2023 Impact Factor
2023 Impact Factor
,1,3* Sezgi Kipcak,2,3 Nur S Gunel,2 Serhat Bor,3 and Eser Y Sozmen1
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Gastroesophageal reflux disease (GERD) is a chronic disease that causes mucosal damage in the esophagus directly and/or indirectly through different noxious agents. GERD is classified based on the appearance of the esophageal mucosa on upper gastrointestinal (GI) endoscopy into erosive reflux disease (ERD) and nonerosive reflux disease (NERD). However, the classification of GERD may not be easy, unlike ERD and NERD. According to nonpathological acid exposure in 24-hour multichannel intraluminal impedance-pH (MII-pH) monitoring, the groups may differ from NERD, such as reflux hypersensitivity (RH) and functional heartburn (FH). According to 24-hour MII-pH monitoring, normal reflux burden but positive symptom association is called RH, and normal reflux burden and negative symptom association is called FH. The modern diagnosis of GERD has accepted FH as the non-GERD group because the patients experience typical heartburn symptoms.1
Two theories have been suggested to explain the pathologic mechanisms. According to the acid-peptic damage approach, which is the traditional theory, acid as well as pepsin, bile acids, and pancreatic enzyme exposure at high concentrations causes damage to the epithelium. While surface epithelial cells are affected by acid-peptic damage, mucosal inflammation begins at the apical surface.2 With the rapid infiltration of neutrophils and eosinophils (granulocytes), the esophageal mucosa becomes more permeable, and the passage of hydrogen ions increases with this damage.3 However, the chemical damage hypothesis is inadequate to explain the pathogenesis of patients with nonacid reflux and NERD since there is no inflammation.
An alternative theory has emerged in recent years that suggests that epithelial damage progresses indirectly from basal cells to apical cells via cytokine mediators.4 The first immune cells to encounter harmful agents are T cells, not granulocytes, and granulocyte infiltration does not occur until erosions of the epithelium appear. In the presence of GERD-associated microscopic esophagitis, other studies have also supported that the dominant cells in the esophageal epithelium are T cells.5 The theory of cytokine-mediated inflammatory damage in the pathogenesis of the disease is more explanatory in patients with weak acid reflux and/or hypersensitivity.
Based on this information, we aim to investigate the inflammatory processes occurring in the esophageal epithelium of reflux phenotypes at the molecular level and the effects of these changes on tissue integrity. We also examined the relationship between cytokines and hypersensitivity symptoms in patients with RH and FH. For this purpose, tissue integrity, cytokine levels and inflammatory pathways were examined from esophageal biopsies of patients with mild ERD (A/B), severe ERD (C/D), NERD, RH, and FH.
Ethical approval was provided by Ege University Clinical Research Ethics Committee, Izmir, Turkey (18-2.1/36, 20 February 2018). All participants gave written informed consent. We included 91 outpatients admitted to the Ege University Hospital, Gastroenterology Department, Gastroesophageal Reflux Outpatient Clinic with typical reflux symptoms, such as pyrosis and/or regurgitation at least once a week or more frequently, and 25 healthy controls (HCs) (n = 116). Twelve out of 91 patients and 5 out of 25 HCs were excluded due to multiple polyps observed in upper GI endoscopy, bleeding during biopsies, or sedation problems. Finally, a total of 79 patients, including 21 ERD A/B, 18 ERD C/D, 17 true NERD, 11 RH, and 12 FH, and 20 HC, were included in the study (Table 1). The study protocol is shown in Figure 1.
Figure 1. Study Schema. GERDQ, gastroesophageal reflux disease questionnaire; QoLRAD, Quality of Life in Reflux and Dyspepsia; PPI, proton pump inhibitor; GI, gastrointestinal; HRM, high-resolution manometry; 24-hr MII-pH, 24-hour multichannel intraluminal impedance-pH; ERD A/B, erosive reflux disease grade A and B; ERD C/D, erosive reflux disease grade C and D; NERD, nonerosive reflux disease; RH, reflux hypersensitivity; FH, functional heartburn; HC, healthy controls; qRT-PCR, quantitative real-time polymerase chain reaction; ELISA, enzyme-linked immunosorbent assay.
Table 1 . Demographics
| Subjects | Gender | Age (yr) | BMI (kg/m2) |
|---|---|---|---|
| ERD A/B (n = 21) | 13 M, 8 F | 43.0 ± 12.0 | 27.1 ± 2.5 |
| ERD C/D (n = 18) | 12 M, 6 F | 50.1 ± 12.4 | 26.7 ± 2.9 |
| NERD (n = 17) | 12 M, 5 F | 37.6 ± 13.0 | 24.5 ± 3.4 |
| RH (n = 11) | 3 M, 8 F | 39.6 ± 10.7 | 22.5 ± 1.9 |
| FH (n = 12) | 3 M, 9 F | 37.7 ± 10.4 | 21.4 ± 4.3 |
| HC (n = 20) | 7 M, 13 F | 40.5 ± 9.8 | 23.6 ± 3.5 |
BMI, body mass index; ERD A/B, erosive reflux disease grade A and B; ERD C/D, erosive reflux disease grade C and D; NERD, nonerosive reflux disease; RH, reflux hypersensitivity; FH, functional heartburn; HC, healthy controls; M, male; F, female.
No differences found between age and BMI. P > 0.05.
Data are presented as mean ± SD.
All patients completed the GERD questionnaire (GERDQ, Validated Mayo Clinic) and Quality of Life in Reflux and Dyspepsia questionnaires (QoLRAD). HCs and patients underwent high-resolution esophageal manometry (MMS-Laborie, Enschede, the Netherlands) and a MII-pH catheter (MMS-Laborie). The MII-pH catheter was placed 5 cm above the lower esophageal sphincter. The acid exposure time (AET), symptom association probability, and symptom index of the subjects were calculated. An AET value of < 4% indicated physiological GERD, and AET > 6% indicated pathologic GERD.1 Baseline impedance values were calculated by measuring the longest period during the sleeping period at night without any reflux episode or swallowing. All subjects were instructed to avoid eating for at least 8 hours and not to take proton pump inhibitors, histamine H2 receptor antagonists and nonsteroids for 1 week before the procedure.
