Hyperosmotic neurosecrtion was used to measure basal and insulin-stimulated amino acid and myoinosital transport into rat motor nerve terminals. L-Alanine and a(methy1amino)-isobutyricacid (a nonmetabolizable system A-specific analog) transport was rapid into motor nerve terminals innervating a fast-twitch muscle, the extensor digitorum longus, and slow into motor nerve terminals innervating the soleus, a slow-twitch muscle. A physiological concentration of insulin, 10 p,U/mL, increased L-alanine and a(methylamino)-isobutyric acid transport into motor nerve terminals in the soleus. Large doses of insulin, 100 or 1000 p,U/mL, had no effect on L-alanine or a(methylamin0)-isobutyricacid transport into nerve terminals in the extensor digitorum longus. There was negligible basal or insulin-stimulatedtransport of D-alanine or myoinositol into nerve terminals of the soleus or extensor digitorum longus. These studies show that insulin regulates sterospecific amino acid transport into soleus motor axons, but has no effect on the rapid amino acid transport into extensor digitorum longus motor axons. Differences in basal and insulin-stimulated tranpsorf suggest that motor axons differ in their metabolism, and might be selectively vulnerable to disease processes. Key words: amino acid biological transport insulin motor nerve myoinositol MUSCLE & NERVE 15:761-767 1992
INSULIN INCREASES AMINO ACID TRANSPORT INTO RAT SOLEUS JACKSON 6. PICKETT, MD
A motor unit consists of an anterior horn cell and the muscle fibers it innervates. Motor units have been classified according to the speed of contraction of their muscle fibers, rapid for type F and slow for type S. Type F and S muscle fibers differ in their histochemical, biochemical, and physiological properties.' Sickles and OblakI4 discovered that the concentration of oxidative enzymes in anterior horn cells paralleled that in their muscle fibers. Also, glucose transport, as measured by an indirect method developed by Shimoni and Rahamimoff,13 is rapid into motor axons innervating a fast-twitch muscle, the rat extensor digitorum longus, and slow into those innervating a slow-twitch
From the Veterans Affairs Medical Center and Department of Neurology, Medical University of South Carolina, Charleston, South Carolina. Acknowledgments. Research supported by Veterans Administration and the Juvenile Diabetes Foundation. Thanks to Mrs. Linda M. Wade for preparation of the manuscript, and Drs. Henry F Martin and Darrell D. Wheeler for review of the manuscript. Address reprint requests to Jackson B. Pickett, MD Veterans Affairs Medical Center (127). 109 Bee Street, Charleston, SC 29401-5799. Accepted for publication September 1 , 1991 CCC 0148-639X/92/070761-07$04.00 0 1992 John Wiley & Sons, Inc.
Insulin Increases Amino Acid Transport
muscle, the rat so leu^.^^'^" Shimoni and Rahamimoff13 were the first to show insulin increases glucose transport into frog motor axons. 'These findings have been extended by a recent reportg suggesting that insulin improves rat soleus motor axon function by increasing glucose transport, and has no effect on extensor digitorum longus axon function or glucose transport. These studies suggest anterior horn cells and their axons are as metabolically specialized as their muscle fibers. Because insulin regulates glucose transport into soleus motor axons, but not extensor digitorum longus motor axons, insulin might also increase amino acid and myoinositol transport into soleus motor axons, but not extensor digitorum longus motor axons. Of the many amino acid transport systems, insulin regulates only system A, which transports small neutral amino acids such as alanine, glycine, proline, and ~ e r i n e . ~ has reported that a small neutral amino acid transport system exists in frog sciatic nerve. This report investigated basal and insulin-stimulated amino acid and myoinositol transport into motor axons innervating rat soleus and extensor digitorum longus muscles. An abstract of this work has been published.'
MUSCLE & NERVE
July 1992
761
METHODS
Nerve-muscle preparations were dissected from male Sprague-Dawley rats (60 to 230 g) and pinned in a small transilluminated recording chamber, which was kept at 31.5 to 32.5" C. The bathing solution was continuously exchanged between the recording chamber and a 200-mL reservoir at a rate of 10 mL/min, so that solution changes were complete within 2 minutes, as measured by dye washout. Solution changes were accomplished by changing the reservoir from which the bathing solution was drawn. The bathing solution contained 137 mmolfL NaCl, 5 mmol/L KC1, 2 mmol/L CaCl,; 1 mmol/L NaH,PO,, 1 mmol/L MgCl,, 24 mmol/L NaHC03, and 11 mmol/L D-glucose; after bubbling with 95% 0, and 5% CO,; the pH was 7.3 to 7.4. Purified pork regular insulin (Iletin 11) was obtained from Eli Lilly and Co.; a(methy1amino)-isobutyric acid, myoinositol, and alanine from the Sigma Chemical Co. Miniature end-plate potentials were detected with 5 to 20-Mfl microelectrodes and conventional methods for intracellular recording. After suitable amplification, the resting potential and miniature end-plate potentials were recorded on film. Recordings were limited to cells where: (1) the miniature end-plate potential rose to a peak in less than 0.5 ms; and (2) the miniature end-plate potential amplitude was well above the amplifier noise (usually 200 pV peak to peak). Increasing the osmotic gradient across the motor nerve terminal membrane caused the miniature end-plate potential frequency cf, to increase. This relation was quantitated by increasing the osmotic gradient across the motor nerve terminal by adding 25, 50, 7 5 , or 100 mmoVL of sucrose to the bathing solution. Next, the miniature end-plate potnetial frequency was measured every 5 minutes for up to 1 hour cf3 in sucrose. A plot of ln('jJfo), where fo was the resting miniature end-plate potential frequency, versus osmotic gradient, was linear with osmotic gradient (mmol/L) = A l n ~ & o )+ B, where A and B are constants estimated by least squares linear regression. Amino acid and myoinositol transport experiments consisted of (1) measuring the resting miniature end-plate potential frequency, fo, for 10 minutes; then (2) increasing the osmotic pressure of the bathing solution by adding 25, 50, or 75 mmol/L of L-alanine, D-alanine, a(methy1amino)isobutyric acid, or myoinositol to the bathing solution; and (3) measuring the miniature end-plate potential frequency after 20 minutes, f20, in the test amino acid or myoinositol. T o prevent small
762
Insulin Increases Amino Acid Transport
minature end-plate potentials from being lost in the amplifier noise, the data from any cell where the resting potential fell by > 10% were discarded. The expression above, A In(fJfo) + B , was used to estimate the osmotic gradient across the motor nerve terminal membrane at 20 minutes, or osm20. Because a known amount of the amino acid or myoinositol was added to the bathing solution, the difference between the millimoles per liter added to the bathing solution and osm20 was the amount of the test amino acid or myoinositol that had entered the motor nerve terminal. For example, if the miniature end-plate potential frequency ratio, fzo/f0, after 20 minutes in 50 mmol/L of L-alanine was 1.3, then the expected osmotic gradient across the motor nerve terminal membrane would be 17 mmol/L. Because the bathing solution's osmotic pressure does not change, this implies that 33 mmoVL of L-alanine had been transported into the motor nerve terminal over 20 minutes. As miniature end-plate potential frequenc is determined by the presynaptic nerve terminal,Bthese experiments can only detect a change in amino acid or myoinositol concentration across the nerve terminal membrane. The time course of amino acid transport was evaluated by plotting f J f o versus time, where f i was the miniature end-plate potential frequency t minutes after adding L-alanine to the bathing solution, and t varied from about 5 to 60 minutes. Next, curves were fitted to the initial decaying part of thefJfo versus time curve. Log and power curves were fitted to the initial decaying part of the f J f o would approach infinty as time approached zero, and were discarded. Linear and exponential curves differed in their predictions about how the rate of transport would vary as a function of osmotic gradient across the motor nerve terminal membrane. The relationship between fJfo and time for an exponential curve was fJfo = CeDt; and, for a linear curve, fJo = E - Ft, where t is time and C, D, E, and F are constants. The relationship between osmotic gradient and fJfo was osmotic gradient (mmol/L) = A ln(CeDt)+ B or osmotic gradient (mmol/L) = A ln(C) + ADt B and for a linear curve is osmotic gradient (mmol/L) = A ln(E - Ft) + B. Both exponential and linear curves predict the osmotic gradient across the motor nerve terminal membrane falls as time increases. The derivative of osmotic gradient with respect to time, or transport rate, for an exponential curve is a constant, AD mmol/L per minute and for a linear curve is a function of time, -AF/(E - Ft) mmol/L per minute. A linear
+
MUSCLE & NERVE
July 1992
curve predicts that, as time increases and the osmotic gradient across the motor nerve terminal membrane falls, the rate of transport increases. This analysis suggests that a choice between linear and exponential curves could be made by measuring the rate of transport as a function of osmotic gradient across the motor nerve terminal membrane. Least squares linear regression was used to estimate the slope and y-intercept. T h e number of time points in the initial, or transport, phase was determined by including as many values of f J f o as possible without reducing the correlation coefficient between lnqJfo) or f J f o and time, or the accuracy of the predicted amino acid transport rate
..
