Essential Role of Dynamin in Internalization of M2Muscarinic Acetylcholine and Angiotensin AT1AReceptors*

Most G protein-coupled receptors (GPCRs), including the M1 muscarinic acetylcholine receptor (mAChR), internalize in clathrin-coated vesicles, a process that requires dynamin GTPase. The observation that some GPCRs like the M2 mAChR and the angiotensin AT1A receptor (AT1AR) internalize irrespective of expression of dominant-negative K44A dynamin has led to the proposal that internalization of these GPCRs is dynamin-independent. Here, we report that, contrary to what is postulated, internalization of M2mAChR and AT1AR in HEK-293 cells is dynamin-dependent. Expression of N272 dynamin, which lacks the GTP-binding domain, or K535M dynamin, which is not stimulatable by phosphatidylinositol 4,5-bisphosphate, strongly inhibits internalization of M1 and M2 mAChRs and AT1ARs. Expression of kinase-defective K298M c-Src or Y231F,Y597F dynamin (which cannot be phosphorylated by c-Src) reduces M1 mAChR internalization. Similarly, c-Src inhibitor PP1 as well as the generic tyrosine kinase inhibitor genistein strongly inhibit M1 mAChR internalization. In contrast, M2 mAChR internalization is not (or is only slightly) reduced by expression of these constructs or treatment with PP1 or genistein. Thus, dynamin GTPases are not only essential for M1 mAChR but also for M2 mAChR and AT1AR internalization in HEK-293 cells. Our findings also indicate that dynamin GTPases are differentially regulated by c-Src-mediated tyrosine phosphorylation.

For most G protein-coupled receptors (GPCRs), 1 receptor internalization is thought to be initiated by phosphorylation of the receptor by G protein-coupled receptor kinases and binding of the cytosolic protein ␤-arrestin to the phosphorylated receptor (1). ␤-Arrestin then sterically inhibits further interaction of the receptor with heterotrimeric G proteins and binds with high affinity to clathrin heavy chains (1). Through this interaction, GPCRs are believed to be targeted to clathrin-coated pits. Following transformation of the clathrin-coated pit into a clathrin-coated vesicle, the clathrin-coated vesicle pinches off from the plasma membrane. This process is catalyzed by the 100-kDa GTPase dynamin, which probably activates (as yet largely unknown) effectors of the fission machinery (2). Three closely related mammalian dynamin isoforms have been identified: neuronal dynamin-1, ubiquitously expressed dynamin-2, and dynamin-3, which is expressed in testes, neurons, and lung (3). Comparison of the primary sequence shows that all three dynamin isoforms contain three highly conserved GTP-binding motifs (i.e. elements I, II, and III). A Lys 44 3 Ala substitution in the first of the three putative GTP-binding motifs yields a dominant-negative dynamin mutant, which displays strongly impaired GTPase activity and is predicted to have a greatly reduced GTP binding affinity (4). The two other GTP-binding motifs in dynamin are likely to be involved in GTP binding as well. Mutation of the third GTP-binding motif (substitution Lys 206 3 Asp in element III) or removal of all three GTPbinding motifs (amino acids 1-271 in dynamin-1) drastically reduces clathrin-coated vesicle-mediated internalization (4 -6). A second important regulator of dynamin function is phosphatidylinositol 4,5-bisphosphate (PIP 2 ) (6 -9). All three dynamin isoforms contain a pleckstrin homology domain that is able to bind PIP 2 . Binding of PIP 2 to dynamin not only strongly increases the GTPase activity of dynamin but may also serve to target dynamin to the plasma membrane, allowing subsequent dynamin self-assembly at the neck of the clathrin-coated vesicle (6 -9). Expression of the dynamin mutant K535M, which is not stimulatable by PIP 2 , effectively blocks transferrin receptor internalization in clathrin-coated vesicles (6).