All the HCs had normal upper GI endoscopy, intraesophageal 24-hour MII-pH, and high-resolution esophageal manometry, and they had a negative history of upper GI disease or surgery. The exclusion criteria for the subjects were primary esophageal motility disorders, Barrett’s esophagus, upper GI surgery, and other disorders that may affect the study, such as cancer (except nonmelanomal skin cancer), chronic renal disorders, severe coronary artery disease, chronic obstructive pulmonary disease, and uncontrolled diabetes mellitus.
All upper GI endoscopies and biopsies were performed by the same gastroenterologist and the same technician. Esophageal biopsy specimens were taken endoscopically from the mucosa without erosions 3-5 cm above the Z-line with biopsy forceps (Radial Jaw 4, opening diameter of 2.8 mm; Boston Scientific, Massachusetts, USA). Seven biopsies were immediately placed in ice cold preoxygenated Ringer’s solution (275-285 mOsmol and pH 7.5). Three biopsies were separated for Ussing chamber studies to measure transepithelial electrical resistance (TEER) and permeability. The remaining biopsies were frozen in liquid nitrogen and then stored at –80℃ for gene expression and protein measurements.
The apical and serosal side of the Ussing chamber (Scientific Instruments, Simmerath, Germany) was filled with Ringer’s solution (pH 7.4 and 37℃). The solution was continuously bubbled with 5% CO2/95% O2 gas. Three biopsies were mounted into 3 mL Ussing chambers modified with a 0.017 cm2 adapter under light microscopy after 30 minutes of calibration. We excluded the chamber if the TEER was < 50 Ω · cm2 at baseline. All experiments were performed under open-circuit conditions. TEER was calculated according to Ohm’s law automatically by the data acquisition system.
After calibration, a fluorescein molecule (376 Da; Sigma Aldrich, St. Louis, MO, USA) at a concentration of 100 mg/mL was added to the apical side of the chamber into 3 mL of Ringer’s solution (the final concentration 1 mg/mL). Every 30 minutes, 100 μL of sample was taken from the basolateral side of the chamber, which was refilled with the same amount of solution after each sampling. We also excluded the chamber in cases showing visual or excessive leakage of fluorescein. Every 30 minutes for 120 minutes, the permeability samples were measured fluorometrically in a fluorescence reader (FLUOstar Omega; BMG Labtech, Ortenberg, Germany) at 480 nm excitation and 520 nm emission. At the end of the experiment, the averaged values of the permeability and TEER measurements of the chambers were calculated for each subject.
Total RNA was isolated from the biopsies using the Aurum Total RNA Mini Kit (Bio-Rad Laboratories, Inc, Hercules, CA, USA) according to the manufacturer’s instructions with a homogenizer (Bioprep-6 Homogenizer; Hangzhou Allsheng Instruments Inc, Hangzhou, Zhejiang, China) after the samples were thawed in RNAzol RT (Sigma-Aldrich, St. Louis, MO, USA) at –20℃. The concentrations and purity of the obtained total RNA samples were determined by measuring their absorbance at 260/280 nm on a NanoDrop Spectrophotometer (Thermo Scientific, Wilmington, DE, USA) using 2 µL of each sample.
Complementary DNA was obtained from total RNA via reverse transcriptase enzyme using an iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Inc) according to the manufacturer’s instructions. Real-time polymerase chain reaction was performed on a LightCycler 480 (Roche Diagnostics Inc, Basel, Switzerland) using iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories, Inc), Human Janus Kinase (JAK)/Signal Transducer and Activator of Transcription (STAT) Signaling Primer Library (Real Time Primers, LLC) and Human NFKappaB Primer Library (Real Time Primers, LLC) according to the manufacturer’s specifications. The actin-beta (ACTB), beta-2-microglobulin (B2M), and ribosomal protein L13a (RPL13A) genes were used as housekeeping genes. Supplementary Table shows the list of the genes.
A Bio-Plex TM Cell Lysis Kit (Bio-Rad Laboratories, Inc) and a homogenizer (Bioprep-6 Homogenizer, Hangzhou Allsheng Instruments, Inc) were used for protein extraction according to the manufacturer’s instructions. The supernatants were divided into aliquots, and the protein amounts were determined by the Lowry method.6 Protein concentration was quantified using bovine serum albumin (200 mg/mL; Sigma-Aldrich). The levels of chemokines and 10 phospho-cell signaling proteins in the aliquots were measured using Bio-Plex Multiplex Immunoassays (Human Chemokine 40-Plex panel, Pro Cell Signaling Phospho 7-plex panel, Pro Cell Signaling Phospho NF-kB p65, and Pro Cell Signaling Phospho p38 MAPK; Bio-Rad Laboratories, Inc). This method uses a capture sandwich immunoassay format via magnetic antibody-coupled beads.
The 2-ΔΔCt method was used for the quantitative analysis of genes. The corresponding gene expression levels in each group were compared. Normalizations were made with 3 housekeeping genes. As a result of pairwise comparisons, genes with P-values < 0.05 and fold change ≥ 1.5 were included in the evaluation. Statistical analyses were performed using ANOVA, Student’s t test (parametric data) and the Mann–Whitney U test (nonparametric data) (SPSS statistics version 25.0; IBS Corp, Armonk, NY, USA). A P-value of < 0.05 was considered statistically significant. The mean ± SD values were used for parametric values because median and variance values were given for nonparametric tests.
The TEER values of ERD A/B and ERD C/D were significantly lower than those of HCs. No significance was found in other groups compared to HCs. Additionally, the mucosal permeability of the ERD C/D and RH groups was significantly increased compared to that of HCs (Table 2).
Table 2 . Transepithelial Resistance and Permeability Results
| Subjects | TEER | Permeability (pmol) |
|---|---|---|
| ERD A/B | 132.0 ± 34.9a | 48.3 ± 25.2 |
| ERD C/D | 135.2 ± 44.6a | 58.0 ± 33.6a |
| NERD | 172.3 ± 48.3 | 46.4 ± 16.6 |
| RH | 174.9 ± 56.2 | 65.2 ± 23.8a |
| FH | 178.8 ± 43.6 | 33.5 ± 15.0 |
| HC | 176.9 ± 40.3 | 39.0 ± 14.0 |
aVersus HC. P < 0.05. Student’s t test.
TEER, transepithelial resistance; ERD A/B, erosive reflux disease grade A and B; ERD C/D, erosive reflux disease grade C and D; NERD, nonerosive reflux disease; RH, reflux hypersensitivity; FH, functional heartburn; HC, healthy controls.
