Introduction

The interaction between adipose and muscle tissues has been highlighted in numerous studies in mammals and competition between the growth of both tissues has been reported1,2,3,4,5,6. For example, in mouse post-embryonic stage, an increase in adipose tissue mass has a negative effect on muscle mass7. In fish, evidences for these interactions remain limited. In zebrafish (Danio rerio), a defect in muscle development has been shown to increase intramuscular adipocyte infiltration8. Genetic selection of rainbow trout (Oncorhynchus mykiss) for muscle lipid content, also indicates that higher adiposity is associated with a higher proportion of large muscle fibers and a lower proportion of small fibers9,10.

Adipose tissue is mainly composed of mature adipose cells called adipocytes, filled with lipids contained in a unique droplet. In rainbow trout, two main adipose tissue deposits have been identified (perivisceral or subcutaneous)11, with specific proteomes and gene expression profiles11,12. Numerous in vitro studies prefer to use adipocytes derived from differentiated fibro-adipogenic precursors (FAP)13,14,15,16,17, although there are some differences with primary mature adipocytes (MAs), such as a smaller droplet size or an immature differentiation state18. However, the extraction and culture of MAs has been reported in gilthead sea bream (Sparus aurata)19, tilapia (Oreochromis mossambicus)20 and rainbow trout (Oncorhynchus mykiss)21. Despite the difficulty of extraction due to their buoyancy and their usually limited survival in culture22,23, in rainbow trout, the extraction process allows the use of mature cells capable of responding to stimuli such as the hormone leptin or exhibiting typical functions like lipolysis24.

Large and fast-growing fish such as rainbow trout exhibit continuous growth resulting from fiber hypertrophy and the formation of new muscle fibers known as hyperplasia25, at least during the exponential growth phase (up to 1 year). Fiber hypertrophy and hyperplasia require the presence of muscle stem cells, which are called satellite cells26. Satellite cells are quiescent in normal adult muscle and express pax7 gene, notably involved in their renewal. Satellite cell activation is rapidly followed by the onset of myoD (myoblast determination protein) expression, while myogenin marks the start of differentiation and myosin the end27. In fish, protocols have been developed to isolate mononucleated muscle cells (MMCs) from muscle tissue enabling the study of their proliferation and myogenic differentiation in vitro. Characterization of the extracted MMCs in trout indicates that approximately 60% of the cells are MyoD+ and thus are myogenic cells, while the identity of the remaining cells is still unknown28. The MMCs have been characterized in mammals, and it's noteworthy that, in addition to myogenic precursors, a significant proportion of FAPs expressing pdgfrα (Platelet-derived growth factor receptor alpha) are also found as muscle resident cells29,30.

Coculture of cells from muscle and adipose tissue is now common in mammals for investigating cellular communication31,32,33,34,35. Over the past decade, it has become clear that adipose and muscle cells communicate through multiple secreted factors, known as adipokines and myokines36,37,38, respectively. In mammals, adiponectin, a key adipokine, play a critical role in lipid metabolism39 and stimulates glucose uptake and fatty acid oxidation in muscle40. Other adipokines, such as leptin, are involved in the development of insulin resistance41 in muscle. Among the myokines, myostatin inhibits preadipocyte myogenic differentiation in vitro, and limits the formation of lipid deposits42. Additionally, muscle-derived interleukin-6 is known to increase uptake and oxidation of fats43 as well as adipocyte lipolysis44,45.

In fish, our understanding of the mechanisms underlying the interactions between adipose and muscle tissues is very limited. Numerous myokines and adipokines have been identified in fish, including in rainbow trout46,47,48, but there is limited evidence for their role in the cross-talk between these two tissues49. For example, in rainbow trout, receptors for adiponectin are found in muscle with differential regulation of their expression depending on situations such as fasting, suggesting a possible cross-talk between adipose tissue and muscle50.

Although primary cultures of MAs and MMCs from fish have been the subject of monoculture studies regarding their development, co-culture techniques have never been used to study cross-talk between these cell types. The aim of this study was to determine whether mature adipocytes influence the in vitro growth of rainbow trout mononucleated muscle cells. We cocultured these cells in a transwell system to avoid physical cell–cell interactions, but to allow cell–cell communication via soluble molecules, which is particularly relevant given that perivisceral and dorsal subcutaneous adipose tissues have no direct contact with skeletal muscle. Comparison of mononucleated muscle cell proliferation and myogenic differentiation in the absence or in presence of adipocytes evidenced a specific cross-talk from adipocytes to fibro-adipogenic progenitors and myogenic cells derived from rainbow trout muscle.

