東北大学機関リポジトリTOUR

著者

Kiichi Takahashi, Naoki Asano, Akira Imatani,

Yutaka Kondo, Masashi Saito, Akio Takeuchi,

Xiaoyi Jin, Masahiro Saito, Waku Hatta,

Kiyotaka Asanuma, Kaname Uno, Tomoyuki Koike,

Atsushi Masamune

journal or

publication title

Carcinogenesis

volume

41

number

11

page range

1543-1552

year

2020-02-14

URL

http://hdl.handle.net/10097/00131033

For Peer Review

Sox2 induces tumorigenesis and angiogenesis of early stage esophageal

squamous cell carcinoma through secretion of Suprabasin

Kiichi Takahashi, Naoki Asano, Akira Imatani, Yutaka Kondo, Masashi Saito, Akio Takeuchi, Xiaoyi Jin, Masahiro Saito, Waku Hatta, Kiyotaka Asanuma, Kaname Uno, Tomoyuki Koike, Atsushi Masamune

Division of Gastroenterology, Tohoku University Graduate School of Medicine, Sendai, Miyagi, Japan

Correspondence to: Akira Imatani,

Division of Gastroenterology, Tohoku University Graduate School of Medicine. 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan.

e-mail: [email protected]

Telephone: +81-22-717-7171 / Fax: +81-22-717-7177

Running Head: Sox2 induced angiogenesis via Suprabasin in ESCC

Keywords: Sox2, esophageal cancer, Suprabasin, carcinogenesis, angiogenesis

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Abbreviations

DAPI: 4',6'-diamidino-phenylindole Dox: doxycycline

ESCC: esophageal squamous cell carcinoma HUVEC:human umbilical vein endothelial cell IPCL: intra-papillary capillary loop

MAPK: mitogen-activated protein kinase MVD: microvascular density

Nrp1: neuropilin1

PDGF: platelet-derived growth factor

Q-PCR: quantitative polymerase chain reaction VEGF: vascular endothelial growth factor

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Abstract

Early stage of esophageal squamous cell carcinoma (ESCC) is known to be accompanied by angiogenesis and morphological changes of microvessels. Transcription factor Sox2 is amplified in various cancers including ESCC, but the role of Sox2 in the carcinogenesis and angiogenesis has not been determined. Hence, we aimed to investigate the role of Sox2 in the early stage of ESCC. We found that the expression of Sox2 was significantly higher in early stage ESCC tissues than that in their adjacent normal tissues. We then established Sox2-inducible normal human esophageal squamous cell line (HetSox2) to investigate the role of Sox2 in esophageal carcinogenesis and angiogenesis in vitro. Sox2 overexpression led to increased cell proliferation and spheroid formation. The culture supernatant of Sox2-overexpressing HetSox2 induced migration and sprouting of endothelial cell line HUVEC. As for the mechanism, we found that the expression of secreted protein Suprabasin was directly induced by Sox2. Suprabasin enhanced proliferation of normal human esophageal squamous cells and HUVEC when added to the culture. Moreover, Suprabasin enhanced migration and sprouting of HUVEC cells, which were observed with the culture supernatant of Sox2-oversxpressing HetSox2. This angiogenic effect of Suprabasin was abolished by inhibiting AKT phosphorylation, which suggested its dependence on AKT signaling. Finally, we showed that Suprabasin expression and the density of microvessels were significantly higher in ESCC tissues with high Sox2 expression. Our study suggested that increased Sox2 expression in esophageal squamous cells induced Suprabasin expression, and as a result initiated the

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

carcinogenesis via increased cell proliferation and angiogenesis.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Introduction

Esophageal cancer is the sixth leading cause of cancer death partly due to its difficulty of diagnosis in its early stage (1). Squamous cell carcinoma, which is common in Eastern Asia, and adenocarcinoma are the two major subtypes of esophageal cancer (2,3). In recent years, endoscopic submucosal dissection (ESD) or endoscopic mucosal resection (EMR) has become the major treatment for early esophageal cancers since the frequency of lymph node metastasis is very limited in their early stage (4). However, recognizing the cancerous lesions and sorting out endoscopically resectable lesions by normal white-light endoscopy could be challenging. Therefore, narrow-band imaging (NBI) endoscopy and iodine-staining endoscopy are commonly used for detecting esophageal cancers at their early stages (5,6). NBI is the optical digital method frequently used to detect cancerous lesions by emphasizing the images of microvessels. As for esophageal squamous cell carcinoma (ESCC), magnifying observation under NBI endoscopy allows us to observe the morphological changes of intra-papillary capillary loop vessels, which help us detect the presence of ESCC and predict their depth of infiltration (7,8). These morphological changes of microvessels suggest that some angiogenic factors are playing a role in the early stage of ESCC.

Recent genomic comprehensive analyses of advanced ESCC have revealed that there are genomic mutations and/or amplifications of p53, Notch1, Ccnd1, Pik3ca, and Sox2 (9-11). Sox2 is one of the four transcription factors crucial for producing induced pluripotent stem (iPS) cells (12). Sox2 also

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

contributes to the development of organs such as brain, lung, and stomach (13). In terms of esophagus, Sox2 contributes to its development from the foregut, and to the maintenance of its stratified squamous epithelium (14). On the other hand, recent studies revealed the involvement of Sox2 in carcinogenesis in various organs. Although the suppressed expression of Sox2 is considered to lead to carcinogenesis in gastric cancer (15), enhanced Sox2 expression is found in various cancers such as advanced ESCC (16,17), cutaneous squamous cell carcinoma (18), lung squamous cell carcinoma (16), and breast cancer (19). A previous study reported that Sox2 promoted cell proliferation through AKT/mTORC1 signaling in ESCC (20), and another study reported that Sox2 overexpression was associated with poor prognosis in advanced ESCC patients (21). In addition, Yokota et al reported that suppression of Sox2, using zinc finger-based artificial transcription factor, inhibited tumor growth in ESCC cell lines (22).

