Amniotic fluid modifies esophageal epithelium differentiation and inflammatory responses

Abstract

The interplay between genetic and environmental factors during pregnancy can predispose to inflammatory diseases postnatally, including eosinophilic esophagitis (EoE), a chronic allergic disease triggered by food. Herein, we examined the effects of amniotic fluid (AF) on esophageal epithelial differentiation and responsiveness to proallergic stimuli. Multiplex analysis of AF revealed the expression of 66 cytokines, whereas five cytokines including IL-4 and thymic stromal lymphopoietin (TSLP) were not detected. Several proinflammatory cytokines including TNFα and IL-12 were highly expressed in the AF from women who underwent preterm birth, whereas EGF was the highest in term birth samples. Exposure of esophageal epithelial cells to AF resulted in transient phosphorylation of ERK1/2 and the transcription of early response genes, highlighting the direct impact of AF on esophageal epithelial cells. In a three-dimensional spheroid model, AF modified the esophageal epithelial differentiation program and enhanced the transcription of IL-13-target genes, including CCL26 and CAPN14, which encodes for a major genetic susceptibility locus for eosinophilic esophagitis. Notably, CAPN14 exhibited upregulation in spheroids exposed to preterm but not term AF following differentiation. Collectively, our findings call attention to the role of AF as a potential mediator of the intrauterine environment that influences subsequent esophageal disorders.

NEW & NOTEWORTHY The interaction between amniotic fluid and the esophageal epithelium during pregnancy modifies esophageal epithelial differentiation and subsequent responsiveness to inflammatory stimuli, including interleukin 13 (IL-13). This interaction may predispose individuals to inflammatory conditions of the esophagus, such as eosinophilic esophagitis (EoE), in later stages of life.

Listen to this article’s corresponding podcast at https://ajpgi.podbean.com/e/got-guts-the-micro-version-effect-of-amniotic-fluid-on-the-esophageal-epithelium/.

INTRODUCTION

Human amniotic fluid (AF) is a complex biological fluid enveloping the fetus, offering mechanical protection, nutrients, and metabolites that are essential for fetal development. Comprising water, proteins, peptides, carbohydrates, hormones, lipids, and uric acid, amniotic fluid is a vital component in the gestational environment. Enteral uptake of amniotic fluid is a major predictor of fetal growth during late gestation, especially of the gastrointestinal tract, including the small intestine and the stomach (1–3).

After midgestation in humans, the main sources of amniotic fluid are fetal urinary excretion and the secretion of fluids from the oral, nasal, tracheal, and pulmonary sources, whereas the elimination of amniotic fluid occurs primarily through fetal swallowing (4). The fetal membranes also contribute to amniotic fluid exchange. On the basis of its proximal location in the gastrointestinal tract, the esophagus is exposed in utero to amniotic fluid. Indeed, human and monkey amniotic fluid contains esophageal-derived proteins at levels that positively correlate with gestational age (5). Nonetheless, the impact of amniotic fluid on the esophageal epithelium has not been examined. Deciphering the impact of amniotic fluid on the esophagus may be particularly important in the emerging allergic disease eosinophilic esophagitis (EoE), which has been linked with intrauterine exposure to maternal fever and antibiotic usage, delivery by caesarian section, and co-occurrence of esophageal atresia (6–8). Prior studies have shown the effects of amniotic fluid on fibroblasts and keratinocytes from the skin and cornea (9–12), but effects on esophageal epithelium have not been reported.

EoE is a developing, chronic, allergic disease triggered by food, distinguished by significant eosinophilia specifically in the esophagus, and linked to esophageal dysfunction, such as dysphagia in adults (13). The development of EoE involves an intricate interplay of genetic, environmental, and immunologic elements. Twin studies support a genetic link, as monozygotic twins have a 40% concordance and exhibit a twofold increase in disease concordance compared with that of dizygotic twins. But what is even more striking was a 10-fold increase in disease concordance in dizygotic twins compared with nontwin siblings, strongly supporting a priming role for in utero exposures (14).

Dysregulation of genes specific to the esophagus constitutes a crucial aspect of EoE pathogenesis. For example, several genetic and functional studies have linked the expression of esophageal-specific intracellular cysteine protease calpain 14 (CAPN14) to the increased risk for EoE, including very early onset of EoE (15). Notably, CAPN14 is encoded by a chief EoE susceptibility locus (16). In the esophagus, CAPN14 disrupts the integrity of the esophageal epithelium, impairing its barrier function through a mechanism associated with the downregulation of the desmosome component desmoglein-1 (DSG1) (16). Despite strong emerging data that early life exposures, especially those in utero, are involved in EoE development (17), there is a limited understanding of how this exposure-to-development is mediated. Herein, we hypothesized that amniotic fluid may be a key part of the mechanism, as exposure to amniotic fluid may influence esophageal epithelial cells. If this hypothesis is proven correct, it has the potential to provide a new mechanistic paradigm for how prenatal factors influence the development of diseases, such as EoE.

