Inhibition of thyrotropin binding to receptor by synthetic human thyrotropin beta peptides.

In order to study the structure and function relationships of the thyrotropin (TSH)-specific beta-subunit, we produced 11 synthetic overlapping peptides containing the entire 112-amino acid sequence of human beta TSH and tested them for activity in TSH radioreceptor assay using both human and porcine thyroid membranes. Synthetic peptides representing four regions of the beta-subunit demonstrated the ability to inhibit binding of 125I-bovine TSH to crude thyroid membranes. The peptide representing the -COOH terminus of the subunit (beta 101-112) possessed highest binding activity, inhibiting binding of labeled TSH with an EC50 of 80 microM. The remaining active peptides were: beta 71-85 (104 microM), beta 31-45 (186 microM), beta 41-55 (242 microM), and beta 1-15 (331 microM). Specificity of the binding activity was shown by the inability of the peptides representing the remainder of the subunit to inhibit binding of label and by the inability of any of the peptides to inhibit binding of 125I-epidermal growth factor to the same thyroid membranes. The low affinity of the peptides as compared with native hormone is in agreement with previous studies of synthetic alpha-subunit peptides and, further, suggests that the interaction of beta TSH with receptor is multifaceted, requiring cooperative binding of these sites for the observed high affinity of the whole hormone. These studies are in agreement with previous predictions of active regions by chemical modification but add two regions to the list, showing the utility of the synthetic peptide strategy in the study of peptide hormone structure-activity relationships.

In order to study the structure and function relationships of the thyrotropin (TSH)-specific B-subunit, we produced 11  Specificity of the binding activity was shown by the inability of the peptides representing the remainder of the subunit to inhibit binding of label and by the inability of any of the peptides to inhibit binding of '261-epidermal growth factor to the same thyroid membranes.
The low affinity of the peptides as compared with native hormone is in agreement with previous studies of synthetic a-subunit peptides and, further, suggests that the interaction of @TSH with receptor is multifaceted, requiring cooperative binding of these sites for the observed high affinity of the whole hormone. These studies are in agreement with previous predictions of active regions by chemical modification but add two regions to the list, showing the utility of the synthetic peptide strategy in the study of peptide hormone structure-activity relationships.
Thyrotropin (TSH)' exerts its effects by interaction with specific surface receptors found on the plasma membrane of thyroid follicular cells. Although the amino acid sequence of TSH and the structurally similar glycoprotein hormones follitropin, lutropin (LH), and choriotropin have been known for several years (l), description of the three-dimensional structure of the hormones and the specific hormone regions responsible for receptor interaction and exertion of biologic activity remains limited.
The glycoprotein hormones are heterodimers consisting of a common a-subunit, the sequence of which is identical between the four hormones, and a P-subunit, the sequence of which is hormone-specific.
Previously, using a comprehensive synthetic peptide approach, we identified two hormone sites within the common a-subunit (~~26-46 and &l-92) that interact with TSH receptors on human and rat thyroid membranes (2) and with the LH-hCG receptor on rat ovarian membranes (3). These synthetic peptides proved to be inhibitors of TSH bioactivity, and the former ((~26-46) was shown to inhibit the bioactivity of TSH receptor autoantibodies (or thyroid-stimulating immunoglobulins (4) from patients with Graves' hyperthyroidism.
However, the affinity of these synthetic peptides for receptor was quite low (10-4-10-5 M) suggesting that multiple hormone sites were involved in receptor interaction.
Because the a-subunit is shared between all the glycoprotein hormones, the P-subunit must, in some as yet unknown fashion, confer hormone-receptor specificity (1,5). We herein describe studies in which we have extended the synthetic peptide approach to PTSH. The studies support the hypothesis of interaction of both cvTSH and PTSH with the TSH receptor. Further, they suggest that the interaction of this glycoprotein hormone with its receptor is multifaceted, with the overall affinity of hormone-receptor interaction resulting from the simultaneous effort of several hormone binding sites.

MATERIALS AND METHODS
Peptide Synthesis-Eleven synthetic peptides were assembled representing the entire 112-amino acid sequence of the @-subunit of human TSH. As shown in Fig. 1  Numbers represent the positions of the residues in intact human BTSH.
Underlined regions represent those residues that overlap with the adjacent peptide.
was separated from free by precipitation with polyethylene glycol and centrifugation as above.

