Diagram of typical motor unit belonging to an ankle extensor muscle in the cat hindlimb, including the motoneuron lying in the spinal cord, its axon leaving the cord via a ventral root and peripheral nerve to reach the target muscle, where it branches profusely to innervate a set of muscle fibers (muscle unit) distributed within the anatomical confines of the muscle (see also Figs. 12, 13).

Histochemical profiles of fibers in heterogeneous (A, C) and homogeneous (B, D) cat muscles. Photomicrographs of representative serial sections of heterogeneous lateral gastrocnemius (LG; panels A, C) and homogeneous soleus (SOL; panels B, D) muscles in the same cat processed as a single block of tissue and photographed under identical conditions. Panels A and B are from a section stained for myofibrillar ATPase activity after incubation in an acidic buffer at pH 4.65 (see Table 1; AC‐ATPase). Panels C and D are from a serial section stained for oxidative enzyme NADH‐dehydrogenase (Table 1). When comparing AC‐ATPase and NADH staining patterns for individual fibers (the histochemical profile), at least 3 fiber types can be recognized in LG but only 1 in SOL. Note that appearance of an LG fiber with the profile characteristic of type S muscle units (panel C, arrow) is not identical to appearance of SOL fibers. Calibration bar in C is 100 μm.

Ultrastructural features of histochemically identified muscle fibers. Scatter diagram comparing Z line width (abscissa) and mitochondrial volume (ordinate, expressed as percentage of fiber core volume) measured from electron micrographs of guinea pig medial gastrocnemius muscle fibers that had been frozen, thawed, and then fixed. Serial cryostat sections of the same fibers were stained for myofibrillar ATPase at alkaline pH (mATP) and for the oxidative enzyme succinic dehydrogenase (see Table 1), permitting comparison of ultrastructural features in histochemically typed fibers. Using nomenclature of Table I, crosses are type I, filled circles are type IIA, and open circles are type IIB.

Physiological profiles of motor unit samples from 3 different cat muscles. Left, comparison of units in medial gastrocnemius (MG) and soleus (SOL). Scatter diagrams show tetanic tension outputs (ordinate) and isometric twitch contraction times (abscissa). Note that data scatter is much greater for MG sample than for SOL. Motor unit types denoted by the following symbols: type FF units, open circles; type FR, filled circles; type S, filled triangles. Right, 3‐dimensional diagram comparing physiological profiles of motor units in cat superficial lumbrical muscle (filled circles) with units in cat medial gastrocnemius (open circles). Tetanic tension output (vertical ordinate) is normalized as percentage of the largest unit in each sample. Note precise, monotonic relation between all 3 variables in data from lumbrical sample compared to greater scatter of points in gastrocnemius sample. The range of variation for tetanic tensions and contraction times was approximately the same in the 2 muscles (although with different absolute values), but the range of axonal conduction velocities was much wider for lumbrical than for gastrocnemius units.

Multivariate physiological profiles of motor units in cat medial gastrocnemius (MG) muscle, displayed on 3‐dimensional graphs. Left, data from a sample of 81 MG units studied in 3 cats. Stippled circles denote units without “sag” in unfused tetani (type S units); open circles denote units with sag (type F). Type F units are divided according to fatigue index as follows: type FF, fatigue index less than 0.25; type F(int), fatigue index between 0.25 and 0.75; type FR, fatigue index greater than 0.75. Increasing values of fatigue index denote increasing resistance to fatigue during a standardized sequence of tetani. Right, diagram same as left but including data from an additional 28 MG units, each of which was also studied histochemically by the glycogen‐depletion method (see text). Histochemical profiles of the 4 units denoted by arrows (1 of each motor unit type) are illustrated in Figure 6.

