High-Content Analysis Applied To Autophagy

Automating Cell Imaging with Thermo Scientific HCA systems

(See a list of the products featured in this article.)

High-content analysis (HCA)—also known as high-content screening (HCS), image cytometry, quantitative cell analysis, or automated cell analysis—is an automated method for identifying substances that alter the phenotype of a cell in a desired manner. Primarily used in biological research and drug discovery, this technology combines fluorescence microscopy and automated cell calculations with phenotype analysis using image processing algorithms and informatics tools that enable the user to make decisions about the data. In this article, we describe the use of Thermo Scientific™ HCA systems (for example, see a description of the new Cellinsight™ CX7 HCA Platform) and Molecular Probes™ reagents to examine the process of autophagy and the effects of various compounds on that process.

Autophagy defined

Long-lived proteins, damaged cellular organelles, and invading microorganisms are cleared from healthy cells via a bulk-catabolic process known as autophagy. It was originally described by Christian De Duve in the 1960s [1] as a response to nutrient deprivation, but the last decade has provided more insight into the process, including characterization of receptors that mediate specific forms of autophagy, the autophagosome that engulfs cargo destined for clearance, and the mechanisms that allow fusion of the autophagosome with the lysosome [2]. Multiple scenarios have been described whereby aberrant mechanisms lead to a variety of disease states, and autophagy is now recognized as an area for therapeutic modulation [3].

Interrogating autophagy as a therapeutic target requires high-throughput analytical techniques. Current cell-based methods include identification of the characteristic double-membrane autophagosome through electron microscopy (EM) or localization of microtubule-associated proteins 1A/1B light chain 3B (LC3B) to the autophagosomal membrane using either fluorescence microscopy or western blotting. Of these three analysis platforms, fluorescence microscopy has proven to be the most amenable to high-throughput automation. Automated imaging and analysis platforms—integrated as high-content analysis [4]—have been used effectively to elucidate the hallmarks and regulators of autophagy [5,6].

An important feature of the assays presented here is the use of appropriate controls to determine whether or not the phenotypes being measured are a product of autophagy (with siRNA knockdown controls), as well as to ascertain if these phenotypes are a result of stimulating autophagy or inhibiting autophagic progress or flux (with chloroquine controls). Such controls are essential to validate assay protocols and to characterize candidate autophagy-moderating compounds under study.

Image segmentation and analysis

High-content analysis of autophagy requires identification of cells (via nuclear segmentation) and subsequent identification of LC3B puncta, which is efficiently accomplished using the Autophagy (or similar Compartmental Analysis) BioApplication software available on Thermo Scientific™ Cellinsight™ and ArrayScan™ HCS platforms. Figure 1 shows immortalized baby mouse kidney (IBMK) epithelial cells stably expressing GFP-LC3B and labeled with the blue-fluorescent nuclear stain Hoechst™ 33342. Once nuclei were identified, LC3B puncta were segmented and quantified using the Cellinsight CX5 HCS Platform.

A number of measurements can be taken to quantify autophagosome formation. Figure 2A shows the dose-dependent accumulation of LC3B puncta, as detected with anti-LC3B primary antibody and Alexa Fluor™ 647 secondary antibody, following inhibition of autophagic flux with bafilomycin A1. The relationship between LC3B puncta intensity and the number of LC3B puncta per cell is identical across bafilomycin A1 concentrations. The dose-response curves for five autophagy modulators were rapidly analyzed on the Cellinsight CX5 HCS Platform using mean LC3B spot intensity as a measure of autophagy (Figure 2B).

Figure 1. High-content image segmentation and analysis strategies.(A) Autophagy BioApplication analysis software on the Thermo Scientific™ Cellinsight™ CX5 HCS Platform. IBMK cells stably expressing GFP-LC3B were labeled with 1 μg/mL Hoechst™ 33342 (B); subsequent identification of nuclei by the software allowed segmentation (C). Once cells were identified, the autophagosomes expressing GFP-LC3B (D) were identified using the spot detector, and then segmented for analysis (E). Images were acquired using a 20x objective on the Cellinsight CX5 HCS Platform.

