Neuromuscular correlates to the evolution of flapping flight in birds.

The neuromotor pattern (i.e. the onset/offset of muscle contraction within the locomotor cycle) is conserved for some homologous muscles of the tetrapod shoulder but not others in the transition from terrestrial locomotion to flight. Here we test for three shoulder muscles of the European starling (Sturnus vulgaris) to determine whether retention of, or deviation from, a conserved neuromotor pattern can be predicted on the basis of the location of the muscle's motor nucleus within the motor column and the histochemical profile of its constituent muscle fibers. The M. supracoracoideus, the major wing elevator, illustrates a neuromotor pattern that has shifted in its timing within the limb movement cycle. Of the two heads of the triceps, the electrical activity pattern of M. humerotriceps is conserved during the transition, whereas that of the M. scapulotriceps is not. We reacted serial sections of each muscle for myosin adenosine triphosphotase (ATPase), nicotinamide adenine dinucleotide diaphorase (NADH-D), and alpha-glycerophosphate dehydrogenase (alpha-GPD) to characterize all muscles into two fiber types: fast glycolytic (FG) and fast oxidative glycolytic (FOG). We used retrograde axonal tracers to determine the longitudinal distribution and topographical organization of the motoneurons within the motor column in the spinal cord. The histochemical profile of each muscle studied is unique and is statistically different from its homologue in non-avian tetrapods. Compared to non-avian tetrapods, the spatial location of the motor nucleus of the supracoracoideus is conserved. The topology of the two heads of the triceps is fundamentally conserved relative to the other test muscles, but relative to one another there is some spatial segregation which might reflect their respective functional specializations. These data indicate that an evolutionary change in neuromotor pattern can occur without a corresponding topological reorganization of a muscle's motor nucleus within the motor column. Nor can the histochemical profile of homologous muscles be used to predict their neuromotor pattern in the transition from terrestrial locomotion to flight. These findings suggest that evolutionary change in neuromotor outflow relates to altered synaptic input from supraspinal or segmental sources or by alteration of factors intrinsic to individual motoneurons.

Neuromuscular organization of avian flight muscle: Morphology and contractile properties of motor units in the pectoralis (pars thoracicus) of pigeon (Columba livia).

We used acid digestion and glycogen depletion to determine fascicle organization, fiber morphology, and physiological and anatomical features of individual motor units of an in-series muscle, the pectoralis (pars thoracicus) of the pigeon (Columba livia). Most fascicles are attached at one end to connective tissue. Average fiber length in the four regions examined range from 42% to 66% of average fascicle length. More than 65% of fibers are blunt at one end of a fascicle and taper intrafascicularly. Fibers with blunt-blunt endings range from 13% to 31% of the population in different regions; taper-taper fibers range from 2% to 17%. Pigeon pectoralis fibers are distinguished histochemically into fast-twitch glycolytic (FG) and fast-twitch oxidative-glycolytic (FOG) populations. Three units composed of FG fibers (FG units) contract more quickly than three units composed of FOG fibers (FOG units) (range 31-37 vs 47-62 msec), produce more tetanic force (0.11-0.32 vs 0.02-0.05 N) and are more fatigable (<18% initial force vs >50% after repeated stimulation). Most motor units are confined to one of the four muscle regions. Territory of two FOG units is <30% of parent fascicle length. Territories of other units spanned parent fascicles; most fibers in these units do not extend the full fascicle length. Compared to FG units, FOG units have lower maximum innervation ratios and density indices (ratio of depleted/total FOG fibers in territory 8-14% vs 58-76% for FG units). These differences support the hypothesis that FG units are organized to produce substantial force and power for takeoff, landing and other ballistic movements whereas FOG units are suited for sustained flight when power requirements are reduced. Implications of findings for understanding the control of in-series muscles and the use of connective tissue elastic elements during wing movements are discussed. J.Morphol. 236:179-208, 1998. © 1998 Wiley-Liss, Inc.

