A R T I C L E
Small molecule inhibitors of HDM2 ubiquitin ligase activity
stabilize and activate p53 in cells
Yili Yang,1Robert L. Ludwig,2Jane P. Jensen,1Shervon A. Pierre,1Maxine V. Medaglia,1Ilia V. Davydov,3
Yassamin J. Safiran,3Pankaj Oberoi,3John H. Kenten,3Andrew C. Phillips,4Allan M. Weissman,1,*
and Karen H. Vousden2
1Laboratory of Protein Dynamics and Signaling, Center for Cancer Research, National Cancer Institute at Frederick, NIH, 1050 Boyles
Street, Frederick, Maryland 21702
2The Beatson Institute for Cancer Research, Switchback Road, Glasgow G61 1BD, United Kingdom
3Meso-Scale Discovery, Meso-Scale Diagnostics, LLC, 9238 Gaither Road, Gaithersburg, Maryland 20877
4Institute of Molecular Medicine and Genetics, Medical College of Georgia, Augusta, Georgia 30912
The p53 tumor suppressor protein is regulated by its interaction with HDM2, which serves as a ubiquitin ligase (E3) to
target p53 for degradation. We have identified a family of small molecules (HLI98) that inhibits HDM2’s E3 activity. These
compounds show some specificity for HDM2 in vitro, although at higher concentrations effects on unrelated RING and
HECT domain E3s are detectable, which could be due, at least in part, to effects on E2-ubiquitin thiol-ester levels. In cells,
the compounds allow the stabilization of p53 and HDM2 and activation of p53-dependent transcription and apoptosis,
although other p53-independent toxicity was also observed.
ability of p53 to induce apoptosis plays an important role in
tumor suppression (Schmitt et al., 2002; Symonds et al., 1994),
additional p53 activities are also important in regulating tumor
development (Fazeli et al., 1997; Liu et al., 2004; Ryan et al.,
2004). p53 functions both as a transcription factor (Vogelstein
et al., 2000) and through transcriptionally independent mecha-
nisms (Chipuk and Green, 2003). Many p53-inducible genes
have been identified that play a role in mediating the different
responses to p53. The choice of response to p53 depends,
at least in part, on which p53-responsive genes are activated
following induction of p53 (Vousden and Lu, 2002). There is
evidence that transformed cells are more sensitive to p53-
induced apoptosis than their normal counterparts, leading to
the suggestion that activation of p53 may cause tumor-specific
cell killing. As such, activation of the p53 response becomes
an attractive therapeutic goal (Lane and Lain, 2002).
The importance of p53 in tumor suppression is highlighted
by the observation that almost all human cancers show evi-
dence for loss of normal p53 function. In about half of all can-
cers, this occurs through mutation within the p53 gene, leading
to the expression of a mutant p53 protein that is defective for
growth and tumor inhibition (Selinova, 2001). In many of the
cancers that retain wild-type p53, there is evidence for defects
Regulation of protein stability through ubiquitin-dependent pro-
teolysis plays important roles in controlling the function of
many proteins, including critical regulators of cell proliferation
and apoptosis (Clarke, 2002; Yang and Yu, 2003). This proteo-
lytic system depends on the conjugation of ubiquitin onto pri-
mary amino groups on substrate proteins, with appropriately
linked polyubiquitin chains serving as targeting signals for pro-
teasomal degradation. Ubiquitylation occurs through a multien-
zyme cascade, where recognition is specified by the ubiquitin
ligase (E3). Numerous E3s have been identified, including a
large family of proteins that contain a structural motif known
as the RING finger (Fang and Weissman, 2004). RING finger
E3s mediate the direct transfer of ubiquitin from the ubiquitin
conjugating enzyme (E2) to substrate, and in some cases these
E3s have also been shown to regulate their own stability
through autoubiquitylation (Fang et al., 2000; Fang and Weiss-
The tumor suppressor protein p53 plays an important role in
preventing cancer development by inhibiting the proliferation
of cells undergoing tumorigenic stress, such as DNA damage
or oncogene activation (Vousden and Lu, 2002). Although the
S I G N I F I C A N C E
Many tumors that retain wild-type p53 show evidence of alterations that prevent efficient activation of p53 in response to stress,
linked to a failure to inactivate HDM2. In these tumors, inhibition of HDM2 and reactivation of p53 is an attractive therapeutic strategy.
While many chemotherapeutics stabilize p53, the HLI98 compounds, which directly target HDM2’s ubiquitin ligase activity, should
function to reactivate p53 without accompanying deleterious genotoxic damage that contributes to the toxicity of current therapeu-
tic drugs. Our data establish ubiquitin ligases such as HDM2 as viable targets for drug discovery, and identify a family of compounds
that may provide a structural basis for generation of drugs that could be used in the treatment of tumors retaining wild-type p53.
CANCER CELL : JUNE 2005 · VOL. 7 · COPYRIGHT © 2005 ELSEVIER INC.DOI 10.1016/j.ccr.2005.04.029547
A R T I C L E
in the mechanisms that allow activation of p53 (Vousden,
2002). The strong growth-suppressive activities of p53 mean
that p53 must be kept tightly regulated to allow normal cell
growth, and this is achieved to a large extent by regulating the
stability of the p53 protein. One of the principal E3s responsible
for targeting the degradation of p53, and keeping p53 levels
low during normal growth and development, is HDM2 (Mdm2
in mice) (Michael and Oren, 2003). Since HDM2 is a transcrip-
tional target of p53, a feedback loop exists in which p53 drives
expression of the protein that downregulates p53 activity. In
response to oncogenic stress, the degradation of p53 by
HDM2 is inhibited, leading to a rapid increase in p53 protein
levels in the cell. Several pathways leading to the inhibition of
HDM2 have been described, including the activation of DNA
damage-induced kinases that phosphorylate p53 and HDM2
(Ljungman, 2000; Meek and Knippschild, 2003), and the in-
teraction of HDM2 with proteins involved in ribosome assembly
or function; these include ARF, L11, L23, and L5 (Dai and Lu,
2004; Dai et al., 2004; Jin et al., 2004; Lohrum et al., 2003;
Sherr and Weber, 2000; Zhang et al., 2003). The binding of
these small proteins to HDM2 does not inhibit the interaction
of HDM2 with p53, but appears to block the ubiquitylation and
degradation of p53. These HDM2 binding proteins can also
prevent HDM2 autoubiquitylation in in vitro assays. Defects in
the pathways that control the stabilization and activation of p53
in response to stress can contribute to cancer development,
without the requirement for mutation within the p53 gene itself.
We have therefore sought to identify small molecule inhibitors
of HDM2’s E3 activity that may have efficacy in activating p53
in those tumors that retain wild-type p53.
rose beads was incubated with cell lysates, unbound material
removed, and ubiquitylation reactions carried out in the pres-
ence of the compounds (Figure 2A). Both HLI98C and the less
potent HLI98D exhibited dose-dependent inhibition of p53
ubiquitylation that becomes apparent for HLI98C at 20–50 ?M.
To evaluate the specificity for HDM2 relative to other E3s, we
assessed the inhibitory effects of HLI98 compounds against
both HDM2 and Nedd4, a structurally distinct HECT domain
E3 (Figure 2B). GLI97H dramatically blocked both HDM2 and
Nedd4 activity. This effect is due to irreversible inactivation of
both E1 and E2 (data not shown)—a feature not shared with
the HLI98 compounds. The HLI98 series also displayed inhibi-
tion of both Nedd4 and HDM2 at 100 ?M, although the inhibi-
tion of HDM2 was more prominent (Figure 2B and data not
shown). Using a different synthesis of HLI98C (Figure 2C),
dose-dependent inhibition of both HDM2 and Nedd4 was ob-
served, with HDM2 inhibition manifest at a lower concentration
(10 ?M compared to 50 ?M for Nedd4). On the other hand,
another RING finger E3, Siah1, showed little evidence for inhi-
bition by HLI98C under the same conditions, although some
inhibition of Siah1 was observed in other experiments where
less Siah1 was used (data not shown). The difference in inhibi-
tion of HDM2 in Figures 2B (100%) and 2C (89%), is reflective
of variation in the range of both E3 activity and inhibition in
these semiquantitative in vitro assays. While screening was
carried out using the full-length HDM2 molecule, the RING fin-
ger by itself is inhibited by HLI98C to a similar degree as the
full-length protein (Figure 2D); thus, more N-terminal regions,
including those involved in interactions with p53, are dispens-
able for the function of this inhibitor. Since the RING finger is
all that is required for the inhibition of HDM2 E3 activity, and
the same regions of E2s that interact with HECT E3s interact
with RING fingers, this may provide an explanation for the
overlap in inhibition between two disparate classes of E3s. Un-
fortunately, stable binding of HDM2 to E2 has yet to be repro-
ducibly observed; thus, formal testing of effects of the HLI98
series on this binding is currently problematic.
