American Journal of Obstetrics and Gynecology Founded in 1920

volume 149 number 6 .JULY 15, 1984

GYNECOLOGY

Transplantation of normal and ectopic human endometrial

tissue into athymic nude mice

Nezaam M. Zamah, M.D., F.R.C.S.(C),* Melvin G. Dodson, M.D., Ph.D.,

L. Clifton Stephens, D.V.M., Ph.D., Veasy C. Buttram, Jr., M.D., Paige K. Besch, Ph.D.,

and Raymond H. Kaufman, M.D.

Houston, Texas

Implants or tiny circumscribed nodules of endometrial tissue were found in all female nude mice given

intraperitoneal injections of fragments of human normal (proliferative and secretory) or ectopic

(endometrioma) endometrium. Half of these animals received estrogen supplementation and the other half

received none. The endometriosis tissue present in these animals at 28 or 56 days after inoculation

consisted of glands and stroma with an infiltration of hemosiderin-laden macrophages. Glands in tissue

transplants of animals given supplemental estrogen tended to be larger, and the secretory endometrium

tended to revert to a proliferative pattern. Palpable nodules at the site of subcutaneous inoculations of

proliferative endometrium became undetectable grossly and microscopically within 24 to 32 days, whereas

endometrioma tissue remained detectable for up to 70 days and resembled the intraperitoneal tissue

microscopically. This study demonstrates that human endometrial tissue can be successfully transplanted

into the nude mouse and will retain its basic morphology. (AM. J. OssTET. GYNECOL. 149:591, 1984.)

In 1880, Breus first recognized endometriosis, and

16 years later Von Recklinghausen 1 first described en­

dometriosis as a pathologic entity. In the first quarter

of this century, several theories regarding the etiology

From the Department of Obstetrics and Gynecology, Baylor College of Medicine, the Reproductive Research Laboratory, St. Luke's Epis­copal Hospital, and the Section of Veterinary Pathology, M. D. Anderson Hospital Tumor Institute.

Supported in part by a grant from the Women's Fund of Houston, Texas.

Presented at the Thirty-first Annual Clinical Meeting of the American College of Obstetricians and Gynecologists, Atlanta, Georgia, May 7-12, 1983, and raeived the first-prize award for the best basic science research paper from The Upjohn Co., Kalamazoo, Michigan.

Received for publication December 14, 1983; accepted january 3, 1984.

Reprint requests: Paige K. Besch, Ph.D., Department of Obstetrics and Gynecology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030.

*Research completed during the tenure of a Postdoctoral Fellowship in Reproductive Endocrinology and Infertility.

of endometriosis were proposed. The three most

commonly accepted theories are those of retrograde menstruation and implantation, coelomic metaplasia,

and vascular dissemination. However, the last hundred

years have not fully clarified the pathogenesis of en­dometriosis. One of the limitations has been the diffi­

culty in doing controlled experiments in human

subjects. Endometriosis occurs naturally in the monkey, which

happens to be the only other animal that has cyclic

menstrual periods.2 Observations and experiments

with the use of monkeys and other species of animals

that do not naturally have the disease have concen­

trated on evaluation of the proposed etiologic theories.

Scott et al.~ demonstrated the development of endome­

triosis in five of 10 Macaca monkeys, but some animals required as long as 3 years before endometriosis was

noted. Merrili,4 using autologous rabbit endometrium

placed inside Millipore diffusion chambers, produced

growth of ectopic endometrial-like tissue in the adja-

591

592 Zamah et al. July 15, 1984 Am. J. Obstet. Gynecol.

Fig. l. A nodule from a nude mouse given intraperitoneal fragments of proliferative endometrium and supplemented with estradiol. The acini are lined by malaligned ciliated columnar epithelium and the lower right area is necrotic. (x250.)

