Material preparation and characterization

A highly stable reproducible pristine C₆₀ fullerene aqueous colloid solution (C₆₀FAS) at a concentration of 0.15 mg/ml was prepared according to a previous protocol [20, 21]. Briefly, for the preparation of C₆₀FAS we used a saturated solution of pure C₆₀ fullerene (purity >99.99%) in toluene with a C₆₀ molecule concentration corresponding to maximum solubility near 2.9 mg/ml, and the same amount of distilled water in an open beaker. The two phases formed were treated in ultrasonic bath. The procedure was continued until the toluene had completely evaporated and the water phase became yellow colored. Filtration of the aqueous solution allowed to separate the product from undissolved C₆₀ fullerenes. The pore size of the filter during the filtration of the aqueous solution was smaller than 2 µm (Typ Whatmann 602 h1/2). The purity of prepared C₆₀FAS (i.e., the presence/absence of any residual impurities, for example carbon black, toluene phase) was determined by HPLC and GC/MS analysis. The maximal concentration of C₆₀ fullerenes in water 0.15 mg/ml was obtained by this method.

The state of C₆₀ fullerenes in aqueous solution was monitored using atomic force microscopy (AFM). Under AFM analysis, the sample was deposited onto a cleaved mica substrate (V-1 Grade, SPI Supplies) by precipitation from an aqueous solution droplet. Sample visualization was performed in semi-contact (tapping) mode (Fig. 1a, b). AFM measurements were performed after the complete evaporation of the solvent.

Small-angle neutron scattering (SANS) measurements (Fig. 1c) were carried out at the YuMO small-angle diffractometer at the IBR-2 pulsed reactor (JINR, Dubna, Russia) in the time-of-flight mode with the two-detector setup [22]. Treatment of the raw data was performed by the SAS program [23].

Procedure and experimental groups

Male Wistar rats, weighing 280–350 g, were used in the study. The use of the animals was approved by the Ethics Committee of the Institute and performed in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC).

The animals were divided into 4 groups. In the experiments, the m. triceps surae fatigue was induced by electrical stimulation of n. tibialis. Saline solution (group 1, n = 6) or C₆₀FAS (F-injection) 0.1–0.15 mg/kg (group 2, n = 6) was administered into the left TS after the development of fatigue. Then, fatigue of the right TS was induced. The data obtained from the ipsilateral (left) side were considered to be the control values vs. those obtained from the contralateral side. The dose range of 0.1–0.15 mg/kg C₆₀FAS does not present any acute or subacute toxicity in rats [13]. The rats of group 3 (n = 6; animals with fatigue of both TS without any injections) and group 4 (n = 6; intact animals) were used only for biochemical studies. After the experiment, the TS of all animals in all groups were removed for biochemical analysis.

It is important to note that a dose of 0.1–0.15 mg/kg C₆₀FAS applied in our experiments does not present any acute or subacute toxicity in animals: it was significantly lower than the maximum tolerated dose of C₆₀ fullerene, which was found to be 5 g/kg both for oral or intraperitoneal administration to rats [13]. No toxic effects or death have been fixed under the action of C₆₀ fullerenes after their oral administration to rats in total dosage of 2 g/kg for 14 days [24]. Finally, it was shown [13] that water-soluble C₆₀ fullerenes administered intraperitoneally to rats (0.5 mg/kg) were subjected to clearance from the organism within 2–4 days.

The animals in groups 1 and 2 were anaesthetized with ketamine (100 mg/kg “Pfizer”, USA) combined with xylazine (10 mg/kg, “Interchemie”, Holland), tracheostomized and artificially ventilated (out of necessity). The left and right TS muscles were separated from the surrounding tissue, and their tendons were detached at the distal insertions. The n. tibialis was separated from the tissue and cut proximally, and all branches of the nerve, except those innervating the TS, were cut. This nerve was mounted on a bipolar platinum wire electrode for electrical stimulation. The hindlimb muscles and nerves were covered with paraffin oil in a pool formed from skin flaps. The TS muscle was connected via the Achilles tendon to the servo-control muscle puller. The muscle tension was measured by semi-conductor strain gauge resistors glued on a stiff steel beam mounted on the moving part of a linear motor.