Angiogenesis is an essential process in carcinogenesis and ESCC is no exception (23). Angiogenic factors such as vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), and interleukin-8 play a crucial role in the process of endothelial cell proliferation, migration, and vascular formation and cause angiogenesis (24-27). Angiogenic factors are known to induce angiogenesis via activation of mitogen-activated protein kinase (MAPK) signaling and/or AKT signaling (28,29). The correlation between these angiogenic factors and clinical prognosis in advanced ESCC has been reported previously (30,31).

Hence, we aimed to investigate the role of Sox2 in carcinogenesis and angiogenesis in early stage ESCC.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Materials and methods

Patients' information and tissue samples

Forty-seven specimens of early stage ESCC were obtained from patients who underwent ESD at Tohoku University Hospital (Sendai, Japan) between March 2014 - October 2017. Clinicopathological information on the samples is summarized in Supplementary Table 1. This study was conducted under the approval of the Ethical Committee Tohoku University Graduate School of Medicine and examined after obtaining informed consent.

Cell culture

Human normal esophageal squamous cell line Het-1A was obtained from American Type Culture Collection (ATCC, Manassas, VA) where the cell lines are authenticated by short tandem repeat (STR) profiling. Human umbilical vein endothelial cell HUVEC, human embryonic kidney cell HEK293, human ESCC cell lines KYSE30, KYSE70, KYSE180, and KYSE450 (32), were purchased from Health Science Research Resources Bank (Osaka, Japan) where the cell lines are authenticated by STR profiling. Het-1A was cultured in Airway Epithelial Cell Basal Medium (Promo Cell, Heiderburg, Germany) supplemented with Growth Medium Supplement Pack (Promo Cell), 4% Tet system approved fetal bovine serum (FBS) (Clontech, Mountain View, CA). HUVEC was cultured in F-12K medium (Invitrogen life technologies, Carlsbad, CA) supplemented with 10% FBS. HEK293 cells were

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

cultured in DMEM medium (Invitrogen) supplemented with 10% FBS. The other cell lines were cultured in F-12 medium and RPMI1640 (Invitrogen) mixed medium supplemented with 2% FBS. All cell lines were incubated at 37 °C in a 5% CO2 atmosphere. Cell lines were carefully checked for

morphologic consistency by microscope.

Establishment of Doxycycline-inducible Sox2 overexpressing Het-1A cells (HetSox2)

As Sox2 is a gene composed of one exon without intron, we obtained the coding region of Sox2 by PCR using the genome extracted from Het-1A cells as a template with the primer pair

5’-GGGAATTCATGTACAACATGATGGAGACGGAGCTG-3’ and

5’-GGGGGATCCTCACATGTGTGAGAGGGGCAGTGTGC-3’. After confirming the sequence of the acquired PCR product, Sox2 coding region was cloned into pTetOne-Vector (Clontech), so that Sox2 could be induced by Doxycycline (Dox). We named the constructed vector pTetOneSox2. Het-1A cells were transfected with pTetOneSox2 and linear puromycin marker (Clonech), using X-fect Transfection Reagent (Clontech). Twenty-four hours after the transfection, the cells were selected by puromycin (1µg/ml) (Sigma-Aldrich, St. Louis, MO). The induction of Sox2 expression in the obtained stable clone (HetSox2) treated with Dox was confirmed by Western blot.

Total RNA isolation and quantitative PCR

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Total RNA was isolated with TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. After removing the contaminating DNA by DNase (Invitrogen) from the isolated RNA, cDNA was synthesized using SuperScript III Reverse Transcriptase (Invitrogen). The synthesized cDNA was subjected to quantitative PCR (Q-PCR) using StepOnePlus Real-Time PCR systems (Applied Biosystems, Foster City, CA). Primer sets for Sox2, VEGF, PDGF, Neuropilin1 (Nrp1), Notch1, and

Suprabasin were purchased from Applied Biosystems. Relative expression was calculated by ∆∆CT method using StepOneSoftware Version 2.2 (Applied Biosystems).

Western blot

Western blot was performed according to the standard method described previously (33) with the following primary antibodies: Sox2 (#3579, Cell Signaling Technology, Danvers, MA), anti-Suprabasin (#sc-515037, Santa Cruz Biotechnology, Dallas, TX), anti-AKT (#9272, Cell Signaling Technology), anti-phospho-AKT (#4060, Cell Signaling Technology), anti-p38 (#8690, Cell Signaling Technology), anti-phospho-p38 (#4511, Cell Signaling Technology), anti-p44/42 (#4695, Cell Signaling Technology), anti-phospho-p44/42 (#4370, Cell Signaling Technology), anti-JNK (#9258, Cell Signaling Technology), anti-phopho-JNK (#4668, Cell Signaling Technology), anti-β-actin (#sc-1616R, Santa Cruz Biotechnology). The membranes were then incubated with appropriate secondary antibody for an hour. Protein bands were detected with ECL prime reagent (GE Healthcare, Little Chalfont,

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Buckinghamshire, UK) and were visualized using a lumino-image analyzer (VersaDoc5000; Bio-Rad, Hercules, CA).

