integrin α5  (Addgene inc)

Journal: Nature Communications

Article Title: HCS-3DX, a next-generation AI-driven automated 3D-oid high-content screening system

doi: 10.1038/s41467-025-63955-5

Figure Lengend Snippet: a Representative images of the 2D annotated dataset, including monoculture (purple) and co-culture (green) spheroids with the smallest and largest volumes, and least and greatest sphericity values. The scale bar represents 50 µm. b Representative image of a monoculture spheroid composed of HeLa Kyoto cells (green - EGFP-alpha-tubulin; red - H2B-mCherry; grey - actin). Monoculture spheroids were generated in 3 independent experiments by 3 experts, resulting in n = 223 for 2D and n = 110 for 3D. The scale bar represents 50 µm. c Representative image of a co-culture spheroid comprising HeLa Kyoto and MRC-5 cell lines (blue - DAPI, green – EGFP-alpha-tubulin; red - H2B-mCherry; grey - actin). Co-culture spheroids were generated in 3 independent experiments by 3 experts, resulting in n = 203 for 2D and n = 114 for 3D. The scale bar represents 50 µm. d Assembled HCS plate in use. e 3D printed heat-resistant mould element that was used to form the FEP foil. f Vacuum-formed transparent FEP foil where each cuvette is suitable for one sample. For visualisation, each cuvette was filled with cell culture medium. g The insert element that fits into the cuvettes of the foil. The insert element only secures the position of the samples at the bottom of the cuvettes. h The grid element that fits into the base prevents the movement of the FEP foil after filling up with the detection solution. Two versions are available. i The base of the plate is an additional component that provides volume for the detection solution and space inside the plate for the water immersion objectives. Two versions are available. j Representative co-culture spheroids showing a spherical and an irregular example from the dataset. Images from left to right show the brightfield images with the corresponding annotation, the maximum intensity projection (MIP) images acquired with the light-sheet microscope, whole spheroid segmentation, and cytoplasm and nucleus segmentation combined with machine learning-based classification. The MRC-5 cells are illustrated in orange, while the HeLa Kyoto cells are illustrated in purple. The scale bar represents 50 µm.

Article Snippet: Spheroid monocultures were generated using HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry cervical cancer cells (Cell Lines Service; CLS-300670, Eppelheim, Germany).

Techniques: Co-Culture Assay, Generated, Cell Culture, Microscopy

Journal: Nature Communications

Article Title: HCS-3DX, a next-generation AI-driven automated 3D-oid high-content screening system

doi: 10.1038/s41467-025-63955-5

Figure Lengend Snippet: a Fluorescent images of T-47D spheroids stained with DRAQ5 were used as a reference to compare image quality. Randomly selected spheroids screened in the HCS plate (left image) and in Petri dish (right image) were visualised where images were selected from the top, middle, and bottom regions. Scale bars represent 100 µm. b To characterise the general blurriness of the whole z-stack, the intensity variance metric was used both for the inside circle and for the whole spheroid. Results are shown in a boxplot, showing that no significant differences were measured between the HCS plate and the Petri. n = 5 for each group. The Kolmogorov-Smirnov test was used. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001. c Preparation and screening times were measured for both techniques and displayed on a stacked bar plot. Starting from the darkest colour the 4 categories are Sample preparation, Change sample (this category does not apply to the HCS plate), Calibration, and Imaging. The total time was calculated for 10 samples for each method. d Example image of a cleared T-47D spheroid labelled with DRAQ5 for nuclei (purple) and Phalloidin 488 for actin (green) and optically cleared with Sucrose protocol. A total of 5184 nuclei were segmented using BIAS . Scale bar represents 100 µm. e By co-culturing T-47D breast cancer cells with MRC-5 fibroblasts and EA.hy926 endothelial cells, a hydrogel-based tumour model was created to track the tumour microenvironment in real time. CellTracker dyes (orange, deep red, and green) are used to detect cells. The tumour, endothelial, and fibroblast cells are marked by red, green, and white labels, respectively. Scale bar represents 50 µm. f A ground truth dataset of HeLa Kyoto - MRC-5 co-cultures was created using the HCS plate. Scale bar represents 50 µm. Box plots were visualised using Tukey’s method, with medians shown as central lines, boxes representing interquartile ranges, and whiskers extending to 1.5× interquartile range.

Article Snippet: Spheroid monocultures were generated using HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry cervical cancer cells (Cell Lines Service; CLS-300670, Eppelheim, Germany).

