Forelimb unloading impairs glenohumeral muscle development in growing rats

Proper joint loading is essential for healthy musculoskeletal development. Many pediatric neuromuscular disorders cause irreversible muscle impairments resulting from both physiological changes and mechanical unloading of the joint. While previous studies have examined the effects of hindlimb unloading on musculoskeletal development in the lower limb, none have examined solely forelimb unloading. Thus, a large deficit in knowledge of the effect of upper limb unloading exists and must be addressed in order to better understand how the glenohumeral joint adapts during development. Two forelimb unloading models were developed to study the effects of varying degrees of unloading on the glenohumeral joint in growing rats: forelimb suspension (n=6, intervention 21 days post-natal) with complete unloading of both limbs via a novel suspension system and forearm amputation (n=8, intervention 3-6 days post-natal) with decreased loading and limb use in one limb after below-elbow amputation. After 8 weeks of unloading, changes in muscle architecture and composition were examined in ten muscles surrounding the shoulder. Results were compared to control rats from a previous study (n=8). Both methods of altered loading significantly affected muscle mass, sarcomere length, and optimal muscle length compared to control rats, with the biceps long head and triceps long head observing the most marked differences. Forearm amputation also significantly affected muscle mass, sarcomere length, and optimal muscle length in the affected limb relative to the contralateral limb. Muscle composition, assessed by collagen content, remained unchanged in all groups. This study demonstrated that forearm amputation, which was administered closer to birth, had greater effects on muscle than forelimb suspension, which was administered a few weeks later than amputation.


Introduction
four muscles (biceps long head, biceps short head, upper and lower subscapularis) were harvested 115 bilaterally for composition analysis. The proximal end of each muscle was embedded in optimum 116 cutting temperature compound and set in 2-methylbutane cooled by liquid nitrogen, and the entire 117 muscle was then snap frozen and stored at -80°C until sectioning. Muscle mass and muscle length were measured for the muscles stored at 4°C. After blotting 121 excess ethanol, muscles were weighed on a digital scale (resolution of 0.01 g). For each muscle, 9 122 muscle fibers were extracted, 3 each from the proximal, middle, and distal regions of the muscle.

123
Sarcomere lengths were measured via a 5.0-mW HeNe laser with a wavelength of 633 nm 124 (Thorlabs, Newton, NJ) using an established laser diffraction method 26 . All muscle lengths and 125 distances between each diffraction band were measured using digital calipers (resolution of 0.01 126 mm). The 9 sarcomere measurements were averaged to find the mean sarcomere length for each 127 muscle. To determine the excursion capacity of the muscles and account for possible stretch in the 128 fixed muscle as indicated by sarcomere length, optimal muscle length was calculated 40 : where is muscle length and is sarcomere length. The optimal sarcomere length corresponded with Masson's trichrome (American MasterTech, Lodi, CA) to identify collagen I deposition, a 138 measure of fibrosis and muscle stiffening, and imaged at 20X magnification with light microscopy 139 (EVOS ® FL Cell Imaging System, Thermo Scientific, Halethorpe, MD) In three sections per 140 muscle, collagen content was calculated as the ratio of collagen area to muscle tissue area using a 141 custom image processing protocol (MATLAB ® , The MathWorks, Inc., Natick, MA). To verify whether side-to-side differences were insignificant for the forelimb suspension 145 and control groups (as expected) and to identify whether differences existed in muscle metrics 146 between the affected and unaffected forelimbs for the amputation group, paired t-tests were used.

147
Muscle architecture (mass, sarcomere length, optimal muscle length) and composition ( Qualitative analysis of histologic images revealed minimal differences across the groups 202 in collagen staining for the 4 analyzed muscles (biceps long head, biceps short head, upper and 203 lower subscapularis muscles) (Fig. 6). Quantitative analysis of these images showed that the ratio 204 of collagen area to total muscle area did not differ significantly across the three groups for any of 205 the 4 muscles examined (Table 2).

