Establishment and characterization of a new immortalized human adenomyosis epithelial-like cell line, tAEC21†

This study involving human subjects was performed in accordance with the principles set out in the Declaration of Helsinki. Written informed consent was obtained from the subject. The study was approved by the Institutional Review Board (IRB) of Indiana University (Indianapolis, IN, USA; IRB protocol #1812764043). The subject was a multiparous 43-year-old Latina female who had pathology-proven superficial pelvic endometriosis diagnosed at the time of her previous bilateral tubal ligation. The indication for hysterectomy was abnormal uterine bleeding and chronic pelvic pain, thought to be caused by endometriosis. Endometriotic tissue procurement was the indication for targeted procurement. Targeted tissue procurement and sample deidentification were performed by the Biospecimen Collection and Banking Core (BC 2 , Indiana University Melvin and Bren Simon Comprehensive Cancer Center, Indianapolis, IN, USA) with written informed consent. At the time of the hysterectomy, a fresh endometrial specimen was obtained from the uterine fundus. The patient did not take any medications. At the time of surgery, powder-burn lesions consistent with old endometriosis were biopsied. These biopsies revealed benign fibrous tissue and smooth muscle on final pathology, and no endometriosis was diagnosed in these lesions. The final pathology report of the uterine corpus indicated proliferative-type endometrium and focal adenomyosis. The fallopian tubes contained metallic clips from tubal ligation, and the ecto- and endocervix showed no pathological change. No leiomyomata were noted in the surgical pathology or operative report. Fresh uterine endometrial tissue was transferred to the lab in Hanks Balanced Salt Solution (HBSS; #14175095, Gibco) with 1% P/S. It was mechanically and enzymatically digested with collagenase/hyaluronidase (#07912, StemCell Technologies, Vancouver, BC, Canada) in DMEM/F12 supplemented with 1% Anti-Anti. The dissociated cells were passed through a sterile 40-micron filter. The primary cultures were grown in HES. Cells were tagged with the phycoerythrin (PE)-fluorochrome-conjugated CD326 (EpCam) antibody according to the manufacturer's protocol specifications. Briefly, cultured cells were trypsinized, collected, and reconstituted to 10 6 cells per sample in 98 μl sorting buffer (1% charcoal-dextran FBS + 1x PBS + 2 mM ethylenediamine tetraacetic acid) and supplemented with 2 μl of EpCam-PE antibody (#130-110-999, Miltenyi Biotec, Gaithersburg, MD, USA) or 2 μl of REA isotype control (S)-PE IgG1 antibody (#130-113-438, Miltenyi Biotec). Cells were incubated in a 1:50 antibody/buffer solution for 10 min in the dark at 4°C, washed with 1 ml of fresh buffer, filtered, and counterstained with propidium iodide (#PI304MP, Thermo Fischer Scientific) at a 1:500 concentration in 1 ml of sorting buffer for the exclusion of dead cells. Cells were immediately sorted with the BD FACSAria Fusion cell sorter and BD FACSDiva Software (BD Biosciences, Franklin Lakes, NJ, USA) in the Flow Cytometry Resource Facility (FCRF, Indiana University Melvin and Bren Simon Comprehensive Cancer Center, Indianapolis, IN, USA). Cells were immortalized via retroviral transduction with the TERT using a pLXSN-hTERT vector (kindly provided by Dr. Hari Nakshatri, Indiana University [ 32 ]). For viral production, 2.5 × 10 6 Phoenix-Ampho packaging cells were transfected with 20 μg of pLXSN-hTERT DNA with 25 μl of Lipofectamine-2000 reagent (#11668019, Thermo Fisher Scientific) in Opti-MEM (#31985070, Thermo Fisher Scientific) per the manufacturer’s instructions. After 24 h of incubation at 37°C and 5% CO 2 , the medium was replaced. After 48 h (72 h post-transfection), the virus-containing medium supernatant was collected, centrifuged to remove the debris, filtered through a 0.45 μm filter, and kept at −80°C until use. Primary EpCam+ cell cultures were sub-cultured in six-well plates until 50% confluence and infected with viral supernatant in the presence of 4 μg/ml polybrene (#TR-1003-G, Millipore, Burlington, MA, USA). Transformed cells were selected with 300 μg/ml geneticin (#10131027, Thermo Fisher Scientific). A cryovial of passage two cells in freezing medium [charcoal-dextran FBS + 10% dimethyl sulfoxide (DMSO; #D8418, Sigma-Aldrich)] was sent to IDEXX BioAnalytics (Columbia, MO 65201, USA) for CellCheck16-Human Test (human 16-marker STR profile and interspecies contamination test). Simple karyotypic analysis was completed for early and late passages using KaryoLogic services (Durham, NC 27713, USA). Cells at early and late passages (p.6 and p.53) were prepared according to their specimen preparation and shipping instructions. The tAEC21 cells were seeded in triplicate (25 000 cells/well) in 6-well plates and cultured in various growth media ( Supplementary Table S1 ) for 7 days, with the medium replaced daily. Daily cell imaging was performed with the AMG EVOS FL Imaging System. Cells were plated at 500 cells per well of a 96-well plate with each medium. Proliferation was assessed using CellTiter 96 AQueous One Solution Cell Proliferation Assay (#G3581, Promega, Madison, WI). Absorbance readings were taken at 490 and 630 nm using a Synergy H1 Hybrid Reader (BioTek, Winooski, VT) and Gen5 3.11 software. Background absorbance was subtracted, and data were normalized to day 1 post-seeding. GraphPad Prism software version 9.2.0 (Dotmatics Platform, Boston, MA, USA) was used for analysis, graphing, and calculation of tAEC21 doubling time in each medium type. All experimental conditions were repeated in biological triplicates. One-way ANOVA with Tukey multiple comparison statistical analysis was used to demonstrate a significant difference in the proliferation of tAEC21 cells across different media, based on a 95% confidence interval. Cells were cultured on Lab-Tek II 8-well chamber slides (#154534PK, Thermo Fisher Scientific), fixed with 4% paraformaldehyde (#J61899-AP, Thermo Fisher Scientific) in PBS for 15 min, rinsed with PBS 3 × 3 min, and permeabilized with ice-cold acetone (#A929SK4, Thermo Fisher Scientific) at −20°C × 10 min. Following permeabilization, cells were rinsed with PBS 3 × 3 min at RT, blocked with 5% normal goat serum (#S-10000, Vector Laboratories, Inc., Burlingame, CA) for 1 h at RT, and incubated with primary antibody ( Supplementary Table S1 ) in a blocking solution of 1% BSA (#sc-2323, Santa Cruz Biotechnology, Dallas, TX) with 0.05% Triton X-100 (#T8787, Sigma-Aldrich) in PBS overnight at 4°C. After 3 × 3 min washes with PBS, samples were further incubated with the fluorescent-conjugated secondary antibody at 1:200 dilution, incubated in the blocking solution for 1 h at RT in the dark, and washed again 3 × 3 min with PBS. Slides were further disassembled to remove the wells, DAPI (diamidino-2-phenylindole)-counterstained, and mounted using Fluoromount G (#00–4959-52, Invitrogen) overnight. Imaging was conducted using the Advanced Microscopy Group (AMG; Mill Creek, WA) EVOS FL imaging system ( Supplementary Table S1 ). For monotypic spheroid culture, tAEC21 cells were seeded at 14 000 cells/well in a U-bottom ultra-low adhesion 96-well plate (S-bio, MS-9096 UZ, Hudson, NH) and centrifuged at 2000 rpm × 3 min. For heterotypic spheroid co-culture, tAEC21 and THESC cells were mixed in a 1:1 cell ratio, seeded in U-bottom plates, and centrifuged. 