A STUDY OF THE TOXIC PROPERTIES OF TUBERCULO- PROTEINS AND POLYSACCHARIDES By FLORENCE R. SABJN, M.D., FRANKLIN R. MILLER, M.D., CHARLES A. DOAN, M.D., AND BRUCE K. WISEMAN, M.D. (From the Laboratories of The Rockefeller Irtstitute fw Medical Research) (Received for publication, Jury 25, 1930) It was reported by White in 1928 (1) and confirmed by Sabin, Doan, and Forkner (2, 3)' that a polysaccharide isolated from the tubercle bacillus by Anderson (4) was toxic for tuberculous guinea pigs. With doses of 10 mg. the guinea pig either died within a few hours, in which &se there was a precipitous fall in temperature; or it survived and showed only a temporary fall in temperature with subsequent rise above the normal level. There was also a change in the blood cells characterized by a rise in the neutrophilic leucocytes and a fall in lymphocytes. Similar phenomena follow the administration of tuberculo-protein. Since the polysaccharide, as prepared, has a certain nitrogen content, and since all the protein preparations contain carbohydrate, one must resort to a biological titration in order to determine responsibility for the intoxication. By giving the polysaccharide and the protein to tuberculous animals, in decreasing doses, we have ascertained that the temperature reaction is due to the protein, since amounts of poly- saccharide which do not contain as much nitrogen as is present in a temperature-altering dose of the protein fail to elicit the change. In the course of our testing of fractions from tubercle bacilli, it has become clear that we must consider not only fractions from merent types of bacilli, according to the plan of the Research Committee of the National Tuberculosis Association (S), but also the.varying mani- festations of the disease in different species of animals. To show the relations of these Werent factors we have drawn up an outline (Table I). The attempt at any quantitative comparative estimation of the factors of the disease must be considered as only an approxima- tion, subject to marked variations in individual animals. For the 51 TUBERCULO-PROTEINS AND POLYSACCHARIDES F. R. SABIN, F. It. MILLER, C. A. DOAN, B. K. WISE&MN 53 , Tp)LE I Phenomenu of Tuberculosis as Rektd to Chemical Fractions from the Bacilli = I Production of lesions Phenomena of the disease CaseatiOtt Probably protein Hemorrhage Protein Tuber&a Non-specific connative tissue Wax and protein - 1 . - Phosphatide Relation of phos hatide to nti-p osphatidt `I, ? ? Protein Relation of chemical fractions from the bacilli to these phammena Protein and polysaccharide rypc of bacillus Fowls Guinea Piss Rabbits Monkey Rat Avian S (Petroff) Avian R (Petroff) BI andfrac- tions of &r + +++ +++ ++++ ++++ ++++ Reversed ++++ slight age Reversed ++ ++++ ++ Stlld Average Average Average Reversed ++++ Average I ++++ ++ Large / 1-y. .l .,J 1 ++++ ++++ h3e Reversed ++++ Extreme ACUte ++++ ++ Minimal +++ f + f ++++ ++++ -I f ++ +++ ++ Slight with phosphatide in normal ++++ Epithelioids f3led with fat 1 - iuong the upper transverse line of the table are listed the phenomena of the disease itself and ! normal and tuberculous animals as far as they are known. The lower part of the chart records th kactions from the bacilli. subacute and dUOUiC Acute or dUOtiC ACUte Chronic Protein 10-2.5 mg. and poly- sac&ride Protein up to .OOOl mg. Sensitive to 0. T. ++++ Lymph nodes ++ ++++ ++ he second line are recorded the effects of chemical fractions isolated from tuberde bacilli on ,arying manifestations of the disease in animals, viewed in the light of studies with chemical 54 TTJBERCULO-PROTEINS AND POLYSACCHARIDES data on the temperature reactions in the monkey we are indebted to Dr. G. P. Berry whose work will soon be published. The phase of the subject upon which we have done the most intensive work (3) is that of the chemical factors causing the production of new tissue. Both lipoids and proteins in repeated doses have given rise in rabbits to the production of new connective tissue cells, but a phosphatide designated A-3 by Anderson (6) has proved to be so specific in the production of epithelioid cells and epithelioid giant cells, the essential elements of the tubercle, that we may speak of this element in the disease as the phosphatide reaction. In the tuberculous animal the phos- phatide is probably liberated only from the bodies of dead bacilli. The amount of the reaction to the phosphatide is governed by two different factors, first, by the number of dead bacilli, which can only be estimated; and second, by the amount of antiphosphatide produced in the disease (7,8,9). The most striking new material shown in the table is represented by the pre- hinary analysis of the effects of the dissociation of the avian strain of bacilli by Petroff (10, 11, 12). These two strains were given to us by Dr. Petroff for path- ological studies. As shown in Table I, the two dissociated strains of the bacilli produce different types of tuberculosis, that is, two different combinations of the manifestations of the disease. Roosters with the avian S strain had no tubercles large enough to be seen with the unaided eye, but under the microscope there were numerous tiny clumps of pure epithelioid cells, containing very large numbers of tubercle bacilli. Smears stained for tubercle bacilli had too many organisms per oil immersion field to be counted. From the small phosphatide reaction, we estimate that there had been a relatively slight death rate of bacilli. The animals had negative skin tests but became extremely emaciated. With the avian R strain, on the other hand, the skin test was marked and the loss in weight only nominal, while tuber&s were comparable in sixe to those seen in rabbits. Such striking diEerences in reaction to bacterial dissociation demon- strate the significance of this method in the further study of tuberculosis. A comparison of the reaction with the avian S in roosters