Muscle remodeling in relation to blood supply: implications for seasonal changes in mitochondrial enzymes

We investigated if seasonal changes in rainbow trout muscle energetics arise in response to seasonal changes in erythrocyte properties. We assessed if skeletal muscle mitochondrial enzymes changed (1) acutely in response to changes in erythrocyte abundance, or (2) seasonally when we altered the age profile of erythrocytes. Rainbow trout were treated with pheynylhydrazine,causing a 75% reduction in hematocrit within 4 days. After erythropoiesis had returned hematocrit to normal, treated and control fish were subjected to a seasonal cold acclimation regime to assess the impact of erythrocyte age on skeletal muscle remodeling. Anemia (i.e. phenylhydrazine treatment) did not alter the specific activities (U g^-1 tissue) of mitochondrial enzymes in white or red muscle. Anemic pretreatment did not alter the normal pattern of cold-induced mitochondrial proliferation in skeletal muscle,suggesting erythrocyte age was not an important influence on seasonal remodeling of muscle. Anemia and cold acclimation both induced a 25–30%increase in relative ventricular mass. The increase in relative ventricular mass with phenylhydrazine treatment was accompanied by a 35% increase in DNA content (mg DNA per ventricle), suggesting an increase in number of cells. In contrast, the increase in ventricular mass with cold temperature acclimation occurred without a change in DNA content (mg DNA per ventricle), suggesting an increase in cell size. Despite the major increases in relative ventricular mass, neither anemia nor seasonal acclimation had a major influence on the specific activities of a suite of mitochondrial enzymes in heart. Collectively, these studies argue against a role for erythrocyte dynamics in seasonal adaptive remodeling of skeletal muscle energetics.

Many different physiological and environmental stimuli induce mitochondrial proliferation in vertebrate muscles (see Hood, 2001; Moyes and Hood, 2003). Regulation of mitochondrial biogenesis may be mediated by both intrinsic(intracellular) and extrinsic (neuro-hormonal) regulation. Increases in mitochondrial content frequently accompany hypermetabolic challenges but direct links between bioenergetics and bioenergetic gene expression remain elusive. Reactive oxygen species (ROS) may act as a regulatory surrogate of energy metabolism under some conditions (see Leary and Moyes, 2000). Energetic limitations arising from pathological defects in OXPHOS complexes increase ROS production (e.g. Turner and Shapira, 2001). Exercise programs that trigger mitochondrial proliferation also appear to induce ROS production (see Moyes and Hood, 2003). It is important to recognize that the increase in ROS cannot be explained simply by increased mitochondrial respiration, as this in itself reduces the production of superoxide by mitochondria (Korshunov et al., 1997). It is likely that elevated ROS production is due at least in part to oxygen limitations(Pearlstein et al., 2002). While no direct links between intracellular ROS production and mitochondrial gene expression have been established, many of the ROS-sensitive transcription factors (e.g. AP-1, NFκB) can regulate mitochondrial genes under some conditions (see Leary and Moyes,2000; Scarpulla,2002; Jackson et al.,2002) and ROS production can vary under conditions that alter mitochondria biogenesis, such as exercise (see Moyes and Hood, 2003).

One model that has been used to explore the mitochondrial response to environmental stress is cold acclimation in fish. Depending upon the species and fiber type, muscle mitochondrial enzyme activities can more than double(e.g. Johnston and Maitland,1980; Johnston,1982; Egginton and Sidell,1989; Battersby and Moyes,1998). Paradoxically, the increase in mitochondrial content coincides with a decrease in absolute metabolic rate due to reduced temperature. In salmonids, skeletal muscle mitochondrial enzyme specific activity increases to the same extent in exercise training(Farrell et al., 1991) and cold acclimation (Battersby and Moyes,1998). Cold acclimation is usually also accompanied by an increase in capillarity (Egginton and Cordiner,1997), frequently in parallel with changes in mitochondrial content (Johnston, 1982). The genetic basis of remodeling of striated muscle energetics with cold acclimation, both cardiac and skeletal, remains largely unexplored.

