Report of Dr. Avery with Drs. Stillman, Goebel, Dubos, Francis, Kelly, Babers, Goodner, and Alloway. I. The decomposition of the Capsular Polysaccharide of Type III Pneumococcus by a Bacterial Enzyme. 1. Methods for obvtaining potent, non-toxin preparations of the enzyme. 2. The protective action of the specific enzyme against Type III Pneumococcus infection in mice. Il. Isolation of other microorganisms decomposing the capsu- lar polysaccharides of different types of Pneumococcus. III. Chemo-immunological studies on Carbohydrates. 1. The determination of the molecular size of the cap- - sular polysaccharide of Type III Pneumococcus. 2. The specific carbohydrate of two strains of Pfeiffer's bacillus. : 3. The somatic carbohydrate of Pneumococcus. ! . 4, Studdes on synthetic carbohydrate derivatives. a) Synthesis of .x para-amino phenol glucoside. | b) Glucuronic acid. >” ’ ry. Chemical Nature of Type Specific, Capsular antigen of Pnevumococcus. V¥. Significance of the Skin Test as a guide to Serum Therapy in Pneumonia. VI. Studies on Natural Resistance and the Immunity induced by R pneumococci. ey Km “3 5 H 92 VII. Principles Governing the Precipitin and Agglutinin reactions with Pneumococcus. VIII. Studies on Epidemiology of Pneumonia. 1. Antibody response to imuunization by different routes. 2. Production of experimental pneumonia in aninals. IX. Antipneunococms Immune Reactions of Normal Hog Serun. X. Pathogenesis of Experimental Intradermal Pneumococcus Infection. 1. Intradernmal infection with a rabbit virulent Type II!I Pneumococcus. XI. Studies on the Transfornation of the Specific Types of Pneumococcus, XII. Significance of Oxidation-Reduct ion Processes in Bacterial Growth. A. Purification of a brand of connercial peptone. B. Preparation of peptones from pure proteins. XIII. Publications. * SAMS RETR pees ae rT eS Te cyte EE oe RY cen era erst es enon aH H 93 I. The decomposition of the capsular polysacchar- ide of Type III Pneunococcus dy a bacterial enzyme. (Drs. Avery and Dubos). A microorganism has been isolated, and from it an enzyne extracted, which deconposos tne purified capsular polysaccharide obtained from Type ITI Pneumococcus. This nicroorganisn, a pleonorphic aerobic bacillus, was iso- lated fron a enil rich in hemicelluloses under-going decan- position, by the use of a synthetic mineral mnediun contain- ing the capsular polysaccharido as sole source of carton, As alrendy described in a provious report, the endocellular enzyne extracted from this microorganism exhibits a rerark- able specificity in its action against the capsular polysac- charide of Type II] Pneumococcus. The work of this year may be conveniently describ- ed under two cain headings - ly Attempts at obtaining potent and non-toxie enzyme pre- parations. . 2. Protective action of a specific enzyme against Type Ii! Pneumococcus infection in nice. Before going on with a description of these stud- des, it may be worth while to state that the work has been controlled by a rapid and fairly accurate method of titra-~- tion of the potency of the enzyme preparations in vitro; this method, based on the existence of a quantitative rela- tionship between amount of substrate decomposed and amount of enzyme used, consists in determining the amount of cap- 7 at ayy H 94 sular polysaccharide decomposed by a definite amount of enzyme preparation. 1. Attempts at obtaining potent and non-toxic enzyme preparations. The amounts of specific enzyme obtain- ed from a culture of the bacillus decomposing the capsular polysaccharide of Type III Pneumococcus fs not a direct function of the number of cells from which the enzyme is ex- tracted. It is conditioned by a series of factors among which the following have been recognized:- A. Past history of the culture. B. Composition of the culture medium, with special reference to ~ a) The presence or absence of the capsular polysaccharide in the mediun; bd) The presence or absence of substances acting as a more available source of energy for the bacillus: c) The form in which nitrogen is supplied in the medium; a) the presence of catalysts affecting the rate of decomposition of the capsular polysaccharide; and e) The presence of substances retarding autolysis of the cells. °:. iG. Bnvironmental factors, especially - temperature and.. conditions of aeration. 9 9). * In brief, the following points must be observed in order to obtain a maximum yield of enzyme. The culture must be maintained at 4 high degree of specific activity by con- tinued passages in the specific synthetic medium containing the capsular polysaccharide as sole source of energy. The enzyme 4s usually not formed, or only in small amounts, un- pee cae se sla neas BE Tate “cline A AGRA OAT nee WME RE Te, hy PSA PMOL ici nent a 28 H 95 less the capsular polysaccharide (or the aldobionic acid de- rived from it), is present in the medium. However, the con- centration of capsular polysaccharide must not be too high; it must be such that the decomposition is completed in 2-3 days at the most. This condition may be accounted for as follows:- The enzyme can be demonstrated free in the medium, as soon as the cells begin to age. In case there is still some polysaccharide undecomposed at this time, the free en- zyme decomposes the residual polysaccharide and is therefore neutralized without contributing in any way to the anabolic processes of tne cell. This results of course ina loss of both enzyme and capsular polysaccharide. This loss may be reduced in two ways: - a) by increasing the rate of bacteri- al growth (use of a large inoculum, presence of enough nitro- gen, proper conditions of aeration, and the addition of growth catalysts, such as soil extract and yeast extract), and by retarding autolysis and thus preventing the libera- tion of the enzyme (addition of yeast extract to the medium). Under these conditions, the yield of enzyme has been increas- ed almost tenfold over that previously obtained. Unfortunately some of the new preparations have exhibited a definite primary toxicity. The production of non-toxic preparations is being attempted along three dif- ferent lines?- a) modification of the medium used for the growth of the bacillus; and b) modification of the process of extraction of the enzyme; and c) by purification of the se caPV Ace Lie eR Sede eee oe o 7 - il ern i re a ree a9 HK 96 enzyme preparations. 