Author response: Transferred mitochondria accumulate reactive oxygen species, promoting proliferation
Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Recent studies reveal that lateral mitochondrial transfer, the movement of mitochondria from one cell to another, can affect cellular and tissue homeostasis. Most of what we know about mitochondrial transfer stems from bulk cell studies and have led to the paradigm that functional transferred mitochondria restore bioenergetics and revitalize cellular functions to recipient cells with damaged or non-functional mitochondrial networks. However, we show that mitochondrial transfer also occurs between cells with functioning endogenous mitochondrial networks, but the mechanisms underlying how transferred mitochondria can promote such sustained behavioral reprogramming remain unclear. We report that unexpectedly, transferred macrophage mitochondria are dysfunctional and accumulate reactive oxygen species in recipient cancer cells. We further discovered that reactive oxygen species accumulation activates ERK signaling, promoting cancer cell proliferation. Pro-tumorigenic macrophages exhibit fragmented mitochondrial networks, leading to higher rates of mitochondrial transfer to cancer cells. Finally, we observe that macrophage mitochondrial transfer promotes tumor cell proliferation in vivo. Collectively these results indicate that transferred macrophage mitochondria activate downstream signaling pathways in a ROS-dependent manner in cancer cells, and provide a model of how sustained behavioral reprogramming can be mediated by a relatively small amount of transferred mitochondria in vitro and in vivo. Editor's evaluation This important work demonstrates that the transfer of dysfunctional mitochondria stimulates proliferation in recipient cancer cells by serving as a signal to induce reactive oxygen species production that in turn activates signaling pathways that control cell cycle. Compelling cell biology assays including rigorous microscopy with elegant reporters track the function and fate of transferred mitochondria in recipient cells. The work is relevant to the study of mitochondria, cancer, and immune cells and will be of broad interest to cell biologists and biochemists. https://doi.org/10.7554/eLife.85494.sa0 Decision letter Reviews on Sciety eLife's review process Introduction It has been previously described that mitochondria can undergo lateral transfer between cells (Torralba et al., 2016; Antanavičiūtė et al., 2014; Lou et al., 2012; Rebbeck et al., 2011; Tan et al., 2015; Wang and Gerdes, 2012; Wang and Gerdes, 2015; Lampinen et al., 2022). Mitochondria are dynamic organelles, known to provide energy for the cell, but more recently shown to have a variety of additional essential cellular functions (Zong et al., 2016). In animal models, a series of seminal studies revealed that cancer cells void of mitochondrial DNA still form tumors by obtaining mitochondria from stromal cells, thereby restoring cancer cell mitochondrial function, cellular respiration, and tumor formation (Tan et al., 2015; Dong et al., 2017). Other experiments suggest that mitochondrial transfer not only restores bioenergetics, but can alter the metabolic state of recipient cells (Brestoff et al., 2021; Nicolás-Ávila et al., 2020; Phinney et al., 2015; Saha et al., 2022; Crewe et al., 2021; Korpershoek et al., 2022; Liu et al., 2022; van der Vlist et al., 2022; Yang et al., 2022; Liu et al., 2021), allowing recipient cells to adapt to stressors or changes in the environment, prompting the development of methods targeting mitochondrial dysfunction in disease (Patel et al., 2023; Caicedo et al., 2015). Although these studies elegantly demonstrate that mitochondrial transfer alters recipient cellular behavior, many aspects of this process remain unclear. For instance, the rescue of cellular function is commonly attributed to enhanced mitochondrial energetic or metabolic profiles; however, the fate and function of transferred mitochondria in recipient cells are under-explored. Furthermore, it is unclear how cells respond to laterally transferred mitochondria if the recipient cells already have a fully functioning mitochondrial network, and in particular, if the transferred mitochondria only comprise a small subset of the overall mitochondrial network in the recipient cell. Given that metastasis is a low-frequency event and is the consequence of changes in cellular behavior on the single-cell level, we aimed to examine the function and behavior of transferred mitochondria within individual recipient cells that