
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
Elsevier

S2405-8440(24)13378-6
10.1016/j.heliyon.2024.e37347
e37347
Research Article
Metformin dampens the progression of cholangiofibrosis induced by thioacetamide using deep learning
Li Chaofu lichaofu000123@126.com
a1
Deng Yating 2675133140@qq.com
b1
Pan Yating yumiko_pan@163.com
b1
Liao Xinyi cynthia_lxy628@163.com
b1
Xie Huadong 583386793@qq.com
c
Xue Xiaoli 476062360@qq.com
b
Yu Shaoqing 1300209@tongji.edu.cn
d⁎⁎⁎
Yu Wenlong 909240168@qq.com
e⁎⁎
Yu Guanzhen qiaoshanjushi@aliyun.com
df⁎
a Department of Oncology, Liuzhou Worker's Hospital, Guangxi, 545005, China
b Department of Oncology, Longhua Hospital Affiliated to Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China
c Department of Surgery, Liuzhou Worker's Hospital, Guangxi, 545005, China
d Allergy and Cancer Research Center, Tongji Hospital, School of Medicine, Tongji University, Shanghai, 200065, China
e Department of Biliary Tract Surgery, Shanghai Eastern Hepatobiliary Surgery Hospital, Naval Medical University, Shanghai, 200438, China
f Medical Artificial Intelligence Laboratory, Zhejiang Institute of Digital Media, Chinese Academy of Science, Shaoxing, 312366, China
⁎ Corresponding author. Allergy and Cancer Research Center, Tongji Hospital, School of Medicine, Tongji University, Shanghai, 200065, China. qiaoshanjushi@aliyun.com
⁎⁎ Corresponding author. 909240168@qq.com
⁎⁎⁎ Corresponding author. 1300209@tongji.edu.cn
1 These authors contributed equally to this work.

07 9 2024
30 9 2024
07 9 2024
10 18 e3734722 5 2024
29 8 2024
2 9 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Purpose

The consistent use of metformin has been linked to a reduced incidence of neoplastic diseases among diabetic populations. As a preventive intervention, metformin may offer a more favorable risk-benefit profile. Here, we explored the efficacy of metformin in the primary prevention of cholangiofibrosis, which can precede the carcinogen-induced development of cholangiocarcinoma (CCA). Our objective was to assess the potential of metformin to act as an intervention prior to the onset of these conditions.

Methods

A rat model of thioacetamide (TAA)-induced cholangiofibrosis was utilized to assess the impact of metformin on the induction process of cholangiocarcinoma (CCA). Liver tissues were harvested and analyzed histologically using light microscopy, complemented by a deep-learning convolutional neural network for enhanced evaluation. Additionally, RNA sequencing (RNA-seq) was performed to investigate the genetic alterations associated with metformin treatment in this TAA-induced cholangiofibrosis model.

Results

In the rat model, the TAA control group exhibited an increased incidence and average count of cholangiofibrosis cases in the liver, with rates of 100 % and an average of 12.0, compared to the metformin-treated group, which showed an incidence of 70 % and an average of 3.3. Notably, the progression from normal cholangioles to cholangiofibrosis was associated with the upregulation of several proteins critical for metabolic processes and the tumor microenvironment. These alterations were significantly mitigated by metformin treatment.

Conclusions

Long-term metformin use may offer protective benefits against cholangiofibrosis, partially by regulating metabolic processes and improving the tumor microenvironment.

