Choose timezone
Your profile timezone:
Powered by Indico
v3.3.5
We welcome you to the 2025 Workshop on Cell Plasticity in Cancer Evolution registration page. This workshop will be held at the Max Planck Institute for Evolutionary Biology from May 19-22, 2025. Registration is free, but contributing participants must arrange their travel and accommodation. Please contact Maren Lehmann if you have questions.
Plasticity enables cancer cells or cells of the tumor microenvironment to change their identity, possibly affecting proliferation, quiescence, and cell motility. These mechanisms can thus counteract external pressures, e.g., during inflammation or treatment, and serve as a hotbed for unwanted adaptation, resistance evolution, and metastasis. This workshop aims to bring together experimentalists and theorists to discuss the state-of-the-art and future directions that better account for cell plasticity in cancer evolution and elucidate respective improvements in cancer treatment.
Main speakers:
Peter Friedl (Nijmegen)
Gaetano Gargiulo (Berlin)
Marco Gerlinger (London)
Sara Hamis (Uppsala)
Purificacion Muñoz Moruno (Barcelona)
Susanne Sebens (Kiel)
Heike Siebert (Kiel)
Fabian Spill (Birmingham)
For further questions, please get in touch with the scientific organizers Philipp, Qianci, Gustav, or Arne.
Cellular plasticity describes the ability of cells to switch their phenotype in response to
microenvironmental changes. It is a key phenomenon in pancreatic ductal adenocarcinoma (PDAC) esentially adding to the tumors high degree of heterogeneity, aggressiveness and therapy resistance. Furthermore, PDAC is characterized by a profound inflammatory stroma which is an important driver of cellular phenotypic changes. We could identify different PDAC cell subpopulations exhibiting defined functional phenotypes and being dependent on distinct environmental conditions. For example, in a physiological liver microenvironment, disseminated PDAC cells exhibited a quiescent phenotype, which could be reversed into a proliferative cell stage in the presence of an inflammatory hepatic microenvironment. This stromal as well as the
associated phenotypic PDAC cell switch resulted in rapid PDAC cell expansion, metabolic alterations and metastatic outgrowth. Notably, a high cellular plasticity is observed not only within cancer cells but also within stromal cells. Accordingly, different phenotypes and associated effector functions have been described for e.g. carcinoma-associated fibroblasts (CAF), macrophages or T cell populations. Tumor cells foster phenotypic switching of stromal cells to create a tumor-favoring microenvironment that further drives tumor evolution and therapy resistance. For example, we could demonstrate that macrophages acquire an immunosuppressive phenotype in the presence of PDAC cells by which they in turn contribute to immune evasion of the tumor cells. Although exhibiting immunecheckpoint molecules, their blockade was not effective in enhancing T cell mediated killing of PDAC cells, supporting the role
of macrophages in PDAC cell resistance towards immune checkpoint inhibitors. Furthermore, quiescent hepatic stellate cells were able to control growth of PDAC cells, while upon activation to a myofibroblastic phenotype they promoted PDAC cell proliferation. Altogether, these findings emphasize that plastic cells are key drivers of PDAC evolution and highlight the importance to better understand the various shades of plasticity, as well as the underlying dynamics and mechanisms in order to efficiently improve PDAC therapy.
