Photoaffinity analogues of farnesyl pyrophosphate transferable by protein farnesyl transferase

Article abstract of DOI:10.1021/ja0124717

Farnesylation is a posttranslational lipid modification in which a 15-carbon farnesyl isoprenoid is linked via a thioether bond to specific cysteine residues of proteins in a reaction catalyzed by protein farnesyltransferase (FTase). We synthesized analogues (3-6) of farnesyl pyrophosphate (FPP) to probe the range of modifications possible to the FPP skeleton which allow for efficient transfer by FTase. Photoaffinity analogues of FPP (5, 6) were prepared by substituting perfluorophenyl azide functional groups for the ω-terminal isoprene of FPP. Substituted anilines replace the ω-terminal isoprene in analogues 3 and 4. Compounds 3-5 were prepared by reductive amination of the appropriate anilines with 8-oxogeranyl acetate, followed by ester hydrolysis, chlorination, and pyrophosphorylation. Additional substitution of three methylenes for the β-isoprene of FPP gave photoprobe 6 in nine steps. Preparation of the analogues required TiCl₄-mediated imine formation prior to NaBH(OAc)₃ reduction for anilines with a pK_a < 1. The azide moiety was not affected by Ph₃PCl₂ conversion of allylic alcohols 13-16 into corresponding chlorides 17-20. Analogues 3-6 are efficiently transferred to target N- dansyl-GCVLS peptide substrate by mammalian FTase. Comparison of analogue structures and kinetics of transfer to those of FPP reveals that ring fluorination and para substituents have little effect on the affinity of the analogue pyrophosphate for FTase and its transfer efficiency. These results are also supported with models of the analogue binding modes in the active site of FTase. The transferable azide photoprobe 5 photoinactivates FTase. Transferable analogues 5 and 6 allow the formation of appropriately posttranslationally modified photoreactive peptide probes of isoprene function.

Full text of DOI:10.1021/ja0124717

Published on Web 06/18/2002
Photoaffinity Analogues of Farnesyl Pyrophosphate
Transferable by Protein Farnesyl Transferase
Kareem A. H. Chehade,†,§ Katarzyna Kiegiel,†,§ Richard J. Isaacs,†,§
Jennifer S. Pickett,| Katherine E. Bowers,| Carol A. Fierke,|
Douglas A. Andres,†,§ and H. Peter Spielmann*,†,‡,§
Contribution from the Department of Molecular and Cellular Biochemistry,
Department of Chemistry, Kentucky Center for Structural Biology,
UniVersity of Kentucky, Lexington, Kentucky 40536-0084, and
Department of Chemistry, UniVersity of Michigan, Ann Arbor, Michigan 48109
Received November 5, 2001
Abstract: Farnesylation is a posttranslational lipid modification in which a 15-carbon farnesyl isoprenoid is
linked via a thioether bond to specific cysteine residues of proteins in a reaction catalyzed by protein
farnesyltransferase (FTase). We synthesized analogues (3-6) of farnesyl pyrophosphate (FPP) to probe
the range of modifications possible to the FPP skeleton which allow for efficient transfer by FTase.
Photoaffinity analogues of FPP (5, 6) were prepared by substituting perfluorophenyl azide functional groups
for the ω-terminal isoprene of FPP. Substituted anilines replace the ω-terminal isoprene in analogues 3
and 4. Compounds 3-5 were prepared by reductive amination of the appropriate anilines with 8-oxo-
geranyl acetate, followed by ester hydrolysis, chlorination, and pyrophosphorylation. Additional substitution
of three methylenes for the â-isoprene of FPP gave photoprobe 6 in nine steps. Preparation of the analogues
required TiCl4-mediated imine formation prior to NaBH(OAc)3 reduction for anilines with a pKa < 1. The
azide moiety was not affected by Ph3PCl2 conversion of allylic alcohols 13-16 into corresponding chlorides
17-20. Analogues 3-6 are efficiently transferred to target N-dansyl-GCVLS peptide substrate by mammalian
FTase. Comparison of analogue structures and kinetics of transfer to those of FPP reveals that ring
fluorination and para substituents have little effect on the affinity of the analogue pyrophosphate for FTase
and its transfer efficiency. These results are also supported with models of the analogue binding modes in
the active site of FTase. The transferable azide photoprobe 5 photoinactivates FTase. Transferable
analogues 5 and 6 allow the formation of appropriately posttranslationally modified photoreactive peptide
probes of isoprene function.
The Ras family of proteins play a pivotal role in the control
of cellular growth and differentiation through their interaction
with a variety of cellular effectors. Ras is a monomeric GTPase
which is associated with the inner leaflet of the plasma
membrane and functions as a binary molecular switch that cycles
between inactive GDP and active GTP bound states.1 Posttrans-
lational modification of Ras with the 15-carbon farnesyl
isoprenoid is essential for the protein to become properly
localized to the inner leaflet of the plasma membrane.2-4 Protein
farnesyltransferase (FTase) catalyzes the formation of a thioether
between the farnesyl isoprenoid from farnesyl pyrophosphate
1 (FPP) and a conserved cysteine in protein substrates.2,5 The
reactive cysteine is located in the C-terminal Ca1a2X motif in
which C is the cysteine to be modified, a1 and a2 are often
aliphatic residues, and X is one of various residues including
Ser, Met, Ala, or Gln.2 Mutated forms of cellular Ras genes
are among the most common genetic abnormalities in human
cancer, occurring in 30% of all neoplasms.6-11 Furthermore,
farnesylation of Ras is required to transform cells.7,12,13 Con-
sequently, a number of selective FTase inhibitors (FTIs) that
prevent Ras farnesylation have been developed as anticancer
therapeutics.14-21
(5) Hightower, K. E.; Fierke, C. A. Curr. Opin. Chem. Biol. 1998, 3, 176.
(6) Casey, P. J.; Solski, P. A.; Der, C. J.; Buss, J. E. Proc. Natl. Acad. Sci.
U.S.A. 1989, 86, 8323-8327.
* To whom correspondence should be addressed. Telephone: 859-257-
4790. Fax: 859-257-8940. E-mail: hps@pop.uky.edu.
† Department of Molecular and Cellular Biochemistry, University of
Kentucky.
(7) Hancock, J. F.; Magee, A. I.; Childs, J. E.; Marshall, C. J. Cell 1989, 57,
1167-1177.
‡ Department of Chemistry, University of Kentucky.
§ Kentucky Center for Structural Biology, University of Kentucky.
| University of Michigan.
(8) Schafer, W. R.; Kim, R.; Sterne, R.; Thorner, J.; Kim, S. H.; Rine, J. Science
1989, 245, 379-385.
(9) Bos, J. L. Cancer Res. 1989, 49, 4682-4689.
(10) Clark, G. J.; Der, C. J. Ras Proto-oncogene ActiVation in Human
Malignancy; Humana Press: Totowa, NJ, 1995; pp 17-52.
(11) Khosravi-Far, R.; Der, C. J. Cancer Metastasis ReV. 1994, 13, 67-89.
(12) Schafer, W. R.; Rine, J. Annu. ReV. Genet. 1992, 30, 209-237.
(13) Barbacid, M. Annu. ReV. Biochem. 1987, 56, 779-827.
(14) Gibbs, J. B.; Oliff, A. Annu. ReV. Pharmacol. Toxicol. 1997, 37, 143-
166.
(1) Scheffzek, K.; Ahmadian, M. R.; Wittinghofer, A. Trends Biochem. Sci.
1998, 23, 257.
(2) Zhang, F. L.; Casey, P. J. Annu. ReV. Biochem. 1996, 65, 241-269.
(3) Reiss, Y.; Goldstein, J. L.; Seabra, M. C.; Casey, P. J.; Brown, M. S. Cell
1990, 62, 81-88.
(4) Glomset, J. A.; Farnsworth, C. C. Annu. ReV. Cell Biol. 1994, 10, 181-
205.
9
8206
J. AM. CHEM. SOC. 2002, 124, 8206-8219
10.1021/ja0124717 CCC: $22.00 © 2002 American Chemical Society

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
Although our understanding of the biological consequences
of Ras farnesylation has increased significantly in recent years,
there is relatively little known about the molecular details of
the interactions that give rise to the lipid’s functional role and
of the processes downstream of farnesylation. Major questions
such as whether the lipid participates in protein-protein
recognition or the exact mechanism by which the modified Ras
is directed only to the inner leaflet of the plasma membrane
remain unanswered. While farnesylation of Ras is absolutely
required for its biological effects and membrane localization, it
is unclear whether the prenyl group functions as a hydrophobic
membrane association signal,7,22-26 a protein recognition signal,
or both.27,28 Recently, several X-ray crystal structures of the
Ras related small G-proteins Rac and Cdc42 in complex with
the GDP dissociation inhibitor protein RhoGDI have become
available. In contrast to Ras, Rac and Cdc42 are geranylgera-
nylated and are cytosolically located in a complex with RhoGDI
in resting cells. The X-ray crystal structures of the complexes
show that the geranylgeranyl group is completely inserted into
a hydrophobic binding pocket of RhoGDI.29-31
Heterobifunctional photoprobe analogues of farnesyl pyro-
phosphate that are transferred by FTase to Ras should enable
the dissection of these interactions through the identification
of specific molecular partners. Analogues of FPP that are
alternative transferable substrates for FTase have been utilized
to study farnesylation and some aspects of the subsequent
downstream processing events.26,32-37 FPP analogues stripped
of most isoprenoid features such as methyl groups and unsat-
uration to more closely resemble simple fatty acids are
transferred to Ras by FTase and allow Ras to function in a
Xenopus signal transduction model system.26 Photoaffinity
labeling has been used extensively in the study of protein
structure and function,38,39 and photoaffinity analogues of FPP
derivatized with diazoketone and benzophenone photoaffinity
moieties have been synthesized.40-51 Unfortunately, none of
the FPP photoaffinity analogues described in the literature are
efficiently transferred by FTase to protein substrates.40-50
A
significant impediment to the development of heterobifunc-
tional FPP photoprobes is that the FTase-catalyzed transfer of
analogues to substrate is very sensitive to the structure of
the lipid.26,36,37,40-50 FPP analogues containing diazoketone
functionality are not transferred by FTase and instead act
as FTIs.40-44 The best transferable FPP photoaffinity probe
developed to date is derived from an ether linked benzo-
phenone moiety in place of the â- and ω-isoprene units of FPP.50
This analogue is transferred to protein substrate very slowly
with low efficiency and requires prolonged UV irradiation
to photochemically inactivate yeast FTase. Recently, the iso-
prenoid portion of this analogue S-linked to cysteine has been
employed as a photoprobe of the isoprenoid binding protein
RhoGDI.49
Substitution of the terminal isoprene of FPP by an aniline
group gives 8-anilinogeranyl pyrophosphate (2, AGPP, Figure
1), which is transferred to Ras by FTase with the same kinetics
as FPP.36 Transferable FPP analogues 38-42 were also obtained
by replacing the ω-isoprene of FPP with a benzyloxy group
and substitution of the â-isoprene with a variable length
methylene chain (Figure 2).37 These studies indicate that an
aromatic ring is an acceptable isostere for the terminal isoprene
of FPP in the FTase-catalyzed transfer of lipid to substrate.
Consequently, transferable FPP photoprobes based on perfluo-
rophenyl azides are an attractive alternative to diazoketone and
benzophenone photoaffinity reagents because they have different
steric and electronic properties.
We report the synthesis and biochemical characterization of
FPP analogues 3 and 4 which contain substituted anilines and
of the heterobifunctional perfluorophenyl azide FPP photoana-
logues 5 and 6. Pyrophosphates 3-6 are transferred by FTase
to dansyl-GCVLS peptide substrate, and analogue 5 inactivates
FTase in a light-dependent manner. Comparison of the analogue
structures and kinetics of analogue transfer with those of FPP
reveal that ring fluorination and para substituents have little
effect on the affinity of the analogue pyrophosphate for FTase
and its transfer efficiency. These results are also supported with
models of the analogue binding modes in the active site of
FTase. Additionally, we have found that weakly basic anilines
having a pKa g 1 are successfully alkylated by standard
reductive amination conditions, while alkylation of anilines
(15) Leonard, D. M. J. Med. Chem. 1997, 40, 2971-2990.
(16) Sebti, S. M.; Hamilton, A. D. Pharmacol. Ther. 1997, 74, 103-114.
(17) Rowinsky, E. K.; Windle, J. J.; Von Hoff, D. D. J. Clin. Oncol. 1999, 17,
3631-3652.
(18) Sebti, S. M.; Hamilton, A. D. Expert Opin. InVest. Drugs 2000, 9, 2767-
2782.
(19) Prendergast, G. C.; Rane, N. Expert Opin. InVest. Drugs 2001, 10, 2105-
16.
(20) Singh, S. B.; Lingham, R. B. Curr. Opin. Drug DiscoVery DeV. 2002, 5,
225-44.
(21) Ganguly, A. K.; Doll, R. J.; Girijavallabhan, V. M. Curr. Med. Chem. 2001,
8, 1419-36.
(22) Hancock, J. F.; Cadwallader, K.; Paterson, H.; Marshall, C. J. EMBO J.
1991, 10, 4033-4039.
(23) Hancock, J. F.; Cadwallader, K.; Marshall, C. J. EMBO J. 1991, 10, 641-
646.
(24) Leevers, S. J.; Paterson, H. F.; Marshall, C. J. Nature 1994, 369, 411-
414.
(25) Stokoe, D.; Macdonald, S. G.; Cadwallader, K.; Symons, M.; Hancock, J.
F. Science 1994, 264, 1463-1467.
(39) Schuster, D. I.; Probst, W. C.; Ehrlich, G. K.; Singh, G. Photochem.
Photobiol. 1989, 49, 785.
(26) Dudler, T.; Gelb, M. H. Biochemistry 1997, 36, 12434-12441.
(27) Marshall, C. J. Science 1993, 259, 1865-1866.
