The relationships between growth rate and mitochondrial metabolism varies over time

The experiments were approved by the French Ethics Committee in charge of Animal Experimentation (no.2019072411491441) and were in accordance with institutional and ARRIVE guidelines.

Animal collection and husbandry

In May 2019, juvenile European sea bass, Dicentrarchus labrax (Linnaeus 1758) (6 months old, mass 5 g), were transferred from a fish farm (Turbot Ichtus, Trédarzec, France) to the Ifremer rearing facility (Plouzané, France). Fish were kept in a common tank for 5 months, maintained under a 12 L: 12 D photoperiod, and fed at satiety three times a week using commercial pellets (Neo Start, Le Gouessant, Lamballe, France).

In October 2019, fish (n = 40) were anaesthetized (Tricaïne; 125 mg L⁻¹), weighed (41.5 ± 1.8 g, MCE11201S-2S00-0, Sartorius, Göttingen, Germany), and implanted subcutaneously with an identification tag (RFID; Biolog-id, Bernay, France). The fish were then randomly allocated to ten replicate 400 L tanks supplied with open-flow, fully aerated seawater (oxygen saturation > 95%, salinity 32 ppt), thermo-regulated during winter to avoid falling below 13 °C, and fed at satiety three times a week. Temperature was recorded weekly. To account for the potential effect of temperature variation over the duration of the trial (15.5 ± 0.5 °C, range: 13.1–17.9 °C) on growth, a correlations analysis was performed between temperature and specific growth rate (SGR). No statistical relationship was found between SGR and temperature (Spearman R² = 0.060, P = 0.596). Additional fish (n = 40) were present in the tanks (final density: n = 8 per tank) for the need of another project.

Growth measurements

Body mass (BM) was measured about every four weeks from October 2019 to June 2020. The fish were fasted for 48 h and anesthetized before each BM measurement (± 0.1 g). The specific growth rate (% day^-1) was estimated as follows:

In March 2020, a red muscle biopsy sample was collected from fish to measure the mitochondrial metabolic traits. Past growth was defined as specific growth rates before the analysis of mitochondrial metabolic traits (past specific growth rate, SGR_past). SGR_past were calculated using the BM at the muscle biopsy as the final BM and the BM at 7, 11, 16, and 20 weeks before the muscle biopsy as the initial BM (Fig. 1). Future growth was defined as specific growth rates after analysis of mitochondrial metabolic traits (future specific growth rates, SGR_future). SGR_future were calculated using the BM at 4, 8, and 12 weeks after the muscle biopsy as the final BM and the BM at the muscle biopsy as the initial BM. In European sea bass, most of the somatic growth occur within the first 3 to 5 years of life, so several months of growth measurement at the juvenile stage might be representative of the overall growth of the animal.

Muscle biopsy procedure

Muscle biopsy was performed as a non-lethal means of sampling tissue for the mitochondrial assay while allowing us to determine future growth rate. Fish were anaesthetized with tricaine (as above), weighed (76.7 ± 3.6 g), and biopsied. A skin incision (< 10 mm in length) was made with a scalpel below the lateral line and between the dorsal and caudal fins. Then, a core of the red muscle was collected using a biopsy punch (2 mm LCH-PUK-20, Kai Medical, Solingen, Germany). The core of red muscle was immediately cleaned of white muscle with a scalpel. The red muscle sample was then weighed (5.6 ± 0.2 mg; AC210P-0F1, Sartorius, Göttingen, Germany), and transferred to ice-cold respiration buffer (20 mmol L⁻¹ Taurine, 10 mmol L⁻¹ KH₂PO₄, 20 mmol L⁻¹ HEPES, 110 mmol L⁻¹ D-sucrose, 60 mmol L⁻¹ K-lactobionate, 1 g L⁻¹ BSA fatty acid free, and pH 7.0 at 13 °C). The incision was disinfected (Vétédine®Solution, Vetoquinol, Magny-Vernois, France) and filled with powdered bandage (ORAHESIVE, ConvaTec®, Deeside, UK). The fish were placed in a recovery tank before returning to their original tank.

The newly developed sampling procedure for mitochondrial assay required several preliminary works. A pilot experiment on a different set of fish evaluated the immunological consequences of biopsy on inflammation resulting from bacterial infection or damaged tissue. There was no significant effect of the biopsy on the immune system (see the supplementary materials and methods for details of all assay protocols and results Fig. S1b). This experiment also demonstrated that biopsy slightly reduced the growth rate but fish continued to gain mass (Fig. S1a). Finally, fish needed anesthesia for the biopsy procedure, and tricaine is the most judicious to study mitochondria in fish muscle. Tricaine acts by blocking voltage-gated sodium channels and blocks neural action potentials²³. Muscle sodium channels are relatively insensitive to tricaine²³, which suggest that tricaine are unlikely to affect the physiology of muscle cells. However, it would be relevant to test for potential effect of tricaine on mitochondrial function in fish muscle.

Tissue homogenate for mitochondrial assay

Immediately after biopsy, muscle samples were shredded using micro-dissecting scissors to obtain a homogenous solution with a particle size less than 0.5 mm (tested by pipetting through a 1 mL tip), an homogenization procedure adapted from²⁴. The tissue preparation by shredding used here allowed a rapid and efficient muscle permeabilization, with no loss of tissue, while maintaining the quality of the mitochondria. Validations of the permeabilization method are described in^11,25. The homogenates were diluted further in respirometry buffer to obtain a final concentration of 1 mg mL⁻¹ (mean ± s.e = 1.0 ± 0.2 mg mL⁻¹). Red muscle homogenizations were carried out on ice. Only four fish could be run simultaneously for the mitochondrial assay (one measurement per fish), and two runs per day were performed; therefore, five days were required to analyze the forty fish.