Keywords

Cholangiofibrosis
Metformin
Thioacetamide
Artificial intelligence
Abbreviations

Cholangiocarcinoma CCA

Thioacetamide TAA

Convolutional neural network CNN

Whole slide image WSI

Immunohistochemistry IHC

Epithelial-mesenchymal transition EMT
==== Body
pmc1 Introduction

Worldwide, the incidence of cholangiocarcinoma (CCA) is fluid and complex [[1], [2], [3]]. The overall prognosis for most cases of CCA is dismal, and the median survival is 11.3 months for extrahepatic cholangiocarcinoma and 6.2 months for intrahepatic cholangiocarcinoma [4,5]. Several risk factors contribute to CCA, including primary sclerosing cholangitis, biliary stone disease, liver flukes, hepatitis C, and unhealthy lifestyles (smoking and alcohol) [[6], [7], [8]]. Therefore, chemoprevention of the damage caused by these risk factors has great potential in decreasing the incidence and mortality of CCA. Evidence from retrospective clinical studies suggests that the use of metformin is linked to a decreased risk of cancer development [9,10]. Metformin exerts its anti-inflammatory effects and modulates macrophage polarization through the activation of AMPK inhibits nuclear factor-κB activation, reduces the expression of inflammatory genes, and ultimately mitigates inflammatory injury [11,12]. Prevention with metformin in cohorts associated with a high risk for CCA may offer additional advantages concerning the risk-versus-benefit profile. However, the safety and duration of metformin as a chemopreventive agent are debated because prolonged use of metformin may lead to a deficiency in vitamin B12 (megaloblastic anemia) if left unchecked, which could lead to irreversible nerve damage [13,14]. This study was conducted to evaluate both the efficacy and safety of metformin in the prevention of CCA development.

To assess the impact of metformin on cancer development, we enrolled a rat model of cholangiofibrosis preceding the development of CCA, which was orally administered thioacetamide (TAA). This results in a chemically induced natural transition from normal cholangioles to biliary dysplasia and ultimately to cholangiofibrosis, a risk factor for CCA [15,16]. This model is useful for designing strategies for preclinical chemoprevention and therapeutic trials. Here, we utilized the TAA-induced cholangiofibrosis model to investigate whether metformin could prevent or diminish the incidence of cholangiofibrosis in vivo. Moreover, the potential mechanisms underlying metformin's inhibition of tumor cell growth were further investigated through cDNA microarray and RNA-seq analysis.

2 Materials & methods

2.1 Animal experimental protocol

10-week-old male Sprague-Dawley (SD) rats were used. To reduce potential variability associated with hormonal cycles, only male SD rats were selected for the metformin experiments aimed at preventing cholangiofibrosis. These animals were categorized into two groups: a control group receiving TAA alone (n = 16) and an experimental group receiving TAA in combination with metformin (n = 16). All rats were given 300 mg/L TAA (CAS: 27366-72-9, Hangyubiotech, Changzhou, China) mixed in their drinking water daily until they were sacrificed. Additionally, the experimental group received 250 mg/kg of metformin via intragastric administration every day, as informed by previous studies [17,18], whereas the control group was administered an equal volume. of regular water via intragastric administration for 20 weeks. Eight weeks after the initiation of TAA alone or TAA combined with metformin, two animals were harvested each month to assess the tumor-promoting effects of TAA and the safeguarding effects of metformin. After a total of 20 weeks, all rats were sacrificed for further evaluation. All rats were maintained under standard housing conditions.

2.2 Liver harvesting procedures and histopathological evaluation

At the designated time points, the animals were anesthetized with pentobarbital sodium (40 mg/kg, intraperitoneally). A midline laparotomy was carried out to thoroughly inspect and assess all liver lobes for morphological changes and the incidence of cholangiofibrosis. Blood samples were obtained for subsequent analyses, including liver function tests and vitamin B12 deficiency assessments. The liver was then rinsed with normal saline and cut into sections measuring 3–5 mm at 5 mm intervals. Half of the liver tissue was stored at −80 °C for future studies, while the remaining portion was fixed in 4 % paraformaldehyde for histopathological examination.

2.3 Area ratio calculation

All slides of liver samples were digitized by Ningbo Jiangfeng biological information technology Co., Ltd. The tissue regions were extracted using the Otsu thresholding method, followed by morphological operations to refine the results, allowing for the calculation of the tissue areas (Atissue). To enhance lesion detection, we established an automated framework utilizing convolutional neural networks (CNNs) designed to identify and localize fibrosis and cholangiofibrosis regions within the whole slide images (WSIs). The detailed procedures were described in our previous report [19]. Finally, a ratio of liver lesions for each slide was generated: ratio = Alesion/Atissue.

