Satellite Cells and Neuromuscular Spindles in Long Term Denervated Skeletal Muscles

Satellite cells (SC) are quiescent cell located between the sarcolemma and basal lamina of the skeletal muscle bers. The SC can get activated contributing to regeneration and/or growth of muscle. The neuromuscular spindles are mechanoreceptors located within the skeletal muscle and are considered as contractile regulatory unit. It is composed of intrafusal muscle bers (IF), surrounded by a sheath and is parallel to extrafusal bers. Denervation cause changes in skeletal muscles both in the SC and neuromuscular spindles. This study analyzed quantitatively the IF and SC in Wistar rats denervated for long period. The animals were divided into normal and denervated groups. The soleus and extensor longus digitorum longus were denervated experimentally during periods of 0, 12, 16, 19, 30 and 38 weeks. The percentage of SC immediately after denervation increases when compared to normal group and later decreases in both the groups. During the process of denervation, there was an increase in IF when compared with normal group. The percentage of SC reduces signicantly between the periods of denervation in both the groups. The smaller percentage of SC corresponds to higher number of IF. Besides that the number of SC decreases after denervation. As for IF, with the increase in time in normal group, the number of bers was unaltered. However, in the experimental group, with increase in the time of denervation, the percentage of SC decreases while there is increase in the number of IF signicantly. In denervated muscles for long period, there is decrease in the percentage of SC and increase in IF. Our results suggest that the period between 16th and 19th week post denervation is the best time for reinnervation of denervated muscle.


Introduction
The skeletal muscle has been the subject of studies and research in various different elds of science.
Various types of pathologies may affect this tissue and there has been increasing interest in recovery of their structural and functional organization in the scienti c world.
Skeletal muscle development is a highly coordinated process that involves the myogenesis and differentiation of primary myoblasts. The skeletal muscle exhibits satellite cells. The satellite cells are small myogenic cells in a quiescent state located between the sarcolemma and basal lamina of muscle bers [1,2].
It is postulated that these cells are protagonists in the process of reconstitution of the injured muscle tissue.
Disruption of the sarcolemma induce the migration of these cells to the injured area where they can blend into the muscle bers which are still viable or position themselves in the tissue precursor cells (myotube) differentiating into muscle cells. The fusion of satellite cells with the consequent transfer of its nucleus allows an increase in protein synthesis thereby increasing the chances of recovery of this cell. On the other hand, the differentiation in myotubes contributes to the emergence of new bers and ultimately recovering the injured area [12][13][14][15].
Neuromuscular spindles are sensory receptors, which are sensitive to stretch and tension. The modi cation of the muscle bers in relation to the tension and length are the main physical parameters of muscle activity, which are perceivable through the proprioceptive system located deep inside the muscles and tendons [16].
The neuromuscular spindles are mechanoreceptors that are situated among the muscle bers. These are the main sensory organ of the muscle belly, also considered as the regulatory contractile unit of muscle and is composed of intrafusal bers (IF), surrounded by a sheath of connective tissue, containing liquid in its inside and are parallel to the extrafusal ber (EF). The spindle is connected to the extrafusal bers, so whenever the muscle is stretched, elongation of the spindle also takes place. The process of excitation of the neuromuscular spindle occurs when a stimulus of elongation is applied [17].
Previous studies have shown that postural muscles contain a high proportion of neuromuscular spindles [18]. Furthermore, the neuromuscular spindles are responsible for monitoring the speed and duration of stretching and detect changes in the length of the muscle. These bers are sensitive to the velocity with which a muscle is stretched [19].
The absolute number of spindles in a muscle depends to some extent on the size of the muscle; for example, the absolute numbers of spindles in the rst lumbrical and latissimus dorsi muscles in humans are 51 and 368, respectively [20]. Therefore, in order to compare the relative richness of muscles in spindle content, the mass of muscle must be taken into account [20,21].
Methods to observe the structural behavior of skeletal muscle have been used in last few decades, especially the techniques of denervation. Several studies on this regard, such as morphological and morphometric analysis of skeletal muscles at various post-denervation periods, observations of proliferation of myonuclei and satellite cells and apoptosis of muscle bers have been conducted [22][23][24].
In skeletal muscle, the denervation causes in increased expression of various genes that are involved in a variety of aspects of autophagy [25]. Furthermore, denervation causes changes, which can be observed by the decrease in trophism along with increase in adipose and connective tissue in the affected muscle [26].
When there is damage to skeletal muscle, a series of processes takes place resulting in tissue repair. Muscle injury can lead to in ammation, activation of satellite cells, myogenesis, proliferation of broblasts and reorganization of connective tissue and extracellular matrix [26].
Slow-twitch muscles (red), like soleus for instance, when subjected to denervation produces greater muscle atrophy than in fast-twitch muscles (white), such as the Extensor digitorum longus (EDL) [4]. This atrophy due to denervation affects both the slow and fast muscle bers, resulting in reduction of the muscle ber diameter and muscle strength [26].
Some authors have reported a signi cant increase of satellite cells after denervation [27,28], however, other studies have demonstrated that immediately after denervation, satellite cell population does not seem to change [29]. In order to collaborate in understanding the phenomena resulting from denervation in striated muscles with different morphological characteristics (soleus and EDL), denervation of these muscles in rats were performed for a long period of time, analyzing the percentage of satellite cells and the number of intrafusal bers. Further this study also evaluates the best time period for reinnervation of muscle after experimental denervation. Initially, trichotomy of the right lateral thigh of each animal was performed. Following this, an incision between the ischial tuberosity and the greater trochanter was made to expose the sciatic nerve.
Approximately 10 mm of the sciatic nerve was excised from its middle third. The proximal stump of the sciatic nerve was ligated, bent backwards and sutured to the adjacent muscles to prevent possible reinnervation.
All the animals which were going to be sacri ced after 4 months of surgery were re-operated at least every three months and the proximal stump was sectioned again and sutured to the adjacent muscle as described above thus aiming to prevent reinnervation of the muscles to be analyzed.
After surgery, all the animals were placed in plastic cages with a maximum of ve animals per cage, with food and water "ad libitum" with light / dark cycles of 12 hours each, with controlled temperature and humidity.
In all the normal and experimental groups, muscles were dissected and removed after time intervals of 0, 12, 16, 19, 30 and 38 weeks for histological analysis.

