VCAM‐1 upregulation accompanies muscle remodeling following resistance‐type exercise in Snell dwarf (Pit1dw/dw) mice

Abstract Snell dwarf mice (Pit1dw/dw) exhibit deficiencies in growth hormone, prolactin, and thyroid stimulating hormone. Besides being an experimental model of hypopituitarism, these mice are long‐lived (>40% lifespan extension) and utilized as a model of slowed/delayed aging. Whether this longevity is accompanied by a compromised quality of life in terms of muscular performance has not yet been characterized. In this study, we investigated nontrained and trained muscles 1 month following a general validated resistance‐type exercise protocol in 3‐month‐old Snell dwarf mice and control littermates. Nontrained Snell dwarf gastrocnemius muscles exhibited a 1.3‐fold greater muscle mass to body weight ratio than control values although muscle quality, maximum isometric torque normalized to muscle mass, and fatigue recovery were compromised. For control mice, training increased isometric torque (17%) without altering muscle mass. For Snell dwarf mice, isometric torque was unaltered by training despite decreased muscle mass that rendered muscle mass to body weight ratio comparable to control values. Muscle quality and fatigue recovery improved twofold and threefold, respectively, for Snell dwarf mice. This accompanied a fourfold increase in levels of vascular cell adhesion molecule‐1 (VCAM‐1), a mediator of progenitor cell recruitment, and muscle remodeling in the form of increased number of central nuclei, additional muscle fibers per unit area, and altered fiber type distribution. These results reveal a trade‐off between muscle quality and longevity in the context of anterior pituitary hormone deficiency and that resistance‐type training can diminish this trade‐off by improving muscle quality concomitant with VCAM‐1 upregulation and muscle remodeling.

with a recessive mutation in Pit1 (Pou1f1), an anterior pituitary transcriptional factor (Snell, 1929). PIT1 is among the most common genes mutated in genetic cases of patients with combined pituitary hormone deficiencies with more than 30 distinct PIT1 mutations identified to date (Stieg, Renner, Stalla, & Kopczak, 2017;Takagi et al., 2017). Such hypopituitarism results as a consequence of compromised anterior pituitary development, and clinical manifestation includes several various forms of muted development overall and short stature (Lee et al., 2011). Studies regarding the Pit1 mutation in Snell dwarf mice have also demonstrated remarkable lifespan extension of >40% relative to littermates (Flurkey, Papaconstantinou, & Harrison, 2002;Flurkey, Papaconstantinou, Miller, & Harrison, 2001). Concomitant with this longevity is an apparent delay in aging in terms of data regarding T-cell function, collagen cross-linking, incidence of cataracts, resistance to cancer, and kidney disease (Alderman et al., 2009;Flurkey et al., 2001;Vergara, Smith-Wheelock, Harper, Sigler, & Miller, 2004).
Despite enduring scientific interest in Snell dwarf mice, whether hypopituitary-induced longevity in these mice compromises quality of life has not been fully tested especially in regard to skeletal muscle performance. Such data, in a limited manner, have been addressed for the closely related long-lived Ames Dwarf mutant (Prop1 df/df ) mice which also exhibit deficiencies in growth hormone, thyrotropin, and prolactin (Brown-Borg, Borg, Meliska, & Bartke, 1996). For these mice, percent lean body mass values were determined to be comparable (at 2 months of age) or increased (4.5-6 months of age) relative to agematched controls (Heiman, Tinsley, Mattison, Hauck, & Bartke, 2003).
In addition, in a wire hang test administered at middle age to old age (19-to 30-month-old), Ames dwarf mice were able to maintain grip for a longer duration than age-matched controls demonstrating prevention of age-related neuromusculoskeletal decline in a distinct test (Arum, Rasche, Rickman, & Bartke, 2013). Although these studies provide some insight into the effects of anterior pituitary hormone deficiency, further evaluation of skeletal muscle mass specifically and muscle performance under high activation especially at young age is required for a more complete characterization. Furthermore, the response to mechanical loading in the form of resistance-type exercise training has not been tested.
The purpose of this study was to investigate skeletal muscle of young (3-month-old) Snell dwarf mice and their littermate controls at the onset and completion of 1 month of resistance training with stretch-shortening contractions (SSCs), contractions typical during resistance-type exercise training (Vaczi et al., 2014(Vaczi et al., , 2011. The SSC protocol consisted of 80 maximally activated SSCs (8 sets with 10 repetitions per set) and has been validated repeatedly to increase performance for muscles of rats and mice Rader et al., 2016;Rader, Naimo, Ensey, & Baker, 2017). Because of the potential for opposing training-induced results for agonist vs. antagonist muscles, the agonist gastrocnemius and antagonist tibialis anterior (TA) muscles were evaluated . In conclusion, to determine whether muscle remodeling involving vascular endothelial growth factor (VEGF) and VCAM-1, a mediator of progenitor cell recruitment following exercise, was possibly influential in the response, levels of these proteins were assessed (Stromberg, Rullman, Jansson, & Gustafsson, 2017). The results provide further insight into the role of anterior pituitary hormones on muscle mass and quality in nontrained and resistance-type trained agonist muscles and the potential influence of VCAM-1 as compensatory. The findings also confirm the concern regarding antagonist muscle atrophy following isolated muscle training for specific muscles. 2.77 ± 0.03 mg/mm, p < 0.0001) and 38% of control values for GTN muscles (Figure 1c). The decreased muscle mass in the nontrained state for Snell dwarf mice was attributable to decreased muscle fiber size (Figures 2c and 3a-c). When expressed per gram body weight, nontrained normalized GTN muscle mass was elevated for Snell dwarf mice demonstrating that the muscles of Snell dwarf mice were large for their size (Figure 1d). This finding was confirmed for nontrained TA muscles with normalized muscle mass per gram body weight values for Snell dwarf vs. control mice of 13.1 ± 4.0 mg mm −1 g −1 vs. 9.0 ± 2.2 mg mm −1 g −1 , respectively, p < 0.0001. However, nontrained muscles of Snell dwarf mice were weak in terms of both plantarflexion peak dynamic torque (1.5 ± 0.2 mN-m vs. 13.9 ± 0.8 mN-m, Snell dwarf vs. control, p < 0.0001) and maximum isometric torque, even after expressing per gram body weight (Figure 1e,f). Therefore, plantarflexor muscle quality of nontrained Snell dwarf muscle was 25% that of controls ( Figure 1g). Fatigue measures were also assessed in the nontrained state, and a compromised recovery from fatigue was observed for Snell dwarf mice (Supporting Information Figure S1A-E).

