Experimental Murine Amyloid. 111. Amyloidosis Induced with Endotoxins By WERNER F. Bmm, M.D., JAMXS T. WILLERSON, M.D., RICHARD ASOFSKY, M.D., JOHN N. SHEAGREN, M.D., AND A new model for the study of experimental amyloid is described using parenteral administration of endotoxin. The differential responses in the various strains of mice examined suggested that genetic factors may be involved in murine A SHELD~N M. WOLFF,M.D. amyloid formation. Many similarities were noted between casein- and endotoxin-induced amyloidosis. It is suggested that endotoxin may be involved in the pathogenesis of several forms of experimental amyloidosis. is a fibrillar glycoprotein found are reported. The immunoglobulin rein the tissues of man and animals sponses and reticuloendothelial activity under a variety of clinical and laboratory after repeated administration of endotoxin conditions.l Numerous procedures have were also examined. Commercially obbeen used for the induction of experimental tained and a highly purified preparation of amyloid. These include the administration Salmonella endotoxin were studied for their of live or killed bacteriaz4 and the repeated ability to induce amyloid. The possible injection of various macromolecular sub- role of endotoxin in some of the models of stances such as serum proteins,6s6 ribo- experimental amyloid induction is disnucleic acid,7Vs adjuvant?JO and ~ a s e i n . ~ J l cussed. Recently, we reported that injection of E. coli endotoxin induced amyloid in MATERIALSAND METHODS mice.12 The histologic findings and strain Animals differences noted simulated the amyloid Mice were obtained from the Rodent and Rabbit produced in mice by repeated injections Production Section, Animal Production Unit of casein.13 In this study, evidence is presented which (NIH). The strains included CSH/Hen, AKR, AL/N, C57BL/6JN, Balb/c AnN, and White suggests that endotoxins may be responsi- Swiss (G.P.). Male mice, 6 to 7 weeks old, were ble for the amyloidosis induced by repeated used in each of the experiments. New Zealand parenteral administration of casein. Our albino rabbits of either sex, weighing approxifindings in six different strains of mice using mately 2 Kg., were used for bioassay of endotoxin E. coli endotoxin for induction of amyloid preparations. MYLOID WERNERF. BARTH,M.D.: Senior Investigator, Senior Investigator, Laboratory of Clinical InvextiArthritis and Rheumatism Branch, National In- gation, National Institute of Allergy and Infecstitute of Arthritis and Metabolic Diseams, tious Diseases, Bethesda, Maryland. SHELDON M. M.D.: WOLFF,M.D. : Clinical Director, National InstiBethesda, Maryland. JAMES T. WILLERSON, Clinical Associatq Arthritis and Rheumatism tute of Allergy and Infectiaus Diseases, Bethesda, Branch, Natiorurl I&itute of Arthritis and Maryland. Metabolic Diseases, Bethesda, Maryland. RICHARD Reprint requests should be addressed to DT. ASOFSKY,M.D.: Head, Experimental Pathology Werner F. Barth, Head, Arthritis Division, DeSection, Laboratory .f Microbial Immunity, NG partment of Medicine, University of Maryland tional Institute of Allergy and Infeetiow Diseases, School of Medicine, Baltimore, Maryland 21201. Bethesda, Maryland. JOHN N. SHEAGREN, M.D.: ARTHRITIS AND RHEUMATISM, VOL. 12, No. 6 (DECEMBER, 1969) 615 616 BARTH ET AL. Materials Escherichiu coli 0127:B8 and Salmonella typhosu 0901 endotoxins (Difco Laboratories, Detroit, Mich. ) were dissolved in phosphate-buffered saline (pH 7.4) at a concentration of 2.0 mg./ml. and frozen at -2OOC until used. A highly purified endotoxin derived from Salmonella enteritidis' was also employed.14 Hammersten Casein ( Nutritional Biochemical Company, Cleveland, Ohio) was slowly dissolved in 0.05 M NaHC03 at a concentration of 8.5 Gm. per cent and then allowed to pervaporate in dialysis tubing for 48 hours in the cold to a final concentration of approximately 12 Gm. per cent. The casein preparations were stored at 4OoC and assayed within 1 week for pyrogenicity. The casein was prepared in sterile glassware and pyrogen-free reagents to a concentration of 8.5 Gm. per cent both with and without added Merthiolate ( 1:10,000). Induction of Amyloid Mice were injected subcutaneously with endotoxin in alternate sites on the back, five times per week. The administration of 500 pg. daily caused a significant number of deaths in White Swiss mice during the first 3 days of this schedule. When the initial dose was decreased to 400 fig. for the first 3 days, followed