Interactions between Conserved Residues in Transmembrane Helices 1, 2, and 7 of the Thyrotropin-releasing Hormone Receptor*

(Received for publication, November 12, 1996, and in revised form, January 16, 1997)

Jeffrey H. Perlman Dagger §, Anny-Odile Colson , Wei Wang Dagger , Kendra Bence Dagger , Roman Osman and Marvin C. Gershengorn Dagger

From the Dagger  Division of Molecular Medicine, Department of Medicine, Cornell University Medical College and The New York Hospital, New York, New York 10021 and  Department of Physiology and Biophysics, Mount Sinai School of Medicine of the City University of New York, New York, New York 10029

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

The roles of conserved residues in transmembrane helices (TMs) of G protein-coupled receptors have not been well established. A computer-generated model of the thyrotropin-releasing hormone receptor (TRH-R) indicated that conserved Asp-71 (TM-2) could interact with conserved asparagines 316 (TM-7) and 43 (TM-1). To test this model, we constructed mutant TRH-Rs containing polar or alanine substitutions of these residues. The maximal activities of N43A and N316A TRH-Rs were diminished, whereas D71A (Perlman, J. H., Nussenzveig, D. R., Osman, R., and Gershengorn, M. C. (1992) J. Biol. Chem. 267, 24413-24417) and N43A/N316A TRH-Rs were inactive. Computer models of D71A and N43A/N316A TRH-Rs show similar changes from native TRH-R in their TM bundle conformations. The inactivity and the similarity of the computer models of D71A and N43A/N316A TRH-Rs are consistent with the idea that Asp-71 bridges Asn-43 and Asn-316 and suggest that activity is critically dependent on these interactions. The conservation of these residues suggests these specific interactions involving TMs 1, 2, and 7 may be structurally important for all members of the rhodopsin/beta -adrenergic receptor subfamily of G protein-coupled receptors.


INTRODUCTION

The thyrotropin-releasing hormone receptor (TRH-R)1 (1) is a member of the rhodopsin/beta -adrenergic receptor subfamily of GPCRs (2). It is thought that all GPCRs have in common a transmembrane bundle that is composed of seven helices. It has been proposed that residues that are highly conserved in these helices are important in maintaining the structure of the bundle and thereby producing a receptor that exhibits high affinity and activity (3, 4).

Bacteriorhodopsin is the only protein that contains seven transmembrane-spanning helices (TMs) for which a crystallographic analysis at an atomic level of resolution is available (5). However, bacteriorhodopsin is not a GPCR and does not share sequence homology with GPCRs. Baldwin (3) predicted a structure for GPCRs of the beta -adrenergic/rhodopsin family in which the TMs are arranged in a counterclockwise bundle (as viewed from the extracellular surface) based on sequence homology, conservation, and polarity of residues. A recently reported two-dimensional projection map of rhodopsin is consistent with this structure (6). The resolution of the projection map was not sufficiently high, however, to allow assignment of specific helical positions and could not identify residues that form interhelical interactions. Indirect evidence, therefore, has been used to suggest proximity of TMs 2, 3, and 7 in rhodopsin (7), of TMs 1, 2, 3, and 6 to TM-7 in adrenergic receptors (8, 9), of TMs 1 and 7 (10) in muscarinic receptors, and of TMs 5 and 6 of the neurokinin NK-1 receptor (11).

Proximity of two of the three highly conserved residues, Asn in TM-1, Asp in TM-2, and Asn in TM-7, has been suggested for several GPCRs by demonstrations that single substitutions of Asp by Asn or of Asn by Asp lead to diminished receptor function and double substitutions restore function. For example, an interaction has been suggested between Asp of TM-2 and Asn of TM-7 in the serotonin 5HT2A receptor (12). Interestingly, the positions of Asp and Asn residues are switched in the GnRH receptor, which contains an Asn in TM-2 and an Asp in TM-7, and these residues were suggested to interact (13). Recently, it was suggested that Asn in TM-1, Asp in TM-2, and Asn (and a conserved Tyr) in TM-7 are near each other and form a polar pocket within the alpha 1B-adrenergic receptor (14).

We have constructed a model of the complex of TRH and TRH-R (15, 16) based on a GPCR template that is different from that of bacteriorhodopsin. The model of the TRH·TRH-R complex predicts proximity of Asn 43 in TM-1, Asp 71 in TM-2, and Asn 316 in TM-7 of TRH-R. The putative binding pocket for TRH is between TMs 3, 6, and 7 and is not close to the region occupied by Asn-43, Asp-71, and Asn-316. In this study, we have examined the roles of Asn-43, Asp-71, and Asn-316 of TRH-R. We present experimental and computational evidence that Asp-71 interacts with Asn-43 and Asn-316. Our data indicate that these interactions are critical for activation of TRH-R and provide support for the proposed TM bundle topology of GPCRs.