Not only are interaction sites on E2s shared between HECT
and RING finger E3s, but the regions on E2s that bind RING
fingers and HECT domains are also crucial for interactions of
E2s with E1. Thus, it might be the case that the HLI98s also
inhibit the transfer of ubiquitin from E1 to E2. Consistent with
this, HLI98C also shows some inhibition of E2-ubiquitin thiol-
ester linkages (Figure 2E). Overall, these results suggest that
there is some selectivity for the HLI98 compounds in the inhibi-
tion of HDM2 E3 activity, but that an effect on unrelated RING
and HECT domain E3s could also be measured. This could be
due, at least in part, to effects on loading of E2 with ubiquitin.
High-throughput screening for E3 ligase inhibitors
Using a previously described high-throughput assay developed
for the identification of ubiquitin ligase inhibitors (Davydov et
al., 2004), we screened libraries of small molecules for inhibi-
tion of HDM2 autoubiquitylation. A library of 10,000 com-
pounds was screened in duplicate, yielding forty compounds
that inhibited HDM2 autoubiquitylation by more than 50%.
These compounds were further tested in previously described
in vitro gel-based assays (Fang et al., 2000) that identified at
least four compounds that significantly inhibited E2-dependent
HDM2 autoubiquitylation. Three of the compounds, 10-(3-chloro-
4-dione (HLI98D), and 10-(4-methyl-phenyl)-7-nitro-10H-pyri-
mido[4,5-b]quinoline-2,4-dione (HLI98E), comprise a family of
closely related 7-nitro-5-deazaflavin compounds (Figure 1A). A
fourth compound, anthra[1,2-c][1,2,5]oxadiazole-6,11-diol
(GLI97H), showed a distinct structure. Using the plate-based
assays, IC50s for inhibition of HDM2 autoubiquitylation were
determined to be w20 ?M for each of the three 7-nitro-5-dea-
zaflavin compounds (Figure 1B and data not shown).
Activity of the HLI98 inhibitors in cells
The in vitro evidence suggests that the HLI98 family com-
pounds could function as E3 inhibitors, with at least some de-
gree of specificity for HDM2. To determine whether these ef-
fects are manifest in cells, we initially tested all 40 compounds
identified in the high-throughput screen for effects on cellular
p53 and HDM2. Primary human fibroblasts were treated with
50 ?M of each compound for 6 hr, after which levels of p53
and HDM2 were determined by Western blotting (an example
of this is shown in Figure 3A). Of the 40 compounds tested,
the HLI98 series exhibited the greatest ability to increase p53
and HDM2 protein levels, functioning with efficiency similar to
In vitro evaluation of inhibitors
To determine whether these compounds have the capacity to
inhibit ubiquitylation of p53 by HDM2, p53 was induced by
doxocycline in SAOS cells stably transfected with p53 cDNA
under the control of Tet-responsive promoter (Nakano et al.,
2000). GST-HDM2 that had been bound to glutathione-Sepha-
CANCER CELL : JUNE 2005
A R T I C L E
Figure 1. Small molecule inhibitors of HDM2
A: The structures of 10-(3-chloro-phenyl)-7-nitro-
10H-pyrimido[4,5-b]quinoline-2, 4-dione (HLI98C),
quinoline-2,4-dione (HLI98D) and 10-(4-methyl-
B: Dose-dependent inhibition of HDM2 auto-
ubiquitylation by HLI98C and HLI98D.
that seen with MG132, a peptide aldehyde proteasome inhibi-
tor. A direct comparison of each of the compounds with an-
other peptide aldehyde proteasome inhibitor (ALLN) and adria-
mycin, a DNA-damaging chemotherapeutic agent known to
induce stabilization of p53, showed that each efficiently in-
creased cellular p53 (Figure 3B). As expected, the HLI98 com-
pounds and ALLN also inhibited the degradation of HDM2,
leading to an increase in HDM2 levels. However, treatment of
cells with adriamycin led to the stabilization of p53 without
increasing the expression of HDM2, reflecting the increase in
HDM2 autodegradation in response to DNA-damage induced
kinase activation (Stommel and Wahl, 2004). Characteristic of
proteasome inhibitors, treatment with ALLN resulted in accu-
mulation of p53 primarily in its nonubiquitylated form, but with
clear evidence of a characteristic ladder of higher molecular
weight forms of the protein representative of ubiquitylation
(Figure 3B). In contrast, and consistent with the predicted
mechanism of action in inhibiting ubiquitylation rather than pro-
teasome function, no evidence for ubiquitylated p53 was de-
tected with the HLI98 compounds. Adriamycin, which reduces
expression of HDM2, also showed p53 accumulation without
ubiquitylation (Figure 3B).
An increase in both p53 and HDM2 is the predicted effect
of an HDM2 inhibitor. However, a possibility that must be
excluded is that the increase in HDM2 is a result of p53-
dependent transcriptional activation of HDM2. To rule out this
possibility, p53/mdm2 null mouse embryo fibroblasts were
transiently transfected with plasmid encoding HDM2 driven by
a p53-independent CMV promoter. As expected, adriamycin
had no effect on the level of transfected HDM2 (Figure 3C).
However, both proteasome inhibition (with ALLN) and treat-
ment with HLI98 compounds significantly increased HDM2
levels (Figure 3C). These findings indicate that the compounds
can lead to an accumulation of HDM2 by inhibition of HDM2
autoubiquitylation, as well as through the activation of p53.
Specificity of the HLI98 compounds in cells
As in vitro assays showed only relative selectivity of the HLI98s
toward HDM2, and potential for inhibition at the level of E2-
ubiquitin thiol-ester bond formation, several approaches were
undertaken to assess their specificity in cells. In vitro studies
showed that the compounds more efficiently inhibited HDM2
E3 activity than that of the unrelated RING finger E3 Siah1. The
stabilizing effect of HLI98C and HLI98D on HDM2 and Siah1 in
cells was therefore examined in transfected p53−/−mdm2−/−cells
(Jones et al., 1995). Under these conditions, no stabilization of
Siah1 was observed with any of the HLI98 compounds,
whereas a RING finger mutant form of Siah1 was stable even
without proteasome inhibitor (Figure 4A and data not shown).
To further examine the effect of the compounds on E3 activ-
CANCER CELL : JUNE 2005549
A R T I C L E
Figure 2. Inhibition of HDM2-mediated ubiqui-
tylation by HLI98 compounds in vitro
A: Inhibition of p53 ubiquitylation by HLI98C and
HLI98D. Transfected p53 expressed in SAOS cells
after induction with doxocycline was bound to
GST-HDM2 and excess material removed by
washing. Ubiquitylation reactions were then car-
ried out in the presence of indicated concen-
trations of HLI98C and HLI98D. Ubiquitylated p53
was detected by Western blotting using the
anti-p53 antibody DO-1.
B: Inhibition of HDM2 and Nedd4 autoubiqui-
tylation in a gel-based assay. Autoubiquitylation
of HDM2 and Nedd4 was detected using32P-
labeled ubiquitin. 20 ?M HLI98C and 50 ?M
GLI97H were used.
C: Evaluation of relative inhibition of HDM2,
Nedd4, and Siah 1 using a gel-based assay and
32P-labeled ubiquitin as in B.