A large number of recent studies indicate that most GPCRs, including M 1 , M 3 , and M 4 muscarinic acetylcholine receptors (mAChRs) in HEK-293 cells, internalize in clathrin-coated vesicles in a dynamin-dependent manner. This evidence is primarily based on the inhibitory effect of the dominant-negative inhibitor of dynamin-mediated internalization, K44A dynamin (10 -13). In contrast, M 2 mAChRs internalize in a clathrinindependent manner and irrespective of expression of K44A dynamin in HEK-293 cells (10,12). Likewise, internalization of angiotensin AT 1A receptors (AT 1A Rs) (13), dopamine D 2 receptors (14), and secretin receptors (15) is also insensitive to expression of K44A dynamin. This has led to the proposal that internalization of these GPCRs is dynamin-independent. However, in light of the notion that the binding of GTP to dynamin probably involves binding to all three GTP-binding motifs in the GTP-binding pocket, we reasoned that a dynamin mutant lacking all three GTP-binding motifs might be a more appropriate dominant negative dynamin mutant to determine whether internalization of a particular GPCR is dynamin-dependent. Indeed, we here demonstrate that internalization of M 2 mAChR and AT 1A R is strongly inhibited by expression of N272 dynamin, which lacks the complete GTP-binding domain. Also, expression of K535M dynamin, which lacks PIP 2 -stimulated GTPase activity, significantly blocks internalization of these GPCR species. were purchased from Biomol and Calbiochem, respectively. The cDNA constructs encoding wild-type c-Src and K298M c-Src were gifts from Dr. J. T. Parsons. Rat wild-type dynamin-1aa and rat K535M dynamin-1aa cDNA in pCMV96-7 (6) were generously provided by Dr. J. P. Albanesi. Rat Y231F,Y597F dynamin-1aa in pCMV96-7 (16) was a gift from Dr. R. J. Lefkowitz. The mouse AT 1A R cDNA in pBC12BI was a gift from Dr. L. Hein. The cDNAs encoding HA-tagged wild-type and K44A human dynamin-1aa (4) in pRK5 were gifts from Dr. S. Schmid. Rat dynamin-1aa N272 was generated by digestion of rat wild-type dynamin-1aa in pCMV96-7 with BglII and EcoRV. The product was filled in with Klenow DNA polymerase and religated with T4 DNA ligase. The authenticity of the N272 dynamin mutant was confirmed by dideoxy DNA sequencing and Western blot analysis. Rabbit anti-c-Src polyclonal antibody (N-16), mouse anti-Tyr(P) antibody (PY20), and goat anti-dynamin-1 antibody (C-16) were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse anti-Myc 9E10 antibody, mouse antidynamin antibody (clone 41), and peroxidase-conjugated goat anti-mouse antibody were obtained from Calbiochem, Transduction Laboratories, and Dianova, respectively. Peroxidase-conjugated rabbit anti-goat antibody, peroxidase-conjugated goat anti-rabbit antibody, and fluorescein isothiocyanate-labeled anti-mouse antibody were from Sigma.
Cell Culture, Plasmid Construction, and DNA Transfection-HEK-293 tsA201 and HEK-293 cells were grown in Dulbecco's modified Eagle's medium (DMEM)/F-12 medium supplemented with 10% fetal calf serum, penicillin G (100 units/ml), and streptomycin (100 g/ml) in an atmosphere of 5% CO 2 . Cells plated on 150-mm plates were transiently transfected with mAChR/pCD-PS or AT 1A R/pBC12BI DNA, together with one of the vectors listed above as described before (10). For epitope tagging of M 2 mAChRs, the Myc sequence EQKLISEEDL was inserted after the initiator methionine in the extracellular amino terminus of the receptor by the polymerase chain reaction method with Pfu DNA polymerase (Stratagene). The complete receptor DNA sequence was verified by dideoxy DNA sequencing and subcloned in pcDNA3 (Invitrogen). Stably transfected HEK-293 cells expressing Myc-tagged M 2 mAChR were selected after culturing in medium containing 500 g/ml G418 (Life Technologies, Inc.).
