78 MUNRO AND FLECK: RECENT DEVELOPMENTS IN THE jlifzaiyst, i’o1. 91 Recent Developments in the Measurement of Nucleic Acids in Biological Materials A Supplementary Review* B Y H. N. MUKRO AND A. FLECK? (Department of Biochemistry, The University of Glasgow) SUMMARY OF CONTENTS Introduction Preparation of tissue samples for nucleic acid determinations Precautions during removal of tissues Extraction of acid-soluble compounds Extraction of lipids 1. Methods of determining nucleic acids in the tissue residue The procedure of Schmidt and Thannhauser The use of alkaline hydrolysis to separate RNA from DNA The RNA fraction of the alkaline digest The DN-4 fraction of the alkaline digest Recommendations for the use of the Schmidt - Thannhauser method 2. The Schneider procedure 3.4. Other procedures The procedure of Ogur and Rosen General recommendations for nucleic acid determination. IN 1961, Hutchison and Munrol reviewed the literature on methods of determining nucleic acids in biological materials. This article is an account of developments that have taken place since 1961 in the field of nucleic acid analysis. For convenience, the main topics will be dealt with in the same order as that used in the earlier review. Under each heading the conclusions reached in the 1961 review will be summarised, and subsequent developments will then be discussed. A comprehensive survey of nucleic acid analysis based on the literature covered by both reviews is to be published elsewhere.2 Since 1961, most investigators have relied on modifications of the procedure of Schmidt and Thannhauser3 as the method of choice for determining the nucleic acid content of tissues.Consequently, in this survey of recent developments we shall devote more space to this method than to the other two main procedures, namely the Schneider method4 and the method of Ogur and R o ~ e n . ~ Nucleic acids contain 3 distinct components ( a ) purine and pyrimidine bases, ( b ) ribose or deoxyribose and (c) phosphorus. The principal methods of determining nucleic acids have therefore been based on the strong ultraviolet absorption of the bases, or on specific colour reactions for the pentoses, or on the determination of phosphorus. Procedures depend- ing on ultraviolet absorption or on phosphorus determination are clearly common to both RNA and DNA and therefore demand preliminary separation of the two nucleic acids.This is the objective in the original Schmidt - Thannhauser3 method in which phosphorus deter- mination is used, and in the Ogur - Rosen5 procedure which depends on ultraviolet-absorption measurement of the separated nucleic acids. On the other hand, in the Schneider method4 RNA and DNA are measured on the same tissue extract by specific colour reactions for ribose and deoxyribose, respectively. Before these specific methods for nucleic acid determination are used, certain procedures have first to be applied to the tissue. Usually this preliminary treatment involves the removal of small molecules (e.g., free nucleotides) and if necessary, lipids, since these may interfere with later chemical determinations. This initial treatment is followed by extraction of the * Reprints of this paper will be available shortly.For details sce Summaries in advertisement pages. t Present address : Departmcnt of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Mass.February, 19661 MEASUREMENT OF KUCLEIC ACIDS IK BIOLOGICAL MATERIALS 79 nucleic acids from the tissue residue and their subsequent determination. Consequently, the main methods for nucleic acid determination can be considered as having 3 stages- (a) Preliminary preparation of the tissue samples for nucleic acid determinations. This generally involves removal of small molecules and sometimes lipids; this stage is common to all three main procedures for nucleic acid measurement.(b) A procedure for the extraction and, if necessary, separation of the nucleic acids. In the Schmidt - Thannhauser method, hydrolysis of RNA in alkali is used to extract it from the residue and separate it from DNA, which resists attack by alkali and can therefore still be precipitated with acid. In the Schneider procedure, both nucleic acids are extracted from the tissue simultaneously with hot acid. In the Ogur - Rosen method, the RNR is extracted with cold perchloric acid and the DNA is then obtained by applying hot perchloric acid to the tissue. Application of a specific procedure for determining the nucleic acid present in the extract. (c) THE PREPARATION OF TISSUE SAMPLES FOR NUCLEIC ACID DETERMINATIONS PRECAUTIONS DURING REMOVAL OF TISSUES- It was pointed out in the earlier review1 that considerable losses of nucleic acids can occur through enzymic action if a tissue is not rapidly cooled on removal and kept chilled during the preliminary stages of the analytical procedure.In order to avoiasuch losses, it has long been the practice to excise and homogenise the tissue at 0" C. In most instances, rapid excision, homogenisation in cold water and addition of a cold protein precipitant have been found to be adequate. However, some workers have considered it necessary to take additional precautions for special purposes. If the free nucleotide pool is to be examined, Saukkonen6 recommends freezing the tissue instantly in liquid nitrogen, a procedure for which he gives full details and references. This method was used by Logan, Mannell and Rossiter' for determining nucleic acids in nervous tissue, as the time required for dissecting out this tissue can be quite lengthy.Despite this, no such special precautions were recorded in some recent investigations into the concentrations of nucleic acids in the b ~ a i n . ~ ~ ~ In a review of nucleic acid determination in plants, Markhamlo suggested that the tissue should be plunged into boiling ethanol in order to inactivate nucleases. Several authors studying plant material1l9l2 9 1 3 and ova14 have recently reported that this procedure is satis- factory, and in unpublished studies we have found the method to be readily applicable to rat