Báo cáo khoa học: Internalization of the human CRF receptor 1 is independent of classical phosphorylation sites and of b-arrestin 1 recruitment potx

Internalization of the human CRF receptor 1 is independent of classical phosphorylation sites and of b-arrestin 1 recruitment Trine N. Rasmussen 1,3 , Ivana Novak 3 and Søren M. Nielsen 2 1 Department of Molecular Biology and 2 Department of Molecular Pharmacology, H. Lundbeck A/S, Valby, Denmark; 3 August Krogh Institute, University of Copenhagen, Denmark The corticotropin releasing factor receptor 1 (CRFR1) belongs to the superfamily of G-protein coupled receptors. Though CRF is involved in the aetiology of several stress- related disorders, including depression and anxiety, details of CRFR1 r egulation such a s internalization remain unchar- acterized. In the present study, agonist-induced internal- ization of CRFR1 in HEK293 cells was v isualized by confocal microcopy and quantified using t he radioligand 125 I-labelled sauvagine. Recruitment of b-arrestin 1 in re- sponse to receptor activation was demonstrated by confocal microscopy. The extent of 125 I-labelled sauvagine stimulated internalization was significantly impaired by sucrose, indi- cating the involvement of clathrin-coated pits. No effect on the extent of internalization was observed in the presence of the second messenger dependent kinase inhibitors H-89 and staurosporine, indicating that cAMP-dependent protein kinase and protein kinase C are not prerequisites for CRFR1 internalization. Surp risingly, deletion of all putative phos- phorylation sites in the C-terminal tail, as well as a cluster of putative phosporylation sites in the third intracellular loop, did not affect receptor internalization. However, these mutations almost abolished the recruitment of b-arrestin 1 following receptor activation. In conclusion, we demonstrate that CRFR1 internalization is independent of phosphory- lation sites in the C-terminal tail and third intracellular loop, and the degree of b-arrestin 1 recruitment. Keywords: b-arrestin 1; CRFR1; GPCR; receptor internal- ization. Corticotropin r eleasing factor (CRF) is a 41-residue neuro- peptide first isolated in 1981 [1] as the main stimulator of the release of adenocorticotropin from the pituitary. By activa- tion of CRF Receptor 1 (CRFR1) [2], CRF regulates not only the endocrine, but also the autonomic, behavioural and immune responses to stress [3]. Moreover, accumulating evidence indicates that CRF and its receptors play a prominent role in the aetiology o f several stre ss-related disorders, such as depression and anxiety [4]. CRFR1 belongs to family II of the G-protein coupled receptors (GPCRs). This family is composed of several distinct peptide receptors, such as secretin, parathyroid hormone and VPAC (vasoactive intestinal polypep tide and pituitary adenylate cyclase activating polypeptide receptors). Despite the impact of CRFR1 on the stress-response and stress-related dis- orders, little is known about its regulation. In general, the sensitivity of G PCRs to extracellular stimuli d epends upon various regulatory mechanisms including receptor desensi- tization, internalization and resensitization. Although recent studies on CRFR1 desensitization revealed the involvement of G-protein coupled receptor kinase 3 (GRK3) [5] and protein kinase C (PKC) [6], details of CRFR1 internalization remains uncharacterized. Internalization following agon ist activation is a common phenomenon amongGPCRs. Theprocess s erves a s the initial step of either receptor resensitization [7] or down-regulation by lysosomal degradation. Furthermore, re ports have been made about internalization-induced signal transduction [8]. According to s tudies mainly on the b-adren ergic receptors, belonging to t he family I GPCRs, t he process of agonist- induced GPCR internaliz ation is facilitated by the same proteins as those involved in receptor desensitization. Following activation, receptors are phosphorylated by specific GRKs at se rine/threonine residues in t he third intracellular loop and/or C-terminal tail. GRK-mediated receptor phosphorylation promotes the binding of cytosolic b-arrestin s, which not only uncouple receptors from their cognate G-proteins, but also target receptors for internal- ization through the subsequent interaction between the receptor/b-arrestin complex and proteins of the endocytic machinery such as clathrin [9] and the clathrin adaptor protein AP-2 [10]. Though these features can b e applied to several GPCRs [11],notable exceptions exist andmechanisms involved in the process of agonist-induced GP CR internal- ization is still a subject of controversy. For example, whereas the b 2 -adrenergic receptor is internalized via clathrin coated pits [12], the cholecystokinin receptor [13], the bradykinin receptor [14] and the b 1 -adrenergic receptors [15] have been Correspondence to S. M. Nielsen, Department of Molecular Pharma- cology, H. Lundbeck A/S, 9 Ottiliavej, DK-2500 Valby, Denmark. Fax: +45 3643 8253, Tel.: +45 3643 2096, E-mail: smn@lund beck.com Abbreviations: b-arr1, b-arrestin 1; CRF, corticotropin releasing factor; CRFR1, corticotropin releasing factor receptor 1; EGFP, enhanced green fluorescence protein; GPCR, G-protein coupled receptor; GRK, G-protein coupled receptor kinases; HEK, human embryonic kidney; PKA, cAMP-dependent protein kinase; PKC, protein kinase C. (Received 9 July 2004, revised 31 August 2004, accepted 20 September 2004) Eur. J. Biochem. 