Reflex and muscular adaptations in rat soleus muscle after hindlimb suspension

The soleus reflex excitability of the motoneuron is usually tested by the quantification of the ‘H reflex’, i.e. the electromyogram elicited by an electrical stimulation of the sciatic nerve. A mechanical tap applied to the Achilles tendon gives another reflex response measurable as an electromyogram, the ‘T reflex’, which involves all the components of the reflex pathway, including the muscle spindles. The H and T reflexes can be obtained in conscious rats (Pérot and Almeida-Silveira, 1994) and changes in these reflexes after a training period are consistent with those observed in trained humans. The soleus muscles of subjects submitted to endurance training show higher H reflex (Ginet et al., 1975; Rochcongar et al., 1979) and T reflex (Dirks and Hutton, 1984) responses than sedentary subjects. Conversely, power training programs lead to lower soleus H reflex responses (Casabona et al., 1990; Nielsen et al., 1993). Changes in reflex excitability of the soleus muscle could be a direct consequence of training, since they are observed when comparing the reflexes obtained before and after the period of training (Pérot et al., 1991; Voigt et al., 1998). Type I fibres are innervated by motoneurons that are more excitable than those innervating type II fibres (Henneman and Olson, 1965) and these reflex changes could therefore reflect adaptations of the motor units to the period of training. Our previous studies on the rat are consistent with this hypothesis: rats where muscles were trained by subjecting them to repeated stretch–shortening cycles showed a lower H reflex in the ankle plantarflexor muscles, and their soleus muscle had mechanical and histochemical properties consistent with a slow-to-fast fibre-type transition phenomenon (Almeida-Silveira et al., 1996a). Opposite results (an increase in reflex amplitude and a relative enhancement of slow fibres) were observed in rats trained to run (Pérot and Almeida-Silveira, 1995). Thus, changes in functional demand on a muscle, induced by hyperactivity (exercise), initiate paired adaptations of the reflex excitability and of the muscle’s mechanical and morphological properties. In this paper, we examine whether such paired adaptations are also present after a period of hypoactivity (muscle disuse).

Hindlimb suspension (HS) of rats was chosen as the model of disuse. This model is known to disturb the mechanical properties of the postural muscles. After HS, the slow-twitch soleus muscle exhibits an increase in contractile speed (Diffee et al., 1991; Fitts et al., 1986) as well as a decrease in stiffness (Canon and Goubel, 1995). These mechanical adaptations are accompanied by a decrease in the proportion of type I relative to type II fibres (Desplanches et al., 1987). With respect to the effects of training, a decrease in reflex excitability is expected from muscles with an increased proportion of type II fibres. Even if mechanical and histochemical results confirm such an enhancement, the reverse changes in H and T reflex responses are evidence in support of different adaptations at the peripheral and central sites of the reflex pathway.

Sprague–Dawley rats were chosen for this study because of their lack of aggressiveness during the reflex tests. However, some adult Sprague–Dawley rats did not endure the suspension treatment: they became depressed and their state prevented us from continuing the experiments. Therefore, for these experiments, immature, more lively, rats were used. Furthermore, it is well known that the suspension treatment has a greater effect on the muscle properties in younger rats (Asmussen et al., 1989; Steffen et al., 1990).

5-week-old male rats (mass 189±26 g, N=12) were used in the experiments. At this age, they were subjected, for the first time, to the reflex methodology described below (test 1: 5-week-old rats). They were then randomly divided into either the hindlimb-suspended (HS) or the control group (C). In the HS group, 6-week-old rats were suspended by the tail to prevent the hindlimbs from coming into contact with any supporting surface, according to the method of Morey (1979). The suspension device was placed so as to avoid tail or hindlimb pain. During the 3 weeks of suspension, animals had free access to food and water. After the suspension period, rats in the HS and C groups were submitted again to reflex tests (test 2: 9-week-old rats). This protocol was approved by the University’s hygiene, safety and ethics committee.

The excitability of the right and left ankle extensor muscles was assessed by measuring their T and H reflexes, in response to an Achilles tendon tap and to a transcutaneaous submaximal electrical stimulation of the sciatic nerve, respectively.

The reflex methodologies were applied to conscious rats lying on a specific device which held the limb such that the knee and ankle joints were approximately at 90 ° (for details see Pérot and Almeida-Silveira, 1994). The rats were held in place by hand, and the sole of the foot was placed on a plate covered with adhesive tape. Leg withdrawal and foot movements were prevented by the experimenter placing his/her finger on the rat’s foot.

