Qin Huang1
Shucan Wei1Jiahui Yu1Yancen Wu1Hao Lai2
Wene Wei1
Linhai Yan2Chenlin Su1
Wei Shi3,4*
Zijie Su1,4*- 1Department of Experimental Research, Guangxi Key Laboratory of Basic and Translational Research for Colorectal Cancer, Guangxi Medical University Cancer Hospital, Nanning, China
- 2Department of Colorectal Surgery, Guangxi Clinical Research Center for Colorectal Cancer, Guangxi Medical University Cancer Hospital, Nanning, China
- 3Department of Laboratory Animal Science, Laboratory Animal Center, Guangxi Medical University, Nanning, China
- 4Guangxi Engineering Research Center of China-ASEAN Laboratory Animal Science and Innovation, Guangxi Medical University, Nanning, China
Colorectal cancer (CRC) remains a major global health burden. While precision therapies like anti-PD-1 and anti-EGFR antibodies show remarkable efficacy, their application is constrained by stringent biomarker requirements, limiting patient benefit. Diverse animal models—including chemically induced, genetically engineered, and transplantation-based systems—have advanced our understanding of CRC pathogenesis but exhibit limited power in predicting therapeutic outcomes for defined patient subgroups. A central challenge is their imperfect recapitulation of key aspects of human CRC biology, specifically anatomical tumor localization, faithful representation of the tumor immune microenvironment (TME), and a frequent lack of rigorous molecular characterization. This gap underscores the urgent need for advanced models that better mirror human disease to support translational research. This review critically evaluates the establishment, advantages, and limitations of prevalent CRC models, focusing on their capacity to replicate key immunological features of human CRC, such as the complex immune landscape and response to immunotherapies. We examine how discrepancies in anatomical site, immune cell composition, and host immunity between animal models and human patients compromise predictive accuracy, particularly for evaluating immune-checkpoint inhibitors (ICIs) in microsatellite-stable (MSS) tumors. By synthesizing these critiques, we aim to provide a framework for developing immunologically relevant models to accelerate the discovery of effective, personalized immunotherapies for CRC.
1 Introduction
Colorectal cancer (CRC) is a major global public health challenge, ranking as the third most prevalent malignancy and the second leading cause of cancer-related mortality worldwide (1). The emergence of precision oncology has begun to transform cancer treatment, moving beyond a one-size-fits-all approach to embrace therapies tailored to the genetic and molecular profile of an individual’s tumor. This paradigm shift is exemplified by the success of targeted agents in molecularly defined subsets of cancers, such as immune checkpoint inhibitors in mismatch repair-deficient (dMMR) CRC and anti-EGFR antibodies in wild-type RAS tumors. However, the broader application of these therapies in CRC is often hampered by intrinsic and acquired resistance, tumor heterogeneity, and the lack of predictive biomarkers for many emerging agents.
The successful translation of novel therapeutic concepts from bench to bedside is critically dependent on preclinical models that faithfully recapitulate the complexity of human disease. Traditional CRC animal models, including chemically induced (e.g., AOM/DSS), genetically engineered (e.g., APCMin/+mice), and transplantation-based (e.g., CDX, PDX) systems, have provided invaluable insights into carcinogenesis pathways and initial drug efficacy screening (Figure 1). Yet, these models often fall short in mirroring the nuanced tumor microenvironment (TME), the interplay with the immune system, and the specific clinical scenarios of human CRC, particularly in the context of precision oncology. For instance, while the APCMin/+ mouse has been instrumental in elucidating the Wnt pathway, its predominant development of small intestinal tumors limits its translational relevance to human colorectal cancer, which arises in a distinct anatomical and immunological context. This discrepancy highlights a significant gap in our preclinical toolbox—the lack of models that can accurately predict patient-specific responses to increasingly targeted therapies.
Figure 1. Construction strategies for four classic mouse tumor models. This schematic illustrates the key approaches to generating chemically induced, genetically engineered, transplantation-based, and spontaneous tumor models, highlighting their distinct methodologies and applications in cancer research.
Furthermore, the rising incidence of CRC associated with modern lifestyle factors, such as high-fat diets and dysbiosis of the gut microbiota, necessitates the development of models that incorporate these environmental and microbiome influences to study inflammation- and metabolism-driven tumor genesis. The limitations of current models become particularly apparent when attempting to evaluate combination therapies, assess immune-modulating agents, or model the metastatic cascade—all key areas in modern drug development.
This review seeks to critically evaluate the current landscape of laboratory animal models for CRC through the lens of precision medicine. We will first detail the foundational principles, construction methods, and applications of classical model systems. We will then specifically examine emerging models designed to study the role of critical risk factors, including chronic inflammation, dietary components, and specific microbial pathogens. A central focus of our discussion will be a critical appraisal of the advantages and inherent limitations of each model system in addressing the specific challenges of precision oncology, such as modeling tumor heterogeneity, predicting immunotherapy responses, and facilitating the study of rare molecular subtypes. Finally, we will discuss the persistent hurdles in modeling the human CRC immune landscape and disease progression, and explore future directions aimed at refining these essential preclinical tools to better guide therapeutic decisions and improve clinical outcomes.
2 Precision therapy for CRC: current status
Precision therapy is now a cornerstone in the management of CRC management, fundamentally relying on molecular subtyping to define tumor biology and uncover therapeutic targets. Current classification systems are primarily based on three key oncogenic pathways: chromosomal instability (CIN, ~85% prevalence), microsatellite instability (MSI, ~15% prevalence), and CpG island methylator phenotype (CIMP, ~22% prevalence), along with their associated driver mutations (2, 3). This molecular framework underpins individualized treatment strategies.
The molecular subtyping of CRC relies on distinct detection methodologies for each classification:
CIN is assessed through techniques that evaluate large-scale genomic alterations. Karyotyping provides a cytogenetic overview but offers lower resolution (5–10 Mb). For higher resolution screening of copy number variations, single nucleotide polymorphism (SNP) arrays (100–500 kb) are commonly employed (4). Ultimately, whole-genome sequencing (WGS) offers the most precise characterization (<100 kb resolution), enabling the detailed mapping of specific amplification/deletion sites and the determination of their frequencies in tumors suspected of having a high CIN burden (5).
MSI status is primarily determined by analyzing short, repetitive DNA sequences. The standard method involves polymerase chain reaction (PCR) amplification of a classic panel of five mononucleotide and dinucleotide markers (BAT25, BAT26, NR-21, NR-22, and NR-24) (6). Instability in two or more loci defines a tumor as MSI-High (MSI-H). Alternatively, immunohistochemistry (IHC) for the four mismatch repair (MMR) proteins (MLH1, MSH2, MSH6, PMS2) serves as a reliable surrogate; loss of nuclear expression for any of these proteins indicates dMMR, which shows 90-95% concordance with MSI-H status. In cases of discordant results, sequencing-based verification is recommended. For instance, absent MLH1 expression by IHC is often associated with promoter hypermethylation, a finding that warrants confirmation via direct assessment of microsatellite status.
CIMP, defined by a high degree of promoter CpG island methylation, is assessed using various detection methods range from targeted assays, such as methylation-specific PCR (MSP) and quantitative pyrosequencing, to comprehensive genome-wide methylation sequencing. In clinical practice, pyrosequencing is frequently used to quantify methylation levels at specific loci, such as the MLH1promoter. This distinction is critical for guiding therapy, as it helps distinguish sporadic MSI-H tumors (often MLH1-methylated and CIMP-high) from those associated with Lynch syndrome (7).
Several biomarkers critically guide clinical decision-making. RAS mutations (~40% prevalence) determine eligibility for anti-EGFR therapies (e.g., cetuximab), effective only in wild-type cases (8). The BRAF V600E mutation (< 10% prevalence), associated with poor prognosis, typically requires targeted combination regimens (encorafenib combined with cetuximab) (9). MSI-H/dMMR tumors (~15% prevalence) demonstrate exceptional responsiveness to immune checkpoint inhibitors like pembrolizumab, while MSS tumors remain largely resistant (10). For patients with HER2 amplification (2-3% prevalence), trastuzumab-based regimens show efficacy (11). Those with rare NTRK fusions (0.2-2.4% prevalence) benefit from TRK inhibitors such as larotrectinib and entrectinib (12).
Despite guideline recommendations for universal molecular profiling, its implementation remains suboptimal. Testing rates reported as low as 20-30% in some countries and regions, including China, largely due to logistical and educational barriers (13). Beyond limited testing access, a more profound challenge is the marked interpatient heterogeneity in treatment response. Many patients either lack targetable alterations or develop resistance to targeted therapies. Consequently, a substantial proportion cannot benefit from current precision strategies. This dual challenge-limited access to molecular profiling and the inherent limitations of existing targeted agents-highlights a critical gap between the theoretical potential of precision oncology and its real-world clinical application. Thus, robust preclinical CRC models that faithfully recapitulate human disease biology are urgently needed to validate biomarker-driven therapeutic strategies, understanding resistance mechanisms, and developing effective treatments for a broader patient population.
3 Current animal models in CRC research
3.1 Carcinogen-induced models
Carcinogen-induced models (CIMs) are widely used in CRC research to simulate tumor development under controlled conditions (14). These models employ chemical carcinogens administered via oral, intraperitoneal/subcutaneous/intramuscular injection, or rectal routes to induce malignant transformations in the intestinal epithelium (15). CIMs are valuable for elucidating molecular pathways and identifying potential therapeutic targets (16). Commonly used compounds including: heterocyclic amines (e.g., PhIP), aromatic amines (e.g., DMAB), alkylating agents (e.g., MNU, MNNG), and dimethylhydrazine derivatives (e.g., DMH, AOM) (15, 17). These carcinogens are classified as either indirect-acting (requiring metabolic activation, such as DMAB and AOM) or direct-acting (such as MNU and MNNG) agents (Table 1).
Table 1. Comparative characterization of chemically-induced CRC models.
Among indirect-acting agents, PhIP combined with dextran sulfate sodium (DSS) promotes inflammation-driven carcinogenesis, mimicking human sporadic CRC and enabling studies on DNA damage and prevention (18, 19). DMAB requires metabolic activation to generate reactive intermediates that foster tumor formation from ROS-induced inflammation, and it is often enhanced by high-fat diets (15). In contrast, MNU and MNNG mutations via rectal administration, yielding short tumor latency but requiring technical expertise (20) (15, 21). AOM, a procarcinogen activated through hepatic metabolism, induces point mutations (e.g., O6-methylguanine lesions) and, when combined with DSS, accelerates tumorigenesis by synergizing genetic mutations with inflammatory microenvironments (e.g., NF-κB activation), closely mimicking human colitis-associated CRC (22).
These models exhibit genetic and pathological similarities to human CRC but face challenges such as prolonged experimental durations due to stochastic tumor development and variability influenced by rodent sex, age, genetic background, gut microbiota, and immune conditions (16). The CIMs tumors are mainly concentrated in the distal part of the colon and rectum, which differs from the widespread distribution of human colorectal cancer (with a high incidence in the rectum and sigmoid colon); the driving mutations are single, the incidence of MSI-H is low (5%-10%), and the molecular subtype matching is poor; the response prediction of EGFR inhibitors and immune checkpoint inhibitors is disconnected from clinical practice, which is prone to cause treatment failure (23, 24). These differences may be related to the induction method of the model, intestinal physiology, as well as human lifestyle and environmental factors.
CIM, represented by AOM/DSS model, has obvious shortcomings compared with humanized model in simulating TME. Its immune cells are all derived from the mouse host, and it is difficult to reproduce the infiltration pattern of CD8+T cells, the proportion of regulatory T cells and the secretion of IFN-γ and IL-10 in CRC TME (25). The matching degree of immunosuppressive cell components to human is less than 30%, and it is difficult to reproduce the immunosuppressive TME of MSS tumors. Specifically, the composition of T cells, immunosuppressive markers and the degree of infiltration of related cells were significantly different from the TME of human MSS colorectal cancer. At the same time, due to the random effects of carcinogens and the differences in host immune status, the interindividual TME variability is large, and it is difficult to dynamically monitor immune-related indicators. However, the humanized model can reconstruct the human immune cell system in mice and reproduce the complex cellular and cytokine characteristics of human CRC TME. This further highlights the application limitations of the CIM model. Despite these limitations, CIMs provide critical insights into inflammation-cancer transitions and subtype-specific drug screening. For example, AOM/DSS models recapitulate “inflammation-driven” (CMS4-like) and “Wnt-driven” (CMS2-like) subtypes, facilitating the evaluation of targeted therapies like iNOS or KRAS inhibitors (26, 27). In the PhIP/DSS and DMAB models, PD-1 inhibitors, cetuximab, celecoxib, and CD73 inhibitors also exert targeted therapeutic effects (28–30).
Regarding the response mechanism of the AOM/DSS model to PD-1 inhibitors, it is hypothesized to involve inflammation-driven upregulation of PD-L1, potentially via NF-κB pathway activation, which can increase PD-L1 transcription levels by approximately 2.5 to 3-fold (31). Although the MSI status in this model has not been unequivocally validated in certain studies, this mechanism fundamentally differs from the immune response triggered by high tumor mutational burden characteristic of human MSI-H tumors. In contrast, patient-derived xenograft (PDX) models, which retain the molecular subtypes and genetic mutations of the original patient tumors, have demonstrated utility in predicting responses to various targeted therapies. This includes forecasting the efficacy of EGFR-targeted treatments, BRAF inhibitor-based combination regimens, and PARP inhibitors, thereby offering valuable guidance for therapeutic decision-making.
3.2 Genetically engineered models
Genetically engineered mouse (GEM) models represent a sophisticated approach to studying CRC pathogenesis by introducing targeted modifications in key driver genes, thereby enabling the investigation of spontaneous tumorigenesis with high genetic stability (32). These models effectively recapitulate significant genetic alterations associated with both sporadic and inherited forms of CRC, facilitating the exploration of gene-environment interactions (33) and molecular mechanisms underlying disease progression (34).
