,T~~mS ON lJ!IE CHEMICAL NATURE OF THE glJJ3STANCE L\q~~ING TRANSFO~ATION OF PNEUMOCOCCAT, TYPES 111. .`a IarpRoVED &hmoD FOR TEE ISOLATION OF TIRE 'IYRAXQ~~MING sL3STA-a m ITS h?LItiTIoN To hmococms -Es n, m, AND VI* BY MACLYN M&ARTY, m.,' Lieutenad Cam&g, ddedti Corps, United States Naval Resene, ~~StiALDT.AVER+f.D.. +--. (pm the United States Navy Remarch Unit at the Hospital of The Rochefelkr In.&&?`, . for Medical Remuch) -.. ..a -(Received for publication, October 1~1945) h he first paper of thii series a method was described for obtaining the u;msfo&g substance of Pneumococcus Type III in the form of a desoxy- ,.ibonu&ic acid fraction (1). The yield of active material was not high, and for this reason a more efficient procedure has been devised which is based upon recently acquired knowledge concerning the properties of the transforming substance and the enzyme that destroys it. In addition, it has been possible to adapt the new method to the isolation and purification of the transforming .&&.nce from pneumococci other than those of Type III and to show that b this case also the active substance is contained in a desoxyiibonucleic acid fraction. . . ,-AL Preparation of Transforming Sdstuke from Pneunwc~~cus Type III.--In earlier work it was found necessary to heat-kill .the bacterial cells at 6Y'C. immediately after they had been collected by centrifugation (1) This step was required because of the presence in the pneumococcal cell of an enzyme which destroys the activity of the transforming substance. ' Consequently, if early inactivation of this enzym< was',neglected, the biological activity of the extract was greatly impaired or completely lost. On the other hand, cells heated at a temperature known to inactivate the enzyme are so altered that complete extraction of the active material is difficult as shown by the fact that the residual Type III cells after iextraction by the method previously Outhued still contain an appreciable amount of the active desoxyribonucleic acid fraction. . ( I .. _ ._ Evidence has accumulated which indicates'that the pneumococcal enzyme' I. *The Bureau of Medicine and Surge& of the U. S. Navy does not necessarily undertake ~e~dorse the views or opinions which are expressed in this paper. 97 98 TUNSFO~TION 6T &EUXOCOCCAL TYPES. IU , ., . responsible for the inactivation of the transforming substance is a desoxy. ribonuclease which attacks the highly polymerized form of desoxyribouudeic acid (l-3). Although no puritied preparation of pneumococcal desor,,. ribonuclease has been obtained, its enzymatic behavior appears to be analogous to that of the desoxyribon&lease'~repared from .beef pancreas, the propees of which have been described in `a separate communication (2). For example, the heat lability, optimum pI& and dependence upon activation by magnesium ion (or manganese ion) are the Same for the pneumococcal enzyme as for that obtained from the pancreas.' ,It has been shown that the enzyme from both these sources has little, if any, activity in the absence of magnesium ions, and that citrate acts as an effective inhibitor of desoxyribonuclease by virtue of its capacity to form a soluble complex with magnesium. Citrate inhibition of the pneumococcal enzyme forms the basis of the methoil to be descriid. The presence of citrate id sticient concentration to inhibit desoxyriboau. clease completely has no retarding action on autolysis of living pneumococcal cells. The effect of bile salts in promoting rapid lysis seems to be accelerated rather than retarded. Thus, by allowing lysis to proceed in the presence of desoxycholate and citrate, it is possible to release quickly the transforming substance from living cells and at the same time to preserve its full activity by inhibiting the destructive action of the enzyme. The procedure for ob- taining the active material in puriiied form from the lysate is described iu the following section. Pre#arntioc df&od.-The da from SO liters of a 12 hour culture of Pheumococcus `f@ ITI are collected in a steam-driven Sharples centrifuge as previously described (1). `lk packed cells are resuspended in soo cc. of a solution containing 011 Y sodium chloride aad 0.1 M sodium citrate. 