with the disease in rats is also interesting. Both show a large number of bacilli, but in the case of the rat, we estimate an increased death rate of bacilli from the extent of the phosphatide reaction, the rat finally dying from the amount of epithelioid cells in the lungs rather than from any toxic effect. The detailed account of the pathological di.iIerences with dissociated strains will be published subsequently. The data in Table I suggest, for the further testing of chemical fractions, the selection of animals showing certain extreme reactions, such as the loss of weight after the Petroff avian S in roosters, or the caseation in monkeys. The hypersensitivity of the guinea pig to tuberculo-protein has been the reason that this animal has been used so extensively for the study of allergy, Through the studies reported F. R. SABIN, F. I(. MILLER, C. A. DOAN, B. K. WISEMAN 55 here, it seems probable that the temperature in tuberculosis as indi- cated in Table I, is related to proteins liberated from the bacilli. The tuberculo-proteins give a rise in temperature in the normal animal but the reaction is increased in the sensitized animal. The two types of temperature curves, one the fall which precedes death, and the other the reaction from which the animal recovers, are shown in Charts 1 to 7 for both proteins and polysaccharides. The results of .the titrations of these two substances in tuberculous guinea pigs are shown in Tables II and III. Most of the experiments with the protein have been made with a preparation from the H. K. Mulford Company, designated MA-100, which gfves an active skin test in tuberculoua guinea pigs hi doses of 0.0001 mg. One guinea pig received the Protein 304 and another an alkali-soluble protein, both prepared from the bacilli by Johnson, Coghill and Renfrew (13,14). Four preparations of tuberculo-polysaccharides have been used. First, a sugar isolated by Anderson (15) from the water separated from the ether-alcohol extraction of lipoids from H-37, and designated A-8; second, an analogous sugar from the bovine strain; third, preparations from the Bacflli H-37, prepared by Heidelberger, Hbg 511 and 514; and fourth, similar preparations made by the H. K. Mulford Company, MB-200 (16). The guinea pigs used in these experi- ments were all tested with tuberculin (0.02 cc. 0. T. Saranac) and were negative before the inoculation with tubercle bacilli. They ah received l,OOO,OOO counted organisms (l/50 mg.) of a 10 day culture of human Strain H-37 obtained from Dr. S. A. Petroff. They were inoculated either intraperftoneahy or subcu- taneously in the groin, as shown in Tables II and III. The shin test was positive to 0. T. in each instance before the injections of protein and polysaccharide were begun. The rabbits were given 0.5 mg. of the Strain B-l undissociated, intravenously from a 13 day culture. The experiments with the proteins and the polysaccharides were begun about 6 weeks after inoculation and were continued through 4 months. One (R 1344) of the 20 gufnea pigs of the first series (inoculated Feb. 14, 1930), died 18 weeks after fnocufation without having received any injection, and showed caaeous fnguinal lymph nodes and extensive tuberculosis of liver, spleen, omentum and lungs. Other anfmals that died following the injection of both protein and polysaccharlde have shown marked lesions of the liver and spleen. .In every case, with one exception, the routine was to take a rectal temperature first, then the specimens for the blood count and then give the injection. The exception was the iirst experiment on Chart 3, Guinea pig R 1323, in which the injection was given first. The amount of fluid injected was in each instance 1 cc. The poly- saccharide and the water-soluble Protein 304 were dissolved in freshly distilled water. The alkali-soluble protein was dissolved in dlstllled water and then DO!% ml. 10 5 2.5 1 17 2/14/3( Mulford R 1329 (i.p.) MA-100 9 2/14/3( R 1321 (i.p.) 40 3/22/3( R 1398 (i.p.) 41 3/22/d R 1399 (i.p.) 48 3/22/3 R 1406 (i.p.) 3 2/i4/3 R 1315 (i.p.) " `I " `L 304 TABLE II Effects of Tubem&-Proteins in Tubercuhu Guinea Pigs Type of protein mjected Date .of n jpm pc$e+ of protein mpctmoa -- :/11/30 l/10/30 i/ 5/30 i/ 5/30 i/ 5/30 i.v. i.p. i.p. i.p. i.p. t/25/30 In tra- testic- ular -10" -7" -3O fl" -1.3" +l" -3" Blood count I I before infection. if White blood cells per c. mm. iii 97c `651 03( - - 17ti - ii - 63: - LY - 13! 28! la! - - 221 - 321 27( - - 34 - llood count before ejection of protein. White blood cells per c. mm. - `MN - LY i- 6985 - - l651 - - - 153' - - 10% - - - 104! - - 274 - - - 590. - Result Death in 2f hours Death in 16 hours Death in 51 hours Death withii 24 hours Death within 24 hours Survival 0.1 0.02 0.01 0.001 30 2/14/30 R 1342 (s.c.) 16 2/14/30 R 1328 (i.p.) Alkali soluble protein (John- son) Mulford MA-100 3 2/14/30 R 1315 (i.p.) 30 2/14/30 R 1342 (s.c.) 31 _ 2/14/30 R 1343 (s.c.) 31 2/14/30 R 1343 (s.c.) 36 3/22/30 R 1394 (s.c.) 30 2/14/30 R 1342 (s.c.) 7 2/14/30 R 1319 (i.p.) u `8 " `C 3/25/30 Intra- testic- ulaF 4/16/30 i.p. 4/21/30 i.p. . 4/22/30 i.p. 4/22/30 i.p. 3/25/30 Intra- tt?&c- uiar 4/23/3Q i.p. 4/25/30 i.p. 5/ l/30 Intra- denaai f4.5" -1.8' +2.2" -3.9" - .8O +3o -5O -1.5" +3.40 +2.9" +3.2" +2.5O -.7" j-3.7" Negative -1.3" +1.40 588333301554 5200 588333301554 5200 91525616~197121888 507 Survival 41465934 1380 2782 1575 735 8532 756 1404 41465934 1380 2782 I` 1768 1224 340 2366 1820 364 3960 176 110 1768 1224 340 2366 0 5883 3330 1554 12198 147 5883 3330 1554 12198 14766 2568 19276 2254 483 ,, 67322856 408 5016 5700 684118081065 1152 67322856 408 5016 5700 684118081065 1152 " 6732 2856 4Of3 12243 6237 4389 14700 2952 980 6732 2856 4Of3 12243 6237 4389 14700 2952 980 ,` --- --- 1527 1560 162 4972 526 396 1527 1560 162 4972 526 396 `4 58833330 1554 5475 9417 3285 16942 1576 985 58833330 1554 5475 9417 3285 16942 1576 985 " 1260 714 126 4380 1440 180 7296 498 1462 1260 714 126 4380 1440 180 7296 498 1462 `I * These are serial numbers of the work of the department covering a term of years. TABLE II--CofKh&d WK. 0.0001 O.OOOC 31 R 134 3 R 131 30 R 134 I IVliifOKl I/14/30 ~-100 4/25/30 i.p. (S.C.) t/14/30 " 6/ 4/30 i.p. !