Initial Endotoxin Shock Expression of Anaphylactic on ...dm5migu4zj3pb.cloudfront.net/manuscripts/104000/... · histamine assays on six dogs to determine the effects of anesthesia,

Journal of Clinical InvestigationVol. 43, No. 4, 1964

The Initial Stage of Canine Endotoxin Shock as an Expressionof Anaphylactic Shock: Studies on ComplementTiters and Plasma Histamine Concentrations *

WESLEYW. SPINK, RICHARD B. DAVIS,t RUTH POTTER, ANDSANDRACHARTRAND

(From the Department of Medicine, University of Minnesota Hospitals and Medical School,Minneapolis, Minn.)

Within 30 to 60 seconds after endotoxin is in-jected into adult mongrel dogs there is a decline insystemic blood pressure, a rise in portal veinpressure, a reduction in the venous return of bloodto the heart, and a decrease in renal blood flow(1, 2). Endotoxin shock is comparable inmany ways to acute anaphylactic shock in thedog (3). Several observations have suggestedthat an immune mechanism causes the initial vas-

cular changes. It was demonstrated that theaction of endotoxin on blood vessels was medi-ated through a heat labile factor in plasma or

serum (4, 5). Gilbert and Braude (6) in studieson Escherichia coli endotoxin shock in rabbits re-

ported that doses of endotoxin greater than theLD50 caused a decline in the titer of complementand a decrease in the serum concentration of E.coli antibody. Spink and Potter (7) also ob-served a prompt decrease in plasma complementvalues in canine endotoxin shock.

The immediate effect of endotoxin on smallervessels suggested that a vasoactive substance or

substances were also implicated. If an immunemechanism were involved, histamine could be one

of the substances that was liberated by an antigen-antibody mechanism, and studies do point to hista-mine as a factor causing the altered vascular ac-

tivity (3, 8-10).This report concerns the study of complement

titers, plasma histamine concentrations, and bloodpressure changes in a series of dogs given a lethal

* Submitted for publication August 15, 1963; acceptedDecember 12, 1963.

Supported by U. S. Public Health Service grantAI 04415-02.

t Recipient of U. S. Public Health Service ResearchCareer Program Award 5-K3-HE-14919 from the Na-tional Heart Institute.

dose of endotoxin. In addition, since epsilon-aminocaproic acid and cortisol protect dogsagainst endotoxin, we were interested in ascer-taining the effect of these two agents upon com-plement and histamine values. The data supportthe concept that the initial hemodynamic phaseof canine endotoxin shock is related to an immunemechanism of the immediate anaphylactic type.

MethodsAnimals. A total of 85 adult mongrel dogs was used

in a series of 10 experiments.Endotoxin. The same lot of E. coli endotoxin was

used throughout and was prepared as described else-where (11). The LD100 of the endotoxin was first es-tablished in mice and then in dogs. The LD,., for thelatter species was 0.55 mg per kg.

Complement assay. Titers were expressed as 50%hemolytic units, employing the method outlined by Kabatand Mayer (12). Blood was collected before the in-jection of endotoxin and then 10 minutes, 1, 3, and 6hours later. After the blood had clotted at room tem-perature, serum was removed by centrifugation andstored at - 200 C overnight. Control studies showedthat this temporary storage did not result in significantdeterioration of complement.

Histamine assay. Serial plasma histamine concentra-tions were determined by a modification (13) of thefluorometric assay described by Shore, Burkhalter, andCohn (14). The reliability of this method was deter-mined by bioassay on duplicate samples through thecourtesy of Dr. Charles Code of the Mayo Foundationfor Medical Research. A good correlation of the twotechniques was found.

Epsilon-aminocaproic acid (EACA).1 One g containedin sterile vials was infused intravenously in 100 ml of5% dextrose and distilled water over a period of 15 min-utes before the administration of endotoxin. Comple-ment and histamine assays were carried out before andafter the injection of EACA, and after endotoxin hadbeen given.

1 Supplied by Lederle Laboratories, American Cyana-mid Co., Pearl River, N. Y.

696

ANAPHYLAXIS AND CANINE ENDOTOXINSHOCK

Cortisol.2 One hundred mg contained in sterile vialswas infused intravenously in 100 ml of 5%o dextrose andwater over a period of 15 minutes before the administra-tion of endotoxin. Complement and histamine assays

were likewise obtained after this infusion, as well as

after endotoxin.EACA plus cortisol. Quantities of these two agents

as given above were infused simultaneously into animals,and complement and histamine values were obtainedduring the same periods before and after the injection ofendotoxin.

