Search

Recent advances with deep neural networks have shown the feasibility of acquiring brightfield images with transmitted light and applying in-silico labeling to predict fluorescent images. We have developed a novel in-silico labeling method based on a generative adversarial network and outperforms the state-of-the-art Unet method in generating realistic fluorescent images and quantitatively recapitulating real staining signals, as demonstrated in a complex co-culture myelination assay. Furthermore, we have performed the assay in live mode with multiple kinetic points, applied in-silico labeling to predict fluorescent images from brightfield and quantified the kinetic phenotypic changes. Thus, the proposed approach provides a potential tool to study the kinetics of cellular phenotypic changes with brightfield imaging.

Recent development of novel methods based on deep neural networks has transformed how high-content microscopy cellular images are analyzed. Nonetheless, it is still a challenge to identify cellular phenotypic changes caused by chemical or genetic treatments and to elucidate the relationships among treatments in an unsupervised manner, due to the large data volume, high phenotypic complexity and the presence of a priori unknown phenotypes. Here we benchmarked five deep neural network methods and two feature engineering methods on a well-characterized public data set. In contrast to previous benchmarking efforts, the manual annotations were not provided to the methods, but rather used as evaluation criteria afterwards. The seven methods individually performed feature extraction or representation learning from cellular images, and were consistently evaluated for downstream phenotype prediction and clustering tasks. We identified the strengths of individual methods across evaluation metrics, and further examined the biological concepts of features automatically learned by deep neural networks.

Gene and compound functions are often interrogated by perturbation. However, we have limited methods to capture associated phenotypes in an unbiased and holistic manner. Here, we describe Fluopack screening as a novel platform enabling the profiling of subcellular phenotypes associated with perturbation. Our approach leverages imaging of a panel of fluorescent chemical probes to survey cellular processes in an unbiased and high throughput fashion. Segmentation-free, whole image analysis applied to Fluopack images identifies probes revealing distinct phenotypes upon perturbation, thereby informing on the function and mechanism of action of perturbagens. This chemical biology approach allows to interrogate phenotypes that tend to be overlooked by other methods, such as lipid trafficking and ion concentration inside the cell. Fluopack screening is a powerful approach to study orphan protein function, as exemplified by the characterization of TMEM41B as novel regulator of lipid mobilization.

High-throughput screening generates large volumes of heterogeneous data that require a diverse set of computational tools for management, processing, and analysis. Building integrated, scalable, and robust computational workflows for such applications is challenging but highly valuable. Scientific data integration and pipelining facilitate standardized data processing, collaboration, and reuse of best practices. We describe how Jenkins-CI, an "off-the-shelf," open-source, continuous integration system, is used to build pipelines for processing images and associated data from high-content screening (HCS). Jenkins-CI provides numerous plugins for standard compute tasks, and its design allows the quick integration of external scientific applications. Using Jenkins-CI, we integrated CellProfiler, an open-source image-processing platform, with various HCS utilities and a high-performance Linux cluster. The platform is web-accessible, facilitates access and sharing of high-performance compute resources, and automates previously cumbersome data and image-processing tasks. Imaging pipelines developed using the desktop CellProfiler client can be managed and shared through a centralized Jenkins-CI repository. Pipelines and managed data are annotated to facilitate collaboration and reuse. Limitations with Jenkins-CI (primarily around the user interface) were addressed through the selection of helper plugins from the Jenkins-CI community.

High throughput screening has become a basic technique with which to explore biological systems. Advances in technology, including increased screening capacity, as well as methods that generate multiparametric readouts, are driving the need for improvements in the analysis of data sets derived from such screens.This article covers the recent advances in the analysis of high throughput screening data sets from arrayed samples, as well as the recent advances in the analysis of cell-by-cell data sets derived from image or flow cytometry application. Screening multiple genomic reagents targeting any given gene creates additional challenges and so methods that prioritize individual gene targets have been developed. The article reviews many of the open source data analysis methods that are now available and which are helping to define a consensus on the best practices to use when analyzing screening data.As data sets become larger, and more complex, the need for easily accessible data analysis tools will continue to grow. The presentation of such complex data sets, to facilitate quality control monitoring and interpretation of the results will require the development of novel visualizations. In addition, advanced statistical and machine learning algorithms that can help identify patterns, correlations and the best features in massive data sets will be required. The ease of use for these tools will be important, as they will need to be used iteratively by laboratory scientists to improve the outcomes of complex analyses.

