Biology subjects

Karlsson, J.

Publications and source records attributed to Karlsson, J..

12 recordsLinked to original sources

bioRxiv2025

Bayesian inference of cell-to-cell interactions in low-nutrient neuroblastoma cell cultures

Neuroblastoma is characterised by significant intratumoural heterogeneity which complicates treatments. Phenotypic plasticity, i.e., the ability of cells to alter their phenotype without genetic mutations, is a key factor contributing to this heterogeneity. In this study, we quantify cell actions and cell-to-cell interactions that lead to phenotypic adaptation in vitro under stress, specifically low-nutrient conditions. We initially record dynamic cell counts under various nutrient conditions for two neuroblastoma cell lines with known phenotype differences. Subsequently, we compile a list of plausible biological processes that could transpire within the aforementioned systems. We then construct candidate models employing mass-action kinetics. To quantitatively infer processes that lead to phenotypic adaptation, we perform a model selection based on computational Bayesian inference methods.Our results suggest that cell-to-cell interactions promote phenotypic adaptation and that the rate of this adaptation increases with decreasing nutrient concentrations. We perform flow cytometry-based experiments to corroborate changes in phenotype composition in response to both nutrient-deprived and treatmentinduced stress conditions.

cancer biology↗

bioRxiv2024

Spatial transcriptomics exploration of the primary neuroblastoma microenvironment unveils novel paracrine interactions

High-risk neuroblastomas exhibit a high degree of intratumoral heterogeneity. Single-cell RNA sequencing has greatly improved our understanding of these tumors, but the method lacks cellular tissue context and spatial information about local signaling dynamics. To address this gap, we profiled untreated and chemotherapy-treated high-risk neuroblastomas from 2 patients using spatial transcriptomics. We confirmed the transcriptional and cellular heterogeneous nature of the neuroblastoma microenvironment and identified several unique spatial niches and patterns. In one of the treated tumors, a spatially constrained cluster of undifferentiated and 11p-gained cancer cells was identified, surrounded by a rim of macrophages. A signaling interaction between the chemokine CCL18 and its receptor PITPNM3 was predicted between these cells and we experimentally demonstrated that CCL18 increases neuroblastoma cell migration. In the other tumor, we identified a stromal cluster with high transcriptional similarity to the adrenal cortex. These adrenocortical-like cells expressed the ALK ligand ALKAL2 and were predicted to communicate with neighboring ALK expressing cancer cells. We demonstrated a unique developmental pattern of adrenal medulla-specific expression of ALK and adrenocortical-specific expression of ALKAL2, suggesting a role of these signaling interactions in neuroblastoma carcinogenesis.

cancer biology↗

bioRxiv2024

Gene regulatory network inference from single-cell data using optimal transport

Modelling gene expression is a central problem in systems biology. Single-cell technologies have revolutionised the field by enabling sequencing at the resolution of individual cells. This results in a much richer data compared to what is obtained by bulk technologies, offering new possibilities and challenges for gene regulatory network inference. Here we introduce GRIT -- a method to fit a differential equation model and to infer gene regulatory networks from single-cell data using the theory of optimal transport. The idea consists in tracking the evolution of the cell distribution over time and finding the system that minimises the transport cost between consecutive time points.

bioinformatics↗

bioRxiv2023

Cell Painting-based bioactivity prediction boosts high-throughput screening hit-rates and compound diversity

Efficiently identifying bioactive compounds towards a target of interest remains a time- and resource-intensive task in early drug discovery. The ability to accurately predict bioactivity using morphological profiles has the potential to rationalize the process, enabling smaller screens of focused compound sets. Towards this goal, we explored the application of deep learning with Cell Painting, a high-content image-based assay, for compound bioactivity prediction in early drug screening. Combining Cell Painting data and unrefined single-concentration activity readouts from high-throughput screening (HTS) assays, we investigated to what degree morphological profiles could predict compound activity across a set of 140 unique assays. We evaluated the performance of our models across different target classes, assay technologies, and disease areas. The predictive performance of the models was high, with a tendency for better predictions on cell-based assays and kinase targets. The average ROC-AUC was 0.744 with 62% of assays reaching [≥]0.7, 30% reaching [≥]0.8 and 7% reaching [≥]0.9 average ROC-AUC, outperforming commonly used structure-based predictions in terms of predictive performance and compound structure diversity. In many cases, bioactivity prediction from Cell Painting data could be matched using brightfield images rather than multichannel fluorescence images. Experimental validation of our predictions in follow-up assays confirmed enrichment of active compounds. Our results suggest that models trained on Cell Painting data can predict compound activity in a range of high-throughput screening assays robustly, even with relatively noisy HTS assay data. With our approach, enriched screening sets with higher hit rates and higher hit diversity can be selected, which could reduce the size of HTS campaigns and enable primary screening with more complex assays.

