Vol. 26, No. 2INFECTION AND IMMUNITY, Nov. 1979, p. 728-7350019-9567/79/11-0728/08$02.00/0

Experimental Chlamydial Salpingitis in ImmunosuppressedGuinea Pigs Infected in the Genital Tract with the Agent of

Guinea Pig Inclusion ConjunctivitisHAROLD J. WHITE,' ROGER G. RANK,2 BERNARD L. SOLOFF,3 AND ALMEN L. BARRON2*Departments ofPathology,' Microbiology and Immunology,2 and Anatomy,3 University ofArkansas forMedical Sciences, Little Rock, Arkansas 72201, and Medical Research Service, Veterans Administration

Medical Center, Little Rock, Arkansas 72206

Received for publication 21 August 1979

At necropsy indication of spread of infection to fallopian tubes was found in 25of 41 (60%) female guinea pigs infected in the genital tract with the chlamydialagent of guinea pig inclusion conjunctivitis and immunosuppressed with cyclo-phosphamide. Eighteen were examined histologically, and the diagnosis of acutesalpingitis was confirmed in 10, based on inflammatory reaction, detection ofguinea pig inclusion conjunctivitis in tissue, and formation of cysts (pyosalpinxand hydrosalpinx). Infection of fallopian tube tissue was confirmed by indirectimmunofluorescence and electron microscopy. Infection of endometrial tissue andperitoneum was also recognized. Data suggested that the immunosuppressionmediated by cyclophosphamide resulted in a prolonged genital tract infection andconcomitant ascending infection leading to salpingitis.

In recent years Chlamydia trachomatis hasbeen well established as a significant cause ofsexually transmitted diseases (4, 13). In the fe-male although infection may be prolonged, it isoften asymptomatic and clinically rather benign.However, recent reports implicate this organismin salpingitis (6) and possibly peritonitis andperihepatitis (10). An animal model for the studyof chlamydial genital tract infections has beendeveloped employing the Chlamydia psittaciagent of guinea pig inclusion conjunctivitis(GPIC), which has been used for the study ofpathogenesis and immunity of chlamydial geni-tal tract infections (1, 5, 9). During the course ofinvestigation on the effect of immunosuppres-sion employing cyclophosphamide (Cy), alteredclinical response was observed in the externalgenitalia; at necropsy, spread of infection to thefallopian tubes was suggested (11a). Documen-tation of salpingitis and infection of the uterus(endometritis) was obtained by histopathology,immunofluorescence, and electron microscopy,and is the subject of this report.

MATERLALS AND METHODSMost of the procedures used in this study have been

described in detail recently (1) and in other publica-tions (8, 9).

Inoculation of guinea pigs. Mature Hartleystrain female guinea pigs were purchased from Simon-sen Laboratories, Gilroy, Calif. Animals were inocu-lated with a semipurified suspension of GPIC agent

(1). The 50% egg lethal dose was 5 x 1045/ml. Animalswere inoculated with 0.05 ml of freshly prepared sus-pension containing about 1.4 x 104 50% egg lethaldoses using a syringe fitted with a 1 inch (ca. 2.54 cm),23-gauge needle, the tip of which was blunted withsolder. For detection of infection, smears of vaginalscrapings were stained with Giemsa and examined forGPIC inclusions in cells (8).Animals studied. A summary of animals studied,

which were infected with GPIC and received differentregimens of Cy, is shown in Table 1. Some animalsdied during the course of an experiment and otherswere selectively killed. Controls consisted of groups ofinfected guinea pigs that did not receive Cy as well asother groups that received Cy but were not infected.

Histopathology. Tissues were collected at ne-cropsy from animals that had died or were killed underanesthesia at selected times. All tissues were fixed inbuffered Formalin before sectioning. In addition toconventional paraffin sections stained as describedpreviously (1), tissues from selected animals wereembedded in plastic by the method of Chi and Smuck-ler (2). Sections were cut at 2.0-jm thickness with aglass knife mounted on a JB-4 microtome (DupontSorvall, Newtown, Conn.).Immunofluorescence microscopy. Frozen sec-

tions were cut on a Cryo-cut microtome unit set at 3.0,um (American Optical, Buffalo, N.Y.). Sections andsmears of pus and cyst fluids were fixed with acetoneand stored at -70'C before staining by the indirectprocedure. Fluorescein-labeled rabbit antiserum toguinea pig gamma globulin was purchased from MilesLaboratories, Elkhart, Ind.

Electron microscopy. Fallopian tube material se-cured from buffered Formalin was soaked for 24 h in

728

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EXPERIMENTAL CHLAMYDIAL SALPINGITIS 729

a 2% glutaraldehyde-3.5% sucrose solution buffered topH 7.3 with 0.1 M cacodylate. Postfixation withbuffered OS04 and subsequent processing was accom-plished as described previously (1).

Cy. Cy, a gift of Mead-Johnson, Evansville, Ind.,was dissolved immediately in phosphate-buffered sa-line, pH 7.4, (PBS) at 30 mg/ml and injected intraper-itoneally, depending on the dose.

RESULTS

Effect of Cy on GPIC genital infection.The general effect of treatment with Cy, eitherby daily administration or on a 9-day intervalschedule (NDI; days 1, 9, 18, etc.), resulted in amarked prolongation of genital tract infection(hla). Histopathological examination of sectionsof cervical tissues obtained from Cy-treated in-fected animals revealed intense chlamydial in-fection of the superficial cells of the exocervix.Whole rows of infected cells could be identified(Fig. 1), while in lower layers scattered polymor-phonuclear leukocytes and occasional mononu-

clear cells were noted. Endocervical mucus-se-creting cells were apparently also involved, butinfection in this region was more difficult toevaluate because deposition of mucus was gran-ular in appearance and had microscopic similar-ity to chlamydiae.Infection ofthe uterus. A major finding was

extension ofinfection into the endometrial canal.In the canal there was occasional sloughing ofcells into the lumen along with a few polymor-phonuclear leukocytes. On careful inspection,chlamydiae could be identified in the endome-trial cells facilitated by staining with toluidineblue (pH 4.0). Organisms were detected in theepithelial lining near the ostium of the gland(Fig. 2A). Further evidence for infection of en-dometrium was obtained by immunofluores-cence (Fig. 2B). However, obvious infection ofthe uterus was not a frequent finding, even whenthere was evidence for salpingitis, as describedbelow.

Infection of fallopian tubes (salpingitis).

_. A,

.. _* ,* w i

.Ov"1a>. la

*.e

.,0 -4

lk

- or_ ) l v - ,_ ' -

FIG. 1. Superficial epithelial cells ofguinea pig cervix containing chlamydiae; some sloughed infected cellsare also present (experiment 104, GP 0058, 25 mg of Cy daily, died day 30). Stained with toluidine blue,pH 4.0.x240. Inset, x600.

VOL. 26, 1979

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

730 WHITE ET AL.

'5.I

.a

I- %

* N.4 4i6-*

A

'-'6

""X

A; it l4 o

FIG. 2. Section of uterus of guinea pig infected with GPIC. (A) Chlamydiae (T) are present in epitheliallining cell near ostium ofgland (t) (experiment 104, GP 0058, 25 mg of Cy daily, died day 30). Stained withtoluidine blue, pH 4.0. x240. (B) Section of uterus of guinea pig infected with GPIC stained by indirectimmunofluorescence (experiment 103, GP 0031, 250 mg of Cy NDI, killed day 30). x400.

Of a total of 41 infected animals treated with Cyexamined by gross inspection at necropsy (Table1), indication of involvement of the fallopiantubes was obtained for 25 (60%). This consistedof unilateral or bilateral enlargement of the fal-lopian tubes with prominent serosal vasculari-zation. On section the most frequent pathologi-cal observation was the presence of slightly di-lated cystic structures which contained pus or

markedly dilated salpinges which containedclear fluid. The fallopian tubes of 18 of these 25animals were examined histologically (Table 2).Acute inflammation was found in 4 of 7 animalsreceiving Cy daily and in 6 of 11 receiving CyNDI (Fig. 3). Thus, microscopically, a diagnosisof acute salpingitis was confirmed in 10 of the 18animals studied. Furthermore, GPIC was de-tected in the luminal epithelial lining in 6 ofthese 10 animals (Fig. 4). Confirmatory evidencefor presence of GPIC was also obtained by im-munofluorescence. Chlamydiae were demon-strated both in the lining epithelial cells as well

