J Biol Chem, Vol. 274, Issue 43, 30550-30556, October 22, 1999

Casein Kinase II Activity Is Required for Transferrin Receptor Endocytosis^*

Laura Fraser Cotlin, Masood A. Siddiqui, Fiona Simpson^§, and James F. Collawn^¶

From the  Department of Cell Biology, University of Alabama at Birmingham, MCLM 392, UAB Station, Birmingham, Alabama 35294-0005 and the ^§ Department of Cell Biology, The Scripps Research Institute, La Jolla, California 92037

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The effect of protein kinase inhibitors on transferrin receptor (TR) internalization was examined in HeLa, A431, 3T3-L1 cells, and primary chicken embryo fibroblasts. We show that TR endocytosis is not affected by tyrosine kinase or protein kinase C inhibitors, but is inhibited by one serine/threonine kinase inhibitor, H-89. Inhibition occurred within 15 min, was completely reversible after H-89 withdrawal, and was specific for endocytosis rather than pinocytosis since a TR mutant lacking an internalization signal was not affected. Interestingly, H-89 also inhibited the internalization of a TR chimera containing the major histocompatibility complex class II invariant chain cytoplasmic tail, indicating that the effect was not specific for the TR. Since H-89 inhibits a number of kinases, we employed a permeabilized cell endocytosis assay to further characterize the kinase. In permeabilized 3T3-L1 cells, addition of pseudosubstrate inhibitor peptides of casein kinase II (CKII) blocked TR internalization by more than 50%, whereas pseudosubstrates of cyclic AMP-dependent kinase A, protein kinase C, and casein kinase I had no effect. Furthermore, addition of purified CKII to the cell-free reactions containing CKII pseudosubstrates reversed the endocytosis block, suggesting that CKII or a CKII-like activity is required for constitutive endocytosis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The transferrin receptor (TR)¹ binds the serum iron transport protein transferrin (Tf), internalizes through clathrin-coated pits, and facilitates Tf iron release in the sorting endosome. Efficient TR internalization requires a cytoplasmic tail tyrosine-containing motif, Tyr-Thr-Arg-Phe (1, 2). Studies by Ohno et al. (3) using yeast two-hybrid analysis demonstrate that the TR cytoplasmic tail signal interacts with one of the four subunits of the AP-2 adaptor complex, the µ₂ chain. This interaction provides a mechanism for promoting TR clustering in clathrin-coated pits and subsequent internalization. The AP-2 adaptor complex is also required for clathrin recruitment (4-6) and lattice assembly (7) and, together with its direct interaction with receptor cytoplasmic tails, links the cell surface receptors to the clathrin-based endocytic machinery (reviewed in Refs. 8 and 9).

Despite the extensive characterization of many of the proteins involved in endocytosis, little is known about how the clathrin-based endocytic machinery is regulated. What is evident, however, is endocytosis via clathrin-coated pits is blocked during mitosis (10), starting at prophase and continuing through telophase (11, 12). In A431 cells, the mitotic block appears to arrest clathrin assembly at various stages of invagination (13), suggesting that the continuous activity of an enzyme (or enzymes) is required for vesicle formation (reviewed in Ref. 14). Using in vitro reconstitution assays, mitotic cytosol has been shown to inhibit invagination of clathrin-coated pits and one of the factors responsible is cdc2 kinase (15).

The role of kinases in receptor trafficking has been demonstrated in studies on the asialoglycoprotein receptor which show that a tyrosine kinase is required for endocytosis (16) and for phosphorylation of the receptor cytoplasmic tail (17). In vitro studies have also shown that components of the clathrin-coated vesicles can be phosphorylated by vesicle-associated kinases, but the functional significance of these results remains unclear (18). Furthermore, a number of cell surface proteins including CD4 (19), Fc  receptor (20), CD3 (21), _v5 integrin receptor (22), and TR (23) are rapidly endocytosed following phorbol ester-mediated stimulation of protein kinase C (PKC). However, the PKC phosphorylation site in TR, Ser²⁴, is not required for this effect (24, 25), suggesting that PKC is acting at the level of the internalization machinery rather than the receptor itself (26, 27). Additionally, cyclic AMP-dependent protein kinase A has also been reported to be required for the internalization of the urokinase-type plasminogen activator (28), indicating that a number of kinases are somehow involved in receptor internalization.

In yeast, casein kinase I (CKI) has been shown to be required for constitutive endocytosis of the pheromone receptor Ste3p (29). Although a similar function for CKI has not been described in mammalian cells, one intriguing model proposed (29) is that CKI activity regulates endocytosis by phosphorylating the adaptor complex, thereby activating the complex to promote coat assembly and endocytosis. However, since the adaptor complex in yeast is homologous to the recently identified AP-3 complex found at the endosome (30, 31), it remains possible that CKI is required for vacuolar targeting rather than endocytosis.

In the present study, we examined the role of protein kinases in TR internalization. Using pharmacological approaches and in vitro assays that reconstitute endocytosis, we found that compounds that inhibit tyrosine kinases and a number of serine/threonine kinases including PKA, PKC, and CKI have no effect on TR endocytosis. In contrast, we provide experimental evidence that suggests casein kinase II, or a closely related kinase, is required.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cells and Reagents-- HeLa and 3T3-L1 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (HyClone, Logan, UT), 2 mM L-glutamine, penicillin, and streptomycin. A431 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (Atlanta Biologicals, Norcross, GA) and 2 mM L-glutamine. Chicken embryo fibroblasts (CEFs) were grown in Dulbecco's modified Eagle's medium supplemented with 1% chicken serum and 1% fetal bovine serum (Atlanta Biologicals, Norcross, GA), 2% (v/v) tryptose phosphate broth (Difco, Detroit, MI), 2 mM L-glutamine, penicillin, and streptomycin.

Biotinylated transferrin (B-Tf) was prepared as described previously (32) using biotin-XX, SSE (6-((6-biotinoyl)amino)hexanoic acid, sulfosuccinimidyl ester, sodium salt) obtained from Molecular Probes (Eugene, OR). Pseudosubstrate peptides for protein kinase C (RFARKGALRQKNVHEVKN) (33), casein kinase I (DDDEESITRR) (34), and casein kinase II (RRREEETEEE) (35) were synthesized by Quality Controlled Biochemicals (Hopkinton, MA). The protein kinase A pseudosubstrate peptide TYADFIASGRTGRRNAI (36) was obtained from Calbiochem (La Jolla, CA). Control peptides containing the same amino acid composition but lacking the consensus phosphorylation sites were also prepared for CKI (STIRRDDDEE) and CKII (EEEEEERRRT) by Quality Controlled Biochemicals (Hopkinton, MA). Peptides were prepared as 20 mM stocks in KSHM buffer (100 mM potassium acetate, 85 mM sucrose, 20 mM Hepes, 1 mM magnesium acetate), the pH was adjusted to 7.4, and the peptides were stored at 4 °C. Genistein and H-89 were obtained from Biomol (Plymouth Meeting, PA). Genistein was prepared fresh for each experiment and H-89 was stored as a 48 mM stock solution in dimethyl sulfoxide at 20 °C. Herbimycin A was obtained from Life Technologies, Inc. (Gaithersburg, MD) and stored as a 1.75 mM stock solution in dimethyl sulfoxide at 20 °C. All other inhibitors were purchased from Calbiochem (La Jolla, CA). Calphostin incubation was performed in the presence of fluorescent light as described by the manufacturer (Calbiochem, La Jolla, CA).

Expression of Wild-type and Mutant Transferrin Receptors, and an MHC Class II Invariant Chain-Transferrin Receptor (Ii-TR) Chimera in CEFs-- Wild-type, S24A, 3-18, 29-59 TR, 3-59 TR, and Ii-TR chimeras were expressed in CEF as described previously (1, 37).

Internalization Assay in Intact Cells-- Cells were plated in triplicate at a density of 7.5 × 10⁴ cells/cm² in 24-well tissue culture plates 24 h before the assay (Costar Corp., Cambridge, MA). Cells were then incubated in serum-free media for 1 h at 37 °C. Inhibitors were included during this incubation period for times ranging from 15 min to 1 h. Internalization rates were determined using the IN/SUR method (38) as described previously (37).

ATP Analysis-- ATP levels were determined using an ATP detection kit (Calbiochem). Briefly, HeLa cells were incubated in the absence or presence of inhibitors for 60 min. An energy-depleting system consisting of 5 mM 2-deoxyglucose and 5 mM sodium azide was used as a control. Treated cells were placed on ice, scraped at 4 °C in HEPES buffer (25 mM HEPES, 10 mM MgCl₂, and 0.02% NaN₃), and aliquots were transferred to a cuvette and the volume adjusted to 200 µl with HEPES buffer. 200 µl of releasing agent was added to each sample and the reaction was started by adding 100 µl of the luciferin-luciferase enzyme mixture. Luminescence was measured at 560 nm on a Turner Designs Luminometer (TD-20/20) and the ATP concentrations were determined from a standard calibration curve. The protein concentration of the lysate was determined using the Bradford Assay (Bio-Rad) and total ATP was measured as micromoles of ATP per mg of protein lysate.

Rat Brain Cytosol Preparation-- Rat brains snap-frozen in liquid nitrogen were obtained from Pel-Freeze (Rogers, AR), sliced into small pieces, and homogenized in a Dounce homogenizer containing an equal volume of KSHM buffer and protease inhibitors (Complete mini-inhibitors, Roche Diagnostic Corp. (Indianapolis, IN)). The homogenate was centrifuged for 20 min at 11,200 rpm in a Beckman J20 rotor, and the supernatant was centrifuged for an additional 45 min at 39,000 rpm using a Beckman Ti75 rotor. The resultant supernatant was snap-frozen in liquid nitrogen in 250-µl aliquots and stored at 80 °C. Stock cytosol protein concentrations ranged from 15 to 30 mg/ml.

Permeabilized Cell Assay for Transferrin Receptor Endocytosis-- Permeabilized cells were prepared and endocytosis assays were preformed as described (32, 39). Permeabilized 3T3-L1 cells were used in place of A431 cells and were prepared by submerging plates in liquid nitrogen before scraping. After cytosol depletion, 10 µl of cells were incubated with pseudosubstrates or inhibitors for 10 min before addition to the reaction mixtures. Briefly, incubations were performed for 20 min at 37 °C in 40-µl reaction volumes containing an ATP-regenerating system (1 mM ATP, 1 mM GTP, 8 mM creatine phosphate, and 40 units/ml creatine phosphokinase), 6 mg/ml cytosol, and 2 µg/ml B-Tf and 0.2% bovine serum albumin in KSHM assay buffer. Reactions were stopped on ice, cells were pelleted in a microcentrifuge at 4 °C, and the supernatants were carefully removed. Control reactions contained an ATP-depleting system (5 mM glucose and 40 units/ml hexokinase) in place of the ATP-generating system and 6 mg/ml bovine serum albumin in place of the cytosol.

Measurement of Avidin Inaccessibility-- The cell pellets were resuspended in avidin (50 µg/ml in 0.2% bovine serum albumin in KSHM buffer) and processed as described (32, 39). Results are presented as the total cell-associated B-Tf (determined from cells not treated with avidin) endocytosed in an ATP- and cytosol-dependent manner.

In Vitro Phosphorylation Reactions-- 200 µg of dephosphorylated casein (Sigma) was incubated with either 62.5 ng of CKII (Calbiochem) or 28 ng of CKI (Promega) and 20 µM [-³²P]ATP (NEN Life Science Products Inc.) in a total reaction volume of 40 µl of KSHM buffer. Inhibitors and pseudosubstrates were included at various concentrations, as specified in the figure legends. The phosphorylation reactions were performed at room temperature for 10 min and terminated with the addition of 10 µl of SDS sample buffer (1 M Tris-HCl, 50% glycerol, 10% SDS, 0.5% -mercaptoethanol, 1% bromphenol blue, pH 6.8). Samples were analyzed by SDS-polyacrylamide gel electrophoresis, and casein phosphorylation was quantitated on a PhosphorImager (Molecular Dynamics).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

H-89 Inhibits TR Endocytosis-- To determine the role of kinases in TR internalization, endocytosis was monitored in HeLa and A431 cells after pretreatment with tyrosine kinase and PKC inhibitors. Using the IN/SUR method for measuring the internalization rate (38), TR endocytosis was found to be rapid in both HeLa and A431 cells under control conditions (k_e = 0.123 min¹ and 0.99 min¹, respectively, Table I) and comparable to TR internalization rates measured in other cells types (37). After pretreatment of the HeLa and A431 cells at 37 °C for 1 h or more with a number of tyrosine kinase and PKC inhibitors, the results indicate that TR endocytosis was unaffected (Table I). Interestingly, however, when the genistein concentration was increased to 500 µM (the concentration required to block asialoglycoprotein receptor endocytosis (16)), 5-fold higher than is normally used to inhibit tyrosine kinases (40), TR endocytosis was inhibited by more than 50% (data not shown). This suggests that a kinase, not necessarily a tyrosine kinase, was required for TR endocytosis.

When a serine/threonine kinase inhibitor, H-89, was tested, TR endocytosis was dramatically inhibited in both HeLa and A431 cells (Fig. 1). After as little as a 15-min pretreatment with 150 µM H-89, TR internalization decreased by ~50% (average of 47 and 55% in HeLa and A431, respectively). Inhibition of endocytosis was confirmed in steady-state distribution assays at 37 °C in HeLa cells as well, which demonstrated that the TR surface expression increased from 20.4 ± 0.2 to 38.8 ± 6.1% (mean ± S.D. from three experiments). Calculation of the estimated internalization rate from the steady-state distribution (1) suggests that after H-89 treatment, the TR internalization rate was 44% of the untreated control, indicating that recycling was not affected.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   H-89 inhibits transferrin receptor internalization in HeLa and A431 cells. Equivalent numbers of HeLa (A) or A431 (B) cells were incubated in serum-free media (Control) or serum-free media containing 50 or 150 µM H-89 for 15 min at 37 °C. The media was removed and the cells were incubated with prewarmed (37 °C) 125I-labeled Tf for the indicated times. The amount of internalized (Internal Tf) and surface-associated (Surface Tf) radiolabel was determined as described under "Experimental Procedures." Data are plotted using the IN/SUR method, in which the slope of the line equals the endocytic rate constant, ke (38). The data represent the mean ± S.E. of four independent experiments for each cell type.

Inhibition of TR Endocytosis by H-89 Is Reversible-- To confirm that H-89 was not indirectly inhibiting TR internalization by altering cell viability, the capacity of the cells to recover after H-89 treatment was tested. In this experiment, HeLa cells were treated for 1 h in 150 µM H-89, after which the media was exchanged with fresh inhibitor-free media every hour. After a 4-h recovery period, TR internalization was then compared with untreated cells. As shown in Fig. 2A, TR internalization returned to 116% of control values, demonstrating that the H-89 inhibition was reversible. To assess the possible cytotoxic effects of H-89 using another approach, we measured cellular ATP concentrations immediately after a 1-h pretreatment in 150 µM H-89. As shown in Fig. 2B, the treated cells had 95% of the control levels of ATP. Taken together, the results suggested that H-89 treatment was not causing a general cytotoxic effect that indirectly inhibited TR endocytosis. As a further control, we treated cells with 5 mM 2-deoxyglucose and 5 mM sodium azide and found that intracellular ATP levels dropped 55%, confirming that changes in ATP levels were easily monitored (Fig. 2B).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   H-89 inhibition of transferrin receptor endocytosis is reversible. A, equivalent numbers of HeLa cells were incubated at 37 °C in media (Control) or media containing 150 µM H-89 (H-89) for 1 h. The media was then removed and the cells were incubated in prewarmed 125I-labeled Tf for the indicated times. For the recovery experiment, the cells were incubated in complete media (without inhibitor) for an additional 4 h after H-89 treatment before incubation in 125I-labeled Tf (H-89 Recovery). The amounts of internalized (Internal Tf) and surface-associated (Surface Tf) radiolabel were determined as described under "Experimental Procedures." Data are plotted as described in the legend Fig. 1. The data represent the mean ± S.E. of three independent experiments. B, parallel cultures were treated using the same conditions described in A and analyzed for ATP levels after a 1-h H-89 treatment or an energy depleting system (5 mM 2-deoxyglucose and 5 mM sodium azide) as described under "Experimental Procedures." Total ATP is described as micromoles of ATP/mg of protein from the cell lysate. The data represent the mean ± S.E. of three independent experiments.

H-89 Inhibits the Internalization of a MHC Class II Invariant Chain-TR Chimera-- To determine if H-89 treatment also affected TR internalization in nontransformed cells, endocytosis was examined in primary CEFs. This experimental system also allowed us to test human TR mutants and chimeras (Fig. 3A) in order to determine the structural features of the TR cytoplasmic tail required for H-89 inhibition. Pretreatment of cells expressing the wild-type TR with 150 µM H-89 inhibited TR endocytosis by more than 80% (Fig. 3B). The only phosphorylation site in the TR cytoplasmic tail is serine 24, a PKC phosphorylation site (41). To confirm that this site was not important for the H-89 inhibition, internalization assays were performed on CEF expressing a mutant TR lacking this site (Fig. 3A). As shown in Fig. 3C, S24A was inhibited to the same extent as wild-type TR, suggesting that the presence of the phosphorylation site on the receptor was not required for the inhibitory effect.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   H-89 inhibits endocytosis of the TR and a MHC class II invariant chain-TR chimera, but not a TR mutant lacking an internalization signal. A, schematic diagram of the wild-type (WT) TR, S24A TR, 3-18, 29-59 TR, an Ii-TR chimera, and 3-59 TR. The amino acid sequences of the cytoplasmic tails (CT) of each are shown. All of the constructs contain the wild-type TR transmembrane (TM) and extracellular domains. Constructs are referred to in the text by the corresponding names shown at the left. B, CEFs expressing wild-type TR were incubated at 37 °C in serum-free media (Control) or serum-free media containing 150 µM H-89 for 15 min (H-89). The media was removed and the cells were incubated in prewarmed 125I-labeled Tf for the indicated times. The amounts of internalized (Internal Tf) and surface-associated (Surface Tf) radiolabel were determined as described under "Experimental Procedures." Data are plotted as described in the legend to Fig. 1. A representative experiment of three is shown. C, CEFs expressing S24A, 3-18, 29-59 TR, Ii-TR, and 3-59 TR were tested as described in B and the internalization rates were determined with and without H-89 treatment. The data represent the mean ± S.E. of three or more independent experiments. The only construct not affected by H-89 treatment was the TR lacking an internalization signal, 3-59.

To determine if the H-89 inhibition was specific for the TR, we tested two additional mutants, a TR cytoplasmic tail deletion mutant, 3-18, 29-59 TR, that lacks most of the cytoplasmic tail but contains the internalization signal (1), and a TR chimera that contains the MHC class II invariant chain cytoplasmic tail in place of the TR cytoplasmic tail (Invariant chain-TR, Fig. 3A) (37). As shown in Fig. 3C, endocytosis of both the deletion mutant and the MHC class II invariant chain chimera were dramatically inhibited by H-89 treatment, suggesting that the wild-type TR is not the only cell surface molecule affected. Furthermore, since the MHC class II invariant chain contains a di-leucine internalization signal rather than a tyrosine-based signal (37, 42), this indicates that receptors containing either class of internalization signal are equally inhibited by H-89 treatment.

One possible concern with the H-89 inhibition was that the H-89 was having an indirect effect on membrane dynamics, in which case pinocytosis would be affected as well. To test for this, we monitored the internalization of a TR mutant that lacked the endocytosis signal, 3-59 TR (Fig. 3A), after H-89 treatment. The results in Fig. 3C show that internalization of the 3-59 TR mutant was not affected by this treatment, indicating that inhibition was not due to nonspecific membrane effects.

H-89 Inhibits TR Internalization in Permeabilized 3T3-L1 Cells-- H-89 inhibits a number of kinases including PKA, PKC, CKI, and CKII by competing for the ATP-binding site (43). To determine which kinase was being inhibited, we tested pseudosubstrate peptide inhibitors of several serine/threonine kinases in an in vitro assay that reconstitutes endocytosis (32, 39, 44). Upon addition of exogenous cytosol and an ATP-regenerating system, B-Tf becomes sequestered in deeply invaginated coated pits and vesicles. Under these conditions, the B-Tf is inaccessible to avidin and is scored as internalized Tf (32).

For the initial characterization, we confirmed that H-89 inhibited B-Tf internalization in permeabilized 3T3-L1 cells. At 150 µM H-89, B-Tf uptake was inhibited by 38.1 ± 4.9% (mean ± S.D). At 300 and 500 µM, B-Tf internalization was inhibited by 66.2 ± 5.8 and 95.7 ± 3.5%, respectively (Fig. 4A). Next, we tested TR endocytosis after H-89 treatment in intact 3T3-L1 cells using the standard IN/SUR assay and ¹²⁵I-labeled Tf. As shown in Fig. 4B, TR endocytosis was blocked by ~54% after a 15-min pretreatment with 150 µM H-89, demonstrating that H-89 inhibited TR endocytosis in 3T3-L1 cells in both assays. Since the IC₅₀ of H-89 for CKII is 137 µM (43) (compared with 0.048 µM for PKA, 32 µM for PKC, and 38 µM for CKI (43)), similar to the concentration that inhibited TR internalization in HeLa and A431 cells, in vitro phosphorylation reactions were performed in the presence and absence of H-89 using purified CKII and casein as the substrate (Fig. 4C). The results show that H-89 at concentrations of 150, 300, and 500 µM inhibited CKII activity 43, 56, and 62%, respectively. This indicated that the loss of CKII activity in the in vitro phosphorylation assay paralleled the loss of B-Tf internalization after H-89 treatment, suggesting a possible link.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   H-89 inhibits transferrin receptor internalization in 3T3-L1 cells in vitro and in vivo. A, H-89 inhibits TR internalization in vitro. The sequestration of B-Tf was measured in perforated 3T3-L1 cells under standard assay conditions as described under "Experimental Procedures" in the presence of increasing concentrations of CKII inhibitor, H-89. The sequestration of B-Tf into constricted-coated pits and coated vesicles was detected by the inaccessibility to avidin. The data represent the mean ± S.E. of five experiments. B, H-89 inhibits TR internalization in vivo. Equivalent numbers of 3T3-L1 cells were incubated in serum-free media (Control) or serum-free media containing 150 µM H-89 for 15 min at 37 °C. The media was removed and the cells were incubated with prewarmed (37 °C) 125I-labeled Tf for the indicated times. The amount of internalized (Internal Tf) and surface-associated (Surface Tf) radiolabel was determined as described under "Experimental Procedures." Data are plotted as described in the legend to Fig. 1 and represent the mean ± S.E. of three independent experiments. C, H-89 inhibits casein kinase II phosphorylation of casein in vitro. Phosphorylation reactions were performed as described under "Experimental Procedures" in the presence of increasing concentrations of H-89. Quantitation of casein phosphorylation levels is shown below and was determined as described under "Experimental Procedures." The data are representative of two independent experiments.

Pseudosubstrates of CKII Block TR Internalization in Permeabilized 3T3-L1 Cells-- Next, we tested the ability of pseudosubstrate peptide inhibitors of PKA, PKC, CKI, and CKII to block B-Tf internalization in permeabilized 3T3-L1 cells. During a 10-min preincubation period when cytosol, ATP, GTP, and an ATP-regenerating system are added, we introduced pseudosubstrate peptide inhibitors of PKA, PKC, CKI, or CKII at 2-fold higher concentrations than the IC₅₀ of the peptides for the kinases. As shown in Fig. 5A, the PKA, PKC, and CKI pseudosubstrate peptides had little effect on B-Tf endocytosis, whereas the CKII peptide inhibited endocytosis by more than 70%. Comparison of the pseudosubstrate peptide to a control peptide containing the same amino acid composition but lacking the consensus site for CKII, indicated that the pseudosubstrate peptide effect was specific and significantly different from the control peptide (p < 0.05 at 1 mM peptide), although the control peptide did inhibit B-Tf uptake. In fact, the control peptide inhibited B-Tf uptake by 29, 36, and 51% at 0.5, 1, and 2 mM concentrations, respectively (Fig. 6B), indicating that although it was not as effective as the pseudosubstrate peptide, it still inhibited B-Tf endocytosis to a significant degree (p < 0.05 at 0.5 mM peptide concentration).

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Peptide pseudosubstrates of casein kinase II inhibit transferrin receptor internalization in perforated 3T3-L1 cells. A, the sequestration of B-Tf was measured in perforated 3T3-L1 cells under standard assay conditions as described under "Experimental Procedures" in the presence of peptide pseudosubstrates for PKA, PKC, CKI, and CKII. The concentration used is 2-fold higher than the IC50 for each peptide. The data represents the average ± S.E. of three independent experiments. The CKII pseudosubstrate was the only one significantly different from the control (***p < 0.001). B, the sequestration of B-Tf was measured as described in A in the presence of increasing concentrations of pseudosubstrate and control peptides for CKII. Control peptides contain the same amino acid composition but lack the consensus phosphorylation sites (see "Experimental Procedures"). The sequestration of B-Tf into constricted-coated pits and coated vesicles was detected by the inaccessibility to avidin. The data represent the mean ± S.E. of three experiments. The IC50 for the CKII pseudosubstrate is 0.5 mM (35). Comparison between the pseudosubstrate and control peptides indicated significant differences (*, p < 0.05; and **, p < 0.01).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Pseudosubstrates of casein kinase II inhibit CKII phosphorylation of casein. Phosphorylation reactions were performed as described under "Experimental Procedures" in the presence of increasing concentrations of CKII pseudosubstrates and CKII control substrates. Control peptides contain the same amino acid composition but lack the consensus phosphorylation sites. A, in vitro phosphorylation of casein by CKII in the presence of increasing concentrations of pseudosubstrate peptides. B, casein phosphorylation in A was quantitated on Molecular Dynamics ImageQuant as described under "Experimental Procedures." Gel and data are representative of three independent experiments.

To confirm that the CKII pseudosubstrate inhibited CKII activity, we performed in vitro phosphorylation reactions using the CKII pseudosubstrate and control peptides in reactions containing purified CKII. As shown in Fig. 6, both the pseudosubstrate and control peptides inhibited CKII phosphorylation of casein in a dose-dependent manner. More importantly, the inhibition of TR internalization demonstrated in the cell-free assays using the CKII pseudosubstrate and control peptides correlated well with the in vitro inhibition of casein phosphorylation by CKII. This suggested the CKII control peptide inhibited B-Tf endocytosis by inhibiting CKII activity; however, 4-fold less effectively than the pseudosubstrate (compare Figs. 5B and 6B; i.e. ~4-fold more control peptide was required for the same level of inhibition as the pseudosubstrate). In contrast, while the CKI pseudosubstrate peptide inhibited CKI phosphorylation of casein (not shown), this peptide had no effect on B-Tf endocytosis (Fig. 5A), suggesting that CKI activity was not required for TR endocytosis.

And finally, to demonstrate that the inhibition by the CKII pseudosubstrate peptide was specific, we repeated the inhibition experiment in the presence of exogenously added purified CKII. As can be seen in Fig. 7A, addition of CKII restored TR endocytosis in a dose-dependent manner. Comparing CKII addition to addition with CKI and albumin at the highest concentration tested (Fig. 7B), showed that the recovery of TR endocytosis with CKII addition was specific (% Tf uptake increased from 26 to 45% after CKII addition; p < 0.02), confirming that the loss of TR endocytosis in the presence of the CKII pseudosubstrate was due to the loss of CKII activity.

View larger version (34K): [in this window] [in a new window]   Fig. 7.   Addition of purified CKII to perforated cells reverses the inhibitory effect of the CKII pseudosubstrates on TR internalization. A, the sequestration of B-Tf was measured in perforated 3T3-L1 cells under standard assay conditions as described under "Experimental Procedures" in the presence of 1 mM CKII pseudosubstrate alone or 1 mM CKII pseudosubstrate plus 50, 100, or 200 ng of purified CKII. Sequestration of B-Tf was detected by inaccessibility to avidin. B, sequestration of B-Tf was measured in the presence of 1 mM CKII pseudosubstrate alone or 1 mM CKII pseudosubstrate plus 200 ng of CKII, CKI, or bovine serum albumin. The data represents the mean ± S.E. of four experiments. Only addition of CKII was significantly different from the control (*, p < 0.02).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Regulation of Transferrin Receptor Endocytosis by a Kinase-- In this study we report that CKII or a CKII-like kinase regulates TR internalization. Supporting evidence comes from pharmacological and biochemical studies in four different cell lines. Using H-89, a serine/threonine kinase inhibitor, we demonstrated that loss of TR internalization (in vivo and in vitro) correlated with a loss of CKII activity in vitro. H-89 inhibition did not affect ATP levels or cell viability and the inhibitory effects were reversible, suggesting that we were monitoring a general regulatory mechanism rather than a nonspecific effect. Furthermore, we showed that endocytosis but not pinocytosis was being affected by this treatment, suggesting that the process being inhibited was clathrin-based endocytosis.

Although H-89 is not specific for CKII, its IC₅₀ for CKII is 137 µM (43), very close to the concentration required to inhibit TR endocytosis in HeLa, A431, and 3T3-L1 cells. Furthermore, both CKII pseudosubstrate peptides and H-89 inhibited TR endocytosis in an in vitro reconstitution assay, indicating that two different inhibitors with different mechanisms of action block TR endocytosis. The specificity of the phosphorylation site in the CKII pseudosubstrate peptide (35) suggests that the kinase being inhibited is either CKII or a closely related kinase with similar substrate specificity. As a final proof that the CKII pseudosubstrate was not inhibiting TR endocytosis in a nonspecific manner, we demonstrated that the addition of purified CKII to the permeabilized cell assay partially reversed the inhibitory effects of the CKII pseudosubstrate, confirming that CKII activity is required for efficient TR endocytosis.

A question raised by these studies is how TR internalization is regulated by phosphorylation. Our data suggests that phosphorylation of the cytoplasmic tail is not required since a deletion mutant that lacks the only phosphorylation site in the TR cytoplasmic tail, S24A, was internalized normally and inhibited to the same extent as the wild-type receptor in the presence of H-89. Even a chimera lacking the TR cytoplasmic tail was inhibited, suggesting that phosphorylation of some other component in this pathway is likely to be required for TR endocytosis.

Potential Phosphorylation Targets-- If the TR cytoplasmic tail is not the target of the CKII activity, what are the potential targets? Since the assays employed here monitor the early steps in the endocytic pathway, the likely targets are in the clathrin-coated pits themselves. Endocytosis is known to be inhibited during mitosis by cdc2 kinase (15), a finding of particular interest as cdc2 kinase is phosphorylated by CKII (45) and CKII is phosphorylated by cdc2 kinase (46). Furthermore, phosphorylation of coat proteins by clathrin-coated vesicle-associated kinases destabilizes the coat of isolated brain clathrin-coated vesicles, suggesting that coat assembly is regulated by phosphorylation (15, 47).

Previous studies demonstrate that phosphorylation of adaptor complexes inhibits adaptor association with clathrin (48), suggesting that phosphorylation reactions inhibit clathrin assembly. Although the  subunits are the principal target for serine phosphorylation, the µ chain is also phosphorylated (48) and has been shown to be a target of a kinase activity associated with the clathrin-coated vesicles (49). Interestingly, phosphorylation of µ chain by the vesicle-associated kinase did not affect its in vitro interaction with the internalization signal of TGN38 (49). Additionally, CKII has been shown to phosphorylate dynamin (50), synaptotagmin (51), syntaxin (51), and clathrin light chain (52), however, the functional significance of these reactions remains unclear.

Identity of the Kinase Involved in Endocytosis-- A number of studies have suggested the existence of a clathrin-associated vesicle kinase (53-57). Although the identity of the kinase and its physiological role have never been established, the vesicle-associated kinase that has been described is CKII-like in its target specificity. Furthermore, CKII has been shown to be associated with the plasma membrane (58-60). However, recent in vitro studies have shown that CKII does not phosphorylate µ2 (49), implying that the vesicle-associated kinase is not CKII, but perhaps a closely related kinase. In our studies, we cannot eliminate the possibility that the kinase involved in TR internalization is a related kinase with a specificity and sensitivity to H-89 similar to that of CKII.

We carefully examined the role of CKI in TR endocytosis since this kinase has been shown to be required for endocytosis in yeast (29, 61). Even though H-89 inhibits both CKI and CKII, and the pseudosubstrates for CKI and CKII are quite similar (acidic residues preceding or following serine/threonine residues, respectively), several experiments suggest that CKI is not involved in TR endocytosis. First, the IC₅₀ of H-89 for CKI is 38 µM (43). At 50 µM H-89, TR endocytosis was not affected in HeLa and only slightly affected in A431 cells. Second, in vitro phosphorylation reactions confirmed that the CKI pseudosubstrate peptide inhibited CKI phosphorylation of casein, yet this peptide, when tested at a 2-fold higher concentration than its IC₅₀, had no effect on TR endocytosis. Third, the pseudosubstrate peptide for CKII does not inhibit CKI (35), yet it was effective at blocking TR endocytosis. And fourth, in the add-back experiment, addition of CKII, but not CKI, partially restored TR endocytosis. Taken together, the results suggest that CKI activity is not required for TR endocytosis.

Although protein phosphorylation of the clathrin-based machinery has often been associated with disassembly rather than the assembly of endocytic coat complexes (62), our studies support the idea that phosphorylation of one or more proteins in the assembly pathway is required for TR endocytosis. This would suggest that the endocytic pathway is tightly regulated by a set of complex phosphorylation-dephosphorylation reactions, and our studies provide evidence that CKII or a CKII-like kinase is one of the necessary regulatory components. Clearly, identification of the kinase involved in this process will be necessary before any of the in vivo substrates can be determined. Furthermore, it will be interesting to determine if other cell surface molecules require the action of this kinase for internalization, and if this kinase is regulated in a cell cycle-dependent manner.

	ACKNOWLEDGEMENTS

We thank Doug Cyr, Danise Rogers, and Sandy Schmid for critical reading of the manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health, NIDDK Grant R29-DK47339 (to J. F. C.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be sent. Tel.: 205-934-1002; Fax: 205-975-5648; E-mail: jcollawn@uab.edu.

	ABBREVIATIONS

The abbreviations used are: TR, transferrin receptor; Tf, transferrin; CEF, chicken embryo fibroblast; B-Tf, biotinylated transferrin; CKI, casein kinase I; CKII, casein kinase II; PKC, protein kinase C; PKA, cyclic AMP-dependent protein kinase A; MHC, major histocompatibility complex; Ii-TR, MHC class II invariant chain-transferrin receptor chimera.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES