This article describes a detailed protocol for producing a reliable and reproducible thin endometrium with a very low mortality rate and minimal intrauterine adhesions by injecting 95% ethanol into the mouse uterus within 1-3 min.

Thin endometrium (TE) has been widely recognized as a critical cause of infertility. However, the pathogenesis of TE remains unclear, and satisfactory treatment options are still urgently needed. Several animal models of TE have been developed, but the mouse model involving abdominal surgery and injection of 95% ethanol presents a formidable challenge due to the high mortality rate and risk of intrauterine adhesions if not performed correctly. Here, we describe a detailed protocol that produces reliable and reproducible TE with a very low mortality rate and minimal intrauterine adhesions by injecting 95% ethanol into the mouse uterus with varying infusion times. The results showed that all of the mice successfully developed TE with infusion times ranging from 1-3 min, characterized by a typical reduction in endometrial thickness and the number of glands, as well as excessive endometrial fibrosis. These findings suggest that this mouse model is suitable for studying thin endometrium and can serve as a platform for developing future TE treatments.

Thin endometrium (TE) is a serious condition in obstetrics and gynecology that often affects women of childbearing age. TE is diagnosed when the endometrium thickness measures less than 7 mm on an ultrasound scan, accompanied by a normal uterine cavity, and is closely associated with pregnancy failure^1,2. It is estimated that approximately 1.5%-9.1% of women undergoing in vitro fertilization (IVF) treatment will experience TE, making it a growing challenge in reproductive medicine^3,4,5. The most common causes of TE include improper endometrium repair following surgical separation of intrauterine adhesions (IUA) and curettage, which are often accompanied by disrupted blood vessel distribution and sparse glands^6,7,8. To date, the cellular and molecular mechanisms underlying TE remain unclear. Recovery of the endometrium in TE patients is time-consuming, although estrogen treatment and low-dose aspirin therapy have been explored as potential interventions⁹. Therefore, an in-depth study of the pathogenesis of TE is the most direct approach to addressing the challenges of this condition. However, studies on the pathogenesis of TE rely on animal models, making the selection of an appropriate model crucial.

Most histopathological studies of thin endometrium (TE) have shown that impaired proliferation of epithelial cells and macrophages, decreased expression of ovarian steroid hormone receptors, excessive deposition of extracellular matrix, and cellular senescence are the most significant pathological features of TE^6,9,10. Currently, various rat models have been developed to mimic TE, including models induced by scratching^11,12,13, ischemia¹⁴, thermal injury¹⁵, and chemical injury^16,17,18,19. A rat model is induced by scratching the endometrium with a needle or catheter, which often results in intrauterine adhesions (IUA) rather than TE^{11,12,20,21,22}. An ischemia-induced TE model in rats has also been reported, where endometrial ischemia is achieved by performing bilateral uterine artery ligation, leading to reduced endometrial thickness. However, this method is time-consuming, requiring three months, and is not widely used for research¹⁴. In the thermal injury-induced model, an artificial insemination tube is used to infuse 85 °C preheated water into one side of the uterine horn through the confluence of the two sides, making it more complex than other methods¹⁵. The chemical-induced model involves injecting 95% ethanol into the exposed uterine horn to damage the entire endometrium. This model offers the advantages of low cost and a short experimental period for studying the pathological mechanisms and treatments in TE, but it is primarily used in rats rather than mice^16,17,23. Although various animal models exist, especially rat models, each has its limitations. They can only simulate certain aspects of TE, and few mouse models closely replicate the disease characteristics of TE while also being convenient and versatile for research.

In this context, we developed a novel thin endometrium (TE) mouse model by infusing 95% ethanol into the uterine cavity using a time-gradient approach with a modified method²⁴ (Figure 1). The results showed that all the mice successfully developed TE when the infusion time ranged from 1-3 min, displaying typical characteristics such as reduced endometrial thickness, gland reduction, and increased endometrial fibrosis. These findings suggest that our mouse model is a suitable tool for studying TE and can serve as a platform for developing future TE treatments.