have functioning endogenous mitochondrial networks. Using a combination of in vitro high-resolution microscopy, optogenetics, imaging flow cytometry, and in vivo tumor models, we demonstrate a previously undescribed mechanism of mitochondrial transfer-associated cellular reprogramming. Collectively, our data explain how a relatively small amount of transferred mitochondria can impact cellular behavior in the recipient cell with fully functioning endogenous mitochondria – Transferred macrophage mitochondria in cancer cells are dysfunctional, ROS accumulates at the site of transferred mitochondria, promoting ERK-mediated cancer cell proliferation. Results Cancer cells with macrophage mitochondria exhibit increased proliferation We previously reported that macrophages transfer cytoplasmic contents to cancer cells in vitro and in vivo (Roh-Johnson et al., 2017), and hypothesized that a macrophage/cancer cell system would be ideal for probing mitochondrial transfer in cells with functioning mitochondrial networks. Our studies employed blood-derived human macrophages and a human breast cancer cell line, MDA-MB-231 (231 cells), stably expressing a mitochondrially localized mEmerald or red fluorescent protein (mito-mEm or mito-RFP, respectively; Figure 1a). We observed mitochondrial transfer from macrophages to 231 cells using live cell confocal microscopy (Figure 1b, arrowheads) and flow cytometry (Figure 1c–d; flow cytometry scheme in Figure 1—figure supplement 1a). Control gates were set to 0.2%, based on confirmation of mitochondrial transfer by FACS-isolation of distinct mEmerald+ populations (see methods for more information). With these methods, a range of transfer efficiencies were observed, which we attribute to donor-to-donor variability (Figure 1d), yet mitochondrial transfer was consistently observed in 231 cells, as well as to another breast cancer line, MDA-MB-468, and a melanoma cell line, A375 (Figure 1—figure supplement 1b). To determine whether macrophage mitochondrial transfer was unique to cancer cells, we tested a non-malignant breast epithelial cell line, MCF10A. We observed reduced mitochondrial transfer efficiencies to MCF10A cells, with no significant differences compared to control (Figure 1—figure supplement 1c), suggesting that macrophages exhibit higher mitochondrial transfer efficiencies to malignant cells. Transferred mitochondria contain a key outer mitochondrial membrane protein, TOMM20 (Figure 1—figure supplement 1d, arrowhead) and mitochondrial DNA (Figure 1—figure supplement 1e, arrowhead), suggesting that intact organelles are transferred to 231 cells. To better define the requirements for transfer, we performed trans-well experiments in which we cultured 231 cells either physically separated from macrophages by a 0.4 μm trans-well insert or in contact with macrophages (scheme in Figure 1—figure supplement 1f), or with conditioned media (Figure 1—figure supplement 1g, h). These data showed that mitochondrial transfer increased dramatically under conditions where 231 cells could contact macrophages directly (Figure 1—figure supplement 1g and h). Taken together, these results suggest that macrophage mitochondrial transfer to cancer cells likely requires cell-to-cell contact. Furthermore, while mitochondrial transfer may not be unique to cancer cells, macrophages transfer mitochondria to cancer cells at higher frequencies. Thus, due to the low rates of mitochondrial transfer across macrophage donors (0.84%, Figure 1d), we subsequently took advantage of single-cell, high-resolution approaches – rather than bulk approaches – to follow the fate and functional status of transferred mitochondria. Figure 1 with 3 supplements see all Download asset Open asset Cell-contact-mediated transfer of macrophage mitochondria leads to increased cancer cell proliferation. (a) CD14+ monocytes harvested from human blood are transduced and differentiated for 6 days. Mito-mEm +macrophages (green) are co-cultured with MDA-MB-231 cells (231 cells) expressing mito-RFP (magenta; right image). (b) Confocal image showing transferred mitochondria (green, arrowhead) in a 231 cell (magenta, cell outline in white). (c) Representative flow cytometry plots depicting mitochondrial transfer (black box) within a population of co-cultured mito-RFP+ 231 cells (right) compared to monoculture control (left) with background level of mEmerald (mEm) fluorescence set at 0.2%. (d) Aggregate data of mitochondrial transfer rates across macrophage donors. Each data point represents one replicate (N=14 donors). (e) Analysis of proliferative capacity by