Technical note: Effects of pegylation and route of administration on leptin kinetics in newborn lambs

Abstract

Chronic energy insufficiency resulting from inadequate feed intake or increased nutrient demand reduces plasma leptin in ruminants. Treatment of energy-deficient ruminants with exogenous leptin has identified some physiological consequences of reduced plasma leptin, but their full complement remains unknown. Additional leptin-dependent responses could be identified by using strategies that interfere with leptin signaling such as administration of leptin mutants that act as competitive antagonists. The effectiveness of these antagonists depends on their fold excess over endogenous leptin, and this condition can be achieved under in vivo conditions by extending the half-life (t_1/2) of the antagonist by addition of a polyethylene glycol (PEG) molecule (pegylation). Use of this approach in ruminants, however, is limited by the lack of information on the t_1/2 of native and pegylated leptin and on the most effective route of administration. To answer these questions, newborn lambs (n = 3) were injected with an intravenous (i.v.) bolus of 150 µg of human leptin followed by blood sampling over the next 12 h. Analysis of the semilog plasma leptin concentration over time yielded a t_1/2 of 43 ± 4.9 min; an i.v. bolus of 276 µg of bovine leptin yielded a comparable t_1/2 (P > 0.05). Next, newborn lambs (n = 4) received a single dose of 229 μg/kg of metabolic body weight (BW^0.75) of pegylated super human leptin antagonist (PEG-SHLA) via the i.v. or subcutaneous (s.c.) route. Plasma PEG-SHLA concentration reached a peak of 1,528 ± 78 ng/mL after 1 min and a nadir of 71 ± 9 ng/mL after 24 h with the i.v. route versus a peak of 423 ± 43 ng/mL after 300 min and a nadir of 146 ± 22 ng/mL after 24 h for the s.c. route; the t_1/2 of PEG-SHLA was 394 ± 29 min for the i.v. route and 433 ± 58 min for the s.c. route. Finally, plasma concentration of PEG-SHLA was modeled when given either i.v. or s.c. at a dose of 229 μg/kg BW^0.75 every 12 h. Once a steady state was reached, peak and lowest concentrations PEG-SHLA over the 12-h windows were 2,269 and 403 ng/mL for the i.v. route and 814 and 555 ng/mL for the s.c. route. Weighted PEG-SHLA concentrations over the 12-h period were 1,455 and 713 ng/mL for the i.v. and s.c. route, translating into 364- and 178-fold excess over endogenous plasma leptin. These data confirm the effectiveness of pegylation in extending the t_1/2 of leptin antagonists in newborn lambs and in increasing their circulation in fold excess over endogenous leptin.

Introduction

The discovery of the Ob gene in 1994 revolutionized our understanding of central mechanisms regulating energy metabolism. The mouse Ob gene is transcribed nearly exclusively in white adipose tissue leading to the secretion of the hormone leptin (Myers and Leibel, 2000; Friedman, 2014). This was followed by the identification of the db gene encoding 2 major receptor isoforms known as Ob-Rb and Ob-Ra (Myers and Leibel, 2000; Allison and Myers, 2014). The receptor Ob-Rb is found in high density almost exclusively in the hypothalamic nuclei which control feed intake and metabolism, and it is the isoform capable of activating all signaling pathways attributed to leptin. In contrast, Ob-Ra accounts for nearly all leptin receptors in peripheral tissues and has little signaling activity. Studies of mouse strains harboring inactivating mutations of these genes revealed the essential role played by leptin in regulating feed intake, body composition, and energy-dependent functions such as reproduction and adaptive metabolism (Myers and Leibel, 2000; Allison and Myers, 2014).

Many elements of the leptin system are conserved in ruminants, including restricted spatial expression of leptin and its receptor isoforms (Boisclair et al., 2006; Thorn et al., 2007). Nevertheless, the physiological roles of leptin in ruminants remain ill-defined, due in part, to the absence of animals with inactivating ob or db mutations. So far, virtually all functional studies have involved exogenous leptin administration to animals with an intact leptin system (e.g., central leptin delivery through intra-cerebroventricular cannula or peripheral administration of gram amounts of leptin) (Henry et al., 1999; Morrison et al., 2002; Reicher et al., 2012). Plasma leptin is reduced in ruminants suffering from chronic energy insufficiency, and exogenous leptin treatment of these animals identified responses such as stimulation of LH and thyroid hormones secretion (Maciel et al., 2004; Ehrhardt et al., 2016). Additional leptin-dependent responses could be identified by using strategies that interfere with leptin signaling. This strategy is appealing because leptin is predominantly a signal of energy insufficiency with the corollary that most physiological responses are triggered by falling rather than increasing plasma leptin (Ravussin et al., 2014; Rosenbaum and Leibel, 2014).

In this context, a significant advance is our development of leptin variants with antagonistic properties (Niv-Spector et al., 2012; Gertler and Elinav, 2014). These antagonists bind the leptin receptor with equal or increased avidity but are completely unable to trigger signaling owing to alanine substitution mutations of amino acid residues 39 to 41. They act as competitive inhibitors of naturally occurring leptin, so accordingly, their effectiveness depends on their fold excess over endogenous leptin. Fold-excess conditions can be achieved under in vivo conditions by extending the half-life (t_1/2) of antagonists through the addition of a polyethylene glycol (PEG) molecule (hereafter referred to as pegylation) (Shpilman et al., 2011; Gertler and Elinav, 2014). Our long-term goal is to use these pegylated antagonists in sheep to identify the physiological roles of leptin on productive functions such as growth, body composition, and lactation. At the initiation of this work, however, pegylated versions of ovine leptin were not available. Accordingly, we used the pegylated version of the super human leptin antagonist (PEG-SHLA) to assess the effects of pegylation and route of administration on the pharmacokinetics of leptin in newborn sheep.

MATERIALS AND METHODS

Animals and Management

All experimental procedures were approved by the Cornell University Institutional Animal Care and Use Committee and followed care and use standards outlined in the FASS Guide for Care and Use of Agricultural Animals in Research and Teaching. Male Finn × Dorset lambs were collected at birth and fed 60 g/kg body weight (BW) artificial colostrum (Land O’Lakes, Arden Hills, MN). Animals were then transported to the Cornell University Large Animal Research and Teaching Unit and housed in individual steel cages (75 cm width × 80 cm length × 80 cm height) at a constant temperature (25 to 27 °C) and photoperiod (lights on between 0700 and 1900 h). Lambs were reared on a milk replacer fed at 18.6% DM and containing 253 g protein, 302 g fat, and 24.6 MJ of gross energy per kg DM. The milk replacer was served twice daily and available in unlimited amounts at all times.

Experimental Design

The first experiment was used to estimate the t_1/2 of human leptin in newborn lambs and to determine whether this estimate differed from that of bovine leptin which is 99.4% identical to ovine leptin. The experiment involved 3 lambs consuming 217 ± 14 g/d DM intake and gaining 258 ± 25 g/d on average over the first 10 d of life. Lambs were allocated to a crossover design with treatments administered on day 10 or 13 of postnatal age when they weighed 6.2 ± 0.2 and 6.6 ± 0.1 kg, respectively. Treatments consisted of a single intravenous (i.v.) bolus administration of either human leptin (150 µg; National Hormone and Peptide Program, Torrance, CA) or bovine leptin (276 µg; Protein Laboratories Rehovot Ltd, Rehovot, Israel). Experimental procedures were as follows: on day 9, each lamb was fitted with a single intrajugular catheter filled with heparinized phosphate-buffered saline (PBS, 100 U heparin/mL for maintenance and 10 U/mL during sampling). Recombinant hormones were prepared as 100 µg/mL solutions in 0.4% NaHCO₃ as recommended by the suppliers. The solution was brought up to 5 mL with PBS and administered as a bolus, followed by a 10 mL PBS flush in less than 15 s. Blood samples (2 mL each) were taken at fixed times relative to bolus (−15, −5, 1, 5, 10, 20, 40, 60, 90, and 120 min), immediately mixed with heparin (20 U/mL), and centrifuged at 3,000 × g for 15 min at 4 °C. Resulting plasma was stored at −20 °C until analyzed for plasma leptin.

