,Ttpa ON THE CHElbdXXL NATURE OF THE SUBSTANCE Le~UCmG TRANSFORMATION OF PNEUMOCOCCAL TYPES lI ETFEa OF DES~-O~~A% ON THE BIOLOGICAL A~MTY or THE ?-RAIvSFOmG &BSTANCE* BY iVfACL.YN MCCARTY, M.D., AM, OSWALD T. AVERY, M.D. tF,,,,,, tb u&ted States Nav hmwd Unit d the Hospitd oj The R~~ht?jelk In&& for bfediizd Resewch) (Received for publication, October 10,194S) ne subst.ame inducing transformation of pneumococcal types has been blated from pneumococcus `l&e .JII in the form of a desoxyribonucleic acid fraction (1). The data obtained by chemical, enzymatic, and serological ,a+$ as well as by electrophoresis and ultracentrifugation of the purified mted strongly suggest that the nucleic acid is itself responsible for the bi&giad aCthhy. ne e-tic analysis was begun early in the course of the attempts to determine the nature of the transforming substance. Relatively unpurified peumocorxal extracts were subjected to enzymatic study in the hope that by this approach some clue might be obtained as to the identity of the biologi- cally active constituent. Crystalline trypsm, chymotrypsin, and ribonuclease bad no effect on the transforming substance, but it was found that certain crude enzyme preparations were able to bring about complete loss of trans- forming activity. When the possible importance of desoqribonucleic acid wu suggested by chemical fractionation, the experiments with crude enzyme preparations were extended to determine whether their ability to destroy the activity of the transforming principle could be correlated with any enzymatic sction on authentic samples of desoxyribonudeic acid. of non-bacterial origin. A Variety of crude enzymes were tested both for their abiity to inactivate the trpDsforming substance, and for theii &ect on desoqribonucleic acid from mimal tissues (1). Of the enzyme preparations tested only. those capable of dcpolYme~g authentic desoxyribonucleic acid were eflective in destroying thC transforming activity of' pneumococcal extracts. Other parallelisms be- hpm the two actions were observed: for exam&, sodium fluoride inhibited both *e depolymetimg action and the inactivation of the transforming sub- `The Btmu of Medicfue and Surgery of the U. S. Navy does not nd~ u.u&ru& ta tadorse the views or opinions which are expreaxd in t+3 paper. 89 stance. Thus, iudirect of the thesis that the evidence was obtained by enzymatic analysis in supl,,,rt acid (1). active transforming substance is' a desoebouuchic Since no purified preparation of: desoxyribonuclease was available, pui&+ tioti of the enzyme was undertaken in this laboratory in order that the eaty. matic evidence concerning the nature of the transforming substance could k ma.de mom direct and conclusive;.-. The enzyme has been isolated from beef pancreas. in highly active form~,relatively free from other enzymes, and in properties have been investigated in some detail (2). It is the purpose of the present paper to describe the action of purified desoxyribonuclease on th biological-activity of the transforming substance of Pneumckoccus Type III. 3. `, `_ EXPEPJXfENTAi ' E$tit of Citrdc Inh&itti of the &sm.LIt has been found that dew ribonuclease from beef pancreas, as well as from all other sources studied, requites the presence of a metallic activator (2). Magnesium ion appear to be the naturally occurring activator, although manganese at the same molar concentration is equally &ective. It has been shown that sodium citrate is I potent inhibitor of magnesium-activated desoxyribonu~ease, presumably because citrate forms a soluble complex with magne-sium, and thus prevent8 activation of the enzyme. On the other hand, the manganese-activated CP zyme is not appreciably inhibited by the presence of citrate. Thus citrate inhibition is dependent upon the nature of the metallic activator (2). In order to determine whether citrate inhibits the action of desoxylibonuclease on the transforming principle, and whether a similar relationship exists with rcspcCt to the nature of the metallic activator, the following experiment was W ried out. A 2 mg. per cc. solution of ptied transformiug substance isolated from ~neumococlm Type III WI) ~88 P@ in 0.025 Y veronal buffer, pH 7.5. To two 0.5 cc. portiom d this sol~~cm 0.1 EC. of 0.03 16 MgSOr was added. To one of these tubes 0.1 cc. of 0.1 Y sodim citrate ~8s alsO added. TWO additional tubes were prepared containing the same UIIO~ td tramformlng substance, but using instead of MgSo, a similar amount ad concentrados d bfn%. AS in the first series, citrate was added to one of these tubes. The volume of aa of the four tubes was brought to 0.9 cc. with veroual buffer, pH 7.5, and 0.1 ix. of a soluw d partially purified desoxyribonuckase containing 0.001 mg. of proteziu was added to all tuba Thus, the hd CO~DU~SS~~OII of Mg* and I&* was 0.003 M, and the m concentirioaydd Citrate 0.01 Y. The -matic reaction was allowed to proceed at 30oC. for 15 minute% then was stopped by heating all tubes at 60oC. for 15 minutes. &rid tenfold dilud@ d the reaction mixtures were then made and tested for transforming activity according to * method of titration previously described (1). 