The rodent HU model was originally devised to investigate the astronauts' musculoskeletal response to weightlessness or low-gravity conditions and has since been widely used as a muscle-wasting model mimicking the condition of muscle-wasting disease, inactivity, bed rest, and immobilization (Table 2).[23] Since weightlessness has been predicted to yield deficits in the principal tissues needed for structure and movement on Earth, primarily muscle and bone, the National Aeronautics and Space Administration (NASA) Ames Research Center (ARC) set up the HU model to study the mechanisms, responses, and treatments for the adverse consequences of spaceflight in the mid-1970s. After inception of the HU model at NASA, many laboratories have used the HU model to simulate weightlessness and subsequently used it as a muscle-wasting model. Since the standard operating procedure for applying the HU model to young and adult rodents was updated and approved by the NASA ARC Institutional Animal Care and Use Committee on August 8, 2001, more than 1,500 papers have published data that used this model system.[24]

The primitive HU model tested in 1975 was very simple, but led to suggestions for modifications ultimately incorporated into succeeding designs. The first HU model described in a full-length paper in 1979 used a hexcelite back harness and a cantilevered rotating beam that allowed the head-down animal to move in a 360° arc.[25] Data were compared with the weight and bone changes found in the Cosmos 782 and 936 biological satellites, and the author concluded that the changes of body weight, food consumption, and bone-formation rates in HU rats were very similar to spaceflight. So the HU model closely mimics results from rat and man exposed to near-weightlessness during orbital spaceflight and will allow preliminary answers to questions posed by spaceflight experiments.

Musacchia and colleagues used a modified first HU model.[26] In this model, they used a denim harness and a rotating beam that allowed the animal to move in a 140° arc. Body weight and food consumption in the HU rats were significantly less than those of the control group, and the animals exhibited adrenal hypertrophy at the end of the 7-day experiment. The authors concluded that the muscle changes were similar to those found during spaceflight and recovery from spaceflight. To find out whether the cephalad fluid shift contributed to changes in metabolism, Deavers and colleagues included a horizontal control and found that the head-down position was required for the diuresis and natriuresis that occurred during HU.[27] Stump and colleague [28] modified the model to measure muscle changes and blood flow. Deavers et al. [27] and Bouzeghrane et al. [29] advocated for a horizontal control, particularly for studies investigating fluid shifts. Hargens and colleagues [30] addressed the importance of the unloading angle. They found that the angle of unloading determined the amount of weight supported by the forelimbs as well as the tension applied to the tail, and showed that the HU rat applies 50% of its body weight to its forelimbs when the angle between the torso and the floor of the cage is 30°. As the angle increased, mechanical loading of the forelimbs declined and traction on the tail increased. If the angle was too steep, then the animals appeared stressed. A 30° angle of unloading was recommended, because it provided normal weight bearing on the forelimbs, unloaded the lumbar vertebrae but not the cervical vertebrae,[31] and induced a cephalad fluid shift.[30]

The HU model has not changed conceptually from the beginning. Data from all laboratories that used the model showed differential muscle atrophy, a cephalad fluid shift, animals having the freedom to move, eat, and groom with the forelimbs, and unloading of the hindlimbs without paralysis so that animals could recover from unloading. However, the harness system and degree of mobility differed significantly between laboratories. One problem possibly related to these differences was the reduced weight gain in growing rats or weight loss in adult animals that persisted throughout the experimental period. Harnesses tested at ARC included a combination of elastic and Velcro straps and hexcelite bonded with an epoxy resin to the back of the rat.[25] Each of these harnesses was only partially successful, and less stressful harness designs were sought. The concept of a tail harness originated with our Russian colleagues, who used a plaster of Paris mold for tail traction. In the early 1980s, orthopedic surgeons from the University of Southern California toured our laboratory and recommended that the tail cast be replaced with the tape that they used for placing human limbs in traction. Traction tape could be applied to an unanesthetized animal and would allow the tail to grow without restriction. In fact, the body weights of growing rats unloaded with the use of tail traction remained comparable to those of controls fed the same amount of food (i.e., group-mean-fed controls), in contrast to rats unloaded with the use of back harnesses. Tail traction appears to be less stressful to animals than are whole-body harnesses, as assessed by corticosterone levels and adrenal, thymus, and body weights.[32]

Unlike other rat models, the HU model did not require confinement of animals in small cages, limb casting, or flaccid paralysis by nerve section or surgical tenotomy. None of these techniques produced the differential muscle atrophy characteristic of spaceflight, i.e., a decreased in mass of the extensor muscles but not of other muscles associated with movement. In addition, recovery from disuse was difficult or impossible with the existing surgical models. But the many studies that used the HU model clearly showed that independent variables can influence results obtained and, ultimately, the analog's validity in terms of understanding the mechanisms and physiological responses to spaceflight. Additional HU variables known to influence experimental results include age (growing vs. adult), sex, species (rat vs. mouse), and strain.

