Introduction

ICCA represents a fatal digestive malignancy that originates from the intrahepatic bile duct epithelium and peribiliary glands [1, 2]. Although the global incidence of iCCA is relatively low, there has been a rapid increase in its occurrence over the past decade [2,3,4,5]. Furthermore, the long-term survival of iCCA patients remains unsatisfactory. For those with unresectable tumors, the untreated median survival is approximately 3–6 months [6], while even for those who undergo radical surgery, the clinical outcomes are still discouraging, with a 5-year survival rate ranging from 20 to 35% [7, 8]. Despite recent advancements in targeted or immunotherapeutics for iCCA, which may provide survival benefits to patients, the overall efficacy of these treatments remains pessimistic, and the mortality continues to rise [9]. Therefore, further investigation into the intrinsic molecular mechanisms underlying iCCA progression is imperative for the identification of novel therapeutic targets and the enhancement of patient prognosis.

Actually, the malignant biological behaviors of tumors are actuated by a complex cascade of factors and steps, the intricate cross-regulatory association between cancer cells and the tumor microenvironment plays an essential role in cancer development. Platelets are small anucleate cells split off from bone marrow megakaryocytes that circulate in the bloodstream [10]. Numerous studies have demonstrated that the reciprocal regulatory relationship between tumor cells and platelets is a key factor in the malignant progression of tumors. Platelets actively participate in tumor growth and metastasis, and tumor cells can also activate platelets to amplify this feedback loop [11, 12]. In the clinic, high platelet counts tightly associate with increased risk of various gastrointestinal tumors, such as esophageal, stomach and colon carcinomas [13]. Platelet-related parameters also correlate with specific clinicopathological features and contribute to tumor diagnosis and prognosis prediction [14,15,16]. Therefore, targeting platelets that engage in interactions with tumor cells is considered an innovative strategy for the prevention and treatment of cancer.

In this work, by analyzing the correlation between platelet-related parameters and the characteristics of iCCA patients, we have identified a significant association between platelets and the prognosis of iCCA patients. In vitro and in vivo experiments demonstrated that co-culturing iCCA cells with platelets significantly enhanced tumor cell motility and viability. Moreover, through proteomic analysis, we discovered two downstream targets ILK and ITGB3, which were up-regulated by platelet stimulation and participated in platelet-induced iCCA progression through activating FAK/PI3K/AKT signaling, suggesting that both ILK and ITGB3 could serve as promising therapeutic targets for iCCA patients, particularly those with abnormally elevated platelet counts.

Materials and methods

Patient specimens and clinical information

Human iCCA and adjacent non-tumorous specimens were procured from patients who underwent surgical resection of the bile duct at the Department of Biliary-Pancreatic Surgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China). The corresponding clinicopathological information, including preoperative and postoperative tests, surgical methods, pathological data, as well as follow-up and survival information, were also acquired individually. All the aforementioned clinical samples and case data were obtained with prior informed consent from patients and in accordance with the principles outlined in the Declaration of Helsinki. Additionally, the study also received approval from the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology.

Cell culture

Two human iCCA cell lines, HuCCT1 and RBE, as well as the HEK-293T cells were obtained from the Meisen Chinese Tissue Culture Collections (MeisenCTCC, Zhejiang, China). All cell lines have undergone short tandem repeat (STR) test and have been confirmed negative for mycoplasma contamination. The two iCCA cell lines were cultured in RPMI-1640 (Servicebio, Wuhan, China) supplemented with 10% fetal bovine serum (FBS; Vazyme, Nanjing, China), while the HEK-293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM). All cells were maintained at a temperature of 37℃ in an atmosphere containing 5% CO2.

Platelet separation and co-culture system

The isolation of human peripheral blood platelets and their co-culture with tumor cells were conducted following previously established protocols [17].

SiRNA construction, cell transfection, and lentivirus production

Small interfering RNAs (siRNAs) targeting both ILK and ITGB3, respectively, were designed and synthesized by Sangon Biotech (Shanghai, China), and their primer sequences were provided in Table S1. Cell transfection was performed via lipofectamine 2000 (Invitrogen, CA, USA) according to the manufacturer’s instructions. For lentivirus construction, the effective siRNA sequences were initially utilized for shRNAs synthesis by Sangon biotech, which were then integrated into the pLV3 vector to generate targeted knockdown plasmids. Subsequently, both the lentiviral plasmids and packaging plasmids (psPAX2 and pMD2.G) were co-transfected into HEK-293T cells. Approximately 8 h after transfection, the medium was replaced with fresh DMEM, and 72 h post-transfection, the viral supernatant was harvested and filtered through a 0.45 μm filter (Millipore, MA, USA). The naked HuCCT1 and RBE cells were infected with viral supernatant supplemented with polybrene transfection reagent (Yeasen, Shanghai, China) for 8 h and subsequently screened in the presence of puromycin for at least one week.

CCK-8, colony formation and EdU tests

For the CCK-8 test, differently treated iCCA cells were initially digested and seeded into 96-well plates at a density of approximately 2000 cells per well. After allowing the cells to adhere, the original medium was replaced with 100 ul of CCK-8 reaction solution (composed of RPMI-1640 and CCK-8 solution at a ratio of 9:1). Subsequently, the cells were incubated at 37℃ for 2 h under light protection, followed by measuring the absorbance at 450 nm. Five distinct time points were designated in a single experimental trial, and six repeated measurements were performed within each temporal interval across all experimental cohorts. For the colony formation test, the processed iCCA cells were seeded into 6-well plates at an approximate cell density of 1000 cells per well and allowed to grow continuously for 14 days with regular medium changes. The cell colonies were subsequently fixed with 4% paraformaldehyde (Servicebio) and stained with 1% crystal violet solution (Servicebio), and finally enumerated with a microscope. And for the EdU assay, 2 × 104 cells were initially seeded and allowed to grow on coverslips. Then, the cells were immobilized and stained using the BeyoClick™ EdU-555 cell proliferation test kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. The images were scanned via a fluorescence microscope, and five fields were randomly selected to analyze the relative cell proliferation rate with ImageJ.

Transwell invasion and migration, wound healing assays

For the transwell invasion and migration assays, differently treated iCCA cells were starved in serum-deprived RPMI-1640 for a duration of 6 h. Subsequently, a total of 5 × 104 cells in 200 µl of serum-deprived RPMI-1640 were planted onto the upper chambers (pre-coated with Matrigel for the invasion assay), and 700 µl of serum-replenished medium was added to the exterior compartments, then incubated at 37℃ and 5% CO2 for 24 h. After that, the chambers were fixed and stained as previously described, with subsequent removal of cells on the inner surface of the chambers while preserving those on the outer surface for further analysis. Similarly, five randomly selected fields were chosen to determine the relative rates of cell invasion or migration using ImageJ. For the wound healing assay, iCCA cells were seeded into 6-well plates and subjected to scratch assay using a 200 µl pipette tip upon reaching over 95% confluence. Then, the wells were washed with PBS twice and photographed at 0, 24, and 48 h, and the healing rate was eventually evaluated with ImageJ.

Flow cytometric analysis

Two milliliters of peripheral blood from healthy volunteers and iCCA patients was respectively collected into ACD anticoagulant vacuum tubes. The first tube of blood was discarded while the second tube was retained for further detection. Subsequently, the platelets were isolated following established protocols and then labeled with corresponding flow cytometry antibodies, including anti-human CD61 (FITC-labeled, pan-platelet surface biomarker), anti-human CD62P (P-selectin) (PE-labeled, platelet activation biomarker), anti-human CD41/CD61 (PAC-1) (Alexa Fluor 647-labeled, platelet activation biomarker), as well as the isotype control FITC Mouse IgG1, PE Mouse IgG1, and Alexa Fluor 647 Mouse IgM. The samples were then incubated at room temperature for 30 min protect from light before detection. The activation of platelets was evaluated by Beckman Coulter CytoFLEX flow cytometer (Suzhou, China), and the results were analyzed using FlowJo software. The main antibodies employed for flow cytometry were documented in Table S1.

Western blotting and real-time PCR analyses

For the western blotting (WB) assay, tumor cells were harvested and lysed with RIPA buffer containing both phosphatase and protease inhibitors. The lysate was then subjected to ultra-sonication and centrifugation, followed by collection of the supernatant, which was subsequently denatured by boiling with 5× loading buffer. Finally, the extracts were analyzed through SDS-PAGE and immunoblotting as previously described [17]. In the case of real-time PCR (RT-PCR), tumor cells were harvested and lysed using RNA-easy isolation reagent (Vazyme), then total RNA extraction was performed according to the manufacturer’s instructions. Subsequently, the extracted RNA was quantified and reverse-transcribed into cDNA using the HiScript II 1st strand cDNA synthesis kit (Vazyme). The cDNA templates, forward and reverse primers, as well as the fluorescent dye SYBR green, were added into a 10 µl reaction system, respectively. Then the RT-PCR was conducted using a Quantagene q225 qualified PCR system (Kubo Technology, Beijing, China). The relative mRNA levels of the target genes were determined via the 2−∆∆Ct formula, with β-actin serving as the internal control. All primers involved in this section were synthesized by Sangon Biotech, and a summary of the antibodies and primer sequences applied in WB and RT-PCR was also provided in Table S1.

Immunofluorescence assay

For the immunofluorescence (IF) assay, tumor cells were seeded onto coverslips and co-cultured with or without platelets for 24 h. Then, the cells were washed three times with PBS and immobilized with 4% paraformaldehyde for 15 min. Subsequently, the cells were permeabilized with 0.1% Triton X-100 for another 15 min, followed by blocking with 3% BSA for 2 h at room temperature. After that, the cells were incubated with indicated primary antibodies (Table S1) overnight at 4 °C, followed by the incubation with corresponding secondary antibodies for 2 h at room temperature. Finally, the coverslips were washed three times, and the images were acquired using a positive fluorescence microscope (DM2500, Leica, Germany).

In vivo experiments

The in vivo experiments in this study have received approval from the Laboratory Animal Welfare and Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (Ethical number: TJH-202206004). For the establishment of xenograft subcutaneous tumor model, 4-week-old female BALB/c nude mice were procured from Vital River (Beijing, China). Negative control, platelet-treated or ILK- and ITGB3-knockdown HuCCT1 cells were injected into the right axilla of nude mice at approximately 5 × 106 cells per mouse (n = 5 randomly each group), and the tumors were permitted to grow for a period of 24 days. For the lung metastasis model, differently treated HuCCT1 cells were injected into the mice via the tail vein at approximately 1 × 106 cells per mouse (n = 6 randomly each group), and the tumors were permitted to grow for a period of 28 days. During the construction phase of these models, the weights of the mice were recorded every 2 days, and the subcutaneous tumor volumes were measured every 3 days starting from the 7th day post-injection. Upon the conclusion of the experiments, all nude mice were euthanized with high concentrations of CO2. The subcutaneous tumors were weighed, and the volumes were calculated follow the formula: 1/2 × length × width2. Furthermore, each mice’s lungs were dissected, followed by fixation with 4% formaldehyde. Finally, the lung metastasis was evaluated by H&E staining.

Hematoxylin-eosin (H&E) and immunohistochemical (IHC) staining

The H&E and IHC analyses were conducted by Biossci Biotechnology (Wuhan, China). For human iCCA tissues and subcutaneous tumors of nude mice, the IHC staining was performed with indicated antibodies (Table S1), and the quantification of staining performance was achieved through images processing with Caseviewer 2.4.0 software. The IHC scores (ranging from 0 to 12) were determined by multiplying the scores of the staining intensities (0-negative; 1-weak; 2-moderate; 3-strong) with the proportion of positively stained cells (0, 0-4%; 1, 5-24%; 2, 25-49%; 3, 50-74%; and 4, 75-99%). Mouse lung tissues were embedded in paraffin and sectioned before H&E staining, and the number of lung metastatic nodules was quantified using an inverted microscope (Mshot, Guangzhou, China).

