Human Growth Hormone-Releasing Factor (hGRF)1–29-Albumin Bioconjugates Activate the GRF Receptor on the Anterior Pituitary in Rats: Identifi
http://www.100md.com
内分泌学杂志 2005年第7期
Research Department, ConjuChem Inc., Montreal, Quebec, Canada, H2X 3Y8
Address all correspondence and requests for reprints to: Roger Léger, Ph.D., Department of Research, ConjuChem Inc., 225 President-Kennedy Avenue, Suite 3950, Montreal, Quebec, Canada, H2X 3Y8. E-mail: leger@conjuchem.com.
Abstract
In vivo bioconjugation to the free thiol on Cys34 of serum albumin by a strategically placed reactive group on a bioactive peptide is a useful tool to extend plasma half-life. Three maleimido derivates of human GH-releasing factor (hGRF)1–29 were synthesized and bioconjugated to human serum albumin ex vivo. All three human serum albumin conjugates showed enhanced in vitro stability against dipeptidylpeptidase-IV and were bioactive in a GH secretion assay in cultured rat anterior pituitary cells. When the maleimido derivatives were individually administered sc to normal male Sprague Dawley rats, an acute secretion of GH was measured in plasma. The best compound, CJC-1295, showed a 4-fold increase in GH area under the curve over a 2-h period compared with hGRF1–29. CJC-1295, a tetrasubstituted form of hGRF1–29 with an added N-3-maleimidopropionamide derivative of lysine at the C terminus, was selected for further pharmacokinetic evaluation, where it was found to be present in plasma beyond 72 h. A Western blot analysis of the plasma of a rat injected with CJC-1295 showed the presence of a CJC-1295 immunoreactive species on the band corresponding to serum albumin, appearing after 15 min and remaining in circulation beyond 24 h. These results led to the identification of CJC-1295 as a stable and active hGRF1–29 analog with an extended plasma half-life.
Introduction
HUMAN GH-RELEASING factor (hGRF) is a 44-amino acid endocrine peptide secreted by the hypothalamus that reaches the anterior pituitary, where it binds to its receptor, setting off a cascade of signals that leads to the release of GH into the plasma (1). The hormone, unfortunately, suffers from a very short half-life in vivo (2, 3). Recently developed long-lasting derivatives of hGRF, such as polyethylene glycol derivatives, have demonstrated the ability to activate the GRF receptor of rats and pigs, leading to the secretion of GH in the plasma (4, 5). Furthermore, the GH secretion pattern remains pulsatile when these derivatives are sustained at an elevated concentration in plasma, in agreement with the results from continuous hGRF infusion experiments (6).
In vivo bioconjugation to serum albumin is a useful tool to increase the half-life of small molecules (7, 8, 9) or peptides (10, 11, 12, 13, 14) in plasma. In vivo bioconjugation occurs when a strategically placed reactive group on a bioactive peptide reacts with a nucleophilic entity found in blood or in sc interstitium to form a stable bond. The foremost nucleophile is the thiol, and its most abundant source in these fluids is Cys34 on albumin. The thiol on Cys34 reacts with a Michael acceptor, such as a maleimido derivative, leading to a new bioactive protein construct that will adopt an extended half-life due to stabilization from enzymatic degradation (11, 12, 13) or reduced elimination through the kidney (14). It therefore became logical to combine the long-lasting effect of bioconjugation with the proper GRF analog.
We initiated a research program to make a number of maleimido derivatives of hGRF1–29 and hGRF analogs and thought that the rat physiology was deemed adequate to screen the products of in vivo bioconjugation with serum albumin. However, it has never been shown that a hGRF albumin conjugate or a fusion protein can activate the rat GRF receptor. For this reason, before initiating in vivo studies, the demonstration that a hGRF1–29-albumin conjugate retains in vitro activity on cultured rat anterior pituitary cells becomes essential.
The relative potency of rat GRF, which is in the order of 3- to 6-fold more active (15), as well as a 70% homology (16), when compared with the hGRF, is also of concern and needs to be addressed early to establish the rat as a viable screening model.
Another point of interest is the demonstration that bioconjugation helps to stabilize the active peptide portion of the new construct from degradation by plasma enzymes. Dipeptidylpeptidase-IV (DPP-IV) is the primary culprit with regard to the metabolism and deactivation of GRF in plasma. It recognizes alanine and proline residues at the P1 site and, in the case of GRF1–29, causes the release of the Tyr-Ala dipeptide fragment generating the inactive GRF3–29 (17).
We report herein the results from three sets of experiments on three hGRF1–29 analogs possessing the ability to bioconjugate to serum albumin: 1) the in vitro bioactivity using a cultured rat anterior pituitary cell assay and stability in the presence of DPP-IV of preformed albumin conjugates, 2) the activation of anterior pituitaries and subsequent release of GH in rats upon in vivo bioconjugation of the hGRF1–29 analogs, and 3) comparison of the plasma pharmacokinetic profiles in rats of one of the derivatives to the native peptide hGRF1–29 amide.
Materials and Methods
Synthetic peptides
The compounds prepared for this study are shown on Fig. 1. They include the native GRF1–29 amide, CJC-1288, which has a 3-maleimidopropionic acid (MPA) unit added to the amine of an extra lysine at the C terminus of GRF1–29. When initiating these studies, it was not known whether bioconjugation to serum albumin would totally stabilize the peptide from enzymatic degradation. For this reason, a D-alanine was introduced to replace the natural L-alanine at the 2-position (11), followed by the insertion of MPA as in the previous compound leading to CJC-1293. The third tetrasubstituted analog, CJC-1295, has a D-Ala at the 2-position, a Gln at the 8-position to reduce asparagine rearrangement or amide hydrolysis to aspartic acid, an Ala at position 15 to enhance bioactivity, and a Leu at position 27 to prevent methionine oxidation along with the MPA-Lys at the C terminus (17, 18, 19, 20). The two amino acid replacements at positions 8 and 27 were essentially done to confer manufacturing stability.
FIG. 1. Molecular structures of hGRF1–29 amide and the three maleimido derivatives CJC-1288, CJC-1293, and CJC-1295, prepared by solid-phase synthesis. CJC-1288 is hGRF1–29 with an extra lysine (Lys) at the 30-position to accommodate a MPA; CJC-1293 is equivalent to CJC-1288 but with a D-alanine (D-Ala) at the 2-position; and CJC-1295 is a tetrasubstituted analog of CJC-1288 but with a D-Ala at the 2-position, a glutamine (Gln) at the 8-position, an Ala at the 15-position, and a leucine (Leu) at the 27-position.
The reference GRF1–29 amide and the three maleimido derivatives CJC-1288, CJC-1293, and CJC-1295 were synthesized using the Fmoc amino acid strategy on Ramage resin, followed by orthogonal deprotection, to attach the MPA as described previously (11, 12, 13). All derivatives were lyophilized as trifluoroacetic acid salts and had more than 95% purity by HPLC.
Human serum albumin (HSA) bioconjugates
Compounds CJC-1288, CJC-1293, and CJC-1295 were bioconjugated to HSA (Cortex Biochem, Inc., San Leandro, CA) according to the protocol previously described (11, 12, 13). The bioconjugates were weighed as lyophilized powders before use.
In vitro DPP-IV stability assay
The DPP-IV (pig) stability assay was performed in the same manner as already described (11). Briefly, test compounds GRF1–29 amide (150 μM) along with (CJC-1288)-HSA, (CJC-1293)-HSA, and (CJC-1295)-HSA conjugates (10 mg) were individually solubilized in 1 x PBS (pH 7.1) (250 μl) and added directly to a vial of DPP-IV enzyme (5 mU, porcine; Calbiochem, La Jolla, CA). The mixture was incubated at 37 C under mixing conditions, and aliquots (25 μl) were taken at 0 and 24 h and immediately frozen at –80 C. Samples were thawed just before liquid chromatography (electrospray) mass spectrometry (LC/(ES)MS) analysis. Percentages are reported as a ratio of the test compound area under the peak or abundance after deconvolution for the desired protein mass at the 24-h time point relative to a reference solution that does not contain DPP-IV but submitted to identical incubation and sampling conditions. The entire experiment was repeated twice.
Primary rat pituitary cell culture
Anterior pituitary cells were harvested and prepared as previously described (21, 22, 23). Male Sprague Dawley rats (150–300 g; Charles River Canada Inc., Quebec, Canada) were anesthetized and subsequently decapitated. Whole pituitaries were removed and placed immediately in ice-cold sterile MEM containing 0.1% BSA, 100 μg/ml streptomycin, and 100 U/ml penicillin (Invitrogen Canada Inc., Burlington, Ontario, Canada). Cells were prepared freshly within 1 h post pituitary collection.
