Structure and function of muscle fibers and motor units - 캠브리지 대학 교과서

[문형철]

muscle의 흥분-수축, 이완의 최소단의 motor unit

위대한 근육생리학자 셰링턴이 제시한 개념

꼭 기억해야할 개념

1. 하나의 근육에 motor unit는 5~1500개. 평균 150개

2. 하나의 motor unit에는 1~160개의 운동종판이 있음. 

3. motor unit size는 차이가 있고, 작은 motor unit부터 흥분, 수축에 동원

4. 통증, 손상, 습관에 따라 근육 동원의 패턴이 깨짐

Structure and function of muscle fibers and motor units.pdf

Introduction

The term “motor unit” was introduced by Sir Charles Sherrington, a founder of modern neurophysiology, who observed that force occurred in discrete steps when a muscle contracted in the stretch reflex [1]. He postulated that each step was produced by the all-or-none action of a single motor neuron upon the muscle fibers it innervated. Sherrington’s concept of the motor unit assumed that each muscle fiber receives innervation from only one motor neuron, and that the muscle fiber faithfully responds to every impulse of the motor neuron. These assumptions have subsequently been shown to be true in healthy adult skeletal muscles. 

The motor unit has become a fundamental concept in understanding the physiology of muscle and the control of movement. A motor unit consists of one motor neuron and all the muscle fibers it innervates. The term muscle unit has been

introduced to refer to the group of muscle fibers innervated

by a given motor neuron [2]. The motor neuron and its

muscle unit are inseparable in function because each action

potential in the neuron activates all fibers of the muscle unit.

Thus motor units are the indivisible quantal elements in all

movements. The electrophysiological, metabolic, mechanical,

and anatomical properties of the motor neuron and its

muscle unit are coordinated in a manner that allows efficient

muscle contraction over a wide range of motor behaviors.

The coordinated expression‎ of the proteins that govern these

properties reflects the interplay between the trophic control

that motor neurons exert over their muscle fibers through

activity patterns and chemical trophic factors, as well as

trophic feedback from the muscle fiber to the motor neuron.

Although most of the properties of a given motor unit

become specified during the early postnatal period of development,

physical activity and disease processes can modify

certain properties to a limited extent. In this chapter, the

basic structural and physiological properties of motor units

and muscle fibers will be introduced, with a particular

emphasis on humans and other mammals.

Anatomy of motor units

Motor neurons

Motor neurons are the only central neurons with axons that

leave the central nervous system (CNS) to innervate nonneuronal

tissue. Their cell bodies are located in the anterior

horn of the gray matter of the spinal cord (Figure 1.1). The

motor neurons that innervate the same muscle cluster together

in motor nuclei that form elongated columns that generally

extend over several spinal cord segments [3]. The number of

motor neurons innervating each muscles varies, ranging from

the estimates of 30–40 motor neurons innervating the delicate

tenuissimus muscle in the cat [4] to estimates of 100–200

motor neurons innervating human thenar muscles [5, 6]. In

the lumbar and cervical enlargements of the spinal cord, the

motor neurons that innervate distal limb muscles are located

most laterally within the anterior horn, and motor neurons

innervating proximal muscles lie more medially [7, 8]. The

axons of motor neurons exit the spinal cord through the

adjacent anterior roots. When motor axons innervating the

same muscle exit from roots of several segments, they rejoin in

a muscle nerve after traversing peripheral plexuses and nerve

trunks. The muscle nerve contains motor axons innervating

the muscle and the sensory axons arising from receptors

within the muscle, such as the muscle spindles and tendon

organs.

In mammals, there are three kinds of motor neurons in the

motor nucleus. Alpha motor neurons are large cells [9, 10] that

innervate the striated muscle fibers that make up the bulk of

skeletal muscle tissue (extrafusal fibers). Gamma, or fusimotor,

neurons are considerably smaller [11] and exclusively innervate

one or more of the three types of specialized muscle fibers

within the muscle spindle – stretch receptor organs that are

present in virtually all somatic muscles [12, 13]. A third class of

motor neuron, called skeleto-fusimotor or beta motor

neurons, innervates both intra- and extrafusal muscle fibers

[14]. Beta motor neurons have been found in higher primates

[15] and probably also occur in humans. Because beta motor neurons are difficult to identify in physiological experiments,

there is little direct evidence about their properties. What

little is known indicates that the properties of beta motor

neurons and their extrafusal muscle fibers are essentially the

same as those of alpha motor neurons [16]. For this reason,

alpha and beta motor neurons will not be distinguished in this

chapter.

