| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
EXERCISE AND RESPIRATORY PHYSIOLOGY
Department of Nutrition and Exercise Sciences, Oregon State University, Corvallis, Oregon
Submitted 22 January 2008 ; accepted in final form 26 August 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
muscle atrophy; rehabilitation; hindlimb suspension; eccentric contractions; muscle damage
Effective strategies for rehabilitating muscles atrophied by prolonged bed rest, spaceflight, or other types of nonweight bearing will require knowledge of this injury process and its impact on physical performance. The ability of a force-producing muscle to shorten and produce power is arguably one of the most critical aspects of skeletal muscle performance. Thus, the primary goal of this study was to assess changes in the force-velocity-power relationships of soleus muscles during recovery from HU. We hypothesized that recovering muscles would show reductions in force, shortening velocity, and power qualitatively similar to normal muscles injured by lengthening contractions (44). We also evaluated how other changes characteristic of the HU model, particularly soleus fiber length, may predispose the muscle to reloading injury. Finally, we studied physiologically relevant periods of HU (15 days) and recovery (9 days) that represent the time at which net protein degradation peaks (36) and a time at which the atrophied soleus mass has undergone partial restoration of mass (20, 37).
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Treatment groups. One week after arrival, animals were randomized by body mass to one of the following treatments: a group that was hindlimb unloaded for 15 days (day 15 HU), a group that was hindlimb unloaded for 15 days followed by 9 days of weight bearing reloading (day 24 HU-R), a weight-bearing (WB) group time matched to the day 15 HU group (day 15 WB), a weight-bearing group time-matched to the day 24 HU-R group (day 24 WB).
Hindlimb-unloading procedure. The hindlimb-unloaded rats were tail suspended using the procedure and apparatus described by Morey-Holton and Globus (24). Briefly, the tail was sprayed with tincture of benzoin. A strip of Skin-Trac tape (cat. no. 3874–03; Zimmer, Dover, OH) was formed into a loop and affixed to the tail. The loop was attached to an overhead trolley system so that the weight-bearing activity of the hindlimbs was eliminated. The animal was able to use the forelimbs to move about the enclosure but was unable to contact any part of the enclosure with the hindlimbs. Waste fell through a plastic grid that made up the floor of the enclosure. Animals were monitored twice daily for overt signs of stress and to ensure that the tail harness did not occlude tail blood flow. Caloric consumption over the previous 24 h was determined by weighing the chow remaining in the cage each day.
Recovery procedure. At day 15 of the study, animals in the HU-R group were removed from the hindlimb-unloading apparatus and housed individually in rodent cages for 9 days under the standard conditions used prior to unloading.
Pair-feeding. The day 15 WB group was pair fed to match the day 15 HU group. The day 24 WB group was pair fed to match the daily chow intake of the HU-R group during both unloading and reloading periods.
Tissue collection. Animals were anesthetized by an intraperitoneal injection of pentobarbital sodium (80 mg/kg body mass). Animals in the day 15 HU group were anesthetized while still hindlimb unloaded to eliminate any muscle damage that might occur with the resumption of weight-bearing activity. Both solei were dissected under anesthesia, and the animals were euthanized.
Muscle function.
The left soleus was pinned out at slack length on a Sylgaar base in a small dish containing a modified bicarbonate buffer (composition, in mM: 137 NaCl, 11 glucose, 5 KCl, 1.25 CaCl2, 1 MgSO4, 1 NaH2PO4, 24 NaHCO3, 0.025 tubocurarine chloride) maintained at 
20°C and continuously aerated with 95% O2 and 5% CO2. Silk suture (4–0) was attached to both tendons. The muscle was transferred to a glass tissue bath containing bicarbonate buffer aerated with 95% O2 and 5% CO2. There, the muscle was mounted between an immovable post and the arm of a dual-mode muscle lever system (model 305C-LR, Aurora Scientific, Aurora, Ontario, Canada). A water jacket surrounding the bath maintained buffer temperature at 22°C.