EXPERIMENTAL PROCEDURES

Materials

[3H]MeTRH was obtained from DuPont. myo-[3H]Inositol was obtained from Amersham Corp. TRH was from Calbiochem and MeTRH from Sigma. Restriction endonucleases were from New England Biolabs. The cloning vector pBluescript was from Stratagene and the expression vector pCDM8 from Invitrogen. Dulbecco's modified Eagle's medium and fetal calf serum were from Collaborative Research.

Mutagenesis

The full-length, mouse TRH-R cDNA in pBluescript (pBSmTRHR) (1) or pCDM8 (pCDM8mTRHR) (17) was used for mutation. Mutants were prepared by the polymerase chain reaction and were subcloned directly into pCDM8mTRHR (N43A, N43D, and D71N) or were subcloned into pBSmTRHR (N316A and N316D) and then subcloned into pCDM8mTRHR after digesting with XhoI and NotI. Double mutants were constructed in pCDM8. Construction of D71A TRH-R was described previously (18). Mutant TRH-R sequences were confirmed by the dideoxy chain termination method.

Cell Culture and Transfection

COS-1 cells were maintained and transfected as described previously (1). In brief, cells were seeded 1 or 2 days prior to transfection at 0.7 to 1.5 × 106 cells/100-mm dish. Cells were transfected using the DEAE-dextran method as described and maintained in Dulbecco's modified Eagle's medium with 10% fetal calf serum for 1 day at which time cells were harvested and seeded into 12-well plates at 100,000 cells/well in Dulbecco's modified Eagle's medium with 5% fetal calf serum.

Receptor Binding Studies

One day after reseeding into 12-well plates, binding experiments were carried out in buffer with cells in monolayer for 1 h at 37 °C as described elsewhere (18). Inhibitory constants (Ki) were derived from competition binding experiments for which curves were fitted by nonlinear regression analysis and drawn with the PRISM program (GraphPad Inc.).

Inositol Phosphate Formation

One day after transfection, cells in monolayer in 12-well plates were labeled with 1 µCi of myo-[3H]inositol/ml. Stimulation of IP formation was measured 1 day later for 1 h at 37 °C by methods previously described (19). None of the TRH-Rs studied stimulated IP formation in the absence of TRH. Maximal activity of a mutant TRH-R was defined as the ratio of IPs formed in response to maximally effective concentrations of TRH in cells expressing a mutant TRH-R compared with WT. The maximal activity of WT TRH-R was set at 100%. Maximal activity was assessed only for those mutant receptors which were shown to be expressed at levels that for WT TRH-Rs were sufficient for maximal activity.

Computer Modeling

We have previously constructed models of the TRH·TRH-R complex using energy minimization techniques (15) and novel mixed mode Monte Carlo/Stochastic Dynamics simulations (16). The first approach produced an initial model of the binding pocket that was later refined through the Monte Carlo/Stochastic Dynamics simulations. These models agree with experimental findings. We have now constructed a model of TRH-R that includes the ECLs. The ECLs were added to a previously constructed model of the TM domain of TRH-R (16). The disulfide bridge between Cys residues in ECL-1 and ECL-2, which has been shown to be necessary to constrain TRH-R in a high affinity conformation (20), was maintained throughout all simulations. The model with frozen helices was energy-minimized with the program CHARMM (21) and used to generate 14 energy-minimized average structures through a simulated annealing protocol (22). The details of these simulations will be presented in a separate report.2 In brief, the minimized structure was heated to 1500 K, and 14 initial structures were extracted from a trajectory at this temperature. Each of these structures was annealed to 300 K over 60 ps followed by 100 ps constant temperature simulation at 300 K. The resulting minimized average structures clustered according to their pairwise root-mean square deviation (rmsd), producing a major family of seven members at rmsd of 1.9 Å. One of these structures was retained and used to conduct a simulation of the receptor-TRH complex. TRH was manually placed in the binding pocket as described previously (16). All helical constraints were removed, and the complex was energy-minimized, heated to 300 K in 23 ps, and subjected to 1 ns of molecular dynamics simulation at 300 K. An average structure calculated from the last 600 ps of the trajectory was energy-minimized in preparation for the construction of the structures described below. To test the relationship among Asn-43, Asp-71, and Asn 316, NMR nuclear Overhauser effect distance constraints as encoded in CHARMM (Kmax = 20 kJ/mol/Å2, Kmin = 40 kJ/mol/Å2, rmin = 1.5 Å, rmax = 2.0 Å) were applied during a 2000-step minimization to induce the formation of hydrogen bonds between the three residues. Subsequently, the constraints were removed, and the structure was further minimized for 3000 steps. Only the Asn-43-Asp-71 and Asp-71-Asn-316 hydrogen bonds remained in the unconstrained structure. Several minimizations were attempted with other constraints which were later removed, to explore alternative hydrogen bonding among the three residues. The only stable hydrogen bonding pattern was Asn-43-Asp-71 and Asp-71-Asn-316. This structure was subsequently heated to 300 K and subjected to 200 ps of molecular dynamics simulation at 300 K. It was also used in preparing the four mutant receptors, i.e. N43A, D71A, N316A, and N43A/N316A TRH-Rs. Each mutant was constructed by substituting Ala for Asn or Asp and energy-minimized for 2000 steps. Subsequently, each structure was heated to 300 K and subjected to 200 ps of molecular dynamics simulation at 300 K. The energy minimized structures averaged over 100-140 ps of the corresponding simulations are presented here. In all calculations, a distant dependent dielectric function was used to approximate the effect of the environment.


RESULTS

Our revised model of the TRH·WT TRH-R complex indicates proximity of conserved residues Asn-43 in TM-1, Asp-71 in TM-2, and Asn-316 in TM-7. This model does not differ significantly in the TM bundle binding pocket from that described in our previous publication (16) but includes a newly constructed domain consisting of the three ECLs.3 A closeup of domains including TM-1, TM-2, and TM-7 (Fig. 1) indicates that Asp-71 forms hydrogen bonds to both Asn-43 and Asn-316. On the other hand, no interaction between Asn-43 and Asn-316 could be maintained in the simulations, suggesting that these residues are involved in maintaining receptor structure through their common bridge to Asp-71. To begin to assess the roles of Asn-43, Asp-71, and Asn-316, substitutions were made with Asp, Asn, and Asp, respectively. The results of binding and activation studies of WT and mutant TRH-Rs transiently expressed in COS-1 cells are shown in Table I. The affinities of N43D and D71N were 8- and 57-fold lower than WT TRH-R, respectively, and the affinity of N316D TRH-R was 6-fold higher than WT TRH-R. These changes in affinities are small compared with substitution of residues that we have shown directly contact TRH. These data are consistent with the model which predicts that these residues do not directly contact TRH and that the changes in affinities are secondary to conformational changes in the TM bundle (see below). N43D, D71N, and N316D TRH-Rs exhibited 57, 52, and 98% of WT TRH-R maximal activity, respectively. Thus, N43D and D71N TRH-Rs exhibited lower maximal activities than WT TRH-R, whereas N316D was as active as WT TRH-R.


Fig. 1. Model of the interactions between Asn-43, Asp-71, and Asn-316 in the wild type occupied TRH receptor. The structure presented is the extract from an averaged minimized structure derived from the last 100 ps of a 200-ps simulation described under "Experimental Procedures." The view is from the extracellular surface; extracellular loops have been deleted for clarity.
[View Larger Version of this Image (44K GIF file)]

Table I. Effects of substitution of Asn-43 by Asp, Asp-71 by Asn, and Asn-316 by Asp in single and double mutants


TRH-R Kda Bmax/well EC50b Maximal activity

nM dpm × 10-3 nM % WT
WT 2.8 20 0.59 100
(2.1-3.6) (±1.8) (0.51-0.69)
(n = 21) (n = 41) (n = 19)
N43D 21 4.5 19 57
(18-24) (±0.65) (17-22) (±3.5)
(n = 6) (n = 25) (n = 8) (n = 19)
D71N 160 47 540 52
(140-180) (±8.0) (430-680) (±5.8)
(n = 8) (n = 14) (n = 2) (n = 10)
N316D 0.46 5.7 0.87 98
(0.1-1.0) (±0.79) (0.66-1.2) (±6.1)
(n = 5) (n = 10) (n = 3) (n = 6)
D71N/N316D 330 10 250 115
(250-420) (±1.6) (200-310) (±13)
(n = 5) (n = 11) (n = 3) (n = 8)
N43D/N316D NSBc 91  ---d
(71-120)
(n = 7)
N43D/D71N 300 22  ---e 7.6
(250-390) (±2.4) (±1.2)
(n = 10) (n = 17) (n = 16)

a Kd for MeTRH.
b EC50 for TRH.
c No specific binding measurable.
d Activity at maximally effective concentrations of TRH (up to 1 mM) was measured as 15% but Bmax could not be determined.
e Maximal activity too low to assess EC50.