Alpha motor neurons have extensive dendritic trees that

receive synaptic input over their entire extent [17, 18, 19].

Their myelinated axons have large diameters with correspondingly

fast conducting velocities, ranging from 40 to 60 m/s

in human motor nerves [20]. Faster conduction velocities,

50–120 m/s, have been reported in cats and smaller mammals

[21]. The axons of motor neurons can be extremely long, up to

a meter in length for those motor neurons innervating the

distal foot muscles of a tall adult. The length and diameter

of the motor axons mean that the volume of axoplasm may

exceed the volume in the cell body and dendrites by tenfold or

more (Figure 1.2). The large metabolic demands of maintaining

the peripheral axon presumably account for the large size

of the motor neuron cell body. 

Neuromuscular junctions

As the myelinated motor axons near their target muscle, they

begin to divide into tens or hundreds of terminal branches,

which lose their myelin sheaths as they near the neuromuscular

junctions (NMJs). The NMJ is a large, highly specialized

synapse between the motor nerve terminal and the muscle

fiber [22]. In somatic muscles there is only one NMJ per

muscle fiber [23], but exceptions are found in some cranial

muscles, such as the laryngeal [24] and extraocular muscles

[25]. On a given muscle fiber, the NMJ is located approximately

equidistant from its ends, allowing action potential

depolarization to spread equally to both ends from the center

of the muscle fiber. The NMJ is a complex structure that

undergoes remodeling during development and aging and in

response to denervation. At the NMJ, the motor nerve terminal

is separated from the postsynaptic muscle membrane by

a synaptic space containing basal lamina with synapse-specific

glycoproteins. On the postsynaptic side, the muscle membrane

is highly folded. Acetylcholine receptors are found on the

crests of the junctional folds apposing the vesicle release sites

on the presynaptic terminal, whereas the voltage-gated sodium

channels responsible for action potential generation are densest

in the depths of the folds [26]. NMJs exhibit structural

specializations related to the size and type of muscle fiber [27].

The structure and function of NMJs will be covered more fully

in Chapter 23.

Muscle fibers

The skeletal muscle fiber is a cylindrical, multinucleated cell

that is formed by the fusion of myoblast cells during development.

The muscle fiber has a highly organized structure, with

several distinct spatial domains. Nuclei are positioned along

the periphery of the fiber beneath the plasma membrane, or

sarcolemma. The center of the muscle fiber is packed with the

contractile apparatus, which consists of longitudinally oriented

myofibrils and scaffolding proteins. The contractile apparatus

is encircled by a network of sarcoplasmic reticulum (SR), a

form of endoplasmic reticulum specialized for calcium release

and reuptake. The sarcolemma has numerous narrow infoldings,

called T-tubules, that penetrate deep into the muscle

fiber, where they become closely apposed to regions of the

SR at specialized junctions called triads or “calcium release

junctions.” The T-tubule membrane is continuous with the

sarcolemma membrane, but it is specifically enriched in certain

membrane proteins, such as voltage-gated calcium channels,

chloride channels, and transporters (Figure 1.3) [28, 29].

The T-tubule “interior” is in continuity with the extracellular

space, although diffusion occurs more slowly from this narrow

space than at the surface membrane. The triads, where

T-tubules meet the SR, are the sites where action potential

depolarization is coupled to the mechanical contraction.

Excitation–contraction coupling occurs through protein–

protein interactions between the sarcoplasmic domains of the

voltage-gated calcium channels on T-tubule membranes and

the calcium release proteins, known as ryanodine receptors, on

the SR membrane [30].

The contractile apparatus of the muscle is organized into a

series of repeated units a few microns long called sarcomeres

[31]. The sarcomere is the smallest unit of contraction. It

consists of highly organized protein assemblies that give the

muscle fiber a characteristic striated appearance (Figure 1.4b).

