Hemodynamic Effects of Scatter Factor in Conscious Rats

Abstract

Scatter factor (SF), also known as hepatocyte growth factor, is a potent mitogen that has been suggested to exhibit greater efficacy than vascular endothelial growth factor (VEGF) in rabbits with hindlimb ischemia. Our study examined the effects of SF on cardiovascular hemodynamics and compared the responses to VEGF. Hemodynamic parameters were monitored before and after administration of SF or VEGF in conscious, instrumented rats. Intravenous injection of SF produced a dose-related reduction in mean arterial pressure (MAP) and increase in heart rate (HR). These responses were significantly attenuated by pretreatment with N^ω-nitro-L-arginine methyl ester, a nitric oxide (NO) synthase inhibitor, suggesting the depressor effect of SF may be mediated by NO. SF (250 μg/kg) reduced stroke volume and cardiac output, but did not affect the maximal first derivation of left ventricular pressure (dP/dt), suggesting that the reduction in cardiac output is caused by decreased stroke volume that probably results from a reduction in venous return. Compared with SF, VEGF produced greater hypotensive and tachycardic responses and greater reductions in stroke volume and cardiac output, indicating that SF has fewer side effects on hemodynamics. Although both growth factors might reduce venous return, SF decreased hematocrit presumably through venodilation, whereas VEGF increased hematocrit as a result of vascular hyperpermeabilty.

Scatter factor (SF), also known as hepatocyte growth factor ^(1,2), is a heparin-binding glycoprotein produced by mesenchymal cells that triggers specific genetic programs of epithelial cells leading to organogenesis and tissue regeneration ^(3,4). SF induces motility, proliferation, and morphogenesis of epithelia by stimulating the tyrosine kinase activity of its specific receptor encoded by the c-met protooncogene ^(5,6). As a cytokine, SF acts not only on epithelial cells, but also on specific cell lineages of mesenchymal cells, including endothelial cells ^(7-12). SF is also synthesized and secreted by vascular smooth muscle and acts on these cells to stimulate migration, protease production, invasion, proliferation, and differentiation into capillary-like tubes in vitro ^(7,13-15). Furthermore, SF has been shown to be a potent angiogenesis factor in vivo ^(13,14,16). The angiogenic response induced by SF in vivo is elicited by stimulation of the SF receptor and requires the presence of both α and β chains ⁽¹⁶⁾.

Vascular endothelial growth factor (VEGF), a major regulator of physiologic and pathologic angiogenesis, is an endothelial cell-specific mitogen. Animal studies have shown that VEGF exerts beneficial angiogenic effects in limb ischemia by systemic injection ^(17,18) and in coronary ischemia by intracoronary ⁽¹⁹⁾ or intramyocardial administration ⁽²⁰⁾. We and other investigators, however, have found that systemic injection of VEGF causes dose-dependent hypotensive and tachycardic responses in rats, rabbits, and pigs ^(21,22). Our previous studies also demonstrate that intravenous injection of VEGF at a dose similar to that used in a rabbit hindlimb ischemic model to stimulate angiogenesis induces a significant reduction in cardiac output and stroke volume ⁽²²⁾. The side effects of VEGF on hemodynamics, including hypotension, tachycardia, and reduced cardiac output, may limit clinical use of VEGF given by systemic bolus injection.

A recent study suggests that, given by intravenous injection at the similar dose, SF exhibits greater efficacy, as evident by collateral formation, regional perfusion, and prevention of atrophy, than does VEGF in a rabbit model of hindlimb ischemia ⁽²³⁾. The hemodynamic effects of SF, however, have not been investigated. Our study was designed to examine the effects of SF on hemodynamics and cardiac performance, to elucidate mechanisms for the effects of SF on cardiovascular hemodynamics, and to compare the hemodynamic responses to VEGF.

MATERIAL AND METHODS

All experimental procedures, which were approved by Genentech's Institutional Animal Care and Use Committee, conform with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication 85-23, revised 1985).

Growth factors

Recombinant human SF was produced at Genentech (South San Francisco, CA, U.S.A.), and endotoxin levels were <2 endotoxin units/mg, as determined in the limulus lysate assay. The 165-amino-acid homodimeric species of recombinant human VEGF was purified from transfected Chinese hamster ovary cells, as previously described ⁽²⁴⁾. Injection volume was 150-200 μl. In a pilot study, intravenous injection of the vehicle of SF or VEGF (200 μl) did not affect mean arterial pressure (MAP), heart rate (HR), or cardiac output in conscious rats.

Systemic arterial pressure and HR in the awake state

Male Sprague-Dawley rats aged 8-9 weeks were anesthetized with ketamine (80 mg/kg) and xylazine (10 mg/kg) given intraperitoneally. Catheters (PE-10 fused with PE-50) filled with heparin/saline (50 U/ml) were implanted into the abdominal aorta via the right femoral artery for measurement of MAP and HR and the right femoral vein for intravenous administration. Catheters were exteriorized and fixed at the back of the neck. All rats were housed individually after surgery.

One day after surgery, the arterial catheter was connected to a model CP-10 pressure transducer (Century Technology Company, Inglewood, CA, U.S.A.) coupled to a polygraph (model 7; Grass Instruments, Quincy, MA, U.S.A.). MAP and HR were measured simultaneously in conscious, unrestrained rats. After a stabilization period, SF (10, 100, 250, or 625 μg/kg) was intravenously injected. MAP and HR were monitored for ≥1 h after injection.

Pretreatment with N^ω-nitro-L-arginine methyl ester (L-NAME)

After anesthesia and cannulation of the femoral artery and vein, MAP and HR were measured as described previously. To determine whether the depressor effects of SF are mediated by nitric oxide, the rats were given an intravenous bolus injection of 75 mg/kg of L-NAME (Sigma Chemical Co., St. Louis, MO, U.S.A.), a specific inhibitor of nitric oxide synthase, dissolved in normal saline. Control animals received an equal volume of saline alone. The result of a pilot study indicated that administration of L-NAME at this dose increased MAP maximally at 4-5 min and that the level remained constant over the next 30 min in conscious rats. SF (100 μg/kg) was injected intravenously 8-10 min after pretreatment with L-NAME or saline vehicle, and MAP and HR were monitored before and after injections. As an additional control in this experiment, rats were pretreated with an i.v. infusion of phenylephrine at 7.5 μg/kg/min for 8 min before SF administration. This dose of phenylephrine produced an MAP response similar to that of L-NAME.

Assessments of cardiac function

Cardiac output was measured by a flowmeter, as described previously ⁽²²⁾. In brief, under anesthesia, rats were intubated and ventilated with a respirator. Through a right-sided thoracotomy, the ascending aorta was exposed. An ultrasonic perivascular flowprobe (Transonic Systems, Inc., Ithaca, NY, U.S.A.) was placed around the ascending aorta. The flowprobe cable was exteriorized and fixed at the back of the neck.

One day after surgery, cardiac output, as shown by the mean flow of the ascending aorta, was digitally obtained by the flowmeter in conscious rats. MAP and HR was measured as indicated. Systemic vascular resistance was calculated as MAP divided by cardiac output, and stroke volume, as cardiac output divided by HR. The hemodynamic parameters were continuously recorded before and after intravenous injection of SF or VEGF at a similar dose (250 μg/kg). This is a therapeutic dose of both growth factors as a bolus for stimulation of angiogenesis in an animal model of limb ischemia ^(18,23).

Assessments of ventricular contractility

After anesthesia, the right carotid artery was cannulated, and a PE-50 catheter advanced into the left ventricle. One day after catheterization, the maximal first derivation of left ventricular pressure (dP/dt) was measured through the ventricular catheter with a pressure transducer connected to a model 7P20 differentiator (Grass Instruments) in conscious rats. Left ventricular dP/dt was measured before and after intravenous injection of SF or VEGF at a similar dose (250 μg/kg). It has been demonstrated that although a fluid-filled catheter system may overestimate dP/dt, the values of dP/dt obtained are useful for making relative comparisons before and after treatment within or between groups of anesthetized and conscious rats ^(22,25-27).

Determination of hematocrit

To avoid the possible influence of blood collection on hemodynamics, hematocrit was measured in a separate group of rats. Blood (0.2 ml) was collected in tubes containing ethylene-diaminetetraacetic acid (EDTA) through the arterial catheter as pretreatment control in conscious, instrumented rats. The animals then received a venous injection of SF (250 μg/kg), an equal volume of vehicle, or VEGF (250 μg/kg). Blood was collected 10 min after injection, the time when maximal decrease in stroke volume and cardiac output were induced by SF or VEGF, and 4 h after injection, the time for recovery as additional controls. Hematocrit was determined by spinning blood-filled capillary tubes in a centrifuge.

Statistical analysis

Results are expressed as mean ± SEM. One-way analysis of variance (ANOVA) was performed to assess differences in parameters at the same time point between groups and to compare changes over time within each group. A value of p < 0.05 was considered to be statistically significant.

RESULTS

Responses of MAP and HR to graded doses of SF

Intravenous injection of 10-625 μg/kg SF resulted in a modest, significant reduction in MAP that began almost immediately, reached a nadir at 3 min, and returned to baseline by 15-20 min after injection. SF concomitantly caused a slight increase in HR, which was rapid and lasted 20 min at the lower doses and 50 min at the higher doses. There were dose-related responses of MAP and HR to SF at 10-100 μg/kg (Fig. 1). The depressor and tachycardic responses to SF at 250-625 μg/kg, however, were not different from those to the dose of 100 μg/kg.

FIG. 1: Maximal responses of mean arterial pressure (MAP, top) and heart rate (HR, bottom) to intravenous injection of scatter factor (SF) at graded doses in conscious rats. Data are presented as the mean of six to eight animals (the number in parentheses is the animal number in each group). *p < 0.05, compared with the lowest dose (10 μg/kg).

Effects of L-NAME on the hemodynamic response to SF

Intravenous injection of L-NAME at 75 mg/kg significantly increased MAP and decreased HR in conscious rats (p < 0.01; Fig. 2). Saline vehicle alone did not alter MAP and HR. Pretreatment with L-NAME significantly attenuated the depressor and tachycardic responses to SF (100 μg/kg; p < 0.01). As a control, the animals were pretreated with phenylephrine at a dose that induced a pressor effect similar to that of L-NAME. In contrast to the results obtained with L-NAME, phenylephrine did not significantly affect the hemodynamic responses to SF (Fig. 2).

FIG. 2: Effect of N ^ω-nitro-L-arginine methyl ester (L-NAME; 75 μg/kg) or phenylephrine (PE; 7.5 μg/kg/min) on the depressor and tachycardic responses to scatter factor (SF; 100 μg/kg). **p < 0.01, compared with the vehicle + SF or PE + SF group (n = 5 per group).

Effects of SF on cardiac function

After intravenous injection of SF (250 μg/kg), cardiac output was transiently increased immediately and rapidly decreased to the basal level at 2 min (Fig. 3, top). However, this increase was not statistically significant. A significant reduction (p < 0.05) in cardiac output in response to SF, which lasted only 5 min, was observed at 5-10 min after injection. SF significantly decreased stroke volume (p < 0.01), which began at 3 min after injection, reached a nadir at 10 min, and lasted 30 min (Fig. 3, bottom). The response pattern of stroke volume was essentially similar to that of cardiac output.

FIG. 3: Effects of scatter factor (SF; 250 μg/kg) on cardiac output (CO, top) and stroke volume (SV, bottom) in conscious rats. *p < 0.05, **p < 0.01, compared with the 0 time point.

SF (250 μg/kg) significantly reduced MAP (Fig. 4, top) and increased HR (Fig. 5, top) as before. A rapid and transient reduction in systemic vascular resistance was found after SF injection (p < 0.01; Fig. 4, bottom). Intravenous injection of SF did not affect left ventricular maximal dP/dt significantly (Fig. 5, bottom).

FIG. 4: Effects of scatter factor (SF; 250 μg/kg) on mean arterial pressure (MAP, top) and systemic vascular resistance (SVR, bottom) in conscious rats. *p < 0.05, **p < 0.01, compared with the 0 time point.

FIG. 5: Effects of scatter factor (SF; 250 μg/kg) on heart rate (HR, top) and maximal first derivation of left ventricular pressure (LV dP/dt, bottom) in conscious rats. *p < 0.05, **p < 0.01, compared with the 0 time point.

Effects of SF on hematocrit

Intravenous injection of vehicle did not induce a significant alteration in hematocrit. In contrast, SF administration at 250 μg/kg produced a significant decrease in hematocrit (p < 0.01) at 10 min, when maximal decrease in stroke volume and cardiac output were seen (Fig. 6). The hematocrit returned to the pretreatment level at 4 h. No significant difference in hematocrit before treatment or 4 h after treatment was observed between vehicle- and SF-treated groups. However, there was a significant difference in hematocrit at 10 min between the two groups.

FIG. 6: Effect of scatter factor (SF) on hematocrit in conscious rats. **p < 0.01, compared with the respective vehicle group at the same time. ^##p < 0.01, compared with the 0 time point in the same group; n = 6 in each group.

Comparison of hemodynamic effects of SF versus VEGF

There were not significant differences in basal levels of MAP, HR, cardiac output, stroke volume, systemic vascular resistance, left ventricular dP/dt, and hematocrit before treatment between SF- and VEGF-treated animals (Table 1).

TABLE 1: Basal levels of hemodynamic parameters before injection of SF or VEGF

At the same dose (250 μg/kg), VEGF markedly reduced cardiac output and stroke volume. The patterns of the VEGF-induced reduction in cardiac output and stroke volume were similar to those induced by SF (Fig. 7). The decrease in these two parameters in both magnitude and duration was much greater in VEGF-treated than in SF-treated animals. The maximal decrease in cardiac output and stroke volume induced by VEGF was −29.9 ± 2.4% and −38.3 ± 2.8%, respectively, and only −5.4 ± 1.9% and −11.3 ± 0.9% by SF (p < 0.0001, VEFG vs. SF for both parameters). The SF-induced reduction in cardiac output and stroke volume lasted only 10 and 30 min, respectively, whereas VEGF caused a prolonged reduction that did not recover >1 h after injection.

FIG. 7: Comparison of responses of cardiac output (CO, top) and stroke volume (SV, bottom) to scatter factor (SF) versus vascular endothelial growth factor (VEGF) at a similar dose. *p < 0.05, **p < 0.01, compared with the 0 time point in the same group. ^##p < 0.01, comparison between two groups after injection.

Compared with SF at the same dose, VEGF also produced greater hypotensive and tachycardic responses (Fig. 8, top and Fig. 9, top). MAP and HR were maximally changed by −17.6 ± 0.8% and 17.3 ± 2.6%, respectively, after VEGF injection and by only −7.9 ± 0.6% and 6.2 ± 1.6% after SF injection (p < 0.0001, VEGF vs. SF for MAP; p < 0.005 for HR).

FIG. 8: Comparison of responses of mean arterial pressure (MAP, top) and systemic vascular resistance (SVR, bottom) to scatter factor (SF) versus vascular endothelial growth factor (VEGF) at the same dose. *p < 0.05, **p < 0.01, compared with the 0 time point in the same group. ^##p < 0.01, comparison between two groups 0.5-20 min after injection (top) and 10-60 min (bottom).

FIG. 9: Comparison of responses of heart rate (HR, top) and maximal first derivation of left ventricular pressure (LV dP/dt, bottom) to scatter factor (SF) versus vascular endothelial growth factor (VEGF) at the same dose. ^##p < 0.01, comparison between two groups 0.5-20 min after injection.

Both SF and VEGF at 250 μg/kg resulted in a rapid and transient reduction in systemic vascular resistance, which was followed by a significant increase in VEGF-injected rats but not in rats receiving SF (Fig. 8, bottom). There was no difference in the effect of SF and VEGF on left ventricular dP/dt, because neither SF nor VEGF caused a significant alteration in dP/dt (Fig. 9, bottom).

At 10 min after intravenous injection at a similar dose, SF significantly decreased hematocrit, whereas VEGF induced a significant increase in hematocrit (Fig. 10). There was no significant alteration in hematocrit 4 h after administration of SF or VEGF.

FIG. 10: Comparison of responses of hematocrit to scatter factor (SF) versus vascular endothelial growth factor (VEGF) at the same dose. **p < 0.01, compared with the respective group at 4 h.

DISCUSSION

Our study first demonstrated that intravenous administration of SF resulted in a dose-related decrease in MAP and an associated increase in HR in conscious rats. The pattern of the reduction in MAP and systemic vascular resistance measured simultaneously was essentially similar. This suggests that the SF-induced depressor response is most likely caused by a decrease in peripheral vascular resistance as a result of the vasodilatory effect of SF.

