Animal Models – PLOS Currents Muscular Dystrophy http://currents.plos.org/md Wed, 17 Oct 2018 20:45:27 +0000 en-US hourly 1 https://wordpress.org/?v=4.5.3 Exploratory Profiling of Urine MicroRNAs in the dy2J/dy2J Mouse Model of LAMA2-CMD: Relation to Disease Progression http://currents.plos.org/md/article/exploratory-profiling-of-urine-micrornas-in-the-dy2jdy2j-mouse-model-of-lama2-cmd-relation-to-disease-progression/ http://currents.plos.org/md/article/exploratory-profiling-of-urine-micrornas-in-the-dy2jdy2j-mouse-model-of-lama2-cmd-relation-to-disease-progression/#respond Mon, 27 Aug 2018 12:05:33 +0000 http://currents.plos.org/md/?post_type=article&p=11302 Circulating microRNAs (miRNAs) are being considered as non-invasive biomarkers for disease progression and clinical trials. Congenital muscular dystrophy with deficiency of laminin α2 chain (LAMA2-CMD) is a very severe form of muscular dystrophy, for which no treatment is available. In order to identify LAMA2-CMD biomarkers we have profiled miRNAs in urine from the dy2J/dy2J mouse model of LAMA2-CMD at three distinct time points (representing asymptomatic, initial and established disease). We demonstrate that unique groups of miRNAs are differentially expressed at each time point. We suggest that urine miRNAs can be sensitive biomarkers for different stages of LAMA2-CMD.

]]>

Introduction

Laminin α2 chain-deficient congenital muscular dystrophy, or LAMA2-CMD, is a severe form of muscular dystrophy caused by mutations in the LAMA2 gene. Genotype-phenotype analyses have demonstrated that complete deficiency of laminin α2 chain leads to a more severe phenotype whereas partial absence leads to a milder disease course. The clinical manifestations of complete laminin α2 chain-deficiency include profound hypotonia at birth, widespread muscle weakness, proximal joint contractures, scoliosis and delayed motor milestones. Patients may achieve unsupported sitting but very few children acquire independent ambulation. Individuals with partial deficiency often have later onset of proximal muscle weakness and delayed motor milestones but achieve independent ambulation 1.

There are several mouse models for laminin α2 chain-deficiency that adequately represent the clinical heterogeneity of LAMA2-CMD and phenocopy the skeletal muscle changes. The dy3K/dy3K mouse completely lacks laminin α2 chain and displays a very severe muscular dystrophy with a median survival of three weeks whilst the dy2J/dy2J mouse model has slightly reduced expression and shows a relatively mild muscular dystrophy with a life span of several months. In both models, skeletal muscle is characterized by repeated cycles of degeneration/regeneration and pathological fibrosis. Consequently, dy3K/dy3K and dy2J/dy2J mice weigh less and display impaired skeletal muscle function 2. Diagnosis of LAMA2-CMD and knowledge of underlying pathogenetic mechanisms have greatly improved due to advances in clinical and pre-clinical studies involving LAMA2-CMD patient material and the above mentioned mouse models (as well as other animal models). However, early detection, assessment of disease progression and response to treatment are still major challenges. Hence, it would be important to find novel biomarkers that could facilitate diagnosis and prognosis and aid in evaluating preclinical as well as clinical trial results. The traditional biomarker for muscular dystrophy is creatine kinase (CK), coupled with histological inspection of muscle biopsies. Unfortunately, CK is not very reliable as it is sensitive to age, sex, physical exertion, stress and diet 3, and muscle biopsies are invasive. Therefore, there is an urgent need for more reliable and less invasive biomarkers for LAMA2-CMD and muscular dystrophies in general.

After their discovery in the early nineties, microRNAs (also called miRNAs or miRs) have been extensively studied for their biological roles and biomarker potential. miRNAs are short (18-24 nt) non-coding RNAs that post-transcriptionally regulate protein synthesis by complementary-binding messenger RNA, which leads to degradation of the latter or translation inhibition 4. Their presence in extracellular fluids, such as blood and urine, along with miRNA dysregulation in various diseases, including muscular dystrophy 5, has spurred extensive biomarker research 6,7,8,9,10. Indeed, we have previously demonstrated that laminin α2 chain-deficiency is associated with miRNA dysregulation in skeletal muscle and plasma 11. In this study we aimed at profiling miRNA expression in urine from dy2J/dy2J mice to assess their potential for monitoring disease progression. Three distinct time points (three, four and six weeks of age) were chosen to represent asymptomatic, initial symptoms and established disease, respectively. Here we show that distinct sets of miRNA characterise each time point whilst CK fails to differentiate between them.

Materials and Methods

Ethics Statement

Wild-type and dy2J/dy2J (B6.WK-Lama2dy-2J/J) mice were purchased from Jackson laboratory and bred in the Biomedical Center according to institutional animal care guidelines. Permission was given by the Malmö/Lund (Sweden) ethical committee for animal research (ethical permit number M152-14).

Tissue collection

Three-, four- and six-week-old control and dy2J/dy2J mice (n = 5 per group) were sacrificed by cervical dislocation. Quadriceps muscles were dissected for histology and embedded in paraffin.

Histology and morphometric analyses

Muscle sections were stained with haematoxylin & eosin 12 or picro sirius red/fast green 11. Stained cross-sections were scanned using Aperio’s Scanscope CS2 (with Scanscope console v. 8.2.0.1263) and representative images were created using Aperio software.

Centrally nucleated muscle fibres representing regenerating muscle cells and peripherally nucleated non-regenerating muscle cells were counted in quadriceps femoris using ImageJ software version 1.45i (NIH, Bethesda, MD). The whole area of each muscle cross section was considered.

Creatine kinase assay

Blood was collected from heart puncture and transferred to anticoagulant tubes (EDTA) and centrifuged at 1100 × g for 10 min. at 4°C. Plasma was analysed at Clinical Chemistry Laboratory at Skåne University Hospital. The CK_P_S Cobas method was used to quantify enzyme activity.

Fast green/sirius red quantification

Collagen content was quantified by a colorimetric method as described 13. Briefly, approximately 15 paraffin sections of 15 µm were placed in a plastic tube. Paraffin removal was accomplished by immersing the sections in the following solutions: 5 min. xylene, 5 min. xylene/ethanol (1:1), 5 min. ethanol, 5 min. ethanol/water (1:1), 5 min. water. The sections were then stained with fast green/sirius red for 30 min. (rotating). The tissue was washed with distilled water until excess dye was removed and the solution was clear. One ml of 0.1 N NaOH was added to elute colours. The eluted fraction was analysed at 560 nm and 605 nm to estimate total protein content (fast green) and collagen content (Sirius red), respectively.

Grip strength

Forelimb grip strength was measured on a grip-strength meter (Columbus Instruments, Columbus, OH) as previously described 14. In short, the mouse was held by the base of the tail and allowed to grasp the flat wire mesh of the pull bar with its forepaws. When the mouse got a good grip it was slowly pulled away by its tail until it released the pull bar. Each mouse was allowed to pull the pull bar five times. The two lowest values were rejected and the mean of the three remaining values was counted. Animals were not subjected to any training prior to the experiment.

Urine collection

Individual mice were manually handled on top of a grid placed above a collection plate. The mouse was grabbed by the neck and tail and placed in an upright position. When stressed by the handling the mouse would urinate onto the plate. The urine was pipetted into a tube and stored at -80°C.

Isolation of RNA and RNA sequencing

Total RNA from urine was extracted with Qiagen miRNeasy Mini Kit following the manufacturer’s instructions. One hundred and twenty microliters of urine were pooled from two or three animals into one sample.

Total RNA isolated from urine was sent to the Uppsala Genome Centre for high-throughput sequencing on the IonTorrent platform (ThermoFisher Scientific). The raw sequencing data is deposited in the European Nucleotide Archive under accession number PRJEB23307.

Bioinformatics analyses

Bioinformatics analyses were performed in conjunction with the Bioinformatics Long-term Support (WABI – SciLifeLab). Shortly, raw reads between 18 and 24 nt were kept for further analyses. These were mapped to the mouse hairpin miRNA sequences (miRBase v. 22) using bowtie 15 (v. 1.2, Johns Hopkins University). Read counts were calculated with HTSeq 16 and miRBase annotation (v. 21). Differential expression analysis was performed in the R statistical environment (v. 3.3.2, R Foundation for Statistical Computing) with the Bioconductor package DESeq2 17 (v. 1.12.3), with significance set at adjusted p lower than 5%.

Statistical analysis

The statistical analysis of next-generation sequencing data was done as described above. The remaining analyses were performed in the IPython 18 environment using the SciPy (v. 0.18.1) 19 and statsmodels packages (v. 0.61). Difference between groups was assessed by one-way analysis of variance. Significance was set at the 5% level. Data are presented as mean ± SEM.

Results

Characterisation of disease stages

Experimental overview. The arrows (upper part) indicate the different time points when analyses (lower left) were performed.

Fig. 1: Experimental overview. The arrows (upper part) indicate the different time points when analyses (lower left) were performed.

In order to profile miRNA expression in urine from dy2J/dy2J mice at asymptomatic, initial and established stages of the disease, we first assessed body weight, muscle function, muscle histology, and creatine kinase levels in three-, four- and six-week-old dy2J/dy2J and wild-type (WT) littermate mice (Figure 1). Independent of age, dy2J/dy2J male mice weighed less than wild-type mice. This reduction in body weight was only significant at six weeks of age among female dy2J/dy2J mice (Figure 2A). Grip strength to body weight ratio is an indicator of muscle function that allows comparison of animals with different body weights. It is also an indicator of muscle mass change as skeletal muscle is the main tissue that produces force. At three weeks of age there was no difference in normalised grip strength, indicating no functional decline at this age, which is in accordance with the lack of overt symptoms (Figure 2B). At four and six weeks of age the dy2J/dy2J groups had lower normalised grip strength (Figure 2B). A similar result was found by McKee et al. (2017), where dy2J/dy2J mice had similar normalised forelimb grip-strength to wild-type mice at 3 weeks of age, with a sharp decline thereafter 20.

Fig2

Fig. 2: Male dy /dy mice weight less than WT and display impaired muscle function. A: Body weight for males and females WT and dy /dy ; B: Normalised grip strength at indicated time points. * p < 0.05, ** p < 0.01, *** p < 0.001.

In order to assess if impaired muscle function was reflected by abnormal histology, we inspected haematoxylin-eosin and picro-sirius red/fast green stained sections. We found that the first visual signs of muscle pathology appeared at four of age weeks and two weeks later the diseased phenotype was evident (Figure 3A). The only indication of disease at 3 weeks of age was a low degree of inflammation (data not shown). Central nucleation was measured as an index of muscle regeneration. At three weeks of age we did not observe any signs of regeneration in dy2J/dy2J muscle. However, at subsequent time points there was a dramatic increase in the number of regenerating fibres in dy2J/dy2J muscle (Figure 3B).A hallmark of muscular dystrophies is the progressive replacement of skeletal muscle by fibrous tissue. At three and four weeks of age, there was no difference in collagen content (assessed by picro-sirius red absorbance) between dy2J/dy2J and wild-type muscle (Figure 3, C and D). However, at six weeks of age there was a significant increase in collagen content in dy2J/dy2J muscle (Figure 3, C and D).

CK, a classical biomarker for skeletal muscle disease, was elevated in dy2J/dy2J serum at all time points (Figure 3E). Furthermore, CK levels in dy2J/dy2J serum did not differ between three and six weeks of age (not shown).

Fig3. Progressive muscle deterioration and collagen accumulation as evidenced by histology and CK concentration. A: Haematoxylin and eosin staining; B: Central nucleation quantification (as percentage of fibres with central nuclei); C: Fast green/sirius red staining: collagen is coloured red; D: Quantification of collagen as percentage of total protein; E: CK. Scale bars = 100 μm; * p < 0.05, ** p < 0.01, *** p < 0.001.

Fig. 3: Progressive muscle deterioration and collagen accumulation as evidenced by histology and CK concentration. A: Haematoxylin and eosin staining; B: Central nucleation quantification (as percentage of fibres with central nuclei); C: Fast green/sirius red staining: collagen is coloured red; D: Quantification of collagen as percentage of total protein; E: CK. Scale bars = 100 μm; * p < 0.05, ** p < 0.01, *** p < 0.001. Source: Fig. 3C originally published in: Moreira Soares Oliveira B, Durbeej M, Holmberg J (2017) Absence of microRNA-21 does not reduce muscular dystrophy in mouse models of LAMA2-CMD. PLoS ONE 12(8): e0181950. https://doi.org/10.1371/journal.pone.0181950.

In summary, these data revealed that there are no signs of pathology in three-week-old dy2J/dy2J mice, but subsequently a gradual disease progression occurs. Hence, we decided to profile urine miRNAs in three-, four- and six-week-old animals.

MiRNA profiling in urine

We detected more than 700 miRNAs in mouse urine at each time point: 773 at three weeks, 764 at four weeks and 703 at six weeks of age. Among these, the number of differentially expressed miRNAs also varied. Five, 18 and 17 miRNAs, respectively, were differentially expressed at three, four and six weeks of age (Table 1, supplemental figures). We also found that distinct miRNA profiles were associated with each time point, i.e. there was minimal overlap between differentially expressed miRNAs at three, four and six weeks of age. Only miR-1957a and miR-675 were differentially expressed both at three and six weeks of age, and miR-181 was differentially expressed at four and six weeks of age. However, miRNA-181 was down-regulated at four weeks but up-regulated at six weeks of age (Table 1). Furthermore, we found that myomiRs (miR-1, miR-133 and miR-206) and muscle-enriched miRNAs (miR-181a and miR-486) dominate the differentially expressed miRNA panel at six weeks of age.

Discussion

For the past couple of decades miRNAs have been extensively studied for their biomarker potential. However, most investigations use biopsies or blood samples for this purpose, which are invasive. CK assessment (along with morphological analysis) is often part of the standard diagnostic protocols for muscle wasting diseases. The reliability of this method has long been questioned as CK is responsive to age, sex, stress, physical exertion and diet 3. It is largely a binary analysis, indicating recent muscle damage but not cause or severity. Here, we show that CK measurement cannot be used to distinguish disease severity as it was elevated in dy2J /dy2J mice already from three weeks onwards. Furthermore, CK levels did not differ between three- and six-week-old dy2J/dy2J mice. We also observed higher variability in the diseased group compared to controls, making it difficult to establish a cut-off value. Considering the aforementioned limitations, we have opted for a less invasive intervention and profiled miRNAs in urine from a LAMA2-CMD mouse model at three distinct time points (asymptomatic, initial symptoms and established disease). We found that distinct miRNA profiles are associated with each time point. Specifically, we demonstrated that: 1) Some miRNAs are differentially expressed in urine from dy2J/dy2J mice at three weeks of age although skeletal muscles appear histologically and functionally normal at that age with no obvious signs of muscle regeneration or fibrosis; 2) The highest number of differentially expressed miRNAs is seen in urine from four-week-old dy2J/dy2J mice, a time point when these mice display dystrophic characteristics including increased myofibre regeneration (but no fibrosis) and functional decline, and finally: 3) Differentially expressed miRNAs in urine from six-week-old animals are predominantly myomiRs (i.e. miRNAs that are specific for or enriched in skeletal muscle) corresponding to fully developed muscle pathology with a high degree of muscle fibre regeneration and fibrosis. Thus, we suggest that urine miRNAs can be sensitive biomarkers for different stages of LAMA2-CMD.

Our analysis showed that at three weeks of age the miRNA with the largest fold change is miR-675-3p, followed closely by its -5p counterpart. Both miRNA are derived from exon 1 of the long non-coding RNA H19. H19 is highly expressed during embryonic phases but strongly repressed after birth, with significant expression remaining only in skeletal and cardiac muscle 21,22. H19 acts through miR-675 to influence regeneration and differentiation 21,22,23. miR-15b was down-regulated in myasthenia gravis patients and was found to regulate IL-15 expression in a mouse model of the disease 8. It also stimulated cardiomyocyte apoptosis in response to ischaemia/reperfusion injury 24.