The sarcomere contains the myofibrils, longitudinal arrays of

thick and thin filaments that are maintained in a hexagonal

lattice by a scaffolding network (Figure 1.4a). Proteins in the

scaffolding network condense at the ends and middle of

the sarcomere to form transverse bands called Z-disks and

M-bands [32]. The thin filaments consist of filamentous actin

entwined by tropomyosin and troponin, a calcium-binding

protein. Thick filaments consist of myosin, a large molecule

with heavy and light chains. The myosin heavy chains have a

tail region and a globular head. Thick filaments are formed by

the assembly of myosin monomers with their tails centrally

and heads protruding outwards, with an antiparallel orientation

on opposite ends of the filament. Z-disks, which mark the

border between sarcomeres, serve to anchor the thin filaments.

The Z-disks are formed by an ensemble of several proteins,

including alpha-actinin. Titin, a large elastic protein spanning

from the Z-disk to the M-band, binds to the myofibrils, keeping

them centered in the sarcomere, and transmitting tension

to the Z-disk during sarcomere shortening [33]. Titin and

proteins that comprise the M-band essentially form an intrasarcomeric

cytoskeleton that maintains the regular spacing of

the thick and thin filaments [32, 34].

The myosin heads on the thick filament contain an ATPase

activity and binding sites for actin. When contraction is initiated

by a muscle fiber action potential, calcium released from

the SR binds troponin, uncovering binding sites on actin. This

leads to the formation of cross-bridges between actin and

myosin. The ATPase activity of myosin is enhanced by formation

of cross-bridges, and as ATP is hydrolyzed the crossbridge

is broken, freeing the myosin head to swivel to the next

actin-binding site. The repeated formation and cleavage of

actomyosin cross-bridges produces the sliding action of thin

and thick filaments that causes shortening of the sarcomere

and muscle contraction [35, 36]. The actomyosin cross-bridges

serve as the mechanical linkage between thick and thin filaments

for transmitting tension to the insertions of the muscle

fiber. The amount of tension is proportional to the number of

cross-bridges, reaching a maximum at sarcomere lengths when

thick and thin filaments have the greatest overlap [37, 38].

The muscle fiber has a rich cytoskeletal network underlying

the membrane and surrounding the myofibrils. In subsarcolemmal

regions, protein complexes of dystrophin, syntrophins,

and other molecules bind to F-actin and other cytoskeletal

proteins. By binding as well to intracellular domains of

membrane proteins such as sarcoglycans these effect a linkage

between the muscle interior and the extracellular matrix.

Beneath the subsarcolemmal cytoskeleton, networks of intermediate filaments, of which desmin is the most prominent,

play a role in the positioning and morphology of organelles

within the muscle (reviewed in [39, 40]). Desmin connects

Z-disks, SR, myofibrils, and other organelles to the subsarcolemmal

cytoskeleton. Mitochondria are usually found in two locations

within the muscle fiber, beneath the sarcolemma and

among the myofibrils, mostly near the Z-disks. Subsarcolemmal

and interfibrillarmitochondria appear to be functionally distinct,

with differing cytochrome content, capacity for ADP-stimulated

respiration, and susceptibility to apoptotic stimuli [41, 42].

Deficiencies of desmin lead to subsarcolemmal accumulation of

mitochondria in mice, supporting a key role for desmin in

mitochondrial positioning [43]. Intermediate filaments also bind

to proteins on the surface of lysosomes, which are relatively

sparse in normal muscle, but become prominent in some myopathies.

Glycogen particles, sometimes termed glycosomes, are

found in myofibrillar and subsarcolemmal locations.

Extracellular matrix

The muscle fiber is surrounded by an extracellular matrix

which consists of several distinct layers [44]. The innermost

layer, the basal lamina, contains the carbohydrate-rich extracellular

domains of membrane proteins, such as dystroglycan

and integrins, that interact with the muscle cytoskeleton;

secreted glycoproteins such as members of the laminin family; and a variety of ligands and proteoglycans that bind to the

extracellular matrix proteins. The outermost layer is rich in

collagen fibers, forming a connective tissue layer, the endomysium.

The extracellular matrix is specialized at the NMJ, containing

synaptic laminins, ligands such as agrin, and the

enzyme acetylcholinesterase. The basal lamina and the extracellular

matrix molecules play a key role in supporting muscle

fiber development and regeneration after injury. Lying beneath

the basal lamina are satellite cells, myogenic precursors that are

able to proliferate and differentiate into myoblasts [45].

