Site-Directed Fragment-Based Screening for the Discovery of Protein-Protein Interaction Stabilizers

Article abstract of DOI:10.1021/jacs.8b11658

Modulation of protein-protein interactions (PPIs) by small molecules has emerged as a valuable approach in drug discovery. Compared to direct inhibition, PPI stabilization is vastly underexplored but has strong advantages, including the ability to gain se

Full text of DOI:10.1021/jacs.8b11658

Subscriber access provided by UNIV OF OREGON
Article
Site-Directed Fragment-based Screening for the
Discovery of Protein-Protein Interaction Stabilizers
Eline Sijbesma, Kenneth K Hallenbeck, Seppe Leysen, Pim de Vink, Lukasz
Skora, Wolfgang Jahnke, Luc Brunsveld, Michelle R. Arkin, and Christian Ottmann
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b11658 • Publication Date (Web): 01 Feb 2019
Downloaded from http://pubs.acs.org on February 2, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Site-Directed Fragment-based Screening for the Discovery of
Protein-Protein Interaction Stabilizers
†
‡
†
†
§
⊥
⊥
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Eline Sijbesma , Kenneth K. Hallenbeck , Seppe Leysen , Pim J. de Vink , Lukasz Skóra , Wolfgang
Jahnke§, Luc Brunsveld†, Michelle R. Arkin*,‡, Christian Ottmann*,†,#
† Laboratory of Chemical Biology, Department of Biomedical Engineering and Institute for Complex Molecular Systems
(ICMS), Eindhoven University of Technology, Eindhoven, the Netherlands
‡ Department of Pharmaceutical Chemistry and Small Molecule Discovery Centre (SMDC), University of California, San
Francisco, United States
§ Chemical Biology and Therapeutics, Novartis Institutes for Biomedical Research, Basel, Switzerland
# Department of Chemistry, University of Duisburg-Essen, Essen, Germany
ABSTRACT: Modulation of Protein-Protein Interactions (PPIs) by small molecules has emerged as a valuable approach in drug
discovery. Compared to direct inhibition, PPI stabilization is vastly underexplored but has strong advantages, including the ability to
gain selectivity by targeting an interface formed only upon association of proteins. Here, we present the application of a site-directed
screening technique based on disulfide trapping (tethering) to select for fragments that enhance the affinity between protein partners.
We target the phosphorylation-dependent interaction between the hub protein 14-3-3 and a peptide derived from Estrogen Receptor
 (ER), an important breast cancer target that is negatively regulated by 14-3-3. We identify orthosteric stabilizers that increase
14-3-3/ER affinity up to 40-fold and propose the mechanism of stabilization based on X-ray crystal structures. These fragments
already display partial selectivity towards ER-like motifs over other representative 14-3-3 clients. This first of its kind study
illustrates the potential of the tethering approach to overcome the hurdles in systematic PPI stabilizer discovery.
Introduction
We envisioned that disulfide trapping (tethering) would be a
promising technology to develop such a platform. Disulfide
trapping allows site-directed selection of ligands and readily
measures cooperative binding – qualities that address the main
challenges posed by screening for PPI stabilizers. Since the
technology was pioneered by Wells, Erlanson and co-workers11,
disulfide trapping has successfully identified allele-specific
inhibitors of oncogenic KRas (G12C)12, allosteric ligands of
Once considered ‘undruggable’, protein-protein interactions
(PPIs) have been successfully targeted by drug-like molecules
in the past 15-20 years.1–4 In contrast to the fruitful development
of PPI disruptors, examples of targeted small-molecule PPI
stabilizers are relatively scarce, and dedicated screening
approaches for PPI stabilizer identification are virtually
absent.5–7
kinase PDK113 and inhibitors of the IL-2/IL-2receptor PPI14,15
.
Therapeutic proof-of-concept for PPI stabilization has been
provided by natural products, including the anti-tumor drug
paclitaxel and immune suppressants rapamycin and FK506.6,7
Additionally, a number of successes using synthetic molecules
have been reported, such as the BRD4-dimer stabilizer
(biBET)8 and the allosteric stabilizer of the tetramer
transthyretin (tafamidis).9 Synthetic approaches - proteolysis
targeting chimeras (PROTACs) and immunomodulatory drugs
(iMiDs) - apply this principle to drive the association of two
proteins that would not otherwise interact.10 These clinical and
chemical-biology applications justify the development of
technology platforms to allow systematic stabilization of PPI,
especially given the fact that most discoveries of PPI stabilizing
molecules have been serendipitous. The design rules for a good
stabilizer are poorly understood and technical difficulties
complicate assay development. There is thus an unmet need for
approaches that overcome inherent limitations of conventional
ligand screening to identify PPI stabilizers.
Here, we offer the first demonstration of the tethering
technology to identify small-molecule stabilizers of a protein
complex.
We selected the interaction between the hub protein 14-3-3
and the phosphorylated motif derived from the breast-cancer-
associated transcription factor Estrogen Receptor  (ER) as a
suitable and relevant test case. With >300 cellular interaction
partners, including Raf kinases16, heat shock proteins,17
oncogenes18 and tumor suppressors (p53)19, 14-3-3 proteins are
central regulators in many biological processes and
pathologies.20–22 For example, 14-3-3 binding antagonizes
multiple transcription factors that act as oncogenic drivers.
Since inhibition of transcriptional activity is a central
therapeutic challenge in cancer, we have focused our efforts
towards identifying small molecule stabilizers for this PPI class.
