Deep neural net tracking of human pluripotent stem cells reveals intrinsic behaviors directing morphogenesis

The use of artificial intelligence in induced pluripotent stem cell-based technology over 10-year period: A systematic scoping review

BACKGROUND

Stem cell research, particularly in the domain of induced pluripotent stem cell (iPSC) technology, has shown significant progress. The integration of artificial intelligence (AI), especially machine learning (ML) and deep learning (DL), has played a pivotal role in refining iPSC classification, monitoring cell functionality, and conducting genetic analysis. These enhancements are broadening the applications of iPSC technology in disease modelling, drug screening, and regenerative medicine. This review aims to explore the role of AI in the advancement of iPSC research.

METHODS

In December 2023, data were collected from three electronic databases (PubMed, Web of Science, and Science Direct) to investigate the application of AI technology in iPSC processing.

RESULTS

This systematic scoping review encompassed 79 studies that met the inclusion criteria. The number of research studies in this area has increased over time, with the United States emerging as a leading contributor in this field. AI technologies have been diversely applied in iPSC technology, encompassing the classification of cell types, assessment of disease-specific phenotypes in iPSC-derived cells, and the facilitation of drug screening using iPSC. The precision of AI methodologies has improved significantly in recent years, creating a foundation for future advancements in iPSC-based technologies.

CONCLUSIONS

Our review offers insights into the role of AI in regenerative and personalized medicine, highlighting both challenges and opportunities. Although still in its early stages, AI technologies show significant promise in advancing our understanding of disease progression and development, paving the way for future clinical applications.

TBXT dose sensitivity and the decoupling of nascent mesoderm specification from EMT progression in 2D human gastruloids

In the nascent mesoderm, TBXT expression must be precisely regulated to ensure that cells exit the primitive streak and pattern the anterior-posterior axis, but how varying dosage informs morphogenesis is not well understood. In this study, we define the transcriptional consequences of TBXT dosage reduction during early human gastrulation using human induced pluripotent stem cell models of gastrulation and mesoderm differentiation. Multi-omic single-nucleus RNA and single-nucleus ATAC sequencing of 2D gastruloids comprising wild-type, TBXT heterozygous or TBXT null human induced pluripotent stem cells reveal that varying TBXT dosage does not compromise the ability of a cell to differentiate into nascent mesoderm, but instead directly influences the temporal progression of the epithelial-to-mesenchymal transition with wild type transitioning first, followed by TBXT heterozygous and then TBXT null. By differentiating cells into nascent mesoderm in a monolayer format, we further illustrate that TBXT dosage directly impacts the persistence of junctional proteins and cell-cell adhesions. These results demonstrate that epithelial-to-mesenchymal transition progression can be decoupled from the acquisition of mesodermal identity in the early gastrula and shed light on the mechanisms underlying human embryogenesis.

High-volume, label-free imaging for quantifying single-cell dynamics in induced pluripotent stem cell colonies

To facilitate the characterization of unlabeled induced pluripotent stem cells (iPSCs) during culture and expansion, we developed an AI pipeline for nuclear segmentation and mitosis detection from phase contrast images of individual cells within iPSC colonies. The analysis uses a 2D convolutional neural network (U-Net) plus a 3D U-Net applied on time lapse images to detect and segment nuclei, mitotic events, and daughter nuclei to enable tracking of large numbers of individual cells over long times in culture. The analysis uses fluorescence data to train models for segmenting nuclei in phase contrast images. The use of classical image processing routines to segment fluorescent nuclei precludes the need for manual annotation. We optimize and evaluate the accuracy of automated annotation to assure the reliability of the training. The model is generalizable in that it performs well on different datasets with an average F1 score of 0.94, on cells at different densities, and on cells from different pluripotent cell lines. The method allows us to assess, in a non-invasive manner, rates of mitosis and cell division which serve as indicators of cell state and cell health. We assess these parameters in up to hundreds of thousands of cells in culture for more than 36 hours, at different locations in the colonies, and as a function of excitation light exposure.

Exogeneous metal ions as therapeutic agents in cardiovascular disease and their delivery strategies

Metal ions participate in many metabolic processes in the human body, and their homeostasis is crucial for life. In cardiovascular diseases (CVDs), the equilibriums of metal ions are frequently interrupted, which are related to a variety of disturbances of physiological processes leading to abnormal cardiac functions. Exogenous supplement of metal ions has the potential to work as therapeutic strategies for the treatment of CVDs. Compared with other therapeutic drugs, metal ions possess broad availability, good stability and safety and diverse drug delivery strategies. The delivery strategies of metal ions are important to exert their therapeutic effects and reduce the potential toxic side effects for cardiovascular applications, which are also receiving increasing attention. Controllable local delivery strategies for metal ions based on various biomaterials are constantly being designed. In this review, we comprehensively summarized the positive roles of metal ions in the treatment of CVDs from three aspects: protecting cells from oxidative stress, inducing angiogenesis, and adjusting the functions of ion channels. In addition, we introduced the transferability of metal ions in vascular reconstruction and cardiac tissue repair, as well as the currently available engineered strategies for the precise delivery of metal ions, such as integrated with nanoparticles, hydrogels and scaffolds.

