Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Human brain natriuretic peptide (hBNP) is a 32-amino-acid, cardiac-derived peptide hormone with potent cardiovascular and renal actions (Lewicki and Protter, 1995). Studies have shown that hBNP is a vasodilator (Protter et al., 1996) that reduces cardiac filling pressures and produces a diuretic and natriuretic response (Clavell et al., 1993; Clemens et al., 1997). The biological properties of this hormone have prompted testing its potential therapeutic effects in patients with congestive heart failure and hypertension, and beneficial effects have recently been demonstrated. Intravenous administration of hBNP to patients with congestive heart failure results in a decrease in pulmonary capillary wedge pressure and an increase in cardiac output (Hobbs et al., 1996; Marcus et al., 1995; Yoshimura et al., 1991). The peptide nature of hBNP (fig. 1) has focused preclinical and clinical studies on protocols using intravenous modes of delivery that are relatively short in duration. Subcutaneous delivery of hBNP may offer considerable advantages over intravenous protocols in certain clinical settings, particularly those involving prolonged treatment. The effectiveness of subcutaneous delivery of hBNP has not been tested in animals or humans.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Primary amino acid sequence of human BNP. The peptide contains 32 amino acids (MW = 3464 g/mol), with a 17-amino-acid ring structure formed by disulfide bonds. The peptide has an isoelectric point of 10.9.

In rabbits, intravenous hBNP elevates plasma cyclic GMP (Clemens et al., 1997), consistent with activation of the membrane-bound quanylyl cyclase-A receptor for which hBNP has been shown to be a potent ligand (Schoenfeld et al., 1995). In addition, hBNP reduces blood pressure and stimulates diuresis and natriuresis when administered intravenously to anesthetized rabbits (Clemens et al., 1997). In the studies reported here, we use these end points to compare the biological actions of hBNP administered to rabbits by intravenous and subcutaneous delivery protocols. In addition, the plasma hBNP levels achieved with these two delivery protocols were estimated with an hBNP-specific immunoassay. Subcutaneous administration was found to be a surprisingly efficient method of delivering the peptide to the vascular space in a biologically active form.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Male New Zealand White rabbits (1.5-2.0 kg), purchased from R&R Rabbitry (Stanwood, WA), were housed individually for at least 1 week prior to study and allowed food and water ad libitum.

Materials. Recombinant hBNP was expressed in bacteria and purified to >95% homogeneity. The purity and identity of the peptide were assessed by reverse phase liquid chromatography, amino acid sequence analysis and amino acid composition (data not shown). Endotoxin levels, determined by a limulus amebocyte lysate assay, was less than or equal to 5 endotoxin units per milligram (data not shown). Pentobarbital was purchased from Anpro Pharmaceuticals (Arcadia, CA). Trimethylbenzidine and goat anti-murine, Fc-specific IgG were from Sigma (St. Louis, MO). Biotinylated hBNP was synthesized with a single biotin moiety at the amino-terminal position by American Peptide Company (Sunnyvale, CA) and was characterized by mass spectroscopy and amino acid sequence analysis. Plasma cyclic GMP was quantitated with a radioimmunoassay manufactured by NEN Life Sciences (Boston, MA).

Hemodynamic and renal measurements. Animals were anesthetized with pentobarbital infusion (40-60 mg/kg) via ear vein. Throughout the experiment, body temperature was maintained on a circulating heating pad at 36°C. Rabbits were administered a continuous infusion of 0.9% NaCl (6 ml/kg/hr) and pentobarbital (10 mg/kg/hr) via the marginal ear vein for the duration of the experiment. A catheter (PE90) was advanced 10 cm through the femoral artery and connected to a Grass Physiograph and a CODAS (DATAQ, Akron, OH) data capture system. Cardiovascular data were monitored continuously, and blood pressures (mean, systolic, diastolic) were expressed as the average value during 10-min periods. Urine was collected from the bladder with a 12-Fr Foley catheter in 20-min intervals. The rates of urine flow and urine sodium excretion were quantitated by weight and flame photometry (Instrumentation Laboratory, Lexington, MA), respectively. Drug treatment was either vehicle (0.9% NaCl, 1 ml/kg) or hBNP (30 µg/kg, 1 ml/kg). One group of rabbits had hBNP (n = 10) or vehicle (n = 8) delivered via a catheter placed in the left femoral vein (intravenous drug delivery protocol) while the second group of animals had hBNP (n = 12) or vehicle (n = 3) delivered by subcutaneous injection between the shoulder blades (subcutaneous protocol). There was no difference between the hemodynamic and renal responses of animals given vehicle by the two delivery protocols; therefore, the data for statistical analysis was combined.