cent subperitoneum and subcutis which was inter­preted to give support to the "coelomic metaplasia theory." A spontaneous case of widely disseminated endometriosis in a rhesus monkey was concluded to be compatible with spread by both lymphatic and hema­togenous routes." In 1980, Schenken and Asch6 devel­

oped an autologous animal model system similar to that of Scott et a!. with the use of the New Zealand White rabbit. They studied prostaglandin levels and fertility rates and showed that increased prostaglandin E2 levels correlated with the severity of endometriosis. These earlier model systems clearly demonstrated that ex­perimental endometriosis can be produced by direct menstruation into the peritoneal cavity or by direct in­jection, although success rates are surprisingly low. Even though Ridley and Edwards' used human men­strual effluents and human volunteers, their study was quite limited in terms of allowing experimentation and

manipulation. Ideally, an experimental animal model

system with the use of human endometrium or endo­metriosis tissue is needed. During the past 5 years

many experiments with a diversity of emphasis have

been done on normal endometrium and endometrio­sis. Workers have been successful in the cultivation of

endometrial epithelial and stromal cells in vitro.8 The

advantages of an in vivo model system for the study of human endometriosis to allow experimentation and manipulation are obvious.

Several workers, using the nude mouse system, have

serially passed human or animal tumors, 9·

10 main­tained their tumorigenicity and morphology, 11 and, in

the case of endometrial adenocarcinoma, maintained its responsiveness to hormones even with serial passage into these animals. 12 Grafts of benign tumors grow less

well than those of malignant tumors, but biopsy speci­mens of human vaginal diethylstilbestrol-associated

adenosis and synovial membranes from patients with rheumatoid arthritis have been successfully grafted into nude mice. 1

'L 14 Normal tissue or grafts can also be

transplanted into these mice but they generally do not proliferate.

Experimental evidence in rhesus monkeys has impli­cated cell-mediated immunity in contributing to a fa­milial predisposition to endometriosis. 15 The immune­deficient characteristics of the nude mouse have been well documented. 16 Reported here is an animal model system with the use of endometriosis or normal pro­

liferative or secretory endometrium injected into the

nude mouse.

Material and methods

Athymic mice. Twenty-nine 8-week-old female BALB/c nu/nu "nude" mice (Harlan Industries, Madi­

son, Wisconsin) were used. Five mice were housed per cage under laminar-flow HEPA filtered hoods in rooms

maintained at 26.7° C with a 12-hour light/12-hour dark cycle. All housing materials, as well as food and water, were autoclaved prior to use. The mice were fed

Volume 149 Number 6

Transplantation of endometrial tissue into athymic mice 593

Fig. 2. Nodule firmly attached to peritoneum and fat from a nude mouse given intraperitoneal fragments of secretory endometrium. Numerous spaces are surrounded by compact stroma. (X25.)

ad libitum with laboratory chow and acidified water.

Handling was done under laminar-flow hoods with

sterile equipment. The injections of human tissues were administered with tuberculin syringes through

18-gauge needles. Intraperitoneal injections were given

on the ventral midline just caudal to the umbilicus. The

subcutaneous inoculations were done on the dorsal

midline between the scapulae to restrict access from

self-inflicted trauma. The subcutaneous inoculation of

tissue fragments immediately produced a palpable nodule at the site. These sites were palpated twice weekly to determine if the nodules 1·egressed or re­mained unchanged. Approximately one half of the animals received 30 p.g/ kg of estradiol cypionate (The

Upjohn Co., Kalamazoo, Michigan) intramuscularly on the date of tissue inoculatio n and weekly thereafter.

A closed carbon dioxide chamber was used for eutha­

nasia. Tissue processing. Human endometrium (pro­

liferative and secretory) and endometriosis tissue were

collected in a sterile fashion from patients undergoing elective hysterectomy for various benign gynecologic

indications. The specimens were taken for pathologic

examination in cold, sterile Dulbecco's phosphate­

buffered saline. After adequate samples were taken fo r

routine histologic processing, po rtions of selected en­dometrial tissue were removed by gentle scraping of

the endometrium. Endometriosis tissue was obtained from endometriomas. The tissue was transported in a mixture of Dulbecco's phosphate-buffered saline, 200