2. Protective Action of the Specific Enzyme on Type III Pneumococcus Infection in Mice. The earlier stud- jes comprised observations on the action of the enzyme on the capsular sugar of Pneumococcus removed and separated from the bacterial cells. In the natural state in which this specific polysaccharide exists, it forms the capsular structure which determines the antigenic and serological specificity of the cell as a wnole, and conditions its pow- er to invade and multiply in animal body. It was of special interest, therefore, to determine what effect the specific enzyme would have upon the encapsulated forms of type IIf pneumococcus growing in vitro and in vivo:- whether in a medium containing the active enzyme, the pneumococci would fail to grow at all or would grow merely deprived of their capsules; whether 4n the body of a susceptible animal, the administration of the enzyme would in any way modify the course of experimental {infection with virulent Type III pneu- mococci. The results of these studies form the subject mat~- ter of a paper which is. now in manuscript forn. Briefly, the work shows that, when a sterile ex- tract of the enzyme is added to a culture medium seeded with Type II! pneumococci, growth occurs but the bacteria are de- prived of their capsules; they are no longer specifically agglutinabdle, and the soluble specific substance is no long- er demonstrable in the culture fluid. Experiments of this Nd EH 97 kind have shown quite clearly that the enzyme by itself is neither bactericidal nor bacteriolytic; that by decomposing the specific carbohydrate, it merely strips the pneumococci of their capsules without impairing the viability of the bacteria. Moreover, the action of the enzyme on the living cell does not result in a loss of the function of elaborat- ing the capsular substance, since pneumococci so treated, regain their capsules and continue to elaborate the soluble specific substance when transferred to fresh medium free of the enzyme. In more recent studies it has been found that the enzyme has a distinct protective action in the animal body against infection with virtlent Type III pneumococci. Mice receiving a single intraperitoneal injection of an active preparation of the enzyme may survive infection with a mil- lion times the fatal number of virulent organisms. The pro- tection afforded by the enzyme is type-specific; just as in the test tubes the ensyme acts only on the Type III polysac- charide, so in the animal body it is effective only against infection with Type III pneumococcus. A protocol of an ex- periment illustrating the degree and the specificity of the protective action of the enzyme in mice is given in Table I, SOE H 98 Table it. Specificity of the Protective Action of Enzyme. Infecting Enzyme (lot 4a) 0.5ce. No Enzyme dose of { | Pneumococcus | Type _ I | Type II | Type Ill Viruéence controls ces | Type Ij it Til el - a Ss 7 - : = - t 01 - - s - - 7 | .001 - ~ § - - - .0001 p20 D34 s - | - - 00001 D24 D34 s p22 ! p36 | D34 .000001 D34 D34 s DS4 | D36 D34 .0000001 - - - D34 ! s20 | p72 | § = Survived. D = Death of animal; the numeral indicates the number of hours before death, or the time at which the animals were found dead. ~ = Not done. The active principle responsible for the protection of mice is destroyed by exposure to 70° Cc. for ten minutes. The inactivation by heat of the protective power for mice of this enzyme, parallels the loss of its activity when tested in vitro after exposure to temperatures of 60° C. or higher. The fact that heat destroys the power of the enzyme to act either in vitro or in vivo, justifies the assumption that the same principle is involved in the mechanism of both reactions. Repeated tests have shown that within the limits of the reaction-capacity of the mouse, the protective action of the enzyme is a function of the concentration of the ac- tive principle in any given preparation; and further, that the protection afforded by different preparations bears & definite relation to their activity in decomposing a known quantity of the capsular polysaccharide in vitro. The evi- dence is that, to be effective in animal protection, the con- centration of enzyme in the body must be maintained in ex- cess so that the rate of decomposition of the capsular sub- stance is always greater than the rate of its production by the living bacteria. Experimental evidence indicates that in mice the enzyme has a curative action when administered in the course of an infection already established at the time of treatment. A single injection of enzyme as late as 18 hours after the onset of infection has brought about the recovery of mice infected with a dose of virulent pneumococci 100 times great- er than that fatal for the untreated controls. It must be borne in mind, however, that the curative action of the en- zyme suffers the limitations imposed by the variations which occur in: the’ eeliular ‘response: of the host, since phagocyto- sis is apparently the Amportant’ ‘process involved in the pro- tection afforded by. the “ontyme. 7 oe ‘better ‘understanding of the mechanism under- ly- ing the protective action of ‘the enzyme in infected mice has been gained by following the course of the bacteremia by means of blood cultures and by a study of the cellular reac- tions in the peritoneal exudates of treated and untreated animals. The results of a typical experiment are graphical- ly illustrated in the accompanying photomicrograph (fig.1). ia SENSOR ie EER REE Rt HR a AE RO ond H 100 In this experimenta, 12 mice were infected with one million fatal doses of Type III pneumococci. Six of these mice re- ceived at the time of infection 0.5 cc. of an active pre- paration of the enzyme. Theo ther six mice served as un- treated controls. At intervals following infection, one mouse of each group was sacrificed and autopsied. Cultures of the heart's blood were made, and stained films of the peritoneal exudate were examined to compare differences in the morphology of the organisms and in the occurrence of phagocytosis in the treated and untreated animals. Two hours after infection the blood cultures of both the treated and untreated mice were positive. Micro- scopic examination of the exudate of the control animal at the end of 2 hours shows the presence of many well encapsu- lated cocci in every field without any evidence of phagocy- tosis. (fig. -la). On the other hand, the pneumococci pre- sent at this same period in the éxudate of the enzyme-treat- ed mouse, (figs 1b), are smaller in size and devoid of cap- sules; only the naked cells are. visible and many of these have already been taken up by the leucocytes. At the end of four hours, the bacteria are more numerous in