Synthesis of farnesyl diphosphate analogues containing ether-linked photoactive benzophenones and their application in studies of protein prenyltransferases

Article abstract of DOI:10.1021/jo991130x

Protein prenylation is a posttranslational lipid modification in which C₁₅ and C₂₀ isoprenoid units are linked to specific protein-derived cysteine residues through a thioether linkage. This process is catalyzed by a class of enzymes called prenyltransferases that are being intensively studied due to the finding that Ras protein is farnesylated coupled with the observation that mutant forms of Ras are implicated in a variety of human cancers. Inhibition of this posttranslational modification may serve as a possible cancer chemotherapy. Here, the syntheses of two new farnesyl diphosphate (FPP) analogues containing photoactive benzophenone groups are described. Each of these compounds was prepared in six steps from dimethylallyl alcohol. Substrate studies, inhibition kinetics, photoinactivation studies, and photolabeling experiments are also included; these experiments were performed with a number of protein prenyltransferases from different sources. A X-ray crystal structure of one of these analogues bound to rat farnesyltransferase illustrates that they are good substrate mimics. Of particular importance, these new analogues can be enzymatically incorporated into Ras-based peptide substrates allowing the preparation of molecules with photoactive isoprenoids that may serve as valuable probes for the study of prenylation function. Photoaffinity labeling of human protein geranylgeranyltransferase with ³²P-labeled forms of these analogues suggests that the C-10 locus of bound geranylgeranyl diphosphate (GGPP) is in close proximity to residues from the β-subunit of this enzyme. These results clearly demonstrate the utility of these compounds as photoaffinity labeling analogues for the study of a variety of protein prenyltransferases and other enzymes that employ FPP or GGPP as their substrates.

Full text of DOI:10.1021/jo991130x

J . Org. Chem. 2001, 66, 3253-3264
3253
Syn th esis of F a r n esyl Dip h osp h a te An a logu es Con ta in in g
Eth er -Lin k ed P h otoa ctive Ben zop h en on es a n d Th eir Ap p lica tion
in Stu d ies of P r otein P r en yltr a n sfer a ses
Tammy C. Turek, Igor Gaon, and Mark D. Distefano*
Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455
Corey L. Strickland
Department of Structural Chemistry, Schering-Plough Research Institute, Kenilworth, New J ersey 07033
distefan@chemsun.chem.umn.edu
Received J uly 15, 1999 (Revised Manuscript Received December 21, 2000 )
Protein prenylation is a posttranslational lipid modification in which C15 and C20 isoprenoid units
are linked to specific protein-derived cysteine residues through a thioether linkage. This process is
catalyzed by a class of enzymes called prenyltransferases that are being intensively studied due to
the finding that Ras protein is farnesylated coupled with the observation that mutant forms of
Ras are implicated in a variety of human cancers. Inhibition of this posttranslational modification
may serve as a possible cancer chemotherapy. Here, the syntheses of two new farnesyl diphosphate
(FPP) analogues containing photoactive benzophenone groups are described. Each of these
compounds was prepared in six steps from dimethylallyl alcohol. Substrate studies, inhibition
kinetics, photoinactivation studies, and photolabeling experiments are also included; these
experiments were performed with a number of protein prenyltransferases from different sources.
A X-ray crystal structure of one of these analogues bound to rat farnesyltransferase illustrates
that they are good substrate mimics. Of particular importance, these new analogues can be
enzymatically incorporated into Ras-based peptide substrates allowing the preparation of molecules
with photoactive isoprenoids that may serve as valuable probes for the study of prenylation function.
Photoaffinity labeling of human protein geranylgeranyltransferase with 32P-labeled forms of these
analogues suggests that the C-10 locus of bound geranylgeranyl diphosphate (GGPP) is in close
proximity to residues from the â-subunit of this enzyme. These results clearly demonstrate the
utility of these compounds as photoaffinity labeling analogues for the study of a variety of protein
prenyltransferases and other enzymes that employ FPP or GGPP as their substrates.
In tr od u ction
geranylgeranyltransferase type II (PGGTase-II).5,6 Re-
cently, there has been an explosion of interest in studying
protein prenylation since oncogenic Ras is farnesylated
and mutant forms have been detected in 30% of all
human cancers7 including 90% of pancreatic, 50% of
colon, and 30% of lung cancers.8 Since farnesylation is
necessary for membrane association and cellular trans-
formation of Ras oncoproteins, preventing this modifica-
tion may eliminate the activity of oncogenic Ras and
hence may serve as a possible cancer chemotherapy. This
notion has generated an intense research effort to design
inhibitors of PFTase as potential anticancer agents.9-17
However, geranylgeranylation of cellular proteins is five
Protein prenylation is a posttranslational lipid modi-
fication that involves attachment of a farnesyl (C15) group
from farnesyl diphosphate (FPP, 1) or, more commonly,
a geranylgeranyl (C20) group from geranylgeranyl diphos-
phate (GGPP, 2) to a specific cysteine residue of a protein
substrate through a thioether linkage.
(3) Moomaw, J . F.; Casey, P. J . J . Biol. Chem. 1992, 267, 17438-
17443.
(4) Yokoyama, K.; Gelb, M. H. J . Biol. Chem. 1993, 268, 4055-4060.
(5) Seabra, M. C.; Goldstein, J . L.; Sudhof, T. C.; Brown, M. S. J .
Biol. Chem. 1992, 267, 14497-14503.
(6) Armstrong, S. A.; Seabra, M. C.; Sudhof, T. C.; Goldstein, J . L.;
Brown, M. S. J . Biol. Chem. 1993, 268, 12221-12229.
(7) Lerner, E. C.; Qian, Y.; Hamilton, A. D.; Sebti, S. M. J . Biol.
Chem. 1995, 270, 26770-26773.
(8) McNamara, D. J .; Dobrusin, E.; Leonard, D. M.; Shuler, K. R.;
Kaltenbronn, J . S.; Quin, J ., II; Bur, S.; Thomas, C. E.; Doherty, A.
M.; Scholten, J . D.; Zimmerman, K. K.; Gibbs, B. S.; Gowan, R. C.;
Latash, M. P.; Leopold, W. R.; Przybranowski, S. A.; Sebolt-Leopold,
J . S. J . Med. Chem. 1997, 40, 3319-3322.
(9) Buss, J . E.; Marsters, J . C. Chem. Biol. 1995, 2, 787-791.
(10) Gibbs, J . B.; Oliff, A.; Kohl, N. E. Cell 1994, 77, 175-178.
This process is catalyzed by protein prenyl trans-
ferases; at present, several enzymes have been identified
from yeast and mammalian sources including two distinct
protein farnesyl transferases (PFTase),1,2 protein gera-
nylgeranyltransferase type I (PGGTase-I),3,4 and protein
(1) Reiss, Y.; Goldstein, J . L.; Seabra, M. C.; Casey, P. J .; Brown,
M. S. Cell 1990, 62, 81-88.
(2) Vogt, A.; Sun, J .; Qian, Y.; Tan-Chiu, E.; Hamilton, A. D.; Sebti,
S. M. Biochemistry 1995, 34, 12398-12403.
10.1021/jo991130x CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/14/2001