Individual anterior pituitaries were minced, and tissue fragments were digested in 10 ml sterile medium containing 0.15% trypsin (Difco Laboratories, Inc., Detroit, MI) under gentle and continuous agitation for 45 min at 37 C. Tissue pieces were mechanically disrupted using a Pasteur pipette and incubated for an additional 15 min. Cells were then washed twice and resuspended in DMEM (Invitrogen Canada Inc., Quebec, Canada) supplemented with 0.1% BSA and antibiotics. Cells were seeded in 24-well plates at a density of 2.5 x 105 cells/50 μl medium·well in DMEM and allowed to attach for 1 h in a humidified CO2 (5%) air incubator at 37 C. After cell attachment, 1 ml fresh DMEM, supplemented with 10% horse serum (Invitrogen Canada Inc., Quebec, Canada), was added.
In vitro GH secretion assay
After 72 h of culture, primary cells were washed twice with serum free DMEM media and incubated for 1 h at 37 C. After this equilibrium period, medium was replaced with fresh medium, and cells were treated with test compound (0, 10–13, 10–9, and 10–6 M final concentrations as prepared previously) (12) or HSA (0, 3 x 10–12, 3 x 10–8, and 3 x 10–5 M) for 4 h. After incubation, supernatants were collected and centrifuged to remove nonadherent cells and were stored at –20 C until analysis. Quantitative determination of rat GH was performed by RIA using a commercial kit (Linco Research, Inc., St. Charles, MO) as described in the manufacturer’s instructions. Each analysis was replicated at least three times in duplicate. The intraassay coefficient of variation for the rat GH measurement was less than 10%.
Acute in vivo GH secretion
The acute GH secretory profile of the four test compounds was evaluated using catheterized (carotid), freely moving, 7- to 8-wk-old male Sprague Dawley rats (all rats were from Charles River Canada, Saint-Constant, Quebec, Canada). The animals received a single bolus sc injection of test compound (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg) in the dorsolumbar area (n = 7 rats per group; CJC-1293 and CJC-1295 had their own individual saline control groups of n = 7; GRF1–29 amide and CJC-1288 had a single saline control group for the two). Serial blood samples were taken up to 2 h post administration. GH plasma levels were determined using the same RIA method described above. Onset of experimentation was only between 0900 and 1100 h.
Pharmacokinetic analysis in Sprague Dawley rats
Seven- to 8-wk-old male Sprague Dawley rats (n = 4 per test compound) were administered a single bolus sc dose of either GRF1–29 amide or CJC-1295 (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg) in the dorsolumbar region. Serial blood samples were collected from each animal at predose (before injection); at 5 and 30 min; at 1, 2, 4, 8, 24, 48, and 72 h in tubes containing EDTA and DPP-IV inhibitor (Linco Research, Inc.). The samples were then centrifuged (2500 rpm for 10 min at 4 C), and the plasma was aliquoted and kept frozen until analysis.
RIA method for pharmacokinetic analysis
The labeled tracer antigen 125I-GRF1–29 amide was obtained from the Douglas Hospital Research Center (Montreal, Quebec, Canada), and rabbit anti-hGRF1–44 antibody was obtained from Accurate Chemical and Scientific Corporation (Westbury, NY). The inability of this antibody to cross-react with rat GRF has been certified by the manufacturer.
The assay was a disequilibrium RIA (24). The GRF1–29 amide and CJC-1295 calibrators were prepared by adding known concentrations of GRF1–29 amide or CJC-1295 to rat plasma containing EDTA and DPP-IV inhibitor, respectively. The rat serum albumin (RSA) conjugate thus obtained was characterized by LC/(ES)MS, and the amount of free unreacted CJC-1295 was measured (<1%). Rabbit anti-hGRF1–44 antibody was found to cross-react with hGRF1–29 amide and CJC-1295. The calibrators and test samples were incubated with the antibody overnight at 2–8 C. This incubation was followed by the addition of the tracer (125I-GRF1–29). After another overnight incubation at 2–8 C, the bound and unbound GRF1–29 amide or CJC-1295 fractions were separated by a second antibody precipitation. The bound GRF1–29 amide or CJC-1295 was collected after centrifugation, whereas the unbound fraction was removed by decanting the liquid phase. The bound fraction was counted in a -counter, and the data were analyzed using a four-parameter logistic model algorithm. The immunoreactive GRF1–29 amide and CJC-1295 in the test samples were calculated by interpolation from the GRF1–29 amide and CJC-1295 calibration curves, respectively.
Generation of antibodies for Western blotting
The specific antibody was raised using a peptide immunogen consisting of CJC-1295 but with the N-(MPA)lysine at position 37 replaced with a C terminus free cystein. This peptide was prepared using conventional solid-phase synthesis as described above, purified, and conjugated to keyhole limpet hemocyanin using m-maleimidobenzoyl-N-hydroxysuccinimide ester as cross-linking agent. Rabbits were injected with immunogen-keyhole limpet hemocyanin conjugate (at Alpha Diagnostic International, Inc., TX) at 15-d intervals, at multiple sc sites and one im site. The conjugate was administered in the presence of complete Freund’s adjuvant for the first injection, and all subsequent injections were given incomplete adjuvant. Terminal bleed was performed at wk 13, 7 d after the last immunization. The selected polyclonal serum R6853 TB was highly specific for CJC-1295 conjugated to plasma proteins.
Western blotting analysis
A single male Sprague Dawley rat (359 g) was administered sc CJC-1295 (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg). Blood samples were taken at predose and at 2-, 15-, and 30-min and 1-, 2-, 6-, and 24-h time points. The samples were treated in the same manner as described in the pharmacokinetic study, and the plasma samples were aliquoted and stored frozen.
The diluted (1:10) plasma samples (10 μl/well) were separated under nonreducing conditions using SDS-PAGE (25). Proteins were subsequently transferred onto nitrocellulose membrane using a semi-dry transfer apparatus. The transfer efficacy was verified by reversible staining of the membrane with 1% red Ponceau solution. The membrane was blocked with TBS-0.01% Tween 20–5% milk for 2 h at ambient temperature. Immunochemical detection was done by a primary incubation (1.4 μg/ml affinity-purified rabbit polyclonal anti-CJC-1295 R6853 TB for 1 h at ambient temperature), followed by a secondary incubation (1/200,000 Goat Anti-Rabbit IgG-HRP; Jackson ImmunoResearch Laboratories, Inc. catalog no. 111-035-144, for 1 h at ambient temperature), and revealed with SuperSignal West Femto (Pierce Chemical Co., Rockford, IL). An extra lane containing plasma sample from the 24-h time point was cut out before the preceding immunochemical detection and incubated with peroxidase-labeled rabbit anti-RSA antibody (diluted 1/400000; Accurate Chemical and Scientific Corporation catalog no. YN-RRaALBP) for 1 h at ambient temperature.
Ethics and statistical analyses
The experimental protocols were performed according to the Canadian Council on Animal Protection, following approval by the Université du Québec à Montréal Institutional Committee on Animal Protection. All GH secretion results are expressed as means ± SEM. The in vitro data were analyzed by ANOVA using Keppel’s modified Bonferroni correction. The comparison of area under the curve (AUC) results with the control was done using the bootstrapping technique (26), with a Bonferroni adjustment to maintain a confidence of 95%.
Results
In vitro stability
The stability of the HSA bioconjugates in the presence of DPP-IV was assessed in vitro. The results are shown in Table 1.
TABLE 1. In vitro stability of hGRF1–29 amide and the three hGRF-HSA bioconjugates in the presence of DPP-IV (pig)
In this assay, all of the free hGRF1–29 amide was cleaved by DPP-IV, and the directly proportional appearance of the GRF3–29 metabolite was observed. When the conjugate (CJC-1288)-HSA was tested, 58 ± 4% of the intact peptide conjugate was observed after 24 h. In this case, the metabolite corresponding to the loss of Tyr-Ala was also observed at the 24-h time point. (CJC-1293)-HSA and (CJC-1295)-HSA conjugates were stabilized 99.2 ± 0.1% and 90.3 ± 0.5%, respectively, owing to the fact that they both possessed a D-Ala at the 2-position. In the case of (CJC-1295)-HSA, no metabolite corresponding to the loss of Tyr-Ala was observed, and the 9.7% reduction in LC peak abundance after deconvolution was attributed to assay conditions.
In vitro activity
The evaluation of the in vitro activity of the bioconjugate was necessary because of differences between rat and hGRF structures (16). For this purpose, freshly excised rat anterior pituitaries were prepared for culture. The cells were incubated in the presence of the test compounds for 4 h, after which the level of GH found in the supernatant was directly measured. The results are shown in Fig. 2.
