METHODS OF BIOCHEMICAL ANALYSIS · METHODS OF BIOCHEMICAL ANALYSIS VOLUME VI CONTRIBUTORS ANITA J. ASPEN, Department of Biochemistry, Tufts University School of Medicine, Boston,

METHODS OF BIOCHEMICAL ANALYSIS

Edited by DAVID GLICK Professor of Physiological Chemistry University of Minnesota, Minneapolis

VOLUME VI

I N T E R S C I E N C E P U B L I S H E R S , I N C . , N E W Y O R K INTERSCIENCE PUBLISHERS LTD., LONDON

M E T H O D S O F B I O C H E M I C A L A N A L Y S I S Volume VI

Advisory Board:

S. BERGSTROM, University of Lund, Sweden A. M . BRUES, Argonne National Laboratory, Lemod, Illinois G. 0. BURR, Experiment Slation, Hawaiian Sugar Planters’ Associalion, Hono-

R. CONSDEN, The Canadian Red Cross Memorial Hospital, Taplow, Maiden-

A. B. HASTINGS, Haraard Medical School, Boston H. HOLTER, Carkberg Laboratory, Copenhagen, Denmark R. D. HOTCHKISS, The Rockefeller Iwtitule for Medical Research, New York J. K. N. JONES, Queen’s University, Kingston, Ontario, Canada C. G. KING, The Nutrition Foundation, New York H. A. LARDY, University of Wisconsin, Madison H. C. LICHSTEIN, Uniuersity of Minnesda, Minneapolis G. F. MARRIAN, University of Edinburgh, Scotland B. L. OSER, Food Research Laboratories, New York J . ROCHE, Collbge de France, Paris W. C. ROSE, University of Illinois, Urbana A. TISELIUS, University of Uppsala, Sweden D. D. VAN SLYKE, Brookhaven National Laboratory, Upton, Long Island,

lulu

head, Berkshire, England

New York

METHODS OF BIOCHEMICAL ANALYSIS

Edited by DAVID GLICK Professor of Physiological Chemistry University of Minnesota, Minneapolis

VOLUME VI

I N T E R S C I E N C E P U B L I S H E R S , I N C . , N E W Y O R K INTERSCIENCE PUBLISHERS LTD., LONDON

Copyright @ 1958 by Interscience Publishers, Inc.

Library of Congress Catalog Card Number 54-7232

VEB ARZNdMlT I t L W t ~ 1 ( UKESDEN - BlblIethok -

Interscience Publishers, Inc., 250 Fifth Avenue, New York 1, N. Y. For Great Brilain and Northern Ireland:

Interscience Publishers Ltd., 88/90 Chancery Lane, London, W. C. 2

PRINTED IN THE UNITED STATES OF AMERICA BY MACK PRINTINQ CO., EASTON. PA.

RIE1'IIODS OF BIOCHEMICAL ANALYSIS VOLUME V1

PREFACE TO THE SERIES

Annual review \-olumes dealing with many different fields of science have proved their value repeatedly and are now widely used and well established. These reviews have been concerned primarily with the results of the developing fields, rather than with the techniques and methods employed, and they have served to keep the ever expanding scene within the view of the investigator, the applier, the teacher, and the student.

It is particularly important that review services of this nature should now be extended to cover methods and techniques, because it is becoming increasingly difficult to keep abreast of the manifold experimental innovations and improvements which constitute the limiting factor in many cases for the growth of the experimental sciences. Concepts and vision of creative scientists far outrun that which can actually be attained in present practice. Therefore an emphasis on methodology and instrumentation is a fundamental need for material achievement to keep in sight of the advance of useful ideas.

The current volume is the first of a series which is designed to try to meet this need in the field of biochemical analysis. The topics to be included are chemical, physical, microbiological and, if necessary, animal assays, as well as basic techniques and instrumentation for the determination of enzymes, vitamins, hormones, lipids, carbohydrates, proteins and their products, minerals, antimetabolites, etc.

Certain chapters will deal with well established methods or techniques which have undergone sufficient improvement to merit re- capitulation, reappraisal, and new recommendations. Other chapters will be concerned with essentially new approaches which bear promise of great usefulness. Relatively few subjects can be included in any single volume, but as they accumulate these volumes should comprise a self-modernizing encyclopedia of methods of biochemical analysis. By judicious selection of topics it is planned that most subjects of current importance will receive treatment in these volumes.

V

vi PREFACE

The general plan followed in the organization of the individual chapters is a discussion of the background and previous work, a critical evaluation of the various approaches, and a presentation of the procedural details of the method or methods recommended by the author. The presentation of the experimental details is to be given in a manner that will furnish the laboratory worker with the complete information required to carry out the analyses.

Within this comprehensive scheme the reader may note that the treatments vary widely with respect to taste, style, and point of view. It is the editor’s policy to encourage individual expression in these presentations because it is stifling to originality and justifiably annoying to many authors to submerge themselves in a standard mold. Scientific writing need not be as dull and uniform as it too often is. In certain technical details a consistent pattern is followed for the sake of convenience, as in the form used for reference citations and indexing.

The success of the treatment of any topic will depend primarily on the experience, critical ability, and capacity to communicate of the author. Those invited to prepare the respective chapters are scientists who have either originated the methods they discuss or have had intimate personal experience with them.

It is the wish of the Advisory Board and the editor to make this series of volumes as useful as possible and to this end suggestions will always be welcome.

