Eosinophilic esophagitis (EoE) is a clinicopathologic condition that is immune- or antigen- driven and characterized by infiltration of eosinophils into the esophageal epithelial layers.^1,2 The incidence and prevalence of EoE have risen rapidly in recent years, making EoE one of the leading causes of dysphagia and food impaction.^2,3 Genome-wide association studies have revealed a strong type 2 immune response in EoE.^4,5 Related chemokines and factors, including eotaxin-3 (also known as CCL26) and filaggrin (FLG) appear to play essential roles in the pathogenesis of EoE.^4,6,7

Inflammatory cells are also involved in the pathogenesis of EoE.⁸ In addition to eosinophils, mast cells are increased in EoE patients, and are thought to be associated with esophageal symptoms and tissue remodeling.^9,10 T lymphocytes are likewise increased in EoE patients.^11,12

Among T lymphocytes, regulatory T cells (Tregs) regulate autoimmune responses and play important roles in many allergic and inflammatory conditions.^13,14 Tregs can inhibit excessive immune response by inhibiting the activity and proliferation of effector T cells, enhancing immune tolerance and maintaining immune homeostasis.¹⁴ However, the number of Tregs infiltrating into the esophageal epithelial layers of EoE patients remains controversial,^15-18 and studies focusing on Tregs in EoE are limited. We therefore aim to evaluate Tregs and their correlations with other inflammatory cells in EoE patients.

Human Endoscopic Biopsies

This was a retrospective, observational, single-center study. EoE was diagnosed according to the guidelines.^1,2 We used a previously described method to process samples.¹⁹ Briefly, controls were patients who underwent endoscopy due to clinical complaints, with 0 eosinophils per high-power field (HPF) and no histologic or endoscopic evidence of EoE.²⁰ Biopsies of esophagus from controls and patients with EoE before and after treatments were taken. Histologic remission after treatment was defined as a count of < 15 eosinophils/HPF. Endoscopic findings (rings, exudates, edema, furrows, and strictures) were graded according to the EoE endoscopic reference score (EREFS).²¹ The total EREFS was calculated by combining individual values, ranging from 0 to 10. Two separate samples were taken from the middle and distal esophagus of every control/patient. Samples were collected from specific locations displaying characteristic endoscopic findings of EoE. One sample was used for measurement of messenger RNA (mRNA) levels, and another one from same region was used for immunohistochemical and hematoxylin and eosin staining.¹⁹ This study was performed with approval from the Ethics Committee of Hyogo Medical University (No. 174). Written informed consent was provided by every control/patient.

Reverse Transcription-quantitative Polymerase Chain Reaction

Total mRNA was extracted using Trizol reagent (Thermo Fisher Scientific, Waltham, MA, USA). Synthesis of cDNA was performed using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific). Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was performed within PCR master mix and QuantStudio 12K Flex real-time PCR system (Thermo Fisher Scientific). TaqMan probes and primers for FLG (Hs00856927_g1) and eotaxin-3 (encoded by CCL26) (Hs00171146_m1) were used, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Hs 02758991_g1) gene was an internal reference. The ΔΔCT method was used to compare relative expression levels.

Immunohistochemical Staining

Formalin-fixed paraffin-embedded samples were sectioned at 4 ^μm. After antigen retrieval by autoclaving in tris-ethylenediaminetetraacetic acid based buffer (S2375; Dako, Carpinteria, CA, USA), sections were incubated with 3% hydrogen peroxide and protein-free blocking solution (X0909; Dako). Sections were stained with rabbit anti-human CD117 (1:500, A4502; Dako), rabbit anti-human CD3 (1:150, ab16669; Abcam, Cambridge, UK) or mouse anti-human Forkhead box P3 (FOXP3) (1:500, ab20034; Abcam), sequentially. Sections were then incubated with EnVision+ Dual Link System-horseradish peroxidase (HRP) (K4063; Dako), EnVision rabbit-HRP (K4003; Dako) or EnVision mouse-HRP (K4001; Dako) and developed with 3,3´-Diaminobenzidine substrate using the chromogen system (K3468; Dako). Finally, specimens were counterstained with Mayer’s hematoxylin. The peak mast cells, T cells and Tregs counts were obtained by counting at least 3 representative non-overlapping HPFs using a microscope (Olympus BX51; Tokyo, Japan) at 400× magnification in a blinded manner. Samples were evaluated by JC and TO in a blind manner. In the event of discordance in histological evaluation, the specimen underwent further review until a consensus was reached on the assessment.

