| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
The Pacific Biosciences Research Center (A.K., G.D.B.-G.), University of Hawaii, Honolulu, Hawaii 96822; and Department of Obstetrics and Gynecology (A.I.A.), Baylor College of Medicine, Houston, Texas 77030
Address all correspondence and requests for reprints to: András Kern, Developmental and Reproductive Biology JABSOM, University of Hawaii, 651 Ilalo Street, Bioscience Building, Honolulu, Hawaii 96813. E-mail: kerna{at}pbrc.hawaii.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human relaxins H1 (RLN1) and H2 (RLN2) are well characterized and their genes (RLN1 and RLN2) are 82% identical at the amino acid level (4). The third human relaxin gene (RLN3) is the most recent to be identified, but appears to be the ancestral gene from which the others subsequently arose by gene duplication (5). RLN1 has only been identified in higher primates, suggesting that RLN1/RLN2 gene duplication event only occurred during primate evolution (6). Human relaxin H1 has a 5-fold lower affinity for binding to the relaxin receptor compared with relaxin H2 (7). It also has less bioactivity compared with relaxin H2 (8). The RLN1 gene is expressed at the mRNA level in the decidua, placenta, prostate, and breast and appears to be associated with the autocrine/paracrine roles of relaxin in these tissues (9). Although the relaxin H1 protein has been localized in the cytoplasm of decidual cells (10), it remains to be demonstrated that this protein is present in blood or prostatic fluid. In contrast, relaxin H2 is the major stored and circulating form of the hormone from the human ovary and is the classical hormone of pregnancy (11). Recent work has expanded this view of relaxin to nonreproductive functions and it now appears to be important in cardiovascular, renal, and lung disease (12, 13).
Recent studies have identified the cognate receptor of relaxin H2 as a member of subgroup C of the leucine-rich repeat (LRR) containing G protein-coupled receptors (GPCR), originally named LRR containing GPCR (LGR7) (14), and recently redesignated as relaxin family peptide receptor (RXFP) 1 (15). The LGRs are a subfamily of the rhodopsin-like GPCR family, which are mosaic receptor proteins containing a transmembrane (TM) domain and an extracellular domain with multiple LRRs important for ligand binding (16). The subgroup C contains two related receptors, LGR7 (RXFP1) and LGR8 (RXFP2). They share 60% amino acid sequence identity and are distinguished from the other LGRs by a unique low-density lipoprotein class A module (LDL-A) at the N terminus of the ectodomain, followed by 10 leucine-rich repeats and seven TM helices. The LGR8 is the primary receptor for INSL3 (17). Both receptors have putative N-glycosylation sites and activate adenylate cyclase to cause cAMP production and the stimulation of protein kinase A (5). Chimeric LGR7/LGR8 receptors were used to investigate the pharmacological characteristics of a range of relaxin family peptides. This showed that both the ectodomain and TM domain are required for optimal binding and signal transduction. These studies also revealed a two-site binding model for LGR7 and LGR8, with the LRRs forming the primary high-affinity binding site and the TM region forming a secondary low-affinity binding site for the cognate ligands relaxin H2 and INSL3 (18, 19). In a recent study, a three step model of LGR7 activation was proposed, in which the LDL-A module directs ligand-activated cAMP signaling (20).
Among the human GPCR family of receptors, only LGR7 and LGR8 contain the LDL-A module (15). The sequence similarity between the LDL-A modules of LGR7 and LGR8 and other LDL-A module-containing proteins are shown in Fig. 1
. The LDL-A modules were first described and characterized in the LDL receptor (LDLR), containing seven nonidentical repeats of this module. They are involved in sequestering and clearance of apolipoprotein B and apolipoprotein E (21). A number of other proteins also contain such LDL-A modules, implicating it in a variety of biological functions, including the very-low-density lipoprotein receptor, the LDL-receptor-related protein (LRP), renal glycoprotein gp330, the C9 component of complement, Tva the receptor for Rous sarcoma virus (22), and retina- and brain-specific NETO1 TM protein (23). Mutations in the LDL-A module of LDLR can cause familial hypercholesterolemia and led to a significant amount of work on the characterization of the structure of these modules (24, 25). A common feature of the LDL-A module is a sequence containing six conserved cysteines that form three disulfide bonds necessary to achieve proper folding (26). Another key structural element is five typically acidic or polar residues forming the conserved motif (Fig. 1). The side chains of these residues contribute to the coordination of Ca2+ binding, essential for structural and functional integrity. Mutations causing familial hypercholesterolemia frequently alter residues in this conserved calcium-binding motif. In experiments performed with isolated LDL-A modules of LDLR, these mutations altered the folding of the module without altering the native three-dimensional structure (27, 28). Recently, a recombinant LDL-A module of the human LGR7 receptor has been produced and used for NMR studies in Escherichia coli (29). The authors demonstrated that Ca2+ is required to form a stable and correctly folded structure. In addition, the LDL-A module of LGR7 requires the formation of native disulfide bonds before Ca2+ binding, consistent with the LDL-A module of LDLR.
|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DNA constructs
The full-length LGR7 cDNA was subcloned into a pCR3.1 expression vector. The LGR7 expression vector was modified by inserting the Kozak sequence (cgccaccatgg, start codon underlined), generated with a primer pair: forward LGR7Koz primer (5'-tcgctagcgatatcgccaccatggcatctggttctgtc, NheI and EcoRV restriction sites underlined) and reverse LGR7Rev609 primer (5'-cagccattctagtctgtgaagatct, BglII restriction site underlined). This generated a PCR product that could be digested with NheI and BglII. PCR was performed in a final volume of 50 µl containing 1x high fidelity buffer from Applied Biosystems (Foster City, CA), 0.2 mM of each dNTP (Applied Biosystems), primers (15 pmols), and 2.5 U high-fidelity enzyme (Applied Biosystems). PCR amplification was performed according to the following schedule: after the hot start (94 C for 3 min) 30 cycles of 94 C for 30 sec, 56 C for 40 sec, and 72 C for 60 sec, followed by final extension at 72 C for 10 min. The product was digested with NheI and BglII and cloned into the pCR3.1 vector containing the LGR7 cDNA, resulting in a LGR7 cDNA including the desired Kozak sequence (Koz-LGR7). To detect receptor expression, we inserted the hemagglutinin (HA)-tag sequence (YPYDVPDYA) into the cDNA of LGR7 after the signal sequence. The HA-tag was created by overlap extension PCR strategy (30) using the LGR7Koz primer combined with signHA-rev-LGR7 primer (5'-ggcatagtctggtacatcatacggatatccacccccatgaga, the HA-tag sequence underlined) and LGR7Rev609 combined with signHA-forw-LGR7 primer (5'- tatccgtatgatgtaccagactatgcccaggatgtcaagtgct, the HA-tag sequence underlined). The product was digested with NheI and BglII, and the fragment was cloned into the Koz-LGR7 expression vector construct generating the wild-type (WT) receptor with an HA-epitope tag at the N terminus (WT-LGR7; Fig. 1
).