MATERIALS AND METHODS

Amniotic Fluid Samples

Thirty-one samples of amniotic fluid were obtained from the Genomic and Proteomic Network (GPN) for the Preterm Birth Research study https://dash.nichd.nih.gov/study/13. The original samples were kept at −80°C. After being thawed, samples were cleared by passing through the 0.22-µm filter, divided into 100-µL aliquots, and kept at −80°C until use. Each aliquot was used once. Relevant clinical and experimental information for the samples used in this work is summarized in Supplemental Table S1.

Generation of Esophageal Epithelial Spheroids

The esophageal hTERT-immortalized human epithelial cell line EPC2 was a gift from Anil Rustgi (Columbia University, New York, NY). Esophageal epithelial spheroids were generated essentially as described previously (18, 19) with minor modifications. In brief, EPC2 cells were passed through 0.7-µm cell strainers and counted. At day 0 (D0) of the culture, a 35-µL droplet of undiluted Matrigel (354234, Corning) containing ∼1,000–2,000 cells/µL was placed in each well of a prewarmed 12-well plate using a cold wide bore tip. The plate was kept at 37°C for 30 min before adding 1 mL of the keratinocyte serum-free medium (17005042, Thermo Fisher) with 0.6 mM CaCl₂ (KSFMC) containing 10 µM Y27632 (688002, InSolution in DMSO, Calbiochem). The medium was replenished with KSFMC without Y27632 every other day starting from day 3 (D3) of the culture. At the indicated time, the Matrigel drop with spheroids was washed out of the plate in 750 µL of cold PBS/1% BSA into the Eppendorf tubes with a wide bore tip and processed. Special care was taken not to collect adherent cells that were attached to the plastic.

For bulk RNA sequencing, the spheroids were differentiated for 11 days, and RNA was collected at D0 before the cells were seeded and at D3 and day 11 (D11) of the culture. To assess the effects of amniotic fluid on the differentiation, amniotic fluid was added at a final volume of 10% to the medium at D0 and D3 of the culture. The spheroids were collected at day 5 (D5). To assess the effect of amniotic fluid on IL-13 response, amniotic fluid was added at a final volume of 10% to the medium at D3. On D5, the medium containing amniotic fluid was removed, and KSFMC medium with IL-13 (200-13, PeproTech, Rocky Hill, NJ) at the final concentration of 1 ng/mL was added for 48 h. The spheroids were collected at day 7 (D7).

Spheroids were collected in the low-adherent Eppendorf tubes (AM12450, Thermo Fisher Scientific) in washing buffer (PBS/1% BSA) using wide bore tips by washing out the Matrigel drops and avoiding plate scratching and spun down at 500 g at 4°C for 5 min; afterward, the washing buffer was carefully removed. For downstream analyses, the spheroids were recovered from Matrigel in 1 mL of cell recovery solution (354253, Corning) for 30 min on ice with occasional mixing, spun down at 500 g at 4°C for 5 min, and washed once with the washing buffer. Pelleted spheroids were resuspended in the washing buffer by tapping rather than by mixing with tips to avoid loss of spheroids due to adhesion to plastic.

RNA Isolation, cDNA Preparation, RT-PCR, and Bulk RNA Sequencing

For RNA isolation, spheroids were washed once with 1 mL of cold PBS and lysed in Tripure Isolation Reagent at 700 µL/sample (11667165001, Sigma). Chloroform was added at a volume of 140 µL per sample, and the tubes were shaken vigorously for 15 s and allowed to stand at room temperature for 10–15 min. Samples were centrifuged at 12,000 g for 15 min at 48°C, after which 350 µL of the upper aqueous phase was collected and mixed at a 1:1 ratio with 100% ethanol. Further isolation was performed by using the Quick-RNA MicroPrep Kit (R1051, Zymo Research, Irvine, CA) per the manufacturer’s instructions. The RNA concentration was measured by Nanodrop, and the RNA integrity number was determined by the Gene Expression Core at CCHMC by using the Agilent 2100 Bioanalyzer. cDNA synthesis was performed according to the protocol of the ProtoScript II Reverse Transcriptase kit (NEB, Ipswich, MA, M0368).

