INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ALTHOUGH IT HAS LONG BEEN established that the internal intercostal muscles of the parasternal region of the rib cage (the so-called parasternal intercostals) inflate the lungs when they contract (4), it has only recently been recognized that their ability to do so is not uniform. Specifically, in supine dogs, the inspiratory mechanical advantage of the parasternal intercostal in a given interspace is greatest in the muscle bundles situated near the sternum and decreases rapidly toward the costochondral junction (5, 12). The inspiratory mechanical advantage of the medial parasternal bundles also increases from the first to the second interspace, peaks in the third interspace, and then decreases gradually to the seventh interspace (7). On the other hand, the triangularis sterni, a well-known expiratory muscle covering the inner aspect of the parasternal intercostals and the costal cartilages (8), has a large expiratory mechanical advantage in all interspaces (6).

The contribution of a particular muscle to lung volume expansion (or deflation) during breathing, however, depends not only on its mechanical advantage but also on the magnitude of its neural input. To evaluate the distribution of neural input among the canine parasternal intercostals, we have initially recorded the inspiratory electromyographic (EMG) activity in different bundles during spontaneous breathing, and inspiratory activity in each bundle was quantified relative to the activity measured during supramaximal, tetanic stimulation of the corresponding internal intercostal nerve (maximal activity). These recordings indicated that inspiratory activity in a given parasternal intercostal muscle decreases markedly from the sternum to the costochondral junction (5). Inspiratory activity in the medial parasternal bundles also decreased from the third to the seventh interspace (11), thus suggesting that the topographic distribution of activity among these muscles matches the topographic distribution of mechanical advantage. Maximal activity in the parasternal intercostal of the first interspace and in the triangularis sterni, however, could not be established, so the amplitude of neural drive to these muscles could not be assessed.

To test further the validity of this concept, we have therefore examined the distribution of metabolic activity across the canine parasternal intercostals and triangularis sterni during breathing by measuring the uptake of the glucose tracer analog [¹⁸F]fluorodeoxyglucose (FDG). This glucose analog has long been used to generate positron emission tomographic images of the heart and brain in humans and large animals and to assess regional glucose metabolism in these organs (15, 20). More recently, FDG in limb muscles has also been shown to augment when muscle activity is increased by electrical stimulation (13, 14) or by exercise (22) and to diminish when muscle activity is reduced by muscle relaxants (1). If the distribution of neural input among the respiratory muscles during breathing matched the distribution of mechanical advantage, one would therefore expect that 1) FDG uptake in the parasternal intercostals shows a mediolateral gradient, 2) FDG uptake in the medial parasternal bundles is greatest in the third interspace and decreases both in the cranial and the caudal direction, and 3) FDG uptake in the triangularis sterni is about the same in all interspaces.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The experiments were carried out on seven adult mongrel dogs (body weight 14-25 kg). After a 16-h fast, the animals were anesthetized with pentobarbital sodium (initial dose, 25 mg/kg iv), placed in the supine posture, and intubated with a cuffed endotracheal tube. A venous cannula was inserted in the forelimb to give maintenance doses of anesthetic (1-2 mg/kg iv), and the rib cage and intercostal muscles were exposed on the right side of the sternum from the first to ninth interspace by deflection of the skin and underlying muscle layers. The medial portion (pars supracostalis) of the scalene muscle, however, was kept intact; indeed, this muscle is known to be inactive during breathing in the dog (3), and it was therefore used as a control, resting muscle.

The animal was allowed to recover for 30 min after surgery, at which time it was given an intravenous infusion of glucose (1 g/kg) over a 15-min period. Twenty minutes later, a bolus of FDG (0.129-0.322 mCi/kg) prepared according to the method of Hamacher et al. (9) was injected intravenously, and the animal was connected to a heated Fleisch pneumotachograph attached to a Validyne differential pressure transducer to measure lung volume, inspiratory time (TI), and expiratory time (TE); three runs of spontaneous room-air breathing were recorded over 30 min. Intravenous blood samples were drawn for determination of plasma glucose concentration immediately before and after the infusion of glucose, before the injection of FDG, and at the end of the recording period, and the mean ± SE values thus obtained in the seven animals are shown in Fig. 1. Plasma glucose concentration was substantially increased by the infusion of glucose and then decreased progressively until the end of the study. With this procedure, we could therefore ensure that the animals had a definite release of insulin, and hence we could expect that the uptake of circulating glucose (and FDG) in the respiratory muscles would be enhanced.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Outline of the experimental protocol and changes in plasma glucose concentration during the study. Values are means ± SE from 7 animals. FDG, [18F]fluorodeoxyglucose.