I-
4
.
"I '
-.i
in comparison with the observed transport rate, which was calculated as described above. The amino acid transport rate was calculated by subtracting the calculated osmotic gradient at 20 minutes, osm20, from the known amount of ~-alanine added to the bathing solution. T h e value of f20/f0 was found by evaluating CeDPominfor an exponential curve or E - F (20min) for a linear curve. Then, f20/f0 was associated with an osmotic gradient across the motor nerve terminal membrane or osm20 as descriobed above. Least squares linear regression was also used to estimate the slope and y-intercept of the relationship between In(fjfo) and osmotic gradient. Most of the data, which consisted of 8 to 11 groups of 4 to 12 cells (see Figs. I-4),were subjected to analysis of variance." If no significant difference was detected, the data were pooled and an overall mean calculated. Usually analysis of variance detected a significant difference, and then subgroups were compared by analysis of variance. Subgroups were chosen by rate of transport, rapid versus slow, as in Figures 2-4; or, os-
Sdeus ( 2 hours1 50 mlnfmathylamino) ISMUlY1IC Acid
2'5
37.5
1 58
T
7b
'T
Osmotic Gradient ImM)
FIGURE 1. The rate of amino acid transport into soleus motor axons as a function of osmotic gradient. The rate of amino acid transport in the 11 groups was not the same (F = 3.8, df = 10 and 46, P < 0.01, analysis of variance). When tested with osmotic gradients of 25 and 37.5 mmol/L, there was no significant difference in the rate of amino acid transport (F = 0.050, df = 4 and 20, P = NS; analysis of variance). However, at both 50 and 75 mmol/L, a signifciant difference was detected (50 mmol/L: f = 5.61, df = 2 and 14, P < 0.05; mmol/L: F = 6.57, df = 2 and 12, P < 0.05; analysis of variance) with i-alanine transport after 10 >mU/mL of insulin being more rapid than either L-alanine without insulin (50 mmollL: t = 2.57, df = 11, P < 0.05; 75 mmol/L: t = 3.90, df = 8, P < 0.01; Student's t test) or D-alanine after 100 kU/mL of insulin (50 rnmol/L: t = 2.34, df = 10, P < 0.05; 75 rnmol/L: t = 3.04, df = 8, P < 0.02; Student's t test). There was no significant difference between the rate of L-alanine transport after 10 wU/mL of insulin at osmotic gradients of 50 and 75 mmol/L (t = 1.8, df = 11, P > 0.05; Student's t test). Also, at osmotic gradients of 25, 50, and 75 mmol/L, there was no significant difference between the rate of L-alanine transport with no insulin and o-alanine transport after 100 pU/mL of insulin (F = 0.58, df = 5 and 23,P = NS; analysis of variance). This analysis shows that alanine transport is sterospecific, most obvious with osmotic gradients of 50 and 75 mmol/L, and increased by a physiological dose of insulin. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
Insulin Increases Amino Acid Transport
. 0
5
10 Insulin IuUIml)
50
100
FIGURE 2. The rate of amino acid transport into soleus motor axons atter 2 hours of insulin treatment. The rate of amino acid transport in the 8 groups was not the same (F = 5.4, dt = 7 and 40, P < 0.01; analysis of variance). The 4 groups with slow amino acid transport, a(methylamin0)-isobutyric acid with 0 and 5 pU/mL of insulin, L-alanine with no insulin, and D-alanine with 100 pU/mL of insulin, were not significantly different (F = 0.16, df = 3 and 21, P = NS; analysis of variance). Likewise, there was not significant difference among the 4 groups with rapid amino acid transport, a(methylamin0)-isobutyric acid with 10, 50, and 100 pU/mL of insulin, and L-alanine with 10 pU/mL of insulin (F = 0.71, df = 3 and 19, P = NS; analysis of variance). The smallest difference between slow and rapid amino acid transport groups, a(methylamin0)-isobutyricacid with 0 or 10 pU/mL of insulin, was significant (t = 4.5, df = 14, P < 0.01; Student's t test). This analysis shows that 10 pU/mL of insulin increases stereospecific amino acid transport into soleus motor axons. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
MUSCLE & NERVE
July 1992
763
-
E
30
0
8
N
.