De Vries-van Leeuwen et al reported that ER is
phosphorylated at the penultimate residue T594 and that
binding of this site to 14-3-3 reduces its estradiol-dependent
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
mutated (C38N; N being the most common residue at this
transcriptional activity. Inhibition of ER activity is enhanced
by the natural product Fusicoccin A (FC-A), which binds at the
14-3-3/ER interface (Figure 1a).18 Stabilizing this PPI was
proposed to be a valid alternative strategy for interfering with
ER-positive breast cancer.
1
2
3
4
5
6
7
8
position) and a cysteine introduced at positions 42 or 45, one or
two -helix turns towards the ER binding site, respectively
(Figure 1a). The three 14-3-3 cysteine constructs were
screened both in apo form and in complex with a 15-mer
phosphopeptide representing the 14-3-3-binding motif of ER
Here, we identified several disulfide fragments that each
bound cooperatively to a complex of 14-3-3 and ER-derived
phosphopeptide (ER-pp). Hits selectively increased the
binding affinity between ER-pp and 14-3-3 by as much as 40-
fold; multiple x-ray co-structures suggested the mechanism of
stabilization. Disulfide tethering is a promising approach to
identify starting points to specifically stabilize protein-peptide
interactions and provides a first and long-needed, systematic
screening platform for PPI-stabilizing molecules.
(ER-pp;
KYYITGEAEGFPApT594V;
Figure
1b).
Phosphorylated motifs derived from 14-3-3 client proteins
recapitulate key interactions of the PPI, and mutating the single
phosphorylation site can completely abrogate the interaction in
vitro and in cells.24 Short 14-3-3 client-derived
phosphopeptides can thus be used in vitro as surrogates for the
PPI; this approach has been used to characterize FC-A/14-3-
3/client complexes and to screen for inhibitors25, e.g., of the 14-
3-3/Tau PPI.26,27
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
Apo-14-3-3 or the 14-3-3/ER-pp complex was screened
against a 1600-member disulfide library under mildly reducing
conditions (100 M betamercaptoethanol; ME). Conjugate
formation for each individual reaction was analyzed by intact
protein MS. Three peaks observed in mass spectra
corresponded to apo, ME-capped, and fragment-conjugated
14-3-3. The ‘percent tethering’, defined as the intensity of the
fragment-specific conjugate protein peak divided by the sum of
the intensities for all protein peaks, was calculated for each
individual experiment using an automated pipeline.28
Tethering uses a cysteine on the target protein as a reactivity
handle to trap disulfide-containing fragments that have an
inherent (weak) binding affinity for a target pocket near the
cysteine. The bound fragments can then be detected by intact
protein mass spectrometry (MS).11,23 Our disulfide trapping
approach was designed to target FC-A’s hydrophobic pocket at
the 14-3-3/ER interface. Sigma is the only one of seven
human 14-3-3 isoforms that contains a native surface-exposed
cysteine (C38) at the edge of this pocket. We further designed
two protein constructs in which the wildtype cysteine was
Figure 1. a) The target pocket for stabilizing the interaction of 14-3-3 (white surface) and ER-pp (green sticks) bound by FC-A (yellow
sticks) (PDB: 4JDD). Indicated are the native cysteine (C38; red surface) and introduced C42 (orange) or C45 (yellow). b) Schematic
illustration of the approach for selecting stabilizers by disulfide trapping. The cysteine-containing protein is incubated with an arrayed
disulfide-fragment library under reducing conditions in the apo state (i) or bound to ER-pp (ii). The equilibrium is shifted to the fragment-
conjugated protein only when it has an inherent affinity for the nearby target pocket. c) 2D-plots illustrate the correlation between % tethering
of individual fragments ( ) for apo 14-3-3 and 14-3-3/ER-pp complex. The hit-selection threshold (mean + three standard deviations;
+3) in each screen is indicated by a green dashed line. Fragment 1 ( ) and 2 ( ) are indicated for each screen; both only score as a ‘hit’
for C42.
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
For each screen, hits were categorized as competitive (only a
1, 2, and seven additional fragments that bound to the 14-3-
3(C42)/ER-pp complex were selected for follow-up
experiments (see Figure S1 for mass spectra).
hit in the apo screen), cooperative (preferentially a hit for the
protein-peptide complex), or neutral (a hit both in the apo and
the 14-3-3ER-pp complex; Figure 1c). C38 yielded the
highest fraction of cooperative hits, but the maximal percent
tethering was low (<55% conjugated), suggesting that
fragments bound to C38 had a low affinity for the pocket.
Conversely, C45, closest to the target pocket, yielded a large
fraction of hits with >75% conjugation, with more competitive
than cooperative hits. Satisfyingly, C42, with an intermediate
position, yielded hits for both apo and ER-pp bound 14-3-3,
suggesting an optimal distance and orientation towards the
target pocket to identify potent and cooperative fragments.
The 14-3-3(C42)-binding hits were confirmed in dose-
response experiments detected by intact protein MS. Both 1 and
2 demonstrated strong preferential binding to the 14-3-3/ER-
pp complex over the 14-3-3 protein alone (Figure 2c). Binding
of the tethered fragment 2 was improved ~300-fold, from an
effective concentration (EC50) ~1 mM for apo to EC50 = 3 M
for the ER-pp bound 14-3-3. Fragment 1, the N-methylated
version of 2, showed an EC50 ~100 M for binding to apo but
remained >80% tethered to 14-3-3 bound to ER-pp down to
100 nM fragment, even in the stringent disulfide-reducing
condition of 1 mM ME. Cooperative binding was less
pronounced for the other primary screening hits (data not
shown).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For C42, the most cooperative fragment was 1; tethering
increased 2.3-fold, from 26% (apo) to 60% (protein-peptide
complex) (Figure 2a). A nearly identical compound, 2, was
identified in both screens, with 46% tethering to the apo 14-3-
3 and 59% tethering to the complex (Figure 2b). Compounds
Figure 2. a) LC/MS spectra of tethering screen results for 1 conjugated to 14-3-3(C42) apo (left) or ER-pp bound (right), resulting
in 26% and 60% tethering, respectively. 14-3-3(C42) expected mass: 26509 Da, ME capped mass: 26585 Da, protein-disulfide
conjugate mass: 26795 Da. b) LC/MS spectra of tethering screen results for 2 conjugated to 14-3-3(C42) apo (46%) or ER-pp
bound (59%); protein-disulfide conjugate mass: 26781 Da. c) LC/MS dose-response curves for 1 (top) and 2 (bottom) showing
percentage of fragment-protein conjugate formation for titrations of disulfides to 14-3-3(C42) apo ( / ) and bound to ER-pp ( ).