DAVID GLICK

Minneapolis, Minnesota

METHODS OF BIOCHEMICAL ANALYSIS VOLUME VI

CONTRIBUTORS

ANITA J. ASPEN, Department of Biochemistry, Tufts University School of Medicine, Boston, Massachusetts

FELIX BERGMANN, Department of Pharmucology, The Hebrew Uni- versity-Hadassah Medical School, Jerusalem, Israel

BERNARD B. BRODIE, Laboratory of Clinical Biochemistry, N a t i m l Heart Institute, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland

SHABTAY DIKSTEIN, Department of Pharmacology, The Hebrew Uni- versity-Hadassah Medical School, Jerusalem, Israel

WILHELM R. FRISELL, Department of Biochemistry, The University of Colorado School of Medicine, Denver, Colorado

SVEN GARDELL, Chemistry Department 11, Karolinska Institutet, Stockholm, Sweden

ALEXANDER KOLIN, Department of Biophysics, University of California School of Medicine, Los Angeles, California

HILTON B. LEVY, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland

COSMO G. MACKENZIE, Department of Biochemistry, The University of Colorado School of Medicine, Denver, Colorado

ALTON MEISTER, Department of Biochemistry, Tufts University School of Medicine, Boston, Massachusetts

OLAF MICKELSEN, Laboratory of Nutrition and Endocrinology, In- stitute of Arthritis and Metabolic Diseases, National Institutes of Health, Public Health Senrice, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland

HERBERT K. MILLER, General Medical Research Division of the Veterans Administration Hospital, Bronx, New York

vii

... v1u CONTRIBUTORS

NORMAN S . R ~ D I N , Biochemistry Department, The Medical School, Northwestern University, Chicago, Illinois

SIDNEY UDENFRIEND, Laboratory of Clinical Biochemistry, National Heart Institute, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland

JUNIUS M. WEBB, Division of Pharmacology, Food and Drug Ad- ministration, U. s. Department of Health, Education, and Welfare, Washingtn, D. C. (formerly National Institute of A l h g y and Infectious Diseases, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland)

HERBERT WEISSBACH, Laboratory of Clinical Biochemistry, National Heart Institute, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland

RICHARD S. YAMAMOTO, Laboratory of Nutrition and EndocrinJogy, Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Public Health Service, U. S. Department of Health, Education, and Welfare, Bethesda, Maryland


CONTENTS

New Developments in the Chemical Deterniination of Nucleic Acids. By Junius M . W.ebb and Hilton B. Levy . .

The hficrobiological Assay of Nucleic Acids and Their Deriva- tives. By Herbert K. Mil ler . . . . . . . . . . .

The Determination of Formaldehyde and Serine in Biological Systems. By Wilhelm R. Frisell and Cosmo G. Alackenzie . . . . . . . . . . . . . . . . . . .

New Methods for Purification and Separation of Purines. By Felix Berymann and Shabtay Dikstein . . . . . . .

Assay of Serotonin and Related Metabolites, Enzymes, and Drugs. By Sidney Udenfriend, Herberl Weissbach, and Bernard B. Brodie . . . . . . . . . . . . . .

By Anita J . Aspen and Alton Meister . . . . . . . . . . . . . . . . . . . .

Glycolipide Determination. By Norman S. Radin . . . . . Methods for the Determination of Thiamine. By Oluj Mickel-

sen and Richard S. Yamamoto . . . . . . . . . . . Rapid Electrophoresis in Density Gradients Combined with

pH and/or Conductivity Gradients. By Alexander Kolin . . . . . . . . . . . . . . . . . . . . .

Determination of Hexosamines. B y Sven Gardell . . . . . . Author Index . . . . . . . . . . . . . . . . . . . . . . Subject Index . . . . . . . . . . . . . . . . . . . . . Cumulative Index . . . . . . . . . . . . . . . . . . . .

Determination of Transaminase.

1

31

63

79

95

131

163

191

259

289

319

337

351

ix


New Developments in the Chemical Determination of Nucleic Acids

JUNIUS M. WEBB* AND HILTON B. LEVY, National Inslitules of Health, Belhesda, Maryland

I. Introduction.. . . . . . . . . . 11. Chemical Composition of 111. Principles Involved in Nucleic Acid Assay

. . . . . . . . . . . . . . . . . . .

1. General.. . . . . . . . . . . . . . . 2. Methods Dependent on th

A. Schmidt-Thannhauser B. Ogur-Rosen hlethod.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3. Methods Not Requiring Separation of RNA from DNA

B. Colorimetric Methods for DNA

. . , . . . . . .

4. Ultraviolet Absorption Methods 5. Comparison of Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Reference Standards. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . References . . . .. . . . , . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 2 3 3 4 4 7 9 9

10 16 16 18 19 20 24 26 27 27

I. INTRODUCTION

The purpose of this review is to cover the literature which has appeared since 1953 on chemical methods for the determination of nucleic acids, the period before having been adequately covered in the review by Volkin and Cohn (100). Since that time a number of reports dealing with the difficulties and inadequacies of available

* Present address, Division of Pharmacology, Food and Drug Adminis- tration, Washington, D.C.

1

2 JUKIUS M. WEBB AND HILTON B. LEVY

methods have appeared. In some instances suggestions for over- coming these difficulties have been presented. In addition, several new methods and modifications of older methods for determining the sugar components have been reported which, because of greater sensitivity, specificity, or both, offer possibilities of supplanting the more established methods.

A brief discussion of the composition of nucleic acids and principles involved in their estimation in biological material is followed by a discussion of methods. Detailed directions for performing some of the newer colorimetric assays for deoxyribonucleic acid .(DNA) and ribonucleic acid (RNA) based on the sugar components are included.

11. CHEMICAL COMPOSITION OF NUCLEIC ACIDS

Both RNA and DNA are high molecular weight polymers of nucleotides. The nucleotide is a combination of a purine or pyrimidine base with a phosphorylated sugar. RNA differs compositionally from DNA in the pentose-ribose, instead of deoxyribose-and a pyrimidine base-uracil, instead of thymine. The purines adenine and guanine, and the pyrimidine cytosine, are common to both RNA and DNA. In addition, some DNA's contain 5-methylcytosine, and the DNA of the Teven bacteriophage contains 5-hydroxymethyl- cytosine. In the intact nucleic acid the different nucleotides are linked together through the phosphate group. Hydrolysis of the nucleotide may yield a nucleoside (in which the base remains attached to the sugar) or a free base, a sugar (ribose or deoxyribose), and phos- phoric acid. Pyrimidine nucleotides are difficult to hydrolyze, where- as purine nucleotides are hydrolyzed with relative ease (20,48). As a consequence, in most methods for RNA and DNA based on color reactions for the sugars, only the sugar from the purine nucleotides reacts.

