Research Spotlight: Dr. Charlie Bond

The endosomal-lysosomal system consists of a series of dynamic membrane-bound compartments that regulate sorting, trafficking, and degradation of cellular materials to maintain cellular homeostasis. Dysfunction in the endosomal-lysosomal system is linked to aging and multiple diseases, including Alzheimer's disease, cardiovascular disease, and various cancers (1,2). Late endosomes and lysosomes (LELs) are increasingly recognized as playing a diverse array of roles within the cell, from autophagy to scaffolding mTOR signaling (1). While over 100 lysosomal membrane proteins have been identified, it remains unclear whether each protein is present at similar levels in every LEL or if there are distinct LEL subtypes with unique combinations of surface proteins. Previous studies investigating the molecular composition of endosomes and lysosomes have been limited by their use of techniques like traditional light microscopy that lack the spatial resolution and sensitivity necessary to effectively characterize differences between individual organelles, and by the low throughput and high cost of higher-resolution methods like electron microscopy. A better means of understanding LEL heterogeneity is therefore needed. Unlike electron microscopy and traditional light microscopy, super-resolution light microscopy allows for the visualization of the inner architecture of cells with both nanoscale spatial resolution and relatively high throughput (3), allowing for resolution of individual proteins on individual organelles. 

DNA Point Accumulation in Nanoscale Topography (DNA-PAINT) is one example of super-resolution light microscopy. DNA-PAINT uses antibodies barcoded with short DNA oligonucleotides to detect, image, and quantify target proteins with single-molecule detection efficiency (4). In DNA-PAINT, fluorescent signal above background levels occurs when fluorescently-tagged imager oligos bind to their complementary oligo on the target antibody. The imager oligos float freely in solution, and randomly and stochastically bind to their complementary oligos. This transient binding creates a blinking effect, whereby only a few spatially distinct imager oligos are bound and in focus at any given time, thus allowing for clear visualization of individual fluorophores. Localization data obtained over the course of multiple rounds of imaging can then be reconstructed to create a higher-resolution image of protein localization, which allows for resolution below the diffraction limit (3). Given the quantitative nature of DNA-PAINT and its single molecule detection efficiency, recent CAMB-CPM graduate Dr. Charlie Bond from the Lakadamyali lab therefore sought to develop a quantitative, multiplexed DNA-PAINT super-resolution imaging pipeline that could be used to assess protein abundance and localization at the single LEL level and examine LEL heterogeneity under native conditions. 

Dr. Bond validated the suitability of the quantitative DNA-PAINT imaging analysis pipeline to identify protein abundance on individual LELs using the highly abundant and commonly studied LEL membrane proteins LAMP1 and LAMP2. He confirmed that both LAMP1 and LAMP2 predominately localized to vesicular compartments resembling LELs. He then developed a novel object-based colocalization analysis pipeline to determine the extent of colocalization between different proteins (i.e., LAMP1 and LAMP2) within a single object (i.e., a single LEL), as most existing colocalization methods are unable to provide information about colocalization with respect to a specific individual object. Briefly, he segmented individual LELs into a reference channel using either LAMP1 or LAMP2 positive signal in combination with a minimum size filter of 250 nm (representing a small LEL) to denote individual LELs. The segmented compartments identified as LELs were then used to denote the regions of interest for assessing the localization of the other LEL target proteins, with signal inside the region of interest above that of the signal outside the region of interest being deemed positive colocalization. Using this method, Dr. Bond observed over 90% of LAMP1-positive LELs overlapped with LAMP2-positive LELs in two different cell types regardless of whether LAMP1 or LAMP2 was used as the reference channel for the colocalization analysis. As LAMP1 and LAMP2 are known to be highly abundant on LELs, these findings suggest DNA-PAINT and the novel object-based colocalization analysis pipeline are capable of localizing proteins to the correct subcellular compartment. Importantly, there were no significant differences in LAMP1 abundance across five distinct biological replicates, further suggesting that the quantitative analysis pipeline is robust. There was also minimal colocalization between LAMP1 and early endosome marker EEA1, verifying that this method is capable of distinguishing lysosomes from early endosomes.

Dr. Bond then employed the quantitative DNA-PAINT pipeline to examine the abundance and localization of five additional lysosomal proteins (Cathepsin D, CD63, LAMTOR4, TMEM192, and NPC1) using either LAMP1 or LAMP2 as a marker of LELs. He found that the degradative enzyme and lysosomal marker Cathepsin D or its precursor localized to over 80% of LAMP2-positive LELs in two different cell types, suggesting that LAMP1, LAMP2, and Cathepsin D mark the same population of organelles. Unlike Cathepsin D, the highly abundant lysosomal membrane protein CD63 was present on 87 ± 6.8% of LAMP1-positive LELs in one cell type but varied between individual cells from 40% to nearly 100% in a different line. These data suggest that different cell types may contain different LEL subtypes, which could reflect cell-type-specific differences in the maturity or function of LELs. 