(28) Casey, P. J. Science 1995, 268, 221-225.
(40) Baba, T.; Allen, C. M. Biochemistry 1984, 23, 1312-1322.
(41) Baba, T.; Muth, J.; Allen, C. M. J. Biol. Chem. 1985, 260, 10467-10473.
(42) Allen, C. M. Methods Enzymol. 1985, 110, 117-124.
(43) Bukhtiyarov, Y. E.; Omer, C. A.; Allen, C. M. J. Biol. Chem. 1995, 270,
19035-19040.
(29) Grizot, S.; Faure, J.; Fieschi, F.; Vignais, P. V.; Dagher, M. C.; Pebay-
Peyroula, E. Biochemistry 2001, 40, 10007-13.
(30) Hoffman, G. R.; Nassar, N.; Cerione, R. A. Cell 2000, 100, 345-56.
(31) Scheffzek, K.; Stephan, I.; Jensen, O. N.; Illenberger, D.; Gierschik, P.
Nat. Struct. Biol. 2000, 7, 122-6.
(44) Das, N. P.; Allen, C. M. Biochem. Biophys. Res. Commun. 1991, 181, 729-
735.
(45) Gaon, I.; Turek, T. C.; Weller, V. A.; Edelstein, R. L.; Singh, S. K.;
Distefano, M. D. J. Org. Chem. 1996, 61, 7738-7745.
(46) Gaon, I.; Turek, T. C.; Distefano, M. D. Tetrahedron Lett. 1996, 37, 8833-
8836.
(32) Gibbs, R. A.; Krishnan, U.; Dolence, J. M.; Poulter, C. D. J. Org. Chem.
1995, 60, 7821.
(33) Mu, Y. Q.; Gibbs, R. A.; Eubanks, L. M.; Poulter, C. D. J. Org. Chem.
1996, 61, 8010.
(34) Shao, Y.; Eummer, J. T.; Gibbs, R. A. Org. Lett. 1999, 1, 627-630.
(35) Gibbs, B. S.; Zahn, T. J.; Mu, Y. Q.; Sebolt-Leopold, J. S.; Gibbs, R. A.
J. Med. Chem. 1999, 42, 3800.
(47) Turek, T. C.; Gaon, I.; Distefano, M. D. Tetrahedron Lett. 1996, 37, 4845-
4848.
(48) Edelstein, R. L.; Distefano, M. D. Biochem. Biophys. Res. Commun. 1997,
235, 377-382.
(36) Chehade, K. A. H.; Andres, D. A.; Morimoto, H.; Spielmann, H. P. J.
Org. Chem. 2000, 65, 3027-3033.
(49) Kale, T. A.; Raab, C.; Yu, N.; Dean, D. C.; Distefano, M. D. J. Am. Chem.
Soc. 2001, 123, 4373-4381.
(37) Micali, E.; Chehade, K. A.; Isaacs, R. J.; Andres, D. A.; Spielmann, H. P.
Biochemistry 2001, 40, 12254-65.
(50) Turek, T. C.; Gaon, I.; Distefano, M. D. J. Org. Chem. 2001, 66, 3253-
3264.
(38) Bayley, H. Photogenerated Reagents in Biochemistry and Molecular
Biology; Elsevier: New York, 1983.
(51) Zhang, Y. W.; Koyama, T.; Marecak, D. M.; Prestwich, G. D.; Maki, Y.;
Ogura, K. Biochemistry 1998, 37, 13411-20.
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8207

A R T I C L E S
Chehade et al.
Results and Discussion
Design of Photoaffinity Analogues. The synthetic route
outlined in Scheme 1 provides a convenient method for the
incorporation of a wide variety of functionalized anilines into
the 8-anilinogeranyl skeleton through reductive amination.36
Examination of the six available FTase X-ray crystal structures
with and without bound substrates suggests that transferable FPP
analogues containing azide substituted aromatic groups might
be accommodated in the active site of the enzyme.50,52-56 To
this end, compounds 3-7 (Schemes 1 and 2) were synthesized
to test the range of chemical functionalities, that when incor-
porated into FPP analogues, would still allow appropriate
transfer of lipid to protein substrate by FTase.
Synthesis of FPP Analogues 3 and 4 and Photoprobe 5.
We have previously shown that acetate 9 was obtained in good
yield by the addition of a slight excess of NaBH(OAc)3 to a
solution of aldehyde 8 and aniline in 1,2-dichloroethane at 25
°C.36 Substitution of p-nitroaniline for aniline in this reaction
afforded allylic amine 10 in 64% yield. However, reaction of
pentafluoroaniline under these conditions only resulted in
reduction of aldehyde 8 without formation of aniline 11. We
found that synthesis of allylic amine 11 required preformation
of the imine intermediate prior to hydride reduction. The desired
imine was formed in a few hours by condensation of aldehyde
8 with pentafluoroaniline in the presence of TiCl4 and excess
pyridine.57 The perfluoro imine was isolated by column chro-
matography and characterized by NMR, IR, and mass spec-
troscopy and can be stored indefinitely at -20 °C under argon.
In practice, the imine-TiCl4 complex was treated in situ with a
slight excess of NaBH(OAc)3 and glacial acetic acid for 16 h
to provide amine 11 in 60% yield after chromatography.
Attempts to preform the imine with trimethylorthoformate58 or
titanium(IV) isopropoxide59-61 failed. Formation of the imine
in the presence of dehydrating agents (molecular sieves or
anhydrous CaSO4) or by the azeotropic removal of water
required long reaction times (80 °C, 1 week) and gave very
low yields.
Figure 1. Structure of farnesyl pyrophosphate (1, FPP), 6-(4-azido-
tetrafluorobenzylester)-3-methyl-2-hexene pyrophosphate (7), and transfer-
able FPP analogues 8-anilinogeranyl pyrophosphate (2, AGPP), 8-(p-nitro-
anilino)geranyl pyrophosphate (3, NAGPP), 8-(pentafluoroanilino)geranyl
pyrophosphate (4, PFAGPP), 8-(p-azido-N-ethyl-tetrafluoroanilino)geranyl
pyrophosphate (5, ETAZAGPP), and 6-(p-azido-N-methyl-tetrafluoro-
anilino)-3-methyl-2-hexenyl pyrophosphate (6, MTAZA2PP).
Successful reductive amination of the pentafluoroaniline
model system allowed us to proceed with the synthesis of the
target heterobifunctional photoprobes. The perfluorophenyl
azides were chosen as the light-activated moiety in this design
because the nitrene produced upon photolysis is likely to remain
in its singlet state and be more reactive toward C-H bond
insertion.62 This is important because the biologically relevant
binding sites for prenyl groups are anticipated to be hydrophobic.
We envisioned introduction of the azide function into the
(52) Park, H.-W.; Boduluri, S. R.; Moomaw, J. F.; Casey, P. J.; Beese, L. S.
Science 1997, 275, 1800-1804.
(53) Dunten, P.; Kammlott, U.; Crowther, R.; Weber, D.; Palermo, R.; Birktoft,
J. Biochemistry 1998, 37, 7907-7912.
(54) Long, S. B.; Casey, P. J.; Beese, L. S. Biochemistry 1998, 37, 9612-9618.
(55) Long, S. B.; Casey, P. J.; Beese, L. S. Structure 2000, 8, 209.
(56) 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 1998, 37,
16601.
Figure 2.
(57) Carlson, R.; Larsson, U.; Hansson, L. Acta Chem. Scand. 1992, 46, 1211-
having a pKa < 1 requires imine preformation by TiCl4 prior
to NaBH(OAc)3 reduction. The development of heterobifunc-
tional photoprobes 5 and 6 should allow for the study of protein
prenyltransferases and other enzymes that employ FPP or GGPP
as their substrates. In addition, incorporation of this analogue
into peptides or proteins may allow for the detection of other
unknown isoprenoid binding proteins.
1214.
(58) Look, G. C.; Murphy, M. M.; Campbell, D. A.; Gallop, M. A. Tetrahedron
Lett. 1995, 36, 2937-2940.
(59) Mattson, R. J.; Pham, K. M.; Leuck, D. J.; Cowen, K. A. J. Org. Chem.
1990, 55, 2552-2554.
(60) Bhattacharyya, S. Tetrahedron Lett. 1994, 35, 2401-2404.
(61) Bhattacharyya, S.; Chatterjee, A.; Williamson, J. S. Synlett 1995, 1079-
1080.
(62) Leyva, E.; Young, M. J. T.; Platz, M. S. J. Am. Chem. Soc. 1986, 108,
8307-8309.
9
8208 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
Scheme 1 a
a Reaction conditions: (1a) p-nitroaniline, NaBH(OAc)3, HOAc, (CH2Cl)2, or (2a) pentafluoroaniline or p-azidotetrafluoraniline, TiCl4, NaBH(OAc)3,
HOAc, (CH2Cl)2; (b) K2CO3, H2O, MeOH; (c) (Ph)3PCl2, Hu¨nig’s base, CH3CN; (d) [(n-Bu)4N]3HP2O7, CH3CN.
Scheme 2 a
a Reaction conditions: (a) MsCl, pyridine; (b) NaI, acetone; (c) sodium ethylacetoacetate; (d) (EtO)2POCH2CO2Et, NaH, THF; (e) DIBAL-H, THF; (f)
TBDMS-Cl, imidazole; (g) DDQ, CH2Cl2/pH 7.2 buffer; (h) (COCl)2, DMSO, Et3N; (i) p-azidotetrafluorobenzoyl chloride, pyridine; (j) (Ph)3PCl2, CH2Cl2;
(k) [(n-Bu)4N]3HP2O7, CH3CN; (l) (1) N-Me-p-azidotetrafluoroaniline, TiCl4, (2) HOAc, NaBH(OAc)3.
8-anilinogeranyl backbone by reductive amination of p-azido-
tetrafluoroaniline. However, using conditions developed for
pentafluoroaniline 11, we found it was not possible to isolate
pure azidoaniline 13. Azidoaniline 13 was found to be very
unstable, rapidly decomposing into deeply colored materials
during attempted isolation and purification. Previously, we had
shown that p-azidotetrafluoroaniline itself readily decomposes.63
We reasoned that introduction of an alkyl substituent onto the
aniline nitrogen would increase the stability of the desired
azidoaniline. Inspection of the six crystal structures of FTase
(63) Chehade, K. A. H.; Spielmann, H. P. J. Org. Chem. 2000, 65, 4949-4953.
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8209

A R T I C L E S
Chehade et al.
Scheme 3 a
a Reaction conditions: (a) MeI, LHDMS, THF; (b) 1 N HCl in HOAc.
with and without bound substrates50,52-56 suggested that small
alkyl substituents on the aniline nitrogen of 13 might yield
transferable analogues. Gribble and co-workers have described
mild conditions for the ethylation of secondary amines in situ
by alkoxy reducing agents, including NaBH(OAc)3.64,65 N-
Ethylation of secondary amines by NaBH(OAc)3 is slower than
the reduction of imines, suggesting that the desired tertiary amine
12 could be obtained in one pot by modification of the
conditions to form 11. Reaction of the p-azidotetrafluoroaniline
imine-TiCl4 complex with a 3-fold excess of NaBH(OAc)3 for
48 h gave N-ethyl azidoaniline 12 in 56% yield after chroma-
tography. Under these conditions, no azidoaniline 13 was
isolated from the reaction mixture, and, more importantly, the
tertiary amine dramatically enhanced the stability of azide 12.
Integration of the 1H NMR spectrum of amines 10-12
indicated that a mixture of cis/trans isomers about the 6,7 double
bond was formed in a 7:93 ratio. This was expected for amine
10, because an identical cis/trans mixture was previously
observed for amine 9.36 Interestingly, preformation of the
perfluoro imines and reduction in the presence of titanium did
not lead to any alteration of the cis/trans ratio for amines 11
and 12. The cis and trans isomers of acetate 11 were easily
separated by employing reverse-phase HPLC.
The allylic acetates 9-12 were elaborated into pyrophos-
phates 3-5 as previously described.36 The acetates 9-12 were
first saponified with K2CO3 in MeOH/H2O to generate alcohols
14-16, respectively, in high yield (89-95%). Allylic alcohols
14-16 were converted into their corresponding chlorides 17-
19 by dichlorotriphenylphosphorane (Ph3PCl2) in MeCN. Chlo-
ride 19 was obtained without any decomposition of the azide
group, highlighting the mildness of Ph3PCl2 as a chlorinating
agent. Because allylic chlorides have been found to be un-
stable,33,36 intermediates 17-19 were immediately converted to
pyrophosphates 3-5 by reaction with tris(tetra-n-butylammo-
nium) hydrogen pyrophosphate.66,67 Following purification by
ion-exchange chromatography and reverse-phase HPLC, pyro-
phosphates 3-5 were obtained in moderate yields (40-60%)
from alcohols 14-16, respectively.
An excess of ethylene glycol was monoalkylated with
p-methoxybenzyl chloride and KOH to provide ether 20 in 66%
yield after distillation. Alcohol 20 was transformed into its
mesylate 21 with methanesulfonyl chloride and pyridine fol-
lowed by conversion to iodide 22 in high yield via the
Finkelstein reaction. Ketone 23 was obtained in 62% yield after
distillation by reacting iodide 22 with sodium ethylacetoacetate
followed by saponification of the resulting ester and subsequent
decarboxylation. Condensation of ketone 23 with sodium
triethylphosphonoacetate afforded olefin 24 in 90% yield after
1
chromatography. Integration of H NMR spectra and analysis
of NOE experiments indicated that olefin 24 was produced in
a 3:1 ratio of E/Z isomers. Repeated attempts to separate the
isomers of ester 24 by column chromatography failed to improve
the E/Z ratio.