Since cold acclimation also induces an increase in relative ventricular mass (Graham and Farrell,1990; Taylor et al.,1996) we considered the possibility that each aspect of cold-induced muscle remodeling (cardiac hypertrophy, skeletal muscle angiogenesis, mitochondrial proliferation) could be attributed to changes in hemodynamics, such as the ability of erythrocytes to penetrate the peripheral vasculature. As water temperatures cool in the Fall, erythrocyte properties change in ways that could influence perfusion. First, cooling an erythrocyte,or any cell, causes the cell membrane to become more rigid. This reduces erythrocyte deformability and, as a consequence, makes it more difficult for the cell to penetrate the peripheral vasculature(Hughes et al., 1982; Kikuchi et al., 1982). Second,erythrocyte perfusion may also be influenced by cell age(Linderkamp and Meiselman,1982). Many temperate fish experience a burst of erythropoiesis in Spring and by the time Fall cooling begins, most of the erythrocytes are approaching the end of their lifespan (see Nikinmaa, 1990). The cell membranes of old erythrocytes are more rigid due to lipid damage and aggregation of membrane-associated protein. Consequently, the onset of Fall cooling may reduce the capacity of the erythrocytes to penetrate the muscle vasculature. This could explain the stimulation of angiogenesis, a response that is often linked to regional hypoxia(Maxwell and Ratcliffe, 2002). While there is no evidence that mitochondrial gene expression is directly sensitive to oxygen levels, erythrocytes have important antioxidant roles and may be an important element of peripheral antioxidant defense by metabolizing ROS (Gabbianelli et al., 1998; Aoshiba et al., 1999; Fedeli et al., 2001). While the antioxidant capacities of erythrocytes do not deteriorate with cell age(e.g. Moyes et al., 2002),reduced penetration of the vascular beds could impair erythrocyte-dependent antioxidant capacities. Thus, seasonal changes in erythrocyte properties could contribute to the remodeling of both the vasculature and energetics in skeletal muscle.

In the present study we examined the impact of erythrocyte dynamics on muscle mitochondrial biogenesis. First, we induced an anemic state to reduce the number of erythrocytes. We assume that this would create a situation where fewer erythrocytes passed through the muscle vasculature. Second, we assessed if the age profile of erythrocytes could influence the effects of seasonal cooling on mitochondrial enzyme changes. Animals made anemic were able to replenish their erythrocyte compliment over several weeks at constant temperature. By the time Fall cooling began, their hematocrit had returned to normal levels, but the cells were largely young cells. Collectively, these studies assessed the impact of perfusion on muscle mitochondrial biogenesis.

Rainbow trout (Oncorhynchus mykiss Walbaum) of undetermined sex were obtained from Pure Springs Trout Farm (Belleville, Ontario) were held in flow-through tanks. Fish were fed standard trout chow to satiety five times per week. Fish were held under a constant 16 h:8 h light:dark photoperiod, and thus we investigated the effects of thermal acclimation rather than seasonal acclimatization. The photoperiod was chosen to reflect local midsummer conditions. Water temperatures were allowed to vary with season, and monitored continuously. Experiments began in September, at which point water temperatures had ranged from 16°C to 20°C for at least 10 weeks(Fig. 1).

At the onset of experiments, fish averaged about 75 g(±6 g s.e.m.). Fish were made anemic by injection of phenylhydrazine(protocol approved by Queen's University Animal Care Committee), as described by Gilmour and Perry (1996). They were anesthetized in bicarbonate-buffered MS222 (0.4 g NaHCO₃and 0.2 g MS-222 per litre water) and injected with phenylhydrazine (10 μg g^-1). Control fish were anesthetized but not injected. Injections occurred when water temperature was 18°C. The phenylhydrazine treatment had no effect on mortality; over the 25 week period, no fish died in either treated or control group. There was also no significant effect on growth rates in treated and untreated fish at either 1 month (95±7 g vs88±6 g) or 6 months (122±6 g vs 115±8 g)post-treatment.

At the onset of the study, 10 untreated fish were sampled as a pre-treatment group (designated Week 0). Groups of five treated fish were sampled at 1, 2, 4, 8 days post-treatment, and compared with pre-treatment fish. For subsequent time points (weeks to 6 months) five fish were sampled from both control and phenylhydrazine-treated groups. Fish were anesthetized in MS222, blood samples were collected, then fish were decapitated and tissues sampled. Cardiac ventricle mass was measured in relation to body weight,giving relative ventricular mass. Tissues (red muscle, white muscle, heart)were flash frozen, powdered in liquid nitrogen, and stored at–80°C.

