Homo-PROTACs: Bivalent small-molecule dimerizers of the VHL E3 ubiquitin ligase to induce self-degradation

Article abstract of DOI:10.1038/s41467-017-00954-1

E3 ubiquitin ligases are key enzymes within the ubiquitin proteasome system which catalyze the ubiquitination of proteins, targeting them for proteasomal degradation. E3 ligases are gaining importance as targets to small molecules, both for direct inhibition and to be hijacked to induce the degradation of non-native neo-substrates using bivalent compounds known as PROTACs (for 'proteolysis-targeting chimeras'). We describe Homo-PROTACs as an approach to dimerize an E3 ligase to trigger its suicide-type chemical knockdown inside cells. We provide proof-of-concept of Homo-PROTACs using diverse molecules composed of two instances of a ligand for the von Hippel-Lindau (VHL) E3 ligase. The most active compound, CM11, dimerizes VHL with high avidity in vitro and induces potent, rapid and proteasome-dependent self-degradation of VHL in different cell lines, in a highly isoform-selective fashion and without triggering a hypoxic response. This approach offers a novel chemical probe for selective VHL knockdown, and demonstrates the potential for a new modality of chemical intervention on E3 ligases.

Full text of DOI:10.1038/s41467-017-00954-1

A R T I C L E
DOI: 10.1038/s41467-017-00954-1
OPEN
H om o - PR O T AC s: b i v a l e n t s m a l l - m o l e c u l e
d im e r i z e rs o f t h e V H L E 3 u b i q u i t i n l i g a s e t o i n d u c e
s e l f - d e g r a d a t i o n
1 , 2
1
1
1
1
3
C
h
i
a
r
a
M
a
n
i
a
c
i
,
S
co
t
t
J
.
H
u
g
h
e
s
,
A
nd
r
e
a
T
es
t
a
,
W
e
n
z
h
a
n
g
C
h
e
n
,
D
o
u
g
l
a
s
J
.
L
a
m
o
n
t
,
S
o
n
i
a
R
o
ch
a
,
2
4
1
D
ar
i
o
R
.
A
l
e
s
s
i
,
R
o
b
e
r
t
o
R
o
m
e
o
&
A
le
s
s
i
o
C
i
u
l
li
E 3 u b iq u i t i n l ig a s e s a r e k e y e n z y m e s w i th i n t h e u b i q u i t i n p ro t e a s o m e s y s t e m w hi c h c a t a l y z e
t h e u b iq u i t i na t i o n o f p r o t e i n s , t a r g e t i n g t h e m f o r p r o t e a s o m a l d eg r a d a t i on . E 3 l i g a s e s a r e
g a i n i n g i mp o r ta n c e a s t a r g e t s t o s m a l l m o l e cu l e s , b o th f or d i re c t i n h i bi ti on a n d t o b e h i j a c k ed
t o i n d u c e t h e d e g ra d a t i o n o f n o n - n a t i v e n e o - s u b s tr a t es u s i n g b i v a l e n t c o m p o u nd s k n o wn a s
P R O T A Cs ( f o r ‘p r ot e o l y s i s - t a r g et i n g c h i m e r a s ’) . W e d e s c r i be H om o - P R O T A C s a s a n
a p pr o ac h t o d im e r iz e a n E 3 l i g a s e t o t r i g g e r i t s s u i ci d e - t y pe c h e m i c a l k n o c k d o wn i n s i d e c e l l s .
W e p r ov i de p r oo f - o f - c o n c e p t o f H o m o - P R O TA C s u s i n g d i v e r s e m o l e c u l e s c o m p o s e d o f t wo
i n s t a n c e s o f a l ig a n d f o r t h e v o n H i p p e l - L i n d a u ( V HL ) E 3 l i g a s e. T h e m o s t a c t i v e c o m p o u nd ,
C M 11 , d im e r iz e s V H L w i th h i g h a v i d i ty i n v i tr o a n d i n d u c e s p ot e nt , r a pi d a n d p ro t ea s o m e -
d e pe nd e nt s e lf -d e g ra d a t i o n o f V H L i n d i f f e r e n t c e l l l i n e s , i n a h i g h l y i s of o rm - s e l e c t i ve f a s h i o n
a n d w i t h ou t t r ig g e r in g a h y p o x i c r e s p o n s e . T h i s a p p r o a c h o ff er s a n ov e l c h e m i c a l p ro b e f or
s e le c t i ve V H L k n o c k d o w n , a n d d e m o n s t r a t e s t h e p o t e n t i a l f or a n e w m o d a l i t y o f c h e m i c a l
i n t e r v e nt io n o n E 3 l ig a s e s .
1 Division of Biological Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, James Black Centre, Dow Street, Dundee, Scotland DD1
5EH, UK. 2 Medical Research Council Protein Phosphorylation and Ubiquitylation Unit, School of Life Sciences, University of Dundee, James Black Centre,
Dow Street, Dundee, Scotland DD1 5EH, UK. 3 Centre for Gene Regulation and Expression, School of Life Sciences, University of Dundee, James Black Centre,
Dow Street, Dundee, DD1 5EH Scotland, UK. 4 Dipartimento di Scienze Chimiche, Biologiche, Farmaceutiche ed Ambientali, University of Messina, Polo
Universitario Viale SS. Annunziata SNC, Messina, 98168, Italy. Correspondence and requests for materials should be addressed to
A.C. (email: a.ciulli@dundee.ac.uk)
N
A
T
U
R
E
C
OM
M
U
N
I
C
A
T
I
ON
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
46
7
-
0
1
7
-
0
0
9
5
4
-
1
w
ww
.n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
re
c
o
m
m
u
n
ic
a
t
io
n
s
1
|
|
|

A R T I C L E
N A T U R E C O M MU N ICA T IO N S | D O I: 1 0 . 1 0 3 8 / s 4 1 46 7 - 0 17 - 0 0 9 5 4 -1
3 ubiquitin ligases are emerging as attractive targets for depletion can have a more sustained cellular effect compared to
small-molecule modulation and drug discovery1–3. E3s target inhibition, and can overcome compensatory cellular feed-
bring a substrate protein and ubiquitin in close proximity to back mechanisms, such as increase in target levels45. Crucially,
E
each other to catalyze the transfer of a ubiquitin molecule to the work from us and others have shown that PROTAC molecules
substrate4, 5. Substrate ubiquitination can trigger different cellular can exhibit an added layer of selectivity for protein degradation
outcomes6, of which one of the best characterized is beyond the intrinsic binding selectivity of the warhead ligand35,
poly-ubiquitination and subsequent proteasomal degradation7, 8
.
38, 41. Our recent structural work with Brd4-selective PROTACs
The human genome comprises > 600 predicted E3 ligases that targeting CRL2-VHL revealed the importance of specific ligand-
play important roles in normal cellular physiology and disease induced PPIs between the ligase and the target, which contribute
states, making them attractive targets for inhibitor discovery9. to cooperative formation of stable and highly populated ternary
However, E3 ligases do not comprise deep and ‘druggable’ active complexes38.
sites for binding to small molecules2. Blockade of E3 ligase
We hypothesized that it could be possible to trigger an E3
activity therefore requires targeting of protein–protein interac- ligase to induce its own degradation, by designing tailored
tions (PPIs), and the often extended, flat and solvent-exposed PPI homo-bivalent PROTACs that recruit two molecules of the same
surfaces make it a challenge for drug design10. Only few potent E3 ligase. The idea was simple, namely that this compound class
inhibitors have been developed to date, mostly compounds that could act as chemical inducers of dimerization (CID)46–48
,
bind to the E3 substrate recognition site2, 11, for example forming a ternary complex in which the E3 acts as the enzyme
MDM212, 13, inhibitor of apoptosis proteins (IAPs)14–16, the von and the neo-substrate at the same time. To provide proof-of-
Hippel-Lindau (VHL) ligase17–19 and KEAP120, 21. Inhibitors of concept for the approach, we designed, synthesized and tested
E3:substrate interaction can exhibit a discrepancy in effective homo-bivalent molecules aiming to target CRL2-VHL, the E3
concentrations between biophysical binding and cellular effi- ligase that targets for ubiquitination and degradation the hypoxia
cacy19, 22, due to competition from high-affinity endogenous inducible factor alpha subunit (HIF-α) under normoxic condi-
substrates that markedly increase their cellular concentration as a tions49. We show the most active compound, CM11, made of two
consequence of the inhibition. This poses limitations, such as instances of a potent hydroxyproline (Hyp) containing VHL
incomplete blockade of enzyme activity and the need to use high ligand19, avidly forms a 1:2 complex with VHL, and induces
inhibitor concentrations, which can lead to off-target effects and potent and preferential isoform-selective degradation of VHL. We
cytotoxicity. Moreover, E3 ligases are multi-domain and multi- call these Homo-PROTACs, as a new modality to induce che-
subunit enzymes, and targeting an individual binding site leaves mical knockdown of E3 ligases.
other scaffolding regions untouched and other interactions
functional. As a result, E3 ligase inhibition may be ineffective or
fail to recapitulate genetic knockout or knockdown. New che- Results
mical modalities to target E3 ligases are therefore demanded.
Rational design. Design of VHL Homo-PROTACs began with
E3 ligases are not merely targets for inhibition. Compounds of careful consideration of the position of derivatization on two
natural or synthetic origin have been discovered that bind to E3 potent VHL ligands recently characterized by our group, VH032
ligases and promote target recruitment. These interfacial com- and VH298 (Fig. 1a, b)18, 19. To retain the strong binding
pounds induce de novo formation of ligase-target PPIs, effectively affinity that characterizes the ligand, co-crystal structures were
hijacking E3 activity towards the neo-substrates, for targeted analyzed to identify solvent-exposed regions from where the
protein degradation23, 24. One class of hijackers of E3 ligase ligands could be derivatized without perturbing their binding
activity comprises monovalent compounds. These so-called modes (Fig. 1a). This analysis and consideration of previous
‘molecular glues’ include the plant hormone auxin, which binds VHL-targeting PROTACs pointed to the methyl group of the
to the Cullin RING ligase (CRL) CRL1-TIR1 to target tran- left-hand site (LHS) terminal acetyl group of VH032 as a suitable
scriptional repressor proteins of the Aux/IAA family;25 the point of connection for a linker35, 36. A second solvent-exposed
immunomodulatory drugs (IMiDs) thalidomide, lenalidomide, position available for derivatization was the phenyl group on the
pomalidomide and analog CC-885, that all share binding to right-hand side (RHS), as previously employed with PROTACs
cereblon (CRBN), a subunit of the CRL4-CRBN ligase, and targeting the Halotag50. To investigate the impact of derivatiza-
redirect CRBN activity to different substrates26–31. More recently, tion, we designed three classes of Homo-PROTACS: (a) sym-
sulfonamide anti-cancer drug indisulam was found to induce metric via the LHS acetyl group of each ligand (Fig. 1c); (b)
degradation of the splicing factor RBM39 via recruiting CRL4- symmetric via the RHS phenyl group (Fig. 1d); and
DCAF15 activity32, 33. Another class of degrader compounds that (c) asymmetric via the acetyl group in one warhead and the
display a similar mechanism of action comprises bivalent mole- phenyl in the other (Fig. 1e). In the cases b and c, at the
cules known as Proteolysis-Targeting Chimeras (PROTACs). underivatized terminal LHS we decided to retain either an acetyl
PROTACs comprise two warheads—one for ligase recruitment (as in VH032) or a cyano-cyclopropyl moiety (as in VH298), a
and a second one for target-binding—joined by a linker34. For- modification that led to increased binding affinities, cell perme-
mation of a ternary complex between the PROTAC, the ligase ability and cellular activities in the context of the VHL inhibitor
and the target triggers proximity-induced target ubiquitination alone19. To evaluate the potential impact of linker length, linkers
and degradation. Potent and cell-active PROTACs have been comprised of polyethylene glycol chains with either three, four or
developed for recruiting different ligases, including five ethylene glycol units were chosen to connect the two VHL
CRL2-VHL35–38, CRL4-CRBN39–42, and IAPs43, 44. Targets ligands.
successfully degraded by PROTACs include BET proteins Brd2,
It is known that the trans epimer of Hyp is an absolute
Brd3 and Brd435, 37–40, FKBP39, protein kinases36, 41, amongst requirement for VHL binding, and that the corresponding cis
others36, 43. An attractive feature of bivalent degrader molecules is epimer abrogates binding to VHL, both within the context of a
their sub-stoichiometric catalytic activity36, which does not native HIF substrate peptide51, and VHL ligands19, 36. We
require full occupancy of the target-binding site as with con- therefore designed two different PROTACs based on the structure
ventional inhibitors, leading to degrading concentrations that can of the first series (Fig. 1c), with the aim to use them as controls: a
be orders of magnitude lower than the inhibitory concentrations cis-cis epimer, expected to be completely inactive, and a cis-trans
of their constitutive parts alone. Furthermore, induced target epimer compound, expected to retain binding to a single VHL
2
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D O I : 1 0 . 1 0 3 8 / s 4 1 4 6 7 - 0 1 7 - 0 0 9 5 4 - 1 w ww .n a t u r e . c o m / n a t u re c o m m u n i c a t i o n s
|
|
|

