Propagation of a rapid cell-to-cell H₂O₂ signal over long distances in a monolayer of cardiomyocyte cells

Abstract

Cell-to-cell communication plays a cardinal role in the biology of multicellular organisms. H₂O₂ is an important cell-to-cell signaling molecule involved in the response of mammalian cells to wounding and other stimuli. We previously identified a signaling pathway that transmits wound-induced cell-to-cell H₂O₂ signals within minutes over long distances, measured in centimeters, in a monolayer of cardiomyocytes. Here we report that this long-distance H₂O₂ signaling pathway is accompanied by enhanced accumulation of cytosolic H₂O₂ and altered redox state in cells along its path. We further show that it requires the production of superoxide, as well as the function of gap junctions, and that it is accompanied by changes in the abundance of hundreds of proteins in cells along its path. Our findings highlight the existence of a unique and rapid long-distance H₂O₂ signaling pathway that could play an important role in different inflammatory responses, wound responses/healing, cardiovascular disease, and/or other conditions.

Graphical abstract

Highlights

• •Wounding induces an H₂O₂ cell-to-cell signal in a monolayer of cardiomyocytes.
• •The cell-to-cell signal requires H₂O₂ and O₂^.- accumulation along its path.
• •The signal propagates over several centimeters changing the redox state of cells.
• •Changes in the abundance of hundreds of proteins accompanies the signal.
• •The cell-to-cell signal requires paracrine and juxtacrine signaling.

1. Introduction

Cell-to-cell communication, including paracrine, juxtacrine, and autocrine signaling, plays a cardinal role in the development, maintenance, repair, and milieu/stimuli responses of multicellular organisms. Among the many signals that mediate cell-to-cell signaling, H₂O₂ plays a key role as a regulator of cellular and extracellular redox responses [[1], [2], [3], [4]]. In most cell-to-cell H₂O₂ signaling studies to date, cellular changes in calcium levels, and/or other signals, trigger the production of H₂O₂ by plasma membrane localized dual oxidase (DUOX) and/or NADPH oxidase (NOX)/superoxide dismutases (SODs) [[5], [6], [7], [8], [9], [10], [11], [12], [13]]. Peroxiporins [14,15] can then facilitate the entry of extracellular H₂O₂ into the cells that produced it, or neighboring cells, coordinating their responses to the stimuli [[1], [2], [3], [4]]. H₂O₂-producing complexes/mechanisms packed into exosomes, or proteins secreted by one cell, can also trigger H₂O₂ responses in neighboring cells [[16], [17], [18]]. H₂O₂ signals produced by different cells have multiple functions. For example, in response to wounding, H₂O₂ produced at the injury boundary generates a gradient that attracts leukocytes to the site of injury [[19], [20], [21]]. In other cases, H₂O₂ sensed by neighboring cells via redox-regulated receptors triggers in the receiving cells multiple responses that include changes in calcium levels and the activation kinases [22]. H₂O₂ can also alter the redox state of the producing and receiving cells, and/or the extracellular matrix, and trigger multiple responses [[1], [2], [3], [4]]. It is also possible that peroxiporins facilitate the diffusion of H₂O₂ produced inside the cell by different organelles/compartments contributing to cell-to-cell H₂O₂ signaling [[1], [2], [3], [4],18,22].

Overall, unless H₂O₂-generating mechanisms are packed into exosomes, it is believed that the distance an extracellular H₂O₂ signal can travel in mammalian tissues is in the range of 100–200 μm, and potentially up to 1 mm^19-21,23. This limitation is mainly attributed to the H₂O₂ scavenging activity of the extracellular matrix/neighboring cells, that generates a gradient from the producing cells, all the way up to the point H₂O₂ levels drop to normal levels at the tissue(s) they are generated at [[19], [20], [21],23]. Recent studies in plant and mammalian cells have, however, challenged this concept showing that H₂O₂ signals can propagate within minutes over long distances (measured in centimeters) in different plant tissues, as well as in monolayers of human MDA-MB-231 breast cancer and rat H9c2 cardiomyocyte cells, and isolated mice hearts [22,[24], [25], [26], [27], [28], [29]]. The mechanism that was proposed to drive this ‘long-distance’ H₂O₂ signal was termed the reactive oxygen species wave (‘ROS wave’) and it is based on the enhanced production of H₂O₂ by each cell along the path of the signal. An initiating cell would therefore produce H₂O₂ that is sensed by a neighboring cell and cause it to actively generate H₂O₂, triggering a cascade of cell-to-cell ‘enhanced H₂O₂ production’ state [22,[24], [25], [26], [27], [28], [29]]. In this respect, it is important to note that it is not H₂O₂ that diffuses over long distances, rather, it is a state of ‘enhanced H₂O₂ production’ that is propagating in a cell-to-cell fashion from the initiating cell(s) through the tissue, sometimes in a directional manner, with a rate of 0.15–0.4 cm/min [22,[24], [25], [26], [27], [28], [29]]. Although H₂O₂ does not diffuse over long distances per se in this cell-to-cell signaling mechanism, it is nonetheless required for the signal to propagate, as external application of catalase, NOX inhibitors, or the introduction of a loss-of-function NOX mutation into cells, block it [22,[24], [25], [26], [27], [28], [29]].

Although the ROS wave has been extensively studied in plants, only one report described it in monolayers of mammalian cells, as well as other eukaryotic organisms such as algae and amoeba (living in a community or film on an agar plate) [24]. Moreover, changes in protein/mRNA expression associated with the ROS wave were only determined in plants and algae [24,25,30]. Many questions regarding this mode of long-distance H₂O₂ signaling in mammalian cells remain, therefore, unanswered. Here, we provide evidence that the long-distance H₂O₂ signal in a monolayer of cardiomyocyte cells is accompanied by changes in protein abundance, enhanced cytosolic H₂O₂ levels, and altered cytosolic redox state, along its path, and that it depends on extracellular production of superoxide, as well as gap junction function.

2. Materials and methods

Cells culture. H9c2 (2–1) rat cardio myoblast (CRL-1446) cells were obtained from ATCC (https://www.atcc.org/) and grown in 37 °C 5% CO₂ RPMI medium 1640 supplemented with 10% FCS, l-glutamine, and antibiotics (ThermoFisher). A day before the experiment the media was replaced with FluoroBrite DMEM (ThermoFisher) to reduce background fluorescence [24,31]. For roGFP2 and roGFP2-Orp1 expressing cells, H9c2 cells were transfected with pMF1707 or pMOS016, obtained from Addgene (https://www.addgene.org/), using the Sigma's gene juice protocol, and selected using a Cytek Aurora cell sorter (Tecan) followed by growth in the presence of the appropriate antibiotics [[32], [33], [34]]. Cells were plated at an initial density of 500,000 per well for the 6-well plate, 25,000 per well for the 96-well plate, and 2,500,000 per dish for the 100 mm dishes and imaged using the IVIS (Revvity) or Lionheart FX (Agilent) apparatuses [24].

IVIS Imaging. Cells were imaged after 2 h of incubation with 20 μM Peroxy Orange 1 (Tocris Bioscience), or 5 μM Dihydroethidium (DHE; Sigma) under the conditions described above [24]. For inhibitor assays, 500U superoxide dismutase (SOD; Sigma), 500U catalase (CAT; Sigma), 50 μM APX115 (MedChemExpress) and/or 50 μM Setanaxib (MedChemExpress), or 600 μM Carbenoxolone (CBX; Sigma) and/or 180 μM and β-glycyrrhetinic acid (βGA; Sigma) were prepared as stock solutions in growth media and diluted (1:1000) into the media culture 1 h before imaging [24]. Injury was performed using an open-flame-heated metal rod (3 mm wide) that touched the local injury area for 1 s as described in Ref. [24]. Untreated, or injured cells, in the presence or absence of inhibitors, were imaged in the same plate/field of view. Images were acquired every 30 s for 30 min. Data analysis was performed with Living Image software using the Math functions. For signal accumulation images, the initial time image was subtracted from all other time images (hence, the absence of signal in the initial image). A signal intensity threshold was implemented based on the untreated field. Total radiant efficiency [24] was quantified by the software in selected regions of interest (ROIs) for local and systemic cells. Statistical analysis was performed as described in Ref. [24]. Briefly, data was obtained from at least 3–5 independent experiments, each including all treatments and controls in the same field of view with 3–5 biological repeats. Data was calculated as percent of 0 min for each treatment and presented in box plots with mean as X ± SE; median is line in the box and box borders are 25th and 75th percentiles; whiskers are the 1.5 interquartile range. Paired Student t-test was conducted using Microsoft Excel (*P < 0.05; **P < 0.01; ***P < 0.001).