Measurement of mitochondrial metabolic traits

Oxygen consumption and magnesium green fluorescence were detected simultaneously in four respirometry chambers (2 mL, a chamber per fish) equipped with fluorescent sensors and recorded using DatLab software (Oroboros Instruments, Innsbruck, Austria) at 13 °C with continuous stirring. Immediately after homogenisation and dilution, 2.2 mL of the homogenate was added to one of the four chambers. Mitochondrial metabolic traits were determined using the method described by Chinopoulos, et al.²⁶. First, an adenylate kinase inhibitor (P1,P5-Di(adenosine-5′) pentaphosphate pentasodium salt, 0.1 mmol L⁻¹), complex I-linked substrates (pyruvate 5 mmol L⁻¹ and malate 0.5 mmol L⁻¹), and MgGreen (2.2 µmol L⁻¹) were added to each well. After addition of EGTA (0.1 mmol L⁻¹) and EDTA (15 µmol L⁻¹), the fluorescent signal of MgGreen was calibrated using ten successive injections of MgCl₂ (1 mmol L⁻¹). Succinate (10 mmol L⁻¹) was added to supply complex II with the energy substrates. OXPHOS respiration and ATP production were measured by adding ADP (2 mmol L⁻¹). Leak respiration was measured by inhibiting ATP synthesis by adding carboxyatractyloside (4 µmol L⁻¹). The rate of ATP disappearance owing to ATPase activity was also measured under these conditions. This measurement was subtracted from ATP production measured during OXPHOS respiration. An inhibitor of coenzyme Q-cytochrome c reductase (CIII), antimycin A (2.5 µmol L⁻¹), was then added. This measure was subtracted from other mitochondrial respiratory rates to correct for oxygen consumption unrelated to the respiratory chain activity. Finally, mitochondrial density was estimated with COX activity by adding ascorbate (8 mmol L⁻¹) followed by N,N,N,N-tetramethyl-p-phenylenediamine (TMPD; 0.5 mmol L⁻¹) to the chambers. The rate of ATP production and ATP/O ratio were calculated as described in Salin, et al.²⁷. Briefly, we converted the fluorescent signal of the free magnesium concentration to the ATP concentration. We used the same binding affinity (K_d) for Mg²⁺ bound to ATP and ADP as in Thoral, et al.²⁸ because the mitochondrial analyses were performed under the same conditions of temperature and homogenate concentration (K_d-ATP = 0.266 mmol L⁻¹, K_d-ADP = 1.803 mmol L⁻¹). The ATP/O ratio was calculated as the ratio of ATP production to OXPHOS respiration. Rates of oxygen consumption and ATP production were expressed as pmol s^− 1 mg^− 1 wet tissue.

In an additional experiment, we determined the technical repeatability of mitochondrial metabolic traits. Duplicated measurements of mitochondrial metabolic traits of two biopsies from the same individual were significantly reproducible (Intraclass correlation coefficients: OXPHOS respiration r = 0.41, p < 0.001; LEAK respiration r = 0.48, p < 0.001; COX activity: r = 0.40, p < 0.001; n = 19; ATP production r = 0.44, p = 0.001; n = 14).

Statistical analysis

Intra-class correlation (ICC) was used to test the consistent differences in growth rates among individuals before and after the mitochondrial assay. Individual consistency of growth rate was tested between the most similar duration of the growth trial, that is, SGR_past at 7 and 11 weeks was correlated to SGR_future at 8 and 12 weeks, respectively. Linear mixed models (LMMs) were used to determine the relationships between fish growth rate and mitochondrial metabolic traits. Each SGR (SGR_past over 7, 11, 16, and 20 weeks, and SGR_future over 4, 8, and 12 weeks) was run in separate models. The models included SGR as the dependent variable, OXPHOS respiration, ATP production, LEAK respiration, and COX activity as continuous predictors, with the fish tank as a random factor. To control for the effect of initial body mass on growth rate, initial body mass was included as a covariate in each model. The day and the processing run of the mitochondrial assay, as well as the Oroboros device, and the respirometry chamber used for the analysis of mitochondrial assay were initially included as potential random factors in each model but were not significant and were subsequently removed. We also tested whether the degree of mitochondrial efficiency in generating ATP, as determined by the ATP/O ratio, explained individual variation in growth, which had to be examined in a separate linear model to prevent problems associated with multicollinearity between ATP/O, OXPHOS respiration, and ATP production.

Since mitochondrial density can influence mass-specific rates of mitochondrial metabolic traits (Pearson correlations: all r > 0.37 and p < 0.05) and SGR. Estimate of COX-independent mitochondrial traits were used in preliminary LMM analyses of SGR. Residuals from separate linear regressions relating mitochondrial metabolic traits (e.g. OXPHOS and LEAK respiration and ATP production) to COX activity were calculated. As before, we used LMM with SGR as the dependent variable and COX-independent OXPHOS and LEAK respiration, COX-independent ATP production, and COX activity as continuous predictors. Since the patterns of the results of growth rate analyses were the same whether mitochondrial metabolic traits were measured in terms of mass-specific or COX-independent tissues, only those for mass-specific mitochondrial metabolic traits are reported here.

All analyses were based on a sample size of 40 fish and all statistical analyses were performed using R (4.0.3; package lme4; package partR2). Data are presented as means ± SEM, the significance level was set to p < 0.05, and inclusive R² (IR²) was added to illustrate the variance explained by a predictor in a model²⁹.

Source: Ecology - nature.com