Statistical Methods

Categorical data are expressed as frequency and bivariate analysis was performed using Fischer’s exact test. Continuous data are expressed as mean ± SD, whereas continuous variables with skewed distributions are reported as median and interquartile range (IQR). Bivariate analysis was evaluated with unpaired t tests (two-tailed) or the Mann–Whitney U test between controls and EoE patients before treatment. The Wilcoxon signed-rank test was used to evaluate findings before and after treatments. Correlations between inflammatory cells were calculated using the Spearman test. A P-value of < 0.05 was deemed statistically significant. Statistical analyses were performed with GraphPad Prism version 9 software (GraphPad Software, La Jolla, CA, USA).

Demographic Characteristics

Seventeen patients with EoE and 17 Controls were included in this study. No age or sex differences were observed between EoE patients and controls (Table 1, Supplementary Tables 1 and 2). None of the EoE patients were found to have any concomitant allergic disease or atopic condition.

Table 1 . Demographic Characteristics of Controls and Eosinophilic Esophagitis Patients

	Control n = 17	EoE n = 17	P-value
Age (yr)	48.5 ± 11.3	42.9 ± 11.7	0.167
Sex (male/female)	8/9	9/8	0.999

EoE, eosinophilic esophagitis; SD, standard deviation.

Data are presented as mean ± SD or n/n.

Regulatory T Cell and Mast Cell Infiltration in Eosinophilic Esophagitis Patients

Eosinophils, mast cells (CD117+), T cells (CD3+), and Tregs (FOXP3+) were significantly increased in the esophageal epithelial layers of EoE patients compared with those of controls (Fig. 1). T cells and Tregs were observed in both controls and EoE patients, while eosinophils and mast cells were only observed in EoE patients (Fig. 2).

Figure 1. Numbers of eosinophils, mast cells, T cells, and regulatory T cells (Tregs) in controls (Cont) and patients with eosinophilic esophagitis (EoE). Numbers of eosinophils (A), mast cells (B), T cells (CD3+) (C), and Tregs (Forkhead box P3 [FOXP3]+) (D) in the esophageal epithelia of controls and patients with EoE are shown. ****P < 0.0001. HPF, high power field.

Figure 2. Infiltration of eosinophils, mast cells, T cells, and regulatory T cells (Tregs) in controls and patients with eosinophilic esophagitis (EoE) before and after treatment. (A-C) Hematoxylin and eosin (H&E) staining of biopsy sections. (D-F) CD117 staining of esophageal biopsy sections. (G-I) CD3 staining of esophageal biopsy sections. (J-L) Forkhead box P3 (FOXP3) staining of esophageal biopsy sections. Representative staining of controls (A, D, G, J), EoE patients before treatment (B, E, H, K), and EoE patients after treatment (C, F, I, L) are shown. Bar = 50 ^μm.

Among controls and EoE patients before treatment, eosinophils correlated positively with mast cells (Fig. 3A). Further, T cells correlated positively with Tregs (Fig. 3B). T cells also correlated positively with both eosinophils and mast cells (Fig. 3C and 3D). In addition, Tregs correlated positively with eosinophils and mast cells (Fig. 3E and 3F). Eosinophils, mast cells, T cells, and Tregs correlated positively with EREFS (Fig. 4).

Figure 3. Correlations between inflammatory cells in esophageal epithelia of controls and patients with eosinophilic esophagitis (EoE). Correlations between eosinophils and mast cells (A), T cells (CD3+) and regulatory T cells (Tregs) (Forkhead box P3 [FOXP3]+) (B), eosinophils and T cells (C), mast cells and T cells (D), eosinophils and Tregs (E), and mast cells and Tregs (F) are shown (n = 34). The Spearman test is used for determining correlations between data sets. ****P < 0.0001. HPF, high power field.

Figure 4. Correlations between inflammatory cells in esophageal epithelia and eosinophilic esophagitis (EoE) endoscopic reference score (EREFS) of controls and patients with EoE. Correlations between eosinophils and EREFS (A), mast cells and EREFS (B), T cells (CD3+) and EREFS (C), and Tregs (FOXP3+) and EREFS (D) are shown (n = 34). The Spearman test is used for determining correlations between data sets. ****P < 0.0001. HPF, high power field.

Regulatory T Cells and Mast Cells in Eosinophilic Esophagitis Patients After Treatment

After treatment using acid suppressants and steroids, 7 EoE patients did not achieve histologic remission (≥ 15 eosinophils/HPF). Similarly, mast cells, T cells, and Tregs were not significantly decreased after treatment in these EoE patients when histologic remission failed to be achieved (Fig. 5A, Supplementary Table 3).