initially bound (Figure 3b). Fragments binding to the 14-3-
3(C42)/ER-pp complex increased the anisotropy, and hence
the bound fraction, of fluorescein-ER-pp. Additionally, from
these experiments we observed that the kinetics of disulfide
formation (i.e. stabilizing effect on ER-pp binding) was
dependent on the disulfide-fragment concentration, as
evidenced by an increase in anisotropy values over time for 0.1-
10 M (Figure S2). The maximum effect was instantaneous at
a saturating concentration (100 M). Compound 1 displayed
slightly more cooperative behavior, as reflected in a lower EC50
(87 nM) compared to 2 (EC50 = 209 nM). Notably, both were
slightly more potent compared to FC-A (EC50 = 216 nM).
The effect of 1 and 2 on the binding affinity between ER-pp
and 14-3-3(C42) was studied in fluorescence anisotropy
experiments. 14-3-3(C42) was titrated into fluorescein-
labeled ER-pp in the presence of DMSO or saturating
concentrations of fragments (Figure 3a). The apparent
dissociation constant of 14-3-3(C42)/ER-pp (Kd,app) was 1.3
M for the DMSO control, and decreased to 32 nM in the
presence of 1, 92 nM in the presence of fragment 2, and 4.2 nM
in the presence of the positive control FC-A. Thus, 1 and 2
stabilized the 14-3-3/ER-pp complex by 40- and 14-fold,
respectively.
We observed the same trend when we titrated the fragments
into a mixture of 1 M 14-3-3(C42) and 100 nM fluorescein-
ER-pp, conditions under which half of the peptide was
Whereas this cooperative behavior for 1 was expected based
on the initial criteria for hit selection, 2 was only slightly
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
cooperative in the primary screen. To evaluate whether the
could suggest that the electronic nature of the ring and/or the
stability of the ring orientation are critical for productive
interactions with both 14-3-3 and ER-pp.
1
2
3
4
5
6
7
8
single-concentration screen was reproducible, we further
evaluated ten additional fragment hits, including three selected
as ‘competitive’ from the screen (Figures S1, S3). While
moderate cooperativity was observed for most of the
‘cooperative’ fragments, the competitive and neutral fragments
generally had no effect on ER-pp binding, except for one
‘neutral’ fragment that modestly inhibited peptide binding.
Thus, while single-concentration screening yielded
reproducible cooperative fragments, screening at multiple doses
could be advantageous.
Comparison of hits from screening C42 and C45 revealed 6,
which differed from 2 only in linker length (propyl vs ethyl,
respectively) between the fragment and the disulfide-forming
thiol. We solved the structure of 6 conjugated to 14-3-3(C45)
in complex with ER-pp and found electron density for the
expected tethered fragment and parts of the linker (Figure 3e).
An overlay of 6-C45 with 2-C42 showed that the chloride
moiety was positioned in the same pocket of 14-3-3, but the
chlorophenyl ring was tilted so that the edge, rather than the
face, of the phenyl ring was pointed towards ER-pp V595.
Indeed, 6 displayed low cooperativity when bound to 14-3-
3(C45) (Figure S4). Interestingly, 6 bound to 14-3-3(C42)
showed similar cooperativity and binding affinity compared to
2 (EC50 value of < 1 M in the presence of ER-pp; Figure S4).
The convergent positioning of the C42- and C45-targeted
analogs 2 and 6 suggested that the fragments were selected
based on their compatibility with the pocket formed by 14-3-3
and ER-pp; however, the conformational restriction imposed
by the anchoring residue determined how productively the
fragment interacted with ER-pp.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To elucidate the molecular mechanism for cooperativity,
fragments were soaked into co-crystals of 14-3-3 an 8-mer
ER-pp. In addition to 1 and 2, electron density was resolved
for three other C42 hits (see Table S1 for XRD statistics).
Fragments 1-5 each contained an aromatic ring pointed into the
back of the 14-3-3 pocket, oriented to make hydrophobic
contact with the C-terminal V595 of ER-pp (Figure 3c+d).
The chlorophenyl substitutions in 1, 2, and 3 were fully buried
in the pocket. Whereas for 1, 2, and 4 continuous electron
density could be traced from the bound cysteine, 3 and 5 had
less complete density, perhaps suggesting disorder in their
linker region. Interestingly, while all fragments had a phenyl
ring in an analogous location, 3-5 showed significantly less
cooperativity compared to 1 and 2 (Figure S3a). These data
Figure 3. Quantification and mechanism of 14-3-3/ER-pp stabilization. a) Schematic of experimental design and plot of anisotropy (mean
+ SD) for 14-3-3(C42) titrations to fluorescein-ER-pp and saturating (100 M) 1 ( ), 2 ( ), FC-A ( ) or DMSO control ( ), reporting a
40-fold increase of the 14-3-3(C42)/ER-pp binding affinity in the presence of 1 (green arrow). b) Schematic of experimental design and
plot of anisotropy (mean + SD) for titrations of 1 ( ), 2 ( ), FC-A ( ) or DMSO ( ) to fluorescein-ER-pp and 1 M 14-3-3(C42). c+d) X-
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
ray crystal structures of fragments 1 - 5 (colored sticks) in complex with 14-3-3(C42) (white surface; C42 yellow) and ER-pp (green
sticks). e) Fragment 6 conjugated to C45 and overlay with 2 bound to C42. All 2Fo-Fc electron density maps contoured at 1.
