Abstract

Since the late 1990s, high-content screening (HCS) approaches have contributed greatly to cell and chemical biology, drug discovery, and toxicology research. Despite the numerous advantages of multi-parameter interrogation of cell models using HCS approaches, these assay formats are also susceptible to artifacts and interference. This chapter describes 1) autofluorescence interference contributed by media, cells, and tissues which may complicate or preclude HCS assay development, 2) interference by environmental or microorganism contaminants, and 3) compound-mediated autofluorescence, fluorescence quenching, and cellular injury/cytotoxicity which may obscure whether compounds modulate the desired target or cellular phenotype or may produce false-positives or -negatives. It provides best-practice recommendations for identifying and mitigating such interference through the application of experimental design strategies, statistics, orthogonal, and counter-screens, and the use of reference compounds. Adoption of these recommendations should enhance the quality of chemical matter identified by HCS assays.

Abbreviations

DHT: dihydrotestosterone; ECM: extracellular matrix; FAD: flavin adenine dinucleotide; FRET: Förster/fluorescence resonance energy transfer; GFP: green fluorescent protein; HCS: high-content screening; HTS: high-throughput screening; IAF: image-based autofocus; LAF: laser-based autofocus; MOA: mechanism of action; NADH: nicotinamide adenine dinucleotide; RFP: red fluorescent protein; ROS: reactive oxygen species; SAR: structure-activity relationship; YFP: yellow fluorescent protein

Introduction

The selection of cellular assay format, experimental design, and components of the follow-up testing paradigm are all critical decisions that influence the ultimate success of a small-molecule high-throughput screening (HTS) campaign (1-5). Deploying a target-based or phenotypic screening approach requires the development, optimization, and validation of an assay that is biologically appropriate with a robust and reproducible signal window that has sufficient throughput and capacity to screen the compound library of interest (3). All HTS assays are subject to artifacts and interference from compounds and other sources, and high-content screening (HCS) assays are no exception (Figure 1) (1-3, 6-11). HCS assays detect perturbations in cellular targets, function, and phenotypic readouts, regardless of whether they arise from desirable or undesirable mechanisms. Additionally, since HCS imaging requires the transmission and reflectance of light for fluorescent signal generation and detection, optically active substances can alter readouts independent of a true biological effect.

Figure 1.

Summary of HCS artifacts and interference. Interferences can be broadly divided into compound-related and non-compound-related categories. Non-compound-related interferences include endogenous materials and contaminants. Compound-dependent interferences (more...)

HCS assays, like all cell-based assay approaches, are susceptible to interference from endogenous substances in cells, tissues, and culture media, as well as with the reagents and probes used for detection. When unaccounted for, the contributions of endogenous substances can preclude, adversely impact, or complicate the development of HCS assays, either by elevating fluorescent background levels sufficiently to make it challenging to detect bioactive responses, or by depressing or quenching fluorescent signals to an extent where they may be indistinguishable from background signal. Similarly, exogenous contaminants from the environment including lint, dust, plastic fragments from labware, pipette filter and lab coat fibers, and microorganisms may cause image-based aberrations such as focus blur and image saturation that can complicate downstream image analysis and impair the identification of subtle phenotypes (11). Exogenous contaminants may also perturb the underlying biology.

However, the major source of artifacts and interference in HCS assays are the test compounds themselves. Compound-dependent assay interference can be broadly divided into fluorescence detection technology-related and non-technology-related cell cytotoxicity or morphology issues, although there is considerable overlap between these two categories (Figure 1) (7).

Compounds that do not modulate the biological target or phenotype of interest but instead interfere with assay detection because of autofluorescence or fluorescence quenching may produce artifactual bioactivity readouts and/or phenotypes in HCS assays, or mask bioactivity depending on the assay construction (12-14). Similarly, compounds that alter light transmission or reflection by a variety of mechanisms, including quenchers, insoluble compounds, and colored (pigmented) compounds may also interfere with HCS assays (Figure 1).

Autofocusing methods employed by HCS instrumentation can affect the acquisition and identification of cell images, as well as impact data quality during image acquisition, segmentation, and post-processing steps. This topic is described in detail in later sections.

Compound interference or artifacts due to autofluorescence or quenching can often be identified by statistical analysis of the fluorescence intensity data because the values produced by these compounds will typically be outliers relative to the normal distribution ranges of these measurements in control wells not exposed to compounds and in wells treated with compounds that are optically transparent or inert (13-16). Compound interference flagged by statistical methods can then be confirmed by 1) manually reviewing the images, 2) implementation of orthogonal assays that utilize a fundamentally different assay detection technology, and/or 3) counter-screens for assay interferences or target selectivity.

Non-technology-related compound interference occurs when compounds alter or modulate the biology without directly interfering with the HCS fluorescence detection technology. For morphological profiling, these effects could be considered desirable, as they produce quantifiable effects on cells, but there are many examples where unique phenotypes result from undesirable and artifactual properties (17). The effects of such compounds are most frequently manifested as cellular injury/loss (“cytotoxicity”) or dramatic changes in cell morphology including cell shape, spreading and/or attachment (Figures 2, 3). Observations of cytotoxic dead cells may include loss of signal but more often as dead cells round-up, the fluorescence from some probes, e.g., nucleic acid probe concentrates fluorescence that may saturate camera or detector dynamic range directly comprising image-based autofocusing methods. Laser-based autofocusing methods are impacted by cytotoxic dead cell fluorescence. Selecting an optimal cell seeding density is one of the critical components of the assay development and optimization process for cell based HCS imaging assays (3,14,18,19). During HCS assay development, there are many variables that can impact the assay performance, including cell seeding density, media, poly-D-lysine (PDL), and extracellular matrix (ECM) or other microplate coatings, the objective magnification (4x, 10x, 20x, 40x, etc.), and the number of fields of view captured per well will determine the total number of cells that will be included in the image analysis. For more details on HCS assay development, please see Assay Development Guidelines for Image-Based High Content Screening, High Content Analysis and High Content Imaging (20). Compound-mediated cytotoxicity may obscure small-molecule activity at the target of interest or can be scored as false positives or negatives depending upon whether the assay is designed to identify inhibitors or activators (1,2,4-8,10). Similarly, compounds that substantially reduce or disrupt the adhesion of cells to assay plate surfaces can also produce significant cell loss which may invalidate the image analysis algorithm (3,13,17,18). For HCS assays, the robustness and reproducibility of the multiparameter data extracted by the image analysis algorithm typically depends upon the number of cells captured and analyzed. In most HCS assays there is a threshold number of cells below which the coefficients of variation (CVs) of the multiparameter data will increase dramatically and the corresponding Z-factor coefficients for the assay signal window separation between maximum and minimum control populations declines precipitously (3,13,17,18). An adaptive image acquisition process where multiple fields of view are acquired until a preset threshold number of cells have been imaged is one strategy that can mitigate the impact of compound mediated cell loss. However adaptive image acquisition may prolong the image acquisition process considerably to a point where it becomes prohibitive, especially if the screening library contains a high percentage of compounds that are cytotoxic or induce extensive cell loss due to altered adhesion properties. Additionally, adaptive image acquisition may not prove to be effective in wells where the compound mediated cell loss is substantial. Substantial reductions in the number of cells captured and analyzed may reduce or impair the statistical significance of the image analysis parameters being measured, and dramatic alterations to cellular morphology may eliminate or disrupt the ability of the image analysis algorithm to accurately segregate objects and/or regions of interest. Substantial loss of cells due to compound mediated cytotoxicity or loss of adherence can readily be identified by statistical analysis of the nuclear counts and nuclear stain fluorescence intensity data because the values of such compounds will be outliers relative to the others.

Figure 2.

Cytotoxicity compound interferences in HCS assays. (A) Schematic of interference in HCS assays by cytotoxic compounds. (B) Example of cytotoxicity encountered in an AR-RFP/TIF2-GFP biosensor HCS assay. Note that cells can be identified by including a (more...)

Figure 3.

Example of dead cell interference in HCS. Non-viable cells are typically round and on a higher Z-plane relative to most adherent cells. They are often out of focus, and consequently, they can produce “significant” morphologies. They can (more...)

Examples of undesirable compound mediated mechanisms of action / mode of action (MOAs) include nonspecific chemical reactivity, colloidal aggregation, redox-cycling, chelation, or denaturation mediated by surfactants or detergents (1,2,4,5,7,8,10). Other examples related to specific organelles or molecular pathways are lysosomotropic agents (cationic amphiphilic drugs, ‘CADs’), cytoskeletal toxins (tubulin poisons), mitochondrial toxins (electron transport chain poisons), and genotoxins (DNA intercalators, alkylating agents) (7). Of course, for the discovery of mechanism-based antineoplastics or antimicrobials, cellular injury may be the appropriate and intended phenotypic readout. Furthermore, reaction- or chelation-based MOAs with a high degree of specificity could represent viable chemical probes/drugs for some screens. Biological targets and systems can exhibit enhanced sensitivities to certain undesirable MOAs. For example, the actives identified in screening assays for targets with catalytic cysteine residues or metal cofactors may be enriched for electrophilic and chelating agents, respectively. In cellular models with redox-sensitive pathways, screening active lists may be enriched for oxidants or redox-active compounds if the pathway is sufficiently dominant in terms of phenotypic contribution (1). For any HCS imaging (or HTS) campaign, a testing paradigm of appropriate counter screens and orthogonal assays can be deployed to ascertain whether active compounds can be confirmed as hits or whether they should be eliminated or “flagged” from further consideration because they are cytotoxic and/or alter cell morphology.

