Molecular Cell 21, 737–748, March 17, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.molcel.2006.02.018
Differential Regulation of EGF Receptor
Internalization and Degradation by
Multiubiquitination within the Kinase Domain
Fangtian Huang,1,3Donald Kirkpatrick,2,3
Xuejun Jiang,1Steven Gygi,2and Alexander Sorkin1,*
1Department of Pharmacology
University of Colorado Health Sciences Center
Aurora, Colorado 80045
2Department of Cell Biology
Harvard Medical School
Boston, Massachusetts 02115
Ubiquitination of the EGF receptor (EGFR) is believed
to play a critical role in regulating both its localization
and its stability. To elucidate the role of EGFR ubiqui-
tination, tandem mass spectrometry was used to iden-
tify six distinct lysine residues within the kinase
domain of the EGFR, which can be conjugated to ubiq-
uitin following growth factor stimulation. Substitution
of these lysine residues with arginines resulted in
a dramatic decrease in overall ubiquitination but pre-
served normal tyrosine phosphorylation of EGFR.
atratescomparable tothose ofwild-type receptors. Fi-
nally, quantitative mass spectrometry demonstrated
that more than 50% of all EGFR bound ubiquitin was
in the form of polyubiquitin chains, primarily linked
through Lys63. Taken together, these data provide di-
rect evidence for the role of EGFR ubiquitination in re-
ceptor targeting to the lysosome and implicate Lys63-
linked polyubiquitin chains in this sorting process.
The intensity, duration, and specificity of signal trans-
ulated by ligand-mediated endocytosis and postendo-
cytic trafficking of activated receptors. Internalization
of activated RTK and subsequent targeting to degrada-
tive organelles result in downregulation of RTK activity
and function as a negative feedback on downstream
signaling processes (Dikic and Giordano, 2003). Inter-
nalized RTK can form signaling complexes in endo-
somes, which may trigger qualitatively different signals
than receptors located at the plasma membrane (Miac-
zynska et al., 2004; Sorkin and Von Zastrow, 2002). De-
spite the important role of endocytosis in the biological
activities of RTK, the molecular mechanisms controlling
this process remain poorly understood.
Epidermal growth factor (EGF) receptor (EGFR) has
been the primary experimental model used to study
the mechanisms of endocytosis. Ligand binding to
EGFR at the cell surface leads to rapid internalization
of ligand-receptor complexes through clathrin-coated
pits (Gorden et al., 1978). When enough activated
EGFR accumulates at the cell surface to saturate the
clathrin-dependent pathway, the resulting pool of acti-
vated receptors is instead internalized through slow,
clathrin-independent pathways (Sigismund et al., 2005;
Wiley, 1988). In either case, internalized ligand-receptor
complexes are transferred to early endosomes and sub-
sequently recycled back to the plasma membrane (Hop-
kins et al., 1985; Sorkin et al., 1991). After each round of
internalization, a significant fraction of internalized re-
ceptors are sorted from early endosomes to multivesic-
ceptors can be targeted to lysosomes for degradation
(Miller et al., 1986).
Ubiquitination of EGFR has been implicated in both li-
gand-mediated endocytosis and endosomal sorting.
Following EGF binding, EGFR is ubiquitinated by the
RING domain E3 ubiquitin (Ub) ligase, Cbl (Levkowitz
et al., 1998). The binding of Cbl to EGFR occurs directly
via phosphorylated Tyr1045 of EGFR, as well as indi-
rectly through Grb2. Whereas both interaction modes
are necessary for efficient ubiquitination of the receptor,
binding of Cbl to EGFR through a direct interaction with
Tyr1045 is dispensable for clathrin-dependent internali-
zation of EGFR (Huang and Sorkin, 2005; Jiang and Sor-
kin, 2003; Waterman et al., 2002).
Ubiquitination refers to the conjugation of Ub to the
3-amino group of a lysine residue within a substrate.
Monoubiquitination has previously been shown to be
necessary for ligand-mediated endocytosis of the a fac-
tor receptor in yeast (Hicke and Dunn, 2003). Previous
reports suggested that EGFR is exclusively modified
by monoubiquitin, consistent with the importance of
this modification in receptor sorting at the plasma mem-
brane and endosomes (Haglund et al., 2003; Mosesson
et al., 2003). While some substrates are modified only
by mono-Ub, poly-Ub chains can form through each of
the seven lysine residues within Ub itself (Peng et al.,
2003). Lys48-linked poly-Ub chains are generally be-
lieved to target proteins to the proteasome for degrada-
tion, while Lys63 linkages have been shown to regulate
processes such as DNA repair and kinase activation
through proteasome independent mechanisms.
Despite a wide array of experimental approaches, ev-
idence supporting the importance of EGFR ubiquitina-
tion in receptor trafficking remains indirect. Although
studies utilizing Tyr1045 mutation have shown this resi-
due to play a critical role in EGFR ubiquitination and
trafficking (Levkowitz et al., 1999), its effects can alter-
natively be attributed to impaired phosphorylation of
neighboring serine residues (Oksvold et al., 2003). Fur-
thermore, a series of endosomal proteins including Hrs
and ESCRT complexes I, II, and III, each containing
a Ub binding domain(s), are thought to participate in
trapping EGFR within internal vesicles of MVBs, al-
though the sequence of protein-protein interactions be-
tween ubiquitinated EGFR and ESCRT proteins during
this process has not been experimentally elucidated. It
is possible to inhibit EGFR degradation by RNAi knock-
down of Hrs and ESCRT proteins in human cells or ge-
netic knockout in mice (Babst et al., 2000; Bache et al.,
2003, 2004; Bishop et al., 2002). Since the Ub ligation
3These authors contributed equally to this work.
and binding proteins used in this process are shared by
many related pathways, knockdown strategies that
eliminate these components would also be expected
to produce multiple direct and indirect effects. Similarly,
it cannot be ruled out that the effects seen with Cbl mu-
tants on endocytosis result from deficient ubiquitination
of proteins besides EGFR (Jiang and Sorkin, 2003; Thien
et al., 2001). As such, direct identification of the ubiqui-
tion of these residues provides an opportunity to isolate
and directly test the functional significance of EGFR
ubiquitination in receptor internalization and trafficking.
In the current study, we have used tandem mass
spectrometry (LC-MS/MS) to identify specific lysine res-
idues within EGFR that are modified by Ub. Using the
Ub-AQUA technique, it was further possible to examine
the forms of Ub bound to the substrate. Surprisingly, we
found that EGFR is markedly polyubiquitinated, with
poly-Ub chains linked primarily through Lys63 of Ub.