liver. Rapid dissolution of tissue with inactivation of enzymes has also been achieved by immersing the sample in detergent solutions, such as deoxycholate in a solution of urea16 and 1 per cent.Triton X 100.16 If the tissue is not immediately analysed, the investigator usually freezes the sample. The conditions of storage can sometimes be critical. May and Grenelll' reported large losses of RNA from rat brain stored in a deep-freeze cabinet, but Smithls found.no change in liver RNA content under these conditions. However, Curriel9 has demonstrated that the xanthine content of rat liver increases during storage at -13" C, suggesting that some enzyme activity continues even at this temperature. I t is therefore essential that the analyst should ensure, by comparison of results obtained from both fresh and stored tissue samples, that the conditions of storage are adequate to prevent serious losses of nucleic acids.EXTRACTION OF ACID-SOLUBLE COMPOUNDS- Three groups of tissue components of low molecular weight may interfere with the estimation of tissue nucleic acids : ( a ) free nucleotides ; (b) carbohydrates ; (c) inorganic phos- phate and organic phosphorus compounds of low molecular weight. For example, it has been calculated2 that failure to remove adenine nucleotides would result in a 10 per cent. over-estimation of the RNA content of rat liver. The usual procedure for removal of such compounds of low molecular weight is by precipitation with cold acids, usually trichloroacetic acid or perchloric acid. Various concentrations of acid and conditions of extraction have been used; these were discussed, in detail, in the earlier review.1 In general, maximal removal of acid-soluble phosphorus has been used as the criterion of satisfactory extraction of small molecular weight compounds.Recently, Hallinan, Fleck and Munro20 have used a modified Schmidt - Thannhauser method of nucleic acid determination in order to ascertain the optimal conditions of extraction of low molecular weight contaminants with trichloroacetic acid and80 MUNRO AND FLECK: RECENT DEVELOPMENTS IN THE [Analyst, lTOl. 91 perchloric acid. After homogenisation of liver samples in cold water, cold acid was added to give various concentrations of acid, and the tissue residue was then washed twice with the same concentration of acid. They found that various concentrations of trichloroacetic acid, from 5 per cent.to 20 per cent., all gave maximal recovery of RNA. With perchloric acid, the same recovery of RNA was also obtained after precipitation with cold 0.2 N acid, but with concentrations above 0-3 N, the recovery of RNA fell progressively. This latter observation is in agreement with the findings of Ogur and R o ~ e n . ~ It is consequently important not to use high concentrations of perchloric acid to remove small molecules. In unpublished experi- ments we have shown that the use of 0.2 N perchloric acid at 0" C also results in maximal precipitation of DNA and protein from rat liver. The general use of this concentration of perchloric acid in the initial stages of nucleic acid determination can therefore be recom- mended, especially since perchloric acid has the advantage over trichloroacetic acid that it does not absorb ultrcviolet light and thus does not interfere with subsequent nucleic acid deter- minations involving ultraviolet absorption.However, it should be noted that the standard conditions of extraction with 0.2 N perchloric acid or with 10 per cent. trichloroacetic acid have been found to give rise to some difficulties with certain tissues. This has been reported by two authors2lyz2 studying yeast, and detailed discussion of difficulties in the use of cold acid to extract plant tissues is provided by Smillie and K r ~ t k o w . ~ ~ Finamore and Volkin24 reported earlier that half of the RNA isolated with phenol from amphibian eggs is soluble in 0.5 N perchloric acid, but this material has since been shownz5 to consist of small oligo- nucleotides." The claim by Levy and Lynt26 that some RNA in HeLa cells remains unpre- cipitated by 1.0 N perchloric acid may merely imply that the expected degree of breakdown of RNA at this concentration of perchloric acid had occurred in their experiments.The technique of cold-acid precipitation for removal of small molecules can be applied in two ways. We prefer to begin by homogenising the tissue in cold water and then adding cold acid to samples of the homogenate. The alternative approach is to homogenise the tissue directly in the cold-acid precipitant, but this often leads to clogging of the homogeniser with precipitate, specially in the Potter - Elvehjem type of homogeniser, and, in addition, sampling of the homogenate becomes difficult.A specially designed homogeniser for rapid tissue disintegration and simultaneous addition of acid has been recently de~cribed.~' EXTRACTION OF LIPIDS- The original procedures of Schmidt and Thannhauser3 and of Schneider4 include a stage at which tissue lipids are extracted, in order to remove the phosphorus present in the tissue in the form of phospholipids. As most modern methods of nucleic acid determination do not depend on phosphorus determination, this step is no longer obligatory, but is never- theless still extensively used. A considerable variety of lipid-extraction procedures have been used in the past. They are discussed in detail in the earlier review,l in which a standard extraction procedure was offered involving ethanol, then ethanol - chloroform, ethanol - ether and finally ether.The question of adverse effects of lipid solvents on recoveries of nucleic acids was raised by Venkataraman and Lowe,28~29 who treated liver samples with 5 per cent. trichloroacetic acid to remove small molecules and then extracted the lipids with cold 95 per cent. ethanol. Under these conditions, they observed losses of up to 30 per cent. of the tissue RNA. In our earlier review,l it was pointed out that several other authors4~30~31~32 had reported that no losses occurred when lipid solvents were used to extract tissues