%J " /6 n/j01 ip. 16 z/14/30 (( R 132 0. p.) 11 z/14/30 u R 132 0.p.) 20 2114130 " R133 (i.p.) 7 2/14/30 " R 131 0.p.) 51 2130 i.p. 6/U/30 i.p. 6/10/30 i.p. 6/10/30 i.p. PMN Ly M PYN Ly * Negative 98914867 785 9398 76204318 2 F. a. SABIN, F. R. MILLER, C. A. DOAN, B. IL WISEMAN 59 enough 1 per cent NaOH added to make the solution neutral to litmus. The Mulford Protein MA-100 was received in sterile salt solution, IO mg. per cubic centimeter, and it was diluted with freshly made, sterile salt solution. Kjllilzg Power of Tuberculo-Protins and Polysaccharides in Tubercdous Guinea Pigs As is shown in Table II, all five tuberculous guinea pigs which received from 10 to 2.5 mg. of the protein died within 24 hours; and all receiving less than 2.5 mg., seventeen injections in all, survived. Thus the minimal lethal dose for tbis protein lies between 2.5 and 1 mg. iu tuberculous guinea pigs. A report of the killing power of protein prepared from the filtrate has been given by Seibert (17). In Chart 1 is shown the type of temperature curve following a lethal dose of the protein, for Guinea pig R 1329. The injection was intravenous; the animal showed irregular and rapid breathing within 2 minutes of the injection. Subsequently there was involuntary defecation. The temperature fell precipitously 10'. The irregular breathing continued until death 41 hours later. The postmortem e xamination showed the omentum, spleen, and liver massively involved with tuberculosis; the lymph nodes were markedly caseous; the lungs had a moderate number of tube&s. There were numerous fresh hemorrhages especially marked around the tuberculous lesions in the omen- turn and in the lung. The lungs were markedly congested. 60 TWERCULO-PROTEINS AND POLYSAC-ES A more gradual fali in temperature, 7" in 7 hours, is shown in Chart 2, of Guinea pig R 1321, which received 5 mg. of the Protein MA-100 intraperitoneally. . The animal died in 16 hours after the injection. The postmortem examination showed extensive involvement of omentum, spleen, liver, and inguinal lymph nodes; there were a few tuber&s in the lung. Fresh hemorrhages were not apparent at autopsy, but in sections there was marked congestion of the vessels of the lung and some hemorrhage into the tubercles. It will be noted in the chart that there was a fall in neutrophiic leucocytes, lymphocytes, and monocytes during the 7 hours that counts were made. All but one of the animals, R 1399, that died within 24 hours after the injection, had the marked fall in temperature. 6WO - * VXQ- 4cm- 3ooo,- 2owL loM)- I I Th- 1 Feh 5 13 14 Ibc4llpr8 9 10 11 1930 bA.RT 2 In this instance the changes in temperature were but slight; there was a drop of a degree and then a rise to the original level; the temperature had begun to fall when last taken and the animal died in the night. All of the animals that showed the fall in temperature were obviously quite sick, showing a marked toxic effect of the protein. The other animal that died after protein showed the same extensive involvement of liver, omentum, spleen, and lymph nodes. The lungs had a few tuber&s and were markedly congested and in one case (R 1406) hemorrhagic. The polysaccharides had also a certain kiiling power in sensitized guinea pigs, though it was by no means as consistently related to dosage as with the protein. In the original studies of Sabin, Doan, and Forkner (2, 3) six tuberculous guinea pigs received 10 mg. of the Polysaccharide A-8 intraperitoneally; two died within a few hours, two during the night after the experiment, while the others survived. F. R. SABIN, F. R. =-R,` c. A. DOAN, B. K. WISER 61 The plysaccharides in the present series did not show as marked a l;iKng power for guinea pigs. *is shown in Table III, one guinea pig, R 1324, died in 4f hours after an intra- Pritoneal injection of 20 mg. of the A-8. The liver, spleen, omentmn, and lymph nodes were on& moderately involved; the lungs were markedly congested; but no hemorrhages were found in sections. This animal had survived two in+&- on the 2 preceding &YS of 10 WC. each of this polysaccharide, None of the animals receiving 10 mg. died; but R 1316 which had an intraperitoneal dose of 5 mg. of the Heidelberger polysaccharide died in 19 hours, the fall in tem- perature and the changes in the blood cells being shown in Chart 1. At post- mortem there was extensive involvement of liver, omentum, and spleen. The lungs showed many tuberdee, but were only moderately congested and had no hemorrhages. It is important to consider whether there was sufficient protein in this dose of the polysaccharide to account for the killing power and for the fall in temperature. The problem with the polysaccharine is complicated through not knowing the actual state of the nitrogenous contamination or its potency. Dr. Heidelberger found the nitrogen content of this sugar to he 0.34 per cent, 90 that if ah of the nitrogen were calculated as protein, 1 mg. of polyssccharide would contain 0.021 mg. of protein; and 5 mg. would have only 0.105 mg. protein, wh&$ is below the minimal lethal dose of the protein whose potency we have been testing. On this basis it would take 100 mg. of the polysac&&de to cone 2.1 mg. of protein, 2.5 mg. having killed a guinea pig (Table II). The question of a possible killing power of the polysaccbaride here employed needs further study, with a larger series of experiments; but these experiments suggest tbat it either has a certain killing power in sensitized animals or it may enhance the killing power of a dose of protein too small to kill by itself. It should be stated tbat the manner of death is exactly like that of the lethal action