Content uploaded by Yan Chuangye
Author content
All content in this area was uploaded by Yan Chuangye on Mar 05, 2024
Content may be subject to copyright.
Structure
Short Article
Structural Basis of RIP1 Inhibition by Necrostatins
Tian Xie,
1,4
Wei Peng,
1,4
Yexing Liu,
1,4
Chuangye Yan,
1
Jenny Maki,
2
Alexei Degterev,
2
Junying Yuan,
3
and Yigong Shi
1,
*
1
Ministry of Education Protein Science Laboratory, State Key Laboratory of Biomembrane and Membrane Biotechnology,
Tsinghua-Peking Joint Center for Life Sciences, Center for Structural Biology, School of Life Sciences and School of Medicine,
Tsinghua University, Beijing 100084, China
2
Department of Biochemistry, Tufts University School of Medicine, 136 Harrison Avenue, Boston, MA 02111, USA
3
Department of Cell Biology, Harvard Medical School, Boston, MA 02115, USA
4
These authors contributed equally to this work
*Correspondence: shi-lab@tsinghua.edu.cn
http://dx.doi.org/10.1016/j.str.2013.01.016
SUMMARY
Necroptosis is a cellular mechanism that mediates
necrotic cell death. The receptor-interacting serine/
threonine protein kinase 1 (RIP1) is an essential
upstream signaling molecule in tumor-necrosis-
factor-a-induced necroptosis. Necrostatins, a series
of small-molecule inhibitors, suppress necroptosis
by specifically inhibiting RIP1 kinase activity. Both
RIP1 structure and the mechanisms by which necros-
tatins inhibit RIP1 remain unknown. Here, we report
the crystal structures of the RIP1 kinase domain indi-
vidually bound to necrostatin-1 analog, necrostatin-3
analog, and necrostatin-4. Necrostatin, caged in a
hydrophobic pocket between the N- and C-lobes of
the kinase domain, stabilizes RIP1 in an inactive
conformation through interactions with highly con-
served amino acids in the activation loop and the
surrounding structural elements. Structural compar-
ison of RIP1 with the inhibitor-bound oncogenic
kinase B-RAF reveals partially overlapping binding
sites for necrostatin and for the anticancer compound
PLX4032. Our study provides a structural basis for
RIP1 inhibition by necrostatins and offers insights
into potential structure-based drug design.
INTRODUCTION
Necroptosis, also known as programmed necrosis, is a cellular
mechanism that mediates necrotic cell death (Christofferson
and Yuan, 2010;Vandenabeele et al., 2010). Necroptosis may
be activated in response to the stimulation of death receptors
by their cognate ligands in the absence of caspase activity, the
essential mediators of apoptosis. Necroptotic cell death is char-
acterized by typical morphological features of necrosis,
including early plasma membrane permeabilization, swollen
organelles, dilated nuclear membrane, and condensed chro-
matin (Vandenabeele et al., 2010). Necroptosis has been impli-
cated in the pathology of a number of diseases, such as ischemic
injury, neurodegeneration, and viral infection (Vandenabeele
et al., 2010). Inhibition of necroptosis represents an attractive
therapeutic strategy for preserving cell viability and functions in
these diseases.
The receptor-interacting serine/threonine protein kinase 1
(RIP1) contains an N-terminal kinase domain, an RIP homotypic
interaction motif (Sun et al., 2002), and a C-terminal death
domain (Stanger et al., 1995). Activation of tumor necrosis factor
receptor 1 (TNFR1) by TNFamay lead to multiple downstream
signaling events, including NF-kB activation, apoptosis, and
necroptosis. The Ser/Thr kinase activity of RIP1 is essential for
necroptosis and ripoptosome-mediated caspase-dependent
apoptosis (Feoktistova et al., 2011;Tenev et al., 2011) but is
dispensable for NF-kB activation, which relies on the ubiquitina-
tion of the central intermediate domain of RIP1. During necropto-
sis, RIP1 kinase activity is involved in mediating the transition
from receptor-anchored complex I to cytosolic complex II, which
is defined by the interaction of RIP1 with RIP3 (Christofferson
and Yuan, 2010;Vandenabeele et al., 2010). The kinase domains
of RIP1 and RIP3 share 33% sequence identity and 53%
sequence similarity.
Necrostatins, isolated as specific and potent small-molecule
inhibitors of necroptosis, directly inhibit the kinase activity of
RIP1 (Degterev et al., 2005,2008). Inhibition of RIP1 by necros-
tatins has been shown to ameliorate tissue damage in animal
models of ischemic brain injury (Degterev et al., 2005), retina
ischemia-reperfusion (Rosenbaum et al., 2010), myocardial
infarction (Lim et al., 2007), and traumatic brain injury (You
et al., 2008). The specificity and activity of three necrostatin-1
analogs have been examined thoroughly both in vitro and in vivo,
providing important information about necrostatin-1 in disease
models (Takahashi et al., 2012).