Muscles

Most mammalian muscle fibers are only a few centimeters

long, much shorter than the length of most muscles. The

length of a muscle fiber is thought to be limited by the need

for sarcomeres to be activated nearly simultaneously to produce

an effective contraction, which in turn is limited by the

time needed for an action potential to travel the length of

the muscle fiber. The conduction velocity of muscle fibers is

relatively slow, in the range of 2–10 m/s [46, 47]. To achieve an

effective mechanical action over a larger length, groups of

muscle fibers, called fascicles, are bound together by perimysial

connective tissue to form a muscle. Muscle fascicles are

arranged in various ways that allow a common direction of

force to be delivered to the muscle’s points of origin and

insertion [48]. There are two general schemes [49]: pinnate,

in which the muscle fibers are oriented at an angle to the

muscle’s primary direction of force; and parallel, in which

the orientation of muscle fibers is the same as the force vector.

In pinnate muscles, the fascicles are arranged in parallel

bundles, often in a feather-like pattern along one or more

tendinous aponeuroses (Figure 1.5a). Muscles with pinnate

architecture have relatively limited distensibility, but can

deliver large output forces. Pinnation is commonly seen in

muscles with relatively short lever arms that operate over a

limited range of physiological lengths, for example the gastrocnemius

muscles of the leg. At the other extreme are muscles

with parallel arrangements of interdigitated muscle fascicles,

staggered at different longitudinal locations along a web-like

intramuscular stroma (Figure 1.5b; [47, 50]). This arrangement

allows a small amount of slippage of fascicles past each

other, and is commonly seen in muscles that span multiple

joints or undergo large changes in length during movement.

As might be expected, some muscles exhibit mixtures of these

designs (e.g., tibialis anterior in the cat; [51]). A few long,

strap-like muscles, such as the biceps femoris, have two or

more bellies arranged in series separated by tendinous inscriptions

that create distinct anatomical compartments [52].

Most muscles have an optimal range of working lengths.

When muscles are stretched during natural movements, they

offer some resistance. Most of the tension is related to the

number of cross-bridges between overlapping thick and thin

filaments [37, 38]. Additional contributions from tendons and

internal connective tissue enter into consideration primarily

when a muscle is stretched beyond its optimal working range.

Because connective tissue is less elastic than muscle fibers,

tension rises quickly at these lengths. Contributions from

connective tissue to muscle length–tension curves are referred

to as passive, in contrast to the active contributions from

the myofibrillar cross-bridges. Passive contributions to muscle

tension differ between healthy and diseased muscle. 

Degenerative muscle diseases, or even the prolonged disuse of

muscles, such as after a stroke, may result in markedly

increased connective tissue within the muscle with stiffness

and increased resistance to stretch [53].

Functional organization of motor units

Distribution of motor unit fibers

The spatial distribution of muscle fibers belonging to an individual

motor unit has been studied experimentally with the

glycogen depletion technique [54]. In this method, prolonged

stimulation of a motor axon is used to deplete muscle fibers of

endogenous glycogen stores, enabling the depleted fibers to be

identified histochemically. The glycogen depletion method

showed that muscle fibers belonging to the same motor unit

were arranged in a mosaic fashion among muscle fibers

belonging to other motor units [54, 55]. Relatively few muscle

fibers from the same unit occurred immediately adjacent to

one another [56, 57]. Statistical studies suggest that the distribution

of fibers in single units is basically random [58]. Nevertheless,

the arrangement of the muscle unit’s fibers must

accommodate to the internal architecture of the parent muscle

to produce a meaningful pattern of force. In pinnate muscles,

fibers from one motor unit were found to be scattered more or

less evenly through territories that were relatively large, but

smaller than the total cross-section of the muscle (Figure 1.6).

In multicompartment muscles, motor unit fibers were usually

distributed only within one compartment [59]. However, there

are examples, such as the extensor digitorum muscle of the monkey forelimb, in which fibers of one motor unit are

distributed among several compartments to exert a common

force on multiple tendons [60].

