. The cell suspension was passed thrice through the pressure chamber of a Manton- Gaulin lab homogenizer, which was maintained at 7000 p.s.i. Pancreatic DNase I (0.5 ug per ml) was added to the cell extract and the extract was stirred for 15 min. Following centrifugation at 16,000 x g for 60 min, the resulting supernatant fraction was carefully decanted and diluted to 840 ml with the above buffer (Initial Extract). Sedimentation.--The initial extract fraction was centrifuged at 80,000 x g for 4 hours. The supernatant fraction was carefully removed by suction and discarded, and the fluffy pellet was stored at -20° overnight (80,000 x g Pellet). Solubilization.--The 80,000 x g pellet fraction was thawed and homogenized in buffer (100 mM glycine-NaOH [pH 2The pH of buffers used in the purification were measured at 0-4°. 51 52 10.6], 5 mM 2-mercaptoethanol, and 0.1 mM PMSF) using a Potter-Elvejnem homogenizer. After adjusting the volume to 420 ml, the suspension was centrifuged at 80,000 x g for 2 hours. The supernatant fraction was carefully decanted (pH 10.6 Supernatant). pH Fractionation.--Solid KCl was added slowly to the pH 10.6 supernatant fraction to a final concentration of l M. Cold 1 M HCl was then added dropwise with stirring until pH 5.0 was reached. The suspension was stirred for 2 hours, then centrifuged at 16,000 x g for 20 min. The supernatant fraction was dialyzed overnight against 14 liters of 10 mM Tris-HCl (pH 8.0), 5 mM 2-mercaptoethanol, and 0.1 mM PMSF (pH 5 Supernatant). pH Concentration.--Cold 1 M HCl was added dropwise with stirring to the pH 5 supernatant fraction until pH 3.2 was reached. After stirring for 15 min, the suspension was centrifuged at 16,000 x g for 15 min. The supernatant fraction was poured off and discarded. The pellet was dissolved in 100 mM glycine-NaOH (pH 10.1), 5 mM 2- mercaptoethanol, and 0.1 mM PMSF by stirring for 4 hours. After the volume of the solution was adjusted to 300 ml, solid KCl was added to a final concentration of l M. The solution was dialyzed overnight against 12 liters of 10 mM Tris-HCl (pH 8.0), l M KCl, 5 mM 2-mercaptoethanol, and ’0.1 mM PMSF (pH 3.2 Precipitate). 53 Bio-Gel P-200 Gel Filtration.-—Bio-Gel P-ZOO, previously equilibrated with 10 mM Tris-HCl (pH 8.0), l M KCl, 5 mM 2-mercaptoethanol, and 0.1 mM PMSF, was used to prepare a column 10 x 87 cm. The column was rigged for ascending flow and thoroughly washed before use. The pH 3.2 precipitate fraction was applied to the column and eluted by the addition of the same buffer. The effluent pattern of the column, Figure 1, shows good resolution between ATP-incorporating activity and the majority of the A280- absorbing material. Some ATP-incorporating activity, how- ever, did elute with the A280-absorbing material. The amount varied up to 15% of the total enzyme activity recovered from the column, depending upon the purification trial. The peak fractions of ATP-incorporating activity, cor- responding to elution volumes from 3.6 to 4.3 liters, were pooled and dialyzed against 28 liters of 10 mM Tris-HCl (pH 9.1), 5 mM 2-mercaptoethanol, and 0.1 mM PMSF (Bio-Gel P-ZOO). DEAE-Sephadex Chromatography.--A column (2.5 x 20 cm) of DEAE-Sephadex was prepared, then washed thoroughly with 10 mM Tris-HCl (pH 9.1), 5 mM 2-mercaptoethanol, and 0.1 mM PMSF. The Bio-Gel P-200 enzyme fraction was applied to the column. The column was then washed with 150 ml of the above buffer. Upon application of a 500 ml linear gradient from 0 to 1 M KCl in the above buffer, the ATP- incorporating activity was eluted as a single peak at 0.2 M 54 KCl, separated from A -absorbing material. The peak 260 fractions of the ATP-incorporating activity were pooled and dialyzed against 12 liters of 10 mM potassium phosphate (pH 7.1), 250 mM KCl, 5 mM 2-mercaptoethanol, and 0.1 mM PMSF (DEAE-Sephadex). Phosphocellulose Chromatography.--A column (1.2 x 8 cm) of phosphocellulose was prepared and washed thoroughly with 10 mM potassium phosphate (pH 7.1), 250 mM KCl, 5 mM 2-mercaptoethanol, and 0.1 mM PMSF. After applying the DEAE-Sephadex enzyme fraction to the column, the column was washed with 25 ml of the above buffer. ATP-incorporating activity was eluted from the column by application of a 100 ml linear gradient from 0.25 to l M KCl in the above buffer. As shown in Figure 2, two peaks of ATP-incorporating activity were eluted from the column, one at 0.3 M KCl, designated A, and the other at 0.6 M KCl, designated B. This elution pattern was reproducible, but depending upon the purifi- cation trial, varying amounts of peak A activity were obtained, ranging from S to 50% of the total activity eluted from the column. The fractions containing the activity of peaks A and B were pooled separately and dialyzed against 4 liters of 10 mM potassium phosphate (pH 7.1), 250 mM KCl, 5 mM 2-mercaptoethanol, and 0.1 mM PMSF (Phosphocellulose A and B). Summary and Comments on Purification Through Phosphocellulose Chromatography.--A summary of a typical 55 ACV woumpompoucw E5. Hoe: mam XIV moEoEepsmmoE xufl>fluosecou E95 onHEEouow mm Eogmwuqoocou Hum 02E .mmammuomm 4euom mcwumuompoucw-me< mo xmmmm as» pom .wmuuofiaou who: come HE N.N mo mEOMuompm cm was :ME pom HE H um pemofio>op mm; EEDHoo och .oExNEo esp ousfio ow com: me: mmzm 2E H.o can .HocmcuooummoonIN 2E m .mH.E may oumnmmoca Esflmmmuom 2E OH EM HUM E o.H ow mm.o Eowm ucofiwmpm hmoEflH HE OOH < .Emmwzn oEmm ecu mo HE mm cue: eozmmz mm: EEEHou one .Eofluowwm oExmce opp mo :oflumUMHQQm MEEZOHHOE .am2a as H.o .Hoeegaaepaeuaae-~ as m .HUE as omN .AH.E may aaegameaa Esflmmmpom SE oH new: eeumpnwfiflsco omoHDHHmuonmmoza mo AEo w x N.HV EEEHou m ow :flE won HE mn.o um woflamam mm: scepomhw eExuao xowmnmom-mzexm one :Qwaoommoh .E: oav cam omo um AHHmo -wwuoEouocmouuuomm oemE mm: unosamwo one EM canofimoEo: @Em Emhuxea oaam mo :ofluoouow och .mcofipfivcou Hmoflucoew have: EEEHoo oEmm any so nonempwoumE -ownu mmz Anny Ewnofiono: paw .flzaqv ommcomOwanow oumuuma .mo>v :mauxom osam mo Eoflpsaom m .onwm AmasooHoEazwaEsaoo on“ oumhnfiflmu op povho EH ACV wouwhompoocw as? HoE: cam ”Adv E: 33 um cousmmoE oocmnAOmn< .mmsmmuomm 4fiuom mcflumhomwoocw-me< may you .mEOwuomhm HE-HN ova kn vezofiaow .wmuuoHHoo who: oEsao> efio> HE comm ecu mo HE omnfi .EoAE .homwsn oEmm one mo EOMHEwwm Hmsceucoo ecu xn EME Mom HE m~.H mo mush m we onm wcwecoumm mo mEOMHchoo have: womofio>ow mm: :Esaoo och .mmZm 2E H.c use .HocmnuoouamouoEIN as m .HUE 2 H .Ao.w may Hu:-mase 2e OH sea: easeseaaasca oom-a aeu-oam mo nEu mm x oav EEEHoo m ow wowfimam mm: :owuumwm oERNEo oumuwawoohn N.m In one .oumufiafiuohm N.m In :ofiuumhw mo :ofiumpuaww Haw oo~-m HooIOEm maocmm ummHVII.H owsmfim 56 QQ'QO " (I/Iow)[Io>I] ‘ 1 l J 0 V N HOW“) NOILVHOdHOONI dWV “I“ 092 1.17 EONVBHOSSV HOW“) NOHVUOdUOONI dVW I20 ‘ 60 3O EFFLUENT VOLUME ( ml ) EFFLUENT VOLUME ( I l 57 purification is presented in Table I. The purification procedure used was similar to that described by Payne and Boezi (16). The scale of the purification, however, was increased over ten-fold, from 40 g to 500 g of cells. Gel filtration was on Bio-Gel P-200 rather than Sephadex G-100; all buffers used after DEAE-Sephadex chromatography con- tained at least 250 mM KCl; and 5 mM 2-mercaptoethanol and 0.1 mM PMSF were present in all buffers throughout the purification. The activity of enzyme fraction phosphocel- lulose B, but not of fraction phosphocellulose A, was detected by Payne and Boezi (16). Those authors described some of the catalytic properties of fraction phosphocellulose B, but did not examine its molecular structure. TABLE I.--Purification of ATP:RNA Adenylyltransferases A and B. Enzyme Fraction Specific Activity Total Units Initial Extract l4 1 260,000 pH 10.6 Supernatant 16 96,000 pH 5 Supernatant 60 70,000 pH 3.2 Precipitate 120 70,000 Bio-Gel P-200 440 58,000 'DEAE-Sephadex 990 30,000 Phosphocellulose A 2,000 5,500 Phosphocellulose B 6,900 8,300 58 Bio-Gel P-100 Gel Filtration.--Bio-Gel P-lOO gel filtration was used as the last step in the purification procedure. When enzyme fraction phosphocellulose A or B was chromatographed on a Bio-Gel P-100 column, however, most of the enzyme activity was lost. Furthermore, the protein concentrations of the peak fractions that contained the enzyme activities A and B which were recovered after gel filtration were not measureable by the method of Lowry et al. (25). Consequently, specific enzymatic activities for A and B through Bio-Gel P-100 gel filtration cannot be reported. Significant purifications of the enzymes, however, were undoubtedly achieved by gel filtration. For example, as analyzed by SDS polyacrylamide gel electrOphoresis, poly- peptide chains of molecular weights in the 10,000 to 20,000 range, which were present in fraction phosphocellulose B and which amounted to 60% (by weight) of the protein (see Figure 8), were removed from enzyme activity B by Bio-Gel P-100 gel filtration. If Bio-Gel P-lOO gel filtration was performed in the presence of 100 pg per ml bovine serum albumin, most of the enzyme activities of A and B were recovered after gel filtration. When this procedure was to be used in the purification, fractions phosphocellulose A and B were made 100 pg per ml with respect to bovine serum albumin and Concentrated 20- to 40-fold to approximately 1 ml using 59 an Amicon Ultrafiltrator, Model 8 MC with PM—lO filter (Concentrated Phosphocellulose A and B). Bio-Gel P-100, previously equilibrated with 10 mM potassium phosphate (pH 7.1), l M KCl, 5 mM Z-mercaptoethenol, 0.1 mM PMSF and 100 pg per ml bovine serum albumin, was used to prepare a column 0.9 x 90 cm. The concentrated phosphocel- lulose A enzyme fraction (1 ml) was applied to the column and eluted by the addition of the above buffer. In a subse- quent run of the column, the concentrated phosphocellulose B enzyme fraction (1 ml) was chromatographed. Enzyme activi- ties A and B eluted as single peaks of ATP-incorporating activity, as seen in Figure 3. The peak fractions of ATP- incorporating activity for A and B were separately pooled ’and concentrated by dialysis against 50% glycerol, 10 mM potassium phosphate (pH 7.1), 250 mM KCl, 5 mM Z-mercapto- ethanol, and 0.1 mM PMSF (Adenylyltransferase A and B). The adenylyltransferase A and B enzyme fractions could be stored at -20° for at least three months without significant loss of enzymatic activity. Fractions adenylyl- transferase A and B were free of ribonuclease activity. The sedimentation of 3H-labeled R17 RNA through sucrose gradients remained unchanged following incubation of the labeled RNA with either fraction adenylyltransferase A or B. Properties of the Reaction Catalyzed by Adenylyltransferases A and B Characteristics of the Reaction Catalyzed by Adenylyl- transferase A.