271, 4366–4374 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04371.x demonstrated to internalize, at least in part, by the caveolae- endocytic pathway. Likewise, studies on the secretin receptor indicate that GRK phosphorylation is not sufficient to facilitate or mediate receptor internalization [16], suggesting that kinases other than GRKs may play a greater role i n GPCR endocytosis than previously appreciated. These suggestions are supported by the recent finding that not only GRK, but also cAMP-dependent protein kinase (PKA) are involved in agonist-induced internalization of b 1 AR [15]. In the present study, we seek to reveal molecular and cellu- lar mechanisms responsible for agonist-induced internali- zation of CRFR1. By use o f confocal microscopy cellular distribution of C RFR1 and b-arrestin 1 are explored. R adio- ligand binding is applied to quantify receptor internalization and the effect of second-messenger kinases are explored using the PKA inhibitor H-89 and the broad spectrum kinase inhibitor staurosporine. Finally, a series of receptors with mutations in the third intracellular loop and C-termin al tail are constructed and examined for their ability to internalize and r ecruit b-arrestin 1. Interestingly, we find that agonist- induced internalization of CRFR1 does not depend on puta- tive phosphorylation sites in the third intracellular loop and C-terminal tail. M oreover, our results indicate that recruit- ment of b-arrestin 1 to the membrane following receptor activation is not a prerequisite for CRFR1 internalization . Experimental procedures cDNA constructs and mutagenesis The coding sequence for CRFR1 was inserted into the mammalian expression vector pCI (Promega) between EcoRI and XbaI restriction sites. The sequence encoding the EQKLISEEDL peptide from c-myc was inserted between residue 31 and 32 of the N-terminus. The insertion of the c-myc epitope at this position in the N-terminal region of mouse CRFR1 has previously been demonstrated not to alter t he binding or signalling properties [17]. An enhanced green fluorescence protein (EGFP)-conjugated CRFR1 construct (CRFR1–EGFP) was created deleting t he stop codon of CRFR1 and situating EGFP in frame directly after CRFR1. The fusion of EGFP t o the C-terminus of numerous other GPCRs reportedly does not alter receptor functionality [18,19]. Bovine b-arrestin 1 was a kind gift from T. Schwartz (University of Copenhagen, Den mark). An EGFP c onjugate (b-arr1–EGFP) was constructed by overlap PCR and the product inserted into pCI-neo (Promega). All mutant recep tors were generated by PCR following standard procedures [20]. CRFR1-stop384 (C-terminally truncated receptor after residue 384) was created using 3¢ antisense primers introducing a stop codon at the relevant position followed by an XbaI restriction site. CRFR1-IC3 was created by site-directed mutagenesis replacing the Ser, Thr, Thr, Ser, Glu, Thr residues at positions 301–306 with alanine residues. Likewise, substitution of serine at position 372 for alanine was achieved by site-directed mutagenesis. All sequences were confirmed by automatic sequencing. Cell culture and transfection Human embryonic k idney (HEK)293 cells (ATCC) were grown in Dulbecco’s modified Eagle’s medium (DMEM) with Glutamax I supplemented with 10% (v/v) heat inac- tivated fetal bovine serum, 5 m M sodium pyruvate and penicillin/streptomycin (100 lgÆmL )1 )at37°C in a humidi- fied incubator with 5% CO 2 . All products for cell culture were from Gibco . For transfection, cells were plated in 90-mm plastic dishes at a density of 3.0 · 10 6 cells per dish in medium without antibiotics for 18–24 h before use. Transient transfection was carried out using t he LipofectAMINE 2000 (Invitrogen Life Technologies, Carlsbad, CA) method according to the manufacturer’s instructions. For studies visualizing r eceptor localization, 4 lg plasmid containing cDNA encoding CRFR1–EGFP was use d. In coexpres- sion studies, 2 lg b-a rr1–EGFP was used in addition to 4 lg of the relevant CRFR1 construct. Twenty micro- grams of all cDNA constructs were used for radioligand binding and functional assays. Transfected cells were cultured for 48 h to allow expression. Twenty-four hours before experiments cells were seeded in appropriate dishes or plates precoated with 20 lgÆmL )1 poly D -lysine (Sigma). Drug treatment For studies including H-89, N-[2-((p-Bromocinnamyl)ami- no)ethyl]-5-isoquinolinesulfonamide, 2HCl (Calbiochem, VWR Denmark) and staurosporine (Sigma) drugs were added to the culture medium 20 min in advance in a concentration of 5 l M and 100 n M , respectively, and these concentrations were maintained throughout the experiment. Human/rat CRF (BACHEM, Germany) was used at a final concentration of 10 )8 M . Immunocytochemistry For colocalization studies of the CRF receptor with b-arrestin 1, cells expressing c-myc epitope-tagged CRFR1 and b-arr1–EGFP were g rown on poly D -lysine coated glass cove rslips (Menzel-glaser) in 35-mm dishes at a density of 5.0 · 10 5 cells per dish. After incubation with 10 )8 M CRF for various times, cells were fixed in 4% (v/v) paraformaldehyde for 10 min, washed t wice in NaCl/P i and permeabilized with 0.1% (v/v) Triton X-100 in blocking buffer (1% BSA/NaCl/P i ). Subsequently, cells were incubated for 45 min with monoclonal mouse c-myc antibody (clone 9E10, Sigma) diluted 1 : 1000 (5.3 lgÆmL )1 ) in blocking buffer. Following five washes in blocking buffer, Cy3 (Indocarbocyanine)-conjugated antibody against mouse (Jackson) 1 : 200 was applied for another 45 min before the cells were washed twice in blocking buffer, twice in NaCl/P i and mounted using Vectashield (Vector Laboratories, Burlingame, CA). Coverslips were sealed with nail polish. Fluorescence was detected using confo cal microscopy as described b elow. Unspecific binding was tested by excluding either the primary or secondary antibody. Detection of EGFP in living cells To visualize CRFR1–EGFP and b-arr1–EGFP in living cells, HEK293 transiently transfected with the relevant cDNA were plated on glass-bottomed culture dishes Ó FEBS 2004 Internalization of CRFR1 (Eur. J. Biochem. 