Data are presented as mean ± SD.
In the ERD A/B group, epidermal growth factor receptor (EGFR) (4.9-fold), and colony stimulating factor 2 (granulocyte-macrophage; CSF2) (2.7-fold) expression was increased compared to that in HCs, but interferon-gamma (IFN-G) (5.2-fold), interferon (alpha, beta, and omega) receptor 1 (IFNAR1), minichromosome maintenance complex component 5 (MCM5), protein inhibitor of activated STAT-1 (PIAS1), suppressor of cytokine signaling 2 (SOCS2), and tumor necrosis factor-alpha–induced protein 3 (TNF-AIP3) expression was decreased compared to that in HCs (Fig. 2A). In the ERD C/D group, matrix metallopeptidase 9 (MMP9) (15.1-fold), Harvey Rat sarcoma virus (HRAS), interleukin-4 receptor (IL-4R), proteasome (prosome, macropain) subunit beta type-5 (PSMB5), tumor necrosis factor receptor superfamily member 1A (TNFRSF1A), and TNF-AIP3 expression was increased compared to that in HCs, while Fas ligand (TNF superfamily, member 6) (FASLG) (4-fold), B-cell lymphoma 2 (BCL2), CREB binding protein (CREBBP), FASLG, epidermal growth factor (EGF), fibroblast growth factor 2 (basic) (FGF2), Heme oxygenase (decycling) 1 (HMOX1), MCM5, V-myc myelocytomatosis viral oncogene homolog (avian) (MYC), and prolactin receptor (PRLR) expression was significantly reduced compared to that in HCs (Fig. 2B). In the NERD group, MYC, MCM5, Fas (TNFRSF6)-associated via death domain (FADD), mitogen-activated protein kinase kinase kinase 1 (MAP3K1), IFNAR1, interferon gamma receptor 1 (IFNGR1), JAK1, proteasome (prosome, macropain) 26S subunit, ATPase, 5 (PSMC5), and PIAS1 expression was significantly decreased compared to that in HCs (Fig. 2C). In the RH group, EGFR expression (5-fold) was significantly increased compared to that in HCs, but BCL2; chemokine (C-C motif) ligand 2 (CCL2); inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase beta (IKBKB); TANK binding kinase 1 (TBK1); transmembrane emp24 protein transport domain containing 4 (TMED4); FGF2; interleukin 1 receptor, type I (IL1R1); protein inhibitor of activated STAT, 2 (PIAS2); TNF receptor superfamily, member 10a (TNFRSF10A); signal transducer and activator of transcription 3 (acute-phase response factor) (STAT3); and TNFRSF1A expression was significantly decreased compared to that in HCs (Fig. 2D). In the FH group, EGFR expression (5.1-fold) was significantly increased compared to that in HCs, but CSF3 (19.3-fold), IFNGR1, IL10, CHUK, PIAS2, JUN, TBK1, BIRC2, CASP1, MALT1, TNFRSF1A, and FOS expression was significantly decreased compared to that in HCs (Fig. 2E). Supplementary Table shows the meaning of the abbreviations.
Figure 2. Messenger RNA expressions of phenotypes. All comparisons are compared with healthy controls. Graphing data expressed as fold changes, P < 0.05. (A) Significant differences of erosive reflux disease grade A and B (ERD A/B), (B) significant differences of erosive reflux disease grade C and D (ERD C/D), (C) significant differences of nonerosive reflux disease (NERD), (D) significant differences of reflux hypersensitivity (RH), (E) significant differences of functional heartburn (FH).
In the ERD A/B group (mild erosive reflux group), the IL-8 (P < 0.001), IL-6, 6CKINE/CCL21, granulocyte-macrophage colony-stimulating factor (GM-CSF), TARC/CCL17, MPIF1/CCL23, MCP1/CCL2, and SDF1-a+b/chemokine (C-X-C motif) ligand 12 (CXCL12) levels were significantly increased compared to those in the HC group (P < 0.05) (Table 3). In the ERD C/D group (severe erosive reflux group), the IL-8 (P < 0.001), IL-6 (P < 0.01), IL-1B, TARC/CCL17, GROA/CXCL1, and MCP4/CCL13 levels were significantly higher than those in the HC group (P < 0.05) (Table 3). In addition, the SCYB16/CXCL16 levels were significantly decreased compared to those in the HC group (P < 0.05).
Table 3 . Protein Levels of Cytokines
| Mean ± SD (pg/mL) | Median | Variance | ||
|---|---|---|---|---|
| Homeostatic and regulatory cytokines | ||||
| 6CKINE/CCL21 | ERD AB | 148.0 ± 152.2 | 77.0a | 23 174.1 |
| ERD CD | 74.0 ± 111.0 | 29.9 | 12 317.4 | |