Reduction of ester 24 with DIBAL-H furnished allylic alcohol
25 in 95% yield. Alcohol 25 was converted into the TBDMS
ether 26 with TBDMS-Cl and imidazole. Alcohol 27 was
revealed in good yield by DDQ deprotection of methoxybenzyl
ether 26. In contrast to ester 24, the E and Z isomers of 27
were easily separated by flash chromatography. To prepare
photoprobe 6, oxidation of 27 to the corresponding aldehyde
followed by reductive amination was required. Swern oxidation
of the pure E isomer of alcohol 27 gave aldehyde 28 in 70%
yield after chromatography.
N-Methyl-p-azidotetrafluoroaniline 31 was prepared in 81%
overall yield by lithium bis(trimethylsilyl) amide deprotonation
of carbamate 29,63 alkylation with methyl iodide to form the
methyl carbamate 30, followed by removal of the N-BOC group
with HCl in acetic acid (Scheme 3). Reaction of equimolar
amounts of aniline 31 and aldehyde 28 with TiCl4, followed by
NaBH(OAc)3 reduction, afforded amine 34 in 60% yield after
isolation. Alcohol 27 also provides the entrance to photoana-
logue 6 using this scheme. Acylation of the E isomer of alcohol
68
27 with p-azidotetrafluorobenzoyl chloride in pyridine gave
benzoate 32 in high yield.
Alkyl and allylic TBDMS ethers have been converted directly
into their corresponding bromides using Ph3PBr2 or CBr4 and
PPh3, respectively.69,70 We have previously found that allylic
bromides are incompatible with the aniline functionality in these
molecules. Consequently, we utilized Ph3PCl2 in CH2Cl2 to
convert the allylic TBDMS ethers 32 and 34 into their
corresponding chlorides in 87% yield. As expected, no degrada-
tion of the aryl azides by the Ph3PCl2 reagent was detected.
Immediately after purification by flash chromatography, pyro-
phosphorylation of allylic chlorides 33 and 35 was effected with
(n-Bu4N)3HP2O7.66,67 Following ion-exchange chromatography
Synthesis of Photoprobes 6 and 7. The preparation of the
second class of FPP photoprobes 6 and 7 in which a flexible
methylene linker chain replaces the â-isoprene of FPP is outlined
in Scheme 2. Analogue 6 was designed to maximize structural
overlap with FPP by replacing the terminal isoprene with
N-methyl-p-azidotetrafluoroaniline and substitution of three
methylenes for the â-isoprene.
(64) Gribble, G. W.; Lord, P. D.; Skotnicki, J.; Dietz, S. E.; Eaton, J. T.; Johnson,
J. L. J. Am. Chem. Soc. 1974, 96, 7812.
(65) Gribble, G. W.; Nutaitis, C. F. Org. Prep. Proced. Int. 1985, 17, 317-
384.
(66) Davisson, V. J.; Woodside, A. B.; Poulter, C. D. Methods Enzymol. 1985,
110, 130-144.
(67) Davisson, V. J.; Woodside, A. B.; Neal, T. R.; Stremler, K. E.; Muehlbacher,
M.; Poulter, C. D. J. Org. Chem. 1986, 51, 4768-4779.
(68) Keana, J. F. W.; Cai, S. X. J. Org. Chem. 1990, 55, 3640-3647.
(69) Aizpurua, J. M.; Cossio, F. P.; Palomo, C. J. Org. Chem. 1986, 51, 4941-
4943.
(70) Mattes, H.; Benezra, C. Tetrahedron Lett. 1987, 28, 1697-1698.
9
8210 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
Table 1. Comparison of pKa of Aniline and Its Derivatives to
Table 2. HPLC Retention Time Comparison of Isoprenyl
Pyrophosphate and N-Dansyl-GC(lipid)VLS Peptide
Reaction Conditions and Yielda
Lewis
acid
HPLC tR N-dansyl-GC(lipid)VLS
aniline derivative
pK
yield
substrate
FPP (1)
AGPP (2)
PFAGPP (3)
NAGPP (4)
HPLC tRa (in min)
product
(in min)
a
aniline
4.6b
none 86%
none 64%
24.8
13.8
24.4
22.7
26.7
19.9
43
44
45
46
47
48
27.1
18.8
25.5
22.5
25.0
21.3
p-nitroaniline
pentafluoroaniline
p-azidotetrafluoroaniline
N-Me-p-azidotetrafluoroaniline
1.01b
(-0.277 ( 0.014),c -0.3d TiCl4 60%
-0.52e
TiCl4 56% f
TiCl4 60%
0.44e
ETAZAGPP (5)
MTAZA2PP (6)
a After isolation. b Reference 71. c Reference 72. d Reference 73. e Es-
timated, see Experimental Section. f N-Ethyl derivative.
a HPLC tR N-dansyl-GCVLS ) 13.4 min. Purified by analytical HPLC
performed on a Waters HPLC controller and a Waters 2487 dual λ
absorbance detector monitoring at 214 and 254 nm with a Western
Analytical 250 × 4.6 mm, 5 µm BioBasic8 C8 column. Compounds were
eluted under the following gradient at a flow rate of 1 mL/min: 0-5 min
100% A, 5-37 min 10% A, 37-40 min 10% A, 40-45 min 100% A,
45-50 min 100% A. Solvents: A ) 25 mM NH4HCO3; B ) CH3CN.
and reverse-phase HPLC, the photoprobes 6 and 7 were obtained
in 35% yield each from 32 and 34, respectively.
Reductive Amination Conditions Depend on Amine Basic-
ity. The basicity of the arylamine is key in choosing the
appropriate reductive amination conditions to prepare com-
pounds 2-6. Table 1 lists the anilines used in this study along
with their respective pKa values. Aniline and p-nitroaniline have
pKa’s greater than 1 and did not require imine preformation prior
to NaBH(OAc)3 reduction. These results suggest that conven-
tional reductive amination methodology should be employed
in the formation of allylic amines from R,â-unsaturated alde-
hydes when the reactant aniline has a pKa g 1.
Table 3. Steady-State Kinetic Parametersa
b
substrate
Km (nM)
kcat (s-1) × 10-2
k
cat/Km (M-1 s-1) × 105
V
rel
FPP (1)
AGPP (2)
PFAGPP (3)
NAGPP (4)
46 ( 2
46 ( 3
110 ( 8
82 ( 7
10 ( 0.3
8.5 ( 0.3
4.8 ( 0.2
4.2 ( 0.2
22 ( 1
1.0
18 ( 1
0.85
0.48
0.42
4.5 ( 0.4
5.1 ( 0.5
a Kinetic parameters for FPP, AGPP, PFAGPP, NAGPP measured as
described in the legend of Figure 3. Vrel refers to kcat/Km with respect to
b
FPP.
Analogues 2-6 Are Substrates for Farnesyltransferase In
Vitro. HPLC and mass spectrometric analysis were employed
to confirm that FPP and analogues 2-6 were transferred by
FTase to the N-dansyl-GCVLS peptide substrate. The study of
analogue 7 was discontinued because the ester was found to be
hydrolytically unstable. Samples for mass spectrometry were
obtained from 1 nmol scale transfer reactions for each of the
pyrophosphate substrates. Reverse-phase HPLC analysis of the
reaction mixtures indicated quantitative conversion of the parent
peptide to the N-dansyl-GC(lipid)VLS peptides 43-47 from FPP
1 and analogues 2-5, respectively. Analogue 5 was also
completely transferred to N-biotin-GCVLS peptide to form
N-biotin-GC(lipid)VLS 49. Only a 50% conversion to alkylated
peptide product 48 was achieved when analogue 6 was
employed under these conditions. The new peaks in the HPLC
chromatograms corresponding to the modified peptides were
isolated, and lipid transfer was confirmed by high-resolution
ESI mass spectrometry. These results indicate that transferable
molecules based on the anilinogeranyl skeleton can be created
that have additional functionality on the aromatic ring. As
expected, the products of analogue transfer catalyzed by FTase
to the peptide have increased hydrophobicity relative to the
unmodified peptide which is reflected in longer retention times
as compared to the parent peptide (Table 2).
Figure 3.
Lineweaver-Burke plots for FPP, AGPP, PFAGPP, and
NAGPP. Each value is the average of quadruplicate incubations in the
presence of the indicated concentration of FPP (9), AGPP (b), PFAGPP
(2), or NAGPP ([), and is representative of one experiment. These assays
were measured at 30 °C in 50 mM Tris‚HCl, pH 7.5, 5 mM MgCl2, 10 µM
ZnCl2, 5 mM DTT, 0.0125 mg/mL BSA, 0.04% n-dodecyl-â-D-maltoside
with 0.15 µM FTase, 1.0 µM dansyl-GCVLS, and varying concentrations
of pyrophosphate substrate distributed around the Km (Table 3). The initial
velocity was determined from the change in fluorescence. FLU represents
an arbitrary fluorescence unit.
Analogues 3 and 4 Are Transferred with Slightly Lower
Efficiency than FPP. We have utilized a continuous fluores-
cence assay to determine the steady-state kinetic parameters for
transfer of FPP and analogues 2-4 to the N-dansyl-GCVLS
pentapeptide (Table 3).74,75 FTase-catalyzed transfer of the
analogue to the N-dansyl-GCVLS peptide results in a time-
dependent increase in fluorescence of the dansyl group because
of a local increase in hydrophobic environment around the
chromophore. Initial reaction velocities were calculated from
the time-dependent increase in fluorescence for reactions of
FTase incubated with various concentrations of lipid pyrophos-
phates and 1.0 µM N-dansyl-GCVLS peptide (Figure 3).
These data were used to determine the kinetic parameters
Km, kcat, and kcat/Km of FPP and analogues 2-4 (Table 3).37 As
expected, AGPP (2) was found to transfer with near the same
value of (kcat/Km) as FPP.36 The analogues 3 and 4 are only
slightly poorer substrates than FPP and are transferred with
(71) Dean, J. A. Lange’s Handbook of Chemistry, 14th ed.; McGraw-Hill: New
York, 1992; pp 8.19-8.71.
(72) Chiang, Y.; Grant, A. S.; Kresge, A. J.; Paine, S. W. J. Am. Chem. Soc.
1996, 118, 4366-4372.
(73) Shoute, L. C. T.; Mittal, J. P.; Neta, P. J. Phys. Chem. 1996, 100, 11355-
11359.
(74) Pompliano, D. L.; Gomez, R. P.; Anthony, N. J. J. Am. Chem. Soc. 1992,
114, 7945-7946.
(75) Cassidy, P. B.; Dolence, J. M.; Poulter, C. D. Methods Enzymol. 1995,
250, 30-43.
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8211

A R T I C L E S
Chehade et al.
nearly identical kinetic parameters despite structurally different
ω-termini. There is no simple relationship between hydopho-
bicity and either the Km or the kcat for analogues 2-4. However,
it is interesting to note that while analogue 2 and FPP share
nearly identical kcat and Km, their hydrophobicities are quite
different. Although analogues 5 and 6 are transferred to
substrate, no increase in fluorescence was detected during the
assay. This observation is not unexpected, as previous studies
have shown that the fluorescence enhancement is highly
dependent on the structure of the transferred lipid.45
Ring Fluorination and Para Substituents Have Little
Effect on the Transfer Efficiency of the Analogue Pyro-
phosphates as Compared with FPP. For mammalian FTase,
substrate binding and product release are rate-limiting for steady-
state catalysis under conditions of subsaturating and saturating
substrate concentrations, respectively.76,77 The Km for FPP does
not simply reflect KD for FPP when peptide concentration is
saturating, assuming that FTase follows an ordered binding
scheme.78 The Km for FPP may reflect the binding of FPP to
the E‚product complex to accelerate product dissociation.77
Perfluorination of the aniline ring in 2 to give the isosteric
PFAGPP (3) has only a small effect on the reaction Km and
kcat, lowering the kinetic efficiency (kcat/Km) of transfer relative
to analogue 2 by a factor of 5. The higher Km and the smaller
kcat of pyrophosphate 3 relative to 2 suggest that formation of
the FTase‚3 complex is less favored and that the dissociation
rate constant of the enzyme-product complex is decreased
(Table 3). Although fluorocarbons are known to interact
unfavorably with both hydrocarbons and polar solvents,79 it
appears that fluorination does not significantly adversely affect
the intimate contacts between the ω-terminus of the lipid and
the FTase amino acid side chains in the binding site. NAGPP
4 is also accommodated by the FTase binding site in a
productive fashion despite conversion of the aniline from
electron donating to electron withdrawing and the increased
length relative to AGPP 2.
The quantitative FTase-catalyzed transfer of ETAZAGPP 5
to N-dansyl-GCVLS peptide indicates that fluorine substitution,
N-ethylation, and the addition of a para substituent allow the
analogue to bind to the enzyme in a productive fashion.
Interestingly, both ETAZAGPP 5 and MTAZA2PP 6 contain a
p-azidotetrafluoroanilino moiety in place of the FPP ω-isoprene,
but these analogues are transferred to the peptide substrate with
very different efficiencies. The 50% conversion of parent peptide
to MTAZA2PP 6 modified product is in sharp contrast to the
complete conversion that was observed for the benzyloxy
substituted variable length methylene chain series of analogues
38-42 under similar conditions.37 This implies that the flexible
linker chain and the electronic properties of the fluorinated
aniline ring compromise the kinetic efficiency of the reaction.
