Keras R-CNN: library for cell detection in biological images using deep neural networks.

BACKGROUND: A common yet still manual task in basic biology research, high-throughput drug screening and digital pathology is identifying the number, location, and type of individual cells in images. Object detection methods can be useful for identifying individual cells as well as their phenotype in one step. State-of-the-art deep learning for object detection is poised to improve the accuracy and efficiency of biological image analysis. RESULTS: We created Keras R-CNN to bring leading computational research to the everyday practice of bioimage analysts. Keras R-CNN implements deep learning object detection techniques using Keras and Tensorflow ( https://github.com/broadinstitute/keras-rcnn ). We demonstrate the command line tool's simplified Application Programming Interface on two important biological problems, nucleus detection and malaria stage classification, and show its potential for identifying and classifying a large number of cells. For malaria stage classification, we compare results with expert human annotators and find comparable performance. CONCLUSIONS: Keras R-CNN is a Python package that performs automated cell identification for both brightfield and fluorescence images and can process large image sets. Both the package and image datasets are freely available on GitHub and the Broad Bioimage Benchmark Collection.

Probing Plasmodium falciparum sexual commitment at the single-cell level.

Background: Malaria parasites go through major transitions during their complex life cycle, yet the underlying differentiation pathways remain obscure. Here we apply single cell transcriptomics to unravel the program inducing sexual differentiation in Plasmodium falciparum. Parasites have to make this essential life-cycle decision in preparation for human-to-mosquito transmission. Methods: By combining transcriptional profiling with quantitative imaging and genetics, we defined a transcriptional signature in sexually committed cells. Results: We found this transcriptional signature to be distinct from general changes in parasite metabolism that can be observed in response to commitment-inducing conditions. Conclusions: This proof-of-concept study provides a template to capture transcriptional diversity in parasite populations containing complex mixtures of different life-cycle stages and developmental programs, with important implications for our understanding of parasite biology and the ongoing malaria elimination campaign.

Lysophosphatidylcholine Regulates Sexual Stage Differentiation in the Human Malaria Parasite Plasmodium falciparum.

Transmission represents a population bottleneck in the Plasmodium life cycle and a key intervention target of ongoing efforts to eradicate malaria. Sexual differentiation is essential for this process, as only sexual parasites, called gametocytes, are infective to the mosquito vector. Gametocyte production rates vary depending on environmental conditions, but external stimuli remain obscure. Here, we show that the host-derived lipid lysophosphatidylcholine (LysoPC) controls P. falciparum cell fate by repressing parasite sexual differentiation. We demonstrate that exogenous LysoPC drives biosynthesis of the essential membrane component phosphatidylcholine. LysoPC restriction induces a compensatory response, linking parasite metabolism to the activation of sexual-stage-specific transcription and gametocyte formation. Our results reveal that malaria parasites can sense and process host-derived physiological signals to regulate differentiation. These data close a critical knowledge gap in parasite biology and introduce a major component of the sexual differentiation pathway in Plasmodium that may provide new approaches for blocking malaria transmission.

Applying Faster R-CNN for Object Detection on Malaria Images.

Deep learning based models have had great success in object detection, but the state of the art models have not yet been widely applied to biological image data. We apply for the first time an object detection model previously used on natural images to identify cells and recognize their stages in brightfield microscopy images of malaria-infected blood. Many micro-organisms like malaria parasites are still studied by expert manual inspection and hand counting. This type of object detection task is challenging due to factors like variations in cell shape, density, and color, and uncertainty of some cell classes. In addition, annotated data useful for training is scarce, and the class distribution is inherently highly imbalanced due to the dominance of uninfected red blood cells. We use Faster Region-based Convolutional Neural Network (Faster R-CNN), one of the top performing object detection models in recent years, pre-trained on ImageNet but fine tuned with our data, and compare it to a baseline, which is based on a traditional approach consisting of cell segmentation, extraction of several single-cell features, and classification using random forests. To conduct our initial study, we collect and label a dataset of 1300 fields of view consisting of around 100,000 individual cells. We demonstrate that Faster R-CNN outperforms our baseline and put the results in context of human performance.