111. PRINCIPLES INVOLVED IN NUCLEIC ACID ASSAY

Chemical methods for the determination of nucleic acids may be based on phosphorus content, sugar content, or purine and pyrimidine content.

Whatever method is adopted should be preceded by an efficient

CHEMICAL DETERMINATION OF NUCLEIC ACIDS 3

disintegration of the material to be analyzed. This is important since the efficiency of removal of interfering substances and of the hydrolysis procedure to follow is directly dependent upon it. Various types of homogenizers, such as the Waring Blendor or, for small amounts of materials, glass or teflon tissue grinders, are employed. The material being homogenized should be kept cold during this process. For microorganisms sonic disintegration (77) or grinding with glass powder has been used for cellular disintegration. As Bonnar and Duggan (8) have shown, the conditions under which tissues are collected, stored, and treated prior to extraction of the nucleic acids have a critical bearing on the results. Consequently, in order to minimize effects of autolysis, tissues should be handled immediately on removal from the animal. Where this is impractical, quick freezing is employed. Similar considerations are to be given to operations with microorganisms.

All methods include a preliminary removal of interfering substances which involves acid extraction and treatment with alcohol and ether or other organic solvents (1) to remove lipids and, in plants, pigment materials. Details of these procedures have been described by Vol- kin and Cohn (100) and will not be repeated here.

Another consideration is the selection of suitable reference materials. Generally, purified preparations of DNA from calf thymus or fish sperm and RNA from yeast are used. One factor leading to the choice of materials from these sources is that these nucleic acids are readily obtainable in essentially pure form. For some purposes, one might consider the use of reference nucleic acids from other sources. Reference standards are discussed in Section IV. 6.

IV. DISCUSSION OF METHODS

1. General

Since phosphorus and the ultraviolet absorption of purines and pyrimidines are common to both RNA and DNA, it is clear that there first must be made a quantitative separation of the two nucleic acids, in order that these characteristics be a measure of each nucleic acid. The methods most widely used for this purpose are the Schmidt- Thannhauser (90) and the Ogur-Rosen (81) procedures.

On the other hand, methods based on the differences between the

4 JUNIUS M. WEBB AND HILTON B. LEVY

carbohydrates of RNA and DNA do not require separation of the two nucleic acids. This approach is exemplified by the Schneider procedure (91) in which simultaneous acid solubilization of both types of sugar is used.

2. Methods Dependent on the Separation of DNA from RNA A. SCHMIDT-THANNHAUSER METHOD (go)

Following the preliminary extraction to remove interfering substances, the residual material is treated with 1 N alkali a t 37” for 16-20 hours. During the alkaline hydrolysis the RNA is converted to acid-soluble nucleotides, while the DNA is not greatly degraded and precipitates when the solution is made acid. Phosphorus determinations on the acid-soluble and on the acid-precipitable fractions usually are used as measures of RNA and DNA, respectively, but ultraviolet absorption measurements can be used.

Since this procedure involves the determination of the amount of phosphorus in the separated RNA and DNA, it is possible for errors to arise from at least two sources: ( I ) the presence in the tissue of other phosphorus-containing compounds which are not removed during the preliminary treatment of the tissue and (2) inadequacy of the separation of DNA from RNA. Difficulties of both types have been reported. In some cases where the sole interest is the determination of the quantity of RNA and DNA, the errors introduced may be of secondary importance; but where the procedure is used primarily to separate the components of RNA and DNA preparatory to the determination of the amount of radioactivity in the separated components, the presence of a small amount of highly radioactive contaminant may be extremely important.

Logan et al. (60), for example, using the Schmidt-Thannhauser method, found in studying the white matter of brain that 66% of the total ‘inucleic acid” phosphorus was actually non-nucleotide phosphorus occurring in the alkali-hydrolyzable fraction. In brain gray matter about 38y0 of the phosphorus that was calculated to be nucleic acid phosphorus was of this non-nucleotide type. Davidson and Smellie (22), through the use of paper electrophoresis, showed that the alkali-hydrolyzable fraction from rat liver had five phosphorus- containing compounds in addition to those attributable to RNA. As pointed out by Leslie (55), marked differences are found in the ratio of RNA phosphorus to DNA phosphorus in rat kidney and spleen


when the results obtained by different methods are compared. When the Schmidt-Thannhauser procedure is used for analysis, the ratios for kidney and spleen are 2 and 1, respectively (45,66,67,86) while with the Schneider procedure the corresponding ratios are 1 and 0.3 (61,88,92). With kidney, the absolute amounts of DNA found are about the same by both methods, while the RNA as found by the Schmidt-Thannhauser method is higher than that found with the Schneider technique. This latter finding is consistent with the view that in the alkali-hydrolyzable fraction of the former procedure there are phosphorus-containing compounds that are not associated

TABLE I Comparison of Schmidt-Thannhauser (S-T) and Schneider (S) Methods on

Mammary Tumor Tissue (31) (Results express as pg./mg. dry tissue)

Tiasue samples

1A 1B 2A 2B 3A 3B

Total phosphorusb S T alkali hydrolyzate 13.3 12.3 7.4 S trichloroacetic acid ex-

tract 10.6 11.1 7.3 Ribonucleic acid (RNA)‘

Phosphorus on S-T supernatant 7.9 7 .6 4.3

Orcine on S-T supernatantd 49.6 47.6 29.2 Orcine on S extract 44.1 41.5 36.6

Deoxyribonucleic acid (DNA)‘ Phosphorus on S-T precip-

Diphenylamine on S-T pre-

Diphenylamine on S ex-

Residual phosphorus, S resi-

itate 5.0’ 5.2’ 2.5’

cipitate 45.8’ 48.1’ 26.5’

tract 76.6 66.4 52.8

due 1 . 1 0 . 5 0.0

8 . 1

8 . 1

4 .2 30.0 37.0

4.1‘

40.9‘

52.4

0.0

12.7

11.5

10.4 33.8 37.6

2.1’

25.7‘

103.0

1 .2

11.7

12.1

10.0 31.6 38.8

1 .v

32. O‘

110.8

1.3 ~~

a Samples 1, 2, and 3 were from mammary tumors of d8erent stages of de-

Phosphorus determined by the method of Fiske and SubbaRow (modified). RNA determined by the Mejbaum orcinol method. Average of two determinations.