The highest number of differentially expressed miRNAs was found at four weeks of age and it is the only time point with down-regulated miRNAs. It is also the first time point with a differentially expressed muscle-enriched miRNA, i.e. miR-181a-1 and miR-181a-2, both of which are down-regulated at four weeks of age. Mir-181a was previously associated with the degree of muscle wasting following high-risk cardiothoracic surgery, with a high predictive value of 91%, despite some limitations in study design and low sensitivity (56%) 9. Apart from being a myomiR, miR-181a is also one of the mitochondria-associated miRNAs, also called mitomiRs. They bind to the mitochondrial outer membrane to regulate its metabolism, gene expression and function 25. Besides taking part in energy metabolism, mitochondria also have prominent roles in cell longevity and apoptosis. In line with this our group has previously shown that most differentially expressed proteins in dy3K/dy3K (a severely affected LAMA2-CMD mouse model) muscle are coupled to energy and calcium metabolism/signalling 26. The most up-regulated miRNA at four weeks of age was miR-495, which is also up-regulated in various cardiac diseases, including cardiomyopathies associated with muscular dystrophies 27. Another differentially expressed miRNA involved in cardiopathy is miR-154, which is associated with increased fibrosis and reduced apoptosis 28. With an almost 6-fold increased expression in dystrophic muscle, miR-182 is involved in skeletal muscle atrophy 29, myocardial hypertrophy 30, muscle glucose utilisation 31 and the response to hormone replacement therapy in women 32. The most down-regulated miRNA at this time point was miR-155. It was reported to be involved in various processes in skeletal and cardiac muscle, such as pathological cardiac hypertrophy 33, skeletal muscle differentiation 34 and regeneration 35. Moreover, miR-155 regulates macrophage transition from a pro- to an anti-inflammatory state in skeletal muscle 35, which is an important step in muscle regeneration.

At six weeks of age myomiRs dominate the differentially expressed miRNA panel. It may suggest that compensatory mechanisms are at play and degeneration/regeneration cycles are intensified. Our lab has previously shown that miR-1, miR-133 and miR-206 are altered in quadriceps and plasma of dy2J/dy2J and dy3K/dy3K mice 6. Levels of miR-1 and miR-133 changed in opposite directions in muscle and plasma, i.e. both were down-regulated in quadriceps whilst up-regulated in blood plasma. MiR-206 was up-regulated in both muscle and plasma. MiR-1 and miR-133 are involved in differentiation and proliferation, respectively 31,32,33. MiR-206 on the other hand seems to be an important hub in gene networks in skeletal muscle given its involvement in fundamental processes such as muscle cell differentiation 32,34 and regeneration 35. Interestingly, work by Böttger et al. 36 presented evidence that the miR-206/133b cluster is in fact dispensable for skeletal muscle development and regeneration. MiR-486, a muscle-enriched miRNA, is also involved in various processes relevant for LAMA2-CMD. Hitachi et al. 37 showed that myostatin, a well-known negative regulator of muscle mass, acts via miR-486 to regulate the IGF-1/Akt/mTOR pathway; others have found similar results 38,39. MiR-486 also affects myoblast differentiation along with miR-206 34. One of the most interesting findings at six weeks of age is that miR-181a is up-regulated, given that it was down-regulated at four weeks. This makes it an interesting target for further investigation, coupled with its purported role in ageing, inflammation, and muscle and mitochondrial metabolism. Validating its targets in skeletal muscle would provide valuable insight into its function in this tissue.

Despite our interesting findings care must be taken when interpreting NGS results. NGS library preparation is known to induce biases that may favour certain sequences and thus compromise further analyses. For this reason, ideally, biomarkers should be validated with an orthogonal method, such as qPCR for example. We must also bear in mind that the clinical reality is quite different from a laboratory one. Our mice had standardised housing, diet, light-cycle, genetic background, etc., all of which differ amongst patients. Future work with clinical samples will have to deal with much higher data variability. Considering the low incidence of LAMA2-CMD it will be very difficult to run clinical trials with age- and sex-matched subjects. One of our goals was to match disease severity to miRNA profile. In the clinical setting this goal is likely to be hampered by the lack of standardised clinical outcomes for muscular dystrophies, another limitation in the field.

In summary, we were able to follow disease progression in LAMA2-CMD by analysing three distinct time points. Three-week-old dy2J/dy2J muscle appears histologically normal with no functional deficit. Yet, CK is elevated and a few miRNAs are differentially expressed. At four weeks of age, muscles are histologically abnormal and show increased regeneration and functional decline (but no fibrosis). CK is increased and several differentially expressed miRNAs are detected. Finally, six-week-old dy2J/dy2J muscle displays histological and functional impairment along with increased CK and the differentially expressed myomiRs. We would like to propose that miRNAs have the potential to distinguish disease stages and should be further investigated as biomarkers for LAMA2-CMD.

Competing Interests

The authors declare that no competing interests exist.

Data Availability

All data and access information are contained in the article.

Corresponding Author

Bernardo Moreira Soares Oliveira: bernardo.moreira_soares_oliveira@med.lu.se

Appendix

Table 1: Differentially expressed miRNAs at the selected time points. Adjusted p < 0.05 and log2FoldChange > 1.

miRNA log2FoldChange p-adj
mmu-miR-675-3p 4.4063 0.0005 3wk
mmu-miR-675-5p 4.1857 0.0003 3wk
mmu-miR-1957a 2.9201 0.0003 3wk
mmu-miR-15b-5p 2.4532 0.0001 3wk
mmu-miR-320-5p 2.4024 0.0063 3wk
mmu-miR-495-3p 3.8713 0.0052 4wk
mmu-miR-369-3p 3.6786 0.0108 a4wk
mmu-miR-337-3p 3.6545 0.0028 4wk
mmu-miR-154-3p 3.6308 0.0028 4wk
mmu-miR-376b-3p 2.9790 0.0028 4wk
mmu-miR-182-5p 2.7268 0.0031 4wk
mmu-miR-127-3p 2.7131 0.0489 4wk
mmu-miR-148a-3p 2.5092 0.0028 4wk
mmu-miR-31-5p 2.1256 0.0012 4wk
mmu-miR-1839-5p 1.8521 0.0167 4wk
mmu-miR-21a-5p 1.3860 0.0108 4wk
mmu-miR-155-5p -2.4041 0.0389 4wk
mmu-miR-615-3p -2.1658 0.0028 4wk
mmu-miR-204-5p -1.7558 0.0389 4wk
mmu-miR-187-3p -1.7246 0.0477 4wk
mmu-miR-181a-1-5p -1.3911 0.0389 4wk
mmu-miR-181a-2-5p -1.3911 0.0389 4wk
mmu-miR-378c -1.1433 0.0389 4wk
mmu-miR-486a-5p 4.7623 0.0002 6wk
mmu-miR-486b-5p 4.7524 0.0002 6wk
mmu-miR-5108 4.5049 0.0076 6wk
mmu-miR-206-3p 4.1917 0.0031 6wk
mmu-miR-8101 4.0483 0.0127 6wk
mmu-miR-675-5p 3.9570 0.0031 6wk
mmu-miR-133b-3p 3.7315 0.0221 6wk
mmu-miR-1a-2-3p 3.5710 0.0103 6wk
mmu-miR-1a-1-3p 3.5710 0.0103 6wk
mmu-miR-133a-1-3p 2.8598 0.0103 6wk
mmu-miR-133a-2-3p 2.8598 0.0103 6wk
mmu-miR-1957a 2.5676 0.0401 6wk
mmu-miR-5100 1.9848 0.0127 6wk
mmu-miR-7a-2-5p 1.7627 0.0204 6wk
mmu-miR-7a-1-5p 1.7603 0.0204 6wk
mmu-miR-181a-1-5p 1.5137 0.0190 6wk
mmu-miR-181a-2-5p 1.5137 0.0190 6wk

Supplemental Figures

pca_heat_3

Supplemental Fig. 1: Diagnostic plots: principal component analysis (PCA) shows how well samples group based on global gene expression; heatmaps are color-coded values from an expression matrix, which may or may not include all the genes. A: PCA of three-week-old samples; B: Heatmap of the 20 mostly expressed miRNAs at three weeks of age; C: Heatmap of differentially expressed miRNAs at three weeks of age.

pca_heat_4

Supplemental Fig. 2: A: PCA of four-week-old samples; B: Heatmap of the 20 mostly expressed miRNAs at four weeks of age; C: Heatmap of differentially expressed miRNAs at four weeks of age.

Sup3. A: PCA of six-week-old samples; B: Heatmap of the 20 mostly expressed miRNAs at six weeks of age; C: Heatmap of differentially expressed miRNAs at six weeks of age.

Supplemental Fig. 3: A: PCA of six-week-old samples; B: Heatmap of the 20 mostly expressed miRNAs at six weeks of age; C: Heatmap of differentially expressed miRNAs at six weeks of age.

]]>
http://currents.plos.org/md/article/exploratory-profiling-of-urine-micrornas-in-the-dy2jdy2j-mouse-model-of-lama2-cmd-relation-to-disease-progression/feed/ 0
Greater Colo-Rectal Activation Phenotype in Exercised mdx Mice http://currents.plos.org/md/article/md-17-0007r1-greater-colo-rectal-activation-phenotype-in-exercised-mdx-mice/ http://currents.plos.org/md/article/md-17-0007r1-greater-colo-rectal-activation-phenotype-in-exercised-mdx-mice/#respond Wed, 02 May 2018 09:25:22 +0000 http://currents.plos.org/md/?post_type=article&p=10671 Introduction: Duchenne Muscular Dystrophy is a genetic disease that is caused by a deficiency of dystrophin protein. Both Duchenne Muscular Dystrophy patients and dystrophic mice suffer from intestinal dysfunction.

Methods: The present study arose from a chance observation of differences in fecal output of dystrophic vs. normal mice during 20­minutes of forced continuous treadmill exercise. Here, we report on the effects of exercise on fecal output in two different dystrophic mutants and their normal background control strains. All fecal materials evacuated during exercise were counted, dried and weighed.

Results: Mice of both mutant dystrophic strains produced significantly more fecal material during the exercise bout than the relevant control strains.

Discussion: We propose that exercise-­induced Colo-­Rectal Activation Phenotype test could be used as a simple, highly sensitive, non­invasive biomarker to determine efficacy of dystrophin replacement therapies.

]]>

INTRODUCTION

Duchenne Muscular Dystrophy (DMD) is an X-­linked genetic disease characterized by the lack of the Dystrophin protein in the muscle tissue of affected individuals. While the predominant problem associated with DMD lies with the striated skeletal and cardiac muscles, a high percentage of DMD patients also suffer from gastrointestinal dysfunction. Constipation, gastric hypomotility, gastric dilation, and delayed gastric emptying are among the gastrointestinal motor abnormalities seen in DMD patients 1, 2, 3, 4.

mdx mice, the animal model of DMD, also present symptoms of digestive dysfunction. Dystrophic mice exhibit delayed intestinal transit time, decreased fecal output, and gastrointestinal contractility disturbances 5, 6, 7. These demonstrations of differences in gastrointestinal function of dystrophic vs. normal animals have employed variably invasive methods that preclude their use as routine investigational tools for serial monitoring of individual mice.

We noticed a gross difference in fecal output between the Exon52­-null mdx mouse (mdx52) and its background control C57BL/6 (BL6) strain mice when subjected to a 20­-minute forced treadmill run. To investigate whether this difference was simply (1) a strain difference, arising perhaps from genetic drift between the dystrophic and control colonies, or (2) was related to the lack of dystrophin, we subsequently repeated the comparison on the C57BL/10ScSn­-mdx (mdx), which carries a natural nonsense point mutation in exon 23, and its background control C57BL/10ScS (BL10). Dystrophin expression is eliminated in both dystrophic strains, yet numerous variations can be found between the dystrophic strains that differentially affect functional and structural outcomes17, such as expansion of revertant fibers19, expression of shorter isoforms of dystrophin in the mdx strain20, and abnormal electroretinograms in mdx52 similar to those in DMD patients16, 18.

Here, we present a simple protocol that arose from a chance observation from an unrelated study of the effects of routine enforced exercise on dystrophic and normal mice. This protocol is simple and sensitive enough to be potentially used as a routine measure of success in restoration of dystrophin by following variations in fecal output.

MATERIALS AND METHODS

Animals

All animal procedures were approved and performed according to the Children’s National Health System’s Institutional Animal Care and Use Committee (IACUC). Animals were housed at a density of up to 5 animals per cage and maintained in pathogen-­free conditions under 12­-hour light/12-­hour dark cycles at a temperature of 18­-23°C and 40­-60% humidity. The mice were allowed ad-­libitum access to standard mouse chow and drinking water except during exercise or resting protocol periods. All animals used in the study were male mice between 1-­5 months of age, and no animals were euthanized throughout the course of these experiments. Both wild-­type strains, C57BL/10ScSn (BL10) and C57BL/6 (BL6), as well as the dystrophic C57BL/10ScSn-Dmdmdx (mdx) strain, were originally obtained from Jackson Laboratories and bred in-house at The Children’s National Health System Research Animal Facility. The exon 52-deficient X-chromosome linked dystrophic mouse model generated on the C57BL/6 background (mdx52) was originally obtained from the laboratory of Dr. Shin’ichi Takeda (National Center of Neurology and Psychiatry, Japan) and bred in-house at Children’s National Health System.

Experimental Procedures

Exercise Protocol: Normal (BL10, n = 10; BL6, n = 6) and dystrophic (mdx, n = 9; mdx52, n = 6) male mice were exercised on a treadmill (Columbus Instruments, Columbus, Ohio) at 0° incline for 20-­minutes without rest at 12 meters/minute, as previously described8. This protocol was conducted within the animal house within its set temperature range and during the 12 hour standard ‘daylight period’ employed in our animal unit. All animals used for the exercise protocol were studied longitudinally, following the same BL10 and mdx male mice and exercising them at 8, 9, and 20 weeks of age. Similarly, the same BL6 and mdx52 male mice were followed and exercised at 4, 8, and 12 weeks of age. Animals were continuously monitored during this exercise period; animals that drifted onto the back platform of the treadmill were pushed gently back onto the belt with a wad of tissue paper.

Resting Protocol: Normal (BL10, n = 10; BL6, n = 8;) and dystrophic (mdx, n = 8; mdx52, n = 9) mice were individually placed in plastic cages for 20-­minutes undisturbed. Animals of various ages were used to determine age effect on resting fecal output (BL10: 7, 14, 18 weeks; BL6: 8, 15 weeks; mdx: 7, 15 weeks; mdx52: 6, 11, 16 weeks).

Fecal Sample Collection: Fecal pellets excreted during the 20-­minute exercise or resting period were collected, counted, and weighed after being dried at 37°C for 10­-minutes.

All exercise and resting protocols were performed between 8:00 A.M. and 11:00 A.M. in an isolated room maintained at 21°C, 65% humidity, and <80 watts of light.

Statistical Analyses

All data are reported as mean values ± standard error of the mean (SEM). The consistently zero values obtained from BL6 control mice limited us to non­parametric statistical analysis. The non­parametric Spearman correlation (correlation coefficient r) was performed between both fecal output rate and weight per fecal pellet over age within the same protocol and strain. Statistical significance between strains within a protocol was assessed by the non­parametric Wilcoxon rank sum (Mann­-Whitney) test. The level of significance was set at P < 0.05.

RESULTS

Fecal output remains constant across all ages regardless of exercise/resting protocol or strain.

To determine age effect on fecal output, animals of various ages were studied. No correlation was found between fecal output rate (Fig. 1A-­B; r = ­-1.00 to 0.87) or weight per fecal pellet (Fig. 1C­-D; r = -­0.80 to 0.50) over age. Correlation for exercised BL10 and BL6 animals were not calculated due to the absence of fecal output from some animals at certain time­-points. Due to the absence of age effect on fecal output, subsequent analysis was performed from pooled data across all ages.