3254 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
to ten times more prevalent than farnesylation in eu-
karyotic cells and tissues;18 therefore, selective inhibition
of PFTase is an important issue for inhibitors to be
clinically useful.
these enzymes suggesting that the peptide binding site
was located at the R/â subunit interface. Pellicena and
colleagues also incorporated a benzophenone group into
their peptide analogue and identified a specific amino
acid on the R subunit of rat PFTase (rPFTase) important
in peptide binding, demonstrating the utility of the
benzophenone moiety at identifying important residues
involved in substrate recognition.25 Benzophenone-
containing analogues have also been prepared and used
to study several enzymes involved in the synthesis of
isoprenoids including farnesyl synthase, hexaprenyl syn-
thase, and undecaprenyl synthase.26 In subsequent work,
photolabeling with one of these compounds was used to
show that component I of heptaprenyl synthase contained
the isoprenoid diphosphate binding site.27 We have
previously reported work employing compounds incor-
porating photoactive benzophenones into FPP and GGPP
analogues to identify the isoprenoid binding site and
shown that the analogues specifically label the â subunit
of yeast PFTase and human PGGTase-I. These results
provide further evidence for this subunit’s involvement
in prenyl group recognition as was found with the DATFP
analogues.28,29 Recently, the X-ray crystal structure of
rPFTase was solved, thus identifying the location of
binding of FPP in the enzyme active site on the â
subunit.30-33 While there is substantial sequence identity
among the mamalian PFTases (93% identity between the
human and rat enzymes), only 37% amino acid identity
between the â subunit of yeast PFTase and mammalian
PFTases and 30% sequence similarity between the â
subunit of PGGTase-I and mammalian PFTases ex-
ists.19,34 Therefore, identifying similarities and differences
between these enzymes in their prenyl substrate binding
sites may facilitate the design and optimization of selec-
tive PFTase inhibitors.
Identifying active site residues of protein prenyltrans-
ferases may be useful in designing more potent inhibitors
that are selective for PFTase and that do not affect other
metabolic pathways. Photoaffinity labeling has provided
important information concerning the active sites of these
enzymes through investigating the substrate binding and
recognition pattern. Incorporating a diazotrifluoropropi-
onate (DATFP) cross-linking group into a FPP analogue,
DATFP-GPP, specifically labeled the â subunit of human
and yeast PFTase.19-21 Preparation of a GGPP analogue,
DATFP-FPP, labeled the â subunit of human and bovine
brain PGGTase-I, implicating the involvement of the â
subunit in prenyl substrate binding.20,22 However, there
are no reports in the literature of specific residues that
have been identified to be important in isoprenoid bind-
ing using these DATFP photoactive analogues. This may
be due to the difficulty in handling the diazoester moiety
as well as the stability of the photolabel. Diazoester
isoprenoid analogues are allylic esters with multiple
pathways for degradation. Benzophenone analogues offer
an attractive alternative to diazo- and azide-containing
analogues due to their increased chemical stability,
activation at higher wavelengths which are less damag-
ing to the protein, ease of manipulation in ambient light,
and preferential reactions with unreactive C-H bonds
even in the presence of solvent and water molecules.23
Ying and co-workers studied the PFTase protein sub-
strate binding site by incorporating benzophenone groups
into several peptide analogues.24 When one benzophenone
group was incorporated, they labeled the â subunit of
human, rat, and bovine PFTase; with a peptide substrate
analogue appended with two benzophenone moieties,
they observed labeling of both the R and â subunit of
In this paper, the syntheses of two new FPP analogues,
3a and 3b, are described along with substrate studies,
inhibition kinetics, photoinactivation studies, and pho-
tolabeling experiments.
These experiments were performed on a number of
protein prenyltransferases from different sources and the
results from these findings are discussed. Crystallization
of one of these analogues with rat farnesyltransferase
illustrates that they are good substrate mimics. Of
particular importance, these analogues can be enzymati-
cally incorporated into Ras-based peptide substrates,
allowing us to prepare molecules with photoactive iso-
prenoids that may serve as valuable probes for the study
(11) Gibbs, J . B.; Pompliano, D. L.; Mosser, S. D.; Rands, E.;
Lingham, R. B.; Singh, S. B.; Scolnick, E. M.; Kohl, N. E.; Oliff, A. J .
Biol. Chem. 1993, 268, 7617-7620.
(12) Manne, V.; Ricca, C. S.; Brown, J . G.; Tuomari, A. V.; Yan, N.;
Patel, D.; Schmidt, R.; Lynch, M. J .; Ciosek, C. P.; Carboni, J . M.;
Robinson, S.; Gordon, E. M.; Barbacid, M.; Seizinger, B. R.; Biller, S.
A. Drug Dev. Res. 1995, 34, 121-137.
(13) Aoyama, T.; Satoh, T.; Yonemoto, M.; Shibata, J .; Nonoshita,
K.; Arai, S.; Kawakami, K.; Iwasawa, Y.; Sano, H.; Tanaka, K.;
Monden, Y.; Kodera, T.; Arakawa, H.; Suzuki-Takahashi, I.; Kamei,
T.; Tomimoto, K. J . Med. Chem. 1998, 41, 143-147.
(14) Garcia, A. M.; Rowell, C.; Ackermann, K.; Kowalczyk, J . J .;
Lewis, M. D. J . Biol. Chem. 1993, 268, 18415-18418.
(15) Lerner, E. C.; Qian, Y.; Blaskovich, M. A.; Fossum, R. D.; Vogt,
A.; Sun, J .; Cox, A. D.; Der, C. J .; Hamilton, A. D. J . Biol. Chem. 1995,
270, 26802-26806.
(16) Breslin, M. J .; deSolms, S. J .; Giuliani, E. A.; Stokker, G. E.;
Graham, S. L.; Pompliano, D. L.; Mosser, S. D.; Hamilton, K. A.;
Hutchinson, J . H. Bioorg. Med. Chem. Lett. 1998, 8, 3311-3316.
(17) Manne, V.; Yan, N.; Carboni, J . M.; Tuomari, A. V.; Ricca, C.
S.; Brown, J . G.; Andahazy, M. L.; Schmidt, R. J .; Patel, D.; Zahler,
R.; Weinmann, R.; Der, C. J .; Cox, A. D.; Hunt, J . T.; Barbacid, M.;
Seizinger, B. R. Oncogene 1995, 10, 1763-1779.
(25) Pellicena, P.; Scholten, J . D.; Zimmerman, K.; Creswell, M.;
Huang, C. C.; Miller, W. T. Biochemistry 1996, 35, 13494-13500.
(26) Marecak, D. M.; Horiuchi, Y.; Arai, H.; Shimonaga, M.; Maki,
Y.; Koyama, T.; Ogura, K.; Prestwich, G. D. Bioorg. Med. Chem. Lett.
1997, 7, 1973-1978.
(27) Zhang, Y.-W.; Koyama, T.; Marecak, D. M.; Prestwich, G. D.;
Maki, Y.; Ogura, K. Biochemistry 1998, 37, 13411-13420.
(28) Gaon, I.; Turek, T. C.; Weller, V. A.; Edelstein, R. L.; Singh, S.
K.; Distefano, M. D. J . Org. Chem. 1996, 61, 7738-7745.
(29) Turek, T. C.; Gaon, I.; Gamache, D.; Distefano, M. D. Bioorg.
Med. Chem. Lett. 1997, 7, 2125-2130.
(30) Park, H.-W.; Boduluri, S. R.; Moomaw, J . F.; Casey, P. J .; Beese,
L. S. Science 1997, 275, 1800-1804.
(31) Dunten, P.; Kammlott, U.; Crowther, R.; Weber, D.; Palermo,
R.; Birktoft, J . Biochemistry 1998, 37, 7907-7912.
(32) Long, S. B.; Casey, P. J .; Beese, L. S. Biochemistry 1998, 37,
9612-9618.
(33) 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, 16601-16611.
(34) Mayer, M. L.; Caplin, B. E.; Marshall, M. S. J . Biol. Chem. 1992,
267, 20589-20593.
(18) Maltese, W. A. FASEB J . 1990, 4, 3319-3328.
(19) Omer, C. A.; Kral, A. M.; Diehl, R. E.; Prendergast, G. C.;
Powers, S.; Allen, C. M.; Gibbs, J . B.; Kohl, N. E. Biochemistry 1993,
32, 5167-5176.
(20) Bukhtiyarov, Y. E.; Omer, C. A.; Allen, C. M. J . Biol. Chem.
1995, 270, 19035-19040.
(21) Edelstein, R. L.; Distefano, M. D. Biochem. Biophys. Res.
Commun. 1997, 235, 377-382.
(22) Yokoyama, K.; McGeady, P.; Gelb, M. H. Biochemistry 1995,
34, 1344-1354.
(23) Dorman, G.; Prestwich, G. D. Biochemistry 1994, 33, 5661-
5673.
(24) Ying, W.; Sepp-Lorenzino, L.; Cai, K.; Aloise, P.; Coleman, P.
S. J . Biol. Chem. 1994, 269, 470-477.

Synthesis of Farnesyl Diphosphate Analogues
J . Org. Chem., Vol. 66, No. 10, 2001 3255
F igu r e 1. Superposition of 3b with FPP (1) and GGPP (2).
The structures of 1 and 2 are indicated in bold while the
structure of 3b is given as dashed bonds.
methyl benzophenones were prepared by radical bromi-
nation with NBS using AIBN as the radical initiator
following the procedure developed by Rajagopalan et al.35
Deprotection of 6a and 6b was accomplished with PPTs
in ethanol to give 7a and 7b in quantitative yield. In
previous work with benzoylbenzoate esters, allylic alco-
hols similar to 7a and 7b were converted to the chlorides
by treatment with N-chlorosuccinimide and dimethyl
sulfide followed by pyrophosphorylation.28,36 Subsequent
work with benzoylbenzyl ethers used bromination instead
of chlorination to obtain the activated species employed
in the subsequent pyrophosphorylation.37 In these experi-
ments the brominated products were produced in higher
yield than the earlier chlorides. Thus, bromination of 7a
and 7b with PPh3 and CBr4 followed by displacement of
the allylic bromides 8a and 8b with tris(tetra-n-butyl-
ammonium) hydrogen pyrophosphate generated 3a and
3b in 32% and 30% yields, respectively. Diphosphates 3a
and 3b were purified by reversed-phase chromatography
of prenylation function. These results clearly demonstrate
these compounds’ utility as photoaffinity labeling ana-
logues for the study of a variety of protein prenyltrans-
ferases and other enzymes that employ FPP or GGPP as
their substrates.
Resu lts a n d Discu ssion
1
and characterized by H NMR, 31P NMR, UV, and HR-
Design a n d Syn th esis. Photoreactive analogues of
FPP (1) and GGPP (2) are shown above. These analogues
incorporate a meta- or para-substituted benzoylbenzyl
photophore [4-BB-DMAPP (3a ) and 3-BB-DMAPP (3b)]
linked to a dimethylallyl unit via an acid stable ether
linkage. They were designed based on the observation
that these analogues had significant overlap with the
structures of 1 and 2 as illustrated in Figure 1 where 1
and 2 are superimposed on 3b. The compounds, 3a and
3b, were synthesized in six steps starting from com-
mercially available dimethylallyl alcohol as shown in
Scheme 1.
The hydroxyl group of dimethylallyl alcohol was pro-
tected by reaction with 3,4-dihydropyran and pyridinium
p-toluenesulfonate (PPTs) to generate a tetrahydropyra-
nyl ether in high yield. The resulting ether, 4, was
oxidized with tert-butyl hydroperoxide and catalytic H2-
SeO3 to yield 5 with E-stereoselectivity as evidenced by
the disappearance of the C-4 methyl group in the 13C
NMR spectrum of 5. Further support for E-stereoselec-
tivity was achieved when a portion of 5 was reacted under
Mitsunobu conditions with phthalimide to generate 6c.
As diagrammed in Scheme 1, NOE analysis of this
compound shows a significantly stronger NOE between
the C-2 vinyl proton and the C-4 methylene group (1.7%)
than that observed between the C-2 vinyl proton and the
C-3 methyl group (0.2%) consistent with a trisubstituted
alkene with E-stereochemistry. Compound 5 was then
treated with NaH and 4-(bromomethyl)benzophenone or
3-(bromomethyl)benzophenone to generate 6a and 6b in
78% and 68% yields, respectively. The requisite bromo-
MALDI mass spectrometry. Compounds 3a and 3b are
stable for several years when stored as lyophilized
powders at -80 °C.
Su bstr a te Stu d ies. Initially, a continuous fluores-
cence assay using N-Ds-GCVIA (9) was employed to
determine whether compounds 3a and 3b were sub-
strates for yeast PFTase (yPFTase) and human PFTase
(hPFTase).38,39 When the natural substrate, 1, was as-
sayed, a time-dependent increase in fluorescence (ap-
proximately 10-fold) was observed; however, when 1 (10
µM) was replaced with 10 µM or 20 µM 3a or 3b, only a
slight increase in fluorescence was seen, suggesting that
these compounds were slow alternate substrates. To
further explore the extent to which the analogues were
substrates, the expected products of the enzymatic pre-
nylation of 9 with 3a and 3b were prepared by chemical
synthesis using the method of Xue et al.40 illustrated in
Scheme 2.
The chemically synthesized products 10a and 10b were
purified by reversed-phase HPLC and analyzed by ESI
(35) Rajagopalan, K.; Chavan, A. J .; Haley, B. E.; Watt, D. S. J .
Biol. Chem. 1993, 268, 14230-14238.
(36) Turek, T. C.; Gaon, I.; Distefano, M. D. Tetrahedron Lett. 1996,
37, 4845-4848.
(37) Gaon, I.; Turek, T. C.; Distefano, M. D. Tetrahedron Lett. 1996,
37, 8833-8836.
(38) Pompliano, D. L.; Gomez, R. P.; Anthony, N. J . J . Am. Chem.
Soc. 1992, 114, 7945-7946.
(39) Bond, P. D.; Dolence, J . M.; Poulter, C. D. Methods Enzymol.
1995, 250, 30-43.
(40) Xue, C.-B.; Becker, J . M.; Naider, F. Tetrahedron Lett. 1992,
33, 1435-1438.

