The Detection of sAMY1α in Ovarian Cancer Ascites During First-Line Chemotherapy Predict the Response of Patients to Treatment


 Background: Ovarian cancer ascites are a clinical conundrum in the process of diagnosis and treatment, and chemotherapy is the main treatment. The efficacy and prognostic assessment of chemotherapy has been the subject of continuous exploration. Methods: Affymetrix HTA2.0 microarray analysis of ovarian cancer ascites without and with chemotherapy, as well as metabolomics techniques, showed that AMY1A gene expression in the chemotherapy was significantly higher than that without chemotherapy.Results: The analysis of ovarian cancer tissues, ascitic precipitated cells and supernatants, showed that the high expression of sAMY-1α was negatively correlated with HE4 (p<0.01), and was also negatively correlated with the expression of the Beclin1 and LC3 (p<0.05, p<0.001) in group with chemotherapy. The results of ovarian cancer cell lines OVCAR3 and A2780, which were treated with autophagic inhibitor (3MA), inducer (rapamycin) or cisplatin (CDDP), were consistent with it. The results of 3MA group showed that as the drug concentration increased, there were decreased in the levels of Beclin1 and LC3II, and sAMY1α levels increased. There was opposite result in rapamycin group. CDDP group revealed significantly decreased levels of LC3II, Beclin1, and increased levels of sAMY1α; when rapamycin was added, autophagy inhibition was alleviated, and slightly reduced levels of sAMY1α. The results indicated that chemotherapy induced autophagic death of tumor cells, and the expression of sAMY1α in ascites of chemotherapy reflected the degree of autophagic death of tumor cells. Conclusions: During the diagnosis and treatment of ovarian cancer ascites, the expression level of sAMY1α was detected at any time, compared with the levels of HE4, Beclin1 and LC3 in the ascites with chemotherapy, and compared with the serum level of CA125 in the patients, so that the recurrence of ascites could be predicted. Therefore, the adjustment of treatment regimen has important guiding significance.


Background
Ovarian cancer ascites is a common complication of advanced ovarian cancer with poor prognosis 1 , and the formation of which is due to the peritoneal dissemination of cancer cells 2 . Targeted killing of cancer cells in ascites, reducing or inhibiting the recurrence of ascites, can improve the quality of life and survival time of patients 3,4 . The preferred clinical treatment for ovarian cancer with ascites is surgery combined with chemotherapy. However, not all patients can suppress the recurrence of ascites after chemotherapy. The reaccumulation or recurrent of ascites is the outcome which the failure of chemotherapy or drug resistance 5 .
So, it is crucial to nd biomarkers in ascites that can predict the prognosis of ovarian cancer ascites.
Genechip technology is a new molecular biology technology emerging in the early 1990s with the advantages of high throughput, high parallelism, high sensitivity and strong speci city 6 . Metabolomics is a fast and effective way to identify new cancer biomarkers. It integrates high-throughput and high-resolution analytical technology with bioinformatics, studies the level of biological metabolism, provides a unique perspective on understanding organisms, and has increasingly mature applications in cancer research 7,8 . Our team performed a genechip test on ovarian cancer ascites samples, and found that AMY1A-C (salivary alpha-amylase, sAMY) and AMY2A(pancreatic alpha-amylase) genes were highly expressed in the ascites. Then metabolomics tests were performed on samples of ovarian cancer tissues, and the obtained differential metabolite, maltotriose. Maltotriose was catalyzed by α-amylase encoding the gene AMY. The TCGA database data show that all patients with serous ovarian cancer III-IV can express AMY1A-C and AMY2A, 2B genes.
sAMY is the main component of salivary protein. It is normally expressed in salivary gland, thyroid, lung, pancreas and testis tissues, while expressed in patients with ovarian, thyroid and lung adenocarcinomas 9 .
Tsunashima detected sAMY in ovarian tumor cells in 1976 10 . In 1978, Corlette MB et al. also detected sAMY in ovarian cancer tissue and ascites 11 . What is the relationship between the high expression of AMY gene in ovarian cancer tissue and ascites and the prognosis of ovarian cancer patients? In particular, can it be a marker for evaluating the e cacy of chemotherapy in patients with ovarian cancer ascites? The results of this study are for reference by peer experts. Through the previous cytotoxicity assay 41 , we found that the IC50 of A2780 and OVCAR3 respectively was 30 μM and 45 μM to cispaltin. A2780 as relative sensitivity, and OVCAR3 as relative resistance.