Materials and methods

Animals

Rainbow trout (Oncorhynchus mykiss) were reared in a recirculating rearing system under natural simulated photoperiod and at 12 ± 1 °C (pH 7.8–8.4; NH4 < 0.1mg/L). Fish were fed daily ad libitum with a commercial diet (Le Gouessant) and reared at the INRAE Fish Physiology and Genomic Laboratory (LPGP) experimental facilities (https://doi.org/10.15454/45d2-bn67, permit number D35-238-6, Rennes, France), delivered by French veterinary services. For tissue collection, fish were anesthetized with tricaine at 50 mg/L and euthanized with tricaine at 200 mg/L. All experimental procedures were carried out in strict accordance with the European Directive 2010/63/EU on the protection of animals used for scientific purposes. The euthanasia procedure was approved by the Ethical Committee for Animal Experimentation of Rennes (CREEA) and received the approval of French minister of national education, research and innovation under the authorization number: APAFIS # 2015121511031837. This study was carried out in compliance with the ARRIVE 2.0 guidelines.

Isolation and culture of mononucleated cells derived from trout muscle

For all studies, mononucleated cells were isolated from the dorsal part of the white muscle of juvenile rainbow trout (5 to 30 g body weight) as previously described28. Briefly, 20 to 80 g of white muscle were mechanically dissociated with scalpels and enzymatically digested by collagenase (Sigma #C9891) and trypsine (Sigma #T4799) prior to filtration (Falcon CellStrainer 100 μm #2360 and 40 μm #2340,). The cells were seeded onto poly-L-lysine and laminin precoated glass coverslips (Knittel, 13mm diameter) placed in a 24-well plate (Nunc, #142475) at a density of 80,000 cells/cm2 and incubated at 18 ◦C. Cells were cultured for 2 days in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FCS) (Sigma, #F7524) and 1% antibiotic–antimycotic solution (Sigma, #A5955). Cells were washed daily with DMEM 1% of antibiotics. The medium was then changed to a 1:1 DMEM and Leibovitz’s L-15 medium containing 10% FCS for last 3 days of monoculture or coculture. Finally, cells were washed twice phosphate buffered saline (PBS, pH 7.4) (Sigma, #P4417) and fixed with ethanol/glycine buffer (100% ethanol, 50 mM glycine, pH 2) or fixed 30min in 4% paraformaldehyde in PBS for in situ hybridization and then preserved in 100% ethanol.

Isolation and culture of mature adipocytes derived from trout adipose tissue

For all studies, mature adipocytes were isolated from two deposits of adipose tissue: perivisceral (PVA) or dorsal subcutaneous (SA) of rainbow trout (150 to 500 g body weight), as previously described21. Briefly, 5 to 40 g of adipose tissue were collected, cut into thin pieces (0.5–2 mm3), and incubated for 90 min in Krebs–Hepes buffer (consisting of a 1X dilution of Krebs solution buffered with HEPES buffer, NaHCO3, and D-Glucose) containing collagenase type II (125 U/mL; Sigma, #C6885) and 1% of bovine serum albumin (BSA) in a shaking platform at 17 °C. The cell suspension was then filtered at 300 µm for perivisceral tissue (pluriStrainer, 43-50300) and 200 µm for subcutaneous tissue (pluriStrainer, #43-50200). After two washes by flotation in Krebs-Hepes 1% BSA and two washes by flotation in Krebs-Hepes 2% BSA, cells were counted and cultured at 18 °C in DMEM/L15 (1:1) 10% FCS and 1% antibiotic–antimycotic solution, directly in a transwell (Corning, #3413) at a number of 5 × 106 (counted with an hemocytometer), unless another number is specified, until monoculture or coculture.

Adipocyte size distribution

At the end of cell extraction, a portion of the cells was placed between a microscope slide and a coverslip for microscopic imaging (Nikon digital camera coupled to an Olympus IX70 microscope). Ten images were captured per preparation. A Fiji macro was then used to automatically measure the diameter of each adipocyte. To avoid considering free lipid droplets, structures below 10 µm were excluded from the analyses.

Coculture of mononucleated muscle cells and mature adipocytes

After extraction, mononucleated muscle cells were cultured at 18 °C, at the density of 80,000 cells/cm2 in DMEM containing 10% FCS until day 2. After extraction, 5 × 106 adipocytes were cultured directly in a transwell (0.33cm2) with a 0.4 µm porous membrane in DMEM/L15 (1:1) with 10% FCS for 1 day. On day 2 of the muscle cell culture, transwells containing the mature adipocytes (24 h) were placed on top of wells containing MMC for 72 h of coculture (day 2–day 5) in DMEM/L15 (1:1) with 10% FCS. Finally, the MMCs were washed twice with PBS and fixed. No difference in adipocyte viability (calcein labeling) in presence or absence of myogenic cells was observed after 3 days of coculture (Supplemental Fig. S1). Moreover, we observed no morphological changes in our cells over time.

Preparation of conditioned medium (CM)

As illustrated in Fig. 3a, medium was collected at day 5 (72 h of coculture) to obtain coculture conditioned medium (CM CC). The media from the six wells per condition were pooled without filtration. Mature adipocytes and MMC were cultured separately in DMEM/L15 (1:1) 10% FCS until day 5 to obtain monoculture conditioned medium (CM MMC, CM MA). All conditioned media were frozen at − 80 °C after collection.