The point mutants (C47A/C53A-LGR7, D58E-LGR7, and N36Q-LGR7; Fig. 1) and the deleted LDL-A module construct (
LDL-LGR7; Fig. 1) were generated in the WT-LGR7 sequence by the overlap extension PCR technique. The HA-tag was produced by overlap extension PCR (30) and inserted as for the WT-LGR7. The PCR primer pair for creating the C47A/C53A-LGR7 was: C47AC53Aforw primer (5'-gctaacggtgtggacgacgccgggaatcaggc, mutagenesis sites underlined) and C47AC53rev primer (5'-ggcgtcgtccacaccgttagcgtgcaggagct, mutagenesis sites underlined), for creating D58E-LGR7, the primer pair was: D58Eforw primer (5'-tcaggccgaggaggacaact, mutagenesis site underlined) and D58Erev primer (5'-agttgtcctcctcggcctga, mutagenesis site underlined), for creating N36Q-LGR7 the primer pair was: N36Qforw primer (5'-gtgggcagatcacaaagtgc, mutagenesis site underlined) and N36Qrev primer (5'-gcactttgtgatctgcccacaggggaa, mutagenesis site underlined) and for creating the LDL-LGR7 the primers were sign-HA-delLDLforw primer (5'- gtaccagactatgcccaggatgtcaagggagacaacaatgga, HA-tag sequence underlined) and signHA-rev-LGR7 primer.
The chimeric LDL-A/LGR8-LGR7 receptor (Fig. 1) was created by overlap extension PCR. The PCR product A was generated from a PCR containing the LGR7Koz primer and signHA-rev-LGR7 primer using the Koz-LGR7 vector as a template. The PCR product B was generated from a PCR containing LGR7Forw268 primer (5'-gcagaaacacctgaatgtttggtcggt) and LGR7Rev609 primer using the Koz-LGR7 vector sequence as a template. The PCR product C was generated from a PCR containing HALDL/8forw primer (5'-gtaccagactatgcccaggatgtcaagccttcatgccaaaaagga) and LDL/8rev primer (5'-acattcaggtgtttctgccacgctgttagcatttcca) primer using the template vector containing the cDNA sequence of LGR8 (31). In the next step, the PCR product D was generated by overlap extension PCR using the PCR product A and PCR product C. In the final PCR, the PCR product E was generated from a PCR containing PCR product D and PCR product B. The final PCR product E was digested with NheI and BglII, and the fragment was cloned into the WT-LGR7 expression vector construct generating the chimeric LDL-A/LGR8-LGR7 receptor (Fig. 1).
The receptor variant bearing the HA-tag sequence at the C terminus before the stop codon (LGR7-HA; Fig. 1) was generated by using the HA-tag sequence containing the fragment from a PCR with primer pair LGR7forw1918 (5'-tgctggatacccatttttgtag) and LGR7revHA (5'-tgcagaattcaggcatagtctggtacatcatacggatactcgagtgaataggaattgagtctcgttg, EcoRI site underlined). The PCR conditions were the same as above for generating the Koz-LGR7 construct. The PCR product was digested with KpnI and EcoRI and cloned into the Koz-LGR7 vector resulting in the LGR7-HA receptor (Fig. 1).
The C-terminal green fluorescent protein (GFP)-labeled WT receptor (WT-LGR7-GFP; Fig. 1B) was created by subcloning the LGR7 cDNA from LGR7-HA digested with NheI and XhoI into the pEGFP-N1 vector (BD Biosciences Clontech, Mountain View, CA) after digestion with NheI and XhoI. We also labeled the N36Q-LGR7 receptor with GFP on its C terminus (N36Q-LGR7-GFP) by subcloning the fragment digested with NheI and BglII from N36Q-LGR7 to the WT-LGR7-GFP. Sequencing was performed to confirm all DNA constructs and to ensure the absence of additional mutations.
Cell culture and transient transfection
HEK293 cells were maintained in DMEM supplemented with 10% fetal bovine serum and 100 µg/ml penicillin/streptomycin in a humidified atmosphere at 37 C and 95% air/5% CO2. Transient transfections of HEK293 cells grown in eight-well chambers and 24-well plates were carried out using Lipofectamine2000 transfection reagent (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. Transient transfections of HEK293 cells grown in petri dishes and T75 flasks were carried out using Effectene transfection reagent (Qiagen, Valencia, CA), according to the manufacturer’s instructions. Transfected cells were used 48 h after transfection. To correct for transfection efficiency, for intracellular cAMP determination and receptor expression assays, the receptor DNA constructs were cotransfected with 50 ng GFP expression vector (pEGFP-N1 expression vector; BD Biosciences Clontech), and data were normalized based on GFP fluorescence measured by a Victor2 fluorometer (PerkinElmer Life Sciences, Wellesley, MA).
