Pre-hibernation diet alters skeletal muscle relaxation kinetics, but not force development in torpid arctic ground squirrels

Abstract

During the hibernation season, Arctic ground squirrels (AGS) experience extreme temperature fluctuations (body temperature, T_b, as low as − 3 °C), during which they are mostly physically inactive. Once T_b reaches ~ 15 °C during interbout arousals, hibernators recruit skeletal muscle (SkM) for shivering thermogenesis to reach T_b of ~ 35 °C. Polyunsaturated fatty acids (PUFA) in the diet are known to influence SkM function and metabolism. Recent studies in the cardiac muscle of hibernators have revealed that increased levels of ω-6 and the ω-6:ω−3 PUFA ratio correlate with sarco/endoplasmic reticulum calcium ATPase (SERCA) activity and hibernation status. We hypothesized that diet (increased ω-6:ω-3 PUFA ratio) and torpor status are important in the regulation of the SERCA pump and that this may improve SkM performance during hibernation. Ex vivo functional assays were used to characterize performance changes in SkM (diaphragm) from AGS fed the following diets. (1) Standard rodent chow with an ω-6:ω-3 ratio of 5:1, or (2) a balanced diet with an ω-6:ω-3 ratio of 1:1 that roughly mimics wild diet. We collected diaphragms at three different stages of hibernation (early torpor, late torpor, and arousal) and evaluated muscle function under hypothermic temperature stress at 4 °C, 15 °C, 25 °C, and 37 °C to determine functional resilience. Our data show that torpid animals fed standard rodent chow have faster SkM relaxation when compared to the balanced diet animals. Furthermore, we discovered that standard rodent chow AGS during torpor has higher SkM relaxation kinetics, but this effect of torpor is eliminated in balanced diet AGS. Interestingly, neither diet nor torpor influenced the rate of force development (rate of calcium release). This is the first study to show that increasing the dietary ω-6:ω-3 PUFA ratio improves skeletal muscle performance during decreased temperatures in a hibernating animal. This evidence supports the interpretation that diet can change some functional properties of the SkM, presumably through membrane lipid composition, ambient temperature, and torpor interaction, with an impact on SkM performance.

Graphical abstract

Introduction

Muscle strength decreases during hypothermia (Rantala and Chaillou 2019), but the mechanisms by which hypothermia affects muscle strength are not well known. Hibernating animals, such as the AGS, experience natural variation in body temperature (T_b; during hibernation, T_b can drop as low as − 3 °C (Barnes 1989), while in summer (during their active state), T_b can reach up to 40 °C (Long et al. 2005). Hibernators improve their chances of surviving the winter by lowering their metabolic rate and T_b, with concomitant decreases in respiratory rate and heart rate. Metabolic changes are critical for the torpid hibernator to have functional skeletal (respiratory, diaphragm) and cardiac muscles. Although respiratory and cardiac functions are reduced during torpor, efficient muscle contraction and relaxation mechanisms are still required. AGS and other hibernating animals have less muscle loss, muscular weakening, and protein loss than non-hibernating animals despite sustained periods of immobility and extreme hypothermia during hibernation (Gao et al. 2012; Bodine 2013; Nowell et al. 2011). As a result, small hibernators like AGS are a promising model for investigating hypothermia’s impact on muscle function (Bertile et al. 2021). Contractile performance is the primary metric used to assess muscle function, and it has received little attention during hypothermia, particularly during hibernation (Cotton 2016). There are several adaptations specific to mammalian hibernators that may explain skeletal muscle’s capacity to operate at hypothermic temperatures during and after hibernation (Tessier and Storey 2016).

Calcium homeostasis is critical to contractile performance and muscle function and calcium overload in skeletal muscle (SkM) can induce muscle contractile dysfunction, atrophy, and mitochondrial apoptosis, among other processes (Hyatt and Powers 2020). For example, excessive Ca²⁺ buildup in cardiac cells can result in arrhythmias and even death (Frommeyer et al. 2012). Muscle movement and shivering involve contraction and relaxation mediated by Ca²⁺ cycling from sarcoplasmic reticulum (SR) stores. Ryanodine receptors (RyR1) on SR membranes release Ca²⁺ into the cytosol, facilitating muscle contraction via actin–myosin cross-bridge formation. SERCA then catalyzes the transport of Ca²⁺ from the cytosol back into the SR utilizing ATP, completing the Ca²⁺ cycle and eliciting muscle relaxation. Sarcolipin (SLN), and phospholamban are short peptides in the SR membrane. These peptides regulate Ca²⁺ transport by inhibiting Ca²⁺ binding to SERCA but still allow hydrolysis of ATP to occur, resulting in wasted energy release in the form of non-shivering thermogenesis (Bal et al. 2012; Oliver et al. 2018). The function and purpose of these peptides are still not completely understood but suggest a role in changing the efficiency of calcium transport (Oliver et al. 2018). Brauch et al. (2005) discovered elevated SERCA 2a gene expression in hibernating ground squirrel cardiac muscle, implying that improved calcium clearance preserved cardiac tissue and highlighting the potential importance of this ATPase in hibernating muscle.