For measurements with roGFP2 and roGFP2-Orp1, ratiometric imaging was conducted based on previous studies [28]. RoGFP molecules were excited in two different wave lengths for reduced (490 nm) and oxidized (420 nm) molecules. Fluorescence was collected through emission filter of 520 nm. Ratiometric images were generated using MATLAB (Matworks) using the RatioImage tool where each time point oxidized channel image was divided by the reduced channel image. Data was normalized by “total reduced” cells image (with 1 mM DTT) and “total oxidized” cells image (with 1 mM H₂O₂) [28].

Lionheart FX imaging. Cells were grown as described above and either incubated with 1.25 μM of CellROX orange (ThermoFisher) for 30 min or incubated with 5 μM DHE for 1 h, in the Lionheart FX apparatus under the growth conditions described above. After incubation cells were wounded using an open-flame-heated metal rod (0.5 mm wide) that touched the local injury area for 1 s at one edge of the well. Data was acquired via the Gen5 software (Agilent). Images were acquired simultaneously from the different treatments and controls every 1 min for 2 h. Image analysis was performed with Gen5 and data was exported to Prism 9.4.1 (GraphPad). Experiments were repeated at least 3 times with 3 biological repeats. Line graphs show mean fluorescence in a. u. With S.E. Results were tested for significance with one-way ANOVA.

Protein extraction and proteomics. H9c2 cells were grown as monolayers in 100 mm diameter petri dishes and incubated with DHE or PO1 as described above. The monolayers were then untreated or injured using an open-flame-heated metal rod (3 mm wide) that touched the local injury area for 1 sec²⁴. Control and wounded plates were then imaged using the IVIS apparatus for 30 min, as described above, and sampled, based on the imaging results, for proteomics analysis. Cells were scraped from the plates at defined areas (based on signal intensity and location) using cells scrapers (ThermoFisher) in PBS, spun down, immediately resuspended in 1X Laemmli buffer, heated to 95 °C for 10 min, and centrifuged at 16,000 g for 10 min. The supernatant was precipitated with cold acetone. Protein pellets were recovered by centrifugation at 16,000 g for 10 min and washed three times with 80% acetone. The final protein pellets were suspended in 6 M urea, 2 M thiourea, and 100 mM ammonium bicarbonate and protein concentration was determined using the Pierce 660 nm Protein Assay (ThermoFisher). Ten micrograms of protein from each sample were then cleaved with trypsin and subjected to proteomics analysis as described in Ref. [24]. The Uniprot ID UP000002494 database was used for protein searching and identification and significant changes (up or down) in protein abundance were determined using unpaired t-tests. Post-analysis based on both MS level and a significance threshold of q-value <0.05, and an absolute average log2 (fold change) > 0.58 were used to identify significant candidates [24]. Proteomics data are available via ProteomeXchange with identifier PXD047131.

Statistical analysis. All experiments were repeated at least three times with at least 3 biological repeats. Box plots graphs are presented with mean as X ± SE; median is line in the box and box borders are 25th and 75th percentiles; whiskers are the 1.5 interquartile range. Paired Student t-test was conducted using Microsoft Excel (*P < 0.05; **P < 0.01; ***P < 0.001).

3. Results and Discussion

The rapid long-distance H₂O₂signal is accompanied by changes in protein abundance in cells along its path. Monolayers of H9c2 Rat cardiomyocytes cells were injured, imaged, and sampled for proteomics analysis. Cells 0.1–1 cm away from the injury site (termed ‘local’ cells), as well as cells 3–4 cm away from it (termed ‘systemic’ cells), were sampled from injured, and uninjured (control) monolayers (Fig. 1A). In parallel, and to confirm that the cells we sampled were producing more ROS, changes in H₂O₂ and O₂^.- levels in response to the injury were measured (Fig. 1A). Proteomics analysis of cells sampled from the local and systemic locations of treated and untreated monolayers revealed that, compared to control unwounded cells from these locations, over 3400 and 1800 proteins significantly changed in their abundance in the local and systemic cells within 30 min, in response to wounding, respectively (Fig. 1B–S1; Tables S1–S3). These rapid and statistically significant changes in protein abundance could result from changes in the rate of protein translation, protein degradation, transcription, mRNA processing, or any other changes in the mechanism(s) that control protein steady-state levels in the cells participating in mediating the signal. Importantly, over 870 proteins that changed in abundance in local cells also changed in abundance in systemic cells (Fig. 1B; Table S3). These proteins could be directly associated with the ‘enhanced H₂O₂ production state’ triggered in the local and systemic cells in response to wounding (Fig. 1A). This group of proteins was enriched in molecular chaperone, oxidoreductase acting on sulfur, and membrane insertase activities (Fig. S1C), and included proteins such as ferredoxin, thioredoxin, MnSOD, catalase, peroxiredoxin, ferritin, and different proteins involved in glutathione metabolism (Table S3), suggesting that they are associated with altered redox/H₂O₂ levels in cells. An overlap between the proteins identified in response to wounding by our study and zebrafish proteins altered in response to H₂O₂ was also found (Fig. 3C) [35].