of the protein. In relation to the fall in temperature, on the other hand, it is interesting to note that the animal (Guinea pig R 1315, Table II) which received 0.1 mg. of the protein, t&e computed amount in the 5 mg. of polysaccharide, showed a fkl drop of 5" in temperature, so that if the fall in temperature in Guinea pig R 1316 (Table III) which succumbed to the polysaccharide was a direct eflect of the substance injected rather than an indirect toxic effect on the animal, it might have been due to the protein introduced with the sugar. It will be noted that one other animal which received 5 mg. of the sugar, R 1317, showed a fall in temperature and died 4 days later. D@Sl 5 I-I- 12 2/14/30 A-8 original 3/13/30 i.p. R 1324 (i.p.) (Anderson) 31 2/14/30 Bovine (Au- 4/17/30 i.p. R 1343 (SC.) de&on) 15 2/14/30 R 1327 (i.p.) 12 2/14/30 R 1324 (i.p.) 12 2/14/30 R 1324 (i.p.) 15 2/14/30 R 1327 (i.p.) 15 2/14/30 R 1327 (i.p.) 18 2/14/30 R 1330 (i.p.) A-8 purified 3/12/30 i.p. (Anderson) A-8 purified 3/11/30 i.p. (Anderson) A-8 original 3/12/30 i.p. (Anderson) Heidelberger 3/21/30 i.p. mg 514 Bovine crude 3/13/30 i.p. II (Ander- -=I Bovine crude 3/13/30 i.p. I (Ander- =n) 11 2/14/30 R 1323 (i.p.) Mulford MB- 4/ 8130 i.v. -0.9" +4.2" 98914876 98914876 785 No couuta 785 No couuta No No counts counts 200 I I TABLE IU Effects of Tubercdo-Po~yzccharid~ in Tuberculous Guinea Pigs Effect on t.empemtLue I I I I Blood count Blood count before injection of Maximum change before infection. in cells folloyi~ White blood cells &e$& &zi$$&& per e. mm. pcrC.lIUU. perc.mm. B 3 4 ii a 2 a 2 3% E 3% E a a 3 % 3 % _---- _---- ----- ----- 26182796 476 5400 7501125 5525 378 252 26182796 476 5400 7501125 5525 378 252 6732 2856 408 8008.5434 572 20976 456 6732 2856 408 8008.5434 572 20976 456 0 0 -3O +4.10 +2.7" -0.8' +3.2" -2.0" +2.g0 -1.2O +1.4" Negative +0 .6" +2.7" 5376 7056 2352 12954 4000 2095 5760 976 525 5376 7056 2352 12954 4000 2095 5760 976 525 26182796 476 522628211202 6732 624 288 26182796 476 522628211202 6732 624 288 26182796 476 4416 1449 966 2NHO 412 8800 26182796 476 4416 1449 966 2NHO 412 8800 5376 7056 2352 2300 5635 3220 6240 8.55 380 5376 7056 2352 2300 5635 3220 6240 8.55 380 5376 7056 2352 14418 1424 1246 2860 553 220 5376 7056 2352 14418 1424 1246 2860 553 220 21601480 21601480 240 586s 1870 595 8733 817 30 240 586s 1870 595 8733 817 30 Radt Death in 4f ~ houn Survival I, u " `I I` u 5 2.5 1.0 0.1 0.01 4 2/14/30 R 1316 (ip.) 7 2/14/30 R 1319 (i.p.) 5 2/14/30 R 1317 (i.p.) 6 2/14/3O R 1318 (i.p.) IO 2/14/30 R 1322 (i.p.) 7 2 /I4130 R 1319 (i.p.) 19 2/14/30 R 1331 (i.p.) 11 2/14/30 R 1323 (i.p.) 20 2/14/30 R 1332 (i.p.) 7 2/14/30 R 1319 (i.p.) Heidelberger 4/ 2/30 i.p. Hbg 514 Mulford MB- 4/ 7130 ip. 2cul Mulford MB- 3/31/30 i.p. 200 I I Heide&erger 41 4/30 i.;. Hbg 514 Mulford MB- 4/14/30 i.p. 200 Heidebrger 4128130 i.p. Hbg514 Heideberger S/22/30 ip. wg 514 Heide.lberger 5/29/30 i.p. Hbg 514 Heidelberger S/23/30 ip. Hbg 514 Heidelberger 6/ 3/20 i.p. Hhg 514 zhfrt blood -4.5" 4160 1728 448 10803888 216 15164 6244 669 +4.5' 1260 714 126 2982 1008 126 5172 232 1170 -3.7O 2K'S2709 516 5775 1750 1225 9625 5161625 +2.1" 3168 1485 198 3136 1617 98 7830 130 720 -4.8O +4.7' 58564309 773 8532 5372 173826080 1810 530 +1.8' 1260 714 126 33211968 738 4009 594 346 With tiesondary rise 4129 +2.3' 1971 1022 584 46745535 615 9130 2300 1660 Negative 98914867 785 13296 9418 2770 9747 2565 3591 +2.3" 1026 3564 810 5960 7003 1937 7802 1034 470 -1" +l" 1260 714 126 3752 728 1064 8034 8241960 Death in 13 houra Survival Died 4 day latex sunrival I` `I `I 64 TUBERCULO-PROTEINS AND POLYSACCHARIDES At the postmortem examina tion, extreme involvement of liver, spleen, omentum, and peripheral lymph nodes wss found together with many tuber&s in the lungs. There were extensive hemorrhages. They involved many of the lymph nodes and some of the organs, as well as the vessels in the skin and in the subserosal layers of the peritoneal lining. It may be that a sustained fall in temperature after the injection of either protein or polysaccharide indicates, even if the animal survives 24 hours, that it is not likely to survive the disease for many days. In general the animals which died after both protein and poly- saccharide had extensive involvement of the liver, spleen, omentum, and lymph nodes, and a moderate number of tubercles in the lungs. They all showed congestion of the vessels of the lung but no edema; most of them had hemorrhages around the tuber&s in the lung and around the caseous areas in the omentum. The lymph nodes were congested. Reactiom to Sublethal Doses of Protein and Polysaccharide ilz Tuberculous G&ma Pigs The animals which survived the injection of either of these sub- stances did not show any marked effects except on the temperature and on the blood cells. They were not clinically ill. The results were the same for both substances regardless of the route of injection, whether intravenous, intraperitoneal, or intratesticular. The type of the changes in the blood cells and the temperature are shown in Tables II and III; but the type of reaction is better indicated by curves ss shown in Charts 3 to 5. A typical temperature reaction involved a preliminary fall lasting from + to 2 hours, followed by a gradual rise of from 2 to 4.5' with a return to the original level in approxi- mately 5 hours. The positive reaction was not only a change in temperature but a specific type of curve. The fall occurred at once after the injection; it was, however, sometimes much shortened or even suppressed, but the subsequent rise and return to the original level were constant. The sublethal reaction for the polysaccharide is shown by the first curve in Chart 3, R 1323, the animal having received 5 mg. of the Mulford polysaccharide intravenously. The injection was made before a prehminary temperature had been taken and no blood counts were taken. The animal had a slight chill