1
2
3
4
5
6
7
8
9
Together, the data for disulfide hits 1, 2 and 6 supported the
hypothesis that the binding affinity of the fragments to the
protein-peptide complex was driven by non-covalent
interactions, which was further enhanced by the linker. To
confirm, we tested a non-covalent analogue for binding to 14-
3-3/ER-pp by ligand-observed NMR in T1 and waterLOGSY
experiments (Figure S5). T1 relaxation was significantly
enhanced in the presence of 14-3-3/ER-pp (Figure S5a) and a
positive wateLOGSY signal was seen in the presence, but not
in the absence, of 14-3-3/ER-pp (Figure S5b). These data
demonstrated that the fragment bound to the complex even in
the absence of a covalent linkage.
of 7 M and 3 M, respectively. By contrast, for ExoS or TAZ-
pp, the dose-response curve for binding of 2 was shifted to the
right. Even at 500 M, 2 just reached ~40% bound, compared
to 80% bound to apo 14-3-3(C42), and 100% bound to 14-3-
3(C42)/ER-pp or TASK3-pp (Figure 4b). The effect of 2 on
the binding affinity of 14-3-3(C42) for the different peptide
partners was further quantified by fluorescence anisotropy,
where 2 was titrated into a solution of fluorescently labeled
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TASK3-pp, ExoS, or TAZ-pp and 14-3-3(C42) at
a
concentration that allowed 20% binding of the peptide initially
(Anisotropy (r) of ~40 mAU; see Figure S6 for 14-3-3 binding
curves). FC-A and DMSO were included as controls. In
alignment with MS data, 2 increased the affinity between 14-3-
3 and TASK3-pp (EC50 = 2 M) and showed a destabilizing
effect on ExoS and TAZ-pp binding (IC50 = 1.4 M and 2M,
respectively) (Figure 4c). Interestingly, the maximal anisotropy
value for TASK3-pp was lower when 2 was titrated compared
to FC-A; this difference was not observed when 2 and FC-A
were titrated to ER-pp (Figure 3b). It might indicate a reduced
stabilization of the distal regions of TASK3-pp. Furthermore,
there was a 10-20 fold shift in the EC50 for the stabilizing vs
inhibiting effect of 2 on the binding of ER-pp (100-200 nM)
compared to TASK3-pp, ExoS and TAZ-pp (1-2 M),
indicating already partial selectivity for the hit fragment that can
be further exploited by chemical optimization.
Finally, to investigate the selectivity of disulfide fragments
for 14-3-3/ER-pp, we selected the binding motifs of ExoS and
TAZ as representative alternative 14-3-3 clients (Figure 4a).29,30
The TAZ phosphopeptide (TAZ-pp) extends through the
druggable pocket, thereby restricting the space for fragment
binding. ExoS is one of the few reported non-phosphorylated
clients of 14-3-3, and also occupies almost the full length of the
amphipathic groove, including the target pocket. We also
included TASK3, which contains a C-terminal phosphoSV
nearly identical to the phosphoTV motif in ER.31 In dose-
response analysis by MS, a shift to the left was observed for the
binding curve of 2 in the presence of TASK3 phosphopeptide
(TASK3-pp) compared to apo 14-3-3(C42), indicating a
similar ability for 2 to bind cooperatively to the 14-3-
3(C42)/TASK3-pp and ER-pp complexes, with EC50 values
Figure 4. Selectivity of hit fragment 2. a) Schematic illustration of the challenge posed by achieving selectivity in PPI network modulation.
b) Dose-response curves obtained by MS, analyzing % tethering for titrations of 2 to 14-3-3 apo (- peptide; ) or bound to different
interaction partner-derived peptide motifs; ER-pp ( ), TASK3-pp ( ), ExoS ( ) or TAZ-pp ( ), starting from 1 mM. c) Overlays of crystal
structures of 14-3-3 (white surface) bound by 2 (Figure 3c), and TASK3-pp (PDB: 3P1N), ExoS (PDB: 2O02) or TAZ-pp (PDB: 5N75)
illustrating (in)compatibility of binding surface areas. Fragment (blue) and peptides (green) in space-filling representation. Fluorescence
anisotropy data (mean + SD; triplicates) and non-linear fit for titration of 2 ( ) to 14-3-3. FC-A ( ) and DMSO ( ) are included as controls.
Conclusions
strategies. The major and unmet challenge in this field is the
lack of starting points for small-molecule stabilizer
development. In contrast to conventional screening techniques,
Small-molecule PPI stabilization has diverse therapeutic
applications, justifying the pursuit of novel drug discovery
ACS Paragon Plus Environment

Journal of the American Chemical Society
we find disulfide trapping to be highly suitable for early strategies to optimize fragments into chemical probes and/or
Page 6 of 9
1
2
3
4
5
6
7
8
stabilizer discovery, likely because the technology is site-
directed and the disulfide bond allows the fragment to fully
saturate the binding site. We have validated the disulfide
screening paradigm by selecting fragments that enhance the
interaction between 14-3-3 and an ER-derived
phosphopeptide (ER-pp) and crystallized these fragments to
learn the molecular requirements to achieve stabilization.
drug leads.
