CRISPR-Engineered Akkermansia muciniphila Synergizes with Intratumoral Microbiome Remodeling to Overcome Immunotherapy Resistance in KRAS-Mutant Colorectal Cancer
Kianoosh Soltani,1,*Asal Tajik,2Shiva Farahani,3Mahdiyeh Rahimzadeh,4
1. Faculty of Basic Sciences, Islamic Azad University, Karaj Branch 2. Faculty of Basic Sciences, Islamic Azad University, Islamshahr Branch 3. Faculty of Basic Sciences, Islamic Azad University, Islamshahr Branch 4. Faculty of Basic Sciences, Islamic Azad University, Islamshahr Branch
Introduction: Colorectal cancer (CRC) remains the third most common malignancy worldwide, with KRAS mutations present in approximately 45% of patients and driving resistance to anti-EGFR therapies. Although immune checkpoint inhibitors (ICIs) have revolutionized treatment paradigms for several solid tumors, microsatellite stable (MSS) CRC exhibits primary resistance to immunotherapy with response rates below 5%. Emerging evidence implicates the gut microbiome as a critical determinant of ICI efficacy, yet the mechanistic interplay between tumor-specific somatic mutations and intratumoral microbiota remains poorly understood. Specifically, whether KRAS mutations actively reshape the tumor microbial niche and whether this ecosystem can be therapeutically engineered using CRISPR-based tools have not been systematically addressed.
Methods: We conducted an integrated multi-omic analysis involving 120 KRAS-mutant CRC patients and 60 wild-type controls. Tumor microbiomes were profiled using deep shotgun metagenomic sequencing, and host transcriptional landscapes were assessed via RNA-seq. Functional validation was performed using isogenic CRC cell lines harboring mutant versus wild-type KRAS and intestine-specific KRAS^G12D murine models. We genetically engineered Akkermansia muciniphila using a non-viral CRISPR-Cas9 system to secrete a tumor-lytic payload and an anti-PD-1 nanobody, constructing a precision "living therapeutic." Therapeutic efficacy was evaluated in orthotopic CRC models, with tumor immune microenvironment dynamics characterized by single-cell RNA sequencing and flow cytometry.
Results: KRAS mutations significantly reshaped the intratumoral microbial community, driving enrichment of Bacteroides acidifaciens and Fusobacterium nucleatum, which correlated with poor prognosis and immunosuppressive microenvironments. Mechanistically, B. acidifaciens surface protein SusF bound and stabilized ARHGEF2, potentiating active KRAS and RAF-MEK-ERK signaling, thereby establishing a feedforward oncogenic circuit. Conversely, administration of wild-type A. muciniphila suppressed tumor-associated macrophage efferocytosis and enhanced CD8+ T cell effector function, sensitizing MSS CRC to anti-PD-1 therapy. Building on these findings, our CRISPR-engineered A. muciniphila strain achieved robust tumor colonization (>10⁶ CFU/g tumor tissue), delivered immunomodulatory payloads locally, and reduced tumor burden by 68% in KRAS^G12D murine models (p < 0.001 versus mock). Single-cell profiling revealed that engineered bacteria remodeled the immune microenvironment, increasing CD8+ T cell infiltration by 3.2-fold and reducing M2-like macrophages by 54%. Importantly, this therapeutic strategy restored responsiveness to anti-PD-1 therapy in previously resistant MSS CRC models, with a complete response rate of 42% in combination cohorts.
Conclusion: This study provides the first evidence that KRAS-driven reprogramming of the intratumoral microbiome establishes an immunosuppressive niche that can be therapeutically reversed using CRISPR-engineered probiotics. Our "living therapeutic" platform offers a localized, sustained therapeutic approach with reduced systemic toxicity, addressing a critical unmet need in immunotherapy-resistant CRC. These findings establish a foundation for clinical translation and support the development of microbiome-informed precision oncology strategies.