FIG. 2. In vitro GH secretion by cultured rat anterior pituitary cells (2.5 x 105 cells/well) exposed to hGRF1–29 amide and hGRF-HSA bioconjugates. HSA was evaluated in the assay to confirm that its presence did not influence the GH secretion. The maximum level of GH secreted for these HSA conjugates was comparable with the native hGRF1–29 amide in the nM range, whereas they showed no activity in the pM range (n 3 in duplicate). The statistical analysis was done using ANOVA vs. the control that contained no test compound. The values are expressed as means ± SEM. *, Statistically significant effect vs. control (P < 0.05).
The conjugates were active in the nanomolar range, causing the release of GH from the rat anterior pituitary cells, whereas the hGRF1–29 remained active in the sub-picomolar range. The maximum level of GH being secreted by the cells in the presence of the conjugates was comparable with the hGRF1–29 amide.
In vivo activity
An in vivo experiment was performed by administering a single bolus sc dose of GRF1–29 amide, CJC-1288, CJC-1293, or CJC-1295 (1 μmol/kg) to male Sprague Dawley rats and monitoring the rat GH released into plasma over a period of 2 h. The results of this experiment are shown on Fig. 3.
FIG. 3. Acute rat GH secretion after a single bolus sc administration of 1 μmol/kg of each test compound to Sprague Dawley rats (n = 7–8 per group) relative to individual 0.9% saline controls. The GH secretion profile for hGRF1–29 amide and CJC-1288 are shown in A, CJC-1293 in B, and CJC-1295 in C with the individual control groups. The GH concentrations were measured directly from blood plasma using a RIA. Values are expressed as means ± SEM. The total secreted rat GH AUC over a 2-h period, post administration, of 1 μmol/kg hGRF1–29 amide and the three test compounds is shown in D. Statistical analysis (bootstrapping technique) was done on compound vs. control for the AUCs. Values are expressed as medians ± 95% confidence interval. *, Significant difference with control.
An increase in GH was measured for GRF1–29 amide as well as the three analogs CJC-1288 (Fig. 3A), CJC-1293 (Fig. 3B), and CJC-1295 (Fig. 3C) post administration, relative to individual saline control groups. These control groups were necessary because experiments were done independently. The response was visibly greater for the tetrasubstituted analog CJC-1295 when comparing the AUC of the total GH secreted over the 2-h period, as shown on Fig. 3D.
Pharmacokinetics
The pharmacokinetic profiles of hGRF1–29 amide and CJC-1295 were evaluated in rats after sc administration (1 μmol/kg). Blood samples were taken at different time points up to 96 h, and the GRF level was measured using an available hGRF1–44-recognizing rabbit antibody. Figure 4 shows the prolonged presence of CJC-1295, where it was detectable after 72 h. The antibody used in this assay does not cross-react with rat GRF. The hGRF1–29 amide could not be detected in plasma beyond 1 h.
FIG. 4. Pharmacokinetic profiles after sc administration of 1 μmol/kg hGRF1–29 amide and CJC-1295 in male Sprague-Dawley rats (n = 4 per compound). The values were determined by RIA using a rabbit anti-hGRF1–44 antibody that does not cross-react with rat GRF.
Detection of CJC-1295 bound to serum albumin
Plasma samples were obtained after the sc administration of CJC-1295 to a male Sprague Dawley rat. Western blot analysis was attempted using the commercial anti-hGRF1–44 antibody used in the pharmacokinetic study. Unfortunately, no detection could be achieved because of the low predilution of this antibody. It was therefore necessary to generate a new antibody specific to CJC-1295 in rabbits. The rabbit polyclonal anti-CJC-1295 antibody R6853 TB was found to be highly selective to CJC-1295 (when this one was added to rat or human plasma). Western blot analysis was done with this antibody on the plasma samples collected from the rat, and results are shown on Fig. 5.
FIG. 5. Identification of CJC-1295 bound to rat albumin, post sc administration, of 1 μmol/kg CJC-1295 using Western blot analysis. Blood samples were taken at the indicated time points, diluted 1:10, subjected to SDS-PAGE, and detected using a CJC-1295-specific antibody (R6853 TB). The second 24-h lane on the right was revealed with a RSA-specific antibody.
The Western blot revealed that some cross-reactivity of the second antibody with rat Igs could be seen but that no band corresponding to albumin was found in the predose and 2-min lanes. A band corresponding to albumin became visible at 15 min and persisted in all samples taken beyond this time. In fact, the albumin band was clearly visible at the 24-h time point. An additional lane from the same blot containing plasma from the 24-h sample was revealed using an anti-RSA specific antibody. The high background was attributed to the large amount of proteins loaded and to the fact that anti-RSA recognizes monomers and polymers of RSA. The RSA band was clearly identified and was aligned with the CJC-1295 band. This is strong evidence that CJC-1295 is attached to RSA in plasma after sc administration.
Discussion
GRF is known to be unstable in the presence of plasma DPP-IV, which has been identified as one of its most important routes of elimination. The in vitro stability of GRF-albumin conjugates was done to demonstrate that the attachment of GRF to albumin protects the active peptide portion from DPP-IV-mediated degradation. The bioconjugation of CJC-1288 to HSA did indeed stabilize the GRF peptide portion from DPP-IV degradation but not completely. The metabolite corresponding to the loss of Tyr-Ala was detected by LC/(ES)MS analysis, confirming that the normal degradation pathway is still not completely offset. It became necessary to substitute the L-alanine at the 2-position for a D-Ala to further stabilize the peptide portion. This variation resulted in a high degree of stability, as exemplified by (CJC-1293)-HSA and (CJC-1295)-HSA conjugates both with more than 90% remaining intact in solution after 24 h exposure to DPP-IV.
The in vitro experiment on cultured rat anterior pituitary cells showed that all three HSA bioconjugates of modified or unmodified hGRF1–29 amide are able to activate the rat GRF receptor in the nM range but not in the sub-pM range. The response in the presence of HSA was also measured, and no GH secretion was observed, indicating that the albumin alone could not activate the receptor. Although there was a loss of activity, relative to the free hGRF1–29 amide, this experiment confirmed that the rat model could be used for our in vivo screen.
When male Sprague Dawley rats were injected with a single bolus sc dose of hGRF1–29 amide, CJC-1288, CJC-1293, and CJC-1295, acute GH secretions were observed in all cases. The latter gave the largest plasma AUC of GH over a 2-h period and was selected for further experiments.
Through a Western blot analysis of plasma samples taken after the administration of CJC-1295, it was demonstrated that CJC-1295 was present on the band corresponding to albumin from 15 min to beyond 24 h. There was no band detected at predose and 2 min post administration. Subjecting the plasma samples to SDS-PAGE assured us that the CJC-1295 was covalently bound to the albumin.
The pharmacokinetic profile of CJC-1295 showed a prolonged residence time in plasma compared with hGRF1–29 amide. The mean residence time of the immunoreactive GRF in plasma went from less than 0.5 h for hGRF1–29 amide to more than 30 h for CJC-1295. It was shown earlier that the conjugate was stabilized against enzyme degradation, removing this route from the overall elimination process. The explanation for the prolonged residence time was therefore attributed almost exclusively to reduced kidney clearance.
There was a good correlation between the amount of immunoreactive GRF found in the plasma and the acute GH response for both hGRF1–29 amide and CJC-1295. The maleimido group has been shown in the literature to react with the free thiol on albumin (8, 9, 10, 11, 12, 13, 14). Albumin is found in the sc extravascular fluid in a concentration of 16 g/liter (27).1 The prolonged plasma residence time, combined with the Western blot analysis, convinced us that in vivo bioconjugation occurred and that the observed biological response was due to the interaction of the conjugate with the GRF receptor on the anterior pituitary gland.
After 2 h of exposure to CJC-1295, the level of GH eventually returns to the baseline value although the immunoreactive GRF level remains elevated. The lowering of GH concentration can be attributed to a number of reasons, such as down-regulation of the GRF receptor (22), a drop in pituitary GH content (28), or the multicomponent feedback loop regulated by somatostatin (29) and IGF-1/insulin (30). At this point, no further experiments were undertaken to explain the reduction in plasma GH, because it is well established that there is receptor desensitization to GRF observed in rat cells, and the physiological relevance has not been correlated to receptor down-regulation in humans (31). It has also been clearly established that GH secretion is pulsatile in the presence of sustained plasma concentrations of GRF (6, 32, 33). The pulsatile response to the administration of CJC-1295 will need to be demonstrated in a more appropriate model (34). This subject will be addressed in a subsequent paper in this series.