Observations on shock. The animals, anesthetized withsodium pentobarbital, were studied in a manner previ-ously described (1). Continuous femoral arterial bloodpressure determinations and urinary output were recordedup until 6 hours after endotoxin, or until death.

Control animals. To assess the role of anesthesia, theloss of blood necessary for the tests, trauma due to fe-moral catheterization, and the effect of intravenous in-jections of physiologic solutions, the following studieswere carried out: 1) complement titers on three dogsanesthetized and given 100 ml of 5% dextrose and water(this was done to control the effects on complementfollowing the infusion of EACAand cortisol); 2) serialhistamine assays on six dogs to determine the effects ofanesthesia, the trauma of catheterization, and the injec-tion of 2 ml of saline, which was the amount of solu-tion required for the administration of endotoxin; 3)complement titers in six dogs after a lethal dose of thesnake venom, Crotalus terrificus,3 since the initial hemo-dynamic alterations in the dog could be produced by sub-stances other than-endotoxin, such as snake venom.

Results1. Complement titers and histamine concentra-

tions in dogs given a lethal dose of endotoxin.2 Supplied as Solu-Cortef (hydrocortisone sodium suc-

cinate) by Upjohn Co., Kalamazoo, Mich.3 Supplied by University of Sao Paulo, Sao Paulo,

Brazil.

(158)

z

W100w

E 800

(-)

o 60(I4

z 40

PRE- 10 2 3 4 5 61ENDO. Min.

-PPOST ENDOTOXIN

FIG. 1. DECREASEIN COMPLEMENTTITERS IN EACH OF

TEN DOGSGIVEN 0.55 MGPER KG ENDOTOXIN. There were

no survivors.

Ten dogs, weighing from 8.1 to 12.5 kg, were anes-

thetized and given 0.55 mg per kg of endotoxin.Survival times varied from 2- to 16 hours. Thedecline in complement is shown graphically inFigure 1. Preliminary studies on complementtiters at 1, 5, 10, and 30 minutes after a lethal doseof endotoxin had shown that the average maximaldecline occurred at 10 minutes. After this periodthere was a tendency for the complement titersto rise, especially in dogs that recovered. Al-though arbitrary, the time of apparent maximaldecline of complement was chosen for comparisonin most studies. The decline of complement inanimals 10 minutes after endotoxin (Table I)

TABLE I

Percentage of change in complement titer in dogs 10 minutesafter lethal dose of endotoxin (0.55 mg/kg)*

Pretreated SurvivorsControl, Pretreated Pretreated with given 2nd

5% Control, with with EACA dosedextrose endotoxin EACA cortisol and cortisol endotoxin

-34 -26 -17 -21 -50-18 -33 -13 -17 -26

0 -57 -12 -32 -25 -47-2 -42 - 6 -60 -23 -57

+21 -36 -16 -39-36 -26 -27 - 2 -73-33 -27 -40 -44 -21-51 -30 -27 - 8 -68-31 -40 -40 - 1 -74-34 -10 -21 -24 -65

Mean i SE +6 i 7 -37 + 3 -23 + 4 -29 + 5 -18 + 5 -52 + 6

* EACA= epsilon-aminocaproic acid.

697

W. W. SPINK, R. B. DAVIS, R. POTTER, AND S. CHARTRAND

t-. 00 00 0 en

00 -o o

0n0 C\ 00 0o 06 64 6 6

o en CN CN C10 o o) o o

06

I- t- 'I It C0) 00C 0 0 06 6 6 6 6 0

Uf) C\0 (DI C \0 0I " - -

en so m ~o 00cl m C> 0 c"I

eq cq I -4

0 04 en0 0 0 0 06 6 6 6 6

00

-4

a'05

- N f) 0C 00

6 6 6 6 6 0

00 0 oN en o

- - 00 \0o en00 o 0o en "

q en U) t-o

66 6 66 6

06

00

06

00 o 0C% c0 000)-4 1- 0 06 6 6 6 6

00 N- n

00 mt o- 4-4 -4 1-

r-U) Ul 00 Cioo N -6 C66 6

C) o o. o o-

00

0

C'

06

0) UO) _ _ t- o

6 o6 6 6 6"

en) 00

0a'

0 Itt

0"

6N

*; 6

Uf) a'- 0

-4 t) 00

- t- -aCi O 06 6 6

)o C0 I- If)

(f W

cd 0 0 04U) a)

Q

00 0 0* *

4) x.

was highly significant compared to saline con-trol animals (p < 0.01).