Powdered tissue (50–100 mg) was weighed and homogenized in 20 volumes of extraction buffer consisting of 20 mmol l^-1 Hepes (pH 7.0), 1 mmol l^-1 EDTA, and 0.1% Triton X-100, using a ground glass tissue homogenizer. Enzyme activities were assayed using a Molecular Devices Spectramax 250 spectrophotometer at 25°C at 340 nm unless otherwise noted. After the assays for COX, CPT and HOAD, the homogenates were frozen at–80°C prior to analyses of other enzymes. Chemicals were purchased from Sigma-Aldrich Canada, Oakville, Canada.

The COX assay was performed within 60 min following homogenization. In brief, homogenate was added to a mixture of Tris-HCl (50 mmol l^-1)containing 50 μmol l^-1 reduced cytochrome c. After rapid mixing,the absorbance (550 nm) was followed for up to 90 s. Homogenate volumes were chosen to ensure that the rate of change in absorbance fell within the range of 0.06 to 0.10 absorbance units per minute. Above this rate, the reaction depleted cytochrome c concentrations enough to reduce reaction rates.

The assay contained (in mmol l^-1):5,5′-dithiobis-(2-nitrobenzoic acid) (0.1), acetyl CoA (0.3),oxaloacetate (0.5), in Tris-HCl (50), pH 8.0. The increase in absorbance at 412 nm was measured. A control well lacking oxaloacetate was used to correct for background thiolase activity.

The assay contained (in mmol l^-1): acetoacetylCoA (0.1), NADH(0.15) in imidazole (50) at pH 7.2. The assay was started with enzyme and no NADH oxidation was evident in the absence of acetoacetylCoA.

The assay contained (in mmol l^-1): ADP (5), KCl (100),MgCl₂ (10), NADH (0.15), fructose 1,6 bis-phosphate (0.01),phosphoenolpyruvate (5) and excess lactate dehydrogenase (free of PK) in 50 mmol l^-1 Mops 7.4. The assay was started with enzyme but was strictly dependent upon phosphoenolpyruvate.

The assay contained (in mmol l^-1): pyruvate (1), NADH (0.15) in Hepes (50) at pH 7.0.

The assay contained (in mmol l^-1):5,5′-dithiobis-(2-nitrobenzoic acid (0.1), palmitoyl CoA (0.1) and carnitine (5) in Tris-HCl (50) at pH 8.0. Control wells lacking carnitine were used to correct for background thiolase. Absorbance was monitored at 412 nm. Since freezing inactivates CPT I, it is presumed that the activity measured in the CPT assay is CPT II.

Homogenates were also used to measure the levels of DNA. A small volume of homogenate (50 μl) was added to 5 volumes of proteinase K digestion buffer(10 mmol l^-1 Tris, 100 mmol l^-1 NaCl, 25 mmol l^-1 EDTA, 0.5% SDS, 0.2 mg ml^-1 proteinase K) in the presence of RNase (Battersby and Moyes 1998). After 16 h at 55°C, and without further purification,the DNA concentration was measured using Picogreen (Molecular Probes) and a standard curve constructed using purified trout genomic DNA.

Although tissues were blotted prior to freezing, we did not perfuse the tissues to expel erythrocytes. However, erythrocyte DNA levels in whole blood(∼0.3 mg g^-1 blood; Moyes et al., 2002) are much lower than heart DNA levels (3 mg g^-1 tissue; Leary et al.,1998). Similarly, the levels of mtDNA do not appreciably influence total DNA levels in these tissues. In skeletal muscles, mtDNA is less than 1%of total DNA (Battersby and Moyes,1998). Thus, neither blood contamination nor mtDNA would substantially affect the DNA determinations.

Time courses were analyzed by analysis of variance (ANOVA) followed by a Tukey's test post-hoc. Differences with P<0.05 were considered significant. The analyses compared treated animals to time-matched controls at all sampling points of two weeks post-treatment or longer. However, the samples collected in the short time course (1, 2, 4, 8 days) were compared with a group untreated animals collected at the start of the experiment as designated as week 0, or pre-treatment.