A second experiment was performed to assess the effects of pegylation and route of administration on leptin kinetics. This involved a second group of 4 lambs consuming 221 ± 11 g/d DM intake and gaining 252 ± 26 g/d on average over the first 10 d of life. They were fitted with a single intrajugular catheter on day 9 of postnatal age and randomly allocated to a crossover design with treatments consisting of i.v. or subcutaneous (s.c.) administration of the pegylated version of the super human leptin antagonist (PEG-SHLA; Protein Laboratories Rehovot Ltd). Both treatments were administered at the dose of 229 μg/kg of metabolic body weight (BW^0.75) on day 10 or 13 of postnatal life when lambs weighed 5.7 ± 0.6 and 6.7 ± 0.7 kg, respectively. The leptin variant SHLA has the same sequence as human leptin except for a leucine substitution mutation of aspartic acid 23 (D23L) and alanine substitution mutations of amino acid residues 39 to 41 (L39A/D40A/F41A) (Shpilman et al., 2011). The leptin variant SHLA was produced in a bacterial expression system, refolded, purified to homogeneity, and converted to PEG-SHLA by attaching the linear reagent methoxy-polyethylene glycol-propionylaldehyde-20 kDa at its N terminus (Shpilman et al., 2011). The s.c. and i.v. doses were prepared from a solution of PEG-SHLA dissolved in 0.4% NaHCO₃ and brought up to a final volume of 5 mL with PBS. The s.c. dose was injected in the subscapular region, whereas the i.v. dose was administered as a bolus followed by a 10 mL PBS flush in less than 15 s. Blood samples (2 mL each) were taken on each treatment day at fixed times relative to PEG-SHLA administration (−15, −5, 1, 10, 30, 60, 180, 300, 420, 540, 720, 1,080, and 1,440 min), and plasma was prepared as described above.

Analysis of Plasma Leptin

Plasma concentrations of bovine and human leptin were measured with an assay capable of recognizing ovine, bovine, and human leptin (Multi-Species Leptin RIA, Millipore Inc., Billerica, MA) (Delavaud et al., 2000; Delavaud et al., 2002; Smith et al., 2018). This RIA uses human leptin as tracer and standards and has cross-reactivity with bovine leptin. The assay was performed with an additional set of standards prepared from recombinant bovine leptin (Protein Laboratories Rehovot Ltd). The plasma concentrations of human and bovine leptin were then calculated from the homologous standard curve. Sensitivity of the Multi-Species Leptin assay was 1.56 ng/mL when using the human leptin standards and 4 ng/mL when using the bovine leptin standards. The plasma concentration of PEG-SHLA was measured using a human leptin RIA that does not recognize ovine leptin (Millipore Inc.). The assay was performed exactly as recommended by the supplier except that standards were prepared with PEG-SHLA. Sensitivity of the human leptin RIA was 2 ng/mL when using the PEG-SHLA standards. For both assays, samples were diluted with assay buffer when necessary. Intra-assay coefficients of variation for both assays were <9%.

Analysis of Leptin Kinetics

For the experiment comparing the kinetics of human and bovine leptin (Exp. 1), the concentration of post-bolus samples was corrected for by subtracting plasma leptin concentration detected in basal samples by the Multi-Species Leptin RIA. These basal leptin values were determined using the human standard curve for the human leptin bolus and the bovine leptin standard curve for the bovine leptin bolus. For the experiment comparing routes of administration of PEG-SHLA (Exp. 2), a correction corresponding to PEG-SHLA present in basal samples was applied to samples obtained on day 13; such a correction was not necessary on day 10 because the human RIA does not recognize ovine leptin.

Plasma leptin concentrations versus time data for individual lambs were plotted on linear and semilogarithmic graphs for analysis. Data were subjected to noncompartmental pharmacokinetic analysis using commercially available software (PK Solutions 2.0, Summit Research Services, Montrose, CO). The initial hormone concentration (C_i) was estimated through linear extrapolation of the first 2 plasma concentration values to time zero. Area under the concentration-time curve was estimated by the trapezoidal method up to the last measured concentration. The area was extrapolated to infinity (AUC_0-∞) by addition of the last measured concentration divided by the apparent terminal disposition rate constant (λ_z), as determined by linear regression analysis of the terminal portion of the log plasma concentration-time curve. Elimination half-life (t_1/2) was calculated from the terminal disposition rate constant as ln(2)/λ _z. The volume of distribution (Vd) was estimated as D/AUC_0-∞·λ _z where D is the ratio of dose to body weight. Clearance rate (Cl) for the i.v. bolus was estimated as D/AUC_0-∞. Mean residence time (mRT) was estimated as AUMC_0-∞C/AUC_0-∞ where AUMC_0-∞ is the area under the concentration·time-time curve extrapolated to infinity.