3256 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
Sch em e 1
Sch em e 2
were collected, concentrated, and analyzed by ESI MS
and ESI MS/MS. The ESI MS for 10a obtained via
yPFTase (Figure 4A) showed an [M + H]+ peak of 973.4
as well as the an [M + 2H]2+ at 486.8 giving a calculated
[M]+ value of 971.99 ( 0.57. For 10b (Figure 4B), we
observed an [M + H]+ peak at 973.4 and the doubly
charged ion at 487.0 giving a [M]+ value of 972.17 ( 0.31
corresponding to the expected peptide products in Scheme
2.
The ESI MS/MS for the enzymatic reactions (data not
shown) gave similar fragmentation patterns as seen with
the authentic products. For the enzymatic reactions
MS and ESI MS/MS to reveal the fragmentation pattern.
The ESI MS and ESI MS/MS for 10a are displayed in
Figures 2A and 2B.
For 10a , an [M + H]+ peak of 973.2 was observed as
well as the doubly charged ion [M + 2H]2+ at 487.0.
Similar peaks and values were obtained for 10b (data
not shown). Interestingly, the chemically synthesized
products show b-type fragmentation up to the modified
cysteine amino acid at which point the C-S bond of the
â carbon cysteine is cleaved removing the entire analogue
including the sulfur atom. Another important fragmenta-
tion observed is the loss of the dansyl group and cleavage
of the C-S bond formed between the cysteine thiol and
the benzoylbenzyl adduct. With the authentic products
characterized, the products from the large-scale enzy-
matic reactions (10 mL) between 3a or 3b and 9 with
yPFTase or hPFTase were analyzed by HPLC and
compared to those of the chemically synthesized products.
The HPLC chromatograms monitored by UV and fluo-
rescence are shown in Figures 3A and 3B for the reaction
between 3a and 9 with yPFTase.
F igu r e 2. ESI MS and ESI MS/MS of chemically synthesized
product 10a . Panel A: ESI MS of Ds-GC(4-BBDMA)VIA. Panel
B: ESI MS/MS of Ds-GC(4-BBDMA)VIA.
The fluorescent product peaks, with elution times
similar to those of the chemically synthesized products

Synthesis of Farnesyl Diphosphate Analogues
J . Org. Chem., Vol. 66, No. 10, 2001 3257
F igu r e 4. ESI MS of the enzymatic prenylation of yPFTase,
9, and 3a or 3b. Panel A: ESI MS from a 10 mL reaction
mixture containing 9 and 3a with yPFTase and generation of
the enzymatic product Ds-GC(4-BBDMA)VIA. Panel B: ESI
MS from a 10 mL reaction mixture containing 9 and 3b with
yPFTase and generation of the enzymatic product Ds-GC(3-
BBDMA)VIA.
F igu r e 3. HPLC analysis of reactions containing yPFTase,
9, and 3a . Panel A: UV analysis, chromatogram A is from a
10 mL reaction mixture containing 9 and 3a . Chromatogram
B is from a sample of chemically synthesized 10a . The amount
of 10a injected is equal to the amount of product that would
be formed in the enzymatic reaction assuming 100% conver-
sion. Panel B: fluorescence (480-520 nm) detection, chro-
matogram A is from a 10 mL reaction mixture containing 9
and 3a . Chromatogram B is from a sample of chemically
synthesized 10a . The amount of 10a injected is equal to the
amount of product that would be formed in the enzymatic
reaction assuming 100% conversion.
formed were less than 1%, and hence the analogues were
not substrates for yPFTase; therefore, these results
indicate that the shorter chain benzoylbenzyl ether
analogues more closely mimic the natural substrate 1
than the longer C10 ester analogues.28 Based on the above
discussion, it is clear that 3a and 3b are not efficient
alternate substrates. However, using larger amounts of
enzyme and long reaction times it is possible to obtain
efficient incorporation of 3b into a peptide and presum-
ably into a protein. This provides a means of preparing
proteins that contain photoactive isoprenoids which could
be useful in studying the role of protein prenylation.
In h ibition Kin etics. Since formation of products 10a
and 10b from the enzymatic reaction of 3a and 3b with
9 by yPFTase and hPFTase were undetected using low
concentrations PFTase (5.1 nM) in the fluorescence assay,
their rates of reaction are negligible compared to that of
FPP under typical assay conditions; therefore, the fluo-
rescence assay was used to measure the inhibition
kinetics of PFTase by analogues 3a and 3b. The rate of
yPFTase- and hPFTase-catalyzed farnesylation of N-
Ds-GCVIA was measured in the presence of fixed con-
centrations of FPP at varying concentrations of 3a and
3b. The concentrations of 3a and 3b used were chosen
based on their IC50 values (Table 1) which were deter-
mined in preliminary experiments to be 12 µM and 3.4
µM. Double reciprocal plots of the data for 3a with
yPFTase (Figure 5) and 3a with hPFTase (Figure 6) both
give patterns of lines that intersect on the 1/ν axis,
consistent with competitive inhibition with respect to
FPP.
containing hPFTase, the results are similar to those
obtained with yPFTase and the chemically synthesized
products, giving a [M]+ peak of 972.36 ( 0.17 for 10a
and a [M]+ peak of 972.35 ( 0.14 for 10b as well as
similar ESI MS/MS patterns (data not shown). To
quantify the amount of product formed in all four cases,
the chemically synthesized products were injected in an
amount equal to that formed in the enzymatic reactions
assuming 100% conversion and the peak areas integrated
using fluorescence detection. The amount of enzymatic
product formed by yPFTase was 23% for 10a and 81%
for 10b; however, the large scale reactions were per-
formed with 18-fold greater amounts of enzyme and were
allowed to react for 6-fold longer times than would be
required with the natural substrate, FPP. For 10a ,
enzymatically prepared by hPFTase, a yield of 8% was
achieved and for 10b a 20% yield was observed. These
reactions were accomplished with 4-fold greater amounts
of enzyme and reacted for 6-fold longer reaction times
than required with FPP; thus, these analogues are poor
substrates for the human enzyme. It is useful to compare
the results described here with earlier observations with
related analogues. With the C10 benzoylbenzoate ester
analogues, 3f and 3g, and yPFTase, the yields of product
Similar results were achieved with 3b using the yeast

