Biology subjects

Bozek, K.

Publications and source records attributed to Bozek, K..

13 recordsLinked to original sources

bioRxiv2025

Single recipient cell tracking of tellurium-labeled extracellular vesicle proteomes (TeLEV) identifies EV-driven immunomodulation

Extracellular vesicles (EVs) mediate tumor-immune cell communication by carrying protein cargo that can immediately modulate signaling and antigen presentation. Yet mapping the uptake of primary EV proteomes by human immune cells at single-cell resolution has been constrained by a lack of labeling strategies. We show here that TeLEV, a tellurium-based metabolic mass tagging approach that incorporates L-2-tellurienylalanine (TePhe) into EV proteomes, can produce a biologically rare monoisotopic signal, which is detectable by mass cytometry, imaging mass cytometry, and nanoscale SIMS, without perturbing EV morphology, yield, or proteome composition. We applied TeLEV to label primary malignant B-cell-derived EVs (MBC-EVs) from chronic lymphocytic leukemia (CLL) patients and could follow EV uptake by distinct cell populations of healthy donor peripheral blood mononuclear cells. MBC-EV uptake occurred predominantly in cells of myeloid lineages. In direct control experiments with matched secreted proteins, a machine learning approach identified CD123, CD127, and CD25 as key markers distinguishing primary MBC-EV recipients from matched secreted protein recipient cells. Nanoscale imaging enabled localization of EV-delivered proteins within heterochromatin, whereas Te-labeled secreted proteins accumulated in the cytoplasm of recipient cells. We then generated a pan-immune EV uptake atlas by tracing the uptake of primary and cell-line EVs from nine cell lines and six donors with chronic lymphocytic leukemia into 2,977,094 recipient cells across 43 cell types and subpopulations. We found that the uptake of MBC-EVs by myeloid recipients induced monocyte-derived dendritic-cell polarization characterized by the co-expression of the interleukin-receptor triad (IL-RT: CD123, CD127, CD25) identified above. Time-resolved EV uptake analysis showed a rapid, transient expression of CD123/CD127 followed by CD25, both tightly coupled to MBC-EV uptake by myeloid cells. The intensity of IL-RT expression correlated with that of PD-L1 and BCL-2. Using different STAT degraders to bidirectionally modify the EV-induced STAT5 signal, we observed that MBC-EV uptake and IL-RT, PD-L1, and BCL-2 expression increased with STAT3 degradation and decreased with STAT5 degradation. To investigate the functional consequences of the MBC-EV-induced changes, we showed that MBC-EVs in the presence of IL-2 induced a high-CD25 immune state with low cytotoxic and high B cell proliferation. Taken together, TeLEV represents a novel tool for single-cell tracking of EV proteomes, revealing STAT5-dependent immune remodeling of recipient cells.

cancer biology↗

bioRxiv2025

Unsupervised Representation Learning of C. elegans Poses and Behavior Sequences From Microscope Video Recordings

Caenorhabditis elegans (C. elegans) is an important model system for studying molecular mechanisms in disease and aging. The nematode can be imaged in highly parallel phenotypic screens resulting in large volumes of video data of the moving worm. However converting the rich, pixel-encoded phenotypical information into meaningful, quantitative description of behavior is a challenging task. There is a range of methods for quantification of the simple body shape of C. elegans and the features of its motion. These methods however are often multi-step and fail in the case of highly coiled and self-overlapping worms. Motivated by the recent development of self-supervised deep learning methods in computer vision and natural language processing, we propose an unbiased, label-free approach to quantify worm pose and motion from video data directly. We represent worm posture and behavior as embedding vectors and visualize them in a unified embeddings space. We observe that the vector embeddings capture meaningful features describing worm shape and motion, such as the degree of body bend or the speed of movement. Importantly, using pixel values directly as input, our method captures coiled worm behaviors which are inaccessible to methods based on keypoint tracking or skeletonization. While our work focuses on C. elegans, the ability to quantify behavior directly from video data opens possibilities to study organisms without rigid skeletons whose behavior is difficult to quantify using keypoint-based approaches.

bioinformatics↗

bioRxiv2025

ATR and PKMYT1 inhibition re-sensitize a subset of TNBC patient-derived models to carboplatin inducing mitotic catastrophe