0.2 cc. of each dilution was added to `Pd. mplicate tubes containing 2.0 cc. each of special serum broth. `ocbd The tubes were m with 0.05 cc. of a 10-' dilution of a broth culture of R cells (R36A) derived from Pneom@ Tp II. Aft= 20 ho& inc~bsti~n at 37'C., each culture was plated on blood agar~ ad& presence of Type III cells was determined by bacteriological and serological tests. TbereaJt' are recorded iu Table I. -uYN MCCARTY AND OSWALD T. AVERY 91 lt~beseenfromtheresult+s presented in Table I that in the presence of nw 0.003 M MgSOr or 0.003 Y MnSO+ the enzyme caused complete in- 4,3~on of the transforming substance. However, 0.01 M sodium citrate Wt,t about almost complete inhibition of the magnesium-activated eve, - Mg++ activated enzyme Mn++ divated -me Mgtt activated enzyme + sodium citrate Mn++ activated enzyme ., + sodium citrate .:. No enzyme (control) ' - - 10-l lo-' lo-' l(r 10-t 1fF 10- lo-4 IO+ 101 lo-' l(r lo-1 10-r IO- lo+ lO+ lo- l(P 1oJ - 1 R only* R sg R `( R I8 R only R (` R u R " sru* sm sm R only SIII SIII SIII SIII 1 R only R " R u R SC R only R (' R SC R a* SIII SIII SIII SIII R only R u R (' RU &II SUI SIII SIII 3 R only R `( R I` R M SIII SIII SIII R onIy SIII SIII SIII SIII 4 . R only R " R " R " SIII SIII SIII RonIy SIII SIII SIII SIII * "S III,' indicates the c&-urrence of transformation as evidenced by the recovery of en- capsulated cds of heumococcus Type III, while the term "R only" means that transforma- tion has not taken place, and only unencapsulated R variants were recovered. . and the titer df the t&form&g &&&& after' trek&e& with the enzyme in the presence of citrate and magnesium was only sli&tly lower than that of the control material. On the other hand, citrate had no apparent &ect on the manganese-activated enzyme, Ad the transforming substance was com- ple~lydestro,,~. ,,: `., _ `." -`:, : "' _; .' .t. -.::;. .`T. I. The above results parallel those obtaikd in a study of .the effect of citrate on the action of'the esmy& on' QsfxyrSmucleic acid from calf thyjnus (2) The data emphasize the close relationship between the abiity of the & preparation to inactivate the transforming substance on the one hand, a t,, depolymerize desoxyribonuckic acid'on tbkather. It therefore appears tbst the tie enzymatic process is involved in both instances, since in each case tin conditions of activation and inhibition of the enzyme are identical. Titration of the Activity af Pur@& DesoxyGb~&~~e.-3'ur&d prepamtir,a, of desoxyribonuclease from beef pancreas have no demonstrable phosphat+ hpase, or ribonuclease activity Q). _ Traces of a proteolytic enzyme are present, but in order to detect proteolytic activity an enzyme concentration of mm than 0.2 mg. per cc. must be used, while a definite measurable effect on tic `viscosity of sodium desoxyribonucleate solutions .can be demonstrated at con. centrations of less than 0.01 microgram per cc. (2). Thus, in terms of enzymatic activity the proteolytic enzyme is a minor contaminant and does not complicatl the interpretation of results obtained with appropriate concentrations of desoxyribonuclease. In the following titration of the activity of the enzyme on the transformiag substance, the concentrations of the various components in the reaction system were identical with those used in the standard method for measuring tbt activity of the enzyme on calf thymus nucleic acid `(2). However, the total volume of the reaction system was smaller, in order to conserve the available supply of the pneumococcal nucleate. Consequently, it was not feasible to measure quantitatively the fall in viscosity of the solution of transform@ substance. A 5 mg. per cc. solution of the d&d, puritied enzyme (preparation 16) was prepared ifs water. To obtain an appropriate concentration for the activity titration, a lO,OOO- fold dh- tion (containing 0.5 /ig per cc. of enzyme protein) of tbis enzyme solution was prepad is volumetric flasks. Further twofold dilutions were made and used in tbe test, Since it hu been shown that gelatin retards denaturation of the eizyme iu dilute solution (2), the dilU@ fluid consisted of an aqueous solution of 025 per cent gelatin and 0.075 Y MgSOt. 