3258 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
Ta ble 1. Su m m a r y of In h ibition P a r a m eter s for
Ben zop h en on e-Ba sed P h otoa ffin ity La belin g An a logu es
w ith Yea st a n d Hu m a n P F Ta ses
yFPTase
photoinact
yFPTase
compd IC50 (µM)
yFPTase
KI (µM)
after
hFPTase
hFPTase
KI (µM)
12 h (%) IC50 (µM)
3a
3b
12 ( 0.55
1.0 ( 0.12
51
53
74
63
-
44
30
-
2.3 ( 0.32 0.34 ( 0.04
3.6 ( 0.18 0.61 ( 0.14
3.4 ( 0.41 0.61 ( 0.14
3ca
3d a
3eb
3fc
5.1
2.0
37
3.0
2.3
3.7
1.6
0.85
2.9
0.96
6.0
0.91
0.38
0.70
0.045
0.049
0.075 (KD)e
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
3gc
3h b
3id
3jd
40
40
FPP
0.012 (KD)f
a
d
From ref 33. b From ref 27. c From ref 26. From ref 34. e From
ref 20. f From ref 21.
F igu r e 6. Double reciprocal plot showing the competitive
inhibition of hPFTase by 3a . The concentrations of 3a are as
follows: (O) 0 µM, (b) 1.0 µM, (0) 2.0 µM, (9) 3.0 µM, (4) 4.0
µM. Inset: replot of slopes from double reciprocal plot versus
[3a ].
may be due to steric interactions with the protein or it
may reflect the greater rigidity of the ester and amide-
containing analogues. Comparison of Ki values of 3a and
3b with those for 3i and 3j indicates that the presence
of the second isoprene unit significantly increases the
affinity of these compounds for PFTase. However, this
enhanced affinity is accompanied by a commensurate
decrease in incorporation efficiency that is consistent
with size limitations imposed by the PFTase isoprenoid
binding pocket as evidenced by the X-ray crystal struc-
ture. As we have noted with previous compounds,28 the
diphosphate functionality present in 3a and 3b is es-
sential for binding. No inhibition at concentrations of up
to 20 µM was observed with the free alcohols 7a and 7b.
X-r a y Cr ysta llogr a p h ic Stu d ies. To gain additional
insight into how these isoprenoid diphosphate analogues
interact with a prenyltransferase, crystals of 3a bound
to rPFTase were obtained, and the structure of the
resulting complex was determined via X-ray diffraction
experiments. Electron density maps clearly show the
bound conformation and location of 3a in the PFTase
active site cavity. The presence of the higher molecular
weight phosphate atoms in the molecule assists in
unambiguous definition of the binding mode. Comparison
of the protein in the structure of the FPT:3a complex and
the FPT:FPP:SCH66336 ternary complex shows a 0.2 Å
rms deviation for all protein atoms, indicating minimal
movement of protein side-chains.33,41 The position of the
diphosphate atoms are well conserved between 3a and
FPP while the benzophenone occupies the binding site
of the second and third isoprenoid units of FPP. A
superposition of 3a and FPP in the isoprenoid binding
pocket of rPFTase is shown in Figure 7. These structural
data combined with the observations that 3a is a slow
alternative substrate and modest inhibitor suggest that
3a is a good structural analogue of FPP.
F igu r e 5. Double reciprocal plot showing the competitive
inhibition of yPFTase by 3a . The concentrations of 3a are as
follows: (O) 0 µM, (b) 3.0 µM, (0) 6.0 µM, (9) 9.0 µM, (4) 12
µM. Inset: replot of slopes from double reciprocal plot versus
[3a ].
and human enzymes (data not shown). The rates were
further analyzed by the Eadie-Hofstee method to deter-
mine Ki values for each inhibitor. A Ki of 1.0 µM was
calculated for 3a , while 3b gave a Ki value of 0.61 µM
using yPFTase. For inhibition studies on hPFTase, a Ki
of 0.34 µM for 3a and 0.61 µM for 3b were obtained. By
comparing these findings to values obtained with other
analogues studied (Table 1), it is clear that structural
features within the analogues affect their binding affini-
ties for the protein prenyltransferases. This includes
variations in the isoprenoid chain length, type of linkage
(amide, ester or ether), and benzophenone regiochemistry
(meta- versus para-substituted benzophenone). Compari-
son of Ki values of 3a and 3b with yPFTase and hPFTase
reveals small differences in their binding affinity. For all
yPFTase Ki values, the meta-linked analogues are more
efficient or at least as efficient as the para-linked
analogues at binding; however, with hPFTase a change
in efficiency occurs, suggesting that the shapes of the
isoprenoid binding pocket of these two enzymes may be
different. By comparing the C5 ether-linked analogues
3a and 3b with the C5 ester-linked compounds 3c and
3d , and amide-linked analogue, 3e, it can be seen that
the carbonyl groups diminish the extent of binding. This
(41) Strickland, C. L.; Weber, P. C.; Windsor, W. T.; Wu, Z.; Le, H.
V.; Albanese, M. M.; Alvarez, C. S.; Cesarz, D.; Rosario, J . d.; Deskus,
J .; Mallams, A. K.; Njoroge, F. G.; Piwinski, J . J .; Remiszewski, S.;
Rossman, R. R.; Taveras, A. G.; Vibulbhan, B.; Doll, R. J .; Girijavall-
abhan, V. M.; Ganguly, A. K. J . Med. Chem. 1999, 42, 2125-2135.

Synthesis of Farnesyl Diphosphate Analogues
J . Org. Chem., Vol. 66, No. 10, 2001 3259
F igu r e 7. Stereoview of the binding of 3a (black) compared with FPP (gray) to rPFTase. The non-carbon atoms of 3a are shown
as small spheres. The rPFTase active site zinc is shown as a large black spot on the enzyme surface.
P h otoin a ctiva tion of yP F Ta se. Compounds 3a and
3b were tested for their ability to inactivate yPFTase
upon UV irradiation at 350 nm by performing experi-
ments containing mixtures of the inhibitors and the
enzyme, withdrawing aliquots at regular intervals, and
assaying the resulting samples for enzyme activity (data
not shown). Samples containing enzyme alone were
irradiated for up to 12 h with a low pressure 350 nm lamp
with no appreciable loss in enzyme activity. In contrast,
samples irradiated for 2 h containing 3a or 3b at
saturating concentrations resulted in a 12% decrease in
enzyme activity with 3a and a 9% decrease with 3b.
Furthermore, the enzyme was protected from inactivation
by 3a or 3b; upon addition of 1 (100 µM), only 8%
inactivation was observed with 3a and 1% inactivation
was obtained with 3b after 2 h irradiation. Greater levels
of inactivation were achieved with longer incubation
times. Photolysis of yPFTase in the presence of 3a
resulted in 51% inactivation after 12 h and with 3b
resulted in 53% inactivation. It should be noted that no
inactivation occurred when yPFTase was incubated with
3a or 3b in the dark, and that inactivation ceased when
irradiation of the reaction mixtures was stopped. Plots
of enzyme activity versus irradiation time in the presence
of 3a or 3b at saturating concentrations (data not shown)
lead to values for the apparant kinact of 0.060 ( 0.002 h-1
and 0.061 ( 0.006 h-1, respectively. These values are
relatively slow; for example Colman and co-workers
reported a value of 15 h-1 for kinact for photoinactivation
of glutathione S-transferase by a benzophenone-contain-
ing inhibitor.42 While this may be due in part to the low
intensity of the UV source used in our experiments, we
believe it is also related to the type of residues that
comprise the FPP binding site in PFTase. It is interesting
to note from the crystal structure of the rPFTase-3a
complex that all of the residues within 6 Å of the
benzophenone carbonyl group (W102, W106, W303, Y361,
and Y365) contain aromatic side chains. Hydrogens
attached to sp2-hybridized carbons are among the least
reactive C-H bonds in proteins.23 From the structure of
the protein complex shown in Figure 7, it is also apparent
that the two aromatic rings of the benzophenone chro-
mophore in 3a are orthogonal to one another. Thus, for
light absorption at 350 nm to occur, the molecule must
adopt a planar, higher energy conformation, thereby
further decreasing the rate of photoinactivation. Com-
parison of photoinactivation results (Table 1) of yPFTase
in the presence of 3a and 3b to 3i and 3j indicates these
shorter chain ether analogues have a slightly higher
cross-linking efficiency. Comparing these analogues to
the C5 ester compounds 3c and 3d , which gave the best
cross-linking efficiencies reported to date for all our
analogues, suggests that the conformational rigidity
provided by the ester group may be important in fixing
the analogues in the active site area relative to the C5
ether analogues which have greater rotational flexibility.
Additional crystallographic studies of the ester complexes
and rPFTase will be useful in clarifying this issue.
P h otolabelin g of P r otein P r en yltr an sfer ases with
[32P ]3a a n d [32P ]3b. To examine the ability of 3a and
3b to label yPFTase, hPFTase, and hPGGTase-I upon
photolysis, the radiolabeled analogues [32P]3a and [32P]-
3b were synthesized and purified using a method previ-
ously described,43 adapted from a procedure developed
by Bukhtiyarov.20 Photolysis of prenyltransferases (380
nM) with 350 nm light for 2 h in the presence of [32P]3a
or [32P]3b at concentrations of [32P]3a and [32P]3b that
were 10-fold above Ki for yPFTase resulted in preferential
labeling of the â subunit for each enzyme (Figure 8:
yPFTase and [32P]3a , Figure 9: hPFTase and [32P]3b,
and Figure 10: hPGGTase-I and [32P]3a ; lane 2′ in each
Figure).
Addition of the natural substrate, 1 or 2, to the
photolysis reactions containing [32P]3a or [32P]3b resulted
in protection from labeling (Figure 8: yPFTase and [32P]-
3a , Figure 9: hPFTase and [32P]3b, and Figure 10:
hPGGTase-I and [32P]3a ; lane 3′ in each Figure). To
quantitate the relative labeling efficiencies of each sub-
unit for each reaction, phosphorimaging analysis was
employed. For photolabeling of yPFTase with [32P]3a and
[32P]3b, the â subunit was labeled 1.6-fold over the R
subunit. Addition of 1 in the photolysis reaction resulted
(42) Wang, J .; Bauman, S.; Colman, R. F. Biochemistry 1998, 37,
15671-15679.
(43) Turek, T. C.; Gaon, I.; Distefano, M. D. J . Labeled Compds.
Radiopharm. 1997, 39, 139-146.