The lever system was interfaced to a personal computer via a data acquisition board (model 6229; National Instruments, Austin, TX). A custom program written in LabView (ver. 7.1; National Instruments) controlled lever arm movement and the output of a biphasic constant current muscle stimulator (model 701; Aurora Scientific), while simultaneously recording muscle force and lever position (1,000 Hz).
Muscles were stimulated using 200-µs square wave pulses delivered to platinum electrodes flanking the muscle. Before data collection, stimulation current and muscle length were adjusted to maximize tetanic force during a 1,000-ms stimulus train at 100 Hz. The length of the muscle was measured with digital calipers and recorded as optimal length (Lo). All subsequent data were collected at this pulse duration, stimulation current, and muscle length.
Twitch force was recorded after stimulation by a single stimuli. Twitch force records were filtered with a Butterworth filter (100-Hz low-frequency cutoff). Peak force, contraction time (time from onset of force to peak force), relaxation time (time for force to drop from peak to peak), the peak rate of tension development (+dP/dt, maximal positive slope of the force record determined over a 10-ms moving window) and the peak rate of tension decline (–dP/dt, maximum negative slope of the force record measured over a 10-ms moving window) were determined on two twitch records, and the mean values were used for subsequent analysis.
A force-frequency relationship was obtained by stimulating the muscle with 1,000-ms trains at 7 different frequencies ranging from 5 to 150 Hz. Maximum force at each frequency was measured and normalized to the peak isometric force of the muscle.
A force-velocity relationship was determined using isovelocity contractions, as previously performed in our laboratory (44) and illustrated in Fig. 1. Relaxed muscles were extended to 110% of resting fiber length, stimulated to attain peak isometric force, and allowed to shorten at a constant velocity to a final length of 95% fiber length. A rapid release of up to 3% fiber length, to reduce series elasticity, proceeded the period of isovelocity shortening. For purposes of data collection only, fiber length was estimated as 0.6 optimal muscle length (46).
|
Force and velocity data were fit by the hyperbolic Hill equation (47) using an iterative curve-fitting procedure (Marquardt-Levenberg algorithm; SigmaPlot, ver. 9.0, Systat Software, Point Richmond, CA). Peak isometric force (averaged across the force-velocity test), maximal shortening velocity (Vmax, defined as the velocity axis intercept of the relationship), and a/Po (the parameter describing the shape of the force-velocity relationship) were used to calculate peak power (47).
Determination of fiber length. At the conclusion of the experiment, the muscle was removed from the lever system, and the tendons were carefully trimmed. The muscle was blotted, and its mass was determined on an analytical balance. The muscle was then pinned to a piece of rigid plastic and fixed in 10% formaldehyde. The fixed muscle was macerated in 20% HNO3 and stored in a 50% glycerol solution. Small bundles of fibers were dissected from each muscle under a stereomicroscope. The length of the muscle and 5–10 fibers were determined using a digital micrometer at x10–20 magnification. Care was taken so that only complete, unbroken fibers were measured. The average fiber length-to-muscle length ratio and the muscles' Lo were used to calculate the average fiber length of each muscle.
Normalization of muscle functional properties. Physiological cross-sectional area (CSA) was calculated by dividing muscle mass by the product of fiber length and muscle density (assumed to be 1.06 mg/mm3). Force was normalized to physiological CSA, shortening velocity to fiber length, and power to muscle volume.
Measurement of fiber area.
The right soleus was pinned out at approximately in vivo length, frozen in isopentane, cooled to its freezing point with liquid nitrogen, and stored at –80°C. Later, serial cross sections (10 µm thick) were cut from the midbelly of the muscle on a cryostat (–20°C) and collected on poly-L-lysine-coated glass slides. Sections were fixed in ice-cold acetone for 5 min, air dried for 5 min, and rehydrated in ice-cold PBS for 5 min. Sections were incubated with a monoclonal mouse antidystrophin antibody (1:20; NCL-DYS2; Novocastra Laboratories, Newcastle, UK) for 1 h. After a 5-min wash in PBS, the sections were incubated with biotinylated donkey anti-mouse IgG (1 µg/ml PBS; Rockland Immunochemicals, Gilbertsville, PA) for 1 h. Following a 5-min wash in PBS, sections were incubated in a preformed avidin and biotinylated horseradish peroxidase solution (Vectastain ABC; Vector Laboratories, Burlingame, CA) for 30 min. Sections were washed in PBS for 5 min and incubated in diaminobenzidine solution until the desired color intensity had developed. Sections were then gently rinsed with tap water, dehydrated in 95% and 100% ethanol, cleared in xylene, and mounted with acrylic resin.