RESULTS
Human Thyroid RRA- Fig.  3 and Table I show the results of the human thyroid radioreceptor assays. Four regions of the human PTSH subunit were represented by five active peptides. The highest activity was found in the peptide representing the -COOH terminus of the subunit @lOl-112 which inhibited ['251]iodo-bTSH binding with an EC5,, of 80 f 9 PM. Two peptides representing the region of the subunit between residues 71 and 95 were also active, @71-85 with an EC& of 104 + 15 PM and @81-95 with an EC& of 1196 + 139 PM. Because these two peptides overlap each other and because both show activity, the active region might be assumed to involve the region of overlap (residues 81-85) and, perhaps, extend somewhat toward the amino end of the region in order to explain the higher activity of the more proximal peptide /371-85.
Activity was also observed in the two peptides ,f331-45 and p41-55. Cysteine residues at 31 and 52 of PTSH correspond to those at 38 and 57 of PLH and hCG (1) (11) suggesting that this region may be less important for TSH binding than it is for the gonadotropins.
Finally, activity was noted in the peptide representing the amino terminus of PTSH, @l-15 which inhibited binding of the label with an EC& of 331 & 45 PM.
The peptides representing the remainder of the PTSH sequence possessed no ability to inhibit hormone-receptor interaction, indicating that the effect of the active peptides is sequence-specific.
In order to determine if the inhibition of binding of labeled TSH might be mediated by dissociation of the subunits of the hormone by the peptide, we performed preincubation experiments in which the active peptides were incubated for 1 h with labeled TSH prior to the addition of the membrane preparation.
No increase or enhancement of activity was noted in these experiments, suggesting that the peptides did not interact directly with the label. In additional experiments, labeled TSH was incubated with the two most active peptides (PlOl-112 and /371-85) under the same conditions as the RRA but without thyroid membranes.
The mixture was then examined by Sephadex G-100 (Pharmacia LKB Biotechnology Inc.) chromatography. The 1.6 X 50-cm column was preequilibrated with the RRA incubation buffer and calibrated with [?]iodo-bTSH and with ['251]iodo-ahCG. There was no appearance of free (Y-or P-subunit in these experiments, suggesting that no dissociation of labeled hormone occurred. Thus, it is quite unlikely that the binding inhibition seen was secondary to peptide-induced dissociation of subunits of the radiolabel.
Porcine Thyroid RRA-- Fig.  4 shows the activity of the synthetic /3TSH peptides in the TSH RRA using porcine thyroid membranes. Four regions of activity similar to those found in the human thyroid RRA were found. Again PlOl-112 and @71-85 were most potent showing EC&, values of 106 + 8.7 and 137 f 14 PM, respectively.
The aminoterminal peptide pl-15 was also active with an EC& of 1158 The peptide fill-25 showed differential activity in the two TSH RRAs. Although it was inactive in the human membrane assay, it showed significant inhibitory activity in the porcine  Fig. 3. Assay conditions were identical to those of the human assay except that 10 mg eq of porcine thyroid membranes were utilized as the receptor source.
assay with an ECso of 186 -C 28 pM. This may represent a region of PTSH that is recognized by the porcine TSH receptor but is not recognized by the human receptor. This may also, at least in part, explain the higher binding activity of unlabeled TSH in the porcine RRA uersus the human membrane assay (Table I).
As noted in Fig. 4, the remaining peptides had no activity, again demonstrating the sequence specificity of the effect. Epidermal Growth Factor RRA-Studies utilizing the EGF RRA were designed to determine the specificity of the synthetic ,f3TSH peptides for interaction with the TSH-specific receptor and to exclude a nonspecific membrane effect as the cause of the activity. As noted in Fig. 5 high affinity binding of EGF could be demonstrated to the porcine thyroid membrane preparations as has been previously noted by others (10,12). The EC& of unlabeled EGF in this assay (252 PM) is quite similar to that of unlabeled TSH in the TSH RRA using the same porcine membranes (114 PM). As shown in the figure, none of the active synthetic @TSH peptides significantly inhibited binding of ['251]iodo-EGF to its thyroid membrane receptor.