This chapter describes compound-mediated interference that impacts the imaging detection approaches and biological/cellular integrity of HCS assays. Guidelines and strategies to identify and mitigate the impact of compound interference in HCS assays are presented together with tips for experimental design, helpful statistical methods for flagging potential interference, the use of reference interference compounds, and sample orthogonal and counter-screen protocols.

This chapter is organized into sections by the HCS imaging workflow process:

• Cells and media
• Seeding and growing cells
• Labels and probe choices
• Image acquisition considerations
• Analysis
• Observations and insights from compound treatments
• Confirmation and follow-up

Cells and media

There are several potential cell model sources of HCS assay interference that are independent of compound exposure. A variety of endogenous substances in culture media, cells or tissues, probes, and vessels or microplates can interfere with HCS fluorescent readouts that will need to be addressed during assay development (Figure 4). This section reviews examples of these endogenous sources of interference and recommends best practices to mitigate their impact.

Figure 4.

Endogenous sources of fluorescence. (A) Cells, tissues, and culture media contain a variety of endogenous substances that fluoresce in common fluorophore channels, often with relatively broad excitation and emission peaks. Adapted with permission (75). (more...)

Media components. Although tissue culture media is generally not a major source of artifacts, some components of culture media have fluorescent properties that can potentially interfere with image acquisition, analysis, and data interpretation (Figure 4A). Riboflavins (co-factors for flavoprotein enzyme reactions) fluoresce in the ultraviolet through green fluorescent protein (GFP) variant spectral ranges (ex. 375-500 nm and em. 500-650 nm) and within these ranges may elevate the fluorescent backgrounds acquired in live-cell imaging applications (Figure 4B, 4C). Typically, fluorescence interference contributed by media components is only a problem when the HCS assay signal of interest is relatively weak and falls within that spectral range. Media with lower riboflavin concentrations (e.g., F-12K, EBM) can be an alternative if the biology is not substantially perturbed (Figure 4D) (21). For short-term live-cell imaging, solutions lacking riboflavin such as HEPES-buffered HBSS or other commercially available products may be substituted. Protein supplements in media to maintain cell health and growth from calf, bovine, or other animal-derived serum contain rich mixtures of albumin, amino acids, and other growth factors and components that may increase background in fluorescent imaging experiments. Reducing the concentration of sera used in media can circumvent the risk of background fluorescence exposure.

Cells and tissues with high fluorescent background. Similarly, intrinsic autofluorescence of biological substances contained in normal and diseased tissues can sometimes interfere with fluorescence microscopy analysis of these samples. Cell types that are rich in pigment (e.g., retinal, melanocytes, etc.) can exhibit a high degree of light absorbance, light scatter, and/or autofluorescence that may confound HCS imaging approaches. Autofluorescence is often more apparent in highly metabolic cells such as hepatocytes, adipocytes, and connective tissues (22,23). Primary isolated hepatocytes are highly metabolic and intrinsically exhibit autofluorescence, leading to an increased fluorescent background in the blue-green part of the visible spectrum. This is particularly evident with excitation light sources between 360-488 nm and emission in the 500-550 nm range from metabolic redox cofactors NADH (nicotinamide adenine dinucleotide) and FAD (flavin adenine dinucleotide). Both can mask the detection of low-intensity fluorescent probe signals within these spectral ranges (24,25). The impact of cellular autofluorescence will depend on its relative intensity and distribution compared to the probe signal intensities and patterns being measured in the HCS assay. Cells/tissues with high autofluorescence coupled with relatively weak object staining intensities (e.g., low protein expression) may substantially reduce the effective signal window and sensitivity of the assay. To determine the extent and potential impact of endogenous autofluorescence in a particular sample, it is recommended that background images of “unstained and untreated” control samples without any reporters, probes, or dyes, together with secondary detection antibody-only controls, be acquired in the relevant HCS channels during assay development. Again, the degree of interference contributed by cell and tissue autofluorescence only constitutes a problem for HCS assay development when the signal of interest is relatively weak and falls within the same spectral range.

Cellular autofluorescence can be altered by cellular metabolism and perturbations that change the concentrations of intrinsically fluorescent molecules such as nicotinamide adenine dinucleotide (NADH), or that modify cellular proteins and metabolites such as lipofuscins (24, 25). Intrinsic fluorescent cellular molecules often exhibit heterogeneous expression levels and generally have broad excitation and emission spectra (Figure 4A). The amount of endogenous autofluorescence in cells is dependent upon cell type, their metabolic state, and the nature of the culture model. The measurement of cellular autofluorescence can sometimes be exploited to diagnose diseased tissues (26-30), and changes in autofluorescence as a function of disease state or treatment can provide “label free” opportunities to quantify treatment-induced effects. For example, dual monitoring of NADH and flavin adenine dinucleotide (FAD) fluorescence lifetimes in patient-derived breast cancer organoids generated from the primary cells of various breast cancer subtypes exposed to therapeutic treatments exhibited differential effects in proliferating cells at the periphery of the organoid versus the lower metabolic activity of cells in their cores (30). FAD and NADH fluorescence correlated with DNA damage after radiation treatment in primary cancer cells and immortalized cancer cell lines (31), while increased FAD and NADH green fluorescence were observed in bacteria, yeast, and human cells exposed to cytotoxic or chemo static treatments (33). An increase in cell or tissue autofluorescence produced by agents that induce DNA damage, cell death, or generate reactive metabolites could be misinterpreted as an increase in fluorescent signal from a reporter or probe being monitored in the same spectral range when the HCS assay signal is relatively weak.

Seeding and growing cells

Other potential sources of HCS interference include exogenous substances (“contaminants”) and microplate vessel materials or coatings (Figure 5A). This section reviews examples of potential contaminants and microplate effects, and recommended practices for mitigating their impact.

Figure 5.

Contamination interferences in high-content screening assays. (A) Summary of microbial, environmental and other contaminants. (B) Environmental contaminants in HCS. Note that lint fibers from laboratory coats or paper towels produce characteristic large (more...)

The selection of microplates and their composition is an important decision for HCS assay development. Black-walled microplates are generally preferred for fluorescence microscopy assays because of low autofluorescence and decreased well-to-well crosstalk from light scatter (34). The composition of different microplate bottom surfaces such as polymers (polystyrene, cyclic olefin copolymer, cyclic olefin polymer) can exhibit varying levels of autofluorescence and may have profound effects on cell attachment and spreading. Most microplate materials specialized for imaging applications are designed and fabricated to increase optical light penetration for transmittance as well as reduce autofluorescence and light scatter to improve image resolution.

Hydrogels and extracellular matrix reagents, such as Matrigel, are typically gelatinous substances containing proteins, growth factors, and other constituents that are commonly used as a substrate overlay for primary cells as a matrix for 2D cells or as a scaffold to support 3D cell models. Some Matrigel components naturally fluoresce but overall, they tend to exhibit low autofluorescence for most applications (35,36). Furthermore, the increase in overall thickness of Matrigel coating of cells or the resulting embedding of 3D cells can interfere with or impede light penetration, scatter, and absorbance that may reduce image quality (37). Synthetic hydrogels such as polyethylene glycol (“PEG”), and natural hydrogels such as hyaluronic acid and plant-derived alginates have also been reported to exhibit autofluorescence (38). Therefore, it is recommended that microplate supplemental coatings and matrices should always be assessed for optical interference and/or autofluorescence. In practice, this requires obtaining quality imaging microplates with desired coatings (PDL, collagen, fibronectin, laminin, hydrogel, etc.) and testing each cell model for autofluorescence and adherence under reference control conditions. In addition to cell adherence, coated surfaces can alter cellular characteristics (size, shape, morphology, texture), which should be thoroughly evaluated during assay development before screening. For additional details on microplates, types and choices for HCS imaging assays, consult the AGM chapter, Microplate Selection and Recommended Practices in High-throughput Screening and Quantitative Biology (34).

Labels and probe choices

The selection of fluorescent labels and probes is a critical component of HCS assay development that can also have a profound impact on the type and pervasiveness of interference in an HCS campaign. This section reviews potential sources of artifacts and interference related to the selection of HCS labels and probes, and recommended practices for mitigating their impact.

The number of compounds that show substantive autofluorescence is generally lower at red-shifted excitation and emission wavelengths (6,10). Selecting fluorescent reporters or probes with red-shifted excitation and emission wavelengths is one tactic that can reduce the incidence of compound-dependent autofluorescence interference in screening assays (9,43). Examples include mCherry and RFP derivatives (relative to GFP) and newer near-infrared (NIR) synthetic dyes. Red nuclear dyes such as DRAQ5 or Sytox Deep Red are accepted alternatives to UV-excited blue nuclear stains like DAPI and the Hoechst dyes.

Image acquisition considerations

The autofocusing method used during the automated image acquisition can impact image quality and data generation. In brief, all HCS imaging systems are equipped with autofocusing hardware and software components to rapidly locate and identify objects of interest in the correct focal plane. The two most common autofocus methods are laser-based autofocus (LAF) or image-based autofocus (IAF). It will be important for HCS Practitioners to match microplate types with AF routines for proper focal plane of cells, otherwise, out-of-focus images may be captured affecting data generation. For additional details on image autofocusing, consult the Assay Guidance Manual chapter Assay Development Guidelines for Image-Based High Content Screening, High Content Analysis and High Content Imaging (20).

