Published online before print May 20, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,2
Departments of Medical Technology and Physiology, Medical College of Georgia, Augusta;
* Noll Physiological Research Center, Pennsylvania State University, University Park; and
Wyeth Research, Andover, Massachusetts
2Correspondence: School of Allied Health Sciences, AA-2028, Medical College of Georgia, 1120 15th Street, Augusta, GA 30912-0100. E-mail: jcannon{at}mcg.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: macrophage • satellite cell • dendritic cell • myotube
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Rather than eliminate blood-cell formation, the present study sought to block cell homing to damaged muscle. The first step in this process involves the selectin family of adhesion molecules. One member of this family, P-selectin, appears to be functional in skeletal muscle vasculature [8 , 9 ]. P-selectin is stored in endothelial cell granules and can be expressed rapidly on the luminal surface in response to stimuli such as thrombin or histamine. Surface P-selectin binds a cognate ligand on leukocytes, P-selectin glycoprotein ligand-1 (PSGL-1). Transient, successive P-selectin-binding events mediate the transition of leukocyte movement from free flow to rolling along the endothelium. This brings the leukocytes in contact with chemokines, which in turn activate integrins, the adhesion molecules that form the anchor points for leukocyte extravasation between endothelial cells into the tissue [10 ]. In the present study, soluble PSGL-1 (sPSGL-1) was injected intravenously (i.v.) before muscle damage to saturate P-selectin and prevent rolling and subsequent extravasation of cells into the damaged muscle tissue. P-selectin function was also inhibited in alternative experiments by injecting a monoclonal antibody specific for murine P-selectin.
Previous studies of macrophage-mediated regeneration have used invasive methods for inducing muscle damage such as crushing exposed muscle with forceps [4 ] or grafting minced muscle into surgically produced cavities in extant muscle [5 ]. An alternative, in vivo method used in the present study induced damage in murine skeletal muscle by lengthening contractions [11 ]. This mechanical stimulus causes limited, well-tolerated muscle fiber damage with a significant influx of macrophages within 24 h [12 , 13 ]. Fibers undergoing regeneration recapitulate developmental isoforms of myosin heavy chain (dMHC) with expression detectable immunohistochemically by 3 days post-injury [12 ]. The contraction-induced damage induced in the present study represents a mild, physiologically relevant model of injury that is not confounded by surgical stress or local infection.
The purpose of this study was to test the hypothesis that interfering with P-selectin-mediated adherence would reduce the influx of blood-borne cells into damaged muscle and as a result, inhibit muscle regeneration. If P-selectin were to be identified as a specific control point in the extravasation process, which is critical to muscle regeneration, it would represent a target for future therapeutic interventions directed at muscle trauma and disease.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immediately before the contraction protocol, mice were anesthetized by inhalation of methoxyflurane and then injected with a recombinant sPSGL-1 expressed in COS cells [14 ] (I316, Wyeth Research, Andover, MA) at doses of 50 or 200 µg/mouse in phosphate-buffered saline (PBS) with 1% bovine serum albumin (BSA; Sigma Chemical Co.) or PBS/BSA vehicle alone via lateral tail vein. A limited number of experiments were performed using 5 µg/mouse, as described in Results. Vehicle was used as a control for sPSGL-1, as the low-affinity form of sPSGL-1 [15 ] was not available in sufficient quantities at the time of these studies. In later experiments, anti-P-selectin (553741, BD PharMingen, San Diego, CA) or isotype control (559157, BD PharMingen) was delivered via lateral tail vein at a dose of 20 µg in 100 µl PBS.