TABLE 1. Summary offemale guinea pigs studiedfor chlamydial salpingitis

GPIC Treatment Dosea No. of No.b ex-infected schedule doses/expt amined

+ Daily 25 13;13;27;36 14+ NDI 100 6 4

150 4;5;6;7 13250 3;3;4 10

+ -c - - 17d Daily 25 13;51 6- NDI 150 6 3- 250 3 2

a Milligrams of Cy per kg.Gross pathology at necropsy.c, Cy not given.

d Inoculated intravaginally with normal yolk sacsuspension.

as in the luminal pus (Fig. 5). Some of theacutely inflamed tubes were moderately dilated,while hydrosalpinx formation with flattened ep-ithelium in the absence of inflammation was

INFECT. IMMUN.

kJ. vv i.% ",

ieV %

i

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

TABLE 2. Histopathological confirmation of salpingitis in guinea pigs infected in genital tract with GPICand treated with Cy

Histopathology

Guinea pig Cy treatment Day killed (K) Inflam-Expt no. Guneapi schedule

o id()mtrno. (dose, mg) or died (D) react GPIC Cyst

tionsa102 3186 Daily (25) 29(D) - - -103 0029 Daily (25) 35(D) - - +103 0048 Daily (25) 38(K) _ _ +103 0028" Daily (25) 51(K) + - +103 0027" Daily (25) 18(K) + +104 0058b Daily (25) 30(D) ++c ++ +104 0069 Daily (25) 37(D) + +104 0053 NDI (100) 78(K) - - +104 0054 NDI (100) 78(K) - - ++104 0050 NDI (150) 57(D) - - +105 0076 NDI (150) 50(D) ++ + +105 0077 NDI (150) 29(K) ++ + +101 0016 NDI (250) 30(K) - - -101 0018 NDI (250) 26(D) + + -101 0019 NDI (250) 36(D) + - +101 0020 NDI (250) 29(D) + - -103 0031b NDI (250) 22(K) + - -103 0032 NDI (250) 22(K) - - -

a Inflammation in wall, pyosalpinx, or both.b Tubo-ovarian abscess.'Marked or striking.

ft.~~~~~~~~~~~~~'4)i 0.1N

dh O~~~~

'1~~~ *....pO.*~~~~-It ~~~~''J

4'.a~~~~~~~*Vo*.,~~~At. ~ ~ ~ ~.%**M

FIG. 3. Section of fallopian tube of guinea pig infected with GPIC (experiment 101, GP 0020, 250 mg of CyNDI, died day 29). Pyosalpinx, lumen filled with polymorphonuclear leukocytes and some histiocytes. Thewall is infiltrated with polymorphonuclear leukocytes (t). and mononuclear cells are present deeper in thewall. Stained u'ith hematoxylin and eosin. x60.

731

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

732 WHITE ET AL.

- N. aI 10i~i* ~~~~11w 0

.

FIG. 4. Section of fallopian tube of guinea pig infected with GPIC (experiment 104, GP 0058, 25 mg of Cydaily, died day 30). Chlamydiae (T) are present in luminal columnar cells. Stained with toluidine blue, pH 4.0.x600.

observed in a total of five animals (Fig. 6). Elec-tron microscopy confirmed the presence of chla-mydiae in epithelial cells of the fallopian tubetissue (Fig. 7).Of additional interest was the fact that in-

volvement of the serosa was found in four ani-mals manifesting salpingitis, and in two(GP0058, GP0069) GPIC was found in the peri-toneal mesothelial cells. Furthermore, four ani-mals with salpingitis also had tubo-ovarian ab-scesses (Table 2).The possibility of disturbance of normal bac-

terial flora and invasion ofnormally sterile tissuewas considered in view of the heavy doses of Cythat were being administered in these experi-ments. Vaginal swabs yielded a variety of gram-negative and gram-positive organisms. However,there was no difference qualitatively or quanti-tatively between Cy-treated and untreated ani-mals (experiment 101). Blood cultures were neg-ative, and, at necropsy, scrapings from the lu-minal aspect of tissue were negative upon culture(experiment 103). Cultures of uterus and cystfluid collected at necropsy were also negative. Ofimportance was the fact that bacterial infectionwas not detected in any of the histological sec-tions of the uterus or fallopian tubes examined.