Dietary fatty acids, such as PUFA, have been shown to influence SkM function and metabolism in non-hibernating models (Jeromson et al. 2015), SERCA activity in hibernating cardiac tissue (Giroud et al. 2013), and are well documented to influence physiological temperatures and depth of torpor cycles in hibernators (Giroud et al. 2013; Giroud et al. 2018; Geiser and Kenagy 1987; Ruf and Arnold 2008; Rice et al. 2021; Florant et al. 1993; Frank et al. 2004). The physiological effects of dietary PUFAs on hibernation vary depending on the type and relative amounts of fatty acid present. ω-6 PUFA levels rise before hibernation and can positively influence the depth of torpor temperatures and increase duration of torpor in heterotherms (Giroud et al. 2013; Giroud et al. 2018; Geiser and Kenagy 1987; Ruf and Arnold 2008). High levels of dietary ω-3, on the other hand, increase torpid T_b and brown adipose tissue in AGS (Rice et al. 2021), and may (Giroud et al. 2013, 2018; Ruf and Arnold 2008; Florant et al. 1993) or may not (Rice et al. 2021; Frank et al. 2004) have an adverse effect on torpor and hibernation in other species. While it has been shown, the ratio of dietary ω-6:ω-3 can modify SERCA activity in hibernating cardiac tissue, little is known about how dietary ω-6:ω-3 ratios affect skeletal muscle function, calcium homeostasis, or SERCA expression at the different physiological temperatures hibernating ground squirrels experience.

We focused our analysis in this study on the physiological effects of torpor, diet, and temperature (as observed during hibernation) on SkM performance and select calcium handling protein (CHPs) expression. This study specifically looked at the effects of two different dietary PUFA ratios on SkM contractile function during hypothermic temperature using an ex vivo muscle tissue preparation in diaphragm (DIA) muscle sampled from AGS at three different stages of hibernation—early torpor, late torpor, and arousal. We further hypothesized that when compared to AGS-fed standard rodent chow (control diet, CD), the test diet (TD, based on wild diet) AGS would have slower relaxation kinetics. The TD referred to as a balanced diet previously (Rice et al. 2021; Mikes et al. 2022) was designed to mimic the free-ranging diet of AGS. We assessed SkM contractile function by measuring the rates of force development and relaxation, as well as the maximum force production at predefined time intervals and bath temperatures. A secondary goal was to see if diet and torpor bout length influenced SERCA expression. To the best of our knowledge, this is the first study to look at the contractile function of SkM in hibernators at different stages of hibernation in animals fed diets with altered ω-6:ω-3 fatty acid ratios.

Materials and methods

Animals

All animal experiments were carried out in accordance with the Guide for the Care and Use of Laboratory Animals, 8th edition (Guide for the Care and Use of Laboratory Animals 2011), and were approved by the University of Alaska Fairbanks animal care and use committee (IACUC #904180), Fairbanks, Alaska. AGS were caught in the wild as juveniles in the summer of 2018. In the diet studies conducted in December 2018, 11 female and 12 male AGS were used.

Trapping

Early in July, juvenile AGS (under 1 year old) were captured in the northern foothills of the Brooks Range, Alaska, 40 miles south of Toolik Field Station (68° 38 N, 149° 38 W; elevation 809 m) and transported to the University of Alaska Fairbanks, Fairbanks, Alaska, under an Alaska Fish and Game permit.

PUFA diet

Mazuri rodent chow (PMI Nutrition International, Rich-mond, IN, #5663-Mazuri^®) CD (control diet) contains ω-6:ω-3 in a 5:1 ratio, whereas the test diet (TD) (9GU5-LabDiet^® 5008, St. Louis, MO) contains ω-6:ω-3 in a 1:1 ratio. Further information on fatty acid ratios can be found in supplemental Table 1 as well as Rice et al. (2021).

Husbandry

AGS were housed in cages at an ambient temperature of 18–20 °C and a 16L:8D hour light/dark cycle until mid-August when they were transferred to cold chambers at 2 °C and a 4L:20D cycle. During the captive, euthermic period (from mid-July until onset of hibernation, typically a 1.5–2 month period (Rice et al. 2021), animals were offered 47 g of either CD or TD and water ad libitum daily. Individual cages (12″ × 19″ × 12″) with cotton nests were used to house the animals. Food and water were removed from the animals once they demonstrated consistent torpor, and they were placed in polycarbonate cages (8.5″ × 17″ × 8.5″) with wood shavings, cotton bedding, and gel hydration packets.

Adult male Sprague–Dawley (SD) rats (250 ± 50 g), n = 30, were purchased from the in-house breeding colony maintained by the Animal Resource Center, University of Alaska Fairbanks (Fairbanks, Alaska). The Sprague Dawley rat colony was derived from Simonsen laboratory (Gilroy, CA). Animals were housed in pairs in polycarbonate cages (8.5″ × 17″ × 8.5″) in an environmentally controlled room (20–23 °C, ~ 44% humidity, 12L:12 D cycle, 280–300 lx, lights on at 0600 h). Standard rodent chow and water were provided ad libitum. Rat diaphragm was used as a positive control for the initial phase of the study.

Tissue collection

Once animals entered hibernation and until tissue collection, torpor length and arousal periods were monitored daily and recorded in a hibernation logbook based on physical observations of the animals using a “shavings added” method, as previously described (Rice et al. 2021). In brief, on the first day of torpor when breath rate was below 5 breaths per minute, they were inactive and in a curled posture, shavings were lightly placed on the animal’s backs. Animals were checked daily and when the shavings were disturbed, an arousal period was recorded. Torpor and arousal periods were verified by abdominal i-button temperature data after tissue collection (Rice et al. 2021).

