(Hypertension. 2001;37:1375.)
© 2001 American Heart Association, Inc.
Scientific Contributions |
From Howard Florey Institute of Experimental Physiology and Medicine, University of Melbourne, Victoria, Australia.
Correspondence to Dr M.J. McKinley, Howard Florey Institute, University of Melbourne 3010, Victoria, Australia. E-mail mmck{at}hfi.unimelb.edu.au
| Abstract |
|---|
|
|
|---|
Key Words: angiotensin renin receptors, angiotensin renal nerves
| Introduction |
|---|
|
|
|---|
We have observed that ICV infusion of either Ang II or hypertonic saline suppresses PRC in sodium-depleted sheep without increasing arterial pressure.8 9 This contrasts with the pressor effects of such ICV infusions in sodium-replete sheep,6 suggesting that the reduced plasma renin levels are not a reflex response to baroreceptor activation. As well as suppressing PRC, ICV infusion of either Ang II or hypertonic NaCl reduces renal sympathetic nerve activity considerably in conscious sheep,6 showing that the reduced PRC may have resulted from reduced renal nerve activity, a conclusion reached earlier from studies that used renal denervation.4
In regard to the brain regions mediating effects of ICV Ang II on PRC, the lamina terminalis may be crucial. This region of the brain is rich in AT1 receptors,10 and transsynaptic neural tracing studies with pseudorabies virus show that the lamina terminalis is neurally linked to the renal nerves.11 Moreover, high levels of PRC result if the lamina terminalis is ablated,12 13 suggesting the disruption of an inhibitory influence on renin secretion. The purpose of this study was to characterize more fully the mechanisms by which central Ang II suppresses renin secretion. Sodium-depleted sheep, having elevated PRC, have been studied so that the full extent of central inhibitory influences of Ang II on renin secretion could be observed in the absence of any change in arterial pressure. Our aims were (1) to determine whether an AT1 receptor mediates the suppression of PRC caused by ICV Ang II, (2) to determine, by studying the effects of brain lesions, whether tissue in the lamina terminalis has a role in this response, (3) to study the role of vasopressin in the suppression of PRC in response to ICV Ang II, and (4) to investigate whether renal nerves have a role in the ICV Ang IIinduced reduction of PRC when blood pressure does not rise. Sheep with denervated kidneys were used for this purpose.
| Methods |
|---|
|
|
|---|
At least 2 months before experimentation, sheep were given a general anesthetic induced with intravenous (IV) sodium thiopentone (17 mg/kg) and maintained with an isoflurane/oxygen gas mixture. The carotid arteries were surgically enclosed in skin loops in the neck, and the ovaries were extirpated. At a second operation (with the same general anesthesia), a 17-gauge stainless steel guide tube was implanted over each lateral cerebral ventricle as described previously.14 During the same operation, in 7 of the sheep, a stainless steel electrode (insulated except for 6 to 7 mm at the tip) was implanted along the anterior wall of the third ventricle, and in another 5 sheep, bilateral stainless steel electrodes were implanted over the median eminence. Electrode positions were verified radiographically. Lesions were made by heating these sites to 58° to 60°C for 4 minutes by passage of radiofrequency current with a Grass LM-4 instrument. These methods have been described previously.14 After ablation of the lamina terminalis, sheep usually showed temporary reductions in daily fluid intake and were maintained with daily ruminal water loads.
In another 6 sheep, both kidneys were surgically denervated, as described previously.15 Briefly, all visible nerves to the kidney were removed, and the renal artery and vein were swabbed with 10% phenol/ethanol after stripping of the adventitia. Catecholamine-containing nerve terminals as shown by formaldehyde-fluorescence histochemistry are totally absent in kidneys for at least 1 month after this procedure.15 Animals were allowed 1 to 2 weeks to recover.