Among the most pivotal GEM mice are those targeting the APC gene, a key element involving multiple cellular processes (e.g., β-catenin degradation, cytoskeletal organization, cell cycle progression, apoptosis, and cell adhesion) (35), as well as a critical tumor suppressor regulating Wnt/β-catenin signaling (36). The classic APCMin/+ mouse model develops numerous intestinal adenomas within 15 weeks and increases tumor burden when combining the KRAS activation, mimicking human familial adenomatous polyposis (FAP) (37, 38).
Some GEM models (such as APC deletion combined with KRAS mutation) can replicate the tumor immunosuppressive TME of MSS, but the defects are significant: TME is a mouse immune environment, lacks the immune regulatory network of human intestinal cancer flora, and the dendritic cells are not mature enough, which affects the evaluation of immune drug efficacy 70% of human CRC occurs in the colon and rectum, while most GEM tumors are located in the small intestine. The anatomical differences prevent GEM models from reproducing key features of human distal CRC and overestimate the efficacy of combination therapy. Rather than conventional models where spontaneous tumors primarily localize to the small intestine, tissue-specific biallelic knockout models exhibit colonic tumors with improved anatomical relevance, for examples, APCflox/flox mice harboring nonsense mutations at specific codons (Table 2) (e.g., T850, T242, T716, or T1638) crossing with Cre+ mice (Villin-Cre or CDX2-Cre) (39, 40). However, these models face limitations such as predominant small intestinal tumor localization, rare metastasis, and prolonged breeding timelines, as well as oversimplification of human tumor complexity through single-gene modifications.
Table 2. The phenotypic disparity in tumor spectrum between APC mutant mouse models and human CRC.
TP53-mutant models, for examples, APCMin/+p53−/− or APCΔ716+Trp53 R270H, demonstrate enhanced tumor malignancy and accelerated adenoma-to-carcinoma progression, though they require compound genetic alterations and exhibit variable penetrance (41). KRAS-mutant models, frequently combined with APC loss (e.g., APCMin/+ KRASG12D), recapitulate the adenoma-carcinoma sequence and are valuable for studying advanced disease and combination therapies (41, 42). Similarly, GEM models targeting the PI3K/PTEN pathway (e.g., APCMin/+ PTEN+/−) (43) or the TGF-β/SMAD signaling (44), highlight roles in tumor invasion, metastasis, and therapy resistance, but often depend on synergistic mutations for robust phenotypes. In BRAF V600E mutation-driven models (e.g., Vil-Cre; BrafV637E/+; p53LSL-R172H/+ mice), tumor metastasis to local lymph nodes, pancreas or lungs occasionally develops, with an incidence ranging from 12% to 25% (45). These models replicate alternative serrated pathway activation and metastatic potential (46), yet still require multiple genetic alterations.
A significant immunological limitation of GEM models is their murine-derived immune system, which lacks human MHC molecules (such as HLA-A and HLA-B). This prevents the simulation of human tumor antigen presentation. Common tumor antigens in human CRC (such as CEA and MUC1) need to be presented through HLA-A to activate CD8+ T cells, but murine MHC molecules cannot effectively bind human antigen peptides, leading to differences in T cell activation efficiency between the model and humans (47). In addition, the T cell receptor (TCR) repertoire of murine origin is significantly different from that of humans (48, 49). The diversity of the human TCR repertoire is about 1012, while that of murine origin is only 108, resulting in insufficient diversity of immune responses in the model and inability to simulate the complex antigen-specific T cell responses in human CRC.
These GEM models have become indispensable for precision oncology research of CRC. In response to APC mutation, the APCMin/+ model has validated PRAP inhibitor in modulating immunosuppressive pathways and enhancing anti-PD-1 efficacy in colon cancer treatments (50, 51). KRAS-driven models demonstrate the efficacy of MEK inhibitors and their synergy with immunotherapy (52). Compound models like the APC/KRAS/Trp53 triple mutation (AKP) reveal the so-called “synthetic lethality” effect with PARP inhibition and chemotherapy (53, 54). BRAF V600E mutation models support combined targeting of MAPK and immune pathways to overcome resistance (55–57), while in TGF-β receptor inhibitor may suppress tumor metastasis in the SMAD4-deficient high-metastatic model. These models provide more efficient tool for the research of high-risk CRC (58, 59).
The GEM model has significant limitations in the preclinical research of human CRC: over 70% of human CRC occurs in the distal colon (with a high BRAF mutation rate and a more immunosuppressive tumor microenvironment), while the classic GEM model has over 80% of tumors concentrated in the small intestine. There are significant differences in anatomy, microbiota, immunity, and gene expression between the two. Specifically, the immune simulation bias of the GEM model is obvious, with T-cell receptor diversity being only one-tenth of that in humans, MHC homology being lower than 50%, and the lack of the co-evolution process between the human microbiota and the immune system. The molecular subtypes are not fully covered, only including the CMS/4 subtype, while missing key subtypes such as MSI-H; the accuracy of efficacy prediction is poor, with the response rate to EGFR inhibitors being much higher than that in clinical trials, seriously interfering with the evaluation of targeted therapy and immunotherapy (48). Although GEM model cannot precisely replicate the pathogenesis mechanism of distant CRC, it needs to be optimized through combining patient-derived organoids (PDO), humanized mice and other models. However, it also has unique advantages--it can precisely simulate the molecular subtypes of CRC, identify specific mutation treatment targets and verify targeted treatment plans. By leveraging the “gene mutation - targeted drug - effect verification” system, it closely connects experimental research with clinical translation.
3.3 Transplantation models of CRC
Transplantation models, which involve implanting human or rodent tumor tissues or cells into immunodeficient or immunocompetent mice, are widely used due to their rapid tumor formation, high reproducibility, and cost-effectiveness. These models are invaluable for large-scale drug screening and tumor microenvironment studies (60). They can be classified based on implant type (cell, organ, or tumor fragment), transplantation site (orthotopic or ectopic), and host type (allogeneic or xenogeneic) (61, 62).
Common immunodeficient hosts include nude mice, NOD/SCID, NSG, and inbred strains like BALB/c (63, 64). The nine commonly used CRC cell lines—Caco-2, HT29, SW480, SW620, DLD-1, HCT116, LoVo, LS174T, RKO—collectively cover the main molecular subtypes of CRC (Table 3). Cell lines are selected to reflect CRC molecular heterogeneity: MSS lines (e.g., SW480/HT-29) simulate microsatellite-stable subtypes; MSI-H lines (e.g., HCT116/LoVo) model instability subtypes; angiogenesis-associated lines (e.g., LS174T) study vascular mechanisms; and BRAF-mutant lines (e.g., Colo205) target V600E-driven CRC (65, 66). Key approaches include subcutaneous injection for tumor growth monitoring, orthotopic implantation for invasion/metastasis studies, and intraperitoneal injection for modeling peritoneal metastasis, and evaluation of anti-metastatic therapies (67, 68).
Table 3. Molecular and phenotypic characterization of CRC cell lines for precision medicine research.
These models are tailored to specific CRC molecular subtypes in precision medicine. MSS subtypes use cell lines like SW480 to assess the efficacy of anti-angiogenic therapies (e.g., bevacizumab) or combination chemotherapies (69). MSI-H models (e.g., HCT116) assess immune checkpoint inhibitors (e.g., anti-PD-1 antibodies), leveraging their high tumor mutational burden and enhanced immunotherapy sensitivity (70). Angiogenesis-associated subtypes employ lines like LS174T) to screen angiogenesis inhibitors (e.g., sunitinib) (71), while BRAF-mutant models (e.g., Colo205) test BRAF inhibitors (e.g., vemurafenib) and its combination strategies (72).
Humanized mouse models, often utilizing NSG strains, are engineered to constitute a human immune system via CD34+ hematopoietic stem cells or peripheral blood mononuclear cells (PBMCs), enabling co-transplantation with PDX models (73–75). These systems simulate human-specific immune processes, providing a platform for immune-related research, especially for poorly immunogenic subtypes like MSS (76, 77). They replicate clinical responses to immunotherapy (e.g., different PD-1 inhibitor effects in MSI-H against MSS tumors) and support the evaluation of novel therapies like CAR-T cells and bispecific antibodies, aiding immune biomarker identification and immune crosstalk analysis (78–80). However, limitations include: (1) complex immune system reconstruction (2), high costs, (3) prolonged experimental cycles, (4) incomplete reconstruction efficiency, (5) graft-versus-host disease (GVHD) risks in PBMC models (4–5 week experimental window), (6) murine stromal influence, and (7) ethical/sample scalability concerns (81, 82). Autologous humanized models use patient-matched immune and tumor tissues to predict immunotherapy efficacy and test combination therapies (e.g., cytokines with checkpoint inhibitors) for resistant “cold” tumors (Figure 2) (83). Humanized mouse models also have key shortcomings that limit their application in immunotherapy evaluation: 1) Low immune reconstitution efficiency: The chimerism rate of human immune cells after CD34+ hematopoietic stem cell transplantation is usually only 30%-50%, and T cell maturation is insufficient, lacking immune memory function; 2) Abnormal cytokine signaling: Cytokines such as IL-2 and IFN-γ secreted by human immune cells cannot effectively activate murine stromal cells, leading to abnormal immune-stromal interactions; 3) Murine myeloid cell bias: Murine macrophages and dendritic cells still dominate in humanized models (60%-70%), affecting the authenticity of immune response evaluation (84, 85).
Figure 2. Schematic overview of humanized mouse models and their therapeutic applications in cancer immunotherapy. Immunodeficient mice are humanized through three primary approaches: (1) engraftment of human PBMCs to generate the human peripheral blood leukocyte (Hu-PBL) model; (2) transplantation of CD34+ hematopoietic stem cells (from sources such as bone marrow, fetal liver, or umbilical cord blood) to create the human CD34+ (Hu-CD34) model; and (3) combined implantation of fetal liver, thymus fragments (under the renal capsule), and matched hematopoietic stem cells to form the bone-liver-thymus (BLT) model. These humanized systems, when combined with PDXs, enable the study of human tumor-immune interactions and support diverse therapeutic applications, including: (1) testing cell-based immunotherapies (e.g., CAR-T, TCR-T), (2) evaluating immune checkpoint inhibitors (e.g., anti-PD-1, anti-CTLA-4), (3) assessing ADCC, bispecific antibodies, and DARPins, (4) investigating cytokine-based therapies (e.g., IL-15), (5) modeling differential responses to immunotherapy in MSS vs. MSI-H tumors, and (6) predicting patient-specific therapeutic responses.(Created with Figdraw.com). CAR-T, Chimeric Antigen Receptor T cells; TCR-T, T Cell Receptor-Engineered T cells; CTLA-4, Cytotoxic T-Lymphocyte-Associated Protein 4; ADCC, Antibody-Dependent Cellular Cytotoxicity; DARPins, Designed Ankyrin Repeat Proteins; IL-15, Interleukin-15; MSS, Microsatellite Stable; MSI-H, Microsatellite Instability-High.
Transplantation models closely mimic clinical tumor characteristics and molecular heterogeneity of patients, serving as key platforms for precision treatment research. They replicate the natural TME to evaluate anti-tumor and anti-metastatic efficacy of agents like Wnt inhibitors and anti-angiogenic drugs (86). The issue of immune rejection in transplantation models can be circumvented by using severely immunodeficient mice such as NSG (lacking T, B, and NK cells), but the interaction between murine stromal cells and human tumor cells may still influence the response to immunotherapy. For example, CXCL12 secreted by murine fibroblasts can promote PD-L1 expression in tumor cells (upregulated ~1.8-fold) (87).
Cell-line-derived xenograft (CDX) models rapidly screen drugs, confirming the specificity of BRAF inhibitors and revealing mechanisms of new AKT inhibitors like costunolide (CTD) (88, 89). PDX retain molecular subtypes and gene mutations, predicting the efficacy of EGFR-targeted therapy, BRAF inhibitor combination regimens (e.g., dasatinib), and PARP inhibitors to guide treatment decisions (89–91). Although murine mesenchyme retains some tumor-cell characteristics, secreted murine-specific cytokines such as murine IL-6 alter human tumor-cell signaling. Most myeloid cells are derived from mice, forming “mouse myeloid bias”, which is not conducive to simulating human immunosuppressive TME (92). HER2 antibody-drug conjugates exhibit synergistic anti-tumor activity when combined with chemotherapy for advanced CRC (93). PDO-sourced transplantation models combine organoid heterogeneity and in vivo evaluation, enabling individualized chemotherapy and HER2 targeted therapy screening (94). Humanized models address low immune response in MSS CRC treatment, demonstrating that TGF-β inhibitors combined with PD-1 blockade reverses immunosuppressive TME (70, 95–97). They also validate BRAF-targeted therapy and immunotherapy synergism. Ectopic and metastasis models simulate clinical scenarios like liver metastasis, supporting development of anti-metastasis drugs (such as MET inhibitors).
The use of xenograft tumor animal models for human colorectal cancer precision medicine has limitations: There is no clear systematic correlation with CRC molecular typing (CMS, MSI/MSS), making it difficult to meet the needs of precision medicine for patient heterogeneity and molecular characteristics. It can only simulate the basic growth of tumors and cannot stably reproduce the characteristics of each CMS subtype, such as the high immunogenicity of CMS1 type, the interstitial immunosuppressive microenvironment of CMS4 type, and the stable recurrence of MSI-H mutation features. It also fails to reflect the differences in BRAF/KRAS mutations, immune microenvironment, and treatment responses of patients. Moreover, there are currently no clear systematic requirements for addressing the limitations of this model, making it difficult for the research results to directly serve CRC precision diagnosis and treatment.
These models cover drug screening, efficacy prediction and drug resistance mechanism studies, building a translational bridge from clinical tumor characteristics to molecular targets and precise therapy, thus optimizing individualized CRC treatment strategies.
However, PDX models also have unavoidable limitations: 1) The success rate of modeling is greatly affected by the quality of tumor tissue. The tumor tissue of advanced CRC patients has severe fibrosis and high necrosis rate, and the success rate of modeling is only 40-60%, significantly lower than that of early patients; 2) Tumor heterogeneity may be lost during long-term passage, especially the disappearance of rare subclones; 3) Murine stromal cells may replace human stromal cells, affecting the integrity of the tumor microenvironment.