bioinformatics↗

bioRxiv2022

Inherent evolutionary unpredictability in cancer model system

Despite the advent of personalized medicine, it is still difficult to predict how a cancer develops over time at the level of the individual patient or even in cancer model systems which begs the question whether certain aspects of cancer can ever be predicted or if there is an inherent unpredictability in cancer, similar to other complex biological systems, We demonstrate by a combination of agent-based mathematical modelling and analysis of data from patient-derived xenograft systems from multiple cancer types that certain conditions may invoke chaotic fluctuations in the clonal landscape of cancer cells. Our findings indicate that under those conditions, the cancer genome behaves as a complex dynamic system, making its long-term evolution inherently unpredictable.

evolutionary biology↗

bioRxiv2022

Clinically-relevant treatment of PDX models reveals patterns of neuroblastoma chemoresistance

Chemotherapy resistance and relapses are common in high-risk neuroblastoma (NB), an aggressive pediatric solid tumor of the sympathetic nervous system. Here, we developed a clinically-relevant in vivo treatment protocol mimicking the first line five-chemotherapy treatment regimen of high-risk NB and applied this protocol to mice with MYCN-amplified NB patient-derived xenografts (PDXs). Genomic and transcriptomic analyses were used to reveal the genetic and non-genetic mechanisms involved in NB chemoresistance. We observed convergent and parallel evolution of key NB genetic aberrations over time. Intrinsic resistance to chemotherapy was associated with high genetic diversity and an embryonic phenotype. Relapsed NB PDX tumors with acquired resistance showed an immature mesenchymal-like phenotype resembling multipotent Schwann cell precursors that are found in the adrenal gland. NBs with a successful treatment response presented a lineage-committed adrenergic phenotype similar to normal neuroblasts, reduced cell cycle gene expression, and negative regulation of the mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) cascade. NB organoids established from relapsed PDX tumors retained drug resistance, tumorigenicity, and transcriptional cell states ex vivo. This work sheds light on mechanisms involved in NB chemotherapy response in vivo and ex vivo using a clinically-relevant protocol, and emphasizes the importance of transcriptional cell states in treatment response. Detailed characterization of resistance mechanisms is essential for the development of novel treatment strategies in non-responsive or relapsed high-risk NB. One Sentence SummaryCOJEC chemotherapy treatment of neuroblastoma PDX models uncovers patterns of transcriptional plasticity and chemoresistance.

cancer biology↗

bioRxiv2021

Epigenetic therapy to enhance therapeutic effects of PD-1 inhibition in uveal melanoma

Targeted therapy and immunotherapy have revolutionized the treatment of metastatic skin melanoma but none of the treatments are approved for patients with metastatic uveal melanoma (UM). Here we hypothesized that the poor responses to immunotherapy of UM can be enhanced by epigenetic modulation using HDAC or BET inhibitors (BETi). Cultured uveal melanoma cells were treated with the HDAC inhibitor (HDACi) entinostat or BETi JQ1. Entinostat induced HLA expression and PD-L1, but JQ1 did not. A syngenic mouse model carrying B16-F10 melanoma cells were treated with PD-1 and CTLA-4 inhibitors, which was curative. Co-treatment with the bioavailable BETi iBET-726 impaired the immunotherapy effect. Monotherapy of a B16-F10 mouse model with anti-PD-1 resulted in a moderate therapeutic effect that could be enhanced by entinostat. Mice carrying PD-L1 knockout B16-F10 cells were also sensitive to entinostat. This suggests HDAC inhibition and immunotherapy could work in concert. Indeed, co-cultures of UM with HLA-matched melanoma-specific tumor-infiltrating lymphocytes (TILs) resulted in higher TIL-mediated melanoma killing when entinostat was added. Further exploration of combined immunotherapy and epigenetic therapy in metastatic UM is warranted.

cancer biology↗

bioRxiv2020

Convolutional neural networks for automated CMR image segmentation in rats with myocardial infarcts

In translational cardiovascular research, delineation of left ventricle (LV) in magnetic resonance images is a crucial step in assessing hearts function. Performed manually, this task is time-consuming and prone to inter- and intra-reader variability. Here we report first AI-based tool for segmentation of rat cardiovascular MRI. The method is an ensemble of fully convolutional networks and can quantify clinically relevant measures: end-diastolic volume (EDV), end-systolic volume (ESV), and ejection fraction (EF) automatically. Overall, our method reaches Dice score of 0.93 on the independent test set. The mean absolute difference of segmented volumes between automated and manual segmentation is 22.5L for EDV, 13.6L for ESV, and for EF 2.9%. Our work demonstrates the value of AI in development of tools that will significantly reduce time spent on repetitive work and result in increased efficiency of reporting data to project teams.

bioengineering↗

bioRxiv2020

Tracing the evolution of aneuploid cancers from multiregional sequencing

Clonal deconvolution of mutational landscapes is crucial to understand the evolutionary dynamics of cancer. Two limiting factors for clonal deconvolution that have remained unresolved are variation in tumor cell fraction (TCF) and chromosomal copy number across different samples of the same tumor. We developed a semi-supervised algorithm that tracks variant calls through multi-sample spatiotemporal tumor data. While normalizing allele frequencies based on TCF it also adjusts for copy number changes at clonal deconvolution. Absent a priori copy number data, it renders in silico copy number estimations from bulk sequences. Using published and simulated tumor sequences, we reliably segregated clonal/subclonal variants even at low sequencing depth ([~]50x). Given at least one pure tumor sample (>70% TCF), we could normalize and deconvolve paired samples down to a TCF of 40%. This renders a reliable clonal reconstruction well adapted to multi-regionally sampled solid tumor, which are often aneuploid and contaminated by non-cancer cells.