208
Unloading with the two models had different effects on the growth of forelimb muscles.

209
The suspension group did not affect muscle mass oroptimal muscle lengths relative to control, 210 except for an increased optimal length for biceps short head. In contrast, the amputation 211 intervention led to lower muscle mass and optimal muscle length for several muscles in the 212 affected forelimb compared to the contralateral limb, suggesting that unloading via forearm 213 amputation during postnatal development can inhibit muscle growth. Specifically, muscle mass 214 was lower in the acromiodeltoid, biceps long head, and triceps long head, and optimal muscle 215 length was shorter in the acromiodeltoid and biceps long head following amputation. The 216 amputated group also had lower muscle mass and shorter optimal muscle length in the affected 217 limb compared to the forelimb suspension (right affected limb) and control (unaffected limb) 218 groups. On average, the amputated biceps and triceps long head muscles were approximately half 219 the mass and 75% of the optimal length of the corresponding muscles in the suspended and control 220 groups.

221
The amputation procedure provides an explanation for the specific affected muscles in this 222 group. The biceps long head, biceps short head, and triceps long head originate at the scapula and 223 insert to the proximal radius or ulna. During the amputation procedure, the severing at the insertion 224 point releases the muscles and causes widespread denervation and atrophy, leading to reduced 225 muscle mass 52 . In other studies found that denervated extensor digitorum longus muscle mass in 226 growing rats increased after initial atrophy, following similar growth patterns as the control 227 contralateral limbs, but soleus muscle mass decreased relative to the control 53 . The authors 228 suggested that the increased growth was due to elevated protein synthesis after continued lengthening of the muscle, while the decrease in growth was attributed to a reduction in protein 230 synthesis after continued shortening of the muscle. Although the biceps short head was denervated, 231 it likely experienced extended periods of lengthening, causing it to grow similar to the suspension 232 and control groups. The biceps and triceps likely experienced shortening over the duration of the 233 study due to the release at amputation, which contributed to muscle mass loss and shortening. The 234 other forelimb muscles were not affected by the amputation procedure and therefore there was no 235 marked differences in muscle architecture.

236
The suspension group also exhibited changes in muscle architecture relative to the control 237 group, which may be explained by the relative immobilization of the limbs. For example, 238 immobilization in innervated lower limb muscles in growing rats found that a decrease in muscle 239 mass compared to a control was attributed to higher levels of protein breakdown and reduced 240 protein synthesis in the affected muscles when the muscles were held in a shortened position 54 .

241
When held in a lengthened position, immobilized muscles in the lower limbs of growing rats 242 exhibited slightly increased muscle mass compared to the control, which was attributed to the anteriodeltoid, spinodeltoid, subscapularis, supraspinatus, and teres major were markedly 263 different in the unloading groups, optimal muscle lengths for these muscles remained the same 264 compared to the control group. Since optimal muscle length is a ratio of sarcomere length to 265 measured muscle length, the unchanged optimal muscle length across groups is likely due to 266 similar changes in sarcomere and optimal muscle lengths. Based on this, the biceps long head and 267 triceps long head, which displayed remarkably lower muscle mass and longer sarcomere lengths 268 that translated to shorter optimal muscle lengths, experienced the greatest decrease in muscle-269 tendon force production across the board.

270
These results are consistent with a previous study that investigated changes in muscle 271 architecture in growing rats after neonatal injury to the brachial plexus nerve 51 . When comparing 272 the affected limb to the contralateral limb, muscle mass in the same ten muscles as in this study 273 was significantly less in all but one observed muscle, including the biceps long head and triceps 274 long head, similar to the amputation group in this study. Concurrent to the previous study, sarcomeres in the amputation group were significantly longer in the teres major and biceps long 276 head, along with the teres major in the suspension group relative to the control group. Unlike the 277 injury groups, the suspension group, however, did exhibit shorter sarcomeres in the biceps long 278 head compared to the control group. This comparison shows that muscle mass and sarcomere 279 length in the amputation group more closely mimicked those of injury groups seen in the literature.

280
Optimal muscle length was shorter in the biceps long head for both unloading groups, and biceps 281 short head for the suspension group, which showed that the suspension group more closely 282 resembled the injury groups seen in literature. The triceps long head was significantly affected by 283 both unloading methods, but not by nerve injury, which could mean that the triceps long head 284 muscle length is more sensitive to changes in loading than denervation.

