Induced Protein Dimerization in Vivo through Covalent Labeling

Article abstract of DOI:10.1021/ja037883p

We present here a novel approach to induce protein dimerization in living cells through covalent labeling of fusion proteins with ligands which are interacting with other proteins. The covalent labeling is based on the reaction of the DNA repair protein O

Full text of DOI:10.1021/ja037883p

Published on Web 11/13/2003
Induced Protein Dimerization in Vivo through Covalent Labeling
Susanne Gendreizig,§,† Maik Kindermann,§,† and Kai Johnsson*,§,†
Institute of Molecular and Biological Chemistry, Institute of Biomolecular Sciences, EÄ cole Polytechnique Fe´de´rale
de Lausanne (EPFL), CH-1015 Lausanne, Switzerland
Received August 12, 2003; E-mail: kai.johnsson@epfl.ch
The dynamic dimerization and dissociation of pairs of proteins
plays an important role in various biological processes.1 As a
chemical approach to study processes that depend on protein
dimerization the groups of Schreiber and Crabtree have introduced
“chemical inducers of dimerization” (CIDs).2 CIDs are cell-
permeable molecules which can bind simultaneously to two different
proteins, thereby inducing their dimerization. Various biological
processes have been controlled and studied with this approach,
including signal transduction and control of transcription in
eukaryotic and prokaryotic cells.1-3 Currently used CIDs rely on
noncovalent interactions, and we present here a new approach to
induce protein dimerization in vivo which is based on the covalent
labeling of fusion proteins with ligands capable of interacting with
other proteins. The high specificity of the covalent labeling reaction
should allow the approach to become an important tool for studying
and controlling protein dimerization in biological processes.
The specific labeling of one of the partners participating in the
induced dimerization is based on the unusual mechanism of the
human DNA repair protein O6-alkylguanine-DNA alkyltransferase
(hAGT). In its natural DNA repair function, hAGT irreversibly
transfers the methyl group from O6-alkylated guanine-DNA to a
reactive cysteine.4 Using cell-permeable O6-benzylguanine (BG)
derivatives of the type 1, we have shown that hAGT fusion proteins
can be labeled in living cells with a variety of different reporter
groups (Figure 1A).5 To convert 1 into a CID, we envisioned the
use of BG-methotrexate heterodimers 2 or 3 in which BG is
coupled to methotrexate (Mtx) using different linkers (Figure 1B).
Mtx is a tight-binding inhibitor of dihydrofolate reductase (DHFR),
Figure 1. Labeling of hAGT fusion proteins with synthetic molecules.
and heterodimers of Mtx and dexamethasone, a ligand of the
(A) Mechanism of labeling reaction; (B) BG derivatives used for the labeling
of hAGT fusion protein with Mtx; (C) hAGT-based three-hybrid system.
Mtx in the BG-based CID is represented as a red ball.
glucocorticoid receptor (GR), have been used as CIDs to control
transcription in yeast.6 In this so-called three-hybrid system, a DNA-
binding domain and a transcriptional activation domain were
expressed as DHFR and GR fusion proteins, respectively, and
transcription initiated through the addition of the CID. To test if
BG derivatives 2 or 3 can act as CIDs in such a three-hybrid system,
we envisioned the construction of fusion proteins of hAGT with
the DNA-binding domain LexA and of DHFR with the transcrip-
tional activation domain B42 (Figure 1C). The in vivo labeling of
the hAGT fusion protein with methotrexate using 2 or 3 should
then induce the dimerization of the hAGT and DHFR fusion
proteins, leading to stimulation of transcription of a reporter gene
(Figure 1C).
We synthesized the different BG-Mtx heterodimers 2 and 3
(Supporting Information), as it is known that the nature of the linker
can significantly influence the efficiency of CIDs.6b For the
construction of hAGT fusion proteins, we used a previously
described hAGT mutant with increased activity against BG
compared to wild-type.5b In addition, we introduced three mutations
that have been shown to disrupt DNA binding of hAGT: Lys125Ala,
Ala127Thr, and Arg128Ala.7 The resulting hAGT mutant carrying
the five mutations described above was abbreviated as 3HYAGT.
To verify that 3HYAGT possesses activity against 2 and 3, we
expressed 3HYAGT as fusion protein with glutathione S-transferase
(GST-3HYAGT) in Escherichia coli, purified the protein, and
determined its activity in an in vitro assay (Supporting Information).
GST-3HYAGT reacts rapidly with both 2 and 3, possessing second-
order rate constants of about (2600 ( 460) s-1 M-1 for the reaction
with 2 and of about (1800 ( 430) s-1 M-1 for the reaction with 3.