In summary, in vivo bioconjugation of maleimido hGRF1–29 derivatives to serum albumin occurs after a single bolus sc administration and is a useful tool to increase the plasma residence time of these particular bioactive peptides. The results reported herein demonstrate that a hGRF-albumin conjugate is able to interact with, and activate, the rat GRF receptor on the anterior pituitary to produce a high level of GH in plasma over time, leading to the discovery of CJC-1295 as a long-lasting hGRF analog.
Acknowledgments
The authors are grateful to the following people: Julie Carette and Peter Bakis for their skillful assistance in the preparation of the peptides, Omar Quraishi for providing the HSA bioconjugates, Dominador Calamba for his technical expertise in the RIA, and Nadia Sourial for performing the statistical analysis.
References
Giustina A, Veldhuis JD 1998 Pathophysiology of the neuroregulation of growth hormone secretion in experimental animals and the human. Endocr Rev 19:717–797
Frohman LA, Jansson JO 1986 Growth hormone-releasing hormone. Endocr Rev 7:223–253
Frohman LA, Downs TR, Williams TC, Heimer EP, Pan Y-CE, Felix AM 1986 Rapid enzymatic degradation of growth hormone-releasing hormone by plasma in vitro and in vivo to a biologically inactive product cleaved at the NH2 terminus. J Clin Invest 78:906–913
D’Antonio M, Louveau I, Esposito P, Bertolino M, Canali S 2004 Pharmacodynamic evaluation of a PEGylated analogue of human growth hormone releasing factor in rats and pigs. Growth Horm IGF Res 14:226–234
Esposito P, Barbero L, Caccia P, Caliceti P, D’Antonio M, Piquet G, Veronese FM 2003 PEGylation of growth hormone-releasing hormone (GRF) analogues. Adv Drug Deliv Rev 55:1279–1291
Vance ML, Kaiser DL, Evans WS, Furlanetto R, Vale W, Rivier J, Thorner MO 1985 Pulsatile growth hormone secretion in normal man during a continuous 24-hour infusion of human growth hormone releasing factor (1–40). Evidence for intermittent somatostatin secretion. J Clin Invest 75:1584–1590
Warnecke A, Kratz F 2003 Maleimide-oligo(ethylene glycol) derivatives of camptothecin as albumin-binding prodrugs: synthesis and antitumor efficacy. Bioconjug Chem 14:377–387
Kratz F, Muller-Driver R, Hofmann I, Drevs J, Unger C 2000 A novel macromolecular prodrug concept exploiting endogenous serum albumin as a drug carrier for cancer chemotherapy. J Med Chem 43:1253–1256
Kratz F, Warnecke A, Scheuermann K, Stockmar C, Schwab J, Lazar P, Drückes P, Esser N, Drevs J, Rognan D, Bissantz C, Hinderling C, Folkers G, Fichtner I, Unger C 2002 Probing the cysteine-34 position of endogenous serum albumin with thiol-binding doxorubicin derivatives. Improved efficacy of an acid-sensitive doxorubicin derivative with specific albumin binding properties compared to that of the parent compound. J Med Chem 45:5523–5533
Kim J-G, Baggio LL, Bridon DP, Castaigne J-P, Robitaille MF, Jette L, Benquet C, Drucker DJ 2003 Development and characterization of a glucagon-like peptide 1-albumin conjugate: the ability to activate the glucagon-like peptide 1 receptor in vivo. Diabetes 52:751–759
Léger R, Thibaudeau K, Robitaille M, Quraishi O, van Wyk P, Bousquet-Gagnon N, Carette J, Castaigne J-P, Bridon DP 2004 Identification of CJC-1131-albumin bioconjugate as a stable and bioactive GLP-1(7–36) analog. Bioorg Med Chem Lett 14:4395–4398
Léger R, Benquet C, Huang X, Quraishi O, van Wyk P, Bridon D 2004 Kringle 5 peptide-albumin conjugates with anti-migratory activity. Bioorg Med Chem Lett 14:841–845
Léger R, Robitaille M, Quraishi O, Denholm E, Benquet C, Carette J, van Wyk P, Pellerin I, Bousquet-Gagnon N, Castaigne J-P, Bridon D 2003 Synthesis and in vitro analysis of atrial natriuretic peptide-albumin conjugates. Bioorg Med Chem Lett 13:3571–3575
Holmes DL, Thibaudeau K, L’Archêveque B, Milner PG, Ezrin AM, Bridon DP 2000 Site specific 1:1 opioid:albumin conjugate with in vitro activity and long in vivo duration. Bioconjug Chem 11:439–444
Baird A, Wehrenberg WB, Ling N 1986 Relative potencies of human, rat, bovine/caprine, porcine and ovine hypothalamic growth hormone-releasing factors to release growth hormone by the rat anterior pituitary in vitro. Neuroendocrinology 42:273–276
Cervini LA, Donaldson CJ, Koerber SC, Vale WW, Rivier JE 1998 Human growth hormone-releasing hormone hGHRH(1–29)-NH2: systematic structure-activity relationship studies. J Med Chem 41:717–727
Su C-M, Jensen LR, Heimer EP, Felix AM, Pan Y-CE, Mowles TF 1991 In vitro stability of growth hormone releasing factor (GRF) analogs in porcine plasma. Horm Metab Res 23:15–21
Campbell RM, Stricker P, Miller R, Bongers J, Liu W, Lambros T, Ahmad M, Felix AM, Heimer EP 1994 Enhanced stability and potency of novel growth hormone-releasing factor (GRF) analogues derived from rodent and human GRF sequences. Peptides 15:489–495
Bongers J, Heimer EP, Lambros T, Pan Y-CE, Campbell RM, Felix AM 1992 Degradation of aspartic acid and asparagine residues in human growth hormone-releasing factor. Int J Pept Protein Res 39:364–374
Campbell RM, Bongers J, Felix AM 1995 Rational design, synthesis, and biological evaluation of novel growth hormone releasing factor analogues. Biopolymers 37:67–88
Wilfinger WW, Larsen WJ, Downs TR, Wilbur DL 1984 An in vitro model for studies of intercellular communication in cultured rat anterior pituitary cells. Tissue Cell 16:483–497
Aleppo G, Moskal 2nd SF, De Grandis PA, Kineman RD, Frohman LA 1997 Homologous down-regulation of growth hormone-releasing hormone receptor messenger ribonucleic acid levels. Endocrinology 138:1058–1065
Sugihara H, Emoto N, Tamura H, Kamegai J, Shibasaki T, Minami S, Wakabayashi I 1999 Effect of insulin-like growth factor-I on growth hormone-releasing factor receptor expression in primary rat anterior pituitary cell culture. Neurosci Lett 276:87–90
Rafferty B, Poole S, Clarke R, Schulster D 1985 Growth hormone-releasing factor analogue (hGRF1–29NH2): immunoreactive-GRF plasma levels after intravenous and subcutaneous administration. J Endocrinol 107:R5–R8
Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685
Zhou XH, Dinh P 2005 Nonparametric confidence intervals for the one- and two-sample problems. Biostatistics 6:187–200
Peters Jr T 1995 All about albumin: biochemistry, genetics and medical applications. New York: Academic Press; 230–232
Turner JP, Tannenbaum GS 1995 In vivo evidence of a positive role for somatostatin to optimize pulsatile growth hormone secretion. Am J Physiol 269:E683–E690
Sugihara H, Minami S, Okada K, Kamegai J, Hasegawa O, Wakabayashi I 1993 Somatostatin reduces transcription of the growth hormone gene in rats. Endocrinology 132:1225–1229
Leung K, Rajkovic IA, Peters E, Markus I, Van Wyk JJ, Ho KK 1996 Insulin-like growth factor I and insulin down-regulate growth hormone (GH) receptors in rat osteoblasts: evidence for a peripheral feedback loop regulating GH action. Endocrinology 137:2694–2702
Vance ML, Kaiser DL, Martha Jr PM, Furlanetto R, Rivier J, Vale W, Thorner MO 1989 Lack of in vivo somatotroph desensitization or depletion after 14 days of continuous growth hormone (GH)-releasing hormone administration in normal men and a GH deficient boy. J Clin Endocrinol Metab 68:22–28
Kovács M, Halmos G, Groot K, Izdebski J, Schally AV 1996 Chronic administration of a new potent agonist of growth hormone-releasing hormone induces compensatory linear growth in growth hormone-deficient rats: mechanism of action. Neuroendocrinology 64:169–176
Kovács M, Fáncsik A, Mez? I, Teplán I, Flerkó B 1994 Effects of continuous and repetitive administration of a potent analog of GH-RH(1–30)NH2 on the GH release in rats. Neuroendocrinology 59:371–379
Dubreuil P, Brazeau P, Moreau S, Farmer C, Coy D, Abribat T 2001 The use of pigs as an animal model to evaluate the efficacy, potency and specificity of two growth hormone releasing factor analogues. Growth Horm IGF Res 11:173–186(Lucie Jetté, Roger Léger,)
Address all correspondence and requests for reprints to: Roger Léger, Ph.D., Department of Research, ConjuChem Inc., 225 President-Kennedy Avenue, Suite 3950, Montreal, Quebec, Canada, H2X 3Y8. E-mail: leger@conjuchem.com.