3.4 Spontaneous models
Spontaneous CRC models utilize inbred strain rodents such as aged C57BL/6 mice and WF-Osaka rats—which are genetically predisposed to diseases—to mimic natural pathogenesis of human colorectal cancer without external carcinogens exposure or genetic engineering (98, 99). These models recapitulate key aspects of tumor development and are particularly valuable for studying innate tumorigenesis under controlled genetic backgrounds. For instance, WF-Osaka rats exhibit a tumor incidence rate of 30-40%, making them suitable for basic research on spontaneous tumorigenesis under controlled genetic backgrounds (100–102).
The clinical translation and application value of spontaneous animal models in precision medicine are limited due to several drawbacks: the tumor incidence rate is only 30%-40%, the tumors are randomly distributed in the digestive tract or concentrated in the proximal colon, which does not match the characteristic of FAP in humans where the lesions are concentrated in the colon and rectum. Therefore, it is impossible to verify the efficacy of colon-specific treatment plans; moreover, this model has a large difference from the human genetic background, with an MSI incidence rate lower than 3%, lacking the driver mutation spectrum and subtypes of human CRC, and due to the differences in drug metabolism and immune mechanisms between species, the clinical reference value for predicting treatment responses is limited.
Moreover, their applicability is constrained by low tumor incidence rates (some below 10%), prolonged latency periods (over eight months), unpredictable lesion location and growth patterns, and poor experimental reproducibility. As a result, spontaneous models are unsuitable for high-throughput drug screening or therapeutic evaluation (100, 103), and are primarily limited to basic mechanistic studies that leverage their natural disease progression for pathophysiological insight.
3.5 Non-murine models of CRC
Complementing traditional murine systems, non-murine models—including zebrafish, pigs, dogs, and fruit flies—offer distinct advantages for CRC research. These non-murine models enhance experimental tractability, physiological relevance, and translational value, offering complementary insights into CRC pathogenesis, therapeutic response, and TME dynamics.
3.5.1 Zebrafish models
The zebrafish (Danio rerio) exhibit genomic homology with humans (70-82%) and conserves numerous disease-related genes and epigenetic regulators (104). Key advantages include high fecundity, rapid embryonic development, and optical transparency during larval stages enabling real-time observation, and efficient genetic manipulation through CRISPR-Cas9 and Tol2 transposon systems. These characteristics support shortened experimental timelines, reduced costs, and enhanced suitability for high-throughput drug screening, making zebrafish as a cost-effective complement to rodents (105). In CRC precision medicine, it takes 3–6 months to establish mouse PDX model (6–8 weeks for drug susceptibility testing), and the molecular and microenvironment are single, which are difficult to connect CMS typing, and there is a gap of “rapid personalized testing”. However, Zebrafish PDX (zPDX) only needs about 100 tumor cells, and can establish the model in 1–2 weeks, complete high-throughput drug sensitivity in 7 days, and observe the inhibition of metastasis in real time, so as to fill the gap in the mouse model and meet the needs of rapid personalized treatment for advanced patients.
Two primary model types are used:
1. Transgenic models utilize tissue-specific promoters to simulate human CRC drivers, revealing tumor heterogeneity and angiogenic-immune crosstalk (106).
2. Xenograft models involve injecting fluorescently labeled human CRC cells into immunodeficient larvae, enabling real-time tracking of metastasis and drug response (107).
In precision medicine applications, PDX zebrafish models allow individualized therapy testing. Genetic models (e.g., APC-deficient models for Wnt pathway, KRASG12V/BRAFV600E mutants) facilitate targeted therapy evaluation (108, 109). ERPP-derived polysaccharides and the triazin sulfonamide MM-129 inhibit molecular targets through signaling cascades in the xenograft models, providing novel approach for chemotherapy of CRC (110–112). However, limitations include gene duplication complicating genetic studies, temperature-dependent cell proliferation differences, and incomplete TME immune recapitulation (113, 114). Future improvements require advanced gene editing and humanized microenvironment.
The zebrafish animal model of colorectal cancer has a genomic homology of 70-82% with the human genome. It is quick to establish and has low cost. The zPDX model can complete drug sensitivity testing during the waiting period for patient treatment, which helps in the research of individualized and targeted therapies for CRC. However, zebrafish have a short lifespan and cannot simulate anatomical location and subtype heterogeneity. There is no clear systematic correlation with molecular typing, and it cannot stably reproduce the characteristics and mutation status of relevant typing. Moreover, there are problems such as gene replication and temperature affecting cell proliferation, and there are no clear norms, making it difficult to fully support precise diagnosis and treatment.
3.5.2 Porcine models
Pigs share 98% genomic similarity with human beings and mirror human anatomy, physiology, and body/organ size, enabling clinical procedures such as colonoscopy (115, 116). They serve as high-fidelity translational bridges between rodents and humans (117). The anatomical structure and intestinal flora of the pig model are 85% similar to those of humans and are highly consistent. The distal colon can be adapted to clinical endoscopy and radiotherapy equipment to simulate minimally invasive/surgical scenarios, and fill the gap in the translational research of minimally invasive and radiotherapy combined therapy for CRC in mice. At the same time, its immune system has 90% homology with human, which can reproduce the adverse reactions related to immunotherapy, which cannot be achieved in mice.
Porcine CRC models include:
1. Spontaneous/induced models: Germline APC mutants mimicking familial adenomatous polyposis, DSS-induced colitis-associated cancer models, and germ-free pigs colonized with human microbiota for studying diet-microbiome interactions (118, 119).
2. Genetically engineered models: such as, APC1311/+ line, LGR5-H2BGFP reporters for tracking intestinal stem cells, and the “Oncopig” model with Cre-inducible oncogenic mutations for dynamic progression monitoring (120–122).
Advantages comprise compatibility with clinical equipment, relatively longer lifespan than rodents, human-like immune responses, and biomarker validation utility (123–125). The TME of pig models is highly similar to that of humans in terms of immune cell subset ratio and cytokine profile (such as IL-10 and IFN-γ levels). The number of goblet cells and tight junction structure in the colonic mucosa are consistent with those of human CRC, and can accurately simulate the immunosuppressive TME of MSS tumors (126). Limitations involve high costs, small cohorts, prolonged tumor latency (115, 127, 128). Applications include validate endoscopic ablation techniques, radiotherapy regimens, and chemotherapeutic protocols (116, 120). Future directions focus on miniature breeds (for cost reduction, etc.), optimized gene editing (for distal CRC), and integrated “mouse-screen, pig-validation” pipelines for commercial applications (129).
The structure of the colon, the intestinal microbiota, and humans are highly matched. It can be directly used with clinical equipment, precisely simulating the immunosuppressive microenvironment of MSS tumors. There are multiple model types that can verify the treatment plans, filling the gap in the translational research of mouse models. However, this model has problems such as high cost, small cohort size, long tumor latency period, single molecular subtype, insufficient high-throughput analysis, and there is no clear specification for addressing its limitations, making it difficult to fully support precise CRC diagnosis and treatment.
3.5.3 Canine models
Pet dogs (Canis lupus familiaris) develop spontaneous CRC with similarities to humans in size, physiology, and environmental exposures (130). Incidence is below 1%, but breed predisposition can exist. Tumors often involve Wnt pathway dysregulation and progress from adenomas to adenocarcinomas in distal colon/rectum (131).
These models enable metastatic studies and align with clinical imaging and treatment protocols (132). Their immune resemblance supports the evaluation of immunotherapies and NSAIDs (133, 134). Limitations include low incidence rates, cohort assembly challenges, and inability to induce tumors experimentally (135). Dogs contribute to pharmacokinetic and metastasis researches via clinical trial participation (136). Future efforts require multi-institutional molecular profiling, canine PDX models, and breeding of susceptible lineages (137, 138).
The canine model of colorectal cancer can spontaneously develop CRC, with tumors and physiological conditions similar to those in humans, which can assist in diagnosis and treatment research. However, its incidence is low, the construction of the cohort is difficult, and the matching degree of molecular subtypes is low. Species differences limit its clinical reference value.
3.5.4 Drosophila models
The fruit fly (Drosophila melanogaster) serves as an invertebrate model for CRC research, by leveraging conserved signaling pathways (Wnt, RAS/MAPK, and JNK, etc.) for CRC mechanistic studies (139). Its midgut contains regenerative stem cells (140, 141), and models multi-step carcinogenesis when APC deletion combines with RAS activation and polarity gene deletion (142, 143).
Strengths include powerful genetic tools for large-scale screens, rapid generation of complex genotypes, and feasibility for small-scale drug testing (143–145). Limitations arise from evolutionary distance, missing human CRC features, and incomplete gene conservation. Drosophila aids target identification and preliminary drug assessment in precision oncology.
The fruit fly model can be used to study the mechanisms of CRC, simulate carcinogenesis, and assist in the research of drugs and targets. However, it can only simulate the basic carcinogenic pathways and cannot reproduce heterogeneity and the immune microenvironment. It has a weak correlation with CRC molecular typing and shows significant differences from humans, making it difficult to support precise diagnosis and treatment of CRC.
3.6 Modeling CRC risk factors
Accumulating clinical evidence implicates chronic intestinal inflammation, high-fat diet (HFD), and pathogenic gut microbiota as of CRC risk factors (146). These are recapitulated in laboratory animal models, where they may accelerate tumorigenesis in genetically predisposed backgrounds (147, 148).
3.6.1 Inflammation-associated CRC models
Chronic inflammation elevates CRC risk, particularly in inflammatory bowel diseases (IBD) such as ulcerative colitis and Crohn’s disease (149). The DSS model induces ulcerative colitis like injury, mostly characterized by bloody stools, ulceration, and granulocyte infiltration, closely mimicking human ulcerative colitis (150, 151). When combined with AOM, it shortens tumor development from 10–20 weeks to 6–10 weeks. This AOM/DSS model reproduces distal colon tumors with progressive pathology (adenomas to adenocarcinomas), enabling study of dietary, pharmaceutical, and microbial influences on inflammation-driven cancer (152). However, the AOM/DSS model has significant limitations that seriously affect its clinical translation value: 1) Low genomic fidelity: Its tumor mutation spectrum is dominated by inflammation-related mutations (such as TP53 R248W), lacking common hot-spot mutations of APC and KRAS in human CRC (APCT876, KRASG12D), and the mutation coincidence rate is only 40%-50%; 2) Random tumor occurrence: The number of tumors and size in the same batch of models vary greatly, resulting in an experimental reproducibility variation coefficient of 25%-30%; 3) Weak metastatic ability: Only less than 5% of models will have local lymph node metastasis, unable to simulate the liver and lung metastasis characteristics of human CRC (clinical metastasis rate 30%-40%).
3.6.2 High-fat diet models
HFD with >30% fat promote metabolic and inflammatory changes linked to CRC (153) (154),. Models identify therapeutic targets (e.g., TLR4/CXCL10, STAT3, and GPR65) and validate agents like evodiamine and naringenin that regulate inflammation, EMT, and microbiome balance. HFD models also reveal immunosuppressive microenvironments and test combinational therapies (e.g., anti-PD-L1 antibodies with CSF-1R inhibitors) (155–158). The microbiota and metabolism exert their effects through intestinal microbiota modulation, bile acid metabolism, or tumor energy supply, expanding the scope of non-pharmacological precision interventions (159). Synergy of short-term HFD and PD-L1 antibody, enhanced by the addition of CSF-1R inhibitor and chemotherapy, addresses challenges related to HFD-induced immunosuppressive microenvironments in CRC and chemotherapy resistance (160, 161). These strategies provide a platform for precision intervention spanning target validation to combination strategy optimization.
3.6.3 Gut microbiota in CRC
Gut microbiota has emerged as a critical environmental factor in CRC initiation and progression, influencing tumor development through complex host-microbial interactions. Specific microorganisms—including Fusobacterium nucleatum, enterotoxigenic Bacteroides fragilis, Porphyromonas gingivalis, and others—correlate with CRC recurrence, metastasis, and poor prognosis (162). These interactions can be modeled via fecal microbiota transplantation in genetically engineered or chemically induced CRC models.
F. nucleatum promotes metastasis in APCMin/+ mice through autophagy activation and modulates the tumor microenvironment via CCL20-mediated macrophage recruitment and M2 polarization. Notably, it may enhance anti-PD-1 efficacy in microsatellite-stable CRC by increasing butyric acid and reducing CD8+T cell exhaustion (163, 164). Parvimonas micra, an oral pathobiont, colonizes the colon via the oral-gut axis, upregulates Cyclin D1, inhibits p53 signaling, and activates NF-κB through Th17 cell induction. Porphyromonas gingivali correlates with reduced overall survival in CRC patients and accelerates tumorigenesis in APCMin/+ mice by recruiting tumor-infiltrating myeloid cells, activating the NLRP3 inflammasome, and increasing tumor burden (165).
Probiotics such as Lactobacillus gallinarum and Lactobacillus plantarum enhance immunotherapy response through metabolic modulation of the IDO1/Kyn/AHR axis and CD8+ T cell activation (166). Potential mechanisms include competitive exclusion of pathogens, production of short-chain fatty acids, reinforcement of the intestinal barrier, and improved chemotherapy tolerance, highlighting the microbiome as a biomarker and therapeutic target for CRC treatment.
Inflammatory, hyperlipidemic, and microbial animal models can respectively simulate the occurrence of CRC triggered by a single pathogenic factor, each having its own research value. However, they also have significant limitations: they can only simulate a single pathogenic factor, deviating from human diseases and classifications, and have low genomic fidelity.
4 Challenges of animal models in meeting precision medicine requirements for CRC
Precision medicine for CRC requires models that heterogeneity and treatment response. Current animal models face significant challenges in replicating key clinical, molecular, and immunological features of human CRC, limiting their utility in biomarker discovery and therapeutic development for clinical transformation (Tables 4, 5).
Table 4. Strengths, limitations, and precision medicine applications of major CRC development models.
Table 5. Comparative analysis of multi-species animal models for CRC research: strengths, limitations, and precision oncology applications.
