Download Analysis of the Juxtamembrane Dileucine Motif in the Insulin Receptor

0013-7227/98/$03.00/0
Endocrinology
Copyright © 1998 by The Endocrine Society
Vol. 139, No. 4
Printed in U.S.A.
Analysis of the Juxtamembrane Dileucine Motif in the
Insulin Receptor*
CAROL RENFREW HAFT, MARIA DE LA LUZ SIERRA, ISABELLE HAMER,
JEAN-LOUIS CARPENTIER, AND SIMEON I. TAYLOR
Diabetes Branch, National Institutes of Diabetes and Digestive and Kidney Diseases, National
Institutes of Health (C.R.H., M.d.L.L.S., S.I.T.), Bethesda, Maryland 20892; and the Department of
Morphology, University of Geneva (I.H., J.-L.C.), Geneva, Switzerland
ABSTRACT
Dileucine-containing motifs are involved in trans-Golgi sorting,
lysosomal targeting, and internalization. Previously, we have shown
that the dileucine motif (EKITLL, residues 982–987) in the juxtamembrane region of the insulin receptor is involved in receptor
internalization. Substitution of alanine residues for Leu986 and
Leu987 led to a 3- to 5-fold decrease in the ability of the receptors to
mediate insulin uptake. In the current study, we show that mutation
of the same motif to Met986Ser987, the sequence found in the homologous position in the type I insulin-like growth factor receptor, did not
affect insulin uptake. Therefore, we inquired whether the sequence
EKITMS as an isolated motif could mediate the targeting of a reporter
molecule to endosomes and then lysosomes, as was shown previously
with the EKITLL motif of the normal receptor. Chimeric molecules
containing Tac antigen fused to different hexapeptide sequences
showed distinct patterns of subcellular localization by immunofluo-
W
HEN INSULIN binds its receptor, this triggers a redistribution of ligand-receptor complexes from microvilli to the nonvillous region of the cell surface. This
redistribution requires ligand-induced autophosphorylation
of the b-subunit of the insulin receptor and kinase activation
(1, 2). The second step in endocytosis, interaction of the
ligand-receptor complexes with clathrin-coated pits, involves other structures in the intracellular domain (3, 4).
Subsequently, insulin-receptor complexes are internalized
into the cell, and both insulin and its receptors are either
recycled back to the cell surface or targeted to lysosomes for
degradation. In many cell types, insulin stimulates receptor
endocytosis and accelerates the rate of receptor degradation.
This process, called down-regulation, results in a decrease in
the number of cell surface receptors and may be important
for attenuating signals initiated by insulin binding (5–7).
Over the past few years, progress has been made toward
understanding the signals involved in internalization and
down-regulation of transmembrane receptors. Efficient internalization of transmembrane receptor molecules requires
one or more signal sequences in the cytoplasmic domain of
Received June 16, 1997.
Address all correspondence and requests for reprints to: Simeon I.
Taylor, M.D., Ph.D., National Institutes of Health, Building 10, Room
9S-213, 10 Center Drive, MSC-1829, Bethesda, Maryland 20892-1829.
* This work was supported in part by Grant 3100 – 043409-95 from the
Swiss National Science Foundation.
1
The numbering system used throughout the text for describing the
amino acid positions in the insulin receptor are based on the cDNA that
lacks exon 11 (29).
rescence microscopy. Tac-EKITLL and Tac-EKITAA were found predominantly in lysosomes and the plasma membrane, respectively. In
contrast, Tac-EKITMS was found at the plasma membrane, in the
trans-Golgi network, and in endosomes, but only small amounts were
found in lysosomes. Thus, the dileucine motif (EKITLL) plays an
important role in directing endocytosis of the intact insulin receptor
and in mediating efficient endocytosis and lysosomal targeting as an
isolated motif. Substitution of AA for LL inhibits endocytosis and
lysosomal targeting in both systems. In contrast, substitution of MS
for LL permits rapid endocytosis in the intact receptor, but mediates
modest endocytosis and very little targeting to lysosomes as an isolated motif. Our observations support the idea that sorting signals are
recognized at multiple steps in the cell, and that specific amino acid
substitutions may differentially affect each of these sorting steps.
(Endocrinology 139: 1618 –1629, 1998)
the protein (8, 9). To date, two different types of internalization signals have been described: tyrosine-based motifs
and dileucine-based motifs (10 –12). These sorting motifs are
often clustered together in the juxtamembrane domain of the
receptor molecule in close apposition to the plasma membrane (13, 14). For the insulin receptor, the sequences GPLY
(residues 950 –953)1 and NPEY (residues 957–960) have been
reported by several laboratories to be required for rapid
endocytosis (15, 16). These tyrosine-based motifs are thought
to form type I turns that then interact with clathrin-associated
adaptins (17). In addition, extensive site-directed mutagenesis of residues including and surrounding the tyrosinebased motifs suggests that additional information required
for rapid endocytosis is found in the juxtamembrane domain
of the receptor (18). We have recently identified a six-amino
acid dileucine-containing sequence (EKITLL, residues 982–
987) that is also involved in mediating efficient internalization of the receptor (19, 20).
The type I insulin-like growth factor I (IGF-I) receptor is
another member of the tyrosine kinase family of receptors
that includes the insulin receptor and the insulin receptorrelated receptor (21, 22). Like the insulin receptor, the IGF-I
receptor undergoes ligand-stimulated internalization, and
the integrity of the juxtamembrane domain of the IGF-I receptor is required for rapid endocytosis (23, 24). However, it
has been suggested that the IGF-I receptor is internalized
more slowly than the insulin receptor (25). Interestingly, the
dileucine motif found in the juxtamembrane domain of the
insulin receptor is not conserved in the IGF-I receptor. In
1618
REPLACING LeuLeu IN hIR WITH MetSer
place of the EKITLL motif found in the juxtamembrane domain in the insulin receptor, the IGF-I receptor contains the
hexapeptide sequence EKITMS. The substitution of MetSer
for LeuLeu is notable because there are so few amino acid
differences in the juxtamembrane domains of these two
growth factor receptors (69% identical; 84% similar over 55
amino acids; Fig. 1) (26).
