Journal: Cell Reports Methods

Article Title: High-content imaging as a tool to quantify and characterize malaria parasites

doi: 10.1016/j.crmeth.2023.100516

Figure Lengend Snippet: 20× image analysis Parasites were stained with Hoechst, indicated in blue, for parasite nuclei and MitoTracker Deep Red (MDR), indicated in red, for functional mitochondria. Red blood cells (RBCs) were visualized by bright field and stained with CellMask Orange plasma membrane stain, which is false colored in purple. The bright-field channel was used to identify RBCs in this analysis, while CellMask Orange plasma membrane stain was only included to better visualize parasitized RBCs. Cells were imaged with a 20× air objective on an Operetta CLS. (A–F) Listed is (A) the workflow to quantify parasitemia using Harmony High-Content Imaging and Analysis software with PhenoLOGIC (v.4.9). The (B) bright-field input image was first filtered using a Gaussian smoothing filter with a width of 3.5 pixels (2.1 μm), yielding the output image (C) “Gaussian smoothed.” This image was then (D) inverted so that RBCs would be highlighted, yielding the output image “inverted.” Using the “inverted” output image, (E) RBCs were identified using “method M,” indicated by rainbow circles. Using bright-field intensity and cell morphology properties, (F) background and cells on a different focus height were excluded, indicated by pink circles, and RBCs were selected, filled in green. (G and H) From the selected RBC population, outlined in white, (G) Hoechst spots and (H) MDR spots were identified, circled in yellow. (I–K) RBCs with at least one Hoechst spot and at least one MitoTracker spot were classified as “infected RBCs,” indicated by yellow circles. Scale bar: 50 μm. To validate this image analysis, asynchronous GC03 and Dd2 parasites were treated for 72 h over a range of chloroquine (CQ) concentrations. For each replicate, parasitemia was determined by flow cytometry (FC) and high-content imaging (20×). At least 10,000 RBCs were analyzed per treatment concentration by each method. Shown are (J) dose-response curves and (K) mean IC 50 values ± standard error of the mean (SEM) from three independent experiments. No significant difference (ns) was detected between FC and high-content imaging using a Student’s t test.

Article Snippet: Harmony® High Content Imaging and Analysis Software with PhenoLOGIC™ Machine Learning (version 4.9) , PerkinElmer , Part# HH17000010.

Techniques: Staining, Functional Assay, Imaging, Software, Infection, Flow Cytometry, Concentration Assay

Journal: Cell Reports Methods

Article Title: High-content imaging as a tool to quantify and characterize malaria parasites

doi: 10.1016/j.crmeth.2023.100516

Figure Lengend Snippet: 40× image analysis Cells were stained with Hoechst, indicated in blue, for parasite nuclei; MDR, indicated in red, for respirating parasite mitochondria; SYTO RNASelect, indicated in green, for parasite cytoplasm; and CellMask Orange plasma membrane stain, indicated in purple, for RBC plasma membranes. Ten planes were captured for each field on an Operetta CLS using a 40× water objective to create a maximum projection. (A–D) Listed is (A) the workflow used to quantify parasite asexual blood stages with Harmony High-Content Imaging and Analysis software with PhenoLOGIC (v.4.9). To identify RBCs, the (B) input image was first filtered using a (C) median smoothing filter to smooth RBC edges. Then, a (D) Gaussian smoothing filter was applied to smooth RBC faces. (E) RBCs were identified using “method P,” using the smoothed output image. Identified cells are indicated in rainbow coloring. (F–H) Using cell morphology and intensity properties, background staining and doublet cells were excluded, circled in pink, allowing single RBCs to be selected, indicated in filled green. To identify RBCs containing viable parasites, (G) Hoechst spots and (H) MDR spots were selected using “method D.” The regions of interest (selected RBCs) are outlined in white, and selected spots are circled in yellow. (I) RBCs that contained at least one Hoechst spot and at least one MDR spot were classified as infected RBCs, circled in yellow. (J) Machine learning was used to classify these RBCs as infected with: single rings (SRs), circled in green; multiple rings (MRs), circled in red; single trophozoites (STs), circled in blue; multiple trophozoites (not shown); single schizonts (SSs), circled in pink; or multiple schizonts (not shown). Scale bar: 25 μm.

Article Snippet: Harmony® High Content Imaging and Analysis Software with PhenoLOGIC™ Machine Learning (version 4.9) , PerkinElmer , Part# HH17000010.

Techniques: Staining, Imaging, Software, Infection

Journal: Cell Reports Methods

Article Title: High-content imaging as a tool to quantify and characterize malaria parasites

doi: 10.1016/j.crmeth.2023.100516

Figure Lengend Snippet: Differentiation of asexual stages After establishing a training set for machine learning, “goodness” scores and scatterplots were generated with Harmony High-Content Imaging and Analysis software with PhenoLOGIC (v.4.9). The “goodness” score is a measure of signal-to-noise ratio, which indicates the separation, but not distribution, between two populations. The scatterplot provides a 2D depiction of the differences between. (A) Single rings vs. multiple rings; (B) single rings vs. single trophozoites; (C) single rings vs. multiple trophozoites; (D) single rings vs. single schizonts; (E) single rings vs. multiple schizonts; (F) multiple rings vs. single trophozoites; (G) multiple rings vs. multiple trophozoites; (H) multiple rings vs. single schizonts; (I) multiple rings vs. multiple schizonts; (J) single trophozoites vs. multiple trophozoites; (K) single trophozoites vs. single schizonts; (L) single trophozoites vs. multiple schizonts; (M) multiple trophozoites vs. single schizonts; (N) multiple trophozoites vs. multiple schizonts; and (O) single schizonts vs. multiple schizonts. Single rings, multiple rings, single trophozoites, multiple trophozoites, single schizonts, and multiple schizonts are shown in green, dark red, dark blue, yellow, pink, and light blue, respectively. Solid circles indicated manually selected parasites, and open circles indicated machine-selected parasites.

Article Snippet: Harmony® High Content Imaging and Analysis Software with PhenoLOGIC™ Machine Learning (version 4.9) , PerkinElmer , Part# HH17000010.

Techniques: Generated, Imaging, Software

Journal: Cell Reports Methods

Article Title: High-content imaging as a tool to quantify and characterize malaria parasites

doi: 10.1016/j.crmeth.2023.100516

Figure Lengend Snippet:

Article Snippet: Harmony® High Content Imaging and Analysis Software with PhenoLOGIC™ Machine Learning (version 4.9) , PerkinElmer , Part# HH17000010.

Techniques: Recombinant, Staining, SYBR Green Assay, Software, Imaging