Additional parameters specific to the s.c. injection were obtained as follows: maximum concentration (C_max) and time for maximum concentration (T_max) after s.c. injection were recorded as observed. The apparent absorption rate constant (λ_a) was determined by linear regression analysis of the log plasma concentration-time curve between the times of initial appearance and peak plasma concentration. Absorption half-life (t_1/2) was calculated from the apparent absorption rate constant as ln(2)/λ _a. Bioavailability after subcutaneous injection (F) was calculated as F (%) = (AUC_0-∞)_s.c./(AUC_0-∞)_i.v.·Dose_i.v./Dose_s.c.·100. Clearance rate (Cl) for the s.c bolus was estimated as F·D/AUC_0-∞.

The single-dose pharmacokinetic data were used to simulate predicted steady-state plasma pegylated leptin concentrations following repeated i.v. or s.c. administration. The multiple-dose pharmacokinetic program operates as an extension of the single-dose pharmacokinetic data. It draws on the pharmacokinetic results calculated for single-dose information (e.g., elimination half-life) and combines this with user input of a multiple dosing interval to compute and graph estimates of multiple-dose pharmacokinetic parameters. For any selected time, it assumes that the dosing interval is regular and equal, that repeated doses are administered during the post-distributive or elimination phase, and that the drug behavior is characterized by linear pharmacokinetics.

Statistical Methods

Kinetic parameters obtained with human and bovine leptin were analyzed by ANOVA using the mixed procedure of SAS (SAS Institute, Raleigh, NC). The model accounted for species (human versus bovine leptin) as the fixed effect and lamb as the random effect. Statistical significance was set at P < 0.05. Disposition data and PEG-SHLA kinetics are presented as means ± SE.

RESULTS

Experiment 1: Kinetics of Leptin in Newborn Lambs

Disposition characteristics of leptin were estimated by injecting an i.v. bolus of 150 µg of human leptin into newborn lambs. The semilog representation of the mean plasma leptin concentration over time is shown in Fig. 1. The estimated concentration of human leptin at time 0 (C_i) was 197 ± 24 ng/mL and remained detectable in all lambs for 120 min. Analysis of the concentration curve indicated an elimination pool with a t_1/2 of 43 ± 4.9 min. Other relevant parameters are shown in Table 1.

Figure 1.

Disposition of leptin in newborn lambs. Lambs received a single intravenous (i.v.) bolus of human (150 µg) or bovine (276 µg) on day 10 or 13 of postnatal life. Blood samples were collected under basal condition at −15 and −5 min before bolus and at 1, 5, 10, 20, 40, 60, 90, and 120 min after bolus. Samples were analyzed by a leptin assay recognizing human and bovine leptin. The leptin concentration of post-bolus samples was corrected for by subtracting basal plasma leptin concentration (1.7 ± 0.2 ng/mL for the human leptin bolus and 4.4 ± 1.0 ng/mL for the bovine leptin bolus). Each point represents the mean ± SEM of 3 lambs.

Table 1.

Pharmacokinetic parameters of human and bovine leptin in newborn lambs

	Leptin source1
Parameter2	Human	Bovine	SE	P-value3
C i, ng/mL	197	333	24	0.02
Vd, mL/kg	397	416	40	NS
AUC0-∞, ng·h/mL	3,835	4,885	257	0.04
t1/2, min	43	32	4.9	NS
mRT, min	46	31	6.4	NS
Cl, mL/min/kg	6.3	8.7	0.4	0.01

¹Lambs received a single intravenous (i.v.) bolus of either human (150 µg) or bovine leptin (276 µg) on day 10 or 13 of postnatal life (n = 3).