Lengthening contraction-induced damage
Ashton-Miller et al. [11
] developed the protocol used in these studies. Percutaneous electrodes were inserted posterior of the midpoint of the left femur (near the tibial nerve). The left foot was strapped into a shoe, and the left knee was stabilized by an adjustable bar. Muscle contraction was stimulated by 400-ms pulses at 150 Hz and 4 V, one stimulation every 2 s for three 5-min periods (total=450 contractions). The shoe was attached to a servomotor (Model 300B, Cambridge Technology, Watertown, MA), which rotated the ankle joint from a neutral (90°) position to a dorsiflexed position (+36°) during the final 200-ms of each stimulation. Each 5-min stimulation period was followed by a 5-min rest period during which time the anesthetized mouse remained immobilized in the apparatus. A Macintosh Quadra 650 computer, GW Instruments analog/digital interface board, and Superscope II software (GW Instruments, Somerville, MA) were used to control the servomotor and collect data. Animals recovered from anesthesia within 1 h and survived until the time of sacrifice. Normal daily activities were not impaired following the muscle damage protocol.
Tissue processing
Groups of mice (n=4–6 mice/dose/time-point, unless otherwise indicated) were killed after 1, 3, 5, and 7 days after injury (with some additional sampling at 1, 3, 6, 12, and 15 h in some experiments). At the time of sacrifice, eccentrically loaded soleus muscles were dissected out of the hind limbs of all animals, and contralateral control solei were sampled from a subset of animals as described in Results. Muscles were rinsed with sterile saline, embedded (OCT, Miles, Naperville IL), frozen in isopentane cooled to the temperature of liquid nitrogen, and transferred to polypropylene containers for storage at –70°C.
Morphological staining and immunohistochemistry
Serial, transverse, 10 µm sections were cut from the midsections of all muscles using a cryostat (Cryocut 1800, Reichert-Jung/Leica, Nussloch, Germany) maintained at –20°C. The sections were applied to poly-L-lysine-coated glass slides (Sigma Chemical Co.). Immunohistochemical staining was conducted using rat anti-mouse CD11b (M1/70, BD PharMingen), rat anti-mouse ER-BMDM1 (Accurate Chemical and Scientific, Westbury, NY), rat anti-mouse 7/4 (Serotec, Raleigh, NC), rat anti-mouse P-selectin (RB40.34, BD PharMingen), and anti-dMHC (RNMY2/9D2, Novocastra Laboratories Ltd., Newcastle, UK). For this procedure, sections were fixed in acetone, blocked with 2% BSA in PBS (pH 7.5), and then rinsed in PBS. The primary antibodies were diluted to optimum concentrations (1:20–1:500) in PBS. Tissue sections were incubated with primary antibody in a humidified chamber for 3 h, rinsed in PBS, and incubated with the secondary antibody, biotinylated goat anti-rat immunoglobulin G (IgG; Vector Laboratories, Burlingame, CA), diluted 1:200, for 1 h in a humidified chamber. Endogenous peroxidase activity was quenched by immersing the slides in 0.3% hydrogen peroxide in methanol for 20 min. Slides were rinsed, then incubated for 30 min with Vectastain ABC peroxidase reagent (Vector Laboratories), rinsed, and then incubated with 3,3'-diaminobenzidine (Sigma Chemical Co.) for 8 min. The sections were dehydrated in 95% and 100% ethanol, cleared with Hemo De (Fisher Scientific, Pittsburgh, PA), and mounted under glass coverslips with nonaqueous medium (Cytoseal, VWR, West Chester, PA). Sections incubated in PBS in the absence of primary antibody served as controls. Some sections were fixed in methanol and then stained with hematoxylin and eosin (H&E) for morphological analysis.
Myeloperoxidase (MPO) activity
Frozen sections were thawed, dried, and then immersed in acetone for 10 min, followed by 8 min in PBS. The sections were then blotted, and a drop of True Blue peroxidase substrate solution (Kirkegaard and Perry, Gaithersburg, MD) was applied to each section and allowed to react for 5 min. The sections were rinsed in distilled water, then dehydrated, cleared, and mounted as described above.
Myoblast cultures
C2C12 myoblasts (American Type Culture Collection, Manassas, VA) were cultured in 10 cm polystyrene dishes (Corning, NY) in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.2 mM L-glutamine (all from Sigma Chemical Co.) until achieving 100% confluence. Differentiation was induced by substituting 5% horse serum for the FBS. At this time, recombinant sPSGL-1 (Wyeth Research) was added to achieve final concentrations of 0.1, 1.0, and 10 µg/ml. An equivalent volume of media alone was added to control cultures. The cells were incubated at 37°C in an atmosphere of 5% CO2 and assessed for differentiation into myotubes with an inverted microscope periodically (every 6–12 h) for 72 h.
