'J-HE OCCURRENCE DURING ACUTE INFECTIONS OF A PROTEIN NOT NORMALLY PRESENT IN THE BLOOD II. ISOLATION AND PROPERTIES OF THE REACTIVE PROTEIN BY COLIN M. MA~LEOD, M.D., AND OSWALD T. AVERY, M.D. (From tke Hos@ital of The Rockefeller Institute for Medical Research) (Received for publication, October 7, 1940) The studies of Abemethy and Avery1 concerning the reactive substance precipitated by the C polysaccharide of Pneumococcus from serum obtained from human beings and monkeys during certain acute infections, have indi- cated that the reactive material is of protein nature. In order to gain fur- ther knowledge of this "reactive protein," attempts have been made to obtain an active fraction as free as possible from other.serum constituents by methods which do not involve the use of the C polysaccharide as pre- cipitant. The present report deals with the isolation of the reactive protein from acute phase sera and with certain of the properties which distinguish it from normal serum proteins. Throughout this paper, for the sake of convenience, the term "acute phase" serum is used when referring to serum obtained from patients acutely ill with an infectious disease. The material precipitated from acute phase serum upon the addition of the C polysac- charide is referred to as reactive protein. EXE'ERWENTAL In the preceding paper of this series it was shown that the reactive protein in acute phase serum is present in the albumin fraction, and that the serum globulins removed by half saturation with ammonium sulfate do not react with the C polysaccharide. The following experiments were performed to determine whether the reactive protein is distributed uniformly throughout the total albumin or whether it is limited to a particular fraction of the al- bumin precipitated between half z&d full saturation with ammonium sulfate. _ Fractionation of Serum Albumin with Ammottium Stiuate.- 10 cc. of acute phase serum were diluted with an equal volume of physiological saline, tid 20 cc. of a saturated solution of ammonium sulfate added. The mixture was allowed to stand at room temperature for 1 hour, and the precipitated globulin removed by TAbemethy, T. J., and Av&, 0. T., J. Lfi..M$l, 1941,73, 173. ). 183 184 ABNORMAL PROTEIN IN BLOOD DURING INlkCTION. II centrifugation. To the solution of albumin, suflicient saturated ammonium sulfate solution was added to bring the salt concentration to 75 per cent saturation. After standing for 1 hour the precipitate (50 to 75 per cent fraction) was collected by filtration. The filtrate was then brought to full saturation and the final precipitate (75 to 100 per cent fraction) collected by fltration. The globulin and the two albumin fractions were resuspended in 5.0 cc. of distilled water and dialyzed in cellophane sacs against physio- logical saline until free of sulfate. On removal from the dialyzing sacs, the volume was . made up to 10.0 cc. in each case by the addition of saline. To 0.5 cc. portions of each serum fraction an equal volume of C 1:20,000 was added to determine the presence or absence of the reactive protein. The results of these tests are shown in Table I. TABLE I L%st&utbn of Reactive Protein in Fractions of Act& Phase Seruna Fraction of acute phue ~aum Wholeserum ............................................................ +++++ Globulin - ............................................................... Albumin (fraction precipitated between 50 and 75 per cent saturation)t ......... ++-I-* Albumk (fraction precipitated between 75 and 100 per cent saturation)$. ....... f *In this and the following tables the degree of precipitation is expressed by the symbols: ++++ - flocculent precipitate, clear supernatant. f = slight turbidity. t Fraction of albumin precipitated betwekn 50 and 75 per cent saturation with am- monium sulfate (active albumin. fraction). $ Fraction of albumin precipitated by full saturation with ammonium sulfate follow- ing removal of albumin precipitated by 75 per cent saturation. From the data shown in Table I it can be seen that the reactive protein is restricted to that fraction of the serum albud precipitated between 50 and 75 per cent saturation with ammonium sulfate. The fraction of the albumin precipitated between 75 and 100 per cent saturation is only slightly reactive with the C polysaccharide, and as previously show, the globulin fraction does not contain reactive protein. By fractional precipitation considerable separation of inert protein ca.n be readily effected, and for further purification only the fraction of the albumin precipitated between 50 and 75 per cent saturation with ammonium sulfate or sodium sulfate was used. This fraction Will be referred to as the "active albumin fraction." Separation of Reactive Protein by Djdysjs of Active Ah&n Fracliotr against Tap Water.