Surface silver electrodes (2 mm in diameter) were attached using adhesive to the skin of the shaved leg in the area overlying the soleus muscle, in a mid-leg position. This electrode position was carefully reproduced during each experiment. The use of surface electrodes was prefered to implant electrodes placed into the soleus itself. Indeed, such an implantation would considerably stress the animals, impeding the quiet state of the rat that is necessary to conduct the reflex testing. Furthermore, the maximal twitch induced with inserted electrodes would be painful (as in humans) and would incite the rat to escape from the stimulus. A reference electrode was inserted into the tail. The recorded electromyograms (EMGs) were representative of the activation of both the gastrocnemius lateralis and soleus muscles (Pérot and Almeida-Silveira, 1994). A complementary experiment was conducted to verify the major contribution of the soleus muscle to the reflex responses detected with surface electrodes. This experiment entailed recording maximal H and motor (M) reponses before and 4 days after the soleus nerve had been carefully cut without damaging the gastrocnemius and soleus muscles. This surgical treatment was carried out on both legs of three rats.

To elicit the response of electrical stimulation, the experimenter pressed stimulation electrodes on the skin overlying the femur. The electrical stimuli were 0.2 ms in duration and the intensity was adjusted to obtain the maximal direct motor response M_max (approximately 15 mA) and the maximal H response H_max (approximately 7 mA). The maximal T reflex (T_max) was obtained as follows: supramaximal mechanical stimuli were delivered by an electromagnetic hammer, placed 1 mm from the Achilles tendon. A pressure sensor was attached to the hammer. The output signal of this sensor was used to determine the time of contact with the tendon and also to measure the reflex latency. The amplitude of the pressure sensor signal was taken as a measure of the mechanical stimulus intensity, and only responses to stimuli of the same amplitude were analysed. The EMG signal was controlled on-line by an oscilloscope to apply the electrical or mechanical stimuli only when the rat was inactive. The EMG data were digitised over 30 ms (sampling frequency 50 kHz). For each leg testing, approximately 30 responses to electrical stimulation and 20 to mechanical stimulation were recorded. H_max and T_max responses were defined, respectively, as the mean value of the ten highest H and T responses stored. The M_max response was defined as the highest direct motor response obtained during the experiment. H_max and T_max responses were normalised with respect to the M_max response. Because the T reflex amplitude reflects both activation of the muscle spindle and activation of the motoneuron, whereas H reflex amplitude depends only on motoneuronal activation, the T_max/H_max ratio was also calculated and used as an index of the relative adaptations of the peripheral and central components of the reflex pathway.

These experiments were always conducted in the morning, as it is known that the rat H reflex differs between the morning and the afternoon (Chen and Wolpaw, 1994). Left and right legs were tested on the same day, with a 45 min resting period between the two tests. Rats from the HS group were tested immediately after the suspension device had been removed.

As soon as the reflex responses had been recorded, the rat was anaesthetised using intraperitoneal injection of urethane (0.12 g 100 g⁻¹ body mass). The mechanical study was limited to the soleus muscle, as it is the only muscle among the triceps surae to possess a proximal tendon that is easily removable without muscle damage. Furthermore, it is known that after HS the gastrocnemius shows little atrophy (Simard and Lacaille, 1988) except for fibres in its red part (Gardetto et al., 1989). The soleus muscle was carefully isolated and placed in a chamber containing oxygenated (95 % O₂ and 5 % CO₂, pH 7.3) physiological salt solution at a temperature of 25 °C.

Biomechanical experiments were performed using a servo-controlled ergometer, described in detail elsewhere (Lensel-Corbeil and Goubel, 1989). The proximal tendon was tied to a displacement transducer using a stainless-steel wire and the distal tendon was similarly tied to a force transducer (compliance 4×10^–5 m N⁻¹). The muscle was held at its reference length (L₀), i.e. the length at which isometric twitch tension (P_t) was maximal, in response to a single supramaximal stimulus. The maximal tetanic tension (P₀) was obtained using a 80 Hz stimulus train of 1 ms pulse duration. From the plateau value, reached around 800 ms after the beginning of the stimulation, an initial release (ΔL) was performed at very high velocity (30 cm s⁻¹), inducing a decrease in tension. Tension/extension curves, characterising the series elastic component (SEC), were obtained by plotting the shortening (ΔL/L₀) versus the tension (P/P₀) for ten releases of different amplitudes ranging from 0.01 to 0.16 cm. The normalised data were fitted to a third-order polynomial function, and the normalised maximal extension value (ΔL_max/L₀) was measured, i.e. the amplitude of release required to obtain zero tension. To complete this classical dual-controlled release, the initial fast release was immediately followed by an isovelocity shortening (V) (Canon and Goubel, 1995; Asmussen and Maréchal, 1989). These controlled releases were repeated for different shortening velocities to establish the hyperbolic force–velocity relationship, characterising the contractile component (CC). A linearised form of this force–velocity relationship was fitted to the data allowing extrapolation to zero force and hence estimation of the maximum shortening velocity, V_max.