Figure 1

Fig. 1: Fecal output remains constant across all ages regardless of exercise/resting protocol or strain. (A) Resting and (B) exercised fecal output rate over age. (C) Resting and (D) exercised weight per fecal pellet over age. All data are reported as mean values ± SEM with level of significance set at P < 0.05.

Fecal pellet weights are constant between all strains regardless of exercise/resting protocol.

To determine whether the altered GI function in dystrophic mice correlates with altered fecal weight, the mean fecal weight per pellet was calculated for normal and dystrophic animals within each protocol group. (Fig. 2) No significant differences in weight per fecal pellet were found between strains or protocol (Resting vs. Exercise; BL10: 29.33 ± 8.59 vs. 16.83 ± 4.075; mdx: 29.17 ± 14.17 vs. 26.48 ± 1.779; BL6: 21.33 ± 3.27 vs. No data; mdx52: 18.39 ± 4.019 vs. 19.40 ± 3.00; n = 2­6). Absence of fecal output from exercised BL6 animals precluded calculation of weight per pellet from this group.

Figure 2

Fig. 2: Fecal pellet weights are constant between all strains regardless of exercise/resting protocol. Weight per fecal pellet in resting (open circles) and exercised (closed circles) animals. Data for exercised BL6 animals are not shown due to the complete absence of fecal output. All data are reported as mean values ± SEM with level of significance set at P < 0.05.

Resting animals show comparable fecal output rates across all strains.

To eliminate the theory that the higher fecal output in dystrophic mice is due to lower fecal output during times of rest that result in higher output during periods of exercise-­induced stress, resting baseline fecal output rate was studied in normal and dystrophic animals subjected to the resting protocol. (Fig. 3) Normal and dystrophic mice of both strains defecated at similarly low rates that were statistically indistinguishable from one another (BL10: 0.90 ± 0.35; mdx: 0.63 ± 0.42; BL6: 1.25 ± 0.45; mdx52: 1.78 ± 0.62; n = 8­10).

Figure 3

Fig. 3: Resting animals show comparable fecal output rates across all strains. Rate of fecal output in resting animals. All data are reported as mean values ± SEM with level of significance set at P < 0.05.

Exercised dystrophic mice show significantly higher rates of fecal output than normal mice.

To determine the effect of exercise-­induced stress on fecal output, fecal output rate was studied in normal and dystrophic animals subjected to the exercise protocol. (Fig. 4) The mean fecal pellet number was significantly higher in dystrophic animals of both strains than in normal animals (mdx: 6.11 ± 0.37 vs. BL10: 0.16 ± 0.07, P = 0.0079, n = 5; mdx52: 2.89 ± 0.33 vs. BL6: 0.00 ± 0.00, P = 0.0022, n = 6). Interestingly, BL6 animals ceased defecation completely during exercise, while normal BL10 animals still had minimal amounts of defecation.

Figure 4

Fig. 4: Fecal output remains constant across all ages regardless of exercise/resting protocol or strain. (A) Resting and (B) exercised fecal output rate over age. (C) Resting and (D) exercised weight per fecal pellet over age. All data are reported as mean values ± SEM with level of significance set at P < 0.05.

DISCUSSION

The aim of this study was to further investigate the differences in fecal output of dystrophic (mdx52 and mdx) vs. normal (BL6 & BL10) mice. Therefore, we compared the number of fecal pellets and total fecal weight from dystrophic and normal mice.

Numerous studies have confirmed that a high proportion of DMD patients suffer from gastrointestinal dysfunction such as constipation, gastric hypomotility and delayed gastric emptying 1, 3, 4, 9, 10, 11, 12. This has been linked tentatively to the fatty tissue infiltration, atrophy, fibrosis and interstitial edema in the gastrointestinal tract in the absence of dystrophin within smooth muscles 13.

While our findings of hyperactive gastrointestinal activity in dystrophic mice do not fit straightforwardly with the observations in humans, neither are they contradictory – they may be interpreted as a response to exercise-­induced stress superimposed on a background of fecal retention. This idea is supported by the fact that, in an empty cage, where human handling is the only stress, no difference was noted in fecal output between dystrophic and normal mice. It is intriguing that the BL6 control mice ceased even the low incidence of defecation seen in empty cages when forced to exercise, whereas both dystrophic strains showed distinctly greater fecal output than their normal counterparts in the latter conditions.

Our findings also seem superficially at variance with the reports of Mule et al, of significantly lower fecal pellet output, as well as gastric emptying and longer intestinal transit times in dystrophic mice 5. We suggest that this may reflect the differences between the stresses generated by oral gavage or by the 24­hour food deprivation in the Mule et al. studies, compared with the 20-­minute forced exercise regimen that preceded collection of fecal samples in our study. We suspect that the substantial difference between fecal outputs of exercised dystrophic and normal animals have not been noticed previously because most protocols do not involve exercising the two strains side by side.

As for other tests of behavioral features associated with lack of dystrophin, the basic mechanisms have yet to be elucidated. Morphological changes in the smooth muscle cells of dystrophic mice have been reported, including a reduction in the thickness of intestinal walls 14, 15 that might account for their abnormal gastrointestinal functions. However, the disentanglement of neurological from muscular components of what appears to be a stress response is likely to prove challenging.

We also acknowledge the unnatural requirement for mice – who are naturally nocturnal – to perform a forced exercise protocol during the day. While we believe that both exercise and stress increases fecal output, the consistency of exercising the dystrophic and normal mice in the same environmental conditions provides assurance that the difference we observe is significant. If the test were to be employed in an animal facility that used a day-night reversal protocol, it would be important to test whether this has any effect on the outcome.

CONCLUSIONS

Irrespective of the underlying mechanisms, this study provides evidence that dystrophic mice show alterations in gastrointestinal function that are easily measured by imposition of a low impact procedure. The size and consistency of the difference between dystrophic and normal mice of the two strains we have tested provides a clear and sensitive signal with little or no overlap, and also emphasizes the need to test any given dystrophin mutation against its background strain of origin. We therefore propose the exercise­induced Colo­-Rectal Activation Phenotype (CRAP) test as a simple, non­invasive biomarker that would be useful for serial longitudinal sampling as part of any test to determine the treatment efficacy of therapeutic restoration of dystrophin.

Corresponding Author

Marie Nearing, MNearing@childrensnational.org and Terence Partridge, TPartridge@childrensnational.org

Data Availability

All relevant data can be found at Figshare: https://doi.org/10.6084/m9.figshare.5738352.v1.

Competing Interests

The authors have declared that no competing interests exist.

]]>
http://currents.plos.org/md/article/md-17-0007r1-greater-colo-rectal-activation-phenotype-in-exercised-mdx-mice/feed/ 0
The FVB Background Does Not Dramatically Alter the Dystrophic Phenotype of Mdx Mice http://currents.plos.org/md/article/the-fvb-background-does-not-dramatically-alter-the-dystrophic-phenotype-of-mdx-mice/ http://currents.plos.org/md/article/the-fvb-background-does-not-dramatically-alter-the-dystrophic-phenotype-of-mdx-mice/#respond Tue, 10 Feb 2015 16:05:36 +0000 http://currents.plos.org/md/?post_type=article&p=7709

Introduction

Duchenne muscular dystrophy (DMD) is an X-linked lethal progressive muscle wasting disorder mainly affecting boys. It is caused by mutations in the dystrophin gene, one of the largest and most conserved genes in the genome (reviewed in 1 ). Numerous mouse models have been used to study dystrophin function and DMD pathogenesis (reviewed in 25). Among these, the mdx mouse is the most frequently used model (reviewed in 6). First described in 1984 by Bulfield and colleagues as a spontaneous myopathy model in C57/BL10 (BL10) mice, mdx mice carry a nonsense mutation in the exon 23 of the dystrophin gene 7,8.

Transgenesis is one of the most powerful technologies to investigate gene function in animal models (reviewed in 9,10). Studies conducted in transgenic mdx mice have laid the foundation for our current understanding of the structure-function relationship of dystrophin and DMD gene therapy (reviewed in 11). We recently began to use the transgenic approach to characterize the function of the dystrophin nNOS-binding domain and to explore cardiac unique features of the dystrophin gene 12,sup>-15. The FVB mouse has been the preferred inbred strain for the production of transgenic mice because of its robust reproductive performance, unusually large pronuclei of the fertilized oocytes (easy for microinjection) and excellent nurturing characteristics 16. To minimize the influence of the genetic background, in the past we have to backcross our FVB background transgenic mice for many generations to the BL10 background before use them in the study 1214. With more transgenic lines being developed, it becomes an extremely labor-intensive and time-demanding task to backcross every FVB transgenic line to the mdx background. To solve this problem, we decided to generate FVB background mdx (mdx/FVB) mice. We backcrossed inbred FVB mice with the original BL10 background mdx mice for seven generations. The resulting mdx/FVB mice had white coat color. However, they showed the characteristic histological and physiological changes as the original mdx mice. Our results suggest that the mdx/FVB mouse may represent a useful model to study DMD.

Materials and Methods

Experimental Animals. All animal experiments were approved by the institutional animal care and use committee and were in accordance with NIH guidelines. Parental FVB (FVB/NJ, Jackson stock number 001800) and mdx (C57BL/10ScSn-Dmdmdx/J, Jackson stock number 001801) mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were maintained in a specific-pathogen free animal care facility on a 12-hour light (25 lux): 12-hour dark cycle with access to food and water ad libitum. For histological and physiological studies, only male mice were used.

Generation of mdx/FVB mice. The mdx/FVB mouse was generated by seven generations of backcrossing. Briefly, female mdx mice were crossed with male FVB mice to obtain the F1 progeny. Heterozygous females were bred with FVB males to get F2 progeny. Heterozygous F2 females were identified by PCR according to our published protocol and crossed with male FVB mice to get F3 progeny 17. In subsequent rounds of breeding, the heterozygous females were used to cross with FVB males until a total of seven generations of backcrossing was finished. Dystrophin-deficient males and heterozygous females from the last round of backcrossing were interbred to generate experimental mdx/FVB mice.

Morphological studies. General histology was examined by hematoxylin and eosin (HE) staining. Central nucleation was quantified on 3 to 5 random 20x fields for each muscle. Fibrosis was examined by Masson trichrome staining as we described before 18. Dystrophin expression was evaluated by immunofluorescence staining using the dystrophin C-terminal domain specific Dys-2 antibody (1:100; clone Dy10/12B2, IgG2a; Novocastra, Newcastle, UK) 19,20. Slides were viewed at the identical exposure setting using a Nikon E800 fluorescence microscope. Photomicrographs were taken with a Qimage REtiga 1300 camera 18.

Serum creatine kinase (CK) activity assay. Fresh serum was collected by tail vein bleeding. The CK activity was determined using CK liqui-UV test kit from Stanbio Laboratory (Boerne, TX) according to the manufacturer’s guidelines.

Fore limb grip strength measurement. Fore limb grip strength was measured with a computerized grip strength meter (Columbus Instruments, Columbus, OH) as we described previously 21,22. The grip strength meter has a pulling bar attached to a force transducer and a digital display. Briefly, the mice were first checked for any sores in the limbs and toes prior to the experiment. Only mice without apparent skin injury were used in the study. The mice were first acclaimed to the apparatus for approximately 5 min. Mouse was then allowed to grab the pulling bar by holding it from the tip of the tail. The mouse was gently pulled away from the grip bar. When the mouse can no longer grasp the bar, the reading was recorded. Protocol was repeated five times with at least 30 sec rest between trials. The highest three values were averaged to obtain the absolute grip strength. Normalized grip strength was obtained by dividing the absolute grip strength with the body weight.

EDL muscle function evaluation. EDL muscle force was determined ex vivo according to our published protocol 21,23. Briefly, mice were anesthetized via intra-peritoneal injection of a cocktail containing 25 mg/ml ketamine, 2.5 mg/ml xylazine and 0.5 mg/ ml acepromazine at 2.5 µl/g body weight. The EDL muscle was gently dissected and mounted to an intact muscle test system (Aurora Scientific, Inc., Aurora, ON, Canada) containing oxygenated (95% O2 and 5% CO2 at 30ºC) Ringer’s buffer. After 10 min equilibration, the muscle length (Lm) of the EDL muscle was measured with an electronic digital caliper (Fisher Scientific, Waltham, MA, USA). This length is defined as the optimal muscle length (L0). The maximum isometric tetanic force (Po) was measured at 150 Hz. The muscle cross-sectional area (CSA) was calculated according to the following equation, CSA = (muscle mass, in gram)/[(optimal fiber length (Lf), in cm) × (muscle density, in g/cm3)]. A muscle density of 1.06 g/cm3 was used in calculation. The optimal fiber length is calculated as 0.44 x Lo. 0.44 represents the ratio of the fiber length to the Lo of the EDL muscle 21. Specific muscle force was determined by dividing the maximum isometric tetanic force with the muscle CSA. After tetanic force measurement, the muscle was rested for 10 min and then subjected to ten rounds of eccentric contraction injury according to our previously published protocol 21,23. Briefly, following a tetanic contraction EDL muscle was stretched 10%L0 at a rate 0.5L0/sec. The muscle was allowed to rest 2 min between each eccentric cycle. The percentage of force drop following each round of eccentric contraction was recorded. Data were processed using the Lab View-based DMC and DMA programs (Version 3.12, Aurora Scientific, Inc.).

TA muscle function evaluation. The TA muscle force was measured in situ according to our published protocol 21,23. Briefly, mice were anesthetized as described above. The TA muscle and the sciatic nerve were exposed. The mouse was transferred to a customer-designed thermo-controlled platform of the footplate apparatus. Sciatic nerve was stimulated using a custom-made 25G platinum electrode to elicit muscle contraction. Subsequently, twitch and tetanic forces and the eccentric contraction profile were measured with a 305C-LR dual-mode servomotor transducer (Aurora Scientific, Inc.). Data recording and analysis were identical to methods described for the EDL muscle. In TA muscle cross-sectional area calculation, the optimal fiber length was calculated as 0.60 x L0. 0.60 represents the ratio of the fiber length to the L0 of the TA muscle 21.

Statistical analysis. Data are presented as mean ± standard error of mean (s.e.m.). Statistical significance between FVB and mdx/FVB was determined by the Student t-test. Difference was considered statistically significant when P < 0.05.

Results

Adult mdx/FVB mice show dystrophic muscle pathology and elevated serum CK. The coat color of mdx mice is black while that of FVB mice is white. After crossing mdx mice with FVB mice for seven generations, we obtained the expected white colored mdx/FVB mice (Figure 1A). To determine whether mdx/FVB mice had myopathy, we first examined histology in the TA muscle of 3 and 6-m-old mice (Figure 1B). Dystrophin expression was observed in the FVB muscle but not in the mdx/FVB muscle. On HE staining, we observed inflammation, degeneration/regeneration and necrosis in the mdx/FVB muscle (Figure 1B). The FVB muscle had the uniform myofiber size but in the mdx/FVB muscle, we noticed the presence of extremely large and small myofibers (Figure 1B). In wild type FVB mice, central nucleation was <1%. In mdx/FVB mice, central nucleation reached 66-71% (Table 1, N=11 mice/group, ~ 12,000 myofibers quantified per strain).

Fig. 1: The tibialis anterior (TA) muscle of the mdx/FVB mouse displays characteristic histological changes of muscular dystrophy.

A, Representative photographs of experimental mice. Left panel, BL10 and mdx mice; Right panel, FVB and mdx/FVB mice. B, Representative photomicrographs of dystrophin immunofluorescence staining (top panel), HE staining (middle panel) and Masson trichrome staining (bottom panel) of the FVB and mdx/FVB TA muscles.

CK level elevation is a salient feature in mdx mice. Consistently, the CK level in mdx/FVB mice was also significantly higher than that of FVB mice (Figure 2A).

Muscle function is significantly compromised in mdx/FVB mice. Three methods were used to evaluate muscle function in 3 and 6-m-old FVB and mdx/FVB mice. Forelimb grip strength was quantified in awaken intact mice. Compared to that of FVB mice, body-weight normalized grip strength was reduced by ~ 50% in mdx/FVB mice (Figure 2B).

