Patient-derived scaffolds and analyses of adapting cancer cell lines can monitor malignant properties of a cell-free cancer microenvironment highlighting distinct links between scaffold influences and clinical aggressiveness in cancer. The protein composition of the cell-free cancer microenvironments influencing adapting cancer cells have been defined by quantitative mass spectrometry and results indicate clear clustering of PDS differing in extracellular matrix related proteins as well as immunoregulatory and metabolic regulators. Interestingly both the defined clusters as well as individual imprinted proteins in the cell-free scaffolds are linked to clinical behaviours of the cancer and data from breast cancer and colorectal cancer will be presented. The multitude of proteins imprinted in the cell-free cancer scaffolds representing various tumour biological activities and cell types, supports the importance of the cancer microenvironment in influencing varying disease behaviours. The in vivo identified proteins can be used for improved disease subtyping, cancer drug targeting and importantly to construct optimal synthetic 3D-models that can be used for human-like drug discovery and validation of novel cancer treatments. Data from the first prototypes of human-mimicking synthetic scaffolds will be presented and discussed in relation to other 3D-based growth models available for cancer discovery.

Blood vessels are fundamental to life and have critical roles in many diseases, such as chronic inflammation, diabetes and cancer. We previously developed self-organizing 3D human blood vessel organoids from human pluripotent stem cells (hPSCs) that exhibit morphological, functional and molecular features of the human microvasculature. When exposed to diabetic conditions in vitro and in vivo, blood vessel organoids display prominent microvascular changes as observed in diabetic patients, whereas common diabetic mouse models fail to do so. Thus, human blood vessel organoids provide a novel window into the mechanism of diabetic vasculopathy, which is critical for the development of novel therapeutics. We are currently developing and applying immunocompetent human vascular models for preclinical drug testing of molecule efficacy and safety. Self-assembled 3D blood vessels on-a-chip mimic physiological immune adhesion to the capillary wall and extravasation into the tissue as well as molecule transport through an endothelial barrier in healthy an inflamed conditions. The increased usage and validation (such as clinical back translation) of such human in vitro models in pharma aims to improve the translation of pre-clinical research as well as the acceptance of in vitro data by regulatory authorities and thus to ultimately decrease the number of animal experiments.

Organoids are a complex 3D model able to encompass the majority of the heterogeneity, the phenotypic and genotypic features of the original patient sample they are derived from. A platform has been setup in AstraZeneca to derive patient derived organoids (PDO) and patient derived xenograft organoids (PDXO). PDOs are derived from cancerous and matched normal lung resections from the Royal Papworth Hospital. These models will be shared with the Wellcome Sanger Institute to perform whole genome and RNA sequencing. Different organoids were used to set up all the protocols. The metabolic fingerprinting illustrated the metabolic differences between organoids and different tissues. This confirmed the heterogeneity of metabolite composition of each organoids in a same model. A screening pipeline has been established with two different analysis methods. An engineering protocol to genetically modify these organoids was created with a success rate of 60 to 80%. A live imaging protocol was set up to visualise the CAR-T or Antibody-drug conjugate (ADC) activity in co-culture with organoids. To mimic the tumour micro-environment (TME), embedded T-regs with the organoids were able to reduce the activity of CAR-Ts. Organoids are a relevant model to mimic the patient context and would be more efficient to increase the number of successful molecules in clinical assays. [Etienne De Braekeleer1, Sarah Consonni1, Romain Lara1, Lewis Chaytor2, Gemma Everitt3, Monica Patel3, Gregory Hamm6, Chen Qian4, James Pilling5, Stewart Jones6, Matthew Garnett7, Doris Rassl9, Robert Rintoul8, Jonathan Orme3, Ultan McDermott9, Benjamin Taylor1 AstraZeneca : (Discovery Sciences (1 Cell Biology & Immunology, 2 Cell Enginneering, 3 Cell Bank, 4 Imaging IT, 5 Functional Genomics), CPSS (6 Imaging Science) 9 Oncology Bioscience, Oncology TTD), 7 Wellcome Sanger Institute, 8 Royal Papworth Hospital]

Advanced 3D Cellular Models

The dynamic and complex crosstalk between target cells and their microenvironment modulates physiological functions, pathological processes, and response to therapy. This crosstalk is mediated by direct cell-cell contact and indirect interactions via secreted soluble mediators, extracellular membrane components and modulators, extracellular vesicles. The difficulties in recapitulating these heterogeneous microenvironments with human cell models, with mature cell functionality and without the confounding effects of heterologous extracellular matrices, are a bottleneck in deciphering the crosstalk mechanisms and predicting their effect on therapeutic efficacy. To overcome these limitations, our team develops 3D cell models of disease, applying advanced cell culture approaches (3D culture, co-culture, cell immobilization) and systems (perfusion, bioreactors) to human pluripotent stem cells, other patient-derived cells, and human cell lines. By integrating cell biology, biochemical, imaging, transcriptomics, and proteomics approaches, we can depict the dynamic modulation of a specific cell microenvironment over time or in response to a therapeutic challenge. Recent advances in innate immune-competent 3D cell models of the central nervous system and carcinomas, and their application in the recapitulation of pathological disease features and utility to address the influence of the immune microenvironment in response to advanced therapeutics, such as gene therapy vectors and biologicals.

In this talk I will give an overview of the computational steps in the analysis of a single cell-based large-scale microscopy experiments. First, I will present a novel microscopic image correction method designed to eliminate illumination and uneven background effects which, left uncorrected, corrupt intensity-based measurements. New single-cell image segmentation methods will be presented using differential geometry, energy minimization and deep learning methods (www.nucleAIzer.org) (Hollandi et al. 2022). I will discuss the Advanced Cell Classifier (ACC) (www.cellclassifier.org), a machine learning software tool capable of identifying cellular phenotypes based on features extracted from the image. It provides an interface for a user to efficiently train machine learning methods to predict various phenotypes. For cases where discrete cell-based decisions are not suitable, we propose a method to use multi-parametric regression to analyze continuous biological phenomena. To improve the learning speed and accuracy, we propose an active learning scheme that selects the most informative cell samples. Our recently developed single-cell isolation methods, based on laser-microcapturing and patch clamping, utilize the selection and extraction of specific cell(s) using the above machine learning models (Brasko et al. 2018). I will show that we successfully performed DNA and RNA sequencing, proteomics, lipidomics and targeted electrophysiology measurements on the selected cells (Mund et al. 2022).

Light-sheet fluorescence microscopy (LSFM) provides low out-of-plane photobleaching and phototoxicity, but usually requires two microscope objective lenses orientated at 90? to one another - one for fluorescence excitation and one for fluorescence detection - making it harder to image samples prepared using conventional mounting methods. Oblique plane microscopy (OPM) is a type of LSFM that has been developed in our laboratory and uses a single high numerical aperture microscope objective to provide both fluorescence excitation and detection whilst maintaining the advantages of LSFM, enabling it to provide high-speed 3D imaging for a range of applications on a conventional fluorescence microscope frame. The speed of OPM imaging can be applied to image a single sample at video volumetric imaging rates. It can also be used to enable higher throughput and time-lapse 3D imaging of arrays of samples arrayed in multi-well plates. This talk will present examples of the application of OPM for high-speed 3D imaging of isolated cardiomyocytes and also examples where the system is being applied to study arrays of multicellular spheroids and organoids in 3D over multiple conditions and over time.

Organoids generated from human pluripotent stem cells provide experimental systems to study development and disease, but we lack quantitative measurements across different spatial scales and molecular modalities. Here we use a single-cell multimodal approach to reconstruct spatial protein maps over a retinal organoid time course and primary adult human retinal development. We develop a toolkit to visualize progenitor and neuron location, the spatial arrangements of extracellular and subcellular components, and global patterning in each organoid and primary tissue. In addition, we generate a single-cell transcriptome and chromatin accessibility time course dataset and infer a gene regulatory network underlying organoid development. We integrate genomic data with spatially segmented nuclei into a multi-modal atlas to explore organoid patterning and retinal ganglion cell (RGC) spatial neighborhoods, highlighting pathways involved in RGC cell death and show that mosaic genetic perturbations in retinal organoids provide insight into cell fate regulation.