| NERD | 91.6a ± 71.9 | 86.9 | 5165.3 | |
| RH | 96.1 ± 95.1 | 44.7a | 9042.6 | |
| FH | 124.9a ± 114.9 | 96.6 | 13 206.2 | |
| HC | 48.9 ± 67.7 | 22.4 | 4588.6 | |
| GMCSF | ERD AB | 40.4 ± 41.1 | 23.3a | 1692.4 |
| ERD CD | 51.5 ± 100.6 | 13.8 | 10 119.3 | |
| NERD | 13.0 ± 10.5 | 10.2 | 110.0 | |
| RH | 35.1 ± 44.7 | 15.1a | 2001.8 | |
| FH | 27.6 ± 20.0 | 27.1a | 398.0 | |
| HC | 13.0 ± 8.6 | 11.1 | 74.0 | |
| IL-6 | ERD AB | 3.3a ± 4.1 | 2.3 | 17.2 |
| ERD CD | 3.3b ± 1.8 | 2.7 | 3.4 | |
| NERD | 2.2 ± 1.9 | 2.0 | 3.4 | |
| RH | 2.4a ± 1.4 | 2.0 | 2.0 | |
| FH | 2.8 ± 1.5 | 2.6a | 2.3 | |
| HC | 1.6 ± 0.7 | 1.6 | 0.5 | |
| MIP3-b/CCL19 | ERD AB | 52.7 ± 50.1 | 39.8 | 2508.9 |
| ERD CD | 60.4 ± 89.4 | 25.8 | 7988.2 | |
| NERD | 27.5 ± 26.4 | 20.6 | 696.6 | |
| RH | 47.7 ± 34.8 | 35.8 | 1211.9 | |
| FH | 57.5 ± 42.3 | 54.6a | 1787.4 | |
| HC | 31.8 ± 23.3 | 23.4 | 544.5 | |
| SCYB16/CXCL16 | ERD AB | 16.8 ± 15.3 | 10.5 | 233.3 |
| ERD CD | 8.3a ± 9.3 | 4.8 | 86.3 | |
| NERD | 11.7 ± 12.7 | 7.3 | 161.4 | |
| RH | 14.5 ± 12.4 | 8.4 | 153.5 | |
| FH | 15.9 ± 11.3 | 14.0 | 127.5 | |
| HC | 13.9 ± 7.7 | 12.2 | 59.8 | |
| SDF1-a+b/CXCL12 | ERD AB | 85.3 ± 91.9 | 53.0a | 8441.1 |
| ERD CD | 89.8 ± 159.4 | 42.8 | 25 397.2 | |
| NERD | 41.9 ± 27.6 | 34.0 | 759.3 | |
| RH | 115.5 ± 94.8 | 84.1a | 8982.9 | |
| FH | 70.2 ± 59.2 | 51.2 | 3501.7 | |
| HC | 40.4 ± 20.2 | 33.8 | 408.6 | |
| Proinflammatory and inflammatory cytokines | ||||
| GRO-a/CXCL1 | ERD AB | 13.2 ± 17.5 | 4.1 | 307.9 |
| ERD CD | 36.2 ± 52.1 | 11.5a | 2715.5 | |
| NERD | 24.7 ± 44.3 | 3.5 | 1962.9 | |
| RH | 16.8 ± 7 | 16.8 | 49.3 | |
| FH | 7.1 ± 3.3 | 6.1 | 11.0 | |
| HC | 3.9 ± 3.1 | 2.4 | 9.5 | |
| MCP1/CCL2 | ERD AB | 21.3 ± 18.9 | 18.5a | 358.3 |
| ERD CD | 16.5 ± 12.4 | 12.6 | 154.1 | |
| NERD | 13.6 ± 9.5 | 10.3 | 90.3 | |
| RH | 17.3 ± 12.7 | 11.5 | 162.6 | |
| FH | 23.1 ± 24.5 | 15.2 | 599.9 | |
| HC | 13.0 ± 20.9 | 7.7 | 437.9 | |
| MIP1d/CCL15 | ERD AB | 8.5 ± 6.9 | 5.4 | 47.8 |
| ERD CD | 9.4 ± 10.0 | 4.3 | 100.0 | |
| NERD | 4.4a ± 3.5 | 2.9 | 12.5 | |
| RH | 6.4 ± 5.2 | 4.7 | 26.6 | |
| FH | 5.0 ± 3.4 | 2.9 | 11.6 | |
| HC | 7.4 ± 5.3 | 6.8 | 28.4 | |
| IL-1B | ERD AB | 6.3 ± 3.9 | 5.1 | 15.4 |
| ERD CD | 8.6a ± 7.0 | 5.7 | 49.0 | |
| NERD | 5.5 ± 3.0 | 4.7 | 9.0 | |
| RH | 6.7 ± 3.1 | 6.3 | 9.5 | |
| FH | 6.9 ± 5.7 | 5.1 | 32.5 | |
| HC | 5.6 ± 2.5 | 5.1 | 6.1 | |
| IL-2 | ERD AB | 2.0 ± 1.8 | 1.5 | 3.4 |
| ERD CD | 1.7 ± 1.4 | 1.3 | 1.9 | |
| NERD | 2.0 ± 1.1 | 2.0 | 1.2 | |
| RH | 1.8 ± 1.3 | 1.5 | 1.6 | |
| FH | 3.2a ± 1.8 | 3.4 | 3.2 | |
| HC | 1.8 ± 0.9 | 1.7 | 0.8 | |
| IL-8 | ERD AB | 35.6 ± 48.8 | 20.7c | 2382.2 |
| ERD CD | 60.0 ± 114.8 | 11.6c | 13 187.1 | |
| NERD | 43.2 ± 113.7 | 5.2 | 12 926.6 | |
| RH | 26.2 ± 35.4 | 9.2 | 1252.4 | |
| FH | 10.6 ± 14.0 | 3.3 | 196.5 | |
| HC | 7.5 ± 16.9 | 3.0 | 286.6 | |
| MPIF1/ CCL23 | ERD AB | 15.7 ± 15 | 9.3 | 224.3 |
| ERD CD | 9.4 ± 8.3 | 9.3 | 69.4 | |
| NERD | 8.5 ± 6.2 | 7.9 | 38.9 | |
| RH | 17.9 ± 24.4 | 11.3a | 595.1 | |
| FH | 21.4 ± 26.1 | 14.6a | 682.5 | |
| HC | 7.8 ± 4.9 | 7.7 | 24.1 | |
| Anti-inflammatory cytokines | ||||
| IL-4 | ERD AB | 1.8 ± 1.0 | 2.0 | 0.9 |
| ERD CD | 2.5 ± 1.6 | 2.5 | 2.5 | |
| NERD | 1.6 ± 0.9 | 1.5 | 0.9 | |
| RH | 3.0a ± 2.1 | 2.6 | 4.6 | |
| FH | 3.0 ± 1.7 | 2.7a | 2.7 | |
| HC | 1.8 ± 0.7 | 1.8 | 0.5 | |
| MCP3/ CCL7 | ERD AB | 12.2 ± 8.0 | 8.8 | 64.5 |
| ERD CD | 15.1 ± 6.7 | 15.2 | 44.9 | |
| NERD | 11.4 ± 5.1 | 12.0 | 26.2 | |
| RH | 18.7a ± 7.0 | 15.2 | 48.7 | |
| FH | 18.1 ± 10.5 | 18.3 | 109.3 | |
| HC | 13.5 ± 6.4 | 12.0 | 41.5 | |
| MCP4/ CCL13 | ERD AB | 5.0 ± 2.5 | 5.5 | 6.2 |
| ERD CD | 8.6 ± 8.5 | 4.9 | 71.7 | |
| NERD | 4.4 ± 2.8 | 4.6 | 7.9 | |
| RH | 5.2 ± 3.9 | 4.3 | 15.6 | |
| FH | 9.0 ± 6.8 | 5.7a | 46.3 | |
| HC | 4.2 ± 2.4 | 3.5 | 5.7 | |
| TARC/ CCL17 | ERD AB | 0.6 ± 0.5 | 0.4 | 0.3 |
| ERD CD | 0.7a ± 0.6 | 0.7 | 0.3 | |
| NERD | 0.5 ± 0.3 | 0.4 | 0.1 | |
| RH | 0.5 ± 0.3 | 0.5 | 0.1 | |
| FH | 0.6a ± 0.3 | 0.6 | 0.1 | |
| HC | 0.4 ± 0.3 | 0.3 | 0.1 | |
aP < 0.05, bP < 0.01, cP < 0.001.