Phenylhydrazine injection caused a rapid anemia, with a reduction in hematocrit of more than 75% by four days post-injection (P<0.001, Fig. 1). By 4 weeks post-injection, hematocrit had returned to normal (∼38%) and water temperatures remained above 15°C. Hematocrit remained between 36 and 40%for the duration of the experiment in both anemic and control groups, with no significant differences between groups at any sampling point beyond 4 weeks(P>0.05).

Relative ventricular mass changed rapidly in response to anemia(Fig. 1B). By 2 weeks post-injection, ventricular mass had increased from 0.085% to 0.11% of body mass (P=0.001). Relative ventricular mass had decreased by 9 weeks post-injection. By the time Winter cooling had occurred, the control fish had experienced enough cardiac growth to match the phenylhydrazine treated fish. At the lowest winter temperatures, phenylhydrazine treated fish and control fish had similar relative ventricular masses of about 0.1% of body mass. Thus,anemic history had no effect on the relative ventricular mass in acclimated fish.

DNA levels were also measured in ventricle to assess the impact of (1)anemia, (2) seasonal acclimation and (3) an anemic history on ventricular remodeling. Although we consider the primary effect of phenylhydrazine to be anemia, other effects are possible (see Discussion). Anemia alone had no effect on the DNA concentration per gram ventricle(Fig. 2B) but the DNA content in the entire ventricle increased 28% (Fig. 2C). Conversely, cold acclimation alone caused a 30% decline in DNA concentration per gram ventricle (Fig. 2B, dark bars), but DNA content per ventricle did not change(Fig. 2C, dark bars). Finally,there was no evidence that an anemic history influenced the effects of cold acclimation. By 25 weeks acclimation, the treated and untreated fish had similar relative ventricular masses (Fig. 2A), DNA concentrations per gram ventricle(Fig. 2B) and DNA contents per ventricle (Fig. 2C).

Ventricular enzyme activities were also assessed in these fish(Fig. 3). In heart, the effects of phenylhydrazine treatment on enzymes must be interpreted with consideration of the effects on relative ventricular mass(Fig. 2). There was no significant effect of anemia on HOAD specific activity (P=0.08). Similarly, the specific activities of COX, CS and CPT did not change with anemia. The relative maintenance of specific activity required active synthesis of enzymes to compensate for the increase in relative ventricular mass.

We investigated the effects of phenylhydrazine treatment on enzymes in white muscle, red muscle and cardiac ventricle. Figs 4 and 5 show the effects of phenylhydrazine treatment on enzymes over 4 weeks, which we interpret to be the impact of anemia. Fig. 6shows the enzyme patterns that occurred in treated and untreated fish as a function of thermal acclimation. Upon acclimation, the phenylhydrazine-treated fish had lived with a normal hematocrit for several months. Thus, the difference between control and phenylhydrazine-treated fish arises from the`anemic history', which we presume to be due to the influence of red blood cell age.

In white muscle (Fig. 4) CS and COX activities were largely unaffected by phenylhydrazine treatment. We also measured the activities of the glycolytic enzymes LDH and PK in white muscle to assess the response of glycolytic genes to the treatment. Neither LDH nor PK showed a significant response to phenylhydrazine treatment.

The effects of phenylhydrazine treatment on red muscle enzymes were less clear (Fig. 5). Individual enzymes appeared to increase or decrease within a range of about 15% in the first week of anemia. None showed systematic changes in activity with anemia.

The anemic history (i.e. a younger erythrocyte population) had little effect on enzymes in skeletal muscle upon cold acclimation(Fig. 6). It did not blunt the increases in COX and CS in red or white muscle. However, an anemic history did appear to enhance the effects on PK and LDH. Specifically, the anemic history caused a reduction in PK levels upon cold acclimation, whereas cold acclimation alone had no significant effect. Cold acclimation alone depressed LDH activity, but the anemic history further depressed LDH specific activity.