Triple negative breast cancer (TNBC) is associated with poor prognosis and is mainly treated with chemotherapy-based regimens, often including carboplatin. Resistance to carboplatin is a common clinical issue that is either initially present or develops with treatment. Overcoming this resistance is a significant clinical challenge, which highlights the need for novel therapeutic strategies. We used a pooled shRNA screening approach with a chemoresistant TNBC patient-derived xenograft (PDX) cell (PDXC) line to identify targets whose knockdown would enhance the efficacy of carboplatin. This screening led to the identification of the ATR (ataxia telangiectasia and Rad3-related) gene as a key therapeutic vulnerability. Inhibiting ATR with BAY1895344 or AZD6738 re-sensitized carboplatin-resistant PDXCs and PDXs to carboplatin, resulting in an increase in DNA damage, and apoptosis. ATR inhibition disrupts the dependence of carboplatin-resistant cells on the S and G2/M checkpoints for DNA repair, leading to mitotic catastrophe. We further found that the addition of ATR inhibitors to carboplatin reversed a FOXM1-targeted gene program enabling premature passage into mitosis. Moreover, targeting PKMYT1, a regulator of cyclin-dependent kinase 1 (CDK1) controlling the G2/M checkpoint, through knockdown or with the novel PKMYT1 inhibitor RP-6306, also enhanced carboplatin efficacy in our TNBC PDXC. Molecular factors associated with response to the ATR inhibitor/carboplatin combination included low RNA levels of PKMYT1. These results underscore the pivotal roles of ATR and PKMYT1 in mediating resistance to carboplatin in TNBC and support targeting these pathways to overcome carboplatin resistance in this disease.

cancer biology↗

bioRxiv2024

Deep learning-based interpretable prediction of recurrence of diffuse large B-cell lymphoma

BackgroundThe heterogeneous and aggressive nature of diffuse large B-cell lymphoma (DLBCL) presents significant treatment challenges as up to 50% of patients experience recurrence of disease after chemotherapy. Upfront detection of recurring patients could offer alternative treatments. Deep learning has shown potential in predicting recurrence of various cancer types but suffers from lack of interpretability. Particularly in prediction of recurrence, an understanding of the models decision could eventually result in novel treatments. MethodsWe developed a deep learning-based pipeline to predict recurrence of DLBCL based on histological images of a publicly available cohort. We utilized attention-based classification to highlight areas within the images that were of high relevance for the models classification. Subsequently, we segmented the nuclei within these areas, calculated morphological features, and statistically analyzed them to find differences between recurred and non-recurred patients. ResultsWe achieved an f1 score of 0.83 indicating that our model can distinguish non-recurred from recurred patients. Additionally, we found that features that are the most predictive of recurrence include large and irregularly shaped tumor cell nuclei. DiscussionOur work underlines the value of histological images in predicting treatment outcomes and enhances our understanding of complex biological processes in aggressive, heterogeneous cancers like DLBCL.

bioinformatics↗

bioRxiv2024

Deep Imputation for Skeleton Data (DISK) for Behavioral Science

AbstractPose estimation methods and motion capture systems have opened doors to quan- titative measurements of animal kinematics. However, these methods are not perfect and contain missing data. Our method, Deep Imputation for Skeleton data (DISK), leverages deep learning algorithms to learn dependencies between keypoints and their dynamics to impute missing tracking data. We developed an unsupervised training scheme, which does not rely on manual annotations, and tested several neural network architectures for the imputation task. We found that transformer outperforms other architectures including graph con- volutional networks that were developed specifically for skeleton-based action recognition. We demonstrate the usability and performance of our imputation method on seven different animal skeletons including two multi-animal set-ups. With an optional estimated imputation error, DISK enables behavior scien- tists to assess the reliability of the imputed data. The imputed recordings allow to detect more episodes of motion, such as steps, and to obtain more sta- tistically robust results when comparing these episodes between experimental conditions. While animal behavior experiments are expensive and complex, track- ing errors make sometimes large portions of the experimental data unusable. DISK allows for filling in the missing information and for taking full advantage of the rich behavioral data. This stand-alone imputation package, freely available at https://github.com/bozeklab/DISK.git, is applicable to results of any track- ing method (marker-based or markerless) and allows for any type of downstream analysis.

