| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROHUMORAL CONTROL OF CARDIOVASCULAR FUNCTION
1Howard Florey Institute of Experimental Physiology and Medicine; and 2Department of Anatomy and Cell Biology, University of Melbourne, Parkville, Victoria Australia
Submitted 5 January 2005 ; accepted in final form 9 June 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
33°C) caused a low level of ongoing activity. After chemical inhibition of raphé neurons, however, neither strong cooling nor chemical stimulation of RVLM neurons activated tail sympathetic fibers. Electrical stimulation of the RVLM elicited tail sympathetic fiber volleys before and after local anesthesia of the raphé (150–500 nl of 5% tetracaine), demonstrating the existence of an independent descending excitatory pathway from the RVLM. The data show that neurons in both the medullary raphé and the RVLM, acting together, provide the essential drive to support vasomotor tone to the tail. Inputs from these two premotor nuclei interact in a mutually facilitatory manner to determine tonic, and cold-induced, tail sympathetic activity. cutaneous vasomotor nerve; rostral ventrolateral medulla; raphé; muscimol
Other evidence, however, suggests that the medullary raphé is not the only brain stem cell group involved in the control of cutaneous vessels. When the neural pathways to the rat’s tail were traced by transneuronal transport of pseudorabies virus, neurons in other regions, including the rostral ventrolateral medulla (RVLM), were labeled at the same time as premotor neurons of the medullary raphé (27). In cats and rabbits, moreover, experiments using local injections of excitatory or inhibitory amino acids indicate that neurons in the RVLM may supply vasoconstrictor drive to cutaneous vascular targets (6, 11, 21), whereas in rats, injections of D,L-homocysetic acid into the RVLM can prevent the tail vasodilator response to preoptic warming (32). Thus, despite the powerful role of medullary raphé neurons in determining the vasomotor outflow to cutaneous vascular targets, it appears that neurons in the RVLM can also contribute to its control.
One interpretation of the results of viral tracing studies is that all sympathetic vasomotor outflows receive inputs from more than one supraspinal cell group (28–30). It is still unknown, however, what characteristics this basic arrangement confers upon their function. We therefore set out to elucidate the functional consequences of a defined vasomotor outflow receiving drive from two supraspinal sources. We examined the interactions and relative contributions of the RVLM and medullary raphé to the tonic, and cold-induced activity of the sympathetic supply to rat tail vessels.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Experiments were performed on 22 male Sprague-Dawley rats (293–476 g). All experiments complied with the code of practice of the National Health and Medical Research Council of Australia and were approved by the Animal Experiment Ethics Committee of the Howard Florey Institute. Animals were initially anesthetized with an intraperitoneal administration of Brietal sodium (methohexitone sodium, 41 to 111 mg/kg; Eli Lilly Indianapolis, IN) or Nembutal (pentobarbitone sodium, 70 mg/kg; Boehringer Ingelheim). After shaving the trunk and limbs, the trachea was cannulated and the animal was ventilated artificially with 2% isofluorane (Forthane, Abbott) in 100% O2. End expiratory PCO2 concentration was monitored continuously (FM1, Analytical Development, Edinburgh, UK) and was maintained between 3.5% and 4.5% by adjusting ventilation volume. The bladder was cannulated suprapubically, and the bladder was kept drained. The right femoral artery and vein were cannulated for measurement of systemic arterial pressure (AP) and intravenous drug infusion, respectively. The arterial line was filled with heparin sulfate (50 U/ml in 0.9% saline). Instantaneous arterial pressure and airway pressure in tracheae were monitored continuously in all experiments with a transducer to an amplifier (NL108, Digitimer). The animal was mounted supine in a Kopf stereotaxic frame. The trachea and esophagus were cut and reflected rostrally. The medulla was then exposed by a ventral craniotomy. On completion of the surgery, the isofluorane anesthesia was replaced during 30 min with urethane (1–1.5 g/kg iv, Sigma St. Louis, MO). The level of anesthesia was checked at all stages and maintained at a depth sufficient to abolish corneal and withdrawal reflexes. Additional doses of urethane (0.1–0.2 g/kg) were given, if necessary. Neuromuscular paralysis (pancuronium bromide or rocuronium bromide, 1 mg/kg iv) was used only in five experiments where the medulla was stimulated electrically. In these cases, a stable, deep level of anesthesia was established before the paralyzing agent was given, and the animal was allowed to recover between doses so that adequate anesthesia could be confirmed before paralysis was reestablished.