Figure 5. Numbers of eosinophils, mast cells, T cells, and regulatory T cells (Tregs) in patients with eosinophilic esophagitis (EoE) before and after treatment. Numbers of eosinophils, mast cells, T cells (CD3+) and Tregs (Forkhead box P3 [FOXP3]+) in the esophageal epithelia of EoE patients after treatment that did not achieve histologic remission (A), and after treatment that did achieve histologic remission (B). *P < 0.05, ***P < 0.001. HPF, high power field. Pre, EoE patients before treatment; Post, EoE patients after treatment.

Conversely, 7 EoE patients achieved histologic remission (< 15 eosinophils/HPF) after treatments by acid suppressants and steroids. Mast cells, T cells, and Tregs were all significantly decreased in EoE patients after treatment when histologic remission was achieved (Fig. 5B, Supplementary Table 3).

Among EoE patients before and after treatment achieving histologic remission, eosinophils correlated positively with mast cells (Fig. 6A). T cells also correlated positively with T regs (Fig. 6B). Further, T cells correlated positively with eosinophils and mast cells (Fig. 6C and 6D). In addition, Tregs correlated positively with eosinophils and mast cells (Fig. 6E and 6F). Mast cells and Tregs correlated positively with EREFS (Fig. 7B and 7D), whereas no correlation was observed between eosinophils or T cells and EREFS (Fig. 7A and 7C).

Figure 6. Correlations between inflammatory cells in esophageal epithelia of patients with eosinophilic esophagitis (EoE) before and after treatment achieving histologic remission. Correlations between eosinophils and mast cells (A), T cells (CD3+) and Tregs (Forkhead box P3 [FOXP3]+) (B), mast cells and T cells (C), mast cells and regulatory T cells (Tregs) (D), eosinophils and T cells (E), and eosinophils and Tregs (F) are shown (n = 14). The Spearman test is used to determine correlations between data sets. **P < 0.01, ***P < 0.001. HPF, high power field.

Figure 7. Correlations between inflammatory cells in esophageal epithelia and eosinophilic esophagitis (EoE) endoscopic reference score (EREFS) of patients with EoE before and after treatment achieving histologic remission. Correlations between eosinophils and EREFS (A), mast cells and EREFS (B), T cells (CD3+) and EREFS (C), and Tregs (Forkhead box P3 [FOXP3]+) and EREFS (D) are shown (n = 14). *P < 0.05, **P < 0.01. HPF, high power field.

Gene Expressions in Controls and Eosinophilic Esophagitis Patients Before and After Treatment

Median levels of eotaxin-3 mRNA in EoE patients (4.41; IQR, 1.70-10.40) were significantly increased compared with controls (0.02; IQR, 0.01-0.04; P < 0.0001). However, mRNA levels of FLG were significantly decreased in EoE patients (0.04; IQR, 0.01-0.09) compared with controls (1.23; IQR, 0.60-1.81; P < 0.0001). When histologic remission was achieved after treatment, levels of eotaxin-3 mRNA were significantly downregulated, whereas levels of FLG mRNA were significantly upregulated. Even when histologic remission was not achieved after treatment, levels of FLG mRNA were significantly upregulated, while levels of eotaxin-3 mRNA were not downregulated (Table 2).

Table 2 . Gene Expression in Biopsy Samples From Eosinophilic Esophagitis Before and After Treatment

	Before treatment	After treatment	P-value
With histologic remission	n = 7	n = 7
CCL26	8.08 (4.09, 12.77)	0.51 (0.08, 4.11)	0.016
FLG	0.01 (0.01, 0.01)	0.74 (0.39, 0.84)	0.016
Without histologic remission	n = 7	n = 7
CCL26	6.19 (3.05, 20.11)	3.26 (1.31, 17.83)	0.672
FLG	0.01 (0.01, 0.08)	0.11 (0.03, 0.47)	0.031

CCL26, eotaxin-3; FLG, filaggrin.

Data are presented as median (interquartile range).

In the present study, infiltration of Tregs in esophageal epithelium layers of adult EoE patients was significantly increased compared to controls and was significantly decreased after treatment. The number of Tregs infiltrating into epithelial layers correlated significantly with numbers of eosinophils and mast cells. Although Tregs are thought to function to regulate inflammation, infiltration correlated with active inflammation in the present study.