animal behavior and cognition↗

bioRxiv2023

WormSwin: Instance segmentation of C. elegans using vision transformer

The possibility to extract motion of a single organism from video recordings at a large-scale provides means for the quantitative study of its behavior, both individual and collective. This task is particularly difficult for organisms that interact with one another, overlap, and occlude parts of their bodies in the recording. Here we propose WormSwin - an approach to extract single animal postures of Caenorhabditis elegans (C. elegans) from recordings of many organisms in a single microscope well. Based on transformer neural network architecture our method segments individual worms across a range of videos and images generated in different labs. Our solutions offers accuracy of 0.990 average precision (AP0.50) and comparable results on the benchmark image dataset BBBC010. Finally, it allows to segment challenging overlapping postures of mating worms with an accuracy sufficient to track the organisms with a simple tracking heuristic. An accurate and efficient method for C. elegans segmentation opens up new opportunities for studying of its behaviors previously inaccessible due to the difficulty in the worm extraction from the video frames.

bioinformatics↗

bioRxiv2022

Predicting the HER2 status in esophageal cancer from tissue microarrays using convolutional neural networks

BackgroundFast and accurate diagnostics are key for personalized medicine. Particularly in cancer, precise diagnosis is a prerequisite for targeted therapies which can prolong lives. In this work we focus on the automatic identification of gastroesophageal adenocarcinoma (GEA) patients that qualify for a personalized therapy targeting epidermal growth factor receptor 2 (HER2). We present a deep learning method for scoring microscopy images of GEA for the presence of HER2 overexpression. MethodsOur method is based on convolutional neural networks (CNNs) trained on a rich dataset of 1,602 patient samples and tested on an independent set of 307 patient samples. We incorporated an attention mechanism in the CNN architecture to identify the tissue regions in these patient cases which the network has detected as important for the prediction outcome. Our solution allows for direct automated detection of HER2 in immunohistochemistry-stained tissue slides without the need for manual assessment and additional costly in situ hybridization (ISH) tests. ResultsWe show accuracy of 0.94, precision of 0.97, and recall of 0.95. Importantly, our approach offers accurate predictions in cases that pathologists cannot resolve, requiring additional ISH testing. We confirmed our findings in an independent dataset collected in a different clinical center. ConclusionsWe demonstrate that our approach not only automates an important diagnostic process for GEA patients but also paves the way for the discovery of new morphological features that were previously unknown for GEA pathology.

bioinformatics↗

bioRxiv2021

Deep learning-based segmentation and quantification of podocyte foot process morphology

The kidneys constantly filter enormous amounts of fluid, with almost complete retention of albumin and other macromolecules in the plasma. Diseases of podocytes at the kidney filtration barrier reduce the intrinsic permeability of the capillary wall resulting in albuminuria. However, direct quantitative assessment of the underlying morphological changes has previously not been possible. Here we developed a deep learning-based approach for segmentation of foot processes in images acquired with optical microscopy. Our method - Automatic Morphological Analysis of Podocytes (AMAP) - accurately segments foot processes and robustly quantifies their morphology. It also robustly determined morphometric parameters, at a Pearson correlation of r > 0.71 with a previously published semi-automated approach, across a large set of mouse tissue samples. The artificial intelligence algorithm wasWe applied the analysis to a set of human kidney disease conditions allowing comprehensive quantification of various underlying morphometric parameters. These results confirmed that when podocytes are injured, they take on a more simplified architecture and the slit diaphragm length is significantly shortened, resulting in a reduction in the filtration slit area and a loss of the buttress force of podocytes which increases the permeability of the glomerular basement membrane to albumin.

pathology↗

bioRxiv2020

Flow-Mediated Olfactory Communication in Honey Bee Swarms

Honey bee swarms are a landmark example of collective behavior. To become a coherent swarm, bees locate their queen by tracking her pheromones, but how can distant individuals exploit these chemical signals which decay rapidly in space and time? Here, we combine a novel behavioral assay with the machine vision detection of organism location and scenting behavior to track the search and aggregation dynamics of the honey bee Apis mellifera L. We find that bees collectively create a communication network to propagate pheromone signals, by arranging in a specific spatial distribution where there is a characteristic distance between individuals and a characteristic direction in which individuals broadcast the signals. To better understand such a flow-mediated directional communication strategy, we connect our experimental results to an agent-based model where virtual bees with simple, local behavioral rules, exist in a flow environment. Our model shows that increased directional bias leads to a more efficient aggregation process that avoids local equilibrium configurations of isotropic communication, such as small bee clusters that persist throughout the simulation. Our results highlight a novel example of extended classical stigmergy: rather than depositing static information in the environment, individual bees locally sense and globally manipulate the physical fields of chemical concentration and airflow.