Histochemical profiles of 4 physiologically identified muscle units from cat medial gastrocnemius (MG; see Fig. 5), representing each of the unit types present in that muscle. Muscle fibers belonging to the studied units were identified in each case by glycogen depletion (not illustrated, but see Fig. 12) and are indicated by stars in left column of photomicrographs. Left column, sections stained for myofibrillar ATPase (M‐ATPase) activity at pH 9.4, in which the F units are all heavily stained (type II) and the S unit is relatively light (type I; see Table 1). Distinction between the various type II fibers can be made on the basis of ATPase activity after acidic preincubation (AC‐ATPase, pH 4.6, middle column, in which 3 staining levels can be seen) and by relative staining for oxidative enzyme, NADH dehydrogenase (NADH‐D; right column), for which all of the unit types except type FF display rather heavy staining. With reference to Table 1, FF unit fibers are of the IIB histochemical type, F(int) unit fibers are type IIAB, FR fibers are type IIA, and S units fibers are type I.

Histochemical profiles (determined in glycogen‐depletion experiments) representative of physiologically studied muscle units in rat soleus (SOL) muscle, separated according to their isometric twitch contraction times (in milliseconds; numbers above each column). Shading denotes staining intensity and size of circles represents relative fiber area. Top 3 rows represent ATPase staining under different conditions (alkaline and acid pH as in Table 1; FIX, after cold formaldehyde fixation). Bottom row shows staining for the oxidative enzyme, succinic dehydrogenase (SDH; see Table 1). Note parallel staining pattern between alkaline ATPase and SDH and inverse relation of both to fiber area, which is different from the pattern found in most other heterogeneous muscles (see Table 1, Fig. 6).

A particularly clear example of dissociation of EMG decrement from mechanical fatigue during intermittent tetanization of type FF muscle unit in cat MG muscle. Mechanical (A) and electromyographic (B; note faster time base) responses during repeated tetanization (13 pulses at 40 Hz repeated every 1 s) show marked slowing of mechanical response after only 26 tetani (26‐s record) and almost complete mechanical fatigue after 70 tetani (910 pulses in 70 s), when EMG responses are almost unchanged in amplitude, although individual spikes are widened. Up to this point, fatigue process appears to be primarily, if not exclusively, intrinsic to muscle fibers themselves. Markedly diminished EMG spike after 376 tetani could be due either to change in muscle fiber action potentials or to failure of neuromuscular transmission, or both (see subsection FATIGABILITY AND METABOLIC PROFILES, p. 358). (From R. Burke, D. Levine, P. Tsairis, and F. Zajac, unpublished records; see ref. 111.)

Three‐dimensional diagrams illustrating physiological profiles of motor units sampled from 2 different human muscles. Upper graph: motor units sampled by spike‐triggered averaging from human 1st dorsal interosseous muscle (1st DI). Lower graph: motor units sampled by intramuscular stimulation from human medial gastrocnemius. Symbols and shading denote units and data regions interpreted to be equivalent to the type FF (filled circles), type FR (open circles), and type S (triangles) motor units described in the cat MG (see Fig. 5). Definition of fatigue index differs in the 2 studies but, as in Fig. 5, increasing values denote increasing fatigue resistance. Note that pattern of correlations between contraction time, force output, and fatigue resistance are generally similar to those found in cat muscle (compare with Fig. 5), although the data distribution in the interosseous sample seems more continuous, in keeping with the cat lumbrical muscle (see Fig. 4, right).

Three‐dimensional diagram illustrating physiological profiles (isometric‐twitch‐rise time and fatigue index) of motor units sampled from various hindlimb muscles (see symbol key) of cats of different postnatal age (left abscissa). Note that most units found during the 1st 2 postnatal wk are slowly contracting and relatively fatique resistant, and that adult pattern becomes evident (despite small sample sizes) between 40 and 70 days after birth.

Physiological evidence for innervation of extrafusal and intrafusal muscle fibers by a single skeletofusimotor (β) axon (conduction velocity 41 m/s) in rabbit lumbrical muscle. Upper trace in each record is isometric lumbrical force; lower trace is discharge of a primary (group Ia) muscle spindle afferent from the same muscle. Stimulation of the motor axon at 10/s (record 1) produces extrafusal twiches and brief bursts of Ia activity during each relaxation phase; tetanization at 150/s (record 2) produces fused mechanical response and silencing of afferent discharge, as would be expected for the α‐innervation pattern. However, after partial curarization (record 3) sufficient to block extrafusal (note mechanical failure) but not intrafusal neuromuscular transmission, the same tetanus produces acceleration of Ia discharge, showing that motor axon must also inactivate intrafusal muscle fibers.

