INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Myosin is one of the basic structural components of skeletal muscles, including respiratory muscles. The different myosin heavy-chain (MyHC) isoforms determine the muscle fiber type (1) and modulate the functional characteristics of the muscle (2). The predominant type of myosin in a muscle determines not only the velocity of shortening of the muscle but also the mechanochemical efficiency and economy related to the different kinds of contraction of the muscle. An interesting aspect of muscles is their capacity to modify their cellular structure in response to a wide variety of stimuli (3) that include both neural (4) and mechanical (5) activity as well as hormonal factors (6).

Muscle adaptation has been explored mainly in the field of sports medicine. In response to increased activity, muscles display an increase in both the quantity of the slow MyHC isoform and the number of slow fibers (7). Reducing the muscle load has the opposite effect (8, 9). MyHC expression has previously been evaluated in the respiratory muscle of some mammals, and similar effects have been observed after muscle stimulation or unloading protocols (10). However, only slight changes in the fibrillar and molecular composition were found in the diaphragm of rats that had undergone whole- body training (13). Very few studies have examined the effect of respiratory loading on MyHC gene expression in respiratory muscles (2, 14, 15), and whether clinical conditions may change gene expression patterns in these tissues.

The purpose of the present study was to explore this topic. We examined the effect of moderate inspiratory resistive breathing (IRB) on the expression of MyHC isoforms in the respiratory muscles of dogs. The resistances and flows at the IRB level chosen for study corresponded to those observed in actual clinical conditions, such as obstructive sleep apnea or heavy snoring (16).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eleven mongrel dogs, weighing 25 ± 4 kg (mean ± SD), were included in the study. Ethical approval was obtained from the Committee for Animal Research of the Centre de Recherche du CHUM.

Experimental Protocol

MyHC messenger RNA (mRNA) expression was analyzed in the diaphragm, external intercostal, and limb (left peroneus longus) muscles of all animals prior to the IRB period. The distribution of MyHC mRNAs in the different portions of the diaphragm (costal versus crural, right versus left) was also assessed. In a subset of seven dogs, the effect of inspiratory resistive loads on MyHC mRNA expression was examined in the same muscles. Awake and spontaneously breathing animals were subjected to 2 h of IRB per day for four consecutive days.

Surgical Procedure

All animals were anesthetized with 60 mg/kg sodium pentobarbital given intravenously, and samples of the diaphragm (left costal and crural portions, accessed via a small laparotomy), external intercostal muscle (left side, 4th intercostal space, taken by local biopsy), and limb muscle (taken by local biopsy on the left side) were obtained to quantify the baseline expression of MyHC mRNA in different muscles and to compare its expression in costal and crural portions of the diaphragm. In four of the animals (controls), samples of the diaphragm were obtained in duplicate (adjacent parts of the same muscle portions) as well as from the right side. This allowed us to make an approximate analysis of inter- and intraregional distributions of MyHC mRNA. In the remaining seven animals, a tracheotomy was performed after the biopsy. On Days 3 to 7 after surgery, these seven dogs were subjected to the IRB protocol. After the 4 d of loaded breathing, the animals were again anesthetized, and samples from the contralateral diaphragm (right costal and crural) and external intercostal and limb muscles were taken. The animals were then euthanized.

Inspiratory Resistive Loads

The dogs were subjected to intermittent periods of IRB. More specifically, a two-way valve was connected to the endotracheal tube, and inspiratory resistive loads (~ 80 cm H₂O/L/s) were applied to the inspiratory port of the valve according to the schedule described in the experimental protocol. The force developed by the inspiratory muscles was measured in the form of swings in mean tracheal pressure. (). Flow was measured with a Fleisch No. 2 pneumotachograph (Fleisch, Lausanne, Switzerland) and end-tidal CO₂ from the tracheal tube was measured with a Godart capnograph (Statham Instruments, Oxnard, CA) (Figure 1). These variables were continuously monitored and recorded on an eight-channel strip-chart recorder (Model HP 7558A; Hewlett-Packard).