3D cultures were incubated at 37°C, and spheroid assembly was imaged at 24 h using an AMG EVOS FL imaging system. The spheroid diameter was determined using ImageJ software (NIH free download) [ 33 , 34 ]. Statistical analysis was performed using a two-sample unpaired Welch t-test (statistical significance defined as p < 0.05; n ≥ 39) in GraphPad Prism software version 9.2.0. Green 5-chloromethylfluorescein diacetate (CMFDA; #C7025, Invitrogen by Thermo Fisher Scientific) and red 4-([4-(chloromethyl)phenyl]carbonylamino)-2-(1,2,2,4,8,10,10,11-octamethyl-10,11-dihydro-2H-pyrano[3,2-g:5,6-g′]diquinolin-1-ium-6-yl)benzoate (CMTPX; # C34552 , Invitrogen by Thermo Fisher Scientific) CellTrackers were utilized for short-term fluorescent labeling of tAEC21 and THESC cells, respectively, as published previously [ 35 ]. Briefly, each CellTracker was diluted with dimethyl sulfoxide (DMSO; #D8418, Sigma-Aldrich, Saint Louis, MO, USA) to a 2 mM stock concentration. tAEC21 and THESC cell monolayers were grown to 70% confluence, washed with serum-free medium (SFM), and incubated in SFM with green CMFDA (1:500 ratio; 4 μM final dye concentration) or red CMTPX (1:1000 ratio; 2 μM final dye concentration), respectively. After 30 min at 37°C, the CellTracker was removed from the cells. The cells were then incubated for an additional 30 min at 37°C in complete media, and fluorescence was confirmed using the AMG EVOS FL imaging system. To generate 3D heterotypic spheroids, green tAEC21, and red THESC cells were trypsinized, mixed in a 1:1 cell ratio, seeded (14 000 cells/well) in a U-bottom ultra-low adhesion 96-well plate (S-bio, MS-9096 UZ), and centrifuged at 2000 rpm × 3 min. Co-cultures were incubated at 37°C and imaged at 0 h (immediately after centrifugation), 24, 48, and 72 h using an AMG EVOS FL imaging system. Spheroids (three days post-seeding) were collected, briefly rinsed with 1× PBS, fixed in 4% paraformaldehyde (Thermo Fisher Scientific) for 30 min, re-rinsed with 1× PBS, and stored in 70% ethanol for an additional 30 min or until embedding. For embedding, spheroids were placed in 1.5 ml Eppendorf tubes, covered with 300 μl of pre-warmed Epredia HistoGel specimen processing gel (#22110678, Thermo Fisher Scientific), and allowed to solidify for 20 min on ice. The HistoGel plug with trapped spheroids was then gently removed from the tube with a spatula, immediately placed in a tissue embedding cassette in 70% ethanol, and transferred to the Histology Core of the Indiana Center for Musculoskeletal Health (Indiana University School of Medicine, Indianapolis, IN USA) for processing, paraffin embedding, 5 μm sectioning, and hematoxylin and eosin (H&E) staining. Additional unstained sections were sent for immunostaining. Spheroid tissue immunostaining was performed by the Immunohistochemistry Research Core (Indiana University School of Medicine, Indianapolis, IN, USA) on an Agilent Dako Omnis platform (Santa Clara, CA, USA). Briefly, sections on slides underwent deparaffinization, rehydration, heat-induced epitope retrieval (Agilent PT Link), antigen retrieval (Dako EnVision FLEX Target Retrieval Solution, high pH), followed by quenching with 3% H 2 O 2 × 3 min for endogenous peroxidase (HRP) activity elimination. Primary antibody incubation, optional signal amplification (Dako Envision FLEX + Mouse or Rabbit LINKER), detection (Dako Envision FLEX/HRP), 3,3′-diaminobenzidine visualization (Dako EnVision FLEX DAB+ Chromogen), and counterstaining (Dako EnVision FLEX Hematoxylin) were performed as specified in Supplementary Table S1 , followed by dehydration, clearing, and coverslip mounting. Imaging was performed using a Zeiss Axio Lab.A1 Microscope and Labscope software (Carl Zeiss Microscopy LLC; Oberkochen, Germany; Supplementary Table S1 ). 12Z and tAEC21 cells were seeded at 1.5 × 10 5 cells/ml in serum-free media overnight. Cells were treated with 15 ng/ml tumor necrosis factor-alpha (TNF-α; #PHC3011, Gibco) or vehicle (DI water) for 24 h. For two-dimensional (2D) experiments, Ishikawa or tAEC21 cells were seeded at 1.5 × 10 5 cells/ml in complete media overnight. For 3D experiments, tAEC21 cells were first subjected to monotypic spheroid assembly (14 000 cells/spheroid) in U-bottom dishes for 72 h as described above. 2D cell monolayers or 3D spheroids were then treated with 10 nM 17β-estradiol (E2; #E1024 or #E4389, Sigma-Aldrich), or 1 μM progesterone (P4; #P7556 or #P0130, Sigma-Aldrich), or vehicle (DI water/Ethanol) for 24 h. RNA was isolated using the miRNeasy mini kit (#217004, Qiagen) with an on-column RNase-Free DNase Set (#79254, Qiagen) according to the manufacturer’s protocol. RNA concentration and purity were confirmed with NanoDrop ND-1000 (Thermo Fisher Scientific). Using 1 μg total RNA, complementary DNA (cDNA, High-Capacity cDNA Reverse Transcription Master Mix with RNase Inhibitor, Thermo Fischer Scientific) was created ( Supplementary Table S1 ). Samples were diluted to 100 μl, and 2 μl was used for qPCR. qPCR was performed using SYBR Green Universal PCR Master Mix (#4309155, Thermo Fisher Scientific) and gene-specific primers ( Supplementary Table S1 ). Reactions were run on a QuantStudio-3 Real-Time PCR System (Thermo Fisher Scientific). The primer oligonucleotide pairs were obtained from Millipore Sigma (Burlington, MA, USA). The relative quantification (RQ) of mRNA expression was calculated according to the 2 -ΔΔCt method [ 36 ]. All experiments were conducted in technical and biological triplicates; data were analyzed with a Mann–Whitney U-test or unpaired Student t-test as specified in figure legends (statistical significance defined as p < 0.05) using GraphPad Prism version 9.2.0.