5.0 cc. of 10 per cent solution of sodium dqcholate are added* Lysis begins ahnost immediately and within a few minutes the creamy suspension is chaoeal into a viscous, translucent solution and no formed elements can be found in tilms Prepand and stained by the Gram technique. After 30 minutes at room temperature, one-third v01~ of chloroform and one-tenth volume of amyl alcohol are added to the lysate and the mixture is shaken mechanically for % hour. The material is centrifuged and the supematant fluid is separated by pipetting from -the chloroform emulsion. Shaking with chloroform is re ptated two additional times after which the extract is quite clear and greenish-yellow in cfkrr* The addition to the extract of two volumes of alcohol with stirring results in the formsnos of two distinct types of precipitate, one &rely granular, and the other in the form of a hesvY mass of fibrous material which rapidly falls to the bottom. The less rapidly sedimaw granular precipitate is immediately decanted and the compact fibrous precipitate which coD tainq among other constituents, the active principle is removed and washed with akohoh A considerable amount of active material is entrained in the chloroform-protein gel pnd can be recovered in large part by extracting the combined gels with 300 cc. of the citrate-aslioc solution. After centrifugation this extract is also precipitated by alcohol and the fibmuJ precipitate is combined with that already obtained. The combined fibrous precipitates are dissolved in 300 cc. saline and the solution is &hen twice with chloroform and amyl alcohol. The solution is now clear and colorless Wi volumes of alcohol are added and the fibrous precipitate formed is again lifted out, wssh MACLYN MCCARTY AND OSWALD T. AVERY 99 si,, &,-,hol, and redissolved in 100 :O 150 cc. of 0.85 per cent Nag. 5 mg. of the puSed bctd~ enzyme capable of b~lyzlng aBe Trpe el[ cqfdar polysaccharide (4) and 10 mg. d cQ5&e ribonucleaw are added.; ?gestion 1~ carned out in a cellophane sac during @,+ at 37oC. against a solution conslstmg of equal parts of saline and 0.05 y venmal bufIer, pH).g. The decomposition of the caps* polysaccharfde is followed by serological tests, rJ pre,;o~Jly described (1). After drgestron has proceeded for 6 to 8 hours at 37T., the mtca is placed in the refrigerator and dialysis is continued in the cold overnight to complete Lhe I+mo~d of the dialyaable end-products of enzymatic action. The solution fs mixed with tfll) volmes of alcohol and the fibmus precipitate thus obtained is redissolved in 75 to 1~10 cc. oi tie. Final deproteinization 19 effected by 2 to 3 successive treatments with chloroform lad smyl alcohol in order to remove the added ensyme protein and any remaining traces of peumOCOCCal protein. At &is point the paddy purified material contains a desoxyribonu&ic acid and a sero- lkd~ active substance which has been identikd as the somatic C! polysaccharide. The c f,olysaccharide is apparently in a more highly polymerized form than that encountered in be oig~al procedure, and the alcoholic fractionation method previously described does not ,e,,srate it from the daxyribonucleic acid. However, use has been made of the fact that in the umsence of caki~m, desoxyribonuckic acid `is preeipitable by small amounts of alcohol, *tie be C poiysaccharide remains soluble under these conditions. Thus, an almost quanti- uuve separation as measured by serological techniques is readily elected. The procedure is &q lollers- rifter final deproteinization, as described above, one-tenth volume of 10 per cent &CL b s&&-l to the saline sOhn.iOn of partially purified trsnsforming substance. Upon the addition ,,f two-tenths volume of alcohol, the deso~bonucleic acid is precipitated and with it afl the transforming activity. The fibrous precipitate is washed in saline containing Cacls and &oh01 in the same concentration as the precipitating mixture, and redissolved in saline. pmcipitation by calcium and alcohol is repeated to remove the last traces of ,verologi&ly active c polywxharide. The fibrous precipitate is redissolved in saline and preserved by storage io the cold. Propertie.s of the Puri$ed iKate&d.