Collection of samples
After the experimental period the animals were again anesthetized as described above and the hind limbs were perfused via the abdominal aorta with Ringer's solution containing 1% procaine hydrochloride and 5,000 units / liter of heparin followed by 2.5% glutaraldehyde in 0.1 M phosphate buffer.
The soleus and EDL muscles were dissected, removed and were subsequently post xed with osmium tetroxide and embedded in 1% epoxy resin (Araldite®).

Histological preparation and data collection
The muscles collected were xed in 10% formalin for 24 hours. The samples were then washed for 24 hours in running water and were stored in 70% alcohol for more 24 hours. For inclusion in historesin, the samples were immersed in 95% alcohol for 2 hours followed by another 4 hours in a mixture of 95% alcohol and historesin. After this period the material was nally immersed for 24 hours in historesin liquid. Subsequently, the samples were placed in molds and the historesin was catalyzed to enable hardening of the blocks.

Satellite Cells
Sections for electron microscopy observation were obtained from the same blocks, collected on suitable mesh grids and stained with uranyl acetate and lead citrate. Each grid has approximately 300 square-shaped holes. Were used covering Sect. 30-50 of the grid holes and made of paper in a schematic section in such low magni cation. The holes were covered by Section numbered. After that, at high magni cation the myonuclei and satellite cell nuclei were counted. All were considered myonuclei ber core within myo laments while presenting cell nuclei were taken of all the satellites within cell nuclei located beneath the basal lamina of the muscle ber, or more precisely between the sarcolemma and the basal lamina of the striated muscle bers. Serial sections were obtained at least 20 µm in distance to avoid the possibility of counting the same myonuclues twice. An area containing at least 400 myonuclei was investigated for each muscle and the percentage of SC was calculated.

Neuromuscular spindles
For analysis of the neuromuscular spindles, transverse semi-thin sections of 1.3 µm thickness were made from the middle third of each muscle and were stained with parafenildiamine (ppd). The cross sections were then assessed with the help of an Olympus® BX-50 microscope equipped with a camera (Olympus® DP-71).
Photographs from each sample were taken and were analyzed with the help of computer software (Image Pro-Plus® 6.0). Intrafusal bers from spindles of all the animals were counted and an average was taken. All muscles with signs of reinnervation such as the presence of myelinated axons or groups of myo bers incompatible with dimensions were discarded.

Statistical analysis
Descriptive statistics (mean, standard deviation) were calculated for all variables using the statistical package SPSS® (version 13.7, SPSS Inc.).
For the satellite cells and intrafusal bers counts in both Soleus and EDL, we performed a comparative analysis between the groups (Normal control and Experimental) and periods of denervation using analysis of variance with two criteria and the Tukey Test. For all analysis, value p < 0.001 were considered signi cant For correlation between the number of satellite cells, intrafusal bers and the periods of denervation, we used the Pearson correlation coe cient. For all analysis, value p < 0.05 were considered signi cant.