A handcrafted water-perfused Silastic jacket was placed around the animal’s shaved trunk. Three thermocouples were glued to different sites on the trunk skin under the jacket, and the mean of their readings was taken as skin temperature. A fourth thermocouple was inserted 3 cm into the rectum to measure body core temperature. Warm water (36–48°C) was circulated at a constant rate (120–200 ml/min) through tubes in the water jacket. Circulating the jacket with cold water (4–10°C) for 1- to 4-min episodes reversibly lowered skin temperature by 2–7°C, and this was followed by a smaller, delayed fall in core (rectal) temperature.
Chemical Injections
The following solutions were microinjected into the medulla in volumes of 40–200 nl: glycine (0.5 M, Sigma), muscimol (2.1–8 mM, Sigma), L-glutamic acid (50 mM, Sigma), tetracaine (5%, Sigma), all dissolved in artificial cerebrospinal fluid (128 mM NaCl, 2.6 mM KCl, 0.9 mM MgCl2·6H2O, 1.3 mM CaCl2·2H2O, 1.3 mM NaH2PO4, 20 mM NaHCO) (31). A small quantity (1–2%) of rhodamine- or fluorescein-tagged latex microspheres was added to the injectate. A glass micropipette (ID 0.58 mm x OD 1.0 mm) was pulled, and the tip broken back to a diameter of 20–30 µm. It was filled with the drug solution. The pipette was mounted in a micromanipulator and inserted vertically into the RVLM or angled laterally to access the ventral medullary raphé, while avoiding the basilar artery. The latter injections were aimed at the midline, 0.3 mm deep to the ventral surface of the medulla. Injections were made by air pressure, controlled manually by a solenoid valve, and the volume was monitored by the movement of the meniscus viewed through a dissecting microscope with a graticule.
Tail Sympathetic Nerve Fiber Recording
The animal’s tail was placed ventral surface upward and incised approximately half way along the tail to expose the ventral artery. The tail was placed in a pool made from a plastic half tube and sealed around the ends with cotton wool soaked in 4% agar. The pool was filled with liquid paraffin. The ventral collector nerves were identified next to the artery, cleaned, and split into fine filaments with a broken razor blade or sharp insect pins. The central cut end of one filament was placed on one pole of a silver wire electrode pair. An inert strand of connective tissue was placed over the second (reference) electrode. The signal was recorded differentially, amplified (NL104, Digitimer), and filtered (high pass: 50–200 Hz; low pass: 600–2,000 Hz). Fibers were split until the active filament contained a small number of discernible active spikes (few-fiber preparation) which were activated by cooling (22). The active spikes were discriminated with a window discriminator and counted.
Experimental Procedure
The animal was first kept in a warm resting condition (skin temperature 
38.3°C, rectal temperature 37.3°C).
Experiment 1. Circulating cold water (5–10°C) into the water jacket for one or two cooling episodes (each cooling episode is 1–4 min) lowered skin temperature by 2–7°C with delayed fall in core temperature (1–2°C). Then warm water (38–40°C) was reintroduced until core temperature gradually returned to precooling levels. After the cooling procedure, glycine was microinjected bilaterally into the RVLM. The cooling procedure was continued in sequence for the following 60–90 min until the response recovered to preinjection levels.
Experiment 2. Modified cooling procedure was performed. Circulating cold water into the water jacket lowered skin temperature. Then, warmer water, but 5–6°C colder water than precooling water, was reintroduced, so that core temperature returned but it was maintained at 1–2°C lower than precooling levels. After the cooling procedure, muscimol was microinjected into either the medullary raphé or RVLM to produce a long-lasting inhibition of function. In each case the intact region (RVLM or raphé) was then microinjected with L-glutamate to activate its neurons and test whether that activation could drive tail fiber activity. Then, the animal was deeply cooled by the modified cooling procedure to test whether this deep cooling stimulus would reactivate the tail fibers.