by 500 pg. daily thereafter, mortality was negligible. This dose schedule was used in all experiments unless otherwise indicated. Animals injected with equal volumes of phosphatebuffered saline (0.25 ml.) served as controls. Groups of animals were sacrificed at intervals of 1-2 weeks. Spleen tissue was obtained for histologic examination and serum for determination of immunoglobulin levels. Frequently liver and kidney tissues were also examined. All slides were examined by hematoxylin and eosin stain. Virtually all slides were also examined by the Congo red stain. The presence of green birefringence using polarized light on Congo red-stained sections was considered specific for amyloid.15 Grading of Amyloid ' A new system was devised to grade the amount of amyloid in the spleens of mice: Grade 1 . A definite perifollicular rim of amyloid in one or more of the splenic follicles. Grade 2 . More than half of all follicles in the random section of the spleen examined show a rim of amyloid. This grade also included those sections *Kindly supplied by Dr. J. A. Rudbach, Rocky Mountain Laboratory NIAID, Hamilton, Mont. in which a small area of bridging between closely adjacent follicles was present. Grade 3. All splenic follicles show a rim of amyloid with significant bridging of amyloid between neighboring follicles. Grade 4. Extensive splenic amyloid with bridging between almost all follicles, some distortion of the follicular architecture, and involvement of the e n tral areas of follicles. Grade 5. Fifty per cent or more of the spleen replaced by amyloid; splenic architecture completely distorted. All slides were examined by two different observers who had no knowledge of the tissue source. The grading system used proved rapid and reproducible. Interobserver differences were never more than a grade and occurred in less than 5 per cent of the slides examined. Mouse Immunoglobulins Serum immunoglobulin levels were measured by a modification16 of the method of radial immunodiffusion17 using antisera specific for each of the five classes of mouse immunoglobulins.l~J9A mixture of mouse myeloma sera, previously calibrated with purified myeloma proteins, was employed as a standard. Mouse blood was obtained by retroorbital puncture at the times indicated and sera stored at -2OOC until used. Reticuloendothelial (RES) Phagocytic Function RES phagocytic activity was evaluated using a carbon suspension (Pelikan Ink Cll-l431a, Guenther-Wagner, New York) as the colloidal test substance using a standard method.20 The carbon suspension was diluted 2 parts carbon to 1 part saline, and an amount determined which gave a half-time of clearance ( T - %) of approximately 10 min. in control animals.21 This amount was found to be 0.01 ml. of the suspension per gram of mouse. Clearances were not determined serially in the same mice, but groups of 6 to 16 individually numbered mice were studied at each point in time. Immediately after completion of the clearance studies on any given group, the animals were sacrificed, tissues examined histologically, and the degree of splenic amyloid determined. Bioassay of Casein for Endotoxic Activity The casein solutions were injected in rabbits for appraisal of pyrogenicity and effect on circulating white blood cells. Details of the housing, recording of temaerature., m - lotting - of fever curves, induction of pyrogenic tolerance, and injections are recorded 617 EXPERIMENTAL MURINE AMYLOID. 111 Amyloid in Table 1.-Splenic G.P. Mice Following Administration of E. coli Endotoxin Grade of Amyloid a No. with Amyloidosis 5 10 15 20 injections injections injections injections ( 1 week) (2 weeks) (3 weeks) ( 4 weeks ) Saline E . coli Endotoxin 0/6 1/6 8/10 17/18 20/20 0/10 0/8 0/24 0.16 1.7 2.5 3.4 * Derived by adding the numerical grade assigned to the amount of amyloid in each animal and dividing by the total number of mice in the group. of Autoclaving Endotoxin on Amyloid Induction in G.P. Mice Table 2.