Medium magnification photomicrograph of glycogen‐depleted fibers belonging to a type FF muscle unit in cat medial gastrocnemius (MG), showing relatively low density of unit fibers and rather uniform scattering through the unit territory. A full reconstruction of this muscle unit is shown in Fig. 13 (Left MG). Area illustrated measures approximately 8.0 × 5.4 mm.

Reconstructions of the territories of a type FF muscle unit in cat medial gastrocnemius (MG, left panel; see also Fig. 12) and of a type S unit in cat soleus (SOL; right panel). In each case, glycogen‐depleted muscle fibers (dots) are plotted on outlines of cross sections taken along the muscle at different levels, as indicated on whole muscle diagrams on right of each panel. The angulation of fiber bundles within MG and SOL is shown on longitudinal section diagrams; shading denotes approximate extent of unit territory projected onto muscle surface in each view. Counts of glycogen‐depleted fibers in each cross section are indicated on section maps. Territory of SOL unit appears to occupy a larger fraction of the whole muscle volume than does the MG unit.

Three‐dimensional diagram showing dependence of isometric tetanic force from a muscle unit on unit innervation ratio, mean fiber area, and specific tension per unit fiber area. The 2 planes in the diagram represent loci for different specific tensions: 2.2 kg/cm² (the larger slope plane, for type F units) and 0.6 kg/cm² (the lower plane for type S units: see Table 2 for derivation of these estimates for cat medial gastrocnemius muscle). Innervation ratio for an individual motor unit can be estimated when its tetanic force output and average fiber area is known and a reasonable estimate of the specific force output per unit fiber area can be assumed. Plotted points represent different medial gastrocnemius (MG) motor units for which force and fiber area data are available 121; open circles denote FF units; filled circles are type FR units; and filled triangles (plotted on the 0.6 kg/cm² plane) represent type S units (see Table 2). Numbers next to 3 of the points are innervation ratio estimates from reconstructions of glycogen‐depleted units 121. If the values assumed for specific tension per unit area are correct, these data suggest that innervation ratios for cat MG units vary over a 4‐ to 5‐fold range, with little systematic difference between fast‐ and slow‐twitch units.

Reconstructions of medial gastrocnemius (MG) and soleus (SOL) motor nuclei in a cat spinal cord, showing a dorsal view (right, with lateral spinal cord margins indicated by solid lines) and 5 levels of cross sections (left, levels indicated by letters and dashed lines on dorsal view diagram). Motoneurons (dots) were identified after retrograde transport of horseradish peroxidase injected into left MG and right SOL muscles of the same animal. Reconstructions were made from serial sagittal sections and show all labeled cells on the dorsal view. Cells indicated on cross‐section reconstructions were only those within 300 μm of the selected level. The MG and SOL nuclei share the same motor cell column and are about the same overall size even though SOL contains only about one‐half the number of motoneurons.

Comparison of average soma sizes of cat medial gastrocnemius (MG) and soleus (SOL) motoneurons. Histogram on left shows average soma diameter of MG and SOL motoneurons labeled by retrograde transport of horseradish peroxidase (HRP) from same preparation as illustrated in Fig. 15. The smaller peak in this bimodal size distribution is assumed to represent γ‐motoneurons. Superimposed symbols denote average soma measurements of HRP‐labeled cells (9 gastrocnemius and 1 SOL) like those shown in Fig. 17, with unit type as follows (○ = FF; ⊗ = F(int); • = FR; ▴ = S). Histograms on right are a more extensive set of data on type‐identified, HRP‐labeled triceps surae motoneurons. These data indicate that, in general, the motoneurons of type S units are somewhat smaller than those of F units, but there is considerable overlap between the groups, at least in some dimensions.

