https://orcid.org/0000-0001-7536-6099
Abstract
Is the ratio of endometrial T-box expressed in T cell (T-bet) and GATA-binding protein 3 (GATA3) changed in patients with recurrent miscarriage (RM) compared to fertile controls?
Our study showed a significantly higher T-bet/GATA3 ratio in patients with RM compared with fertile controls.
The endometrial T-bet (Th1 lineage-committed transcription factor)/GATA3 (Th2 lineage-committed transcription factor) ratio could represent the Th1/Th2 balance, which is particularly important for healthy pregnancy. However, a reliable reference range for the T-bet/GATA3 ratio during the peri-implantation period has not yet been established for use in clinical practice.
This was a retrospective study carried out in a private fertility center. The control group included 120 women in couples undergoing IVF treatment for male infertility, who had experienced a live-birth baby following the first IVF cycle. The study group included 93 women diagnosed with RM that experienced at least two consecutive clinically spontaneous miscarriages before gestational week 12. The ratio of T-bet/GATA3 was calculated in the control group and RM group.
Endometrium samples were collected at mid-luteal phase of the menstrual cycle prior to IVF treatment or pregnancy. The percentage of T-bet+ and GATA3+ cells in total endometrial cells was analyzed using immunohistochemical staining and quantitative analysis.
Using the 95th percentile to define the upper limits of the endometrial T-bet/GATA3 ratio during the mid-luteal phase, the reference range of control fertile women was ≤0.22. Compared with the control group, the RM group exhibited a significantly higher T-bet/GATA3 ratio (P = 0.02), and 19.4% (18/93) women with RM exhibited a T-bet/GATA3 ratio above the reference range in the mid-luteal phase.
All patients were recruited from a single center. The stability and clinical value of the endometrial T-bet/GATA3 ratio require further investigation.
The present study suggests that an abnormal endometrial T-bet/GATA3 ratio may be one of the risk factors of RM. Further studies are needed to follow up the pregnancy outcomes in patients with RM with normal and abnormal endometrial T-bet/GATA3 ratio according to the reference range.
This work was supported by Shenzhen Fundamental Research Program (JCYJ20180228164631121, JCYJ20190813161203606, JCYJ20220530172817039). There are no conflicts of interest to declare.
N/A.
Introduction
Recurrent miscarriage (RM) is defined as two or more consecutive pregnancy losses before 20 weeks of gestation, affecting 1–2% of couples (Practice Committee of the American Society for Reproductive Medicine, 2013). Following the exclusion of potential causes, such as abnormal karyotype, anatomy, infection, or endocrine disorders, 40–50% of patients experience RM with an unknown cause (Rai and Regan, 2006). Pregnancy is considered a dynamic process rather than a single event, characterized by distinct immune states at each biological process (Mor and Cardenas, 2010; PrabhuDas et al., 2015). Numerous studies have suggested that dysregulation of the maternal immune system may be associated with the pathogenesis of RM (Abdolmohammadi Vahid et al., 2019; Ali et al., 2020; Yu et al., 2023).
A T helper (Th)1/Th2 balance represents the inflammatory state of the maternal–fetal interface, and an imbalance of Th1/Th2 is considered a probable cause of miscarriage (Romagnani et al., 2000). Th1 cells produce interferon (IFN)-γ and tumor necrosis factor (TNF)-α and are involved in cell-mediated immunity. In contrast, Th2 cells produce IL-4 and IL-10 and participate in humoral immunity. An upregulation of IFN-γ and TNF-α may lead to Th1/Th2 imbalance at the maternal interface, subsequently resulting in miscarriage (Chaouat et al., 2004). Previous studies have demonstrated that evaluated peripheral TNF-α/IL-4 and TNF-γ/IL-10 may reflect the systemic contribution of Th1 immune responses to RM (Kwak-Kim et al., 2003; Kuroda et al., 2021). Notably, a local immune reaction occurs in the endometrium at the time of implantation, and embryonic signals act on endometrial immune cells to induce immune tolerance. However, there is insufficient evidence for modulation of the endometrial Th1/Th2 cell ratio during the mid-luteal phase in patients with RM.
T-box expressed in T cell (T-bet) and GATA-binding protein 3 (GATA3) are key transcription factors that direct Th1 and Th2 cell differentiation, and play an important role in regulating the Th1/Th2 balance (Kanhere et al., 2012). Increased peripheral T-bet and decreased GATA3 mRNA levels were observed in RMs compared with healthy women (Abdolmohammadi Vahid et al., 2019). In a mouse model of RM, the T-bet/GATA3 ratio increased in the maternal circulation and decidual tissues (Luo et al., 2021). Results of our previous study demonstrated that an elevated endometrial T-bet/GATA3 ratio may predict adverse pregnancy outcomes in infertile patients, suggesting that the endometrial T-bet/GATA3 balance may play a key role in the maternal–fetal interface (Li et al., 2022). However, the role of T-bet/GATA3 ratio in the endometrium of patients with RM remains to be fully elucidated. In addition, a reliable reference range for the T-bet/GATA3 ratio during the peri-implantation period has not yet been established for use in clinical practice.