Modeling Indicates that the â-Isoprene of FPP and
Transferable Analogues 2-5 Are Essential for Efficient
Transfer. We have examined the potential FTase‚analogue
binding geometry of pyrophosphates 2-6 by docking them into
the active site of the FTase crystal structure using computer
Figure 4. Cross section of FTase active site surface with pyrophosphate
coordinating region in red, terminal isoprene “slot” in magenta, stick
rendering of FPP in black, and of ETAZAGPP 5 in white. Note how the
azide group of 5 (green) occupies the slot and how the aryl group
superimposes with the terminal isoprene of FPP. Fluorines are colored
yellow.
modeling as previously described.37 The procedure used to
model FPP into the FTase active site reproduced the conforma-
tion and geometry of the lipid found in the X-ray crystal
structures of the binary complexes.54,56 In these complexes,
electrostatic interactions between the pyrophosphate of FPP and
FTase residues Lys 164â, His 248â, Arg 291â, Lys 294â, and
Tyr 300â are calculated to be the dominant source of binding
energy, fixing the position of the pyrophosphate with respect
to the catalytic zinc ion. These electrostatic interactions were
also found to be the dominant source of binding energy for the
FTase models with analogues 2-6. In all of the models, the
pyrophosphate atoms of the analogues were found in essentially
the same location as those for FPP 1 in the X-ray crystal
structures. Examination of the X-ray crystal structures of the
two binary FTase‚FPP complexes shows that the terminal
isoprene methyl groups of FPP fill a slot defined by side chains
from Trp 102â, Tyr 205â, Phe 253â, Cys 254â, Trp 303â and
the aliphatic stretch of Arg 202â at the bottom of the active
site (Figure 4).54,56 Between the slot and the pyrophosphate
binding site, the FPP chain is fully extended and bound against
the side of the active site opposite that where the incoming
peptide or protein C-terminus binds. The location and confor-
mation of the R- and â-isoprene units of analogues 2-6 in these
models are coincident with those of the farnesyl chain in the
binary complex. Except for analogue 6, models of the complexes
with analogues 2-5 have the terminal edge or the para
substituents of the aniline group occupying the slot at the floor
of the FTase active site (Figure 4). Although analogue 6 appears
to be too short for the azidoaniline group to occupy the slot
when its pyrophosphate is in register with the zinc ion, solvent
access to the slot is blocked in the complex. The loss of these
van der Waals interactions and the subsequent void in the
complex may contribute to the lower efficiency of FTase-
catalyzed transfer of 6 to substrate.
(76) Furfine, E. S.; Leban, J. J.; Landavazo, A.; Moomaw, J. F.; Casey, P. J.
Biochemistry 1995, 34, 6857-6862.
(77) Tschantz, W. R.; Furfine, E. S.; Casey, P. J. J. Biol. Chem. 1997, 272,
9989-9993.
(78) Saderholm, M. J.; Hightower, K. E.; Fierke, C. A. Biochemistry 2000, 40,
12398-12405.
(79) Dunitz, J. D.; Taylor, R. Chem.-Eur. J. 1997, 3, 89-98.
9
8212 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
In each model of the analogue complexes, the face of the
aniline ring makes contact with numerous residues of the â-chain
of FTase. The aromatic ring is found in the volume occupied
by the terminal isoprene of FPP in the FTase‚FPP complex.
Interestingly, there are no qualitative differences between the
aniline ring-FTase contacts in AGPP 2 and the perfluoro
analogue 3. The perfluoro aromatic ring of ETAZAGPP 5 made
intimate contacts with side chains of FTase, but the N-ethyl
group protruded into the pocket in the low energy models
(Figure 4). There were no apparent specific hydrogen bonds
formed between the aniline ring nitrogen of analogues 2-6 and
the various functional groups in the pocket.
Utilizing the fluorescence transfer assay, we found that FTase
does not transfer the 20-carbon GGPP 36 to protein substrate
even though the longer isoprenoid is a competitive inhibitor of
the enzyme.37,80 However, previous studies with radiolabeled
GGPP have shown that detectable amounts of lipid can be
transferred to peptide substrate by FTase, albeit with very low
efficiency.80 Interestingly, NAGPP 4 and ETAZAGPP 5 are
transferred by FTase with high efficiency despite being one to
two atoms longer than FPP. In a previous study, we prepared a
series of flexible FPP analogues of varying length to examine
the origins of FTase discrimination for FPP 1 over the longer
GGPP 36.37 The flexible analogues 38-42 were constructed
by replacing the â-isoprene of FPP with a variable length
methylene chain and substituting a benzyloxy group for the
terminal isoprene (Figure 2). These analogues lacked the â and
γ double bonds and branched methyl groups of the isoprenoids
and ranged in length from the 10-carbon GPP 37 to one atom
shorter than GGPP 36 in their fully extended conformations.
All of these flexible analogues were transferred by FTase to
substrates, but with significantly reduced rates relative to FPP
1.
Figure 5. Photoinactivation of FTase with ETAZAGPP (5). FTase (10
ng) was photolyzed in the presence of 1.0 µM ETAZAGPP or 1.0 µM AGPP
with 366 nm light for 30 min at 0 °C. The UV light was removed, and the
samples were then incubated with 5 µM Ras and 1 µM [3H]FPP (33 000
dpm/pmol) at 37 °C for 30 min in a final volume of 25 µL. The amount of
[3H]-prenyl transferred to Ras was determined. The 100% of [3H]farnesyl
transferred was 1.1 pmol. Each value is the average of duplicate experiments.
efficiency than the flexible analogues of the same extended
length reinforces the suggestion that the conformational restric-
tions imposed on the lipid by the â-isoprene may act to pre-
organize the lipid in the pocket for transfer. X-ray structural
analysis has demonstrated that the peptide in the ternary com-
plex is found in intimate contact with the farnesyl lipid.55,56
Additional stabilization of the ternary complex for productive
transfer could result from van der Waals contacts between the
a2 and X residues of the peptide and the â-isoprene unit that
may be absent in the flexible analogues.
FPP 1, AGPP 2, PFAGPP 3, and the flexible analogue B3PP
39 all have the same length in their extended conformations.
However, B3PP 39 is transferred to substrate only 3% as
efficiently as FPP, while 1-3 are transferred with similar
efficiency (Table 2).37 Both NAGPP 4 and ETAZAGPP 5 retain
the conformationally restricted â-isoprene unit, and NAGPP 4
is transferred with twice the efficiency of the longer flexible
analogues B4PP 40 and B5PP 41.
Previously, the structural features that allowed the flexible
BnPP analogues 38-42 to be transferred while preventing
transfer of GGPP 36 were examined by modeling FPP 1, GGPP
36, and the flexible analogues into the active site of the FTase
crystal structure.37 The modeling indicated that within the
confines of the FTase pocket, the double bonds and branched
methyl groups of the geranylgeranyl chain significantly restrict
the number of possible conformations relative to the more
flexible lipid chain of the analogues. These conformational
restrictions prevent transfer of the GGPP by forcing it to partially
occlude the space occupied by the Ca1a2X peptide. It was further
concluded that the reduced efficiency of transfer for the flexible
analogues was likely due to the larger number of conformations
available to them as compared with the more rigid FPP isoprene
and that the lack of branched methyl groups may result in
suboptimal interactions in the ternary complex. The observation
that analogues 2-5 are transferred with a significantly higher
Transferable Azide Photoprobe 5 Photoinactivates FTase.
Photoprobe 5 was evaluated for its ability to inactivate FTase
upon UV irradiation with 366 nm light for 30 min at 0 °C. The
resulting samples were assayed for residual activity by monitor-
ing [3H]-farnesyl transfer from [3H]-FPP (Figure 5) to Ras. As
expected, decreased FTase activity was observed when either
analogue 2 or analogue 5 was incubated with Ras in the presence
of FTase without UV exposure because both are alternative
substrates for the enzyme. Transfer activity was further reduced
by exposure of analogue 5 and FTase to longwave UV light.
Irradiation of FTase at a saturating concentration of 5 (1 µM)
led to a 70% overall decrease in enzyme activity. No decrease
in enzyme activity was observed when FTase was irradiated
with 366 nm light for 30 min in the absence of lipid analogue.
These experiments indicate that FPP analogue 5 is a true
photoaffinity labeling reagent because 5 inhibited FTase in a
light-dependent manner.
Experimental Procedures
General Procedures. Caution! Organic azides should be considered
explosiVe, and all manipulations should take place behind a blast shield!
Aryl azides are light sensitiVe, and all reactions and flash chromatog-
raphy procedures should be conducted under diminished light. All
reactions were conducted under dry argon and stirred magnetically,
except as noted. Reaction temperatures refer to the external bath
temperatures, except as noted. Analytical TLC was performed on
precoated (0.25 mm) silica gel 60F-254 (Merck) plates and developed
with 30% ethyl acetate in hexane, except where noted otherwise.
(80) Reiss, Y.; Brown, M. S.; Goldstein, J. L. J. Biol. Chem. 1992, 267, 6403-
6408.
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8213

A R T I C L E S
Chehade et al.
Visualization was achieved either by UV irradiation, anisaldehyde-
sulfuric acid spray followed by heating, or by subjecting the plates to
a 5% ethanolic phosphomolybdic acid solution followed by heating.
Melting points are uncorrected. Flash chromatography was performed
on Merck silica gel 60 (230-400 mesh ASTM). All chromatography
solvents were purchased from VWR (EM Science-Omnisolv high purity
grade) and used as received. Anhydrous acetonitrile, pyridine, and DME
were purchased from Aldrich, and anhydrous THF was purchased from
Fluka; all other reagents were purchased either from Aldrich or from
Pfaltz and Bauer, unless otherwise noted. NMR spectra were obtained
in CDCl3 (unless otherwise noted) at 200 or at 400 MHz. Chemical
shifts for the following deuterated solvents are reported in ppm
downfield using the indicated reference peaks: CDCl3 (CDCl3 internal
peak, 7.27 ppm for H, 77.4 ppm for 13C), D2O (TSP, 0 ppm H and
13C; H3PO4 as an external reference, 0 ppm for 31P). Hexafluorobenzene
(C6F6) was used as an internal reference (-162.9 ppm, 19F). Electron
impact, FAB, and MALDI mass spectra were performed at the
University of Kentucky Mass Spectra Facility. Combustion analyses
were performed by Atlantic Microlabs, Inc., Norcross, GA. HPLC
pyrophosphate purifications were monitored at 254 nm. HPLC analysis
of fluorescence product studies was carried out using a Hewlett-Packard
Series 1100 system equipped with a UV (monitoring at 254 nm) and
fluorescence detector (340 nm excitation, 505 nm emission).
(m, 2F), -165.65 (m, 2F), -172.56 (tt, 1F, J ) 6.1 Hz). LRMS (EI)
m/z (relative intensity): 377 (5%, M+), 317 (10%, M+ - AcOH), 249
(100%). HRMS calcd for C18H20F5NO2, 377.1409; found, 377.1388.
8-(p-Azido-N-ethyl-tetrafluoroaniline)-3,7-dimethyl-1-acetoxyl-
2,6-octadiene (12). Into a 100 mL round-bottom flask was introduced
30 mL of CH2Cl2 and 1.5 mL of pyridine. The solution was then
deoxygenated by bubbling dry nitrogen through it for 15 min. After
the addition of 5 mL of TiCl4 (1 M solution in toluene, 5 mmol), the
mixture was stirred for 10 min before the addition of 1.08 g (5.24 mmol)
of p-azidotetrafluoroaniline.63 The mixture was stirred for 5 min at room
temperature, and a solution of 1.00 g (4.76 mmol) of aldehyde 8 in 2
mL of CH2Cl2 was then added in one portion to the reaction mixture.
The reaction was stirred for 8 h at room temperature (TLC: Rf 0.54,
imine) after which time 3 mL of glacial acetic acid was added. Stirring
was continued for an additional 10 min at room temperature before
the addition of 3.03 g (14.3 mmol) of NaBH(OAc)3. After the mixture
was stirred at room temperature for 24 h, an additional 3.03 g (14.3
mmol) of NaBH(OAc)3 and 1 mL of glacial acetic acid was added,
and the mixture was stirred for another 24 h at room temperature. The
reaction mixture was then quenched with 10 mL of water and stirred
for 30 min. The mixture was poured into 300 mL of Et2O and washed
(100 mL of water, 5% NaHCO3, water, and brine). The organics were
then dried (MgSO4), filtered, and concentrated to give a blood-red oil.
Purification by flash chromatography (5% EtOAc in hexane) yielded
1.15 g (56%) of a pale-yellow oil that rapidly turned pale-orange.
TLC: (Rf 0.62). 1H NMR (400 MHz): δ 5.32 (m, 2H), 4.57 (d, 2H, J
) 6.8 Hz), 3.61 (s, 2H), 3.09 (q, 2H, J ) 7.2 Hz), 2.33 (m, 2H), 2.05
(s, 3H), 2.04 (m, 2H), 1.69 (s, 3H), 1.59 (s, 3H), 1.01 (t, 3H, J ) 7.2
Hz). 13C NMR (100.7 MHz): δ 171.33, 146.24 (m), 143.79 (m), 142.45
(m), 141.93, 140.06 (m), 132.69, 127.88, 125.88 (m), 118.73 (s + m),
61.51, 61.23 (m), 46.95 (m), 39.28, 26.06, 21.22, 16.56, 14.32, 13.01.
19F NMR (376.7 MHz): δ -148.66 (m, 2F), -154.32 (m, 2F). IR
(KBr): 2122, 1740, 1492. LRMS (EI) m/z (relative intensity): 428
(45%, M+), 402 (100%, M+ - N2 + 2H), 249 (70%). HRMS calcd for
C20H24F4N4O2, 428.1830; found, 428.1821.
1
1
8-(p-Nitroaniline)-3,7-dimethyl-1-acetoxyl-2,6-octadiene (10). Into
a 250 mL three-neck flask was introduced 150 mL of 1,2-dichloro-
ethane, followed by 7.88 g (37.48 mmol) of 8, 5.69 g (41.22 mmol) of
p-nitroaniline, and 2.58 mL (44.89 mmol) of glacial acetic acid. After
stirring for 1 min at room temperature, 11.12 g (52.47 mmol) of NaBH-
(OAc)3 was added to the above solution. The solution was stirred at
room temperature overnight, then quenched by pouring into a separatory
funnel containing 200 mL of 5% NaHCO3. The product was extracted
with ether (3 × 75 mL), dried (MgSO4), concentrated, and purified by
flash chromatography (10% EtOAc in hexane) yielding 7.97 g (64%)
1
of a bright lemon yellow oil. TLC: (Rf 0.36). H NMR (CDCl3, 400
MHz): δ 8.06 (m, 2H), 6.53 (m, 2H), 5.33 (m, 2H), 4.82 (bs, 1H),
4.57 (d, 2H, J ) 6.8 Hz), 3.74 (d, 2H, J ) 5.2 Hz), 2.18 (m, 2H),
2.10-2.03 (m, 5H), 1.69 (s, 3H), 1.66 (s, 3H). 13C NMR (CDCl3, 100.7
MHz): δ 171.36, 153.87, 141.47, 137.89, 131.09, 126.65, 126.47,
118.99, 111.34, 61.55, 51.02, 39.10, 25.90, 21.19, 16.56, 14.69. LRMS
(EI) m/z (relative intensity): 332 (20%, M+), 204 (100%). HRMS calcd
for C18H24N2O4, 332.1736; found, 332.1751.