N A T UR E C O M M UN I CA T I O NS | D OI : 1 0. 1 0 3 8 / s 4 1 46 7 - 01 7 - 009 5 4 - 1
A R T I C L E
molecule in a 1:1 fashion, thus potentially acting as inhibitor but
not as degrader.
a
Synthesis. For the synthesis of the first class of Homo-PROTACs
(Fig. 1c), symmetric PEG linkers 4, 5 and 6 bearing free
carboxylate groups at either ends were obtained by reaction of
tert-butyl bromoacetate with tri-, tetra- and penta-ethylene glycol
in the presence of NaH in dioxane and followed, after purifica-
tion, by treatment with 50% TFA in DCM (Fig. 2).
The final compounds CM9, CM10 and CM11 were obtained by
amide coupling of the VHL ligand 7 (prepared as previously
described)35 with linkers 4, 5 and 6, in a 2:1 ratio, respectively, in
the presence of HATU as the coupling agent and DIPEA as the
base (Fig. 2). For the synthesis of the symmetric cis-cis compound
CMP98, compound 835 was coupled with linker 6 to afford the
desired product (Fig. 2). For the preparation of the asymmetric
cis-trans compound CMP99, a synthetic route toward the
synthesis of the monoprotected di-carboxylate linker was estab-
lished. Pentaethylene glycol was the linker of choice because of
ease of purification compared to longer PEGs, and at the same
time yielding a control compound of average linker length (PEG-
4 in this case). Pentaethylene glycol was converted in monobenzyl
ether 9 in 71% yield, which was reacted with tert-butyl bro-
moacetic acid under biphasic conditions (DCM / 37% aq. NaOH
and stoichiometric tetrabutyl ammonium bromide). After
deprotection of the benzyl group by catalytic hydrogenation,
formation of the carboxylic acid moiety was achieved by oxida-
tion with TEMPO and bis-acetoxy iodobenzene (BAIB), deliver-
ing compound 11 in 65% yield (Fig. 3). Compound 7 was then
coupled with linker 11 using the condition described above,
affording compound 12. Deprotection of the tert-butyl group
using TFA and subsequently coupling with 8 afforded CMP99 in
66% yield (Fig. 3).
For the synthesis of the second class of symmetric
Homo-PROTACs (Fig. 1d), it was decided to utilize compounds
17 and 18 as VHL warheads. Common precursor 16 was synthesized
following a previously reported procedure50, with minor modification
that led to yield and purity improvements (see Supplementary
Notes 1–3). Indeed, we observed that the use of HATU in
combination with HOAT for the coupling steps of both Boc-L-
Hyp and Boc-tert-leucine led to the formation of only the desired
products, avoiding the formation of a bis-acylate secondary
product50, instead prominent when HATU was used alone.
Compound 17 or 18 were obtained by treatment of compound 16
with 1-cyanocyclopropanecarboxylic acid in presence of HATU,
HOAT and DIPEA or acetylimidazole and TEA (see Supplementary
Methods). Synthesis of 17 was also performed using acetic anhydride,
but during this reaction it was observed the formation of a secondary
product di-acetylated, not only at the desired position but also at the
hydroxyl group of the phenyl ring, which could however be
separated. The PEG linkers for this class of compound were designed
to contain a methanesulfonate group at either end, which could be
coupled in a single step with the phenol of the VHL ligand. Linker 19
was prepared by mesylation of pentaethylene glycol and reacted with
either compounds 17 or 18 in a 1:2 ratio in the presence of K2CO3 to
afford CMP106 and CMP108, respectively, in good yield (Scheme 4).
For the synthesis of asymmetric Homo-PROTACs, PEG 10 was
converted in to the mesylated derivative 20 and reacted with 17 or 18
to obtain 21 and 22, respectively in good yield (see Supplementary
Notes 1–3). Final compounds CMP112 and CMP113 were obtained
in good yield upon deprotection of the tert-butyl group and amide
coupling with compound 7 (see Supplementary Notes 1–3).
b
OH
O
OH
O
O
O
N
N
NC
S
S
NH
NH
N
N
O
O
H
H
N
N
VH032
VH298
c
OH
O
N
S
H
N
NH
O
O
O
linker
N
N
O
HN
H
N
S
N
O
n = 3,4,5
(Acetyl-acetyl)
OH
OH
d
O
R
N
N
O
H
O
NH
O
N
O
S
N
S
linker
O
HN
HO
O
H
N
N
R
O
n =5
(Phenyl-phenyl)
OH
e
N
O
N
S
O
NH
S
linker
N
H
O
O
N
HN
HO
O
O
H
N
N
R
O
n =5
(Phenyl-acetyl)
,
R= CH3
O
linker
=
n
NC
Fig. 1 S tr u c tu re - g u i d e d d e s i gn o f H o m o - PR OT AC s t o i n d u c e V H L
d i m e ri z a ti on . a C r y s t a l s t r u c t u r e s o f V HL i n c o m p l ex w it h V H2 9 8 ( P D B
c o de 5 L L I ) . V H L i s s h o wn a s a p i n k s u r fa c e a n d t h e b o u n d l ig a n d a s s t i ck s
r e p re s e nt a ti o n w i th p u r p l e c a r b o n s , n i t r o g en a t o m s i n b l u e , o x y g e n i n r e d
a n d s u l p h u r i n d a r k y e l l o w . b C h e m i c al s t r u c t u r e o f V HL i n h i b i t o r s V H 0 3 2
a n d V H 2 98 . c–e G e n e r a l c h em i c a l s t r u c t u r e s a n d d e s i g n o f H o m o -
P RO T A C s c o m p o u n d s . L i n k a ge s i t e s a t t h e a c e t y l a n d p h e n y l g r o u p s a r e
i n di c a t e d i n r e d
1
9
Biological evaluation. We next tested all our Homo-PROTACs
by monitoring protein levels after 10 h of compound treatment at
1 µM concentration in HeLa cells (Fig. 4a). We observed the
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
46
7
-
0
1
7
-
0
0
9
5
4
-
1
w
ww
.n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
re
c
o
m
m
u
n
i
c
a
t
io
n
s
3
|
|
|

A R T I C L E
N A T U R E C O M MU N ICA T IO N S | D O I: 1 0 . 1 0 3 8 / s 4 1 46 7 - 0 17 - 0 0 9 5 4 -1
O
O
tert-butyl bromoacetate,
TFA,
DCM, 1 h, r.t.
Quantitative yield
O
OH
O
O
O
HO
O
O
O
NaH ,Dioxane, 16 h, r.t.
HO
H
3-5
n
O
n
O
1 n =3 19%
2 n =4 10%
3 n =5 14%
4 n =3
5 n =4
6 n =5
X
Y
DIPEA, HATU,
DMF,
rt, 30 min
O
N
H
N
O
Y
S
O
N
O
NH
O
HN
N
O
O
H
S
N
N
n
O
X
OH
CM09 X=OH Y=H
CM10 X=OH Y=H
CM11 X=OH Y=H
n =3 40%
n =4 30%
n =5 50%
n =5 45%
N
S
NH
CMP98 X=H Y=OH
H2N
N
O
O
7
8
trans -Hyp
cis -Hyp
Fig. 2 S y n t h e s i s o f s y m m et r ic h o m o -P R OT AC c o m p o u n d s d e r i v a t i z e d f r o m t h e t e r m i n a l a c e t y l g r o u p . T h e s h o w n r o u t e y i e l d e d s y m m e t r ic trans-trans
C M 0 9, C M 1 0 a n d C M 1 1 a n d n e g a t i ve c o n t r o l s y m m e t r ic cis- cis c o m p o u n d C M P 9 8
striking effectiveness of CM09, CM10 and CM11 in inducing
It is known that VHL RNAi is insufficient to induce significant
VHL depletion in cells (Fig. 4a), and a remarkably selective HIF stabilization (see also Fig. 4b) and does not cause detectable
degradation for the closely migrating bands corresponding to the upregulation in HIF activity56. On the basis of these considera-
long isoform of VHL52, 53, preferentially over the short isoform. tions, and the relatively low HIF stabilization observed with active
The VHL gene includes three exons and it encodes two major Homo-PROTACs CM09–11 (Fig. 4a), it was presumed that
isoforms of VHL: a 213 amino-acid long, 30 kDa form (pVHL30) compound treatment would not induce a HIF-dependent hypoxic
and a 160 amino-acid long, 19 kDa form (pVHL19). pVHL19 response. To confirm this, we first used a luciferase reporter
lacks a 53 amino-acid long amino-terminal domain or N-terminal assay57. Hypoxia response element (HRE) -luciferase reporter
tail (pVHL-N), which is instead present in pVHL30. Although HeLa-HRE and U2OS-HRE cells were treated with different
both isoforms are expressed in human cells, pVHL19 is the more concentrations of CM11 and at different times, and no increase in
prominent form in human tissues54. The most active compounds luciferase activity was detected relative to DMSO control
are symmetrically linked from the terminal LHS acetyl group of treatment (Supplementary Fig. 1). These results were confirmed
VH032. Linkage at different positions proved ineffective, sug- in a quantitative PCR with reverse transcription assay, where no
gesting a critical role played by the linking pattern. Control upregulation of mRNA levels of the known HIF-target genes CA9
compounds CMP98 and CMP99 were unable to induce degra- was detected (Supplementary Fig. 2). Together the data suggest
dation of VHL (Fig. 4a), demonstrating that Homo-PROTAC that un-degraded pVHL19 is sufficient to maintain low levels of
activity is dependent on productive bivalent recruitment of VHL HIF-1α.
by the trans epimer. The length of the linker also seemed to affect
We next turned our attention to further characterize the mode
cellular potency. Indeed, a decrease in effectiveness was observed of action of the protein degradation induced by the active Homo-
at shorter linker lengths, with CM10 and CM11 being the most PROTACs CM09–11. To interrogate their relative cellular
active compounds achieving total knockdown of pVHL30, fol- potency, dose-dependent treatments were performed at two
lowed by CM09 depleting 82% of the target protein. Interestingly, different time points, 4 and 24 h before harvesting. All
some degradation of the short isoform pVHL19 was also compounds confirmed preferential degradation of pVHL30 in a
observed, albeit low (around 10% depletion). Levels of Cullin2, concentration-dependent manner, relative to the corresponding
the central subunit of the CRL2-VHL complex55, were also DMSO control (see Fig. 5a for CM11, and Supplementary Fig. 3
reduced upon treatment with CM10 and CM11 by up to 22% for CM09 and CM10). CM11 proved the most potent compound,
(Fig. 4a). Treatments with CM10 and CM11 also showed inducing complete depletion of pVHL30 after 4 h already at 10
detectable albeit low increase in protein levels of the hydroxylated nM (DC99 = 10 nM, Fig. 5a). Selective pVHL30 knockdown was
form of HIF-1α (Hdy-HIF-1α, Fig. 4a). As the parent inhibitor retained after 24 h, with half-degrading concentration (DC50)
VH032 is completely ineffective at the same concentration of 1 between 10 and 100 nM. The effective degrading concentrations
µM (see ref. 19 and vide infra, Fig. 5), this effect cannot be due to of CM11 are > 3 orders of magnitude lower than the inhibitory
VHL inhibition and is thought to be the result of compound- concentrations of the constitutive ligand VH032 alone, which is
induced protein degradation. Levels of HIF-1α were, however, only active in cells at ~ 100 µM, underscoring the difference in
significantly lower than observed when VH032 was used at cellular efficacy between the two mode of actions. Cellular levels
concentrations > 100 µM (Fig. 4a, see also ref. 19). VHL knock- of Cullin2 decreased by up to 73% upon treatment with CM11
down by siRNA experiments in three different cell lines was (Fig. 5a). As previously observed, selective pVHL30 knockdown
consistent with CM11-induced knockdown, and confirmed that by Homo-PROTACs resulted in only minor increase in levels of
the bands observed indeed correspond to VHL (Fig. 4b).
HIF-1α, relative to hypoxia-inducing controls CoCl2, PHD
4
N
A
TU
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D O I : 1 0 . 1 0 3 8 / s 4 1 4 6 7 - 0 1 7 - 0 0 9 5 4 - 1 w ww .n a t u r e . c o m / n a t u re c o m m u n i c a t i o n s
|
|
|

