Phone: 149-341-8652430, fax: 149-341-8651143
Improvements in high-throughput, high-content analysis of single cells†
Version of Record online: 20 MAR 2013
Copyright © 2013 International Society for Advancement of Cytometry
Cytometry Part A
Volume 83A, Issue 4, pages 331–332, April 2013
How to Cite
Tárnok, A. (2013), Improvements in high-throughput, high-content analysis of single cells. Cytometry, 83A: 331–332. doi: 10.1002/cyto.a.22285
The work presented in this editorial was made possible by funding from the German Federal Ministry of Education and Research (BMBF, 1315883).
- Issue online: 20 MAR 2013
- Version of Record online: 20 MAR 2013
- Manuscript Accepted: 27 FEB 2013
- Manuscript Received: 13 FEB 2013
High-throughput and high-content analysis of cultured cells or even small organisms are of imminent importance in drug discovery, toxicology and in individualized medicine. In medicine, the identification of tailored patient-specific therapy is important in cancer, chronic inflammation, infection and other diseases (e.g.1,2). Substances that are screened are traditional drugs like antibiotics or cytostatics but also new small molecule drug classes bioactive peptides, or RNAi. Frequently, analysis is done after treatment of the cells by end-point measurement of changes in the cell's function and morphology. Flow cytometry is, by nature, the high-content technology as it measures a multitude of markers simultaneously. Nevertheless, imaging and image cytometry are standard for high content-screening in many drug companies and in medicine because imaging needs lower amounts of biological material, is well suited for high-throughput and yields more information on the cell's morphology (3).
The entire work flow for high-content analysis requires great care in standardization of the pre-analytical procedures (cell culturing, stimulation, staining), a fully automated measurement process (wide field or confocal fluorescence microscopy) and fully automated data analysis and result presentation. In the past, high-content screening analyzed only one or a few fields of vision and the readouts were averaged for these fields in order to speed up measurement and analysis. However, this approach sacrifices cell resolution for increased throughput and does not yield information on the cell population's heterogeneity. Therefore, image cytometry is the method of choice for more detailed information on drug action on individual cells and changes in the cell population. Innovative technologies and workflows fulfilling this task are becoming increasingly available.
By fluorescence driven flow cytometry, cell cycle progression cell death onset can be characterized by 10 or more markers simultaneously. This high number of markers is even topped by the new technology of mass cytometry where the combination of 30 or more phenotypic and cell cycle markers can be analyzed in a single run (4). Image cytometry is still below this record but new approaches get closer to it. Particularly, Furia and colleagues from Milano, Italy report in two companion papers (this issue, Parts I & II on pages 333 and 344) about their success in developing a workflow for high-content fully automated analysis of the cell cycle. In the Part I paper (this issue, page 333), the authors describe their new microscopy platform and data analysis workflow termed A.M.I.C.O. (Automated Microscopy for Image-Cytometry). They adapted staining methods for the flow-cytometric detection of cell cycle distribution measuring total DNA content in combination with cell cycle checkpoint markers such as cyclins. The microscopic analysis combines image cytometry of the whole cell population using wide field microscopy. Subsequently, detailed target-driven analysis of individual cells can be performed using a confocal setup. The optimization method is detailed and includes different filter settings, labeling, microscope objectives, and image pre-processing, among others. The results are similar to those obtained by flow cytometry with regard to population distribution and quality of measurement (CV-values, resolution). The authors emphasize, that their approach is extendable to other high-content analysis including life cell imaging.
In their Part II companion work (this issue, page 344), Furia and colleagues performed with their image cytometry technique, a seven color assay for analyzing spontaneous DNA damage in cultured cells. The staining combined stoichiometric DNA labeling, proliferation monitoring by Ki67, and histone phosphorylation check-point activation markers p21 and p52. They demonstrate the combination of population-wide cytometric analysis with detailed co-expression profiling and structural information on the single cell bases for the foci of checkpoint-activation markers.
Microfluidic image cytometry is an alternative tool for high-content and throughput measurements of cells as it combines measurement principles of flow cytometry with imaging capabilities as presented recently for toxicological applications by Yoo and colleagues from Seoul, Republic of Korea (3). The group extended their previous work on MTT assays to fluorescence-based cell cycle analysis (this issue, page 356) and present the whole workflow for their experimental setup. Cells are cultivated in numerous parallel microchannels; treatment with toxins is done in the channels applying a toxin concentration gradient. Cells in the channels are measured by fluorescence microscopy along the concentration gradient. Therewith, cells treated with different toxin concentrations can be analyzed in a single microchannel in a single run. By simplification of the optical setup, this technology could become a cost effective way for high-content analysis requiring low amounts of reagents and producing minimal waste.
Tissue sections (5) and even whole small animals such as the zebra fish (6) are also important targets for high-content analysis and require an appropriate workflow. For improving the automated analysis of histopathological material, Heindl and colleagues from Vienna, Austria, and Cardiff, Wales United Kingdom (this issue, page 363.) developed new image analysis tools. The authors addressed the question of how to remove in silico, contaminating, highly autofluorescent events like erythrocytes from tissue section images by using their new tool, ARETE. As model tissues they investigated the efficacy of their approach on sections from human placenta and colon. Placental sections were stained for RAGE (receptor for advanced glycation end products), cytokeratin 7 and actin; colon for nuclear DNA, KI67 and keratin 8. For both tissue types, three different fluorochromes were applied. ARETE was “taught” on transmission images by machine learning to identify erythrocytes. Its results were in good agreement with those from human expert observers. This trained program then automatically and efficiently removed erythrocyte images thereby clearly reducing the rate of false positive cells.
In summary, high-content analysis of cell cultures and tissues by image cytometry has many facets and involves a great variety of methodology. This technology is more and more approaching the performance of flow cytometry with respect to throughput, complexity of measurements and data quality and is a useful alternative for cell population analysis.