-Effect Group A B C Endotoxin Given Daily, pg. No. of Injections 500 500 14 14 14 500 Autoclaved,* hr. - No. Animals with Amybid/ Total Receiving Endotoxin 15 43 '250°F. f The single animal with splenic amyloid had a small amount with a few perifollicular foci (grade 1). elsewhere.22 Antibody titers were measured in serial twofold dilutions of serum using the bentonite flocculation technic.23 The number of white blood cells in blood obtained by ear vein puncture were counted in duplicate in an electronic counter (Model B, Coulter Electronics, Hialeah, Fla.) after lysis of red blood cells with saponin. Differential counts of 100 white blood cells were performed on Wright-stained blood smears. RESULTS Amyloidosis in White Swiss (G.P.) Mice With repeated administration of E. coli endotoxin, amyloid was detected as early as 1 week after start of the injections (Table 1). Subsequently, in this mouse strain, amyloid appeared in the spleens of almost all animals after 10 or more injections. The earliest deposition was perifollicular; later the red pulp and germinal follicle were progressively involved. The amount deposited in the spleen increased with the longer injection schedule, and amyloidosis was then also detected in the liver and kidneys. None of the 48 animals injected with buffered saline alone developed amyloidosis. When the endotoxin was autoclaved for prolonged periods, such that pyrogenicity in rabbits was almost completely abolished, the ability to induce amyloid was also lost (Table 2).' Amyloidosis in Different Strains of Mice Using the same injection schedule, amyloid formation in different strains of mice was evaluated. In some strains (AKR and G.P. ) , amyloidosis developed early and in virtually all animals (Table 3). I n other strains, such as the Balb/c, C3H/Hen, and C57BL, amyloid was of a lesser degree, appeared later, and in fewer animals. Consistency with respect to the degree and rapidity of development of amyloid was noted within each of the three strains in which several experiments were performed. In three separate experiments, the C3H/Hen strain developed less amyloid, in fewer animals, and after delay when compared with the G.P. mice (Table 2 and reference 12). After 10 injections of endotoxin in G.P. mice, 8 out of 10 animals showed amyloid in two different experi'A few rabbits became febrile when 4 mg. of endotoxin was administered intravenously, a dose uniformly lethal prior to autoclaving. No rabbits developed fever following an intravenous injection of 400 pg. of endotoxin subjected to prolonged autoclaving. 618 BARTH ET AL. Table 3.-Splenic Amyloid in Different Strains of Mice Following Administration of E. coli Endotoxin Saline, 30 Injections E . coli Endotoxin Strain 10 Injections (grade) AL/N C57BL 4/21 Balb/c C3H/Hen 0/10 1/11 AKR G.P. * 5 of 0/8 10/10 8/10 (2.1) (1.9) 20 Injections (grade) 0/3 l0/15 9/9 2/11 (1.5) (2.1) 9/9 10/10 (4.2) (4.3) 30 Injections (grade) - 0/10 9/9 9/11 7/10 9/9 11/11 * 0/8 (3.2) (1.8) (1.6) (4.2) (4.5) 0/10 0/8 0/10 0/10 these mice were sacrificed after 10 injections of saline. ments; the same grade of amyloid was found on both occasions (Tables 1 and 2). The AL/N strain was examined twice because of the high mortality noted in the first experiment. Twenty-five out of 35 mice died in the first 2 weeks of administration of E. coli endotoxin. Most deaths occurred during the first 3 days. There was equally high mortality in a second experiment using 35 more mice; hence there are no data available after 2 weeks' treatment. Virtually none of the C3H/Hen mice and very few of the G.P., Balb/c, or AKR mice died on this injection schedule. Higher mortality than was observed in these 4 strains was seen in the C57BL strain. Consequently, in this strain, on this schedule, the later time points could not be studied. None of the saline-treated animals, in any strain, developed amyloidosis. purified S. enteritidb endotoxin was studied. Three of 4 mice that received 15 injections of 0.5 mg./day of this preparation for 3 weeks developed splenic amyloid. Immunoglobulin Responses Following Administration of Endotoxin The immunoglobulin responses in G.P. and C3H/Hen mice following repeated administration of endotoxin are shown in Figs. 1-4. In the G.P. mice, 7s-yl levels were only slightly increased with high doses of endotoxin but not increased with the lower doses (Fig. 1 ) . In this strain, 7S-y2,, levels were decreased but other immunoglobulins unchanged after 5 weeks' treatment with 0.005 mg. of endotoxin (Fig. 2 ) . At the higher doses, there was an early and sustained increase in levels of 7 S - y ~and ~ yM and a decrease in levels of yA (Figs. 1 and 2). The immunoglobulin responses in this strain contrasted with the C3H mice which showed a marked increase in 7Syl levels after high doses of endotoxin for 6 weeks but little increase with the smaller doses (Fig. 3). The changes in 7S-yzB, yM, and y,, noted in the G.P. mice were not found in the C3H mouse (Fig. 4 ) . Production of Amyloid by S . typhosa and Highly Purified S . enteritidis Endotoxins Twenty-one of 25 G.P. mice examined after 15 injections (3 weeks) of S . typhosa endotoxin showed splenic amyloidosis. The limited study with this endotoxin suggested that it was qualitatively similar to E . coli RES Phagocytic Function During Adminendotoxin in the type of histological lesions istration of Endotoxin induced. In order to determine whether Table 4 shows serial changes in the rate amyloidosis induced by the commercial of clearance of colloidal carbon in G.P. and products was not secondary to some other contaminating nonendotoxin material," the *Aerobic and anaerobic cultures of the comeffect of repeated injections of a highly mercial Difco endotoxin were negative. 6 19 EXPERIMENTAL MURINE AMYLOID. 111 G.l? MICE WITH AMYLOIDOSIS 16.0 - 12.0 - 0 1 CONTROL ENDOTOXIN ENDOTOXIN ENDOTOXIN ENDOTOXIN 0.005mg 0.5mg 0.5mg 0.5mg ( 2 wks.) (4 wks.) (5 wks.) (5 wks.) Fig. 1.-The 7S-yl immunoglobulin response following administration of E. coli endotoxin to White Swiss (G.P.) mice. Each point represents the level in an individual mouse. Mean levels are indicated by the horizontal bars. - G.P. < 6.0- UJ 5.0 - -P 5 -J 0 0 0 2 4.0. A 0 0 COO0 ooo 0" 0 0 v 0 0 00 00. -co- 080 +O CONTROL ti ENDOTOXIN 0.005mg ( 5wks.) naoa 0 e. M 0 .+ * so A ?? AA W 0 " : &?A !: - i a A a4 4 - 5 ENDOTOXIN 0.5 mg (2wks ) ENDOTOXIN 0.5mg (4w k s . ) ENDOTOX IN 0.5mg (5wks.) Fig. 2.-Serum levels of 7 s - y ~ S~ -~~ ZIgA, B , and IgM in G.P. mice after repeated injections of E. coli endotoxin. Each point represents the level in an individual mouse. Mean levels are indicated by the horizontal bars. CSH/Hen mice. Acceleration of the clearance rate was seen at the end of 1 week in the G.P. mice. By the end of the second week, the clearance rate had returned to control levels, where it remained despite progressively increasing amounts of amy- loid in the spleen and liver. Similar alterations in the rate of carbon clearance were demonstrated in the C3H/Hen mice, but over a more extended period of time. Acceleration was seen at 3 weeks in this strain with return to control levels at 6 weeks. 620 BARTH ET AL, C3" MICE WITH AMYLOIDOSIS 0 0 '6 0 '/10 '10 18 00 28.0 0 00 24.0 0 T 20.0 'SY, (mg/ml) 16.0 0 0 12.0 0 0 8.0 4.c 4 0 - 0 0 0 0 0 0 0 0 1 ~ :NDOTOXIN ENDOTOXll :NDOTOX II :NDOTOXIN 0.005mg 0.00Smg 0.5 mg 0.5mg (3wkr.) (6 wkr.) (6 wkr.1 (3 wkr.) CONTROL Fig. 3.-The 7S-yl immunoglobulin response to various doses of E. coli endotoxin in C3H mice. Each point represents the level in an individual mouse. Mean levels are indicated by the horizontal bars. .*o 0 O h == A*& A A 0 ENDOTOX IN 0.005mg (3wks. 1 ** . - 0= * = *-. 8 + 0 ENDOTOX I N 0.005mg (6 wks.) I 1 ENDOTOX IN 0.5 mg (3w k s . ) ENDOTOXIN 0.5 mg (6wks.) Fig. 4.-Responses of 7S-y2*, ~ S - YIgA, ~ ~ and , IgM to repeated administration of E. coli endotoxin in the C3H mouse. Each point represents the level in an individual mouse. Mean levels are indicated by the horizontal bars. 621 EXPERIMENTAL MURINE AMYLOID. 111 of Clearance of Colloidal Carbon at Various Times During the Development of Endotoxin-Induced Amyloidosis in Two Strains of Mice Table 4.-Rates Half-Time of Clearance (TM) Performed at: Mouse Group 0 Baseline Control G.P. Endotoxin $ treated G.P. Control C3H/Hen Endotoxin $ treated C3H/Hen 10.5+0.9t 13.1+1.6 10.5 f 0.9 2 Weeks 3 Weeks 4 Weeks 12.1+0.9 10.4f0.8 12.1f0.6 1 Week 5.2 + 0.4 10.9 + 1.3 + 0.6 8.8 + 0.6 8.8 + 1.0 13.2 -I- 1.0 10.4 + 1.7 4.2 + 0.3 Q 6 Weeks 11.3 + 1.5 8.2 + 1.0 9.3 Each group consisted of a minimum of 6 animals. (min.) f S. E. $ E. coli endotoxin in dose of 0.5 mg. subcutaneously five times per week. $ Significantly different from control: p 0.01 ( t test). O t Mean T 95 < Table 5 . F e b r i l e Responses in Rabbits Following Intravenous Administration of Casein and E. coli Endotoxin Agent Administered 0 NaHCO, 8.5 Gm. % casein 0.85 Gm. % casein 8.5 Gm. % casein E. coli endotoxin E. coli endotoxins 8.5 Gm. % casein 11 Amount, ml. 1.0 1.0 1.o 1.0 ml. daily for 6 days 0.5 pg. 0.5 p g . 