* DNA determined by the Dische diphenylamine method. Schmidt-Thannhauser precipitate extracted with hot trichloroacetic acid (in

aamDle 2A extraction was short and did not extract all of the DNA from the

velopment.

pre6pitate). 8 Schmidt-Thannhauser precipitate hydrolyzed with NaOH.

6 JUNIUS M. WEBB AKD HILTON B. LEVY

with nucleotides. With spleen, the Schmidt-Thannhauser method shows a lower DNA and a higher RNA content than does the Schneider method, suggesting that with this tissue the alkaline hydrolysis converts some of the DNA into acid-soluble components. This is in conformity with the observations of Drasher (31) who showed that application of the Schmidt-Thannhauser procedure to mammary tumor tissue from C3H mice gave low DNA values in comparison with those obtained by the Schneider technique. Some of Drasher’s data on these procedures are reproduced in Table I.

It can be seen that the total phosphorus as determined in both instances is approximately the same; also orcinol determinations of RNA in the Schmidt-Thannhauser supernatant and in the Schneider extract agree reasonably. However, diphenylamine determinations show that the DNA in the Schmidt-Thannhauser precipitate is mark- edly lower than that in the Schneider extract. These findings could be attributed to some degradation of the DNA to acid-soluble components by the alkali treatment. It would have been of interest to have determined the deoxyribose content of the acid-soluble supernatant (the RNA fraction). It might be worth considering that, with some DNA preparations, pretreatment with trichloracetic or perchloric acids, particularly if extended or if the temperature is allowed to rise, might lead to loss of some purines. The resultant material, on its way to becoming apurinic acid, might be expected to show some of the increased alkali-lability associated with apurinic acid and might yield acid-soluble fragments after treatment with alkali.

Sherrat and Thomas (94) found that a very large fraction of the DNA of Streptococcus faecalis is not soluble in alkali, the exact amount being a function of the stage of growth of the bacteria. They found that this bound DNA was associated with a polysaccharide material, probably from the cell wall. It could be partly solubilized by boiling in 1 N alkali. Its base composition was not distinguishable from the alkalisoluble DNA. They found also that in the Schmidt-Thann- hauser “RNA fraction” about 15% of the total phosphorus could not be accounted for on the basis of the purine and pyrimidine content. On the same basis the polysaccharide-bound DNA had 22% excess phosphorus, while the soluble DNA had only 5% excess.

According to some, the strength of the alkali and the length of time of hydrolysis are important. De Lamirande et al. (23), following the observations of Daoust (21), found that with 0.3 N alkali as the hydrolyzing agent only 70% of the total nucleotides of the RNA of


rat liver were released, as determined by chromatography; 42 hours of treatment with 1.5 N KOH were required for complete hydrolysis. The proportions of nucleotides released did not change as time or alkali concentration changed.

Downing and Schweigert (30) , using the Schmidt-Thannhauser method for the separation of nucleic acids as a preliminary to enzy- mic hydrolysis of the DNA to deoxyribonucleotides, found the metbod satisfactory for E. coli and for animal tissues, but not for E. gracilis or L. leichmannii. In the latter instances poor yields of deoxyribonucleotides were obtained, and these were contaminated with ribonucleotides. I n addition, two-thirds of the DNA from these latter organisms were not alkali soluble (cf. Sherrat and Thomas (94)). Volkin and Astrachan (98) found that the Schmidt-Thannhauser procedure could not be applied to the lysates of bacteria that had been attacked by bacteriophage.

Winder and Denneny (104) reported that the Schmidt-Thann- hauser procedure could not be applied to mycobacteria because of their high lipid content arid the difficulty of its removal (54). They experienced a low recovery of DNA and found that some of the DNA remained in the original alkali-insoluble fraction. This DNA could be solubilized with hot trichloracetic acid. After several modifications were introduced, they obtained fair agreement between their modified Schmidt-Thannhauser method and other analytical procedures. Loring (62) reported some adsorption of RNA by the Schmidt-Thannhauser precipitate of protein and DNA. The difficulty presented could be overcome by redissolving and reprecipitat- ing this acid-insoluble material. MouM (78) found that the alkaline hydrolyzate of rat liver contained some acid-soluble peptides showing optical absorption in the ultraviolet.

The findings of these investigators should not be interpreted as meaning that satisfactory results are not obtainable with this procedure. However, as Drasher has pointed out (31), the application of the Schmidt-Thannhauser method to new tissue should be preceded by a determination of the applicability of the method to that tissue.

B. OGUR-ROSEN METHOD

The Ogur-Rosen (81) method of separation of RNA from DNA depends upon the observation that RNA is solubilized by treatment with 1 N perchloric acid for 18 hours at 4"C., while DNA is not. The


latter, however, becomes solubilized by 20-minute treatment a t 70°C. in 0.5N perchloric acid. Quantification of the amounts of RNA and DNA in the fractions is customarily made by ultraviolet spectropho- tometry, but in some cases phosphorus or sugar content might be used. The procedure, which was originally developed for determining small amounts of RNA and DNA in root tips and pollen, has been applied to other biological materials. A number of difficulties have been reported.