Data analysis
All slides were coded and then analyzed without knowledge of sample identity. Three outcome variables were assessed in this study. Invaded fibers: Myofibers containing CD11b+ cells well within their borders were classified as "invaded". The total number of invaded fibers in the entire cross-section was counted and reported as a percentage of the total fiber number. Cellular densities in the perimysium and endomysium [i.e., extracellular matrix (ECM)]: The densities of CD11b+ and ER-BMDM1+ cells were determined by counting cells in a 0.0625-mm2 field (encompassing 
36 fibers) using an ocular grid. As the muscle damage was focal in nature, the areas of greatest density in the cross-section were selected. Three to five fields were analyzed, and the highest count was reported, expressed per mm2. Cells with segmented nuclei, 7/4+ cells, and MPO activity were counted over the entire cross-section and were expressed per mm2. Muscle cross-sectional areas were determined on images captured with a Sony CCD camera linked to a Macintosh G3 computer using National Institutes of Health Image software. The number of myofibers in the entire section that expressed dMHC was classified as "regenerating myofibers" and was reported as a percentage of the total fiber number: We have published photomicrographs of these features in damaged murine skeletal muscle in previous reports [12
, 13
]. Data throughout are presented as mean ± SE. dMHC data were normalized by log transformation. ANOVA was performed using SuperANOVA software (Abacus Concepts, Berkeley, CA). Post-hoc analyses of differences between groups were conducted using Fisher’s protected least significant difference test.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the damaged muscle, the number of fibers invaded by CD11b+ mononuclear cells was maximal 1 day after injury (8±3 fibers/section or 0.66±0.28%; Fig. 1
). Myofibrillar invasion diminished over time and resolved by 7 days post-injury. sPSGL-1 had no significant influence on the number of fibers invaded with CD11b+ cells (P>0.8). Accumulation of CD11b+ cells in the ECM reached maximal, focal densities of 200 cells/mm2 by 24 h post-injury. These cells remained evident throughout the 7 days of observation, gradually declining in number over time. sPSGL-1 treatment had no significant influence on extracellular density of CD11b+ cells (P>0.4; Fig. 2
).
 View larger version (21K): [in a new window] |
Figure 1. Percentage of muscle fibers invaded by CD11b+ mononuclear cells following contraction-induced damage. No significant differences were observed among vehicle-injected controls (solid bars) or mice injected i.v. with 50 µg (open bars) or 200 µg sPSGL-1 (hatched bars). Sample size = four to six mice per group, except 200 µg at 5 days; n = 2.
|
 View larger version (30K): [in a new window] |
Figure 2. Density of CD11b+ cells in the ECM of soleus muscle following contraction-induced damage. Bar designations and sample sizes are the same as in Figure 1
. Two outliers (>5 standard deviations above the group means)—one in the 50 µg, 5-day group and the other in the 200 µg, 7-day group—were not included in the data plotted in Figure 3
so that the bars present an unskewed visualization of trends. No significant differences were observed between vehicle and sPSGL-1-treated animals regardless of whether the outliers were included or excluded from statistical analysis.
|
 View larger version (19K): [in a new window] |
Figure 3. Number of cells expressing MPO, antigen 7/4, or exhibiting segmented nuclei [polymorphonuclear neutrophil (PMN)] in soleus muscle within the first 24 h following contraction-induced injury. Only MPO showed a statistically significant increase over time (P=0.0004). Sample size = four animals per time-point except at 15 h; n = 2.
|
 View larger version (17K): [in a new window] |
Figure 4. Number of myofibers expressing dMHC. One outlier was omitted from the 50 µg, day-5 group, as the value was >10 standard deviations above the group mean. A dose-dependent inhibition of dMHC expression by sPSGL-1 was observed across all time-points, regardless of whether the outlier was in included (P=0.021) or excluded (P=0.003) from statistical analysis. Bar designations and sample sizes are the same as Figure 1
.