Experimental Protocols
Sheep were sodium-depleted by loss of saliva from a
cannula inserted into a parotid duct (as described
previously).9 After 2 days of
loss of saliva, food and water were removed from the cage. A
polyethylene cannula filled with heparinized isotonic saline and
attached to an 18-gauge needle was then inserted into a carotid artery
and fixed there. In some experiments, arterial pressure was
recorded by connecting the arterial cannula to a
pressure transducer coupled to a polygraph. A 3-way tap enabled blood
samples to be drawn from the carotid artery. Two samples of blood were
obtained
15 minutes apart; a needle was then inserted into the
lateral ventricle (as evidenced by withdrawal of a few drops of
cerebrospinal fluid [CSF]), and an infusion of test solution (Ang II,
losartan, 0.75 mol/L NaCl, or artificial CSF) at 1 mL/h into
the lateral ventricle was commenced and continued for 1 hour. Blood
samples (10 mL) were obtained at 30 and 60 minutes after the start of
the ICV infusion and at 30 and 90 minutes after the ICV infusion was
terminated. In some experiments (in which either artificial CSF or 0.75
mol/L NaCl had been infused first into lateral ventricle), a second
infusion of a different test substance was then made 2 hours after the
completion of the first infusion. Blood samples were again obtained at
30 and 60 minutes after commencement and 30 and 90 minutes after
termination of infusions. In some experiments, instead of an ICV
infusion, an infusion of test solution was made into a jugular vein at
1 mL/h for 1 hour, with similar time sequence of blood sampling. At
least 2 weeks were allowed to elapse between successive experiments,
except in sheep with renal denervation. In these animals, a second ICV
infusion was made on the day after the first infusion while the sheep
were still depleted of sodium, so that experiments were completed
within 1 month of the denervation.
Solutions Infused
All solutions infused into the lateral ventricle were
prepared in an artificial CSF, based on the composition of sheep CSF.
The composition of this artificial CSF was NaCl 150 mmol/L; KCl
2.8 mmol; CaCl2 1.2 mmol/L;
MgCl2 1.0 mmol/L; and
Na2HPO4 0.5 mmol/L.
Test solutions infused into a lateral ventricle at 1 mL/h for 1 hour
were Ang II (Peninsula) at 3 µg/h, losartan (Dupont-Merck) at
1 mg/h, or artificial CSF. Losartan and Ang II were also
infused in combination for 1 hour at these doses preceded by ICV
infusion of losartan for 15 minutes. Infusion of Ang II (3
µg/h) or losartan (1 mg/h) for 1 hour were also made into the
jugular vein.
Analysis of Blood Samples
Blood samples were divided with 2.5 mL going into a
tube with EDTA as anticoagulant and 7.5 mL into another tube containing
heparin. These tubes were chilled, then centrifuged and the
plasma obtained stored at either -20°C (for radioimmunoassays) or
4°C (for ionic analysis). Sodium and potassium concentrations
were measured by ion-selective electrode on a Beckman Clinical
Analyzer. Plasma renin concentration was measured by an
antibody capture technique. Briefly, 60 µL of unknown plasma sample,
50 µL of exogenous renin substrate, and 20 µL of Ang I antibody
were incubated at 37°C for 60 minutes. At the end of the incubation,
1 mL of [125I]Ang I was added. The Ang I
generated was determined by radioimmunoassay and expressed as picomoles
of Ang I generated per milliliter of plasma per hour. Intra-assay and
interassay coefficients of variation were 4% and 8%, respectively.
Arginine vasopressin (AVP) was measured by radioimmunoassay with
antiserum raised in rabbits. It had <1% cross-reactivity with
oxytocin, the sensitivity of the assay was 0.4 pg/mL, and the
intra-assay and interassay coefficients of variation were 3% and
12%.
When experiments were completed in a particular animal, it was killed by injection of IV sodium pentobarbital (100 mg/kg). In sheep with a brain lesion, the head was perfused with saline and fixative, and either coronal or sagittal sections of the hypothalamus were prepared as described previously.14 The extent of lesions was mapped with a microfilm reader (Zeiss, Jena).
Statistical Treatment of Results
Results are expressed as mean±SEM. For
analysis of PRC and plasma sodium, potassium, or protein
concentrations, results were subjected to repeated-measures ANOVA.