Authors: M. Andersen, H. Hasselbalch, T. Stiehl, J.T. Ottesen
Human blood cell production is maintained by hematopoietic stem cells (HSC) which give rise to all types of mature blood cells. Experimental observation of HSC in their physiologic bone-marrow microenvironment, the so-called stem cell niche, is challenging. Therefore, the details of HSC dynamics and the cellular interactions in the stem cell niche remain elusive. Mutations that lead to a competitive advantage are the cause of clinical challenges when treating HSC-derived malignancies such as acute myeloid leukemia or the myeloproliferative neoplasms (MPN). To investigate the significance of the interaction between the HSC and the stem cell niche in these malignancies, we propose and analyse a mechanism-based mathematical model of HSC dynamics within the bone-marrow microenvironment. The JAK2V617 mutation is a key driver for overproduction of blood cells in MPN with chronic inflammation acting as an additional driver of the disease. Long-term treatment with interferon-alfa (IFN) can reduce the disease burden of MPN patients. Determining individual patient responses to IFN therapy may allow for efficient personalized treatment, reducing both drop-out and disease burden. The mathematical model is calibrated to data of the randomized trial study DALIAH. Through comprehensive modeling of the effects of IFN, the model was related to individualized patient-data consisting of longitudinal hematologic and molecular measurements. We believe that this approach could have direct clinical relevance, offering expert guidance for clinical decisions about IFN treatment of MPN patients. If time permits we will address recent findings of mathematical modeling and data of pre-disease dynamics (CHIP) as well as experimental combination therapy for MPN patients receiving IFN combined with inflammation inhibitors.
The traditional clonal evolution model integrates genetic mutations and clonal selection but often neglects the critical role of phenotypic plasticity. In this study, we combine longitudinal whole-genome sequencing and flow cytometry data to investigate the progression of chronic lymphocytic leukaemias with bimodal CD49d expression, a prime example of complex clonal evolution. We introduce a novel population genetics framework that incorporates reversible phenotypic switching into clonal dynamics, enabling the reconstruction of evolutionary trajectories for CD49d+ and CD49d- subpopulations. By leveraging mutation data and white blood cell counts, we quantify patient-specific CD49d expression heritability and selection parameters. Our findings demonstrate that transitioning to CD49d+ frequently confers a selective advantage, driving differential growth rates between subpopulations. The heritability of CD49d expression varies across patients, reflecting a spectrum of selective pressures and plasticity. This enhanced clonal deconvolution and phylogenetic analysis underscores the interplay of genetic and epigenetic factors in shaping tumour clonal architecture, providing a robust foundation for developing personalised therapeutic strategies targeting phenotypic heterogeneity.
Authors: Lisa-Marie Philipp, Axel Künstner, Anne-Sophie Mehdorn, Charlotte Hauser, Jan-Paul Gundlach, Olga Will, Sören Franzenburg, Hendrike Knaack, Udo Schumacher, Hauke Busch, Susanne Sebens
Pancreatic ductal adenocarcinoma (PDAC) is mostly diagnosed at advanced or even metastasized stages limiting patient´s prognosis and overall survival. Metastasis requires strong cancer cell plasticity and high tumor cell heterogeneity implying phenotypic switching, notably, Epithelial-Mesenchymal-Transition (EMT) being associated with the gain of cancer stem cell (CSC) properties, in response to changing environments. To analyze whether CSC-properties are related to distinct EMT-phenotypes and if this CSC-EMT axis impacts malignancy-associated properties, this study aimed at characterizing mesenchymal-like and epithelial PDAC cell variants, with the focus on CSC populations. Single-cell cloning of PDAC cells revealed CSC (Holoclone) and non-CSC (Paraclone) clones from both mesenchymal-like (Panc1) and epithelial (Panc89) cells. Comparatively analyzing parental Panc1 and Panc89 as well as related Holo- and Paraclone cells, it was found that Panc1 and Panc89 cell variants exhibit differences in their colony formation ability, reflecting the potential to self-renew, as well as distinct transcriptional CSC and EMT signatures. While Panc1 cell variants show increased signatures of EMT, Panc89 cell variants exhibit increased self-renewal capacity and CSC signatures. Isolated Panc1 Holoclone cells show a mesenchymal phenotype dominated by high expression of the CSC marker Nestin, while Panc89 Holoclone cells exhibit a Sox2-dominated epithelial stemness phenotype. Functional analysis revealed decreased cell growth for Panc1 cell variants compared to Panc89 cell variants, while response to chemotherapy was overall higher Panc89 cell variants. Panc1 Holoclones show the weakest response to treatment, while Panc89 Holoclones showed the strongest affection to therapy. Invasion assays presented Panc1 cell variants to have an increased invasion potential compared to Panc89 cell variants, while Panc89 cell populations show stronger migration ability. Further detailed, Panc1 Holoclone cells are highly invasive in a mesenchymal-like invasion manner, while Panc89 Holoclone cells show pronounced cell migration in clusters. In vivo, Panc1 and Panc89 cell variants essentially differ with respect to their metastatic capacity, as Panc1 and Panc89 Holo- and Paraclone tumors varied regarding number and size of metastases formed as well as organ manifestation, leading to different survival outcomes. Overall, these data support the view of EMT-related plasticity and heterogeneity within cancer (stem) cells in PDAC, differentially impacting metastatic propensities.