the peri- toneal exudate of the infected control (fig. le); they are surrounded by well defined capsules and none are found with- in the leucocytes; cultures of heart's blood of the untreat- ed animals indicate a progressively increasing bacteremia. In the treated animals at this period, however, (fig. 14) OO5 H 101 only an occasional naked cell is seen outside the leucocytes and frequently at this time and invariably by the fifth hour the organisms are no longer demonstrable in the blood. The evidence obtained from repeated experiments of this nature supports the view that the protective action of the enzyme lies in its capacity to decompose the capsular polysaccharide of Pneumococcus Type III. The process of de- capsulation, brought about by the direct action of the en- zyme, strips the bacteria of an important defense mechanism and thereby exposes their naked, unprotected bodies to di- rect attack by the phagocytes of the host. In this sense the action of the enzymes may be said to initiate phagocy-~ tosis: not, as in the case of antibodies, by specifically sensitizing the bacteria, but by the process of decapsula- tion. In the former instance, the reaction 46 an immuno- logical one, whereby the specific substance of the capsule is altered by union with the type-specific antibodies; in the latter case, the reaction is a chemical one in which the capsular polysaccharide itself is decomposed by the en- zyme. Although the mode of action is aifferent in each in- stance, the end result, so far as the fate of the pneumo- coccus is concerned, is the same in both instances. The enzyme, like the type-specific antibody, serves to initiate a protective reaction, the completion of which, however, is ultimately dependent for its successful issue upon the effective cellular response of the host, This study GOS H 102 suggests that the capsule - long recognized as a defense meé- chanism of virulent bacteria - is a decisive factor in deter- mining the fate of pneumococci in the animal body, and that this structure is vulnerable to attack by specific agents other than antibodies. II. Isolation of other microorganisms decomposing the capsular polysaccharides of different types of Pneumo- coceus. (Dr. Dubos). The search for other microorganisms decomposing tne specific polysaccharides of Pneumococcus types is being continued. A new spore forming, Gram positive, thermophilic bacillus, decomposing tne capsular polysaccharide of Type III Pneumococcus, has just been obtained from horse manure. Nothing is known as yet of the biochemical properties of this organism except that it decomposes the Type III poly- saccharide when incubated at 50° C. From a corn cob compost, there has been isolated an unidentified microorganism, probably pelonging to the Mucorales, which decomposes the Type II capsular polysac- charide, This organiem does not attack the capsular poly- saccharides of Type I and Type III Pneumococcus. It may be recalled that the capsular polysacchar- ide of Type II Pneumococcus exhibits the same serological reactions as that of Type B Friedldnder bacillus. It is a remarkable fact that the same microorganism which specifi- cally decomposes the Type II polysaccharide of Pneumococcus, oh? H 103 also decomposes the immunologically related capsular substance of Type B FriedlAnder bacillus. These observations again con- firm the fact that the capsular polysaccharides determine the specificity of these serological types. III. Chemo-immunological Studies on Carbohydrates. (Dr. Goebel and Mr. Babers). 1. Determination of the molecular size of the capsular polysaccharide of Type III Pneumococcus. A preliminary account of tne determination of the molecular size of the specific polysaccharide of Type III Pneumocococcus was given in the last report. This work has since been com- pleted. The diffusion coefficient of the sodium salt of the specific carbohydrate has been found to be 0.0415 om.” per day. From this value the radius of the molecule has been calculated as 2.975 x 107’em. and subsequently a molecular weight of 118,000 has been ascertained. The molecular weight of hemoglobin and of the carbohydrate from ovomcoid were de- termined in control experim nts. Values of 67,200 and of 2,200 respectively, were found. | These numerical values are in agreement with the accepted values of other investigators, reported in the literature. 2. The specific carbohydrate of two strains of Pfeiffer's bacillus. The specific carbohydrate from a strain of Type A Pfeiffer's bacillus has been isolated from cultures, grown on dextrose-agar. The medthod of isolation was similar to that which has been used in the preparation of other bac- terial polysaccharides. Three preparations have been secur- nen 33 H 104 ed which react with homologous. immune rabbit serum in dilu- tions of 1:2,000,000. The carbohydrate has a levo-rotation of -32°, it is acidic, and in the present state of purifi- cation contains only slight traces of nitrogenous constituents. The carbohydrate from a strain of Type B Pfeiffer's bacillus is also being studied. Thus far a material has been secured which poseésses only a fourth of the serologi- cal activity of that of the Type A strain. This carbohy- drate appears to be much less stable than that of Type A. Both the A and B types of Pfeiffer's bacillus elaborate only small quantities of specific polysaccharides, about one- tenth the quantities obtained from Pneumococcus, A more de- tailed account awaits the preparation of sufficient quanti- ties of the materials for chemical investigation. 3. The somatic carbohydrate of Pneumococcus. The "Cc" substance, or species-specific somatic carbohydrate of Pneumococcus has been identified as a nitrogenous’ polysac- charide which yields 30 per cent of reducing sugars, calcu- lated as glucose, on hydrolysis. at ‘the suggestion of Dr.” Francis, a phosphorous analysis waa made on this substance by Dr. Heidelberger at the Presbyterian Hospital. The car- vohydrate contains organically bound phosphorous to the ex- tent of approximately 3.3 per cent. We have also found this material to contain about 5 per cent of nitrogen. The car- bohydrate is hydrolyzed by nitrous acid in the cola, with an accompanying loss in specificity. We have attempted to split ko H 105 off the organic phosphorous of this carbohydrate by means of a powerful phosphatase, but the enzyme causes no liberation of inorganic phosphorous. At present sufficient quantities of this material are being prepared for Dr. Francis in order that he may carry out certain clinical studies. 