Quantitative Colocalization Analysis in Fluorescence Microscopy

Software and datasets

1. Developed MATLAB-based script for colocalization analysis (https://github.com/NanLabMyxo/coAnalyzer), DOI: 10.5281/zenodo.21516437

2. MATLAB R2018b and later versions (The MathWorks, Inc. https://www.mathworks.com/help/install/ug/install-products-with-internet-connection.html) or MATLAB Online (https://www.mathworks.com/products/matlab-online.html)

3. Operating systems: Microsoft Windows 11 Enterprise version 10.0 (Build 26200, 26100), Microsoft Windows 10 Enterprise version 10.0 (Build 19045)

3. Image Processing Toolbox (https://www.mathworks.com/products/image-processing.html, version 26.1)

4. Optimization Toolbox (https://www.mathworks.com/products/optimization.html, version 26.1)

5. Statistics and Machine Learning Toolbox (https://www.mathworks.com/products/statistics.html, version 26.1)

Note: The MATLAB toolboxes indicated are free and available on the MATLAB Online platform. Users must use compatible MATLAB toolboxes for the MATLAB program (e.g., MATLAB toolbox version 26.1 is compatible with MATLAB R2026a).

Procedure

A. Load and label images

1. Start MATLAB and load the “co3.fig” file. The interface of the GUI of coAnalyzer will be displayed (Figure 1).

Note: Users should add the downloaded coAnalyzer folder to the MATLAB path and run the “co3.fig” file to launch the GUI of coAnalyzer.

Critical: The Image Processing Toolbox and the Optimization Toolbox must be installed before launching the software package.

Figure 1. GUI of coAnalyzer

2. In the Color 1 folder and Color 2 folder text boxes, type or copy and paste the folder paths for images from the first and second imaging channels.

Notes:

1. A list of input file formats supported and not supported by the software package is included in Table 1.

2. A list of output files, their locations, descriptions, and file formats after image analysis is included in Table 2.

3. Image files in both folders should contain matched and identically named/ordered images. The program automatically pairs images across both channels.

4. Images used in the procedure section were acquired using emission wavelengths corresponding to the respective fluorophores and fluorescent dyes. Single image files of 8- or 16-bit depth are supported by the software package.

Table 1. Input file formats supported and not supported by coAnalyzer. Input file formats with their extensions and software compatibility.

Note: Hyperstack files are not compatible with coAnalyzer.

Table 2. Output files generated by coAnalyzer during image analysis.

The table lists each output file with its storage location, description, and file format.

Note: Color 1 label and Color 2 label refer to the user-defined names assigned to the two fluorescence channels. Numbered output files correspond to individual ROIs analyzed by coAnalyzer.

3. In the Color 1 label and Color 2 label text boxes, type the color labels of the respective imaging channels. These labels will be displayed in the analysis result windows and used as identifiers for M1 and M2.

Note: Labels of image channels will be displayed in the result windows and in the names of a few output files (Table 2).

4. In the Image Number text box below the Select ROI button, type “1” to begin the image analysis for the first image.

Notes:

1. Once the user follows the steps below and finishes analyzing the first image, type “2” in the Image Number text box to analyze the second image. Repeat these steps to analyze all images in the folder. To allow users to control the analysis step-by-step, coAnalyzer requires manual image indexing.

2. Users must track which image number they are on. In the case where a non-existent image number is typed, an error message will be displayed: “error, image ‘x’ does not exist.”

5. Adjust the contrast for images for Color 1 and Color 2 by dragging the scroll thumbs.

Note: This step allows the user to adjust the contrast of both channels. The contrast adjustment is for visualization only and does not affect the quantitative results (M1/M2/PCC).

B. Select ROIs

6. Click the Select ROI button. A ROI selection window (Figure 2) will appear. To select each ROI, draw a polygon to enclose it: click once to mark the first vertex of the polygon, and a dot will appear. Drag the cursor to the second vertex and click again. Repeat this process to close the polygon, double-click to confirm the selection of a ROI, and proceed to the next ROI (Figure 2). Each selected ROI will be marked with red lines and a red number. After selecting all the ROIs, press Enter to proceed to the next image number or step 7.