2.4 RNA-seq and data analysis

Sequencing and analysis were conducted on liver samples from four rats treated with TAA alone and five rats treated with both TAA and metformin. Total RNA was extracted using TRIzol reagent (Invitrogen) and purified with the RNeasy Mini Kit (Qiagen). RNA-seq libraries were prepared following the Illumina Standard Library Preparation Kit guidelines. The sequencing was performed on the Illumina HiScan™ 2500 platform at Shanghai Oe Biotech Co., Ltd. For quality control of the raw sequencing data, we utilized FastQC and NGS QC Toolkit v2.3.3. Clean reads were then aligned to the B73 reference genome (RefGen_v3) and the reference gene model dataset (FGS 5b) using TopHat/Bowtie2 (ccb.jhu.edu/software/tophat/). Gene expression levels were normalized as fragments per kilobase of exon model per million mapped reads (RPKM). A gene was deemed expressed if its PFKM value exceeded zero across all biological replicates. Differentially expressed genes were identified using a false discovery rate (FDR) threshold of <0.05, as determined by the DESeq software package (Bioconductor.org/) [19].

2.5 Immunohistochemistry (IHC) and assessment of liver specimens

Paraffin-embedded sections of TAA-induced tumors, cut to a thickness of four microns, were prepared for immunohistochemical analysis. Antibodies targeting S100A6 (H-55) and S100A10 (4E7E10) were sourced from Santa Cruz Biotechnology. Visualization of antibody binding was achieved using a streptavidin-biotin kit (#KIT-9720; Maixin-Bio). Two evaluators, G. Yu (a researcher) and Y. Chen (a pathologist), independently assessed all samples. The final outcomes were aggregated according to a semiquantitative scoring system, as outlined in previous studies [18,20].

2.6 Western blotting

Whole-cell lysates were obtained from the livers of rats in both the TAA and TAA + metformin groups. The lysates were subjected to SDS/PAGE for resolution and subsequently transferred to PVDF membranes (Bio-Rad Laboratories, Hercules, CA, USA) via electrophoresis. Following the transfer, the membranes were incubated with the specified antibodies, and the resulting signals were detected using an enhanced chemiluminescence (ECL) kit from Santa Cruz (CA, USA).

2.7 Statistics

All statistical analyses were conducted using SPSS software (Chicago, IL) and GraphPad Prism (version 5.0). Categorical data were evaluated with χ2 tests. To compare protein expression levels between normal and tumor tissues, we employed a two-tailed Student's t-test [20]. A P value below 0.05 was considered to indicate statistical significance.

3 Results

3.1 Systemic effects of TAA and TAA + metformin

The experimental outline for the evaluation of metformin-induced inhibition of TAA-induced cholangiofibrosis is shown in Fig. S1A. Throughout the 20-week study period, there were no occurrences of mortality related to TAA or metformin. Additionally, no significant differences were noted in body weight, liver condition, liver volume, or liver weight between the TAA group and the TAA + metformin group (Figs. S1B–D). Biochemistry test results, including those for AST, ALT, and vitamin B12, were similar for both groups (Fig. S1E).

3.2 The rate of nodule formation in the TAA and TAA + metformin groups

We began by carefully inspecting the white nodules on the liver's surface before proceeding to slice the liver into sections (Fig. 1A). The slices were subsequently evaluated for the number of white nodules present (Fig. 1B). Finally, we recorded the total number of nodules on both the liver surface and the slices. In the TAA group, the average number of macroscopic nodules was 7.4 ± 5.7, which was notably greater than the 1.2 ± 2.8 observed in the TAA + metformin group. (Fig. 1C).Fig. 1 The morbidity of cholangiolar proliferation and cholangiofibrosis in each group at 20 weeks. A, At 20 weeks, all rats were sacrificed and the livers were sliced into 5-mm pieces. B, Nodules on the surface of the liver and inside the liver were counted. C, Graphical representation of the numbers of visible nodules of each liver in the two groups. D, Representative images of the liver lesions from the two groups. Small figures are enlargement of the yellow box. HE, Hematoxylin-eosin staining; AI, Artificial intelligence. Small figures, 20 × ; large figure, 1 × . E, The ratio of liver lesions evaluated by AI in the two groups. F, The average numbers of microscopic cholangiofibrosis of the liver in the two groups. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 1