Although cold-induced mitochondrial proliferation in fish was first shown more than 20 years ago (e.g. Johnston and Maitland, 1980), its regulatory origins remain enigmatic. Muscle mitochondrial proliferation can be induced several ways, but each can be related directly (e.g. exercise, electrical stimulation) or indirectly(thyroid hormone) to elevated metabolic rates. Thus, mitochondrial proliferation is usually envisioned as a compensatory response to energetic shortfalls. Cold temperatures reduce the metabolic rate of poikilotherms, yet,paradoxically, cold acclimation induces mitochondrial proliferation. Our analyses attempted to distinguish between the effects of seasonal changes in temperature versus erythrocyte dynamics. We chose to maintain a midsummer photoperiod throughout the experiment. Photoperiod could influence muscle energetics, possibly through activity patterns or hormonal signaling pathways linked to seasonal remodeling. For example, continuous light can alter muscle growth properties (Johnston et al.,2003), but to the best of our knowledge, the influence of photoperiod on seasonal remodeling of muscle energetics has not been addressed.

Many factors influence the ability of erythrocytes to penetrate the peripheral vasculature. Erythrocyte content of the blood can influence oxygen delivery to the periphery by changing O₂ content of the blood. Direct manipulation of erythrocyte O₂ content can compromise

max,decreasing it at low hematocrit and enhancing it at high hematocrit in fish(Gallaugher et al., 1995) and mammals (Davies et al., 1982). Viscoelastic properties of the erythrocyte also influence penetration into the capillaries. Membrane elasticity, as well as cytoskeletal rigidity, influences the deformability of the erythrocyte. When erythrocytes are cold, the reduced fluidity of the lipid bilayer reduces membrane elasticity, altering deformability and the ability to penetrate capillary beds(Hughes et al., 1982; Kikuchi et al., 1982). Cellular aging also increases membrane rigidity, reducing deformability(Linderkamp and Meiselman,1982). Since Winter-acclimatized fish would experience both of these changes (aging and temperature), we examined if alterations in erythrocyte properties could induce the change in muscle mitochondrial content. Neither of our approaches to altering peripheral oxygen delivery caused changes in the activity of mitochondrial enzymes.

Anemia, induced by phenylhydrazine, failed to alter the levels of any mitochondrial enzyme in either white muscle(Fig. 4) or red muscle(Fig. 5). Pre-acclimation anemia, with the resultant rejuvenation of erythrocyte population, addressed the impact of erythrocyte age on seasonal changes in muscle energetics. Experimental temperatures cooled long after the fish had recovered from the anemia. Cytochrome oxidase increased about 50% in white muscle and about 30%in red muscle (Fig. 6), in line with previous studies on cold acclimation(Battersby and Moyes, 1998). The treated fish, with anemic history, exhibited the same pattern of cold-induced changes in mitochondrial enzymes.

Cold acclimation in trout can cause an increased reliance on lipids for muscle during exercise (Kieffer et al.,1998). In this study, analysis of red muscle revealed no significant changes in the activities of fatty acid oxidizing enzymes (CPT and HOAD) as a result of cold acclimation or anemic history.

Both anemia and cold-acclimation induced ventricular remodeling, with important differences. Norman and McBroom(1958) found that phenylhydrazine treatment induced cardiac growth in rats. The myocardial changes associated with phenylhydrazine treatment likely arise from anemia and the resultant cardiovascular effects, however, it is important to recognize other potential effects of phenylhydrazine treatment. Meerson and Evsevieva(1985) found that the effects of phenylhydrazine on cardiac growth could be largely prevented by co-treatment with an antioxidant, suggesting that the phenylhydrazine might be acting through direct effects on the heart. Nonetheless, it is clear that the cardiac remodeling with phenylhydrazine treatment differed from that seen with cold acclimation, despite similar effects on relative ventricular mass.

Previous studies in salmonids and other species have shown that relative ventricular mass increases by about 30% in cold acclimated fish(Graham and Farrell, 1990; Taylor et al., 1996; Farrell et al., 1988). Our sampling protocol also allowed us to follow the time course of change. Relative ventricular mass remained constant as temperature dropped from 20°C to 10°C. However, within 2 weeks of temperatures reaching the seasonal low (1.8°C), relative ventricular mass had increased about 30%. A similar degree of ventricular growth was seen in response to anemia(Fig. 2). Cold temperature is likely to be accompanied by increased peripheral resistance(Taylor et al., 1996; Farrell et al., 1988),possibly due to elevated plasma viscosity and reduced erythrocyte deformability (see Egginton,2002). In contrast, ventricular remodeling in anemic fish probably occurred as a response to a chronic increase in cardiac output to maintain oxygen delivery to the periphery.