Where a significant effect occurred, values for each time period after
the commencement of an ICV or IV infusion were then compared with the
value obtained immediately before infusion by Dunnetts test for
multiple comparisons.
| Results |
|---|
|
|
|---|
|
|
|
Effect of Lamina Terminalis Lesions on Renin
Responses to ICV Infusions in Sodium- Depleted Sheep
Extent of Brain Tissue Ablated
Of the 7 sheep in which lesions were made in the
lamina terminalis, histological examination showed that
5 incurred ablation of >95% of tissue in the lamina terminalis; these
are termed lamina terminalis lesion sheep (LTL sheep). In these sheep,
the subfornical organ and median preoptic nucleus were completely
ablated, whereas there was total ablation of the organum vasculosum of
lamina terminalis (OVLT) in 2 of the sheep, and in the other 3 animals
the OVLT was largely destroyed with a small amount (
10%) of the
most ventral part of the OVLT left intact
(Figure 2). In addition to the lamina terminalis, small
amounts of tissue close to the midline were damaged ventral to the
anterior commissure, and this was mainly the most medial parts of the
medial preoptic region, periventricular preoptic nucleus,
and vertical limb of the diagonal band. Dorsal to the anterior
commissure, lesions extended into the midline part of the medial septal
nucleus. The anterior hypothalamus, bed nucleus of the stria
terminalis, and the supraoptic, suprachiasmatic, and hypothalamic
paraventricular nuclei were not damaged. In the other 2
sheep, the lamina terminalis incurred less damage. The subfornical
organ and median preoptic nucleus dorsal to the anterior commissure as
well as this fiber tract and the medial septal nucleus were severely
damaged (>90% ablated). However, considerable parts (
50%) of the
median preoptic nucleus ventral to the anterior commissure and OVLT
were left intact in these sheep
(Figure 2).
|
Changes in Plasma Renin Concentration
Loss of saliva in LTL sheep increased PRC to
levels similar to or greater than those observed in intact sheep. No
significant change in PRC was observed with ICV infusion of either Ang
II or 0.75 mol/L NaCl in sodium-depleted LTL sheep
(Figure 1C and 1D). No significant changes in plasma sodium,
potassium, or total protein concentrations were observed.
Arterial pressure was not measured. In 2 sheep with smaller
lesions in the lamina terminalis, after 1 hour of infusion PRC fell
from 2.7 to 2.5 and 2.1 to 1.2 pmol Ang I/mL per hour with ICV Ang II
and from 2.8 to 1.4 and 2.4 to 1.8 pmol Ang I/mL per hour with ICV
infusion of 0.75 mol/l NaCl. When ICV infusions of losartan (1
mg/h) were made into sodium-depleted LTL sheep, there were no
significant changes in PRC
(Figure 1C) or plasma sodium, potassium, or protein
concentrations
(Tables 1 and 2).
Effect of ICV Infusion of Ang II in
Sodium-Depleted Sheep With Median Eminence Ablated
Loss of saliva for 2 days caused PRC to increase to
levels similar to those observed in sodium-depleted intact sheep. ICV
infusion of Ang II (3 µg/h for 1 hour) caused PRC to fall
significantly by the end of the infusion
(Figure 1C) and remain at low levels for the duration of the
experiment.
Plasma AVP concentration was measured before and at the end of the ICV infusion of Ang II. It was below the level of detectability (<0.4 pg/mL) and remained so in 3 of the sheep, whereas in the other 2 sheep, small increases to 0.6 and 0.8 pg/mL were measured. No significant change in plasma sodium or potassium concentration occurred, but a significant reduction of plasma protein was observed after the ICV Ang II infusion (Tables 1 and 2). Arterial pressure was not measured.
Effect of Renal Denervation on Renin Response
to ICV Infusion of Either Ang II or Losartan
Plasma renin levels in sheep with denervated kidneys
increased to levels similar to those that occur in sodium-depleted
sheep with intact renal nerves. However, with ICV infusion of Ang II (3
µg/h for 1 hour), no significant reduction in PRC was observed until
30 and 90 minutes after the cessation of infusion. Thus, in comparison
to sheep with intact renal nerves, reduction in PRC induced by ICV Ang
II was delayed by at least 1 hour
(Table 3). Plasma AVP levels increased with ICV Ang II
infusion in these sheep
(Table 3). Plasma electrolyte concentrations did not change.
Plasma protein levels fell at 30 and 90 minutes after the conclusion of
the ICV Ang II infusion.
|
In 5 sheep with kidneys denervated, ICV infusion of the AT1 receptor antagonist losartan caused a significant increase in PRC at 30 and 90 minutes after the termination of the ICV infusion. No changes in MAP or plasma sodium concentration were observed (Table 3). Plasma protein concentration increased significantly 90 minutes after the completion of the infusion.
