HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 281, Issue 43, 32485-32495, October 27, 2006

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 281/43/32485    most recent M606179200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dean, T. Articles by Gardella, T. J. Search for Related Content PubMed PubMed Citation Articles by Dean, T. Articles by Gardella, T. J. Social Bookmarking                      What's this?

Role of Amino Acid Side Chains in Region 17–31 of Parathyroid Hormone (PTH) in Binding to the PTH Receptor^*

Thomas Dean, Ashok Khatri, Zhanna Potetinova, Gordon E. Willick, and Thomas J. Gardella¹

From the Endocrine Unit, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts 02114 and the Institute for Biological Sciences, National Research Council, Ottawa, Ontario K1A 0R6, Canada

Received for publication, June 28, 2006 , and in revised form, August 9, 2006.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The principal receptor-binding domain (Ser¹⁷–Val³¹) of parathyroid hormone (PTH) is predicted to form an amphiphilic -helix and to interact primarily with the N-terminal extracellular domain (N domain) of the PTH receptor (PTHR). We explored these hypotheses by introducing a variety of substitutions in region 17–31 of PTH-(1–31) and assessing, via competition assays, their effects on binding to the wild-type PTHR and to PTHR-delNt, which lacks most of the N domain. Substitutions at Arg²⁰ reduced affinity for the intact PTHR by 200-fold or more, but altered affinity for PTHR-delNt by 4-fold or less. Similar effects were observed for Glu substitutions at Trp²³, Leu²⁴, and Leu²⁸, which together form the hydrophobic face of the predicted amphiphilic -helix. Glu substitutions at Arg²⁵, Lys²⁶, and Lys²⁷ (which forms the hydrophilic face of the helix) caused 4–10-fold reductions in affinity for both receptors. Thus, the side chains of Arg²⁰, together with those composing the hydrophobic face of the ligand's putative amphiphilic -helix, contribute strongly to PTHR-binding affinity by interacting specifically with the N domain of the receptor. The side chains projecting from the opposite helical face contribute weakly to binding affinity by different mechanisms, possibly involving interactions with the extracellular loop/transmembrane domain region of the receptor. The data help define the roles that side chains in the binding domain of PTH play in the PTH-PTHR interaction process and provide new clues for understanding the overall topology of the bimolecular complex.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Parathyroid hormone (PTH)² plays a key role in calcium and phosphate homeostasis and has potent effects on the bone-remodeling process. PTH interacts with a Class 2 G protein-coupled receptor that is prominently expressed in bone osteoblasts and in cells located in the proximal and distal portions of the renal convoluted tubules. The PTH receptor (PTHR) is also expressed in the primordia of developing long bones, heart, mammary glands, and other tissues, where it mediates the morphogenic actions of PTH-related protein (PTHrP) (1).

For both PTH and PTHrP, the bioactive portions of the molecule reside within the first 34 amino acids of the processed polypeptides. Within region 1–34, the principal determinants of receptor-binding affinity and receptor-signaling activity map to the C- and N-terminal domains, respectively (2, 3). Solution-phase NMR studies of PTH-(1–34)-based ligands typically show a well formed -helix in the region of the C-terminal binding domain (4, 5). This C-terminal -helix, extending approximately from Ser¹⁷ to Val³¹ (5), exhibits strong amphiphilic character (6–8). PTH analog substitution studies have shown that the side chains of Trp²³, Leu²⁴, and Leu²⁸, which form the hydrophobic face of this -helix, are particularly important for efficient interaction with the receptor (9–11).

The mechanism by which PTH interacts with its receptor has been investigated via the approaches of ligand analog design, receptor mutagenesis, and photochemical cross-linking (reviewed in Ref. 12). The view that has emerged from these studies is that the overall mechanism consists of two principal and, to some extent, autonomous components: 1) an interaction between the C-terminal helical domain of the ligand and the N-terminal extracellular domain (N domain; spanning Tyr²³ to approximately Ile¹⁹⁰) of the mature receptor and 2) an interaction between the N-terminal portion of the ligand and the juxtamembrane domain (J domain) of the receptor containing the extracellular loops and seven transmembrane helices. The N domain component of the interaction is thought to provide the major portion of binding energy and stability to the complex, and the J domain component is thought to mediate the conformational changes involved in receptor activation (13). It now seems likely that most, if not all, of the 15 or so other Class 2 G protein-coupled receptors utilize a similar two-site binding mechanism for interacting with their cognate peptide ligands (14–16).

Consistent with such a two-site binding mechanism for PTH and the PTHR, we have shown that N-terminal PTH peptide fragments, such as PTH-(1–14), bind only extremely weakly to the receptor, but can nevertheless induce at least measurable increases in cAMP levels in PTHR-expressing cells (17). The potency of such N-terminal PTH fragments can be significantly enhanced by introducing substitutions that improve the affinity of the ligand for the PTHR J domain (18–20). Thus, the analog [Aib^1,3,Gln¹⁰,Har¹¹,Trp¹⁴]PTH-(1–14)-NH₂ is equipotent to PTH-(1–34) in stimulating cAMP as well as inositol phosphate (IP) production in PTHR-expressing cells. Moreover, such optimized N-terminal PTH analogs retain full potency in cells expressing PTHR-delNt, a PTHR construct that lacks most (Ala²⁴–Arg¹⁸¹) of the N domain, whereas unmodified PTH-(1–34) exhibits at least 100-fold reductions in potency and affinity for PTHR-delNt compared with its actions on the intact PTHR (18–20).