quantifying Ki-67 levels and DNA content in co-cultured 231 cells after 24 hr. Percentage of cancer cells within a specific cell cycle phase with or without transfer is shown. A significantly different percent of recipient cells occupies G2/M (black) phases of the cell cycle compared to non-recipient cells (N=4 donors; statistics for G2/M only). (f) Co-cultured recipient 231 cells have a significantly higher specific growth rate compared to non-recipients (N=60 cells (control), 115 (recipient) over 4 donors indicated as shades of gray). (g) Schematic of mitochondrial isolation and bath application on MDA-MB-231 cells. Mitochondria are isolated from mito-mEmerald expressing THP-1 monocytes and bath applied at 20–30 µg/mL for 24 hr. Cancer cells which had taken up mEm+ mitochondria are then FACS-isolated and plated for 48 hr for further analyses. (h) Representative confocal image showing mito-RFP-expressing 231 cell (magenta) that had taken up macrophage mitochondria (green, grey arrow). (i) 48 hr after FACS-isolating 231 cells with macrophage mitochondria, flow cytometry was used to determine percent of daughter cells which still contain mEm+ mitochondria. N=3 biological replicates. (j), Cell cycle analysis of daughter cells 48 hr after FACS-isolation of 231 cells that had taken up macrophage mitochondria. N=3 biological replicates. For all panels, standard error of the mean (SEM) is displayed and scale bars are 10 µm. Mann-Whitney (d), two-way ANOVA (e, j), Welch's t-test (f, i), *p<0.05; **p<0.01; ****p<0.0001. To determine the effects of macrophage mitochondrial transfer on cancer cells, we performed single cell RNA-sequencing on cancer cells that received macrophage mitochondria. These data revealed that mitochondrial transfer induced significant changes in canonical cell proliferation-related pathways (Figure 1—figure supplement 2a). To follow up on the RNA-sequencing results, we used flow cytometry to evaluate proliferation changes, and found significant increases in the percent of cells within the G2 and Mitotic (M) phases of the cell cycle in recipient cells, as compared to their co-cultured counterparts that did not receive mitochondria (Figure 1e; Figure 1—figure supplement 2b-d). These cells were not undergoing cell cycle arrest, as we found that recipient cells completed cytokinesis at rates equivalent to their co-cultured non-recipient counterparts (Figure 1—figure supplement 2e). For further confirmation of this proliferative phenotype, we used quantitative phase imaging (QPI) to detect changes in dry mass of co-cultured 231 cells over time (Zangle and Teitell, 2014). With this approach, we could obtain growth rate information of a large number of cancer cells over time (n=60 control cells; n=115 recipient cancer cells). Consistent with the flow cytometry-based cell cycle analysis, the specific growth rates increased significantly in 231 cells with macrophage mitochondria compared to 231 cells that did not receive mitochondria (Figure 1f). To examine whether the effects of mitochondrial transfer was sustained in recipient cells, we also measured the growth rates of daughter cells born from recipient 231 cells containing macrophage mitochondria (Zangle et al., 2014). We identified five 'parent' cancer cells with macrophage mitochondria, for which we were able to reliably follow both daughter cells upon division. Daughter cells that inherited the 'parent's' macrophage mitochondria exhibited an increase in their rate of change of dry mass over time versus sister cells that did not inherit macrophage mitochondria (Figure 1—figure supplement 3a-c). These experiments indicate that the proliferation phenotype in recipient cancer cells is sustained. Our results so far suggest that either macrophage mitochondrial transfer increases cancer cell proliferation, or that more proliferative cells are simply more capable of receiving macrophage mitochondria. Thus, to test between these hypotheses, we first blocked cells in the G1-phase of the cell cycle by treating co-cultures with a CDK4/6 inhibitor, Palbociclib (Figure 1—figure supplement 3d), and we observed no changes in mitochondrial transfer rates (Figure 1—figure supplement 3e). These data indicate that the enhanced proliferation observed in recipient cells is not due to proliferative cells more readily receiving transfer. We then performed experiments to rigorously test whether transferred macrophage mitochondria causes cancer cell proliferation, rather than mitochondrial receipt and proliferation being correlative events in cancer cells. We also wanted to determine whether