The contractile properties of the M. supracoracoideus In the pigeon and starling: a case for long-axis rotation of the humerus

Wing upstroke in birds capable of powered flight is kinematically the most complicated phase of the wingbeat cycle. The M. supracoracoideus (SC), generally considered to be the primary elevator of the wing, is a muscle with a highly derived but stereotyped morphology in modern flying birds. The contractile portion of the SC arises from a ventral sternum, but its tendon of insertion courses above the glenohumeral joint to insert on the dorsal surface of the humerus. To clarify the role of the SC during wing upstroke, we studied its contractile and mechanical properties in European starlings (Sturnus vulgaris) and pigeons (Columba livia), two birds with contrasting flight styles. We made in situ measurements of isometric forces of humeral elevation and humeral rotation and, in addition, measured the extent of unrestrained humeral excursion during stimulation of the muscle nerve. We also generated passive and active length-force curves for the SC of each species. Stimulation of the SC at humeral joint angles of elevation/depression and protraction/retraction coincident with the downstroke-upstroke transition and mid-upstroke produced substantially higher forces of long-axis rotation than elevation. When the humerus was allowed to move (rotate/elevate) during stimulation, we observed rotation about its longitudinal axis of up to 70-80 degrees , but humeral elevations of only 40-60 degrees above the horizontal (as measured in lateral view). In the active length-force experiments, we measured mean (+/-s.d.) maximal tetanic forces of 6.5+/-1.2 N for starlings (N=4) and 39.4+/-6.2 N for pigeons (N=6), unexpectedly high forces approximately 10 times body weight. The working range of the SC in both species corresponds to the ascending limb (but not the plateau) of the active length-force curve. The potential for greatest active force is high on the ascending limb at joint angles coincident with the downstroke-upstroke transition, a time when the humerus is depressed below the horizontal and rotated forward maximally. As the SC shortens to counterrotate and elevate the humerus during early upstroke, the potential for active force at shorter lengths declines at a relatively rapid rate. These findings reveal that the primary role of the SC is to impart a high-velocity rotation of the humerus about its longitudinal axis, which rapidly elevates the distal wing. This rapid twisting of the humerus is responsible for positioning the forearm and hand so that their subsequent extension orients the outstretched wing in the parasagittal plane appropriate for the subsequent downstroke. We propose that, at the downstroke-upstroke transition, variable levels of co-contraction of the M. pectoralis and SC interact to provide a level of kinematic control at the shoulder that would not be possible were the two antagonists to work independently. The lack of a morphologically derived SC in Late Jurassic and Early Cretaceous birds precluded a high-velocity recovery stroke which undoubtedly limited powered flight in these forms. Subsequent evolution of the derived SC capable of imparting a large rotational force to the humerus about its longitudinal axis was an important step in the evolution of the wing upstroke and in the ability to supinate (circumflex) the manus in early upstroke, a movement fundamental to reducing air resistance during the recovery stroke.

The functional anatomy of the shoulder in the European starling (Sturnus vulgaris).

The excursions of wing elements and the activity of eleven shoulder muscles were studied by cineradiography and electromyography in European starlings (Sturnus vulgaris) flying in a wind tunnel at speeds of 9-20 m s-1 . At the beginning of downstroke the humerus is elevated 80-90° above horizontal, and both elbow and wrist are extended to 90° or less. During downstroke, protraction of the humerus (55°) remains constant; elbow and wrist are maximally extended (120° and 160°, respectively) as the humerus passes through a horizontal orientation. During the downstroke-upstroke transition humeral depression ceases (at about 20° below horizontal) and the humerus begins to retract. However, depression of the distal wing continues by rotation of the humerus and adduction of the carpometacarpus. Humeral retraction (to within about 30° of the body axis) is completed early in upstroke, accompanied by flexion of the elbow and carpometacarpus. Thereafter the humerus begins to protract as elevation continues. At mid-upstroke a rapid counterrotation of the humerus reorients the ventral surface of the wing to face laterad; extension of the elbow and carpometacarpus are initiated sequentially. The upstroke-downstroke transition is characterized by further extension of the elbow and carpometacarpus, and the completion of humeral protraction. Patterns of electromyographic activity primarily coincide with the transitional phases of the wingbeat cycle rather than being confined to downstroke or upstroke. Thus, the major downstroke muscles (pectoralis, coracobrachialis caudalis, sternocoracoideus, subscapularis, and humerotriceps) are activated in late upstroke to decelerate, extend, and reaccelerate the wing for the subsequent downstroke; electromyographic activity ends well before the downstroke is completed. Similarly, the upstroke muscles (supracoracoideus, deltoideus major) are activated in late downstroke to decelerate and then reaccelerate the wing into the upstroke; these muscles are deactivated by mid-upstroke. Only two muscles (scapulohumeralis caudalis, scapulotriceps) exhibit electromyographic activity exclusively during the downstroke. Starlings exhibit a functional partitioning of the two heads of the triceps (the humerotriceps acts with the pectoralis group, and does not overlap with the scapulotriceps). The biphasic pattern of the biceps brachii appears to correspond to this partitioning.