Satellite cells
The percentage of satellite cells in relation to myonuclei decreases with age in normal muscles, and on average there is a higher incidence of satellite cells in Soleus muscle (Fig. 3A) compared to the EDL muscle (Fig. 3B). It is also observed that the percentage of satellite cells in muscles immediately after denervation increases compared to normal muscle, especially in the EDL muscle and then decreases in both muscles from the 12th week of denervation ( Fig. 3A and 3B). Another interesting observation is that during the 16th weeks of denervation of the experimental group, the percentage of satellite cells is similar to the percentage of satellite cells in the normal group, in both muscles studied (Fig. 3A and 3B). This was not seen during other time periods.

Neuromuscular spindles
The neuromuscular spindles in normal groups have on average four to ve intrafusal bers in the Soleus muscle; however the number in the denervated group increases 16 week after surgery (Fig. 3C). In the EDL muscle such change could be observed only 19 weeks after surgery (Fig. 3D).
This increase in the number of intrafusal bers was directly proportional to the duration of denervation, ie during the progression of the denervation period of both the Soleus and EDL muscles, there was an increase in the number of intrafusal bers, compared to the normal group ( Fig. 3C and 3D).

Statistical Analysis: Comparative (Tukey test)
Satellite cells: Statistical analysis comparing the number of satellite cells between the periods of denervation in both the experimental and normal group in the soleus muscle shows a signi cant difference between most periods. No signi cant difference was seen between the periods: Normal 0 X 12 Denervated, Normal 16 X 19 Normal, Denervated 16 X 30 Normal, Denervated 19 X Normal 38 (Table 1 and Fig. 3A).

SC (satellite cells), IF (intrafusal bers)
In the EDL muscle, statistically signi cant differences were seen between all the groups except between Denervated 16, Normal 19 and Normal 30, where the percentage of satellite cells were in close proximity to each other. Comparison between the rests of the groups showed statistically signi cant results (Table 1 and

Neuromuscular spindles
Through statistical analysis it can be observed that in Soleus muscle, there is no signi cant difference in the number of Intrafusal bers between the different intervals of the normal group (Table 1 and Fig. 3C).
While in the denervated group, there was no statistically signi cant difference between the groups: Denervated 0 X 12 Denervated, Denervated 16 X 19 Denervated and Denervated 30 X 38 Denervated (Table 1 and Fig. 3C).
However, when comparing the normal group with the experimental, statistically signi cant difference between all periods were seen except between Normal 0 X 0 Denervated (Table 1 and Fig. 3C).
In EDL muscle, the statistical analysis shows similar results to that of the soleus muscle, ie, no signi cant difference in the number of Intrafusal bers between the different intervals of the normal group were seen (Table 1 and Fig. 3D).
Meanwhile in the denervated group, statistically signi cant difference was not seen between the groups Denervated 16 and 19 (Table 1 and Fig. 3D).
When compared with the normal group, there were no statistically signi cant differences between: Normal 0 X 0 Denervated and Denervated 12 X 12 Normal (Table 1 and Fig. 3D).

Statistical Analysis: Correlation (Pearson Test)
Correlating of the percentage of satellite cells with the number of intrafusal bers, there was a statistically signi cant negative correlation to both muscles examined (soleus and EDL) only in the denervated group, ie, the more reduced the number of satellite cells is, greater the number of IF ( Table 2, Fig. 4C and 4D).

SC (satellite cells), IF (intrafusal bers)
In the control group it was found a negative statistical correlation between denervation time and the number of satellite cells and intrafusal bers in both of studied muscles. It means that when the denervation time is increased the SC number decreases. On the other hand when the denervation time increases the number of IF remains the same (Table 3, Fig. 4A and 4B).

SC (satellite cells), IF (intrafusal bers)
Furthermore, in the experimental group when compared with the time of denervation, satellite cells demonstrated a negative correlation while the correlation was positive with intrafusal bers (Table 4, Fig. 4C and 4D).

Discussion
In order to determine the most appropriate period to attempt recovery of a nerve and the corresponding innervated muscle, it is important to consider the changes that occur in both of them.
In nerve bers which have undergone axotomy, it is known that the production of Nerve Growth Factor (NGF) reaches a maximum level after 24 hours and high levels of NGF are maintained at least two weeks after injury. The denervated muscle bers however suffer atrophy and gradually lose tissue mass [22].
In muscle tissue denervated for a longer period of time i.e over 25 weeks, there is a signi cant reduction in the number of satellite cells [22]. Therefore muscles with a longer period of denervation do not present conditions for a possible reinnervation.
This was supported by another study [30] who conducted denervation of the pectoralis muscle of frog and observed, after nearly four years, the complete absence of muscle bers.
Studies have been conducted evaluating the optimum time period for reinnervation in muscles with sectioned motor nerves. Some ndings have suggested that the best time to receive new contact of motor nerve it would be between 12 and 16 weeks after surgery [23]. However, when there is no tissue loss, the faster the nerve repair attempt, better is the regeneration [23].
Classically, attempts of reinnervation of muscle would be performed as soon as possible. However on analyzing denervated muscles, muscle bers are seen morphologically with characteristics of young bers only after 12 weeks of denervation. The nerves play an important role in the modulation of such muscle bers by changing its constitution or by modulating the different ber types to better reach the functional demands of the muscle. Therefore in observing the natural development of striated skeletal muscles, a delay of several weeks may be more accurate than the immediate attempt of reinnervation, in particular if there is nerve tissue loss.
Recently, studies in rats have revealed that the gradual increase in muscle activity after targeted muscle reinnervation (TMR) takes place within 4 weeks. In comparison to normal muscle, the electromyography activity of the reinnervated muscle indicates an innerveration between the transferred nerve and the targeted muscle within four weeks after TMR surgery. In cases of amputees after TMR surgery, the muscle needs a long time to reinnervate and heal [31].
In this study, we conducted experimental lesions in rat sciatic nerve (with tissue loss) without recovery. The muscles chosen for morphological analysis of its recovery were the Soleus and EDL. The soleus muscle is considered as a postural muscle, which is rich in red bers (type SO and FOG), and therefore more homogeneous, while EDL on the other hand is rich in white bers (FG).
The present study demonstrated that the percentage of satellite cells decreases in soleus and EDL after denervation. The initial reaction in the EDL muscle after denervation is an increase in the percentage of satellite cells for a short period and then continues to decrease to near zero after 30 weeks. The increase in the percentage of satellite cells in soleus muscle also occurred in a short period of time after surgery.
The satellite cell population represents considerable potential for the postnatal growth of skeletal muscle [32] and the amount of satellite cells is dependent on nerve or muscle activity mediated by innervation. Their percentage is normal in reinnervated and regenerated muscles [33], but declines rapidly during regeneration deprived of innervation [34]. The events that occur in a denervated muscle can be described as slow and not synchronous [30,34,35]. Such events also affect the intrafusal bers, where there is increase in their number [36].
We believe that the poor recovery of muscle strength after a signi cant delay in the surgical repair of the motor nerve is not only due to prevention of axonal regeneration through endoneurium, but also due to the decline of satellite cells in the muscle while it remained denervated for a long period.
Furthermore, our results show that the percentage of satellite cells in denervated muscle is similar to the percentage of satellite cells in normal muscle between the 16th and 19th week post-surgery.
From the results obtained in this research and with the aid of the literature it may be appropriate to infer that the optimum time period for reinnervation of a denervated muscle is between 16th and 19th weeks after surgery. During this time, the denervated muscle contains a similar amount of satellite cells to normal muscle, ultimately favoring muscle recovery.
In this study, the muscles in all the normal groups at the 38th week showed on an average 4 to 5 Intrafusal ber. However, in the denervated groups of both the soleus and EDL muscles there is a statistically signi cant increase in the number of intrafusal bers, directly proportional to the experimental time.
However, studies in humans on age related changes of the proprioceptive system with focus on muscle spindle showed a loss of intrafusal bers per neuromuscular spindle which theoretically may be related to its own degeneration process [37].
Furthermore, the authors also suggest that these aforementioned changes related to aging are connected to an increased thickness of the connective capsule of the neuromuscular spindle. This is contrary to what we observed in our work, where only in the denervated groups; the connective capsule of neuromuscular spindles was altered showing a thickening of tissue around the intrafusal bers.
This study demonstrates that there is an increase in the number of intrafusal bers in the denervated group over a long period of time (38 weeks), where this increase is inversely proportional to the percentage of satellite cells. This correlation is considered statistically signi cant, suggesting that the increase of intrafusal bers may be an attempt to maintain the muscular proprioceptive function.
We emphasize that this research can contribute to other studies on recovery of an injured muscle, as it suggests that the optimum period of reinnervation of a denervated muscle is between 16th and 19th week after denervation, during which the percentage of the satellite cells is similar that of the normal muscles. The results also demonstrate that with the increase in the period of denervation, the number of intrafusal bers increases, suggesting that even during degeneration of muscle, intrafusal bers increase in number in order to maintain proprioceptive and sensory functions.
In summary, further research is necessary to complement our ndings in developing an ideal time frame for reinnervation for recovery of injured muscle and motor nerves.