After mechanical and enzymatic dissociation of the fresh endometrial tissue, cultured primary cells displayed morphological heterogeneity with secluded islets of epithelial-like polygonal cells surrounded by fibroblast-like stroma ( Figure 1A , left panel). Cell sorting using epithelial cell adhesion molecule (EpCAM), a cell surface marker for epithelial cells, allowed successful separation of epithelial (EpCAM+) and non-epithelial (EpCAM-) cell subpopulations ( Figure 1A , center and right panels; Supplementary Figure S1A ) [ 37 ]. Epithelial-like EpCAM+ cells underwent retroviral transduction of the human telomerase (hTERT) gene, followed by selection with geneticin and cell expansion. The cell line was designated tAEC21 (TERT-immortalized adenomyosis epithelial cells from patient 21) cell line. Via qPCR, hTERT expression in the tAEC21 cells was as high as THESC (365 ± 47-fold expression increase) ( Supplementary Figure S1B ). To date, the tAEC21 cell line has been successfully passaged >80 times with no proliferation decline or loss of epithelial-like morphology in vitro (data not shown). As indicated in the figure legends, early (p.3–p.15) or late (p.53–p.60) passage tAEC21 cells were used. A newly established tAEC21 cell line possesses morphology and marker expression consistent with an epithelial-like phenotype. (A) A heterogeneous primary cell population dissociated from fresh endometrial tissue from a uterus with focal adenomyosis (pre-sort, left panel). Post-sort, EpCam+ (epithelial-like, middle panel) and EpCam- (right panel) cell subpopulations. Micrographs: AMG EVOS FL imaging system; scalebar: 200 μm. (B) Primary AEC21 (p.5–7), early passage hTERT-immortalized tAEC21 (p.5–10), epithelial-like 12Z, and endometrial stromal fibroblast THESC cells were cultured on 8-well chamber slides and evaluated for cytokeratin-7 (KRT7), cytokeratin-5 (KRT5), and CD10 immunofluorescence, nuclei-counterstained with DAPI (blue), and imaged using an AMG EVOS FL imaging system; scalebar: 200 μm. Morphologically, the tAEC21 cells in vitro appeared very epithelial-like ( Figure 1A , center panel). To further examine the molecular epithelial features, we performed immunofluorescent staining of primary cultures and the immortalized tAEC21 line. We used published data from the endometriotic epithelial-like 12Z line to guide our cytokeratin(KRT)-7 marker selection [ 38 ]. Initial immunofluorescent staining of primary cells and the immortalized tAEC21 line showed no KRT7 expression ( Figure 1B , left panel). Subsequent panel expansion to address other commonly explored endometrial epithelial cytokeratins KRT5, KRT8, KRT19, and KRT20 (reviewed in [ 39 ]) detected the expression of a basal epithelial marker KRT5 in both primary and tAEC21 cells ( Figure 1B , center panel). As the most likely contaminating cell type would be stromal cells, we examined the stromal cell marker CD10 using published data from the stromal cell line, THESC [ 38 , 40 ]. Both primary cells and the immortalized tAEC21 line were CD10-negative ( Figure 1B , right panel). The tAEC21 cell line exhibits properties of epithelial cells and not stromal cells. To further characterize the tAEC21 cell line, we examined additional markers. Again, we used published data from the endometriotic epithelial-like cell line 12Z to guide our marker selection [ 25 , 41 ]. In human adenomyosis, neuronal cadherin (Ncad) is expressed in both eutopic and ectopic endometrial tissues at higher levels compared to normal endometrium [ 42 , 43 ]. Additionally, the decrease or loss of epithelial cadherin (Ecad) is observed in both eutopic and ectopic adenomyosis tissues in parallel with increased expression of Ncad [ 43–45 ]. This cadherin switch is associated with the epithelial-to-mesenchymal transition (EMT) and is typically accompanied by upregulated expression of vimentin, a canonical marker of mesenchymal lineages or EMT-reprogrammed epithelial cells with enhanced invasive properties [ 46 ]. Assessed via immunostaining, both primary culture and immortalized tAEC21 cells were strongly positive for junctional Ncad ( Figure 2 ). This Ncad staining was consistent with the staining in 12Z cells ( Figure 2 ), as previously published [ 25 , 47 ]. Primary and immortalized tAEC21 cells were negative for Ecad ( Figure 2 ). Similar to previously reported, 12Z cells were also negative ( Figure 2 ) [ 25 , 47 ]. Further, primary cultures exhibited strong cytoplasmic and predominantly perinuclear vimentin expression in roughly 50% of the cells and moderate vimentin staining in the other half ( Figure 2 ). The tAEC21 cells exhibited less frequent, highly positive cytoplasmic vimentin staining, with most cells displaying weak staining. 12Z cells showed a small portion of strongly positive cells, and the rest of cells were weakly positive for perinuclear vimentin ( Figure 2 ). A non-invasive breast cancer MCF-7 cell line with poor metastatic potential served as an E-cadherin-positive, N-cadherin- and vimentin-negative antibody control ( Figure 2 ) [ 48 ]. Immunostaining of late passage tAEC21 cells (p.58–p.60) confirmed retained expression of Ncad and vimentin, with continued absence of KRT7, Ecad, and CD10 ( Supplementary Figure S1C ). The only notable difference was an almost complete loss of KRT5 expression in late passages, with fewer than 1% of cells remaining positive ( Supplementary Figure S1C ). The tAEC21 cell line exhibits epithelial-mesenchymal transition markers. Primary AEC21 (p.5–7), hTERT-immortalized tAEC21 (p.5–10), epithelial-like 12Z, and MCF-7 (antibody control) cells were cultured on 8-well chamber slides and evaluated for N-cadherin (Ncad), E-cadherin (Ecad), and vimentin (Vim) immunofluorescence, nuclei-counterstained with DAPI (blue), and imaged using an AMG EVOS FL imaging system; scalebar: 200 μm. Short-tandem repeat profiling is a rigorous method of authenticating a new cell line [ 49 ]. We performed CellCheck 16-Human Test (IDEXX) on the tAEC21 cell line. This commercially available service showed that the tAEC21 is a human-derived cell line, without interspecies contamination. IDEXX provided the 16-marker short tandem repeat (STR) profile for tAEC21 ( Supplementary Table S2 ) with a comparison to the DSMZ STR database [ 50 ]. The STR profile of tAEC21 did not match any lines in the database and was not a cross-contaminant or misidentified cell line. Therefore, tAEC21 is a novel, immortalized cell line. Typical karyotyping involves creating a metaphase spread of the chromosomes, aligning each chromosome pair by its characteristic banding pattern, and counting the number of chromosomes in metaphase. A normal human cell would be expected to have 23 pairs or 46 chromosomes [ 51 , 52 ]. We prepared and shipped tAEC21 cells at passages 6 (early) and 53 (late) to KaryoLogic. KaryoLogic could only obtain 10 metaphase spreads from both the early and late passages. Typically, 20 metaphase spreads are examined for a typical karyotype. Data output from KaryoLogic showed that multiple chromosomal rearrangements were observed, making the alignment of pairs impossible. Further, all 10 of the metaphase spreads from early and late passages showed more than 46 chromosomes. The 10-metaphase chromosome counts at early passage were (76, 77, 79, 81, 81, 81, 82, 82, 83, and > 100) and at late passage were (67, 69, 71, 72, 73, 74, 75, 76, 77, and > 100) passages. All of the cells were polyploid. Representative metaphase spread images are displayed ( Supplementary Figure S2 ). The development of a novel cell line requires establishing cell growth patterns and optimizing culture conditions suitable for diverse experimental applications [ 53 ]. The tAEC21 cells were maintained in the most commonly used, commercially available basal cell culture media ( Supplementary Table S1 ) supplemented with 10% FBS and 1% P/S. Cellular morphology and confluency were monitored daily. Microscopically, the tAEC21 cells demonstrated consistent and comparable growth in DMEM/F12 and RPMI-based complete media. In both media types, cells avidly propagated, forming a fully confluent monolayer by day 7 (168 h), while retaining regular epithelial morphology ( Figure 3A , left and middle panels). In contrast, tAEC21 cells cultured in the MEM-based media lost their characteristic polygonal morphology, acquired a spindle-like cell shape, and proliferated more slowly ( Figure 3A , right panel). To examine the effects of specific serum derivations, tAEC21 cells were grown in DMEM/F12 basal medium supplemented with charcoal-dextran-stripped fetal bovine serum (c/d FBS). DMEM/F12 c/d FBS-cultured tAEC21 cells retained their original epithelial morphology and exhibited a steady growth rate, reaching confluency by day 7 ( Supplementary Figure S3A ). The tAEC21 cell line grows in common commercially available media. (A) Early passage (p.6–10) tAEC21 cells were seeded (25,000 cells/well) in a 6-well plate in DMEM/F12, MEM, or RPMI supplemented with 10% FBS and 1% P/S. Cell growth/morphology was monitored for 168 h (7 days) with daily fresh medium exchange (n = 3). Images were taken using AMG EVOS FL imaging system; scale bar: 400 μm. (B) tAEC21 were plated (500 cells/well) in a 96-well plate and subjected to MTS proliferation assay. Data are presented as relative quantification (RQ) with absorbance values normalized to 24 h post-seeding. Statistical analysis (one-way ANOVA with Tukey multiple comparison; ****, p < 0.0001; n = 3) and (C) cell doubling time acquisition (nonlinear fit) were performed in GraphPad Prism version 9.2.0. To more objectively measure cellular proliferation, the CellTiter 96 AQueous One Solution Cell Proliferation Assay (#G3581, Promega, Madison, WI) was used. The highest tAEC21 proliferation rate was found in DMEM/F12-based medium compared to other media types (p < 0.0001, Turkey multiple comparison test) ( Figure 3B ; Supplementary Figure S3 ). DMEM/F12 showed the fastest doubling time at 21.2 h, while MEM-based media was nearly double that doubling time at 46 h ( Figure 3C ; Supplementary Figure S2C ). Adenomyosis is a hormone-dependent disease associated with increased estrogen receptor expression (predominantly estrogen receptor 1, ESR1; also known as ERα) and simultaneous progesterone resistance [ 54–56 ]. Assessing the estrogen and progesterone receptor (PGR) status in a newly established cell line is essential for evaluating its hormonal sensitivity and potential utility in disease modeling. The tAEC21 cells cultured in monolayer showed no ESR1 or PGR expression, and findings mirrored those in the endometriotic epithelial-like 12Z cell monolayer culture ( Figure 4A ). As previously published, MCF-7 breast cancer cells stained strongly positive for both receptors ( Figure 4A ) [ 48 ]. To expand the tAEC21 cell line molecular characterization outside the endometrium, we evaluated the expression of paired box gene 8 (PAX8), a transcriptional factor implicated in embryonic development of Müllerian tissues and known to be expressed in both normal eutopic and ectopic (adenomyotic/endometriotic) epithelial endometrium, as well as Müllerian neoplasms [ 57–59 ]. Separately, the status of AT-rich interactive domain-containing protein 1A (ARID1A), a chromatin remodeler and a tumor suppressor often mutated in deep infiltrating endometriosis and endometriosis-associated malignancies, was assessed [ 60–66 ]. Both tAEC21 and 12Z cell lines displayed high nuclear expression of PAX8 and ARID1A proteins ( Figure 4B ), with endometrial stromal THESC and ovarian clear cell carcinoma TOV21G cells utilized as respective PAX8- and ARID1A-negative controls [ 57 , 67 , 68 ]. All immunostaining assays were also performed using late passage tAEC21 cells, which confirmed the absence of ESR1 and PGR expression and continued presence of PAX8 and ARID1A (up to passage 60) ( Supplementary Figure S3D ). The tAEC21 cell line is negative for estrogen and progesterone receptors. Early passage (p.5–10) tAEC21, 12Z, MCF-7 (positive ERα/PGR antibody control), THESC (negative PAX8 antibody control), TOV21G (negative ARID1A antibody control) cells were subcultured on 8-well chamber slides and processed for (A) estrogen (ESR1/ERα) and progesterone (PGR) receptors, (B) paired box (PAX)8, and AT-rich interaction domain 1A (ARID1A) immunofluorescence, nuclei-counterstained with DAPI (blue), and imaged using an AMG EVOS FL imaging system; scalebar: 200 μm. Abnormal levels of pro-inflammatory mediators, such as tumor necrosis factor (TNF)-α, interleukin (IL)6, C-X-C motif chemokine ligand 8 (CXCL8; also known as IL8), and C-C motif chemokine ligand 2 (CCL2; also known as monocyte-chemoattractant protein 1, MCP-1), are consistently elevated in the adenomyotic uterus, and peritoneal fluids of women with adenomyosis [ 55 , 69 , 70 ]. To determine the initial suitability of a newly created cell line for modeling disease-associated inflammation, we evaluated the tAEC21 line’s sensitivity to TNF-α stimulation. TNF-α-treated tAEC21 cell monolayers showed notably enhanced IL6 (8.8 ± 1.4-fold increase), CXCL8 (11.8 ± 2.1-fold increase), and CCL2 (14.4 ± 1.6-fold increase) gene expression (mean ± SEM, p < 0.0001) ( Figure 5 ). A moderate but statistically significant upregulation of mucin 1 ( MUC1 ), a TNF-α-induced transmembrane mucin constitutively expressed in normal uterine and ectopic endometrium, was observed (3.0 ± 0.2-fold increase, p < 0.0001) ( Figure 5 ) [ 71 , 72 ]. The induced expression pattern was consistent with a previously reported TNF-α-sensitive 12Z cell line, which was a TNF-α response control in current experiments ( Figure 5 ) [ 38 , 41 ]. The tAEC21 cell line elicits an immune response to TNF-α cytokine pressure. Early passage (p.7–12) tAEC21 and 12Z cell monolayers were treated with TNF-α (15 ng/ml) or vehicle (water) for 24 h, and RNA was extracted and processed for qPCR. Data are presented as relative quantification (RQ) of indicated gene transcript expression in TNF-α-treated cells compared to vehicle-treated cells, normalized to β-actin endogenous control (mean ± SEM, n ≥ 3); ****, p < 0.0001, Mann–Whitney U-test. Statistical analysis was performed in GraphPad Prism (version 9.2.0). IL6 —interleukin 6; CXCL8 –C-X-C motif chemokine ligand 8; CCL2 —C-C motif chemokine ligand 2; MUC1 —mucin 1. 2D monolayer cell cultures have been foundational for in vitro disease modeling, offering simplicity and high-throughput screening capabilities. However, 3D culture approaches are more relevant for recapitulating the complex multicell-type tissue environment, providing a more accurate representation of cell–cell and cell-matrix interactions [ 73 ]. To explore 3D dynamics and physiological relevance, tAEC21 cells were grown in 3D. The tAEC21 cells were grown alone as monotypic epithelial spheroids and with THESC as heterotypic spheroids. In 3D, tAEC21 cells formed larger (>1500 μm in diameter), loosely attached multicellular aggregates ( Figure 6A , left and right panels). Co-culture of tAEC21 with THESC endometrial stromal fibroblasts resulted in spatially organized, cohesive spheroid structures of significantly smaller diameter (roughly 500 μm in diameter) ( Figure 6A , center and right panels). Epithelial-like tAEC21 cells and supporting stroma assemble into a biologically relevant pattern in a 3D co-culture model. (A) Early passage tAEC21 (p.10–13) were seeded into spheroids alone (14 000 cells/well) or in co-culture with stromal THESC cells (14 000 cells/well; 1:1 cell ratio) and imaged with the EVOS FL imaging system at 24 h; scalebar: 1000 μm. (B) CMTPX-tagged (red) stromal THESC and CMFDA-tagged (green) tAEC21 cells were co-cultured (14 000 cells/well, 1:1 cell ratio) in a U-bottom 96-well plate, and spheroid conglomeration was monitored with the EVOS FL imaging system at 0-72 h in transmitted, green fluorescent protein (GFP) and red fluorescent protein (RFP) modes. Experiments were performed in triplicate, and representative images are shown; scalebar: 1000 μm. ****p < 0.0001 (two-sample unpaired Welch t-test). Immunofluorescent imaging of live heterotypic 3D spheroids composed of transient fluorophore-tagged epithelial-like, tAEC21 (green, CellTracker) and stromal, THESC (red, CellTracker) over time showed a physiologic response of cells. Over time, the epithelial cells formed a shell around the stromal core. At 0 h, the green epithelial cells were intermingled with the red stromal cells. By 24 h, a green epithelial shell was forming around the red stromal core. Within 48-72 h, the green epithelial shell and the red stromal core were well-defined ( Figure 6B ). To evaluate further, monotypic tAEC21 spheroids and heterotypic tEAC21 with THESC spheroids were fixed, embedded, and underwent molecular immunohistochemistry (IHC). During histological processing, loose monotypic tAEC21 multicellular aggregates failed to maintain structural integrity and were disassembled into unorganized clusters ( Figure 7A and B ). Heterotypic spheroids maintained structural integrity and were tightly compacted. Hematoxylin and eosin-stained (H&E) sections from heterotypic spheroids showed that the epithelial-like tAEC21 component (evident by round morphology, lightly eosinophilic cytoplasm, and basophilic nuclei) populated predominantly along the periphery around the stromal (spindle-shaped, eosinophilic cells with elongated nuclei) THESC supportive background. Immunostaining confirmed cytokeratin-positive and CD10-negative tAEC21 localization in a spheroid shell ( Figure 7B ). Some epithelial cells are interspersed among the stroma, exhibiting incomplete cell type segregation ( Figure 7A and B ). In alignment with endometrial epithelium, as monotypic spheroids, tAEC21 cells were strongly positive in nearly all cells for hepatocyte nuclear factor 1 beta (NHF1β) and constituted an NHF1β-positive outer shell in the heterotypic co-cultures ( Figure 7B ) [ 74 ]. Further, tAEC21 immunostaining for Napsin A, a marker for endometrial-associated malignancies, detected no expression ( Figure 7B ) [ 75 ]. Importantly, in both monotypic and heterotypic 3D tissue structures, tAEC21 cells expressed ESR1 but not PGR ( Figure 7B ). All monotypic and heterotypic 3D culture H&E and IHC experiments were repeated using late-passage tAEC21 cells, and the results were consistent with those obtained from early-passage cells ( Supplementary Figure S4 ). tAEC21/stromal 3D co-cultures exhibit histology and molecular expression patterns consistent with endometrial-like tissue. (A) Early passage tAEC21 (p.12–15) were seeded into spheroids alone or in a mixture with stromal THESC cells (1:1 cell ratio); assembled 3D spheroids were harvested at 72 h, fixed, paraffin-embedded, sectioned, and stained with H&E. Scalebar: 50 μm. (B) 3D spheroid tissue sections were subjected to immunohistochemistry and visualized with 3,3′-diaminobenzidine (brown, positive) using a Zeiss Axio Lab.A1 Microscope. Positive (+) control tissues: skin (keratin; KRT+), tonsil (CD10+), renal cell carcinoma (hepatocyte nuclear factor 1 homeobox B; HNF1β+), lung adenocarcinoma (Napsin A+), breast (estrogen receptor; ERα+), breast (progesterone receptor; PGR+). Scalebar: 50 μm. Given the importance of steroid hormones in adenomyosis [ 54–56 ], we evaluated the expression of the steroid hormone receptors, ESR1 and PGR, in 2D and 3D cultures of tAEC21 compared to monolayer cultures of Ishikawa, a known ESR1 + /PGR + cell line [ 22 ]. In 2D culture, tAEC21 cells displayed markedly low ESR1 expression (2.8-fold lower than Ishikawa). This low ESR1 transcript expression is consistent with the undetectable ESR1 protein expression ( Figure 4A , Supplementary Figure S3D ). In contrast, when grown in 3D culture, tAEC21 cells demonstrated robust ESR1 expression, roughly an 8.5-fold increase compared to the tAEC21 2D monolayer and 3-fold higher than Ishikawa ( Figure 8A , left panel). This high ESR1 transcript expression is consistent with the high ESR1 protein expression in 3D culture ( Figure 7B , Supplementary Figure S4B ). When cultured in 2D or 3D, tAEC21 exhibited very low PGR expression compared to Ishikawa cells, with tAEC21 3D cultures exhibiting nearly undetectable PGR expression ( Figure 8A , right panel). This low PGR transcript expression is consistent with the undetectable PGR protein in the 2D and 3D cultures ( Figure 4A , Figure 7B , Supplementary Figure S3D , Supplementary Figure S4B ). Steroid hormone receptor expression and steroid hormone responsiveness of tAEC21 cells in 2D and 3D culture. (A) Ishikawa ( ESR1 and PGR positive control) and early passage (p.6–11) tAEC21 cell monolayers or tAEC21 3D monotypic spheroids were subjected to RNA isolation and qPCR to assess the status of hormone receptors. Data are presented as relative quantification (RQ) of indicated gene transcript expression in tAEC21 cells compared to Ishikawa, normalized to β-actin endogenous control (mean ± SEM, n ≥ 3). (B-C) tAEC21 cell monolayers or tAEC21 3D monotypic spheroids were treated with 17β-estradiol (E2, 10 nM), progesterone (P4, 1 μM), or vehicle (water/ethanol) for 24 h, and RNA was extracted and processed for qPCR. Data are presented as relative quantification (RQ) of indicated gene transcript expression in hormone-treated cells compared to vehicle-treated 2D cells, normalized to β-actin endogenous control (mean ± SEM, n ≥ 3). Statistical analysis was performed in GraphPad Prism (version 10.6.0); ns, non-significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001, unpaired Student t test. ESR1 – estrogen receptor 1; PGR —progesterone receptor; SCGB1D2 —secretoglobin family 1D member 2; IHH —Indian hedgehog signaling molecule; FOXO1 —forkhead box O1; MUC1 —mucin 1; EGFR —epidermal growth factor receptor. To evaluate steroid hormone molecular response, we utilized a set of epithelial endometrium-relevant estrogen-responsive [( PGR ; secretoglobin family 1D member 2 ( SCGB1D2 ))] and progesterone-responsive [Indian hedgehog signaling molecule ( IHH ), forkhead box O1 ( FOXO1 ), MUC1 , and epidermal growth factor receptor ( EGFR )] genes [ 76–78 ]. Following 17β-estradiol treatment, 2D cultures of tAEC21 responded appropriately with a 10-fold increase in PGR expression. However, 17β-estradiol was not sufficient to induce PGR expression significantly in the 3D cultures of tAEC21, which have undetectable PGR at baseline ( Figure 8B , left panel). Both 2D and 3D tAEC21 cultures responded with robust increases in SCGB1D2 expression ( Figure 8B , right panel). While there was a 15-fold increase in IHH expression in 3D cultures compared to 2D cultures of tAEC21, there was no significant increase in IHH with progesterone treatment in either 2D or 3D cultures ( Figure 8C ). Similarly, there was a 2-fold increase in FOXO1 in 3D compared to 2D cultures. There was a nearly 2-fold reduction in FOXO1 expression in the 2D cultures with progesterone treatment ( Figure 8C ). There was a 15-fold higher expression of MUC1 in 3D cultures compared to 2D cultures. There was a statistically significant but slight decrease in MUC1 expression in the 2D cultures treated with progesterone ( Figure 8C ). For EGFR expression, there was a modest increase in expression in 3D cultures compared to 2D. Only in 2D cultures was there a significant decrease in EGFR expression with progesterone treatment ( Figure 8C ). The 3D cultures did not respond to progesterone treatment in the limited subset of genes examined.