Effect of IV Infusions of Either Ang II or
Losartan on PRC of Sodium-Depleted Intact Sheep
IV infusion of either Ang II at 3 µg/h or
losartan at 1 mg/h for 1 hour did not significantly change PRC
(Figure 1, A and B) or plasma sodium or potassium (not shown)
of intact sodium-depleted sheep.
| Discussion |
|---|
|
|
|---|
Ablation of the lamina terminalis of sheep prevented the changes in PRC that occur in response to ICV infusion of either Ang II or losartan, suggesting that this region of the brain plays an important role in the central influences of Ang II on renin secretion. The lamina terminalis, which comprises the subfornical organ, median preoptic nucleus, and OVLT, is rich in angiotensin receptors in all mammalian species studied, including sheep.10 In addition, ICV infusion of Ang II has been shown to result in the activation of neurons (as indicated by Fos production) throughout the lamina terminalis of the rat.16 Therefore, it seems likely that the suppression of PRC resulting from ICV infusion of Ang II is due to activation of neural pathways that originate in the lamina terminalis. The observation in 2 sheep with partial lesions, in which some of the ventral lamina terminalis was left intact, that ICV Ang II or hypertonic saline still caused some albeit less inhibition of PRC emphasizes the importance of the ventral lamina terminalis in the response. Weekley7 found that ablation of the medial basal forebrain of rats (producing relatively larger lesions than those made in the present study in sheep) prevented the inhibition of renin secretion in response to ICV Ang II in anesthetized rats. The data in sheep are consistent with this result because the lamina terminalis was included in the region of medial basal forebrain ablated in the rat.
The observation that ablation of the lamina terminalis also disrupted the inhibition of PRC in response to ICV hypertonic NaCl was not unexpected because this inhibitory response is probably mediated by a central angiotensinergic pathway,9 which as discussed above, is also disrupted by such a brain lesion. Previous work in a number of species has shown the importance of the ventral lamina terminalis in other centrally mediated responses to Ang II or hypertonic saline such as drinking, vasopressin secretion, and natriuresis.12 13 14 16 17
Ablation of more caudal midline tissue in the median eminence, which interrupts the AVP-containing axons of the hypothalamo-neurohypophysial tract, almost abolished the increase in plasma AVP concentration resulting from ICV infusion of Ang II. However, this infusion still reduced PRC considerably in these sheep with MEL, suggesting that increased AVP levels do not mediate the central influence of Ang II on PRC, as suggested previously.5
It is possible that the reduced renin secretion in response to ICV Ang II that was reported previously3 4 5 6 7 is due to reduced renal sympathetic nerve activity resulting from baroreceptor activation caused by the pressor effect of ICV Ang II. Arterial pressure did not increase with ICV infusion of Ang II in sodium-depleted sheep, showing that the observed changes in PRC were not secondary to alterations in arterial pressure and baroreceptor activity. ICV infusions of Ang II increase arterial pressure in sodium-replete sheep6 ; therefore, it seems likely that this centrally mediated pressor response is dependent on the animals sodium status, analogous to the pressor action of systemically administered Ang II being inhibited by sodium depletion.18 19 Studies performed with an inhibitor of vascular nitric oxide (NO) production in sodium-depleted sheep showed that NO had a role in the reduced pressor responsiveness to peripheral Ang II.19 We speculate that increased vascular NO may also have a role in the lack of a pressor response to ICV Ang II in sodium-depleted sheep.
The lack of any significant reduction of PRC during the hour of ICV infusion of Ang II in sodium-depleted sheep with denervated kidneys suggests that reduced renal nerve activity may a play a role in the inhibitory influence of ICV Ang II on renin secretion in conscious sodium-deplete sheep, consistent with earlier results in anesthetized cats.4 It has been shown previously that ICV infusion of Ang II causes a pronounced and long-lasting inhibition of renal sympathetic nerve activity in sodium-replete sheep, which is partly independent of baroreceptor activation.6 This effect of ICV Ang II on renal nerve activity, like that on renin secretion, is also prevented by ablation of the lamina terminalis.20 Therefore, we propose that ICV infusion of Ang II acts on neurons in the lamina terminalis to activate polysynaptic neural pathways leading to the eventual inhibition of renal sympathetic nerve activity and renin secretion. We have recently provided a neuroanatomic basis for such a proposal by showing that the neurotropic virus pseudorabies is transported retrogradely from injection sites in the kidney of the rat to the lamina terminalis through at least 3 synaptic passages.11 Because Ang II almost invariably has a direct stimulatory influence on neurons in the lamina terminalis,21 an inhibitory synapse must be involved at one point in the polysynaptic neural pathway for the observed inhibition of renal nerve activity and renin secretion to occur.