Analysis of proliferation and myogenic differentiation of mononucleated muscle cells

Cells were cultured in presence of 10 μM BrdU during 24 h before fixation at day 5. The cells were fixed with ethanol/glycine buffer. After three washes with PBS, MMC were saturated for 1 h with 3% BSA, 0.1% Tween-20 in PBS (PBST). Cells were incubated for 30 min at 37 °C with mouse anti-BrdU (Roche, #11,296,736; dilution 1/10) then washed before incubation at room temperature for 3 h with the primary antibody anti-myosin heavy chain (Hybridoma Bank, MF20; dilution 1/50;). Finally, cells were incubated with two secondary antibody anti-mouse for 1 h (Fisher anti-mouse IgG1 Alexa 488 #A21121, anti-mouse IgG2b Alexa 594 #A21145; dilution 1/1000). Nuclei were stained with a solution of 0.1 µg/mL DAPI (Sigma #D8417,) in PBS applied to the cells for 5 min. Cells were then mounted in Mowiol and photographed using a Nikon digital camera coupled to a Nikon Eclipse 90i microscope. Five images were taken per well and the number of BrdU positive nuclei, the number of nuclei in the myosin positive cells and the total number of nuclei was automatically calculated using FIJI software51 (version 2.14, https://imagej.net/software/fiji/).

RNAscope in situ hybridization and BrdU detection

Detection by in situ hybridization of pax7, myoD1/2, myogenin and pdgfrα transcripts in fixed MMCs was performed as previously described52. Briefly, MMCs were fixed with 4% PFA overnight at 4 °C and stored in 100% (v/v) ethanol at − 20 °C until use. Hybridization was performed using the RNAscope Multiplex Fluorescent Assay v2 (Bio-Techne, #323100) according to the manufacturer’s protocol. After rehydration, cells were placed in hydrogen peroxide solution (Bio-Techne, #322335) for 10 min, followed by Protease III solution (1/15) (Bio-Techne, #322337) at 40 °C for 10 min. Due to the presence of two myoD genes in the rainbow trout genome, we designed a set of probes targeting myoD1 and myoD2 mRNA. This probe set, as other probes, was hybridized at 40 °C for 2 h. The pax7, myoD1/2 (condition 72 h), myogenin or pdgfrα transcripts were detected using the fluorescent dyes Opal 520 (Akoya Biosciences, #OP-001001) and myoD1/2 (condition 24 h) was detected using the fluorescent dyes Opal 620 (Akoya Biosciences, #OP-001004). For the cells under the 72 h condition, after two washes with PBS, proliferation staining was followed by in situ labeling. Cells were saturated with 3% BSA in 0.1% Tween-20 in PBS (PBST) for 1 h. Cells were incubated with rabbit anti-BrdU (Akoya Biosciences, #PA5-32256; dilution 1/750) for 4 h at room temperature, washed, and then incubated with anti-rabbit the secondary antibody (Fisher, anti-rabbit IgG Alexa 594 #A21122; dilution 1/1000) for 1 h at room temperature. Cell nuclei of all conditions were stained with a solution of 0.1 µg/mL DAPI (Sigma, #D8417) in PBS applied to the cells for 5 min. The cells were then mounted in Mowiol and photographed using a Nikon digital camera coupled to a Nikon Eclipse 90i microscope. Five images were taken per well and 4 to 6 wells were used per condition.

Automated quantification of cells labeled by in situ hybridization

To automatically quantify the number of cells expressing these gene, we adapted a macro-command in the Fiji software to quantify puncta corresponding to the RNAscope labeling, per cell52. A cell was considered positive if at least 5 puncta were detected in a cell. Our quantification method is available at https://gitlab.univ-nantes.fr/SJagot/fijimacro_rnascopecells.

Statistical analyses

For analyses comparing proliferation or differentiation across multiple conditions and experimental repetitions (Figs. 1, 2, 3, 4), statistical analyses were conducted using the following approach: when sample size (n) exceeded 10 and the assumptions of parametric tests (such as normal distribution and homogeneity of variances) were met (confirmed by Shapiro–Wilk and Levene tests, respectively), a two-way ANOVA (Conditions and Experiments) followed by Tukey post hoc tests was applied. Otherwise, the non-parametric Scheirer-Ray-Hare test followed by Dunn's post hoc test with Bonferroni correction was utilized. The p-values reported in the text correspond to comparisons between different experimental conditions. The comparison of mean adipocyte diameters (n > 10 and meeting the test conditions validated by Shapiro–Wilk and Levene tests) was conducted using a t-test. Statistical analysis was performed using the chi-square test to compare the proportions of adipocytes with a diameter greater than 25µm between the two extraction in Fig. 2. The comparison of our two groups with sample sizes below 10 was conducted using the non-parametric Wilcoxon test. Specifically, the test was applied to compare the expression of genes between MMC and MMC + MA (Figs. 5, 6). For each figure, the n (as the number of wells for a condition or the number of adipocytes in Fig. 2b, c) is the same for proliferation and differentiation of a same condition and is indicated in the legend. All statistical tests were two-tailed, and the significance level (alpha) was set at 0.05. All the statistical analyses were performed with R (version 4.1.3, https://www.r-project.org/).

