Arthus was the first to show that the intradermal injection of horse serum in appropriately sensitized rabbits resulted in severe hemorrhagic and necrotic lesions in the skin areas injected (1). Later histopathologic studies have indicated that the characteristically severe and extensive tissue damage occurring during the course of the Arthns phenomenon is due to some form of vascular injury (2, 3), but there is little available information as to the mechanism by which this injury is produced.The macroscopic and microscopic appearance of the skin lesions of the Arthus phenomenon show certain striking similarities to those of the Shwartzman phenomenon, and the possibility has been considered that similar mechanisms may be operating in both cases (4, 5). The techniques employed for eliciting the two phenomena are ordinarily quite different; the Shwartzman phenomenon is produced by two injections of an appropriate bacterial endotoxin, one given intradermally and the other given intravenously after an interval of several hours, while the Arthns phenomenon is elicited by a single intradermal injection of antigen in a sensitized animal. It has been shown, however, that under appropriate conditions the Shwartzman phenomenon can be elicited by antigen-antibody interaction in ~ivo (5). have furthermore shown that bacterial products active in eliciting the Shwartzman phenomenon are also capable of greatly intensifying the local tissue damage resulting from repeated intradermal injections of antigen under circumstances leading to the development of the Arthus phenomenon. These demonstrations of the interrelationship between the two phenomena support the concept that the pathogenesis of the skin lesions in both cases may involve certain common factors which have not as yet been clearly defined.Recent studies on the mechanism of the Shwartzman phenomenon (7-9) have shown: (a) that skin areas prepared for this phenomenon exhibit a marked degree of aerobic glycolysis; (b) that this alteration in metabolism is largely due to the exudate polymorphonuclear leucocytes which migrate into such

In an earlier study (1) an attempt was made to determine whether there might be a demonstrable biochemical abnormality in skin prepared for the Shwartzman phenomenon. Investigation of the carbohydrate metabolism of such prepared skin sites revealed the presence of a striking and consistent deviation from normal, in the form of a pronounced degree of aerobic glycolysis. As a result of this atypical form of glucose utilization, such skin sites contain large amounts of lactic acid. It was suggested that the ultimate blood vessel disruption might be due to the action of proteolytic enzymes or "cathepsins," and it was pointed out that the occurrence of aerobic glycolysis provided local conditions favoring the action of such enzymes. Evidence was subsequeutly presented (2) to indicate that the observed aerobic glycolysis in prepared skin sites was largely due to the influx of exudate polymorphonuclear leucocytes, rather than to a direct alteration of the metabolism of intrinsic skin elements by the intradermally injected toxin.In the present study, attention was turned to the changes which occur following the intravenous challenging injection, in an effort to ascertain the nature of the mechanisms involved in the final blood vessel damage and the production of hemorrhagic necrosis. Evidence will be presented to indicate that the capillaries and veins in skin areas prepared for the Shwartzman phenomenon exhibit a marked vulnerability to a peculiar sort of cellular thrombosis induced by the challenging injection. The resulting damage to the involved blood vessels and the surrounding tissue elements constitutes a type of tassue damage which, it is believed, has not been previously described. Materials and MethodsAniraals.--Hybrid male and female rabbits, weighing 1500 to 2000 gin., from the Rockefeller Institute stock were used throughout the experiments. Litter-mate controls were used

No abstract

The intradermal injection of certain bacterial products in rabbits, followed after a suitable interval of time by the intravenous injection of similar material, results in an intense hemorrhagic-necrotic lesion at the site of the intradermal injection. This phenomenon, first described by Shwartzman (1), may also be produced by an intmdermal injection of a protein antigen in an appropriately sensitized rabbit, followed by an intravenous injection of bacterial products (2). Local infection with certain bacteria or with vaccinia virus also prepares the skin for hemorrhagic necrosis, which can be elicited by the subsequent intravenous injection of bacterial culture filtrates. The phenomenon thus provides a useful laboratory tool for the study of immune reactions occurring in vivo as well as of certain host responses to infectious agents. Numerous investigations of the nature of the phenomenon have been undertaken, but at the present time little is known of the basic mechanisms involved.In an earlier communication, it was shown that a characteristic metabolic abnormality exists in rabbit skin sites prepared for the Shwartzman phenomenon by the intradermal injection of a meningococcal culture filtrate (3). This abnormality consists of a marked degree of aerobic glycolysis, with a consequent local accumulation of lactic acid. It is known that such skin areas contain appreciable numbers of polymorphonuclear leacocytes as a result of the local inflammatory reaction to the bacterial filtrate (4), and it has further been shown that such "exudate leucocytes" exhibited a high degree of aerobic glycosis (5). Because of the difficulty in enumerating the inflammatory cells in Shwartzman-prepared skin sites, however, it was impossible to determine whether the over-all metabolic abnormality could be explained simply on the basis of the leucocyte content.In 1948 Becker (6) reported that the Shwartzman phenomenon could be

The prominence of thrombosis in the pathogenesis of the Arthns phenomenon (1, 2), the alterations in the clotting mechanism seen in experimental anaphylaxis (3), and the occurrence of "fibrinoid" in lesions generally considered to be due to hypersensitivity suggest that the interaction of antigen and antibody in vivo may in some manner initiate coagulation. Experiments to be reported here indicate that such an effect can, in fact, be demonstrated in vitro under appropriate conditions. Although the mechanism of this phenomenon is not yet clear, it seems worth while to report the basic observations because of their apparent pertinence to various irnmunopathological problems. Materials and MethodsAntigens.--The antigens used were ovalbumin (2X crystallized, obtained from Worthington Biochemical Corp.), bovine gamma globulin and bovine serum albumin (obtained from Armour Laboratories), native dextran (kindly supplied by Dr. E. Hehre), and preparations of the somatic antigen ('qlpopolysaccharide endotoxin") of Escheri~hia eoli strain 0111 :B4 (obtained from Difco Laboratories, Inc., Detroit). Rabbits were immunized against nvalbumin or the bovine proteins by repeated intradermal and/or intravenous injections of antigen in doses which ranged from i to 50 rag. Serum obtained from immunized and normal animals was stored in the frozen state and allquots were inactivated by heating at 56°C. for a halfhour before use.Coagulation Time Measurement.--Whole blood was obtained from normal and immunized rabbits by cardiac puncture and immediately distributed in 2.0 or 4.0 ml. amounts in 15 x 75 mm. siliconized glass test tubes. The tubes were kept at room temperature, and the endpoint of coagulation was taken as that time at which no flow occurred upon inversion of the tube. The effect of various test materials on the coagulation time was determined by adding these materials in a total volume of 0.1 or 0.2 ml. to the tubes before addition of the blood.Under these conditions the coagulation time of rabbit blood was usually prolonged to from 20 to 40 minutes as compared to 8 to 12 minutes in clean but non-siliconized glassware. This prolongation of clotting time permitted an evaluation of the coagulation-accelerating effect