4. Chemo-immunological studies on synthetic carbo- hydrate derivatives. In previous reports we have described the method for the preparation of synthetic carbohydrate (hexose) -protein antigens, as well as the preparation of a polysaccharide-protein antigen which elicit specific immune responses wnen injected into animals. From these studies we have learned first, that carbohydrates, both simple and com- plex, can be rendered antigenic when they are coupled to a protein, and second, that the specific immune response (in the case of hexoses) is determined by the configuration of the hexose molecule. Jn the case of large molecules, such as the specifically reacting bacterial polysaccharides, the . ultimate immune response is such that the induced antibody will not only agglutinate microorganisms from which the car- pohydrate was derived, but will actually confer passive pro- tection on mice infected with an homologous type of organism. We are inclined to believe that in order to secure immunity against a disease in which the chief immune response is de- pendent upon the elaboration of anticarbohydrate antibodies, it is necessary only to have as the antigenic agent, a com- bined carbohydrate, of an exact and correct chemical confi- iD H 106 guration, capable of stimulating the formation of these an- tibodies. In a molecule as complex as that of Type III Pneu- mococcus polysaccharide there are naturally many factors which enter into the orientation of its ultimate specificity. We know, for example, that the molecule is studded with highly polar carboxyl groups. We know that there are three free hydroxyl groups per unit of hexose, and we know that the molecule is a polymer of an aldobionic acid (glucurono- glucose). We do not know, however, the mode of linkage of one aldobionic acid unit to the next; we do not know whether the elucurono-glucose is an x7 or 6 glucuronoside, nor do we even know if an A sugar Can give rise to a different antibody than its « homologoue. However, these fundamental problems in the under- standing of the specificity of immunological reactions in the case of carbohydrates are not entirely unapproachable, for we have at our disposal methods based on synthetic chen- istry, whereby we Can analyze the influence, not only of in- termolecular stereochemical relationships, but the influence of polar groups in the carbohydrate molecule as well. a) The synthesis of % para-amino phenol giuco- side. We have demonstrated in previous work that the alter- ation of the stereo-chemical configuration of one carbon atom in different hexoses, suffices to elicit totally differ- ent and specific antibody response. Our next problem is to oad H 107 ascertain whether the A and @ glucosides of the same hex- ose would give rise to separate and distinct antibodies, or merely to antibodies showing but slight aifferences. We have already synthesized the ¢ -p-nitro phenol glucoside of glucose. The homologous & glucoside has been synthesized in the following manner. C HOH jf HODOCH, H-0-C1 / | ‘| yt / C HOH / C HOCOOH, /, HOCOC., fs 4 (CH, co) 2° f | {| 0 ¢ HOH LS 2 of © HOCOCH, 4 PC1l, Q GHOCOH \ \ | nan | HOH . ° \ ¢ HOCOCH, + NI \ \ i 6 cH ¢ # oH | C HOH C H OCOCH, H,0COCH, Glucose B-pentacetyl glucose 2 trichloracetyl 3,4,6 triacetyl 2 glycosyl chloride Poll, H-¢- Cl / / 3 HCl NH ¢ HOH L 7 J \ 3 POlg | HOCOCH, + Ago’_/NO, \ CH,0H . -—> CH,COC1 4G ROCOCHS (Walden inversion) - H : co eHg000CH ) (4,6 triacetyl _ °3)4,6 triacetyl - 6: glucogyl chloride - | N0o 4 oH 0 “OH CO i 4 if a OTTO glucuron secured fr H - Cy This derivative has veen acetylated and ~ and 8 - triacetyl glucuron have been separ lization from alcohol. glucuron will yield a stable bromo derivative, but we have J@ HOW C HOH | (CHgzC0)20 C OR Ni ‘oO “CH we glucuron tives now m ated by fractional crystal- Neither the A nor the -triacetyl om both derivatives a stable diacetyl chloroglucuron. /GHOCOCHS - ococH, __--—--—-} - > 90008 E ¢ cote,” . No triacetyl glucuron l-chlor diacetyl glucuron Though yet obtained in a crystalline state, these halogen acetyl derivatives give correct analyses. map TS TRE RE te em tet Le The syn- but the thesis has not as yet progressed beyond this point; preparation of these two chloro diac etyl glucuron deriva-~ ake possible the completion of the synthesis of the para amino phenol glucoside of glu hope to. accomplish in the near future. curonic acid, which we _ a¥: The Chemical Nature of the Type- specific capsular sntigen pneumococcu of Pneumococcus. (Drs. Avery and doebel). The importance of determinin of the type- g the chemical hature specific antigen of Pneumococcus becomes evident e substances in anti- when one considers that the protectiv s serum are precisely those antibodies which are lar antigen. By speci- produced in response to this particu fic union with these antibodies, the capsular substance of. the virulent Pneumococcus bocomes 80 altered that the bac- torial cell as a whole is rendered susceptible to phagocy- tosis. And, so far ae our knowledge at present goes, the therapeutic effect of antipneumococcus serum is largely, if not exclusively, dependent upon the presence of these type- specific antibodies. Indeed, the application of serum thera- py in Pneumococcus today rests upon our knowledge of these type-specific relationships. It would appear, therefore, that an understanding of the chemical nature of this Lpor- tant antigenic constituent of the Pneumococcus might contri- bute much to the improvement and perfecting of methods useful in the production of potent antisera. Interestingly onough, al though the exact nature of this antigen itself eludes us, we have considerable knowledge of one of its component parts, namely, the type-specific capsular polysaccharide. For instance, we know that this specific carbohydrate is given off from the bacteria growing in the focus, of disease and circulates in soluble form in the vody fluids; that in this form it still retains its capa- city to utite with and thus to neutralize the immune substan- ces of the serum} that even in minute amount Lt inhibits phagocytosis, and that because of these reactions its pre- sence tends to interfere with the normal processes of recov- ery from the disease. We know also something of the chemt- stry and {mmunological specificity of this capsular substance. wo H 113 We have succeeded in chemically identifying a polysaccharide in the capsular substance of each of the different types of Pneumococcus and have found in each instance that the carvo- hydrate is as chemically