Introduction

Top Abstract Introduction Materials and Methods Results Discusssion References

Endothelin-1 (ET-1) is a 21 amino acid peptide with potent cardiovascular effects (Yanagisawa et al., 1988; Miller et al., 1989; Vierhapper et al., 1990; Lerman et al., 1991b; Pernow et al., 1991, 1996; Weitzberg et al., 1991; Wagner et al., 1992). Although elevated circulating levels of this peptide occur in various diseases (Marguulies et al., 1990; Saito et al., 1990; Lerman et al., 1991a, 1992; Cody et al., 1992; Rodeheffer et al., 1992; Stewart et al., 1992; Underwood et al., 1992; Omland et al., 1994; Wei et al., 1994), the kinetics of clearance of this peptide remain poorly understood to date. It remains to be determined if elevated circulating levels of ET-1 in the setting of disease represents an increase in production and release or a reduction of clearance. An improved understanding of ET-1 kinetics also has important implications for understanding the mechanism of action of ET-1 receptor antagonists and endothelin-converting enzyme inhibitors that may significantly alter ET-1 clearance (Hemsen et al., 1996; Plumpton et al., 1996). It is also important to the interpretation of the effects of angiotensin-converting enzyme inhibitors and -adrenergic receptor blockers on plasma levels of ET-1 (Clavell et al., 1996; Krum et al., 1996). Previous human studies of ET-1 have suggested a plasma half-life of 1.4 to 3.6 min (Vierhapper et al., 1990; Weitzberg et al., 1991). One report has suggested a two-compartment elimination process exhibiting a second ("terminal") half-life of 35 min (Weitzberg et al., 1991). These kinetic data are inconsistent with the prolonged vasoconstrictor effects of exogenous ET-1, which have been shown to persist up to several hours (Vierhapper et al., 1990; Lerman et al., 1991b; Pernow et al., 1991; Weitzberg et al., 1991). Importantly, these studies involved infusions of large amounts of recombinant ET-1, producing major hemodynamic perturbations and significant changes in circulating levels of ET-1, both of which may have important effects on kinetics of this peptide (Shichiri et al., 1990; Sirvio et al., 1990; Stewart et al., 1992; Wagner et al., 1992; Mangieri et al., 1997). In addition, infusions of unlabeled ET-1 do not permit the determination of its true kinetics because the infused peptide cannot be detected and differentiated from endogenous ET-1 at low levels.

The kinetics of a vasoactive substance are best described by techniques that avoid hemodynamic changes and alterations of detectable circulating levels, and permit the detection of the exogenous substance at very low levels (Esler et al., 1979). Radiotracer techniques meet these important criteria and have been widely used to determine the human kinetics of other endogenous compounds like norepinephrine (Esler et al., 1979). Although radiotracer techniques have been applied to develop models of ET-1 spillover and extraction across specific vascular beds (Dupuis et al., 1996a,b), total body kinetics of clearance have yet to be described using this approach. To accurately describe the pharmacokinetics of ET-1, we developed a methodology of infusion of ¹²⁵I-labeled ET-1 (¹²⁵I-ET-1) with frequent sampling up to 240 min. We find that a three-compartment model of clearance best fits the observed elimination of ET-1 from plasma. ET-1 has a large volume of distribution and an extremely long terminal half-life. This suggests that ET-1 must be extensively taken up by various tissues throughout the body and provides an explanation for its prolonged vasoconstrictor effect.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discusssion References

Subjects. Five young, healthy male volunteers were recruited through advertisements. Individuals were excluded if they had a history of any significant medical conditions (including hypertension), if they had been taking any medications, or if they had a history of smoking or ethanol or drug abuse.

Study Protocol. Studies were performed at 8:00 AM after the subjects had fasted overnight and refrained from vigorous exercise and ethanol ingestion for 24 h. An i.v. line for ¹²⁵I-ET-1 infusion and a radial arterial line for blood pressure monitoring and blood sampling was inserted. For measurement of the right atrial pressure, a 61-cm Intracath i.v. catheter (Becton Dickinson, Sandy, UT) was inserted into the brachial vein and passed to the right atrium. The catheter was passed without the use of fluoroscopy and its position within the right atrium was confirmed by a typical pressure wave form and evidence of respiratory variation. Catheters were placed while subjects were under local anesthesia without sedation. Pressure recordings were obtained through a Perceptor Morse Manifold Pressure Transducer (Namic Inc., Glens Falls, NY). Individuals rested in a quiet supine position for 45 min after catheters were placed to equilibrate to their surroundings. Subjects remained in the supine position for the entire study.