Montage photomicrographs of 3 gastrocnemius motoneurons, identified as to motor unit type and labeled by intracellular iontophoresis of horseradish peroxidase (see ref. 124 for methods). Left panel shows, at low magnification, a type FR motoneuron in thick section, superimposed on Nissl‐stained cell bodies of other ventral horn cells. The complete dendritic tree is not evident, but it is clear that dendrites extend out of the ankle extensor cell column into the columns of toe plantar flexors (dorsal) and hamstring (ventral) nuclei. Right, higher magnification shows, in another cat, a comparison in which type S motoneuron is clearly smaller and has fewer main stem dendrites than the FF neuron. These differences are not always so pronounced, as illustrated in Fig. 16.

Photomontage reconstruction of 5 HRP‐labeled α‐motoneurons: 4 medial gastrocnemius (MG) and 1 soleus (SOL), showing appearance of longitudinal dendritic associations between neighboring cell despite the basic radial arrangement of dendrites of any 1 cell. Reconstruction made by superimposing low‐power photomicrographs, each enlarged onto photolithographic film, of 5 serial sections (pseudosagittal plane of sectioning shown in inset on lower left). Sections were each 75 μm thick so total reconstruction includes about 475‐μm thickness. (From R. E. Burke, M. J. O'Donovan, M. J. Pinter, and A. Lev Tov, unpublished experiments; see ref. 124 for methods.)

Graph showing relation between diameter of intraspinal motor axon (abscissa) and number of swellings (presumably synaptic terminations) on recurrent collaterals from same axons, measured from type‐identified motoneurons labeled by intracellular horseradish peroxidase iontophoresis (as in Fig. 17). •. Gastrocnemius type FF; ⊗, gastrocnemius FR; ▿, gastrocnemius S; ○, SOL type S. Distribution of axonal diameters fits axonal conduction velocities for same unit types (see Figs. 20, 21). The FF units as a group exhibit significantly more recurrent collateral swellings than the other types that, in contrast, display a positive linear correlation between axon diameter and number of collateral swellings.

Relation between axonal conduction velocity (abscissa) and motoneuron input resistance (ordinate; note logarithmic scale) in a sample of type F (open circles) and type S (crossed circles) motoneurons of cat gastrocnemius, plus a group of type S units from soleus (filled circles). There is a clear negative correlation despite considerable scatter. Note that input resistance values of gastrocnemius S units is, on the average, somewhat less than those of soleus, as fits with more recent anatomical data; see subsection Motoneuron Anatomy in Relation to Unit Type, p. 377, especially Figure 16.

Three‐dimensional diagram illustrating interrelations between muscle unit tetanic‐force output, motor axon conduction velocity, and amplitude of homonymous (i.e., medical gastrocnemius nerve stimulation) group Ia EPSPs in a sample of MG motor units. There is a rather good negative correlation between tetanic tensions and Ia EPSP amplitudes, but neither of these variables is well correlated with axonal conduction velocities. The type F group (denoted by circles) exhibits a narrow range of conduction velocities but broad ranges of both tension output and Ia EPSP size. Open circles, type FF; crossed circles, type F (int); filled circles, type FR; cones, type S.

Comparison of polysynaptic potentials (PSP) produced in medial gastrocnemius (MG) motoneurons by stimulating ipsilateral sural nerve at 5× threshold for the most excitable fibers. Inset records show computer averages (20 sweeps) of sural PSPs produced in 3 motoneurons recorded sequentially in the same cat within a 90‐min period, showing marked difference in amount of early excitation between the F and S units. Brackets indicate 20‐ms epochs over which voltage‐time integrals of depolarizing and hyperpolarizing components were integrated for display in lower graph (data points indicated by arrows). Each of the 105 points in the graph came from a different MG motoneuron, with membrane potentials ranging between −60 and −75 mV. Open circles, type FF; crossed circles, type F(int); filled circles, type FR; triangles, type S. Despite some overlap in scatter, type S motoneurons have significantly smaller early polysynaptic excitation than type F units.