precipitated with cold trichloroacetic acid, and it was concluded that the findings of Vcnkataraman and Lowe did not have general application to nucleic acid determination. However, we subsequently noted that all the investigators who observed no loss of RNA had used 10 per cent.trichloroacetic acid, whereas Venkataraman and Lowe had used 5 per cent. trichloroacetic acid. Accordingly, Hallinan, Fleck and Munro20 re-investigated the situation and found that the concentration of cold acid applied initially to the tissue was critical in determining the extent of loss of RNA into ethanol and other lipid solvents. They showed that about 40 per cent. of rat-liver RNA was extracted when ethanol was applied to liver samples after precipitation with 5 per cent. trichloroacetic acid, whereas after treatment with 10 per cent. trichloroacetic acid the loss is only about 10 per cent., and after using 15 per cent. trichloroacetic acid it fell to 5 per cent.Hallinan, Fleck and MunroZO also demonstrated that a similar effect occurred when tissues precipitated with perchloric acid were subsequently extracted with ethanol. After precipitation with 0-2 N perchloric acid, ethanol extracted about 25 per cent. of the RNAFebruary, 19661 MEASUREMENT OF NUCLEIC ACIDS IN BIOLOGICAL MATERIALS 81 from liver samples, whereas after treatment with 0.7 N perchloric acid, the loss into ethanol was small. However, when perchloric acid is used, degradation of RNA by the cold perchloric acid becomes significant at concentrations above 0.3 N, so that the reduced loss into lipid solvents resulting from higher acid concentrations is offset by degradation at the stage of cold-acid precipitation. As pointed out in the earlier review, Marko and Butler33 observed that DNA was often degraded if the tissue samples were treated with cold trichloroacetic acid and then extracted with hot lipid solvents, owing to the retention of acid by the lipid solvents.This effect could be prevented by buffering the first lipid solvent with potassium acetate. Steele, Okamura and BuschN have shown that this procedure can be successfully used to prevent loss of RNA into lipid solvents after treatment of the tissue with cold acid. They recommend using ethanol buffered with 2 per cent. sodium acetate as the first lipid solvent after precipitation with perchloric acid, and using ethanol containing 10 per cent. potassium acetate as first lipid solvent after precipitation with trichloroacetic acid. I t has been found by our colleagues that 1 per cent.potassium acetate is effective in preventing losses of RNA into ethanol from samples of adrenal gland35 and kidney (Halliburton, unpublished results) after treatment of these tissues with cold 0.2 N perchloric acid. Investigators who choose to use acetate buffers in this way should ensure that they are using them under optimal conditions, since these have not yet been fully defined. I t would therefore be desirable for the analyst to check recoveries of RNA following the addition of various concentrations of acetate buffer to ethanol. As pointed out earlier, the use of lipid solvents is usually not obligatory, and the analyst should therefore consider using procedures such as described by Fleck and M ~ n r o ~ ~ and Fleck and Begg37 in which lipid extraction is not part of the technique.However, if it is thought desirable to remove lipids, as with plant nuclei38 which contain an ultraviolet-absorbing lipid component, there are two possible procedures. First, the investigator can use cold acid to extract small molecules and then use ethanol buffered with acetate for the initial step in the removal of lipids. Alternatively, he can carry out lipid extraction as his first step, with subsequent cold-acid treatment. The latter approach was first described by Ogur and Rosen5 and has since been used by several other investigators, mainly working with plants, whose findings were reviewed in the earlier artic1e.l The use of lipid solvents as the initial step can be combined with inactivation of enzymes by first immersing the tissue in boiling ethanol ; this procedure has been adopted in several recent investigations.ll J2 J3 914 METHODS OF DETERMINING NUCLEIC ACIDS IN THE TISSUE RESIDUE 1.THE PROCEDURE OF SCHMIDT AND THANNHAUSER THE USE OF ALKALINE HYDROLYSIS TO SEPARATE RNA FROM DNA- In this method, the tissue residue is digested in alkali, which hydrolyses the RKA to products that are no longer precipitable on acidification. The DNA resists attack by alkali and is consequently precipitated when the digest is acidified. As pointed out in the earlier review,l a considerable variety of conditions of alkaline hydrolysis has been used. The conclusions reached at that time remain valid, namely- (a) Incubation in 0.3 N alkali for 1 hour at 37" C is adequate to extract all the RNA from mammalian tissues in an acid-soluble form, although these conditions do not degrade the RNA quantitatively to mononucleotides. ( b ) More prolonged alkaline incubation and the use of stronger alkali have the dis- advantages of rendering acid-soluble increasing amounts of tissue protein and, with N alkali, leading to de-amination of cytidylic acid. These effects can interfere seriously with subsequent measurement of RNA by ultraviolet absorption.(c) There is good reason to believe that DNA is not degraded under the usual conditions of alkaline incubation. Consequently, we recommend a 1-hour period of digestion in N potassium hydroxide at 37" C for mammalian tissues. IJnder these circumstances the RNA can be determined by measuring the ultraviolet absorption of the acid-soluble fraction obtained on acidifying the dige~t.369~7 For plant tissues, two 1-hour periods39 and a 3-hour period38 of digestion in alkali have been found necessary to extract all the RNA.82 MUNRO AND FLECK: RECENT DEVELOPMENTS IN THE [AWZjJl'st, VOl.91 THE RNA FRACTION OF THE ALKALINE DIGEST- The original Schmidt - Thannhauser determination depended on measurement of RNA as phosphorus. However, since only 80 per cent. or less of the phosphorus in the RNA fraction can be accounted for by ribonucleotides,l this method of assessing RNA content is not generally used. RNA has frequently been measured by ribose determination, usually by the orcinol reaction. The conditions for this determination have been fully discussed in the earlier review.l New colorimetric reactions for pentoses are occasionally described, such as those with dimethyl- phenol and chloro-m-creso140 and with benzofurane derivative^.