Analysis of the C-Terminal Tail of the Rat Thyrotropin- Releasing Hormone Receptor-1 in Interactions and Cointernalization with b-Arrestin 1-Green Fluorescent Protein

Coexpression of the rat thyrotropin releasing hormone receptor-1 with b-arrestin 1-green fluorescent protein (GFP) in human embryonic kidney 293 cells results in agonist-dependent translocation of the arrestin to the plasma membrane followed by its cointernalization with the receptor. Truncations of the receptor C-terminal tail from 93 to 50 amino acids did not alter this. Truncations to fewer than 47 amino acids prevented such interactions and inhibited but did not fully eliminate agonistinduced internalization of the receptor. Deletion and site-directed mutants of the C-terminal tail indicated that separate elimination of a potential casein kinase II phosphorylation site or clathrin/clathrin adapter motifs was insufficient to prevent either internalization of the receptor or its cointernalization with b-arrestin 1-GFP. Alteration of sites of acylation reduced internalization and prevented interactions with b-arrestin 1-GFP. Combinations of these mutants resulted in lack of interaction with b-arrestin 1-GFP and a 10-fold reduction in internalization of the receptor. Despite this, the receptor construct that lacked the three protein sequence motifs was fully functional. These studies map sites that contribute the interactions of the thyrotropin releasing hormone receptor-1 C-terminal tail required for effective contacts with b-arrestin 1-GFP and indicate key roles for these interactions in agonist-induced internalization of the receptor. Thyrotropin releasing hormone (TRH) is a hypothalamic tripeptide that mediates its function via a small group of G protein-coupled receptors (Gershengorn and Osman, 1996). In the rat, three distinct TRH receptors are derived from two genes. The long and short isoforms of the rat TRH receptor-1 (TRHR-1) derive from alternative splicing and are identical through most of their sequence, differing only in their Cterminal tails (de la Pena et al., 1992a,b; Sellar et al., 1993). The rat TRH receptor-2 is derived from a distinct gene and is only ;50% identical with the TRHR-1 sequences (Cao et al., 1998; Itadani et al., 1998). All of these receptors produce their major effects via activation of members of the Gq-family of heterotrimeric G proteins, resulting in stimulation of phosphoinositidase activity and the elevation of intracellular [Ca] (Aragay et al., 1992; Hsieh and Martin, 1992; Lee et al., 1995; O’Dowd et al., 2000). As with many GPCRs, exposure to agonist results in rapid internalization and subsequent recycling of the receptor. Using a C-terminally GFPtagged form of the long isoform of TRHR-1 stably expressed in HEK293 cells, Drmota et al. (1998) demonstrated that agonist-induced internalization of this receptor was substantially blunted in the presence of hyperosmolar sucrose and thus was likely to proceed via a clathrin-dependent mechanism. Such mechanisms frequently involve the interaction of the receptor with members of the arrestin family because the nonvisual arrestins can interact directly with clathrin (Krupnick et al., 1997); in many cases, the coexpression of various mutant forms of arrestins limits agonist-induced receptor internalization (Goodman et al., 1998). Recently, direct monitors of agonist-induced interactions between GPCRs and arrestins have been provided by studying rapid, agonist-induced translocation of arrestin-GFP constructs from the cytoplasm to the plasma membrane in cells coexpressing the construct and an appropriate GPCR (Barak Financial support for this work was provided by the Biotechnology and Biosciences Research Council. 1 Current address: Department of Physiology, Tufts University School of Medicine, Boston, Massachusetts. ABBREVIATIONS: TRH, thyrotropin-releasing hormone; TRHR-1, thyrotropin-releasing hormone receptor-1; GFP, green fluorescent protein; HEK, human embryonic kidney; GPCR, G protein-coupled receptor; VSV, vesicular stomatitis virus; PCR, polymerase chain reaction; HEK, human embryonic kidney cells; DMEM, Dulbecco’s minimum essential medium; RT, room temperature; PBSGG, PBS containing 0.1% goat serum and 0.2% gelatin; KRH, Krebs-Ringer-HEPES; FLIPR, fluorometric imaging plate reader; GnRH, gonadotropin-releasing hormone. 0026-895X/01/5902-375–385$3.00 MOLECULAR PHARMACOLOGY Vol. 59, No. 2 Copyright © 2001 The American Society for Pharmacology and Experimental Therapeutics 490/881213 Mol Pharmacol 59:375–385, 2001 Printed in U.S.A. 375 at A PE T Jornals on O cber 8, 2017 m oharm .aspeurnals.org D ow nladed from Fig. 1. C-terminal truncation of TRHR-1 prevents agonist-induced interaction with b-arrestin 1-GFP. A, VSV-tagged forms of full-length (a) and the DStop (b) V-Stop (c), V-Stop (d), I-Stop (e), T-Stop (f), S-Stop (g), K-Stop (h), and NStop (i) mutants of the TRHR-1 were transiently transfected into HEK293 cells stably expressing b-arrestin 1-GFP. Cells were exposed to TRH (1 mM, 60 min), permeabilized, and prepared for microscopy after treatment with an anti-VSV monoclonal antibody and an Alexa 594-labeled secondary antiserum. Images were taken, for all truncations, after merging of the red (antibody) and green (fluorescent protein) signals, with the yellow punctate pattern observed in a–e representing the overlapping distribution of cointernalized receptor and b-arrestin 1-GFP. No such pattern was observed in f–i. A number of cells in the fields did not express receptor and in these the distribution of b-arrestin 1-GFP remained cytoplasmic and even [see Groarke et al. (1999) and Milligan (1999) for further details]. B, resolved signals for b-arrestin 1-GFP (left column) and the receptor (center column) from cells treated with vehicle (top) or TRH (1 mM, 60 min) (bottom) and the merged signals (right column) are shown after expression of VStop (top) and the T-Stop mutant (bottom). 376 Groarke et al. at A PE T Jornals on O cber 8, 2017 m oharm .aspeurnals.org D ow nladed from et al., 1997; Vrecl et al., 1998; Zhang et al., 1998; Dery et al., 1999; Ferrari et al., 1999; Groarke et al., 1999; McConalogue et al., 1999; Yu and Hinkle, 1999; Zhang et al., 1999; for reviews, see Ferguson et al., 1998; Milligan, 1999). In many but not all cases, such translocation is followed by the cointernalization of the receptor and arrestin-GFP into intracellular vesicles (Zhang et al., 1999). Such cointernalization has previously been observed for the TRHR-1 and b-arrestin 1-GFP (Groarke et al., 1999). The C-terminal tail of GPCRs often plays a key role in agonist-induced internalization. Indeed, C-terminal truncations of a number of GPCRs, including the TRHR-1, are known to slow ligand-induced internalization (Nussenzveig et al., 1993; Yu and Hinkle, 1999; Drmota and Milligan, 2000). In the case of the gonadotropin releasing-hormone (GnRH) receptors from mammalian species, the absence of a C-terminal tail seems responsible for their very slow rates of agonist-induced internalization (Vrecl et al., 1998). The equivalent GPCR from catfish has a C-terminal tail and both this GPCR and a mammalian version with the tail of the rat TRHR-1 appended internalize rapidly in response to agonist and now display a sensitivity to b-arrestin dominant-negative mutants (Heding et al., 2000). Furthermore, swapping the C-terminal tails between GPCRs can be sufficient to determine whether agonist-induced b-arrestin translocation is followed by cointernalization with the GPCR (Oakley et al., 1999). Recent studies on the CXCR4 receptor have begun to identify key residues in the C-terminal tail of this receptor involved in agonist-induced, arrestindependent internalization (Orsini et al., 1999). Herein we explore the role of distinct protein motifs in the C-terminal tail of the TRHR-1 for both interactions with b-arrestin 1-GFP and for receptor internalization. Experimental Procedures