Immunoblot Analysis of c-Src and Dynamin Expression-Forty-eight hours after DNA transfection, cells on 35-mm plates were washed twice with phosphate-buffered saline (150 mM NaCl, 6.5 mM Na 2 HPO 4 , 2.7 mM KCl, pH 7.4) and lysed by the addition of 0.5 ml of boiling lysis buffer (1% SDS, 10 mM Tris-HCl, pH 7.4). Lysate was transferred to a microcentrifuge tube and boiled for 5 min. After five passages through a 25-gauge needle, samples were centrifuged for 5 min to remove insoluble material and diluted to an equal amount of protein as measured by the BCA method (Pierce) with lysis buffer. Fifty microliters of electrophoresis sample buffer (250 mM Tris-HCl, pH 6.8, 4% SDS, 10% glycerol, 0.006% bromphenol blue, 2% 2-mercaptoethanol) were added to 50 l of the diluted samples and boiled for another 5 min. After SDS-polyacrylamide gel electrophoresis on 10% polyacrylamide gels, protein was blotted onto nitrocellulose. Nitrocellulose was then blocked with 150 mM NaCl, 10 mM Tris-HCl, pH 7.5, containing 5% bovine serum albumin (BSA; fraction V; Sigma) (dynamin) or 5% skin milk (c-Src). After washing three times for 5 min in 150 mM NaCl, 10 mM Tris-HCl, pH 7.5, 0.1% Tween 20, the blot was incubated with either anti-dynamin antibody (clone 41, 0.125 g/ml; or C-16, 0.4 g/ml) or anti-c-Src antibody N-16 (0.1 g/ml) in blocking buffer for 1 h. Following four washes for 5 min in wash buffer and incubation in blotting buffer for 10 min, the blot was incubated with horseradish peroxidase-conjugated goat anti-mouse antibody (0.2 g/ml), horseradish peroxidase-conjugated rabbit anti-goat antibody (diluted 1:1000), or horseradish peroxidase-conjugated goat anti-rabbit antibody (diluted 1:1000) at room temperature. After 1 h, the blot was washed again, and immunoreactivity was visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech).
Immunocytochemical Localization of Myc-tagged M 2 mAChR-HEK-293 cells stably expressing Myc-tagged M 2 mAChR at a density of 350 fmol/mg of protein and grown on poly-L-lysine-coated 18 ϫ 18-mm glass coverslips were incubated in 25 mM HEPES-buffered DMEM/F-12 medium (pH 7.4) in the presence and absence of 1 mM carbachol for 60 min. Then cells were washed twice with phosphate-buffered saline, fixed, and permeabilized in methanol for 5 min at 4°C. Cells were washed three times with phosphate-buffered saline for 5 min each and incubated in TBS (10 mM Tris-HCl, pH 7.4, 150 mM NaCl) containing 0.5% fatty acid-free BSA for 45 min at room temperature. After incubation with mouse anti-Myc 9E10 antibody (5 g/ml) in TBS plus 0.5% BSA for 60 min at room temperature, cells were washed three times with TBS plus 0.5% BSA for 5 min each and subsequently incubated with fluorescein isothiocyanate-labeled anti-mouse antibody (10 g/ml) in TBS plus 0.5% BSA for 60 min at room temperature in the dark. After three washes with TBS plus 0.5% BSA for 5 min each and once in TBS for 5 min, coverslips were mounted using Moviol (Calbiochem). Immunofluorescence was detected using a Zeiss Axiophot fluorescence microscope equipped with standard fluorescein filter. Immunofluorescence was marginal in nontransfected cells and cells expressing wild-type M 2 mAChRs (without Myc tag), as well as in Myc-tagged M 2 mAChR-expressing cells when second antibody without the first antibody was used.