Bioinformatic analysis

For quantitative proteomic profiling, including screening for differential proteins, functional annotation, enrichment analysis, and construction of interaction networks, were conducted and provided by PTM-Biolab (Hangzhou, China). The raw data for each part of the analysis have been uploaded in the form of supplementary files in the corresponding sections of the paper. For gene expression analysis, two independent RNA-seq datasets containing integrated expression files were downloaded from The Cancer Genome Atlas (TCGA, TCGA-CHOL) and the Gene Expression Omnibus (GEO, GSE107943). The relative expression levels were analyzed and presented utilizing GraphPad Prism 9.

Statistical analysis

The statistical analysis was conducted via GraphPad Prism 9. Each experiment was independently repeated at least three times unless otherwise stated, and the results were presented as Means ± SDs. Survival differences were compared using the log-rank test. Discrepancies between categorical variables were assessed using the chi-square test. Cox regression analyses, including univariate and multivariate models, were employed to determine the associations between different clinical variables and patient prognosis. Spearman’s correlation test was utilized to assess associations between individual variables. Two-way ANOVA or Student’s t-test was employed for comparisons among groups. A significance level of p < 0.05 was considered statistically significant unless otherwise specified.

Results

Platelet-related characteristics in patients with iCCA

Overall, a total of 209 iCCA patients who met the inclusion criteria were gathered in this work. Based on the median preoperative platelet count (PLTc), these patients were classified into high- and low-PLTc subgroups. Subsequently, we investigated the relationships between PLTc and clinical characteristics of iCCA patients within different subgroups. The results illuminated that high PLTc (> 221 × 109/L) was significantly associated with aberrant CA-199 (> 34 U/mL, p < 0.001), vascular invasion (p < 0.001), and advanced tumor-node-metastasis (TNM) stage (TNM III-IV, p < 0.001) in iCCA patients (Table 1). Meanwhile, an obvious elevation in preoperative PLTc levels was discovered in the iCCA cohort compared to the control group consisting of 69 patients with benign biliary tract diseases, such as gallstones and gallbladder polyps (Fig. 1A). Kaplan-Meier (K-M) survival analysis revealed that high-PLTc patients suffered worse overall survival (OS) and recurrence-free survival (RFS) compared to those with low PLTc levels (Fig. 1B). To further evaluate the influence of different preoperative PLTc levels on patient prognosis, the entire iCCA cohort was divided into three subgroups according to the normal reference range of platelets (100 ~ 300 × 109/L). K-M analysis indicated that patients with increased PLTc ( > = 300 × 109/L) had poorer OS compared to both the normal and decreased subgroups, while this discrepancy was not statistically significant in RFS outcomes (Fig. S1A). Additionally, a drastically positive association between PLTc and CA-199 was also observed, indicating that increased PLTc might be connected with the malignancy of iCCA, similar to the conventional cancer biomarker CA-199 (Fig. S1B). Furthermore, univariate cox regression analysis identified a close correlation between.

patients’ OS and preoperative levels of CA-199 (> 34 U/mL, p = 0.035) and PLTc (> 221 × 109/L, p < 0.001), positive lymph nodes (p = 0.029), vascular invasion (p < 0.001), as well as TNM stage III-IV (p = 0.002). Subsequent multivariate analysis further determined preoperative PLTc (> 221 × 109/L, p = 0.017), vascular invasion (p = 0.007), and TNM stage III-IV (p = 0.029) as independent risk factors for iCCA patients’ outcomes (Table S2).

Table 1 Association of PLTc with different clinicopathological features in iCCA
Full size table

The above findings suggested a strong association between abnormally elevated preoperative PLTc and poorer survival outcomes in patients with iCCA. Additionally, the infiltration levels of platelets in tumor tissues were further analyzed using western blotting and immunohistochemical detection with platelet-specific biomarkers CD41, CD62P, and CD42b. The results discovered significantly higher IHC scores of platelet markers in tumors compared to adjacent normal tissues based on the analysis of 20 iCCA samples (Fig. 1C, Fig. S1C), which were further validated by WB analysis (Fig. 1D). These findings illuminated the increased presence of platelet infiltration within the tumor microenvironment of iCCA. Moreover, to determine the correlation between platelet infiltration and patient prognosis, a total of 70 iCCA patients were recruited and divided into different subgroups according to the staining intensity of the platelet-specific marker CD42b. K-M survival analysis demonstrated that patients with higher expression of CD42b exhibited worse OS and RFS (Fig. S1D). Not surprisingly, a significant positive association was also observed between CD42b expression and the PLTc (Fig. S1E), which further bolstered the credibility of platelets in prognostic prediction for iCCA. Similarly, an investigation into the correlation between CD42b expression and the clinicopathological parameters of 70 iCCA patients was also conducted. As shown in Table 2, the results demonstrated that high CD42b expression was positively related to elevated preoperative PLTc (> 221 × 109/L, p < 0.001), positive lymph nodes (p = 0.005), vascular invasion (p < 0.001), as well as TNM stage III-IV (p = 0.008). Univariate and multivariate cox analyses further identified preoperative PLTc (> 221 × 109/L, p = 0.034), vascular invasion (p = 0.046), TNM stage III-IV (p = 0.019), and high CD42b expression (p = 0.013) as independent risk factors for the prognosis of this iCCA cohort (Table S3).