Critical: The Image Processing Toolbox and the Optimization Toolbox must be installed in MATLAB to perform this step.

Notes:

1. ROIs can be selected in either imaging channel. Once selected, the program automatically reproduces the same selection in the other channel.

2. There are no universal standards on the minimum number of images and ROIs to be analyzed. Users should use statistical methods to justify their sampling choices.

3. Users can terminate ROI selection by pressing “F”.

Figure 2. ROI selection. After clicking the Select ROI button, this window appears. Draw polygons to select ROIs on one image. Here, we used the model bacterium Myxococcus xanthus [6] that expresses two fluorescence-labeled motility proteins, AglZ-GFP and AgmU-mCherry. While AglZ is a cytoplasmic protein, AgmU localizes to both the cytoplasm and periplasm [7–9].

C. Merge images

7. The GUI of coAnalyzer (Figure 1) will reappear. Type “1” in the ROI Number text box under the Merge images button to begin the image analysis for the first ROI.

Notes:

1. This number defines which ROI to be merged for analysis. There is no maximum number of ROIs that can be analyzed by the program.

2. Users should type the same ROI number consistently in ROI Number text boxes across Merge images, Isolate signal, and Analyze to merge images and perform 1D and 2D analysis for a specific ROI number.

8. Under the Merge images button, select the colors of the first and second imaging channels from the drop-down menus.

Note: These colors should correspond to the emission wavelengths used during imaging.

9. Click the Merge images button to merge the images from the two imaging channels. A window (Figure 3A, D, G, J) will appear, which allows the user to assess the quality of the merged ROI. Close this window to continue.

Critical: To use the line scanning function (step 11), check the Line scanning box before clicking on the Merge images button.

10. The GUI of coAnalyzer (Figure 1) will reappear. Type “2” in the ROI Number text box under the Merge images button to merge the second ROI. Repeat steps 7–9 to merge multiple ROIs.

Figure 3. 1D analysis. (A, D, G, J) After clicking the Merge images button, images from two channels will be merged. Each channel is displayed in the color defined by the user. A merged image allows the user to assess image quality. If the Line scanning box is checked before clicking the Merge images button, the user can define a line along which the fluorescence signals will be scanned. (B–F, H–L) Scanning results along individual lines. Panels B, E, H, and K show the fluorescence profile of each fluorophore, while panels C, F, I, and L display the correlation between the fluorescence intensities of each imaging channel. While FrzCD, a cytoplasmic chemoreceptor of M. xanthus, is labeled by GFP, chromosomal DNA (nucleoid) is stained by Hoechst. Lbx1 is a transcription factor marker found in the nucleus of expressing neurons in the mouse brain. A mouse brain tissue was used for the colocalization analysis between Lbx1-mCherry and DAPI-stained neurons. Scale bars, 5 μm.

D. 1D analysis

11. If the Line scanning function is selected, draw a straight line on the merged image (Figure 3A, D, G, J) across the region the user wishes to analyze, and double-click to confirm. A red line will appear on the merged image (Figure 3A, D, G, J). The line scanning results will be shown as the fluorescence profiles of the fluorophore from each fluorophore (Figure 3B, E, H, K) and their correlation along the line (Figure 3C, F, I, L). The results of each ROI are saved as files named “LineScan1, LineScan2, …” and “LineTwoColor1, LineTwoColor2, …” in the Color 1 folder under the data directory assigned for the specific image number. The LineScan file shows a plot of pixel intensities of the two fluorophores against their positions (pixel), and the LineTwoColor file shows a plot of the correlation between pixel intensity of one fluorophore against the other fluorophore. These files are saved in the TIFF file format.