Loeb and Dickinson (58), working with c6.I mouse thymus, studied the relative base and phosphorus content of the RNA fraction (soluble in cold HC1O4) and the DNA fraction (soluble in hot HC1O4). They found that the phosphorus in the RNA fraction was less than that expected from the content of ultraviolet absorbing material, while the reverse was true of the DNA fraction. This suggested that some DNA purines were solubilized by the cold HC104 treatment. This had already been pointed out by Ogur et al. (80) for yeast.

Loeb and Dickinson also compared the values they obtained, using the Ogur-Rosen separation coupled with phosphorus determinations, with the data obtained by Ceriotti (14) ,who extracted the total nucleic acids from the same tissue using hot HC104. Loeb and Dickinson concluded that Ceriotti's procedure gave higher values. Winder and Denneny (104) modified the Ogur-Rosen procedure by substi- tuting 5% trichloroacetic acid for HC1o4. Using purified DNA from M . phlei they found that three extractions with 5% trichloroacetic at 19", totalling 72 hours, removed purines but not phosphorus or pentose. The resultant apurinic acid was hydrolyzable and extractable by heating with 5% trichloroacetic acid for 15 minutes a t 90'. The optical absorption of this extract had an &f,#)/[email protected] ratio of 0.77. Purified RNA, on the other hand, treated similarly, had a ratio (E260/E268.6) of 1.05. Comparing these ratios with those obtained from whole organisms ( M . smegmatis), they concluded that the RNA and the DNA purines of organisms were extracted by the 19" treatment, while the remainingapurinic acid (from the DNA) was extracted at the higher temperature. Using an E(P) (see Section IV.4) a t 268.5 mp of 5700, they estimated DNA phosphorus on the heated extract. Similarly RNA phosphorus was estimated on the combined 19°C. extracts, using an E(P) at 268.5 mp of 9800, and making allow- ance for the absorption of the purines from DNA.

Logan et al. (60) found that perchloric acid was not as good a protein precipitant as trichloracetic acid. Cassel (12) found, with

CHEMICAL DETERMINATION OF NTJCLEIC ACIDS 9

Racilli~s cereus, that 30 hours of extraction with perchloric acid were needed to remove all the R N A from fised bacteria. Prior extraction of lipids did not shorten this time requirement. Koenig and Stah- lecker (53) developed a modification of the Ogur-Rosen method for fixed tissue sections. Basler and Commoner (6,7) showed that the Ogur-Rosen procedure did not lead to adequate separation of RNA from DNA in tobacco leaf tissue. They extracted the total nucleic acid from the tissue by 30 minutes heating at 80" in 0.5 M perchloric acid.

Perhaps a good example of the problems that are encountered in nucleic acid determinations when applying methods developed for one substance to another substance is seen in studies of the nucleic acid in influema virus. In earlier work on a purified preparation of the virus, Knight (51) found about 5% nucleic acid. The determination was based on the assumption that the total phosphorus minus lipid phosphorus was nucleic acid phosphorus. Later work by Knight (52) suggested that both RNA and DNA were present. Other workers presented conflicting evidence (40,96). Recently, Ada and Perry (1,2) using a modified Schmidt-Thannhauser method on the intact virus, found 1.0% RNA and 0.04% DNA. They found that adequate removal of the phospholipid required the use of a mixture of chloroform, methanol, and ether. This could account for the higher total amount of nucleic acid found by Knight. The Ogur-Rosen procedure applied directly to the virus also gave evidence of a trace of DNA. Ada and Perry then extracted the total nucleic acids from the virus with hot 10% sodium chloride solution. Using the Schmidt- Thannhauser procedure, they found this extract to contain no DNA and concluded that the virus contains only RNA. However, on the basis of a positive microbiological test for thymine, Miller (76) has cautioned against too ready a judgment on the absence of DNA.

3. Methods Not Requiring Separation of RNA from DNA

A. SCHNEIDER METHOD

I n this procedure, following the cold trichloroacetic acid, alcohol, and alcoholether extractions, the tissue residue is heated in 5% trichloroacetic acid 15 minutes a t 90-95". Both RNA and DNA are hydrolyzed, while most of the tissue protein remains insoluble. According to McIndoe and Davidson (74), working with free nuclei, the hydrolysis procedure splits off all the reactive sugar components.


but appreciable amounts of phosphorus are left bound to protein. This retention of phosphorus probably is different among biological materials, since the findings of Drasher (see Table I) and Patterson and Dackerman (83) show negligible amounts of phosphorus in various residues and close to theoretical amounts of nucleic acid phosphorus in the hydrolyzates. The findings of Webb (101; see Table 111) are also consistent with the observations of Drasher and Patter- son and Dackerman.

Certain difficulties with the Schneider procedure are worth noting. Webb and Levy (103) observed that, following acid hydrolysis, more protein material could be precipitated, in some instances, by the addition of more trichloroacetic acid. This protein material, if left in solution, could be a source of error in the various assays performed on the extracts. Lindigkeit and Rapoport (57) claimed that 15-minute hydrolysis at 90" with 5% trichloroacetic acid was insufficient to extract the nucleic acids from blood erythrocytes and brain white matter. In another investigation Hirtz and Fayet (41) have reported difficulties encountered in the application of the procedure to cow- tongue epithelium and have suggested performing the alcohol and alcohol-ether extractions prior to the cold trichloroacetic acid extraction. The interference of the absorption of trichloroacetic acid in the ultraviolet measurement of nucleic acids has led to the substitution of perchloric acid extraction by some investigators (81).

The important advantage of the Schneider procedure is that no separation of DNA from RNA is required, since the determinations are based on characteristic color reactions for pentoses and deoxy- pentoses. The following sections are devoted to a discussion of modifications of older colorimetric methods and to new methods that have appeared since 1952.

R. COLORIMETRIC METHODS FOR DNA

Methods used for DNA estimation, based on the sugar component., are not specific for 2deoxyribose but are more or less specific for 2- deoxy sugars. The occurrence of 2-deoxy sugars or their compounds in nature, other than in DNA, is relatively rare.