Electromyographic (EMG) studies of single motor units in

humans suggest a similar spatial organization of muscle unit

fibers. Using a technique called scanning EMG, in which a

motor unit action potential is recorded as an electrode is

advanced in successive steps of 50 μm through the muscle,

Stalberg and colleagues [61, 62] recorded territories with

cross-sectional areas of 2–10mm for single motor units in

the biceps and tibialis anterior muscles. Within the same

region of muscle, they found that several dozen motor units

had overlapping territories. For an individual motor unit, at

some places the muscle fiber action potentials were grouped,

and separated from other regions, suggestive of fractions of the

muscle unit innervated by different branches of the motor

axon (arrows, Figure 1.7).

One way to describe the size of a motor unit is according to

its innervation ratio: the number of muscle fibers innervated

by a given motor neuron. The number of muscle fibers innervated by one motor neuron varies widely between different

muscles. In humans, innervation ratios have been estimated

by dividing an estimate of the total number of muscle

fibers in a muscle by counts of the number of large axons in

cross-sections of the muscle nerve. Such calculations have

produced estimates of innervation ratios ranging from less

than a dozen for the extraocular muscles to over a thousand

for motor units of large limb muscles (Table 1.1) [63, 64].

Physiological methods have also been used to estimate the

number of motor units innervating certain muscles, and these

studies have also shown similar ranges [6]. However, using the

glycogen depletion method to identify the fibers of individual

motor units in animals, Burke and Tsairis [56] found considerable

variation in the innervation ratios for different units

within a given muscle. The innervation ratio of the motor unit

is a major factor governing its force output. Variation in

innervation ratios is likely to provide much of the variability

in force output produced by different motor units within a

muscle [65, 66].

Muscle fiber types

For more than a century, it has been recognized that mammalian

muscles fall into two general groups: dark “red” muscles

with slow contraction times and lighter “white” muscles with

fast contraction times. Histological and physiological studies

have shown that most muscles contain a mixture of muscle

fibers with differing contraction speeds and force outputs;

muscles composed of purely fast or slow muscle fibers are

exceptional (for reviews see [67, 68]). The isoform of the

myosin heavy chain (MHC) expressed in the muscle fiber is

one of the most important factors influencing the speed of

contraction, because the rate of ATP hydrolysis determines the

speed of cross-bridge cycling and sarcomere shortening [69,

70]. Other factors affecting the contractile speed of muscle

fibers include the isoforms of the calcium reuptake and release

proteins expressed and the density of the SR [71, 72, 73, 74].

There are three major isoforms of MHC expressed in adult

human limb muscles: MHC I, also called slow myosin; and the

two fast isoforms, MHC IIA and MHC IIX (also called MHC

IID). Subtypes of these isoforms, as well as embryonic and

neonatal forms of MHCs, generate further diversity. The fast

and slow isoforms of myosin were first able to be distinguished

histochemically because of their differing amounts of ATPase

activity at acid and alkaline pH [75]. This histochemical difference

allowed fast and slow muscle fibers to be classified into

two types. Fast and slow muscle fiber types are further subdivided

by their dependence on aerobic or anaerobic metabolic

pathways. Muscle fibers that utilize oxidative metabolism for

energy needs have abundant mitochondria and lipid droplets.

In contrast, muscle fibers using anaerobic pathways for energy

tend to be richer in glycolytic enzymes with more abundant

glycogen stores. Histochemical methods for demonstrating

mitochondrial enzymes combined with myosin ATPase activity

have traditionally been used to define three major types of

muscle fiber in adult human limb muscles, described below.

The histochemical properties of different fiber types correspond

fairly well to their contractile properties, allowing muscle

fibers to be grouped into a small number of types by either

histochemical or physiological measures. It should be recognized,

however, that qualitative and quantitative differences in

expression‎ of fiber-type-specific proteins generate a continuous

range of physiological properties.