Figure 1
figure 1

Mature adipocytes stimulate proliferation but inhibit differentiation of mononucleated muscle cells. (a) Diagram of the isolation, extraction, culture and co-culture of mononucleated muscle cells (MMC) and mature adipocytes (MA) from perivisceral adipose tissue of rainbow trout. (b, c) Images of immunocytofluorescence analysis for myosin (MF20, red), BrdU (green) in MMC alone (MMC) or cocultured with MA (MMC + MA) after fixation on day 5. Cell nuclei are stained with Dapi (blue). (d, f) Quantification of cell proliferation, measured by BrdU incorporation, and myogenic differentiation (e, g), measured by immunocytofluorescence labeling of myosin (MF20). Compared to monocultured MMCs (n = 58), MMCs cocultured with MA (n = 47) for 72 h display higher proliferation (d) and lower differentiation (e). (f, g) Quantitative effect of adipocyte number (0 (n = 15), 5 × 105 (n = 11), 1 × 106 (n = 11), 5 × 106 (n = 11), 8 × 106 (n = 6)) on proliferation (BrdU) and differentiation (MF20) of MMCs. The different symbols correspond to different experiments. Statistical significance was determined by two-way ANOVA followed by Tukey post hoc tests, or Scheirer-Ray-Hare tests with Dunn post hoc and Bonferroni correction. Significance levels: ns (not significant), ** (p < 0.01), *** (p < 0.001).

Full size image
Figure 2
figure 2

Comparison of the effect of mature subcutaneous and perivisceral adipocytes on MMC. (a) Bright field images of mature adipocytes (MA) extracted from subcutaneous (SA) (n = 19,912) (left, red) and perivisceral (PVA) (n = 28,408) (right, orange) adipose tissue. (b) Comparison of the mean adipocyte diameter (> 10µm) (mean+/− SD µm) between the two extractions, showed a significantly higher adipocyte diameter in PVA. (c) Distribution of adipocyte size (> 10µm) from SA (red) and PVA (orange) showed a higher proportion of adipocytes with a diameter > 25 µm in PVA compared to those from SA. (d) MMC proliferation, measured by BrdU quantification and (e) myogenic cell differentiation, measured by immunocytofluorescence labeling of myosin (MF20). Compared to MCC monoculture (n = 16), MMC cocultured for 72 h with MA extracted from perivisceral adipose tissue (MMC + PVA) (n = 12), displayed higher proliferation and lower differentiation. When MMCs were cocultured with MA extracted from subcutaneous adipose tissue (MMC + SA) at an equivalent cell number (5 × 106) (n = 12) and duration (72 h), no significant differences were observed compared to MMCs cultured alone. However, an increase to 3 × 107 cells (n = 12) showed a similar effect on both proliferation and differentiation parameters, suggesting a response dependent on cell quantity. The different symbols correspond to different experiments. Statistical significance was determined by two-factor ANOVA followed by Tukey post hoc tests, or Scheirer-Ray-Hare tests with Dunn post hoc and Bonferroni correction. Significance levels: ns (not significant), * (p < 0.05), ** (p < 0.01), *** (p < 0.001).

Full size image
Figure 3
figure 3

Conditioned medium with mature adipocytes alone is sufficient to influence the development of mononucleated muscle cells in vitro. (a) Diagram of the isolation, extraction, culture for media conditioning, and diagram of isolation, extraction, and culture of mononucleated muscle cells (MMC) with either mature adipocytes (MMC + MA) or conditioned media: conditioned medium with MMC alone (CM MMC), conditioned medium with coculture (CM CC), conditioned medium with mature adipocytes (CM MA). (b) MMC proliferation, measured by BrdU quantification. Compared to MMC monoculture (MMC) (n = 22) or MMC culture with conditioned medium with MMC (CM MMC) (n = 17), exposure to conditioned medium with mature adipocytes (CM MA) (n = 16) or with coculture (CM CC) (n = 17) led to an increase in MMC proliferation (p < 0.001). This effect mirrors that observed in the presence of freshly extracted mature adipocytes from perivisceral tissue (MMC + MA) (n = 12). (c) Opposite effects were observed on myogenic cell differentiation, as revealed by immunocytofluorescence labeling of myosin (MF20). No significant change was observed with CM MMC, whereas CM MA and CM CC showed a marked reduction, similar to the effect of MA. The different symbols correspond to different experiments. Statistical significance was determined by two-factor ANOVA followed by Tukey post hoc tests, or Scheirer-Ray-Hare tests with Dunn post hoc and Bonferroni correction. Significance levels: ns (not significant), * (p < 0.05), ** (p < 0.01), *** (p < 0.001).