Although the increase in relative ventricular mass with anemia was similar to the change with cold acclimation, cellular changes in DNA content imply different mechanisms of ventricular growth. While the exact nature of cold-induced changes in relative ventricular mass is unclear, the analyses of DNA content are consistent with hypertrophy; relative ventricular mass increased but the total DNA content of the ventricle was unchanged. The most parsimonious explanation is that increases in relative ventricular mass arise from increases in cardiomyocyte size. In contrast to the situation with cold-acclimation, the increase in relative ventricular mass that accompanied anemia was accompanied by an increase in DNA content. While these observations are consistent with cardiomyocyte hyperplasia, it is important to acknowledge the other potential explanations for increase DNA content per ventricle. Cardiomyocytes comprise the bulk of ventricular mass, but they are not the most abundant type of cell within the heart. Fibroblasts, immune cells,vascular smooth muscle, endothelial cells and other cell types collectively account for more than 70% of the cells in a vertebrate heart (e.g. Alder et al., 1996). Thus, proliferation of non-cardiomyocytes could account for the increased DNA content of ventricle. Another potential explanation is cardiomyocyte polyploidy. While most cardiomyocytes are generally thought to be terminally differentiatied, mononucleated cells(Zak, 1974), under some conditions the cardiomyocytes can re-enter karyokinesis and become polyploid(Kellerman et al., 1992). While we cannot discount these potential explanations for the increase in both ventricular DNA content and ventricular growth, recent studies have established that cardiomyocyte hyperplasia contributes to ventricular growth under disease conditions. These cardiomyocyte precursor cells may be pre-cardiomyocytes that proliferate and differentiate, or a pool of circulating stem cells that invade cardiac tissue(Anversa and Kajstura, 1998; Nadal-Ginard et al., 2003). Based on estimates of trout cardiomyocyte volume, Clark and Rodnick(1998) concluded that hyperplasia was also necessary to account for the increase in relative ventricle mass seen during fish growth.

The patterns seen in ventricle enzymes, in relation to mass and DNA can be interpreted in terms of both causes and consequences. The cause of the change in total enzyme content of the ventricle (specific activity × mass)reflects a genetic event that culminates in a change in synthesis or degradation of enzymes. Thus, the 40–50% increase in total activity of COX and CS upon cold acclimation (Fig. 3) reflects a stimulation of mitochondrial proliferation.

The increase in mitochondrial enzyme content that accompanies the increase in ventricular mass served to maintain nearly constant the enzyme specific activity. Across each treatment (i.e. anemia, anemic history, temperature acclimation) the specific activities of each enzyme varied less than 15%. The specific activity of CPT, an index of the propensity for fatty acid oxidation,was relatively unaffected by anemia (treated vs untreated at 4 weeks), cold acclimation (untreated at 4 vs 25 weeks), or anemic history (treated vs untreated at 25 weeks). The same was true for HOAD, another index of fatty acid oxidation(Fig. 3). Changes in indices of fatty acid oxidation often accompany cardiac hypertrophy. For example, HOAD specific activities increase in the cardiac hypertrophy that accompanies sexual maturation of trout (Clark and Rodnick, 1998). In mammalian models, hypertensive hypertrophy is accompanied by a change in fuel preference away from fatty acids and toward carbohydrates (Barger and Kelly,1999).

In summary, we found little evidence that our treatments, which targeted erythrocyte properties (hematocrit, age profile), affected the mitochondrial enzyme profile in skeletal muscle. Our manipulation of erythrocyte age profile did not alter the seasonal response of skeletal muscle enzymes. We cannot rule of the effects of temperature-dependent effects on erythrocyte deformability,but even severe anemia failed to alter mitochondrial biogenesis. While the effects of anemia on COX activity were significant, the rapidity of the response suggests a post-translational route of enzyme activation, rather than induced expression. It is possible that elevating activity levels of the fish could exacerbate oxygen delivery limitations, and enhance the seasonal compensatory response.

We would like to thank Tony Farrell for his helpful comments on the manuscript. These studies were supported by an NSERC Discovery Grant, with the assistance of personnel support from the Ontario Ministry of Science and Technology through a Premier's Research Excellence Award (C.D.M.).