Tissues were collected mid-December at times corresponding to early torpor, late torpor, and arousal as previously described (Rice et al. 2021). Actual bout length of each animal was used to define early torpor (ET) as 10–25% of the average of the previous two torpor bouts (n = 7; 3 females and 4 males), late torpor (LT, defined as 89–100% of torpor bout based on the average of the previous two torpor bouts, n = 8; 4 females and 4 males), and arousal (AR, T_b > 33 °C, n = 8; 4 females and 4 males) (Fig. 1). Arousal was induced by handling between 0730 and 0830 h for tissue collection from the AR group, and tissues were collected at 1300 h. Animals were left in the cold room during the arousal process. Between 0900 and 1100 h, all hibernating animals (ET and LT group) were euthanized, in accordance with AVMA guidelines. Torpid animals were not aroused for tissue collection. All animal rectal temperatures were taken directly prior to euthanization, verifying arousal animals were above 33 °C and torpid animals were below 5 °C. Temperatures were also verified post-tissue collection via abdominal i-button temperature recordings as previously described (Rice et al. 2021). Prior to tissue collection, euthermic AGS were deeply anesthetized with isoflurane (5% mixed with medical-grade oxygen at a flow rate of 1.5 L/min) to a surgical plane, confirmed with toe-pinches (Rice et al. 2021). In less than a minute after the AGS were decapitated, the diaphragm (DIA) was collected and immersed in preoxygenated Ringer’s solution.

Fig. 1.

Diet and hibernation schematics. Arctic ground squirrel (AGS) hibernation cycle: The black solid line represents the core body temperature as measured by a surgically implanted transmitter (described in Laughlin et al. 2018). The air temperature in the environmental chambers is indicated by the blue solid line (hibernaculum). The highlighted area depicts an individual torpor bout (~ 21 days); time points indicated by a red asterisk represent tissue collections for functional assay/western blotting

Preparation of skeletal muscles and assessment of functional performance

Oliver et al. (2008) provided the experimental methods for the muscle contractile functional assay, which were carried out in a tissue organ bath system (TOBS, Radnoti, LLC., CA). Each animal’s DIA was used to create four DIA strips (5–6 mm wide) with the central tendon and rib intact. The central tendon was attached to a plastic frame with a small amount of cyanoacrylate gel adhesive to a force transducer (#1030, UFI, Moro Bay, CA) via 3–0 silk suture. A suture loop was used to secure the ribbed end to a glass frame. After that, each strip was placed in a separate 40 mL water-jacketed tissue bath with preoxygenated Ringer’s solution kept at 37 °C. For each muscle strip, the optimal length (L₀) and stimulation voltage resulting in maximum twitch force were determined. For 30 min (initial phase), all strips were exposed to 37 °C. Following 30 min of respective variable temperature exposure (4 °C, 15 °C, 25 °C, or 37 °C), the temperature was then returned to 37 °C for 30 min (recovery phase) (Fig. 2). Ringer’s solution was replaced with a new solution every 30 min. After each 30-min phase, a force–frequency curve (FFC) was created by using twitch force (0.05 Hz) and tetanic contractions of 400-ms trains of supramaximal stimuli of 0.2 ms duration at 20, 30, 60, and 150 Hz, with 1-min rest in between stimulations. For all functional assays, data from force transducers were recorded using LabChart 8 (AD Instruments, CO) with the peak analysis add-on. The DIA strips were named after the temperatures to which the tissues were exposed during the variable temperature phase of 37 °C, 25 °C, 15 °C, and 4 °C. Initial and recovery phases were maintained at 37 °C in all groups.

Fig. 2.

Timeline of the experimental protocol. Diaphragm (DIA) were harvested with intact ribs and strips were created before being placed in a water-jacketed tissue bath with Ringer’s solution (37 °C). All strips were exposed to 37 °C for 30 min before and after each variable temperature (4 °C, 15 °C, 25 °C, or 37 °C). After each 30-min phase, a force–frequency curve (FFC) (0.05 Hz, 20, 30, 60, and 150 Hz) was performed, and Ringer’s solution was replaced with a fresh solution. Except for the control diet (CD) group of early torpor (ET), which had n = 7, all control diet (CD) and test diet (TD) torpor groups had n = 8

Inhibition of SERCA and RyR

We hypothesized that measuring the rate of relaxation would directly correlate with SERCA 1a (re-uptake of Ca²⁺ back to SR) activity. We also measured the force development rate (as an approximate measure of RyR1 activity) to assess any potential differences but recognize that some literature suggests a more tenuous link (Maffiuletti et al. 2016). To validate this correlation, RyR1 and SERCA 1a were inhibited one at a time pharmaco-logically during the ex vivo functional assay. More details on this protocol are provided in the supplemental section. Raw data were analyzed for correlations using the SlopeRise and SlopeFall derivative functions in LabChart 8’s peak analysis add-on. SlopeRise is the line that connects the crossing point from the peak’s baseline (L₀) to the peak’s maximum height, whereas SlopeFall is the line that connects the peak’s maximum height to the peak’s baseline (L₀). We used dantrolene (Villanueva et al. 2018) or MgCl₂ (Smith et al. 2013) to inhibit the RyR channel, and cyclopiazonic acid (CPA) (Du et al. 1994, 1996) to inhibit the SERCA pump (see the supplemental section for dosage and details). With the addition of the drugs, we observed a linear dose-dependent decrease in RFD and rate of relaxation (Supplemental Fig. 1). MgCl₂ inhibitory effects were reversed (normal RFD was observed) when it was replaced with a fresh Ringer’s solution (after two-to-three times rinsing with Ringer’s solution). Whereas the inhibitory effects of both dantrolene and CPA were sustained (irreversible inhibition of contractile properties) even after 2–3x rinsing with fresh Ringer’s solution (data not shown). This proof of concept allowed us to use RFD and rate of relaxation to assess RyR1 activity and SERCA 1a activity.