View larger version (29K): [in this window] [in a new window]   Figure 1.   Animal preparation used in the IRB protocol. The loads were applied at the level of the tracheal tube. Mouth pressure, flow, and ETCO2 were continuously monitored. Control samples were taken from the left side of the diaphragm, 4th left external intercostal muscle, and left peroneus longus, whereas sampling after resistive breathing was performed from the right side. Circles represent samples from crural diaphragm.

Processing of Muscle Samples

Tissue biopsies were fixed in 4% paraformaldehyde and subsequently washed in sucrose. After immersion in standard ornithyl carbamyltransferase and snap-freezing in N₂-isopentane, muscle samples were cut into 10-µm slices at 24° C in a cryostat and were mounted on poly-L-lysine-precoated slides.

In Situ Hybridization

Specific probes for the mRNA of canine fast (type II) and slow (type I) MyHC were used. These represented exons 6 to 8, including the hypervariable region of the adenosine triphosphatase (ATPase) site, which is known to differ in the two isoforms. VIDFAST and VIDSLOW complementary DNAs (cDNAs), obtained by polymerase chain reaction from canine muscles in the laboratory of one of us (G.G.), were subcloned into the EcoRI site of plasmid pCRII (Invitrogen). ³²P-labeled sense and antisense cRNA probes were produced by T3 and T7 polymerase-driven transcription of the corresponding cDNAs. Each muscle sample was processed in quadruplicate (twice with each MyHC isoform probe). Briefly, slides were sequentially washed in phosphate-buffered saline (PBS), glycine-PBS, Triton-PBS, protease K, 4% paraformaldehyde-PBS, photographic emulsion, and formamide. They were then gradually dehydrated with increasing ethanol concentrations and submitted to hybridization with the probes. In the presence of ribonuclease, the slides were sequentially washed at room temperature under increasingly stringent conditions (from 4× standard saline-citrate [SSC] to 0.9× SSC). The slides were then dehydrated once again and submitted to autoradiography for visualization of hybridization. In total, 60 fibers were evaluated for each muscle sample, and the evaluation was done in a double-blind manner. Only diaphragm samples were analyzed through in situ hybridization.

Slot-Blot Hybridization

To validate the quantification of in situ hybridization of diaphragm muscle MyHC mRNA, and to investigate mRNA expression in the external intercostal and limb muscles, the same cloned cDNA probes were used for slot-blot hybridization as for in situ hybridization. Briefly, total RNA was isolated from frozen tissue samples with TRIzol reagent (Life Technologies, Inc., Gaithersburg, MD) according to the manufacturer's instructions. The RNA samples were slot-blotted onto Nytran-Plus membranes (Schleicher & Schuell, Keene, NH), using The Convertible Filtration Manifold System Series 1055 (Life Technologies), and were fixed by UV crosslinking. The blotted samples were then sequentially probed with each cDNA probe. The latter were ³²P-labeled by random priming, using Ready-To-Go labeling beads (Pharmacia Biotech, Piscataway, NJ). The first (either VIDFAST or VIDSLOW) radioactive probe that was used was purified by Sephadex G-50 column separation, and was then denatured and allowed to hybridize overnight to the blotted RNAs at 42° C in the presence of 50% formamide, 5× Denhardt's solution, 6× SSC, 0.5% sodium dodecyl sulfate (SDS), and 100 µg/ml of denatured salmon sperm DNA. Membranes were rinsed and washed to a stringency of 0.2× SSC and 0.1% SDS at 62° C. Densitometric analysis of the resulting autoradiographs was done with the Alpha Image 2000 documentation and analysis system (Alpha Inotech Corporation). The blots were then stripped of radioactive probe through a 4-h incubation at 80° C in 1 L of 1 mM Tris-Cl (pH 8.0), 1 mM ethylenediamine tetraacetic acid (pH 8), and 0.1× Denhardt's solution. The membrane was then ready for reprobing with the second cDNA probe under the same conditions. The results obtained for each RNA sample were compared, and the relative ratios of slow- to fast-MyHC mRNAs were calculated.