The disulfide trapping strategy can be generalized to any 14-
3-3/client pair. In addition to ER, several other important
transcription factors, including TAZ, Myc, RelA, and FOXO-1,
are clients of 14-3-3, and this approach could conceivably
develop modulators of multiple transcription factors.
Furthermore, small molecules might be able to induce unnatural
14-3-3/protein complexes, allowing the exploration of synthetic
biology. As a site-directed binding methodology, disulfide
trapping is an ideal technology for such a platform approach.
Systematic discovery of novel PPI stabilizers has the potential
to access ‘undruggable’ targets and provide opportunities for
intervention in previously inaccessible pathways.
Disulfide-bound fragments bind cooperatively with ER-pp
to 14-3-3(C42), providing as much as a 40-fold increase in
affinity for the 14-3-3(C42)/ER-pp complex. Both the
binding affinity and the degree of PPI stabilization depend on
the chemical structure of the fragments and their orientation in
the binding site. In particular, stabilization of the 14-3-3/ER-
pp complex correlated with the presence of para-chlorophenol
ring oriented with its face towards the terminal valine of the
peptide. Taken together, the biochemical data and crystal
structures support the hypothesis that binding of the fragments
was driven by the non-covalent interactions with the protein-
peptide complex, which was confirmed in ligand-observed
NMR experiments for a non-covalent analogue.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Protein expression and purification. The 14-3-3  isoform
with a truncated C-terminus after T231 (C; to enhance
crystallization) and an N-terminal His6-tag was expressed in
NiCo21 (DE3) competent E.coli (New England biolabs Inc)
from
a pPROEX HTb expression vector. Site-directed
Towards development of the platform, we compared
differential hits from three cysteine constructs of 14-3-3 and
observed that the appropriate stringency of screening is
essential for selecting fragments can engage the targeted pocket.
In addition to the differences in the degree of tethering, we also
observed that different types of hits (i.e. cooperative, neutral,
competitive towards peptide binding) were more likely at
different positions. Whereas we initially were very stringent
towards selecting cooperative hits for follow-up, we found that
‘neutral’ hits displaying high intrinsic affinity for protein could
also induce a cooperative effect when studied in more detail.
Hence, future efforts could include multiple-dose screening to
maximize the window between binding to 14-3-3 and to 14-3-
3/peptide complexes.
mutagenesis to obtain double mutants C38N/N42C and
C38N/S45C was performed using the QuickChange Lightening
site-directed mutagenesis kit (Agilent Technologies) following
manufacturer’s instructions. C38N was selected since
asparagine is the most prevalent amino acid at that position
across the 14-3-3 family. Primer sequences are listed in SI Table
S2. Constructs were confirmed by DNA sequencing. After
transformation following manufacturer’s instructions, single
colonies were picked to inoculate 30 mL pre-cultures (LB),
which were added to 1.5 L 2XYT medium after overnight
growth at 37°C, 250 rpm. Expression was induced upon
reaching OD600 0.5-0.6 by adding 400 M IPTG. After
overnight expression at 18°C, 140 rpm, cells were harvested by
centrifugation at 8000 rpm and resuspended in lysis buffer (50
mM Tris, pH 8.0, 300 mM NaCl, 10 mM imidazole, 5 mM
MgCl2, 1 mM PMSF, 250 M TCEP). The His6-tagged proteins
were first purified by Ni-affinity chromatography (HisTrap HP
column, GE) (Elution buffer 50 mM Tris, pH 8.0, 300 mM
NaCl, 250 mM imidazole, 250 M TCEP), followed by His-tag
cleavage by TEV protease during dialysis (25 mM HEPES pH
7.5, 200 mM NaCl, 5% glycerol, 10 mM MgCl2, 250 M TCEP)
overnight at 4°C. The flow-through of a second HisTrap column
was subjected to final purification step by size-exclusion
chromatography (Superdex75, GE) (SEC buffer 25 mM HEPES
pH 7.5, 100 mM NaCl, 10 mM MgCl2, 250 M TCEP). The
protein was concentrated to ~60 mg/mL, analyzed for purity by
SDS-PAGE and Q-Tof LC/MS and aliquots flash-frozen for
storage at -80°C.
One challenge when targeting PPI of proteins with many
binding partners, such as 14-3-3, is client selectivity.
Opportunities for selectivity result from the significant
variation in phosphoprotein sequences and 14-3-3-binding
modes, giving rise to differences in the protein-protein interface
that small molecules could exploit. As demonstration, 2 shows
the strongest cooperativity towards ER-pp, secondly towards
the ER-like-peptide TASK3-pp, and at higher concentrations
also partially influences the structurally unrelated 14-3-3/TAZ-
pp or ExoS protein-peptide complexes. These differences could
be further exploited by optimizing contacts with ER-pp and
tuning the ratio of intrinsic binding of the fragments to apo-14-
3-3 versus binding to the 14-3-3/peptide complex.
It is important to note that 14-3-3 client proteins are much
larger than the peptides studied here. However, 14-3-3 proteins
exert their regulatory role specifically via phosphorylation-
induced PPIs, and the phosphate group on a binding partner is
usually the primary driver of the binding affinity. Therefore,
even in the context of differential secondary interactions, a
stabilizing effect on this primary interaction site will result in
an overall increased stability of the full-length protein complex.