Techniques: Staining, Sample Prep, Imaging

Journal: Nature Communications

Article Title: HCS-3DX, a next-generation AI-driven automated 3D-oid high-content screening system

doi: 10.1038/s41467-025-63955-5

Figure Lengend Snippet: a Vertical projection of a spherical and irregular spheroid that displays the number of HeLa Kyoto (purple) and MRC-5 (orange) cells. The centre of each spheroid was defined using whole spheroid segmentation. Starting from the centre position, spheres with increasing radius were used to create shells and count the number of objects in each section. Line plots show the number of cells per class based on the shell analysis. b Spider plot representation of the Spearman’s correlation between 2D and 3D features. Paired spheroids ranked by Solidity 3D (grey, n = 34) are divided into a Spherical (dark blue, n = 18) and Irregular (red, n = 16) groups. c Correlation between 2D and 3D features ( n = 34). Previously displayed spherical (dark blue) and irregular (red) spheroid, along with the paired samples are displayed. Outliers (black and red) were not used for the evaluation. All scatter plots were generated using a fitted regression line, with shaded bands indicating the standard error of the mean (SEM). Source data are provided as a Source Data file.

Article Snippet: Spheroid monocultures were generated using HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry cervical cancer cells (Cell Lines Service; CLS-300670, Eppelheim, Germany).

Techniques: Generated

Journal: Nature Communications

Article Title: HCS-3DX, a next-generation AI-driven automated 3D-oid high-content screening system

doi: 10.1038/s41467-025-63955-5

Figure Lengend Snippet: a 3D visualisation of the Spherical and Irregular spheroids where each circle represents a cell displayed based on their position within the spheroid and class (HeLa Kyoto - purple and MRC-5 - orange). Each class is also visualised separately. b Collected 2D, 3D, and Neighbourhood (NH) features for both spheroids are displayed. c 3D representation of the Spherical and Irregular co-culture spheroids. Both top and side views show the nuclei of the HeLa Kyoto (purple) and MRC-5 (orange) cells. For visualisation, a square was placed around the spheroids indicating the side view. The scale bar represents 50 µm for the top and 100 µm for the side view. Leica’s LAS X microscope software was used to display the images.

Article Snippet: Spheroid monocultures were generated using HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry cervical cancer cells (Cell Lines Service; CLS-300670, Eppelheim, Germany).

Techniques: Co-Culture Assay, Microscopy, Software

Journal: Nature Communications

Article Title: HCS-3DX, a next-generation AI-driven automated 3D-oid high-content screening system

doi: 10.1038/s41467-025-63955-5

Figure Lengend Snippet: a Cell ratio for the 114 co-culture spheroids showed 72 spheroids with MRC-5 cells in the majority (orange), 28 spheroids containing a majority of HeLa Kyoto (purple), and 14 spheroids with an equal number (grey). Ratio of classified objects with the exact number are displayed. The 114 spheroids were divided into 3 groups based on the cell type majority (i.e. HeLa Kyoto Majority, Equal, MRC-5 Majority). The size of the doughnut charts and the numbers in their centre represent the average cell numbers for each group. The average number of cells for each class is displayed with orange for HeLa Kyoto and purple for MRC-5. b Comparison of the two cell lines based on Volume 3D (µm 3 ), the average inter-nuclear, and cell distance of the same class in µm. In this experiment, 114 co-culture spheroids were utilised, and average values were derived by aggregating the results across all constituent cells. Each group showed a statistically significant difference ( p < 0.001). c Features ( Solidity , Volume 3D , number of cells) collected from the spheroids with different cell ratio were displayed based on the 3 groups (HeLa Kyoto majority n = 28, Equal n = 14, MRC-5 majority n = 72). Dark grey colour represents the average value of the whole dataset ( n = 114). Statistically significant differences were identified in two cases, with p -values of <0.0023 (**) and <0.001 (***), respectively. d Spearman’s correlation of the total number of cells and Volume 3D were visualised based on HeLa Kyoto majority (purple), MRC-5 majority (orange), and Equal (grey) groups separately. The variation in sample size (n) across groups was a direct outcome of the cell ratio experiment. n = 114. For the statistical analysis, a two-sided non-parametric Kruskal-Wallis test was conducted, followed by Dunn’s multiple comparison test. When comparing only two groups, the two-sided Kolmogorov-Smirnov test was utilised. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001. Box plots were visualised using Tukey’s method, with medians shown as central lines, boxes representing interquartile ranges, and whiskers extending to 1.5× interquartile range. Source data are provided as a Source Data file.

Article Snippet: Spheroid monocultures were generated using HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry cervical cancer cells (Cell Lines Service; CLS-300670, Eppelheim, Germany).

Techniques: Co-Culture Assay, Comparison, Derivative Assay