Sympathoexcitatory Response to Cyclosporin A and Baroreflex Resetting
We postulate that the sympathoexcitatory response associated with the immunosuppressive agent cyclosporin A is due to an upward resetting of the arterial baroreflex. We performed studies in conscious intact and sinoaortic-denervated rabbits instrumented with catheters and renal nerve electrodes. In intact rabbits, cyclosporin A (20 mg/kg IV, 30 minutes) produced significant increases in renal sympathetic nerve activity (100% to 269±74%, P<.05) but did not increase mean arterial pressure. In intact rabbits, we determined arterial baroreflex curves relating renal sympathetic nerve activity and heart rate to mean arterial pressure by producing ramp increases (intravenous phenylephrine) and decreases (intravenous nitroprusside) in mean arterial pressure. Cyclosporin A treatment produced a shift of the midrange of the baroreflex control of heart rate (78.0±4.1 to 84.6±4.7 mm Hg, P<.05) and renal sympathetic nerve activity (74.6±3.9 to 87.0±4.8 mm Hg, P<.05). Vehicle administration produced no effects on arterial baroreflex curves relating renal sympathetic nerve activity and heart rate to mean arterial pressure. Compared with vehicle treatment, cyclosporin A reduced the maximum gain of heart rate (−5.6±0.6 versus −3.1±0.8 beats per minute per millimeter of mercury, P<.05) but had no effect on the maximum gain of renal sympathetic nerve activity. In conscious sinoaortic-denervated rabbits, cyclosporin A had no effect on mean arterial pressure (95.7±7.3 to 91.8±10.8 mm Hg), renal sympathetic nerve activity (100% to 110±6%), and heart rate (287±10 to 279±8 beats per minute). However, when the same sinoaortic-denervated rabbits were anesthetized with sodium pentobarbital, cyclosporin A (20 mg/kg IV) produced increases in renal sympathetic nerve activity (100% to 189±27%). These data indicate (1) that the sympathoexcitatory response to cyclosporin A depends on baroreceptor afferent input in the conscious state and (2) that this response involves an upward resetting of the arterial baroreflex.
Treatment with CsA, one of the most effective immunosuppressive agents, prolongs survival after organ transplantation but also results in some serious side effects.1 In addition to inhibiting T-cell activation,2 CsA causes hypertension and renal dysfunction.1 3 4 5 6 Studies in both humans and animals suggest that these side effects are predominantly due to an increase in SNA7 8 9 10 11 12 13 ; however, the mechanisms responsible for this increase are not understood. The sympathoexcitatory effects could be a result of the direct effect of CsA on the central or peripheral nervous system to increase SNA. In this case, the magnitude of the response would be buffered by the arterial baroreflexes. Another possibility is that the sympathoexcitatory effects of CsA are due to an alteration in the gain and/or operating point of the arterial baroreflex. Previous studies from this laboratory have shown that resetting the operating point of the arterial baroreflex can determine the level of SNA.14 15 Resetting the operating point of the arterial baroreflex toward higher arterial pressures results in an increase in sympathetic outflow,14 and resetting it toward lower pressures produces a sympathoinhibitory response.15
A previous study with anesthetized rats16 reported that the increase in RSNA induced by CsA depended on subdiaphragmatic reflexes and was still present after SAD. More recently, preliminary data obtained from conscious rabbits suggest that SAD prevented the apparent sympathoexcitatory response to CsA.17 Since anesthesia is known to alter neural factors regulating the cardiovascular system,17 18 19 we designed additional experiments to determine the mechanisms responsible for the sympathoexcitatory action of CsA. We determined the effects of CsA on MAP, HR, and RSNA in conscious rabbits with and without functional arterial baroreflexes. We also determined the effects of CsA on the arterial baroreflex by recording HR and RSNA during ramp increases and decreases in MAP before and after treatment with CsA or vehicle (Cremophor EL).