N A T UR E C O M M UN I CA T I O NS | D OI : 1 0. 1 0 3 8 / s 4 1 46 7 - 01 7 - 009 5 4 - 1
A R T I C L E
Benzyl chloride
tert-butyl bromoacetate,
O
O
O
O
HO
H
O
H
NaH, DMF
0°C to r.t, O/N
71%
37% NaOH, TBABr, DCM
O
5
5
O
16 h, r.t.
80%
5
10
9
I. Pd/C, H2,
EtOH, r.t, 1 h
Quantitative yield
II.BAIB, TEMPO
ACN/H2O 1:1
3 h, r.t
OH
65%
O
O
N
O
O
S
7, DIPEA, HATU,HOAT
O
O
NH
HO
O
O
R'
N
DMF,
rt, 30 min
44%
O
4
O
4
H
O
N
O
12 R'= -t-butyl
TFA,
11
DCM, 1 h, r.t.
Quantitative yield
13 R'= –H
8, DIPEA, HATU,
DMF,
rt, 30 min
66%
OH
O
N
H
N
O
S
O
N
O
NH
O
HN
N
O
O
H
S
N
N
4
O
CMP99
OH
Fig. 3 S y n t h e s i s o f n e g a t i v e c o n t r o l h o m o -P RO TA C c o m p o u n d d e r i v a t i z e d f r o m t h e t e r m i n a l a c e t y l g r o u p . C o m p o u nd C M P 9 9 h a s cis-trans c o nfig ur at i on
inhibitor IOX2 and VH032 (Fig. 5a). However, when tested at levels of VHL in the two cell lines. CM09 and CM10 achieved
high micromolar concentrations, Homo-PROTACs acted prefer- complete degradation of the target protein after 2 h of treatment.
entially as VHL inhibitors over VHL degraders, consistent with CM11 confirmed to be the most potent compound also in U2OS,
the so-called ‘hook-effect’ whereby formation of binary achieving complete degradation of pVHL30 already after 1 h.
1:1 complexes competes with and eventually supersedes the Interestingly CM09 lost its cellular efficacy after 36 h. In contrast,
formation of the productive catalytic 2:1 complex58. Stabilization both CM10 and CM11 retained their efficacy even at these longer
of Hdy-HIF-1α upon treatment with all three compounds at time points (Supplementary Fig. 4).
100 µM was indeed comparable with the effect obtained with
To gain mechanistic insights in the mode of action of
VH032 alone (Fig. 5a for CM11, and Supplementary Fig. 3 for Homo-PROTACs, the dependency on CRL2-VHL and protea-
CM09 and CM10). To confirm the cellular activities of some activities was examined. The reliance of the
Homo-PROTACs in a different cell line, U2OS cells were treated Homo-PROTAC-induced protein degradation on CRL2-VHL
for 10 h with CM09, CM10 and CM11 using the same range of was assessed by inhibiting neddylation of Cullin2 using the NAE1
concentrations (1 nM–100 µM). A consistent profile of cellular inhibitor MLN4924, which blocks the activity of CRLs, including
activity was observed, confirming that the effects observed are CRL2-VHL59. Proteasome-dependency was interrogated by
independent from cell type (Supplementary Fig. 4).
treating cells with the proteasome inhibitor MG132. To limit
We next interrogated the time-dependent activity of the known cytotoxicity of MLN4924 and MG132, HeLa cells
Homo-PROTACs. Progressive removal of VHL protein over where pre-treated with MLN4924 for 3 h followed by MG132 for
time was observed, confirming selective depletion of pVHL30 30 min before adding CM11 to the media, and cells were
over the short isoform (Fig. 5b for CM11 and Supplementary incubated for further 4 h before collecting. Single treatments with
Fig. 5 for CM09 and CM10). In particular, CM11 was confirmed DMSO, MLN4924, MG132 and CM11 and combinations thereof
to be the most effective compounds, decreasing pVHL30 level by were performed to disentangle the individual and combined
more that 70% already after 2 h of treatment, and essentially to effects of compound treatments. Degradation of pVHL30 induced
completion after 8 h. The depletion effect was retained up to 12 h; by CM11 was completely abrogated when cells were pre-treated
however, interestingly, pVHL30 levels up to 11% were detected with MG132, establishing the expected proteasome-dependence
after 24–36 h treatment, to then decrease again after of the chemical intervention (Fig. 6). CM11-induced degradation
48 h. Incomplete degradation of pVHL was observed upon was also prevented by pre-treatment with MLN4924, confirming
treatment with CM09, even in the longer time points (Supple- the dependency on the activity of CRL2VHL (Fig. 6). The same
mentary Fig. 5). The results obtained treating U2OS cells were effect was observed when cells where co-treated with MLN4924
consistent with what observed in the previous experiment. and MG132 before CM11 (Fig. 6). Immunoblots of Cullin2 levels
However, in this cell line all three compounds induced complete confirmed the effective blockade of Cul2 neddylation by
degradation of pVHL30 over time (Supplementary Fig. 4). We MLN4924 (Fig. 6). To assess if CM11 degrading activity was
hypothesize that this could be due to the different expression dependent on VHL binding, a competition experiment was
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
5
|
|
|

A R T I C L E
N A T U R E C O M MU N ICA T IO N S | D O I: 1 0 . 1 0 3 8 / s 4 1 46 7 - 0 17 - 0 0 9 5 4 -1
a
<-- pVHL30
<-- pVHL19
20
15
100 127 19
3
1 108 62 121 110 126 114
100 113 102 92 91 104 104 111 89 97 93
100
75
Cullin2
100 126 112 96 78 123 112 123 117 105 104
150
100
Hdy-HIF-1α
150
100
HIF-1α
β-actin
U2OS
50
37
b
HEK 293
HeLa
<-- pVHL30
<-- pVHL19
20
15
150
Hdy-HIF-1α
β-actin
100
50
37
Fig. 4 H o m o - P R O T AC s C M0 9 , C M 10 a n d C M 11 i n d u c e s e l e c t i v e k n o c kd ow n o f t h e l o n g i s o fo r m o f V H L . a H e La c e l l s w e r e t r e a te d w i t h 0 .1 % D M SO ,
V H 03 2 ( 1 5 0 µM ) a n d 1 µM o f t h e i n d i c a t ed c o m p o u n d s f o r 1 0 h . A b u n d a n c e o f i n d i v i d u a l p r o t ei n s w a s a n a l y z e d b y w e s t e r n b l o tt i n g u s i n g c o r r e s p on di ng
s pe c ific a n t i b o d ie s a c c o r d in gl y a f t e r S DS - P A G E . T h e d a t a d e m o n s t r at e d i s a r e p r e s e n t a ti v e e x a m p l e o f e x p e r i m e n t s p e r f o r m e d i n b i o l o g i ca l d up li c a t es . b
D i f f e r e n t c e l l s l i n e s w e re t r e a t e d w it h s i R N A t a r g e t in g V HL p r o t e i n s o r n e g a t i v e c o n t r o l s i R NA ( f o r 4 8 h ) , a s w e l l a s w i t h C M 1 1 ( 1 µM ) o r 0 .1 % v/ v D M SO
5 2, 5 3
f o r 1 0 h . T h e b a n d s o b s er ve d t o s e l e c t i v el y d i s a p p e a r i n t h e p r e s e n ce o f C M 0 9–1 1 a r e c l o s e l y m i g r a t i n g i s o fo r m s o f p V HL 3 0 ( s e e r e f s
) . T he d at a
d e m o n s t rat ed i s a r ep r es e n t a t i v e e x a m p l e o f e x p e r im e n t s p e r fo r m e d a s f o u r b i o l o g i ca l r e p l i c a te s
performed using the VHL inhibitor VH03219. HeLa cells were CM11 mainly induces depletion of pVHL30 and Cul2 but not
pre-treated with VH032 at 150 µM for 30 min before adding other proteins. The proteomics data further evidenced no increase
CM11 into the media. The plates were incubated for further in protein levels of HIF-1α with Homo-PROTACs (but a small
4 h before harvesting. As expected, VH032 blocked pVHL decrease with CM09) relative to DMSO (Supplementary Fig. 6).
degradation (Fig. 6) consistent with the hypothesis that VHL In contrast, treatment with 150 µM VH032 led to increased HIF-
induces degradation of itself. In contrast, pre-treatment with 1α levels, as expected (Supplementary Fig. 6).
IOX4, a PHD2 inhibitor, did not impact CM11 activity (Fig. 6).
To evaluate the specificity of Homo-PROTAC-induced degra-
dation, and identify potential off-targets, we performed isobaric Biophysical evaluation. Key to the catalytic mode of action of
tagging mass spectrometry proteomics to quantify degradation at PROTACs is the formation of a ternary complex36, 38. With
the proteome level in an unbiased fashion. Amongst the 6,450 Homo-PROTACs, VHL is presumed to act as both the E3 ligase
detected proteins that passed filtering criteria, no proteins other and the substrate. Therefore, we next sought to monitor and
than Cul2 were substantially depleted by Homo-PROTACs biophysically characterize the ternary complex VHL:Homo-
CM09, CM10 or CM11 (1 µM for 10 h) compared to DMSO or PROTAC:VHL that is thought to underlie cellular activity. To
VHL inhibitor treatment (Supplementary Data 1). Crucially, no assess the formation of this ternary complex species in solution,
effect on protein levels of other Cullins (Cul1, Cul3, Cul4A, isothermal titration calorimetry (ITC), size-exclusion chromato-
Cul4B, Cul5 and Cul7) or CRL-associated subunits was observed graphy (SEC) and AlphaLISA proximity assays were performed
(Supplementary Fig. 6). Together, the data demonstrate that (Fig. 7). In ITC titration of CM11 against the VCB complex (VHL
6
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D O I : 1 0 . 1 0 3 8 / s 4 1 4 6 7 - 0 1 7 - 0 0 9 5 4 - 1 w ww .n a t u r e . c o m / n a t u re c o m m u n i c a t i o n s
|
|
|

N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
A
R
T
I
C
L
E
CM11
Conc. (μM)
CM11 (1 μM)
1 2 4 8 122436 48 Time (h)
<-- pVHL30
24
4
4
24
Time (h)
24
4
<-- pVHL30
<-- pVHL19
20
15
20
15
<-- pVHL19
100 82 71 112 138 100 88 102
100 71 72 92 104 100 102 100 92 91 112 87 146 127 148 121 126 96
2
11
2
89 106
0
0
15 80 77
100 106 84 127128100 85 26 10
2
1
4
9
2
100 120 91 91 108100 94 129 90 105110 93 81 79
100
75
100
75
Cullin2
Cullin2
100 101 100 130 120 100 86 72 76 105 124 107 118 41 27 98 132 101
100 126 96 81 106 100 80 90 70 96 87 66 55 69
150
100
Hif-1α
Hif-1α
100
100
150
100
150
100
Hif-1α
(longer exp.)
Hif-1α
(longer exp.)
150
100
150
100
Hdy-Hif-1α
Hdy-Hif-1α
150
100
Hdy-Hif-1α
(longer exp.)
Hdy-Hif-1α
(longer exp.)
50
37
50
37
β-actin
β-actin
Fig. 5 C M 1 1 i n d u c es p V H L 3 0 d e p l e t i o n i n a c o n c e n t ra t i o n a n d t i m e - d e p e n de n t f a s h i o n . ( l ef t ) D o s e - r e s p o n se p r o fil e o f H e La c e l l s t r e a te d w i t h i nc r e a s i ng
c o nc e n tra t i o n o f C M 1 1 f o r 4 o r 2 4 h . ( r i g h t ) T im e - c o u rse i m m u n o b l o t s o f l y s a t e s f r o m H e La c e l l s t r e a te d 1 µM o f C M 1 1 u p t o 4 8 h . C o n t r o l t r e a t m e n t s o f
0 . 1 % D M SO , C o C l ( 1 0 0 µM ) , I O X 2 ( 1 5 0 µM ) , a n d V H0 3 2 ( 2 5 0 µM o r 1 µM ) a r e i n c l u d e d . A l l d a t a d e m o n s t r a t e d i s a r e p r e s e n t a ti v e e x a m pl e o f
2
e x p e r i m en t s p e r f o rm e d i n b i o l o gi c a l d u p li c a t e s
Time (h) μM
binding affinity. Consequently, this analysis suggests that we may
be underestimating the binding affinity of CM11, i.e., we can
conclude that Kd is ≤11 nM. This corresponds to an avidity (also
known as cooperativity α) of > 18-fold when compared to
VH032. Such large avidity of homo-bivalent molecules has been
observed previously with other systems, for example the BET
inhibitor MT148. The binding interaction between CM11 and
VHL was driven by a large apparent binding enthalpy (ΔΗ
= –12.3 kcal mol−1), whereas the entropic term was slightly
unfavorable (–TΔS = 1.4 kcal mol−1). This observation underlines
how the thermodynamic signature of CM11 is also very different
when compared with that of VH032, in which case the binding
ΔH was around half that observed with CM11, and both the
enthalpic and entropic term contributed favorably to the ΔG of
binding (Table 1). By contrast, the thermodynamic values
obtained for CMP99 binding were entirely consistent with the
ones of VH032 (Table 1). Specifically, CMP99 bound to VHL in a
1:1 ratio, as expected due to the presence of the cis-Hyp in one of
the two moieties, and it exhibited comparable ΔH and Kd values
to VH032. As anticipated, binding was not detected with CMP98,
the inactive cis-cis epimer. Superposition of integrated heat curves
of CM11, CMP98 and CMP99 is shown in Fig. 7a and visually
highlights the different behaviors of the three compounds.
MLN4924 7.5
3
–
–
–
–
–
+
– – – + + + + – – – – + + + + – – – –
– + + – – + + – – + + – – + + – – – –
– – – – – – – + + + + + + + + + + + +
– – – – – – – – – – – – – – – + + – –
– – – – – – – – – – – – – – – – – + +
+ + + + + + + + + + + + + + + – – – –
MG132
CM11
VH032
IOX4
4.5 50
4
1
4.5 150
4.5 50
7.5
DMSO
pVHL30 -->
pVHL19 -->
20
15
150
Hdy-HIF-1α
Cullin2
100
100
75
50
37
β−actin
V
H
L
Fig. 6 C M 1 1 a c t i v i t y i s C R L 2
a
n
d
p
r
o
t
e
a
s
o
m
e
-
d
e
p
e
n
d
e
n
t
.
H
e
L
a
c
e
l
l
s
t re a t e d w i th C M 1 1 i n t h e a b s e n c e o r p r e s e n c e o f p r o t e a s om e i nh i b i t o r
M G 1 3 2 , n e d d y l a t io n i n h i b it o r M L N4 9 2 4 , V HL i n h i b i t o r V H0 3 2 o r P H D 2
i n hi b i t o r I O X 4 a s n e g a t i v e c o n t r o l . T h e d a t a d e m o n s t r a t e d h e r e i s
r e p re s e nt a t i ve o f o n e b i o l o g ica l r e p l ic a te
with Elongin B and Elongin C)60 the stoichiometry of binding (n CM10 showed similar thermodynamic binding parameters rela-
value) was found to be 0.6, compared to n = 1 with VH032 tive to CM11, with n value equal to 0.7 and a low Kd of 32 nM. A
(Fig. 7a, Table 1). This result is consistent with CM11 binding to stoichiometry close to 1 was instead found for CM09, suggesting
VHL in a 1:2 molar ratio, in contrast to VH032 that binds to VHL that at the end of the titration this system was primarily popu-
in a 1:1 ratio18. Notably, the Kd value measured for CM11 was 11 lated by 1:1 complexes (Supplementary Figs 7 and 8), consistent
nM (Table 1). Closer examination of the titration curve revealed with its lower avidity (Table 1)58, 61. To compare and contrast
that only one point features during the inflection of the curve. Homo-PROTACs binding to the long VHL isoform, we expressed
Indeed, because the protein concentration used in the experiment and purified pVHL30 in complex with Elongin C and Elongin B
was 20 µM, the c value (defined as [P]tot/Kd) calculated for this (V30CB) and performed ITC titrations (Table 1 and Supple-
experiment is 2500, which is above the upper limit of c mentary Fig. 9). The thermodynamic parameters of binding of
(~500–1000) that is a prerequisite for precise measurement of CMP99 to V30CB were comparable to those to V19CB, suggesting
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
7
|
|
|