Despite rigorous effort, there is no atomic-resolution structure
for RIP1 or its homolog RIP3. The lack of structural information
hinders functional and mechanistic understanding of the RIP
kinases and limits the optimization of necrostatins and rational
drug design. In this study, we present the crystal structures of
the RIP1 kinase domain bound to three different necrostatins.
Our results reveal the structural basis of necrostatin-mediated
RIP1 inhibition and may facilitate the design and development
of RIP1-specfic small-molecule inhibitors.
RESULTS AND DISCUSSION
Characterization and Crystallization of the RIP1 Kinase
Domain
The human RIP1 kinase domain (residues 1–312) was ex-
pressed in insect cells and purified to homogeneity. To
characterize the activities of necrostatins, we reconstituted an
Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved 493
autophosphorylation assay as previously described (Degterev
et al., 2008). Using this assay, a chemically improved derivative
of necrostatin-1, (R)-5-([7-chloro-1H-indol-3-yl]methyl)-3-meth-
ylimidazolidine-2,4-dione (Nec-1a) (Figure 1A) (Degterev et al.,
2008), exhibited an inhibitory constant (IC
50
) of 0.32 mM for
RIP1 (Figure 1B; Figure S1A available online). By contrast,
necrostatin-4, (S)-N-(1-[2-chloro-6-fluorophenyl]ethyl)-5-cyano-
1-methyl-1H-pyrrole-2-carboxamide (Nec-4) (Figure S2A), and
a necrostatin-3 analog, 1-([3S,3aS]-3-[3-fluoro-4-[trifluorome-
thoxy]phenyl]-8-methoxy-3,3a,4,5-tetrahydro-2H-benzo[g]inda-
zol-2-yl)-2-hydroxyethanone (Nec-3a) (Figure S2B), exhibited
IC
50
values of about 0.37 and 0.44 mM, respectively (Figure S1A).
Despite rigorous effort, the human RIP1 kinase domain, both
by itself and in complex with necrostatins, defied crystallization.
Reasoning that flexible sequences might hinder crystallization,
we generated a number of constructs by trimming the hydro-
philic sequences at the N and/or C terminus. One such construct
(residues 1–294) gave rise to small crystals only in the presence
of Nec-1a. However, these crystals diffracted X-rays weakly.
Replacement of four cysteine residues by alanine in the RIP1
kinase domain led to marked improvement of crystal size and
morphology. The engineered RIP1 kinase domain (residues
1–294, C34A, C127A, C233A, and C240A) was also crystallized
in the presence of Nec-4 and Nec-3a. All crystals appear in the
space group P2
1
2
1
2
1
, with different cell dimensions (Table 1).
There are two molecules of RIP1 kinase domain in each asym-
metric unit. Because these two molecules display identical
features relevant for discussion, we only focus on one such
molecule.
Overall Structure of the RIP1 Kinase Domain
We attempted, and were successful at, structure determination
by molecular replacement using the atomic coordinates of
B-RAF kinase domain (Protein Data Bank [PDB] code 3C4C)
(Tsai et al., 2008). The X-ray structure of the RIP1 kinase domain
(referred to as RIP1 hereafter) bound to Nec-1a was refined at
2.25 A
˚resolution (Figure 1C; Table 1). RIP1 exhibits a canonical
kinase fold, with an N-lobe, a C-lobe, and an intervening activa-
tion loop (also known as the T-loop). Nec-1a is bound between
the N- and C-lobes, in close proximity to the activation loop
(Figure 1C). Notably, the Nec-1a-bound RIP1, where Nec-1a
binds largely outside of the ATP-binding pocket, does not
contain any nucleotide. Similar to B-RAF and other protein
kinases, the N-lobe comprises an antiparallel, five-stranded
bsheet and an activation helix (commonly known as the alpha
C-helix) (Figure 1C; Figure S1B). The C-lobe contains six ahelices
and a pair of bstrands (Figure 1C; Figure S1B). All essential
amino acids for ATP binding and hydrolysis in canonical kinases
are conserved in RIP1, including the catalytic triad residues
Lys45/Glu63/Asp156 and key residues in the P-loop (residues
24–31) and the catalytic loop (residues 136–143) (Figure S1B).
Most of these amino acids are visible in the electron density
map and are placed in the vicinity of the ATP-binding site in RIP1.
Recognition of Nec-1a
Nec-1a is buried in a relatively hydrophobic pocket between the
N-lobe and the C-lobe (Figure 1D; Figure S1C). The indole ring of
Nec-1a interacts with six amino acids, Met67, Leu70, Val75,
Leu129, Val134, and His136, through van der Waals contacts,
whereas the five-membered ring is surrounded by hydrophobic
amino acids Val76, Leu78, Leu90, Met92, Leu157, Leu159, and
Phe162 (Figure 1D). These hydrophobic interactions likely
provide the major driving force for binding. Consistent with the
structural observations, mutation of Phe162 to Glu led to
decreased inhibition of RIP1 by Nec-1 (Degterev et al., 2008).