the observed proliferative phenotype is due to macrophage mitochondria, and not other molecules that are passed along with the macrophage mitochondria. Thus, we biochemically purified mitochondria from a macrophage cell line, THP-1, and directly applied these macrophage mitochondria to cancer cells for 24 hr (Figure 1g). We then FACS-isolated cancer cell populations that contained purified macrophage mitochondria, and allowed this population to undergo additional rounds of cell division, and then reanalyzed the proliferative capacity of cancer cells that had retained the macrophage mitochondria versus cancer cells that had lost the macrophage mitochondria over this time. We first confirmed that cancer cells retained the macrophage mitochondria by imaging (Figure 1h). We also found that cancer cells that had retained the macrophage mitochondria exhibited an increased percentage of cells in the G2/M phase of the cell cycle compared to cancer cells that had lost the macrophage mitochondria (Figure 1i-j). Together with the QPI results, these results support the model that macrophage mitochondrial transfer promotes a sustained pro-growth and proliferative effect in both recipient and subsequent daughter cells. Transferred mitochondria are dysfunctional and accumulate ROS We next sought to understand how donated mitochondria can stimulate a proliferative response in recipient cells. We performed time-lapse confocal microscopy on co-cultures and found that in cancer cells with macrophage mitochondria, macrophage-derived mito-mEm+ mitochondria remained distinct from the recipient host mitochondrial network. Cancer cells were cocultured with macrophages for 12 hr and subjected to an additional 15 hr of timelapse microscopy, and we observed no detectable loss of the fluorescent signal at transferred mitochondria throughout the course of imaging (Figure 2a, arrowhead; Video 1). Thus, transferred macrophage mitochondria did not appear to fuse with the existing endogenous mitochondrial network in recipient cells. To probe the functional state of the donated mitochondria, we performed live imaging with MitoTracker Deep Red (MTDR), a cell-permeable dye that is actively taken up by mitochondria with a membrane potential (Poot et al., 1996). To our surprise, all of the transferred mitochondria were MTDR-negative (Figure 2b, top left). This was also confirmed using a different mitochondrial membrane potential-sensitive dye, Tetramethylrhodamine Methyl Ester (TMRM; Figure 1—figure supplement 1e). These results suggested that the transferred mitochondria lacked membrane potential. To determine whether these membrane potential-deficient transferred mitochondria were subjected to lysosomal degradation, we labeled lysosomes and acidic vesicles with a dye, LysoTracker, and found that the majority of transferred mitochondria (57%) did not co-localize with the LysoTracker signal (Figure 2b, top right). The status of transferred mitochondria was unexpected because mitochondria typically maintain strong membrane potentials, and dysfunctional mitochondria that lack membrane potential are normally degraded or repaired by fusion with healthy mitochondrial networks (Phinney et al., 2015). Next, we utilized another dye which stains cellular membranes, MemBrite, and observed that 91% of transferred mitochondria were not encapsulated by a membranous structure, thus also excluding sequestration as a mechanism for explaining the lack of degradation or interaction with the endogenous mitochondrial network (Figure 2—figure supplement 1a). These data, taken together with the long-lived observation of the transferred mitochondria in Figure 2a, suggest that transferred macrophage mitochondria lack membrane potential, yet remain as a distinct population in recipient cancer cells, not fusing with the endogenous host mitochondrial network nor subjected to degradation. Figure 2 with 2 supplements see all Download asset Open asset Transferred macrophage mitochondria are long-lived, depolarized, and accumulate reactive oxygen species, promoting cancer cell proliferation. (a) Stills from time-lapse imaging depicting the longevity of the transferred mitochondria (green, arrowhead) within a 231 cell (magenta, cell outline in white). Time elapsed listed in left corner. (b) Confocal image of a mito-RFP+ 231 cell (magenta) containing macrophage mitochondria (green, arrowhead) stained with MTDR (yellow) and LysoTracker (teal). MTDR does not accumulate in 100% of donated mitochondria (N=25 cells, 5 donors). Majority (57%) of donated mitochondria do not colocalize with LysoTracker signal (N=24 cells, 4 donors). (c) Ratiometric quantification of mito-Grx1-roGFP2 biosensor mapped onto the recipient 231 cell with fire LUT (top panel). Confocal image of mito-Grx1-roGFP2-expressing 231 cell (bottom right, green and yellow) containing a macrophage mitochondria (bottom left, red, arrowhead). (d) Ratiometric measurements of the mito-Grx1-roGFP2 sensor per 231 cell (paired dots) at a region of interest containing the host mitochondrial network (host) or a transferred mitochondria (transfer). Cells were co-cultured for 24 hr (N=27 cells, 3 donors indicated in shades of gray). (e) Exogenous purified macrophage mitochondria (green) is void of mitochondrial membrane potential (MitoTracker Deep Red-negative, yellow, arrowhead) in cancer cells. (f) Cell cycle analysis of cancer cells with exogenous purified macrophage mitochondria versus sister cells that did not take up exogenous purified mitochondria, either treated with vehicle or 100 μM mitoTEMPO (mitochondrially-targeted superoxide scavenger. N=3 donors; statistics for G2/M only). (g) Schematic of optogenetic experiments to generate data in (h). Cells expressing mito-KillerRed are photobleached in a specific ROI containing either cytoplasm only (left) or mito-KillerRed+ mitochondria (right). Following photobleaching, cells are imaged over time to quantify the amount of cell division. (h) Quantification of cell division after photobleaching. Each data point is the average within a field of view (N=13 experiments), with control (cyto) and experimental (mito) data shown as paired dots per experiment. Scale bars are 10 µm. Wilcoxon matched-pairs signed rank test (d, h), two-way ANOVA (f), *p<0.05; ****p<0.0001. Video 1 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Macrophage mitochondria are long-lived and remain distinct in recipient cancer cells. Video depicting a recipient mito-RFP expressing 231 cell (magenta) that contains mito-mEm macrophage mitochondria (green in magenta cell, center of frame). 231 cells were co-cultured with macrophages for 7 hr prior to the start of imaging for a duration ~15 hr with a time interval of 5 min. Maximum intensity projections of images are displayed at 12 frames per second, timestamp in upper left corner in hours (h), and scale bar is 10 μm. Given the surprising observation that transferred mitochondria lack membrane potential, we hypothesized that instead of providing a metabolic or energetic advantage, the donated mitochondria may act as a signal source to promote sustained changes in cancer cell behavior. This hypothesis could offer insight into how this rare event, in which a relatively small amount of mitochondria is transferred, could mediate sustained changes in the proliferative capacity of recipient cancer cells. One signaling molecule associated with mitochondria is reactive oxygen species (ROS), which occur normally as byproducts of mitochondrial respiration, and can be produced at high levels during organellar dysfunction (Schieber and Chandel, 2014). Using a genetically encoded biosensor, mito-Grx1-roGFP2, as a live readout of the mitochondrial glutathione redox state (Gutscher et al., 2008), we found that after 24 and 48 hr, significantly higher ratios of oxidized:reduced protein were associated with the transferred mitochondria versus the host network (Figure 2c–d; Figure 2—figure supplement 1b). These data indicate that transferred macrophage mitochondria in recipient cells are associated with higher levels of oxidized glutathione, suggesting that they are accumulating higher amounts of ROS. Consistent with these results, a second biosensor that is specific for the reactive oxygen species H2O2, mito-roGFP2-Orp1 (Gutscher et al., 2009), also reported more oxidation at the transferred mitochondria compared to the host network (Figure 2—figure supplement 1c–d) after 48 hr of co-culture. At 24 hr, we observed a similar trend, but no statistically significant difference (Figure 2—figure supplement 1d). These results indicate that ROS accumulate at the site of transferred mitochondria in recipient cancer cells. It is unclear whether the observed ROS accumulation is generated by the transferred mitochondria themselves, or generated elsewhere in the recipient cancer cell and accumulating locally at transferred mitochondria. Regardless of the source, we observed robust ROS accumulation specifically at the site of transferred mitochondria and with this unexpected finding, we next tested whether this ROS accumulation could serve as a molecular signal, regulating cell proliferation. To rigorously test the model that transferred