Materials And Methods
Processing of ovarian cancer ascites The cases of cancerous ascites were centrifuged for 10 min at 3500 rpm The supernatants of cancerous ascites were stored at -80 °C for subsequent analysis by ELISA. Then, moderate ACK lysis (Leagene Biotechnology, China) buffer was added to the precipitates, which were collected into 1.5-mL centrifuge tubes. Subsequently, RNA, electron microscopy, light microscopy, and protein analyzes were performed.

Genechip
No chemotherapy (n=3) and chemotherapy (n=1) ovarian cancer ascites were selected. 1 mL TRIzol was added to the ascitic precipitate, and mixed evenly. The samples were sent to Beijing Kangpson Biological Company for expression pro le analysis of Affymetrix HTA2.0 Genechip (Speci c steps in supplementary materials).

Metabolomics techniques
Metabolite extraction from ovarian cancer tissues was performed as described by references 42,43 , and the abnormal metabolites were selected and mapped to KEGG database (http://www.genome.jp/kegg/). Speci c steps in supplementary materials. AMY genes screened by genechip and metabolomics were compared with epithelial serous ovarian cancer data in TCGA (The Cancer Genome Atlas) database (https://cancergenome.nih.gov/).

Pathological techniques
Ovarian cancer tissues or precipitated ascites cells were immobilized with 10% formaldehyde, alcohol dehydration, routinely processed into para n blocks and then sectioned. The sections were then stained with H&E and observed under light microscope.
Precipitated ascites cells were xed in 2.5% glutaraldehyde in PBS (pH 7.4), rinsed with PBS, and then xed in 1% osmium tetroxide (pH 7.4). Samples were dehydrated in a graded series of acetone, and then embedded in Epon 812. Samples were cut into ultrathin sections, double-stained with uranyl acetate and lead citrate and examined with an electron microscope (H-7650).

Flow cytometry
After cell counts were performed, 1×10 6 cells were added to a 1.5mL tube, rinsed twice with PBS and centrifuged for 5 min at 1500 rpm. Cells were then rinsed once with staining buffer, and 1 µL of CA125 (Bioss, bs-0091R) antibody was added (staining buffer as negative control), resulting in a 100 µL total volume. After 30 min of incubation in the dark and xation in 500 µL of 1% paraformaldehyde, the detection assay was performed. This assay was repeated thrice.

Western blot
The total protein from cell lines and precipitated ascites cells was extracted, and the protein concentration was determined by a BCA protein concentration detection kit. Equal amounts of cellular proteins were loaded into each well and resolved using 15% or 12% SDS-PAGE gels. Polyvinylidene uoride membrane blotting was subsequently performed under standard conditions. The following primary antibodies were used against sAMY1α (1:1000), HE4 (1:100), CA125 (1:500), Beclin1 (1:1000), and LC3 (1:500) for overnight at 4°C. The membrane was added a 1:5000 dilution of the secondary antibody. Then, the blot was developed with an ECL kit. GAPDH (1:5000) used as an internal reference.

Statistical analysis
Statistical analysis SPSS 17.0 software was used for the statistical analysis. One-way analysis of variance (ANOVA) was used to analyze the differences between groups. An independent t-test or u-test for differences in the mean value was used for comparison. Spearman's test was used to evaluate the associations. A p<0.05 was considered signi cant, and values of p<0.01 or p<0.001 were considered statistically signi cant. GraphPad Prism was used for mapping and curve tting.