In view of the histopathological findings de-scribed, sections of genital tract tissues and otherorgans were extensively examined from unin-fected animals receiving Cy only. No apparenthistopathological lesions were seen in any of thetissues examined, including cervix, uterus, fallo-pian tube, ovary, lung, liver, and kidney. Despiteprolonged administration of Cy, the histologicalfeatures of these tissues were essentially normal.

DISCUSSIONInfection of the female guinea pig genital tract

with GPIC agent has been documented to be arather self-limiting manifestation (1, 8). The ma-jor target site has been shown to be the squa-mous epithelium of the exocervix and thesquamocolumnar junction (1). Infection is usu-ally resolved after approximately 20 days withthe absence of detection of GPIC inclusions invaginal smears and presentation of an essentiallynormal histological picture. It should be empha-sized that no involvement of uterus or fallopiantubes had ever been observed in our previousexperiments. In the presence of immunosuppres-sion, however, ascending genital tract infectionculminating in salpingitis was observed. Thissalpingitis varied from an early acute inflam-

INFECT. IMMUN.

A&-

.,ior,

4.'

.... AE,

.'a,.71016-1, -I.

I

1

%,.. 'o

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

FIG. 5. Fallopian tube of guinea pig infected with GPIC stained by indirect immunofluorescence. (A)Section offallopian tube ofguinea pig infected with GPIC (experiment 103, GP 0027, 25 mg of Cy daily, killedday 19). xl,000. (B) Smear ofpus from fallopian tube cyst (experiment 105, GP 0076, 150 mg of Cy NDI, diedday 50). xI000.

733

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

734 WHITE ET AL.

I

i_..

-7o

9.

i

IsII if .

i { a

I /

FIG. 6. Section offallopian tube and ovary ofguinea pig infected with GPIC (experiment 104, GP 0054, 100mg of Cy NDI, killed on day 78). Fallopian tube is markedly dilated, and a cyst is present in the ovary.Stained with hematoxylin and eosin. x3.

FIG. 7. Electron micrograph of fallopian tube epithelial cell (experiment 105, GP 0077, 150 mg of Cy NDI,killed day 29) showing typical stages in the life cycle of the GPIC organism. x34,000.

matory reaction in the wall of the fallopian tube pyosalpinx and hydrosalpinx parallels observa-with presence of chlamydiae in the mucosal tions in women with gonorrheal salpingitis. Inepithelium progressing to a pyosalpinx and even- the latter, it has been stated that hydrosalpinxtual hydrosalpinx formation. The presence of represents the end stage or resolution of pyo-

INFECT. IMMUN.

--a.76Z. _

Ir I...f.1.

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EXPERIMENTAL CHLAMYDIAL SALPINGITIS 735

salpinx (11). The argument for ascending infec-tion was supported by detection of GPIC in theendometrial epithelium even extending beyondthe confines of the fallopian tube to its serosa(peritoneum).Attention has been drawn to salpingitis as a

potential complication of human chlamydialgenital tract infection (6, 14). In these reports,evidence for C. trachomatis as the etiologicalagent was obtained by culture at laparoscopyand by antibody studies. However, direct dem-onstration of C. trachomatis in the mucosal cellsof fallopian tubes by histological means was notobtained. Ripa et al. (12) have reported on ex-perimental induction of acute salpingitis ingrivet monkeys inoculated with C. trachomatis.Two animals were inoculated directly by inject-ing C. trachomatis in the lumen through thefallopian tube wall and one into the uterinecavity via the cervical canal. Although C. tra-chomatis was isolated and serological evidencewas obtained for infection, no histological dem-onstration of C. trachomatis was provided tosupport the claim of chlamydial salpingitis. Theresults reported here using the guinea pig modeland GPIC clearly show that chlamydiae canindeed infect fallopian tube tissue. The devel-opment of salpingitis in the guinea pig was de-pendent experimentally upon immunosuppres-sion (11a). These data should stimulate consid-eration of human salpingitis in the light of adisturbed or inefficient immune response.