^^ The first two reactions have been used as a basis of methods for RNA determination, but the last has not so far been used €or this purpose and might warrant further exploration.As pointed out in the earlier review,l many carbohydrates can interfere with colorimetric reactions for ribose. With animal tissues, the error is usually not serious, but particularly with plant tissues and yeast, the use of these reactions to determine the RNA in the untreated acid-soluble fraction of the digest is commonly subject to gross errors. Smillie and K r ~ t k o w ~ ~ and de Deken- Grenson and de D ~ k e n ~ ~ recommend using ion-exchange resins to separate the ribonucleotides from orcinol-reacting contaminants.Subsequent workers43 have also found this procedure t o be advantageous. The characteristic intense ultraviolet absorption of RNA has been widely used for determining ribonucleotides in the acid-soluble fraction of the alkaline digest, It is, however, subject to error when appreciable amounts of protein-degradation products are released by prolonged alkaline digestion. The error due to the ultraviolet absorption of peptides present in the RNA fraction can be dealt with in three ways. First, the period of digestion can be limited to 1 hour in 0.3 N potassium hydroxide at 37" C as recommended earlier. These conditions of digestion have generally been found to cause minimal solubilisation of tissue protein.36 There are, however, some exceptions. With the thyroid gland, even this short period of digestion results in significant peptide contamination of the RNA fraction.44 Barker and Hollin~head~~ found an appreciable contamination with peptide material when they digested plant tissue with alkali for two periods of 1 hour; this may have arisen through their use of acid ethanol to precipitate the DNA-protein fraction, since some tissue protein is soluble in this s0lvent.~5 Alternatively, one can correct the observed ultraviolet absorption of the RNA fraction for the presence of peptide material in one of two ways.First, the protein can be deter- mined directly46 and a suitable adjustment made in the observed ultraviolet absorption of the acid-soluble fraction.36 939 Secondly, measurements of ultraviolet absorption can be made at two wavelengths, one of which is that of maximum absorption of RNA, namely 260 mp.It has been common practice to select a second wavelength around 280 mp, since this is in the region of maximum absorption of protein. Tsanev and T~4arkov~~ explored this approach extensively and describe a procedure claimed to correct for the extensive protein contamina- tion of the RNA fraction that occurred after 18 hours of alkaline digestion. El~ewherel,~~ we have shown that this procedure results in considerable errors owing to the difficulty of standardising the protein correction. Fleck and Regg37 have recently shown that 232 mp, the absorption minimum in the RNA spectrum, is a sensitive point at which to determine the presence and amount of protein contamination. These auth0rs,~7 as well as Tsanev and M a ~ k o v , ~ ~ Fleck and R1unr0~~ and Warburg and Christian,48 provide a discussion of the principles and correct application of two-wavelength procedures in the analysis of mixtures. I n order to use such methods, the ultraviolet-absorption characteristics of the tissue RXA must be used, not that of another species of RNA, such as yeast RXA.47 Thirdly, the ultraviolet absorption due to protein can be eliminated on separating the ribonucleotides from contaminant peptides by passage through an ion-exchange resin, as recommended by Smillie and K r ~ t k o w ~ ~ and by de Deken-Grenson and de Deken.42 Several recent investigators measuring RNA in plant tissues have found this method to give reliable ~ a l u e ~ .~ ~ ~ ~ ~ ~ ~ ~ The peptides have also been removed with charcoal,51 but with plant-issue digests this procedure has been found unsatisfactory (Holdgate and Goodwin ; private com- munication). Finally, the nucleotides can be separated from the contaminating peptides by ele~trophoresis5~ or possibly by paper ~hromatography.~~ THE DNA FRACTION OF THE ALKALINE DIGEST- Difficulties have frequently been encountered in determining the DNA contained in the precipitate formed on acidifying the alkaline digest. The DNA can be determined from itsFebruary, 19661 MEASUREMENT OF NUCLEIC ACIDS IN BIOLOGICAL MATERIALS 83 phosphorus content by deoxyribose assay, by its ultraviolet absorption or by some miscel- laneous methods.As pointed out in the earlier review,l determination of the total phosphorus content of the precipitate has usually given acceptable values for the DNA content of the tissue, provided that phospholipids have first been removed. However, with nervous tissue, even defatting leaves phospho-inositides which increase the phosphorus content of the DNA Measurement of DNA by phosphorus determination has the advantage that the whole pre- cipitate can be digested and used for the determination. Other methods of measuring DNA, such as colorimetric determinations of deoxypentose or ultraviolet-absorption measurements, first involve either dissolving the precipitate in alkali or extracting the DNA by means of hot acid or enzymes.The DNA-protein precipitate can usually be rapidly and completely dissolved in 0.3 N potassium hydroxide at room temperature or at 37" C. This solution will clearly also contain the tissue protein which can interfere with DSA determination, especially by ultraviolet absorption. Consequently, many investigators extract the DNA with hot acid, usually either with 5 per cent. trichloroacetic