4.1 Anatomical and genetic relevance
Most current transgenic CRC models develop tumors primarily in the small intestine, contrasting with the predominant rectal and colonic localization in humans. The anatomical mismatch between models and human CRC not only affects the quantitative evaluation of treatment effects but also may lead to biomarker misjudgment. For example, the proportion of MSI-H subtypes in small intestine-derived tumor models is about 25%-30%, which is significantly higher than that in human distal CRC (12%-15%). If drugs targeting MSI-H are screened based on this model, the efficacy may be lower than expected during clinical translation. Furthermore, commonly used mutation sites (e.g., APCT850) often differ from clinically prevalent mutations (e.g., APCT876,T1309,T1450) (167), reducing translational relevance. Combining genetic engineering with chemical induction (e.g., AOM/DSS treatment in APCMin/+ mice) or employing large animal models such as pigs may better mimic pathogenesis and improve metastatic modeling, representing a promising future direction.
4.2 Molecular and immunological annotation
Although some models have shown responses to PD-1 inhibitors and anti-EGFR therapies in the studies, the validation standards for MSI and EGFR status in different laboratories are inconsistent - for instance, some studies did not verify the expression of MMR proteins through IHC, but relied solely on PCR to detect the stability of microsatellite sites, resulting in reduced comparability of the results. Therefore, the current molecular characterization of the models faces the challenge of lack of a unified standard, rather than being completely lacking in annotations. For another example, although the AOM/DSS model responds to immune checkpoint inhibitors like anti-PD-1 or anti-EGFR therapy, its MSI or EGFR status often remains unverified, complicating mechanistic interpretation (28, 29). Enhanced genomic, transcriptomic, and proteomic profiling, along with established model databases, would strengthen validity and reproducibility.
4.3 Personalized immunocompetent models
While PDOs and PDXs enable drug sensitivity testing, they typically lack functional immune microenvironment. Humanized models, generated via hematopoietic stem cell transplantation, offer partial solutions but face technical challenges including low engraftment efficiency and extended timelines. Alternative approaches, such as immunocompetent PDXs in humanized pigs, may provide more physiologically relevant platforms for immunotherapy evaluation (Figure 3).
Figure 3. Schematic of PDX models for CRC research. Tumors from CRC patients are engrafted into immunodeficient mice to establish PDX models (spanning F1 to F3 generations). These models were subsequently used for in vivo drug screening. The integration of various patient-derived models and deep sequencing-based data analysis facilitates the correlation of preclinical results with individual patient profiles, supporting the development of personalized therapeutic strategies.
4.4 High-throughput in vivo platforms
Conventional PDX models require prolonged serial passaging, delaying drug testing rather than using PDO models. zPDX system overcome this bottleneck by using minimal cell numbers (approximately 100 cells) and leveraging rapid embryonic development, enabling rapid in vivo drug sensitivity assays. Emerging clinical evidence suggests zPDX-guided therapies may improve disease-free survival in CRC patients, supporting its potential in clinical decision-making and preclinical trials (105, 168–170).
4.5 Expanding therapeutic targeting
Current targeted therapy for CRC address only a narrow subset of biomarkers (e.g., MMR/MSI, RAS, BRAF, and HER2). Broader therapeutic strategies are needed, particularly for commonly altered genes such as APC, which remains undrugged due to its large protein size and mutational complexity (167). Developing models that better represent diverse molecular subtypes can accelerate novel target identification and validation (Figures 4, 5).
Figure 4. Selecting preclinical models for CRC consensus molecular subtypes. This schematic aligns the four CMS of CRC with their recommended preclinical models, based on shared biological features. CMS1 (immune), defined by dMMR/MSI status and high immunogenicity, is best modeled using CIMs or humanized mice to recapitulate its immune-active tumor microenvironment. CMS2 (canonical), driven by WNT pathway activation, is most accurately simulated with GEM models. CMS3 (metabolic), associated with KRAS mutations and metabolic reprogramming, can be effectively studied using GEMs or PDOs to maintain its distinct metabolic phenotype. CMS4 (mesenchymal), characterized by an EMT-like signature and immunosuppressive stroma, is suitably modeled in PDX systems to reconstruct its complex tumor-stromal-immune interactions. This framework provides a rationale for selecting context-specific models in precision oncology research. (Created with Figdraw.com).
Figure 5. Biomarker-guided precision treatment matching and corresponding preclinical models for CRC. This schematic presents a decision framework for personalized CRC management, aligning patient-specific biomarkers with optimal therapeutic strategies and preclinical validation platforms. The diagram links key molecular determinants—including RAS/BRAF wild-type status, BRAFV600E mutation, MSI-H/dMMR, HER2 amplification, and NTRK fusions—to their respective precision therapies (e.g., anti-EGFR antibodies, targeted combinations, immune checkpoint inhibitors). Corresponding preclinical models, such as PDX, GEMs, humanized mice, and risk-factor-driven systems, are recommended for evaluating each treatment modality, thereby establishing a direct translational bridge from molecular profiling to clinical strategy. (Created with Figdraw.com).
These challenges highlight the need for more clinically aligned, molecularly annotated, and immunologically functional animal models to advance CRC precision medicine.
4.6 Limited biomarker-guided model selection
Currently, only a limited number of animal models are well-clarified for studying specific molecular subtypes of CRC. For investigating immunotherapy resistance in MSS tumors, PDX models are preferred due to their ability to recapitulate key immunosuppressive features such as Treg enrichment and elevated PD-L1 expression, making them suitable for probing resistance mechanisms mediated by pathways like TGF-β and IL-6. In cases of MSS tumors with concurrent BRAFV600E mutation, CDX or BRAF-mutant PDX models are recommended for evaluating combination therapies targeting BRAF and EGFR. For the 2–3% of CRC cases characterized by HER2 amplification, PDX or CDX models that retain HER2 amplification (e.g., HER2 copy numbers ≥6) are appropriate for assessing agents like trastuzumab and DS-8201; humanized PDX models further enable the evaluation of combining HER2-targeted therapy with immunotherapy, achieving synergistic inhibition rates of 65–70%. For the more common KRAS-mutant subtypes (~40% incidence), GEM models or relevant PDX models can model MAPK pathway activation and are useful for testing combinations of MEK inhibitors with immunotherapy, showing combined inhibition rates of 45–50%. These examples underscore that model availability remains constrained for many CRC molecular subtypes, highlighting a critical gap in precision oncology research.
4.7 Challenges in metastatic fidelity and immune repertoire compatibility
Most preclinical models fail to faithfully recapitulate the metastatic progression of human CRC. A primary limitation is the discrepancy in metastatic patterns: GEM models typically exhibit metastasis rates below 5%, with dissemination primarily to local lymph nodes. This contrasts sharply with the clinical reality in patients, where liver (50%–60%) and lung (20%–30%) are the most common sites of distant metastasis (171). Furthermore, the TME of metastatic lesions in these models often differs from that in humans; for instance, CD8+ T cell infiltration in model metastases tends to be higher than the 15%–20% typically observed in human metastatic sites.
These models are further constrained by a fundamental mismatch in immune repertoire between mice and humans. The diversity of the human TCR repertoire (~1012) vastly exceeds that of mice (~108), limiting the spectrum of simulated immune responses. Additionally, key differences exist in antibody gene rearrangement, particularly in the diversity of the CDR3 region. These inherent immunological disparities compromise the ability of current models to accurately predict human-specific responses to immunotherapy and to study associated resistance mechanisms (Table 6).
Table 6. Comparative overview of preclinical models for immuno-oncology research in CRC.
5 Conclusion and future perspectives
Animal models remain indispensable for advancing CRC precision medicine, yet their clinical relevance is limited by anatomical, molecular, and immunological disparities. Future efforts should focus on developing integrated models that incorporate human-relevant genetic mutations, physiologically accurate microenvironments, and functional immune components. The expanded use of multi-omics profiling and non-murine systems—such as porcine or zebrafish-based platforms—can improve molecular annotation and experimental throughput. By closing these translational gaps, next-generation models will accelerate the discovery of biomarkers and tailored therapies, thereby enhancing personalized treatment strategies for CRC patients, ultimately.
Author contributions
QH: Writing – original draft, Writing – review & editing, Conceptualization, Data curation, Investigation, Software. SW: Investigation, Writing – original draft, Writing – review & editing. JY: Data curation, Writing – original draft, Writing – review & editing. YW: Writing – original draft, Writing – review & editing, Data curation. HL: Writing – original draft, Writing – review & editing, Data curation. WW: Writing – original draft, Writing – review & editing. LY: Writing – original draft, Data curation. CS: Writing – original draft, Data curation. WS: Supervision, Writing – review & editing. ZS: Conceptualization, Project administration, Supervision, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was partially supported by Natural Science Foundation of Guangxi (2025GXNSFDA069008 and 2022GXNSFAA035636), National Natural Science Foundation of China (32460163) and Guangxi Science and Technology Department Project (GuikeAB18221086).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Abbreviations
ADCC, Antibody-Dependent Cellular Cytotoxicity; AOM, Azoxomethanes; CAR-T, Chimeric Antigen Receptor T cells; CDX, Cell line-derived xenografts; CMS, Consensus Molecular Subtype; CRC, colorectal cancer; CTLA-4, Cytotoxic T-Lymphocyte-Associated Protein 4; DARPins, Designed Ankyrin Repeat Proteins; DMBA, 7, 12-dimethylbenz[alpha]anthracene; DMH, 1, 2-dimethylhydrazine; dMMR, Deficient mismatch repair; DSS, Dextran sulfate sodium; FAP, Familial adenomatous polyposis; Fn, Fusobacterium nucleatum; GVHD, Graft-versus-host disease; IL-15, Interleukin-15; MSS, Microsatellite stable; MSI-H, Microsatellite instability-high; MNU, N-methyl-N-nitrosourea; PDO, Patient-derived organoid; PDX, Patient-derived tumor xenografts; PhIP, 2-amino-1-methyl-6-phenylimidazopyridine; TCR-T, T Cell Receptor-Engineered T cells; TME, Tumor microenvironment.
References
1. Chen G, Wu J, Huang S, Gong Z, and Wang H. Global, regional, and national trends in colorectal cancer from 2010 to 2021: an analysis of the global burden of disease study 2021. Ann Med. (2025) 57:2534098. doi: 10.1080/07853890.2025.2534098
2. Ciardiello F, Ciardiello D, Martini G, Napolitano S, Tabernero J, and Cervantes A. Clinical management of metastatic colorectal cancer in the era of precision medicine. CA Cancer J Clin. (2022) 72:372–401. doi: 10.3322/caac.21728
3. Mármol I, Sánchez-de-Diego C, Pradilla Dieste A, Cerrada E, and Rodriguez Yoldi MJ. Colorectal carcinoma: A general overview and future perspectives in colorectal cancer. Int J Mol Sci. (2017) 18:197. doi: 10.3390/ijms18010197
4. Watkins TBK, Lim EL, Petkovic M, Elizalde S, Birkbak NJ, Wilson GA, et al. Pervasive chromosomal instability and karyotype order in tumour evolution. Nature. (2020) 587:126–32. doi: 10.1038/s41586-020-2698-6
5. Wang R, Li J, Zhou X, Mao Y, Wang W, Gao S, et al. Single-cell genomic and transcriptomic landscapes of primary and metastatic colorectal cancer tumors. Genome Med. (2022) 14:93. doi: 10.1186/s13073-022-01093-z
6. Shimozaki K, Hayashi H, Tanishima S, Horie S, Chida A, Tsugaru K, et al. Concordance analysis of microsatellite instability status between polymerase chain reaction based testing and next generation sequencing for solid tumors. Sci Rep. (2021) 11:20003. doi: 10.1038/s41598-021-99364-z
7. Cortes U, Guilloteau K, Rouvreau M, Archaimbault C, Villalva C, and Karayan-Tapon L. Development of pyrosequencing methods for the rapid detection of ras mutations in clinical samples. Exp Mol Pathol. (2015) 99:207–11. doi: 10.1016/j.yexmp.2015.07.003
8. Jensen BV, Schou JV, Yilmaz M, Johannesen HH, Skougaard K, Linnemann D, et al. Cetuximab plus irinotecan administered biweekly with reduced infusion time to heavily pretreated patients with metastatic colorectal cancer and related ras and braf mutation status. Int J Cancer. (2021) 148:2542–56. doi: 10.1002/ijc.33448
9. Tabernero J, Grothey A, Van Cutsem E, Yaeger R, Wasan H, Yoshino T, et al. Encorafenib plus cetuximab as a new standard of care for previously treated braf V600e-mutant metastatic colorectal cancer: updated survival results and subgroup analyses from the beacon study. J Clin Oncol. (2021) 39:273–84. doi: 10.1200/jco.20.02088
10. Le DT, Kim TW, Van Cutsem E, Geva R, Jäger D, Hara H, et al. Phase ii open-label study of pembrolizumab in treatment-refractory, microsatellite instability-high/mismatch repair-deficient metastatic colorectal cancer: keynote-164. J Clin Oncol. (2020) 38:11–9. doi: 10.1200/jco.19.02107
11. Ross JS, Fakih M, Ali SM, Elvin JA, Schrock AB, Suh J, et al. Targeting her2 in colorectal cancer: the landscape of amplification and short variant mutations in erbb2 and erbb3. Cancer. (2018) 124:1358–73. doi: 10.1002/cncr.31125
12. Pietrantonio F, Di Nicolantonio F, Schrock AB, Lee J, Tejpar S, Sartore-Bianchi A, et al. Alk, ros1, and ntrk rearrangements in metastatic colorectal cancer. J Natl Cancer Inst. (2017) 109:djx089. doi: 10.1093/jnci/djx089
13. Li H, Guo L, Wang C, Hu X, and Xu Y. Improving the value of molecular testing: current status and opportunities in colorectal cancer precision medicine. Cancer Biol Med. (2023) 21:21–8. doi: 10.20892/j.issn.2095-3941.2023.0293
14. Neto Í, Rocha J, Gaspar MM, and Reis CP. Experimental murine models for colorectal cancer research. Cancers (Basel). (2023) 15:2570. doi: 10.3390/cancers15092570
15. Rosenberg DW, Giardina C, and Tanaka T. Mouse models for the study of colon carcinogenesis. Carcinogenesis. (2009) 30:183–96. doi: 10.1093/carcin/bgn267
16. Yuan C, Zhao X, Wangmo D, Alshareef D, Gates TJ, and Subramanian S. Tumor models to assess immune response and tumor-microbiome interactions in colorectal cancer. Pharmacol Ther. (2022) 231:107981. doi: 10.1016/j.pharmthera.2021.107981
17. Tanaka T. Colorectal carcinogenesis: review of human and experimental animal studies. J Carcinog. (2009) 8:5. doi: 10.4103/1477-3163.49014
18. Cheung C, Ma X, Krausz KW, Kimura S, Feigenbaum L, Dalton TP, et al. Differential metabolism of 2-amino-1-methyl-6-phenylimidazo[4,5-B]Pyridine (Phip) in mice humanized for cyp1a1 and cyp1a2. Chem Res Toxicol. (2005) 18:1471–8. doi: 10.1021/tx050136g
19. Demeyer D, Mertens B, De Smet S, and Ulens M. Mechanisms linking colorectal cancer to the consumption of (Processed) red meat: A review. Crit Rev Food Sci Nutr. (2016) 56:2747–66. doi: 10.1080/10408398.2013.873886
20. Narisawa T, Sato T, Hayakawa M, Sakuma A, and Nakano H. Carcinoma of the colon and rectum of rats by rectal infusion of N-methyl-N’-nitro-N-nitrosoguanidine. Gan. (1971) 62:231–4.