For each section, a rectangular image comprising a known area (typically 523,557 µm2) was captured using a light microscope, digital camera, graphics board, personal computer, and Image J software (National Institutes of Health, Bethesda, MD, http://rsb.info.nih.gov/ij/). For the weight-bearing and reloading groups, each image contained 300 fibers. Images of hindlimb-unloaded muscles contained 400 fibers. A minimum of two nonoverlapping images were captured and analyzed for all muscles.
Areas of a section not comprising fibers were highlighted using a pen-based input device. These areas were then summed to give the nonfiber CSA of the section. Total fiber CSA was calculated as the known section area minus the nonfiber CSA. Fiber number was determined by counting individual fibers within the section. Partial fibers boarding the top and left sides of the image were counted as whole fibers, while partial fibers bordering the bottom and right sides were excluded from the count. Average fiber CSA was determined by dividing the total fiber area by the fiber count.
Tibia length. Tibias were cleaned of soft tissue. Tibial length was measured using calipers as the distance between the intercondylar eminence and the medial malleoulus.
Statistical analysis. Variables were analyzed using a one-way ANOVA, except for the force-frequency data, which were analyzed using a repeated-measures ANOVA. In the event of a significant F ratio, a false discovery rate-adjustment (7, 8) was used to evaluate the following four comparisons (and associated research questions): 1) day 15 WB vs. day 24 WB (was there a difference between the weight-bearing control groups?), 2) day 15 WB vs. day 15 HU (what was the effect of hindlimb unloading?), 3) day 15 HU vs. day 24 HU-R (did reloading effect muscles subjected to unloading?), and 4) day 24 WB vs. day 24 HU-R (did reloading result in recovery to the weight-bearing value?).
Analyses were conducted using the GLM (for one-way ANOVA), ANOVA (for repeated-measures ANOVA), and MULTTEST procedures of SAS (ver. 9.1, SAS Institute, Cary, NC) with a type I error rate of P < 0.05. All data are presented as means ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Animal mass and caloric intake. Body mass did not differ among experimental groups at the beginning of the study (Table 1). Hindlimb-unloaded animals showed a reduction in body mass over days 1–15 despite consuming more kilocalories than the weight-bearing controls. Animals recovering from hindlimb unloading also consumed more kilocalories per day than time-matched weight-bearing controls, and this partially reversed the body mass loss that occurred during HU.
|
86% of the soleus CSA under both WB and HU conditions (Fig. 3A, bottom, 3B, bottom, and 3E).
|
|
|
|
|
|
relaxation time (137 ± 8 vs. 124 ± 14 ms, P < 0.05). This speeding occurred despite the fact that the rate of twitch tension development (+dP/dt, Fig. 8E) and the rate of twitch tension relaxation (–dP/dt, Fig. 8F) were reduced 41% and 25%, respectively. Thus, the post-HU twitch contraction time and relaxation times were "faster" not because of an alteration in kinetics, but because the muscle required a shorter period of time to attain a lower peak force and to relax from this force. Consistent with these changes, HU shifted the force-frequency relationship to lower forces at all frequencies up to and including 60 Hz (Fig. 9).
|
|
Reloading restored soleus Lo, and the Lo-to-tibia length ratio, to the weight-bearing values (Fig. 4, A and B). Because the soleus fiber length-to-soleus Lo ratio did not differ between day 24 HU-R and day 24 WB groups (Fig. 4C), soleus fiber length was also restored to normal (Fig. 4D). As a result of these changes in mass and fiber length, soleus physiological CSA recovered to within 11% of the time-matched WB value (Fig. 5).