A
R
T
I
C
L
E
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
a
0.0
b
c
6.0×104
CM09
DMSO
–2.0
–4.0
100
75
50
25
0
VH032
CMP98
CMP99
CM11
CM10
5.0×104
4.0×104
3.0×104
2.0×104
1.0×104
CM11
CMP98
–6.0
–8.0
–10.0
–12.0
12
14
16
18
1
10
100
Elution volume (ml)
[Compound] (nM)
0.0
0.5
1.0
1.5
Molar ratio
Fig. 7 B i o p h y s i c al s t u d i e s o f H o m o - PR OT A C s b i n d i n g t o V HL . a S u p e r p o s i t io n o f t h e i n t e g r a t e d I TC h e a t c u r v e s o f C M 1 1 ( b l u e ) , C M P 9 9 ( b l a c k ) o r C MP 9 8
( g r e e n ) t i t r a t i o n s a g a i n s t V C B . b S EC a s s a y o f c o m p l e x f o r m a t i o n a f t e r i n c u b a t i o n o f C M 1 1 ( r e d ) , C M P 9 8 ( p a l e g r e e n ) , C M P 9 9 ( p u r p l e ) , V H 0 32 ( do t t e d
p u rp l e ) o r D M S O ( b l a c k ) w it h V C B . c A l p h a L I S A : i n t e n s i t y v a l u e s t i t r a t i n g C M 0 9 , C M 1 0 , C M 1 1 a n d C M P 9 8 a g a i n s t V C B . E a c h p o i n t i s m e a n ( S E M)
i n te n si t y o f f o u r t ec h n i c al r e p l ic a t e s
identical binary target engagement in vitro between the two iso- Discussion
forms. CM11 also exhibited large ΔH of binding and n < 1, albeit We describe Homo-PROTACs, a small-molecule approach to
interestingly lower avidity for pVHL30 compared to pVHL19 effectively dimerize an E3 ubiquitin ligase to induce its own self-
(Table 1 and Supplementary Fig. 9).
destruction. Using potent ligands for the E3 ligase VHL, we
SEC experiments showed that V19CB migrates more quickly in developed a series of homo-bivalent molecules that induce
the presence of the active compound CM11 (2:1 protein:ligand remarkably rapid, profound and selective degradation of the long
ratio), relative to vehicle control (Fig. 7b). The shifted peak eluted isoform of pVHL at nanomolar concentrations. Compound-
at a volume corresponding to a species of ~ 90 kDa molecular induced degradation was exquisitely dependent on the linkage
weight, based on a calibration run with globular proteins of pattern on the VHL ligand. The most active Homo-PROTAC,
known molecular weight (see Methods), suggesting the peak CM11, induces complete depletion of pVHL30 after 4 h already at
corresponds to the ternary complex (VCB)2:CM11. Formation of 10 nM. Potent and selective degradation of pVHL30 was long-
a 2:1 complex was also observed with V30CB (Supplementary lasting, with a half-degrading concentration (DC50) of <100 nM,
Fig. 10). Mixing V19CB, V30CB and CM11 resulted in a mixed a remarkable increase in cellular activity of >1000-fold compared
population of the three species (V19CB)2:CM11, (V19CB):CM11: to the parent inhibitor VH032. Mechanistically, we show that
(V30CB) and (V30CB)2:CM11, suggesting that both VHL isoforms CM11 activity is strictly dependent on proteasome activity, Cul2
can form ternary complexes with CM11. No shift in VCB was neddylation, and on VHL binding, and specifically on the for-
observed following incubation with inactive CMP98, CMP99 or mation of an avid 2:1 complex with VHL. Our data therefore
ligand VH032. Only in the sample containing CMP99 a small supports a model in which a highly cooperative ternary complex
peak eluted at 13.5 ml (green curve, Fig. 7b). It is possible that (VHL)2:CM11 functions as the key species responsible for the
such peak could be due to the formation of a lowly populated induced degradation of VHL itself (Fig. 8). Future structural
ternary complex. It is interesting that Schofield and colleagues studies of this ternary species are warranted. Interestingly, CM11
observed weak binding of a cis-hydroxyprolyl containing HIF-1α also led to a decrease in cellular levels of Cullin2, which we
peptide to VHL51. This weak binding, potentially enhanced by hypothesize to be the result of direct ubiquitination and degra-
high avidity in the ternary complex, could be responsible for the dation of Cullin2 as part of the CRL2VHL complex. To our
small decrease of VHL levels observed during biological tests in knowledge, it is unprecedented that a PROTAC can induce the
cells (Fig. 3a). CM10 and CM09 showed formation of a ternary degradation of a protein forming part of the same complex with
complex eluting at identical retention volume when compared to the protein targeted directly.
CM11 (Supplementary Fig. 11). No evidence of aggregation was
seen with any of the compounds evaluated, as all observed peaks short VHL isoform was unanticipated and is an intriguing result
eluted well after the void volume. of this work. This observation adds to recent evidence from us
The preferential induced degradation of pVHL30 over the
Lastly, we employed an AlphaLISA proximity assay to compare and others that chemical degraders designed from inhibitors
ternary complex formation by CM09, CM10 and CM11. The recruiting more than a single protein can add a layer of target
assay showed the highest intensity signal for CM11, whereas degradation selectivity independently of target engagement35, 38,
negligible levels of complex formation were detected for CM09 41. Biophysically, the VHL warhead was found not to distinguish
and CM10 (Fig. 7c). Since SEC detected ternary species with all between the two VHL isoforms at the level of binary target
three compounds, the minimal intensity detected in the engagement. It was also found that CM11 does not considerably
AlphaLISA likely reflects the inability of CM09 and CM10 to distinguish between pVHL19 and pVHL30 within ternary com-
yield a significant ternary population at the low concentrations plexes. We therefore view it unlikely that the remarkable selec-
required for the assay. These results indicate that CM11 is the tivity of VHL degradation is due to differences in molecular
most effective Homo-PROTAC at driving ternary complex recognition. We also consider unlikely that preferential and more
formation, consistent with CM11 exhibiting the highest avidity efficient lysine ubiquitination could play a role, because the extra
and full 2:1 stoichiometry in ITC. Together, the biophysical data region present in the long isoform (1–53) does not contain a
supports CM11 as the most cooperative Homo-PROTAC in vitro, single lysine residue. On the other hand, this region is predicted
and provides a molecular rationale explaining its potent VHL- as intrinsically disordered62, and it has been shown that proteins
degrading activity inside cells.
containing disordered N-terminal regions are more prone to
8
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
|
|
|

N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
A
R
T
I
C
L
E
Table 1 Thermodynamic binding parameters of complex formation of Homo-PROTACs measured by isothermal titration
calorimetry
Protein
Compound
No. of replicates
n
Kd (nM)
α
–
1 8
–
4
6
7 .5
–
ΔG (kcal mol−1
)
ΔH (kcal mol−1
)
−TΔS (kcal mol−1
)
–3 . 6 5 0 .0 2
1 . 4 0 .8
–3 . 1 0 .7
–3 . 5 0 .5
–0 .8 0 .2
–0 .9 0 .1
–4 .3 0 .1
1
8
p
V
H
L
1
9
V H 0 3 2 ( r ef .
C M 1 1
)
1 . 03 0 0 . 0 0 1
0 . 6 0 . 0 1
0 . 9 6 4 0 . 0 0 5
0 . 9 8 0 . 0 9
0 . 7 3 0 . 0 1
1 8 8
1 1
6
–9 . 1 7 0 .0 2
–1 0 . 9 0 .1
–9 . 3 3 0 .0 6
–1 0 . 3 0 .2
–1 0 . 2 0 .1
–1 0 . 4 0 .1
–9 . 5 1 0 .0 5
–5 .5 3 0 .0 1
–1 2 . 3 0 .7
–6 . 2 3 0 .0 5
–6 . 9 0 .3
–9 . 4 0 .1
–1 1 .3 0 .1
2
1
2
2
1
2
a
C
M
P
9
9
1
4
6
2
C M 0 9
C M 1 0
4
1
1
5
3 2
2 5
5
3
a
p
V
H
L
3
0
C
M
1
1
0 . 8 6 6 0 . 0 0 3
1 . 05 0 0 . 0 0 4
a
C
M
P
9
9
1
1
0
6
1
0
–5 .1 9 0 .0 3
A
E
l
r
l
t
i
t
r
r
s
a
t
a
i
r
o
e
n
s
g
w
e
e
r
r
e
a
p
e
d
r
f
b
o
y
r
m
e
e
d
a
O
t
r
2
5
n
°
p
C
r
.
o
E
g
r
r
r
a
o
m
r
v
a
a
n
l
d
u
e
r
s
e fl
r
e
e
p
o
r
t
t
h
e
e
d
a
r
a
e
l
t
h
y
e
o
m
e
h
a
e fi
n
s
1
s
.
e
e
.
m
.
,
u
n
e
l
e
n
s
o
s
o
t
n
h
e
e
a
r
r
w
l
i
e
s
a
e
s
s
p
s
e
q
c
u
ifi
a
e
d
s
.
R
c
a
u
w
I
e
T
a
C
n
d
d
a
t
t
h
a
e
a
e
r
x
e
p
s
e
h
o
m
w
e
n
n
f
o
l
r
e
a
t
c
a
h
.
t
i
t
r
a
t
i
o
n
i
n
F
i
g
.
7
a
a
n
d
S
u
p
p
l
e
m
e
n
t
a
r
y
F
i
g
s
7
–
9
.
a
r
o
e
n
t
e
t
h
i
g
i
c
t
q
u
i
t
f
t
t
b
e
t
w
e
n
t
h
n
l
i
t
-
re
r
v
r
i
t
a
d
a
chemically using the current Hyp-based VHL inhibitors19.
Selective and acute knockdown of pVHL30 by CM11 therefore
offers a new opportunity to address these questions.
VHL
Target
VHL
VHL
In summary, we present CM11, a chemical probe for rapid
and selective pVHL30 knockdown. CM11 provides a chemical
tool alternative to conventional knockdown RNAi approaches
and gene editing knockout technologies such as CRISPR-Cas9.
Relevant information to the use of CM11 will be made available
in the newly established ‘Chemical Probes Portal’ (http://www.
chemicalprobes.org/)72. We anticipate CM11 will find wide use
amongst chemical and cell biologists alike interested in inves-
tigating and dissecting the pleiotropic biological functions of
pVHL. More generally, we provide proof-of-concept that
bivalent molecules can be designed to induce an E3 ligase to
destroy itself. This strategy opens powerful new avenues to
drugging E3 ligases in ways that are not possible with inhibitors
alone.
VHL
Homo-PROTAC
avid 2:1
Ternary complex
Uncomplexed/binary 1:1 complexes
No target degradation
VHL self-degrades
Fig. 8 P r o p o s e d m o d e l f o r t h e m e c h a n i s m o f a c t i o n o f H o m o - P R O T A C
C
M
1
1
proteasomal degradation63. It is also known that VHL is resistant
to proteasomal degradation when in complex with ElonginB and
ElonginC64, so the form observed to be preferentially depleted
Methods
may be free VHL, i.e., not bound to Elongins, or other Synthesis of CM09, CM10, CM11 and controls CMP98 and CMP99. General
proteasome-sensitive forms65. Previous work revealed different
cellular localization patterns of pVHL19 and pVHL30, the former
method A. PEG (1 eq.) was solubilised in dioxane anhydrous and NaH (2 eq.) was
added under stirring. The resulting mixture was stirred at r.t. for 3h. The mixture
was cooled down to 0 °C using ice bath and tert-butylbromo acetate (2 eq.) was
being primarily nuclear, while the latter predominantly cyto-
plasmic56. Crucially, pVHL30 was found to associate to micro-
tubules, a function that is thought to be independent from its
ligase activity66. We therefore speculate that the selective degra-
dation of pVHL30 observed in cells could be the result of
pVHL19 acting preferentially as part of CRL2-VHL, and pVHL30
as the ‘neo-substrate’. Obtaining a more detailed biochemical and
mechanistic understanding of the cellular mode of action of VHL
Homo-PROTACs is of clear importance for future investigation.
Selective degradation of pVHL30 by CM11 led to minimal
stabilization of HIF-α in cells, and as a result did not trigger
HIF-dependent activity. This is consistent with complete knock-
down of all VHL isoforms being required to achieve effective HIF
stabilization in cells, as observed in vhl–/– cells such as
added drop by drop. The resulting mixture was stirred at r.t O/N. The precipitate
was filtered off and the organic phase evaporated to dryness. The resulting oil was
taken up with ethyl acetate, washed with water, dried over MgSO4 and evaporated
to dryness. The resulting oil was purified by column chromatography using a
gradient of ethyl acetate from 50% to 100% v/v in heptane.
General method B. tert-butyl esters 1, 2, 3 or 12 were dissolved in a solution
of 50% v/v trifluroacetic acid in DCM. The resulting solution was stirred for 1 h
or until complete conversion of starting material. The solvent was removed
under high vacuum. The resulting carboxylic acid was used as crude in the next
step without any further purification. To a solution of carboxylic acid in 1 ml
DMF were added HATU (1 eq.) and HOBT (1 eq.) and the pH of the reaction
mixture was adjusted to > 9 by addition of DIPEA (3 eq.). The resulting solution
was stirred at room temperature for 5 min and then amine 7 or 8 was added. The
mixture was stirred at room temperature until no presence of the starting
materials was detected by LC-MS. Water was added and the mixture was
extracted with ethyl acetate (×3). The combined organic phases were washed
with brine (×2), dried over MgSO4 and evaporated under reduced pressure to
VHL-deficient renal carcinoma cells49. Our results underscore the give the corresponding crude, which was purified by HPLC using a gradient of
20% to 95% v/v acetonitrile in 0.1% aqueous solution of ammonia to yield the
potential benefit of using CM11 to interrogate the biological
function of specific VHL isoforms, without the masking down-
desired compound.
stream effects of a hypoxic response. Not much is known about
the individual roles of VHL isoforms. Studies have highlighted
di‐tert‐butyl 3, 6, 9, 12 – tetraoxatetradecanedioate (1)
how the 53-residue extra region of pVHL30 is not needed for
tumor suppression53, and how both isoforms can have HIF-
dependent tumor suppressor functions in vivo67, 68. Other HIF-
independent roles of pVHL have been proposed69, including a
role for pVHL in collagen assembly70. However, the individual
Following general method A, from triethylene glycol (1.125 g, 1 ml, 7.49 mmol,
1 eq.) in 10 ml of dioxane, NaH 60% in mineral oil (595.75 mg, 14.9 mmol, 2 eq.)
and tert-Butyl bromoacetate (2.905 g, 2.19 ml, 14.9 mmol, 2 eq.), compound 1 was
obtained as an oil after high vacuum. Yield: 538 mg, 1.42 mmol (19%). ¹H NMR
roles of the different isoforms in these biological functions remain
elusive. Moreover, many HIF-independent roles are thought to be
independent upon Hyp recognition71, and thus cannot be probed
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
9
|
|
|