TBXT dose sensitivity and the decoupling of nascent mesoderm specification from EMT progression in 2D human gastruloids

In the nascent mesoderm, levels of Brachyury (TBXT) expression must be precisely regulated to ensure cells exit the primitive streak and pattern the anterior-posterior axis, but how this varying dosage informs morphogenesis is not well understood. In this study, we define the transcriptional consequences of TBXT dose reduction during early human gastrulation using human induced pluripotent stem cell (hiPSC)-based models of gastrulation and mesoderm differentiation. Multiomic single-nucleus RNA and single-nucleus ATAC sequencing of 2D gastruloids comprised of WT, TBXT heterozygous (TBXT-Het), or TBXT null (TBXT-KO) hiPSCs reveal that varying TBXT dosage does not compromise a cell's ability to differentiate into nascent mesoderm, but that the loss of TBXT significantly delays the temporal progression of the epithelial to mesenchymal transition (EMT). This delay is dependent on TBXT dose, as cells heterozygous for TBXT proceed with EMT at an intermediate pace relative to WT or TBXT-KO. By differentiating iPSCs of the allelic series into nascent mesoderm in a monolayer format, we further illustrate that TBXT dose directly impacts the persistence of junctional proteins and cell-cell adhesions. These results demonstrate that EMT progression can be decoupled from the acquisition of mesodermal identity in the early gastrula and shed light on the mechanisms underlying human embryogenesis.

Artificial Intelligence Supports Automated Characterization of Differentiated Human Pluripotent Stem Cells

Revolutionary advances in AI and deep learning in recent years have resulted in an upsurge of papers exploring applications within the biomedical field. Within stem cell research, promising results have been reported from analyses of microscopy images to, that is, distinguish between pluripotent stem cells and differentiated cell types derived from stem cells. In this work, we investigated the possibility of using a deep learning model to predict the differentiation stage of pluripotent stem cells undergoing differentiation toward hepatocytes, based on morphological features of cell cultures. We were able to achieve close to perfect classification of images from early and late time points during differentiation, and this aligned very well with the experimental validation of cell identity and function. Our results suggest that deep learning models can distinguish between different cell morphologies, and provide alternative means of semi-automated functional characterization of stem cell cultures.

Artificial intelligence and improvement of stem cell delivery in healthcare

Artificial intelligence (AI) is critical for improving the quality of stem cell manufacturing and delivery. AI can assist in determining the viability, effectiveness, efficacy, and safety of stem cells. Furthermore, in stem cell and regenerative medicine, AI is utilized to streamline simulation and model-building processes and find connections between cellular activities and their microenvironments. However, thoughtful consideration is required to minimize unwanted implications of AI incorporation for stem cell-based treatment.

Loss of TJP1 disrupts gastrulation patterning and increases differentiation toward the germ cell lineage in human pluripotent stem cells

Biological patterning events that occur early in development establish proper tissue morphogenesis. Identifying the mechanisms that guide these patterning events is necessary in order to understand the molecular drivers of development and disease and to build tissues in vitro. In this study, we use an in vitro model of gastrulation to study the role of tight junctions and apical/basolateral polarity in modulating bone morphogenic protein-4 (BMP4) signaling and gastrulation-associated patterning in colonies of human pluripotent stem cells (hPSCs). Disrupting tight junctions via knockdown (KD) of the scaffolding tight junction protein-1 (TJP1, also known as ZO1) allows BMP4 to robustly and ubiquitously activate pSMAD1/5 signaling over time, resulting in loss of the patterning phenotype and marked differentiation bias of pluripotent stem cells to primordial germ cell-like cells (PGCLCs). These findings give important insights into how signaling events are regulated and lead to spatial emergence of diverse cell types in vitro.

High precision tracking analysis of cell position and motion fields using 3D U-net network models

Cells are the basic units of biological organization, and the quantitative analysis of cellular states is an important topic in medicine and is valuable in revealing the complex mechanisms of microscopic world organisms. In order to better understand cell cycle changes as well as drug actions, we need to track cell migration and division. In this paper, we propose a novel engineering model for tracking cells using cell position and motion fields (CPMF). The training sample does not need to be manually annotated, and we modify and edit it against the ground truth using auxiliary tools. The core idea of the project is to combine detection and correlation, and the cell sequence samples are trained by a U-Net network model composed of 3D CNNs, which can track the migration, division, and entry and exit of cells in the field of view with high accuracy in all directions. The average detection accuracy of the cell coordinates is 98.38% and the average tracking accuracy is 98.70%.

