INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE IMPORTANCE OF THE venous system in mammalian cardiovascular function has been well documented (9, 19), whereas the physiological significance of these vessels in fish remains enigmatic. The gravity-free environment in which fish live, plus the absence of parietal valves, certainly argues against the need for a venous pump, and venous return in many fish has been attributed to cardiac aspiration (23, 24). However, the ability of bluefish to withstand head-up tilting when out of water (13, 14) and the rapid posthemorrhagic restoration of arterial blood pressure by trout (3) is suggestive of active venoreflexes. Recent in vivo studies have, in fact, shown that venous tone and/or venous compliance in trout may be selectively affected by exogenous hormones such as arginine vasotocin and natriuretic peptides (NPs), whereas the tonically pressor renin-angiotensin system is solely an arterial constrictor (1, 16, 17, 27). There is no evidence to date, however, to either support or refute tonic regulation of venous function in fish.

The sympathetic nervous system (SNS) is a well-known effector of the fish cardiovascular system, and its contribution to systemic and branchial resistances and cardiac function have been well documented (11). Neural adrenergic tonus maintains systemic arterial resistance in most teleosts, and both neural and humoral stimuli affect cardiac function (11). Although in vitro studies have shown that systemic veins are contracted by catecholamines (2, 18), there is no information on SNS regulation of venous function in vivo.

In the present study, ventral aortic, dorsal aortic, and central venous pressure (P_VA, P_DA, and P_VEN, respectively) and cardiac output (CO) were monitored in unanesthetized trout during epinephrine (Epi) and norepinephrine (NE) infusion to evaluate the effects of these catecholamines in different regions of the circulation. Because Epi was found to be a potent venopressor, a second series of experiments were conducted to specifically examine venous function during catecholamine infusion and blockade of autonomic reflexes. The results show that infusion of Epi, but not NE, increases venous tone and decreases venous compliance, and that inhibition of endogenous SNS activity decreases tone and increases compliance. Thus venous return in fish appears to be tonically regulated, and the SNS plays an important role in this process through its ability to mobilize blood from the unstressed compartment and elevate P_VEN.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals

Pressure Flow Experiments

The pericardial cavity was exposed with a midline ventral incision, and both the right horn of the ductus Cuvier and the bulbus arteriosus were cannulated with 5-cm long, 0.51-mm-ID silicone tubing (Dow Corning veterinary grade; Konigsberg Instruments, Pasadena, CA). The free ends of the tubing were connected to 60 cm of PE-90. All three cannulas were filled with heparinized saline (100 USP units/ml heparin in 9.0 g/l NaCl) and connected to Gould P23 pressure transducers. A 3S Transonic flow probe (Transonic Systems, Ithaca, NY) was placed around the ventral aorta, distal to the cannula, and connected to a Transonic T206 flowmeter. The incision was closed with interrupted silk sutures and sealed with cyanoacrylate gel. Venous and ventral aortic cannulas and the flow probe lead were secured to the fish with silk sutures. The fish were revived and placed in black plastic tubes and immersed in a 1,500-liter experimental aquarium with aerated, through-flowing well water at 15°C. Experiments were conducted 24-48 h after surgery.

Analog pressure signals were displayed with Hewlett-Packard 7853A patient monitors (Palo Alto, CA). Digitized signals of pressure and flow were collected at 0.1-s intervals, and 1-s averages were stored on computer. Resting pressure and cardiac output were visually monitored for 1-2 h before experimentation to ensure that they were stable. Cardiovascular variables were then continuously recorded for a 5-min control period, during 20 min of catecholamine infusion, and for an additional 15-min recovery period. Catecholamines were infused into the dorsal aortic cannula with a syringe infusion pump (model 22; Harvard Apparatus, South Natick, MA). The dead space of the cannula (~0.2 ml) was flushed by a 50-s priming infusion of 10⁴ mol/l catecholamine at 0.3 ml/min. Thereafter, 10⁴ mol/l catecholamines were infused at 2 ml · h¹ · kg body wt¹ (3.3 nmol · min¹ · kg body wt¹).

Pressure transducers were calibrated with a water manometer, and the flow probe was calibrated in situ at the end of the experiment by saline perfusion of the ventral aorta. Mean P_DA, P_VA, and P_VEN were calculated as the arithmetic averages of their systolic and diastolic pressures. Heart rate (HR) was calculated from the pulse interval of either dorsal or ventral aortic pressure. CO was normalized to body weight (kg), and stroke volume (SV) was calculated as SV = CO/HR. Gill (R_G) and systemic (R_S) resistances were calculated from the pressure drop across the respective vasculature relative to CO; i.e., R_G = (P_VA  P_DA)/CO and R_S = (P_DA  P_VEN)/CO.