Type 1 muscle fibers have a slow twitch and use oxidative

metabolism. Type 1 fibers express MHC I, the slow isoform of

myosin, and contain many mitochondria. These muscle fibers

can be visualized histochemically by strong myosin ATPase

activity at low pH and by dense staining for mitochondrial

enzymes such as NADH dehydrogenase (i.e., nicotinamide

adenine dinucleotide, reduced) and SDH (i.e., succinate dehydrogenase)

(Table 1.2). Compared to Type 2 fibers, their SR is

less abundant, and it contains a slower isoform of the SR

calcium ATPase. Type 1 fibers contain myoglobin, a protein

that binds oxygen and confers a red color, and have a rich

capillary blood supply [76]. The metabolic profile and vascularization

render Type 1 muscle fibers highly resistant to

fatigue, and thus suitable for sustained contraction under

aerobic conditions. The acronym “SO,” slow oxidative, is used

by some to denote these fibers.

Type 2 muscle fibers are fast-twitch fibers, expressing fast

isoforms of myosin which exhibit strong ATPase activity at

alkaline pH. There are several subtypes of Type 2 fibers, but

two major subtypes occur in human limb muscles. Type 2A

fibers express the MHC IIA isoform of myosin. Compared to

Type 1 fibers, their SR is denser, and expresses isoforms of

calcium handling proteins that allow a more rapid cycling of

calcium ions from SR [71, 72, 73]. Mitochondria are relatively

abundant in Type 2A fibers. In addition Type 2A fibers contain glycolytic enzymes, such as phosphorylase, and have

abundant glycogen stores. These metabolic properties allow

Type 2A to function under aerobic and anaerobic conditions,

and provide them with a fairly high resistance to fatigue. Type

2A fibers have been denoted by the acronym “FOG” because

they are fast twitch with oxidative and glycolytic metabolic

capabilities. The third major muscle fiber type that occurs in

human limb muscles is the Type 2B fiber. Type 2B fibers

express the fastest isoform of myosin, MHC IIX (also known

as IID). Their SR is dense and contains a fast isoform of SR

calcium ATPase. Type 2B fibers have relatively sparse mitochondria,

but contain glycolytic enzymes and stores of glycogen.

Type 2B muscle fibers fatigue easily, but are suitable for

short bursts of anaerobic exercise. The acronym “FG,” fast,

glycolytic, is sometimes used.

Other isoforms of myosin are found in specialized muscle

or at different developmental stages. In a number of animal

species, Type 2B fibers express a very fast form of myosin, the

MHC IIB isoform, particularly in muscles with very fast speeds

of contraction [77, 78, 79, 80]. In humans, MHC IIB expression‎

has been reported in some cranial muscles [81] but it is

not expressed to a significant extent in limb muscles. Immature

forms of myosin are expressed by muscle fibers prior to

completing their differentiation during development [82, 83].

Fibers expressing immature forms of myosin that stain for

ATPase activity at acid and alkaline pH, Type 2C fibers, are found

in small numbers in normal adult limb muscles. The Type 2C

profile occurs in regenerating fibers, which can be common in

several muscular dystrophies. Muscle spindles also express a

mixture of immature and slow isoforms of myosin [84].

The classification of the major muscle fiber types by their

pattern of MHC expression‎ agrees relatively well with the

histochemical classification of fiber Types 1, 2A, and 2B that

is based on myosin ATPase activity at differing pH. However,

histochemical methods are relatively insensitive to hybrid

muscle fibers expressing more than one MHC isoform. Hybrid

muscle fibers can be demonstrated with immunocytochemical

methods or in-situ hybridization for different isoforms of

MHCs [80, 85]. Combinations of MHC IIA with IIx expression‎

are relatively common in Type 2 fibers, for example [85,

86]. In some muscles hybrid fibers make up a sizeable fraction

of the muscle fibers [78, 79, 85]. Hybrid fibers may play a role

in the ability of muscle fibers to undergo rapid adaptations in

response to training and use [87, 88, 89, 90].