Notes:

1. The program automatically creates the data directories assigned for the image numbers in the Color 1 folder.

2. Line scanning is a tool for analyzing protein colocalization in one dimension. For analysis in two dimensions, skip this step and go to step 12.

E. 2D analysis

12. Close the line scanning results window, and the GUI of coAnalyzer (Figure 1) will reappear. Type “1” in the ROI Number text box under the Isolate signal button to isolate the fluorescence signals for the first ROI.

Note: This number defines which ROI will be analyzed.

13. Adjust the threshold to isolate the fluorescence signals from images for Color 1 and Color 2 by dragging the scroll thumbs.

Critical: These adjustment buttons set a threshold to isolate fluorescence signals from images for Color 1 and Color 2. The threshold value is a normalized value from 0 to 1. For the best result, set the threshold at a value that best isolates the fluorescence signals from images without highlighting the image background as a signal. For consistency, use the same threshold value for all ROIs/images. The threshold value is displayed numerically in the text boxes above the scroll thumbs and can be applied across all ROIs, provided it is not changed manually by dragging the scroll thumbs.

14. Click the Isolate signal button. A result window (Figure 4A, D, G, J) will appear. Evaluate the fluorescence signals that were isolated from the current ROI. Close the signal isolation result window. The GUI of coAnalyzer (Figure 1) will reappear. Type “2” in the ROI Number text box under the Isolate signal button to continue the analysis for the second ROI. Repeat steps 13 and 14 to isolate the fluorescence signals from multiple ROIs.

Pause point: The user cannot proceed to Analyze (step 16) without isolating the fluorescence signals. When this step is skipped, the program warns the user by displaying the error message “error, ROI number ‘x’ does not exist.”

15. After signal isolation, the GUI of coAnalyzer (Figure 1) will reappear. Type “1” in the ROI Number text box under the Analyze button to analyze the first ROI.

Note: This number defines which ROI to be analyzed.

16. Click the Analyze button. 2D scatterplots (Figure 4B, E, H, K) and colocalization coefficients (Figure 4C, F, I, L) will appear. The results of each ROI are saved as files named “Scatterplot1, Scatterplot2, …” and “barplot1, barplot2, …” in the Color 1 folder under the data directory assigned for the specific image number. The “Scatterplot” file shows a scatterplot of the pixel intensity of one fluorophore against the other fluorophore, while the “barplot” file contains M1, M2, and PCC measurements of the two fluorescence signals. These files are saved in TIFF and Encapsulated PostScript file formats.

Note: The program automatically creates the data directories assigned for the image numbers in the Color 1 folder.

17. Close the results window. The GUI of coAnalyzer (Figure 1) will reappear. Type “2” in the ROI Number text box under the Analyze button to continue the analysis for the second ROI. Repeat steps 16 and 17 to analyze multiple ROIs.

Note: The quantitative results of each ROI will be displayed separately. To combine the results from multiple ROIs, go to step 18.

Figure 4. 2D analysis of single ROIs. (A, D, G, J) Signal isolation results. (B, E, H, K) scatterplots of the pixel intensities of one fluorophore against the other. (C, F, I, L) Bar plots presenting the M1, M2, and PCC measurements of the two fluorescence signals. For additional details, see the Validation of protocol section. Scale bars, 5 μm.

F. Combine results from multiple ROIs/images

18. After analyzing all the ROIs, the GUI of coAnalyzer (Figure 1) will reappear. Type “1” in both the Image from and image to text boxes under the Combine button to combine the results from all ROIs within image number 1. To combine the results from multiple images, type the numbers of the first and last images into the Image from and image to text boxes, respectively.

Note: For instance, to combine all the ROIs in images 1 and 2, type “1” and “2” into the Image from and image to text boxes.

Pause point: All images must be fully analyzed before combining results. If unanalyzed images are combined, the program will display the error message “error, data for image ‘x’ does/do not exist”.