|
 View larger version (13K): [in a new window] |
Figure 5. Dose-response for sPSGL-1 inhibition of dMHC expression by myofibers at 7 days post-injury. Sample size = six for all groups except vehicle control (0 µg sPSGL-1); n = 5. *, P < 0.03, compared with vehicle control.
|
To verify that the observed changes in dMHC were a result of inhibition of P-selectin, additional mice were injected i.v. with 20 µg anti-P-selectin or isotype-control IgG. The plantarflexor muscles were subjected to lengthening contractions and then were removed at 7 days post-injury along with uninjured, contralateral control muscles. Mononuclear cells were characterized in the muscles using ER-BMDM1, which recognizes mature macrophages and dendritic cells (DC). As shown in Figure 6 , the maximal focal densities of ER-BMDM1+ cells in the injured muscles from anti-P-selectin-treated mice were approximately half that in isotype-treated controls (197±41 per mm2 vs. 377±83 per mm2, P=0.019). The undamaged, contralateral control muscles had significantly lower ER-BMDM1+ cell densities than the injured muscles (135±13 per mm2, P=0.008). Developmental MHC was expressed in 4.14 ± 1.53% of the fibers of injured soleus muscle from isotype-injected mice, but only 0.68 ± 0.27% of the fibers in muscle from anti-P-selectin-injected mice (P=0.006). Maximal focal densities of CD11b+ cells were not significantly different in damaged muscles from anti-P-selectin versus isotype-treated mice (150±44 vs. 143±19 cells/mm2, P=0.83) but were significantly higher than uninjured, contralateral control muscles (29±5 cells/mm2, P<0.0001).
 View larger version (12K): [in a new window] |
Figure 6. Maximal focal density of cells expressing ER-BMDM1 (left panel) and percent of fibers expressing dMHC (right panel) at 7 days after contraction-induced injury. The contralateral soleus muscles were not subjected to contractions. Shaded bars, Isotype-treated, control mice; n = 7. Solid bars, Mice treated with 20 µg anti-P-selectin; n = 6. *, P<0.02, compared with isotype control.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4% of the fibers expressed dMHC, indicating active myofiber regeneration. Injection (i.v.) of sPSGL-1 immediately before the muscle-contraction protocol inhibited subsequent muscle regeneration in a dose-dependent manner. Anti-P-selectin also significantly inhibited dMHC expression and reduced the accumulation of ER-BMDM1+ cells. These results are consistent with the concept that mononuclear cells are critical for muscle regeneration, and these cells require the P-selectin adhesion system for homing to or activation in the damaged tissue. Macrophages secrete factors that are chemotactic and mitogenic for muscle precursor cells, as described in the Introduction. Furthermore, macrophages are important sources of the enzymes responsible for protein catabolism in skeletal muscle following blunt trauma [18
] and are sources of fibronectin, which contributes to ECM repair [19
]. Macrophages typically express CD11b, but the maximal density of CD11b+ mononuclear cells observed infiltrating the muscle was not significantly affected by P-selectin inhibition. In contrast, anti-P-selectin treatment reduced the maximal density of cells reacting with ER-BMDM1, an antibody that recognizes a 160-kDa aminopeptidase associated with the plasma membranes of mature macrophages and DC [20 ]. In previous studies, we found that cells expressing this antigen were significantly increased in muscle 5–7 days after injury [13 ]. The reduced focal densities of ER-BMDM1+ cells but not CD11b+ cells in the present study suggest that P-selectin may selectively recruit a population of monocytes already committed to a particular maturation pathway, or alternatively, P-selectin directly influences monocyte maturation and function. In vitro experiments have shown that P-selectin binding stimulates monocyte expression of tissue factor [17 ], interleukin-1 receptor type II, and insulin-like growth factor-I [21 ]. It also modulates macrophage phagocytosis [22 ] and expression of surface antigens associated with DC [23 ].