Metabolic stress is a frequent adverse event in tumors caused by mutations, malperfusion, hypoxia, and nutrition deficit. The resulting bioenergetic deprivation triggers signaling, mechanical and metabolic adaptation responses in tumor cells to secure survival and adjust migration activity. Using 3D invasion models and preclinical intravital microscopy in mouse models of breast cancer, we identified kinetic responses of cancer cells to energy deficit. Upon challenge with glucose deprivation, mild acidosis and hypoxia signaling, invading cancer cells switched from energy-consuming collective to energy-efficient amoeboid migration and an enhanced capability for distant metastasis. In this seminar, I will discuss the molecular mechanisms of collective to single-cell plasticity, the role of autophagy and proteasomal degradation, and adaptive programs of cell-cell and cell-matrix adhesions. As emerging concept, low-adhesive, amoeboid dissemination represents a critical route to metastasis and therapy resistance. Understanding the biomechanics and energetic requirements of amoeboid and other dissemination strategies offers rationales for improving therapeutic targeting of metastatic cancer progression.
Authors: Raafat Chalar, Yujie Xiao, Joon-Hyun Song, Naheel Khatri, Jowana Obeid, Andrew Chen, Fabiola Velazquez, Daniel Canal, Yusuf Hannun, Giovanni D'Angello, Mehdi Damaghi
Phenotypic plasticity and metabolic reprogramming in cancer cells are essential for their adaptation to their harsh tumor microenvironment (TME). Pre-existing cell states and phenotypic plasticity can define the evolutionary trajectories and eventually cell fates under TME selection pressures. Here we used spatial multi-omics approaches at single cell resolution followed by principles of ecology analysis to decipher the impact of the sphingolipid metabolic plasticity on the evolution of breast cancer cells under selection of acidic microenvironment. Cancer cells exposed to acute and chronic acid stress regulate their sphingolipid metabolism with an impressive level of plasticity to choose the ceramide metabolic pathway. Acute acid exposure induces ceramides accumulation that is toxic to the cells and needs to be cleared so the cells survive. Remarkably, cancer cells with long term acid exposure efficiently suppress ceramide accumulation by utilizing various ceramide metabolic pathways. CRISPR/Cas screening revealed the ceramide-> Sphingosin-1-P (S1P) pathway is essential for cancer cell viability under acidic conditions. Notably, multiple Ceramidases (CDases) and Sphingosine Kinase-1 (SPHK1) were significantly upregulated in both in vitro and in vivo acidic environments. When S1P was blocked, cancer cells still survived implying their plasticity. We then performed single cell multi-omics (RNA/ATAC) sequencing data to capture heterogeneity of response in response to acid seeking for alternatives for S1P path.Transcriptome analysis revealed over expression of several enzymes of glucosylceramide pathway in acid adapted cells. We then blocked both S-1-P and glucosylceramide pathways and cancer cells still survived. To explore the plasticity of ceramide metabolism, we investigated all four major metabolic pathways in 3D spheroids and patient-derived organoids. Using MALDI mass spectrometry, we confirmed the mechanisms of action of targeted inhibitors and observed pathway switching when one metabolic route was blocked. Our findings revealed a high degree of metabolic plasticity in cancer cells, showcasing their ability to adapt by redirecting metabolic flux under selective pressure. Leveraging this plasticity, we employed sphingolipid metabolism inhibitors in patient-derived organoids to manipulate cancer cell behavior and evolutionary trajectories. This approach aimed to shift the adaptive landscape of cancer cells as part of an evolutionary therapy strategy, offering a promising new therapeutic avenue that targets cancer cell resilience and adaptability.