Immunohistochemistry

Tissue samples were fixed in 10% buffered formaldehyde and embedded in paraffin. Serial sections cut from paraffin blocks (3µm) were dewaxed. After inhibiting the endogenous peroxidase activity, antigen retrieval was carried out by autoclaving the sections in 10mmol/l citrate buffer. For primary antibodies, we applied anti-Sox2 1:100 (#14962, Cell Signaling Technology), anti-CD34 1:100 (#M7165, Dako Cytomation, Carpintera, CA), and anti-Suprabasin 1:100 (#sc-515037, Santa Cruz Biotechnology), and incubated sections overnight at 4 °C. After the application of appropriate secondary antibodies for 30 minutes at room temperature, bindings were visualized with Liquid DAB+ Substrate Chromogen System (Dako). Sections were counterstained with haematoxylin in the standard fashion.

The expression of Sox2 was evaluated by TissueFAXS (TissueGnostics, Vienna, Austria). Sox2 positive cells were defined as cells whose nuclei were positive for Sox2. Sox2 positive rate in ESCC tissues were compared with that of their adjacent normal tissues. Sox2 positive rate in ESCC tissues was defined as "high" when it was greater than the average, while Sox2 positive rate in ESCC tissues was defined as "low" when it was below the average. Microvessels in ESCC tissues were counted in 5 areas at 200 folds magnification on the basis of CD34 staining. Microvascular density (MVD) was indicated

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

according to the international consensus (34). Immunostaining of Suprabasin in ESCC tissues was evaluated in a double-blinded manner by two independent observers using a semiquantitative reactivity scoring system for both staining intensity (0: negative, 1: weak, 2: moderate, 3: intense) and staining percentage of positively stained cancer cells (0: none, 1: 1-10%, 2: 11-50%, 3: 51-80%, 4: >80%). The final immunohistochemistry (IHC) scores ranged from 0 to 12 as previously reported (35).

Cell proliferation Assay

Cells were seeded on 96-well plates (7.0×103 _{cells/well) and were cultured for 3 days with or}

without Dox. Twenty µl of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carcoxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) was added to the culture and the plate was incubated for another 4 hours at 37℃. Then, the absorbance at 490nm was measured in each well to evaluate cell proliferation. The average absorbance of 8 wells was compared. Suprabasin recombinant protein (OriGene, Rockville, MD) was added to the culture at the final concentration of 5ng/ml in the indicated experiments.

Spheroid Assay

Cells were embedded in Type I collagen gels (Cellmatrix Type I-A; Nitta Gelatin, Osaka, Japan) (3.0×104 _{cells/well) on 24-well plates to form spheroids. Spheroids were cultured in the culture}

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

supernatant of HetSox2 treated with or without Dox for 6 days. The number of spheroids was calculated by ImageJ (National Institutes of Health, Bethesda, MD) after staining with 0.1% crystal violet. Average number of total spheroids in randomly selected three fields per well was calculated.

Migration Assay

HUVEC cell migration was investigated using 24-well plates and 8µm PET membrane (Corning). Culture supernatant of HetSox2 was added to the lower chamber as a chemoattractant. HUVECs were seeded on the upper chamber (5.0×104 _{cells/chamber), and plates were incubated for 4 hours at 37 ℃.}

Cells were then fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Unmigrated cells were removed with cotton bud. Number of migrated HUVECs was calculated using BZ-9000 (KEYENCE, Osaka, Japan). Average number of migrated HUVECs in randomly selected three fields per insert was calculated.

Sprouting Assay

HUVECs were seeded on a 96-well Ultra-low attachment plate (Corning) (2.0×103 _{cells/well) and}

incubated for 16 hours at 37 ℃ to form macrosphere. HUVEC macrosphere was embedded in Type I collagen gels (Nitta Gelatin), applied to 24-well plate, and incubated for 24 hours at 37 ℃ in the culture supernatant of Dox treated or untreated HetSox2 to form sprouts. HUVEC macrospheres were fixed

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

with 4% paraformaldehyde and sprout length was measured by ImageJ (National Institutes of Health). Average cumulative length of sprouts in three macrospheres was calculated.

Immunofluorescence

Dox treated or untreated HetSox2 cells were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton-X in PBS. Cells were incubated overnight at 4˚C with anti-Suprabasin antibody (#sc-515037, Santa Cruz Biotechnology). After being incubating with Alexa Fluor 555-labeled anti-mouse IgG secondary antibody (#A28180, Invitrogen), the cells were then incubated with Alexa Fluor 488-labeled anti-Sox2 antibody (#53-9811-82, eBioscience, San Diego, CA) for 3 hours at room temperature. Nuclei were counterstained with 4',6'-diamidino-2-phenylindole (DAPI)(Invitrogen). Images were captured using a laser scanning confocal microscope (Nikon C2si; Tokyo, Japan).

Luciferase Assay

Transcriptional factor-binding site in the 5'-promoter region of human Suprabasin was analyzed using JAPAR (http://jaspar.genereg.net). Then, the 5'-promoter region of Suprabasin was obtained by PCR and ligated into Luciferase Reporter Vector-pGL3 (Promega, Madison, WI) using DNA Ligation Kit (Takara Bio, Shiga, Japan). The constructed luciferase vector, pTetOneSox2 and pRL-TK vector (Promega) were transfected into HEK293 cells using TransIT-X2 transfection reagent (Mirus Bio,

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Madison, WI). The transfected cells were cultured for 48 hours in the presence or the absence of Dox, and then were lysed with Passive Lysis Buffer (Promega). The luciferase activity was measured using Dual-Luciferase Reporter Assay System (Promega).