macrophage mitochondria accumulate ROS, promoting cancer cell proliferation, we turned toward purified macrophage mitochondria approaches as in Figure 1g and sought approaches to reduce ROS levels. First, to better model the macrophage mitochondrial transfer to cancer cells that occurs in coculture conditions, we determined conditions for cancer cells to internalize exogenous macrophage mitochondria at rates similar to in vitro mitochondrial transfer conditions at 24 hr – 0.68% ± 0.36% internalization rate, n=3 biological replicates (compare to Figure 1d). We next determined that purified mitochondria taken up by cancer cells remain distinct, are not encapsulated by membranes after 24 hr (Figure 2—figure supplement 1e), and do not exhibit membrane potential (Figure 2e). Similar to our previous proliferation results with purified macrophage mitochondrial uptake at longer time points (Figure 1j), we found that cancer cells with internalized purified macrophage mitochondria (which, under these conditions, comprise of the exhibited a significant increase in proliferative cells in the G2/M phase of the cell compared to sister cells that did not internalize mitochondria (Figure bars in and that this increase was ROS is with a mitochondrially localized superoxide mitoTEMPO (Figure bars in cancer cells that did not internalize mitochondria were not by ROS (Figure bars in These results indicate that transferred mitochondria promote proliferation in a ROS-dependent To test whether ROS accumulation can induce cancer cell proliferation we stably a mitochondrially localized which ROS photobleached with et al., mito-KillerRed+ of interest induced ROS et al., Figure 2—figure supplement 2a). We then mito-KillerRed+ of interest that the of macrophage mitochondrial transfer to induce ROS in cancer cells, and the rate of cell division by imaging these cells over hr (Figure We found that cells with induced ROS mito-KillerRed+ exhibited an increased percentage of cells compared to control photobleached cells Figure Figure 2—figure supplement These results indicate that of mitochondrially localized ROS can directly promote cancer cell proliferation. ROS accumulation leads to proliferation We next aimed to determine how ROS may cell proliferation. ROS is known to induce downstream signaling pathways (Schieber and Chandel, 2014; et al., 2021), including signaling, a known to proliferation and et al., Thus, we sought to determine if cancer cells that had received macrophage mitochondria exhibited increased ERK We stably the et al., which from the to the cytoplasm ERK is in 231 cells cells with we used the imaging flow to in of cells that had or had not received macrophage mitochondria quantification and ERK signaling described in Figure supplements These data show that cancer cells with macrophage mitochondria have significantly higher cytoplasmic to ratios compared to cells that did not receive mitochondrial transfer, increased ERK (Figure Figure supplement Figure 3 with 4 supplements see all Download asset Open asset cancer cells exhibit proliferation. (a) was used to the of an in the or cytoplasm of co-cultured 231 cells that did (right) or did not (left) receive mitochondria (green, arrowhead). of and (b) ERK from data displayed in (d) mean fluorescence intensity N=3 donors indicated as shades of gray). (c) Confocal images of 231 cells expressing (green) and (magenta) with after control cytoplasmic or mito-KillerRed+ right). of (green) and (d) Quantification of to time Each represents a from a single cell. (e) Analysis of proliferative capacity by quantifying Ki-67 and DNA levels of co-cultured 231 cells treated with vehicle or ERK with or without transfer or (f), mitochondrial internalization after mitochondrial bath application donors; statistics for G2/M only). bars and scale bars are 10 Welch's t-test Mann-Whitney (d), two-way ANOVA *p<0.05; **p<0.01; ****p<0.0001. to our that cells that receive macrophage mitochondria exhibit increased ERK and that ROS is to induce cell proliferation, we then whether cancer cell mitochondrial ROS would directly ERK expressing both mito-KillerRed and in 231 cells, we induced ROS by mito-KillerRed+ and found that ROS increased that ROS is to increase ERK in cancer cells (Figure Figure supplement We next tested whether ERK signaling is for the mitochondrial cancer cell proliferation. We first determined an of an ERK that still ERK but does not dramatically affect 231 proliferation, as we sought to determine whether ERK mitochondrial proliferation, not proliferation more We first confirmed that with this of
Read more