The present study aimed to determine a reference range for the T-bet/GATA3 ratio in the peri-implantation endometrium of control fertile women and compare this with the ratio in patients with RM.
Materials and methods
Subjects
Participants were recruited from a population of women admitted to the Fertility Center of Shenzhen Zhongshan Urology Hospital (SZUH) during February 2017 and December 2021. All eligible patients agreed to undergo endometrial biopsy in order to be assessed for chronic endometritis and received a detailed informed consent that they signed to allow scientific research on their endometrial tissue. The information of enrolled patients was recorded in the Hospital Information System of SZUH.
Two groups of women were enrolled in the study: Control fertile women in couples undergoing IVF-embryo transfer (IVF-ET) treatment for male infertility (oligospermia, asthenospermia, teratospermia), who had experienced a live-birth following the first IVF cycle (n = 120); and patients with RM that experienced at least two consecutive clinically spontaneous miscarriages before gestational week 12 (n = 93).
The flowchart for recruitment of control fertile women is presented in Fig. 1. The inclusion criteria of the control group were as follows: aged <40 years; exhibited a regular menstrual cycle of between 28 and 32 days; underwent endometrial biopsy prior to controlled ovarian hyperstimulation, and subsequent hematoxylin–eosin (H&E) staining revealed endometrial tissue in the mid-luteal phase; and underwent ET within 3 months after endometrial sampling. The exclusion criteria of the control group were as follows: experienced any spontaneous miscarriage or biochemical pregnancy; chromosome abnormality; abnormal uterus, such as endometriosis, endometrial polyps, adenomyosis or scarred uterus; exhibited PCOS; and salpingitis or chronic endometritis.
Flowchart for recruitment of control fertile women. Eventually, 174 control fertile women were enrolled in the study for establishing the reference range (n = 124), and observing the changes in the endometrial T-bet/GATA3 ratio during the menstrual cycle (n = 50). T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3; ET: embryo transfer.
A further 50 control fertile women were retrospectively enrolled to assess the endometrial T-bet/GATA3 ratio at different menstrual phases, including 14 women biopsied in the proliferative phase (Pro), nine women biopsied in the early secretory phase (ES), 18 women biopsied in the mid-secretory phase (MS), and nine women biopsied in the late secretory phase (LS). The histologic dating of the specimens was subsequent verified using H&E staining.
The inclusion criteria of the RM group were as follows: aged <40 years; exhibited a regular menstrual cycle of 28–32 days; underwent endometrial biopsy, and subsequent H&E staining revealed endometrial tissue in the mid-luteal phase. The exclusion criteria were: chromosome abnormality; salpingitis or chronic endometritis; and exhibited antiphospholipid syndrome.
Ethical approvals
The present study was approved by the Ethics Committee of Shenzhen Zhongshan Urology Hospital (approval no. SZZSECHU-F-2021040).
Baseline sex hormone level assays
Peripheral bloods from each participant were collected on Days 3–5 of a natural menstrual cycle. Serum concentrations of LH, FSH, estradiol, progesterone, prolactin (PRL), and total testosterone were examined using competitive electrochemiluminescence immunoassays on the Elecsys 2010 autoanalyzer according to the manufacturer’s instructions (Roche Diagnostics, Indianapolis, IN, USA). The results are listed in Table 1.
Characteristics of the control fertile women and patients with recurrent miscarriage.
