The authors state that they have no conflicts of interest.
Functional Effects of Monoclonal Antibodies to the Purified Amino-Terminal Extracellular Domain of the Human Ca2+ Receptor†
Article first published online: 22 JAN 2007
Copyright © 2007 ASBMR
Journal of Bone and Mineral Research
Volume 22, Issue 4, pages 601–608, April 2007
How to Cite
Hu, J., Reyes-Cruz, G., Goldsmith, P. K., Gantt, N. M., Miller, J. L. and Spiegel, A. M. (2007), Functional Effects of Monoclonal Antibodies to the Purified Amino-Terminal Extracellular Domain of the Human Ca2+ Receptor. J Bone Miner Res, 22: 601–608. doi: 10.1359/jbmr.070111
- Issue published online: 4 DEC 2009
- Article first published online: 22 JAN 2007
- Manuscript Accepted: 19 JAN 2007
- Manuscript Revised: 28 NOV 2006
- Manuscript Received: 25 SEP 2006
- human Ca2+ receptor;
- extracellular domain;
- Venus flytrap-like domain;
- monoclonal antibodies;
- chimeric receptor
We generated three functionally unique monoclonal antibodies to the purified human CaR extracellular domain. Flow cytometry studies of chimeric receptors localized their epitopes to lobe 2 of the VFT domain. These results lead us to propose a mechanism for the functional effects of these antibodies.
Introduction: The human Ca2+ receptor (CaR), which plays a central role in the regulation of [Ca2+]0 homeostasis, has a distinctively large extracellular domain that consists of a bilobed Venus flytrap (VFT) domain, involved in agonist binding, and a cysteine-rich domain. Functional antibodies that specifically bind to this domain would have therapeutic potential and could be used as a tool to gain insights into receptor activation as well.
Materials and Methods: We generated three monoclonal antibodies (mAbs), 7F8, 5C8, and 1A8, to the purified human CaR extracellular domain. Functional characterization of these antibodies included Ca2+ stimulation of phosphoinositide hydrolysis to examine effects of intact or protease digested antibodies on sensitivity of the receptor to extracellular Ca2+ and flow cytometry assay of binding of the antibodies to HEK-293 cells expressing chimeric receptors to map antibody epitopes.
Results: We found these mAbs specifically recognize native but not denatured human CaR or homologous native Fugu CaR. Sensitivity of the human CaR to extracellular calcium was increased by binding of 5C8 but decreased by binding of 1A8. A chimeric receptor FCFCF, with lobe 2 region of the human CaR VFT domain in the Fugu CaR backbone, bound all three mAbs, and the sensitivity of this chimeric CaR to extracellular Ca2+ was also increased by binding of 5C8 and decreased by binding of 1A8.
Conclusions: The epitopes of these mAbs reside in the lobe 2 region of the human CaR VFT domain. 5C8 might activate the receptor by facilitating closure and/or rotation of the VFT domains on agonist binding, whereas 1A8 might inhibit the receptor by impeding such agonist-induced conformational changes. Recombinant antibodies with antigen binding domains of 5C8 and 1A8 could be useful in the treatment of hyperparathyroidism and osteoporosis, respectively.
The ca2+ receptor plays a central role in the regulation of [Ca2+]0 homeostasis.(1,2) It is a member of family 3 of the G protein–coupled receptor (GPCR) superfamily that also includes eight subtypes of metabotropic glutamate receptors (mGluRs), two subtypes of GABAB receptors, and certain taste and pheromone receptors.(3) Family 3 members are typically characterized by a large extracellular amino terminus comprised of a bilobed Venus flytrap (VFT) domain, homologous to bacterial periplasmic amino acid binding proteins and involved in agonist binding, and a cysteine-rich domain, with the exception of GABAB receptors that lack such a domain.