Plasma histamine concentrations are presentedin Table II. Many preliminary observations re-vealed that the maximal rise in plasma histamineoccurred between 30 and 60 seconds after theinjection of endotoxin. An explosive rise oc-curred in five dogs at the end of 60 seconds, re-vealing an average concentration 8 times the valueobtained before endotoxin was injected. Thiselevation was also confirmed in selected animalsby the bioassay method used by Dr. Code. Themean increase in plasma histamine was signifi-cantly greater than in saline-injected dogs 5 min-utes after endotoxin (p < 0.05), but was not sta-tistically significant after 30 and 60 seconds. Indogs that survived 4 hours, histamine values ap-proached those of the control pre-endotoxinperiod.

It has been repeatedly observed in the dog thatan abrupt drop in systemic pressure occurred 1 to5 minutes after endotoxin. Within 15 to 30 min-utes pressures usually approached the pre-endo-toxin base line. There was a good correlationbetween the abrupt rise in histamine and theinitial period of hypotension (Table II), althoughtwo animals failed to show an appreciable in-crease in histamine concentrations, and in threeinitial hypotension preceded a demonstrated in-crease in histamine concentrations.

2. Complement titers in dogs given an LD50,, injection of endotoxin. Six animals given an

o LD5o injection revealed a distinct difference inthe values of complement on the basis of survivalor death as seen in Figure 2. All animals had

m4 an initial drop in complement. In the three ani-r mals that survived the mean titer was 77%o of

1 the control, whereas in the three that expired the, mean titer was 23%o of control after 3 hours.Cd This difference is highly significant (p < 0.01).0

3. Complement titers and histamine concen-Cd trations in dogs pretreated -with EACA and then

given a lethal dose of endotoxin. In previouso. studies the majority of dogs pretreated withn EACAsurvived a lethal dose of endotoxin (15).

Ten dogs were given 1 g of EACAas an infusion'< in 100 ml of 5% dextrose and water. At the end* of this time 0.55 mg per kg of endotoxin was

injected. Seven of the 10 dogs survived.

698

U

~In

a4

:r

(n

LI)

VI)

a

avI r.Y

5̀_

cn

::6

:4(n

*

6-

0'a0

0

to-

*'a

4Iaa

a

*a

a

a

.0U

0

'a

.aa~w

lc-

IC 4)£Uto c-

0

E v8

-

o ¢A

-1

1. u

UW@

4)8.

W

0 =I..

.6i v

ANAPHYLAXIS AND CANINE ENDOTOXINSHOCK

Complement titers in this group of animalswere not appreciably altered by the initial infusionof 1 g of EACA in 100 ml of 5 % dextrose andwater, the mean decrease in complement being19%o in animals receiving dextrose and wateralone, and 25%o in those given EACAin dextroseand water. The mean complement titer that waspresent after the EACA infusion and before theinjection of endotoxin was not significantly dif-ferent from the valuies in those control animalsreceiving endotoxin alone. In the dogs pre-treated with EACA there was no difference inthe fall of complement after 10 minutes in thosethat survived and those that died.

Mean plasma histamine concentrations of 10animals pretreated with EACA and then givena lethal dose of endotoxin are shown in Table II.The average values for this treated group areonly slightly lower than the untreated controlgroup. Coincident with the rise in plasma hista-mine there was a moderate fall in blood pressure.The infusion of EACAalone causes a rise in sys-temic pressure, an observation that has been con-firmed by others (16).

This experiment showed that pretreatment ofdogs with EACA significantly reduced the mor-tality rate from endotoxin and also reduced thefall in complement when compared to animals re-ceiving endotoxin alone (p < 0.02), as noted inTable I. In EACA-treated animals plasma his-tamine concentrations rose significantly 60 sec-onds after endotoxin (p < 0.05).

4. Complement titers and histamine concentra-tions in dogs pretreated with cortisol and thengiven a lethal dose of endotoxin. Pretreatmentwith cortisol protects dogs against a lethal doseof endotoxin. Ten dogs were given 100 mg ofcortisol intravenously in 100 ml of 57o dextroseand saline. Seven of the 10 dogs survived. Asseen in Table I this amount of cortisol did notsignificantly alter the initial decline in comple-ment titers 10 minutes after endotoxin, as com-pared to the titers in control endotoxin animals(p = 0.2). The mean pre-endotoxin titer aftercortisol infusion (66 U) was lower than in en-dotoxin-treated dogs (p < 0.01).