| Plantarflexion training induced an exceptional improvement in Snell dwarf plantarflexor muscle quality and fatigue recovery
For control mice, plantarflexion SSC training did not alter GTN muscle mass (Figure 1b-d). Peak dynamic torque (13.9 ± 0.8 mN-m to 17.0 ± 0.8 mN-m, p = 0.002), maximum isometric torque, and muscle quality increased by 20% (Figure 1e Table 1). For Snell dwarf mice, absolute performance measures-peak dynamic torque (1.5 ± 0.2 mN-m and 1.8 ± 0.1 mN-m, nontrained vs. trained) and maximum isometric torque-were maintained following plantarflexion SSC training (Figure 1e,f). It is an interesting fact this maintenance of performance occurred despite a decrease in GTN muscle mass so that normalized muscle mass per gram body weight values became comparable to control values (Figure 1b-d). Therefore, muscle quality increased exceptionally (twofold) for Snell dwarf mice (Figure 1h). For both groups of mice, SSC training improved the absolute torque values generated during the SSC session without increasing fatigue rate (Supporting Information Figure S1A-D). Furthermore, fatigue recovery was enhanced by training especially for Snell dwarf mice-a threefold increase (Supporting Information Figure S1E). These training-induced performance outcomes in Snell dwarf mice were accompanied by alterations in muscle composition apparent with hematoxylin and eosin staining ( Figure 2). Quantitative morphology analysis demonstrated an increase in GTN muscle interstitium (   mice and, unlike the agonist GTN muscle, no indication of increased muscle fiber number was observed (Figure 3d-f). Protein levels for six cytokines were investigated, interferon gamma, interleukin-6, interleukin-10, interleukin-12, interleukin-17, and tumor necrosis factor alpha, and no training-induced change was observed (Table 1 and   Supporting Information Table S1). These results suggest that the TA atrophy was not accompanied by a response in these cytokines.