3260 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
F igu r e 10. Analysis of photolabeling of hPGGTase-I with [32P]-
3a by SDS-PAGE. Lanes 1 and 1′ contain samples of hPG-
GTase-I and [32P]3a which were not irradiated. Lanes 2 and
2′ contain samples of hPGGTase-I irradiated at 350 nm in the
presence of [32P]3a . Lanes 3 and 3′ contain samples of
hPGGTase-I irradiated in the presence of [32P]3a and 2. Lanes
1, 2, and 3 show silver-stained proteins. Lanes 1′, 2′, and 3′
show the radiolabeled proteins.
F igu r e 8. Analysis of photolabeling of yPFTase with [32P]3a
by SDS-PAGE. Lanes 1 and 1′ contain samples of yPFTase
and [32P]3a which were not irradiated. Lanes 2 and 2′ contain
samples of yPFTase irradiated at 350 nm in the presence of
[32P]3a . Lanes 3 and 3′ contain samples of yPFTase irradiated
in the presence of [32P]3a and 1. Lanes 1, 2, and 3 show silver-
stained proteins. Lanes 1′, 2′, and 3′ show the radiolabeled
proteins.
agreement with the extent of protection obtained in the
photoinactivation experiments described above; in that
case inactivation in the presence of 3b and 1 decreased
to 11% of that obtained in the absence of the 1. In
contrast, total labeling of yPFTase in the presence of 3a
and 1 decreased to only 54% of that obtained in the
absence of the competitor. Again rough agreement with
the extent of protection obtained in the photoinactivation
experiments was observed; inactivation in the presence
of 3a and 1 decreased to 75% (or 50% in a different
experiment) of that obtained in the absence of the 1.
Interestingly, the greater labeling specificity obtained
with 3b compared with 3a (in terms of protection from
labeling of yPFTase in the presence of 1) was also
observed in photolabeling experiments with hPFTase.
Total labeling of hPFTase in the presence of 3b and 1
decreased to 25% of that obtained in the absence of the
1 while a decrease to 44% was observed in similar
experiments performed with 3a and 1. Last, photolysis
of hPGGTase-I by [32P]3a and [32P]3b gave 8.0-fold
labeling of â over R with [32P]3a and a 6.4-fold â over R
subunit labeling for [32P]3b. Addition of 2 in the pho-
tolysis reaction with [32P]3a resulted in a 8.1-fold de-
crease in â subunit labeling and a 1.3-fold decrease in R
subunit labeling. Similar experiments with [32P]3b re-
sulted in a 6.5-fold decrease in â subunit labeling and a
1.3-fold decrease in R subunit labeling. These results
suggest that the â subunit is involved in prenyl group
recognition for all three enzymes and are consistent with
our work previously reported using other benzophenone
analogues with yPFTase and hPGGTase-I as well as
work obtained using other photoactive analogues on all
three enzymes.19-22,28,29,44 Moreover, these data provide
additional information concerning the isoprenoid binding
sites of these enzymes. Earlier work with benzophenone
analogues prepared from geraniol (3f-j) suggested that
the C-14 locus of GGPP was proximal to the â subunit
in the isoprenoid/PGGTase complex as shown in Figure
11. This is based on the observation that C-14 of GGPP
is close to the ketone carbon of 3f-j when these ana-
logues are superimposed with GGPP. In the present
F igu r e 9. Analysis of photolabeling of hPFTase with [32P]3b
by SDS-PAGE. Lanes 1 and 1′ contain samples of hPFTase
and [32P]3b which were not irradiated. Lanes 2 and 2′ contain
samples of hPFTase irradiated at 350 nm in the presence of
[32P]3b. Lanes 3 and 3′ contain samples of hPFTase irradiated
in the presence of [32P]3b and 1. Lanes 1, 2, and 3 show silver-
stained proteins. Lanes 1′, 2′, and 3′ show the radiolabeled
proteins.
in a 3.3-fold decrease in â subunit labeling and a 1.4-
fold decrease in R subunit labeling for [32P]3a . Similar
experiments with [32P]3b resulted in a 4.0-fold decrease
in â subunit labeling and a 3.1-fold decrease in R subunit
labeling. In the case of hPFTase photolabeling by [32P]-
3a and [32P]3b, the â subunit was labeled 4.3-fold and
8.3-fold over the R subunit, respectively. Addition of 1
resulted in 2.0-fold decrease in â labeling and a 3.8-fold
decrease in R labeling for [32P]3a . A 6.2-fold decrease in
â labeling and a 1.0-fold decrease in R labeling occurred
with [32P]3b in the presence of 1. In general, the labeling
using the meta-substituted analogue, 3b, appears to be
more specific. Total labeling of yPFTase (R + â) in the
presence of 3b and 1 decreased to 28% of that obtained
in the absence of the competitor, 1. This is in rough

Synthesis of Farnesyl Diphosphate Analogues
J . Org. Chem., Vol. 66, No. 10, 2001 3261
Con clu sion
Compounds 3a and 3b are analogues of FPP that can
be efficiently prepared in six steps from dimethylallyl
alcohol. Both compounds are modest competitive inhibi-
tors of yPFTase and hPFTase, but at higher enzyme
concentrations and longer reaction times 3b can be
effectively transferred to a peptide substrate. A crystal
structure with rat farnesyltransferase illustrates that 3a
is a good substrate mimic. Photolysis of 3a and 3b with
several different prenyltransferases results in the specific
labeling of these proteins and suggests that the C-10 loci
of FPP and GGPP are bound at sites on the â subunits
of their respective transferases. Analogues 3a and 3b also
label FPP synthase. Given their ease of synthesis, higher
stability relative to ester analogues and good photo-
labeling properties, these compounds should be useful for
a variety of studies with other isoprenoid-utilizing en-
zymes.
F igu r e 11. Model for the interaction of benzophenone-based
photoaffinity labeling analogues with prenyltransferases.
Exp er im en ta l Section
Meth od s a n d Ma ter ia ls. All reactions were conducted
under dry nitrogen and stirred magnetically. Analytical TLC
was performed on precoated (0.25 mm) silica gel 60F-254
plates purchased from E. Merck. Visualization was accom-
plished under UV irradiation or by subjecting the plates to
ethanolic phosphomolybdic acid solutions followed by heating.
Flash chromatography silica gel (60-120 mesh) was obtained
by E. M. Science. CH2Cl2 and CH3CN were distilled from CaH2,
and THF was distilled from sodium/benzophenone ketyl.
Trifluoroacetic acid (TFA) was purchased from Fisher. NMR
J values are given in Hz. UV spectra were obtained using a
Hewlett-Packard 8452A spectrophotometer, and fluorescence
measurements were performed with a Perkin-Elmer LS 50B
luminescence spectrometer. HPLC analysis was carried out
using a Beckman Model 127/166 instrument equipped with a
diode array UV detector and a Beckman 6300A fluorescence
detector (305-395 nm excitation filter and 480-520 nm
emission filter). Preparative HPLC separations were per-
formed with a Rainin Dynamax Microsorb C18 column (2.14 ×
25 cm with a 5 cm guard column), while analytical separations
employed a Rainin Dynamax Microsorb C18 column (5 µm, 4.6
× 250 mm with a 15 mm guard column). Phosphorimaging
analysis was performed with a Molecular Dynamics 445 SI
Phosphorimager. Sep-pak columns were obtained from Waters
and Dowex 50W-X8 resin was obtained from Bio-Rad. N-
dansylglycine was purchased from Sigma. N-Dansyl-GCVIA
was purchased from Bio-Synthesis, Inc. and purified by
reversed-phase preparative HPLC. [32P]H3PO4 (specific activity
8500-9120 Ci/mmol) was obtained from DuPont NEN. E. coli
DH5R/pGP114 and purified FPP synthase were generous gifts
from C. D. Poulter, Department of Chemistry, University of
Utah. E. coli BL21(DE3)/pRD577 and purified human PFTase
were provided generously to us from C. A. Omer, Merck
Research Laboratories, and P. J . Casey, Department of Mo-
lecular Cancer Biology, Duke University Medical Center,
respectively.
F igu r e 12. Analysis of photolabeling of FPP synthase with
[32P]3b by SDS-PAGE. Lanes 1 and 1′ contain samples of FPP
synthase and [32P]3b which were not irradiated. Lanes 2 and
2′ contain samples of FPP synthase irradiated at 350 nm in
the presence of [32P]3b. Lanes 3 and 3′ contain samples of FPP
synthase irradiated in the presence of [32P]3b and 1. Lanes 1,
2, and 3 show silver-stained proteins. Lanes 1′, 2′, and 3′ show
the radiolabeled proteins.
study, the subunit labeling results obtained with 3a and
3b suggest that the C-10 regions of FPP and GGPP also
interact with the respective â subunits of these enzymes.
While this was already known with PFTase (from the
X-ray structure), it is a new piece of structural informa-
tion in the case of hPGGTase.
P h otola belin g of F P P Syn th a se w ith [32P ]3a a n d
[32P ]3b. FPP synthase catalyzes the condensation of
isopentenyl pyrophosphate and either dimethylallyl py-
rophosphate or geranyl pyrophosphate generating FPP
(1). Although this enzyme catalyzes a similar prenyl
transfer reaction to those catalyzed by prenyltrans-
ferases, there is little sequence homology between these
enzymes.19 To determine if [32P]3a and [32P]3b are
capable of photolabeling FPP synthase, the enzyme (465
nM) was irradiated with 350 nm light for 2 h in the
presence of [32P]3a (13 µM) or [32P]3b (6.6 µM) which
resulted in labeling FPP synthase (Figure 12; lane 2′,
labeling with [32P]3b).
1-[(Te t r a h yd r o-2H -p yr a n -2-yl)oxy]-3-m e t h yl-2-b u -
ten e (4). To a solution of 3-methyl-2-buten-1-ol (3.4 g, 40
mmol) in CH2Cl2 (40 mL) were added 3,4-dihydropyran (5.0
g, 60 mmol) and PPTs (1.0 g, 4.0 mmol); the resulting solution
was stirred for 4 h at room temperature. The reaction mixture
was partially evaporated, diluted with Et2O, and washed with
saturated NaHCO3. The organic layer was dried over Na2SO4
and concentrated to yield a colorless oil (6.3 g, 93%). Rf 0.62
(silica gel, toluene/EtOAc, 5:2, v/v); 1H NMR (200 MHz, CDCl3):
δ 1.40-1.85 (m, 6H), 1.64 (s, 3H), 1.70 (s, 3H), 3.40-3.51 (m,
1H), 3.79-3.86 (m, 1H), 3.92 (dd, 1H, J ) 6.0, 12.0), 4.17 (dd,
1H, J ) 6.0, 12.0), 4.58 (t, 1H, J ) 4.0), 5.28 (t, 1H, J ) 6.0);
13C NMR (50.3 MHz, CDCl3): δ 17.7, 19.5 (primary), 25.2, 25.6,
30.5, 62.0, 63.4 (secondary), 97.6, 120.8 (tertiary), 136.6
Addition of 1, to the photolysis reactions containing
[32P]3a or [32P]3b resulted in protection from labeling
(Figure 12; lane 3′, labeling with [32P]3b). To quantitate
the relative labeling efficiencies for each reaction, phos-
phorimaging analysis was employed (data not shown).
For photolabeling with [32P]3a , addition of 1 in the
photolysis reaction resulted in a 8.1-fold decrease in
labeling and for [32P]3b resulted in a 12-fold decrease in
labeling. From these studies it is apparent that these
analogues are capable of binding other enzymes and
should be useful for studying other enzymes that utilize
isoprenoid diphosphates as substrates.