For control experiments we also constructed a hAGT mutant in
which the reactive Cys145 was mutated to alanine (A145AGT).
Pairs of plasmids encoding LexA and B42 fusion proteins were
transformed into the yeast strain L40 in which the dimerization of
LexA and B42 fusion proteins leads to transcription of the reporter
genes HIS3 and lacZ (Supporting Information). The expression of
each fusion protein after transformation in L40 was verified by
Western blotting using appropriate antibodies. We then tested if
the expression of different combinations of fusion protein in the
presence of either 2 or 3 complemented the histidine auxotrophy
of the yeast L40 (Table 1).
§ Institute of Molecular and Biological Chemistry.
† Institute of Biomolecular Sciences.
9
14970
J. AM. CHEM. SOC. 2003, 125, 14970-14971
10.1021/ja037883p CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Transcriptional Activation of HIS3 and lacZ in Yeast L40
This hypothesis is supported by the observation that the transcrip-
tional activation of lacZ in this system increases with the concentra-
tion of 2 and 3, whereas a quantitative labeling would result in a
decrease of transcriptional activation of lacZ with increasing
concentration of CID.9 At 1, 10, and 100 µM of 3, the measured
ONPG hydrolysis in cell extracts was 39, 830, and 8100 nM min-1
mg-1, respectively (Supporting Information). For 2, the measured
â-galactosidase activity in cell extracts was 13, 865, and 2600 nM
min-1 mg-1 at concentrations of 0.1, 1, and 10 µM of 2, respectively
(Supporting Information). The relatively low solubility of 2 made
measurements at higher concentrations impossible.
To evaluate the achieved transcriptional activation using CID 2,
we measured the transcriptional activation of lacZ by coexpression
of LexA-Fos and B42-Jun fusion proteins, a protein pair known
to yield a strong interaction signal in the two-hybrid system.10
â-Galactosidase activity in cell extracts of L40 expressing LexA-
Fos and B42-Jun was below the activity measured for LexA-
3HYAGT and B42-DHFR in the presence of 1 µM CID 2 (Table
1). Furthermore, the observed 200-fold transcriptional activation
at 1 µM of 2 over background is comparable to those achieved
with other noncovalent CIDs in yeast.6,9
growth
ratea
ONPG
hydrolysisb
protein pair
CID
LexA-3HYAGT/
B42-DHFR
1 µM 2
++
870 ( 410
1 µM 3
+
-
-
-
+
39 ( 19
4 ( 2
-
100 µM BG
100 µM Mtx
1 µM 2, 100 µM BG,
100 µM Mtx
1 µM 2, 100 µM BG
1 µM 2, 100 µM Mtx
1 µM 2
4 ( 4
6 ( 4
19 ( 2
+
++
-
41 ( 16
820 ( 210
1 ( 0.3
LexA-A145AGT/
B42-DHFR
LexA-Fos/ B42-Jun
-
++
260 ( 130
a ++: detection of colonies within 3 days; +: detection of colonies
within 6 days; -: no colonies within 7 days. b nM of o-nitrophenol formed
per min per mg total protein. Each value represents the average of at least
three independent experiments.
Coexpression of LexA-3HYAGT and B42-DHFR allowed yeast
L40 to grow on plates lacking histidine but containing either 2 or
3, indicating transcription of HIS3. No growth was observed in the
absence of 2 or 3 or in the presence of only BG or Mtx. For yeast
L40 expressing LexA-3HYAGT and B42-DHFR, 2 proved to be
more effective as a CID than 3, most likely due to differences in
cell permeability of the two CIDs. Yeast L40 coexpressing LexA-
A145AGT and B42-DHFR did not grow on plates lacking histidine
and supplemented with 2, indicating that the growth of the yeast
depends on the labeling of the hAGT fusion protein. For CID 2,
we also investigated if the growth of yeast expressing LexA-
3HYAGT and B42-DHFR on plates lacking histidine but containing
2 could be suppressed by the addition of free Mtx and BG (Table
1). Although the growth rate was significantly lower in the
simultaneous presence of excess Mtx and BG, growth was not
completely suppressed. Addition of only Mtx did not significantly
affect the growth rate. The competition experiments indicate that
the intracellular concentration of the small molecules, and in
particular that of Mtx, is below that of the fusion proteins.
We then examined the activation of transcription of lacZ by the
BG-based CIDs in yeast L40 expressing LexA-3HYAGT and B42-
DHFR fusion proteins (Table 1). In this assay, the activity of the
product of the lacZ gene, â-galactosidase, was determined by
measuring the hydrolysis rate of the chromogenic substrate o-ni-
trophenyl-â-D-galactopyranoside (ONPG) in cell extracts of liquid
cultures.8 The data obtained for the transcriptional activation of
lacZ confirm those observed for HIS3 described above. At 1 µM
CID, 2 leads to about 20-fold higher levels of â-galactosidase
activity compared to 3, and no significant activity was measured
in the absence of CID or when CID was replaced by BG or Mtx.
No transcriptional activation in the presence of 2 was observed
when 3HYAGT was replaced by A145hAGT. The addition of free
Mtx and BG to the culture medium did not lead to complete
suppression of â-galactosidase activity. As observed in the HIS3
assay, adding only Mtx to the medium did not significantly affect
the activity of CID 2, whereas the addition of only BG leads to a
reduction of â-galactosidase activity by a factor of 20. Again, the
lack of (complete) suppression of transcription of lacZ indicates
that the covalent and noncovalent labeling of the 3HYAGT and
DHFR fusion proteins under these conditions is not quantitative.
The work presented here demonstrates that BG derivatives can
be used as CIDs to control transcription in yeast. The specificity
of the labeling of hAGT fusion proteins and its independence of
the nature of the ligand should make the approach a valuable tool
to control protein dimerization in vivo.
Acknowledgment. This work was financed by the Swiss
National Science Foundation, the Bundesamt fuer Bildung und
Wissenschaft (Project X-TB), and the Boehringer Ingelheim
Foundation. We thank Nils Johnsson for valuable advice.
Supporting Information Available: Detailed procedures for
synthesis of 2 and 3 and biochemical assays (PDF). This material is
available free of charge via the Internet at http:// pubs.acs.org.
References
(1) Klemm, J. D.; Schreiber, S. L.; Crabtree, G. R. Annu. ReV. Immunol. 1998,
16, 569-592.
(2) Spencer, D. M.; Wandless, T. J.; Schreiber, S. L.; Crabtree, G. R. Science
1993, 262, 1019-1024.
(3) (a) Rivera, V. M.; Clackson, T.; Natesan, S.; Pollock, R.; Amara, J. F.;
Keenan, T.; Magari, S. R.; Phillips, T.; Courage, N. L.; Cerasoli, F., Jr.;
Holt, D. A.; Gilman, M. Nat. Med. 1996, 2, 1028-1032. (b) Licitra, E.
J.; Liu, J. O. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 12817-12821. (c)
Althoff, E. A.; Cornish, V. W. Angew. Chem., Int. Ed. 2002, 41, 2327-
2330.
(4) Pegg, A. E. Mutat. Res. 2000, 462, 83-100.
(5) (a) Keppler, A.; Gendreizig, S.; Gronemeyer, T.; Pick, H.; Vogel, H.;
Johnsson, K. Nat. Biotechnol. 2003, 21, 86-89. (b) Juillerat, A.;
Gronemeyer, T.; Keppler, A.; Gendreizig, S.; Pick, H.; Vogel, H.;
Johnsson, K. Chem. Biol. 2003, 10, 313-317. (c) Kindermann, M.;
George, N.; Johnsson, N.; Johnsson, K. J. Am. Chem. Soc. 2003, 125,
7810-7811.
(6) (a) Lin, H. N.; Abida, W. M.; Sauer, R. T.; Cornish, V. W. J. Am. Chem.
Soc. 2000, 122, 4247-4248. (b) Abida, W. M.; Carter, B. T.; Althoff, E.
A.; Lin, H.; Cornish, V. W. ChemBioChem 2002, 3, 887-895.
(7) (a) Lim, A.; Li, B. F. EMBO J. 1996, 15, 4050-4060. (b) Daniels, D. S.;
Mol, C. D.; Arvai, A. S.; Kanugula, S.; Pegg, A. E.; Tainer, J. A. EMBO
J. 2000, 19, 1719-1730.
(8) Burke, D.; Dawson, D.; Stearns, T. Methods in Yeast Genetics: A Cold
Spring Harbor Laboratory Course Manual; Cold Spring Harbor Labora-
tory Press: Plainview, NY, 2000.
(9) Hussey, S. L.; Muddana, S. S.; Peterson, B. R. J. Am. Chem. Soc. 2003,
125, 3692-3693.
(10) Mitton, K. P.; Swain, P. K.; Chen, S.; Xu, S.; Zack, D. J.; Swaroop, A.
J. Biol. Chem. 2000, 275, 29794-29799.
JA037883P
9
J. AM. CHEM. SOC. VOL. 125, NO. 49, 2003 14971