New developments and novel applications in high throughput and high content imaging

Since the first microscopes of Van Leeuwenhoek about 300 years ago, microscopy has continued to provide novel insights about the fundamental elements of life. Images are by nature rich in information and recent times have seen an increased effort to move beyond their descriptive value and to quantify this information. Over the past twenty-five years, significant advances in computational power, robotics and automation, and image detectors have made possible a huge increase in the throughput and content obtainable with light microscopy methods. Though the modalities that have been developed as platforms for high-content (HC) and high-throughput (HT) imaging vary somewhat, the resulting output is essentially the same: morphological data obtained from a large number of fields of cells. The term “high-throughput microscopy” first appeared in PubMed relatively recently 1 and in this case it referred to the use of an image-based screen of a cDNA library to identify genes that induced altered nuclear/cytoaplasmic ratios of TORC1-eGFP. The following year, a chapter discussing the use of different screening strategies in RNAi based screens was the first use of the term “high-content microscopy” 2. In general, HC assays consider the morphological phenotypes of individual cells, rather than the average signals measured from all cells 3. Interestingly, the first example of imaging flow cytoametry (IFC) appears much earlier, in 1979 4. The authors developed a unit capable of detecting cellular particles, triggering a flashlamp if particles were in the preselected size range of interest, and feeding 16-mm film for acquisition, documentation, and image analysis. IFC has since been used to automate the analysis of many rare cellular events and those requiring data from large populations of cells, such as analysis of internalized bacteria 5, phagocytoasis 6, nuclear–cytoaplasmic translocation events 7, and yeast cell cycle analysis 8. Since this time the number of publications relying on HT/HC imaging methods has, predictably, grown exponentially (Fig. 1). This has led to a simultaneous explosion in the scale and content of image data and, most importantly, the type of scientific questions that can be asked when interrogating these datasets. These technological advances have brought imaging into the realm of “omics,” in which large datasets permit the quantification of molecular and morphological phenotypes that can inform as to the structure, function, and dynamics of a biological system. Number of publications by year for the listed PubMed search terms. There are essentially four areas in which technological advances have had a significant impact on HT/HC imaging: (i) Reagents and assay tools, e.g. advanced fluorescent probes, genome-wide RNAi libraries, and gene editing techniques (e.g., TALENs and CRISPR/Cas9); (ii) Microscopy and optical advancements, e.g. high sensitivity detectors (sCMOS cameras and GaAsp detectors) and superresolution microscopy; (iii) Automation and robotics, such as the robotic microscope 9 and commercially available HT/HC platforms such as the Operetta (PerkinElmer) and the IN Cell Analyzer (GE Lifesciences); and (iv) Computational advancements, since one of the major bottlenecks for most HT/HC imaging platforms is how to handle storage, processing, and analysis of ever larger datasets. Improvements in the speed and capacity of desktop computing, tools for the rendering and analysis of three-dimensional image data, and platforms for archiving and analyzing large image databases (such as OMERO, Bisque, and Cell Profiler) have all facilitated this. In this special issue, we will provide a broad snapshot of some of these rapidly evolving technologies, as well as some of the specific applications that have been facilitated by these tools. This will be a Special Issue in two parts; this first part includes 7 articles, 2 of which are reviews. Imaging cytoametry remains a strong approach for HT/HC imaging, and we have a number of articles focusing on this, including three original articles describing novel applications of this technique. Nascimento et al. (this issue, page 708) have developed a method for detecting rare mitochondrial fusion events in intact cells by IFC. As many recent findings have spawned increased interest in mitochondrial dynamics, this should prove to be a valuable new tool in the field. There are many potential clinical applications for HT/HC imaging, as pathology labs need to extract rare diagnostic markers from vast arrays of patient samples. Fuller et al. (this issue, page 720) have developed an Immuno-S-FISH method using IFC that makes possible the identification of specific chromosomal abnormalities from large populations of cells. This can be combined with phenotypic analysis, permitting the simultaneous screen for pathological morphologies in combination with the chromosomal determinants. Algorithm development is another critical aspect of these technologies, and Girault et al. (this issue, page 731) describe a novel algorithm for detecting cell edges that dramatically improves the precision of cell size detection in both single cells and cellular clusters in cells in flow. In another example of modifying specific imaging technologies to use for HT/HC, Kim and Grailhe (this issue, page 742) describe the adaptation of intramolecular bioluminescence resonance energy transfer (BRET) for HT/HC imaging of protein–protein interactions in live cells using an automated widefield microscope. In their Technical Note, Ong et al. (this issue, page 747) introduce Neuroncytoa II, a powerful software package for the automated quantitation of neurites. This solution has integrated a number of previously published methods for addressing the “crossover” problem of neurite quantitation that employed in combination, show a significant improvement in the accuracy of morphological analysis of neurons. As this Special Issue covers a relatively broad field of technological advancements, we have also included two review articles in order to permit a wider scope. From a more clinical perspective, Ang and Pethe (this issue, page 755) discuss the significant contribution of HC imaging to studying the cellular mechanisms of host–pathogen interactions and the development of therapeutics. These advancements in screening approaches promise to deliver specifically targeted drug hits for the rational and rapid development of effective antimicrobials. In further exploration of specific technical advancements, Grecco et al. (this issue, page 761) provide a review of how different multiplexed imaging approaches and the use of optical advancements in spectral unmixing and novel fluorescent probes have been applied to the study of intracellular protein networks. In controlling for cell-to-cell variability (due to cell type, cell cycle stage, environmental factors, etc.) multiplexed methods for monitoring multiple molecular components within single cells on an HT scale provide distinct advantages for understanding complex cellular processes. Just as with the earlier advent of IFC, the relatively recent adaptation of conventional microscopes to HT/HC instruments has significantly expanded their use as screening platforms. Both types of systems yield databases of cellular images which can subsequently be probed for complex morphological phenotypes on a population-wide level. The next step in taking biological imaging to the “omics level” will be to couple this data with the information gained from genomic, proteomic, and metabolomic datasets and provide additional depth and dimensionality to cellular informatics. This is starting to be realized as genome-wide expression maps, such as the Allen Brain atlases and the Berkeley Drosophila Transcription Network Project (BDTNP), have been developed that provide precise 3D biochemical map of protein expression at the organismal level. As the torrent of image data obtained by HT/HC imaging techniques swells, so too does the need for collaborative interactions between imaging scientists and bioinformaticists. In response to this, we have seen the recent emergence of Bioimage informatics, a new field poised to facilitate the refinement of these techniques, the development of best practices, and the continued advancement of resources 10, 11. We look forward to watching this field mature; we also hope you will enjoy this special Edition as much as we enjoyed putting it together.