Pimorady-Esfahani and co-workers [24 ] have identified a population of putative DC expressing MHC class II, which accumulates in crush-injured tibialis anterior muscle over a time-course independent of CD11b+ accumulation. Some subclasses of DC do not express CD11b [25 ] but do express PSGL-1 [26 ] and therefore may be influenced by P-selectin. Matrix degradation products [27 ] and necrotic cells [28 ] can activate DC to produce cytokines that modulate the response to injury.
It is possible that P-selectin could influence other types of cells implicated in muscle regeneration. Ferrari and co-workers [29 ] reported that marrow-derived cells can be recruited from the bloodstream and become incorporated into regenerating tibialis anterior muscle after cardiotoxin-induced damage. Primitive marrow-derived precursor cells express PSGL-1 [30 ] and thus may represent another class of cells whose migration into damaged muscle may have been affected by P-selectin inhibition.
The role of P-selectin has been investigated in other models of tissue damage and wound repair. By pretreating rats with the same sPSGL-1 formulation used in the present study, Dulkanchainun et al. [31 ] reduced neutrophil infiltration and hepatic injury significantly following ischemia/reperfusion. Similarly, Wang et al. [32 ] demonstrated that pretreating dogs with a PSGL-1–Ig chimera before ischemia/reperfusion reduced neutrophil influx as well as myocardial injury. A substantial portion of the damage following ischemia/reperfusion is thought to be mediated by neutrophils; hence, these two studies confirm that blocking neutrophil influx reduces tissue injury. In contrast, the damage resulting from the present lengthening-contraction model had no neutrophil involvement, as shown in Figure 3 and reported previously by Lapointe et al. [33 ].
In another study, PSGL–Ig pretreatment of rats had no influence on histology or tensile strength development during the healing process following surgical incisions of the skin [34 ] (in contrast to the reduced muscle regeneration observed in the present paper). Different tissues and types of wounds may require different healing mechanisms. For example, E-selectin is an important adhesion molecule in the skin [35 ] but is not constitutively expressed in skeletal muscle nor is it induced in skeletal muscle by local application of antigen [9 ] or cytokines [36 ].
The observed CD11b+ cellular infiltration into skeletal muscle in the present study suggests that P-selectin-independent recruitment mechanisms were functioning. These results are consistent with the observation that macrophage recruitment into damaged skeletal muscle is maintained in mice deficient in P- and E-selectins [37 ]. L-selectin-mediated adhesion has been demonstrated in muscle [38 , 39 ]. Furthermore, mononuclear cell recruitment can be mediated by integrins in the absence of selectins [40 ].
In conclusion, inhibition of P-selectin caused significant reductions in myofiber regeneration following contraction-induced muscle damage. Although differences in CD11b+ cellular infiltration into damaged muscle were not detected, significant reductions in ER-BMDM1+ cells were observed. The results are consistent with the concept that P-selectin is involved in the extravasation, development, and/or activation of cells that are critical for muscle fiber repair. Identification of interventions that enhance P-selectin function and macrophage/DC development may lead to new treatments for acute muscle injury and degenerative muscle diseases.
Received November 5, 2002; revised April 8, 2004; accepted April 9, 2004.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4 integrin J. Immunol. 163,3441-3448This article has been cited by other articles:
 |
|
 |
|
  L. Pelosi, C. Giacinti, C. Nardis, G. Borsellino, E. Rizzuto, C. Nicoletti, F. Wannenes, L. Battistini, N. Rosenthal, M. Molinaro, et al. Local expression of IGF-1 accelerates muscle regeneration by rapidly modulating inflammatory cytokines and chemokines FASEB J, May 1, 2007; 21(7): 1393 - 1402. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. K. Shireman, V. Contreras-Shannon, O. Ochoa, B. P. Karia, J. E. Michalek, and L. M. McManus MCP-1 deficiency causes altered inflammation with impaired skeletal muscle regeneration J. Leukoc. Biol., March 1, 2007; 81(3): 775 - 785. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
