In this issue
In this issue
Avoiding pitfalls in MRD detection by Flow Cytometry
Minimal residual disease (MRD) has emerged as a major prognosis factor to adapt the treatment intensity in patients with acute lymphoblastic leukemia (ALL). Flow cytometry (FC) measurement of MRD is becoming a very promising challenger to a molecular approach. It is based on the identification of leukaemia specific phenotypes (LAPs) that are minimally expressed on bone marrow normal B cells. Solly and coworkers provide useful information about the reliability of a new marker, CD304, expressed in 40% of B-ALL. CD304 is not expressed by normal B cells at every stage of medullar maturation and allows monitoring of the persistence of B lymphoblasts at low levels (10−4). It is well known that LAPs may vary during treatment but the authors show that CD304 expression on leukemic cells is stable during treatment and at relapse and thus constitutes a very promising marker for FC MRD assessment.
In this issue, page 17
Hierarchy of cell populations automatically
Hierarchy of gated populations is a standard way of looking at flow cytometry data. In this issue, Fišer and coworkers show how to build a population hierarchy automatically using hierarchical clustering analysis (HCA). They developed a new HCA algorithm and validated the approach by monitoring acute lymphoblastic leukemia samples, showing excellent concordance with standard gating approaches. HCA not only builds cellular hierarchy automatically from the single cell level, but also displays all measured parameters on a single image - heatmap. Seeing all events and all measured parameters on one plot helps significantly with selection of the right cell populations. To improve the method for minimal residual disease detection, the authors combined HCA with support vector machines learning (SVM). SVM was, after training on a diagnostic sample, able to fully and automatically dissect the MRD population in ALL follow up samples.
In this issue: page 25
Hyperspectral cytometry at the single-cell level
In their latest report, Grégori and coworkers further describe a new technology able to simultaneously collect 32 narrow bands of fluorescence (in addition to forward and right angle light scatter) from single particles flowing across the laser beam of a flow cytometer. These 32 discrete values provide a proxy of the full fluorescence emission spectrum for each single bioparticle in the absence of any dichroic or band-pass filters. Demonstration of the technology has been performed with fluorescent microspheres and human lymphocytes labeled with a cocktail of antibodies (CD45/FITC, CD4/PE, CD8/ECD, CD3/Cy5). This hyperspectral capability does not increase data-analysis complexity, as a number of well-established data reduction techniques can be employed for simple exploratory analysis and visualization. The authors were able to identify cell clusters without compensation using 32-channel spectral patterns formed by combination of the four fluorochromes as inputs for clustering and visualization.
In this issue: page 35
Quantification and Visualization of Cell Divisions in 3D Tissues
3D time-lapse microscopy is a powerful technique to non-invasively study dynamic biological processes in living tissues. Fluorescently tagged histones are widely used to track cells and visualize cell cycle phases. To measure and visualize the cell cycle dependent changes in DNA packaging in Drosophila embryos, Chinta and Wasser developed a 3D image segmentation method that addressed two challenges in image analysis: accuracy and speed. While many related studies focused on centroid detection, this work also validated the accuracy of volume detection which is crucial to monitor cell cycle progression. The authors showed that their method was robust to variations in morphology and brightness of chromatin and performed accurately throughout interphase and mitosis. Speed was sufficient to allow segmentation of large datasets on standard computers. The authors implemented their method as a stand-alone software application that is freely available to download.
In this issue, page 52
High throughput flow cytometry for proteomics study
A key focus of proteomics efforts is in the analysis of protein-protein interactions. For large-scale analysis, yeast two-hybrid (Y2H) systems have been the most commonly used method, particularly in conjunction with increased access to automation which overcomes time-consuming manipulations. While the Y2H array approach has been previously automated, it remains difficult due to intricate colony handling and replication requirements. Chen and coworkers report an integrated liquid handling Y2H array approach carried out in a 96-well plate format by employing high throughput flow cytometry analysis of the yEGFP reporter. This approach offers the advantages of automated liquid handling, quantitative reporter analysis, and self-activators detection. This high throughput flow cytometry system shows promise for future large-scale analysis of protein-protein interactions.
In this issue: page 90