Although renal denervation severely disrupted the inhibitory effect of ICV Ang II on renin release, the increase of PRC resulting from ICV infusion of the angiotensin antagonist losartan was still evident in sheep with denervated kidneys. This suggests that an additional factor to renal nerve activity may also have a role in mediating central angiotensinergic influences on renin secretion. Presumably, this effect of ICV losartan is humorally mediated, because no change in arterial pressure occurs with ICV losartan in sodium-depleted sheep. Involvement of such a humoral factor may explain the slowly occurring reduction of PRC that was observed at 30 and 90 minutes after the start of the ICV infusion of Ang II in sheep with denervated kidneys. Plasma AVP levels were elevated after ICV infusion of Ang II in sheep with denervated kidneys, and AVP could have a role in this slow reduction of PRC. It should be noted that a significant fall in plasma protein concentration occurred after the end of the ICV infusion of Ang II in these sheep, indicating a probable increase in blood volume at these times, which may also have influenced PRC.
It is not clear at present why blood volume should increase after ICV infusion of Ang II. It is unlikely that this effect is due to the increased plasma AVP levels causing marked fluid retention by the kidneys because plasma protein concentration also fell in sheep with an MEL, which prevented plasma AVP levels from increasing in response to ICV Ang II. We have observed that ICV infusion of Ang II reduces the copious saliva flow from the parotid gland of the sheep (M.J. McKinley and M.L. Mathai, unpublished observations). This could result initially in less fluid leaving the bloodstream during ICV infusion of Ang II. Thus, blood volume could increase before fluid reabsorption from the gastrointestinal tract was adjusted.
Conclusion
These data suggest that as well as a direct
influence on the kidney of circulating Ang II to inhibit renin
secretion,18 there may also
be an inhibitory influence of brain angiotensin
on renal renin secretion, which is probably signaled to the kidney as a
reduction in renal sympathetic nerve activity.
Angiotensin-receptive neurons in the lamina terminalis may
initiate these changes; however, the polysynaptic neural pathway from
lamina terminalis to kidney, with an inhibitory synapse,
remains to be determined. We speculate that angiotensin may
have actions on neurons in the lamina terminalis that connect to
spinally projecting neurons of the paraventricular
nucleus, to influence preganglionic neurons linked to the renal nerves
to inhibit renal nerve activity and renin secretion. Because parts of
the lamina terminalis lack a blood-brain barrier, it may be influenced
by increases in endogenous circulating Ang II.
Angiotensinergic nerve terminals may innervate the
median preoptic nucleus,17
which cannot be accessed by circulating Ang II. Therefore, it is
possible that either peripherally or centrally generated
Ang II could have a role in the central regulation of renin secretion
by the
kidney.
| Acknowledgments |
|---|
Received September 22, 2000; first decision October 31, 2000; accepted November 20, 2000.
| References |
|---|
|
|
|---|
2. Schramm LP, Strack AM, Platt KB, Loewy AD. Peripheral and central pathways regulating the kidney: a study using pseudorabies virus. Brain Res. 1993;616:251262.[Medline] [Order article via Infotrieve]
3. Eriksson L, Fyhrquist F. Plasma renin activity following central infusions of angiotensin and altered CSF sodium concentration in the conscious goat. Acta Physiol Scand. 1976;48:209216.
4. Lokhandwala MF, Buckley JP, Jandhyala BS. Reduction of plasma renin activity by centrally-administered angiotensin II in anesthetized cats. Clin Exp Hypertens. 1978;1:16775.
5.
Malayan SA, Keil
LC, Ramsay DJ, Reid IA. Mechanism of suppression of plasma renin
activity by centrally administered angiotensin II.
Endocrinology. 1979;104:672675.