Authors: Irina Kareva, Jana Gevertz
Despite the revolutionary impact of immune checkpoint inhibition on cancer therapy, the lack of response in a subset of patients, as well as the emergence of resistance, remain significant challenges. Here we explore the theoretical consequences of the existence of multiple states of immune cell exhaustion on response to checkpoint inhibition therapy. In particular, we consider the emerging understanding that T cells can exist in various states: fully functioning cytotoxic cells, reversibly exhausted cells with minimal cytotoxicity, and terminally exhausted cells. We hypothesize that inflammation augmented by drug activity triggers transitions between these phenotypes, which can lead to non-genetic resistance to checkpoint inhibitors. We introduce a conceptual mathematical model, coupled with a standard 2-compartment pharmacometric (PK) model, that incorporates these mechanisms. Simulations of the model reveal that, within this framework, the emergence of resistance to checkpoint inhibitors can be mitigated through altering the dose and the frequency of administration. Our analysis also reveals that standard PK metrics do not correlate with treatment outcome. However, we do find that levels of inflammation that we assume trigger the transition from the reversibly to terminally exhausted states play a critical role in therapeutic outcome. A simulation of a population that has different values of this transition threshold reveals that while the standard high-dose, low-frequency dosing strategy can be an effective therapeutic design for some, it is likely to fail a significant fraction of the population. Conversely, a metronomic-like strategy that distributes a fixed amount of drug over many doses given close together is predicted to be effective across the entire simulated population, even at a relatively low cumulative drug dose. We also demonstrate that these predictions hold if the transitions between different states of immune cell exhaustion are triggered by prolonged antigen exposure, an alternative mechanism that has been implicated in this process. Our theoretical analyses demonstrate the potential of mitigating resistance to checkpoint inhibitors via dose modulation.
Authors: A. Schilhabel, M. Khouja, S. Neumann, T. Stadler, H. Ahmed, E. Dazert, T. Gemoll, U. Günther, I. König, M. Kotrova, H. Busch, C. Baldus, C. Pott, N. von Bubnoff, M. Brüggemann, C. Khandanpour
MM is a rare hematological malignancy in Europe and is not curable(1). It relapses and the time interval from each line to the next one decreases, requiring frequent changes of regime(2). Bispecific antibody have the potential to transform the treatment paradigm of MM(2, 3). However, also for Bispecific antibody current practice mandates continuous therapy for all patients in complete remission, even though the impact on disease control remains uncertain and potential side effects may worsen over time. Moreover, prolonged administration of Bispecific antibody can exhaust T-cells, potentially impairing their therapeutic efficacy (Fig. 1). Hence, we aim to alter the application method of Bispecific antibody. Continuous therapy with bispecific antibodies might exhaust T-cells worsens prognosis and contributes to cloneal evolution. It also increases infectious complications and is associated with high cost. From (4). Assessment of response rate (either CR or CR combined with MRD-negativity) could identify patients with sustainable response, who do not require continuous therapy but potentially a reduced rate of application. It could even improve outcome of patients by reducing T-cell exhaustion and other negative effects without hampering disease control (3, 5). Simultaneously, therapy will be promptly restarted if MRD relapse is indicating imminent loss of CR. Current practice for determining MRD relies on invasive and painful bone marrow punctures. In addition, single bone marrow samplings do not accurately reflect the whole disease status, given the MM heterogeneity, high clonal evolution and sampling errors. Liquid biopsies are promising non-invasive alternative sources for MRD assessment that may overcome these limitations.