RT-qPCR was performed using a QuantStudio 7 Flex PCR system from Applied Biosystems with PowerUP SYBR Green Master Mix (A25778, Thermo Fisher Scientific). Primer sequences are available upon request.

Pair-end next-generation RNA sequencing was performed by the CCHMC Genomics Sequencing Facility Core using the standard Illumina protocol, and ∼20 million reads per sample were targeted for the sequencing. Data analysis and visualization were performed by using the CLC Genomics Workbench, version 23.0.5 (Qiagen, Hilden, Germany). Gene ontology (GO) enrichment analysis, which uses statistical methods to determine functional pathways associated with differentially expressed genes, was performed with the ToppGene suite14 (https://toppgene.cchmc.org) (20). For the STRING analysis, a stringency of 0.4 was applied to generate the connectome (21).

Western Blot

EPC2 cells were seeded at 200,000 cells/well in a 48-well plate in the KSFM medium. The next day, cells were washed three times with PBS, and 405 µL of PBS was added to the wells. Cells were rested for 2 h before stimulation with amniotic fluid, which was added to PBS to a final volume of 10% (45 µL). After 10 min, proteins were extracted with NuPAGE LDS Sample Buffer (4×) (NP0007, Thermo Fisher Scientific) premixed with radioimmunoprecipitation assay (RIPA) buffer (89900, Thermo Fisher Scientific) to a final concentration of 1× and supplemented with 5% β-mercaptoethanol and protease/phosphatase inhibitors (No. 5872, Cell Signaling). Samples were sonicated with a probe sonicator three times for 10 s each with 30% output and boiled. Protein lysates were subjected to electrophoresis on a 4%–12% protein gel and probed with rabbit mAb anti-phospho-p44/42 MAPK (ERK1/2) (Thr202/Tyr204) (D13.14.4E) XP (No. 4370, Cell Signaling, RRID: AB_2315112) and with mouse mAb p44/42 MAPK (ERK1/2) (3A7) (No. 9107, Cell Signaling, RRID: AB_10695739). Scanning was performed on an Odyssey M imager.

Immunofluorescence

After the recovery from Matrigel, the spheroids were washed once with PBS and fixed with 4% paraformaldehyde for 20 min at room temperature. The spheroids were washed once with 1 mL of PBS/5% BSA/0.5% Triton X-100 and then permeabilized and blocked in the same solution for 4 h at room temperature in the Eppendorf tube with gentle shaking. Primary antibodies against E-cadherin (AF648, R&D, RRID: AB_355504) and MKI567 (ab15580, Abcam, RRID: AB_443209) were added to the spheroids in the blocking solution for 16 h at 4°C with shaking. Spheroids were washed three times with 1 mL of the blocking solution by spinning at 1,000 g for 5 min, and the solution was replaced. The secondary antibodies conjugated to fluorophores (Jackson Laboratories) and Hoechst 33342 (H3570, Thermo Fisher Scientific, final concentration 1 µg/mL) were added together for 2 h at room temperature with shaking. The spheroids were washed three times with PBS/5% BSA/0.5% Triton X-100, placed on the slide in a small drop of the blocking solution, and covered with the coverslip. Imaging was performed with a Nikon A1 inverted confocal microscope in the Confocal Imaging Core at CCHMC.

Multiplex Analysis

Multiplex analysis of amniotic fluid was performed by Eve Technologies (Calgary, AB, Canada) using the Human Cytokine 71-Plex Discovery Assay with sensitivities ranging from 0.14 to 55.8 pg/mL.

Statistical Analysis

Statistical analysis using combined data from several experiments was performed with GraphPad Prism, version 9, software. For RT-PCR experiments with three groups, the Kruskal–Wallis test with Benjamini–Hochberg false discovery rate (FDR) correction was used, and an adjusted P value <0.05 was considered significant. When comparing between two groups, a two-tailed Mann–Whitney test with Holm–Sidak correction was used, and P < 0.05 was considered significant. For bulk RNA sequencing, either a t test or ANOVA with FDR correction was applied. Differentially expressed genes with an adjusted P value <0.05, minimal expression of 1 transcript per million, and fold change >2 were considered significant.