Ulml of penicillin, 200 p.g/ ml of stre ptomycin, and 20

U/ ml o f Mycostatin on ice to the tissue culture labora­tory. Under a laminar-flow hood, the tissues were

chopped into 1 mm'1 fragments, was hed twice in

phosphate-buffered saline to remove red blood cells

and debris, and then divided into two portions: (l)

coarse fragments and (2) material for single-cell prep­

aration. The inoculum of fragments for intraperito­neal or subcutaneous injection consisted of 0.2 ml of

chopped tissues in 0 .2 ml of phosphate-buffered saline

(total volume, 0.4 ml) . T he single-cell suspension was prepared from

chopped fragments of proliferative or secretory endo­metrium placed in 15 ml conica l centrifuge tubes con­

taining 10 ml of tissue culture medium with enzymes added. The medium consisted of Ham's F-12 (Flow Labs, Inc., McLean, Virginia) that contained 2 mmoi/L

of L-glutamine, 1.0 ml/500 ml of phenol red, and di­

gestive enzymes consisting of 2.5 mg/ml of collagenase

in Dulbecco's phosphate-buffered saline without Ca++

and Mg++, 0.25% trypsi n, and 1 gm /500 ml of ethyl­enediaminetetraacetic acid. Fragments were incubated

at 37° C for 4 hours, then centrifuged, washed in

phosphate-buffered saline , counted in a hemacytome­

ter, and resuspended in phosphate-buffered saline at a

concentration of 2 x 107 viable cells per milliliter. A

volume of 0.4 ml of suspended cells was used for intra­peritoneal inoculation.

The proliferative endometrium was taken from two

patients. The date of the menstrual cycle was not

594 Zamah et al. July 15, 1984 Am. J. Obstet. Gynecol.

Fig. 3. Higher magnification of implanted secretory endometrium from a mouse supplemented with estradiol. Flattened columnar cells line acini and the stroma is fairly vascular and compact. (X250.)

noted, but histologically the specimens had the charac­

teristic morphologic features of the early proliferative

phase. The endometrial glands were straight and lined by malaligned , tall , columnar, nonvacuolated cells. Mi­

totic figures were numerous in both epithelial and stromal layers. All of the secretory endometrium (cycle

day 24) was taken from the uterus that was removed

from a 38-year-old patient. The endometrial glands

were dilated. The columnar epithelial cells were well

aligned and had vacuoles above and below the nuclei.

Mitotic figures were not found in the mucosal epithe­

lium or the underlying mildly edematous stroma. The endometriosis tissue that was given to mice intra perito­

neally was from the ovary removed from a 33-year-old patient. The serosa of the ovary had several 1 to 2 em

reddish brown foci. Sectioning through these foci re­

vealed that some had cystic centers filled with brown fluid. Microscopically, the endometriotic tissue con­

sisted mainly of fibrous stroma with hemorrhagic

areas and numerous hemosiderin-laden macrophages.

There were a few small glands that were lined by

disorganized cuboidal or low-columnar cells. The en­

dometriosis tissue that was administered to mice subcu­

taneously was from the serosa o f an ovary from a 15-

year-old patient. The serosa of this ovary was covered

by hemorrhagic fibrotic material that contained mul­

tiple cysts that were up to 4 em in diameter. This tissue

had microscopic features like those described for the

other endometriosis specimen . Study groups. Nude mice given intraperitoneal in-

jections of fragments of proliferative endometrium, se­cretory endometrium, or endometriosis tissue were put

to death at 28 or 56 days after injection . Sacrifice of the

mice given proliferative endometrium subcutaneously

was begun on day 24 whe n the originally palpable

nodules regressed to 1 mm in size. One mouse died on

day 28 and the remaining mice were killed on the

thirty-second day. Two of the mice that received subcu­taneous fragments of endometriosis tissue were put to

death after 56 days, while the others were held for 70

days to determine if palable nodules regressed. Both of

the mice given cell suspensions of proliferative endo­metrium intraperitoneally were put to death after 56 days. The ones given suspensions of secretory endome­

trium were put to death at either 28 or 56 days after

injection. Approximately one half of the animals that

were given tissue fragments received intramuscular in­

j ec tions of estradiol on the day of tissue inoculation and

weekly thereafter. All of the mice given cell suspen­

sions received the estradiol injections according to the

same schedule. At necropsy, the sites of subcutaneous inoculation

with or without palpable nodules and lesions in the

abdo men of intraperitoneally injected mice were har­

vested. These tissues and others, including caudal ab­

dominal fat, urogenital system, and abdominal wall,

were placed in neutral buffered 10% formalin. The

tissue was processed by standard methods, and par­

affin-embedded sections were cut at 3 J.Lm and stained

with hematoxylin and eosin.