A
R
T
I
C
L
E
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
(500 MHz, CDCl3): δ 3.81 (s, 4H), 3.51–3.46 (m, 12H), 1.26 (s,18H). 13C NMR
(126 MHz, CDCl3): δ 169.1, 80.9, 70.1, 70.0, 68.5, 27.5. Analytical data matched
57.7, 56.03, 55.98, 42.1, 35.6, 34.6, 25.4, 14.8. HRMS (ESI) m/z: [M + H]+
calculated for: C56H78N8O13S2: 1134.51; observed: 1135.5538.
those previously reported73
.
N1,N20-bis((S)-1-((2S,4R)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidin-1-yl)-3,3-dimethyl-1-oxobutan-2-yl)-3,6,9,12,15,18-
hexaoxaicosanediamide (CM11)
di‐tert‐butyl 3,6,9,12,15-pentaoxaheptadecanedioate (2)
Following general method A, from tetrathylene glycol (1.125 g, 1 ml, 5.49 mmol,
1 eq.) in 10 ml of dioxane, NaH 60% in mineral oil (463 mg, 11.5 mmol, 2 eq.) and
tert-Butyl bromoacetate (2.25 g, 1.7 ml, 11.5 mmol, 2 eq.), compound 2 was
obtained as an oil after high vacuum. Yield: 500 mg, 1.18 mmol (10%). ¹H NMR
(500 MHz, CDCl3): δ 3.86 (s, 4H), 3.55–3.49 (m, 16H), 1.31 (s, 9H). Analytical data
matched those previously reported73
.
di‐tert‐butyl 3,6,9,12,15,18-hexaoxaicosanedioate (3)
Following general method B, from compound 3 (8.39 mg, 0.018 mmol, 1 eq.),
compound 7 (20 mg, 0.045 mmol, 2.5 eq.), HATU (17 mg, 0.045 mmol, 2.5 eq),
HOAT (6.12, 0.045 mmol, 2.5 mmol), DIPEA (6.98 mg, 0.054 mmol, 3 eq)
compound CM11 was obtained as a white solid. Yield: 11.74 mg, 0.0099 mmol
(55%). ¹H NMR (400 MHz, CDCl3): δ 8.61 (s, 2H), 7.41–7.38 (m, 2H), 7.29
(t, J = 7.6 Hz, 10H), 4.66–4.61 (m, 2H), 4.49–4.41 (m, 6H), 4.35–4.29
(m, 2H), 3.98–3.91 (m, 6H), 3.62–3.50 (m, 24H), 2.45 (s, 6H), 2.42–2.35
(m, 2H), 2.11–2.06 (m, 2H), 0.88 (s, 18H); 13C NMR (101 MHz, CDCl3):
δ 171.2, 170.9, 170.4, 150.3, 148.5, 138.3, 131.6, 130.9, 129.5, 128.1, 71.2,
70.61, 70.59, 70.5, 70.4, 70.3, 58.6, 57.0, 43.2, 36.5, 35.6, 26.4, 16.1. HRMS
(ESI) m/z: [M + H]+ calculated for: C58H82N8O14S2: 1178.54; observed:
1179.6015.
Following general method A, from pentaethylene glycol (1.126 g, 1 ml, 4.72
mmol, 1 eq.) in 10 ml of dioxane, NaH 60% in mineral oil (377 mg, 9.45 mmol, 2
eq.) and tert-Butyl bromoacetate (1.872 g, 1.7 ml, 11.5 mmol, 2 eq.), compound 3
was obtained as an oil after high vacuum. Yield: 300 mg, 0,641 mmol (14%). 1H
NMR (400 MHz, CDCl3): δ 3.94 (s, 4H), 3.66–3.56 (m, 20H), 1.40 (s, 18H).
Analytical data matched those previously reported73
.
N1,N20-bis((S)-1-((2S,4S)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidin-1-yl)-3,3-dimethyl-1-oxobutan-2-yl)-3,6,9,12,15,18-
hexaoxaicosanediamide (CMP98)
N1,N14-bis((S)-1-((2S,4R)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidin-1-yl)-3,3-dimethyl-1-oxobutan-2-yl)-3,6,9,12-tetraoxatetradecanediamide
(CM09)
Following general method B, from compound 3 (7.12 mg, 0.028 mmol, 1 eq.),
compound 8 (18.06, 0.040 mmol, 2.1 eq.), HATU (15.2 mg, 0.040 mmol, 2 eq.),
HOAT (5.44 mg, 0.040 mmol, 2 eq.), DIPEA (7.45 mg, 0.0010 ml, 3 eq.), compound
CMP98 was obtained as a white solid. Yield: 10.58 mg, 0.0089 mmol (45%). ¹H
NMR (400 MHz, CDCl3): δ 9.09 (s, 2H), 8.02 (s, 2H), 7.31 (d, J = 8.5 Hz, 4H), 7.22
(d, J = 8.0 Hz, 4H), 7.16 (d, J = 9.2 Hz, 2H), 4.75–4.64 (m, 4H), 4.51 (d, J = 8.9 Hz,
2H,), 4.41–4.37 (m, 2H), 4.24–4.17 (m, 2H), 3.94 (d, J = 3.2 Hz, 4H), 3.84–3.81
(m, 4H), 3.62–3.54 (m, 20H), 2.49–2.47 (m, 2H), 2.44 (s, 6H), 2.26–2.17 (m, 4H),
0.93 (s, 18H); 13C NMR (101 MHz, CDCl3): δ 173.2, 171.5, 169.7, 151.8, 138.8,
132.9, 129.5, 129.2, 128.3, 71.2, 71.1, 70.6, 70.48, 70.45, 70.4, 70.3, 59.9, 58.5, 56.5,
43.2, 35.6, 35.2, 26.4, 15.0. HRMS (ESI) m/z: [M + H]+ calculated for:
Following general method B, from compound 1 (6.80 mg, 0.018 mmol, 1 eq.),
compound 7 (20 mg, 0.045 mmol, 2.5 eq.), HATU (17 mg, 0.045 mmol, 2.5 eq),
HOAT (6.12, 0.045 mmol, 2.5 mmol) and DIPEA (6.98 mg, 0.054 mmol, 3 eq)
compound CM09 was obtained as a white solid. Yield: 8 mg, 0.007 mmol (40%). ¹H
NMR (400 MHz, CDCl3): δ 8.61 (s, 2H), 7.48–7.45 (m, 2H), 7.31–7.27 (m, 8H),
7.23 (d, J = 10.2 Hz, 2H), 4.64–4.59 (m, 2H), 4.52–4.46 (m, 4H), 4.41–4.38 (m, 2H),
4.31–4.25 (m, 2H), 4.01–3.94 (m, 4H), 3.82 (d, J = 15.7 Hz, 2H), 3.62–3.52 (m,
12H), 2.45 (s, 6H), 2.42–2.34 (m, 2H), 2.12–2.06 (m, 2H), 1.19 (s, 2H), 0.89 (s,
18H); 13C NMR (101 MHz, CDCl3): δ 170.2, 169.9, 169.6, 149.3, 147.5, 137.3,
130.6, 129.9, 128.4, 127.1, 69.9, 69.5, 69.3, 69.1, 57.6, 56.1, 55.9, 42.2, 35.5, 34.6,
25.4, 15.1. HRMS (ESI) m/z: [M + H]+ calculated for: C54H74N8O12S2: 1090.49;
observed: 1091.5093.
C
58H82N8O14S2: 1178.54; observed: 1179.6087.
N1,N17-bis((S)-1-((2S,4R)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidin-1-yl)-3,3-dimethyl-1-oxobutan-2-yl)-3,6,9,12,15-
pentaoxaheptadecanediamide (CM10)
1‐phenyl‐2,5,8,11,14-pentaoxahexadecan-16-ol (9)
Pentaethylene glycol (9.53 g, 50 mmol, 5 eq.) was added dropwise to a
suspension of NaH 60% in mineral oil (800 mg, 20 mmol, 2.5 eq.) in 20 ml of DMF
at 0 °C. The resulting mixture was stirred at r.t for 1 h. The reaction mixture was
cooled to 0oC, benzyl chloride (1 ml, 1.1 g, 8.72 mmol, 1 eq.) was added. The
resulting mixture was stirred O/N at r.t. The reaction was quenched with a
saturated solution of NH4Cl and the aqueous phase was extracted with ethyl acetate
(×3). The combined organic phases were dried over MgSO4 and evaporated to
dryness. The resulting oil was purified by column chromatography (from 0 to 60%
of ethyl acetate in heptane) to afford the title compound as a oil. Yield: 2.055 g,
6.25 mmol (71%). ¹H NMR (400 MHz, CDCl3): δ 7.28–7.19 (m, 5H), 4.50 (s, 2H),
3.66–3.52 (m, 20H), 2.50 (s, 1H). 13C NMR (101 MHz, CDCl3): δ 138.2, 128.3,
127.8, 127.6, 73.2, 72.7, 70.61, 70.58, 70.53, 70.51, 70.2, 69.4, 61.7
Following general method B, from compound 2 (7.60 mg, 0.018 mmol, 1 eq.),
compound 7 (20 mg, 0.045 mmol, 2.5 eq.), HATU (17 mg, 0.045 mmol, 2.5 eq),
HOAT (6.12, 0.045 mmol, 2.5 mmol) and DIPEA (6.98 mg, 0.054 mmol, 3 eq)
compound CM10 was obtained as a white solid. Yield: 6 mg, 0.005 mmol
(30%). ¹H NMR (500 MHz, CDCl3): δ p.p.m., 7.58–7.55 (m, 2H), 7.35–7.30
(m, 10H), 4.72–4.67 (m, 2H), 4.57–4.47 (m, 6H), 4.38–4.33 (m, 2H), 4.15–3.91
(m, 8H), 3.68–3.56 (m, 20H), 2.51 (s, 6H), 2.45–2.38 (m, 2H), 2.20–2.14 (m, 2H),
0.95 (s, 18H). 13C NMR (126 MHz, CDCl3): δ 170.3, 169.8, 169.6, 149.6,
146.9, 137.5, 129.5, 128.47, 128.40, 127.1, 70.0, 69.56, 69.52, 69.5, 69.2, 69.1,
tert-butyl 1-phenyl-2,5,8,11,14,17-hexaoxanonadecan-19-oate (10)
10
N
A
TU
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
|
|
|