19. Click the Combine button. The result window for combined results from multiple ROIs/images (Figure 5) will appear. This graphical result is automatically saved as a file named “hist.tif.” under the “sum” directory in the Color 1 folder. Numerical data are saved in files named “M1.txt.” and “M2.txt.” for Manders’ colocalization coefficients and “PCC.txt.” for Pearson’s correlation coefficient (PCC). These files contain the colocalization coefficient values calculated for each ROI in all the analyzed images in the Color 1 folder (one value per ROI). Colocalization coefficients for individual ROIs in each image can be found under the data directory assigned for the specific image number in the Color 1 folder. These files are saved in the .txt file format.

Note: The program automatically creates the sum directory and the data directories assigned for the image numbers in the Color 1 folder.

Figure 5. Results for combining multiple ROIs/images. Here, we show the colocalization of GFP-labeled FrzCD and Hoechst-stained DNA (nucleoid). Histograms of M1, M2, and PCC are calculated from 42 cells in 20 images.

Validation of protocol

We used the colocalization between the proteins AglZ and AgmU (alias GltD) in the rod-shaped gliding bacterium Myxococcus xanthus to validate coAnalyzer [9]. Previous reports showed that fluorescence-labeled AglZ and AgmU both assembled in the gliding motor complexes, which formed clusters along the cell body [7]. We imaged M. xanthus cells that express AglZ-GFP and AgmU-mCherry and visualized their colocalization using coAnalyzer. Merged images show that the localization patterns of the GFP-labeled AglZ and mCherry-labeled AgmU largely overlap in M. xanthus cells (Figure 3A and Figure 5 in [7]), which is confirmed by the fluorescence profiles generated by the line scan function (Figure 3B and Figure 5 in [7]). Using the same function, the correlation between the fluorescence intensities of AglZ-GFP and AgmU-mCherry along the user-defined line was further calculated, where R² > 0.5 indicates significant correlation between these two proteins but nonidentical localization (Figure 3C and Figure 5 in [7]). In contrast, AgmU and FrzCD, a cytoplasmic chemoreceptor, occupy mutually exclusive positions in M. xanthus cells [7]. Merged images and line scanning results indicate that the localization patterns of FrzCD-GFP and AgmU-mCherry do not overlap (Figure 3D–F and Figure 5 in [7]). A study by Moine et al. [10] reported that FrzCD binds the nucleoid of M. xanthus with its N-terminal domain. We imaged FrzCD-GFP and Hoechst-stained nucleoid in M. xanthus cells and tested coAnalyzer for its ability to analyze the colocalization between proteins and subcellular structures. Consistent with the reported finding, both the merged images and the line scanning results indicate significant colocalization between FrzCD and nucleoids (Figure 3G–I).

Using 2D analysis, scatterplots show significant intensity correlations between AglZ-GFP and AgmU-mCherry and between FrzCD-GFP and nucleoids (Figure 4B, H). These correlations are further confirmed by the high values of PCC, M1, and M2 (Figure 4C, I). In contrast, the scatterplot obtained from FrzCD-GFP and AgmU-mCherry does not show significant correlation (Figure 4E), which is consistent with the low values of PCC (0.43), M1 (0), and M2 (0) (Figure 4F). In contrast, the mean values of M1 (FrzCD, 0.6), M2 (nucleoid, 0.5), and PCC (0.75) from 42 cells in 20 images confirm significant overlap between FrzCD-GFP and nucleoid on the population level (Figure 5). Using coAnalyzer, we have validated the conclusion regarding colocalization in several publications [11–13].

Lbx1 is a transcription factor marker of dI4-dI6 neurons. In post-mitotic neurons at 10 days in vitro, neurons expressing Lbx1 only account for approximately 15% of DAPI-stained neurons [14]. Figures 3J and 4J show the localization between Lbx1-mCherry and DAPI-stained neurons in a mouse brain tissue [14]. While the scatterplot from individual ROIs shows a significant intensity correlation (Figure 3L), such a correlation does not exist at the population level (Figure 4K), which is confirmed by the low value of PCC (0.37) (Figure 4L). However, the calculated M1 (Lbx1) is 0.80, indicating that most of Lbx1-mCherry expressed neurons overlap with DAPI-stained neurons. In contrast, the calculated M2 (DAPI) is 0.14 (Figure 4L), indicating that only a small population (14%) of the DAPI-stained neurons express Lbx1-mCherry. Taken together, coAnalyzer generates reliable scatterplots and colocalization coefficients in both single-cell and multicellular samples. Importantly, besides colocalization analysis, coAnalyzer could be used to determine the expression patterns of proteins in multicellular samples.

To demonstrate the ability of coAnalyzer on highly complex localization patterns, we analyzed the colocalization between PDHA1 and mitochondria in Caenorhabditis elegans. PDHA1 is a subunit of pyruvate dehydrogenase E1, which is a component of the pyruvate dehydrogenase complex that localizes in the mitochondrial matrix. To validate the ability of coAnalyzer in analyzing signals in complex, irregular shapes, we obtained confocal microscopy images where PDHA1 and mitochondria were labeled with YFP and CFP, respectively [15]. Due to the cylindrical shape of C. elegans, both PDHA1 and mitochondria were illuminated unevenly (Figure 6A). It is thus difficult to analyze colocalization using the entire illuminated regions. To separate the fluorescence signals from their local backgrounds, we selected 16 small ROIs (Figure 6A). coAnalyzer can calculate M1, M2, and PCC for each ROI and combine all the data together (Figure 6B). The mean values of M1 (mitochondria) and M2 (PDHA1) are 0.60 and 0.66, respectively (Figure 6B), indicating that there is significant overlap between PDHA1 and mitochondria. However, the mean value of PCC is 0.53, indicating that PDHA1 does not occupy the entire mitochondria. Taken together, our results indicate that through the selection of multiple ROIs, coAnalyzer can quantify colocalization on complex images.

The results described above were generally consistent with those obtained using other colocalization analysis tools, such as Coloc2 (https://imagej.net/plugins/coloc-2) and JACoP (https://imagej.net/plugins/jacop). However, colocalization coefficients may differ across software packages due to variations in ROI selection, thresholding methods, and other analytical parameters. Therefore, to ensure reliable statistical comparisons among samples, we strongly recommend performing all analyses using the same software and processing workflow.

Figure 6. Colocalization quantification of PDHA1 and mitochondria by multi-ROI selection. (A) Fluorescent images of PDHA1-YFP and CFP labeled mitochondria in live C. elegans [15]. PDHA1 localizes in the mitochondrial matrix, where it serves as the catalytic E1 alpha subunit of the pyruvate dehydrogenase complex. Sixteen manually selected ROIs are outlined in red color. (B) Histograms of M1, M2, and PCC from the 16 ROIs.

General notes and troubleshooting

General notes

1. Due to the diffraction of light, the resolution of fluorescence microscopy is limited to 200–250 nm [16], much larger than the distance between interacting macromolecules [17]. For this reason, overlapping fluorescence signals from different macromolecules do not guarantee interactions. Thus, conventional fluorescence microscopy usually serves as an auxiliary or screening tool in the studies of molecular interactions. Interactions at the molecular level must be confirmed by genetic, biochemical, biophysical, and super-resolution microscopy methods [18].

2. Although coAnalyzer provides an all-in-one package for image analysis, it is critical to select the right methods for the presentation of colocalization. For qualitative visualization, the image merging and line scanning functions will be sufficient. If proteins and/or structures show irregular patterns and no single line reflects their distributions, researchers could select ROIs and use the scatterplots from ROIs to present intensity correlations.

3. In contrast, M1, M2, and PCC provide quantitative information for colocalization. These coefficients become the only choices when researchers combine data from multiple ROIs and images. Compared with M1 and M2, which directly measure the spatial overlap, PCC also quantifies the correlation between the intensities of two fluorophores. While high PCC values confirm strong colocalization, low PCC values do not suggest the lack of colocalization. As shown in Figures 4 and 6, partial colocalization typically results in low PCC values. In such cases, M1 and M2 become better choices in the presentation and quantification of colocalization.

Troubleshooting

Problem 1: The GUI of coAnalyzer fails to launch.

Possible cause: The coAnalyzer folder is not in the MATLAB path.

Solution: Add the folder path of coAnalyzer to the MATLAB path and run the “co3.fig” figure file.

Problem 2: Unable to select ROI(s) in the ROI selection window.

Possible cause: The Image Processing Toolbox and the Optimization Toolbox are not installed.

Solution: Install the Image Processing Toolbox and the Optimization Toolbox in MATLAB before running the program.

Problem 3: The program displays the error message “error, image ‘x’ does not exist” when the user clicks the Select ROI button.

Possible cause: There is a mistake in the folder paths for images from one or both imaging channels.

Solution: Type the right folder paths for the Color 1 folder and the Color 2 folder.

Problem 4: There is a mismatch in the image files across image channels in the ROI selection window (Figure 2).

Possible cause: There is an error in the order of images in the image folders.

Solution: Correct the order of images in the image folders by renaming image files correctly in the right order of choice.

Problem 5: The program displays the error message “error, data for image ‘x’ does/do not exist” or displays partial results when the user clicks the Combine button.

Possible cause: The user has typed a wrong image range, or the user has not analyzed specific image/ROI numbers within the image range before clicking the Combine button.

Solution: The user should type the correct image range, or the user should analyze all image/ROI numbers within the image range to display the full combined results.

Acknowledgments

Part of our research was supported by the National Institutes of Health under Award Number R01GM129000. We received financial support from Dr. David R. Zusman, who played no role in the design, execution, or presentation of this work. Images of mouse brain tissue and C. elegans were provided by Drs. Jennifer Dulin and L. René García.

Author contributions

Conceptualization, C.J.M. and B.N.; Investigation, C.J.M.; Writing—Original Draft, C.J.M. and B.N.; Writing—Review & Editing, C.J.M. and B.N.; Funding acquisition, B.N.; Supervision, B.N.

Competing interests

The authors declare no conflicts of interest.

Ethical considerations

While this protocol is applied to microscopy images derived from human and animal sources, it does not involve direct experimentation on human or animal subjects.