DISCUSSION
Current evidence for TSH (13)(14)(15) as well as LH-hCG (16) suggests that both the LY-and P-subunits interact with the receptor. However, direct studies of the binding activity of the individual subunits have been difficult to perform due to problems related to the apparent very low affinity binding of the isolated subunits and to the detection of this activity in preparations of uncertain purity (5). Thus, in order to study the interaction of BTSH with the TSH receptor we have extended the comprehensive synthetic peptide approach we used previously for the study of the glycoprotein a-subunit (2, 3) and, thereby, prepared and tested overlapping peptides representing the entire sequence of human /3TSH. Four regions of activity within PTSH are represented by synthetic peptides. The most potent activity was noted in the peptide representing the -COOH terminus of the subunit, /3101-112. Activity of this region of PTSH has been proposed previously by Cheng et al. (17) using enzymatic modification. Carboxypeptidase treatment of isolated PTSH resulted in reduced receptor binding activity of intact hormone when the modified P-subunit was recombined with cy. The authors suggested that the residues Ser"' and Tyruz were responsible for the activity. Our synthetic peptide data strongly support the importance of this region. Further, because the synthetic peptide /391-105 had no activity, the active residues would be B/B0 60 c (%' 60 1 \ ,I -IbTi-* j h..,,., ~_., /3TSH Peptides Inhibit TSH Binding expected to lie after residue 105, in agreement with the carboxypeptidase findings. The region between cysteine residues at positions 31 and 52 of BTSH corresponds to the region termed the "intercysteine loop," believed to be a looped region in the native hormone that is formed by disulfide bonding between the cysteines (11). The intercysteine loop region of PLH and PhCG (PLH38-57) has been shown, by a similar synthetic peptide approach, to possess binding activity in LH-hCG radioreceptor assay using ovarian membrane preparations (11). The synthetic peptide PLH38-57 also caused stimulation of testosterone production in the Leydig cell bioassay, further indicating its role in lutropin-receptor interaction. These data support our finding of activity in the peptides representing the intercysteine loop region of PTSH. However, the activity noted in the two peptides /331-45 and fi41-55 was significantly lower than that noted for the /3LH peptides in the previous studies, suggesting that the region, although still possessing binding activity for TSH, may not be as important in the overall activity of the intact hormone as it appears to be for LH-hCG.
Activity of the amino terminus of PTSH has not been previously reported. However, chemical modification studies have suggested activity for the amino terminus of PLH (1); thus activity in this region of PTSH should not be surprising. The amino terminus of BTSH has not proved susceptible to similar chemicai modification; thus the synthetic peptide approach has demonstrated activity in this region where more prototypic studies failed. The peptide /371-85 was second only to the -COOH-terminal peptide in its potency in the TSH RRA. Activity in this region has not been previously alleged for PTSH, as the region is relatively devoid of residues susceptible to chemical modification, and those residues which might be susceptible have not proven to be so in the whole hormone (1,18). This further confirms the utility of the synthetic peptide approach for study of ligand-receptor interaction. Several authors have noted higher TSH binding activity of porcine thyroid membranes as compared with human (10,19). The current studies have shown that the region 11-25 of the PTSH subunit has significantly greater activity when tested with porcine receptor than with human receptor, suggesting that the binding activity of this region may be responsible, at least in part, for the higher activity of the porcine receptor. In addition, reduced activity of @l-15 in the porcine assay suggests that the binding site of TSH in this region for the porcine receptor is shifted toward the -COOH end of 11-25 uer.sus the site recognized by the human receptor. The significance of these findings is increased when one notes the similarity of the activity of the remaining peptides between the two assays (Table I). Additional studies of this region are needed to characterize further which residues are responsible for this apparent species-dependent activity. As we previously noted with the synthetic a-subunit peptides (2,3), the affinity of interaction of the synthetic fisubunit peptides with the receptor is quite low when compared with that of the native hormone (10-5-10-4 M versus 10-i' M). If one assumes that four regions of the P-subunit and two regions of the a-subunit participate in receptor binding, the affinity of each interaction must be rather low; otherwise the overall affinity of the hormone for receptor would greatly exceed that observed. Some negative cooperativity must therefore occur if all these sites interact with the receptor simultaneously. Thus, the noted activity of the peptides suggests that the interaction of TSH with its receptor is multifaceted, not only involving each subunit but also multiple sites within each subunit, the final affinity thereby resulting from a simultaneous cooperative contribution of a combination of these sites.