¹²⁵I-ET-1 Infusion. ¹²⁵I-ET-1 with a specific activity of 780 µCi/µg (2200Ci/mmol) was purchased from New England Nuclear (Boston, MA). Fifty µCi (64 ng) of ¹²⁵I-ET-1 was mixed in 50cc of 5% dextrose and water and infused i.v. at 10 ml/min for 5 min using a Harvard 33 Pump (Harvard Apparatus, St. Laurent, Canada). Heart rate, blood pressure, right atrial pressure, and arterial blood samples for the counting of ¹²⁵I and measurement of ET-1 were obtained at 15, 0, 1, 2, 3, 4, 5, 6, 8, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 100, 120, 150, 180, 210, and 240 min after the start of the infusion.

Measurement of Plasma ET-1 and ¹²⁵I-ET-1. Blood was collected into prechilled tubes containing EDTA (1 mg/ml) and aprotinin (500 kallikrein inactivator units/ml) and placed in ice. Plasma was separated through centrifugation at 4000 RPM at 4°C and stored at 70°C until analysis. On the day of analysis, 2 ml of plasma was acidified with 2.0 ml 4% acetic acid and applied to a conditioned silica C₁₈ cartridge (Sep-Pak, WAT051910, Waters, Milford, MA) through a modification of a previously reported technique (Stewart et al., 1992). The cartridge was conditioned by successively washing with 5 ml 87% methanol, 5 ml 100% methanol, 10 ml ion-free water and 5 ml 4% acetic acid. After washing with 2× 10 ml ion-free water, the ET-1 was eluted with 3 ml of 87% methanol. The flow rate for the conditioning and loading of the cartridge was 5 ml/min and for loading and eluting of the sample was 0.2 ml/min. The eluent was dried in a centrifugal evaporator (Speed Vac Systems AES 1010, Savant Instruments, Inc., Holbrook, NY) and reconstituted in 0.25 ml of sample diluent (Parameter Human Endothelin-1 Immunoassay Kit; R&D Systems, Inc., Minneapolis, MN). After vortexing the reconstituted residue at high speed for 5 s (Maxi Mix-1; Thermoline Co., Dubuque, IA), the tubes were incubated at 4°C for 30 min and counted on a gamma counter (LKB 1270; Wallac Oy, Turku, Finland) for 5 min each. Counts are expressed after background correction. The mean recovery of ¹²⁵I-ET-1 spiked into control plasma was 88.2 ± 11.3% by the Sep-Pak procedure and did not change significantly with incubation of the spiked plasma at 37°C for 3 h (data not shown). This provided a control for the extraction procedure.

Pharmacokinetic Analysis. The plasma I-125 count concentrations were analyzed by means of multiexponential equations (Reich et al., 1972). This was done via a nonlinear computer fitting program as originally described by D'Argenio and Schumitsky (1979). Optimal fits were obtained using procedures suggested by Motulsky and Ransnas (1987). The simplest statistically adequate equation was identified as that which yielded a minimal sum of squares. Optimal/goodness of fit was assessed further via the F-test (Boxenbaum et al., 1974), the Akaike information criterion (Akaike, 1974), and the nearest neighbor criterion (Reich et al., 1972). All pharmacokinetic parameters were determined from the computer fitting or by model-independent or -dependent calculations (Gibaldi and Perrier, 1982).

Statistical Analysis. All data are presented as mean ± S.E.M. Comparisons of the effects of ¹²⁵I-ET-1 infusion on hemodynamics, circulating ET-1 levels, electrolytes, blood urea, and creatinine were made with Wilcoxon signed rank test (SigmaStat 2.0, Jandel Scientific Software Corp., San Rafael, CA). A P value of < 0.05 was required for statistical significance.

	Results

Top Abstract Introduction Materials and Methods Results Discusssion References

Five male volunteers, aged 26 ± 2 years, height 173 ± 2 cm, and weight 66.2 ± 2.0 kg were studied.

Hemodynamic Effects of ¹²⁵I-ET-1 Infusion Although hemodynamics were recorded at all time points, only data obtained at 0, 5, 30, and 240 min hemodynamics are presented (Table 1). The ¹²⁵I-ET-1 infusion caused no change in any hemodynamic parameter.

Plasma ET-1 Level. Plasma ET-1 levels measured at 0, 3, 5, 16, 60, and 240 min did not change during the ¹²⁵I-ET-1 infusion (Table 2).