distinct as 4t is serologically specific for each type. We have demonstrated the remark- able specificity of these carbohydrates in the phenomenon of anaphylaxis and in the reactions they incite in the skin of convalescents from pneumonia, We have evidence that these specific carbohydrates, when separated from the bacteria, still retain unimpaired the property of combining with anti- podies, but lose more or less completely their power to in- cite the formation of these same antibodies when injected alone into the animal body. Tis very fact indicates that in their native state, these specific polysaccharides exist in the cell not as ha@tens but as part of a more complex antigen from which they may be more or less readily disso- ciated either beforo or after the injection of the whole antigen into the animal body. The rate and extent to which this dissociation occurs apparently varies with the differ- ent types. The theory of antigenic dissociation and its re- lation to the production of specific sera have been aiecuss-_ ed in a preceding report. It is referred to here merely to emphasize the importance of knowing the chemical nature of the type-specific antigen, and to point out again the prac- tical bearing which such knowledge might have upon the prob- lems of active and passive 4mmunity to Pneumococcus infec- tion. Progress in the solution of this problem has been oi E114 impeded by two major technical difficulties:- one, the diffi- culty of extracting the intact antigen in active form because of the ease witn which dissociation of the complex occurs, and the other, the difficulty of {dentifying in such extracts the particular substance or chemo-specific groups which con- fer antigenicity upon the capsular polysaccharide. The first of these aifficulties has been largely overcome, and the second has been attacked through an indi- rect but promising approach. Following the leads suggested by the work of Perlzweig and of Day, filtered solutions con- taining the type-specific antigen in active form have been obtained by extracting the encapsulated cells at an acid re- action with heat. The antigenic activity of the extracts has been tested by injecting them into mice and determining whether the treated animals have acquired specific {mmunity to infection with pneumococel of the homologous type. In an acid buffer solution (pH 6), the antigen resists heating in a boiling water bath for ten minutes; two injections total- ing 0.4 cc. of a solution containing as little as 0,003: mes. N per CCe have proved effective: in immunizing mice against. an homologous strain of Type J pneumococcus. The extraction of the antigen, however, is only the first step in the anal- ysis of 4ts chemical nature. Is tho type-specific polysac- charide bound to a protein, or to some other substance? Is 4t even necessary to assume that the effective complex con- sists of a protein coupled to the carbohydrate? This seems ‘ wo co the mast reasonable assumption, especially since we have at found that an artificial antigen prepared by the coupling of a diazonium derivative of the Type Ill polysaccharide with animal protein stimulates in rabbits the formation of type- a LAIST SATO specific antibodies which agglutinate living pneumococci of weep OBES BS nee aed the homologous type and protect mice against infection with virulent Type II!l pneumococci. Despite this analogy, may it not be that the- polysaccharide carrying the specific binding groups is but part of a still larger and nore complex carbo- hydrate rolecule fro. which the antigenic groups may be eas- Aly split off, leaving the antigenically inert but specifi- cally reactive polysaccharide intact. In the case of the pectins for example, partial hydrolysis suffices to eplit off pectinic acid from the complex and a polysaccharide then remains as 4 product of hydrolysis. As Wells points out, there is theoretically no reason why @ carbohydrate should not function as. an antigen provided 4t ie colloidal in nature and of sufficient molecu- lar size, However, to assume that the lack of antigenicity of the capsular polysaccharides 4 48 attriputabie to denatura-. nt tion in the process of. fsolation is to overlook the fact that the more ‘those substances are chemically purified the more specifically reactive they become and the more closely they an conform on chemical analysis to the theoretical values for al pure carbohydrates. The mere prediction of theoretical possibilities Ee IEE MEST OD ID t BAe H 116 however, is not paramount to the actual jdaentification of the true antigen. The amount of substance required for chenm- ical analysis is large and this is difficult to obtain. De- termining whother a given extract contains the effective an- tigen involves the process of {immunization and this requires time. It seemed possible, however, that an indirect approach and one which might furnish a clue as to the nature of the substance to be looked for, might be made vy determining if possible, what enzyme or class of enzymes would destroy the antigenicity of a vacterial extract known to contain the whole antigen. Whether for example, the antigenicity of an extract is destroyed by trypsin, or pepsin, etc., oF by li- pases or nucleases, etc., OF whether enzymes, such as pecti- nase, emulsion, and others which are known to hydrolize par- tially the more complex carbohydrates would destroy the ef- fective antigen without impairing the specific binding pro- perties of the polysaccharide. tn collaboration with Dr. Goebel this work ie now in progress, and the results thus far obtained are promising, but not as yet sufficiently established to warrant presenta- tion at this time. y. The Significance of _the Skin Dest with specific Polysaccharide as_a Guide to Serum Therapy in Pneumonia. (Dre Francis). During the study of the cutaneous reaction of pneu- monia patients to the type specific polysaccharides, 4{t was soon observed that all of the cases of Type I pnevmonia which recovered gave a positive reaction to the Typa I polysaccha- ride, The majority of Type I patients received specific se- rum therapy, while those sufforing from infections due to Type If and III Pneumococcus do not. Approximately only 50 per cent of tne Latter group reacts to the specific soluble substance of the respective types. In none of the patients was this reaction elicited before the onset of recovery, even when type specific therapeutic serum had been adminis- tered. Furthermore, dn certain {ngtances, When the patient still appeared sick and the problem of additional serum