-.- 20 ;10 0
c n
9 0
E 20 min
1 hour
2 hours
Time in 10 OUlml of Insutln
4 -1c
25
37.5
I I
50
75
Osmolic Gradient ImM)
FIGURE 3. Time course of the effect of 10 p U h L of insulin on amino acid transport into soleus motor axons. The rate of amino acid transport in the 8 groups was not the same (F = 6.2, df = 7 and 37, P < 0.01 ; analysis of variance). The 3 groups with slow amino acid transport, a(methylamin0)-isobutyric acid after 0 and 20 minutes of insulin and L-alanine after 0 minutes, were not significantly different (F = 0.038, df = 2 and 13, P = NS; analysis of variance). Also, there was no significant difference in the 5 groups with rapid amino acid transport, L-alanine after 20 minutes, 1 and 2 hours of 10 pU/mL of insulin and a(methy1amino)isobutyric acid after 1 and 2 hours of 10 pU/mL of insulin ( F = 0.67, df = 4 and 24, P = NS: analysis of variance). The smallest difference between the slow and rapid amino acid transport groups, L-alanine after 0 minutes and a(methylamin0)-isobutyric acid after 1 hour of 10 pUhnL of insulin, was significant (f = 3.1, df = 9, P < 0.02; Student's t test). This analysis shows 10 pU/mL of insulin increases amino acid transport in 1 hour or less. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
motic gradient, as in Figure 1. The rapid and slow transport subgroups were not significantly different by analysis of variance, and the significance of the smallest difference between the subgroups with rapid and slow transport was evaluated with a Student's t test." In Figure 1, the subgroups at larger osmotic gradients were significantly different by analysis of variance, and individual groups were compared using a Student's t test. RESULTS
A total of 259 units were studied. The mean resting potential was 69 mV, miniature end-plate potential amplitude was 1.01 mV, with a range of 0.52 to 1.82 mV, and resting miniature end-plate potential frequency was 2. Us. Rise in MEPP Frequency Versus Osmotic Gradient and Time. Miniature end-plate potential fre-
quency was measured every 5 minutes from 5 to 60 minutes ( f 5 - 6 0 ) after increasing the osmotic gradient across the motor nerve terminal membrane by adding 0 to 100 mmol/L of sucrose to the
764
Insulin Increases Amino Acid Transport
FIGURE 4. The rate of amino acid transport into extensor digitorum longus motor axons as a function of osmotic gradient. The rate of amino acid transport in the 8 groups was not the same ( F = 9.04, df = 7 and 49, P < 0.01; analysis of variance). The 5 groups with slow amino acid transport, 25 and 37.5 mmol/L of L-alanine with no insulin and 25, 50, and 75 mmol/L of D-alanine with 1000 pU/mL of insulin were not significantly different (F = 1.93, df = 4 and 31, P = NS; analysis of variance). Likewise, there was no significant difference among the 3 groups with rapid amino acid transport, 50 and 75 mmol/L of L-alanine with no insulin, and 50 mmol/L of L-alanine with 1000 @/mL of insulin (F = 0.91, df = 2 and 18, P = NS; analysis of variance). The smallest difference between the slow and rapid amino acid transport groups, 37.5 versus 75 rnmol/L L-alanine with no insulin, was significant (t = 2.93, df = 15, P < 0.05; Student's f test). This analysis shows that alanine transport is stereospecific, most obvious at 50 and 75 mmol/L, and unchanged by insulin. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
bathing solution. The rise in miniature end-plate potential frequency at 5 to 60 minutes was divided by resting miniature end-plate potential frequency yb),and (f5-6O)lfO was plotted against the osmotic gradient across the motor nerve terminal membrane. A t each osmotic gradient, 0 to 100 mmol/L, The there was no significant change in (f5-salfo. plot of ln(f5-60/fo) versus osmotic gradient was linear with a correlation coefficient of 0.992, osmotic gradient (mmol/L) = 67.8 ln(f$-60/fo) gradLerlt) . T h ese 0.92 andf5-,,/fo = 1.01e0.015(osmotic results were based on recordings from the soleus and extensor digitorum longus after 2 hours in 1000 pU/mL of insulin, and are similar to those reported for diaphragm without insulin treatment7 and other studies.4x6These results suggest that hyperosmotic neurosecretion can be used to study the time course of amino acid transport. As expected from the relationship between osmotic gradient and the rise in miniature end-plate potential frequency, the addition of 25 to 50 mmoVL of L-ala-
Time Course of Amino Acid Transport.
MUSCLE & NERVE
July 1992
nine to the bathing solution caused an initial rise in miniature end-plate potential frequency. The rise in miniature end-plate potential frequency, as measured by f J f o , was not sustained. T h e decline in MEPP frequency implies that the osmotic gradient across the motor nerve terminal membrane has declined. Because the bathing solution concentration of L-alanine remains constant, the decline in f J f o suggests L-alanine had been transported into the motor nerve terminal, and dissipated the osmotic gradient across the motor nerve terminal membrane. The amount of transport was usually calculated by finding f 2 J f 0 , and using the relationship between fpo/f0 and osmotic gradient to find the osmotic gradient across the motor nerve terminal membrane (osm20). The observed transport rate in mmol/L per 20 minutes was the difference between the amount of L-alanine added to the bathing solution and osm20. Another way to estimate transport was to fit curves to fdfo versus time. Miniature end-plate potential frequency declined for about 5 to 25 minutes and did not decline any further between 25 and 60 minutes. These phases were called transport and equilibrium. Linear and exponential curves fit the f J f o versus time data for the transport phase equally well (linear r = 0.636, exponential r = 0.633, mean correlation coefficient of 9 cells). Also, linear and exponential curves predicted observed transport with similar accuracy (linear r = 0.800, exponential r = 0.812, correlation coefficient between observed and calculated transport rates). Linear and exponential curves differed in their predictions of how the rate of L-alanine transport in mmoVL per minute would vary as a function of the osmotic gradient across the motor nerve terminal membrane (see METHODS). Linear curves predict that the rate of transport should decline as the osmotic gradient across the motor nerve terminal membrane increases, and exponential curves predict the rate of transport will remain constant as the osmotic gradient across the motor nerve terminal membrane increases. The transport rate did not decline as the osmotic gradient was increased (see next 2 sections) and, for this reason, exponential curves were chosen. T h e mean rate of transport into the motor nerve terminal as estimated by exponential curves was 1.03 mmoVL per minute (9 cells).
ure 1. As can be seen, L-alanine is not transported into soleus motor axons until insulin is added to the bathing solution. Insulin, 10 pU/mL, increases L-alanine transport into soleus motor axons at osmotic gradients of 50 and 75 mmol/L, but not at 25 or 37.5 mmol/L. A larger dose of insulin, 100 kU/mL, does not change the rate of D-alanine transport into soleus motor axons. This result implies amino acid transport into soleus motor axons is stereospecific, and suggests insulin increases the rate of amino acid transport. An insulin dose-response experiment is shown in Figure 2. A low dose of insulin, 10 kU/mL but not 5 pU/mL, increases the rate of transport of both L-alanine and a(methy1amino)-isobutyric acid into soleus motor axons. The time course of the effect of 10 kU/mL of insulin on amino acid transport is shown in Figure 3. Insulin increases the rate of L-alanine transport after 20 minutes and a(methy1amino)-isobutyric acid transport after 60 minutes. These experiments suggest insulin regulates amino acid transport into soleus motor axons. Amino Acid Transport into Extensor Digitorum Longus Motor Axons. The effect of osmotic gradient
on the rate of alanine transport into motor axons in the extensor digitorum longus is shown in Figure 4.At 50 and 75 mmoVL, but not 25 or 37.5 mmoVL, L-alanine rapidly entered motor axons in the extensor digitorum longus. Like the soleus, alanine transport into motor axons in the extensor digitorum longus was stereospecific with L-alanine, but not D-alanine, entering motor axons. Large doses of insulin, 100 or 1000 kU/mL, do not increase the rate that L-alanine or a(methy1amino)isobutyric acid are transported into motor axons in the in the extensor digitorum longus. In an attempt to evaluate the sodium dependence of amino acid transport, sodium in the bathing solution was replaced with choline. In a bathing solution containing sodium, the mean miniature endplate potential amplitude was 0.92 mV in 10 cells in both the extensor digitorum longus and soleus. Two minutes after replacing sodium with choline, no miniature end-plate potentials could be seen. While the sodium dependence of amino acid transport could not be evaluated, it is clear that amino acid transport into motor axons in the extensor digitorum longus is rapid and not increased by large doses of insulin.
