This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Naranda, T. Articles by Olsson, L. Search for Related Content PubMed PubMed Citation Articles by Naranda, T. Articles by Olsson, L.

Activation of Erythropoietin Receptor through a Novel Extracellular Binding Site

Receptron, Inc. (T.N., R.I.K., J.L., K.W., A.B., L.O.), Mountain View, California 94043; Biovitrum, previously a Division of Pharmacia Corporation (D.H.), SE11276 Stockholm, Sweden; Department of Pharmaceutical Sciences (K.Y.C.F., M.W.D.), University of Colorado Health Sciences Center, Denver, Colorado 80262; Department of Chemistry (N.A.), University of Washington, Seattle, Washington 98195; and Stanford University (A.G.), Stanford, California 95305

Address all correspondence and requests for reprints to: Lennart Olsson, Receptron, Inc., 835 Maude Avenue, Mountain View, California 94043. E-mail: . lolsson{at}mcimail.com

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Activation of erythropoietin (EPO) receptor (EPOR) by a small peptide (ERP) was reported previously. ERP binds to a different receptor site than EPO, and binding of ERP does not change the dissociation constant and maximal binding for EPO binding to the receptor. The extracellular binding site for ERP is now characterized. The site is located in the membrane proximal, extracellular part of the receptor. ERP binds to a region on the EPOR that contains the same sequence as ERP. It is speculated that ERP binds to its identical sequence on EPOR, as ERP self-interacts. ERP is specific for EPOR and associates noncovalently with EPOR in a ratio 1:1. Peptide binding to the receptor results in receptor-mediated cellular proliferation, intracellular signaling, and erythroid colony-forming unit formation in bone marrow cells. The activity is comparable to that of EPO. Recognition of such receptor sites represents a new and important concept in receptor function.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
ERYTHROPOIETIN (EPO) CONTROLS the proliferation and differentiation of erythroid progenitor cells into red blood cells. It is essential for the maintenance of red blood cells. Deficits in EPO production result in anemia in humans and in animal models. EPO induces its biological effect after binding to a cell surface receptor [EPO receptor (EPOR)]. Biological responses associated with EPO include activation of many intracellular signaling molecules, i.e. transcription factors like signal transducer-and-activator-of-transcription (STAT) proteins, leading to cellular growth and differentiation. EPOR belongs to a family of homodimerizing receptors that form a 2:1 complex between receptor chains and hormone. Although it was originally thought that dimerization of EPOR was all that was required to trigger the biological responses associated with EPO (1, 2, 3, 4, 5, 6, 7, 8, 9), more recent reports show that EPOR exists as a preformed dimer, and EPO binding results in closer association of the two chains and intracellular signaling (10, 11). In addition, a constitutively active (hormone-independent) EPOR was first isolated with an arginine-to-cysteine mutation at position 129 (1). Similar mutant proteins were than created by introducing a cysteine residue in parts of the putative EPOR dimerization interphase (12, 13).

It was previously reported that a peptide, designated EPOR peptide (ERP), with sequence identical to a site on the EPOR, activates receptor signaling in the absence of EPO and can also act in synergy with EPO (14).

We now report on characteristics of the binding site for ERP on EPOR and of the complex formed by EPOR-ERP. The ERP sequence (amino acids 194–216 of human EPOR) is contained in the same extracellular domain of EPOR as that known, from x-ray crystallographic studies, to participate in dimerization of activated receptor chains (10, 11, 15). Further, bindings studies show that ERP binds to a membrane proximal region on EPOR, between amino acids 174–223. The binding sites for EPO and ERP on the EPOR are different, with the ERP site being more membrane proximal than EPO.

EPOR activation is now described to be a result of ERP binding in the region where two receptor chains normally interact in a scissor-like model in response to hormone binding. The stable complex between ERP and EPOR has a 1:1 stoichometry. ERP binding to EPOR results in receptor- mediated signal transduction, cell proliferation, and CFU-e colony formation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Signal transduction
Signal transduction was evaluated in TF-1 cell line (erythroleukemia, human, ATCC CRL-2003). Cells were grown and processed as for the binding experiments, resuspended to the original density in medium without serum or GM-CSF; 2 ml (1 x 10⁷ cells) were used for each stimulation experiment. Cell suspension was treated with ERP or EPO for 35 min at 37 C in 5% CO₂, then processed as described (14). For each experimental condition, lysates were split between two immunoprecipitations. After washes, one set of immunoprecipitations (for nondenaturing/nonreducing gel conditions) was treated with Tris-Glycine native sample buffer, and proteins were analyzed on NuPAGE Tris-Acetate 7% gels (Novex, San Diego, CA) according to manufacturer’s recommendations. For the other set of immunoprecipitations (reducing gel conditions) SDS sample buffer (50 µl) was added, and the proteins were separated on 8% polyacrylamide minigel. Proteins were transferred in Hoefer TE 22 Mighty small transfer unit (Amersham Pharmacia Biotech, Chicago, IL), for 1 h at 300 mA, in a buffer of 10 mM (3-[cyclohexylamino]-propanesulfonic acid), pH 11, and 10% methanol. The polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA) were treated for 1 h in blocking buffer (10 mM Tris, pH 7.4; 0.15 M NaCl; 0.075% Tween-20; 0.02% NaN₃) with 0.5% dry skim milk powder, and incubated with the appropriate primary antibody (1:1000 dilution) overnight at 4 C. The membrane was washed, incubated with appropriate alkaline phosphatase-conjugated secondary antibody (1:2000 dilution) at room temperature for 2 h. Membranes were washed, and color was developed with nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate (Sigma, St. Louis, MO). The following antibodies were used: polyclonal rabbit antibody against ERP was from Zymed Laboratories, Inc. (South San Francisco, CA); antiphosphotyrosine monoclonal antibody PY-99 (PY) and rabbit polyclonal antibodies against EPO, STAT5, and EPOR were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Western blots were scanned, and intensity of the protein bands was quantitated.