RESULTS

Amniotic Fluid Is Characterized by an Abundance of Cytokines

Amniotic fluid samples were collected from women with a gestational age ranging from 24 to 40 wk during preterm and term cesarean and vaginal deliveries (Supplemental Table S1). To elucidate the chemokine/cytokine profile of amniotic fluid, we analyzed 26 amniotic fluid samples for 71 analytes. Based on the Euclidean distance clustering, analytes were categorized into three groups: not detected, variably expressed, and highly expressed (Fig. 1 and Supplemental Table S2). The group of five not detected cytokines comprised IL-3, IL-4, IL-7, VEGFα, and TSLP (the minimum detectable concentration for the cytokines was <1 pg/mL, except TSLP for which it was ∼3 pg/mL). Undetectable TSLP levels are consistent with the recent study using a highly sensitive immunoassay that reported low levels of TSLP in the amniotic fluid of women experiencing preterm labor without intra-amniotic inflammation (22). The highly expressed group included 17 cytokines/chemokines with an average concentration of 70–3,000 pg/mL, some of which were previously identified as highly expressed in amniotic fluid (23). These analytes included tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), IL-6, IL-8, IL-1α, macrophage inflammatory protein-1 δ (MIP1δ), and interleukin-1 receptor α (IL-1RΑ), which were detected at a high level in most samples. The remaining 49 cytokines/chemokines exhibited variable expression levels across the amniotic fluid samples. This group included pro- and anti-inflammatory cytokines TNFα and IL-10; pro-atopy cytokines IL-5, IL-13, and IL-33; and the eosinophilic chemoattractant eotaxin-3 [Chemokine (C-C motif) ligand 26 (CCL26)].

Figure 1.

Cytokine/chemokine profile of amniotic fluid. The heatmap shows the expression of 71 cytokines/chemokines in amniotic fluid (n = 26). Expression was log (10) transformed, and samples were clustered using Euclidian distance and single linkage parameters. The gestational age (gest age) for each sample is indicated above the heatmap. Yellow and blue colors represent high- and low-expressed analytes, respectively.

Labor-associated changes in cytokine concentrations have been well characterized (24–26). We therefore compared the levels of cytokines/chemokines in the amniotic fluid obtained from women with vaginal deliveries (n = 5) to those who delivered by cesarean section (n = 21) independent of the gestational age (Supplemental Fig. S1A and Supplemental Table S1). Ten analytes were significantly elevated in the amniotic fluid from women who had labor (vaginal delivery), including IL-10, IL-17, CCL26, IL-2, and IL-1β, consistent with prior reports (25, 26).

Although amniotic fluid samples from women undergoing preterm (<37 wk) and term (38–40 wk) deliveries clustered together, the expression levels of several cytokines/chemokines correlated with the gestational age of the pregnancy (Supplemental Fig. S1B and Supplemental Table S2, yellow background). Notably, the only cytokine with a strong positive correlation with gestational age was EGF, whereas the expression of proinflammatory and immunomodulatory cytokines, including IL12-p70, TNFα, IL-1α, and thymus and activation-regulated chemokine [TARC (CCL17)], showed a negative correlation with gestational age. These findings collectively emphasize the abundance of cytokines/chemokines in amniotic fluid, influenced by the stage of pregnancy.

Use of Antibiotics in Pregnancy

Given that EoE has been associated with early-life exposure to antibiotics (17, 27), we assessed the use of antibiotics by women during pregnancy and in delivery from the questionnaires provided by the Genomic and Proteomic Network (GPN) for the Preterm Birth Research study (Supplemental Table S1). We found that women with preterm delivery (<37 wk) used antibiotics in a higher proportion both during pregnancy and in delivery than did women who delivered in term (38–40 wk) (Supplemental Fig. S1C, P = 0.02, Fisher’s extract test).

Exposure of Esophageal Epithelial Cells to Amniotic Fluid Leads to the Intracellular Signaling and Induction of Early Response Genes

To gain insights into the potential response of the esophageal epithelium to amniotic fluid, we investigated whether amniotic fluid would directly induce signaling in esophageal epithelial cells. Exposure of esophageal epithelial cells (EPC2) to amniotic fluid did not induce phosphorylation of either STAT6 or p65 NFκB but led to increased phosphorylation of extracellular signal-regulated kinase (ERK)1/2 (Supplemental Fig. S2). The application of amniotic fluid independent of preterm and term deliveries to EPC2 cells rapidly increased the phosphorylation of ERK1/2 in a time-dependent manner (Fig. 2A). This signaling event was accompanied by an increase in the expression of the early response genes EGR1 and EGR2, but not EGR3, FOS, and JUN following 1 h of exposure (Fig. 2B).

Figure 2.