Cortisol has a pronounced effect in suppressingthe rise in plasma histamine concentrations result-ing from endotoxin (Table II). The control

140

120

z 100w

w(] 80

0

, 600

u 40

20

- Dead--- Survivor

PRE- 10 2 3 4 5 6FENDO MPTin

-POST ENDOTOXIN

FIG. 2. DECREASEIN COMPLEMENTTITERS IN EACH OF

SIX DOGS GIVEN LD30 DOSE OF ENDOTOXIN (0.25 MG PERKG). The decline in complement was less in the survi-vors than in those that died.

animals receiving endotoxin alone had an eight-fold rise, whereas animals pretreated with cortisolhad slightly less than a fourfold elevation. Inanimals pretreated with cortisol and given endo-toxin the plasma histamine concentrations wereno greater than in those animals given saline solu-tion alone. In addition, the initial fall in sys-temic pressure was somewhat less in cortisol-treated animals than in the control group. Sixtyseconds after injection of endotoxin, when themaximal average level of 0.19 ug per ml of hista-mine was reached in the plasma, the average sys-tolic blood pressure was 136 mmHg. Compar-able values in the control endotoxin group were0.40 Mig per ml of histamine and a pressure of 91mmHg. Although the differences in plasma his-tamine were not statistically significant after 30and 60 seconds, 5 minutes after endotoxin plasmahistamine in the cortisol-treated group was 0.09Mtg per ml, and in the endotoxin controls, 0.27;this difference was significant (p < 0.02). Therespective mean systolic blood pressures were 123and 95 (Table II).

5. Complement titers and histamine concentra-tions in dogs pretreated zwith both EACA andcortisol and then given a lethal dose of endotoxin.Studies reported elsewhere (15, 17) indicated thatdifferent mechanisms were involved for EACAand for cortisol in the protection afforded dogsagainst endotoxin. For this reason 10 dogs were

699

W. W. SPINK, R. B. DAVIS, R. POTTER, AND S. CHARTRAND,

0% 0-o c -

o 1)m

00O O0

c1)

0% 0

1f)

. 0 00 61')000

Xf)~~~~~~~~~Y

0 C'))Oto oo

0. 0

c '1) 0

oo Xf n4

U'. 11) Co4

- 0 0 0

6o 6o i

\00 0

oo Cl

6o o 6v

Cl C' C')

* U. U.

61p 6o 61,

11o) oo

6o 60 6Cl)oo0O

Cl) U).

6o r' 6o

0%o

011)

Cl~~~~~~~~~C

O l1 V)

U'.

00) E f<c0 v. n U)

011

U'- CU-

Co-0 0 >

0%

0 CdC

OO ri . >

1c) 00 _

U Cl

Cl)~~~~~~~~~C0 .U>

6o o 1ioc)u

6o.g4I,) C 00 >-

0- U)

Cl)

o o U U) >Ul)

0~~~~'

cd CE ca cd Cd E2E E8

3-. (n I 0 Clr COm

0 - cd

00 0 0 0 ~~~~~0 CdU'u a4 11)~ 4 cu W~*J). a4 ) U))

700

(n ot-

m0

Cl

(14

0

cx;401.4

00CU" 0

ON00c; tU'-

00 CS*o 000

; 0Cl4

0 0rCx;6 C;

o

i r0

*0r;

*0;0

0I'

0

*)

U)

1.) *.I

I-

0

c)

0

."

ICb4

41

x.0

bo

)z0

in) -4o.1-

6o0Cl4

c;0CN

\0 -

M-4 0 IC'

X11

Cx;

CU Cl

Co 00

11) Olo- 0I- 0

C.)CdU

0.CUC)0

EcU0

aCl),11I.1

U)

U2cu0.~00

,.M'I

*

1:

H

ANAPHYLAXISAND CANINE ENDOTOXINSHOCK

infused simultaneously with 1 g EACAand 100mg cortisol contained in 100 ml 5%dextrose andwater. Seven of 10 animals survived 0.55 mgper kg of endotoxin, which was the same outcomein an equal number of animals pretreated withEACAor cortisol alone.