In the present work, we demonstrated that substitution of
MetSer for Leu986Leu987 in the full-length insulin receptor did
not impair endocytosis of the mutant receptor. This raised
the question of whether the MetSer sequence was capable of
functioning as a dileucine motif even though it lacked leucine
residues. We addressed this question using the original assay
for a dileucine motif, i.e. the ability to target a Tac chimera
to lysosomes (12). We constructed chimeric molecules consisting of Tac fused to either the IGF-I receptor juxtamembrane domain sequence, EKITMS, or the insulin receptor
juxtamembrane domain sequence, EKITLL. In addition, we
constructed a chimera with the EKITAA sequence that lacks
the ability to target the chimera to lysosomes. These chimeras
were constructed to isolate the hexapeptide sequences from
other targeting motifs in the cytoplasmic domain of the fulllength receptor (e.g. tyrosine-based motifs and other
dileucine motifs). Interestingly, the EKITMS chimera did not
possess the full activity of the EKITLL sequence to target the
chimeric molecule efficiently to lysosomes. A substantial
fraction of the EKITMS chimeras was detected in endosomes
and the trans-Golgi network (TGN), whereas the EKITAA
chimera was found almost exclusively at the plasma
membrane.
Materials and Methods
Cells and medium
NIH-3T3 cells transfected with wild-type (WT) and mutant insulin
receptors were maintained as previously described (27). HeLa cells were
maintained in DMEM (high glucose, 4.5 mm glucose; Biofluids, Gaithersburg, MD), supplemented with 10% (vol/vol) FBS, 4 mm glutamine,
100 U/ml penicillin, and streptomycin at 37 C in 5% CO2.
Construction of juxtamembrane dileucine pair mutants of
the human insulin receptor complementary DNA (cDNA)
A plasmid containing a PstI/EcoRI fragment of the human insulin
receptor cDNA (nucleotides 2744 – 4037) (28, 29) was digested with BanII
and XhoI to remove a 41-nucleotide fragment. A double stranded insert
was formed from the following oligonucleotide primers to introduce the
MetSer for LeuLeu substitutions (mutations denoted by underlining
nucleotide sequence): 59-TCGAGAGAAGATCACCATGTCTCGAGAGCTGGGGCAGGGCT (nucleotides 3070 –3111 of the sense strand)
and 59-CTGCCCAGCTCTCGAGACATGGTGATCTTCTC (nucleotides
3074 –3108 of the antisense strand). The double stranded insert was then
1619
substituted for the WT BanII/XhoI fragment. The construction was verified by determination of its nucleotide sequence. Introduction of an
AlaAla pair for the same juxtamembrane leucine residues has been
described previously (19).
A plasmid containing the full-length insulin receptor cDNA was then
digested with SspI and the combination of SspI and BstXI. The subclones
were also cut with SspI and BstXI, and the 441-nucleotide fragments
containing the mutant sequences were isolated. After ligation of each
mutant fragment into the WT insulin receptor cDNA, the complete
insulin receptor cDNAs encoding either the AlaAla pair or the MetSer
pair were ligated into a bovine papilloma virus-based expression vector
(Pharmacia, Piscataway, NJ) in which insulin receptor cDNA expression
is driven by the murine metallothionein promoter (27, 30). The constructions were verified by restriction digestion and by determination of
the nucleotide sequences.
Expression of receptors by transfection of cDNAs into NIH3T3 cells
NIH-3T3 cells (;2 3 106 cells) were stably transfected with a bovine
papilloma virus-based expression vector encoding WT human insulin
receptor and the various mutant receptors as described previously (27).
Expression of insulin receptors was assayed by measuring [125I]insulin
binding and/or immunoblotting (31). We have chosen to use this expression vector in NIH-3T3 cells because this method has consistently
provided the highest level of stable expression among the methods we
have investigated in our laboratory.
Construction of chimeras
Double stranded inserts containing the insulin receptor juxtamembrane dileucine motif or the corresponding mutant motifs followed by
a stop codon were ligated into a modified version of TacDKQTLL(CD3g) as previously described (19). The sequences of the
inserts are available upon request. The constructions were verified by
restriction digestion and determination of their nucleotide sequence.
Transient transfection of chimeras into HeLa cells
Two to 3 h before the start of the transfection, subconfluent cultures
of HeLa cells were placed in 9 ml fresh growth medium (see above). For
each 10-cm plate, 10 –15 mg DNA were mixed with 64 ml 2 m CaCl2 and
brought to a final volume of 0.5 ml with water. The DNA/calcium
phosphate mixture was then slowly added to 0.5 ml 2 3 HEPES-buffered
saline (280 mm NaCl, 1.5 mm Na2HPO4z7H2O, and 50 mm HEPES, pH
7.05). After 20 min at room temperature, 1 ml of the DNA complex was
added dropwise to each plate. After 14 –18 h, the cells were washed twice
with PBS and then placed in standard growth medium for 40 – 48 h
before experiments were performed. We have chosen to express the
chimeras in HeLa cells because this facilitates high quality morphological studies and allows for comparisons with previous published work
using Tac chimeras (12).
Immunofluorescence microscopy
HeLa cells (;1 3 104/chamber) were grown overnight on two-chambered glass slides (Nunc, Naperville, IL). The cells were then fixed for
15 min in 2% (vol/vol) formaldehyde in PBS at 25 C, and immunofluorescent staining was carried out using an anti-Tac monoclonal antibody
FIG. 1. Amino acid sequence alignment of the juxtamembrane domains of the insulin and IGF-I receptors. A best-fit alignment is shown for
the juxtamembrane domain sequences of the insulin receptor (IR; amino acids 940 –994) and the IGF-I receptor (IGF-IR; amino acids 928 –979).
Two previously identified tyrosine-based internalization motifs are shown in boldface (15, 16) as well as the dileucine-containing motif of the
IR and the homologous amino acid sequence in the IGF-IR. Vertical bars are shown between identical amino acids, double dots mark conservative
amino acid substitutions with best-fit matrix scores greater than 0.5, and single dots mark conservative amino acid substitutions with positive
best-fit scores less than 0.5 (53, 54).
1620
REPLACING LeuLeu IN hIR WITH MetSer
as previously described (19). For colocalization studies, the following
polyclonal antibodies were used: anti-bCOP (coatomer protein) antibody (1:500), provided by Dr. J. Lippincott-Schwartz (NIH, Bethesda,
MD); antiendoplasmic reticulum (anti-ER) antibodies (1:100), provided
by Dr. D. Louvard (Pasteur Institute, Paris, France); antihuman transferrin antibody (1:500; Sigma Chemical Co., St. Louis, MO); or antilysosome-associated membrane protein I (anti-LAMPI) antibody (1:100),
provided by Dr. M. Fukuda (Cancer Research Center, La Jolla, CA). After
the appropriate washes, cells were incubated for 1 h with fluorescein
isothiocyanate (FITC) swine antirabbit IgG (1:50 dilution; Dako, Carpenteria, CA). Samples were then mounted in Vectashield mounting
medium (Vector Laboratories, Burlingame, CA) and examined under a
Zeiss Axiophot microscope (Zeiss, New York, NY) using 340 –100 oil
lenses. For transferrin uptake studies, cells were first incubated for 10
min at 37 C with iron-loaded transferrin [10 mg/ml in DMEM and 0.2%
(wt/vol) BSA; Sigma Chemical Co.), washed three times, and then fixed
(32). For antibody uptake studies, cells were incubated for 30 min at 37
C with anti-Tac antibody [2 mg/ml in DMEM and 0.5% (wt/vol) BSA],
provided by Dr. Thomas Waldmann. The cells were then washed and
fixed (33). The above immunofluorescence protocol was then followed.