1.2 cc. samples of a 0.1 per cent solution of purified Type III transform&g substance (RCn in 0.025 LI vemd buffer, pH 7.5, were introduced into each of five tubes. 0.05 cc. of lk falliug twofold dilutions of enzyme was added to the kst four tubes. The dilutions Of the enzyme used contained 0.25,0.125,0.062, and 0.031 pg. per cc. so that the final enzyme cob centrations in the reaction system were 0.01, 0.005, O.CW5, and 0.00125 Fg. per CC- * tively. 0.05 cc. of the gelatin-MgSO, diluent alone wss added to tbe tit& tube which d as control. The enzyme was allowed to act at 30oC. for 30 minutes after which the tubes were hertrd in a water bath at 60oC. for 10 minutes to stop the reaction. Serial tenfold dilution Of tk reaction mixtures were then made in saline, and tested for transforming activity by the uarJ &xxdum The results are recorded in Table II. It can be seen from Table II that the enzyme in concentrations of 0.01 ssd 0.005 pg. per cc. completely destroyed the activity of the transforming a@s' MACLYN M`32.ARTY AND OSWiUD T. AVERY 93 in the 30 minute period. Inactivation was practically complete when 0.0025 ~ per cc. of enzyme was used, ad even with tbe smallest concentration of the a;tume (0.00125 Pg. per cc.) more than a 90 per cent reduction in the activiq ol tie transforming substance resulted. lt h apparent, then, thao the purified desoxyribonuclease is exceedingly rc~,,e jn destroymg the pneumococcal transforming substance. It is of in- TABLE II The A&m of Dti~ontu.kase on the Transforming ..Wstatue Tiaation of Activity of P&W Enzyme 0.0025 pg. per cc. 0.00125 reg. per cc. None (control) Dilutfoa of T&q& 10-t 109 10-a 10-i lO+ l(r lo-1 lo-' lO+ lo-~ la? 101 lO+ 10-t IO+ 1oJ l(r T 1 R only+ R " R `( R only R *( R `( R only R `I R " SIII R only R `L R " SIII SIII SIII SIII Quadmpliate tats 2 R only R I` R `L R only R " R " R only R " R Ld SIII SIII R only R " SIII SIII SIII ROY 3 R only R I` R " sIII* R only R " SIII SIII R only R " SIII SIII SIII R only 4 R only R (` R (' R only R " R " R only R " R " SIII SIII R only R " SIII SIII SIII R only o Symbols same as in Table I. terest that the titration reveals almost tbe same end-point as that obtained when the activity of the identical preparation is measured by the viscosimetric , method on calf tbymus desoxyribonucleic acid (2). The fall in viscosity of animal nucleate solution in the presence of 0.00125 j4g. per cc. of enzyme, the smallest amount used in the above test, is slow but measurable, and approxi- mately a 10 per cent decrease in viscosity occurs in 30 minutes under the conditions of the test. For reasons stated above, the effect on viscosity was not measured quantitatively in the experiments using the transforming substance 94 lXANSFORMATION, OF PNEU3fOCWCAL .fYPES. II / t as sub&ate. ._ Howwer, the ,&al$kive tiect was readily mognk&le, snd in those-tubes .in which `complete enzymatic destruction of the transforming activity had occurred, there was a corresponding loss in viscosity when com- pared with the control tube containing no enzyme. , ..' :. :; DISCUSSION : The fact that a purified &par&ion if desoxyribonucle&e .m exceedingly low concentration is capable of destroying irreversibly the Type III trat&orr+ sub&ante provides strong confumatory evidence for the view that hiolo$~ activity is a property of the desoxyribonucleic acid. In this connection it h of interest that irreversible inactivation of the transforming agent by enzymatic digestion differs from that form of inactivation brought about by ascorbic add and certain related compounds, since in the latter instance, the reaction b reversible and full activity can be restored by the use of sulfhydryl ccm. `pounds (3). `I The possibility has heen recognized that the activity of the transfom@ agent might be referable to minute amounts of some other substance such BS protein in the purified preparations. The results of the present investigation show that in order to detect proteolytic activity, it is necessary to use an amount of puSed desoxyribonuclease 100,000 times greater than that required to cause rapid and complete destruction of activity of the transforming sub stance. This evidence, in conjunction with the data previously reported on the chemical and physical properties of the active principle, leaves little doubt that the abiity of a pneumococcal extract to