To further analyze the roles of these conserved residues, mutant receptors containing two substitutions of the three pairs of residues were constructed (Table I). The affinities of N43D/D71N and D71N/N316D TRH-Rs were 110- and 120-fold lower, respectively, than WT TRH-R, and no binding was detectable with N43D/N316D TRH-R. The maximal activities of N43D/D71N and D71N/N316D TRH-Rs were 7.6 and 115% that of WT TRH-R, respectively. The maximal activity of N43D/N316D could not be determined because the Bmax could not be measured. Thus, the mutant receptor containing Asn at position 71 was made to revert to WT TRH-R levels of activity by substituting Asp at position 316. This suggests that positions 71 and 316 are proximate and that an interaction between Asp-71 and Asn-316 occurs in WT TRH-R that is important for activation. We cannot determine whether the loss in activity of the double mutant N43D/D71N TRH-R was additive, which would suggest that these two residues were functionally independent, or more than additive, which would be consistent with these residues being functionally interdependent, compared with the two receptors with single mutations. Therefore, it is not clear from these data whether Asn-43 and Asp-71 are proximate (see below). As we could not determine the activity of N43D/N316D TRH-R, we could not assess whether Asn-43 and Asn-316 are functionally interdependent from these data also (see below).

To analyze our model further, alanine residues, which would minimize the possibility of introducing non-native interactions, were substituted at Asn-43, Asp-71, and Asn-316 (Table II). The affinity of N43A TRH-R was similar to WT TRH-R, and the affinities of D71A and N316A TRH-Rs were 3- and 11-fold lower, respectively, than WT TRH-R. The maximal activities of N43A and N316A TRH-Rs were 37 and 47% of WT TRH-R, respectively; D71A TRH-R was inactive. These data indicate that Asn-43, Asp-71, and Asn-316 are important for activation of TRH-R. The larger effect noted with D71A TRH-R compared with either N43A or N316A TRH-Rs is consistent with the idea that Asp-71 interacts with both Asn-43 and Asn-316. To further test this prediction, the double mutant N43A/N316A TRH-R was constructed. The affinity of N43A/N316A TRH-R was 22-fold lower than WT. More importantly, N43A/N316A TRH-R was inactive. Thus, both D71A and N43A/N316A TRH-Rs are inactive receptors. These data are consistent with our model in which Asp-71 interacts with both Asn-43 and Asn-316.

Table II. Effects of substitution of Asn 43, Asp 71 and Asn 316 by Ala


TRH-R Kda Bmax/well EC50b Maximal activity

nM dpm × 10-3 nM % WT
WT 2.8 20 0.59 100
(2.1-3.6) (±1.8) (0.51-0.69)
(n = 21) (n = 41) (n = 19)
N43A 3.1 2.8 3.2 37
(2.0-4.4) (±0.37) (2.2-4.6) (±4.3)
(n = 5) (n = 10) (n = 4) (n = 14)
D71A 8.6c 9.2 NAc,d
(±1.5)
(n = 3)
N316A 32 23 140 47
(26-38) (±3.0) (100-180) (±3.1)
(n = 3) (n = 13) (n = 3) (n = 15)
N43A/N316A 61 8.6 NAd
(47-79) (±1.3)
(n = 8) (n = 8)

a Kd for MeTRH.
b EC50 for TRH.
c Data are from Perlman et al. (18).
d NA, no activity measurable with TRH up to 100,000 nM.

In our model of WT TRH-R, Asn-316 plays an important role in addition to hydrogen bonding to Asp-71. An averaged minimized structure from the simulation shows that the hydrogen bond to Asp-71 orients the amide group of Asn-316 to also form a hydrogen bond with its own backbone N-H group throughout 90% of the simulation. This introduces a major perturbation in the helicity of TM-7 as shown schematically in Fig. 2. The helicity in this area is weakened by the presence of Pro-317, which cannot form a hydrogen bond to Ser-313. The interaction of Asn-316 with its own backbone prevents the formation of a hydrogen bond to Asn-312, which is positioned above it. This local disruption of helical hydrogen bonds unwinds the helix between Asn-312 and Asn-316. Consequently, Ser-313 forms a bifurcated hydrogen bond to Ile-309 and Tyr-310, Ala-314 hydrogen bonds to Leu-311, leaving the carbonyls of Asn-312 and Ala-314 without hydrogen bonding partners. Hydrogen bonding patterns are also disrupted below Asn-316 because the helix is kinked and unwound. Consequently, Asn-321 hydrogen bonds to Asn-316 and Ile-319 to Ile-315 leaving Val-318 and Tyr-320 without hydrogen bonds. A similar disruption of helicity in TM-7 was proposed on the basis of the observation that an Asn-Pro motif (as in Asn-316-Pro-317) is a helix breaker (23).