Chromatin immunoprecipitation Assay

Chromatin immunoprecipitation (ChIP) was performed for Dox-treated and untreated HetSox2 with anti-Sox2 antibody (#2748, Cell Signaling Technology) using ChIP-IT Express (Active motif, Carlsbad, CA) following manufacturer's protocol. Briefly, HetSox2 cells were cultured with or without Dox for 3 days. Cells were then fixed with 1% formaldehyde (Sigma-Aldrich) and chromatin was extracted from their nuclei. Immunoprecipitation using either anti-Sox2 antibody or normal rabbit IgG (#2729, Cell Signaling Technology) was performed and the immunoprecipitated chromatin was subjected to PCR using PrimeStar GXL (Takara Bio) with Primers 5’-TTTGGGTGGAGTCAAGTGTCCAGG-3’ and 5’-TTGGTGCCAAGGGGAGAGACTGAG-3’. The obtained PCR products were electrophoresed on a 0.9% agarose gel containing Tris-acetate EDTA buffer and ethidium bromide.

Statistical Analysis

All experiments were performed three to five times. Student’s t-test was performed to evaluate the

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

significance of the differences between two groups using Microsoft Excel 2016. A value of p<0.05 was considered as statistically significant. Results are indicated as means ± standard deviation.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Results

Immunohistochemistry of Sox2 in early stage ESCC

We investigated the expression of Sox2 by immunohistochemistry (IHC) in early stage ESCC obtained by endoscopic resection. Forty-seven ESCC samples consisted of 37 mucosal cancers and 10 submucosal cancers. In the adjacent normal squamous epithelial tissue, Sox2 immunoreactivity was observed in the nuclei of cells in the basal layer, but the immunoactivity decreased toward the princkle layer and granular layer (Figure 1A). In the mucosal cancer tissues, Sox2 immunoreactivity showed stronger staining than the adjacent normal squamous tissue, especially in the basal layer. In the submucosal cancer tissues, Sox2 was diffusely stained almost throughout the layer, mainly in the infiltrating lesions (Figure 1A). Quantitative image analysis using TissueFAXS showed that Sox2-positive rate in cancer tissues (64.7%) was significantly higher than that in adjacent normal tissues (53.0%) (Figure 1B).

Enhanced cell proliferation and tumorigenicity by Sox2 induction

To investigate the role of Sox2 in esophageal carcinogenesis, we established Dox-inducible Sox2 overexpressing Het-1A cells (HetSox2, see Materials and methods). After confirming that Sox2 was overexpressed in Dox-treated HetSox2 (Figure 2A), we performed MTS assay to evaluate whether cell proliferation was enhanced by Sox2 overexpression in these cells. Dox-treated HetSox2 showed

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

increased proliferation compared to Dox-untreated HetSox2 (Figure 2B). Next, we performed spheroid assay to evaluate the effect of Sox2 on tumorigenicity. Total spheroid area of Dox-treated HetSox2 was significantly larger than that of Dox-untreated HetSox2 (Figure 2C). These results suggested that overexpression of Sox2 in normal esophageal squamous cells induced cell proliferation and tumorigenicity.

HUVEC migration and sprouting by the induction of Sox2

Since early stage ESCC is known to be accompanied by transformation of microvessels (7), we next investigated the influence of Sox2 overexpression on angiogenesis. To this end, we employed HUVEC and performed migration assay and sprouting assay. In migration assay, the culture supernatant of treated HetSox2 induced HUVEC migration to a greater degree compared to that of Dox-untreated HetSox2 (Figure 2D). In sprouting assay, the culture supernatant of Dox-treated HetSox2 formed significantly more sprouts compared to that of Dox-untreated HetSox2 (Figure 2E). These results suggested that some kind of growth factors related to proliferation, angiogenesis, and differentiation were secreted from Sox2 overexpressing HetSox2, and that they were involved in the migration of endothelial cells and the formation of new vascular structures.

Identification of Suprabasin as a growth factor secreted from Sox2 overexpressing HetSox2

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

To identify the growth factors related to angiogenesis and differentiation induced by Sox2 overexpressing HetSox2, we extracted the RNA of Dox-treated and untreated HetSox2 and performed Q-PCR for the candidate genes. Unexpectedly, VEGF, the major angiogenic factor previously reported to be expressed in advanced ESCC (30,31), was not upregulated in Dox-treated HetSox2 (Figure 3A). In addition, PDGF and Nrp1, which are also factors related to angiogenesis, and Notch1, which is a factor related to differentiation, were not induced by Sox2 (Figure 3A). On the other hand, we found that Suprabasin expression was induced to 2.58 ± 0.27folds by Sox2 induction (Figure 3A). This induction of Suprabasin by Sox2 was confirmed by immunofluorescence study, which showed that Dox-treated HetSox2 cells expressed nuclear Sox2 and cytoplasmic Suprabasin, while these expressions were limited in Dox-untreated HetSox2 cells (Figure 3B). Similarily, Suprabasin was expressed in Sox2-expressing ESCC cell lines, whereas its expression was scarce in Sox2-deficient Het-1A cells (Figure 3C). These findings suggested that Sox2 induced Suprabasin. Since Suprabasin is a secreted protein specifically expressed in stratified squamous epithelium and is known to be involved in the differentiation of keratinocytes (36,37), we decided to focus on this molecule .