Figure 2. Dose-dependent LC3B labeling.(A) Dose-dependent correlation between the intensity of anti- LC3B staining and the number of autophagosomes detected via LC3B puncta. (B) Dose-response curves for five modulators of autophagy using mean LC3B spot intensity as an indicator of autophagy. Images were acquired using a 20x objective on the Thermo Scientific™ Cellinsight™ CX5 HCS Platform, and anti- LC3B staining was quantified using the Autophagy BioApplication analysis software.

Mechanism of action and specificity

When examining the nature of autophagosome formation, it is critical to confirm that the signal being measured is specific to autophagy. Knockdown of key autophagy genes, such as ATG5 or ATG7, is the best way to establish a given autophagy reporter is specific. Chloroquine is a known inhibitor of autophagic flux through the pathway, and chloroquine treatment results in an increase of LC3B puncta in otherwise normal cells. In contrast, siRNA-mediated knockdown of ATG7 expression results in a loss of LC3B puncta after chloroquine treatment (Figure 3).

An increase in LC3B puncta in the cytoplasm can arise either from agents that stimulate autophagy or from those that block autophagic flux. To identify which of these possibilities is occurring, a known autophagic flux inhibitor, such as chloroquine, may be used. Figure 4 shows this approach in practice. Two compounds, A and B, are applied in the presence or absence of chloroquine. Both compound A and B cause an increase in LC3B labeling; however, only compound B shows additional LC3B staining in the presence of chloroquine. Therefore, compound B is stimulating autophagy, whereas compound A is acting on the autophagy pathway at a point similar to that of chloroquine (i.e., inhibiting autophagic flux).

Figure 3. siRNA-mediated knockdown of key autophagy genes to confirm specificity of the label. LC3B puncta significantly increased upon treatment of A549 cells with chloroquine (CQ), as compared with treatment with vehicle alone (vh, in this case water). This increase is much reduced when cells are transfected with siRNA against ATG7, but not when cells are transfected with the scrambled version of the siRNA (scramble). Anti-LC3B staining was quantified using the Compartmental Analysis BioApplication software on a Thermo Scientific™ ArrayScan™ VTI HCS Reader.

Figure 4. LC3B puncta resulting from induction of autophagy or inhibition of autophagic flux. Treatment of cells with compound A or B significantly increases anti-LC3B staining. Combining chloroquine with compound A or B causes a further increase only with compound B. Anti-LC3B staining was quantified using the Compartmental Analysis BioApplication software on a Thermo Scientific™ ArrayScan™ VTI HCS Reader.

Measuring autophagy with transient expression of GFP-based biosensors

An alternative to antibody-based monitoring of autophagy is the use of cells transformed with a biosensor that is a fusion of a fluorescent protein (FP) and an autophagy marker. Stable cell lines, once generated, are ready to assay; however, the process of generating stably transformed cell lines is expensive and time consuming, and assaying different cell types requires the creation of new cell lines. Fortunately, the delivery of FP-based biosensors can be performed quickly and easily using the BacMam gene delivery and expression system, and the transiently transduced cells can be used in autophagy assays the next day. Figure 5 shows U2OS cells transduced with either Green Fluorescent Protein (GFP)-LC3B or GFP-p62 using BacMam 2.0 technology. High-content imaging demonstrates the dose-dependent accumulation of both markers when autophagic flux is inhibited using chloroquine.

Figure 5. Measuring autophagy with transient expression of GFP-based biosensors. U2OS cells were transduced using either the Premo™ Autophagy Sensor LC3B-GFP or the Premo™ Autophagy Sensor GFP-p62. High-content imaging of autophagy is shown in cells expressing fluorescent protein–based biosensors for either (A) LC3B or (B) the autophagy receptor p62/SQSTM1. Images were quantified using the Compartmental Analysis BioApplication software on a Thermo Scientific™ ArrayScan™ VTI HCS Reader.

References

1. Klionsky DJ (2008) Autophagy 4:740–743.
2. Feng Y, He D, Yao Z et al. (2014) Cell Res 24:24–41.
3. Jiang P, Mizushima N (2014) Cell Res 24:69–79.
4. Joachim J, Jiang M, McKnight NC et al. (2015) Methods 75:96–104.
5. McKnight NC, Jefferies HB, Alemu EA et al. (2012) EMBO J 31:1931–1946.
6. Strohecker AM, Joshi S, Possemato R et al. (2015) Oncogene doi:10.1038/onc.2015.23.