ACKNOWLEDGMENTSWe thank Laura Cloud, Claudia McRaven, Danny Smith,

and Robin Storey for excellent technical assistance.This investigation was supported by Public Health Service

grant AI 13069 from the National Institute of Allergy andInfectious Diseases.

LITERATURE CITED1. Barron, A. L., H. J. White, R. G. Rank, and B. L.

Soloff. 1979. Target tissues associated with genital in-fection of female guinea pigs by the chlamydial agent ofguinea pig inclusion conjunctivitis. J. Infect. Dis. 139:

60-68.2. Chi, E. Y., and E. A. Smuckler. 1976. A rapid method

for processing liver biopsy specimens for 2 micron sec-tioning. Arch. Pathol. Lab. Med. 100:457462.

3. Hanna, L., L Schmidt, M. Sharp, D. P. Stites, and E.Jawetz. 1979. Human cell-mediated immune responsesto chlamydial antigens. Infect. Immun. 23:412-417.

4. Hobson, D., and K. K. Holmes. 1977. Nongonococcalurethritis and related infections. American Society forMicrobiology, Washington, D.C.

5. Lamont, H. C., D. Z. Semine, C. Leveille, and R. LNichols. 1978. Immunity to vaginal reinfection in fe-male guinea pigs infected sexually with Chlamydia ofguinea pig inclusion conjunctivitis. J. Infect. Dis. 19:807-813.

6. Mardh, P.-A., T. Ripa, L Svensson, and L Westrom.1977. Chlamydia trachomatis infection in patients withacute salpingitis. N. Engl. J. Med. 296:1377-1379.

7. Modabber, F., S. E. Bear, and J. Cerny. 1976. Theeffect of cyclophosphamide on the recovery from a localchlamydial infection. Guinea pig inclusion conjunctivi-tis (GPIC). Immunology 30:929-933.

8. Mount, D. T., P. E. Bigazzi, and A. L Barron. 1972.Infection of genital tract and transmission of ocularinfection to newborns by the agent of guinea pig inclu-sion conjunctivitis. Infect. Immun. 5:921-926.

9. Mount, D. T., P. E. Bigazzi, and A. L Barron. 1973.Experimental genital infection of male guinea pigs withthe agent of guinea pig inclusion conjunctivitis andtransmission to females. Infect. Immun. 8:925-930.

10. Miller-Schoop, J. W., S. P. Wang, J. Munzinger, H.U. Schlapfer, M. Knoblauch, and R. Wammann.1978. Chlamydia trachomatis as possible cause of per-itonitis and perihepatitis in young women. Br. Med. J.1:1022-1024.

11. Novak, E. R., G. S. Jones, and H. W. Jones, Jr. 1975.Pelvic infections, p. 414. In E. R. Novak (ed.), Textbookof gynecology, 9th ed. The Williams & Wilkins Co.,Baltimore.

lla.Rank, R. J., H. J. White, and A. L Barron. 1979.Humoral immunity in the resolution of genital infectionin female guinea pigs infected with the agent of guineapig inclusion conjunctivitis. 26:573-580.

12. Ripa, K. T., B. R. Moller, P.-A. Mardh, E. A. Freundt,and F. Melsen. 1979. Experimental acute salpingitis ingrivet monkeys provoked by Chlamydia trachomatis.Acta Pathol. Microbiol. Scand. 87:65-70.

13. Schachter, J., and C. R. Dawson. 1978. Human chla-mydial infections. PSG Publishing Co., Inc., Littleton,Mass.

14. Treharne, J. D., K. T. Ripa, P.-A. Mardh, L. Svensson,L. Westrom, and S. Darougar. 1979. Antibodies toChlamydia trachomatis in acute salpingitis. Br. J. Ven.Dis. 55:26-29.

VOL. 26, 1979

on May 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from