Didehydrogeranylgeranyl (ΔΔGG): A fluorescent probe for protein prenylation

Article abstract of DOI:10.1021/ja0119144

The first intrinsically fluorescent analog of geranylgeraniol, (2E,6E,8E,10E,12E,14E)-geranylgeraniol (all-trans-ΔΔGGOH·1) has been synthesized stereoselectively and shown to substitute for the geranylgeranyl (GG) moiety in prenyl transferase reactions and in protein-ligand binding assays. All-trans-ΔΔGGOH 1 showed blue fluorescence in methanol, with λ_ex = 310 nm and λ_em = 410 nm (ε₃₁₀ = 2.4 × 10⁴ M^-1 cm^-1), but was only weakly fluorescent in aqueous solution. The prenyl transferase efficiency for ΔΔGGPP 2 as a substrate for yeast protein geranylgeranyl transferase (PGGTase-I) was 60% relative to that for GGPP. The binding of ΔΔGG-AcCysMe 3 to the recombinant Rho GTPase dissociation inhibitor (RhoGDI) had a K_D of 15.1 ± 1.2 μM, 6-fold lower than the affinity of GG-AcCysMe. Thus, the ΔΔGG moiety is a novel fluorophore suitable for studying the interaction and subcellular localization of prenylated small GTPase proteins in signaling complexes. Copyright

Full text of DOI:10.1021/ja0119144

Published on Web 12/06/2001
Didehydrogeranylgeranyl (∆∆GG): A Fluorescent Probe for Protein
Prenylation
Xiao-hui Liu and Glenn D. Prestwich*
Department of Medicinal Chemistry and the Center for Cell Signaling, 30 South 2000 East, Room 201,
The UniVersity of Utah, Salt Lake City, Utah 84112-5820
Received August 8, 2001
Chart 1. Structure of ∆∆GGOH 1, ∆∆GGPP 2, and
∆∆GG-AcCysMe 3.
Protein prenylation involves the covalent attachment of a farnesyl
or geranylgeranyl moiety via a thioether linkage to the cysteine
residue of a C-terminal CAAX motif. This posttranslational modifi-
cation of small G proteins is required for cellular trafficking, mem-
brane association, and formation of protein-protein signaling com-
plexes that trigger a wide variety of cellular responses.1 Specifically,
the GTPase farnesylated Ras is essential for normal cell growth
and regulation of differentiation.2 Oncogenic H-Ras is well-known
for its Akt-1 activation and transforming ability.3 Geranylgeranyl-
ated Rho GTPases play key roles in the reorganization of the actin
cytoskeleton and the control of gene transcription,4 while bis-
geranylgeranylated Rab GTPases regulate vesicular trafficking and
exocytosis.5 Recently, farnesylation and geranylgeranylation have
emerged as important new targets in cancer therapy6 (Chart 1).
We describe herein the first intrinsically fluorescent analogue
of the geranylgeranyl moiety, and we demonstrate its utility in cell-
based and in vitro protein-ligand interactions. The design of the
parent compound, (2E,6E,8E,10E,12E,14E)-geranyl-geraniol (all-
trans-∆∆GGOH 1), its diphosphate (∆∆GGPP 2), and its N-acetyl
cysteine methyl ester adduct (∆∆GG-AcCysMe 3) required stereo-
selective introduction of 8E- and 12E-alkene bonds to produce a
conjugated pentaene. The pentaene was predicted to fluoresce by
analogy with the naturally occurring pentaene, trans-retinol (vitamin
A1).7 Importantly, the intrinsic fluorophore of ∆∆GGOH 1 does
not add additional steric bulk, although pentaene will be more rigid
than GGOH. Crystallographic and NMR data support an extended
conformation for isoprenoid chains, both in solution and when
bound to PFTase or RhoGDI.8 Moreover, yeast protein geranylgera-
nyltransferase (PGGTase-I) was capable of transferring isoprenoid
analogues to substrate peptides or proteins.9 We hypothesized that
the ∆∆GG moiety would substitute on a molecular level for the
geranylgeranyl group for both biochemical and physiological func-
tions. In particular, we anticipated that ∆∆GG-modified peptides
would provide new ligands for in vitro assays, and that ∆∆GG
derivatives could be delivered intracellularly to study protein trans-
location and protein-protein interactions of signaling complexes
in living cells.
Scheme 1. Synthesis of ∆∆GGOH 1, ∆∆GGPP 2, and
∆∆GG-AcCysMe 3a
a (a) SeO2; (b) KOH; (c) TBDMS-Cl, Et3N; (d) CH3OH, conc. H2SO4;
(e) NBS; (f) P(OEt)3; (g) n-BuLi, 0 °C; (h) LiAlH4; (i) PCC; (j) n-BuLi,
15; (k) TBAF (l) Ph3P, CBr4; (m) (n-Bu4N)3HP2O7; (n) AcCysMe,
Zn(OAc)2, H2O.
purity. The remaining 5% 8Z-isomer was readily removed by SiO2
separation of either the alcohol 11 or aldehyde 12 isomers.
Reports on the stereoselective synthesis of trans di-substituted
alkenes using isoprenyl-substituted phosphorus reagents are scarce.12
Strong bases such as t-BuOK, or NaOMe, which are required to
ionize Wittig-Horner reagents without electron-withdrawing sub-
stitutes, would quickly decompose the sensitive aldehyde 12, even
at -78 °C. On the basis of the mechanism of the Wittig reaction,
we reasoned that a smaller phosphine ligand would minimize the
steric interaction with the two alkyl groups, R1 and R2,13 giving
rise to the desired E-isomer as the dominant product (see Supporting
Information). Three isoprenyl trialkyl phosphites (diphenyl 13,
dibutyl 14, and diethyl 15) were employed at both -78 and 0 °C.
At 0 °C, 13 and 14 gave 64 and 85% E-isomer 16, respectively. At
-78 °C, 14 and 15 afforded 16 with 91 and >99% selectivity. It
appeared that the Wittig reaction was subject to both steric and
kinetic control, since the lower temperature resulted in a slower
collapse of the intermediate and also amplified the effect of energy
difference between the two transition states.14
All-trans-∆∆GGOH 1 showed blue fluorescence in methanol,
with λex ) 310 nm and λem ) 410 nm (ꢀ310 ) 2.4 × 104 M-1
cm-1), but was only weakly fluorescent in aqueous solution. This
environmental sensitivity was confirmed by titration (see Supporting
Information). The fluorescence intensity at 410 nm was quenched
4-fold when the percentage of water reached 50% (v/v). The relative
quantum yield (Φ) in methanol was 0.0075 using trans-retinol (Φ
) 0.0298, with λex ) 355 nm) as the standard;7,15 retinol was
selected for having solubility and spectral properties similar to those
of 1.
The primary challenge in preparing all trans-∆∆GGOH was the
stereoselective installation of the 8E and 12E double bonds. Scheme
1 illustrates the synthetic route to compounds 1, 2, and 3, beginning
with selective oxidation of geranyl acetate with selenium dioxide.
Attempted selective oxidation of the geranyl TBDMS ether resulted
in lower yield and poor selectivity.10 The 8E-alkene was introduced
using a Wittig-Horner reagent, allylic diethylphosphonate 9.11
Moreover, if the proper isomeric allylic diethylphosphonate 9 in
the mixed isomers was 1.2 equiv of the geranyl-TBDMS aldehyde
6, then the 8E-alkene 10 was obtained with over 95% geometric
The ability of ∆∆GGPP 2 to serve as a substrate for yeast
PGGTase-I was measured using a continuous fluorescence method.16
All-trans-∆∆GGPP 2 had a KM value of 0.33 µM (3.45 µM for
* Address for correspondence. Telephone: 801 585-9051. Fax: 801 585-9053.
E-mail: gprestwich@deans.pharm.utah.edu.
9
20 VOL. 124, NO. 1, 2002 J. AM. CHEM. SOC.
10.1021/ja0119144 CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
in subcellular distribution of ∆∆GG-modified proteins will be
presented in due course.
Acknowledgment. Bacteria containing the pGEX2T-RhoGDI
vector were a generous gift from Dr. A. Hall (University College
London). Yeast PGGTase-I expression plasmid was obtained cour-
tesy of Dr. C. D. Poulter (The University of Utah, Utah) and was
expressed by Dr. D.-Y. Suh (UUtah). This work was supported by
an award from the NIH (GM 44836), and by the Center for Cell
Signaling, a member of the Utah Centers of Excellence Program.
Supporting Information Available: Preparative procedures and
analytical data for 1, 2, and 3. Protocols for expression and purification
of PGGTase-I and RhoGDI; enzymatic, fluorescence quenching, fluor-
escence polarization assay procedures. Equations to calculate the quan-
tum yield and for fitting of fluorescence polarization data (PDF). This
material is available free of charge via the Internet at http://pubs.acs.org.
Figure 1. Titration of RhoGDI (5.62 µM) with ∆∆GG-AcCysMe (O 0
µM, 9 5.62 µM, 2 16.9 µM, [ 56.2 µM), AcCysMe (b 48 µM). Inset:
Titration with GG-AcCysMe (0 16.9 µM, 2 56.2 µM).
GGPP), and the prenyl-transfer efficiency was 60% relative to that
for GGPP.