Results
The content and characteristics of tumor cells in EOC ascitic precipitated cells In without chemotherapy group, the tumor cells were clustered under light microscope; electron microscope results showed that the ratio of tumor nucleus to plasma increased and more autophagosome or autophagic vacuoles were seen in the cytoplasm; meanwhile, ow cytometry (CA125 label tumor cells) results indicated that the tumor cells content accounted for 52.75% ± 11.8% (Fig. 1A). In the chemotherapy group, tumor cells became loose after chemotherapy; most of tumor cells appeared apoptosis in electron microscope; and the tumor cell content was reduced to 19.45%±5.96% (Fig. 1B). In the cases of re-ascites after chemotherapy, the number of tumor cells increased signi cantly and arranged with a visible glandular structure. Tumor cells proliferate actively, cell membrane protuberance is rich, and cytoplasmic cytoskeleton is increased. Tumor cell expressed reaching 83.9%±13.4% (Fig. 1C, showed in ref 12 ).
Affymetrix HTA2.0 Genechip, Metabolic Pro ling and TCGA At the beginning of the experiment, we collected 4 samples for genechip. By screening the differential genes among the 4 RNA quality control samples, it was preliminarily found that the largest differential gene was AMY ( Fig. 2A). In order to further con rm the role of AMY, we conducted metabonomics and TCGA analysis. Metabonomic data analysis of ovarian cancer tissues (n=22) found that the expression of maltotriose was up-regulated but not statistically signi cant in group with chemotherapy. Furthermore, the comparison between the chemosensitive (sensitive) group and the chemoresistant (resistance) group showed that maltotriose, the differential metabolite, was increased in the drug resistance group with statistical signi cance (Fig. 2B). Maltotriose was mapped to KEGG database and found to be involved in carbohydrate digestion and absorption, catalyzed by the amylase which encoding the gene AMY (map04973). RNAseq data were collected from 379 cases of serous ovarian cancer in TCGA database, in which AMY gene subtypes were expressed as AMY1A(173/379), AMY1B(356/379), AMY1C(60/379), AMY2A(379/379), and AMY2B(277/379) (Fig. 2C).
Expression of sAMY-1α, HE4, CA125, Beclin1, and LC3 in ovarian cancer tissues Pathologic features of serous adenocarcinoma (Fig.4A). The expression of sAMY1α in group without chemotherapy was lower than that with chemotherapy (Fig.4B). HE4 and CA125 were highly expressed in cancer tissues of group without chemotherapy, and expressed lower in chemotherapy group (Fig.4B). There was no signi cant difference of Beclin1 in ovarian cancer tissue that without and with chemotherapy (Fig.4B). The expression of LC3 was signi cantly higher in ovarian cancer tissues that without chemotherapy, and was signi cantly lower in chemotherapy group (Fig.4B). Image J quanti es the histological samples which show in Fig.4C. In conclusion, the expression of sAMY1α in both group without and with chemotherapy showed a contrary trend to that of HE4, CA125, Beclin1 and LC3.
Correlation between the expression of sAMY1α, HE4, CA125, Beclin1, LC3 in ovarian cancer ascites and clinical pathological data In the 74 cases of ovarian cancer ascites, the expression of sAMY-1α was signi cantly difference in group with and without chemotherapy (p<0.001). The expression of HE4 was statistically signi cant with chemotherapy, histological type and differentiation (p<0.001). Serum CA125 was signi cantly difference with chemotherapy (p<0.05), differentiation (p<0.001), and lymph node metastasis (p<0.05). The expression of ascitic CA125 had no correlation with any of these clinicopathologic data. Beclin1 was expressed lower in group with chemotherapy than that without (p<0.01) and highly expressed in patients less than 50 years of age (p<0.01). The expression of LC3 was statistically signi cant with chemotherapy and lymph node metastasis (p<0.001). Show in Table 1.
Correlation among the expression of sAMY1α, Beclin1, and LC3 The ovarian cancer cell lines OVCAR3(relative resistance) and A2780(sensitivity) were treated with autophagic inhibitor (3MA) and inducer (rapamycin). OVCAR3 and A2780 were treated with different concentration of 3MA(5mM, 7.5mM, 10mM) and rapamycin (50nM,100nM, 150nM). The results of 3MA group showed that as the drug concentration increased, there were decreased in the protein levels of LC3II and Beclin1, and sAMY1α levels increased. There was opposite result in rapamycin group, there were increased in the protein levels of LC3II and Beclin1, and sAMY1α levels decreased (Fig. 6A,B).
Subsequently, OVCAR3 and A2780 cells were treated with cisplatin (CDDP) (20 μM) alone or then add rapamycin. Compared with control group, cells that were treated with CDDP revealed signi cantly decreased levels of LC3II, Beclin1, and increased levels of sAMY1α; CDDP+ rapamycin showed that autophagy inhibition was alleviated, and slightly reduced levels of sAMY1α. (Fig. 6A,B). This nding indicates that inhibition or induction autophagy can affect the expression level of sAMY1α. Addition, chemotherapy inhibits the process of autophagy activation and induces autophagic death of tumor cells.