1.0 ml. Number of Rabbits Mean F.I.,f cm.2 3 3 3 4 4 4 3 3.7 36.3 16.6 17.0 43.5 28.3 21.0 Mean 3 Hr. Temp.,* C 0.13 1.63 0.40 0.18 1.95 1.28 0.33 * All agents given I.V. t F.I. = 5 hr. fever index, cm.2 Fever index is area under a fever curve in which 1 degree and 1hour equals 1 inch. $ 3 hr. temp. degrees C above baseline 3 hours after agent injected. $ Rabbits had received 1 ml., 8.5 Gm. %, casein daily (I.V.) for 7 days prior to challenge with endotoxin. I I Rabbits had received 7 daily I.V. injections of 3 pg. E. coli endotoxin. Amyloid did not develop in these C3H mice but was present in virtually all of the G.P. mice after 10 injections (Table 1). “‘Endotoxic” Properties of Casein The febrile responses of rabbits to the intravenous injection of casein are presented in Table 5. Every sample of casein tested was pyrogenic. The febrile responses to 8.5 Gm. per cent of casein were higher than to 0.85 Gm. per cent of casein. All rabbits which received a single injection of the higher dose of casein had biphasic fever curves typical of those seen after administration of endotoxin. The daily administration of casein resulted in pyrogenic refractoriness ( “tolerance”) to subsequent challenge with casein characterized by a lower febrile response and a monophasic fever curve. Cross tolerance developed between E. coli endotoxin and casein; E. coli-tolerant rabbits showed diminished responses to casein, and casein-tolerant rabbits had similarly decreased febrile responses to E. coli endotoxin. Results of studies shown in Table 5 were derived with casein solutions prepared within 1-2 days of testing. Similar experiments were 622 BAR= Table 6.-Effect Time, hr. 0 0.5 2 4 6 24 ET AL. of Casein on Blood Leukocyte Count in Rabbits' WBC Granulocytes t 9425 6200 4600 6375 8500 8325 3420 2873 1628 5252 6130 4391 Mononuclears t 6005 3327 2172 1123 2370 3934 'Rabbits received 1 ml. of 8.5 Cm. % casein I.V. at 0 time. The results are the mean values of 4 rabbits. t Granulocytes =polymorphonuclear leukocytes and basophiles. Mononuclears = lymphocytes and monocytes. performed with stock casein solutions, and the results obtained were essentially identical to those in Table 5. An aliquot of a casein solution was autoclaved at 250°F for 2 hours and tested €or pyrogenicity. The material was pyrogenic prior to autoclaving but not afterward. Such a change is characteristic of endotoxins .24 Further evidence that the casein was contaminated with endotoxin is presented in Table 6. Casein induced a prompt leukopenia, followed by a granulocytosis. Such results are similar to those reported following the administration of bacterial endotoxins to rabbits.25 One week following the intravenous administration of 1 ml. of casein, all 4 rabbits tested showed low titers of circulating antibody to either the somatic antigens of E. coli or S. typhosa. Antibodies detectable by the bentonite flocculation test are rarely present prior to the administration of endotoxic polysaccharides.26Such cross-reacting antibodies as appear after administration of endotoxin are regarded as nonspecific and do not serve as an indicator of the type of endotoxin present in casein. different microorganisms as well as a highly purified endotoxin to induce amyloid suggests that amyloid induction may be a general property of endotoxins and not due to some unidentified contaminants in the commercial preparations. Amyloid was initially deposited in the spleen and only later in the liver and kidney. There were progressively increasing amounts of amyloid in these organs after longer periods of endotoxin administration. Differences were detected in the ease of induction and amount of amyloid deposited in the spleen among the six strains of mice examined. For example, AKR mice had more splenic amyloid after 10 injections than did C3H mice after 30 injections. This strain difference resembled that seen in a previous study of spontaneous amyloid in aged mice in which amyloid was more extensive and frequent in A strain mice than in C3H mice.27 Strain differences were also observed in the immunoglobulin responses to repeated administration of endotoxin. Whereas C3H mice had a marked increase in 7S-yl levels with little increase in the other immunoglobulins, G.P. mice had increases in 7S-y1, 7S-yZB, and yy levels. Both the serum DISCUSSION 7S-yl levels and the total immunoglobulin This study demonstrates that the re- levels were lower in the G.P. than in the peated administration of commercially C3H mice, however. It was noted that available E. coli endotoxin rapidly and con- serum yA levels fell during treatment in sistently induces amyloidosis in mice. The the G.P. mice. Possible mechanisms for a ability of endotoxins derived from several fall in levels of some of the immunoglob- EXPERIMENTAL MURINE AMYLOID. III ulins with administration of endotoxin are increased immunoglobulin catabolism or some reciprocal regulation