This study was approved by the Institutional Animal Care and Use Committee of Shenzhen Zhongshan Obstetrics & Gynecology Hospital (formerly Shenzhen Zhongshan Urology Hospital), where all animal experiments were conducted. Female C57BL/6 mice (age 6-8 weeks, weight 18-20 g) were used in this study. All animals were housed in a specific pathogen-free (SPF) environment in the same room and acclimatized for one week before the experiments in a room without specific pathogens at (22 ± 1) °C under a 12-h light/dark cycle. They were provided with free access to food and water. The details of the reagents and equipment used in this study are listed in the Table of Materials.

1. Identification of estrus

NOTE: The estrus phase in mice is comparable to the late luteal phase in humans, with a relatively thick endometrial lining at this time. Selecting mice in estrus for model induction ensures that they are in a relatively consistent physiological state. Additionally, the results of endometrial thinning are more pronounced when the endometrial thickness is relatively thick. For details on this procedure, refer to previously published reports^25,26.

1. Prepare 20 µL of saline using a pipette. By grasping both sides of the mouse's ears and fixing the tail of the mouse, gently insert the tip of the pipette into the vaginal orifice to a depth of approximately 2 mm. Infuse 20 µL of saline into the vagina and flush it gently for 5 cycles.
2. Withdraw the saline and transfer it evenly onto a clean slide. Allow it to evaporate and dry naturally at room temperature.
3. Apply 30 µL of ethanol to each slide to fix the cells. Allow the ethanol to evaporate for 5 min until the slide is dry.
4. Apply 50 µL of crystal violet staining solution to each slide and stain the cells at room temperature for 10 min.
5. Rinse each slide with deionized water to remove excess dye and reduce non-specific staining.
6. Observe the cell morphology and determine the proportion of cell types under a microscope.
7. Observe the estrous cycles of mice twice consecutively to ensure stability. Select mice in the third estrus for the experiment.
   NOTE: The estrous cycle in mice typically lasts 4-5 days and is divided into four stages: proestrus, estrus, metestrus, and diestrus. Proestrus lasts about 24 h and is characterized by round to oval nucleated epithelial cells (Figure 2A). Estrus lasts between 12 h and 48 h and is characterized by predominately anucleated keratinized epithelial cells (Figure 2B). Metestrus lasts up to 24 h and is characterized by a mix of anucleated keratinized epithelial cells and neutrophils (Figure 3C). Diestrus lasts between 48 h and 72 h and is characterized by a combination of neutrophils, nucleated epithelial cells, and a few anucleated keratinized cells (Figure 2D).

2. Group designing

NOTE: To assess the impact of various durations of 95% ethanol infusion on the stability of the TE model, six treatment groups were established. One group underwent sham surgery without ethanol administration to serve as a control for the procedural effects. Mice with ethanol-induced TE were randomly divided into five treatment groups based on the infusion time (n = 5 per group) as follows:

1. 1-min group: Inject ethanol into the uterine cavity for 1 min.
2. 2- min group: Inject ethanol into the uterine cavity for 2 min.
3. 3- min group: Inject ethanol into the uterine cavity for 3 min.
4. 4- min group: Inject ethanol into the uterine cavity for 4 min.
5. 5- min group: Inject ethanol into the uterine cavity for 5 min.