We then investigated the molecular mechanism involved in Suprabasin induction by Sox2. Examination of the promoter region of Suprabasin by JASPAR predicted a probable binding site of Sox2 in the position -1566/-1559 (Figure 3D). Hence, we hypothesized that Sox2 directly regulated Suprabasin expression, and constructed a luciferase reporter vector containing this binding site to verify

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

this hypothesis (Figure 3D, see Materials and methods). We found that overexpression of Sox2 led to increased luciferase activity (6.36 ± 0.92folds, Figure 3E). In addition, ChIP assay revealed that Sox2 bound to this binding site located in the promoter region of Suprabasin (Figure 3F). These results suggested that Sox2 bound to the position -1566/-1559 of Suprabasin promoter region, and that it directly induced Suprabasin expression.

The effect of Suprabasin on cell proliferation and angiogenesis

We then investigated whether Suprabasin had an effect on cell proliferation by MTS assay. We found that Het-1A cells cultured in the presence of Suprabasin (5ng/ml) showed increased proliferation compared to the cells cultured in the absence of Suprabasin (Figure 4A). This result suggested that Suprabasin induced cell proliferation of esophageal squamous cells.

Since the supernatant of Sox2-overexpressing HetSox2 induced the migration and sprouting of HUVEC (Figure 2D and E), we evaluated the angiogenic effect of Suprabasin by HUVEC migration assay and sprouting assay. In migration assay, Suprabasin induced migration of HUVEC to a greater degree (Figure 4B). Furthermore, in sprouting assay, Suprabasin induced significantly longer sprout elongation (Figure 4C). These results suggested that Suprabasin had an angiogenic effect.

AKT signaling is involved in the angiogenic effect of Suprabasin

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Since a previous report showed that Suprabasin activated AKT signaling (38), and since AKT signaling and MAPK signaling are the major pathways downstream of VEGF (28,29), we examined whether these signaling pathways were involved in angiogenesis. First of all, we investigated whether Suprabasin activated AKT and MAPK signaling by Western blot and found that AKT phosphorylation and p38 phosphorylation were enhanced by the addition of Suprabasin (Figure 5A, upper and middle panel). On the other hand, phosphorylation of p42/44 did not differ by the addition of Suprabasin (Figure 5A, lower panel).

LY294002 is a known inhibitor of AKT signaling (39). After confirming that AKT phosphorylation could be inhibited by LY294002 (10μM), we investigated whether AKT signaling is involved in angiogenesis. We found that LY294002 (10μM) significantly inhibited the migration of HUVEC (Figure 5B). On the other hand, p38 MAPK inhibitor SB202190 (10μM) had no influence on HUVEC migration (Figure 5C). These results suggested that Suprabasin acted as an angiogenic factor via AKT signaling.

Sox2, Suprabasin, and CD34 expression in early stage ESCC.

We then evaluated the expression of Sox2, Suprabasin, and endothelial cell marker CD34 in human early stage ESCC tissues by immunohistochemistry. We observed that Sox2 was diffusely stained in the nuclei of ESCC cells and that Suprabasin was highly expressed in the cytoplasm of these cells compared

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

to the adjacent normal squamous cells (Figure 6A). In addition, the number of CD34-positive microvessels was greater in the Sox2 and Suprabasin expressing ESCC tissues (Figure 6A). Next, we evaluated the expression of Suprabasin and the density of CD34-positive microvessels by a semiquantitative immunohistochemical scoring to examine their correlation with Sox2 expression in ESCC cases. We found that the expression of Suprabasin (Figure 6B) and the density of CD34-positive microvessels (Figure 6C) were significantly higher in early stage ESCC cases with high Sox2 expression. These results suggested that Suprabasin was induced in Sox2-overexpressing ESCC cells, and as a result, caused angiogenesis in this region.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Discussion

The morphological changes of microvessels have become one of the major characteristics important for detecting the existence and speculating the depth of invasion of early stage ESCC, but the molecular mechanism underlying these changes have not been determined. In this study, we attempted to elucidate the molecular mechanism involved in the carcinogenesis and angiogenesis of early stage ESCC from the perspective of Sox2, the transcription factor required for the development and maintenance of the esophagus. We found that the expression of Sox2 was enhanced even in early stage ESCC, as was reported with advanced ESCC (9). This suggested that Sox2 could be playing a pivotal role in the carcinogenesis of ESCC. By using Sox2-overexpressing esophageal squamous cell line HetSox2 and endothelial cell line HUVEC, we showed that Sox2 overexpression led to enhanced cell proliferation and tumorigenicity as well as accelerated angiogenesis. Furthermore, we identified Suprabasin as a direct target of Sox2 and found that Suprabasin induced cell proliferation and angiogenesis. Finally, we showed that Sox2-overexpressing ESCC tissues also expressed Suprabasin and that the density of microvessels was high in ESCC tissues with high Sox2 expression.