The vast majority of these amino acids are highly conserved
among RIP3 and RIP1 orthologs in mouse, frog, and fish
(Figure S1B). This binding feature is consistent with the hydro-
phobic nature of Nec-1a, which, compared to a solubility of at
least 500 mM in dimethyl sulfoxide (DMSO), is soluble only up
to about 2 mM in aqueous solution.
In addition to the hydrophobic interactions, there are three
specific hydrogen bonds (H-bonds) at the interface, which
Figure 1. Overall Structure of Nec-1a-Bound RIP1 and Recognition
of Nec-1a by RIP1
(A) Chemical structure of Nec-1a.
(B) Nec-1a inhibits the autophosphorylation activity of RIP1 kinase domain
in vitro in a dose-dependent manner (see also Figure S1A).
(C) Two perpendicular views of the overall structure of Nec-1a-bound RIP1.
The N- and C-lobes of RIP1 kinase domain are colored cyan. The activation
loop (residues Asp156–Glu196) is colored red; residues 172–187 are disor-
dered in the crystal. Nec-1a is colored green. All structural figures were
prepared with PyMOL (DeLano, 2002).
(D) Two close-up views of the interactions between Nec-1a and surrounding
residues in RIP1. Nec-1a is shown in green ball-and-stick. The hydrophobic
residues around Nec-1a in RIP1 kinase domain are shown in sticks. H-bonds in
this and all other figures are represented by red or yellow dashed lines.
See also Figure S1.
Structure
Structure of Necrostatin-Bound RIP1
494 Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved
appear to anchor the orientation of Nec-1a in the greasy pocket
of RIP1. The indole ring of Nec-1a contributes one H-bond,
between its nitrogen atom and the hydroxyl oxygen of Ser161
on the activation loop (Figure 1D). The other two H-bonds are
mediated by the five-membered ring of Nec-la, involving the
carbonyl oxygen of Val76 and the amide nitrogen of Asp156
(Figure 1D). Consistent with a dominant role by hydrophobic
contacts, mutation of Ser161 to Ala only led to a 4-fold increase
of the IC
50
value (Figure S1D). The moderate increase is caused
by loss of one H-bond; Ala can be nicely accommodated in the
Nec-1a binding pocket. By contrast, mutation of Ser161 to a
bulkier, negatively charged amino acid Glu resulted in a 10-fold
decrease in efficiency of inhibition by Nec-1a (Figure S1D).
Nonetheless, the S161E mutation is tolerated by Nec-1a binding,
suggesting a remarkable conformational adjustability in the
binding pocket of RIP1.
Inactive Conformation of RIP1
Binding by Nec-1a presumably locks RIP1 in an inactive confor-
mation. Structural comparison between RIP1 and the catalytic
subunit of protein kinase A (PKA; PDB code 2CPK) (Knighton
et al., 1991a,1991b) confirms this notion (Figure 2). Most
notably, the activation helix in RIP1 is rotated by approximately
40relative to that in PKA, and the space vacated by the RIP1
activation helix is partially occupied by the activation loop and
inhibitor Nec-1a (Figure 2A). Consequently, one of the catalytic
triad residues, Glu63 in RIP1, is about 15 A
˚away from Lys45.
Figure 2. Nec-1a-Bound RIP1 Exists in an Inactive Conformation
(A) Structural comparison of PKA (PDB code 2CPK)(Knighton et al., 1991a) and
Nec-1a-bound RIP1. PKA is colored light gray, with its alpha C-helix in yellow
and T-loop in green. RIP1 is coloredcyan, with its alpha C-helix colored magenta
and T-loop colored red. Nec-1a is shown in green sticks. The side chains
of Lys72 and Glu91 of PKA and Lys45 and Glu63 of RIP1 are shown in sticks.
(B) A network of conserved H-bonds around the DFG motif in PKA of the active
configuration.
(C) H-bonds between Nec-1a and the T-loop reorient the DLG motif in RIP1.
(D) A stereo view of the stacked spine structure in PKA and distorted spine
structure in RIP1. PKA and RIP1 are colored light gray and cyan, respectively.
The side chains of the spine structures are shown in sticks.