This combination alone had no effect on thecomplement titer compared to dextrose-infuseddogs. The mean control titer after EACA andcortisol (87 U) did not differ significantly fromthe endotoxin control group (93 U). Ten min-utes after the injection of endotoxin the averagedecrease in titer was less than that observed incontrol endotoxin animals (p < 0.01) or in thosepretreated with either EACAor cortisol (TableI) .

The average concentration of plasma histaminein this group did not exceed 0.09 pg per ml(Table II). Five minutes after endotoxin themean plasma histamine concentration was less(0.08) than in animals given only endotoxin(0.27); the difference was significant (p < 0.02).It was again observed that the systemic pressurerose after the infusion of EACAand cortisol, andthe average level in systolic pressure after endo-toxin in this group did not decline below 100 mmHg.

6. Complement titers and histamine concentra-tions in dogs given a second lethal dose of endo-toxin. A group of dogs treated previously by avariety of methods had survived a lethal dose ofendotoxin (Table III). After a period of sur-vival varying between a few days to over 3months, a second lethal dose of 0.55 mg per kgof endotoxin was given to each of these animals.Although there was an immediate and severe re-action following this second injection, eight ofthe ten animals survived. There was a markeddecline in the titer of complement (Table I),which exceeded that in other groups and wassignificantly greater than in animals which re-ceived one dose of endotoxin (p < 0.05), but thedecrease had no relation to the subsequent deathor survival of the animal.

The highest sustained concentrations of hista-mine were observed in this group of animals(Table III). Within 60 seconds after endotoxinwas injected there was a 13-fold increase. In-

creased concentrations were present 4 hours laterin five animals.

The remarkable feature about these animalswas the prompt fall in complement, the onset ofhypotension, and a marked rise in plasma hista-mine, all of which suggested an antigen-antibodyreaction of- the immediate anaphylactic type.Nevertheless, eight of ten animals recovered.

7. Histamine concentrations in dogs pretreatedwith histamine and then given a lethal dose of en-dotoxin. Although dogs receiving the second doseof endotoxin exhibited the initial manifestations ofsevere anaphylactic shock and had sustained in-creases of histamine, only two of the ten animalsdied (Table II). This suggested that histaminemight actually protect animals against the lethaleffect of endotoxin, although one of its propertiesis the production of hypotension.

Ten adult dogs were each given an infusion ofhistamine phosphate, equivalent to 1 mghistaminebase, in 100 ml 5 % dextrose and water over aperiod of 15 to 20 minutes. Although some hypo-tension was induced with histamine, the averagesystolic blood pressure did not decrease below 90mmHg. After this infusion the animals werethen given a lethal dose of endotoxin (0.55 mgper kg). The initial reaction to the endotoxinwas mild, and eight of the ten animals survived.There was no rise in plasma histamine afterendotoxin (Table II).

8. Control animals. a) Complement titers indogs given Cr. terrificus venom. Six adult ani-mals were given 50 mg of snake venom con-tained in 0.5 ml of saline. Although lethal shockwas produced in all of the dogs the mean com-plement titer was 96%o of the control titer 10minutes after venom, and changed little there-after. b) Complement titers in animals given100 ml 5%o dextrose in distilled water. SinceEACAand cortisol were given intravenously in

-a solution of 5%o dextrose and 100 ml of dis-tilled water, this amount was infused into threeanesthetized dogs. Complement titers fell ap-proximately 19%o during the infusion of dextrose,but thereafter complement changed little, and nochange in blood pressure occurred. c) Plasmahistamine concentrations in dogs given 2 ml ofsaline solution intravenously. After anesthesia,arterial catheterization, and iv injection of 2 ml

701

W. W. SPINK, R. B. DAVIS, R. POTTER, AND S. CHARTRAND

of saline, plasma histamine was measured seriallyin six dogs. No significant changes in concentra-tion were obtained, with the mean (60 determina-tions) being 0.05 fug per ml.

Blood pressure always fell in animals given en-

dotoxin, and the fall was associated with variablerises in histamine and decreases in complementtiters. Nevertheless, calculation of correlation co-

efficients between the maximal fall of blood pres-

sure and maximal change of histamine or com-

plement in each of the six groups showed no sta-tistically significant correlation.