After completion of the recording period, the animal was given a lethal dose of pentobarbital sodium (30-40 mg/kg iv), and muscle samples (0.1-0.3 g) were removed for determination of FDG uptake. Bundles of parasternal intercostals were obtained first from the medial, middle, and lateral portions of the muscles in each interspace from the third to the sixth; the precise location of the medial, middle, and lateral parasternal bundles has been previously defined (5). Bundles of parasternal intercostals were then obtained from the medial portion of the muscle in the first, second, seventh, and eighth interspace, and the triangularis sterni was completely exposed. Bundles of this muscle were subsequently removed in all interspaces from the second to the eighth (the muscle was never present in the first interspace). In each interspace, the bundle thus selected for investigation was that inserted into the lateral aspect of the costal cartilage of the rib above. Finally, samples were removed from the pars supracostalis of the scalene and from the right ventricular myocardium.

After removal, the muscle samples were kept on ice and weighed, and they were placed in a gamma counter (Cobra model, Packard Instruments, Downers Grove, IL). Radioactive content was expressed first as counts per minute (cpm) per gram of wet weight. To allow comparison between the different animals of the study, it was then normalized for the number of microcuries injected per gram of body weight. Individual results were therefore expressed as counts per minute per gram of wet weight per number of microcuries injected per gram of body weight. Data were finally averaged for the animal group and expressed as means ± SE. Statistical comparisons between the different muscle bundles were obtained by using a repeated-measures ANOVA and, whenever appropriate, Student-Newman-Keuls tests. The criterion for statistical significance was taken as P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Parasternal FDG uptake. The mean ± SE values of FDG activity in the medial, middle, and lateral parasternal bundles in the third through sixth interspaces are compared with the FDG activity in the scalenes in Fig. 2. Parasternal activity decreased markedly from the medial to the middle bundle (P < 0.01) in each interspace, and it decreased further from the middle to the lateral bundle (P < 0.05). Activity in the lateral bundles, however, ranged from 1.05 ± 0.14 to 1.23 ± 0.07 cpm · g¹ · µCi¹ · g¹ and was consistently greater than activity in the scalenes (0.73 ± 0.10 cpm · g¹ · µCi¹ · g¹).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   FDG activity in the medial, middle, and lateral parasternal bundles in the third through sixth interspaces. Values are means ± SE from 7 animals. Hatched areas, mean ± SE activity in the pars supracostalis of the scalene muscle (control). cpm, Counts/min.

1

1

1

1

1

1

1

1

1

View larger version (23K): [in this window] [in a new window]   Fig. 3.   FDG activity in the medial parasternal bundles and the triangularis sterni in the first through eighth interspaces. Values are means ± SE from 7 animals. There is no triangularis sterni muscle in the first interspace. Hatched area, mean ± SE activity in the pars supracostalis of the scalene muscle (control).

Triangularis sterni FDG uptake. In contrast to the parasternal intercostals, FDG activity in the triangularis sterni showed no gradient from the second to the seventh interspace and decreased only from the seventh to the eighth interspace (P < 0.05; Fig. 3). Also, triangularis sterni activity was greater than activity in the medial parasternal bundles in all interspaces (Fig. 3). However, tidal volume in the period of FDG uptake was 294 ± 24 ml, and TI and TE averaged 1.02 ± 0.08 and 2.12 ± 0.17 s, respectively. Because the triangularis sterni contracts throughout expiration and the parasternal intercostals contract exclusively during inspiration, this timing implies that if the two muscles had the same metabolic activity per unit muscle mass and per second of contraction, FDG uptake in the former would be about twofold greater than in the latter. To assess the influence of this confounding factor, we therefore corrected in each individual animal the measured values of parasternal and triangularis sterni FDG activity for TI and TE, respectively. With this correction, medial parasternal activity was ~60% greater than triangularis sterni activity (P < 0.02) in all interspaces from the second to the sixth (Fig. 4).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Time-corrected FDG activity in the medial parasternal bundles and the triangularis sterni in the first through eighth interspaces. Values are means ± SE from 7 animals. Parasternal activity was corrected for inspiratory time, and triangularis sterni activity was corrected for expiratory time.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The first important result of these studies is the confirmation that neural input to the canine parasternal intercostals decreases markedly from the sternum to the costochondral junctions (5). Indeed, in each interspace in every animal, FDG activity was greater in the medial than in the middle bundles and was lowest in the lateral bundles (Fig. 2). However, the lateral parasternal bundles in the dog are electrically silent during breathing, even when the work of breathing is augmented by CO₂-enriched gas mixtures or by inspiratory airflow resistance (5), yet FDG uptake in these bundles was slightly but consistently greater than uptake in the control scalene muscle. The reason for this difference is uncertain but could be related to a combination of two factors. First, it has been established that glucose uptake is greater in slow-twitch oxidative (type I) fibers than in fast-twitch (type II) fibers (2, 10, 18, 19), and the fiber-type composition of the canine parasternal intercostals appears to be different from that of the scalenes. Specifically, the parasternal intercostals contain ~60 percent of type I fibers, whereas the scalenes have 62% of type IIa fibers (11, 16). Second, studies by James et al. (10) have provided evidence that blood flow in skeletal muscles is an important determinant of glucose uptake independent of muscle activity. Because the medial and middle portions of the canine parasternal intercostals contract during inspiration (5), blood flow to these muscles overall must be greater than flow to the scalenes. In addition, the intercostal arteries provide blood supply not only to the parasternal intercostals but also to the triangularis sterni. The contraction of this muscle during expiration might further enhance blood flow and, with it, FDG activity in the lateral parasternal bundles.