| Training-induced alterations in VCAM-1 and fiber type distribution indicative of robust GTN muscle remodeling correlated with improved performance for Snell dwarf mice
To investigate further the remodeling apparent in trained GTN muscles especially for Snell dwarf mice, the protein levels of VEGF and VCAM-1 were evaluated. For nontrained muscles, VEGF was elevated twofold for Snell dwarf mice relative to control mice (Figure 4a). Training had no effect on VEGF levels. However, training had a pronounced effect (i.e., fourfold increase) on VCAM-1 levels specifically for Snell dwarf mice ( Figure 4b). Immunofluorescence labeling for laminin, nuclei, and VCAM-1 was undertaken to determine the distribution of VCAM-1 ( Figure 5). The twofold increase in the number of muscle fibers per unit area observed for trained Snell dwarf muscle was accompanied by a comparable increase in a number of nodes (i.e., anatomical features encircled by laminin and adjacent to muscle fibers-characteristics indicative of capillaries) per unit area (Figure 6a,b). Furthermore, these increases in muscle fibers and nodes coincided with increases in the incidence of VCAM-1 + staining in these features (Figure 6c,d). Close inspection of the immunostaining revealed that VCAM-1 staining tended to circle or F I G U R E 2 Transverse sections of nontrained and plantarflexion SSC-trained muscles stained with hematoxylin and eosin for control (a-d) and Snell dwarf mice (e-h). Scale bar = 50 µm cluster adjacent to nuclei within muscle fibers, nodes, and the interstitium ( Figure 5). Therefore, an analysis of the nuclei was undertaken which demonstrated an overall increase in nuclei number per unit area for trained Snell dwarf mice in the various tissue compartments analyzed (Table 2). Concomitant with this training-induced increase in nuclei number was an increase in percentage of VCAM-1 staining associated with these nuclei (Table 2).
Stretch-shortening contraction training-induced GTN muscle remodeling for Snell dwarf mice also extended to fiber type alterations consistent with a shift to a more oxidative phenotype as apparent in immunofluorescence labeling (Supporting Information Figure S2). Type IIb fibers decreased in cross-sectional area to 40% of nontrained area and percentage of muscle tissue composed of type IIb fibers decreased to 55% of control value (Supporting Information Figure S3A,C). Meanwhile, the number of type IIx fibers per unit area increased by sevenfold and percentage of IIx muscle fibers (relative to a total number of muscle fibers) increased by fourfold (Supporting Information Figure S3C,D). This altered fiber type distribution from type IIb to IIx fibers with training was confirmed by chi-square analysis (Supporting Information Figure S3E,F). This SSC-induced remodeling at the fiber type level for Snell dwarf mice was coordinated with VCAM-1 distribution as suggested by the correlation between these factors (Supporting Information Fig. S4). Both high percentages of VCAM-1 + muscle fibers and nodes were accompanied by a high percentage of type IIx fibers (Supporting Information Figure S4F,G). This extensive muscle remodeling evident by VCAM-1 and fiber type distribution correlated with the improved fatigue recovery gained by Snell dwarf muscle following training (Supporting Information Figure S4H,I,J). Pearson product correlation analysis was also performed for Snell dwarf muscle quality vs. percentage VCAM-1 + nodes (r = 0.584, p = 0.046), type IIx muscle fibers (r = 0.507, p = 0.092), and VCAM-1 + muscle fibers (r = 0.485, p = 0.11). Overall, the findings indicated that the Snell dwarf muscle tissue remodeling impacted performance in general.  Although the training was beneficial in regard to muscle quality and fatigue recovery capacity for Snell dwarf mice, responsiveness was muted for absolute performance measures, maximum isometric torque, and peak dynamic torque. This demonstrated the trade-off between growth vs. longevity with anterior pituitary hormone deficiency Sharples et al., 2015). Such a trade-off is consistent with reports regarding IGF-1, a key hormone regulated by growth hormone. Individuals with IGF1 variants associated with high IGF-1 blood levels gain more strength following resistance training (Hand et al., 2007;Kostek et al., 2005). Yet, genetic polymorphisms in the gene for IGF-1 receptor, which result in decreased IGF-1 plasma levels, associate with longevity (Bonafe et al., 2003;Suh et al., 2008). Although the responsiveness to resistance-type exercise training may be muted, absolute performance gains may still be possible as evident by research regarding growth hormone deficient adults (Werlang Coelho et al., 2002). As demonstrated with previous research regarding aging, careful design and refinement of resistance-type exercise training consisting of SSCs has great potential for benefiting muscle even for muscle in an anabolic-compromised state Rader et al., 2016). Exercise prescription must also account for effects on surrounding muscles as evident by

| DISCUSSION
The SSC-induced increase in muscle fibers nodes per unit area and for Snell dwarf GTN muscles was accompanied by increased VCAM-1 + staining. Total number of (a) muscle fibers and (b)  the finding of 35% atrophy for the antagonist TA muscles of both control and Snell dwarf mice following training in the present study.
The observation of antagonist muscle atrophy confirmed an earlier report regarding the same plantarflexion SSC training to C57BL/6 mice . This response was attributed to a possible neuromuscular imbalance imposed on antagonist non-weight-bearing muscles such as the TA muscle when training is isolated to weightbearing muscles . The present study establishes that this training-induced muscle imbalance is independent of anterior pituitary hormone levels. Overall, our findings support the further refinement and utilization of balanced resistance-type exercise training for benefiting skeletal muscle in cases of hypopituitarism.