Vascular Capacitance

In vivo vascular capacitance curves were obtained in conscious trout during conditions of zero-flow CO. Ventricular outflow was prevented by either ventricular fibrillation, as described previously (17, 27) and summarized in the following paragraph, or by a new method of occluding the ventral aorta as described in a subsequent paragraph.

Ventricular fibrillation. Trout were anesthetized in benzocaine, and the dorsal aorta and ductus Cuvier were cannulated as described above. Two coiled stainless steel wire stimulating electrodes (0.126 mm diameter) were placed in the pericardial cavity on either side of the ventricle and exteriorized and secured to the fish along with the venous cannula. The fish were revived and placed in black plastic tubes suspended in the experimental aquarium. Zero-flow conditions were produced by electrical fibrillation of the heart for 6-8 s with a 3.5- to 5.0-V, 40-ms-duration pulse administered at 50 Hz. This method was used for all Epi, low-dose NE (2.6 nmol · min¹ · kg body wt¹), and autonomic blockade studies.

Ventral aortic occlusion. A method to produce zero-flow conditions by occluding the ventral aorta was subsequently developed and employed in an additional group of experiments in which the effects of a high rate of NE infusion (10.4 nmol · min¹ · kg body wt¹) were examined. A sleeve was constructed from a 7.5-mm length of 6-gauge stainless steel tubing. A 2.5-mm-diameter hole was drilled through one wall in the middle of the sleeve, and a 1- to 2-mm-wide notch was cut down the length of the sleeve within 2-3 mm of the hole. A 90° bend was made 5 mm from the end of a 20-mm length of 20-gauge stainless steel tubing, and a piece of heat-flared polyethylene tubing (PE-90) was placed over the short end of the 20-gauge tubing. A piece of latex rubber was cut from a condom and secured over the flared end of the PE tubing with a silk ligature, and a 1-m length of PE-90 tubing was attached to the long end of the stainless steel tubing. The free end of the PE tubing was then inserted into the hole in the sleeve from the luminal side, and the tubing was pulled through the sleeve until the flared end was seated in the hole on the luminal side. A 1-ml syringe was attached to the other end of the PE-90 tubing, and the volume of air required to inflate the latex to the point where it occluded the sleeve was noted. The dorsal aorta and ductus Cuvier were cannulated as described above and the occluder sleeve was fitted over the anterior bulbus and ventral aorta by passing the vessel through the notch. Tubing from the occluder was exteriorized and secured to the fish along with the venous cannula. Inflation of the occluder produced changes in P_DA and P_VEN essentially identical to ventricular fibrillation without stimulation of nearby skeletal muscle.

P_DA and P_VEN were measured in unanesthetized fish before and within 5-7 s after initiation of zero-flow conditions. P_VEN during zero-flow was assumed to be equal to MCFP. Blood pressures were restored within 2-3 s after cessation of ventricular fibrillation or ventral aortic occlusion. Vascular capacitance curves were obtained by measuring MCFP during normovolemia (30-35 ml/kg body wt; Ref. 15), and then again as blood volume was adjusted up or down in 10% increments between 120 and 80% of resting volume. Whole blood from a donor fish was used for volume expansion. Cardiac arrest or ventral aortic occlusion was initiated within 30 s after each volume manipulation, and blood volume was restored to 100% within 30 s after zero-flow pressure measurement. Zero-flow conditions did not exceed 15 s. The interval between each volume perturbation, during which the fish were normovolemic, was 15 min. Control capacitance curves were obtained from fish infused with saline at 0.25 ml/h. The fish were then infused, at the same flow rate, with either Epi (1 nmol · min¹ · kg body wt¹) or NE (2.6 or 10.4 nmol · min¹ · kg body wt¹), and the entire pressure-volume protocol was repeated 15 min after onset of catecholamine infusion. The effects of -adrenoceptor blockade with prazosin (0.47 mol/kg body wt) or phentolamine (1 mol/kg body wt), or nicotinic receptor blockade of autonomic ganglion with hexamethonium (3.7 mol/kg body wt), were examined 20-40 min after a single bolus of the drug.