3262 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
(quaternary); LR-CI MS (NH3): m/z (relative intensity) 85.07
(100), 103.07 (34), 170.13 ([M]+, 14), 171.14 ([M + H]+, 40);
HR-CI MS: for C10H18O2 [M]+, calcd 170.1302, found 170.1299,
for C10H19O2 [M + H]+, calcd 171.1380, found 171.1387.
(E)-1-[(Tetr ah ydr o-2H-pyr an -2-yl)oxy]-3-m eth yl-2-bu ten -
4-ol (5). Compound 4 (6.3 g, 37 mmol) and tert-butyl hydro-
peroxide (14 mL, 120 mmol) were stirred in the presence of
H2SeO3 (480 mg, 3.4 mmol) and salicylic acid (480 mg, 3.4
mmol) in CH2Cl2 (40 mL) for 10 h at room temperature. The
CH2Cl2 was removed under reduced pressure and the tert-butyl
hydroperoxide was removed by repeated washing with toluene.
The residue was dissolved in Et2O, washed with NaHCO3 to
remove H2SeO3, dried over Na2SO4, and filtered. The organic
layer was concentrated and the crude product purified by flash
chromatography on silica gel (toluene/EtOAc, 5:2, v/v) to afford
compound 5 as a colorless oil (2.3 g, 33%). Rf 0.25 (silica gel,
toluene/EtOAc, 5:2, v/v); 1H NMR (300 MHz, CDCl3): δ 1.48-
1.81 (m, 6H), 1.65 (s, 3H), 3.44-3.49 (m, 1H), 3.80-3.86 (m,
1H), 3.97 (s, 2H), 4.02 (dd, 1H, J ) 6.0, 12.0), 4.24 (dd, 1H, J
) 6.0, 12.0), 4.44 (t, 1H, J ) 4.0), 5.58 (t, 1H, J ) 6.0); 13C
NMR (75.4 MHz, CDCl3): δ 13.7 (primary), 19.3, 25.3, 30.5,
62.1, 63.2, 67.5 (secondary), 97.8, 120.7 (tertiary), 139.2
(quaternary); LR-CI MS (NH3): m/z (relative intensity) 85.06
(100), 169.12 (56), 187.13 ([M + H]+, 32); HR-CI MS: for
(106 mg) and DEAD (90 mg) were added. The mixture was
then stirred for 24 h at room temperature followed by a third
aliquot of the same quantities of PPh3 and DEAD. After a final
18 h of reaction, hexane (10 mL) was added to the mixture to
precipitate triphenylphosphine oxide and diethyl hydrazine-
dicarboxylate. The flask was stored overnight at 4 °C, and the
solid byproducts were removed by filtration. The filtrate was
evaporated, and the residue was applied to a silica gel column
which was eluted with toluene/EtOAc (5:2, v/v) giving 6c in
79% yield as white crystals. mp: 56-58 °C; Rf 0.57 (silica gel,
toluene/EtOAc, 5:2, v/v); 1H NMR (300 MHz, CDCl3): δ 1.54-
1.86 (m, 6H), 1.74 (s, 3H), 3.49-3.52 (m, 1H), 3.83-3.86 (m,
1H), 4.04-4.11 (m, 1H), 4.23-4.33 (m, 1H), 4.25 (s, 2H), 4.62
(t, 1H, J ) 6.0), 5.55 (t, 1H, J ) 6.0), 7.73-7.81 (m, 2H), 7.90-
7.99 (m, 2H); 13C NMR (75.5 MHz, DEPT, CDCl3): δ 15.0
(primary), 19.4, 25.5, 30.6, 44.5, 62.2, 63.2 (secondary), 97.8,
123.4, 133.4, 134.0 (tertiary), 132.0, 133.9, 168.1 (quaternary);
LR-FAB MS (MNBA): m/z (relative intensity) 154 (45), 214.2
(100), 232.2 (8), 316.2 ([M + H]+, 12); HRFAB MS: for C18H22
-
NO4 [M + H]+, calcd 316.1542, found 316.1569; NOE from H
at C-2 (5.55 ppm) to H at C-4 (4.25 ppm): 1.7%; NOE from H
at C-2 (5.55 ppm) to CH3 at C-3 (1.74 ppm): 0.2%.
(E)-4-(4′-Ben zoylben zyloxy)-3-m eth yl-2-bu ten -1-ol (7a ).
To a solution of 6a (380 mg, 1.0 mmol) in EtOH (6.0 mL) was
added PPTs (25 mg, 0.10 mmol) and the reaction stirred at 55
°C for 8 h. The solvent was partially evaporated in vacuo, the
resulting solution was diluted with Et2O, and washed with
half-saturated brine to remove the catalyst. The organic layer
was dried over Na2SO4 and concentrated to yield 7a (300 mg,
100%). Rf 0.29 (silica gel, toluene/EtOAc, 5:2, v/v); 1H NMR
(300 MHz, CDCl3): δ 1.75 (s, 3H), 3.98 (s, 2H), 4.24 (d, 2H, J
) 6.0), 4.57 (s, 2H), 5.73 (t, 1H, J ) 9.0), 7.47 (d, 2H, J ) 9.0),
7.49 (t, 2H, J ) 6.0), 7.60 (t, 1H, J ) 6.0), 7.80 (d, 4H, J )
9.0); 13C NMR (75.4 MHz, CDCl3): δ 14.1 (primary), 59.0, 71.2,
75.8 (secondary), 126.7, 127.2, 128.3, 130.0, 130.3, 132.5
(tertiary), 135.0, 136.7, 137.6, 143.3, 196.6 (quaternary); LR-
FAB MS (MNBA): m/z (relative intensity) 167.2 (22), 195.2
(100), 211.1 (11), 279.2 (10), 289.2 (5), 297.2 ([M + H]+, 5);
HR-FAB MS: for C19H21O3 [M + H]+, calcd 297.1485, found
297.1484.
C
10H18O3 [M]+, calcd 186.1256, found 186.1260, for C10H19O3
[M + H]+, calcd 187.1334, found 187.1340.
(E)-4-(4′-Ben zoylben zyloxy)-1-[(tetr a h yd r o-2H-p yr a n -
2-yl)oxy]-3-m eth yl-2-bu ten e (6a ). Compound 5 (190 mg, 1.0
mmol) was dissolved in THF (2.0 mL) and cooled to 0 °C prior
to sodium hydride (90 mg of 60% dispersion in oil, 2.3 mmol)
addition. After 1 h, 4-benzoylbenzyl bromide35 (450 mg, 1.6
mmol) was added and stirred for 1 h at 0 °C followed by 24 h
at room temperature. The reaction mixture was poured over
water and extracted with EtOAc. The combined extracts were
dried over Na2SO4 and filtered. The filtrate was evaporated
and the residue purified by flash chromatography on silica gel
(toluene/EtOAc, 5:2, v/v). Evaporation of the solvent gave 6a
as a light yellow oil (300 mg, 78%). Rf 0.42 (silica gel, toluene/
EtOAc, 5:2, v/v); 1H NMR (300 MHz, CDCl3): δ 1.53-1.86 (m,
6H), 1.78 (s, 3H), 3.53-3.57 (m, 1H), 3.89-3.92 (m, 1H), 4.02
(s, 2H), 4.13 (dd, 1H, J ) 6.0, 12.0), 4.36 (dd, 1H, J ) 6.0,
12.0), 4.59 (s, 2H), 4.68 (t, 1H, J ) 3.6), 5.72 (t, 1H, J ) 6.0),
7.49 (d, 2H, J ) 6.0), 7.50 (t, 2H, J ) 9.0), 7.61 (t, 1H, J )
6.0), 7.82 (d, 4H, J ) 9.0); 13C NMR (75.4 MHz, CDCl3): δ
14.2 (primary), 19.6, 25.5, 30.7, 62.3, 63.3, 71.2, 75.9 (second-
ary), 98.1, 124.1, 127.2, 129.3, 130.0, 130.3, 132.4 (tertiary),
135.9, 136.7, 137.7, 143.3, 196.4 (quaternary); LR-FAB MS
(MNBA): m/z (relative intensity) 167.1 (25), 195.1 (100), 279.5
(8), 297.2 (7), 381.2 ([M + H]+, 2); LR-FAB MS: for C24H29O4
[M + H]+, calcd 381.2, found 381.2.
(E)-4-(3′-Ben zoylben zyloxy)-3-m eth yl-2-bu ten -1-ol (7b).
The procedure described for the preparation of 7a was used
for the synthesis of 7b. The product was obtained in 100% yield
(300 mg). Rf 0.30 (silica gel, toluene/EtOAc, 5:2, v/v); 1H NMR
(300 MHz, CDCl3): δ 1.70 (s, 3H), 3.95 (s, 2H), 4.21 (d, 2H, J
) 6.0), 4.54 (s, 2H), 5.69 (t, 1H, J ) 6.0), 7.46 (t, 1H, J ) 6.0),
7.48 (t, 2H, J ) 9.0), 7.58 (d, 1H, J ) 9.0), 7.59 (t, 1H, J )
9.0), 7.71 (d, 1H, J ) 9.0), 7.79 (s, 1H), 7.80 (d, 2H, J ) 6.0);
13C NMR (75.4 MHz, CDCl3): δ 14.1 (primary), 58.9, 71.4, 75.7
(secondary), 126.7, 128.3, 128.4, 129.1, 129.4, 130.1, 131.7,
132.6 (tertiary), 135.0, 137.5, 137.7, 138.7, 196.8 (quaternary);
LR-FAB MS (MNBA): m/z (relative intensity) 154 (45), 195.1
(100), 213.1 (10), 279.2 (17), 297.1 ([M + H]+, 5); HR-FAB
MS: for C19H21O3 [M + H]+, calcd 297.1485, found 297.1494.
(E)-4-(4′-Ben zoylb en zyloxy)-3-m et h yl-2-b u t en -1-b r o-
m id e (8a ). To the alcohol 7a (180 mg, 0.50 mmol) in CH2Cl2
(4.0 mL) were added CBr4 (190 mg, 0.60 mmol) and PPh3 (160
mg, 0.60 mmol). The reaction was stirred under N2 for 4 h at
room temperature. Hexanes were added to precipitate PPh3O
which was subsequently removed by filtration. The filtrate was
concentrated under reduced pressure to afford the bromide in
quantitative yield. Rf 0.97 (silica gel, toluene/EtOAc, 5:2, v/v);
1H NMR (300 MHz, CDCl3): δ 1.80 (s, 3H), 4.01 (s, 2H), 4.06
(d, 2H, J ) 6.0), 4.58 (s, 2H), 5.87 (t, 1H, J ) 9.0), 7.49 (t, 2H,
J ) 6.0), 7.51 (d, 2H, J ) 6.0), 7.60 (t, 1H, J ) 6.0), 7.82 (d,
4H, J ) 9.0); 13C NMR (75.4 MHz, CDCl3): δ 13.7 (primary),
28.0, 71.4, 75.1 (secondary), 122.7, 127.2, 128.3, 130.0, 130.3,
132.4 (tertiary), 136.2, 138.0, 138.2, 142.9, 200.1 (quaternary);
LR-FAB MS (MNBA): m/z (relative intensity) 195.2 (20),
279.2 (100), 359.0/361.0 ([M + H]+, 3); HR-FAB MS: for
(E)-4-(3′-Ben zoylben zyloxy)-1-[(tetr a h yd r o-2H-p yr a n -
2-yl)oxy]-3-m eth yl-2-bu ten e (6b). Compound 6b was pre-
pared and purified using the procedure described for the
synthesis of 6a . After chromatography, 6b was obtained as a
light yellow oil (260 mg, 68%). Rf 0.43 (silica gel, toluene/
EtOAc, 5:2, v/v); 1H NMR (200 MHz, CDCl3): δ 1.52-1.83 (m,
6H), 1.71 (s, 3H), 3.45-3.55 (m, 1H), 3.81-3.91 (m, 1H), 3.95
(s, 2H), 4.07 (dd, 1H, J ) 6.0, 12.0), 4.29 (dd, 1H, J ) 6.0,
12.0), 4.52 (s, 2H), 4.62 (t, 1H, J ) 4.0), 5.66 (t, 1H, J ) 6.0),
7.46 (t, 1H, J ) 8.0), 7.50 (t, 2H, J ) 8.0), 7.56 (d, 1H, J )
6.0), 7.57 (t, 1H, J ) 6.0), 7.69 (d, 1H, J ) 8.0), 7.76 (s, 1H),
7.79 (d, 2H, J ) 6.0); 13C NMR (75.4 MHz, CDCl3): δ 14.1
(primary), 21.3, 25.5, 30.7, 62.2, 63.2, 71.3, 75.7 (secondary),
98.0, 124.0, 127.6, 128.3, 129.0, 129.3, 130.0, 131.7, 132.4
(tertiary), 135.9, 137.5, 137.7, 138.9, 196.4 (quaternary); LR-
FAB MS (MNBA): m/z (relative intensity) 165.1 (23), 195.2
(100), 279.2 (16), 298.2 (30), 381.2 ([M + H]+, 3); LR-FAB
MS: for C24H29O4 [M + H]+, calcd 381.2, found 381.2.
(E)-1-[(Tet r a h yd r o-2H -p yr a n -2-yl)oxy]-3-m et h yl-4-N-
p h th a la m id o-2-bu ten e (6c). To a stirred flask containing 5
(186 mg, 1.0 mmol), phthalimide (145 mg, 1.0 mmol), and PPh3
(212 mg, 1.0 mmol) dissolved in 5.0 mL of THF was added a
solution of DEAD (174 mg, 1.0 mmol) in 2.0 mL of THF
dropwise over 30 min. The reaction was allowed to proceed
for 18 h at room temperature at which time additional PPh3
C
19H20O2Br [M + H]+, calcd 359.0641, found 359.0652.
(E)-4-(3′-Ben zoylb en zyloxy)-3-m et h yl-2-b u t en -1-b r o-
m id e (8b). Compound 8b was prepared from 7b using the
procedure outlined for the synthesis of 8a . The desired bromide
8b was obtained in quantitative yield. Rf 0.98 (silica gel,

Synthesis of Farnesyl Diphosphate Analogues
J . Org. Chem., Vol. 66, No. 10, 2001 3263
1
i
toluene/EtOAc, 5:2, v/v); H NMR (300 MHz, CDCl3): δ 1.78
in PrOH/NH4OH/H2O (6:3:1, v/v/v) followed by phosphorim-
aging analysis and was found to be 35%. The sample also
contained 32% monophosphate, 23% triphosphate, and 10%
inorganic phosphate.
(s, 3H), 3.99 (s, 2H), 4.05 (d, 2H, J ) 9.0), 4.56 (s, 2H), 5.85 (t,
1H, J ) 9.0), 7.49 (t, 1H, J ) 6.0), 7.50 (t, 2H, J ) 6.0), 7.60
(d, 1H, J ) 9.0), 7.59 (t, 1H, J ) 9.0), 7.74 (d, 1H, J ) 6.0),
7.81 (s, 1H), 7.83 (d, 2H, J ) 9.0); 13C NMR (75.4 MHz,
CDCl3): δ 13.7 (primary), 28.0, 71.5, 75.0 (secondary), 122.8,
128.3, 128.4, 128.7, 130.1, 131.6, 132.1, 132.5 (tertiary), 137.5,
137.7, 138.5, 138.9, 196.0 (quaternary); LR-FAB MS
(MNBA): m/z (relative intensity) 195.2 (50), 279.2 (100), 359.0/
361.0 ([M + H]+, 8); HR-FAB MS: for C19H20O2Br [M + H]+,
calcd 359.0641, found 359.0649.
(E)-4-(4′-Ben zoylben zyloxy)-3-m eth yl-2-bu ten -1-diph os-
p h a te (3a ). The allylic bromide 8a (130 mg, 0.50 mmol) was
pyrophosphorylated with [(n-Bu)4N]3HP2O7 (1.1 g, 1.2 mmol)
in anhydrous CH3CN (4.0 mL) for 5 h at room temperature.
The product was converted to the NH4 salt using an ion-
exchange column (Dowex 50W-X8 resin, NH4 form) equili-
[r,â(n )32P ]-(E )-4-(3′-Be n zoylb e n zyloxy)-3-m e t h yl-2-
bu ten -1-d ip h osp h a te ([32P ]3b). Compound [32P]3b was pre-
pared as described above for [32P]3a . After reversed-phase
chromatography, [32P]3b was obtained in 1% yield (specific
activity 1300 Ci/mol) and 76% radiochemical purity. The
amount of inorganic phosphate was 24%.
N-Da n syl-GC(4-BBDMA)VIA (10a ). Compound 10a was
prepared by reacting 8a with 9 under acidic conditions
containing DMF/butanol/0.025% TFA (2:1:1, v/v/v) and 1 M
Zn(OAc)2 as a catalyst following the procedure described by
Xue et al.40 The crude product was purified by reversed-phase
HPLC employing a H2O/CH3CN (containing 0.2% TFA) gradi-
ent: ESI MS and ESI MS/MS were obtained. ESI MS for
+
+
brated in and eluted with 25 mM NH4HCO3/2-propanol (49:1,
v/v), and the salt was obtained by lyophilization. The final
product was purified employing a reversed-phase C18 column
(Sep-pak cartridge) eluted with a step gradient containing 25
mM NH4HCO3 and CH3CN. After lyophilization, 3a (160 mg,
32%) was obtained as a white solid. Rf 0.56 (iPrOH/NH4OH/
H2O, 6:3:1, v/v/v); 1H NMR (300 MHz, D2O, adjusted to pH
8.5 with ND4OD): δ 1.62 (s, 3H), 3.94 (s, 2H), 4.42 (t, 2H, J )
6.0), 4.51 (s, 2H), 5.63 (t, 1H, J ) 6.0), 7.43 (d, 2H, J ) 9.0),
7.44 (t, 2H, J ) 9.0), 7.59 (t, 1H, J ) 6.0), 7.66 (d, 4H, J )
6.0); 31P NMR (121.4 MHz, D2O, pH 8.5 with ND4OD): δ -6.87
(d, 1P, J ) 22), -10.79 (d, 1P, J ) 23); UV (25 mM NH4HCO3):
λmax ) 262 nm, ꢀ ) 17,100 M-1‚cm-1; LR-ESI MS (CH3OH/
H2O/TFA): m/z (relative intensity) - Ve mode, 455.0 ([M -
H]-, 100); + Ve mode, 242.2 (100), 279 (35), 457.1 ([M + H]+,
17); HR-MALDI MS (DHBA/CH3CN/H2O/TFA): for C19H21O9P2
[M-H]+, calcd 455.0653, found 455.0669; 97% chemical purity
based on HPLC integration analysis. HPLC analysis was
performed with a C18 Phenomenex Luna column (250 × 4.6
mm) eluted with a flow rate of 0.4 mL/min using a gradient of
solvent A (25 mM NH4HCO3) and solvent B (CH3CN) and
detecting at 262 nm. Elution was performed by isocratic elution
with 20% solvent B for 10 min followed by a 25 min linear
gradient to 60% solvent B and finally a 5 min linear gradient
to 100% solvent B with isocratic elution at 100% B for 10 min.
Under these conditions, 3a eluted at 25.1 min.
(E)-4-(3′-Ben zoylben zyloxy)-3-m eth yl-2-bu ten -1-diph os-
p h a te (3b). The allylic bromide 8b was pyrophosphorylated
as described for the preparation of 3a to give 3b as a white
powder (150 mg, 30%). Rf 0.58 (iPrOH/NH4OH/H2O, 6:3:1, v/v/
v); 1H NMR (300 MHz, D2O, adjusted to pH 8.5 with ND4OD):
δ 1.54 (s, 3H), 3.86 (s, 2H), 4.35 (t, 2H, J ) 6.0), 4.41 (s, 2H),
5.54 (t, 1H, J ) 6.0), 7.39 (t, 4H, J ) 6.0), 7.54 (d, 2H, J )
6.0), 7.59 (d, 2H, J ) 9.0), 7.63 (s, 1H); 31P NMR (121.4 MHz,
D2O, pH 8.5 with ND4OD): δ -6.89 (d, 1P, J ) 22), -10.81
(d, 1P, J ) 22); UV (25 mM NH4HCO3): λmax ) 260 nm, ꢀ )
26,000 M-1‚cm-1; LR-ESI MS (CH3OH/H2O/TFA): m/z (relative
intensity) - Ve mode, 219.5 (31), 245.0 (16), 454.9 ([M - H]-,
100); + Ve mode, 242.2 (100), 279 (76), 457.0 ([M + H]+, 26);
HR-MALDI MS (DHBA/CH3CN/H2O/TFA): for C19H21O9P2 [M
- H]+, calcd 455.0653, found 455.0654; 72% chemical purity
based on HPLC integration analysis using the conditions
described for 3a . Under these conditions, 3b eluted at 25.1
min. The major impurity eluting at 43.6 min (22% based on
integration) present in 3b is the alcohol 7b or isomer thereof.
[r,â(n )32P ]-(E )-4-(4′-Be n zoylb e n zyloxy)-3-m e t h yl-2-
bu ten -1-d ip h osp h a te ([32P ]3a ). Alcohol 7a (1.9 mg, 6.3 µmol)
was reacted with anhydrous [32P]H3PO4 (1.2 mg, 13 µmol) in
CH3CN (200 µL) containing 20% (v/v) CCl3CN and triethy-
lamine (2.5 mg, 26 µmol) for 2 h. The volatile components from
the reaction were evaporated, and the resulting residue
purified by reversed-phase chromatography using a C18 Sep-
pak cartridge employing a NH4HCO3/CH3CN stepwise gradi-
ent. The product [32P]3a was obtained in 16% yield (specific
activity 2100 Ci/mol); the concentration of solutions containing
[32P]3a were determined by UV using the wavelengths and
extinction coefficients reported for 3a . The radiochemical
purity of [32P]3a was assessed by thin-layer chromatography
C
50H64O10N6S2 [M]+, calcd 972.41, found 971.99 ( 0.57,
C50H65O10N6S2 [M + H]+, calcd 973.4, found 973.2, C50H66
-
O
10N6S2 [M + 2H]2+, calcd 487.2, found 487.0; UV (H2O/CH3-
CN/TFA, 50/50/0.2) λmax ) 320 nm, ꢀ ) 1780 M-1‚cm-1
;
fluorescence (H2O/CH3CN/TFA, 50/50/0.2) λex ) 322 nm, λem
) 530 nm, intensity ) 0.37 (relative to N-dansylglycine, 1.0).
N-Da n syl-GC(3-BBDMA)VIA (10b). Compound 10b was
prepared from 8b and 9 as described for 10a : ESI MS and
ESI MS/MS were obtained. ESI MS for C50H64O10N6S2 [M]+,
calcd 972.41, found 972.17 ( 0.31, C50H65O10N6S2 [M + H]+,
calcd 973.4, found 973.2, C50H66O10N6S2 [M + 2H]2+, calcd
487.2, found 487.0; UV (H2O/CH3CN/TFA, 50/50/0.2) λmax
)
320 nm, ꢀ ) 1780 M-1‚cm-1; fluorescence (H2O/CH3CN/TFA,
50/50/0.2) λex ) 322 nm, λem ) 530 nm, intensity ) 0.24
(relative to N-dansylglycine, 1.0).
P u r ifica tion of yP F Ta se a n d h P GGTa se-I. yPFTase was
purified as described by Mayer et al.45 and published in our
earlier work.28 hPGGTase-I was purified using modifications
of published procedures.19,46,47
En zym e Assa ys. A spectrophotometric assay was employed
to monitor yPFTase activity as previously described by Gaon
et al.28 modified from the original published procedure.38,39
Su bstr a te Stu d ies. Large-scale enzymatic reactions were
performed to determine if 3a and 3b were alternate substrates
for yPFTase and hPFTase. The reactions (10 mL) contained
10 µM 3a or 3b, 45 nM yPFTase, or 100 nM hPFTase, and
other assay components at the same concentration omitting
the natural substrate, 1, and detergent, n-dodecyl-â-D-malto-
side. The reactions were desalted and concentrated by applying
to a Sep-pak C18 cartridge. The samples were washed with
H2O/CH3CN/TFA (95:5:0.2, v/v/v) and eluted with CH3CN/TFA
(100:0.2, v/v). The fluorescent fractions were concentrated in
vacuo, dissolved in H2O/CH3CN/TFA (50:50:0.2, v/v/v), and
purified by HPLC using a reversed-phase C18 column with a
H2O/CH3CN/TFA gradient. The product peaks were collected
and analyzed by ESI MS and ESI MS/MS.
En zym e In h ibition Exp er im en ts. To determine if 3a and
3b were competitive inhibitors of yPFTase and hPFTase,
duplicate assays were run in which the natural substrate, 1,
was maintained at fixed concentrations (1.0, 1.5, 3.0, 7.0 µM),
and 3a (0, 3.0, 6.0, 9.0, 12 µM) or 3b (0, 1.0, 2.0, 3.0, 4.0 µM)
for yPFTase; 3a or 3b (0, 1.0, 2.0, 3.0, 4.0 µM) for hPFTase
were varied at each substrate concentration. The analogue
concentrations were chosen based on their IC50 values per-
formed first. The concentration of yPFTase and hPFTase in
these experiments were 2.5 nM and 25 nM, respectively. The
reaction rates were determined from initial velocity measure-
ments using the fluorescence assay and Ki values were
(44) Das, N. P.; Allen, C. M. Biochem. Biophys. Res. Commun. 1991,
181, 729-735.
(45) Mayer, M. P.; Prestwich, G. D.; Dolence, J . M.; Bond, P. D.;
Wu, H.-Y.; Poulter, C. D. Gene 1993, 132, 41-47.
(46) Zhang, F. L.; Diehl, R. E.; Kohl, N. E.; Gibbs, J . B.; Giros, B.;
Casey, P. J .; Omer, C. A. J . Biol. Chem. 1994, 269, 3175-3180.
(47) Stirtan, W. G.; Poulter, C. D. Arch. Biochem. Biophys. 1995,
321, 182-190.