Dynamin Immunoprecipitation-HEK-293 cells on 150-mm plates stably expressing M 1 mAChRs (17) were serum-starved in DMEM/F-12 medium. After 16 h, cells were preincubated in 25 mM HEPES-buffered DMEM/F-12 medium for 30 min and then stimulated for 5 min with 100 M carbachol. After rapid suction of medium from the plates, 1.5 ml of ice-cold radioimmune precipitation assay buffer (10 mM Tris-HCl, pH 7.4, 10 mM EDTA, 500 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate, 20 mM NaF, 1 mM NaVO 4 , 1 mM dithiothreitol, 2.5 g/ml leupeptin, 2.5 g/ml aprotinin, 100 M phenylmethylsulfonyl fluoride) was added to the cell monolayer. After 10 min at 4°C, cells were lysed by repeated aspiration through a 21-gauge and 25-gauge needle. The cell lysate was centrifuged at 14,000 ϫ g for 10 min at 4°C. The supernatant was incubated with mouse anti-dynamin monoclonal antibody (3.75 g) for 1 h at 4°C followed by incubation with 50 l of Protein G Plus/Protein A-agarose beads (Calbiochem) for 2 h at 4°C while rotating. Then immunoprecipitates were collected by centrifugation at 14,000 ϫ g at 4°C. The pellets were washed five times with ice-cold radioimmune precipitation buffer and once with ice-cold buffer A (110 mM NaCl, 3 mM KCl, 7 mM Na 2 HPO 4 , 2 mM KH 2 PO 4 , and 20 mM NaF, pH 7.4). The pellets were resuspended in 30 l of 2ϫ Laemmli sample buffer. After boiling for 5 min, samples were centrifuged, and protein in the supernatant was analyzed on a 10% SDS-polyacrylamide gel and Western blotting with anti-Tyr(P) antibody (0.2 g/ml) and, after stripping, with anti-dynamin antibody (clone 41; 0.125 g/ml). Immunoreactivity was visualized with peroxidase-conjugated goat antimouse antibody.
c-Src Immunoprecipitation and c-Src Kinase Activity Assay-After serum depletion for 16 h, HEK-293 cells on 150-mm plates stably expressing M 1 or M 2 mAChRs (17) were stimulated for 5 min at 37°C with 100 M carbachol in 25 mM HEPES-buffered DMEM/F-12 medium. Then the medium was rapidly removed from the plates, and the cells were lysed in 1.5 ml of lysis buffer (10 mM Tris-HCl, pH 7.4, 500 mM NaCl, 10 mM EDTA, 20 mM NaF, 1 mM NaVO 4 , 1.0% Nonidet P-40, 1 mM dithiothreitol, 2.5 g/ml leupeptin, 2.5 g/ml aprotinin, and 100 M phenylmethylsulfonyl fluoride). From this lysate, c-Src was immunoprecipitated with anti-Src antibody N-16 (1.5 g) with 50 l of Protein A Plus/Protein G-agarose beads. Immunoprecipitates were washed five times with lysis buffer and once in c-Src kinase buffer (100 mM Tris-HCl, pH 7.4, 125 mM MgCl 2 , 25 mM MnCl 2 , 2 mM EGTA, 250 M NaVO 4 , 2 mM dithiothreitol), followed by resuspension in 85 l of Src kinase buffer with prepared c-Src substrate peptide and [␥-32 P]ATP (125 M, 10 -20 Ci/vial) according to the manufacturer's instructions (Upstate Biotechnology, Inc., Lake Placid, NY). The mixture was incubated at 30°C for 12 min while shaking, and the reaction was stopped by the addition of 40 l of 40% trichloroacetic acid. Thirty-microliter aliquots of the reaction mixture were spotted on P81 phosphocellulose paper in duplicate, washed five times for 5 min each with 0.75% phosphoric acid and once with acetone for 3 min, followed by radioactivity counting.