Volume 149 Number 6

Results

Proliferative endometrium. Six nude m1ce were

given fragments of proliferative endometrium intra­

peritoneally. Two were put to death after 28 days and the remaining four, at 56 days after inoculation. At necropsy, five of the six nude mice injected had single or multiple discrete, tan nodules 1 to 3 mm in diameter and loosely attached to the omentum, mesentery, or caudal abdominal fat. A nodule was found in histologic

sections of the caudal abdominal fat from the sixth mouse. Microscopically, the loosely attached nodules

harvested at 28 or 56 days from five of the mice were similar. The periphery of the nodules had a thin mem­

branous capsule overlying densely cellular stroma that surrounded variable numbers of small acinar struc­

tures. The nodules from mice that received estradiol had more acini that were larger than those from the

ones that were unsupplemented, but the microscopic

features of the acini in both groups of mice were similar. These structures were lined by malaligned

columnar cells that commonly had a ciliated luminal border (Fig. 1). Small numbers of lymphocytes and

plasma cells infiltrated the stroma. The centers of nodules from estradiol-supplemented and unsup­

plemented mice were, in general, necrotic. The nodule found microscopically in an estradiol-supplemented mouse (C4) put to death at 56 days was embedded within the caudal abdominal fat and surrounded by

congested capillaries. This nodule was composed of several cystically dilated acini. Ciliated columnar epi­thelium lined these structures which were separated by thin fibrous trabeculae. This tissue contained no areas of necrosis.

The nodules produced by subcutaneous inoculation

of fragments of proliferative phase endometrium into

six mice began to regress within 3 to 4 weeks after inoculation. Tissues taken from the inoculation site 24 to 32 days after inoculation either had no microscopic lesions or contained only mild diffuse infiltrates of lymphoid cells and hemosiderin-laden macrophages.

No gross or microscopic lesions were found in the two mice put to death 28 days after being given an intraperitoneal cell suspension made from proliferative

endometrium. Secretory endometrium. Four mice were given in­

traperitoneal fragments of secretory endometrium.

Two were put to death at 28 days and the other two at

56 days after inoculation. All had discrete nodules 1 mm in diameter in the abdomen. In addition to loosely

attached nodules in fat, such as were seen in mice given

proliferative endometrium, nodules of tissue were firmly attached to caudal abdominal fat or the serosa of abdominal viscera (Fig. 2). One mouse had a nodule on

the serosa of the stomach, another had a nodule at­tached to the capsule of the spleen, and a third had a

Transplantation of endometrial tissue into athymic mice 595

Fig. 4. Endometriosis tissue implanted on the serosa of the cecum of a nude mouse. (Original magnification x25.)

nodule attached to the peritoneum of the abdominal

wall. Microscopically, all mice had nodules without ap­preciable necrosis. These nodules had capillaries at the

interface of attachment to fat or serosa. The loosely

attached nodules had necrotic centers. The stroma was compact and densely cellular rather than loosely ar­

ranged and edematous, as would be expected in the secretory phase endometrium. The columnar cells lin­

ing the acini were flatter and better aligned than in the transplanted proliferative endometrium and the cyto­plasm was sometimes vacuolated (Fig. 3). Lymphoid

cells infiltrated all of the nodules but these cells were present in greatest numbers in nodules harvested 56

days after inoculation. Considerable variation in acinus size was noted in both estradiol-supplemented and un­supplemented mice. However, the two mice given es­tradiol injections and put to death at 28 days had acini

that were at least twice as large as any in either unsup­

plemented mouse put to death at 56 days. The larger acinar size in the estradiol-supplemented mice might be a reflection of estrogen stimulation and/or the shorter duration of the transplant.

No gross or microscopic lesions were found in the four mice given intraperitoneal cell suspensions of se­cretory endometrium.