--Some of the characteristics of the reaction 60 Figure 3.--Bio-Gel P-100 gel filtration of fractions phos- phocellulose A and phosphocellulose B. One ml of the concentrated phosphocellulose A enzyme fraction was applied to a column (0.9 x 90 cm) of Bio-Gel P-100 equilibrated with 10 mM potassium phosphate (pH 7.1), l M KCl, 5 mM Z-mercaptoethanol, 0.1 mM PMSF, and 100 pg per ml bovine serum albumin (panel A). The column was developed at a flow rate of 0.1 ml per min by the continual addition of the same buffer and sixty 0.7 ml fractions were collected. For the assay of ATP- incorporating activity (A), 25 111 samples of the fractions were assayed in the reaction mixture described under EXPERIMENTAL PROCEDURE. Similarly, one ml of fraction concentrated phosphocellulose B was applied to the same Bio-Gel P-100 column, then developed as described above (panel B). For the assay of ATP-incorporating activity (0), 5 ul samples of the fractions were assayed in the reaction mixture described under EXPERIMENTAL PROCEDURE. In or st to calibiate the column for molecular size, a solution of Blue Dextran (V0), alkaline phosphatase (AP), hemoglobin (Hb), and cytochrome c (Cyt c) was chromatographed on the same column under identical conditions. The detection of Blue Dextran, hemoglobin, and cytochrome c in the effluent was made spectrophotometrically at 650, 410 and 410 nm, respectively. Alkaline phosphatase was assayed spectrophotometrically by measuring the appearance of p-nitrophenol at 410 nm (41). AMP INCORPORATION (nmol) 61 ; I I I .. A, a I I, - I “a. M 4 W I —r I I Vo AP Hb Cytc vo AP Hb Cytc - III I I0 20 so 40 EFFLUENT VOLUME (ml) 62 catalyzed by adenylyltransferase A are presented in Table II. Incorporation of ATP into a trichloroacetic acid-insoluble form was dependent on added enzyme, an exogenous RNA primer, 2+ 2+ and Mg . Mn2+ could substitute for Mg The rate of AMP incorporation into acid-insoluble form at 1 mM Mn2+ was about 15% of that observed in the presence of 20 mM Mg2+ These concentrations of the divalent metal ions were those that gave the optimal rates of AMP incorporation. If 10 mM sodium pyrophosphate was added to the reaction mixture, no AMP was incorporated. On the other hand, the addition of 10 mM sodium phosphate to the reaction mixture had no effect on the rate of AMP incorporation. Adenylyltransferase A was relatively specific in its requirement for ATP as the substrate. The rate of reaction using CTP as the substrate was about 12% of that for ATP. GTP, UTP, 2'-dATP or ADP did not serve as substrates. For comparison, the charac- teristics of the reaction catalyzed by adenylyltransferase B are also presented in Table II. These characteristics are in agreement with those reported by Payne and Boezi (16). The results of a study of the primer requirements for adenylyltransferase A are presented in Table III. Ribosomal RNA and transfer RNA, but not QB RNA or denatured calf thymus DNA were effective primers. Of the synthetic polyribonucleotides tested, poly(A) and poly(C) were 'minimally effective as primers. Poly(U) and poly(dT) did not function as primers. The oligonucleotides, Ap(Ap)ZA 63 TABLE II.--Characteristics of the Reactiog Catalyzed by Adenylyltransferase A. Relative Incorporation of Mononucleotides by Components of the Reaction Mixture Adenylyltransferase A Adenylyltransferase B Complete 100 (2.61 :_0.01) 100 (7.78 :_0.04) minus enzyme minus ribosomal RNA minus Mg2+ minus Mth, plus 1 mM Mn2+ 15 15 plus 10 mM sodium perphosphate 0 0 plus 10 mM sodium phosphate 100 100 minus ATP, plus CTP 12 12 minus ATP, plus GTP minus ATP, plus UTP minus ATP, plus 2'-dATP minus ATP, plus ADP 0000 GOOD * The complete reaction mixture (0.25 ml) contained 30 mM glycine-NaOH (pH 9.5), 20 mM MgC12, saturating amounts of P. putida ribosomal RNA, 100 mM KCl, 20 ul adenylyltransferase A or 10 ul adenylyltransferase B, and 0.8 mM labeled nucleoside di- or triphos- phate as indicated. The labeled nucleoside di- or triphosphates were of the following specific radioactivities (cpm per nmol): [3H]ADP, 1.0 x 103; [3H12'-dATP. 1.0 x 103; [3H1GTP, 1.9 x 103; [3H1UTP, 2.3 x 103; and [3H]CTP, 2.0 x 103. The nmol of [3H1nuc1eotide monophosphate incorporated into acid-insoluble product after 10 min of incubation at 37° is given in parentheses. 64 9 TABLE III.--Primer Specificity for Adenylyltransferase A. Relative Incorporation of AMP by Primer Adenylyltransferase.A Adenylyltransferase B It putida ribosomal RNA 100 (6.82 :_0.03) 100 (5.11 i 0.03) E. cali ribosomal RNA 114 114 B. subtilus ribosomal RNA 80 92 E. coli transfer RNA 288 361 S. cerevisiae transfer RNA 441 436 QB RNA 11 30 Denatured calf thymus DNA 3 0 Poly(A) 18 12 Poly(C) 15 26 Poly(U) Poly(dT) 0 0 A 78 111 ‘5‘ 192“ ApApA 38 47 ApA 0 0 NOne 0 0 8 Reaction mixtures (0.25 ml) contained 30 mM glycine-NaOH (pH 9.5), 0.8 mM L3HIATP (1.8 x 103 cpm per nmol), 20 mM MgCl , 100 mM KCl, 20 ul adenylyltransferase A or 10 ul adenyl 1transferase and a saturating amount of primer. The nmol of [ HIAMP incorporated into acid-insoluble product after 10 min of incubation at 37° is given in parentheses. 65 and ApApA, but not ApA functioned as primers for adenylyl- transferase A. In parallel experiments, similar results were obtained for adenylyltransferase B. In additional experiments designed to characterize the reaction catalyzed by adenylyltransferase A, the pH optimum was found to be 9.5. The Km3 for ATP was found to be 1 i 0.5 x 10.4 M and Km for P. putida ribosomal RNA, in 4 M (P). In terms of nucleotide phosphate, was 2 i l x 10- parallel experiments, results identical within experimental error were obtained with adenylyltransferase B. The charac- teristics of adenylyltransferase B were identical to those reported for the less highly purified enzyme fractions used by Payne and Boezi (16). The Effect of ATP Structural Analogs on Adenylyl- transferases A and B.--Four structural analogs of ATP were tested for their effect on AMP incorporation by adenylyl- transferase A and B. The rate of AMP incorporation was measured at various concentrations of ATP in the absence and in the presence of each analog which was at a concen- tration of 0.7 mM. The data were then analyzed using the Lineweaver-Burk double reciprocal plots. Three of the structural analogs of ATP, e-ATP, a fluorescent derivative of ATP, and 2'-dATP and AmTP, analogs of ATP altered at the 3A linear least squares analysis of the data provided the source for the reported values of Km or Ki. The variances following the values represent the error from multiple determinations. 66 2'-position of the ribose moiety, had no effect on AMP incorporation by either adenylyltransferase A or B. The fourth analog of ATP, 3'-dATP, the triphosphate derivative of the drug cordycepin, was an effective competitive inhi- bitor of ATP for adenylyltransferase A and B. The data for adenylyltransferase B is presented in Figure 4. The Ki for 3 3'-dATP with adenylyltransferase A was 1 i 0.1 x 10_ M and for adenylyltransferase B was 6 1 l x 10.4 M. The Km for ATP was 1 i 0.5 x 10.4 M for adenylyltransferase A and 2 i 0.5 x 10.4 M for adenylyltransferase B. The Effect of Rifamycin Derivatives on Adenylyl- transferases A and B.--The rifamycin family of antibiotics are effective inhibitors of bacterial RNA polymerases (32- 34). Only one derivative, rifampicin, has been tested and was found inactive as an inhibitor of bacterial poly A polymerases (14,15,35). In the present study fourteen derivatives of rifamycin, all of them effective inhibitors of P. putida DNA-dependent RNA polymerase,4 were tested for their effects on AMP incorporation by adenylyltransferase A and B. The rifamycin derivatives were dissolved in dimethyl sulfoxide and added to the reaction mixtures at a concentration of 40 ug per ml. As a consequence of these additions, the concentration of dimethyl sulfoxide in the reaction mixtures was 16% (v/v). When this concentration 4H. C. Towle, personal communication. 67 Figure 4.