118 Report of Dr. Avery (assisted by Drs. Hotchkiss and Taylor) Studies on transformation of pneumococci. In order to demonstrate the transformation of pneumococcal types, the following minimal requirements must be supplied: (1) pneumococci of a competent R strain inoculated into (2), a medium adequate for growth, containing (3) anti-B agglutinins, (4) a factor contained in serous fluids, end in addition (5) a specific transforming principle. A proportion of the pneumococci growing in this system are found to be in the smooth phase, which have been "trensformed" to the specific serological type that had served as source of transforming principle. It has been the objective of recent work of this leboratory to achieve maximum definition of the ‘chemiczl and biological nature of these five factors, anc the mode of their interaction. In recent studies progress in this direction has been mado by the application of two principles: First, in the detailed investigation of any single fector, this component is made the limiting one in the system by. supplying the other four components in the optimal form end quantity insofar as this is known. Secondly, the success of transforma- tion is judged semi-quantitatively by observing the rate of eaeppezarance, and quantity, of the transformed cells produced. These principles have made it possible to assess on a quantitative basis the effect of chemical procedures end treatments that aro reflected only ambiguously in an all- or-none end point titration. This approach hes led to certain findings which will be discussed first in terms of the transforming principle, then of the serum factor, and finally with respect to the interaction of these two factors. aemnen ghee eames ne 119 Chemicel nuture of the transforming principle (Dr. Hotchkiss). Although previous work of this leboratory has strongly indicated that the purified transforming principle of pneumococcus is « highly polymerized desoxyribonucleic acid, chemical charscterization of the matericl wis still in meny ways incomplete. Chemical analysis has therefore been undertaken of an initial concentrate (containing coproximately 300 milli- groms of the finelly purified material) representing a pool of several lots of active transforming material derived from 225 liters of Type Til oncumococcus culture. The materiel was investigated throughout, and especiully ct the Inter stages of a rather extensive purification pro- cedure.’ The analytical data are fruitful in suggesting further possi- bilities for purification, but they will be summarized insofer es they have throvm light on the nature of tho material. The purified material finally recovered shows activity at 0.01 ug por ce., end 0.07 vig per ce. is cpproximately the level which consistently gives transformation in 50 percent of the tests in a system optimum in other respects. It is of especial interest that during the treatments which this active material has received, its amino acid content (determined by a method to be described later in this report) has steadily fallen until less than 0.5 percent of the nitrogen present is amino acid nitrogen. Although a protein material of exceedingly high activity might still be present, it appeurs quite unlikely that non-specific fractionation procedures which are successfully removing protein material in general should simultaneously be causing enrichment of a hypothetical active protein. In the last stage of the purifica- tion, the cmino acid content was reduced by one half. Accordingly, while awaiting further development of this work, it is possible to state fem ate a ir he BE a Aa ced eat Mette ae Nw. definitely that a highly active purified trensforming principle contains not more than 0.6 percent of its weight of protein, and gives promise of being rendered still more neariy free of a proteinaceous constituent. Other analytical data indicate that certain steps in the purification ore highly effective in removing inert moterial, Total nitrogen, orgenic phosphorus, and ultraviolet-absorbing constituents are removed rapidly during treatment with ribonuclease, and then fell more slowly toward an apparently steady value during many reprecipitations. As should be expected, the nitrogen-to-phosphorus ratio demonstrates neither sensitivity nor reliability as an indication of protein content, but fluctuctes considerebly during the processing. In the final product, however, this ratio does huve the velue 1.72, close to the "theoretical" value 1.69 calculated for 2 desoxyribotetranucleotide. In contrast to the constituents just mentioned, desoxyribose wes but little diminished during the ourification process, nearly all of that present in the original crude preparr.tions being found in the highly purified product. Furthermore, the greatly diminished levels of nitrogen, phosphorus and ultraviolet-absorption which remain now bear a relation to the desoxyribose content. However, the ratios are not in all cases precisely the same as those of thymus desoxyribonucleic acid. In particular, the ultraviolet-absorption per unit of phosphorus is about 5 percent lower than that of the thymus preparation. Present conceptions of the nature of nucleic acids re of two divergent kinds. On the one hend, it has long been stated that all are polymers of one of two simple structures, irrespective of source. Hither they ere composed of units containing one residue each of adenine, guanine, cytosine, and uracil combined with four ribose end phosphate residues; or they have the