6.
May CN, McAllen RM.
Baroreceptor-independent renal nerve inhibition by
intracerebroventricular
angiotensin II in conscious sheep.
Am J Physiol. 1997;273:R560R567.
7. Weekley LB. Renal renin secretion rate and norepinephrine secretion rate in response to centrally administered angiotensin-II: role of the medial basal forebrain. Clin Exp Hypertens. 1992;14:923945.
8. McKinley MJ, Evered M, Mathai M, Coghlan JP. Effects of central losartan on plasma renin and centrally mediated natriuresis. Kidney Int. 1994;46:14791482.[Medline] [Order article via Infotrieve]
9. McKinley MJ, Mathai ML. Centrally administered losartan inhibits the reduction in plasma renin concentration caused by intracerebroventricular hypertonic saline in Na- depleted sheep. Regul Pept. 1996;66:3740.[Medline] [Order article via Infotrieve]
10. Allen AM, McKinley MJ, Mendelsohn FA. Localisation of angiotensin receptors in the nervous system. In: Quirion R, Bjorklund A, Hokfelt T, eds. Handbook of Chemical Neuroanatomy, Vol 16. Peptide Receptors, Part 1. Amsterdam, Holland: Elsevier; 2000:79124.
11. Sly DJ, Colvill L, McKinley MJ, Oldfield BJ. Identification of neural projections from the forebrain to the kidney, using the virus pseudorabies. J Auton Nerv Syst. 1999;77:7382.
12. Schrager EE, Johnson AK. Anteroventral third ventricle (AV3V) region ablation: chronic elevations of plasma renin concentration. Brain Res. 1980;190:554558.[Medline] [Order article via Infotrieve]
13. McKinley MJ, Coghlan JP, Congiu M, Denton DA, Fei DT, Park RG. Augmented plasma renin levels in dehydrated sheep with periventricular lesions. Brain Res. 1986;376:416419.[Medline] [Order article via Infotrieve]
14. McKinley MJ, Mathai ML, Pennington GL, Rundgren M, and Vivas L. The effect of individual or combined ablation of the nuclear groups of the lamina terminalis on water drinking in sheep. Am J Physiol. 1999;276:R673R683.
15. Park RG, Clevers J, McKinley MJ, Rundgren M. Renal denervation does not prevent dehydration-induced natriuresis in sheep. Acta Physiol Scand. 1989;137:199206.[Medline] [Order article via Infotrieve]
16. McKinley MJ, Badoer E, Vivas L, Oldfield BJ. Comparison of c-fos expression in the lamina terminalis of conscious rats after intravenous or intracerebroventricular angiotensin. Brain Res Bull. 1995;37:131137.[Medline] [Order article via Infotrieve]
17. Johnson AK, Cunningham JT, Thunhorst RL. Integrative role of the lamina terminalis in the regulation of cardiovascular and body fluid homeostasis. Clin Exp Pharmacol Physiol. 1996;23:183191.[Medline] [Order article via Infotrieve]
18. Blair-West JR, Coghlan JP, Denton DA, Funder JW, Scoggins BA, Wright RD. Inhibition of renin secretion by systemic and intrarenal angiotensin infusion. Am J Physiol. 1971;220:13091315.
19. Tresham JJ, Boutros P, Wintour EM, May CN. Role of nitric oxide in the attenuated pressor responses of pregnant or sodium deplete sheep. J Cardiovasc Pharmacol. 1996;27:908915.[Medline] [Order article via Infotrieve]
20. May CN, McAllen RM, McKinley MJ. Inhibition of renal nerve activity by central sodium and angiotensin is abolished by lesions of the lamina terminalis. Am J Physiol.. 2000;279:R1827R1833.
21. Ferguson AV, Bains JS. Electrophysiology of circumventricular organs. Front Neuroendocrinol. 1996;17:440475. [Medline] [Order article via Infotrieve]
This article has been cited by other articles:
![]() |
R. Frithiof, R. Ramchandra, S. Hood, C. May, and M. Rundgren Hypothalamic paraventricular nucleus mediates sodium-induced changes in cardiovascular and renal function in conscious sheep Am J Physiol Regulatory Integrative Comp Physiol, July 1, 2009; 297(1): R185 - R193. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hypertension Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2001 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |