| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cell and Cancer Biology Branch and Vascular Biology Faculty, National Cancer Institute (A.M., T.W.M., F.C.); Section of Pharmacology, National Institute of Mental Health (C.B.); and Laboratory of Retinal Cell and Molecular Biology, National Eye Institute (L.N.), National Institutes of Health, Bethesda, Maryland 20892; and Department of Chemical Sciences, San Pablo-Centro de Estudios Universitarios (CEU) University (M.J.), 28668 Madrid, Spain
Address all correspondence and requests for reprints to: Dr. Alfredo Martínez, Cell and Cancer Biology Branch, National Cancer Institute, National Institutes of Health, Building 10, Room 13N262, Bethesda, Maryland 20892. E-mail: martinea{at}mail.nih.gov.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These activities are mediated by a receptor system encompassing the calcitonin receptor-like receptor (CL), a receptor activity-modifying protein (RAMP), and the receptor component protein (2, 3). Three RAMPs have been identified whose coexpression with CL results in different binding affinities, with RAMP1 eliciting a characteristic type 1 CGRP receptor response, whereas RAMP2 or RAMP3 results in a specific AM receptor (2).
AM levels are dysregulated in many human pathologies, such as hypertension, heart failure, sepsis, cancer, and diabetes (1). This observation together with experimental data in animal model systems suggest an involvement of this molecule in the pathophysiology of such conditions. Modifications of AM levels have apparently paradoxical effects on a patient’s health. For instance, elevated AM expression exerts a protective role in renal and cardiovascular diseases (4), sepsis (5), and central nervous system ischemia (6). In other circumstances, elevated AM expression worsens the progression of type 2 diabetes and cancer (7, 8).
Given this scenario, small molecules that regulate the physiological effects of AM may constitute attractive pharmacological tools for disease management. Several peptidic antagonists have been proposed, including monoclonal antibodies (9), and inhibitory peptide fragments such as AM22–52, AM16–31, and AM11–26 (10, 11). Although these molecules are effective as research tools, they have significant limitations as potential drugs given the lack of humanized blocking antibodies and the short biological half-life of fragmentary peptides. To overcome these problems, we have developed a new screening paradigm to identify nonpeptidic molecules able to interact with AM from a large library of small molecules. Our method is based on the assumption that a blocking monoclonal antibody binds to an epitope on the peptide that is critical for receptor recognition (12). Thus, molecules that disrupt peptide-antibody binding may be good candidates for modulators of peptide-receptor interactions. In an effort to ensure that this screen did not select nonspecific modulators, we conducted a parallel library screen using neutralizing antibodies to gastrin-releasing peptide (GRP). Selected small molecules were characterized for their physiological impact on AM-mediated responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Reagents
Synthetic human AM and GRP were purchased from Peninsula Laboratories (San Carlos, CA). Synthetic CGRP and forskolin were obtained from Sigma-Aldrich Corp. (St. Louis, MO). Neutralizing monoclonal antibodies against AM and GRP were produced in-house (14, 15) and labeled with peroxidase using EZ-Link Plus Activated Peroxidase (Pierce Chemical Co., Rockford, IL).
Primary screening for AM and GRP
Human AM was attached to PVC 96-well plates by passive absorption, which involved incubating 50 µl AM (at 1 nmol/µl)/well for 1 h. To cross-link GRP into the plates, these were pretreated with glutaraldehyde as previously described (16). After discarding the coating solution, wells were blocked with 200 µl 1% BSA in PBS. After 1 h, this solution was aspirated off, and 50 µl containing 1 µM of one of the compounds of the library in PBS were added. Immediately thereafter, 50 µl peroxidase-labeled antibody (at 2.4 µg/ml) were added, and the solution was allowed to react for 1 h. After thorough washes with 1% BSA in PBS, peroxidase activity was developed using o-phenylenediamine dihydrochloride (Sigma-Aldrich Corp.) as a substrate. The reaction product was quantified with a plate reader (Spectra Rainbow; Tecan, Raleigh, NC) at 450 nm. Each plate contained several internal controls, including wells without any coating that are used to calculate nonspecific binding; wells in which no potential antagonists were added, which provided maximum binding; and wells in which the unlabeled antibody (at 1.2 µg/ml) substituted for the small molecule, as a positive inhibition control (Fig. 1B
). Each compound was added to duplicate wells in the same plate. A positive hit was defined as a compound able to significantly reduce the amount of reaction product in three independent plates. The intraassay variation was 6%, and the interassay variation was 13%. The sensitivity of the assay, as calculated with the cold antibody, was 12 nM, and the dynamic range was 12–54 nM.
|
A cell line expressing the CGRP receptor was generated by transfecting HEK 293 cells with rat CL and human RAMP1 as previously reported (19) (a gift from Dr. Debbie Hay, Hammersmith Hospital, London, UK). The analysis was performed as described above, but using CGRP instead of AM as the main agonist. In both cases, 50 µM forskolin was used as a positive control.