-The properties of the desoxyribonucleic ncid thus prepared are the same as those reported previously for preparations obtained from heat-killed cells (1). Solutions are highly viscous, give a strong diphenylamine test for desoxyribose, and have a nitrogen and phosphorus content characteristic of nucleic acid. The material is serologically inactive when tested with high titer Type III anti-pneumococcal serum, mci gives negative qualitative tests for ribonucleic acid and `protein. ,Z'he transforming activity is of the same order as that of. preparations obtained by &e earlier method, and less than 0.01 pg. suffices to induce transformation. The chief advantage of the method ig that the yield of purified m&&l ig ~rkedly increased. Thus, 60 to 80 mg. Bre recovered from 50 liters of culture, representing a yield fivefold greatkr than that obtained by the previotis method. Isol&n of Transford~g Substance from Pnemwcocci of Types @her T.han Type III `. _`j. Attempts to obtain t&r transforming substance from pneumocoki of Types I, II, and XIV by the extraction of heat-&iIled.cells have yielded only relatively 100 TMNSFOFtMAlTON >OB' PNEIJ?.fOCOCC& TYPES.. m I I impure and weakly ,active extrac$.,y, It appeared likely, therefore, that lysh of living pneumococd in the presence of citrate would yield larger amounts of active material and afford a better `opportunity .of purifying the trs,nsfo~ substance. , jj'. A second difkulty encountered `m attempts to &&e the transfix : `. .L : material from other types, in a degree of purity comparable to that of at Type III substance, was the fact that no method was available for separating the capsular polysaccharide from the active material. In the case of Tyf-,r HI, the soluble bacterial enzyme which hydrolyzes the capsular polysacchsde served as a very e&&t tool for this purpose. However, since enzymes js purified form capable of decomposing the capsular polysaccharide of other types of pneumococci are not available, the separation of these carbohydrater from the active desoxyribonucleic acid fraction can be achkved O&TTY ch&. cal means. It has now been found that precipitation with calcium and alcohol, as descriied in the preceding section for the removal of the somatic C PC@. saccharide, can aho be used in the same way to remove the capsular poly- saccharides of certain types. For example, in the presence of calcium the capsular polysaccharides of Pneumococcus Types II and VI are not precipitated by low concentrations of alcohol which throw down &e desoxyribonucleic add fraction, and in these instances the transforming substance can be separated from the carbohydrate by this technique. On the other hand, with certain other types (e.g. Types III and XIV), the calcium salt of the capsular poly sac&ride precipitates with the desoxyribonucleic acid fraction and no repark tion can be achieved by this means, By using the procedures mentioned, i.e. lysis of living cells in the presenti of citrate and separation of the polysaccharides with calcium and alcohol, the transforming substance has been prepared from Type II pneumococd m s state of purity comparable to that of the Type III material. PrCparahn of the Transforming Subdance fk Pnsnrfn-~~ Ty#e II.-A stock strain d Pneumococcus Type II (D39) was used as source material. The cells from 50 liters of sa 8 hour culture are collected and the initial procedures involving the use of citrate in pr& an active lysate of the living cds an4 the technique of depro&jni&g the crude material thus obtained are carried out by the method described above for preparing the Type IIIsub* Since no purified preparation of the enzyme capable of decomposing the Type II cse polysac&ride is available, enzymatic digestion of the extract with ~bonuclease alone n carried out under the same conditions as those just described m the use of tbe combm+-j uT zymes in the Type III extract. Thfs procedure facixitates further purification, since the spfit products of ribonucleic acid are much more readily separable from the desoxyribonu~~c lad fraction than is tbe intact ribonuckic acid. After enzymatic treatment the mater% is e" one or more additional times tith chloroform and amyl alcohol to complete the deprot? tion. Two volumes of alcohol are then added to the protein-free solution, and the resulti fibrous precipitate is redissolved in 80 cc. saline. The solution of partially purified material is made up &i&y of desoxyrfbonud~c rdh Type II capsular polysaccharide, and the somatic C polysac&ride. In order to Wsrs" MMLYN MCCARTY AND OSW+LD T. AVERY 101 ,Jw pu~jc sdd from these mbohyhtes the calcium-alcohol method is employed w ft&ows: I.pon Ihead&cOn of qnytentpvolumeof 10 percent CaCLandtwo-tenthsvolume of ethyl . dbol, s fibrous prypl?fe 1s for!nd Which ests of the desoxyribonu&ic add. This dpimte is washed m salme c0nT.g 1 per cent CaCla and 20 per cent alcohol, redissolved z -dne, & further purified by PryPlaW a second time with calcium and alcohol. This d pmipit&e is redissolved m sahue. AS ix&&d by the diphenykmine test, do of the Jesor?