Photomicrographs were taken using identical conditions for exposure,
development, and printing.
Measurement of [125I]insulin internalization by NIH-3T3
cells expressing WT or mutant insulin receptors
To determine the internalization rate constant for WT and mutant
receptors, transfected cells were grown to confluence in six-well plates.
The cells were washed once with PBS and then incubated at 37 C for 2–10
min in 1 ml internalization buffer [RPMI 1640 medium, 25 mm HEPES,
and 0.1% (wt/vol) BSA] containing 0.1 ng/ml [125I]insulin (20,000-50,000
dpm/ml). At 2-min intervals, the cells were placed on ice, washed twice
with ice-cold PBS to remove unbound insulin, and washed twice with
PBS, pH 3.0, containing 0.1% (wt/vol) BSA to remove any surface-bound
radioactivity. The residual cell-associated radioactivity (internalized insulin) was then quantified after dissolving the acid-washed cells in 1 ml
1 n NaOH (19, 31). The internalization rate constant for each receptor
was determined during a 2- to 10-min incubation period as previously
described (34, 35). The data reflect the mean 6 the se for three experiments, each performed in triplicate. The rate constant for the WT receptor was compared with those of the dileucine mutants using unpaired Student’s t test. P , 0.05 was considered significant.
Insulin receptor and endogenous substrate phosphorylation
Transfected NIH-3T3 cells were grown to confluence in 10-cm dishes.
After 12–18 h of serum starvation, cells were incubated in the absence
or presence of insulin (0 –1026 m) for 1 min at 37 C. The incubation
medium was then removed, and the cell monolayers were rapidly frozen
with liquid nitrogen. The cells were solubilized, and proteins containing
phosphotyrosine were then detected by immunoblotting or immunoprecipitation with an a-subunit-specific antibody (B7/B10), followed by
immunoblotting as described previously (36, 37). Insulin receptors were
also detected by immunoblotting with an a-subunit-specific antibody
(Upstate Biotechnology, Lake Placid, NY).
Quantitative electron microscopic autoradiography
After incubation at 37 C in the presence of [125I]insulin (10211 m), cells
were fixed, dehydrated, and quantified as previously described (1). For
each time point studied, three Epon blocks were prepared and sectioned.
About 450 – 600 grains were analyzed from all morphologically intact
cells. Grains within 250 nm of the plasma membrane were considered
associated with the cell surface. Grains inside the cytoplasm and more
than 250 nm from the plasma membrane were considered internalized.
Grains present at the plasma membrane fell into the following classes:
microvilli, clathrin-coated pits, nonvillous nonclathrin-coated pit segments, and unclassifiable. Grains were considered associated with microvilli or clathrin-coated pits if the center was less than 250 nm from
the surface domain.
Endo • 1998
Vol 139 • No 4
Results
Characterization of insulin receptors mutated in the
dileucine sequence of their juxtamembrane domain
To investigate the importance of the dileucine pair at positions 986 and 987 in the juxtamembrane domain of the
insulin receptor, we mutated the LeuLeu residues to MetSer
and AlaAla, and expressed the various recombinant receptors in NIH-3T3 cells. Metabolic labeling studies confirmed
that the mutant receptors were synthesized, processed, and
transported to the plasma membrane normally (data not
shown) (19). We used immunoblotting to estimate the number of receptors expressed in each cell line. As judged by an
antibody directed against the a-subunit, the cells expressing
the mutant receptors (Met986Ser987 and Ala986Ala987) contained approximately equal numbers of receptors. In contrast, the cells expressing the WT receptors (Leu986Leu987)
contained 2- to 3-fold more receptors. Furthermore, all three
receptors appeared to bind insulin with similar affinities.
When we plotted [125I]insulin binding as a function of the free
insulin concentration, half-maximal inhibition of [125I]insulin
binding was obtained at free insulin concentrations of 0.9, 1.1,
and 1.0 nm for the WT, Ala986Ala987, and Met986Ser987-mutant
receptors, respectively (data not shown). Moreover, in the
presence of tracer concentrations of [125I]insulin (50 pg/ml),
the ratios of bound/free insulin were 0.54 6 0.01
(Ala986Ala987), 0.70 6 0.01 (Met986Ser987), and 1.38 6 0.14
(WT) (19). If one assumes that all three receptors bind insulin
with the same affinity constant, then these binding data
imply that the ratios of WT receptors to mutant receptors are
2.6 and 2.0 for the Ala986Ala987 and Met986Ser987 mutant receptors, respectively. These ratios are entirely consistent with
the results of the immunoblotting studies, thus supporting
the hypothesis that the affinity of insulin binding (Ke, as
defined in Ref. 38) is normal in the two mutant receptors.
Phosphorylation of insulin receptors and endogenous
substrates in intact cells
To assess the in situ autophosphorylation of the MetSer
mutant insulin receptor, intact cells were incubated in the
presence or absence of insulin. Phosphorylated insulin receptors were detected by immunoblotting with antiphosphotyrosine antibody. Both the MetSer mutant and the WT
receptor molecules underwent rapid phosphorylation after
the addition of 100 nm insulin (Fig. 2A). When the amount
of b-subunit phosphorylation was normalized for the total
receptor number present in each sample, as determined by
Western blot analyses, the MetSer mutant showed no significant change in the amount of b-subunit phosphorylation
compared with the WT receptor. Furthermore, the MetSer
mutant receptors exhibited normal tyrosine kinase activity
toward endogenous substrates. Two prominent phosphorylated bands were visualized in cells expressing both WT
and mutant receptors after insulin addition (Fig. 2B). The
95-kDa band corresponds to the b-subunit of the insulin
receptor, and the 185-kDa band corresponds to pp185/insulin receptor substrate-1 and/or -2 (39). Neither b-subunit
nor insulin receptor substrate-1 and/or -2 phosphorylation
was detectable in the parental cell line (data not shown) (40).
REPLACING LeuLeu IN hIR WITH MetSer
FIG. 2. Phosphorylation of insulin receptors and endogenous substrates stimulated by insulin in intact cells. NIH-3T3 cells expressing
either WT or mutant (Met986Ser987) insulin receptors were incubated
in the absence or presence of insulin for 1 min at 37 C. A, A portion
of the cell lysate was immunoprecipitated with an antiinsulin receptor antibody directed against the a-subunit, followed by Western
blotting with an antiphosphotyrosine antibody. A portion of lysate
was also used to determine the total amount of receptor present in
each sample by Western blotting with antibody directed against the
a-subunit of the receptor. B, Western blots of whole cell lysates were
probed with an antiphosphotyrosine antibody. The positions of insulin
receptor substrate-1 and the insulin receptor precursor (prec) a- and
b-subunits are indicated. The total amount of cell surface receptor
was estimated by quantifying the a-subunit. The amount of phosphorylated receptor was determined by quantifying the tyrosinephosphorylated b-subunit as described in Materials and Methods.
Autoradiograms representative of two experiments are shown. Note
that we adjusted the loading of the gel so as to load approximately
equal numbers of WT and MetSer mutant receptors: 40 ml extract
from cells expressing MetSer mutant receptor, and 20 ml extract from
cells expressing WT receptor.
Internalization of WT and mutant receptors
To investigate whether the juxtamembrane domain mutation affected events on the cell surface, we analyzed quantitatively the surface localization and redistribution of the
various receptors. As shown in Fig. 3A, both the WT and
MetSer mutant receptors associated normally and preferentially with microvilli after a 2-h incubation at 4 C in the
presence of [125I]insulin. At 37 C, both receptors redistrib-
1621
FIG. 3. Surface redistribution of [125I]insulin in NIH-3T3 cells expressing WT or mutant insulin receptors. The results presented are
the mean 6 SE of the analysis of sections from each of seven different
Epon blocks (n 5 7). Three blocks were obtained from one experiment,
and the remaining four blocks were obtained from two separate experiments (two blocks from each experiment). Cells were incubated at
37 C in the presence of [125I]insulin for the indicated periods and then
processed for EM autoradiography. A shows the number of grains
associated with the microvilli (expressed as a percentage of the total
number of grains) as a function of time. B shows the number of grains
associated with clathrin-coated pits (expressed as a function of the
number of grains associated with the nonvillous surface). Because the
association with clathrin-coated pits did not change over time, we
pooled the values obtained at 0, 5, 15, and 30 min.
uted toward the nonvillous domain of the cell surface (Fig.
3A) and segregated similarly in coated pits (Fig. 3B).
Next we examined the delivery of mutant and WT receptors into endosomes. NIH-3T3 cells expressing WT or mutant
receptors were incubated with [125I]insulin for 0 –10 min at 37
C. The cells were then chilled to 4 C and washed with acidic
PBS (pH 3) to determine the amount of acid-dissociable radioactivity (i.e. surface-bound insulin) and the amount of
residual cell-associated radioactivity (i.e. internalized insulin). Figure 4 shows the ratio of intracellular to cell surface
[125I]insulin as a function of time for the mutant and WT
receptor molecules. Over a 10-min time frame, when receptor
recycling and degradation were negligible, MetSer mutant
receptors internalized similar amounts of ligand as the WT
receptor. In contrast, a mutant in which the dileucine pair
was changed to a pair of alanine residues internalized approximately 80% less ligand than the WT receptor. Using the
data shown in Fig. 4, we then determined the rate constant
for insulin receptor internalization for each cell line from the
slope of a plot of internalized insulin vs. the integral of the
surface-bound ligand measured at each time point (34, 35)
The slope was calculated by linear regression (41) using
1622
REPLACING LeuLeu IN hIR WITH MetSer
Endo • 1998
Vol 139 • No 4
viously in rat basophilic leukemia cells (19), the chimeras
containing the WT insulin receptor sequence (Tac-EKITLL)
were associated with intracellular vesicular structures (Fig.
5A), whereas molecules expressing the mutated sequence
(Tac-EKITAA) were localized to the plasma membrane (Fig.
5B). In contrast, cells transiently transfected with chimeras
containing the IGF-I receptor sequence (Tac-EKITMS)
showed a heterogeneous pattern. Staining was detected both
at the plasma membrane and intracellularly (Fig. 5C). When
the immunostaining was carried out in the absence of de-
FIG. 4. Insulin internalization by NIH-3T3 cells expressing WT or
mutant insulin receptors. NIH-3T3 cells expressing either WT or
mutant insulin receptors were exposed to [125I]insulin (0.01 nM) for
2–10 min at 37 C. At the indicated times after insulin addition,
surface-bound and internalized [125I]insulin were determined as described in Materials and Methods. The ratio of the intracellular insulin to surface-bound insulin is plotted as a function of time for
NIH-3T3 cells expressing either WT (solid squares), AlaAla mutant
(Ala986Ala987; solid triangles), or MetSer mutant insulin receptors
(Met986Ser987; open triangles). The results are the mean 6 SE of three
experiments, each performed in triplicate.
experimental data; the regression coefficients were greater
than 0.95. Cells expressing WT insulin receptors internalized
0.01 nm insulin rapidly, with a rate constant of ke 5 0.159 6
0.068 min21. The rate constant for MetSer mutant receptors
was slightly, but not significantly, lower, with a value of ke
5 0.133 6 0.078 min21. In contrast, the AlaAla mutant receptors internalized insulin with a rate 5-fold slower (ke 5
0.031 6 0.012 min21) than that estimated for WT receptors.
Similar results were obtained with three independent clones
of cells expressing the mutant receptors.
Pulse-chase labeling studies showed that there were no
significant differences in the turnover of the WT and mutant
receptor molecules after extended insulin treatment (data not
shown) (19).
Localization of Tac chimeras
In the context of the full-length insulin receptor molecule,
the EKITMS sequence permits normal receptor endocytosis
and degradation. There are at least two possible explanations
for this observation. It is possible that the isolated EKITMS
sequence is a functional equivalent of the dileucine motif
(EKITLL) as a sorting signal. Alternatively, there may be
other sequences in the insulin receptor that are capable of
substituting for the EKITLL targeting sequence. To address
this question, we separated the hexapeptide sequences from
other potential internalization motifs present in the intact
receptors by constructing chimeric proteins. The chimeras
consisted of the human Tac antigen (interleukin-2R, a-chain)
(42) fused to the various hexapeptide motifs. After transient
transfection of HeLa cells with the Tac chimeras, the subcellular localization of the chimeric molecules was evaluated
by indirect immunofluorescence microscopy. As shown pre-
FIG. 5. Immunofluorescence microscopy of HeLa cells expressing
Tac-insulin receptor chimeras. HeLa cells were transiently transfected with Tac-EKITLL (A), Tac-EKITAA (B), and Tac-EKITMS (C).
Forty-eight hours after transfection, the cells were fixed, permeabilized, and stained for immunofluorescence with anti-Tac antibody
and rhodamine-conjugated second antibody.
REPLACING LeuLeu IN hIR WITH MetSer
tergent, no specific signal was observed for Tac-EKITLL, and
the intracellular signal for Tac-EKITMS was eliminated.
However, the cell surface staining observed for the TacEKITAA and Tac-EKITMS was unchanged (data not shown).
We next carried out a series of colocalization studies to
define the subcellular localization of the various Tac chimeras. Tac-EKITLL showed extensive colocalization with
LAMPI (43), an endogenous HeLa cell protein that is located
in late endosomes and lysosomes (Fig. 6, A and B). In contrast, Tac-EKITMS showed only slight colocalization with
LAMPI (Fig. 6, C and D). Tac-EKITAA was not detectable in
LAMPI-containing structures (data not shown). When cells
were incubated for 10 min at 37 C with diferric transferrin (10
mg/ml) before fixation to load early endosomes with tracer,
a different colocalization pattern was seen. Tac-EKITMS
showed extensive colocalization with transferrin (Fig. 7, A
and B), whereas Tac-EKITAA (Fig. 7, C and D) and TacEKITLL (Fig. 7, D and E) showed only small amounts of
colocalization with early endosomes. Compared with the
distribution of an ER marker in double immunofluorescence
labeling experiments, the punctate intracellular stainings for
Tac-EKITMS and Tac-EKITLL were distinct from the diffuse
reticular structures labeled by an ER-specific antibody (Fig.
1623
8, A and B, data not shown). However, double labeling for
Tac and the coatomer protein, bCOP, a marker for the intermediate compartment and TGN (44 – 46), showed partial
colocalization with both Tac-EKITLL and Tac-EKITMS (Fig.
8, C–F).
Figure 9 summarizes the subcellular localizations of the
various Tac-hexapeptide chimeras obtained by quantifying
which organelles contained detectable immunofluorescence
signal for each chimera. Four or five independent experiments were examined. x2 analyses of the data showed that
the distributions obtained for Tac-EKITAA, Tac-EKITMS,
and Tac-EKITLL were significantly different from each other
with P # 0.001.
Internalization of Tac chimeras
The immunofluorescence studies demonstrated that a
large fraction of Tac-EKITLL molecules are located in lysosomes. Although we did not detect immunofluorescent staining at the plasma membrane, it is possible that the steady
state level of the Tac-EKITLL chimera at the plasma membrane is below the limit of detection due to extremely rapid
endocytosis. Therefore, we incubated living cells transfected
FIG. 6. Colocalization of Tac chimeras with endogenous LAMPI. HeLa cells were transiently transfected with Tac-EKITLL (A and B) or
Tac-EKITMS (C and D). Forty-eight hours after transfection, the cells were fixed, permeabilized, and processed for immunofluorescence.
Endogenous LAMPI was stained with polyclonal anti-LAMPI antibody (left), and the Tac chimeras were stained with monoclonal anti-Tac
antibody (right). FITC-conjugated swine antirabbit and rhodamine-conjugated donkey antimouse second antibodies were used.
1624
REPLACING LeuLeu IN hIR WITH MetSer
Endo • 1998
Vol 139 • No 4
FIG. 7. Colocalization of Tac chimeras with transferrin-loaded endosomes. HeLa cells were transiently transfected with Tac-EKITMS (A and
B), Tac-EKITAA (C and D), and Tac-EKITLL (E and F). Forty-eight hours after transfection, the cells were incubated for 10 min at 37 C with
10 mg/ml diferric transferrin to label early endocytic compartments. The cells were then washed, fixed, permeabilized, and processed for
immunofluorescence. Transferrin was stained with polyclonal antitransferrin antibody (left), and the Tac chimeras were stained with monoclonal anti-Tac antibody (right). FITC-conjugated swine antirabbit and rhodamine-labeled donkey antimouse second antibodies were used.
Arrows point to representative collections of vesicles where transferrin and the various Tac chimeras colocalize. Arrowheads point to vesicular
structures that contain Tac-EKITLL, but lack transferrin.
with the various Tac chimeras in the presence of anti-Tac
antibody for 30 min at 37 C. Under these conditions, antibody
can bind to and be internalized by the chimeras even if they
only appear transiently on the cell surface. After washing,
fixing, and permeabilizing the cells, the cell-associated antiTac antibody was visualized by immunofluorescence using
a labeled secondary antibody. When cells transfected with
Tac-EKITAA were examined, they exhibited bright surface
REPLACING LeuLeu IN hIR WITH MetSer
1625
FIG. 8. Colocalization of Tac chimeras with endogenous ER antigens and bCOP. HeLa cells were transiently transfected with Tac-EKITMS
(A–D) or Tac-EKITLL (E and F). Forty-eight hours after transfection, the cells were washed, fixed, permeabilized, and processed for immunofluorescence. Tac chimeras (B, D, and F) were stained with anti-Tac monoclonal antibody followed by rhodamine antimouse IgG (right).
Endogenous HeLa cell bCOP (C and E) was stained with an anti-bCOP polyclonal antibody and endoplasmic reticulum (ER) antigens were
stained with anti-ER (A) followed by FITC antirabbit IgG (left). Arrows point to representative structures where bCOP and Tac-EKITMS or
Tac-EKITLL chimeras colocalize. Arrowheads point to vesicular structures that contain the designated chimeras, but lack bCOP.
labeling and small amounts of vesicular staining near the cell
periphery (Fig. 10, A and A9). The Tac-EKITMS chimeras had
cell surface staining that was not as extensive as that seen
with Tac-EKITAA (Fig. 10B). In addition, endosomal labeling
was more extensive in cells expressing the EKITMS chimera
(Fig. 10B9). In contrast, with cells expressing Tac-EKITLL, the
anti-Tac antibody was seen in endosomal and lysosomal
structures that were often found in the perinuclear region of
the cell; the chimera was not detectable on the cell surface
(Fig. 10, C and C9). These findings demonstrate that at least
a portion of the Tac-EKITLL chimeras was delivered to the
cell surface, where it then bound antibody and was rapidly
1626
REPLACING LeuLeu IN hIR WITH MetSer
Endo • 1998
Vol 139 • No 4
nonconservative substitutions (i.e. MetSer for Leu986Leu987)
is located in the juxtamembrane dileucine motif that we had
previously demonstrated to be required for rapid endocytosis of the insulin receptor (19, 20). Because the insulin
receptor has been reported to undergo 3-fold more rapid
endocytosis than the IGF-I receptor (25), we inquired
whether the substitution of MetSer for LeuLeu in the EKITLL
sequence of the insulin receptor might account for the reduction of the endocytosis rate. However, the MetSer substitution did not affect the endocytosis of the receptor, as
measured by uptake of [125I]insulin or by quantification of
both constitutive and insulin-stimulated internalization of
receptors. The ability to substitute Met for the first Leu was
not surprising, in that MetLeu has been shown to function as
an internalization signal in the major histocompatibility complex-associated invariant chain (14). However, substitution
of Ser (a hydrophilic amino acid residue) for the second Leu
in the insulin receptor juxtamembrane dileucine motif is
clearly a nonconservative substitution. In any case, preservation of the ability of the sequence motif, EKITMS in the
juxtamembrane domain of the IGF-I receptor, to direct a
protein toward endocytosis suggests that there are evolutionary constraints that conserved the functionality of this
targeting sequence.
The substitution of MetSer for LeuLeu does not explain the
difference in the rates at which the receptors for insulin and
IGF-I undergo endocytosis. Thus, it is likely that other structural differences between the two receptors provide an explanation for the observed differences in the rates of endocytosis. In this regard, it is noteworthy that the insulin
receptor has three additional dileucine motifs in its cytoplasmic domain (19). Although two of these are conserved in
the IGF-I receptor, the C-terminal dileucine motif (EIVNLL,
amino acids 1246 –1251) is not present in the IGF-I receptor.
FIG. 9. Quantitation of the subcellular distribution of Tac chimeras.
HeLa cells were transiently transfected with expression vectors encoding each of three Tac-hexapeptide chimeras, and immunofluorescence staining was performed with anti-Tac antibody. Each transfected cell was classified according to which organelles were stained
(plasma membrane, endosomes, lysosomes, and/or TGN). Staining
was scored as positive if a signal was visually detectable. At the time
the pictures were scored, the observer was blinded with respect to the
identity of the construct. The number of cells counted for each construct (n) is indicated in the upper righthand corner of each panel.
internalized and delivered to lysosomes. However, although
Tac-EKITMS is internalized from the cell surface, it does so
more slowly than Tac-EKITLL. Furthermore, once internalized into endosomes, Tac-EKITMS is not efficiently targeted
to lysosomes.
Discussion
Sequence similarities in the juxtamembrane domains of the
insulin receptor family of tyrosine kinases
Comparison of the amino acid sequences of the insulin
receptor with the homologous IGF-I receptor reveals that the
juxtamembrane domains are highly conserved (69% identical; 84% similar over 55 amino acids) (26). One of the few
Comparison of observations in full-length receptor and
Tac chimeras
We used two complementary experimental approaches in
this manuscript. First, we analyzed the effects of mutations
in the juxtamembrane dileucine motif in the full-length insulin receptor. This provides insight into the function of the
sequence in a physiologically relevant context, but is limited
by interactions with other targeting motifs elsewhere in the
cytoplasmic domain of the receptor. This approach led to the
conclusion that the Met986Ser987 substitution did not impair
endocytosis of the full-length insulin receptor. Second, we
have assayed the ability of hexapeptide sequences to target
chimeric proteins to lysosomes. Although this is an artificial
experimental system, it has the advantage of isolating a specific motif from other sequences in the insulin receptor. This
approach demonstrated that MetSer does not substitute for
LeuLeu in the dileucine motif when the hexapeptide sequence is placed at the C-terminus of the Tac antigen (interleukin-2 receptor, a-chain). In the context of the Tac chimera, MetSer does not function as efficiently as LeuLeu in
mediating endocytosis and targeting of the chimeric protein
to lysosomes. Nevertheless, the Met986Ser987 substitution allowed for more efficient internalization of both the fulllength receptor and the Tac chimera than the corresponding
REPLACING LeuLeu IN hIR WITH MetSer
1627
FIG. 10. Surface distribution and endocytosis of Tac chimeras. Forty-eight hours after transfection of HeLa cells with Tac-EKITAA, TacEKITMS, and Tac-EKITLL, living cells were incubated for 30 min at 37 C with 2 mg/ml anti-Tac monoclonal antibody. The cells were then
washed, fixed, and permeabilized, and the surface-bound or endocytosed anti-Tac monoclonal antibody was detected with rhodamine antimouse
IgG. Two focal planes are shown for each field of cells examined (A–C, a plane near the surface of the cell; A9–C9, a plane deeper in the cell).
All pictures were taken using identical conditions for exposure, development, and printing. Tac-EKITAA was found predominantly at the cell
surface (A and A9), whereas Tac-EKITLL was found exclusively in intracellular structures resembling late endosomes and lysosomes (C and
C9). In contrast, Tac-EKITMS was found both at the plasma membrane (B) and in vesicular structures resembling early endosomes (B9).
Ala986Ala987 substitution. The differences in sorting efficiencies seen with the three hexapeptide motifs are probably due
to differences in the specificity and avidity with which these
sequences interact (either directly or indirectly) with the
plasma membrane adaptor complex (AP2), the TGN adaptor
complex (AP1), and/or the recently described adaptor complex (AP3), which may serve as an adaptor complex on
endosomes (17, 47–50). According to this interpretation, our
findings could be explained by assuming that the motif
EKITLL has a high affinity interaction with AP1, and thus,
much of Tac-EKITLL would be directly targeted from the
TGN to lysosomes. The small amount of Tac-EKITLL that
reaches the plasma membrane would then be recognized by
AP2, rapidly internalized into endosomes, and subsequently
targeted to lysosomes, presumably via high affinity interactions with a adaptor complex on endosomes, perhaps AP3
(50, 51). In contrast, Tac-EKTIAA and Tac-EKITMS would
have lower affinity interactions with AP1 than Tac-EKITLL
and thus leave the TGN and travel to the plasma membrane.
Tac-EKITAA would then be internalized slowly from the
1628
REPLACING LeuLeu IN hIR WITH MetSer
plasma membrane due to weak interactions with AP2. In
contrast, we propose that Tac-EKITMS has a higher affinity
interaction with AP2 and thus would be internalized into
endosomes more rapidly than Tac-EKITAA. However, in
light of the fact that very little Tac-EKITMS is found in
lysosomes, it is possible that the MS substitution weakened
the interactions with the adaptor complex on endosomes.
Although several tyrosine-based motifs have been shown to
bind differentially to the m-chains of the AP1, AP2, and AP3,
to date no direct interaction between dileucine-containing
motifs and the m-chains has been demonstrated (17, 47–50).
Perhaps, dileucine motifs interact with other components of
the adaptor complexes or with molecules such as Nef HIV
early protein that are thought to serve as connectors between
adaptor complexes and receptor molecules (52).
In conclusion, it is likely that dileucine motifs as well as
tyrosine-based signals interact with multiple adaptor molecules. The localization of a protein within the cell is, therefore, determined by the net effect of interactions of its various
targeting sequences with multiple adaptor complexes. Mutations would thus have differential effects on each of these
binding interactions, leading to complex effects of mutations
upon protein trafficking.
Acknowledgments
We are grateful to Dr. Axel Ullrich for the generous gift of insulin
receptor cDNA. In addition, we thank Drs. Juan Bonifacino, Julie
Donaldson, Mickey Marks, and Hagai Agmon-Snir for advice and helpful discussions, and Mrs. G. Porcheron-Berthet for skilled technical
assistance. Finally, we thank Dr. Valarie Barr for critical reading of the
manuscript.
References
1. Carpentier J-L, Paccaud J-P, Gordon P, Rutter WJ, Orci L 1992 Insulin-induced
surface redistribution regulates internalizaton of the insulin receptor and requires its autophoshorylation. Proc Natl Acad Sci USA 89:162–166
2. Carpentier J-L, Paccaud J-P, Backer J, Gilbert A, Orci L, Kahn CR 1993 Two
steps of insulin receptor internalization depend on different domains of the
b-subunit. J Cell Biol 122:1243–1245
3. Backer JM, Kahn CR, White MF 1990 Tyrosine phosphorylation of insulin
receptors during insulin-stimulated internalization in rat hepatoma cells. J Biol
Chem 264:1694 –1701
4. Thies RS, Webster NJ, McClain DA 1990 A domain of the insulin receptor
required for endocytosis in rat fibroblasts. J Biol Chem 265:10132–10137
5. Gavin III JR, Roth J, Neville Jr DM, De Meyts P, Buell DN 1974 Insulindependent regulation of insulin receptor concentrations: a direct demonstration in cell culture. Proc Natl Acad Sci USA 71:397– 401
6. Kasuga M, Kahn CR, Hedo JA, Van Obberghen E, Yamada KM 1978 Insulininduced receptor loss in cultured human lymphocytes is due to accelerated
receptor degradation. Proc Natl Acad Sci USA 78:6917– 6921
7. Gorden P, Carpentier J-L, Freychet PO, Orci L 1980 Internalization of
polypeptide hormones: mechanism, intracellular localization and significance.
Diabetologia 18:263–274
8. Trowbridge IS, Collawn JF, Hopkins CR 1993 Signal-dependent membrane
trafficking in the endoctyic pathway. Annu Rev Cell Biol 9:129 –161
9. Bonifacino JS, Marks M, Ohno H, Kirchhausen T 1996 Mechanisms of signalmediated protein sorting in the endocytic and secretory pathways. Proc Assoc
Am Physicians 108:285–295
10. Goldstein JL, Brown MS, Anderson RG, Russell DW, Schneider WJ 1985
Receptor-mediated endocytosis: concepts emerging from the LDL receptor
system. Annu Rev Cell Biol 1:1–39
11. Davis CG, Van Driel GIR, Russel DW, Brown MS, Goldstein JL 1987 The low
density lipoprotein receptor: identification of amino acids in the cytoplasmic
domain required for rapid endocytosis. J Biol Chem 262:4075– 4082
12. Letourneur F, Klausner RD 1992 A novel di-leucine motif and a tyrosine-based
motif independently mediate lysosomal targeting and endocytosis of CD3
chains. Cell 69:1143–1157
13. Johnson KF, Kornfeld S 1992a The cytoplasmic tail of the mannose 6-phosphate/insulin-like growth factor-II receptor has two signals for lysosomal
enzyme sorting in the Golgi. J Cell Biol 119:249 –257
Endo • 1998
Vol 139 • No 4
14. Bremnes B, Madsen T, Gedde-Dahl M, Bakke O 1994 An LI and ML motif
in the cytoplasmic tail of the MHC-associated invariant chain mediate rapid
internalization. J Cell Sci 107:2021–2032
15. Rajagopalan M, Neidigh JL, McClain DA 1991 Amino acid sequences glypro-tyr and asn-pro-glu-tyr in the submembranous domain of the insulin
receptor are required for normal endocytosis. J Biol Chem 266:23068 –23073
16. Backer JM, Shoelson SE, Weiss MA, Hua QX, Bentley Cheatham R, Haring
E, Cahill DC, White MF 1992 The insulin receptor juxtamembrane region
contains two independent tyrosine/b-turn internalization signals. J Cell Biol
118:831– 839
17. Pearse BM, Robinson MS 1990 Clathrin, adaptors, and sorting. Annu Rev Cell
Biol 6:151–171
18. Kaburagi Y, Momomura K, Yamamoto-Honda R, Tobe K, Tamori Y, Sakura
H, Akanuma Y, Yazaki Y, Kadowaki T 1993 Site-directed mutagenesis of the
juxtamembrane domain of the human insulin receptor. J Biol Chem
268:16610 –16622
19. Renfrew Haft C, Klausner RD, Taylor SI 1994 Involvement of dileucine motifs
in the internalization and degradation of the insulin receptor. J Biol Chem
269:26286 –26294
20. Hamer I, Renfrew Haft C, Paccaud J-P, Maeder C, Taylor SI, Carpentier J-L
1997 Dual role of a dileucine motif in insulin receptor endocytosis. J Biol Chem
272:21685–21691
21. Hanks SK, Quinn AM, Hunter T 1988 The protein kinase family: conserved
features and deduced phylogeny of the catalytic domains. Science 241:42–52
22. Shier P, Watt VM 1992 Tissue-specific expression of the rat insulin receptorrelated receptor gene. Mol Endocrinol 6:723–729
23. Yamasaki H, Prager D, Melmed S 1993 Structure-function of the human
insulin-like growth factor-I receptor: a discordance of somatotroph internalization and signaling. Mol Endocrinol 7:681– 685
24. Prager D, Li HL, Yamasaki H, Melmed S 1994 Human insulin-like growth
factor I receptor internalization. Role of the juxtamembrane domain. J Biol
Chem 269:11934 –1193
25. Zapf A, Hsu D, Olefsky JM 1994 Comparison of the intracellular itineraries
of insulin-like growth factor-I and insulin and their receptors in Rat-1 fibroblasts. Endocrinology 134:2445–2452
26. Ullrich A, Gray A, Tam AW, Yang-Feng T, Tsubokawa M, Collins C, Henzel
W, Le Bon T, Kathuria S, Chen E, Jacobs S, Francke U, Ramachandran J,
Fujita-Yamaguchi Y 1986 Insulin-like growth factor I receptor primary structure: comparison with insulin receptor suggests structural determinants that
define functional specificity. EMBO J 10:2503–2512
27. Kadowaki T, Bevins CL, Cama A, Ojamaa K, Marcus-Samuels B, Kadowaki
H, Beitz L, McKeon C, Taylor SI 1988 Two mutant alleles of the insulin
receptor genein a patient with extreme insulin resistance. Science 240:787–790
28. Ebina Y, Ellis L, Jarnagin K, Edery M, Graf L, Clauser E, Ou JH, Masiarz F,
Kan YW, Goldfine ID, Roth RA, Rutter WJ 1985 The human insulin receptor
cDNA: the structural basis for hormone-activated transmembrane siganalling.
Cell 40:747–758
29. Ullrich A, Bell JR, Chen EY, Herrera R, Petruzzelli LM, Dull TJ, Gray A,
Coussens L, Liao YC, Tsubokawa M, Mason A, Seburg P, Grunfeld C, Rosen
OM, Ramachandran J 1985 Human insulin receptor and its relationship to the
tyrosine kinase family of oncogenes. Nature 313:756 –761
30. Pavlakis GN, Hammer DH 1983 Regulation of a metallothionein-growth hormone hybrid gene in bovine papilloma virus. Proc Natl Acad Sci USA
80:397– 401
31. Kadowaki T, Kadowaki H, Cama A, Marcus-Samuels B, Rovira A, Bevins C,
Taylor SI 1990 Substitution of lysine for asparagine at position 15 in the alpha
chain of the human insulin receptor. A mutation that impairs transport of
receptors to the cell surface and decreases the affinity of insulin bindiing. J Biol
Chem 265:21285–21296
32. Peters PJ, Hsu VW, Ooi CE, Finazzi D, Teal SB, Oorschot V, Donaldson JG,
Klausner RD 1995 Overxpression of wild-type and mutant ARF1 and ARF6:
distinct perturbations of nonoverlapping membrane compartments. J Cell Biol
128:1003–1017
33. Uchiyama T, Broder S, Waldmann TA 1981 A monoclonal antibody (anti-Tac)
reactive with activated and functionally mature human T cells. I. Production
of anti-Tac monoclonal antibody and distribution of Tac (1) cells. J Immunol
126:1393–1397
34. Lund KA, Opresko LK, Starbuck C, Walsh BJ, Wiley HS 1990 Quantitative
analysis of the endocytic system involved in hormone-induced receptor internalization. J Biol Chem 265:15713–15723
35. Backer JM, Shoelson SE, Haring E, White MF 1991 Insulin receptors internalize by a rapid, saturable pathway requiring receptor autophosphorylation
and an intact juxtamembrane domain. J Cell Biol 115:1535–1545
36. Levy-Toledano R, Caro HP, Hindman N, Taylor SI 1993 Streptavidin blotting:
a sensitive technique to study cell surface proteins; application to investigate
autophosphorylation and endocytosis of biotin-labeled insulin receptors. Endocrinology 113:1803–1808
37. Levy-Toledano R, Caro HP, Acilli D, Taylor SI 1994 Investigations of the
mechanism of the dominant negative effect of mutations in the tyrosine kinase
domain of the insulin receptor. EMBO J 13:835– 842
38. DeMeyts P, Roth J 1975 Cooperativity in ligand binding: a new graphic
analysis. Biochem Biophys Res Commun 66:1118 –1126
REPLACING LeuLeu IN hIR WITH MetSer
39. Sun XJ, Rothenberg P, Kahn CR, Backer JM, Araki E, Wilden PA, Cahill DA,
Goldstein BJ, White MF 1991 Structure of the insulin receptor substrate IRS-1
defines a unique signal transduction protein. Nature 352:73–77
40. Renfrew Haft C, Taylor SI 1994 Deletion of 343 amino acids from the carboxyl
terminus of the b-subunit of the insulin receptor inhibits insulin signaling.
Biochemistry 17:9143–9151
41. Opresko LK, Wiley HS 1987 Receptor-mediated endocytosis in Xenopus oocytes. II. Evidence for two novel mechanisms of hormonal regulation. J Biol
Chem 262:4116 – 4123
42. Leonard WJ, Depper JM, Crabtree GR, Rudikoff S, Pumphrey J, Robb RJ,
Kronke M, Svetlik PB, Peffer NJ, Waldmann TA, Green WC 1984 Molecular
cloning and expression of cDNAs for the human interleukin-2 receptor. Nature
311:626 – 631
43. Fukuda M 1991 Lysosomal membrane glycoproteins. Structure, biosynthesis
and intracellular trafficking. J Biol Chem 266:21327–21330
44. Rothman JE, Orci L 1992 Molecular dissection of the secretory pathway.
Nature 355:409 – 415
45. Pepperkok R, Scheel J, Horstmann H, Hauri HP, Griffiths G, Kreis T 1993
b-COP is essential for biosynthetic membrane transport from the endoplasmic
reticulum to the Golgi complex in vivo. Cell 74:71– 82
46. Lippincott-Schwartz J 1993 Bidirectional membrane traffic between the endoplasmic reticulum and Golgi apparatus. Trends Cell Biol 3:81– 88
1629
47. Ohno H, Stewart J, Fournier M-C, Bosshart H, Rhee I, Miyatake S, Saito T,
Gallusser A, Kirchhausen T, Bonifacino JS 1995 Interaction of tyrosine-based
sorting signals with clathrin-associated proteins. Science 269:1872–1875
48. Ohno H, Fournier M-C, Poy G, Bonifacino JS 1996 Structural determinants
of interaction of tyrosine-based sorting signals with adaptor medium chains.
J Biol Chem 271:29009 –29015
49. Boll W, Ohno H, Songyang Z, Rapoport I, Cantley LC, Bonifacino JS, Kirchhausen T 1996 Sequence requirements for the recognition of tryosine-based
endocytic signals by clathrin AP-2 complexes. EMBO J 15:5789 –5795
50. Dell’Angelica EC, Ohno H, Eng Ooi C, Rabinovich E, Roche KW, Bonifacino
JS 1997 AP-3: an adaptor-like protein complex with ubiquitous expression.
EMBO J 16:917–928
51. Luzio JP, Banting G 1993 Eukaryotic membrane traffic: retrieval and retention
mechanisms to achieve organelle residence. Trends Biochem Sci 18:395–398
52. Mangasarian A, Foti M, Aiken C, Chin D, Carpentier J-L, Trono D 1997 The
HIV-1 nef protein acts as a connector with the sorting pathway in the Golgi
and at the plasma membrane. Immunity 6:67–77
53. Smith TF, Waterman MS 1981 Identification of common molecular subsequences. J Mol Biol 147:195–197
54. Lipman DJ, Pearson WR 1985 Rapid and sensitive protein similarity searches.
Science 227:1435–1441