ET-1 Kinetics. The composite plasma ¹²⁵I-ET-1 counts concentration versus time profile is presented in Fig. 1. There was an initial rapid rise in counts during the 5-min infusion. Upon completion of the infusion, counts initially fell rapidly and then more gradually over time. The plasma counts concentration versus time profiles after the ¹²⁵I-ET-1 infusion were not adequately described by a single exponential disappearance function. Rather, it appeared that a multiexponential function best described the data. Using the combined data, computer fitting was used to identify the most appropriate relationship. As seen in Fig. 1, a biexponential function, reflecting a two-compartment model, also appeared to be unsatisfactory. The nearest neighbor residual test indicated a systematic, statistically significant (P < .01) deviation that pointed to the requirement for an additional exponential. As shown, a triexponential function, associated with a three-compartment pharmacokinetic model, was the simplest mathematical expression that adequately described the data set. Detailed pharmacokinetic parameters for the three-compartment model are presented in Table 3.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Composite data for all subjects, concentration of counts [(Cpm)/l] versus time from start of infusion at 0 min. The infusion was terminated at 5 min. Computer fits, using unity weighting, via the two-compartment model and the three-compartment model.

	Discusssion

Top Abstract Introduction Materials and Methods Results Discusssion References

This is the first detailed total body kinetic analysis of the fate of exogenous ¹²⁵I-ET-1 in animals or humans. The extent and length of sampling has uncovered new findings that considerably alter the understanding of both the distribution and elimination of this peptide. The kinetics of ET-1 clearance are best described by a three-compartment model with a large steady-state volume of distribution (34.9 ± 4.5 l/kg) and a prolonged terminal half-life (455 ± 59 min). Previous estimates of the half-life of ET-1 were much shorter (Vierhapper et al., 1990; Weitzberg et al., 1991) and appear to reflect only the  and  phases of the overall disappearance profile that we have described (Table 3). The overall profile is important for three fundamental reasons. First, it is the basis for an accurate estimate of ET-1 clearance. Failure to obtain the entire profile underestimates the total area under the curve and consequently overestimates clearance. Such information shapes our understanding of the efficiency of clearance processes and sets the stage for a comparison of disease states in which clearance may be altered. Second, it enables a more accurate assessment of equilibrium or steady-state distribution volume. This provides a better estimate of tissue uptake and could clarify the fate of ET-1 in different patient groups. Third, it provides accurate kinetic data necessary for any attempt to attain steady-state concentrations of ET-1 and for the interpretations of the dynamics of physiologic change (Stewart et al., 1992). For example, exogenous ET-1 infusions will not reach steady-state concentrations in a matter of minutes, as might have been assumed based on previous reports. Rather, continuous infusions will ultimately approach steady state according to the terminal () half-life.

Using compartmental models is a convenient way of expressing multiexponential profiles and defining zones of equilibration. It appears that ET-1 equilibrates in three such zones. The plasma, associated with the sampling compartment (compartment one), is part of the early phase. Compartments two and three reflect slower or more delayed equilibration. They contribute strongly to the estimates of steady-state distribution space. For ET-1, we describe a much larger distribution space than previously reported data would suggest (Lerman et al., 1991b; Weitzberg et al., 1991). The observed steady-state volume of distribution (34.9 l/kg or 2429 l/70 kg normal human) suggests that ET-1 must be extensively taken up into cells throughout the body, and that relatively little peptide is found in the initial zone, which includes the plasma. ET-1 binding to the ET_B receptor and internalization of ET-1 have been demonstrated in many tissue beds (Hirata et al., 1988; Wagner et al., 1992; Fukuroda et al., 1994; Dupuis et al., 1996a,b). It is likely that the large volume of distribution represents binding and internalization ET-1 into cells of various organ systems.

Among the pharmacokinetic information presented (Table 3), we have also calculated values for mass transfer (k₂₁ and k₃₁) from slower equilibrating zones back to the initial space. As the peptide is being cleared from the body, the concentrations in zone I drop; this is followed by a return of ET-1, as defined by k₂₁ and k₃₁. This return is comparatively slow and has a sizeable impact on the so-called "terminal" half-life. We have shown this half-life to be at least 15 times greater than previously estimated (Vierhapper et al., 1990; Weitzberg et al., 1991).

The total plasma clearance of ET-1 in the average range of 65 ml/kg/min is high, given that normal cardiac output is about 85 ml/kg/min. This may be a reflection of the fact that the lung is a major site of ET-1 clearance, able to remove over 40% of the ET-1 presented to it (Dupuis et al., 1996a,b). Ordinarily such a high clearance rate would result in a comparatively rapid disappearance of the peptide from the body. However, due to a high degree of tissue uptake, as defined by the steady-state volume of distribution, the terminal half-life is comparatively long.