Following the mechanical experiments, the soleus muscles were frozen in liquid nitrogen. Serial sections (10 μm) were cut at −20 °C. The myosin adenosine triphosphatase (ATPase) activities of these muscle sections were determined using methodology described by Brooke and Kaiser (1970), using an acidic pre-incubation (pH 4.2). The microscopic observations of the sections were analysed using a camera linked to a microcomputer. A specific software package was used to count the number of positively (black) and negatively (white) stained fibres, respectively classified as type I (slow-twitch) fibres and type II (fast-twitch) fibres, and to evaluate semi-automatically the mean cross-sectional area of each type of fibre and the total fibre cross-sectional area (CSA). This CSA was taken as an index of muscle section to calculate the isometric stress.

For the two groups of rats, all variables (reflex, mechanical and histochemical) were expressed as the mean value ± S.D. (N=12). The statistical test of Cochran (1964) allowed the assessment of differences between the two populations for a level of significance of either 5 % or 1 %.

All control and suspended rats accepted the electrical and mechanical stimuli without trying to escape the experiment and without any aggressive reaction.

M_max responses were not affected by growth and by HS since the two groups of rats presented slight non-significant changes (P>0.05) in M_max responses between the 5-and 9-week-old rats: M_max for the C group was 30.1±7.9 mV at 5 weeks and 34.5±9.5 mV at 9 weeks (N=12); M_max for the HS group was 27.7±6.6 mV at 5 weeks and 25.4±5.5 mV at 9 weeks (N=12).

The mean H_max/M_max ratio calculated for both age groups was significantly higher (P<0.05) in the 9-week-old rats than at the start of the experiment. However, the mean of individual gains for the H_max/M_max ratios was higher for the HS group (+120 %) than for the C group (+50 %) (Fig. 1).

The H_max/M_max ratio calculated after denervation of the soleus muscle fell drastically to 43 % of pre-surgical value (P<0.01, N=6).

For the C group, the mean T_max/M_max ratios calculated during both tests did not differ significantly (mean of individual gains: +23 %, P>0.05). However after HS, the reflex responses to a tendon tap decreased, and the mean T_max/ M_max ratios were significantly lower in 9-week-old rats than in 5-week-old rats (mean of individual gains: −24 %, P<0.05) (Fig. 2).

Furthermore, the T_max/H_max ratios calculated for each leg were found either to increase or to decrease for the C group, as illustrated by both negative and positive changes (Fig. 3A). Thus, the mean of individual gains of the T_max/H_max ratios calculated for the C group was not changed (−12 %, P>0.05) between 5 and 9 weeks. Every rat of the HS group exhibited a decrease in the T_max/H_max ratio between 5 and 9 weeks (Fig. 3B), and the mean of individual changes was significantly negative (−62 %, P<0.01).

The contractile and elastic characteristics of the soleus muscles of the C and HS groups are summarised in Table 1. These results highlight significant differences between the two populations in the amplitudes of twitch tension (−60 %, P<0.01), tetanic tension (−77 %, P<0.01) and of isometric stress, P₀/CSA, (−34 %, P<0.01). These lower values of tensions are in accord with the muscular atrophy induced by disuse (the mean soleus muscle masses were 171±8.9 mg (N=12) and 65.5±10.2 mg (N=12) for the C and HS groups respectively; mean rat masses at 9 weeks were 391.5±29.0 g and 277.5±27.0 g).

Taking V_max as representative of the shortening velocity of the contractile component, the significantly higher values of V_max for the HS rats (increased by 74.5 %, P<0.01) attested to an increase in muscle contractile speed.

Taking ΔL_max/L₀ as representative of the characteristics of the series elastic component, these values were significantly higher for the HS than for the C group (increased by 23 %, P<0.01), indicating a decrease in the stiffness of the series elastic component.

The soleus muscle atrophy was confirmed by the quantification of the mean cross-sectional areas for the type I fibres, 2385±209 μm² for the C group (N=12) and 1190±131 μm² for the HS group (N=12) and the type II fibres, 2317±366 μm² for the C group (N=12) and 1299±295 μm² for the HS group (N=12). This atrophy was accompanied by a fibre-type transition phenomenon. The percentage of type I fibres was 80.3±9.9 % for the C group and 47.1±4.6 % for the HS group. The percentage of type II fibres was 19.7±9.9 % for the C group and 52.8±4.7 % for the HS group. Thus, in the HS group, the soleus muscle had a significantly higher proportion of type II fibres than type I fibres compared with the C group.

The reflex excitability, mechanical properties and fibre content of rat muscles submitted to HS were analysed to test whether paired changes in reflex and muscular variables could be observed after a period of disuse, as occurs following a period of training.