Table 1. Statistics of Data Collection and Refinement
Data
Nec-1a-Bound
RIP1
Nec-4-Bound
RIP1
Nec-3a-Bound
RIP1
Integration package HKL2000 HKL2000 HKL2000
Space group P2
1
2
1
2
1
P2
1
2
1
2
1
P2
1
2
1
2
1
Unit cell (A
˚) 47.13, 93.60,
129.07
46.84, 97.14,
128.12
82.63, 91.41,
103.33
Unit cell () 90, 90, 90 90, 90, 90 90, 90, 90
Wavelength (A
˚) 1.0000 0.9793 0.9793
Resolution (A
˚)402.25
(2.332.25)
401.80
(1.861.80)
402.90
(3.002.90)
R
merge
(%) 5.9 (20.1) 6.0 (22.1) 7.3 (81.8)
I/sigma 21.2 (5.2) 24.8 (6.6) 35.0 (3.8)
Completeness (%) 99.3 (98.7) 98.8 (98.7) 99.7 (100.0)
Number of measured
reflections
144,351 235,759 171,368
Number of unique
reflections
51,908 54,270 18,400
Redundancy 2.8 (2.7) 4.3 (4.4) 9.3 (9.7)
Wilson B factor (A
˚
2
) 40.2 20.1 96.6
R
work
/R
free
(%) 21.90/24.80 20.03/21.98 22.70/26.65
No. atoms
Overall 4,233 4,907 4,375
Protein 4051 4278 4313
Ligand 38 42 62
Water 132 578 0
Other entities 12 9 0
Average B value (A
˚
2
)
Overall 44.12 29.05 93.41
Protein 44.30 28.37 93.43
Ligand 31.75 17.67 91.91
Water 41.38 35.61 0
Other entities 54.09 30.43 0
rmsd
Bonds (A
˚) 0.010 0.009 0.010
Angle () 1.249 1.261 1.366
Ramachandran plot statistics (%)
Most favorable 89.1 91.6 85.8
Additionally
allowed
9.6 7.8 13.0
Generously
allowed
0.7 0.6 1.3
Disallowed 0.7 0.0 0.0
Values in parentheses are for the highest-resolution shell. R
merge
=
S
h
S
i
jI
h,i
-I
h
j/S
h
S
i
I
h,i
, where I
h
is the mean intensity of the iobservations
of symmetry-related reflections of h.R=SjF
obs
-F
calc
j/SF
obs
, where F
calc
is the calculated protein structure factor from the atomic model (R
free
was calculated with 5% of the reflections selected).
Structure
Structure of Necrostatin-Bound RIP1
Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved 495
In the active kinase PKA, Glu91 in the alpha C-helix (which corre-
sponds to Glu63 in RIP1) forms a salt bridge with Lys72 (which
corresponds to Lys45 in RIP1), which then stabilizes ATP
through H-bonds to the a- and b-phosphates.
There are additional important structural differences between
RIP1 and active protein Ser/Thr kinases. In PKA, the DFG motif in
the activation loop forms a characteristic H-bond network both
within and with a neighboring amino acid, Arg165 (Figure 2B).
In this network, the side chain of Asp184 accepts an H-bond
from the amide nitrogen of Gly186, whereas the carbonyl oxygen
of Phe185 accepts an H-bond from the amide nitrogen of Ala188
(the DFG + 2 residue) and the carbonyl oxygen of Phe187 (the
DFG + 1 residue) accepts an H-bond from the side chain of
a highly conserved residue, Arg165, in the catalytic loop (Kornev
et al., 2006)(Figure 2B). These three consecutive H-bonds, with
the same dipolar orientation, help to maintain PKA and other
protein Ser/Thr kinases in an active conformation. By contrast,
the corresponding DLG motif in RIP1 exists in a quite different
conformation, and Asp156 and Ser161 of the DLG motif interact
with the Nec-1a inhibitor through two H-bonds (Figure 2C).
An active protein Ser/Thr kinase usually contains an assem-
bled hydrophobic spine structure both to stabilize the kinase
fold and to help the kinase to go through open and closed
conformations during catalysis (Kornev et al., 2006). In PKA,
the hydrophobic spine comprises Leu106, Leu95, Phe185, and
Tyr164, which appear in a spatially linear order and stack against
each other through van der Waals interactions (Figure 2D). In
RIP1, the corresponding residues Leu78, Met67, Leu157, and
His 136, no longer form a linear spine. Rather, Met67 and
Leu157 move away from the spine and from each other, resulting
in an approximate square distribution (Figure 2D).
RIP1 Bound to Nec-4 and Nec-3a
The crystal structures of the RIP1 kinase domain bound to Nec-4
and Nec-3a were determined at 1.8 and 2.9 A
˚, respectively
(Figures S2A and S2B; Table 1). These two RIP1 structures
are very similar to that of Nec-1a-bound RIP1. Nec-1a-bound
RIP1 exhibits root-mean-squared deviation (rmsd) values
of 0.555 A
˚over 223 aligned Caatoms with Nec-4-bound
RIP1 and 0.635 A
˚over 222 aligned Caatoms with Nec-3a-bound
RIP1 (Figure 3A). All structural elements important for catalysis
and inhibitor binding are invariant among these three structures.
Notably, despite their very different chemical formulas, all three
necrostatins are bound in the same general location of RIP1
(Figure 3A). The conformation of the P-loop in Nec-4-bound
RIP1 adopts a well-defined conformation and is quite different
from that in the ATP-bound CDK2 structure (PDB code 1FIN)
(Jeffrey et al., 1995)(Figure S2C). The P-loop in Nec-4-bound
RIP1 occupies the same general location as that required for
ATP binding (Figure S2D), which explains why a nucleotide can
no longer bind to Nec-4-bound RIP1. In addition, the Nec-4-
binding pocket is far away from the ATP-binding site (Figures
S2C and S2D).