271) 4367 (0.17-mm Delta T dish, Bioptechs Inc., Butler, PA) and k ept at 35 °C on a heated microscope stage during the experi- ment. C RF was added to the culture dish and images were collected at the indicated time points using confocal microscopy. Confocal microscopy Confocal microscopy was performed on a Biorad Radi- ance2000 confocal laser scanning microscope (CLSM) using a Nikon 60 · NA 1.4 oil immersion objective to examine immuno-stained coverslips or a Nikon 60 · NA1.2 water immersion objective to examine living cells. EGFP was excited with 488-nm Ar laser and the fluorescent signal was collected with an emission filter set comprising a 560-nm long-pass dichroic mirror and a 500– 530-nm barrier filter. For Cy3 detection the 543-nm Green HeNe laser was used along with a 555–625-nm emission filter. In addition to the m agnification provided by the objectives an additional zoom factor of 1.8–3.0 was applied. Images were collected in 512 · 512 pixels with a scan speed of 50 Hz. P inhole was set to achieve the optimal confocal sectioning, which is determined by the Airy disk diameter. However, sometimes the pinhole was opened in order to improve light collection of preparations with a weak fluorescence. To adjust detector gain and offset, a false- colour look-up table was applied. To improve the signal- to-noise ratio each image was an average of three or four scans. Images were subsequently processed using Adobe PHOTOSHOP 5.0. Lineprofile was performe d u sing the LASERPIX software (Bio-Rad). 125 I-labelled sauvagine internalization The day prior to t he experiment transfected cells were plated at a density of 200 000 cellsÆwe ll )1 in 24-well dishes (Nunc A/S Denmark). One hour prior to the assay, medium was changed to assay medium (DMEM supplem ented w ith 20 m M HEPES and 0.1% BSA). Internalization was initiated by incubating with rad iolabelled CRFR agonist 125 I-labelled s auvagine (PerkinElmer) diluted to 100 000 c.p.m. in 0.5 mL assay buffer for various times at 37 °C. Subsequently, cells were transferred to ice and washed twice in ice-cold NaCl/P i . To remove surface-bound radioligand, cells were washed with 1 mL of acid solution (50 m M acetic acid, pH 3) for 10 min. The acid supernatant, containing surface-bound radioactivity, was collected and measured. Subsequently, cells were solubilized in lysis buffer ( 0.2 M NaOH, 2% NP40) and internalized radioactivity was measured. Nonspecific binding, d etermined in the presence of 10 )7 M unlabelled CRF, was subtracted and the radio- activity internalized was expressed as a percentage of the sum of the surface radioactivity and the internalized radioactivity. In experiments w here the effect of hyp ertonic medium was tested, cells were pretreated with 0.4 M sucrose for 20 min and this concentration was maintained during radioligand incu bation. Data were analysed using Graph- Pad PRISM software. Results Visualization of agonist-mediated internalization of CRFR1 The C-terminally EGFP-conjugated version of the receptor, CRFR1–EGFP, was used in order to visualize the cellular localization and trafficking of CRFR1 following agonist exposure. Fluorescence was detected in transiently trans- fected HEK293 cells by confocal microscopy. I n unstimu- lated cells, the fluorescence was almost exclusively confined to and sharply outlining the contours of the plasma membrane, as shown in Fig. 1A1. Following CRF expo- sure, a time-dependent increase in the appearance of small fluorescent intracellular vesicles was observed (Fig. 1A2 and A3). After 20 min of incubation, an aggregate of fluores- cence started t o appear near the nucleus of each cell (A4) and after 40 min, these aggregates became even more distinct (A5). However, fluorescence was still detectable at Fig. 1. Agonist-induced redistribution of CRFR1–EGFP visualized by confocal microscopy. HEK293 cells transiently expressing CRFR1-EGFP were exposed to 10 )8 M CRF for 0 (A1), 5 (A2), 10 (A3), 20 (A4) and 40 min (A5). In unstimulated cells (A1), CRFR1–EGFP was evenly distributed and concentrated at the plasma membrane. Following agonist exposure, a time-dependent increase in intracellular CRFR1–EGFP was observed (A2–5). As a c ontrol, H EK293 c ells t ransiently e xpressing EGFP alone were e xposed to the same c onditions (B1–5). The images are representative of multip le indepe ndent stu dies. Bar ¼ 10 lm. 4368 T. N. Rasmussen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 the membrane, although in a more punctuate pattern than in unstimulated cells. As a control, HEK293 ce lls transiently expressing EGFP showed a relatively e ven distribution of fluorescence throughout the cell and this distribution remained unaltered during CRF exposure (Fig. 1B). Interestingly, the morphology of the CRFR1–EGFP- expressing cells changed during the experiment. The unstimulated cells appeared unaltered compared to un- transfected cells, whereas following stimulation the cells seemed to loose the attachment to the bottom of the dish and shrink (Fig. 1A1 vs. A5). This change in morphology was not observed for cells expressing only EGFP stimulated with CRF (Fig. 1B). Quantification of agonist-mediated CRFR1 internalization To quantify receptor internalization, we used the CRF analogue 125 I-labelled sauvagine for i nternalization studies. The use of radioligand internalization to assay receptor internalization is based on the assumption that receptor and ligand are endocytosed together and that the intracellular receptor–ligand complex can be determined by measuring intracellular radioactive labelled ligand. Following 125 I-labelled sauvagine stimulation, within 5 min more than 50% of the cell specific associated radioligand was internalized (Fig. 2). The internalization reached a plateau after 10–20 min, with a maximal radio- ligand internalization of 69%. Consistent with the translo- cation of CRFR1–EGFP observed by confocal microscopy, these results demonstrate that CRFR1 is internalized following agonist exposure and that internalization can take place w ithin minutes. Agonist-induced endocytosis of many G PCRs occurs via clathrin-coated pits, a process that can be inhibited by hypertonic sucrose [21]. Thus, in order to determine if CRFR1 was internalized via clathrin-coated pits, 125 I-labelled sauvagine incubation was performed in a medium containing 0.4 M sucrose. Under these circum- stances, almost no internalization of radioligand was observed during the initial 10 min and the maximal extent of internalization was reduced to 39% following 40 min of incubation (Fig. 2). This result indicate s the involvement of clathrin-coated pits in the agonist-induced internalization of the C RFR1. PKA and PKC inhibitors do not abolish agonist-induced CRFR1 internalization As CRFR1 activation is thought to signal through both the cAMP and the PLC pathway [22], the second messenger product of either or both of these pathways could be invo lved in agonist-stimulated internalization of this receptor. The effect of PKA and PKC was investigated by use of the specific PKA inhibitor H-89 and the broad-spectrum protein kinase inhibitor stauroporine. Radioligand inter- nalization assay was performed as described, including H-89 or staurosporine in the culture medium before and during 125 I-labelled sauvagine incubation. As shown in Fig. 3, no significant alteration of the extent of receptor-mediated internalization o f 125 I-labelled sauvagine could be observed in the presence of any of these inhibitors. These findings indicate that neither the activation of PKA nor the activation of PKC is necessary for agonist-induced CRFR1 internalization. Involvement of putative phosphorylation sites in agonist-induced receptor internalization To explore further the impact of various serine/threonine residues on r eceptor internalization, we constructed a series of CRFR1 with mutations in intracellular loop 3 and the C-terminal tail. cAMP measurements (data not shown) indicate that these mutations do not significantly a lter CRF- induced cAMP production. The various constructs are Fig. 2. Quantification of agonist-induced CRFR1 internalization. Transiently transfected HEK293 cells were incubated with 125 I-labelled sauvagine for various times at 37 °C. Cell surface-bound and cytosolic radioligand was determined and the specific i nternalization w as determined as described in Experimental p rocedures. To examine the effect of hypertonic media, 0.4 M sucrose was added before and during incubation. (A) Internalization of CRFR1 in the absence (j)andin the presence of sucrose (m) is shown. (B) The proportion of 125 I- labelled sauvagine internalized after 40 m in was 69% (± 2%), but only 39% ( ± 3%) in the presence o f s ucrose (Stu de nts t-test: ***P < 0.005). Data represent the mean (± SEM) of four se parate experiments performed in triplicate. Fig. 3. The effect of the PKA inhibitor H-89 and the broad spectrum kinase inhibitor staurosporine on radioligand internalization. Transiently transfected HEK293 cells were incubated with 125 I-labelled sauvagine at 37 °C for 40 min without or in the presence of 5 l M H-89 or 100 n M staurosporine. The proportion of 125 I-labelled sauvagine internalized after 40 min was 69% (± 2%), 66% (± 1%) in the presence of H-89 and 66% (± 2%) in the presence of staurosporine. The extent of receptor internalization without inhibitors was set to 100%. Data represent the mean (± SE M) of four independent experiments per- formed in triplicate. Ó FEBS 2004 Internalization of CRFR1 (Eur. J. Biochem. 271) 4369 depicted schematically in Fig. 4A. As shown in Fig. 4B, truncation of the C-terminal tail at position 384 (CRFR1- stop384), where all but one of the putative phosphorylation sites in the C-terminal tail was removed, did not cause a reduction in radioligand-induced internalization. Neither did an additional mutation o f t he remaining serine in close proximity to the expected seventh transmembrane domain (CRFR1-stop384; S372A). Furthermore, a construct with substitution of a cluster of serines and threonines to alanines in the third intracellular loop (CRFR1-IC3) retained its ability to internalize to the same extent as the wild-type CRFR1. Likewise a combination of this modification and CRFR1-stop384 (CRFR1-stop384; IC3) did not affect internalization. Taken together, these data indicate the ability of CRFR1 to internalize in the absence of the majority of putative phosphorylation sites in the C-terminal tail and third intracellular l oop. b-arrestin 1 is recruited to the plasma membrane following CRFR1 activation To examine the cellular trafficking of b-arrestin 1 following receptor activation, an EGFP conjugate of b-arrestin 1 (b- arr1–EGFP) was used. HEK293 cells were transiently cotransfected with b-arr1–EGFP and CRFR1 and cellular localization of b-arr1–EGFP was visualized by confocal microscopy. In unstimulated cells, b-arr1–EGFP fluoresc- ence was evenly distributed throughout the cytoplasm with no apparent enhanced plasma membrane localization (Fig. 5A1). However, within minutes following receptor stimulation b-arr1–EGFP was rapidly redistributed to the plasma membrane and after 5 min the cytosol was almost depleted (Fig. 5A3). No significant b-arr1–EGFP translo- cation was observed in cells lacking overexpressed CRFR1 (Fig. 5B). This demonstrates that b-arr1–EGFP is recruited specifically to th e plasma membrane in response to CRFR1 activation. b-arr1–EGFP was never observed associated with intracellular vesicles following receptor activation. This was readily demonstrated in a colocalization study of b-arr1–EGFP with c-myc-tagged CRFR1 (Fig. 6). In unstimulated cells, CRFR1 immunofluoresence (red) was confined to the plasma membrane, whereas the b-arr1– EGFP fluorescence (green) w as distributed in the cytoplasm (Fig. 6A1). This is illustrated in the overlay a nd quantified in the line profile by the peaks of CRFR1–Cy3, where the line crosses the plasma membrane and the even distribution of b-arr1–EGFP fluorescence throughout the cytosol (Fig. 6A2). In response to agonist activation of the CRFR1, b-a rr1–EGFP was translocated to the membrane where it colocalizes with CRFR1 as visualized by the appearance of yellow spots (Fig. 6B1). The colocalization is a lso demon- strated in the line profile in that the intensity of CRFR1– Cy3 and b-arr1–EGFP fluorescence peaks at the same position (Fig. 6B2). However, yellow spots and colocaliza- tion of fluorescence intensity in the line profile could only be observed at or near the cell membrane. No colocalization of b-arr1–EGFP with CRFR1 was observed in the CRFR1- containing intracellular vesicles in the cytoplasm of the cell (Fig. 6B1). This is also illustrated in the line profile, where the increase in intracellular CRFR1–Cy3 fluorescence representing internalized CRFR1 is not colocalized with b-arr1–EGFP ( Fig. 6B2). Involvement of putative phosphorylation sites in recruitment of b-arrestin1 To examine if removal of potential phosphorylation sites in the C-terminal tail and the third intracellular loop had an effect on recruitment of b-arrestin 1 following receptor acti- vation, trafficking of b-arr1–EGFP in response to CRF was visualized by live confocal microscopy in HEK293 cells coex- pressing b-arr1–EGFP and CRFR1-stop384:IC3. Fig. 5C1 demonstrates that in unstimulated cells fluorescence is Fig. 4. Putative phosphorylation sites in C-terminal tail and third intracellular loop are not necessary for agonist-induced receptor internalization. (A) Schematic representation of C RFR1. Mutated re sidues are shown in grey circles. (B) HEK293 cells transiently expressing the wild-type or mutant CRFR1 were incubated with I 125 -labeled sauvagine for 40 min at 37 °C. Internalization of radioligand with the wild-type receptor was set to 100%. Each bar represents the mean ± SEM of four independent experiments performed in triplicate. (***, P < 0.005 Student’s t-test). 4370 T. N. Rasmussen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 evenly distributed in the cytosol corresponding to the distribution of b-arr1–EGFP coexpressed with the wild-type CRFR1 (Fig. 5A1). When exposed to CRF, only few clusters appear at the cell surface of some cells. No profound translocation of cytosolic b-arr1–EGFP, as observed when coexpressed with the wild-type CRFR1 (Fig. 5A3), is A1 B1 0 50 100 150 Intensity Distance (µm) A2 0 20 0 50 100 150 Intensity Distance (µm) B2 25 Fig. 6. b-arrestin1 does not colocalize with CRFR1 in endocytic vesicles. HEK293 cells transiently coexpressing c-myc tagged CRFR1 and b-arr1- GFP were incubated with or without CRF for 10 min and fixed with paraformaldehyde. b-arr1–EGFP was visualized by its intrinsic fluorescence (green) and CRFR1 was visualized by immunocytochemistry with 9E10 anti-c-myc primary and Cy3-conjugated secondary antibodies (red). CRFR1/b-arr1–EGFP colocalization appears as yellow in the o verlap image. In unstimulated cells (A1), CRFR1 is confined to the plasma membrane, whereas b-arr1–EGF P is diffusely distributed in the cytosol. After exposure to CRF for 10 m in (B1), accumulation of CRFR1- containing intracellularvesiclesisobserved.b-arr1–EGFP is translocated to the plasma membrane where it colocalizes with the remaining CRFR1, but no c olocalization o f b-arr1–EGFP and intracellular CRFR1 is observed. The im ages are representative of t hree independent experiments. Bar ¼ 10 lm. Profiles of the fluorescence intensity along the lines depicted in the overlap images are shown (A2, B2). Fig. 5. Live confocal microscopy of HEK293 cells coexpressing b-arr1–EGF P and either CRFR1 or CRFR1-stop384; IC3. In unstimu- lated cells, b-arr1–EGFP cotransfected with CRFR1 was evenly distributed in the cyto- plasm (A1). Upon re ceptor stim ulation, b-arr1–EGFP was rapidly (within minutes) redistributed to the plasm a membrane (A2) and af ter 5 min this redistribution was almost complete (A3). As a cont rol, the s ame experi- ment was performed on cells expressing b-arr1–EGFP alone (B1–3). No redistribution upon CRF stimulation was observed in cells not coexpressing CRFR 1. When coexpressed with the modified receptor, CRF R1-stop384; IC3, b-arr1–EGFP was likewise distributed evenly in the cytoplasm in unstimulated cells (C1). However, shortly after stimulation (2 min, C2) not much effect was observed and after 5 min only a very small fraction of b-arr1–EGFP was observed in clusters at the plasma memb rane of some cells (C3 , a rrows). No further translocation was observed at later time points. The images are representative of three independent experiments. Bar ¼ 10 lm. Ó FEBS 2004 Internalization of CRFR1 (Eur. J. Biochem. 271) 4371 detectable (Fig. 5C2,3). These observations indicate that CRFR1-stop384; IC3 has a very limited capability of recruiting b-arr1–EGFP to t he membrane. Discussion To date, most GPCRs characterized internalize following agonist exposure. However, despite the growing in terest of CRF and CRFR1 as essential components in the aetiology of depression and other stress-related disorders, internal- ization of CRFR1 has so far not been demonstrated. I n the present study, we show agonist-mediated internalization of CRFR1. Furthermore, we demonstrate internalization to be independent of the second messenger-dependent protein kinases PKA and PKC. The extent of internalization remained unaltered in mutated CRFR1 lacking putative phosphorylation sites in the C-terminal tail and third intracellular loop as compared to the wild-type receptor. However these mutations abolished the profound recruit- ment of b-arrestin 1 to the plasma membrane observed in response to stimulation of the wild-type receptor, indicating that receptor phosphorylation and recruitment of b-arrestin might be part of an event separate from that of CRFR1 internalization. Agonist-mediated internalization of CRFR1 was dem- onstrated by visualizing CRFR1–EGFP conjugated recep- tor and by quantification of radioligand internalization. As determined by use of radioligand, the extent of receptor internalization reached alm ost its maximum after 5–10 min of agonist exposure (Fig. 2). Likewise, in response to agonist stimulation, the CRFR1–EGFP fusion protein moved from a homogenous plasma membrane distribution to a punctuate distribution and location in small intra- cellular vesicles. This redistribution was detected after 5–10 min as visualized by confocal microscopy (Fig. 1). After 20–40 min, many of the small vesicles seemed to fuse into larger structures observed near the nucleus (Fig. 1). The nature of these fluorescent