Based upon data from LC-MS/MS analysis, it was pos-
sible to develop a series of EGFR mutants that were
deficient in growth-factor-stimulated ubiquitination. In
these mutants, Ub-mediated receptor internalization
was uncoupled from lysosomal degradation, revealing
distinct Ub-dependent mechanisms for these two pro-
Results and Discussion
Ubiquitin Conjugation Sites Are Located
in the Kinase Domain of EGFR
To identify ubiquitination sites on the EGFR, large
amounts of receptors were purified from porcine aortic
endothelial (PAE) cells stably expressing wild-type (wt)
human EGFR. Cells were untreated or treated with
EGF for 2 min at 37ºC to stimulate maximum EGFR ubiq-
uitination. EGFR was immunoprecipitated using an anti-
body to the extracellular domain of the receptor and
stringently washed to minimize coprecipitation of other
proteins. Immunopurified EGFR was separated by
SDS-PAGE and visualized by Coomassie blue or silver
staining (Figure 1A). In order to analyze ubiquitinated
EGFR by mass spectrometry, the gel region containing
Ub-modified protein was excised, digested with trypsin,
and analyzed by LC-MS/MS using a hybrid-linear ion
trap Fourier transform mass spectrometer. This instru-
ment made it possible to collect a high accuracy precur-
sor ion measurement (<10 ppm) and an MS/MS spectra
for each peptide ion. Database searching of these MS/
MS spectra revealed that EGFR was the predominant
protein identified in these samples.
Numerous reports have demonstrated that trypsin di-
gestionofaubiquitinated protein produces unique 2GG
signature peptides, which can be analyzed by standard
LC-MS/MS techniques and used to identify substrate
ubiquitination sites (Kirkpatrick et al., 2005). These sig-
nature peptides can be identified in database searches
by modifying search parameters to allow for an addi-
tional mass of 114.0429 Da (corresponding to the 2GG
Ub remnant) on lysine residues. Using this approach,
we identified six distinct ubiquitination sites on EGFR.
MS/MS spectra obtained during this analysis matched
of EGFR at lysines 2692, 2713, 2730, 2843, 2905, and
are shown that demonstrate ubiquitination at lysine 843
(Figure 1C) and lysine 905 (Figure 1D). In the case of the
six 2GG signature peptides identified, the mass devia-
tion between the observed and expected precursor
ions was found to be between 2 and 10 ppm. In each
case, manual validation independently confirmed the
ubiquitination site assignment.
Sequence analysis revealed that each of the ubiquiti-
nation sites identified map to the kinase domain of
EGFR. Using a previously reported structure of the
EGFR kinase domain (Stamos et al., 2002), we found
that these six lysines are distributed throughout the sur-
face of the domain (Figure 2A). Although Western blot-
ting revealed that the total amount of Ub conjugated to
the stably expressed Y1045F mutant EGFR was approx-
imately an order of magnitude lower than in wt EGFR
(Figure 1A), LC-MS/MS analysis revealed that at least
two of these sites, Lys713 and Lys843, could still be
ubiquitinated in the Y1045R mutant. In contrast, no
2GG signature peptides were identified in samples con-
taining unstimulated wt EGFR.
To investigate the effects of EGFR ubiquitination on
receptorfunction,lysine residuescorresponding tospe-
cific EGFR ubiquitination sites were mutated to argi-
nines. Single and multiple site-point mutants (nKR) of
five, six, or nine lysines (Figure 2B) were generated
and transiently expressed in PAE cells. In the 9KR mu-
tant, besides the six lysines found to be modified by
Ub, three additional lysines (Lys689, Lys690, and
Lys733) situated in close proximity to confirmed ubiqui-
tination sites were also substituted with arginines (Fig-
ure 2B). Western blot analysis revealed that single muta-
tions of all individual sites and various combinations of
two or three mutations did not result in a significant de-
crease in the ubiquitination of EGFR (data not shown).
However, ubiquitination of 5KR, 6KR, and 9KR mutants
was reduced by w70%–80% as compared to wtEGFR
(Figure 2B). Similarly, cells stably expressing 5KR or
Y1045F mutant forms of the receptor displayed weak
ubiquitination (about 20% compared to wtEGFR)
(Figure 2C). In all of the mutants, residual ubiquitination
was observed following EGF stimulation.
Despite multiple Lys/Arg mutations within the ki-
nase domain, the extent of tyrosine phosphorylation
for KR mutants was essentially similar to that of wt re-
ceptors (Figures 2B and 2C). All nKR mutants stably ex-
pressed in PAE cells were capable of activating MAPK/
a comparable extent as in cells expressing wt EGFR
(data not shown, see also Figure 5 below). These data
suggest that multiple lysine-to-arginine substitutions
within the kinase domain had a direct effect on ubiquiti-
nation but did not affect the tyrosine kinase activity of
EGFR or its signaling capacity.
EGFR Is Polyubiquitinated
It hasbeen previously reported thatEGFRubiquitination
occurs exclusively in the form of mono-Ub (Haglund
et al., 2003; Mosesson et al., 2003). As a part of LC-
MS/MS analysis of ubiquitinated EGFR samples, we ob-
and K11-linked poly-Ub chains. To quantify the various
forms of Ub bound to EGFR, samples were analyzed us-
ing the newly developed Ub-AQUA method (D.K. and
S.G., unpublished data). This method utilizes a series
of isotope-labeled internal standard peptides and se-
lected reaction monitoring (SRM) experiments to quan-
tify each poly-Ub chain linkage as well as mono-Ub.
Analysis of samples from unstimulated cells revealed
a low background level of Ub. Following EGF stimula-
tion, 2.91 pmol total Ub was measured in the gel region
immediately above EGFR (>170 kDa), equating to an in-
crease of more than 20-fold over unstimulated receptor
than half of all EGFR-associated Ub was found not as
mono-Ub but rather in the form of poly-Ub chains.
Poly-Ub chains linked through K63 were more abundant
than all other linkages combined, making up 41.8% of
the total Ub. In addition to 1.22 pmol K63 branched pep-
tide, 0.20 pmol K48, 0.12 pmol K11, and 0.03 pmol K29
chains were also detected (Figure 3B).
Because of weak ubiquitination of 5KR and Y1045F
mutants, Ub-AQUA analysis required higher amounts
of starting material than corresponding analysis of wt
EGFR. Despite less efficient overall ubiquitination, the
linkage profile for Ub bound to mutant forms of EGFR
was strikingly similar to what was seen in wt EGFR
The detection of EGFR polyubiquitination was sur-
prising in light of published data showing that EGFR is
exclusively monoubiquitinated (Haglund et al., 2003;
Mosesson et al., 2003). It is difficult to reconcile our re-
sults with the latter reports, which based their conclu-
sions on the inability of FK1/2 antibodies to recognize
both mono-Ub and ubiquitinated EGFR and on the use
of Ub mutants deficient in polyubiquitination. This result
ognize K63-linked Ub chains. Alternatively, since trans-
fection of mutant Ub does not eliminate the abundant
pools of endogenous Ub, it is likely that mutant Ub
Figure 1. Identification of Multiple Ubiquitination Sites in the Kinase Domain of the EGFR by LC-MS/MS
(A) PAE cells stably expressing wtEGFR or Y1045F mutant were untreated or treated with 20 ng/ml EGF for 2 min at 37ºC, and EGFR was immu-
noprecipitated. Representative gels analyzed by LC-MS/MS for the identification of EGFR ubiquitination sites are shown. Western blots against
Ub, run in parallel with Coomassie- or silver-stained gels, were used as guides while cutting gel regions containing Ub-EGFR.