Full size image
Figure 4
figure 4

Time-dependent effects of mature adipocytes on the development of mononucleated muscle cells in vitro. (a) Diagram of the isolation, extraction and coculture of mononucleated muscle cells (MMC) and mature adipocytes (MA) from perivisceral adipose tissue, with fixation of MMC at different time points after the onset of coculture (24 h, 48 h, 72 h). (b) MMC proliferation (n = 18), measured by BrdU quantification, showed no significant difference at 24 h (n = 18) in the presence of MA, whereas a marked increase in proliferation was observed at 48 h (n = 18) and 72 h (n = 18) of coculture. (c) Myogenic cell differentiation, measured by immunocytofluorescence labeling of myosin (MF20), was slightly decreased at 24 h in the presence of MA, followed by a pronounced decrease at 48 h and 72 h of coculture. (d) Diagram of the isolation, extraction and culture of MMCs with adipocyte-conditioned medium (CM MA) using different exposure times (24 h, 48 h, 72 h) but fixation on day 5 for each condition. (e) MMC proliferation, measured by BrdU quantification, showed that a 24 h exposure to conditioned medium between day 4 and day 5 (n = 12) is not sufficient to increase proliferation compared to monocultured cells (n = 12) (p = 0.3), in contrast to 48 h (n = 12) or 72 h (n = 11) exposure to conditioned medium. (f) Myogenic cell differentiation, measured by immunocytofluorescence labeling of myosin (MF20), showed a significant reduction in differentiation already after 24 h of exposure to conditioned medium. The different symbols correspond to different experiments. Statistical significance was determined by two-factor ANOVA followed by Tukey post hoc tests, or Scheirer-Ray-Hare tests with Dunn post hoc and Bonferroni correction. Significance levels: ns (not significant), * (p < 0.05), ** (p < 0.01), *** (p < 0.001).

Full size image
Figure 5
figure 5

Early inhibition of myogenic differentiation of mononucleated muscle cells by mature adipocytes in vitro. (a) Images showing in situ hybridization analyses of pax7 (green) or myod1/2 (red) or myogenin (green) expression on monocultured mononucleated muscle cells (MMC) or MMC cocultured for 24 h with mature adipocytes (MA) from perivisceral adipose tissue. Cell nuclei are stained with Dapi (blue). (b) Quantification of MMC percentage expressing pax7, myod1/2 and myogenin revealed that matured adipocytes significantly reduced only the percentage of myogenin+ cells (n = 6 for each). Statistical significance was determined by the Wilcoxon test. Significance levels: ns (not significant), * (p < 0.05), ** (p < 0.01).

Full size image
Figure 6
figure 6

Mature adipocytes stimulate in vitro proliferation of fibro-adipogenic progenitors but not of myogenic cells derived from mononucleated muscle cells in vitro. (a) Images showing immunocytofluorescence detection of BrdU (red) associated with in situ hybridization analysis of myod1/2 (green) or pdgfrα (green) expression on monocultured mononucleated muscle cells (MMC) or MMC cocultured for 72 h with mature adipocytes (MA) from perivisceral adipose tissue. Cell nuclei are stained with Dapi (blue). (b) Quantification of MMC percentage expressing myod1/2 and pdgfrα indicated that mature adipocytes significantly increased pdgfrα+ cell proportion. (c) Proliferation of each cell type (myod1/2+ or pdgfrα+), was analyzed by detection of BrdU incorporation and showed no increase of BrdU+ percentage in myod1/2+ cells in coculture, but a significant increase in proliferation of pdgfrα+ (n = 6 for each). Statistical significance was determined by the Wilcoxon test. Significance levels: ns (no significance), * (p < 0.05), ** (p < 0.01).

Full size image

Results

Mature adipocytes influence the growth of mononucleated muscle cells in vitro

To determine whether mature adipocytes (MAs) could influence the in vitro growth of mononucleated muscle cells (MMCs), we measured the proliferation (BrdU+) and the myogenic differentiation (myosin+) of MMCs by immunofluorescence in the presence or absence of MA extracted from perivisceral adipose tissue (Fig. 1a). After 72 h of coculture or MMC monoculture, results showed the presence of numerous BrdU+ nuclei as well as mononucleated and multinucleated (myotubes) cells expressing myosin (Fig. 1b, c). Measurement of the percentage of BrdU+ nuclei showed a significant (p < 0.001) increase in the proliferation of MMCs in the presence of MAs (5 × 106), with approximately 39% of cells having proliferated in the last 24 h, compared to approximately 26% in the absence of MAs (Fig. 1d). In contrast, the percentage of nuclei expressing a late myogenic differentiation marker, myosin, decreased in the presence of MAs (5 × 106) compared to MMC monoculture (25% vs. 15%; p < 0.001) (Fig. 1e).