Next, we conducted light microscopy and CNNs to detect and evaluate the liver sections. Generally, the liver size decreased, and the liver surface became rougher in the TAA group, whereas the liver surface in the TAA + metformin group appeared remarkably smooth (Fig. 1D). There were two obvious characteristics in the liver sections: cholangiolar proliferation and cholangiofibrosis, characterized by CK20-positive staining (Fig. S2A). We developed an AI algorithm to evaluate the ratio of cholangiolar proliferation as Alesion/Atissue using deep-learning CNNs. This AI algorithm precisely located all cholangiolar proliferation (Fig. 1D) and provided an exact ratio (Fig. 1E). The ratio of cholangiolar proliferation in the TAA group was 19.6 ± 9.01, significantly greater than that in the TAA + metformin group (6.7 ± 7.4; Fig. 1D and E) (P < 0.001). Due to limited time, we didn't develop our AI algorithm to recognize cholangiofibrosis as a unique characteristic. Instead, we calculated the number of cholangiofibrosis visually. At 20 weeks, all rats in the TAA group exhibited cholangiofibrosis, and with 12 ± 10.1 microscopic cholangiofibrosis observed per rat. Strikingly, 70 % (7/10) of rats in the TAA + metformin group exhibited a rate of microscopic cholangiofibrosis of 3.3 ± 6.7 per rat, whereas the other 30 % did not display any development of cholangiofibrosis (P < 0.05) (Fig. 1F).

3.3 Metformin inhibits TAA-induced cholangiolar proliferation

Next, we analyzed microscopic lesions of the liver during the 20-week study period. We mainly focused on the cholangiolar proliferation recognized through the AI algorithm. The proportion of microscopic cholangiolar proliferation in the TAA and TAA + metformin groups gradually increased with the administration of TAA revealed by both histological examination and AI (Fig. 2A and B). The baseline ratio of lesions in the TAA + metformin group (3.88 ± 2.78 at 12 weeks, 3.33 ± 1.91 at 16 weeks, 6.7 ± 7.4 at 20 weeks) was remarkably lower than that of the TAA group (5.98 ± 5.62 at 12 weeks, 27.57 ± 23.57 at 16 weeks, 19.6 ± 9.01 at 20 weeks) (Fig. 2C). These data further confirm that metformin retards the development of liver lesions induced by TAA.Fig. 2 The microscopic profiles of the liver lesions during TAA induction and/or metformin administration. A and B, Representative images of microscopic (upper) and AI-recognized (lower) cholangiolar proliferation of the liver in the TAA alone induction group (A) and in the TAA + metformin group (B). C, The average numbers of AI-recognized liver lesions from the TAA alone and the TAA + metformin group at the indicated time. ***P < 0.001.

Fig. 2

3.4 Metformin inhibits TAA-induced activation of multiple signaling pathways

Next, we explored the possible mechanisms underlying the metformin-dependent decrease in TAA-induced cholangiofibrosis using RNA-seq analysis. A number of genes were altered after long-term use of metformin compared with TAA induction alone (Fig. 3A and B). Metformin significantly influenced several signaling pathways, particularly those involving genes related to glycolysis, focal adhesion, and cancer progression (Fig. S2B). Notably, the ERBB2 signaling pathway was inhibited by metformin treatment. Furthermore, metformin treatment in TAA-induced cholangiofibrosis led to a 3- to 6-fold downregulation of key enzymes in glycolysis, such as hexokinase (HK), phosphofructokinase (PFK), and pyruvate kinase (PKM). Proteins in the HIF-1 signaling pathway, including Serpine1, Timp1, and Edn1, were also decreased by metformin treatment. Considering that the AMPK signaling pathway is the main target of metformin, we also examined the expression profiles of genes within this pathway. Three members, including Acacb, Irs3, and Pck1, showed increased expression upon metformin treatment, whereas another five members showed decreased expression upon metformin treatment (Fig. 3C).Fig. 3 RNA-seq analysis reveals differential gene expression between the livers from rats in the TAA group and the TAA + metformin group. A, Four livers from the TAA group and five livers from the TAA + metformin group were subjected to RNA-seq analysis and distributions of all genes are presented in the volcano map. B, Heatmap of significantly altered genes between the TAA group and the TAA + metformin group. C, Several proteins from multiple signaling pathways were significantly decreased upon metformin treatment compared that those without metformin treatment.