Several lines of evidence suggest that the prolonged terminal half-life that we have observed does represent persistent, circulating ET-1. First, all plasma samples were passed through a Sep-Pak C₁₈ column before the counting of ¹²⁵I. This solid-phase extraction would separate hydrophobic molecules like ET-1 from ionic particles like free ¹²⁵I that may result from cleavage and degradation of ¹²⁵I-ET-1. Second, anti-ET-1 antibody was able to identify plasma ¹²⁵I as ¹²⁵I-ET-1 in a similar ratio to control stock ¹²⁵I-ET-1 identification. Third, incubation of ¹²⁵I-ET-1 in human plasma resulted in no loss of recovery of ¹²⁵I-ET-1 with Sep-Pak extraction, suggesting that ¹²⁵I-ET-1 has insignificant degradation in plasma. Fourth, our prolonged half-life is consistent with the persistent elevation in ET-1 levels above baseline for many hours in studies where unlabeled ET-1 was infused and the failure to achieve steady-state plasma concentrations during prolonged infusions of unlabeled ET-1 (Vierhapper et al., 1990; Weitzberg et al., 1991; Wagner et al., 1992; Pernow et al., 1996). Finally, these results are also consistent with the prolonged vasoconstrictor properties of ET-1 (Vierhapper et al., 1990; Lerman et al., 1991b; Pernow et al., 1991; Weitzberg et al., 1991).

An important difference in the present protocol from previous human studies (Vierhapper et al., 1990; Weitzberg et al., 1991) is that we used significantly smaller doses of ET-1, which avoided appreciable effects on hemodynamics (Table 1). In the present study the highest dose delivered was only 0.2 ng/kg/min, 12-fold lower than the lowest dose demonstrated to have any hemodynamic effects (Vierhapper et al., 1990). The observed lack of hemodynamic effects is important because previous studies have shown that changes in hemodynamics can alter circulating ET-1 levels (Shichiri et al., 1990; Stewart et al., 1992; Mangieri et al., 1997). The mechanism of changes in circulating ET-1 levels with these hemodynamic stressors remain undefined but alterations in clearance kinetics cannot be excluded (Stewart et al., 1992).

The present study also avoided detectable changes in circulating ET-1 levels (Table 2). It has been suggested that higher infusion rates of ET-1 may result in changes in the proportional contribution of different organ systems to overall ET-1 clearance (Wagner et al., 1992). Sirvio et al. (1990) have demonstrated that at higher doses, a greater proportion of ET-1 is cleared through the kidney with a reduction in lung clearance. Thus, techniques using tracer doses, avoiding detectable changes in circulating levels, may provide more accurate estimates of the pharmacokinetics of endogenous ET-1.

Our methodology also consisted of more frequent sampling over longer time periods than previous studies (Vierhapper et al., 1990; Weitzberg et al., 1991), permitting us to precisely define the clearance rates in a three-compartment model and to appreciate the actual duration of the terminal half-life. Weitzberg et al. (1991) have noted that ET-1 may have two half-lives of 1.4 and 35 min, similar to the half-lives we would have seen by using only a two-compartment model to our data (Fig. 1). As seen in Fig. 1, a two-compartment model fails to explain the slow decline of ¹²⁵I-ET-1 counts beyond 60 min and fails to describe the actual terminal half-life. In their study, Weitzberg et al. (1991) note that ET-1 levels remained significantly elevated above baseline even 180 min after the end of their infusion, consistent with our finding that ET-1 actually has a terminal half-life far in excess of the 35 min they report. Animal studies of ET-1 clearance using radiotracer techniques of ¹²⁵I-ET-1 infusion have demonstrated a multicompartment kinetic model (Anggard et al., 1989; Shiba et al., 1989) with a prolonged terminal half-life (Anggard et al., 1989), consistent with our findings in humans.

Limitations of our study should be considered. We have used a radiotracer infusion technique where ¹²⁵I-ET-1, as opposed to synthetic unlabeled ET-1, was infused. The ¹²⁵I is incorporated at the ¹³tyrosine site of the ET-1 molecule. Although it is possible that substitution of ¹²⁵I for a hydrogen molecule normally present at that site may alter the stereochemistry of the molecule, a study by Anggard et al. (1989) and data available from the manufacturer (data on file, New England Nuclear, Wilmington, DE) show that both the labeled and unlabeled species display the same chromatographic mobilities on HPLC. ¹²⁵I-ET-1 has the same biological activity and displays similar binding affinities for antibody sites (Anggard et al., 1989). As well, unlabeled ET-1 is able to displace ¹²⁵I -ET-1 from ET-1-specific antisera in similar concentrations (Anggard et al., 1989), all of which suggest that using ¹²⁵I-ET-1 to determine the kinetics of endogenous ET-1 is reasonable.