Western blotting

SDS-PAGE and Western blotting of CHPs such as SLN and SERCA-1a and 2a were performed as previously described (Oliver et al. 2018) using isolation of SR vesicles. Frozen tissues were weighed and adequate (1 mL) of homogenization buffer was added before homogenizing using Tekmar homogenizer, followed by centrifugation 2Kg at 4 °C for 10 min; supernatant was centrifuged 10Kg at 4 °C for 5 min. Used pellets were discarded. The supernatant was transferred to an Eppendorf tube containing 4 mg of KCL, vortexed, and incubated in an orbital shaker for 30 min at 4 °C, Ultracentrifugation was then conducted at 26Kg for 15 min at 4 °C. Enriched protein samples of SLN and SERCA-1a and 2a were extracted and pellets diluted with Radioimmunoprecipitation assay (RIPA) lysis buffer containing 10% SDS, followed by Tricine sample buffer for sarcolipin detection, or Laemmli sample buffer for SERCA detection, to a final concentration of 2 mg/mL (protein quantification using a modified Lowry’s method) prior to electrophoresis and protein blotting. Each well was loaded with 10 μg of sample protein and run in an electrophoresis unit with a muscle control sample (assorted rat SkM) to normalize samples run on multiple gels. To estimate the size of proteins resolved by gel electrophoresis, we used MagicMark XP Western Protein Standard (Invitrogen) as a protein molecular weight marker. The samples were then transferred to 0.2-μm nitrocellulose membranes at 190 V for 3 h. Ponceau staining was used to determine the overall expression of total proteins, and the blot was then washed several times with TBS before blocking. The membrane was then blocked for 2 h with 5% bovine serum albumin (BSA), and then washed with 1% BSA Tris-Buffered Saline (TBS) and 0.1% Tween^® 20 Detergent (TBST) (3×). Primary antibodies were used at recommended concentrations (1:1000 dilution) to detect SERCA1a (Abcam Cat# ab105172; Cambridge, MA.), SERCA2a (Abcam Cat# ab150435; Cambridge, MA), and SLN (# ab1055172, Abcam) and were applied overnight at 4 °C. Membranes were incubated in secondary antibody (1:10,000 dilution), horseradish peroxidase-conjugated goat anti-rabbit (#32460) for 4 h at 4 °C (Invitrogen, Carlsbad, CA). Membranes were then washed in TBS before being developed with the SuperSignal West Pico PLUS Chemiluminescent kit (#34577 Thermo Scientific, Grand Island, NY). To acquire and quantify band intensity, a camera-based digital ChemiDoc imaging system (UVP, Upland, CA) was used with appropriate exposure settings as recommended by the developer system (#34577 Thermo Scientific, Grand Island, NY) to avoid the saturation of bands. Using FIJI imaging software, chemiluminescence blot results were normalized with Ponceau staining data within gels and loading control protein to normalize multi-gel comparison. Data were expressed in normalized density. For all time points, N = 3 for both CD (control diet) and TD (test diet).

Data analysis

During calibration with analytical weights (0.5–200 g), raw data (in mV) were converted to grams. Based on inhibition protocol experiments, we concluded that the slope of the muscle contraction (i.e., RFD) represents the indirect activity of the RyR1 channel-induced Ca²⁺ release, and the relaxation slope (i.e., rate of relaxation) represents the activity of the SERCA pump. All three contractile properties were normalized to specific force to eliminate the possibility that differences in force generation were due to muscle strip size. Specific force is defined as: N/cm² = ((force (g)) × (muscle length (cm) × 1.06) × 0.00981/(muscle weight (g)/1000 (Park et al. 2012). To better understand changes in muscle contraction and relaxation, percent loss in rates of force development or relaxation was calculated for the variable temperature phase and recovery phases relative to the initial phase. The assessment of residual muscle tension was not included in our current study, because it was outside the scope of our investigation.

Statistical analysis

The data are presented as means ± SEM. Before proceeding with the statistical analysis, all data sets (muscle contraction and relaxation, percent loss in rates of force development or relaxation, and western blot relative intensity densitometry data) were tested for normality and found to be deviating from the normal distribution using the Kolmogorov–Smirnov, and Shapiro–Wilk tests. Transformations failed to normalize the data. Because our groups were independent (comprised of different animals), we used the Mann–Whitney U test to assess differences between groups; two groups were evaluated concurrently among the three, and then, a similar process was repeated with the third group in a combination approach across all data sets. The significance level was set at P < 0.05 (two-tailed). A priori power analysis based on data from Oliver et al. (2018) showed that a sample size of n = 7 would yield 95% power to detect a difference between groups at P < 0.05. While both males and females are included in the study, the number of animals available was not sufficient to power the study to detect a sex difference. Sexes were combined for statistical analysis, but results for males and females are provided in supplemental data (Supplemental Fig. 2). GraphPad Prism-8.3.1 (San Diego, CA.) and SPSS/PC+, Version 25 were used for all statistical analyses (SPSS Inc., Chicago, IL, USA).