Cell proliferation
TF-1 cells were grown as described previously. Before proliferation experiments, the cells were starved overnight in media with 5% FCS, washed, and resuspended to a density of 5 x 10⁵ cells/ml in a media with 5% serum. Fifty thousand cells were used per well; triplicates were performed per experimental condition. Peptide or EPO was added in desired final concentrations. Plates were incubated for 68 h at 37 C, 5% CO₂. Proliferation solution (tetrazolium salt solution, CellTiter 96, Promega Corp., Madison, WI) was added, and cell proliferation was evaluated by reading at A 490 nm.

CFU-e formation
Bone marrow cells were isolated from mouse femurs. Cells were washed twice with PBS and resuspended to a density of 6.5 x 10⁶ cells/ml in media; 300 µl of cells were mixed with 3.0 ml of methylcellulose media (StemCell Technologies, Vancouver, British Columbia, Canada), to which the desired concentration of ERP (1–30 µM) or the EPO (10 U/ml) was added. The mixture was poured onto 35-mm dishes; all the conditions are performed in duplicates. The incubation was for 8 d at 37 C, 5% CO₂. The colonies were scored on d 7, and cells were stained with May-Grunwald-Giemsa (StemCell Technologies) solution for cell lineage determination.

ERP binding to TF-1 cells
Cells were maintained in RPMI-1640 supplemented with 5 mM glutamine, 100 U/ml penicillin, 10 mg/ml streptomycin, 10% FBS (HyClone Laboratories, Inc., Logan, UT), and 1 ng/ml GM-CSF. Cells were grown in RPMI-1640 (supplemented as above) to a density of approximately 1 x 10⁶/ml, centrifuged (200 x g, 5 min), and resuspended in medium with 5% serum and no GM-CSF. Cells were starved 12–16 h at 37 C in 5% CO₂, centrifuged, and resuspended to 2 x 10⁷/ml in medium without serum and GM-CSF. Cell suspension (1 ml) was incubated with 1 x 10⁷ cpm [¹²⁵I]ERP in the presence or absence of 10 µM ERP, or alternatively with 1 x 10⁶ cpm [¹²⁵I]EPO in the presence or absence of 10 U/ml EPO. Cell suspension was incubated with end-over-end rotation overnight at 4 C. Cells were washed three times with PBS and lysed in 200 µl of a 2x lysis buffer for 30 min at 4 C. SDS sample buffer (without reducing agent) was added to cleared cell lysates, and the proteins were separated on 7% SDS-PAGE. Gels were exposed to film.

ERP binding to purified receptor
Extracellular binding protein (EBP) (recombinant human soluble receptor (sR), EPO-sR) was purchased from R&D Systems (Minneapolis, MN). Binding of ERP and of EPO to EBP was performed in PBS buffer overnight at 4 C. Unless specified otherwise, 160 nM EBP was incubated with 300 pM [¹²⁵I]EPO (specific activity 300 Ci/mmol; Amersham Pharmacia Biotech) in the absence or presence of 10 U/ml EPO. Alternatively, EBP was incubated with 200 pM [¹²⁵I]ERP in the absence or presence of 1 µM ERP. ERP was synthesized, characterized, and ¹²⁵I-labeled (specific activity 10–15 Ci/mmol) as previously described (14). After overnight incubation, SDS sample buffer without reducing agent was added to the mixture, and samples were separated on SDS-PAGE 8%, 10%, or 12%, depending on which complexes were to be separated. Gels were exposed on film.

Expression of EBP in Escherichia coli
The cDNA encoding human EPOR gene was constructed as previously described (14). The extracellular domain of the human receptor was amplified by PCR from a full-length human EPOR, using the following 5' and 3' primers, respectively: 5'-GGAATTCCATCGAAGGTCGTGCGCGCCCCCGCCTAAC and 5'-GGGGCGGCCGCCTAGGGGTCCAGGTCGCT. The amplified DNA was digested with EcoRI and NotI and ligated into the corresponding sites in pGEX-4T-3 vector. The extracellular portion of the EPOR recombinant protein was expressed in E. coli as a glutathione S-transferase (GST) fusion protein, and purified on GSH-Sepharose beads. This construct allows induction of a GST-EPOR extracellular domain (EBP) fusion protein by IPTG. Factor Xa cleavage site was engineered between GST and EPOR extracellular domain. Fusion protein was first eluted from GSH beads, than cleaved with Factor Xa overnight at 4 C in buffer (10 mM Tris, pH 7.6, 10 mM CaCl₂, 100 mM NaCl). After cleavage of GST protein with Factor Xa, the remaining EBP is a mature 226-residue protein (amino acids 25–250, SwissProt numbering). Purification of EBP was done by affinity chromatography on GSH-beads to remove GST protein, and streptavidin beads to remove biotinylated Factor Xa. The binding of [¹²⁵I]EPO and [¹²⁵I]ERP to the purified protein expressed in E. coli was comparable with the binding obtained with purchased fully glycosylated EBP (R&D Systems, Minneapolis, MN).

Calorimetric analysis
The isothermal titration calorimetry (ITC) experiments were performed in a Thermal Activity Monitor (Thermometric AB, Järfälla, Sweden), equipped with nanowatt-amplifiers and 1-ml titration vessels. The experiments were performed at 15 C, 20 C, 30 C, and 35 C. The calorimeters were calibrated electrically using both static and dynamic calibrations. EBP was dialyzed against 50 mM sodium phosphate and 100 mM sodium chloride at pH 7.4. ERP was dissolved in water and diluted with the dialyzing buffer to 150–200 µM. The EBP was diluted with water to have the same buffer condition for titrant and titrand. Concentration determination of EBP was done by absorbance at 280 nm, using extinction coefficient = 41400 M^-1cm.

An EBPsolution (0.9 ml; 10 µM) was placed in the titration cell, which was stirred at a rate of 60 rpm. At each titration, 25–35 additions of ERP (150–200 µM) where made at intervals of 4–8 min between each addition. Experimental control, data collection, and data analysis were done using the Digitam Software (Scitech Software, Järfälla, Sweden).

Matrix-assisted laser desorption ionization (MALDI)-TOFMS (time-of-flight mass spectrometer) analysis
The experiments were performed after incubation of ERP with EBP, overnight at room temperature. In these experiments, all buffers were free of albumin and other proteins that may interfere with the assay. Recombinant EPO-sR (EBP) was incubated overnight (Tris buffer, pH 8.0), at room temperature, with varying amounts of ERP in the ratios of 1:0.5; 1:1; 1:3; 1:6; and 1:12. All were in aqueous solution. EBP/ERP, incubated in the ratios 1:6 and 1:12 (100 pmol EBP with 600 pmol ERP or with 1200 ERP), provided evidence of formation of EBP-ERP complexes. EBP without addition of ERP was used as control. Mass analyses of EBP only, of ERP only, and of EBP-ERP complex were performed on a Voyager DE-STR MALDI-TOFMS analyzer (Perspective Biosystems, Framingham, MA). Mass determinations were made in positive ion linear mode with delayed extraction. Sinapinic acid (Sigma) was used as matrix (5 mg/ml; 80% CH₃CN, 0.1% trifluoroacetic acid). The average mass of the ERP was determined by internal 2-point calibration to substance P (Sigma) and bovine insulin (Sigma). The average molecular mass of EBP, both pre- and post incubation, was determined by external 2-point calibration to lysozyme and BSA (Fisher Scientific, Houston, TX).

CNBr (cyanogen bromide) digest
Approximately 50 µg E. coli-purified EBP were incubated with 10 µM ERP, overnight at 4 C, in PBS. Protein was digested in 75% formic acid with 0.2M CNBr, for 24 h at room temperature, in the dark. Digested protein solution was dried and resuspended in 50 µl PBS. SDS sample buffer (without reducing agent) was added, and the proteins were separated on 7–15% gradient or 10% NuPAGE BT/MES gels (Novex). Duplicate halves of the gel were prepared. After transfer, one half of the Immobilon-P membrane was blocked and then incubated with anti-ERP antibody (1:1000 dilution), and the other half of the membrane was stained with Coomassie Blue. The Western blot served as a guide in identifying the Coomassie Blue band containing the EBP fragment with bound peptide. That band was cut out of the stained membrane and sequenced, at Molecular Structure Laboratory at the University of California, Davis, on the ABI 477 (Foster City, CA) sequencing machine.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Signal transduction
Signal transduction was studied in a human erythroleukemia cell line (TF-1) as described in Materials and Methods. To examine activation, i.e. phosphorylation, of substrates by EPOR, the cells were incubated with EPO and/or ERP, and lysed, and tyrosine phosphorylated molecules were immunoprecipitated. The results of Western blot analysis of the immunoprecipitated material, using antibodies to the ERP, EPO, EPOR, and STAT5 protein, respectively, are shown in Fig. 1, A–D.

View larger version (49K): [in this window] [in a new window]   Figure 1. Signaling in TF-1 cells activated by ERP and/or EPO. In all four panels, immunoprecipitation of cell lysate was by an antiphosphotyrosine antibody followed by Western blotting (WB), with antibodies against ERP (A), EPO (B), EPOR (C), or STAT5 (D). Arrows at right indicate positions of specified complexes. Stimulation, as indicated across the top, was by low and high concentrations of ERP and EPO, alone or in combination. Gels in A and B were under nonreducing and nondenaturing conditions, those in C and D under normal reducing and denaturing conditions.

In Fig. 1B, the gel was again run under nonreducing, nondenaturing conditions. The Western blotting antibody (EPO) binds to EPO. The bands in lanes 3–4 correspond, as expected, to one EPO molecule binding to a dimerized receptor. The band at approximately 120 kDa, representing a complex of EPO and monomer chain of EPOR, is much weaker; this may be attributable to low stability of the 1:1 complex of EPO and EPOR, to lower affinity of the EPO antibody, or to a dominant amount of dimer complexes at low EPO concentrations. Figure 1B, lanes 5–6, also illustrates the well-known reduction of signaling (measured by phosphorylation) by a high concentration of EPO (bell-shaped dose-response curve). In this case, only a weak band is seen at the molecular mass of a complex of EPO+2EPOR, and no band is seen for a complex of EPO+EPOR.

Analysis of phosphorylation of the EPOR itself is seen in Fig. 1C, using an antibody specific for EPOR. Thus, because only phosphorylated molecules were immunoprecipitated, the Western blot shows phosphorylated EPORs. The gel for Western blot was in contrast to panels A and B, run under reducing and denaturing conditions. It is seen that stimulation of TF-1 cells with either ERP or EPO results in phosphorylation of the EPOR.

Finally, phosphorylation of the intracellular substrate STAT5 was studied in panel D, using an antibody to STAT5 for the Western blot. Again, both ERP and EPO result in phosphorylation of STAT5. Lane 5 also shows, as expected, a reduction in phosphorylation of STAT5 at high EPO concentrations, in agreement with the bell-shaped dose-response curve. It is interesting, and in line with previous results (14), that ERP eliminates any sign of a bell-shaped curve effect of EPO (lane 6), i.e. high concentrations of EPO do not result in decreased substrate phosphorylation when ERP is present. This abolition of the bell-shaped form of the dose-response curve of EPO-induced EPOR activation suggests that ERP binding results in stabilization of the active conformation of the EPOR chain, even when EPO is bound to a single receptor chain.

Cell proliferation and colony formation
The ability of ERP to stimulate cell proliferation was evaluated. Figure 2A shows that ERP promotes TF-1 cell proliferation in a dose-responsive manner. ERP promotes cell proliferation but with a significant right shift in a dose-response curve (about 10-fold), compared with its effects on substrate phosphorylation. The difference in ERP’s effectiveness in these two systems may be caused by the fact that almost all of the peptide is degraded within 24 h (data not shown). In the system used to measure cell signaling via substrate phosphorylation, the cells are incubated with peptide for only 20–60 min. Cell proliferation, on the other hand, requires that the cells be incubated with peptide for 48–72 h. Peptide degradation almost certainly plays a role in this assay.

View larger version (24K): [in this window] [in a new window]   Figure 2. Effect of ERP in cell proliferation and colony formation assay. A, Cell proliferation in TF-1 cells. Experiment was performed as described in Materials and Methods. Proliferation was measured 68 h after the addition of ERP (closed squares) or EPO (open circles). Data are mean ± SEM of three independent experiments, each done in triplicate. Different scales for EPO and ERP are indicated. MTS, 3-(4,5 Dimethylthiazol-2-yl)-5-(-3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium. B, Colony formation in murine bone marrow. Cells were prepared as described. Fifty thousand cells were seeded in the presence of EPO or ERP and colonies were scored at d 7. Data are mean ± SEM of three independent experiments, each done in duplicate. Colonies were stained and erythroid lineage formation was confirmed.

Colonies were scored 7 d after the seeding and were stained with May-Grunwald-Giemsa, and formation of erythroid colonies was confirmed morphologically. Maximal numbers of colonies were approximately the same for EPO and ERP at their maximal effect.

ERP binding to EPOR
The interaction between ERP and EPOR was analyzed by several methods. First, binding of [¹²⁵I]EPO and [¹²⁵I]ERP to intact receptor was studied in whole cells. Upon binding of the labeled ligand to whole cells, the latter were lysed and analyzed by PAGE (Fig. 3A) under nonreducing conditions. Radiolabeled EPO binds to a molecule of apparent molecular mass corresponding to that of dimerized EPOR, whereas radiolabeled ERP binds to a molecule of size corresponding to that of an EPOR monomer chain. Competition by nonradioactive EPO and ERP (lanes 2 and 4) shows that the binding of both ligands is specific. Surprisingly, SDS does not affect the complex. Bound ¹²⁵I-labeled ERP can be competed out with unlabeled peptide (see further below), excluding covalent binding through cysteine bridges. On the other hand, the strong binding of ERP concurs with the extremely slow off-rate of the peptide upon binding to the receptor. It should be noted that the PAGE-gel system used to separate proteins in lysates of whole cells was different (12%) from the PAGE-gel system for immunoprecipitated proteins in Fig. 1 (8% PAGE). Such differences may have caused apparent differences in mobility of otherwise identical protein complexes.

View larger version (42K): [in this window] [in a new window]   Figure 3. Binding of ERP to EPO-sR or EBP. A, Specific binding of [125I]EPO and [125I]ERP to whole cells. Total cell lysates were separated on 12% SDS-PAGE under nonreducing conditions and the gel analyzed by autoradiography. Lanes 1 and 2 show binding of [125I]EPO, and lanes 3 and 4 of [125I]ERP. Competition with unlabeled EPO or ERP is indicated on the top of the figure (lanes 2 and 4). Arrows at right indicate positions of formed complexes: labeled EPO (single molecule) bound to receptor dimer, and of labeled ERP bound to receptor monomer. Bands at bottom represent free radioligands. Standard molecular mass markers (kDa) based on approximate gel mobilities are indicated at right. B, Binding of radiolabeled [125I]EPO and [125I]ERP to EBP. The binding protein was purchased from R&D Systems. The complexes were separated on 12%SDS-PAGE under nonreducing conditions and the gel analyzed by autoradiography. Lane 1 shows expected positions of all possible complexes, based on gel mobility. Lanes 2–4 and 5–7 show specific binding of [125I]EPO and of [125I]ERP, respectively, to EBP. Competition with unlabeled EPO or ERP is indicated on the top of the figure. Arrows to the left of the figure indicate the position of specific complexes 2EBP+EPO and EBP+ERP. Heavy dark bands are free radioligands and are indicated with arrows to the right. Molecular markers are also indicated to the right. C, Competition curve for binding of [125I]ERP to EPO-sR and to EBP. The autoradiograms were scanned by densitometry for quantitative purposes. Binding of [125I]ERP only was given a value of 100%. Apparent Kd is about 100 nM. Data are means ± SEM of nine independent experiments with EPO-sR and three with EBP. Inset shows that three other peptides did not compete. GRP, GH receptor peptide; TRP, TPO receptor peptide; 4RP, IL-4 receptor peptide. D, Estimation of molar ratio between ERP/EBP upon binding of ERP to EBP. Exothermic enthalpy, Q, is defined as positive. The plot represents processed experimental data from binding experiments of ERP to EBP at 35 C in PBS at pH 7.4, where corrections for enthalpy of dilution of ERP have been made. ERP/EBP is the equimolar ratio of ERP and Q the individual nonnormalized enthalpy obtained at each injection step. In the experiment, the calorimetric vessel was charged with 900 µl 10.6 µM EBP and 21 consecutive additions of 7.5 µl 150 µM ERP was made. The plot shows the processed ITC data at 35 C in PBS at pH 7.4, where corrections for enthalpy of dilution of ERP have been made. The points represent means of experimental measurement values, and the dotted line corresponds to the best-fit curve obtained by nonlinear regression.

The competition curve presented in Fig. 3C was performed by incubating 160 nM EBP with ¹²⁵I-ERP and cold ERP at the concentrations indicated. The competition binding curve for [¹²⁵I]ERP to EBP (Fig. 3C) indicates an apparent dissociation constant (K_d) of approximately100 nM. The inset in Fig. 3C shows that similar peptides (used at 10 µM concentrations), derived from GH receptor (GRP), thrombopoietin receptor peptide (TRP), and IL-4 receptor (IL4P), did not compete with [¹²⁵I]ERP for binding to EBP, although each of these peptides competes with its own cognate peptide when binding to its respective native receptor (data not shown). Therefore, the binding of ERP to the EPOR binding protein is selective. Cross-linking experiments between ERP and either EPOR (whole cells) or EBP also demonstrated the presence of one dominant complex corresponding to a 1:1 ratio between ERP and the receptor (data not shown). Homobifunctional sulhydryl-reactive cross-linking reagent, 1,4-di-[3'-(2'-pyridyldithio)-propionamido]butane (DPDPB) (3), was used.

Binding of ERP to EBP was also studied by ITC. At 35 C, the calorimetric data could be rationalized to a 1:1 binding model (Fig. 3D). The apparent K_d is 240 ± 100 nM, calculated from nonlinear regression; and the enthalpy of binding, H°, = 31.6 ± 1.4 kJ mol^-1. The entropy of binding at this condition is positive with the value S° = 229 ± 6 J K^-1 mol^-1. Both entropy and enthalpy data suggest that hydrophobic interactions are dominating features of the binding. Heat capacity data (Cp) for the binding could not be obtained, because ITC experiments at lower temperatures (5 C and 20 C) were hampered by dominating nonideality effects of ERP in solution. These effects could not be rationalized into any simple model. Consequently, a calorimetric analysis of ERP binding to EBP was not feasible at lower temperatures. Interestingly, the nonideality of ERP was not present at 35 C, permitting a calorimetric study at that temperature. In agreement with this, analysis of ERP by NMR in PBS at pH 7.4 revealed a distinct line broadening of the ¹H-NMR spectra at lower temperatures (data not shown). This line broadening, attributable to slow tumbling of the peptide in solution, is consistent with formation of peptide aggregates at low temperature.

Finally, as seen in Table 1, MALDI-TOFMS analysis demonstrated that ERP binds to a monomer form of EBP (a complex of 1:1 ratio between peptide and EBP), and that this complex persists in the gas phase. No evidence was observed for formation of an EBP-ERP complex when the same samples were analyzed by electrospray liquid chromatography-mass spectroscopy. This result is consistent with a noncovalent association between ERP and EBP, forming a complex that dissociates during chromatography. Similar noncovalent interactions of this type have recently been established by MALDI (17).

View larger version (27K): [in this window] [in a new window]   Figure 4. Binding site of ERP on EBP. A, EBP, the extracellular domain of EPOR, was cloned, expressed, purified, and shown to bind radiolabeled EPO and ERP (Fig. 3B). ERP was bound, and the receptor-peptide complex was digested with CNBr, and then separated by gradient SDS-PAGE under nonreducing conditions. After transfer, one half of the membrane was incubated with -ERP antibody membrane, and the Western blots shown. The other half of the membrane was stained with Coomassie Blue. A fragment in lane 5 corresponding to an 8-kDa band was excised and subjected to amino acid sequencing. Presence of ERP, EBP, and CNBr is indicated on the top of the figure for each lane. The arrows to the right indicate nondigested EBP/ERP fragment and the fragment subjected to sequencing. Bottom band represents excess of ERP. B, Cartoon at top represents EBP, showing (in black) the domain of identity to ERP (residues 194–216). Residue 25 (SwissProt numbering) follows the signal peptide. Arrows at residues 174 and 223 indicate the two methionine residues that flank ERP. The sequence of the 5-kDa band that binds ERP, and which itself contains the sequence of ERP, is shown. Two independent experiments were carried out under nonreducing conditions, yielding identical data.

This shows unambiguously that, under nonreducing conditions, ERP binds to EBP in the receptor domain exactly where the sequence of homology is detected. However, because the Edman procedure only detects the sequences of the peptide and the 5-kDa fragment from the N-terminal end, and residue by residue, the alignment of ERP with the 5-kDa fragment cannot be established. ERP seems to bind only to the 5-kDa fragment in a 1:1 ratio. The residues of crucial importance for this binding are not yet established but are currently being studied through structural and mutagenesis experiments.

Experiments with radiolabeled ERP (data not shown) confirmed that the size of the CNBr fragment to which [¹²⁵I]ERP binds corresponds to the same fragment identified by Western analysis using antibody to ERP. The dependence of ERP binding on nonreducing conditions indicates the possibility of a disulfide bond between the cysteine residues of ERP and EBP (see sequences). Although such covalent association could not be excluded as a secondary interaction, our data indicate that it is not necessary for peptide binding, because radiolabeled or biotinylated ERP can be displaced with unlabeled peptide in pulse and chase experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
EPOR is a typical homodimerizing receptor: one EPO molecule is bound to two receptor chains. Receptors for GH and colony-stimulating factor belong to the same family. The EPOR may also exist as a preformed inactive dimer (10, 11). In that case, EPO binding changes the conformation of the inactive dimer into a conformation that promotes intracellular signaling (8, 10, 11). The interaction between the two receptor chains takes place between similar regions on the two receptor chains, and the dimerization is assumed to stabilize a receptor conformation that facilitates intracellular signaling (2, 11).

Here, we describe the activation of EPOR by ERP, a small peptide that binds to the receptor at a site that is different from the hormone binding site: 1) this bioactive peptide does not affect the binding (either in terms of number of binding sites or of affinity) of the natural ligand (EPO) to the receptor; and 2) ERP and EPO act in synergy in in vitro signaling (14), in vivo experiments, and on CFU-e formation from bone marrow. In contrast, EPO mimetic peptides with binding sites identical to the ligand binding site have been reported (reviewed in Ref. 18).

Sequence analysis of the EPOR/ERP complex suggests that the peptide is binding to an extracellular portion of the receptor close to the membrane. The binding site is between amino acid residues 174–223, and the transmembrane domain starts at amino acid residue 250. The 8-kDa band (lane 5 in Fig. 4A), representing a complex of the 3-kDa ERP with the 5-kDa EBP fragment, indeed contained the ERP sequence, as shown by the Edman degradation analysis. It is of considerable interest that the 5-kDa EBP fragment is localized precisely in the receptor region (aa 174–223) that also contains the sequence of the ERP (aa 194–216). Therefore, it seems most likely that the binding site for ERP to the receptor is in this region. Further, the ERP molecules self-interact, and we therefore speculate that ERP binds to the 5-kDa fragment by interacting with its identical sequence in this fragment. This is supported by cross-linking experiments, as well as by data from a previous report on other similar peptides (19). The 5-kDa receptor fragment comprises a region that is part of the receptor dimerization domain. It is suggested that the ERP binding site on EPOR also has a functional role in the activation of EPOR in the presence of normal ligand, because this site is localized in the dimerization region of the receptor.

The binding studies indicate a slow on-rate and a particularly slow off-rate for the ERP/EBP interaction. In fact, ERP remains bound to the receptor, even upon lysis of the cells and PAGE analysis. Radiolabeled ERP can be displaced from the receptor with unlabeled ERP in a bind-and-chase experiment (data not shown), indicating that the binding is noncovalent. The affinity of ERP to the receptor seems to be in the nanomolar range but is difficult to determine precisely because of: 1) self-aggregation, whereby ERP is not available for binding to the receptor; 2) different conformations of ERP, where some may be active and others inactive; and 3) uncertainty as to whether equilibrium is achieved. The calorimetric measurements indicate a K_d of approximately 200 nM. This value for binding affinity is in line with previous data from binding studies of labeled ERP to purified EPOR binding protein or to cells (14). The estimated EC₅₀ value for biological activity, as measured by activation of signaling, is approximately 100 nM.