Whether Tregs are increased in EoE patients has remained controversial. In several previous studies examining pediatric EoE patients, Tregs were increased,^16,18 while others focusing on adult EoE patients have found a decreased ratio of Tregs in patients with active EoE.^15,17 The present study observed increased Tregs in adult EoE patients before treatment, while Tregs were decreased after treatments achieving histologic remission. As other factors assessed were similarly modified in EoE samples and after treatment,^5,7,20 the subjects assessed in the present study were considered typical and the findings were similar to data from pediatric EoE patients. A previous study likewise showed that Tregs were decreased in patients with acute-stage atopic dermatitis, but increased in patients with chronic-stage atopic dermatitis.²² Since EoE is a chronic inflammatory disease, Tregs must be somewhat increased in EoE patients. Despite the increased number of Tregs in EoE patients, the role of Tregs remains unclear. Traditionally, Tregs have been considered to suppress inflammatory responses, such as by inhibiting effector T-cell activation.¹⁴ However, some previous studies have found that Tregs from patients with atopic diseases have less suppressive capacity than those from controls.^23-25 Similar conditions might be involved in EoE patients. Even though the number of Tregs was increased, impaired suppressive capacity could no longer suppress the inflammatory responses caused by T helper 2 cells, suggesting that Tregs from EoE patients are unable to stop or reverse the pathogenesis of EoE. We now speculate that increased Tregs might represent an attempt by the body to inhibit inflammation via negative feedback loops, although with impaired function they could not prevent disease in the present study.

FLG is a structural protein expressed in esophageal and skin mucosa, where the protein contributes to stratified epithelial barrier function.^7,26,27 Levels of FLG mRNA were decreased in EoE patients, consistent with previous studies.^6,19 However, even without histologic remission, levels of FLG mRNA were upregulated in EoE patients after treatments, indicating that the genes related to barrier function tend to recover first.

In the present study, mast cells were increased in esophageal epithelium of EoE patients, consistent to previous studies.^9,19,28 Mast cells that infiltrate into the esophageal epithelia release mediators involved in the remodeling of those layers.^20,29 Previous studies have found that mast cell counts could be used as a marker of EoE activity in addition to eosinophil count.^9,19,29

Infiltrated eosinophils correlated with Tregs in EoE patients before and after treatment in the present study. A previous study using both in vivo and in vitro models found that eosinophils could regulate expansion of gastrointestinal Tregs, and eosinophils could induce FOXP3 expression in naïve T cells upon co-culture.³⁰ Another in vivo study indicated that eosinophils were able to recruit Tregs.³¹ Such findings suggest that eosinophils can regulate the accumulation of Tregs, which may explain how eosinophils correlated with infiltration of Tregs in patients with EoE.

In this study, both mast cells and Tregs were increased in patients with EoE before treatment, and both were decreased after treatment. Further, a positive correlation was seen between mast cells and Tregs. Traditionally, Tregs are thought to suppress the expansion of mast cells,¹³ but this would not explain why Tregs and mast cells showed a positive correlation. However, similar to the present study, mast cells and Tregs were found to be increased in patients with eosinophilic gastritis.³² In addition, several in vivo studies have found that mast cells can promote the expansion of Tregs in mouse models.^33,34 Further, Tregs from patients with atopic diseases show less suppressive capacity.^23-25 These interactions could be why mast cells and Tregs correlated positively in the present study. Further studies are required to determine the meanings of these correlations between mast cells and Tregs in EoE patients.

EREFS is a validated grading and classification system developed by Hirano et al,²¹ consistent in the description of endoscopic findings of EoE patients. EREFS has also been validated for its responsiveness to treatment in previous studies involving EoE cases.^35,36 In the present study, the positive correlation between Tregs and EREFS was observed in controls and EoE patients, as well as before and after treatment, suggesting a potential association between Tregs and the endoscopic findings of individuals with EoE.

Several limitations to the present study merit consideration. First, our sample size was small due to the low prevalence of EoE in Japan.³ However, our data share several similarities with findings from previous reports.^5,7,20 The subjects included in the analysis were deemed representative of the target population. Additionally, the limited number of patients available in previous comparable studies resulted in a similar sample size.^37,38 Second, only 2 samples were taken, from the middle and distal esophagus. This is less than recommended in the guidelines.^1,2 Evaluations should be performed with more samples from each patient.

In conclusion, Tregs were increased in the esophageal epithelium of patients with EoE. Infiltration of Tregs was decreased in EoE patients achieving histologic remission after treatment. Peak Treg counts correlated with mast cell and eosinophil counts.

Note: To access the supplementary tables mentioned in this article, visit the online version of Journal of Neurogastroenterology and Motility at http://www.jnmjournal.org/, and at https://doi.org/10.5056/jnm23040.

We thank Ms Mayumi Yamada and Dr Takashi Nakanishi for technical assistance.

This study was supported by a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (JSPS) (19K08480 and 22K08066 to Tadayuki Oshima).

None.

Junji Chen conducted experiments, analyzed data, prepared figures, and wrote the original draft; Tadayuki Oshima handled project administration, interpreted results of experiments, and reviewed the manuscript; Toshihiko Tomita contributed to interpretation of data; Hirokazu Fukui was the supervisor in this study and revised the paper critically for important intellectual content; and Shinichiro Shinzaki designed the research study and contributed to interpretation of data. All authors approved the final version of the manuscript.