Statistical Analysis

Data appear as mean ± SD. The MyHC mRNA expression level was evaluated with a current semiquantitative method (17) for in situ hybridization (with a score ranging from 0 = absent to 1 mild, 2 = moderate, 3 = strong, and 4 = very strong), and laser densitometry was used for slot-blot hybridization. The normality for the distribution of each variable was evaluated with the Kolmogorov-Smirnov test. Wilcoxon's signed-ranks test for paired data was used for comparisons of variables with nonnormal distribution. The absolute and relative percentages of change were used to describe modifications in MyHC mRNA expression after IRB. A level of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mechanics of Breathing

During the IRB protocol, the applied resistance elicited a of 35.3 ± 3 cm H₂O (2.2 ± 0.3 cm H₂O at baseline), with a PET_CO2 of 34 to 46 mm Hg. The duty cycle (TI/Ttot) increased from 0.45 ± 0.02 to 0.55 ± 0.01 (p < 0.01). Changes in and respiratory-times ratio resulted in an increase in the pressure- time index

(1)

(where Ptr_max is the maximal tracheal pressure), which increased from 0.01 ± 0.003 cm H₂O at baseline to 0.11 ± 0.004 cm H₂O (p < 0.001) during IRB.

MyHC mRNA Expression in Costal and Crural Diaphragm before IRB (In Situ Hybridization)

The VIDFAST and VIDSLOW probes hybridized to mRNA within all the fibers of every sample. MyHC mRNAs were detected principally in the periphery of diaphragm fibers, close to the nuclear membranes (Figure 2). Slow-MyHC mRNA expression (1.50 ± 0.54 in the costal portion and 1.81 ± 0.37 in the crural portion of the diaphragm) was lower (p < 0.05 for both portions) than was expression of the corresponding fast-MyHC mRNA (2.13 ± 0.35 in the costal portion and 2.13 ± 0.64 in the crural portion of the diaphragm).

View larger version (135K): [in this window] [in a new window]   Figure 2.   In situ hybridization. Autoradiograph shows hybridization of VIDSLOW RNA probe in the periphery of diaphragm fibers.

Homogeneity of MyHC mRNA Expression across the Diaphragm (In Situ Hybridization)

No significant differences were observed in expression of mRNAs for fast and slow MyHC between costal (fast-MyHC mRNA = 2.1 ± 0.4, slow-MyHC mRNA = 1.5 ± 0.5) and crural portions of the diaphragm (fast-MyHC mRNA 2.1 ± 0.6, slow-MyHC mRNA = 1.8 ± 0.4) or between the left (fast- MyHC mRNA = 1.8 ± 1.0 and slow-MyHC mRNA = 1.4 ± 0.2) and right sides (fast-MyHC mRNA = 1.8 ± 0.5 and slow- MyHC mRNA = 1.6 ± 0.2) of the diaphragm. The mRNA expression patterns were also similar in adjacent zones of the same diaphragmatic portions.

Effect of Inspiratory Resistive Breathing on MyHC mRNA Expression in Different Respiratory Muscles

Changes in the relative expression of fast- and slow-MyHC mRNA were assessed by semiquantitative evaluation of their respective in situ hybridization results, as well as by densitometric scanning of slot-blot hybridizations of mRNA specimens from IRB-loaded animals. Prior to IRB, the estimated in situ hybridization score for fast-MyHC mRNA was 1.93 ± 0.13 and that for slow-MyHC mRNA was 1.14 ± 0.09 in the costal portion of the diaphragm, and that for fast-MyHC mRNA was 2.17 ± 0.11 and for slow-MyHC mRNA was 1.75 ± 0.17 in the crural portion of the diaphragm. The expression of fast-MyHC mRNA remained unchanged in the dogs subjected to the IRB protocol. However, the expression of slow-MyHC mRNA increased significantly in both the costal (to 3.14 ± 0.18, p = 0.02) and crural (to 3.20 ± 0.20, p = 0.04) portions of the diaphragm. When expressed as a percentage of total detected MyHC mRNA, expression of the slow MyHC isoform rose from 37% to 60% (an absolute increase of 23%) in the costal portion of the diaphragm, and from 45% to 59% (an absolute increase of 14%) in the crural portion. These represent relative increases of 62% (23/37) and 31% (14/45), respectively.