292
Results from previous unloading studies vary, depending on animal age and method of 293 unloading. Aprevious study investigating the effects of zero gravity on muscle in 3-month old mice 294 found that mass in three leg muscles were not significantly affected by 30 days in space where 295 observed grooming rate was high 9 . Since the mice maintained daily grooming activity, the muscles 296 were activated throughout unloading, and these results are similar to our forelimb unloading 297 condition with limited muscle effects. Previous hindlimb unloading studies reported muscle atrophy and decreased muscle mass . One study examined the effect of 30-day space flight on 19-

299
week old mice and found that hindlimb muscle mass was not significantly affected by 300 weightlessness but trended towards decreased soleus and extensor digitorum longus mass in the 301 unloading groups 9 . This could be close to the cut-off of growing and adult. Another study using a 302 tail-casting hindlimb unloading model in young adult female rats found that the addition of 303 combined isometric, concentric, and eccentric muscle stimulation dampened muscle mass loss 304 compared to the untrained contralateral limb, and muscle mass was unchanged from the regular 305 weight-bearing group 44 . This could help adult rats maintain their muscle mass during unloading.

306
In a partial weight-bearing study, 10-week old adult female mice gastrocnemius muscle was found 307 to be significantly lower mass than that of the control groups 36 . Another hindlimb unloading study 308 with adult male rats found that soleus, plantaris, adductor longus, gastrocnemius, and tibialis 309 anterior muscle mass was significantly reduced after hindlimb unloading compared to typical 310 weight-bearing. Isometric exercise attenuated the effects of unloading in the soleus by 54%.

311
Isometric exercise, however, did not aid the gastrocnemius and plantaris in maintaining muscle 312 mass, as they were significantly less than control by 15% 45 . Hindlimb unweighting was further 313 determined as a cause for reduced muscle mass in adult rats 47 . Soleus muscle mass was 314 significantly reduced in hindlimb unloaded growing rats compared to control rats after 17 days of 315 unloading 46 . This effect was reversed after a 28-day reambulation period. The authors noted that 316 during hindlimb unloading, the ankle was plantarflexed, which caused shortening of the soleus and 317 reduced muscle mass 54 . The mechanism of unloading largely affects muscle properties. If the limb 318 is held in place by a cast, it could be immobilized in a shortened position, which has been shown 319 to have detrimental effects on muscle. In a model in which the unloaded limbs are exposed, they 320 can be held at a natural, optimal position, which may not have as much of an effect of muscle.
the suspended limbs are non-weight bearing, they still experience some non-weight-bearing 323 loading and muscle activity during daily grooming and eating activities. Because muscle mass and 324 optimal muscle length were, for the most part, similar between the suspension and control groups, 325 this small amount of loading seems sufficient to stimulate normal forelimb muscle growth.

326
However, our forearm amputation group experienced both reduced weight bearing and reduced 327 limb use following amputation and were unable to walk on or groom with the amputated limb 328 normally. The affected limb served only as an occasional weight-bearing stabilizer, and forelimb 329 muscle use during these daily activities was greatly reduced. Therefore, the reduced muscle mass 330 observed in this group, compared both to the contralateral limb and to the suspension and control 331 groups, may result from limb disuse rather than direct unloading of the muscles. but the rats progressed normally after the wounds healed. In the future, an additional layer of breathable fabric should be placed between the harness and rat to reduce the amount of chaffing 345 and discomfort over the long unloading period. The suspension system, while removing weight 346 bearing from the forelimbs, did not completely eliminate loading, as the animals were able to 347 continue normal grooming and feeding activities, as noted above. With forearm amputation, 348 because the affected limb experienced reduced weight bearing and overall use, the contralateral 349 limb likely was loaded more throughout the study, potentially augmenting the side differences this a more suitable model to assess isolated muscle effects due to forelimb unloading.       forelimb), and forearm amputation (right forelimb) groups. Mean ± standard deviation. 547 control forelimb suspension forearm amputation biceps long head 5.9 ± 1.3 5.9 ± 0.8 a 5.9 ± 0.0 b biceps short head 6.0 ± 1.3 4.7 ± 0.3 a 7.9 ± 0.0 b upper subscapularis 5.8 ± 2.0 4.1 ± 0.5 5.5 ± 1.7 lower subscapularis 8.1 ± 2.0 a 4.3 ± 1.1 4.4 ± 1.0 a Not all specimens could to be imaged. b Not all specimens could be sectioned.