Fig. 1.

The wound induced H₂O₂ accumulation signal is accompanied by changes in protein abundance in cells along its path. A. Monolayers of rat cardiomyocytes (H9c2) grown in culture were wounded with a heated metal rod to induce injury (average injury diameter is 120 μm) [20] and O₂^.- and H₂O₂ accumulation were imaged using DHE and PO1 respectively. Representative time-lapse images of whole plate O₂^.- and H₂O₂ accumulation in treated and untreated plates are shown alongside bar graphs of combined data from all plates used for the analysis at the 0- and 30-min time points (local and systemic; Background levels of O₂^.- and H₂O₂, measured at time 0, were subtracted from all other time points). Local and systemic sampling locations are indicated with solid and dashed circles, respectively. B. Venn diagrams showing changes in protein abundance in local and systemic cells (Following the experimental design shown in A; Local and systemic sampling locations are indicated with solid and dashed rectangles, respectively), 30 min following wounding of the local cells as described above. C. Venn diagrams showing overlap between the proteins that changed in abundance in local (Left) or systemic (right) rat cardiomyocytes (B) and proteins that changed in abundance in response to H₂O₂ application in Zebrafish [31]. All experiments were repeated at least 3 times with 10 plates per experiment. Data is presented in A as box plot graphs; X is mean ± S.E., N = 30, **P < 0.01, *P < 0.05, Student t-test. Scale bar, 2 cm. Abbreviations: DHE, dihydroethdium; PO1, peroxy orange 1; TRE, total radiant efficiency. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3.

Inhibition of the wound induced O₂^.- accumulation signal with catalase (CAT), superoxide dismutase (SOD), or a mixture of two NADPH oxidase (NOX) inhibitors. A. Monolayers of rat cardiomyocytes (H9c2) grown in culture were wounded with a heated metal rod to induce injury [20] and O₂^.- accumulation was imaged using DHE in the presence or absence of SOD (added 60 min prior to wounding). Representative time-lapse images of whole plate O₂^.- accumulation in treated and untreated plates are shown alongside bar graphs of combined data from all plates used for the analysis at the 0- and 30-min time points (local and systemic; Background levels of O₂^.- measured at time 0, were subtracted from all other time points). Local and systemic sampling locations are indicated with solid and dashed circles, respectively. B. Same as in A, but instead of SOD, CAT was applied [20]. C. Same as in A, but instead of SOD, a mixture of two NOX inhibitors [APX115 (50 μM) and Setanaxib (50 μM)] [20] was applied (application of APX115 or Setanaxib did not inhibit the O₂^.- accumulation signal; Fig. S4). All experiments were repeated at least 3 times with 10 plates per experiment. Data is presented in A-C as box plot graphs; X is mean ± S.E., N = 30, *P < 0.05, Student t-test. Scale bar, 1 cm. Abbreviations: CAT, catalase; DHE, dihydroethdium; NOX, NADPH oxidase; SOD, superoxide dismutase; TRE, total radiant efficiency.

In addition to imaging H₂O₂ levels with the IVIS apparatus (Fig. 1A), we also used an automated fluorescent microscope to measure the rates of ROS accumulation in cells along the path of the signal (Fig. S2). This analysis revealed that the highest rates of ROS accumulation occurred close to the injury site, followed by lower rates at the middle and opposite edge of the long-distance H₂O₂ signal path (Fig. S2). This analysis further supported the model of enhanced H₂O₂ production by different cells along the path of the signal [4,24,25]. The identified proteins/pathways associated with the state of enhanced H₂O₂ production (Fig. 1, S1; Tables S1–S3) could be used in future studies to dissect the process of long-distance cell-to-cell H₂O₂ signaling, especially since some of them were not previously associated with responses to H₂O₂/ROS (Fig. 1C; Tables S1–S3). Although it would be ideal to study the proteomics responses of H9c2 monolayers treated with an external antioxidant, or composed of NOX/DOUX mutants, these treatments/manipulations would affect both the initiation as well as the propagation of the H₂O₂ signal and would not provide a high-resolution analysis of this process. Instead, methods to generate or print monolayers composed of sections that contain NOX/DOUX mutations flanked by sections that do not (like our studies in Chlamydomonas) [24] should be developed. Single-cell proteomics and/or transcriptomics studies should also be conducted in future studies, as they may uncover additional proteins and transcripts involved in this process.