aggregates is unknown, but a similar pattern of fluo rescence distribu- tion was observed when EGFP-conjugated thyrotropin- releasing hormone receptor was stim ulated with agonist for 15–30 min [23]. EGFP has a long half-life and the formation of large fluorescence aggregates in the perinuclear region could represent accumulation of EGFP-tagged CRFR1 in lysosomes [24]. Interestingly, a profound change in cell morphology is observed following stimulation with agonist (Fig. 1). The change is characterized by shrinkage of the cells and detachment from the surface of the culture dish. A similar change in the shape of the cell following agonist stimulation is observed with the Substance P receptor [25]. Detachment of the cells from the surface of the culture dish could r eflect disassembly of the actin cytoskeleton in the me mbrane regions as a consequence of excessive membrane receptor activation. Internalization of membrane proteins may occur by at least two different mechanisms: receptor endocytosis via clathrin-coated pits; or caveolae-mediated internalization. The clathrin-coated pits pathway is the best characterized endocytic route and is used by numerous GPCRs [11]. The formation of the clathrin lattice is disrupted by hypertonic sucrose and this is an established method used to block internalization processes that involves clathrin-coated pits and vesicles [21]. In t he present study, sucrose in the media inhibited CRFR1 internaliz ation by 50% indicating (Fig. 2) that a p art o f CRFR1 internalization may be clath rin dependent. Similar partial inhibition of internalization is observed for the parathyroid hormone receptor after treatment with sucrose [26]. However, it is not clear if the remaining 50% reflects an alternative mechanism of inter- nalization, such as internalization via caveolae as seen for b 1 -adrenergic receptor [15] a nd cholecystekinin receptor [13], or an incomplete inhibition of clathrin-coated pit formation. Depletion of c holesterol levels prior to experi- ments could reveal the involvement of caveolae. Activation o f m any G PCRs, i ncluding CRFR1, leads to signalling via the cAMP and/or PLC pathways, which involve activation of the second messenger-dependent protein kinases PKA and PKC. However, even though internalization o f some family II GPCRs are dependent on PKA [16] and PKC [27], not much attention has been drawn to the involvement of these kinases in GPCR internaliza- tion. CRFR1 possesses several putative P KA/PKC sites in the third intracellular loop and C-terminal tail. PKA preferentially phosphorylates serine and threonine residues with the basic residues arginine or lysine at positio n )2, )3 [28] and four serine residues in this consensus sequence are present in the third intracellular loop and in the C-terminal tail (Ser301 Ser396, Ser405 and Ser412). Two potential PKC phosphorylation sites (Ser386 and Ser408) are present in the C-terminal tail of CRFR1 [6]. In our study, truncation of the C-terminal tail at position 384 and substitution of serine/ threonine residues f or alanine in the third intracellular loop removed all putative PKA/PKC phosphorylation sites. However, this did not alter the extent of receptor internal- ization ( Fig. 4). In addition, the PKA inhibitor H-89 and the broad spectrum kinase inhibitor staurosporine had no effect on the extent o f agonist-induced CRFR1 internal- ization (Fig. 3). Thus, internalization of the family II GPCR CRFR1 does not seem to depend upon the second messenger kinases PKA and PKC. Recruitment of b-arrestin to the plasma membrane following receptor activation has been demonstrated for several GPCRs [29]. The function of b-arrestin is to uncouple the receptor f rom the G-protein, thereby mediating recepto r desensitization, and furthermore to couple the receptor to the endocytic machinery and thus mediate internalization. In this study, activation of CRFR1 led to th e recruitment of b-arrestin 1 to the plasma membrane also (Fig. 5A). The initial redistribution of b-arrestin 1 was observed already after 2 min and no further changes in translocation was observed after minutes of agonist stimulation. With the rate of CRFR1 internalization reaching maximum after 5–10 min, this result is in agreement with recruitment of b-arrestin 1 preceding receptor internalization. For many GPCRs, b-arrestin is subsequently transferred along with receptor-containing endocytic vesicles [30]. Although we demonstrated both CRFR1–EGFP and myc-tagged CRFR1 to redistribute from a diffuse mem- brane localization to intracellular vesicles, b-arr1–EGFP was never observed i n association with intracellular v esicles following receptor activation (Fig. 6). These findings indi- cate that b-arrestin 1 is recruited to the receptors at the plasma membrane following receptor activation, but that 4372 T. N. Rasmussen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 the CRFR 1/b-arrestin 1 complex dissociates at or near the plasma membrane before reaching the endocytic vesicles. This phenomenon has b een observed for other GPCRs, such as the b 2 -adrenergic receptor [30]. The difference in stability of the b-arrestin/rec eptor complex for different GPCRs appears to d epend on a cluster of serine/threonine residues in the C-ter minal tail [31]. Because such a cluster is not present in the C-terminal tail of CRFR1, it might explain why stable b-arrestin/receptor complex is not detected in our experiments. The exact mechanisms respon- sible for the dissociation of b-arrestin and receptor a re currently unknown, but the biological effect seems to be rapid receptor dephosphorylation and recycling [32]. Binding of visual arrestin to rhodopsin requires GRK phosphorylation of rhodopsin. Several other studies have confirmed the role of GRK phosphorylation in binding of b-arrestin to GPCRs [33]. The s ubstrate of GRK phos- phorylation is usually serine and threonine residues within the C-terminal tail and/or the third intracellular loop. In accordance with this theory, cells coexpressing CRFR1- stop384:IC3 