Affinity-based labeling of cytohesins with a bifunctional SecinH3 photoaffinity probe

Article abstract of DOI:10.1002/anie.200803962

(Chemical Equation Presented) The specific binding of the cytohesin inhibitor SecinH3 to Sec7 domains of cytohesins was shown by the use of a bifunctional photoaffinity probe (see structure) on guanine nucleotide exchange factors (GEFs) and GTPases. This probe should be useful for the proteome-wide profiling of cytohesin complexes and the identification of the binding site.

Full text of DOI:10.1002/anie.200803962

Angewandte
Chemie
DOI: 10.1002/anie.200803962
Photoaffinity Labeling
Affinity-Based Labeling of Cytohesins with a Bifunctional SecinH3
Photoaffinity Probe**
Xihe Bi, Anton Schmitz, Alaa M. Hayallah, Jin-Na Song, and Michael Famulok*
Dedicated to Professor Manfred T. Reetz on the occasion of his 65th birthday
We recently identified the 1,2,4-triazole derivative SecinH3
(see Scheme 1) as the first small-molecule inhibitor of a class
of cytoplasmic regulatory proteins called cytohesins by using
an aptamer-displacement assay.[1,2] Cytohesins are small
guanine nucleotide exchange factors (GEFs) that stimulate
ADP (adenosine diphosphate) ribosylation factors (Arfs),
ubiquitously expressed Ras-like GTPases which control
various cellular regulatory networks ranging from vesicle
trafficking to integrin activation.[3] The four highly homolo-
gous cytohesins known in mammals share the same domains:
a Sec7 domain, which harbors the GEF activity, a pleckstrin-
homology (PH) domain, and a coiled-coil (CC) domain.
SecinH3 targets the Sec7 domain of cytohesins 1, 2 (also
known as “Arf nucleotide-binding-site opener” (ARNO)),
and 3, and inhibits their guanine-nucleotide-exchange activ-
ity. Its application in human liver cells, flies, and mice led to
impaired insulin signaling in each case. Thus, cytohesins
associated with the insulin-receptor complex appear to be
essential components of this central cellular signaling path-
way.[1a,4]
ligands.[6c,d] Reporter-tagged derivatives of small molecules
can also serve as probes for activity-based protein profiling
(ABPP).[7]
Herein we report the design, synthesis, and application of
SecinH3 photoaffinity probes. By applying them to a wide
range of GEFs and their small GTPase substrates, we
demonstrate that Secin derivatives, which enable the direct
detection of the covalent linkage between the photoaffinity
probe and the cytohesin Sec7 domain, exhibit high specificity
for cytohesins.
Preliminary structure–activity-relationship (SAR) studies
of SecinH3 showed that the terminal thiophenyl group is
critical for inhibitory activity.[8] The deletion of this residue in
XH1009 (Scheme 1) resulted in a dramatic loss of inhibition
Two classes of Arf-GEFs are known and can be distin-
guished by their size. Cytohesins belong to the small 47 kDa
Arf-GEFs; the large Arf-GEFs, which also contain a Sec7
domain, have a mass of about 200 kDa.[5] Unlike the large
Arf-GEFs, the small SecinH3-responsive cytohesins are
insensitive to the fungal metabolite brefeldin A (BFA). The
specificity of SecinH3 for cytohesin GEFs has so far been
tested only for a limited set of proteins by means of
isothermal titration calorimetry (ITC).[1a]
A powerful method for assessing the specificity of small-
molecule modulators for their targets is photoaffinity label-
ing,[6] as it enables the analysis of a large number of individual
proteins or protein mixtures in parallel.[6a,b] The combination
of photoaffinity labeling and mass spectrometry has been
established as an efficient approach for the identification of
the binding site of modulator compounds to their protein
[*] Dr. X. Bi, Dr. A. Schmitz, Dr. A. M. Hayallah, Dr. J.-N. Song,
Prof. Dr. M. Famulok
LIMES Institute Program Unit Chemical Biology &
Medicinal Chemistry, Universitꢀt Bonn
Gerhard-Domagk-Strasse 1, 53121 Bonn (Germany)
Fax: (+49)228-735-388
Scheme 1. Synthesis of the Secin derivatives used: a) KSCN, acetone,
608C; b) R1OH, acetone, 508C, 83%; c) 4-nitrophenylhydrazine, EtOH,
808C, 70%; d) Fe/HCl, NH4Cl, MeOH/H2O, 658C, 64%;
E-mail: m.famulok@uni-bonn.de
e) R2CH2COOH, HBTU/NEt3, DMF, 228C, 57%; f) TFA/CH2Cl2, 94%;
g) d-(+)-biotin, HBTU/NEt3, DMF, 69%. Boc=tert-butoxycarbonyl,
DMF=N,N-dimethylformamide, HBTU=O-benzotriazole-N,N,N’,N’-
tetramethyluronium hexafluorophosphate, TFA=trifluoroacetic acid.
[**] This research was supported by the SFBs 645 and 704 (M.F.) and by
the Alexander von Humboldt Foundation (X.B.).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803962.
Angew. Chem. Int. Ed. 2008, 47, 9565 –9568
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9565

Communications
of Sec7-catalyzed guanine-nucleotide exchange on [D17]Arf1
(Figure 1). In contrast, modification of the phenyl ring had
little effect on the inhibitory activity, or was even beneficial.
with Bio-SecinPP (5 mm) at ambient temperature for 2–3 min.
Afterwards, the samples were subjected to UV irradiation at
366 nm for varying amounts of time (0–5 min) at an energy
output of 200 mW. After separation of the proteins by 12.5%
SDS-PAGE, the modified protein was detected with a horse-
radish peroxidase–streptavidin conjugate. As all proteins used
in this study contained a hexahistidine (His6) tag, a His6-
specific antibody was used to normalize the amount of loaded
protein. The amount of labeled protein increased steadily as a
function of time; after irradiation for 3 min, the amount of
labeled protein was 30 times greater than that observed for
the nonirradiated background (Figure 2B). The photoaffinity
labeling at a fixed concentration of the protein (2.5 mm) and
constant UV irradiation for 3 min was found to be concen-
tration-dependent when Bio-SecinPP was used at concentra-
tions between 0 and 5 mm (Figure 2C). When Bio-SecinH3
(2.5 mm) was subjected to the same conditions, the signal
intensity was similar to that of the background signal
(Figure 2C, lane 7), which shows that the labeling depends
on the presence of the photophore. We also performed a
competition experiment (Figure 2D) to verify that the bind-
ing mode of Bio-SecinPP was similar to that of SecinH3. For
this purpose, SecinH3 (2.5 mm) was first incubated with
ARNO-Sec7 for 3 min at ambient temperature, and then
the same amount of Bio-SecinPP was added; the resulting
mixture was irradiated for 1 min and then analyzed by
western blotting. In the presence of an equimolar amount of
SecinH3, the amount of labeled protein decreased signifi-
cantly (Figure 2D, lane 4). In contrast, XH1009 at this
concentration did not affect the efficiency of photolabeling
Figure 1. Inhibition by SecinH3, SecinPP, and XH1009 of guanine-
nucleotide exchange on [D17]Arf1 catalyzed by the Sec7 domain of
cytohesin 2 (ARNO; tryptophan fluorescence assay).[10] c=inhibitor
concentration, va,rel =relative initial rate.
To convert SecinH3 into a photoaffinity probe, we took
advantage of this finding by substituting a photolabile
benzophenone (Bp) group for the terminal phenyl ring, as
in SecinPP (Scheme 1). The advantages of Bp as the photo-
phore are its chemical stability, specificity of insertion sites
during photoactivation, and ability to be excited repeatedly
by UV light without loss of reactivity.[9] The attachment of the
biotin group with the aminoethoxyethanol spacer in
Bio-SecinPP (Scheme 1) was based on preliminary SAR
studies, according to which the methoxy group at the 3-
position of the triazole ring can be extended without
significant loss of inhibitory activity.[8] As a control, we
synthesized biotin-labeled Bio-SecinH3.
The synthesis of the Secin derivatives (Scheme 1)
begins with the transformation of the benzodioxole
carboxylic chloride 1 into the thiocarbamoyl acid O-
alkyl ester 2 with KSCN and the respective alcohol.
Treatment with 4-nitrophenylhydrazine and subsequent
reduction afforded the 1,2,4-triazole compounds 3. The
attachment of phenylsulfanyl- or benzophenonesulfa-
nylacetic acid derivatives onto 3a gave SecinH3 and
SecinPP, respectively, whereas compounds 4 and 5,
respectively, were obtained by the attachment of the
same derivatives onto 3b. The acetylation of 3a afforded
XH1009. The deprotection of 4 and 5, followed by
acylation with d-(+)-biotin, gave Bio-SecinH3 and Bio-
SecinPP in good yields.
SecinPP showed a similar inhibition curve to that of
SecinH3 in the ARNO Sec7-catalyzed GDP/GTP-
exchange assay on [D17]Arf1, which suggests that the
two compounds bind the ARNO Sec7 domain similarly.
Compound XH1009 was used as a negative control
(Figure 1). Next, we examined the concentration- and
time-dependent efficiency of the labeling of ARNO-
Sec7 by Bio-SecinPP. For this purpose, the Sec7 domain
Figure 2. A) Structures of Bio-SecinPP and Bio-SecinH3. B),C) Photoaffinity
labeling of ARNO-Sec7 with Bio-SecinPP with increasing irradiation time (B)
and increasing concentration (C). D) Competition experiment with an equal
amount of SecinH3 or XH1009.
of ARNO was incubated at a concentration of 2.5 mm
9566 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9565 –9568