D: Inhibition of GST-HDM2 and GST-HDM2 RING
finger autoubiquitylation. Assays were carried
out as in B and C.
E: The effect of HLI98C on the formation of
E1-dependent E2 thiol-ester conjugates with
ity in cells, expression of the RING finger E3 Cbl-b was exam-
ined. Autoubiquitylation and degradation of Cbl-b is seen fol-
lowing stimulation of cells with EGF, a response that requires
the E3 activity of Cbl-b and is blocked by MG132 (Ettenberg
et al., 2001) (Figure 4B). However, treatment of cells with
HLI98D failed to prevent EGF-induced Cbl-b loss, suggesting
that the HLI98 inhibitors do not block Cbl-b ubiquitin ligase
activity. In similar experiments, the HLI98 compounds also
failed to block TNFα-induced ubiquitin-dependent degradation
of IκBα (Chen et al., 1995b) under conditions where an E1 in-
hibitor under development did inhibit IκBα degradation (Y.Y.
and A.M.W., unpublished observations). Taken together, these
studies suggest that the HLI98 compounds show specificity in
vivo, inhibiting the autoubiquitylation of HDM2 more efficiently
than the activity of other RING finger E3s.
To determine whether the compounds also inhibit the degra-
dation of other proteins regulated by ubiquitylation, we eval-
uated the effect of the compounds on the stability of en-
dogenous p21WAF1/CIP1(Figure 4C), a protein targeted to the
proteasome by both ubiquitin-dependent and ubiquitin-inde-
pendent mechanisms (Bendjennat et al., 2003; Bloom et al.,
2003; Coulombe et al., 2004; Sheaff et al., 2000). As expected,
proteasome inhibition by treatment with ALLN resulted in the
stabilization of HDM2, p53, and p21WAF1/CIP1. In contrast, treat-
ment of cells with adriamycin resulted in stabilization of p53,
but had no effect on levels of either HDM2 or p21WAF1/CIP1. It
should be noted that the failure of adriamycin to elevate levels
of HDM2 or p21WAF1/CIP1protein in these wild-type p53 ex-
pressing cells is a reflection of the early time point (6 hr) follow-
ing treatment at which these cells were evaluated. Since both
HDM2 and p21WAF1/CIP1are transcriptional targets of p53, sta-
bilization of p53 by adriamycin leads to the transcriptional acti-
vation of both HDM2 and p21WAF1/CIP1(Figure 6B), which is
seen as an increase in protein levels at later time points. The
absence of stabilization of p21WAF1/CIP1further supports a de-
gree of selectivity of the compounds for HDM2 E3 activity. In
addition to HDM2, other E3s such as Pirh2 (Leng et al., 2003)
and COP1 (Dornan et al., 2004) can target p53 for degradation.
Since the HLI98 compounds can potentially inhibit E3s other
than Mdm2 we examined the contribution of Mdm2 expression
to the ability of these compounds to stabilize p53 by compar-
ing p53 null or p53/mdm2 null mouse embryo fibroblasts that
CANCER CELL : JUNE 2005
A R T I C L E
zation of p53 in response to DNA damage. However, HLI98C
treatment showed little stabilization of p53 in the absence of
Mdm2, supporting a selectivity for inhibition of Mdm2 activity.
Finally, to determine whether the compounds could affect
the degradation of p53 mediated by E3s other than HDM2, an
isogenic cell system was utilized. In this system, p53 degrada-
tion is mediated either by HDM2 (RKO cells) or by stably ex-
pressed HPV-E6 from an oncogenic strain of HPV that func-
tions with the cellular HECT domain E3, E6-AP (RKO/E6 cells)
(Scheffner et al., 1993) (Figure 4E). In this system, proteasome
inhibitor resulted in the stabilization of p53 regardless of which
E3 was targeting degradation. When HLI98C was evaluated,
selective stabilization of HDM2-mediated p53 loss was ob-
served at 10 ?M, although when the concentration was in-
creased to 20 ?M, evidence of inhibition of E6-AP-mediated
degradation was also observed. Collectively, these results are
consistent with the in vitro data in demonstrating relative selec-
tivity for HDM2, but as demonstrated in Figure 4E, there is
clearly the potential to inhibit a HECT E3, in this case E6-AP.
Recently, small molecules that can inhibit the HDM2-medi-
ated degradation of p53 by blocking the HDM2-p53 interaction
have been described (Issaeva et al., 2004; Vassilev et al.,
2004). Although the HLI98 compounds were selected for their
ability to inhibit HDM2 directly, and also inhibit the isolated
HDM2 RING finger (Figure 2D), we wished to confirm that these
compounds do not interfere with the binding of HDM2 to p53.
Both transfected p53 and transfected HDM2 were stabilized by
treatment with MG132 or HLI98C (Figure 5A, lower two panels).
HDM2 was found to be efficiently coimmunoprecipitated with
Flag-p53 in cells cotransfected with p53 and HDM2 and
treated with either MG132 or HLI98C (Figure 5A). These results
are consistent with a role for the HLI98 compounds that does
not involve inhibition of the p53-HDM2 interaction.
Many drugs that induce p53 stabilization also activate stress
responsive kinases that lead to N-terminal phosphorylation of
p53 (Xu, 2003). These phosphorylation events have been pro-
posed to play a role in allowing both the stabilization and acti-
vation of p53. The induction of p53 in response to both DNA
damage and oxidative stress is accompanied by phosphoryla-
tion of p53 on serine 15 (Chen et al., 2003; Shieh et al., 1997;
Siliciano et al., 1997). As shown previously, this modification
can be detected, using a serine 15 phosphospecific antibody,
in response to camptothecin treatment (Figure 5B), but not in
p53 stabilized by proteasome inhibition (MG132) or by treat-
ment with low levels of actinomycin D (Ashcroft et al., 2000).
Importantly, treatment of cells with the HLI98 compounds also
failed to stimulate serine 15 phosphorylation. Furthermore,
while the presence of the nitro group in each of the HLI98 com-
pounds raises some possibility for the activation of oxidative
stress, we have recently synthesized and tested a number of
HLI98 analogs lacking this nitro group that show no reduction
in the ability to stabilize p53 in cells (D. Robins, G. Henderson,
J. Wilson, R.L.L., and K.H.V., unpublished observations). Taken
together, these results indicate that the means by which these
compounds stabilize p53 and increase HDM2 levels is not
through genotoxic or oxidative stress.
Figure 3. HLI98 family compounds accumulate p53 and HDM2 in cells
A: Screening for compounds that increase p53 and HDM2 in cells. Hits from
high-throughput screening were added to cultured human MRC5 fibro-
blasts for 6 hr at a final concentration of 50 ?M. Cells treated with 50 ?M
MG132 were used as a control. HDM2 and p53 in the cells were examined
by Western blotting with anti-HDM2 and anti-p53 antibodies.
B: Human RPE cells were treated with 20 ?M HLI98C, 50 ?M HLI98D, 50 ?M
HLI98E, 1 ?g/ml adriamycin, and 50 ?M ALLN for 6 hr. HDM2 and p53 in cells
were assessed by Western blot using anti-HDM2 and anti-p53 antibodies.
C: Fibroblasts from p53−/−mdm2−/−mice were transfected with HDM2
cDNA under the control of a CMV promoter. After 24 hr, the cells were
treated with the indicated concentrations of HLI98E, 50 ?M MG132, or 0.2
?g/ml adriamycin for 8 hr, and the level of HDM2 was determined by West-
ern blot. Expression of β-actin was monitored as a control for loading.
had been transfected with p53. In the cells expressing Mdm2,
both HLI98C and adriamycin led to p53 stabilization (Figure
4D, left panels). By comparison, in cells lacking Mdm2, the
basal levels of p53 were significantly higher, as previously de-
scribed (Kubbutat et al., 1997). Adriamycin treatment could fur-
ther stabilize p53, despite the absence of Mdm2, consistent
with a contribution of inhibition of Pirh2 or COP1 in the stabili-
Induction of p53 transcriptional activity in response
to treatment with the inhibitors
One of the principal activities of p53 is as a transcription factor,
inducing expression of a number of genes that can contribute
CANCER CELL : JUNE 2005 551
A R T I C L E
Figure 4. HLI98 family compounds show specific-
ity toward HDM2-mediated ubiquitylation
A: Fibroblasts from p53−/−mdm2−/−mice were
transfected with cDNAs encoding HDM2 or
Siah1. Twenty-four hours after transfection, cells
were treated with 50 ?M HLI98D, HLI98E, or
MG132 for 8 hr. The levels of HDM2 and Siah1
were determined by Western blot using anti-
HDM2 and anti-Flag antibodies.