Association of motor unit types with muscle

fiber types

All muscle fibers belonging to the same motor unit have the

same type, as judged from their staining for ATPase activity

[54, 91, 92] and MHC isoforms [93, 94, 95]. Within a muscle

unit the fibers also appear to have similar metabolic enzyme

capacities [94, 96]. It is, therefore, assumed that muscle fibers

within the motor unit also have essentially identical mechanical

properties. Edstrom and Kugelberg [54] were the first to

use the glycogen depletion method to examine the association

between the mechanical properties and histochemical characteristics

of the muscle fibers of individual motor units for two

types of fast-twitch motor unit in rats. Burke and coworkers

[55, 56] later used the same approach to examine the histochemistry

of muscle fibers within the full range of physiologically

identified motor units in the cat gastrocnemius muscle. In

these studies, motor neurons were characterized physiologically

with intracellular recordings, including stimulation with

short stimulus trains while measuring force output and prolonged stimulation to deplete glycogen stores in active

muscle fibers. Burke and coworkers found that motor units

differed in several mechanical properties, not just the speed of

contraction. These properties included the magnitudes of force

produced by individual twitches (twitch force) and the maximal

force produced by repetitive stimulation (tetanic force),

resistance to fatigue during sustained activation, and the ratio

of the twitch to the tetanic force [67]. These properties each

exhibited continuous distributions that initially made it problematic

to define distinct groups of motor units. However, two

criteria were found that permitted relatively clear clustering of

motor units into fast and slow groups in the cat: a “fatigue

index” based on the decline in force output during a defined

sequence of intermittent tetanization and a “sag property”

based on the shape of unfused isometric tetanic contractions

(Figure 1.8) [55, 91, 92, 97, 98]. Using these criteria, Burke and

colleagues were able to define three main types of motor units:

Type FF (fast twitch, fatigable), Type FR (fast twitch, fatigue

resistant) and Type S (slow twitch, fatigue resistant). Some

fast-twitch units exhibited fatigue resistance intermediate

between those of FF and FR units and were, therefore, referred

to as F(int) or FI [56, 92, 99]. Physiologically, there was a

perfect match between S, FR and FF motor units with the

histochemically defined muscle fiber Types 1, 2A, and 2B,

respectively (Table 1.2; see also [97, 98, 100]). They also found

some evidence that fibers in the minority F(int) unit type were

histochemically distinct from the three main types [56, 98].

These same physiological criteria have been used with somewhat

more variable success in classifying motor units in rat

muscles (e.g., [101, 102]). It is possible that some of the

variation in properties such as contraction time within a given

motor unit type are associated with hybrid combinations of

myosin isoforms, but this remains to be investigated

systematically.

Motor units in human muscles

There is a wealth of information available from EMG studies

in humans about the behavior of motor units in normal and

diseased muscle, and it has been known for some time that

fast- and slow-twitch muscle fibers coexist in human muscle

[103]. However, for obvious technical reasons, it is difficult to

examine the mechanical responses of individual motor units

under the controlled conditions possible in animal experiments.

Denny-Brown and Pennybacker [104] were the first

to record individual twitches from the fasciculations of motor

units in patients with motor neuron disease, using an indirect

pneumatic transducer. Buchthal and Schmalbruch [105] used a

mechanical transducer attached to a needle inserted into

tendons, plus intramuscular stimulation of small nerve

branches, to demonstrate that small groups of human motor

units in normal muscles generate a wide range of twitch

speeds, which varied in relation to the predominant local fiber

type (see also [106]). The introduction of spike-triggered computer

averaging into clinical neurophysiology made it possible

to record the responses of individual motor units with greater

assurance [107]. In this technique, discharges of single motor

units during steady voluntary contractions are used to trigger

an averaging computer while measuring the force produced by

an appendage (e.g., a finger) attached to a force transducer.

There are two limitations of this technique. First, the recorded

twitch responses are not isolated twitches but rather components

of unfused tetani, leading to errors in estimating the

twitch forces and contraction times [108, 109]. Intra-neural

stimulation of single motor axons to produce twitches has

been used in an attempt to overcome this problem [110, 111,

112]. Secondly, the mechanical responses measured can be

significantly degraded by the compliance of components

between the active muscle fibers and the force transducer,

including tendons of various lengths. Despite these technical

limitations, most of the contractile properties measured from

human motor units are generally similar to those from animals

[113, 114]. There is disagreement about whether fatigability

and “sag” can be used to classify human motor unit types in

the same manner as in animals, and whether force measurements

relate to the fatigability in the same way [110, 112, 115].

However, when motor units have been identified by glycogen

depletion in muscle biopsy samples, these properties were

consistent with histochemical identification [116]. Overall,

the available physiological evidence and correspondence with the histochemical classification strongly suggest that the basic

characteristics of Types S, FF, and FR human motor units are

similar to those described for the cat and rat.