3264 J . Org. Chem., Vol. 66, No. 10, 2001
Turek et al.
calculated from Eadie-Hofstee plots obtained from a Macintosh
PowerMac G3 computer running Kaleidagraph (v. 3.0.1)
software.
0.01% bromophenol blue) was then added to each sample,
samples were heated to 70 °C for 5-10 min, and analyzed by
SDS-PAGE with a 10% Tris-tricine gel. Gels were silver
stained and exposed to a phosphorimaging screen for 14-22
h. The relative intensities of photolabeled products were
determined by phosphorimaging analysis.
X-r a y Cr ysta llogr a p h ic Stu d ies. Crystals of the FPT:3a
complex were prepared by soaking 3a into preformed crystals
using methods previously described.41 X-ray diffraction data
were collected on a Rigaku rotating anode generator equipped
with osmic mirrors and a Raxis-IV image plate detector. With
the detector set at 135 mm, data were collected in 236
contiguous 0.30° oscillation images each exposed for 9 min.
The data extend to 2.1 Å resolution and have a Rmerge of 3.5%
with a 4.4-fold multiplicity. The structure was refined using
XPLOR98.1 (Molecular Simulations Inc.) to an Rfactor of 19.9%.
[A figure showing the electron density contoured at 1.0σ for
3a can be found in the Supporting Information section. ]
P h otoin a ctiva tion Stu d ies. Photolysis reactions were
performed in silanized quartz test tubes (10 × 45 mm) and
contained 52 mM Tris‚HCl pH 7.0, 5.8 mM DTT, 12 mM
MgCl2, 12 µM ZnCl2, 25 mM NH4HCO3, 13 µM 3a or 6.6 µM
3b which were chosen to be 10-fold above their Ki values, and
3.8 nM yPFTase in a final volume of 0.75 mL. For substrate
protection studies, 1 was added at a final concentration of 100
µM. All reactions were photolyzed at 350 nm for up to 12 h
using a UV Rayonet minireactor equipped with 8 RPR-3500°
lamps and a circulating platform at 4 °C. At various photolysis
times, 50 µL aliquots were removed, placed on ice, and assayed
for activity.
P h otolysis Rea ction of F P P Syn th a se w ith [32P ]3a a n d
[
32P ]3b. The photolysis reactions were accomplished as de-
scribed above with the prenyltransferases except 470 nM FPP
synthase was used, and substrate protection studies were
completed with 1 at 100 µM. Reactions were irradiated for 2
h using the same apparatus. Loading buffer was then added
to each sample, samples were heated to 70 °C for 5-10 min,
and analyzed by SDS-PAGE with a 10% Tris-tricine gel. Gels
were silver stained and exposed to a phosphorimaging screen
for 42 h. The relative intensity of photolabeled products were
determined by phosphorimaging analysis.
Ack n ow led gm en t. The authors would like to thank
Dr. Leo Bonilla and Nancy Raha, University of Min-
nesota Cancer Institute, for obtaining the mass spectral
data. This research was supported by funds from the
American Cancer Society (BE-222 and J FRA-584) and
the National Institutes of Health (GM58842).
P h otolysis Rea ction of P r en yltr a n sfer a ses w ith [32P ]-
Su p p or tin g In for m a tion Ava ila ble: Copies of 1H NMR,
31P NMR and mass spectral data for 3a and 3b and intermedi-
ates in their syntheses as well as HPLC chromatograms of 3a
and 3b are included. Plots of photoinactivation kinetics,
photolabeling efficiencies, and a figure showing the electron
density for 3a from the rPFTase-3a complex contoured at 1.0σ
are also enclosed. This material is available free of charge via
the Internet at http://pubs.acs.org.
3a a n d
[
32P ]3b. The photolysis reactions (100 µL) were
performed in silanized quartz test tubes (10 × 45 mm) and
contained 52 mM Tris‚HCl pH 7.0, 5.8 mM DTT, 12 mM
MgCl2, 12 µM ZnCl2, 25 mM NH4HCO3, 13 µM [32P]3a or 6.6
µM [32P]3b, and 380 nM enzyme (yPFTase, hPFTase, hPGG-
Tase-I). For substrate protection studies, 1 or 2 was added to
a final concentration of 100 µM. Reactions were irradiated for
2 h using the above apparatus. Loading buffer (50µL; 4% SDS,
12% glycerol (m/v), 50 mM Tris, 2% mercaptoethanol (v/v),
J O991130X