+t,onu&ic acid is present.in the cslcium-alcohol precipitate. daon.~~n~&ic acid fraction ti be defined below. The prope&a of ho E&tiDmgs of th Crrl~i~?dkdd ikfdhod of Frdh.-The m&A rem-g h Je auvatant fluid after remOVal of the calcium-alcohol precipitate was recovered by tiding crcess of alcohol. The fraction thus obtained was redissolved in the me volume of afine u tit & in the solution of the desoxyribonucleic acid fraction, and 88 wfll be &own con- &ted largely of the cap;* po!y=ch=ide and ~0mat.i~ C carbohydrate. In order to demon- ,trate &e sharp separatmn aclue& by the use of calcium-alcohol fractionation, both of these ticom were tested for serological prope-rties and transforming a~titit~, wal dilutions of be two fractions were ma$ in saline and precipitin reactions using Type II anti-pnemo- - horse serum were uvned out by the techniques described in previous cowdations from w laboratory. The results are Presented in Table I. It can be seen from the resuhs of precipitin reactions presented iu Table I bat the material soluble in two-tenths vohme of alcohol in the presence of csc& reacted with immune serum when the sulution was diluted 1:1000, whereas in the same antiserum the calcium-alcohdl iqAuble fraction gave only s questionable reaction in the lowest dilution tested. From the results of the p&pitin tests it is evident, therefore, that a sharp and ahnost quantitat,&' separation of the desoxyribonucleic acid fraction from the serologically active polysaccharides is &ected by calci~-alcohol fractionation. Tests of the biolo&al activity of these same fractions in the. transforming system were carried out by the method ptiously describ+. The results of the transformation tests are shown in Table II. From the data recorded in Table II it is evident that the fraction containing the serolo&a& active material possessed no transforming activity even at the highest COnmtratiOn used &I the &her hand, the d~xyribonu&ic acid fraction was active in a l(r-r dilution.' The transforming tests in con- junction with the serological data emphasix the dectiveness of &e chemical procedure used in removing the polysaccharide fraction from the active trans- forming substance. ,,..*. It should be noted that the above experiment is not, strictly speaking, an e-pie of pneumococcus transformation since both the extract and the R Strain were derived from Pneumococcus Type II. This fact; however, does not ah the si~cance of the.results, since this particular R strain, w&h has been under continuous cultivation and study in this Iaboratory during the Pst *veml years, has never been observed. to revert spontaneousIy to iti ' ~~~ O-2 CC Of this dilution w-.&d& to 2 CC. of SIXUIII broth the final &uhn h the transforming system was approximately tenfold greatex than thatIndicated, 102 , .`- lXNVS.FO~~ON 02 PNEtl?bfOW~ TYPES. Xl origikd specific type. Indeed, alI attempts to cause it to revert by repeated animal passage. and by serial subcultures in anti--R serum have invar$blY failed. Moreover, revedon Canbe induced, as in thy present instance, o* when the specific transforming substance is used., *ABLE I..,- _ ,_ Swologkal Tests with F~pcfiorr Obtaikd by Prctipizoting with 02 VoJums of Akohoj ifi h Ph?.ret8ceofcaloirmrIol' ,, , Fd `., Initial dillJtion of frecthl 13 1:10 MO 1:mo lsoo 1:1m ---- Precipitateat 0.2 volumeof alcohol Very faint - - - - - (Deaoryribonucleic acid) trace . Material soluble at 0.2 volume of ++++ +++A +++ ++ + f alcohol ++++, pronounced precipitate with clear supernate. -, no precipitate. TABLE II Test of TIarrrforming Adidy of F~acth of Ty#s II E&ad Obtuimd by PIcciptihg with 0.2 Volume of AL&w1 in Presence of Calcium Ion Tnluformillg activity lhctiml Dilution* Qludmpliutc tests Precipitate at 0.2 volume of al- 101 S II sn SII L3x.l cohol (desoxyribonucleic acid) *l(r SII SII SII SII 101 SII Ronly SII R O~Y 1F R only R only R only R dy Material soluble at 0.2 &me of lo-' R mly R only R ouly R o&' alcohol (serologically active) 10-r R " R " R " R " Tubes in which only the R variant was recovered are designated as "R only.`! S n bdi* cates the occurrence of en~psulated Type II cells. * Dilution of fraction in saline. Se-e footnote 1. Proptdes of the Type I.. Tratzsfomitzg St&arm.