Colorectal cancer (CRC) represents one of the leading global health burdens, with the liver being its predominant site of distant metastasis. A significant proportion of patients with colorectal liver metastases (CRLM) are diagnosed with synchronous lesions (sCRLM), identified at or shortly after diagnosis of the primary tumor, a clinical scenario linked to aggressive disease and poor outcomes. A major clinical challenge in this setting is that the majority of CRCs with liver metastasis are microsatellite stable or low instable (MSS or MSI-L, respectively), tumor types largely unresponsive to immune checkpoint (IC) therapies. Indeed, MSS/MSI-L tumors display "immune cold" microenvironments dominated by immunosuppressive mechanisms, which severely constrain treatment options. This unmet clinical need highlights the importance of dissecting the immune microenvironment of sCRLM, to identify early drivers of metastatic progression and highlight immunological features with therapeutic potential, ultimately supporting a new immune-weighted stratification of patients for novel therapeutic approaches. To this end, we combined single-cell transcriptomic with paired V(D)J profiling to generate a multi-compartmental innate immune map of synchronous CRC and matched sCRLM, together with circulating blood samples. Our analyses revealed that Tumor-Associated Macrophages (TAMs) and Dendritic Cells (DCs) emerged as dominant players of the tumor microenvironment (TME) along divergent tumor-specific trajectories. In CRC, TAMs were enriched in inflammatory states that paradoxically supported immune activation while simultaneously laying the groundwork for metastatic dissemination. Conversely, CRLM was dominated by proliferative CTLA4high immunoregulatory TAMs, whose higher frequencies were associated with shorter disease-free survival and worse prognosis. Furthermore, we identified two CRLM-specific TAM subsets characterized by IL7R expression and a cytotoxic/NK-like signature, suggesting unconventional differentiation trajectories or functional mimicry through trogocytosis. In parallel, DCs displayed a CRLM-specific developmental program in which cDC1 subsets differentiated into immunoregulatory LAMP3+ DCs (mregDCs) enriched for IDO1 expression and immune checkpoint ligands. Natural Killer (NK) and Innate Lymphoid Cells (ILCs) showed strikingly distinct patterns between blood and tumor tissues. Circulating NK cells followed a canonical CD56bright-to-CD56dim trajectory, which bifurcated into a conventional cytotoxic-high branch and into a disease-specific cytokine hepatic-primed one. In sharp contrast, within CRLM the CD56bright population expanded dramatically compared to circulation, and diversified into IFNG-activated, cytotoxic-enriched, and stress-adapted states, with no developmental connection to hepatic CD56dim NK cells. Importantly, CRLM-associated CD56bright NK cells displayed concurrent upregulation of effector programs and multiple inhibitory receptors, including TIGIT, CXCR4, and SIGIRR, reflecting a paradoxical state of activation restrained by layered inhibitory checkpoints. Finally, integrative analyses resolved a highly heterogeneous γδ T cell landscape across blood, CRC, and CRLM. Vδ1 T cells predominated in all compartments, with an inversion of the expected Vδ2 dominance in circulation, indicating disease-associated remodeling. Gene co-expression analysis revealed tissue-specific functional signatures, including IFNG-driven activation programs in tumors, cytotoxic/TEMRA modules in blood, and residency/regulatory features in CRC. By integrating TCR clonotype tracking, we uncovered direct evidence of active trafficking of identical γδ T clonotypes from blood to CRLM. Crucially, these clonotypes underwent compartment-specific functional rewiring, with cytotoxic programs preserved in circulation, whereas intrahepatic counterparts with the acquisition of exhaustion and stress signatures, highlighting the liver metastatic niche as a site of immune restraint. Altogether, these findings delineate possible mechanisms involved in the early immunological remodeling that occurs in synchronous CRLM, showing how the hepatic niche actively reprograms myeloid, NK, and γδ T cell compartments toward states of restrained activation and immune escape. By linking clonal fate mapping with transcriptional states, this work provides a mechanistic framework for characterizing early immune influences in synchronous CRLM, reframing the liver not as a passive metastatic destination but as an active architect of immune escape. Such mechanistic insights not only expand the understanding of metastatic immune biology but also point toward new opportunities for therapeutic innovation, by paving the way for novel immunotherapeutic strategies specifically targeting IC-resistant CRLM.

