ERK5 Is Required for Tumor Growth and Maintenance Through Regulation of the Extracellular Matrix in Triple Negative Breast Cancer

Conventional mitogen-activated protein kinase (MAPK) family members regulate diverse cellular processes involved in tumor initiation and progression, yet the role of ERK5 in cancer biology is not fully understood. Triple-negative breast cancer (TNBC) presents a clinical challenge due to the aggressive nature of the disease and a lack of targeted therapies. ERK5 signaling contributes to drug resistance and metastatic progression through distinct mechanisms, including activation of epithelial-to-mesenchymal transition (EMT). More recently a role for ERK5 in regulation of the extracellular matrix (ECM) has been proposed, and here we investigated the necessity of ERK5 in TNBC tumor formation. Depletion of ERK5 expression using the CRISPR/Cas9 system in MDA-MB-231 and Hs-578T cells resulted in loss of mesenchymal features, as observed through gene expression profile and cell morphology, and suppressed TNBC cell migration. In vivo xenograft experiments revealed ERK5 knockout disrupted tumor growth kinetics, which was restored using high concentration Matrigel™ and ERK5-ko reduced expression of the angiogenesis marker CD31. These findings implicated a role for ERK5 in the extracellular matrix (ECM) and matrix integrity. RNA-sequencing analyses demonstrated downregulation of matrix-associated genes, integrins, and pro-angiogenic factors in ERK5-ko cells. Tissue decellularization combined with cryo-SEM and interrogation of biomechanical properties revealed that ERK5-ko resulted in loss of key ECM fiber alignment and mechanosensing capabilities in breast cancer xenografts compared to parental wild-type cells. In this study, we identified a novel role for ERK5 in tumor growth kinetics through modulation of the ECM and angiogenesis axis in breast cancer.

An Open-Hardware Insemination Device for Small-Bodied Live-Bearing Fishes to Support Development and Use of Germplasm Repositories

Small-bodied live-bearing fishes attract broad attention because of their importance in biomedical research and critical conservation status in natural habitats. Artificial insemination is an essential process to establish hybrid lines and for the operation of sperm repositories. The existing mouth-pipetting technique for artificial insemination of live-bearing fishes has not been substantially upgraded since the first implementation in the 1950s. The goal of this work was to develop a standardized artificial inseminator device (SAID) to address issues routinely encountered in insemination by mouth-pipetting, including lack of reproducibility among different users, difficulty in training, and large unreportable variation in sample volume and pressure during insemination. Prototypes of the SAID were designed as relatively inexpensive ( 0.99) between the piston position and volume. Pressure generation from eight mouth-pipetting operators and SAID prototypes were assessed by pressure sensors. The pressure control by SAID was superior to that produced by mouth-pipetting, yielding lower pressures (31−483 Pa) and smaller variations (standard deviation <11 Pa). These pressures were sufficient to deliver 1−5 μL of fluid into female reproductive tracts yet low enough to avoid physical injury to fish. Community-level enhancements of the SAID prototype could enable standardized insemination with minimal training and facilitate the participation of research communities in the use of cryopreserved genetic resources.

A practical evaluation of machine learning for classification of ultrasound images of ovarian development in channel catfish (Ictalurus punctatus)

Machine learning is a powerful tool to improve efficiency of industrial processes, but it has not yet been well utilized in aquacultural and hatchery applications. The goal of the present study was to evaluate the feasibility of using a broad array of machine learning approaches (testing of > 200 vectorization and model combinations, reporting on 20) to classify ultrasound images spanning annual ovarian development (i.e., from undeveloped to mature) of channel catfish (Ictalurus punctatus). The specific objectives were to: 1) establish dataset preprocessing to standardize image features; 2) develop and train image classification models with deep learning methods; 3) develop and train models with traditional machine learning methods; 4) compare performance of deep learning and traditional methods on two classification problems (2-class and 5-class), and 5) propose insights to deploy models in practical aquaculture applications for research and hatchery use. A total of 931 ultrasound images of catfish ovaries were used to train and evaluate models for a 2-class problem (as a 'yes' or 'no' answer) to support hormone-injection decisions for spawning management in hatcheries, and a 5-class problem for classifying gonadal development stages for research. By using feature extraction, cropping, dimension reduction, and histogram normalization, a preprocessing method was created to standardize images to develop traditional (i.e., vector input), and deep learning convolutional neural network (CNN) (i.e., image input) approaches. Traditional machine learning models with image classification achieved 100% median accuracy on the 2-class problem (with the models RN-50 and RN-152), and 96% median accuracy for the 5-class problem (with VGG-19 image vectorization). The deep learning approach for the 2-class problem had a median accuracy of > 98% for 15models. The 5-class deep learning models produced a steady increase in median accuracy with training net size, achievable through expansion of the dataset. These models can be developed further, but traditional models (using CNN architectures to simply calculate image vectors) outperformed the deep learning approach. These models can be directly applicable to aquaculture in hatcheries and reproductive biology research, in addition to a wide variety of other image-based applications.