The 12Z cell line was obtained from Applied Biological Materials, Inc. (ABM; #T0764, Vancouver, Canada) and maintained in Dulbecco Modified Eagle Medium: Nutrient Mixture F12 (DMEM/F12; #11330-032, Thermo Fisher Scientific, Waltham, MA, USA), supplemented with 10% fetal bovine serum (FBS; #S12450, Atlanta Biologicals, Flowery Branch, GA, USA) and 1% penicillin/streptomycin (P/S; #15140-122, Thermo Fisher Scientific) [ 25 ]. The THESC cell line was purchased from American Type Culture Collection (ATCC; #CRL-4003, Manassas, VA, USA) and maintained in phenol red(−) DMEM/F12 medium (#21041-025, Thermo Fisher Scientific), supplemented with 10% charcoal-dextran-treated FBS (# S11650 , R&D Systems, Minneapolis, MN, USA), 1% ITS + Premix Universal Culture Supplement (containing human recombinant insulin, human transferrin, selenous acid, bovine serum albumin, and linoleic acid) (#354352, Corning, Durham, NC, USA), and 500 ng/ml puromycin (#J67236-XF, Thermo Fisher Scientific) [ 28 ]. The tAEC21 cells were cultured in human epithelial-stromal [HES, phenol red(−) DMEM/F12 medium +10% charcoal-dextran treated FBS +1% antibiotic-antimycotic (Anti-Anti; #15240062, Gibco) + 0.075% sodium bicarbonate (#25080094, Gibco)]. The TOV21G cell line was purchased from ATCC (#CRL-3577) and was maintained in a 1:1 ratio of Medium-199 (#M4530, Sigma-Aldrich, St. Louis, MO, USA) and MCBD-105 (#M6395, Sigma-Aldrich), supplemented with 15% FBS and 1% P/S [ 29 ]. The MCF-7 cell line was purchased from ATCC (#HTB-22) and was maintained in RPMI-1640 (#11875093, Gibco) supplemented with 10% FBS, 1% P/S, and 1% Glutamax (#35050061, Gibco) [ 30 ]. The Ishikawa cell line was obtained from the Cytogenetics and Cell Authentication Core at MD Anderson Cancer Center (Houston, TX) and maintained in RPMI-1640 with 10% FBS and 1% P/S [ 22 ]. The Phoenix-Ampho cell line was purchased from ATCC (#CRL-3213) and was maintained in DMEM (#11965092, Gibco), supplemented with 10% FBS and 1% P/S [ 31 ]. All cell lines were routinely passaged and maintained in a humidified incubator at 37°C and 5% CO2. Cell line authentication was confirmed using CellCheck 9 Plus (IDEXX BioAnalytics, Westbrook, ME, USA). All cell lines were free from mycoplasma contamination (MycoAlert Mycoplasma Detection Kit, #LT07-318, Lonza, Basel, Switzerland).

The etiopathogenesis of adenomyosis remains obscure. Leading theories include: (a) repetitive microtraumas and disruption of the junctional zone (JZ) between the endometrium and the myometrium, followed by invagination of basal endometrial cells into the muscular layer with their subsequent migration and proliferation; (b) development of intramyometrial endometrial-like lesions from displaced Müllerian system embryonic pluripotent residues; (c) aberrant migration and differentiation of adult endometrial epithelial and stromal progenitor cells, relocated from the endometrium basalis stem cell niches through the disrupted JZ; (d) “outside-to-inside” theory, wherein transcoelomic endothelial cells, mislaid after retrograde menstruation or migrating from ectopic endometriosis lesions, penetrate inward through uterine serosa into the myometrium to establish endometrial-like implants [ 45 , 79 , 80 ]. Intramyometrial adenomyotic lesion formation, occurring through one or more of the proposed mechanisms, is exacerbated by a dysregulated steroid hormone background. This is propagated by an imbalance between elevated estrogen receptor levels and estrogen response, and the loss of PGR and/or PGR activity [ 55 , 56 ]. Increased estrogen exposure and persistent tissue injury and repair promote chronic local and systemic inflammatory response and neoangiogenesis, facilitating disease progression [ 80–82 ]. Importantly, adenomyosis shares overlapping pathophysiology with endometriosis, as both conditions are characterized by ectopic growth of histologically similar endometrial-like lesions, possibly arising from the same origin (uterine endometrium), harbor common molecular abnormalities (in particular, KRAS mutations and various epigenetic and immune alterations), possess estrogen dependence and progesterone resistance, share clinical presentation, and are often concurrent [ 83–85 ]. Multiple next-generation sequencing studies have identified frequent KRAS mutations in adenomyosis epithelial cells and adjacent basalis endometrial glands, supporting the invagination theory of pathogenesis [ 86 ]. These mutations are also observed in endometriosis, suggesting a similar disease process in both conditions [ 86 ]. Our newly created hTERT-immortalized adenomyosis-derived epithelial-like cell line tAEC21 exhibits features of epithelial cells, including polygonal morphology, tight packing during monolayer growth, and cell surface expression of the epithelial marker EpCam ( Figure 1A ; Supplementary Figure S1A ) [ 37 ]. The cytoplasmic expression of cytokeratin and the absence of CD10 confirm tAEC21 epithelial-like status ( Figure 1B ) [ 39 , 40 ]. The cells also stain strongly positive for nuclear PAX8 and nuclear HNF1β, markers exclusive to endometrial glandular epithelium, and not stroma, confirming their epithelial origin from the endometrium ( Figure 4B ; Figure 7B ) [ 57–59 , 74 ]. In addition to epithelial characteristics, the tAEC21 cell line possesses hallmarks of EMT including complete loss of Ecad, strong junctional Ncad expression, and strong-to-moderate cytoplasmic vimentin expression ( Figure 2 ). In human normal proliferative-phase endometrium, Ncad is present in 11%–20% of epithelial cells and highlights a subset of cells with high self-renewal capacity and proliferative potential, serving endometrial regeneration [ 44 , 87 ]. Ncad-positive cells are predominantly localized in the endometrium basalis, adjacent to the myometrium, and can differentiate into gland-like structures in vitro [ 87 ]. Ncad expression has also been linked to increased migration and invasiveness in endometrial-related pathologies through the mechanisms of EMT [ 22 , 39 ]. On the other hand, Ecad is a key epithelial cell marker that maintains the non-invasive epithelial cell phenotype, epithelial cell–cell adhesion, and layer integrity (reviewed in [ 88 ]). Loss of Ecad expression contributes to the EMT shift and the acquisition of a more invasive, mesenchymal cell phenotype [ 88–90 ]. Immunohistochemical studies demonstrate evident membranous Ecad staining in human normal endometrial epithelium, with strong expression consistent across the proliferative and secretory phases of the menstrual cycle [ 42 , 44 ]. Immunostaining data on vimentin expression in human normal endometrial epithelium vary. Dabbs et al. reported perinuclear vimentin staining in normal human proliferative endometrial glands and vimentin absence in the secretory phase [ 91 ]. Nonwitz et al. showed consistent co-expression of vimentin and cytokeratin in surface and glandular epithelial cells across all menstrual cycle phases [ 92 ]. Cyclic expression of epithelial vimentin, predominantly in the proliferative phase with highly variable staining intensity, was also reported in both eutopic and ectopic endometrial tissues of women with adenomyosis [ 93 ]. On the contrary, immunostaining performed by Zhou et al. showed low epithelial vimentin in normal uterine endometrial glands and increased expression in eutopic endometrium and, more significantly, ectopic lesions from adenomyosis patients [ 43 ]. In alignment with human data, EMT events are observed in murine models of adenomyosis [ 94 ]. In genetically engineered mice with constitutive activation of uterine β-catenin, epithelial cells of adenomyosis lesions showed Ecad suppression together with expression of vimentin and other EMT-related markers compared to control mice uteri [ 15 ]. Similarly, decreased Ecad and increased vimentin immunostaining were observed in ectopic adenomyotic tissues of tamoxifen-induced adenomyosis mice [ 95 ]. The described molecular switch is observed in the tAEC21 cells and suggests acquisition of EMT-related cell properties by the cell line, such as increased motility, invasiveness, and altered adhesion [ 25 , 35 , 44 , 46 , 90 , 96 ]. This invasive behavior may explain why epithelial-origin primary cells were initially able to attach to the tissue culture plastic within the pool of stromal cells and were ultimately isolated. Typically, endometrial epithelial cells require extracellular matrix coatings, such as Matrigel, on culture dishes to attach and grow effectively, as this mimics their dependence on the basement membrane for in vivo adhesion [ 97 , 98 ]. Epithelial-like cells that have transitioned through EMT often gain enhanced cell-substrate adhesion and invasiveness, similar to mesenchymal cells and metastatic carcinoma cells, allowing them to attach more easily to rigid surfaces like uncoated tissue culture plastic [ 25 ]. An unexpected finding in the tAEC21 cell line characterization is its abnormal polyploid karyotype with multiple chromosomal rearrangements ( Supplementary Figure S2 ). Limited studies involving the one published adenomyosis-derived cell line and patient-derived organoids (from three different adenomyosis patients) report a normal 46, XX karyotype [ 27 , 99 ]. Nevertheless, chromosomal alterations have been investigated in the context of adenomyosis with the goal of discerning contributing genetic factors. Comparative genomic hybridization (CGH) analysis of frozen tissues from 25 cases of pathologically proven adenomyosis did not detect recurrent chromosomal gains or losses, suggesting that large chromosomal changes may be rare [ 100 ]. Further, research by Chao et al. suggests a strong link between chromosomal instability and the pathogenesis of adenomyosis [ 101 ]. A whole exome and RNA sequencing study identified polyploidy (whole genome doubling, 4n) in 19 out of 57 human adenomyosis tissue biopsies, which correlated with early disease onset and higher sensitivity to estrogen treatment [ 101 ]. Alternatively, a polyploid karyotype may suggest a metaplastic potential of our adenomyosis-derived cells. Metaplasia of displaced embryonic Müllerian remnants or stem cells is one of the theories of adenomyosis development [ 102 , 103 ]. Karyotype abnormalities, including aneuploidy and non-random chromosomal breaks, are also found in adenomyosis patient-derived endometrial mesenchymal stem cells [ 104 ]. Interestingly, the established tAEC21 cell line expresses cytokeratin (KRT)-5 ( Figure 1B ), a reported progenitor or stem cell marker, particularly in glandular reproductive and mammary tissues, orchestrating cell maturation into luminal or myoepithelium [ 105 , 106 ]. Immunohistochemical analysis by Stefansson et al. demonstrates low, focal expression of KRT5 in normal endometrium as opposed to its enhanced presence in endometrial hyperplasia and squamous metaplasia, which in turn is linked to endometrial epithelial malignancies [ 107 , 108 ]. KRT5 positivity is associated with the loss of E-cadherin expression and suppression of ESR1 and PGR, which is consistent with our observations in tAEC21 [ 109 , 110 ]. Furthermore, in reproductive cancers, KRT5+ cell percentage increases following cisplatin treatment, suggesting chemotherapy-induced enrichment of cancer stem cell population [ 105 ]. Notably, the possibility that the polyploid karyotype observed in endometrial epithelial cells represents an artifact of the immortalization process cannot be excluded. Similar karyotypic abnormalities, along with morphological changes and EMT-like features, have been reported in other immortalized epithelial cell lines of various tissue sources [ 20 , 111–114 ]. An earlier study that followed long-term human normal endometrial fibroblastic and epithelial cultures reported quick acquisition of polyploidy by epithelial cultures simply as a result of in vitro culture [ 115 ]. In our work, we chose hTERT-induced immortalization due to its reported advantages in reducing karyotypic changes, maintaining relatively higher genomic stability, and imparting characteristic phenotypic properties to the cells [ 116 , 117 ]. It is important to note that the morphological and EMT marker expression profile observed in tAEC21 cells closely mirrors that seen in the original primary, non-immortalized AEC21 adenomyotic epithelial cell cultures ( Figure 1B and 2 ). This consistency strongly suggests that the EMT-like phenotype is truly intrinsic to the original cells and was retained rather than induced by the immortalization process. Altered response to high cytokine pressure contributes to adenomyosis symptomatology and progression, and it can be a promising target for therapeutic intervention. Consistent with the biological properties of the disease, tAEC21 cells exhibit a significant increase in the production of multiple immune response mediators upon treatment with TNF-α ( Figure 5 ), a key pro-inflammatory cytokine implicated in adenomyosis [ 69 ]. This response is characterized by elevated expression of IL6 and CXCL8 (IL8), cytokines known to amplify inflammatory signals and recruit immune cells, both of which are found at higher levels in both eutopic and ectopic endometrial tissues of adenomyosis patients [ 69 , 118–120 ]. TNF-α stimulation of TAEC21 cells also led to a marked upregulation of CCL2 , a chemokine critical for monocyte recruitment and subsequent macrophage infiltration, which is also elevated in adenomyosis tissues [ 69 , 118 , 119 ]. Macrophages are highly enriched in the eutopic and ectopic endometrium of adenomyosis patients, and their dysregulated activity, largely attributed to M2 polarization, facilitates EMT of endometrial epithelium, enhances invasiveness of adenomyotic lesions, and contributes to disease-associated infertility through the disruption of immune tolerance and subsequent embryo implantation failure (reviewed in [ 121 ]). Finally, TNF-α-stimulated tAEC21 cells increased expression of MUC1 , a glycoprotein that is normally expressed on apical surfaces of uterine epithelial cells, forming a protective barrier, but depending on the context, may promote or suppress inflammation, result in adaptive anti-MUC1 immunity, and alter the state of immune tolerance necessary for successful implantation [ 122–125 ]. Observed findings represent the initial characterization of the tAEC21 cell line, and future studies will investigate a broader cytokine/chemokine panel to establish this cell line’s inflammatory profile. 3D cell culture systems have become increasingly valuable in mimicking the natural architecture of tissues and organs, faithfully representing cell–cell and cell-matrix interactions, and providing more accurate drug responses than cell monolayers. In this work, we screened tAEC21 cells for their propensity to self-organize and interact with endometrial stromal fibroblasts, confirming the spontaneous formation of physiologically relevant 3D spheroids with an epithelial tAEC21 shell surrounding a stromal core ( Figure 6 ). Our observations mirrored results by Song et al., who presented similar spatiotemporal organization and epithelial-stromal interaction between epithelial-like 12Z cells and uterine or endometriotic stromal cells in an elegant spheroid model for endometriotic lesions [ 126 ]. In the absence of established adenomyosis cell lines, in vitro adenomyosis tissue 3D modeling relies on patient-derived organoid technology, often coupled with synthetic biomaterial matrices and microfluidic platforms (comprehensively reviewed by Gnecco and colleagues) [ 127 , 128 ]. Our described tAEC21/THESC epithelial/stromal spheroid model presents a valuable alternative to adenomyosis-derived organoids, offering greater simplicity, faster generation, and cost-efficiency without the need for post-hysterectomy tissue explants and artificial scaffolds. Given our experience with the model, we predict that spheroid size and cellular complexity can be easily manipulated by incorporating various cell subpopulations and that straightforward drug testing