induce transformation depends upon the presence of a highly polymerized and specific form of desoxyribosu- cleic acid, and that this constituent is the fundamental unit of the transformiog principle. The objection can be raised that the nucleic acid may merely serve BS s "carrier" for some hypothetical substance, presumably protein, which possesseS the spec& transforming activity. Depolymerisation of the nucleic acid would according to this hypothesis, destroy the effectiveness of the essential ~amfl and thus result in loss of biological activity. There is no evidence in favor of such a hypothesis, and it is supported chietly by the traditional view t-hat nucleic acids are devoid of biological specificity. On the contrary, there fl indications that even minor disruptions in the long-chain nucleic acid molecule have a profound effect on biological activity. Thus, treatment of the tranS- forming substance with concentrations of desoxyribonuclease. so small that odY a slight fall in viscosity occurs causes a marked loss of biological activity- It is suggested that the initial stages of enzymatic depolymerisation which are reflected only by minimal changes in the physical properties of the nucleate are sufficient to bring about destruction of specific activity. Although the results of enzymatic studies provide additional evidence for MACLYN MCCARTY AND OSWALD T. AVERY 95 the spe c$c r8le of desoxyribonucleic acid in pneumococcal transformation, Ihey throw no light on fhe possible chemical basis for this specikity. It re- ds one of the .challengmg problems for future research to determine what cOrt of codiguration~ or structural drfferences can be demonstrated between &,,Vribonu&ates of separate specificities. h this connection, it should be Pihtd out that m .a" probability only a relatively small number of the total . . molecule 111 an active prepafatmn of desoxyribonucleate from Pneumococcus Tj.l,c 111 are capable of mducmg transformation. This is suggested by the fact (h3t extnxtion of unencapstdated R pneumococci yields a similar desoxyribonu- cleate fraction which at present can be distinguished from the Type III material oaly by the fact that the former is inactive in the transforming system. It is #ble that the nucleic acid of the R pneumococcus is concerned with innurner- &le other functions Of the bacterial cell, in a way similar to that in which cap- sular development is controlled by the transforming substance. The desoxyri- baucleic acid from Type III pneumococci would then necessarily comprise not only molecules endowed with transforming activity, but in addition a variety of ohem which determine the structure and metabolic activities possessed in common by both the encapsulated (S) and unencapsulated (R) forms. If these considerations have any foundation in fact, the task of discovering the chemical basis of biological speciscity of desoxyribonucleic acids becomes atremely complex, since a given preparation will represent a mixture of a large sun&r of entities of diverse specificity. An example of a roughly analogous situation is aBorded by the gamma globulin fraction of immune sera. In this instance, although the preparation is chemically and physically homogeneous, it has been shown by immunological techniques to consist of a variety of antibody molecules of diverse specificity. Determination of differences in specificity of these various protein molecules which are chemically indistinguishable one from another is made possible only by the use of corresponding specific antigens, selectively reactive with homol- ogous antibody protein. In the case of deso@bonucleic acid, the techniques of~pneumococca.l tram+ formation provide at the present time the only available method for deter-n&- ing differences in selective activities and specificities of this biologically im- portant group of chemical substances. It has been shown that .extremely minute amounts of puri&d preparations of desoxyribonuclease are capable of briiqiug about the complete and irreversi- ble inactivation of the transforming substance of Pneumococcus Type III. The significance of the effect of the enzyme, and its bearing on the chemical pature of the transforming substance, together with certain considerations 96 TRANSFORMATTON bP.PNWBfO&cAL l&PEG. II concerning the bihgical specificity of desoxyribonucleic acids in general, a discussed. iIBLIOGRiPH+ '. 1. Avery, 0. T., MacLeod, C. M., `Ad McCarty,` M., J. Exp. Med., 1944,79,13X 2, McCarty, M., J. &+a. Physiol., 1946, !29, 123.. 3. McCarty, M., J. Exp. Med., 1945,81,501. :'