Results

Contractile properties during progression through a torpor bout

At 37 °C, we assessed RFD, rate of relaxation, and maximal force production to investigate contractile characteristics during torpor development. We produced a line graph for 0.05 Hz, 20 Hz, 30 Hz, 60 Hz, and 150 Hz of RFD, rate of relaxation, and maximal force output for the purpose of visualization of the force–frequency curves (Fig. 3A–E). All parameter analyses were at 150 Hz (Fig. 3B–F), since that was the frequency that produced the most force. RFD and maximum force production did not differ between torpor stages (Fig. 3B, D), within DIA from ET, LT, and AR time points when stimulated at 150 Hz (Table 1). The rate of relaxation of muscle tissue (Fig. 3F) on the other hand was significantly faster during LT compared to the other time points (P ≤ 0.05, LT over ET; P ≤ 0.0001, LT over AR; P ≤ 0.05, ET over AR; n = 3–4). Within the CD group, LT tissues were able to produce more force and achieve faster relaxation. To compare inherent contractile properties, SD rats (Fig. 3B–F) were used as a non-hibernator control and were not fed with TD. Both RFD and maximum force production were significantly different between AGS and SD rats (P ≤ 0.05; n = 23–30, Kruskal–Wallis one-way ANOVA); however, hibernating AGS had a significantly higher relaxation rate than SD rats at all time points (Fig. 3) (P ≤ 0.0001, ET, LT over SD rats; P ≤ 0.01, AR over SD rats; n = 23–30). The LT ex vivo tissues exhibited the fastest rate of relaxation.

Fig. 3.

A high ω-6:ω-3 ratio in the diet increased the relaxation rate during torpor. A Force-frequency relationships obtained after 30 min of exposure to 37 °C. Force is measured in specific force (N/cm²), peak twitch force is 0.05 Hz, and tetanic force is 150 Hz. B Tetanic force at 150 Hz as a bar graph. C Rate of force development (RFD) frequency relationships obtained after 30 min of exposure to 37 °C. D RFD in bar graph format at 150 Hz. E Relaxation rate frequency relationships obtained after 30 min of 37 °C exposure. F Bar graph representation of the relaxation rate at 150 Hz. The data presented are means ± SEM. Except for the control diet (CD) group of early torpor (ET), which had n = 7 (12 strips), all CD and test diet (TD) torpor groups had n = 8 (16 strips). Sprague–Dawley rat (SD) was used as a non-hibernator reference point, n = 30 (57 strips). Mann–Whitney U test. Each animal can produce multiple muscle strips, so if the same group had two strips from the same animal, it was still an n = 1, but both dots were shown on the graph. The analysis presented in this figure utilized the initial functional assessment of animals before treatment with different temperatures. The n is a reflection of those total animals. *P ≤ 0.05 (CD LT over TD LT; CD ET over TD ET; CD ET over CD LT n = 7 to 8), ****P ≤ 0.0001 (CD ET over CD AR; CD LT over CD AR; n = 7–8)

Table 1.

Contractile properties during a torpor bout progression in A control diet and B test diet animals at 37 °C

Species/torpor status	Rate of force development (N/cm2)	Maximum force (N/cm2)	Rate of relaxation (N/cm2)
A
Early torpor (ET)	34.1 ± 5.9	11.4 ± 1.3	− 347.4 ± 24.2
Late torpor (LT)	34.0 ± 3.6	13.1 ± 0.78	− 449.9 ± 24.2
Aroused (AR)	33.6 ± 5.7	11.1 ± 1.8	− 160.3 ± 18.0
Sprague–Dawley (SD) rats	18.9 ± 0.6	10.1 ± 0.57	− 107.7 ± 0.5
B
Early torpor (ET)	26 ± 3.5	11.5 ± 1.29	− 188.4 ± 18.4
Late torpor (LT)	27.7 ± 2.7	10.6 ± 1.0	− 164.3 ± 17.2
Aroused (AR)	32.8 ± 2.97	13.6 ± 0.87	− 147.8 ± 11.5

Characterization of SkM contractile function at 150 Hz in hibernating AGS along with SD rats as a non-hibernator reference point. The data presented are means ± SEM (ET n = 7 for CD, n = 8 for TD, late torpor (LT), and Aroused (AR) n = 8 for both CD and TD). Sprague–Dawley (SD) n = 14. The analysis presented in this table utilized the initial functional assessment of animals before treatment with different temperatures. The n is a reflection of those total animals

Contractile and dietary properties

Functional analysis revealed significant differences in maximum force production and rate of relaxation between CD and TD AGS during the initial phase of testing at 37 °C (Table 1, Fig. 3). At 150 Hz, the maximum force produced in tissue collected during ET, LT, and AR (Fig. 3A, B) had small, but statistically significant differences (only CD LT compared to TD LT; P ≤ 0.05, n = 4, Kruskal–Wallis one-way ANOVA) between diet groups. Similarly, at all stimulation frequencies, the diet had no significant effect on RFD (Fig. 3C, D). Between ET and LT time points, the relaxation rate in the TD group statistically lower than in the CD group (P ≤ 0.05, for both ET and LT; n = 3–4), but not during AR (Fig. 3E, F). When compared to CD animals, the relaxation rates of the TD group were not different comparing LT and ET. The CD group’s LT ex vivo tissues exhibited the fastest rate of relaxation. Overall, the relaxation rate was the most telling response and indicated a significant alteration of the calcium re-uptake machinery. Furthermore, the data showed no sex specific response (Supplemental Fig. 2).

The effect of diet on protein expression

We discovered that the normalized density of SERCA 1a protein expression (supplemental Fig. 6a) in CD and TD animals was similar across different stages of torpor. In contrast, SERCA 1a and SERCA 2a expression in CD animals was substantially higher than in TD animals during AR (P ≤ 0.05; n = 4) (supplemental Fig. 6b). SLN expression was similar without any significance between diet groups at each stage of torpor and arousal (supplemental Fig. 6c). When both diet groups (CD and TD) were combined, we identified a substantial increase in SLN expression as torpor progressed (P ≤ 0.01, ET over LT, LT over AR, and AR over ET; n = 6). We discovered that SERCA 1a, SERCA 2a, and SLN expression did not differ significantly across any of the animal groups studied. As a result, diet and torpor state may have had minimal influence on the expression of these proteins during torpor or arousal.

Temperature effects on inherent contractile properties

Supplementary Figures 3–5 show the FFCs of maximum force production, RFD, and rate of relaxation at various stages of torpor fed CD or TD. All contractile properties were obtained after 30 min of exposure to 37 °C, variable temperatures (4 °C, 15 °C, 25 °C, and 37 °C), and a recovery phase at 37 °C. Throughout all steps, all strips had relatively similar RFD and maximum force production, with no significance regarding hypothermic temperature exposure. Except for LT AGS CD animals, none of the animals demonstrated any significant diet based contractile performance changes and were functionally inactive at 4 °C during the variable temperature phase (Fig. 4). The recovery of all DIA strips from hypothermic temperatures to 37 °C was very similar, regardless of diet or torpor. Except for the rate of relaxation, no other contractile properties differed significantly during the recovery phase in terms of diet and torpor. CD AGS, on the other hand, demonstrated a significant increase in relaxation rate in relation to diet and torpor at all stimulation frequencies (Fig. 4). It was interesting to note that the functional recovery of 15 °C and 4 °C tissues was very drastic, and SkM performance was almost identical to or even higher than the initial state (Fig. 5). Tissue recovery was subtle at 37 °C and 25 °C. Among contractile properties, diet and torpor only affected the rate of relaxation. Furthermore, regardless of ambient temperature, the rate of relaxation among LT animals was significantly faster than that of the AR group during the progression of torpor bout, whereas maximum force production and RFD remained consistent across all groups.

Fig. 4.

Temperature resilience does not affect loss in contractile kinetics. Data shown are the loss in rate of force development (RFD)/max force/relaxation rate after 30 min of exposure at the temperature indicated on the x-axis. Results are expressed as a percentage of RFD/max force/relaxation rate measured at 37 °C, 30 min before the variable temperature phase. The data presented are means ± SEM. Except for the control diet (CD) group of early torpor (ET), which had n = 3 (1 female, 2 males), all CD and test diet (TD) torpor groups had n = 4 (2 female, 2 males)

Fig. 5.

Temperature resilience does not affect functional recovery. Data shown are the rate of force development (RFD)/max force/relaxation rate during the recovery phase at a bath temperature of 37 °C after 30 min of exposure at the temperature indicated on the x-axis. After allowing the DIA strips to recover from variable temperature to 37 °C for 30 min and tissues were stimulated at 150 Hz, the percentage loss (with respect to the initial phase) in terms of force development, maximum force, and relaxation rate was calculated. The data presented are means ± SEM. Except for the control diet (CD) group of early torpor (ET), which had n = 3 (1 female, 2 males), all CD and test diet (TD) torpor groups had n = 4 (2 female, 2 males)

No change in temperature resilience between diet and torpor

There is little-to-no literature that discusses the effects of diet and torpor on SkM contractile performance during hibernation. Temperature resilience is defined as the muscle’s ability to maintain contractile properties during temperature changes (in this case hypothermic temperature). Contractile properties were expressed as a percentage loss from an initial 37 °C exposure and demonstrated a proportional drop with temperature, with very low values expressed at 4 °C. The temperature was then increased to 37 °C during the recovery period. Regardless of diet or torpor, all tissues were found to have similar temperature resilience in functional recovery (Fig. 5).

Discussion

This study shows that a diet with a 1:1 ω-6:ω-3 ratio, that mimics the diet of free-ranging animals, decreases relaxation kinetics when compared to a standard rodent chow with a 5:1 ω-6:ω-3 ratio. This suggests that a 1:1ω-6:ω-3 ratio, which is associated with a similar ω-6:ω-3 ratio in the blood (Rice et al. 2021; Mikes et al. 2022), may decrease calcium re-uptake. The influence on relaxation kinetics observed during torpor but not during AR suggests that the ω-6:ω-3 ratio plays a greater rate-limiting role in muscle relaxation during torpor than during other phases of hibernation. Interestingly, the 1:1 ratio of ω-6:ω-3, which was expected to mimic a free-ranging diet (Rice et al. 2021; Mikes et al. 2022), did not improve relaxation kinetics during the other phases of hibernation tested.

We interpret these findings as evidence of resistance to Ca²⁺ overload in SkM of AGS. Decreased relaxation kinetics in DIA muscle at cold-temperature torpor argues for decreased SERCA activity and persistently elevated cytoplasmic Ca²⁺. Other tissues, particularly the brain, have been shown to be resistant to injury during ischemia/reperfusion (Drew et al. 2001). Resistance might be attributed in part to processes downstream of elevated cytoplasmic Ca²⁺ (Zhao et al. 2006; Bhowmick et al. 2017). Reduced SERCA activity during cold-temperature torpor would contribute to energy savings. Slower DIA muscular relaxation is adequate to satisfy the respiratory demands of hibernating AGS. Resistance to injury, which is frequently linked with elevated cytoplasmic Ca²⁺, may potentially play a role in the adaptation in AGS that allows for cost-effective respiratory function during hibernation.