bioinformatics↗

bioRxiv2020

Deep learning models for lipid-nanoparticle-based drug delivery

Large-scale time-lapse microscopy experiments are useful to understand delivery and expression in RNA-based therapeutics. The resulting data has high dimensionality and high (but sparse) information content, making it challenging and costly to store and process. Early prediction of experimental outcome enables intelligent data management and decision making. We start from time-lapse data of HepG2 cells exposed to lipid-nanoparticles loaded with mRNA for expression of green fluorescent protein (GFP). We hypothesize that it is possible to predict if a cell will express GFP or not based on cell morphology at time-points prior to GFP expression. Here we present results on per-cell classification (GFP expression/no GFP expression) and regression (level of GFP expression) using three different approaches. In the first approach we use a convolutional neural network extracting per-cell features at each time point. We then utilize the same features combined with: a long-short-term memory (LSTM) network encoding temporal dynamics (approach 2); and time-series feature extraction using the python package tsfresh followed by principal component analysis and gradient boosting machines (approach 3), to reach a final classification or regression result. Application of the three approaches to a previously unanalyzed test set of cells showed good predictive performance of all three approaches but that accounting for the temporal dynamics via LSTMs or tsfresh led to significantly improved performance. The predictions made by the LSTM and tsfresh applications were not significantly different. The results highlight the benefit of accounting for temporal dynamics when studying drug delivery using high content imaging.

cell biology↗

bioRxiv2019

Molecular profiling of driver events and tumor-infiltrating lymphocytes in metastatic uveal melanoma

Uveal melanoma (UM) is a rare form of melanoma with a genetics and immunology that is different from skin melanoma. Previous studies have identified genetic driver events of early stage disease when the tumor is confined to the eye. However due to lack of a clinical rationale to biopsy metastatic disease, access to tumor material to perform molecular profiling of metastases has been limited. In this study, we have characterized genomic events in UM metastases using whole-genome sequencing of fresh frozen biopsies from thirty-two patients and profiled the transcriptomes of individual tumor infiltrating lymphocytes in eight patients by single-cell sequencing. We find that 91% of the patients have metastases carrying inactivating events in the tumor suppressor BAP1 and this coincided with somatic alterations in GNAQ, GNA11, CYSLTR2, PLCB4, SF3B1 and/or CDKN2A. Mutational signature analysis revealed a rare subset of tumors with prominent signs of UV damage, associated with outlier mutational burden. We study copy number variations (CNV) and find overrepresented events, some of which were not altered in matched primary eye tumors. A focused siRNA screen identified functionally significant genes of some of the segments recurrently gained. We reintroduced a functional copy of BAP1 into a patient-derived BAP1 deficient tumor cell line and found broad transcriptomic changes of genes associated with subtype distinction and prognosis in primary UM. Lastly, our analysis of the immune microenvironments of metastases revealed a presence of tumor-reactive T cells. However, a majority expressed the immune checkpoint receptors TIM-3, LAG3 and TIGIT, and to a lesser extent PD-1. These results provide an updated view of genomic events represented in metastatic UM and immune interactions in advanced lesions.

cancer biology↗

bioRxiv2019

A regulatory RNA is associated to invasive meningococcal disease in Europe

The strictly human pathogen Neisseria meningitidis is a commensal bacterium but can occasionally turn lethal causing septicaemia and meningitis. The mechanisms of how the meningococcus shifts to invasive infection remain poorly understood. Here we demonstrate that an eight base-pair tandem repeat deletion in the 5-untranslated region of the polysaccharide capsular biosynthesis operon results in a hypercapsulation phenotype in clinical isolates. The increased capsule production significantly improves the bacterium survival in human serum while impairing its ability to adhere and colonise human pharyngeal cells. Among 4501 reported meningococcal cases in Europe from 2010-2018, the loss of an eight base-pair tandem repeat is three times more prevalent in invasive isolates (16.3%) compared to carrier isolates (5.1%).Combined results indicate that polymorphisms in this regulatory RNA contributes to meningococcal virulence.\n\nImportanceIn this study we report a regulatory RNA to be directly involved in clinical manifestation of meningococcal disease. Using readily accessible WGS of meningococcus, we have now demonstrated that regulatory RNAs directly contribute to the progression of invasive meningococcal infection. We believe this novel combination of molecular and comparative regulatory RNA study could be used for the identification of additional RNAs involved in not only meningococcus but also pave the way for similar studies in other important bacterial pathogens. The identification of specific regulatory RNAs will no doubt facilitate clinicians, microbiologists, and public health practitioners to adjust their diagnostic techniques and treatments to best fit the condition of the patients.

microbiology↗