Experiments were performed on 29 New Zealand White rabbits weighing 1.85 to 3.00 kg. All rabbits were anesthetized with a mixture of 43% ketamine, 14% chlorpromazine, and 43% xylazine (0.5 to 1 mL/kg IM) and instrumented with a silicone elastomer catheter positioned in the abdominal aorta via the right femoral artery for measurement of arterial pressure, a catheter in the right femoral vein for CsA infusion, and one in the right external jugular vein for phenylephrine and nitroprusside infusions. Rabbits that had previously undergone the SAD procedure or sham operation were allowed 7 to 10 days of recovery before catheters were implanted. All catheters were heparinized every other day until experiments were completed. After a minimum of 1 week of recovery from catheterization, the rabbits were again anesthetized with a mixture of 43% ketamine, 14% chlorpromazine, and 43% xylazine, and a retroperitoneal incision was made for implantation of stainless steel recording electrodes (Medwire, Sigmund Cohn Corp) around the renal nerves. The nerves and electrodes were covered with silicone gel (Wacker silicone 604A and 604B, Wacker-Chemie GmbH). Ampicillin (6.0 mg/kg) was administered for 2 days after each surgery, and nalbuphine (2.0 mg/kg) was administered immediately after each surgery. All surgical procedures were performed in accordance with institutional guidelines.
SAD was performed by cutting the carotid sinus, aortic depressor, and superior laryngeal nerves.19 20 The adventitia was stripped from the carotid sinus and other small vessels in the region. The vessels were then painted with 10% phenol.20 The completeness of the denervation was confirmed by the lack of reflex increases in HR and RSNA in response to acute reductions in MAP produced by intravenous injections of sodium nitroprusside.
Experimental Instrumentation and Data Acquisition and Analysis
Each rabbit was conditioned to the laboratory environment before testing. After a minimum 2-day recovery period from implantation of the renal nerve electrodes, the rabbit was placed in an acrylic rabbit restrainer. Blood pressure was determined by connecting the femoral arterial catheter to a pressure transducer (CDXIII, Cobe Laboratories Inc) coupled to a dynograph (Beckman 611). MAP was obtained electronically with the use of a filter with a 0.5-second time constant. HR was determined with a cardiotachometer (Beckman 9857 B) triggered by the arterial pressure pulse. RSNA was amplified with an AC preamplifier (Grass P5) with the use of a 30- to 3000-Hz bandwidth. Whole-nerve activity was rectified and integrated with an Analog Devices root-mean-square-to-direct current converter. Mean RSNA maintained a 0.5-second time constant. Background noise was determined by an intravenous injection of either trimethaphan camsylate (ganglionic blocker) in the SAD group or phenylephrine to increase arterial pressure in the other groups.
The electronic signals representing MAP, HR, integrated RSNA, and mean RSNA were fed into an analog-to-digital converter (MacLab, AD Instruments) and analyzed with a Macintosh microcomputer. The digitized data were analyzed by a logistic sigmoid curve function.20 21 22 RSNA values were recalculated as (RSNA)×100/(Mean RSNA of Baseline Data in the Pre-CsA or Pre-Vehicle Infusion Period) and expressed as a percentage.
Rabbits were given at least 30 minutes to stabilize before the experimental protocols were initiated.
MAP, HR, and RSNA dose-response curves to CsA were determined in conscious rabbits (n=5). After 15 minutes of stable baseline data collection, CsA (Sandimmune, Sandoz Pharmaceuticals) was administered intravenously. Intervals of 20 minutes were allowed for each dose, with data collection occurring in the last 10 minutes of each administration. The dose at each interval totaled 0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 10.0, and 20.0 mg/kg, respectively.
The effects of CsA versus vehicle on blood pressure, HR, and RSNA were observed. After 15 minutes of stable baseline data collection, CsA (n=6) or an equivalent volume of vehicle (Cremophor EL, Sandoz Pharmaceuticals) (n=6) was infused over 30 minutes to a total dose of 20 mg CsA/kg body wt. MAP, HR, and RSNA were measured continuously before and during CsA infusion. Each rabbit received both treatments with a 3- to 4-day recovery period between each infusion. The order of the treatments was randomized.
To determine how CsA modulates the baroreflex, we constructed arterial baroreflex curves relating HR and RSNA to MAP before and after a 30-minute infusion of either CsA or vehicle by producing ramp increases (intravenous phenylephrine) and decreases (intravenous nitroprusside) in intact rabbits (n=6). Both treatments were performed in the same rabbits, with 3 to 4 days between each. The order of the treatments was randomized. Results before and after CsA infusion were compared with each other and with results before and after vehicle infusion.
To determine the role of the arterial baroreflex in altering RSNA, we infused CsA in SAD (n=6) and sham-operated (n=6) rabbits for 30 minutes to a total dose of 20 mg/kg and compared the effects.
In a previous study,16 CsA increased RSNA in SAD rats under anesthesia; however, in the present study, CsA caused no increases in RSNA in conscious SAD rabbits. To determine whether the anesthetics could influence the RSNA response to CsA, we anesthetized (20 mg/kg pentobarbital IV) the same SAD rabbits (n=6) included in protocol 4 and administered CsA. The order of treatments observed in protocols 4 and 5 was randomized.
In conscious SAD rabbits, the basal RSNA level might be increased to near maximum levels, explaining the reason why RSNA did not increase in conscious SAD rabbits. To determine whether the basal RSNA level in SAD rabbits is near the maximum level, we evoked the nasopharyngeal reflex in SAD (n=6) and sham-operated (n=6) rabbits. One to 3 days later, these rabbits were used in protocol 4. After control levels of MAP, RSNA, and HR were obtained, a puff of cigarette smoke was blown close to the rabbit's nose over 1 to 2 seconds. MAP, HR, and RSNA responses were recorded during and the 2 minutes after this maneuver.23 24
Data are presented as mean±SE. MAP, HR, and RSNA values obtained before and during CsA infusion in the same rabbits were analyzed by repeated measures two-way ANOVA. Significant effects determined by ANOVA were evaluated with the Newman-Keuls multiple range post hoc test. The baroreflex parameters obtained before and after CsA or vehicle infusion and the MAP, HR, and RSNA values in protocol 4 between SAD and sham operation were analyzed by Wilcoxon's nonparametric test. A value of P<.05 was considered statistically significant.
Dose-response curves relating MAP, HR, and RSNA responses to increasing intravenous doses of CsA are shown in Fig 1⇓. No significant increases in MAP were observed during progressive increases in the CsA dose. However, increasing the CsA dose to 20 mg/kg caused a dose-dependent increase in HR from 245±12 to 272±9 beats per minute (bpm) (Fig 1⇓). A similar dose-dependent increase was also observed in RSNA. RSNA increased from 100% to 208±14% at a CsA dose of 20 mg/kg (P<.005).
Serial changes in MAP, HR, and RSNA in protocol 2 are depicted in Fig 2⇓. Control MAP and HR during the control periods did not differ significantly between CsA and vehicle groups. Although infusions of CsA or vehicle caused no significant changes in HR, CsA increased MAP to 85.5±3.7 mm Hg at 20 minutes (P<.05), and RSNA progressively increased to 269±74% at 30 minutes (P<.05). Vehicle infusion had no significant effect on RSNA from 0 to 20 minutes. However, at 25 and 30 minutes, RSNA significantly increased to 112±4% (P<.05) and 122±6% (P<.05), respectively.