21. MaChado VF, Feitosa MR, da Rocha JJR, and Féres O. A review of experimental models in colorectal carcinogenesis. J Coloproctology. (2016) 36:053–7. doi: 10.1016/j.jcol.2015.09.001
22. Neufert C, Becker C, and Neurath MF. An inducible mouse model of colon carcinogenesis for the analysis of sporadic and inflammation-driven tumor progression. Nat Protoc. (2007) 2:1998–2004. doi: 10.1038/nprot.2007.279
23. Bürtin F, Mullins CS, and Linnebacher M. Mouse models of colorectal cancer: past, present and future perspectives. World J Gastroenterol. (2020) 26:1394–426. doi: 10.3748/wjg.v26.i13.1394
24. Srivatsa S, Paul MC, Cardone C, Holcmann M, Amberg N, Pathria P, et al. Egfr in tumor-associated myeloid cells promotes development of colorectal cancer in mice and associates with outcomes of patients. Gastroenterology. (2017) 153:178–90.e10. doi: 10.1053/j.gastro.2017.03.053
25. Sun Y, Wang Q, Jiang Y, He J, Jia D, Luo M, et al. Lactobacillus intestinalis facilitates tumor-derived ccl5 to recruit dendritic cell and suppress colorectal tumorigenesis. Gut Microbes. (2025) 17:2449111. doi: 10.1080/19490976.2024.2449111
26. Stastna M, Janeckova L, Hrckulak D, Kriz V, and Korinek V. Human colorectal cancer from the perspective of mouse models. Genes (Basel). (2019) 10:788. doi: 10.3390/genes10100788
27. Corpet DE and Pierre F. Point: from animal models to prevention of colon cancer. Systematic review of chemoprevention in min mice and choice of the model system. Cancer Epidemiol Biomarkers Prev. (2003) 12:391–400.
28. Zhao Y, Liu X, Huo M, Wang Y, Li Y, Xu N, et al. Cetuximab enhances the anti-tumor function of macrophages in an il-6 dependent manner. Life Sci. (2021) 267:118953. doi: 10.1016/j.lfs.2020.118953
29. Kim M, Min YK, Jang J, Park H, Lee S, and Lee CH. Single-cell rna sequencing reveals distinct cellular factors for response to immunotherapy targeting cd73 and pd-1 in colorectal cancer. J Immunother Cancer. (2021) 9:e002503. doi: 10.1136/jitc-2021-002503
30. Ravoori S, Feng Y, Neale JR, Jeyabalan J, Srinivasan C, Hein DW, et al. Dose-dependent reduction of 3,2’-dimethyl-4-aminobiphenyl-derived DNA adducts in colon and liver of rats administered celecoxib. Mutat Res. (2008) 638:103–9. doi: 10.1016/j.mrfmmm.2007.09.003
31. Jia M, Yuan Z, Yu H, Feng S, Tan X, Long Z, et al. Rapamycin circumvents anti pd-1 therapy resistance in colorectal cancer by reducing pd-L1 expression and optimizing the tumor microenvironment. BioMed Pharmacother. (2024) 176:116883. doi: 10.1016/j.biopha.2024.116883
32. Jackstadt R and Sansom OJ. Mouse models of intestinal cancer. J Pathol. (2016) 238:141–51. doi: 10.1002/path.4645
33. Johnson RL and Fleet JC. Animal models of colorectal cancer. Cancer Metastasis Rev. (2013) 32:39–61. doi: 10.1007/s10555-012-9404-6
34. Wang T, Chen Z, Zhang Y, Liu M, Sui H, and Tang Q. Recent advances in the development and application of colorectal cancer mouse models. Front Pharmacol. (2025) 16:1553637. doi: 10.3389/fphar.2025.1553637
35. Hankey W, Frankel WL, and Groden J. Functions of the apc tumor suppressor protein dependent and independent of canonical wnt signaling: implications for therapeutic targeting. Cancer Metastasis Rev. (2018) 37:159–72. doi: 10.1007/s10555-017-9725-6
36. Sharma A, Zalejski J, Bendre SV, Kavrokova S, Hasdemir HS, Ozgulbas DG, et al. Cholesterol-targeting wnt-Β-catenin signaling inhibitors for colorectal cancer. Nat Chem Biol. (2025) 21:1376–86. doi: 10.1038/s41589-025-01870-y
37. Luongo C, Moser AR, Gledhill S, and Dove WF. Loss of apc+ in intestinal adenomas from min mice. Cancer Res. (1994) 54:5947–52.
38. Moser AR, Pitot HC, and Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science. (1990) 247:322–4. doi: 10.1126/science.2296722
39. Hinoi T, Akyol A, Theisen BK, Ferguson DO, Greenson JK, Williams BO, et al. Mouse model of colonic adenoma-carcinoma progression based on somatic apc inactivation. Cancer Res. (2007) 67:9721–30. doi: 10.1158/0008-5472.Can-07-2735
40. el Marjou F, Janssen KP, Chang BH, Li M, Hindie V, Chan L, et al. Tissue-specific and inducible cre-mediated recombination in the gut epithelium. Genesis. (2004) 39:186–93. doi: 10.1002/gene.20042
41. Sakai E, Nakayama M, Oshima H, Kouyama Y, Niida A, Fujii S, et al. Combined mutation of apc, kras, and tgfbr2 effectively drives metastasis of intestinal cancer. Cancer Res. (2018) 78:1334–46. doi: 10.1158/0008-5472.Can-17-3303
42. Hadac JN, Leystra AA, Paul Olson TJ, Maher ME, Payne SN, Yueh AE, et al. Colon tumors with the simultaneous induction of driver mutations in apc, kras, and pik3ca still progress through the adenoma-to-carcinoma sequence. Cancer Prev Res (Phila). (2015) 8:952–61. doi: 10.1158/1940-6207.Capr-15-0003
43. Shao J, Washington MK, Saxena R, and Sheng H. Heterozygous disruption of the pten promotes intestinal neoplasia in apcmin/+ Mouse: roles of osteopontin. Carcinogenesis. (2007) 28:2476–83. doi: 10.1093/carcin/bgm186
44. Fleming NI, Jorissen RN, Mouradov D, Christie M, Sakthianandeswaren A, Palmieri M, et al. Smad2, smad3 and smad4 mutations in colorectal cancer. Cancer Res. (2013) 73:725–35. doi: 10.1158/0008-5472.Can-12-2706
45. Rad R, Cadiñanos J, Rad L, Varela I, Strong A, Kriegl L, et al. A genetic progression model of braf(V600e)-induced intestinal tumorigenesis reveals targets for therapeutic intervention. Cancer Cell. (2013) 24:15–29. doi: 10.1016/j.ccr.2013.05.014
46. Kane AM, Liu C, Fennell LJ, McKeone DM, Bond CE, Pollock PM, et al. Aspirin reduces the incidence of metastasis in a pre-clinical study of braf mutant serrated colorectal neoplasia. Br J Cancer. (2021) 124:1820–7. doi: 10.1038/s41416-021-01339-4
47. Nair R, Lannagan TRM, Jackstadt R, Andrusaite A, Cole J, Boyne C, et al. Co-inhibition of tgf-Β and pd-L1 pathways in a metastatic colorectal cancer mouse model triggers interferon responses, innate cells and T cells, alongside metabolic changes and tumor resistance. Oncoimmunology. (2024) 13:2330194. doi: 10.1080/2162402x.2024.2330194
48. Patel SA, Nilsson MB, Yang Y, Le X, Tran HT, Elamin YY, et al. Il6 mediates suppression of T- and nk-cell function in emt-associated tki-resistant egfr-mutant nsclc. Clin Cancer Res. (2023) 29:1292–304. doi: 10.1158/1078-0432.Ccr-22-3379
49. Gutierrez WR, Scherer A, McGivney GR, Brockman QR, Knepper-Adrian V, Laverty EA, et al. Divergent immune landscapes of primary and syngeneic kras-driven mouse tumor models. Sci Rep. (2021) 11:1098. doi: 10.1038/s41598-020-80216-1
50. Ghonim MA, Ibba SV, Tarhuni AF, Errami Y, Luu HH, Dean MJ, et al. Targeting parp-1 with metronomic therapy modulates mdsc suppressive function and enhances anti-pd-1 immunotherapy in colon cancer. J Immunother Cancer. (2021) 9:e001643. doi: 10.1136/jitc-2020-001643
51. Kaur A, Lim JYS, Sepramaniam S, Patnaik S, Harmston N, Lee MA, et al. Wnt inhibition creates a brca-like state in wnt-addicted cancer. EMBO Mol Med. (2021) 13:e13349. doi: 10.15252/emmm.202013349
52. Zhang X, Lao M, Xu J, Duan Y, Yang H, Li M, et al. Combination cancer immunotherapy targeting tnfr2 and pd-1/pd-L1 signaling reduces immunosuppressive effects in the microenvironment of pancreatic tumors. J Immunother Cancer. (2022) 10:e003982. doi: 10.1136/jitc-2021-003982
53. Tang J, Feng Y, Kuick R, Green M, Green M, Sakamoto N, et al. Trp53 null and R270h mutant alleles have comparable effects in regulating invasion, metastasis, and gene expression in mouse colon tumorigenesis. Lab Invest. (2019) 99:1454–69. doi: 10.1038/s41374-019-0269-y
54. Li J, Lan Z, Liao W, Horner JW, Xu X, Liu J, et al. Histone demethylase kdm5d upregulation drives sex differences in colon cancer. Nature. (2023) 619:632–9. doi: 10.1038/s41586-023-06254-7
55. Sun Z, Shi M, Xia J, Li X, Chen N, Wang H, et al. Hdac and mek inhibition synergistically suppresses hoxc6 and enhances pd-1 blockade efficacy in braf(V600e)-mutant microsatellite stable colorectal cancer. J Immunother Cancer. (2025) 13:e010460. doi: 10.1136/jitc-2024-010460
56. Yoshida Y, Takahashi M, Taniguchi S, Numakura R, Komine K, and Ishioka C. Tretinoin synergistically enhances the antitumor effect of combined braf, mek, and egfr inhibition in braf(V600e) colorectal cancer. Cancer Sci. (2024) 115:3740–54. doi: 10.1111/cas.16280
57. Feng D, Qin B, Pal K, Sun L, Dutta S, Dong H, et al. Braf(V600e)-induced, tumor intrinsic pd-L1 can regulate chemotherapy-induced apoptosis in human colon cancer cells and in tumor xenografts. Oncogene. (2019) 38:6752–66. doi: 10.1038/s41388-019-0919-y
58. Gao Y, Li H, Wang P, Wang J, and Yao X. Sik1 suppresses colorectal cancer metastasis and chemoresistance via the tgf-Β Signaling pathway. J Cancer. (2023) 14:2455–67. doi: 10.7150/jca.83708
59. Ma C, Liu M, Feng W, Rao H, Zhang W, Liu C, et al. Loss of setd2 aggravates colorectal cancer progression caused by smad4 deletion through the ras/erk signalling pathway. Clin Transl Med. (2023) 13:e1475. doi: 10.1002/ctm2.1475
60. Golovko D, Kedrin D, Yilmaz ÖH, and Roper J. Colorectal cancer models for novel drug discovery. Expert Opin Drug Discov. (2015) 10:1217–29. doi: 10.1517/17460441.2015.1079618
61. Georges LMC, De Wever O, Galván JA, Dawson H, Lugli A, Demetter P, et al. Cell line derived xenograft mouse models are a suitable in vivo model for studying tumor budding in colorectal cancer. Front Med (Lausanne). (2019) 6:139. doi: 10.3389/fmed.2019.00139
62. De Angelis ML, Francescangeli F, Nicolazzo C, Xhelili E, La Torre F, Colace L, et al. An orthotopic patient-derived xenograft (Pdx) model allows the analysis of metastasis-associated features in colorectal cancer. Front Oncol. (2022) 12:869485. doi: 10.3389/fonc.2022.869485
63. Uronis JM, Osada T, McCall S, Yang XY, Mantyh C, Morse MA, et al. Histological and molecular evaluation of patient-derived colorectal cancer explants. PLoS One. (2012) 7:e38422. doi: 10.1371/journal.pone.0038422