Authors: Madhuri Majumder and Samares Pal
An HIV-COVID-19 co-infection dynamics is modelled mathematically assimilating vaccination mechanism that incorporates endogenous modification of human practices generated by the COVID-19 prevalence, absorbing the relevance of treatment mechanism in suppressing the co-infection burden. Envisaging COVID-19 situation, HIV-subsystem is analysed by introducing COVID-19 vaccination for the HIV infected population as a prevention. It has been observed that Co-infection treatment needs to be emphasized parallelly with single infection medication under dual-epidemic situation. Further, an optimization technique is introduced to the co-infection model integrating vaccination and treatment control mechanisms, which approves the strategy combining vaccination with awareness and medication as the ideal one for epidemic and economic gain.
CAR-T therapy, where T cell are genetically modified to respond to CD19-presenting B cells, has revolutionized the treatment of several blood cancers. The immediate effect of this immunotherapy is associated with a significant inflammatory response in the patient, sometimes leading to a tumor that seemingly swells in the days following CAR-T cell injection. This is referred to as pseudo-progression, and in this project we present a mathematical model of CAR-T cell therapy where pro-inflammatory cytokines play a crucial role. The model is capable of capturing a wide range of behaviors, including long-term successful treatments, failed treatments and pseudo-progression. The parametrization of the model maps on to measurable patient characteristics, and we discuss how these parameters impact the long term behavior of the model.
The tumour microenvironment is characterized by alterations in the molecular and cellular composition, such as the infiltration of immune cells, reprogramming of fibroblasts, changes in ECM composition, changes in oxygen concentrations and pH, and related cell phenotype changes such as metabolic reprogramming or the emergence of migratory and invasive cancer cells. Critically, many, if not all, of these changes are related to physical changes. For example, the ECM changes its mechanical properties (e.g. stiffness, viscosity, porosity), cells modify their adhesive properties (e.g. changes in cell-cell and cell-matrix adhesion during EMT), and the forces they generate (e.g. during immune infiltration and during cancer invasion). Moreover, these physical changes are associated with spatial and geometric signatures (such as alterations of cell or nuclear morphology in cancer progression, or changes in tissue patterning resulting from immune invasion).
I will present several projects that introduce physical and geometrical changes that occur at various scales of the tumour microenvironment, ranging from molecular pathways, subcellular geometry (e.g. organelle shape), cell shape to cellular and ECM organization. I will highlight the integrative and predictive role mechanistic mathematical models can play in uncover the bi-directional links between molecular and physical features in regulating cell functions, including cell migration, proliferation and metabolism.
Treatment resistance is a major obstacle in cancer treatment. While whole-genome sequencing revealed that single mutations are sometimes sufficient to explain resistance, recent studies demonstrate that more often resistance is a multi-factorial process including non-genetic mechanism. In particular, epigenetic and phenotypic changes that take place at higher frequencies than the acquisition of genetic aberrations are now known to contribute treatment resistance. Adaptive therapies try to overcome these resistance obstacles with adaptive drug dosing that exploits cell-cell competition and balances tumour burden with drug sensitivity. Yet, a systematic study of the impact of phenotypic plasticity is lacking. In this presentation, we analyse birth-death processes and their deterministic counterparts including cell-cell competition and phenotypic switching. We determine parameter ranges for which cancers go extinct or can be contained for infinite time. In case, neither extinction nor containment is possible, we investigate the time to progression. We find that if cell-cell competition is absent, parameter ranges for infinite containment are equivalent to those for extinction. Our results show how biases and convexity in phenotypic switching rates impact phenotype distributions and treatment outcomes. Eventually, we discuss how epigenetic and transcriptomic data can be used as proxies to determine phenotype distributions in patients, and how treatments that affect switching rates could be used to improve treatment outcomes.