desoxyribonucleic structure mace up of one residue cach of adenine, guanins, cytosine and thymine, with four desoxy- ribose and phosphate groups. Contrasting with this view is the opinion that, for exanple, desoxyribonucleic acids from biolocically different sources possess structures characteristic of their sources, differing quantitatively and perhaps even qualitatively from the composition of the classical structure. This letter view Cerives its support from a few observations of ratios of purine beses deviating materially from the "theoretical" and in one or two reports of the discovery of unusual bases in nucleic acid vreparations. Also, the conception that the pneumococcus transforming principle is essentially a desoxyribonucleic acid implies the corollary that even each such preparation from a distinct serological type of pneumococcus must hive & unicue chemical, or physical, structure. fnalyses of nucleic acids other then those from yeast and calf thymus, however, have been too scattered and incomplete to furnish convincing evidence for either of the opposing views. With the aid of a new analyticzl tool to be described below, it has been possible to determine quantitatively the individual purine end pyrimidine bases of the various preparations used in this work. "Profiles" of three of them are represented in the accompanying figure. It may be noted that purified transforming principle, like calf thymus desoxyribonucleic acid, contains guanine (and d-guenine: see below), cytosine, adenine, and thymine. Cruder material, on the other hand, contains also uracil, and has a much smaller proportion of thymine. Most significant of these preliminary analys2s appears to be the abnormally high content of thymine, one which clearly indicates that at least a portion of the pneumococcus transforming factor must contain two or more PNEUMOCOCCUS TRANSFORMING PRINCIPLE original concentrate ( apssoepanu) PRINCIPLE highly purified PNEUMOCOCCUS TRANSFORITING auimhys ou uapo auisoho auiupn& auruynb-% THYMUS DESOXYRIBONUCLEIC ACID auruhyy aulwapo aussogh>-» aursopho & ( tadod fo “uses wad) 09¢ ~ Nolld¥0S9Y LITOINVELIO 200 mm. 400 DISTANCE ALONG PAPER 123 thymine residues per tetranucleotide "unit," a definite departure from the classical structure. This pyrimidine may in part take the place of some residues of adenine, 2 purine base, and of cytosine, both of which appear to be appreciably below the "theoretical" level. The lower content of adenine and higher content of thymine, 4 base of lower ultra- violet absorption, probably explains satisfactorily the low absorption per unit of phosphorus shown by this nucleic acid as compared with the thymus preparation. In terms of total base recovered, it may be noted from the figure that uracil, a constituent typicel of ribonucleic acids, is com- pletely removed during purification. Calculations show further that adenine and cytosine are reduced during purification, but reveal the noteworthy fact that all of the thymine present in the original material is recovered in the purified preparation. In summary, these findings force one to the conclusion that the effective steps in the purification of pneumococcus transforming prin- ciple are steps which remove proteins and ribonucleic acid, and retain almost without loss the desoxyribose and thymine of the original material. The final product is a desoxyribonucleic acid containing typical con- stituents, but differs from the classical calf thymus desoxyribonucleic acid in having more thymine, less adenine, less cytosine, and lower ultraviolet absorption per unit of phosphorus. Chemical methods for the study of nucleic acids (Hotchkiss). Two methods which were developed for the study, the preliminary phases of which have just been summarized, are sufficiently unusual to require some cescription. First, the assay of small amounts of amino acids in nucleic 124 acid oreparations is complicated by the fact that adenine partially decom- poses to give glycine when heated in acid solution. This side reaction has been controlled by precisely fixing the acid concentration, tempera- ture, ond time of hydrolysis in an autoclave. fAcznine in all its usual compounds behnves quantitatively identically since it is very rapidly liber:ted as free bese. The other purines and pyrimidines do not give rise to amino acids. Under suitabls conditions only 0.5 percent of the nitrogen of an ordinary nucleic acid is converted to glycino. Using the actual adcning content of the nucleic acid preparation us a besis, an necurate correction for the cmino acid libsrated from adenine can b> anplied to the total amino ecid detormined by tho Ven Slyke ninhydrin procedur:. sceondly, and of far mors general interost, a peper chromato- graphic method has been davoloped in this laboratory which makes possible. the quantitative determination of individual purine and pyrimidine bases in a hydrolysite of one or two milligrams of nucleic acid. Briefly, as in amino acid chromatography, a moist organic solvent is caused to flow evenly along « strip of filter paper bearing at one spot the mixture to be analysed. It is found that a base the solubility of which in the organic solvent is high, relative to that in water, travels as a well- defined spot in the moving solvent, and travels mors rapidly the highor its solubility. After a suitable tims, ths paper contains separated narrow bands of individual, highly purified purine and pyrimidine besss. Thes2 substances are dotected by virtue of their ultraviolet absorption