In both structures, necrostatin is bound in a hydrophobic
pocket, similar to that in Nec-1a-bound RIP1. The same set of
hydrophobic amino acids is involved in binding to Nec-1a,
Nec-4, and Nec-3a, except that the exact van der Waals interac-
tions are adjusted to account for the different chemical struc-
tures of the three necrostatins. The RIP1 amino acids that are
involved in the specific H-bonds with necrostatins are also
similar in these three structures. Nec-4 forms two H-bonds
with the carbonyl oxygen of Ile154 and the amide nitrogen of
Asp156 (Figure 3B), and Nec-3a makes three H-bonds with
the amide nitrogen and carbonyl oxygen of Asp156 and the
side chain of Ser161 (Figure 3C). These conserved structural
features are likely to serve as an important guide for improve-
ment of these RIP1-specific inhibitors. Nonetheless, the
chemical groups in these necrostatins that are involved in the
interactions are quite different among the three structures
(Figures 3B and 3C).
Comparison between RIP1 and PLX4032-Bound B-RAF
The human Ser/Thr protein kinase B-RAF, encoded by the proto-
oncogene BRAF, is a key signaling molecule in the RAF-MEK-
ERK pathway for the regulation of cell growth, proliferation,
and differentiation (Robinson and Cobb, 1997;Wan et al.,
2004). A number of activating B-RAF mutations have been found
to cause cancers including melanoma, ovarian cancer, and colo-
rectal carcinoma (Ro
¨ring and Brummer, 2012). A small-molecule
inhibitor of B-RAF, named Vemurafenib, has been successfully
used in the clinic to treat late-stage melanoma with a V600E
mutation in the B-RAF protein. Vemurafenib (also known as
Figure 3. Structures of RIP1 Bound to Nec-4 and Nec-3a
(A) Structural comparison of Nec-1a-bound RIP1, Nec-4-bound RIP1, and
Nec-3a-bound RIP1. Nec-1a-, Nec-4-, and Nec-3a-bound RIP1 are colored
dark green, pink, and yellow, respectively. Nec-1a, Nec-4, and Nec-3a are
colored green, magenta, and yellow, respectively.
(B) A close-up view of the interactions between Nec-4 and surrounding resi-
dues in RIP1. Nec-4 is shown in magenta ball-and-stick. The hydrophobic
residues around Nec-4 in RIP1 kinase domain are shown in sticks.
(C) A close-up view of the interactions between Nec-3a and surrounding
residues in RIP1. The hydrophobic residues around Nec-3a in RIP1 kinase
domain are shown in sticks.
See also Figure S2.
Structure
Structure of Necrostatin-Bound RIP1
496 Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved
PLX4032), discovered by a scaffold-based drug design
approach (Bollag et al., 2010;Tsai et al., 2008), potently inhibits
the kinase activity of B-RAF-V600E (Bollag et al., 2010).
The human RIP1 kinase domain shares 28% sequence iden-
tity and 47% similarity with the kinase domain of B-RAF (Fig-
ure S1B). Consistent with the degree of sequence conservation,
the Nec-1a-bound RIP1 structure is quite similar to that of
PLX4032-bound B-RAF, with an rmsd of 1.041 A
˚over 140
aligned Caatoms (Figure 4A). Most notably, necrostatins and
PLX4032 bind to the same general locations in RIP1 and
B-RAF, respectively. A portion of the more extended and larger
molecule, PLX4032, overlaps with Nec-1a (Figure 4A, inset).
However, the binding interactions for necrostatins are quite
different from those for PLX4032. Compared to the hydrophobic
environment for necrostatins (Figure 4B), PLX4032 appears to
be bound in a relatively hydrophilic pocket that also contains
a number of hydrophobic amino acids (Figure 4C). Conse-
quently, PLX4032 forms four H-bonds with the main-chain
groups of amino acids Gln530, Cys532, Gly596, and Asp594
(Figure 4A, inset).
Given its important role in diseases, RIP1 has been pursued as
a target for potential therapeutic intervention. The advent of the
crystal structure may facilitate rational drug design and develop-
ment. Interestingly, unlike the B-RAF binding pocket for
PLX4032, which is extended and fits PLX4032 snugly (Figure 4C),
the RIP1 binding pocket for necrostatins is considerably larger
than the van der Waals surface of the bound necrostatin mole-
cules (Figure S3). In fact, the van der Waals volumes of the
secluded binding cavities in Nec-1a-, Nec-4-, and Nec-3a-
bound RIP1 are approximately 403, 643, and 754 A
˚
3
, respec-
tively (calculated by Chimera) (Pettersen et al., 2004). These
volumes are approximately 1.7, 2.5, and 2.3 times the volumes
for Nec-1a, Nec-4, and Nec-3a, respectively. This structural
feature strongly suggests a potential design strategy to exploit
the empty space between the bound necrostatins and
surrounding amino acids in the binding pocket of RIP1. For
example, a number of main-chain groups from RIP1 residues
are positioned in the cavity and may be potential candidates
for H-bonds with improved necrostatins. In addition, the
surrounding hydrophobic amino acids provide a rich source for
van der Waals interactions with a larger necrostatin molecule.