Fig. 2: Serum creatine kinase (CK) and forelimb grip in FVB and mdx/FVB mice.

A, Quantification of the serum CK level. n=8 and 9 for 3-m-old and n=4 and 8 for 6-m-old FVB and mdx/FVB mice, respectively. B, Forelimb grip strength. n=7 and 8 for 3-m-old and n=4 and 6 for 6-m-old FVB and mdx/FVB mice, respectively. The absolute grip force is normalized to the body weight. Asterisk, significantly different from that of FVB mice.

The TA muscle force was analyzed in situ in anesthetized mice (Figure 3, Table 1). The specific twitch force was marginally reduced in mdx/FVB mice (p=0.05) (Figure 3A). However, the specific tetanic force was significantly decreased in mdx/FVB mice. It only reached ~ 80% of the normal (Figure 3B). Mdx/FVB mice were significantly more susceptible to eccentric contraction damage (Figure 3C). From 3 to 6 months, the eccentric contraction profile of FVB mice did not change much. Interestingly, compared to that of 3-m-old mdx/FVB mice, 6-m-old mdx/FVB mice showed a much sharper force drop during the first three rounds of eccentric contraction. The residual force was also lower in 6-m-old mdx/FVB (~21% of the starting force; this value was ~ 35% in 3-m-old mdx/FVB).

Fig. 3: In situ analysis of the contractile properties of the tibialis anterior (TA) muscle in FVB and mdx/FVB mice.

A, Specific twitch force. n=10 and 8 for 3-m-old and n=8 and 9 for 6-m-old FVB and mdx/FVB mice, respectively; p=0.51 and p=0.50 for 3 and 6-m-old comparisons respectively. B, Specific tetanic force. n=10 and 8 for 3-m-old and n=8 and 9 for 6-m-old FVB and mdx/FVB mice, respectively. C, Percentage of force drop during ten cycles of eccentric contractions. Absolute force generated during the first cycle is set as the baseline (100%) and the percentage of force drop following each cycle of eccentric contraction is calculated relative to the baseline. n=7-10 and 7-8 for 3-m-old and n=7-8 and 6-9 for 6-m-old FVB and mdx/FVB mice, respectively. Asterisk, significantly different from that of FVB mice.

The EDL muscle force was analyzed ex vivo (Figure 4, Table 1). The overall trend was similar to that of the TA muscle with mdx/FVB mice showing significantly more compromised contractility compared to that of normal mice. Interestingly, a statistically significant difference in the specific twitch force was found between FVB and mdx/FVB mice (Figure 4A). Further, the eccentric contraction profile of 3-m-old mdx/FVB mice was similar to that of 6-m-old mdx/FVB (Figure 4C).

Table 1. Comparison of body weight, muscle weight and centronucleation in FVB and mdx/FVB mice

Data shown as mean ± standard error of mean. a, significantly different from the 3-m-old FVB control group; b, significantly different from the 6-m-old FVB control group.

Abbreviations: EDL, extensor digitorum longus; CSA, cross sectional area; TA, tibialis anterior; CN, central nucleation.

Sample size: Body weight n=11-14 for FVB, n=11-19 for mdx/FVB; EDL weight/CSA n=14-24 for FVB, n=12-18 for mdx/FVB; TA weight/CSA n=8-10 for FVB, n=8-9 for mdx/FVB; CN analysis n= 10-11 for each strain.

FVB mdx/FVB
3-m-old 6-m-old 3-m-old 6-m-old
Body weight (g) 28.67±0.60 32.52±0.44 34.06±0.58a 39.19±1.20b
EDL weight (mg) 11.08±0.27 12.36±0.16 14.03±0.14a 13.41±0.19b
EDL CSA (mm2) 1.81±0.04 1.95±0.03 2.13±0.05a 2.07±0.04b
TA weight (mg) 52.92±1.45 65.38±1.04 62.24±3.01a 73.99±2.10b
TA CSA (mm2) 5.63±0.15 6.90±0.12 6.08±0.29 7.29±0.14b
Central nucleation (%) 0.28±0.15 0.19±0.03 65.80±1.44a 70.77±0.82b

Fig. 4: Ex vivo analysis of the contractile properties of the extensor digitorum longus (EDL) muscle in FVB and mdx/FVB mice.

A, Specific twitch force. n=14 and 12 for 3-m-old and n=24 and 18 for 6-m-old FVB and mdx/FVB mice respectively. B, Specific tetanic force. n=14 and 12 for 3-m-old and n=24 and 18 for 6-m-old FVB and mdx/FVB mice respectively. C, Percentage of force drop during ten cycles of eccentric contractions. Absolute force generated during the first cycle is set as the baseline and (100%) and the percentage of force drop following each cycle of eccentric is calculated relative to the baseline. n=9-12 and 9-12 for 3-m-old and n=14-15 and 12-14 for 6-m-old FVB and mdx/FVB mice respectively. Asterisk, significantly different from FVB mice.

Discussion

To meet the practical needs of our transgenic studies, we crossed the BL10-background mdx mice with FVB/NJ mice. Recent studies suggest that the so-called “wild type” inbred mice may actually carry various changes in their genome. For example, the commonly used A/J mice were recently show to display progressive muscular dystrophy due to a mutation in the dysferlin gene 24. The FVB strain was also found to carry mutations in several genes of the visual system 25. It is thus important to determine whether the FVB background alters the dystrophic phenotype of the original mdx mice. After seven generations of backcross, we obtained white mdx/FVB mice. These mice showed classic dystrophic changes including elevated serum CK, myofiber centronucleation, muscle inflammation and fibrosis, force reduction and enhanced sensitivity to eccentric contraction injury. In young adult mdx mice, the specific twitch and tetanic force for EDL muscle range from 26.6±1.2 to 29.0±1.5 and 129.6±10.5 to 138.5±5.6 mN/mm2, respectively. In young adult C57Bl/10 (BL10) mice, the specific twitch and tetanic force for EDL muscle range from 33.2±1.8 to 46.5±3.7 and 185.4±5.7 to 245.0±1.4 mN/mm2, respectively 12,2628. Muscle force drops from 100% (baseline) to 53.5-29.4% (after 10cycles of eccentric contraction) in young mdx mice. Muscle force drops from 100% (baseline) to 73.1-68.0% (after 10 cycles of eccentric contraction) in BL10 mice 26,27. The values we observed in EDL muscle of mdx/FVB mice were comparable to these of mdx mice. The contractile properties of the TA muscle in mdx/FVB mice and FVB mice also fall within the range of those reported in mdx/mdx4cv mice and BL10/BL6 mice, respectively 2931. Tables 2 and 3 show a comprehensive comparison of contractile properties of limb muscles in FVB vs BL10 and mdx/FVB vs mdx/BL10 mice. Elevated levels of serum CK and myofiber centronucleation are hallmarks of muscle diseases in mdx and mdx4cv mice 27,3133. The mdx/FVB mice showed the similar trend. Based on the phenotypic similarity between mdx/FVB mice and the original mdx mice, we conclude that mdx/FVB may serve as a good control for studying FVB-background transgenic mdx mice.

Table 2. Contractile properties of FVB versus BL10 limb muscles

a, values are from the Duan lab studies

b, values are from 4-m-old BL6 mice

FVB
(3 to 6-m-old)
BL10
(2 to 8-m-old)
References
Extensor digitorum longus
Specific twitch force (mN/mm2) 34.8±1.7 to 40.8±1.5 33.2±1.8a to 46.5±3.7a 26, 27, 28
Specific tetanic force (mN/mm2) 189.0±8.7 to 202.0±3.4 185.4±5.7a to 245.0±1.4a 26, 27, 28
Percent force decrease following 10 cycles of eccentric contractions 26.9±2.0 to 28.1±1.5 26.8±0.25a to 32.0±3.2a 26, 27
Tibialis anterior
Specific twitch force (mN/mm2) 67.7±3.6 to 75.3±3.7 41.7±1.6b 31
Specific tetanic force (mN/mm2) 241.9±8.2 to 255.0±7.0 ~225 to 268.5±4.0b 12, 29, 30, 31
Percent force decrease following 10 cycles of eccentric contractions 39.0±1.9 to 44.2±1.9 23.8±3.6b 31

Table 3. Contractile properties of mdx limb muscles on the FVB (mdx/FVB) versus BL10 background (mdx)

a, values are from the Duan lab studies

b, Data are from 4-m-old mdx4cv mice

mdx/FVB
(3 to 6-m-old)
mdx
(2 to 8-m-old)
References
Extensor digitorum longus
Specific twitch force (mN/mm2) 24.5±2.3 to 27.3±1.2 26.6±1.2a to 29.0±1.5a 26, 27, 28
Specific tetanic force (mN/mm2) 120.4±9.3 to 142.2±7.4 129.6±10.5a to 138.5±5.6a 26, 27, 28
Percent force decrease following 10 cycles of eccentric contractions 47.5±2.5 to 55.5±1.8 46.5±4.5a to 70.6±2.7a 26, 27
Tibialis anterior
Specific twitch force (mN/mm2) 63.5±4.7 to 72.2±2.3 27.7±3.0b 31
Specific tetanic force (mN/mm2) 192.5±10.3 to 206.8±7.1 129.5±10.5b to ~200 12, 29, 30, 31
Percent force decrease following 10 cycles of eccentric contractions 65.1±5.4 to 78.6±2.0 66.4±5.2b 31

Over the last two decades, mdx mice have been backcrossed to the background of at least six different inbred strains including albino, BALB/C, C3H, BL6, DBA/2 and FVB 3440. Except for the DBA/2 background mdx mice 37, the dystrophic phenotype is rarely characterized in other backgrounds. It has become apparent that genetic background can significantly modulate the phenotype of single gene mutation in mice (Reviewed in 4144). This feature has been used in genome-wide genetic analysis to identify the genetic modifiers that may account for the phenotypic differences in muscular dystrophy (reviewed in 45,46). The mdx/FVB strain described here may add in the research in this direction.

Corresponding Authour

Dongsheng Duan Ph.D.

Department of Molecular Microbiology and Immunology

One Hospital Drive

Columbia, MO 65212

Phone: 573-884-9584

Fax: 573-882-4287

Email: duand@missouri.edu

Competing Interests

The authors have declared that no competing interests exist.

]]>
http://currents.plos.org/md/article/the-fvb-background-does-not-dramatically-alter-the-dystrophic-phenotype-of-mdx-mice/feed/ 0
Absence of a Major Role for the Snai1 and Snai3 Genes in Regulating Skeletal Muscle Regeneration in Mice http://currents.plos.org/md/article/absence-of-a-major-role-for-the-snai1-and-snai3-genes-in-regulating-skeletal-muscle-regeneration-in-mice/ http://currents.plos.org/md/article/absence-of-a-major-role-for-the-snai1-and-snai3-genes-in-regulating-skeletal-muscle-regeneration-in-mice/#respond Fri, 08 Nov 2013 03:30:05 +0000 http://currents.plos.org/md/?post_type=article&p=5733

Introduction

The Snail gene family encodes zinc finger proteins that function as transcriptional repressors 1,2. Three members of the Snail gene family have been described in mammals, encoded by the Snai1 (also termed Snail), Snai2 (Slug), and Snai3 (Smuc) genes. While the Snai1 and Snai2 genes and proteins have been studied extensively in both mice and humans, much less is known about the functions of the Snai3 gene. Snai3 (originally termed Smuc, for Snail related gene from skeletal MUscle Cells) was isolated using a degenerate PCR-amplification protocol as a Snail family gene expressed in adult mouse skeletal muscle 3. Northern blot analysis revealed that the Snai3 gene was highly expressed in adult mouse skeletal muscle and thymus, at lower levels in adult heart, lung and spleen, and was also expressed during embryogenesis 3. Analysis by in situ hybridization during mouse embryogenesis revealed that Snai3 transcripts were first observed at embryonic day (E)13.5 in skeletal muscle and diaphragm 4. At E15.5, in addition to skeletal muscle and diaphragm expression, Snai3 transcripts also were expressed in the thymus. Skeletal muscle and thymus remained the dominant sites of Snai3 expression through the early postnatal period 4. However, both our laboratory 5 and the Weis laboratory 6 described recently the absence of an obvious phenotype in Snai3 null mice. Here we report the analysis of Snai3 mutant mice, and of Snai1/Snai3 double mutant mice, using the cardiotoxin injury model of hindlimb skeletal muscle regeneration.

Materials and methods

Mice

Generation and genotyping of the Snai3null 5, Snai3-EYFP 5 and Snai1-flox 7 mice have been described previously. Myf5-Cre mice 8 were obtained from the Jackson Laboratory, and Mef2c-73k-Cre mice 9 were obtained from Dr. Brian Black. All mouse protocols followed the guidelines of the US National Research Council Guide for the Care and Use of Laboratory Animals, and were approved by the Maine Medical Center Institutional Animal Care and Use Committee.

Hindlimb regeneration assay, histological analyses, and skeletal preparations

Tibialis anterior (TA) muscles of anesthetized mutant and control littermate mice at 11 to 18 weeks of age were injected with 100 ul of 10 uM cardiotoxin (Naja mossambica cardiotoxin; catalog number C9759, Sigma-Aldrich). Mice were euthanized 10 to 12 days after injection to collect the injured TA muscles for histological analysis. Some TA muscles were embedded in paraffin, sectioned at 7 um, and stained with hematoxylin and eosin. Other TA muscles were snap frozen and cryosectioned before staining with hematoxylin and eosin. All regeneration experiments were repeated at least three times. Myofiber cross sectional area was measured using Zeiss Axiovison software, and differences of means of the genotype groups were tested for statistical significance using the two tailed, unpaired Student’s t test. Alcian blue-alizarin red-stained skeletal preparations were generated as described previously 10. Mice were euthanized, skinned, eviscerated, and fixed in 100% ethanol (EtOH). They were then stained in 0.015% Alcian blue, 0.005% Alizarin red in 5% acetic acid/70% EtOH. Clearing was performed in 2% potassium hydroxide, followed by 1% potassium hydroxide/20% glycerol, after which they were brought into 80% glycerol for photography and storage. For conditional deletion experiments, efficient gene deletion in skeletal muscle by the Myf5-Cre and Mef2c-73k-Cre drivers was confirmed by quantitative PCR on genomic DNA isolated from the gastrocnemius or TA muscle.

Results

Snai3 null mutant mice exhibit normal skeletal muscle regeneration after injury

We recently described the generation and analysis of two different null alleles of the Snai3 gene, Snai3null and Snai3-EYFP 5. Due to the fact that the Snai3 gene was originally cloned from mouse skeletal muscle RNA, and is expressed at high levels in that tissue, we paid particular attention to possible pathological changes in skeletal muscle in the Snai3 mutant mice. Histological analysis of multiple skeletal muscles of Snai3null/Snai3null or Snai3-EYFP/Snai3-EYFP homozygous mice through 13 months of age did not reveal any obvious skeletal muscle pathology, such as muscle hypotrophy, aberrant muscle fiber size, centrally located nuclei, or infiltration of fibrotic or adipose tissue (Figure 1).

Fig. 1: Snai3null/Snai3null mice have normal skeletal muscle.

(A-D) Sections of the lower hindlimb of wildtype (A, C) and Snai3-null/Snai3-null (B, D) littermate mice at 13 months of age. No differences were observed between the Snai3 mutants and their wildtype littermates. Sections from paraffin-embedded legs were stained with hematoxylin and eosin. Abbreviations: EDL: extensor digitorum longus muscle; F: fibula; PE: peroneus longus muscle; T: tibia; TA: tibialis anterior muscle. Magnifications: (A, B) 2.5X; (C, D) 20X.

To additionally assess formation and integrity of the musculoskeletal system, we examined Alcian blue-alizarin red-stained skeletons of Snai3null/Snai3null or Snai3-EYFP/Snai3-EYFP homozygous and control littermate mice. No skeletal defects were observed in the Snai3 mutant mice (Figure 2, and data not shown).

Fig. 2: Snai3null/Snai3null mice do not exhibit altered bone formation.