The detection and compensation of certain types of interference may be specific to one image acquisition process but not to another. Compounds that exhibit autofluorescence or quenching in any of the acquired fluorescent channels may impact the image analysis algorithm from accurately segmenting the images in those channels to define objects and/or regions of interest and to extract the desired quantitative data. The impact of compound autofluorescence or quenching on ‘normal’ image acquisition results from various factors, including out-of-focus images, segmentation algorithm failures, and obscuration of small-molecule activity at the target of interest. Depending on the desired activity profile (activation or inhibition), this can lead to false positives or negative scoring. This is observed more often in IAF acquisition processes when an identifiable artifact is in one or more focal planes, creating problems previously mentioned.

Artifacts and autofluorescence from compounds will rarely directly affect LAF image acquisition. However, technical challenges arise from interferences of the microplates or vessels used to host cell models. If these devices are too thick or thin, or constructed with components that impede NIR light scatter, then these can create unwanted out-of-focus images. Selection of a suitable high-quality microplate is therefore a critical component of the assay development process. Additionally, the use of different LAF thresholding strategies (i.e., single, or double peak) may be useful to identify the correct focal plane containing the cells of interest. And in some cases, a combination of LAF and IAF can be used. The use of transmission brightfield imaging on the HCS imager or conventional microscope can augment the fluorescent image acquisition approach for morphological evaluation and can be very useful to help determine if the source of the artifact is lint, microbes, compound precipitant, or other debris. This label-free approach has many advantages including no interference from fluorescent molecules or compounds, and with the advances in artificial intelligence (AI) and machine learning (ML) with improved segmentation of a brightfield channel, offers value-added information from cell images.

The following are some best practices to enable autofocusing strategies for imaging:

• Use high quality validated microplates or vessels intended for HCS imaging experiments.
• Determine the focal plane (i.e., offset) of the cell model to optimize the image autofocus routine.
• When necessary, such as natural product or other suspect libraries, collect transmission brightfield images to augment and confirm potential sources of artifacts.

Special considerations for live-cell imaging

Compound autofluorescence can present unique challenges for live-cell imaging, mainly because cells are continuously exposed to compounds throughout the experiment (Figure 6A).

Figure 6.

Compound-mediated interferences in live-cell imaging assays. Since compounds are not removed during the assay, live-cell assays are more susceptible to certain types of interference. A major clue of compound-mediated light interferences are positive readouts (more...)

• Poorly soluble compounds can precipitate and form crystals or fine precipitants that can adhere to microplate surfaces (Figure 6B, C).
• Large, circular fluorescent patterns can form at the compound delivery site (Figure 6C). These artifacts can sometimes be addressed by adjusting the mechanics of the acoustic or pin tool dispensing system including xy centering or offset and Z-height, including adding a gentle mixing step, or optimizing compound concentration/solvent/delivery volume.
• More prominently, compounds that fluoresce in the fluorophore channels can increase background fluorescence. Even with various background subtraction and segmentation smoothing filter methods, fluorescent compounds can still produce localized signal within cellular compartments or entire cells (Figure 6D).
• Cellular debris from damaged cells can also interfere with certain automated confluence measurements (Figure 6E). This phenomenon is associated with surfactants, such as saponins, or protein precipitants, such as tannins. Cellular debris interference from cytotoxic compounds is usually less of an issue in assays that include washing steps prior to fixation.

In live-cell imaging, continuous or pulse exposure of fluorescent light to cells, compounds, or other constituents in the well may inadvertently result in cytotoxicity from photon exposure (phototoxicity) or photo-bleaching. To reduce potential issues from excitation light during live-cell imaging, reduce the number of times the same field of view is imaged. Assay development parameters should be established, typically for fluorescent proteins, lower excitation power with longer exposure times results in lower photobleaching than high energy excitation power and shorter exposure times.

Analysis

The sheer volume of feature data generated from HCS necessitates automated image and data analyses. In addition to helping identify bioactive compounds, well-designed analyses are crucial for flagging potential artifacts and interferences amongst the large number of images generated by HCS. These approaches are critical for prioritizing the highest quality compounds for follow-up and the design of appropriate orthogonal experiments. This section describes guidance for image analysis, recommended statistical methods and HCS readouts for outlier analyses, and cheminformatic tools for flagging potential interferences.

Types of readouts

Assay endpoints can be simple (object number, target intensity such as for a kinase site-sensitive antibody), algebraic (nuclear/cytoplasmic ratio, cell cycle distribution) or complex (high-dimensional algorithms that include texture measurements). Fluorescence interference can affect all of these, as well as how such measurements are presented. In terms of increasing subtlety:

• Object counts. Fluorescent artifacts such as compound precipitates may be identified as a large number of fluorescent objects compared to controls in a given imaging channel if they fall within the object identification parameters (typically pixel area but can include shape parameters as well).
• Outliers among scaled measurements. Autofluorescent and quenching artifacts may produce unintended outliers from high background fluorescence intensity measurements. For example, the mean and integrated object fluorescent intensities in each HCS channel can be calculated, and compounds that result in higher Z-scores (e.g., > +3.0) would be flagged as likely autofluorescent compounds. Conversely, compounds that result in lower Z-scores (e.g., < -3.0) would be flagged as likely light absorbing/fluorescent quenching compounds (Figure 7).
• Well-level readouts versus object-level readouts versus image-level analyses. Analyzing readouts at the object level (e.g., per nucleus, per cell, per colony or cluster) can help to reduce the impact of background signal from fluorescent extracellular compounds. Measuring the fluorescence per well typically calculates summary statistics (mean, median, etc.) for the objects in a well and shares the same advantages for eliminating aggregate and contamination-based fluorescent artifacts. Image-level analysis (object- or segmentation-free analysis of the image), such as deep-learning algorithms, will include fluorescent signal from extracellular compounds or other fluorescent substances. Also note that for object-based assays, post illumination flatfield correction to remove vignetting effect, algorithms that use background correction may inappropriately offset the intensity measurement if autofluorescent debris is outside the defined object area. Careful assessment of reference control wells to established background intensity parameters should be verified.

Figure 7.

Example of using Z-score to identify compound interferences. Z-scores can be calculated for HCS features/readouts that may be informative for particular interferences (fluorescence, brightness; cytotoxicity, nuclei counts; quenching; dimness). (Left) (more...)

Observations and insights from compound treatments

Compounds are the major source of HCS artifacts and interferences. The selection of test compounds, details of how they will be tested, and the inclusion of compound controls can all have significant downstream consequences in HCS not just in the general context of a screening campaign, but with regards to how fluorescence interference is detected and addressed. This section describes the two main categories of compound dependent HCS interference: compound autofluorescence and quenching, and nonspecific cytotoxicity or reduced cell adherence. Also described are reference compounds that can be used to identify or flag potential sources of compound interference in HCS assays.

Interference magnitude

When viewing HCS images and analyzing for autofluorescence interference, it is critical to consider the magnitude of fluorescence interference (Figure 8A). For compounds with low intrinsic fluorescence, the potential for interference increases if the corresponding dye is weak (Figure 8B). The dynamic range of fluorescent proteins (particularly in the green range) is extremely wide, so robust options are typically available (48). Weak staining also occurs when only suboptimal primary antibodies are available for a target protein, and for cases where expression of a labeled target protein is deliberately kept to the lowest level achievable (for example keeping the expression of a GFP-fusion protein low to minimize effects in the cell that may include results from a subpopulation biological response). HCS assays with large signal intensities should be inherently more resistant to fluorescence interferences. It is also important to note that some compounds (or their associated impurities) are strongly fluorescent, with intensities far above those of the fluorescent dyes. These can be flagged as significantly higher well-level and object-level intensities (e.g., by Z-score) and most likely affecting other features (such as texture) if these are part of the assay metric.

Figure 8.

Impact of compound fluorescence depends on magnitude of interference. (A) Summary of fluorescent compound interference interpretation. Due to the high sensitivity of modern imagers, the magnitude of compound fluorescence should be interpreted with respect (more...)

Direct compound-mediated fluorescent artifacts

Compound autofluorescence can result in image sets that are out of focus, create image analysis failures, may obscure small-molecule activity at the target of interest, and can be scored as false positives or negatives depending upon the desired activity profile, activation or inhibition. This is a significant issue because many HCS assay formats measure an increase in fluorescent signal intensity, while spectroscopic profiling of small-molecule HTS libraries have shown that about 10% of compounds are fluorescent (8,10). As a practical example, the EPA ToxCast compound library was screened in a 3 channels (blue, green, & red) biosensor assay for agonist and/or antagonist activity against DHT-induced androgen receptor protein-protein interactions with transcription intermediary factor 2 (TIF2/SRC2) a p160 coactivator (13,14) and was evaluated for fluorescent artifacts as part of the hit evaluation process (19). A detailed review of this process is included as Appendix 2. A key lesson learned from this study is that fluorescent anomalies resulting from cytotoxicity are frequent and their discovery can be made through reviewing wells with increased fluorescence in each channel and a correspondingly high Z-score.