Each of the three necrostatin molecules used for cocrystalliza-
tion contains two ring structures connected by one to four linear,
covalent bonds, and this structural motif might also serve as
a target for improvement.
In summary, we report the crystal structures of the RIP1 kinase
domain at atomic resolutions. These structures were determined
in the presence of bound necrostatin inhibitors. Analysis of the
inhibitor-kinase interactions reveals intriguing clues for potential
drug discovery and development. Our study serves as an impor-
tant framework for mechanistic understanding of RIP1.
EXPERIMENTAL PROCEDURES
Chemicals and Reagents
Nec-1a, Nec-4, and Nec-3a were kindly provided by Stephen M. Condon of
the TetraLogic Pharmaceuticals, and all of them are single enantiomers.
Necrostatins were dissolved in 100% DMSO with a stock concentration of
200 mM. All other chemicals were purchased from Sigma-Aldrich.
Protein Preparation and Crystallization
The RIP1 constructs were subcloned into the Nde I and Xho I sites of pFastBac
(Invitrogen) with an N-terminal 103His-tag and an engineered cleavage site for
the caspase drICE. The cleavage site of drICE is Asp-Glu-Val-Asp-Ala, and
cleavage occurs between Asp and Ala. The linker between the drICE cleavage
site and RIP1 is Gly-Ser-Gly. Bacmids were generated in DH10Bac cells, and
the resulting baculoviruses were generated and amplified in Sf-9 insect
cells. After infection by baculoviruses for 48 hr, the cells were harvested in
a buffer containing 25 mM Tris (pH 8.0) and 150 mM NaCl. The RIP1 kinase
domain was purified to homogeneity by nickel affinity chromatography
(QIAGEN), anion-exchange chromatography (Source-15Q, GE Healthcare),
and gel-filtration chromatography (Superdex-200, GE Healthcare). An addi-
tional step of drICE cleavage was performed to remove the 103His tag just
prior to gel filtration. The purified RIP1 was in a buffer containing 25 mM Tris
(pH 8.0), 150 mM NaCl, and 5 mM dithiothreitol.
About 20 baculoviruses, each expressing an RIP1 kinase domain with
a distinct N- and C-terminal boundary, were generated for screening of the
optimal RIP1 construct for crystallization. On the basis of protein solution
behavior, the wild-type RIP1 kinase domain (residues 1–294) was chosen for
extensive crystallization screening. The purified RIP1 (residues 1–294) was
concentrated to 9 mg/ml, and the necrostatin analogs were added at a final
concentration of 2.5–5 mM (Nec-1a, 2.5 mM; Nec-4, 5 mM; Nec-3a, 4 mM).
Crystals were grown at 18C using the hanging-drop vapor-diffusion method.
Unfortunately, these crystals diffracted X-rays weakly. To help improve diffrac-
tion, we attempted dehydration, seeding, crystal aging, construct reengineer-
ing, and cysteine mutation. In the end, introduction of four cysteine mutations
Figure 4. Structural Comparison between RIP1 and PLX4032-Bound
B-RAF
(A) Structural comparison of B-RAF (PDB code 3OG7) (Bollag et al., 2010) and
Nec-1a-bound RIP1. PLX4032-bound B-RAF is colored blue; Nec-1a-bound
RIP1 is colored cyan. PLX4032 and Nec-1a are colored in light gray and green,
respectively. The H-bonds in B-RAF and RIP1 are shown in yellow and red
dashed lines, respectively.
(B) The pocket for Nec-1a in RIP1. The surface of RIP1 is shown by electro-
static potential; Nec-1a is shown in green sticks.
(C) The pocket for PLX4032 in B-RAF. The surface of B-RAF is shown by
electrostatic potential; PLX4032 is shown in green sticks.
See also Figure S3.
Structure
Structure of Necrostatin-Bound RIP1
Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved 497
into the RIP1 kinase domain (C34A, C127A, C233A, and C240A) resulted in
significant improvement of diffraction.
All mutations were generated with two-step PCR and verified by plasmid
sequencing. Crystals of the Nec-1a-bound RIP1 kinase domain appeared after
2 days in a well buffer containing 0.25 M NH
4
I, 20% polyethylene glycol (PEG)
3350, and 0.03 M glycyl-glycyl-glycine. Crystals of the Nec-4-bound RIP1
kinase domain appeared after 1 day in a well buffer containing 0.15 M NH
4
I
and 17% PEG 3350. Crystals of the Nec-3a-bound RIP1 kinase domain
appeared in a well buffer containing 0.1 M HEPES, pH 7.0, and 15% PEG
20,000.
Data Collection and Structure Determination
Despite different crystallization conditions, all crystals of necrostatin-bound
RIP1 kinase domain are in the same space group, P2
1
2
1
2
1
, but have different
unit-cell dimensions. There are two molecules of RIP1 in each asymmetric unit.