Receptor Internalization Assays-Internalization of mAChRs was determined 48 h after DNA transfection by [ 3 H]NMS binding assays to intact cells in 25 mM HEPES-buffered saline, pH 7.4 (HBS), containing 113 mM NaCl, 6 mM glucose, 3 mM KCl, 3 mM MgCl 2 , 2 mM MgSO 4 , and 1 mM NaH 2 PO 4 at 4°C as described in detail previously (10). Expression levels of M 1 and M 2 mAChRs varied between 100 and 750 fmol/mg of protein. Where indicated, transfected HEK-293 tsA201 cells were serum-starved for an additional 16 h in DMEM/F-12 medium. AT 1A R internalization was measured following incubation in 25 mM HEPESbuffered DMEM/F-12 medium buffer containing 1 mg/ml BSA with 1 M unlabeled human angiotensin II (Sigma) for 60 min at 37°C. Cells were then washed twice with ice-cold HBS; twice with ice-cold 20 mM 2-morpholinoethanesulfonic acid, 300 mM NaCl (pH 5.0); and twice with ice-cold HBS buffer (3 min each) to remove angiotensin II from receptor. Thereafter, cells were incubated in HBS buffer (with 1 mg/ml BSA) at 4°C with 4 -5 pM [ 125 I]Tyr 4 -Sar 1 -Ile 8 -angiotensin II with or without 10 M angiotensin II to determine nonspecific and total binding, respectively. After 4 h, cells were washed three times with HBS buffer and solubilized in 0.1% Triton X-100, and radioactivity was counted. Depletion of [ 125 I]Tyr 4 -Sar 1 -Ile 8 -angiotensin II was limited to maximally 20% of total added radioligand by the inclusion of 4 nM unlabeled angiotensin II. Receptor internalization is expressed as (1 Ϫ quotient of cell surface receptors of agonist-treated and untreated cells) ϫ 100%.

Subcellular Redistribution of M 2 mAChR in HEK-293 Cells in Response to
Carbachol-Internalization of M 2 mAChRs in HEK-293 cells was monitored by indirect immunofluorescence of M 2 mAChRs tagged with a c-Myc epitope at the extracellular amino terminus. For this, we used stably M 2 mAChR-expressing cells instead of transiently expressing HEK-293 tsA201 cells. In the latter cell type, there was a significant preexisting intracellular pool of M 2 mAChRs that did not permit unequivocal demonstration of receptor translocation from the plasma membrane into the cytosol upon carbachol treatment. Control [ 3 H]NMS binding experiments demonstrated that the Myctagged M 2 mAChRs sequestered with similar characteristics as the wild-type M 2 mAChRs in either HEK-293 tsA201 or HEK-293 cells (data not shown). As shown in Fig. 1A, M 2 mAChRs in untreated cells were found predominantly at the cell surface. During 60 min of incubation with 1 mM carbachol, M 2 mAChRs translocated into the cytoplasm (Fig. 1B). These results indicate that M 2 mAChRs like M 1 mAChRs (18) internalize into the cell interior of HEK-293 cells.
Effect of N272 Dynamin on M 1 -M 4 mAChR Internalization in HEK-293 Cells-To investigate the role of dynamin in M 2 mAChR internalization, N-terminal deletion dynamin-1 mutant N272 was expressed with either M 1 or M 2 mAChR in HEK-293 tsA201 cells. Fig. 2 shows the overexpression of the various transfected dynamin constructs used in this study. Expression of all dynamin forms (with the exception of N272 dynamin) was determined with a dynamin antibody recognizing the N-terminal part of dynamin-1 and dynamin-2. N272 dynamin-1, which migrates with an apparent molecular mass of ϳ72 kDa instead of ϳ100 kDa, lacks the greater part of this antibody-binding epitope. Expression of N272 dynamin was therefore detected by a dynamin-1 antibody that specifically recognizes a C-terminal domain of dynamin-1. Fig. 3 shows the effect of expression of N272 dynamin on M 1 and M 2 mAChR internalization in HEK-293 tsA201 cells. Expression of N272 dynamin inhibited internalization of M 1 and M 2 mAChRs in response to receptor stimulation with 100 M or 10 M carbachol for 60 min by 68 and 55%, respectively. Also, sequestration of M 3 and M 4 mAChRs in response to 100 or 10 M carbachol for 60 min was reduced from 20 Ϯ 5 to 1 Ϯ 1% and from 35 Ϯ 3 to 12 Ϯ 4%, respectively, by co-expression of N272 dynamin in HEK-293 tsA201 cells (n ϭ 3 experiments; data not shown). In contrast, as reported earlier by us (10) and others (12), expression of K44A dynamin inhibited M 1 but not M 2 mAChR internalization (Fig. 3).