ERP has a self-interacting property. The ITC and ¹H-NMR data suggest that the degree of aggregation is strongly temperature-dependent and that the aggregation process has cooperative features. The cooperative properties are manifested by the dominating contributions of heat of dilution of the peptide in the binding ITC experiments at lower temperatures, but there are no (or nondetectable) contributions at higher temperatures. We therefore speculate that the same features resulting in peptide-peptide interaction also are important for binding of ERP to EPOR binding site that contains the same sequence as ERP.

The peptide (ERP) can, by itself, initiate the signal transduction cascade as generally understood for EPO, i.e. phosphorylation of EPOR, of STAT-5 or JAK-2 (14), stimulation of cell proliferation, and CFU-e formation in bone marrow. It is unclear, however, whether the ERP induces dimerization of the receptor. In fact, the following lines of evidence suggest that the traditional dimerization may not occur: 1) Binding of ¹²⁵I-labeled peptide suggests a stable 1:1 ratio. Minor presence of 2:1 or 2:2 complexes shows the possibility of 2:1 complex formation. The calorimetric (and, to some extent, also MALDI-TOFSM) analyses indicate a ratio of 1:1 between ERP and the receptor, i.e. one ERP binds to one EPOR chain. 2) In vitro signaling experiments suggest that the activated complex contains one receptor chain (immunoprecipitation and Western blot analysis under the nondenaturing conditions). 3) ERP eliminates the bell-shaped curve for activation of the receptor by EPO (14). The bell-shaped curve is assumed to be caused by a lack of dimerization, because each receptor chain is occupied by a ligand, thereby preventing formation of an active dimer. However, if each chain is in a stable active conformation because of the binding of ERP, the bell-shaped curve is eliminated, even at high concentrations of EPO.

This seems to somewhat contradict the notion that the EPOR must be homodimeric to signal. However, it may not be a significant contradiction. One may envision that ERP binds to one or both chains in the dimerization region. This could bring one or both chains into an active conformation in the absence of EPO, and signaling may take place because the two chains are in proximity to each other. Depending on the concentration of ERP, and assuming that the EPOR preexists as an inactive dimer, ERP may activate just one chain in the inactive dimer. In that case, it is possible that the active conformation of one chain will drive the other chain toward an active conformation, and thereby result in formation of a homodimer receptor conformation, associated with one peptide. The synergy between EPO and ERP could, in that case, be explained by the assumption that EPO binding to such a complex results in a highly stable dimer in its active conformation.

The method to identify receptor agonist or antagonist peptide sequences such as ERP was first reported several years ago (14, 20). To our surprise, the major histocompatibility complex class I-derived peptides (19, 21, 22) were discovered to have significant homology to sequences in the extracellular region of numerous receptors; such regions are almost exclusively close to the cell membrane. For each receptor, a different sequence has been identified (14, 20).

We do not yet understand the evolutionary principles underlying the observations that the 1-domain of major histocompatibility complex class I can be used to identify such receptor sequences, but we do note the generality of our findings. Evidently, the method of homology search is widely applicable; we have applied it successfully to 33 other receptors, including some in the 7-helix family of receptors (unpublished observations). The identification of such extracellular receptor sites with a significant role in regulation of receptor function (and different from the binding site of the natural hormone) are of essential importance for our understanding of molecular and conformational processes during receptor activation and deactivation.

The findings reported here also have important practical implications in drug discovery and in the use of such drugs in novel, receptor-directed pharmacotherapies.

	Acknowledgments

 
We thank the Protein Structure Laboratory, at the University of California, Davis, for protein and peptide sequence analyses; J. Hohmann and Dr. G. Hendrick, at Receptron, Inc., for contributions to the presented work; and R. Matthew Fesinmeyer, at the University of Washington, for suggestions concerning the dimerization interfaces of homodimeric receptors and spectroscopic studies of ERP. Professor David E. Pettijohn (University of Colorado Health Sciences Center, Denver, CO) and Professor Rodney Cotterill (Danish Technical University, Copenhagen, Denmark) are thanked for carefully reading and commenting on the manuscript. The encouragement and suggestions by the late Dr. S. Ohno (Beckman Coulter, Inc. Institute of City of Hope, Duarte, CA) are much appreciated.

	Footnotes

 
Abbreviations: CFU-e, Erythroid colony-forming unit; CNBr, cyanogen bromide; EBP, extracellular binding protein; EC₅₀, 50% maximal effect; EPO, erythropoietin; EPO-sR, EPO soluble receptor; ERP, EPOR peptide; GST, glutathione S-transferase; ITC, isothermal titration calorimetry; MALDI, matrix-assisted laser desorption ionization; STAT, signal transducer and activator of transcription; TOMFS, time-of-flight mass spectrometer; TRP, thrombopoietin receptor peptide.

Received November 6, 2001.

