scholarly journals Investigating heterogeneities of live mesenchymal stromal cells using AI-based label-free imaging

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Sara Imboden ◽  
Xuanqing Liu ◽  
Brandon S. Lee ◽  
Marie C. Payne ◽  
Cho-Jui Hsieh ◽  
...  
Keyword(s):  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
High Throughput Screening ◽  
Imaging Method ◽  
Significant Heterogeneity ◽  
Label Free ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Integrative Assessment ◽  
Conventional Cell

AbstractMesenchymal stromal cells (MSCs) are multipotent cells that have great potential for regenerative medicine, tissue repair, and immunotherapy. Unfortunately, the outcomes of MSC-based research and therapies can be highly inconsistent and difficult to reproduce, largely due to the inherently significant heterogeneity in MSCs, which has not been well investigated. To quantify cell heterogeneity, a standard approach is to measure marker expression on the protein level via immunochemistry assays. Performing such measurements non-invasively and at scale has remained challenging as conventional methods such as flow cytometry and immunofluorescence microscopy typically require cell fixation and laborious sample preparation. Here, we developed an artificial intelligence (AI)-based method that converts transmitted light microscopy images of MSCs into quantitative measurements of protein expression levels. By training a U-Net+ conditional generative adversarial network (cGAN) model that accurately (mean $$r_s$$ r s = 0.77) predicts expression of 8 MSC-specific markers, we showed that expression of surface markers provides a heterogeneity characterization that is complementary to conventional cell-level morphological analyses. Using this label-free imaging method, we also observed a multi-marker temporal-spatial fluctuation of protein distributions in live MSCs. These demonstrations suggest that our AI-based microscopy can be utilized to perform quantitative, non-invasive, single-cell, and multi-marker characterizations of heterogeneous live MSC culture. Our method provides a foundational step toward the instant integrative assessment of MSC properties, which is critical for high-throughput screening and quality control in cellular therapies.

scholarly journals Multi-pronged approach to human mesenchymal stromal cells senescence quantification with a focus on label-free methods

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Weichao Zhai ◽  
Jerome Tan ◽  
Tobias Russell ◽  
Sixun Chen ◽  
Dennis McGonagle ◽  
...  
Keyword(s):  
Flow Cytometry ◽  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
Preclinical Model ◽  
Label Free ◽  
Current Standard ◽  
Differentiation Potential ◽  
Human Mesenchymal Stromal Cells ◽  
Endogenous Fluorophores

AbstractHuman mesenchymal stromal cells (hMSCs) have demonstrated, in various preclinical settings, consistent ability in promoting tissue healing and improving outcomes in animal disease models. However, translation from the preclinical model into clinical practice has proven to be considerably more difficult. One key challenge being the inability to perform in situ assessment of the hMSCs in continuous culture, where the accumulation of the senescent cells impairs the culture’s viability, differentiation potential and ultimately leads to reduced therapeutic efficacies. Histochemical $$\upbeta $$ β -galactosidase staining is the current standard for measuring hMSC senescence, but this method is destructive and not label-free. In this study, we have investigated alternatives in quantification of hMSCs senescence, which included flow cytometry methods that are based on a combination of cell size measurements and fluorescence detection of SA-$$\upbeta $$ β -galactosidase activity using the fluorogenic substrate, C$${_{12}}$$ 12 FDG; and autofluorescence methods that measure fluorescence output from endogenous fluorophores including lipopigments. For identification of senescent cells in the hMSC batches produced, the non-destructive and label-free methods could be a better way forward as they involve minimum manipulations of the cells of interest, increasing the final output of the therapeutic-grade hMSC cultures. In this work, we have grown hMSC cultures over a period of 7 months and compared early and senescent hMSC passages using the advanced flow cytometry and autofluorescence methods, which were benchmarked with the current standard in $$\upbeta $$ β -galactosidase staining. Both the advanced methods demonstrated statistically significant values, (r = 0.76, p $$\le $$ ≤ 0.001 for the fluorogenic C$${_{12}}$$ 12 FDG method, and r = 0.72, p $$\le $$ ≤ 0.05 for the forward scatter method), and good fold difference ranges (1.120–4.436 for total autofluorescence mean and 1.082–6.362 for lipopigment autofluorescence mean) between early and senescent passage hMSCs. Our autofluroescence imaging and spectra decomposition platform offers additional benefit in label-free characterisation of senescent hMSC cells and could be further developed for adoption for future in situ cellular senescence evaluation by the cell manufacturers.

scholarly journals Deep inverse reinforcement learning for structural evolution of small molecules

2020 ◽  
Author(s):  
Brighter Agyemang ◽  
Wei-Ping Wu ◽  
Daniel Addo ◽  
Michael Y Kpiebaareh ◽  
Ebenezer Nanor ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
High Throughput Screening ◽  
Structural Evolution ◽  
Search Space ◽  
New Drugs ◽  
Inverse Reinforcement Learning ◽  
Generative Adversarial Network ◽  
Entropy Maximization ◽  
Reward Function ◽  
Adversarial Network