The present study confirmed for the soleus muscle the classical effects of HS: atrophy, an increase in the proportion of fast fibres (Desplanches et al., 1987), decreases in twitch and tetanic tensions, decrease in the isometric stress, increase in V_max (Diffee et al., 1991; Fitts et al., 1986) and an increase in the stiffness of the SEC (Canon and Goubel, 1995). We expected these mechanical and morphological changes to be accompanied by neural adaptations.

Muscles atrophied by HS should present lower M_max responses, as classically observed in disused muscles (Duchateau and Hainaut, 1990). However M_max responses did not differ significantly for suspended and control rats. This result could be explained if the EMGs recorded activity from both the superficial gastrocnemius and the deeper soleus muscles. The small amount of atrophy of the gastrocnemius muscle after HS could explain why the amplitude of M_max responses was not significantly modified.

The second expected neural change was a decrease in reflex excitability from muscles enhanced in fast, less excitable, fibres (see Introduction). Such a decrease was observed, but only when considering the T reflex. The increase in H reflexes observed for both groups of rats could reflect a maturation process since the rats were immature at the first test. Young children also present lower H_max/M_max ratios than adult subjects (Vecchierini-Blineau and Guilheneuc, 1981).

HS had a particular effect on the changes in H response since the H_max/M_max ratio increased more significantly for the HS group than for the C group without any significant decrease in direct motor response. This result indicates that the H reflex changes concern mainly the weight-bearing soleus muscle, even though EMG activity was detected in both the gastrocnemius lateralis and soleus muscles. Other arguments can be given to confirm that H reflexes originate mainly from the soleus muscle. First, in humans, the fast gastrocnemius muscle demonstrates a significantly smaller H response than the soleus muscle (Mongia, 1972). More generally, it seems that the H reflex is greater when the muscle slow fibre content is higher (Pérot et al., 1991). Furthermore, the complementary experiment showed that the denervation of the soleus muscle induced a drastic fall of the H_max/M_max ratio.

The significant increase in H reflex amplitude after HS was not compatible with our initial hypothesis of a decrease in motoneuronal excitability in soleus muscles with a higher content of fast fibres. This result indicates that hypoactivity induced by HS initiates changes in synaptic efficiency, as is the case in other types of disuse programme. Such an increase in Ia fibre efficacy on the presynaptic side due to disuse has been reported in humans (Gallego et al., 1979) and was also suspected to be responsible for the higher H_max/M_max ratio calculated in man after a long period of bed-rest (Duchateau, 1995) or space flight (Rüegg et al., 1997). Of the two factors responsible for H reflex amplitude (synaptic efficiency and motoneuronal excitability; Delwaide, 1973), changes in synaptic efficiency seem to predominate in all types of disuse and interfere with the eventual decrease in motoneuronal excitability suspected for muscles with a higher fast-fibre content.

Despite the proposed increase in synaptic efficiency, the T responses were strongly depressed after HS. Since T and H reflexes of approximately the same amplitude involve the same motor units (Van Boxtel, 1986), these reversed T and H changes in the HS group cannot be caused by the motoneuronal component. The systematic negative changes in the T_max/H_max ratios calculated for the HS rats suggest that the peripheral components of the stretch reflex pathway are largely responsible for the changes in the T response. These peripheral changes could be occurring at the level of the muscle spindles. Jozsa et al. (1988) have reported structural changes in the intrafusal fibres as a result of disuse by immobilisation. This change, a decrease in intrafusal fibre diameter, could modify the spindle activation. The sensitivity of muscle spindles could be changed after HS, as is the case following chronic de-efferentation (Arutiunian, 1981).

The observed decrease in the stiffness of the SEC (Table 1; Canon and Goubel, 1995) and also of tendinous elements (Almeida-Silveira et al., 1996b) will affect the T response. This decrease in stiffness implies a lower spindle solicitation after mechanical stimulation (tendon tap), as the spindle experiences a lower degree of muscle stretch if the mechanical stimulus is applied to a more compliant tendon (Rack et al., 1983).

In conclusion, HS was found to produce mechanical and histochemical changes typical of a slow-to-fast fibre-type transformation. This increase in the proportion of fast fibres was not paired with a decrease in H reflex amplitude, as observed after a period of training for explosive exercise. Muscles disused through HS exhibited higher H reflexes than the controls, a result explicable in terms of an increase in synaptic efficiency. When the reflex pathway was solicited by a mechanical stimulus, a decrease in reflex excitability was observed after HS and paired with a decrease in the stiffness of the SEC, leading to lower spindle solicitations. This result illustrates the importance of the mechanical properties of the musculo–tendinous elements in determining the reflex response when mechanically evoked.

We thank Professor Francis Goubel for his constructive criticisms on this manuscript.