Because the capacitance curve is not linear (27), C and USBV were determined at three blood volume intervals, 80-90-100%, 90-100-110%, and 100-110-120%, by regression analysis of the three consecutive pressure-volume data points within each interval. By convention, MCFP is treated as the independent variable; therefore, the slope of the resultant volume-pressure line is equal to C, and the intercept of this line with the blood volume axis at MCFP = 0 is the percent of the total blood volume in the unstressed compartment (22). The product of percent blood volume times estimated actual blood volume (30-35 ml/kg body wt) permits conversion of C and USBV into actual volumes, i.e., milliliters per millimeter mercury per kilogram body weight and milliliters per kilogram body weight, respectively.

Chemicals

Statistics

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The effects of catecholamine infusion on cardiovascular parameters in unanesthetized trout are shown in Figs. 1 and 2. Epi infusion increased P_VA, P_DA, and P_VEN, whereas NE increased both arterial pressures but did not affect P_VEN. Epi also transiently decreased R_G and increased R_S. CO, HR, and SV appeared to change, but this was not statistically significant because of the variable responses of individual fish, as has been previously reported (7).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Effects of 3.3 nmol · min1 · kg body wt1 epinephrine (Epi) infusion on cardiovascular parameters in unanesthetized trout. Thick lines represent means and thin lines indicate 95% confidence limits for n = 5 trout; thick horizontal line indicates significantly different from control (P  0.05). Dotted vertical line indicates onset of infusion; infusion ended at dashed vertical line. Values above thick horizontal line are significantly (P  0.05) different from preinfusion. Ventral aorta pressure (PVA; A), dorsal aorta pressure (PDA; C), and central venous pressure (PVEN; E) are in mmHg. Cardiac output (CO; B) is in ml · min1 · kg body wt1. Heart rate (HR; D) is in beats/min. Stroke volume (SV; F) is in ml · beat1 · kg body wt1. Branchial resistance (RG; G) and systemic resistance (RS; H) are in mmHg · ml1 · min · kg body wt.

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Effects of 3.3 nmol · min1 · kg body wt1 norepinephrine (NE) infusion on cardiovascular parameters in unanesthetized trout. A: PVA. B: CO. C: PDA. D: HR. E: PVEN. F: SV. G: RG. H: RS. Symbols as in Fig. 1; n = 5 trout.

The effects of catecholamine infusion on P_DA before and during zero-flow conditions and on vascular capacitance are shown in Fig. 3, and the effects on P_VEN during normal ventricular outflow are listed in Table 1. Blood volume expansion above 100% in saline-infused trout slightly increased P_DA, whereas volume depletion did not appear to affect pressure (Fig. 3, A, C, and E). The effects of volume expansion on P_DA were more pronounced and statistically significant during catecholamine infusion. Changes in P_DA during zero-flow conditions were qualitatively similar to those observed when CO was uninterrupted, albeit at lower pressures. Epi infusion increased P_VEN at all blood volumes (Table 1), whereas NE infusion did not significantly affect P_VEN at any blood volume (not shown).

View larger version (38K): [in this window] [in a new window]   Fig. 3.   PDA (A, C, and E) and vascular capacitance curves (B, D, and F) during infusion of saline control (Con; circles), 1 nmol · min1 · kg body wt1 Epi (triangles; A and B), 2.6 nmol · min1 · kg body wt1 NE (triangles; C and D) or 10.4 nmol · min1 · kg body wt1 NE (triangles; E and F). Closed symbols in A, C, and E indicate pressure during zero-flow conditions; open symbols indicate pressure before ventricular fibrillation (A and C) or occlusion of ventral aorta (E). Values are means ± SE; n = 10 (A-D) or 6 (E-F) trout for each catecholamine. Volume expansion from 100% to 110% or 120% significantly (P  0.05) increased PDA during Epi and low-dose NE infusion and from 80% to 100% expansion increased PDA during high-dose NE infusion. Epi and low-dose NE infusion significantly increased PDA before and during zero-flow conditions at blood volumes 100%, and high-dose NE infusion increased PDA at all blood volumes.

Epi infusion displaced the vascular capacitance curve down and to the right (Fig. 3B) and significantly decreased USBV and C over the blood volume range of 80-100 and 90-110% (Table 2). At 100-120% blood volume, Epi reduced USBV but did not affect C. NE infusion did not affect either USBV or C at any blood volume. The lack of NE effect on MCFP was noted even when the rate of NE infusion exceeded that of Epi to the extent that NE had a greater effect on P_DA than Epi (Fig. 3, E and F).