A recent comprehensive molecular analysis of advanced ESCC reported that Sox2 copy number is increased in around half of the advanced ESCC (9). Another report showed that Sox2 was genetically amplified in ESCC and that cases with Sox2 amplification overexpressed Sox2 (16). In cutaneous squamous cell carcinomas, aberrant Sox2 expression was reported to directly regulate the key genes

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

involved in cancer-cell proliferation and stemness (18). Similarly, Sox2 overexpression was reported to induce cancer cell proliferation and anti-apoptotic protein expression in head and neck cancers, which is another squamous cell carcinoma (40). These reports suggested that Sox2 plays an important role in cancer initiation in various squamous cell carcinoma. In the present study, we also showed that early stage ESCC tissues exhibited Sox2 expression at a higher degree compared to their adjacent normal tissues. In addition, Sox2 overexpression in normal human esophageal squamous cell line significantly enhanced proliferation and spheroid formation. These results suggested that abnormally enhanced Sox2 expression could lead to carcinogenesis, though it is an essential transcription factor for the maintenance of the esophagus (14).

Early stage ESCC is usually accompanied by angiogenesis and morphological changes of microvessels, but the molecular mechanism underlying this phenomenon has not been determined. In the present study, the culture supernatant of Sox2-overexpressing HetSox2 induced migration and sprouting of endothelial cell line HUVEC. This suggested the enhancing effect of Sox2 overexpression on angiogenesis. There has been very limited reports about the influence of Sox2 on angiogenesis. Among these few reports, Kautzman et al reported that Sox2 deletion in retinal astrocytes led to delayed angiogenesis (41). On the other hand, Siegle et al reported that overexpression of Sox2 induced the expression of angiogenic factors such as VEGF, Nrp1, and Epiregulin in cutaneous squamous cell carcinoma (42). These two reports are in accordance with our result in that Sox2 induced angiogenesis.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Previously, angiogenic factor VEGF was reported to be overexpressed in advanced ESCC (30,31), but we were not able to detect its enhanced expression in our Sox2-overexpressing HetSox2. Instead, we identified Suprabasin as a factor induced by Sox2-overexpression.

Suprabasin was identified by Park et al as a gene involved in the differentiation of cutaneous stem

cells (36). It is a secreted protein which is expressed in stratified epithelium of skin and esophagus (37). To date, there are only two previous reports regarding the regulation of Suprabasin: one showed that it was induced by 12-O-tetradecanoylphorbol-13-acetate treatment (36) and the other showed its induction by interferon-γ (43). In the present study, we firstly showed that Sox2 induced the activation of

Suprabasin promoter, which contained a Sox2 binding site, by luciferase assay. Moreover, we showed

that Sox2 bound to its binding site located in the position -1566/-1559 in Suprabasin promoter by ChIP assay. Our findings suggested that Suprabasin was directly induced by Sox2 in the esophagus.

Aberrant Suprabasin expression has been reported in non-small cell lung cancer (44), salivary adenoid cystic carcinoma (45), and advanced ESCC (46). In the present study, Suprabasin induced cell proliferation of Het-1A. Moreover, it induced cell migration and sprouting of HUVEC, indicating that Suprabasin has an angiogenic effect. According to a previous report, VEGF is not expressed in the early phase of ESCC (7), which lets us suspect that Suprabasin but not VEGF is inducing angiogenesis in the early stage of ESCC.

Regarding the signal transduction pathway downstream of Suprabasin, Alam et al reported that

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

knocking down Suprabasin in mouse tumor endothelial cells led to reduced AKT activation (38). In the present study, Suprabasin induced phosphorylation of AKT and p38, the typical signal transducers downstream of VEGF signaling (28,29). Since we found that inhibiting AKT but not p38 led to reduced migration of HUVEC, we assume that AKT signaling is playing a pivotal role in Suprabasin induced angiogenesis, but further studies are warranted.

Finally, immunohistochemical analysis revealed that Suprabasin expression and microvessel density were higher in early stage ESCC cases with high Sox2 expression, which supported our findings obtained from in vitro studies.

In conclusion, we found that Sox2 was overexpressed in early stage ESCC and that Sox2 directly induced Suprabasin expression, which led to enhanced cell proliferation in esophageal squamous cells, as well as cell migration and sprout formation in endothelial cells. This Sox2-Suprabasin pathway could be one of the pathways involved in the carcinogenesis and the angiogenesis of early stage ESCC.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Acknowledgement

The authors would like to thank Biomedical Research Unit of Tohoku University Hospital for their technical support. The authors would also like to thank Ms. Kiiko Ogashiwa for her help in histology.

The authors declare no conflict of interest.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

References

1. Pennathur, A. et al. (2013) Oesophageal carcinoma. Lancet, 381, 400-412.

2. Pickens, A et al. (2003) Geographical distribution and racial disparity in esophageal cancer. Ann

Thorac Surg, 76, S1367-S1369.

3. Rustgi, A.K. et al. (2014) Esophageal carcinoma. N Engl J Med, 371, 2499-2509.

4. Japan Esophageal Society. (2017) Japanese Classification of Esophageal Cancer, 11th Edition: part I. Esophagus, 14, 1-36.

5. Muto, M. et al. (2010) Early detection of superficial squamous cell carcinoma in the head and neck region and esophagus by narrow band imaging: a multicenter randomized controlled trial.

J Clin Oncol, 28, 1566-1572.

6. Sugimachi, K. et al. (1992) Endoscopic diagnosis of early carcinoma of the esophagus using Lugol's solution. Gastrointest Endosc, 38, 657-661.

7. Kumagai, Y. et al. (2002) Dynamism of tumour vasculature in the early phase of cancer progression: outcomes from oesophageal cancer research. Lancet Oncol, 3, 604-610.

