THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 265, No. 17, Issue of June 15, pp. 1013%10137,199O Printed in U.S. A.

A Domain of the Insulin Receptor Required for Endocytosis in Rat Fibroblasts”

(Received for publication, January 26, 1990)

R. Scott Thies@, Nicholas J. Webster& and Donald A. McClain#YlII From the *Department of Medicine, Uniuersity of California, San Diego, La Jolla, California 92093 and the TlVeterans Administration Medical Center, Medical Research Service, San Diego, California 92161

To study the mechanism and role of ligand-dependent endocytosis, we have engineered a mutant insulin receptor that retains its insulin binding and insulin- stimulated tyrosine kinase activities but does not ex- hibit ligand-induced internalization. The mutant has a deletion of the 16th exon which encodes 22 amino acids (residues 944-965) on the cytoplasmic side of the transmembrane region of the receptor &subunit. When the cDNA is transfected in Rat 1 cells, the mutant receptor (HIRAexl6) is processed to a glycosylated cv& heterotetramer and expressed at the cell surface. HIRAexl6 receptors bind insulin with lower affinity than normal receptors (EDso for insulin competition = 1.1 nM compared with 0.2 XIM for normal receptors), but binding is normal in detergent solution. The mutant HIRAexl6 receptor undergoes insulin-dependent au- tophosphorylation and activation as a tyrosine kinase toward exogenous substrates in vitro. In viva, the receptor is also enzymatically active, as assessed 1) by the ability of antiphosphotyrosine antibodies to precip- itate equivalent proportions (58-60%) of occupied wild type or mutant receptors and 2) by immunoblotting extracts of insulin-stimulated cells using antiphospho- tyrosine antibodies. In the latter experiment, cells ex- pressing HIRAexl6 receptors exhibit tyrosine phos- phorylation of insulin receptor p-subunits as well as of ~~185, a putative substrate of the receptor. Despite the ability to bind insulin and activate as a tyrosine kinase, HIRAexl6 receptors do not internalize in Rat 1 cells. Whereas normal surface receptors covalently labeled with the photoaffinity reagent “‘1-NAPA-DP insulin are 36% intracellular after 1 h at 37 “C, only background levels of internalization are seen when HIRAexl6 receptors are labeled. The HIRAex16 recep- tors mediate no internalization or degradation of lz51- insulin compared with control untransfected Rat 1 cells, and they do not down-regulate after long expo- sure to saturating concentrations of insulin. We con- clude that the 16th exon encodes a domain necessary for ligand-dependent endocytosis.

Insulin action is mediated by the insulin receptor, a trans- membrane glycoprotein consisting of two extracellular cy- and two transmembrane P-subunits. Analysis of the structure and

* This work was supported by the Research Service of the Veterans Administration. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement ” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3 Current address: Genetics Inst., Cambridge, MA 02140. 11 To whom correspondence should be addressed: Division of En-

docrinology, University of Alabama at Birmingham, UAB Station, Birmingham, AL 35294.

function of the insulin receptor has been greatly facilitated by the cloning of the receptor cDNA (1, 2) and gene (3). For example, analysis of the cDNA sequence has revealed the existence of several structural domains within the receptor, including a cysteine-rich domain in the a-subunit, a single transmembrane domain, and a region in the P-subunit with homology to tyrosine-specific protein kinases. The insulin receptor gene consists of 22 exons, and there is good corre- spondence between the domains described above and individ- ual exons (3).

With the exception of the tyrosine kinase region itself, functional domains of the insulin receptor are less well char- acterized, and in particular the regions of the receptor respon- sible for signaling insulin action are incompletely understood. Site-directed mutagenesis has been used to begin this char- acterization. For example, a COOH-terminal region of the p- subunit has been shown to be required for signaling the metabolic but not mitogenic actions of insulin (4-6). The structural and functional requirements for another activity of the insulin receptor, namely endocytosis, have not been de- fined, however. There is evidence that the tyrosine kinase activity of the receptor is required for normal receptor inter- nalization and down-regulation in fibroblastic cells (7-g), but this is apparently not the case for receptor internalizztion in other cell types (10). Analysis is made difficult not only by cell- or tissue-specific differences in the endocytotic itinerary of insulin receptors, but also by the multiplicity of pathways that a receptor can take even within a single cell (11, 12).

To begin to analyze the necessary and sufficient conditions for insulin receptor endocytosis, we have created by site- directed mutagenesis receptors that have altered endocytotic itineraries. In this paper we report the identification of a cytoplasmic domain of the insulin receptor that is required for internalization of the receptor in fibroblasts.

EXPERIMENTAL PROCEDURES

Materials-Porcine insulin and ‘Y-insulin, monoiodinated at the alanine 14 position (300-400 &i/rg), were kindly provided by Eli Lilly, Inc. Cell culture reagents and dialyzed fetal calf serum were purchased from GIBCO, and cells were cultured in dishes from Costar (Cambridge, MA). Wheat germ agglutinin coupled to agarose was from Vector Labs, Inc. (Burlingame, CA). Protein concentrations were determined using Bio-Rad protein assay dye. [32P]ATP labeled in the y position (-5000 Ci/mmol, 10 mCi/ml) was from Amersham Corp. Other routine reagents were purchased from Sigma. A mono- clonal anti-insulin receptor antibody was kindly provided by Dr. Kenneth Siddle (University of Cambridge, Cambridge, United King- dom).

Construction of the Mutant-Deletion of the nucleotides encoding exon 16 of the HIR’ was achieved by oligonucleotide-directed muta-

1 The abbreviations used are: HIR, human insulin receptor; HIRc, human insulin receptor (wild type); HIRAexl6, human insulin recep- tor with the 16th exon deleted; SDS-PAGE, sodium dodecylsulfate- polyacrylamide gel electrophoresis; HEPES, 4-(2-hydroxyethyl)-l- piperazineethanesulfonic acid; l”“I-NAPA-DP-insulin, iodinated B2 (2-nitro-4-azidophenylacetyl)-des-Phes’-insulin.

10132

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Endocytosis-defective Insulin Receptors 10133

genesis using the method of Kunkel et al. (13) and standard molecular biologic techniques (14). Briefly, the 2.5-kilobase pair BarnHI-SpeI fragment of the HIR cDNA was subcloned into the phagemid pBluescript SK+ (Stratagene). The HIR cDNA was kindly provided by Axe1 Ullrich (Max Planck Institute, Munich) and is the HIR-A variant lacking the alternately spliced exon 11 (1). Single-stranded template was produced by transfecting the phagemid into Escherichia coliCJ236 and infecting-with helper phage Mi3K07. One-half pmol of mutagenic olieonucleotide CTGAGAAAGAGGGTGTTTCC- ATGC, which bridges the nucleotides flanking exon 16, was annealed to 0.5 pmol of the template and extended with T7 DNA polymerase (Sequenase, United States Biochemicals). Following transfection into E. coli XL-1 blue cells, the resultant colonies were screened by restriction endonuclease digestion. Clones positive for the deletion were sequenced using a primer 50 base pairs upstream of the mutation to verify the construct. The 2.4 kilobase pair BumHI-SpeI fragment containing the deletion was then subcloned back into the expression vector pCVSVE-HIRc, which has been described previously (7). This vector encodes the mouse dihydrofolate reductase gene that is, like the cDNA, under the control of the SV40 early promoter.

Generation and Culture of Z’ransfected Cell Lines-Rat 1 cells were contransfected with pSVEneo (1 rg) and either pCVSV HIRc (10 rg) or PCVSV HIRSexl6 bv the calcium nhosphate nrecipitation method (15). Following 24-h exposure to the DNA; the cells were trypsinized, replated at 1:4, and placed under selection (Dulbecco’s modified Eagle’@12 medium, without thymidine and hypoxanthine with 10% fetal calf serum, 100 nM methotrexate, and 100 rg/ml GeneticinTM (all GIBCO)). After l-2 weeks the cells attained confluence and were passed at a 1:2 split and the methotrexate increased to 300 nM. After a further 1-2 weeks the methotrexate was increased to 1 FM. At this point, tracer insulin binding studies revealed that the average receptor number had increased in the population, so the cells were subcloned by limiting dilution in 500 nM methotrexate in 96-well culture dishes. Clones were analyzed for insulin binding, and depending on the success of the transfection, 20-80% of these clones had clearly ele- vated levels of insulin receptors.

Insulin Binding Assay-Insulin binding to whole cells was quanti- tated using A-14 monoiodinated ““I-insulin as described (8). Cells in 35.mm dishes were exposed to 17 pM ““I-insulin and various concen- trations of unlabeled insulin for 2.5 h at 16 “C, and specific binding was determined by subtracting the amount of ““I-insulin bound in the presence of excess (330 nM) unlabeled insulin. Under these conditions, bound insulin is >98% acid-extractable and thus still at the cell surface. Insulin receptors were purified by lectin affinity chromatoeraphv from 1% Triton X-100 extracts of -lOa cells as described-(g). Insulin binding to solubilized receptors was measured using polyethylene glycol precipitation (16).

Insulin Receptor Tyrosine Kinase Actiuity-Autophosphorylation of partially purified solubilized receptors was accomplished by expos- ing 200 fmol of insulin binding activity in 40 ~1 to different concen- trations of insulin overnight at 4 “C (17). 10 ~1 of a solution were then added to give the receptor solution a final concentration of 5 mM M&l,. 4 mM MnCb. 50 uM ATP (includine 2 uCi of l~-~‘Pl

- - , . 1

ATP), and 5 mM cytidine triphosphate. After 1 h on ice, the reaction was stopped by the addition of 50 ~1 of 2 x SDS-PAGE sample buffer and the receptors fractionated on 7.5% acrylamide gels. Bands visu- alized on autoradiograms were excised and counted in a scintillation counter. Assay of Glu:Tyr polypeptide phosphorylation has been described (4).

Insulin receptors were also covalently labeled using 40 rig/ml iZ61- NAPA-DP-insulin (2 h at 10 “C) as described previously (18). After UV cross-linking of the insulin to its receptor, the dishes were warmed to 37 “C for 4 min to allow autophosphorylation to proceed. The cells were rinsed and solubilized in 1% Triton containing protease and phosphatase inhibitors: 0.5 mg/ml bacitracin, 0.2 mhlphenylmethyl- sulfonyl fluoride, 1000 kIU/ml aprotinin, 50 mM NaF, 2 mM EDTA, 2 mM Na?VO+ 10 mM sodium pyrophosphate, 15 mM benzamidine, and 2 mM dichloroacetic acid. Receptors from one 60 mM dish were split equally and incubated with either a monoclonal antibody specific for human insulin receptors or a polyclonal antiserum specific for phosphotyrosyl residues. The generation and specificity of the latter antibody has been described previously (19). Immunoblotting of stim- ulated cell extracts was performed by exposing l-4 x lo6 cells in a 60-mm dish to insulin for 5 min at 37 “C. Cells were then lysed with SDS sample buffer and the lysates fractionated by SDS-PAGE. The proteins were then electrophoretically transferred to nitrocellulose and exposed to the anti-phosphotyrosine antibody and “‘I-protein A according to the method of Towbin et al. (20).

Insulin and Insulin Receptor Endocytosis and Metabolism-Cells were covalently labeled with iz51-NAPA-DP-insulin as described above and then warmed to 37 “C to allow internalization to proceed. At various times, cultures were removed and either exposed to trypsin (L-1-tosylamido-2-phenylethyl chloromethyl ketone, 1 mg/ml, 1 h on ice) or not. Trypsinization was stopped by the addition of 2.5 mg/ml soybean trypsin inhibitor, and the cells were solubilized in 1% Triton containing the inhibitors listed above. The human insulin receptors were immunoprecipitated using an antibody specific for the human receptor and fractionated by 7.5% reducing SDS-PAGE. The propor- tion of intact 135-kDa HIR o-subunit was determined by excising and counting gel bands visualized by autoradiography.

To study insulin internalization and degradation, the culture me- dium was aspirated from 35-mm dishes of cells (-2 X lo” cells/dish) and replaced with 1 ml of modified Eagle’s medium (bicarbonate- free), 1% bovine serum albumin, 10 mM HEPES, pH 7.4. Cultures were held at 37 “C in a shaking water bath. ‘2”I-Insulin (0.07 nM) was added, and at various times, cultures were assayed for insulin degra- dation and internalization as described (8). Degraded insulin was estimated by measuring trichloroacetic acid-soluble material, and internalized insulin was defined as that not removed by pH 4.0 buffer (21).

To study down-regulation, 2-4 x lo” cells in 35-mm dishes were exposed to 170 nM insulin for 24 h at 37 “C. Cells were then washed twice (2 min each at 4 “C) with pH 4.0 buffer to remove bound insulin and returned to 37 ’ for 1 h. During this period internalized receptors return to their pretreatment distribution, i.e. 80-90% at the cell surface so that the total cellular component of receptors can be estimated by surface binding of ““I-insulin.

RESULTS

Transfection of Cells and Characterization of Insulin Bind- ing-A mutant insulin receptor with the 16th exon deleted was engineered as described under “Experimental Proce- dures.” The deletion encompasses 22 amino acids from resi- dues 944-965 inclusive according to the numbering of Ullrich et al. (1). It should be pointed out that the wild type template was the HIR-A isoform of the receptor, that is the receptor that lacks the 12-amino acid insert encoded by the alternately spliced exon 11 (22, 23). After verifying the construct by sequencing, the wild type human insulin receptor cDNA (HIRc) or the cDNA encoding the receptor with the deletion (HIRAexl6) were transfected into Rat 1 fibroblasts. After selection by growth in methotrexate, cells were cloned by limiting dilution and clones expressing high levels of insulin binding selected for further analysis. The host Rat 1 cells do express endogenous insulin receptors (24), but at levels suf- ficiently low (-2000 rat insulin receptors/cell) to make the selection of transfectants relatively straightforward.

Competition of tracer labeled insulin by cold insulin was studied in two clones of cells expressing similar levels of insulin receptors (4 x lo5 HIRc, 6 X lo5 HIRAexl6). Fig. 1 reveals that cells expressing the HIRAexl6 receptors bind insulin with an affinity somewhat lower than cells expressing wild type receptors. Half-maximal competition of tracer bind- ing was seen at 0.2 nM insulin for the wild type HIR but at 1.1 nM for the mutant receptors. This result was verified in several other independent clones; five clones of HIRc (of the HIRc-A or minus 12-amino acid isoform) had average EC& values for competition of 0.24 nM, whereas four clones of HIR ex16 cells all had EC,,, values of greater than 1 nM. Scatchard analysis of these competition curves reveals that the HIRAexl6 receptors, like normal receptors, still exhibit cur- vilinear binding but that the high affinity component of insulin binding is decreased in the cells carrying the mutant (not shown). Interestingly, when insulin binding is assayed in solution, the mutant receptors are very similar to normal HIR, with EC,, values for competition of tracer insulin being 0.1 nM for both receptors. Both mutant and wild type recep- tors were processed to a& glycoproteins, as determined by

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

10134 Endocytosis-defective Insulin Receptors

FIG. 1. Insulin binding to transfected cell lines. HIRc (0) and HIR ex16 (Cl) cells were exposed to 8.3 PM ““I-insulin and varying concentrations of cold insulin for 3 h at 12 “C. Competition of the label (as the percent maximum bound/free (B/F) ratio of lY”I-insulin) is plotted as a function of total insulin concentration. Results are the means of two experiments each performed in triplicate. Experiments were performed at B/F z lo%, and the amount of labeled insulin bound in the presence of 330 nM cold insulin was subtracted to yield specific binding.

-HIRc- -HIRAex 16-

?!T?m - 200

-116

- 68

- + - +

Insulin

FIG. 2. Autophosphorylation of wild type and mutant insu- lin receptors. Equal numbers (-400 fmol) of solubilized wheat germ agglutinin-agarose purified wild type or HIRlexl6 receptors were exposed to 330 nM insulin or not and then allowed to autophospho- rylate as described. The reaction was stopped with 2 x SDS sample buffer and the receptors fractionated on a 7.5% gel. The positions of the wild type (* ) and HIRlexl6 ( a) p-subunits are indicated.

their mobility on nonreducing SDS gels (not shown), their ability to bind to lectins, and the sizes of their subunits (see below).

Insulin Receptor Tyrosine Kinase Activity-The ability of the HIRAexl6 receptor to autophosphorylate and to mediate tyrosine phosphorylation of other substrates was verified in a number of ways. First, receptors from transfected cells were solubilized in detergent and partially purified by lectin affinity chromatography. Receptor concentration was determined by measuring insulin binding activity, and equal numbers of HIRc and HIRAexl6 receptors were exposed to insulin and then autophosphorylated in the presence of [y-“‘PIATP. The radiolabeled P-subunits were visualized after SDS-PAGE and autoradiography. As can be seen in Fig. 2, both normal and mutant receptors do autophosphorylate. The apparent size of the mutant P-subunit is 92 kDa, consistent with the loss of 22 amino acids from the wild type 95-kDa species. The amount of phosphate incorporation per mutant receptor is 83% of that seen by the wild type receptors. The small amount of radiolabeled material in the HIRlexl6 cells that comigrates with the wild type insulin receptor P-subunit represents the native rat receptors that are still expressed in the transfected cells; these receptors are not recognized by a human-specific anti-receptor antibody (not shown).

The mutant receptors also had kinase activity toward ex- ogenous substrates in vitro. Insulin stimulation of the wild type receptor caused a 2.9 + O.&fold increase in phosphoryl-

ation of a random copolymer of Glu:Tyr (4:1), whereas the HIRAexl6 led to a 2.0 + 0.2-fold increase. Insulin sensitivities were similar, with half-maximal stimulation occurring at 0.7- 1 nM insulin (not shown).

Autophosphorylation and kinase activity of the receptors were also intact in living cells. To measure autophosphoryla- tion, receptor a-subunits were first covalently labeled in the cold (10 “C) with a photoreactive ‘““I-insulin derivative, NAPA-DP insulin (18). The cells were then briefly warmed to 37 “C to allow autophosphorylation of the occupied recep- tors to proceed and then solubilized in the presence of phos- phatase inhibitors. The receptors were immunoprecipitated using either human insulin receptor-specific antibody or an antibody reactive to phosphotyrosine residues (19). As shown in Fig. 3, both HIRc and HIRAexl6 receptors have identical n-subunits that label with the ““I-insulin. Similar proportions of the labeled receptors are precipitated by the anti-phospho- tyrosine antibody; in two experiments an average of 60% of HIRc receptors and 58% of HIRAexl6 receptors were recog- nized by the antiphosphotyrosine antibody. Analysis of the supernatants from the precipitation confirmed that the insu- lin receptor antibody recognized both normal and mutant receptors quantitatively.

Kinase activity was also assayed by Western blotting ex- tracts of insulin-stimulated cells using the anti-phosphotyro- sine antibody. In Fig. 4, it is clear that both HIRc and

HIRC HIRAex 16 alR upY alR ClpY

135 kDa-+

FIG. 3. Autophosphorylation of wild type and mutant insu- lin receptors in situ in living cells. Cells were covalently labeled with ““I-NAPA-DP insulin and warmed briefly (37 ‘C, 4 min) as described. After solubilization, labeled receptors were precipitated either with an anti-insulin receptor (nlR) antisera that recognized >95% of the total receptors or an anti-phosphotyrosine antibody (cupy). Shown is an autoradiogram of a reducing SDS gel that reveals the precipitated labeled n-subunits.

Tyrosine Phosphorylation

Rat 1 HlRc HIR Aex 16

0 1.7 17 0 1.7 17 0 1.7 17

Insulin (nM)

FIG. 4. Tyrosine phosphorylation in intact cells. Cells in cul- ture dishes were exposed to insulin (5 min, 37 “C) and rapidly lysed in SDS sample buffer. Equal amounts of cellular protein (as deter- mined in parallel dishes) were fractionated by SDS-PAGE and West- ern blotted using an antiphosphotyrosine antibody followed by ““I- protein A. The insulin receptor p-subunits (HI&) as well as a phos- phoprotein of 185 kDa (pp 185) are indicated.

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Endocytosis-defective Insulin Receptors 10135

HIRlexl6 receptors tyrosine phosphorylate themselves and a previously described 185-kDa phosphoprotein (25) in intact cells. Equal amounts of protein from untransfected Rat 1 cells, cells expressing 4 x lo” normal HIR, or cells expressing 6 X lo” mutant HIRlexl6 were loaded on the gel. The fact that little insulin receptor or pp185 tyrosine phosphorylation is seen in untransfected Rat 1 cells with their relatively few recept.ors confirms that the phosphorylation seen in the other cells is mediated by the human receptors encoded by the transfected cDNAs.

Ligand and Receptor Internalization-Having demon- strated that the HIRJ.exlG receptors are expressed at the cell surface, bind insulin, and exhibit ligand-dependent tyrosine kinase activity, we next examined the endocytosis of normal and mutant receptors. Cell surface receptors were covalently labeled with ““I-NAPA-DP-insulin at 10 “C to prevent inter- nalization during the labeling procedure. Cells were then warmed to 37 “C to allow endocytosis to proceed, and at different times the cells were trypsinized to degrade any receptors remaining on the cell surface. The number of intact (intracellular) labeled 135kDa a-subunits was then quanti- tated after SDS-PAGE. As can be seen in Fig. 5, surface receptors can be labeled and prior to warming these receptors are >95% accessible to trypsin degradation. With time at 37 “C, however, normal receptors internalize and become tryp- sin-resistant such that 36% of the receptors are intracellular at 60 min. The HIRAexl6 receptors do not undergo endocy- tosis to any appreciable degree, however. After 60 min at 37 “C, only 5% have become trypsin-resistant.

A similar conclusion is suggested by the lack of ““I-insulin internalization and degradation mediated by HIRLexl6 cells (Fig. 6). Internalization of tracer insulin (0.07 nM) (A) pro- ceeds rapidly in cells expressing 4 x lo” normal HIRc, as measured by the inability of the internalized insulin to be

A HIRC HIR Aex 16

Trypsin - + + Time 0 0 ,t, 6’ 0 0 7:s 6’

o’-o --------_-_-__ 0

,

I , I I , I

15 30 45 60

Time (min)

FIG. 5. Internalization of labeled insulin receptors in trans- fected cells. A, an autoradiogram, with the intact tr-subunit indi- cated, showing the total initial level of receptors (- trypsin, time O), the initial trypsin sensitivity of all of these receptors (+ trypsin, time 0). and the internalization or escape from trypsin sensitivity of these receptors with time at 37 “C (+ trypsin, 75 and GO min) R, plot of internalization or escape from trypsin sensitivity of these receptors with titne at 37 “C. The cu-subunit bands from the experiment in A were excised and counted. The total amount of receptor initially present on the cells (- trypsin, time 0) was taken to represent 100%.

5 15 30 5 15 30 Time (mln)

FIG. 6. Insulin internalization and degradation by trans- fected cells. 2 x lo” cells, either HIRc (0) or HIRlexl6 (0) were equilibrated at 37 “C in 35mm dishes in 1 ml of huffer with 0.5% albumin. At time 0, ““I-insulin (0.06 nM) was added as described under “Experimental Procedures.” A, the amount of intracellular insulin (counts/min), as determined by the cell-associated counts not removable by an acid wash, is plotted as a function of time. H, the amount. of intracellular insulin was normalized as the percentage of total cell-associated insulin (cell bound courts/min after saline rinse but without an acid wash). The small amount of insulin internalized by untransfected Rat 1 cells (~200 cpm) has been subtracted to control for insulin metabolism by the endogenous rat insulin recep- tors. C, insulin degradation as estimated by the appearance of tri- chloroacetic acid-soluble counts/min in the medium. 11, insulin deg- radation normalized as in H, that is, degraded counts/min expressed as a fraction of total cell-associated counts, after subtracting the counts/min degraded by an equal number of untransfected Rat 1 cells (<200 cpm).

TABLE I Down-re&ation of normal and mutant HIR

Cells were exposed to saturating concentrations of insulin (170 nM) and then ““I-insulin binding assayed as described. Insulin bind- ing was studied after an acid wash and an incubation for 1 h at 37 “C in the absence of insulin to remove the bound insulin and to allow any internalized receptors to recycle and relax to their pretreatment distribution at the cell surface (11).

Receptor type Clone Receptors/cell Down-regulation

HIRc 8-3 4 x 10” 26% HIRc B 1.3 x 10” 41% HIRLexl6 14 5.9 x lo” 0 HIRlexl6 19 6.4 X lo” 0

released from the cells by acid pH. Very few counts are internalized in the HIRAexl6 cells, however. Because of the lower affinity of insulin for the mutant receptor, it was necessary to normalize these results for the lower total cell- associated counts in the HIRAexl6 cells. After subtracting the counts internalized in untransfected Rat 1 cells, the proportion of total cell-associated counts/min that are acid- resistant (intracellular) has been plotted as a function of time in panel B. By this analysis the HIRAexl6 receptors are defective in their ability to internalize insulin (B).

Wild type insulin receptors also deliver insulin to a degra- dative compartment, and the amount of degradation can be estimated by the appearance in the culture medium of trichlo- roacetic acid-soluble I”,’ I. C shows that HIRc but not HIRlexl6 cells mediate significant amounts of insulin deg- radation. Normalizing the data for cell-associated insulin after subtracting the degradation caused by control untransfected cells confirms this conclusion (D). Nearly identical results (after normalization) were seen in another independently isolated clone each of HIRc and HIRAexl6 cells (not shown).

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

10136 Endocytosis-defective Insulin Receptors

As a final index of receptor internalization, HIRc and HIRAexl6 cells were exposed to saturating concentrations of insulin for prolonged periods to assess whether receptor down- regulation occurred (26). As shown in Table I, the mutant HIRAexl6 receptors do not down-regulate. Whereas total receptor number decreases by 26-41% in cells with normal receptors exposed to 170 nM insulin for 24 h, no decrease is seen in HIRAexl6 cells.

DISCUSSION

The data presented here indicate that deletion of the im- mediately submembranous region of the insulin receptor re- sults in a molecule that maintains ligand specific activation of tyrosine kinase activity but does not undergo endocytosis in Rat 1 cells. That this receptor maintains normal transmem- brane signaling of its kinase despite the deletion of the region connecting its extracellular and kinase domains is of interest. These results argue that exons of the insulin receptor may indeed encode structurally and functionally discrete domains as has been hypothesized for insulin (3) and low density lipoprotein (27) receptors.

The properties of the HIRAexl6 receptor have not been completely unaltered by the deletion, however. The insulin binding properties of the mutant receptors in cells (but not in solution) were changed from those of normal receptors. The decreased affinity of the HIRAexl6 receptor for insulin suggests that the submembranous conformation of the recep- tor can affect ligand binding. A similar conclusion was reached by studying chimeras of the insulin and epidermal growth factor receptors, wherein it was demonstrated that the cyto- plasmic region did affect the binding properties of the extra- cellular ligand-binding domain (28). This could occur because of intramolecular (i.e. fl-subunit-p-subunit) or intermolecular interactions. A similar pattern of binding properties, that is, lower affinity on cells but normal affinity in solution, has also been reported for a kinase-defective receptor with a point mutation in the ATP binding site (8). The basis for these binding properties is unknown. Although both the HIRAexl6 and the kinase defective receptors are processed correctly into (Y& heterotetramers (data not shown), further details of their structure and their association with other receptors is not known.

Also interesting is the failure of the HIRAexl6 receptors to internalize. Given the otherwise near-normal properties of the mutated receptor, this suggests that the deleted region includes a domain involved in ligand-dependent endocytosis. Chen, Goldstein, and Brown have recently described an amino acid sequence, NPXY, that is required for normal internali- zation of the low density lipoprotein receptor (35). One and only one copy of this sequence exists in the insulin receptor and that NPEY is in the deleted exon of the mutant reported here. This supports the hypothesis of the above-named work- ers that this sequence is important for signalling some (but not necessarily all) endocytotic pathways, although it should be emphasized that the current studies do not directly impli- cate the NPEY sequence itself in causing the defect of the HIRAexl6 receptor. The receptor for insulin-like growth fac- tor I has an NPEY sequence in a similar submembranous position (29). As pointed out by Chen et al. (35) however, other receptors of the tyrosine kinase family either have NPXY sequences in their cytoplasmic tails (e.g. the receptor for epidermal growth factor, Ref. 30) or have no NPXY sequence (e.g. the receptors for platelet-derived growth factor or for colony-stimulating factor I, that have interrupted ty- rosine kinase domains (31)). Thus, the requirement for a submembranous NPXY sequence is not universal for all

receptors; alternate endocytotic mechanisms may operate in different cell types for different receptors. Interestingly, an epidermal growth factor receptor mutant lacking the cyto- plasmic tail, which deletes that receptor’s NPXY sequences, has been described (32). This mutant retains biologic and kinase activity but does not internalize.

The domain encoded by the 16th exon (and by implication possibly the NPEY sequence) is required for endocytosis in Rat 1 cells. However, understanding insulin receptor endo- cytosis will require the integration of a much larger and often contradictory body of evidence. For example, in these same cells, tyrosine kinase activity is also required for ligand- dependent endocytosis (7-9). In other cell types, notably liver- derived cells, kinase activity may not be required, however. For example, Backer et al. (10) have found that depletion of cellular ATP by treatment with 2,4-dinitrophenol sufficient to block receptor autophosphorylation does not block ligand- dependent endocytosis of insulin receptors. The mechanism underlying these differences among cell types in their require- ment for kinase activity to initiate internalization is not known. Ligand-dependent internalization of receptors does proceed through different pathways that can be differentiated on kinetic (33) or pharmacologic (12) grounds. Ultrastruc- turally, insulin receptor endocytosis is largely via noncoated pit pathways in adipocytes and liver-derived cells (ll), whereas embryonic cells (11) and, to different degrees, fibro- blastic lines (34) may make more use of coated pit pathways. Such data, combined with the fact that the NPXY sequence is required for coated pit-mediated internalization of the low density lipoprotein receptor, which is devoid of kinase activity suggest: 1) there are coated pit-dependent and -independent modes of endocytosis for insulin receptors; 2) there are tyro- sine kinase-dependent and -independent pathways. However, for insulin receptors the interrelationships among these path- ways and the requirement for the deleted domain (or the NPXY sequence) are unclear. Further experiments, including both expressing the mutant receptor in other cell types and combining the deletion with other mutations such as those leading to tyrosine kinase defects, are needed. These studies, as well as assessing the ability of the noninternalizing recep- tors to activate and deactivate different aspects of insulin action, should shed further light on the mechanism and role of ligand-dependent endocytosis of insulin receptors.

Acknowledgments-We would like to acknowledge Dr. Axe1 Ullrich (Max Planck Institute, Munich) for making the wild type HIR cDNA available and Dr. Jerrold Olefsky (Veterans Administration Medical Center and University of California, San Diego) for helpful discus- sions. Dr. Dietrich Brandenburg provided the NAPA-insulin. Cleon Tate prepared the manuscript.

REFERENCES

1. Ullrich, A., Bell, J. B., Chen, E. Y., Herrera, R., Petruzzelli, L. L. M., Dull, T. J., Gary, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeberg, P. H., Grunfeld, C., Rosen, 0. M., and Ramachandran. J. (1985) Nature 313. 756-761

2. Ebina, Y., Ellis, L., Jarmagin, K., Eder;, M., Graf, L., Clauser, E.. Ou. J.-H.. Masiarz. F.. Kan. Y. W.. Goldfine. I. D.. Roth. R.: and Rut&, W. (1985) ‘Cell 46, 747-758

3. Seino, S., Seino, M., Nishi, S., and Bell, G. I. (1989) Proc. Natl. Acad. Sci. U. S. A. 86. 114-118

4. McClain, D. A., Maegawa, H., Levy, J., Huecksteadt, T., Dull, T. J.. Lee. J.. Ullrich. A.. and Olefskv. J. M. (19881 J. Sol. Chem. 2ii3,8904-8911 ’ ’

“I .

5. Maegawa, H., McClain, D. A., Freidenberg, G., Olefsky, J. M., Napier, M., Lipari, T., Dull, T. J., Lee, J., and Ullrich, A. (1988) J. Biol. Chem. 263,8912-8917

6. Thies. R. S.. Ullrich. A.. and McClain. D. A. (1989) J. Biol. Chem. 264, 128iO-12824

7. Russell, D. S., Gherzi, R., Johnson, E. L., Chou, C-K., and Rosen,

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Endocytosis-defective Insulin Receptors 10137

0. M. (1987) J. Biol. Chem. 262, 11833-11840 8. McClain, D. A., Maegawa, H., Lee, J., Dull, T. J., Ullrich, A., and

Olefsky, J. M. (1987) J. Biol. Chem. 262, 14663-14671 9. Hari, J., and Roth, R. A. (1987) J. Biol. Chem. 262,15341-15344

10. Backer, J. M., Kahn, C. R., and White, M. F. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 3209-3213

11. Smith, R. M., and Jarett, L. (1988) Lab. Inuest. 58, 613-629 12. McClain, D. A., and Olefsky, J. M. (1988) Diabetes 37, 806-815 13. Kunkel, T. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,488-492 14. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman,

J. G., Smith, J. A., and Struhl, K. (eds) (1989) Current Protocols in Molecular Biology, John Wiley and Sons, New York

15. Wigler, M., Pellicer, A., Silverstein, S., and Axel, R. (1978) Cell 14,725-732

16. Marshall, S. (1985) J. Biol. Chem. 260,4136-4144 17. Freidenberg, G. R., Henry, R. R., Klein, H. H., Reichard, D. R.,

and Olefskv. J. M. (1987) J. Biol. Chen. 79.240-250 18. Heidenreich,“i(. A., Brandenburg, D., Berhan;, P., and Olefsky,

J. M. (1984) J. Biol. Chem. 259, 6511-6515 19. Kamps, M. P., and Sefton, B. M. (1988) Oncogene 2,305-315 20. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

Sci. U. S. A. 76, 4850-4354 21. Caro, J. F., Muller, G., and Glennon, J. A. (1982) J. Biol. Chem.

257,8459-8466 22. Seino, S., and Bell, G. I. (1989) Biochem. Biophys. Res. Commun.

159,312-316 23. Moller, D. E., Yokota, A., Caro, J. F., and Flier, J. S. (1989) Mol.

Endocrinol. 3, 1263-1269 24. Maegawa, H., Olefsky, J. M., Thies, S., Boyd, D., Ullrich, A., and

McClain, D. A. (1988) J. Biol. Chem. 263, 12629-12637 25. White. M. F.. Maron. R.. and Kahn. C. R. (1985) Nature 318.

183-186 26. Gavin, J. R., Roth, J., Neville, D. M., DeMeyts, P., and Buell, D.

N. (1974) hoc. N&l. Acad. Sci. U. S. A. 71, 84-88 27. Sudhot, T. C., Goldstein, J. L., Brown, M. S. and Russell, D. W.

(1985) Science 228, 815-822 28. Riedel, H., Dull, T. J., Honegger, A. M., Schlessinger, J., and

Ullrich, A. (1989) EMBO J. 8, 2943-2954 29. Ullrich, A., Gray, A., Tam, A. W., Yang-Feng, T., Tsubokawa,

M., Collins, C., Henzel, W., LeBon, T., Kathuria, S., Chen, E., Jacobs, S., Franche, U., Ramachandran, J., and Fujita-Yama- guchi, Y. (1986) EMBO J. 5,2503-2512

30. Ullrich, A., Coussens, L., Hayflick, J. S., Dull, T. J., Gray, A., Tam, A. W., Lee, J., Yarden, Y., Libermann, T. A., Schlessin- ger, J., Downward, J., Mayes, E. L. V., Whittle, N., Waterfield, M. D., and Seeburg, P. H. (1984) Nature 309, 418-425

31. Yarden, Y., Escobedo, J. A., Kuang, W. J., Yang-Feng, T. L., Daniel, T. O., Tremble, P. M., Chen, E. Y., Ando, M. E., Harkins, R. N., Francke, U., Fried, V. A., Ullrich, A., and Williams, L. T. (1986) Nature 323, 226-232

32. Chen, W. S., Lazar, C. S., Lund, K. A., Welsh, J. B., Chang, C.- P., Walton, G. M., Der, C. J., Wiley, H. S., Gill, G. N., and Rosenfeld, M. G. (1989) Cell 59, 33-43

33. Wiley, H. S. (1988) J. Cell Biol. 107, 801-810 34. Pilch, P. F., Shia, M. A., Benson, R. J. J., and Fine, R. E. (1983)

J. Cell Biol. 96, 133-138 35. Chen, W.-J., Goldstein, J. L., and Brown, M. S. (1990) J. Biol.

Chem. 265,3116-3123

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

R S Thies, N J Webster and D A McClainA domain of the insulin receptor required for endocytosis in rat fibroblasts.

1990, 265:10132-10137.J. Biol. Chem. 

  http://www.jbc.org/content/265/17/10132Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/17/10132.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on August 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from