INFECTION AND IMMUNITY, Aug. 1970, p. 175-182Copyright ©) 1970 American Society for Microbiology

Vol. 2, No. 2Printed in U.S.A.

Prevention of Experimental Hematogenous andRetrograde Pyelonephritis by AntibodiesAgainst Enterobacterial Common Antigen

GERALD DOMINGUE, ALI SALHI, CARL ROUNTREE, AND WALTER LITTLE

Tulane University School of Medicine, Department of Surgery, Section of Urology, and Department of Micro-biology and Immunology, New Orleans, Louisiana 70112

Received for publication 24 March 1970

Rabbits were immunized with common antigen (CA) derived from Salmonellatyphimurium. Animals with CA hemagglutinin titers of 1-320 to 1-81,920 were in-jected with 107 viable Proteus mirabilis via the retrograde and hematogenous routes.Nonimmune control groups were challenged similarly. From the retrogradelychallenged groups sacrificed at 4 weeks, pyelonephritis was found in 89%, of thecontrol animals but not in those immunized. Bacteriuria was present in 89% of thecontrols but in only 61% of the immunized group. Hematogenously challenged,immune animals sacrificed at 6 weeks did not show histological evidence of renalpathology, and only 6% had bacteriuria. Eighty-six per cent of the nonimmunecontrols showed both pyelonephritis and bacteriuria. The protective effect isspecific because of the following. (i) Active and passive CA immunization did notprevent pyelonephritis due to Pseudomonas aeruginosa which does not produce thisantigen. (ii) Passive immunization with a CA antiserum conferred protection againsta P. mirabilis challenge. (iii) Passive administration of an antiserum from which CAantibodies had been differentially absorbed abolished the protective activity.

There is very limited information available onthe immune response and the role of0 antibodiesin protection against urinary tract infections (1,3, 9). Also, because of the persistence or re-currence of urinary tract infections with a strainof the same serological type in patients withhigh titers of homologous 0 antibodies, thevalue of host immunoglobulins versus thesomatic 0 antigen of such bacteria has beenquestioned by many investigators. Numerousenteric bacteria, including Escherichia, Enterobac-ter, Proteus, Serratia, Salmonella, and Shigella,share a common antigen (CA), in addition tothe well known 0, H, and Vi antigens whichthey may have. The common antigen first de-scribed in 1962 by Kunin et al. (8) had probablyescaped detection for many years because itcould be demonstrated only by means of thehemagglutination and hemolysis tests and not bybacterial agglutination or precipitation pro-cedures. Antibodies against the CA have beenshown to opsonize enteric bacteria containingthis antigen (4, 5). Since phagocytosis is one ofthe major defense mechanisms of the host, it isconceivable that such antibodies may play aprotective role in infections including pyelone-

phritis of man and animals. In 1968, Gorzynskiand Neter (E. A. Gorzynski and E. Neter, Bac-teriol. Proc., p. 86, 1968) reported that C57BL/6Ha black mice immunized with the CA ofEscherichia coli 014 had a death rate of 47 to57% as compared with 72 to 80% for controlswhen challenged with Salmonella enteritidis.Swiss white mice, which produced CA antibodiesless effectively than black mice, were not pro-tected. The protective effects of antibodies againstthe enterobacterial CA in experimental hematog-enous and retrograde pyelonephritis of rabbitsare described in this paper.

MATERIAL AND METHODSCommon antigen preparation. CA was prepared by

the previously described procedures of Suzuki et al.(10). The various bacterial strains were grown onBrain Veal Agar (Difco) in Kolle flasks for 18 hr at37 C. The growth was suspended in hemagglutinationbuffer (Difco) and boiled for 1 hr. After centrifugationat 23,500 X g for 20 min, the supematant fraction wasexposed to ethyl alcohol (final concentration of 85%)for 18 hr. The mixture was centrifuged; the super-natant was air-dried for 24 hr and reconstituted indistilled water to the original volume.

Serum titrations: indirect hemagglutination (IHA).175

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

DOMINGUE ET AL.