Inositol 1,4,5-triphosphate (IP3) and Ca2+ analysis for GRP.
The lung cancer cell line H-1299 has been shown to contain specific GRP receptors (20). The signal transduction pathway for GRP includes elevation of intracellular levels of IP3 and Ca2+, and these were investigated as previously shown (21).
Measurement of blood pressure in vivo
AM is a potent and long-lasting vasodilator (1); therefore, we expected negative modulators of AM to elevate blood pressure, and positive regulators to decrease it further. We analyzed suspected negative regulators in normotensive rats (10-wk-old Lewis/ssncr males, Science Applications International Corp., Frederick, MD; n = 20) and potential positive modulators in hypertensive animals (10-wk-old SHR males, Taconic Farms, Germantown, NY; n = 16).
Animals were anesthetized with 3% halothane, intubated, and maintained with 1% halothane in 70% nitrous oxide and 30% oxygen (VMS anesthesia machine, Matrx Medical, Inc., Orchard Park, NY) at 82 strokes/min. A PE50 catheter was placed on the right femoral artery, and arterial blood pressure was recorded through a P23XL transducer (Grass Instruments, Quincy, MA). Peptides and small molecules were injected into the right femoral vein through another catheter in the amounts of 3 ng/kg (for AM) or 20 nmol/kg (for the small molecules). All procedures were performed under a protocol approved by the NIH.
Receptor-binding assays
Binding of [125I]AM to Rat2 cells was performed as previously described (22). Briefly, 5 x 104 cells were placed in 24-well plates coated with fibronectin (20 µg/well). When a monolayer was formed, the cells were washed three times in transferrin, insulin, and selenium medium, followed by incubation with receptor-binding medium (transferrin, insulin, and selenium medium, 1% BSA, and 1 mg/ml bacitracin) with 0.2 nM [125I]AM (2200 Ci/mmol; Phoenix Pharmaceuticals, Belmont, CA) in the presence or absence of competitors (cold AM or small molecules). After 2 h at 4 C, free peptide was removed by washing three times in receptor binding medium. Peptide bound to the cells was solubilized in 0.2 N NaOH and counted in a 
-counter.
Surface plasmon resonance assays
Characterization of the binding between AM and the small molecules was performed immobilizing 3 µg AM on a CM5 sensor chip by 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC)/N- hydroxysuccinimide activation, followed by covalent amino coupling of the peptide to the surface, using BIAcore 3000 (Piscataway, NJ). The remaining free surface was blocked with 1 mM ethanolamine, and the matrix was washed with 0.5 M NaCl solution and then reequilibrated with binding buffer (1:200 DMSO in PBS). Eight different dilutions of each small molecule were prepared in binding buffer with concentrations ranging from 0–10 µM and injected from low to high concentration. Each injection was followed by a matrix regeneration step. Mass transfer control experiments were performed by injecting the same concentration of each small molecule at different flow rates (5, 15, and 75 µl/min). The data were then fitted to several models for a kinetic analysis, and the binding constants were calculated. The best fittings were obtained with a simple 1:1 Langmuir model with drift baseline correction.
Statistics
Different treatments were compared with two-tailed t test. P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Internal controls were placed in every plate as described in Materials and Methods. In many cases, the active compounds could be identified by the naked eye, even before colorimetric quantification (Fig. 1B). The 2000 parental compounds of the library were screened using this methodology for AM, and 121 of them (6%) caused a statistically significant inhibition of color intensity. To evaluate the specificity of our methodology, we screened the same compounds with a blocking monoclonal antibody against GRP, a peptide similar to AM in size and chemical characteristics. Screening the same library, 109 compounds (5.4%) were identified that inhibited color formation to a significant degree. Only five of them were also present among the molecules able to interfere with AM, indicating that, in fact, different combinations of peptide-antibody complexes pulled out distinct sets of small molecules.
Secondary screening
It is possible that not all the molecules that prevent binding between the peptide and its antibody would also modify binding between the peptide and its receptor. Therefore, all positive compounds were analyzed for their ability to modify the levels of intracellular second messengers. From the initial 121 compounds, 24 were able to significantly modulate the amount of cAMP induced by 100 nM AM in Rat2 cells. Interestingly, some of these compounds reduced the cAMP levels (negative modulators), whereas others actually elevated intracellular cAMP levels over those induced by AM alone, identifying them as positive modulators (Fig. 2A and Table 1). In the absence of AM, none of the compounds elicited any response (Fig. 2A), suggesting that the mechanism of action includes binding of the small molecule to AM rather than to the receptor. These responses were dose dependent, with significant responses seen with drug concentrations as low as 10 nM (Fig. 2B).
|
|
), suggesting that grouping compounds based on their chemical similarity could be useful to predict their potential biological activity.
The difference between AM and CGRP receptors consists only in the nature of the particular RAMP that is associated with CL (2). When the active compounds for AM were added to a CGRP receptor-containing cell in the presence of synthetic CGRP, no effect was observed, although the small molecule did modify the AM response in the same cells (Fig. 2D), thereby demonstrating the specificity of these compounds in regulating AM physiology.
In a similar approach the small molecules that were identified in the primary screening with the GRP antibody were characterized by their ability to modify IP3 or Ca2+ levels induced by synthetic GRP in cells containing its receptor (Fig. 3). Again, both positive and negative modulators were identified. As was the case with modulators of AM, the GRP-interfering small molecules by themselves did not produce any change in IP3 levels (Fig. 3A). In the Ca2+ assay, 1 nM GRP produced a marked elevation of intracellular Ca2+ in H1299 cells (Fig. 3B), but preexposure of the cells to the identified negative modulators greatly reduced the Ca2+ spike amplitude (Fig. 3C). The compounds that showed consistent behavior with either the AM or GRP system have been summarized in Table 1.
|
, A and B) compared with basal levels. These differences were 62 ± 21 mm Hg (P < 0.05) for compound 128911 and 55 ± 24 mm Hg (P < 0.05) for compound 145425. Vehicle alone (DMSO in PBS) at the same concentration did not alter blood pressure (Fig. 4A). On the other hand, when screen-selected, negative modulators of AM were injected into normotensive animals, also at 20 nmol/kg, an elevation in blood pressure was observed (Fig. 4C). In the case of compound 16311, the difference from basal levels was 127 ± 47 mm Hg (P < 0.01).
|
), allowing for a kinetic analysis of binding (Table 2). The calculated Kd values varied from 4.14 x 10–6 for compound 128911 to 7.76 x 10–9 for compound 16311. One of the molecules identified for its binding to GRP (54671) was used as a control and was shown not to bind to immobilized AM (Table 2).
|
|
), and a mass transfer limitation influence could therefore be ruled out. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As suspected, not all the molecules interfering with binding of the antibody to the peptide were active in a peptide receptor assay. In our hands, only 19.8% (for AM) and 4.6% (for GRP) of the initially identified compounds were useful for modulating receptor-mediated responses. The rest of the molecules may bind directly to the antibody or even interfere with peroxidase activity, therefore providing false positive results. This is why our method requires a secondary screening based on the production of second messengers. Nevertheless, the primary antibody-based screening process is extremely fast and considerably reduces the number of compounds that must be tested with the more expensive and time-consuming cell-based screen.
The specificity of the selected compounds toward AM binding was confirmed by dose dependency in functional assays as well as by their inability to modify CGRP receptor activity. To test the specificity of the method, we screened the same library using a different ligand/antibody complex. GRP was chosen because it has a somewhat similar molecular weight to AM and is also amidated at its carboxyl terminus (23). The primary screening of the library with this neutralizing antibody identified 109 compounds, and only five of them had also been pulled out with the anti-AM antibody. This result clearly shows that this methodology is able to discriminate between target molecules. The five common compounds may bind nonspecifically, or they may specifically recognize common motifs, such as the terminal amide shared by both AM and GRP. In any case, none of these five compounds modified the second messengers for either peptide.
GRP was used in our study as a control for the specificity of the method. Nevertheless, interesting compounds able to interfere with GRP biology were identified. GRP is a 27-amino acid peptide, initially identified as the human counterpart of bombesin, a peptide found in the frog’s skin (24). GRP has antimicrobial properties, reduces food intake (25), and is involved in respiratory development (26) and the regulation of short-term memory (27). We used a neutralizing monoclonal antibody against GRP in phase I/II clinical trials of previously treated small cell lung cancer patients (28), thus indicating that efficient GRP antagonists may have clinical value.