64. Fidler IJ. Orthotopic implantation of human colon carcinomas into nude mice provides a valuable model for the biology and therapy of metastasis. Cancer Metastasis Rev. (1991) 10:229–43. doi: 10.1007/bf00050794
65. De’ Angelis GL, Bottarelli L, Azzoni C, De’ Angelis N, Leandro G, Di Mario F, et al. Microsatellite instability in colorectal cancer. Acta BioMed. (2018) 89:97–101. doi: 10.23750/abm.v89i9-S.7960
66. Yaeger R and Saltz L. Braf mutations in colorectal cancer: clinical relevance and role in targeted therapy. J Natl Compr Canc Netw. (2012) 10:1456–8. doi: 10.6004/jnccn.2012.0148
67. Xia W, Geng Y, and Hu W. Peritoneal metastasis: A dilemma and challenge in the treatment of metastatic colorectal cancer. Cancers (Basel). (2023) 15:5641. doi: 10.3390/cancers15235641
68. Lund-Andersen C, Torgunrud A, Kanduri C, Dagenborg VJ, Frøysnes IS, Larsen MM, et al. Novel drug resistance mechanisms and drug targets in braf-mutated peritoneal metastasis from colorectal cancer. J Transl Med. (2024) 22:646. doi: 10.1186/s12967-024-05467-2
69. Liu R, Ji Z, Wang X, Zhu L, Xin J, Ma L, et al. Regorafenib plus sintilimab as a salvage treatment for microsatellite stable metastatic colorectal cancer: A single-arm, open-label, phase ii clinical trial. Nat Commun. (2025) 16:1481. doi: 10.1038/s41467-025-56748-3
70. Küçükköse E, Heesters BA, Villaudy J, Verheem A, Cercel M, van Hal S, et al. Modeling resistance of colorectal peritoneal metastases to immune checkpoint blockade in humanized mice. J Immunother Cancer. (2022) 10:e005345. doi: 10.1136/jitc-2022-005345
71. Guo XL, Lin GJ, Zhao H, Gao Y, Qian LP, Xu SR, et al. Inhibitory effects of docetaxel on expression of vegf, bfgf and mmps of ls174t cell. World J Gastroenterol. (2003) 9:1995–8. doi: 10.3748/wjg.v9.i9.1995
72. Yang H, Higgins B, Kolinsky K, Packman K, Bradley WD, Lee RJ, et al. Antitumor activity of braf inhibitor vemurafenib in preclinical models of braf-mutant colorectal cancer. Cancer Res. (2012) 72:779–89. doi: 10.1158/0008-5472.Can-11-2941
73. Hasgur S, Aryee KE, Shultz LD, Greiner DL, and Brehm MA. Generation of immunodeficient mice bearing human immune systems by the engraftment of hematopoietic stem cells. Methods Mol Biol. (2016) 1438:67–78. doi: 10.1007/978-1-4939-3661-8_4
74. Blümich S, Zdimerova H, Münz C, Kipar A, and Pellegrini G. Human cd34(+) hematopoietic stem cell-engrafted nsg mice: morphological and immunophenotypic features. Vet Pathol. (2021) 58:161–80. doi: 10.1177/0300985820948822
75. Lanis JM, Lewis MS, Strassburger H, Larsen K, Bagby SM, Dominguez ATA, et al. Testing cancer immunotherapeutics in a humanized mouse model bearing human tumors. J Vis Exp. (2022) 16:10.3791/64606. doi: 10.3791/64606
76. Suto H, Funakoshi Y, Nagatani Y, Imamura Y, Toyoda M, Kiyota N, et al. Microsatellite instability-high colorectal cancer patient-derived xenograft models for cancer immunity research. J Cancer Res Ther. (2021) 17:1358–69. doi: 10.4103/jcrt.JCRT_1092_20
77. Jin Z and Sinicrope FA. Mismatch repair-deficient colorectal cancer: building on checkpoint blockade. J Clin Oncol. (2022) 40:2735–50. doi: 10.1200/jco.21.02691
78. Teng R, Zhao J, Zhao Y, Gao J, Li H, Zhou S, et al. Chimeric antigen receptor-modified T cells repressed solid tumors and their relapse in an established patient-derived colon carcinoma xenograft model. J Immunother. (2019) 42:33–42. doi: 10.1097/cji.0000000000000251
79. Chuprin J, Buettner H, Seedhom MO, Greiner DL, Keck JG, Ishikawa F, et al. Humanized mouse models for immuno-oncology research. Nat Rev Clin Oncol. (2023) 20:192–206. doi: 10.1038/s41571-022-00721-2
80. Kanikarla Marie P, Sorokin AV, Bitner LA, Aden R, Lam M, Manyam G, et al. Autologous humanized mouse models to study combination and single-agent immunotherapy for colorectal cancer patient-derived xenografts. Front Oncol. (2022) 12:994333. doi: 10.3389/fonc.2022.994333
81. King MA, Covassin L, Brehm MA, Racki W, Pearson T, Leif J, et al. Human peripheral blood leucocyte non-obese diabetic-severe combined immunodeficiency interleukin-2 receptor gamma chain gene mouse model of xenogeneic graft-versus-host-like disease and the role of host major histocompatibility complex. Clin Exp Immunol. (2009) 157:104–18. doi: 10.1111/j.1365-2249.2009.03933.x
82. Matsuda M, Ono R, Iyoda T, Endo T, Iwasaki M, Tomizawa-Murasawa M, et al. Human nk cell development in hil-7 and hil-15 knockin nod/scid/il2rgko mice. Life Sci Alliance. (2019) 2:e201800195. doi: 10.26508/lsa.201800195
83. Ouyang P, Wang L, Wu J, Tian Y, Chen C, Li D, et al. Overcoming cold tumors: A combination strategy of immune checkpoint inhibitors. Front Immunol. (2024) 15:1344272. doi: 10.3389/fimmu.2024.1344272
84. Chen H, Pan Y, Zhou Q, Liang C, Wong CC, Zhou Y, et al. Mettl3 inhibits antitumor immunity by targeting M(6)a-bhlhe41-cxcl1/cxcr2 axis to promote colorectal cancer. Gastroenterology. (2022) 163:891–907. doi: 10.1053/j.gastro.2022.06.024
85. Li S, Liu T, Li C, Zhang Z, Zhang J, and Sun D. Overcoming immunotherapy resistance in colorectal cancer through nano-selenium probiotic complexes and il-32 modulation. Biomaterials. (2025) 320:123233. doi: 10.1016/j.biomaterials.2025.123233
86. Klose J, Eissele J, Volz C, Schmitt S, Ritter A, Ying S, et al. Salinomycin inhibits metastatic colorectal cancer growth and interferes with wnt/Β-catenin signaling in cd133(+) human colorectal cancer cells. BMC Cancer. (2016) 16:896. doi: 10.1186/s12885-016-2879-8
87. Capasso A, Lang J, Pitts TM, Jordan KR, Lieu CH, Davis SL, et al. Characterization of immune responses to anti-pd-1 mono and combination immunotherapy in hematopoietic humanized mice implanted with tumor xenografts. J Immunother Cancer. (2019) 7:37. doi: 10.1186/s40425-019-0518-z
88. Huang H, Park S, Zhang H, Park S, Kwon W, Kim E, et al. Targeting akt with costunolide suppresses the growth of colorectal cancer cells and induces apoptosis in vitro and in vivo. J Exp Clin Cancer Res. (2021) 40:114. doi: 10.1186/s13046-021-01895-w
89. Rubio-Cuesta B, Carretero-Puche C, Llamas P, Sarmentero J, Gil-Calderon B, Lens-Pardo A, et al. Co-targeting src overcomes resistance to braf inhibitors in colorectal cancer. Br J Cancer. (2025) 133:404–19. doi: 10.1038/s41416-025-03058-6
90. Scott AJ, Song EK, Bagby S, Purkey A, McCarter M, Gajdos C, et al. Evaluation of the efficacy of dasatinib, a src/abl inhibitor, in colorectal cancer cell lines and explant mouse model. PLoS One. (2017) 12:e0187173. doi: 10.1371/journal.pone.0187173
91. Bajpai P, Agarwal S, Afaq F, Al Diffalha S, Chandrashekar DS, Kim HG, et al. Combination of dual jak/hdac inhibitor with regorafenib synergistically reduces tumor growth, metastasis, and regorafenib-induced toxicity in colorectal cancer. J Exp Clin Cancer Res. (2024) 43:192. doi: 10.1186/s13046-024-03106-8
92. Iranpour S, Bahrami AR, Nekooei S, Sh Saljooghi A, and Matin MM. Improving anti-cancer drug delivery performance of magnetic mesoporous silica nanocarriers for more efficient colorectal cancer therapy. J Nanobiotechnology. (2021) 19:314. doi: 10.1186/s12951-021-01056-3
93. Liu H, Zhou D, Liu D, Xu X, Zhang K, Hu R, et al. Synergistic antitumor activity between her2 antibody-drug conjugate and chemotherapy for treating advanced colorectal cancer. Cell Death Dis. (2024) 15:187. doi: 10.1038/s41419-024-06572-2
94. Xu J, Gong J, Li M, Kang Y, Ma J, Wang X, et al. Gastric cancer patient-derived organoids model for the therapeutic drug screening. Biochim Biophys Acta Gen Subj. (2024) 1868:130566. doi: 10.1016/j.bbagen.2024.130566
95. Wang X, Fang Y, Liang W, Wong CC, Qin H, Gao Y, et al. Fusobacterium nucleatum facilitates anti-pd-1 therapy in microsatellite stable colorectal cancer. Cancer Cell. (2024) 42:1729–46.e8. doi: 10.1016/j.ccell.2024.08.019
96. Zhang SL, Mao YQ, Zhang ZY, Li ZM, Kong CY, Chen HL, et al. Pectin supplement significantly enhanced the anti-pd-1 efficacy in tumor-bearing mice humanized with gut microbiota from patients with colorectal cancer. Theranostics. (2021) 11:4155–70. doi: 10.7150/thno.54476
97. Liao Y, Yang R, Wang B, Ruan Y, Cui L, Yang J, et al. Mevalonate kinase inhibits anti-tumor immunity by impairing the tumor cell-intrinsic interferon response in microsatellite instability colorectal cancer. Oncogene. (2025) 44:944–57. doi: 10.1038/s41388-024-03255-2
98. Evans JP, Sutton PA, Winiarski BK, Fenwick SW, Malik HZ, Vimalachandran D, et al. From mice to men: murine models of colorectal cancer for use in translational research. Crit Rev Oncol Hematol. (2016) 98:94–105. doi: 10.1016/j.critrevonc.2015.10.009
99. Miyamoto M. Tani Y. A study on colon cancer-prone rats of wf-osaka strain. Med J Osaka Univ. (1989) 38:1–12.
100. Nascimento-Gonçalves E, Mendes BAL, Silva-Reis R, Faustino-Rocha AI, Gama A, and Oliveira PA. Animal models of colorectal cancer: from spontaneous to genetically engineered models and their applications. Vet Sci. (2021) 8:59. doi: 10.3390/vetsci8040059
101. Newmark HL, Yang K, Kurihara N, Fan K, Augenlicht LH, and Lipkin M. Western-style diet-induced colonic tumors and their modulation by calcium and vitamin D in C57bl/6 mice: A preclinical model for human sporadic colon cancer. Carcinogenesis. (2009) 30:88–92. doi: 10.1093/carcin/bgn229
102. Rowlatt C, Franks LM, Sheriff MU, and Chesterman FC. Naturally occurring tumors and other lesions of the digestive tract in untreated C57bl mice. J Natl Cancer Inst. (1969) 43:1353–64.
103. Kobaek-Larsen M, Thorup I, Diederichsen A, Fenger C, and Hoitinga MR. Review of colorectal cancer and its metastases in rodent models: comparative aspects with those in humans. Comp Med. (2000) 50:16–26.