The 3D structures of the mGluR1 VFT with or without ligand binding show that it is an intermolecular disulfide-linked homodimer.(4) Agonist binding to the cleft of the VFT leads to closure of the two lobes and a 70° rotation of one protomer relative to the other about an axis perpendicular to the dimer interface. Current models of human CaR VFT domain are based on these crystal structures of the mGluR1 VFT. It is noteworthy that most naturally occurring activating mutations identified in the human CaR VFT in subjects with autosomal dominant hypocalcemia (ADH) appear at the dimer interface in our homology model of the CaR.(5) We speculated that these ADH mutations cause activation of the CaR by facilitating the agonist-induced conformational changes in the VFT domain. However, at this point, how the signal of conformational changes in the VFT domain is transmitted through the cysteine-rich domain to the seven-transmembrane domain (7TM) leading to receptor activation remains an open question.
Functional monoclonal antibodies (mAbs), which have exquisite specificity for their antigens, not only have therapeutic potential but also offer insights into structure and function of the proteins they bind to. We reported earlier that we purified the dimeric extracellular domain of the human calcium receptor from the medium of HEK-293 cells stably transfected with a human CaR cDNA containing an isoleucine 599 nonsense mutation.(6) Here we report the generation of monoclonal antibodies to this purified human CaR extracellular domain (ECD) and characterization of different functional effects of these mAbs on response of the receptor to [Ca2+]0.
MATERIALS AND METHODS
Monoclonal antibodies to the human CaR ECD
Two Balb/c mice were immunized with ECD antigen, 1:1 in Titer Max. Mice were boosted 4 and 7 weeks later. Test bleeds were taken before each boost. The first in a series of boosts before a fusion was fatal to mouse 2. The splenocytes from mouse 1 were isolated, fused with NSO myeloma cells, and seeded on ten 96-well plates by standard procedure.(7) Positive-growth wells were screened for antibody production by ELISA on ECD-coated plates. Hybrids positive by ELISA were expanded into 24-well plates, rescreened by ELISA and isotyped. Six hybrids were positive as assessed by flow cytometry, and hybrids 7F8, 5C8, and 1A8 were cloned by two rounds of limiting dilution at one cell/well and 5 cells/well. Clones and subclones were screened against ECD by ELISA. One cell line for each hybrid was chosen for ascites production. Ascites were purified by Protein-A affinity chromatography, dialyzed against PBS, and 0.2 μm filtered. Concentration was measured by A280, and purity was assessed by high-performance liquid chromatography (HPLC).
Fab and F(ab)2 fragments of 5C8 and 1A8 were prepared by using the ImmunoPure IgG1 Fab and F(ab\E ′)2 preparation kit (Pierce Biotechnology, Rockford, IL, USA) according to the manufacturer's directions. Briefly, the intact antibody was digested with ficin under conditions optimized for production of Fab or F(ab)2 fragments from murine IgG1, and the fragments were separated from whole IgG and Fc fragments by affinity chromatography with immobilized protein A (Pierce Biotechnology). Concentrations of Fab and F(ab)2 fragments were determined by the modified Bradford method (Bio-Rad). SDS-PAGE of the F(ab)2 and Fab fragments followed by staining with Coomassie brilliant blue showed major protein bands of the expected molecular weights.
Construction of mutant and chimeric CaRs
The full-length human CaR (hCaR) cDNA cloned in the pCR3.1 expression vector was described previously.(8) Site-directed mutagenesis was used to insert HA tag into loop 3 region of VFT in CaR constructs. Our previous study showed that as many as 21 residues (365–385) of the protruding loop 3 region of the human CaR (residue 356–416) could be deleted without impairing receptor expression or activation.(9) Our results indicate that the loop 3 regions in both the human CaR and Fugu CaR are tolerant of insertion of a HA tag. The nine residue HA tag (YPYDVPDYA) was inserted after residue 371 of the human CaR and residue 367 of the Fugu CaR by using the Quickchange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA), according to the manufacturer's instructions. Parental CaR cDNA in pCR3.1 vector was amplified using pfu Turbo DNA polymerase with mutagenic oligonucleotide primers (sequences available on request) for 16 cycles in a DNA thermal cycler (Perkin-Elmer, Norwalk, CT, USA). After digestion of the parental DNA with DpnI for 1 h, the amplified DNA with incorporated nucleotide substitution was transformed into E. coli (DH-5α strain). Chimeric receptors were constructed by either subcloning approach or multistep PCR strategy using overlapping primers.(10) Chimera L2a contains human CaR lobe 2a region (residues 187–325) in the Fugu CaR backbone, chimera L2b contains human CaR lobe 2b region (residues 491–536) in the Fugu CaR backbone, and chimera FCFCF contains both human CaR lobe 2a and 2b region in the Fugu CaR backbone. The sequence of mutant receptors was confirmed by automated DNA sequencing using a dRhodamine Terminator Cycle Sequencing kit and ABI PRISM-373A DNA sequencer (PE Applied Biosystems, Foster City, CA, USA).
Transient transfection of wildtype and mutant receptors in HEK-293 cells
Transfections were performed using 12 μg plasmid DNA for each transfection in a 75-cm2 flask of HEK-293 cells. DNA was diluted in serum-free DMEM (BioFluids, Rockville, MD, USA), mixed with diluted Lipofectamine (Invitrogen, Carlsbad, CA, USA), and the mixture was incubated at room temperature for 30 minutes. The DNA-Lipofectamine complex was further diluted in 6 ml serum-free DMEM and was added to 80% confluent HEK-293 cells plated in 75-cm2 flasks. After 5 h of incubation, 15 ml complete DMEM containing 10% FBS (BioFluids, Rockville, MD, USA) was added. Twenty-four hours after transfection, transfected cells were split and cultured in complete DMEM. The transfected cells were assayed 48 h after transfection.
Phosphoinositide hydrolysis assay
Phosphoinositide (PI) hydrolysis assay has been described previously.(8) Briefly, 24 h after transfection, transfected cells from a confluent 75-cm2 flask were split and plated in two 12-well plates in complete DMEM medium containing 3.0 μCi/ml of [3H]myoinositol (New England Nuclear, Beverly, MA, USA) and cultured for another 24 h. Culture medium was replaced by 1>PI buffer (60 mM NaCl, 2.5 mM KCl, 2.8 mM glucose, 0.2 mM MgCl2, 10 mM LiCl in 12.5 mM PIPES buffer, pH 7.2) and incubated for 1 h at 37°C. After removal of PI buffer, cells were incubated for an additional 1 h with different concentrations of Ca2+ in 1>PI buffer with or without treatment of mAbs. The reactions were terminated by addition of 1 ml of acid-methanol (1: 1,000 vol/vol) per well. Total inositol phosphates were purified by chromatography on Dowex 1-X8 columns, and radioactivity for each sample was counted with liquid scintillation counter. Graphs of concentration dependence for stimulation of PI hydrolysis by [Ca2+]0 for each transfection were drawn by using GraphPad Prism version 2.0 software. Each value on a curve is the mean of duplicate determinations. Graphs shown in this paper are representative of at least three independent experiments.
Confluent cells in 6-well plates were rinsed with ice-cold PBS and scraped on ice in lysis buffer containing 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, and freshly added protease inhibitors cocktail (Boehringer Mannheim, Indianapolis, IN, USA). For immunoblotting of full-length receptors, 50 μg of protein per lane reduced with β-mercaptoethanol (5%) was separated on 5% SDS-PAGE gel. The proteins on the gel were electrotransferred onto nitrocellulose membrane and incubated with 0.1 μg/ml of protein A-purified mouse monoclonal anti-hCaR antibody ADD (raised against a synthetic peptide corresponding to residues 214–235 of hCaR protein) or 0.5 μg/ml monoclonal antibodies raised against purified hCaR ECD. Subsequently, the membrane was incubated with a secondary goat anti-mouse antibody conjugated to horseradish peroxidase (Amersham Pharmacia Biotech) at a dilution of 1:2000. The hCaR protein was detected with an enhanced chemiluminescence System (ECL; Amersham Pharmacia Biotech).
Forty-eight hours after transfection, one million HEK-293 cells were stained with the control versus test antibodies in a volume of 100 μl PBS. Staining was performed for detection of HA-tagged human CaR, Fugu CaR, or chimeric receptors with monoclonal antibodies 7F8, 5C8, or 1A8 at the concentration of 2 μg/million cells. Goat polyclonal anti-mouse IgG F(ab)2 fragment conjugated with phycoerythrin (PE; Novus Biologicals, Littleton, CO, USA) was used at the concentration of 0.1 μg/million cells for secondary staining. The cells were additionally stained with fluorescein isothiocyanate (FITC)-conjugated, monoclonal anti-HA (clone HA-7; Sigma, St Louis, MO, USA) at a concentration of 0.1 μg/million cells. Mouse isotypic IgG1κ-antibodies were used for primary and secondary staining controls. Cellular size and granularity were used for gating, and a minimum of 15,000 cells were analyzed in each sample. A Coulter Epics Elite Flow cytometer (Beckman Coulter, Hialeah, FL, USA) was used for all analyses. Cells having fluorescence levels at least 2 SD above the isotypic controls were scored as being positive.
Generation of monoclonal antibodies against purified human CaR ECD
We generated three mouse mAbs, 7F8, 5C8, and 1A8, to purified hCaR ECD. The subtypes of these three mAbs were determined as IgG1. Immunoblotting experiments under reducing conditions indicated that none of these antibodies, at the concentration of 0.5 μg/ml, recognized the denatured human CaR blotted on nitrocellulose membrane (Fig. 1), suggesting that the epitope of these antibodies are likely not linear. In comparison, another monoclonal antibody, ADD, raised against a synthetic peptide corresponding to residues 214–235 of the human CaR, at a concentration of 0.1 μg/ml, detected two major bands in size of ∼130 and 150 kDa for the wildtype (WT) human CaR (Fig. 1, right). The ∼150-kDa band represents the receptor forms expressed at the cell surface and modified with N-linked, complex carbohydrates, and the ∼130-kDa band represents high mannose-modified forms, trapped intracellularly and sensitive to Endo-H digestion.(8) mAbs 7F8, 5C8, and 1A8 efficiently bound native human CaR in immunoprecipitation experiments. We published earlier the use of 7F8 in immunoprecipitation of biotin-labeled human CaR expressed at the cell surface.(11,12) We also observed that chimeric receptor Ca-Glu-Ca with mGluR1 cysteine-rich domain in human CaR backbone(11) was efficiently bound by 7F8, 5C8, and 1A8 in immunoprecipitation experiments (data not shown). Because there are 40 residue differences in the ∼60 amino-acid long cys-rich region between human CaR and mGluR1, the epitopes of these three mAbs likely reside in the VFT and not the cys-rich domain. Consistent with this speculation, 7F8 efficiently bound a mutant human CaR with deletion of entire cys-rich domain in an immunoprecipitation experiment.(11)
Effects of the monoclonal antibodies on response of human CaR to Ca2+
Because the VFT domain of the human CaR is the agonist-binding domain, which undergoes structural changes on agonist binding to the cleft of two lobes of the VFT, we tested if the binding of these mAbs to the VFT has any functional effects on the response of the receptor to Ca2+. We measured PI hydrolysis as a function of extracellular Ca2+ concentration in HEK-293 cells transfected with WT human CaR cDNA in the presence or absence of these monoclonal antibodies. We found that, at a concentration of 10 μg/ml, 7F8 exhibited minimum effects. However, interestingly, 5C8 increased sensitivity of the receptor to Ca2+, whereas 1A8 decreased the sensitivity of the receptor to Ca2+ (Fig. 1, left). The effective concentration for 50% maximal response (EC50) value for the WT human CaR in the absence of antibody treatment was 3.08 ± 0.06 (SE) mM (n = 5), whereas the EC50 value for the receptor with 5C8 treatment was 0.92 ± 0.08 mM and the EC50 value for the receptor with 1A8 treatment was 6.64 ± 0.12 mM. The same amount of isotype control IgG1 antibody did not affect calcium response by the HEK-293 cells transfected with the WT human CaR or other CaR constructs mentioned below (data not shown).
Functional activities of 5C8 and 1A8 did not depend on the Fc portion or the covalent bonds between two Fab in these mAbs
Fab and F(ab)2 fragments of 5C8 and 1A8 were prepared by ficin digestion, and the fragments were separated from whole IgG and Fc fragments by affinity chromatography with immobilized protein A. SDS-PAGE followed by staining with Coomassie brilliant blue showed major Fab and F(ab)2 bands of the expected molecular weights (Fig. 2, right). When tested at the same concentration (10 μg/ml), Fab and F(ab)2 fragments of 5C8 stimulate response of the human CaR to [Ca2+]0, whereas Fab and F(ab)2 fragments of 1A8 retain inhibitory effects (Fig. 2, left). These results suggest that the Fc portion of the antibody is not required for the function of 5C8 or 1A8. Although gel staining shows some contaminant Fab in the F(ab)2 preparation as well as some F(ab)2 in the Fab preparation, the potent activating effects of the 5C8 Fab sample and inhibitory effects of the 1A8 Fab sample suggested that monomeric Fab fragments of these mAbs are capable of exerting the same functional effects as the intact antibodies.
7F8, 5C8, and 1A8 did not bind to Fugu CaR
Human CaR and Fugu CaR share substantial sequence homology, and a chimeric receptor with Fugu CaR cys-rich domain in human CaR backbone, causing 20 amino acid changes in the cys-rich domain, was fully functional.(11) However, none of these mAbs bound to Fugu CaR in our ELISA (data not shown) and flow cytometry experiments (Figs. 3C, 4, and 5), and thus neither 5C8 nor 1A8 affected response of the Fugu CaR to Ca2+ in PI hydrolysis assays (data not shown). We hypothesize that the nonlinear epitope of these mAbs likely comprise the human CaR residues in the VFT different from those of the Fugu CaR.
7F8, 5C8, and 1A8 bind to chimeric receptor FCFCF with human CaR lobe2 sequence in Fugu CaR backbone
In an effort to locate the epitopes of these three mAbs, we constructed chimeric receptors with human CaR sequences in Fugu CaR backbone. We found that a chimeric receptor, FCFCF, with human CaR lobe2 sequence in Fugu CaR backbone exhibited excellent binding of the three mAbs (Figs. 3C and 5), whereas other chimeras, including the one with human CaR lobe 1 sequence in the Fugu CaR backbone, were not recognized by any one of these mAbs (data not shown). Because lobe2 of the human CaR VFT is made up of two fragments of a sequence that are termed lobe2a and lobe2b (Fig. 3C), we further constructed chimeric receptors with either human CaR lobe2a or lobe2b in the Fugu CaR backbone. Lobe2b chimera was expressed well at the surface, but none of the three mAbs bound to the receptor as shown in our flow cytometry results (Fig. 5), leaving lobe2a a promising site for binding of these mAbs. However, the lobe2a chimera was barely expressed at the cell surface (Fig. 5). Interestingly, however, mAbs 7F8 and 5C8 but not 1A8 exhibited some degrees of binding to this poorly expressed chimera (Fig. 5).
Effects of the monoclonal antibodies on response of chimeric CaR FCFCF to Ca2+
Because the chimeric CaR FCFCF was well expressed at the cell surface and all three mAbs bound to the mutant (Fig. 5), we set to examine if these mAbs affect response of the FCFCF receptor to Ca2+ in the pattern seen for WT human CaR. Figure 6 shows that, compared with the WT human CaR (Figs. 1 and 2), the basal response of the FCFCF chimera to Ca2+ was increased. Similar to the effects on the WT hCaR, 5C8 activated the mutant receptor, 1A8 inhibited the receptor (to the extent less than it inhibited the WT), whereas 7F8 did not significantly alter function of the receptor. These results confirmed that the epitopes of the three mAbs reside in the lobe 2 region of the human CaR VFT and suggested similar, if not identical, folding of human and Fugu CaR, which allowed action of 5C8 and 1A8 in the context of FCFCF chimera.