Experiment 3. The medullary raphé and the RVLM were both stimulated electrically with concentric bipolar electrodes (0.5–1 mA, 3 pulses, 0.5-ms pulse duration, 5-ms separation). The tail sympathetic fiber response to the stimuli was measured. Stimuli to RVLM and raphé were alternated every 5 s, and each stimulus was repeated once every 10 s. Microinjections of local anesthetic (5% tetracaine, 1–5 injections, 150–500 nl) were then made into the raphé, after which the stimulus protocol was repeated. Then, the animal was given cooling procedure to test whether this stimulus would reactivate the tail fibers.
At the end of each experiment, hexamethonium bromide (10–50 mg/kg), was injected intravenously to confirm that the fibers recorded were sympathetic postganglionic.
Data Analysis
Arterial pressure, airway pressure, tail sympathetic nerve activity, and skin and core temperatures were recorded on a videotape-based instrumentation recorder. These data and discriminated sympathetic unit activity counts were recorded on computer (nerve activity digitized at 10 KHz, other channels at 200 Hz), either online or from tape playback, using the CED Power 1401 interface and Spike 2 software (Cambridge Electronic Design, Cambridge, UK). Few-fiber spike activity was discriminated from the digitized signal using the spike recognition facility in the Spike2 program. The few-fiber discriminated spikes were plotted as 15-s spike counts.
Thresholds and thermal sensitivities of tail nerve records were estimated by plotting 15-s spike counts against mean skin temperature or mean rectal temperature. Linear regression lines of these relations were calculated from appropriate sections of the data, as detailed in our previous paper (22). Statistical analysis of the results was carried out using Student’s t-test for paired data. The level of significance was taken as P < 0.05. Results were expressed as a means ± SE.
Histology
At the end of each experiment, rats were killed with an overdose of pentobarbitone (120 mg/kg iv). The brain was then removed and immersed for 2–4 days in 10% formalin followed by 1 day in 20% sucrose in phosphate buffer. Serial 50-µm coronal sections were cut through the medulla and used to reconstruct injection sites. The sections were mounted on a glass slide, coverslipped, and observed with a fluorescence microscope to assess the location and spread of injections. The spread of each injection was measured from the distribution of fluorescent beads from the injectate. Its full extent was mapped in both coronal and sagittal planes. Spread in the coronal plane was assessed directly by plotting fluorescence distribution from superimposed sections. Rostrocaudal spread was reconstructed by plotting these measurements onto a sagittal map of the medulla. These were referenced to the location of the caudal pole of the facial nucleus.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Using established methods (20, 22), few-fiber activity was recorded from sympathetic fibers supplying the rat’s tail. In every case, these were identified as cutaneous vasoconstrictor fibers by their excitatory response to skin cooling (20, 22) and the fact that their activity was eliminated by hexamethonium bromide (10–50 mg/kg iv) at the end of the experiment. Tail sympathetic spike activity was low or absent when skin and core temperatures were kept warm (38.5°C). Switching the blanket perfusion briefly to cold water typically caused a biphasic activation of tail sympathetic activity, with an early phase responding to the fall in trunk skin temperature and a later phase responding to the delayed fall in core temperature (Fig. 1) (22). Episodes of skin cooling were used as the test stimulus in subsequent experiments.
|
Glycine injections into the RVLM.