In Tables 1 and 2⇓⇓, the parameters and maximum gains of the baroreflex control of HR and RSNA are shown before and after CsA or vehicle infusion. CsA treatment produced a shift of the midrange of the baroreflex control of HR (72.6±4.0 mm Hg after vehicle to 84.6±4.7 after CsA, P<.05) and RSNA (73.1±4.4 to 87.0±4.8 mm Hg, P<.05). CsA reduced the maximum gain of HR (−5.6±0.6 to −3.1±0.8 bpm/mm Hg, P<.05). CsA also increased the range of RSNA response compared with the range before CsA (P<.05), but this increase was not significant when compared with the range observed with vehicle infusion. CsA and vehicle had no effect on the maximum and minimum HR and RSNA nor on the maximum gain of RSNA. Fig 3⇓ shows the reconstructed logistic function curves and gains for baroreflex functions of HR and RSNA before and after CsA or vehicle infusion.
Fig 4⇓ shows the effect of CsA on MAP, HR, and RSNA in SAD and sham-operated rabbits. CsA caused no increase in RSNA (100% to 110±6% at 30 minutes), MAP (95.7±7.3 to 91.8±10.8 mm Hg at 30 minutes), and HR (287±10 to 279±8 bpm) in SAD rabbits. Compared with data observed in sham-operated rabbits, SAD abolished the increases in RSNA.
Fig 5⇓ shows the effect of sodium pentobarbital (20 mg/kg IV) on MAP, HR, and RSNA in CsA-treated SAD rabbits compared with conscious rabbits. With rabbits under anesthesia, CsA caused increases in RSNA in SAD rabbits (100% to 183.2±23.1% at 30 minutes, P<.05).
Fig 6⇓ shows the nasopharyngeal reflexes in SAD and sham-operated rabbits. Increases in RSNA in SAD rabbits were not different from those of sham-operated rabbits.
The present study demonstrates (1) that intravenous infusion of CsA causes increases in RSNA in conscious intact rabbits but not in conscious rabbits without functional arterial baroreflexes and (2) that CsA shifts the midranges of baroreflex control of RSNA and HR to higher pressures.
If the sympathoexcitatory response to CsA is caused by a direct effect of CsA to increase sympathetic outflow, we would expect the response to be buffered by the arterial baroreflex. Accordingly, the sympathoexcitatory response to CsA, as judged by the HR, MAP, and RSNA responses, should be augmented in SAD rabbits. However, in the present study, SAD abolished the sympathoexcitatory response to CsA, indicating that the effects of CsA depended on afferent input from the arterial baroreceptors. As shown in protocol 3, CsA caused a resetting of the midpoint of the arterial baroreflex function (HR and RSNA components) to higher arterial pressures. Previous studies from this laboratory have shown that resetting of the operating point of the arterial baroreflex function determines the SNA level relative to arterial pressure.14 15 For example, the sympathoexcitatory response at the onset of exercise appears to be responsible for an upward resetting of the operating point of the arterial baroreflex, whereas the sympathoinhibitory effect of arginine vasopressin appears to involve a resetting of the operating point toward lower pressures.14 15 In the absence of the arterial baroreceptor inputs, these sympathoexcitatory and sympathoinhibitory responses were abolished. The results of this study suggest that resetting of the operating point of the arterial baroreflex toward higher pressures is a major contributing factor for the sympathoexcitatory effects of CsA.
CsA infusion in rats produces increases in MAP, RSNA, and lumbar SNA.8 9 16 In the present study, although CsA increased RSNA, it had little or no effect on MAP. The lack of a pressor response accompanying the increases in RSNA may indicate that CsA effects were limited to the renal vascular beds. In rats,8 lumbar SNA increased gradually and continued increasing after CsA infusion. However, in rabbits, RSNA increased immediately after the onset of CsA infusion. Therefore, it may not be possible to achieve enough vasoconstriction during a 30-minute infusion. Another possibility is species difference in the vasoconstrictive effect of CsA. In our study, data obtained during vehicle infusion alone suggest that the vehicle (Cremophor EL) may produce a vasodilation that opposes any vasoconstrictor response to CsA. Because of the variability of the decrease in MAP during Cremophor EL infusion, the decrease in MAP at the end of the infusion was not significant compared with the initial values. However, MAP did differ significantly during CsA and vehicle treatments at 20 and 25 minutes. This observation is compatible with studies in dogs which show that Cremophor EL produces decreases in MAP.25 The observation that Cremophor EL decreases MAP is not surprising because it is known to act on diacylglycerol to inhibit binding to protein kinase C.26
As noted previously, CsA infusion caused an upward resetting of the operating point of the arterial baroreflex, resulting in an increase in HR and RSNA but not in MAP. This suggests that the central effects of CsA to reset the operating point were greater than its effects to directly or indirectly increase MAP. Consequently, the increases in HR and RSNA were in part due to an attempt by the baroreflex to adjust the arterial pressure to the new prevailing operating point.