General Saponification Conditions for Compounds 14-16. To a
100 mL round-bottom flask were added the acetate (5.56 mmol) and
50 mL of methanol. Potassium carbonate (2.31 g, 16.7 mmol) dissolved
in 10 mL of water was added, and the reaction mixture was stirred at
room temperature overnight. Water was added (20 mL), and the mixture
was extracted with CH2Cl2 (3 × 100 mL). The combined extracts were
washed once with brine (100 mL), dried (MgSO4), and concentrated.
8-(p-Nitroaniline)-3,7-dimethyl-2,6-octadien-1-ol (14). 1.50 g (93%)
8-(Pentafluoroaniline)-3,7-dimethyl-1-acetoxyl-2,6-octadiene (11).
Into a 100 mL round-bottom flask was introduced 30 mL of CH2Cl2
and 1.5 mL of pyridine. The solution was then deoxygenated by
bubbling dry nitrogen through it for 15 min. After the addition of 5
mL of TiCl4 (1 M solution in toluene, 5 mmol), the mixture was stirred
for 10 min before the addition of 0.96 g (5.24 mmol) of pentafluoro-
aniline. The mixture was stirred for 5 min at room temperature, and a
solution of 1.00 g (4.76 mmol) of aldehyde 8 in 2 mL of CH2Cl2 was
then added in one portion to the reaction mixture. The reaction was
stirred for 8 h at room temperature (TLC: Rf 0.54, imine) after which
time 3 mL of glacial acetic acid was added. Stirring was continued for
an additional 10 min at room temperature before the addition of 3.03
g (14.3 mmol) of NaBH(OAc)3. After the reaction mixture was stirred
at room temperature overnight, it was quenched with 50 mL of saturated
NH4Cl and stirred for 30 min. The mixture was then poured into 300
mL of Et2O and washed (100 mL of water, 5% NaHCO3, water, and
brine). The organics were dried (MgSO4), filtered, concentrated, and
purified by flash chromatography (5% EtOAc in hexane) affording 1.08
1
of a bright lemon yellow oil. TLC: (Rf 0.10). H NMR (CDCl3, 400
MHz): δ 8.07 (m, 2H), 6.53 (m, 2H), 5.38 (m, 2H), 4.90 (bs, 1H),
4.15 (d, 2H, J ) 6.4 Hz), 3.73 (s, 2H), 2.18 (m, 2H), 2.05 (t, 2H, J )
7.6 Hz), 1.66 (bs, 6H). 13C NMR (CDCl3, 100.7 MHz): δ 153.88,
138.99, 137.92, 131.06, 126.97, 126.52, 123.99, 111.34, 59.47, 51.12,
39.15, 25.98, 16.37, 14.78. LRMS (EI) m/z (relative intensity): 290
(70%, M+), 204 (100%). HRMS calcd for C16H22N2O3, 290.1630; found,
290.1640.
8-(Pentafluoroaniline)-3,7-dimethyl-2,6-octadien-1-ol (15). 1.78 g
1
(96%) of a pale-yellow oil. TLC: (Rf 0.20). H NMR (400 MHz): δ
5.37 (m, 2H), 4.15 (d, 2H, J ) 7.0 Hz), 3.79 (s, 2H), 2.31 (bs, 2H),
2.14 (q, 2H, J ) 7.0 Hz), 2.03 (t, 2H, J ) 7.2 Hz), 1.67 (s, 3H), 1.65
(s, 3H). 13C NMR (100.7 MHz): δ 139.71 (m), 139.47, 137.30 (m),
135.23 (m), 132.77 (s + m), 127.61 (s + m), 124.08 (s + m), 59.70,
54.33 (m), 39.39, 26.29, 16.58, 14.70. 19F NMR (376.7 MHz): δ
-159.82 (d, 2F, J ) 22 Hz), -165.61 (m, 2F), -172.37 (m, 1F). LRMS
(EI) m/z (relative intensity): 335 (15%, M+), 249 (85%), 196 (100%).
HRMS calcd for C16H18F5NO, 335.1303; found, 335.1311.
8-(p-Azido-N-ethyl-tetrafluoroaniline)-3,7-dimethyl-2,6-octadien-
1-ol (16). 2.08 g (97%) of a viscous, orange colored oil. TLC: (Rf
0.20). 1H NMR (400 MHz): δ 5.35 (m, 2H), 4.14 (d, 2H, J ) 6.8 Hz),
3.61 (s, 2H), 3.09 (q, 2H, J ) 7.2 Hz), 2.13 (m, 4H), 1.66 (s, 3H),
1.60 (s, 3H), 1.34 (bs, 1H), 1.01 (t, 3H, J ) 7.2 Hz). 13C NMR (100.7
1
g (60%) of a bright lemon yellow oil. TLC: (Rf 0.56). H NMR (400
MHz): δ 5.32 (m, 2H), 4.58 (d, 2H, J ) 7.1 Hz), 3.79 (s, 2H), 3.71
(bs, 1H), 2.14 (q, 2H, J ) 7.9 Hz), 2.06 (s, 3H), 2.04 (t, 2H, J ) 7.0
Hz), 1.69 (s, 3H), 1.65 (s, 3H). 13C NMR (100.7 MHz): δ 171.52,
141.88, 139.69 (m), 137.22 (m), 135.18 (m), 132.89, 132.75 (m),
127.40, 127.36 (m), 124.07 (m), 119.08, 119.02 (m), 61.68, 54.29 (m),
39.37, 26.20, 21.40, 16.77, 14.68. 19F NMR (376.7 MHz): δ -159.98
9
8214 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
MHz): δ 146.17 (m), 143.72 (m), 142.44 (m), 139.97 (m), 139.39,
132.57, 128.03, 125.86 (m), 123.79, 114.13 (m), 61.24 (m), 59.51, 46.98
(m), 39.30, 26.15, 16.35, 14.30, 13.02. 19F NMR (376.7 MHz): δ
-149.35 (m, 2F), -155.05 (m, 2F). IR (KBr): 2122, 1492. LRMS
(EI) m/z (relative intensity): 386 (15%, M+), 358 (M+ - N2, 5%),
135 (60%), 93 (100%). HRMS calcd for C18H22F4N4O, 386.1730; found,
386.1737.
General Halogenation Conditions for Compounds 17-19. Into a
100 mL round-bottom flask were added the allylic alcohol (5.05 mmol),
N,N-diisopropylethylamine (1.41 mL, 8.09 mmol), and 30 mL of dry
acetonitrile. The solution was stirred at 0 °C for 10 min. Solid dichloro-
triphenylphosphorane (2.45 g, 7.58 mmol) was then added evenly to
the reaction mixture over a 7 min period. After the final addition of
Ph3PCl2, the reaction was allowed to stir at 0 °C for an additional 40
min. The mixture was then loaded directly onto a silica gel column
and purified by flash chromatography (5% EtOAc in hexane).
8-(p-Nitroaniline)-3,7-dimethyl-1-chloro-2,6-octadiene (17). 1.17
g (75%) of a bright lemon yellow oil. TLC: (Rf 0.56).
-9.16 (m, 1P, J ) 22 Hz), -9.71 (m, 1P, J ) 21 Hz). LRMS (FAB-,
glycerol matrix): calculated (M - H)-, 494.1; measured, 494.1.
8-(p-Azido-N-ethyl-tetrafluoroaniline)-3,7-dimethyl-2,6-octadi-
ene Pyrophosphate (5). White solid (57%). 1H NMR (D2O, 400
MHz): δ 5.40 (t, 1H, J ) 7.0 Hz), 5.35 (t, 1H, J ) 6.2 Hz), 4.47 (t,
2H, J ) 6.2 Hz), 3.62 (s, 2H), 3.11 (q, 2H, J ) 7.3 Hz), 2.08 (q, 2H,
J ) 6.8 Hz), 1.94 (t, 2H, J ) 7.3 Hz), 1.67 (s, 3H), 1.60 (s, 3H), 0.99
(t, 3H, J ) 7.1 Hz). 13C NMR (D2O, 100.7 MHz): δ 149.77 (m), 147.42
(m), 145.53, 144.95 (m), 142.51 (m), 134.66, 132.77 (m), 125.67 (m),
122.36 (m), 119.27 (m), 65.64 (d, J ) 5.3 Hz), 65.10 (m), 51.42 (m),
41.20, 28.21, 18.41, 17.01, 14.72. 19F NMR (D2O, 376.7 MHz): δ
-147.07 (m, 2F), -153.99 (m, 2F). 31P NMR (D2O, 162.1 MHz): δ
-8.76 (m, 1P, J ) 20 Hz), -9.75 (m, 1P, J ) 21 Hz). LRMS (FAB-,
glycerol matrix): calculated (M - H)-, 545.1; measured, 545.1. HRMS
(ESI): calculated (M + H)+ C18H25F4N4O7P2, 547.1135; measured,
547.1124.
(p-Methoxybenzyloxy)ethanol (20). Solid potassium hydroxide
(86%, 2.4 g, 3.8 mmol) was added to ethylene glycol (21 mL, 38 mmol)
and stirred until dissolution. The temperature was then increased to
130 °C (internal temperature) and maintained at this temperature until
the water stopped collecting in the air condenser (ca. 2-3 h). The
reaction was cooled to room temperature, and p-methoxybenzyloxy
chloride (6 g, 3.8 mmol) was added. The temperature was then raised
to 35 °C and kept at this point overnight. The reaction mixture was
then cooled, diluted with water (80 mL), and extracted with Et2O. The
extract was washed with water, dried (MgSO4), filtered, and evaporated.
The residue was purified by distillation under reduced pressure to afford
4.6 g (66%) of 20 as a colorless oil: bp 132 °C/0.35 mmHg (lit.71
8-(Pentafluoroaniline)-3,7-dimethyl-1-chloro-2,6-octadiene (18).
1.48 g (83%) of a yellow oil. TLC: (Rf 0.56).
8-(p-Azido-N-ethyl-tetrafluoroaniline)-3,7-dimethyl-1-chloro-2,6-
octadiene (19). 1.74 g (85%) of a pale-yellow oil which rapidly turned
orange. TLC: (Rf 0.68).
General Pyrophosphorylation Conditions for Compounds 3-5.
Into a 50 mL glass centrifuge tube was added 7.45 g (7.6 mmol) of
tris(tetrabutylammonium) hydrogen pyrophosphate and 12 mL of dry
acetonitrile. After vortexing, the milky white solution was then
centrifuged at 2000 rpm for 10 min. The clear supernatant was decanted
into a flame-dried, 100 mL, round-bottomed flask charged with the
allylic chloride (3.8 mmol). The solution was allowed to stir at room
temperature for 4 h. Solvent was removed, and the residue was dissolved
in 3 mL of ion-exchange buffer (ion-exchange buffer was generated
by dissolving ammonium bicarbonate (2.0 g, 25.3 mmol) in 1.0 L of
2% (v/v) isopropyl alcohol/water). The resulting milky solution was
loaded onto a preequilibrated 2 × 30 cm column of Dowex AG 50W-
X8 (100-200 mesh) cation-exchange resin (NH4+ form). The flask was
washed twice with 5 mL of buffer, and both washes were loaded onto
the column before elution with 190 mL (two column volumes) of ion-
exchange buffer. The cloudy eluant was collected in a 600 mL freeze-
drying flask, frozen, and lyophilized. A portion of the crude pyrophos-
phate was dissolved in 25 mM NH4HCO3 to a final concentration of
ca. 5 mM. The solution was loaded onto a preparative Western
Analytical 250 × 10 mm, 5 µm BioBasic8 C8 column and eluted under
the following gradient at a flow rate of 4 mL/min: 0-11 min 100%
A, 11-12 min 80% A, 12-24 min 80% A, 24-27 min 5% A, 27-31
min 5% A, 31-36 min 100% A, 36-45 min 100% A. Solvents, A )
25 mM NH4HCO3; B ) 2-propanol:acetonitrile (1:1 v/v). The desired
peak was collected, frozen, and lyophilized.
1
145-146 °C/5 mmHg). H NMR (200 MHz): δ 7.32-7.20 (m, 2H),
6.94-6.80 (m, 2H), 4.48 (s, 2H), 3.79 (s, 3H), 3.75-3.70 (m, 2H),
3.58-3.53 (m, 2H), 2.26 (bs, 1H). 13C NMR (50 MHz): δ 159.65,
130.41, 129.77, 114.19, 73.26, 71.44, 62.18, 55.59.
(p-Methoxybenzyloxy)ethyl Methanesulfonate (21). To a cooled
(0 °C) solution of the alcohol 20 (14.8 g, 81 mmol) in anhydrous
pyridine (34 mL) was added methanesulfonyl chloride (9.4 mL, 121.5
mmol). The reaction mixture was warmed to room temperature, stirred
for 2 h, and then diluted with water and extracted with Et2O. The
organic solution was washed (10% HCl, then water, then 10% NaHCO3,
then water), dried (MgSO4), filtered, and evaporated to obtain 16.85 g
(80%) of a colorless oil that was used in the next step without further
purification. The analytical sample was obtained by flash chromatog-
1
raphy (20% EtOAc in hexane) to give 21 as a colorless oil. H NMR
(400 MHz): δ 7.32-7.21 (m, 2H), 6.94-6.84 (m, 2H), 4.51 (s, 2H),
4.41-4.36 (m, 2H), 3.81 (s, 3H), 3.73-3.69 (m, 2H), 3.03 (s, 3H).
