A simple method for quantitating confocal fluorescent images

Abstract

Western blotting (WB), enzyme-linked immunosorbent assay (ELISA) and flow cytometry (FC) have long been used to assess and quantitate relative protein expression in cultured cells and tissue samples. However, WB and ELISA have limited ability to meaningfully quantitate relative protein levels in tissues with complex cell composition, while tissue dissociation followed by FC is not feasible when tissue is limiting and/or cells difficult to isolate. While protein detection in tissue using immunofluorescent (IF) probes has traditionally been considered a qualitative technique, advances in probe stability and confocal imaging allow IF data to be easily quantitated, although reproducible quantitation of relative protein expression requires careful attention to appropriate controls, experiment design, and data collection. Here we describe the methods used to quantify the data presented in Shihan et al. Matrix Biology, 2020 which lays out a workflow where IF data collected on a confocal microscope can be used to quantitate the relative levels of a molecule of interest by measuring mean fluorescent intensity across a region of interest, cell number, and the percentage of cells in a sample "positive" for staining with the fluorescent probe of interest. Overall, this manuscript discusses considerations for collecting quantifiable fluorescent images on a confocal microscope and provides explicit methods for quantitating IF data using FIJI-ImageJ.

Keywords: Cell counting; Confocal microscopy; ImageJ; Immunofluorescence; Mean fluorescence intensity (MFI); Protein quantitation.

Graphical abstract

Fig. 1

Acquisition of a confocal image suitable for quantification Panel A is a representative confocal image of a section taken through the adult mouse lens epithelium stained for E cadherin (Alexa Fluor 488, green) and DNA (DRAQ5, blue) imaged with laser, master gain, and offset settings appropriate for quantitation as shown in the “range indicator” view of this image shown in Panel B which shows that areas lacking tissue have a very little signal (blue background), while few to no red pixels (signifying saturation of the detector) are seen. Panel C shows a negative control section through the lens epithelium that was created by omitting the primary antibody in the experiment, and Panel D shows the range indicator view of the negative control image Panel E is also a confocal image of a section taken through the adult mouse lens epithelium stained for E cadherin (Alexa Fluor 488, green) and DNA (DRAQ5, blue), however, the settings used have oversaturated the green channel as shown by the large number of “red” pixels seen in the range indicator view shown in Panel F. C- Lens Capsule, LC- Lens cells, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Quantitation of signal intensity in a region of interest using the Mean fluorescence intensity (MFI) method (Approach 1 (3.2.1)) A. Original confocal image of a section cut through an adult lens epithelium ((E cadherin (Alexa Fluor 488- green), αSMA (Cy3-Red) & DNA (DRAQ5-blue)). B. View showing the channel containing the E cadherin to be quantified. C. Selection of the region of interest by tracing the tissue using a high-resolution drawing tablet, and output from MFI measurement. D. Determination of the MFI of the background from a rectangular area of the image that lacks tissue and the output from that measurement. C- Lens Capsule, LC- Lens cells, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Mean fluorescence intensity (MFI) measurement of a non-nuclear protein using an automated region of interest selection based on a signal threshold (Approach 2 (3.2.2)) Panel A. Original confocal image of a section cut through an adult mouse lens epithelium (E cadherin (Alexa Fluor 488- green), αSMA (Cy3-Red) & DNA (DRAQ5-blue)). Panel B. View showing the channel containing the E cadherin to be quantified. Panel C. Automated selection of the region of interest using the “threshold” method. Panel D Clicking on the ‘Reset’ button to get back to the original image and the resulting output from the MFI measurement. C- Lens Capsule, LC- Lens cells, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Mean fluorescence intensity (MFI) measurement of cell nuclei using automated region of interest selection based on a signal threshold (Approach 2 (3.2.2)) A. Original confocal image of a section cut through a WT mouse lens capsule at 3 days post cataract surgery (three-color imaging; (pSmad3- Alexa Fluor 568- red), αSMA (FITC- green) & DNA (DRAQ5-blue)). B. pSmad3 channel alone only. C. Adjusting the red channel to select for cell nuclei stained for pSmad3 protein by using the ‘Threshold’ button. D. Nuclei exhibiting pSmad3 staining above ‘Threshold’ are outlined in yellow and are ready for MFI quantitation. E. Click on the ‘Reset’ button to get back to the original image and results of MFI quantitation of pSmad3 levels in selected cell nuclei. C- Lens Capsule, LC- Lens cells, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Automated cell counting based on quantification of the number of nuclei in an image A. Original confocal image of a section taken through a WT lens capsule at 3 days post cataract surgery (three-color imaging; (Ki 67- Alexa Fluor 568- red), αSMA (FITC- green) & DNA (DRAQ5-blue)). B. The 633 channel showing only the DNA staining. C. Identification of ROIs stained for DNA using the ‘Threshold’ feature. D. The results from the use of the ‘Fill holes’ feature to correct for the detection of only a portion of some nuclei. E. The use of the ‘Watershed’ feature to separate “overlapping” nuclear signals to allow for automated counting. F. Automated cell counting is based on the number of nuclei identified by the prior image processes steps. C- Lens Capsule, LC- Lens cells, ROI- Region of interest, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Automated counting of the number of nuclei in a section positive for a signal of interest A. Original confocal image of a section through a WT lens capsule 3 days post cataract surgery (three-color imaging; (Ki 67- Alexa Fluor 568- red), αSMA (FITC- green) & DNA (DRAQ5-blue)). B. Red channel that shows the Ki 67 staining alone. C. Identification of the Ki 67 positive nuclei using the ‘Threshold’ function. D. The use of the ‘Fill holes’ feature to ensure that the “positive” nuclei are detected as a single particle. E. Application of the ‘Watershed’ feature to distinguish areas where the signals for two or more nuclei overlap in the image. F. The ROIs exhibiting Ki 67 staining identified by the prior image processing steps are counted by the software. C- Lens Capsule, LC- Lens cells, ROI- Region of interest, Scale bar- 35 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)