treatment was 4mmediate, a positive skin test was found to inditate recovery and that treatment could be safely discon- tinued. In the case of 4naividuals, in whom no positive re~- action Was obtained at any time during the course of the jliness, a fatal termination occurred, even though a subsi- dence of fever, sterilization of the blood of the invading organisms and a comparatively high concentration of serum antibodies were observed. Consequently a negative cutaneous reaction to the polysaccharide was thought to be an indica- tion for continuation of serum therapy. . The study has been extended with the idea of e8- tablishing the true value of this reaction as a guide to proper serum therapy. At present, data have been obtained in 35 cases of Type I pnevmonia. Tour of these were not treated with serum but recovered and gave positive skin reac~- tions; they serve as controls for the cases which received ae ee Chee 118 specific treatment. The results are tabulated below. \ Rocovered Died 4 Total Positive Negative| Positive | Negative | _ Nuaber! Skin Reaction Reaction|Skin Reaction Reaction | Serum Treated 31 25 i* 0 5 5 Untreated i | Controls ' 4 4 0 0 oO : *Empyema From the table it can be seen tnat of 26 recovered cases which received serum, only one failed to give a posi- tive response to an injection of the specific polysaccharide and in the exception a purulent complication had developed. In the remaining 25, all of whom gave 4 positive skin test, recovery took place despite such complications a8 sterile pleural effusions, delayed resolution, furunculosis, and re- current acute nephritis. In 5 fatal cases no skin reaction was obtained at any time. The extended observations have, therefore, con- firmed the impression that the skin test with type specific polysaccharide is a valuable guide in the serum treatment of pneumonia. Its value lies in the fact that, when positive, 4t indicates that recovery has begun and serum therapy may pe safely discontinued, but a negative respomse is an “andi- . cation that the administration of serum should be continued. “If, in the face of continued serum therapy with definite clinical improvement a negative reaction persists, a suppur- ative complication shoulda be suspected. If none is found serum therapy should be maintained, although the outlook is unfavorable. oe POD i Cred : H 119 VI. Studies on Natural Resistance and on the In- munity induced in Rabbits by Injections of R Pneumococci. (Dr. Francis). A study is being made of tho natural resis- tances of the rabbit to Type III Pneumococcus and of the form of immunity produced in rabbits by injections of R pneumococci. The method of Robertson and Sia, using serum- leucocyte mixtures, yas boen shown to be quite sensitive in detecting slight degrees of {mmunity. For this reason the * ae same method, as well as a modification employing whole hepa- rinized blood, Was adopted in the present study. Although tho work is in its early stages, several interesting facts have been revealed. While the normal rab- bit is resistant to the rabbit-avirulent strain of Type II! Pneumococcus, no pneumococeidal powers against this organism is demonstrable 4n the animal's blood. Against the R form of Pneumococcus, however, the normal rabbit exhibits the power to inhibit the growth of a large number of organisms. The blood of a rabbit containing anti-R antibodies, produced vy immunisation with R pnoumococel, has an even greater bac- tericidal power for the R organisms than has the normal ani- mal. In addition, it possesses & moderate but definite ca- pacity to prevent the growth of the avirulent encapsulated Type II! Pneumococcus. So far no experiments have been done with the blood of immune rabbits containing type specific antibodies. The state of immunity has been simulated, how- ever, by the addition of type specific horse anti-serum to the blood of the normal: rabbit. Under these conditions a or, ped E 120 marked pneumococcidal effect on the rabbit avirulent Type III pneumococci is also observed. Gertain conclusions appear justifiable. The nor- mal rabbit, resistant to the avirulent encapsulated Type Iil Pneumococcus, shows no evidence in the blood of a phagocytic mechanism for its disposal. In the R immune rabbdit this ca- pacity has been developed to a moderate degree, while the blood of the animal containing type specific immune podies has a still greater pneumococcidal power. VII. The Principles Governing the Agglutinin and Precipitin Roactions- (Dr. Francis). It was observed in the course of precipitin reactions that when a prozone Was pro- duced with an excess of soluble specific substance of the pneumococcus, the serum had also lost its capacity to agelu- tinate the homologous type-specific organism. This observa- tion brought further evidence to indicate the unity of agglu- tinogen and precipitinogen... Consequently the following e6x- periments were devised in an attempt to correlate the mechan- isms of the reactions of agglutination and precipitation. To an immune serum & suspension of the homologous type specific pneumococcus was added. After agglutination had occurred the compact mass Was placed ina ailute concen- tration of the specific polysaccharide of the same type. Under these circumstances precipitation occurred, with re- moval of the polysaccharide from solution, indicating that an excess of antibody was attached to the pacterial cells ser C hare H 121 and was able to unite with additional amounts of antigen, in this case the specific soluble substance. Now, if still more of the free soluble substance was added, the agzlutinat- ed organisms became dispersed and remained diffusely suspend- ed in solution. When the bacteria 80 dispersed were removed from the mixture by centrifugation they were found to be in a normal serological state as shown by the fact that they could be agglutinated again, but only by immune serum for the homologous type of Pneumococcus. Had antibody remained bound to them, the vacteria might conceivably aggl utinate spontaneously in heterologous serum, or in salt solution. This, however, Was not the case. Similarly, organisms which were added to a serum in which a prozone had been created with specific soluble substance were serologically unaltered. Tne resulta show that the type-specific precipitin and agglutinin reactions are dependent upon the reaction of antibodies with the same substance; in one case, present in the bacterial cell; in the other, chemically purified and separate from the bacteria themselves. The two forme of the type