To induce muscle fatigue, 1–3 (30 min duration) series intermittent high-frequency electrical stimulation was used (Fig. 2a), separated by rest intervals of 10–20 min. Each series consisted of trains of 0.2-ms rectangular pulses at a rate of 40/s at 12.4 s duration and separated by 5 s intervals of rest (Fig. 2b). The stimulus current was set to 1.3–1.4 times the motor threshold. Note, if muscle fatigue developed in less than 30 min, stimulation was interrupted (it was predicted that fatigue development occurred when there was a muscle force decrease of more than 50% of the initial data). After the end of the 12.4-s-stimulation, the muscle was stretched, and the change in length had a bell-shaped form (one period of 4 Hz sinusoidal signal with corresponding phase locking) of 3.5 mm amplitude and 2 s duration (Fig. 2b; bottom row). The muscle reaction to the stretches appeared as a tension increase after continuous stimulation. These stretches were applied before the post-stimulation twitches to remove, or at least diminish, the after-effects remaining from the continuous stimulation [1]. The signals (stimulus pulses, muscle tension and other) were sampled via DAC-ADC device (CED Power 1401).

Biochemical experiment

For biochemical analysis, the excised m. triceps surae (soleus and gastrocnemius) were rapidly dissected, free of fat and tendon, divided into several portions and stored in liquid N₂. For reduced glutathione (GSH) analysis, tissue samples were transferred into a medium containing 1 N perchloric acid (1:10 w/v) and homogenized with a motor-driven Potter–Elvehjem glass homogenizer. The resultant homogenate was centrifuged at 10,000g for 10 min (4 °C). The GSH content was spectrophotometrically measured [25]. For the enzyme activity assays and H ₂ O ₂ and lipid peroxidation assays, the muscle samples were thawed and homogenized in 50 mM phosphate buffer with 2 mM EDTA (pH 7.4) at 4 °C (1:9 w/v). Homogenates were then centrifuged for 15 min at 15,000g (4 °C), and the post mitochondrial supernatant was stored at −70 °C.

Oxidative damage in the tissue was measured using the thiobarbituric acid reactive substances (TBARS) assay. TBARS were isolated by boiling tissue homogenates for 15 min at 100 °C with thiobarbituric acid reagent (0.5% 2-thiobarbituric acid/10% trichloroacetic acid/0.63 M/dm³ hydrochloric acid) and measuring the absorbance at 532 nm. The results are expressed as nM TBARS/mg protein, using ɛ = 1.56 × 10⁵ dm³/M¹/cm¹ [26].

The H ₂ O ₂ concentration in the tissue homogenates was measured using the FOX method, which is based on the peroxide-mediated oxidation of Fe²⁺, followed by the reaction of Fe³⁺ with xylenol orange (o-cresolsulphonephthalein 3′,3″-bis[methylimino] diacetic acid, sodium salt). This method is extremely sensitive and is used to measure low levels of water-soluble hydroperoxide present in the aqueous phase. To determine the H ₂ O ₂ concentration, 500 μl of the incubation medium was added to 500 μl of assay reagent (500 μM ammonium ferrous sulphate, 50 mM H₂SO₄, 200 μM xylenol orange, and 200 mM sorbitol). The absorbance of the Fe³⁺-xylenol orange complex (A₅₆₀) was detected after 45 min. Standard curves of H ₂ O ₂ were obtained for each independent experiment by adding variable amounts of H ₂ O ₂ to 500 μl of basal medium mixed with 500 μl of assay reagent. Data were normalized and expressed as μM H ₂ O ₂ per mg protein [27].

Catalase activity was measured by the decomposition of hydrogen peroxide, determined by a decrease in the absorbance at 240 nm [28].

GSH was determined using Ellman’s reagent. One millilitre of supernatant was treated with 0.5 ml of Ellman’s reagent (5.5′-dithio-bis-nitrobenzoic acid in abs. ethanol) and 0.4 M Tris HCl buffer with 2 mM EDTA, pH 8.9. The absorbance was read at 412 nm in a spectrophotometer [25].

The protein concentration was estimated using the method of Bradford with bovine serum albumin as a standard. All chemicals were purchased from Sigma, Fluka and Merck and were of the highest purity.

Data analysis

In the electrophysiological part of the study, each stimulation series (30 min) was divided into three equal portions (Fig. 2a), which were averaged (maximum 33 stimulation in one portion). The average value of the first portion was set to 100%, and the other series were normalized in relation to this (for each hindlimb). The peak amplitudes of the front (P ₁ ) and rear of the front (P ₂ ) (maxima amplitudes at the site, duration of 1 s, Fig. 2b) of the muscle strength of each single series (12.4 s) were identified and the difference between P ₁ and P ₂ (ΔP) was calculated. This difference determines the dynamic component of the muscle force decrease in a short period of continuous stimulation. Mean values (mean ± SD) of the TS muscle strength before and after F-injection were compared using a two-way statistical analysis of variance (ANOVA). The factors of variation included two conditions, time and the effects of the C₆₀FAS. A Bonferroni post hoc analysis was used to determine the differences between groups. The level of significance was set at p < 0.001.

Biochemical data are expressed as the means ± SEM for each group. The differences among experimental groups were detected by one-way ANOVA followed by Bonferroni’s multiple comparison test. Values of p < 0.05 were considered significant.