Bilateral microinjections of glycine (0.5 M, 120–200 nl per side) were made into the ventrolateral medulla of five rats to cause reversible inactivation of RVLM neurons. The injections in three of these animals were accurately located bilaterally in the RVLM at the rostrocaudal level of the caudal pole of the facial nucleus. In these cases, mean arterial pressure fell to 40 ± 2 mmHg, returning to preinjection levels after 28 ± 10 min. Tail fiber activity was promptly silenced in two of these rats and reduced to 8% of preinjection levels in the third. The tail fiber response to the following cooling episode was substantially reduced (15 ± 8% of the preinjection increase, n = 3) (Fig. 2), and this recovered gradually to preinjection levels over 80 ± 9 min after the injection. In the remaining two rats, one or both glycine injections were placed rostral to the RVLM, level with the rostral part of the facial nucleus. In these cases the falls in blood pressure (
10 mmHg) and tail fiber activity (30% and 100% of the preinjection increase) were modest.
|
Bilateral injection of muscimol into RVLM. Nine rats were subjected to a modified cooling protocol (experiment 2 protocol), during which muscimol was injected to cause prolonged inactivation of RVLM neurons. Figure 3 shows the procedure. After 2 or 3 cooling episodes, the temperature of the water perfusing the jacket was lowered to maintain a steady stimulus and sustained activation of tail sympathetic fibers. Mean trunk skin temperature was then 37.0 ± 0.3°C and mean core temperature 37.4 ± 0.2°C. Then muscimol (2.1 mM, 50–160 nl per side) was injected bilaterally into the RVLM. The muscimol injection reduced tail sympathetic fiber activity by 95 ± 2% (P < 0.01, n = 9). Mean arterial pressure fell to 44 ± 4 mmHg and stayed below 70 mmHg for more than 1 h in all cases.
|
Experiment 2b
Muscimol injections into raphé.
Three rats were subjected to the same temperature manipulations as described in experiment 2 (above), to compare the effects of microinjecting muscimol into the medullary raphé instead of the RVLM. An example is shown in Fig. 4. Before muscimol injections, mean skin temperature was maintained at 36.0 ± 0.3°C, and mean rectal temperature was maintained at 36.1 ± 0.2°C, to produce a strong, steady level of tail sympathetic activity. Muscimol injections (2.1–8 mM, 40–120 nl) into the raphé completely abolished tail nerve activity in all three rats but had little effect on blood pressure [105 ± 21 mmHg before muscimol vs. 96 ± 21 mmHg at 5 min after muscimol (98 vs. 87 mmHg, 72 vs. 72 mmHg, 145 vs. 138 mmHg in each case)]. Further cooling episodes lowered skin temperature by a further 6.8 ± 0.5°C and core temperature by a further 2.6 ± 0.3°C, but failed to activate the tail nerve (Fig. 4). The skin was then rewarmed to 34.0°C, and glutamate injections (50 mM, 120 nl) were made into the left or right RVLM. These raised arterial pressure (peak response +19 ± 5 mmHg, n = 3), but tail sympathetic fibers remained silent.
|
Electrical stimuli (0.5–1 mA, 3 pulses, 0.5-ms pulse, 5-ms separation, every 10 s) were delivered alternately to the RVLM and the raphé before and after local anesthetic was injected into the medullary raphé of 6 rats. Volleys of tail fiber spike activity were recorded in response to electrical stimuli delivered alternately to the RVLM and the medullary raphé. Their response latencies were 350 ± 12 ms (from RVLM) and 393 ± 12 ms (from raphé), which were significantly different (P < 0.05, n = 6) (Fig. 5). Local anesthetic (1–5 injections of up to 150 nl of 5% tetracaine at the same site or 2–3 sites, which were 0.5 mm apart) was then injected into the raphé region to block conduction in cells and fibers of the region. After these injections, tail sympathetic fiber activity fell silent for the rest of the experiment and no longer responded to cooling (Fig. 6). The tail fiber response to electrical stimulation of the raphé was also abolished by tetracaine. The response to RVLM stimulation remained, with no consistent change in latency (from 350 ± 12 to 364 ± 29 ms, P > 0.05, n = 6) or numbers of response spikes (88 ± 28% of preinjection spikes, P > 0.05, n = 6). The tetracaine injection did not cause any consistent change in arterial pressure (110 ± 8 mmHg before injection vs. 94 ± 10 mmHg after tetracaine, P > 0.05, n = 6).
|
|
The locations of effective microinjections made bilaterally into the RVLM and into the medullary raphé were mapped by the distribution of fluorescent beads, as illustrated in Fig. 7. The effective sites match those previously established for the RVLM (7, 15) and the appropriate level of the ventral medullary raphé (20).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
We interpret our results in the following way. Descending pathways from raphé and RVLM neurons converge in the spinal cord, presumably on tail vasoconstrictor preganglionic neurons. Normally, preganglionic tail vasoconstrictor neurons are depolarized beyond their threshold only when they receive synaptic inputs from both RVLM and raphé. The influence of the medullary raphé is the more powerful of the two (23) and probably carries temperature-sensitive drive (1, 4, 17, 20, 24, 32). Neurons in the RVLM (and possibly also other sources) provide essential background drive that maintains preganglionic neuron excitability (11, 21).