Screens using high-throughput, information-rich technologies such as microarrays, high-content screening (HCS), and next-generation sequencing (NGS) have become increasingly widespread. Compared with single-readout assays, these methods produce a more comprehensive picture of the effects of screened treatments. However, interpreting such multidimensional readouts is challenging. Univariate statistics such as t-tests and Z-factors cannot easily be applied to multidimensional profiles, leaving no obvious way to answer common screening questions such as "Is treatment X active in this assay?" and "Is treatment X different from (or equivalent to) treatment Y?" We have developed a simple, straightforward metric, the multidimensional perturbation value (mp-value), which can be used to answer these questions. Here, we demonstrate application of the mp-value to three data sets: a multiplexed gene expression screen of compounds and genomic reagents, a microarray-based gene expression screen of compounds, and an HCS compound screen. In all data sets, active treatments were successfully identified using the mp-value, and simulations and follow-up analyses supported the mp-value's statistical and biological validity. We believe the mp-value represents a promising way to simplify the analysis of multidimensional data while taking full advantage of its richness.

KSR-1 is a scaffold protein that is essential for Ras-induced activation of the highly conserved RAF-MEK-ERK kinase module. Previously, we identified a close homolog of KSR-1, called KSR-2, through structural homology-based data mining. In order to further understand the role of KSR-2 in MAPK signaling, we undertook a functional proteomics approach to elucidate the dynamic composition of the KSR-2 functional complex in HEK-293 cells under conditions with and without TNF-α stimulation. We found nearly 100 proteins that were potentially associated with KSR-2 complex and 43 proteins that were likely recruited to the super molecular complex after TNF-α treatment. Our results indicate that KSR-2 may act as a scaffold protein similar as KSR-1 to mediate the MAPK core (RAF-MEK-ERK) signaling but with a distinct RAF isoform specificity, namely KSR-2 may only mediate the A-RAF signaling while KSR-1 is responsible for transducing signals only from c-RAF. In addition, KSR-2 may be involved in the activation of many MAPK downstream signaling molecules such as p38 MAPK, IKAP, AIF, and proteins involved in ubiquitin–proteasome, apoptosis, cell cycle control, and DNA synthesis and repair pathways, as well as mediating crosstalks between MAPK and several other signaling pathways, including PI3K and insulin signaling. While interactions with these molecules are not known for KSR-1, it's reasonable to hypothesize that KSR-1 may also play a similar role in mediating these downstream signaling pathways.

We have previously shown that genetically engineered adult mesenchymal stem cells (AMSCs) expressing recombinant human bone morphogenetic protein –2 (rhBMP-2), under tet-regulation, can induce bone formation and regeneration. We showed that these cells induce bone formation via paracrine and autocrine effect of the secreted rhBMP-2 protein. To distinguish between these two effects, and to develop a platform for continuous delivery of rhBMP-2 by engineered cells protected from the immune system, we have used hydrogel polymer (alginate) in order to encapsulate the AMSCs. Mixing of the cells with potassium alginate, followed by sedimentation in Ca2+ solution, results in polymerization of the alginate around the cells, forming microcapsules composed of a membrane allowing diffusion of small molecule and proteins. Encapsulated engineered AMSCs were able to survive inside the capsules in vitro and in vivo and secrete rhBMP-2 under tet-regulation. Transplantation of capsules both subcutaneously and into bone defect elicited physiological response manifested in osteogenic tissue composed of bone trabeculae and cartilage. Inside the capsules, engineered AMSCs differentiated to chondrocytes (autocrine effect), but not to osteoblasts. Newly formed bone has developed around the polymeric external layer without any observed intermediate layer of tissue. There was no evidence of immune response in the transplants area. We therefore conclude that engineered AMSCs can be efficiently encapsulated within polymeric alginate microcapsules, maintain viability, differentiate by autocrine effect, secrete rhBMP-2 under exogenous regulation, and induce bone formation by paracrine effect, with no adverse or immune response to the transplanted capsules. Copyright © 2003 John Wiley & Sons, Ltd.