but no other symptoms. Similar curves for the different proteins are shown in Chart 4, R 1342. The animal received an injection of 1 mg. of the alkali-soluble protein (Johnson) intratesticularly 5 weeks after inoculation. It was highly sensitized; the temperature fell slightly for an hour and then rose 4.5' by five o'clock. It was still above 103" the next morning, the original level having been 102'. It will be noted that the leucocytes rose in this instance while both lymphocytes and monocytes fell. IIll I I I I I I I I I I II I I I I I I I I I lb57 !a l30 ml OtrJJiw 910 910 lb36 Ml Ml 93l llJ0 No m BM F&S 13 Wr4 8 grrs;a 28 2Jm &? - 14000 PllN - 13000 - l2000 - 11000 104. - 10000 10s L - 9000 102' '\ - 0000 \ - 7000 \ - 6000 \ - 5oMl - Km --o -3ooo 2oM 1000 I I I I I II I I II I I I I l l I l l l l I l l P 0SART3 td F 4 R E 66 TUBERCULO-PROTEINS AND POLYSACCHARIDES Bidogitxd Titration of the Tubercdo-Protein and Polysacchade in Tubercdous Guinea Pigs. -Since the reaction of the temperature and of the blood cells to these t&o substances was the same and since each was contaminated with the other, it was necessary to see if one substance could give the reaction in such small doses as to exclude the other. The experiment with titrations of these two substances is shown in Tables II and III. All but one of the guinea pigs &eiving injections of the protein of from 1 to 0.9001 mg. gave the &amcte&tic temperature curve and fall in lymphocytes. The one negative reaction was in Guinea pig R 1319, Chart 5, following the injection of 0.001 rug. of the Mulford protein given by the intradermal route. With the next dilution, containing 0.00001 mg. of protein, the ,=&ion became varied; two were negative, one moderately and one markedly positive. This dosage thus approaches the limits of the reaction to the protein. On the other hand, fn Table III it is clear that the inconstant temperature reaction was obtained with the dose of 0.01 mg. of the polysaccharide; one animal was negative, a second showed a slight reaction and the third had a rise of 2.3". On the basis that 1 mg. of the Heidelberger polysaccharide contains 0.021 mg. of protein or protein degradation products, 0.01 mg. would contain 0.00021 mg. of protein, which is twice the dose of protein which gave a consistently positive reaction. I There was probably enough protein, or its degradation products, to account for the temperature reaction in all of the experiments with the polysaccharides, provided the protein was in an active form. Variations in Temperature Induced in Tuber&o&s Guinea Pigs by Protein and Polysacchade Certain interesting points can be made out in this connection. In Chart 3 it will be noted that the guinea pig was highly sensitive to polysac- charide (protein) as shown by the first temperature reaction after inoculation with tubercle bacilli. During the months of May and June the control temperature curves were constant. These control periods are an important part of the experi- ment for they show that there is for the most part relatively little fluctuation in temperature from the disease itself. In only one snimal of the entire series was there an afternoon temperature during one of these control periods; in this animal, R 1332, on one occasion the temperature was steady for the morning hours but rose suddenly 3.2' between one-thirty and two-thirty and then fell gradually to the original level. The cause of this rise was not determined. In Chart 3 it will be seen that the temperature remained normal both after 0.01 mg. of the Heidel- we 10s 104 10s 1OZ lOi' 106 low 6oa . ---l&-J $000 ("-., h &ART4 5000- SOUI- 4Ot@- 5ow- 2om- t IWO- '\ III III I I I II I i I I I III 11% \,m II I II II II I I II I I It ~50530 m lK!il J40 3% . B;y) ml 930 lm 16 rl 16 28 Ml $30 y No .yll 350 850 lm .M $30 30 . CE~RT 5-Continucd0n Next Page to-r- - --- d M E 70 TKJBERCULO-PROTEINS AND POLYSACCEARIDES berger polysaccharide and after 0.00801 mg. of the Mulford protein; but onboth of these days, the lymphocytes fell markedly and the neutrophilic leucocytes were irregular. Thus the reactions on the blood cells are not to be correlated with the change in temperature. Chart 4 represents a long experiment with Guinea pig R 1342. The first experiment shows that there was no rise in temperature, but rather a fall after an injection of 1 cc. distilled water intraperitoneally. The water was freshly distilled from glass, boiled and used as soon as sufiiciently cooled. A second control with distilled water in another guinea pig, R 1332, showed a steady reaction. The intratesticular route of injection was employed m R 1342 to study the tissue reaction as reported by Long (18). 21 days later the testis was removed and showed the marked degeneration desaii by Long. By the twenty-first of April the control temperature was moderately steady, and the following day the animal gave a positive reaction to 0.02 mg. of the Mulford protein. The rise in leucocytes and the fall in lymphocytes were extreme; it is interesting that the lymphocytes had returned to 8000 cells per cubic milli- meter by the next morning, though not to the original level On the twenty-fourth of April a test was made with salt solution which had been sterilixed and sealed several days previously; there was a slight rise in temperature in the afternoon which did not occur with freshly made salt solution. The rest of the experiment had to do with the minute doses of the protein, 0.001 and 0.0001 mg., with less reaction in temperature to the smaller dose. In the intervening days, it will be noted that the level of the temperature of the guineapigwasadegreehigheronthethirdof Junethanontheprecedingday. This phenomenon has occurmd both with normal and tuberculous guinea pigs. The mean temperatures for New York City, together with-three readings for humidity, reported at 8 am., at 12 noon, and at 8 p.m., by the Weather Bureau, United States Department of Agriculture, are included in the charts. In Chart 4 it will be noted that with a rise of 6oF. between June 2 and 3, tbe temperature of the guinea pig was a degree higher on the second day. Comparing the humidities on June 2 and 3 it will be noted that on June 2 there was a fall of 13'points between 8 a.m. and noon, and on the next day a rise of about the same amount, namely 15 points, so that the change in temperature of the animal in this instance has run parallel with the change in atmospheric temperature rather than with the variation in humidity. Chart 5 represents another long series of experiments. This guinea`pig, R 1319, was markedly sensitive to an injection of 5 mg. of the Mulford polysaccharide on April 7, about 7 weeks after infection. The original control temperature was somewhat irregular, as frequently happens the first thne temperatures are taken. The change in humidity was not great. It will be noted that the temperature of the guinea pig was entirely steady on the day after the injection, but on the 14th and 15th of April was irregular and ran nearly 2" lower on the 15th. There was a drop in atmospheric temperature of 11oF. The temperature of the animal became F. R. SABIN, F. R. MILLER, C. A. DOAN, B. K. WISEI&AN 71 quite level again and on the 28th of April, it showed an average temperature reaction to 1 mg. of the Heide-lherger polysaccharide. The next day there was a ~condary rise, pra~tid~ a duplicate of the curve of the preceding day with no change in the weather, and for 3 days the temperature of the animal was at a TABLE IV ~ffec& of Tubercdo-Protean and Pd~sadaride on Nomwd Guinea Pigs m. 20 15 10 5 1 0.1 0.01 0.001 0.0001 O.ooool Numbu Dsteot & llljecti -- R 1433 6/ 6/3 R 1430 R 1432 6/ 6/3 R 1425 +1.3o f1.90 R 1425 R14.30 R1430 R 1430 4/30/3 R 1432 R1433 R 1436 R 1425 +1 so R 1435 5/ l/3 R 1432 5/ 213 +1.6" 0 +2.4.' R 1433 5/ Z/30 Negative R 14365/ l/30 + .5" +l" 66 63 48 32 R 14345/ Z/30 NegatiE 76 70 61 6.5 77 69 higher level than before the injection. We have seen no other secondary rise of this magnitude, but in three instances it has taken 3 days for the temperature to come down to the previous level Q 1323, Chart 3, after 5 mg. of Mulford poly- sac&ride; R 1331, (19) after 0.1 mg. of the Heidelberger polysaccharide; and Rabbit R 1365, Chart 8, after the Heidelberger polysaccharide). On May 1, R 1319, Chart 5, there was a negative reaction to an injection of 0.001 mg. of protein MA-100 given intradermally. On the third of June there was a slight 99'- 6030 - 5OQO- 4OMJ- 3C@O- ZOIM - rho. JON!- Illllllllllllllr I lllllllllllllll 930 lkJ0 l?il 330 WI ll3n Pso 330 JIM ll30 PJO 330 MO $30 IlQ MO 400 930 Ml Pa3 3% 4 AFEW 11 14 15 m 1930 F. R. SABIN, F. R. MILLER, C. A. DOAN, B. K. WISE116;AN 73 reaction to 0.01 mg. of the Heidelberger polysaccharide, but 7 days later the animal proved to be highly sensitive to the Mulford protein in a dilution of 10". It will be noted also that the temperature of the guinea pig during the control period of June 9, was lower by 2oC. than the initial temperature of June 3. Between these.2 days the environment diEered by lOoF. This guinea pig showed low lymphocytes thoughout the experiment but they fell with every injection except after the 0.01 mg. of the polysaccharide. Ejects of Tuberczclo-Protein and Polysaccharides om the Temperature of Nowd Gkzea Pigs The results of experiments with two types of fractions are shown in Table IV and Chat 6. For ail of the work the Mulford Protein MA-100 and the Heidelberger Poly- sac&wide 514 were used. It wiii be seen that with the normai animal the rise of temperature was .more moderate throughout than with the sensitised animai; the type of curve was, however, the same. It is likewise ciear that the biological titres of polysaccharide and protein exhibit the same relationship in the normal guinea pig as in the sensitized; however the necessary amounts for normai animals are 10 to 100 times greater. For the polysac&aride, the reaction was either negative or siight with doses of 0.1 and 0.01 mg.; while with the protein the iimit of reaction wss 0.001 mg. and the dose of 0.0001 mg. was negative. In Chart 6 it will be noted that during the control periods the temperatures were more irregular but were not at a lower level than with the tubercuious animais. There were more of these irreguiar control temperatures in normai than in tubercuious guinea pigs. EJects of Tuberdo-Proteins &d Polysacchardes an Tubercdous and Non& Rabbits Fewer experiments were made with rabbits than with guinea pigs. With two tubercuious rabbits, it wss determined that the protein from the Strain H-37 gave a rise in temperature in small doses. Rabbit R 1366 was given 1 mg. of the Protein MA-100 10 weeks after inocuiation with 0.5 mg. bovine Strain B-l; the temperature feii 0.5" and then rose 2"; 2 days later the reaction to 0.001 mg. of the same protein was similar but with a rise of only 1.2". Rabbit R 1367, in- ocuiated with tuberculosis at the same time as R 1366, was given 0.01 mg. of the same protein in 10 weeks and showed a fail of 1.2O and a subsequent rise of 2.1"; while 2 days later the rise in temperature after the injection of O.ooOl mg. was of the same magnitude, 2". After all four injections in these two animals, the leucocytea rose and the lymphocytes aud monocytes fell. Considering the difference in size of rabbit and guinea pig, the tubercuIous rabbit inoculated with the bovine strain of organisms proved to be quite sensitive to the protein from the human strain of tuberde bacillus. II II II I"' B 39 936 WJ la 336 WI IlM ml 360' 566 ' n F. R. SABIN, F. It. MILLER, C. A. DOAN, B. K. WISEM.AN 75 In two other tuberculous rabbits, R 1364 and R 1362, a comparison was made between the reaction of a polysacchan'de isolated by Anderson from the bovine organism, analogous to the A-8 from the human strain and the polysaccharide isolated by Heidelberger from