We can make a further inference from the special case in which tail sympathetic fibers were reactivated by strong cooling after inhibition of the RVLM. Then, not only was the threshold temperature for activation shifted, but the gain of the response to cooling was much reduced. This implies that normally, RVLM neurons enhance the gain of the tail sympathetic response to raphé neurons. In other words, the two drives show a greater than additive, or mutually facilitatory, interaction.
Before accepting this model, two further points should be considered. First, do both premotor cell groups tonically drive tail sympathetic activity? Yes, because tail sympathetic activity was silenced when either RVLM or raphé neurons were inhibited by agents (muscimol, glycine), which act on cell bodies but not passing fibers. Second, is the tonic drive from each cell group mediated by its own descending pathway, or do RVLM neurons act via the raphé (or vice versa)? The fact that chemical stimulation of raphé neurons could still activate tail sympathetic fibers after inhibition of RVLM neurons, shows that raphé neurons do not act via RVLM neurons. The putative reverse circuit (RVLM 
raphé spinal cord) cannot be eliminated by the same logic, because chemical stimulation of RVLM neurons was ineffective after inhibition of the raphé. Electrical stimulation of the RVLM was effective, however, even after large injections of local anesthetic into the raphé. This shows the existence of a descending pathway from the RVLM, which was not via the medullary raphé; its shorter latency compared with that to raphé stimulation (before tetracaine) further supports this view. Tetracaine was used to block both cells and fibers in the raphé region to prevent "axon reflex" excitation of raphé-spinal pathways from collateral branches, which might project into the RVLM (12). Although other nonraphé pathways [e.g., paraventriculospinal, (26)] could have contributed to the response to RVLM electrical stimulation, we believe it is clear that two independent descending pathways exist from the brain-stem to drive the tail sympathetic supply: one from the medullary raphé and one from the RVLM, presumably used by RVLM neurons.
Why should chemical stimulation of RVLM neurons be unable to activate tail sympathetic fibers after raphé inhibition? We consider the most likely reason is that sympathetic preganglionic neurons controlling tail vasomotion are in a strongly hyperpolarized state after removal of their dominant synaptic input from the raphé. Synchronous volleys from electrical stimuli would cause synchronous EPSPs in preganglionic neurons, giving a high chance of summation to threshold; asynchronous inputs from chemically stimulated RVLM neurons would summate poorly and thus fail to reach threshold.
Although recent work has emphasized the role of the medullary raphé in the control of cutaneous vasoconstrictor outflows such as the rat’s tail (1, 9, 17, 20, 32), other evidence has also implicated RVLM neurons. First, in anatomical studies, in which pathways projecting polysynaptically to the tail were traced with pseudorabies virus, after an appropriate time for transport and replication, virally labeled neurons were found in the RVLM, as well as the raphé (27). RVLM and raphé neurons thus both possess anatomical connections by which they could affect the tail sympathetic supply. Further, in functional studies, it has been shown that experimental activation of neurons in the RVLM can drive the tail vasoconstrictor supply (8), albeit less powerfully than raphé neurons (23) and can inhibit the vasodilator response to preoptic warming (32).The present data are the first to show that the RVLM-tail sympathetic pathway plays a tonic, physiological role in rat, however. This is concordant with evidence from other species, indicating that RVLM neurons provide tonic drive to cutaneous vessels in the cat’s hindlimb (11) and the rabbit’s ear (21).