Intracellular cAMP determination
Cells were plated into 24-well plates coated with poly-D-lysine (Becton Dickinson Labware, Frankin Lakes, NJ) and transfected with 0.5 µg of the appropriate DNA construct. Forty-eight hours later, cells were washed with PBS and pretreated with 0.25 mM 3-isobutyl-1-methylxanthine (Sigma-Aldrich) for 20 min at 37 C and stimulated for 30 min at 37 C with different concentrations of human relaxins H1 and H2 or INSL3. After treatment, cells were lysed and the intracellular cAMP was determined with the Amersham cAMP Biotrak EIA System (GE Healthcare, Little Chalfont, Buckinghamshire, UK) according to the manufacturer’s instructions. The cAMP results were expressed as the maximum response (percent) compared with the maximum stimulation of cAMP achieved with relaxin H1, relaxin H2, or INSL3. In each experiment, to obtain the maximal responses and EC50 values, curve fitting was performed using GraphPad Prism software (GraphPad Software Inc., San Diego, CA). Data were normalized based on GFP fluorescence, as described above. Each experiment was performed in duplicate and expressed as means ± SE of the mean for three to four observations.
Membrane preparations
To obtain membrane preparations, HEK293 cells were transfected with 2 µg of the appropriate DNA constructs in petri dishes. After 48 h, the cells were washed with PBS, detached with PBS containing 5 mM EDTA, and collected by centrifugation (1500 x g, 10 min at 4 C). Cells were washed twice with ice-cold Tris-mannitol buffer (50 mM Tris, pH 7.4; 300 mM mannitol; and 0.5 mM phenylmethylsulfonyl fluoride). The cells were lysed and homogenized on ice using a glass-Teflon tissue homogenizer in lysis buffer [50 mM Tris, pH 7.4; 50 mM mannitol; 2 mM EGTA; 0.5 mM phenylmethylsulfonyl fluoride; 1 mM dithiothreitol, and protease inhibitors (Complete protease inhibitor mix; Roche Diagnostic GmbH, Mannheim, Germany)]; the undisrupted cells and nuclear debris were removed by centrifugation (1000 x g, 10 min at 4 C). The supernatant was then centrifuged for 60 min at 32,000 x g, 4 C, and the pellet containing the membranes was resuspended in lysis buffer to obtain 1–2 mg/ml protein concentration. The membranes were aliquoted and stored at –70 C. The protein concentration was determined by the Bradford protein assay (Bio-Rad, Hercules, CA).
Western blot analysis
For SDS-PAGE (7.5% separating gel), membrane preparations (5 µg), were suspended in SDS-PAGE loading buffer (50 mM Tris, pH 6.8; 2% SDS; 15% glycerol; 2% ß-mercaptoethanol; 1 mM EDTA; and 0.02% bromphenol blue) and denatured by heating at 95 C for 5 min. The proteins were resolved in SDS-PAGE and were transferred electrophoretically to nitrocellulose membrane (Amersham Biosciences). Blotted samples were blocked with 5% nonfat milk (Bio-Rad) in TBS-Tween buffer (50 mM Tris, pH 7.4; 200 mM NaCl; and 0.1% Tween 20) for 1 h at room temperature. Blocked membranes were incubated with an HA antibody (Covance HA.11 monoclonal antibody, Berkeley, CA; at 1:1000 dilution) or with mouse monoclonal calnexin antibody (Abcam, Cambridge, MA; at 1:4000 dilution) for 1 h at room temperature, washed twice with TBS-Tween buffer for 15 min, then incubated with horseradish peroxidase-conjugated mouse antibody (Amersham Biosciences; at 1:2000 dilution) for 1 h at room temperature and washed three times with TBS-Tween buffer for 20 min. The proteins were visualized with the chemiluminescence Western blotting detection reagent (Amersham Biosciences).
Enzymatic digestion of membrane preparations
For glycosidase treatment, membrane preparations (5 µg) from transfected cells were denatured in glycosidase denaturing buffer (New England Biolabs) for 10 min at 95 C. Denatured samples were treated overnight with either endoglycosidase H (Endo H; New England Biolabs), to remove high mannose-type oligosaccharides or peptide N-glycosidase F (PNGase F; New England Biolabs), to remove all N-glycans, according to the manufacturer’s instructions. Proteins were analyzed by SDS-PAGE and Western blotting.
Total and cell surface expression ELISA
The expression of WT and mutant LGR7 receptors was assayed by cell surface ELISA (32) with modifications. Briefly, cells were transfected with the appropriate DNA receptor construct in 24-well plates coated with poly-lysine (Becton Dickinson Labware). Forty eight hours later, the cells were washed with TBS (20 mM Tris, pH 7.4; 150 mM NaCl; and 2 mM CaCl2) and fixed in 4% paraformaldehyde (for cell surface analysis) or 4% paraformaldehyde with 0.25% Triton X-100 (for total receptor expression). Cells were washed three times in TBS, incubated for 45 min with TBS containing 1% BSA (TBS/BSA) and incubated with HA antibody (Covance HA.11 monoclonal antibody, 1:1000 dilution) for 1 h at room temperature. Cells were washed three times with TBS, blocked with TBS/BSA for 15 min at room temperature, and incubated with secondary mouse antibody conjugated with alkaline phosphatase (Jackson ImmunoResearch, West Grove, PA; 1:2000 dilution) for 1 h at room temperature. Cells were washed four times with TBS, and a colorimetric alkaline phosphatase substrate pNPP (p-nitrophenyl phosphate; Sigma-Aldrich) was added for 30 min at room temperature in the dark. The reaction was stopped by adding 50 µl of 3 M NaOH. The color reaction was measured at 405 nm using a microplate reader (Versa Max microplate reader; Molecular Devices, Sunnyvale, CA). The value, obtained from empty vector-transfected cells was subtracted from the values obtained from cells transfected with the receptor variants. Each experiment was performed in duplicate and expressed as means ± SE of the mean for four to six observations. Statistical analysis was performed to include the variance of transfection efficiency between experiments. Preliminary calibrations experiments with the increasing amount of plasmid used for transfection ensured that surface HA expression measurements were conducted in the linear range.
Immunocytochemistry
HEK293 cells were grown in poly-D-lysine-coated eight-well chambers (Becton Dickinson Labware) and transfected with 0.3 µg DNA constructs per well using Lipofectamine2000. After 48 h, they were washed with PBS and for permeabilization were fixed for 10 min with 4% paraformaldehyde containing 0.25% Triton X-100 and washed three times with PBS. For both nonpermeabilized and permeabilized conditions, cells were blocked with PBS/BSA (PBS containing 1% BSA) for 1 h and incubated with HA antibody (Covance HA.11 monoclonal antibody, 1:1000 dilution) for 1 h and washed three times with PBS. For nonpermeabilized conditions, cells were fixed with 4% paraformaldehyde for 10 min and washed three times with PBS. Both permeabilized and nonpermeabilized cells were incubated with AlexaFluor 488 mouse antibody (Molecular Probes, Eugene, OR; 1:1000 dilution) and washed three times with PBS. Slides were mounted and laser scanning confocal microscopy was performed using a LSM-5 system (Zeiss, Oberkochen, Germany).
For colocalization of the receptor and the plasma membrane, HEK293 cells were grown in poly-D-lysine-coated eight-well chambers and transfected with WT-LGR-GFP or N36Q-LGR7-GFP constructs as above. Cells were stained after permeabilization as described above, except the first monoclonal antibody was against Na+/K+ ATPase (Upstate, Lake Placid, NY; 1:200 dilution) and the second antibody was mouse AlexaFluor 546 (Molecular Probes; 1:1000 dilution). Laser scanning confocal microscopy was performed using a Zeiss LSM-5 system.
Cell surface protein biotinylation analysis
Cell surface protein biotinylation and isolation were performed using the Pinpoint cell surface protein isolation kit (Pierce, Rockford, IL). Briefly, cells were grown in T75 flasks and transfected with 2.5 µg DNA. After 48 h they were washed twice with ice-cold PBS and incubated with 10 ml sulfo-NHS-SS-biotin solution (0.25 mg/ml) for 30 min at 4 C. Excess biotin was quenched by adding 0.5 ml quenching solution, and the cells were scraped off and collected by centrifuge at 500 x g for 3 min. Cells were washed once with TBS, solubilized with lysis buffer, sonicated for 1 sec five times, and incubated on ice for 30 min. Samples were then cleared by centrifugation at 10,000 x g for 2 min at 4 C. The supernatants were incubated with immobilized NeutrAvidin biotin binding protein beads on a rotating rack for 1 h at room temperature to precipitate the biotinylated proteins. Unbound proteins (flow-through samples) were removed by centrifugation and the beads were washed three times with washing buffer. The protein was eluted from the beads by incubation with SDS-PAGE loading buffer (supplemented with 50 mM dithiothreitol) for 1 h at room temperature (eluted samples). The eluted samples and flow-through samples were separated by SDS-PAGE and detected by Western blotting as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
LDL-LGR7 receptors in HEK293 cells, upper panel). The activity of relaxin H2 was slightly higher (EC50 = 0.07 ± 0.01 nM) compared with the relaxin H1 (EC50 = 0.15 ± 0.1 nM), these data correlated well with previous studies (33, 34). There was no cAMP response to added relaxin H2 with an empty expression vector as a control (Fig. 2A, upper panel). To test whether the HA-tag impacted LGR7 receptor function, we used an LGR7 expression construct without the HA-tag and measured the cAMP response of cells treated with relaxins H1 or H2. This had similar activity compared with the HA-tagged LGR7 receptor when treated with relaxins H1 (EC50 = 0.2 ± 0.11 nM) and H2 (EC50 = 0.1 ± 0.08 nM), indicating that incorporation of the HA-tag sequence into the N terminus had no influence on receptor function (data not shown).
|
) generating one double mutant construct, C47A/C53A-LGR7, and a second mutant construct, D58E-LGR7, thus altering the Ca2+ binding motif. A third construct was produced by deleting the LDL-A module (LDL-LGR7) (Fig. 1). Each construct was then used for the transient transfection of HEK293 cells and their functional characterization by measuring cAMP production after treatment with relaxins H1 or H2. As shown in Fig. 2A (middle and bottom panels) all three mutant constructs did not trigger the cAMP production after treatment with relaxins H1 or H2, demonstrating the importance of the LDL-A module in the signaling mechanism by these ligands.
HEK293 cells transfected with WT-LGR7, C47A/C53A-LGR7, and D58E-LGR7 expressed two molecular mass forms, analysis by treatment with endoglycosidase
The expression of WT and mutant constructs in transiently transfected HEK293 cells was tested using total membrane preparations and Western blot analysis with an antibody against the HA-tag. As seen in Fig. 2B, the HA antibody recognized two protein species with apparent molecular masses of 80 and 95 kDa in cells expressing WT-LGR7, C47A/C53A-LGR7, and D58E-LGR7, the relative intensities of which varied from one experiment to another. None of these species was detected in cells transfected with the empty vector, confirming that the detected bands represent the expressed receptors.
The identities of the two receptor species were studied further by treating the membrane preparations with endoglycosidases (Fig. 2C). The enzyme Endo H glycosidase selectively removes unprocessed high mannose-type N-linked oligosaccharides from glycoproteins, but does not cleave complex fully processed glycans (35). Treatment of the membrane-bound proteins with Endo H reduced the 80-kDa receptor species to 70 kDa in membrane preparations from cells transfected with the WT-LGR7, C47A/C53A-LGR7, and D58E-LGR7 receptors. The small decrease in the 95-kDa band after this digestion suggested the inclusion of some high-mannose complex structures. Therefore, digestion with Endo H demonstrated that the lower, 80-kDa molecular mass species represents an intact receptor polypeptide bearing high mannose-type oligosaccharides, whereas the upper 95-kDa band mostly contains complex-type N-linked oligosaccharides and, therefore, would be transported through the Golgi complex. This was further supported by the observation using PNGase F, an enzyme that removes all types of N-linked oligosaccharides from glycoproteins (35). Treatment with PNGase F increased the total electrophoretic mobility of the 95-kDa molecular mass species (upper band) and the 80-kDa molecular mass species (lower band) of the WT and mutant receptors (Fig. 2C). Thus, the 95-kDa species (upper band) represents the mature form of the LGR7 receptor, whereas the 80-kDa species (lower band) represents the immature or precursor LGR7 receptor form. Furthermore, treatment with PNGase F resulted in a 70-kDa form for the WT-LGR7, C47A/C53A-LGR7, and D58E-LGR7 receptors, representing the completely deglycosylated form of the receptors (Fig. 2C).
Deletion of the LDL-A module of the LGR7 receptor resulted in the expression of only one receptor form
We investigated whether deletion of the LDL-A module of the LGR7 receptor would affect the receptor expression in transfected HEK293 cells. Total membrane preparations were used for Western blot analysis. The HA-antibody recognized only one band of 80 kDa (Fig. 2B). Endo H treatment of membrane preparations from HEK293 cells transfected with the LDL-LGR7 receptor confirmed the identity of this 80-kDa band and showed it to be mostly resistant to Endo H digestion. Only a slight shift of the 80-kDa band was detected, which indicated that it contained some high mannose-type oligosaccharide structures (Fig. 2C). Digestion with PNGase F led to full degradation of the 80-kDa band to 60 kDa, indicating complete deglycosylation of the LDL-LGR7 receptor (Fig. 2C). These results show that only the 80-kDa form containing predominantly complex-type oligosaccharides was expressed by HEK293 cells transfected with the LDL-LGR7 receptor construct, and that this probably represents the mature receptor.
Functional characterization of the LDL-A module and its importance in ligand-specific signal transduction
To show the specificity of the LDL-A module for the signaling by relaxins H1 and H2, we created a chimeric receptor by switching the LDL-A module of LGR8 with that of the LGR7 receptor (LDL-A/LGR8-LGR7; Fig. 1). To determine whether this was expressed, we used total membrane preparations from cells transiently transfected with the chimeric LDL-A/LGR8-LGR7 receptor construct for Western blot analysis. Using an HA antibody, two molecular mass species (95 and 80 kDa) were detected (Fig. 3A), as previously identified for the WT-LGR7 (Fig. 2B). Their identities were determined by endoglycosidase treatment (Fig. 3B). Treatment with Endo H of membrane preparations from cells transfected with the chimera led to digestion of the 80-kDa species to the 70-kDa species, indicating that this contained high mannose-type oligosaccharides. The small reduction in the upper (95 kDa) band after Endo H treatment suggested it contains some high-mannose complex structures. Digestion with Endo H demonstrated that the lower (80 kDa) species (sensitive to Endo H digestion) represented the immature (precursor) receptor containing high mannose-type oligosaccharide structures, whereas the upper (95 kDa) species (mostly resistant to Endo H digestion) represented the mature receptor with complex-type oligosaccharides. We further investigated the nature of the two bands by PNGase F, which resulted in deglycosylation represented by a 70-kDa band, as for the WT-LGR7 (Fig. 3B). Thus, these data on the chimeric LDL-A/LGR8-LGR7 receptor were identical to the data shown (above) for the WT-LGR7 receptor.
|
, upper panel). The natural ligand for the LGR8 receptor is INSL3; therefore, we tested whether this could activate the chimeric receptor construct. As shown in Fig. 3C (lower panel), cells expressing this receptor showed no cAMP response to INSL3, whereas treatment of cells expressing the WT-LGR8 with INSL3 led to a dose-dependent increase in cAMP production (EC50 = 0.26 ± 0.05 nM). Furthermore, INSL3-failed to activate the WT-LGR7 receptor, as expected (Fig. 3C, lower panel). These results show that the LDL-A module alone is insufficient for specific ligand activation and suggest that, for efficient signal transduction by relaxins H1 and H2, in addition to the LRR and the TM domains, the LDL-A module is required.
The WT-LGR7, C47A/C53A-LGR7, D58E-LGR, LDL-LGR7, and LDL-A/LGR8-LGR7 receptors were delivered to the cell surface in HEK293 cells
The cell surface expression for the different receptor constructs expressed by HEK293 cells was determined by an ELISA using recognition of the HA-epitope sequence at their N termini. Amino acid substitution of the conserved cysteine residues Cys47 and Cys53 (C47A/C53A-LGR7 construct) or the conserved Asp58 (D58E-LGR7 construct) led to 102 ± 17.5% and 96 ± 9.15% cell surface expression, respectively, compared with the WT-LGR7 receptor cell surface expression level (Fig. 4A). Deletion of the LDL-A sequence from LGR7 (LDL-LGR7) led to 98 ± 6.4% of the cell surface expression compared with the WT-LGR7 constructs (Fig. 4A). However, expression of the chimeric receptor increased the cell surface delivery to 170 ± 10.5% compared with the WT-LGR7 receptor (P < 0.001) (Fig. 4A). Thus, these results show that complete deletion of the LDL-A module or alteration of the conserved Cys47, Cys53, or Asp58 residues in the LDL-A module had no effect on their cell surface delivery. However, the switching of the LDL-A module between LGR8 and LGR7 increased the cell surface delivery compared with the WT-LGR7. We determined the total receptor expression of the WT-LGR7 and mutant receptor constructs after permeabilization of cells with Triton-X. For each of the mutants compared with expression of the WT, C47A/C53A-LGR7 the total expression level was 104 ± 4.3%; for the mutant D58E-LGR7, the total expression level was 92 ± 9.6%; for LDL-LGR7, the total expression level was 78 ± 6%; and for the chimeric LDL-A/LGR8-LGR7, the total expression level was 110 ± 9.25% (Fig. 4A). The relative amounts of cell surface expression compared with the total expression were 46% for WT-LGR7, 45% for C47A/C53A-LGR7, 47% for D58E-LGR7, 57% for LDL-LGR7, and 71% for the chimeric mutant receptor.
|
). This would only be exposed to the intracellular surface of the cells and, therefore, would be undetectable with the antibody used in cell surface ELISA (Fig. 4A). We only detected the cell surface expression of the LGR7-HA in 3 ± 1% compared with the WT in intact cells (P < 0.001), whereas the total expression level of the LGR7-HA receptor was 99.5 ± 8% of the WT-LGR7 detected in Triton X-100 permeabilized cells.
We then examined the expression of the WT-LGR7 and the modified receptors by confocal microscopy. The cells were fixed under nonpermeabilized or permeabilized conditions and stained with HA antibody followed by visualization with AlexaFluor 488-conjugated secondary antibody. In nonpermeabilized HEK293 cells, the membrane localization of the WT and mutant receptors was clearly observed (Fig. 4B, upper row). However, in permeabilized HEK293 cells, the WT and mutant receptors were detected intracellularly except for the nuclei, demonstrating that a portion of these receptors was retained within the intracellular compartments (Fig. 4B, lower row). In control experiments in HEK293 cells transiently transfected with LGR7-HA constructs, no signal was detected in nonpermeabilized conditions, although we detected LGR7-HA receptor protein staining in the permeabilized cells as expected (Fig. 4B). These results support the cell surface expression data and confirm that the localization of the mutated receptors in the cell membrane was unaffected.
For further investigation of the cell surface expression of the WT and mutant receptor proteins, we transfected HEK293 cells and then assayed the cell surface proteins using biotinylation followed by capture to avidin beads and analyzed the eluant from the beads by Western blotting with the antibody against HA-tag. As shown in Fig. 4C, this resulted in the visualization of a 95-kDa band indicating that the WT-LGR7 and mutant receptors were indeed expressed at the cell surface. This result also supports the suggestion that the 95-kDa species represents the mature receptor protein expressed at the cell surface, whereas the 80-kDa protein detected by analyzing the cell membrane preparations (Figs. 2B and 4C) represents the immature (precursor) receptor form. The experiments using protein biotinylation of the cells transfected with the deleted LDL-A module variant LGR7 receptor, resulted in detection of a smaller 80-kDa band (Fig. 4C), showing that the LDL-LGR7 receptor is expressed at the cell surface and supporting the data that this represents the mature receptor form of LDL-LGR7 (Figs. 2B and 4C). The specificity of the biotinylation was demonstrated by the inability to detect the intracellular protein (calnexin) by immunoblotting in the avidin eluted samples (Fig. 4C, lower panel), whereas an antibody to calnexin recognized the strong 90 kDa band in flow-through samples (Fig. 4C, lower panel). This confirmed that the plasma membrane fraction was not contaminated by intracellular proteins.
The glycosylation mutant N36Q-LGR7 had decreased activity to relaxins H1 and H2 compared with the WT-LGR7
It was reported previously that the LDL-A module of the LGR7 receptor contains one putative N-linked glycosylation site at Asn36 (14, 36). To investigate the role of this we created the N36Q-LGR7 (Fig. 1). The N36Q-LGR7 mutant receptor expressed in HEK293 cells was tested for cAMP production after treatment with relaxins H1 and H2. As seen in Fig. 5A, this resulted in a dose-dependent decrease in cAMP production after treatment with relaxins. Relaxin H1 and H2 both caused significantly (P < 0.05) lower maximal activation of N36Q-LGR7 (61.1 ± 8.1% and 60 ± 3%, respectively) compared with the WT-LGR7. However, its cell surface delivery was compromised (Fig. 5C). When the cAMP production was normalized to receptor surface expression, the maximal cAMP response of this mutant receptor was even higher than that of the WT-LGR7 receptor (165.2 ± 21% and 162 ± 8.1% for relaxin H1 and H2 activation, respectively, compared with WT-LGR7). However, the N36Q-LGR7 mutant showed a significantly (P < 0.05) increased EC50 value after relaxin H1 activation (1.21 ± 0.2 nM) compared with that of WT-LGR7 (0.173 ± 0.08 nM). This was also significant (P < 0.05) after relaxin H2 activation (0.35 ± 0.1 nM) compared with the WT-LGR7 (0.07 ± 0.01 nM), indicating that glycosylation at Asn36 is required for optimal efficacy of relaxin action (Fig. 5A). Thus, cells transfected with this glycosylation mutant showed decreased signal transduction after treatment with relaxins H1 and H2.
|
). Immunoblot analysis also showed a decrease in the level of expression of the mutant receptor compared with the WT receptor (Fig. 5B). When the membrane preparations of the N36Q-LGR7 were treated with endoglycosydases, the lower molecular mass band was completely digested with Endo H leading to a 70-kDa band (Fig. 5B). This indicated that the lower molecular mass band contained high mannose-type oligosaccharides. The small reduction in the upper band after Endo H treatment (Fig. 5B) suggested that it also contained some high-mannose complex structures. These results suggest that the smaller band of N36Q-LGR7 represented the immature (precursor) receptor bearing high mannose-type oligosaccharide structures, whereas the upper band represented the mature receptor containing mostly complex-type oligosaccharides. Digestion with PNGase F led to degradation of both molecular forms to a 70-kDa band, representing the completely deglycosylated form of the receptor (Fig. 5B). This molecular mass form was identical to that of the deglycosylated form of the WT receptor after PNGase F treatment (Fig. 5B).
Cell surface delivery of N36Q-LGR7 is compromised
We performed the cell surface ELISA to test whether the glycosylation at the Asn36 site was critical for cell surface trafficking of the LGR7 receptor. The total level of expression of the mutant N36Q-LGR7 receptor was decreased to 72 ± 9.5% of that of the WT-LGR7 receptor (Fig. 5C). The cell surface expression of the N36Q-LGR7 receptor in HEK293 cells was decreased to 37 ± 7% compared with that of the WT-LGR7 receptor (P < 0.001) (Fig. 5C). This result showed that N-linked glycosylation at Asn36 is important in the delivery of the receptor to the plasma membrane. In the control experiments, the total expression level of the LGR7-HA receptor was 99 ± 8% of the WT receptor level and the cell surface expression of LGR7-HA was less than 1 ± 0.9% of that of the WT-LGR7 level (P < 0.001) (Fig. 5C). The relative amount of cell surface expression of N36Q-LGR7 compared with its total expression was 23%.
We also investigated the cell surface expression of the N36Q-LGR7 receptor by confocal microscopy. To determine protein localization, we performed immunochemical analysis under both nonpermeabilized and permeabilized conditions. In nonpermeabilized conditions, we observed plasma membrane localization of the WT-LGR7, whereas, in the HEK293 cells transfected with the glycosylation mutant N36Q-LGR7, there was much less staining (Fig. 5D, upper row). As anticipated, we were unable to detect expression of the LGR7-HA at the cell surface under nonpermeabilized conditions (Fig. 5D). In permeabilized cells, all three proteins were detected in the intracellular network (Fig. 5D, lower row).