Chronic hindlimb suspension (HS) (unweighting) has been shown to limit growth and result in significant losses in hindlimb muscle mass, slow-twitch properties, and show myosin content.[33,34,35,36,37] This loss in muscle mass is more extensive in those muscles predominantly composed of slow-twitch (type I) fibers.[33,35,36,38] Although weight-bearing activity (mechanical stress) appears to be a primary factor in maintaining muscle weight in the context of HS,[37] evidence is lacking as to whether anabolic steroid treatment can serve as an independent stimulus to preserve muscle weight in the absence of weight-bearing activity.

It is thought that complex degeneration of the neuromuscular system contributes to dynapenia.[2,39,40,41,42] Neuromuscular changes contributing to myofiber denervation occur within the central and peripheral nervous systems as well as within skeletal muscle tissue. Changes include diminished function or loss of neurons in the brain and spinal cord, demyelination of nerves, and progressive degeneration of NMJs.[43,44] Skeletal muscle denervation, caused by such problems as traumatic peripheral nerve injury, disease, pharmacologic intervention, and aging (Table 3), diminishes the function leads to immediate muscle atrophy.[14,45,46] Early muscle atrophy could be restored by a timely and appropriate reinnervation occurrence, but without one, myofiber atrophy progresses to irreversible changes in the muscle with muscle fibrosis and myofiber death.[47,48] Denervation is a common phenomenon in an aged NMJ. The tibial- or sciatic-nerve transection model to induce the denervation is commonly employed and a well-validated model in rodents. Only a single dose of analgesic is necessary in the immediate postoperative period. With the use of proper sterile technique, soft-tissue infection is rare.[49] This model allows the investigator to use genetically engineered mice to study the process of muscle atrophy in vivo in the absence of proteins crucial to the regulation of muscle mass.[50,51]

The tibial-nerve transection model is a validated, reproducible, and well-tolerated model of denervation-induced skeletal muscle atrophy in rodents, and is used to study the physiologic, cellular, and molecular biologic mechanisms that underlie muscle atrophy in vivo in the gastrocnemius and soleus muscle. The tibial nerve is a mixed motor-sensory peripheral nerve in the rodent hindlimb and is 1 of the 3-terminal branches of the sciatic nerve. Transection of the tibial nerve denervates the gastrocnemius, soleus, and plantaris muscles (and the 3 small deep flexor muscles of the foot, including the tibialis posterior, flexor digitorum longus, and flexor hallicus longus), and is a well-standardized and validated model in rats.[52,53] Also, various knockout (KO) and transgenic (Tg) mice allow us to assess the specific functions of proteins in the induction, development, and maintenance, or alternatively the resolution of muscle atrophy and fibrosis in vivo in this model. The tibial nerve supplies the gastrocnemius, soleus, and plantaris muscles, so its transection permits the study of denervated skeletal muscle composed of fast twitch (type II) fibers and/or slow twitch (type I) fibers. The gastrocnemius muscle is a mixed-fiber muscle (type I and type II, although predominantly type II), and the soleus muscle is composed of a large proportion of type I fibers, thereby providing both fast- and slow-twitch muscles for assessment.[54,55] The tibial-nerve transection model is suitable for studying the process of denervation-induced muscle atrophy in both the short term (days) [50] and long term (weeks to months).[51,52]

The gastrocnemius and soles muscles, both deserved in denervation model, can be easily and rapidly dissected with minimal handling, thus providing excellent-quality mRNA and protein for subsequent molecular analyses. Similarly, because of the size of the muscles, they can be split, providing tissue from the same animal for concomitant histologic and morphometric analyses. If hindlimb functional assessment is required, walking-track analysis can be serially done. The feet are dipped in ink, and the mouse is walked through an enclosure with paper on the bottom. Characteristics of the prints can be reliably measured and scored to indicate the extent of neuromuscular disability and gait compromise, since footprint characteristics reflect the functional muscle groups.[56,57] Although originally developed and validated in rats,[56] walking-track analysis can also be used with mice.[58]

The tibial nerve of only one hindlimb is transected, and since mice bear weight almost equally on both hindlimbs, the musculature from the contralateral un-operated limb can be used as an internal control within each animal.[50,51,59,60] This is not necessarily the case in the sciatic transection model, where more significant abnormalities of gait can induce a hypertrophic response in the contralateral limb muscle. In the tibial-nerve transection model, we typically use the gastrocnemius and soleus muscle from the un-operated limb as our control muscle.[50,51] If the investigator chooses to use separate animals from which to harvest control muscle, then sham surgery should be performed. Sham surgery would consist of the administration of anesthesia, splitting of the skin to expose the tibial nerve, but no transection. The skin would simply be closed following nerve exposure.