All diffraction data sets were collected at Shanghai Synchrotron Radiation
Facility beamline BL17U and processed using HKL2000 (Otwinowski and
Minor, 1997). Further data processing was carried out using programs from
the CCP4 suite (CCP4, 1994). Human B-RAF (PDB code 3C4C) (Tsai et al.,
2008) was selected as the search model for molecular replacement. To
make a more accurate model, the program Chainsaw (Stein, 2008) was applied
to make a modification of the structure. The sequence alignment between
B-RAF and RIP1 was used as input to Chainsaw. The newly modified structure
was used as the initial search model for molecular replacement. Molecular
replacement was performed with the program PHASER (McCoy et al., 2007).
Manual model rebuilding and refinement were iteratively performed with
COOT (Emsley and Cowtan, 2004) and PHENIX (Adams et al., 2002), respec-
tively. Data collection and refinement statistics are summarized in Table 1.
In Vitro Kinase Activity Assay
For the in vitro kinase activity assay, wild-type or mutants of the RIP1 kinase
domain (residues 1–312) were used. RIP1 protein at 3 mM was incubated in
50 ml reaction buffer containing 25 mM HEPES (pH 7.0), 10 mM MgCl
2
,
50 mM NaCl, and 1 mM dithiothreitol for 15 min at 25C in the presence of
varying concentrations of necrostatins. For these assays, compounds were
diluted to appropriate concentrations in DMSO and added in the assay
systems with a final concentration of 1% DMSO. Kinase reactions were initi-
ated by addition of 10 mM cold ATP and 1 mCi of [g-
32
P] ATP, and the reactions
were carried out at 25C for 30 min. Reactions were stopped by adding SDS-
PAGE sample buffer and subjected to 16% SDS-PAGE. RIP1 band was visu-
alized by autoradiography.
ACCESSION NUMBERS
The atomic coordinates and structure factor files of RIP1 bound to Nec-1a,
Nec-4, and Nec-3a have been deposited in the Protein Data Bank with the
accession codes 4ITH, 4ITJ, and 4ITI, respectively.
SUPPLEMENTAL INFORMATION
Supplemental Information contains three figures and can be found with this
article online at http://dx.doi.org/10.1016/j.str.2013.01.016.
ACKNOWLEDGMENTS
We thank J. He from the Shanghai Synchrotron Radiation Facility beamline
BL17U for assistance. This work was supported by funds from the National
Natural Science Foundation of China (projects 31021002 and 31130002).
Received: November 19, 2012
Revised: January 13, 2013
Accepted: January 23, 2013
Published: March 5, 2013
REFERENCES
Adams, P.D., Grosse-Kunstleve, R.W., Hung, L.W., Ioerger, T.R., McCoy, A.J.,
Moriarty, N.W., Read, R.J., Sacchettini, J.C., Sauter, N.K., and Terwilliger, T.C.
(2002). PHENIX: building new software for automated crystallographic struc-
ture determination. Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954.
Bollag, G., Hirth, P., Tsai, J., Zhang, J., Ibrahim, P.N., Cho, H., Spevak, W.,
Zhang, C., Zhang, Y., Habets, G., et al. (2010). Clinical efficacy of a RAF
inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature
467, 596–599.
CCP4 (Collaborative Computational Project, Number 4). (1994). The
CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol.
Crystallogr. 50, 760–763.
Christofferson, D.E., and Yuan, J. (2010). Necroptosis as an alternative form of
programmed cell death. Curr. Opin. Cell Biol. 22, 263–268.
Degterev, A., Huang, Z., Boyce, M., Li, Y., Jagtap, P., Mizushima, N., Cuny,
G.D., Mitchison, T.J., Moskowitz, M.A., and Yuan, J. (2005). Chemical inhibitor
of nonapoptotic cell death with therapeutic potential for ischemic brain injury.
Nat. Chem. Biol. 1, 112–119.
Degterev, A., Hitomi, J., Germscheid, M., Ch’en, I.L., Korkina, O., Teng, X.,
Abbott, D., Cuny, G.D., Yuan, C., Wagner, G., et al. (2008). Identification of
RIP1 kinase as a specific cellular target of necrostatins. Nat. Chem. Biol. 4,
313–321.
DeLano, W.L. (2002). The PyMOL Molecular Graphics System. http://www.
pymol.org.
Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular
graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132.
Feoktistova, M., Geserick, P., Kellert, B., Dimitrova, D.P., Langlais, C., Hupe,
M., Cain, K., MacFarlane, M., Ha
¨cker, G., and Leverkus, M. (2011). cIAPs block
Ripoptosome formation, a RIP1/caspase-8 containing intracellular cell death
complex differentially regulated by cFLIP isoforms. Mol. Cell 43, 449–463.
Jeffrey, P.D., Russo, A.A., Polyak, K., Gibbs, E., Hurwitz, J., Massague
´, J., and
Pavletich, N.P. (1995). Mechanism of CDK activation revealed by the structure
of a cyclinA-CDK2 complex. Nature 376, 313–320.
Knighton, D.R., Zheng, J.H., Ten Eyck, L.F., Ashford, V.A., Xuong, N.H., Taylor,
S.S., and Sowadski, J.M. (1991a). Crystal structure of the catalytic subunit of
cyclic adenosine monophosphate-dependent protein kinase. Science 253,
407–414.
