New rat to mouse xenograft transplantation of endometrium as a model of human endometriosis

Abstract Background Endometriosis can lead to infertility. Since there is no definitive treatment for endometriosis, animal modelling seems necessary to examine the possible treatments. Mouse endometrium cannot be separated for endometriosis induction. In addition, transplantation of uterus into the abdominal viscera to induce endometriosis causes organ damage. In this study, we defined a new model of endometriosis leading to separability of endometrium and a safe anatomical region for transplantation. Methods Forty female mice were allocated to 5 groups: 1, sham; 2, allograft uterus transplantation of mice to anterior abdominal wall of mice; 3, allograft uterus transplantation of mice to mesentery of mice; 4, xenograft endometrial transplantation of rat to anterior abdominal wall of mice; 5, xenograft endometrial transplantation of rat to mesentery of mice. Adult female rats with a previous pregnancy experience were selected and placed in the vicinity of male rats for 2 weeks to induce estrogen secretion and increase endometrial thickness. Results In the 4th group of animals, compared to sham, the peritoneal concentrations of VEGF‐A, TNF‐α, NO, MDA, and serum levels of CA‐125 and IL‐37 were increased and total body weight was decreased, while weight and size of endometrial lesions were increased significantly (P < .05). Genes expression of HOXA10 and HOXA11 were decreased significantly (P < .05) in groups 2 and 4 compared to sham. Conclusions Xenograft transplantation of endometrium from rat to anterior abdominal wall of mice can potentially mimic human endometriosis morphologically, histologically, and genetically.


| INTRODUC TI ON
Endometrium is the inner epithelial layer of mammalian uterus and is divided into thin basal and thick functional layers. The presence of a functional layer strictly depends on the menstrual (in humans) or estrous (in rodents) cycles.
Endometriosis is a pathological condition in which the endometrial cells grow outside the uterus in abdominal or pelvic cavities. These cells reside on the peritoneum layer located on organs (such as ovaries, fallopian tubes, uterus, and intestine), attract adjacent blood vessels (angiogenesis), and grow by triggering inflammation. These pathological conditions lead to formation of new endometrial lesions. Thus, angiogenesis and inflammation are two crucial biological phenomena in endometriosis. Complications of this disease include formation of scar tissue, organ adhesions, pus-filled cysts, and adhesion-related obstruction (in intestines or uterus lumina). In addition, following scar formation, anatomical disposition, and uterus obstruction, the occurrence of female infertility is probable, and is considered an end-stage outcome of endometriosis. 1 The most common theory of human endometriosis is that endometrial cells follow a retrograde route from the intrauterine cavity to the peritoneal fossa. Although most studies in the field of infertility and reproduction are performed in mouse models rather than other animals, the presence of intrinsic endometriosis in mice is impossible because there are no menstrual cycles and endometrium abscission in rodents and these integrated cells are absorbed by adjacent cells. Thus, in mice induction of endometriosis via a surgical procedure is necessary to assess aspects of human endometriosis. 2 The most common animal models for endometriosis are rodents, monkeys, and rabbits. 3 In 1985 Vernon introduced a rat model of endometriosis through transplantation of uterus fragments to the peritoneal layer, 4 and in 1995 Cummings achieved a mouse model of endometriosis through transplantation of uterus to the mesentery layer. 5 In these models uterus fragments were used for endometriosis induction, but based on the histopathology of endometriosis, this pathological condition is caused by the attachment of endometrial cells to the peritoneal layer. Thus, it seems that isolation of the endometrium from the two other layers of the uterus (myometrium and perimetrium) can probably model human endometriosis conditions more closely. 6 Endometrium in mice is a thin microscopic layer that is often impossible to dissect out, and besides the endometrium is a fragile tissue which cannot be sutured. However, the endometrial layer in the proliferative phase of of the estrous cycle is thick enough to be dissected and transplanted. In addition, transplantation of endometrium to vital organs such as mesentery, as a routine protocol of endometriosis induction, can potentially damage vessels, leading to internal hemorrhage and animal death. These difficulties make endometriosis induction a complicated process in mice.
Thus developing a new animal endometriosis model and finding a suitable anatomical site for endometrial implantation that completely mimics human endometriosis conditions, with fewer complications, seems necessary. In the present study, we compared three models of animal endometriosis (rat-rat and mouse-mouse allografts, and rat-mouse xenografts) with transplants to the anterior abdominal wall and mesentery layer. In this experimental animal study, we assessed the recipient animals using various biomarkers, including rate of angiogenesis (serum levels of VEGF-A and vessels count), serum levels of CA-125, inflammatory biomarkers (TNFα in peritoneal fluid and IL-37 in blood serum), oxidative stress biomarkers (NO and MDA in peritoneal fluid) and histological assessments to ensure that the new model accurately imitated human endometriosis.