Role of c-Src in Dynamin-mediated mAChR Internalization in HEK-293 Cells-The results presented above strongly suggested that dynamin is not only required for internalization of M 1 , M 3 , and M 4 mAChRs but also essential for M 2 mAChR internalization in HEK-293 tsA201 cells. We therefore set out to analyze whether dynamin function in the M 1 and M 2 mAChR internalization pathways is differentially regulated. Recently, it was reported that internalization of ␤ 2 -adrenergic Equal amounts of total cell lysates (50 g of protein/lane) were subjected to SDS-polyacrylamide gel electrophoresis and immunoblotting. Expression of endogenous dynamin (lane 1), wild-type dynamin, and the dynamin mutants K44A, K535M, and Y231F,Y597F was determined using a mouse anti-dynamin monoclonal antibody recognizing the N terminus of dynamin. N272 dynamin expression was determined using a rabbit anti-dynamin polyclonal antibody directed against the C terminus of dynamin-1. Control immunoblot experiments with the rabbit anti-dynamin polyclonal antibody showed that expression of N272 dynamin was similar to expression of the other transfected dynamin constructs (data not shown). receptors in HEK-293 cells requires c-Src-mediated phosphorylation of dynamin on two tyrosine residues (i.e. Tyr 231 and Tyr 597 ) (16). c-Src is activated by ␤-arrestin, which is bound to the agonist-occupied ␤ 2 -adrenergic receptor and targets the receptor to the clathrin-coated pit (16). As M 1 mAChRs in HEK-293 tsA201 cells internalize into clathrin-coated vesicles in a ␤-arrestin-dependent manner (11), we first investigated the role of c-Src in M 1 mAChR internalization in HEK-293 cells. As shown in Fig. 4A, activation of M 1 mAChRs in HEK-293 cells with 100 M carbachol for 5 min increased c-Src kinase activity by 208 Ϯ 16%. Basal and receptor-stimulated c-Src activity was effectively blocked by treatment with the selective c-Src inhibitor PP1 (1 M). Stimulation of M 1 mAChRs led to an 82 Ϯ 29% increase in tyrosine phosphorylation of endogenously expressed dynamin (Fig. 4B) (mean Ϯ S.E. of four independent experiments), an increase that is comparable with the increases observed in other experimental systems (16,19). As shown in Fig. 5A, transfection of HEK-293 tsA201 cells with wild-type c-Src or K298M c-Scr cDNA led to a strong overexpression of the corresponding c-Src protein over endogenous c-Src. While the expression of wild-type c-Src had no effect, the expression of catalytically defective K298M c-Src reduced M 1 mAChR internalization by 52% (Fig. 5B). Also, expression of Y231F,Y597F dynamin, which cannot be tyrosine-phosphorylated by c-Src (16), inhibited M 1 mAChR internalization by 76% (Fig. 5B). In contrast, expression of a catalytically defective mutant of another tyrosine kinase, Pyk2 (K457A), did not affect M 1 mAChR internalization (data not shown). Like M 1 mAChRs, M 2 mAChRs stimulate c-Src activity in HEK-293 cells, albeit to a smaller extent than M 1 mAChRs (i.e. 78 Ϯ 34%) (Fig. 4A). However, expression of K298M c-Src did not inhibit M 2 mAChR internalization, while expression of Y231F,Y597F dynamin slightly reduced M 2 mAChR internalization. These results are supported by the observation that M 2 mAChR internalization in HEK-293 tsA201 cells was not inhibited by pretreatment of the cells with PP1 (10 M), while the generic tyrosine kinase inhibitor genistein (100 M) inhibited M 2 mAChR internalization by 16% (Table I). In contrast, internalization of M 1 mAChRs was reduced by 41 and 48% following treatment of the cells with 10 M PP1 or 100 M genistein, respectively (Table I).
Effect of K535M Dynamin on M 1 -M 4 mAChR Internalization in HEK-293 Cells-Since PIP 2 has recently been implicated as an important regulator of dynamin function (6 -9), we tested the role of PIP 2 binding in dynamin-mediated mAChR internalization by coexpression of K535M dynamin. Western blot analysis of K535M dynamin expression is shown in Fig. 2. As depicted in Fig. 6  Effect of N272 and K535M Dynamin on AT 1A R Internalization in HEK-293 Cells-Since internalization of AT 1A Rs in HEK-293 cells has been previously reported to be insensitive to overexpression of K44A dynamin (13), we also determined whether N272 dynamin or K535M dynamin blocks AT 1A R internalization in HEK-293 tsA201 cells. As shown in Fig. 7, expression of N272 dynamin and K535M dynamin inhibited AT 1A R internalization by 63 and 71%, respectively. In accordance with the aforementioned study on AT 1A R internalization (13), expression of K44A dynamin did not affect AT 1A R internalization (Fig. 7). DISCUSSION In the past few years, the question whether dynamin plays an essential role in the internalization of a particular GPCR has been mostly analyzed by using K44A dynamin as dominant-negative mutant. While internalization of most GPCRs is blocked by expression of K44A dynamin, some GPCRs like the M 2 mAChRs, D 2 dopamine receptors, secretin receptors, and AT 1A Rs internalize irrespective of K44A dynamin expression, suggesting that internalization of these GPCRs is dynaminindependent (10,(12)(13)(14)(15). We now report that, contrary to what is currently postulated, internalization of M 2 mAChR and AT 1A R is dynamin-dependent. Coexpression of the dominantnegative dynamin mutants N272 and K535M strongly inhibited M 2 mAChR and AT 1A R internalization in HEK-293 tsA201 cells. These findings imply that N272 and K535M dynamin are more appropriate dominant-negative dynamin mutants than K44A dynamin. In this context, it will be interesting to determine whether fluid-phase endocytosis (5,20) and internalization of ricin (21) are affected by expression of N272 or K535M dynamin also, because these trafficking processes are not blocked by K44A dynamin and are thus considered to be dynamin-independent. It is intriguing that N272 dynamin, which lacks all three GTP-binding motifs, inhibits internalization of both M 1 and M 2 mAChR, while K44A dynamin, which lacks only the first GTP-binding motif, blocks only M 1 mAChR internalization. It is possible that K44A dynamin selectively sequesters away an essential component of the M 1 but not of the M 2 mAChR internalization pathway. Another potential explanation relates to the fact that K44A dynamin is able to coassemble with wild-type dynamin (22). Since dynamin assembly and interaction of dynamin with other proteins requires the C terminus of the dynamin, which varies among the dynamin isoforms (23), different internalization pathways may use different dynamin isoforms. As a result, different internalization pathways may display differential sensitivity toward interference of K44A dynamin. Perhaps assembled GTP-bound K44A dynamin is sufficiently active to catalyze the budding of M 2 mAChR-and AT 1A R-containing vesicles from the plasma membrane but is not able to support internalization of M 1 mAChRs in clathrin-coated vesicles.  In the present study, we observed that mAChR and AT 1A R internalization in HEK-293 cells is strongly inhibited by expression of K535M dynamin, a dynamin mutant, which lacks the putative PIP 2 binding site (6). At present, it is unknown at which stage of the vesicle budding process PIP 2 binding to dynamin is essential. It has been postulated that, after recruitment of dynamin to the clathrin-coated pit, dynamin's interaction with the plasma membrane is strengthened by the binding of dynamin's pleckstrin homology domain with PIP 2 in the plasma membrane (8). In addition, PIP 2 binding might promote self-assembly of the dynamin molecules at the neck of the clathrin-coated pit and stimulate dynamin's GTPase activity (8,9). An alternative possibility is that lysine 535 in the pleckstrin homology domain of dynamin serves to promote interaction of dynamin with proteins rather than with PIP 2 (6). Regardless of the mechanism, our study clearly underscores the relevance of dynamin's lysine 535 residue in GPCR internalization. Identification of the binding partners of dynamin's pleckstrin homology domain will be important for understanding the mechanisms of GPCR internalization.