of the levels among the various immunoglobulin classes. It is apparent that amyloidosis was induced in some mice with little or no increase in serum immunoglobulin levels, and that two strains of mice manifested different immunoglobulin responses following the same antigenic stimulus. Many of the features of amyloidosis induced by endotoxin simulated those noted following administration of casein. Sites of deposition and order of appearance in the various organs in the two models were similar. As with amyloid induced by casein,13strain differences were present and there was no direct relationship between amyloidosis and an increase in serum immunoglobulin levels. Accelerated RES phagocytosis of carbon followed by a return to normal has been described with amyloidosis induced by c a ~ e i n ~and ~-* was ~ duplicated in the present studies by administration of endotoxin. The impairment in ribosome function during the early phase of amyloid induction using casein30 has also been produced by repeated injections of endotoxin.3l A significant difference in the two methods of amyloid induction is the relative dose of inducing agent necessary. Forty to fifty milligrams of casein administered daily for several weeks are often necessary to produce amy10id.l~In contrast, a total dose of 0.125 mg. of E. coli endotoxin given over a 5 week period12 or 2.3 mg. given in 1 week as in this study have induced amyloid. In view of the many similarities between these two models of induction of amyloid, it was postulated that endotoxin contamination might account for casein-induced amyloidosis. The presence of endotoxin in casein was shown by the following: (1)A typical biphasic febrile response in rabbits was observed following administration of casein; ( 2 ) a dose response relationship be- 623 tween amount of casein administered and febrile reactivity was found; ( 3 ) febrile tolerance with casein and cross tolerance between both casein and endotoxin was produced. Other evidence of the presence of endotoxin in casein was: (4)heating in the autoclave destroyed the pyrogenicity of casein; (5) the hematological responses in rabbits after casein administration were the same as noted after endotoxin; and ( 6 ) antibodies to the somatic polysaccharides of two different Enterobacteraceae were detected after administration of casein. Febrile responses in mice following repeated administration of casein were studied over 40 years ago, and the results were consistent with contamination by endotoxin.6 Every lot of casein we examined showed endotoxin activity in the rabbit, but quantitation has not been possible. The type of endotoxin( s ) present in casein is unknown. It is probably not E. coli endotoxin since the strain responses to E. coli endotoxin and casein were not the same in the different strains of mice. The C57BL strain developed little amyloid after E. coli endotoxin, but had the highest incidence in response to casein.13 Conversely, AKR mice developed amyloidosis as readily as did G.P. mice following administration of endotoxin but not after administration of casein.13 Recently spores of Bacillus cereus have been found as contaminants in the type of casein we used.32 Possibly endotoxin from these organisms is responsible for the amyloid reaction induced by casein. The mechanism of amyloid formation following administration of endotoxin is unknown, but our results suggest that immunoglobulins are not directly involved in its pathogenesis. That genetic factors may play a role is suggested by the different responses in the strains examined. Variations in leukocyte responses, resistance, and lethality following the administration of endotoxin have previously been observed in 624 BARTH ET AL. diflerent strains of It has been Endotoxin may be implicated in other suggested that chronic stimulation of cells forms of experimental amyloidosis. It is in the reticuloendothelial system is involved ubiquitous and difficult to inactivate within the pathogenesis of amy1oidosis.l The out such vigorous procedures as prolonged ability of endotoxin to induce amyloidosis autoclaving at high temperature^.