the human shin. In both instances the reaction was pmct,ically identical and is shown in Chart 7 for R 1364. A moremarked rise in temperature following the Heidelberger polysaccharide was shown by another rabbit, R.1365, Chart 8. The chart illustrates VW wd the delayed return of the temperature to the oti@d already noted. co?ztrozs The polysaccharides, the Protein 304, and the alkali-soluble protein were all injected into the guht%t pig in 1 CC. water freshly distilled from glass. In the rabbits 2 cc. of fluid were used. The Protein MA-100 was received from the H. K. i\Iulford Company in sealed ampules in sterile salt solution. It was thus necessary to study the effects of these dihmnts. The freshly distilled water or salt was boiled and used as soon as sticiently cooled. Neither the water nor the freshly made salt solution gave any rise in temperature. The curves of temperature for distilled water and for a stock salt scduti~n are shown in Chart 4, and the records are given in Table V. In general both the distilled water and the salt solution did not give a positive curve in temperature. The distilled water was either negative or showed a slight fall as on Chart 4. The stock salt solution, which had been kept in sterile flasks gave a slightly more irregular curve, Chart 4, than the freshly prepared solution. In these experiments it has been found that control temperatures at hourly intervals should be taken for 2 or 3 days pmceding each experiment and after each experiment until the temperature reaches the or&al level. This may take 4 or 5 days. Concerning these control temperature reactions, two factors must be considered; fkst, the level at which the temperature runs, and second, the steadiness of the reaction during the day. Our charts show diEerences in level of as much as 2". This diBerence in the animal is probably related to differences in atmospheric temperature, which, of course, could only be tested when the animals were not in artificially heated rooms. The irregular temperature reactions are more difhlt to evaluate because more factors may be invokd. It is probable that variations in humidity a affect the temperature of the animal, but abnormal conditions, such as accidental infections must also be taken into consideration. Animal oumbcr 30 R 1342 14 R 1326 20 R 1332 30 R 1342 31 R 1343 57 R 1425 Date of ofection with ubercle bacilli Ifar 1,060,006 orgaoisma 2/13/30 2/13/30 2/13/30 2/13/30 2/13/30 Normal TABLE V Effects of Distilled Water and Salt Solution in Tuberculmu Guinea Pigs TypL$id Date>' Route in au00 (1 cc.) 1 of *II- o fluid j&ion -- Distilled 3/24/30 i.p. i.p. DiStilled HSO stock salt solution Fresh salt 0OlUtiOIl Distilled w i.p. i.p. i.p. i.p. Effect on temperature Irregular -1.9" +1.70 Negative +o.v Negative -0.7' +0.40 Irregular Al.00 Negative Negative Bhodco~co~nefom Blood count before Maximom change in cells foUowi white blood &Is in'ection of fluid. White blood cells injstion of flu1 % per e. mm. per e. mm. White bhd cells Result per c. mm. PMN Ly Mono PMN Ly Mono PbfN Ly Mono --------- 5883 3330 1554 10089 5564 1770 - - - Survival 9638 13806 208411712 4224 2880 14980 2041 1365 " 1026 3564 810 5746 1774 929 7169 1422 1148 " 5883 3330 1554 4260 6390 1988 15870 2760 1525 " 6732 2856 408 6840 3720 1320 11160 1878 552 " 3504~ 16791 19711 - 1 - 1 - 1 - 1 - 1 - 1 " F. R. SABIN, F. R. MILLER, C. A. DOAN, B. K. WISEMAN 77 Ejects of the Injection of the Tubetculo-Prohens and PoZysacchades on the Blood Cells In this series of experiments only immediate effects of the injection have been considered. The blood counts for tuberculous guika pigs after protein and polysaccharide are shown in Tables II and III. The total numbers of neutrophilic leucocytes, lymphocytes, and monocytes are giveb before the infection with tuber& bacilli, in the tirst cohnnn; in the second cohrmn are the corresponding 8gures for the count taken just before the injection of either protein or sugar; while in the third are recorded the greatest change in these three strains. This third column therefore does not represent a single count a~ was true for the first two cohmms. After the injection of the protein there was a rise in leucocytes in every guinea pig except the second snimal in Table II, R 1321, which died in 16 hours; this was the only one of the fatal cases in which ~~tmts were made. In every animal there was a fall in lymphocytes, which for the most part showed little or no tendency to recover; the monocytes have also fallen in every instance. but have shown a tendency toward a subsequent rise before the end of the 7 hours of the experhnent. The average fall in lymphocytes in tuberculous guinea pigs after the protein, computing from the percentages rather thanfromthetotalnumbers,was63percent. After the injection of the polysac&Gdes the total numbers of lymphocytes fell in every instance but one; this case was the last experiment in Table III, R 1319, shown in Chart 5, after the injection of 0.01 mg. of the Heidelberger polysaccharide, in which the lymphocytes were very low before the injection. In this animal, the total numbers of lymphocytes rose slightly but the percentages fell. In two instances, the percentage of lymphocytes rose while the total numbers fell; (R 1324 after A-8 on 3/13/30, and R 1327 after the bovine polysacckide on 3/13/30). Including these percentages in the average, the lymphocytes fell 45 per cent after the polysaccharides, as is shown in Table III. Six normal guinea pigs showed a fall in lymphocytes of 55 per cent in 7 hours after tuberculo-protein, and eight showed a fall of 65 per cent after tuberculo-polysac&&de. Non.-.!