13C NMR (100 MHz): δ 159.77, 129.83, 129.77, 114.24, 73.35, 69.57,
67.82, 55.64, 38.05. LRMS (EI) m/z (relative intensity): 260 (65%,
M+), 121 (100%). HRMS calcd for C11H16O5S, 260.0699; found,
260.0699. Anal. Calcd for C11H16O5S: C, 50.76; H, 6.19; S, 12.32.
Found: C, 50.60; H, 6.05; S, 12.10.
8-(p-Nitroaniline)-3,7-dimethyl-2,6-octadiene Pyrophosphate (3).
1
(p-Methoxybenzyloxy)ethyl Iodide (22). To a solution of NaI (32
g, 213 mmol) in acetone (140 mL) was added at room temperature a
solution of the methanesulfonate 21 (18.4 g, 71 mmol) in acetone (30
mL). After the mixture was stirred at room temperature overnight, water
was added, and the product was extracted with Et2O. The organic phase
was washed (10% Na2S2O3, then water), dried (MgSO4), filtered, and
evaporated to obtain 18.66 g (90%) of the desired iodide as a slightly
yellow oil that was used without further purification. The analytical
sample was purified by flash chromatography (10% EtOAc in hexane)
Bright lemon yellow solid (54%). H NMR (D2O, 400 MHz): δ 8.09
(m, 2H), 6.69 (m, 2H), 5.42 (m, 2H), 4.46 (t, 2H, J ) 7.0 Hz), 3.81 (s,
2H), 2.21 (q, 2H, J ) 6.8 Hz), 2.10 (t, 2H, J ) 7.3 Hz), 1.70 (s, 3H),
1.65 (s, 3H). 13C NMR (D2O, 100.7 MHz): δ 158.26, 145.37, 138.84,
134.51, 129.85, 128.75, 122.88 (d, J ) 9.2 Hz), 114.50, 65.55 (d, J )
5.3 Hz), 52.45, 41.35, 28.12, 18.47, 16.46. 31P NMR (D2O, 162.1
MHz): δ -7.65 (d, 1P, J ) 22 Hz), -9.62 (d, 1P, J ) 21 Hz). LRMS
(FAB-, glycerol matrix): calculated (M - H)-, 449.1; measured, 449.1.
8-(Pentafluoroaniline)-3,7-dimethyl-2,6-octadiene Pyrophosphate
1
1
to give iodide 22 as a colorless oil. H NMR (200 MHz): δ 7.31-
(4). White solid (60%). H NMR (D2O, 400 MHz): δ 5.37 (dt, 2H, J
7.25 (m, 2H), 6.91-6.86 (m, 2H), 4.51 (s, 2H), 3.81 (s, 3H), 3.71 (t,
2H, J ) 6.8 Hz), 3.26 (t, 2H, J ) 6.8 Hz). 13C NMR (50 MHz): δ
159.81, 130.27, 129.76, 114.26, 72.84, 70.81, 55.58, 3.28. LRMS (EI)
m/z (relative intensity): 292 (52%, M+), 121 (100%). HRMS calcd
for C10H13O2I, 291.9955; found, 291.9947.
) 7.0 Hz), 4.46 (t, 2H, J ) 6.6 Hz), 3.72 (s, 2H), 2.11 (q, 2H, J ) 7.3
Hz), 1.99 (t, 2H, J ) 7.1 Hz), 1.67 (s, 3H), 1.62 (s, 3H). 13C NMR
(D2O, 100.7 MHz): δ 145.66, 143.65 (m), 141.90 (m), 141.23 (m),
139.45 (m), 135.39, 130.99, 125.27 (m), 122.36 (d, J ) 7.7 Hz), 65.68
(d, J ) 5.3 Hz), 56.64 (m), 41.29, 28.24, 18.40, 16.10. 19F NMR (D2O,
376.7 MHz): δ -155.86 (d, 2F, J ) 21 Hz), -164.73 (t, 2F, J ) 21
Hz), -168.47 (t, 1F, J ) 22 Hz). 31P NMR (D2O, 162.1 MHz): δ
5-(p-Methoxybenzyloxy)-pentan-2-one (23). To a solution of
ethylacetoacetate sodium salt (19.5 g, 128 mmol) in anhydrous DME
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8215

A R T I C L E S
Chehade et al.
(180 mL) at room temperature was added a solution of the iodide 22
(18.7 g, 64 mmol) in anhydrous DME (15 mL), and the reaction mixture
was heated at reflux overnight. After cooling to room temperature, a
solution of NaOH (14.2 g, 350 mmol) in water (142 mL) was added,
the mixture was refluxed for 2 h, cooled to room temperature, acidified
with 50% H2SO4 (40 mL, final pH ≈ 2), and then refluxed for 2 h
more. Water was added, and the product was extracted with Et2O. The
organic phase was washed (10% NaHCO3 then water), dried (MgSO4),
filtered, and evaporated to obtain an oily residue. Purification by
distillation under reduced pressure gave 8.81 g (62%) of ketone 23 as
at room temperature overnight. The reaction mixture was then diluted
with water and extracted with hexane. The combined organic layers
were washed with water, dried (MgSO4), filtered, and evaporated.
Purification by flash chromatography (25% Et2O in hexane) afforded
10.25 g (95%) of silyl ether 26 as a colorless oil. 1H NMR (400 MHz)
(isomer E): δ 7.28-7.24 (m, 2H), 6.89-6.86 (m, 2H), 5.33-5.29 (m,
1H), 4.42 (s, 2H), 4.18 (dd, 2H, J ) 6.2 Hz, J ) 0.4 Hz), 3.80 (s, 3H),
3.43 (t, 2H, J ) 6.4 Hz), 2.08-2.04 (m, 2H), 1.76-1.69 (m, 2H), 1.61
(d, 3H, J ) 0.8 Hz), 0.9 (s, 9H), 0.06 (s, 6H). 13C NMR (100 MHz)
(isomer E): δ 159.48, 136.83, 131.09, 129.61, 124.99, 114.13, 72.94,
70.12, 60.66, 55.64, 36.34, 28.16, 26.39, 18.79, 16.66, -4.66. LRMS
(CI) m/z (relative intensity): 365 (5%, MH+), 307 (40%, M - C4H9+),
243 (65%, M - C8H9O+). Anal. Calcd for C21H36O3Si: C, 69.18; H,
9.95. Found: C, 69.23; H, 9.84.
1-(tert-Butyldimethylsilyloxy)-3-methyl-hex-2-en-6-ol (27). Solid
DDQ (480 mg, 2.1 mmol) was added in one portion to a solution of
26 (533 mg, 1.46 mmol) in 5.5 mL of 0.1 M sodium phosphate buffer
(pH 7.2) and CH2Cl2 (55 mL), and the resultant mixture was vigorously
stirred for 2.5 h at room temperature. The mixture was then filtered
through a silica gel pad, and the filter cake was washed with 1:1 (v/v)
ether:hexane. The colorless filtrate was concentrated and purified by
flash chromatography (5% EtOAc in hexane, then 10% EtOAc in
hexane) to give 186 mg (52%) of 27 as Z isomer, and 62 mg (17%) of
27 as E isomer.
1
a pale-yellow oil: bp 110 °C/0.25 mmHg. H NMR (200 MHz): δ
7.27-7.22 (m, 2H), 6.89-6.85 (m, 2H), 4.41 (s, 2H), 3.80 (s, 3H),
3.45 (t, 2H, J ) 6.2 Hz), 2.53 (t, 2H, J ) 7.1 Hz), 2.13 (s, 3H), 1.93-
1.80 (qui, 2H). 13C NMR (50 MHz): δ 209.02, 159.56, 130.90, 129.61,
114.16, 72.91, 69.37, 55.66, 40.74, 30.32, 24.29. IR (KBr): 1716, 1514,
1243. LRMS (EI) m/z (relative intensity): 222 (5%, M+), 121 (100%).
HRMS calcd for C13H18O3, 222.1250; found, 222.1258. Anal. Calcd
for C13H18O3: C, 70.24; H, 8.16. Found: C, 70.31; H, 8.14.
Ethyl 6-(p-Methoxybenzyloxy-3-methyl-hex-2-enoate (24). Tri-
ethylphosphonoacetate (16.6 g, 74 mmol) was added dropwise to a
stirred suspension of NaH (1.8 g, 74 mmol) in dry THF (45 mL) at 0
°C and allowed to stir for 30 min. A solution of the ketone 23 (8.3 g,
37 mmol) in dry THF (30 mL) was then added at the same temperature.
After the reaction was allowed to warm to room temperature and stirred
overnight, it was then diluted with saturated NH4Cl and extracted with
Et2O. The ether layer was washed with water, dried (MgSO4), filtered,
and evaporated to give an oily residue. Purification by flash chroma-
tography on silica gel (20% EtOAc in hexane) gave 9.73 g (90%) of
olefin 24 as a colorless oil and as an E/Z mixture in the ratio ∼75:25
which was used to the next reaction without further separation. 1H NMR
(400 MHz) (isomers Z and E): δ 7.27-7.24 (m, 2H), 6.88-6.86 (m,
2H), 5.67-5.66 (m, 1H), 4.43 and 4.41 (two singlets, 2H, isomers Z
and E, respectively), 4.16-4.09 (m, 2H), 3.79 (s, 3H), 3.49 and 3.43
(two triplets, 2H, J ) 6.6 Hz for isomer Z and J ) 6.2 Hz for isomer
E), 2.70-2.66 (m, 2H from isomer Z), 2.24-2.01 (m, 2H, from isomer
E), 2.15 and 1.88 (two dublets, 3H, J ) 1.2 Hz for isomer E and J )
1.2 Hz for isomer Z), 1.82-1.73 (m, 2H), 1.29-1.22 (m, 3H). 13C NMR
(100 MHz) (isomers Z and E): δ 167.11, 159.73, 159.49, 130.81,
129.56, 116.71, 116.09, 114.09, 114.04, 72.92, 72.88, 70.29, 69.42,
59.77, 55.54, 37.79, 30.47, 30.02, 28.62, 27.82, 25.52, 19.03, 14.64.
IR (KBr): 1708, 1508, 1243. LRMS (EI) m/z (relative intensity): 292
(5%, M+), 121 (100%). HRMS calcd for C17H24O4, 292.1645; found,
292.1645.
Z-1-(tert-Butyldimethylsilyloxy)-3-methyl-hex-2-en-6-ol (Z-27). Rf
1
0.56. H NMR (200 MHz): δ 5.50-5.40 (m, 1H), 4.15 (d, 2H, J )
7.2 Hz), 3.61-3.55 (m, 2H), 2.22 (t, 2H, J ) 7 Hz), 1.73 (s, 3H),
1.72-1.50 (m, 2H), 0.91 (s, 9H), 0.09 (s, 6H).
E-1-(tert-Butyldimethylsilyloxy)-3-methyl-hex-2-en-6-ol (E-27). Rf
1
0.52. H NMR (200 MHz): δ 5.40-5.31 (m, 1H), 4.09 (dd, 2H, J )
6.2 Hz, J ) 0.6 Hz), 3.65 (m, 2H), 2.12-2.05 (m, 2H), 1.72 (m, 2H),
1.64 (s, 3H), 1.57 (bs, 1H), 0.90 (s, 9H), 0.07 (s, 6H). 13C NMR (50
MHz): δ 137.11, 125.25, 63.10, 60.55, 36.17, 30.96, 26.33, 18.72,
16.55, -4.77. LRMS (CI) m/z (relative intensity): 245 (10%, MH+),
243 (5%, M - H+), 229 (7%, M - CH3+), 187 (100%, M - C4H9+).
Anal. Calcd for C13H28O2Si: C, 63.88; H, 11.54. Found: C, 63.73; H,
11.58.
6-(tert-Butyldimethylsilyloxy)-3-methyl-hex-2-enal (28). The E
isomer of alcohol 27 (218 mg, 0.89 mmol) in 3 mL of CH2Cl2 was
added to a premixed solution of oxalyl chloride (0.15 mL, 1.37 mmol)
and DMSO (0.19 mL, 2.67 mmol) in 5 mL of CH2Cl2 at -78 °C. The
resulting mixture was stirred for 30 min, and triethylamine (0.65 mL,
4.7 mmol) was then added. After being stirred for an additional 15
min at -78 °C, the reaction was allowed to warm to room temperature
and was stirred for 30 min longer. The reaction mixture was then
quenched with water and extracted with CH2Cl2. The combined extracts
were washed (water then brine), dried (MgSO4), and concentrated.
