W. H. MANWARING 1081 



EXPERIMENTAL TESTS OF THE EHRLICH THEORY 



It is evidently impossible to prove or disprove the Ehrlich theory by data drawn 

 from test tube experiments, since there is no apparent inconsistency or even paradox 

 in test tube reactions that cannot be harmonized with this theory, or for that matter 

 with any other theory, by a sufficient number of minor hypotheses. Proof or disproof 

 of this theory can come only from physiological studies of hypersensitive and immune 

 tissues. 



When one considers the prominent role the Ehrlich theory has played in the stimu- 

 lation and co-ordination of immunological research one marvels at the few attempts 

 thus far made at its physiological verification. During the last twelve years we have 

 been engaged in such work. The physiological reactions used in most of our tests have 

 been: (a) the sharp fall in arterial blood pressure during acute anaphylactic shock in 

 dogs, (b) the accompanying contraction of the urinary bladder, and (c) the marked 

 changes in perfusion resistances in hypersensitive organs, on the addition of specific 

 antigen to the perfusion fluid. The following are some of our t>'pical findings: 



a) According to the Ehrlich theory of antibody formation, the blood-free tissues 

 of hypersensitive animals have an increased affinity for the specific antigen due to a 

 local multiplication of "sessile receptors." This increased affinity is the assumed cause 

 of the specific hypersensitiveness. Yet repeated perfusions of blood-free organs of 

 hypersensitive animals with Locke's solution containing even minute quantities of 

 the specific protein used in sensitization give no demonstrable reduction in the con- 

 centration of this protein in the perfusate, as determined by specffic titration with 

 rabbit precipitin.^ 



b) According to the Ehrlich theory, the blood-free tissues of immune animals are 

 necessarily hypersensitive, at least during the early stages of immunization, due to 

 an increased number of "sessile receptors." Yet the blood-free lungs of protein-im- 

 mune dogs show no suggestion whatsoever of specific hypersensitiveness on perfusion 

 with Locke's solution containing the specific protein,- nor will the liver of a protein- 

 immune dog transplanted by blood-vessel anastomosis into a normal dog render the 

 normal dog hypersensitive.^ The liver of a hypersensitive dog similarly transplanted 

 renders the recipient tj^Dically hypersensitive.^ 



c) The Ehrlich theory assumes that specific immunity is due solely to circulating 

 antibodies serving as portal defenses to the fixed tissues. Yet if a protein hypersensi- 

 tive dog is exsanguinated as completely as possible, and its blood volume restored to 

 normal by transfusion from a protein-immune donor, the immune blood gives no de- 

 monstrable protection to the hypersensitive tissues as determined by an immediate 

 routine anaphylactic test.^ 



d) According to the Ehrlich theory, the only conceivable effect of immune serum 

 on normal tissues is sensitization of these tissues, due to a local "fixation" of specific 

 receptors. Yet the blood of a hypersensitive dog will render a transfused normal dog 

 typically hypersensitive, while fractional transfusion from an immune donor will give 



'7. Immunol., 2, 511. 1917. 



^J.A.M.A., 8s, 1729. 1925. ^Ibid., 13, 59. 1927. 



i J. Immunol., 13, S9- 1927. i Ibid., 10, $75- 1925. 