N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
A
R
T
I
C
L
E
((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)pyrrolidin-1-yl)-3,3-dimethyl-1-
oxobutan-2-yl)-3,6,9,12,15-pentaoxaheptadecanediamide (CMP99)
To a stirred solution of 9 (2.055 g, 6.25 mmol, 1 eq.) in 12.8 ml of DCM was
added 37% solution of NaOH (12.8 ml), followed by tert-butylbromo acetate
(4.882 g, 25 mmol, 4 eq.) and TBABr (2118 mg, 6.37 mmol, 1.02 eq.). The resulting
solution was stired O/N at r.t. The reaction mixture was extracted with ethyl acetate
(×3). The organic phases were combined and washed with brine (×1), dried over
MgSO4 and concentrate in vacuo. The resulting brow oil was purified by column
chromatography (from 0 to 30% of ethyl acetate in petroleum) to afford the titled
compound as colorless oil. Yield: 2.216 g, 5 mmol (80%). ¹H NMR (500 MHz,
CDCl3): δ 7.28–7.20 (m, 5H), 4.50 (s, 2H), 3.95 (s, 2H), 3.65–3.55 (m, 20H), 1.40
(s, 9H). 13C NMR (126 MHz, CDCl3): δ169.7, 128.4, 127.7, 127.6, 81.5, 73.2, 70.7,
70.7, 70.6, 70.6, 69.4, 69.1, 28.1. MS (ESI) m/z: [(M-tBu) + H]+ calculated for:
Following general method B, from compound 12 (75.6 mg, 0.094 mmol, 1 eq.)
and trifluoroacetic acid (1 ml in 1 ml of DCM), the carboxylic acid derivative was
obtained as oil. The crude was used for the next step without further purification.
Yield: 70 mg, 0.094 mmol (quantitative). MS (ESI) m/z: [M + H]+ calculated for:
C23H38O8: 442.26; observed: 387.20.
C
34H50N4O11S: 722.32; observed: 723.3. Following general method B, from
compound 13 (5.5 mg, 0.006 mmol, 1 eq.), compound 8 (2.77 mg, 0.006 mmol, 1
eq.), HATU (2.28 mg, 0.0.006 mmol, 1 eq.), HOAT (1 mg, 0.0.006 mmol, 1 eq.),
DIPEA (2.23 mg, 0.002 ml, 0.018 mmol, 3 eq.), CMP99 was obtained as a white
solid. Yield: 4.5 mg, 0.004 mmol (66%). ¹H NMR (400 MHz, CDCl3): δ 8.74
(s, 1H), 8.73 (s, 1H), 7.37–7.34 (m, 9H), 7.18 (d, J = 8.9 Hz, 1H), 4.76–4.64 (m, 3H),
4.59–4.44 (m, 5H), 4.37–4.26 (m, 2H), 4.05–3.59 (m, 27H), 2.52 (s, 6H), 2.31–2.11
(m, 4H), 0.96 (s, 9H), 0.95 (s, 9H). 13C NMR (101 MHz, CDCl3): δ 173.0, 171.3,
170.0, 150.7, 145.4, 138.4, 129.53, 129.49, 128.2, 128.16, 71.2, 71.0, 70.54, 70.48,
70.4, 70.3, 58.4, 57.0, 56.7, 43.4, 35.2, 35.0, 26.4, 26.35, 15.9. HRMS (ESI) m/z: [M
+ H]+ calculated for: C56H78N8O13S2: 1134.51; observed: 1135.5814.
19,19-dimethyl-17-oxo-3,6,9,12,15,18-hexaoxaicosanoic acid (11)
10 (2.216 g, 5 mmol, 1 eq.) was dissolved in 75 ml of ethanol, Pd/C (10 wt%)
was added and the resulting mixture was place under hydrogen and stirred at r.t.
until complete conversion of the starting material. The reaction mixture was
filtered through celite, the celite pad was washed few times using ethanol. The
filtrate was concentrated in vacuum to give an oil that was used for the next step
without further purification. Yield: 1764 g, 5 mmol (quantitative). BAIB (3.546 g,
11.01 mmol, 2.2 eq.) and TEMPO (171.87 mg, 1.10 mmol, 0.22 eq.) were added to a
solution of ACN/H2O 1:1 containing previous obtained oil (1.764 g, 5 mmol, 1 eq.).
The resulting mixture was stirred at r.t until complete conversion of the starting
material. The crude was purified using ISOLUTE® PE-AX anion exchange column.
The column was equilibrate with methanol, the reaction mixture poured in the
column and let it adsorbed in the pad. The column was washed with methanol (×3)
to elute all the unbound material. Then, the titled product was eluted using a 50%
solution of formic acid in methanol. The organic phase was evaporated to dryness
to afford the title compound as oil. Yield: 1.200 g, 3.27 mmol (65%). ¹H NMR
(400 MHz, CDCl3): δ 4.12 (s, 2H), 3.98 (s, 2H), 3.72–3.60 (m, 16H), 1.43 (s, 9H).
13C NMR (101 MHz, CDCl3): δ 172.6, 169.7, 81.6, 71.0, 70.59, 70.56, 70.54, 70.46,
70.38, 70.35, 70.30, 68.9, 68.8, 28.1.
Biological and biophysical assays. Cell culture. Human cell lines HeLa, U2OS
and HEK 293, purchased from ATCC, were propagated in DMEM supplemented
with 10% fetal bovine serum (FBS), L- glutamine, 100 μg ml−1 of penicillin/strep-
tomycin at 37 °C and 5% CO2. Cells were maintained for no more than 30 passages.
All cell lines were routinely tested for mycoplasma contamination using MycoAlert
kit from Lonza. For compound treatment experiments, cells were transferred in
six-well plates with either 3 × 105 or 5 × 105 cells per well in 2 ml of media. At 80%
confluence, cells were treated with compounds at the desired concentration,
reaching final DMSO concentration of 0.1% v/v. Cells were incubated at 37 °C and
5% CO2 for the desired time before harvesting.
Small interfering RNA. For siRNA knockdown experiments, 3 × 105 cells were
seeded into each well of a six-well plate in order to achieve 70% of confluence on
the day of transfection. siRNA (SMARTpool: ON-TARGETplus VHL siRNA
L-003936-00-0005) was prepared as a 20 μM solution in RNase-free 1× siRNA
buffer. Negative control siRNA (siRNA from Life Technologies, cat. # 4390843)
was used as negative control. Medium was replaced on the day of transfection.
siRNA solution (5 μL) of both VHL-targeting siRNA and negative control were
added to 250 µl of Opti-mem in 1.5 ml tube. This solution was prepared in
duplicate. Lipofectamine RNAiMax (5 µl) was added to 250 µl of Opti-mem in
another 1.5 ml tube, also prepared in duplicate. The two solutions were combined
and mixed by vortex and incubated at r.t. for 20 min. The whole volume of
transfection mix was added to the six-well plate. Plates were incubated at 37 °C and
5% CO2 for 48 h before harvesting.
tert-butyl (S)-19-((2S,4R)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidine-1-carbonyl)-20,20-dimethyl-17-oxo-3,6,9,12,15-pentaoxa-18-
azahenicosanoate (12)
ML4924 and MG132 co-treatment. Cells were transferred in six-well plates with
5 × 105 cells per well in 2 ml media in order to achieve 80% confluence the day
after. At t = 0, MLN4924 was added into the desired wells at 3 μM final
concentration and 0.1% v/v of DMSO. DMSO (0.1% v/v final conc.) was added to
the remaining wells in order to match identical conc. of vehicle in all wells. At t = 3
h, MG 132 was added into the desired wells at 50 μM final conc. and 0.1% v/v of
DMSO. DMSO (0.1% v/v final conc.) was added to the remaining wells in order to
achieve the same conc. of vehicle in all the wells. At t = 3.5 h, the desired wells were
treated with 1 μM of CM11 in 0.1% v/v DMSO final concentration. DMSO (0.1% v/
v final conc.) was added to the remaining wells to obtain the same conc. of vehicle
in all the wells. The total final concentration of DMSO was therefore 0.3% v/v.
Plates were incubated for 4 h at 37 °C and 5% CO2 before harvesting. For
competition experiments with VH032, cells were treated with VH032 at a final
concentration of 150 μM (or IOX4 at 50 μM) for 30 min before treatment with
CM11 at 1 μM final concentration for 4 h. Plates were incubated for the desired
time at 37 °C and 5% CO2 before harvesting.
Immunoblotting. Cells were lysed in lysis buffer (20 mM Tris pH 8, 150 mM
NaCl, 1% Triton × 100) and a protease inhibitor cocktail (Roche) per 10 ml buffer.
For protein extracts, the dishes were placed on ice. The media was aspirated and
the tissue layer washed twice with ice-cold phosphate buffer saline (PBS). Lysis
buffer (120 μl) was added and the cells detached from the surface with a cell
scraper. After removal of the insoluble fraction by centrifugation, the protein
concentration of the supernatant was determined by Pierce™ Coomassie (Bradford)
Protein Assay Kit. Protein extracts were fractionated by SDS-PAGE on 4–12%
Tris-Acetate NuPage® Novex® (Life Technologies) polyacrylamide gels and
transferred to a nitrocellulose membrane using wet transfer. The membrane was
then blocked with 5 % w/v Bovine serum albumin (BSA) in Tris-buffered saline
(TBS) with 0.1% w/v Tween-20. For detecting proteins the following primary
To a solution of PEG linker 11 (78.8 mg, 0.215 mmol, 1 eq.) in 1.5 ml DMF was
added HATU (81.74 mg, 0.215 mmol, 1 eq.), HOAT (29.26 mg, 0.215 mmol, 1 eq.),
DIPEA (80.13 mg, 0.106 ml, 0.645 mmol, 3 eq.) and the solution was stirred at
room temperature for 5 min. Compound 7 (100 mg, 0.215 mmol, 1 eq.) was added
and the pH of the reaction mixture was adjusted to > 9 by addition of DIPEA
(80.13 mg, 0.106 ml, 0.645 mmol, 3 eq.). The mixture was stirred at room
temperature until no presence of the starting materials was detected by LC-MS.
The solvent was evaporated under reduced pressure to give the corresponding
crude, which was purified by HPLC using a gradient of 20–95% v/v acetonitrile in
0.1% aqueous solution of ammonia to yield the titled compound as white solid.
Yield: 75.6 mg, 0.094 mmol (44%). ¹H NMR (400 MHz, CDCl3): δ 9.00 (s, 1H),
7.45 (t, J = 5.9 Hz, 1H), 7.39–7.33 (m, 4H), 7.29 (d, J = 8.9 Hz, 1H), 4.71 (t, J = 8.0
Hz, 1H), 4.59–4.48 (m, 3H), 4.34 (dd, J = 5.2, 14.6 Hz, 1H), 4.08–3.92 (m, 5H),
3.69–3.61 (m, 18H), 2.52 (s, 3H), 2.47–2.41 (m, 1H), 2.19–2.11 (m, 1H), 1.46 (s,
9H), 0.97 (s, 9H). 13C NMR (101 MHz, CDCl3): δ 171.3, 171.1, 170.5, 170.0, 151.7,
139.1, 129.4, 128.3, 82.0, 71.1, 70.6, 70.4, 70.4, 70.3, 70.3, 70.2, 70.2, 68.9, 58.7, 57.3,
56.8, 43.1, 36.3, 35.1, 28.1, 26.4, 15.1. MS (ESI) m/z: [M + H]+ calculated for:
C38H58N4O11S2: 778.38; observed: 779.4.
N1-((R)-1-((2R,4R)-4-hydroxy-2-((4-(4-methylthiazol-5-yl)benzyl)carbamoyl)
pyrrolidin-1-yl)-3,3-dimethyl-1-oxobutan-2-yl)-N17-((S)-1-((2S,4R)-4-hydroxy-2-
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
11
|
|
|