The present data cannot tell us whether or not RVLM neurons transmit temperature-sensitive drive to the tail sympathetic supply. In cats, McAllen and May (11) recorded from barosensitive bulbospinal neurons in a subregion of the RVLM that, when activated with glutamate injections, briskly drove cutaneous vasomotor fibers in the hindlimb. About a quarter of these cells were inhibited by preoptic warming, suggesting that some RVLM neurons transmit temperature-responsive drive to spinal preganglionic neurons. On the other hand, Tanaka et al. (32) found that while microinjections into the rat’s raphé of the GABAA receptor antagonist, bicuculline, could block the tail vasodilator response to preoptic warming, similar injections into the RVLM did not. This suggests that the preoptic warming signal acts by GABA-ergic inhibition of raphé, but not of RVLM neurons. Additionally, the pattern of Fos expression (indicating neuronal activation) in cold-exposed rats suggests that there is substantial activation of neurons in the medullary raphé but not in the RVLM (4, 15). On balance, therefore, the evidence favors the view that RVLM neurons provide background drive to the tail vasomotor supply without playing a major role in its response to temperature, but further experiments will be necessary to answer this point definitively.
Recent evidence has shown that activity in a second sympathetic outflow that may be activated by cold exposure, the supply to interscapular brown adipose tissue (BAT), depends upon drive from medullary raphé neurons (16, 17). In this case, however, RVLM neurons provide no excitatory drive (other than to a minor, presumed vasomotor, component of BAT nerve activity) and may actually inhibit BAT activity (16). Another interesting parallel arises in the case of the cardiac sympathetic supply: here, both RVLM and raphé neurons can stimulate this outflow when they are activated experimentally (5, 33), but under normal conditions, only the RVLM neurons provide tonic drive (33). The raphé-cardiac pathway remains silent unless activated by stress-related stimuli such as airjet (34), fever (14), or disinhibition of the dorsomedial hypothalamic nucleus (5, 25). Moreover, tonic drive from the RVLM is sufficient to maintain cardiac sympathetic activity without any contribution from raphé neurons (33).
In common with the present findings, however, is evidence for a facilitatory interaction between raphé and RVLM drives to the cardiac sympathetic outflow. The cardiac sympathetic nerve response to disinhibition of raphé neurons with bicuculline survived bilateral inactivation of RVLM neurons; the percentage increase in cardiac sympathetic activity remained the same as before RVLM inactivation, but the absolute increase was reduced in amplitude by about two-thirds (5). This implies that when raphé and RVLM drives are present together, they interact in a greater-than-additive manner to stimulate cardiac sympathetic activity.
The findings of the present study are novel, in that they show how descending inputs from two premotor cell groups are essential for the normal function of a defined sympathetic vasomotor supply. The fact that both RVLM and raphé neurons act together to determine tail vasomotor activity, means that the characteristics of both of these cell groups will be reflected in the properties of the tail outflow. The raphé component of this descending drive probably confers temperature sensitivity. We cannot be certain which characteristics are conferred by the drive from RVLM neurons, but a reasonable possibility is that it is this influence that makes the tail sympathetic output moderately barosensitive (22). The way in which these two drives interact—mutual facilitation with an obligatory requirement for each to be active—results in the output pathway being tuned to respond to small increases or decreases in the activity of either premotor cell group. This may be the mechanism whereby the tail vasomotor supply is rendered sensitive to both general cardiovascular demands (i.e., the support of blood pressure), as well as to specific thermoregulatory demands. In this respect, it differs both functionally and anatomically from the sympathetic supply to BAT, whose thermoregulatory drive also depends on premotor neurons in the same part of the medullary raphé (16, 17), but which receives no drive from the RVLM, presumably reflecting its lack of any cardiovascular role.
A further point is that an arrangement in which mutually facilitatory inputs from more than one premotor cell group converge to drive a single sympathetic outflow, is readily overlooked: if inhibition of one area blocks the response, why look further? It is not the general practice to systematically eliminate the contributions of every potential contributing premotor cell group that might drive a given sympathetic outflow. It therefore remains to be established whether or not an organization similar to that described here for the control of tail vasomotion might apply more generally to other sympathetic vasomotor pathways.
|
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 |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