acid for 15 minutes at 90" C, or with 0.5 N perchloric acid for 10 minutes at 80" C.l However, as pointed out in the earlier review, these extraction procedures may either be insufficient to remove all the DNA, or, if too vigorous, can lead to destruction of deoxypentose and to extraction of significant amounts of protein degradation products by the hot acid. This is the dilemma also inherent in Schneider's method for extracting nucleic acids from tissues with hot trichloroacetic acid or perchloric acid, which will be discussed later.Few investigators have studied in detail the optimum conditions for extraction of DNA with hot acid. Recently, Threlfall (private communication) obtained maximal recoveries of deoxyribose from liver samples by heating in N perchloric acid at 70" C, but the optimum temperature for kidney samples was found to be 65" C. At 85" C, destruction of deoxyribose was serious, indicating the importance of carefully controlled conditions. Wannemacher et report a study in which they extracted tissue samples with 0.5 N perchloric acid at various temperatures and examined the recoveries of DNA by measurement as deoxyribose, as phosphorus and by ultraviolet absorption.Optimal recovery was obtained after heating for 45 minutes at 96" C. However, as will be discussed in more detail under the heading of the Schneider procedure, Lprvtrup and R W S ~ have demonstrated that the apparent maximal extraction is, in reality, a balance between incomplete extraction and de&-uction of DNA and therefore does not imply full recovery of DNA. A few investigators have attempted to extract DNA from tissue residues by using deoxyribon~clease~~ or a combination of deoxyribonuclease and phosphodiestera~e.~~ The fraction is first incubated with the enzymes and then the protein is removed by precipitation with acid, the products of DKA hydrolysis being left in solution. This procedure can obviously be applied to the DNA-containing fraction of the Schmidt - Thannhauser procedure.How- ever, in a few preliminary studies on rat-liver samples we have not found conditions under which adequate extraction of DXA can be attained; a similar failure to extract all the DNA enzymically from chloroplasts has been recorded by Kirk.58 The DNA contained in the fraction dissolved in alkali or in the extract has frequently been measured by reactions for deoxypentose. In the earlier review, a wide variety of procedures for determining deoxypentose was discussed. In practice, only the Burton59 modification of the diphenylamine method and the indole method of Ceriotti60 are extensively used nowadays. Croft and Lubran6l describe a form of diphenylamine reaction based on Burton's procedure, but having greater sensitivity and not subject to interference by sialic acid.Giles and ;Llyers62 have also improved the sensitivity of the Burton method by 70 per cent. and have reduced the absorption of the reagent blank to one-third after an investigation of the optimal proportions of reagents. Schmid, Schmid and B r ~ d i e ~ ~ have explored optima! conditions for the indole reaction. It was originally claimed by CeriottiG0 that his form of the indole procedure was subject to interference only by arabinose, which is not likely to be present in the DNA fraction of the Schmidt - Thannhauser method. However, glycoproteins and sialic acidel have recently been found to cause erroneously high values by the Ceriotti method. Samples of thyroid glandM have been shown to give an unsatisfactory absorption spectrum for the coloured complex formed in the Ceriotti procedure; this is presumably due to the carbohydrate contained in thyroglobulin.As to determination of DNA by ultraviolet absorption, the solution of the DNA-con- taining precipitate in alkali includes such a large amount of protein that it is not practicable84 MCNRO -4ND FLECK: RECEXT DEVELOPMENTS IN THE [Amdyst, vol. 91 to measure the DNA with adequate accuracy, even when a correctly designed two-wavelength method is applied. Consequently, many authors have extracted the DNA with hot trichloro- acetic acid or perchloric acid, as described in the earlier review.l The problem involved in this approach is to secure adequate extraction without also solubilising some of the tissue protein that absorbs ultraviolet light.In consequence, Scott, Fraccastoro and TaftG4 describe conditions that are a compromise between these opposing effects. Wannemacher, Banks and W ~ n n e r ~ ~ used 0-5 N perchloric acid to extract the DXA and compared the ultraviolet absorp- tion at 265 and 290mp in order to detect the presence of peptide. At temperatures of extraction up to 96" C, the ratio of 265 to 290 was constant, but at higher temperatures a fall in the ratio was interpreted as being due to peptide contamination. Trace amounts of protein were detected at lower temperatures by the Lowry method.46 A more sensitive method of detecting contamination by the products of protein degradation would be to examine the ultraviolet spectrum of the extract in the region of the absorption minimum of pure DNA (around 233 mp), as suggested for RNA by Fleck and Begg.37 A correctly designed two-wavelength procedure should provide an accurate procedure for correcting the hot acid extract for peptide contaminants, provided that appropriate standards for tissue DSA and contaminant protein can be prepared.65 This approach was taken by Santen and Agranoffg when measuring brain D1JA4; however, their use of salmon-sperm DNA as a standard may have led to erroneous values.In the earlier review1 we referred to two sensitive fluorimetric methods for measuring DSA4.6C 9 G 7 Although no further methods for Auorimetric measurement of DXA have since been described, new fluorescent reactions have been reported for adenine,68 and after treatment of RNA and DNA with berberineC9; these might become the basis of quantitative procedures.The high sensitivity of fluorescent methods has been utilised recently by workers studying the DlVA content of ova and e m b r y o ~ . l ~ 9 ~ ~ DSA can also be determined