In analogy with recent studies on the regulation of internalization of ␤ 2 -adrenergic receptors in HEK-293 cells (16,24), we show that internalization of M 1 mAChRs is strongly reduced by inhibition of c-Src activity and by overexpression of Y231F,Y597F dynamin, which cannot be phosphorylated by c-Src. Since M 1 mAChRs internalize in clathrin-coated vesicles in a ␤-arrestindependent manner, we propose that, in analogy to ␤ 2 -adrenergic receptors, internalization of M 1 mAChRs involves ␤-arrestin-mediated targeting of receptor in the clathrin-coated pit and activation of c-Src by ␤-arrestin. c-Src then phosphorylates dynamin, a process that is required for M 1 mAChR internalization in HEK-293 cells. Whether tyrosine phosphorylation activates dynamin or allows activation of dynamin by other molecules remains to be determined. In contrast, dynaminmediated internalization of M 2 mAChRs was found not to be inhibited by expression of kinase-defective K298M c-Src or treatment of the cells with the specific c-Src inhibitor PP1. Thus, c-Src does not play a role in M 2 mAChR internalization. These findings are supported by recent studies showing that M 2 mAChR internalization in HEK-293 cells is ␤-arrestinindependent (11,12). However, treatment of the cells with the generic tyrosine kinase, genistein, or coexpression of Y231F,Y597F dynamin did slightly reduce M 2 mAChR internalization. This suggests that M 2 mAChR internalization is regulated to a very limited extent by phosphorylation of dynamin by tyrosine kinases other than c-Src. On the basis of the present and previous findings (11), we propose that M 2 mAChR internalization in HEK-293 cells is catalyzed by a dynamin isoform that differs from the dynamin isoform involved in clathrin-mediated M 1 mAChR internalization. Much remains to be learned about the internalization pathway of M 2 mAChR (and AT 1A R) in HEK-293 cells. We have observed that pretreatment of HEK-293 cells with 0.45 M sucrose fully blocks M 2 (and M 1 ) mAChR internalization in HEK-293 cells. 2 Yet expression of a dominant-negative clathrin mutant or ␤-arrestin V53D, which inhibits clathrin-mediated internalization of GPCRs, blocks M 1 but not M 2 mAChR internalization (11). Similarly, expression of the amphiphysin SH3 domain, which blocks the targeting of dynamin to clathrin-coated pits, inhibits M 1 but not M 2 mAChR internalization in HEK-293 cells. 2 On the basis of these findings, we conclude that M 2 mAChR internalization in HEK-293 cells is clathrin-independent and that the inhibitory effect of sucrose on receptor internalization is not specific for clathrin-mediated internalization. In this respect, it is important to note that hypertonic sucrose treatment of HEK-293 cells does not only block vesicle formation at the plasma membrane but may also induce other cellular responses including MAP kinase activation (25), which may inhibit receptor internalization indirectly (26). Interestingly, similar findings have been obtained recently with the secretin receptor. Internalization of secretin receptors in HEK-293 cells is unaffected by expression of dynamin K44A and ␤-arrestin V53D but is sensitive to sucrose pretreatment (15). Presently, it is unknown through which vesicles AT 1A R internalizes in HEK-293 cells. In adrenal glomerulosa and Chinese hamster ovary cells, AT 1A R may internalize through clathrin-coated vesicles, although this has only been inferred from biochemical (and not morphological) experiments using hypertonic sucrose treatment or potassium depletion (27). In vascular smooth muscle cells, however, AT 1A Rs have been found to internalize through noncoated pits, possibly caveolae, as well as coated pits (27,28). Thus, the internalization of AT 1A R and other GPCRs (10) appears to differ among different cell types. The identification of the budding vesicles through which AT 1A R and M 2 mAChR internalize in HEK-293 cells and other cell types will provide important information on the molecular mechanisms of GPCR internalization.