3. Induction of TE model in mice

1. Anesthetize the mice using an anesthesia machine with 5% isoflurane and 100% oxygen for 3 min (following an institutionally approved protocol). Maintain anesthesia at 1%-3% isoflurane using a nose cone. Apply veterinary ointment to the mice's eyes to prevent dryness during anesthesia. Monitor the respiratory rate with a toe pinch throughout the procedure.
2. Use medical tape to secure the mice's limbs and keep them in a supine position on the operating table preheated with an electronic blanket. Gently pull the mice's tongue slightly to prevent airway obstruction.
   NOTE: Keeping mice warm during surgery helps maintain immune function and blood circulation, and prevents clotting. Additionally, pulling out the mice's tongues ensures clear respiratory tracts to avoid suffocation during anesthesia.
3. Apply animal hair removal cream to the inferior abdomen of the mice for 3 min. Remove the hair with clean wipes. Disinfect the area with 75% ethanol and iodophor.
4. Make a 1 cm incision in the abdomen, 1 cm from the urethral orifice. Use sterile scissors and forceps to extend the incision down to the peritoneal cavity.
5. Gently move the entrails aside with sterile forceps to locate the uterus and expose the right uterine horn. Apply hemostatic clamps at the proximal end near the cervix and the distal end near the ovary.
   NOTE: Clamping both ends of the uterine horn ensures no leakage during ethanol injection, preventing damage to the cervix, vagina, and ovary.
6. Instill approximately 50 µL of 95% ethanol²⁷ into the uterine cavity using a 25 G syringe, with the needle parallel to the direction of the uterine horn. Insert the needle into the uterine cavity at a 30-degree angle. Maintain the syringe for 1 min, 2 min, 3 min, 4 min, or 5 min, as appropriate. Withdraw the ethanol and flush the uterine cavity with sterile saline 5 times to remove any remaining ethanol. Similarly, instill an equal volume of sterile saline into the left uterine horn as a control, following the same procedures.
   NOTE: Ensure the syringe needle is parallel to the direction of the uterine horn. Avoid inserting the needle at an angle that is too steep to prevent it from traversing the entire uterine horn. Withdraw the ethanol as completely as possible to avoid persistent damage and experimental instability.
7. Remove the clamps and return the uterus and entrails to their original position.
8. Suture the peritoneum with 5-0 absorbable polyglycolide surgical sutures, followed by suturing the epidermis with 5-0 non-absorbable polyglycolide surgical sutures.
9. Disinfect the incision with 75% ethanol and iodophor, and then return the animals to their cage. After the surgery, closely observe whether the heart rate of the mouse gradually and steadily increases until the mouse fully recovers and can move normally (Figure 3A-H).
   NOTE: Oral analgesic (3.2 mg/mL acetaminophen in the drinking water) was provided on the day before surgery and continued throughout the remainder of the study.
10. Observe the mice at 9:00 and 15:00 each day and disinfect the incision with iodophor to ensure they are in good condition.

4. Sample collection

1. Euthanize the mice by cervical spine dislocation while under deep anesthesia 7 days after 95% ethanol treatment (following an institutionally approved protocol).
2. Cut the sutures to expose the uterus using surgical scissors.
3. Carefully and slowly extract the entire uterus using surgical forceps. Detach the uterus from the surrounding entrails.
4. Immediately remove both uterine horns and divide each horn into 3-4 small segments.
5. Add 1 mL of 4% paraformaldehyde to a 1.5 mL microcentrifuge tube using a pipette. Preserve the uterine tissue in 4% paraformaldehyde for 24 h for fixation.

5. Tissue dehydration

NOTE: Tissue dehydration is achieved using an automatic tissue processor (see Table of Materials).

1. Place the formalin-fixed samples into tissue cassettes.
2. Transfer the tissue cassettes through a series of increasing ethanol concentrations to dehydrate the tissue as follows: 70% ethanol for 1 h, 85% ethanol for 1 h, 95% ethanol twice (1 h each), and 100% ethanol twice (1 h each).
3. Replace the ethanol with 100% xylene (135 min for the first immersion, 20 min for each subsequent immersion twice) to ensure that the paraffin penetrates the tissue.
4. Immerse the tissue cassettes in paraffin in an incubator at 60-65 °C (3 times, 140 min each), and keep the cassettes warm until embedding.

6. Paraffin embedding

NOTE: Perform these steps using a modular tissue embedding center (see Table of Materials).

1. Fill a wax mold with molten paraffin wax heated to 55 °C.
2. Use heated forceps to quickly transfer the infiltrated segments from the embedding cassette into the wax mold, ensuring that the cross-section of each segment rests parallel to the base of the mold.
3. Once the segments are properly positioned, place the bottom of the embedding cassette on top of the histology mold.
4. Place the mold directly on a cooling plate set to 4 °C for at least 20 min to allow the paraffin to solidify.
5. After the wax has solidified, remove the sample from the mold and use a paraffin trimmer to trim any excess wax around it.

