INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

G PROTEIN-COUPLED CELL MEMBRANE RECEPTORS (GPCRs) respond to a wide variety of stimuli. For the purpose of our primary interest, the regulation of cardiac muscle biology, these receptors can be considered in terms of their effects on cAMP-dependent protein kinase, where they may couple to adenyl cyclase either via the stimulatory G protein G_s to increase cAMP, via the inhibitory G protein G_i to decrease cAMP, or instead act on phospholipase C independently of cAMP via the G_q/11 protein family. The homeostatic compensatory roles of GPCRs become especially important in diseased myocardium, where the activity of the G_s-coupled receptors tends to be decreased, and those of G_i-coupled and G_q/11-coupled receptors are less consistently altered (2, 3, 5). Any reduction in activity is mediated in the short term by phosphorylation of GPCRs during either agonist-dependent homologous desensitization or agonist-independent heterologous desensitization (5). In the longer term, as in cardiac disease states, reduced GPCR activity is also mediated by diminished receptor density, or downregulation, which is especially well defined for the -adrenoceptor (2, 26). While a number of causes for GPCR downregulation have been established (3), the potential role of diminished GPCR recovery after agonist-induced GPCR internalization has not been fully defined.

In substantially pressure-overloaded, hypertrophied, and failing myocardium, there is a striking increase in the density and stability of the cardiocyte microtubule network (10, 34), and these microtubules are heavily decorated by microtubule-associated protein 4 (MAP 4) (30). Given this observation, together with the facts that activated GPCRs are recycled within endocytic vesicles (5), that vesicular transport occurs at least partly via motor ATPases along microtubules (13, 16), and that microtubule decoration by structural microtubule-associated proteins inhibits this trafficking (4, 11), we wondered whether the GPCR downregulation seen in hypertrophied and failing myocardium might be based to some extent on the abnormalities of the extramyofilament cytoskeleton that we had observed.

As an initial simplifying approach to this specific question, we applied it here to a well-characterized model of GPCR dynamics, muscarinic acetylcholine receptor (mAChR) recovery in neuroblastoma cells (7, 19, 20, 22, 27) by examining the effects of MAP 4 decoration of microtubules on this process. Furthermore, as an approach to a more comprehensive question, we wished to use this model system to ask whether altered cytoskeleton-based intracellular trafficking might be a potential mechanism for the regulation of some facets of GPCR behavior in general.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MAP 4 antibody. To generate the MAP 4 antibody, we made a bacterial expression construct using the 1-740 NH₂-terminal residues of human MAP 4 (8). The recombinant protein, which had a hexahistidine tag inserted at the COOH terminus, was overexpressed in Escherichia coli, purified on a nickel-chelate affinity column, and submitted to SDS-PAGE. The purified protein band was excised, eluted from the gel, and sent to Lampire Biological Laboratories for preparation of a rabbit polyclonal MAP 4 antibody.

Cell culture. Murine neuroblastoma N1E-115 cells from American Type Culture Collection (passages 5-10) were grown in a 5% CO₂ atmosphere at 37°C on tissue culture dishes in DMEM supplemented with 10% fetal bovine serum, 4 mM L-glutamine, 4.5 g/l glucose, and 100 U/ml each of penicillin and streptomycin. Cells were seeded into six-well plates at a density of 10⁵ cells/35-mm well, grown for 3 days until 80-90% confluent, and maintained in serum-free medium thereafter.

[N-methyl-³H]scopolamine binding assays. Cell surface mAChR density was measured in N1E-115 cells via ligand binding assays using the hydrophilic, plasmalemma-impermeant muscarinic antagonist [N-methyl-³H]scopolamine (NMS; 78 mCi/mM) (6, 18). The binding assays were performed at 4°C to avoid both receptor endocytosis and reappearance of the receptors on the cell surface after internalization (14). Nonspecific binding was determined by the addition of 2 µM atropine, a high-affinity muscarinic receptor antagonist, to triplicate wells and represented <5% of total counts. This value was subtracted from the total binding to obtain specific plasmalemma-bound [³H]NMS.

Immunofluorescence confocal microscopy. N1E-115 cells were extracted in 1% Triton X-100 for 1 min and rinsed three times in microtubule stabilization buffer (2 mM EGTA, 0.1 mM EDTA, 1 mM MgSO₄, and 100 mM MES; pH 6.75) (25). The cells were fixed and stained as previously described (30).

Construction of recombinant adenoviruses. The pTG3602 plasmid system for generating replication-defective recombinant adenoviruses via homologous recombination with a bacterial system (9, 29) was used to construct adenoviruses for MAP 4 and -galactosidase (-Gal) overexpression in N1E-115 cells. The MAP 4 cDNA construct was generated by PCR using specific oligonucleotide primers and full-length human MAP 4 cDNA (a gift from J. C. Bulinski, Columbia University, Columbia, NY) as a template (4). The primer for the NH₂ terminus also contained a sequence for an epitope tag of amino acids 410-419 of human c-Myc. In addition, KpnI restriction sites were included at the end of each primer. The sequence of the full-length MAP 4 construct was confirmed by DNA sequencing. The KpnI fragment was subcloned into an adenovirus shuttle vector designed to produce a high level of constitutive expression by replacing E1 adenoviral sequences with a gene cassette consisting of the cytomegalovirus promoter element, the multiple cloning site, and the simian virus 40 polyadenylation signal. After homologous recombination of a NheI-ApaI fragment of this shuttle vector with adenovirus type 5 DNA whose EIA region was deleted (pTG3602), the recombinant adenovirus DNA containing the MAP 4 expression cassette was linearized and transfected into the human epithelial kidney (HEK) cell line HEK-293 using lipofectamine reagent. The adenovirus was plaque purified, expanded, and titered via the detection of visible plaque formation in HEK-293 monolayers.

Adenovirus-mediated gene transfer. N1E-115 cells were infected with MAP 4 adenovirus (AdMAP 4) in serum-free medium (2 ml/35-mm plate) at the indicated multiplicity of infection (MOI) for 24 h. The cells were rinsed in serum-free medium and incubated for a further 48 h to permit transgene expression. Companion dishes were infected at the same MOI with -Gal adenovirus (Ad-Gal), a similarly constructed adenovirus expressing the -Gal gene (29). A MOI range of 10-100 plaque-forming units (pfu)/cell of AdMAP 4 or Ad-Gal caused >90% infection after 48 h, as determined by immunofluorescence microscopy using antibodies to the c-Myc epitope tag on the transfected MAP 4 or to -Gal; there was no observable cytotoxicity.

Immunoblots. Cell lysates were prepared for immunoblotting as we previously described (30). The blots were incubated for 16 h with the primary antibody. After blots were incubated with biotinylated secondary antibody, specific protein bands were detected using avidin-biotinylated horseradish peroxidase in conjunction with enhanced chemiluminescence.

Data analysis. Means ± SE are given for the numerical data. Any statistical comparisons are specified in the individual figures.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Characteristics of the cell model. To establish the location of the mAChR in unstimulated N1E-115 neuroblastoma cells, the cells were exposed both to tetramethylrhodamine-Con A, a plasma membrane marker (Fig. 1A), and to an antibody to the M₂ subtype of the mAChR (Fig. 1B). Colocalization of the resultant signals to the plasma membrane is evident in Fig. 1, C and D.

View larger version (52K): [in this window] [in a new window]   Fig. 1.   The M2 muscarinic acetylcholine receptor (mAChR) is localized to the plasmalemma of neuroblastoma cells. The plasma membrane of N1E-115 cells was visualized via a 2-min incubation of the cells on ice with 10 µM concanavalin A followed by a wash of 3 times with PBS (A; red). M2 mAChRs were visualized in the same cells using C-18 M2 mAChR primary antibody (1:50) and FITC-conjugated goat polyclonal IgG secondary antibody (1:200) (B; green). Localization of the M2 AChR to the plasma membrane is shown by overlapping red and green signals (C; yellow), which are also superimposed on a modulation contrast image of the cell (D). Bar, 5 µm.

Role of microtubules and microfilaments in mAChR recovery. To determine whether microtubules and/or microfilaments are required for mAChR recycling, N1E-115 cells were treated with carbachol, a nonsubtype-selective mAChR agonist, to activate and internalize these receptors. Recovery of cell surface receptor density was then followed over time in the control state and after microtubule or microfilament depolymerization. The normal microtubule network in these cells (Fig. 2A) was reduced to a scattering of residual microtubules after a 2-h exposure to 1 µM colchicine (Fig. 2B). The normal actin microfilament architecture (Fig. 2C) was completely disrupted after a 2-h exposure to 1 mM cytochalasin B (Fig. 2D). The summary data in Fig. 2 show that incubation of N1E-115 cells with 2 mM carbachol for 16 h led to a decline in the density of cell surface mAChR to 24 ± 1% of that in control cells incubated without carbachol; mAChR density recovered to 80 ± 4% of the control value 9 h after carbachol withdrawal. When instead the microtubules were depolymerized before carbachol withdrawal, mAChR density recovered to only 35 ± 9% of control, as shown previously for another GPCR, the -adrenoceptor (21). In contrast, with microfilament disruption before carbachol withdrawal, receptor density recovered to 79 ± 7% of control, a value concordant with that for cells with an intact cytoskeleton. Thus microtubules have a critical role in the recovery of endocytosed mAChRs, whereas microfilaments do not.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Microtubules, but not microfilaments, are involved in M2 mAChR recovery. N1E-115 cells were plated on coverglasses, incubated for 1 day in culture medium, treated for 2 h with an equal volume of either 0.9% saline (A) or 1 µM colchicine (B), and processed using B-5-1-2 -tubulin primary antibody (1:200) and FITC-conjugated donkey anti-mouse IgG secondary antibody (1:200). Other N1E-115 cells were treated for 2 h with an equal volume of either DMSO vehicle (C) or 1 mM cytochalasin B (D), fixed with formaldehyde, blocked for 30 min in 1% BSA-PBS, incubated for 20 min at room temperature with 1 unit rhodamine phalloidin/slide, and washed 3 times with PBS. Bottom: N1E-115 cell cultures were preincubated with or without 2 mM carbachol for 16 h. Fresh medium with 2 mM carbachol and either 1 µM colchicine or 1 mM cytochalasin B was then added. After a 2-h incubation, the cells were washed with carbachol-free DMEM, and incubation was continued for 9 h to assay receptor recovery to the plasmalemma. [N-methyl-3H]scopolamine (NMS) binding was determined as described in MATERIALS AND METHODS. cpm, Counts per minute. Each point is the mean of 6 determinations. Statistical comparisons were by one-way ANOVA followed by Scheffé's S-procedure. Bar, 5 µm. * P = not significant (NS) for a difference from carbachol alone; P < 0.01 for a difference from recovery alone; P = NS for a difference from carbachol alone; §P = NS for a difference from recovery alone.

Adenovirus-mediated MAP 4 expression and microtubule decoration. Before the effects of microtubule decoration with MAP 4 on mAChR recovery were investigated, it was necessary to ascertain in neuroblastoma cells whether MAP 4 protein expressed by the transgene decorated the microtubules efficiently and to identify an adenoviral MOI sufficient for efficacious MAP 4 translation. Normal microtubule network architecture was retained in N1E-115 cells infected with Ad-Gal at a MOI of 100 pfu/cell (Fig. 3A). In cells infected with AdMAP 4 at a MOI of 100 pfu/cell, there was extensive microtubule network decoration by transfected human MAP 4 (Fig. 3B). The data shown in Fig. 3C, generated using antibodies to the c-Myc epitope tag on the transfected MAP 4 and to endogenous glyceraldehyde-3-phosphate dehydrogenase (GAPDH), show that robust and essentially equivalent expression of MAP 4 protein occurs in N1E-115 cells at an AdMAP 4 MOI range of 10-100. As shown in Fig. 4, with the use of a MAP 4 antibody in cells infected with either Ad-Gal or AdMAP 4, there was virtually no MAP 4 detected in the -Gal-infected cells. Thus the MAP 4 protein expressed at an AdMAP 4 MOI of 100 was almost exclusively the product of the transgene. Because no cytotoxicity was seen at the higher MAP 4 adenovirus concentration, an AdMAP 4 MOI of 100 was used for the studies that follow.

View larger version (46K): [in this window] [in a new window]   Fig. 3.   Microtubule-associated protein (MAP) 4 microtubule decoration and protein level are related to the titer of adenoviruses containing MAP 4 (AdMAP 4). N1E-115 cells were infected with AdMAP 4 at a multiplicity of infection (MOI) of 100 plaque-forming units (pfu)/cell for 24 h and allowed to express the protein for a further 48 h. c-Myc tag primary antibody (1:100) and FITC-conjugated secondary antibody (1:200) were used to visualize overexpressed MAP 4 via its c-Myc tag (green); B-5-1-2 -tubulin primary antibody (1:200) and Cy3-conjugated secondary antibody (1:100) were used to visualize microtubules (red). Comparison of a control cell (A) with an AdMAP 4-infected cell (B) shows overlapping immunoreactivity representing MAP 4 microtubule decoration solely in the latter. C: immunoblot analysis of samples from N1E-115 cells infected with AdMAP 4 over the indicated range of MOI; the blot was probed with c-Myc tag antibody and, after stripping, with glyceraldehyde- 3-phosphate dehydrogenase (GAPDH) (6C5) antibody (insets). The bar graph represents densitometric quantification of the immunoblot after GAPDH normalization. Bar, 5 µm.

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Level of endogenous versus AdMAP 4-mediated MAP 4 expression in neuroblastoma cells. N1E 115 cells were infected with either -galactosidase (-Gal)-containing adenoviruses (Ad-Gal; A and inset, lane 1) or AdMAP 4 (B and inset, lane 2) at a MOI of 100 pfu/cell. For the immunoblot, cells were harvested and homogenized in MAP 4 lysis buffer consisting of 2% NP-40, 100 mM Tris · HCl, 10 mM EGTA, 0.35 M NaCl, 1 mM Na3VO4, 0.5 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 2 µg/ml pepstatin A, 2 µM trans-epoxysuccinyl-L-leucylamido-(4-guanidino) butane, and 10 mM dithiothreitol. After incubation on ice for 20 min, the cell lysates were centrifuged at 16,000 g for 10 min at 4°C. The supernatants, after centrifugation a second time at 16,000 g for 30 min at 4°C, were mixed with an equal volume of 2× SDS sample buffer and boiled. An equal amount of total protein (20 µg) was loaded in each lane. Our MAP 4 antibody (see MATERIALS AND METHODS) was used to visualize MAP 4 in both the micrographs and immunoblot. There is minimal MAP 4 expression and minimal microtubule decoration by MAP 4 in Ad-Gal-infected cells. The opposite obtains for AdMAP 4-infected cells. Bar, 5 µm.

mAChR recovery. The studies to this point using a radiolabeled mAChR ligand clearly indicate a decrease in cell surface receptor density after agonist exposure. But it is also possible that this decrease is due to some extent to a change in mAChR affinity rather than being solely due to a change in mAChR number. First, for this reason, second, to establish visually agonist-induced mAChR internalization and subsequent recovery after agonist withdrawal, and, finally, to preclude significant nonspecific effects of adenoviral infection on this process, N1E-115 cells were infected with Ad-Gal at a MOI of 100 pfu/cell, allowed to express the transgene for 48 h, and then probed with mAChR antibody. At this point, the mAChRs were localized to the plasma membrane (Fig. 5, A and B). When instead the cells were exposed to carbachol for 16 h before fixation, the internalized mAChRs were distributed throughout the cytoplasm (Fig. 5, C and D). Finally, 9 h after the carbachol was withdrawn, mAChRs were once again present at the plasma membrane (Fig. 5, E and F). This response of the mAChR to agonist stimulation and withdrawal is identical to that observed in control N1E-115 cells not exposed to adenovirus (data not shown).

View larger version (59K): [in this window] [in a new window]   Fig. 5.   Effect of a muscarinic agonist on subcellular M2 mAChR distribution in neuroblastoma cells: overexpressed -Gal does not influence receptor trafficking. N1E-115 cells were infected with Ad-Gal at a MOI of 100 pfu/cell for 24 h and allowed to express the protein for a further 48 h. Cells were then treated with 2 mM carbachol, a mAChR agonist, for 16 h. A, C, and E: confocal micrographs of N1E-115 cells stained with C-18 M2 mAChR primary antibody (1:50) and FITC-conjugated secondary antibody (1:200) to illustrate M2 mAChR distribution before carbachol (A), immediately after carbachol treatment (C), and 9 h after carbachol withdraw (E). B, D, and F: modulation contrast images of the same cells overlaid with their fluorescence images. Membrane mAChRs internalized after carbachol treatment (C and D), but after agonist withdrawal they returned to the plasmalemma (E and F). Bar, 5 µm.

View larger version (62K): [in this window] [in a new window]   Fig. 6.   Effect of a muscarinic agonist on subcellular M2 mAChR distribution in neuroblastoma cells: overexpressed MAP 4 inhibits outward transport of internalized M2 mAChRs. N1E-115 cells were infected with AdMAP 4 at a MOI of 100 pfu/cell for 24 h and allowed to express the protein for a further 48 h. The cells were then treated with 2 mM carbachol, a mAChR agonist, for 16 h. A, C, and E: confocal micrographs of N1E-115 cells stained with C-18 m2 mAChR primary antibody (1:50) and FITC-conjugated secondary antibody (1:200) to illustrate M2 mAChR distribution before carbachol (A), immediately after carbachol treatment (C), and 9 h after carbachol withdraw (E). B, D, and F: modulation contrast images of the same cells overlaid with their fluorescence images. Membrane mAChRs internalized after carbachol treatment (C and D), just as in the cells overexpressing -Gal (Fig. 5). However, in contrast to the extensive outward transport of internalized M2 mAChRs seen after agonist withdrawal in N1E-115 cells overexpressing -Gal (Fig. 4, E and F), the cells overexpressing MAP 4 show here very little clearing of the M2 mAChR signal (E and F). Bar, 5 µm.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Extent of M2 mAChR recycling varies inversely with the level of overexpressed MAP 4. N1E-115 cells were infected with Ad-Gal at a MOI of 100 pfu/cell or with AdMAP 4 at a MOI of 0, 102, 101, 100, 101, or 102 pfu/cell for 24 h and allowed to express the proteins for a further 48 h. Cells were either not further treated or treated with 2 mM carbachol for 16 h. As shown in A, agonist exposure caused the number of cell membrane M2 mAChRs, measured as specific binding of [3H]NMS, to decrease by ~78% in cells infected with both the -Gal and MAP 4 viruses. As shown for the MAP 4-infected cells in A, the extent of mAChR recycling at 9 h after agonist withdrawal in AdMAP 4-infected cells varied inversely with the AdMAP 4 titer. Regression analysis showed a statistically significant inverse linear relationship between mAChR recycling and AdMAP 4 titer (B).

View larger version (47K): [in this window] [in a new window]   Fig. 8.   Rate of M2 mAChR outward transport after agonist-induced internalization is reduced by MAP 4 overexpression. N1E-115 cells were infected with either Ad-Gal or AdMAP 4 at a MOI of 100 pfu/cell. The cells were then treated with 2 mM carbachol for 16 h, washed 3 times, and allowed to recover for 0, 3, 6, or 9 h at 37°C before determination of [3H]NMS binding. A: the results show that overexpressed MAP 4 reduces the rate of cell surface mAChR recovery significantly and to a similar degree at each time point 3 h (* P < 0.01 for a difference between the Ad-Gal and AdMAP 4 groups via one-way ANOVA followed by Scheffé's S-procedure). B: for the immunoblots, identically treated cells were washed 3 times and incubated in DMEM with (bottom) or without (top) 20 µg/ml cycloheximide and then harvested at the times indicated. They were immunoblotted as described using C-18 M2 mAChR primary antibody.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Recovery of cell surface mAChR binding after carbachol treatment in nontransfected N1E115 cells with and without inhibition of protein synthesis. The cells were treated with 2 mM carbachol for 16 h, washed 3 times, and allowed to recover for 0, 1, 2, 3, 6, or 9 h at 37°C with or without 20 µg/ml cycloheximide. Cell surface mAChRs were then determined as described before. Cell surface mAChR density in control cells (1,389 ± 67 cpm/dish) treated with neither carbachol nor cyclohexmide was taken as 100% (* P < 0.05 for a difference from the group treated with carbachol alone via a repeated-measures ANOVA). The early phase of receptor recovery is protein synthesis independent; the later phase of receptor recovery is protein synthesis dependent.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The intent of this study was to address the question of whether extensive microtubule decoration with a structural MAP would inhibit the recovery of an activated GPCR. For the cell-receptor combination studied here, this question was answered in the affirmative. That is, the neuroblastoma cell plasmalemmal mAChRs (Fig. 1) internalize upon agonist exposure and then return to the cell surface upon agonist withdrawal (Fig. 2). This recovery process is microtubule dependent but not microfilament dependent (Fig. 2). Decoration of the microtubules with MAP 4 (Figs. 3 and 4) inhibits this recovery process (Figs. 5 and 6) in terms of both the extent of receptor recycling (Fig. 7) and the rate at which this recovery occurs (Fig. 8). This altered recovery is due to the exogenous MAP 4 (Fig. 4), because, as would be expected for the undifferentiated neuroblastoma cells used here (23), expression of native MAP 4 is virtually undetectable, but there is robust expression of the MAP 4 transgene and microtubule decoration by its protein product.

This data set is specifically informative of M₂ mAChR properties. That is, although N1E-115 cells express both the M₁ and M₂ mAChR subtypes (28), we found here in N1E-115 cells that, as in other cell types (17), the M₂ but not the M₁ subtype exhibits internalization in response to agonist exposure (data not shown). This differing internalization behavior of the two mAChR subtypes is based on five amino acids that are present in the M₂ subtype but absent from the M₁ subtype and that are essential for agonist-induced dynamin-independent receptor internalization (31). In contrast to other GPCRs that are transported along microfilaments but not along microtubules (36), we show here that M₂ mAChRs are transported along microtubules, as has again been found in another cell type (27), but not along microfilaments.

More interesting in the context of our original question, however, are the effects of MAP 4 microtubule decoration on GPCR trafficking. The recovery of cell surface M₂ mAChRs after agonist exposure tends to be rather slow (6), with 50% recovery requiring ~8 h in N1E-115 cells (32). This recovery has two phases (6). The first phase of protein synthesis-independent recovery reflects the recycling of internalized M₂ mAChRs to the cell surface. In the present data, Fig. 8 shows that MAP 4 overexpression impedes the reappearance of M₂ mAChRs on the cell surface at the earlier time points of 3 and 6 h, such that receptor recycling was inhibited. The second phase of protein synthesis-dependent recovery accounts for the balance of the reappearance of M₂ mAChRs on the cell surface after agonist exposure. In the present data, Fig. 9 shows that inhibition of protein synthesis does, in fact, inhibit the later but not the earlier phase of mAChR recovery. In Fig. 8, it is apparent that MAP 4 overexpression also inhibits mAChR recovery to the same extent throughout this later phase. Thus both phases of the trafficking of mAChRs to the cell surface after agonist-induced receptor internalization were inhibited both by microtubule depolymerization and by MAP 4 decoration of the microtubules. Such recovery required an intact microtubule network, and the extent of recovery was inversely related to the level of expression of the MAP 4 transgene.

The M₂ mAChR internalizes after agonist stimulation via mechanisms that are common to a number of other GPCRs (12): agonist binding to a GPCR causes both the relevant signaling event via heterotrimeric G protein interaction and receptor desensitization via GPRC kinase-mediated receptor phosphorylation that leads to -arrestin binding to the receptor, and -arrestin interaction with clathrin-coated vesicles then leads directly to receptor translocation to an endosomal compartment wherein the GPCR is dephosphorylated. If not degraded, the GPCR is then transported outward to the plasmalemma in its resensitized native state via mechanisms that have heretofore been poorly understood (12). It is the extent of the microtubule dependence of this outward transport of the M₂ mAChR that has been of particular interest here. In this regard, two aspects of microtubule properties are relevant.

First, it is now well established that microtubule dynamics are controlled in part by structural MAPs (15). MAP 4, which in contrast to other MAPs is expressed ubiquitously, has a COOH-terminal microtubule-binding domain consisting of three to five repeats of an 18-amino acid motif that is well conserved among structural MAPs (8, 15). We show here that overexpressed MAP 4 binds to microtubules in the N1E-115 cell, and it is clear that microtubule-bound MAP 4 causes a marked increase in microtubule stability (24). However, as shown in another study of microtubule-based transport (4), we found that microtubule stabilization by taxol had no effect on M₂ mAChR recovery (data not shown). Thus MAP 4 effects on microtubule stability do not appear to have an important role in the inhibitory effects of MAP 4 on AChR recovery that we see in N1E-115 cells.

Second, it is also well established that microtubules serve as tracks for intracellular transport via motor ATPases of the kinesin and dynein superfamilies (1, 16). Decoration of microtubules by tau, a neuronal MAP, decreases the attachment frequency of these motor proteins to the microtubule (33), likely via steric inhibition (4), such that the run length of these motor proteins along the microtubule, and thus net vectorial transport of their associated cargoes, are decreased. Especially pertinent in the context of the present study of GPCR outward transport, this effect is most pronounced for kinesin-dependent microtubule plus-end directed transport to the cell periphery (33).

In summary, recovery of the activated mAChR is microtubule dependent, and it is inhibited by MAP 4 decoration of the microtubules. Of further general interest is the question of whether this behavior applies to other GPCRs in other cell types, either in vitro or more importantly in vivo. That is, altered cytoskeleton-based intracellular trafficking, whether via microtubules or via other cytoskeletal polymers, may well be one mechanism for the regulation of GPCR behavior in general. Of further specific interest for our understanding of cardiac hypertrophy and failure, the present data provide a rational basis for asking the question of whether the GPCR downregulation seen there might be related in part to the dense, stabilized, and heavily MAP 4-decorated cardiocyte microtubule network that we find in pressure-overload cardiac hypertrophy and failure.