In some situations, it may be possible to identify compound autofluorescence by incorporating a fluorescent pre-read of the compound library in a plate reader (or low magnification whole well or wells imaging) at the defined excitation and emission wavelengths for the channels being acquired prior to compound transfer into the HCS assay plate. However, there are numerous examples where autofluorescence only becomes detectable and/or apparent after compounds enter cells and bind to proteins, nucleic acids, and/or lipid macromolecules. The use of counter-screens to identify fluorescent artifacts usually involves cells treated with compounds but omitting the stains, probes and/or other reporters, this may include a counter-screen to identify cellular fluorescent artifacts by using a control cell line that lacks the biology or reporter of interest. It follows that compounds that fluoresce in the HCS imaging channels in the absence of stains/reagents would do so by an autofluorescent mechanism (Figure 9A). This would not be practical for an HCS campaign of compound libraries greater than a few hundred compounds but could be used to confirm autofluorescence identified by statistical methods. Compounds that fluoresce in cells can constitute between 2.5% (blue) to 0.5% (red) of a screening library, depending on the wavelength (6). In general, the incidence of fluorescent artifacts decreases with red-shifted (higher wavelength) channels. The exact impact of autofluorescent compounds ultimately depends on the assay format. In gain-of-signal assays, fluorescent compounds can present as false-positives (Figure 9B). Fluorescence will usually exhibit concentration-dependence, so in general decreasing compound concentrations will help mitigate this interference, though this must be balanced with the desire to identify weakly potent actives. As such, primary HTS campaigns are frequently run at compound concentrations of 10 - 20 µM (Figure 9C), oftentimes as a matter of policy and experience by the lab, and by the hit follow-up/confirmation plan post-HTS. The distribution of fluorescent compounds in cells varies between compounds, from nuclear patterns (often intercalators) to cytoplasmic to punctate (Figure 9D) and is difficult to predict reliably.

Figure 9.

Fluorescent compound interferences in HCS assays. (A) Schematic of interference in HCS assays by fluorescent compounds. Fluorescent compounds will produce “active” readouts in the absence of stains/reporters, and are usually more prevalent (more...)

Some factors affecting the incidence of autofluorescent compound interference include:

• The concentration and spectral properties of a given compound.
• The spectral properties of assay reagents and image acquisition settings.
• The intensity of the assay readout relative to the magnitude of compound fluorescence.
• The potential for the interfering agent to interact with one or more cellular components, increasing the fluorescence intensity or altering the spectra of the fluorescence.

In an HCS campaign to identify inhibitors of miR-21 biogenesis (8), all 1,130 primary hits from a screen of 315 K compounds were found to optically interfere with the enhanced green fluorescent protein (EGFP) reporter. The fluorescence interference was determined by counter-screening with the primary cell line minus the EGFP reporter. Methods for predicting compound fluorescence based on chemical structure are imperfect, especially in cellular contexts.

In addition to direct autofluorescence, some compounds that undergo degradation or are metabolized in situ can become fluorophores. This challenge is growing for image-based screening because morphological changes in assays like Cell Painting rely on defined pixel measurements of fluorescent regions within cells (49). Interfering fluorescent compounds can obscure these regions of interest, making it difficult to distinguish them from cell compartment bioprobes. As an example, the pro-drug pentamidine (DB289), a therapeutic for African sleeping sickness, has been reported to produce a cascade of five metabolites, including the product (DB75) which is highly fluorescent in the green spectrum (50,51).

Quenchers

Broadly defined, fluorescent quenchers are compounds that attenuate fluorescence intensity of a given substance, and this phenomenon can occur by a variety of fundamental mechanisms such as chemical and physical quenching (Figure 10A), or by direct competition at the binding site for the fluorescent probe in use. In FRET formats, quenching can occur at one or more steps of the signal cascade between donor fluorophore excitation to acceptor fluorophore emission. The exact mechanisms of FRET fluorescence interferences are usually not determined during HCS triage.

Figure 10.

Quenching compound interferences in HCS assays. (A) Schematic of interference in HCS assays by quenching compounds. Quenching compounds will produce “active” readouts in loss-of-signal assays by attenuating the expected fluorescence intensity (more...)

In HCS assays, quencher detection may already be “built into” the assay and compounds that reduce the intensity of nuclear stains such as DAPI, DRAQ5, or Hoechst are flagged as potential quenchers (Figure 10B). Counter-screens for quencher artifacts usually involve brief compound treatment times to minimize the chance of biological perturbation. It follows that compounds that lead to reductions in fluorescent intensities would do so by quenching. Like fluorescence, quenching will usually exhibit concentration-dependence, so in general decreasing compound concentrations will help mitigate this interference.

Some factors affecting the incidence of quenching interference include:

• The concentration and associated fluorescent spectral properties of a given compound quencher.
• The spectral properties of assay reagents and image acquisition settings.
• The intensity of the assay readout relative to the magnitude of compound quenching.

Quenchers have been used as tools for many cell-based assays including live-cell imaging for decades as they can be used to remove or reduce non-cell-based fluorescent signal. Many quenchers used in live-cell imaging are cell impermeant and therefore only block extracellular fluorescent signal, thereby effectively increasing the contrast of cell-specific signaling. While these quenchers can have great utility in live-cell imaging, small-molecules with quenching activity could prevent detection of a fluorescent signal and be considered a false positive or false negative depending on the nature of the HCS assay.

For additional details on compound-mediated light interferences, consult the Assay Guidance Manual chapters Interference with Fluorescence and Absorbance (52) and Compound-Mediated Assay Interferences in Homogenous Proximity Assays (53).

Compound informer sets

A practical approach to investigate phenotypes of interest for possible nuisance behaviors is the use of compound informer libraries. In many cases, such compound collections are composed of compounds with well-characterized activities to explore signaling pathways (‘MOA-box’) and previous clinical use (repurposing libraries) (54). In the context of assay artifacts and interference, such an informer set is a small collection of prototypical interference compounds (Figure 11) (7). Employing this approach from the onset helps to estimate the likelihood of specific interferences for a particular assay and characterizes the interference phenotypes which are difficult to predict a priori. In this way, screening compounds that produce similar phenotypes as informer nuisance compounds are clustered and can be triaged or investigated by additional counter-screens. Many industrial screening operations employ nuisance informer sets (55). The following are additional tips for using nuisance compound informer sets:

Figure 11.

Schematic for the design of an HCS nuisance compound informer set. It is recommended to use multiple chemotypes per interference MOA, multiple analogs per chemotype, and testing at several concentrations. Note that compounds can both interfere with the (more...)

• Informer sets can include technology-related light generated interference compounds (fluorescent compounds, quenchers, light scattering compounds) and non-technology interference compounds (nonspecific electrophiles, aggregators, redox cyclers, surfactants, and various cytotoxins), among other classes.
• An example informer set has been reported (see Supplemental Information of reference (7)).
• Test compounds at multiple concentrations, as interferences are often concentration-dependent and can vary from assay-to-assay.
• Include multiple compounds (both analogs and distinct chemotypes) from each interference class, and inactive analogs when possible.
• Interference informer sets can be used during the assay development/validation phase and/or during the primary HCS.
• We recommend including high-quality chemical probes and reference compounds in parallel, since nuisance and “on-target” phenotypes can potentially overlap.

An informer collection focused on known fluorescent compounds, particularly across spectral wavelengths and mechanisms (such as those that change fluorescence as a function of their local environment) could alert a group on the sensitivity of an assay to fluorescence interference and which types of interference are most difficult to deconvolute.

Confirmation and follow-up

After identifying compounds of interest by appropriate data analyses, it is crucial to perform a series of experiments on HCS-active compounds to dually select for desirable compounds and triage undesirable compounds. This screening cascade should assess for artifacts, irreproducible hits, and undesirable MOAs, and triaging as indicated. It should include orthogonal assay formats and counter screens designed and selected to confirm activity at the target or phenotype of interest, evaluate selectivity and/or specificity, and evaluate cytotoxicity. This section details experimental counter-screens for compound fluorescence and quenching in HCS assays and summarizes other high-yield secondary assays in the screening cascade for these purposes.

Optical interference counter-screens

A recommended counter-screen for fluorescent compound interference involves testing active compounds in the absence of stains, labels, or reporters (Figure 12A). For example, in a fluorescent protein assay (i.e., GFP), one can use a parental cell line without transfecting the GFP reporter (19,48). A recommended counter-screen for quenching compound interference involves testing active compounds in the presence of stains or reporters, but with short incubation times to preclude actual bioactivity (Figure 12B). Both significant fluorescent and quenching compounds should generally show similar readouts as the primary HCS readout.

Figure 12.

HCS assay counter-screens for compound fluorescence and quenching interferences. (A) Example workflow for fluorescence compound interference. Hits from the primary HCS are carried through the primary HCS procedure, except that the stains/reporter reagents (more...)

An overview of these process steps including technical information is highlighted in Table 1.

Table 1.

Generic protocol for fluorescence compound interference counter-screen in high-content screening assays

Orthogonal assays

Compounds active in an HCS assay (especially a primary screen) should have their bioactivity confirmed in at least one orthogonal assay. The purpose of orthogonal assays is two-fold: to de-risk for technology-based interferences and to enhance overall confidence in the original bioactivity. An orthogonal assay should involve a different technology that is not susceptible to the same technology interferences and is usually lower throughput but more sensitive and specific. For example, simply swapping a GFP reporter with a mCherry reporter would not constitute a high-quality orthogonal assay, since both could be susceptible to fluorescent interferences. For HCS assays, potential orthogonal methods include more direct measurement of specific pathway components (e.g., immunoassays, mass spectrometry) or gene expression profiling of specific pathway components (e.g., qPCR).

Can fluorescent compounds still be viable HCS hits/leads or why do technology interferences persist despite washing steps? These and other frequently asked questions are summarized in Table 2.

Table 2.

Frequently asked questions about HCS artifacts and interferences

Conclusions

Interference and artifacts can be a significant burden in the application of HCS for chemical probe and drug discovery. Endogenous autofluorescent substances found in cells, tissues, and culture media may prevent the development and optimization of an effective HCS assay, especially when the specific signal of interest is not robust. Consistent with other cell based HTS assay formats, contaminants such as microbes and airborne particles can also represent a substantial interference problem for HCS assays. Autofluorescent compounds confound HCS assays by interfering with both the detection and analysis components of the technology (usually optical based), while other compounds compromise the assay by altering cell morphology, reducing cell adherence, or inducing cytotoxicity. Several best practices that can mitigate interference in HCS assays include the use of sound tissue culture practices, careful technical decisions, and including the measurement of cell number into the primary HCS experiment. Recommended practices for identifying interference in HCS assays are conceptually like best practices in biochemical or other cell based HTS assay formats and include the use of statistical tools such as Z-scores, counter-screens for optical interference (quenching, autofluorescence), informer sets that incorporate known interferants, and orthogonal assays. The use of these best practices should lead to more efficient identification of higher-quality chemical matter from HCS assays.

Appendix 1: Derivation of the Z-score

A relatively straightforward and simple way to identify compound autofluorescence is to apply a simple Z-score statistical data processing analysis to the fluorescence intensity data collected for each of the channels acquired (13-17, 19, 20, 59-71). The Z-score (Equation 1) is a plate-based statistical score that is calculated only from the sample values (no controls) on an assay plate (16, 17, 66).

Where X_i is the raw measurement on the i^th compound, $\bar{X}$ and s_x are the mean and standard deviation (SD) of all the sample measurements on a plate, respectively. A deviation of ±3 s_x from the sample average $\bar{X}$ is frequently utilized as a statistical threshold or cut-off for active compounds. The method assumes that in a randomly distributed diverse compound library, fluorescent compounds are relatively rare, and that most of the compounds on a plate are non-fluorescent and serve as controls. Compound measurements are rescaled relative to within-plate variation by subtracting the average of the plate compound well values and dividing the difference by the SD estimated from all sample measurements of the plate. Z-score analysis of multiparameter HCS data provides a statistical method to identify compounds that are auto-fluorescent in any of the acquired channels or that are cytotoxic or disrupt cell adhesion to substantially reduce the numbers of cells that are acquired and are available for analysis. (13-15, 19, 20, 59-66).

Appendix 2: Analysis of cytotoxic and autofluorescent compounds in the ToxCast compound collection

It is possible to identify compounds that are likely to be fluorescent artifacts using standard HTS screening data processing methods, including the Z-score analysis (see appendix 1, AGM chapter or primary ref (16).

Glossary

artifacts

Compounds that modulate an assay readout by interfering with some component of the assay technology, as opposed to modulating the intended target.

autofluorescence (compound)

Emission of light from compounds when they have absorbed light; this is unrelated to modulation of the assayed biology.

compound interference

Compounds that confound the interpretation of a bioassay readout by interfering with the assay technology or perturbing the assayed biology by undesirable mechanism(s) of action, ultimately impeding the identification of quality chemical matter.

cytotoxicity

The property of a compound to cause injury to cells; can include a variety of cellular fates including necrosis, decreased proliferation, or apoptosis.

high-content screening

Cell- or organism-based imaging assays that measure multiple parameters from the same set of images; often performed with multiple fluorophore channels.

nuisance compounds

Compounds with apparent bioactivity that is often due to interference with the assay technology and/or modulation of the assayed biology by undesirable mechanisms of action.

phenotype

The composite observable characteristics of a cell or organism; can include morphology, biochemical composition, and/or physiological properties.

promiscuous compounds

Compounds statistically-enriched in bioactivity across a variety of unrelated biological assays.

quenchers (fluorescence)

Compounds that decrease the observed fluorescence intensity that is unrelated to modulation of the assayed biology; can occur by different chemical mechanisms such as FRET, collision, and static processes.

Suggested readings (alphabetical order)

1. JL Dahlin, DS Auld, I Rothenaigner, S Haney, JZ Sexton, JWM Nissink, J Walsh, JA Lee, JM Strelow, FS Willard, L Ferrins, JB Baell, MA Walters, BK Hua, K Hadian, BK Wagner. Nuisance compounds in cellular assays. Cell Chem. Biol. 2021, 28 (3), 356-370. PMID 33592188.

A summary of nuisance compounds in cellular assays from academia, industry, and government experts. Includes recommended practices for mitigating and identifying cellular nuisance compounds, including those found in HCS assays.

2. G Ibáñez, PA Calder, C Radu, B Bhinder, D Shum, C Antczak, H Djaballah. Evaluation of compound optical interference in high-content screening. SLAS Discov. 2018, 23 (4), 321-329. PMID 28467117.

An informative case report demonstrating the high hit rate of artifactual autofluorescent compounds in a GFP reporter readout in a phenotypic HTS.

Suggested resources

1. InterPred: Prediction of Chemical-Assay Interference.

An open-source cheminformatics tool for predicting compound fluorescence in blue, green, and red fluorescent channels. QSAR models are based on several compound fluorescence counter-screens with and without cells using the Tox21 library. Users can submit compound queries by SMILES (56).

Acknowledgements

Authors acknowledge the following financial support: BKW (Ono Pharma Foundation; NIH High-End Instrumentation Program, S10-OD026839; NIDDK, U01-DK123717). The Assay Guidance Manual is supported by the Intramural Research Program of the National Center for Advancing Translational Sciences at the National Institutes of Health (ZIA TR000340). The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

Conflicting interests

All authors hereby declare no conflicting interests pertaining to the material in this manuscript.

References

1.

Johnston PA. Redox cycling compounds generate H2O2 in HTS buffers containing strong reducing reagents—real hits or promiscuous artifacts? Curr Opin Chem Biol (Internet). 2011 Feb (cited 2023 Sep 18);15(1):174–82. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S1367593110001584. [PMC free article: PMC3040250] [PubMed: 21075044]

2.

Inglese J, Johnson RL, Simeonov A, Xia M, Zheng W, Austin CP, et al. High-throughput screening assays for the identification of chemical probes. Nat Chem Biol (Internet). 2007 Aug (cited 2023 Sep 18);3(8):466–79. Available from: https://www​.nature.com​/articles/nchembio.2007.17. [PubMed: 17637779]

3.

Johnston PA, Johnston PA. Cellular platforms for HTS: three case studies. Drug Discov Today (Internet). 2002 Mar (cited 2023 Sep 18);7(6):353–63. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S1359644601021407. [PubMed: 11893544]

4.

Thorne N, Auld DS, Inglese J. Apparent activity in high-throughput screening: origins of compound-dependent assay interference. Curr Opin Chem Biol (Internet). 2010 Jun (cited 2023 Sep 18);14(3):315–24. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S1367593110000463. [PMC free article: PMC2878863] [PubMed: 20417149]

5.

Thorne N, Inglese J, Auld DS. Illuminating Insights into Firefly Luciferase and Other Bioluminescent Reporters Used in Chemical Biology. Chem Biol (Internet). 2010 Jun (cited 2023 Sep 18);17(6):646–57. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S1074552110001973. [PMC free article: PMC2925662] [PubMed: 20609414]

6.

Borrel A, Huang R, Sakamuru S, Xia M, Simeonov A, Mansouri K, et al. High-Throughput Screening to Predict Chemical-Assay Interference. Sci Rep (Internet). 2020 Mar 4 (cited 2023 Sep 18);10(1):3986. Available from: https://www​.nature.com​/articles/s41598-020-60747-3. [PMC free article: PMC7055224] [PubMed: 32132587]

7.

Dahlin JL, Auld DS, Rothenaigner I, Haney S, Sexton JZ, Nissink JWM, et al. Nuisance compounds in cellular assays. Cell Chem Biol (Internet). 2021 Mar (cited 2023 Sep 18);28(3):356–70. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S2451945621000477. [PMC free article: PMC7979533] [PubMed: 33592188]

8.

Ibáñez G, Calder PA, Radu C, Bhinder B, Shum D, Antczak C, et al. Evaluation of Compound Optical Interference in High-Content Screening. SLAS Discov (Internet). 2018 Apr (cited 2023 Sep 18);23(4):321–9. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S2472555222068538. [PubMed: 28467117]

9.

Imbert PE, Unterreiner V, Siebert D, Gubler H, Parker C, Gabriel D. Recommendations for the Reduction of Compound Artifacts in Time-Resolved Fluorescence Resonance Energy Transfer Assays. ASSAY Drug Dev Technol (Internet). 2007 Jun (cited 2023 Sep 18);5(3):363–72. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2007.073. [PubMed: 17638536]

10.

Simeonov A, Jadhav A, Thomas CJ, Wang Y, Huang R, Southall NT, et al. Fluorescence Spectroscopic Profiling of Compound Libraries. J Med Chem (Internet). 2008 Apr 1 (cited 2023 Sep 18);51(8):2363–71. Available from: https://pubs​.acs.org/doi/10​.1021/jm701301m. [PubMed: 18363325]

11.

Bray MA, Carpenter AE. Quality Control for High-Throughput Imaging Experiments Using Machine Learning in Cellprofiler. In: Johnston PA, Trask OJ, editors. High Content Screening (Internet). New York, NY: Springer New York; 2018 (cited 2023 Sep 18). p. 89–112. (Methods in Molecular Biology; vol. 1683). Available from: http://link​.springer​.com/10.1007/978-1-4939-7357-6_7. [PMC free article: PMC6112602] [PubMed: 29082489]

12.

Dudgeon DD, Shinde SN, Shun TY, Lazo JS, Strock CJ, Giuliano KA, et al. Characterization and Optimization of a Novel Protein–Protein Interaction Biosensor High-Content Screening Assay to Identify Disruptors of the Interactions Between p53 and hDM2. ASSAY Drug Dev Technol (Internet). 2010 Aug (cited 2023 Sep 18);8(4):437–58. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2010.0281. [PMC free article: PMC2929144] [PubMed: 20662736]

13.