Three‐dimensional diagram showing interrelation between motoneuron input resistance (A_N) and maximum amplitudes of homonymous (MG) and heteronymous (LG‐SOL) group Ia EPSPs in a sample of MG motoneurons. Open circles denote type F units; filled circles denote type S units. Trapezoid drawn through data points indicates least‐squares fit to data. Although all 3 variables display positive correlations, that between MG and LG‐SOL EPSP amplitudes is much stronger (r = 0.64) than the correlation between either set of EPSP data and R_N (0.44 and 0.40, respectively). If the F and S groups are considered separately, there is little correlation between R_N and EPSP amplitude, whereas that between the EPSP amplitudes themselves remains evident. The dissociation between EPSP size and R_N suggests that input resistance per se is not a critical factor in regulating synaptic efficacy in this system.

Relations between muscle unit mechanical properties, obtained by spike‐triggered averaging method)see subsection Motor Units in Human Muscle, p. 359), and functional thresholds of their motoneurons in a human hand muscle (first dorsal interosseous). The ordinate for both graphs is the “twitch” force, which is positively correlated with the voluntary threshold force required for steady firing (left graph) and negatively correlated with twitch contraction time (right graph). Points plotted in < 20 column on left graph all had force thresholds less than 20 g but were not more precisely characterized. Comparison of the 2 graphs indicates that motor units with lowest voluntary thresholds tend to produce small individual forces, and many have slow contraction as well. The highest threshold units all tend to have both rapid contraction and large force output.

A “recruitment model” for cat medial gastrocnemius (MG) motor pool using Ia EPSP amplitude as the “recruitment rule” to generate a sequential ordering of units. The MG motor units included in this 3‐dimensional diagram are arrayed incrementally along the right‐hand abscissa (% MG Pool Recruited) in strict accord with decreasing amplitude of the MG group Ia EPSP measured in each motoneuron (data from ref. 118; see also Fig. 21). The isometric tetanic force produced by each individual unit is plotted against the left‐hand abscissa (Unit Tetanic Tension). Vertical dimension is cumulative isometric force that would be produced as each unit is recruited in sequence, expressed as percentage of total force (i.e., 100%) expected from the entire MG muscle when all units are added together. Filled symbols indicate fatigue‐resistant unit types, with types FR and F(int) indicated by filled ovals and type S units by filled triangles. Type FF units are indicated by open ovals. The ranges of forces actually produced by MG muscle in intact cats during standing, walking, running, and maximum vertical jumping are indicated along the left‐hand vertical ordinate (data from ref. 707), expressed as percentage of maximum MG output during fused isometric tetani. When ordered according to Ia EPSP amplitude, the first‐recruited half of the MG population consists almost entirely of fatigue‐resistant units [types S, FR, and F(int)], which together produce only about 25% of the maximum force available from the MG unit population. See text for further discussion.

Examples of differential modulation of functional thresholds of motor units within the same pool. A: motor axons recorded in a fine filament of the medial gastrocnemius (MG) muscle nerve during tonic vibration reflex (decerebrate cat). One α‐motoneuron axon (unit A, conduction velocity about 73 m/s) showed low‐threshold firing to stretch and vibration but was slowed (i.e., inhibited) by a light pinch to the ipsilateral ankle at the same time that 2 other MG axons with much higher functional thresholds (B and C, conduction velocities 79 and 85 m/s, respectively; both unresponsive to stretch and tendon vibration) were excited to discharge. The whole soleus (SOL) muscle (lower traces) was inhibited at the same time as MG unit A. B: EMG recordings of 2 motor units in human first dorsal interosseous (1st DI) muscle (upper traces) during isometric ramp contractions (lower traces, irregular line) in which subject repeatedly tracks a ramp force target (regular sawtooth). Note small EMG spike of low‐threshold unit (A) and larger spike of much higher threshold unit (B) in top set of traces. Second set of traces, during delivery of continuous electrical stimuli (50 pulse/s at strength 4 times perceptible threshold) to the index finger skin, show that the force threshold for unit B decreased whereas that of A increased, so that B occasionally fired before A. After skin stimulation ended (lower sets of traces), relative thresholds of units A and B returned slowly to control values. Force calibration is 2 N for all records. C: graph of voluntary force thresholds for 1st DI units before (left ordinate, mean values) and during skin stimulation (right ordinate, maximum or minimum values, depending on trend of overall change), from experiments as in B. Skin stimulation raised thresholds for most low‐threshold units and decreased them in high‐threshold units, resulting in an overall narrowing of the threshold differences. Simultaneous enhancement of excitability in some units and depression of others within the same motor unit population can be regarded as a major change in the pattern of functional thresholds.