Amniotic fluid induces signaling and transcriptional response in esophageal epithelial cells. A: representative Western blot for total and phosphorylated (phospho) p44/42 MAPK (ERK) following exposure of esophageal epithelial cells (EPC2) cells to amniotic fluid (AF) for 10 and 45 min. PBS, 10% vol/vol of PBS was added to the cells. Molecular weight markers are indicated on the right (kDa). Quantification of the phospho-ERK signal relative to the total ERK signal for each sample is shown on the graph (n = 7), Mann–Whitney test. B: expression level of early response genes was assessed by RT-qPCR 1 h after exposure to amniotic fluid (squares) or PBS [control (Ctrl, circles)]. Each marker represents an individual sample (n = 4 control, n = 10 amniotic fluid). Expression was normalized to the housekeeping gene GAPDH. The data are presented as means ± SE; ns, not significant using the Mann–Whitney test.

Amniotic Fluid Affects Esophageal Epithelial Differentiation in a Three-Dimensional Spheroid Model

Seeking to investigate the impact of amniotic fluid exposure on esophageal epithelial differentiation, we established a three-dimensional (3-D)-organoid culture system derived from EPC2 cells. This approach was previously used to mimic esophageal epithelial differentiation and changes associated with esophageal diseases in humans and mice (18, 19). Single-cell suspensions from EPC2 cells were embedded in Matrigel and cultured for 11 days, resulting in the formation of multilayered spheroids ∼100 μm in diameter. E-cadherin expression was consistently observed throughout the epithelium of these spheroids, and the outer layer contained proliferating cells marked by MKI67 positivity (Supplemental Fig. S3, A and B), consistent with previous reports (18, 19, 28). Bulk RNA sequencing analysis highlighted significant transcriptional changes during the differentiation process. Comparing D3 and D0, 1,971 genes were differentially expressed, and comparing D11 and D0, 4,255 genes were differentially expressed; there was an overlap of 1,462 genes between these datasets (FDR < 0.05, fold change (FC) 2, Supplemental Tables S3 and S4). Functional analysis of the 100 most differentially expressed genes across all conditions revealed upregulation of genes associated with epithelial development and differentiation biological processes, accompanied by a simultaneous downregulation of genes linked to cell proliferation (Supplemental Fig. S3, C–E). Together, these results established that esophageal epithelial cells, cultured as 3-D spheroids, provide a model system for studying the effects of amniotic fluid exposure on esophageal epithelial differentiation.

To investigate the impact of amniotic fluid on esophageal epithelial differentiation in the spheroid model, we differentiated esophageal spheroids for 5 days in the presence (10% vol/vol) or absence of amniotic fluid (n = 26). In the absence of amniotic fluid, spheroids were exposed to their basal serum-free medium supplemented with human recombinant epidermal growth factor and bovine pituitary extract in the presence of 0.6 mM of calcium. We then used RT-qPCR to evaluate the expression of several differentiation markers identified by RNA sequencing as significantly upregulated during the differentiation process. Findings revealed a downregulation in the expression of CRNN, DSG1, KRT1, KRT78, SPRR2E, and SLURP1 at D5 of the differentiation process in the presence of amniotic fluid compared with the KSFMC medium alone. Notably, the expression of SERPINB3 and KLK7 remained unaffected (Fig. 3). Although amniotic fluid from term deliveries exerted a trend for a less deleterious effect on the expression of some differentiation markers, it did not reach statistical significance (data not shown). Collectively, these results indicate that exposure of the esophageal epithelium (in vitro) to amniotic fluid can have a profound effect on the differentiation program.

Figure 3.

Effect of amniotic fluid on esophageal epithelial differentiation. Esophageal spheroids were differentiated either in the presence or in the absence of amniotic fluid from day 0 (D0) to day 5 (D5). Untreated (UT), spheroids differentiated in KSFMC medium (n = 12); amniotic fluid (AF), spheroids differentiated in the presence of 10% of amniotic fluid [total n = 26; preterm (24–37 wk), n = 12, and term (38–40 wk), n = 14]. The expression level of the indicated genes was assessed by RT-qPCR and normalized to the housekeeping gene GAPDH. For the box and whisker plots, the box represents the 50th percentile of the data, the whiskers show minimum and maximum values, and the line in the box represents the median. ns, not significant. Mann–Whitney test, ****P < 0.0001.