Infected erythrocyte-derived extracellular vesicles alter vascular function via regulatory Ago2-miRNA complexes in malaria.

Malaria remains one of the greatest public health challenges worldwide, particularly in sub-Saharan Africa. The clinical outcome of individuals infected with Plasmodium falciparum parasites depends on many factors including host systemic inflammatory responses, parasite sequestration in tissues and vascular dysfunction. Production of pro-inflammatory cytokines and chemokines promotes endothelial activation as well as recruitment and infiltration of inflammatory cells, which in turn triggers further endothelial cell activation and parasite sequestration. Inflammatory responses are triggered in part by bioactive parasite products such as hemozoin and infected red blood cell-derived extracellular vesicles (iRBC-derived EVs). Here we demonstrate that such EVs contain functional miRNA-Argonaute 2 complexes that are derived from the host RBC. Moreover, we show that EVs are efficiently internalized by endothelial cells, where the miRNA-Argonaute 2 complexes modulate target gene expression and barrier properties. Altogether, these findings provide a mechanistic link between EVs and vascular dysfunction during malaria infection.

Ensuring transmission through dynamic host environments: host-pathogen interactions in Plasmodium sexual development.

A renewed global commitment to malaria elimination lends urgency to understanding the biology of Plasmodium transmission stages. Recent progress toward uncovering the mechanisms underlying Plasmodium falciparum sexual differentiation and maturation reveals potential targets for transmission-blocking drugs and vaccines. The identification of parasite factors that alter sexual differentiation, including extracellular vesicles and a master transcriptional regulator, suggest that parasites make epigenetically controlled developmental decisions based on environmental cues. New insights into sexual development, especially host cell remodeling and sequestration in the bone marrow, highlight open questions regarding parasite homing to the tissue, transmigration across the vascular endothelium, and maturation in the parenchyma. Novel molecular and translational tools will provide further opportunities to define host-parasite interactions and design effective transmission-blocking therapeutics.

Designer lipid-like peptides: a class of detergents for studying functional olfactory receptors using commercial cell-free systems.

A crucial bottleneck in membrane protein studies, particularly G-protein coupled receptors, is the notorious difficulty of finding an optimal detergent that can solubilize them and maintain their stability and function. Here we report rapid production of 12 unique mammalian olfactory receptors using short designer lipid-like peptides as detergents. The peptides were able to solubilize and stabilize each receptor. Circular dichroism showed that the purified olfactory receptors had alpha-helical secondary structures. Microscale thermophoresis suggested that the receptors were functional and bound their odorants. Blot intensity measurements indicated that milligram quantities of each olfactory receptor could be produced with at least one peptide detergent. The peptide detergents' capability was comparable to that of the detergent Brij-35. The ability of 10 peptide detergents to functionally solubilize 12 olfactory receptors demonstrates their usefulness as a new class of detergents for olfactory receptors, and possibly other G-protein coupled receptors and membrane proteins.

A robust and rapid method of producing soluble, stable, and functional G-protein coupled receptors.

Membrane proteins, particularly G-protein coupled receptors (GPCRs), are notoriously difficult to express. Using commercial E. coli cell-free systems with the detergent Brij-35, we could rapidly produce milligram quantities of 13 unique GPCRs. Immunoaffinity purification yielded receptors at >90% purity. Secondary structure analysis using circular dichroism indicated that the purified receptors were properly folded. Microscale thermophoresis, a novel label-free and surface-free detection technique that uses thermal gradients, showed that these receptors bound their ligands. The secondary structure and ligand-binding results from cell-free produced proteins were comparable to those expressed and purified from HEK293 cells. Our study demonstrates that cell-free protein production using commercially available kits and optimal detergents is a robust technology that can be used to produce sufficient GPCRs for biochemical, structural, and functional analyses. This robust and simple method may further stimulate others to study the structure and function of membrane proteins.