Amino Acid Transport Into Soleus Motor Axons.
The effect of osmotic gradient on amino acid transport into soleus motor axons is shown in Fig-
Insulin Increases Amino Acid Transport
Myoinositol Transport into Motor Axons. There was no evidence of myoinositol transport into
MUSCLE 8. NERVE
July 1992
765
motor axons in the soleus or extensor digitorum longus. Neither myoinositol concentration (25, 50, or 75 mmol/L) or insulin (100 or 1000 pU/ mL), influence the rate of myoinositol transport. Indeed myoinositol appears to be an osmotic agent with a mean transport rate of 0.85 mmol/L per 20 minutes. The rate of myoinositol transport is the same as that of sucrose, -0.78 mmol/L per 20 minutes, and D-alanine, 1.2 mmol/L per 20 minutes (F = 0.66, df = 2 and 108, P-NS, analysis of variance). These experiments suggest that either myoinositol is not transported into motor axons, or it is transported at a rate that is too slow to be detected. DISCUSSION
Shimoni and Kahaminoff13 have developed an indirect method to estimate the rate of glucose transport into single frog motor nerve terminals. In this study, their method was used to estimate the rate of amino acid and myoinositol transport into rat motor axons. This method has a number of potential and real limitations. First, miniature end-plate potential frequency varies greatly from minute to minute. This large variability could conceal small amounts of transport, and might explain why no myoinositol transport was detected. The equations suggest another limitation. The transport rate, as estimated by the derivative (see METHDS for details), does not increase as amino acid concentration was increased from 25 to 50 mmol/L. This suggests that at these substrate concentrations transport is at its maximal velocity and thus other parameters of transport can not be estimated. Another real limitation of this method is sodium-dependence of amino acid or myoinositol transport could not be tested. When sodium was replaced by choline in the bathing solution, the MEPPs vanished. A possible limitation of this method is that insulin might change the relationship between osmotic gradient across the motor nerve terminal membrane and the rise in MEPP frequency as measured by lnf5-solfo0). Fortunately, there is no discernible difference between the results with 1000 pU/mL of insulin and previously published curves without insulin.’ Metabolism of a substrate after it was transported into a motor nerve terminal might falsely reduce estimates of substrate transport. Happily, no significant difference was detected between the rate of transport of metabolizable and nonmetabolizable substrates, such as 3-O-methyl-~-glucose versus D-glUCoSe’’g and ol(methy1amino)-isobutyric acid versus 1.-alanine (Figs. 2 and 3 ) . Thus, metabolism
766
lnsutin Increases Amino Acid Transport
of a substrate after transport does not appear to be a significant problem. Despite these limitations, hyperosmotic neurosecretion is the only way to estimate transport into single motor axon terminals. Amino acid and glucose transport into rat motor nerve terminals are very similar. Both glucoseg and A-system amino acid transport are (1) rapid into extensor digitorum longus motor nerve terminals with or without insulin (Fig. 4);(2) slow into soleus motor nerve terminals without insulin (Fig. 1); and (3) rapid into soleus motor nerve terminals after a low-physiological dose of insulin, 10 pU/mL, for 1 hour (Figs. 2 and 3). These differences in transport into motor axons innervating the extensor digitorum longus, a fast-twitch muscle, and soleus, a slow-twitch muscle, are part of the metabolic specializations of motor units. Motor axons and muscle fibers in the same motor unit are similar in their oxidative enzyme a~tivity’.’~ and in the sensitivity of their amino acid and glucose transport to i n ~ u l i n , ~ ~ but ’ , ’ ~they differ in their basal amino acid and glucose transport.5,’,10,11,15 Difference in basal transport might reflect different energy sources, with muscle using lipid,* and nerve using glucose.12 Except for basal transport, axons and muscle fibers in the same motor unit are similar in their metabolic specializations. Metabolic specialization of motor axons implies motor axons should be selectively involved by disease processes.
REFERENCES
1. Burke RE, Rudorriin P. Spinal neurons and synapses, in Handbook of Physiology, The Nervous System. Neurophysiology. Cellular Biology of Neurons. Bethesda, MD, American Physiological Society, 1977, vol I , pp 885-886. 2. Felig P, Wahren J : Fuel homostasis in exercise. N Engl J Med 1975;293:1078- 1084. 3. Guidotti GG, Borghetti AF, Gazzola GC: The regulation of amino acid transport in animal cells. Bzochim Bzophys Actu 1978;515:329-366. 4. Hubbard JI, Jones SF, Landau EM: An examination of the effects of osmotic pressure changes upon transmitter release from mammalian motor nerve terminals. J Physiol Lond 1968; 197:639-657. 5. James DE, Jenkins AB, Kraegen EW: Heterogeneity of insulin action in individual muscles in vivo: Euglycemic clamp studies in rats. Am J Physiol 1985;248:E567-E574. 6. Martin AR: Junctional transmission 11. Presynaptic mechanisms, in Handbook of Physiology, The Newow. System. Neurophysiology. Cellular Bzology of Neurons. Bethesda, MD, American Physiological Society, 1977, vol 1, pp 329-355. 7. Pickett JB: Nerve terminals are as metabolically different as the muscle fibers they innervate. Science Wash DC 1980;2 10:927- 928. 8. Pickett JB: Insulin increase amino acid transport into soleus motor axons. Neurology 1988;38(suppl 1):223. 9. Pickett JB: Insulin rat soleus axon function by increasing glucose transport. A m J PhyJiol 1988;255:660-69.
MUSCLE & NERVE
July 1992
10. Rosenheimer JL, Smith DO: Electrophysiological measurements of glucose uptake into motor nerve terminals of mature and aged rats. Brain Res 1985;330:373-377. 11. Rosenheimer JL: Electrophysiological examination of glucose transport into motor nerve terminals in mature and aged rats, in Rahaminoff R, Katz B (eds): Calcium, Neuronal Function and Transmitter Release. Boston, Nijhoff, 1986, p p 575-582. 12. Schwartz WJ, Smith CB, Davidsen L, Savaki H, Sokoloff L, Mata M, Fink DJ, Gainer H: Metabolic mapping of functional activity in the hypothalmo-neurohypophysical system of the rat. Science Wash DC 1979;205:723-725. 13. Shimoni Y, Rahamimoff R: Stereospecific glucose transport across motor nerve terminal membrane: An electrophysiological study. A m J Physzol 1983;245:C308-C3 15.
Insulin Increases Amino Acid Transport
14. Sickles DW, Oblak TG: Metabolic variation among cx-mo-
15. 16. 7.