specific substance may be substituted for one another without altering the phase of reaction. Heidelberger and Kendall recently studied the pre- cipitin reaction quantitatively and concluded that the var- fous phases of reaction could be expressed in terms of the laws of mass action. That is, antibody and specific sub- stance combine in multiple proportions, depending upon the re - : » ce tea - : po ieee ta Te ART ART TY GENT DEES POI nee ee 2H H 122 concentration of the two reagents. The present studies con- form qualitatively with theirs, and it seems that the same laws govern both the agglutinin and precipitin reactions. VIII. Studies on Epidemiology. (Dr. Stillman). TEER Zt gO ES : crepe GEE: 1, Antibody response to immunization by different routes. The work on the epidemiology of pnevmonia is being pursued f along two lines. Data is being accumulated as to the deve- 4 Bem ete nS lopment of immunity following immunization of rabbits with aifferent types of pneumococes by various routes. The length of time that agglutinins and protective antibodies persist in the serum of rabbits {mmunized with heat-killed Type I pneumococci has been determined. Agglutinins soon disappeared from the sera of these rabbits. Protective an- tibodies, on the other hand, could be demonstrated in high concentrations in thoir sera for prolonged periods. It ap- pears that there is a definite relationship between the to- tal amount of antigen administered and the length of time antibodies may be demonstrated in the blood or rabbits. Furthermore, there is a close relationship between the route of administration of antigen and the character and persist- ence of the antibody response in rabbits. 2, Production of experimental pneumonia in ani- mals» Work is also in progress on the production of exper- {mental lobar pneumonia in laboratory animals. This has a two-fold object. Once the complicated mechanism of the fac- y tors necessary to cause an animal to localize pneumococcus infection is understood, we will be able to not only treat the disease of lobar pneumonia in man more efficiently but also know better what precautions to observe in order to escape the disease. Although experimental lobar pneumonia has been successfully produced in mice, the various factors entering into the production of pneumonia could not be stud- ted in such a small animal. Ever since Wadsworth claimed to have produced pneumonia in partially 4mmunized rabbits in 1904, it has been assumed that pneumonia could be induced in these animals. It has already been shown that pneumonic consolidation may be produced in partially immunized alco- holized mice by the inhalation method. But we have been un- able to cause partially {mmunized rabbits to localize the infection which develops following inhalation of virulent pneumococci. The inability of rabbits to localize the in- fection may possibly be due to the anatomy of the rabbits! lung. Pneumococei in rabbits rapidly pass through the pleura and cause empyema and pericarditis. There seems to be no at- tempt to confine the infection within the lung itself. IX. The Antipneumococcus Immune Reactions of Normal Hog Serum. (Dr. Kelly). It was reported by Bull and McKee in 1921 that the blood serum of normal chickens when injected into susceptible animals confers a notable degree of passive antipneumococcus immunity. ‘They also showed that the factor responsible for this immunity is associated with serum globulin, and that it is type specific in action. Pe BO e 2k BH 124 In 1924 Robertson and Sia by means of serum-leuco- cyte mixtures were able to demonstrate the presence of nat- urally occurring antipneumococcus opsonins in the vlood of resistant animals. By this technic the serum of normal pigs showed a high content of these opsonins. Sia in 1929 found that pig serum injected intra- peritoneally into mice affords these animals a marked degree of protection against pneumococcus infection. This antipneu- mococcus protective action was shown by absorption exneri- ments to be type specific in nature. At variance with the widely accepted theory that natural immunity is dependent solely on the cellular defense of the host, aré the two examples, nere cited, of a natural immunity to pneumococcus infection that is passively trans- ferable through the blood serum, and that appears to be type-specific in nature. The present study 4s an attempt to repeat the ex- periments of Sia, and to further analyze the antipnoumococ~ cus immune reactions of the serum of normal hogs and of other resistant animals. In confirmation of Slats results” it has been found that the serum of normal hogs when in- jected intraperitoneally confers upon white mice a marked degree of immunity to Pneumococcus, Types Iand II. i cece, the optimal protective dose of tne sorun, usually affords protection against 10,000 to 100,080 lethal doses of pneu- mococcus culture. Tne serum yields {ts maximal protective ee SRNR IT TT ni see SE ee or eee eR Rane i Eee PEE ee Soe Te Ie BH 125 action when injected four to eight hours pefore {noculation of the mice with culture. The degree of protection against Type If is consistently higher than that against Type I. The serum gives no protection against pneumococcus infection when injected subcutaneous ly- Ynder the same conditions the serum from other animals - rabbit, guinea pig, OX, and sheep, - when injected intraperitoneally, have conferred no protection on mice. The protective principle of hog serum fs associ- ated with the serum globulin obtained by precipitation with ammonium sulphate. Hog serum gradually loses its antipneu- mococcus protective action on standing, 80 that after two to four months in the ice-box the protective action 18 entirely, or almost entirely, lost. At room temperaturé this loss oc- curs more rap idly. In a paraffin-lined tube, under & vase- line seal the serum loses its protective power in about the game time 88 {t does when unsealed. The protective action ‘of the serum against Type I is lost usually in shorter time than that against Type It.. Phe protective power is consid- erably reduced by heating. the serum to 56° G. for. 30 minutes, and 4s entirely destroyed by exposure to a temperature of. e4e CG. for the same time. The protective action when lost, either through heat, or atanding, 4g not restored to the se- yum by the addition of @& small amount of fresh normal serum. The antipneumococcus protective principle for either Typ® I or Type II can be specifically absorbed. The SS ory wed H 126 quantity of pneumococci per volume of serum