中文导读
胰腺导管腺癌以丰富的基质反应为特征。本文介绍胰腺癌类器官模型及其在肿瘤-基质相互作用研究中的应用。
以下为英文全文（English full text）
Abstract
Pancreatic ductal adenocarcinoma (PDAC) is characterized by an exceptionally dense desmoplastic stroma that constitutes up to 90% of tumor mass and contributes to therapeutic resistance. Patient-derived pancreatic cancer organoids (PC-PDOs) have emerged as essential models for studying cancer-stroma interactions and developing stroma-targeted therapies. This article provides detailed protocols for establishing PC-PDOs from surgical and biopsy specimens, strategies for co-culturing with stromal components, and applications in evaluating anti-stromal combination therapies.
1. Introduction
Pancreatic ductal adenocarcinoma (PDAC) has a 5-year survival rate of approximately 12%, making it one of the most lethal solid tumors. The genomic landscape is dominated by KRAS mutations (present in >90% of cases), with frequent mutations in TP53, CDKN2A, and SMAD4. The hallmark of PDAC is its extensive desmoplastic stroma, composed of cancer-associated fibroblasts (CAFs), pancreatic stellate cells (PSCs), immune cells, and extracellular matrix (ECM) components. This stroma creates a physical barrier to drug delivery, promotes tumor cell proliferation through paracrine signaling, and induces chemoresistance [1].
Pancreatic cancer organoids were first established by Boj et al. (2015), who demonstrated that both normal and malignant pancreatic ductal epithelial cells could be expanded as organoids with remarkable fidelity to the original tissue. Normal pancreatic organoids require Wnt signaling, while tumor organoids often grow independently of Wnt, reflecting the common Wnt pathway activation in PDAC [2].
2. Pancreatic Cancer Organoid Establishment Protocol
2.1 Tissue Collection and Processing
PDAC tissue is obtained from surgical resections (Whipple procedure, distal pancreatectomy) or endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA). For surgical specimens, the tissue is immediately placed in ice-cold Advanced DMEM/F12 with 1% P/S and 10 mM HEPES. EUS-FNA samples can be processed directly or collected in cell culture medium for transport.
The tissue is washed three times in ice-cold PBS. Necrotic and fatty tissue is removed with scissors. The remaining tissue is minced into 1 mm³ fragments and transferred to a 15 mL conical tube containing 5 mL digestion medium: Advanced DMEM/F12 with 1 mg/mL collagenase IV, 0.5 mg/mL dispase, and 10 μM Y-27632. The tube is incubated at 37°C on a rocker for 30–60 minutes [3].
After digestion, the tube is allowed to settle for 2 minutes. The supernatant is transferred to a new tube containing 10 mL cold wash medium (Advanced DMEM/F12 with 10% FBS). The remaining tissue fragments are subjected to a second digestion if necessary. The cell suspension is centrifuged at 300 × g for 5 minutes, washed twice, and resuspended in cold Matrigel.
2.2 Matrix Embedding and Initial Culture
The cell pellet is resuspended in cold basement membrane matrix (Matrigel or GBiowit Pancreatic Organoid Matrix) at a density of 10,000–50,000 cells per 50 μL. The suspension is plated as 40–50 μL droplets in pre-warmed 24-well plates and polymerized at 37°C for 20–30 minutes.
The complete pancreatic cancer organoid medium (PC-OGM) contains:
- Advanced DMEM/F12 (base)
- 1× B-27 supplement (without vitamin A)
- 1× N-2 supplement
- 1.25 mM N-acetylcysteine
- 10 mM nicotinamide
- 50 ng/mL human recombinant EGF
- 100 ng/mL human recombinant FGF10
- 10 nM [Leu15]-gastrin
- 1 μM A83-01
- 10 μM SB202190
- 10 μM Y-27632 (first 3 days)
- 100 μg/mL Primocin
Note: Wnt3a and R-spondin 1 are typically omitted for tumor organoids, as most PDACs have autocrine Wnt activation or RSPO3 overexpression. However, these factors should be included for normal pancreatic organoids [4].
2.3 Success Rates and Expansion
Pancreatic cancer organoids achieve establishment success rates of 80–90% from surgical resections and 60–70% from EUS-FNA biopsies. The time to first passage is typically 10–14 days. Organoids are passaged every 7–10 days at a 1:3 to 1:5 split ratio. A critical parameter for PDAC organoids is the initial cell density; higher densities (closer to 50,000 cells per droplet) improve establishment success, particularly for biopsy samples. The morphology of PDAC organoids varies from dense, irregular structures to cystic formations, reflecting the histological diversity of the parental tumors [5].
3. Characterization of PC-PDOs
3.1 Histological and Immunohistochemical Validation
PDAC organoids are validated by comparing H&E sections with the original tumor. Key features include nuclear pleomorphism, prominent nucleoli, and irregular glandular architecture. Important immunohistochemical markers include:
- Cytokeratin 7 (CK7) and CK19: ductal epithelial markers
- CA19-9: pancreatic cancer-associated antigen
- p53: abnormal accumulation indicates mutation
- SMAD4 (DPC4): loss indicates mutation/deletion
- Ki-67: proliferation index
- MUC5AC: mucin production
- CDX2: negative (distinguishes from intestinal metaplasia)
Genomic validation includes targeted sequencing of KRAS, TP53, CDKN2A, and SMAD4, with the expectation that PDOs maintain the driver mutations present in the original tumor. Whole-exome sequencing has confirmed >95% concordance for driver mutations between PDAC tumors and derived organoids [6].