Table 2 Association of CD42b expression with different clinicopathological features in iCCA
Full size table

In addition to the elevated infiltration of platelets in the tumor microenvironment, aberrant platelet activation was also closely associated with iCCA development. Here we assessed the relative platelet activation levels in 20 iCCA patients compared to 20 healthy adults by collecting their peripheral blood and analyzing it using flow cytometry. The results revealed that compared with healthy controls, preoperative platelet activation was significantly enhanced in iCCA patients, as evidenced by the augmentation of the incidence of PAC-1- and CD62P-positive cell populations. (Fig. 1E, Fig. S1F).

Fig. 1
Fig. 1
Full size image

Platelet-related characteristics in patients with iCCA. (A) Comparison of preoperative platelet counts (PLTc) between patients with iCCA and benign biliary tract diseases (including gallstones and gallbladder polyps). (B) Comparison of overall survival (OS) and recurrence-free survival (RFS) in patients at high- and low-PLTc groups divided by median preoperative PLTc (221 × 109/L). (C) Comparison of immunohistochemical (IHC) scores of platelet-related markers CD41, CD62P, and CD42b in 10 pairs of matched iCCA and para-tumoral tissues. (D) Relative expression of CD41, CD62P, and CD42b in tumors and paired adjacent normal tissues detected by western blotting assay. (E) Comparison of platelet activation levels (PAC-1+, CD62P+, and PAC-1+/CD62P+) between iCCA patients and healthy adults using flow cytometry. Data were described as means ± SD, ***p < 0.001

Taken together, the aforementioned findings have unveiled the potential value of platelet-related parameters in the clinical diagnosis and prognostic prediction of patients with iCCA, encompassing abnormally increased preoperative PLTc, as well as heightened levels of platelet infiltration and activation, which were tightly associated with poor prognosis of iCCA patients. Consequently, further investigation into the platelet-tumor interaction could be instrumental in elucidating the pro-tumoral mechanism of platelets and developing potential intervention targets for iCCA.

Platelet stimulation induces iCCA growth and metastasis

Numerous studies have demonstrated the crucial role of platelets in various aspects of human cancer development, including tumor growth and metastasis promotion, stimulation of proliferation and inhibition of apoptosis, as well as induction of angiogenesis [18]. These effects are mediated through a diverse range of mechanisms involving direct contact with platelets or dependence on platelet-derived cytokines or extracellular vesicles. More importantly, tumor cells can also “educate” platelets to promote their aggregation, activation and secretion functions, thereby forming a reciprocal cross-talk [19, 20]. In this section, the influence of platelet stimulation on iCCA cell proliferation, invasion and metastasis will be examined both in vitro and in vivo by co-culturing platelets and tumor cells.

Epithelial-mesenchymal transition tightly associates with tumor metastasis and can be triggered by platelet-derived factors such as transforming growth factor β (TGF-β), lysophosphatidic acid (LPA), as well as platelet-derived growth factor (PDGF), and microparticles via eliciting the activation of specific signals [21, 22]. Here, we employed WB, RT-PCR, and IF tests to confirm the impacts of platelet stimulation on the epithelial-mesenchymal characteristics of iCCA cell lines RBE and HuCCT1. As shown in Fig. 2A, by co-culturing HuCCT1 cells with platelets for 24 h, the cell morphology noticeably changed from round to fusiform shape, accompanied by an increase in the inter-cellular space and the acquisition of a mesenchymal appearance. The WB and RT-PCR results demonstrated a significant increase in the expression of mesenchymal markers N-cadherin (N-Ca), Vimentin (Vim), and transcriptional factor Snai1 in RBE and HuCCT1 cells co-cultured with platelets compared to the control without platelet stimulation, while an obvious down-regulation of E-cadherin (E-Ca) as indicated by the RT-PCR results, indicating that platelet-induced EMT occurred in these cell lines (Fig. 2B, Fig. S2A). Meanwhile, IF assay also observed the down-regulation of E-Ca and up-regulation of Vim after co-culturing with platelets (Fig. 2C). Additionally, considering the established associations between tumor metastasis and the matrix metalloproteinase (MMP) system as well as the integrin family [23, 24], the expression of related proteins was also detected via WB. The results showed that the expression of MMP3, MMP9, as well as platelet-related integrins CD41, CD42b, CD61, and CD62P were significantly up-regulated upon platelet stimulation (Fig. S2B). The above findings strongly indicated that platelet stimulation could induce EMT and upregulate the expression of tumor metastasis-related proteins in iCCA cells.

Therefore, it was reasonable to hypothesize that platelet stimulation could facilitate the invasion and migration of iCCA cells. To validate this hypothesis, we performed motility-associated functional assays using co-cultured and control RBE and HuCCT1 cells both in vitro and in vivo. Transwell invasion and migration assays confirmed a significant enhancement of iCCA cell invasion and migration upon platelet stimulation (Fig. 2D). The wound healing test consistently demonstrated that co-culturing with platelets could promote the motility of tumor cells (Fig. 2E). Furthermore, to investigate the impact of platelet stimulation on iCCA metastasis in vivo, we established lung metastatic mouse models by intravenously injecting platelet-treated HuCCT1 cells alongside control cells. The results discovered that compared to the negative control, co-culturing with platelets obviously promoted the lung metastasis of iCCA cells, as represented by the increase of pulmonary metastatic nodes (Fig. 2F). Taken together, these findings indicated that co-culturing iCCA cells with platelets markedly enhanced aggressiveness of tumor cells both in vitro and in vivo through actuating EMT.