Serum was obtained from all rabbits before im-munization, before bacterial challenge, and at sacrifice.Sheep erythrocytes were exposed to the test CApreparation for 0.5 hr at 37 C. After three washings inhemagglutination buffer, 0.05 ml of 2.5% suspensionof sensitized erythrocytes was added to 0.5 ml of testserum dilution from 1: 5 to 1: 81,920. After incubationfor 0.5 hr at 37 C, the test mixtures were centrifugedat 2,000 rev/min for 2 min and hemagglutination wasread as the sediments were shaken. The highest serumdilution producing visible hemagglutination wastaken as the IHA titer.

Immunization. Rabbits (adult New Zealand whites)weig'hing 4 to 7 lb were actively immunized by intra-venous injection on alternate days as follows: threeinijections (0.5-ml doses of S. typhimurium heat-killedsupernatant) for priming to CA followed by twoinjections (0.5-ml doses of ethanol-soluble CA derivedfrom S. typhimurium). Blood samples were obtainedbefore and after immunization, and serum antibodieswere determined by the IHA test. Three groups ofrabbits were passively immunized by a single intra-venous injection of immune serum. One group wasinjected intravenously with 0.5 ml of rabbit CA anti-serum with an IHA titer of 1:10,240 and challenged4 hr later with 107 viable Proteus mirabilis organisms;another was injected intravenously with 1 ml of rabbitCA antiserum with an IHA titer of 1:640 and chal-lenged 4 hr later with 107 viable P. mirabilis organisms;and a third group was injected intravenously with I mlof rabbit CA antiserum with an IHA titer of 1:640and challenged 4 hr later with 107 viable Pseudomonasaerutginosa organisms.Hematogenous challenge. Seven groups of rabbits

were fed 1.2% oxamide mixed in their diet for 6 days(7) and on the sixth day were injected intravenouslywith 107 live P. mirabilis or P. aeruginosa organisms.These strains were obtained from patients with acutepyelonephritis. The next day, oxamide was dis-continued from the diet. During the feeding of oxam-ide, microscopic intraductal calcifications are found.When the oxamide administration is stopped, thecalcifications are soon absent. Pyelonephritis thusinitiated, however, continues.

Retrograde challenge. Female rabbits were anes-thetized with ether. The urinary bladders were emptiedby needle aspiration and a sterile wound clip wassutured to the bladder wall near one of the ureteraljunctions. To establish a chronic urinary tract infec-tion, a foreign body must be applied; otherwise spon-taneous cure of the infection occurs. Only thoseanimals with intact wound clips at autopsy wereincluded in the final analysis of the results. Afterclosing the bladder, 1 ml of the bacterial suspension(containing ten million viable P. mirabilis organisms)was injected through the bladder wall. Refluxing urinewas expected to carry bacteria to the kidneys forcolonization.

Absorption of antiserum. Sheep erythrocytes weremodified with ethanol-soluble CA derived from E. coli04. Sediment of the CA-modified erythrocytes wasmixed with S. typhimnurium antiserum (containing0 + CA antibodies) which had been inactivated at56 C for 30 min, and absorption was allowed to

proceed for 30 min in a water bath at 37 C followed byovernight incubation in a cold room. This antiserumhad a heterologous hemagglutinin CA titer of 1:640before absorption. The absorption process was re-peated three times and the absorbed antiserum wasshown to have a CA hemagglutinin titer of less than 1to 2.

Passive transfer of CA absorbed antiserum. A 0.5-mlamount of absorbed antiserum (with 0 antibodiesremaining) was injected intravenously into 10 rabbitsfollowed by the iv administration of 5 X 107 viableP. mirabilis organisms. A 0.5-ml amount of un-absorbed serum (containing 0 + CA antibodies) wasinjected into nine control rabbits and the animals werelikewise challenged.