8.
toneurons innervating differnt muscle-fiber types. I. Oxidative enzyme activity. J Neurophysiol 1984;51:529-537. Turinsky J: Glucose and amino acid uptake by exercising muscles in vivo: Effect of insulin, fiber population, and denervation. Endocrinology 1987; 12 1:528-535. Wheeler DD: Amino acid transport in peripheral nerve: Specificity of uptake. J Neurochem 1975;24:97- 104. Wheeler DD: Amino acid transport in peripheral nerve: Energy and sodium dependence. J Neurochem 1975;24: 1197- 1202. Zar JH: Biostatistical Analysis. Englewood Cliffs, NJ, Prentice-Hall, 1984.
MUSCLE & NERVE
July 1992
767
INSULIN INCREASES AMINO ACID TRANSPORT INTO RAT SOLEUS JACKSON 6. PICKETT, MD
A motor unit consists of an anterior horn cell and the muscle fibers it innervates. Motor units have been classified according to the speed of contraction of their muscle fibers, rapid for type F and slow for type S. Type F and S muscle fibers differ in their histochemical, biochemical, and physiological properties.' Sickles and OblakI4 discovered that the concentration of oxidative enzymes in anterior horn cells paralleled that in their muscle fibers. Also, glucose transport, as measured by an indirect method developed by Shimoni and Rahamimoff,13 is rapid into motor axons innervating a fast-twitch muscle, the rat extensor digitorum longus, and slow into those innervating a slow-twitch
From the Veterans Affairs Medical Center and Department of Neurology, Medical University of South Carolina, Charleston, South Carolina. Acknowledgments. Research supported by Veterans Administration and the Juvenile Diabetes Foundation. Thanks to Mrs. Linda M. Wade for preparation of the manuscript, and Drs. Henry F Martin and Darrell D. Wheeler for review of the manuscript. Address reprint requests to Jackson B. Pickett, MD Veterans Affairs Medical Center (127). 109 Bee Street, Charleston, SC 29401-5799. Accepted for publication September 1 , 1991 CCC 0148-639X/92/070761-07$04.00 0 1992 John Wiley & Sons, Inc.
Insulin Increases Amino Acid Transport
muscle, the rat so leu^.^^'^" Shimoni and Rahamimoff13 were the first to show insulin increases glucose transport into frog motor axons. 'These findings have been extended by a recent reportg suggesting that insulin improves rat soleus motor axon function by increasing glucose transport, and has no effect on extensor digitorum longus axon function or glucose transport. These studies suggest anterior horn cells and their axons are as metabolically specialized as their muscle fibers. Because insulin regulates glucose transport into soleus motor axons, but not extensor digitorum longus motor axons, insulin might also increase amino acid and myoinositol transport into soleus motor axons, but not extensor digitorum longus motor axons. Of the many amino acid transport systems, insulin regulates only system A, which transports small neutral amino acids such as alanine, glycine, proline, and ~ e r i n e . ~ has reported that a small neutral amino acid transport system exists in frog sciatic nerve. This report investigated basal and insulin-stimulated amino acid and myoinositol transport into motor axons innervating rat soleus and extensor digitorum longus muscles. An abstract of this work has been published.'
MUSCLE & NERVE
July 1992
761
METHODS
Nerve-muscle preparations were dissected from male Sprague-Dawley rats (60 to 230 g) and pinned in a small transilluminated recording chamber, which was kept at 31.5 to 32.5" C. The bathing solution was continuously exchanged between the recording chamber and a 200-mL reservoir at a rate of 10 mL/min, so that solution changes were complete within 2 minutes, as measured by dye washout. Solution changes were accomplished by changing the reservoir from which the bathing solution was drawn. The bathing solution contained 137 mmolfL NaCl, 5 mmol/L KC1, 2 mmol/L CaCl,; 1 mmol/L NaH,PO,, 1 mmol/L MgCl,, 24 mmol/L NaHC03, and 11 mmol/L D-glucose; after bubbling with 95% 0, and 5% CO,; the pH was 7.3 to 7.4. Purified pork regular insulin (Iletin 11) was obtained from Eli Lilly and Co.; a(methy1amino)-isobutyric acid, myoinositol, and alanine from the Sigma Chemical Co. Miniature end-plate potentials were detected with 5 to 20-Mfl microelectrodes and conventional methods for intracellular recording. After suitable amplification, the resting potential and miniature end-plate potentials were recorded on film. Recordings were limited to cells where: (1) the miniature end-plate potential rose to a peak in less than 0.5 ms; and (2) the miniature end-plate potential amplitude was well above the amplifier noise (usually 200 pV peak to peak). Increasing the osmotic gradient across the motor nerve terminal membrane caused the miniature end-plate potential frequency cf, to increase. This relation was quantitated by increasing the osmotic gradient across the motor nerve terminal by adding 25, 50, 7 5 , or 100 mmoVL of sucrose to the bathing solution. Next, the miniature end-plate potnetial frequency was measured every 5 minutes for up to 1 hour cf3 in sucrose. A plot of ln('jJfo), where fo was the resting miniature end-plate potential frequency, versus osmotic gradient, was linear with osmotic gradient (mmol/L) = A l n ~ & o )+ B, where A and B are constants estimated by least squares linear regression. Amino acid and myoinositol transport experiments consisted of (1) measuring the resting miniature end-plate potential frequency, fo, for 10 minutes; then (2) increasing the osmotic pressure of the bathing solution by adding 25, 50, or 75 mmol/L of L-alanine, D-alanine, a(methy1amino)isobutyric acid, or myoinositol to the bathing solution; and (3) measuring the miniature end-plate potential frequency after 20 minutes, f20, in the test amino acid or myoinositol. T o prevent small
762
Insulin Increases Amino Acid Transport
minature end-plate potentials from being lost in the amplifier noise, the data from any cell where the resting potential fell by > 10% were discarded. The expression above, A In(fJfo) + B , was used to estimate the osmotic gradient across the motor nerve terminal membrane at 20 minutes, or osm20. Because a known amount of the amino acid or myoinositol was added to the bathing solution, the difference between the millimoles per liter added to the bathing solution and osm20 was the amount of the test amino acid or myoinositol that had entered the motor nerve terminal. For example, if the miniature end-plate potential frequency ratio, fzo/f0, after 20 minutes in 50 mmol/L of L-alanine was 1.3, then the expected osmotic gradient across the motor nerve terminal membrane would be 17 mmol/L. Because the bathing solution's osmotic pressure does not change, this implies that 33 mmoVL of L-alanine had been transported into the motor nerve terminal over 20 minutes. As miniature end-plate potential frequenc is determined by the presynaptic nerve terminal,Bthese experiments can only detect a change in amino acid or myoinositol concentration across the nerve terminal membrane. The time course of amino acid transport was evaluated by plotting f J f o versus time, where f i was the miniature end-plate potential frequency t minutes after adding L-alanine to the bathing solution, and t varied from about 5 to 60 minutes. Next, curves were fitted to the initial decaying part of thefJfo versus time curve. Log and power curves were fitted to the initial decaying part of the f J f o would approach infinty as time approached zero, and were discarded. Linear and exponential curves differed in their predictions about how the rate of transport would vary as a function of osmotic gradient across the motor nerve terminal membrane. The relationship between fJfo and time for an exponential curve was fJfo = CeDt; and, for a linear curve, fJo = E - Ft, where t is time and C, D, E, and F are constants. The relationship between osmotic gradient and fJfo was osmotic gradient (mmol/L) = A ln(CeDt)+ B or osmotic gradient (mmol/L) = A ln(C) + ADt B and for a linear curve is osmotic gradient (mmol/L) = A ln(E - Ft) + B. Both exponential and linear curves predict the osmotic gradient across the motor nerve terminal membrane falls as time increases. The derivative of osmotic gradient with respect to time, or transport rate, for an exponential curve is a constant, AD mmol/L per minute and for a linear curve is a function of time, -AF/(E - Ft) mmol/L per minute. A linear
+
MUSCLE & NERVE
July 1992
curve predicts that, as time increases and the osmotic gradient across the motor nerve terminal membrane falls, the rate of transport increases. This analysis suggests that a choice between linear and exponential curves could be made by measuring the rate of transport as a function of osmotic gradient across the motor nerve terminal membrane. Least squares linear regression was used to estimate the slope and y-intercept. T h e number of time points in the initial, or transport, phase was determined by including as many values of f J f o as possible without reducing the correlation coefficient between lnqJfo) or f J f o and time, or the accuracy of the predicted amino acid transport rate
..