Authors: Simon Syga, Haralampos Hatzikirou, Andreas Deutsch
Cancer is a significant global health issue, with treatment challenges arising from intratumor heterogeneity. We study the complex relationship between somatic evolution and phenotypic plasticity, explicitly focusing on the interplay between cell migration and proliferation. We propose that evolution does not act directly on phenotypic traits, like the proliferation rate, but on the phenotypic plasticity in response to the microenvironment. We study this hypothesis using a novel, spatially explicit model that tracks individual cells' phenotypic and genetic states. We assume cells change between mobile and growing states controlled by inherited and mutation-driven genotypes and the cells' microenvironment. We observe that cells at the tumor edge evolve to favor migration over proliferation and vice versa in the tumor bulk. However, this phenotypic heterogeneity can be realized by distinct regulations of the phenotypic switch, which depend on the apoptosis rate and the cells' ability to sense their environment. Emerging synthetic tumors display varying levels of heterogeneity, which we show are predictors of the cancer's recurrence time after treatment. Interestingly, higher phenotypic heterogeneity predicts poor treatment outcomes, unlike genetic heterogeneity.
Cell plasticity refers to the ability of cells to change their identity or function in response to internal or external cues. This remarkable ability is traditionally associated with embryonic development, where stem cells and progenitors choose from multiple downstream fates to form specialized, functional tissues and organs. Under normal conditions, cell fate commitments are stable, ensuring the proper functioning of specialized tissues throughout life. In the last two decades, it has been shown that cell plasticity can be induced experimentally but also occurs naturally and in response to chronic physiological and pathological stresses. In terminally differentiated adult cells, cell plasticity primarily serves as a mechanism for tissue adaptation and repair. While it is certainly advantageous in these contexts, this plasticity also carries risks, as evidenced by its involvement in various disorders, including cancer. Our research explores how cells make decisions during embryonic development and investigates the extent to which these decisions are reversible. To understand the plasticity of cell fate commitment, we employ single-cell sequencing and genetic tracing techniques. This phenomenon is of particular interest to us because of an evolutionary perspective: as organisms became more complex, they lost much of their regenerative capacity compared to simpler organisms. While species like planarians can regenerate large portions of their bodies, most higher vertebrates have a much more limited regenerative ability. We hypothesize that, to compensate for this loss, complex organisms have evolved to utilize cell plasticity, particularly in certain cell types. Infrastructures such as blood vessels and peripheral innervation, which reach virtually every part of the body, harbour specialized cells capable of dedifferentiation, giving rise to a spectrum of other cell types. While this adaptation enables the organism to replenish needed cell types, it is also implicated in pathological conditions such as congenital syndromes, diseases, and cancer. I will summarize our findings on the plasticity of cell fate decisions, emphasizing how these insights can advance our understanding of cellular processes in the context of cancer and regenerative medicine.
Tumor heterogeneity significantly complicates cancer treatment and prognosis.
Our recent findings demonstrate that the transcriptional dosage of the oncogenic form of the KRAS gene drives tumor evolution in lung adenocarcinoma. Using mouse models, CRISPR activation screens, and chromatin profiling, we showed that - despite being already mutated and copy-number amplified - a mild transcriptional amplification of oncogenic KRAS further accelerates tumor progression and metastasis. High and low KRAS dosages integrate inflammatory and TGFβ signaling into transcriptional responses, thereby defining specific cancer cell states. Clinical data confirm that variations in KRAS dosage correlate with patient survival outcomes and distinct cell states. Direct KRAS inhibition suppresses the proliferative, KRAS-high state while sparing a residual ciliated-like, KRAS-low state, providing insights into resistance mechanisms. Collectively, our work highlights how a cancer cell intrinsic feature such as the dosage of a driver oncogene alone dictates tumor heterogeneity and evolution, offering crucial insights into therapeutic strategies targeting KRAS-driven cancers.
