Abstract
<jats:p><p dir="ltr">Although clinical practice for cancer patients has advanced significantly over the past decades, particularly due to the major breakthrough of immunotherapy, a substantial proportion of patients across various cancer types remain refractory to treatment or experience relapse and tumor progression due to primary or acquired resistance. Solid evidence indicates that cancer cells activating epithelial-to-mesenchymal plasticity (EMP) programmes gain a survival advantage that accelerates both cancer progression and colonization of distant sites, while conferring resistance to multiple therapies. Even more notably, EMP has recently emerged as a link between cancer-associated inflammation and immunomodulation, by reshaping both immune microenvironment functionality and cancer cell immunogenicity, with the potential to become a novel node for targeting to improve therapy efficacy in combination therapeutic approaches. Many mechanisms linking EMP to enhanced immune modulation and metastasis are already known, but many others remain to be elucidated.</p><p dir="ltr">Building on the current state of the art for each disease under study, we aimed to uncover the gaps in knowledge that have so far represented unresolved bottlenecks in each of the solid tumors under investigation: lung cancer, breast cancer, and brain cancers. Furthermore, aware of methodological limitations in the field, we aimed to reshape the methodology toward a more comprehensive and universally applicable EMP classification.</p><p dir="ltr">Overall, this thesis aims to elucidate the molecular interactors through which cancer cell plasticity remodels cancer behavior to favor aggressiveness, with the goal of identifying new targetable pathways that could improve therapy efficacy.</p><p dir="ltr">In Paper I, we defined a new method to score plasticity in cells, exploiting dynamic changes in organelle morphology occurring during EMP. This represents a novel, unbiased approach capable of detecting multiple EMP states by relying, for the first time, not on classical quantitative approaches - which carry the disadvantage of being context- dependent and cell-type specific - but rather on common, shared, and universal changes that characterize cells undergoing EMP. The potential of this approach lies in an increased accuracy in detecting intermediate EMP states, which are the most relevant to cancer aggressiveness, and holds promise for other high-throughput applications, such as drug screenings aimed at reverting the EMP phenotype.</p><p dir="ltr">In Paper II, we identified a novel molecular connection between EMP and immune modulation in patients with small cell lung cancer, which led us to explore a novel axis contributing to NK cell resistance. In a disease characterized by an extremely low response rate to immunotherapy and immune checkpoint blockade (ICB), yet a high level of EMP, we first demonstrated as a proof of concept that a common modulator of both plasticity and immunity exists, driving the two programmes in opposite directions. We then explored the effects and implications of manipulating this node on the anti-tumor response. We found that higher levels of the transcription factor ELF3 are associated with decreased infiltration of activated NK cells in clinical samples, and that manipulation of ELF3 leads to increased antigen presentation machinery (APM) genes expression and NK cell resistance in vitro. These findings open a promising window of opportunity for developing strategies aimed at targeting such common modulators to improve overall immunotherapy benefit.</p><p dir="ltr">In Paper III, we showed that cells overexpressing EGFR are preferentially directed toward the lymph nodes in triple-negative breast cancer (TNBC), uncovering a new chemotactic axis driving lymphatic metastasis. EGFR has been implicated in TNBC aggressiveness, but drugs targeting this factor have so far proven ineffective in clinical trials. By studying the plasticity-induced reprogramming of lymphatic endothelial cells, we found that they secrete EGFR ligands, and identified TGF-a among these as a chemotactic driver of metastasis for EGFR-expressing TNBC cells. Blocking the EGFR/TGF-a axis in vivo reduced lymphatic metastasis, revealing that EGFR drives aggressiveness through this mechanism, rather than merely acting as an essential factor for tumor survival. This suggests that EGFR targeting could be repositioned earlier in treatment protocols, within a timeline that allows prevention of lymphatic spread.</p><p dir="ltr">In Paper IV, we used immunogenomic analysis to investigate new targets for adoptive cell therapy (ACT) in pons diffuse midline glioma (pDMG), a highly plastic brain cancer for which therapeutic options remain dismal for a subset of patients, highlighting the need for personalized, targetable approaches. Mutational analysis of a patient cohort revealed several strong candidate targets for TCR-based cell therapy, based on neoantigen predictions that we were able to validate through functional analysis. In this approach, we used plasticity itself as the antigen to be targeted. Overall, we established a new pipeline that, in the future, could enable the relatively fast generation of an off-the-shelf product by using healthy donors as a source of antigen-reactive TCRs, subsequently applied in ACT with the aim of re-educating the patient's own T cell compartment.</p><p dir="ltr">Collectively, the findings in this thesis position cancer cell plasticity as a central driver of tumor aggressiveness, immune evasion, and metastatic progression; by identifying both novel therapeutic targets and a refined methodological framework for EMP classification, this work provides a foundation for designing combination therapy strategies tailored to the plastic, immune-evasive nature of cancer cells. Together, these results support the rationale for targeting EMP-related programmes as a means to overcome therapy resistance and improve clinical outcomes across multiple solid tumors.</p><h3 dir="ltr">List of scientific papers</h3><p dir="ltr">I. Slager, J*, <b>Gatto, F</b> *. , Frey, B., Shi, W., Porebski, B., Carreras-Puigvert, J., Parniewska, M.M., Fuxe, J., 2025. A morphology-based machine learning model for scoring epithelial-mesenchymal plasticity using organelle dynamics. Commun Biol. 9, 59. *Equal contribution. <a href="https://doi.org/10.1038/s42003-025-09326-8" rel="noreferrer" target="_blank">https://doi.org/10.1038/s42003-025-09326-8</a><br></p><p dir="ltr">II. <b>Gatto, F.</b>, Beninato, T., Pernici, D., Portararo, P., Zanichelli, A., Shi, W., Prelaj, A., Sangaletti, S, Fuxe, J. ELF3 is an epithelial regulator of antigen presentation and NK cell immune evasion in Small Cell Lung Cancer. [Manuscript]</p><p dir="ltr">III. Shi, W., Pan, Y., Rathod, B., Wang, Y., Wang, Z., Shen, J., <b>Gatto, F.</b>, Liu, M., Sun, Y., Wilhelm, M., Helleday, T., Ulvmar, M.H., Windahl, S.H., Karlsson, M.C.I., Fuxe, J., 2026. TGF-a/EGFR-mediated lymphatic metastasis reveals a repositionable therapeutic target in breast cancer. Npj Breast Cancer 12, 52. <a href="https://doi.org/10.1038/s41523-026-00941-0" rel="noreferrer" target="_blank">https://doi.org/10.1038/s41523-026-00941-0</a><br></p><p dir="ltr">IV. Rovesti, G., Fernandez Woodbridge, A., Yao, H, Pancaldi, A., Chiavelli, C., <b>Gatto, F.</b>, Sandvick, U., Sällberg, M., Fuxe, J., Dominici, M., Grönlund, H., Nilsson, O., Pasetto, A., Nascimento Silva, D. Immune and Genomic Profiling Reveals Targets for Cellular Therapy in Pediatric Brain Tumors. [Accepted]</p></jats:p>