Analysis of AFM and SANS data

Because the C₆₀ fullerene particle size directly correlates with their biodistribution and toxicity [29, 30], the AFM and SANS studies were performed.

The AFM images (Fig. 1a, b) clearly demonstrate randomly arranged, individual C₆₀ fullerenes (0.7 nm in diameter) and their bulk clusters with a height of 1.5–200 nm. At the same time, some individual C₆₀ fullerene aggregates with a height of >200 nm are also seen in the AFM image (Fig. 1b). The results obtained are consistent with the theoretical calculations and experimental measurements [20, 21, 31, 32] and demonstrate the polydispersity of the C₆₀FAS used in our study.

Experimental SANS curve for C₆₀FAS is shown in Fig. 1c. The scattering curve of C₆₀FAS is well described by the form-factor of polydisperse spherical particles. The mean radius of gyration of the particle cross section, R_g, and pair distance distribution function, P(r), were found by using indirect Fourier transformation (IFT) approach [33]. We can calculate the radius of particles, R, present in the C₆₀FAS according to well-known equation R _g ²  = 0.6R² assuming of homogeneous and spherical of C₆₀ fullerene clusters. This conclusion follows from previous experimental data [20, 21] and the estimates of the average cluster density according to the contrast-variation experiments [31, 32, 34]. The data given by this procedure indicate that C₆₀FAS consists of C₆₀ fullerene sphere-like nanoparticles with an average size of ~56 nm that is in a good agreement with above AFM data.

It is known [35, 36] that the permeability and cytochemical behavior of nanoparticles strongly depend on their size and, correspondingly, mass (number) distribution. In this regard, our previous studies [16, 18, 19, 29] clear demonstrate that the used C₆₀ fullerene nanoparticles can effectively penetrate through the plasma membrane of cells by passive diffusion or endocytosis (depending on the size) and do not exhibit cytotoxic effects.

Electrophysiological experiments

Changes in the TS force reaction under fatigue conditions due to prolonged high frequency stimulation (30 min, 40/s) of the n. tibialis for animal groups 1 (before and after administration of saline solution) and 2 (before the application of C₆₀FAS) did not significantly differ. The analysis was performed by determining the force level at the beginning (P ₁ ) and end (P ₂ ) of single tetanic contractions and the difference between these values (ΔP), which determines the dynamic component of the force decrease during a short period of continuous stimulation (Fig. 2b). The muscle was considered tired if the amplitude of the single tetanic contractions decreased by more than 50% relative to the initial level. When muscle fatigue was reached, the stimulation was stopped and followed by a 10–20 min rest period. Therefore, in the case of one animal, as a result of muscle fatigue stimulation of the left TS, a 50% fatigue level was reached in approximately 12 min; during the next 4 min of stimulation, it continued to decrease [Fig. 3a (I_L), b(I_L)]. After 10 min of rest, a single tetanic contraction force was slightly restored, but it did not reach the initial muscle activity level and continued to decrease rapidly [Fig. 3a (II_L), b(II_L)]. In this case, there was also a simultaneous decrease in the dynamic component of the force drop ΔP. Note that the dynamic component was the most highly expressed at the beginning of the first experimental series and that the P ₁ amplitude was higher relative to the P ₂ amplitude [Fig. 3b (II_L)]. After tetanic contractions for 1.5–2 min, difference between amplitudes P ₁ and P ₂ was reduced to zero, with moderate variations both in one and the opposite direction over the additional fatigue stimulation period. Simultaneously with the decrease in ΔP values, there was a constant decrease in the developed force. In the following stimulation series, after a period of rest, the initial amplitude of the dynamic component was usually decreased [Fig. 3b (II_L, III_L)].

When a predetermined level of muscle fatigue was reached, C₆₀FAS (0.1–0.15 mg/kg) was injected intramuscularly [at 45 min after the beginning of fatigue stimulation; Fig. 3a \(({\text{I}}_{{ {\text{L}}}}^{\text{F}}\), \({\text{II}}_{{ {\text{L}}}}^{\text{F}}\)), b(\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\), \({\text{II}}_{{ {\text{L}}}}^{\text{F}}\))]. At the same time, the dynamic changes in the muscle strength level in response to stimulation reflected the further development of fatigue, and the single contraction forces were reduced rapidly [Fig. 3b(\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\))]. However, F-injection led to the gradual recovery of the isometric force levels (at 32 min after drug application; Fig. 3a [(\({\text{II}}_{{ {\text{L}}}}^{\text{F}}\)), b \(({\text{II}}_{{ {\text{L}}}}^{\text{F}}\))]. The appearance of negative ΔP values (P ₂ amplitude increase compared to P ₁ amplitude) indicated the beginning of the recovery [Fig. 3b (\({\text{II}}_{{ {\text{L}}}}^{\text{F}}\)), 10th min]. In this series of stimulations, the level of the muscle contraction force was recovered to that developed during the initial stages of fatigue stimulation.