Amniotic Fluid Amplifies the Response of the Esophageal Epithelial Cells to IL-13

We investigated the impact of amniotic fluid on the modification of esophageal epithelial response to the pro-allergic cytokine IL-13, a primary contributor to EoE pathogenesis (29). Spheroids underwent amniotic fluid exposure between D3 and D5 of differentiation, followed by IL-13 stimulation. After 48 h, we assessed the expression of established IL-13 target genes using RT-qPCR. As a control, IL-13 stimulation alone upregulated CAPN14, CCL26, SERPINB4, and TNFAIP6 and downregulated DSG1, PADI1, and SLURP1 (Fig. 4A). After amniotic fluid exposure, CAPN14 and CCL26 expression further increased, whereas DSG1 and SLURP1 expression decreased compared with IL-13 alone (Fig. 4B). Other IL-13 targets exhibited no change.

Figure 4.

Amniotic fluid amplifies the response of the esophageal epithelial cells to IL-13. A: esophageal spheroids were exposed to IL-13 for 48 h (IL-13) or left untreated (UT), and the expression level of the indicated genes was assessed by RT-qPCR and normalized to the housekeeping gene GAPDH. The Mann–Whitney test with the Benjamini–Hochberg correction for the false discovery rate was used to calculate significance. B: esophageal spheroids were exposed to amniotic fluid (AF) between day 3 (D3) and day 5 (D5) (n = 22); at D5, spheroids were exposed to IL-13 for 48 h (IL-13 + AF). Spheroids also were differentiated without amniotic fluid and exposed to IL-13 at D5 (IL-13, n = 8). The expression level of the indicated genes was assessed by qRT-PCR and normalized to the housekeeping gene GAPDH. For the box and whisker plots, the box represents the 50th percentile of the data, the whiskers show minimum and maximum values, and the line in the box represents the median. Mann–Whitney test, *P < 0.05, **P < 0.01, and ****P < 0.0001; ns, not significant. C: expression of CAPN14 and CCL26 in the spheroids differentiated in the absence [untreated (UT)] or presence of amniotic fluid but in the absence of IL-13. Amniotic fluid samples were divided into 2 groups by gestational age: preterm (24–37 wk, Pre, n = 12) and term (38–40 wk, Term, n = 14). For the box and whisker plots, the box represents the 50th percentile of the data, the whiskers show minimum and maximum values, and the line in the box represents the median. P values are shown for the Kruskal–Wallis test with the Benjamini–Hochberg correction for the false discovery rate.

We also explored the capacity of the amniotic fluid from women who underwent preterm and term delivery to increase CAPN14 and CCL26 expression independent of IL-13. Amniotic fluid samples were divided into two groups by gestational age: preterm (24–37 wk, Pre, n = 12) and term (38–40 wk, Term, n = 14). After 5 days of the differentiation process, CAPN14 exhibited significant upregulation in spheroids exposed to preterm but not term amniotic fluid. In contrast, the baseline expression level of CCL26 was further decreased following exposure to amniotic fluid from term but not preterm deliveries (Fig. 4C). Notably, the concentration of IL-13 in the amniotic fluid was not significantly different between preterm and term groups based on the multiplex analysis (Supplemental Fig. S4), and exposure of EPC2 cells to the amniotic fluid did not induce STAT6 phosphorylation (Supplemental Fig. S2), consistent with the low concentration of IL-13 in the amniotic fluid (<100 pg/mL). Overall, these findings indicate that amniotic fluid has the capacity to modify pro-atopic responses including IL-13-induced transcriptional effects and the expression of key genes involved in type 2 responses such as CAPN14 and CCL26.

DISCUSSION

The prenatal environment is widely recognized as the initial exposure in a cascade of lifelong exposures that modify the susceptibility to future inflammatory diseases, such as allergies including asthma and EoE (30, 31). Existing studies predominantly focused on factors such as the delivery method, maternal feeding practices, maternal infections, neonatal care, antibiotic and acid suppressant use during infancy, pet ownership, and dysbiosis (8, 32, 33). Of those, the occurrence of EoE is positively associated with such prenatal factors as maternal fever but not with diet or smoking (32), suggesting that intrauterine exposures can predispose to inflammatory responses later in life. Herein, we aimed to uncover mechanisms that may be operational in utero by exploring the potential effect of amniotic fluid on esophageal epithelial cells. Our findings demonstrate a direct effect of amniotic fluid on esophageal epithelial cells, including amniotic fluid’s ability to modify epithelial cell differentiation, response to inflammatory mediators, and expression of select genes key to type 2 allergic inflammation.