Accepted for publication February 21, 2002.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Watowich SS, Yoshimura A, Longmore GD, Hilton DJ, Yoshimura Y, Lodish HF 1992 Homodimerization and constitutive activation of the erythropoietin receptor. Proc Natl Acad Sci USA 89:2140–2144[Abstract/Free Full Text]
2. Youssoufian H, Longmore G, Neumann D, Yoshimura A, Lodish HF 1993 Structure, function, and activation of the erythropoietin receptor. Blood 81:2223–2236[Free Full Text]
3. Livnah O, Stura EA, Johnson DL, Middleton SA, Mulcahy LS, Wrighton NC, Dower WJ, Jolliffe LK, Wilson IA 1996 Functional mimicry of a protein hormone by a peptide agonist: the EPO receptor complex at 2.8 A. Science 273:464–471[Abstract]
4. Ihle JN 1995 Cytokine receptor signaling. Nature 377:591–594[CrossRef][Medline]
5. Heldin CH 1995 Dimerization of cell surface receptors in signal transduction. Cell 80:213–223[CrossRef][Medline]
6. Taniguchi T 1995 Cytokine signaling through nonreceptor protein tyrosine kinases. Science 268:251–255[Abstract/Free Full Text]
7. Tarr K, Watowich SS, Longmore GD 1997 Cell surface organization of the erythropoietin receptor complex differs depending on its mode of activation. J Biol Chem 272:9099–9107[Abstract/Free Full Text]
8. Weiss A, Schlessinger J 1998 Switching signals on or off by receptor dimerization. Cell 94:277–280[CrossRef][Medline]
9. Livnah O, Johnson DL, Stura EA, Farrell FX, Barbone FP, You Y, Liu KD, Goldsmith MA, He W, Krause CD, Pestka S, Jolliffe LK, Wilson IA 1998 An antagonist peptide-EPO receptor complex suggests that receptor dimerization is not sufficient for activation. Nat Struct Biol 5:993–1004[CrossRef][Medline]
10. Livnah O, Stura EA, Middleton SA, Johnson DL, Jolliffe LK, Wilson IA 1999 Crystallographic evidence for preformed dimers of erythropoietin receptor before ligand activation. Science 283:987–990[Abstract/Free Full Text]
11. Remy I, Wilson IA, Michnick SW 1999 Erythropoietin receptor activation by a ligand-induced conformation change. Science 283:990–993[Abstract/Free Full Text]
12. Watowich SS, Hilton DJ, Lodish H 1994 Activation and inhibition of erythropoietin receptor function: role of receptor dimerization. Mol Cell Biol 14:3535–3549[Abstract/Free Full Text]
13. Longmore G, Pharr P, Lodish H 1994 A constitutively activated erythropoietin receptor stimulates proliferation and contributes to transformation of multipotent, committed nonerythroid and erythroid progenitor cells. Mol Cell Biol 14:2266–2277[Abstract/Free Full Text]
14. Naranda T, Wong K, Kaufman RI, Goldstein A, Olsson L 1999 Activation of erythropoietin receptor in the absence of hormone by a peptide that binds to a domain different from the hormone binding site. Proc Natl Acad Sci USA 96:7569–7574[Abstract/Free Full Text]
15. Caravella JA, Lyne PD, Richards WG 1996 A partial model of the erythropoietin receptor complex. Proteins 24:394–401[CrossRef][Medline]
16. Mayeux P, Lacombe C, Casadevall N, Chretien S, Dusanter I, Gisselbrecht S 1991 Structure of the murine erythropoietin receptor complex. Characterization of the erythropoietin cross-linked proteins. J Biol Chem 266:23380–23385[Abstract/Free Full Text]
17. Kiselar JG, Downard KM 2000 Preservation and detection of specific antibody-peptide complexes by matrix-assisted laser desorption ionization mass spectrometry. J Am Soc Mass Spectrom 11:746–750[CrossRef][Medline]
18. Johnson DL, Joliffe J 2000 Erythropoietin mimetic peptides and the future. Nephrol Dial Transplant 15:1274–1277[Free Full Text]
19. Weaver L, Stagsted J, Behnke O, Matthews BW, Olsson L 1997 ß-Sheet models for the ordered filamentous structure formed by a peptide that enhances the action of insulin. J Struct Biol 117:165–172
20. Naranda T, Goldstein A, Olsson L 1997 A peptide derived from an extracellular domain selectively inhibits receptor internalization: target sequences on insulin and insulin-like growth factor 1 receptors. Proc Natl Acad Sci USA 94:11692–11697[Abstract/Free Full Text]
21. Stagsted J, Reaven GM, Hansen T, Goldstein A, Olsson L 1990 Regulation of insulin receptor functions by a peptide derived from a major histocompatibility complex class I antigen. Cell 62:297–307[CrossRef][Medline]
22. Olsson L, Goldstein A, Stagsted J 1994 Regulation of receptor internalization by the major histocompatibility complex class I molecule. Proc Natl Acad Sci USA 91:9086–9090[Abstract/Free Full Text]

This article has been cited by other articles:

				R. M. Stroud and J. A. Wells Mechanistic Diversity of Cytokine Receptor Signaling Across Cell Membranes Sci. Signal., May 4, 2004; 2004(231): re7 - re7. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Naranda, T. Articles by Olsson, L. Search for Related Content PubMed PubMed Citation Articles by Naranda, T. Articles by Olsson, L.