Abstract The size and quality of chemical libraries to the drug discovery pipeline are crucial for developing new drugs or repurposing existing drugs. Existing techniques such as combinatorial organic synthesis and high-throughput screening usually make the process extraordinarily tough and complicated since the search space of synthetically feasible drugs is exorbitantly huge. While reinforcement learning has been mostly exploited in the literature for generating novel compounds, the requirement of designing a reward function that succinctly represents the learning objective could prove daunting in certain complex domains. Generative adversarial network-based methods also mostly discard the discriminator after training and could be hard to train. In this study, we propose a framework for training a compound generator and learn a transferable reward function based on the entropy maximization inverse reinforcement learning (IRL) paradigm. We show from our experiments that the IRL route offers a rational alternative for generating chemical compounds in domains where reward function engineering may be less appealing or impossible while data exhibiting the desired objective is readily available.

An Inhomogeneous Background Imaging Method Based on Generative Adversarial Network

2020 ◽  
Vol 68 (11) ◽  
pp. 4684-4693
Author(s):  
Xiuzhu Ye ◽  
Yukai Bai ◽  
Rencheng Song ◽  
Kuiwen Xu ◽  
Jianping An
Keyword(s):  
Imaging Method ◽  
Generative Adversarial Network ◽  
Adversarial Network

scholarly journals RRG-GAN Restoring Network for Simple Lens Imaging System

Sensors ◽  
2021 ◽  
Vol 21 (10) ◽  
pp. 3317
Author(s):  
Xiaotian Wu ◽  
Jiongcheng Li ◽  
Guanxing Zhou ◽  
Bo Lü ◽  
Qingqing Li ◽  
...  
Keyword(s):  
Imaging System ◽  
Optimization Theory ◽  
Computational Imaging ◽  
Imaging Method ◽  
Generative Adversarial Network ◽  
Application Performance ◽  
Adversarial Network ◽  
Convex Lens ◽  
Evaluation Algorithm ◽  
Experimental Comparisons

The simple lens computational imaging method provides an alternative way to achieve high-quality photography. It simplifies the design of the optical-front-end to a single-convex-lens and delivers the correction of optical aberration to a dedicated computational restoring algorithm. Traditional single-convex-lens image restoration is based on optimization theory, which has some shortcomings in efficiency and efficacy. In this paper, we propose a novel Recursive Residual Groups network under Generative Adversarial Network framework (RRG-GAN) to generate a clear image from the aberrations-degraded blurry image. The RRG-GAN network includes dual attention module, selective kernel network module, and residual resizing module to make it more suitable for the non-uniform deblurring task. To validate the evaluation algorithm, we collect sharp/aberration-degraded datasets by CODE V simulation. To test the practical application performance, we built a display-capture lab setup and reconstruct a manual registering dataset. Relevant experimental comparisons and actual tests verify the effectiveness of our proposed method.

scholarly journals Biomolecular phenotyping and heterogeneity assessment of mesenchymal stromal cells using label-free Raman spectroscopy

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
R. A. Rocha ◽  
J. M. Fox ◽  
P. G. Genever ◽  
Y. Hancock
Keyword(s):  
Raman Spectroscopy ◽  
Cell Fate ◽  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
Complex Disease ◽  
Dermal Fibroblasts ◽  
Statistical Sampling ◽  
Label Free ◽  
Cell Level ◽  
Phenotypic Characterisation

AbstractEasy, quantitative measures of biomolecular heterogeneity and high-stratified phenotyping are needed to identify and characterise complex disease processes at the single-cell level, as well as to predict cell fate. Here, we demonstrate how Raman spectroscopy can be used in the difficult-to-assess case of clonal, bone-derived mesenchymal stromal cells (MSCs) to identify MSC lines and group these according to biological function (e.g., differentiation capacity). Biomolecular stratification is achieved using high-precision measures obtained from representative statistical sampling that also enable quantified heterogeneity assessment. Application to primary MSCs and human dermal fibroblasts shows use of these measures as a label-free assay to classify cell sub-types within complex heterogeneous cell populations, thus demonstrating the potential for therapeutic translation, and broad application to the phenotypic characterisation of other cells.

scholarly journals Exploring the social life of mesenchymal stromal cells by label-free tracking of individual and collective cell behaviour

2018 ◽  
Vol 26 ◽  
pp. S149
Author(s):  
A.L. Barnes ◽  
J. O'Flaherty ◽  
A. Carstairs ◽  
S. Quick ◽  
A.P. Stone ◽  
...  
Keyword(s):  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
Social Life ◽  
Label Free ◽  
Cell Behaviour ◽  
The Social

scholarly journals Adipose-Derived Mesenchymal Stromal Cells Treated with Interleukin 1 Beta Produced Chondro-Protective Vesicles Able to Fast Penetrate in Cartilage