There is an urgent need for scalable microphysiological systems (MPS) that can better predict drug efficacy and toxicity at the preclinical drug screening stage. An ideal preclinical model that can accurately predict clinical response does not exist however the use of MPS can help bridge the translational gap by providing more accurate representation of human susceptibility to drug response. Mera is an automated, modular and scalable system for culturing and assaying microtissues with interconnected fluidics, inbuilt environmental control and automated image capture. The system presented has microfluidic flow control and multiple possible fluidics modes, the primary mode allowing cells to be matured into a desired microtissue type and the secondary mode where the fluid flow can be re-orientated to create a body-on-a-plate-format with recirculating circuits composed of inter-connected channels to allow microtissue communication. We present data demonstrating the prototype system Mera using an Acetaminophen/HepG2 liver microtissue toxicity assay with Calcein AM (CalAM) and Ethidium Homodimer (EtHD1) viability stains. The prototype microtissue culture plate wells are laid out in a 3 x 3 or 4 x 10 grid format with viability (multi-organ models) and toxicity assays demonstrated in both formats. We present the groundwork for the Mera system to be used as a viable option for scalable microtissue culture and assay development for preclinical drug development.

Traditional two-dimensional (2D) cancer cell culture has been used extensively to evaluate the efficacy of drug delivery systems, but it has limitations that can deviate results from a real tumour mass in vivo. The lack of accurate cellular interactions and extracellular matrix (ECM) components are the primary differences between 2D cancer models and an actual tumour in vivo. 2D cell culture also lose tumour heterogeneity. The results obtained from 2D models do not represent the complexity of drug delivery and diffusion that occurs in vivo. In recent years, three-dimensional (3D) tumour models have gained popularity as they can overcome the limitations of 2D models and are physiologically comparable to a real tumour mass. 3D models possess ECM components which resembles in vivo tumours. In 3D culture, drug penetration and distribution are more representative of the tumour microenvironment, allowing us to investigate the efficacy and toxicity of a drug in a more clinically relevant model. Together, this makes 3D culture models suitable for drug delivery research. Phage Cancer Therapy group has employed 3D tumour models to evaluate the efficacy of a gene delivery platform derived from a harmless filamentous bacteriophage (Phage); Transmorphic Phage/AAV (TPA). The TPA particle has many advantages over other gene delivery vectors including low production cost, very low toxicity to normal cells, can be targetable to specific cell type, and does not require low temperature storage. Phage Cancer Therapy group has developed TPA as a multifunctional cancer cell targeted nanocarrier to guide the delivery of therapeutic genes (or drugs) to cancer cells. Capsids of TPA particles were engineered to equip cancer targeting ligand (RGD4C) on the capsids with a therapeutic transgene cassette inside the capsids. The RGD4C ligand specifically binds to ?v?3 and ?v?5, which are specifically overexpressed on tumour cell surface. It is subsequently up taken by, and delivers the therapeutic gene to, cancer cells. We adopted 3D tumour models of various cancer cell types as models to evaluate the efficacy of the novel TPA nanocarrier. We found that cancer stem cell population in our 3D tumour sphere models resemble tumours in vivo. Our novel TPA particle demonstrates promising gene delivery efficiency across many types of 3D cancer models. TPA efficiently penetrated through the ECM and delivered a transgene to targeted cells, ultimately resulting in tumour regression. Furthermore, toxicity tests, in vitro and in vivo, showed that the TPA particle itself does not harm normal cells and organs. We confirmed the efficacy of TPA in an animal model by systemic administration of the particle through the teil vein. TPA can selectively deliver a transgene to tumour cells but spare other organs unharmed. Altogether, the TPA particle has the potential to be a powerful cargo for tumour-targeted gene delivery.