Discussion

The foregoing experiments were primarily con-

cerned with the initial hemodynamic alterationsin the dog that are induced by endotoxin. Thedata support the concept that the initial phase ofcanine endotoxin shock involves an immune mech-anism of the anaphylactic type in which endotoxin,acting as an antigen, reacts with antibody in thepresence of complement, and histamine is liber-ated. This concept does not exclude the appear-

ance of other vasoactive substances in the blood.Other studies also support such a concept. Gil-bert and Braude (6) in immunologic studies on

endotoxin shock in the rabbit concluded that endo-toxin had an anaphylactic effect. Kostka andSterzl (18) observed that the serum of piglets re-

moved from their mothers at birth contained com-

plement but no E. coli antibody. When endotoxinwas added to piglet serum no inactivation ofcomplement occurred. However, the addition ofendotoxin to the sera of adult swine containingantibody did result in a significant decrease incomplement. In contrast to the evidence suggest-ing that endotoxin acts as an antigen in theinitial phase of shock, Franke (19) found that al-though a polysaccharide from Serratia marcescens

was lethal for guinea pigs, the effects did not ap-

pear to be due to anaphylaxis. They measuredonly lung volume and specific gravity, however,and it has been shown in another species thatmany objective signs of toxicity during anaphy-laxis may be absent despite the occurrence ofchemical changes (20).

There is a considerable body of evidence indi-cating that the initial changes in vascular activityin endotoxin shock can be provoked by histamine.

Unique in the dog is the immediate hepatoveno-constriction with an increase in portal vein pres-sure that occurs after endotoxin and is associatedwith a pooling of blood along the intestinal tract.Essex and Thomas (21) observed in dogs that theintravenous injection of histamine resulted inconstriction of the hepatic vein, and Weil andSpink (3) demonstrated a prompt rise in a his-tamine-like substance in the hepatic vein blood ofdogs immediately after the injection of endotoxin.They were able to demonstrate histamine-like ac-tivity in hepatic vein blood but not in femoralvein blood, which is in accord with the variableincreases in histamine found in this study in fe-moral artery samples. Histamine causes veno-constriction in the tissues and organs of otherspecies (22, 23). Endotoxin produces a venousconstriction of pulmonary veins in the cat and dog(24). Finally, the hypoxia and reduction of oxy-gen content in the dog's blood that follows endo-toxin has been produced by histamine (25).

While these studies are consistent with thehypothesis that the early phase of canine endo-toxin shock is related to an antigen-antibody re-action involving complement, resulting in hista-mine liberation, the precise role of histamine inthe over-all genesis of irreversible shock is notclear. Schayer (26) has advanced the thesis thatexcessive amounts of histamine or epinephrineare deleterious to the microcirculation of the host.He has concluded that histamine appears in theblood after the onset of shock as a result of theliberation of preformed histamine and also becauseof continued synthesis of histamine from histidinethrough the action of the enzyme, histidine de-carboxylase (27). The continued synthesis ofhistamine is more important in the genesis ofshock than is the liberation of preformed hista-mine. Others (10) support this dual origin ofhistamine, although Waton (28) questions thecontinuing synthesis of histamine in cats, dogs,and man. Since the present experiments are con-cerned with acute and lethal endotoxin shock, noconclusive data on the continued elevation ofplasma histamine over a period of several hourshave been obtained except in those surviving ani-mals given a second injection of endotoxin.

Several provocative points relating to the sig-nificance of histamine in endotoxin shock havebeen brought out in the present studies. First,

702

ANAPHYLAXIS AND CANINE ENDOTOXINSHOCK

the highest sustained concentrations of plasmahistamine were detected in those dogs that hadbeen given a second lethal dose of endotoxin.This elevation of histamine was accompanied bythe manifestations of severe anaphylactic shock.However, the majority of these animals survived.A state of tolerance to endotoxin was associatedwith acquired endotoxin hypersensitivity. Thenature of this relationship is not clear. It is dif-ficult to correlate the initial severe anaphylacticreaction with the mechanism or mechanisms bywhich recovery ensues, if indeed, a correlationexists. It is possible that the state of tolerance tothe lethal effect of endotoxin is related to thereticuloendothelial system, as the studies of otherswould suggest (29, 30). Second, dogs given aninfusion of histamine over a period of 15 minutesare rendered highly tolerant to the lethal effect ofendotoxin, an observation that others have madein mice (31). An explanation for this phenome-non is not easily forthcoming. The injection ofendotoxin was not followed by significant eleva-tions of plasma histamine after histamine infusion.It is known that histamine stimulates the secre-tion of epinephrine from the adrenal medullae.Is it possible that the injection of appropriateamounts of histamine depletes or diminishes thepotential secretion of epinephrine after the in-jection of endotoxin? This aspect is under in-vestigation. Third, EACA or cortisol protectsdogs against a lethal dose of endotoxin, althoughthe mechanism by which each of these agents ac-complishes this effect is probably different.EACAbut not cortisol pretreatment significantlydecreased the consumption of complement afterendotoxin, when compared to the results in con-trol animals (Table I). The rise in plasma his-tamine from endotoxin was no different in EACApretreated animals, but was much less in corti-sol-treated animals than in control animals givenendotoxin (Table II). Liberation of histaminefrom cells is probably inhibited by cortisol at thecell surface, since the liberation of several cellularenzymes by endotoxin is considerably diminishedboth in vitro and in vivo in the presence ofpharmacologic amounts of cortisol (32, 33).