Musculotopic innervation of the primary flight muscles, the pectoralis (Pars thoracicus) and supracoracoideus, of the pigeon (Columba livia): a WGA-HRP Study.

The distribution of motoneurons innervating the primary depressor and elevator muscles of the wing of the domestic pigeon (Columba livia) was studied by using the retrograde axonal tracer lectin-conjugated horseradish peroxidase (WGA-HRP). Injection of WGA-HRP into the pectoralis (pars thoracicus) labeled neurons in the ventromedial corner of the lateral motor column of the spinal cord. These neurons were arranged in a column extending from spinal segment X or XI to spinal segment XII or XIII. The pectoralis, the primary depressor muscle of the wing, consists of two parts which are anatomically and functionally distinct, the sternobrachialis (SB) and thoracobrachialis (TB). Injection into the SB labeled neurons in the rostral and middle regions of the pectoralis motoneuron column. In contrast, injection into the TB labeled neurons in the middle and caudal regions of the pectoralis motoneuron column. Injection into the primary elevator muscle of the wing, the supracoracoideus, labeled neurons in the lateral motor column in spinal segments X and XI. These motoneurons were located dorsolateral to motoneurons labeled following pectoralis injection. These data demonstrate musculotopic segregation of the motoneurons innervating the primary flight muscles in the pigeon and, further, illustrate that the SB and TB subregions of the pectoralis are innervated by discrete aggregations of motoneurons.

Neuromuscular organization of the pectoralis (pars thoracicus) of the pigeon (Columba livia): implications for motor control.

The pectoralis (pars thoracicus) of the domestic pigeon (Columba livia) is divisible into two anatomical parts, the pars sternobrachialis (SB) and the pars thoracobrachialis (TB). Innervation to this complex is from rostral and caudal branches of the brachial ventral cord. In four anesthetized pigeons, the distribution of muscle units associated with each nerve branch was mapped after prolonged stimulation of each nerve and subsequent analysis for muscle fiber glycogen. An additional three animals were used to analyze the morphology, distribution, and histochemical profiles of the muscle fibers in the SB and TB subregions. Fibers were characterized on the basis of their reactions for myofibrillar adenosine triphosphates (alkaline and acid preincubation) and reduced nicotinamide adenine dinucleotide diaphorase (NADH-D). The SB is primarily innervated by the rostral nerve branch and the TB by the caudal nerve branch. For two-thirds of the muscle's length, the SB is separated from the TB by an aponeurosis, the membrana intermuscularis (MI). SB and TB fibers located posteroventral to the caudal margin of the MI are innervated variously by both nerves. Two populations of fibers were recognized, distinguishable primarily by 1) fiber diameter and 2) density of the NADH-D reaction product. Compared to the TB, the SB possesses a higher average percentage of large fibers. Within the SB but not the TB the percentage of large fibers increases from deep to superficial. These data support our previous findings that the pars thoracicus of the pigeon is partitioned into at least two functional subunits, each with a potential for independent action on the wing during flight.

A cineradiographic analysis of bird flight: the wishbone in starlings is a spring.