| Study animal groups
Forty female mice were divided into 5 experimental groups (n = 8 in each group): 1, sham (application of laparotomy followed by abdominal wall suture with no experimental treatments); 2, mouse-mouse allograft -mouse uterus transplantation to mouse anterior abdominal wall; 3, mouse-mouse allograft -mouse uterus transplantation to mouse mesentery; 4, rat-mouse xenograft -rat endometrial transplantation to mouse anterior abdominal wall; 5, rat-mouse xenograft -rat endometrial transplantation to mouse mesentery.
In all treatment groups, the laparotomy was applied in the ventral midline region of the animals. Due to the thinness and inseparability of mouse endometrium, all layers of mouse uterus (endometrium, myometrium, and perimetrium) were cut and transplanted as allografts. While in rats, the endometrium was separated from total uterus tissue and transplanted to recipient mice in a xenograft procedure. Two anatomical transplantation locations were used: anterior abdominal wall and mesentery of small intestine. In all sites of transplantation, the uterine epithelium was sutured directly to the peritoneal layer.

| Preparation of donor animals
To increase the endometrium thickness in donor rats and facilitate endometrial dissection, a principle modification in this method, three main conditions were established: all female rats or mice were of adult age (to increase their uterine response sensitivity to proliferation following estrogen secretion), all animals had a previous pregnancy experience (to produce separatable endometrium from other uterine layers in the xenograft procedure), and the animals were exposed to male animals for 2 weeks (a ratio of one male/ two females for sexual stimulation). Animals were housed in special plastic cages divided by a metal grid to prevent mating. This indirect contact caused visual sexual stimulation, leading to estrogen secretion in donor female rats or mice. Also, the animals' straw was not replaced with a new bedding while they were in the cages, which resulted in pheromone secretion and urine odor. These special conditions caused sexual stimulation in female rats or mice leading to estrogen secretion and endometrial thickening. 7

| Hormonal and estrous cycle synchronization in recipient animals
According to the Whitten effect and based on the estrogendependent nature of endometriosis in rodents, the female recipient mice were exposed to males (indirectly through metal nested cages).
The female mice were housed in the vicinity of males in special cages divided by metal grids before (2 weeks) and during (4 weeks) endometrial induction. As with donor animals, these conditions caused the estrogen secretion necessary for successful implantation of endometrial fragments. Also, for estrus cycle synchronization, vaginal cytology was checked daily for a week before the surgery. A cotton swab impregnated with normal saline was used to collect a vaginal smear from the vaginal orifice. The samples were stained using the Papa-Nicola staining method, and endometriosis induction was performed during the estrus stage for all animals. The mice were sexually receptive in estrus cycle as characterized by estrogen secretion.
In this study, the estrus cycle was detected by the presence of whole superficial cells of different types during microscopic examination of vaginal cytology. 8

| Implantation of endometrium to abdominal wall and mesentery
Donor animals (mice and rats) were anesthetized through intraperitoneal injection of 25 IU ketamine-xylazine/25g animal (10 IU ketamine/90 IU xylazine) and then they were euthanized through cervical dislocation. Uterus (of mouse) and endometrium (of rats) were dissected, and all surrounding attached connective tissues were removed under a loop microscope ( Figure 1A). Dissected tissues were placed in DMEMF12/FBS 5% cell culture solution to preserve cell viability ( Figure 1B). Round grafts were prepared using a 3-mm diameter punch and were sutured (nylon, 5-0 USP, F I G U R E 1 Rat-mouse xenograft transplantation of endometrial segments (A and B) to abdomonal wall (C) and small intestine mesentery (D) for endometriosis induction (E). Arrows represent intestine (red), uterus horn (black), abdominal wall (green), angiogenesis (blue), mesentery artey (purple), endometrial lesions (yellow), and endometrial tissue attachments (white) SUPA medical devices, Iran) to the anterior abdominal wall or mesentery layer of small intestine in recipient animals ( Figure 1C,D).
Recipient mice were weighed and anaesthetized with 25 IU ketamine-xylazine/25 g animal. The small intestine was explored, and the inner epithelial surface (containing endometrium) of grafts was sutured in direct contact with the peritoneal surface of mesentery or abdominal wall. Also, for preparation of blood supply, the grafts were transplanted as close as possible to mesentery vessels.
Peritoneal and muscular layers of abdominal wall were sutured using absorbable threads (chromic 5-0 USP, SUPA medical devices, Iran), and the skin was closed with nylon (Nylon, 5-0 USP, SUPA medical devices, Iran). A day after recovery from the surgery, the recipient animals were exposed to males (for 4 weeks) in separate metal cages (to prevent mating) for endometriosis induction ( Figure 1). Prior to tissue implantation, the size and weight of whole grafts were recorded. 9