Abstract
In vivo bioconjugation to the free thiol on Cys34 of serum albumin by a strategically placed reactive group on a bioactive peptide is a useful tool to extend plasma half-life. Three maleimido derivates of human GH-releasing factor (hGRF)1–29 were synthesized and bioconjugated to human serum albumin ex vivo. All three human serum albumin conjugates showed enhanced in vitro stability against dipeptidylpeptidase-IV and were bioactive in a GH secretion assay in cultured rat anterior pituitary cells. When the maleimido derivatives were individually administered sc to normal male Sprague Dawley rats, an acute secretion of GH was measured in plasma. The best compound, CJC-1295, showed a 4-fold increase in GH area under the curve over a 2-h period compared with hGRF1–29. CJC-1295, a tetrasubstituted form of hGRF1–29 with an added N-3-maleimidopropionamide derivative of lysine at the C terminus, was selected for further pharmacokinetic evaluation, where it was found to be present in plasma beyond 72 h. A Western blot analysis of the plasma of a rat injected with CJC-1295 showed the presence of a CJC-1295 immunoreactive species on the band corresponding to serum albumin, appearing after 15 min and remaining in circulation beyond 24 h. These results led to the identification of CJC-1295 as a stable and active hGRF1–29 analog with an extended plasma half-life.
Introduction
HUMAN GH-RELEASING factor (hGRF) is a 44-amino acid endocrine peptide secreted by the hypothalamus that reaches the anterior pituitary, where it binds to its receptor, setting off a cascade of signals that leads to the release of GH into the plasma (1). The hormone, unfortunately, suffers from a very short half-life in vivo (2, 3). Recently developed long-lasting derivatives of hGRF, such as polyethylene glycol derivatives, have demonstrated the ability to activate the GRF receptor of rats and pigs, leading to the secretion of GH in the plasma (4, 5). Furthermore, the GH secretion pattern remains pulsatile when these derivatives are sustained at an elevated concentration in plasma, in agreement with the results from continuous hGRF infusion experiments (6).
In vivo bioconjugation to serum albumin is a useful tool to increase the half-life of small molecules (7, 8, 9) or peptides (10, 11, 12, 13, 14) in plasma. In vivo bioconjugation occurs when a strategically placed reactive group on a bioactive peptide reacts with a nucleophilic entity found in blood or in sc interstitium to form a stable bond. The foremost nucleophile is the thiol, and its most abundant source in these fluids is Cys34 on albumin. The thiol on Cys34 reacts with a Michael acceptor, such as a maleimido derivative, leading to a new bioactive protein construct that will adopt an extended half-life due to stabilization from enzymatic degradation (11, 12, 13) or reduced elimination through the kidney (14). It therefore became logical to combine the long-lasting effect of bioconjugation with the proper GRF analog.
We initiated a research program to make a number of maleimido derivatives of hGRF1–29 and hGRF analogs and thought that the rat physiology was deemed adequate to screen the products of in vivo bioconjugation with serum albumin. However, it has never been shown that a hGRF albumin conjugate or a fusion protein can activate the rat GRF receptor. For this reason, before initiating in vivo studies, the demonstration that a hGRF1–29-albumin conjugate retains in vitro activity on cultured rat anterior pituitary cells becomes essential.
The relative potency of rat GRF, which is in the order of 3- to 6-fold more active (15), as well as a 70% homology (16), when compared with the hGRF, is also of concern and needs to be addressed early to establish the rat as a viable screening model.
Another point of interest is the demonstration that bioconjugation helps to stabilize the active peptide portion of the new construct from degradation by plasma enzymes. Dipeptidylpeptidase-IV (DPP-IV) is the primary culprit with regard to the metabolism and deactivation of GRF in plasma. It recognizes alanine and proline residues at the P1 site and, in the case of GRF1–29, causes the release of the Tyr-Ala dipeptide fragment generating the inactive GRF3–29 (17).
We report herein the results from three sets of experiments on three hGRF1–29 analogs possessing the ability to bioconjugate to serum albumin: 1) the in vitro bioactivity using a cultured rat anterior pituitary cell assay and stability in the presence of DPP-IV of preformed albumin conjugates, 2) the activation of anterior pituitaries and subsequent release of GH in rats upon in vivo bioconjugation of the hGRF1–29 analogs, and 3) comparison of the plasma pharmacokinetic profiles in rats of one of the derivatives to the native peptide hGRF1–29 amide.
Materials and Methods
Synthetic peptides
The compounds prepared for this study are shown on Fig. 1. They include the native GRF1–29 amide, CJC-1288, which has a 3-maleimidopropionic acid (MPA) unit added to the amine of an extra lysine at the C terminus of GRF1–29. When initiating these studies, it was not known whether bioconjugation to serum albumin would totally stabilize the peptide from enzymatic degradation. For this reason, a D-alanine was introduced to replace the natural L-alanine at the 2-position (11), followed by the insertion of MPA as in the previous compound leading to CJC-1293. The third tetrasubstituted analog, CJC-1295, has a D-Ala at the 2-position, a Gln at the 8-position to reduce asparagine rearrangement or amide hydrolysis to aspartic acid, an Ala at position 15 to enhance bioactivity, and a Leu at position 27 to prevent methionine oxidation along with the MPA-Lys at the C terminus (17, 18, 19, 20). The two amino acid replacements at positions 8 and 27 were essentially done to confer manufacturing stability.
FIG. 1. Molecular structures of hGRF1–29 amide and the three maleimido derivatives CJC-1288, CJC-1293, and CJC-1295, prepared by solid-phase synthesis. CJC-1288 is hGRF1–29 with an extra lysine (Lys) at the 30-position to accommodate a MPA; CJC-1293 is equivalent to CJC-1288 but with a D-alanine (D-Ala) at the 2-position; and CJC-1295 is a tetrasubstituted analog of CJC-1288 but with a D-Ala at the 2-position, a glutamine (Gln) at the 8-position, an Ala at the 15-position, and a leucine (Leu) at the 27-position.
The reference GRF1–29 amide and the three maleimido derivatives CJC-1288, CJC-1293, and CJC-1295 were synthesized using the Fmoc amino acid strategy on Ramage resin, followed by orthogonal deprotection, to attach the MPA as described previously (11, 12, 13). All derivatives were lyophilized as trifluoroacetic acid salts and had more than 95% purity by HPLC.
Human serum albumin (HSA) bioconjugates
Compounds CJC-1288, CJC-1293, and CJC-1295 were bioconjugated to HSA (Cortex Biochem, Inc., San Leandro, CA) according to the protocol previously described (11, 12, 13). The bioconjugates were weighed as lyophilized powders before use.
In vitro DPP-IV stability assay
The DPP-IV (pig) stability assay was performed in the same manner as already described (11). Briefly, test compounds GRF1–29 amide (150 μM) along with (CJC-1288)-HSA, (CJC-1293)-HSA, and (CJC-1295)-HSA conjugates (10 mg) were individually solubilized in 1 x PBS (pH 7.1) (250 μl) and added directly to a vial of DPP-IV enzyme (5 mU, porcine; Calbiochem, La Jolla, CA). The mixture was incubated at 37 C under mixing conditions, and aliquots (25 μl) were taken at 0 and 24 h and immediately frozen at –80 C. Samples were thawed just before liquid chromatography (electrospray) mass spectrometry (LC/(ES)MS) analysis. Percentages are reported as a ratio of the test compound area under the peak or abundance after deconvolution for the desired protein mass at the 24-h time point relative to a reference solution that does not contain DPP-IV but submitted to identical incubation and sampling conditions. The entire experiment was repeated twice.
Primary rat pituitary cell culture
Anterior pituitary cells were harvested and prepared as previously described (21, 22, 23). Male Sprague Dawley rats (150–300 g; Charles River Canada Inc., Quebec, Canada) were anesthetized and subsequently decapitated. Whole pituitaries were removed and placed immediately in ice-cold sterile MEM containing 0.1% BSA, 100 μg/ml streptomycin, and 100 U/ml penicillin (Invitrogen Canada Inc., Burlington, Ontario, Canada). Cells were prepared freshly within 1 h post pituitary collection.