Acute lymphoblastic leukemia (ALL) is a type of cancer in which the bone marrow produces too many lymphocytes without further differentiation into mature blood cells. The primary treatment for most ALL cases is chemotherapy; however, after the intense treatment phase, relapse is observed in some patients. Early-stage relapses might be due to some remaining leukemia cells not being detectable by conventional cytomorphology. The molecular tests on minimal residual disease (MRD) can quantify disease burden at relatively low levels, thus helping track cancer remission and relapse. We investigated the dynamics of residual disease to describe the observed response and relapse kinetics heterogeneity. We developed stochastic, patient-specific models based on longitudinal MRD data to further the understanding of the driving factors of relapse, aiming to quantify better prediction of likelihood and timing of relapse in individual patients based on individual treatment choices.
Authors: B Vibishan, Paras Jain, Vedant Sharma, Kishore Hari, Jason T George, Mohit Kumar Jolly
Cancer is a heterogeneous disease and variability in drug sensitivity is widely documented across cancer types. Adaptive therapy is an emerging modality of cancer treatment that leverages this heterogeneity in drug resistance to achieve better therapeutic outcomes. Current treatments typically eliminate a large fraction of drug- sensitive cells and release drug-resistant cells from competitive inhibition, but adaptive therapy maintains some frequency of drug-sensitive cells that limit the growth of resistant cells through biotic competition. While early clinical trials of such a strategy have shown promise, optimisation of adaptive therapy is still a subject of active study. In this context, current methods largely assume cell phenotypes to remain constant, even though cell-state transitions could allow drug-sensitive and -resistant phenotypes to interchange and thus escape therapy. Here, we address this gap in the literature using a deterministic model of population growth in which sensitive and resistant cells grow under competition as well as cell-state transitions. Based on the model’s steady-state behaviour and temporal dynamics, we identify distinct balances of competition and cell-state transitions that are suitable for effective adaptive versus constant dose therapy. Our data indicate that under adaptive therapy, the frequency of fluctuations varies systematically across models with different levels of competition and cell-state transitions . Our analyses also identify key limitations of applying phenomenological models in clinical practice for therapy design and implementation, particularly when cell-state transitions are involved. These findings provide an overall perspective on the relevance of phenotypic plasticity for emerging cancer treatment strategies using population dynamics as a investigation framework.
Micronenvironmental signals play an important role in controlling cell
phenotypes. A complex network of interactions, both between a cell and
its environment as well as within a cell, govern cell development
processes. Mathematical models can contribute to our understanding of
such processes and support experimental design by generating testable
hypotheses. In this talk, I highlight the use of rather abstract,
so-called logical models that capture systems' mechanisms and behavior
only qualitatively. They still can be highly effective in capturing how
extracellular signals influence cell phenotypes and in identifying
intervention targets in the context of therapeutic strategies. Using an
established model of a cellular network underlying
epithelial-to-mesenchymal transition (EMT) focusing on degrees of cell
adhesion, we can assess how microenvironmental signals shape phenotype
plasticity along the EMT continuum. Going further, we can identify drug
intervention targets by employing methods from control theory. The focus
here is to identify key regulators the manipulation of which forces the
systems to adopt a chosen phenotype, or, as often of interest in a
therapeutic context, avoid a specific phenotype. Altogether, the
mathematical analysis suggests a number of hypotheses for targeted
experimental research.
Phenotypic adaptation, the ability of cells to change phenotype in response to external pressures, has been identified as a driver of drug resistance in cancer. To quantify phenotypic adaptation in BRAFV600E-mutant melanoma, we develop a theoretical model that emerges from data analysis of WM239A-BRAFV600E cell growth rates in response to drug challenge with the BRAF-inhibitor encorafenib. Our model constitutes a cell population model in which each cell is individually described by one of multiple discrete and plastic phenotype states that are directly linked to drug-dependent net growth rates and, by extension, drug resistance. Data-matched simulations reveal that phenotypic adaptation in the cells is directed towards states of high net growth rates, which enables evasion of drug-effects. The model subsequently provides an explanation for when and why intermittent treatments outperform continuous treatments in vitro, and demonstrates the benefits of not only targeting, but also leveraging, phenotypic adaptation in treatment protocols.