7. Paraffin sectioning

1. Prepare a water bath at 42 °C. Place the paraffin blocks on a cold plate before cutting to allow them to cool and become easier to trim.
2. Clamp the paraffin-embedded specimen onto a rotary microtome with the tissue surface facing up.
3. Section 20 µm at a time to remove excess paraffin covering the tissue until the desired tissue plane is reached.
4. Cut 4 µm sections at a time.
5. Transfer the paraffin-sectioned ribbon to the 42 °C warm water bath for 2 min.
6. Pick up the sections from the water surface and place them onto a microscope slide after unfolding.
7. Dry the slides in a 60 °C incubator for 30 min.

8. Hematoxylin and eosin staining

NOTE: Hematoxylin and Eosin staining is performed using an automated slide stainer machine (see Table of Materials).

1. Prepare all reagents in advance before starting the staining procedure.
2. Soak the slides in 100% xylene for 10 min, 5 min, and 3 min, respectively.
3. Rehydrate the slides in a decreasing ethanol series in deionized water as follows: 100% ethanol, 95% ethanol (twice), and 80% ethanol, for 2 min each.
4. Rinse the slides in deionized water.
5. Stain the slides in hematoxylin solution for 10 min. Rinse the slides in deionized water for 1 min or until the water is no longer purple.
6. Dip the slides in 0.5% acid ethanol differentiation solution twice for 3-5 s each time. Rinse the slides in deionized water for 1 min.
7. Place the slides in a bluing buffer (1% ammonia solution) for 1 min. Rinse the slides in deionized water for 1 min.
8. Dehydrate the slides with 80% ethanol and then dip them in an eosin-Y working solution for 10 s.
9. Transfer the slides to 95% ethanol and rinse twice for 10 s each. Then, transfer the slides to absolute ethanol and dehydrate them twice for 10 s each.
10. Dip the slides in 100% xylene to clear the tissue twice, for 2 min each.
11. Place a coverslip over the tissue and seal it with natural resin.

9. Masson staining

1. Deparaffinize the tissue by soaking the slides in 100% xylene for 10 min, 5 min, and 3 min, respectively.
2. After deparaffinizing, soak the slides in a decreasing ethanol series in deionized water: 100% ethanol, 95% ethanol (twice), and 80% ethanol, for 2 min each.
3. Rinse the slides in deionized water.
4. Stain the slides in Weigert's iron hematoxylin solution for 5 min, then wash the slides in deionized water for 30 s.
5. Differentiate the colors of the stained tissue structures by incubating the slides in 1% hydrochloric acid alcohol for 15 s.
6. Rinse the slides in deionized water for 30 s.
7. Incubate the tissue in a Bluing Solution for 3-5 min.
8. Rinse the slides in deionized water for 30 s.
9. Place the slides in Ponceau-Acid Fuchsin Solution for 5-10 min.
   NOTE: Mix deionized water and weak acid solution in a ratio of 2:1 to prepare the weak acid working solution.
10. Rinse the slides with the weak acid working solution for 30 s.
11. Differentiate the tissue in the phosphomolybdic acid solution for 1-2 min, and then rinse the sections with a weak acid working solution for 30 s.
12. Stain the tissue sections with aniline blue solution for 1-2 min.
13. Dehydrate the tissue sections quickly in 95% ethanol for 2-3 s, followed by 100% ethanol twice for 5-10 s each. Clear the slides in xylene twice, for 1-2 min each.
14. Place a coverslip over the tissue and seal it with natural resin.
15. Drain excess resin and allow the slides to dry.

10. Imaging and analysis

1. Capture images of the sections using a slide scanner system.
   NOTE: Use a 4x or 10x objective for an overview of the sections; for more detailed images, use 20x to 100x objectives.
2. Perform quantitative analysis of endometrial thickness, the number of glands, and the extent of tissue fibrosis using the HALO image analysis platform.
   NOTE: Average thickness is obtained by measuring the vertical distance from the uterine cavity to the myometrium in 5 randomly selected fields of view.

11. Statistical analyses

1. Conduct statistical analyses using SPSS version 23.0 (SPSS Inc., Chicago, IL, USA).
2. Assess the statistical significance of experimental differences between the control and model groups by first performing a normal distribution test.
3. Present data that are normally distributed as mean ± SEM.
4. Analyze data from different treatment groups using one-way ANOVA. Adjust for multiple comparisons of the number of tested alleles in each locus using the Bonferroni method. Consider a P-value <0.05 as statistically significant.