Fig. 2. Perturbation in helicity of TM-7 of the TRH receptor. Left, hydrogen bonding network of residues 308 through 320 of TM-7. As described in the text, due to the interaction of Asp-71 with Asn-316, the side chain of Asn-316 forms a hydrogen bond to its own backbone N-H group and the side chain of Asn-312 forms a hydrogen bond to the backbone carbonyl of Cys-308. This local disruption in helical hydrogen bonding results in perturbation of the helix between Asn-312 and Asn-316. Right, a schematic description of the disruption in the helical hydrogen bond network in the form of a helical net. The arrows represent hydrogen bonds from the donor (N-H) to the acceptor (C=O). Filled and unfilled circles denote hydrogen bonded or non-hydrogen bonded C=O groups, respectively. Unfilled triangles denote non-hydrogen bonded N-H groups. Dotted lines denote backbone interactions, and solid lines denote interactions involving side chains.
[View Larger Version of this Image (106K GIF file)]

The effects of substituting Ala for Asp-71 or for Asn-43 and Asn-316 are illustrated in Fig. 3. Introducing an Ala substitution in the position of Asn-316 no longer constrains TM-7 by the hydrogen bond to Asp-71, and its helicity is restored. This occurs both in the N316A (not shown) and the N43A/N316A mutants. Similarly, a substitution of Asp-71 to Ala results in the loss of a hydrogen bond to Asn-316. Unconstrained, Asn-316 turns to form a hydrogen bond with the backbone carbonyl of Ser-313, which is left available due to the imide nitrogen of Pro-317. Finally, the constraining effect of the hydrogen bonding network can be noticed in the N43A/N316A double mutant receptor in which Asp-71, no longer constrained by these residues, turns to form a hydrogen bond with the backbone N-H of Met-51 in TM-1.


Fig. 3. A comparison of the occupied wild type and D71A TRH receptors (A and B) and the occupied D71A and N43A/N316A TRH receptors (C and D). To simplify the view, portions of the receptor are shown: TM-1 through TM-4 in panels A and C, TM-1 and TM-5 through TM-7 in panels B and D. All structures were obtained from an averaged minimized structure as described under "Experimental Procedures." The two receptors in each panel are positioned so that the extracellular portions of the receptors are aligned (that is, Leu-30 to Ile-40; Leu-75 to Thr-84; Cys-100 to Ile-109; Ala-151 to Trp-160; Pro-190 to Phe-199; Trp-279 to Val-288; Asn-299 to Asn-312). Residues 43, 71, and 316 are colored blue. The view is from the side of the receptor with the intracellular portion of the receptor at the top of the panels. Significant structural changes in the backbone of D71A TRH-R versus WT TRH-R are observed for the intracellular portions of TM-1 (A), TM-5 (B) and TM-6 (B). The helicity of TM-7 in the vicinity of Asn-316 is nearly recovered in D71A TRH-R (B). Aside from a slight deflection of the intracellular portion of TM-6 (D), the backbones of the mutant receptors, D71A and N43A/N316A TRH-Rs, show very similar conformations (C and D).
[View Larger Version of this Image (124K GIF file)]

Disruption of hydrogen bonds present in WT TRH-R has major effects on the structure of the helical bundle. As shown in Fig. 3, A and B, mutating Asp-71 to Ala has a major effect on the intracellular portion of the helices. A comparison to WT receptor shows major disturbances in TM-1 (Fig. 3A) and in TM-5 and TM-6 (Fig. 3B). Despite major changes in the helicity of TM-7 (see above), its position in the helical bundle does not change much compared with WT. The extent of the difference between the helical bundles can be expressed by the rmsd between Calpha atoms of WT and D71A TRH-Rs of 1.78 Å. In contrast, the conformations of the backbones of the mutant D71A TRH-R and the double mutant N43A/N316A TRH-R are similar. As can be seen in Fig. 3, C and D, there is excellent overlap of D71A and N43A/N316A TRH-Rs and the rmsd between the Calpha atoms of the helices is only 1.01 Å. Hence, the simulations establish a proposed link between the disturbance of the helical bundle and the similar nonresponding phenotype of the D71A and the N43A/N316A TRH-Rs.


DISCUSSION

The major experimental finding reported herein is that the effect of substituting Asn-43 in TM-1 and Asn-316 in TM-7 with Ala in a doubly mutated TRH-R leads to the same complete loss of activity as the previously reported effect of substituting Asp-71 in TM-2 with Ala (18). These data are consistent with our model of the TRH·TRH-R complex which predicts that Asp-71 interacts with both Asn-43 and Asn-316 forming a bridge between them. Moreover, the modeling shows that the D71A and N43A/N316A TRH-Rs exhibit disruption of hydrogen bonds usually present in the native receptor that results in similar changes in the intracellular aspects of the helices of these mutant receptors. Thus, we conclude that a structural feature of TRH-R that is important for activation involves the bridging of these specific residues in TMs 1, 2, and 7.