Enhanced cytosolic accumulation of H₂O₂and changes in the redox sate of cells along the path of the H₂O₂signal. We previously imaged the long-distance H₂O₂ signal in different organisms using various H₂O₂/ROS-sensitive dyes [24]. These methods could however complicate the interpretation of results [2]. In addition, they did not measure changes in redox levels associated with the signal. We, therefore, generated monolayers of H9c2 rat cardiomyocyte cells expressing cytosolic roGFP2 (for redox), or cytosolic roGFP2-Orp1 (for H₂O₂), imaging (Fig. 2, S3). Analysis of wounded and unwounded H9c2 monolayers expressing these proteins revealed that the long-distance H₂O₂ signal is accompanied by enhanced cytosolic accumulation of H₂O₂ as well as changes in the redox state of cells along its path. These findings provide direct evidence for the putative function of the signal in altering the redox state of cells along its path, thereby, activating different redox responses (also supported by the identification of redox-response proteins in our proteomics analysis; Table S3).

Production of extracellular superoxide and gap junction function are required for the propagation of the long-distance H₂O₂signal. H₂O₂-mediated cell-to-cell signaling in monolayers of H9c2 cells was previously shown to be blocked by catalase application, as well as a mixture of the NOX inhibitors APX115 and Setanaxib (but not each of these inhibitors applied individually; suggesting that NOX1 or NOX2, and NOX4 might be involved in mediating this process) [24]. However, the role of superoxide in this signaling pathway was not substantiated. We therefore used external application of SOD, as well as imaging using DHE in our experimental system. As shown in Fig. 1, Fig. 3A, 3B we could image the progression of the signal using DHE, and this signal was blocked by SOD or catalase application. In addition, a mixture of two NOX inhibitors blocked the signal imaged with DHE (Fig. 3B–S4). These findings are highly intriguing, as they support a model in which both superoxide and H₂O₂ are essential for the propagation of the long-distance H₂O₂ signal. Adding excess SOD, that should have simply converted superoxide to H₂O₂ and supported the signal, was therefore found to, instead, block it (Fig. 3A). This could suggest that a minimal/threshold level of superoxide produced extracellularly is needed for the signal to propagate, and that superoxide could function in other, yet unknown, ways as a signal required for this pathway (as opposed to simply being a substrate for SODs to generate H₂O₂). Further studies are needed to address this intriguing possibility.

The ‘ROS wave’ long-distance H₂O₂ signal of plants was shown to require paracrine as well as juxtacrine signaling (the latter requiring the function of plasmodesmata, the plant equivalents of gap junctions) [22,[24], [25], [26], [27], [28], [29], [30]]. As cardiomyocytes contain multiple types of gap junctions, including some that may be used to secrete compounds from the cell to the extracellular matrix upon wounding [36,37], we tested the impact of different gap junction inhibitors on the cell-to-cell H₂O₂ signal. To test whether gap junction function is required for the long-distance H₂O₂ signal in H9c2 cells, we treated monolayers of H9c2 cells with CBX [38], βGA [39], or a mixture of the two [40], prior to wounding. The combination of CBX+βGA, but not CBX or βGA (potentially due to the presence of multiple gap junction types in cardiomyocytes) blocked the progression of the H₂O₂ signal, imaged with DHE or PO1 (Fig. 4, S5). Our findings (Fig. 2, Fig. 4) [24] support, therefore, a model in which paracrine and juxtacrine signaling, are both required for the propagation of the long-distance H₂O₂ pathway in mammalian cells (Fig. 4C). They further provide a potential mechanistic explanation to the directionality of the signal in H9c2 cells. While in other eukaryotic organisms, such as Chlamydomonas cells (that lacks gap junction) or MDA-MB-231 cells (growing as a monolayer with poor gap junction connectivity), the signal appears to diffuse away from the site of injury, without a clear direction, in H9c2 cells (that are highly connected to each other via multiple gap junctions), it appears to be directional, following the orientation of cells in the monolayer [24]. This directionality could be related to the many different gap junctions that connect H9c2 cells and direct the signal along a specific path [36,37]. In future studies, the polarity, and other aspects of the directionality of the signal, should be addressed, as these could play a key role in the function of the H₂O₂ signal in different tissues. As gap junctions are involved in mediating the long-distance cell-to-cell H₂O₂ signal (Fig. 4, S5), it is also possible that H₂O₂ produced inside the cell (e.g., by NOX 4 at the mitochondria) [41] participates in this pathway.