animal behavior and cognition↗

bioRxiv2020

Markerless tracking of an entire insect colony

We present a comprehensive, computational method for tracking an entire colony of the honey bee Apis mellifera using high-resolution video on a natural honeycomb background. We adapt a convolutional neural network (CNN) segmentation architecture to automatically identify bee and brood cell positions, body orientations and within-cell states. We achieve high accuracy (~10% body width error in position, ~10{degrees} error in orientation, and true positive rate > 90%) and demonstrate months-long monitoring of sociometric colony fluctuations. We combine extracted positions with rich visual features of organism-centered images to track individuals over time and through challenging occluding events, recovering ~79% of bee trajectories from five observation hives over a span of 5 minutes. The resulting trajectories reveal important behaviors, including fast motion, comb-cell activity, and waggle dances. Our results provide new opportunities for the quantitative study of collective bee behavior and for advancing tracking techniques of crowded systems.

bioinformatics↗

bioRxiv2020

OrganoidTracker: efficient cell tracking using machine learning and manual error correction

Time-lapse microscopy is routinely used to follow cells within organoids, allowing direct study of division and differentiation patterns. There is an increasing interest in cell tracking in organoids, which makes it possible to study their growth and homeostasis at the single-cell level. As tracking these cells by hand is prohibitively time consuming, automation using a computer program is required. Unfortunately, organoids have a high cell density and fast cell movement, which makes automated cell tracking difficult. In this work, a semi-automated cell tracker has been developed. To detect the nuclei, we use a machine learning approach based on a convolutional neural network. To form cell trajectories, we link detections at different time points together using a min-cost flow solver. The tracker raises warnings for situations with likely errors. Rapid changes in nucleus volume and position are reported for manual review, as well as cases where nuclei divide, appear and disappear. When the warning system is adjusted such that virtually error-free lineage trees can be obtained, still less than 2% of all detected nuclei positions are marked for manual analysis. This provides an enormous speed boost over manual cell tracking, while still providing tracking data of the same quality as manual tracking.

bioinformatics↗

bioRxiv2019

Pixel personality for dense object tracking in a 2D honeybee hive

Tracking large numbers of densely-arranged, interacting objects is challenging due to occlusions and the resulting complexity of possible trajectory combinations, as well as the sparsity of relevant, labeled datasets. Here we describe a novel technique of collective tracking in the model environment of a 2D honeybee hive in which sample colonies consist of N [~] 103 highly similar individuals, tightly packed, and in rapid, irregular motion. Such a system offers universal challenges for multiobject tracking, while being conveniently accessible for image recording. We first apply an accurate, segmentation-based object detection method to build initial short trajectory segments by matching object configurations based on class, position and orientation. We then join these tracks into full single object trajectories by creating an object recognition model which is adaptively trained to recognize honeybee individuals through their visual appearance across multiple frames, an attribute we denote as pixel personality. Overall, we reconstruct [~] 46% of the trajectories in 5 min recordings from two different hives and over 71% of the tracks for at least 2 min. We provide validated trajectories spanning 3, 000 video frames of 876 unmarked moving bees in two distinct colonies in different locations and filmed with different pixel resolutions, which we expect to be useful in the further development of general-purpose tracking solutions.

animal behavior and cognition↗

bioRxiv2018

Parallel genomic evolution of parasite tolerance in wild honey bee populations

Sudden biotic pressures, such as those from novel diseases and pathogens, require populations to respond rapidly or face potential extinction. How this response process takes place remains poorly understood, particularly in natural environments. In this study we take advantage of unique decade-long data sets of two wild honey bee (Apis mellifera) populations in the United States to reconstruct the evolution of tolerance to a novel parasite, the ectoparasitic mite Varroa destructor. Upon the arrival of Varroa, the two geographically isolated populations simultaneously suffered massive Varroa-induced mortality, but stabilized within two years. Here we sequenced and phased genomes of 465 bees sampled from both populations annually over the decade that spanned Varroas arrival. Remarkably, we found that genetic changes were strongly correlated between the populations, suggesting parallel selective responses to Varroa parasitization. The arrival of Varroa was also correlated with an influx of genes of African origin into both populations, though surprisingly it did not substantially contribute to the overall similarity of the evolutionary response between them. Genes involved in metabolic, protein processing and developmental pathways were under particularly strong selection. It is possible that interactions among highly connected gene groups in these pathways may help channelize selective responses to novel parasites, causing completely unrelated populations to exhibit parallel evolutionary trajectories when faced with the same biotic pressure. Our analyses illustrate that ecologically relevant traits emerge from highly polygenic selection involving thousands of genes contributing to complex patterns of evolutionary change.

evolutionary biology↗