In addition to tumor metastasis, the impact of platelet stimulation on the proliferative capacity of tumor cells was further investigated. The iCCA cells were pre-treated with platelets as described previously, and subsequent CCK-8, colony formation, and EdU assays were conducted to evaluate the influence of platelet stimulation on tumor cell proliferation. Additionally, iCCA mouse xenograft models were established for in vivo investigations. For the in vitro experiments, the EdU assay showed an obvious increase in the number of EdU-positive cells in the platelet treatment group compared to the negative control group, suggesting enhanced proliferation induced by platelet stimulation (Fig. 2G). The CCK-8 and colony formation tests also yielded similar results (Fig. 2H-I). And this effect was further validated through xenograft mouse models in which iCCA cell HuCCT1, pre-treated with or without platelets, was subcutaneously injected. Compared to the negative control, tumors in the platelet-stimulated group exhibited significantly accelerated growth, as evidenced by the obviously increased tumor weight and volume (Fig. 2J). Moreover, the IHC results also confirmed an augmentation in the proliferative capacity of iCCA cells following platelet stimulation. Tumors developed from co-cultured HuCCT1 cells displayed an increased expression of the multiplication marker Ki-67, along with the up-regulation of platelet-related markers CD42b and CD62P (Fig. S2C). These findings concluded that platelet stimulation could also expedite both in vitro and in vivo proliferation of iCCA cells.

Fig. 2
Fig. 2
Full size image

Platelet stimulation induces iCCA growth and metastasis. A. The morphological changes of HuCCT1 cells before and after co-culturing with platelets. Scale bars, 100 μm, 50 μm. B–C. Detection of the changes in expression of epithelial-mesenchymal transition (EMT) related markers in iCCA cells after co-culturing with platelets. Western blotting (B), and immunofluorescence (C) assays, E-Ca/Vim (Green, FITC-labeled), nucleus (Blue, DAPI staining). Scale bars, 20 μm. D–E. Evaluation of invasion and migration capabilities of iCCA cells with or without platelet stimulation. Transwell (D), wound healing (E) tests. Scale bars, 100 μm. F. Effect of platelet stimulation on in vivo metastatic ability of HuCCT1 cells (n = 6). General diagram of lung metastasis in nude mice (left), representative H&E staining images of lung metastasis (middle), statistical diagram of lung metastatic nodules (right). G–I. Evaluation of proliferation capabilities of iCCA cells with or without platelet stimulation. EdU (G), colony formation (H), and CCK-8 (I) assays. Scale bars, 50 μm. J. Effect of platelet stimulation on in vivo proliferation ability of HuCCT1 cells (n = 5). Image of xenograft subcutaneous tumors (left), tumor weight (middle), and growth curves (right). Data were described as means ± SD, **p < 0.01, ***p < 0.001

Proteomic profiling screens the downstream targets of platelet-mediated tumor promotion

The previous investigation has demonstrated that co-culturing iCCA cells with platelets facilitates tumor growth and promotes iCCA metastasis by inducing EMT. Therefore, to explore the intrinsic mechanism by which platelet stimulation influences tumor malignant biological behaviors, we performed 3D-label-free relative quantitative proteomics using RBE cells co-cultured with or without platelets in advance. Generally, a total of 4120 proteins were identified by mass spectrometer, of which 3122 proteins were further quantitatively assessed (Table S4, Fig. S3A). Principal component analysis (PCA) demonstrated that the proteomic profiling of cell samples with platelet stimulation could be clearly distinguished from non-platelet-stimulated cell samples (Fig. S3B). Pearson correlation analysis discovered a strong association among the three biological replicates within each subgroup (Fig. S3C). And the characteristics of the protein mass distribution as well as peptide-protein matching map were also displayed in Fig. S3D. Among the 3122 quantifiable proteins identified by proteomic profiling, 67 and 40 proteins were markedly up- and down-regulated in platelet-stimulated RBE cell samples, respectively (p < 0.01 and|log2 (fold change)| > 1.5, Fig. 3A and Table S5), and the protein-protein interaction (PPI) network involving these dysregulated targets was also described (Fig. 3B). Subsequently, we performed functional analysis based on these differentially expressed proteins (DEPs). The results of the GO enrichment analysis revealed that DEPs were predominantly enriched in the cellular component (CC) term focal adhesion, the molecular function (MF) term TAP binding, and the biological process (BP) term morphogenesis of an epithelium, respectively (Fig. 3C and Table S6). Analogously, separate GO enrichment analyses specifically targeting up- or down-regulated DEPs were also performed, and the results indicated a potential close relationship between platelet stimulation and cell adhesion, given that the overall DEPs or up-regulated DEPs were found to be obviously enriched in the CC term focal adhesion and the BP term positive regulation of cell adhesion (Fig. S3E). Furthermore, the KEGG enrichment analysis disclosed that the focal adhesion was also significantly over-represented in both the entire and up-regulated DEPs in platelet-stimulated cell samples, whereas the down-regulated proteins were found to be relatively enriched in terms of non-alcoholic fatty liver disease and pathways of neurodegeneration-multiple diseases (Fig. 3D and Table S7). Moreover, the GSEA analysis targeted GO-BP terms as well as KEGG pathways further emphasized the consistent observation that the up-regulated DEPs were indeed enriched in cell adhesion of BP term and focal adhesion of KEGG pathway (Fig. S3F).

Given the intimate relationship between tumor progression and focal adhesion signaling [25], as well as the above findings that both the GO, KEGG, and GSEA enrichment analyses highlighted the potential function of activated focal adhesion signaling induced by up-regulated DEPs under platelet stimulation in tumor proliferation and metastasis promotion, indicating that targeting these abnormally expressed DEPs to disturb this signaling pathway might be a promising therapeutic strategy against platelet-mediated iCCA development. Therefore, we prioritized our focus on those overexpressed DEPs stimulated by platelets. By overlapping the up-regulated DEPs enriched in focal adhesion of GO-CC term, positive regulation of cell adhesion of GO-BP term, and focal adhesion of KEGG pathway, three repeated genes ILK, ITGA5, and ITGB3 were obtained, and we further intersected them with the top 10 elevated genes acquired from our proteomic data. Two candidate targets ILK and ITGB3 eventually came into our view (Fig. 3E).