Postchallenge treatment. Six weeks after hematog-enous and four weeks after retrograde challenge withlive bacteria, each rabbit was bled just before sacrifice.Both kidneys were removed aseptically, and urinewas removed by needle puncture of the bladder forquantitative cultures.The individual kidneys were weighed and examined

grossly for abscesses. A section was cut for histo-pathology, and a weighed portion was then groundwith Trypticase Soy Broth by using a motorized tissuehomogenizer. Pour plates of the suspension weremade by using Trypticase Soy Agar. The plates wereincubated at 35 C for 18 to 24 hr, and the viable countwas expressed per gram of kidney tissue. Representa-tive bacterial colonies were chosen, and biochemicaland drug susceptibility patterns and slide agglutina-tion tests with homologous antiserum were used ascriteria to insure identity of the organism with theoriginal challenge strain. Thin sections of both kidneysof each rabbit were stained with hematoxylin-eosinand examined for evidence of renal pathology withoutknowledge by the pathologist as to identity of speci-mens.

RESULTSTable 1 summarizes results of experiments on

the effect of active immunization with CA of S.typhimurium on hematogenous and retrogradechallenge with P. mirabilis. It is evident that CAimmunization provided significant protectionagainst renal infection in the hematogenouslychallenged group. The results also clearly showdifferences (significant at the 99.5% level) be-tween the immune and nonimmune groups withrespect to bacterial recovery rates from urinesand kidneys as well as renal pathology. The datain Table 1 indicate that there was a difference(significant at the 95% level) between the im-mune and controls in the retrogradely challengedrabbits when comparing development of kidneypathology. As for the recovery of bacteria fromthe urine, there was no significant difference.Although immunization protected rabbits fromdeveloping renal pathology, bladder infectionwas found in the majority of these animals.Homologous Proteus serum IHA titers (O anti-

176 INFEC. IMMUN.

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

PROTECTIVE EFFECTS OF ANTIBODIES

40

cd

._4.

g:.a

404obocd

44

ax

I,

C40:1

0

0

44

u70

biD040

vu

CdU)

40

U)

u)40

COd

40bo

r.

U

o ow

tn"

V -

I I

V V

se_I 'i V_1 I-

0 0

( t V_

0i

IIV

V V

0 -

0

o, :oCO,)00 V

.do U 0 d O 0 00)0.

) VI 0 0 00

- e 0Z;0->> ° 9° l-

ce )o- O X )' X X

Cd o 0

0 = E <

444)~ ~ ~ ~~C

4)bC.~~~~C) C) C00

4)0~~~~~~~~~~~~~~0 0

0ed .C -1- r

4.) .-0 -Z)co

| E g Eg~~~E0 0 %Z

P- 4, 00 0

0

0

40o

.0u

CA

0 0C IC4) 1dOD 000 0

E4)9

VOL. 2, 1970

4)4

04.0.,

0

4.

*.05

4..000

4.4.

0

0

o .^

-

0

0-

04

*0

0

I'

-440

*

177

4)CnQ)

CO4L)0.

CO

COCO

4L)

20

4)

E._.

C:_

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

DOMINGUE ET AL.

ao

03

L)

Z:

4a

0co

I..

e

._

ts

4)

3

-'4

It

Ut'Q

EI000

'4

UXCd

0bO

204.4)

'C

ci)Ce

-C

'4.

ceEn

Cd

uI

o o

0 0

-4

-qWl lot

v V

0

*--00 CNI

to

0

O tm00

I4

V

Er)

)) .- .C

U'A0-4) tr .

.0

440U6. 0 C'

4i t~~~~~~ r)C's

_aa

1---B 0~~~~~~~c

~~-~ 4X0tt)

.1 0 t

0c

CA ~~~r- t

SLi

a3aL)

e .a)

0

bodies) were found to rise during the course ofthe experiment in the nonimmune groups butdid not correlate with the presence or absence ofbacteria and apparently conferred no significantprotection.To assess the specificity of the CA antibody

protective effect, the following experiments wereperformed. The first concerned the effect of im-munization with CA on hematogenous pyelone-phritis after challenge with P. aeruginosa, a micro-organism which does not contain CA. There wasno significant difference between the immune andnonimmune groups in bladder and kidney infec-tion rates or occurrence of renal pathology(Table 2). However, a striking discrepancy wasobserved between rate of renal pathology andkidney and bladder infection. P. aeruginosacaused damage to the kidneys but, at the timeof sacrifice, was present only infrequently.Homologous Pseudomonas serum IHA titerswere found not to correlate with the presence orabsence of bacteria. The second series of ex-periments to prove specificity involved passivetransfer of immune serum to animals which werechallenged 4 hr later with P. mirabilis or P.aeruginosa. In the Proteus challenge groups, onegroup was given potent, the other moderatedosages of antiserum. CA antiserum providedsignificant protection, even when low-titeredantiserum was used (Table 3). In addition, it isevident that passive administration of CA anti-serum did not confer protection against pyelone-phritis due to P. aeruginosa, a microorganismdevoid of the common antigen.