| Animal weighing, dissection, and tissue sampling
Four weeks after surgery, the recipient mice were anesthetized and then were euthanized by cervical dislocation procedure.
Immediately, 1 mL of injectable distilled water was injected into the peritoneal cavity. Two minutes later, the peritoneal fluid was aspirated. After laparotomy, the endometrial lesions and uterus were dissected. Thoracotomy was also performed, and blood was aspirated from the right ventricle and centrifuged (3000 g, 15 min) to separated blood serum. All biological samples were frozen in liquid nitrogen for future biochemical and genetic analysis or were fixed in 10% formaldehyde for histopathological assessments. Total body weight was also recorded.

| Morphometric assessments of endometrial lesions
During the tissue sampling procedure, the diameter of endometrial lesions was calibrated (Calliper, Sana't Co., Iran), and the weight of the grafts was calculated (Laboratory scale, Model GR 202, Tajhizat Co., Iran) after complete excision of surrounding connective tissues. The presence or absence of pus-filled cysts was also recorded. These factors were considered as growth markers of endometrial lesions, and they were compared with the primary size (3 mm) and primary weight (0.01 g) of endometrial lesions exactly prior to implantation.

| Assessment of angiogenesis in endometrial lesions and peritoneal fluid
To evaluate angiogenesis rate in endometrial lesions, the number of newly generated vessels around endometrial lesions were counted using histopathological sections (Nicon biology microscope, E200). Also, the peritoneal concentration of VEGF-A, a peritoneal angiogenic biomarker, was assessed biochemically using an ELISA kit (Abcam, ab100662, USA) according to routine procedure based on the manufacturer's instruction. 10

| Assessment of inflammation status in peritoneal fluid and blood serum
The concentration of TNFα (Abcam, ab193687, USA) in peritoneal fluid was considered a marker of inflammation status induced by endometrial lesions and produced by peritoneal macrophages. IL-37 (Abcam, ab213798, USA) was also measured as a serum biomarker for endometriosis diagnosis. These measurements were done using an ELISA kit according to the manufacture's protocol. 10

| Status of oxidative stress in peritoneal fluid following endometriosis induction
Generated oxidative stress following hyper-activation of macrophages and high rate proliferation of endometrial cells was measured in peritoneal fluid using an ELISA kit. In this process, NO concentration (Abcam, ab272517, USA) was evaluated using the Griess test, and MDA (Abcam, ab238537, USA) levels (representing lipid peroxidation status) were also measured. 10

| Serum levels of CA-125
Cancer antigen 125 is a member of the mucin family of glycoproteins. CA-125 is a biomarker which is elevated in the blood of some types of cancers and endometriosis. This factor was measured using an ELISA kit (Abcam, ab108653, USA). 10

| Histopathological assessments using H&E and Perls staining
The right horn of uterus and half of endometriosis lesions were fixed in 10% formaldehyde for H&E and Perls staining. H&E staining was used for glandular and stromal assessments of endometriosis, and Perls staining was also used to assess hemosiderin deposition in macrophages as an accepted factor for confirming endometriosis. Tissue processing was performed, and paraffin blocks were prepared. Thin sections (5 µm) were cut (Microtome, Leica RM 2125, Germany) and stained using H&E and Perls staining. Finally, the slides were assessed using a research microscope (Olympus, BX-51T-32E01) based on histopathological variations in the tissue including epithelium and stroma of endometrium, endometrial glands, blood vessels, and macrophages loaded with hemosiderin. 10

| Statistical analysis
After data extraction, the Kolmogorov-Smirnov test was first conducted to confirm data compliance with normal distribution. One-way analysis of variance (one-way ANOVA) was used for statistical analysis, and the Tukey post hoc test was used to determine the difference between the groups. Statistical Package for the Social Sciences 16 (SPSS Inc, Chicago, IL) was used for data analysis, the results were expressed as mean ± SD, and P < .05 was considered significant.