It has been proposed that four conserved residues in TMs 1, 2, and 7 of the alpha 1B-adrenergic receptor, which includes the residues homologous to Asn-43, Asp-71, and Asn-316 of TRH-R, form a polar pocket that interacts with an Arg residue of the conserved Asp-Arg-Tyr (DRY) motif at the junction between TM-3 and intracellular loop 2 to constrain the receptor in an inactive conformation (14). This does not appear to be the function of these residues in TRH-R. Several differences between TRH-R and alpha 1B-adrenergic receptor regarding activity have been found when similar experiments were performed with homologous mutants in the two systems. 1) We have not been able to demonstrate basal activity, that is, IP formation in the absence of agonist, for N43A TRH-R,4 as was shown for the corresponding N63A alpha 1B-adrenergic receptor. It may be that a more sensitive system is needed to demonstrate constitutive activity of TRH-Rs. Alternatively, it is possible that similar structural changes induced by the mutation of the Asn in TM-1 may result in different phenotypes of the mutant receptors. 2) N43A TRH-R is only approximately half as active as WT TRH-R in the presence of maximally effective concentrations of TRH, whereas N63A alpha 1B-adrenergic receptor was reported to be as active as WT alpha 1B-adrenergic receptor. Thus, the role(s) of Asn in TM-1 in basal and agonist-induced activity is different in TRH-R and alpha 1B-adrenergic receptor. 3) R123A TRH-R retains substantial activity in response to TRH,4 whereas the mutant receptor with the corresponding substitution of Arg within the Asp-Arg-Tyr motif with Ala in alpha 1B-adrenergic receptor is inactive. Thus, the role of this Arg is at least quantitatively different in agonist activation of TRH-R and of the alpha 1B-adrenergic receptor. It appears, therefore, that there are differences in the roles of highly conserved residues in different members of the rhodopsin/beta -adrenergic receptor subfamily of GPCRs (see below).

Substitutions by alanines in mutated proteins may be the most reliable indicators of the roles of substituted native residues. Although polar substitutions for polar residues have the potential for restoring native interactions, they also may introduce strong non-native hydrogen bonding or ionic interactions. Nevertheless, the effects on activation observed with mutant TRH-Rs in which polar residues were substituted for the conserved residues support the findings observed with Ala substitutions. For example, D71N TRH-R exhibited activity that was decreased to 52% of WT TRH-R. However, when Asn-316 was mutated to Asp, the effect of mutating Asp-71 to Asn in the doubly mutated receptor, D71N/N316D TRH-R, was reversed. That is, the maximal activity of D71N/N316D TRH-R was similar to that of WT TRH-R. This restorative effect is consistent with proximity or interaction of the substituted residues (24-26). By switching residues in this way, a similar conclusion was reached with regard to these residues in the 5HT2A receptor (12). The native GnRH receptor contains Asn at the position homologous to Asp-71 of TRH-R and Asp at the position homologous to Asn-316 of TRH-R. A double mutant GnRH receptor in which these residues were switched resulted in restoration of binding though not of activation (13). Thus, these highly conserved Asn residues in TMs 1 and 7 and Asp in TM-2 are important for agonist-stimulated signaling. These data also highlight the fact that similar substitutions in different receptors may give rise to different phenotypes. Such a behavior suggests that another factor, e.g. the conformation of the intracellular loops, may be important in the appearance of a certain phenotype.

The double mutant N43D/N316D TRH-R, at a maximally effective concentration of TRH, exhibited only 15% of the WT TRH-R maximal activity. However, specific binding by this double mutant could not be measured (Table I) and, therefore, we could not ascribe the lowered IP formation to an intrinsic property of the mutant receptor. In fact, as our inability to measure specific binding was likely due to markedly lowered cell surface expression, the apparent decrease in stimulation of IPs was likely due to low receptor number. This is so because binding was detectable in several mutant receptors with lower potencies than N43D/N316D TRH-R (Tables I and II), and there is usually a good correlation between potency and affinity in TRH-R mutants (27). Although there may be several reasons for decreased expression of N43D/N316D TRH-R, it is reasonable to speculate, in conjunction with other findings described above, that it may be due to an abnormal receptor structure caused by a repulsive effect between the introduced Asp residues which would be consistent with proximity of the native residues in TRH-R.

Aside from supporting the idea that Asn-43, Asp-71, and Asn-316 are proximate and interact, the data allow conclusions to be drawn regarding the importance of the side chains of these residues in activation. Analysis of the single Ala mutants indicate that Asn-43, Asp-71, and Asn-316 are all important for activation. Asn can partially substitute for the function of Asp-71 and allow for partial activation suggesting that both oxygens of Asp-71 are important. Asp can fully substitute for Asn-316 with regard to activation which may be interpreted in at least two ways. The carbonyl group of Asn-316 (retained by Asp) may be important for activation (implying protonation of Asp-71) or the Asp introduced at position 316 may be protonated and become a potential hydrogen bond donor (like Asn). The latter mechanism has also been suggested to explain the activity of a similar mutant in the 5HT2A receptor (12).