Angewandte
Chemie
by Bio-SecinPP (Figure 2D, lane 5). These results indicate
that Bio-SecinPP and SecinH3 bind to ARNO at the same
site.
Having verified that Bio-SecinPP could be used as a
bifunctional probe, we examined the labeling specificity by
using a range of Arf-GEF Sec7 domains, GEFs without Sec7
domains, and small GTPases (Figure 3). The Sec7 domains of
Experimental Section
The photoaffinity labeling of ARNO-Sec7 with Bio-SecinPP in a
time-course experiment is described exemplarily: Bio-SecinPP
(2.5 mm) was incubated with ARNO-Sec7 (2.5 mm) in phosphate-
buffered saline (PBS; pH 7.4, 50 mL) with 1% dimethyl sulfoxide at
238C for 2–3 min. The mixture was then transferred into a glass tube
(50 ꢀ 10 mm, thickness: 0.65 mm) and irradiated with UV light
(home-built) for 0–5 min at 366 nm and 100 mW. After UV irradi-
ation, the proteins were denatured by adding 6X
loading buffer and heating for 5 min at 968C. The
proteins were separated by 12.5% SDS-PAGE and
subsequently blotted onto a nitrocellulose mem-
brane (Whatman). For biotin detection, the mem-
brane was treated for 1 h with TBST (Tris-buffered
saline 0.1% Tween (Sigma); pH 7.6, 10 mL; Tris =
tris(hydroxymethyl)aminomethane)
and
5%
bovine serum albumin (BSA) at room temper-
ature, washed twice with TBST (10 mL), and then
incubated with horseradish peroxidase (HRP)
conjugated streptavidin (1:20000; Rockland) in
5% BSA/TBST (10 mL) at 48C for 16 h. For
Figure 3. Photoaffinity labeling of GEF functional domains and small GTPases (2.5 mm)
with Bio-SecinPP (2.5 mm). Irradiation: 100 mW, 1 min.
detection of the His6 tag, the membrane was
blocked in 5% nonfat powdered milk in TBST at
cytohesins 1 and 3 and of the Drosophila melanogaster
cytohesin Steppke (Grp1 Sec7) were all labeled in the same
way as ARNO-Sec7 with high efficiency after treatment with
Bio-SecinPP and UV irradiation for 1 min. Full-length
ARNO and several truncated ARNO constructs, including
ARNO-Sec7-PH and ARNO-CC-Sec7, were labeled as
efficiently as the Sec7 domain alone. Thus, the other cytohesin
domains did not affect the labeling reaction. The efficient
labeling of full-length ARNO indicates that Bio-SecinPP is a
promising ABPP probe for cytohesins.
238C for 1 h, and then incubated at 48C for 16 h with a His6-specific
antibody (1:2000; Santa Cruz Biotechnology). After extensive wash-
ing, the membrane was incubated at 238C for 1 h with a goat
antimouse IgG conjugated to HRP (Santa Cruz Biotechnology;
IgG = immunoglobulin G). Proteins were visualized by enhanced
chemiluminescence (Millipore). Experiments were performed in
duplicate.
Received: August 11, 2008
Published online: October 29, 2008
The Sec7 domains of the medium and large Arf-GEFs,
such as “exchange factor for Arf6” (EFA6), “brefeldin A
inhibited guanine nucleotide-exchange protein 1” (Big1), and
“guanine nucleotide exchange factor for Arfs 2” (Gea2) from
yeast showed no or only background modification. Likewise,
the Sec7-unrelated “Dbl homology” (DH) domain of the Rho
GEF Vav1 did not react with Bio-SecinPP. The DH–PH
domain of the Rac GEF “T-lymphoma invasion and meta-
stasis-inducing protein 1” (Tiam1) and the CDC25-like GEF
domain of the Ras GEF “son of sevenless” (Sos) exhibited a
slight nonspecific interaction with the peroxidase–streptavi-
din conjugate. This interaction was not dependent on
irradiation (Figure 3) or Bio-SecinPP (data not shown). The
small GTPases Arf1, Arf6, and Ras were not labeled with
Bio-SecinPP. Taken together, these results demonstrate that
the labeling by Bio-SecinPP is cytohesin-specific.
In conclusion, we have designed, synthesized, and applied
a photoactive probe for cytohesins by integrating a benzo-
phenone group into SecinH3 in such a way that its inhibitory
activity was maintained. Further functionalization with a
detectable biotin tag led to Bio-SecinPP, an activity-based
protein-profiling (ABPP) reagent for cytohesins. The appli-
cation of this probe to a variety of GEFs and GTPases
provided cogent evidence for the specific binding of SecinH3
to the Sec7 domain of the members of the cytohesin family.
We can now define reaction conditions that might enable the
proteome-wide, affinity-based profiling of cytohesin com-
plexes in whole cells or even organisms.
Keywords: benzophenone · cytohesins · inhibitors ·
photoaffinity labeling · proteins
.
[1] a) M. Hafner, A. Schmitz, I. Grꢁne, S. G. Srivatsan, B. Paul, W.
Kolanus, T. Quast, E. Kremmer, I. Bauer, M. Famulok, Nature
2006, 444, 941; b) M. Hafner, E. Vianini, B. Albertoni, L.
Marchetti, I. Grꢁne, C. Gloeckner, M. Famulok, Nat. Protoc.
2008, 3, 579.
[2] For reviews, see: a) M. Famulok, J. S. Hartig, G. Mayer, Chem.
Rev. 2007, 107, 3715; b) S. G. Srivatsan, M. Famulok, Comb.
Chem. High Throughput Screening 2007, 10, 698; c) M. Famulok,
Curr. Opin. Mol. Ther. 2005, 7, 137; d) M. Famulok, G. Mayer,
ChemBioChem 2005, 6, 19.
[3] a) A. K. Gillingham, S. Munro, Annu. Rev. Cell Dev. Biol. 2007,
23, 579; b) W. Kolanus, Immunol. Rev. 2007, 218, 102; c) J. L.
Bos, H. Rehmann, A. Wittinghofer, Cell 2007, 129, 865.
[4] B. Fuss, T. Becker, I. Zinke, M. Hoch, Nature 2006, 444, 945.
[5] J. G. Donaldson, C. L. Jackson, Curr. Opin. Cell Biol. 2000, 12,
475.
[6] a) G. Dormꢂn, G. D. Prestwich, Trends Biotechnol. 2000, 18, 64;
b) F. Kotzyba-Hibert, I. Kapfer, M. Goeldner, Angew. Chem.
1995, 107, 1391; Angew. Chem. Int. Ed. Engl. 1995, 34, 1296;
c) L. Q. Al-Mawsawi, V. Fikkert, R. Dayam, M. Witvrouw, T. R.
Burke, Jr., C. H. Borchers, N. Neamati, Proc. Natl. Acad. Sci.
USA 2006, 103, 10080; d) A. Sinz, Angew. Chem. 2007, 119, 670;
Angew. Chem. Int. Ed. 2007, 46, 660; e) M. C. Hagenstein, J. H.
Mussgnug, K. Lotte, R. Plessow, A. Brockhinke, O. Kruse, N.
Sewald, Angew. Chem. 2003, 115, 5793; Angew. Chem. Int. Ed.
2003, 42, 5635.
[7] a) B. F. Cravatt, A. T. Wright, J. W. Kozarich, Annu. Rev.
Biochem. 2008, 77, 383; b) N. Jessani, B. F. Cravatt, Curr. Opin.
Angew. Chem. Int. Ed. 2008, 47, 9565 –9568
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9567