HIGH-THROUGHPUT PROFILING OF INNATE IMMUNE CELLS TO UNVEIL TUMOR ESCAPE MECHANISMS IN MATCHED COLORECTAL CANCER AND LIVER METASTASES

MARZANO, PAOLO
2026

Abstract

Colorectal cancer (CRC) represents one of the leading global health burdens, with the liver being its predominant site of distant metastasis. A significant proportion of patients with colorectal liver metastases (CRLM) are diagnosed with synchronous lesions (sCRLM), identified at or shortly after diagnosis of the primary tumor, a clinical scenario linked to aggressive disease and poor outcomes. A major clinical challenge in this setting is that the majority of CRCs with liver metastasis are microsatellite stable or low instable (MSS or MSI-L, respectively), tumor types largely unresponsive to immune checkpoint (IC) therapies. Indeed, MSS/MSI-L tumors display "immune cold" microenvironments dominated by immunosuppressive mechanisms, which severely constrain treatment options. This unmet clinical need highlights the importance of dissecting the immune microenvironment of sCRLM, to identify early drivers of metastatic progression and highlight immunological features with therapeutic potential, ultimately supporting a new immune-weighted stratification of patients for novel therapeutic approaches. To this end, we combined single-cell transcriptomic with paired V(D)J profiling to generate a multi-compartmental innate immune map of synchronous CRC and matched sCRLM, together with circulating blood samples. Our analyses revealed that Tumor-Associated Macrophages (TAMs) and Dendritic Cells (DCs) emerged as dominant players of the tumor microenvironment (TME) along divergent tumor-specific trajectories. In CRC, TAMs were enriched in inflammatory states that paradoxically supported immune activation while simultaneously laying the groundwork for metastatic dissemination. Conversely, CRLM was dominated by proliferative CTLA4high immunoregulatory TAMs, whose higher frequencies were associated with shorter disease-free survival and worse prognosis. Furthermore, we identified two CRLM-specific TAM subsets characterized by IL7R expression and a cytotoxic/NK-like signature, suggesting unconventional differentiation trajectories or functional mimicry through trogocytosis. In parallel, DCs displayed a CRLM-specific developmental program in which cDC1 subsets differentiated into immunoregulatory LAMP3+ DCs (mregDCs) enriched for IDO1 expression and immune checkpoint ligands. Natural Killer (NK) and Innate Lymphoid Cells (ILCs) showed strikingly distinct patterns between blood and tumor tissues. Circulating NK cells followed a canonical CD56bright-to-CD56dim trajectory, which bifurcated into a conventional cytotoxic-high branch and into a disease-specific cytokine hepatic-primed one. In sharp contrast, within CRLM the CD56bright population expanded dramatically compared to circulation, and diversified into IFNG-activated, cytotoxic-enriched, and stress-adapted states, with no developmental connection to hepatic CD56dim NK cells. Importantly, CRLM-associated CD56bright NK cells displayed concurrent upregulation of effector programs and multiple inhibitory receptors, including TIGIT, CXCR4, and SIGIRR, reflecting a paradoxical state of activation restrained by layered inhibitory checkpoints. Finally, integrative analyses resolved a highly heterogeneous γδ T cell landscape across blood, CRC, and CRLM. Vδ1 T cells predominated in all compartments, with an inversion of the expected Vδ2 dominance in circulation, indicating disease-associated remodeling. Gene co-expression analysis revealed tissue-specific functional signatures, including IFNG-driven activation programs in tumors, cytotoxic/TEMRA modules in blood, and residency/regulatory features in CRC. By integrating TCR clonotype tracking, we uncovered direct evidence of active trafficking of identical γδ T clonotypes from blood to CRLM. Crucially, these clonotypes underwent compartment-specific functional rewiring, with cytotoxic programs preserved in circulation, whereas intrahepatic counterparts with the acquisition of exhaustion and stress signatures, highlighting the liver metastatic niche as a site of immune restraint. Altogether, these findings delineate possible mechanisms involved in the early immunological remodeling that occurs in synchronous CRLM, showing how the hepatic niche actively reprograms myeloid, NK, and γδ T cell compartments toward states of restrained activation and immune escape. By linking clonal fate mapping with transcriptional states, this work provides a mechanistic framework for characterizing early immune influences in synchronous CRLM, reframing the liver not as a passive metastatic destination but as an active architect of immune escape. Such mechanistic insights not only expand the understanding of metastatic immune biology but also point toward new opportunities for therapeutic innovation, by paving the way for novel immunotherapeutic strategies specifically targeting IC-resistant CRLM.
27-gen-2026
Inglese
MAVILIO, DOMENICO
MIKULAK, JOANNA
LANDSBERGER, NICOLETTA
Università degli Studi di Milano
169
File in questo prodotto:
File Dimensione Formato  
phd_unimi_R13849.pdf

embargo fino al 22/06/2027

Licenza: Creative Commons
Dimensione 4.16 MB
Formato Adobe PDF
4.16 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/355476
Il codice NBN di questa tesi è URN:NBN:IT:UNIMI-355476