The Ca2+ receptor is a member of family 3 GPCRs. Crystal structures of the mGluR1 VFT domain in both free and agonist-bound forms indicate that it folds as a bilobed homodimer.(4) Figure 3A is a schematic diagram showing human CaR in its inactive and active forms. On agonist binding to the cleft of the two lobes in the VFT, the two lobes close and one protomer rotates ∼70° relative to the other about the axis perpendicular to the dimer interface. Figure 3B shows our homology model of the dimeric human CaR VFT. We built our model of human CaR VFT based on the crystal structures of the mGluR1 VFT and found that most naturally occurring activating mutations of the human CaR ECD aggregate in the dimer interface of both lobe 1 and lobe 2.(5) We speculated that these mutations activate the receptor by facilitating closure and/or rotation of the VFT domains.
We reported earlier expression, purification, and biochemical characterization of the human CaR ECD. Here we report three mAbs generated against this purified ECD. Among them, 5C8 and 1A8 exert functional effects on receptor activation. 5C8 activates the receptor, whereas 1A8 inhibits the receptor. The Fab fragments of 5C8 and 1A8 exerted similar effects to the intact antibodies. It is unlikely, based on these results, that the functional effects of these antibodies on the receptor needs simultaneous binding of 1 molecule of antibody to both protomers of the human CaR dimer.
By testing chimeric receptors, we found that the epitope of these antibodies reside in the lobe 2 region of the VFT. 5C8 might activate the receptor by facilitating closure and/or rotation of the VFT domains, whereas 1A8 might inhibit the receptor by impeding such agonist-induced conformational changes. Our flow cytometry results further suggest that the epitopes of 7F8 and 5C8, but not 1A8, might reside mainly in lobe2a portion. Further studies are needed to delineate the residues comprising the epitope of these mAbs.
Because the CaR plays a central role in the regulation of Ca2+ homeostasis, agents that specifically alter function of the CaR are potential drugs for treatment of disorders of calcium metabolism. Small molecule allosteric modulators that target specifically the 7TM of the human CaR are vigorously explored as potential therapeutic agents.(13,14) Positive allosteric modulators of the CaR increase CaR activation, thereby decreasing secretion of PTH, and thus are potential for treatment of primary and secondary hyperparathyroidism. Among them, Cinacalcet is the first-in-class GPCR allosteric modulator approved by the FDA recently for treatment of secondary hyperparathyroidism in patients with chronic kidney disease on dialysis and hypercalcemia in patients with parathyroid cancer. On the other hand, negative allosteric modulators of the CaR decrease receptor activation, thereby stimulating endogenous PTH secretion. This potentially offers a novel method for the treatment of osteoporosis. Our results show that mAbs 5C8 and 1A8 specifically bind to lobe 2 of the human CaR VFT and alter the response of the receptor to Ca2+. 5C8 potentiates response of the CaR to Ca2+, mimicking the effect of positive allosteric modulators of the CaR, whereas 1A8 decreases sensitivity of the receptor to Ca2+, mimicking the effect of negative allosteric modulators of the CaR. Tremendous progress has been made in recent years in the therapeutic application of functional antibodies.(15) By chimerization or humanization strategies, recombinant antibodies with antigen-binding domains of 5C8 and 1A8 could be useful in treatment of hyperparathyroidism and osteoporosis, respectively.
This research was supported, in part, by the Intramural Research Program of the NIDCD and NIDDK, National Institutes of Health, and through a cooperative research and development agreement with NPS Pharmaceuticals. The authors thank Stefano Costanzi for providing a picture of our human CaR VFT model.