6CKINE/CCL21, chemokine (C-C motif) ligand 21; GMCSF, granulocyte-macrophage colony-stimulating factor; GRO-a/CXCL1, chemokine (C-X-C motif) ligand 1, Growth-regulated alpha protein; IL-1B, interleukin 1 beta; MCP1/CCL2, chemokine (C-C motif) ligand 2, monocyte chemoattractant protein 1; MCP3/CCL7, chemokine (C-C motif) ligand 7, monocyte-chemotactic protein 3; MCP4/CCL13, chemokine (C-C motif) ligand 13, monocyte-chemotactic protein 4; MIP1d/CCL15, chemokine (C-C motif) ligand 15, macrophage inflammatory protein 1d; MIP3-b/CCL19, chemokine (C-C motif) ligand 1, macrophage inflammatory protein 3b; MPIF1/CCL23, chemokine (C-C motif) ligand 23, myeloid progenitor inhibitory factor 1; SCYB16/CXCL16, chemokine (C-X-C motif) ligand 16, small-inducible cytokine B16; SDF1-a+b/CXCL12, chemokine (C-X-C motif) ligand 12, the stromal cell-derived factor 1 a+b; TARC/CCL17, chemokine (C-C motif) ligand 17, thymus and activation-regulated chemokine; ERD, erosive reflux disease; NERD, nonerosive reflux disease; RH, reflux hypersensitivity; FH, functional heartburn; HC, healthy controls.
Mean ± SD values are given for parametrically significant proteins; median and variance values are given for nonparametrically significant proteins. All comparisons were performed versus healthy controls. Mean ± SD values are given for Student’s t test; median and variance values are given for Mann–Whitney U test.
The mitogen activated protein kinase kinase 1 (MEK1) levels were significantly increased in the ERD A/B and ERD C/D groups compared to those in the HC group (P < 0.05) (Table 4). The nuclear factor kappa B (NF-κB) inhibitor alpha (IκB-α) (P < 0.01), Transcription factor Jun (c-JUN), and mechanistic target of rapamycin (mTOR) levels in the ERD C/D group were significantly decreased compared to those in the HC group (P < 0.05).
Table 4 . Levels of Cell-signaling Proteins
| Mean ± SD (FI) | Median | Variance | ||
|---|---|---|---|---|
| IκB-α | ERD AB | 71.4 ± 40.7 | 69.0 | 1659.6 |
| ERD CD | 45.5b ± 26.7 | 35.5 | 713.1 | |
| NERD | 69.1 ± 60.8 | 56.0 | 3702.6 | |
| RH | 59.5 ± 25.8 | 56.0 | 664.6 | |
| FH | 88.3 ± 68.8 | 54.3 | 4730.7 | |
| HC | 86.5 ± 58.6 | 76.5 | 3434.3 | |
| MEK1 | ERD AB | 483.7 ± 707.4 | 281.0a | 500 414.4 |
| ERD CD | 463.6 ± 749.4 | 270.8a | 561 575.2 | |
| NERD | 177.4 ± 164.1 | 116.0 | 26 941.6 | |
| RH | 103.3 ± 88.1 | 99.0 | 7758.8 | |
| FH | 125.2 ± 107.5 | 98.5 | 11 557.2 | |
| HC | 161.9 ± 151.4 | 102.5 | 22 920.5 | |
| c-Jun | ERD AB | 44.9 ± 15.0 | 41.0 | 224.1 |
| ERD CD | 40.9a ± 14.2 | 41.5 | 200.4 | |
| NERD | 42.5a ± 8.2 | 44.0 | 67.7 | |
| RH | 49.8 ± 13.5 | 46.0 | 182.5 | |
| FH | 56.2 ± 26.2 | 47.5 | 687.7 | |
| HC | 50.9 ± 11 | 50.5 | 121.1 | |
| mTOR | ERD AB | 85.1 ± 43.5 | 83.5 | 1895.7 |
| ERD CD | 59.1 ± 34.6 | 48.0a | 1198.6 | |
| NERD | 90.9 ± 68.6 | 64.0 | 4706.5 | |
| RH | 88.0 ± 80.8 | 44.0 | 6522.5 | |
| FH | 132.4 ± 93.6 | 128.5 | 8761.6 | |
| HC | 103.1 ± 80.5 | 75.0 | 6479.8 | |
aP < 0.05, bP < 0.01.
c-Jun, transcription factor Jun; IκB-α, NF kappa B inhibitor alpha; MEK1, mitogen activated protein kinase kinase 1; mTOR, mechanistic target of rapamycin; ERD, erosive reflux disease; NERD, nonerosive reflux disease; RH, reflux hypersensitivity; FH, functional heartburn; HC, healthy controls.
Mean ± SD values are given for parametrically significant proteins; median and variance values are given for nonparametrically significant proteins. All comparisons were performed versus healthy controls. Mean ± SD values are given for Student’s t test; median and variance values are given for Mann–Whitney U test.
The levels of 6CKINE/CCL21 were significantly increased in NERD (P < 0.05). The MIP1d/CCL15 and c-JUN levels were significantly decreased compared to those in the HC group (P < 0.05) (Table 3). There was no significant difference in other chemokines. MCP3/CCL7, IL-4, IL-6, IL-8, SDF1-a+b/CXCL12, MPIF1/CCL23, GM-CSF, and 6CKINE/CCL21 were significantly higher in the RH group than in the control group (P < 0.05) (Table 3). In patients with FH, 6CKINE/CCL21, IL-2, TARC/CCL17, GM-CSF, IL-6, IL-4, MCP4/CCL13, MPIF1/CCL23, MIP3-b/CCL19, and SDF1-a+b/CXCL12 were significantly increased compared to HCs (P < 0.05) (Table 3). Supplementary Table shows the meaning of the abbreviations.