Fig. 2.

The wound induced H₂O₂ accumulation signal is accompanied by changes in cytosolic H₂O₂ and redox levels. A. Monolayers of rat cardiomyocytes (H9c2) expressing roGFP2-Orp1 grown in culture were wounded with a heated metal rod to induce injury [20] and cytosolic H₂O₂ levels were imaged as the ratio between 420/480 nm [24]. Representative time-lapse images of whole plate changes in cytosolic H₂O₂ levels in treated and untreated plates are shown alongside bar graphs of combined data from all plates used for the analysis at the 0- and 30-min time points (local and systemic; Background levels of cytosolic H₂O₂ levels, measured at time 0, were subtracted from all other time points). B. Same as A, but for rat cardiomyocytes (H9c2) expressing roGFP2. Calibration of the roGFP2-Orp1 and roGFP2 signals in rat cardiomyocytes are shown in Fig. S3. All experiments were repeated at least 3 times with 10 plates per experiment. Data is presented in A and B as box plot graphs; X is mean ± S.E., N = 30, **P < 0.01, ***P < 0.001, Student t-test. Scale bar, 1 cm.

Fig. 4.

Inhibition of the wound induced H₂O₂ and O₂^.- accumulation signal with gap junction inhibitors, and a model. A. Monolayers of rat cardiomyocytes (H9c2) grown in culture were wounded with a heated metal rod to induce injury [20] and O₂^.- accumulation was imaged using DHE in the presence or absence of a mixture of Carbenoxolone (CBX; 600 μM) and β-glycyrrhetinic acid (βGA; 180 μM), added 60 min prior to wounding. Representative time-lapse images of whole plate O₂^.- accumulation in treated and untreated plates are shown alongside bar graphs of combined data from all plates used for the analysis at the 0- and 30-min time points (local and systemic; Background levels of O₂^.- measured at time 0, were subtracted from all other time points). Local and systemic sampling locations are indicated with solid and dashed circles, respectively. B. Same as in A, but instead of DHE, PO1 (H₂O₂ accumulation) was used for imaging. C. A hypothetical model of the wound induced H₂O₂ and O₂^.- accumulation cell-to-cell signal. Each sell along the path of the signal is shown to produce H₂O₂ and O₂^.- propagating the signal across long distances in a process that requires NOX and gap junction function (i.e., a mixture of paracrine and juxtacrine signaling). See text for more details. All experiments were repeated at least 3 times with 10 plates per experiment. Data is presented in A and B as box plot graphs; X is mean ± S.E., N = 30, *P < 0.05, Student t-test. Scale bar, 1 cm. Abbreviations: βGA, β-glycyrrhetinic acid; CBX, Carbenoxolone; DHE, dihydroethdium; NOX, NADPH oxidase; PO1, peroxy orange 1; TRE, total radiant efficiency. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Summary and conclusions. Taken together, our findings reveal that the long-distance H₂O₂ cell-to-cell signaling pathway triggered by wounding in a monolayer of H9c2 cells (Fig. 4C) is accompanied by enhanced accumulation of H₂O₂ and altered redox state at the cytosol of cells along its path, as well as significant changes in the abundance of hundreds of proteins. In addition, we show that this pathway requires the production of superoxide as well as the function of gap junctions. As H₂O₂ signaling plays such a key role in mammalian physiology and development [[1], [2], [3]], more studies are needed to dissect this pathway and determine how it is involved in different inflammatory responses, wound healing, development, cardiovascular disease, and other diseases/conditions.

Funding

This work was supported by funding from the National Science Foundation (IOS-1932639, MCB2224839), and the National Institute of Health grant (GM111364, R35GM134919).

Use of AI and AI-assisted technologies-

None.

CRediT authorship contribution statement

Yosef Fichman: Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing. Linda Rowland: Data curation, Formal analysis, Methodology. Thi Thao Nguyen: Data curation, Formal analysis, Methodology. Shi-Jie Chen: Conceptualization, Writing – review & editing. Ron Mittler: Conceptualization, Funding acquisition, Investigation, Resources, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.