-In the previous experiment the albumin fractions were dialyzed against physiological &.ue to remove the ammonium sulfate- When tap water instead of saline was used for dialysis, it was found that the C0LI.N bf6. MMLEOD A.ND OSWALD T. AVERY 185 rsactive protein precipitated whereas the normal albumin remained in Slution. Dialysis against tap water was continued until the solution was iree of sulfate, and the flocculent precipitate which formed was recovered bu centrifugatiOD. On the addition of physiological saline most of the pre- cipitate again became soluble, although a small amount of material generally remained which was insoluble in saline. This appeared to be protein which had become denatured during the course of dialysis at the pH of ammonium sulfate (PH 5-s). Consequently, in subsequent fractionations sodium sulfate was used in place of ammonium sulfate, all procedures up to the Pint of dialysis being carried out at 37oC. Fr~tio?&O?Z of &rum Albumin with Sod&m Suifak- 1~ cc. of serum were obtained at autopsy from a fatal case of Type III Pneumococcus pneumonia. After titration through a Berkefeld V candle the serum was diluted with 1~ cc. of distilled water. The subsequent steps in fractionation were carried out at 37oC. To the diluted serum 200 CC. of a saturated solution of sodium sulfate were added. After standing for 1 hour the precipitated globulin was collected on a titer paper. 355 cc. of the clear tiltrate containing the albumin were brought to 75 per cent saturation by the addition of 177 cc. of a saturated solution of sodium sulfate, and allowed to stand for 40 minutes after which the albumin precipitate was collected on a filter paper. This precipitate was washed off the filter paper with tap water, and the solution was refiltered to remove paper fibers. The finsl volume was 80 cc. The clear solution was transferred to a cellophane sac and dialyzed against running tap water for 3 hours. At the end of this time the solution in the sac had become turbid and a &e flocculent precipitate was beginning to settle out. The dialyzing sac was then transferred to a 10 liter jar of tap water and dialysis continued in the cold room (approximately + 2oC.) for 16 hours. The copious flocculent precipitate was separated by centrifugation for 1 hour in the cold. The packed precipitate was light yellow in color; the supematant albumin solution remained slightly turbid but cleared on the addition of sodium chloride in a concentra- tion of 0.85 per cent. The pH of the albumin solution was 6.97. The precipitate was completely soluble in physiological saline, and when redissolved, reacted strongly with the C polysaccharide. The albumin fraction which precipitated between 75 and 100 per cent saturation WBS dissolved and dialyzed against tap water. No precipitation occurred during dialysis. The salt concentration of the various fractions was adjusted with NaCl to 0.85 per cent. Each fraction was tested for reactivity with the C polysaccharide. The . results of these tests are ahown in Table II. :.c -!? The results shown in Table II indicate that practically all of the reactive protein is precipitated when the active albumin fraction is dialyzed against tap water.. If instead of tap water, calcium-free distilled water is used for dialysis, little or no precipitation occurs. This difference in solubility is attributable to the presence of traces of calcium in tap water; 186 ABIiORMAL PROTEIN IN BLOOD DURING INPECTION. II Further Pw+i.catim ,of Reactive Prot&.-The precipitate of reactive protein obtained by the method described above was further purified by dissolving it in alkaline s&line and reprecipitating by dialysis. The protein was dissolved in saline buffered at pH 7.9 with M/SO borate buffer and re- dialyzed against tap water. Distilled water is unsatisfactory, since as pointed out above, the reactive protein is soluble in water unless a trace of ' calcium is present. : `. _. 90.0 cc. of pooled acute phase `serum were fractionated by means of sodium sulfate according to the method described. The yellowish precipitate of reactive protein ob- tained by dialyzing the active albumin fraction against tap water was dissolved in 25 cc. of saline buffered at pH 7.9 with M/SO borate buffer. The somewhat opalescent solution was dialyzed against tap water for 48 hours in the cold room, and the precipitate collected TABLE11 Preci@atiota of Reactive Pro&in un Dialysis of the Albumin Fra&m against Tap Water SeIum fnctfon Prcdpitation mctum ritb c 1:zo#ooo 1. Whole serum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ++++ 2. Activealbuminfraction*............................................... +++* 3. Proteinprecipitatedfmmfraction2on dialysisagaiusttapwater . . . . . . . . . . . -t++ 4. supematant of fraction 3.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . + *Fraction of albumin precipitated between 50 and 75 per cent saturation with sodium sulfate. by centrifugation in the cold. A small portion of the precipitiite was dissolved in saline and found to be reactive with the C polysaccharide. The bulk of the precipitate was dissolved in 10.0 cc. of distilled water to which was added 0.4 cc. of ~/lo0 NaOH. The lipids were removed by shaking at room temperature in 10 volumes of 50 per cent alcohol- ether. The precipitate was washed once with ether and dried to constant weight. 