Knighton, D.R., Zheng, J.H., Ten Eyck, L.F., Xuong, N.H., Taylor, S.S., and
Sowadski, J.M. (1991b). Structure of a peptide inhibitor bound to the catalytic
subunit of cyclic adenosine monophosphate-dependent protein kinase.
Science 253, 414–420.
Kornev, A.P., Haste, N.M., Taylor, S.S., and Eyck, L.F. (2006). Surface compar-
ison of active and inactive protein kinases identifies a conserved activation
mechanism. Proc. Natl. Acad. Sci. USA 103, 17783–17788.
Lim, S.Y., Davidson, S.M., Mocanu, M.M., Yellon, D.M., and Smith, C.C.
(2007). The cardioprotective effect of necrostatin requires the cyclophilin-D
component of the mitochondrial permeability transition pore. Cardiovasc.
Drugs Ther. 21, 467–469.
McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C.,
and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Cryst. 40,
658–674.
Otwinowski, Z., and Minor, W. (1997). Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 276, 307–326.
Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S., Greenblatt, D.M.,
Meng, E.C., and Ferrin, T.E. (2004). UCSF Chimera—a visualization system
for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612.
Robinson, M.J., and Cobb, M.H. (1997). Mitogen-activated protein kinase
pathways. Curr. Opin. Cell Biol. 9, 180–186.
Ro
¨ring, M., and Brummer, T. (2012). Aberrant B-Raf signaling in human
cancer — 10 years from bench to bedside. Crit. Rev. Oncog. 17, 97–121.
Rosenbaum, D.M., Degterev, A., David, J., Rosenbaum, P.S., Roth, S., Grotta,
J.C., Cuny, G.D., Yuan, J., and Savitz, S.I. (2010). Necroptosis, a novel form of
caspase-independent cell death, contributes to neuronal damage in a retinal
ischemia-reperfusion injury model. J. Neurosci. Res. 88, 1569–1576.
Stanger, B.Z., Leder, P., Lee, T.H., Kim, E., and Seed, B. (1995). RIP: a novel
protein containing a death domain that interacts with Fas/APO-1 (CD95) in
yeast and causes cell death. Cell 81, 513–523.
Structure
Structure of Necrostatin-Bound RIP1
498 Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved
Stein, N. (2008). CHAINSAW: a program for mutating pdb files used as
templates in molecular replacement. J. Appl. Crystallogr. 41, 641–643.
Sun, X., Yin, J., Starovasnik, M.A., Fairbrother, W.J., and Dixit, V.M. (2002).
Identification of a novel homotypic interaction motif required for the phosphor-
ylation of receptor-interacting protein (RIP) by RIP3. J. Biol. Chem. 277, 9505–
9511.
Takahashi, N., Duprez, L., Grootjans, S., Cauwels, A., Nerinckx, W.,
DuHadaway, J.B., Goossens, V., Roelandt, R., Van Hauwermeiren, F.,
Libert, C., et al. (2012). Necrostatin-1 analogues: critical issues on the speci-
ficity, activity and in vivo use in experimental disease models. Cell Death
Dis. 3, e437.
Tenev, T., Bianchi, K., Darding, M., Broemer, M., Langlais, C., Wallberg, F.,
Zachariou, A., Lopez, J., MacFarlane, M., Cain, K., and Meier, P. (2011). The
Ripoptosome, a signaling platform that assembles in response to genotoxic
stress and loss of IAPs. Mol. Cell 43, 432–448.
Tsai, J., Lee, J.T., Wang, W., Zhang, J., Cho, H., Mamo, S., Bremer, R., Gillette,
S., Kong, J., Haass, N.K., et al. (2008). Discovery of a selective inhibitor of
oncogenic B-Raf kinase with potent antimelanoma activity. Proc. Natl. Acad.
Sci. USA 105, 3041–3046.
Vandenabeele, P., Galluzzi, L., Vanden Berghe, T., and Kroemer, G. (2010).
Molecular mechanisms of necroptosis: an ordered cellular explosion. Nat.
Rev. Mol. Cell Biol. 11, 700–714.
Wan, P.T., Garnett, M.J., Roe, S.M., Lee, S., Niculescu-Duvaz, D., Good, V.M.,
Jones, C.M., Marshall, C.J., Springer, C.J., Barford, D., and Marais, R.; Cancer
Genome Project. (2004). Mechanism of activation of the RAF-ERK signaling
pathway by oncogenic mutations of B-RAF. Cell 116, 855–867.
You, Z., Savitz, S.I., Yang, J., Degterev, A., Yuan, J., Cuny, G.D., Moskowitz,
M.A., and Whalen, M.J. (2008). Necrostatin-1 reduces histopathology and
improves functional outcome after controlled cortical impact in mice.
J. Cereb. Blood Flow Metab. 28, 1564–1573.
Structure
Structure of Necrostatin-Bound RIP1
Structure 21, 493–499, March 5, 2013 ª2013 Elsevier Ltd All rights reserved 499