Endometriosis. Both mice that were given intraperi­toneal fragments of endometriosis tissue received es­

tradiol supplementation. The mouse that was put to death at 28 days had a 4 by 2 by 1 mm discrete tan­

mottled red nodule firmly attached to the serosa of the cecum (Fig. 4). The mouse put to death at 56 days had

a slightly larger (7 by 4 by 5 mm) nodule embedded in the caudal abdominal fat. Microscopically, the nodule

on the cecum was well vascularized. It contained no well-formed acini or glands but had scattered slitlike spaces and stroma that contained a mild infiltrate of lymphocytes and plasma cells and a few hemosiderin­laden macrophages. The nodule harvested at 56 days and found in the caudal abdominal fat had numerous

596 Zamah et al. July 15, 1984 Am.]. Obstet. Gynecol.

Fig. 5. An intraperitoneal implant of endometriosis tissue. The stroma surrounding acinar remnants contains a marked infiltrate of lymphoid cells and marcophages. (X 100.)

acini and spaces. The lining of the acini was highly

variable . It was either partially missing, necrotic, or

flattened. This nodule was well vascularized and had

prominent hyalinized stroma that was edematous in

some areas. There was marked infiltration by mono­

nuclear leukocytes with a predominance of hemosid­

erin-laden macrophages that formed several large focal

aggregations (Fig. 5). A few multinucleated giant cells

and neutrophils were seen.

Two mice given supplemental estradiol and subcuta­

neous fragments of endometriosis tissue were put to

death after 56 days. Palpable nodules were present

which in the microscopic sections had multinucleated

giant cells associated with necrotic acini. The nodules

from both mice contained large numbet·s of hemosid­

erin pigment-laden macrophages and focal accumula­

tions of extracellular pigment. The palpable nodules

taken from the three mice that were put to death at 70

days and were not given estradiol supplementation re­

sembled the ones taken from the preceding two mice.

The nodules were mainly stroma that contained extra­

cellular and intracellular hemosiderin. Giant cells were

associated with the remnants of epithelium lined acini,

and in one there were focal accumulations of neutro­

phils associated with these structures.

Comment

Human endometrial tissues (proliferative, secretory,

and endometriosis) were transplanted into the nude

mouse intraperitoneally and subcutaneously in an at-

tempt to observe the biologic behavior of such tissues in

this immune-deficient animal. We have demonstrated

that fragments of all three of these tissue types can be

successfully transplanted into the nude mouse and

survive for varying periods of time, depending on the

tissue type and route of injection. The success of the

intraperitoneal route of injection seemed to depend on

whether the tissue fragments were lodged in a suitable

area of the peritoneal cavity to permit survival. The

caudal abdominal fat pad, the most dependent portion

of the abdomen of the mouse, seems to be very recep­

tive to tissue transplantation. In the human, the cul­

de-sac is the most dependent portion of the peritoneal

cavity and is also a common site of endometriosis.

However, the serosa of the bowel, stomach, and spleen

also seemed receptive to transplantation of endome­

triosis or normal endometrial tissue. When attachment

to fat or serosal surfaces occurred, the tissue became

vascularized and remained viable for at least 56 days

while the normal anatomic relationship of glands to

stroma was maintained. Fragments of endometrioma

can also be successfully transplanted intraperitoneally

but, in addition, these grafts were maintained for up to

70 days in the subcutis whereas normal proliferative

endometrium regressed in the subcutis in a much

shorter period of time. For monitoring purposes the

subcutaneous route of injection of the tissue appears to

be preferred over the intraperitoneal route, since it

allows frequent evaluation of nodule size. However, the

subcutis is not the "normal" site for endometriosis, and

Volume 149 Number 6

so studies of tissue behavior may not be comparable to

the in vivo situation. The effect of estrogen supplementation is striking in

that the tissue showed better morphologic features in the estrogen-stimulated group compared to the non­estrogen-stimulated group with proliferative but not

with secretory endometrium. Mitotic activity also ap­peared to be more common in the estrogen-treated

group of animals. Single-cell intraperitoneal trans­

plants of proliferative endometrium did not survive in the presence or absence of estrogen stimulation.