Individual anterior pituitaries were minced, and tissue fragments were digested in 10 ml sterile medium containing 0.15% trypsin (Difco Laboratories, Inc., Detroit, MI) under gentle and continuous agitation for 45 min at 37 C. Tissue pieces were mechanically disrupted using a Pasteur pipette and incubated for an additional 15 min. Cells were then washed twice and resuspended in DMEM (Invitrogen Canada Inc., Quebec, Canada) supplemented with 0.1% BSA and antibiotics. Cells were seeded in 24-well plates at a density of 2.5 x 105 cells/50 μl medium·well in DMEM and allowed to attach for 1 h in a humidified CO2 (5%) air incubator at 37 C. After cell attachment, 1 ml fresh DMEM, supplemented with 10% horse serum (Invitrogen Canada Inc., Quebec, Canada), was added.
In vitro GH secretion assay
After 72 h of culture, primary cells were washed twice with serum free DMEM media and incubated for 1 h at 37 C. After this equilibrium period, medium was replaced with fresh medium, and cells were treated with test compound (0, 10–13, 10–9, and 10–6 M final concentrations as prepared previously) (12) or HSA (0, 3 x 10–12, 3 x 10–8, and 3 x 10–5 M) for 4 h. After incubation, supernatants were collected and centrifuged to remove nonadherent cells and were stored at –20 C until analysis. Quantitative determination of rat GH was performed by RIA using a commercial kit (Linco Research, Inc., St. Charles, MO) as described in the manufacturer’s instructions. Each analysis was replicated at least three times in duplicate. The intraassay coefficient of variation for the rat GH measurement was less than 10%.
Acute in vivo GH secretion
The acute GH secretory profile of the four test compounds was evaluated using catheterized (carotid), freely moving, 7- to 8-wk-old male Sprague Dawley rats (all rats were from Charles River Canada, Saint-Constant, Quebec, Canada). The animals received a single bolus sc injection of test compound (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg) in the dorsolumbar area (n = 7 rats per group; CJC-1293 and CJC-1295 had their own individual saline control groups of n = 7; GRF1–29 amide and CJC-1288 had a single saline control group for the two). Serial blood samples were taken up to 2 h post administration. GH plasma levels were determined using the same RIA method described above. Onset of experimentation was only between 0900 and 1100 h.
Pharmacokinetic analysis in Sprague Dawley rats
Seven- to 8-wk-old male Sprague Dawley rats (n = 4 per test compound) were administered a single bolus sc dose of either GRF1–29 amide or CJC-1295 (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg) in the dorsolumbar region. Serial blood samples were collected from each animal at predose (before injection); at 5 and 30 min; at 1, 2, 4, 8, 24, 48, and 72 h in tubes containing EDTA and DPP-IV inhibitor (Linco Research, Inc.). The samples were then centrifuged (2500 rpm for 10 min at 4 C), and the plasma was aliquoted and kept frozen until analysis.
RIA method for pharmacokinetic analysis
The labeled tracer antigen 125I-GRF1–29 amide was obtained from the Douglas Hospital Research Center (Montreal, Quebec, Canada), and rabbit anti-hGRF1–44 antibody was obtained from Accurate Chemical and Scientific Corporation (Westbury, NY). The inability of this antibody to cross-react with rat GRF has been certified by the manufacturer.
The assay was a disequilibrium RIA (24). The GRF1–29 amide and CJC-1295 calibrators were prepared by adding known concentrations of GRF1–29 amide or CJC-1295 to rat plasma containing EDTA and DPP-IV inhibitor, respectively. The rat serum albumin (RSA) conjugate thus obtained was characterized by LC/(ES)MS, and the amount of free unreacted CJC-1295 was measured (<1%). Rabbit anti-hGRF1–44 antibody was found to cross-react with hGRF1–29 amide and CJC-1295. The calibrators and test samples were incubated with the antibody overnight at 2–8 C. This incubation was followed by the addition of the tracer (125I-GRF1–29). After another overnight incubation at 2–8 C, the bound and unbound GRF1–29 amide or CJC-1295 fractions were separated by a second antibody precipitation. The bound GRF1–29 amide or CJC-1295 was collected after centrifugation, whereas the unbound fraction was removed by decanting the liquid phase. The bound fraction was counted in a -counter, and the data were analyzed using a four-parameter logistic model algorithm. The immunoreactive GRF1–29 amide and CJC-1295 in the test samples were calculated by interpolation from the GRF1–29 amide and CJC-1295 calibration curves, respectively.
Generation of antibodies for Western blotting
The specific antibody was raised using a peptide immunogen consisting of CJC-1295 but with the N-(MPA)lysine at position 37 replaced with a C terminus free cystein. This peptide was prepared using conventional solid-phase synthesis as described above, purified, and conjugated to keyhole limpet hemocyanin using m-maleimidobenzoyl-N-hydroxysuccinimide ester as cross-linking agent. Rabbits were injected with immunogen-keyhole limpet hemocyanin conjugate (at Alpha Diagnostic International, Inc., TX) at 15-d intervals, at multiple sc sites and one im site. The conjugate was administered in the presence of complete Freund’s adjuvant for the first injection, and all subsequent injections were given incomplete adjuvant. Terminal bleed was performed at wk 13, 7 d after the last immunization. The selected polyclonal serum R6853 TB was highly specific for CJC-1295 conjugated to plasma proteins.
Western blotting analysis
A single male Sprague Dawley rat (359 g) was administered sc CJC-1295 (1 μmol/kg in 0.9% saline; injection vol, 1 ml/kg). Blood samples were taken at predose and at 2-, 15-, and 30-min and 1-, 2-, 6-, and 24-h time points. The samples were treated in the same manner as described in the pharmacokinetic study, and the plasma samples were aliquoted and stored frozen.
The diluted (1:10) plasma samples (10 μl/well) were separated under nonreducing conditions using SDS-PAGE (25). Proteins were subsequently transferred onto nitrocellulose membrane using a semi-dry transfer apparatus. The transfer efficacy was verified by reversible staining of the membrane with 1% red Ponceau solution. The membrane was blocked with TBS-0.01% Tween 20–5% milk for 2 h at ambient temperature. Immunochemical detection was done by a primary incubation (1.4 μg/ml affinity-purified rabbit polyclonal anti-CJC-1295 R6853 TB for 1 h at ambient temperature), followed by a secondary incubation (1/200,000 Goat Anti-Rabbit IgG-HRP; Jackson ImmunoResearch Laboratories, Inc. catalog no. 111-035-144, for 1 h at ambient temperature), and revealed with SuperSignal West Femto (Pierce Chemical Co., Rockford, IL). An extra lane containing plasma sample from the 24-h time point was cut out before the preceding immunochemical detection and incubated with peroxidase-labeled rabbit anti-RSA antibody (diluted 1/400000; Accurate Chemical and Scientific Corporation catalog no. YN-RRaALBP) for 1 h at ambient temperature.
Ethics and statistical analyses
The experimental protocols were performed according to the Canadian Council on Animal Protection, following approval by the Université du Québec à Montréal Institutional Committee on Animal Protection. All GH secretion results are expressed as means ± SEM. The in vitro data were analyzed by ANOVA using Keppel’s modified Bonferroni correction. The comparison of area under the curve (AUC) results with the control was done using the bootstrapping technique (26), with a Bonferroni adjustment to maintain a confidence of 95%.
Results
In vitro stability
The stability of the HSA bioconjugates in the presence of DPP-IV was assessed in vitro. The results are shown in Table 1.
TABLE 1. In vitro stability of hGRF1–29 amide and the three hGRF-HSA bioconjugates in the presence of DPP-IV (pig)
In this assay, all of the free hGRF1–29 amide was cleaved by DPP-IV, and the directly proportional appearance of the GRF3–29 metabolite was observed. When the conjugate (CJC-1288)-HSA was tested, 58 ± 4% of the intact peptide conjugate was observed after 24 h. In this case, the metabolite corresponding to the loss of Tyr-Ala was also observed at the 24-h time point. (CJC-1293)-HSA and (CJC-1295)-HSA conjugates were stabilized 99.2 ± 0.1% and 90.3 ± 0.5%, respectively, owing to the fact that they both possessed a D-Ala at the 2-position. In the case of (CJC-1295)-HSA, no metabolite corresponding to the loss of Tyr-Ala was observed, and the 9.7% reduction in LC peak abundance after deconvolution was attributed to assay conditions.
In vitro activity
The evaluation of the in vitro activity of the bioconjugate was necessary because of differences between rat and hGRF structures (16). For this purpose, freshly excised rat anterior pituitaries were prepared for culture. The cells were incubated in the presence of the test compounds for 4 h, after which the level of GH found in the supernatant was directly measured. The results are shown in Fig. 2.