Plasma hBNP determination. Blood samples (1 ml) for hBNP determinations were collected into EDTA-coated microcentrifuge tubes (Brinkman) containing 0.01 ml of aprotinin (1.8 mg/ml). Plasma was isolated by centrifugation and stored at 80°C. Venous blood samples were taken immediately prior to and 5, 15, 30, 60, 120 and 180 min following drug delivery in the subcutaneous delivery protocol and immediately prior to and 2, 5, 10, 15, 30, 60, 120 and 180 min following drug delivery in the intravenous delivery protocol. Plasma hBNP levels were analyzed from 6 animals each from the subcutaneous and intravenous protocols.

Plasma cyclic GMP determinations. Plasma cyclic GMP levels were determined by radioimmunoassay. The labeled antigen was a succinyl tyrosine-(¹²⁵I)-methyl ester derivative of cyclic GMP. Separation of bound cyclic GMP from free antigen was achieved by the use of a prereacted primary and secondary antibody complex. Prior to the assay, the plasma samples were extracted with ethanol, and the supernatants were evaporated to dryness in a Speed Vac concentrator (Savant Instruments, Holbrook, NY). The dried samples were reconstituted with sodium acetate buffer prepared according to the manufacturer's instructions. Plasma cyclic levels were determined by interpolation from the standard curve. Any sample with levels above the range of the assay (10 pmol/ml) was diluted appropriately and reassayed. The lowest level of detection was 0.01 pmol/ml. Interassay and intra-assay coefficients of variation were 6.8% and 10.4%, respectively.

Data analysis. Area under the plasma hBNP concentration-time curve was determined using the integrate-area function in Kaleidagraph v3.0.4 (Synergy Software, Reading, PA). Plasma hBNP values resulting from intravenous treatment were best fitted to a two-compartment model assuming drug concentrations decline biexponentially as the sum of two first-order processes using the formula: C_t = A exp^(t) = B exp^(t). Plasma hBNP values resulting from subcutaneous treatment were best fitted to a one-compartment model assuming drug concentrations decline exponentially using the formula: C_t = A exp^(t). Values for t_1/2 and t_1/2 were calculated from 0.693/ and 0.693/, respectively.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Plasma hBNP levels resulting from subcutaneous and intravenous hBNP delivery protocols. Administration of hBNP by either intravenous or subcutaneous delivery protocols resulted in significant levels of hBNP-immunoreactive material in the plasma (see fig. 2; peak values of 26.6 ± 4.3 and 2.3 ± 0.6 nmol/liter were achieved by the intravenous and subcutaneous protocols, respectively). Values for plasma hBNP-immunoreactive material prior to hBNP administration were less than 0.012 nmol/liter. The calculated area under the plasma concentration-time curve (180 min) for hBNP delivered by the intravenous and subcutaneous routes were 310 ± 20 and 187 ± 47 nmol×mins/liter, respectively. The plasma decay curves for hBNP administered by the intravenous protocol were best fitted to a two-compartment model with computed t_1/2 = 5.5 ± 0.9 min and t_1/2 = 27.4 ± 9.7 mins. Plasma hBNP-immunoreactive material following subcutaneous hBNP administration achieved a maximum level between 15 and 30 min following treatment and then declined exponentially (best fit to a one-compartment model) with a t_1/2 of 28.7 ± 2.4 min.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Immunoreactive hBNP in plasma after intravenous and subcutaneous administration of 30 µg/kg hBNP. Inset shows same data plotted on a log scale. Results are mean ± S.E.M. Data from n = 6 for each group.

Plasma cyclic GMP levels resulting from subcutaneous and intravenous hBNP delivery protocols. Bolus administration of 30 µg/kg hBNP by either intravenous or subcutaneous delivery protocols resulted in a time-related increase in plasma cyclic GMP (fig. 3). Following subcutaneous hBNP treatment, plasma cyclic GMP levels were maximally elevated within 20 min and remained elevated for 60 min. Following intravenous hBNP administration, plasma cyclic GMP values were maximally elevated within 5 min and then quickly declined. By 60 min following hBNP treatment, plasma cyclic GMP levels were higher in the subcutaneous administration group than in the intravenous group (P < .005).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Plasma cyclic GMP following 30 µg/kg hBNP treatment using subcutaneous (n = 6) and intravenous (n = 6) delivery protocols. The drug was administered at t = 0. Data are mean ± S.E.M. * P < .05 and ** P < .05 comparing values following intravenous hBNP treatment to the value at t = 0. @ P < .05 comparing values following subcutaneous hBNP treatment to the value at t = 0. # P < .005 comparing values from the intravenous group to the subcutaneous group.

Cardiovascular and renal effects in anesthetized rabbits. Treatment with 30 µg/kg hBNP delivered by either subcutaneous or intravenous protocols resulted in a significant increase in the rates of urine flow and sodium excretion (fig. 4). When delivered by the intravenous route, most of the renal response to hBNP occurred within the first 20-min collection period. When delivered by the subcutaneous route, most of the renal response to hBNP occurred during the first two 20-min collection periods (P < .05). There was no change in renal function following treatment with saline.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Rates of urine flow (A) and urine sodium (B) in rabbits treated with 30 µg/kg hBNP by intravenous (, n = 10) or subcutaneous (, n = 12) delivery protocols compared with 0.9% NaCl treated animals ( , n = 11). Drug was administered at t = 0. Data are mean ± S.E.M. * P < .05 and ** P < .01.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Change from base line of systolic (A) and diastolic (B) blood pressures in rabbits treated with 30 µg/kg hBNP by intravenous (, n = 10) or subcutaneous (, n = 12) delivery protocols compared with 0.9% NaCl-treated animals (, n = 11). Drug was administered at t = 0. Data are mean ± S.E.M. * P < .05 and ** P < .01.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This is the first report demonstrating that hBNP can be efficiently delivered by a subcutaneous route of administration and elicit significant hemodynamic and renal responses. These data suggest that subcutaneous treatment might be an effective method for delivering the peptide in human clinical trials.

Previous studies have demonstrated that intravenous administration of hBNP to rabbits results in an elevation of plasma cyclic GMP, reduced systolic blood pressure, natriuresis and diuresis (Clemens et al., 1997). The time-related increase in plasma cyclic GMP is consistent with activation of the guanlyl cyclase-A receptor. Reduced blood pressure following hBNP treatment of normotensive animals has been shown to result from reduced cardiac preload resulting in reduced cardiac output (Clavell et al., 1993). Natriuresis and diuresis following hBNP treatment is generally believed to result from increased glomerular filtration rate and reduced reabsorption of tubular sodium.

Subcutaneous administration of hBNP to normotensive rabbits resulted in reduced systolic blood pressure, similar in magnitude but prolonged in duration when compared to the effects of hBNP given intravenously. A greater reduction in systolic rather than diastolic blood pressure was seen with hBNP administered by both protocols, consistent with the preload effects of hBNP, which have been described in previous studies in dogs (Clavell et al., 1993). Subcutaneous treatment with hBNP resulted in a significant diuresis and natriuresis. While the overall magnitude of the renal effect of hBNP was similar in the two delivery protocols, the effect was more prolonged in the subcutaneous treatment group.

Significant circulating concentrations of immunoreactive-hBNP were seen in animals given the peptide subcutaneously. Area under the plasma concentration-time curves of hBNP delivered by the two protocols suggests that up to 60% of the hBNP delivered by the subcutaneous route is found in the plasma. This assumes that with the intravenous administration protocol, 100% of the hBNP was delivered to the plasma compartment, all of the immunoreactive material detected in the plasma is intact and/or biologically active and metabolism of hBNP via peptidases specific to the subcutaneous pathway does not result in the formation of a hBNP species with enhanced affinity for the antibody used in the immunoassay. As the biological response to subcutaneous hBNP was comparable in magnitude and more prolonged in duration than intravenous hBNP, it is clear that significant circulating levels of hBNP were achieved by the subcutaneous protocol.

Subcutaneous administration of hBNP resulted in plasma concentrations of hBNP-immunoreactive material 15, 30 and 60 min after treatment of 2.2 ± 1.0, 2.4 ± 0.6 and 1.7 ± 0.5 nM, respectively. As hBNP activates the rabbit GC-A receptor with an ED₅₀ of 7.2 ± 1.9 nM (A. Protter, data not shown), these circulating levels are biologically meaningful. While plasma hBNP levels 60 min following subcutaneous delivery remain significantly elevated, the plasma concentration at this time after intravenous treatment is only 0.12 ± 0.03 nM. The more rapid loss of circulating hBNP following intravenous delivery compared with subcutaneous delivery is consistent with the shorter duration of biological effects of hBNP given by the former protocol.

The effectiveness of subcutaneous delivery of hBNP demonstrated here in rabbits suggests that this mode of administration might be applied therapeutically in humans. Conflicting results of subcutaneous delivery of the structurally related peptide, synthetic human ANP, have been reported (Crozier et al., 1987; Osterode et al., 1995). Characterizing ANP delivery to the circulation by area under the curve analysis of immuno-reactive material, bioavailability estimates of 3% (Crozier et al., 1987) and 22% (Osterode et al., 1995) were reported. No significant renal or hemodynamic effects were reported following ANP subcutaneous treatment studies, although one report demonstrated that subcutaneous ANP induced a significant increase in plasma cyclic GMP, an effect consistent with activation of the biological receptor for ANP.

Bolus intravenous administration of hBNP (Hobbs et al., 1996) to patients with congestive heart failure has demonstrated beneficial hemodynamic effects, including decreased pulmonary capillary wedge pressure and increased cardiac index. Subcutaneous administration may increase the duration of hBNP's beneficial effects thereby simplifying treatment protocols. In addition, a subcutaneous delivery method might allow testing for beneficial effects of long term hBNP treatment.