Prazosin, phentolamine, and hexamethonium reduced P_DA at all blood volumes during both uninterrupted and zero-flow conditions (Fig. 4, A, C, and E). P_DA did not change when blood was withdrawn below 100% before drug treatment; however, after treatment, P_DA continued to fall as blood volume was decreased below 100%. Similar responses were observed during zero-flow conditions. Hexamethonium and prazosin reduced P_VEN in unfibrillated fish (Table 1), whereas phentolamine was ineffective (not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 4.   PDA (A, C, and E) and vascular capacitance curves (B, D, and F) before (Con; circles) and after (triangles) injection of 0.47 mol/kg body wt prazosin (Praz; A and B), 1 mol/kg body wt phentolamine (Phent; C and D), or 3.7 mol/kg body wt hexamethonium (Hexameth; E and F). Solid symbols in A, C, and D indicate pressure during cardiac fibrillation during control or experimental treatment; open symbols indicate pressure before fibrillation. Values are means ± SE; n = 15, 12, and 8 trout for Praz, Phent, and Hexameth, respectively.

Prazosin shifted the capacitance curve to the left (Fig. 4B) and significantly increased USBV and C at 90-110% blood volume (Table 3). At 80-100% blood volume it increased C, and at 110-120% USBV increased. Similarly, hexamethonium produced a leftward shift in the capacitance curve (Fig. 4F), increased C at all blood volumes, and increased USBV at all volumes except 100-120%. Phentolamine did not affect the capacitance curve, USBV, or C. 

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The results of the present study show that infusion of Epi into trout increases both arterial and central venous pressures, whereas NE infusion primarily affects arterial pressures. In a similar manner, Epi alone increases MCFP by increasing venous tone (which shifts blood from the unstressed into the stressed vascular compartments) and by decreasing C (which directly increases venous pressure at a constant SBV). The opposite effects, i.e., a decrease in venous tone and an increase in C, are produced by inhibition of ₁-adrenoceptors with prazosin or by autonomic ganglionic blockade with hexamethonium. These studies indicate that the SNS is an important effector of venous function in trout and they provide the first evidence, in any fish, of tonic regulation of venous tone and compliance. Thus in fish, as in mammals, venous pressure appears to be an important, and regulated, determinant of cardiac filling and therefore CO.

The two primary determinants of CO in mammals are the heart and venous return. Although these two are intimately interrelated and their actions are difficult to separate, numerous studies (reviewed in Refs. 8, 20) support the hypothesis that the latter is perhaps more important in most situations, except for those that result from impaired cardiac function. In this context, the mammalian heart is considered to be more of a sump pump (i.e., filled through extracardiac factors) than a suction pump.

A number of arguments have been made for the opposite situation in fish, i.e., that the heart is the primary determinant of CO (reviewed in Refs. 5, 24). These have been largely based on three observations. 1) Many fish have a rigid pericardium that could support a substantially negative pericardial pressure during ventricular contraction and thereby promote venous filling of the atrium. 2) Central venous pressure is near, or slightly below atmospheric, suggesting cardiac aspiration. 3) Fish in water are in an essentially gravity-free environment and therefore they will not experience orthostatic venous pooling. However, before the present study, active venous regulation in fish had not been directly examined.

The effects of Epi, prazosin, and hexamethonium on venous tone and compliance are consistent with the SNS acting as a tonically active anti-drop regulator of arterial blood pressure in trout. SNS activation can be significant under two circumstances (20): 1) if blood volume is reduced by hemorrhage or other factors, SNS activation will mobilize blood from the USBV to the SBV and thereby restore MCFP, venous return, and CO, and 2) if tissue metabolism is elevated in normovolemic fish, CO can be increased by SNS-stimulated elevation of MCFP and the resulting augmentation of venous return. SNS regulation of venous capacitance in concert with tonic SNS regulation of arteriolar resistance will enable the fish to adjust both systemic resistance and CO commensurate with the desire to maintain arterial blood pressure.

Infusion of Epi mobilizes ~2 ml/kg body wt of blood from the USBV in normovolemic fish (Table 2), which is nearly 7% of the total blood volume and, more importantly, ~15% of the hemodynamically active SBV. This increases central venous pressure in unfibrillated trout by 30-60% and accounts for nearly a 1-mmHg increase in pressure in normovolemic trout. A 1-mmHg increase in central venous pressure will increase CO by ~25% (6) in the isolated trout heart. In trout with an intact pericardium this increase in preload may have an even greater effect on CO (4).