We also studied the subcellular localization of the WT and glycosylated mutant receptor by using the colocalization of the Na+/K+ ATPase plasma membrane protein marker. The HEK293 cells were transiently transfected with the receptor variant labeled at the C terminus with GFP and were also stained for Na+/K+ ATPase plasma membrane protein. Confocal microscopy showed that the WT-LGR7-GFP colocalized with the plasma membrane protein marker (Fig. 5E). Only some of the glycosylation mutant N36Q-LGR7-GFP was colocalized with the plasma membrane marker and was mostly localized intracellularly (Fig. 5E), showing that cell surface delivery was influenced by the N-linked glycosylation at Asn36.
We also confirmed these results using the N36Q-LGR7 mutant receptor by cell surface protein biotinylation. As seen in Fig. 5F, in HEK293 cells transfected with the WT receptor, we detected the 95-kDa band (mature receptor form), whereas in cells transfected with the N36Q-LGR receptor, cell surface delivery was only just detectable. In the controls, the specificity of biotinylation was demonstrated by the inability to detect the intracellular protein (calnexin) by immunoblotting of the eluted samples (Fig. 5F, lower panel), whereas the antibody to calnexin strongly recognized the 90-kDa band in the flow-through samples (Fig. 5F, lower panel). These results support the view that glycosylation at Asn36 is critical for the proper delivery of the LGR7 receptor to the cell surface.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The three-dimensional structure of the LDL-A module of LDLR shows that these conserved cysteines are cross-linked with three disulfide bonds, and the acidic residues for calcium ion coordination lie at the C-terminal end of the domain (25). In addition, it has been shown that the C-terminal region, generated by the last two disulfide bonds is required for the formation of its native structure and calcium ion coordination (26). Recently, the isolated domain of the LDL-A module of LGR7 has been studied by NMR; this confirmed the formation of the disulfide bonds and calcium ion coordination and showed it to be like that of the LDL-A module of LDLR (29). Based on these data, we mutated the two cysteines in the LGR7 receptor, which are likely to be important in forming the disulfide bonds in the C terminus of the LDL-A module. Functional analysis of this C47A/C53A-LGR7 showed a complete loss of cAMP production after treatment with relaxins H1 and H2, confirming the importance of the formation of disulfide bonds in the C-terminal part of the module and the overall importance of the LDL-A module for signal transduction.
Another key structural element of the LDL-A module is the conserved calcium ion binding motif; residues in this motif are involved in calcium-dependent folding of the module. Mutations causing familial hypercholesterolemia frequently alter residues in this motif of the LDL-A module of the LDLR (37). To investigate the role of this conserved calcium ion binding motif, we generated the point mutant D58E-LGR7. This receptor showed a loss of activity in the functional assay with relaxins H1 or H2, indicating the importance of this conserved acidic residue in the LDL-A sequence and, thus, the importance of the LDL-A module in signaling. Furthermore, we investigated the function of the WT protein by complete deletion of the LDL-A module (LDL-LGR7). This showed complete loss of activity in the functional assay with relaxins H1 or H2, showing its importance in signal transduction. In summary, mutations in the conserved cysteine residues, mutations of the residue for the conserved calcium binding motif, or complete removal of the module from the LGR7 sequence all led to the loss of function of this receptor, showing the importance of the LDL-A module of LGR7 in receptor activation with relaxins H1 and H2. Similar studies using site-directed mutagenesis of the conserved residues of the LDL-A module of the LGR8 receptor [point mutations of the conserved cysteine residue (C71Y) and the conserved aspartic acid residue (D70Y)] abolished INSL3 stimulated cAMP production (32), supporting our results that the LDL-A module is important for LGR7 receptor function.
The LDL-A module of the LGR7 and LGR8 receptors share 72% sequence homology, allowing us to exchange the LDL-A module between them forming a chimeric receptor (LDL-A/LGR8-LGR7). This caused a complete loss of activity in the functional assay with relaxins H1 and H2 and INSL3. These data again show the importance of the LDL-A module of LGR7 for signal transduction and for providing specificity in hormone activation. Interestingly, in previous studies when the ectodomain of LGR8 was switched to the LGR7 receptor, relaxin H2 and INSL3 retained activation but with reduced efficacy (18), but when we exchanged only the LDL-A module of the ectodomain, activity was abolished, suggesting that the LDL-A module of LGR7 needs the LRR and/or TM domain of LGR7 for activation. Indeed, it has been shown that both the LRR region and the TM domain of LGR7 are important for receptor activation (19). In a recent study (20) it was demonstrated that, although an LDL-A module-deleted LGR7 receptor retained ligand binding, ligand-induced signaling was abolished, as shown in our studies. The binding was shown to be the same as that of the WT and implies that deletion of the LDL-A module or introduction of point mutations at conserved residues does not alter the conformation of the extracellular domain, and has no influence on ligand binding. However, in our study, we used different epitope tag and signal peptide constructs for expression of the LDL-A-deleted LGR7 receptor. Thus, the ligand binding characteristics could potentially differ in our study compared with that of Scott et al. (20). Therefore, further investigation is needed to compare the impact of different expression constructs with the function of the receptor. In our study, we also used point mutants of the LDL-A module and a chimeric LGR7 receptor. These modifications might alter the conformation of the receptor and disrupt ligand binding. In any event, the results show that the LDL-A module of LGR7 is obligate for receptor activation, although further investigation is needed to understand the intramolecular interactions involved.