| Animals
The young (

| Plantarflexion SSC training
Plantarflexor muscles were exposed to SSC training based on a previous procedure demonstrated to induce muscle mass and performance gains in young C57BL/6 J mice . For each training session, the mouse was anesthetized with isoflurane gas, placed in dorsal recumbency on a heated table, and the left foot secured to a footplate of a dual mode muscle lever system (Whole Mouse Test System, 1300A; Aurora Scientific). Platinum electrodes were placed subcutaneously to activate the tibial nerve, and muscle stimulation was set at parameters (8-V magnitude, 0.2-ms pulse duration, and 150-Hz frequency) for maximal contraction. Prior to the 80 SSC training, static and dynamic performance was assessed.
A single maximal isometric contraction with the ankle at 90°(angle between tibia and foot) was utilized for static performance, and a single SSC test was utilized for dynamic performance consisting of an isometric contraction for 200 ms at a 110°ankle angle followed by rotation to 70°at 500°per second and back to 110°at 500°per second while continuing activation for 200 ms.
At 2 min following the single SSC test, the training session was administered. Each training session consisted of 8 sets (2-min intervals between sets) with 10 SSCs per set (3-s intervals between SSCs). For each SSC, the muscles were maximally activated at ankle angle 90°for 100 ms, rotated to 70°at 60°/s, returned to 90°at 60°/s, and deactivated 100 ms later. Chronic exposure to SSCs with a velocity of 60°/s induces adaptation without overt muscle inflammation and degeneration in young wild-type rodents days to weeks into training (Baker, Hollander, Kashon, & Cutlip, 2010;Baker, Hollander, Mercer, Kashon, & Cutlip, 2008;Cutlip et al., 2006;. At 5 min following the 80 SSCs, a maximal isometric tetanic contraction was measured and compared with pre-80 SSC exposure values to determine recovery from fatigue

| Immunofluorescence
Transverse cryosections of gastrocnemius (GTN) muscles were analyzed for myosin heavy chain (MHC) staining using a previously described method . Sections were blocked (10% P36931; ThermoFisher Scientific). The investigator was blinded to section identification for image capture. Midpoint of the muscle section was identified, and nonoverlapping images were captured at the sites of most area in both the lateral and medial regions of the muscle section. An overlay graticule (with 0.04 mm 2 square boundary) was placed at the center of each image, and the investigator counted various features provided the topmost region of each feature resided within the graticule boundary. Features included muscle fibers, nodes, and nuclei. A muscle fiber was counted as VCAM-1 + if any VCAM-1 staining was apparent within the muscle fiber laminin border. Anatomical features encircled by laminin and adjacent to muscle fibers, characteristics indicative of capillaries, were classified as nodes and considered VCAM-1 + when any VCAM-1 staining was observed within the laminin border. Nuclei were counted within the sarcoplasm (central or peripheral), nodes, and interstitial regions and considered associated with VCAM-1 when any VCAM-1 staining colocalized with or was directly adjacent to DAPI staining. A number of muscle fibers, nodes, and nuclei were normalized by total muscle section area sampled.
After centrifugation at 1,500 rcf for 15 min at 4°C, the supernatant was collected for ELISA analysis for cytokines (Aushon Ciraplex Cytokine 1 Array Kit; #107-17F-1-AB) and growth factors (Aushon Ciraplex Custom Array Kit for VCAM-1 and VEGF, #100-0286) per standard kit instructions. Images of the arrays were taken using an Aushon Cirascan Imaging System. Total protein was determined using a standard colorimetric bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL, USA).

| Statistical analysis
Data were analyzed using ANOVA (JMP version 11; SAS Institute, Inc., Cary, NC, USA) with the variable of animal identification as a random factor to account for repeated measures when appropriate.
Post hoc comparisons were performed using Fisher's least significant difference method. Correlations were assessed by Pearson product correlation analysis (SigmaPlot version 12.5; Systat Software, Inc., San Jose, CA, USA). Chi-square analysis (SigmaPlot version 12.5) was utilized to determine training-induced differences in absolute frequency distributions of fiber type. With the exception of frequency distribution data, all data are expressed as means ± standard error.
p < 0.05 was considered statistically significant.

ACKNOWLEDG MENTS
This study was supported by internal National Institute for Occupational Safety and Health funds. The funding body did not have a role in the design of the study, collection, analysis, interpretation of data, and writing of the manuscript.

CONFLI CT OF INTEREST
The authors declare no conflict of interest.

PUBLI CATION DISCLAIMERS
The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention.

AUTHOR CONTRIBU TI ONS
E.P.R. and B.A.B. designed the study, interpreted the data, and wrote the manuscript; E.P.R., M.A.N., J.E., and B.A.B. performed the experiments. All authors approved the final version of the manuscript.