will be possible in the future. Immunolocalization studies report strong nuclear ESR1 expression in human adenomyotic glandular epithelium and stroma with varying levels across menstrual cycle phases [ 129 , 130 ]. PGR expression reports are controversial. Some studies confirm its nuclear-cytoplasmic presence in the glandular epithelial and stromal cells of human adenomyosis tissue, similar to that in normal myometrium [ 129 ]. Other studies show loss of PGR expression in adenomyotic samples, especially stroma basalis and myometrium [ 130 ]. Our initial assessment of 2D-cultured tAEC21 hormone receptor status showed undetectable levels of both ESR1 (ERα) and PGR via immunostaining. Importantly, our work showed a gain of ERα expression by tAEC21 under 3D culture conditions only ( Figure 4A ; Figure 7 and 8 ). This observation suggests that the 3D context of cells fosters spatial cell–cell interactions and activates signaling pathways conducive to ESR1 expression [ 131 , 132 ]. The lack of PGR expression in tAEC21 cells was maintained under 2D and 3D culture conditions ( Figure 4A ; Figure 7 and 8 ). Our results also demonstrate that 3D-cultured tAEC21 cells adopt an ESR1 + /PGR − phenotype, which is estrogen-responsive but progesterone-unresponsive, reflecting a dysregulated hormonal signaling pattern that closely resembles that observed in adenomyosis tissue ( Figure 8 ). Interestingly, while not explicitly shown in adenomyosis, high Ncad expression negatively correlates with PGR expression in endometriotic ectopic lesions and the endometriotic 12Z cell line [ 47 ]. Furthermore, in a tamoxifen-induced adenomyosis mouse model, the progression of ectopic lesions and an increase in EMT markers coincided with a progressive loss of PGR in immunostained adenomyotic glandular epithelium [ 95 ]. Future detailed profiling of all hormone receptor subtypes and downstream target genes is required. In summary, our novel endometrial-like epithelial cell line derived from a patient with pathologically-confirmed adenomyosis and a remote history of endometriosis exhibits characteristics of invasiveness, karyotypic abnormalities, accurately mimics immune responses observed in vivo, faithfully recapitulates physiologically relevant spatial rearrangement with supporting stroma in a heterotypic epithelial/stromal 3D spheroid model, and bears a conditional ESR +/− and a consistent PGR − status accompanied by avid estradiol responsiveness and progesterone resistance. The adequate growth rate in multiple commonly used commercially available media, supplemented with standard or charcoal/dextran-stripped FBS, makes the tAEC21 cell line a flexible, inexhaustible, and time- and cost-effective in vitro model. This model additionally reduces media-related experimental constraints and removes ethical concerns associated with human and animal tissue use. It can serve as a valuable tool for exploring the underlying mechanisms of adenomyosis/endometriosis and associated complications, as well as testing new diagnostic and treatment strategies.

Adenomyosis is a common, non-malignant uterine condition in which endometrial glands and stroma are present within the uterine myometrium. Clinical symptoms of adenomyosis include abnormal uterine bleeding, pelvic pain, and infertility. These symptoms are similar to those of other common gynecologic diseases, including endometrial cancer, endometriosis, primary dysmenorrhea, and uterine leiomyomata. For example, in the evaluation of abnormal uterine bleeding using the PALM-COEIN strategy, adenomyosis is a common structural cause of heavy menstrual and intermenstrual bleeding along with Polyps, Adenomyosis, Leiomyomata, and uterine Malignancy and endometrial hyperplasia [ 1 ]. Patients with adenomyosis may have varying degrees of pelvic pain, ranging from intermittent mild pelvic discomfort or abdominal bloating to chronic pelvic pain (defined as more than six months of pelvic pain that affects daily activities), dysmenorrhea (defined as painful menstruation), dyspareunia (defined as painful sexual intercourse), dyschezia (defined as painful defecation), and dysuria (defined as painful urination). Up to 30% of patients present with no symptoms and are diagnosed histologically during hysterectomy for other gynecologic indications [ 2–5 ]. Other gynecological comorbidities with similar symptomatology frequently co-occur with adenomyosis. In particular, there is up to 90% concurrence with endometriosis and up to 57% concurrence with uterine leiomyomata [ 6–8 ]. The definitive pathological criteria for adenomyosis include the presence of endometrial glands and stroma within the myometrium, extending at least 2.5 mm deep from the endometrium-myometrium junction or deeper than 25% of the total myometrial thickness [ 9 ]. Histologically, it is divided into diffuse (i.e. multiple ectopic endometrial foci with undefined borders widely scattered throughout the myometrium) or focal (i.e. isolated endometrial gland nodules, localized and well-circumscribed by myometrium) [ 10 ]. Furthermore, adenomyosis affects pregnancy rates and obstetrical outcomes. In patients thought to have adenomyosis by imaging findings at the time of infertility evaluation, adenomyosis is associated with lower embryo implantation rates, reduced pregnancy rates, and higher miscarriage rates. The hypothesized pathophysiology involves a hostile uterine environment, impaired adhesion molecule expression, and local inflammation [ 11 , 12 ]. Data from mouse models of adenomyosis support this pathophysiology [ 13–15 ]. Previous work has shown that the endometrium from uteri with adenomyosis exhibits a distinct molecular profile that may be responsible for uterine dysfunction [ 16 , 17 ]. A significant hurdle in adenomyosis research is the scarcity of representative disease models. While patient samples from surgeries are valuable for histological analyses, biomarker discovery, and validating diagnostic tools, they are unsuitable for studying cell behavior, disease progression mechanisms, or testing therapeutic interventions. In that regard, cultured primary epithelial and stromal cells from endometrial or adenomyosis biopsies are preferable. However, primary cultures are challenging to isolate, purify, and culture, and they have limited scalability compared to immortalized cell lines [ 18 , 19 ]. Additionally, in vitro propagation and immortalization of benign endometrial epithelial cells have been tricky [ 19–21 ]. In the absence of available immortalized adenomyosis-derived cell lines, researchers mimic endometrium with classic epithelial features using Ishikawa, an epithelial-type human endometrial adenocarcinoma cell line positive for the epithelial marker, E-cadherin, and low/negative for mesenchymal markers of invasiveness, N-cadherin and vimentin [ 22–24 ]. To model ectopic lesions, the endometriotic epithelial-like 12Z cell line, derived from a red peritoneal endometriotic lesion and expressing invasive markers, including N-cadherin and vimentin, is commonly employed [ 25 , 26 ]. No adenomyosis-derived epithelial cell lines existed until recently. The only published immortalized human adenomyosis ectopic cell line (ihAMEC) was created by simian vacuolating virus 40 (SV40) lentiviral infection and is not commercially available or widely shared [ 27 ]. Our work presents a novel adenomyosis-derived epithelial-like cell line immortalized via retroviral transduction of a human telomerase reverse transcriptase ( TERT ) gene. This new cell line retains epithelial morphology, expresses both epithelial and mesenchymal markers, indicating epithelial origin and acquired invasive properties, and exhibits a steroid hormone receptor profile and response characteristic of adenomyosis. The cell line is sensitive to immune pressure and retains biologically relevant spatial architecture in a three-dimensional (3D) co-culture model. This cell line offers an additional in vitro research resource to explore the mechanisms of adenomyosis.