Lastly, CA antibodies were differentially ab-sorbed from an S. typhimurium antiserum con-taining 0 + CA antibodies and the absorbedantiserum was passively administered to rabbits.Removal of CA antibody abolished the protec-tive activity of the antiserum, since 100% of theanimals challenged with P. mirabilis in this grouphad bacteria in the kidneys with evidence ofrenal histopathology at 4 weeks postchallenge.Animals injected with unabsorbed antiserumcontaining 0 + CA antibodies demonstratedsignificant protection.The histological changes noted in the kidneys

were categorized into four rather distinct groupsas follows. (i) Calcific deposits as well as focalscar formations were localized largely in thecortex; (ii) foci of acute pyelonephritis were

, found in several instances; (iii) pyelitis was° common to a large proportion of cases; and (iv)

a peculiar vacuolar nephropathy was also notedoccasionally. These characteristics were com-

monly seen in the nonimmune control groups

178 INFEC. IMMUN.

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

PROTECIVE EFFECTS OF ANTIBODIES 179

TABLE 3. Ccmparison of pyelonephritis in passively immunized rabbits challenged with Proteus mirabilis

or Pseudomonas aeruginosa

Passive immunization andchallenge bacteria

High-titered CA serumc andProteus

Low-titered CA serumd andProteus

None and Proteus

Low-titered CA serum andPseudomontas

None and Pseudomonias

Rabbits with recovery of bacteria at sacrifice from

Urine

Ratio ofno. positiveper total

1/7

1/7

6/7

1/5

1/5

Range oforganisms/mlb

(14 X 103)

(lO6)

105-3 X 106(1.5 X 106)(33 X 102)

(3 X 106)

Kidneys

Ratio ofno. positiveper total

0/7

1/7

6/7

1/5

2/5

Range oforganisms/gb

(13)

27 X 104-101(5.2 X 105)(24 X 102)

5 X 103-3 X 106(1.1 X 106)

a Histopathology-acute pyelonephritis, focal scar formation, pyelitis.I Value for mean given in parentheses.c Group given high-titered (1:10,240) common antigen (CA) antiserum.d Group given low-titered (1:640) CA antiserum.

Rabbitswith renal

histo-pathology"(ratio of

no. positiveper total)

0/7

1/7

6/7

4/5

4/5

TABLE 4. Effects of common antigen (CA) antibody absorption on passive immunization in animalschallenged with Proteus mirabilis

Recovery of P. mirabilis from kidneys at sacrifice Rabbits with renal

Serum administered histopathologya(ratio o0 no. positiveRatio of no. Ragaraimsge ol

positive per total Range organisms/gb per tot

CA absorbed serumc 10/10 106-3 X 106 10/10(2.3 X 106)

Unabsorbed serumd 0/9 0/9

None 6/7 7 X 104-106 6/7(5.2 X 105)

a Histopathology: acute pyelonephritis, focal scar formation, pyelitis.b Value for mean given in parentheses.c Heterologous CA hemagglutinin titer, <1 to 2; homologous 0 titer before absorption, 1:640; ho-

mologous 0 titer after absorption, 1:640.d Containing heterologous 0 antibodies; heterologous CA hemagglutinin titer, 1:640.

after challenge and absent in the kidneys of im-mune animals.

Calcific deposition in renal tissue was char-acterized by marked variation in degree fromanimal to animal. In most animals, the degreeof involvement of the two kidneys was approxi-mately equal. The material was localized largelyin the convoluted tubules and, in many cases,was grossly visible as rather discrete white areas

in the cortex.