Early cutaneous squamous cell carcinomas (cSCCs) generally show epithelial differentiation features and good prognosis, whereas advanced cSCCs present mesenchymal traits associated with tumor relapse, metastasis, and poor survival. The mechanisms involved in cSCC progression are unclear, and the established markers are suboptimal for accurately predicting the clinical course of the disease. Immunotherapy based on immune checkpoint inhibitors (ICIs) was recently approved for the treatment of advanced cSCCs. However, half of cSCC patients remain unresponsive or develop resistance. We show that during cSCC progression, cancer cells acquire epithelial/mesenchymal plasticity (EMP). The emergence of epithelial plastic cancer cells expressing integrin αV (ITGAV) promotes cSCC progression to a mesenchymal state. Consistently, ITGAV expression allows the identification of patients at risk of cSCC relapses above the currently employed clinical histopathological parameters. The activation of insulin-like growth factor-1 receptor (IGF1R) pathway in epithelial cancer cells is necessary to induce EMP and mesenchymal state acquisition in response to tumor microenvironment-derived factors, while promoting ITGAV expression. Likewise, ITGAV knockdown in epithelial plastic cancer cells blocks EMP, generating epithelial tumors. Finally, cancer cells change their IC ligand profile according to their features, which dictates the IC pathways involved in immune evasion. Mouse epithelial cancer cells, through the PD-1/PD-L1 pathway, and mesenchymal cancer cells, through the CTLA-4/CD80 and TIGIT/CD155 pathways, differentially block antitumor immune responses and determine the response to ICI strategies. Analysis of E-cadherin/Vimentin/CD80/CD155 proteins in cSCC, HNSCC and melanoma patient samples shows their value as predictive biomarkers of anti-PD-1 response. Collectively, our findings suggest that the selection of ICIs therapies should be based on the epithelial/mesenchymal features of cancer cells.
Authors: Frederick J.H. Whiting, Maximilian Mossner, Calum Gabbutt, Christopher Kimberley, Chris P Barnes, Ann-Marie Baker, Andrea Sottoriva, Richard A. Nichols, Trevor A Graham
Effective cancer treatment frequently fails due to the evolution of drug resistant cell phenotypes caused by underlying genetic or non-genetic changes. The origin of these adaptations, their timing and rate of spread is key information for distinguishing the mechanism(s) of drug resistance, yet the dynamics cannot be observed directly. Here, we construct a mathematical framework to infer the dynamics of drug resistance without the need for direct measurement of the resistance phenotype using only genetic lineage tracing and population size data. The veracity of the framework is demonstrated through experimental evolution to 5-Fu chemotherapy in two common colorectal cancer cell lines: SW620 and HCT116. In SW620 cells, a stable pre-existing resistant subpopulation was inferred. In HCT116 cells resistance emerged through phenotypic switching into a slow growing resistant state with stochastic exiting into a fully resistant phenotype. Extensive functional assays, including scRNA-seq and scDNA-seq confirmed these distinct evolutionary routes and their molecular nature. Our mathematical framework can be extended to diverse experimental designs to infer the evolutionary dynamics of cancer cell therapy resistance evolution from readily obtained experimental data, enabling more rapid characterisation of resistance mechanisms.
Authors: Chen M. Chen, Rosemary Yu
Plasticity is the potential for cells or cell populations to change their phenotypes and behaviours in response to internal or external cues. Plasticity is fundamental to many complex biological processes, yet to date there remains a lack of mathematical models that can elucidate and predict molecular behaviours in a plasticity programme. Here we report a new mathematical framework that models cell plasticity as a multi-step completion process, where the system moves from the initial state along a path guided by multiple intermediate attractors until the final state (i.e. a new homeostasis) is reached. Using omics time-series data as model input, we show that our method fits data well, and identifies attractor states by their timing and molecular markers which are well-aligned with domain knowledge. Importantly, our model can make non-trivial predictions such as the molecular outcomes of blocking a plasticity programme from reaching completion, in a quantitative and time-resolved manner.