Discussion
Due to the anatomical position of the ovary, early diagnosis of ovarian cancer is di cult. Once the patient has ascites, the cancer has entered the advanced stage 3,4,13 . The clinical treatment of patients with ovarian cancer ascites generally involves ascites pumping, rst-line chemotherapy, and resection of ovarian cancer. However, some patients will have ascites again. The occurrence of repeated ascites indicates that the patient's condition is more serious at this stage. At this stage, it is not only necessary to adopt corresponding treatment methods to reduce the disease, but also to nd valuable markers that can predict the recurrence of ascites.
High throughput screening (HTS) technology is to obtain a large amount of information through one experiment and nd valuable information from it 14,15 . Genechip is a highly integrated intelligent gene chip that uses micro-samples for detection. It can perform simultaneous research on the obtained large-scale and high-throughput thousands of genes 16 . Metabolomics technology has the advantage of high throughput which can detect thousands of metabolites and screen for differential metabolites. Compared with genomics, transcriptomics and proteomics, metabolomics is more connected with life activities 17 . The cancer marker test also has certain signi cance for the follow-up of patients after surgery. It can be judged whether the drug has a signi cant effect and whether to continue the examination or treatment. Therefore, metabolomics is expected to become a "weapon" for cancer screening [18][19][20] . In this study, AMY1A / AMY1B and AMY2B genes were screened from ovarian cancer tissues and ascites sedimentary cells by genechip and metabolomics techniques. Comparison with TCGA database also veri ed that serous ovarian cancer can express AMY genes. What is the clinical signi cance of the expression of AMY gene in ovarian cancer tissue or ascites, especially after chemotherapy? Little is known so far.
AMY is a general term for enzymes that hydrolyze starch and glycogen. The amylase in the body can be divided into pancreatic type and salivary gland type. The former is secreted by the exocrine glands of the pancreas and released into the blood; the latter is mainly secreted by the salivary glands. Amylase activity is also found in extracts from other tissues of the human body, such as ovaries, fallopian tubes, lungs, testes, semen and breast 21,22 . High expression of sAMY can be detected in the serum of patients with esophageal cancer, ovarian cystadenocarcinoma, hepatocellular carcinoma, prostate cancer bone metastases, lung bronchoalveolar adenocarcinoma, etc. [23][24][25] . In recent years, with increasing awareness of the effectiveness and reliability of sAMY activity detection, sAMY has become a sensitive biomarker. As a new type of tumor marker, sAMY has been widely used and concerned in the auxiliary diagnosis of various malignant tumors [26][27][28] .
The team selected sAMY1α protein to verify ovarian cancer tissues, ovarian cancer ascites supernatant and precipitated cells, and ovarian cancer cell lines. It was found that the increase in sAMY1α after chemotherapy was negatively correlated with the level of HE4 / CA125 in patients with ovarian cancer ascites, suggesting that the level of sAMY1α and HE4 expression in the ascites can be measured during the rst-line chemotherapy to predict the effect of chemotherapy. Previous studies have suggested that HE4 levels increased signi cantly in ascites after ovarian cancer chemotherapy. The expression level of HE4 in ovarian cancer ascites may re ect the treatment effect of ovarian cancer patients and high levels of HE4 may predict the resistance of chemotherapy and the possibility of ascites formation 29,30 . HE4 is a relatively new tumor marker, and has been used as a monitoring index for epithelial ovarian cancer in combination with CA125. The sAMY1α found in this study is expected to be used as a guideline for clinical diagnosis and treatment of patients with ovarian cancer ascites during the rst-line chemotherapy treatment.