Osteoporosis is a disease manifested in drastic bone loss resulting in osteopenia and high risk for fractures. This disease is generally divided into two subtypes. The first, post-menopausal (type I) osteoporosis, is primarily related to estrogen deficiency. The second, senile (type II) osteoporosis, is mostly related to aging. Decreased bone formation, as well as increased bone resorption and turnover, are thought to play roles in the pathophysiology of both types of osteoporosis. In this study, we demonstrate in murine models for both type I (estrogen deficiency) and type II (senile) osteopenia/osteoporosis that reduced bone formation is related to a decrease in adult mesenchymal stem cell (AMSC) number, osteogenic activity, and proliferation. Decreased proliferation is coupled with increased apoptosis in AMSC cultures obtained from osteopenic mice. Recombinant human bone morphogenetic protein (rhBMP-2) is a highly osteoinductive protein, promoting osteogenic differentiation of AMSCs. Systemic intra-peritoneal (i.p.) injections of rhBMP-2 into osteopenic mice were able to reverse this phenotype in the bones of these animals. Moreover, this change in bone mass was coupled to an increase in AMSCs numbers, osteogenic activity, and proliferation as well as a decrease in apoptosis. Bone formation activity was increased as well. However, the magnitude of this response to rhBMP-2 varied among different stains of mice. In old osteopenic BALB/c male mice (type II osteoporosis model), rhBMP-2 systemic treatment also restored both articular and epiphyseal cartilage width to the levels seen in young mice. In summary, our study shows that AMSCs are a good target for systemically active anabolic compounds like rhBMP-2. J. Cell. Biochem. 86: 461–474, 2002. © 2002 Wiley-Liss, Inc.

Background Among the approximately 6.5 million fractures suffered in the United States every year, about 15% are difficult to heal. As yet, for most of these difficult cases there is no effective therapy. We have developed a mouse radial segmental defect as a model experimental system for testing the capacity of Genetically Engineered Pluripotent Mesenchymal Cells (GEPMC, C3H10T1/2 clone expressing rhBMP-2), for gene delivery, engraftment, and induction of bone growth in regenerating bone. Methods Transfected GEPMC expressing rhBMP-2 were further infected with a vector carrying the lacZ gene, that encodes for β-galactosidase (β-gal). In vitro levels of rhBMP-2 expression and function were confirmed by immunohistochemistry, and bioassay. Differentiation was assayed using alkaline phosphatase staining. GEPMC were transplanted in vivo into a radial segmental defect. The main control groups included lacZ clones of WT-C3H10T1/2-LacZ, and CHO-rhBMP-2 cells. New bone formation was measured quantitatively via fluorescent labeling, X-ray analysis and histomorphometry. Engrafted mesenchymal cells were localized in vivo by β-gal expression, and double immunofluorescence. Results In vitro, GEPMC expressed rhBMP-2, β-gal and spontaneously differentiated into osteogenic cells expressing alkaline phosphatase. Detection of transplanted cells revealed engrafted cells that had differentiated into osteoblasts and co-expressed β-gal and rhBMP-2. Analysis of new bone formation revealed that at fout to eight week post-transplantation, GEPMS significantly enhanced segmental defect repair. Conclusions Our study shows that cell-mediated gene transfer can be utilized for growth factor delivery to signaling receptors of transplanted cells (autocrine effect) and host mesenchymal cells (paracrine effect) suggesting the ability of GEPMC to engraft, differentiate, and stimulate bone growth. We suggest that our approach should lead to the designing of mesenchymal stem cell based gene therapy strategies for bone lesions as well as other tissues. Copyright © 1999 John Wiley & Sons, Ltd.

Background Human mesenchymal stem cells (hMSCs) are pluripotent cells that can differentiate to various mesenchymal cell types. Recently, a method to isolate hMSCs from bone marrow and expand them in culture was described. Here we report on the use of hMSCs as a platform for gene therapy aimed at bone lesions. Methods Bone marrow derived hMSCs were expanded in culture and infected with recombinant adenoviral vector encoding the osteogenic factor, human BMP-2. The osteogenic potential of genetically engineered hMSCs was assessed in vitro and in vivo. Results Genetically engineered hMSCs displayed enhanced proliferation and osteogenic differentiation in culture. In vivo, transplanted genetically engineered hMSCs were able to engraft and form bone and cartilage in ectopic sites, and regenerate bone defects (non-union fractures) in mice radius bone. Importantly, the same results were obtained with hMSCs isolated from a patient suffering from osteoporosis. Conclusions hMSCs represent a novel platform for skeletal gene therapy and the present results suggest that they can be genetically engineered to express desired therapeutic proteins inducing specific differentiation pathways. Moreover, hMSCs obtained from osteoporotic patients can restore their osteogenic activity following human BMP-2 gene transduction, an important finding in the future planning of gene therapy treatment for osteoporosis. Copyright © 2001 John Wiley & Sons, Ltd.