necessary for specific absorption, is small. The absorption requires only a short time for completion. Absorption of the serum with avirulent R pneumococci, and with bacteria other than pneu- mococci, docs not consistently alter the protective power of the serum. Absorption with charcoal and Kieselgur does not appreciably lower the protective power. Sia in 1922 noted that S pneumococci, when sensi- tized by large amounts of hog serun, show agglutination af- ter separation from the sensitizing serum by centrifugation. It has been found in the present study that hog serum agglu- tinates S pneumococci. The agglutinated pneumococci from a firm, thin disc similar to that produced by type specific antipneumococcus serum. In comparison with specific anti- pneumococcus serum large amounts of hog serum are necessary for demonstration of the agglutination. It is necessary also, except occasionally in case of Type II, to use Gate's ‘technic for this demonstration. ‘The agglutinins for Type II are present in hog serum in greater concentration, than are those for Type Il. Similar to the antipneumococcus protec- tive action in mice, the agglutinins for § pneumococci are specifically absorbable and are destroyed by heating the se- rum at 64° C. for 30 minutes. It has also been observed that avirulent R pneu- mococcl are agglutinated by hog serum. This agglutination is of typical R character. The agelutinated pneumococci peer Tee? Sp ah: ae pe Spee tog CRE Coen ae. om oP- epee. B 127 form a granular precipitate leaving the supernatant clear. The maximum R agglutinin titre is usually 1:250 to 1:500. So far no falling off of the R agglutinin titre has been seen even after the antipneumococcus protective action of the serum is lost. Although sensitization of living pneumococci with hog serum does not alter their virulence, heat-killed pneu- mococci when sensitized by hog serum provoke a very meagre specific antibody response in rabbits. With the concentrations used, no change in colony form has been brought about vy growing either S or R, pneu- mococci in dilutions of hog serum. The R pneumococci cul- tured in presence of the sorum grow in clumps. X. Pathogenesis of Experimental Intradermal Pneu- mococeus Infection. (Dr. Goodner). The intradermal pneumo- coccus infection in rabbits offers an experimental disease which bears many analogies to lobar pneumonia. The general nature of this symptom-complex has been studied for a num- ber of years and is now suffid ently well understood to serve as a basic method for more specific projects. With Dr. Rhoads the histological pathogenesis of this lesion has been studied and found to be comparable to that of lobar pneumonia in man. The findings are described in detail in another section. The outstanding feature of the early lesion 4e the edema fluid; its accumulation occurs before any change in the cellular picture, and in tne pro- woe aw 128 gressing or moving lesion its passage through the tissue precedes any other sign of reaction between tissue and micro- orgmism. It seems probable that the advancing fluid car- ries with it the first infecting organisms and consequently {noculates all tissues which it reaches. The resultant in- fection appears to be duc, not to an active invasion, but rather to a progressive passive 4noculation. Edema fluid has been removed from the developing lesion and its propertics studied in vitro. If obtained free from blood and tissue elements such fluid does not clot even after long standing. That this property is not due to a deficiency of fibrinogen 4s shown by the fact that a firm clot promptly appears on the addition of thromhoplastin. Bvidence points to a high antithrombic content since the edema fluid has the property of retarding the clotting of normal rabvit blood. ime clotting time of the blood in these rapoite infected with pneumococcs is prolonged, par- tieularly. in ‘Severe cases, but. a effect is never as ‘pro- nounced as 4n the exudative fluid where the “Goncentration of the antithrombic factor. appears to be ‘much ° greater. The na- ture and source of this antithronbic substance are now being investigated. | | . So A study has been made of the factors which have to do with the development, movement, final localization, and extent a the dermal lesion in the rabbit. The direction of movement is determined by gravity and is independent of the distribution and course of blood vessels and lymphatics. The lesion seems always to seek to reach the more directly dependent aréa- The direction of movement may be modified by changing the position of the animal or by arre ging ar- tificial guiding structures such as scars. If the skin of the abdomen is the final site of localization there is 4 magsive collection of edema fluid. Such edematous lesions are not obtained in areas where the skin is tense, such as over an extremity. Drainage distance largely determines the volume of edema fluid accumulating in a given area. Numerous factors modify the development and rate of movement. Perhaps the most striking of those yet observ- ed ia the accelerating or synergistic co-infection with Be influenzae. This seems of particular significance since _——- — some form of the preiffer bacillus is found in a large per- centage of sputa from cases of lobar pneumonia. If B. in- fluenzae and pneumococci (Type I) are injected together in- tradermally the effect is a marked acceleration in the de- velopment and movement of the lesion. The lag period of the pneumococcic infection is almost entirely eliminated. Other studies on the associative relationships of these organisms are now in progress. I.Intradermal Infection with a Rabbit Virulent Type III Pneumococcus. (Dr. Goodner). The character of the in- tradormal infection with: a rabbit virulent strain of Type III pneumococcus has been studied and found to follow essen- TE ST ere vas TMT eS neers 4 HZ Loe tially the same course as with Type I, although the Gisoase 48 much more severe and the fatality rate higher. Cases terminated more often by lysis than by crisis. This is pos- sibly explained by tne low grade active immunity which fol- lows artificial {mmunization or recovery from the disease. Recovered animals show but slight resistance to reinfection even at short intervals. The immunity acquired from a course of the disease is of sufficiont quantitative character to prevent deata after reinfection but not sufficient to pre- vent a severe infection. The use of Type IIT antiserum (by ourself anda by others)in treating experimental infections has yielded poor results.