| Characteristics | Control | RM | P value |
|---|---|---|---|
| n | 120 | 93 | |
| Female age (years) | 32.0 (28.5, 34.0) | 34.0 (30.5, 36.0) | <0.001 |
| Female age ≥35 years | 22 (18.3%) | 39 (41.9%) | <0.001 |
| Female BMI (kg/m2) | 20.61 ± 2.21 | 21.3 (19.4, 22.5) | 0.092 |
| Basal LH (mIU/ml) | 4.4 (3.2, 6.2) | 4.5 (3.1, 6.2) | 0.701 |
| Basal FSH (mIU/ml) | 5.4 (2.9, 7.0) | 5.8 (4.7, 6.8) | 0.065 |
| Basal Estradiol (pg/ml) | 45.8 (30.7, 127.9) | 38.9 (29.7, 58.4) | 0.081 |
| Basal Progesterone (ng/ml) | 0.7 (0.4, 9.4) | 0.4 (0.3, 5.0) | 0.071 |
| PRL (ng/ml) | 16.8 (11.3, 27.7) | 14.2 (10.6, 19.9) | 0.090 |
| T (ng/ml) | 0.3 (0.2, 0.4) | 0.3 (0.2, 0.5) | 0.073 |
| RM classification based on previous miscarriages | |||
| RM2 | 59 (63.4%) | ||
| RM3 | 30 (32.3%) | ||
| RM4&5 | 4 (4.3%) |
| Characteristics | Control | RM | P value |
|---|---|---|---|
| n | 120 | 93 | |
| Female age (years) | 32.0 (28.5, 34.0) | 34.0 (30.5, 36.0) | <0.001 |
| Female age ≥35 years | 22 (18.3%) | 39 (41.9%) | <0.001 |
| Female BMI (kg/m2) | 20.61 ± 2.21 | 21.3 (19.4, 22.5) | 0.092 |
| Basal LH (mIU/ml) | 4.4 (3.2, 6.2) | 4.5 (3.1, 6.2) | 0.701 |
| Basal FSH (mIU/ml) | 5.4 (2.9, 7.0) | 5.8 (4.7, 6.8) | 0.065 |
| Basal Estradiol (pg/ml) | 45.8 (30.7, 127.9) | 38.9 (29.7, 58.4) | 0.081 |
| Basal Progesterone (ng/ml) | 0.7 (0.4, 9.4) | 0.4 (0.3, 5.0) | 0.071 |
| PRL (ng/ml) | 16.8 (11.3, 27.7) | 14.2 (10.6, 19.9) | 0.090 |
| T (ng/ml) | 0.3 (0.2, 0.4) | 0.3 (0.2, 0.5) | 0.073 |
| RM classification based on previous miscarriages | |||
| RM2 | 59 (63.4%) | ||
| RM3 | 30 (32.3%) | ||
| RM4&5 | 4 (4.3%) |
Continuous data with normal distribution are reported as mean ± SD and analyzed using Student’s t test, and non-normally distributed data are presented as the median (quartiles) and analyzed using a Kruskal–Wallis test. Categorical variables presented as the number (percentage) and analyzed using a Chi-square (χ2) test. RM: recurrent miscarriage; D3: Day 3; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3; PRL: prolactin; T: total testosterone. Hormones were measured in peripheral blood samples taken from each participant on Days 3–5 of a natural menstrual cycle.
Characteristics of the control fertile women and patients with recurrent miscarriage.
| Characteristics | Control | RM | P value |
|---|---|---|---|
| n | 120 | 93 | |
| Female age (years) | 32.0 (28.5, 34.0) | 34.0 (30.5, 36.0) | <0.001 |
| Female age ≥35 years | 22 (18.3%) | 39 (41.9%) | <0.001 |
| Female BMI (kg/m2) | 20.61 ± 2.21 | 21.3 (19.4, 22.5) | 0.092 |
| Basal LH (mIU/ml) | 4.4 (3.2, 6.2) | 4.5 (3.1, 6.2) | 0.701 |
| Basal FSH (mIU/ml) | 5.4 (2.9, 7.0) | 5.8 (4.7, 6.8) | 0.065 |
| Basal Estradiol (pg/ml) | 45.8 (30.7, 127.9) | 38.9 (29.7, 58.4) | 0.081 |
| Basal Progesterone (ng/ml) | 0.7 (0.4, 9.4) | 0.4 (0.3, 5.0) | 0.071 |
| PRL (ng/ml) | 16.8 (11.3, 27.7) | 14.2 (10.6, 19.9) | 0.090 |
| T (ng/ml) | 0.3 (0.2, 0.4) | 0.3 (0.2, 0.5) | 0.073 |
| RM classification based on previous miscarriages | |||
| RM2 | 59 (63.4%) | ||
| RM3 | 30 (32.3%) | ||
| RM4&5 | 4 (4.3%) |
| Characteristics | Control | RM | P value |
|---|---|---|---|
| n | 120 | 93 | |
| Female age (years) | 32.0 (28.5, 34.0) | 34.0 (30.5, 36.0) | <0.001 |
| Female age ≥35 years | 22 (18.3%) | 39 (41.9%) | <0.001 |
| Female BMI (kg/m2) | 20.61 ± 2.21 | 21.3 (19.4, 22.5) | 0.092 |
| Basal LH (mIU/ml) | 4.4 (3.2, 6.2) | 4.5 (3.1, 6.2) | 0.701 |
| Basal FSH (mIU/ml) | 5.4 (2.9, 7.0) | 5.8 (4.7, 6.8) | 0.065 |
| Basal Estradiol (pg/ml) | 45.8 (30.7, 127.9) | 38.9 (29.7, 58.4) | 0.081 |
| Basal Progesterone (ng/ml) | 0.7 (0.4, 9.4) | 0.4 (0.3, 5.0) | 0.071 |
| PRL (ng/ml) | 16.8 (11.3, 27.7) | 14.2 (10.6, 19.9) | 0.090 |
| T (ng/ml) | 0.3 (0.2, 0.4) | 0.3 (0.2, 0.5) | 0.073 |
| RM classification based on previous miscarriages | |||
| RM2 | 59 (63.4%) | ||
| RM3 | 30 (32.3%) | ||
| RM4&5 | 4 (4.3%) |
Continuous data with normal distribution are reported as mean ± SD and analyzed using Student’s t test, and non-normally distributed data are presented as the median (quartiles) and analyzed using a Kruskal–Wallis test. Categorical variables presented as the number (percentage) and analyzed using a Chi-square (χ2) test. RM: recurrent miscarriage; D3: Day 3; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3; PRL: prolactin; T: total testosterone. Hormones were measured in peripheral blood samples taken from each participant on Days 3–5 of a natural menstrual cycle.
Endometrial biopsy
Endometrial samples in the Pro phase were obtained using biopsy forceps during hysteroscopy. Endometrial biopsies in the secretory phase were collected using a standard approach with a pipelle catheter (Jiangxi Nuode Medical Healthy Science Company, Jiangxi, China). Collected samples were immediately placed into 10% neutral buffered formalin for 4–6 h fixation at room temperature, and the fixed tissues were subsequently dehydrated and embedded into paraffin wax.
H&E staining
H&E staining and histologic dating of the endometrium were used to precisely divide specimens into the Pro, ES, MS, and LS phases. Paraffin-embedded tissue samples were sectioned into 4-μm-thick slices. One tissue section was selected for H&E staining performed on a Tissue-Tek Prisma Automated slide stainer (Sakura Finetek, CA, USA). Histologic dating of the endometrium was determined by a pathologist using criteria described previously (Noyes et al., 1950).
Immunohistochemical staining of T-bet and GATA3
A total of two tissue slides were randomly selected for immunohistochemical (IHC) staining using a Leica Bond III automated IHC stainer (Leica Microsystems, Wetzlar, Germany) and BOND Polymer Refine Detection (DS9800, Leica Microsystems, Wetzlar, Germany). Each slide was incubated with anti-human monoclonal mouse T-bet antibody (clone 4B10, BD Biosciences, San Diego, CA, USA) and anti-human monoclonal mouse GATA3 antibody (clone 634913, R&D Systems, MN, USA) for 30 min. The slides were subsequently washed and incubated with the corresponding secondary antibody. Following the addition of 3,3’-diaminobenzidine and a hematoxylin counterstain, the stained slides were cover-slipped using Tissue-Tek Film Automated Coverslipper (Sakura Finetek, CA, USA). The mouse IgG1 κ isotype control (clone MOPC-31C, BD Biosciences, San Diego, CA, USA) for T-bet antibody and mouse IgG2B isotype control (clone 20116, R&D Systems, MN, USA) for GATA3 antibody were used (Supplementary Fig. S1). No non-specific interactions between antibodies and the sample were detected.
Cell counting methodology
As shown in Fig. 2, the slides were initially scanned at a low magnification, and 30 fields (×200 magnification) were randomly selected and captured using the Perkin-Elmer system (Perkin-Elmer, MA, USA). The number of positive (T-bet+ and GATA3+) and total endometrial cells from each panel were characterized and counted using the cell segmentation and phenotype tools of the Inform Cell Analysis system (Perkin-Elmer, MA, USA), and subsequently checked by a professional pathologist. The staining for T-bet and GATA3 was presented as a percentage of total endometrial cells for each image and reported as an average of all 20 fields. The T-bet/GATA3 ratio was calculated as the percentage of T-bet+ cells/the percentage of GATA3+ cells.
Photomicrographs and quantitative analysis path for human endometrial T-bet and GATA3. (A) Panoramic scanning of endometrium tissue sections that stained with T-bet and GATA3 using immunohistochemistry (×40 magnification). (B) Randomly selecting fields at low magnification (×40 magnification). (C) Identification of endometrial cells (blue) and positive immune cells (T-bet+ and GATA3+: red) in the selected field (×200 magnification; 0.0625 mm2/field), and each panel was characterized and counted using the cell segmentation and phenotype tools of the Inform Cell Analysis system. (D) Staining patterns of T-bet (upper) and GATA3 (lower) in the endometrium (×200 magnification). Both T-bet and GATA3 were expressed in the cell nucleus, and presented as individual cells throughout the stroma. T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
Interobserver variability
To determine the interobserver variability using the Inform Cell Analysis system, the counting of T-bet+ and GATA3+ cells was evaluated by two independent observers on six slides (two slides with high, two slides with medium, and two slides with low expression) for each, using the protocol described above.
Quantification platform verification
A total of 20 randomly selected slides were scanned at ×200 magnification using the Olympus SLIDEVIEW™ VS200 (Olympus Corporation, Tokyo, Japan). HALO Image Analysis System (Indica Labs, Albuquerque NM, USA) was used for the quantification of T-bet+ and GATA3+ cells under the supervision of at least one trained pathologist. For both T-bet and GATA3 IHC analysis, a classifier program was used to identify tissue regions, and a digital analysis program was used to quantify cells based on color deconvolution. The population of T-bet+ and GATA3+ cells was measured as the percentage in all endometrial cells of whole tissue.
Multicolor immunofluorescence
Formalin-fixed paraffin-embedded (FFPE) slides were processed automatically for multi-color immunofluorescence by using a Leica Bond RX automated IHC stainer (Leica Microsystems, Wetzlar, Germany) and tyramide signal amplification Opal kit (PerkinElmer, Waltham, MA, USA). The antibodies and detection reagents used are presented in Supplementary Table S1. After primary antibody and secondary antibody incubation, the slides were labeled with fluorophores in combination with tyramide signal amplification, according to manufacturer’s protocol. The nuclei were subsequently visualized with DAPI, and the section was coverslipped using antifade mounting medium. The stained slides were imaged using the Olympus SLIDEVIEW™ VS200 (Olympus Corporation, Tokyo, Japan).
IHC double-staining
For each patient, two tissue slides were randomly selected for IHC double-staining by using a two-polymer detection system: Bond Polymer Refine Detection Kit (peroxidase-based detection, DAB brown chromogen; DS9800, Leica Microsystems, Wetzlar, Germany) and Bond Polymer Refine Red Detection Kit (alkaline phosphatase-based detection, red chromogen; DS9390, Leica Microsystems, Wetzlar, Germany). The staining panels were CD4-brown/T-bet-Red and CD4-brown/GATA3-Red. The automatic IHC was performed on a Leica Bond III automated IHC stainer (Leica Microsystems, Wetzlar, Germany). Following a hematoxylin counterstain, the stained slides were cover-slipped using Tissue-Tek Film Automated Coverslipper (Sakura Finetek, CA, USA). The slide scanning and cell counting were performed as described above.
Statistical analysis
All statistical analyses were performed using SPSS software (version, 22.0; IBM Corp., Armonk, NY, USA). Outliers were removed using the Dixon test. The normality of distribution of continuous variables was examined using a one-sample Kolmogorov–Smirnov test. The reference range was defined by the 95th percentiles, according to the recommendations of the CLSI C28-A3 guidelines, which was published by the Clinical and Laboratory Standards Institute (CLSI, previously National Committee for Clinical Laboratory Standards (NCCLS)) for Defining, Establishing and Verifying Reference Intervals in the Clinical Laboratory and provides recommendations for determining reference values and reference intervals for quantitative clinical laboratory tests (Cals, 2008). Continuous data with normal distribution were reported as mean±SD and analyzed using Student’s t test, and non-normally distributed data were presented as the median (quartiles) and analyzed using a Kruskal–Wallis test. Categorical variables were presented as the number (percentage) and analyzed using a Chi-square (χ2) test. Pearson’s correlation coefficient was used to calculate the correlation coefficient between two quantitative parameters and presented in a scatter diagram. (van den Berg et al., 2021). P < 0.05 was considered to indicate a statistically significant difference.
Results
Staining patterns and quantitative analysis path
The IHC staining for T-bet+ and GATA3+ cells in the peri-implantation endometrium is displayed in Fig. 2. Both T-bet and GATA3 were present in the cell nucleus and presented as individual cells throughout the stroma.
Characteristics and baseline of the participants
The demographics and characteristics of the control fertile women and patients with RM are summarized in Table 1. No significant variations were observed in female BMI, Day 3 (D3) LH, D3 FSH, D3 estradiol, D3 progesterone, PRL, and testosterone between two groups, but the female age of patients with RM was significantly higher than control fertile women (P<0.001). We further evaluated whether T-bet/GATA3 ratio is affected by female age and found no correlation between female age and T-bet/GATA3 ratio (Pearson correlation (PC) = 0.120, P = 0.080, Supplementary Fig. S2).
T-bet/GATA3 ratio in control fertile women
After removing four outliers using the Dixon test, 120 control subjects were enrolled to establish the reference range for the T-bet/GATA3 ratio. The percentile distributions of the T-bet/GATA3 ratio are displayed in Table 2. With no lower limit, and the 95th percentile to define the upper limits of the endometrial T-bet/GATA3 ratio during the mid-luteal phase, the reference range of control fertile women was established as ≤0.22.
T-bet/GATA3 ratio in control infertile women.
| Group | n | Median | Range | Reference internal (<95th) |
|---|---|---|---|---|
| Fertile controls | 120 | 0.13 | 0.03–0.28 | <0.22 |
| Group | n | Median | Range | Reference internal (<95th) |
|---|---|---|---|---|
| Fertile controls | 120 | 0.13 | 0.03–0.28 | <0.22 |
T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
T-bet/GATA3 ratio in control infertile women.
| Group | n | Median | Range | Reference internal (<95th) |
|---|---|---|---|---|
| Fertile controls | 120 | 0.13 | 0.03–0.28 | <0.22 |
| Group | n | Median | Range | Reference internal (<95th) |
|---|---|---|---|---|
| Fertile controls | 120 | 0.13 | 0.03–0.28 | <0.22 |
T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
T-bet/GATA3 ratio in women with RM
The median T-bet/GATA3 ratio in women with RM (n = 93) was 0.13 (range, 0.05–0.50). Compared with the fertile controls, there was a significantly higher T-bet/GATA3 ratio in women with RM (P = 0.02, Fig. 3). Patients with different numbers of previous miscarriages were marked using dots with different color (RM2 as green dot, RM3 as blue dot, RM4&5 as orange dot). Moreover, using the 95th percentile (0.22) as the upper limit to define the reference range of the T-bet/GATA3 ratio, 18 of the 93 (19.4%) women with RM were above the reference range. We also compared the T-bet/GATA3 ratio in patients with two, three, and more than four miscarriages with control women, and the results suggested an increased T-bet/GATA3 ratio with increasing miscarriages, but these differences were not statistically significant (Supplementary Fig. S3).
Endometrial T-bet/GATA3 ratio in control fertile women and women with recurrent miscarriage. Compared with control fertile women, a significantly increased T-bet/GATA3 ratio was observed in women with RM (P = 0.02). The horizontal dotted line depicts the 95th percentile (0.22) in the control group, as the upper limit to define the reference range of the T-bet/GATA3 ratio. Notably, 18 of 93 (∼19.4%) women with RM were above the reference range. Patients with different numbers of previous miscarriages are presented as dots with different colors (patient RM2 as green dot, RM3 as blue dot, RM4&5 as orange dot). Ctrl: Control; RM: recurrent miscarriage; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
Interobserver variability
To determine the interobserver variability for measuring the percentage of T-bet+ and GATA3+ cells, six slides were selected for each marker, and analyzed by two independent observers. As shown in Fig. 4, there was a significant correlation (T-bet, PC = 0.998, P < 0.001; GATA3, PC = 0.996, P < 0.001) between the measurements of observer A and B.
Interobserver variability for T-bet and GATA3 cell count measurements in endometrium. There was a significant correlation between the percentage measurements of two independent observers in recognizing the T-bet+ (A: PC = 0.998, P < 0.001) and GATA3+ cells (B: PC = 0.996, P < 0.001). PC: Pearson correlation; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
Quantification platform verification
The quantification of T-bet+ and GATA3+ cells in IHC images was analyzed using the Inform Cell Analysis system. In order to assess the accuracy and variability of the quantification platform, images of 20 specimens in the control group were measured using the HALO Image Analysis System. As shown in Fig. 5, results of the present study demonstrated that the T-bet+ percentage, GATA3+ percentage, and T-bet/GATA3 ratio determined using the Inform Cell Analysis system exhibited a strong correlation with the results obtained using the HALO Image Analysis System (T-bet, PC = 0.933, P < 0.001; GATA3, PC = 0.969, P < 0.001; T-bet/GATA3 ratio, PC = 0.939, P < 0.001).
Correlation between two immunohistochemical quantification platforms. The T-bet+ percentage, GATA3+ percentage, and T-bet/GATA3 ratio determined using the Inform Cell Analysis system exhibited a strong correlation with the results obtained using the HALO Image Analysis System (A: T-bet, PC = 0.933, P < 0.001; B: GATA3, PC = 0.969, P < 0.001; C: T-bet/GATA3 ratio, PC = 0.939, P < 0.001). PC: Pearson correlation; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
Endometrial T-bet/GATA3 ratio in different phases of the menstrual cycle
In order to observe the changes of the endometrial T-bet/GATA3 ratio during the menstrual cycle, the expression levels of T-bet and GATA-3 in the human endometrium during a normal menstrual cycle were determined. Following H&E staining and histologic dating, IHC analysis of T-bet and GATA3 was carried out using specimens in the Pro (n = 14), ES (n = 9), MS (n = 18), and LS (n = 8) phases. The results showed that the T-bet/GATA3 ratio was significantly increased during the Pro phase, compared with the ES (P < 0.05), MS (P < 0.0001), and LS phases (P < 0.001). Moreover, the T-bet/GATA3 ratio in the ES phase was significantly increased compared with the MS (P < 0.01) and LS phases (P < 0.001). However, the difference between the MS and LS phases was not significant (Fig. 6).
Fluctuations of the T-bet/GATA3 ratio across four phases of the menstrual cycle. The T-bet/GATA3 ratio was evaluated in the Pro (n = 14), ES (n = 9), MS (n = 18), and LS (n = 8) phases, and the ratio was significantly increased during the Pro phase, compared with the ES (P < 0.05), MS (P < 0.0001), and LS phases (P < 0.001). Moreover, the T-bet/GATA3 ratio in the ES phase was significantly increased compared with the MS (P < 0.01) and LS phases (P < 0.001). Pro: proliferative phase; ES: early secretory phase; MS: mid-secretory phase; LS: late secretory phase; T-bet: T-box expressed in T cell; GATA3: GATA-binding protein 3.
T-bet and GATA3 expressed in various immune cell types
In order to identify T-bet and GATA3 expression in immune cell types, multiple color immunofluorescence was performed on FFPE tissue sections (Supplementary Fig. S4). The results showed that the T-bet and GATA3 were not only expressed in CD4+ Th cells, but also in CD8+ cytotoxic T lymphocytes (CTL) and CD56+ natural killer (NK) cells. In order to analyze whether T-bet/GATA3 ratio can represent Th1/Th2 balance, we performed IHC double-staining with CD4 and T-bet/GATA3 and analyzed the expression pattern of T-bet and GATA3 in CD4+T cells (Supplementary Fig. S5A and B). The results show that the ratio of T-bet/GATA3 is consistent with the ratio of CD4+T-bet+/CD4+GATA3+ (Supplementary Fig. S5C, PC = 0.941, P<0.001).
Discussion
In the present study, a standardized image analysis platform was used to determine endometrial T-bet+ and GATA3+ cells, and a group of 120 control fertile women was used to establish a reference range for the T-bet/GATA3 ratio in the peri-implantation endometrium, following the CLSI C28-A3 guidelines. Compared with the control group, we demonstrated a significantly higher T-bet/GATA3 ratio in women with RM, and 19.4% patients with RM exhibited a T-bet/GATA3 ratio above the reference range, which may be one of the etiologic factors of RM.
In our previous study, the predictive value of the T-bet/GATA3 ratio was demonstrated for live birth in infertile patients, suggesting the potential clinical application of the endometrial T-bet/GATA3 ratio (Li et al., 2022). We have also found much higher T-bet/GATA3 ratio both in non-clinical pregnancy and miscarriage groups than in clinical pregnancy and ongoing pregnancy groups (unpublished data); the results indicated that endometrial T-bet/GATA3 ratio has predictive potential for embryo implantation and maintenance of pregnancy in infertile women. The data also suggested that more Th1 and less Th2 might alter the immune environment needed for implantation and pregnancy maintenance. However, a reference range for the endometrial T-bet/GATA3 ratio derived from a large number of fertile women has not been previously established during mid-luteal phase. In the present study, a reliable reference range for the T-bet/GATA3 ratio was established using results from 120 control fertile women, according to the recommendations of the CLSI C28-A3 guidelines, and the reference limit was defined using 95th percentiles. The reference range was subsequently applied in patients with RM.
Maternal-embryonic immune defects were suggested to be an important cause of RM (Garrido-Gimenez and Alijotas-Reig, 2015), and previous investigations in both peripheral and decidual tissue link impaired levels of Th1/Th2 cytokines with RM status (Verma et al., 2019). T-bet and GATA3, as transcription factors that respectively drive the differentiation of Th1 and Th2 cells, were used to evaluate the balance of Th1/Th2 in tissues (Inman et al., 2008; Chen et al., 2012). Jasper et al. (2006) demonstrated no notable difference in endometrial T-bet and GATA3 mRNA levels between controls and patients with primary infertility. Moreover, Abdolmohammadi Vahid et al. (2019) demonstrated an increase in the mRNA expression levels of T-bet, and a decrease in those of GATA3 in the peripheral blood of patients with RM. However, changes in the endometrial T-bet/GATA3 ratio have not been previously examined in patients with RM. Results of the present study revealed that the endometrial T-bet/GATA3 ratio in patients with RM was significantly higher than fertile controls, and there was a Th1-dominant immunity in the uterus of patients with RM, which is consistent with the results of previous studies, suggesting that patients with RM exhibit a Th1 bias (Raghupathy et al., 1999; Ng et al., 2002; Michimata et al., 2003). We found 19.4% of patients with RM were defined as exhibiting an abnormal T-bet/GATA3 ratio using the aforementioned reference range. These data suggested that an abnormal T-bet/GATA3 ratio may contribute to the etiology of RM and highlighted that strategies or treatment options that inhibit pro-inflammatory responses and promote Th2-induced immunity may improve pregnancy outcomes in patients with RM. As the risk of miscarriage showed a strong pattern of recurrence (Magnus et al., 2019), we compared the T-bet/GATA3 ratio in patients with RM with different numbers of miscarriages, and the results revealed the higher T-bet/GATA3 ratio did not change with the number of miscarriages.
In the present study, the Inform Cell Analysis system was used to quantify the percentage of T-bet+ and GATA3+ cells in the endometrium, as described in our previous study (Li et al., 2022). The present study used an average of 20 fields (×200 magnification), rather than whole image. In order to confirm the reliability and accuracy of this counting method, a further standard image analysis platform was used for quantitative tissue analysis, namely, the HALO Image Analysis System (Friedrich et al., 2021; Voabil et al., 2021), which was used to analyze T-bet and GATA3 expression in 20 women. The results exhibited a strong correlation between the two counting methods. Furthermore, two trained pathologists performed interobserver variation analysis in the detection of T-bet and GATA3 percentages using six selected slides. The results showed an acceptable variation, and both exhibited strong correlation, suggesting that the analysis methods used in the present study may provide reasonable estimates of T-bet and GATA3 percentages in IHC slides.
The cellular components of the maternal immune system are under hormonal control (Shah et al., 2018), and endometrial T cells exhibit menstrual cycle-dependent fluctuations (Flynn et al., 2000; Ohta et al., 2013). Inman et al. (2008) previously reported that the mRNA expression levels of T-bet and GATA3 exhibited menstrual cyclic regulation, however, the study lacked quantitative protein expression comparisons. The present study demonstrated a significant downward trend in the T-bet/GATA3 ratio dependent on the menstrual cycle, suggesting that the Th1/Th2 balance was under direct or indirect hormonal control. The T-bet/GATA3 ratio in the luteal phase was significantly decreased, compared with the Pro phase. This may be related to the increased secretion of progesterone by the corpus luteum following ovulation, as progesterone may block Th1 activity and induce the release of Th2 cytokines (Nardo and Sallam, 2006; AbdulHussain et al., 2020). In vitro experiments using human endometrial stromal cells further demonstrated that the gene expression of T-bet and GATA3 changed under hormonal conditions mimicking female physiological conditions (Lu et al., 2013). Implantation of the embryo occurs during the MS phase, which is known as the ‘implantation window’ (Hewitt and Korach, 2011).This peri-implantation period seems to be the earliest period in which the mother can recognize that she is pregnant (Afshar et al., 2007), which can better reflect the state of the endometrium at the moment of embryo implantation. Therefore, the MS phase (LH 7–9) was used for tissue collection in the present study.
Although T-bet and GATA3 were initially described in Th cells (Zheng and Flavell, 1997; Szabo et al., 2000), more advances have demonstrated that T-bet and GATA3 play a much broader role in multiple adaptive and innate lymphocytes, including CTLs, NK cells, and other cells (Lighvani et al., 2001; Miao et al., 2012; Hoyler et al., 2013; Yashiro et al., 2015). However, there are few reports on the expression of T-bet and GATA3 in endometrial immune cells. The present study identified that T-bet and GATA3 were not only expressed in CD4+ Th cells, but also CD8+ CTL and CD56+ NK cells in endometrium. Thus, the T-bet/GATA3 ratio might reflect the balance of immune regulatory signals among multiple immune cells in endometrium. However, the double-staining IHC indicated a significant correlation between T-bet/GATA3 ratio and CD4+T-bet+/CD4+GATA3+. Overall, this study suggested that the T-bet/GATA-3 ratio could provide a useful marker to evaluate the Th1/Th2 status of the uterus during the peri-implantation period.
Based on the establishment of a standardized protocol of IHC and cell counting to obtain the endometrial T-bet/GATA3 ratio, our study is the first to report a reliable reference range of T-bet/GATA3 ratio derived from a large group of 120 fertile women, and speculate that an abnormal T-bet/GATA3 ratio may be one of the etiologic factors of RM, which may be of use in clinical practice in the future. However, the present study has some limitations. Notably, all patients were recruited from a single center. The stability and clinical value of the endometrial T-bet/GATA3 ratio require further investigation, which can be achieved by designing a multi-center clinical trial. Moreover, follow-up of pregnancy outcomes in patients with RM and an abnormal endometrial T-bet/GATA3 ratio are required.
Conclusion
A reference range was established for the T-bet/GATA3 ratio in the peri-implantation endometrium, using the 95th percentile from 120 fertile women. In addition, we demonstrated a significantly increased T-bet/GATA3 ratio in women with RM, and 19.4% of women with RM exhibited an abnormal T-bet/GATA3 ratio. Collectively, these findings highlight the potential application of endometrial T-bet/GATA3 ratio detection in the clinical diagnosis, and perhaps treatment, of patients with RM.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Acknowledgements
The authors are grateful to all women who consented and donated samples for this study.
Authors’ roles
Y.L. and Y.Z. designed and supervised the study. S.Y. conducted the study and write the manuscript. L.D. and R.L. conducted patient recruitment and collected samples. C.C. and X.L. performed experiments. C.H. performed data clean up and data analyses. All authors reviewed and agreed to the final version of the manuscript.
Funding
Shenzhen Fundamental Research Program (JCYJ20180228164631121, JCYJ20190813161203606, JCYJ20220530172817039).
Conflict of interest
The authors declare no conflict of interest.
References
Author notes
Yuye Li and Yong Zeng jointly directed the study.