As noted above, we have proposed that the putative binding pocket for TRH is between TMs 3, 6, and 7 in TRH-R which is not close to the region occupied by Asn-43, Asp-71, and Asn-316. Therefore, effects on affinity resulting from substitution of these residues are due to indirect effects on the binding pocket. This is consistent with the idea that conserved residues would not be involved in forming a high affinity, highly specific binding pocket. The binding affinities of N43A, D71A, and N316A TRH-Rs for MeTRH (an analog of TRH with 5-10-fold higher affinity than TRH that is used as radioligand) were little changed from WT TRH-R (Table II), which is consistent with the idea that these residues do not directly contact the ligand. In fact, the larger effects on binding of substituting Asp for Asn-43 or Asn for Asp-71 appear to be deleterious effects due to the introduced substituents rather than loss of the native side chains. Thus, these data are consistent with the idea that these conserved residues do not constitute the TRH-R binding pocket.

The model appears to support several interesting features of the experimental data. In WT TRH-R, the hydrogen bonding network obtained after exploring various possible patterns through the use of distance constraints, supports the idea that Asp-71 forms hydrogen bonds with both Asn-316 and Asn-43, while Asn-316 and Asn-43 do not interact with each other. In the D71A and N43A/N316A mutant receptors, however, the hydrogen bonds that maintain this network are lost. Comparison of the structures of WT and D71A TRH-Rs shows that the helical bundle undergoes a significant change which is localized to the intracellular portion of the helices. The similarity of these changes for the mutant receptors and their difference from WT TRH-R suggests that a possible mechanism for loss of activity may be related to the rearrangement of helices. These rearrangements will induce conformational changes in the intracellular loops which enable the interaction of the activated receptor with the G-protein. However, our simulations only provide part of the answer since the intracellular loops are not included. Further work will be required in constructing a model for the intracellular loops before this question can be addressed. Nevertheless, it appears that the phenotypic differences between similar mutations in different receptors (e.g. N43A in TRH-R versus N63A in alpha 1B-adrenergic receptor (14)) may result from different consequences in the conformations of the intracellular loops caused by similar changes in the TM portions.

The highly conserved Asp in TM-2 has been implicated in a number of functional roles in GPCRs including activation, binding and allosteric regulation by Na+ (28). The conserved Asn in TM-7, which is part of a highly conserved Asn-Pro-Xaa-Xaa-Tyr (NPXXY) motif, has been shown to be important for activation in the 5HT2A (12), angiotensin AII (29), and beta 2-adrenergic receptors (30). Interestingly, an extensive review of GPCR site-specific substitutions shows that residues in TM-1 have rarely been targeted (28) (see above).

In conclusion, our experimental observations are consistent with the idea that Asn-43 of TM-1, Asp-71 of TM-2, and Asn-316 of TM-7 are proximate to one another in the native TRH-R and with a bifunctional role of Asp-71 in interacting with both Asn-43 and Asn-316. Our computational data support this idea and predict that Asp-71 forms a bridge between Asn-43 and Asn-316 that may be important in holding TRH-R in a conformation in which TM 1, 2, and 7 are apposed. The conservation of these residues suggests that these interactions may be important for the structural integrity of all members of the rhodopsin/beta -adrenergic receptor GPCR subfamily.


FOOTNOTES

*   This work was supported by National Institutes of Health Physician Scientist Award DK 02101 (to J. H. P.) and Grant DK 43036 (to M. C. G. and R. O.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   To whom correspondence should be addressed: Cornell University Medical College, 1300 York Ave., Rm. A328, New York, NY 10021. Tel.: 212-746-6280; Fax: 212-746-6289.
1   The abbreviations used are TRH, thyrotropin-releasing hormone; TRH-R, thyrotropin-releasing hormone receptor; GPCR, guanine nucleotide-binding protein-coupled receptor; TM, transmembrane spanning helix; GnRH, gonadotropin-releasing hormone; WT, wild type; MeTRH, thyrotropin-releasing hormone in which N-tau -methylhistidine is substituted for histidine; ECL, extracellular loop; N316D TRH-R, for example, TRH-R in which Asn at position 316 is substituted by Asp; IP, inositol phosphate; rmsd, root-mean square deviation.
2   A.-O. Colson, J. H. Perlman, A. Smolyer, M. C. Gershengorn, and R. Osman, manuscript in preparation.
3   Atomic coordinates of this model can be obtained by contacting Roman Osman at the following E-mail address: osman{at}inka.mssm.edu.
4   J. H. Perlman, W. Wang, K. Bence, and M. C. Gershengorn, unpublished observations.