After observing the effect at a given number of MAs, we studied the effect with lower and higher quantities of MAs. Using different amounts of MAs (5 × 105 up to 8 × 106), we observed that the highest quantity of MAs induced the strongest stimulation of MMCs proliferation (Fig. 1f) and the strongest inhibition of myogenic differentiation (Fig. 1g). Moreover, these results indicated that even as few as 1 × 106 MAs were sufficient to affect both the proliferation and the myogenic differentiation of MMCs.

Adipose tissue origin influences the size of mature adipocytes and their effect on mononucleated muscle cells

To characterize the MAs that have been extracted from perivisceral and subcutaneous adipose tissue of rainbow trout, analysis of bright field images (Fig. 2a) showed a lower mean diameter in the subcutaneous (SA) compared to the perivisceral (PVA) extraction (21.5 µm vs. 24.3 µm, p < 0.001) (Fig. 2b). Overall, we observed a lower proportion of MAs > 25 µm in SA compared to PVA (25.5% vs. 36.5%, p < 0.001) (Fig. 2c).

We wondered if MAs extracted from subcutaneous adipose tissue would have the same effects on MMCs as previously observed with perivisceral MAs. To address this question, we established cocultures with MMCs and different amounts of subcutaneous MAs. For a same amount of MAs added (5 × 106), the percentage of MMC proliferation did not show a significant difference in the presence or absence of subcutaneous MAs (32% vs. 35%; p = 0.13), whereas a clear increase is observed with perivisceral MAs (32 vs. 44%; p < 0.001) (Fig. 2d). When examining the percentage of nuclei in myosin+ cells, no significant reduction in myogenic differentiation was observed in MMCs cocultured with 5 × 106 subcutaneous MAs compared to monoculture of MMCs (25% vs. 28%; p = 0.8). In contrast, with the same amount (5 × 106) of perivisceral MAs, a significant decrease in the percentage of nuclei in myosin+ cells was observed (28% vs. 20%; p = 0.002) (Fig. 2e). However, a sixfold increase (3 × 107) in the number of subcutaneous MAs gave results comparable to 5 × 106 perivisceral MAs, i.e. an increase in proliferation (32% vs. 42%; p < 0.001) (Fig. 2d) and a decrease in myogenic differentiation (28% vs. 21%; p = 0.008) (Fig. 2e) compared to MMC monoculture.

Mature adipocyte-derived soluble factor(s) influence the in vitro development of mononucleated muscle cells

In our previous experiments, we used indirect cocultures, in which both cell types share a common culture medium but are physically separated by a porous membrane (0.4 µm). The factor(s) contributing to the observed effects on MMCs should be soluble, smaller than the transwell’s pores, and able to diffuse through the culture medium. In order to confirm this hypothesis, we cultivated MMCs with different conditioned media as shown in Fig. 3a. Our analyses showed no significant effect on both proliferation (31% vs. 32%; p = 0.99) (Fig. 3b) and myogenic differentiation (32% vs. 31%; p = 0.88) (Fig. 3c) of MMCs cultured with medium conditioned by a previous MMC culture (CM MMC) alone compared to fresh monoculture of MMCs (MCC). In contrast, medium conditioned by either coculture (CM CC) or by MAs alone (CM MA) increased proliferation (31% vs. 38%; p < 0.001, 31% vs. 39%; p < 0.001) (Fig. 3b) and decreased myogenic differentiation (32% vs. 25%; p < 0.001, 32% vs. 23%; p < 0.001) (Fig. 3c) of MMCs. This effect of conditioned medium was comparable to that obtained with freshly extracted adipocytes cells (proliferation: 31% vs. 40% p < 0.001, myogenic differentiation: 32% vs. 19%; p < 0.001) (Fig. 3b, c). We observed the same effect with a medium conditioned by coculture of adipocytes and MMCs (CM CC) as with mature adipocytes alone (CM MA).

Mature adipocytes inhibit early differentiation of myogenic cells in vitro

To investigate the dynamics of the interaction between MAs and MMC development in vitro, we established coculture kinetics to determine the time required to observe a significant effect on proliferation and on myogenic differentiation (Fig. 4a). While a clear increase in proliferation of MMCs was observed after 48 h (29% vs. 37%; p < 0.001) and 72 h (30% vs. 39%; p < 0.001) of coculture, 24 h of exposure was not sufficient to observe a significant effect on MAs (19% vs. 23%; p = 0.051) (Fig. 4b). In contrast, a small but significant decrease in myogenic differentiation was observed already after 24 h of adipocyte exposure (9% vs. 5%; p = 0.0157), and was further enhanced at 48 h and 72 h (Fig. 4c).