104. Sanders JL, Zhou Y, Moulton HM, Moulton ZX, McLeod R, Dubey JP, et al. The zebrafish, danio rerio, as a model for toxoplasma gondii: an initial description of infection in fish. J Fish Dis. (2015) 38:675–9. doi: 10.1111/jfd.12393
105. Fontana CM and Van Doan H. Zebrafish xenograft as a tool for the study of colorectal cancer: A review. Cell Death Dis. (2024) 15:23. doi: 10.1038/s41419-023-06291-0
106. Yan C, Yang Q, Do D, Brunson DC, and Langenau DM. Adult immune compromised zebrafish for xenograft cell transplantation studies. EBioMedicine. (2019) 47:24–6. doi: 10.1016/j.ebiom.2019.08.016
107. Kobar K, Collett K, Prykhozhij SV, and Berman JN. Zebrafish cancer predisposition models. Front Cell Dev Biol. (2021) 9:660069. doi: 10.3389/fcell.2021.660069
108. Yagdi Efe E, Mazumder A, Lee JY, Gaigneaux A, Radogna F, Nasim MJ, et al. Tubulin-binding anticancer polysulfides induce cell death via mitotic arrest and autophagic interference in colorectal cancer. Cancer Lett. (2017) 410:139–57. doi: 10.1016/j.canlet.2017.09.011
109. Bousquet MS, Ma JJ, Ratnayake R, Havre PA, Yao J, Dang NH, et al. Multidimensional screening platform for simultaneously targeting oncogenic kras and hypoxia-inducible factors pathways in colorectal cancer. ACS Chem Biol. (2016) 11:1322–31. doi: 10.1021/acschembio.5b00860
110. Sun Q, Tao Q, Ming T, Tang S, Zhao H, Liu M, et al. Berberine is a suppressor of hedgehog signaling cascade in colorectal cancer. Phytomedicine. (2023) 114:154792. doi: 10.1016/j.phymed.2023.154792
111. Hermanowicz JM, Szymanowska A, Sieklucka B, Czarnomysy R, Pawlak K, Bielawska A, et al. Exploration of novel heterofused 1,2,4-triazine derivative in colorectal cancer. J Enzyme Inhib Med Chem. (2021) 36:535–48. doi: 10.1080/14756366.2021.1879803
112. Liu M, Wang W, Wang H, Zhao S, Yin D, Zhang H, et al. Antitumor activity of ruditapes philippinarum polysaccharides through mitochondrial apoptosis in cellular and zebrafish models. Mar Drugs. (2025) 23:304. doi: 10.3390/md23080304
113. Lu JW, Ho YJ, Ciou SC, and Gong Z. Innovative disease model: zebrafish as an in vivo platform for intestinal disorder and tumors. Biomedicines. (2017) 5:58. doi: 10.3390/biomedicines5040058
114. Cabezas-Sáinz P, Pensado-López A, Sáinz B Jr., and Sánchez L. Modeling cancer using zebrafish xenografts: drawbacks for mimicking the human microenvironment. Cells. (2020) 9:1978. doi: 10.3390/cells9091978
115. Joshi K, Katam T, Hegde A, Cheng J, Prather RS, Whitworth K, et al. Pigs: large animal preclinical cancer models. World J Oncol. (2024) 15:149–68. doi: 10.14740/wjon1763
116. Schachtschneider KM, Schwind RM, Newson J, Kinachtchouk N, Rizko M, Mendoza-Elias N, et al. The oncopig cancer model: an innovative large animal translational oncology platform. Front Oncol. (2017) 7:190. doi: 10.3389/fonc.2017.00190
117. Gonzalez LM, Moeser AJ, and Blikslager AT. Porcine models of digestive disease: the future of large animal translational research. Transl Res. (2015) 166:12–27. doi: 10.1016/j.trsl.2015.01.004
118. Flisikowski K, Perleberg C, Niu G, Winogrodzki T, Bak A, Liang W, et al. Wild-type apc influences the severity of familial adenomatous polyposis. Cell Mol Gastroenterol Hepatol. (2022) 13:669–71.e3. doi: 10.1016/j.jcmgh.2021.11.002
119. Pistol GC, Bulgaru CV, Marin DE, Oancea AG, and Taranu I. Dietary grape seed meal bioactive compounds alleviate epithelial dysfunctions and attenuates inflammation in colon of dss-treated piglets. Foods. (2021) 10:530. doi: 10.3390/foods10030530
120. Yim JJ, Harmsen S, Flisikowski K, Flisikowska T, Namkoong H, Garland M, et al. A protease-activated, near-infrared fluorescent probe for early endoscopic detection of premalignant gastrointestinal lesions. Proc Natl Acad Sci U.S.A. (2021) 118:e2008072118. doi: 10.1073/pnas.2008072118
121. Schaaf CR, Polkoff KM, Carter A, Stewart AS, Sheahan B, Freund J, et al. A lgr5 reporter pig model closely resembles human intestine for improved study of stem cells in disease. FASEB J. (2023) 37:e22975. doi: 10.1096/fj.202300223R
122. Sikorska A, Flisikowska T, Stachowiak M, Kind A, Schnieke A, Flisikowski K, et al. Elevated expression of P53 in early colon polyps in a pig model of human familial adenomatous polyposis. J Appl Genet. (2018) 59:485–91. doi: 10.1007/s13353-018-0461-6
123. Rogalla S, Flisikowski K, Gorpas D, Mayer AT, Flisikowska T, Mandella MJ, et al. Biodegradable fluorescent nanoparticles for endoscopic detection of colorectal carcinogenesis. Adv Funct Mater. (2019) 29:1904992. doi: 10.1002/adfm.201904992
124. Troya J, Krenzer A, Flisikowski K, Sudarevic B, Banck M, Hann A, et al. New concept for colonoscopy including side optics and artificial intelligence. Gastrointest Endosc. (2022) 95:794–8. doi: 10.1016/j.gie.2021.12.003
125. Schaaf CR and Gonzalez LM. Use of translational, genetically modified porcine models to ultimately improve intestinal disease treatment. Front Vet Sci. (2022) 9:878952. doi: 10.3389/fvets.2022.878952
126. Xiong W, Favier S, Wu T, Ponce F, Dumontet C, Albaret MA, et al. Beyond rodents: alternative animal models in colorectal cancer research. Int J Mol Sci. (2025) 26:10874. doi: 10.3390/ijms262210874
127. Flisikowska T, Merkl C, Landmann M, Eser S, Rezaei N, Cui X, et al. A porcine model of familial adenomatous polyposis. Gastroenterology. (2012) 143:1173–5.e7. doi: 10.1053/j.gastro.2012.07.110
128. Lunney JK, Van Goor A, Walker KE, Hailstock T, Franklin J, and Dai C. Importance of the pig as a human biomedical model. Sci Transl Med. (2021) 13:eabd5758. doi: 10.1126/scitranslmed.abd5758
129. Pan Z, Yao Y, Yin H, Cai Z, Wang Y, Bai L, et al. Pig genome functional annotation enhances the biological interpretation of complex traits and human disease. Nat Commun. (2021) 12:5848. doi: 10.1038/s41467-021-26153-7
130. Yoshizaki K, Hirata A, Nishii N, Kawabe M, Goto M, Mori T, et al. Familial adenomatous polyposis in dogs: hereditary gastrointestinal polyposis in jack russell terriers with germline apc mutations. Carcinogenesis. (2021) 42:70–9. doi: 10.1093/carcin/bgaa045
131. McEntee MF and Brenneman KA. Dysregulation of beta-catenin is common in canine sporadic colorectal tumors. Vet Pathol. (1999) 36:228–36. doi: 10.1354/vp.36-3-228
132. Wang J, Wang T, Sun Y, Feng Y, Kisseberth WC, Henry CJ, et al. Proliferative and invasive colorectal tumors in pet dogs provide unique insights into human colorectal cancer. Cancers (Basel). (2018) 10:330. doi: 10.3390/cancers10090330
133. Pang LY, Argyle SA, Kamida A, Morrison KO, and Argyle DJ. The long-acting cox-2 inhibitor mavacoxib (Trocoxil™) has anti-proliferative and pro-apoptotic effects on canine cancer cell lines and cancer stem cells in vitro. BMC Vet Res. (2014) 10:184. doi: 10.1186/s12917-014-0184-9
134. Chow L, Wheat W, Ramirez D, Impastato R, and Dow S. Direct comparison of canine and human immune responses using transcriptomic and functional analyses. Sci Rep. (2024) 14:2207. doi: 10.1038/s41598-023-50340-9
135. Regan D, Garcia K, and Thamm D. Clinical, pathological, and ethical considerations for the conduct of clinical trials in dogs with naturally occurring cancer: A comparative approach to accelerate translational drug development. Ilar J. (2018) 59:99–110. doi: 10.1093/ilar/ily019
136. Lin Z, Zhang J, Chen Q, Zhang X, Zhang D, Lin J, et al. Transcriptome analysis of the adenoma-carcinoma sequences identifies novel biomarkers associated with development of canine colorectal cancer. Front Vet Sci. (2023) 10:1192525. doi: 10.3389/fvets.2023.1192525
137. Dobson JM. Breed-predispositions to cancer in pedigree dogs. ISRN Vet Sci. (2013) 2013:941275. doi: 10.1155/2013/941275
138. Frazier JP, Beirne E, Ditzler SH, Tretyak I, Casalini JR, Thirstrup DJ, et al. Establishment and characterization of a canine soft tissue sarcoma patient-derived xenograft model. Vet Comp Oncol. (2017) 15:754–63. doi: 10.1111/vco.12215
139. Lucchetta EM and Ohlstein B. The drosophila midgut: A model for stem cell driven tissue regeneration. Wiley Interdiscip Rev Dev Biol. (2012) 1:781–8. doi: 10.1002/wdev.51
140. Zhai Z, Boquete JP, and Lemaitre B. A genetic framework controlling the differentiation of intestinal stem cells during regeneration in drosophila. PLoS Genet. (2017) 13:e1006854. doi: 10.1371/journal.pgen.1006854
141. Pinal N, Calleja M, and Morata G. Pro-apoptotic and pro-proliferation functions of the jnk pathway of drosophila: roles in cell competition, tumorigenesis and regeneration. Open Biol. (2019) 9:180256. doi: 10.1098/rsob.180256
142. Martorell Ò, Merlos-Suárez A, Campbell K, Barriga FM, Christov CP, Miguel-Aliaga I, et al. Conserved mechanisms of tumorigenesis in the drosophila adult midgut. PLoS One. (2014) 9:e88413. doi: 10.1371/journal.pone.0088413
143. Munnik C, Xaba MP, Malindisa ST, Russell BL, and Sooklal SA. Drosophila melanogaster: A platform for anticancer drug discovery and personalized therapies. Front Genet. (2022) 13:949241. doi: 10.3389/fgene.2022.949241
144. Gondal MN, Butt RN, Shah OS, Sultan MU, Mustafa G, Nasir Z, et al. A personalized therapeutics approach using an in silico drosophila patient model reveals optimal chemo- and targeted therapy combinations for colorectal cancer. Front Oncol. (2021) 11:692592. doi: 10.3389/fonc.2021.692592
145. Bangi E, Ang C, Smibert P, Uzilov AV, Teague AG, Antipin Y, et al. A personalized platform identifies trametinib plus zoledronate for a patient with kras-mutant metastatic colorectal cancer. Sci Adv. (2019) 5:eaav6528. doi: 10.1126/sciadv.aav6528
146. Wang Z, Dan W, Zhang N, Fang J, and Yang Y. Colorectal cancer and gut microbiota studies in China. Gut Microbes. (2023) 15:2236364. doi: 10.1080/19490976.2023.2236364
147. Wong SH and Yu J. Gut microbiota in colorectal cancer: mechanisms of action and clinical applications. Nat Rev Gastroenterol Hepatol. (2019) 16:690–704. doi: 10.1038/s41575-019-0209-8
148. Shahgoli VK, Noorolyai S, Ahmadpour Youshanlui M, Saeidi H, Nasiri H, Mansoori B, et al. Inflammatory bowel disease, colitis, and cancer: unmasking the chronic inflammation link. Int J Colorectal Dis. (2024) 39:173. doi: 10.1007/s00384-024-04748-y
149. Shah SC and Itzkowitz SH. Colorectal cancer in inflammatory bowel disease: mechanisms and management. Gastroenterology. (2022) 162:715–30.e3. doi: 10.1053/j.gastro.2021.10.035
150. Wirtz S, Popp V, Kindermann M, Gerlach K, Weigmann B, Fichtner-Feigl S, et al. Chemically induced mouse models of acute and chronic intestinal inflammation. Nat Protoc. (2017) 12:1295–309. doi: 10.1038/nprot.2017.044
151. Azuma YT, Matsuo Y, Kuwamura M, Yancopoulos GD, Valenzuela DM, Murphy AJ, et al. Interleukin-19 protects mice from innate-mediated colonic inflammation. Inflammation Bowel Dis. (2010) 16:1017–28. doi: 10.1002/ibd.21151
152. Tanaka T, Kohno H, Suzuki R, Hata K, Sugie S, Niho N, et al. Dextran sodium sulfate strongly promotes colorectal carcinogenesis in apc(Min/+) mice: inflammatory stimuli by dextran sodium sulfate results in development of multiple colonic neoplasms. Int J Cancer. (2006) 118:25–34. doi: 10.1002/ijc.21282
153. Jin H and Zhang C. High fat high calories diet (Hfd) increase gut susceptibility to carcinogens by altering the gut microbial community. J Cancer. (2020) 11:4091–8. doi: 10.7150/jca.43561
154. Yang J, Wei H, Zhou Y, Szeto CH, Li C, Lin Y, et al. High-fat diet promotes colorectal tumorigenesis through modulating gut microbiota and metabolites. Gastroenterology. (2022) 162:135–49.e2. doi: 10.1053/j.gastro.2021.08.041
155. Chung KS, Heo SW, Lee JH, Han HS, Kim GH, Kim YR, et al. Protective potential of nodakenin in high-fat diet-mediated colitis-associated cancer: inhibition of stat3 activation and wnt/Β-catenin pathway, and gut microbiota modulation. Int Immunopharmacol. (2025) 157:114734. doi: 10.1016/j.intimp.2025.114734
156. Guo H, Zhuang K, Ding N, Hua R, Tang H, Wu Y, et al. High-fat diet induced cyclophilin B enhances stat3/lncrna-pvt1 feedforward loop and promotes growth and metastasis in colorectal cancer. Cell Death Dis. (2022) 13:883. doi: 10.1038/s41419-022-05328-0
157. Sun J, Shi L, Xu F, Sun H, Liu Y, Sun J, et al. Naringenin inhibits colorectal cancer associated with a high-fat diet through modulation of gut microbiota and il-6/stat3 pathway. J Microbiol Biotechnol. (2025) 35:e2412029. doi: 10.4014/jmb.2412.12029
158. Zhu LQ, Zhang L, Zhang J, Chang GL, Liu G, Yu DD, et al. Evodiamine inhibits high-fat diet-induced colitis-associated cancer in mice through regulating the gut microbiota. J Integr Med. (2021) 19:56–65. doi: 10.1016/j.joim.2020.11.001
159. McFarlin BK, Deemer SE, and Bridgeman EA. Oral spore-based probiotic supplementation alters post-prandial expression of mrna associated with gastrointestinal health. Biomedicines. (2024) 12:2386. doi: 10.3390/biomedicines12102386