It is well known that the nude mouse still has the

innate ability to mount a limited immune response be­cause of its intact natural "killer" T-cell function. It was

not surprising, therefore, to see a marked lymphocytic

and plasma cell infiltrate in some of the tissues recov­ered. Interestingly, endometriosis tissue survived for a much longer period of time than did normal prolifera­

tive endometrium when both were administered subcu­

taneously. This marked variation in survival time in the nude mouse raises some very important questions re­

garding the "normality" of the glands and stroma of endometriosis and suggests that ectopic endometrium is biologically different. Perhaps adaptation to an aber­rant site in the patient resulted in this tissue becoming more autonomous and, as a result, more susceptible to

transplantation. It is to be noted that none of the single-cell suspen­

sion preparations resulted in endometrial cellular pro­

liferation grossly or microscopically. Perhaps repeated injections of the same preparation over a prolonged

period of time are necessary before such changes can

occur. Further studies will be necessary to determine the

usefulness of the nude mouse system for the study of

endometriosis and possibly other benign gynecologic diseases.

We wish to thank David Fink, B.A., B.S., Research Assistant, Department of Obstetrics and Gynecology, Baylor College of Medicine, Charles Castro, B.A., De­nise Martin, B.A., Willie Virgil, and the staff of the M. D. Anderson Hospital and Tumor Institute Animal Research Facilities for their technical assistance.

Transplantation of endometrial tissue into athymic mice 597

REFERENCES

l. Von Recklinghausen F. Die Edenomyone und Cystadeno­myone der uterus und Tukenwandrug; ihre Akkunft von Restur des Wolff'scher Korpers Berlin, Berlin: Hirsch­wald, 1896.

2. MacKenzie WF, Casey HW. Animal model of human dis­ease: endometriosis in rhesus monkeys. Am J Pathol 1975;80:341.

3. Scott RB, TeLinde RW, Wharton LR Jr. Further studies on experimental endometriosis. AM J OBSTET GYNECOL 1953;66:5.

4. Merrill JA. Endometrial induction of endometriosis across Millipore filters. AM J 0BSTET GYNECOL 1966;94:780.

5. McClure HM, Graham CE, Guilloud NB. Widespread endometriosis in a rhesus monkey (Macaca mulatta). In: Hofer HO, ed. Proceedings of the Second International Congress of Primatology. New York: S. Karger AG, 1969:155. (vol 3).

6. Schenken RS, Asch RH. Surgical induction of endome­triosis in the rabbit: effects on fertility and concentrations of peritoneal fluid prostaglandins. Fertil Steril 1980;34:6.

7. Ridley JH, Edwards JK. Experimental endometriosis in the human. AM J 0BSTET GYNECOL 1958;76:4.

8. Varma VA, Adamac TA, Dorman BH, Kaufman DG. Human endometrial epithelial and stromal cells in cul­ture (Abstract). Lab Invest 1982;46:86A.

9. Rygaard J, Povlsen CO. Heterotransplantation of a human malignant tumor to "nude" mice. Acta Pathol Mi­crobial Scand 1960;77:758.

10. Fogh J, Fogh JM, Orfed T. One hundred and twenty­seven cultured human tumor cell lines producing tumors in nude mice. J Nat! Cancer Inst 1977;59:221.

11. Stiles CD, Desmond W, Chuman LM, Sato G, Saier MH. Growth control of heterologous tissue culture cells in the congenitally athymic nude mouse. Cancer Res 1976; 36:1353.

12. Satyaswaroop PG, Zaino RJ, Mortel R. Human endome­trial adenocarcinoma transplanted into nude mice: growth regulation by estradiol. Science 1983;219:58.

13. Pienkowski MM, Mann LC, Rosloniec EF, Welch CW. Accretion biopsy specimens of vaginal adenosis from pa­tients exposed in utero to diethylstilbesterol, when trans­planted to nude mice. JNCI 1979;62:521.

14. Brinckerhoff CE, Harris ED. Survival of rheumatoid synovium implanted into nude mice. Am J Pathol 1981; 103:411.

15. Dmowski WP, Steele RW, Baker GF. Deficient cellular immunity in endometriosis. AMj 0BSTET GYNECOL 1981; 141:377.

16. Wortis HH. Immunological studies of nude mice. Con­temp Top Immunobiol 1974;3:243.