The key features of thin endometrium (TE) are decreased endometrial thickness and glandular density, along with increased endometrial fibrosis. This method successfully replicated these characteristics in the model mice. Data analysis revealed a significant decrease in endometrial thickness in the 1-min group (222.3 µm ± 13.96 µm vs. 359.2 µm ± 12.41 µm, P < 0.05), the 2-min group (168.7 µm ± 17.57 µm vs. 359.2 µm ± 12.41 µm, P < 0.05), and the 3-min group (131.8 µm ± 3 µm vs. 359.2 µm ± 12.41 µm, P < 0.05) when compared to the control group. However, no significant difference in endometrial thickness was observed between the 2-min and 3-min groups (168.7 µm ± 17.57 µm vs. 131.8 µm ± 3 µm, P > 0.05) (Figure 3A,B).

The number of endometrial glands was significantly reduced in the 1-min group (4.4 ± 2.38 vs. 24.4 ± 1.6, P < 0.05), the 2-min group (0.8 ± 0.49 vs. 24.4 ± 1.6, P < 0.05), and the 3-min group (4.4 ± 3.20 vs. 24.4 ± 1.6, P < 0.05) when compared to the control group (Figure 3A,C). However, no significant difference in the number of endometrial glands was observed among the 1-min, 2-min, and 3-min groups.

Since TE occurred after treatment with 95% ethanol for 7 days, it was hypothesized that endometrial thickness might be associated with the infusion duration. To investigate this, two additional groups were treated with 95% ethanol for 4 min and 5 min, respectively. Interestingly, the results showed severe uterine adhesion when the infusion time increased. (Figure 4), with the uterine cavity becoming almost invisible. Additionally, no endometrial glands were observed in the mouse uterus with the prolonged infusion time (Figure 4).

Moreover, the level of endometrial fibrosis increased with the duration of 95% ethanol infusion, as confirmed by Masson staining, indicating excessive fibrosis in the damaged endometrium (Figure 5A,B).

Figure 1: Flowchart of the TE mouse model establishment. The entire procedure includes the following steps: First, mice were allowed to acclimate for 1 week before the experiment. Next, two estrous cycles were monitored. Mice in the third estrus were then selected for model establishment by injecting 95% ethanol into the uterine cavity. Seven days after the ethanol injection, the mice were sacrificed, and the uterine horns were collected for further HE staining and Masson staining analysis. Please click here to view a larger version of this figure.

Figure 2: Identification of estrus in mice. (A) Proestrus is characterized by the predominance of nucleated epithelial cells (arrow). (B) Estrus is characterized by the predominance of anucleated keratinized epithelial cells (asterisks). (C) Metestrus is characterized by a combination of anucleated keratinized epithelial cells and neutrophils (triangle). (D) Diestrus is characterized by a combination of neutrophils, nucleated epithelial cells, and a low number of anucleated keratinized cells. Objective magnification of 20x. Scale bars = 50 µm. Please click here to view a larger version of this figure.

Figure 3: Procedure for TE induction by 95% ethanol infusion. (A) Mice are anesthetized using an anesthesia machine. (B) Abdominal hair is removed. (C) A 1 cm incision is made 1 cm away from the urethral orifice. (D) The uterus is located, and one side of the uterine horn is exposed. (E) Hemostatic clamps are applied on both sides of the uterine horn. (F) 95% ethanol is instilled into the uterine horn and then flushed with sterile saline. (G) The clamps are removed, and the uterus is placed back. (H) The incision is sutured. Please click here to view a larger version of this figure.

Figure 4: Observation of endometrial morphology in mice using HE staining under microscopy. (A) The morphology of the endometrium on the injured and control sides was examined. The right-side uterus was injected with 95% ethanol, while the left-side uterus was injected with sterile saline as a control. Scale bars = 200 µm. (B) Comparison of endometrial thickness among the Con group, 1 min, 2 min, and 3 min groups. (C) Comparison of the number of glands in the endometrium among the Con group, 1 min, 2 min, 3 min, 4 min, and 5 min groups. Con, control; **P < 0.01, ****P < 0.0001 with one-way ANOVA. Data are represented as mean ± standard error of the mean. Please click here to view a larger version of this figure.