To fully validate the utility of our fragments will require
chemical optimization and characterization of the PPI in a
biological environment. The principle innovation of these
studies is the systematic platform for discovery of PPI
stabilizing fragments that are then suitable for tried-and-true
Peptide sequences. Peptides for disulfide trapping were
purchased from Elim Biopharmaceuticals, Inc. (Hayward, CA)
Sequences were as follows: Ac-KYYITGEAEGFPA{pT}V-
COOH (ER-pp); Ac-RRK{pS}V-COOH (TASK3-pp); Ac-
RSH{pS}SPASLQLGT-CONH2
(TAZ-pp);
Ac-
SGHGQGLLDALDLAS-CONH2 (ExoS). ER-pp for X-ray
crystallography and fluorescein-labeled peptides were ordered
from GenScript Biotech Corp. Sequences were: Ac- or 5-FAM-
AEGFPA{pT}V-COOH (8mer ER-pp) and 5-FAM-labeled
sequences as above for TASK3-pp, TAZ-pp and ExoS.
Disulfide Tethering screening and data processing. The
primary screening was performed by incubating the target with
individual compounds in a 384-well plate format. A custom
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
The Supporting Information is available free of charge on the ACS
Publications website.
library of 1600 disulfide-containing fragments of the UCSF
1
2
3
4
5
6
7
8
Small Molecule Discovery Center (SMDC), synthesized as
previously reported, was available as 50 mM stock solutions in
DMSO.32,33 For screening, 14-3-3 wild-type and Cys-mutants
were diluted to 100 nM in buffer (10 mM Tris, 100 M
betamercaptoethanol (ME), pH 8.0) and plated in 384-well
plates (15 L/well). 30 nL of each fragment was pinned from
the library master plates into the protein samples using a
Biomek FX (Beckman) to give a final concentration of 100 M.
The duplicate experiments additionally contained 200 nM ER-
pp. The reaction mixtures were incubated for 3 hours at RT
before being subjected to LC/MS (I-class Acquity UPLC / Xevo
G2-XS Quadrupole Time of Flight mass spectrometer, Waters).
Data collection and automated processing followed a custom
workflow, as previously described.28 All compounds described
in the text were from the same lot as the original screening
material.
XRD experimental details, data collection and refinement statistics;
synthetic procedure; NMR spectroscopy methods; primer
sequences for mutagenesis; deconvoluted MS screening data;
anisotropy titration curves for 1 – 5 and seven additional screening
hits; dose-response data (MS and anisotropy) for 6; ligand-
observed NMR spectra; and 14-3-3 titration curves to various
binding partners. (Tables S1-S2 and Figures S1-S6) (PDF)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Authors
*c.ottmann@tue.nl
*michelle.arkin@ucsf.edu
Author Contributions
Dose-Response LC/MS experiments. Disulfide tethering
dose-response analysis used the same procedures as primary
screening, with the exception that the ME concentration was 1
mM, and compounds were titrated from 5-50 mM in 2-fold
serial dilutions in DMSO, then 400 nL of the compound was
transferred to 10 L protein solution for final concentrations
0.1-2000 µM and 4% DMSO. For the dose-response of 2 in the
presence of TAZ-pp (150 M), a 5 minute chromatography step
was employed to separate the hydrophobic peptide from the 14-
3-3 before ionization.
⊥These authors contributed equally to this work.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENTS
We thank the Renslo laboratory for synthesis of the SMDC
disulfide library, Connie Merron and Julia Davies for automated
MS data processing infrastructure at the SMDC, Jim Wells and
Adam Renslo for helpful discussions. Joris Adriaans and Dongju
Wang are thanked for synthesis and technical support. Bert de Boer
is kindly acknowledged for his contributions in the initial stage of
the 14-3-3/ER project.
The research described was funded by the National Institutes of
Health (R01AG044515); the Initial Training Network TASPPI
(H2020 Marie Curie Actions of the European Commission, Grant
Agreement 675179); and the Netherlands Organization for
Scientific Research (NWO) (Gravity Program 024.001.035 and
Vici grant 016.150.366). E.S. acknowledges the Nora Baart
Foundation (NVBMB; KNCV) and the Hendrik Muller program.
Fluorescence Anisotropy. Fluorescein-labeled peptides, 14-
3-3 protein, FC-A (10 mM stock solution in DMSO) and
disulfide fragments (50 mM stock solutions in DMSO) were
diluted in buffer (10 mM HEPES pH 7.5, 150 mM NaCl, 0.1%
TWEEN-20, 1 mg/mL Bovine Serum Albumine (BSA; Sigma
Aldrich)). Final DMSO concentration in the assay was always
1%. Dilution series of 14-3-3 protein or fragments were made
in black, round-bottom 384-microwell plates (Corning) in a
final sample volume of 10 L in triplicates. Fluorescence
anisotropy measurements were performed directly and after
overnight incubation at room-temperature, using a Tecan
Infinite F500 plate reader (filter set ex: 485±20 nm, em: 535±25
nm). Data reported are at endpoint. EC50 values were obtained
from fitting the data with a four-parameter logistic model (4PL)
in GraphPad Prism 6.
REFERENCES
1.
2.
3.
4.
Berg, T. Modulation of protein-protein interactions with small
organic molecules. Angew. Chem. Int. Ed Engl. 42, 2462–81
(2003).
Wells, J. A. & McClendon, C. L. Reaching for high-hanging fruit
in drug discovery at protein-protein interfaces. Nature 450, 1001–
9 (2007).
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).
Scott, D. E., Bayly, A. R., Abell, C. & Skidmore, J. Small
molecules, big targets: drug discovery faces the protein-protein
interaction challenge. Nat. Rev. Drug Discov. 15, 533–550
(2016).