Power reaction of the right TS was significantly different from the left TS. Notably, the TS of the right limb was not previously fatigued before the F-injection (Fig. 3c, d). At 52 min after drug administration, a certain force muscle decrease was observed. In this case, the P ₁ amplitude was higher than the P ₂ amplitude, as indicated by the increase in ΔP values [Fig. 3c (\({\text{I}}_{{ {\text{R}}}}^{\text{F}}\)), d (\({\text{I}}_{{ {\text{R}}}}^{\text{F}}\))]. However, at 6 min after the beginning of fatigue stimulation, the force developed by the muscle appeared at a certain stationary level, which was held during the experimental series. The difference between the P ₁ and P ₂ amplitudes disappeared (value of ΔP decreased to zero), which may indicate a constant force level at the time of loading. It is significant that the muscle maintained the developed force level for an additional 1 h [Fig. 3c (\({\text{II}}_{{ {\text{R}}}}^{\text{F}}\), \({\text{III}}_{{ {\text{R}}}}^{\text{F}}\)), d(\({\text{II}}_{{ {\text{R}}}}^{\text{F}}\), \({\text{III}}_{{ {\text{R}}}}^{\text{F}}\))]. For this muscle, the total time of the decrease of the isometric force contraction by 50% was 120 min after drug administration. For comparison, the control duration of the fatigue occurrence period was 42 min.

In Fig. 4a, b, a comparison of the force level changes developed by the left TS before (I_L) and after (\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\)) F-injection in two different experiments is presented. The statistical analysis showed a significant (p < 0.001) force decrease during the series of fatigue stimulation before C₆₀FAS administration [Fig. 4a (I_L), b(I_L)]. After F-injection, a recovery of the muscle force for the first animal [Fig. 4a (\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\))] and its holding for the second animal [Fig. 4b (\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\))] was observed. At the end of the experimental series, the recovery of the active muscle force response was significant compared to that at the beginning and nearly reached the control values. The analysis showed a significant effect for the factors: drug administration (D) and time (T) after administration and their interaction. The corresponding results of this analysis were as follows: F(D) = 2904.47, p(D) < 0.001, F(T) = 42.420, p(T) < 0.001, F(DxT) = 1350.58, p(DxT) < 0.001 (first animal) and F(D) = 122.80, p(D) < 0.001, F(T) = 1058.29, p(T) < 0.001, p(DxT) = 1287.35, p(DxT) < 0.001 (second animal). The data obtained in all experiments (Fig. 4c–h) indicate that the decrease in the developed force after C₆₀FAS administration (\({\text{I}}_{{ {\text{L}}}}^{\text{F}}\), \({\text{II}}_{{ {\text{R}}}}^{\text{F}}\)) was almost two times slower than in controls. The maximum significant reduction of the muscle force developed during the entire period of fatigue stimulation was 44% after drug administration, whereas in the control this was 85%. For all experimental animals, similar dynamics of the force level decrease in the control and its more gradual decrease after F-injection were observed.

Biochemical experiments

During long-term stimulation of the muscle, metabolic processes change and are a main factor of muscle fatigue. As a result of the fatigue test, the accumulation of lipid peroxidation secondary products and changes in the levels of antioxidants in the tissue of the fatigued muscle were determined. The data clearly demonstrate the increased level of peroxidation and oxidative stress marker TBARS and H ₂ O ₂ after fatigue stimulation (Fig. 5a, b). This increase was significant in relation to the intact muscle (‘norm’) and was 23% (p < 0.05) for TBARS and 38% (p < 0.05) for H ₂ O ₂ . After C₆₀FAS administration into the left TS, the TBARS concentration was significantly reduced compared to fatigue as follows: 29% (p < 0.05) for the left TS and 12% (p < 0.05) for the right one. The H ₂ O ₂ level decreases in comparison to the ‘fatigue’ group (by 6% for the left TS and 7% for the right one), although the H ₂ O ₂ level remained higher in relation to the intact group (p < 0.05). In turn, in response to such changes in the working muscle, an activation of endogenous antioxidants occurred. During fatigue stimulation, the amount of muscle GSH quantitatively increased more than two-fold (p < 0.05) and the activity of the antiperoxide enzyme CAT also increased. After C₆₀FAS administration, the GSH and CAT activities were significantly decreased compared to the group ‘fatigue’ by 41.8 and 15.4% for GSH and 53 and 43% for CAT (p < 0.05) for the left and right TS, respectively (Fig. 5c, d).