Diagram of a recruitment model in which functional thresholds of motoneurons (MNs, labeled A–D) are controlled by the efficacy of synaptic input to them, represented by width of arrows. Synaptic input

Steady firing frequencies (ordinate) observed for individual motor units in human extensor digitorum communis muscle, recorded throughout a wide range of voluntary forces (abscissa). Note that virtually all units begin firing at about the same steady rate (8 Hz) and increase in rate to 16–24 Hz as total force output increases. Note also that highest threshold units display smallest range of variation but have high‐force frequency slopes.

Comparison of firing patterns (upper traces in A, B) from an individual motor unit in human ankle dorsiflexor, tibialis anterior, during a slow ramp contraction reaching 10 kg of force in 5 s (record A; lower trace is the force record) versus a ballistic contraction reaching 10 kg in 0.11 s. (record B; note change in time scale). The unit of interest (large spike) began to fire before any measurable force output in the ballistic contraction (i.e., its force “threshold” was 0), but during the slow ramp contraction its force threshold was about 2 kg. Graphs in C, D, and E illustrate the mean instantaneous frequency (ordinate; open circles) attained at the output force indicated on the abscissa during 10‐kg ramp contractions lasting 1 s, 5 s, and 10 s, respectively. Filled circles in C show the mean instantaneous frequency attained during 5 ballistic contractions like that in record B. The firing patterns are similar in all of the ramp contractions, with gradually increasing instantaneous frequency (10–30 Hz), but the pattern is reversed in the ballistic contractions, with the highest frequency at the onset of activity (60 Hz).

Isometric force records from a type F (A, B) and a type S (C, D) motor unit showing (A, C) superimposed responses to 1, 2, 3, and 4 pulses at interpulse intervals noted at the end of each record (in ms) and (B, D) unfused tetanic responses to constant‐frequency pulse trains, again with intervals noted at right of each record. Note presence of “sag” (see also subsection MOTOR UNIT POPULATION OF CAT MEDIAL GASTROCNEMIUS MUSCLE, p. 353) in all of the F unit tetani but only in the slowest (140‐ms interval) tetanus to type S unit. Records in A, C show that at shortest stimulus interval (10 ms), the 2nd pulse in a train adds the greatest increment of force and force‐time integral (see Fig. 31).

Graphs showing force‐time integrals (ordinates) of mechanical responses from same units illustrated in Figure 30, plotted in each case against the interval between pulses in the input trains (abscissae). Graphs A and C show, for the F and S units respectively (sample records in Fig. 30A, C), the force‐time integral added to the already developed isometric force by the 2nd (A₂), 3rd (A₃), and 4th (A₄) pulses at the indicated intervals, with integral expressed relative to that produced by the single twitch (A₁). In general, A₂ is predominant when 2nd pulse arrives at intervals less than the isometric twitch contraction time (T_C); at longer intervals, each pulse produces about the same increment in force‐time integral. Lower graphs (B, D) show cumulative force‐time integral for 1st 15 pulses (A_T; plotted relative to left‐hand ordinates) during different trains (sample records in Fig. 30B and D). The abscissa scale is exactly the same as in A and C but indicates absolute interpulse interval in ms. Note that shape of the A_T curves roughly match those of A₃ curves above. Curves labeled P_T indicate absolute isometric force attained at peak of 15th response during tetani at the various frequencies (referred to the right‐hand ordinate in each case). Maximum slope of P_T curve occurs with intervals that produce the largest force‐time integrals.

The “catch” effect in a type S muscle unit from cat medial gastrocnemius (MG). Superimposed traces of 3 isometric responses to pulse trains with average frequencies of about 12.5 pulses/s but different interval structures (diagrams a–c below). The addition of a single extra pulse (arrow) within 10 ms of the 1st pulse in the basic train (pattern a) produced marked and persistent enhancement of force output (record a) even though the basic train rate is unchanged. Widening one interval slightly during this catch enhancement (pattern b) caused a drop in output force that was then “caught” again at a new force level (record b). A modest shortening of a single interval (pattern c; arrow) produced an equivalently modest catch enhancement (record c).

Effect of posttetanic potentiation (PTP) on mechanical responses of a type F muscle unit in cat medial gastrocnemius (MG). Top records are superimposed traces of a single twitch and a fused tetanus before (A) and after (C) repeated high‐frequency (200 Hz) tetanization. Note that there is a marked increase in twitch force and duration and more rapid rising phases of the twitch and tetanus after PTP, but that there is no change in fused‐tetanus plateau force. Lower records illustrate unfused tetanic responses to low‐frequency (approximately 20 Hz) stimulation with (larger response) and without a single extra pulse within 10 ms of the 1st (a “doublet,” to illustrate the catch effect; see Fig. 32). In contrast to the S unit in Fig. 32, catch enhancement is not sustained (record C); instead, force output decays with a time course similar to that of the “sag” of tension in the basic tetanus. After PTP the initial doublet produces much less catch enhancement, and the tetani with and without the extra pulse converge within 200 ms.

Input‐output graph (as in Fig. 31B and D) for a type F muscle unit in cat medial gastrocnemius (MG), showing the effect of posttetanic potentiation (PTP) on mechanical output during short tetani at various frequencies. Continuous lines denote force‐time integrals under 1st 14 responses (A_T) in tetani under initial conditions [i.e., before repeated tetanization to produce maximal PTP; curve A_T1, filled squares] and after tetanization [i.e., during PTP; curve A_T2, open squares]. Plateau force reached by the 14th response during the same tetani (P_T; dashed curves) is shown for initial conditions [P_T1; filled circles] and during maximum PTP (P_T 2; open circles). Both curves converge to the same maximum force in fused tetani (P_O), but the potentiated tetani generate more peak force at all lower frequencies (see also Fig. 33). The PTP produces a greater increase in twitch force (arrows labeled P_tw, referred to right hand ordinate) than in twitch contraction time (arrows labeled T_C, referred to lower abscissa). The PTP has the effect of lowering and broadening the range of input frequencies over which muscle unit output is most effectively modulated by input rate (compare with Fig. 31 and attendant discussion). [From Burke et al. 116.l

Three‐dimensional diagram summarizing the interrelations between a variety of features of the physiological, morphological, and histochemical profiles of motor units found in the cat medial gastrocnemius (MG) population. Muscle unit properties include tetanic force output, twitch contraction time, resistance to fatigue, myofibrillar ATPase staining at alkaline pH, metabolic enzyme patterns (anaerobic and oxidative), and average muscle fiber diameter. Motoneuron properties included are axonal conduction velocity and duration of the afterhyperpolarization. Contrasting patterns of synaptic organization are represented by amplitude of monosynaptic group Ia EPSP (right abscissa) and by the apparent strength of polysynaptic excitatory input from distal hindlimb skin (sural nerve; left abscissa). The diagram is based on displays such as in Figs. 2 and 5 and on material like that in Tables 1–3. It should be clear that this summary somewhat distorts reality in that the various features specified are not all distributed in exactly the way displayed, but the general trends are as illustrated. The diagram is not intended to be all inclusive (other characteristics could be added to one or another axis) but rather emphasizes the fact that a great many properties of motoneurons and their muscle units display correlated variation, which can be communicated by the notion of motor unit “types.” The types identified in the cat MG, along with their percentages in that population, are indicated on the shaded boxes, which denote the approximate loci of the clusters of data points seen in the actual data displays, such as in Figure 5.