We have analyzed whether the LDL-A sequence influences receptor expression and cell surface delivery by using our point mutants, the chimeric receptor, and the LDL-A module-deleted receptor variant. We found that the WT-LGR7 receptor was expressed as two molecular weight species in HEK293 cells. Endoglycosidase treatment and cell surface expression studies showed that the smaller molecular weight form was the immature/precursor receptor containing high mannose-type N-linked glycans and the larger form was the mature receptor delivered to the cell surface. Expression of two molecular weight forms (mature and immature/precursor) has also been demonstrated for other LGR family members: the LH receptor expresses two molecular weight forms in HEK293 cells and in natural tissues (38, 39, 40), the FSH receptor also expresses two molecular weight forms in HEK293 cells (41, 42). Previous studies have shown that a substantial portion of LGR receptors in a heterologous expression system and natural tissues exist in an immature form containing high mannose-type N-linked oligosaccharides, which are typical for proteins located in the endoplasmic reticulum (43, 44). Recent studies with the LH receptor have shown that a majority of the newly synthesized receptors are targeted for degradation, some fail to acquire the correct conformation and only a fraction are transported to the cell surface and bind hormone (39). Inefficient maturation occurs for other GPCRs (45, 46, 47) and may be a common phenomenon among this group of proteins.
The point mutants of the LDL-A module produced in our study and the chimeric receptor were all expressed as two molecular weight forms, like that of the WT receptor, indicating that the mutagenesis of the conserved residues of the LDL-A module or the exchange with another LDL-A module sequence had no influence on receptor expression. The cell surface delivery of these mutants was not compromised when their cell surface expression was characterized using both subcellular localization and a cell surface protein biotinylation assay. However, contrary to our results, the cell surface expression of similar point mutants of the LDL-A module of LGR8 (C71Y and D70Y) was shown to be drastically decreased (32), indicating that the maturation and delivery of the LGR8 receptor to the cell surface might differ from that of LGR7. It is also possible that different expression constructs/mutations used in two experiments may influence receptor surface expression. Interestingly, the LDL-A deleted LGR7 receptor variant only expressed one molecular weight form in transfected cells, which indicated that the LDL-A contains a sequence that actually interferes with expression of this receptor. Endoglycosydase treatment and cell surface protein biotinylation showed that this form represented the mature receptor delivered to the cell surface. Thus, the lack of the LDL-A sequence allowed the receptor to pass the endoplasmic reticulum quality control mechanism and be expressed at the cell surface. This cell surface delivery was confirmed by ELISA and the subcellular localization was confirmed by confocal microscopy. In addition, switching the LDL-A module sequence of LGR7 with the LDL-A sequence of the LGR8 (chimeric receptor LDL-A/LGR8-LGR7) resulted in an increase in cell surface delivery to 170 ± 10.5% compared with the WT receptor. These data suggest that the LDL-A sequence influences both receptor maturation and cell surface expression, and show that the N-terminal sequence of LGR7 contains important elements inhibiting its expression at the cell surface. Further work is needed to define the mechanism of this effect.
We investigated the putative N-linked glycosylation site in the LDL-A sequence of LGR7 by creating the point mutant N36Q-LGR7. In signal transduction with relaxins H1 and H2, the glycosylation mutant retained functional activity, but efficacy was decreased compared with the WT. This glycosylation mutant showed a reduction in molecular weight on Western blots compared with the molecular weight of the WT receptor, indicating that the Asn36 site is indeed glycosylated. Expression analysis and glycosidase treatment showed that the N36Q-LGR7 mutant receptor was expressed as two molecular weight forms (mature and precursor) in HEK293 cells, but that its level of expression was decreased compared with the WT receptor, showing that glycosylation at Asn36 is important for receptor expression. Cell surface delivery of this mutant receptor was compromised and showed only 37 ± 7% of the WT cell surface expression by quantitative ELISA. Compromised cell surface delivery was also confirmed by confocal microscopy and cell surface protein biotinylation assay. This showed the importance of glycosylation at Asn36 for delivery to the cell surface. Thus, although deglycosylation compromised cell surface delivery, the removal of the entire LDL-A module resulted in more rapid delivery to the cell surface. This can be explained by the latter resulting in production of novel protein that could pass the quality control mechanism and accelerate delivery to the cell surface. Contrary to our results, a recent publication showed (20) that the deglycosylation mutant of the LDL-A module of the LGR7 was expressed at the cell surface to the same degree as the WT receptor and responded to relaxin H2 with similar efficacy. The authors have also shown that the maximum cAMP response was lower than that of the WT receptor, similar to our data. A possible explanation for these differences might be again related to the use of different expression constructs with endogenous LGR7 signal sequence and HA tag in our experiments vs. signal sequence of bovine prolactin and FLAG epitope (20).
In summary, we have shown the importance of the LDL-A module of the LGR7 receptor for receptor expression and signaling. Mutagenesis of the conserved residues in the LDL-A module, deletion of the LDL-A sequence or use of a chimeric receptor all abolished the cAMP production caused by addition of relaxins. This shows the importance of the LDL-A module in signal transduction events. Future studies on how the LDL-A interacts and activates this receptor are clearly important to an understanding of a working model for the LGR7 receptor. Our work shows that the LDL-A module influences receptor expression and cell surface delivery, but further work is need to define the exact sequences implicated in these processes and the mechanism for these effects.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online December 7, 2006
Abbreviations: Endo H, Endoglycosidase H; GFP, green fluorescent protein; GPCR, G protein-coupled receptor; HA, hemagglutinin; HEK293, human embryonic kidney 293; IBMX, 3-isobutyl-1-methylxanthine; INSL, insulin-like peptide; LDL-A, low-density lipoprotein class A; LDLR, low-density lipoprotein receptor; LGR, leucine-rich repeat containing G-protein coupled receptors; LRR, leucine-rich repeat; PNGase F, peptide N-glycosidase F; RXFP, relaxin family peptide receptor; TM, transmembrane; WT, wild type.
Received August 15, 2006.
Accepted for publication November 27, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) or H2 (
) caused a dose-dependent production of cAMP in HEK293 cells transfected with WT-LGR7. The control (empty vector), relaxin H2 (
) had no effect. Middle panel, Relaxin H1 stimulated a dose-dependent production of cAMP in HEK293 cells transfected with the WT-LGR7 (
), D58E-LGR7 (
), and 
) had no effect with any dose tested. Lower panel, Relaxin H2 stimulated a dose-dependent production of cAMP in HEK293 cells transfected with the WT-LGR7 (