8. Oyama, T. et al. (2017) Prediction of the invasion depth of superficial squamous cell carcinoma based on microvessel morphology: magnifying endoscopic classification of the Japan Esophageal Society. Esophagus, 14, 105-112.

9. Cancer Genome Atlas Research Network. (2017) Integrated genomic characterization of

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

oesophageal carcinoma. Nature, 541, 169-175.

10. Sawada, G. et al. (2016) Genomic Landscape of Esophageal Squamous Cell Carcinoma in a Japanese Population. Gastroenterology, 150, 1171-1182.

11. Liu, X. et al. (2017) Genetic Alterations in Esophageal Tissues From Squamous Dysplasia to Carcinoma. Gastroenterology, 153, 166-177.

12. Takahashi, K. et al. (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell, 126, 663-676.

13. Sarkar, A. et al. (2013) The sox family of transcription factors: versatile regulators of stem and progenitor cell fate. Cell Stem Cell, 12, 15-30.

14. Que, J. et al. (2007) Multiple dose-dependent roles for Sox2 in the patterning and differentiation of anterior foregut endoderm. Development, 134, 2521-2531.

15. Hutz, K. et al. (2014) The stem cell factor SOX2 regulates the tumorigenic potential in human gastric cancer cells. Carcinogenesis, 35, 942-950.

16. Bass, A.J. et al. (2009) SOX2 is an amplified lineage-survival oncogene in lung and esophageal squamous cell carcinomas. Nat Genet, 41, 1238-1242.

17. Gen, Y. et al. (2010) SOX2 identified as a target gene for the amplification at 3q26 that is frequently detected in esophageal squamous cell carcinoma. Cancer Genet Cytogenet, 202, 82-93. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

18. Boumahdi, S. et al. (2014) SOX2 controls tumour initiation and cancer stem-cell functions in squamous-cell carcinoma. Nature, 511, 246-250.

19. Leis, O. et al. (2012) Sox2 expression in breast tumours and activation in breast cancer stem cells. Oncogene, 31, 1354-1365.

20. Gen, Y. et al. (2013) SOX2 promotes tumor growth of esophageal squamous cell carcinoma through the AKT/mammalian target of rapamycin complex 1 signaling pathway. Cancer Sci,

104, 810-816.

21. Wang, Q. et al. (2009) Oct3/4 and Sox2 are significantly associated with an unfavorable clinical outcome in human esophageal squamous cell carcinoma. Anticancer Res, 29, 1233-1241. 22. Yokota, E. et al. (2017) Targeted silencing of SOX2 by an artificial transcription factor showed

antitumor effect in lung and esophageal squamous cell carcinoma. Oncotarget, 8, 103063-103076.

23. Kitadai, Y. et al. (2004) Angiogenic switch occurs during the precancerous stage of human esophageal squamous cell carcinoma. Oncol Rep, 11, 315-319.

24. Lee, S. et al. (2007) Autocrine VEGF signaling is required for vascular homeostasis. Cell, 130, 691-703.

25. Verbeke, H. et al. (2010) Expression of angiostatic platelet factor-4var/CXCL4L1 counterbalances angiogenic impulses of vascular endothelial growth factor,

interleukin-4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

8/CXCL8, and stromal cell-derived factor 1/CXCL12 in esophageal and colorectal cancer. Hum

Pathol, 41, 990-1001.

26. Carmeliet, P. et al. (2011) Molecular mechanisms and clinical applications of angiogenesis.

Nature, 473, 298-307.

27. Goel, H.L. et al. (2013) VEGF targets the tumour cell. Nat Rev Cancer, 13, 871-882.

28. Wagner, E.F. et al. (2009) Signal integration by JNK and p38 MAPK pathways in cancer development. Nat Rev Cancer, 9, 537-549.

29. Manning, B.D. et al. (2007) AKT/PKB signaling: navigating downstream. Cell, 129, 1261-1274. 30. Shih, C.H. et al. (2000) Vascular endothelial growth factor expression predicts outcome and lymph node metastasis in squamous cell carcinoma of the esophagus. Clin Cancer Res, 6, 1161-1168.

31. Lin, E.W. et al. (2016) The tumor microenvironment in esophageal cancer. Oncogene, 35, 5337-5349.

32. Shimada, Y. et al. (1992) Characterization of 21 newly established esophageal cancer cell lines.

Cancer, 69, 277-284.

33. Asonuma, S. et al. (2009) Helicobacter pylori induces gastric mucosal intestinal metaplasia through the inhibition of interleukin-4-mediated HMG box protein Sox2 expression. Am J

Physiol Gastrointest Liver Physiol, 297, G312-G322.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

34. Koide, N. et al. (2001) Coexpression of vascular endothelial growth factor and p53 protein in squamous cell carcinoma of the esophagus. Am J Gastroenterol, 96, 1733-1740.

35. Nan, P. et al. (2019) MTA1 promotes tumorigenesis and development of esophageal squamous cell carcinoma via activating the MEK/ERK/p90RSK signaling pathway. Carcinogenesis. 36. Park, G.T. et al. (2002) Suprabasin, a novel epidermal differentiation marker and potential

cornified envelope precursor. J Biol Chem, 277, 45195-45202.

37. Matsui, T. et al. (2004) Identification of novel keratinocyte-secreted peptides dermokine-alpha/-beta and a new stratified epithelium-secreted protein gene complex on human chromosome 19q13.1. Genomics, 84, 384-397.