I-
4
.
"I '
-.i
in comparison with the observed transport rate, which was calculated as described above. The amino acid transport rate was calculated by subtracting the calculated osmotic gradient at 20 minutes, osm20, from the known amount of ~-alanine added to the bathing solution. T h e value of f20/f0 was found by evaluating CeDPominfor an exponential curve or E - F (20min) for a linear curve. Then, f20/f0 was associated with an osmotic gradient across the motor nerve terminal membrane or osm20 as descriobed above. Least squares linear regression was also used to estimate the slope and y-intercept of the relationship between In(fjfo) and osmotic gradient. Most of the data, which consisted of 8 to 11 groups of 4 to 12 cells (see Figs. I-4),were subjected to analysis of variance." If no significant difference was detected, the data were pooled and an overall mean calculated. Usually analysis of variance detected a significant difference, and then subgroups were compared by analysis of variance. Subgroups were chosen by rate of transport, rapid versus slow, as in Figures 2-4; or, os-
Sdeus ( 2 hours1 50 mlnfmathylamino) ISMUlY1IC Acid
2'5
37.5
1 58
T
7b
'T
Osmotic Gradient ImM)
FIGURE 1. The rate of amino acid transport into soleus motor axons as a function of osmotic gradient. The rate of amino acid transport in the 11 groups was not the same (F = 3.8, df = 10 and 46, P < 0.01, analysis of variance). When tested with osmotic gradients of 25 and 37.5 mmol/L, there was no significant difference in the rate of amino acid transport (F = 0.050, df = 4 and 20, P = NS; analysis of variance). However, at both 50 and 75 mmol/L, a signifciant difference was detected (50 mmol/L: f = 5.61, df = 2 and 14, P < 0.05; mmol/L: F = 6.57, df = 2 and 12, P < 0.05; analysis of variance) with i-alanine transport after 10 >mU/mL of insulin being more rapid than either L-alanine without insulin (50 mmollL: t = 2.57, df = 11, P < 0.05; 75 mmol/L: t = 3.90, df = 8, P < 0.01; Student's t test) or D-alanine after 100 kU/mL of insulin (50 rnmol/L: t = 2.34, df = 10, P < 0.05; 75 rnmol/L: t = 3.04, df = 8, P < 0.02; Student's t test). There was no significant difference between the rate of L-alanine transport after 10 wU/mL of insulin at osmotic gradients of 50 and 75 mmol/L (t = 1.8, df = 11, P > 0.05; Student's t test). Also, at osmotic gradients of 25, 50, and 75 mmol/L, there was no significant difference between the rate of L-alanine transport with no insulin and o-alanine transport after 100 pU/mL of insulin (F = 0.58, df = 5 and 23,P = NS; analysis of variance). This analysis shows that alanine transport is sterospecific, most obvious with osmotic gradients of 50 and 75 mmol/L, and increased by a physiological dose of insulin. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
Insulin Increases Amino Acid Transport
. 0
5
10 Insulin IuUIml)
50
100
FIGURE 2. The rate of amino acid transport into soleus motor axons atter 2 hours of insulin treatment. The rate of amino acid transport in the 8 groups was not the same (F = 5.4, dt = 7 and 40, P < 0.01; analysis of variance). The 4 groups with slow amino acid transport, a(methylamin0)-isobutyric acid with 0 and 5 pU/mL of insulin, L-alanine with no insulin, and D-alanine with 100 pU/mL of insulin, were not significantly different (F = 0.16, df = 3 and 21, P = NS; analysis of variance). Likewise, there was not significant difference among the 4 groups with rapid amino acid transport, a(methylamin0)-isobutyric acid with 10, 50, and 100 pU/mL of insulin, and L-alanine with 10 pU/mL of insulin (F = 0.71, df = 3 and 19, P = NS; analysis of variance). The smallest difference between slow and rapid amino acid transport groups, a(methylamin0)-isobutyricacid with 0 or 10 pU/mL of insulin, was significant (t = 4.5, df = 14, P < 0.01; Student's t test). This analysis shows that 10 pU/mL of insulin increases stereospecific amino acid transport into soleus motor axons. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
MUSCLE & NERVE
July 1992
763
-
E
30
0
8
N
.