Fancher AT, Hua Y, Camarco DP, Close DA, Strock CJ, Johnston PA. High-Content Screening Campaign to Identify Compounds That Inhibit or Disrupt Androgen Receptor-Transcriptional Intermediary Factor 2 Protein-Protein Interactions for the Treatment of Prostate Cancer. ASSAY Drug Dev Technol (Internet). 2018 Aug (cited 2023 Sep 18);16(6):297–319. Available from: https://www​.liebertpub​.com/doi/10.1089/adt.2018.858. [PMC free article: PMC6114076] [PubMed: 30109944]

14.

Hua Y, Camarco DP, Strock CJ, Johnston PA. High Content Positional Biosensor Assay to Screen for Compounds that Prevent or Disrupt Androgen Receptor and Transcription Intermediary Factor 2 Protein-Protein Interactions. In: Johnston PA, Trask OJ, editors. High Content Screening (Internet). New York, NY: Springer New York; 2018 (cited 2023 Sep 18). p. 211–27. (Methods in Molecular Biology; vol. 1683). Available from: http://link​.springer​.com/10.1007/978-1-4939-7357-6_13. [PMC free article: PMC6753931] [PubMed: 29082495]

15.

Malo N, Hanley JA, Cerquozzi S, Pelletier J, Nadon R. Statistical practice in high-throughput screening data analysis. Nat Biotechnol (Internet). 2006 Feb (cited 2023 Sep 18);24(2):167–75. Available from: https://www​.nature.com/articles/nbt1186. [PubMed: 16465162]

16.

Shun TY, Lazo JS, Sharlow ER, Johnston PA. Identifying Actives from HTS Data Sets: Practical Approaches for the Selection of an Appropriate HTS Data-Processing Method and Quality Control Review. SLAS Discov (Internet). 2011 Jan (cited 2023 Sep 18);16(1):1–14. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S2472555222077371. [PubMed: 21160066]

17.

Dahlin JL, Hua BK, Zucconi BE, Nelson SD, Singh S, Carpenter AE, et al. Reference compounds for characterizing cellular injury in high-content cellular morphology assays. Nat Commun (Internet). 2023 Mar 13 (cited 2025 Mar 24);14(1):1364. Available from: https://www​.nature.com​/articles/s41467-023-36829-x. [PMC free article: PMC10011410] [PubMed: 36914634]

18.

Dudgeon DD, Shinde S, Hua Y, Shun TY, Lazo JS, Strock CJ, et al. Implementation of a 220,000-Compound HCS Campaign to Identify Disruptors of the Interaction between p53 and hDM2 and Characterization of the Confirmed Hits. SLAS Discov (Internet). 2010 Aug (cited 2023 Sep 18);15(7):766–82. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S247255522207962X. [PubMed: 20639499]

19.

Fancher AT, Hua Y, Camarco DP, Close DA, Strock CJ, Johnston PA. Reconfiguring the AR-TIF2 Protein–Protein Interaction HCS Assay in Prostate Cancer Cells and Characterizing the Hits from a LOPAC Screen. ASSAY Drug Dev Technol (Internet). 2016 Oct (cited 2023 Sep 18);14(8):453–77. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2016.741. [PMC free article: PMC5067810] [PubMed: 27606620]

20.

Buchser W, Collins M, Garyantes T, et al. Assay Development Guidelines for Image-Based High Content Screening, High Content Analysis and High Content Imaging. 2012 Oct 1 [Updated 2014 Sep 22]. In: Markossian S, Grossman A, Baskir H, et al., editors. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004-. Available from: https://www​.ncbi.nlm​.nih.gov/books/NBK100913/ [PubMed: 23469374]

21.

Mamontova AV, Bogdanov AM, Lukyanov KA. Influence of cell growth conditions and medium composition on EGFP photostability in live cells. BioTechniques (Internet). 2015 May (cited 2023 Sep 18);58(5):258–61. Available from: https://www​.future-science​.com/doi/10.2144/000114289. [PubMed: 25967905]

22.

Abbott KL, Ali A, Casalena D, Do BT, Ferreira R, Cheah JH, et al. Screening in serum-derived medium reveals differential response to compounds targeting metabolism. Cell Chem Biol (Internet). 2023 Sep 21 (cited 2025 Mar 24);30(9):1156-1168.e7. Available from: https://doi.org/. 10.1016/j.chembiol.2023.08.007 . [PMC free article: PMC10581593] [PubMed: 37689063] [CrossRef]

23.

Campbell JM, Mahbub SB, Habibalahi A, Agha A, Handley S, Anwer AG, et al. Clinical applications of non‐invasive multi and hyperspectral imaging of cell and tissue autofluorescence beyond oncology. J Biophotonics (Internet). 2023 Apr (cited 2023 Sep 18);16(4):e202200264. Available from: https:​//onlinelibrary​.wiley.com/doi/10.1002/jbio.202200264. [PubMed: 36602432]

24.

Kolenc OI, Quinn KP. Evaluating Cell Metabolism Through Autofluorescence Imaging of NAD(P)H and FAD. Antioxid Redox Signal (Internet). 2019 Feb 20 (cited 2023 Sep 18);30(6):875–89. Available from: https://www​.liebertpub​.com/doi/10.1089/ars.2017.7451. [PMC free article: PMC6352511] [PubMed: 29268621]

25.

Burkhardt M, Vollmar B, Menger MD. In Vivo Analysis of Hepatic NADH Fluorescence. In: Hudetz AG, Bruley DF, editors. Oxygen Transport to Tissue XX (Internet). Boston, MA: Springer US; 1998 (cited 2025 Mar 20). p. 83–9. (Advances in Experimental Medicine and Biology; vol. 454). Available from: http://link​.springer​.com/10.1007/978-1-4615-4863-8_10. [PubMed: 9889879]

26.

Croce AC, Ferrigno A, Bottiroli G, Vairetti M. Autofluorescence-based optical biopsy: An effective diagnostic tool in hepatology. Liver Int (Internet). 2018 Jul (cited 2023 Sep 18);38(7):1160–74. Available from: https:​//onlinelibrary​.wiley.com/doi/10.1111/liv.13753. [PubMed: 29624848]

27.

Yuanlong Y, Yanming Y, Fuming L, Yufen L, Paozhong M. Characteristic autofluorescence for cancer diagnosis and its origin. Lasers Surg Med (Internet). 1987 (cited 2023 Sep 18);7(6):528–32. Available from: https:​//onlinelibrary​.wiley.com/doi/10.1002/lsm.1900070617. [PubMed: 3431331]

28.

Hung J, Lam S, Leriche JC, Palcic B. Autofluorescence of normal and malignant bronchial tissue. Lasers Surg Med (Internet). 1991 (cited 2023 Sep 18);11(2):99–105. Available from: https:​//onlinelibrary​.wiley.com/doi/10.1002/lsm.1900110203. [PubMed: 2034016]

29.

Umeno A, Biju V, Yoshida Y. In vivo ROS production and use of oxidative stress-derived biomarkers to detect the onset of diseases such as Alzheimer’s disease, Parkinson’s disease, and diabetes. Free Radic Res (Internet). 2017 Apr 3 (cited 2023 Sep 18);51(4):413–27. Available from: https://www​.tandfonline​.com/doi/full/10.1080/10715762​.2017.1315114. [PubMed: 28372523]

30.

Thorling CA, Crawford D, Burczynski FJ, Liu X, Liau I, Roberts MS. Multiphoton microscopy in defining liver function. J Biomed Opt (Internet). 2014 Sep 8 (cited 2023 Sep 18);19(9):090901. Available from: http:​//biomedicaloptics​.spiedigitallibrary​.org/article.aspx?doi=10​.1117/1.JBO.19.9.090901. [PubMed: 25200392]

31.

Walsh AJ, Cook RS, Sanders ME, Aurisicchio L, Ciliberto G, Arteaga CL, et al. Quantitative Optical Imaging of Primary Tumor Organoid Metabolism Predicts Drug Response in Breast Cancer. Cancer Res (Internet). 2014 Sep 15 (cited 2023 Sep 18);74(18):5184–94. Available from: https:​//aacrjournals​.org/cancerres/article​/74/18/5184/596318​/Quantitative-Optical-Imaging-of-Primary-Tumor. [PMC free article: PMC4167558] [PubMed: 25100563]

32.

Schaue D, Ratikan JA, Iwamoto KS. Cellular Autofluorescence following Ionizing Radiation. Launikonis BS, editor. PLoS ONE (Internet). 2012 Feb 22 (cited 2023 Sep 18);7(2):e32062. Available from: https://dx​.plos.org/10​.1371/journal.pone.0032062. [PMC free article: PMC3284545] [PubMed: 22384140]

33.

Surre J, Saint-Ruf C, Collin V, Orenga S, Ramjeet M, Matic I. Strong increase in the autofluorescence of cells signals struggle for survival. Sci Rep (Internet). 2018 Aug 14 (cited 2023 Sep 18);8(1):12088. Available from: https://www​.nature.com​/articles/s41598-018-30623-2. [PMC free article: PMC6092379] [PubMed: 30108248]

34.

Auld DS Ph.D., Coassin PA B.S., Coussens NP Ph.D., et al. Microplate Selection and Recommended Practices in High-throughput Screening and Quantitative Biology. 2020 Jun 1. In: Markossian S, Grossman A, Baskir H, et al., editors. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004-. Available from: https://www​.ncbi.nlm​.nih.gov/books/NBK558077/

35.

Reinhart PH, Kaltenbach LS, Essrich C, Dunn DE, Eudailey JA, DeMarco CT, et al. Identification of anti-inflammatory targets for Huntington’s disease using a brain slice-based screening assay. Neurobiol Dis. 2011;43(1):248–56. [PMC free article: PMC3104027] [PubMed: 21458569]

36.