^^ The supports this hypothesis, since endotoxin literature contains many instances in which rapidly localizes to these cells following contamination with endotoxin was later parenteral a d m i n i s t r a t i ~ n ~and ~ - ~stimu~ shown to be responsible for a variety of lates RES phagocytic function as ~ e l l . ~ biological ~ 9 ~ ~ effects.25 Materials other than Chronic stimulation of lymphocytes is ap- casein that have been used to induce parently inadequate, since prolonged ad- amyloid in animals include such diverse ministration of phytohemagglutinin did not substances as live or killed bacterial induce amyloid.3l organisms, serum proteins, adjuvant, and Studies thus far do not permit distinction ribonucleic acid. It is possible that endobetween endotoxin, a fraction of endotoxin, toxin could be involved in these models. or an endogenous product released followEndotoxin may be of pathogenetic iming endotoxin administration as a cause of portance in some forms of human amyloid, amyloid formation. It is of interest that particularly those associated with chronic loss of pyrogenicity by prolonged autoinfections. The persistent presence of endoclaving abolished the amyloidogenic properties of endotoxin. Efforts are underway toxin in chronically infected tissues in man in our laboratory to identify the amyloido- may be analogous to the murine model degenic principal in endotoxin and to study scribed here. its subcellular localization. ACKNOWLEDGMENTS The authors thank Mrs. Kay Gordon, Mr. S. B. Ward, and Mr. J. B. Edelin for their expert tech- nical assistance. We would also like to thank Dr. John L. Decker and Dr. Leon Sokoloff (NIH) for their advice and encouragement in the course of these studies. SUMMARYAND CONCLUSIONS The repeated administration of commercially available E. co2i endotoxin to mice rapidly and consistently induced amyloidosis. S. typhosa endotoxin and a highly purified S. mtrn’tidhs endotoxin were also able to induce formation of amyloid. The development of amyloidosis was related to the mouse strain used and in some mice was not accompanied by an increase in serum immunoglobulin levels. Injections of endotoxin initially produced an acceleration in reticuloendothelial clearance of carbon, whereas prolonged administration was associated with a return to control levels. Large amounts of endotoxin were found to be present in casein. There were many similarities in the amyloidosis induced by administration of endotoxin and casein. It is suggested that endotoxin may be responsible for the amyloidosis induced by repeated parenteral administration of casein, and that endotoxin may be implicated in other forms of experimental amyloidosis. SUMMARIOIN INTERLINGUA Le administration repetite de commercialmente disponibile endotoxina de Escherichia coli a muses resultava in le rapide e consistente induction d e amyloidosis. Endotoxina de Salmonella typhosa e u n altemente purificate endotoxina de Salmonella enteritidis etiam se monstrava capace de inducer le formation de amyloide. Le disveloppamento de amyloidosis esseva relationate con le racia murin usate, e in certe muses ill0 non esseva accompaniate de un augment0 in l e nivellos sera1 de irnmunoglobulina. Injectiones de endotoxina produceva initialmente u n acceleration in le clearance reticulo- 625 EXPERIMENTAL MURINE AMYLOID. 111 endothelial de carbon, durante que le administration prolongate esseva associate con un retorno a1 nivellos de controlo. Grande quantitates de endotoxina esseva retrovate in caseina. Multe similitudes esseva notate inter le amyloidosis inducite per le administration de endotoxina e ill0 inducite per caseina. Es suggestionate q u e endotoxina es possibilemente responsabile pro le amyloidosis inducite per repetite administrationes parenteral de caseina e q u e endotoxina ha possibilemente un rolo in le genese de altere formas de amyloidosis experimental. REFERENCES 1. Cohen, A. S.: The constitution and genesis of amyloid. In: Richter, G. W., and Epstein, M. A. (Eds.) : International Review of Experimental Pathology, Vol. IV. New York Press, 1965. 2. Bailey, C. H.: The production of amyloid disease and chronic nephritis in rabbits by repeated intravenous injections of living colon bacilli. J. Exp. Med. 23:773, 1916. 3. Hass, G. M., Huntington, R., and Krumdieck, N.: Amyloid. 111. The properties of amyloid deposits occurring in several species under diverse conditions. A.M.A. Arch. Path. 35:226, 1943. 4. Howes, E. L., Jr., Pincus, T., McKay, D. G., and Christian, C. L.: A model of amyloidosis. Arthritis Rheum. 6:278, 1963. 5. Jaffee, R. H.: Amyloidosis produced by injections of proteins. Arch. Path. Lab. Med. 1:25, 1926. 6. Dick, G. F., and Leiter, L.: Some factors in the development, localization, and reabsorbtion of experimental amyloidosis in the rabbit. Amer. J. Path. 17:741, 1941. 