@eci$c Protein Fever It has long been known that the injection of proteins causes fever and a leucocytosis. In 1890 Buchner (19) showed that bacterial proteins injected subcutaneously gave a local reaction of aseptic pus. The mechanism of this phenomenon was then studied by Roemer (20, 21) who found that an intravenous injection of bacterial protein gave a leuaxytosk which was maximum in 8 hours. Gold- &eider and Jacobs (22) then found that leucopenla preceded the lencocytosis using an extensive series of stimuli, organ extracts, bacterial proteins as well as 78 TUBERCULO-PROTEINS AND POLYSACCEAIUDES proteins from other sources. Ameth (23) then demonstrated that 8 hours after the injection of peptone intravenously in rabbits, at the height of the leucocytosk, there were 3 per cent myelocytes in the blood stream, showing a replacement of leucocytes from the bone marrow. In 1923 Hussey (24) discovered that the same phenomenon occurred after the injection of a considerable series of salts, sodium chloride, sodium carbonate, potassium phosphate, lithium nitrate, and sodium sulfate, and stressed the fall in mononuclear cells, finding that it was about 70 per cent in 3 hours. More recently Beard and Beard (25) have followed the blood cells in eight rabbits every 10 minutes for 5 hours after the intravenous injection offrom1Oto15cc.ofsaltsolu~nvslyingfrom1to2.5percent. Bymaking the counts at such frequent intervals they demonstrated that all three groups of the white cells, leucocytes, lymphocytes, and monocytes, fall immediately after the injection and that the maximum leucopenia is reached in about an hour. At this time the neutrophilic leucocytes start to rise, while the lymphocytes and monocytes continue to fall. The monocytes then begin to rise but in their experf- ments did not reach their original level in 5 hours, while the lymphocytes were lower by approximately 2000 cells at the end of the experiment. This type of reaction is well shown for the tuberculous guinea pig, Chart 4, R 1342. By taking the count 2 hours after injection, the fall in the leucocytes was missed but the subsequent rise is clear. It will be noted that before each injection shown in this chart, the lymphocytes were high and showed a marked fall with the lowest level, with one exception, at the end of 7 hours. They had recovered in varying degrees by the next morning, but only in one instance did they exceed the level of the time before the injection. The monocytes varied somewhat in reaction in these tuberculous guinea pigs, but as shown in Chart 4, after a fall during the first 2 hours, they showed a tendency toward a temporary recovery in 5 hours with subsequent fall at the seventh hour. As is shown in Table V the eiIect of the injection of distilled water on the blood cells may be slight as in Guinea pig R 1326, or negative as in R 1332; while the injection of salt solution is followed by the characteristic changes already described. It is clear that the injection of proteins, salts and sugars has constant ekcts on the blood cells at quite specifk time intervals. There is an immediate leucopenia followed by a leucocytosis, in which the three strains of circulating white cells fall at difIerent rates and return at different times. The leucocytes start to return first; then the mono- cytes, and finally the lymphocytes. This reaction is elicited in the tuberculous animal by ail the proteins tested; it also follows the injection of polysaccharide in concentrations containing an amount of nitrogen too smali to cause a temperature reaction. The rise F. R. SAFIIN, F. B. WEB, C. A. DOAN, B. K. WISEMAN 79 h Ieucocytes in tuberculous animals is not consistently greater than is recorded in the literature, though several counts of over 20,000 pseudo-eosinophik 1eucoCytes are shown in Tables II and III. The f& in lymphocytes of 65 per cent corresponds with the 70 per cent for all mononuclear cells recorded by Hussey. The question suggests itself of whether the tuberculo-protein and the polysaccharides may not affect the proportion of monocytes to lymphocytes and so be concerned with the cellular factors of resistance to infection. S-Y The temperature reaction in tuberculous and normal guinea pigs and rabbits is elicited by the tuberculo-protein and probably not at all by the polysaccharides. The polysaccharides may have some killing power under certain conditions, but this is not as consistently related to dosage as in the case of the proteins. Both proteins and polysac- &rides cause a change in the white blood cells when introduced by any route. BIBLIOGRAPHY 1. White, W. C., Tram. Asm. Am. Phys., 1928,43,311. 2. Sabin, F. R., Doan, C. A., and Forkner, C. E., Tram. of the 24th Am. iKeeL N. T. A., 1928,253. 3. Sabin, F. R., Doan, C. A., and Forkner, C. E., J. Erp. Me& 1930,52, suppL 3,1-152. 4. Anderson, R. J., Ptm. Sot. I&*. Bid. and Me&, 1930,27,387, 5. White, D. C., N. T. A. Technical Series No. 9. 6. Anderson, R. J., J. Bid. Ch., 1927,74,537; 1929,85,351. 7. Doan, C. A., Trans. 2% Am. Meet. N. T. A., 1929,182. 8. Doan, C. A., Pm. Sot. I&$+. Bid. and Med., 1929,26,672. 9. Doan, C. A., and Moore, D., Trans. 26th Am d&et. N. T. A., 1930, 188. IO. Pet&f, S. A., Proc:Soc. Exp. Bid. ad Med., 1927,24,633,956,958. Il. Petroff, S. A., Branch, A., and Steenken, Wm., Jr., Ibid., 1927,25,14, and Am. Rev. T&m., 1929,19,9. 12. Petroff, S. A., and Steenken, Wm., Jr., J. l&p. itfed., 1930,51,831. 13. Coghill, R. D., J. Biol. Chess., 1926,70,449. 14. Johnson, T. B., and Renfrew, A. G., Am. Rev. T&em., 1928, b, 505. 15. Anderson, R. J., Trans. 26th Am. Meet. N. T. A., 1930, 181. 16. Masucci, P., McAlpine, K. L., and Glenn, J. T., I&i., 1930, 182. 17. Seiirt, F. B., Ibid., 1930,234. 18. Long, E. R., and Seyfarth, Ma&I.., Am. Rev. Tubem., 1924,9, 254. 80 TUBERCULO-PROTEINS AND POLYSACCHARIDES 19. Buclmer, H., Bd. Klin. Wochenschr., 1890,2?, 1084. 20. Roemer, F., Ibid., 1891,28,886,1189. 21. Roemer, F., Virch. Arch., 1892, 128,98. . 22. Golds&eider, A., and Jacobs, P., Zeitschr. f. Klin. Med., 1894,25, 373. 23. Ameth, J., Die Qualitative Blutlehre, Leipzig, 1920. 24. Hussey, R. G., J. Gen. Physiol., 1923,5,359. 25. Beard, L. A., and Beard, J. W., Am. J. PhysioZ., 1928,85,169.