Skeletal preparations were stained with Alizarin red (to stain mineralized bone) and Alcian blue (to stain cartilage). No differences in skeletal formation were observed between the wildtype (A) and Snai3-null/Snai3-null (B) littermate mice. F: fibula; Fe: femur; T: tibia.

To assess whether conditions of stress might reveal a requirement for Snai3 gene function in skeletal muscle, we tested the ability of the hindlimb Tibialis anterior (TA) muscle of Snai3 mutant and control littermate mice to regenerate after injury, using the standard cardiotoxin injury model. Both TA muscles of Snai3null/Snai3null or Snai3-EYFP/Snai3-EYFP and control littermate mice were injected with cardiotoxin, and the TA muscles were isolated 10 to 12 days after injury for histological analysis. These studies did not reveal substantive differences in regeneration between Snai3null/Snai3null or Snai3-EYFP/Snai3-EYFP homozygotes and their heterozygous and wildtype control littermates. As assessed by the presence of centrally located nuclei in the myofibers, both Snai3 mutant homozygotes and wildtype control littermates exhibited extensive myofiber regeneration after cardiotoxin injury (Figure 3A, B). The mean cross sectional area of regenerating myofibers did not differ significantly between Snai3 mutant homozygotes and wildtype control littermates (Fig. 3C). Some, but not all, Snai3 mutant homozygotes exhibited a small amount of fibrosis in the regenerating TA muscle. However, TA muscle regeneration still proceeded efficiently in the Snai3 mutant mice.

Fig. 3: Snai3null/Snai3null mice exhibit no obvious defects in skeletal muscle regeneration.

Hematoxylin and eosin-stained cryosections of TA muscle from wildtype littermate (A) and Snai3-null/Snai3-null (B) mice 10 days after cardiotoxin-mediated injury revealed no obvious differences between the two genotypes. Arrowheads indicate examples of centrally-located nuclei in regenerating myofibers. Magnification: 20X. (C) Mean myofiber cross sectional area (CSA) did not differ significantly between the wildtype littermate (n=2) and Snai3null/Snai3null (n=3) genotypes.

Skeletal muscle regeneration in Snai1/Snai3 double mutant mice

We have shown that during chondrogenesis (cartilage development) in mice, the Snai1 and Snai2 genes function redundantly, and that both Snai1 and Snai2 gene function must be removed to detect a phenotype during cartilage formation 10,11,12. The Snai1 gene is induced during skeletal muscle regeneration 13, and a recent paper demonstrated that a Snai1-HDAC1/2 repressive complex bound and excluded the myogenic transcription factor MyoD from its targets 14. We therefore decided to test TA muscle regeneration in Snai1/Snai3 double mutant mice. Since Snai1null/Snai1null homozygotes die early in embryogenesis 15, we utilized our Snai1-flox allele 7 and either the Myf5-Cre 8 or Mef2c-73k-Cre9 driver lines to perform skeletal muscle-specific deletion of the Snai1 gene.

We generated both Myf5-Cre/+; Snai1-flox/Snai1-flox; Snai3null/Snai3null and Mef2c-73k-Cre/+; Snai1-flox/Snai1-flox; Snai3null/Snai3null (or Mef2c-73k-Cre/+; Snai1-flox/Snai1-flox; Snai3-EYFP/Snai3-EYFP) mice (referred to as Snai1/Snai3 double mutant mice). Histological analysis of uninjured TA muscles from the Snai1/Snai3 double mutant mice did not reveal any obvious defects, compared to control littermate mice lacking the Cre allele (i.e., Snai1-flox/Snai1-flox; Snai3-EYFP/Snai3-EYFP mice) (Figure 4A, B).

TA muscles of Snai1/Snai3 double mutant mice and control littermate mice were injected with cardiotoxin, and the TA muscles were harvested 12 days after injury for histological analysis. The regenerating TA muscles of Snai1/Snai3 double mutant and control Cre-negative littermate mice appeared identical histologically (Figure 4C, D). Both control Myf5-Cre-negative; Snai1-flox/Snai1-flox; Snai3null/Snai3null TA muscle (Figure 4C) and Snai1/Snai3 double mutant TA muscle (Figure 4D) exhibited extensive myofiber regeneration after cardiotoxin injury. The mean cross sectional area of regenerating myofibers did not differ significantly between the two genotypes (Figure 4E). We conclude that the Snai3 gene does not play a major role in hindlimb skeletal muscle regeneration in mice, even in combination with the related Snai1 gene.

Fig. 4: Mice with skeletal muscle-specific deletion of the Snai1 gene on a Snai3 homozygous mutant background exhibit no obvious defects in skeletal muscle development or regeneration.

(A, B) Sections of uninjured TA muscle of control (Snai1-flox/Snai1-flox; Snai3-EYFP/Snai3-EYFP) (A) and Snai1/Snai3 double mutant (Mef2c-73k-Cre/+; Snai1-flox/Snai1-flox; Snai3-EYFP/Snai3-EYFP) (B) littermate mice at 5 months of age. (C, D) Sections of TA muscle 12 days after cardiotoxin-mediated injury from control (Snai1-flox/Snai1-flox; Snai3null/Snai3null) (C) and Snai1/Snai3 double mutant (Myf5-Cre/+; Snai1-flox/Snai1-flox; Snai3null/Snai3null) (D) mice. Arrowheads indicate centrally-located nuclei in regenerating myofibers. Sections from paraffin-embedded TA muscles were stained with hematoxylin and eosin. Magnification: 20X. (E) Mean myofiber cross sectional area (CSA) did not differ significantly between the control (n=3) and Snai1/Snai3 double mutant (n=4) genotypes.

Discussion

Our results demonstrate effective skeletal muscle regeneration after cardiotoxin-mediated injury in Snai3 homozygous mutant (Snai3null/Snai3null or Snai3-EYFP/Snai3-EYFP) mice. We further show that mice with skeletal muscle-specific deletion of the Snai1 gene on a Snai3 null genetic background exhibit the same general level of skeletal muscle regeneration as the Snai3 mutant homozygotes. While our histopathological analyses cannot exclude minor regeneration defects in the Snai3 single or Snai1/Snai3 double mutants, it is clear that substantial muscle regeneration occurs after cardiotoxin-mediated injury in these mice.

A recent study utilized ChIP-Seq and gene expression analyses to demonstrate that a Snai1-HDAC1/2 repressive complex bound and excluded the myogenic transcription factor MyoD from its targets 14. These authors further showed that a regulatory network involving myogenic regulatory factors, Snai1/Snai2, and the microRNAs miR-30a and miR-206 acted as a molecular switch controlling entry into myogenic differentiation. It is possible that we did not observe a substantial effect on skeletal muscle development or regeneration in our experiments because our mice were wildtype at the Snai2 locus.

In 2002, we participated in a study demonstrating that Snai2 gene expression is induced during muscle regeneration, and that Snai2-lacZ homozygous null mice exhibit impaired hindlimb skeletal muscle regeneration 13. At the time of those studies, our Snai2-lacZ mice were on a mixed (129S1/SvImJ X C57Bl/6J) genetic background, and approximately 50% of homozygotes survived into adulthood 16. The remainder died postnatally from cleft palate. Since that time, we have maintained the Snai2-lacZ line as a heterozygous backcross to C57BL/6J mice, and our Snai2-lacZ/+ mice are now a congenic line on the C57BL/6J genetic background. We have found that on the C57BL/6J background virtually all Snai2-lacZ/Snai2-lacZ homozygotes now die in the early postnatal period, apparently as an increase in the penetrance of the cleft palate phenotype. We therefore were not able to test Snai2-lacZ/Snai2-lacZ mice, or compound mutants containing the Snai2-lacZ allele, in the current set of experiments. Further work, including the generation and utilization of a Snai2-flox allele for skeletal muscle-specific Snai2 gene deletion, will be required to remove the function of all three Snail family genes in skeletal muscle to definitively assess the requirement for Snail family genes during skeletal muscle development and regeneration.

Correspondence

Thomas Gridley. Email: gridlt@mmc.org

Competing Interests

The authors have declared that no competing interests exist.

]]>
http://currents.plos.org/md/article/absence-of-a-major-role-for-the-snai1-and-snai3-genes-in-regulating-skeletal-muscle-regeneration-in-mice/feed/ 0
The HDAC Inhibitor TSA Ameliorates a Zebrafish Model of Duchenne Muscular Dystrophy http://currents.plos.org/md/article/the-hdac-inhibitor-tsa-ameliorates-a-zebrafish-model-of-duchenne-muscular-dystrophy/ http://currents.plos.org/md/article/the-hdac-inhibitor-tsa-ameliorates-a-zebrafish-model-of-duchenne-muscular-dystrophy/#respond Tue, 17 Sep 2013 17:10:39 +0000 http://currents.plos.org/md/?post_type=article&p=5049 dmd-MO). We use two approaches, muscle birefringence and muscle actin expression, to quantify muscle damage and show that the dmd-MO dystrophic phenotype closely resembles the zebrafish dmd mutant phenotype. We then show that the histone deacetylase (HDAC) inhibitor TSA, which has been shown to ameliorate the mdx mouse Duchenne model, can rescue muscle fiber damage in both dmd-MO and dmd mutant larvae. Our study identifies optimal morpholino and phenotypic scoring approaches for dystrophic zebrafish, further enhancing the zebrafish dmd model for rapid and cost-effective small molecule screening.]]>

Introduction

Muscular dystrophies are genetic disorders characterized by progressive muscle degeneration and impaired muscle function. Many muscular dystrophies are caused by mutations in genes that encode components of the Dystrophin-glycoprotein complex 1,2, including Duchenne muscular dystrophy (DMD), which is caused by mutations in the Dystrophin (DMD) gene 3. Mechanisms that contribute to muscle degeneration in the muscular dystrophies include muscle membrane instability, disrupted calcium homeostasis, and oxidative stress 2. Gene-mediated and cell-mediated therapeutic strategies for DMD hold tremendous promise, but these approaches still face many obstacles 4,5,6. Therefore, many different pharmacological therapies are currently being pursued5,7,8.

The zebrafish, Danio rerio, offers several advantages as a model system for screening for chemical modifiers of the muscular dystrophy phenotype. First, zebrafish can be produced readily in large numbers and provide a rapid and cost-effective system for chemical screening 9. Second, zebrafish represent excellent models of human muscular dystrophies, in particular loss of Dystrophin to model DMD 10,11,12. Third, recent approaches have demonstrated pharmacological rescue of the zebrafish dmd mutant and other zebrafish myopathy models 13,14,15,16.

Even with these significant advantages, limitations in the zebrafish DMD model remain. With the zebrafish dmd mutant strain, about 25% of the embryos from a cross of heterozygote carriers show the recessive dmd muscle lesion phenotype 10,11,13. Therefore, when using zebrafish dmd clutches for chemical screening, phenotypic rescue can be assessed only on these 25% mutants while 75% of the clutches are phenotypically normal. Furthermore, to accurately assess rescue, the mutant individuals must be genotyped. Multiple previous studies have used antisense morpholino oligos to knock down zebrafish dmd, but these studies report achieving a dystrophic phenotype in only about 30% of morpholino-treated animals 10,13,17. We thus wanted to identify an improved zebrafish dmd morpholino knock-down approach for chemical screening and also for rapid use in other zebrafish mutant and transgenic backgrounds.

Here we develop a robust zebrafish dmd morpholino (dmd-MO) knock-down model that closely resembles the zebrafish dmd mutant phenotype and achieves almost 100% penetrance. We show that this dmd-MO model is useful for identifying small molecules that rescue the dmd phenotype by showing that the histone deacetylase (HDAC) inhibitor TSA, which has been shown to ameliorate the mdx mouse DMD model 18,19,20,21, can rescue muscle fiber damage similarly in both dmd-MO and dmd mutant larvae. Our study identifies optimal morpholino and phenotypic scoring approaches for dystrophic zebrafish, further enhancing the zebrafish dmd model for rapid and cost-effective small molecule screening.

Materials and Methods

Zebrafish husbandry

All experiments involving live zebrafish (Danio rerio) were carried out in compliance with IACUC guidelines. Zebrafish were raised and staged as previously described 22. Time (hpf or dpf) refers to hours or days postfertilization at 28.5°C. The wild-type stock used was AB. The Tg(acta1a:gfp) line that labels skeletal muscle actin has been described 23. The dmdta222a mutant line (also known as sapje) has been described and is a null allele 10,24. dmdta222a genotyping was performed as described 25.

Morpholino injections

Morpholino (MO) injections were performed as previously described 26 using a Narishige IM 300 Microinjector. MOs used were dmd-MO1, 5′-TTGAGTCCTTTAATCCTACAATTTT-3′, and dmd-MO6, 5′-GCCATGACATAAGATCCAAGCCAAC-3′ 11. MOs were initially tested individually at several doses injected in a volume of 1 nl into 1-cell stage embryos. For our strong dmd-MO cocktail, MOs were combined at the following working concentrations: dmd-MO1, 4 mg/ml; dmd-MO6, 7.5 mg/ml, and injected in a volume of 1 nl into 1-cell stage embryos. Non-injected sibling embryos served as controls.

Wholemount immunocytochemistry

Whole-embryo immunostaining was performed with the following primary antibodies: anti-Dystrophin, 1:100 (Sigma D8043); anti-GFP, 1:250 (Torrey Pines TP401). Secondary antibodies used were: goat anti-mouse AlexaFluor-568 and goat anti-rabbit AlexaFluor-488 (1:250, Life Technologies/Molecular Probes). Stainings were performed as previously described 27. To quantitate anti-Dystrophin staining (Fig. 1C), staining intensity per animal was assigned a value between 0 and 3, where 0 is no staining and 3 is the brightest control staining.

Phalloidin staining

For visualizing actin, phalloidin staining was performed as previously described 15 using Alexa-488 phalloidin (Life Technologies) on dmdta222a larvae and rhodamine-phalloidin (Cytoskeleton) onTg(acta1a:gfp) larvae.

Scoring and imaging muscle lesions

For scoring numbers of animals with muscle lesions, whole embryos and larvae were either anaesthetized with tricaine 22 or fixed in 4% paraformaldehyde and viewed using an SZX16 Olympus stereomicroscope. Birefringence was imaged as previously described 28. Animals with gaps in acta1a:gfp, phalloidin, or birefringence patterns were scored as affected. For scoring numbers of myotomes with lesions per animal, larvae were mounted in 70% glycerol in PBS under a coverslip and imaged using a Leica TCS SP5 confocal microscope. 10 somites, approximately between the levels of somite 5 and 15, were imaged per animal. Somites with gaps in acta1a:gfp or phalloidin patterns were scored as affected. To image birefringence on the confocal, a single polarizing filter was placed between the filter turret and the condenser of the inverted microscope, and images were collected using the visible laser and brightfield illumination settings.

TSA treatments

A 30 µM stock solution of Trichostatin A (TSA, T8552, Sigma) was made in DMSO. Embryos were treated with 200 nM TSA in Embryo Medium (EM) 22 for three days, beginning at 24 hpf, and were fixed in 4% paraformaldehyde at 96 hpf for scoring muscle lesions. EM containing TSA or DMSO vehicle control was changed every 24 hours. Groups of 30 control or dmd-MO embryos were treated in 35 mm dishes containing 3 ml medium. Groups of approximately 80 embryos from dmdta222a heterozygote matings were treated in 60 mm dishes containing 9 ml medium.

Statistical analyses

Error bars in graphs report standard deviation, and significance was calculated using two-tailed, paired Student’s t tests.