Unlike the highly abundant Cathepsin D and CD63, LAMTOR4, transmembrane protein 192 (TMEM192), and Niemann Pick Disease Type C1 protein (NPC1) were lowly abundant on the surface of LELs. LAMTOR4, which plays a critical role as a scaffold for Rag GTPases crucial for the recruitment and activation of mTORC1 on LEL membranes, was found on over 75% of LAMP1-positive LELs in two different cell lines despite its low abundance, suggesting LAMTOR4 is present at low levels in multiple LEL subpopulations. Interestingly, LAMTOR4 was found to form 83 nm nanoclusters on the LEL membrane. As LAMTOR4 plays a role in the recruitment of mTORC1 to the LEL membrane, these nanoclusters may facilitate efficient mTORC1 recruitment. Unlike LAMTOR4, both TMEM192 and NPC1 localized to only around 45% of LAMP1-positive LELs in both cell lines. The low colocalization of TMEM192 and NPC1 with LAMP1-positive LELs suggest that not all lysosomal proteins are found on every LEL and that these markers may be subpopulation-specific. Notably, NPC1 also localized in nanoscale domains on the LEL membrane, though  the nanoscale domains formed by NPC1 were more tightly packed (median diameter 55 nm) than those formed by LAMTOR4. As NPC1 is known to be involved in cholesterol export from LELs, these nanoclusters may be important for facilitating cholesterol export. Further validations using alternative antibodies, higher antibody concentrations, and an alternative colocalization method for TMEM192 and NPC1 similarly revealed that these proteins were only present in a subset of LELs, suggesting these findings are biologically significant and not an artifact of the study’s methodology. 

Dr. Bond then determined whether various lysosomal perturbations altered protein abundance and localization on LELs. He found that either overexpressing LAMP1 or treating cells with drugs that alter lysosomal pH altered protein abundance, colocalization, and/or nanocluster formation. These data indicate that the protein composition of LELs is sensitive to perturbation and that loss of homeostatic conditions, such as those occurring in disease states, may result in loss or gain of LEL subpopulations. As overexpression of LAMP1 is a common technique used to study lysosomes, these data also suggest that the results of prior studies using overexpression should be interpreted with caution. Moreover, these data highlight the utility of DNA-PAINT for studying lysosomes under native conditions. 

In addition to being sensitive to changing conditions, lysosomal function is influenced by the lysosome’s spatial positioning within the cell. Dr. Bond therefore examined the localization of different LEL subpopulations relative to other organelles. There did not appear to be a significant clustering of any subpopulations relative to the nucleus. However, there was a significant overlap between NPC1-positive LELs and mitochondria compared to NPC1-negative LELs in HeLa cells, but not in ARPE-19 cells. This suggests that subcellular positioning of distinct LEL subpopulations with respect to other organelles may also be cell-type specific. The positioning of NPC1-positive LELs near mitochondria in HeLa cells may also indicate that NPC1-positive LELs play a role in the delivery of cholesterol to the mitochondria in HeLa cells.

A key feature of DNA-PAINT is its capacity to image a large number of distinct targets. Dr. Bond therefore adapted a recently developed workflow for  high-order multiplexing (5) to visualize multiple LEL protein targets together. While DNA-PAINT has the capacity for multiplexing, the number of protein targets able to be imaged at one time has historically been limited by the low availability of high-quality antibodies from unique species and a limited number of spectrally distinct fluorophores. To overcome these barriers, Dr. Bond used primary antibodies preincubated with DNA-PAINT-labeled secondary nanobodies and developed a strategy for precise alignment of targets over multiple rounds of target imaging. With this method, they were able to multiplex imaging for four different markers and found that the predominant LEL subpopulation in HeLa cells definitively contained LAMP1, NPC1, and LAMTOR4, and likely also contained LAMP2 and CD63. They also identified a significant subpopulation of LELs that were LAMP1-positive but lacked NPC1 and LAMTOR4, demonstrating that not all LELs contain the same membrane proteins. Further highlighting the LEL heterogeneity, up to eight different LEL subpopulations were identified in ARPE-19 cells based on differential protein abundance. Notably, there was also variability in LEL protein composition within the same cell line, with some subpopulations being present in some cells but absent in others. This variability may suggest that not all LEL subtypes are functionally significant.

Through his thesis work, Dr. Bond developed a novel colocalization-based imaging analysis pipeline compatible with quantitative and multiplexed DNA-PAINT super-resolution imaging.With this technique, he identified previously unknown diversity in the protein composition of LELs and demonstrated the ability of the image analysis pipeline to characterize protein abundance and localization at the level of individual organelles. This methodology has broad implications for the field of cell biology, as it can be used to assess protein composition and localization within and between different types of organelles beyond just LELs. Future work extending this pipeline to 3D imaging and the incorporation of emerging advancements in the quality of labeling reagents, such as the development of synthetic nanobodies, will allow for a more complete characterization of protein abundance and localization across a variety of organelles in the future, which will better inform our understanding organelle structure and function. 

References

1. Bond, C., Hugelier, S., Xing, J. et al. Heterogeneity of late endosome/lysosomes shown by multiplexed DNA-PAINT imaging. Journal of Cell Biology, 224, 1 (2024). https://doi.org/10.1083/jcb.202403116

2. Cao, M., Luo, X., Wu, K. et al. Targeting lysosomes in human disease: from basic research to clinical applications. Sig Transduct Target Ther 6, 379 (2021). https://doi.org/10.1038/s41392-021-00778-y

3. Bond, C., Santiago-Ruiz, A. N., Tang, Q., & Lakadamyali, M. Technological advances in super-resolution microscopy to study cellular processes. Molecular Cell, 82, 2, 315–332 (2022). https://doi.org/10.1016/j.molcel.2021.12.022

4. Jungmann, R., Avendaño, M., Woehrstein, J. et al. Multiplexed 3D cellular super-resolution imaging with DNA-PAINT and Exchange-PAINT. Nat Methods 11, 313–318 (2014). https://doi.org/10.1038/nmeth.2835

5. Sograte-Idrissi, S., Schlichthaerle, T., Duque-Afonso, C. J. et al. Circumvention of common labelling artefacts using secondary nanobodies. Nanoscale, 12, 18, 10226-10239 (2020).