High-speed x-ray movies of European starlings flying in a wind tunnel provide detailed documentation of avian skeletal movements during flapping flight. The U-shaped furcula (or "wishbone," which represents the fused clavicles) bends laterally during downstroke and recoils during upstroke; these movements may facilitate inflation and deflation of the clavicular air sac. Sternal movements are also coupled with wingbeat, ascending and retracting on downstroke and descending and protracting on upstroke in an approximately elliptical pathway. The coupled actions of the sternum and furcula appear to be part of a respiratory cycling mechanism between the lungs and air sacs.

A functional analysis of the primary upstroke and downstroke muscles in the domestic pigeon (Columba livia) during flight.

In domestic pigeons (Columba livia), the electrical activity of the major depressor muscle of the wing, the pectoralis (pars thoracicus), beings in late upstroke well before the wing begins its downstroke excursion. The two architecturally distinct heads of the pectoralis, the sternobrachialis and the thoracobrachialis, are differentially recruited during take-off, level flight and landing. In addition to wing depression, the sternobrachialis protracts the humerus and the thoracobrachialis retracts the humerus. At the point of transition from wing upstroke to downstroke, the pectoralis EMG signal typically exhibits a reduction in amplitude. The supracoracoideus, in addition to an expected EMG associated with wing elevation, is co-activated with the pectoralis about 50% of the time.

Structure and neural control of the pectoralis in pigeons: implications for flight mechanics.

The pectoralis muscle in pigeons (Columba livia) is composed of two heads (sternobrachialis, thoracobrachialis) that are separately innervated and have different fiber orientations. High-speed film and electromyographic studies of free-flying pigeons reveal that the pectoralis is activated prior to wing depression (the power stroke) and that its two heads are differentially recruited during takeoff, level flight, and landing. The electrical activity patterns of both heads support an interpretation that intramuscular elasticity provides energy storage. The pectoralis is not only the prime wing depressor but is also capable of adjusting the excursion of the wing during different phases of flight.

Kinematics of locomotion in cats with partially deafferented spinal cords: the spared-root preparation.

The spinal cord was partially deafferented in two cats by dorsal rhizotomy sparing the dorsal root, L6. Kinematic methods were used to study movements of the pelvis and the affected hindlimb during treadmill walking. The E2 yield of the ankle 2 days postsurgery was of greater duration and extent than normal. Longer duration remained at 14 days, whereas the extent was normal by 5 days. Knee displacement was limited during E3. Lateral oscillations of the pelvis were exaggerated initially, but normal by 14 days. Pelvic shift over the contralateral limb suggested abnormal loading patterns during early recovery. These qualitative descriptions of motor recovery are consistent with reported central nervous system responses following partial deafferentation.

Motor units of the primary ankle extensor muscles of the opossum (Didelphis virginiana): functional properties and fiber types.

Motor units of the medial gastrocnemius (MG) and the single lateral gastrocnemius/soleus (LG/S) muscles of the opossum (Didelphis virginiana) were found to have uniformly slow contraction times relative to homologous muscles of the cat. Though a broad range of peak tetanic tensions was found among motor units from both muscles, most of the motor units were quite large relative to tension of the whole muscle. Comparison of the relative sizes of motor units showed that those of LG/S are significantly larger and slower than the units of MG. This suggests that the motor units of the two muscles may be differentially recruited during different behaviors. All of the MG and LG/S motor units were highly or moderately resistant to fatigue. Histochemical staining for NADH-diaphorase activity indicated consistently high levels of the enzyme in all of the fibers of both muscles. Apparently, all of the fast motor units consist of fast oxidative/glycolytic (FOG)-type muscle fibers. Our data provide functional evidence that the types of myofibrillar ATPase demonstrated by Brooke and Kaiser ('70), are not necessarily correlated to physiological classification of fiber types as slow oxidative (SO), fast oxidative/glycolytic (FOG), and fast glycolytic (FG) (Peter et al., '72). Perhaps compartmentalization of muscle fiber types may be a first step in the separation of muscles into multiple heads during the evolution of specialization to diverse locomotor habits among the mammals.