Results

Identifying a robust zebrafish dystrophin morpholino knock down

Multiple previous studies using morpholinos to knock down zebrafish dystrophin (dmd) reported achieving a dystrophic phenotype in about 30% of morpholino-treated animals 10,13,17. We wanted to identify an improved zebrafish dmd morpholino knock-down approach that would allow for robust and rapid dmd knock down for chemical screening. Our previous studies, and those of others, have shown that cocktails of non-overlapping morpholinos targeted against a single gene can provide more robust knock down and reduce non-specific morpholino toxicity than by using individual morpholinos 26,29,30,31. Therefore, we obtained two previously described, non-overlapping morpholinos designed to target near the translation start site of zebrafish dmd, dmd-MO1 and dmd-MO6 11 (nomenclature according to ZFIN, http://zfin.org). To identify a dose for each individual morpholino to use in a cocktail, we first determined whether there were doses of the individual morpholinos that would cause non-specific effects. We injected a series of doses of each morpholino and assessed whether the animals formed a swim bladder at 5 days post fertilization (dpf), a phenotype we have previously used to assess morpholino toxicity 26. We find that even a 10 ng dose of dmd-MO1 allows larvae to form swim bladders (Fig. 1A). With dmd-MO6, doses up to 7.5 ng allow for swim bladder formation (Fig. 1A). This swim bladder test thus identifies a maximum dose of 7.5 ng for dmd-MO6. We next asked whether doses of individual MOs could cause a dystrophic phenotype at 2 dpf. We assessed dystrophic muscle by injecting the dmd MOs into embryos carrying the skeletal muscle transgene acta1a:gfp 23, which reveals muscle lesions in dmd mutants 10 (this work, Fig. 3). We observe that, for dmd-MO1, increasing doses lead to increased penetrance, with a plateau at 4 ng. For dmd-MO6, increasing doses also lead to increased penetrance (Fig. 1B). We then tested whether doses of individual MOs could reduce Dystrophin expression at 2 dpf, as assessed with anti-Dystrophin staining in larvae. We observe that, for dmd-MO1, increasing doses lead to increasing reduction of Dystrophin, while for dmd-MO6, strong reduction of Dystrophin only occurs at 7.5 ng (Fig. 1C-H). Taking into account these three analyses, we settled on doses of 4 ng for dmd-MO1 and 7.5 ng for dmd-MO6 to combine for our dmd-MO cocktail. When we inject the dmd-MO cocktail, we find loss of anti-Dystrophin staining in all injected animals (Fig. 1C, 1I). Thus, our approach to determining individual MO doses identifies a cocktail that causes strong zebrafish Dystrophin knock down. Our subsequent dmd-MO experiments refer to this MO1 4 ng/MO6 7.5 ng combination.

Fig. 1: Identification of morpholino doses for zebrafish dmd knockdown.

(A-C) Graphs of dmd morpholino (MO) effects. (A) Swim bladders were scored at 5 dpf. For each bar, n=3 with ≥25 larvae for each replicate. * P<0.04 versus control. (B) Muscle lesions were scored based on acta1a:gfp expression at 2 dpf. For each bar, n=3 with ≥30 larvae for each replicate. (C) Anti-Dystrophin staining was performed at 48 hpf. X-axes show ng of MO injected; c refers to controls. Anti-Dystrophin staining levels are arbitrary units. For each bar, n=2-3 with ≥12 larvae for each replicate. (D-I) Anti-Dystrophin staining in 48 hpf larvae. Lateral views of trunk somites show anterior to the left. Staining accumulates at myotome boundaries (Bassett et al., 2003; Guyon et al., 2003). In (C) and (I), MO1+MO6 refers to MO1 4 ng/MO6 7.5 ng combination.

The dmd-MO muscle lesion phenotype strongly resembles that of dmd mutants

To determine how well the dmd-MO cocktail phenocopies the dmd mutant, we took two approaches to analyzing muscle damage: muscle birefringence pattern and muscle actin labeling. Zebrafish muscle appears bright under polarized light, but zebrafish dmd mutants show disruptions in this birefringence due to muscle lesions caused by detachment of fibers from the myotendinous junctions 24,32,33. Similarly, muscle actin labeling with the transgene acta1a:gfp or with phalloidin reveals muscle lesions in dystrophic zebrafish 10,15. At 24 hpf, the muscle birefringence pattern is not yet developed, but dmd-MO and dmd mutant embryos show normal actin labeling (Fig. 2A-D). However, at 48 hpf and 96 hpf, both dmd-MO and dmd mutant larvae show strongly disrupted birefringence and actin labeling patterns (Fig. 2E-I). By 96 hpf, over 90% of dmd-MO larvae show disrupted birefringence and actin labeling patterns, similar to the 100% affected dmd mutants (Fig. 2E-H). Defects in actin labeling appear in a similar percentage of larvae as the disrupted birefringence (Fig. 2H-I). Thus, in dmd-MO animals, as in dmd mutants, muscle structure initially appears normal and then progressively worsens. These results show that our dmd-MO cocktail strongly phenocopies the dmd mutant muscle lesion phenotype.

Fig. 2: dmd-MO animals have a high penetrance of muscle lesions and resemble dmd mutants.

(A-D) Phalloidin staining in 24 hpf embryos. Lateral views of trunk somites show anterior to the left. (E-G) Birefringence images of 4 dpf larvae. Anterior to the left. (H-I) Quantification of larval birefringence and actin labeling patterns. For each bar, n=3-6 with ≥11 larvae for each replicate. For each condition, P<0.0006 relative to paired control sample. All larvae from dmd crosses were genotyped.

We next wanted to confirm that birefringence and actin labeling are similar reporters for the dmd muscle lesion phenotype. Using confocal microscopy to more closely examine muscle lesions, we find that there is a strong correlation among the birefringence, acta1a:gfp, and phalloidin patterns, and all three approaches reveal the same muscle lesions (Fig. 3). These results show that these three approaches are indeed similar reporters for dmd muscle lesions.

Fig. 3: Correlation among birefringence, acta1a:gfp, and phalloidin patterns.

Confocal images of a single control (A,C,E) and dmd-MO (B,D,F) larva at 4 dpf. Phalloidin staining was imaged using the red channel but false-colored in green in E,F. Lateral views of trunk somites show anterior to the left. The birefringence, acta1a-gfp, and phalloidin muscle lesion patterns strongly correlate in all larvae that were examined for all three patterns (n=8). Abnormal birefringence also correlates with lesions visualized with phalloidin staining in dmd-/- larvae (n=13, not shown).

We noticed that not all muscle segments contained muscle lesions in dmd-MO larvae, consistent with the variability described in dmd mutants 10. To more quantitatively compare muscle damage in dmd-MO larvae with that in dmd mutants, we used confocal microscopy to score the percent myotomes per larva with muscle lesions, similar to the approach previously described 15. At 48 hpf and 96 hpf, both dmd-MO and dmd mutant larvae show a similar percentage of myotomes with lesions based on disrupted actin labeling (Fig. 4). Consistent with our per larva scoring above (Fig. 2), these results show that our dmd-MO cocktail strongly phenocopies the dmd mutant muscle lesion phenotype.

Fig. 4: dmd-MO animals have a similar percentage of affected myotomes as dmd mutants.

(A-D) acta1a:gfp expression. (E-H) phalloidin staining. Lateral views of trunk somites show anterior to the left. (I) Quantification of actin labeling patterns. For each bar at 48 hpf, n=3-6 with ≥ 11 larvae for each replicate. For each bar at 96 hpf, n=3 with ≥8 larvae for each replicate. For each condition, P<0.02 relative to paired control sample. All larvae from dmd crosses were genotyped.

TSA rescues the muscle lesion phenotype of both dmd-MO larvae and dmd mutants

Because our new dmd-MO cocktail induces a high frequency of animals with muscle lesions, we wanted to test whether we could use our dmd-MO to identify small molecules that could rescue the dmd phenotype. The histone deacetylase inhibitor TSA has been shown to ameliorate Duchenne muscular dystrophy in the mouse mdx model 18,19,20,21. We therefore asked whether TSA could rescue the muscle lesion phenotype in zebrafish dmd-MO larvae and dmd mutants. We began the TSA treatments at 24 hpf, which is prior to the appearance of the dmd phenotype and the stage when previous studies have initiated other chemical treatments that rescue the zebrafish dmd phenotype 13,14. Scoring both birefringence and actin labeling, we find that TSA treatment strongly rescues the muscle lesion phenotype in both dmd-MO larvae and dmd mutants (Fig. 5). Interestingly, the TSA rescue effect is similar whether scored per larva by birefringence pattern or per myotome with actin labeling (Fig. 5I-J). We confirmed that the TSA treatment did not restore Dystrophin expression in rescued dmd-MO or dmd mutant larvae (not shown). These results show that both the dmd-MO and dmd mutant zebrafish models are similarly and robustly rescued by TSA treatment and also show that assessing the muscle birefringence pattern in whole larvae, using a stereomicroscope, is a reliable approach to scoring the dmd-MO or dmd mutant phenotype following chemical treatment.

Fig. 5: Treatment with TSA rescues dmd-MO and dmd mutant muscle lesions.

(A-D) acta1a:gfpexpression. (E-H) phalloidin staining. Lateral views of trunk somites show anterior to the left. (I-J) Quantification of larval birefringence and actin labeling patterns. For control/dmd-MO treatments, for each bar, n=6 with ≥10 larvae for each replicate. For dmd+/+/dmd-/-treatments, for each bar, n=3 with ≥8 larvae for each replicate. * P<0.03. ** P<0.02. *** P<0.009. **** P<0.002. All larvae from dmd crosses were genotyped.

Discussion

Our results identify a dmd morpholino cocktail that induces a high penetrance of muscle lesions and strongly resembles the zebrafish dmd mutant phenotype. We also show that the HDAC inhibitor TSA rescues both dmd-MO and dmd mutant muscle lesions. By comparing different approaches to scoring muscle lesions, our study confirms a previous report 13 that simple assessment of the muscle birefringence pattern in whole larvae, using a stereomicroscope, is a reliable approach to scoring the dmd-MO or dmd mutant phenotype following chemical treatment. Thus, our work identifies optimal morpholino and phenotypic scoring approaches for dystrophic zebrafish, further enhancing the zebrafish dmd model for rapid and cost-effective small molecule screening.

Our work provides additional support for using cocktails of non-overlapping morpholinos targeted against a single gene in order to provide more robust knock down and reduce non-specific morpholino toxicity than by using individual morpholinos 26,29,30,31. We previously reported using swim bladder scoring as an approach to identify working concentrations of individual morpholinos in zebrafish 26. Swim bladder scoring helped us to identify a working concentration for one of our dmd MOs (dmd-MO6, see Fig. 1A), but dmd-MO1 does not affect swim bladder formation, even at very high concentrations (Fig. 1A). Therefore, we used two additional approaches, dmd muscle phenotype and Dmd protein expression, to help determine the MO working concentrations. Based on this and our previous studies 26,31, we suggest identifying working concentrations of individual MOs that each have subtle effects on embryonic or larval phenotype for use in cocktails.

Several previous studies have demonstrated quantitative approaches to scoring dystrophic muscle lesions in zebrafish 13,14,15,33. For chemical screening, it is ideal to optimize the cost and speed of the screening approach while still allowing for reliable phenotypic scoring. We find that simple assessment of the muscle birefringence pattern in whole larvae, using a stereomicroscope, provides a comparable measurement of dmd phenotypic rescue by TSA treatment as the more quantitative, but time-consuming, approach of scoring myotome lesions using the confocal microscope. Also, in spite of having to inject embryos to generate dmd-MO larvae, we find that assaying dmd-MO larvae for rescue by TSA saves time and resources over using dmd mutant larvae, which require genotyping. There are, however, three potential disadvantages to using dmd-MO larvae. First, although we have not examined the dmd-MO phenotype past 5 dpf, the dmd-MO larvae likely are not useful for examining longer-term chemical rescue effects because the morpholinos will become diluted. Second, without using more quantitative measures of birefringence 14,33, it is possible we would not be able to distinguish phenotypic rescue by a drug that subtly improved muscle structure but still caused reduced birefringence. Third, even with the high penetrance of the dmd-MO phenotype, there are still approximately 20% or so larvae that appear not affected, whereas 0% of dmd mutants are not affected. This could be an important consideration for drug screening, where, if small numbers of fish were screened, the incomplete dmd-MO penetrance might cause false positive hits. Nevertheless, our work helps identify optimal morpholino and phenotypic scoring approaches for dmd drug screening. Candidate drugs that rescue dmd-MO larvae can be retested on dmd mutants.

The ability of TSA to ameliorate muscular dystrophy in the mouse mdx model may work through more than one mechanism. Initial studies of TSA and mdx rescue suggested that TSA acted through promoting upregulation of follistatin expression in satellite cells 18. A recent study, however, showed that fibroadipogenic progenitor cells mediate the ability of TSA to ameliorate muscular dystrophy in young mdx mice 21. Our demonstration that TSA can rescue the zebrafish dmd model now provides an additional model system for further mechanistic analysis of how TSA, and other small molecules, function to ameliorate dystrophic muscle.

Correspondence

Lisa Maves. Email: lmaves@u.washington.edu

Author Contributions

NMJ, GHF and LM conceived the experiments. NMJ and GHF performed the experiments. NMJ, GHF and LM analyzed the data. LM wrote the paper with contributions from NMJ and GHF.

Competing Interest Statement

The authors have declared that no competing interests exist.

]]>
http://currents.plos.org/md/article/the-hdac-inhibitor-tsa-ameliorates-a-zebrafish-model-of-duchenne-muscular-dystrophy/feed/ 0
Stem Cell Antigen-1 in Skeletal Muscle Function http://currents.plos.org/md/article/stem-cell-antigen-1-in-skeletal-muscle-function/ http://currents.plos.org/md/article/stem-cell-antigen-1-in-skeletal-muscle-function/#respond Thu, 15 Aug 2013 10:28:50 +0000 http://currents.plos.org/md/?post_type=article&p=4095

INTRODUCTION

In response to skeletal muscle damage, resident myogenic progenitors undergo activation to form a pool of proliferating myoblasts. These mononuclear myoblasts differentiate and fuse, forming multinucleated myocytes, which repair or replace the damaged tissue 1,2,3 . This programmed series of events is essential to maintaining tissue homeostasis during exercise and aging, and to ensuring recovery from muscle trauma 4 . While the balance between myoblast proliferation and differentiation is critical to muscle repair, its regulation is incompletely understood.

We previously identified Stem cell antigen-1 (Sca-1; also known as Ly-6A/E) during an expression screen to identify genes regulating myoblast cell cycle withdrawal during differentiation 5 . Sca-1 is a member of the Ly-6 multigene family encoding a number of highly homologous, glycosyl-phosphatidylinositol (GPI)-anchored surface membrane proteins, and is widely used as a marker of murine hematopoietic stem cells 6,7,8. Beyond its role as a stem cell marker, it has been shown that overexpression of Sca-1 inhibits proliferation of CD4+ T-cells (9), as well as differentiation of hematopoietic stem cells 6,10,11,12 . Sca-1-/- mice are viable, however, they exhibit immune and hematopoietic defects 6,10,11,12 . Specifically, these mice demonstrate a lymphocytosis and thrombocytopenia, and isolated Sca-1-/- T-cells undergo prolonged hyperproliferation with stimulation in vitro 11 . Consistent with a role in progenitor cell maintenance, Sca-1-null animals have a reduced ability to re-populate bone marrow after serial transplantation 6,12 and develop age-related failure of osteogenesis 10 .

Sca-1 also is expressed on the surface of muscle-derived stem cells 13,14 and myogenic precursors recruited to sites of skeletal or cardiac muscle injury 13,15,16,17,18 . We previously reported that inhibition of Sca-1 expression by antisense or Sca-1 interference with blocking antibodies stimulated myoblast proliferation and delayed myoblast fusion in vitro19. Subsequently, others observed sustained proliferation in Sca-1-/- myoblasts cultured ex vivo 20 . Using a myonecrotic injury model in Sca-1-/- and Sca-1+/+ mice, we then showed that Sca-1 regulates the tempo of muscle repair by controlling the balance between proliferation and differentiation of activated myoblasts 21 .

Despite its function in stem cell proliferation in general, and more specifically in myoblast expansion during secondary myogenesis, a role for Sca-1 in normal, post-natal muscle function has not been apparent. To explore this, we undertook the systematic comparison of Sca-1-/-and Sca-1+/+mice and hindlimb muscles to elucidate the tissue, mechanical, and functional effects of Sca-1 in young adult and aging animals.