(B) Six ubiquitination sites were identified from within the kinase domain of EGFR. Each peptide was characterized by a 2GG modified lysine,
bearing an additional mass of 114.0429 Da. Peptides matching MS/MS spectra were only accepted if they varied by less than 10 ppm from the
expected monoisotopic mass of the parent ion.
(CandD)RepresentativeMS/MSspectra ofpeptides demonstrating ubiquitination at(C)lysine 843and (D)lysine 905of theEGFR. Peaks match-
ing expected singly and doubly charged (++) b and y ions are labeled.
EGF Receptor Ubiquitination and Endocytosis
acts as an endcap on polyubiquitin chains formed with
endogenous Ub. Regardless of the nature of the dis-
crepancy, mass spectrometry analysis of immuno-
purified EGFR unequivocally demonstrated K63-linked
polyubiquitination of EGFR that was proportionally re-
duced in two distinct EGFR mutants (Figures 2C and
3C). K63-linked polyubiquitin chains are proposed to
function in DNA repair, IkB kinase activation, transla-
tional regulation, and endocytosis in yeast, although
their mechanistic role in these processes is still unclear
(Deng et al., 2000; Galan and Haguenauer-Tsapis, 1997;
Hofmann and Pickart, 1999; Spence et al., 2000). Thus,
predominant K63 polyubiquitination is consistent with
the nonproteosomal function of EGFR ubiquitination.
Ubiquitination Sites Are Essential
for EGFR Degradation
EGFR ubiquitination has been implicated in the regula-
tion of EGFR trafficking. To directly test this hypothesis,
wt and nKR mutant forms of EGFR were transiently
expressed in PAE cells and used to measure growth-
factor-stimulated EGFR turnover rates. Treatment of
cells with EGF led to degradation of wt EGFR with a
half-life of 1.5–2.0 hr (Figures 4A and 4B). In contrast,
mained virtually unchanged after 4 hr incubation of cells
with EGF. Similarly, EGF-induced degradation of the
5KR mutant receptor was much slower than that of the
wt receptor in two separate stable cell lines (Figures
4C and 4D). Moreover, measurements of the number
of125I-EGF binding sites at the cell surface upon EGF
treatment revealed a very low rate of downregulation
of 5KR mutant (Figure 4E).
Increased stability of mutant EGFRs in cells stimu-
lated with EGF suggests that signaling by these EGFR
mutants can be prolonged. The time course of EGFR ac-
tivation (phosphorylation of the one of the major EGFR
sites, Tyr1086) and MEK1/2 was comparable in cells ex-
pressing wt and 5KR receptors (Figure 5A). Probing
EGFR with antibodies to pTyr1068 and pTyr1148 pro-
duced essentially similar results (data not shown). Sig-
nificant differences in the stability of activated EGFR
were observed only in the presence of EGF concentra-
tions that cause downregulation of the total EGFR pro-
tein (R20 ng/ml). Under these conditions, the amount
of active wt EGFR dramatically decreased after 2 hr in-
cubation with EGF, whereas the level of activity of 5KR
mutants remained unchanged (Figure 5B). However, de-
tors, the extent of MEK1/2 activation by both receptors
was similar (Figure 5B). Interestingly, after comparable
initial activation (5 min of EGF stimulation), the amount
of activated 5KR receptors decayed more slowly than
that for wt EGFR (Figure 5C). Dephosphorylation of
MEK1/2 after removal of EGF from the medium was fol-
lowed by a new ‘‘late’’ wave of MEK1/2 activation. The
amplitude of late MEK1/2 activation was higher in cells
expressing 5KR mutant than in wt EGFR-expressing
cells. It is possible that an elevated, late MEK1/2 activity
is due to slow degradation and increased recycling of
activated 5KR receptors. Nevertheless, the data in Fig-
ure 5 suggest that, under physiological conditions of
the continuous presence of EGF, elimination of receptor
ubiquitination does not lead to increased MEK/ERK
Figure 2. Mutations of Ubiquitination Sites in the EGFR Result in Im-
paired Ubiquitination but Normal Tyrosine Phosphorylation of the
(A) Ub-modified lysines were mapped on a previously reported
structure (1M14) of the EGFR kinase domain. Underlined are ubiqui-
tination sites identified in both wtEGFR and Y1045F mutant.
(B) Lysines depicted in (A) were mutated to arginines (5KR and 6KR).
Besides the six sites identified by LC-MS/MS, three structurally ad-
jacent lysines (K689/K690/K733) from the kinase domain were also
mutated (9KR). Wt and mutant receptors were transiently expressed
in PAE cells, and EGFR immunoprecipitates were probed with Ub,
phosphotyrosine, and EGFR antibodies. The mean amount of Ub
and phosphotyrosine normalized to total EGFR (6SD) from four ex-
periments is presented on the graphs.
(C) EGFR was immunoprecipitated from PAE cells stably expressing
wtEGFR (clone B2), 5KR (clone #1), and Y1045F mutants (clone #1),
and the extent of ubiquitination and phosphorylation was deter-
mined in three experiments as in (B). The mean amount of Ub and
phosphotyrosine normalized to total EGFR (6SD) from four experi-
ments is presented on the graphs.
signaling in PAE cells. Presumably, the extent of MAPK
activation is limited by saturable factors and processes
downstream of EGFR.
The inhibition of EGF-induced degradation of nKR
mutants observed in Figures 4 and 5 could be due to
slow internalization or impaired lysosomal degradation
of EGF-receptor complexes. Our previous results have
suggested that transient transfection leads to overex-
pression of the receptor in the majority of cells so that
the pathway responsible for rapid internalization be-
comes saturated (Carter and Sorkin, 1998). Likewise, in-
ternalization of125I-EGF in cells transiently expressing
wt or nKR mutant EGFRs was similarly slow as com-
pared with the maximum high rates measured for
clathrin-dependent EGFR internalization in PAE cells
(Figure 6A). As in the case of transient transfections,
treatment of cells stably expressing EGFR with saturat-
ing EGF concentrations results in receptor internaliza-
tion primarily through a slow, clathrin-independent
pathway. No difference was observed between the re-
ing cell lines under these conditions (Figure 6A). There-
fore, the defect in the turnover of nKR mutants is likely
due to the blockade of lysosomal targeting rather than
the internalization step of EGFR trafficking.