An effective sleeving technique in nerve repair.

This report describes the use of a porous polymeric sleeve (Gore-tex) to direct nerve fiber growth after axotomy. Select nerves of the triceps surae muscles in 5 adult cats were surgically isolated, sectioned, and crossed or self- reunited . A piece of Gore-tex, 15 mm in length, was compressed to 5 mm and sleeved over each distal nerve end. The appropriate proximal and distal ends were stitched together, and the Gore-tex stretched back to its original length over the suture junction. The effectiveness of the Gore-tex sleeve was assessed 4-15 months post-operatively. Electrophysiological measurements of muscle force and dorsal root volleys revealed a complete absence of unintended reinnervation and a regeneration that was more substantial for motor than sensory axons. Finally, serial histological cross-sections were prepared for each nerve above, below and at the cross union. There was no evidence of nerve tissue invading the Gore-tex wall.

The functional anatomy of the shoulder of the savannah monitor lizard (Varanus exanthematicus).

The excursions of the scapulocoracoid and forelimb and the activity of 18 shoulder muscles were studied by simultaneous cineradiography and electromyography in Savannah Monitor lizards (Varanus exanthematicus) walking on a treadmill at speeds of 0.7-1.1 km/hour. During the propulsive phase, the humerus moves anteroposteriorly 40-55° and rotates a total of 30-40°. Simultaneously, the coracoid translates posteriorly along the tongue-and-groove coracosternal joint by a distance equivalent to about 40% the length of the coracoid. Biceps brachii, coraco-brachialis brevis and longus, the middle and posterior parts of the latissimus dorsi and pectoralis, serratus anterior, serratus anterior superficialis, subscapularis, supracoracoideus, and triceps usually become active during the late swing phase and continue activity throughout most or all of propulsion. The anterior part of the latissimus dorsi is active during the transition from propulsive to swing phases. Brachialis, deltoideus scapularis, levator scapulae, the anterior part of pectoralis, scapulo-humeralis posterior, and subcoracoideus are active primarily during the swing phase; they are occasionally active during propulsion. Deltoideus clavicularis, scapulo-humeralis posterior, sternocoracoideus, and the posterior part of the trapezius are biphasic, with activity in both the propulsive and swing phases. A number of shoulder muscles in Varanus exanthematicus and Didelphis virginiana (the Virginia opossum) are similar in attachments, in activity patterns with respect to phases of the step cycle, and in apparent actions. These similarities are interpreted as a pattern inherited from the ancestors of higher tetrapods. The sliding coracosternal joint permits an increase in step length without demanding greater excursion at the shoulder and elbow joints.

Histochemical and physiological properties of cat motor units after self-and cross-reinnervation.

1. This report describes selected histochemical and physiological properties of the motor units of adult cat soleus muscle approximately one year after self- and cross-reinnervation with the nerve of the heterogenous flexor hallucis longus (f.h.l.). Self-reinnervated f.h.l. motor units are also considered. Whole muscles were tested for fibre reaction to alkaline pre-incubated ATPase, alpha-glycerophosphate dehydrogenase (alpha-GPD) and reduced nicotinamide adenine dinucleotide diaphorase (NADH-D). Motor units were isolated and studied by splitting the ventral root in acute preparations.2. The histochemical fibre type profile in the self-reinnervated muscle was comparable to normal muscle as was mean twitch contraction time, twitch-tetanus ratio and fatigue index. The mean tetanic tension of the soleus self- and cross-reinnervated motor units appeared close to a normal soleus whereas the mean tetanic tension of the f.h.l. self-reinnervated units was significantly less than a normal f.h.l.3. An average of 14% of the fibres of the soleus cross-reinnervated muscles had high ATPase and a alpha-GPD staining intensity in contrast to normal and self-reinnervated soleus in which such fibres are absent. Thus alkaline lability of myofibrillar ATPase increased in some fibres of what was originally a homogeneous population. The small increase in the number of densely staining fibres for ATPase at an alkaline pH (14%) was associated with a 73% decrease in (mean) contraction time (41 +/- 11 ms) of the thirty-three cross-reinnervated muscle units studied, with no unit's contraction time greater than 60 ms. Mean contraction times for the self-reinnervated soleus and f.h.l. muscles were 78 +/- 31 ms and 27 +/- 8 ms respectively.4. All fibres of the soleus cross-reinnervated muscles showed intense reaction to NADH-D, as was true of self-reinnervated soleus. This staining pattern is typical of normal soleus. In concordance, these motor units consistently demonstrated a high resistance to fatigue when stimulated for a four-minute period.5. These results suggest that in the adult self-and cross-reinnervated soleus muscle, there is some active mechanism which regulates the eventual size of motor units as reflected by tetanic tension.6. Change in contraction time from that typical for a soleus unit to that similar to an f.h.l. unit remains incomplete one year after cross-reinnervation. Within this time this partial change in single motor units reflects incomplete neural control of this property rather than a mixture of self- and foreign-innervation.7. A greater degree of independence from neural control to conversion of the histochemically demonstrated myofibrillar ATPase activity exists than is the case for contraction time.