The most widely known and most frequently used method for the determination of DNA is that described by Dische (26). It depends upon the formation of a blue color (maximal absorption, 595-600 mN) when DNAis heated at 100°C. with diphenylamine in a mixture of glacial acetic and sulfuric acids. The mech-

Diphenylamine Test.

CHEMICAL DETERMINATION OF SUCLEIC ACIDS 11

anism of the reaction has been extensively investigated by Dische (28) and others (24,82) who have attributed the color to the formation in the reaction mixture of w-hydroxylaevulaldehyde, which subse- quently condenses with diphenylamine.

Various interfering substances have been known since at least 1936 (85). More recently Ogur et al. (79) have reported the presence of a substance in the flower bud of Lilium longiflorum, extractable with petroleum ether, which would interfere with the diphenylamine reaction. Ayala et al. (3) report a mucoprotein in bovine tonsil extracts and serum which gives a purple color (maximal absorption, 530 mp). The reaction was also given by trichloroacetic acid hydrolyzates of these materials. Holden (42) , investigating the application of the reaction to plant materials, found galacturonic acid to give a blue color with diphenylamine, but the color developed more slowly than that from DNA. Among other recent investigations, 1,ogan d al. (60) reported substances in dog brain tissue that inter- fered with the Dische test. In addition, the presence of proteins has been reported to interfere with the reaction (29).

In order to improve the specificity, in some cases, Dische has recommended optical density readings a t two different wavelengths (28). For example, the difference between the absorptions a t 595 and 650 mp can be used to determine small amounts of 2-deoxyribose when the characteristic blue color is obscured by the green color pro- duced by interfering substances. For quantitative purposes, the difference, 0.D.696- O.D.W, was found to give better agreement between duplicate samples than when a single reading a t 595 mp was used as a measure of the concentration of DNA (28).

Recent investigations (1 I ,83) have shown that the color density of the diphenylamine reaction can be increased by carrying out the reaction a t a lower temperature and over a longer period of time. For example, Burton (1 1) , using a modified diphenylamine reagent containing acetaldehyde, incubated the reaction mixture for several hours a t 30". The modified method is claimed to be 3.5 times as sensitive as Dische's original procedure.

Burton used suitable extracts of the biological materials in 0.5 N perchloric acid so that the filial solution contained 6-80 pg. of DNA per nil. Trichloroacetic wit1 extracts may also Le used provided per- c.liloric, acid is adcled to gi1.c :I c*oticentr:ition nl' 0.5 N with respect to the latter before addition of the modified diphenylamine reagent. 'l'he reagent is prepared by dissolving 1.5 g. of steam-distilled di-


phenylamine in 100 ml. of redistilled acetic acid and adding 1.5 ml. of concentrated H$04. The reagent is stored in the dark. At the time it is to be used, 0.10 ml. of aqueous acetaldehyde (16 mg. per ml.) is added for each 20 ml. of reagent required. The assay is performed as follows:

A measured volume of the extract (1 or 2 ml.) is mixed with 2 volumes of the modified diphenylamine reagent and the color is developed by incubating a t 30" for 16-20 hours. Tubes containing known amounts of standard DNA and a blank containiig 0.5 N perchloric acid, but no DNA, are treated in a similar manner. The optical density a t 600 mp is measured against the blank and compared with the values obtained with a standard DNA solution.

Besides greater sensitivity, it is claimed that this modified method is less susceptible to interference by other compounds. The amount of color given by moderate amounts of RNA and certain other substances was negligible, but the presence of certain substances, nota- bly cysteine and ascorbic acid, appreciably reduced the color formed by the reagent with DNA.

Burton has applied his method to purified DNA preparations from calf thymus, E. coli, and bacteriophage T-2. He also used the method in an investigation of nucleic acid metabolism in bacteriophage T-2

Another recently proposed method for DNA assay takes advantage of a reaction between p-nitrophenylhydrazine and deoxyribose when heated in trichloroacetic acid (103). In this procedure the biological material is hydrolyzed 30 minutes, rather than 15 minutes as recommended by Schneider, since slightly greater color intensities resulted in the final solution. Hydrolysis is carried out in centrifuge tubes, the mouths of which are covered with sealed ampoule bulbs to minimize loss from evaporation. Fol- lowing hydrolysis a volume of trichloroacetic acid (5%) equal to the original volume is added to each tube.

The assay is performed as follows:

Two ml. aliquots of the diluted hydrolyzates containing 5-150 pg of DNA per ml. are transferred to 15 ml. glass-stoppered conical Centrifuge t,uhes. Into each are pipetted 2 ml. of 5% trichloroacetio acid and 0.2 ml. of freshly prepared p-nitrophenylhydrazine reagent (0.5% in 95% ethyl alcohol). Tlw tubes are heated 20 minutes in a boiling water bath, using a sealed ampoule hulb for a condenser. After cooling in cold water, the solutions we ex-

(10). p-Nitrophenylhydrazine Test.

CIIEMICAI, DETERMINATION OF NUCLEIC ACIDS 1.7

tracted with 10 ml. of butyl acetate, centrifuged and the greater portion of the organic layer decanted and discarded. 3 ml. of the aqueous phase of ~ a c h tub, taken by dipping a 3 ml. volumetric. pipet beneath the organir phase, :ire transferred to a 5 nil. voliinietric flask. 1 ml. of 2 N NaOH is added to e:tch to develop the color :tnd the solution diluted to volume with water. A purple cdnr tlevclops ininietlintely. The optical densities at 560 mp are measured within a minute after color development against the blank (4 ml. of 5% trichloroacetic acid carried through the same procedure) and compared with the value obtained from 100 pg. of standard DNA.

40 z 0

50

2 60 cn z 9 70 I-

t- 00 z W

'"

90

100

W a

I. p-Nitrophenylhydrozine 2. Diphenylarnine

400 560 700 600

WAVELENGTH ( rnp 1

Fig. 1. Absorption spectra of the products of the p-nitrophenylhydrazine and diphenylamine reactions with hydrolyzed DNA. The initial concentra- tions of hydrolyzed DNA were such that the final colored solution in each case was equivalent to 15 pg. of DNA per ml.