---The general properties of the purified Type II material are in all respects identical with those of the Type III transforming substance, and the two can be differentiated odY On the basis of the specificity of their biological action. Qualitative chemical tests for protein are negative, and the material gives no appreciable reaction n-i& high titer Type II antisera. Elementary chemical analysis reveals a nitrogen content of 15.76 per @!' and phosphorus content of 8.50 per cent. The nitrogen-phosphorus ratio Is M.4~Y.N' hRC!ARTY AND OSWALD T. AVERY 103 thus 1.8j which is slightly higher than that of. most samples of the Type III tasfo&g substance. However, the figures fall well within the range of ,3,+tion encountered among different samples of desoxyribonucleic acid prepar~ from the same source by a single method. \vithout any modScation in preparative procedures, the method outlined ior obtiining p T? II trfmf0-g substance has also been successfully lpplied to the isolation of a biologically active desoxyribonucleic acid fraction from pneumococ~ Type VI. DISCUSSION It is of interest that a study of the properties of the enzyme which inactimtes he transforming substance has led to a more efkient method of obtaining the p&ed material from the pneumococcal cells. Although it is possible to inactivate the pneumococcal enzyme by heating the bacterial suspension, =traction of the active material from heat-hilled cells is diflicult and incom- plete. On the other hand, when release of the active substance is accomplished by @S~S Of the king ce& some hs of activity results from enZyXnatiC action even when the su$ension is rapidly lysed in the cold and then immediately heated as in the procedure described by Alloway (5). The Snding that the ewe in question is activated by magnesium and is effectively inhibited by citrate afforded a means of obtaining the transforming substance from autolyzed ,& without the use of heat. The effect of calcium on the alcohol precipitability of desoxyribonucleic acid has been used in the final puritkation of the active material. In the pres- ence of excess calcium ion the active desoxyribonucleic acid is completely precipitated by as little as two-tenths volume of alcohol. The highly polymeric pneumococcal polysaccharides have physical properties similar to desoxy- ribonucleic acid and thus are difficult to separate from it. However, it has been shown that the precipitability of certain of the polysaccharides is not affected by the presence of excess calcium, and sharp separation from the desoxyribonucleic acid can be achieved. Thus, in addition to increasing the yield of purified transforming substance, the application of the method outlined has provided a means of isolating the substance from pneumococci other than those of Type III. AS in the case of the Type III material, the Type II transforming substance has been found to be associated with the desoxyribonucleic acid fraction. The results are thus confirmatory of the previous studkc5 on the chemical nature of the trans- forming principle and serve to emphasize that in the instances studied, as in probably all other types of pneumococci, the active agents belong to the same general class of chemical substances. Although the individual desoxyribonu- cleates of different types cannot yet be distinguished from one another on chemical grounds, the selective specki@ they exhibit in inducing transforma- 104 TiANSFOEMATION OF ' PNEUMOti~ TYPl% III tion is diikult to interpret save in .terms'of individual di8erences iu i&mid strkture and molecular contiguration. . SmsMABy :, .; I 1. An improved method is outlined for the isolation and purification of t,& pnetiococcal transforming substance. This method makes use of the fact that citrate inhibits the destructive action of the enzyme, desoxyribonucleaq which is released together with the active material during lysis of the li% ba'cterial cells. A fivefold greater yield of pur%ed transforming agent is ob tained by the present method than by the procedure previously destibed. 2. The specific transforming substance has been isolated from pneumococd of types II and VI, in addition to Type III. In each instance the biologically active material has beeti found to consist of desoxyribonucleic acid. BIBLIOGRAPk -. 1. Avery, 0. T., MacT.,eod, .C. M., and McCarty~ M., J. Exp. Med., 1944,79,137. 2. McCarty, M., J. Gen. Physiol., 1946,29,123. 3. McCarty, Id., and Avery, 0. T.;J. Exp. Med., 1946,8s, 89. 4. Dubos, R. J., and Avery, 0. T., J. Ezp. dkd., 1931,64,51. Dubos, R. J., and Bauer, J. H., J. Exp. Med., 193562,271. 5. Mloway, J. L., J. Exp. Med., 1933,67,265.