--Inhibition of the ATP-incorporating activity of adenylyltransferase B in the presence of 3'-dATP. The reaction conditions were the same as those described under EXPERIMENTAL PROCEDURE, except that varying amounts of [3H]ATP (2.4 x 103 cpm per nmol) were present in each reaction mixture. AMP incorporated in the complete reaction mixture (O), or in the complete reaction mixture plus 0.7 mM 3'-dATP (A). Velocity is in nmol AMP incorporated per 10 min; and substrate is in mM ATP. l/V 68 69 of dimethyl sulfoxide was added to the standard reaction mixture, AMP incorporation by adenylyltransferase A or B was increased 2.6-fold. In other experiments in which the effect of the concentration of dimethyl sulfoxide on AMP incorporation was examined, 16% was found to be the concen-1 tration which produced the greatest increase in the rate of AMP incorporation. Of the fourteen rifamycin derivatives tested at 40 pg per ml, two derivatives, AF/013 and AF/DNFI, were effective inhibitors of AMP incorporation by adenylyltransferase B (Table IV). As determined in other experiments, a concen- tration of 5 pg per ml AF/013 inhibited AMP incorporation 50% by either adenylyltransferase A or B. AF/DNFI, at a concentration of 10 ug per m1, inhibited 50%. The Effect of Synthetic Polynucleotides on AMP Incorporation by Adenylyltransferases A and B with Ribosomal RNA as the Primer.--A variety of synthetic polynucleotides were tested for the effects that they would have on the rate of AMP incorporation by adenylyltransferase A and B with ribosomal RNA as the primer. The addition to the standard reaction mixture of poly(A), poly(dA), poly(Am), 4M (P). or poly(Um), at a concentration of about 1 x 10' had no effect on the rate of AMP incorporation by adenylyl- transferase A or B. The addition to the standard reaction 'mixture of poly(C) or poly(I) resulted in a 30% lower rate of AMP incorporation by either enzyme. Little or no AMP 70 TABLE IV.--The Effect of Rifamycin Depivatives on Adenylyl- transferase B. Rifamycin Derivative Added to the Reaction Mixture __ ‘.-~-.._.-.—- Relative Incorporation of AMP None Rifamycin AG Rifamide PR/14 PR/l9 4-Desossi SV AF/APR AF/DEI AF/BO AF/AP AF/DA-AMP AF/AOP AF/ABDP-CIS AF/DNFI AF/013 100 (6.31 1 0.03) 113 106 105 100 83 97 97 92 100 92 87 70 * Reaction mixtures (0.25 ml) contained 30 mM’glycine-NaOH (pH 9.5), 20 mM MgC12, 100 mM KCl, 0.8 mM I3HIATP (1.9 x 10 cpm per nmol), a saturating amount of P. putida ribosomal RNA, 1 ul of adenylyl- transferase B, 16% (v/v) dimethyl sulfoxide and the additions noted. The final concentration of a rifamycin derivative in the reaction mixture was 40 ug per ml. The nmol of [3H1AMP incorporated into acid- insoluble product after 10 min of incubation at 37° is given in parentheses. 71 incorporation was observed when poly(U), poly(dT), or poly(G) was added to the standard reaction mixture. The mechanism by which poly(U) and poly(dT) inhibit AMP incorporation was examined. As shown in Table III, poly(U) and poly(dT) did not function as primers for adenylyltransferase A or B. These polynucleotides inhibit AMP incorporation by functioning as competitive inhibitors of the ribosomal RNA primer (Figure 5). In these experi- ments, the rates of AMP incorporation by adenylyltransferase A and B were measured at various concentrations of ribosomal RNA in the absence and in the presence of either poly(U) or poly(dT). The data were then analyzed using the Lineweaver-Burk double reciprocal plots. The Km for the 4 ribosomal RNA primer was 2 1 l x 10- M (P) for adenylyl- 4 M (P) for adenylyltransferase transferase A and 4 1 1 x 10- B. The Ki for poly(U) or poly(dT) for adenylyltransferase I A was 4 1 2 x 10.6 M (P). The Ki for poly(U) or poly(dT) SM (P). for adenylyltransferase B was 1 1 0.2 x 10- Just as adenylyltransferase A and B bind poly(U), the enzymes can bind to a poly(U)-Sepharose column (Figure 6). Upon application of a linear KCl gradient, adenylyl- transferase A eluted at 0.19 M KCl, while in a separate run of the column, adenylyltransferase B eluted at 0.38 M KCl. In both experiments, nearly 100% of the enzyme activity I applied to the column was recovered upon elution. 72 Figure 5.--Inhibition of the ATP-incorporating activity of adenylyltransferase B in the presence of poly(U) or poly(dT). The reaction conditions were the same as those described under EXPERIMENTAL PRO- CEDURE, except that varying amounts of P. putida ribosomal RNA were present in each reaction mixture. AMP incorporated in the complete reaction mixture (0), in the complete reaction mixture plus 16 IIM(P) poly(dT) (I), or in the complete reaction mixture plus 39 HM (P) poly(U) (A). Velocity is in nmol AMP incorporated per 10 min; and substrate is in mM (P) P. putida ribosomal RNA. WV 30 20 IO ll [rRNA] 1 I20 74 .HIO mucothzmmoE .CHEZLUEOEOU Eopm OoEHEEouoO mm: EoHumHquocoo HUM one .mmammuomm HHuom mEHumwompooEHIOE< mo xmmmm on» pom .Hm Hocmmv o>onm eonHuomoe mm pounHo was EEsHoo amoemnaem-H:O>Hom oEmm on» ow eaHHmam mm: :oHuumEm oExNEo m omoHsHHooozmmoam woumuucoocoo one mo oHQEmm H: mm m .zHEmHHEHm .mmsnmuomm HHuum wEHumwomHoucH Imb< mo Emmmm ecu pom .UouooHHou who: come HE O.O mo mcoHuomnm mm was :HE Hem HE N.O um OoaoHo>oO we: EEsHoo one .mExnco any ousHo op wow: mw3 :HEanm Enhom oEH>on HE hem w: OOH use Hocmcpoouamopee-~ 2E m .HH.E may oumcamosm EsHmmmuom 2E OH EH HUM 2 O.H ow H.O Eoww paonmum AmoEHH HE ON < .Eewmsn oEmm ecu mo HE m.~ :sz eonmmz mm: EEsHou ecu .EoHuomHH oExNEo any mo EoHumuHHmmm wEHonHom .H< Hocmav :HEanm Esuom oEH>on HE you w: OOH OEm HocmnuooummoHoEIN 2E m .Hux 2E OOH .HH.E may aeeaameea asammaeoa as oH EaHs eaueaeHHasee .mmanmuoca aHoa co m omoHsHHooogamocm wopmuuceocoo Ocm < omoHsHHooocmmoaa Oeumhuaouaou mEoHuumhm mo AsamhwoumEowao AuHchm<--.O epzmHm 75 IO .J.JI4J 1.111.. x B IIa\a\ .. / rLLlrL i L 8. 6. 4. 2. O B. 6. 4. 2. O . .58; 78; r . .58.. .78.; . . . 3 I a. 6. 4. 2. O O O 0 2065 2034181002. azd EF F LUENT VOLUME (ml) 76 Structural Properties of Adenylyltransferases A and B Determination of the Sedimentation Coefficient of Adenylyltransferases A and B.--Glycerol gradient centri- fugation was used to determine the sedimentation coeffi- cients of adenylyltransferases A and B. The results are presented in Figure 7. With horse blood hemoglobin @20,w = 4.2 S) as the reference protein, the sedimentation coeffi- cient5 of adenylyltransferase A was calculated to be 9.3 S and that for adenylyltransferase B 4.3 S. For typical globular proteins, sedimentation coefficients (320,w) of 9.3 S and 4.3 S correspond to molecular weights of 185,000 and 58,000, respectively (36). Determination of the Stoke's Radius for Adenylyl- transferase B Using Gel Filtration on Bio-Gel P-lOO.--Kd values for adenylyltransferase B and for the reference proteins alkaline phosphatase, hemoglobin, and cytochrome c were determined for the Bio-Gel P-lOO column that was used as the last step in the purification procedure (Figure 3). The inverse error function complements of the Kd values of the reference proteins were plotted against their Stokes' radii (37). Using this plot as the standard curve, a Stoke's radius of 2.9 nm for adenylyltransferase B was determined from its inverse error function complement of Kd. 5The enzyme concentration was much less than 0.1 mg per m1, so this is essentially a 320 w value; however, 0.01% bovine serum albumin was present. ’ 77 Figure 7.--Glycerol gradient centrifugation of adenylyl- transferases A and B. The protein solutions, 200 ul of either adenylyltransferase A (panel A) or adenylyltransferase B (panel B), were layered on separate 10 to 30% linear glycerol gradients (4.8 ml) prepared in 10 mM potassium phosphate (pH 7.1), 1M KCl, 0.5 mM dithiothreitol and 100 pg per m1 bovine serum albumin. After centri- fugation in a Spinco SW SOL rotor at 46,000 rpm for 21 hours at 4°, fractions (0.15 ml) were collected from the bottom of the centrifuge tube. The ATP-incorporating activity of adenylyltrans- ferase A (‘) and adenylyltransferase B (0) were assayed on 25 ul samples in the reaction mixture described under EXPERIMENTAL PROCEDURE, except that the specific radioactivity for [3H1ATP was 6.1 x 103 cpm per nmol. Hemoglobin (Hb) was used as a marker of known sedimentation coeffi- cient. 78 — 0.9 '- O.6 '- P a s O 0 :95: 29:51:82. as: 0.6 '- 0.3 - 30 20 FRACTION NUMBER l0 79 From a Stoke's radius (r) of 2.9 nm and a sedimenta- tion coefficient (3) of 4.3 S, a molecular weight (M.W.) for adenylyltransferase B of 52,000 can be calculated from the equation6 (38). Determination of the Polypeptide Composition of Adenylyltransferase B Using SDS Polyacrylamide Gel Electro- phoresis.--A sample of the phosphocellulose B enzyme fraction was subjected to SDS polyacrylamide gel electrophoresis (Figure 8) essentially as described by Johnson et al. (39), except that the concentration of acrylamide was 10% (w/v) and that of bis-acrylamide was 0.368% (w/v). In parallel experiments, P. putida DNA-dependent RNA polymerase, bovine serum albumin, pepsin and pancreatic ribonuclease A were analyzed electrophoretically. When compared to the relative migration of these polypeptides of known molecular weight, the four Commassie blue stained7 protein bands of fraction phosphocellulose B had molecular weights of 53,000, 47,000, 14,000 and 12,000 in the weight ratio of 1.2 : 1.1 : 1.0 6Also, where N is the Av0gadro number, the viscosity of the solvent (n) is assumed to be 0.01 poise, the partial specific volume of the protein (v) is assumed as 0.725 cm3 per g, and the value of the density of the solvent (p) is taken as 1.0 g per cm3. 7For an alternative staining procedure to that described (39), see Appendix. 