A
R
T
I
C
L
E
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
antibodies in the given concentrations were used: anti-β-Actin (Cell Signaling
Technology, 4970S, 13E5) 1:2000, anti-VHL (Cell Signaling Technology, #68547)
1:1000, anti-Hif-1α (BD Biosciences, 610959, clone 54) 1:1,000, anti-hydroxy-HIF-
1α (Hyp564) (Cell Signaling Techonology; #3434) 1:1,000, anti-PHD2 (Bethyl
Laboratories; A300-322A) 1:1,000, anti-PHD3 (Bethyl Laboratories; A300-327A)
1:1000, anti-CRBN (Proteintech; 11435-1-AP) 1:1,000. Following incubation with a
horseradish peroxidase-conjugated secondary antibody (Cell Signaling
Technology), the signal was developed using enhanced chemiluminescence (ECL)
Western Blotting Detection Kit (Amersham) on Amersham Hyperfilm ECL film
(Amersham). Band quantification was performed using ImageJ software and
reported as relative amount as ratio of the each protein band relative to the lane’s
loading control. The values obtained were then normalized to 0.1% DMSO vehicle
control. See Supplementary Fig. 12 for full-scanned images of western-blots shown
in the main text figures.
Escherichia coli BL21(DE3) at 20 °C for 18 h using 0.5 mM isopropyl β-D-1-
thiogalactopyranoside (IPTG). E. coli cells were resuspended in binding buffer
(50 mM HEPES, pH 8.0, 500 mM NaCl, 10 mM imidazole and 5 mM β-
mercaptoethanol), lysed using a pressure cell homogenizer (Stansted Fluid Power),
and lysate clarified by centrifugation. The supernatant was loaded onto a HisTrap
FF affinity column (GE Healthcare) equilibrated with binding buffer, and bound
protein was eluted using an imidazole gradient. The His6 tag was removed using
thrombin protease (Sigma), and the untagged complex was dialyzed into binding
buffer. V30CB was then loaded onto the HisTrap FF column a second time,
allowing impurities to bind as the complex eluted without binding. The
flow-through was dialyzed using a low-salt buffer (20 mM HEPES, pH 8.0, 25 mM
NaCl and 1 mM DTT) and then loaded onto a Resource Q anion exchange column
(GE Healthcare); bound protein was eluted across a linear NaCl gradient. V30CB
was then additionally purified by size-exclusion chromatography using a Superdex-
75 16/600 column (GE Healthcare) equilibrated with 20 mM HEPES, pH 7.5,
100 mM sodium chloride and 1 mM TCEP. All chromatography purification steps
were performed using Äkta FPLC purification systems (GE Healthcare).
Isothermal titration calorimetry (ITC). Titrations were performed on an ITC200
micro-calorimeter (GE Healthcare). Homo-PROTACs (CM11, CMP98 or CMP99)
were diluted from a 100 mM DMSO stock solution to 150 µM in a buffer
containing 20 mM Bis-tris propane, 150 mM NaCl, 1 mM tris(2-carboxyethyl)
phosphine (TCEP), pH 7.4. The final DMSO concentration was 0.15% v/v. VCB
protein experiments were carried out in a buffer containing 20 mM Bis-tris
propane, 150 mM NaCl, 1 mM TCEP, 0.15% v/v DMSO, pH 7.4. The titrations
consisted of 19 injections of 2 µL compounds solution (150 µM, in the syringe) at a
rate of 0.5 µl s−1 at 120 s time intervals into the VCB protein solution (20 µM, in the
cell). An initial injection of compound solution (0.4 µL) was made and discarded
during data analysis. All experiments were performed at 25 °C, whilst stirring the
syringe at 600 rpm. The data were fitted to a single binding site model to obtain the
stoichiometry n, the dissociation constant Kd and the enthalpy of binding ΔH using
the Microcal LLC ITC200 Origin software provided by the manufacturer.
Size-exclusion chromatography (SEC). SEC experiments were carried out in a
ÄKTA pure system (GE Healthcare) at room temperature. The oligomeric state of
the VCB complex in solution was analyzed by gel filtration in a buffer containing
20 mM Bis-Tris (pH 7), 150 mM NaCl and 1 mM 1,4-dithiothreitol (DTT) using a
Superdex 200 Increase 10/300 GL column (GE Healthcare) calibrated with globular
proteins of known molecular weight (GE Healthcare, 28-4038-41/42). VCB protein
(50 µM) was incubated with CM11 (30 µM), CMP98 (30 µM), CMP99 (30 µM),
VH032 (30 µM) or DMSO (0.5 %) for 20 min at room temperature prior to
injection. Sample volume for each injection was 200 µl, and the flow rate was 0.5
ml min−1. Peak elution was monitored using ultraviolet absorbance at 280 nm.
Biotinylation of VCB. The VCB complex was mixed with EZ-link NHS-PEG4-
biotin (Thermo Scientific) in a 1:1 molar ratio and incubated at room temperature
for 1 h. The reaction was quenched using 1 M Tris-HCl, pH 7.5, and unreacted
NHS-biotin was removed with a PD-10 MiniTrap desalting column (GE Healthcare)
equilibrated with 20 mM HEPES, pH 7.5, 150 mM NaCl and 1 mM DTT.
AlphaLISA assay. All assays were performed at room temperature in 384-well
plates with a final assay volume of 25 µL per well; plates were sealed with
transparent film between addition of reagents. All reagents were prepared as
5× stocks diluted in 50 mM HEPES, pH 7.5, 100 mM NaCl, 0.1% (w/v) bovine
serum albumin and 0.02% (w/v) 3-[(cholamidopropyl)dimethylammonio]-1-
propanesulfonate (CHAPS). Biotinylated VCB (20 nM final) and His6-VCB (20 nM
final) were incubated with a range of Homo-PROTAC concentrations
Luciferase assay. Assay was performed as described19. Briefly, cells (HeLa and
U2OS) stably expressing an HRE-luciferase reporter were treated for the indicated
times with compounds. Cells were harvested in passive lysis buffer (Promega) and
subjected to three freeze-thaw cycles. The soluble lysate fraction was used for
assays, performed according to the manufacturer’s instructions (Promega) using a
Berthold Lumat LB 9507 Luminometer. Results were normalized for protein
concentration, and reported as mean s.e.m. from three biological replicates.
Quantitative real-time PCR. Assay was performed as described19. Briefly, RNA
was extracted from HeLa cell lysates using the RNeasy Mini Kit (Qiagen) and
reverse transcribed using the iScript cDNA Synthesis kit (Bio-Rad). Real-time PCR
was performed using PerfeCTa SYBR Green FastMix (Quanta Biosciences) in
C1000 Touch Thermal Cycler (Bio-Rad). mRNA levels were calculated based on
averaged Ct values from two technical replicates, normalized to mRNA levels of β-
actin, and reported as mean s.e.m. from three biological replicates.
MS proteomics. Sample preparation. HeLa cells were seeded at 5 × 106 on a
10 cm plate 16 h before treatment. Cells were treated with 0.1% DMSO as vehicle
control, VH032 (150 μM) to control for VHL inhibition, and CM09, CM10 or
CM11 (1 μM), reaching a final DMSO concentration of 0.1% v/v. Cells were
incubated at 37 °C and 5% CO2 for 10 h before harvesting. Cells were placed on ice
and medium was removed and cells were washed twice with cold PBS. Cells were
lysed in 0.5 ml of 100 mM Tris pH 8.0, 4% (w/v) SDS supplemented with protease
inhibitor cocktail (Roche). The lysate was pulse sonicated briefly and then
centrifuged at 17,000 × g for 15 min at 4 °C. The supernatant fraction of cell extract
was snap-frozen in liquid nitrogen and stored in −80 °C freezer. Further sample
processing, digestion, desalting, labeling and fractionation were performed as
previously described38
.
nLC-MS/MS analysis. The TMT 10plex fractions were first analyzed as
previously described38. Briefly, the TMT 10plex fractions were analyzed
sequentially on a Q Exactive HF Hybrid Quadrupole-Orbitrap Mass Spectrometer
(Thermo Scientific) coupled to an UltiMate 3000 RSLCnano UHPLC system
(Thermo Scientific) and EasySpray column (75 μm × 50 cm, PepMap RSLC C18
column, 2 μm, 100 Å, Thermo Scientific). The MS2 isolation window was set to
0.4 Da and the resolution set at 120,000. MS2 resolution was set at 60,000. The
AGC targets for MS1 and MS2 were set at 3e6 ions and 1e5 ions, respectively. The
normalized collision energy was set at 32%. The maximum ion injection times for
MS1 and MS2 were set at 50 and 200 ms, respectively. The dynamic exclusion time
was set to 45 s. After the first analysis, the TMT labeled fractions were re-analyzed
on the Q-Exactive HF using the same instrument settings and a target list generated
from a theoretical trypsin digest of VHL. In this case only precursor ions that
matched peptide ions from the target list were selected for ms/ms. The dynamic
exclusion time was set to 30 s. After the second analysis, two fractions were further
analyzed using a targeted SIM/ddMS2 method on the Q-Exactive HF. The targeted
list for this run was generated from previous theoretical TMT labeled VHL protein
digest. The methods were as follows, one SIM scan is followed by five targeted
dd-MS2 scans. The SIM and MS2 isolation windows were set to 6 Da and 0.4 Da,
respectively and the SIM scan and MS2 resolution set at 120,000 and 60,000
respectively. The AGC targets for MS1 and MS2 were set at 1e5 ions and 2e5 ions,
respectively. The normalized collision energy was set at 32%. The maximum ion
injection times for MS1 and MS2 were set at 200 and 300 ms, respectively. The
dynamic exclusion time was set to 3 s.
Peptide and protein identification. All the raw MS data files from the three
analyses were merged and searched against the Uniprot-sprot-Trembl-Human-
Canonical database by Maxquant software 1.5.3.30 for protein identification and
TMT reporter ion quantitation. The Maxquant parameters were set as follows:
enzyme used Trypsin/P; maximum number of missed cleavages equal to two;
precursor mass tolerance equal to 10 p.p.m.; fragment mass tolerance equal to 20 p.
p.m.; variable modifications: Oxidation (M), Acetyl (N-term); fixed modifications:
Carbamidomethyl (C). The data was filtered by applying a 1% false discovery rate
followed by exclusion of proteins with less than two unique peptides. Quantified
proteins were filtered if the absolute fold-change difference between the two DMSO
replicates was ≥ 1.5.
(0.5–200 nM; three-in-five serial dilution) for 1 h. Anti-His acceptor beads
(PerkinElmer, 10 µg ml−1 final) were added and plates were incubated for another
hour. Streptavidin-coated donor beads (PerkinElmer, 10 µg ml−1 final) were added
and plates were incubated for a final 1 h. Plates were read on a PHERAstar FS
(BMG Labtech) using an optic module with an excitation wavelength of 680 nm
and emission wavelength of 615 nm. Intensity values were plotted against
PROTAC concentration on a log10 scale.
Data availability. Data supporting the findings of this study are available within
the article (and its Supplementary Information files) and from the corresponding
author upon reasonable request. The mass spectrometry proteomics data have been
deposited to the ProteomeXchange Consortium via the PRIDE partner repository
with the dataset identifier PXD007594.
Received: 17 April 2017 Accepted: 1 August 2017
References
1. Skaar, J. R., Pagan, J. K. & Pagano, M. SCF ubiquitin ligase-targeted therapies.
Nat. Rev. Drug Discov. 13, 889–903 (2014).
Protein expression and purification. Human proteins VHL (UniProt accession
number: P40337), ElonginC (Q15369), and ElonginB (Q15370) were used for all
protein expression. The short-isoform-containing VCB complex (V19CB) was
expressed, purified and stored as described previously74. For expression of the
long-isoform-containing VCB complex (V30CB), N-terminally His6-tagged VHL
(1–213), ElonginC (17–112) and ElonginB (1–104) were co-expressed in
2. Bulatov, E. & Ciulli, A. Targeting Cullin-RING E3 ubiquitin ligases for drug
discovery: structure, assembly and small-molecule modulation. Biochem. J. 467,
365–386 (2015).
3. Galdeano, C. Drugging the undruggable: targeting challenging E3 ligases for
personalized medicine. Future Med. Chem. 9, 347–350 (2017).
12
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
|
|
|