The p-nitrophenylhydrazine test shows greater sensitivity than the Dische test.. Figure 1 shows a comparison of extinction coefficients of solutions of the same concentration of DNA, one treated with p- nitrophenylhydrazine and the other with diphenylamine. It can be seen that the color of the former a t its maximum is about 5 times as intense as the latter at its maximum.

Luder- itz (64) has studied the specificity of the reaction in connection with his investigations of the structure of abequose and tyvelose, both of which on oxidation with periodate yield products which give a positive p-nitrophenylhydrazine test. Of a number of carbohydrate sub-

The test appears to be very specific for 2-deoxy sugars.

stances tested, only 2-deoxy sugars and oxymethylfuran gave the test. It is probable that only those sugars capable of forming such a furan derivative in dilute acid give a positive reaction.

Analyses for DNA of rat lung, liver, and kidney; yeast; and hac- teria (I?. coli and P . viclgnris) I)y thc mri hod ronsistcntly shomctl values slightly lower than those found for the same materials by the Dische diphenylamine method. This small difference might be attributed to non-specific color formation in the diphenylamine reaction resulting from the action of the strong acid solution on various substances in the tissue extracts. Dische (26) and Seibert (93) have employed an additional control to compensate for this non-specific color.

The procedure to be discussed here, proposed by Ceriotti (13), is a modification of a reaction first described in 1929 (25). The test depends upon the formation of a yellow-brown color when DNA is heated with indole in HCl solution. The method is reputed to be about 10 times as sensitive as the Dische diphenylamine reaction.

Perchloric acid extracts of tissues are used because trichloroacetic acid was found to inhibit, the reaction with indole (14). The assay is performed as follows:

To 2 ml. of extract containing the equivalent of 2.5-15 pg. per ml. of DNA are added 1 ml. of 0.04% indole solution in distilled water and 1 ml. of concentrated HC1. The tube is placed in a boiling water bath 10 minutes and then cooled in cold water. The solution is extracted 3 times with 4 ml. por- tions of CHClp, the water layer being separated from the organic phase by centrifugation. The intensity of the yellow color left in the water phase is measured in a Beckman spectrophotometer at 490 mp against a blank treated in an identical manner and compared with the value obtained for a standard DNA solution.

Ceriotti states that the purity of the CHC13 is of utmost importance and recommends its purification by extracting with concentrated HBO,, followed by extracting with water, drying over CaC12, and distilling. The product recovered at a boiling point of 61" is used for the assays.

Moderate amounts of RNA or ribose and several other carbohydrates give colors of varying intensity with indole, but the colors are completely extracted by CHC13. According to Dische (28), both galacturonic and arabinose give considerable yellow-brown color in the water phase which color is not extractable with CHC13. Ceriotti

Indole Test.

CHEMICAL DETERMINATION OF KUCLEIC ACIDS 15

has coil timed that the presence of sigriificaiit amouiits of arabinose i t i solution may seriously interfere with the test (13).

The mechanism of the reaction is not known but is probably the same as that of DNA with tryptophan, an indole derivative which also gives a color reaction though much less intense, with DNA (18).

The application of the method to some biological materials has t)een shdied. Ceriotti has compared DNA values for a number of rat and mouse tissues by the indole method with values obtained for the same tissues by the Dische diphenylamine procedure (14). The values found by the two methods were essentially the same. No data for other than mammalian tissues were shown. Durand (33) used the Ceriotti method for determining the DNA content of gametes of Gryllus domesticus and compared the result with that obtained by a11 isotopic dilution method. By the indole method there was 0.024 pg. of DIVA per egg and by the isotopic dilution method, 0.01 pg. per egg. The investigations of Loeb and Dickinson (58), using the Ceriotti method, are described in Section IV. 2. B.

Keck (47) has presented a micro modification of the Ceriotti procedure which allows estimation of DNA in amounts of 0.1 to 1 pg. in 20 pl. of solution. Amy1 acetate, inst,ead of CHC13, is used for extraction ( 2 times). In this modification, i t is claimed, the color given by arabinose is completely extracted and trichloroacetic acid does not interfere with the reaction.

Cysteine, also proposed hy Dische, gives a more or less specific reaction with DNA (2'7). There have been modifications of this reaction, one proposed by Stumpf (95) and another, more recent, by Brody (9). The latter studied the effects of several variables and stressed the need for mork- ing under strictly defined conditions to attain reproducibility. Accord- ing to Manson (68), pyrimidine, as well as purine-bound deoxyribose, is measured in the Brody procedure. Although the cysteine reaction has been known for a number of years, it has never gained much prominence. Part of its poor awept,aiice may be due to the difficult'ies in making :dequat,e correctioti for non-spec:ific color resulting from the action of the strotig I12S04 (75(%,) 0 1 1 t,issue extracts. Further, cysteine ofters no advantage over tliphenylaniiiie as far as specificity is (:on(:eriied and is much less sensitive. IIolden (42), comparing the results of the tryptophan (M), diphenylaniiiie (Sci), and cysteinc: (95) methods as applied to plant tissues, found the cysteine method to give erratic results.

Othw Reactions for D N A Determination.


Another recently proposed method involves the reaction of deoxyribose with anthrone (39). The method is not specific, and correction is necessary for RNA interference. Hexoses also interfere. The DNA product with anthrone shows a maximum at 565 mp, while RNA shows a peak a t 620 mp.

Still another method (38), based on the absorption by DNA of methyl green, also appears nonspecific and to offer no particular advantage over those methods discussed.

C. EVALUATION OF COLORIMETRIC METHODS FOR DNA

Adequate evaluation of the modified older methods or the newer methods must await more extensive critical application to a variety of biological materials. That all of these proposed methods appear to be striving for more sensitivity than the original Dische procedure is a reflection of the fact that presentday nucleic acid investigations often require working with small amounts of material.

Where high sensitivity or specificity is not required, the Dische method offers advantages in simplicity of application and in being a time-tested reaction, the limitations of which are better known. Even in these instances, however, the availability of other methods, which may be more specific, is advantageous, if for no other purpose than checking the results of the Dische method.

D. COLORIMETRIC METHODS FOR RNA

Colorimetric assay methods for RNA lack the specificity of those available for DNA since the former are more or less general reactions of pentoses and certain other carbohydrate substances. I n an attempt to gain some degree of specificity, the reactions are carried out under carefully controlled conditions (acidity, temperature, etc.) which are optimal for conversion of ribose to furfural or furfural derivatives and minimize such conversion for other sugars. The furfural or furfural derivative formed is then reacted with various chromogenic substances.

The most commonly employed methods are the Mejbaum (75) orcinol procedure and a similar method which employs phloroglucinol (37). With the former, the results must be corrected for DNA interference. DNA does not give a color in the phloroglucinol procedure, but the method is less sensitive than that with orcinol. This absence of interference by DNA is probably a result of a prolonged heating

(!HEMICAI, DRTERMINATIOS OF NUC1,EIC ACIDS 17

period which destroys %deoxyribose or any derivatives which can form colored compouiids with phloroglurinol, rather than of any greater specificity of t,he reagent,.

There have been many modifications of the original orcinol (Bial) reactions, some of which, like in the Mejbaum procedure, use FeCl3 (4,81,89) as a catalyst while others use CuC12 (5). In order to increase the sensitivity, specificity, or both, in some of the modifications the green-blue pigment formed is extracted. One of the most recent examples of this method is that proposed by Ceriotti (14) in which the pigment is extracted with isoamyl alcohol. The assay is performed as follows :

To 5 ml. of the solution to be tested, containing 25-200 pg. of RNA, are added 5 ml. of the orcinol reagent (200 mg. of orcinol and 6.1 mg. CuClaH20 in 100 ml. concentrated HCl). The contents of the test tubes are mixed; the tubes are immersed in boiling water for 40 minutes and then cooled under running water. The color is extracted with 5 ml. of isoamyl alcohol and, after centrifugation, is read a t 675 mp against a blank treated in the same manner. The value obtained is compared with that from a standard RNA or ribose solution.

No data were given for interfering substances otherrthan DNA. Interference by DNA was only 0.85% compared to 12% found by others (37,101) employing the Mejbaum procedure (without extraction). It would have been of interest if the extent of interference of various carbohydrate substances known to interfere in the Mejbaum procedure had been shown.

In another recently proposed method (101) the furfural from RNA is trapped in xylene and the xylene extract caused to react with p-bromophenylhydrazine.

p-Bromophenythydrazine Method.

To 1 ml. of the 5% trichloroacetic acid extracts, containing 9-200 pg of RNA, is added 1 ml. of 8 N HC1, followed by 1 ml. of xylene (c.P.) and enough NaC1 crystals to saturate the mixture. The reaction mixtures, in 12 ml. centrifuge tubes are placed in a boiling water bath for 3 hours. After cooling in running water, 2 ml. of xylene are added to the contents of each tube. The tubes are centrifuged and 2 ml. of the xylene layers are transferred to 5 ml. volumetric flasks to which are added 2 ml. of p-bromophenylhydrazine reagent. The reagent is a 2.5% solution of p-bromophenylhydrazine in ethyl alcohol-HC1 solution (2 ml. of HCl, 37%, added to 100 ml. of 95% ethyl alcohol) prepared fresh daily. The color is developed by incubating a t 37" for 1 hour. After diluting to volume with ethyl alcohol-HC1 solution, optical densities are measured at 450 mp against the blank (1 ml. of 5% trichloroacetic

18 JUSIUS M. WEBB AKI) HILTOS €3. LEVY

acid tre:ttetl in the identical manner) and compared with values obtained for a standard RNA solution.

The amount of color given by 1 mg. of DNA was negligible, but several other carbohydrate substances yielded various amounts of color under condit,ions of the test. Galacturonic acid, which gave about the same amount of color as that given by the same weight of RNA, was the most serious potential interfering substance tested.

Galacturonic acid has been shown to interfere with a number of the colorimetric sugar reactions used in nucleic acid estimation. This is important to note since polysaccharides, found in tissues and bacteria, may be bound to protein in the native state (84) and contain uronic acid groups. It is conceivable these uronic acid groups may not be removed by cold extractions but could be released, at least in part, by heating with acid.

Mauritzen et al., in their studies of thymus nuclei, distilled the furfural formed from RNA and measured it by the amount of color given with aniline acetate (73). A similar procedure has been worked out by Dunstan and Gilliam (32) in which the furfural is measured spectrophotometrically a t 278.5 mp. These methods are not well adapted to small amounts of material.

Other Methods for RNA Determination.

E. EVALUATION OF COLORIMETRIC METHODS FOR RNA

The reliability of the results obtained by the methods described is limited by large compositional variations in ribonucleic acids and by the unspecific nature of the methods. The former difficulty would partly be overcome if the total ribose (see Section IV. 6) was measured, but suggestions for accomplishing this (72) have not been satisfactory (48). Another limiting factor is the presence of interfering substances in the nucleic acid extracts which may give a positive test with all the methods since the mechanism of the reactions is essentially the same. Methods in which the furfural formed is isolated prior to treatment with the chromogenic agent appear more specific; a t least those interfering substances which do not yield furfural are eliminated. The sensitivity of the p-bromophenylhydrazine method and the Ceriotti modified orcinol procedure described are about the same. Both of these methods are more sensitive than the Mejbaum method.

Documents

METHODS OF BIOCHEMICAL ANALYSIS · METHODS OF BIOCHEMICAL ANALYSIS VOLUME VI CONTRIBUTORS ANITA J. ASPEN, Department of Biochemistry, Tufts University School of Medicine, Boston,