Derivation of Human Salivary Epithelial Progenitors from Pluripotent Stem Cells via Activation of RA and Wnt Signaling

Derivation of salivary gland epithelial progenitors (SGEPs) from human pluripotent stem cells (hPSCs) has great potential in developmental biology and regenerative medicine. At present, no efficient method is available to generate salivary gland cells from hPSCs. Here, we described for the first time a robust protocol for direct differentiation of hPSCs into SGEPs by mimicking retinoic acid and Wnt signaling. These hPSC-derived SGEPs expressed SOX9, KRT5, and KRT19, important progenitor markers of developing salivary glands. CD24 and α-SMA positive cells, capable of restoring the functions of injured salivary glands, were also present in SGEP cultures. Importantly, RNA-sequencing revealed that the SGEPs resembled the transcript profiles of human fetal submandibular glands. Therefore, we provided an efficient protocol to induce hPSCs differentiation into SGEPs. Our study provides a foundation for generating functional hPSCs derived salivary gland acinar cells and three-dimensional organoids, potentially serving as new models for basic study and future translational research.

Machine learning-based detection of label-free cancer stem-like cell fate

The detection of cancer stem-like cells (CSCs) is mainly based on molecular markers or functional tests giving a posteriori results. Therefore label-free and real-time detection of single CSCs remains a difficult challenge. The recent development of microfluidics has made it possible to perform high-throughput single cell imaging under controlled conditions and geometries. Such a throughput requires adapted image analysis pipelines while providing the necessary amount of data for the development of machine-learning algorithms. In this paper, we provide a data-driven study to assess the complexity of brightfield time-lapses to monitor the fate of isolated cancer stem-like cells in non-adherent conditions. We combined for the first time individual cell fate and cell state temporality analysis in a unique algorithm. We show that with our experimental system and on two different primary cell lines our optimized deep learning based algorithm outperforms classical computer vision and shallow learning-based algorithms in terms of accuracy while being faster than cutting-edge convolutional neural network (CNNs). With this study, we show that tailoring our deep learning-based algorithm to the image analysis problem yields better results than pre-trained models. As a result, such a rapid and accurate CNN is compatible with the rise of high-throughput data generation and opens the door to on-the-fly CSC fate analysis.

3D Polymer Architectures for the Identification of Optimal Dimensions for Cellular Growth of 3D Cellular Models

Organ-on-chips and scaffolds for tissue engineering are vital assay tools for pre-clinical testing and prediction of human response to drugs and toxins, while providing an ethical sound replacement for animal testing. A success criterion for these models is the ability to have structural parameters for optimized performance. Here we show that two-photon polymerization fabrication can create 3D test platforms, where scaffold parameters can be directly analyzed by their effects on cell growth and movement. We design and fabricate a 3D grid structure, consisting of wall structures with niches of various dimensions for probing cell attachment and movement, while providing easy access for fluorescence imaging. The 3D structures are fabricated from bio-compatible polymer SZ2080 and subsequently seeded with A549 lung epithelia cells. The seeded structures are imaged with confocal microscopy, where spectral imaging with linear unmixing is used to separate auto-fluorescence scaffold contribution from the cell fluorescence. The volume of cellular material present in different sections of the structures is analyzed, to study the influence of structural parameters on cell distribution. Furthermore, time-lapse studies are performed to map the relation between scaffold parameters and cell movement. In the future, this kind of differentiated 3D growth platform, could be applied for optimized culture growth, cell differentiation, and advanced cell therapies.

Artificial-Intelligence-Based Imaging Analysis of Stem Cells: A Systematic Scoping Review

This systematic scoping review aims to map and identify the available artificial-intelligence-based techniques for imaging analysis, the characterization of stem cell differentiation, and trans-differentiation pathways. On the ninth of March 2022, data were collected from five electronic databases (PubMed, Medline, Web of Science, Cochrane, and Scopus) and manual citation searching; all data were gathered in Zotero 5.0. A total of 4422 articles were collected after deduplication; only twenty-seven studies were included in this systematic scoping review after a two-phase screening against inclusion criteria by two independent reviewers. The amount of research in this field is significantly increasing over the years. While the current state of artificial intelligence (AI) can tackle a multitude of medical problems, the consensus amongst researchers remains that AI still falls short in multiple ways that investigators should examine, ranging from the quality of images used in training sets and appropriate sample size, as well as the unexpected events that may occur which the algorithm cannot predict.