Fig. 3

Histopathology revealed that TAA-induced cholangiofibrosis were surrounded by intense stromal desmoplasia, along with inflammatory cell invasion [15]. Considering that S100 proteins have been shown to have a strong association with the tumor microenvironment (TME) [21], we investigated whether metformin treatment could influence their expression. Notably, several S100 proteins (S100A1, -4, -6, -8, -9, -10, and −11) were found to be highly expressed in the TAA group. Following metformin treatment, their expression levels were markedly decreased, showing a downregulation of 3- to 8-fold (Fig. 4A). Further IHC revealed that the expression levels of S100A6 and S100A10 was lower in specimens from the TAA + metformin group (66.7 ± 21.6 for S100A6 and 49.2 ± 15.6 for S100A10) than in the TAA group (141.7 ± 49.2 for S100A6 and 126.7 ± 41.8 for S100A10) (P < 0.01) (Fig. 4B). Moreover, metformin could inhibit the expression of S100A6 and S100A10 as revealed by western blotting (Fig. S3).Fig. 4 Metformin improves liver microenvironment by decreasing the expression levels of several proteins in the S100 family. (A) A number of S100 proteins was significantly decreased upon metformin treatment compared that those without metformin treatment. (B) Low expression of S100A6 and S100A10 in liver lesions and tumors from TAA + Metformin group compared that from TAA group by IHC.

Fig. 4

3.5 Mutation profiles between TAA group and TAA + metformin group

Considering the decreased incidence of cholangiofibrosis in rats and improved liver microenvironment induced by metformin, we investigated whether metformin treatment resulted in a low frequency of genetic aberrations. Unfortunately, neither the overall types of mutations nor missense mutations varied significantly between the two groups (Fig. S4). These data suggest that, once the mutations occur, they cannot be reversed, even if the tumor was controlled.

4 Discussion

The association of metformin with antineoplastic activity and its ability to reduce the burden of various cancers positions it as a promising therapeutic strategy for cancer treatment [[22], [23], [24]]. Metformin intake has been proved to improve clinical outcome in advanced cholangiocarcinoma patients after starting chemotherapy [25] and extend the survival of patients with advanced extrahepatic cholangiocarcinoma who have received drainage treatment [26]. However, metformin administration as an adjuvant therapy is debated in advanced solid cancers [27]. In particular, metformin treatment alone is unable to effectively inhibit the growth of various human solid cancer xenografts [17,28]. These results suggest that after malignancies become established, metformin might not offer any therapeutic potential [29]. However, increasing evidence suggests that metformin is supposed to be more effective for the chemoprevention of human solid cancers [29,30]. A phase 3 trial demonstrated the preventive effect of low-dose metformin (250 mg daily) in decreasing both the prevalence and the number of metachronous adenomas and polyps following polypectomy [31]. Given the relationship between metformin use and a decreased risk of ICC [32], we believe that metformin use prior to tumor establishment could decrease the incidence of TAA-induced cholangiofibrosis. Not surprisingly, long-term use of metformin indeed reduces the prevalence and number of cholangiofibrosis induced by TAA.

Metformin effectively inhibits cell growth by inducing cell cycle arrest and skewing macrophage polarization [33,34]. Moreover, not only are genes involved in cell cycle progression influenced by metformin, but genes associated with metabolic processes, response to stress, and protein transport or localization are also targeted by metformin [33]. Therefore, it is essential to explore the chemopreventive role of metformin in CCA. TAA-induced cholangiofibrosis is an ideal model for our study because this model produces a 100 % rate of cholangiofibrosis, a risk factor for CCA, by the 22 nd week of treatment. TAA is a potent hepatotoxin that interferes with ribosomal activity, thus hindering protein synthesis [35] and stimulating DNA synthesis [36].

Metformin use as a chemopreventive agent has been widely explored. Here, we introduced an AI algorithm to assess all H&E slides [37], avoiding the subjective judgment from the pathologists. Surprisingly, prolonged administration of metformin led to a 30 % reduction in TAA-induced cholangiofibrosis in all rats and decreased the incidence of tumor-like nodules in each rat treated with metformin. In addition to visible nodules, another major characteristic of our carcinogenesis model is cholangiolar proliferation [15]. Long-term use of metformin could improve the quality of the liver tissue by reducing the area of cholangiolar proliferation. Thus, the ‘soil’ of the liver destroyed by TAA will be restored by metformin. Importantly, we observed no side effects, such as gastrointestinal disturbances or vitamin B12 deficiency, associated with the long-term use of metformin when vitamin B12 was supplemented through dietary sources.