FIG. 2. In vitro GH secretion by cultured rat anterior pituitary cells (2.5 x 105 cells/well) exposed to hGRF1–29 amide and hGRF-HSA bioconjugates. HSA was evaluated in the assay to confirm that its presence did not influence the GH secretion. The maximum level of GH secreted for these HSA conjugates was comparable with the native hGRF1–29 amide in the nM range, whereas they showed no activity in the pM range (n 3 in duplicate). The statistical analysis was done using ANOVA vs. the control that contained no test compound. The values are expressed as means ± SEM. *, Statistically significant effect vs. control (P < 0.05).
The conjugates were active in the nanomolar range, causing the release of GH from the rat anterior pituitary cells, whereas the hGRF1–29 remained active in the sub-picomolar range. The maximum level of GH being secreted by the cells in the presence of the conjugates was comparable with the hGRF1–29 amide.
In vivo activity
An in vivo experiment was performed by administering a single bolus sc dose of GRF1–29 amide, CJC-1288, CJC-1293, or CJC-1295 (1 μmol/kg) to male Sprague Dawley rats and monitoring the rat GH released into plasma over a period of 2 h. The results of this experiment are shown on Fig. 3.
FIG. 3. Acute rat GH secretion after a single bolus sc administration of 1 μmol/kg of each test compound to Sprague Dawley rats (n = 7–8 per group) relative to individual 0.9% saline controls. The GH secretion profile for hGRF1–29 amide and CJC-1288 are shown in A, CJC-1293 in B, and CJC-1295 in C with the individual control groups. The GH concentrations were measured directly from blood plasma using a RIA. Values are expressed as means ± SEM. The total secreted rat GH AUC over a 2-h period, post administration, of 1 μmol/kg hGRF1–29 amide and the three test compounds is shown in D. Statistical analysis (bootstrapping technique) was done on compound vs. control for the AUCs. Values are expressed as medians ± 95% confidence interval. *, Significant difference with control.
An increase in GH was measured for GRF1–29 amide as well as the three analogs CJC-1288 (Fig. 3A), CJC-1293 (Fig. 3B), and CJC-1295 (Fig. 3C) post administration, relative to individual saline control groups. These control groups were necessary because experiments were done independently. The response was visibly greater for the tetrasubstituted analog CJC-1295 when comparing the AUC of the total GH secreted over the 2-h period, as shown on Fig. 3D.
Pharmacokinetics
The pharmacokinetic profiles of hGRF1–29 amide and CJC-1295 were evaluated in rats after sc administration (1 μmol/kg). Blood samples were taken at different time points up to 96 h, and the GRF level was measured using an available hGRF1–44-recognizing rabbit antibody. Figure 4 shows the prolonged presence of CJC-1295, where it was detectable after 72 h. The antibody used in this assay does not cross-react with rat GRF. The hGRF1–29 amide could not be detected in plasma beyond 1 h.
FIG. 4. Pharmacokinetic profiles after sc administration of 1 μmol/kg hGRF1–29 amide and CJC-1295 in male Sprague-Dawley rats (n = 4 per compound). The values were determined by RIA using a rabbit anti-hGRF1–44 antibody that does not cross-react with rat GRF.
Detection of CJC-1295 bound to serum albumin
Plasma samples were obtained after the sc administration of CJC-1295 to a male Sprague Dawley rat. Western blot analysis was attempted using the commercial anti-hGRF1–44 antibody used in the pharmacokinetic study. Unfortunately, no detection could be achieved because of the low predilution of this antibody. It was therefore necessary to generate a new antibody specific to CJC-1295 in rabbits. The rabbit polyclonal anti-CJC-1295 antibody R6853 TB was found to be highly selective to CJC-1295 (when this one was added to rat or human plasma). Western blot analysis was done with this antibody on the plasma samples collected from the rat, and results are shown on Fig. 5.
FIG. 5. Identification of CJC-1295 bound to rat albumin, post sc administration, of 1 μmol/kg CJC-1295 using Western blot analysis. Blood samples were taken at the indicated time points, diluted 1:10, subjected to SDS-PAGE, and detected using a CJC-1295-specific antibody (R6853 TB). The second 24-h lane on the right was revealed with a RSA-specific antibody.
The Western blot revealed that some cross-reactivity of the second antibody with rat Igs could be seen but that no band corresponding to albumin was found in the predose and 2-min lanes. A band corresponding to albumin became visible at 15 min and persisted in all samples taken beyond this time. In fact, the albumin band was clearly visible at the 24-h time point. An additional lane from the same blot containing plasma from the 24-h sample was revealed using an anti-RSA specific antibody. The high background was attributed to the large amount of proteins loaded and to the fact that anti-RSA recognizes monomers and polymers of RSA. The RSA band was clearly identified and was aligned with the CJC-1295 band. This is strong evidence that CJC-1295 is attached to RSA in plasma after sc administration.
Discussion
GRF is known to be unstable in the presence of plasma DPP-IV, which has been identified as one of its most important routes of elimination. The in vitro stability of GRF-albumin conjugates was done to demonstrate that the attachment of GRF to albumin protects the active peptide portion from DPP-IV-mediated degradation. The bioconjugation of CJC-1288 to HSA did indeed stabilize the GRF peptide portion from DPP-IV degradation but not completely. The metabolite corresponding to the loss of Tyr-Ala was detected by LC/(ES)MS analysis, confirming that the normal degradation pathway is still not completely offset. It became necessary to substitute the L-alanine at the 2-position for a D-Ala to further stabilize the peptide portion. This variation resulted in a high degree of stability, as exemplified by (CJC-1293)-HSA and (CJC-1295)-HSA conjugates both with more than 90% remaining intact in solution after 24 h exposure to DPP-IV.
The in vitro experiment on cultured rat anterior pituitary cells showed that all three HSA bioconjugates of modified or unmodified hGRF1–29 amide are able to activate the rat GRF receptor in the nM range but not in the sub-pM range. The response in the presence of HSA was also measured, and no GH secretion was observed, indicating that the albumin alone could not activate the receptor. Although there was a loss of activity, relative to the free hGRF1–29 amide, this experiment confirmed that the rat model could be used for our in vivo screen.
When male Sprague Dawley rats were injected with a single bolus sc dose of hGRF1–29 amide, CJC-1288, CJC-1293, and CJC-1295, acute GH secretions were observed in all cases. The latter gave the largest plasma AUC of GH over a 2-h period and was selected for further experiments.
Through a Western blot analysis of plasma samples taken after the administration of CJC-1295, it was demonstrated that CJC-1295 was present on the band corresponding to albumin from 15 min to beyond 24 h. There was no band detected at predose and 2 min post administration. Subjecting the plasma samples to SDS-PAGE assured us that the CJC-1295 was covalently bound to the albumin.
The pharmacokinetic profile of CJC-1295 showed a prolonged residence time in plasma compared with hGRF1–29 amide. The mean residence time of the immunoreactive GRF in plasma went from less than 0.5 h for hGRF1–29 amide to more than 30 h for CJC-1295. It was shown earlier that the conjugate was stabilized against enzyme degradation, removing this route from the overall elimination process. The explanation for the prolonged residence time was therefore attributed almost exclusively to reduced kidney clearance.
There was a good correlation between the amount of immunoreactive GRF found in the plasma and the acute GH response for both hGRF1–29 amide and CJC-1295. The maleimido group has been shown in the literature to react with the free thiol on albumin (8, 9, 10, 11, 12, 13, 14). Albumin is found in the sc extravascular fluid in a concentration of 16 g/liter (27).1 The prolonged plasma residence time, combined with the Western blot analysis, convinced us that in vivo bioconjugation occurred and that the observed biological response was due to the interaction of the conjugate with the GRF receptor on the anterior pituitary gland.
After 2 h of exposure to CJC-1295, the level of GH eventually returns to the baseline value although the immunoreactive GRF level remains elevated. The lowering of GH concentration can be attributed to a number of reasons, such as down-regulation of the GRF receptor (22), a drop in pituitary GH content (28), or the multicomponent feedback loop regulated by somatostatin (29) and IGF-1/insulin (30). At this point, no further experiments were undertaken to explain the reduction in plasma GH, because it is well established that there is receptor desensitization to GRF observed in rat cells, and the physiological relevance has not been correlated to receptor down-regulation in humans (31). It has also been clearly established that GH secretion is pulsatile in the presence of sustained plasma concentrations of GRF (6, 32, 33). The pulsatile response to the administration of CJC-1295 will need to be demonstrated in a more appropriate model (34). This subject will be addressed in a subsequent paper in this series.
In summary, in vivo bioconjugation of maleimido hGRF1–29 derivatives to serum albumin occurs after a single bolus sc administration and is a useful tool to increase the plasma residence time of these particular bioactive peptides. The results reported herein demonstrate that a hGRF-albumin conjugate is able to interact with, and activate, the rat GRF receptor on the anterior pituitary to produce a high level of GH in plasma over time, leading to the discovery of CJC-1295 as a long-lasting hGRF analog.