MATERIALS AND METHODS

Animals. All animal procedures were approved by the Institutional Animal Care and Use Committee at the University of California, San Francisco, in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care International guidelines. Mice heterozygous at the Sca-1 locus were graciously provided by Patrick Flood (University of North Carolina) 11 , and backcrossed to BALB/c strain for ten generations. Sca-1+/- littermates were bred to homozygosity. Genotypes were confirmed for all experimental animals by Southern blot analysis, as previously described 21 . All experiments were performed on 12-16 week-old or 47-50 week-old female mice. Wild type (Sca-1+/+) female BALB/c littermates from heterozygous matings were used as controls in all experiments.

Muscle volume analysis. Animals were euthanized with intraperitoneal injection of pentobarbital followed by cervical dislocation. Following euthanization, skinned hindlimbs were fixed in 10% paraformaldehyde, then dehydrated in ethanol and embedded in paraffin. Hindlimbs were sectioned at 5 μm perpendicular to the myofiber through the length of the limb. A series of every tenth section was selected and processed for staining. Sections were deparaffinized and stained with hematoxylin and eosin to identify tissue architecture. Using accepted anatomic boundaries, the relevant muscle was traced at low power (4X) using the live image generated with a Zeiss Axiovision fluorescent microscope (Carl Zeiss; Thornwood, NY) equipped with a Hamamatsu Orca2 digital camera (Hamamatsu; Bridgewater, NJ). Volume estimates were calculated using the Cavalieri principle 22 and contours traced at low power. Both hindlimbs from at least 3 animals were analyzed for each genotype/age.

Myofiber CSA and perimeter analysis. Tissue sections were prepared as described above. A series of every tenth section was selected and processed for staining. Sections were deparaffinized in xylene and dehydrated with ethanol before antigen retrieval with proteinase K for 10 minutes at room temperature (RT). After washing out proteinase K, sections were incubated with 3% H2O2 in methanol for 10 min at RT, Rodent Block M (BioCare; Concord, CA) for 30 min at RT, DAKO antibody diluent (DAKO; Carpinteria, CA) for 30 min at RT, rabbit anti-laminin (Sigma-Aldrich; St. Louis, MO) at 1:100 dilution in DAKO antibody diluent (DAKO; Carpinteria, CA) overnight at 4°C, then washed with phosphate-buffered saline. To develop peroxidase, sections were incubated with Rabbit HRP Polymer (BioCare; Concord, CA), then DAB substrate (BioCare; Concord, CA) according to the manufacturer’s instructions. Sections were counterstained with hematoxylin before dehydration with ethanol (80, 90, 100%). To quantitate myofiber cross-sectional area (CSA) and perimeter within a section, ≥300 myofibers within a centrally located field (20X; ~300,000 µm2) within the tibialis anterior, extensor digitorum longus, or soleus muscle were analyzed using Stereo Investigator software (Microbrightfield; Williston, VT). Both hindlimbs from at least 3 animals were analyzed for each genotype/age.

Pax7 analysis. Paraffin-embedded tissue sections were prepared as described above. Sections were incubated with Rodent Block M (BioCare; Concord, CA) for 30 min at RT, blocking buffer (2% goat serum, 1% bovine serum albumin, 0.1% fish gelatin, 0.1% Triton X-100, 0.1% glycine) for 30 min at RT, 10 µg/ml Alex488-conjugated anti-Pax7 (R&D Systems; Minneapolis, MN) in blocking buffer without glycine for 60 min at RT, then washed with phosphate-buffered saline, counterstained with DAPI, and dehydrated with ethanol (70, 80, 90, 100%). Immunostained sections were imaged by randomly placing the 20X objective within the tibialis anterior muscle within the section. Following imaging, the field was manually moved a fixed distance of approximately 600 µm in the horizontal and then vertical axis, resulting in 4–6 counting images per section. DAPI+Pax7+ nuclei localized beneath the basal lamina of myofibers were confirmed at 60X magnification, then counted and expressed as a percentage of total DAPI+ nuclei. Both hindlimbs from at least 3 animals were analyzed for each genotype/age.

Arteriole number and CSA analysis. Tissue sections were prepared as described above. Sections were incubated with Rodent Block M (BioCare; Concord, CA) for 30 min at RT, DAKO antibody diluent (DAKO; Carpinteria, CA) for 30 min at RT, rat anti-CD31 (BioCare; Concord, CA) at 1:25 dilution in DAKO antibody diluent (DAKO) overnight at 4°C, then washed with phosphate-buffered saline. To develop peroxidase, sections were incubated with Rat HRP Polymer (BioCare; Concord, CA) followed by DAB substrate (BioCare; Concord, CA) according to the manufacturer’s instructions. After rinsing with phosphate-buffered saline, sections were then re-blocked with Rodent Block M (BioCare; Concord, CA) for 30 min at RT and DAKO antibody diluent (DAKO; Carpinteria, CA) for 30 min at RT, and incubated with mouse anti-SMA (BioCare; Concord, CA) at 1:50 dilution for 60 min at RT. Sections were then incubated with Mouse AP Polymer (BioCare; Concord, CA) and developed with Vulcan Red (BioCare; Concord, CA) according to the manufactorer’s instructions. Sections were counterstained with hematoxylin before dehydration with ethanol (70, 80, 90, 100%). Immunostained sections were imaged using the 4X objective. Arteriole number was calculated as the total number of SMA+CD31+ vessels per section. Each vessel was confirmed at 40X. To quantitate arteriolar CSA, all identified arterioles were analyzed using Stereo Investigator software (Microbrightfield; Williston, VT). Both hindlimbs from at least 3 animals were analyzed for each genotype/age.

Fibrosis. Tissue sections were prepared as described above. A series of every tenth section was selected and processed for staining. Sections were deparaffinized in xylene and dehydrated with ethanol before staining with aniline blue, hematoxylin, and scarlet acid fuchsin with Masson’s Trichrome 2000 stain kit (American MasterTech, Lodi, CA) according to the manufacturer’s instructions. Stained sections dehydrated with alcohol, cleared with xylene, and imaged using the 4X objective. To quantitate CSA of fibrosis, all stained areas were analyzed using Stereo Investigator software (Microbrightfield; Williston, VT). Both hindlimbs from at least 3 animals were analyzed for each genotype/age. Differences between genotype/age groups were tested for significance by one-way analysis of variance.

Ki67 immunostaining. Tissue sections were prepared as described above. A series of every tenth section was selected and processed for staining. Sections were deparaffinized in xylene and dehydrated with ethanol before antigen retrieval with citrate buffer for 20 minutes at 37°C. After washing with phosphate-buffered saline, sections were incubated with 3% H2O2in methanol for 20 min at RT, Rodent Block M (BioCare; Concord, CA) for 30 min at RT, DAKO antibody diluent (DAKO; Carpinteria, CA) for 30 min at RT, rat anti-Ki67 (DAKO; Carpinteria, CA) at 1:10 dilution in DAKO antibody diluent (DAKO; Carpinteria, CA) overnight at 4°C, then washed with phosphate-buffered saline. To develop peroxidase, sections were incubated with Rat HRP Polymer (BioCare; Concord, CA), then DAB substrate (BioCare; Concord, CA) according to the manufacturer’s instructions. Sections were counterstained with hematoxylin then washed with phosphate-buffered saline, counterstained with DAPI, and dehydrated with ethanol. Immunostained sections were imaged by randomly placing the 20X objective within the section. Following imaging, the field was manually moved a fixed distance of approximately 600 µm in the horizontal and then vertical axis, resulting in 4–6 counting images per section. DAPI+Ki67+nuclei localized beneath the basal lamina of myofibers were confirmed at 60X magnification, then counted and expressed as a percentage of total DAPI+nuclei. Both hindlimbs from at least 3 animals were analyzed for each genotype/age. Differences between genotype/age groups were tested for significance by one-way analysis of variance.

Exercise. To evaluate voluntary exercise, mice were allowed to run at will during normal light-dark cycles as established at the UCSF Laboratory Animal Resource Center. Distance and duration over consecutive 24h periods were recorded with individual FX10 cycle computers (E3 Cycling; Chapel Hill, NC). To evaluate forced exercise, mice were forced to run on each day over a 37d period on a Lafayette Mouse Forced Exercise Run/Walk Wheel System (Lafayette Instrument; Lafayette, IN) at 8 meters/min for five 10 min intervals with 30s rests between intervals. Over the next five 10 min intervals immediately following during that day’s exercise session, speeds were progressively increased until mice reached their maximum rate, as determined by the maximum speed (meters/min) at which mice continued to run on the wheel without falling off or hanging onto the wheel.

Myomechanical analysis. Ex vivo muscle analysis was performed as previously described 23 . Following euthanization, the EDL muscle was extracted and placed in a bath containing Krebs Hensleit buffer within 15 min. The muscle was attached by opposing tendons to a DMT Model 820MS force transducer (Danish Myo Technology; Ann Arbor, MI) filled with Krebs Hensleit buffer prewarmed to 25oC and bubbled with O2/CO2 (95%/5%) for ≤15 min prior to use. During the mounting process, the muscle was only handled through the suture, without direct contact to the muscle to prevent damage to the muscle fiber. The muscle was stimulated by a Grass Model S48 square pulse electrical stimulator (Grass Technologies; West Warwick, RI) and the data analyzed and projected using a custom acquisition platform (ADInstruments PowerLab Data Acquisition System and LabChart software; ADInstruments; Colorado Springs, CO).

Twitch tension (Pt) was recorded by first stretching the EDL muscle until there was no laxity in the muscle fiber. A square stimulation of 0.5 ms duration was used to induce twitch. The voltage was increase incrementally until maximal twitch tension was achieve and then the voltage was set at 20% above the maximum to induce a supramaximal stimulus (mean supramaximal stimulus was 40 volts). Optimum length of the muscle was determined by carefully stretching the muscle and recording the twitch response after square stimulation until maximal twitch was recorded. The muscle was left to equilibrate at the optimal length for 3 min before another supramaximal stimulus was applied and the output recorded as the twitch force. Tetanic tension (Po) was recorded by applying a train of supramaximal stimuli for 300 ms at 150 Hz.

Force-frequency fatigue was measured by exposing muscle to a supramaximal stimulus train of 3-5 pulses (300 msec duration separated by 3 sec) at successive frequencies (30, 60, 100 and 140 Hz), with 5 min intervals between stimulations. For a given frequency stimulus and muscle, the maximum pulse was chosen and then normalized relative to the peak response over all pulse frequencies. An average force-frequency diagram was then constructed from all normalized muscle responses. Force-frequency stimulation produces a maximum response at a given frequency that may fall off precipitously at other frequencies. The fatigue characteristic of a given group would be interpreted as a significant reduction in the peak relative to the other groups at the various frequencies.

Low-frequency time-fatigue was measured by supramaximal tetanic muscle stimulation at low frequency (60 Hz) for a duration of 300 msec, repeated every 3 seconds for a period of 10 min. The low-frequency time-fatigue curve produces a peak response at a given time that is followed by a multi-exponential decay in contraction force. Fatigue in a given group would be interpreted as a significant reduction in percent maximum contraction force over time, assuming that the different groups experienced the same average peak force at the onset. We sampled the percent of maximal force of contraction at given periods (0.5 , 1, 1.5, 2-10 min in 1 min increments).

Muscle mass, cross-sectional area, and length were measured after the fatigue analysis to avoid handling prior to analysis.

Statistical analysis.For volume, CSA, perimeter, and cell number analyses, differences between genotype/age groups were tested for significance by one-way analysis of variance (ANOVA), followed by unpaired t-test with Bonferroni correction to isolate specific differences. A value of p <0.05 was considered significant. For exercise studies, analysis of co-variance was used to test for differences between animals within groups. Mean slopes of each group were then compared using Tukey’s test.

For myo-mechanical analysis, variables for twitch, tetanus and fatigue were first subjected to a nested ANOVA to evaluate whether left and right leg measurements associated with each genotype (WT vs. KO) and age (Young vs. Aging) were similar. If so, then the left and right leg variables of an animal were averaged and the genotype and age groups were evaluated by two-way ANOVA. For the tetanus analysis, a repeated measure ANOVA was performed on the basis of the stimulus frequency of contraction (30, 60, 100, 160 Hz), which included interactions between the genotype, age and frequency. The fatigue analysis included time of maximal contraction (0.5, 1, 1.5, 2-10 min in 1 min increments). If significant differences were present between the means of groups, a multiple comparison test was performed utilizing multilple t-tests with the number of comparisons adjusted by Bonferroni correction. A value of p <0.05 was considered significant.

RESULTS

Morphometric analysis of hindlimb muscle from wild type and Sca-1 KO mice.To examine the effects of Sca-1 on body mass and muscle homeostasis over the lifespan, we evaluated 12-16 week-old adult Sca-1-/- (KO) and Sca-1+/+ (WT) mice compared with 47-50 week-old aging KO and WT mice. There was no significant difference in body weight (BW) or hindlimb weight (HW) between KO and WT mice at either age (KO young vs. WT young: BW 29.7±0.9 vs. 27.6±1.7 gms, p>0.05, HW 0.41±0.06 vs. 0.49±0.06 gms, p>0.05; KO aging vs. WT aging: BW 28.1±2.1 vs. 28.7±3.4 gms, p>0.05, HW 0.46±0.06 vs. 0.40±0.09 gms, p>0.05), or between young and aging mice (KO young vs. KO aging: BW 29.7±0.9 vs. 28.1±2.1 gms, p>0.05, HW 0.41±0.06 vs. 0.46±0.06 gms, p>0.05; WT young vs. WT aging: BW 27.6±1.7 vs. 28.7±3.4 gms, p>0.05, HW 0.49±0.06 vs. 0.40±0.09 gms, p>0.05) (Fig. 1). We also compared volumes of tibialis anterior (TA) and extensor digitorum longus (EDL) muscles between groups with unbiased stereology using the Cavalieri method. This demonstrated no significant difference between young and aging mice (TA-KO young vs. KO aging: 16.8±2.3 vs. 19.7±1.3 mm3, p>0.05; WT young vs. WT aging: 20.3±0.4 vs. 16.2±2.2 mm3, p>0.05; EDL-KO young vs. KO aging: 9.8±2.2 vs. 10.6±1.8 mm3, p>0.05; WT young vs. WT aging: 7.3±2.0 vs. 8.1±1.3 mm3, p>0.05), or between genotypes at either age (TA-KO young vs. WT young: 16.8±2.3 vs. 20.3±0.4 mm3, p>0.05; KO aging vs. WT aging: 19.7±1.3 vs. 16.2±2.2 mm3, p>0.05; EDL-KO young vs. WT young: 9.8±2.2 vs. 7.3±2.0 mm3, p>0.05; KO aging vs. WT aging: 10.6±1.8 vs. 8.1±1.3 mm3, p>0.05) (Fig. 2).

Fig. 1: Body and hindlimb mass in wild type and Sca-1 KO mice.

Animals were weighed immediately following euthanasia. Then hindlimbs were severed at the knee and ankle joints and weighed. No significant differences were observed between genotypes or ages.

Fig. 2: Hindlimb muscle volume in wild type and Sca-1 KO mice.

Sections were stained with hematoxylin and eosin. A typical hindlimb section is shown (left). Unbiased stereology was used to measure muscle volumes by the Cavalieri method (right). Data shown are mean±s.e.m. (N=6 hindlimbs). No significant differences were observed between genotypes or ages. EDL, extensor digitalis longus; TA, tibialis anterior.