Purification by flash chromatography (hexane/diethyl ether 8:1) afforded
142 mg (66%) of aldehyde 28 as a colorless oil. 1H NMR (400 MHz):
δ 9.78 (t, J ) 1.8 Hz, 1H), 5.32 (m, 1H), 4.18 (m, 2H), 2.56 (m, 2H),
6-(p-Methoxybenzyloxy)-3-methyl-hex-2-enol (25). To the solution
of ester 24 (9.6 g, 33 mmol) in 150 mL of dry THF under argon at
-78 °C was added diisobutyl aluminum hydride (1.5 M solution in
toluene, 88 mL, 132 mmol). The reaction was stirred at -78 °C for 2
h. A 10% (w/v) solution of potassium sodium tartrate was added, and
the reaction mixture was allowed to warm to room temperature. After
dilution with water, the mixture was extracted with Et2O (2×). The
combined organic layers were washed (water then brine), dried over
MgSO4, and evaporated to obtain 7.85 g (95%) of an oily residue that
was used in the next step without further purification. The analytical
sample was purified by flash chromatography (30% Et2O in hexane)
2.34 (t, 2H, J ) 7.6 Hz), 1.65 (m, 3H), 0.89 (s, 9H), 0.06 (s, 6H). 13
C
NMR (50 MHz): δ 216.52, 135.16, 125.88, 60.26, 42.29, 31.93, 26.39,
18.80, 16.86, -4.70. IR (KBr): 1728. LRMS (EI) m/z (relative
intensity): 241 (2%, M - H+), 185 (55%, M - tBu), 75 (100%). HRMS
calcd for C12H23O2Si (M - Me), 227.1462; found, 227.1423. C9H17O2-
1
giving alcohol 25 as a colorless oil. H NMR (200 MHz) (isomer E):
t
δ 7.28-7.24 (m, 2H), 6.89-6.86 (m, 2H), 5.44-5.35 (m, 1H), 4.42
(s, 2H), 4.12 (d, 2H, J ) 7 Hz), 3.80 (s, 3H), 3.43 (t, 2H, J ) 6.6 Hz),
2.14-2.01 (m, 2H), 1.80-1.67 (m, 2H), 1.66 (s, 3H). 13C NMR (50
MHz) (isomer E): δ 159.42, 139.56, 130.95, 129.42, 123.81, 113.98,
72.67, 69.66, 59.49, 55.38, 36.08, 27.85, 16.22. LRMS (EI) m/z (relative
intensity): 250 (3.5%, M+), 121 (100%). Anal. Calcd for C15H22O3:
C, 71.97; H, 8.86. Found: C, 71.76; H, 8.86.
Si (M - Bu), 185.0907; found, 185.0992.
4-(N-Methyl-N-tert-butoxycarbonylamino)tetrafluorophenyl Azide
(30). The carbamate 29 (1.0 mmol) was dissolved in 10 mL of
anhydrous THF and cooled to -78 °C. Next 1.2 mL (1.2 mmol) of
lithium bis(trimethylsilyl)amide (1 M solution in THF) was added. The
solution was stirred for 30 min, and methyl iodide (0.1 mL, 1.55 mmol)
was added. The reaction was warmed to 5 °C and kept at this
temperature overnight. Water was then added, and the reaction was
extracted with Et2O. The combined organic extracts were washed (1 N
HCl, then water), dried (MgSO4), filtered, and evaporated. The residue
was purified by chromatography on silica gel (hexane/diethyl ether 8:1)
1-(tert-Butyldimethylsilyloxy)-6-(p-methoxybenzyloxy)-3-methyl-
hex-2-ene (26). To a solution of allylic alcohol 25 (7.4 g, 29.6 mmol)
in dry DMF was added imidazole (2.4 g, 35.5 mmol) followed by tert-
butyldimethylsilyl chloride (5.4 g, 35.5 mmol). The reaction was stirred
9
8216 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
1
to give 297 mg (93%) of 30. H NMR (400 MHz): δ 3.17 (s, 3H),
colorless oil which was used in the next reaction without further
purification. An analytical sample was prepared by flash chromatog-
raphy (2% EtOAc in hexane). 1H NMR (400 MHz): δ 5.37-5.33 (m,
1H), 4.36 (t, 2H, J ) 6.4 Hz), 4.20 (dd, 2H, J ) 6.4 Hz, J ) 0.8 Hz),
2.16-2.12 (m, 2H), 1.92-1.86 (m, 2H), 1.65 (s, 3H), 0.90 (s, 9H),
0.07 (s, 6H). 13C NMR (100 MHz): δ 159.71, 146.92 (m), 144.35 (m),
142.06 (m), 139.56 (m), 135.63, 125.86, 123.56 (m), 108.35 (m), 66.55,
60.55, 35.83, 26.78, 26.35, 18.78, 16.58, -4.74. 19F NMR (376
MHz): δ -139.94 (m), -152.11 (m). IR (KBr): 2120, 1735, 1483.
LRMS (EI) m/z (relative intensity): 461 (12%, M+), 460 (32%, M -
H+), 433 (100%, M+ - N2). HRMS calcd for C16H18F4N3Si O3 (M -
C4H9+), 292.1645; found, 292.1645.
General Halogenation Conditions for Compounds 33 and 35. To
a solution of Ph3PCl2 (78 mg, 0.23 mmol) in CH2Cl2 (0.5 mL) was
added the silyl ether (0.10 mmol) dissolved in CH2Cl2 (0.75 mL). After
being stirred at room temperature for 1.5 h, the reaction mixture was
loaded directly onto a silica gel column and purified by flash
chromatography (2% EtOAc in hexane) to afford the allylic chloride
which was used immediately in the pyrophosphorylation step.
6-Chloro-4-methyl-4-hexene-(4-azidotetrafluorobenzoate) (33).
Yield (87%). TLC: Rf 0.54.
1.53 and 1.39 (two singlets from minor and major rotamers, respec-
tively, 9H). 13C NMR (100 MHz): δ 154 and 153.8 (minor and major
rotamers, respectively), 144 (m) and 149.82 (m) (minor and major
rotamers, respectively), 143 (m) and 142.83 (m) (minor and major
rotamers, respectively), 142.3 (m), 139.8 (m), 125.91, 119 (m), 82.5
and 82.1 (minor and major, respectively), 37.5 and 36.5 (minor and
major rotamers, respectively), 28.53 and 28.33 (minor and major
rotamers, respectively). 19F NMR (376 MHz): -146.93 (m) and
-147.57 (m) (minor and major rotamers, respectively), -153.69 and
-153.95 (m) (minor and major rotamers, respectively). IR (KBr): 2120,
1716, 1506. LRMS (EI) m/z (relative intensity): 320 (8%, M+), 247
(30%, M - BuO), 57 (100%). HRMS calcd for C12H12F4N4O2 (M+),
t
320.0891; found, 320.0813. C8H3F4N4O (M - tBuO), 247.0238; found,
247.0209.
N-Methyl-4-azidotetrafluoroaniline (31). Carbamate 30 (108 mg,
0.34 mmol) was dissolved in anhydrous CH2Cl2 (5 mL) at 0 °C, and
1.2 mL of 1 N HCl in AcOH (Aldrich) was added. The reaction was
slowly (ca 2 h) allowed to warm to room temperature and stirred
overnight. Next 4 mL of water was added, and the reaction was
extracted with hexane:diethyl ether (1:1 v/v). The organic layers were
washed (5% NaHCO3 then water), dried (MgSO4), filtered through a
pad of silica gel, and concentrated to afford 70 mg (94%) of 31 which
N-(6-Chloro-4-methyl-4-hexenyl)-N-methyl-4-azidotetrafluoro-
aniline (35). Yield (87%). TLC: Rf 0.63.
1
was used in the next reaction without further purification. H NMR
General Pyrophosphorylation Conditions for Compounds 6 and
7. To a solution of the allylic chlorides 33 or 35 (20 mg, 0.06 mmol)
in dry acetonitrile (3 mL) was added tris(tetrabutylammonium)hydrogen
pyrophosphate (140 mg, 0.16 mmol). The reaction was allowed to stir
at room temperature for 4 h. Solvent was removed, and the residue
was dissolved in 3 mL of ion-exchange buffer. The resulting milky
(500 MHz): δ 4.78 (bs, 1H), 3.08 (m, 3H). 13C NMR (100 MHz): δ
142.8 (m), 140.4 (m), 139.1 (m), 136.7 (m), 126.2 (m), 107.7 (m),
33.34 (t). 19F NMR (470 MHz): δ -154.49 (m), -160.85 (m). IR
(KBr): 2116, 1509. LRMS (EI) m/z (relative intensity): 219 (75%, M
- H+), 192 (100%, M - N2).
+
white solution was loaded onto a cation-exchange resin (NH4 form),
N-(6-(tert-Butyldimethylsilyloxy)-4-methyl-4-hexenyl)-N-methyl-
4-azidotetrafluoroaniline (34). A solution of pyridine (0.1 mL) in CH2-
Cl2 (5 mL) was cooled to -78 °C, and 0.3 mL (0.3 mmol) of TiCl4 (1
M solution in toluene) was slowly added. The mixture was then stirred
for 10 min followed by the dropwise addition of amine 31 (66 mg,
0.30 mmol) in dry CH2Cl2 (2 mL). The reaction mixture was stirred
for 5 min at -78 °C, and then a solution of aldehyde 28 (72 mg, 0.30
mmol) in 1 mL of dry CH2Cl2 was added in one portion. The reaction
mixture was allowed to warm to room temperature and was stirred
overnight. After 0.38 mL (6.6 mmol) of glacial acetic acid was added,
the mixture was stirred for 10 min, and NaBH(OAc)3 (263 mg, 1.2
mmol) was added in one portion. The reaction was stirred at room
temperature for 20 h, diluted with 5 mL of hexane, and poured into a
separatory funnel containing 10 mL of hexane and 10 mL of a saturated
solution of NH4F. The aqueous layer was extracted with hexane, and
the combined organic layers were washed (5% NaHCO3, then water),
dried (MgSO4), filtered, and evaporated. The residue was purified by
flash chromatography (4% Et2O in hexane) affording 80 mg (60%) of
eluted with ion-exchange buffer, and the eluant was lyophilized. A
portion of the crude pyrophosphate (ca. 25 mg) was dissolved in 1 mL
of 25 mM NH4CO3. The solution of compound 6 was loaded (in 100
λ portions) onto a Vydac 218TP1010 C18 column (250 mm × 10 mm)
monitoring at 254 nm and eluted under the following gradient at a
flow rate of 4 mL/min: 0 f 6 min 100% A, 6 f 18 min 5% A, 18 f
19 min 5% A, 19 f 24 min 100% A, 24 f 35 min 100% A (A ) 25
mM NH4HCO3, B ) CH3CN). The solution of compound 7 was loaded
onto a Western Analytical 250 × 10 mm, 5 µm BioBasic8 C8 monitoring
at 254 nm and eluted under the following gradient at a flow rate of 4
mL/min: 0 f 6 min 95% A, 6 f 31 min 5% A, 31 f 36 min 5% A,
36 f 43 min 95% A, 43 f 50 min 95% A (A ) 25 mM NH4HCO3
in H2O, B ) CH3CN).
6-(N-Methyl-4-azidotetrafluoroaniline)-3-methyl-2-hexene Pyro-
phosphate (6). Fluffy white solid (40%). tR: 13.8 min. 1H NMR (D2O,
400 MHz): δ 5.18 (m, 1H), 4.23 (t, J ) 6.8 Hz, 2H), 2.81 (t, J ) 7.8
Hz, 2H), 2.61 (s, 3H), 1.82 (t, J ) 7.4 Hz), 1.44 (s, 3H), 1.40-1.30
(m, 2H). 13C NMR (D2O, 100 MHz): δ 146.05 (m), 143.54 (m), 142.38,
142.2 (m), 139.8 (m), 125.04 (m), 119.9, 116.09 (m), 62.91 (d, J )
5.3 Hz), 55.61, 41.64, 36.27, 24.77, 15.39. 19F NMR (D2O, 376
MHz): δ -149.54 (m), -154.47 (m). 31P NMR (D2O, 162 MHz): δ
21.19 (d, 1P, J ) 20.8 Hz), 20.84 (d, 1P, J ) 20.8 Hz). IR (KBr):
2120. LRMS(FAB-, glycerol matrix): (M - H+) 491, (M - N2) 464.
6-(4-Azidotetrafluorobenzylester)-3-methyl-2-hexene Pyrophos-
1
compound 34. H NMR (400 MHz): δ 5.29-5.25 (m, 1H), 4.17 (dd,
2H, J ) 6.4 Hz, J ) 0.8 Hz), 3.08 (t, 2H, J ) 7.6 Hz), 2.89 (t, J ) 2
Hz), 1.99 (t, J ) 7.6 Hz), 1.65-1.59 (m, 5H), 0.89 (s, 9H), 0.06 (s,
6H). 13C NMR (100 MHz): δ 145.01 (m), 142.6 (m), 140.18 (m),
136.27, 127.5 (m), 125.1, 113 (m), 60.40, 55.28 (t, J ) 3 Hz), 41.08
(t, J ) 3.1 Hz), 36.63, 26.18, 25.89, 18.62, 16.42, -4.91. 19F NMR
(376 MHz): δ -150.31 (m), -154.11 (m). IR (KBr): 2120, 1491.
LRMS (EI) m/z (relative intensity): 446 (50%, M+), 418 (44%, M+
-
1
phate (7). Fluffy white solid (40%). tR: 17.8 min. H NMR (D2O:
N2), 246 (100%). HRMS calcd for C20H30F4N4OSi (M+), 446.2125;
found, 446.2164.
CD3CN 2:1, 400 MHz): 5.46 (m, 1H), 4.50-4.30 (m, 4H), 2.19 (m,
2H), 2.01 (m, 2H), 1.72 (s, 3H). 13C NMR (D2O:CD3CN 2:1, 125
MHz): 160.92, 146.72 (m), 144.76 (m), 141.97 (m), 140.24 (m), 140.97,
6-(4-Azido-2,3,5,6-tetrafluorobenzyloxy)-1-tert-butyldimethyl-
silyloxy-3-methyl-2-hexene (32). 4-Azido-2,3,5,6-tetrafluorobenzoyl
chloride63 (119 mg, 0.47 mmol) was dissolved in CH2Cl2 (1 mL), and
pyridine (38 µL, 0.47 mmol) was added. Alcohol 27 (104 mg, 0.42
mmol), dissolved in 4 mL of CH2Cl2, was then slowly added. After
the reaction was stirred at room temperature overnight, water was added,
and the reaction was extracted with hexane. The combined organic
layers were washed (1 N HCl, then 5% NaHCO3, then water), dried
(MgSO4), filtered, and evaporated to give 180 mg (93%) of 32 as a
124.43 (m), 121.75, 107.34 (m), 67.23, 63.05, 35.61, 26.69, 16.19. 19
F
NMR (D2O:CD3CN 2:1, 376 MHz): -140.76 (m), -152.34 (m). 31P
NMR (D2O:CD3CN 2:1, 202.37 MHz): 10.8 (d, 1P, J ) 18.5 Hz),
10.1 (d, 1P, J ) 18.5 Hz). IR (KBr): 2128, 1720. LRMS (FAB-,
glycerol matrix): (M - H+) 506, (M - N2) 479.