| Angiogenesis rate
The angiogenesis rate was assessed by measuring the peritoneal

| Serum concentration of CA-125
CA-125 biomarker, which represented the cancer antigen in blood serum, was significantly (P < .05) elevated in all allograft and xenograft transplantations in comparison with sham animals, while no significant (P > .05) difference was found between the rat-mouse xenograft transplantation groups and the allograft groups (Table 1).  (Table 1).

| Oxidative stress status
The concentration of all oxidative stress markers (NO and MDA) was increased significantly (P < .05) in whole transplanted animals (both allografts and xenograft treatment groups) compared to the sham group. No significant (P > .05) alteration was found among the animals in xenograft groups compared to the allograft groups (Table 1). in rat-mouse xenograft transplantation to anterior abdominal wall compared to allografts (Table 1).

| Histopathological variations
As depicted in Figure 2, the endometrial grafts of rats (with dissected layers of myometrium and perimetrium) were stained using H&E prior to implantation. Obvious epithelium (purple arrow) with compact stromal tissue (green rectangular) was seen located above lamina properia layer (Figure 2A). Also, stroma had coronal sections of endometrial glands. After implantation of mouse uterus to abdominal wall of recipient mice ( Figure 2B), pathological features were found including: epithelial layer ( Figure 2B, purple arrow) with no villi, stroma ( Figure 2B, green rectangular) seen as a thin layer with multiple cavities, and fewer of coronal sections of blood vessels. Also, as shown in Figure 2B, the thick layer of myometrium and perimetrium degenerated following uterus transplantation (orange arrow). These histopathological changes led to less secretion of pus in the luminal space of endometrium, decreased size of grafts, and reduced weight of lesions after a 4-week implantation. Figure 2C shows  Figure 2F represents xenograft transplantation of endometrium to mesentery layer. In these histological sections, many glandular lumina (black circle) were found with no pus in the cavity (yellow star).

| D ISCUSS I ON
In this experimental study, we present a xenograft (rat to mouse) model of endometriosis through implantation of endometrium of rat to anterior abdominal wall of mouse. In the study, estrogen production was stimulated, leading to increased endometrium thickness, through exposure of adult female rats (with previous pregnancy experience) to males (for 2 weeks). The recipient mice were also stimulated (using exposure to males) to produce estrogen to enable successful transplantation of endometrial grafts. Finally, we found that due to the estrous cycle and induction of estrogen production in donor rats, it is possible to increase the thickness of the rat endometrium, which can then be dissected easily. Implantation of these This study aimed to determine the animal model most similar to human endometriosis with regard to molecular, histological, and genetic characteristics. The main modifications to the protocol for using rats as graft donor to accelerate the success rate of xenigraft implantation were: previous pregnancy experience, and exposure to male animals for 2 weeks. The graft recipient mice were also exposed to males for 4 weeks. These essential conditions led to increased estrogen levels and greater thickness of endometrium for effective dissection and successful endometrial implantation. Endometrial lesions isolated from stimulated uterus of a rat implanted in the anterior abdominal wall of mice in a xenograft surgical process could potentially be an appropriate option for endometriosis modelling in rodents. For future studies, additional laboratory assays such as immunoassay and protein immunoblotting are strongly recommended.
Also, inclusion of a human endometriosis sample to compare with rodent endometriosis lesions will be necessary to upgrade the level of study.

ACK N OWLED G EM ENTS
We are grateful to the Research Council of Kermanshah University of Medical Sciences for their financial support (grant number 4000077).

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no competing interests.

AUTH O R S' CO NTR I B UTI O N S
M.B A.A conceived and designed the study, supervised the data collection, interpreted the results, and revised the manuscript. A.A conducted the experimental procedures, data collection analysis, and manuscript preparation. C.J was a scientific advisor for conducting laboratory analysis. K.M was a scientific advisor in the filed of angiogenesis pathways and animal disease modelling. All authors read and approved the final manuscript prior to submission.

CO N S E NT FO R PU B LI C ATI O N
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets used and analyzed for this are available from the corresponding author upon reasonable request.