4. Modeling the Tumor Microenvironment: Stroma and Immune Cells
4.1 Cancer-Associated Fibroblast Co-Culture
CAFs are the dominant stromal cell type in PDAC and contribute to desmoplasia, immunosuppression, and chemoresistance. CAFs can be isolated from the same tumor specimen or from separate patient-matched tissue. CAFs are expanded in DMEM with 10% FBS and then co-cultured with PC-PDOs in mixed matrix systems or in transwell co-culture setups.
In co-culture models, CAFs induce increased organoid growth, ECM deposition, and resistance to gemcitabine. Single-cell RNA sequencing of co-culture systems has revealed that CAFs adopt distinct subtypes (myofibroblastic CAFs, inflammatory CAFs, antigen-presenting CAFs) depending on their proximity to tumor cells, recapitulating the spatial heterogeneity observed in human PDAC [7].
4.2 Pancreatic Stellate Cell (PSC) Co-Culture
PSCs are the primary source of collagen deposition in PDAC stroma. When activated, PSCs transition from a quiescent, vitamin A-storing phenotype to a myofibroblast-like state expressing α-SMA and producing collagen. PSCs can be isolated from PDAC specimens or from adjacent normal pancreas and co-cultured with organoids.
Studies have shown that all-trans retinoic acid (ATRA) induces PSC quiescence and reduces surrounding collagen deposition, thereby improving gemcitabine penetration and efficacy in organoid models. A phase I clinical trial demonstrated that ATRA augments gemcitabine effects in advanced PDAC patients, validating the organoid-predicted mechanism [8].
4.3 Macrophage and Immune Cell Co-Culture
Tumor-associated macrophages (TAMs) are abundant in PDAC and promote immunosuppression. Macrophages can be differentiated from peripheral blood monocytes using M-CSF and then polarized to an M2-like phenotype using IL-4 and IL-13. In co-culture with PC-PDOs, M2 macrophages secrete IL-10 and TGF-β, promoting tumor growth and resistance.
Targeting macrophages with liposomal clodronate in co-culture models induces macrophage apoptosis and reduces tumor organoid growth, providing preclinical rationale for macrophage-depletion strategies in PDAC [9].
5. Drug Screening Applications: Targeting Stroma and Tumor
5.1 Standard Chemotherapy Testing
PC-PDOs have been validated for predicting responses to standard PDAC chemotherapeutics. Gemcitabine, nab-paclitaxel, FOLFIRINOX, and capecitabine are routinely tested using 6–10 point dose-response curves. Studies have shown that PC-PDO responses to gemcitabine correlate with clinical outcomes, with resistant organoids showing elevated expression of cytidine deaminase, the enzyme that inactivates gemcitabine [10].
5.2 Anti-Stromal Combination Therapies
The physical and biological barrier created by the desmoplastic stroma limits chemotherapy delivery. Organoid-based co-culture models have been instrumental in evaluating anti-stromal combination strategies:
- ATRA + gemcitabine: ATRA induces PSC quiescence, reducing collagen and improving drug penetration.
- PEGylated hyaluronidase (PEGPH20) + chemotherapy: Degrades hyaluronic acid in the stroma, enhancing drug delivery.
- FAP-targeted agents: Inhibits fibroblast activation protein-expressing CAFs.
- Hedgehog pathway inhibitors: Target the stromal signaling that promotes desmoplasia.
Organoid-on-chip platforms that incorporate patient-derived cancer cells, CAFs, and macrophages in spatially defined microenvironments have demonstrated that stroma-depleting agents significantly increase chemotherapy efficacy in a patient-specific manner [11].
5.3 Targeted Therapy and KRAS Inhibition
The development of KRAS G12C inhibitors (sotorasib, adagrasib) has transformed the treatment of KRAS-mutant lung cancers, and their application in PDAC (which harbors primarily KRAS G12D and G12V mutations) is under active investigation. PC-PDOs with KRAS G12D mutations are being used to evaluate next-generation KRAS inhibitors and downstream pathway inhibitors (MEK, ERK, SHP2). Organoid models have demonstrated that KRAS inhibition is more effective in combination with EGFR or SHP2 inhibitors, providing a rationale for combination clinical trials [12].
6. Advanced Models: Fused Organoids and Organoid-Derived Xenografts
Fused pancreatic cancer organoids (FPCOs) that incorporate patient-derived PDAC cells with hiPSC-derived endothelial and mesenchymal cells represent a next-generation model that more faithfully recapitulates the TME. These FPCOs exhibit enhanced chemoresistance and CAF heterogeneity compared to monoculture organoids, providing a more physiologically relevant platform for drug screening [13].
Organoid-derived xenografts (PDOX) generated by injecting PC-PDOs into immunocompromised mice provide an in vivo validation step. The transplantation success rate from in vitro organoids to in vivo tumors exceeds 75%, and the resulting xenografts maintain the histological and genetic features of the original patient tumor. PDOX models are particularly valuable for evaluating drug pharmacokinetics and systemic toxicity in conjunction with organoid-based in vitro screening [14].
Conclusion
Pancreatic cancer organoids have become essential tools for studying the complex interplay between tumor cells and the desmoplastic stroma. By enabling patient-specific modeling of cancer-stroma interactions, PC-PDOs are accelerating the development of stroma-targeted combination therapies that may finally improve outcomes for this devastating disease.