Cells ◽  
2021 ◽  
Vol 10 (5) ◽  
pp. 1180
Author(s):  
Alessandra Colombini ◽  
Enrico Ragni ◽  
Leonardo Mortati ◽  
Francesca Libonati ◽  
Carlotta Perucca Orfei ◽  
...  
Keyword(s):  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
High Throughput Screening ◽  
Target Genes ◽  
Inflammatory Responses ◽  
Second Harmonic ◽  
Interleukin 1 ◽  
Time Lapse ◽  
Cartilage Explants

The study of the miRNA cargo embedded in extracellular vesicles (EVs) released from adipose-derived mesenchymal stromal cells (ASC) preconditioned with IL-1β, an inflammatory stimulus driving osteoarthritis (OA), along with EVs-cartilage dynamic interaction represent poorly explored fields and are the purpose of the present research. ASCs were isolated from subcutaneous adipose tissue and EVs collected by ultracentrifugation. Shuttled miRNAs were scored by high-throughput screening and analyzed through bioinformatics approach that predicted the potentially modulated OA-related pathways. Fluorescently labeled EVs incorporation into OA cartilage explants was followed in vitro by time-lapse coherent anti-Stokes Raman scattering; second harmonic generation and two-photon excited fluorescence. After IL-1β preconditioning, 7 miRNA were up-regulated, 4 down-regulated, 37 activated and 17 silenced. Bioinformatics allowed to identify miRNAs and target genes mainly involved in Wnt, Notch, TGFβ and Indian hedgehog (IHH) pathways, cartilage homeostasis, immune/inflammatory responses, cell senescence and autophagy. As well, ASC-EVs steadily diffuse in cartilage cells and matrix, reaching a plateau 16 h after administration. Overall, ASCs preconditioned with IL-1β allows secretion of EVs embedded with a chondro-protective miRNA cargo, able to fast penetrate in collagen-rich areas of cartilage with tissue saturation in a day. Further functional studies exploring the EVs dose-effects are needed to achieve clinical relevance.

Mechanisms of kidney repair by human mesenchymal stromal cells after ischemia: A comprehensive view using label-free MSE

PROTEOMICS ◽  
2014 ◽  
Vol 14 (12) ◽  
pp. 1480-1493 ◽  
Author(s):  
Milene R. da Costa ◽  
Luciana Pizzatti ◽  
Rafael S. Lindoso ◽  
Julliana Ferreira Sant’Anna ◽  
Barbara DuRocher ◽  
...  
Keyword(s):  
Mesenchymal Stromal Cells ◽  
Stromal Cells ◽  
Label Free ◽  
Human Mesenchymal Stromal Cells ◽  
Comprehensive View

scholarly journals Leveraging high-throughput screening data and conditional generative adversarial networks to advance predictive toxicology

2020 ◽  
Author(s):  
Adrian J. Green ◽  
Martin J. Mohlenkamp ◽  
Jhuma Das ◽  
Meenal Chaudhari ◽  
Lisa Truong ◽  
...  
Keyword(s):  
High Throughput ◽  
High Throughput Screening ◽  
Chemical Structure ◽  
Structural Information ◽  
Generative Adversarial Networks ◽  
Toxicity Data ◽  
Generative Adversarial Network ◽  
Structure Information ◽  
Adversarial Network

AbstractThere are currently 85,000 chemicals registered with the Environmental Protection Agency (EPA) under the Toxic Substances Control Act, but only a small fraction have measured toxicological data. To address this gap, high-throughput screening (HTS) methods are vital. As part of one such HTS effort, embryonic zebrafish were used to examine a suite of morphological and mortality endpoints at six concentrations from over 1,000 unique chemicals found in the ToxCast library (phase 1 and 2). We hypothesized that by using a conditional Generative Adversarial Network (cGAN) and leveraging this large set of toxicity data, plus chemical structure information, we could efficiently predict toxic outcomes of untested chemicals. CAS numbers for each chemical were used to generate textual files containing three-dimensional structural information for each chemical. Utilizing a novel method in this space, we converted the 3D structural information into a weighted set of points while retaining all information about the structure. In vivo toxicity and chemical data were used to train two neural network generators. The first used regression (Go-ZT) while the second utilized cGAN architecture (GAN-ZT) to train a generator to produce toxicity data. Our results showed that both Go-ZT and GAN-ZT models produce similar results, but the cGAN achieved a higher sensitivity (SE) value of 85.7% vs 71.4%. Conversely, Go-ZT attained higher specificity (SP), positive predictive value (PPV), and Kappa results of 67.3%, 23.4%, and 0.21 compared to 24.5%, 14.0%, and 0.03 for the cGAN, respectively. By combining both Go-ZT and GAN-ZT, our consensus model improved the SP, PPV, and Kappa, to 75.5%, 25.0%, and 0.211, respectively, resulting in an area under the receiver operating characteristic (AUROC) of 0.663. Considering their potential use as prescreening tools, these models could provide in vivo toxicity predictions and insight into untested areas of the chemical space to prioritize compounds for HT testing.SummaryA conditional Generative Adversarial Network (cGAN) can leverage a large chemical set of experimental toxicity data plus chemical structure information to predict the toxicity of untested compounds.

Sign in / Sign up

Export Citation Format

Share Document