To verify and validate these findings, the more quantitative densitometric analysis was done of slot-blot hybridizations performed with the same cDNA probes and with total RNA extracted from the same tissues (Figure 3). The densitometric analysis confirmed the in situ hybridization data. IRB resulted in an increase in the expression of slow-MyHC mRNA in both the costal and crural portions of the diaphragm. With this method, we found absolute increases of 11% and 6%, and relative increases of 30% and 12%, respectively, in the costal and crural portions of the diaphragm (Figure 4).

View larger version (40K): [in this window] [in a new window]   Figure 3.   Slot-blot hybridization of samples from respiratory (costal diaphragm, crural diaphragm, external intercostal muscle) and peripheral muscles.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Absolute changes in slow-MyHC mRNA expression in different muscles. Results are expressed in as percent increases compared with controls, and were evaluated by both in situ hybridization (ISH) and slot-blot hybridizations (SBH).

Limb and external intercostal muscles were processed only by slot-blot hybridization. Although limb muscles showed no changes in MyHC isoforms, slow-MyHC mRNA expression increased markedly in external intercostal muscles (showing absolute and relative increases of 13% and 27%, respectively). If the ratio of slow- to fast-MyHC mRNA after IRB is divided by the corresponding ratio in control tissue (with a value of 1 in limb muscles signifying no change), the resulting index shows the highest value in the external intercostal muscles (1.74), followed by the costal (1.59) and crural (1.29) portions of the diaphragm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that as few as 4 d of moderate IRB is sufficient to induce changes in the pattern of MyHC expression in canine external intercostal muscle as well as in the costal and crural portions of the canine diaphragm. The increased expression of the slow MyHC isoform is presumably adaptive, acting to provide greater endurance to the loaded respiratory muscles. It is interesting to note that a recent study by Levine and associates (14) demonstrated a similar upregulation of the slow MyHC isoform in the diaphragm of patients with severe COPD.

Skeletal muscles have a high capacity to adapt their phenotype according to their needs and uses (4, 18). This includes stretch as well as repetitive stimulation. However, the threshold level of tension that can trigger this remodeling process, and the stimulus-response relationship in such remodeling, remain unknown. Although changes in the expression of slow-MyHC mRNA were observed in all the respiratory muscles that we studied, it is worth noting that the external intercostal muscles appeared to exhibit the greatest amount of induction. These findings are best explained by the separate contribution of each respiratory muscle to the effort of breathing, particularly when resistive loads are increased: the crural portion of the diaphragm, whose main role is to stabilize the surrounding structures, showed the smallest response. This was followed by changes in the expression in the costal portion of the muscle, which acts as a piston generating an important part of the pressure necessary to effect ventilation. The external intercostal muscles, which play a progressively important role in ventilation during loaded breathing (19, 20), exhibited the greatest amount of isoform switching. It is well established that intercostal muscle activity increases when respiratory resistance and/or ventilatory demands increase.

Mammalian muscles consist of a mixture of different types of fibers, each made up of different myosin isoforms. The type of MyHC is the main factor determining the velocity of muscle contraction, ATPase activity, and hence the rate of muscle fatigue. Fibers consisting of fast MyHC isoforms are adapted to producing rapid, powerful movements, whereas those made up of slow isoforms produce slow, economic contractions. The different MyHC isoforms are encoded by a major gene superfamily.

Critique of the Study Methods

Slot-blot and in situ hybridizations with isomer-specific probes were chosen for our study because they appeared to be the most adequate techniques for the study purposes of detecting, quantifying, and localizing early changes in MyHC expression. These two techniques are in fact complementary. In situ hybridization provides topographic information, such as where the expression of a particular MyHC isoform takes place and whether there is any degree of coexpression of isoforms within the same fiber, whereas slot-blot hybridization is better for assessing the relative amount of this expression.