The foci of early scar formation were composedof chronic inflammation, largely lymphocytic,and early fibroblastic proliferation. In somecases, calciferous depositions and foci of materialappearing to represent hemosiderin were alsofound. Loss of parenchymal tissue with collapseof remaining stroma, as well as the immediatesubcapsular location of many of these areas,caused flat broad-based depressions on the sub-capsular cortical surfaces of the kidneys in

VOL. 2, 1970

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

DOMINGUE ET AL.

several instances. There appeared to be no con-sistent relationship between these cicatrizingareas and calcific deposits, as each phenomenonwas seen to occur independently. The destructionof parenchyma, chronic inflammation, andscarring were considered to constitute evidenceof healing acute pyelonephritis.

In the hematogenously challenged animalsdemonstrating acute pyelonephritis histologically,the lesions were largely of the "cannonball"type. They were, for the most part, rounded dis-crete masses of inflammatory infiltrate. Destruc-tion of preexisting tissue and abscess formationwere a common finding. The rounded discretenature of the lesion was suggestive of hematoge-nous pyelonephritis, rather than "ascending"infection.

Pyelitis, when present, was sometimes unilat-eral, but when bilateral, differences in severity,such as between the two kidneys, were frequentlydemonstrated. When acute inflammatory cellswere present, sections of the renal parenchymausually showed acute pyelonephritis as well.An inconstant finding was a focal vacuolar

nephropathy, the significance of which was notdetermined. It was frequently associated withareas of scarring. In some instances, the cellmembranes of the involved tubular epitheliumappeared to be calcified.

DISCUSSIONEvidence has been obtained suggesting that

immunization with the enterobacterial commonantigen derived from S. typhimurium protectsrabbits against pyelonephritis after hematogenousand retrograde challenge with unrelated entericbacteria containing CA.The following lines of evidence indicate that

this protective effect is specific. (i) CA immuniza-tion did not prevent pyelonephritis due to P.aeruginosa which does not produce this antigen.(ii) Animals passively immunized with CAantiserum did not develop pyelonephritis due toP. mirabilis. (iii) Passive immunization did notprevent a P. aeruginosa pyelonephritis. (iv) Dif-ferential absorption of CA antibodies from anantiserum containing 0 + CA antibodies abol-ished the protective activity of the antiserumwhen passively administered.

Since the vaccines utilized in priming injectionsfor immunizing the animals contained endotoxin,the nonspecific protective activity of endotoxinmight be suggested as a possible explanation forour findings. In addition to the afore mentionedevidence which serves to argue against thishypothesis, it must be emphasized that a com-pletely heterologous 0 system was employed.

There is no known cross-reaction between S.typhimurium 0 antibodies and P. mirabilis. Also,if antirough antibodies, anti-O antibodies, ornonspecific endotoxin protection were playing arole, one would have expected to observe thisphenomenon with the animals challenged with P.aeruginosa. Pseudomonas and Proteus both con-tain endotoxin, yet nonspecific protection wasnot evident in any of the controls. The protectivecapacity of passively transferred antibody ex-cludes not only tolerance but also delayed hyper-sensitivity as the sole mechanism of protection.Since a variety of Enterobacteriaceae have notbeen employed as challenge microorganisms, wehave not eliminated the possibility that theprotective effect may be due to antibodies directedagainst an antigen shared by Salmonella andProteus only, and capable of being absorbedfrom the serum by an ethanol-soluble fractionderived from E. coli 04.The resistance to pyelonephritis as reported by

Sanford et al. (9) was specific, acquired im-munity due to circulating antibodies. Thisresistance was type-specific. Both early non-specific resistance and tolerance are classicallynot type-specific but are shared between variousspecies of gram-negative bacteria. In their study,significant reduction in gross renal abscesses wasnot observed with any of the four heterotypicstrains of E. coli or with a strain of Klebsiellapneumoniae type C when injected into animalswhich had been previously infected with aheterologous strain of E. coli and allowed to heal.Nonspecific protection of endotoxin or antiroughantibodies, or both, did not seem to play a rolein protection in these studies. Furthermore,animals so injected would not be expected todevelop CA antibodies to any appreciable titers.It is therefore difficult to implicate any protectiveactivity due to CA antibodies in this study inview of the experimental design.The frequent absence of bacteria in the urines