To test whether the depressor response to SF was mediated by nitric oxide, we pretreated animals with L-NAME, a potent competitive inhibitor of nitric oxide synthase ^(13,22). We found that the depressor response to SF was significantly attenuated by L-NAME. Consistent with previous findings that nitric oxide synthase inhibition caused marked vasoconstriction and hypertension ^(11,12,19), intravenous administration of L-NAME in this study induced a pressor response. Thus a control group of animals was pretreated with phenylephrine to mimic the induced hypertension. It was found that phenylephrine had no effect on the depressor response to SF. These data indicate that the depressor effect of SF is mediated, at least in part, by release of nitric oxide. Our finding in this regard is consistent with that obtained for VEGF ^(22,28), suggesting that the nitric oxide system may play an important role in both growth factor-induced hypotension and vasodilation.

In our study, chronically indwelling catheters and perivascular flowprobes were used continually to monitor MAP, HR, cardiac output, stroke volume, and systemic vascular resistance simultaneously before and after intravenous injection of both growth factors in conscious and unrestrained rats. SF reduced cardiac output in association with decrease stroke volume and increased HR, indicating that the decrease in cardiac output results from reduced stroke volume. Because SF did not affect myocardial contractility, as shown by left ventricular dP/dt, the decrease in stroke volume was not the result of a direct action of SF on ventricular contractility. Further, SF induced a rapid, transient decrease in systemic vascular resistance, which would have a favorable effect on stroke volume, and systemic vascular resistance then returned to the pretreatment level when stroke volume was reduced, indicating that the SF-induced reduction in stroke volume was not caused by the effect on afterload. Therefore the reduction in stroke volume may be related to a decrease in venous return or preload.

Our study confirms the observation of our previous study that intravenous injection of VEGF results in a substantial decrease in stroke volume and cardiac output in conscious rats ⁽²²⁾. The previous study also demonstrated that this effect of VEGF may be caused by a decrease in venous return ⁽²²⁾. Compared with SF at the same dose, the adverse effect of VEGF on stroke volume and cardiac output was much greater and longer, although the reduction in stroke volume and cardiac output induced by both growth factors displayed a similar pattern. Actually SF at 250 μg/kg, the dose equivalent to the therapeutic dose for stimulation of angiogenesis, caused a small (5%) decrease in cardiac output that lasted for only 5 min. This suggests that with increased HR as a compensatory mechanism, the heart function can maintain at near normal despite a significant decrease in stroke volume after SF administration.

Although the reduction in stroke volume induced by both SF and VEGF was similarly caused by decreased venous return, there were opposite intravascular differences between the two growth factors: SF decreased whereas VEGF increased hematocrit. VEGF increased hematocrit, suggesting that hemoconcentration was achieved through extracapillary fluid migration. VEGF is known also as vascular permeability factor (VPF) because of its ability to promote extravasation ^(29,30) or to increase microvascular permeability ^(31,32), thereby decreasing intravascular volume and venous return. In contrast, SF reduced hematocrit presumably through venodilation, which would decrease postcapillary resistance and capillary hydrostatic pressure and favor transcapillary migration of fluid into the circulation ⁽³³⁾.

Both SF and VEGF produced a rapid, transient decrease in systemic vascular resistance. After the early reduction phase, however, systemic vascular resistance was increased in the VEGF-injected animals but not in animals receiving SF. The increase in systemic vascular resistance may be the result of a reflex response to the prominent reduction in cardiac output induced by VEGF, because the late increase phase corresponded to the decrease in cardiac output.

Our previous and present studies consistently showed that VEGF induced hypotension because of vasodilation and an associated increase in HR ⁽²⁶⁾. The results are also in agreement with previous findings from other investigators that VEGF has hypotensive and tachycardic effects in rabbits and pigs ⁽²¹⁾. SF produced only a small decrease in MAP. The hypotensive and tachycardic responses to VEGF were significantly greater than those to SF.

In summary, short-term administration of SF induces slight hypotension, which may be mediated by nitric oxide, tachycardia, and a transient decrease in cardiac output caused by reduced stroke volume, probably resulting from a decrease in preload. Although both SF and VEGF may decrease venous return, SF reduces hematocrit, probably through venodilation, whereas VEGF increases hematocrit as a result of vascular hyperpermeability. Except hematocrit, both growth factors exhibit a similar hemodynamic profile, but the hemodynamic responses to SF are substantially less and shorter than those to VEGF, suggesting that SF may have fewer side effects on hemodynamics and cardiac function.