80 .mmmv Ammmzmv <-H maze .ommoaosconfip owumohucmm Mow oom.m~ can .cfimmoa how ooo.nm .flon pow ooo.wo .emmpoExHom on .mmmhmE>Hom \2V afi.o ecu .AH.A may opmegmong esaeom 2 H.o :fi How you as w an muse: m.o pow omN um umEuomgoa mm: mammu05Q0Huuon .mpfiEmfixuom-mfln m>\zv wwom.o can owfiEmmxhom m>\3v woa mcficfimucou menuxfle mam-opfismaxpom cm Eopw umhmmohm flEo HHV How mp«Em~>uomz~om-mam m com: popmxma mm: flan oov menuxHE one .Hopooxaw A>\>V wm paw HocmcumOHQmupoE -N fl>\>v ”a .mam fl>\3v wH .AH.A :QV opaggmoga enaeom 2 H.o mcaeamueou :oHpSHom H: ooa m :fl GAE om pom com um woumnzocfl mm: am: vv cofiuumhw mexuco m omoHSHHooozamocm one .Ammv .Nu no comcsow kn umnfiHUmop mm zafiwwucommo woahompom mm: mflmohocmo~uuoao How ewMEmHAHomxHom-mam .m omofisfiaooocmmocm seapomhm we mammHOLQOHuuoHo How opfismfixpumxaom-mam--.w ohswfim 81 .53 2.25 20¢... mozflrma @ v q _ 9! O W“ 099 1' BONVBUOSBV v.0 82 2.0. A sample of the enzymatic activity peak from Bio-Gel P—lOO gel filtration of fraction phosphocellulose B in the absence of bovine serum albumin was subjected to the same SDS polyacrylamide gel electrOphoresis system. Only two polypeptides of molecular weights 53,000 and 47,000 were present. Electrophoresis was also performed on a sample of fraction phosphocellulose A in the same SDS polyacrylamide gel system (data not shown). Of the two possible polypep- tides representing adenylyltransferase activity, only the 53,000 molecular weight species was common to both fractions phosphocellulose A and B. DISCUSSION ATP:RNA adenylyltransferase was purified from P. putida by a procedure similar to that described earlier (16). However, the scale of the purification was increased over ten-fold, from 40 to 500 grams of cells, and another step, Bio-Gel P-lOO chromatography was added. In addition, improvements to the procedure were introduced to increase the recovery of enzymatic activity. PMSF, a protease inhibitor, at a concentration of 0.1 mM and 2-mercaptoethanol at a concentration of 5 mM were added to all buffers used in the purification. The addition of at least 0.25 M KCl to buffers subsequent to DEAE-Sephadex chromatography, and 50% glycerol in the final storage buffer stabilized enzyme activity. Through phosphocellulose chromatography the improved scheme resulted in an increased specific activity of nearly four times over that previously reported. With the inclusion of Bio-Gel P-100 chromatography the enzyme was estimated to be purified another 2.5-fold. As a result of the revised purification scheme, a previously undetected ATP-incorporating activity was dis- covered. It was evident in phosphocellulose chromatography 8 as a distinct peak of activity eluting separately from the previously reported adenylyltransferase (Figure 2). The 83 84 activity was indistinguishable by all kinetic criteria from the adenylyltransferase originally described by Payne and Boezi (16). Both of these ATP-specific, primer-dependent activities were Optimally active at pH 9.5, were inhibited by inorganic perphosphate but not by orthophosphate, and demonstrated the same specificity for primer. The divalent metal ion requirement for both activities was preferentially satisfied by Mg2+, rather than Mn2+, with an optimum at 20 mM. Within experimental error, both enzymes had identical values for the Km of ATP, the Ki for 3'-dATP and the Km for ribosomal RNA. When synthetic polynucleotides were added to ribosomal RNA—primed reactions, the effect on the AMP incorporation by either enzyme was the same. The Ki values determined for poly(U) and poly(dT) were equivalent within experimental error. The amount of rifamycin AF/Ol3 or AF/DNFI required to inhibit 50% the rate of AMP incorporation by either enzyme was identical. Neither activity was inhi- bited by rifampicin. This newly detected ATP-incorporating activity was designated adenylyltransferase A, while the previously reported activity was designated adenylyltrans- ferase B. It was not possible to distinguish between these two adenylyltransferases catalytically. However, an exami- nation of their chromatographic and sedimentation properties ‘ indicates that these enzymes have different structures. A difference in elution from anionic-exchange chromatography 85 provided the basis for the initial detection of the new adenylyltransferase. During chromatography on phosphocell- ulose (Figure 2), adenylyltransferase A eluted at 0.3 M KCl, while adenylyltransferase B eluted at 0.6 M KCl. When subjected to poly(U)-Sepharose chromatography, adenylyl- transferase A eluted at 0.2 M KCl and adenylyltransferase B eluted at 0.4 M KCl (Figure 6). Another difference observed for these enzymes was in their molecular weights. When chromatographed on Bio-Gel P-lOO, adenylyltransferase A was excluded while adenylyltransferase B was included in the gel (Figure 3). From its elution on Bio-Gel P-lOO relative to protein standards, adenylyltransferase B was estimated to have a molecular weight of 52,000. By sedimentation through glycerol gradients, adenylyltransferase A was determined to have a sedimentation coefficient (820,w) of 9.3 S and adenylyltransferase B of 4.3 8 (Figure 7). From these determinations the molecular weight of adenylyltrans- ferase A was estimated to be approximately 185,000 and that of adenylyltransferase B to be 50,000 to 60,000. During the purification of adenylyltransferase, a peak of enzymatic activity eluted at approximately 4 liters in Bio-Gel P-ZOO gel filtration (Figure 1). When compared to the elution volumes of protein standards, the elution of adenylyltransferase activity corresponded to a molecular ' weight of approximately 50,000 to 60,000. Pooling of this activity in the effluent from 3.6 to 4.3 liters during the 86 purification selected against material of molecular weight greater than about 65,000. If a sample of this pooled activity was subjected to glycerol gradient centrifugation under the conditions of Figure 7, there was only a single 4.3 3 peak of adenylyltransferase activity. Thus, only the activity corresponding to adenylyltransferase B was present at this point in the purification. Adenylyltrans- ferase A must then have arisen during the purification procedure, subsequent to Bio-Gel P-200 gel filtration but prior to phosphocellulose chromatography. The generation of adenylyltransferase A during the purification was not the result of a random aggregation. Several facts indicate that this enzyme has some specific structure. Adenylyltransferase A eluted as a sharp peak of activity in both phosphocellulose and poly(U)-Sepharose chromatographies. Glycerol gradient centrifugation of adenylyltransferase A revealed a single, well-defined peak of AMP incorporation. If adenylyltransferase A was first preincubated with ribonuclease, then subjected to the glycerol gradient centrifugation, there was no alteration in the sedimentation profile. This indicated that the specific structure of adenylyltransferase A did not include a bound RNA molecule. Since adenylyltransferases A and B are identical catalytically, it is probable that adenylyltransferase A contains as part of its structure adenylyltransferase B. 87 Although there is no direct evidence, the interpretation of the data available supports this conclusion. The specific structure of adenylyltransferase A could be generated from the 50,000 to 60,000 molecular weight material of Bio-Gel P-200 chromatography, adenylyltransferase B, by several mechanisms. Adenylyltransferase A could be the result of a simple self—association of adenylyltransferase B. It is not likely that adenylyltransferase A would be a dimer of adenylyltransferase B, since a typical globular protein of 100,000 to 120,000 has a sedimentation coefficient of 6-7 8. However, a trimer or a tetramer of adenylyltransferase B is possible since these would have 820,w values close to that of adenylyltranSferase A (9.3 S). Adenylyltransferase A could alternatively be the result of the specific association of a unique protein or proteins with adenylyltransferase B. When SDS polyacrylamide gel electrophoresis patterns of enzyme fractions phosphocell— ulose A and B were compared, the 53,000 molecular weight polypeptide corresponding to adenylyltransferase B was present in both patterns. In addition, several polypeptides of the molecular weight range 50,000 to 80,000 were found in A but not in B. These additional polypeptides are likely candidates for the specific binding to the catalytic poly— peptide, adenylyltransferase B, to form adenylyltransferase A. 88 From the limited amount of structural data on these adenylyltransferases, it is not possible to decide which of these mechanisms is applicable. However, it is clear that the governing process is not readily reversible. Since a simple, readily reversible association is an equilibrium process, once the species have been separated the equilibrium should re-establish. Once adenylyltransferases A and B were separated, however, no further appearance of adenylyltrans- ferase A was detected in fractions containing adenylyltrans- ferase B or vice versa. Poly A polymerase of E. 007i has been implicated by some (35,42) to be a subunit of DNA-dependent RNA polymerase. Sippel (14) compared purified "ATP-RNA adenyltransferase" of E. coli to RNA polymerase by electrophoresis in SDS polyacrylamide gels. None of the polypeptides of RNA polymerase had the same mobility as that for the "adenyl- transferase." Schafer et al. (43) demonstrated that even purified E. coli RNA-primed polynucleotidepyrophosphorylase, which can synthesize poly(A), was not a subunit of RNA polymerase. When adenylyltransferase B is compared to P. putida DNA—dependent RNA polymerase by SDS polyacrylamide gel electrophoresis (Figure 8), it is clear that these two enzymes also do not have any polypeptides in common. P. putida adenylyltransferase B is one of the most highly purified poly A polymerases from the particulate fraction of a prokaryote. Recently, Sippel (14) has 89 characterized a highly purified "adenyltransferase" from the particulate fraction of E. coli. These two particulate enzymes have some similar catalytic, chromatographic and structural ploperties. Both of the ATP-specific, primer- dependent adenylyltransferases have an unusually strong affinity for phosphocellulose. Elution of activity required the presence of greater than 0.5 M KCl. To maintain activity buffers used in the later steps of purification were high in ionic strength. Molecular weight analysis indicated that either activity was a single polypeptide of 50,000 to 60,000. Several catalytic differences, however, demonstrate that these enzymes are distinct from one another. The E. coli enzyme had a pH optimum of 8, had a preference for Mn2+ over Mg2+, was inhibited by inorganic orthophosphate, and was primed as well by Q8 RNA as it was by transfer RNA or ribosomal RNA. In addition, optimal enzymatic activity occurred in the presence of 0.35 M KCl. Although the data was not presented, the P. putida enzyme was optimally active at 20 mM KCl and completely inhibited by 200 mM KCl. The rifamycin family of antibiotics have been used as potent inhibitors of many nucleic acid enzymes. These semisynthetic derivatives of the natural fermentation product of Streptomyces mediterranei have been postulated to interact with the nucleoside triphosphate binding site of the template-requiring DNA and RNA polymerases (for a review of rifamycins, see Riva and Silvestri [34]). Only 90 a few of the derivatives have been examined for their effect on poly A polymerases. One derivative, rifampicin, was ineffective in inhibiting the ATP-incorporating activity of the prokaryotic (14,15,35) and eukaryotic enzymes (2,3, 5,6). On the other hand, AF/AP and AF/ABDMP were effective inhibitors of the HeLa enzyme (3); and AF/Ol3, at 100 pg per m1, inhibited the rat liver nuclear enzyme (2). The effects of fourteen derivatives of rifamycin, all effective inhibitors of P. putida DNA—dependent RNA polymerase, on the ATP-incorporating activity of adenylyltransferase was presented in this report. Of these derivatives only two, AF/013 and AF/DNFI, inhibited the AMP incorporation of P. putida ATP:RNA adenylyltransferase. AF/Ol3 is a strong inhibitor, as 50% inhibition is obtained at only 5 pg per ml. ATP:RNA adenylyltransferase, a non-template-requiring polymerase, can be added to the list of nucleic acid poly- merizing enzymes that are inhibited by AF/013 (32,33,44). The characteristics of the nucleoside triphosphate binding site of ATP:RNA adenylyltransferase were examined. It was demonstrated, as shown in Table II (and [16]), that this enzyme preferentially utilized ATP as substrate. The base specificity for Substrate was also demonstrated by the minimal amount of inhibition of AMP incorporation ob- served when GTP, CTP or UTP was present (16). The amount of inhibition is an indication of the binding efficiency of the three ribonucleoside S'-triphosphates relative to 91 ATP. Additional characteristics of substrate binding were obtained from studies on the inhibition of AMP incorporation by analogs of ATP. Three analogs, 2'-dATP, AmTP and e-ATP did not affect the rate of AMP incorporation. On the other hand, a fourth analog, 3'-dATP, was an effective inhibitor of ATP for the enzyme. The failure of the analogs altered at the 2'-position of the ribose moeity to compete with ATP is interpreted as a failure to bind to the enzyme. Thus, the hydroxyl at the 2'-position seems to be required of a potential substrate for binding. However, the inhi- bition of AMP incorporation and, therefore, competitive binding demonstrated by 3'-dATP indicates that for the 3'-position either a hydrogen or a hydroxyl substituent is satisfactory. The large variability in selectivity of substituents between the 2'- and the 3'-positions may be an expression of the evolutionary pressure on the enzyme to distinguish between the only two naturally-encountered isomers, ATP and 2'-dATP. The ribosomal RNA-primed reaction catalyzed by ATP:RNA adenylyltransferase was inhibited upon the addition of synthetic non-priming homopolynucleotides. A study of this inhibition has lead to a characterization of the primer binding site of the enzyme. The reaction was inhibited to varying extents depending upon the polynucleo- tide added. 'Poly(U), poly(dT) and poly(G) were the strongest, poly(C) and poly(I) were intermediate and poly(A), poly(Am) 92 poly(dA) and poly(Um) were ineffective as inhibitors. Since the kinetics of inhibition by poly(U) and poly(dT) were competitive with ribosomal RNA (Figure 5), these non-priming polynucleotides did not inhibit by irreversibly binding to the enzyme. One explanation for the inhibition is to presume that poly(U) or poly(dT) hydrogen-bonded to the adenylate residues of the partially synthesized product. The resulting complex then was not a primer for the reaction. This would not explain, however, why poly(G), poly(C) or poly(I) inhibited. In addition, this contradicts the fact that poly(Um), which would be expected to bind to the adeny- late residues at least as efficiently as poly(U), did not inhibit the reaction. Another explanation for the inhibition involves the competition of the synthetic polynucleotides with ribosomal RNA directly for the primer binding site on the enzyme. The variation in effectiveness of the synthetic homopolynucleotides for inhibition could be related to their differences in secondary structure. The primer binding site would then seem to prefer single-stranded polynucleo- tides with little secondary structure. However, binding alone is not sufficient criteria for a polynucleotide to prime AMP incorporation. Assuming the Km represents the dissociation constant, the fact that the Km for ribosomal RNA is ten times the Ki for poly(U) indicates that adenylyl- transferase binds poly(U) much more efficiently. However, 93 pMyUU does not prime the reaction. Thus, in addition to ngmncral property of binding nucleic acids, the enzyme Msorequires some 3'-terminal base specificity of the pdymmleotide for primer. The strong affinity demonstrated by adenylyltrans- feuwe for the non-priming poly(U) led to the possibility thatthe enzyme would bind to a column of immobilized poly(U). Whmiadenylyltransferase was subjected to poly(U)-Sepharose chnmmtography (Figure 6), the enzyme was bound. Upon znnflication of a salt gradient, adenylyltransferase activity eluted with nearly 100% recovery. Affinity chromatography of this enzyme on poly(U)-Sepharose could provide an alter- native method for the purification of adenylyltransferase to greater specific activities and higher yields. When enzyme fraction pH 5 supernatant was applied to the column, however, no appreciable amount of activity was retained. The failure of this fraction to bind to the column was probably due to the enzyme's strong association in impure fiwuztions, and at low salt concentrations, with macromolecu- lar‘tcomponents of the cell (16). Thus; for the P. putida enizywne this column would only be successful in later stages of the purification, but poly(U)-Sepharose chromatography might prove valuable in the purification of poly A poly— merase from other sources. The ATP:RNA adenylyltransferase of P. putida has been studied in some detail. Several kinetic and structural 94 characterizations of the enzyme were made. Further kinetic studies on the primer-binding site could determine whether the enzyme also requires a specific 3'-terminal sequence, or if only the 3'-terminal residue is important. Additional investigation on the structure of adenylyltransferase A could lead to deducing the proper mechanism for its for- mation. This may also provide for an understanding of whether the formation of such a protein complex with the catalytic polypeptide, adenylyltransferase B, is of 1 biological significance or merely an artifact of the . purification procedure. REFERENCES 1. Winters, M. A., and Edmonds, M. (1973) J. Biol. Chem. 248, 4756-4762. 2. Niessing, J., and Sekeris, C. E. (1973) Nature New Biol. 243, 9-12. 3. Mans, R. J. (1973) Fed. Eur. Biochem. Soc. Lett. 33, 245—246. 4. Tsiapalis, C. M.; Dorson, J. W.; De Sante, D. M.; and Bollum, F. J. (1973) Biochem. Biophys. Res. Commun. 50, 737-743. 5. Coleman, M. S.; Hutton, J. J.; and Bollum, F. J. (1974) Nature 248, 407-409. 6. Jacob, S. T., and Schindler, D. G. (1972) Biochem. Biophys. Res. Commun. 48, 126-134. 7. Burkard, G., and Keller, E. B. (1974) Proc. Nat. Acad. Sci. U.S.A. 7l, 389-393. 8. Venkataraman, P. R., and Mahler, H. R. (1963) J. Biol. Chem. 238, 1058-1067. 9. Haff, L. A., and Keller, E. B. (1973) Biochem. Biophys. Res. Commun. 5l, 704-710. 10. Twu, J. S., and Bretthauer, R. K. (1971) Biochemistry l0, 1576—1582. 11. Hyatt, E. A. (1967) Biochim. Biophys. Aeta Z42, 246- 253. 12. Hyatt, E. A. (1967) Biochim. Biophys. Acta Z42, 254- 262. 13. August, J. T.; Ortiz, P. J.; and Hurwitz, J. (1962) J. Biol. Chem. 237, 3786-3793. 14. Sippel, A. E. (1973) Eur. J. Biochem. 37, 31-40. 15. Modak, M. J., and Srinivasan, P. R. (1973) J. Biol. Chem. 248, 6904-6910. 95 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 96 Payne, K. J., and Boezi, J. A. (1970) J. Biol. Chem. 245, 1378-1387. Moss, B.; Rosenblum, E. N.; and Paoletti, E. (1973) Nature New Biol. 245, 59-63. Stoltzfus, C. M.; Morgan, M.; Banerjee, A. K.; and Shatkin, A. J. (1974) J. Virol. l3, 1338-1345. Darnell, J. E.; Jelinek, W. R.; and Molloy, G. R. (1973) Science l8l, 1215-1221. r Brawerman, G. (1974) Annu. Rev. Biochem. 43, 621-642. Hirsch, M., and Penman, S. (1973) J. Mol. Biol. 80, 379-391. Hirsch, M., and Penman, S. (1974) J. Mol. Biol. 83, l3l-l42. L, Edmonds, M., and Kopp, D. W. (1970) Biochem. Biophys. Res. Commun. 4l, 1531-1537. Perry, R. P.; Kelley, D. E.; and LaTorre, J. (1972) Biochem. Biophys. Res. Commun. 48, 1593-1600. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; and Randall, R. J. (1951) J. Biol. Chem. Z93, 265-275. Bollum, F. J. (1966) in Procedures in Nucleic Acid Research. Edited by S. L. Cantoni and D. R. Davies. New York: Harper and Row, p. 579. Ayad, S. R., and Blamire, J. (1972) Techniques of Nucleic Acid Fractionation. New York: John Wiley and Sons, p. 42. Ts'o, P. P.; Helmkamp, G. K.; and Sander, C. (1962) Biochim. Biophys. Acta 55, 584-600. Rottman, F.; Friderici, K.; Comstock, P.; and Khan, M. K. (1974) Biochemistry l3, 2762-2771. Poonian, M. S.; Schlabach, A. J.; and Weissbach, A. (1971) Biochemistry Z0, 424-427. Failla, D., and Santi, D. V. (1973) Anal. Biochem. 52, 363-368. Chamberlin, M. J., and Ring, J. (1972) Biochem. Biophys. Res. Commun. 49, 1129-1136. 33. 34. 36. 37. 38. 39. 40. 41. 42. 43. 44. 97 KUpper, H.; McAllister, W.; and Bautz, E. K. F. (1973) Eur. J. Biochem. 38, 581-586. Riva, S., and Silvestri, L. G. (1972) Annu. Rev. Microbiol. 26, 199-224. Ohasa, S., and Tsugita, A. (1972) Nature New Biol. 240, 35-38. Holleman, W. (1966) Ph.D. thesis, Michigan State University, E. Lansing, Michigan. Ackers, G. K. (1967) J. Biol. Chem. 242, 3237-3238. McGuinness, E. T. (1973) J. Chem. Educ. 50, 826-830. Johnson, J. C.; DeBacker, M.; and Boezi, J. A. (1971) J. Biol. Chem. 246, 1222-1232. Jaenicke, R., and Knof, S. (1968) Eur. J. Biochem. 4, 157-163. Garen, A., and Levinthal, C. (1960) Biochim. Biophys. Acta 38, 470-483. Terzi, M.; Cascino, A.; and Urbani, C. (1970) Nature 226, 1052-1054. Schfifer, R.; Zillig, W.; and Priess, H. (1972) Fed. Eur. Biochem. Soc. Lett. 25, 87-90. Gerard, G. F.; Gurgo, C.; Grandgenett, D. F.; and Green, M. (1973) Biochem. Biophys. Res. Commun. 53, 194-201. 5' “‘57:. l ' nu. SECTION III THE EFFECT OF 3'-DEOXYADENOSINE 5'-TRIPHOSPHATE ON THE REACTION CATALYZED BY ATP:RNA ADENYLYLTRANSFERASE FROM NOVIKOFF HEPATOMA 98 ABSTRACT A Mn2+-dependent ATP:RNA adenylyltransferase was purified from the cytoplasm of Novikoff hepatoma (strain Nl-Sl). The activity was primer-dependent, had a pH optimum of 8, and was sensitive to the sulfhydryl reagent, p-hydroxymercuribenzoate. This enzyme had a sedimentation coefficient (S20,w) of 4.1 S. The preferred primers were polyadenylic acid and yeast transfer RNA. Rifamycin SV AF/013 inhibited the enzyme; 50% of the activity remained at 20 Hg per ml. The drug cordycepin (3'-deoxyadenosine) did not affect adenylyltransferase activity. However, the triphosphate derivative of the drug, 3'-deoxyadenosine S'-triphosphate, was a competitive inhibitor with ATP for the reaction. The Km for ATP and the Ki for 3'-deoxy- adenosine S'-triphosphate were 2 x 10'5 M. 99 INTRODUCTION ATP:RNA adenylyltransferase (EC 2.7.7.19), or as it is more commonly known by the trivial name, poly A (Adenylate) polymerase, was first isolated nearly fifteen years ago from calf thymus nuclei (1). Since then there have been numerous reports of the activity in prokaryotes (2-4), eukaryotes (5-10), and viral cores (11,12). Recently, the discovery in eukaryotes of 3'-terminal polyadenylation of messenger RNA (mRNA) and of its precursor, heterogeneous DNA-like nuclear RNA (HnRNA), has lead to the postulation that enzymes of this type catalyze the posttranscriptional synthesis of those polyadenylate (poly[A])segments (for a review, see Darnell et al. [13] and Brawerman [14]). Poly- adenylation of these DNA transcripts is believed to be im- portant for the effective processing and the functionality of mRNA. The in vivo polyadenylation of these RNAs is sensitive to the drug cordycepin. If the eukaryotic adenylyltransferase is responsible for this polyadenylation, then it should also be sensitive to the drug. The AMP incorporation by the adenylyltransferase purified from calf thymus nuclei (5) or from rat liver nuclei (6), how- ever, was not affected by the presence of cordycepin. The triphosphate derivative of cordycepin (3'-dATP) is probably 100 101 the form in which cordycepin would inhibit adenylyltrans- ferase activity. The effect that 3'-dATP has on adenylyl- transferase activity has not been studied previously. This report describes some of the basic properties of the ATP:RNA adenylyltransferase from Novikoff hepatoma cells. In addition, there is presented the first evidence that a eukaryotic adenylyltransferase is effectively inhibited by 3'-dATP. In agreement with previous results, cordycepin itself was found ineffective in inhibiting the in vitro synthesis of poly(A). EXPERIMENTAL PROCEDURE Materials The homopolyribonucleotides, p-hydroxymercuribenzoate and calf thymus DNA were from Sigma Chemical Co. Poly(dT)1 and Q6 RNA were purchased from Miles Laboratories, Inc. Dithiothreitol, ADP, 2'-dATP and all unlabeled 5'-triphosphate f .I;'L":‘I 'HLILJ ; '3:- -_-I1 1‘- PA.- q derivatives of the ribonucleotides were obtained from P-L Biochemicals, Inc. 3H-labeled nucleotides were purchased from Schwarz/Mann. Whatman DEAE-cellulose (DESZ) and cellulose phosphate (P-l) were from H. Reeve Angel, Inc. Nitrocellulose membrane filters (type B-6) were obtained from Schleicher and Schuell, Inc. Nogalomycin, actinomycin, and a-amanitin were from Upjohn, Merck, and Henley and Co., respectively. Rifamycin derivatives were the gifts of Dr. Luigi Silvestri, Gruppo Lepetit, Inc., Milan, Italy. Yeast tRNA, 3'-dA, and 3'-0-methylATP were the gifts of Dr. Fritz Rottman of this department. Chromatographically pure 3'-dATP was a gift of Ron Desrosiers also of this department. Pseudomonas putida ribosomal RNA was prepared according to a procedure previously described (3). QB RNA was purified by sedimentation through a sucrose gradient before use. Calf thymus DNA was activated by a previously described method (15). 1The abbreviations used are listed in the Journal of Biological Chemistry 249, 1 (1974). 102 103 Analytical Methods Protein concentration was determined by the method of Lowry et al. (16) with bovine serum albumin as the standard. The concentrations of calf thymus DNA and Q8 RNA were deter- mined spectrophotometrically based on the extinction coeffi- 1% 260 mine the concentration of ribosomal and transfer RNA at 260 nm were 7.4 x 103 (17) and 7.2 x 103 cient E = 200. The molar extinctions, e(P), used to deter- (18), respectively. The molar extinctions, c(P), for the homopolynucleotides were: 10.5 x 103 at 257 nm, 9.2 x 103 at 260 nm and 6.5 x 103 at 267 nm for poly(A), poly(U), and poly(C), respectively (19) and 8.1 x.1o3 at 260 nm for poly(dT)_(17). IThe molar extinction uSed to determine spectrophotometrically the concentrations of 3'-dA, 3'-dATP and 3'-0-methylATP at 259 nm was 1.5 x 104. Assay_of ATP:RNA Adenylyltransferase The standard assay measured the conversion of [3H] ATP into acid-insoluble form. Unless otherwise indicated, the complete reaction mixture (0.2 ml) contained 100 mM Tris-HCI (pH 8), 1 mM MnCl 1 mM dithiothreitol, 100 pg 2’ per ml bovine serum albumin, 0.1 mM [3H1ATP (100 cpm per pmol), a saturating amount of poly(A) and adenylyltrans- ferase. After incubation for 30 min at 37°, the reaction was terminated by addition of 2 m1 cold 10% trichloroacetic acid -1% sodium pyrophosphate. Bovine serum albumin, 50 pl of a 5 mg per ml solution, was then added as carrier. After 5 min at 0°, the mixture was centrifuged at 10,000 x g for 2 min. The supernatant solution was removed and discarded. 104 The pellet was dissolved in 0.35 ml of 0.5 N NaOH. Following the addition of cold 10% trichloroacetic acid -1%sodium pyrophosphate and incubation for 5 min at 0°, the acid-insoluble material was collected on a membrane filter. The filter was then washed and dried. The radioactivity in the filter was determined by counting in 5 ml of a scintillation fluid containing 4 g of 2,5-bisl2-(5-tert- butylbenzoxazolyl)] thiophene per liter of toluene. One unit of enzymatic activity is defined as one pmol of AMP incorporated into acid-insoluble form per 30 min. Specific activity is expressed as units per mg protein. The amounts of enzyme activity which were added to the reaction mixtures were within the activity range which gave direct proportionality with the rate of AMP incorpo- ration for 30 min. Partial Purification of ATP:RNA Adenylyltransferase The Nl-Sl strain of Novikoff hepatoma cells were grown in Swimm's S-77 media supplemented with 4 mM glutamine and 10% calf serum (20). The cells were grown in an atmos- phere of 5% CO2 in air in sealed screw cap Erlenmeyer flasks or spinner flasks with a doubling time of approximately 12 hours at 37°. Cells were harvested by centrifugation when 5 growth had reached 8 x 10 cells per ml (mid-logrithmic phase). All subsequent procedures were performed at 0-4°. For a typical purification about 1 x 109 cells were used. The cells were washed by suspension in Basal Salt Solution 105 (21), followed by low speed centrifugation (1000 x g for 2 min). The cells were swollen by suspension in 15 ml of 10 mM Tris-HCl (pH 7.6), 6 mM KCl, and 5 mM MgCl2 for 5 min, then lysed by bringing the suspension to 0.5% (v/v) NP-40. The supernatant fraction of a low speed centri- fugation of the lysed cells was centrifuged at 80,000 x g for 60 min. The resulting supernatant fraction was dialyzed against 10 mM Tris-HCl (pH 8) and 5 mM Z-mercaptoethanol (Cytoplasmic Extract). The cytoplasmic extract enzyme fraction was passed through a 4 x 14 cm column of DEAR-cellulose equilibrated with 10 mM Tris-HC1 (pH 8) and 5 mM Z-mercaptoethanol. The enzymatic activity absorbed to the column was eluted by the application of column buffer containing 150 mM KCl. The peak activity fractions were pooled, then dialyzed against 10 mM potassium phosphate (pH 7.1) and 5 mM 2- mercaptoethanol (DEAE-cellulose). The DEAE-cellulose enzyme fraction was absorbed onto a 0.9 x 7 cm column of phosphocellulose equilibrated with 10 mM potassium phosphate (pH 7.1) and 5 mM Z-mercaptoethanol. A linear gradient of 0 to 1 M KCl in column buffer applied to the phosphocellulose column produced a single peak of enzymatic activity, eluting at approximately 0.4 M KCl. Fractions containing peak enzymatic activity were pooled (Adenylyltransferase), and used as a source of enzyme for all assays reported here. Adenylyltransferase had a 106 .specific activity of l x 105 units per mg protein, repre— senting a lS-fold purification and a 60% recovery from the cytOplasmic extract fraction. Enzymatic activity was stable for at least one month when stored at 4°. RESULTS AND DISCUSSION Properties The characteristics of the reaction catalyzed by the soluble enzyme from Novikoff hepatoma purified through phosphocellulose chromatography are presented in Table I. The incorporation of AMP into acid-insoluble form was dependent upon added enzyme, an exogenous RNA primer and 2+ 2+ 2+ n . Mg could not substitute for Mn at either 2.5 mM M or 10 mM. The activity was sensitive to the sulfhydryl reagent, p-hydroxymercuribenzoate. Even though dithiothreitol or bovine serum albumin were not required, their presence stabilized optimal activity. In separate experiments it ‘ was determined that optimal activity was achieved at pH 8 and at 0.5 mM MnClZ. The specificity of the enzyme for ATP as substrate is displayed in Table II. The rate of reaction was greatest when ATP was present. UTP, CTP, or GTP were not utilized as substrates by adenylyltransferase. The small amount of incorporation when 2'-dATP or ADP was present might possibly be a reflection of the impurity of this preparation. The primer specificity of adenylyltransferase was investigated (Table III). 0f the synthetic homopolyribo- nucleotides tested at saturating concentrations, poly(A) was the most effective primer of AMP incorporation. Yeast 107 108 TABLE I.--Characteristics of the Reaction Catalyzed by Adenylyltransferase?” Components of the Relative Incorporation Reaction.Mixture ‘ of AMP Complete 100 (230 :_1) minus enzyme 0 minus poly(A) 0 minus Mhz+ 0 minus lunz”, plus 2.5 mM or 10 mM Mg2+ 0 minus bovine serum albumin 82 minus dithiothreitol 84 minus dithiothreitol, plus 1 mM p-hydroxymercuribenzoate 4 *The complete reaction.mdxture (0.2 m1) contained 100 mM'Tris- HCl (pH 8), 1 mMNMhClz, 1 mM dithiothreitol, 100 pg per ml bovine serum albumin, 0.1 mM’I3HlATP (100 cpm.per pmol), a saturating amount of poly(A) and an apprOpriate amount of enzyme. The pmol of [3H]AMP incorporated into acid-insoluble product after 30 min of incubation at 37° is given in parentheses. 109 * TABLE II.--Nucleotide Specificity for Adenylyltransferase. F1 g! “’ “ ===:: l Relative Incorporation 2 Nucleotide of Mononucleotides ~ ATP 100 (129 :_l) Z'wMJP 19 m ADP 12 CTP 4 GTP 0 UTP 0 * The complete reaction mixture (0.2 ml) contained 100 mM Tris- HCl (pH 8), l mM’MnClz, 1 mM dithiothreitol, 100 ug per ml bovine serum albumin, a saturating amount of poly(A), an appropriate amount of enzyme and 0.1 mM labelled nucleoside di- or triphosphate (100 cpm per pmol) as indicated. The pmol of [3Hlnucleotide monophosphate incorporated into acid-insoluble product after 30 min of incubation at 37° is given in parentheses. 110 * TABLE IlI.--Primer Specificity for Adenylyltransferase. Relative Incorporation Primer of AMP Poly(A) 100 (237 :_l) Poly(C) 21 Poly(U) 3 Poly(dT) 2 S. cerevisiae transfer RNA 79 P. putida ribosomal RNA 27 Q8 RNA 12 Activated calf thymus DNA 0 None 0 * Reaction mixtures (0.2 ml) contained 100 mM Tris-HCl (pH 8), 1 mM MnCl , l mM’dithiothreitol, 100 pg per ml bovine serum albumin, 0.1 mM [3HIATP (100 cpm.per pmol), an appropriate amount of enzyme and.a saturating amount of primer. The pmol of [ HIAMP incorporated into acid-insoluble product after 30 min of incubation at 37° is given in parentheses. lll transfer RNA provided the greatest priming activity of the natural RNAs. Neither polydeoxynucleotide, poly(dT) nor activated calf thymus DNA, functioned as a primer for the reaction. Adenylyltransferase was subjected to centrifugation in 5-20% sucrose gradients containing 50 mM Tris-HCI (pH 7.5), 250 mM KCl and 5 mM Z-mercaptoethanol (data not shown). _When compared to the relative migration of E. coli alkaline phosphatase @20,w = 6.3 S), the single, sharp peak of adenylyltransferase activity was determined to have a 2 of 4.1 s. sedimentation coefficient (320,w) In summary, the properties of this adenylyltrans- ferase are similar to those reported for the Mn2+-dependent enzyme from other eukaryotic sources (8-10). For example, the adenylyltransferase from the cytoplasm of calf thymus had a pH optimum of 8.3, was effectively primed by poly(A) and had a sedimentation coefficient (320 w) of 3.5 S (8). Inhibitors Several rifamycin derivatives, actinomycin, nogalomycin and a-amanitin were tested for their effect on the AMP incorporation by adenylyltransferase. As shown in Table IV, the presence of nogalomycin, actinomycin or a-amanitin had little or no effect on the rate of AMP incorporation. Of 2The enzyme concentration was much less than 0.1 mg per m1, so this is essentially a 330 w value. ’ 112 TABLE IV.--The Effect of Inhibitor Cgmpounds on Adenylyl- transferase. - ”~__...__—-r -...— ..__-_ .-_.~_ ...-_ . ,- ...__. .._._ _m... .____-._...._.—---.._._- _‘.—-~—. .m- - . _-——- .. .-..-.., ... Inhibitor Added to the Relative Incorporation Reaction Mixture of AMP Experiment I NOne 100 (93 1.1) Rifamycin AF/AP 122 Rifampicin 122 Rifamycin PR/l9 100 Rifamycin AF/BO 67 Rifamycin AF/DNFI 22 Rifamycin AF/013 . ll Experiment II None 100 (119 :_1) Nogalomycin 92 Actinomycin 125 a-amanitin ll7 * Reaction mixtures (0.2 md) contained 100 mM Tris-”Cl (pH 8), l mM’MnCl , 1 mM dithiothreitol, 100 pg per ml bovine serum albumin, 0.11nM [3 ]ATP (100 cpm per pmol), a saturating amount of poly(A) and an appropriate amount of enzyme. In addition, in Experiment I the rifamycin derivatives were added to reaction mixtures to a final con- centration of 50 pg per m1. In Experiment II nogalomycin, actinomycin or a-amanitin were added to reaction.mixtures to a final concentration of 50, 50, or 20 pg per ml. The pmol of [3H1AMP incorporated into acid- insoluble product after 30 min of incubation at 37° is given in parentheses. 113 the rifamycin derivatives tested, only AF/DNFI and AF/013 were effective inhibitors of adenylyltransferase. In a separate experiment, the concentration of AF/Ol3 that gave 50% inhibition was determined to be 20 pg per ml. The adenylyltransferase from Novikoff hepatoma, a non-template- requiring polymerase, can be added to the list of nucleic acid polymerizing enzymes that are inhibited by AF/013 (22). Substrate Analogs Several structural analogs of ATP were examined for their effect on the ATP-incorporating activity of adenylyltransferase. The rate of AMP incorporation was measured at various concentrations of ATP in the absence and in the presence of each analog. The data was analyzed using Lineweaver-Burk double reciprocal plots (Figure 1). At a concentration of 1 mM, 3'-dA (cordycepin) had no effect on the rate of AMP incorporation. However, 3'-dATP and 3'-0-methylATP were equally effective as competitive inhi- 3 for ATP as well as the Ki for 3'- dATP or 3'-0-methylATP was 2 1 l x 10.5 M. bitors of ATP. The Km Practically all eukaryotic mRNA (except histone mRNA) isolated contains a 3'-termina1 poly(A) sequence. As a part of the current model for eukaryotic mRNA biogenesis, the poly(A) sequences are proposed to be posttranscriptionally 3A linear least squares analysis of the data provided the source for the reported values of Km or Ki. The variances following the values represent multiple determinations. 114 Figure l.--The effect of ATP structural analogs on AMP incorporation by adenylyltransferase. The complete reaction mixture was that described in the EXPERIMENTAL PROCEDURE, except that varying amounts of [3H1ATP (100 cpm per pmol) were present. AMP incorporated in the complete reaction mixture ( ), in the complete reaction mixture plus 8.8 x 10'5 M 3'-dATP ( ), and in the complete reaction mixture plus 5 x 10’4 M 3'-0-methy1 ATP ( ). The addition of 1 x 10-3 M 3'-dA to the complete reaction mixture did not alter the rate of AMP incorporation. Velocity is in pmol AMP incorporated per 30 min; and substrate is in mM ATP. __.—___-"—__ _ M-.—..—-_———n— 115 '60 IZO WV 40 116 synthesized on the 3'-hydroxyl end of the heterogeneous nuclear RNA (HnRNA) (13). The presence of these sequences on the DNA transcripts is associated with the effective processing of mRNA from HnRNA and for the appearance of mRNA in the cytoplasm. These functions for the poly(A) sequence were derived from observations of the in vivo effects of the drug cordycepin (3'-dA). In cells treated with 3'-dA, HnRNA synthesis continued, but little mRNA appeared with polyribosomes. Also, the poly(A) sequences of HnRNA and mRNA isolated from these cells were signifi- cantly shorter. Similar results have been observed for Novikoff hepatoma cells.4 Since poly A polymerase appeared as a likely candi- date for catalyzing the posttranscriptional synthesis of poly(A), 3'-dA was examined for its effect on this enzyme. When in vitro reactions catalyzed by the enzyme from the nuclei of calf thymus (S) or of rat liver (6) included 3'-dA, no effect on the rate of AMP incorporation was detected. This is reasonable since the triphosphate derivative (3'-dATP) is probably the form in which cordycepin would be active against adenylyltransferase. This report provides the first in vitro evidence that 3'-dATP does inhibit the poly(A) synthesis catalyzed by a eukaryotic polyIXpolymerase. It also confirms the previous observations 4 . . . .. Il.Desr051ers, personal communication. 117 that 3'-dA does not effectively inhibit the in vitro synthe- sis of poly(A). 10. 11. 12. 13. 14. 15. REFERENCES Edmonds, M., and Abrams, R. (1960) J. Biol. Chem. 235, 1142-1149. August, J. T.; Ortiz, P. J.; and Hurwitz, J. (1962) J. Biol. Chem. 237, 3786-3793. Payne, K. J., and Boezi, J. A. (1970) J. Biol. Chem. 245, 1378-1387. Sippel, A. E. (1973) Eur. J. Biochem. 37, 31-40. Winters, M. A., and Edmonds, M. (1973) J. Biol. Chem. 248, 4756-4762. Niessing, J., and Sekeris, C. E. (1973) Nature New Biol. 243, 9-12. Burkard, G., and Keller, E. B. (1974) Proc. Nat. Acad. Sci. U.S.A. 7l, 389-393. Tsiapalis, C. M.; Dorson, J. W.; DeSanta, D. M.; and Bollum, F. J. (1973) Biochem. Biophys. Res. Commun. 50, 737-743. Giron, M. L., and Huppert, J. (1972) Biochim. Biophys. Acta 287, 438-447. Walter, T. J., and Mans, R. J. (1970) Biochim. Biophys. Acta 2Z7, 72-82. Moss, B.; Rosenblum, E. N.; and Paoletti, E. (1973) Nature New Biol. 245, 59-62. Stoltzfus, C. M.; Morgan, M.; Banerjee, A. K.; and Shatkin, A. J. (1974) J. Virol. l3, 1338-1345. Darnell, J. E.; Jelinek, W. R.; and Molloy, G. R. (1973) Science l8l, 1215-1221. Brawerman, G. (1974) Annu. Rev. Biochem. 43, 621-642. Schlabach, A.; Fridlender, B.; Bolden, A.; and Weissbach, A. (1971) Biochem. Biophys. Res. Commun. 44, 879-885. 118 16. 17. 18. 19. 20. 21. 22. 119 Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; and Randall, R. J. (1951) J. Biol. Chem. l93, 1593-1600. Bollum, F. J. (1966) inProcedures in Nucleic Acid Research. Edited by S. L. Cantoni and D. R. Davies. New York: Harper and Row, p. 579. Ayad, S. R., and Blamire, J. (1972) Techniques of Nucleic Acid Fractionation. New York: John Wiley and Sons, p. 42. Ts'o, P. P.; Helmkamp, G. K.; and Sander, C. (1962) Biochim. Biophys. Acta 55, 584-600. Morse, P. A., Jr., and Potter, V. R. (1965) Cancer Res. 25, 499-508. Plagemann, P. G. W., and Roth, M. F. (1969) Biochem. 8, 4782-4789. Gerard, G. F.; Gurgo, C.; Grandgenett, D. P.; and Green, M. (1973) Biochem. Biophys. Res. Commun. 53, 194-201. APPENDIX 120 APPENDIX A RAPID, SIMPLE STAINING TECHNIQUE FOR PROTEINS SEPARATED IN POLYACRYLAMIDE GEL ELECTROPHORESIS Introduction The electrophoresis of proteins in polyacrylamide gels is a quick, simple procedure for the separation of proteins. Sophisticated analytical measurements, for example, the determination of molecular weight (1), can be obtained relatively easily. One drawback to this tech- nique has been the tedious routine to locate separated proteins. By a conventional staining technique employing Coomassie Brilliant Blue R250 in trichloroacetic acid followed by various destaining solutions and procedures (2), a minimum of 20 to 24 hours was required for protein identi- fication. For experiments requiring optimal sensitivity (about 0.1 pg protein) with low, uniform backgrounds such a lengthy procedure has been acceptable. Decreased time for protein band visualization has been obtained by electro- phoretically destaining, but some band mobility has also occurred. However, for many situations such delays, extreme sensitivity or involved techniques were undesirable. This is the case, for example, with the use of preparative gel 121 122 electrophoresis as part of a purification procedure, or in the analysis of large numbers of samples. A new staining procedure has been develOped to achieve the goals of simplicity and quickness without significantly sacrificing sensitivity or background uni- formity. The staining procedure originated from ideas presented in articles from the laboratories of Diezel et al. (3) and Malik and Berrie (4). The first report (3) intro- duced the methyl substituted triphenylmethane dye, Coomassie Brilliant Blue 6250. This dye was less soluble in 12.5% trichloroacetic acid than the conventionally used parent dye, Coomassie Brilliant Blue R250. This fact meant the reduction of dye penetration into non-protein areas of the polyacrylamide gels, requiring less destaining and thus providing shorter times for protein band visualization. As a consequence, interior portions of the gel containing protein were not stained, lowering the protein-stained band intensity and, thereby, stain sensitivity. The second paper (4) described a method using a stain containing sulfuric acid, trichloroacetic acid and Coomassie Brilliant Blue R250. The protein-stained color produced was usually not of sufficient contrast compared to the background for satisfactory protein detection or sensitivity. To reduce backgrounds destaining was required, but destaining also caused the concomitant reduction in intensity of protein- stained areas. The integration of these two procedures has 123 produced a staining technique which is rapid (as quick as 30 min), requires little or no destaining, has a sensitivity of at least 1 pg of protein per band, and is simple to perform. Stain Preparation Coomassie Brilliant Blue GZSO (xylene brilliant cyanin G) was purchased from K and K Laboratories, Inc. To a 0.2% (w/v) aqueous solution of the dye was added an equal volume of 2 N (5.6% [v/v]) H250 After the solution 4. was mixed well, it was allowed to stand for at least three hours. The precipitated solution was filtered through Whatman No. l by gravity flow. To the clear brown filtrate was added one-nineth volume of 10 N KOH. Trichloroacetic acid was added to the resulting dark purple solution to a final concentration of 12% (w/v). The resulting clear, light blue solution was ready for use. The stain can be stored for several months without loss of effectiveness. Upon storage the solution color will change towards-grey, but no change in the resulting protein-stained color or intensity is Observable. The stain may be reused several times, but the pH of the solu- tion must be maintained below 1.0. Comments After electrophoresis, the polyacrylamide gel can be added directly to the staining solution; for example, 124 15 m1 of stain is used in a test tube for a 0.7 x 10 cm cylindrical gel. Within 30 minutes protein bands of 5 to 10 pg are evident, while maximum development is achieved in 5 to 8 hours. Gels can be stored in the staining solution without overstaining or increasing the background level. When stored in water, there is a marked color intensification and background reduction producing optimal sensitivity. A rapid loss of all stain occurs if the gels are placed in solutions containing sulfuric acid, trich- loroacetic acid, or acetic acid. The stained gels can be scanned densimetrically at 640 nm. A linear relationship was observed between absorbance and protein concentration in the concentration range of 1 pg to at least 10 pg per band. When a tracking dye, for example, pyronin B, is used, the dye does not become obscured by the stain, permitting direct measurements of relative mobilities. This staining procedure can be used with SDS as well as non-dissociating polyacrylamide gels. The stain does contain a sufficient concentration of potassium ions to precipitate SDS in the gel, causing a clouded background. However, the resulting background will not obscure the visual detection of most protein-stained bands. When clear back- grounds are desired, especially for densimetric scanning, gels containing SDS should be rinsed with 10% (w/v) trichloroacetic acid, 33% (v/v) methanol, or water for up to 48 hours prior to stain application. If electrophoresis 125 is performed in the presence of less than 0.1% (w/v) SDS, then significant gel clouding is not encountered. Alter— natively, NaOH can be substituted for KOH in preparing the stain, reducing the amount of interference by SDS precipi- tation (5). However, this modification has not been met with complete success. This staining procedure has also been effectively used in other laboratories in this department. A part of the purification procedure of rat pancreas zymogen granule membrane proteins employed preparative slab gel electro- phoresis (5). The rapid identification of the proteins using this staining technique enabled the subsequent purification steps to follow without long delays. Simi- larly, the purification of rat brain hexokinase utilized gel electrophoresis of antibody-enzyme complex (6). This stain provided the quick identification of the protein for subsequent elution, reducing the duration of the purification by one to two days. In another case, the lipoproteins of insect hemolymph were subjected to poly- acrylamide gel electrophoresis as part of a study of the effect of filipin on insects (7). This protein stain was chosen for its fast development and clear, uniform back- grounds, which were well-suited for densimetric scanning. The quick visualization of the results also allowed for many more samples to be analyzed than by the conventional staining procedures. Finally, the simplicity and almost 126 infallibility of this staining technique made it practical for use in the teaching laboratory. The experiment in BCH 805 on gel electrophoresis of proteins was greatly simplified by use of this stain. The problems encountered by the students, that is, proper staining and destaining intervals and procedures, were eliminated by the substitution of the conventional technique with the procedure described here. REFERENCES Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244, 4406-4412. Johnson, J. C.; DeBacker, M.; and Boezi, J. A. (1971) J. Biol. Chem. 246, 1222-1232. Diezel, W.; Kopperschlager, G.; and Hofmann, E. (1972) Anal. Biochem. 48, 617-620. Malik, N., and Berrie, A. (1972) Anal. Biochem. 49, 173-176. MacDonald, R. J. (1974) Ph.D. thesis, Michigan State University, E. Lansing, Michigan. Felgner, P., personal communication. Schroeder, F. (1973) Ph.D. thesis, Michigan State Uni- versity, E. Lansing, Michigan. 127