² C _i, initial concentration; Vd, volume of distribution; AUC_0-∞, area under the concentration-time curve extrapolated to infinity; t_1/2, elimination half-time; mRT, mean residence time; Cl, clearance rate.

³Type I error probability for the effect of leptin source. NS = P > 0.05.

Next, we repeated this analysis with bovine leptin (Fig. 1). Estimated C_i and AUC_0-∞ were significantly greater for bovine than human leptin (Table 1, P < 0.05), reflecting administration of a ~50% larger dose of bovine leptin. More importantly, most dose-independent parameters obtained with bovine leptin did not differ significantly from those derived with human leptin, including an estimated t_1/2 of 32 ± 4.9 min from the elimination pool. The single exception was a greater clearance rate for bovine than human leptin (P < 0.05).

Experiment 2: Effects of Pegylation on Leptin Kinetics

A second group of lambs was used to assess the impact of pegylation on leptin kinetics. This was done by bolus administration of the pegylated version of the super human leptin antagonist (PEG-SHLA) via an i.v. or s.c. route. The semilog representation of the mean plasma PEG-SHLA concentration over time for both modes of injection is depicted in Fig. 2. After i.v. injection, the plasma concentration was 1,528 ± 78 ng/mL after 1 min and remained over 71 ± 9 ng/mL after 24 h (Fig. 2). Analysis of the concentration-time curve indicated that PEG-SHLA distributed into a single pool with a volume of 114 ± 6 mL/kg of body weight and turned over with a t_1/2 of 394 ± 29 min, a 9.2-fold increase over the a t_1/2 of unpegylated human leptin in the vascular compartment (Table 2).

Figure 2.

Disposition of pegylated super human leptin antagonist (PEG-SHLA) in newborn lambs. Lambs received a single intravenous (i.v.) or subcutaneous (s.c.) bolus of PEG-SHLA [229 μg/kg of metabolic body weight (BW^0.75)] on day 10 or 13 of postnatal life. Blood samples were collected under basal conditions at −15 and −5 min before bolus and at 1, 10, 30, 60, 180, 300, 420, 540, 720, 1,080, and 1,440 min after bolus. Samples were analyzed for the concentration of PEG-SHLA by an assay unable to recognize ovine leptin. The leptin concentration of post-bolus samples collected on day 13 was corrected for by subtracting basal plasma leptin concentration (4.0 ± 0.5 ng/mL). Each point represents the mean ± SEM of 4 lambs.

Table 2.

Pharmacokinetic parameters of pegylated super human leptin antagonist (PEG-SHLA) in newborn lambs

	Mode of administration1
Parameter2	i.v. administration	s.c. administration
C i, ng/mL	1,528 ± 78	−
C max, ng/mL	−	423 ± 43
T max, min	−	300 ± 69
Vd, mL/kg	114 ± 6	−
AUC0-∞, ng·h/mL	12,305 ± 533	8,314 ± 952
F, %	−	68
Absorption t1/2, min	−	173 ± 10
Elimination t1/2, min	394 ± 29	433 ± 58
mRT, min	524 ± 27	924 ± 84
Cl, mL/min/kg	0.20 ± 0.01	0.30 ± 0.03

¹Lambs received a single intravenous (i.v.) or subcutaneous (s.c.) bolus of PEG-SHLA [229 μg/kg of metabolic body weight (BW^0.75)] on day 10 or 13 of postnatal life (n = 4).

² C _i, initial concentration; C_max, maximum concentration; T_max, time for maximum concentration; Vd, volume of distribution; AUC_0-∞, area under the concentration-time curve extrapolated to infinity; F, bioavailability after s.c. injection; t_1/2, half-life; mRT, mean residence time; Cl, clearance rate.

The same lambs also received an identical s.c. dose of PEG-SHLA. With this mode of administration, PEG-SHLA was only 26 ± 7 ng/mL 1 min after injection, rose to a peak concentration of 423 ± 43 ng/mL at 300 min, and remained at or near this maximal concentration until 720 min post-bolus (Fig. 2). Plasma concentration of PEG-SHLA declined to 146 ± 22 ng/mL 1,440 min after bolus. After s.c. injection, bioavailability of PEG-SHLA was 68%, and average absorption and elimination t_1/2 were 173 ± 10 min and 433 ± 58 min, respectively (Table 2). When compared to the i.v. route, the s.c. injection led to lower plasma PEG-SHLA concentrations except for the last 6 h of the 24-h period of measurement. A side-by-side comparison of other kinetics parameters obtained with i.v. and s.c. injections is provided in Table 2. Lambs did not show any noticeable changes in appetite or behavior on the day of i.v. and s.c. injections or during the resting period between injections.