Cytokine/chemokine profiling of amniotic fluid revealed a rich cytokine and chemokine presence, emphasizing its complexity and potential role in influencing esophageal epithelial cells. Although the expression profile of the cytokines/chemokines did not separate amniotic fluid samples by gestational age, the heightened expression of proinflammatory cytokines, such as TNFα, IL-12, and CCL17 in amniotic fluid from women who underwent preterm birth suggests a potential proinflammatory environment associated with preterm deliveries. Conversely, EGF, a growth factor crucial for fetal development, exhibited the highest expression in term birth samples, indicating a possible protective role associated with full-term pregnancies. These collective findings suggest that exposure of the esophageal epithelium to amniotic fluid has the potential to modify epithelial differentiation. We also identified labor-associated changes in the composition of amniotic fluid, including elevated expression of IL-1β and IL-2, consistent with previous literature (24–26). We, however, acknowledge that factors other than labor and the use of antibiotics during delivery including the sensitivity of the assay may affect the consequences of cytokine profiling.

We demonstrate that in vitro exposure of esophageal epithelial cells to amniotic fluid triggered direct signaling events, notably the transient phosphorylation of ERK1/2 and the transcription of early response genes (EGR1 and EGR2). ERK phosphorylation is a common signal transduction response that has been previously demonstrated in primary neurons following exposure to amniotic fluid (34, 35). It is, therefore, not surprising that this pathway would be induced in the esophageal epithelial cells by the amniotic fluid that is enriched in cytokines/chemokines and growth factors compared with the defined medium. We, however, demonstrate elevated expression of the specific downstream targets (e.g., EGR1 and EGR2, but not JUN and FOS) and specific effects on the gene expression dependent upon the gestational age when the fluid was harvested (e.g., elevated expression of CAPN14 by the preterm but not term amniotic fluid samples), suggesting that the effects observed have specificity. Previous research has demonstrated the impact of amniotic fluid on various cell types but not esophageal epithelial cells. For instance, in human corneal endothelial cells, amniotic fluid significantly enhanced cellular proliferation, accompanied by increased expression of the tight-junction protein zonula occludens-1 (ZO-1) (10). Similarly, human fetal and adult skin fibroblasts exhibited a stimulatory effect when exposed to amniotic fluid, mediated through activation of MEK/ERK and PI3K/Akt signaling (12). Furthermore, exposure of retinal pigment epithelial cells to amniotic fluid resulted in their transdifferentiation into retinal neurons (36), showcasing amniotic fluid’s ability to directly influence cellular differentiation.

Esophageal epithelium maintains a delicate balance between proliferation and differentiation by continuously guiding differentiated keratinocytes toward the luminal surface and ultimately desquamating cornified and flat keratinocytes (37). Historically, assessing esophageal epithelial differentiation in vitro involved techniques such as air-liquid interphase culture and organotypic culture, where cells grow on membrane supports (38, 39). In recent years, a 3-D spheroid model has emerged as a powerful tool in studying esophageal epithelial biology and pathophysiology with the potential for personalized medicine (40). This system is especially useful in assessing the differentiation process of the esophageal epithelium (18, 19, 41–43). Herein, a 3-D spheroid model of esophageal epithelial differentiation was used, revealing that amniotic fluid modifies the esophageal epithelial differentiation program when compared with the serum-free defined medium. The presence of amniotic fluid altered the expression of differentiation markers, such as CRNN, KRT78, and SLURP1, emphasizing the amniotic fluid’s potential impact on cellular processes crucial for maintaining esophageal epithelium integrity. Notably, a higher level of detection of the esophagus-enriched proteins in the amniotic fluid from more advanced gestational age (5) suggests that amniotic fluid is an active regulator of epithelial maturation. It is tempting to speculate that the level of the esophagus-enriched proteins in the amniotic fluid might serve as an early predictor of the development of EoE and/or other tissue-specific inflammatory responses later in life.

Amniotic fluid is a reservoir of antigens and allergens, including major food allergens and decidual tissue-derived Th2 cytokines, such as IL-13, and IgE, which are swallowed by the fetus (44–47). This environment may induce early in utero sensitization, potentially leading to subsequent childhood allergic reactions to foods and the development of EoE. Investigating the influence of amniotic fluid on the esophageal epithelial response to IL-13, a pro-allergic cytokine driver of EoE, we observed that amniotic fluid enhanced the transcriptional response of genes linked to the IL-13 response, particularly CCL26 and CAPN14. This effect is likely multifactorial and not due merely to the exposure of the spheroids to the very low concentration of IL-13 present in the amniotic fluid, which is insufficient to induce STAT6 phosphorylation (Supplemental Fig. S2). Notably, the upregulation of CAPN14, a gene associated with the susceptibility to EoE, was observed following exposure to preterm (but not term) amniotic fluid samples. These findings suggest that the disruption of the epithelial barrier, the hallmark of the “atopic march,” may begin in utero. Amniotic fluid may amplify the response of epithelial cells to pro-allergic stimuli, providing a potential link between in utero exposures and the development of allergic conditions and signifying the importance of early intervention steps to improve the epithelial barrier, including the use of synthetic ceramide analog (48–50).