and kidneys of the immunized groups challengedhematogenously lends support to our interpreta-tion of contrary findings (absence of pyelone-phritis but presence of bacteriuria) in the urineof animals in the experiments on retrogradepyelonephritis: serum, rather than secretory, CAantibodies neutralize infection depending on themethod of bacterial seeding. We are thereforeproposing that multiplication of bacteria may bemarkedly retarded within the blood or kidneys,or both. It remains to be determined whetherthere is more rapid removal of bacteria from thecirculation in the immune versus nonimmuneanimals, thereby diminishing the bacterial inocu-lation to which the renal tissue is exposed.

180 INRC. IMMUrN

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

PROTECTIVE EFFECTS OF ANTIBODIES

Two possibilities must be considered in inter-preting the observation that Pseudomonas in thecontrol and immune groups caused renal pathol-ogy, but was eliminated from the kidneys in theprocess. The microorganisms may have convertedto persisting L-forms and were, therefore, notdetectable on the routine media employed. Themicroorganisms may have been rapidly destroyedby the animals' normal defense mechanisms andextracellular products released after their destruc-tion may have caused tissue damage. It must beemphasized that the majority of these animalswere sacrificed at six weeks. Those few animalssacrificed at 1 and 2 weeks showed the presenceof the challenging microorganism. In any event,active chronic pyelonephritis without bacteriuriais known to occur in man (2).An explanation for the occurrence of obvious

protection from development of renal pathology,but not from a urinary bladder infection, afterchallenge with P. mirabilis in the retrogradelychallenged immune group may well be found inthe observation that the anti-CA globulins areof the 19S type, which, as far as is known, havenot been reported in the urine. The presence ofbacteria in low count in the renal tissues of 4 ofthe 13 immunized animals could be the result ofcontinual reflux of infected urine into the kidneyswhich is known to occur in the presence of chroniccystitis. This then may be interpreted as internalcontamination since no apparent histologicalchanges were observed. The CA antibody titersby IHA for three of these animals closely paral-leled those of the other immunized animals whosekidneys were negative on culture. Only one ofthe four had a CA antibody titer of less than1:320 at sacrifice.Unfortunately, the immune mechanisms of the

urinary tract which may be highly significant interms of defense against infection are incom-pletely explored. Although the protective mech-anism of CA antibody appears to be humoral,attention deserves to be focused also on cellularimmune mechanisms which are obviously ofconsiderable importance in bacterial infections.It has been previously shown that intravenousinjection of CA leads to the formation of CAantibody-forming cells in the spleen and ad-ministration of CA into a foot pad results in asubstantial increase in antibody-forming cells inthe corresponding lymph node, as tested bymeans of the hemolytic plaque technique ofJerne (6). It will be interesting to learn whethercell-bound antibodies with CA specificity arefound in the kidney and urinary bladder.The finding that the 0 antigen (lipopolysac-

charide) and its lipoid A component interfere

with the immunogenicity of the common antigenis of probable significance here (11, 13). In-dividuals with various enterobacterial infectionshave been shown to develop very low hemag-glutinating titers to CA antibodies with theprobable exception of E. coli 014 (12). As hasbeen experimentally documented, it is only whenCA is separated from the 0 antigen that highhemagglutinating antibody titers develop (10).In a natural infection, CA antibodies may,therefore, seldom reach levels which are protec-tive because of the immunosuppressive action ofthe bacterial llpopolysaccharide.

In addition to the reported protective effects ofCA antibodies, this paper further substantiatesthe suitability of the oxamide-feeding experi-mental model for studying hematogenous pyelo-nephritis. Oxamide forms obstructive oxalatemicrodeposits in the collecting tubules, thusallowing for implantation of bacteria in thekidneys. These deposits disappear soon after theoxamide diet is discontinued. The initial feedingof oxamide allows the study of the disease in anessentially normal urinary tract and the variablesof surgical or traumatic damage to the urinarytract are avoided, thus mimicking human disease.This model undoubtedly lends itself well to variedmanipulation in experimental pyelonephritis.The dynamics of the CA antibody protection

are being further elucidated in our laboratorywith a view to shedding further light on thefollowing: duration of protective effect and long-term effectiveness of the CA antibodies unrelatedto chemotherapy; booster effects on the antibodylevels with relation to protective activity; therelationship between protective effects and variousinfective agents containing CA; and effects ofpassively administered CA antibody on thecourse of an already established infection.Further studies may eventually lead to clinicalapplication.