and b-arrestin1 showed marked impairment of b-arrestin1 recruitment to the membrane following receptor stimulation (Fig. 5C). However, the recruitment was not completely abolished. Some clusters of b-arrestin 1 could still be observed at t he membrane in response to receptor activation even in the absence of the C-terminal tail of the receptor and the cluster of serine/threonine in the third intracellular loop. This could be due to heterologous recruitment to other endogenously expressed receptors in response to CRFR1 activation. However, the possibility remains that C RFR1 contains determinants of b-arrestin 1 interaction apart from the C-terminal tail and/or serine/ threonine in the third intracellular loop that are sufficient for transient association. Nevertheless, regions in the C-terminal tail/third intracellular loop are required for a more profound translocation of b-arrestin1 to the receptor. Interestingly, removal of putative phosphorylation sites in the C-terminal tail and/or IC3 did not alter the extent of receptor internalization (Fig. 4). This is in contrast to the prevailing observations of GPCRs dictating both the process of desensitization and of internalization to depend on receptor phosphorylation [34,35] and recruitment of b-arrestin [36]. In accordance with our studies a report on the p arathyroid hormone receptor indicates that a lysine in the third intracellular loop and an asparagine in the third transmembrane helix are important for receptor internal- ization [37]. Both residues are conserved a mong family II GPCRs, including the human CRFR1. In the VPAC1 receptor, a serine in the C-terminal tail seems to be important for receptor desensitization, but mutation of this serine does not affect agonist-mediated VPAC1 receptor internalization. This observation along with our data suggest that, at least for some family II GPCRs, desensitization and intern alizat- ion might be mediated by two separate mechanisms. Phosphorylation and recruitment of b-arrestin serve to uncouple the receptor from the G-protein, whereas other yet unidentified mechanisms serve t o render the receptors to endocytic vesicles, thus enabling receptor regulatio n by degradation or r ecycling of e ndocytosed receptors. In summary, internalization of CRFR1 in HEK293 cells in response t o agonist has been demonstrated. T he process seems to involve clathrin-coated pits, but to be independent of the activation of PKA and/or PKC. b-arrestin 1 is recruited to the receptors following activation, but no colocalization of b-arrestin 1 with receptor-containing endocytic vesicles could be observed. Whereas profound recruitment o f b-arrestin 1 following CRFR1 activation was dependent on a cluster of serine/threonine residues in the third intracellular loop and/or residues in the C-terminal tail, agonist-induced CRFR1 internalization was not, sug- gesting that CRFR1 internalizes independently of receptor phosphorylation and interaction with b-arrestin 1. Acknowledgements We thank Jacob Flemming Hansen for superior help with figure preparations and Kate Christensen and Louise Tarpø for excellent technical assistance. References 1. Vale, W., Spiess, J., Rivier, C. & Rivier, J. (1981) Characterization of a 41-residue ovine hypothalamic peptide that stimulates secre- tion of corticotropin and beta-endorphin. Science 213, 1394–1397. 2. Chang, C.P., Pearse, R.V., O’Connell, S. & Rosenfeld, M.G. (1993) Identificatio n of a seven transmembrane helix receptor for corticotropin-releasing factor and sauvagine in mammalian brain. Neuron 11, 1187–1195. 3. Grammatopoulos, D.K. & Chrousos, G.P. (2002) Functional characteristics of CRH receptors and potential clinical applica- tions of CRH-receptor antagonists. Trends Endocrinol. Metab. 13, 436–444. 4. Arborelius, L., Owens, M.J., Plotsky, P.M. & Nemeroff, C.B. (1999) The role of corticotropin-releasing factor in depression and anxiety disorders. J. Endocrinol. 160, 1–12. 5. Dautzenberg, F.M., Braun, S. & Hauger, R.L. (2001) GRK3 mediates desensitization of C RF1 receptors: a potential m echan- ism regulating stress adaptation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 280, R935–R946. 6. Hauger, R.L., Olivares-Reyes, J.A., Braun, S., Catt, K.J. & Dautzenberg, F.M. (2003) Mediat ion of corticotropin releasing factor type 1 receptor phosphorylation and desensitization by protein kinase C: a possible role in stress a daptation. J. Pharmacol. Exp. Ther. 306, 794–803. 7. Yu, S.S., Lefkowitz, R.J. & Hausdorff, W.P. (1993) Beta-adre- nergic recepto r sequ estratio n. A pot ential mec hanism o f recep tor resensitization. J. Biol. Chem. 268, 337–341. 8. Daaka, Y., Luttrell, L.M., Ahn, S., Della Rocca, G.J., Ferguson, S.S.G., Caron, M.G. & Lefkowitz, R.J. (1998) Essential role for G protein-coupled receptor endocytosis in the activation of mitogen- activated protein kinase. J. Biol. Chem . 273, 685–688. 9. Goodman, O.B. Jr, Krupnick, J.G., Santini, F., Gurevich, V .V., Penn, R.B., Gagnon, A.W., Keen, J.H. & Benovic, J.L. (1996) Beta-arrestin acts as a clathrin adaptor in endocytosis of the beta2- adrenergic receptor. Nature 383, 447–450. 10. Laporte, S.A., Oakley, R.H., Zhang, J., Holt, J.A., Ferguson, S.S., Caron, M.G. & Barak, L.S. (1999) The beta2-adrenergic receptor/ betaarrestin complex recruits the clathrin adaptor AP-2 d uring endocytosis. Proc. Natl Acad. Sci. USA 96, 3712–3717. 11. Claing,A.,Laporte,S.A.,Caron,M.G.&Lefkowitz,R.J.(2002) Endocytosis of G protein-coupled receptors: ro les o f G protein- coupled receptor kinases and beta-arrestin proteins. Prog. Neuro- biol. 66, 61–79. 12. Zhang, J., Ferguson, S.S.G., Barak, L.S., Menard, L. & Caron, M.G. (1996) Dynamin and beta-arrestin revea l distinc t me chan- isms for G protein-coupled receptor internalization. J. Biol. Chem. 271, 18302–18305. Ó FEBS 2004 Internalization of CRFR1 (Eur. J. Biochem. 271) 4373 13. Roettger, B.F., Rentsch, R.U., Pinon, D., Holicky, E., Hadac, E., Larkin, J.M. & Miller, L.J. (1995) Dual pathways of internaliza- tion of th e cholecystokinin receptor. J. Cell Biol. 128, 1029–1041. 