X-Ray Crystallography. 14-3-3 protein (470 M; 12.5
mg/mL) was mixed with ER-pp (1:2 molar stochiometry; 940
M) and incubated in buffer (20 mM HEPES pH 7.4, 2 mM
MgCl2, 2 mM ME overnight at 4°C before setting up for sitting
drop crystallization in MRC crystallization plates (Swissci)
with a custom crystallization liquor-grid (0.095 M HEPES (pH
7.1, 7.3, 7.5, 7.7), 0.19 M CaCl2, 5% glycerol, 24-29% PEG
400). Crystals grew at 4°C within 4 days. Soaking of crystals
was performed by mixing 0.4 L disulfide fragments from 50
mM stock solutions in DMSO with 2 mM ME in 3.6 L mother
liquor, and adding this to crystal-containing drops. Soaked
crystals were fished after overnight incubation and flash-frozen
in liquid nitrogen. Data collection and processing are described
in Supporting Information.
5.
6.
Fischer, G., Rossmann, M. & Hyvönen, M. Alternative
modulation of protein-protein interactions by small molecules.
Curr. Opin. Biotechnol. 35, 78–85 (2015).
Andrei, S. A., Sijbesma, E., Hann, M., Davis, J., O'Mahony, G.,
Perry, M. W. D., Karawajczyk, A., Eickhoff. J., Brunsveld, L.,
Doveston, R. G., Milroy, L.-G. & Ottmann, C. Stabilization of
protein-protein interactions in drug discovery. Expert Opin. Drug
Discov. 12, 925–940 (2017).
ASSOCIATED CONTENT
Supporting Information
7.
Zarzycka, B., Kuenemann, M. A., Miteva, M. A., Nicolaes, G. A.
F., Vriend, G. & Sperandio, O. Stabilization of protein–protein
interaction complexes through small molecules. Drug Discov.
Today 21, 48–57 (2016).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
8.
9.
Waring, M. J., Chen. H., Rabow, A. A., Walker, G., Bobby, R.,
20. Zhao, J., Meyerkord, C. L., Du, Y., Khuri, F. R. & Fu, H. 14-3-3
proteins as potential therapeutic targets. Semin. Cell Dev. Biol.
22, 705–712 (2011).
21. Aghazadeh, Y. & Papadopoulos, V. The role of the 14-3-3 protein
family in health, disease, and drug development. Drug Discov.
Today 21, 278–287 (2016).
22. Stevers, L. M., Sijbesma, E., Botta, M., MacKintosh, C., Obsil,
T., Landrieu, I., Cau, Y., Wilson, A. J., Karawajczyk, A.,
Eickhoff, J., Davis, J., Hann, M., O'Mahony, G., Doveston, R. G.,
Brunsveld, L. & Ottmann, C. Modulators of 14-3-3 Protein–
Protein Interactions. J. Med. Chem. 61, 3755–3778 (2018).
23. Erlanson, D. A., Wells, J. A. & Braisted, A. C. Tethering:
fragment-based drug discovery. Annu. Rev. Biophys. Biomol.
Struct. 33, 199–223 (2004).
Boiko, S., Bradbury, R. H., Callis, R., Clark, E., Dale, I., Daniels,
D. L., Dulak, A., Flavell, L., Holdgate, G., Jowitt, T. A., Kikhney,
A., McAlister, M., Méndez, J., Ogg, D., Patel, J., Petteruti, P.,
Robb, G. R., Robers, M. B., Saif, S., Stratton, N., Svergun, D. I.,
Wang, W., Whittaker, D., Wilson, D. M. & Yao, Y. Potent and
selective bivalent inhibitors of BET bromodomains. Nat. Chem.
Biol. 12, 1097–1104 (2016).
Bulawa, C. E., Connelly, S., Devit, M., Wang, L., Weigel, C.,
Fleming, J. A., Packman, J., Powers, E. T., Wiseman, R. L., Foss,
T. R., Wilson, I. A., Kelly, J. W. & Labaudinière R. Tafamidis, a
potent and selective transthyretin kinetic stabilizer that inhibits
the amyloid cascade. Proc. Natl. Acad. Sci. U. S. A. 109, 9629–
9634 (2012).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10. Hughes, S. J. & Ciulli, A. Molecular recognition of ternary
complexes: a new dimension in the structure-guided design of
chemical degraders. Essays Biochem. 61, 505–516 (2017).
11. Erlanson, D. A., Braisted, A. C., Raphael, D. R., Randal, M.
Stroud, R. M., Gordon, E. M. & Wells, J. A. Site-directed ligand
discovery. Proc. Natl. Acad. Sci. U. S. A. 97, 9367–9372 (2000).
12. Ostrem, J. M., Peters, U., Sos, M. L., Wells, J. A. & Shokat, K.
M. K-Ras(G12C) inhibitors allosterically control GTP affinity
and effector interactions. Nature 503, 548–51 (2013).
13. Erlanson, D. A., Arndt, J. W., Cancilla, M. T., Cao, K., Elling, R.
A., English, N., Friedman, J., Hansen, S. K., Hession, C., Joseph,
I., Kumaravel, G., Lee, W. C., Lind, K. E., McDowell, R. S.,
Miatkowski, K., Nguyen, C., Nguyen, T. B., Park, S., Pathan, N.,
Penny, D. M., Romanowski, M. J., Scott, D., Silvian, L.,
Simmons, R. L., Tangonan, B. T., Yang, W. & Sun, L. Discovery
of a potent and highly selective PDK1 inhibitor via fragment-
based drug discovery. Bioorg. Med. Chem. Lett. 21, 3078–3083
(2011).