Although tibial-nerve transection does induce sensory paraesthesia on the plantar aspect of the foot, the mice must be inspected daily for signs of auto-mutilation, heel pressure ulcers, or point-of-care endpoints. Although we have negligible mortality with the model, we find that approximately 2% to 5% of mice must be euthanized because of self-inflicted injury to, or pressure ulcers developing on, the operated hind limb. Sciatic-nerve ligation as well as the spared nerve injury model of ligation (where the tibial and common peroneal branches of the sciatic are ligated, but the sural is left intact) serve as models of neuropathic pain.[61,62] Thus, allodynia and thermal hyperalgesia could occur in the foot in our model as well, but we have not seen overt pain behavior in the mice with normal daily activity on soft bedding.

One of the leading theories on mechanisms underlying age-related muscle denervation points to oxidative stress.[63,64,65] Reactive oxygen species (ROS) are natural byproducts of mitochondrial activity involved in respiration and energy production. ROS-mediated oxidative damages to DNAs, proteins, and lipids are normally kept in check by antioxidants. However, excessive ROS production can overwhelm the antioxidant defense, leading to increased oxidative damage of cellular machinery. Two mouse models, one lacking the Cu/Zn superoxide dismutase (Sod1) gene and another harboring the Tg mutant human SOD1 gene, display progressive changes at the NMJ, including muscle-endplate fragmentation, nerve-terminal sprouting, and denervation. These changes at the NMJ share many of the common features observed in the NMJs of aged mice.[66] SOD1 is a cytoplasmic antioxidant enzyme involved in the scavenging of superoxide free radicals. Mice lacking the SOD1 enzyme (Sod1^-/- mice) show increased oxidative damages to proteins, lipids, and DNAs.[67] In addition, these mice display progressive muscle denervation, weakness, and loss, changes seen despite the absence of a spinal cord motor neuron and ventral root axon loss.[68,69,70] NMJ denervation and sprouting are observed in these mice at between one and 4 months of age and precede muscle loss, [69,70] which is observed at between 3 and 4 months of age.[67] Furthermore, muscle denervation and loss are greater in the gastrocnemius and tibialis anterior compared to the soleus.[67,69,70] The G93A SOD1 mouse line Tg for SOD1 containing a point mutation at amino acid position 93 (G −> A) present in patients with familial amyotrophic lateral sclerosis. G93A SOD1 mice recapitulate many of the pathological hallmarks of amyotrophic lateral sclerosis, such as progressive muscle weakness and denervation, motor neuron loss, and paralysis.[71,72,73] It has been demonstrated that muscle denervation is observed as early as at 47 days of age in these mice [74,75,76,77] and precedes both motor neuron loss,[62,63] and muscle atrophy.[78,79] These characteristics are similar to those observed in the rodent models of aging [14,80] as well as in Sod1^-/- mice.[69,70]

Immobilization-induced skeletal muscle atrophy is characterized by a decrease in muscle mass and an increase in the risk of debilitating diseases and orthopedic problems (Table 4). The cast immobilization is the most frequently used model for studying muscle atrophy because it simulates conditions after fractures that require casts and wrap the leg with a plaster bandage or spiral wire; so this model can mimic prolonged immobilization.[12] The cast immobilization model may prove useful in studies on therapeutic interventions of muscle atrophy using Tg and mutant mouse strains.[81] This model could also evaluate the muscle loss, because muscles that are fixed in a contracted state show greater atrophy than do stretched muscle.[82] However, this model is time consuming, needs some skill, and may cause adverse events, including skin injury, local edema or necrosis, probably because of the retention of urine by the cast, and problems of escaping from the cast; so it requires experience, frequent observation, and replacement.[83,84,85] There are several methods for construction of a cast immobilization model. The traditional method is using the plaster cast on a unilateral or bilateral hindlimb.[86,87,88,89,90,91] After anesthesia, one or both hindlimbs was immobilized with a plaster cast and monitored on a daily basis for chewed plaster cast, abrasions, and problems with ambulation. Some modified methods have also been developed. Onda and colleagues used steel bonsai wire, which enables repeated direct access to the immobilized muscle and allows concurrent application and assessment of various therapeutic interventions.[92] In this study, the weight of the soleus and planters muscles were decreased significantly in both bilateral and unilateral immobilization and the mRNA expression of Fbxo32 (MAFbx [also known as atropin-1] protein coding gene) and Trim63 (MuRF1 protein coding gene) were also decreased in both muscles.[92] Another group developed a Velcro immobilization method using the commercially available hook-and-loop fastener that is faster and has less adverse events than does cast immobilization. They insist that Velcro immobilization could substitute for cast immobilization and allow the immobilization-intervention process to be repeated easily.[85] Speecht and colleagues explored disuse-induced muscle atrophy by using a unilateral casting model in conjunction with HS. They showed in the study that a 2-week HS resulted in a significant decrease in gastrocnemius and quadriceps weight of about 9% to 10%, but over a 2-fold greater decrease in HS with casted limb.[93]