ACKNOWLEDGMENTS

The investigation was supported by Public Health Servicegrants 5-ROl-HE04659-10 from the National Heart and LungInstitute and 5-SO1-FR-05377-06 from the Division of ResearchFacilities and Resources, National Institutes of Health.

The helpful suggestions of Erwin Neter, Morris F. Shafrer, andJ. U. Schlegel, and the technical assistance of Carole Richtmyer,Nancy Douglas, and Catherine Ehrhardt are greatly appreciated.

LITERATURE CITED

1. Andersen, H. J. 1966. Studies of urinary tract infections ininfancy and childhood. VII. The relation of E. coli anti-bodies in pyelonephritis as measured by homologous andcommon (Kunin) antigen. J. Pediat. 68:542-550.

2. Angell, H. E., A. S. Relman, and S. L. Robbins. 1968. "Ac-tive" chronic pyelonephritis without evidence of bacterialinfection. N. Engl. J. Med. 278:1303-1309.

VOL. 2, 1970 181

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

DOMINGUE ET AL.

3. Diaz, F., and E. Neter. 1968. Antibody response to the com-

mon enterobacterial antigen of children with Shigellosis,Salmonellosis or urinary tract infection. Amer. J. Med.Sci. 256:18- 24.

4. Domingue, G. J., and E. Neter. 1966. Opsonizing and bac-tericidal activity of antibodies against common antigen ofenterobacteriaceae. J. Bacteriol. 91:129-133.

5. Domingue, G. J., and E. Neter. 1966. Inhibition by lipopoly-saccharide of immune phagocytosis of latex particlesmodified with common antigen of enteric bacteria. Proc.Soc. Exp. Biol. Med. 112:133-137.

6. Domingue, G. J., and E. Neter. 1967. The plaque test for thedemonstration of antibodies against the enterobacterialcommon antigen produced by spleen and lymph node cells.Immunology 13:539-545.

7. Fried, F. A., and R. J. Wong. 1969. Etioiogy of pyelone-phritis: intraductal crystallization as a co-factor. J. Urol.101:786-790.

8. Kunin, C. M., M. U. Beard, and N. H. Halmagyi. 1962.Evidence for a common hapten associated with endotoxin

fractions of E. coli and other enterobacteriaceae. Proc.Exp. Biol. Med. 111:160-166.

9. Sanford, J. P., B. W. Hunter, and L. L. Sonda. 1962. Therole of immunity in the pathogenesis of experimental hema-togenous pyelonephritis. J. Exp. Med. 115:383-410.

10. Suzuki, T., E. A. Gorzynski, and E. Neter. 1964. Separationby ethanol of common and somatic antigen of entero-bacteriaceae. J. Bacteriol. 88:1240-1243.

11. Suzuki, T., H. Y. Whang, E. A. Gorzynski, and E. Neter.1964. Inhibition by lipopolysaccharide (endotoxin) of anti-body response of rabbit to common antigen of entero-bacteriaceae. Proc. Soc. Exp. Biol. Med. 117:785-789.

12. Whang, H. Y., and E. Neter. 1963. Study of heterogenetic(Kunin) antibodies in serum of healthy subjects and chil-dren with enteric and urinary tract infections. J. Pediat. 63:412-419.

13. Whang, H. Y., 0. Luderitz, 0. Westphal, and E. Neter. 1965.Inhibition by lipoid A of formation of antibodies againstcommon antigen of enterobacteriaceae. Proc. Soc. Exp.Biol. Med. 120:371-374.

182 INFEC. IMMUN.

on May 29, 2020 by guest

http://iai.asm.org/

Dow

nloaded from