21-o mg. of dried material were recovered. Chemical analysis of the dried precipitate gave the fokkng results: N, 13.97 per cent; P, 0.05 per cent. It would appear justifiable in the light of this and other experiments to refer to the C-precipitable material in acute phase serum as reactive pro- tein. The nitrogen content of various purified preparations has been of the same order as that found in the above experiment, namely between 13 and 14 per cent. The phosphorus content of material not extracted with alcohol and ether has been found to lie between 0.45 and 0.7 per cent, but upon extraction of the lipids the phosphorus content fell to a very low value as in the above preparation where it was found to be 0.05 per cent. Absmpion of Acute Phase Serum with C Polysaccharide and Subsequent COLIN M. MAcLEOD AND OSWALD T. AVERY 187 ,-,,,~/i~t&t.-It wss of interest to determine whether or not absorption oi %,-ute phase serum with the C Polysaccharide will remove the reactive Protein P recipitable from the active albumin fraction upon dialysis against t3p mater. 20 c~. of acute phase serum were absorbed with the C polysaccharide so that no further nraipitation occurred on the addition of an excess of C. The C-absorbed serum was fn,-conated by means of sodium sulfate as outlined above. The albumin fraction precipitated between 50 and 75 per cent saturation was dialyzed against tap water for 72 hoUs+. No precipitation of reactive protein occurred under these circumstances. It would appear, therefore, that the material precipitated from acute pbsse serum on the addition of the C polysaccharide, and the material precipitated from serum of the same origin by dialysis of the albumin frac- tion against tap water, are identical. R&t&m of Lipids TV Reactive Protein.~In order to determine the rela- tion of lipids to the reactive protein, acute phase serum was treated with &ohol and ether in the cold, and an attempt made to separate the reactive protein from the "defatted" serum. &mat of Lipids from AC& Phase Serum. -100 cc. of acute phase serum were chilled to 0oC. and added slowly, with constant stirring, to a mixture of 750 cc absolute alcohol and 250 cc. of absolute ether chilled to -12oC. in a salt-ice bath. The suspension was kept in the salt-ice bath for 1 hour with frequent shaking, after which it was centrifuged at -1T. to collect the precipitated protein. The precipitate was washed three times with absolute ether at - 12'C., then distributed in a thin layer and dried in V~WU. The defatted and dried serum was taken up in the original volume of normal saline, in which it was completely soluble. On fractionation with sodium sulfate, the albumin which precipitated between 50 and 75 per cent saturation appeared definitely less in amount than in the case of serum from which the lipids had not been removed. After dialysis against tap water for 36 hours only slight turbidity of the solution occurred. The small amount of water-insoluble material was collected by centrifugation, and when dissolved was found to be reactive with the C polysaccharide. The supernatant albumin which still showed marked reactivity with the C polysaccharide, was dialyzed for an additional period of 48 hours, but no further precipitation occurred. The various fractions of the defatted albumin were tested for reactivity with the C polysaccharide as shown in Table III. The results shown in Table III indicate that the serum lipids are of importance in dete r-mining certain of the properties of the reactive protein, particularly its sensitivity to calcium as indicated by increased solubility in tap water. The reactivity of the defatted albumin with the C poly- saccharide remains unchanged. After removal of the serum lipids, although the reactive protein is still almost quantitatively precipitated by 7'5 per 188 ABNORMAL PROTEIN IN BLOOD DURING INFECTION. II cent saturation with sodium sulfate, its solubility in tap water is greatly increased. For this reason only a small amormt of the reactive protein is precipitated upon dialysis against tap water, and the supematant albumin retains practically all of its original reactivity with the C polysaccharide. Moreover, it has been found that the reactive protein remains soluble in the presence of a concentration of calcium sufficient to precipitate it from . undefatted albumin. : TABLE m Eject of Extract&m of Lipids on the Solu.Wty of the Reactive Protein in AC& Phase Serum salml fmctlo6 Whole serum ................................................ Globulin .................................................... Active albumin fraction* ...................................... Supemate of active albumin fraction after dialysis against tap water .................................................... 