Alternative methods were considered inadequate. In this regard, histochemical studies based solely on identification of fiber types provide limited information for various reasons. First, they can only detect consolidated phenotypic changes occurring relatively long after the original stimulus causing a change. In addition, these methods incorrectly assume that muscle fibers are always homogeneous for a specific MyHC isoform. Furthermore, the histochemical classification of muscle fibers is based on the muscle fibrillar phenotype, which has been characterized in only a few species (primarily the laboratory rat). Recent work has shown that the main fast MyHC isoform expressed in the rat is MyHC-IIB, which is not expressed in canine or human muscles (1).

Immunohistochemical methods, which involve the use of specific monoclonal antibodies, improve the resolution of histochemical techniques, but are still unable to detect early phenotypic changes in MyHC expression, and are at present incapable of identifying fibers that predominantly contain MyHC-IIX. This is important for respiratory muscles, since MyHC-IIX appears to be present in high concentrations in the diaphragm of many species (21). MyHC-IIX, together with abundant oxidative enzymes (22), endows respiratory muscles with a great deal of fatigue resistance (23).

Of further importance is that although protein electrophoresis can quantify the final proportions of MyHC isoforms, it is again not appropriate for detecting initial genetic events occurring after a short period of stimulation.

Expression of MyHC Genes

In the present and in previous studies (24), mRNA for MyHC was detected predominantly around the nucleus, in the periphery of muscle fibers. Lower concentrations of MyHC mRNA have been reported close to the myosin assembly site, suggesting that the translation of MyHC takes place in the interfibrillar spaces. Although they are being actively investigated, the transcriptional and translational mechanisms that regulate the expression of MyHC are not well known. It is believed that these mechanisms are regulated by mechanical signals. As a rule, the genes for fast-type MyHC seem to be expressed by default (i.e., during muscle inactivity), whereas the genes encoding slow-type MyHC would be activated in response to changes in loads (18), leading to transformations in muscle fiber composition (25). However, the type of activity of muscle also influences its final phenotype (10). For example, although aerobic training triggers expression of slow MyHC, weightlifting induces a greater expression of the fast isoform (10).

It is believed that the changes in tension generated by a muscle during effort are the initial inducers of the transformation process. These changes could simply act mechanically or could act through the release of a messenger molecule (26), but the link between the mechanical signal and changes in MyHC expression remains unknown (27, 28). A good candidate is the splice variant of insulin-like growth factor-1, which was recently cloned from muscles subjected to stretch or that had sustained damage (29). In this context, we have previously reported data on membrane and sarcomere injury in diaphragm of dogs subjected to IRB loads similar to those applied in the present study (30). It was found that diaphragm muscle fibers showed an increased prevalence of membrane damage and sarcomere disruption after IRB. Interestingly, type II fibers appeared more susceptible to damage. This supports the idea that damage could be a prerequisite for gene switching following IRB, with both IRB and damage participating to promote muscle remodeling.

The changes observed after the IRB period in the present study illustrate just how adaptive respiratory muscles are, and their need for appropriate exercise in maintaining their properties. In this regard, increased expression of genes encoding the slow MyHC isoform would confer multiple metabolic and mechanical advantages. The lower energetic cost necessary to maintain a similar tension (31) would be added to higher resistance to fatigue. The endurance-type challenge offered by IRB in the present study is compatible with a predominant recruitment of type I (slow) fibers. Endurance training has previously been shown to reprogram the genetic expression pattern of skeletal muscle, transforming its components to make it more resistant (32). In fact, an interesting implication of our study is the possibility of eliciting a phenotypic adaptation of respiratory muscles in humans through use of a training protocol designed to achieve increased endurance. On the other hand, Levine and colleagues (14) have recently shown that smaller stimuli, such as those that sustain a patient with severe stable COPD, can also produce similar changes over a longer period of time. Further studies will focus on determining the minimal threshold (in terms of both intensity and time course) needed to trigger this form of muscle remodeling.

In summary, breathing against moderate levels of inspiratory resistance quickly induces an increase in expression of the genes encoding slow MyHC isoforms in canine respiratory muscles. This demonstrates the importance of even transient respiratory overloads in determining the phenotypic characteristics of these muscles.