To determine whether the responsiveness of MMCs to adipocyte-derived factor(s) changes during the culture, we exposed the MMCs to conditioned medium for different time periods and fixed the cells at the same developmental stage (day 5) (Fig. 4d). Exposure to conditioned medium for the last 48 h and the last 72 h, increased proliferation (26% vs. 36%; p = 0.014, 26% vs. 34% p = 0.033) (Fig. 4e) as well as decreased myogenic differentiation (33% vs. 23%; p < 0.001, 33% vs. 21%; p < 0.001) (Fig. 4f). However, while we did not observe a significant difference in proliferation of MMCs cultured in conditioned medium during the last 24 h (26% vs. 27%; p = 0.3) (Fig. 4e), we still found a reduction in myogenic differentiation (33% vs. 27%; p = 0.022) (Fig. 4f).

To determine at which stages adipocyte-derived factor(s) inhibits the myogenic program, we performed in situ hybridization with markers of satellite cells (pax7), myoblasts (myoD1/2) and myocytes (myogenin), after 24 h of coculture (Fig. 5a). As shown in Fig. 5b, 24 h of coculture did not change the percentage of pax7+ (73% vs. 70%; p = 0.55) or the percentage of myoD1/2+ cells (52% vs. 48%; p = 0.41). In contrast, we observed a significant decrease in the percentage of myogenin+ cells after 24 h of coculture compared to monoculture of MMCs (44% vs. 38%; p < 0.001).

Mature adipocytes stimulate proliferation of fibro-adipogenic progenitors but not of myogenic cells in vitro

We aimed to further characterize which population of MMCs proliferates in response to adipocyte-derived soluble factor(s). First, we performed in situ hybridization on the monoculture of MMCs with markers of fibro-adipogenic progenitors (FAPs) and myogenic cells, i.e. pdgfrα and myoD1/2, respectively (Fig. 6a). Our results showed that MMC monoculture contained 53% of myoD1/2+ cells and 55% of pdgfrα + cells, indicating that they represent the two major populations of mononucleated cells derived from white muscle (Fig. 6b). After 72 h of coculture, we observed that the percentage of pdgfrα + cells was increased compared to monoculture (55% vs. 66%, p = 0.0043), whereas no significant difference was observed for the percentage of myoD1/2+ cells (53% v .52%, p = 1) (Fig. 6b). To better determine which cell population was stimulated by MAs, we also performed double labeling with BrdU and in situ hybridization for myoD1/2 or pdgfrα. The results indicated that the percentage of BrdU+ cells within the myoD1/2+ population was similar between coculture and monoculture conditions (19% vs. 19%, p = 0.84), whereas the percentage of proliferative cells (BrdU+) in the pdgfrα + population increased (24% vs. 32%; p = 0.041) (Fig. 6c) when MMCs were cultured in the presence of MAs.

Discussion

The rainbow trout (Oncorhynchus mykiss) is an interesting model to study the communication between adipose and muscle tissues due to its different growth patterns compared to mammals. Indeed, trout exhibit an exponential muscle growth during the post-larval phase, associated with strong hyperplasic and hypertrophic muscle activity. In the specific context of salmonid models, the influence of adipose tissue on muscle growth remains poorly characterized. The aim of this study was to determine whether mature adipocytes from different adipose tissues can influence the proliferation and the myogenic differentiation of mononucleated muscle cells (MMCs) in vitro. Our main results provide direct evidence for the existence of cellular communication between mature adipocytes (MAs), fibro-adipogenic progenitors (FAPs) and myogenic cells in trout.

In vertebrates, two preferred storage sites for adipose tissue have been identified, perivisceral and subcutaneous, which are known to have different mobilization and metabolism53,54,55. In rainbow trout, differences in the size distribution of MAs and the abundance of certain proteins have been observed between the two tissue types, indicating different metabolic activities11,12. Our results confirm the difference in MA size between both adipose tissues, with a higher proportion of larger MAs in visceral adipose tissue compared to subcutaneous tissue. Because of these differences, we also compared the effect of MAs from both adipose tissues on MMC growth.

To assess the influence of trout MAs on the in vitro growth of MMCs, we measured the proliferation and myogenic differentiation in the presence or absence of adipocytes in an indirect coculture system. We used mature primary adipocytes, which brings us closer to in vivo conditions, whereas studies in mammals typically use in vitro differentiated preadipocytes. Indeed, adipocytes differentiated in vitro, due to their less advanced differentiation stage, may exhibit different properties compared to mature adipocytes, including differences in morphology, lipid storage capacity, gene expression, and metabolic activity18. Under our experimental conditions, we observed a strong stimulation of MMCs proliferation by perivisceral mature adipocytes, as well as an inhibited differentiation of myogenic cells. Subcutaneous adipocytes induced the same effect, but at a much higher number, suggesting a difference in secretome between perivisceral and subcutaneous MAs.