Injection of the active compounds into experimental animals produced the expected results in blood pressure regulation, considering AM’s hypotensive function (1). In these experiments the small molecule was provided without AM, because the peptide is already present in the serum (1). The small molecules were injected at 20 nmol/kg, which is on the same order of magnitude as previously reported effective doses of AM (11). Nevertheless, more careful pharmacodynamic studies may find more effective concentrations for these compounds for the regulation of blood pressure.
To investigate the mechanism of action by which the small molecules regulate peptide activity, we performed receptor binding studies in the presence or absence of the small molecules without finding any difference. It is possible that more sensitive receptor binding assays (kinetic binding, use of cell membranes instead of whole cells, etc.) may detect subtle changes in the affinity between AM and its receptor, but we suspected that the molecular mechanism involves actual binding to the peptide. This hypothesis was supported by the fact that with our method, the peptide is the only component that is common in both screening steps. Furthermore, the addition of small molecules in the absence of peptide did not modify the intracellular response. The hypothesis was confirmed by surface plasmon resonance studies that clearly showed specific binding between AM and its small molecule regulators, with dissociation constants as low as 7.76 nM. A similar scenario has been shown for the binding between AM and its serum binding protein, complement factor H. Addition of complement factor H to AM treatment results in higher cAMP induction, higher tumor growth, and further repression of insulin release (18, 29), but the affinity between AM and its receptor remains unchanged in the presence of the binding protein (18). We are beginning to realize that the influence of factor H in AM physiology may involve protection of the peptide from protease degradation (unpublished observations), thus prolonging AM’s half-life. Similar mechanisms may explain the behavior of the small molecules described here.
Many applications can be foreseen for the molecules that regulate AM’s physiology, besides regulating blood pressure. This peptide is a survival factor for tumor cells (8), and it has been shown that inhibition of AM in vivo reduces tumor burden (30). Another disease in which AM plays a pernicious role is type 2 diabetes. AM reduces insulin secretion from ß-cells in the islets of Langerhans (14), and injection of our monoclonal antibody against AM results in a reduction of glycemia in diabetic animals (7). In both cases, nonpeptidic, negative modulators of AM may find a useful niche.
Conversely, there are other diseases in which AM plays an important protective role. These include renal and cardiovascular conditions as well as sepsis and brain ischemia (4, 5, 6). In these cases, the most probable cause of the protective impact of AM is its vasodilator activity. Here we have shown several selected compounds that augment the vasodilatory effect of AM.
Analysis of the chemical structures of some of the active compounds of AM reveals some common characteristics (Fig. 6). The most active negative modulators have in common an aromatic ring separated from a three-substituted nitrogen by four elements. There is also a hydroxy group at two or three elements from the nitrogen (Fig. 6A). The compounds with positive activity share the presence of nitrogenated heterocycles with oxygen atoms at similar distances (Fig. 6B). Before these compounds reach full clinical application, they may require further optimization. This could be attained by combinatorial chemistry, slightly modifying the chemical backbone with different radicals (31). If the three-dimensional structure of the AM-AM receptor complex is eventually resolved, it may help to identify why binding of the small molecules modifies AM physiology and to directly design modifications of our molecules to more closely fit the AM-binding site (32).
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 19, 2003.
Accepted for publication April 12, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Zudaire, S. Portal-Nunez, and F. Cuttitta The central role of adrenomedullin in host defense J. Leukoc. Biol., August 1, 2006; 80(2): 237 - 244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L. Nikitenko, N. Blucher, S. B. Fox, R. Bicknell, D. M. Smith, and M. C. P. Rees Adrenomedullin and CGRP interact with endogenous calcitonin-receptor-like receptor in endothelial cells and induce its desensitisation by different mechanisms. J. Cell Sci., March 1, 2006; 119(Pt 5): 910 - 922. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Zudaire, A. Martinez, M. Garayoa, R. Pio, G. Kaur, M. R. Woolhiser, D. D. Metcalfe, W. A. Hook, R. P. Siraganian, T. A. Guise, et al. Adrenomedullin Is a Cross-Talk Molecule that Regulates Tumor and Mast Cell Function during Human Carcinogenesis Am. J. Pathol., January 1, 2006; 168(1): 280 - 291. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