at 260 millimicrons. fExcisid segments of the pepzsr strip ars soak2d in wator and ths solution is examined in tho ultraviolet spsctrophotometer. The figure already refsorrcd to shows the results of thre> such analysos, SRD in which magnitude of absorption is plotted against position elong the paper strip. This procedure is mors indirect than the simple color reaction by which amino acids are detected, but it has the enormous oadvantaze that it is quantitative. In addition, experience has shown that individual knorm bases may bea identified positively on the brsis of absorption at a few solected wavelongths. Spectrally sure acenine, urecil, thymine, ana eytosine hrvs been repeatedly isolated from complex nucleic ecids in those: regions of the paper strip carrying maximal absorption. Guenine cen slso be recovered in a virtually pure state in most instances. Furthermore, in case ths positions of two substances on the paper do overlay to a certain degrees, ths individusl substances may be quanti- tatively determined in any binury mixture that may be encountered in practice. This cetorminetion cen be cccomplished by making absorption measuraments et three or four wavelengths so chosen that quentitative uceurecy is maximal. In addition, speciel advantage is taken of charac- teristic shifts in absorption occasioned by the addition of alkali. The total absorption dus to a2 given base is an accurate measure of the total amount of that base present in the originel mixture. Recoveries from : known mixtures heve been within 10 percent of tho theoretical. Applications of this analytical tool, other than that already given, have appeared during the testing of the method. The nucleosides, u cdenosine, guanosine, cytidine, and the desoxyribosides of thymine and guenine have been isolated as successfully upon paper strips as have the bases themselves. Partial hydrolysates of nuclscic acid cen be studied, and for example, it has been showvm that on ecid hydrolysis guanine cand adenine are most rapidly split from yeast nucleic acid, followed by 126 uracil und icstly, cytosine. It is anticipated that the study of enzy- matic digests should reveal something of the structure of the nucleic ceids, and also tho neture of the enzymatic degradations catalyzed by the nuclees2s. Finally, any atypicsl bases present in « specific nucleic acid should be recognizable on « papor strip chrometogrem. The figure shows two such bases, "“y-guenine” and "y-cytosine." The former, judged upon its nbsorption spectrum is closely related to guanine, and is tentatively regarded us an artifact formed only in tho hydrolysi: of certain prepera- tions of nucleic ecid. Its neture is under further investigation. ‘-cytosine is clearly releted to, yet distinct from, cytosine, cnd it has been obtained repeatedly us 2 minor constituent of hydrolysates of 2 preparetion of enlf thymus nucizie ucid. It has not boen identified with any known compound as yet, but ths following suggestive observations might be mentioned: (1) The groater mobility of u-cytosine than cytosine in butyl alcohol suggests alkyl substitution. (2) The absorption spec- trum stends in the same relation to that of cytosine us the spectrum of thymine (5-methyl-urecil) does to that of uracil. (3) Johnson and Coghill in 1925 cleimed to have isolated small emounts of 5-nethyl cytosine from the nucleic acid of the tubercle bacillus. Identification of these un- usuzl bases which have been revealed on the paper strip chromatogrem is being sought, since it should throw further light on the chemical struc- ture of certain nucleic acids. Biochemical studies of the accessory serum factor (Hotehkiss, Taylor, cnd Avery). In the preceding report from this laboratory it was indicated that serum albumin of bovine or human origin can replace humen serous fluids as an accessory factor (factor 4 in the list enumereted 127 zbove) in the transforming system. This fact did not at first appear reconcilable with the observation thet serous fluids in general were of extremely variable potency ana that normal sera were usually ineffective. Tle activity of serum albumin may, however, be obscured by certain interferences. The effect of excessive umounts of the albumin itsel® will be ciscussed in a later section of this report. Another kind of interference is found when certain inhibitory substances are added together with slbumin. or:ll amounts of lauric or oleic acid can inhibit transtorm:tion in a system cont: ining fifty times their weight of albumin. It is not known whether the fatty ecids are themselves toxic or whether a they combine with the ol bumin rendering its chemical groups unable to enter into some other comlinution essential for transform:tion. Serum globulin fractions generally tive « similer inhibitory effect not clearly due to their content of fatty ccics. hs uw result of such inhibition, normal sera and many prspare— tions of rlbumin appear to he licking in activity. By fractionation, however, active albumin my be recovered from sera which are themselves inactive, and the only active fractions are those containing albumin. Moreover, when serum fri.ctions are tested quantitatively in . transform- ing system adequate with respect to the other factors, it becomes evident that further treatment of the albumin preparations is often necessary to bring them to full effectiveness. Treatment with trichloro- acetic acic, or heating in acid at pH <.5 have proved to be useful for this purposs. These procedures are such that globulins