References
(1) (a) Sinensky, M. Biochim. Biophys. Acta 2000, 1484, 93-106. (b) Wagner,
To test the interaction of the ∆∆GG moiety with signaling
proteins that bind geranylgeranylated peptides, we prepared ∆∆GG-
AcCysMe 3 as a fluorescent ligand.17 This is the minimum unit
recognized by the binding domain of proteins that bind the
prenylated C-terminal Cys methyl ester. The fluorescence spectrum
of ∆∆GG-AcCysMe 3 (10 mM Tris-Cl pH 7.5) showed λex ) 360
nm and λem ) 460 nm. The 50 nm red-shift in fluorescence exci-
tation and emission further demonstrated the environmental sen-
sitivity of the ∆∆GG fluorophore in this peptide adduct relative to
the free alcohol 1. We examined the binding of ∆∆GG-AcCysMe
to the recombinant Rho GTPase dissociation inhibitor (RhoGDI).18
In cell cytosol, RhoGDI forms a complex with geranylgeranylated
Rho (GDP bound), which, in addition to slowing the rate of GDP
dissociation from the Rho, has been shown to solubilize the GTPase
from subcellular membranes.19 Therefore, RhoGDI regulates
vesicular trafficking between donor membranes to acceptor mem-
branes.
The binding of ∆∆GG-AcCysMe 3 to RhoGDI20 was determined
through the quenching of the intrinsic fluorescence of the tryptophan
residues (Trp192, Trp194, and Trp202) in a hydrophobic binding
pocket (Figure 1).21 Thus, when RhoGDI was titrated with ∆∆GG-
AcCysMe 3 or GG-AcCysMe (Figure 1, inset), but not with non-
prenylated AcCysMe, a significant reduction in fluorescence was
observed within 5 min. A red-shift (10-15 nm) of the Trp fluor-
escence indicated that water was involved in this binding interaction.
After 1 h, the fluorescence of RhoGDI had completely recovered,
further suggesting that quenching was due to dynamic changes in
the Trp microenvironments.21
To measure the binding constant of ∆∆GG-AcCysMe 3
to RhoGDI, fluorescence anisotropy was employed. Binding
of rhodamine-labeled geranylgeranyl cysteine methyl ester
(GG-RhodCysMe) to RhoGDI showed increasing polarization as
the concentration of RhoGDI in the assay buffer increased. A
dissociation constant (KD) was determined to be 2.45 ( 0.14 µM
by curve fitting, in close agreement with the reported value.18
Competitive binding was measured for ∆∆GG-AcCysMe 3, and
curve fitting gave a KD value of 15.1 ( 1.2 µM. This indicated
that the more rigid ∆∆GG fluorophore had approximately 6-fold
lower affinity compared with the geranylgeranyl moiety, equivalent
to the displacement obtained with Farn-AcCysMe.
In conclusion, the in vitro reaction of ∆∆GGPP 2 with yeast
PGGTase-I and the binding between ∆∆GG-AcCysMe 3 and
RhoGDI demonstrate the potential of the ∆∆GG moiety as a novel
fluorophore for studying the interaction and subcellular localization
of prenylated small GTPase proteins in signaling complexes. Cell-
based fluorescence assays22 to monitor the time course of changes
A.; Williams, J. Am. J. Physiol. 1994, 266, G1-14.
(2) Barbacid, M. Annu. ReV. Biochem. 1987, 56, 779-827.
(3) Liu, A. X.; Prendergast, G. C. FEBS Lett. 2000, 481, 205-208.
(4) (a) Mackay, D. J.; Hall, A. J. Biol. Chem. 1998, 273, 20685-20688. (b)
Hall, A. Science 1998, 279, 509-514 (c) Allal, C.; Favre, G.; Couderc,
B.; Salicio, S.; Sixou, S.; Hamilton, A. D.; Sebti, S. M.; Lajoie-Mazenc,
I.; Pradines, A. J. Biol. Chem. 2000, 275, 31001-31008.
(5) (a) Plutner, H.; Schwaninger, R.; Pind, S.; Balch, W. E. EMBO J. 1990,
9, 2375-2383. (b) Kalab, P.; Pu, R. T.; Dasso, M. Curr. Biol. 1999, 9,
481-484.
(6) (a) Crul, M.; deKlerk, G. D.; Beijnen, J. H.; Schellens, J. H. Anticancer
Drugs 2001, 12, 163-184. (b) Bell, I. M. Expert Opin. Therap. Patents
2000, 10, 1813-1831. (c) Adjei, A. A. Drugs Fut. 2000, 25, 1069-1079.
(7) Tsin, A. T. C.; Pedrozo-Fernandes, H. A.; Gallas, J. M.; Chambers, J. P.
Life Sci. 1988, 43, 1379-1384.