It is apparent that endotoxin shock involves achain of biochemical and functional alterations.Because of this complexity shock can be reversedat different stages and by a variety of agents

(34). Further progress in the understanding ofthis type of shock is dependent upon a more pre-cise chemical identification of endotoxin. Addi-tional information is desirable on the quantitativerelationship of histamine and catecholaminesduring the various phases of shock. Other vaso-active substances may also be involved. Dataare especially needed in different species, includ-ing man.

Summary

Experimental data in canine endotoxin shocksupport the concept that the initial stage of hemo-dynamic alterations is due to an anaphylactic typeof immune mechanism involving complement,with the liberation of histamine. However, theseverity of the systemic reaction is not indicativeof the final outcome of animals. Epsilon-amino-caproic acid and cortisol modify the severity ofthe initial reaction, and the majority of animalssurvive lethal doses of endotoxin. On the otherhand, surviving animals given a second lethaldose of endotoxin have a severe initial reaction, asignificant decline in complement, and a markedrise in plasma histamine. The majority of theseanimals also survive. Finally, dogs infused withhistamine survive a lethal dose of endotoxin. Therelationship of the initial anaphylactic activity tothe ultimate outcome of an animal with endotoxinshock is not clear. Tolerance or intolerance tothe lethal action of endotoxin is dependent uponother mechanisms and may be related to acquiredhumoral immunity and conditioning of the reticu-loendothelial system.

References1. Weil, M. H., L. D. MacLean, M. B. Visscher, and

W. W. Spink. Studies on the circulatory changesin the dog produced by endotoxin from gram-nega-tive microorganisms. J. clin. Invest. 1956, 35,1191.

2. Hinshaw, L. B., W. W. Spink, J. A. Vick, E. Mal-let, and J. Finstad. Effect of endotoxin on kid-ney function and renal hemodynamics in the dog.Amer. J. Physiol. 1961, 201, 144.

3. Weil, M. H., and W. W. Spink. A comparison ofshock due to endotoxin with anaphylactic shock.J. Lab. clin. Med. 1957, 50, 501.

4. Vick, J. Studies on the trigger mechanism involvedin the vascular alterations induced by endotoxin.(abstract). J. Lab. clin. Med. 1960, 56, 953.

703

W. W. SPINK, R. B. DAVIS, R. POTTER, AND S. CHARTRAND

5. Spink, W. W., and J. Vick. A labile serum factor inexperimental endotoxin shock: cross-transfusionstudies in dogs. J. exp. Med. 1961, 114, 501.

6. Gilbert, V. E., and A. I. Braude. Reduction of serumcomplement in rabbits after injection of endo-toxin. J. exp. Med. 1962, 116, 477.

7. Spink, W. W., and R. Potter. Endotoxin shock andimmunologic factors: studies on complement. Fed.Proc. 1963, 22, 628.

8. Greisman, S. E. Activation of histamine-releasingfactor in normal rat plasma by E. coli endotoxin.Proc. Soc. exp. Biol. (N. Y.) 1960, 103, 628.

9. Hinshaw, L. B., J. A. Vick, C. H. Carlson, and Y-L.Fan. Role of histamine in endotoxin shock.Proc. Soc. exp. Biol. (N. Y.) 1960, 104, 379.

10. Hinshaw, L. B., M. M. Jordan, and J. A. Vick.Histamine release and endotoxin shock in theprimate. J. clin. Invest. 1961, 40, 1631.

11. Halberg, F., and W. W. Spink. The influence ofBrucella somatic antigen (endotoxin) upon thetemperature rhythm of intact mice. Lab. Invest.1956, 5, 283.

12. Kabat, E. A., and M. M. Mayer. ExperimentalImmunochemistry, 2nd ed. Springfield, Ill.Charles C Thomas, 1961, p. 133.

13. Noah, J. W., and A. Brand. A fluorometric methodto determine levels of histamine in human plasma.J. Allergy 1961, 32, 236.