38. Alam, M.T. et al. (2014) Suprabasin as a novel tumor endothelial cell marker. Cancer Sci, 105, 1533-1540.

39. Zhou, Y. et al. (2015) LY294002 inhibits the malignant phenotype of osteosarcoma cells by modulating the phosphatidylinositol 3kinase/Akt/fatty acid synthase signaling pathway in vitro.

Mol Med Rep, 11, 1352-1357.

40. Schrock, A. et al. (2014) Expression and role of the embryonic protein SOX2 in head and neck squamous cell carcinoma. Carcinogenesis, 35, 1636-1642.

41. Kautzman, A.G. et al. (2018) Sox2 regulates astrocytic and vascular development in the retina.

Glia, 66, 623-636. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

42. Siegle, J.M. et al. (2014) SOX2 is a cancer-specific regulator of tumour initiating potential in cutaneous squamous cell carcinoma. Nat Commun, 5, 4511.

43. Hubackova, S. et al. (2019) Interferon-regulated suprabasin is essential for stress-induced stem-like cell conversion and therapy resistance of human malignancies. Mol Oncol, 13, 1467-1489. 44. Glazer, C.A. et al. (2009) Integrative discovery of epigenetically derepressed cancer testis

antigens in NSCLC. PLoS One, 4, e8189.

45. Shao, C. et al. (2012) Suprabasin is hypomethylated and associated with metastasis in salivary adenoid cystic carcinoma. PLoS One, 7, e48582.

46. Zhu, J. et al. (2016) Overexpression of suprabasin is associated with proliferation and tumorigenicity of esophageal squamous cell carcinoma. Sci Rep, 6, 21549.

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

Legends to Figures

Figure 1. Sox2 was expressed in early stage ESCC tissues. (A) Hematoxylin-eosin staining and

immunohistochemistry for Sox2 of mucosal and submucosal ESCC tissues, and their adjacent normal tissues are shown. bar: 100μm. (B) Quantification of Sox2 positive rates of cancer tissues and adjacent normal tissues are shown. n=47. *p<0.05

Figure 2. Sox2 promoted cell proliferation, spheroid formation, migration, and sprouting.

(A) Dox-induced Sox2 expression of the established HetSox2 was confirmed by Western blot. (B and C) MTS assay (B) and spheroid assay (C) were performed with Dox-treated and untreated HetSox2. n=5. (C: x40) (D and E) Migration assay (D) and sprouting assay (E) of HUVEC cells were performed with the culture supernatant of Dox-treated and untreated HetSox2. n=5. (D: x10, E:x100) *p<0.05

Figure 3. Sox2 directly induced Suprabasin. (A) The expressions of genes related to angiogenesis and

differentiation in Dox-treated and untreated HetSox2 were evaluated by Q-PCR. n=5. (B) The expressions of Sox2 and Suprabasin were assessed by immunofluorescence. Sox2: green, Suprabasin: red, DAPI: blue. (C) Expressions of Sox2, Suprabasin, and β-actin of Het-1A and ESCC cell lines were evaluated by Western blot. (D) The scheme showing the putative Sox2 binding site in the position -1566/-1559 of Suprabasin promoter and the luciferase construct containing this region. TSS;

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

transcription starting site. (E) Suprabasin promoter activity in Dox-treated and untreated HetSox2 was assessed by luciferase assay. n=5. (F) ChIP assay was performed for the Sox2 binding site in Suprabasin promoter using anti-Sox2 antibody. Rabbit IgG was used as an isotype control. *p<0.05

Figure 4. Suprabasin promoted cell proliferation and angiogenesis. (A) Suprabasin was added to

the culture of Het-1A and MTS assay was performed. n=3. (B) HUVEC migration assay was performed with Suprabasin added to the culture medium of the lower chamber. n=3. (x10) (C) Suprabasin was added to the culture of HUVEC and sprout formation was evaluated. n=3. (x100) *p<0.05

Figure 5. AKT signaling was involved in Suprabasin induced angiogenesis. (A) Suprabasin was

added to the culture of HUVEC. Cell lysates were collected at the indicated time points and were subjected to Western blot analyses to evaluate AKT, p38, and ERK activation. (B) HUVEC migration assay was performed with LY294002 added to the lower chamber together with Suprabasin. n=3. (x10) (C) HUVEC migration assay was performed with SB202190 added to the lower chamber together with Suprabasin. n=3. (x100) N.S.; not statistically significant. *p<0.05

Figure 6. Sox2, Suprabasin, and CD34 were expressed in the identical region of early stage ESCC.

(A) Representative immunohistochemical images for Sox2, Suprabasin, and CD34 in early stage ESCC

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

tissues and their adjacent normal tissues. bar: 100μm. (B and C) Comparison of Sox2 expression and Suprabasin expression (B) or CD34-positive microvascular density (C) in ESCC tissues. n=47. *p<0.05

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

For Peer Review

For Peer Review

For Peer Review

For Peer Review

For Peer Review

For Peer Review

Supplementary Table 1. Clinicopathological characteristics of ESCC patient samples

Characteristics Number of cases (%) Gender Male 33 (70.2) Female 14 (29.8) Age ≧65 _{33 (70.2)} ＜65 14 (29.8) T classification T1 47 (100.0) T2 0 (0.0) T3 0 (0.0) T4 0 (0.0) N classification N0 47 (100.0) N1 0 (0.0) N2 0 (0.0) N3 0 (0.0) M classification M0 47 (100.0) M1 0 (0.0) 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58