Notably, the primary pathogenesis of EoE involves the interplay of gene-environment interactions. EoE genetic factors primarily influence EoE pathogenesis through the epithelium rather than eosinophils or the adaptive immune system as revealed by the genetics of human EoE and genetic models of EoE in mice (51–53). Environmental factors such as the use of antibiotics in infancy are associated with the increased risk of pediatric EoE (32). We found that women who underwent preterm delivery used antibiotics in a higher proportion than women with term deliveries. The findings reported herein advance the potential mechanistic understanding of how the intrauterine environment mediates gene-environment interaction, calling attention to the direct effects of amniotic fluid on esophageal epithelial responses involved in innate immunity. We acknowledge, however, the lack of detailed medical history for the participants, including the occurrence of premature rupture of membranes and the presence of Group B Streptococcus, which can trigger the use of antibiotics and subsequently alter amniotic fluid composition and microbiome (54).

In conclusion, this investigation provides evidence for the interplay between amniotic fluid components and esophageal epithelial cells, emphasizing the potential role of amniotic fluid in the development of esophageal disorders, particularly EoE. The findings provide new insight into the effects of intrauterine exposures on the risk and pathogenesis of inflammatory diseases, calling attention to the role of amniotic fluid as a potential mediator of the intrauterine environment that influences esophageal disorders (Fig. 5).

Figure 5.

Schematic illustration delineating potential impacts of the amniotic fluid on the esophageal epithelium. During the gestational period, the interaction between amniotic fluid and the esophageal epithelium has the potential to alter the esophageal epithelium in terms of differentiation and subsequent responsiveness to inflammatory stimuli, notably interleukin 13 (IL-13). These effects are evident from expression changes in the genes linked to the esophageal epithelial differentiation and IL-13 response. Collectively, these effects may consequently predispose to inflammatory conditions of the esophagus, such as eosinophilic esophagitis (EoE), in the later stages of life. Abx, antibiotics; pERK1/2, phosphorylated ERK1/2.

DATA AVAILABILITY

RNA sequencing data have been uploaded to GEO under the accession number GSE256022.

SUPPLEMENTAL MATERIAL

Supplemental Tables S1: https://doi.org/10.6084/m9.figshare.26603914.

Supplemental Table S2: https://doi.org/10.6084/m9.figshare.26603911.

Supplemental Table S3: https://doi.org/10.6084/m9.figshare.26603920.

Supplemental Table S4: https://doi.org/10.6084/m9.figshare.26603917.

Supplemental Figs. S1–S4: https://doi.org/10.6084/m9.figshare.26662801.

GRANTS

This work was supported by the National Institutes of Health Grants R01 AI045898, R01 AI124355, U19 AI070235, and P30 DK078392 (Gene and Protein Expression Core); the Campaign Urging Research for Eosinophilic Disease (CURED); the Buckeye Foundation; and the Sunshine Charitable Foundation and its supporters, Denise A. Bunning and David G. Bunning.

DISCLOSURES

M.E.R. is a consultant for Pulm One, Spoon Guru, ClostraBio, Serpin Pharm, Allakos, Celldex, Uniquitybio, Santa Ana Bio, EnZen Therapeutics, Bristol Myers Squibb, Astra Zeneca, Pfizer, GlaxoSmith Kline, Regeneron/Sanofi, Revolo Biotherapeutics, and Guidepoint and has an equity interest in the first nine listed and royalties from reslizumab (Teva Pharmaceuticals), PEESSv2 (Mapi Research Trust), and UpToDate. M.E.R. is an inventor of patents owned by Cincinnati Children’s Hospital. Y.S. is a consultant to Bio-Rad Laboratories, Inc. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

M.R., Y.S., and M.E.R. conceived and designed research; M.R. and A.M.K. performed experiments; M.R., A.M.K., and J.M.C. analyzed data; M.R., A.M.K., J.M.C., Y.S., and M.E.R. interpreted results of experiments; M.R. prepared figures; M.R. drafted manuscript; M.R., A.M.K., J.M.C., Y.S., and M.E.R. edited and revised manuscript; M.R., A.M.K., J.M.C., Y.S., and M.E.R. approved final version of manuscript.