Nine days of ambulatory recovery had no effect on the restoration of absolute or specific Po (Figs. 6, A and B), that is, the 9-day reloading values were not different from the values observed immediately after HU. Reloading had a detrimental effect on Vmax, as the HU-R value was significantly slower than values observed for the HU and the WB animals (Fig. 6C). The only force-velocity parameter restored to WB levels by reloading was a/Po (Fig. 6D).
After 9 days of reloading, absolute peak power showed no change from the immediate post-HU value (Figs. 7, A and B). Because of the depressive effect that reloading had on shortening velocity, power normalized to muscle volume was 21% lower after 9 days of reloading than immediately after unloading (Fig. 7C).
The recovery period restored over half the HU-induced reduction in absolute and relative twitch force (Figs. 8, A–C). In fact, recovery of twitch tension greatly exceeded any recovery of tetanic force as the twitch-to-tetanic force ratio was significantly elevated after reloading (Fig. 8D). Reloading restored twitch contraction time (HU-R: 115 ± 9 ms; day 24 WB: 111 ± 10 ms; P > 0.05) but prolonged relaxation time (HU-R: 168 ± 11 ms; day 24 WB: 140 ± 12 ms; P < 0.05). The rate of twitch force development was partially restored after 9 days of ambulatory recovery (Fig. 8E), while the rate of twitch force relaxation (Fig. 8F) showed no change from the postunloading value. After reloading, the low-frequency portion of the force-frequency relationship (Fig. 9) was shifted to greater relative forces compared with both the HU animals (at frequencies up to and including 60 Hz) and the WB animals (at frequencies up to and including 30 Hz). At 150 Hz, the relative force of muscles allowed 9 days of reloading was about 4% less (P < 0.05) than the day 15 HU and the day 24 WB values.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
In addition to changes in mass, we found that soleus Lo and fiber length were both reduced 11% after HU. Because sarcomeres are at a consistent length at Lo, these findings translate into an equivalent reduction in the number of sarcomere in series. As proposed by Riley et al. (34), the chronic plantar flexon characteristic of rat HU (as well as humans subjected to bed rest or spaceflight) is likely to be a significant contributor to the histological and ultrastructural changes that occur in this model of nonweight bearing. More recently, this group has shown that tail-suspended rats show an acute 23% reduction in soleus fiber length (33). This is an important feature of nonweight bearing because chronic muscle shortening results in remodeling of the muscle toward shorter fiber lengths (45, 46).
In addition to documenting a HU-induced reduction in muscle fiber length, we found that this variable was very responsive to the resumption of normal weight-bearing activity, showing full recovery within 9 days of reloading. This implies that restoration of sarcomeres in series, to restore fiber length, occurred sooner than restoration of cross-bridges in parallel, to restore fiber CSA. This finding may simply be due to the fact that the HU-induced loss of cross-bridges in parallel exceeded the loss of cross-bridges in series. Alternatively, adding sarcomeres longitudinally may occur over a faster time course than adding sarcomeres radially. In fact, the apparent rate of sarcomerogenesis noted here falls within the rates observed during tibia distraction (4).
The complete or partial restoration of sarcomeres in parallel and in series contributed to a 60% restoration of soleus physiological CSA. Nevertheless, the ability of the soleus to produce force failed to improve, as both absolute and specific Po after 9 days of normal weight-bearing recovery were the same as the immediate post-HU values. Because fibers comprised 84–90% of the soleus CSA regardless of treatment, our physiological CSA measurements were not likely to be confounded by edema or other changes in interstitial CSA. The conclusion is that the mechanisms underlying the observed force deficit are intracellular in origin.
Our results support previous work showing isometric force deficits during ambulatory recovery from HU (2, 13, 27). These results from animal models may be analogous to the electrically stimulated plantar flexor torque deficits reported in humans recovering from spaceflight and bed rest (25, 26). A previous study reported substantial recovery of rat soleus isometric force, on a per wet mass basis, after 1 wk of standing recovery following 14 days of HU (1). Nevertheless, absolute force still remained over 40% below the weight-bearing control value. Thus, the preponderance of data suggests a delay in the restoration of muscle force after periods of prolonged disuse.