The large reduction in affinity/potency that unmodified PTH-(1–34) exhibits for PTHR-delNt can be attributed, according to the two-site interaction model, to the loss of the binding interactions that normally occur between the C-terminal helical domain of the ligand and the N domain of the receptor. The cross-linking approach has indeed established spatial proximities between PTH residues 23, 27, and 28, when substituted with para-benzoyl-L-phenylalanine (Bpa), and the N domain of the receptor (21, 22). On the other hand, [Lys²⁷(Bp)₂]PTH-(1–34)-NH₂, which contains the photoreactive Bp moiety attached to the Lys²⁷ side chain amino groups, was shown to cross-link to the first extracellular loop of the PTHR (23). This observation raised the possibility that the C-terminal binding domain of PTH can functionally interact with the receptor J domain. Consistent with this possibility, we recently showed that methylation of several backbone nitrogen atoms in region 17–31 of PTH-(1–31) impairs, albeit modestly, the capacity of the ligand to stimulate cAMP formation in cells expressing PTHR-delNt (8). Taken together, these observations point to the uncertainty that exists in our understanding of the specific mechanisms by which the C-terminal domain of PTH contributes to the PTHR-binding process.

This study was undertaken to examine further the mode of action used by the C-terminal binding domain of PTH. We sought to address the roles that the side chains in this domain play in the PTHR-binding process, the general functional importance of amphiphilicity, and the potential for binding interactions with the receptor J domain. Our strategy was to introduce a variety of conservative and non-conservative substitutions in region 17–31 of PTH-(1–31)-NH₂ and to assess their effects on binding to the intact PTHR and to PTHR-delNt. The overall results indicate a dominant role for specific interactions between side chains on the hydrophobic face of the C-terminal helix and the receptor N-terminal domain. They also provide evidence for weak interactions between side chains projecting from the hydrophilic face of the helix and the PTHR J domain. The data also shed new light on the overall topology of the bimolecular complex, as they support folding of the bound ligand and proximity of the binding sites in the N and J domains of the receptor.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Peptide Synthesis—Peptides were based on the human PTH-(1–31)-NH₂ sequence (SVSEIQLMHNLGKHLNSMERVEWLRKKLQDV-NH₂). The alanine substitutions were incorporated into this otherwise unmodified PTH-(1–31)-NH₂ scaffold. Subsequently, cyclohexylalanine (Cha) and Glu substitutions were incorporated into the [Ala¹,Arg¹⁹]PTH-(1–31)-NH₂ scaffold, which, because of the Ser¹ Ala and Glu¹⁹ Arg substitutions, exhibits improved affinity for PTHR-delNt (19, 24). These Ala-, Cha-, and Glu-substituted PTH-(1–31)-NH₂ and [Ala¹,Arg¹⁹]PTH-(1–31)-NH₂ peptides and their corresponding parental controls were synthesized by the Massachusetts General Hospital Biopolymer Core facility using conventional methodologies as we described previously (20). Additional analogs of PTH-(1–31)-NH₂ with substitutions at position 20 were prepared as part of a previous study (25). All peptides were verified by analytical HPLC, matrix-assisted laser desorption ionization mass spectrometry, and amino acid analysis, and peptide concentrations of stock solutions were established by amino acid analysis. The radioligands ¹²⁵I-[Nle^8,21,Tyr³⁴]rPTH-(1–34)-NH₂ and ¹²⁵I-[Aib^1,3,Nle⁸,Gln¹⁰,Har¹¹,Ala¹²,Trp¹⁴,Tyr¹⁵]rPTH-(1–15)-NH₂ (henceforth referred to as ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂) were prepared by the oxidative chloramine-T procedure using Na¹²⁵I (specific activity of 2200 Ci/mmol; PerkinElmer Life Sciences) and purified by reversed-phase HPLC.

Circular Dichroism—CD spectra were obtained on a Jasco J-600 spectropolarimeter at 20 °C. Four spectra were averaged, and the data were smoothed by the Jasco software. The instrument was calibrated with ammonium (+)-10-camphorsulfonate. Data are expressed as the number of helical residues/peptide chain as calculated from –[]₂₂₂ x 30/28,000, where []₂₂₂ is the mean residue ellipticity at 222 nm, as we described previously (8).

Cell Culture—Cells were cultured at 37 °C in a humidified atmosphere containing 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (HyClone, Logan UT), 100 units/ml penicillin G, and 100 µg/ml streptomycin sulfate (Invitrogen). For binding and cAMP experiments performed with the intact PTHR, the HKRK-B7 and ROS 17/2.8 cell lines were used. HKRK-B7 cells are derived from the porcine kidney cell line LLC-PK₁ and express, via stable DNA transfection, the wild-type human PTHR at an approximate surface density of 950,000 PTH-binding sites/cell (26). ROS 17/2.8 cells are rat osteosarcoma cells and express the endogenous PTHR at an approximate surface density of 70,000 PTH-binding sites/cell (27). The cells were plated and assayed in 24-well plates.

PTHR-delNt was derived from the human PTHR by site-directed mutagenesis and lacks most (Ala²⁴–Arg¹⁸¹) of the N domain (28). PTHR-delNt was expressed in COS-7 cells via transient DNA transfection. For binding assays, cell membranes were prepared from the transfected COS-7 cells. To increase the maximum binding of ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ to PTHR-delNt in these membranes, the cells were cotransfected with a negative-dominant mutant G_s protein (G_s(₃₅/Gly²²⁶ Ala/Ala³⁶⁶ Ser); hereafter referred to as G_s^ND). This mutant G_s subunit is thought to couple to cognate receptors, and thus stabilize high affinity receptor conformations more efficiently than does wild-type G_s without increasing basal cAMP levels (29). We recently used a precursor of this G_s mutant, G_s(₃₅), to increase the binding of ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ to PTHR-delNt in COS-7 cell membranes (18). We subsequently found that G_s^ND, which contains the same five-amino acid replacement of the corresponding G_i residues in the ₃₅ loop as does G_s(₃₅) plus the point mutations Gly²²⁶ Ala, which increases affinity for G/, and Ala³⁶⁶ Ser, which decreases affinity for GDP (29), yielded 2-fold higher levels of specific binding of ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ than did G_s(₃₅) (data not shown). The COS-7 cells were cotransfected in 6-well plates using plasmid DNA encoding PTHR-delNt (1 µg/well), plasmid DNA encoding G_s^ND (1 µg/well), and FuGENE 6 reagent (6 µl/well; Roche Diagnostics). Control experiments were performed with COS-7 cells similarly cotransfected with G_s^ND and the wild-type PTHR. Cells were harvested 3 days after transfection, and membranes were prepared as described (18).

Receptor Binding—Binding to the intact PTHR in HKRK-B7 and ROS 17/2.8 cells was assessed using ¹²⁵I-[Nle^8,21, Tyr³⁴]rPTH-(1–34)-NH₂ as a tracer radioligand as described (28). In brief, confluent cells in 24-well plates (500,000 cells/well) were incubated in binding buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 5% heat-inactivated horse serum, 0.5% fetal bovine serum, adjusted to pH 7.7 with HCl) containing radioligand (100,000 cpm/well) with or without unlabeled peptide ligand (3 x 10^–9 to 1 x 10^–5 M) for 4 h at 15 °C. The binding mixture was then removed by aspiration' the cells were rinsed three times with binding buffer and lysed in 1 M NaOH; and the entire lysate was counted for -irradiation in a -counter. Binding to PTHR-delNt in COS-7 cell membranes was assessed in 96-well vacuum filtration plates (Multi-Screen-HV Durapore, 0.65-µm membranes; Millipore Corp.) using ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ as a tracer radioligand as described (18). In brief, cell membranes (20 µg/well) were incubated in membrane binding buffer containing radioligand (30,000 cpm/well) with or without unlabeled peptide ligand (3 x 10^–9 to 1 x 10^–5 M) for 90 min at 21 °C (reaction volume of 200 µl). The plates were then subjected to rapid vacuum filtration, and the filters were washed once with buffer, airdried, detached from the plate, and counted for -irradiation in a -counter. Nonspecific binding was defined as the binding observed in the presence of 1 x 10^–6 M PTH-(1–31)-NH₂ for HKRK-B7 and ROS 17/2.8 cells and of 1 x 10^–6 M [Aib^1,3,M]rPTH-(1–15)-NH₂ for PTHR-delNt. Specifically bound radioactivity was calculated as a percentage of the radioactivity specifically bound in the absence of competing ligand.

Stimulation of Intracellular cAMP and IP—The capacities of the ligands to stimulate formation of cAMP were assessed in intact ROS 17/2.8 cells, as described (28). In brief, cells in 24-well plates were incubated in binding buffer containing the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (2 mM) with or without a peptide ligand (3 x 10^–11 to 1 x 10^–6 M) for 30 min at room temperature. The medium was then removed, and the cells were lysed by adding 50 mM HCl and freezing the plate on dry ice. The cAMP in the thawed lysate was quantified by radioimmunoassay. The stimulation of production of inositol phosphates (IP₁ + IP₂ + IP₃) was assessed in COS-7 cells transfected with the intact human PTHR as we described previously (28). In brief, intact transfected COS-7 cells in 24-well plates were labeled with myo-[³H]inositol (specific activity of 25 Ci/mmol; PerkinElmer Life Sciences) for 16 h. The labeled cells were then treated for 30 min with ligand in the presence of LiCl₂ (30 mM), and the medium was removed and replaced with ice-cold trichloroacetic acid (5%). After 2 h on ice, the acid lysates were extracted with ether and processed by ion-exchange chromatography (0.5-ml resin bed), and the ammonium formate-eluted [³H]inositol phosphates were quantified by liquid scintillation counting.

Data and Statistical Calculations—Binding and cAMP data were processed for curve fitting and derivation of IC₅₀ and EC₅₀ values using least-squares nonlinear regression analysis and the following equation: y = y_min + (y_max–y_min/1 + (IC₅₀/x)ⁿ, where y, y_min, and y_max are the observed, minimum, and maximum response values, respectively; x is the ligand concentration; and n is the slope factor. In cases in which incomplete inhibition of binding occurred, e.g. with certain PTH-(1–31)-NH₂ analogs binding to PTHR-delNt, the curves were extrapolated to nonspecific binding. Paired data sets were statistically compared using a two-tailed Student's t test, assuming unequal variance for the two sets.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Alanine Scan of Region 17–31 of PTH—We first individually replaced each residue in region 17–31 of PTH-(1–31)-NH₂ with alanine and assessed the effects of the substitutions on binding to the intact human PTHR stably expressed in HKRK-B7 cells. Binding was assessed by competition methods using ¹²⁵I-[Nle^8,21,Tyr³⁴]rPTH-(1–34)-NH₂ as a tracer radioligand. The parental PTH-(1–31)-NH₂ peptide fully inhibited the binding of this tracer with an IC₅₀ of 68 ± 10 nM (Fig. 1A and Table 1). The various alanine substitutions had a range of effects on this binding. Most dramatic was that of the Arg²⁰ Ala substitution, which abolished detectable binding (Fig. 1A). The alanine substitutions at Trp²³ and Leu²⁴ reduced the apparent binding affinity by 19- and 12-fold, respectively, relative to the parental peptide (p < 0.05). The alanine substitutions at Val²¹, Arg²⁵, Lys²⁷, Leu²⁸, and Val³¹ reduced affinity by 3-fold, and the remaining alanine substitutions altered affinity by 2-fold or less (Fig. 1, A and B; and Table 1). The alanine substitutions at Glu¹⁹, Glu²², and Gln²⁹ each produced a small (2-fold) enhancement of the apparent binding affinity, as did the Glu¹⁹ Arg substitution, which we have shown previously to enhance cAMP-stimulating potency in PTH-(1–34) and PTH-(1–20) peptides (24). The Ala²² substitution was thus paired with Ala¹⁹ as well as with Arg¹⁹, but neither pairing improved affinity further relative to the single substitutions alone (Table 1).

Unlabeled [Aib^1,3,M]rPTH-(1–15)-NH₂ (used as a control peptide) fully inhibited the binding of ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ to these membranes with high apparent affinity (IC₅₀ = 2.2 ± 0.5 nM) (Fig. 1, C and D; and Table 1). As expected from the absence of ligand interactions with the PTHR N domain, unmodified PTH-(1–31)-NH₂ bound to PTHR-delNt with relatively low affinity (IC₅₀ = 3700 ± 400 nM) (Fig. 1, C and D; and Table 1). This binding was nevertheless sufficient to assess the effects of the alanine substitutions on the capacity of the ligand to interact with PTHR-delNt. None of the alanine substitutions altered binding to PTHR-delNt by >5-fold, including the Arg²⁰ Ala substitution, which had strongly diminished binding to the intact PTHR (Fig. 1, A versus C). Similarly, the Ala substitutions at Trp²³ and Leu²⁴, which reduced affinity for the PTHR by 19- and 12-fold, respectively, reduced affinity for PTHR-delNt by only 2-fold. These findings indicate that the mechanisms by which the Ala substitutions at Arg²⁰, Trp²³, and Leu²⁴ impair binding to the intact PTHR are largely independent of interactions with the PTHR J domain.

View larger version (42K): [in this window] [in a new window]   FIGURE 1. Alanine scan of region 17–31 of PTH-(1–31)-NH2. PTH-(1–31)-NH2 and analogs thereof with individual alanine substitutions in region 17–31 were assessed by competition methods for binding to the intact PTHR (A and B) and to PTHR-delNt (C and D). Intact HKRK-B7 cells stably transfected with the PTHR and the 125I-[Nle8,21,Tyr34]rPTH-(1–34)-NH2 (125I-PTH(1–34)) tracer radioligand were used in the assays in A and B; membranes prepared from COS-7 cells transiently transfected with PTHR-delNt and the 125I-[Aib1,3,M]rPTH-(1–15)-NH2 tracer radioligand were used in C and D. To increase the total binding of 125I-[Aib1,3,M]rPTH-(1–15)-NH2, the COS-7 cells were cotransfected with a negative-dominant Gs mutant. For the experiments with PTHR-delNt, unlabeled [Aib1,3,M]rPTH-(1–15)-NH2 () was used as a control and to determine nonspecific binding, to which the curves for the PTH-(1–31) analogs were extrapolated. Data for the parental PTH-(1–31) peptide (• and dotted line) are shown in each panel. Data are the means ± S.E. of three or more experiments, each performed in duplicate. SB, specific binding.

Substitutions with Cha—The strong effects that the alanine substitutions at Trp²³, Leu²⁴, and Leu²⁸ had on binding to the intact PTHR (Fig. 1, A and B) suggested that hydrophobicity per se in region 17–31 of PTH might play a key role in determining the affinity of ligand for the receptor. To evaluate this further, we substituted each residue in region 17–31 of PTH with Cha, an amino acid analog that would preserve bulk hydrophobicity, yet alter the specific chemistry and topology of the side chain at the substituted site. We used [Ala¹,Arg¹⁹]PTH-(1–31)-NH₂ as a scaffold peptide in these experiments, as we wanted to augment our capacity to assess binding to PTHR-delNt, and the Glu¹⁹ Arg and Ser¹ Ala substitutions were known to improve interaction with this truncated receptor (19, 24).

The effects of the Cha substitutions on the intact PTHR in HKRK-B7 cells generally paralleled those of the corresponding alanine substitutions. Thus, relative to the parental [Ala¹,Arg¹⁹]PTH-(1–31)-NH₂ peptide, the Cha substitution at Arg²⁰ had the strongest effect on binding and reduced affinity by 120-fold (p = 0.005). Substitutions at Trp²³ and Leu²⁴ reduced affinity by 14- and 11-fold, respectively (p < 0.002), and those at the remaining positions altered PTHR-binding affinity by 6-fold or less (Fig. 3, A and B; and Table 2). The deleterious effects that the Cha substitutions at Trp²³ and Leu²⁴ had on PTHR-binding affinity indicate that hydrophobicity per se is not the main physicochemical property of these two side chains that underlies their contributions to PTHR-binding affinity.

The Cha substitutions at Arg²⁵ and Lys²⁶ reduced affinity for PTHR-delNt by 5–7-fold, and the remaining Cha substitutions, including those at Arg²⁰,Trp²³, and Leu²⁴, altered apparent affinity by 3-fold or less (Fig. 3, C and D; and Table 2). These results are in agreement with those obtained with the corresponding Ala substitutions in that they suggest that the detrimental effects that substitutions at positions 20, 23, and 24 have on binding to the intact PTHR are not due to altered interactions with the PTHR J domain, but rather involve interactions with the receptor N domain. They also suggest that Arg²⁵ and Lys²⁶ can contribute to PTHR-binding affinity via mechanisms that are not dependent on interactions with the receptor N domain.

View larger version (19K): [in this window] [in a new window]   FIGURE 2. CD spectroscopy of alanine-substituted PTH-(1–31)-NH2 analogs. The parental PTH-(1–31)-NH2 peptide and derivatives thereof altered by a single alanine substitution in region 17–31 were analyzed by CD spectroscopy. The negative deflections in the mean residue ellipticity ([]) in the regions at 209 and 222 nm of the spectra are indicative of -helical structure. The CD spectra obtained for other peptides used in this study are shown in supplemental Fig. 1. For each peptide, the number of helical residues/peptide chain was calculated from []222, and the resulting values are reported in Tables 1 and 2. deg, degrees.

View larger version (44K): [in this window] [in a new window]   FIGURE 3. Cha scan of region 17–31 of [Ala1,Arg19]PTH-(1–31)-NH2. Residues in region 17–31 of [Ala1,Arg19]PTH-(1–31)-NH2 ([A1,R1]PTH(1–31)) were individually replaced with Cha, and the effects on binding to the intact PTHR (A and B) and to PTHR-delNt (C and D) were assessed as described in the legend to Fig. 1. Data are the means ± S.E. of three or more experiments, each performed in duplicate. 125I-PTH(1–34) SB, 125I-[Nle8,21,Tyr34]rPTH-(1–34)-NH2-specific binding.

View larger version (38K): [in this window] [in a new window]   FIGURE 4. Effects of Glu substitutions in region 19–28 of [Ala1,Arg19]PTH-(1–31)-NH2. Residues in region 19–28 of [Ala1,Arg19]PTH-(1–31)-NH2 ([A1,R1]PTH(1–31)) were replaced with Glu, and the effects on binding to the intact PTHR (A and B) and to PTHR-delNt (C and D) were assessed as described in the legend to Fig. 1. Data are the means ± S.E. of three or more experiments, each performed in duplicate. 125I-PTH(1–34) SB, 125I-[Nle8,21,Tyr34]rPTH-(1–34)-NH2-specific binding.

The replacement of Arg²⁰ in PTH-(1–31)-NH₂ with Gln, Glu, Lys, (S)-2-amino-4-((2-amino)pyrimidinyl)butanoic acid (Apa), or L-(4-guanidino)phenylalanine (Gph) resulted in a complete loss of binding to the intact PTHR, and the substitutions with PipGly, Nle, and citrulline reduced affinity by 200-fold relative to the parental peptide (Fig. 5, A and B; and Table 1). None of the position 20 substitutions altered binding to PTHR-delNt by >5-fold (Fig. 5, C and D; and Table 1). These results thus indicate that the effects of the substitutions at position 20 on binding to the PTHR are largely independent of interactions with the PTHR J domain.

Effects on cAMP and IP Signaling—Selected analogs of [Ala¹,Arg¹⁹]PTH-(1–31)-NH₂ with Cha or Glu substitutions that markedly impaired binding to the intact PTHR were assessed for their capacity to stimulate cAMP production in ROS 17/2.8 cells. These cells endogenously express the rat PTHR at a moderate level (70,000 PTHRs/cell) and were thus considered more useful for correlating effects on binding affinity and cAMP-signaling potency compared with HKRK-B7 cells, which express 13-fold higher levels of receptor. Competition binding assays confirmed that the selected substitutions strongly impaired binding to ROS 17/2.8 cells and revealed effects on affinity that paralleled those seen in HKRK-B7 cells (Fig. 6A and Table 3). These effects on affinity in ROS 17/2.8 cell were accompanied by parallel reductions in cAMP-stimulating potency (Fig. 6B and Table 3). Although potency was reduced, each substituted analog (at the highest concentration) produced approximately the same maximum cAMP response as did the parental peptide.

View larger version (32K): [in this window] [in a new window]   FIGURE 5. Substitution analysis of Arg20. The effects of replacing the highly conserved arginine at position 20 of PTH-(1–31)-NH2 with various encoded (Gln, Glu, and Lys) or non-encoded (Nle, citrulline (Cit), ornithine (Orn), Apa, Gph, and PipGly (PipG)) amino acids on binding to the intact PTHR (A and B) and to PTHR-delNt (C and D) were assessed by competition methods as described in the legend to Fig. 1. Data are the means ± S.E. of three or more experiments, each performed in duplicate. 125I-PTH(1–34) SB, 125I-[Nle8,21,Tyr34]rPTH-(1–34)-NH2-specific binding.

View larger version (19K): [in this window] [in a new window]   FIGURE 6. Binding and cAMP-stimulating activities in ROS 17/2.8 cells. The parental peptide [Ala1,Arg19]PTH-(1–31)-NH2 ([A1,R1]PTH(1–31)) and Glu- or Cha-substituted analogs thereof were evaluated for the capacity to bind to the endogenous PTHR in ROS 17/2.8 cells (A) and to stimulate cAMP formation in these cells (B). Competition binding studies were performed in intact cells using 125I-[Nle8,21,Tyr34]rPTH-(1–34)-NH2 (125I-PTH(1–34)) as a tracer radioligand. Data are the means ± S.E. of data of three experiments, each performed in duplicate. SB, specific binding.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The greatest impact on the binding of our PTH-(1–31) peptides to the intact PTHR occurred with the non-conservative Glu substitutions at Arg²⁰, Trp²³, Leu²⁴, and Leu²⁸,as each reduced apparent affinity by 150-fold or more. Such effects are consistent with previous PTH substitution studies showing important roles for these four residues in binding to the intact PTHR (6, 9–11). The same Glu substitutions had little or no effect on binding to PTHR-delNt.

Our binding experiments performed with PTHR-delNt utilized membranes prepared from COS-7 cells that were cotransfected with PTHR-delNt and G_s^ND. (The negative-dominant G_s mutant was utilized to improve the total binding of the ¹²⁵I-[Aib^1,3,M]rPTH-(1–15)-NH₂ radioligand to the membrane preparations.) To assess the possibility that G_s^ND per se could account for the lack of effect that the substitutions had on binding to PTHR-delNt, we performed control experiments in COS-7 cell membranes to evaluate the effects of several key substitutions (Cha²⁰, Glu²⁰, Glu²³, and Glu²⁴) on binding to the intact wild-type PTHR in the presence of G_s^ND. The experiments clearly showed that the substitutions again caused large decreases in binding affinity for the intact PTHR, even in the presence of G_s^ND (supplemental Fig. 3). The lack of an effect of the substitutions on binding to PTHR-delNt must therefore be due to the absence of the receptor N domain and not to the presence of the G_s^ND mutant.

Our CD analyses indicated that the substitutions did not cause major perturbations in the helical structure of the peptide. It thus seems clear from the data that the side chains of Arg²⁰, Trp²³,Leu²⁴, and Leu²⁸ contribute to the PTHR-binding process by mechanisms that are largely, if not completely, dependent on interactions with the receptor N-terminal domain.

The side chains of Trp²³, Leu²⁴, and Leu²⁸ form the hydrophobic face of the amphiphilic -helix predicted to reside within region 17–31 of PTH (6, 7). Our data predict that this hydrophobic helical face contributes to the PTHR-binding process by interacting with the N-terminal domain of the receptor. Our findings do not support a mechanism by which this hydrophobic face contributes to the PTHR-binding process by interacting nonspecifically with the lipid component of the cell membrane, as has been discussed for PTH (32) and for amphiphilic peptide ligands in general (33). If this were the case, then the substitutions would have impacted equally the binding of the ligand to the intact PTHR and to PTHR-delNt, which did not occur. Moreover, the hydrophobic Cha substitutions at Trp²³ and Leu²⁴ reduced binding to the intact PTHR by 12-fold and again had little or no effect on binding to PTHR-delNt (Table 2). Thus, the mechanisms by which the side chains of Try²³ and Leu²⁴ contribute to the PTHR-binding process are not likely to be based simply on nonspecific hydrophobic interactions with the lipid bilayer, but instead involve additional spatial and chemical features of the side chains and their specific interactions with cognate functional groups in the receptor.

It is clear from our data and that presented elsewhere (10, 11, 25) that Arg²⁰ of PTH plays a key role in the PTHR-binding process. Each of 10 substitutions tested at this position in PTH-(1–31)-NH₂ reduced affinity for the intact PTHR by at least 200-fold, and most (seven) abolished detectable binding. Each of these substitutions had only a minor effect on binding to PTHR-delNt. The side chain of Arg²⁰, like those of Trp²³, Leu²⁴, and Leu²⁸, must therefore contribute to the PTHR-binding process via a mechanism that primarily involves interactions with the PTHR N domain. The molecular nature of these interactions is not clear at present, but likely involves multiple components of the arginine side chain, including the cationic and H-bonding nitrogen atoms of the guanidino group and the three methylenes of the linker (25). Our data now imply that these functional groups of the Arg²⁰ side chain fit within a highly specific binding pocket within the N domain of the receptor.

Unlike Arg²⁰, residue 19 of PTH appears to interact predominantly with the PTHR J domain (24). This can be seen in our present data by the 90-fold reduction in binding affinity for PTHR-delNt caused by the Arg¹⁹ Glu substitution, the strongest effect on binding to PTHR-delNt of any substitution in this study. This reduction in binding affinity mirrors the enhancing effect that the Glu¹⁹ Arg substitution in either PTH-(1–34) or PTH-(1–20) analogs has on cAMP-stimulating potency in COS-7 cells expressing PTHR-delNt (24). That residue 19 of the bound ligand is in spatial proximity to the PTHR J domain is further suggested by the cross-linking of Bpa¹⁹-containing PTH analogs to the extracellular end of transmembrane helix 2 of the PTHR (34). When considered together, the data for residues 19 and 20 suggest that the 19/20-position in the ligand marks a point of divergence for the ligand segments that interact predominantly with the N and J domains of the receptor: regions 20–31 and 1–19 of PTH, respectively.

We observed subtle but consistent effects of substitutions at positions 21 and 25–27 in our PTH-(1–31) peptides on interaction with PTHR-delNt. These effects were accompanied by approximately proportional effects on interaction with the intact PTHR. Such findings suggest that the side chains of these residues, although not making major contributions to overall binding energy, can promote binding via mechanisms that involve interactions with the PTHR J domain. The cationic side chains of Arg²⁵, Lys²⁶, and Lys²⁷ form the hydrophilic face of the predicted amphiphilic -helix in the C-terminal domain of the ligand, and Val²¹ lies at the edge of this face (8). It is possible that the side chains projecting from this helical face contribute to binding indirectly, for example, by interacting with the phospholipid head group and/or aliphatic components of the cell membrane (32, 33), as discussed above. Another possibility is that these side chains interact with anionic and/or hydrophobic groups in the extracellular loops and/or transmembrane domain regions of the receptor. The cross-linking of [Lys²⁷(Bp)₂]PTH-(1–34)-NH₂ to the first extracellular loop of the PTHR indeed suggests a physical proximity of this helical face in the ligand and the PTHR J domain (23). The possibility for a modest functional interaction between the C-terminal helix and the PTHR J domain is also supported by our recent finding that backbone methylations at Ser¹⁷,Trp²³, and Lys²⁶ in PTH-(1–31)-NH₂ impair the capacity of the ligand to stimulate cAMP production via PTHR-delNt (8).

If both the C-terminal (region 20–31) and N-terminal (region 1–19) domains of the ligand interact with the PTHR J domain, then a bend would most likely be required between the two domains of the receptor-bound ligand. The tertiary structure of receptor-bound PTH has been a matter of some debate, with linear and folded structures supported (4, 35). A folded helix-turn-helix structure for PTH-(1–34) was indeed predicted early on based on modeling and structure-activity data (36), and solution-phase NMR studies of PTH and PTHrP ligands generally reveal mid-region flexibility or a hinge between the N- and C-terminal domains (4, 5, 37, 38). A bend in receptor-bound PTH-(1–34) has more recently been suggested by the cross-linking of both [Bpa¹¹]PTH-(1–34) and [Bpa²¹]PTH-(1–34) to the same segment (Ala¹⁶⁵–Asn¹⁷⁶) of the PTHR N domain (39).

In addition to the tertiary structure of the bound ligand, our data have implications for the topology of the occupied receptor and the spatial relationship of the N and J domains. If the C-terminal -helical domain (region 20–31) of PTH interacts with both the N and J domains of the receptor via its hydrophobic and hydrophilic faces, respectively, as supported by this study, then the N and J domain-binding sites in the receptor must be near each other. This possibility is again supported by cross-linking data: specifically, the cross-linking of one PTH-(1–34) analog with a photolabile Bp moiety incorporated at position 27 as a Bpa substitution with the receptor N domain (22) and that of another PTH-(1–34) analog with the Bp moiety attached to the distal amino groups of the Lys²⁷ side chain with the first extracellular loop of the receptor (23). Because the spatial positioning of the Bp moieties in these two ligands is likely to be similar, a proximity of the two cross-linked sites in the receptor is also likely.

The available data derived from the functional and cross-linking studies combined thus suggest some intriguing hypotheses regarding the topology and domain architecture of the PTH·PTHR complex. However, the lack of direct structural information on receptor-bound PTH or on the intact PTHR itself hampers our capacity to assimilate such data into a detailed three-dimensional model of the PTH·PTHR complex. One important goal that may be facilitated by our work is to identify sites in the receptor that are used by key ligand residues such as Arg²⁰, Trp²³, Leu²⁴, and Leu²⁸. Our results predict that such interaction sites will be located in the receptor N-terminal domain. For Trp²³, the extreme N-terminal segment (Thr³³–Leu⁴⁰) of the N domain needs to be considered, as it contains the cross-linking site for [Bpa²³]PTHrP-(1–36) (34). For the other ligand residues, few clues are available: cross-linking has not been achieved for positions 20 and 24, and [Bpa²⁸]PTHrP-(1–36) cross-links to a nonessential segment (Ser⁶¹–Gly¹⁰⁵) of the N-terminal domain (34).