B: MDA-MB-468 cells stably expressing Cbl-b
were first treated with 50 ?M MG132 or 50 ?M
HLI98D for 30 min. 100 ?g/ml EGF was then
added to the cultures as indicated for 2 hr. The
levels of Cbl-b and β-actin in cells were as-
sessed by Western blotting with anti-Cbl-b and
C: p21WAF1/CIP1degradation was not blocked
by HLI98 family compounds. RPE cells were
treated with 0.2 ?g/ml adriamycin, 50 ?M ALLN,
and 20 ?M HLI98C or 50 ?M HLI98D or HLI98E for
6 hr. Cellular HDM2, p53, and p21WAF1/CIP1were
analyzed by immunoblotting with specific anti-
D: Fibroblasts from p53−/−or p53−/−mdm2−/−
mice were transfected with p53 and treated
with 20 ?M HLI98C or 0.2 ?g/ml adriamycin.
Levels of p53 and Cdk4 were determined after
16 hr by Western blotting.
E: Carcinoma cell line RKO and RKO cells ex-
pressing E6 from oncogenic HPV were treated
with 10–20 ?M of MG132 or 10–20 ?M of HLI98C
for 6 hr, as indicated. p53 and Cdk4 levels were
determined by Western blotting.
to the cell cycle arrest and apoptotic response. To test the
ability of the HLI98 compounds to activate p53-dependent
transcription, we made use of a cell line expressing endoge-
nous wild-type p53 and a luciferase reporter gene under the
transcriptional control of a synthetic p53-responsive promoter
(Figure 6A). Treatment of these cells with the HLI98 compounds
resulted in a reproducible increase in luciferase induction, al-
beit substantially less than that seen with adriamycin (Figure
6A). This weaker activation of p53-dependent transcription in
response to the HLI98 compounds correlated to some extent
with a lower induction of p53 protein levels compared to that
seen with adriamycin (data not shown). We next examined the
effect of the HLI98 compounds on the transcription of endoge-
nous p53 target genes in cells expressing wild-type p53 (Figure
6B). The transcription of both the cell cycle arrest p53 target
gene p21WAF1/CIP1and the apoptotic p53 target gene PUMA
was efficiently induced in response to treatment of cells with
HLI98C, to an extent similar to that seen following treatment
with camptothecin (Figure 6B). This activation of p21WAF1/CIP1
expression by two independent syntheses of HLI98C is also
seen at the protein level (Figure 6C). These results indicate that
in cells, the stabilization of p53 by the HLI98 compounds re-
sults in an activation of p53-dependent transcription.
Induction of p53-dependent apoptosis
The ability of p53 to induce apoptosis is likely to play an impor-
tant role in tumor suppression. A well-established system for
measuring p53-dependent cell death is in mouse embryo fibro-
blasts (MEFs) transformed with the adenovirus E1A protein and
activated Ha-ras (Lowe et al., 1993). Analysis of caspase acti-
vation in cells following treatment with the HLI98 compounds
showed an efficient induction of DEVDase activity in the 5–20
CANCER CELL : JUNE 2005
A R T I C L E
but little apoptosis in the parental RPE cells. However, as seen
previously with MEFs, expression of E1A strongly sensitizes
these cells to p53-dependent apoptosis in response to adria-
mycin (data not shown). Similarly, treatment of these cells with
the HLI98 compounds only induced significant apoptosis in the
cells expressing E1A (Figure 8A). This death occurred in a
dose-dependent manner, with 50% of the cells undergoing
apoptosis at 10 ?M HLI98C (Figure 8B). These results suggest
that the compounds may lead to p53-dependent death selec-
tively in cancer cells. Interestingly, these studies suggest that
induction of p53-mediated death in cells is achieved at con-
centrations of the HLI98 compounds lower that those neces-
sary for clear inhibition of p53 ubiquitylation in vitro (Figure 2A).
This may reflect differences in levels of HDM2 or threshold ac-
tivity of the compounds required to achieve a detectable effect.
Evidence for the p53 dependence of the growth-inhibitory ef-
fect of the HLI98 compounds was also seen in human tumor
cell lines (Figure 8C). Using a colony assay and low concentra-
tions of the compounds, p53-containing tumor cells such as
RKO and U2OS showed greater sensitivity to growth inhibition
than p53 null cells such as H1299 or RKO cells expressing
E6. However, at higher concentrations, the HLI98 compounds
inhibited the growth of all cells, regardless of their p53 status,
most likely reflecting additional, p53-independent activities of
these compounds. To begin to investigate this, we examined
the effect of the compounds on cell cycle progression in cells
that would not be expected to undergo apoptosis in response
to p53. FACS analysis of HCT116 (p53+/+and p53−/−) and
H1299 cells revealed a complex and variable accumulation of
cells in S phase and G2 that was not correlated to p53 expres-
sion (data not shown). At higher concentrations of the com-
pounds, p53-independent apoptosis also became evident. In
RPE cells, a concentration-dependent difference in response
to the compounds could be observed (Figure 8D). At lower
concentrations (10 ?M HLI98C), where HDM2 is preferentially
inhibited, a classic p53 response of G1 arrest is observed. By
contrast, at higher concentrations (20 ?M HLI98C), the re-
sponse of a strong G2 arrest is similar to that seen following
more general inhibition of ubiquitin-dependent degradation by
treatment with MG132. A similar G2 arrest was also seen in
these cells in response to treatment with the E1 inhibitor (Y.Y.,
R.L.L., K.H.V., and A.M.W., unpublished observations). These
results clearly show that the compounds can have activities
detrimental to cell growth and survival that are independent of
p53, and probably reflect off-target activities such as the inhibi-
tion of other E3s or perhaps a subset of E2s.
Figure 5. Effects of the HLI98 compounds on p53/HDM2 interaction and
A: HLI98C does not inhibit p53/HDM2 interaction. U2OS cells were trans-
fected with HDM2 and Flag-p53, then treated with HLI98C or MG132 for 6
hr as indicated before harvesting. p53/HDM2 complexes were immuno-
precipitated with an anti-flag antibody, and the amount of HDM2 in the
complex determined by Western blotting (top). Total levels of HDM2 and
p53 were assessed by Western blotting the lysate before immunoprecipita-
tion (middle and bottom).
B: HLI98 family compounds do not induce phosphorylation of p53. RPE cells
were treated with 20 ?M MG132, 2.5 ?M actinomycin-D, 50 ?M HLI98D,
and 0.5 ?M camptothecin for 6 or 24 hr as indicated. The overall p53 levels
were determined by Western blotting with the DO1 antibody, and the ex-
tent of serine 15 phosphorylation assessed using a phosphoserine 15-spe-
cific anti-p53 antibody.
?M range (Figure 7A). This activity was dependent on p53,
since a similar activation was not seen in p53-deficient MEFs
expressing Ha-ras and E1A. Similarly, treatment with the HLI98
compounds resulted in p53-dependent PARP cleavage selec-
tively in the p53-expressing cells (Figure 7B). The compounds
also induced cell death in a p53-dependent manner as mea-
sured by trypan blue exclusion (Figure 7C).