In agreement with previous work (10, 12), we observed an elevation in soleus Vmax immediately after HU. However, a novel finding of this study is that ambulatory activity following HU was detrimental to the ability of the muscle to shorten as Vmax fell below weight-bearing values after 9 days of reloading. It is well documented that HU results in increased expression of fast myosin heavy-chain isoforms in the normally slow soleus (3, 10, 11). It may take several weeks of reambulation for the myosin heavy chain isoform expression to return to normal (9, 37). De novo expression of embryonic myosin heavy chain has also been observed during recovery from HU (23), which may confer a faster shortening velocity to fibers compared with the adult, slow myosin heavy chain isoform (31). Because myosin heavy chain is the major regulator of muscle shortening velocity (32), these previous findings imply that soleus shortening velocity should be elevated during recovery from unloading. Our finding that this is not the case suggests that soleus shortening velocity is uncoupled from myosin isoform expression during recovery from unloading.
This reduction in shortening velocity had important functional implications for the recovering soleus. For instance, peak power per soleus volume was 21% below the immediate post-HU value after 9 days of recovery. Because specific force at 9 days of recovery was unchanged from the HU value, and a/Po had returned to normal, this reduction in normalized soleus power can be attributed to the observed fall in shortening velocity during the recovery period.
The most likely explanation for these observed functional deficits is that the atrophied muscle was injured during recovery. The reductions in Po, Vmax, and peak power observed here are qualitatively similar to the acute changes in function that occur when weight-bearing solei are subjected to damaging lengthening contractions (44). These findings are consistent with, and provide a functional context for, previous literature showing increased muscle cell permeability and ultrastructure disruption in HU-atrophied muscles subjected to ambulatory recovery (16, 19, 21, 38, 41).
Lee et al. (22) have reported that soleus shortening velocity, determined at loads of 35% and 40% of Po, as well as the power produced at these velocities, were unchanged during recovery from HU. A significant difference between Lee et al. (22) and the present study is the duration of recovery, 
1 day vs. 9 days, respectively. It is possible that Lee et al. studied time points that preceded any reloading-induced changes to shortening velocity or power. However, other studies have reported reductions in isometric force within the initial 24 h of reloading (13, 27), changes that would be expected to also reduce power output. Thus, other factors that differ between Lee et al. and the present study may also need to be considered, such as the methodology used to assess shortening velocity and power, the procedures used to suspend the animals, and the duration of unloading.
The present study provides clues as to why HU-atrophied muscles display an increased sensitivity to contraction-induced damage. On the basis of biomechanical studies performed on 14-day HU rats (5), the shorter fibers in the HU-atrophied soleus observed here were likely stretched upon reloading. If the shortened, atrophied soleus is stretched back to its length under WB conditions, this implies that the sarcomeres in HU fibers would be extended 10% beyond their "normal" sarcomere length (with normal defined as the remodeled sarcomere length). On the basis of the known dimensions of the mammalian thin and thick filaments (29) and assuming sarcomere length lies somewhere along the plateau of the length-tension relationship under pre-HU weight-bearing conditions, i.e., from 2.45 to 2.65 µm, the average sarcomere length during reambulation can be calculated to fall between 2.70 to 2.92 µm. Thus, the magnitude of post-HU stretch is sufficient to move sarcomeres onto the descending limb of the length-tension relationship. Sarcomeres extended onto the descending limb are inherently weak, unstable, and prone to disruption (28). As a first approximation, our data therefore support the contentions of Riley et al. (33, 34) that fiber length remodeling during HU predisposes the atrophied soleus to reloading damage.
The model proposed above assumes that the shorter, HU soleus is re-extended to a normal length upon reloading. However, Canu et al. (5) observed greater than normal ankle flexion during treadmill locomotion performed after 14 days of HU. These authors contend that HU-atrophied soleus muscles will be stretched beyond their normal WB length during ambulatory reloading activity. This excessive muscle stretch would exacerbate the effects described in the preceding paragraph and presumably contribute to a greatly heightened susceptibility to myofiber disruption. Unloading-induced alterations to the compliance of the muscle-tendon unit, the passive properties of atrophied muscle fibers, and the characteristics of the sarcomere length-tension relationship (14, 15, 40) could also alter the susceptibility of the muscle to overextension. Whether these changes would work to increase or decrease susceptibility to reloading damage is unknown. Even subtle changes in sarcomere length may have important consequences, as only a subpopulation of overextended sarcomeres may be necessary to initiate myofiber disruption (28).
In addition to changes in the number of sarcomeres, HU results in a reduction in relative dystrophin content, a protein involved in the transmission of force across the sarcolemma (6), and a change in the elastic properties of titin, a protein that maintains A-band alignment in the center of the sarcomere (15, 40). It is possible that either of these changes could alter the susceptibility of the atrophied soleus to the mechanical stresses encountered during reambulation. Thus, there may be multiple contributors to the damage that occurs during recovery from HU.
An important observation of this study is the dissociation between recovery of soleus morphology (mass, physiological CSA, fiber CSA) from soleus function. On a practical basis, this means that measurements of soleus mass are inadequate for characterizing the return of the HU-atrophied soleus to normal levels of function and performance. The mechanisms underlying this dissociation are not clear. It is known that not all proteins are restored at the same rate after HU (35). Thus, our data could be explained by a failure of one (or more) key contractile, regulatory, or structural protein(s) to keep pace with increases in total cell protein content during recovery. An attractive candidate protein might be one involved in excitation-contraction (EC) coupling, as the EC coupling process is susceptible to contraction-induced muscle damage (43), EC coupling fails soon after reloading (13), and exercise-induced EC coupling failure may require many days to be restored to normal (18).
A role for EC coupling failure in modulating muscle function during recovery from atrophy would seem to go against the observed partial recovery of twitch force and force at low frequencies of stimulation, as these variables are sensitive to the processes involved in muscle activation (17). However, twitch properties and force-frequency relationships are influenced by muscle myosin isoform expression. The interpretation of the present twitch and low-frequency force data is therefore not straightforward due to increases in fast and developmental myosin isoforms during atrophy and recovery. Finally, the resumption of weight-bearing activity is also associated with an inflammatory response (13), and there is accumulating evidence that cytokines and immune cells can impact not only muscle regeneration (39), but also muscle function (30). Future work will need to directly evaluate the processes involved in muscle contraction to identify the site of contractile dysfunction during recovery from HU.
It should be noted that even though we observed no difference in soleus force for the HU and HU-R animals, some functional recovery may have occurred during the initial 9 days of reloading. On the basis of the work of others (13), it seems likely that soleus Po reached a nadir prior to the 9th day of reloading. This means that the present data may actually have been obtained as Po had started to recover. Whether Vmax and power also show recovery during the initial 9 days of reloading is unknown.
Perspectives and Significance
Soleus Po, Vmax, and peak power either failed to respond, or showed further deficits, during the initial 9 days of reloading after HU. Over the same reloading period, soleus mass, fiber CSA, and physiological CSA showed substantial, and, in some cases, complete, recovery. This dissociation between morphological and functional variables has several implications. First, it shows that changes in soleus mass and fiber CSA are not reliable markers of functional recovery from HU. Second, it supports the hypothesis that atrophied soleus muscles undergo damage upon the resumption of weight-bearing activity. We propose that this hypersensitivity to weight bearing is due to the reductions in soleus fiber length noted here, a loss in the number of sarcomeres in series, and the subsequent stretch of these sarcomeres onto the descending limb of their length-tension relationship upon the resumption of weight-bearing activity.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Trappe, D. Costill, P. Gallagher, A. Creer, J. R. Peters, H. Evans, D. A. Riley, and R. H. Fitts Exercise in space: human skeletal muscle after 6 months aboard the International Space Station J Appl Physiol, April 1, 2009; 106(4): 1159 - 1168. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