Corning FAQ. Corning® Matrigel® Matrix Frequently Asked Questions (Internet). Available from: https://www​.corning.com​/catalog/cls/documents​/faqs/CLS-DL-CC-026.pdf.

37.

Riss T, Trask OJ. Factors to consider when interrogating 3D culture models with plate readers or automated microscopes. Vitro Cell Dev Biol-Anim. 2021;57:238–56. [PMC free article: PMC7946695] [PubMed: 33564998]

38.

Langhans SA. Three-Dimensional in Vitro Cell Culture Models in Drug Discovery and Drug Repositioning. Front Pharmacol (Internet). 2018 Jan 23 (cited 2023 Sep 18);9:6. Available from: http://journal​.frontiersin​.org/article/10​.3389/fphar.2018.00006/full. [PMC free article: PMC5787088] [PubMed: 29410625]

39.

Busch M, Thoma HB, Kober I. Does Your Lab Coat Fit to Your Assay? SLAS Discov (Internet). 2013 Jul (cited 2023 Sep 18);18(6):744–7. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S2472555222075244. [PubMed: 23507575]

40.

Almeida JL, Cole KD, Plant AL. Standards for Cell Line Authentication and Beyond. PLOS Biol (Internet). 2016 Jun 14 (cited 2023 Sep 18);14(6):e1002476. Available from: https://dx​.plos.org/10​.1371/journal.pbio.1002476. [PMC free article: PMC4907466] [PubMed: 27300367]

41.

Reid Y, Storts D, Riss T, Minor L. Authentication of human cell lines by STR DNA profiling analysis. 2013;

42.

Almeida JL, Korch CT. Authentication of Human and Mouse Cell Lines by Short Tandem Repeat (STR) DNA Genotype Analysis. 2023 Jan 17. In: Markossian S, Grossman A, Baskir H, et al., editors. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004-. Available from: https://www​.ncbi.nlm​.nih.gov/books/NBK144066/ [PubMed: 23805434]

43.

Vedvik KL, Eliason HC, Hoffman RL, Gibson JR, Kupcho KR, Somberg RL, et al. Overcoming Compound Interference in Fluorescence Polarization-Based Kinase Assays Using Far-Red Tracers. ASSAY Drug Dev Technol (Internet). 2004 Apr (cited 2023 Sep 18);2(2):193–203. Available from: http://www​.liebertpub​.com/doi/10.1089/154065804323056530. [PubMed: 15165515]

44.

Sexton JZ, Fursmidt R, O’Meara MJ, et al. Machine Learning and Assay Development for Image-based Phenotypic Profiling of Drug Treatments. 2023 Mar 15. In: Markossian S, Grossman A, Baskir H, et al., editors. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004-. Available from: https://www​.ncbi.nlm​.nih.gov/books/NBK589577/ [PubMed: 23469374]

45.

Inglese J, Auld DS, Jadhav A, Johnson RL, Simeonov A, Yasgar A, et al. Quantitative high-throughput screening: a titration-based approach that efficiently identifies biological activities in large chemical libraries. Proc Natl Acad Sci. 2006;103(31):11473–8. [PMC free article: PMC1518803] [PubMed: 16864780]

46.

Joossens E, Macko P, Palosaari T, Gerloff K, Ojea-Jiménez I, Gilliland D, et al. A high throughput imaging database of toxicological effects of nanomaterials tested on HepaRG cells. Sci Data. 2019;6(1):46. [PMC free article: PMC6497662] [PubMed: 31048742]

47.

Schmitt DL, Dranchak P, Parajuli P, Blivis D, Voss T, Kohnhorst CL, et al. High-throughput screening identifies cell cycle-associated signaling cascades that regulate a multienzyme glucosome assembly in human cells. Plos One. 2023;18(8):e0289707. [PMC free article: PMC10403072] [PubMed: 37540718]

48.

Wang M, Da Y, Tian Y. Fluorescent proteins and genetically encoded biosensors. Chem Soc Rev (Internet). 2023 (cited 2023 Sep 18);52(4):1189–214. Available from: http://xlink​.rsc.org/?DOI=D2CS00419D. [PubMed: 36722390]

49.

Bray MA, Singh S, Han H, Davis CT, Borgeson B, Hartland C, et al. Cell Painting, a high-content image-based assay for morphological profiling using multiplexed fluorescent dyes. Nat Protoc (Internet). 2016 Sep (cited 2023 Sep 18);11(9):1757–74. Available from: https://www​.nature.com/articles/nprot​.2016.105. [PMC free article: PMC5223290] [PubMed: 27560178]

50.

Lanteri CA, Trumpower BL, Tidwell RR, Meshnick SR. DB75, a novel trypanocidal agent, disrupts mitochondrial function in Saccharomyces cerevisiae. Antimicrob Agents Chemother. 2004 Oct;48(10):3968–74. [PMC free article: PMC521894] [PubMed: 15388460]

51.

Lanteri CA, Tidwell RR, Meshnick SR. The mitochondrion is a site of trypanocidal action of the aromatic diamidine DB75 in bloodstream forms of Trypanosoma brucei. Antimicrob Agents Chemother. 2008 Mar;52(3):875–82. [PMC free article: PMC2258549] [PubMed: 18086841]

52.

Simeonov A, Davis MI. Interference with Fluorescence and Absorbance. In: Markossian S, Grossman A, Brimacombe K, Arkin M, Auld D, Austin C, et al., editors. Assay Guidance Manual (Internet). Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004 (cited 2023 Sep 20). Available from: http://www​.ncbi.nlm.nih​.gov/books/NBK343429/

53.

Coussens NP, Auld D, Roby P, Walsh J, Baell JB, Kales S, et al. Compound-Mediated Assay Interferences in Homogeneous Proximity Assays. In: Markossian S, Grossman A, Brimacombe K, Arkin M, Auld D, Austin C, et al., editors. Assay Guidance Manual (Internet). Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004 (cited 2023 Sep 20). Available from: http://www​.ncbi.nlm.nih​.gov/books/NBK553584/

54.

Clemons PA, Bittker JA, Wagner FF, Hands A, Dančík V, Schreiber SL, et al. The Use of Informer Sets in Screening: Perspectives on an Efficient Strategy to Identify New Probes. SLAS Discov (Internet). 2021 Aug (cited 2023 Sep 18);26(7):855–61. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S247255522206734X. [PMC free article: PMC8991386] [PubMed: 34130532]

55.

Hansson P, Boyd H, Dale IL, Dahl G, Nicolaus F, Bowen W, et al. A Comparative Study of Fluorescence Assays in Screening for BRD4. ASSAY Drug Dev Technol (Internet). 2018 Oct (cited 2023 Sep 18);16(7):372–83. Available from: https://www​.liebertpub​.com/doi/10.1089/adt.2018.850. [PubMed: 30307314]

56.

Borrel A, Mansouri K, Nolte S, Saddler T, Conway M, Schmitt C, et al. InterPred: a webtool to predict chemical autofluorescence and luminescence interference. Nucleic Acids Res (Internet). 2020 Jul 2 (cited 2023 Sep 18);48(W1):W586–90. Available from: https://academic​.oup​.com/nar/article/48/W1/W586/5838861. [PMC free article: PMC7319558] [PubMed: 32421835]

57.

Hua Y, Shun TY, Strock CJ, Johnston PA. High-Content Positional Biosensor Screening Assay for Compounds to Prevent or Disrupt Androgen Receptor and Transcriptional Intermediary Factor 2 Protein–Protein Interactions. ASSAY Drug Dev Technol (Internet). 2014 Sep (cited 2023 Sep 18);12(7):395–418. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2014.594. [PMC free article: PMC4146499] [PubMed: 25181412]

58.

Hua Y, Strock CJ, Johnston PA. High Content Screening Biosensor Assay to Identify Disruptors of p53–hDM2 Protein-Protein Interactions. In: Meyerkord CL, Fu H, editors. Protein-Protein Interactions (Internet). New York, NY: Springer New York; 2015 (cited 2023 Sep 18). p. 555–65. (Methods in Molecular Biology; vol. 1278). Available from: https://link​.springer​.com/10.1007/978-1-4939-2425-7_37. [PubMed: 25859976]

59.

Johnston PA, Sen M, Hua Y, Camarco D, Shun TY, Lazo JS, et al. High-Content pSTAT3/1 Imaging Assays to Screen for Selective Inhibitors of STAT3 Pathway Activation in Head and Neck Cancer Cell Lines. ASSAY Drug Dev Technol (Internet). 2014 Jan (cited 2023 Sep 18);12(1):55–79. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2013.524. [PMC free article: PMC3934522] [PubMed: 24127660]

60.

Johnston PA, Sen M, Hua Y, Camarco DP, Shun TY, Lazo JS, et al. HCS Campaign to Identify Selective Inhibitors of IL-6-Induced STAT3 Pathway Activation in Head and Neck Cancer Cell Lines. ASSAY Drug Dev Technol (Internet). 2015 Sep (cited 2023 Sep 18);13(7):356–76. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2015.663. [PMC free article: PMC4556090] [PubMed: 26317883]

61.

Johnston PA, Nguyen MM, Dar JA, Ai J, Wang Y, Masoodi KZ, et al. Development and Implementation of a High-Throughput High-Content Screening Assay to Identify Inhibitors of Androgen Receptor Nuclear Localization in Castration-Resistant Prostate Cancer Cells. ASSAY Drug Dev Technol (Internet). 2016 May (cited 2023 Sep 18);14(4):226–39. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2016.716. [PMC free article: PMC4876501] [PubMed: 27187604]

62.