Previous studies have shown that certain oncogenes sensi-
tize cells to p53-induced apoptosis (Lowe et al., 1993). To de-
termine whether this differential is also seen in human epithelial
cells (the cell type from which the majority of human cancers
arise), we established Tert-immortalized human retinal pigment
epithelial cells expressing E1A. Activation of p53 with the gen-
otoxic chemotherapeutic adriamycin induced a cell cycle arrest
Intensive study of the p53 protein has revealed that loss of p53
function is virtually a prerequisite for cancer development and
that induction of p53 is likely to lead to tumor cell specific kill-
ing, two key observations that have prompted several attempts
to identify small molecules that can activate p53 (Haupt and
Haupt, 2004). In tumors that retain wild-type p53, these efforts
have concentrated on the development of drugs that protect
p53 from the inhibitory effects of HDM2. Although other ubiqui-
tin ligases—such as COP1 and Pirh2—can also target p53
(Dornan et al., 2004; Leng et al., 2003), downregulation of
HDM2 appears to be sufficient to activate a p53 response in
cells and in vivo (Mendrysa et al., 2003). Most attempts to in-
CANCER CELL : JUNE 2005 553
A R T I C L E
hibit HDM2 have concentrated on preventing the interaction
with p53, and significant success has been achieved with pep-
tides (Chene, 2003) and more recently with the identification of
small molecules such as the Nutlins (Vassilev et al., 2004) and
RITA (Issaeva et al., 2004). We have taken an alternative ap-
proach, which is to identify small molecule inhibitors of HDM2’s
E3 activity. A potential concern with this approach is that the
binding of HDM2 to p53 can inhibit p53’s transcriptional activ-
ity, and so elevation of HDM2 and p53 protein levels without
perturbing their interaction may not allow full activation of p53.
One approach to circumvent this issue has been to identify
small molecules with selective ability to inhibit p53 ubiquityla-
tion, without blocking HDM2 autoubiquitylation in vitro (Lai et
al., 2002). Our inhibitors, which were selected for their ability
to inhibit HDM2 autoubiquitylation, clearly lead to the activa-
tion of p53 function despite allowing the stabilization of both
HDM2 and p53. There are various potential explanations for
this observation, including the possibilities that p53 levels in-
crease more rapidly than HDM2 levels and so allow for the
accumulation of free p53, or that the p53/HMD2 complex re-
tains some activity. A recent study has shown that the protea-
some inhibitor bortezomib also leads to the stabilization of p53
without impeding the interaction with HDM2, and, as with the
HLI98 compounds, this p53 was transcriptionally active (Wil-
liams and McConkey, 2003). In vivo validation to this approach
is also provided by the identification of several small proteins—
such as p14ARF and the ribosomal proteins L11, L5, and L23.
These can bind HDM2 and inhibit p53 degradation without di-
rectly preventing the HDM2/p53 interaction (Dai and Lu, 2004;
Dai et al., 2004; Lohrum et al., 2003; Sherr and Weber, 2000;
Zhang et al., 2003).
In addition to the regulation of p53 stability, the ubiquitylation
of p53 by HDM2 has also been shown play a role in allowing
nuclear export of p53 (Gu et al., 2001; Lohrum et al., 2001).
Depending on levels of HDM2, p53 can become monoubiquity-
lated and exported from the nucleus, or polyubiquitylated and
degraded (Li et al., 2003). As expected, the p53 that is stabi-
lized by the HLI98 compounds also remains localized to the
nucleus (data not shown). While degradation of p53 clearly in-
hibits p53 activity, the consequences of nuclear export on p53
function are less predictable. While preventing transcriptional
activity, cytoplasmic localization could favor the mitochondrial
function of p53 in promoting apoptosis (Mihara et al., 2003).
Presumably, inhibition of ubiquitylation by the HLI98 com-
pounds would also reduce nuclear export of p53, although our
results suggest that this is not accompanied by a defect in its
ments. The differences between the DMSO-treated cells, and those
treated with adriamycin and HLI98C, D, and E are all significant, with p
values of <0.01. There are no significant differences between HLI98C, D,
B: Induction of p53-responsive genes by HLI98 compounds. RPE cells were
treated with 20 ?M HLI98C or 2 ?M camptothecin for the indicated times.
Expression of p21WAF1/CIP1and PUMA was determined by RT-PCR; GAPDH
expression was monitored as a control for RNA integrity.
C: Induction of p53-reponsive gene expression by two different syntheses
of HLI98C. RPE cells were treated as indicated for 24 hr, then levels of p53
and p21WAF1/CIP1protein analyzed by Western blotting. Cdk4 levels were
monitored as a control.
Figure 6. Activation of p53-dependent transcription by the HLI98 com-
A: HLI98 family compounds activate p53-dependent transactivation. U2OS
cells harboring a luciferase gene under the control of p53-response ele-
ments were treated with 0.3 ?g/ml adriamycin or HLI98 family compounds
(20 ?M HLI98C, 50 ?M HLI98D, and 50 ?M HLI98E) for 20 hr and luciferase
activity assessed. Data represents averages of three independent experi-
CANCER CELL : JUNE 2005
A R T I C L E
for proteasome inhibitors (Adams, 2004), and bortezomib has
now been approved for treatment of multiple myeloma, a par-
ticularly refractory cancer. This is despite their lack of specific-
ity for any specific protein or family of proteins. The explosion
in knowledge of E3s and their many obvious links to disease
has spurred considerable interest in targeting E3 activity as a
much more specific means of developing therapeutics by alter-
ing the stability of critical specific substrates, such as p53. This
enthusiasm is, however, tempered by concerns as to whether
specific inhibitors can be developed given the similarity among
RING fingers, the level of similarity in E2 interaction sites be-
tween and among RING finger and HECT domain E3s, and the
potential to inhibit E2 loading with ubiquitin by E1 by phar-
maceuticals that might affect E2-E3 interactions (Weissman,
2001). Our data reinforce the potential significance of these is-
sues in that, although the HLI98 compounds are relatively spe-
cific for HDM2/Mdm2, we provide evidence for effects on
HECT E3s both in vitro and in cells, which may well reflect the
inhibition of E2 loading seen in vitro. Consistent with such off-
target effects, there is clear p53-independent toxicity in several
different cell lines.
Off-target activity is of course both a bane and a potential
benefit in therapeutic development, and when known before-
hand to be present, is a factor that must be carefully weighted
in making decisions as to whether to move a lead compound
forward. The compounds described here themselves have little
potential as therapeutics due to limited solubility and the need
for micromolar levels. This can potentially be addressed by
identifying more “druggable” analogs. Should this be success-
ful, the issue of specificity will need to be addressed in ani-
mal models and weighted against potential benefits in patients
with wild-type p53 tumors. Regardless, our results should be
viewed as heartening, as they establish proof of principle, in
that relatively selective inhibitors of E3 ligases, such as HDM2,
can in fact be identified, and that they have the predicted cellu-
Figure 7. HLI98 family compounds activate caspases in a p53-dependent
Wild-type or p53-deficient MEFs transformed with E1A and Ha-ras were
treated with HLI98 family compounds for 20 hr.
A: Caspase activity measured by FMC-DEVD.
B: Cleavage of PARP following treatment of cells with 20 ?M ALLN, 5 ?g/ml
etoposide, 20 ?M HLI98C, and 50 ?M HLI98D, revealed by immunoblotting.
C: Cell death measured by trypan blue exclusion.
Proteasome inhibitors ALLN and MG132, creatine phosphokinase, and rab-
bit E1, were from Calbiochem (La Jolla, CA). Adriamycin, actinomycin D,
creatine phosphate, and camptothecin were obtained from Sigma (St.
Louis, MO). Recombinant EGF was from R&D Systems (Minneapolis, MN).
Glutathione Sepharose was from Amersham Biosciences (Piscataway, NJ).
Anti-HDM2 antibodies Ab-1 and Ab-2 were from Oncogene (Boston, MA).
Anti-p53 monoclonal antibody DO-1, anti-p21WAF1/CIP1, anti-PARP, and
anti-Cbl-b polyclonal antibodies were from Santa Cruz Biotechnology
(Santa Cruz, CA). Anti-phospho-p53 (serine15) antibody was from Cell Sig-
naling (Beverly, MA). Anti-β-actin and anti-Flag monoclonal antibodies were
from Sigma (St. Louis, MO).