Communications
Chem. Biol. 2004, 8, 54; c) K. T. Barglow, B. F. Cravatt, Nat.
[8] X. Bi, A. M. Hayallah, M. Famulok, unpublished results.
[9] a) G. Dorman, G. D. Prestwich, Biochemistry 1994, 33, 5661;
b) P. J. A. Weber, A. G. Beck-Sickinger, J. Pept. Res. 1997, 49,
375.
[10] J. Cherfils, J. Menetrey, M. Mathieu, G. Le Bras, S. Robineau, S.
Beraud-Dufour, B. Antonny, P. Chardin, Nature 1998, 392, 101.
Methods 2007, 4, 822; d) S. A. Sieber, B. F. Cravatt, Chem.
Commun. 2006, 2311; e) A. T. Wright, B. F. Cravatt, Chem. Biol.
2007, 14, 1043; f) S. A. Sieber, S. Niessen, H. S. Hoover, B. F.
Cravatt, Nat. Chem. Biol. 2006, 2, 274; g) T. Bꢃttcher, S. A.
Sieber, Angew. Chem. 2008, 120, 4677; Angew. Chem. Int. Ed.
2008, 47, 4600; h) M. C. Hagenstein, N. Sewald, J. Biotechnol.
2006, 124, 56.
9568 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9565 –9568