References

1. Radmacher, N., Chizhik, A. I., Nevskyi, O., Gallea, J. I., Gregor, I. and Enderlein, J. (2025). Molecular Level Super-Resolution Fluorescence Imaging. Annu Rev Biophys. 54(1): 163–184. https://doi.org/10.1146/annurev-biophys-071524-105321
2. Adler, J. and Parmryd, I. (2010). Quantifying colocalization by correlation: The Pearson correlation coefficient is superior to the Mander's overlap coefficient. Cytometry Part A. 733–742. https://doi.org/10.1002/cyto.a.20896
3. Dunn, K. W., Kamocka, M. M. and McDonald, J. H. (2011). A practical guide to evaluating colocalization in biological microscopy. Am J Physiol, Cell Physiol. 300(4): C723–C742. https://doi.org/10.1152/ajpcell.00462.2010
4. Adler, J. and Parmryd, I. (2012). Colocalization Analysis in Fluorescence Microscopy. Methods Mol Biol. 97–109. https://doi.org/10.1007/978-1-62703-056-4_5
5. MANDERS, E. M. M., VERBEEK, F. J. and ATEN, J. A. (1993). Measurement of co‐localization of objects in dual‐colour confocal images. J Microsc. 169(3): 375–382. https://doi.org/10.1111/j.1365-2818.1993.tb03313.x
6. Kroos, L., Wall, D., Islam, S. T., Whitworth, D. E., Muñoz-Dorado, J., Higgs, P. I., Singer, M., Mauriello, E. M., Treuner-Lange, A., Søgaard-Andersen, L., et al. (2025). Milestones in the development of Myxococcus xanthus as a model multicellular bacterium. J Bacteriol. 207(7): e00071–25. https://doi.org/10.1128/jb.00071-25
7. Nan, B., Mauriello, E. M. F., Sun, I. H., Wong, A. and Zusman, D. R. (2010). A multi-protein complex from Myxococcus xanthus required for bacterial gliding motility. Mol Microbiol. 76(6): 1539–1554. https://doi.org/10.1111/j.1365-2958.2010.07184.x
8. Nan, B., Chen, J., Neu, J. C., Berry, R. M., Oster, G. and Zusman, D. R. (2011). Myxobacteria gliding motility requires cytoskeleton rotation powered by proton motive force. Proc Natl Acad Sci USA. 108(6): 2498–2503. https://doi.org/10.1073/pnas.1018556108
9. Nan, B. and Zusman, D. R. (2011). Uncovering the Mystery of Gliding Motility in the Myxobacteria. Annu Rev Genet. 45(1): 21–39. https://doi.org/10.1146/annurev-genet-110410-132547
10. Moine, A., Espinosa, L., Martineau, E., Yaikhomba, M., Jazleena, P. J., Byrne, D., Biondi, E. G., Notomista, E., Brilli, M., Molle, V., et al. (2017). The nucleoid as a scaffold for the assembly of bacterial signaling complexes. PLos Genet. 13(11): e1007103. https://doi.org/10.1371/journal.pgen.1007103
11. Ramirez Carbo, C. A., Faromiki, O. G. and Nan, B. (2024). A lytic transglycosylase connects bacterial focal adhesion complexes to the peptidoglycan cell wall. eLife. 13: e99273. https://doi.org/10.7554/elife.99273
12. Zhang, H., Mulholland, G. A., Seef, S., Zhu, S., Liu, J., Mignot, T. and Nan, B. (2020). Establishing rod shape from spherical, peptidoglycan-deficient bacterial spores. Proc Natl Acad Sci USA. 117(25): 14444–14452. https://doi.org/10.1073/pnas.2001384117
13. Pogue, C. B., Zhou, T. and Nan, B. (2017). PlpA, a PilZ‐like protein, regulates directed motility of the bacterium Myxococcus xanthus. Mol Microbiol. 107(2): 214–228. https://doi.org/10.1111/mmi.13878
14. Dulin, J. N., Adler, A. F., Kumamaru, H., Poplawski, G. H. D., Lee-Kubli, C., Strobl, H., Gibbs, D., Kadoya, K., Fawcett, J. W., Lu, P., et al. (2018). Injured adult motor and sensory axons regenerate into appropriate organotypic domains of neural progenitor grafts. Nat Commun. 9(1): e1038/s41467–017–02613–x. https://doi.org/10.1038/s41467-017-02613-x
15. Goncalves, J., Wan, Y., Guo, X., Rha, K., LeBoeuf, B., Zhang, L., Estler, K. and Garcia, L. R. (2020). Succinate Dehydrogenase-Regulated Phosphoenolpyruvate Carboxykinase Sustains Copulation Fitness in Aging C. elegans Males. iScience. 23(4): 100990. https://doi.org/10.1016/j.isci.2020.100990
16. Goodwin, P. C. (2014). A primer on the fundamental principles of light microscopy: Optimizing magnification, resolution, and contrast. Mol Reprod Dev. 82: 502–507. https://doi.org/10.1002/mrd.22385
17. Kuriyan, J. and Eisenberg, D. (2007). The origin of protein interactions and allostery in colocalization. Nature. 450(7172): 983–990. https://doi.org/10.1038/nature06524
18. Akbarzadeh, S., Coşkun, Ã. and Günçer, B. (2024). Studying protein–protein interactions: Latest and most popular approaches. J Struct Biol. 216(4): 108118. https://doi.org/10.1016/j.jsb.2024.108118