Such differences are far from being studied in fish species and only partially in humans35,56. Thus far however, our approach, by using BrdU incorporation, provides a specific and accurate measurement of proliferation in indirect coculture systems. These results are consistent with other studies in mammals showing an increase in proliferation of MMCs in response to preadipocytes or adipocytes, using less specific measures such as MTT assay reflecting viability57 or by assessing the increase in total cell number33. Our results suggest that mature adipocytes from perivisceral tissue may enhance the proliferation of MMCs in trout in vivo.

In addition, our results clearly showed that MAs inhibited differentiation of myogenic cells. These results are consistent with previous observations in mammalian models showing a decrease in myotube formation during indirect coculture in different models, such as in immortalized cell lines indirect coculture58, in rat muscle progenitors with adipogenic cells derived from rat muscle33 and in indirect cocultures of dedifferentiated chicken intramuscular adipocytes59. Interestingly, the inhibition of myogenic differentiation by MAs was characterized by a decrease in the percentage of myogenin+ and myosin+ cells but not of pax7+ and myoD1/2+ cells showing that the inhibition occurs from the early stage of differentiation, preventing the formation of myotubes. Accordingly, Takegahara et al. (2014) show that rat MAs decrease the percentage of myosin+ cells but not that of MyoD+ cells33. Inhibition of MMC myogenic differentiation, observed as early as 24 h, is earlier than previously reported in the literature for an indirect coculture with MAs whereas coculture durations of 2 to 5 days are generally required to observe such an effect33,58,59. Nevertheless, quantitative RT-PCR analyses show that expression of pax7, myoD, myogenin and myosin59 is reduced as early as 24 h in presence of chicken intramuscular preadipoctyes. This apparent discrepancy, can arise from existing differences between preadipocytes and mature adipocytes, but also to the technique used. Together, the marked and rapid reduction in myotube formation by MAs is probably due to early inhibition of myogenic differentiation.

Considering the absence of cell-to-cell contact in our experiments, the observed effects on muscle cells should be due to soluble factors. However, we cannot exclude the possibility that MMCs induce the production of factors by adipocytes, which in turn may affect their proliferation and myogenic differentiation. Our results showed that cultured MMCs with medium conditioned with both MMCs and MAs or with MAs alone, stimulated proliferation and inhibited myogenic differentiation to the same extend as freshly isolated adipocytes. These results confirm that MAs secrete one or more soluble factors that directly influenced MMCs growth in vitro, and that the production of these factors by adipocytes is independent of the MMCs. The nature of this factor is unknown, but it is known that adipocytes, as many other cells, secrete various molecules such as proteins, lipids, extracellular vesicles, etc. that could stimulate the proliferation of MMCs and inhibit the differentiation of myogenic cells33,58,60.

Since proliferation and myogenic differentiation are mutually exclusive cellular processes, we wondered whether increased proliferation would cause decreased myogenic differentiation, or vice versa, inducing a time lag in the onset of both effects. The kinetic of MMCs proliferation and myogenic differentiation, indicate the effect of MAs on myogenic differentiation was observed as early as 24 h, while the effect on proliferation was not observed until 48 h. Furthermore, incubation of MMCs with adipocyte-conditioned medium during the last 24 h of the culture (from day 4 to day 5) was sufficient to reduce myogenic differentiation but not proliferation. Taken together, these results indicate that the effects of MAs on myogenic proliferation and myogenic differentiation are only slightly time delayed, which cannot directly explain the increased proliferation of MMCs by MAs. We have previously shown that 2 days after MMCs extraction, some cells proliferate while others start to differentiate28, demonstrating the presence of cell subtypes at different stages of the myogenic program in the MMCs extracted from trout muscle. Therefore, we wondered whether adipocyte-secreted factors, in addition to inhibiting myogenic differentiation, would also stimulate the proliferation of the cells that are not yet engaged in the myogenic differentiation. Surprisingly, our results show that the proliferation of myogenic cells (myoD+) does not account for the observed increase in MMCs proliferation induced by MAs in contrast to the proliferation of fibro-adipogenic progenitors (pdgfrα+) that is stimulated. Several works report that preadipocytes or adipocytes enhance the proliferation of primary culture of MMCs57,61, but the identity of the proliferative cells has never been investigated. In contrast, MAs-induced stimulation of FAP proliferation has previously been observed in FAPs derived from adipose tissue in human, but never in muscle derived FAPs62,63. Thus, FAP proliferation in response to MAs-derived factor appears to be a conserved mechanism regardless of FAP origin.

In conclusion, we have demonstrated a cross-talk between mature adipocytes and mononucleated muscle cells in trout based on adipocytes-derived secreted factor(s) that stimulates proliferation of FAPs but inhibits differentiation of myogenic cells in vitro. Despite these findings, much remains to be explored regarding the diverse secretions of adipose tissue in fish, and further studies (proteomics, lipidomics, metobolomics) are needed to determine which specific adipocyte-derived factors may be responsible for the observed effects on mononucleated muscle cells in our experimental context.