-.- 20 ;10 0
c n
9 0
E 20 min
1 hour
2 hours
Time in 10 OUlml of Insutln
4 -1c
25
37.5
I I
50
75
Osmolic Gradient ImM)
FIGURE 3. Time course of the effect of 10 p U h L of insulin on amino acid transport into soleus motor axons. The rate of amino acid transport in the 8 groups was not the same (F = 6.2, df = 7 and 37, P < 0.01 ; analysis of variance). The 3 groups with slow amino acid transport, a(methylamin0)-isobutyric acid after 0 and 20 minutes of insulin and L-alanine after 0 minutes, were not significantly different (F = 0.038, df = 2 and 13, P = NS; analysis of variance). Also, there was no significant difference in the 5 groups with rapid amino acid transport, L-alanine after 20 minutes, 1 and 2 hours of 10 pU/mL of insulin and a(methy1amino)isobutyric acid after 1 and 2 hours of 10 pU/mL of insulin ( F = 0.67, df = 4 and 24, P = NS: analysis of variance). The smallest difference between the slow and rapid amino acid transport groups, L-alanine after 0 minutes and a(methylamin0)-isobutyric acid after 1 hour of 10 pUhnL of insulin, was significant (f = 3.1, df = 9, P < 0.02; Student's t test). This analysis shows 10 pU/mL of insulin increases amino acid transport in 1 hour or less. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
motic gradient, as in Figure 1. The rapid and slow transport subgroups were not significantly different by analysis of variance, and the significance of the smallest difference between the subgroups with rapid and slow transport was evaluated with a Student's t test." In Figure 1, the subgroups at larger osmotic gradients were significantly different by analysis of variance, and individual groups were compared using a Student's t test. RESULTS
A total of 259 units were studied. The mean resting potential was 69 mV, miniature end-plate potential amplitude was 1.01 mV, with a range of 0.52 to 1.82 mV, and resting miniature end-plate potential frequency was 2. Us. Rise in MEPP Frequency Versus Osmotic Gradient and Time. Miniature end-plate potential fre-
quency was measured every 5 minutes from 5 to 60 minutes ( f 5 - 6 0 ) after increasing the osmotic gradient across the motor nerve terminal membrane by adding 0 to 100 mmol/L of sucrose to the
764
Insulin Increases Amino Acid Transport
FIGURE 4. The rate of amino acid transport into extensor digitorum longus motor axons as a function of osmotic gradient. The rate of amino acid transport in the 8 groups was not the same ( F = 9.04, df = 7 and 49, P < 0.01; analysis of variance). The 5 groups with slow amino acid transport, 25 and 37.5 mmol/L of L-alanine with no insulin and 25, 50, and 75 mmol/L of D-alanine with 1000 pU/mL of insulin were not significantly different (F = 1.93, df = 4 and 31, P = NS; analysis of variance). Likewise, there was no significant difference among the 3 groups with rapid amino acid transport, 50 and 75 mmol/L of L-alanine with no insulin, and 50 mmol/L of L-alanine with 1000 @/mL of insulin (F = 0.91, df = 2 and 18, P = NS; analysis of variance). The smallest difference between the slow and rapid amino acid transport groups, 37.5 versus 75 rnmol/L L-alanine with no insulin, was significant (t = 2.93, df = 15, P < 0.05; Student's f test). This analysis shows that alanine transport is stereospecific, most obvious at 50 and 75 mmol/L, and unchanged by insulin. Height of bars is the mean, the T is 1 SE of the mean, and individual data points are shown to the left of the bar.
bathing solution. The rise in miniature end-plate potential frequency at 5 to 60 minutes was divided by resting miniature end-plate potential frequency yb),and (f5-6O)lfO was plotted against the osmotic gradient across the motor nerve terminal membrane. A t each osmotic gradient, 0 to 100 mmol/L, The there was no significant change in (f5-salfo. plot of ln(f5-60/fo) versus osmotic gradient was linear with a correlation coefficient of 0.992, osmotic gradient (mmol/L) = 67.8 ln(f$-60/fo) gradLerlt) . T h ese 0.92 andf5-,,/fo = 1.01e0.015(osmotic results were based on recordings from the soleus and extensor digitorum longus after 2 hours in 1000 pU/mL of insulin, and are similar to those reported for diaphragm without insulin treatment7 and other studies.4x6These results suggest that hyperosmotic neurosecretion can be used to study the time course of amino acid transport. As expected from the relationship between osmotic gradient and the rise in miniature end-plate potential frequency, the addition of 25 to 50 mmoVL of L-ala-
Time Course of Amino Acid Transport.
MUSCLE & NERVE
July 1992
nine to the bathing solution caused an initial rise in miniature end-plate potential frequency. The rise in miniature end-plate potential frequency, as measured by f J f o , was not sustained. T h e decline in MEPP frequency implies that the osmotic gradient across the motor nerve terminal membrane has declined. Because the bathing solution concentration of L-alanine remains constant, the decline in f J f o suggests L-alanine had been transported into the motor nerve terminal, and dissipated the osmotic gradient across the motor nerve terminal membrane. The amount of transport was usually calculated by finding f 2 J f 0 , and using the relationship between fpo/f0 and osmotic gradient to find the osmotic gradient across the motor nerve terminal membrane (osm20). The observed transport rate in mmol/L per 20 minutes was the difference between the amount of L-alanine added to the bathing solution and osm20. Another way to estimate transport was to fit curves to fdfo versus time. Miniature end-plate potential frequency declined for about 5 to 25 minutes and did not decline any further between 25 and 60 minutes. These phases were called transport and equilibrium. Linear and exponential curves fit the f J f o versus time data for the transport phase equally well (linear r = 0.636, exponential r = 0.633, mean correlation coefficient of 9 cells). Also, linear and exponential curves predicted observed transport with similar accuracy (linear r = 0.800, exponential r = 0.812, correlation coefficient between observed and calculated transport rates). Linear and exponential curves differed in their predictions of how the rate of L-alanine transport in mmoVL per minute would vary as a function of the osmotic gradient across the motor nerve terminal membrane (see METHODS). Linear curves predict that the rate of transport should decline as the osmotic gradient across the motor nerve terminal membrane increases, and exponential curves predict the rate of transport will remain constant as the osmotic gradient across the motor nerve terminal membrane increases. The transport rate did not decline as the osmotic gradient was increased (see next 2 sections) and, for this reason, exponential curves were chosen. T h e mean rate of transport into the motor nerve terminal as estimated by exponential curves was 1.03 mmoVL per minute (9 cells).
ure 1. As can be seen, L-alanine is not transported into soleus motor axons until insulin is added to the bathing solution. Insulin, 10 pU/mL, increases L-alanine transport into soleus motor axons at osmotic gradients of 50 and 75 mmol/L, but not at 25 or 37.5 mmol/L. A larger dose of insulin, 100 kU/mL, does not change the rate of D-alanine transport into soleus motor axons. This result implies amino acid transport into soleus motor axons is stereospecific, and suggests insulin increases the rate of amino acid transport. An insulin dose-response experiment is shown in Figure 2. A low dose of insulin, 10 kU/mL but not 5 pU/mL, increases the rate of transport of both L-alanine and a(methy1amino)-isobutyric acid into soleus motor axons. The time course of the effect of 10 kU/mL of insulin on amino acid transport is shown in Figure 3. Insulin increases the rate of L-alanine transport after 20 minutes and a(methy1amino)-isobutyric acid transport after 60 minutes. These experiments suggest insulin regulates amino acid transport into soleus motor axons. Amino Acid Transport into Extensor Digitorum Longus Motor Axons. The effect of osmotic gradient
on the rate of alanine transport into motor axons in the extensor digitorum longus is shown in Figure 4.At 50 and 75 mmoVL, but not 25 or 37.5 mmoVL, L-alanine rapidly entered motor axons in the extensor digitorum longus. Like the soleus, alanine transport into motor axons in the extensor digitorum longus was stereospecific with L-alanine, but not D-alanine, entering motor axons. Large doses of insulin, 100 or 1000 kU/mL, do not increase the rate that L-alanine or a(methy1amino)isobutyric acid are transported into motor axons in the in the extensor digitorum longus. In an attempt to evaluate the sodium dependence of amino acid transport, sodium in the bathing solution was replaced with choline. In a bathing solution containing sodium, the mean miniature endplate potential amplitude was 0.92 mV in 10 cells in both the extensor digitorum longus and soleus. Two minutes after replacing sodium with choline, no miniature end-plate potentials could be seen. While the sodium dependence of amino acid transport could not be evaluated, it is clear that amino acid transport into motor axons in the extensor digitorum longus is rapid and not increased by large doses of insulin.
Amino Acid Transport Into Soleus Motor Axons.
The effect of osmotic gradient on amino acid transport into soleus motor axons is shown in Fig-
Insulin Increases Amino Acid Transport
Myoinositol Transport into Motor Axons. There was no evidence of myoinositol transport into
MUSCLE 8. NERVE
July 1992
765
motor axons in the soleus or extensor digitorum longus. Neither myoinositol concentration (25, 50, or 75 mmol/L) or insulin (100 or 1000 pU/ mL), influence the rate of myoinositol transport. Indeed myoinositol appears to be an osmotic agent with a mean transport rate of 0.85 mmol/L per 20 minutes. The rate of myoinositol transport is the same as that of sucrose, -0.78 mmol/L per 20 minutes, and D-alanine, 1.2 mmol/L per 20 minutes (F = 0.66, df = 2 and 108, P-NS, analysis of variance). These experiments suggest that either myoinositol is not transported into motor axons, or it is transported at a rate that is too slow to be detected. DISCUSSION
Shimoni and Kahaminoff13 have developed an indirect method to estimate the rate of glucose transport into single frog motor nerve terminals. In this study, their method was used to estimate the rate of amino acid and myoinositol transport into rat motor axons. This method has a number of potential and real limitations. First, miniature end-plate potential frequency varies greatly from minute to minute. This large variability could conceal small amounts of transport, and might explain why no myoinositol transport was detected. The equations suggest another limitation. The transport rate, as estimated by the derivative (see METHDS for details), does not increase as amino acid concentration was increased from 25 to 50 mmol/L. This suggests that at these substrate concentrations transport is at its maximal velocity and thus other parameters of transport can not be estimated. Another real limitation of this method is sodium-dependence of amino acid or myoinositol transport could not be tested. When sodium was replaced by choline in the bathing solution, the MEPPs vanished. A possible limitation of this method is that insulin might change the relationship between osmotic gradient across the motor nerve terminal membrane and the rise in MEPP frequency as measured by lnf5-solfo0). Fortunately, there is no discernible difference between the results with 1000 pU/mL of insulin and previously published curves without insulin.’ Metabolism of a substrate after it was transported into a motor nerve terminal might falsely reduce estimates of substrate transport. Happily, no significant difference was detected between the rate of transport of metabolizable and nonmetabolizable substrates, such as 3-O-methyl-~-glucose versus D-glUCoSe’’g and ol(methy1amino)-isobutyric acid versus 1.-alanine (Figs. 2 and 3 ) . Thus, metabolism
766
lnsutin Increases Amino Acid Transport
of a substrate after transport does not appear to be a significant problem. Despite these limitations, hyperosmotic neurosecretion is the only way to estimate transport into single motor axon terminals. Amino acid and glucose transport into rat motor nerve terminals are very similar. Both glucoseg and A-system amino acid transport are (1) rapid into extensor digitorum longus motor nerve terminals with or without insulin (Fig. 4);(2) slow into soleus motor nerve terminals without insulin (Fig. 1); and (3) rapid into soleus motor nerve terminals after a low-physiological dose of insulin, 10 pU/mL, for 1 hour (Figs. 2 and 3). These differences in transport into motor axons innervating the extensor digitorum longus, a fast-twitch muscle, and soleus, a slow-twitch muscle, are part of the metabolic specializations of motor units. Motor axons and muscle fibers in the same motor unit are similar in their oxidative enzyme a~tivity’.’~ and in the sensitivity of their amino acid and glucose transport to i n ~ u l i n , ~ ~ but ’ , ’ ~they differ in their basal amino acid and glucose transport.5,’,10,11,15 Difference in basal transport might reflect different energy sources, with muscle using lipid,* and nerve using glucose.12 Except for basal transport, axons and muscle fibers in the same motor unit are similar in their metabolic specializations. Metabolic specialization of motor axons implies motor axons should be selectively involved by disease processes.
REFERENCES
1. Burke RE, Rudorriin P. Spinal neurons and synapses, in Handbook of Physiology, The Nervous System. Neurophysiology. Cellular Biology of Neurons. Bethesda, MD, American Physiological Society, 1977, vol I , pp 885-886. 2. Felig P, Wahren J : Fuel homostasis in exercise. N Engl J Med 1975;293:1078- 1084. 3. Guidotti GG, Borghetti AF, Gazzola GC: The regulation of amino acid transport in animal cells. Bzochim Bzophys Actu 1978;515:329-366. 4. Hubbard JI, Jones SF, Landau EM: An examination of the effects of osmotic pressure changes upon transmitter release from mammalian motor nerve terminals. J Physiol Lond 1968; 197:639-657. 5. James DE, Jenkins AB, Kraegen EW: Heterogeneity of insulin action in individual muscles in vivo: Euglycemic clamp studies in rats. Am J Physiol 1985;248:E567-E574. 6. Martin AR: Junctional transmission 11. Presynaptic mechanisms, in Handbook of Physiology, The Newow. System. Neurophysiology. Cellular Bzology of Neurons. Bethesda, MD, American Physiological Society, 1977, vol 1, pp 329-355. 7. Pickett JB: Nerve terminals are as metabolically different as the muscle fibers they innervate. Science Wash DC 1980;2 10:927- 928. 8. Pickett JB: Insulin increase amino acid transport into soleus motor axons. Neurology 1988;38(suppl 1):223. 9. Pickett JB: Insulin rat soleus axon function by increasing glucose transport. A m J PhyJiol 1988;255:660-69.
MUSCLE & NERVE
July 1992
10. Rosenheimer JL, Smith DO: Electrophysiological measurements of glucose uptake into motor nerve terminals of mature and aged rats. Brain Res 1985;330:373-377. 11. Rosenheimer JL: Electrophysiological examination of glucose transport into motor nerve terminals in mature and aged rats, in Rahaminoff R, Katz B (eds): Calcium, Neuronal Function and Transmitter Release. Boston, Nijhoff, 1986, p p 575-582. 12. Schwartz WJ, Smith CB, Davidsen L, Savaki H, Sokoloff L, Mata M, Fink DJ, Gainer H: Metabolic mapping of functional activity in the hypothalmo-neurohypophysical system of the rat. Science Wash DC 1979;205:723-725. 13. Shimoni Y, Rahamimoff R: Stereospecific glucose transport across motor nerve terminal membrane: An electrophysiological study. A m J Physzol 1983;245:C308-C3 15.
Insulin Increases Amino Acid Transport
14. Sickles DW, Oblak TG: Metabolic variation among cx-mo-
15. 16. 7.
8.
toneurons innervating differnt muscle-fiber types. I. Oxidative enzyme activity. J Neurophysiol 1984;51:529-537. Turinsky J: Glucose and amino acid uptake by exercising muscles in vivo: Effect of insulin, fiber population, and denervation. Endocrinology 1987; 12 1:528-535. Wheeler DD: Amino acid transport in peripheral nerve: Specificity of uptake. J Neurochem 1975;24:97- 104. Wheeler DD: Amino acid transport in peripheral nerve: Energy and sodium dependence. J Neurochem 1975;24: 1197- 1202. Zar JH: Biostatistical Analysis. Englewood Cliffs, NJ, Prentice-Hall, 1984.
MUSCLE & NERVE
July 1992
767