7. Bohle, A., Hartmann, F., and Pola, W.: Elektrophoretische Serumeiweissuntersuchungen bei experimentellem mauseamyloid (Serum electrophoretic examination in experimental murine amyloid). Virchow Arch. [Path. Anat.] 319:231, 1950. 8. Richter, G. W.: Alterations in serum globulins during the formation and resorbtion of amyloid in rabbits. J. Exp. Med. 104:847, 1956. 9. Rothbard, S., and Watson, R. F.: Amyloidosis and renal lesions induced in mice by injection with Freund-type of adjuvant. Proc. Soc. Exp. Biol. Med. 85:133, 1954. 10. Ram, J. S., DeLellis, R. A., and Glenner, G. G.: Amyloid. 111. A method for rapid induction of amyloid in mice. Int. Arch. Allerg. 34:201, 1968. 11. Smetana, H.: Experimental study of amyloid formation. Bull. Hopkins Hosp. 37:383, 1925. 12. Barth, W. F., Gordon, J. K., and Willerson, J. T.: Amyloidosis induced in mice by Escherichia coli endotoxin. Science 162:694, 1968. 13. Wdlerson, J. T., Asofsky, R., and Barth, W. F.: Experimental murine amyloid. IV. Amy- loidosis and immunoglobulins. J. Immun. In press. 14. Ribi, E., Haskins, W. T.,Landy, M., and Milner, K. C.: Preparation and host reactive properties of endotoxin with low content of nitrogen and lipid. J. Exp. Med. 114:647, 1961. 15. Missmahl, H. P.: Rectal biopsy in the diagnosis of amyloidosis. German Med. Monthly. 9: 101, 1964. 16. Mancini, G., Garbonara, A. O., and Hermans, J. F. : Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry. 2:235, 1965. 17. Fahey, J. L., and McKelvey, E. M.: Quantitative determination of serum immunoglobulins in antibody-agar plates. J. Immun. 94:84, 1965. 18. Fahey, J. L., Wunderlich, J., and Mishell, R.: The immunoglobulins of mice. I. Four major classes of immunoglobulins, 7Sy,, 7Sy,, yIA ( B2*) and 18S-y,, globulins. J. Exp. Med. 120:223, 1964. 19. Fahey, J. L., Wunderlich, J., and Mishell, R.: The immunoglobulins of mice. 11. Two subclasses of mouse 7Sy, globulins, y P A and yZB globulins. J. Exp. Med. 120:243, 1964. 20. Biozzi, G., Benacerraf, G., and Halpern, B. N.: Quantitative study of the granulopectic activity of the reticuloendothelial system. 11. A study of the kinetics of the granulopectic activity of the R.E.S. in relation to the dose of carbon injected. Relationship between the weight of the organs and their activity. Brit. J. Exp. Path. 34: 441, 1953. 21. MacGregor, R. R., Sheagren, J. N., and Wolff, S. M.: Endotoxin induced modification of Plasmodium berghd infection in mice. J. Immun. 102:131, 1969. 22. Wolff, S. M., Mulholland, J. H., and Ward, S. B.: Quantitative aspects of the pyrogenic response of rabbits to endotoxin. J. Lab. Clin. Med. 65:268, 1965. 23. WOW, S. M., Ward, S. B., and Landy, M.: Serologic properties of bentonite particles coated with microbial polysaccharide. Proc. SOC. Exp. Biol. Med. 114:530, 1963. 24. Berger, A., Elenbogen, G. D., and Ginger, L. G.: Pyrogens. Advances Chem. (Amer. Chem. 626 SOC.)16:168, 1956. 25. Bennett, I. L., Jr., and Cluff, L. W.: Bacterial pyrogens. Pharm. Rev. 9:427, 1957. 26. Wolff, S. M., and Ward, S. B.: Personal observations. 27. Dunn, T. B.: Relationship of amyloid infiltration and renal disease in mice. J. Nat. Cancer Inst. 5:17, 1944. 28. Shearing, S. P., Comerford, F. R., and Cohen, A. S.: Effect of an amyloid inducing regimen on phagocytosis of carbon particles. Proc. SOC.Exp. Biol. Med. 119:673, 1965. 29. Ranlov, P.: Phagocytosis in experimental mouse amyloidosis. Acta Path. Microbiol. Scand. 68:19, 1966. 30. Talal, N., and Barth, W. F.: Decreased ribosome function during the induction of experimental amyloidosis. Clin. Res. 16:325, 1968. 31. Talal, N., Daley, G. G., and Barth, W. F.: Altered ribosome function during the induction BARTH ET AL. of experimental amyloid. Submitted for publication. 32. Stora, C., Bariety, J., and Bariety, M.: Role du “Bacillus Cereus” dans l’induction de l’amylose experimentale provoquhe par les injections de caseinate d e soude et d’azo-caseine (Role of Bacillus cereus in the induction of experimental murine amyloid produced by the injections of casein and azocasein). Path. Biol. 16:649, 1968. 33. Sultzer, B. M.: Genetic control of leucocyte responses to endotoxin. Nature 219:1253, 1968. 34. Hill, A. B., Hatswell, J. M., and Topley, W. W. C.: The inheritance of resistance demonstrated by the development of a strain resistant to experimental inoculation with a bacterial endotoxin. J. Hyg. 40:538, 1940. 35. Bohme, D. H.: The reticuloendothelial system and nonspecific resistance. Ann. N. Y. Acad. Sci. 88:172, 1960.