REFERENCES

  1. Straub, R. E., Frech, G. C., Joho, R. H., and Gershengorn, M. C. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 9514-9518 [Abstract]
  2. Dohlman, H. G., Thorner, J., Caron, M. G., and Lefkowitz, R. J. (1991) Annu. Rev. Biochem. 60, 653-688 [CrossRef][Medline] [Order article via Infotrieve]
  3. Baldwin, J. M. (1993) EMBO J. 12, 1693-1703 [Abstract]
  4. Baldwin, J. M. (1994) Curr. Opin. Cell Biol. 6, 180-190 [Medline] [Order article via Infotrieve]
  5. Henderson, R., Baldwin, J. M., Ceska, T. A., Zemlin, F., Beckmann, E., and Downing, K. H. (1990) J. Mol. Biol. 213, 899-929 [Medline] [Order article via Infotrieve]
  6. Schertler, G. F. X., and Hargrave, P. A. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 11578-11582 [Abstract]
  7. Rao, V. R., Cohen, G. B., and Oprian, D. D. (1994) Nature 367, 639-642 [CrossRef][Medline] [Order article via Infotrieve]
  8. Suryanarayana, S., Von Zastrow, M., and Kobilka, B. K. (1992) J. Biol. Chem. 267, 21991-21994 [Abstract/Free Full Text]
  9. Mizobe, T., Maze, M., Lam, V., Suryanarayana, S., and Kobilka, B. K. (1996) J. Biol. Chem. 271, 2387-2389 [Abstract/Free Full Text]
  10. Liu, J., Schöneberg, T., van Rhee, M., and Wess, J. (1995) J. Biol. Chem. 270, 19532-19539 [Abstract/Free Full Text]
  11. Gether, U., Nilsson, L., Lowe, J. A., III, and Schwartz, T. W. (1994) J. Biol. Chem. 269, 23959-23964 [Abstract/Free Full Text]
  12. Sealfon, S. C., Chi, L., Ebersole, B. J., Rodic, V., Zhang, D., Ballesteros, J. A., and Weinstein, H. (1995) J. Biol. Chem. 270, 16683-16688 [Abstract/Free Full Text]
  13. Zhou, W., Flanagan, C., Ballesteros, J. A., Konvicka, K., Davidson, J. S., Weinstein, H., Millar, R. P., and Sealfon, S. C. (1994) Mol. Pharmacol. 45, 165-170 [Abstract]
  14. Scheer, A., Fanelli, F., Costa, T., De Benedetti, P. G., and Cotecchia, S. (1996) EMBO J. 15, 3566-3578 [Abstract]
  15. Perlman, J. H., Laakkonen, L. J., Guarnieri, F., Osman, R., and Gershengorn, M. C. (1996) Biochemistry 35, 7643-7650 [CrossRef][Medline] [Order article via Infotrieve]
  16. Laakkonen, L., Guarnieri, F., Perlman, J. H., Gershengorn, M. C., and Osman, R. (1996) Biochemistry 35, 7651-7663 [CrossRef][Medline] [Order article via Infotrieve]
  17. Gershengorn, M. C., and Thaw, C. N. (1991) Endocrinology 128, 1204-1206 [Abstract]
  18. Perlman, J. H., Nussenzveig, D. R., Osman, R., and Gershengorn, M. C. (1992) J. Biol. Chem. 267, 24413-24417 [Abstract/Free Full Text]
  19. Perlman, J. H., Laakkonen, L., Osman, R., and Gershengorn, M. C. (1994) J. Biol. Chem. 269, 23383-23386 [Abstract/Free Full Text]
  20. Perlman, J. H., Wang, W., Nussenzveig, D. R., and Gershengorn, M. C. (1995) J. Biol. Chem. 270, 24682-24685 [Abstract/Free Full Text]
  21. Brooks, B. R., Bruccoleri, R. E., Olafson, B. D., States, D. J., Swaminathan, S., and Karplus, M. (1983) J. Comp. Chem. 4, 187-217
  22. Kirkpatrick, S., Gelatt, C. D., Jr., and Vecchi, M. P. (1983) Science 220, 671-680
  23. Konvicka, K., Ballesteros, J., and Weinstein, H. (1995) Biophys. J. 62, A446 (Abstract)
  24. Wells, J. A. (1990) Biochemistry 29, 8509-8517 [Medline] [Order article via Infotrieve]
  25. LiCata, V. J., and Ackers, G. K. (1995) Biochemistry 34, 3133-3139 [Medline] [Order article via Infotrieve]
  26. Mildvan, A. S., Weber, D. J., and Kuliopulos, A. (1992) Arch. Biochem. Biophys. 294, 327-340 [Medline] [Order article via Infotrieve]
  27. Gershengorn, M. C., and Osman, R. (1996) Physiol. Rev. 76, 175-191 [Abstract/Free Full Text]
  28. Van Rhee, A. M., and Jacobson, K. A. (1996) Drug Dev. Res. 37, 1-38 [CrossRef]
  29. Hunyady, L., Bor, M., Baukal, A. J., Balla, T., and Catt, K. J. (1995) J. Biol. Chem. 270, 16602-16609 [Abstract/Free Full Text]
  30. Barak, L. S., Ménard, L., Ferguson, S. S. G., Colapietro, A. M., and Caron, M. G. (1995) Biochemistry 34, 15407-15414 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.