We previously had demonstrated that Sca-1 regulates the tempo of myoblast expansion during secondary myogenesis following injury, and that regenerated KO muscle tissue is hyperplastic compared with wild type regenerated tissue 21 . To determine whether there was a difference in myofiber size between genotypes and at different ages, we compared the TA, EDL, and soleus (Sol) myofiber cross-sectional areas and perimeters between KO and WT mice at either age, or between young and aging mice. We observed that myofibers in young KO mice had greater cross-sectional areas (TA-KO vs. WT: 0.13±0.001 vs. 0.06±0.003 nm2, p<0.01; EDL-KO vs. WT: 0.18±0.003 vs. 0.06±0.004 nm2, p<0.01; Sol-KO vs. WT: 0.11±0.002 vs. 0.06±0.004 nm2, p<0.01) and perimeters (TA-KO vs. WT: 11.2±1.2 vs. 8.8±0.3 nm, p<0.05; EDL-KO vs. WT: 14.2±1.3 vs. 8.6±0.7 nm, p<0.01; Sol-KO vs. WT: 10.4±1.2 vs. 8.5±0.9 nm, p<0.05) than young WT mice, but that this difference reversed with aging (cross-sectional areas/TA-KO vs. WT: 0.04±0.002 vs. 0.05±0.001 nm2, p<0.05; EDL-KO vs. WT: 0.03±0.006 vs. 0.06±0.0.002 nm2, p<0.05; Sol-KO vs. WT: 0.04±0.006 vs. 0.06±0.001 nm2, p<0.05; perimeters/TA-KO vs. WT: 5.5±0.5 vs. 7.9±0.7 nm, p<0.05; EDL-KO vs. WT: 5.2±0.2 vs. 8.5±0.7 nm, p<0.05; Sol-KO vs. WT: 6.3±0.3 vs. 8.4±0.8 nm, p<0.05) (Fig. 3).

Fig. 3: Hindlimb myofiber dimension in wild type and Sca-1 KO mice.

Sections were stained with anti-laminin antibody (brown) and nuclear counterstaining with DAPI (blue). A typical hindlimb section (top left), and representative micrographs (10X) from tibialis anterior (TA) muscle (bottom left) are shown. Muscle fiber cross-sectional areas (CSA; top right) and perimeters (bottom right) were measured. Data shown are mean±s.e.m. (N=6 hindlimbs). While KO young myofibers were larger in all muscles examined than WT young fibers, KO aging myofibers were smaller than WT aging fibers. *, p<0.05; **, p<0.01; EDL, extensor digitalis longus; TA, tibialis anterior; Sol, soleus.

Satellite cell number in wild type and Sca-1 KO mice.To determine whether there is an age-dependent difference in the number of resident satellite cells between KO and WT mice, we examined the number of Pax7+ nuclei localized to the satellite cell compartment in the TA muscles of KO versus WT, young versus aging animals by unbiased stereology. There was a trend toward an increase in Pax7 number in young KO versus WT muscle (KO vs. WT: 0.034±0.007 vs. 0.025±0.005 Pax7+/DAPI+ nuclei, p>0.05), this was not significant, and this difference decreased in magnitude in older animals (KO vs. WT: 0.026±0.003 vs. 0.030±0.001 Pax7+/DAPI+ nuclei, p>0.05) (Fig. 4).

Fig. 4: Pax7+ cells in hindlimbs of wild type and Sca-1 KO mice.

Sections were stained with anti-Pax7 antibody. A typical micrograph is shown (20X); arrows indicate Pax7+DAPI+ nuclei (top). Data shown are mean±s.e.m. (N=4 hindlimbs) (bottom). No significant differences were seen between genotypes or ages.

Vascularity in hindlimb muscle of wild type and Sca-1 KO mice.Sca-1 is expressed on the surface of a variety of stem cells 24 , including vascular endothelial progenitors 25 . To determine whether muscle vascularity was affected by Sca-1 expression, we compared the number of arterioles and the total arteriolar cross-sectional area in TA muscles of young versus aging, KO versus WT mice. We found that while the number of arterioles increased with age (KO young vs. KO aging: 28.8±6.7 vs. 81.6±10.5 SMA+CD31+ vessels/section, p<0.001; WT young vs. WT aging: 30.2±4.0 vs. 70.29.5 SMA+CD31+ vessels/section, p<0.001), the total arteriolar cross-sectional area decreased (KO young vs. KO aging: 590.5±116.1 vs. 264.4±88.5 µm2, p<0.01; WT young vs. WT aging: 569.1±72.0 vs. 259.0±72.9 µm2, p<0.01) (Fig. 5). There was no difference, however, between KO and WT mice (KO young vs. WT young: 590.5±116.1 vs. 569.1±72.0 µm2, p>0.05; KO aging vs. WT aging: 264.4±88.5 vs. 259.0±72.9 µm2, p>0.05) (Fig. 5).

Fig. 5: Hindlimb vascularity in wild type and Sca-1 KO mice.

Sections were stained with anti-SMA (brown) and –CD31 (pink) antibodies. Typical micrographs are shown (40X) (left). A significant increase in the number of vessels, accompanied by a decrease in total vascular cross-sectional area (CSA) was observed with age, however, this was independent of genotype (right).

Fibrosis and myoblast proliferation in hindlimb muscle of wild type and Sca-1 KO mice.Previously, we had shown that KO myoblasts activated during the response to injury have a prolonged proliferative phase with resulting hyperplasia during healing 21 . To determine whether muscle homeostasis, or secondary myogenesis with normal use, resulted in a difference in fibrosis with age, we compared the extent of fibrotic tissue in cross-sections from young and aging, KO and WT hind limb muscle following a 37 day course of voluntary and forced exercise (described below; Figs. 8, 9). We found no difference between genotypes or at different ages (Fig. 6). To determine whether upregulated proliferation during secondary myogenesis in KO animals leads to exhaustion of proliferating myoblasts with age, we compared the number of Ki67+ proliferating cells in the TA muscles of these mice, and found no significant differences (Fig. 7).

Fig. 6: Hindlimb fibrosis in wild type and Sca-1 KO mice.

Hindlimbs were skinned, fixed and embedded in paraffin, and sectioned prior to staining with Masson’s trichrome. Typical hindlimb sections are shown; muscle and intercellular fiber (red), collagen (blue), nuclei (black). Unbiased stereology was used to measure scar (blue) volumes in all genotypes and ages by the Cavalieri method (data not shown). No significant differences were observed between groups.

Fig. 7: Ki67+ cells in hindlimbs of hindlimbs of wild type and Sca-1 KO mice.

Hindlimbs were skinned, fixed and embedded in paraffin, and sectioned prior to immunohistochemical staining with anti-Ki67 antibody (brown) and DAPI (blue). A typical micrograph is shown (60X oil); arrow indicates Ki67+DAPI+ nucleus. No significant differences were seen between genotypes or ages.

Exercise capacity in wild type and Sca-1 KO mice.To assess the functional significance of the difference in myofiber size between young and aging KO versus WT mice, we evaluated exercise capacity of these animals during voluntary (Fig. 8) and forced (Fig. 9) exercise. Mice were allowed to run at will during normal light-dark cycles over consecutive 24h periods. For voluntary distance, analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (Day versus Distance: WT young 64.4, KO young 43.6, WT aging 41.9, KO aging 29.5) were statistically significantly different from each other (p<0.0001), except for WT aging and KO young, which were statistically the same (p>0.05) (Fig. 8). For voluntary duration, analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (Day versus Duration: WT young 127.0, KO young 105.5, WT aging 81.5, KO aging 60.0) were statistically significantly different from each other (p<0.01) (Fig. 8). Voluntary exercise showed that young KO mice achieved significantly less distance and duration than their WT counterparts, and resembled aging WT animals. Similarly, aging KO mice achieved significantly less distance and duration than aging WT mice.

Fig. 8: Voluntary exercise in wild type and Sca-1 KO mice.

Mice were allowed to run at will during normal light-dark cycles. Each line represents data from one animal. Distance and duration over consecutive 24 hr periods were recorded. Distance (meters/day; top) and duration (min/day; bottom) were plotted for individual animals over a 37 d period. For voluntary distance (top), analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (Day versus Distance) were statistically significantly different from each other as determined using Tukey’s test, except for WT aging and KO young, which were the same. For voluntary duration (bottom), analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (Day versus Duration) were statistically significantly different from each other as determined by Tukey’s test.

On each day over a 37d period, mice also were forced to run using a protocol where speeds were progressively increased until mice reached their maximum rate. Analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (log(Day) versus Maximum Rate2: WT young 45.6, KO young 41.1, WT aging 32.4, KO aging 21.6) were statistically significantly different from each other (p<0.01) (Fig. 9). Forced exercise similarly demonstrated that young KO mice achieved a lower maximum exercise rate than young WT mice, as did aging KO animals compared to aging WT controls.

Fig. 9: Forced exercise in wild type and Sca-1 KO mice.

On each day over a 37 d period, mice were forced to run at 8 meters/min for five 10 min intervals with 30 sec rests between intervals. Over the next five 10 min intervals immediately following that day’s exercise session, speeds were progressively increased until mice reached their maximum rate, as determined by the maximum speed (meters/min) at which mice continued to run on the wheel without falling off or hanging onto the wheel. Each line represents data from one animal. Analysis of co-variance demonstrated no statistically significant differences between animals within any group, but the mean slopes of all groups (log(Day) versus Maximum Rate2) were statistically significantly different from each other as determined using Tukey’s test.

Myomechanical analysis of hindlimb muscle from wild type and Sca-1 KO mice. To quantify changes in muscle function observed during exercise, we performed myomechanical analysis of isolated EDL muscle from KO and WT young and aging mice. There were no significant differences in muscle mass or muscle cross-sectional area between ages and genotypes, consistent with data from isolated hind limbs (Fig. 1) and unbiased stereology assessment of muscle volumes by Cavalieri method (Fig. 2). Aging animals (KO aging 3.4±0.4 x 104, WT aging 3.9±0.6 x 104) had significantly lower muscle:body mass ratios compared with young animals (KO young 3.9±0.6 x 104, WT young 5.5±0.5 x 104; p<0.05), however, this difference was similar for both KO and WT mice.

Analysis of muscle twitch demonstrated that aging mice (KO aging 115±5 mN, WT aging 101±7 mN) generated higher maximal tension (Pt) than young animals (KO young 84±7 mN, WT young 89±6 mN; p<0.05), but again there was no difference between KO and WT mice. Aging mice (KO aging 23±1 ms, WT aging 24±1 ms) also had shorter contraction times (CT) than young mice (KO young 29±1 ms, WT young 25±1 ms; p<0.05), but with no genotype-specific difference. There were no significant differences in specific maximal twitch tension (sPt) or half-relaxation time (HRT) between animals. Analysis of tetanus revealed similar results, with an age-dependent increase in maximal tetanic tension (P0; KO aging 372±15 mN, WT aging 357±22 mN vs. KO young 297±23 mN, WT young 302±18 mN; p<0.05) and maximal rate of rise of tetanus (MRRT; KO aging 21.1±0.9 mN/ms-1, WT aging 19.5±1.3 mN/ms-1 vs. KO young 16.0±1.3 mN/ms-1, WT young 17.5±1.0 mN/ms-1; p<0.05). There were no significant differences in specific maximal tetanic tension (sP0) or Pt/P0 ratio between animals.

Force-frequency analysis of isolated EDL muscle from young and aging KO and WT mice showed no significant difference between genotypes or ages (Fig. 10 left). A low-frequency fatigue protocol, however, showed a significant difference between KO young muscle and all other groups (p<0.05), with KO young muscle generating a greater percent of maximal force, and less fatigue, at all time points over 10 min (Fig. 10 right). We also examined the percent maximum force generated with relaxation during low-frequency fatigue, to determine whether there was any difference in the return to baseline between contractions between the groups. Although the force generated during relaxation trended upward over time (i.e., incomplete relaxation with fatigue), this showed no significant differences between groups (Fig. 10 right).

Fig. 10: Ex vivo myomechanical analysis of EDL muscle from wild type and Sca-1 KO mice.

Extensor digitorum longus muscle was dissected, mounted in a force transducer, and stimulated as decribed in Methods. Data shown are mean percent maximum force (N=12 young WT; N=2 young KO; N=3 aging WT; N=9 aging KO). No differences were observed in force-frequency relationships between genotypes or ages (left). Young KO animals demonstrated significantly greater maximum contraction (C) force (less fatigue) over time compared with other groups, although relaxation (R) force was the same between groups (right). Absolute peak force at the beginning of the study was statistically similar for all groups (KO Young 198±40 mN, KO Aging 234±66 mN, WT Young 208±52 mN, WT Aging 186±30 mN). *, p

DISCUSSION

While previous studies have provided focused investigations of Sca-1 KO animals 6,10,11,20,21,26 , this represents the first comprehensive analysis of the tissue, mechanical, and functional effects of Sca-1 gene disruption on murine skeletal muscle. Compared with wild type animals, we found that myofiber size was significantly increased in young KO mice, but then significantly decreased with age. Consistent with this difference, we observed greater force generation in young KO hindlimb muscle versus wild type or aging KO muscle. Interestingly, both young and aging KO animals demonstrated decreased conditioning with exercise compared with wild type littermates.

Our observations of myofiber area and perimeter are similar to another study that demonstrated a 19% increase in mean cross-sectional area in 2- to 4-month-old Sca-1 KO TA muscles, and a 13% decrease in mean cross-sectional area in 1-year-old Sca-1 KO TA muscles 20 . Those authors suggested that this difference may be the result of differences in satellite cell number or function, however, we saw no alteration in the number of Pax7+ cells beneath the basal lamina of WT versus KO or young versus aging muscles. In addition, we observed no difference in the number of proliferating, Ki67+ cells between various groups, suggesting that the difference in myofiber size with genotype or age is not due to alterations in number of satellite cells or proliferating myoblasts. More recently, Kafadar et al. showed that Sca-1 KO muscle displays reduced matrix metalloproteinase activity, suggesting that the effect of Sca-1 on myofiber size may be mediated through extracellular matrix interactions 26 .

Sca-1 is normally expressed on vascular endothelial progenitor cells 25 , so we also wanted to evaluate whether an alteration in vascularity in Sca-1 KO muscle might be responsible for the difference in myofiber development in young KO animals. However, we found no difference in vascular cross-sectional area or vessel number between WT and KO animals at either age, although a shift from fewer, larger vessels to more numerous, smaller vessels with age was observed independent of Sca-1 expression. We did not directly examine vascularity or composition of the myocardium in Sca-1 KO animals. We recently showed that Sca-1-expressing cells differentiate into multiple myocardial cell types in vitro, and that these cells induce angiogenesis and differentiate into endothelial and smooth muscle cells in mouse hearts following myocardial infarction 27. It is possible that Sca-1 deficiency in KO animals also may be playing a role in physical conditioning as a function of cardiovascular fitness.

Analysis of twitch and tetanic tension generated by Sca-1 KO muscle compared with WT showed no significant differences. In addition, the response to forced-frequency fatigue was preserved in KO as well as older muscle. However, in the case of time-dependent fatigue at low frequency stimulation, we did observe that young KO muscle generated greater force at most stimulation frequencies compared with young WT or aging WT and KO muscle. This coincided with an increase in myofiber size seen in young KO muscle. Interestingly, the relative decrease in myofiber size in aging KO animals did not lead to a downward shift in the force-frequency curve. It is possible that the myofiber environment may play a greater role in myomechanical function than currently presumed. The positive regulatory effects of Sca-1 signaling on matrix metalloproteinases has been observed in regenerating muscle (26). Further investigation into the impact of Sca-1/extracellular matrix interactions on myomechanical function are needed.

Our experiments showed that the absence of Sca-1 affected exercise performance in vivo, and that age compounded these effects. Young KO mice exercised for shorter time periods and ran shorter distances during voluntary exercise, and achieved slower maximal exercise rates during forced exercise, than young WT mice, and achieved results similar to aging WT animals. Aging KO mice showed an even greater decline in voluntary and forced exercise performance. In addition, KO mice did not improve with practice, while WT animals demonstrated improved performance over time. These findings suggest that although Sca-1 KO did not negatively impact muscle strength ex vivo, Sca-1 expression contributes to the response to exercise, and that the subnormal response observed in Sca-1 KO animals becomes more pronounced with age.

COMPETING INTERESTS

The authors have declared that no competing interests exist.

]]>
http://currents.plos.org/md/article/stem-cell-antigen-1-in-skeletal-muscle-function/feed/ 0