160. Boufaied N, Chetta P, Hallal T, Cacciatore S, Lalli D, Luthold C, et al. Obesogenic high-fat diet and myc cooperate to promote lactate accumulation and tumor microenvironment remodeling in prostate cancer. Cancer Res. (2024) 84:1834–55. doi: 10.1158/0008-5472.Can-23-0519
161. Li Z, Zhang C, Du JX, Zhao J, Shi MT, Jin MW, et al. Adipocytes promote tumor progression and induce pd-L1 expression via tnf-Α/il-6 signaling. Cancer Cell Int. (2020) 20:179. doi: 10.1186/s12935-020-01269-w
162. Cheng Y, Ling Z, and Li L. The intestinal microbiota and colorectal cancer. Front Immunol. (2020) 11:615056. doi: 10.3389/fimmu.2020.615056
163. Zepeda-Rivera M, Minot SS, Bouzek H, Wu H, Blanco-Míguez A, Manghi P, et al. A distinct fusobacterium nucleatum clade dominates the colorectal cancer niche. Nature. (2024) 628:424–32. doi: 10.1038/s41586-024-07182-w
164. Guo S, Chen J, Chen F, Zeng Q, Liu WL, and Zhang G. Exosomes derived from fusobacterium nucleatum-infected colorectal cancer cells facilitate tumour metastasis by selectively carrying mir-1246/92b-3p/27a-3p and cxcl16. Gut. (2020) 71:e1–3. doi: 10.1136/gutjnl-2020-321187
165. Wang X, Jia Y, Wen L, Mu W, Wu X, Liu T, et al. Porphyromonas gingivalis promotes colorectal carcinoma by activating the hematopoietic nlrp3 inflammasome. Cancer Res. (2021) 81:2745–59. doi: 10.1158/0008-5472.Can-20-3827
166. Huang F, Li S, Chen W, Han Y, Yao Y, Yang L, et al. Postoperative probiotics administration attenuates gastrointestinal complications and gut microbiota dysbiosis caused by chemotherapy in colorectal cancer patients. Nutrients. (2023) 15:356. doi: 10.3390/nu15020356
167. Chatila WK, Kim JK, Walch H, Marco MR, Chen CT, Wu F, et al. Genomic and transcriptomic determinants of response to neoadjuvant therapy in rectal cancer. Nat Med. (2022) 28:1646–55. doi: 10.1038/s41591-022-01930-z
168. Xiao J, Glasgow E, and Agarwal S. Zebrafish xenografts for drug discovery and personalized medicine. Trends Cancer. (2020) 6:569–79. doi: 10.1016/j.trecan.2020.03.012
169. Costa B, Estrada MF, Barroso MT, and Fior R. Zebrafish patient-derived avatars from digestive cancers for anti-cancer therapy screening. Curr Protoc. (2022) 2:e415. doi: 10.1002/cpz1.415
170. Dudziak K, Nowak M, and Sozoniuk M. One host-multiple applications: zebrafish (Danio rerio) as promising model for studying human cancers and pathogenic diseases. Int J Mol Sci. (2022) 23:10255. doi: 10.3390/ijms231810255
171. Tsilimigras DI, Brodt P, Clavien PA, Muschel RJ, D’Angelica MI, Endo I, et al. Liver metastases. Nat Rev Dis Primers. (2021) 7:27. doi: 10.1038/s41572-021-00261-6
172. Ochiai M, Ushigome M, Fujiwara K, Ubagai T, Kawamori T, Sugimura T, et al. Characterization of dysplastic aberrant crypt foci in the rat colon induced by 2-amino-1-methyl-6-phenylimidazo[4,5-B]Pyridine. Am J Pathol. (2003) 163:1607–14. doi: 10.1016/s0002-9440(10)63517-1
173. Bardelli A, Cahill DP, Lederer G, Speicher MR, Kinzler KW, Vogelstein B, et al. Carcinogen-specific induction of genetic instability. Proc Natl Acad Sci U.S.A. (2001) 98:5770–5. doi: 10.1073/pnas.081082898
174. Spjut HJ and Noall MW. Experimental induction of tumors of the large bowel of rats. A review of the experience with 3-2’ Dimethyl-4-aminobiphenyl. Cancer. (1971) 28:29–37. doi: 10.1002/1097-0142(197107)28:1<29::aid-cncr2820280107>3.0.co;2-x
175. Venkatachalam K, Vinayagam R, Arokia Vijaya Anand M, Isa NM, and Ponnaiyan R. Biochemical and molecular aspects of 1,2-dimethylhydrazine (Dmh)-induced colon carcinogenesis: A review. Toxicol Res (Camb). (2020) 9:2–18. doi: 10.1093/toxres/tfaa004
176. Guda K, Upender MB, Belinsky G, Flynn C, Nakanishi M, Marino JN, et al. Carcinogen-induced colon tumors in mice are chromosomally stable and are characterized by low-level microsatellite instability. Oncogene. (2004) 23:3813–21. doi: 10.1038/sj.onc.1207489
177. Crist RC, Roth JJ, Baran AA, McEntee BJ, Siracusa LD, and Buchberg AM. The armadillo repeat domain of apc suppresses intestinal tumorigenesis. Mamm Genome. (2010) 21:450–7. doi: 10.1007/s00335-010-9288-0
178. Zeineldin M and Neufeld KL. Understanding phenotypic variation in rodent models with germline apc mutations. Cancer Res. (2013) 73:2389–99. doi: 10.1158/0008-5472.Can-12-4607
179. Colnot S, Niwa-Kawakita M, Hamard G, Godard C, Le Plenier S, Houbron C, et al. Colorectal cancers in a new mouse model of familial adenomatous polyposis: influence of genetic and environmental modifiers. Lab Invest. (2004) 84:1619–30. doi: 10.1038/labinvest.3700180
180. Shibata H, Toyama K, Shioya H, Ito M, Hirota M, Hasegawa S, et al. Rapid colorectal adenoma formation initiated by conditional targeting of the apc gene. Science. (1997) 278:120–3. doi: 10.1126/science.278.5335.120
181. Kuraguchi M, Wang XP, Bronson RT, Rothenberg R, Ohene-Baah NY, Lund JJ, et al. Adenomatous polyposis coli (Apc) is required for normal development of skin and thymus. PLoS Genet. (2006) 2:e146. doi: 10.1371/journal.pgen.0020146
182. Robanus-Maandag EC, Koelink PJ, Breukel C, Salvatori DC, Jagmohan-Changur SC, Bosch CA, et al. A new conditional apc-mutant mouse model for colorectal cancer. Carcinogenesis. (2010) 31:946–52. doi: 10.1093/carcin/bgq046
183. Oshima M, Oshima H, Kitagawa K, Kobayashi M, Itakura C, and Taketo M. Loss of apc heterozygosity and abnormal tissue building in nascent intestinal polyps in mice carrying a truncated apc gene. Proc Natl Acad Sci U.S.A. (1995) 92:4482–6. doi: 10.1073/pnas.92.10.4482
184. Oshima H, Oshima M, Kobayashi M, Tsutsumi M, and Taketo MM. Morphological and molecular processes of polyp formation in apc(Delta716) knockout mice. Cancer Res. (1997) 57:1644–9.
185. Tomita H, Yamada Y, Oyama T, Hata K, Hirose Y, Hara A, et al. Development of gastric tumors in apc(Min/+) mice by the activation of the beta-catenin/tcf signaling pathway. Cancer Res. (2007) 67:4079–87. doi: 10.1158/0008-5472.Can-06-4025
186. Schnitzler M, Koorey D, Dwight T, Tomaras C, Macrae F, Marsh D, et al. Frequency of codon 1061 and codon 1309 apc mutations in Australian familial adenomatous polyposis patients. Hum Mutat. (1998) Suppl 1:S56–7. doi: 10.1002/humu.1380110120
187. Ficari F, Cama A, Valanzano R, Curia MC, Palmirotta R, Aceto G, et al. Apc gene mutations and colorectal adenomatosis in familial adenomatous polyposis. Br J Cancer. (2000) 82:348–53. doi: 10.1054/bjoc.1999.0925
188. Quesada CF, Kimata H, Mori M, Nishimura M, Tsuneyoshi T, and Baba S. Piroxicam and acarbose as chemopreventive agents for spontaneous intestinal adenomas in apc gene 1309 knockout mice. Jpn J Cancer Res. (1998) 89:392–6. doi: 10.1111/j.1349-7006.1998.tb00576.x
189. Niho N, Takahashi M, Kitamura T, Shoji Y, Itoh M, Noda T, et al. Concomitant suppression of hyperlipidemia and intestinal polyp formation in apc-deficient mice by peroxisome proliferator-activated receptor ligands. Cancer Res. (2003) 63:6090–5.
190. Niho N, Mutoh M, Komiya M, Ohta T, Sugimura T, and Wakabayashi K. Improvement of hyperlipidemia by indomethacin in min mice. Int J Cancer. (2007) 121:1665–9. doi: 10.1002/ijc.22872
191. Pollard P, Deheragoda M, Segditsas S, Lewis A, Rowan A, Howarth K, et al. The apc 1322t mouse develops severe polyposis associated with submaximal nuclear beta-catenin expression. Gastroenterology. (2009) 136:2204–13.e1-13. doi: 10.1053/j.gastro.2009.02.058
192. Lewis A, Segditsas S, Deheragoda M, Pollard P, Jeffery R, Nye E, et al. Severe polyposis in apc(1322t) mice is associated with submaximal wnt signalling and increased expression of the stem cell marker lgr5. Gut. (2010) 59:1680–6. doi: 10.1136/gut.2009.193680
193. Fodde R, Edelmann W, Yang K, van Leeuwen C, Carlson C, Renault B, et al. A targeted chain-termination mutation in the mouse apc gene results in multiple intestinal tumors. Proc Natl Acad Sci U.S.A. (1994) 91:8969–73. doi: 10.1073/pnas.91.19.8969
194. Bu XD, Li N, Tian XQ, and Huang PL. Caco-2 and ls174t cell lines provide different models for studying mucin expression in colon cancer. Tissue Cell. (2011) 43:201–6. doi: 10.1016/j.tice.2011.03.002
195. Cizkova K, Birke P, Malohlava J, Tauber Z, Huskova Z, and Ehrmann J. Ht-29 and caco2 cell lines are suitable models for studying the role of arachidonic acid-metabolizing enzymes in intestinal cell differentiation. Cells Tissues Organs. (2019) 208:37–47. doi: 10.1159/000506735
196. Zhang Q, Xiao H, Jin F, Li M, Luo J, and Wang G. Cetuximab improves azd6244 antitumor activity in colorectal cancer ht29 cells in vitro and in nude mice by attenuating her3/akt pathway activation. Oncol Lett. (2018) 16:326–34. doi: 10.3892/ol.2018.8674
197. Chi X, Yang P, Zhang E, Gu J, Xu H, Li M, et al. Significantly increased anti-tumor activity of carcinoembryonic antigen-specific chimeric antigen receptor T cells in combination with recombinant human il-12. Cancer Med. (2019) 8:4753–65. doi: 10.1002/cam4.2361
198. Lyu N, Pedersen B, Shklovskaya E, Rizos H, Molloy MP, and Wang Y. Sers characterization of colorectal cancer cell surface markers upon anti-egfr treatment. Explor (Beijing). (2022) 2:20210176. doi: 10.1002/exp.20210176
199. Zhang BL, Li D, Gong YL, Huang Y, Qin DY, Jiang L, et al. Preclinical evaluation of chimeric antigen receptor-modified T cells specific to epithelial cell adhesion molecule for treating colorectal cancer. Hum Gene Ther. (2019) 30:402–12. doi: 10.1089/hum.2018.229
200. Eshleman JR, Donover PS, Littman SJ, Swinler SE, Li GM, Lutterbaugh JD, et al. Increased transversions in a novel mutator colon cancer cell line. Oncogene. (1998) 16:1125–30. doi: 10.1038/sj.onc.1201629
201. Zhou Y, Wen P, Li M, Li Y, and Li XA. Construction of chimeric antigen receptor−Modified T cells targeting epcam and assessment of their anti−Tumor effect on cancer cells. Mol Med Rep. (2019) 20:2355–64. doi: 10.3892/mmr.2019.10460
202. Botchkina GI, Zuniga ES, Das M, Wang Y, Wang H, Zhu S, et al. New-generation taxoid sb-T-1214 inhibits stem cell-related gene expression in 3d cancer spheroids induced by purified colon tumor-initiating cells. Mol Cancer. (2010) 9:192. doi: 10.1186/1476-4598-9-192
203. Huang Q, Xi J, Wang L, Wang X, Ma X, Deng Q, et al. Correction to: mir-153 suppresses ido1 expression and enhances car T cell immunotherapy. J Hematol Oncol. (2018) 11:90. doi: 10.1186/s13045-018-0633-1
204. Bhattacharyya NP, Ganesh A, Phear G, Richards B, Skandalis A, and Meuth M. Molecular analysis of mutations in mutator colorectal carcinoma cell lines. Hum Mol Genet. (1995) 4:2057–64. doi: 10.1093/hmg/4.11.2057
205. Fricke F, Lee J, Michalak M, Warnken U, Hausser I, Suarez-Carmona M, et al. Tgfbr2-dependent alterations of exosomal cargo and functions in DNA mismatch repair-deficient hct116 colorectal cancer cells. Cell Commun Signal. (2017) 15:14. doi: 10.1186/s12964-017-0169-y
206. Eshleman JR, Lang EZ, Bowerfind GK, Parsons R, Vogelstein B, Willson JK, et al. Increased mutation rate at the hprt locus accompanies microsatellite instability in colon cancer. Oncogene. (1995) 10:33–7.
207. Eshleman JR, Markowitz SD, Donover PS, Lang EZ, Lutterbaugh JD, Li GM, et al. Diverse hypermutability of multiple expressed sequence motifs present in a cancer with microsatellite instability. Oncogene. (1996) 12:1425–32.
208. Le Marchand L. The role of heterocyclic aromatic amines in colorectal cancer: the evidence from epidemiologic studies. Genes Environ. (2021) 43:20. doi: 10.1186/s41021-021-00197-z
209. Roper J and Hung KE. Priceless gemms: genetically engineered mouse models for colorectal cancer drug development. Trends Pharmacol Sci. (2012) 33:449–55. doi: 10.1016/j.tips.2012.05.001
210. Schroeder A, Heller DA, Winslow MM, Dahlman JE, Pratt GW, Langer R, et al. Treating metastatic cancer with nanotechnology. Nat Rev Cancer. (2011) 12:39–50. doi: 10.1038/nrc3180
211. Gao H, Korn JM, Ferretti S, Monahan JE, Wang Y, Singh M, et al. High-throughput screening using patient-derived tumor xenografts to predict clinical trial drug response. Nat Med. (2015) 21:1318–25. doi: 10.1038/nm.3954
212. McEntee MF, Cates JM, and Neilsen N. Cyclooxygenase-2 expression in spontaneous intestinal neoplasia of domestic dogs. Vet Pathol. (2002) 39:428–36. doi: 10.1354/vp.39-4-428
Keywords: chemical carcinogen, colorectal cancer, patient-derived xenograft, precision medicine, transgene animal
Citation: Huang Q, Wei S, Yu J, Wu Y, Lai H, Wei W, Yan L, Su C, Shi W and Su Z (2026) The paradigm shift: re-evaluating preclinical animal models for colorectal cancer in the precision medicine era. Front. Immunol. 16:1744692. doi: 10.3389/fimmu.2025.1744692
Received: 12 November 2025; Accepted: 24 December 2025; Revised: 18 December 2025;
Published: 15 January 2026.
Edited by:
Rahul Shivahare, The Ohio State University, United StatesReviewed by:
Ravi Thakur, Christ University, IndiaShailesh Singh, Centre for Cellular and Molecular Biology (CCMB), India
Copyright © 2026 Huang, Wei, Yu, Wu, Lai, Wei, Yan, Su, Shi and Su. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Zijie Su, emlqaWVzdUAxMjYuY29t; Wei Shi, c2hpd2VpMjAyMkBzci5neG11LmVkdS5jbg==