Modeling Plasma PEG-SHLA Kinetics During Chronic s.c. or i.v. Injection

An important factor determining PEG-SHLA effectiveness is its fold concentration excess over endogenous leptin. Accordingly, we use the kinetic parameters derived above to model plasma leptin concentration in newborn lambs when PEG-SHLA is given either i.v. or s.c. at the dose of 229 μg/kg BW^0.75 every 12 h. The profile of the plasma antagonist reached a steady state after 48 h for both modes of administration. When this steady pattern is reached, the peak and lowest concentrations of the antagonist over a 12-h period were 2,269 and 403 ng/mL for i.v. administration and 814 and 555 ng/mL for s.c. administration (Fig. 3). Overall, s.c. delivery resulted in a more even plasma PEG-SHLA profile than i.v. injection, but concentrations were always lower than those achieved with i.v. injection except for the last hour of each 12-h period. Average plasma concentrations calculated from integration over the 12-h period were 1,455 and 713 ng/mL for the i.v. and s.c. administration, respectively.

Figure 3.

Predicted plasma steady state of pegylated super human leptin antagonist (PEG-SHLA) concentration in newborn lambs. Steady-state plasma pegylated leptin concentrations were simulated for multiple doses of PEG-SHLA [229 μg/kg of metabolic body weight (BW^0.75) every 12 h] by inputting the intravenous (i.v.) or subcutaneous (s.c.) single-dose pharmacokinetic data into the modeling function of PK Solutions 2.0.

Discussion

We know little about the physiological role of leptin in ruminants. Previous studies have identified some responses when leptin was administered to animals experiencing hypoleptinemia as a consequence of energy insufficiency (Maciel et al., 2004; Ehrhardt et al., 2016). Use of exogenous administration, however, has proven ineffective in revealing other expected responses such as reduced feed intake (Morrison et al., 2002; Ehrhardt et al., 2016). In fact, the latter was observed only when leptin was administered via intra-cerebroventricular infusion, an approach raising hypothalamic leptin concentrations 2–3 orders of magnitude above physiological levels (Henry et al., 1999; Foskolos et al., 2015). We now understand that this response is a reflection of the unusual manner in which leptin signals, i.e., falling plasma leptin triggers most of the biologically relevant responses, whereas increasing leptin above normal levels has little or no effect (Ravussin et al., 2014; Rosenbaum and Leibel, 2014). This implies that approaches which antagonize leptin signaling may be required to uncover the full range of leptin actions in intact animals. In this context, we identified a stretch of 3 amino acid residues that are both conserved across species and required for leptin receptor activation (Niv-Spector et al., 2005). These amino acid residues, however, play no role in receptor binding such that their mutation effectively turns mouse, human, and ovine leptin into competitive antagonists. More recently, we have extended the half-life of these antagonists with the addition of a PEG moiety, therefore improving their in vivo potency (Elinav et al., 2009; Shpilman et al., 2011). One factor limiting the use of these reagents in ruminants is the lack of information on the t_1/2 of native and PEG-modified leptin in ruminants.