Influence of diet and torpor

Dietary fatty acids, such as PUFA, have been shown to influence SkM function and metabolism (Jeromson et al. 2015), as well as altering SERCA activity in hibernating animals (Giroud et al. 2013). The ratio of increased ω-6 in ω-6:ω-3 dietary PUFA and not just ω-6 n fatty acid content in the SR can affect the activity of the SERCA 2a pump (Giroud et al. 2013). Our findings show that increased ω-6:ω-3 ratio of PUFAs in the diet and progression of torpor have a significant effect on relaxation kinetics (Ca²⁺ re-absorption), but not on RFD (rate of calcium release) or max force production. Previous research in hibernators such as bats (ex vivo) (Lee et al. 2008) and black bears (in vivo) (Lohuis et al. 2007) found no change in RFD or max force in-between seasons. In an indirect comparison, an in vivo functional assay on tibialis anterior muscle (94% type 2 muscle fibers) in black bear found no significant difference in relaxation rate between early and late hibernation season (Lee et al. 2008) and black bears (in vivo) (Lohuis et al. 2007). At a more acute time scale, our current ex vivo study discovered significant changes in the relaxation rate of the DIA (55% type 1, 45% type 2, mixed fiber type) during the progression of torpor. This contradictory result could be attributed to differences in hibernating species/timing, the protocol used, and/or tissue type. For example, hibernating black bear does not reach core body temperatures as low as AGS nor do they hibernate for the same duration, which could limit the alterations in SERCA efficiency. Additionally, DIA was chosen for this study for its ease of preparation but also because DIA is continually utilized even in extremely low body temperatures, which makes it a very important muscle for continual maintenance.

In comparison to the TD (1:1 ω-6:ω-3 PUFA ratio), our CD animals with a 5:1 ratio of ω-6:ω-3 PUFA demonstrated increased rates of muscular relaxation. Previous research in cardiac muscle has discovered that a higher ω-6:ω-3 dietary ratio can alter SR membrane lipid composition and promote SERCA activity (Giroud et al. 2013, 2018). Considering these findings, we speculate that, despite the lack of direct assessment of SR membrane lipid composition, the increased rate of muscle relaxation reported in our CD group is most likely due to an increase in SERCA activity mediated by a change in SR membrane lipid composition, regardless of muscle type. However, further studies on this need to be completed. Unlike RFD and force production, the rate of relaxation within the CD group was found to increase as torpor progressed (i.e., from ET to AR). Along with decreased SLN expression during hibernation (reviewed in Oliver et al. (2018)), an increased ω-6:ω-3 ratio in the SR membrane will boost SERCA activity during torpor, allowing hibernators to achieve and tolerate T_b closer to ambient temperature (Giroud et al. 2013). Changes in relaxation rates can be influenced by factors other than membrane composition, such as SERCA regulation by peptides (SLN, phospholamban, and myoregulin), SERCA efficiency (Ca²⁺ pumped/ATP spent), and ATP supply. Future studies need to investigate the effects of torpor and dietary change on these parameters by utilizing direct measurements from a functional standpoint.

Our finding also highlight important considerations for how artificial laboratory diets may potentially influence molecular and physiological aspects of hibernation that may not be present in free-range populations (Mikes et al. 2022). We and others have shown that free-range hibernator’s dietary ω-6:3 PUFA consumption is far closer to 1:1, with greater ω-3 consumption (Florant et al. 1990; Arnold et al. 2011), than common laboratory rodent chows often fed to hibernating squirrels (Mikes et al. 2022). Given endogenous products of ω-6 and 3 PUFAs have pro- and anti-inflammatory capabilities (such as the series 3 leukotrienes or series 2 prostaglandins) (Schmitz and Ecker 2008) and the influence dietary ω-6 and 3 PUFAs have on phospholipid membrane composition, there may be unknown and unaccounted molecular and physiological differences between findings of free-range vs. laboratory hibernations due to dietary input. Such differences could influence overall metabolic data as well as muscle function.

Effects of low-temperature exposure

Muscles require a temperature range that is optimal for performance (Petrofsky and Lind 1981; Crowley et al. 1991). The temperature of the surrounding tissue affects the contractile and biochemical properties of skeletal muscle (Edman 1979; Bottinelli et al. 1996). Previous research has shown that below normothermia (37 °C), muscle contractile properties such as RFD and force production are impaired (Holewijn and Heus 1992; Zhou et al. 1998). Our findings are consistent with previous literature. Specifically, hypothermic exposure during the variable temperature phase suppressed contractile function at all torpor stages (Fig. 4). Contractile suppression was directly proportional to tissue temperature, with the lowest values at 4 °C. Despite differences in species/muscle type/experimental timing, hypothermic studies in rodents (Ranatunga and Wylie 1983; Coupland and Ranatunga 2003) and humans (Ranatunga et al. 1987; Cornwall 1994) support our findings. In contrast, only a few studies in rat DIA (Foldes et al. 1978) or SOL (Ziganshin et al. 2017), where twitch force was measured during hypothermic temperature exposure and rewarming, revealed an increase in muscle contraction amplitude. Why these studies showed improved muscle contraction at low bath temperature is unclear. A recent study has been shown that the muscle cross-sectional area has an effect on oxygen transport and, as a result, oxygen availability, which may be mitigated by lowering the experimental temperature (Barclay 2005). Previous research on the effects of hyperthermic stress on SkM contractile performance has found no functional recovery (Oliver et al. 2018; Poel and Stephenson 2002), whereas the present study found complete contractile functional recovery or improved performance in some individual AGS, but only partial recovery in rats. As a result, it appears that there may be species differences in the functional recovery of SkM contractile properties following cold stress.