Electrical activity and relative length changes of dog limb muscles as a function of speed and gait.

Electrical activity and length changes of 11 muscles of the fore- and hind- limbs of dogs walking, running, and galloping on a treadmill, were measured as a function of forward speed and gait. Our purpose was to find out whether the activity patterns of the major limb muscles were consistent with the two mechanisms proposed for storage and recovery of energy within a stride: a 'pendulum-like' mechanism during a walk, and a 'spring-like' mechanism during a run. In the stance phase of the walking dog, we found that the supraspinatus, long head of the triceps brachii, biceps brachii, vastus lateralis, and gastrocnemius underwent only minor length changes during a relatively long portion of their activity, Thus, a major part of their activity during the walk seems consistent with a role in stabilization of the joints as the dog 'pole-vaulted' over its limbs (and thereby conserved energy). In the stance phase of trotting and/or galloping dogs, we found that the supraspinatus, lateral head of the triceps, vastus lateralis, and gastrocnemius were active while being stretched prior to shortening (as would be required for elastic storage of energy), and that this type of activity increased with increasing speed. We also found muscular activity in the select limb flexors that was consistent with storage of kinetic energy at the end of the swing phase and recovery during the propulsive stroke. This activity pattern was apparent in the latissimus dorsi during a walk and trot, and in the biceps femoris during a trot and gallop. We conclude that, during locomotion, a significant fraction of the electrical activity of a number of limbs muscles occurs while they undergo little or no length change or are being stretched prior to shortening and that these types of activities occur in a manner that would enable the operation of pendulum-like and spring-like mechanisms for conserving energy within a stride. Therefore these forms of muscular activity, in addition to the more familiar activity associated with muscle shortening, should be considered to be important during locomotion.

Is resistance of a muscle to fatigue controlled by its motoneurones?

The original experiment of Buller et al. and the many subsequent confirmatory reports clearly show that the time-to-peak tension and many other speed-related parameters of slow and fast muscle fibres are dictated by the motoneurone. It has been concluded that the motoneurone exerts this control of the physiological and associated biochemical properties by the frequency at which it excites the muscle fibre. However, no studies have been reported on the fatigue properties and the associated biochemical characteristics after cross-reinnervation. Based on the 'size principle' of motoneurones, it would be reasonable to assume that a muscle fibre reinnervated by a small motoneurone would be active often and that this would be manifested biochemically as an elevated oxidative capacity. Also, it has been shown repeatedly that the mitochondrial content of a muscle fibre can be modified by daily endurance type exercise. Thus, it would seem that the motoneurone at least indirectly also controls the mitochondrial content of a muscle fibre by controlling the degree of activity. We have now tested this hypothesis using self- and cross-reinnervated muscles in cats. We found that fast- and slow-twitch muscles retained their characteristic fatigue resistance properties regardless of whether the nerve to which they had become connected had originally innervated a fatigue-resistant or relatively fatiguable muscle.