The mechanisms through which metformin exerts its effects have been extensively studied. Notably, metformin has the potential to reverse mesenchymal and epithelial-mesenchymal transition (EMT) characteristics in intrahepatic CCA by activating AMPK-FOXO3-related pathways [24]. Here, we expanded its potential mechanism against TAA-induced liver diseases (Fig. 5). Given that metformin treatment did not lower the frequency of genetic aberrations, its chemopreventive effects may be attributed to the functional alterations resulting from genetic mutations. ERBB2 is overexpressed in invasive CCAs, and upon metformin treatment, ERBB2 was down-regulated in our cDNA and RNA-seq datasets. In addition to ERBB2, two other members of the ERBB signaling pathway, PAK6 and SHC2, were also decreased upon metformin treatment. Similarly, members of the HIF signaling pathway were also down-regulated upon metformin treatment. Tumor cells can exhibit glycolytic rates that are over 200 times greater than those of the normal tissues from which they originate. Interestingly, treatment with metformin resulted in decreased expression of key rate-limiting enzymes in glycolysis, such as HK, PFK, and PKM. This suggests a novel mechanism by which metformin may prevent cancer by reducing glycolysis and effectively starving cancer cells. As an activator of the AMPK signaling pathway, metformin administration influenced the expression of over eight genes within this pathway in rats with TAA-induced cholangiofibrosis. Additionally, the S100 protein family plays diverse roles in both intracellular and extracellular environments, particularly in relation to the tumor microenvironment [21,38]. It is well recognized that the S100 family facilitates communication between cancer cells and stromal cells. Our finding that long-term metformin use significantly reduces the expression of various S100 proteins helps explain the relatively improved liver condition observed in rats from the TAA + metformin group compared to those receiving TAA alone. However, the evidence suggesting metformin's role in inhibiting tumorigenesis remains indirect. Therefore, additional research is required to elucidate the direct impact of metformin on these signaling pathways.Fig. 5 A diagrammatic figure summarizing the main findings and conclusions for the current proposal.

Fig. 5

However, this study has several limitations. First, the dosages of metformin used in the animal protocol are relatively high compared to human clinical exposure. Further research should systematically evaluate a range of dosages to identify the minimum effective dose and better correlate these findings with human clinical levels. Second, the role of sex hormones in mediating the response to metformin treatment has not been investigated. To address this, we plan to conduct follow-up studies that include both male and female mice to examine any sex-specific effects. Additionally, more relevant animal models that closely mimic the progression to cholangiocarcinoma (CCA) [39], including genetically engineered mouse models, are necessary to confirm metformin's preventive and therapeutic effects in both early and advanced stages of the disease. Moreover, the precise mechanism by which metformin inhibits tumor initiation remains unclear. An in-depth investigation of its mechanisms and a more rigorous cohort study will be essential for future research.

In summary, our data demonstrate for the first time that the use of metformin is effective in inhibiting or delaying TAA-induced biliary disease in rats by targeting multiple signaling pathways essential for tumor formation. This evidence provides a valuable scientific basis for the potential contribution of metformin to the chemoprevention of tumors, especially for those with high-risk factors.

Ethics statement

All animal protocols for this study received approval from the Experimental Animal Ethics Committee at Longhua Hospital, which is affiliated with Shanghai University of Traditional Chinese Medicine, China (Approval number: PZSHUTCM210514001).

Funding

This research received partial support from the Self-Funded Research Project of the Guangxi Zhuang Autonomous Region Health Commission (Z-B20221452 ) and the 10.13039/501100001809 National Natural Science Foundation of China (No. 81972721 ).

Data availability statement

The RNA-seq datasets featured in this study are available in online repositories. The repository names and corresponding accession numbers can be accessed at: https://www.ncbi.nlm.nih.gov/bioproject/795823.

CRediT authorship contribution statement

Chaofu Li: Writing – original draft, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Yating Deng: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Yating Pan: Methodology, Investigation, Formal analysis, Data curation. Xinyi Liao: Methodology, Investigation. Huadong Xie: Investigation, Data curation. Xiaoli Xue: Methodology, Data curation. Shaoqing Yu: Supervision, Software, Resources, Methodology, Conceptualization. Wenlong Yu: Software, Project administration, Methodology, Conceptualization. Guanzhen Yu: Writing – review & editing, Writing – original draft, Resources, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A Supplementary data

The following is the supplementary data to this article:

Acknowledgements

We would like to thank Xing Zhao, Wenzheng Fang, Ying Chen, Peilian Wei, Yong Gao, Haiyan Song, and Wenlian Chen for their contributions to the implementation of the animal experiments, data processing, and valuable discussions.

Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.heliyon.2024.e37347.
==== Refs
References

1 Brindley P.J. Bachini M. Ilyas S.I. Cholangiocarcinoma Nat. Rev. Dis. Prim. 7 2021 65 34504109
2 Rodrigues P.M. Olaizola P. Paiva N.A. Pathogenesis of cholangiocarcinoma Annual review of pathology 16 2021 433 463
3 Qurashi M. Vithayathil M. Khan S.A. Epidemiology of cholangiocarcinoma Eur. J. Surg. Oncol. : the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology 2023 107064
4 Flemming J.A. Zhang-Salomons J. Nanji S. Increased incidence but improved median overall survival for biliary tract cancers diagnosed in Ontario from 1994 through 2012: a population-based study Cancer 122 2016 2534 2543 27183133
5 Khan S.A. Genus T. Morement H. Global trends in mortality from intrahepatic and extrahepatic cholangiocarcinoma J. Hepatol. 71 2019 1261 1262 31558288
6 Kodali S. Connor A.A. Brombosz E.W. Update on the screening, diagnosis, and management of cholangiocarcinoma Gastroenterol. Hepatol. 20 2024 151 158
7 McGee E.E. Jackson S.S. Petrick J.L. Smoking, alcohol, and biliary tract cancer risk: a pooling Project of 26 prospective studies J. Natl. Cancer Inst. 111 2019 1263 1278 31127946
8 Fung B.M. Lindor K.D. Tabibian J.H. Cancer risk in primary sclerosing cholangitis: epidemiology, prevention, and surveillance strategies World J. Gastroenterol. 25 2019 659 671 30783370
9 Foretz M. Guigas B. Viollet B. Metformin: update on mechanisms of action and repurposing potential, Nature reviews Endocrinology 19 2023 460 476 37130947
10 Galal M.A. Al-Rimawi M. Hajeer A. Metformin: a dual-role player in cancer treatment and prevention Int. J. Mol. Sci. 25 2024
11 Salminen A. Hyttinen J.M. Kaarniranta K. AMP-activated protein kinase inhibits NF-kappaB signaling and inflammation: impact on healthspan and lifespan J. Mol. Med. 89 2011 667 676 21431325
12 Chiang C.F. Chao T.T. Su Y.F. Metformin-treated cancer cells modulate macrophage polarization through AMPK-NF-κB signaling Oncotarget 8 2017 20706 20718 28157701
13 Aroda V.R. Edelstein S.L. Goldberg R.B. Long-term metformin use and vitamin B12 deficiency in the diabetes prevention program outcomes study J. Clin. Endocrinol. Metab. 101 2016 1754 1761 26900641
14 Grober U. Kisters K. Schmidt J. Neuroenhancement with vitamin B12-underestimated neurological significance Nutrients 5 2013 5031 5045 24352086
15 Yeh C.N. Maitra A. Lee K.F. Thioacetamide-induced intestinal-type cholangiocarcinoma in rat: an animal model recapitulating the multi-stage progression of human cholangiocarcinoma Carcinogenesis 25 2004 631 636 14656942
16 Adams E.T. Auerbach S. Blackshear P.E. Proceedings of the 2010 national toxicology program satellite symposium Toxicol. Pathol. 39 2011 240 266 21177527
17 Yu G. Fang W. Xia T. Metformin potentiates rapamycin and cisplatin in gastric cancer in mice Oncotarget 6 2015 12748 12762 25909163
18 Yu G. Yu W. Jin G. PKM2 regulates neural invasion of and predicts poor prognosis for human hilar cholangiocarcinoma Mol. Cancer 14 2015 193 26576639
19 Li C. Zhao X. Gu X. The preventive role of hydrogen-rich water in thioacetamide-induced cholangiofibrosis in rat assessed by automated histological classification Front. Pharmacol. 12 2021 632045
20 Zhang Y. Zhu Y. Chen Y. Nuclear translocation of cleaved PCDH9 impairs gastric cancer metastasis by downregulating CDH2 expression iScience 27 2024 109011
21 Li M. Cui P. Dai W. S100 protein family: emerging role and mechanism in digestive tract cancer (Review) Int. J. Oncol. 64 2024 59 38666531
22 Saraei P. Asadi I. Kakar M.A. The beneficial effects of metformin on cancer prevention and therapy: a comprehensive review of recent advances Cancer Manag. Res. 11 2019 3295 3313 31114366
23 Yu H. Zhong X. Gao P. The potential effect of metformin on cancer: an umbrella review Front. Endocrinol. 10 2019 617
24 Di Matteo S. Nevi L. Overi D. Metformin exerts anti-cancerogenic effects and reverses epithelial-to-mesenchymal transition trait in primary human intrahepatic cholangiocarcinoma cells Sci. Rep. 11 2021 2557 33510179
25 Casadei-Gardini A. Filippi R. Rimini M. Effects of metformin and vitamin D on clinical outcome in cholangiocarcinoma patients Oncology 99 2021 292 299 33626532
26 Wu J. Zhou Y. Wang G. Metformin use and survival in patients with advanced extrahepatic cholangiocarcinoma: a single-center cohort study in Fuyang, China Gastroenterol Res Pract 2021 2021 9468227
27 Kordes S. Pollak M.N. Zwinderman A.H. Metformin in patients with advanced pancreatic cancer: a double-blind, randomised, placebo-controlled phase 2 trial Lancet Oncol. 16 2015 839 847 26067687
28 Lipner M.B. Marayati R. Deng Y. Metformin treatment does not inhibit growth of pancreatic cancer patient-derived xenografts PLoS One 11 2016 e0147113
29 Bhat A. Sebastiani G. Bhat M. Systematic review: preventive and therapeutic applications of metformin in liver disease World J. Hepatol. 7 2015 1652 1659 26140084
30 Cheung K.S. Chung K.L. Leung W.K. Chemopreventive effect of metformin on gastric cancer development Gut and liver 16 2022 147 156 34158423
31 Higurashi T. Hosono K. Takahashi H. Metformin for chemoprevention of metachronous colorectal adenoma or polyps in post-polypectomy patients without diabetes: a multicentre double-blind, placebo-controlled, randomised phase 3 trial Lancet Oncol. 17 2016 475 483 26947328
32 Chaiteerakij R. Yang J.D. Harmsen W.S. Risk factors for intrahepatic cholangiocarcinoma: association between metformin use and reduced cancer risk Hepatology 57 2013 648 655 23055147
33 El-Arabey A.A. Abdalla M. Ali Eltayb W. Metformin: ongoing journey with superdrug revolution Adv. Pharmaceut. Bull. 9 2019 1 4
34 Wang J.C. Sun X. Ma Q. Metformin's antitumour and anti-angiogenic activities are mediated by skewing macrophage polarization J. Cell Mol. Med. 22 2018 3825 3836 29726618
35 Barker E.A. Smuckler E.A. Altered microsome function during acute thioacetamide poisoning Mol. Pharmacol. 8 1972 318 326 4402748
36 Morley C.G. Boyer J.L. Stimulation of hepatocellular proliferation by a serum factor from thioacetamide-treated rats Biochim. Biophys. Acta 477 1977 165 176 884110
37 Wang X. Chen Y. Gao Y. Predicting gastric cancer outcome from resected lymph node histopathology images using deep learning Nat. Commun. 12 2021 1637 33712598
38 Chen H. Xu C. Jin Q. S100 protein family in human cancer Am. J. Cancer Res. 4 2014 89 115 24660101
39 Li M. Zhou X. Wang W. Selecting an appropriate experimental animal model for cholangiocarcinoma research Journal of clinical and translational hepatology 10 2022 700 710 36062286