Autophagy plays a dual role in promoting and inhibiting the apoptosis of tumor cells [31][32][33] . Autophagy is conducive to maintaining the homeostasis of cells. During stress, cell autophagy prevents the accumulation of toxic or carcinogenic damage proteins and organelles, and inhibits the cell from becoming cancerous; however, once tumors form, cell autophagy provides cancer cells with richer nutrition and promotes tumor growth 34 . The previous research results of the team concluded that autophagy has always been highly expressed to continuously provides nutrition for tumor cells, leading to recurrence or occurrence repeatedly of ovarian cancer ascites after chemotherapy, and also one of the causes of drug resistance 12 . It is known that there are 60 enzymes and many metabolites in lysosomes, among which sialidase is one 35 . The increase of sAMY1α in ovarian cancer ascites after chemotherapy is presumably due to the killing effect after chemotherapy, which caused lysosome collapse to release the enzyme during the process of autophagic cell death of tumor cells. In vitro, cytology further con rmed that inhibition or induction of autophagy in ovarian cancer cells can affect sAMY1α expression. Chemotherapy not only inhibits autophagy activation, but also induces autophagic cell death of tumor cell. Previous research by our team also con rmed that the effects of autophagy inhibitors 3MA and rapamycin on ovarian cancer cells OVCAR3 and A2780 have shown that autophagy inhibitors and chemotherapeutics can increase the sensitivity of ovarian cancer cells to drugs to promote tumor cell death.
The production of cancerous ascites is a very di cult practical problem for the treatment of this patient, which is an indisputable fact. How to predict the recurrence of ascites due to chemotherapy resistance in advance. There are many markers currently used for the diagnosis of ovarian cancer, such as CA125, HE4, CEA, CA199, KLK6 and so on 36,37 . HE4 is a newly proposed tumor marker. It was a new and promising ovarian cancer marker rst discovered in the early 1990s and can be used as an independent indicator to predict the prognosis of ovarian cancer 38,39 . CA125 has been proposed as a speci c marker for the diagnosis of ovarian cancer as early as the early 1980s. However, CA125 has recently been found to have high sensitivity and poor speci city in the detection of ovarian cancer, with some false positives 40 .
After ovarian cancer ascites chemotherapy, sAMY1α increases and Beclin1 and LC3 levels decrease, which can accelerate the formation of autophagosomes into autophagolysosomes. The lysosomes release AMY into the ascites during degradation, which can be used as a marker for autophagy death of cancer cells in ascites. Patients with ovarian cancer ascites follow the detection of sAMY1α levels in ascites during the rstline chemotherapy, combined with the expression level of HE4 and CA125 in the patients' serum, which provides important reference value judge the e cacy of the treatment of patients with ovarian cancer ascites, monitor drug resistance, modify the treatment plan, etc.

Declarations
Availability of data and materials All data generated or analysed during this study are included in this published article and its additional information les.

Ethics approval and consent to participate
This study was approved by the ethics committee of The Third A liated Hospital of Harbin Medical University, and wrote informed consent was obtained from all patients. This study was approved by the ethics committee of Harbin Medical University.

Consent for publication
Not applicable.

Figure 2
The AMY gene in ovarian cancer. (A). Top 9 down-regulated genes (without vs. with) after genechip analysis. (B). Metabonomic data analysis of up-and down-regulated Metabolite (sensitivity/resistance). (C). The expression quantity of AMY gene in TCGA.