14. Shore, P. A., A. Burkhalter, and V. H. Cohn, Jr.A method for the fluorometric assay of histaminein tissues. J. Pharmacol. exp. Ther. 1959, 127,182.

15. Spink, W. W., and J. Vick. Endotoxin shock andthe coagulation mechanism: modification of shockwith epsilon-aminocaproic acid. Proc. Soc. exp.Biol. (N. Y.) 1961, 106, 242.

16. Ramos, A. O., L. F. Chapman, A. P. Corrado, andV. A. Fortes. Sympathomimetic action of theepsilon-amino caproic acid. Arch. int. Pharma-codyn. 1961, 132, 274.

17. Spink, W. W., and J. Vick. Evaluation of plasma,metaraminol, and hydrocortisone in experimentalendotoxin shock. Circulat. Res. 1961, 9, 184.

18. Kostka, J., and J. Sterzl. The action of endotoxin oncomplement. Folia microbiol. (Praha) 1962, 7,155.

19. Franke, F. E. Attempts to produce anaphylaxis inguinea pigs with a polysaccharide from Serratiamarcescens (Bacillus prodigiosus). U. S. Nat.Cancer Inst. J. 1944, 5, 173.

20. Waalkes, T. P., H. Weissbach, J. Bozicevich, and S.Udenfriend. Serotonin and histamine release dur-ing anaphylaxis in the rabbit. J. clin. Invest.1957, 36, 1115.

21. Essex, H. E., and W. D. Thomas. The response ofthe hepatic venous circulation to certain substancesgiven intravenously. Proc. staff Meeting MayoClinic 1950, 25, 34.

22. Gilbert, R. P., L. B. Hinshaw, H. Kuida, and M. B.Visscher. Effects of histamine, 5 hydroxytrypta-mine and epinephrine on pulmonary hemodynamicswith particular reference to arterial and venoussegment resistances. Amer. J. Physiol. 1958, 194,165.

23. Haddy, F. J. Effect of histamine on small andlarge vessel pressures in the dog foreleg. Amer.J. Physiol. 1960, 198, 161.

24. Kuida, H., L. B. Hinshaw, R. P. Gilbert, and M. B.Visscher. The effect of gram-negative endotoxinon the pulmonary circulation. Amer. J. Physiol.1958, 192, 335.

25. Kowalewski, K., and C. Wisniewski. Oxygen con-tent and pH of arterial, venous and portal bloodin dogs after histamine. Proc. Soc. exp. Biol.(N. Y.) 1962, 109, 161.

26. Schayer, R. W. Evidence that induced histamineis an intrinsic regulator of the microregulatorysystem. Amer. J. Physiol. 1962, 202, 66.

27. Schayer, R. W. Induced synthesis of histamine, mi-crocirculatory regulation and the mechanism of ac-tion of the adrenal glucocorticoid hormones inProgress in Allergy. Basel, Switzerland, S. Karger,1963, vol. 7, p. 187.

28. Waton, N. G. Is tissue histamine formation neces-sary in cat, dog and man? Int. Arch. Allergy1963, 22, 87.

29. Carey, F. J., A. I. Braude, and M. Zalesky. Studieswith radioactive endotoxin. III. The effect oftolerance on the distribution of radioactivity afterintravenous injection of Escherichia coli endotoxinlabeled with Cr51. J. clin. Invest. 1958, 37, 441.

30. Fine, J., H. Rutenberg, and F. B. Schweinberg. Therole of the reticulo-endothelial system in hemor-rhagic shock. J. exp. Med. 1959, 110, 547.

31. Fox, C. L., Jr., and S. E. Lasker. Protection byhistamine and metabolites in anaphylaxis, scalds,and endotoxin shock. Amer. J. Physiol. 1962, 202,111.

32. Melby, J. C., R. H. Egdahl, I. C. Bossenmaier, andW. W. Spink. Suppression by cortisol of increasedserum-transaminase induced by endotoxin. Lan-cet 1959, 1, 441.

33. Weissmann, G., and L. Thomas. Studies on lyso-somes. I. The effects of endotoxin, endotoxin tol-erance, and cortisone on the release of acid hy-drolases from a granular fraction of rabbit liver.J. exp. Med. 1962 116, 433.

34. Spink, W. W. Endotoxin shock. Ann. intern. Med.1962, 57, 538.

704