Nickischer D, Laethem C, Trask OJ, Williams RG, Kandasamy R, Johnston PA, et al. Development and Implementation of Three Mitogen‐Activated Protein Kinase (MAPK) Signaling Pathway Imaging Assays to Provide MAPK Module Selectivity Profiling for Kinase Inhibitors: MK2‐EGFP Translocation, c‐Jun, and ERK Activation. In: Methods in Enzymology (Internet). Elsevier; 2006 (cited 2023 Sep 18). p. 389–418. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S0076687906140227. [PubMed: 17110204]

63.

Trask OJ, Baker A, Williams RG, Nickischer D, Kandasamy R, Laethem C, et al. Assay Development and Case History of a 32K‐Biased Library High‐Content MK2‐EGFP Translocation Screen to Identify p38 Mitogen‐Activated Protein Kinase Inhibitors on the ArrayScan 3.1 Imaging Platform. In: Methods in Enzymology (Internet). Elsevier; 2006 (cited 2023 Sep 18). p. 419–39. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S0076687906140239. [PubMed: 17110205]

64.

Trask OJ, Nickischer D, Burton A, Williams RG, Kandasamy RA, Johnston PA, et al. High-Throughput Automated Confocal Microscopy Imaging Screen of a Kinase-Focused Library to Identify p38 Mitogen-Activated Protein Kinase Inhibitors Using the GE InCell 3000 Analyzer. In: Janzen WP, Bernasconi P, editors. High Throughput Screening (Internet). Totowa, NJ: Humana Press; 2009 (cited 2023 Sep 18). p. 159–86. (Methods in Molecular Biology; vol. 565). Available from: http://link​.springer​.com/10.1007/978-1-60327-258-2_8. [PubMed: 19551362]

65.

Williams RG, Kandasamy R, Nickischer D, Trask OJ, Laethem C, Johnston PA, et al. Generation and Characterization of a Stable MK2‐EGFP Cell Line and Subsequent Development of a High‐Content Imaging Assay on the Cellomics ArrayScan Platform to Screen for p38 Mitogen‐Activated Protein Kinase Inhibitors. In: Methods in Enzymology (Internet). Elsevier; 2006 (cited 2023 Sep 18). p. 364–89. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S0076687906140215. [PubMed: 17110203]

66.

Johnston PA, Shinde SN, Hua Y, Shun TY, Lazo JS, Day BW. Development and Validation of a High-Content Screening Assay to Identify Inhibitors of Cytoplasmic Dynein-Mediated Transport of Glucocorticoid Receptor to the Nucleus. ASSAY Drug Dev Technol (Internet). 2012 Oct (cited 2023 Sep 18);10(5):432–56. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2012.456. [PMC free article: PMC3464420] [PubMed: 22830992]

67.

Johnston PA, Sen M, Hua Y, Camarco DP, Shun TY, Lazo JS, et al. High Content Imaging Assays for IL-6-Induced STAT3 Pathway Activation in Head and Neck Cancer Cell Lines. In: Johnston PA, Trask OJ, editors. High Content Screening (Internet). New York, NY: Springer New York; 2018 (cited 2023 Sep 18). p. 229–44. (Methods in Molecular Biology; vol. 1683). Available from: http://link​.springer​.com/10.1007/978-1-4939-7357-6_14. [PMC free article: PMC6507410] [PubMed: 29082496]

68.

Judson R, Houck K, Paul Friedman K, Brown J, Browne P, Johnston PA, et al. Selecting a minimal set of androgen receptor assays for screening chemicals. Regul Toxicol Pharmacol (Internet). 2020 Nov (cited 2023 Sep 18);117:104764. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S0273230020301902. [PMC free article: PMC8356084] [PubMed: 32798611]

69.

Grantab R, Sivananthan S, Tannock IF. The Penetration of Anticancer Drugs through Tumor Tissue as a Function of Cellular Adhesion and Packing Density of Tumor Cells. Cancer Res (Internet). 2006 Jan 15 (cited 2023 Sep 18);66(2):1033–9. Available from: https:​//aacrjournals​.org/cancerres/article​/66/2/1033/526484/The-Penetration-of-Anticancer-Drugs-through-Tumor. [PubMed: 16424039]

70.

Grantab RH, Tannock IF. Penetration of anticancer drugs through tumour tissue as a function of cellular packing density and interstitial fluid pressure and its modification by bortezomib. BMC Cancer (Internet). 2012 Dec (cited 2023 Sep 18);12(1):214. Available from: http://bmccancer​.biomedcentral​.com/articles/10​.1186/1471-2407-12-214. [PMC free article: PMC3407510] [PubMed: 22672469]

71.

Kerr DJ, Kaye SB. Aspects of cytotoxic drug penetration, with particular reference to anthracyclines. Cancer Chemother Pharmacol (Internet). 1987 Feb (cited 2023 Sep 18);19(1):1–5. Available from: http://link​.springer​.com/10.1007/BF00296245. [PubMed: 3545523]

72.

Minchinton AI, Tannock IF. Drug penetration in solid tumours. Nat Rev Cancer (Internet). 2006 Aug (cited 2023 Sep 18);6(8):583–92. Available from: https://www​.nature.com/articles/nrc1893. [PubMed: 16862189]

73.

Shan F, Close DA, Camarco DP, Johnston PA. High-Content Screening Comparison of Cancer Drug Accumulation and Distribution in Two-Dimensional and Three-Dimensional Culture Models of Head and Neck Cancer. ASSAY Drug Dev Technol (Internet). 2018 Jan (cited 2023 Sep 18);16(1):27–50. Available from: http://www​.liebertpub​.com/doi/10.1089/adt.2017.812. [PMC free article: PMC5775115] [PubMed: 29215913]

74.

Tannock IF, Lee CM, Tunggal JK, Cowan DS, Egorin MJ. Limited penetration of anticancer drugs through tumor tissue: a potential cause of resistance of solid tumors to chemotherapy. Clin Cancer Res. 2002;8(3):878–84. [PubMed: 11895922]

75.

Monici M. Cell and tissue autofluorescence research and diagnostic applications. In: Biotechnology Annual Review (Internet). Elsevier; 2005 (cited 2023 Sep 18). p. 227–56. Available from: https://linkinghub​.elsevier​.com/retrieve​/pii/S1387265605110072. [PubMed: 16216779]

76.

Wagnieres GA, Star WM, Wilson BC. In Vivo Fluorescence Spectroscopy and Imaging for Oncological Applications. Photochem Photobiol (Internet). 1998 Nov (cited 2023 Sep 18);68(5):603–32. Available from: https:​//onlinelibrary​.wiley.com/doi/10.1111/j​.1751-1097.1998.tb02521.x. [PubMed: 9825692]

Appendix Figures and Table

Appendix Figure A1.

Compound autofluorescence as indicated by the nuclei Z-scores in the Hoechst (DAPI channel, ~450 nm).

A. Max + DHT (green dots); Compounds (red dots); Min - DMSO (black dots). B: Hoechst-stained nuclei average intensity with cell counts Z-scores > -3 but < 3; 12-bit images. B1 = Compound; B2= Compound: B3 = Max + DHT control; B4 = Min – DMSO control; B5 & B6 = apoptosis inducer compounds. C: Hoechst-stained nuclei Quenchers with average intensity Z-scores < -3. C1 and C2 represent 12-bit image display range of 96-767; C3 & C4 represent a rescaled contrast stretch 8-bit image display range of 97-136 of the same fields of view for visualization.

Appendix Figure A2.

Compound autofluorescence as indicated in the TIF2-GFP positive nucleoli Z-scores (FITC channel, ~525 nm).

A. Max – DHT (green dots); Compounds (red dots); Min – DMSO Z-scores (black dots)

B: TIF2-GFP biosensor positive nucleoli average intensity Z-scores. B1 = Max + DHT control; B2 = Min – DMSO control; B3, B4, B5, B6 = compounds.

Appendix Figure A3.

Images of TIF2-GFP positive nucleoli (pseudocolored green) from concentration response showing compound autofluorescence as compared to DMSO and DHT controls with increasing Z-scores expressed as the average (Ave) or integrated (Int) intensities in the FITC channel.

Appendix Figure A4.

Compound autofluorescence as indicated by the average intensity Z-scores of the AR-RFP biosensor in cells (Texas Red channel, ~600 nm).

A: Z-scores: Max + DHT (green dots); Compounds (red dots); Min – DMSO (black dots)

B: Fluorescent images (10x) of AR-RFP (Texas Red) biosensor average intensity Z-scores. B1 = Max + DHT control (nucleoli); B2 = Min – DMSO control (cytoplasm); B3 and B4 = increased autofluorescence inside cells from compounds; B5 and B6 compound autofluorescence in solution showing saturation (white) that appears blank.

Appendix Figure A5.

Images of AR-RFP nucleoli (pseudocolored red) from concentration response showing compound autofluorescence as compared to DMSO (cytoplasm) & DHT (nucleoli) controls with increasing Z-scores expressed as average (Ave) or integrated (Int) intensity values in the Texas Red channel.

Appendix Figure A6.

Compound-mediated cytotoxicity and/or cell loss and higher-than-average cell counts.

A. Max + DHT (green dots); Compounds (red dots); Min – DMSO Z-scores (black dots)

B: Fluorescent images (10x) of total cell counts from Hoechst-stained nuclei (DAPI channel) Z-scores.

B1 = Max + DHT control; B2 = Min – DMSO control; B3 and B4 compounds with higher-than-average cell counts (Z-scores >3), B5 and B6 compounds with lower-than-average cell counts (Z-scores <-3).