The second important result of the present study is the demonstration that the parasternal intercostals but not the triangularis sterni show definite intersegmental differences in metabolic activity during breathing (Fig. 3). In agreement with our laboratory's initial EMG studies (11), FDG uptake by the medial parasternal intercostals was consistently greater in the third than in the sixth through eighth interspaces. Uptake in our animals was also greater in the third interspace than in the first, thus suggesting that in spontaneously breathing dogs, neural input to the parasternal intercostals peaks in the middle interspaces and decreases both caudally and cranially. On the other hand, neural input to the triangularis sterni would be uniform from the rostral to the caudal end of the muscle. In view of the fact that the inspiratory mechanical advantage of the canine parasternal intercostals is greater in the third interspace than in the first and seventh interspaces (7) and that the expiratory mechanical advantage of the triangularis sterni is relatively uniform throughout the rib cage (6), the present findings overall thus strengthen the concept that the spatial distribution of neural input within a particular set of respiratory muscles is matched with the spatial distribution of mechanical advantage (5, 11).

However, FDG uptake in the medial parasternal bundles was similar in the second through fifth interspaces, and this differs from our laboratory's previous observation that in supine dogs, the medial parasternal bundles in the third interspace have a greater inspiratory mechanical advantage and display greater inspiratory EMG activity during breathing than those in the fifth interspace (7, 11). This discrepancy is difficult to explain. As we have pointed out (see the Introduction), quantitative comparison between the inspiratory EMG activities recorded in the two muscles was made by expressing these activities as percentages of the activities recorded during supramaximal stimulation of the internal intercostal nerves (maximal activity). Because these nerves provide motor supply to both the parasternal intercostals and the triangularis sterni, the parasternal electrodes must have picked some triangularis activity during the procedure. As a result, maximal parasternal activity was overestimated, and it is possible that the overestimation was greater in the fifth interspace than in the third. The greater thickness of the parasternal intercostals relative to the triangularis sterni, however, suggests that the errors in maximal activity were rather small and insufficient to account entirely for the discrepancy between inspiratory EMG activity and metabolic activity. Furthermore, the canine parasternal intercostals in all interspaces contain ~60% type I fibers and ~40% type IIa fibers (11), which suggests that this discrepancy cannot be related to a difference in fiber-type composition.

On the other hand, although our values of FDG activity in the parasternal intercostals and triangularis sterni were primarily related to the act of breathing at the time of the study, they must have been affected by other factors. The parasternal intercostals in several animals of the study had repeated fasciculations throughout the period during which breathing was recorded; these spontaneous contractions, combined with the increased blood flow due to the expiratory contraction of the triangularis sterni, may have obscured the difference in inspiratory activity between the third and the fifth interspace. Perhaps more importantly, energy production by any particular muscle derives not only from plasma glucose but also from the glycogen stored in the muscle (19). Starvation depletes endogenous glycogen, and it is well established that after refeeding, the rate of glycogen deposition in skeletal muscles, including the diaphragm, closely depends on the extent of glycogen depletion (21). Because our animals were studied after a 16-h fast, FDG uptake in the parasternal intercostals was therefore determined by the extent of glycogen depletion at the onset of the study as well as by the metabolic activity during the study. It is possible that the distribution of parasternal inspiratory activity before the study, i.e., when the animals were unanesthetized and moving freely, was different from that during the period of recording. That is, the inspiratory mechanical advantage of the parasternal intercostals and the neural input to the muscles in standing dogs might peak in the fifth, rather than the third, interspace. Alternatively, the fifth parasternal intercostal in such animals might be more involved than the third in producing movements of the trunk. In either case, glycogen depletion in the fifth interspace would be more severe than in the third interspace, leading to an augmented FDG uptake.

Despite these drawbacks, the assessment of metabolic activity by using FDG has a major advantage over electrical measurements in that it allows for quantitative comparisons between respiratory and nonrespiratory muscles as well as between muscles active in different phases of the breathing cycle. Such comparisons in our animals indicate that metabolic activity per unit muscle mass was greater both in the medial parasternal bundles and in the triangularis sterni than in the right ventricular myocardium. Also, metabolic activity per unit contraction time was greater in the medial parasternal bundles of the first through sixth interspaces than in the triangularis sterni (Fig. 4). However, after baseline (scalene) activity was subtracted, triangularis sterni activity in all interspaces still amounted to 30% of the entire (medial plus middle plus lateral) parasternal activity. This finding further supports the idea that in the dog, contraction of the triangularis sterni during expiration is a significant component of the act of breathing (8).