In the present study, the inflammatory processes and electrophysiological characteristics that may play a possible role in the pathogenesis of different phenotypes associated with GERD were investigated. This study demonstrated that even a small amount of acid exposure is sufficient to trigger and propagate the inflammatory response in the human esophageal mucosa. Symptoms appear as a result of inflammation in the esophageal mucosa after irritating harmful agents reach the esophagus. The severity of GERD may be related to pH and pepsin in gastric juice.7 However, these agents are likely to affect epithelial integrity as well.3
Loss of mucosal integrity can be measured in endoscopic biopsies by transepithelial resistance and epithelial permeability in an in vitro Ussing chamber system.8 Although the Ussing chamber is the gold standard for measuring mucosal integrity in vitro, there is no reference range due to variations related to the brands or technical experience. Studies in the literature on tissue resistance and permeability of the esophageal epithelium have been performed in limited numbers of patients, mostly within all 4 groups of patients with erosive esophagitis versus all NERD phenotypes combined in 1 group and HCs. The present study was the first to investigate all GERD phenotypes, including ERD A/B, ERD C/D, NERD, RH, and FH, and compare them to HCs.
In accordance with our data, Rinsma et al9 observed that tissue resistance significantly decreases in the ERD group compared to the HC group but that there is no difference between ERD and NERD. Similarly, it has been reported that there is no significant change in tissue resistance and tissue permeability between NERD and HC.9 Another study, in which erosive patients were excluded, has reported that there is no significant difference between the tissue resistance and permeability of NERD and FH patients.10 In the same study, the authors stated that the tissue permeability of HCs is significantly lower than that of ERD and NERD but that there is no difference between ERD and NERD. In addition, no difference in the TEER has been reported between ERD and NERD.11 The fact that there is higher tissue permeability and lower resistance in the ERD groups indicates that tissue integrity is impaired.9 Tissue resistance in the NERD, RH, and FH groups was similar to that in HCs, revealing that the mucosal barrier is preserved in these groups. Disruption of the mucosal barrier integrity in ERD groups can provide information about the presence of a severe inflammatory process in that region, but “microinflammation” may be the root of the sensitivity in patients with NERD, esophageal sensitivity, and FH. Acid (and possibly pepsin) can directly or indirectly activate acid-sensitive receptors on mucosal afferent nerves and epithelial cells in the esophagus through the acidic microenvironment created by cytokine-induced inflammation, such as the release of protons from exocytic granules and lysosomes of immune cells.12
In the present study, IL-6 and IL-8 protein levels were significantly higher in groups with erosive esophagitis than in controls. IL-8, a potent neutrophil marker, is a highly secreted chemoattractant in oxidative damage caused by oxygen radicals.13,14 In addition to many cell culture and experimental model studies, it is known that serum or tissue levels of IL-8 increase in patients with GERD.15,16 IL-6, which is an important mediator in the inflammatory process, has proinflammatory and anti-inflammatory properties.17 Moreover, IL-6 is secreted from cells due to NF-κB activation via receptors in the case of acute inflammation. Although the IL-8 (ERD A/B and ERD C/D) and MCP-1 (only ERD A/B) levels were significantly increased in ERD groups compared to the HC group, no difference was observed according to the severity of esophagitis. MCP-1/CCL2, a proinflammatory chemokine, is a potent chemoattractant for monocytes and activates macrophages and T cells.18 IL-8 and MCP-1 levels are significantly increased in esophageal biopsies of patients with ERD, but the MCP-1 levels are not proportional to the severity of esophagitis.16 It is known that activated neutrophils increase the proinflammatory process by secreting MPIF1/CCL23. MPIF1/CCL23 acts by upregulating MCP-1/CCL2, which activates lymphocytes, monocytes, and macrophages in the inflammatory process.19 Migrated neutrophils also interacts with MCP-1/CCL2, thereby mediating the attraction of monocytes and lymphocytes to the inflammatory region.20 IL-6 induces MCP-1 secretion to promote the accumulation of mononuclear cells to the damaged area in the inflammatory process.21 All these MCP-1 sources might explain the increase in the ERD A/B group. CXCL12, which was highly expressed in ERD A/B group, is secreted from epithelial cells, and has homeostatic effects. CXCL12 plays an important role in neutrophil activation.22 Moreover, CXCL12 contributes to the increase in MCP-1 and IL-8 levels.23 The significant increase in these parameters showed that the proinflammatory response is constantly triggered in the ERD A/B group.
Both chronic and acute inflammatory processes take place in the esophageal mucosa in GERD. In addition to the proinflammation mentioned above, inflammatory resolution also simultaneously operates in erosive groups. Increasing the number of macrophages in the tissue plays an important role in the resolution of the inflammatory process with the process of chronic inflammation or reorganization.24 Related to this information, we showed that TARC/CCL17 and 6KINE/CCL21 levels were increased in the ERD A/B group. TARC/CCL17 may be secreted from activated macrophages. Although T helper 2 (Th2) cells have anti-inflammatory properties through the activation of M2 macrophages, they can also invoke memory Th1 cells and regulatory T cells.25 With these mechanisms, TARC/CCL17 has both inflammatory and homeostatic chemokine effects.26 6KINE/CCL21 is mainly secreted from lymphatic endothelial cells and is chemotactic for T cells. 6KINE/CCL21 mostly mediates the transport of lymphocytes to secondary lymphoid tissues with CCL19 in chronic inflammation.26,27
EGFR mRNA expression and MEK1/2 (Ser217/221) levels were also significantly increased in the ERD A/B. MEK1/2 is activated when phosphorylated at Ser217 and Ser221, and it is involved in the regulation of growth factors.28 MEK activation may be due to EGFR signaling. EGFR is produced in many epithelial cells throughout the digestive tract and has an important role in cell proliferation, differentiation, migration and repair processes.29 Some proinflammatory cytokines may be secreted from basal epithelial cells after IL-6 stimulation via the EGFR/Raf-1/MEK/extracellular signal-regulated kinase (ERK) pathway.30 EGFR signaling can also be activated in the regeneration phase in the resolution process as well as in ongoing chronic inflammation if resolution is not successful.31 As in chronic inflammatory diseases, the fibrotic process might be active in chronic GERD. After acid peptic damage, extracellular matrix deposition as well as tissue fibroblast activation and changes occur in the esophageal mucosa, leading to chronic inflammation. This process may occur via the EGFR/Raf-1/MEK/ERK pathway.30 In a histological study, Pretto et al29 stated that as the severity of the disease increases, EGFR expression increases in patients with GERD, Barrett’s esophagus and adenocarcinoma.