Effects on the expression of the SERCA/SLN proteins

The relationship between muscle performance and protein expression remains poorly understood. Recent findings (Guo et al. 2017; Zhang 2019) of SERCA (1a and 2a) expression in the soleus, EDL, gastrocnemius, plantaris, and adductor magnus were similar in tissue collected during ET, LT, and IBA but higher in tissue collected from animals in the summer. We found that results in DIA were similar to other muscles. These findings support the notion that neither diet nor phases of hibernation influenced SERCA protein expression and protein expression per se cannot explain differences in relaxation kinetics observed in torpid animals fed the test diet. A recent study by Giroud et al. (2013) on Syrian hamster cardiac muscle, analyzed SR membrane lipid composition from hibernating (torpid), summer, and non-hibernating winter animals found that PUFA affects SERCA activity, not SERCA expression. Although we did not measure SERCA activity, the absence of changes in protein expression suggests that a high ω-6:ω-3 diet influenced SERCA activity rather than SERCA expression.

During arousal from torpor, brown adipose tissue and SkM are primarily responsible for thermogenesis (Fons et al. 1997; Carey et al. 2003; Génin et al. 2003). Fons et al. (1997) have demonstrated that during an IBA, the initial increase in T_b is caused by BAT non-shivering thermogenesis, and SkM is only responsible for shivering thermogenesis (high energy demand) once T_b exceeds 17 °C. Even though SLN expression was not shown to be significantly related to diet in the current study, we did notice a statistically significant trend for SLN expression (both diets combined) to increase from torpor to IBA during hibernation. Similar findings have been found in both SLN gene expression (mRNA, transcriptomics, and microarray) and SLN protein expression in both 13-lined ground squirrel and AGS SkM tissue where SLN is least expressed during torpor, increases toward terminal arousal, and peaks during summer (Oliver et al. 2018; Carey et al. 2003; Hampton et al. 2011; Fedorov et al. 2014). Thus, our findings support the possibility of SLN expression being involved in SkM NST (Bal et al. 2012; Oliver et al. 2018), i.e., increased SLN expression reduces the efficiency of SERCA activity by reducing Ca²⁺ accumulation in the SR without changing the rate of ATP hydrolysis (Bal et al. 2012), resulting in energy released in the form of non-shivering thermogenesis to achieve interbout arousal.

Critique of approach

(1) To avoid muscle function loss, the extraction processes were completed quickly, and the tissues were transferred to a temperature-controlled (37 °C), isotonic, and continuously oxygenated Ringer’s solution. Furthermore, the diet and torpor groups received similar treatments, so variation due to ex vivo analysis is reflected in both groups. (2) We hypothesized that the contraction slope (i.e., RFD) represents the indirect activity of RyR1 channel Ca²⁺ release and the relaxation slope (i.e., relaxation rate) represents the indirect activity of the SERCA pump based on supplementary experimental data (supplemental Fig. 1). Our hypothesis was validated by observing a dose-dependent decrease in RFD with RyR1 receptor inhibition using dantrolene and MgCl₂ (supplemental Fig. 1a). CPA was also used to inhibit the SERCA 1a pump, which resulted in a dose-dependent decrease in the rate of muscle relaxation (supplemental Fig. 1b). (3) We do not go into detail about the differences in contractile properties between AGS and Sprague–Dawley rats, because we focused the current study on AGS contractile properties related to hibernation and torpor, and Sprague–Dawley data were used to show baseline properties from non-hibernating species for a point of reference. (4) One of the major limitations of our current study is the absence of direct measurement of RyR/SERCA activity and the influence of diet on the Ca²⁺ cycle. This limitation is attributed to our approach employing a functional aspect of SkM performance and contractile properties in relation to PUFA diet and torpor, and using the inhibition study as a proof of concept to demonstrate the link between our data and our conclusions.

Conclusion

A high ω-6:ω-3 PUFA ratio in the diet during torpor enhances the inherent SkM relaxation kinetics. This evidence supports the interpretation that diet and hibernation can influence SERCA pump activity and SkM functional properties and thus influence SkM performance. Overall, our current model emphasizes the importance of diet in conjunction with torpor status for contractile performance in hibernation models, as well as the importance of researching diet in myopathic disease therapeutics. Furthermore, due to the striking changes in relaxation rates across hibernation and with diet, studies using hibernating animal models should take care in interpretation of metabolic and muscle function data when feeding only rodent chow. The interpretations may not align with what is occurring in the wild.

Prospectus

Hibernating animals demonstrate metabolic depression and reduced skeletal muscle activity that is enhanced when exposed to ambient temperatures near freezing. Through adaptations of hibernation, these animals maintain muscle mass and adapt to extremely cold environments. Understanding the effects of tissue temperature and diet on muscle physiology and contractile function in hibernators may thus aid in the development of therapeutic interventions for a wide range of diseases that affect skeletal muscle function. These interventions could include but are not limited to (1) muscle weakness (due to osteolytic cancer) and (2) muscle atrophy (as seen in elderly patients and spaceflight crewmembers).

Supplementary Material

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00360-023-01527-z.

Data availability

Raw data is available upon request.