N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
7
-
0
0
9
5
4
-
1
A
R
T
I
C
L
E
4. Morreale, F. E. & Walden, H. Types of ubiquitin ligases. Cell 165, 248–248.e1
(2016).
36. Bondeson, D. P. et al. Catalytic in vivo protein knockdown by small-molecule
PROTACs. Nat. Chem. Biol. 11, 611–617 (2015).
5. Zheng, N. & Shabek, N. Ubiquitin ligases: structure, function, and regulation.
Annu. Rev. Biochem. 86, 129–157 (2017).
6. Komander, D. & Rape, M. The ubiquitin code. Annu. Rev. Biochem. 81,
203–229 (2012).
7. Pickart, C. M. Mechanisms underlying ubiquitination. Annu. Rev. Biochem. 70,
503–533 (2001).
8. Petroski, M. D. & Deshaies, R. J. Mechanism of lysine 48-linked ubiquitin-chain
synthesis by the cullin-RING ubiquitin-ligase complex SCF-Cdc34. Cell 123,
1107–1120 (2005).
37. Raina, K. et al. PROTAC-induced BET protein degradation as a therapy for
castration-resistant prostate cancer. Proc. Natl Acad. Sci. USA 113, 7124–7129
(2016).
38. Gadd, M. S. et al. Structural basis of PROTAC cooperative recognition for
selective protein degradation. Nat. Chem. Biol. 13, 514–521 (2017).
39. Winter, G. E. et al. Phthalimide conjugation as a strategy for in vivo target
protein degradation. Science 348, 1376–1381 (2015).
40. Lu, J. et al. Hijacking the E3 ubiquitin ligase cereblon to efficiently Target
BRD4. Chem. Biol. 22, 755–763 (2015).
9. Nalepa, G., Rolfe, M. & Harper, J. W. Drug discovery in the ubiquitin-
proteasome system. Nat. Rev. Drug Discov. 5, 596–613 (2006).
10. Arkin, M. R., Tang, Y. & Wells, J. A. Small-molecule inhibitors of protein-
protein interactions: progressing toward the reality. Chem. Biol. 21, 1102–1114
(2014).
41. Lai, A. C. et al. Modular PROTAC design for the degradation of oncogenic
BCR-ABL. Angew Chem. Int. Ed. Engl. 55, 807–810 (2016).
42. Lebraud, H., Wright, D. J., Johnson, C. N. & Heightman, T. D. Protein
degradation by in-cell self-assembly of proteolysis targeting chimeras. ACS
Cent. Sci. 2, 927–934 (2016).
11. Lucas, X. & Ciulli, A. Recognition of substrate degrons by E3 ubiquitin ligases
and modulation by small-molecule mimicry strategies. Curr. Opin. Struct. Biol.
44, 101–110 (2017).
12. Vassilev, L. T. et al. In vivo activation of the p53 pathway by small-molecule
antagonists of MDM2. Science 303, 844–848 (2004).
43. Itoh, Y., Ishikawa, M., Naito, M. & Hashimoto, Y. Protein knockdown using
methyl bestatin-ligand hybrid molecules: design and synthesis of inducers of
ubiquitination-mediated degradation of cellular retinoic acid-binding proteins.
J. Am. Chem. Soc. 132, 5820–5826 (2010).
44. Ohoka, N. et al. In vivo knockdown of pathogenic proteins via specific and
nongenetic inhibitor of apoptosis protein (IAP)-dependent protein erasers
(SNIPERs). J. Biol. Chem. 292, 4556–4570 (2017).
13. Vu, B. et al. Discovery of RG7112: a small-molecule MDM2 inhibitor in clinical
development. ACS Med. Chem. Lett. 4, 466–469 (2013).
14. Oost, T. K. et al. Discovery of potent antagonists of the antiapoptotic protein
XIAP for the treatment of cancer. J. Med. Chem. 47, 4417–4426 (2004).
15. Sun, H. et al. Design, synthesis, and characterization of a potent, nonpeptide,
cell-permeable, bivalent Smac mimetic that concurrently targets both the BIR2
and BIR3 domains in XIAP. J. Am. Chem. Soc. 129, 15279–15294 (2007).
16. Chessari, G. et al. Fragment-based drug discovery targeting inhibitor of
apoptosis proteins: discovery of a non-alanine lead series with dual activity
against cIAP1 and XIAP. J. Med. Chem. 58, 6574–6588 (2015).
17. Buckley, D. L. et al. Targeting the von Hippel-Lindau E3 ubiquitin ligase using
small molecules to disrupt the VHL/HIF-1α interaction. J. Am. Chem. Soc. 134,
4465–4468 (2012).
18. Galdeano, C. et al. Structure-guided design and optimization of small molecules
targeting the protein-protein interaction between the von Hippel-Lindau
(VHL) E3 ubiquitin ligase and the hypoxia inducible factor (HIF) alpha subunit
with in vitro nanomolar affinities. J. Med. Chem. 57, 8657–8663 (2014).
19. Frost, J. et al. Potent and selective chemical probe of hypoxic signalling
downstream of HIF-α hydroxylation via VHL inhibition. Nat. Commun. 7,
13312 (2016).
20. Davies, T. G. et al. Monoacidic inhibitors of the Kelch-like ECH-associated
protein 1: nuclear factor erythroid 2-related factor 2 (KEAP1:NRF2) protein-
protein interaction with high cell potency identified by fragment-based
discovery. J. Med. Chem. 59, 3991–4006 (2016).
21. Jiang, Z.-Y., Lu, M.-C. & You, Q.-D. Discovery and development of Kelch-like
ECH-associated protein 1. Nuclear factor erythroid 2-related factor 2 (KEAP1:
NRF2) protein-protein interaction inhibitors: achievements, challenges, and
future directions. J. Med. Chem. 59, 10837–10858 (2016).
45. Durham, T. B. & Blanco, M.-J. Target engagement in lead generation. Bioorg.
Med. Chem. Lett. 25, 998–1008 (2015).
46. Spencer, D. M., Wandless, T. J., Schreiber, S. L. & Crabtree, G. R.
Controlling signal transduction with synthetic ligands. Science 262, 1019–1024
(1993).
47. Corson, T. W., Aberle, N. & Crews, C. M. Design and applications of
bifunctional small molecules: why two heads are better than one. ACS. Chem.
Biol. 3, 677–692 (2008).
48. Tanaka, M. et al. Design and characterization of bivalent BET inhibitors. Nat.
Chem. Biol. 12, 1089–1096 (2016).
49. Maxwell, P. H. et al. The tumour suppressor protein VHL targets hypoxia-
inducible factors for oxygen-dependent proteolysis. Nature. 399, 271–275
(1999).
50. Buckley, D. L. et al. HaloPROTACS: use of small molecule PROTACs to induce
degradation of HaloTag fusion proteins. ACS. Chem. Biol. 10, 1831–1837
(2015).
51. Loenarz, C. et al. Evidence for a stereoelectronic effect in human oxygen
sensing. Angew Chem. Int. Ed. Engl. 48, 1784–1787 (2009).
52. Bangiyeva, V. et al. Differences in regulation of tight junctions and cell
morphology between VHL mutations from disease subtypes. BMC Cancer. 9,
229–18 (2009).
53. Schoenfeld, A., Davidowitz, E. J. & Burk, R. D. A second major native von
Hippel-Lindau gene product, initiated from an internal translation start site,
functions as a tumor suppressor. Proc. Natl Acad. Sci. USA 95, 8817–8822
(1998).
54. Robinson, C. M. & Ohh, M. The multifaceted von Hippel-Lindau tumour
suppressor protein. FEBS Lett. 588, 2704–2711 (2014).
55. Kamura, T. et al. Rbx1, a component of the VHL tumor suppressor complex
and SCF ubiquitin ligase. Science 284, 657–661 (1999).
56. Barry, R. E. & Krek, W. The von Hippel-Lindau tumour suppressor: a
multi-faceted inhibitor of tumourigenesis. Trends Mol. Med. 10, 466–472
(2004).
57. Rapisarda, A. et al. Identification of small molecule inhibitors of hypoxia-
inducible factor 1 transcriptional activation pathway. Cancer Res. 62,
4316–4324 (2002).
22. Swinney, D. C. Biochemical mechanisms of drug action: what does it take for
success? Nat. Rev. Drug. Discov. 3, 801–808 (2004).
23. Huang, X. & Dixit, V. M. Drugging the undruggables: exploring the ubiquitin
system for drug development. Cell Res. 26, 484–498 (2016).
24. Lai, A. C. & Crews, C. M. Induced protein degradation: an emerging drug
discovery paradigm. Nat. Rev. Drug Discov. 16, 101–114 (2017).
25. Tan, X. et al. Mechanism of auxin perception by the TIR1 ubiquitin ligase.
Nature 446, 640–645 (2007).
26. Ito, T. et al. Identification of a primary target of thalidomide teratogenicity.
Science 327, 1345–1350 (2010).
27. Fischer, E. S. et al. Structure of the DDB1-CRBN E3 ubiquitin ligase in complex
with thalidomide. Nature 512, 49–53 (2014).
58. Douglass, E. F., Miller, C. J., Sparer, G., Shapiro, H. & Spiegel, D. A. A
comprehensive mathematical model for three-body binding equilibria. J. Am.
Chem. Soc. 135, 6092–6099 (2013).
59. Soucy, T. A. et al. An inhibitor of NEDD8-activating enzyme as a new approach
to treat cancer. Nature 458, 732–736 (2009).
60. Stebbins, C. E., Kaelin, W. G. & Pavletich, N. P. Structure of the VHL-
ElonginC-ElonginB complex: implications for VHL tumor suppressor function.
Science 284, 455–461 (1999).
61. Mack, E. T., Perez-Castillejos, R., Suo, Z. & Whitesides, G. M. Exact analysis of
ligand-induced dimerization of monomeric receptors. Anal. Chem. 80,
5550–5555 (2008).
62. Minervini, G. et al. Isoform-specific interactions of the von Hippel-Lindau
tumor suppressor protein. Sci. Rep. 5, 12605 (2015).
63. Schrader, E. K., Harstad, K. G. & Matouschek, A. Targeting proteins for
degradation. Nat. Chem. Biol. 5, 815–822 (2009).
64. Schoenfeld, A. R., Davidowitz, E. J. & Burk, R. D. Elongin BC complex prevents
degradation of von Hippel-Lindau tumor suppressor gene products. Proc. Natl
Acad. Sci. USA 97, 8507–8512 (2000).
65. Feldman, D. E., Thulasiraman, V., Ferreyra, R. G. & Frydman, J. Formation of
the VHL-elongin BC tumor suppressor complex is mediated by the chaperonin
TRiC. Mol. Cell 4, 1051–1061 (1999).
28. Lu, G. et al. The myeloma drug lenalidomide promotes the cereblon-dependent
destruction of Ikaros proteins. Science 343, 305–309 (2014).
29. Krönke, J. et al. Lenalidomide causes selective degradation of IKZF1 and IKZF3
in multiple myeloma cells. Science 343, 301–305 (2014).
30. Petzold, G., Fischer, E. S. & Thomä, N. H. Structural basis of lenalidomide-
induced CK1α degradation by the CRL4(CRBN) ubiquitin ligase. Nature 532,
127–130 (2016).
31. Matyskiela, M. E. et al. A novel cereblon modulator recruits GSPT1 to the CRL4
(CRBN) ubiquitin ligase. Nature 535, 252–257 (2016).
32. Han, T. et al. Anticancer sulfonamides target splicing by inducing
RBM39 degradation via recruitment to DCAF15. Science 356, eaal3755 (2017).
33 Uehara, T. et al. Selective degradation of splicing factor CAPERα by anticancer
sulfonamides. Nat. Chem. Biol. 13, 675–680 (2017).
34. Toure, M. & Crews, C. M. Small-molecule PROTACS: new approaches to
protein degradation. Angew Chem. Int. Ed. Engl. 55, 1966–1973 (2016).
35. Zengerle, M., Chan, K.-H. & Ciulli, A. Selective small molecule induced
degradation of the BET bromodomain protein BRD4. ACS. Chem. Biol. 10,
1770–1777 (2015).
N
A
T
U
R
E
C
OM
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
46
7
-
0
1
7
-
0
0
9
5
4
-
1
w
w
w
.n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
13
|
|
|

A
R
T
I
C
L
E
N A T U R E C O M MU N ICA T IO N S | D O I: 1 0 . 1 0 3 8 / s 4 1 46 7 - 0 17 - 0 0 9 5 4 -1
66. Hergovich, A., Lisztwan, J., Barry, R., Ballschmieter, P. & Krek, W. Regulation Author contributions
of microtubule stability by the von Hippel-Lindau tumour suppressor protein
pVHL. Nat. Cell. Biol. 5, 64–70 (2003).
67. Kaelin, W. G. The von Hippel-Lindau tumour suppressor protein: O2 sensing
and cancer. Nat. Rev. Cancer 8, 865–873 (2008).
68. Gossage, L., Eisen, T. & Maher, E. R. VHL, the story of a tumour suppressor
gene. Nat. Rev. Cancer 15, 55–64 (2015).
A.C. supervised the project. C.M., S.J.H., A.T. and A.C. designed experiments. C.M.
performed chemical synthesis, biological evaluation and ITC experiments. S.J.H.
expressed and purified proteins, and performed SEC and AlphaLISA experiments; W.C.
and D.J.L. performed MS proteomics experiments. C.M., S.J.H., A.T., S.R., D.R.A. and
A.C. analyzed the data. S.R. and D.R.A. contributed reagents and tools; C.M., S.J.H. and
A.C. wrote the manuscript with input from all authors.
69. Li, M. & Kim, W. Y. Two sides to every story: the HIF-dependent and
HIF-independent functions of pVHL. J. Cell. Mol. Med. 15, 187–195 (2011).
70. Grosfeld, A. et al. Interaction of hydroxylated collagen IV with the von
hippel-lindau tumor suppressor. J. Biol. Chem. 282, 13264–13269 (2007).
71. Frew, I. J. & Krek, W. Multitasking by pVHL in tumour suppression. Curr.
Opin. Cell Biol. 19, 685–690 (2007).
Additional information
Supplementary Information accompanies this paper at doi:10.1038/s41467-017-00954-1.
Competing interests: The authors declare no competing financial interests.
72. Arrowsmith, C. H. et al. The promise and peril of chemical probes. Nat. Chem.
Biol. 11, 536–541 (2015).
73. Wittmann, V., Takayama, S. & Gong, K. W. Ligand recognition by E-and
P-selectin: chemoenzymatic synthesis and inhibitory activity of bivalent sialyl
Lewis x derivatives and sialyl Lewis x carboxylic acids. J. Org. Chem. 63,
5137–5143 (1998).
Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
74. Van Molle, I. et al. Dissecting fragment-based lead discovery at the von
Hippel-Lindau protein:hypoxia inducible factor 1α protein-protein interface.
Chem. Biol. 19, 1300–1312 (2012).
75 Uehara, T. et al. Selective degradation of splicing factor CAPERα by anticancer
sulfonamides. Nat. Chem. Biol. 13, 675–680 (2017).
Open Access This article is licensed under
a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
Acknowledgements
Financial support for this work was provided by the European Research Council
(ERC-2012-StG-311460 DrugE3CRLs Starting Grant to A.C.), the Italian Ministry of
Education, University and Research (Miur, PhD Studentship to C.M.), the Medical
Research Council (MRC, grant number MC_UU_12016/2 to D.R.A.), and the Wellcome
Trust (Strategic Award 100476/Z/12/Z for biophysics and drug discovery to the Division
of Biological Chemistry and Drug Discovery). S.R. is funded by a CRUK Senior Fellowship
(C99667/A12918). We thank Hannah Tovell and Dr Athanasios Karapetsas (MRC-PPU)
for assistance with functional assays and Julianty Frost (BCDD/GRE) for discussion.
© The Author(s) 2017
14
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
8
:
8
3
0
D O I : 1 0 . 1 0 3 8 / s 4 1 4 6 7 - 0 1 7 - 0 0 9 5 4 - 1 w ww .n a t u r e . c o m / n a t u re c o m m u n i c a t i o n s
|
|
|