Although we have concentrated on the analysis of the HLI98
inhibitors on HDM2’s ability to degrade p53, it is also possible
that these compounds can prevent HDM2 mediated neddyla-
tion of p53 (Xirodimas et al., 2004). Further analysis of the
mechanism of function of these compounds, particularly an
identification of whether and where they interact with HDM2,
will be essential in understanding how p53 stability and func-
tion can be modulated. The HLI98 compounds described here
are relatively insoluble, which has hindered further analysis of
their activity in vivo.
While attempts to identify novel drugs for cancer treatment
have concentrated mainly on kinase inhibitors, there is much
interest in developing the therapeutic potential of modulators
of protein stability. Success in clinical trials has been reported
Bacterial expression plasmids for HDM2, Nedd4, and Siah1 have been de-
scribed previously (Fang et al., 2000; Lorick et al., 1999). CMV-driven mam-
malian expression constructs for HDM2 and Siah1 have also been de-
scribed (Chen et al., 1995a; Hu and Fearon, 1999).
RPE, a human retinal pigment epithelial cell line that stably expresses hu-
man telomerase reverse transcriptase (hTert) was from Clontech (Palo Alto,
CA) and cultured with DME/F-12 (HyClone, Logan, UT) supplemented with
10% FCS, 100 units/ml of penicillin, 100 ?g/ml of streptomycin, and 0.3%
sodium bicarbonate. E1A-expressing RPE cells (Balint et al., 2002) have
been described. Mouse embryo fibroblasts (MEFs) from p53- and mdm2-
CANCER CELL : JUNE 2005 555
A R T I C L E
Figure 8. Transformed cells are susceptible to
HLI98 family compound-induced apoptosis
A: HLI98 family compounds induce apoptosis in
E1A-transformed cells but not untransformed
RPE cells (1 ?g/ml adriamycin, 20 ?M HLI98C,
and 50 ?M HLI98D).
B: HLI98C induces dose-dependent apoptosis of
E1A-transformed RPE cells, but not untrans-
formed RPE cells.
C: Colony assay with human tumor cell lines.
Wild-type p53 expressing cells (RKO and U2OS)
and cells null for p53 (H1299) or expressing E6
(RKO+E6) were grown in the presence of the in-
dicated concentrations of HLI98C for 5–7 days,
before fixing and counting colonies.
D: FACS analysis of RPE cells following treatment
with the indicated concentrations of HLI98C
deficient mice and MEFs transformed with both E1A and Ras were main-
tained in DMEM supplemented with 10% FCS. U2OS-pG13 are a clone of
U2OS cells stably expressing the luciferase gene under the control of the
synthetic p53-responsive promoter (reporter pG13) (Kern et al., 1992). p53-
inducible Saos-2 cells have been described (Nakano et al., 2000). These
human cells were cultured with DMEM supplemented with 10% FCS. MDA-
MB-468 breast cancer cells stably expressing Cbl-b (Ettenberg et al., 2001)
were maintained in RPMI 1640 supplemented with 10% FCS.
synthesis was carried out using 5 ?g total RNA using SuperScript First-
Strand Synthesis Systems for RT-PCR (Invitrogen) or GC-Rich PCR System
(Roche). 10 ng of the reaction was used for PCR using AmpliTaq (Applied
Cells were lysed with RIPA buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 1%
NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 ?g/ml aprotinin, 100 ?g/
ml PMSF, and 5 ?g/ml leupeptin) and centrifuged at 15,000 × g for 20 min.
Superanatants were mixed with 4× SDS-PAGE sample buffer (200 mM Tris
[pH 6.8], 40% glycerol, 8% SDS, 400 mM DTT, and 0.2% bromophenol
blue) and heated at 100°C for 3 min prior to resolution by SDS-PAGE and
Total RNA was isolated from RPE cells grown to 80% confluence and
treated for indicated times using TRIzol reagent (Invitrogen). First strand
CANCER CELL : JUNE 2005
A R T I C L E
transfer to nitrocellulose membranes. Membranes were preincubated with
5% nonfat dry milk in TBST (50 mM Tris [pH 8.0], 150 mM NaCl, 0.05%
Tween-20) before incubation with specific antibody for 2 hr. Specific mole-
cules were visualized with horseradish peroxidase-labeled anti-mouse or
anti-rabbit secondary antibodies and enhanced chemiluminescence (Amer-
sham Biosciences, UK).
We thank Dr. Steve Jones (University of Massachusetts) for the p53/mdm2
null MEFs, Dr. Scott Lowe (Cold Spring Harbor Laboratories) for the E1A/
Ras transformed wild-type and p53 null MEFs, and Dr. Stan Lipkowitz (NCI,
Bethesda) for MDA-MB-468 breast cancer cells expressing Cbl-b. We are
also grateful to Carlton Briggs for help with generation of the U2OS-pG13
reporter cell line, and Dr. John Beutler and Dr. Kevin Lorick (NCI, Frederick)
for helpful discussions.
In vitro autoubiquitylation
Glutathione sepharose-bound GST fusion proteins (GST-HDM2, GST-
Nedd4, and GST-Siah1) were resuspended in 27 ?l of ubiquitylation reaction
buffer (50 mM Tris [pH 7.4], 0.2 mM ATP, 0.5 mM MgCl2, 0.1 mM DTT, 16
?M creatine phosphokinase, and 1 mM creatine phosphate). Ubiquitylation
reactions were carried out by adding 100 ng of rabbit E1, 100 ng of bacteri-
ally expressed UbcH5B, and 2 × 104cpm of32P-ubiquitin (prepared as
described previously [Fang et al., 2000; Scheffner et al., 1994]) to each
mixture and incubating at 30°C for 90 min with agitation. Reactions were
terminated by addition of 10 ?l of 4× reducing SDS-PAGE sample buffer
and heating at 100°C for 3 min. Samples were resolved on 8% SDS-PAGE,
and ubiquitylated HDM2 visualized by Storm PhosphoImager (Molecular
Dynamics, Sunnyvale, CA).
Received: October 21, 2004
Revised: March 2, 2005
Accepted: April 21, 2005
Published: June 13, 2005
Adams, J. (2004). The development of proteasome inhibitors as anticancer
drugs. Cancer Cell 5, 417–421.
HDM2-mediated p53 ubiquitylation
SAOS2 cells stably transfected with p53 cDNA under the control of Tet-on
promoter were treated with 2 ?g/ml of doxocycline for 20 hr to induce the
expression of p53. Cells were then harvested and lysed with lysis buffer (50
mM Tris [pH 7.4], 150 mM NaCl, 0.5% NP-40). After a 15,000 × g centrifuga-
tion to remove cellular debris, the supernatant was added to GST-HDM2
that had been bound to glutathione Sepharose beads and incubated at 4°C
for 2 hr. Following washing 4 times with 50 mM Tris to remove unbound
materials, the beads containing HDM2/p53 complexes were resuspended
in ubiquitylation reaction buffer and ubiquitylation reaction was carried out
as described above. The reaction mixture was separated on 8% SDS-PAGE
and transferred to nitrocellulose membrane for immunoblotting with anti-
p53 antibody DO-1. The blot was developed with enhanced chemilumines-
Ashcroft, M., Taya, Y., and Vousden, K.H. (2000). Stress signals utilize multi-
ple pathways to stabilize p53. Mol. Cell. Biol. 20, 3224–3233.
Balint, E., Phillips, A.C., Kozlov, S., Stewart, C.L., and Vousden, K.H. (2002).
Induction of p57KIP2expression by p73β. Proc. Natl. Acad. Sci. USA 99,
Bendjennat, M., Boulaire, J., Jascur, T., Brickner, H., Barbier, V., Sarasin, A.,
Fotedar, A., and Fotedar, R. (2003). UV irradiation triggers ubiquitin-depen-
dent degradation of p21(WAF1) to promote DNA repair. Cell 114, 599–610.