Functional correlates of fiber properties

and motor unit types

It is clear that many factors contribute to mechanical properties

of the different motor unit types: in addition to the expression‎

of MHC isoforms, there are fiber-type-specific differences

in myosin light chains, troponin and tropomyosin proteins,

proteins involved in calcium release and reuptake, and sarcotubular

structures [72, 73, 74, 117]. It seems likely that the

“sag” property, which differs sharply in fast and slow units, is

produced by interactions among these factors [67, 68, 118].

Resistance to fatigue is directly related to the oxidative capacity

of the different fiber types (Table 1.2; [91, 119]), as well as to

their mitochondrial content [72] and local capillary supply

[120]. These correlations are certainly causally related. The

forces produced by individual motor units can vary by over

two orders of magnitude during tetanization, and this

variation is correlated with motor unit type (Figure 1.8 and

Table 1.2). The force produced by a motor unit is a function of

the effective cross-sectional area of its muscle fibers and the

specific force output of that fiber type per unit area. Estimation

of the effective cross-sectional area must take into account the

effective innervation ratio [121], which may approximate the

actual innervation ratio in pinnate muscles [91] but would be

less in interdigitated muscles which have unit fibers in serial

arrays (Figure 1.5). In general, Type 1 and 2A fibers have

smaller diameters than Type 2B, making fiber area an important

component of the equation. In humans, Type 2 fibers

exhibit the greatest variability in diameter; in general fiber

diameters tend to be larger in men than women [122]. There

is some controversy about whether specific force output, which

cannot be measured directly, differs between units with Type 1

and 2 muscle fibers [65, 69, 91, 92].

Motor neurons and synaptic specializations

In view of the differences between muscle fiber types, it is not

surprising that the motor neurons that innervate them exhibit

corresponding physiological differences (Table 1.3; reviewed

by [67]). In general, motor neurons of Type S motor units have

slower axonal conduction velocities, longer durations of postspike

hyperpolarized after-potentials (AHPs), and higher

whole-cell input resistance values than the cells that innervate

either FR or FF motor units. The AHP duration is particularly

important because it is a key factor that controls the rate of

motor neuron firing; motor neurons of Type S units have the

longest AHPs and generally fire more slowly than those of FR

or FF units. When examined with intracellular labeling

methods, the motor neurons of Type S units tend to be smaller

in membrane area than Type FF cells; Type FR motor neurons

are intermediate in size [9, 10]. There is no systematic difference between axonal conduction velocities of FF and FR

unit groups [123]. Although the distributions of motor neuron

properties are continuous and exhibit large overlaps when

sorted according to muscle unit type, the relative excitability

of the motor neurons to depolarizing currents injected directly,

measured as the rheobase (the amount of current

required to produce action potentials reliably), is more closely

related to unit type than other measures [124, 125]. The

rheobase data imply that intrinsic motor neuron excitability

varies according to the sequence S>FR>FF, which has

important implications for the recruitment order of motor

units (Figure 1.9).

The strength of several synaptic inputs to motor neurons

shows type-related differences that are undoubtedly related to

the way in which the various types of motor units are used

during activity. For example, the average amplitudes of monosynaptic

excitatory postsynaptic potentials (EPSPs) produced

in motor neurons by group Ia muscle spindle afferents, which

are largely responsible for the stretch reflex, are ordered as

S>FR>FF (Table 1.2) [126, 127]. The same ordering is

evident with the disynaptic inhibition produced by stimulation

of group Ia afferents from antagonist muscles [126] and with

disynaptic recurrent inhibition produced by Renshaw interneurons

activated from motor axon collaterals [128]. The

organization of synaptic efficacy is a key factor that controls

the function of motor unit populations [129], and for most

inputs to motor neurons, the ordering of synaptic efficacy

follows the size principle. However, there is evidence that

certain cutaneous inputs and supraspinal systems, notably

the rubrospinal tract, tend to excite relatively high-threshold

motor neurons while inhibiting low-threshold cells [130, 131,

132], a pattern opposite to that found in group Ia excitation.

Although there would be potential advantages to competing

control systems that could bypass low-threshold, slow-twitch

motor units that are slow to relax, the idea that large, ...

더 없음.