14. Hyde, J., Braisted, A. C., Randal, M. & Arkin, M. R. Discovery
and characterization of cooperative ligand binding in the adaptive
region of interleukin-2. Biochemistry 42, 6475–6483 (2003).
15. Raimundo, B. C., Oslob, J. D., Braisted, A. C., Hyde, J.,
McDowell, R. S., Randal, M., Waal, N. D., Wilkinson, J., Yu, C.
H. & Arkin, M. R. Integrating fragment assembly and biophysical
methods in the chemical advancement of small-molecule
antagonists of IL-2: an approach for inhibiting protein-protein
interactions. J. Med. Chem. 47, 3111–3130 (2004).
16. Molzan, M. & Ottmann, C. Synergistic Binding of the
Phosphorylated S233- and S259-Binding Sites of C-RAF to One
14-3-3ζ Dimer. J. Mol. Biol. 423, 486–495 (2012).
17. Sluchanko, N. N., Beelen, S., Kulikova, A. A., Weeks, S. D.,
Antson, A. A., Gusev, N. B. & Strelkov, S. V. Structural basis for
the interaction of a human small heat shock protein with the 14-
3-3 universal signaling regulator. Struct. Lond. Engl. 1993 25,
305–316 (2017).
18. De Vries-van Leeuwen, I. J., Da Costa Pereira, D., Flach, K. D.,
Piersma, S. R., Haase, C., Bier, D., Yalcin, Z., Michalides, R.,
Feenstra, K. A., Jiménez, C. R., De Greef, T. F., Brunsveld, L.,
Ottmann, C., Zwart, W. & De Boer, A. H. Interaction of 14-3-3
proteins with the estrogen receptor alpha F domain provides a
drug target interface. Proc. Natl. Acad. Sci. U. S. A. 110, 8894–9
(2013).
19. Schumacher, B., Mondry, J., Thiel, P., Weyand, M. & Ottmann,
C. Structure of the p53 C-terminus bound to 14-3-3: Implications
for stabilization of the p53 tetramer. FEBS Lett. 584, 1443–1448
(2010).
24. Pennington, K. L., Chan, T. Y., Torres, M. P. & Andersen, J. L.
The dynamic and stress-adaptive signaling hub of 14-3-3:
emerging mechanisms of regulation and context-dependent
protein–protein interactions. Oncogene 37, 5587 (2018).
25. Zhao, J., Du, Y., Horton, J. R., Upadhyay, A. K., Lou, B., Bai,
Y., Zhang, X., Du, L., Li, M., Wang, B., Zhang, L., Barbieri, J.
T., Khuri, F. R., Cheng, X. & Fu, H. Discovery and structural
characterization of a small molecule 14-3-3 protein-protein
interaction inhibitor. Proc. Natl. Acad. Sci. U. S. A. 108, 16212–
16216 (2011).
26. Milroy, L.-G., Bartel, M., Henen, M. A., Leysen, S., Adriaans, J.
M., Brunsveld, L., Landrieu, I. & Ottmann, C. Stabilizer-Guided
Inhibition of Protein–Protein Interactions. Angew. Chem. Int. Ed.
54, 15720–15724 (2015).
27. Andrei, S. A., Meijer, F. A., Neves, J. F., Brunsveld, L., Landrieu,
I., Ottmann, C. & L.-G. Milroy. Inhibition of 14-3-3/Tau by
Hybrid Small-Molecule Peptides Operating via Two Different
Binding Modes. ACS Chem. Neurosci. 9, 2639–2654 (2018).
28. Hallenbeck, K. K., Davies, J. L., Merron, C., Ogden, P.,
Sijbesma, E., Ottmann, C., Renslo, A. R., Wilson, C. & Arkin, M.
R. A Liquid Chromatography/Mass Spectrometry Method for
Screening Disulfide Tethering Fragments. SLAS Discov. 23, 183–
192 (2018).
29. Ottmann, C., Yasmin, L., Weyand, M., Veesenmeyer, J. L., Diaz,
M. H., Palmer, R. H., Francis, M. S., Hauser, A. R., Wittinghofer,
A. & Hallberg, B. Phosphorylation-independent interaction
between 14-3-3 and exoenzyme S: from structure to
pathogenesis. EMBO J. 26, 902–913 (2007).
30. Sijbesma, E., Skora, L., Leysen, S., Brunsveld, L., Koch, U.,
Nussbaumer, P., Jahnke, W. & Ottmann, C. Identification of Two
Secondary Ligand Binding Sites in 14-3-3 Proteins Using
Fragment Screening. Biochemistry 56, 3972–3982 (2017).
31. Anders, C., Higuchi, Y., Koschinsky, K., Bartel, M.,
Schumacher, B., Thiel, P., Nitta, H., Preisig-Müller, R.,
Schlichthörl, G., Renigunta, V., Ohkanda, J., Daut, J., Kato, N. &
Ottmann, C. A Semisynthetic Fusicoccane Stabilizes a Protein-
Protein Interaction and Enhances the Expression of K+ Channels
at the Cell Surface. Chem. Biol. 20, 583–593 (2013).
32. Burlingame, M. A., Tom, C. T. M. B. & Renslo, A. R. Simple
One-Pot Synthesis of Disulfide Fragments for Use in Disulfide-
Exchange Screening. ACS Comb. Sci. 13, 205–208 (2011).
33. Turner, D. M., Tom, C. T. M. B. & Renslo, A. R. Simple Plate-
Based, Parallel Synthesis of Disulfide Fragments using the
CuAAC Click Reaction. ACS Comb. Sci. 16, 661–664 (2014).
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
1
2
3
Table of Contents graphic (TOC)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
ACS Paragon Plus Environment