Acknowledgments
The authors are grateful to the following people: Julie Carette and Peter Bakis for their skillful assistance in the preparation of the peptides, Omar Quraishi for providing the HSA bioconjugates, Dominador Calamba for his technical expertise in the RIA, and Nadia Sourial for performing the statistical analysis.
References
Giustina A, Veldhuis JD 1998 Pathophysiology of the neuroregulation of growth hormone secretion in experimental animals and the human. Endocr Rev 19:717–797
Frohman LA, Jansson JO 1986 Growth hormone-releasing hormone. Endocr Rev 7:223–253
Frohman LA, Downs TR, Williams TC, Heimer EP, Pan Y-CE, Felix AM 1986 Rapid enzymatic degradation of growth hormone-releasing hormone by plasma in vitro and in vivo to a biologically inactive product cleaved at the NH2 terminus. J Clin Invest 78:906–913
D’Antonio M, Louveau I, Esposito P, Bertolino M, Canali S 2004 Pharmacodynamic evaluation of a PEGylated analogue of human growth hormone releasing factor in rats and pigs. Growth Horm IGF Res 14:226–234
Esposito P, Barbero L, Caccia P, Caliceti P, D’Antonio M, Piquet G, Veronese FM 2003 PEGylation of growth hormone-releasing hormone (GRF) analogues. Adv Drug Deliv Rev 55:1279–1291
Vance ML, Kaiser DL, Evans WS, Furlanetto R, Vale W, Rivier J, Thorner MO 1985 Pulsatile growth hormone secretion in normal man during a continuous 24-hour infusion of human growth hormone releasing factor (1–40). Evidence for intermittent somatostatin secretion. J Clin Invest 75:1584–1590
Warnecke A, Kratz F 2003 Maleimide-oligo(ethylene glycol) derivatives of camptothecin as albumin-binding prodrugs: synthesis and antitumor efficacy. Bioconjug Chem 14:377–387
Kratz F, Muller-Driver R, Hofmann I, Drevs J, Unger C 2000 A novel macromolecular prodrug concept exploiting endogenous serum albumin as a drug carrier for cancer chemotherapy. J Med Chem 43:1253–1256
Kratz F, Warnecke A, Scheuermann K, Stockmar C, Schwab J, Lazar P, Drückes P, Esser N, Drevs J, Rognan D, Bissantz C, Hinderling C, Folkers G, Fichtner I, Unger C 2002 Probing the cysteine-34 position of endogenous serum albumin with thiol-binding doxorubicin derivatives. Improved efficacy of an acid-sensitive doxorubicin derivative with specific albumin binding properties compared to that of the parent compound. J Med Chem 45:5523–5533
Kim J-G, Baggio LL, Bridon DP, Castaigne J-P, Robitaille MF, Jette L, Benquet C, Drucker DJ 2003 Development and characterization of a glucagon-like peptide 1-albumin conjugate: the ability to activate the glucagon-like peptide 1 receptor in vivo. Diabetes 52:751–759
Léger R, Thibaudeau K, Robitaille M, Quraishi O, van Wyk P, Bousquet-Gagnon N, Carette J, Castaigne J-P, Bridon DP 2004 Identification of CJC-1131-albumin bioconjugate as a stable and bioactive GLP-1(7–36) analog. Bioorg Med Chem Lett 14:4395–4398
Léger R, Benquet C, Huang X, Quraishi O, van Wyk P, Bridon D 2004 Kringle 5 peptide-albumin conjugates with anti-migratory activity. Bioorg Med Chem Lett 14:841–845
Léger R, Robitaille M, Quraishi O, Denholm E, Benquet C, Carette J, van Wyk P, Pellerin I, Bousquet-Gagnon N, Castaigne J-P, Bridon D 2003 Synthesis and in vitro analysis of atrial natriuretic peptide-albumin conjugates. Bioorg Med Chem Lett 13:3571–3575
Holmes DL, Thibaudeau K, L’Archêveque B, Milner PG, Ezrin AM, Bridon DP 2000 Site specific 1:1 opioid:albumin conjugate with in vitro activity and long in vivo duration. Bioconjug Chem 11:439–444
Baird A, Wehrenberg WB, Ling N 1986 Relative potencies of human, rat, bovine/caprine, porcine and ovine hypothalamic growth hormone-releasing factors to release growth hormone by the rat anterior pituitary in vitro. Neuroendocrinology 42:273–276
Cervini LA, Donaldson CJ, Koerber SC, Vale WW, Rivier JE 1998 Human growth hormone-releasing hormone hGHRH(1–29)-NH2: systematic structure-activity relationship studies. J Med Chem 41:717–727
Su C-M, Jensen LR, Heimer EP, Felix AM, Pan Y-CE, Mowles TF 1991 In vitro stability of growth hormone releasing factor (GRF) analogs in porcine plasma. Horm Metab Res 23:15–21
Campbell RM, Stricker P, Miller R, Bongers J, Liu W, Lambros T, Ahmad M, Felix AM, Heimer EP 1994 Enhanced stability and potency of novel growth hormone-releasing factor (GRF) analogues derived from rodent and human GRF sequences. Peptides 15:489–495
Bongers J, Heimer EP, Lambros T, Pan Y-CE, Campbell RM, Felix AM 1992 Degradation of aspartic acid and asparagine residues in human growth hormone-releasing factor. Int J Pept Protein Res 39:364–374
Campbell RM, Bongers J, Felix AM 1995 Rational design, synthesis, and biological evaluation of novel growth hormone releasing factor analogues. Biopolymers 37:67–88
Wilfinger WW, Larsen WJ, Downs TR, Wilbur DL 1984 An in vitro model for studies of intercellular communication in cultured rat anterior pituitary cells. Tissue Cell 16:483–497
Aleppo G, Moskal 2nd SF, De Grandis PA, Kineman RD, Frohman LA 1997 Homologous down-regulation of growth hormone-releasing hormone receptor messenger ribonucleic acid levels. Endocrinology 138:1058–1065
Sugihara H, Emoto N, Tamura H, Kamegai J, Shibasaki T, Minami S, Wakabayashi I 1999 Effect of insulin-like growth factor-I on growth hormone-releasing factor receptor expression in primary rat anterior pituitary cell culture. Neurosci Lett 276:87–90
Rafferty B, Poole S, Clarke R, Schulster D 1985 Growth hormone-releasing factor analogue (hGRF1–29NH2): immunoreactive-GRF plasma levels after intravenous and subcutaneous administration. J Endocrinol 107:R5–R8
Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685
Zhou XH, Dinh P 2005 Nonparametric confidence intervals for the one- and two-sample problems. Biostatistics 6:187–200
Peters Jr T 1995 All about albumin: biochemistry, genetics and medical applications. New York: Academic Press; 230–232
Turner JP, Tannenbaum GS 1995 In vivo evidence of a positive role for somatostatin to optimize pulsatile growth hormone secretion. Am J Physiol 269:E683–E690
Sugihara H, Minami S, Okada K, Kamegai J, Hasegawa O, Wakabayashi I 1993 Somatostatin reduces transcription of the growth hormone gene in rats. Endocrinology 132:1225–1229
Leung K, Rajkovic IA, Peters E, Markus I, Van Wyk JJ, Ho KK 1996 Insulin-like growth factor I and insulin down-regulate growth hormone (GH) receptors in rat osteoblasts: evidence for a peripheral feedback loop regulating GH action. Endocrinology 137:2694–2702
Vance ML, Kaiser DL, Martha Jr PM, Furlanetto R, Rivier J, Vale W, Thorner MO 1989 Lack of in vivo somatotroph desensitization or depletion after 14 days of continuous growth hormone (GH)-releasing hormone administration in normal men and a GH deficient boy. J Clin Endocrinol Metab 68:22–28
Kovács M, Halmos G, Groot K, Izdebski J, Schally AV 1996 Chronic administration of a new potent agonist of growth hormone-releasing hormone induces compensatory linear growth in growth hormone-deficient rats: mechanism of action. Neuroendocrinology 64:169–176
Kovács M, Fáncsik A, Mez? I, Teplán I, Flerkó B 1994 Effects of continuous and repetitive administration of a potent analog of GH-RH(1–30)NH2 on the GH release in rats. Neuroendocrinology 59:371–379
Dubreuil P, Brazeau P, Moreau S, Farmer C, Coy D, Abribat T 2001 The use of pigs as an animal model to evaluate the efficacy, potency and specificity of two growth hormone releasing factor analogues. Growth Horm IGF Res 11:173–186(Lucie Jetté, Roger Léger,)