As in the ERD A/B, the acute and chronic inflammatory processes worked together in the severe esophagitis (ERD C/D). However, the tissue healing process may be overwhelmed by severe pathological acid exposure and worsened epithelial damage. IL-6 and IL-8 levels were increased in the ERD C/D group, but the difference was more significant than ERD A/B. Tissue damage was the worst in the ERD C/D as a result of the strong proinflammatory effects of IL-6 and IL-8. In addition to the drastic neutrophil chemotaxis of IL-8,13 GRO-A/CXCL1 with IL-8-like functions is also a chemoattractant for many immune cells, especially neutrophils.32
IL-1B, which was increased in the ERD C/D group, is a proinflammatory cytokine expressed in activated macrophages, monocytes and neutrophils, and it plays a role in the proliferation and differentiation of cells.33 IL-1B increases the gene expression of inducible nitric oxide synthase and cyclooxygenase-2, and the expression of adhesion molecules in the inflammatory process exacerbates damage in acute and chronic inflammation.18 IL-1B activation can occur via the classical NF-κB pathway, but it is also a potent activator of the NF-κB.34 IL-1B is a cytokine that has also been studied in esophageal cell lines and GERD patients.12 Exposure of human esophageal cells to bile acid leads to the inflammatory process due to IL-1B and IL-8 secretion, which results in chemotaxis of neutrophils and lymphocytes instead of death.4 This highlighted study by Souza et al4 has formed the basis of the “cytokine leak” reflux model of GERD pathogenesis.
IκB-α protein levels in the ERD C/D group were significantly lower than those in the HC group. Although NF-κB increased at the mRNA and protein levels in the present study, the difference was not significant. However, downregulation of IκB-α, as an activation indicator, is an important factor because NF-κB is associated with many pathways. The H-ras gene, which had significantly increased expression in the ERD C/D, is also required for IκB-α degradation and the transcriptional function of NF-κB. Therefore, upregulation of the H-ras gene is one of the indicators of NF-κB activity.35 In addition, the active Ras gene binds to Raf1 (which was increased in the ERD C/D) activates the MEK/ERK pathway. The MEK/ERK pathway is involved in cell growth, adhesion, differentiation, and cytoskeletal reorganization. In our study, the increase in MEK1 in both the ERD A/B and ERD C/D indicated that the MEK/ERK pathway is active. Activated ERK can enter the nucleus and secrete many cytokines. Therefore, activation of the H-ras gene in these groups, where oxidative and acidic damage is most severe, may have led to the secretion of many cytokines, which was increased in our study via the NF-κB pathway via IL-1B.36 At the same time, this regulation and cytokines may have been released via activation of the ras/raf/MEK/ ERK pathway and the effect of IL-6.30 H-Ras activation increases matrix metallopeptidase 9 (MMP9) expression via the NF-κB and MEK pathways.37 MMP9 is effective in local proteolysis of the extracellular matrix and leukocyte migration.38 In an esophageal epithelial organoid model, low pH human gastric juice has been reported to elevate MMP9 expression.39 MMP9 expression increases as esophageal tissue damage or dysplasia rises. In the present study, the robust increase in MMP9 mRNA expression (15.1-fold compared to control) supported the role of MMP9 in the ERD C/D group, which had the most intense mucosal damage among the phenotypes. MMP9 increases endothelial permeability by degrading occludin, an endothelial tight junction protein, which leads to inflammation of the tissue.38 Severe barrier damage in the inflammatory process in the ERD C/D may have increased permeability through MMP9 activity. However, the inflammatory response may have triggered the increase in MMP9.
MCP4/CCL13 secreted from epithelial, endothelial, and muscle cells induce the secretion of several proinflammatory cytokines. CCL13, which is a chemoattractant for monocytes and T lymphocytes, can be produced via the NF-kB pathway under the influence of proinflammatory cytokines. During inflammation, the activity of CCL13 is regulated by proteases and MMPs.38 Together with IL-6, CCL13 is a key regulator of the uptake of immune cells in chronic inflammation.40 Although acute and chronic inflammation processes progress together, chronic damage was more prominent in the ERD C/D.
Although pathological acid exposure causes erosion in esophageal tissues, some patients have visibly normal epithelium with pathological acid exposure, such as the NERD. Intense acid peptic attack is expected to cause tissue damage, but the integrity of the epithelial barrier is not impaired. The pathogenesis of this group is less known than that of esophageal esophagitis. In some studies, enlargements in the intercellular spaces have also been shown in the NERD group.41 These dilatations may occur due to the immune cells coming to the epithelial surface as a result of proinflammatory chemokines secreted by epithelial cells. Increased protein levels of 6KINE/CCL21 in NERD indicate high T cell chemotaxis. The increase in CCL21 may be responsible for the balance between acid exposure and epithelial integrity in NERD.
MIP1d/CCL15 promotes the recruitment of immune cells into the inflammatory region and the induction of angiogenesis.42 The strongest triggers of this chemokine, which was significantly decreased in the NERD group in the present study, are tumor necrosis factor alpha (TNF-A) and IFN-G. This triggering occurs with the activation of NF-κB and AP-1 (c-Jun and c-fos).43 However, we found that c-Jun, the most important subunit of AP-1, was significantly decreased in the NERD group. MAP3K1 downregulation may be responsible for decreased c-Jun expression.44 A significant decrease in c-Jun levels and stable NF-κB levels in the NERD group were consistent with a decrease in CCL15. This angiogenesis-related protein may be suppressed because there is no mucosal damage in NERD. Therefore, these findings suggested that acid accumulation is not accompanied by proinflammatory processes in the NERD group. Normal physiological processes are maintained because cytoprotective mechanisms may be activated.
RH and FH are the most notable groups in pathogenesis and diagnosis. It remains controversial whether FH is a part of the GERD spectrum, but RH is accepted as a phenotype of GERD in Porto Consensus.45 In the present study, EGFR expression was increased in both the RH and FH groups. It has been reported that EGFR signaling is increased in esophageal cells treated with pH 5.0 gastric + bile acid, and squamous cell differentiation is decreased as a mucosal repair against the reflux content of EGFR.46 In addition, the pleiotropic cytokines, IL-6 and GM-CSF, as well as homeostatically effective CXCL12, chemotactic CCL23, and CCL21 was increased in both groups. Overall, these results suggested that in the FH and RH groups, the migration of immune cells with chemokines was similar to that in the ERD A/B group. However, the fact that MEK1 levels did not increase in these groups explains the absence of MEK1/MAP2K1 activation and, thus, the absence of pathological changes in the cytoskeleton.
Similar to the ERD A/B, the IL-8 levels were significantly increased in the RH. It has been reported that IL-8 is increased in reflux patients without endoscopic erosion.47 Therefore, reflux content can stimulate the production of proinflammatory cytokines without causing mucosal damage. However, IL-1B did not increase in the RH, and its receptor, IL1R, and CCL2 expressions were also significantly decreased in the RH group. The absence of reflux exposure at pathological levels in the RH suggested that the proinflammatory response was not induced excessively in the epithelium.
IL-4 is the key cytokine of the adaptive immune response and an anti-inflammatory cytokine that functions by suppressing the proinflammatory environment.48 IL-4 is frequently expressed in epidermal hypersensitivity.49 In a previous study, there was no difference in the IL-4 levels of reflux patients whose serum cytokine profiles were examined before and after proton pump inhibitor treatment.50 It has been shown that mast cell infiltration, which is an important source of IL-4, increases in patients with reflux hypersensitivity.51 Tryptase and histamine released after mast cell activation can activate enteric nerves, leading to neuronal hyperexcitability.52 Mast cells may be the cause of hypersensitivity in these groups. IL-4 in inflammatory sites in peripheral tissues induces the production of MCP3/CCL7 in Th2 immunity. Although CCL7 mediates the migration of proinflammatory monocytes to the inflammatory region in peripheral tissue, it exhibits more anti-inflammatory properties.27 The arrival of Th2 cells to the inflammatory region creates a positive feedback loop by secreting more IL-4. We found that CCL7 responses, IL-4 and MCP3/CCL7 were also significantly increased in the RH that demonstrates evidence of neuronal hyperexcitability.
Although GM-CSF/CSF2 protein levels increased in the FH group, CSF3 mRNA expression was dramatically decreased (–19.3-fold). While CSF2 induces both macrophage and neutrophil activation,53 CSF3 is more involved in neutrophil activation.54 This decrease in CSF3 expression indicated that the acute inflammatory response was lower in the FH group. In our study, the IL-2 levels were significantly increased only in the FH group. In the present study, high levels of IL-2 were found in FH patients who did not have acid attacks as indicated by the 24-hour pH-impedance monitoring. IL-2 is involved in the growth and activation of T cells, B cells and macrophages. IL-2 also induces the production of proinflammatory cytokines.18 IL-2 and IL-4 are regulators of adaptive immunity and play a role in the activation phase of the T cell-dependent immune response, lymphocyte growth/differentiation and local tissue damage. CCL13 was increased in the FH group. CCL13 induces the secretion of proinflammatory cytokines, and it is also a chemoattractant for monocytes and T lymphocytes. In addition, CCL13 is involved in the Th2 response.27 In addition, CCL17, which has homeostatic and inflammatory effects, was also significantly increased in the FH group. CCL17 can invoke Th1 cells and regulatory T cells, but its main task is the Th2 response.25,26 The high Th2 response suggested that the anti-inflammatory process was dominant in the FH.
In the present study, the accuracy of all patients and HCs was ensured with 24-hour pH-impedance monitoring, and upper GI endoscopy increased the reliability of the study. The well-known cytokines in GERD are IL-1B, IL-8, IL-6, TNF-A, and MCP-1. The majority of our results in patients with pathological reflux supported the literature. However, many inflammatory parameters with significant differences were shown for the first time in the human esophageal epithelium. The limitation of our study was the inability to histologically evaluate the immune cells because the biopsies taken during upper GI endoscopy are not suitable to demonstrate the epithelial layers. In addition, immune cells are concentrated in the lamina propria layer where the capillaries are dense, but biopsies are taken more superficially.
In conclusion, pathological or physiological reflux attacks in the human esophageal epithelium produce a T lymphocyte-dominant inflammatory response. Although tissue integrity is preserved in groups with normal acid exposure, such as hypersensitive esophagus and FH, as in healthy individuals, the increase in cytokines is almost as high as in the pathological reflux groups. “Microinflammation” in these patient groups indicates that even a mild trigger is sufficient for the initiation and progression of inflammatory activity. However, the proinflammatory/anti-inflammatory balance is preserved via low acid exposure, and the process does not lead to tissue damage. But mild triggers may increase sensitization by causing nerve endings to become closer to the surface in the mucosa of these patients. The differences between phenotypes may shed light on the potential role of cytokine-directed therapies in GERD. The inflammatory processes in the RH and FH are similar to those in the mild esophagitis group, suggesting that attention should be focused on these 2 groups. In the future, we will investigate the biochemical parameters, such as IL-4, IL-2, and IgE, which may increase hypersensitivity in these groups.
Note: To access the supplementary table 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/jnm22186.
We thank Yasemin Alev and Esra Yildirim for technical assistance. The results of this study were presented at 30th United European Gastroenterology Week 2022 (MP428, Moderated Poster) and NeuroGASTRO 2021 (NGS15594-96, Oral presentation).
This study was supported by The Scientific and Technological Research Council of Turkey/TUBITAK (Project ID: 118S260) and Ege University Scientific Research Projects Coordination (TGA-2021-22732).
None.
Pelin Ergun, Serhat Bor, and Eser Y Sozmen conceived and designed research, edited and revised manuscript, and approved final version of manuscript; Serhat Bor chose the patients and performed Upper GI endoscopies; Pelin Ergun and Sezgi Kipcak performed experiments; Pelin Ergun, Sezgi Kipcak, Nur S Gunel, and Eser Y Sozmen analyzed data; Pelin Ergun, Sezgi Kipcak, and Nur S Gunel interpreted results of experiments; Pelin Ergun prepared figures; and Pelin Ergun and Eser Y Sozmen drafted manuscript.
Author ORCID Information
Pelin Ergun |
 |