75tolOOpercentalbuminfraction~. .......................... ++++ +++* + f Afte; ~yIr&v&l ++++ +++* +++ f * Fraction of albumin precipitated by 75 per cent saturation with sodium sulfate. t Fraction of albumin precipitated by full saturation with sodium sulfate following removal of albumin precipitated by 75 per cent saturation. DISCUSSION One Of the most unique properties of the reactive protein is its precip- itability by the C polysaccharide of Pneumococcus. Under these condi- tions the reactive protein is almost quantitatively precipitated. The effect of calcium in this reaction has been described.1 It was shown that floccuh+ tion does not occur when the polysaccharide is added to a reactive serum from which the calcium has been removed, but that this capacity is corn-- pletely restored on the addition of calcium in a concentration much lower than that normally present in blood. The precipitate formed in the presence of the polysaccharide becomes soluble when the calcium is split off at acid or alkaline reactions or by the addition of sodium citrate. flow- ever, these procedures have not as yet proved useful in further purification of the reactive protein because of the difficulty in removing the polY- saccharide without at the same time denaturing the protein. Because of the importance of calcium in the precipitation reaction it was thought that this reagent might be used as a direct means of isolating the COLIN M. MAcLEOD AND OSWALD T. AVERY 189 EactiVe protein. It was found that when an excess of calcium is added . to acute phase serum the reactive protein is precipitated even in the absence of the C polysaccharide. This procedure however has certain disadvan- . . * tages since preclpltatlon of normal serum proteins also occurs under these conditions. In addition most acute phase sera contain large amounts of lipoid material which flocculates upon the addition of an excess of calcium &oride. The COpiOUS fatty precipitate is difficult to separate from the reactive protein without causing denaturation of the latter. A further obje&ll to the USA of calcium chloride is that the precipitate formed is not soluh1e at neutral reaction. As a consequence of these observations procedures were sought for ob- taining the reactive protein in a relatively pure state without making use of either the C polysaccharide or an excess of calcium as precipitants. The reactive protein has been found to be present in the fraction of albu- min precipitated by ammonium or sodium sulfate between 50 and 75 per cent saturation. Sodium sulfate was found to be more satisfactory since the slight acidity of ammonium sulfate usu&y caused denaturation of part of the reactive protein, particularly if dialysis were prolonged. By the use of sodium sulfate denaturation is avoided. The reactive protein is only partially precipitated from the active albumin fraction of serum on dialysis against distilled water. However, more complete separation results if tap water is used for dialysis in place of distilled water. This difference is due to the trace of calcium contained in tap water. The reactive protein can then be freed of extraneous material by dissolving the precipitate in physiological saline at alkaline reaction and redialyzing against tap water. The purified material obtained by this means is fully reactive with the C polysaccharide. The property of insolubility in tap water makes it possible to separate the reactive protein from serum albumin. It has been found that this property is due in part to the association of the protein with certain lipids. If acute phase serum is first extracted with alcohol-ether in the cold to remove the lipids the reactive protein is still almost quantitatively precipitated by sodium sulfate between 50 and 75 per cent saturation. However, it now remains soluble upon dialysis against tap ,water and is not precipitable by calcium. It is noteworthy that the reactivity of the protein with the C . plysaccharide is not altered by removal of the lipids. a . It is evident, therefore, that the solubility of the reactive `protein is conditioned by the interaction of two factors; namely, the presence of the lipid or lipids with which it is intimately associated, and the marked sensi- 190 ABNOBhfAL PROTEIN IN BLOOD DURING INTECTION. II tivity of this lipo-protein complex to-calcium ions. The effect of both these factors on solubility is evidenced by the fact that following removal of the serum lipids, the active protein is far less. sensitive to calcium, as shown by its solubility in tap water. Although important in determining the solubility of the protein, the lipids are not essential in the C-reaction, since they can be extracted either before or after fractionation of the serum kvithout impairing the capacity of the protein to react with the poly- saccharide. Studies of the lipoid constituents of the reactive protein have not as yet been made. However, it is probable that a phosphorus-containing lipid is associated with the protein since phosphorus determinations on prepara- tions of the reactive protein after extraction with lipid solvents show a much lower phosphorus content than similar preparations from which the lipids have not been removed. SUMMARY Methods are described for isolating a protein commonly present in the blood of patients during the acute phase of various infections'which, unlike the normal serum proteins, is precipitable by the C polysaccharide of Pneumococcus. The reactive protein is present in the fraction of serum albumin precipi- tated by either ammonium or sodium sulfate between 50 and 75 per cent saturation. From this fraction the reactive protein separates out on dial- ysis against tap water. Following removal of the alcohol-ether-soluble lipids from acute phase serum the reactive protein becomes soluble in tap water, and is no longer precipitable by traces of calcium but still retains its precipitability with the C polysaccharide.