Figure 5: Endometrial fibrosis in the mouse endometrium. (A) Masson's trichrome staining was used to evaluate endometrial fibrosis, where the blue color indicates the distribution of collagen fibers, in the indicated treatment groups. Scale bars = 200 µm. (B) A summary bar chart shows the level of collagen fibrosis among different groups. **P < 0.01, ***P < 0.001, ****P < 0.0001 with one-way ANOVA. Data are represented as mean ± standard error of the mean. Please click here to view a larger version of this figure.

TE is characterized by insufficient cell proliferation and dysfunctional cells, closely linked to infertility, recurrent miscarriage, and placental abnormalities^2,3. Unfortunately, there is currently no effective therapy for TE. Animal models play a crucial role in studying this condition. Between 2014 and 2024, rats were used as model organisms in 16.4% of 208,000 studies (34,200 studies) and mice in 22.7% (47,300 studies). The increasing use of mice in experiments can be attributed to their advantages over other models, particularly their potential in regeneration studies^28,29,30. Given the importance of model consistency, there is a growing need for more compatible tools to conduct studies using mice as models³¹.

Common methods for inducing TE models, as reported in previous studies, include scratching^11,12, ischemia¹⁴, thermal-induced injury¹⁵, and chemical-induced injury^16,17,23. To facilitate the use of animal models in TE studies and lay the foundation for effective treatments, Gao et al.²⁴ first established a TE model in SD rats by injecting absolute ethanol for 5 min to injure the endometrium. Although this method restricts endometrial growth, it often leads to severe adhesions in the abdominal cavity, with a survival rate of only 70%.

Previously, 95% ethanol was predominantly used in rats due to its ease of application, but this method generated few mouse models for TE studies. After clinical observations, a comprehensive literature review, pre-experiments, repeated discussions, and analysis, an improved method with a shorter uterine infusion time was developed. Skilled techniques allow 95% ethanol to be injected into the uterine cavity to create a TE model with varying infusion times.

The mouse model introduced in this study builds on the principles of Gao's model²⁴. However, several modifications were made, including shortening the ethanol infusion time and increasing the saline flushing frequency. In this model, mouse uteri were exposed via abdominal dissection, and 95% ethanol was injected into one side of the uterine horns for various durations: 1 min, 2 min, 3 min, 4 min, and 5 min. Thanks to the precise control over the ethanol infusion time and the standardized procedure, stable data were achieved in mouse experiments, which minimized deviations due to individual differences among the mice.

Histological results demonstrate clear changes in endometrial thickness, the number of endometrial glands, and endometrial fibrosis in varying degrees among the TE mice, confirming the reliability of the model established in this study^3,6,9. Notably, when the ethanol infusion time exceeded 4 min or 5 min, severe intrauterine adhesions (IUA) were observed. These findings suggest that the TE mouse model can be effectively established with 1-3 min of 95% ethanol infusion, while infusion times exceeding 3 min may induce severe IUA.

To date, several animal models have been developed to study TE, including scratching, ischemia, thermal-induced, and chemical-induced models. The scratching-induced model, while easier to perform, often leads to intrauterine adhesions (IUA) rather than TE. The ischemia-induced TE model requires a much longer duration than the chemical-induced model, and the thermally induced method is more complex and difficult to execute compared to chemical induction. Although the chemical-induced model is time-efficient and stable, it does have potential limitations, such as trauma to the animals and risks of infection or death due to improper technique. These issues can be mitigated with careful training and experience.

In conclusion, the results demonstrate that the TE mouse model can be effectively established by injecting 95% ethanol into the uterine cavity for 1-3 min. This approach offers a promising method for studying the pathological mechanisms and repair processes of TE, and for developing effective treatments for patients with this condition.

The authors have nothing to disclose.

We gratefully thank the anonymous referees for their important and helpful comments. This work was supported by the Shenzhen Science and Technology Project (No. JCYJ20220818103207016) and the Guangdong Basic and Applied Basic Research Foundation (No. 2024A1515010478).