14. de Weerd, W.F. & Leeb-Lundberg, L.M. (1997) Bradykinin sequesters B2 bradykinin receptors and the receptor-coupled Galpha subunits Galphaq and Galphai in caveolae in DDT1 MF-2 smooth muscle cells. J. Bio l. C hem. 272, 17858–17866. 15. Rapacciuolo, A., Suvarna, S., Barki-Harrington, L., Luttrell, L.M.,Cong,M.,Lefkowitz,R.J.&Rockman,H.A.(2003) Protein kinase A and G protein-coupled receptor kinase phosphorylation mediates beta-1 adrenergic receptor endocytosis through different pat hways. J. Biol. Chem. 278, 35403–35411. 16. Walker, J.K., Premont, R.T., B arak, L.S., Caron, M.G. & Shetzline, M.A. (1999) Properties of secretin receptor internaliza- tion d iffer from t hose of the beta(2)-adrenergic receptor. J. Biol. Chem. 274, 31515–31523. 17. Qi, L.J., Leung, A.T., Xiong, Y., Marx, K.A. & Abou-Samra, A.B. (1997) Extracellular cysteines of the corticotropin-releasing factor receptor are critical for ligand interac tion. Biochemistry 36, 12442–12448. 18. Barak, L.S., Ferguson, S.S., Zhang, J., Martenson, C., Meyer, T. & Caron, M.G. (1997) Internal trafficking and surface mobility of a functionally intact beta2-adrenergic receptor-green fluorescent protein conjugate. Mol. Pharmacol. 51, 177–184. 19. Kallal, L. & Benovic, J.L. (2000) Using green fluorescent proteins to st udy G-protein-coupled recepto r l ocalization and trafficking. Trends Pharmacol. Sci. 21, 175–180. 20. Ho, S.N., Hunt, H.D., Horton, R.M., Pullen, J .K. & Pease, L.R. (1989) Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77, 51–59. 21. Heuser, J.E. & Anderson, R.G. (1989) Hypertonic media inhibit receptor-mediated endocytosis b yblockingclathrin-coatedpit formation. J. Ce ll B iol. 108, 389–400. 22. Hillhouse, E.W., Randeva, H., Ladds, G. & Grammatopoulos, D. (2002) Corticotropin-releasing hormone receptors. Biochem. Soc. Trans. 30, 428–432. 23. Drmota, T., Gould, G.W. & Milligan, G. (1998) Real time visualization of agonist-mediated redistribution and internaliza- tion of a green fluorescent protein-tagged form of the thyrotropin- releasing hormone re ceptor. J. Biol. Chem. 273, 24000–24008. 24. McLean, A.J. & Milligan, G. (2000) Ligand re gulation of green fluorescent protein-tagged forms of the human b eta(1)- and beta(2)-adrenoceptors; c omparisons with the unmodified recep- tors. Br. J. Pharmacol. 130, 1825–1832. 25. Barak,L.S.,Warabi,K.,Feng,X.,Caron,M.G.&Kwatra,M.M. (1999) Real-time visualization o f t he cellular redistribution of G protein-coupled receptor kinase 2 and beta-arrestin 2 during homologous desensitization of the substance P receptor. J. Biol. Chem. 274, 7565–7569. 26. Malecz, N., Bambino, T., B encsik, M. & Nissenson, R.A. (1998) Identification of phosphorylation sites in the G protein-coupled receptor for parathyroid hormone. Receptor phosph orylation is not required for agonist-induced internalization. Mol. Endocrinol. 12, 1846–1856. 27. Ferrari, S.L., Behar, V., Chorev, M., Rosenblatt, M. & Bisello, A. (1999) Endocytosis of ligand-human parathyroid hormone receptor 1 complexes is protein kinase C-dependent and involves beta-arrestin2. Real-time monitoring by fluo rescence microscopy. J. Biol. Chem. 274, 29968–29975. 28. Blom,N.,Kreegipuu,A.&Brunak, S. (1998) PhosphoBase: a database of phosphorylation sites. Nucleic Acids Res. 26, 382–386. 29. Barak, L.S., Ferguson, S.S., Zhang, J. & Caron, M.G. (1997) A beta-arrestin/green fluorescent Protein biosensor for detecting G protein-coupled receptor activation. J. Biol. Chem. 272, 27497– 27500. 30. Zhang, J., Barak, L.S., Anborgh, P.H., Laporte, S.A., Caron, M.G. & F erguson, S.S. (1999) C ellular trafficking of G protein- coupled rec eptor/beta -arrestin e ndocytic c omplex es. J. Biol. C hem. 274, 10999–11006. 31. Oakley,R.H.,Laporte,S.A.,Holt,J.A.,Barak,L.S.&Caron, M.G. (2001) Molecular determinants underlying the formation o f stable intracellular G protein-coupled receptor-beta-arrestin complexes after receptor endocy tosis. J. Biol. C hem. 276, 19452– 19460. 32. Oakley,R.H.,Laporte,S.A.,Holt,J.A.,Barak,L.S.&Caron, M.G. (1999) Association of beta-arrestin with G protein-coupled receptors during clathrin-mediated endocytosis dictates the profile of receptor resen sitization. J. Biol. Chem. 274, 32248–32257. 33. Claing,A.,Laporte,S.A.,Caron,M.G.&Lefkowitz,R.J.(2002) Endocytosis of G protein-coupled recep tors: ro les o f G protein- coupled receptor kinas es and beta-arrestin protein s. Prog. Neurobiol. 66, 61–79. 34. Bunemann, M. & Hosey, M.M. (1999) G-pro tein coupled receptor kinases as modulators of G-protein signalling. J. Physiol. 517 (1), 5–23. 35. Ferguson, S.S., Menard, L., Barak, L.S., Koch, W.J., Colapie tro, A.M. & Caron, M.G. (1995) Role of phosphorylatio n in agonist- promoted beta 2-adrenergic receptor sequestration. Rescue of a sequestration-defective mutant receptor by beta ARK1. J. Biol. Chem. 270, 24782–24789. 36. Ferguson, S.S., Downey, W.E. III, Colapietro, A.M., Barak, L.S., Menard, L. & Caron, M.G. (1996) Role of beta-arrestin in mediating agonist-promoted G protein-coupled receptor in ter- nalization. Science 271, 363–366. 37. Vilardaga, J.P., Krasel, C., Chauvin,S.,Bambino,T.,Lohse,M.J. & Nissenson, R.A. (2002) Internalization determinants of the parathyroid hormone receptor differentially regulate beta–arres- tin/receptor association. J. Biol. Chem. 277, 8121–8129. 4374 T. N. Rasmussen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Internalization of the human CRF receptor 1 is independent of classical phosphorylation sites and of b-arrestin 1 recruitment Trine N. Rasmussen 1, 3 , Ivana Novak 3 and Søren M. Nielsen 2 1 Department. neuro- peptide first isolated in 19 81 [1] as the main stimulator of the release of adenocorticotropin from the pituitary. By activa- tion of CRF Receptor 1 (CRFR1) [2], CRF regulates not only the endocrine,. phosphorylation and recruitment of b-arrestin might be part of an event separate from that of CRFR1 internalization. Agonist-mediated internalization of CRFR1 was dem- onstrated by visualizing CRFR1–EGFP