The recently reported NMR-derived structure of the N-terminal domain of the related corticotropin-releasing factor receptor (15) may open paths for structure-based analyses of the ligand interaction sites in the N domains of the Class 2 G protein-coupled receptors (14). Even with such an approach, additional functional studies will be needed. The new PTH analogs presented here should be of value in this regard, as they can be used in conjunction with PTHR mutants altered at candidate sites in the N-terminal domain to probe for allele-specific rescue effects. This is a direction that we hope to pursue in future studies.

	FOOTNOTES

 
^* This work was supported in part by National Institutes of Health Grant DK-11794 (to T. J. G.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. 1–3.

¹ To whom correspondence should be addressed: Endocrine Unit, Massachusetts General Hospital, Wellman Bldg., Rm. 503C, 50 Blossom St., Boston, MA 02114. Tel.: 617-726-3966; Fax: 617-726-7543; E-mail: gardella{at}helix.mgh.harvard.edu.

² The abbreviations used are: PTH, parathyroid hormone; PTHR, parathyroid hormone receptor; PTHrP, parathyroid hormone-related protein; N domain, N-terminal extracellular domain; J domain, juxtamembrane domain; IP, inositol phosphate; Aib, -aminoisobutyric acid; Har, homoarginine; Bpa, para-benzoyl-L-phenylalanine; Bp, benzophenone; Cha, cyclohexylalanine; HPLC, high pressure liquid chromatography; Nle, norleucine; rPTH, rat parathyroid hormone; PipGly, (S)-4-piperidyl-N-amidino)glycine; Apa, (S)-2-amino-4-((2-amino)pyrimidinyl)butanoic acid; Gph, L-(4-guanidino)phenylalanine.

	ACKNOWLEDGMENTS

 
We thank Dr. John T. Potts, Jr., for careful review of the manuscript and Dr. Catherine Berlot for kindly providing the plasmid encoding G_s^ND.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Strewler, G. J. (2000) N. Engl. J. Med. 342, 177–185[Free Full Text]
2. Tregear, G. W., Van Rietschoten, J., Greene, E., Keutmann, H. T., Niall, H. D., Reit, B., Parsons, J. A., and Potts, J. T., Jr. (1973) Endocrinology 93, 1349–1353[Abstract/Free Full Text]
3. Nussbaum, S. R., Rosenblatt, M., and Potts, J. T., Jr. (1980) J. Biol. Chem. 255, 10183–10187[Free Full Text]
4. Pellegrini, M., Royo, M., Rosenblatt, M., Chorev, M., and Mierke, D. F. (1998) J. Biol. Chem. 273, 10420–10427[Abstract/Free Full Text]
5. Chen, Z., Xu, P., Barbier, J.-R., Willick, G., and Ni, F. (2000) Biochemistry 39, 12766–12777[CrossRef][Medline] [Order article via Infotrieve]
6. Neugebauer, W., Surewicz, W. K., Gordon, H. L., Somorjai, R. L., Sung, W., and Willick, G. E. (1992) Biochemistry 31, 2056–2063[CrossRef][Medline] [Order article via Infotrieve]
7. Epand, R. E. (1983) Mol. Cell. Biol. 57, 41–47
8. Barbier, J.-R., Gardella, T. J., Dean, T., MacLean, S., Potetinova, Z., Whitfield, J. F., and Willick, G. E. (2005) J. Biol. Chem. 280, 23771–23777[Abstract/Free Full Text]
9. Gardella, T. J., Wilson, A. K., Keutmann, H. T., Oberstein, R., Potts, J. T., Jr., Kronenberg, H. M., and Nussbaum, S. R. (1993) Endocrinology 132, 2024–2030[Abstract/Free Full Text]
10. Oldenburg, K. R., Epand, R. F., D'Orfani, A., Vo, K., Selick, H., and Epand, R. M. (1996) J. Biol. Chem. 271, 17582–17591[Abstract/Free Full Text]
11. Reidhaar-Olson, J., Davis, R., De Souza-Hart, J., and Selick, H. (2000) Mol. Cell. Endocrinol. 160, 135–147[CrossRef][Medline] [Order article via Infotrieve]
12. Gensure, R. C., Gardella, T. J., and Jüppner, H. (2005) Biochem. Biophys. Res. Commun. 328, 666–678[CrossRef][Medline] [Order article via Infotrieve]
13. Hoare, S. R. J., Gardella, T. J., and Usdin, T. B. (2001) J. Biol. Chem. 276, 7741–7753[Abstract/Free Full Text]
14. Tan, Y. V., Couvineau, A., Murail, S., Ceraudo, E., Neumann, J. M., Lacapere, J. J., and Laburthe, M. (2006) J. Biol. Chem. 281, 12792–12798[Abstract/Free Full Text]
15. Grace, C. R., Perrin, M. H., DiGruccio, M. R., Miller, C. L., Rivier, J. E., Vale, W. W., and Riek, R. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 12836–12841[Abstract/Free Full Text]
16. Dong, M., Pinon, D. I., and Miller, L. J. (2005) Mol. Endocrinol. 19, 1821–1836[Abstract/Free Full Text]
17. Luck, M. D., Carter, P. H., and Gardella, T. J. (1999) Mol. Endocrinol. 13, 670–680[Abstract/Free Full Text]
18. Dean, T., Linglart, A., Mahon, M. J., Bastepe, M., Jüppner, H., Potts, J. T., Jr., and Gardella, T. J. (2006) Mol. Endocrinol. 20, 931–942[Abstract/Free Full Text]
19. Shimizu, M., Potts, J. T., Jr., and Gardella, T. J. (2000) J. Biol. Chem. 275, 21836–21843[Abstract/Free Full Text]
20. Shimizu, N., Guo, J., and Gardella, T. J. (2001) J. Biol. Chem. 276, 49003–49012[Abstract/Free Full Text]
21. Mannstadt, M., Luck, M. D., Gardella, T. J., and Jüppner, H. (1998) J. Biol. Chem. 273, 16890–16896[Abstract/Free Full Text]
22. Gensure, R. C., Gardella, T. J., and Jüppner, H. (2001) J. Biol. Chem. 276, 28650–28658[Abstract/Free Full Text]
23. Greenberg, Z., Bisello, A., Mierke, D. F., Rosenblatt, M., and Chorev, M. (2000) Biochemistry 39, 8142–8152[CrossRef][Medline] [Order article via Infotrieve]
24. Shimizu, M., Shimizu, N., Tsang, J. C., Petroni, B. D., Khatri, A., Potts, J. T., Jr., and Gardella, T. J. (2002) Biochemistry 41, 13224–13233[CrossRef][Medline] [Order article via Infotrieve]
25. Barbier, J.-R., MacLean, S., Whitfield, J. F., Morley, P., and Willick, G. E. (2001) Biochemistry 40, 8955–8961[CrossRef][Medline] [Order article via Infotrieve]
26. Takasu, H., Guo, J., and Bringhurst, F. (1999) J. Bone Miner. Res. 14, 11–20[CrossRef][Medline] [Order article via Infotrieve]
27. Yamamoto, I., Shigeno, C., Potts, J. T., Jr., and Segre, G. V. (1988) Endocrinology 122, 1208–1217[Abstract/Free Full Text]
28. Shimizu, M., Carter, P. H., Khatri, A., Potts, J. T., Jr., and Gardella, T. J. (2001) Endocrinology 142, 3068–3074[Abstract/Free Full Text]
29. Berlot, C. H. (2002) J. Biol. Chem. 277, 21080–21085[Abstract/Free Full Text]
30. Shimizu, N., Dean, T., Tsang, J. C., Khatri, A., Potts, J. T., Jr., and Gardella, T. J. (2005) J. Biol. Chem. 280, 1797–1807[Abstract/Free Full Text]
31. Chakrabartty, A., Schellman, J. A., and Baldwin, R. L. (1991) Nature 351, 586–588[CrossRef][Medline] [Order article via Infotrieve]
32. Pellegrini, M., Bisello, A., Rosenblatt, M., Chorev, M., and Mierke, D. F. (1998) Biochemistry 37, 12737–12743[CrossRef][Medline] [Order article via Infotrieve]
33. Sargent, D., and Schwyzer, R. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 5774–5778[Abstract/Free Full Text]
34. Gensure, R. C., Shimizu, N., Tsang, J. C., and Gardella, T. J. (2003) Mol. Endocrinol. 17, 2647–2658[Abstract/Free Full Text]
35. Jin, L., Briggs, S. L., Chandrasekhar, S., Chirgadze, N. Y., Clawson, D. K., Schevitz, R. W., Smiley, D. L., Tashjian, A. H., and Zhang, F. (2000) J. Biol. Chem. 275, 27238–27244[Abstract/Free Full Text]
36. Cohen, A. E., Strewler, G. J., Bradley, M. S., Carlquist, M., Nilsson, M., Ericsson, M., Ciardelli, T. L., and Nissenson, R. A. (1991) J. Biol. Chem. 266, 1997–2004[Abstract/Free Full Text]
37. Barden, J. A., and Kemp, B. E. (1994) Biochim. Biophys. Acta 1208, 256–262[CrossRef][Medline] [Order article via Infotrieve]
38. Peggion, E., Mammi, S., Schievano, E., Silvestri, L., Schiebler, L., Bisello, A., Rosenblatt, M., and Chorev, M. (2002) Biochemistry 41, 8162–8175[CrossRef][Medline] [Order article via Infotrieve]
39. Wittelsberger, A., Corich, M., Thomas, B. E., Lee, B. K., Barazza, A., Czodrowski, P., Mierke, D. F., Chorev, M., and Rosenblatt, M. (2006) Biochemistry 45, 2027–2034[CrossRef][Medline] [Order article via Infotrieve]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				I. Assil-Kishawi, T. A. Samra, D. F. Mierke, and A. B. Abou-Samra Residue 17 of Sauvagine Cross-links to the First Transmembrane Domain of Corticotropin-releasing Factor Receptor 1 (CRFR1) J. Biol. Chem., December 19, 2008; 283(51): 35644 - 35651. [Abstract] [Full Text] [PDF]

				R. Mann, M. J. Wigglesworth, and D. Donnelly Ligand-Receptor Interactions at the Parathyroid Hormone Receptors: Subtype Binding Selectivity Is Mediated via an Interaction between Residue 23 on the Ligand and Residue 41 on the Receptor Mol. Pharmacol., September 1, 2008; 74(3): 605 - 613. [Abstract] [Full Text] [PDF]

				T. J. Gardella Crystal Structure of the Extracellular Domain of the PTH/PTHrP Receptor IBMS BoneKEy, May 1, 2008; 5(5): 171 - 175. [Full Text] [PDF]

				A. A. Pioszak and H. E. Xu Molecular recognition of parathyroid hormone by its G protein-coupled receptor PNAS, April 1, 2008; 105(13): 5034 - 5039. [Abstract] [Full Text] [PDF]

				T. Dean, J.-P. Vilardaga, J. T. Potts Jr., and T. J. Gardella Altered Selectivity of Parathyroid Hormone (PTH) and PTH-Related Protein (PTHrP) for Distinct Conformations of the PTH/PTHrP Receptor Mol. Endocrinol., January 1, 2008; 22(1): 156 - 166. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 281/43/32485    most recent M606179200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dean, T. Articles by Gardella, T. J. Search for Related Content PubMed PubMed Citation Articles by Dean, T. Articles by Gardella, T. J. Social Bookmarking                      What's this?