(8) (a) Gosser, Y.; Nomanbhoy, T.; Aghazadeh, B.; Manor, D.; Combs, C.;
Cerione, R.; Rosen, M. Nature 1997, 387, 814-819. (b) Hoffman, G.;
Nassar, N.; Cerione, R. Cell 2000, 100, 345-356. (c) Strickland, C. L.;
Windsor, W. T.; Syto, R.; Wang, L.; Bond, R.; Wu, Z.; Schwartz, J.; Le,
H. V.; Beese, L. S.; Weber, P. C. Biochemistry USA 1998, 37, 16601-
16611. (d) Zahn, T. J.; Eilers, M.; Guo, Z. M.; Ksebati, M. B.; Simon,
M.; Scholten, J. D.; Smith, S. O.; Gibbs, R. A. J. Am. Chem. Soc. 2000,
122, 7153-7164.
(9) (a) Marecak, D. M.; Horiuchi, Y.; Arai, H.; Shimonaga, M.; Maki, Y.;
Koyama, T.; Ogura, K.; Prestwich, G. D. Bioorg. Med. Chem. Lett. 1997,
7, 1973-1978. (b) Kale, T. A.; Raab, C.; Yu, N.; Dean, D. C.; Distefano,
M. D. J. Am. Chem. Soc. 2001, 123, 4373-4381. (c) Yokoyama, K.;
Zimmerman, K.; Scholten, J.; Gelb, M. H. J. Biol. Chem. 1997, 272,
3944-3952.
(10) Fairlamb, I. J. S.; Dickinson, J. M.; Pegg, M. Tetrahedron Lett. 2001, 42,
2205-2208.
(11) Thus, bromination of tiglic acid with NBS gave 8 (E:Z ) 2:1), and the
Michaelis-Arbuzov reaction gave two isomeric phosphonates; the proper
isomer of 9 was selectively consumed in the next step. Pattenden, G.;
Weedon, B. C. L. J. Chem. Soc. (C) 1968, 1997-2006.
(12) Liu, R.-q.; Schlosser, M. SynLett 1996, 1197-1198.
(13) Tamura, R.; Kato, M.; Saegusa, K.; Kakihana, M.; Oda, D. J. Org. Chem.
1987, 52, 4221-4124.
(14) Briefly, n-BuLi (1.2 equiv) was added into a mixture containing isoprenyl
triethyl phosphite 14 (1 equiv) in anhydrous THF at 0 °C. After 0.5 h,
the mixture was moved to -78 °C, and aldehyde 11 (0.8 equiv) was added
dropwise. Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry,
3rd edition; Plenum Press: New York, 1990.
(15) Eaton, D. F. Pure Appl. Chem. 1988, 60, 1107-1114.
(16) Stirtan, W. G.; Poulter, C. D. Biochemistry 1997, 36, 4552-4557
(17) Xue, C.-B.; Becker, J. M.; Naider, F. Tetrahedron Lett. 1992, 33, 1435-
1438.
(18) Mondal, M. S.; Wang, Z.; Seeds, A. M.; Rando, R. R. Biochemistry 2000,
39, 406-412.
(19) (a) Sasaki, T.; Takai, Y. Biochem. Biophys. Res. Commun. 1998, 245,
641-645. (b) Koch, G.; Tanaka, K.; Masuda, T.; Yamochi, W.; Honaka,
H.; Takai, Y. Oncogene 1997, 15, 417-422.
(20) The soluble GST-RhoGDI fusion protein was expressed using modifica-
tions of the original protocol (pGEX2T-RhoGDI expression vector, E.
coli strain JM109, 25 °C, 20 h, 0.15 mM IPTG) and purified by affinity
chromatography. Hancock, J. F.; Hall, A. EMBO J. 1993, 12, 1915-1921.
(21) (a) Calls, P. R. In Methods of Enzymology; Brand, L., Johnson, M. L.,
Eds.; Academic Press: New York, 1997; Vol. 278, pp 113-150. (b)
Eftink, M. R. In Methods of Enzymology; Brand, L., Johnson, M. L., Eds.;
Academic Press: New York, 1997; Vol. 278, pp 221-257. (c) Chen, Y.;
Barkley, M. D. Biochemistry 1998, 37, 9976-9982.
(22) (a) Choy, E.; Chiu, V.; Silletti, J.; Feoktistov, M.; Morimoto, T.;
Michaelson, D.; Ivanov, I.; Philips, M. Cell 1999, 98, 69-80. (b) Kraynov,
V. S.; Chamberlain, C.; Bokoch, G. M.; Schwartz, M. A.; Slabaugh, S.;
Hahn, K. M. Science 2000, 290, 333-337.
JA0119144
9
J. AM. CHEM. SOC. VOL. 124, NO. 1, 2002 21