A Modified-Herringbone Micromixer for Assessing Zebrafish Sperm (MAGS)

Sperm motility analysis of aquatic model species is important yet challenging due to the small sample volume, the necessity to activate with water, and the short duration of motility. To achieve standardization of sperm activation, microfluidic mixers have shown improved reproducibility over activation by hand, but challenges remain in optimizing and simplifying the use of these microdevices for greater adoption. The device described herein incorporates a novel micromixer geometry that aligns two sperm inlet streams with modified herringbone structures that split and recombine the sample at a 1:6 dilution with water to achieve rapid and consistent initiation of motility. The polydimethylsiloxane (PDMS) chip can be operated in a positive or negative pressure configuration, allowing a simple micropipettor to draw samples into the chip and rapidly stop the flow. The device was optimized to not only activate zebrafish sperm but also enables practical use with standard computer-assisted sperm analysis (CASA) systems. The micromixer geometry could be modified for other aquatic species with differing cell sizes and adopted for an open hardware approach using 3D resin printing where users could revise, fabricate, and share designs to improve standardization and reproducibility across laboratories and repositories.

Programmable peroxidase-assisted signal amplification enables flexible detection of nucleic acid targets in cellular and histopathological specimens

In situ hybridization (ISH) is a powerful tool for investigating the spatial arrangement of nucleic acid targets in fixed samples. ISH is typically visualized using fluorophores to allow high sensitivity and multiplexing or with colorimetric labels to facilitate co-visualization with histopathological stains. Both approaches benefit from signal amplification, which makes target detection effective, rapid, and compatible with a broad range of optical systems. Here, we introduce a unified technical platform, termed 'pSABER', for the amplification of ISH signals in cell and tissue systems. pSABER decorates the in situ target with concatemeric binding sites for a horseradish peroxidase-conjugated oligonucleotide which can then catalyze the massive localized deposition of fluorescent or colorimetric substrates. We demonstrate that pSABER effectively labels DNA and RNA targets, works robustly in cultured cells and challenging formalin fixed paraffin embedded (FFPE) specimens. Furthermore, pSABER can achieve 25-fold signal amplification over conventional signal amplification by exchange reaction (SABER) and can be serially multiplexed using solution exchange. Therefore, by linking nucleic acid detection to robust signal amplification capable of diverse readouts, pSABER will have broad utility in research and clinical settings.

Predicting cell cycle stage from 3D single-cell nuclear-stained images

The cell cycle governs the proliferation of all eukaryotic cells. Profiling cell cycle dynamics is therefore central to basic and biomedical research. However, current approaches to cell cycle profiling involve complex interventions that may confound experimental interpretation. We developed CellCycleNet, a machine learning (ML) workflow, to simplify cell cycle staging from fluorescent microscopy data with minimal experimenter intervention and cost. CellCycleNet accurately predicts cell cycle phase using only a fluorescent nuclear stain (DAPI) in fixed interphase cells. Using the Fucci2a cell cycle reporter system as ground truth, we collected two benchmarking image datasets and trained 2D and 3D ML models-of support vector machine and deep neural network architecture-to classify nuclei in the G1 or S/G2 phases. Our results show that 3D CellCycleNet outperforms support vector machine models on each dataset. When trained on two image datasets simultaneously, CellCycleNet achieves the highest classification accuracy (AUROC of 0.94-0.95). Overall, we found that using 3D features, rather than 2D features alone, significantly improves classification performance for all model architectures. We released our image data, models, and software as a community resource.

Predicting cell cycle stage from 3D single-cell nuclear-stained images

The cell cycle governs the proliferation, differentiation, and regeneration of all eukaryotic cells. Profiling cell cycle dynamics is therefore central to basic and biomedical research spanning development, health, aging, and disease. However, current approaches to cell cycle profiling involve complex interventions that may confound experimental interpretation. To facilitate more efficient cell cycle annotation of microscopy data, we developed CellCycleNet, a machine learning (ML) workflow designed to simplify cell cycle staging with minimal experimenter intervention and cost. CellCycleNet accurately predicts cell cycle phase using only a fluorescent nuclear stain (DAPI) in fixed interphase cells. Using the Fucci2a cell cycle reporter system as ground truth, we collected two benchmarking image datasets and trained two ML models-a support vector machine (SVM) and a deep neural network-to classify nuclei as being in either the G1 or S/G2 phases of the cell cycle. Our results suggest that CellCycleNet outperforms state-of-the-art SVM models on each dataset individually. When trained on two image datasets simultaneously, CellCycleNet achieves the highest classification accuracy, with an improvement in AUROC of 0.08-0.09. The model also demonstrates excellent generalization across different microscopes, achieving an AUROC of 0.95. Overall, using features derived from 3D images, rather than 2D projections of those same images, significantly improves classification performance. We have released our image data, trained models, and software as a community resource.

Efficient and highly amplified imaging of nucleic acid targets in cellular and histopathological samples with pSABER

In situ hybridization (ISH) is a powerful tool for investigating the spatial arrangement of nucleic acid targets in fixed samples. ISH is typically visualized using fluorophores to allow high sensitivity and multiplexing or with colorimetric labels to facilitate covisualization with histopathological stains. Both approaches benefit from signal amplification, which makes target detection effective, rapid and compatible with a broad range of optical systems. Here, we introduce a unified technical platform, termed 'pSABER', for the amplification of ISH signals in cell and tissue systems. pSABER decorates the in situ target with concatemeric binding sites for a horseradish peroxidase-conjugated oligonucleotide, enabling the localized deposition of fluorescent or colorimetric substrates. We demonstrate that pSABER effectively labels DNA and RNA targets in cultured cells and FFPE specimens. Furthermore, pSABER can achieve fivefold signal amplification over conventional signal amplification by exchange reaction (SABER) and can be serially multiplexed using solution exchange. Therefore, by linking nucleic acid detection to robust signal amplification capable of diverse readouts, pSABER will have broad utility in research and clinical settings.