To further explore the nature of the delayed downreg-
ulation of ubiquitination-deficient receptor mutants,
the localization of EGF-occupied 5KR mutant was ana-
lyzed. After 10 min of cell stimulation with EGF conju-
gated with rhodamine (EGF-Rh) or fluorescein (EGF-
FC), both wt EGFR and 5KR mutant were well colocal-
ized with the marker ofearly endosomes, EEA.1 (Figures
7A and 7B). At the same time, very little, if any, localiza-
tionof bothreceptors inlateendosomes and
Figure 3. Polyubiquitinaiton of EGFR Re-
vealed by Ub-AQUA Analysis
(A) PAE/EGFR cells were untreated or treated
with 20 ng/ml EGF for 2 min at 37ºC, and
EGFR was immunoprecipitated. Representa-
tive gel analyzed by LC-MS/MS for the quan-
tification of EGFR ubiquitination and corre-
(B) Absolute amounts (fmol) of total Ub and
each poly-Ub chain linkage measured in the
analysis of the gel region containing multiubi-
quitinated EGFR (Ubn-EGFR) and lower re-
gion containing mainly unmodified EGFR
(C) PAE cell stably expressing wt, Y1045F,
and 5KR mutants of EGFR were treated with
EGF as in (A), and EGFRs were immunopre-
cipitated. The results of ubiqutin-AQUA anal-
ysis are presented for each poly-Ub chain
linkage and mono-Ub/endcaps as a percent-
age of total EGFR-associated Ub.
EGF Receptor Ubiquitination and Endocytosis
lysosomes, marked by LysoTrackerRed, was observed
(Figures 7C and 7D). After 3 hr of continuous incubation
of cells expressing wt EGFR with labeled EGF, about
70% of EGF-FC was accumulated in late endosomes/
lysosomes (Figures 7C and 7D) while very little, if any,
EGF-Rh could be detected in early endosomes (Figures
7A and 7B). In contrast, a significant pool of 5KR mutant
remained in EEA.1-positive compartments (Figures 7A
and 7B), and only a small amount of the mutant receptor
reached late endosomes/lysosomes (Figures 7C and
7D). Together with the poor downregulation of the sur-
that5KR mutant isnot efficiently sorted tothe lateendo-
some/lysosome pathway and instead accumulated in
early endosomal compartments where it is extensively
recycled to the plasma membrane.
The establishment of the role of EGFR ubiquitination
in the lysosomal targeting and degradation of the recep-
tor and the detection of marked EGFR K63 polyubiquiti-
nation raise the question of what role Ub chains play in
EGFR trafficking. Structural studies of K63-linked diubi-
qutin have revealed that each of the two ubiquitins can
bind the Ub binding domain in the same mode as does
monoubiquitin (Varadan et al., 2004). It can, therefore,
be suggested that K63 chains allow for multivalent inter-
somes, such as Hrs and TSG101, which may be neces-
sary to allow an efficient sorting of the ubiquitinated
receptors to internal vesicles of MVBs. A recent study
using mutant Ub lacking Lys63 has proposed that poly-
ubiquitination through this residue can mediate endocy-
tosis of TrkA RTK (Geetha et al., 2005). Although an E3
Figure 4. Mutations of Ubiquitination Sites
Inhibit EGFR Downregulation
(A) wtEGFR, 5KR, 6KR, or 9KR mutants were
transiently expressed in PAE cells. The cells
were serum starved and incubated with EGF
(100 ng/ml) for indicated times before lysis
tected with antibodies 2913.
(B) The amount of EGFR immunoreactivity
was quantitated from three experiments,
and the mean values (6SD) were plotted
(C) Cells stably expressing wt or 5KR mutant
(clones #1 and #4) were incubated with EGF
asin (A) for indicated times. EGFR was immu-
nodetected as in (A).
(D) The amount of EGFR immunoreactivity
was quantitated from three experiments per-
formed as in (C), and the mean values (6SD)
were plotted against time.
(E) Cells stably expressing wt or 5KR mutant
(clone #1) were incubated with EGF as in (A)
for indicated times. The cells were then acid
washed to remove unlabeled EGF from sur-
face receptors, and the number of surface
EGF binding sites was measured using incu-
bation with125I-EGF (50 ng/ml) at 4ºC for 1 hr.
Each data point represents a value averaged
from three experiments (6SD).
ferent than those controlling EGFR ubiquitination, the
hypothesis can be put forward that K63-linked Ub
chains could be a common signal in RTK trafficking.
Major Ubiquitination Sites Are Not Essential
for EGFR Internalization
To compare rates of internalization of wt and mutant
EGFR through the clathrin-dependent pathway, cell
Figure 5. Signaling by 5KR Mutant
(A)PAEcellsstably expressing wtEGFRor5KR(clone#1)wereserumstarved, treatedwith 10ng/mlEGF forindicated timesat37ºC,andlysed in
the presence of OV and NEM. The cell lysates were probed for active EGFR (antibody pY1086), total EGFR (Ab2913), phosphorylated MEK1/2,
and total MEK. The experiment is representative of three independent experiments.
Active EGFR and MEK1/2 were detected as in (A). All steps of the experiment were performed identically for wt EGFR and mutant-expressing
cells. The experiment is representative of three independent experiments. EGF-induced downregulation of EGFR was accompanied by the ap-
pearance of a low molecular mass EGFR form (w160 kDa), a product of partial proteolysis of EGFR that is not detected by antibodies to active
EGFR (A and B). Note that Ab2913 recognized the EGF-activated full-length EGFR (w175 kDa) with a lesser efficiency than the truncated EGFR
EGF. Active EGFR and MEK1/2 were detected in cell lysates as in (A). The mean amounts of active EGFR and MEK1/2 normalized to total EGFR
and MEK1/2, respectively, from three experiments (6SD) plotted against chase time are presented on the graphs at the right.
EGF Receptor Ubiquitination and Endocytosis
lines stably expressing these mutants were used. In
these cell lines, it is possible to measure the rates of
EGFR internalization via rapid, physiological pathway
of endocytosis by using low concentrations of EGF. Be-
cause of different expression levels of receptors in var-
tosis can be saturated at different occupancies of
surface EGFR. Therefore,125I-EGF internalization rates
of wt and mutant EGFR were measured using a range
of EGF concentrations to compare maximum high inter-
nalization rates. Figure 6A shows that, although there
was significant variation in the maximal rates of internal-
ization, all nKR cell lines internalized125I-EGF within the
range of the internalization rates of EGFR reported in
various types of cells (>0.2/min) (Jiang et al., 2003;
Huang et al., 2004; Wiley, 1988).125I-EGF internalization
rate constants in nKR-expressing cells were also com-
parable with these constants measured for the transfer-
clathrin coated pits (Huang et al., 2004). These data sug-
gested that complete EGFR ubiquitination is not essen-
tial for clathrin-dependent internalization of EGFR.
Internalization of Lysine Mutants of the EGFR
Is Mediated by Grb2 and Cbl
If internalization of nKR mutant EGFR does not require
efficient ubiquitination, is it regulated by the same
mechanisms as the internalization of wt receptors?
Using a combination of siRNA and dominant-negative
overexpression approaches, we previously demon-
strated that clathrin-dependent endocytosis requires
Grb2 and its interaction with Cbl proteins in PAE and
other cells (Huang and Sorkin, 2005; Jiang et al., 2003;
proach, we found that Grb2 is necessary for clathrin-
dependent125I-EGF internalization by 5KR and 9KR re-
ceptors (Figure 6B).
To test for the importance of Cbl proteins in EGFR
internalization, siRNA duplexes targeted to c-Cbl and
Cbl-b were generated. When transfected together,
Figure 6. Internalization of EGFR Is Grb2 and
Cbl Dependent but Is Not Affected by Muta-
tions of Ubiquitination Sites
(A) On the left, internalization rate constants
(ke) of125I-EGF (1 ng/ml) were measured in
cells transiently expressing wt, 5KR, 6KR, or
9KR receptors. On the right, kevalues were
measured in cells stably expressing wtEGFR
(clones B2 and B11) or various nKR mutants
treated with 0.5–40 ng/ml125I-EGF and plot-
ted against the number of EGFR occupied
by125I-EGF at the cell surface after 5 min in-
cubation at 37ºC (acid-wash-sensitive pool).
(B) Internalization rates of wtEGFR and stably
expressed 5KR and 9KR mutants were mea-
sured using 1 ng/ml125I-EGF in cells trans-
fected with control siRNA (mock), Grb2
siRNA, or c-Cbl/Cbl-b siRNAs. The extents
of c-Cbl/Cbl-b protein depletion are shown
by Western blotting (on the right).
(C) 5KR cells were transfected with wt c-Cbl-
YFP or 70Z-Cbl-YFP. The cells were incu-
bated with 2 ng/ml EGF-Rh for 5 min at 37ºC
and fixed. Z stacks of images of YFP and rho-
damine fluorescence were acquired through
FITC and Cy3 filter channels, respectively.
Optical sections through the middle of the
cell are shown. Insets represent high-magni-
fication and high-intensity images of EGF-
Rh in the periphery of 70Z-Cbl-expressing
cells that show localization of EGF-Rh in
small clusters at the cell surface under condi-
tions of inhibited endocytosis. Scale bars,
(D) Quantification of the amount of EGF-Rh in
nonexpressing (non-exp), c-Cbl-YFP, and
70Z-Cbl expressing cells from experiments
performed as in (C) (5KR#1) and Figure S2
(9KR#1). Each data point represents a value
averaged from ten to 20 cells (6SD). Cells
containing more than 2 3 108a.l.u.f.i. (arbi-
trary linear units of fluorescence intensity) of
YFP were used for quantification.
these siRNAs caused significant depletion of both Cbl
proteins and a substantial inhibition of125I-EGF internal-
ization in cells expressing wt, 5KR, or 9KR mutant EGFR
(Figure 6B). Together with the observation of essentially
similar kinetics of internalization compared to wt EGFR,
the RNA interference experiments suggested that 5KR
Figure 7. Trafficking of 5KR Mutant Is Delayed in Early Endosomes
Afterfixation,thecellswerestainedwith antibodytoEEA.1.AZstackofimageswasacquired throughCy3(EGF-Rh)andFITC (EEA.1)filterchan-
nels and deconvoluted. Optical sections through the middle of the cell are shown. Arrows indicate examples of colocalization of EGF-Rh and
EEA.1. Scale bars, 10 mm.
(B) The relative amount of EGF-Rh in EEA.1-containing endosomes was calculated from two experiments performed as in (A). Each data point
represents value averaged from ten cells (6SD).
(C) Cells were preincubated with leupeptine and then incubated with EGF-FC (100 ng/ml) as in (A). The cells were incubated with LysoTracker for
of the tail are shown. Arrows indicated on the examples of colocalization of EGF-FC and LysoTracker. Scale bars, 10 mm.
(D) The relative amount of EGF-FC in vesicles containing LysoTrackerRed was calculated from two experiments performed as in (C). Each data
point represents value averaged from ten cells (6SD).
EGF Receptor Ubiquitination and Endocytosis
and 9KR mutant receptors are internalized using the
mechanisms normally operating during rapid internali-
zation of the wt EGFR. These experiments also directly
demonstrated that Cbl is required for clathrin-depen-
dent EGFR internalization. The data obtained using
siRNA Cbl depletion are consistent with the previous re-
ports of dominant-negative effects of Cbl mutants on
ity of Cbl-Grb2 chimera to rescue EGFR endocytosis in
Grb2-depleted cells (HuangandSorkin, 2005).However,
the same data are inconsistent with the observation of
normal endocytosis in mouse embryonic fibroblasts
with targeted knockout of c-Cbl (Duan et al., 2003).
One possible explanation for this discrepancy is the
presenceofCbl-b inc-Cblknockout cells. Infact,insev-
eral types of cells, only highly efficient knockdown of
both c-Cbl and Cbl-b allowed the demonstration of Cbl
requirement for EGFR internalization (data not shown).
We and others have previously demonstrated that the
intact Cbl RING domain is necessary for EGFR internal-
ization (Huang and Sorkin, 2005; Jiang and Sorkin, 2003;
Thien et al., 2001). To confirm that this domain is also re-
quired for internalization of ubiquitination-deficient
EGFR mutants, a naturally occurring c-Cbl mutant lack-
ing RING domain activity, 70Z-Cbl, was overexpressed
to examine its effects on EGFR internalization. Because
the efficiency of transient transfection of PAE cells is not
sufficient for biochemical endocytosis assays, single-
EGF-Rh. As shown in Figure 6C, accumulation of EGF-
Rh in peripheral and perinuclear endosomes of PAE/
5KR cells was not affected by overexpression of wt
c-Cbl. However, endocytosis was impaired in PAE/
5KR and PAE/9KR cells overexpressing 70Z-Cbl mu-
tant, as revealed by very few endosomes containing
EGF-Rh (Figure 6C and Figure S2) and significant reduc-
As in our previous studies using PAE cells expressing
wtEGFR (Jiang and Sorkin, 2003), EGF-Rh in PAE/5KR
cells overexpressing 70Z-Cbl was localized in small
dots, previously identified as clathrin-coated pits, and
also diffusely distributed throughout the plasma mem-
brane (see insets in Figure 6C and Figure S2). Thus,
despite its ubiquitination defect, internalization of the
5KR and 9KR mutants still required an intact RING do-
main of Cbl.
Two models can be proposed to explain the require-
ment of Cbl and its RING domain but not efficient
EGFR ubiquitination in this process. One possibility is
that Cbl RING domain is necessary to recruit proteins
besides EGFR to the pit. In this model, recruitment
of EGFR-associated proteins may or may not be Ub de-
pendent (requiring the RING activity). In fact, it has been
proposed that ubiquitination of a component of endocy-
tosis apparatus is required for endocytosis of a factor
receptor in yeast (Dunn and Hicke, 2001) and growth
hormone receptor in mammalian cells (Govers et al.,
1999). While the participation of other ubiquitination tar-
gets of Cbl may be the case, we believe that our results
point toward a multiple threshold model in which each
step in the internalization and trafficking of EGFR re-
quires more extensive ubiquitination. In this case, resid-
ualubiquitination ofthe receptorinthe absenceofmajor
ubiquitination sites, as was seen here, would be pre-
dicted to be sufficient for receptor internalization but
not further postendocytic trafficking. This model as-
sumes that the strength of the interaction of weakly
ubiquitinated EGFR with Ub binding proteins in clathrin-
coated pits is higher than in endosomes, since the
same residual ubiquitination is not sufficient for EGFR
trafficking to MVBs or its lysosomal degradation. Thus,
we propose that the density of Ub modifications on
EGFRsets regulatory thresholds allowing internalization
and endosomal sorting steps of EGFR trafficking to be
EGF-Rh, EGF-FC, and LysoTrackerRed were purchased from Mo-
lecular Probes (Eugene, Oregon). Monoclonal antibodies to c-Cbl,
phosphotyrosine (PY20), and EEA.1 were from BD Transduction
Laboratories (San Diego, California) and monoclonal antibody to
EGFR (Ab528) from American Type Culture Collection (Manassas,
Virginia).Monoclonal antibody to Ub (P4D1) and polyclonalantibody
to Cbl-b were from Santa Cruz Biotechnology (Santa Cruz, Califor-
nia). Polyclonal antibody to phosphorylated MEK1/2, monoclonal
antibody to total MEK1/2, and polyclonal antibody to EGFR phos-
photyrosine 1086 (pY1086) were from Cell Signaling Technology
(Beverly, Massachustts). Rabbit serum Ab2913 to the intracellular
domain of EGFR was kindly provided by Dr. L. Beguinot (DIBIT Ra-
faele, Milan, Italy) and polyclonal antibody to actin was from Sigma
(St. Louis, Missouri).
Plasmids and Mutations
The c-Cbl-YFP (wt and 70Z) and EGFR (wt and Y1045F mutant) con-
structs were described previously (Jiang et al., 2003; Jiang and Sor-
kin, 2003). Point mutations were generated using the QuikChange
Site-Directed Mutagenesis Kit according to the manufacture proto-
col (Stratagene Cloning Systems, La Jolla, California). All point mu-
tations were verified by automated dideoxynucleotide sequencing.
Cell Culture and Transfections
PAE cells were grown and transiently DNA transfected as described
(Jiang et al., 2003). Transfected cells were split 1 day after transfec-
tion and used for experiments on the second day. PAE cell lines sta-
bly expressing various mutated EGFRs were generated as previ-
ously described (Jiang et al., 2003).
In siRNA knockdown experiments, Grb2 siRNA duplex (Jiang
et al., 2003) or c-Cbl (967–985, target sequence 50-CCTCTCTTCC
AAGCACTGA-30) and Cbl-b (414–432, target sequence 50-GGACAG
ACGAAATCTCACA-30) siRNA duplexes together were resuspended
in 13 siRNA universal buffer (Dharmacon, Inc., Lafayette, Colorado)
to 20 mM prior to transfection. PAE/EGFR cell lines grown in 6-well
plates (50%–60% confluency) were transfected with 10 ml Grb2
siRNA3 duplex or a mixture of 5 ml c-Cbl siRNA and 5 ml Cbl-b siRNA
duplexes in 5 ml DharmaFECT I reagent (Dharmacon) for 3 days.
For mock transfections, control cyclophilin siRNA (Dharmacon)
was used. Cells were replated to 12-well dishes 24 hr prior to experi-
LC-MS/MS Analysis of EGFR Ubiquitination
For mass spectrometry analysis, PAE cells stably expressing wt or
mutant EGFR were grown in 150 mm dishes. Cells were untreated
or treated with 20 ng/ml EGF, washed with Ca2+, Mg2+-free phos-
phate buffered saline (CMF-PBS), and lysed in Triton X-100/glyc-
erol/HEPES solubilization buffer (TGH) (Huang et al., 2004) contain-
ing 1% sodium deoxycholate and 10 mM N-ethyl-maleimide (NEM).
Lysates were cleared by centrifugation (10 min, 14,000 3 g). EGFR
was immunoprecipitated with antibody Ab528. The precipitates
were washed three times with TGH/DOC containing, sequentially,
500 mM, 100 mM, and no NaCl and resolved by SDS-PAGE.
Gel regions containing Ub-EGFR were excised, destained, and di-
gested using 15 ng/ml trypsin. Following extraction, digested pep-
tides were analyzed on an LTQ-FT hybrid linear ion trap Fourier
transform mass spectrometer equipped with a Famos Autosampler
(LC Packings, San Francisco, California) and an Agilent 1100 Series
HPLC pump (Agilent Technologies, Palo Alto, California). Peptides
were first loaded onto fused silica microcapillary column packed
with reversed phase C18 material (Magic C18AQ, Michrom Biore-
dient of 5%–40% buffer B (97.5% acetonitrile/0.15% formic acid) in
buffer A (2.5% acetonitrile/0.15% formic acid). The LTQ-FT was op-
erated in data-dependent top ten mode as has been described
(Haas et al., 2006). To obtain maximum sensitivity, automatic gain
control (AGC) settings of 3 3 106for MS (in the FT-MS) and 2 3
104for MS/MS (in the linear ion trap-MS) were employed. Protein
identifications were made by database searching an indexed ver-
sion of the nonredundant database available from NCBI. To identify
modification with a mass of 114.0429 Da in a simplified form of the
human database containing protein sequences for EGFR, ubiquitin,
and known contaminants.
Immunoprecipitations and Western Blotting
To examine ubiquitination and tyrosine phosphorylation of EGFR by
Western blotting, PAE cells stably expressing EGFR or transiently
transfected with EGFR constructs in 60 mm dishes for 2 days were
pretreated with 20 ng/ml EGF for 2 min at 37ºC and washed with
CMF-PBS. The cells were lysed in TGH/NEM, and EGFR was immu-
degradation, the cells stably or transiently expressing EGFR were
serum starved overnight and incubated with EGF (100 ng/ml) for in-
dicated times. The cells werelysed in TGH asabove, with the excep-
tion that orthovanadate (OV) and NEM were omitted from the lysis
buffer. This was necessary to provide equal efficiency of Western
blot detection of inactive and activated EGFR by Ab2319.
To probe for active EGFR and MEK1/2, cells in 12- or 6-well plates
times, and lysed in TGH buffer containing 1 mM sodium OV and
10 mM NEM.
Western blotting of cell lysates and EGFR immunoprecipitates
was performed as described (Huang et al., 2004). Several X-ray films
were analyzed to determine the linear range of the chemilumines-
cence signals, and the quantifications were performed using densi-
Mouse receptor-grade EGF (Collaborative Research Inc., Bedford,
Massachusetts) was iodinated as described previously (Jiang
et al., 2003).
40 ng/ml125I-EGF, and the specific rate constant for internalization
kewas calculated as described previously (Jiang et al., 2003).
125I-EGF internalization was measured using 0.5–
The number of surface EGFR before and after EGF treatment was
(Carter and Sorkin, 1998).
125I-EGF binding assay as previously described
PAE/EGFR cells grown on glass coverslips were transiently trans-
fected with Cbl-YFP. The cells were treated with 2 ng/ml EGF-Rh
at 37ºC for 5 min, washed with ice-cold phosphate buffer saline,
and fixed with freshly prepared 4% paraformaldehyde (Electron
Microscopy Sciences, Ft. Washington, Pennsylvania). A Z stack of
images was acquired and deconvoluted using a Marianas imaging
workstation and SlideBook 4.1 software (Intelligent Imaging Innova-
tion, Inc, Denver, Colorado) as described previously (Huang and
Sorkin, 2005). The amount of EGF-Rh in individual cells that do not
express or express Cbl constructs was calculated as described
(Huang and Sorkin, 2005).
To determine the extent of colocalization of EGFR with EEA.1, the
cells were preincubated with leupeptine (21 mM) for 1 hr at 37ºC to
block lysosomal degradation and then incubated with 100 ng/ml
EGF-Rh in the same medium at 37ºC for indicated times. The cells
were then fixed as above and stained with EEA.1 antibody followed
by secondary antibody conjugated with fluorescein, all incubations
in the presence of 0.02% saponin. To determine the extent of coloc-
alization of EGFR and LysoTrackerRed, the experiments were per-
formed as above, with the exception that EGF-FC was used instead
of EGF-Rh, and the cells were fixed and imaged without permeabi-
lization. The cells were incubated with LysoTracker (50 nM) for
30 min at 37ºC. The experiments were designed so that the 30 min
incubation ended at the end of cell incubation with EGF-FC. In
both types of experiments, a Z stack of two-dimensional images
was acquired through FITC and Cy3 filter channels and deconvo-
luted using constrained iterative algorithm. The method of quantifi-
cation of the relative amount of EGF-Rh or EGF-FC colocalized
with EEA.1 or LysoTrackerRed, respectively, is described in Supple-
Supplemental Data include Supplemental Experimental Procedures
and two figures and can be found with this article online at http://
structure of the EGFR kinase and Drs. W. Haas and S. Gerber for ad-
vice on using high-accuracy precursor scans to validate spectral
identifications. This work was supported by APRC and CA089151
grants from NCI.
Received: December 5, 2005
Revised: January 23, 2006
Accepted: February 16, 2006
Published: March 16, 2006
Babst, M., Odorizzi, G., Estepa, E.J., and Emr, S.D. (2000). Mamma-
lian tumor susceptibility gene 101 (TSG101) and the yeast homo-
logue, Vps23p, both function in late endosomal trafficking. Traffic
Bache, K.G., Raiborg, C., Mehlum, A., and Stenmark, H. (2003).
STAM and Hrs are subunits of a multivalent ubiquitin-binding com-
plex on early endosomes. J. Biol. Chem. 278, 12513–12521.
and Stenmark, H. (2004). The growth-regulatory protein HCRP1/
hVps37A is a subunit of mammalian ESCRT-I and mediates receptor
down-regulation. Mol. Biol. Cell 15, 4337–4346.
Bishop, N., Horman, A., and Woodman, P. (2002). Mammalian class
E vps proteins recognize ubiquitin and act in the removal of endoso-
mal protein-ubiquitin conjugates. J. Cell Biol. 157, 91–102.
Carter, R.E., and Sorkin, A. (1998). Endocytosis of functional epider-
mal growth factor receptor- green fluorescent protein chimera.
J. Biol. Chem. 273, 35000–35007.
Deng, L., Wang, C., Spencer, E., Yang, L., Braun, A., You, J., Slaugh-
ter, C., Pickart, C., and Chen, Z.J. (2000). Activation of the IkappaB
kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating
enzyme complex and a unique polyubiquitin chain. Cell 103, 351–
Dikic, I., and Giordano, S. (2003). Negative receptor signalling. Curr.
Opin. Cell Biol. 15, 128–135.
Duan, L., Miura, Y., Dimri, M., Majumder, B., Dodge, I.L., Reddi, A.L.,
Ghosh, A., Fernandes, N., Zhou, P., Mullane-Robinson, K., et al.
(2003). Cbl-mediated ubiquitinylation is required for lysosomal sort-
ing of epidermal growth factor receptor but is dispensable for endo-
cytosis. J. Biol. Chem. 278, 28950–28960.
Dunn, R., and Hicke, L. (2001). Multiple roles for Rsp5p-dependent
ubiquitination at the internalization step of endocytosis. J. Biol.
Chem. 276, 25974–25981.
Galan, J.M., and Haguenauer-Tsapis, R. (1997). Ubiquitin lys63 is in-
volved in ubiquitination of a yeast plasma membrane protein. EMBO
J. 16, 5847–5854.
tination of the nerve growth factor receptor TrkA directs internaliza-
tion and signaling. Mol. Cell 20, 301–312.
Gorden, P., Carpentier, J.-L., Cohen, S., and Orci, L. (1978). Epider-
mal growth factor: morphological demonstration of binding
EGF Receptor Ubiquitination and Endocytosis
internalization and lydosomal association in human fibroblasts.
Proc. Natl. Acad. Sci. USA 75, 5025–5029.
Govers, R., ten Broeke, T., van Kerkhof, P., Schwartz, A.L., and
Strous, G.J. (1999). Identification of a novel ubiquitin conjugation
motif, required for ligand-induced internalization of the growth hor-
mone receptor. EMBO J. 18, 28–36.
Haas, W., Faherty, B.K., Gerber, S.A., Elias, J.E., Beausoleil, S.A.,
Bakalarski, C.E., Li, X., Ville ´n, J., and Gygi, S.P. (2006). Optimization
and use of peptide mass measurement accuracy in shotgun proteo-
mics. Mol. Cell. Proteomics, in press.
Haglund, K., Sigismund, S., Polo, S., Szymkiewicz, I., Di Fiore, P.P.,
and Dikic, I. (2003). Multiple monoubiquitination of RTKs is sufficient
for their endocytosis and degradation. Nat. Cell Biol. 5, 461–466.
Hicke, L., and Dunn, R. (2003). Regulation of membrane protein
transport by ubiquitin and ubiquitin-binding proteins. Annu. Rev.
Cell Dev. Biol. 19, 141–172.
Hofmann, R.M., and Pickart, C.M. (1999). Noncanonical MMS2-en-
coded ubiquitin-conjugating enzyme functions in assembly of novel
polyubiquitin chains for DNA repair. Cell 96, 645–653.
Hopkins, C.R., Miller, K., and Beardmore, J.M. (1985). Receptor-
mediated endocytosis of transferrin and epidermal growth factor
receptors: a comparison of constitutive and ligand-induced uptake.
J. Cell Sci. Suppl. 3, 173–186.
Huang, F., and Sorkin, A. (2005). Growth factor receptor binding
dermal growth factor receptor is essential and sufficient to support
receptor endocytosis. Mol. Biol. Cell 16, 1268–1281.
Huang,F., Khvorova, A.,Marshall,W.,andSorkin, A.(2004).Analysis
of clathrin-mediated endocytosis of epidermal growth factor recep-
tor by RNA interference. J. Biol. Chem. 279, 16657–16661.
Jiang, X., and Sorkin, A. (2003). Epidermal growth factor receptor in-
ternalization through clathrin-coated pits requires Cbl RING finger
and proline-rich domains but not receptor polyubiquitylation. Traffic
Jiang, X., Huang, F., Marusyk, A., and Sorkin, A. (2003). Grb2 regu-
lates internalization of EGF receptors through clathrin-coated pits.
Mol. Biol. Cell 14, 858–870.
Kirkpatrick, D.S., Denison, C., and Gygi, S.P. (2005). Weighing in on
ubiquitin: the expanding role of mass-spectrometry-based proteo-
mics. Nat. Cell Biol. 7, 750–757.
Levkowitz, G., Waterman, H., Zamir, E., Kam, Z., Oved, S., Langdon,
W.Y., Beguinot, L., Geiger, B., and Yarden, Y. (1998). c-Cbl/Sli-1 reg-
ulates endocytic sorting and ubiquitination of the epidermal growth
factor receptor. Genes Dev. 12, 3663–3674.
Levkowitz, G., Waterman, H., Ettenberg, S.A., Katz, M., Tsygankov,
A.Y., Alroy, I., Lavi, S., Iwai, K., Reiss, Y., Ciechanover, A., et al.
(1999). Ubiquitin ligase activity and tyrosine phosphorylation under-
lie suppression of growth factor signaling by c-Cbl/Sli-1. Mol. Cell 4,
Miaczynska, M., Pelkmans, L., and Zerial, M. (2004). Not just a sink:
endosomes in control of signal transduction. Curr. Opin. Cell Biol.
Miller, K., Beardmore, J., Kanety, H., Schlessinger, J., and Hopkins,
C.R. (1986). Localization of epidermal growth factor (EGF) receptor
within the endosome of EGF-stimulated epidermoid carcinoma
(A431) cells. J. Cell Biol. 102, 500–509.
Mosesson, Y., Shtiegman, K., Katz, M., Zwang, Y., Vereb, G.,
Szollosi, J., and Yarden, Y. (2003). Endocytosis of receptor tyrosine
kinases is driven by monoubiquitylation, not polyubiquitylation.
J. Biol. Chem. 278, 21323–21326.
Oksvold, M.P., Thien, C.B., Widerberg, J., Chantry, A., Huitfeldt,
H.S., and Langdon, W.Y. (2003). Serine mutations that abrogate
ligand-induced ubiquitination and internalization of the EGF recep-
tor do not affect c-Cbl association with the receptor. Oncogene
Peng, J., Schwartz, D., Elias, J.E., Thoreen, C.C., Cheng, D.,
Marsischky, G., Roelofs, J., Finley, D., and Gygi, S.P. (2003). A pro-
teomics approach to understanding protein ubiquitination. Nat.
Biotechnol. 21, 921–926.
Sigismund, S., Woelk, T., Puri, C., Maspero, E., Tacchetti, C., Transi-
dico, P., Di Fiore, P.P., and Polo, S. (2005). From the cover: clathrin-
independent endocytosis of ubiquitinated cargos. Proc. Natl. Acad.
Sci. USA 102, 2760–2765.
Sorkin, A., and Von Zastrow, M. (2002). Signal transduction and en-
docytosis: close encounters of many kinds. Nat. Rev. Mol. Cell Biol.
Sorkin, A., Krolenko, S., Kudrjavtceva, N., Lazebnik, J., Teslensko,
L., Soderquist, A.M., and Nikolsky, N. (1991). Recycling of epidermal
growth factor-receptor complexes in A431 cells: identification of
dual pathways. J. Cell Biol. 112, 55–63.
Spence,J.,Gali, R.R.,Dittmar, G.,Sherman, F.,Karin,M.,andFinley,
D.(2000). Cellcycle-regulatedmodificationof theribosome byavar-
iant multiubiquitin chain. Cell 102, 67–76.
Stamos, J., Sliwkowski, M.X., and Eigenbrot, C. (2002). Structure of
the epidermal growth factor receptor kinase domain alone and in
complex with a 4-anilinoquinazoline inhibitor. J. Biol. Chem. 277,
Thien, C.B., Walker, F., and Langdon, W.Y. (2001). RING finger mu-
tations that abolish c-Cbl-directed polyubiquitination and downreg-
ulation of the EGF receptor are insufficient for cell transformation.
Mol. Cell 7, 355–365.
Varadan, R., Assfalg, M., Haririnia, A., Raasi, S., Pickart, C., and
Fushman, D. (2004). Solution conformation of Lys63-linked di-ubi-
quitin chain provides clues to functional diversity of polyubiquitin
signaling. J. Biol. Chem. 279, 7055–7063.
Waterman, H., Katz, M., Rubin, C., Shtiegman, K., Lavi, S., Elson, A.,
Jovin, T., and Yarden, Y. (2002). A mutant EGF-receptor defective in
ubiquitylation and endocytosis unveils a role for Grb2 in negative
signaling. EMBO J. 21, 303–313.
Wiley, H.S. (1988). Anomalous binding of epidermal growth factor to
A431 cells is due to the effect of high receptor densities and a satu-
rable endocytic system. J. Cell Biol. 107, 801–810.