Remarkably, these two focal adhesion-related proteins have previously been reported to be associated with HCC malignant behaviors, including tumor metastasis and chemo-resistance [26, 27]. However, relative studies about these two targets are still insufficient in iCCA, further investigations on their expression patterns, influence on iCCA malignant phenotypes, as well as the relationship with platelet stimulation are what we need to elucidate.

Fig. 3
Fig. 3
Full size image

Proteomic profiling screens the downstream targets of platelet-mediated tumor promotion. (A) Differentially expressed proteins (DEPs) identified by 3D-label free relative quantitative proteomics using RBE cells co-cultured with or without platelets (p < 0.01 and|log2 (fold change)| > 1.5). (B) Protein-protein interaction network among DEPs. (C) GO enrichment analysis of DEPs. Cellular component (left), molecular function (middle), and biological process (right). (D) KEGG enrichment analysis of DEPs. Analysis with entire DEPs (left), up-regulated DEPs (middle), and down-regulated DEPs (right). (E) Intersection of up-regulated DEPs enriched in focal adhesion (GO-cellular component), positive regulation of cell adhesion (GO-biological process), and focal adhesion (KEGG pathway), Chordal (left) and Venn diagram (middle). The merge genes were further intersected with the top 10 elevated genes acquired from proteomic data, Venn diagram (right)

ILK and ITGB3 are overexpressed in iCCA and can be further up-regulated by platelet stimulation

To validate the proteomic observations, we initially evaluated the expression levels of the top 10 genes utilizing RNA-seq data obtained from the TCGA and GEO databases (Table S8 and S9). The results indicated that most of these genes were overexpressed in tumor tissues of iCCA patients (Fig. S4), except for CA2 and ARHGDIB in the TCGA-CHOL dataset and CA2 in the GSE107943 dataset. Subsequently, we performed the WB assay to examine the expression changes of these targets upon platelet stimulation. In iCCA cell lines RBE and HuCCT1, platelet stimulation obviously increased the protein expression levels of the targets except for ARHGDIB, and two candidate proteins ILK and ITGB3 displayed an obvious and consistent upregulation tendency (Fig. 4A, Fig. S2B), suggesting that platelet-mediated iCCA progression might depend on these dysregulated proteins, specifically ILK and ITGB3. Meanwhile, we also evaluated the relative expression of ILK and ITGB3 in 12 paired iCCA tumor tissues. As shown in Fig. 4B, the expression of ILK and ITGB3 was indeed up-regulated in tumor samples compared to para-cancerous tissues. The results of the IHC staining also confirmed that compared with matched adjacent tissues, both ILK and ITGB3 were overexpressed in iCCA tumor samples (Fig. 4C).

The above investigations demonstrated that platelet stimulation certainly increased the protein expression of several downstream targets acquired from proteomic profiling, in which two candidate genes ILK and ITGB3 were further verified to be up-regulated in iCCA tumor tissues and negatively associated with patients’ survival outcomes, suggesting that both ILK and ITGB3 might act as oncogenes in iCCA development. Cause there was a close relationship between these two targets with the focal adhesion signaling (Fig. 4D), whether platelet-induced iCCA growth and metastasis depended on these two proteins, as well as their precise correlation with the focal adhesion pathway, remained to be further explored.

Fig. 4
Fig. 4
Full size image

ILK and ITGB3 are overexpressed and can be up-regulated by platelet stimulation in iCCA. (A) Detection of the expression changes of the top 10 genes acquired from proteomic data under platelet stimulation. (B) Detection of the relative expression of ILK and ITGB3 in 12 pairs of matched iCCA and para-tumoral tissues. (C) Representative images of IHC staining of ILK and ITGB3 in 5 pairs of resected iCCA and para-tumoral tissues. Scale bars, 50 μm, 20 μm. (D) Schematic diagram of the focal adhesion signaling pathway

Both ILK and ITGB3 are responsible for platelet-induced iCCA development

Here, to evaluate the intent pro-tumoral functions of ILK and ITGB3 in iCCA progression, the expression of these two genes was firstly knocked down via targeted siRNA transfection. As depicted in Fig. 5A, both the ILK and ITGB3 knockdown (ILK-kd and ITGB3-kd) RBE and HuCCT1 cell lines were successfully constructed and validated by WB and RT-PCR assays (including si-NC (negative control), si-ILK-1, si-ILK-2, si-ITGB3-1, and si-ITGB3-2). Subsequently, the influence of gene silencing on the motility and viability of iCCA cells was investigated both in vitro and in vivo. For tumor metastasis, the transwell assay illustrated that both ILK-kd and ITGB3-kd significantly restrained the migratory and invasive capabilities of iCCA cells, and similar results were also observed in the wound healing test (Fig. 5B-C). More importantly, several EMT biomarkers also presented altered expression under gene silencing treatment, including the down-regulation of N-Ca, Vim, Snai1, MMP3, and MMP9, indicating that ILK- or ITGB3-silencing inhibited the EMT process of iCCA cells (Fig. 5D). Moreover, to evaluate the impacts on in vivo tumor metastasis, we constructed lung metastatic mouse models by injecting with ILK or ITGB3 stably knockdown HuCCT1 cells through the tail veins. The results discovered that compared to the negative control, inhibition of either ILK or ITGB3 markedly suppressed iCCA metastasis in vivo, as reflected by a noticeable reduction in lung metastatic nodules (Fig. 5E). The above findings indicated that both ILK and ITGB3 are involved in the regulation of tumor invasion and migration both in vitro and in vivo, potentially through their effects on modulating the EMT process in iCCA cells.

Fig. 5
Fig. 5
Full size image

ILK and ITGB3 are involved in platelet-induced iCCA metastasis. A. Measurement of silencing efficiency of both ILK and ITGB3 in iCCA cells transfected with two distinct siRNA sequences, respectively. Si-NC was used as internal control, validated by western blotting and RT-PCR assays. B–C. Evaluation of the influence of ILK- or ITGB3-silencing on invasion and migration capabilities of iCCA cells. Transwell (B), wound healing (C) tests. Scale bars, 100 μm. D. Detection of the expression changes of EMT related proteins in iCCA cells transfected with ILK or ITGB3 targeted siRNAs, si-NC was used as internal control, validated by western blotting assay. E. Effect of ILK- or ITGB3-silencing on in vivo metastatic ability of HuCCT1 cells (n = 6). Data were described as means ± SD, **p < 0.01, ***p < 0.001

With respect to tumor viability, the CCK-8 and colony formation tests demonstrated that ILK- and ITGB3-inhibition effectively suppressed in vitro tumor cell proliferation, and the EdU assay further substantiated this impact that the EdU positive cells were significantly reduced in iCCA cells transfected with ILK or ITGB3 siRNAs (Fig. 6A-C). Analogously, this repressive function was further validated in xenograft tumor models, as subcutaneous tumor formation was markedly suppressed by ILK- or ITGB3-silencing, represented by the obviously decreased tumor weight and volume (Fig. 6D).

Collectively, these findings demonstrated that ILK- or ITGB3-silencing could effectively attenuate the motility and viability of tumor cells both in vitro and in vivo, indicating the critical role of both ILK and ITGB3 in platelet-induced tumor growth and metastasis.

Fig. 6
Fig. 6
Full size image

ILK and ITGB3 are involved in platelet-induced iCCA growth. A–C. Evaluation of the influence of ILK- or ITGB3-silencing on proliferation capabilities of iCCA cells. CCK-8 (A), colony formation (B), and EdU (C) assays. Scale bars, 100 μm. D. Effect of ILK- or ITGB3-silencing on in vivo proliferation ability of HuCCT1 cells (n = 5). Image of xenograft subcutaneous tumors (left), tumor weight (middle), and growth curves (right). Data were described as means ± SD, **p < 0.01, ***p < 0.001

Platelet stimulation activates the FAK signaling pathway to promote iCCA progression

According to previous investigations, we have demonstrated that platelet stimulation indeed promotes iCCA malignant behaviors, which is partially attributed to the up-regulation of ILK and ITGB3. Since both ILK and ITGB3 are implicated in the modulation of the focal adhesion kinase (FAK)-related signaling pathway [28, 29], we postulate that platelet stimulation may activate the FAK signaling by elevating the expression of ILK and ITGB3, thereby facilitating iCCA progression. To verify this hypothesis, we subjected negative control, ILK- or ITGB3-silenced iCCA cells to platelet treatment as previously described, respectively. The activation of the FAK pathway was further assessed using WB assay. As shown in Fig. 7A, silencing of ILK or ITGB3 effectively attenuated the phosphorylation levels of FAK (p-FAK), downstream regulators PI3K (p-PI3K) and AKT (p-AKT), without exerting a notable influence on their total protein expression, suggesting that both ILK and ITGB3 were indeed involved in regulating the activation of FAK/PI3K/AKT signaling pathway. Intriguingly, platelet stimulation significantly augmented the expression of ILK and ITGB3, along with a partially reversion of the inhibition of FAK signaling caused by ILK- or ITGB3-silencing (Fig. 7B). Furthermore, the down-regulation of EMT-related markers mediated by ILK- or ITGB3-silencing also exhibited a consistent upward trend, which was evidently rescued upon platelet stimulation (Fig. 7C). Similarly, in vitro functional investigations further validated that the reduction in invasion and migration abilities of iCCA cells induced by ILK- or ITGB3-inhibition could be partially reversed by exogenous platelet stimulation, and the same was also true for tumor proliferation capacity (Fig. 7D-F).

In summary, the above findings indicated that platelet-treated tumor cells could activate FAK/PI3K/AKT signaling pathway to promote iCCA progression both in vitro and in vivo in an ILK- and ITGB3-dependent manner.

Fig. 7
Fig. 7
Full size image

Platelet stimulation activates FAK/PI3K/Akt signaling pathway to promote iCCA progression. (A) Detection of the regulatory effect of ILK- or ITGB3-silencing on FAK/PI3K/Akt signaling pathway, si-NC was used as internal control, validated by western blotting assay. (B) Detection of the influence of platelet stimulation on the regulation of FAK/PI3K/Akt signaling pathway in iCCA cells transfected with ILK or ITGB3 targeted siRNAs, si-NC was used as internal control, validated by western blotting. (C) Detection of the influence of platelet stimulation on the expression of EMT and matrix metallopeptidase (MMP) related proteins in iCCA cells transfected with ILK or ITGB3 targeted siRNAs, si-NC was used as internal control, validated by western blotting. (D) Performing transwell assay to evaluate the influence of platelet stimulation on invasion and migration capabilities of iCCA cells transfected with ILK or ITGB3 targeted siRNAs. Scale bars, 100 μm. E–F. Evaluation of the influence of platelet stimulation on proliferation capabilities of iCCA cells transfected with ILK or ITGB3 targeted siRNAs, EdU (E), and colony formation (F) assays. Scale bars, 50 μm. Data were described as means ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant

Discussion

The complex disease characteristics and poor prognosis of iCCA present a significant challenge for its treatment. The escalating incidence and disappointing clinical outcomes necessitate the expedited development of early diagnostic biomarkers and intervention targets for iCCA. Notably, recent research on malignancies has expanded beyond cancer cells themselves to focus more on the interplay between distinct cells within the tumor microenvironment and their impacts on tumor development. And platelets play crucial roles as one of the key components in the tumor blood microenvironment. However, there is insufficient investigation into the role of platelets in the progression of iCCA.