The first experiment involved an i.v. bolus of either human leptin or bovine leptin for 2 reasons. First, we wanted to compare the effect of pegylation on the kinetics of leptin derived from the same species. Because we used a pegylated human leptin variant, we sought to obtain the t_1/2 of unmodified human leptin. After an initial distribution phase of approximately 4–5 min, human leptin was eliminated with a t_1/2 of 43 min. Second, mature human and ovine leptin differ in 19 of 146 amino acid residues, raising the possibility that human leptin did not provide accurate estimates of ovine leptin kinetics. Accordingly, we also administered a bolus of bovine leptin which is nearly identical to ovine leptin (i.e., 1 difference out of 146 amino acid residues) and obtained a t_1/2 of 32 min. Comparable estimates were expected for both bovine and human leptin because kidney filtration is the major factor driving elimination of proteins < 70 kDa, and both have a calculated molecular weight of 16 kDa (Morath et al., 2015). This result is supported by work in mice where comparable t_1/2 were obtained for human leptin (30 min) and mouse leptin (24 to 40 min) (Klein et al., 1996; Morath et al., 2015; Hart et al., 2016; Burnett et al., 2017). We also obtained a t_1/2 of 30 min in early lactating dairy cows by modeling the rising portion of plasma leptin during a constant i.v. infusion of human leptin (Ehrhardt et al., 2016).

A number of approaches have been developed to extend the t_1/2 of small proteins including genetic fusion with a stabilizing protein such as the immunoglobulin Fc region and conjugation with PEG polymers (Pasut and Veronese, 2012). These methods work by increasing the hydrodynamic molecular volume of modified proteins beyond the pore size of the glomerular filtration apparatus. Our choice of pegylation to extend the t_1/2 of leptin antagonists was based on its well-established chemistry and use for more than 10 approved biopharmaceuticals (Pasut and Veronese, 2012). As a test case, we used the pegylated version of the antagonist SHLA differing from native human leptin in only 4 amino acid residues (D23L/L39A/D40A/F41A). When delivered i.v., PEG-SHLA had a Vd value of 114 mL/kg, nearly identical to the blood volume of newborn lambs estimated at 111 mL/kg (Gotsev, 1939). This observation suggests that i.v.-administered PEG-SHLA was mostly confined to the vascular compartment where it circulated with a t_1/2 of 394 min, representing a 9-fold t_1/2 extension relative to unpegylated human leptin. Despite retention in the vascular compartment, pegylated leptin antagonists are highly effective in blocking hypothalamic Ob-Rb signaling because they competitively inhibit endogenous leptin transport through the blood brain barrier (Elinav et al., 2009). The plasma concentration after s.c. injection is a function of both appearance in and elimination from the vascular system, and not surprisingly, it yielded a slightly longer t_1/2 of 433 min or a 10-fold t_1/2 extension over that of unpegylated human leptin. This t_1/2 prolongation is similar to the 13.4-fold extension of leptin t_1/2 seen in the mouse after s.c. administration of the pegylated mouse homologue of SHLA (Elinav et al., 2009). The PEG moiety becomes the main determinant of t_1/2 when added to small proteins such as leptin, and therefore, a similar t_1/2 extension is expected for the ovine homologue of PEG-SHLA.

An important factor in determining the effectiveness of any competitive antagonist is its fold concentration excess over the targeted agonist. Using kinetic parameters obtained with PEG-SHLA, we estimated this excess over time for a dose of 229 μg/kg BW^0.75 administered every 12 h via either the i.v. or s.c. route. This dose is 50% greater on a metabolic BW basis than the dose of the mouse homologue of PEG-SHLA that produced a 120% body weight increase in mice (Shpilman et al., 2011), and therefore, it has a reasonable likelihood of being effective in newborn lambs. This simulation shows that the i.v. route produced plasma PEG-SHLA concentrations that are always greater than the s.c. route, except for the last hour of the 12-h window, whereas the s.c. route produced a more uniform concentration profile. When integrated over the 12-h window, the plasma SHLA concentration averaged 1,455 and 713 ng/mL for the i.v. and s.c. route, translating into fold excesses of 364 and 178 over the plasma leptin concentration of approximately 4.0 ng/mL prevailing in well-fed newborn lambs (Ehrhardt et al., 2003). Because maximal fold excess aids in countering the reduced receptor binding ability of PEG-leptin derivatives, the i.v. route may be preferable in domestic animals, particularly when the combination of the amount of PEG-leptin needed and cost becomes prohibitive.

In summary, we have obtained estimates for the t_1/2 of both native and pegylated leptin variants in newborn lambs and established the circulating profile of pegylated leptin when given either i.v. or s.c. These data will be useful in selecting the proper dosage and mode of administration of pegylated leptin antagonists in experiments aimed at identifying the consequences of reduced leptin signaling in ruminants.