It has been shown in the in situ-perfused trout heart model that central venous pressure becomes elevated when the pericardium is opened and that opening the pericardium lowers CO if preload is maintained constant. (4). However, resting CO (and arterial blood pressure) in trout with an intact pericardium (26) is essentially the same as that observed in trout with a open pericardium (present study). This suggests that CO (or more probably arterial pressure) is being maintained, and in order for this to be achieved central venous pressure may be adjusted upward to compensate for reduced aspiration into the open pericardium. Thus the increase in venous pressure is more likely the direct result of a change in vascular capacitance, i.e., a peripheral response, and not due to blood passively damming up behind a mechanically inefficient heart. If the latter were the case, CO and arterial pressure would be expected to fall.

It is surprising that vascular capacitance and P_VEN are refractory to NE even when NE is infused at 2.5-10 times the rate of Epi and, at the highest NE infusion rate, NE is a more potent arterial pressor (Fig. 3; Table 2). Differences in venous responsiveness are also evident when the two amines are infused at 3.3 nmol · min¹ · kg body wt¹ (Figs. 1 and 2); i.e., although both have nearly the same effect on P_DA, only Epi appreciably affects P_VEN. These differences may be attributable to one or several factors, including different rates of catecholamine inactivation, different accessibility of the receptors to the two catecholamines, different receptors, or different receptor sensitivities. Radiolabeled NE is removed from the trout circulation somewhat faster than Epi (12) and if removal occurs in prevenous systemic vessels this could account for some of the reduced NE effect, although it seems unlikely that differential inactivation kinetics could explain the total lack of NE effect on P_VEN. Similarly, there does not appear to be a differential sensitivity of large vessels to catecholamines because Epi and NE are equipotent constrictors in propranolol-blocked large arteries and veins from trout (2). However, catecholaminergic control of venous capacitance may be a property of the venules, and their receptors have not been characterized. A significant venular contribution to venous capacitance has been proposed based on other studies in trout that have shown considerable differences between large vein responses, in vitro, and vascular capacitance in vivo to NPs and sodium nitroprusside (17). The fact that neither catecholamine affects rapid compliance of large systemic trout veins (2) provides additional support for a venular role in vivo.

It is also surprising that while both prazosin and phentolamine lowered P_DA, only prazosin affected the capacitance curve and P_VEN (Fig. 4, A-D; Table 1). It is tempting to speculate that in these experiments phentolamine's actions are directed at NE effector sites, hence the consistent effects on P_DA and lack of venous responsivity. Clearly, additional studies are needed to clarify the location of vascular receptors in these fish.

Prazosin's effect on P_DA is consistent with a rapid fall in R_S, and even though R_G is unaffected this is still sufficient to reduce P_VA (Fig. 5). The concomitant reduction in P_VEN is likely due to venous-specific events, as predicted from Fig. 4 and Table 3. However, ventricular aspiration could also augment the reduction in P_VEN because CO was clearly not compromised in these experiments.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Effects of 0.47 mol/kg body wt prazosin injection on cardiovascular parameters in unanesthetized trout. A: PVA. B: CO. C: PDA. D: HR. E: PVEN. F: SV. G: RG. H: RS. Symbols as in Fig. 1; n = 4 trout. Prazosin produced a significant (P  0.05) reduction in PVA, PDA, PVEN, and RS within 5 min of injection.

Perspectives

Although SNS regulation of peripheral resistance and venous capacitance appears similar among trout and mammals, the sites and mechanisms of action of other classical cardiovascular regulatory systems are surprisingly different. For example, the renin-angiotensin system, pressor in both fish and mammals (15), has no effect on vascular capacitance in trout (27), whereas it is an important effector of capacitance in mammals (25). Conversely, arginine vasopressin does not affect capacitance in mammals (25), yet its evolutionary antecedent, arginine vasotocin, increases venous tone in the trout (1). Similarly, vasodilator NPs and the nitric oxide donor, sodium nitroprusside, have distinctly different effects in trout and mammals. In trout, NPs primarily affect venous compliance, and sodium nitroprusside decreases arteriolar resistance without affecting the venous system (17), whereas in mammals, NPs reduce arteriolar resistance and NO donors primarily affect venous function (10). These differences point out the inherent dangers in generalization of cardiovascular control mechanisms across different vertebrate classes. However, they also show that there is a pervasive scheme for integrating cardiovascular function, albeit with different messengers, that is common among vertebrates.