Although elevated PLTc commonly occurs in cases of acute blood loss, infections, and inflammatory diseases, it is worth noting that patients with solid tumors may also exhibit increased PLTc, which can be attributed to tumor-derived interleukin-6 (IL-6) stimulating the secretion of thrombopoietin, thereby facilitating enhanced platelet production [30,31,32]. Actually, elevated PLTc has been reported to be associated with an increased risk of multiple human cancers. Vasily et al. defined abnormally elevated PLTc as exceeding 450 × 109/L and conducted a retrospective analysis of the complete blood cell counts of over 90,000 Canadian residents. The findings revealed a significant correlation between elevated PLTc and increased hazard of colon, lung, ovarian, gastric, esophageal, kidney, breast, and prostate cancers. Conversely, individuals with low PLTc were found to have a reduced risk of breast and prostate cancers. Furthermore, the study also indicated that individuals with high PLTc might be potential candidates for occult cancer presence within a time frame ranging from 6 months to 3 years [13]. In addition, the PLTc can also serve as an indicator of cancer treatment response and disease recurrence. For instance, in ovarian cancer, patients with elevated platelet levels at diagnosis and a decrease of less than 25% after initial treatment have been found to have a poorer prognosis, and the platelet levels tend to rise again following tumor recurrence [33]. In addition to the observed elevation of peripheral PLTc in most tumors, there is also evidence suggesting aberrant platelet infiltration and activation within the tumor microenvironment, which are also correlated with tumor development and patient prognosis. For example, it has been demonstrated that tumor infiltrating platelets (TIPs) can act as prognostic indicators for colorectal cancer, pancreatic ductal adenocarcinoma (PDAC), and pancreatic neuroendocrine tumors [34,35,36]. The compromised integrity of tumor-associated micro-vessels and the heightened permeability of the blood vessel wall may be attributed as significant factors contributing to extensive platelet infiltration [37]. Conversely, activated platelets can secrete angiogenic regulatory factors, such as vascular endothelial growth factor (VEGF) and stroma-derived factor 1 (SDF-1), which promote the formation of tumor-associated blood vessels [38,39,40,41]. Subsequently, the newly formed highly permeable vessels can further enhance platelet exosmosis. Additionally, direct or indirect interactions between tumor cells and platelets can also induce platelet activation, called tumor-cell-induced platelet activation (TCIPA). For instance, tumor cells can release particles containing platelet P-selectin ligands, such as tissue factor (TF) and P-selectin glycoprotein ligand-1 (PSGL-1), which can interact with platelets to trigger their activation [42]. Furthermore, tumor cells are also capable of releasing high mobility group protein 1 (HMGB1) that binds to Toll-like receptor 4 (TLR4) on platelets, leading to their activation [43]. Research has demonstrated that the evaluation of platelet activation status is highly valuable in both diagnosing and prognostication for liver cancer, patients with HCC exhibit elevated levels of platelet activation, which are correlated with a poorer prognosis [44]. Collectively, the assessment of platelet infiltration and activation also holds potential value in predicting tumors. Nevertheless, the available evidence regarding this matter in relation to iCCA remains inadequate.

In the first section of this study, a cohort of patients diagnosed with iCCA was assembled along with their corresponding clinicopathological information to comprehensively investigate the correlation between platelets and clinical characteristics. The findings revealed a significant association between PLTc and tumor burden in individuals with iCCA. Moreover, platelet-related parameters could also serve as predictive markers for adverse patient outcomes, including the link between high PLTc and worse OS and tumor relapse, as well as a notable increase in platelet infiltration, which also exhibited a strong relationship with patient prognosis.

The poor prognosis of iCCA can be attributed to its aggressive biological behavior, characterized by a high propensity for recurrence and metastasis, which is significantly influenced by the tumor microenvironment, where platelets play a pivotal role. Numerous studies have demonstrated that platelets actively contribute to the promotion of growth and metastasis in various malignancies, including breast, colorectal, oral, and prostate cancers [12, 22, 45, 46]. In the following section, we illuminated that platelet stimulation could also enhance the proliferative and invasive abilities of iCCA cells both in vitro and in vivo. Furthermore, the application of proteomic method unveiled the pivotal downstream proteins ILK and ITGB3, which were markedly up-regulated by platelet stimulation. Subsequent silencing experiments targeting these two proteins in iCCA cells demonstrated their active involvement in modulating cell proliferation, invasion and metastasis both in vitro and in vivo, with a direct association to the regulation of FAK/PI3K/AKT signaling pathway. Finally, we conducted rescue assays by co-culturing with platelets to validate the functionality of ILK and ITGB3. The results further substantiated that platelet stimulation partially rescued the decreased proliferative and invasive capabilities of iCCA cells, as well as alleviated the inhibition of FAK signaling axis caused by ILK- or ITGB3-silencing.

Taken together, the cumulative findings suggested that platelet stimulation might contribute to the malignant progression of iCCA by positively modulating FAK/PI3K/AKT signaling pathway, which was partially dependent on the up-regulation of ILK and ITGB3. Meanwhile, our research also exists certain limitations, such as the underlying mechanism by which platelet stimulation up-regulates ILK and ITGB3 expression, and whether the overexpression of ILK and ITGB3, or the facilitation of FAK signaling in iCCA cells can trigger the activation of platelets to form a reciprocal cross-talk, as well as the effect of combined anti-platelet therapy and gene silencing on iCCA treatment. These aspects will be subjects of future investigation.

Conclusions

The peripheral blood platelet counts, as well as platelet activation and aggregation in the microenvironment, were found to be elevated in patients diagnosed with iCCA. Increased platelet counts and aggregation were associated with an unfavorable prognosis. Co-culturing with platelets promoted the EMT, proliferation, and metastasis of iCCA cells. Mechanistically, platelet stimulation activated FAK/PI3K/AKT signaling pathway by up-regulating the expression of the integrin-related kinases ILK and ITGB3 to facilitate malignant biological behaviors in iCCA. Silencing ILK and ITGB3 effectively inhibited the growth and metastasis of iCCA cells both in vitro and in vivo, suggesting their potential therapeutic significance for iCCA.