as nucleotides or bases, as discussed in the previous review.1 Recent advances in the chromatography of these products, including notably the use of thin-layer chromatography6 and of gas chrornat~graphy,~~ increase the potential usefulness and sensitivity of this approach. Isotope-dilution methods for measuring DNA, mentioned in the earlier review, appear not to have been used during the past few years. A review of the measurement of DNA by microbiological methods was provided recently by Lervtrup and Roos.71 RECOMMENDATIONS FOR THE USE OF THE SCHMIDT - THANNHAUSER METHOD- In view of the studies of Hallinan, Fleck and Munro,20 the general application of lipid solvents in the preliminary treatment of the tissue is no longer considered desirable.Further, we now prefer to use cold 0.2 N perchloric acid rather than 10 per cent. trichloroacetic acid for the initial inactivation of enzymes and precipitation of the nucleic acids, especially when the RNA content of the tissue is to be measured by ultraviolet absorption. The digestion of the samples in 0.3 N potassium hydroxide for 1 hour a t 37" C is satisfactory for mammalian tissues. These various recommendations and modifications are incorporated in the Schmidt - Thannhauser procedure described by Fleck and 111unr0~~ and by Fleck and Begg.37 Our final procedure is set out below in a form suitable for the analysis of rat-liver samples.The tissue is homogenised in 19 volumes of ice-cold de-ionised water, and 5 ml (zz 250 mg wet weight of tissue) are transferred by pipette into a 15-ml centrifuge tube. To this are added 2-5 ml Qf ice-cold 0.6 N perchloric acid. After thorough mixing and standing for 10 minutes in ice, the tube is centrifuged, the supernatant (acid-soluble) fraction is dis- carded and the precipitate is then washed twice with ice-cold 0.2 N perchloric acid. Excess of acid is finally drained off by inverting the tube briefly on to filter-paper. After incubation for 1 hour at 37" C (air-oven or water-bath), the digest is cooled in ice, and the protein and DNA are precipitated by adding 2-5 ml of 1-2 N perchloric acid.After 10 minutes' standing in ice, the precipitate is separated centrifugally and washed twice with 0-2 N perchloric acid. The supernatant fluid from the first centrifugation and the washings are combined, 10 ml of 0-6 N perchloric acid are added, and the solution is made up to 100 ml with de-ionised water to give a final solution of 0.1 N perchloric acid. The ultraviolet absorp- tion of this solution at 260 mp (and 232 mp also, if the absence of protein is to be checked) is then measured. An extinction of 1.000 at 260 mp is given by a concentration of 32 pg of The recommendations made in the earlier review1 remain valid in general. To the tissue residue 4 ml of 0.3 N potassium hydroxide are added and mixed.February, 19661 MEASUREMENT OF NUCLEIC ACIDS I N BIOLOGICAL MATERIALS 85 RNA per ml for rat liver.The original figure of 35 pg of RNA per ml given by Fleck and Munro36 is wrong because of an error in calculation. The precipitate containing the DNA is dissolved in 5 ml of 0.3 N potassium hydroxide, if necessary by warming to 37" C for a few minutes, and a further 12 ml of 0.3 N potassium hydroxide are added. The solution is then made up to 50 ml with water to give a solution of DNA in 0.1 N potassium hydroxide; 2-ml samples of this solution are taken for DNA estimation by the Ceriotti indole procedure60 or by Burton's modification59 of the diphenyl- amine method. Analysts using the latter procedure should consult the papers by Croft and Lubran61 and by Giles and Myers62 for further improvements in this method.The procedure described above can be used for analysing samples containing 0.7 to 3.5 mg of RNA and 0.2 to 1 mg of DNA. It can readily be scaled down to measure one-tenth or less of these amounts by reducing the volumes in which the RNA and DNA fractions are made up. 2. THE SCHNEIDER PROCEDURE In the original Schneider procedure,* the nucleic acids are extracted from the tissue residue after the removal of small molecules and lipids by heating in 5 per cent. trichloroacetic acid at 90" C for 15 minutes. The RNA is then measured in the extract by the orcinol method and the DNA by the diphenylamine reaction. Since the original description of this method was published, extraction of the nucleic acids has also been commonly carried out with hot perchloric acid.As pointed out in the earlier review,l both acids have been used at a wide variety of concentrations, temperatures and duration of extraction by investigators who did not appreciate that the conditions of acid extraction were critical. Hutchison, Downie and Munro30 showed that for liver samples extracted with perchloric acid at 70" C, increasing the concentration of the acid resulted in excessive amounts of orcinol-reacting material appearing in the extract. M'ith DNA, two opposing effects were observed. At low concentrations of perchloric acid the DNA was incompletely extracted, but at higher concentrations, especially when the temperature was raised to 90" C, there was destruction of deoxyribose.No condi- tions of hot-acid extraction were found that gave full recovery of the DNA of the tissue, as compared with the values obtained for the Schmidt - Thannhauser method. These findings have been confirmed and amplified by Lovtrup and R o o s ~ ~ ~ ~ ~ ~ ~ ~ in a series of careful studies of the kinetics of DNA extraction and deoxyribose destruction by hot 0.5 N perchloric acid at various temperatures. On the basis of these kinetic studies, they offer a method of deter- mining the tissue DNA, after treatment at 90" C for 1 hour in 0.5 N perchloric acid. The deoxyribose value obtained for the extract is corrected to give 100 per cent. recovery by calculations derived from the kinetic studies. They found that the correction equations differed for each tissue studied, and in consequence no single set of universally applicable conditions can be recommended.Therefore, the use of the Lovtrup - Roos correction method has to be worked out for each tissue independently and would be tedious for ordinary manual determination. However, since the Schneider method has now been made the basis of an automated analytical method for measuring the RNA and DNA in bacteria,74 the incorporation of the correction for deoxyribose loss can be readily carried out as part of the analytical procedure. Automated analysis may thus restore the attractiveness of the Schneider pro- cedure, in spite of the inherent errors in extraction. This does not, however, guarantee the exclusion of errors in measurement of RNA and LISA due to solubilisation by hot acid of tissue constituents that give the reactions for ribose and deoxyribose.The earlier review1 contains numerous examples of erroneous values for KNA and DNA by sugar determination due to this cause. Consequently, the analyst who is embarking on the use of sugar reactions for automated analysis with a new tissue should first examine the spectrum of the coloured product obtained with the tissue extract and compare it with the corresponding spectrum for pure ribose or deoxyribose. 3. THE PROCEDURE OF OGUR AND ROSEN Ogur and Rosen5 described a procedure in which the RNA is extracted with cold per- chloric acid; DKA is subsequently extracted with hot perchloric acid. The ultraviolet absorption of each extract is then used as a measure of the RNA and DNA contents, respec- tively, of the tissue.In the previous review,l numerous reports of incomplete separation of RNA from DNA by this procedure were quoted. During the last few years the method86 MUNRO AND FLECK: RECENT DEVELOPMENTS I N THE [Andyst, Vol. 91 has been used on several occasions, on some of which the authors report unsatisfactory r e ~ ~ l t ~ . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Consequently, we see no reason to modify the conclusion reached in the earlier review that this method is inadequate for the accurate determination of nucleic acids in biological materials. 4. OTHER PROCEDURES In the earlier review,l several miscellaneous methods for nucleic acid determination were discussed. There is little to add to the comments made there. Claims have been made that extraction of DNA from tissues by the phenol procedure can be made quantitati~e.~89~~ Samis, Wulff and FalzoneS0 explored methods of recovering nucleic acids quantitatively after phenol extraction.They report the effects of ethanol concentration on precipitation of nucleic acids and conclude that it is difficult, with this method, to obtain satisfactory quantitative recoveries. Accordingly, they investigated the use of indium salts and were able to obtain conditions under which both nucleic acids were completely precipitated. The precipitate was then treated with alkali to dissolve the RNA and DNA, the indium being discarded as the insoluble hydroxide. The nucleic acids could then be separated by incubation in the alkaline medium, as in the Schmidt - Thannhauser procedure. This approach may be worth further exploration.In less frequently used methods for determining the extracted nucleic acids, recent advances in thin-layer chromatography6 and in gas chr~matography~~ may find applications to quantitative measurements of bases, nucleosides or nucleotides of RNA and DNA. The application of microbiological assays to measurement of nucleic acids has been reviewed by Lprvtrup and R o o s . ~ ~ GENERAL RECOMMENDATIONS FOR NUCLEIC ACID DETERMINATION It was concluded in the earlier review1 that of the three major methods of nucleic acid estimation, the Schmidt - Thannhauser procedure is least subject to analytical error. This conclusion has been strengthened by a study of the recent literature, and we offer in an earlier section of this review an outline of a procedure found satisfactory for many animal tissues.However, investigators continue to report new and unsuspected sources of error in the standard methods of analysis, as for nervous t i ~ s u e , ~ > 8 y 9 thyroid gland,44 gastric mucosa61 and For plant tissues, considerable analytical errors have been encountered. Smillie and K r ~ t k o v ~ ~ made an extensive study of analytical methods for measuring nucleic acids in plant tissues and concluded that the Schmidt - Thannhauser procedure gave the most satisfactory results, provided that contaminants are removed from the RNA fraction by passing it through an ion-exchange resin. Subsequent authors studying nucleic acids39 lg2 ~8~ from plant tissue have confirmed the importance of resin treatment in order to obtain valid results for RNA content by phosphorus or orcinol determination or ultraviolet absorption.Barker and H~llinshead~~ used a short period of incubation in alkali in order to minimise contamina- tion of the RNA fraction with protein. Nevertheless, they found significant amounts of protein in this fraction; they were able to correct for the ultraviolet-absorption error due to this material on determining the amount of protein present by the Lowry method46 and then applying a correction factor to the ultraviolet absorption. In this way they obtained results that were in good agreement with those obtained by treating the RNA fraction with an ion-exchange resin. I t should be pointed out that the unusually large amount of protein made acid-soluble by the short period of alkaline digestion may have occurred because they used acid ethanol to precipitate the Dh’A and protein at the end of digestion; it has been shown that acid ethanol extracts a considerable amount of tissue proteing5 that would subsequently appear in the RNA fraction.From these reports of difficulties in the application of standard nucleic acid procedures to various tissues, it is apparent that no single method can be expected to be universally applicable. Consequently, the analyst making a new study of the nucleic acid content of a tissue should be on his guard against sources of error, and should provide evidence that his method of choice has been rigorously tested. REFERENCES 1. 2. Hutchinson, W-.C., and Munro, H. N., Analyst, 1961, 86, 768; 1962, 87, 303. Munro, H. N., and Fleck, A., in Glick, D., Editor, “Methods of Biochemical Analysis,” Interscience Publishers, a division of John Wilcy & Sons Inc., New York, London and Sydney, 1966, Volume XIV, in the press.February, 19661 MEASUREMENT OF NUCLEIC ACIDS IN BIOLOGICAL MATERIALS 87 3. 4. 5. 6. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 60. 51. c 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. Schmidt, G., and Thannhauser, S. J., J . Biol. Chem., 1945, 161, 83. Schneider, W. C., Ibid., 1945, 161, 293. Ogur, M., and Rosen, G., Archs. Biochem., 1950, 25, 262.Saukkonen, J. J., Chromat. Rev., 1964, 6, 53. Logan, J . E., Mannell, W. A., and Rossiter, R. J., Biochem. J., 1952, 51, 470. Zamenhof, S., Bursztyn, H., Rich, K., and Zamenhof, P. J., J . Neurochem., 1964, 11, 505. Santen, R. J., and Agranoff, B. W., Biochim. Biophys. Acta, 1963, 72, 251. Markham, R., in Paech, K., and Tracy, M. V., Editors, “Moderne Methoden der Pflanzenanalyse,” Springer-Verlag, Berlin, 1955, Volume IV, p. 246. Stange, L., Kirk, M., Bennett, E. L., and Calvin, M., Biochim. Biophys. Acta, 1962, 61, 681. Giles, K. W., and Myers, A., Ibid., 1964, 87, 460. Icessler, B., and Engelberg, N., Ibid., 1962, 55, 70. Baltus, E., and Brachct, J., Ibid., 1962, 61, 167. Kovacs, E., Archs. Biochem. Biophys., 1958, 76, 546. Ledoux, L., Galand, P,, and Huart, R., Biochim.Biophys. Acta, 1962, 55, 97. May, L., and Grenell, R. G., Proc. SOC. Exp. Biol. illed., 1959, 102, 235. Smith, 0. K., Yale J . Biol. filed., 1953, 26, 126. Currie, R., Nature, 1965, 205, 1212. Hallinan, T., Fleck, A., and Munro, H. N., Biochim. Biophys. Acta, 1963, 68, 131. Kuroima, Y., and Hashirnoto, N., Bull. Agric. Chem. SOC. Japan, 1960, 24, 547. Katchman, B. J., and Fetty, W. G., J . Bact., 1955, 69, 607. Smillie, R. M., and Krotkov, G., Can. J . Bot., 1960, 38, 31. Finamore, F. J., and Volkin, E., J . Biol. Chem., 1961, 236, 443. Finamore, F. J., Ibid., 1964, 239, 1882. Levy, H. B., and Lynt, R. K., Biochim. Biophys. Acta, 1963, 72, 529. Nekhorocheff, J., and Cantan, B., Bull. SOC. Chim. Biol., 1964, 46, 805. Venkataraman, P.R., and Lowe, C. U., Biochem. J . , 1959, 72, 430. Venkataraman, P. R., Biochim. Biophys. Acta, 1960, 39, 352. Hutchison, W. C., Downie, E. D., and Munro, H. N., Ibid., 1962, 55, 561. Greenbaum, A. L., and Slater, T. F., Biochem. J . , 1957, 66, 155. Cooper, W. D., and Loring, H. S., J . Biol. Chem., 1957, 228, 813. Marko, A. M., and Butler, G. C., Ibid., 1951, 190, 165. Steele, W. J., Okamura, N., and Busch, H., Biochim. Biophys. Acta, 1964, 87, 490. Imric, R. C., Ramaiah, T. R., Antoni, F., and Hutchison, W. C., J . Endocr., 1965, 32, 303. Fleck, A., and Munro, H. N., Biochim. Biophys. ,4cta, 1962, 55, 571. Fleck, A., and Begg, D. J., Ibid., 1965, 108, 333. McLeish, J., Proc. Roy. Soc., B, 1963, 158, 261. Barker, G. R., and Hollinshead, J. A., Biochem. J., 1964, 93, 78.Staron, T., Xuong, N. D., Allard, C., and Chambre, M. M., Hebd. Sdanc. Acad. Sci., 1962,254, 3048. Mikulaszek, E., Merkel, M., Osowiecki, H., and Zawadowski, T., Bull. Acad. Polon. Sci., B, 1963, de Deken-Grenson, M., and de Deken, K. H., Biochim. Biophys. Acta, 1959, 31, 195. Young, E. G., Can. J . Bot., 1964, 42, 1471. Begg, D. J., McGirr, E. M., and Munro, H. N., Endocrinology, 1965, 76, 171. Munro, H. N., and Downie, E. D., Archs. Biochem. Biophys., 1964, 106, 516. Lowry, 0. H., Kosebrough, N. J., Farr, A. L., and Randall, R. J., J . Biol. Chem., 1951, 193, 265. Tsanev, R., and Markov, G. G., Biochim. Biophys. Acta, 1960, 42, 442. Warburg, O., and Christian, W., Biochem. Z., 1942, 310, 384. Rottger, B., and Fritz, H. G., Biochim. Biophys. Acta, 1962, 61, 621. Marcus, A., and Feeley, J., Ibid., 1962, 61, 830. Kitagawa, T., Schmidt, G., and Thannhauser, S. J., “Abstracts of papers presented a t A.C.S. National Meeting, ” Division of Biological Chemistry, American Chemical Society, Washington, D.C., 1962, Abstract 701. 11, 565. Davidson, J. N., and Smellie, R. M. S., Biochem. J . , 1952, 52, 594 and 599. Gerlach, E., Dreisbach, R. H., and Deuticke, R., J . Chromat., 1965, 18, 81. Wannemacher, R. W., jun., Banks, W. L., jun., and Wunner, W. H., Analyt. Biochem., 1965,11,320. Lavtrup, S., and Roos, K., Biochim. Biophys. Acta, 1961, 53, 1. Haggis, A. J., Devl. Biol., 1964, 10, 358. Iwamura, T., Biochim. Biophys. .4cta, 1962, 61, 472. Kirk, J . T. O., Ibid., 1963, 76, 417. Burton, K., Biochem. J., 1966, 62, 315. Ceriotti, G., J. Biol. Chem., 1955, 214, 59. Croft, D. N., and Lubran, M., Biochem. J . , 1966, 95, 612. Giles, K. W., and Myers, A., Nature, 1965, 206, 93. Schmid, P., Schmid, C., and Brodie, D. C., J . Biol. Chem., 1963, 238, 1068. Scott, J. F., Fraccastoro, A. P., and Taft, E. B., J . Histochem. Cytochem., 1956, 4, 1. Fleck, A,, Ph.D. Thesis, University of Glasgow, 1963. Kissane, J. M., and Robins, E., J . Biol. Chem., 1958, 233, 184. Roberts, D., and Friedkin, M., Ibid., 1958, 233, 483. Estabrook, R. W., and Maitra, P. K., Analyt. Biochem., 1962, 3, 360. Yamagishi, H., .J. Cell Biol., 1962, 15, 589. Hancock, R. I,., and Coleman, D. L., Analyt. Biochem., 1966, 10, 365. Lavtrup, S., and Roos, K., Acta Biochim. Pol., 1963, 10, 73. -- , Biochim. Biophys. Acta, 1063, 68, 425.88 73. 74. 75. 76. 77. 78. 79. 80. 81. MUNRO AND FLECK [Analyst, Vol. 91 Lavtrup, S., Actu Biochim. Pol., 1962, 9, 411. Gerke, J. R., Watson, R W., and Umbreit, W. W., “Continuous Analysis of Microbial Nucleic Acids and Protein with an AutoA4nalyzer Instrumental System,” Paper presented a t the Technicon 25th Anniversary International Symposium, “Automation in Analytical Chemistry,” London, 1964. Brawerman, G., Pogo, A. O., and Chargaff, E., Biochim. Biofihys. Ada, 1962, 55, 326. Ambellan, E., Ibid., 1964, 80, 8. Ambellan, E., and Webster, G., Ibid., 1963, 68, 119. Colter, J . S., Brown, R. A., and Ellem, K. A. O., Ibid., 1962, 55, 31. Lyttleton, J . W., and Petersen, G. B., Ibid., 1964, 80, 391. Samis, H. V., Wulff, V. J., and Falzone, J. A., Ibid., 1964, 91, 223. Holdgate, D. P., and Goodwin, T. W., Phytochem. hTewsl., 1965, 4, 831. Received July 2211d, 1965