General Biological Materials and Methods. [1-3H]-Farnesyl py-
rophosphate ([3H]FPP, 15 Ci/mmol) was obtained from American
Radiochemical Co. Recombinant Ras-CVLS was produced in bacteria
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8217

A R T I C L E S
Chehade et al.
as described previously.81 All other solvents and chemicals were reagent
grade and purchased from standard commercial sources.
N-Dansyl-G-(8-anilinogeranyl-1-S-cysteine)-VLS (44). HRMS
(ESI): calculated for C47H68N7O9S2, 938.4520; measured, 938.4568.
N-Dansyl-G-(8-pentafluoroanilinogeranyl-1-S-cysteine)-VLS (45).
HRMS (ESI): calculated for C47H63F5N7O9S2, 1028.4049; measured,
1028.4029.
N-Dansyl-G-(8-(p-nitroanilino)geranyl-1-S-cysteine)-VLS (46).
HRMS (ESI): calculated for C47H67N8O11S2, 983.4371; measured,
983.4373.
N-Dansyl-G-(8-(N-ethyl-p-azidotetrafluoroanilino)geranyl-1-S-
cysteine)-VLS (47). LRMS (ESI): calculated for C49H68F4N10O9S2,
1080.4993; measured, 1079.62. HRMS (-N2 + H ) 1053.5010,
measured, 1053.5016) (-N2 + 2H ) 1054.5088), measured, 1054.5050
(-N2 + 2H).
N-Dansyl-G-(6-(N-methyl-p-azidotetrafluoroanlino)-3-methyl-2-
hexenyl-1-S-cysteine)-VLS (48). HRMS (ESI): calculated for C45H62-
F4N10O9S2, 1026.4079 (-N2 + H ) 999.4095, measured 999.4081)
(-N2 + 2H ) 1000.4174), measured, 1000.4118 (-N2 + 2H).
N-Biotin-G-(8-(N-ethyl-p-azidotetrafluoroanilino)geranyl-1-S-
cysteine)-VLS (49). HRMS (ESI): calculated for C47H69F4N11O9S2,
1071.4657; measured, 1071.5.
Photoinactivation of FTase. Farnesyltransferase activity was as-
sayed by measuring the amount of [3H]farnesyl transferred from [3H]-
FPP to recombinant Ras, as described previously.83 Unless otherwise
stated, each reaction mixture contained the following components in a
final volume of 25 µL: 50 mM Tris (pH 7.4), 20 mM KCl, 10-20 ng
of recombinant FTase, 3 mM MgCl2, 50 µM ZnCl2, and 0.6 µM ETAZ-
AGPP. The solution was placed underneath a 254 nm hand-held light
source (UVP Model UVG-11) and photolyzed for 30 min at 0 °C. After
photolysis, a solution of 1 mM dithiothreitol, 5 µM Ras, 0.2% octyl
â-glucopyranoside, and 0.6 µM [3H]FPP (33 000 dpm/pmol; American
Radiochemical Co.) was added, and the reaction was incubated for 30
min at 37 °C. The amount of [3H]farnesyl transferred was measured
by ethanol-HCl precipitation and filtration on glass fiber filters with
modification as previously described.81 A blank value was determined
in parallel incubation mixtures containing no enzyme. This blank value
was subtracted from each reaction before calculating the picomoles of
[3H]prenyl transferred.
Biochemical Evaluation: Continuous Fluorescence Assay. The
kinetic constants Km, Vmax, and kcat for FPP and FPP analogues 2-4
were determined using the continuous spectrofluorimetric assay as
previously described.37 The peptide N-dansyl-GCVLS was synthesized
and HPLC-purified by the MSAF facility at the University of Kentucky.
Rat protein farnesyltransferase containing a 6X His-tag with a thrombin
cleavage site on the N-terminus of the â subunit (N-terminus )
MGSSHHHHHHSSGLVPRGSH) was recombinantly expressed in
BL21(DE3) E. coli cells82 and purified using a Ni(II)-charged IMAC
(immobilized metal affinity chromatography, POROS) column with an
imidazole gradient. The purified protein was >90% pure, as judged
by SDS-PAGE analysis. Fluorescence was detected using a time-based
scan at 30 °C for a period of 60 s.
The kinetic constants Km, Vmax, and kcat for FPP and the analogues
2-4 were determined using the continuous spectrofluorimetric assay
originally developed by Pompliano et al.74 and modified by Poulter
and co-workers75 with some additional modification. Utilizing N-dansyl-
GCVLS as the peptide substrate, we measured the linear portion of
the increase in fluorescence versus time with a Hitatchi F2000
spectrofluorimeter (excitation wavelength, 350 nm; emission wave-
length, 505 nm). The assay components {209.6 µL of assay buffer (50
mM Tris-HCl, pH 7.5, 5 mM DTT, 5 mM MgCl2, 10 µM ZnCl2), 40
µL of detergent solution (0.4% n-dodecyl-â-D-maltoside in H2O), 0.4
µL of N-dansyl-GCVLS solution (1 mM in 20 mM Tris-HCl, pH 7.5,
10 mM EDTA), 100 µL of FPP or FPP analogue (2-20 mM stock
solutions in 12 mM NH4HCO3:MeOH (7:3 v/v); final concentration
0.03-3 µM)} were assembled in a 1.5 mL Eppendorf tube in the order
indicated above and incubated at 30 °C for 5 min. A 50 µL FTase
solution (0.1 mg/mL BSA with 0.27 µM FTase (for FPP only) or 1.2
µM FTase in assay buffer) was loaded into a 1.0 mL quartz cuvette
and incubated at 30 °C for 5 min. The reaction was initiated by the
addition of the 350 µL assay buffer/peptide/prenyl pyrophosphate
solution to the 50 µL FTase/BSA solution. Fluorescence was detected
using a time-based scan at 30 °C for a period of 60 s. The velocity
was determined by converting the rate of increase in fluorescence
intensity units (FLU/s) to pmol/s by fitting the data to eq 1.
pKa Prediction. We utilized the Hammett eq 2 to describe the pKa
value of an unsubstituted base84
R × P
V )
(1)
pKa ) pKa0 - F( σ)
(2)
FMAX
∑
V is the velocity of the reaction in pmol/s. R is the rate of the reaction
in FLU/s. P is equal to the number of picomoles of FPP or analogues
2-4 used in the reaction mixture when the number of N-dansyl-GCVLS
picomoles is 2-fold or greater to the number of prenyl pyrophosphate
picomoles. FMAX is the fluorescence intensity of the reaction mixture
after incubation for 60 min at 30 °C when an amount of P was used
for FPP or analogues 2-4. On the basis of HPLC analysis of large-
scale reactions under similar conditions, we assume that each reaction
has gone to completion.
where pKa0 is the pKa of the unsubstituted base, σ is a constant assigned
to a particular substituent, and F is a constant for a particular
equilibrium. Equation 3 describes the expression for the ionization of
substituted anilines at 25 °C84 (Table A.3):
pKa ) 4.58-2.88
σ
(3)
∑
A pKa for pentafluoroaniline was estimated with eq 2 using literature
values of σmeta (0.34),85 para (0.06),85 and apparent σortho (0.47)84 (Table
σ
Product Studies. Large-scale reactions contained 50 mM Tris‚HCl,
pH 7.5, 5 mM MgCl2, 10 µM ZnCl2, 5 mM DTT, 0.04% n-dodecyl-
â-D-maltoside, 2.5 µM N-dansyl-GCVLS, 0.25 µM FTase, and 25 µM
FPP or 2-6 in a final reaction volume of 400 µL. After the samples
were allowed to react for 1 h at 37 °C, each reaction mixture was loaded
onto an analytical Vydac C8 (208TP54) column and eluted with a linear
gradient of 0-40 min 10% B to 100% B at a flow rate of 1 mL/min.
Solvents: A ) 0.01% (v/v) TFA in water; B ) 0.01% (v/v) TFA in
CH3CN. The fraction corresponding to coincident absorbance and
fluorescence peaks was collected and analyzed by mass spectroscopy.
N-Dansyl-G-(farnesyl-1-S-cysteine)-VLS (43). HRMS (ESI): cal-
culated for C46H71N6O9S2, 915.4724; measured, 915.4726.
A.5) for fluorine. The pKa for 4-azidotetrafluoroaniline was then
estimated with eq 2 when the literature values for σmeta (0.34) and
apparent σortho (0.47) for fluorine were used, along with a σpara (0.15)85
value for azide.
pKa ) 4.85 - F
σ
(4)
∑
Equation 4 describes the expression for the ionization of substituted
N-methyl anilines. To determine the pKa of N-methyl-4-azidotetrafluoro-
aniline, we first fit the literature pKa value for N-methyl-pentafluoro-
(83) Andres, D. A.; Goldstein, J. L.; Ho, Y. K.; Brown, M. S. J. Biol. Chem.
1993, 268, 1383-1390.
(84) Perrin, D. D.; Dempsey, B.; Serjeant, E. P. pKa Prediction for Organic
Acids and Bases; Chapman and Hall: London, 1981.
(85) Hansch, C.; Leo, A.; Unger, S. H.; Kim, K. H.; Nikaitani, D.; Lien, E. J.
J. Med. Chem. 1973, 16, 1207-1216.
(81) Crick, D. C.; Suders, J.; Kluthe, C. M.; Andres, D. A.; Waechter, C. J. J.
Neurochem. 1995, 65, 1365-1373.
(82) Zimmerman, K. K.; Scholten, J. D.; Huang, C.-C.; Fierke, C. A.; Hupe, D.
J. Protein Expression Purif. 1998, 14, 395-402.
9
8218 J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002

Photoaffinity Analogues of Farnesyl Pyrophosphate
A R T I C L E S
aniline (0.676 ( 0.012)72 to eq 3 along with the σ values used above
for pentafluoraniline to determine F. The pKa value for N-methyl-4-
azidotetrafluoroaniline was then estimated with eq 3 when the σ values
used above for 4-azidotetrafluoroaniline were employed.
Tyr 251â, Phe 253â, Cys 254â, Trp 303â, Tyr 361â, Tyr 365â. A
two-stage refinement was used to model the binary complexes. The
first round consisted of a Monte Carlo conformational search and
minimization, from which 50 structures were generated. This was done
with the electrostatic component of the force field turned off and the
van der Waals component scaled to 10%. The second round consisted
of a Monte Carlo conformational search and minimization of the 50
generated structures with all components of the force field at 100%
and inclusion of a distance-dependent dielectric component (dielectric
constant ) 1) followed by simulated annealing consisting of 50
increments of 100 fs each to cool the system from 500 to 300 K. The
overall and per residue binding energy of each of the final 50 structures
were then measured via a CDiscover routine. For comparison to these
generated structures, the modified 1FT2 binary cocrystal structure was
also minimized as in the second round of refinement (without simulated
annealing), and the FTase-FPP binding energy was calculated.
Molecular Modeling. The FTase starting structure was created by
adding hydrogen atoms to the binary FTase-FPP crystal structure
structure (PDB code 1FT2). Atom potential types and partial charges
were assigned on the basis of the cff91 force field defaults. The cysteine
and histidine residues coordinating the zinc atom were considered to
be deprotonated, and their charges were modified accordingly. Because
charges for the deprotonated residues are not provided in the cff91
force field library, charges for deprotonated cysteine were taken from
the AMBER 6.0 library, charges for deprotonated histidine (histidinate)
were taken from Pang et al.86 FPP, and analogue starting structures
were constructed in the “Builder” module of Insight II (v2000, MSI,
Inc.). The optimal geometery and atomic charges of the termini of the
analogues were calculated using density functional methods (SVWN)
as implemented in MacSpartan Pro. The DN* basis set and SCF
damping (0.5) were used. The atomic charges of the remainder of the
analogues were assigned on the basis of RESP fitting of FPP. Atom
potential types were assigned automatically on the basis of hybridization
and connectivities.
All ligand starting structures were in the fully extended conformation.
Analogues 5 and 6 were modeled in both cis and trans conformations
with respect to the relative orientation of the azido group and main
chain about the aniline C-N bond. Minimization and molecular
dynamics were set up using the “Docking” module and carried out
with the “CDiscover” module of Insight II (v2000, MSI, Inc.) under
the cff91 force field. Optimal geometry in the analogue termini was
enforced by angle and torsional restraints. The coordinates of all protein
residues were fixed, except for the following active site residues, which
were allowed to move: Tyr 200R, His 201R, Trp 102â, Ala 151â, Tyr
154â, Met 193â, Asp 200â, Arg 202â, Tyr 205â, Cys 206â, Gly 250â,
Acknowledgment. We thank Mark Meier for assistance with
MacSpartan, Robert Dickson and Robert Lester for assistance
with the HPLC fluorescence analysis, and Louis Hersh for his
advice with the enzyme kinetics. This work was supported in
part by the American Heart Association grant (to D.A.A.), the
National Science Foundation grant MCB-9808633 (to H.P.S.),
the Kentucky Lung Cancer Research Program (to H.P.S. and
D.A.A.), Kentucky-NSF EPSCoR Program (EPS-9452895) (to
H.P.S.), and the National Institutes of Health (GM40602 to
C.A.F.).
Supporting Information Available: 1H, 13C, 31P, 19F NMR,
IR, and MS spectra as appropriate are available for compounds
reported (PDF). This material is available free of charge via
the Internet at http://pubs.acs.org.
(86) Pang, Y.-P.; Xu, K.; El Yazal, J.; Prendergast, F. G. Protein Sci. 2000, 9,
1857-1865.
JA0124717
9
J. AM. CHEM. SOC. VOL. 124, NO. 28, 2002 8219