It is important to note that after CsA infusion, resting RSNA was elevated even though resting MAP was unaltered. If the only effect of CsA was to increase the operating point of the arterial baroreflex curve, we would have expected that resting RSNA would remain constant and that MAP would have increased to the same extent as the shift of P3 (see Table 1⇑ legend). However, HR and RSNA were increased while MAP was unchanged. The vasodilator action of the vehicle may oppose the vasoconstrictor effect of CsA, thereby limiting the increase in MAP. Since the resting MAP after CsA infusion is less than the new operating point, the elevated RSNA is mediated via the arterial baroreflex.
CsA reduced the gain of the HR-MAP baroreflex curve. The exact mechanisms responsible for the action of CsA on the HR-MAP baroreflex cannot be determined from the present study; in addition to a central action to alter parasympathetic and/or sympathetic outflow to the heart, CsA may exert a direct effect on the sinoatrial node. In anesthetized SAD rabbits, CsA caused a decrease in HR. The slowing of HR in the absence of arterial baroreflexes may involve either a direct action of CsA on the sinoatrial node or a central action to increase the parasympathetic outflow to the heart. The direct effect of CsA to decrease HR is supported by a recent study in heart transplant recipients in which mild HR slowing was correlated with CsA plasma level.27
CsA had no effects on RSNA in conscious SAD rabbits but did increase RSNA when they were anesthetized (Figs 4 and 5⇑⇑). Many reports suggest that anesthetics alter baroreflex function.17 18 19 Since we did not measure cardiac output in the anesthetized SAD rabbits during CsA infusions, we do not know the mechanisms responsible for the decrease in MAP. One possibility is that the decrease in MAP is the result of the direct vasodilator actions of the vehicle. However, this seems unlikely in view of the response in conscious SAD rabbits. A second possibility is that cardiac output is decreased secondary to the bradycardia. The effect of the decrease in HR on cardiac output may be exaggerated since peripheral resistance is likely to be increased, as reflected by the elevated RSNA.
In the conscious SAD rabbits, the response of RSNA to CsA could potentially be influenced by the RSNA level and/or a decreased responsiveness of the sympathetic nervous system. For these reasons, all of the experiments were performed at least 2 weeks after SAD. At that time, SNA levels are thought to have returned toward the normal range.28 To show that SAD does not alter the responsiveness of the sympathetic nervous system to other stimuli, we determined the RSNA response to activation of the nasopharyngeal reflex in conscious SAD and sham-operated rabbits.23 24 As shown in Fig 6⇑, RSNA increased to similar degrees in conscious SAD and sham-operated rabbits.
During the dose-response curve (Fig 1⇑), HR increased significantly; however, during a 30-minute CsA infusion, HR did not increase (Fig 2⇑). In anesthetized rats,16 the increase in SNA is caused by activation of excitatory neural reflexes arising in the subdiaphragmatic region. In protocol 1, we injected CsA every 20 minutes. These repeated sudden increases in concentration might stimulate neural reflexes, resulting in significant increases in HR; however, the exact mechanism involved could not be determined in the present study.
In conclusion, the sympathoexcitatory response to CsA depends on baroreceptor afferent input and involves an upward resetting of the arterial baroreflex.
Selected Abbreviations and Acronyms
|MAP||=||mean arterial pressure|
|RSNA||=||renal sympathetic nerve activity|
|SNA||=||sympathetic nerve activity|
This work was supported by grants HL-36080 and HL-12415 from the National Heart, Lung, and Blood Institute, National Institutes of Health. We thank Anne Pack for her expert technical assistance and Sue Garner for her secretarial assistance in the preparation of this manuscript.
Reprint requests to Dr Vernon S. Bishop, Department of Physiology, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dr, San Antonio, TX 78284-7756. E-mail firstname.lastname@example.org
- Received April 17, 1996.
- Revision received May 13, 1996.
- Revision received August 27, 1996.
Cohen DJ, Loertscher R, Rubin MF, Tilney NL, Carpenter CD, Strom TB. Cyclosporine: a new immunosuppressive agent for organ transplantation. Ann Intern Med. 1984;101:667-682.
Rogers AJ, Kahan BD. Mechanism of action and clinical applications of cyclosporine in organ transplantation. Clin Immunol Allergy. 1984;4:217-258.
Laskow DA, Curtis JJ. Post-transplant hypertension. Am J Hypertens. 1990;3:721-725.
Bennett WM, Pulliam JP. Cyclosporine nephrotoxicity. Ann Intern Med. 1983;99:851-854.
Morgan BJ, Lyson T, Scherrer U, Victor RG. Cyclosporine causes sympathetically mediated elevations in arterial pressure in rats. Hypertension. 1991;18:458-466.
Lyson T, Ermel LD, Belshaw PJ, Alberg DG, Schreiber SL, Victor RG. Cyne-Cyclosporine- and FK506-induced sympathetic activation correlates with calcineurin-mediated inhibition of T-cell signaling. Circ Res. 1993;73:596-602.
Murray BM, Paller MS. Beneficial effects of renal denervation and prazosin on GFR and renal blood flow after cyclosporine in rats. Clin Nephrol. 1986;25(suppl 1):s37-s39.
Chiu PJS, Vemulapalli S, Sabin C, Rivelli M, Bernardino V, Sybertz EJ. Sympathoadrenal stimulation, not endothelin, plays a role in acute pressor response to cyclosporine in anesthetized rats. J Pharmacol Exp Ther. 1992;261:994-999.
Moss NG, Powell SL, Falk RJ. Intravenous cyclosporine activates afferent and efferent renal nerves and causes sodium retention in innervated kidneys in rats. Proc Natl Acad Sci U S A. 1985;82:8222-8226.
DiCarlo SE, Bishop VS. Onset of exercise shifts operating point of arterial baroreflex to higher pressures. Am J Physiol. 1992;262:H303-H307.
Nishida Y, Bishop VS. Vasopressin-induced suppression of renal sympathetic outflow depends on the number of baroafferent inputs in rabbits. Am J Physiol. 1992;263:R1187-R1194.
Lyson T, McMullan MD, Ermel LD, Morgan JB, Victor RG. Mechanism of cyclosporine-induced sympathetic activation and acute hypertension in rats. Hypertension. 1994;23:667-675.
Matsukawa K, Ninomiya I. Anesthetic effects on tonic and reflex renal sympathetic nerve activity in awake cats. Am J Physiol. 1989;256:R371-R378.
Bishop VS, Hasser EM, Nair UC. Baroreflex control of renal nerve activity in conscious animals. Circ Res. 1987;61(suppl I):I-76-I-81.
White SW, McRitchie RJ. Nasopharyngeal reflexes: integrative analysis of evoked respiratory and cardiovascular effects. Aust J Exp Med Sci. 1973;51:17-31.
McRitchie RJ, White SW. Role of trigeminal, olfactory, carotid sinus and aortic nerves in the respiratory and circulatory responses to nasal inhalation of cigarette smoke and other irritants in the rabbit. Aust J Exp Med Sci. 1974;52:127-140.
Saito M, Terui N, Numao Y, Kumada M. Absence of sustained hypertension in sinoaortic-denervated rabbits. Am J Physiol. 1986;251:H742-H747.