Bloom, J., Amador, V., Bartolini, F., DeMartino, G., and Pagano, M. (2003).
Proteasome-mediated degradation of p21 via N-terminal ubiquitinylation.
Cell 115, 71–82.
Chen, J., Lin, J., and Levine, A.J. (1995a). Regulation of transcription func-
tions of the p53 tumor suppressor by the mdm-2 oncogene. Mol. Med. 1,
In vitro E2 ubiquitin conjugation
The reaction was carried out by mixing bacterially expressed UbcH5B (100
ng) with rabbit E1 (100 ng) and 2 × 104cpm of32P-ubiquitin in ubiquitylation
reaction buffer. After incubating the mixture at room temperature for 5 min,
the reaction was stopped by addition of 4× nonreducing SDS sample buffer
and heated at 100°C for 3 min. The samples were then separated on 18%
SDS-PAGE and visualized by autoradiography.
Chen, Z., Hagler, J., Palombella, V.J., Melandri, F., Scherer, D., Ballard, D.,
and Maniatis, T. (1995b). Signal-induced site-specific phosphorylation
targets I kappa B alpha to the ubiquitin-proteasome pathway. Genes Dev.
Chen, K., Albano, A., Ho, A., and Keaney, J.F.J. (2003). Activation of p53 by
oxidative stress involved platelet-derived growth factor-beta receptor medi-
ated ataxia telangiectasia mutated (ATM) kinase activation. J. Biol. Chem.
DEVDase activity was determined as previously described (Yang et al.,
1998). Briefly, the cells were lysed with 50 mM HEPES buffer (pH 7.4) con-
taining 10% sucrose and 0.1% Triton X-100. After centrifugation, the super-
natants were incubated with 50 ?M ac-DEVD-AFC (Biomol, Meeting, PA) in
the presence of 10 mM DTT for 30 min at room temperature. Fluorescence
was measured using a CytoFluor multiwell plate reader (PerSeptive Biosys-
tems, Foster City, CA).
Chene, P. (2003). Inhibiting the p53–MDM2 interaction. Nat. Rev. Cancer 3,
Chipuk, J.E., and Green, D.R. (2003). p53’s believe it or not: Lessons on
transcription-independent death. J. Clin. Immunol. 23, 355–361.
Clarke, D.J. (2002). Proteolysis and the cell cycle. Cell Cycle 1, 233–234.
Coulombe, P., Rodier, G., Bonneil, F., Thibault, P., and Meloche, S. (2004).
N-terminal ubiquitination of thet extracellular signal-regulated kinase 3 and
p21 directs their degradation by the proteasome. Mol. Cell. Biol. 24,
Cells were lysed with the reporter lysis buffer (Promega, Madison, WI). Fol-
lowing centrifugation, luciferase activity in the supernatant was measured
with reagents in the luciferase assay system according to manufacturer’s
instructions (Promega, Madison, WI).
Dai, M.S., and Lu, H. (2004). Inhibition of MDM2-mediated p53 ubiquitina-
tion and degradation by ribosomal protein L5. J. Biol. in press.
Cell death assays
In some experiments, cell death was directly assessed by trypan blue exclu-
sion under microscope. In others, the percentage of cells with sub-G1 DNA
content was determined as previously described (Rowan et al., 1996).
Briefly, cells were harvested, washed with PBS, and fixed with 75% ethanol
overnight at −20°C. Following centrifugation and removal of residual etha-
nol, cell pellets were resuspended in 0.5 ml PBS and incubated with 0.2
mg/ml RNase A and 20 ?g/ml propidium iodide for 1 hr at room temperature
in the dark. The samples were then analyzed with a FACSCalibur (Becton
Dickinson, San Jose, CA).
Dai, M.S., Zeng, S.X., Jin, Y., Sun, X.X., David, L., and Lu, H. (2004). Ribo-
somal protein L23 activates p53 by inhibiting MDM2 function in response
to ribosomal perturbation but not to translation inhibition. Mol. Biol. Cell 24,
Davydov, I.V., Woods, D., Safiran, Y.J., Oberol, P., Fearnhead, H.O., Fang,
S., Jensen, J.P., Weissman, A.M., Kenten, J.H., and Vousden, K.H. (2004).
Assay for ubiquitin ligase activity: High-throughput screen for inhibitors of
HDM2. J. Biomol. Screen. 9, 695–703.
Dornan, D., Wertz, I., Shimizu, H., Arnott, D., Frantz, G.D., Dowd, P.,
CANCER CELL : JUNE 2005557
A R T I C L E
O’Rourke, K., Koeppen, H., and Dixit, V.M. (2004). The ubiquitin ligase
COP1 is a critical negative regulator of p53. Nature 429, 86–92.
K.H. (2003). Regulation of HDM2 activity by the ribosomal protein L11. Can-
cer Cell 3, 577–587.
Ettenberg, S.A., Magnifico, A., Cuello, M., Nau, M.M., Rubinstein, Y.R., Yar-
den, Y., Weissman, A.M., and Lilowitz, S. (2001). Cbl-b-dependent coordi-
nated degradation of the epidermal growth factor receptor signaling com-
plex. J. Biol. Chem. 276, 27677–27684.
Lorick, K.L., Jensen, J.P., Fang, S., Ong, A.M., Hatakeyama, S., and Weiss-
man, A.M. (1999). RING fingers mediate ubiquitin-conjugating enzyme (E2)-
dependent ubiquitination. Proc. Natl. Acad. Sci. USA 96, 11364–11369.
Lowe, S.W., Ruley, H.E., Jacks, T., and Housman, D.E. (1993). p53-depen-
dent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74,
Fang, S., and Weissman, A.M. (2004). A field guide to ubiquitylation. Cell.
Mol. Life Sci. 61, 1546–1561.
Fang, S., Jensen, J.P., Ludwig, R.L., Vousden, K.H., and Weissman, A.M.
(2000). Mdm2 is a RING finger-dependent ubiquitin protein ligase for itself
and p53. J. Biol. Chem. 275, 8945–8951.
Meek, D.W., and Knippschild, U. (2003). Posttranslational modifications of
MDM2. Mol. Cancer Res. 1, 1017–1026.
Mendrysa, S.M., McElwee, M.K., Michalowski, J., O'Leary, K.A., Young,
K.M., and Perry, M.E. (2003). mdm2 is critical for inhibition of p53 during
lympopoiesis and the response to ionizing irradiation. Mol. Cell. Biol. 23,
Fazeli, A., Steen, R.G., Dickinson, S.L., Bautista, D., Dietrich, W.F., Bronson,
R.T., Bresalier, R.S., Lander, E.S., Costa, J., and Weinberg, R.A. (1997). Ef-
fects of p53 mutations in apoptosis in mouse intestinal and human colonic
adenomas. Proc. Natl. Acad. Sci. USA 94, 10199–10204.
Michael, D., and Oren, M. (2003). The p53-Mdm2 module and the ubiquitin
system. Semin. Cancer Biol. 13, 49–58.
Gu, J., Nie, N., Wiederschain, D., and Yuan, Z.-M. (2001). Identification of
p53 sequence elements that are required for MDM2-mediated nuclear ex-
port. Mol. Cell. Biol. 21, 8533–8546.
Mihara, M., Erster, S., Zaika, A., Petrenko, O., Chittenden, T., Pancoska, P.,
and Moll, U.M. (2003). p53 has a direct apoptogenic role at the mito-
chondria. Mol. Cell 11, 577–590.
Haupt, S., and Haupt, Y. (2004). Improving cancer therapy through p53
management. Cell Cycle 3, 912–916.
Nakano, K., Balint, E., Ashcroft, M., and Vousden, K.H. (2000). A ribonucleo-
tide reductase gene is a transcriptional target of p53 and p73. Oncogene
Hu, G., and Fearon, E.R. (1999). Siah-1 N-terminal RING domain is required
for proteolysis function, and C-terminal sequences regulate oligomerization
and binding to target proteins. Mol. Cell. Biol. 19, 724–732.
Rowan, S., Ludwig, R.L., Haupt, Y., Bates, S., Lu, X., Oren, M., and Vous-
den, K.H. (1996). Specific loss of apoptotic but not cell cycle arrest function
in a human tumour derived p53 mutant. EMBO J. 15, 827–838.
Issaeva, N., Bozko, P., Enge, M., Protopopova, M., Verhoef, L.G., Masucci,
M., Pramanik, A., and Selinova, G. (2004). Small molecule RITA binds to
p53, blocks p53-HDM-2 interaction and activates p53 function in tumors.
Nat. Med. 10, 1321–1328.
Ryan, K.M., O'Prey, J., and Vousden, K.H. (2004). Loss of nuclear factor-
kappaB is tumor promoting but does not substitute for loss of p53. Cancer
Res. 64, 4415–4418.
Jin, A., Itahana, K., O'Keefe, K., and Zhang, Y. (2004). Inhibition of HDM2
and activation of p53 by ribosomal protein L23. Mol. Cell. Biol. 24, 7669–
Scheffner, M., Huibregtse, J.M., Vierstra, R.D., and Howley, P.M. (1993). The
HPV-16 E6 and E6-AP complex functions as a ubiquitin-protein ligase in
the ubiquitination of p53. Cell 75, 495–505.
Jones, S.N., Roe, A.E., Donehower, L.A., and Bradley, A. (1995). Rescue of
embryonic lethality in Mdm2-deficient mice by absence of p53. Nature 378,
Scheffner, M., Huibregtse, J.M., and Howley, P.M. (1994). Identification of a
human ubiquitin-conjugating enzyme that mediates the E6-AP-dependent
ubiquitination of p53. Proc. Natl. Acad. Sci. USA 91, 8797–8801.
Kern, S.E., Pietenpol, J.A., Thiagalingam, S., Seymour, A., Kinzler, K.W., and
Vogelstein, B. (1992). Oncogenic forms of p53 inhibit p53-regulated gene
expression. Science 256, 827–830.
Schmitt, C.A., Fridman, S.J., Yang, M., Baranov, E., Hoffman, R.M., and
Lowe, S.W. (2002). Dissecting p53 tumor suppressor functions in vivo. Can-
cer Cell 1, 289–298.
Kubbutat, M.H.G., Jones, S.N., and Vousden, K.H. (1997). Regulation of
p53 stability by Mdm2. Nature 387, 299–303.
Selinova, G. (2001). Mutant p53: The loaded gun. Curr. Opin. Investig. Drugs
Lai, Z., Yang, T., Kim, Y.B., Sielecki, T.M., Diamond, M.A., Strack, P., Rolfe,
M., Caligiuri, M., Benfield, P.A., Auger, K.R., and Copeland, R.A. (2002).
Differentiation of Hdm2-mediated p53 ubiquitination and Hdm2 autoubiqui-
tination activity by small molecular weight inhibitors. Proc. Natl. Acad. Sci.
USA 99, 14734–14739.
Sheaff, R.J., Singer, J.D., Swanger, J., Smitherman, M., Roberts, J.M., and
Clurman, B.E. (2000). Proteasomal turnover of p21Cip1 does not require
p21Cip1 ubiquitination. Mol. Cell 5, 403–410.
Lane, D.P., and Lain, S. (2002). Therapeutic exploitation of the p53 pathway.
Trends Mol. Med. 8, S38–S42.
Sherr, C.J., and Weber, J.D. (2000). The ARF/p53 pathway. Curr. Opin.
Genet. Dev. 10, 94–99.
Leng, R.P., Lin, Y., Ma, W., Wu, H., Lemmers, B., Chung, S., Parant, J.M.,
Lozano, G., Hakem, R., and Benchimol, S. (2003). Pirh2, a p53-inducible
ubiquitin-protein ligase, promotes p53 degradation. Cell 112, 779–791.
Shieh, S.-Y., Ikeda, M., Taya, Y., and Prives, C. (1997). DNA damage-
induced phosphorylation of p53 alleviates inhibition by MDM2. Cell 91,
Li, M., Brooks, C.L., Wu-Baer, F., Chen, D., Baer, R., and Gu, W. (2003).
Mono- versus polyubiquitination: Differential control of p53 fate by Mdm2.
Science 302, 1972–1975.
Siliciano, J.D., Canman, C.E., Taya, Y., Sakaguchi, K., Appella, E., and Kas-
tan, M.B. (1997). DNA damage induces phosphorylation of the amino termi-
nus of p53. Genes Dev. 11, 3471–3481.
Liu, G., Parant, J.M., Lang, G., Chau, P., Chavez-Reyes, A., El-Naggar, A.K.,
Multani, A., Chang, S., and Lozano, G. (2004). Chromosomal stability, in the
absence of apoptosis, is critical for suppression of tumorigenesis in Trp53
mutant mice. Nat. Genet. 36, 63–68.
Stommel, J.M., and Wahl, G.M. (2004). Accelerated MDM2 auto-degrada-
tion induced by DNA-damage kinases is required for p53 activation. EMBO
J. 23, 1547–1556.
Symonds, H., Krall, L., Remington, L., Saenzrobles, M., Lowe, S., Jacks, T.,
and Van Dyke, T. (1994). p53-dependent apoptosis suppresses tumor
growth and progression in vivo. Cell 78, 703–711.
Ljungman, M. (2000). Dial 9-1-1 for p53: Mechanisms of p53 activation by
cellular stress. Neoplasia 2, 208–225.
Lohrum, M.A.E., Woods, D.B., Ludwig, R.L., Bálint, E., and Vousden, K.H.
(2001). C-terminal ubiquitination of p53 contributes to nuclear export. Mol.
Cell. Biol. 21, 8521–8532.
Vassilev, L.T., Vu, B.T., Graves, B., Carvajal, D., Podlaski, F., Filipovic, Z.,
King, N., Kammlott, U., Lukacs, C., Klein, C., et al. (2004). In vivo activation
of the p53 pathway by small-molecular antagonists of MDM2. Science 303,
844–848.Lohrum, M.A., Ludwig, R.L., Kubbutat, M.H.G., Hanlon, M., and Vousden,
CANCER CELL : JUNE 2005
A R T I C L E Download full-text
Vogelstein, B., Lane, D., and Levine, A.J. (2000). Surfing the p53 network.
Nature 408, 307–310.
(2004). Mdm2-mediated NEDD8 conjugation of p53 inhibits its transcrip-
tional activity. Cell 118, 83–97.
Vousden, K.H. (2002). Activation of the p53 tumor suppressor gene. Bio-
chim. Biophys. Acta 1602, 47–59.
Xu, Y. (2003). Regulation of p53 responses by post-translational modifica-
tions. Cell Death Differ. 10, 400–403.
Vousden, K.H., and Lu, X. (2002). Live or let die: The cell's response to p53.
Nat. Rev. Cancer 2, 594–604.
Yang, Y., and Yu, X. (2003). Regulation of apoptosis: The ubiquitous way.
FASEB J. 17, 790–799.
Weissman, A.M. (2001). Themes and variations on ubiquitylation. Nat. Rev.
Mol. Cell Biol. 2, 169–178.
Yang, Y., Liu, Z., Tolosa, E., Yang, J., and Li, L. (1998). Triptolide induces
apoptotic death of T lymphocytes. Immunopharmacology 40, 139–149.
Williams, S.A., and McConkey, D.J. (2003). The proteasome inhibitor bortez-
omib stabilizes a novel active form of p53 in human LNCaP-Pro5 prostate
cancer cells. Cancer Res. 63, 7338–7344.
Zhang, Y., Wolf, G.W., Bhat, K., Jin, A., Allio, T., Burkhart, W.A., and Xiong,
Y. (2003). Ribosomal protein L11 negatively regulates oncoprotein MDM2
and mediates a p53-dependent ribosomal-stress checkpoint pathway. Mol.
Cell. Biol. 23, 8902–8912. Xirodimas, D.P., Saville, M.K., Bourdon, J.C., Hay, R.T., and Lane, D.P.
CANCER CELL : JUNE 2005 559