Grignard-mediated synthesis and preliminary biological evaluation of novel 3-substituted farnesyl diphosphate analogues

Article abstract of DOI:10.1016/S0960-894X(00)00337-1

A series of substituents was installed at the 3 position of farnesyl diphosphate through a copper-cyanide mediated coupling of a vinyl triflate with various Grignard reagents. These novel FPP mimetics were then evaluated as inhibitors of or substrates for mammalian protein farnesyl transferase. The IC₅₀ values for these compounds range from 18 to 10,100 nm, with the 3-isopropenyl analogue being one of the most potent FPP-mimetic mFTase inhibitors yet synthesized. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00337-1

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1763±1766
Grignard-Mediated Synthesis and Preliminary Biological
Evaluation of Novel 3-Substituted Farnesyl Diphosphate
Analogues
Todd J. Zahn,a Carolyn Weinbaumb and Richard A. Gibbsa,*
aDepartment of Pharmaceutical Sciences, College of Pharmacy and AHP, Wayne State University, Detroit, MI 48202, USA
bDepartment of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, NC 27710, USA
Received 21 March 2000; accepted 8 June 2000
AbstractÐA series of substituents was installed at the 3 position of farnesyl diphosphate through a copper-cyanide mediated cou-
pling of a vinyl tri¯ate with various Grignard reagents. These novel FPP mimetics were then evaluated as inhibitors of or substrates
for mammalian protein farnesyl transferase. The IC50 values for these compounds range from 18 to 10,100 nm, with the 3-isopro-
penyl analogue being one of the most potent FPP-mimetic mFTase inhibitors yet synthesized. #2000 Elsevier Science Ltd. All
rights reserved. # 2000 Elsevier Science Ltd. All rights reserved.
Protein farnesyl transferase (FTase) is an enzyme that
carries out the isoprenoid modi®cation of proteins con-
taining a CAAX box motif. In particular, this ca. 90
kDa heterodimeric enzyme catalyzes the addition of a
15-carbon farnesyl moiety to the terminal cysteine sulf-
hydryl of several key signal transduction proteins,
including Ras proteins. This farnesylation of Ras is the
®rst and obligate step for its participation in a signal
transduction cascade. Mutated Ras proteins have been
implicated in ca. 30% of all human cancers, with the
highest incidence found in pancreatic tumors (ca. 90%).1
Therefore, inhibition of FTase has become the object of
intense research e€orts.1 Signi®cant progress has been
made in the development of peptidomimetic and non-
substrate based FTase inhibitors, and in vivo evaluation
of some of these compounds has indicated promising
potential for their development into anticancer agents.2
However, less work has been initiated on farnesyl
diphosphate based inhibitors of FTase (FPP; 1a, Figure
1). Consequently, less is known regarding the speci®city
of FTase for its isoprenoid substrate. Recently, we have
found that 3-vFPP (1b) is an alternative substrate for
mammalian FTase, while 3-alFPP (1c) is a potent and
selective inhibitor of this enzyme.3 The introduction of a
tert-butyl group at the 3-position leads to one of the
most potent FPP-based mFTase inhibitors, 3-tbFPP
(1d). Moreover, the alcohol precursors of 1b and 1c
inhibit the growth of ras transformed cells, apparently
via two di€erent mechanisms.3 The fact that this subtle
change in functionality leads to large di€erences in activity
has provided the impetus for the preparation and testing
of a wide variety of 3-substituted FPP analogues. Herein,
we report the Grignard-mediated synthesis and pre-
liminary biological evaluation of several novel 3-sub-
stituted FPP analogues as potential mFTase inhibitors.
The natural biosynthetic route to isoprenoids is well-
characterized, but is of little use for the synthesis of
analogues. Therefore, synthetic chemical routes must be
used instead. The procedures used for the synthesis of
isoprenoids via various synthetic methodologies have
been continually improved and updated. Methods that
have been employed for chain elongation of isoprenyl
derivatives include Wittig condensations,4 Biellman
couplings,5 and palladium-catalyzed cross-coupling
reactions.6 We have developed a novel and ecient
synthetic route leading to a wide variety of isoprenoids,
in particular 3-substituted farnesyl and geranylgeranyl
analogues.7a,b This method is based on a vinyl phos-
phate route used to synthesize isoprenoids developed by
Sum and Weiler.8
Previously we have reported the coupling of 3 with a
wide variety of organotin,7 organoboron,9 and organo-
copper10 reagents. We have now enhanced our arsenal
of synthetic methodologies with a copper-cyanide medi-
ated coupling of 3 with Grignard reagents. This new
method has been used to introduce various substituents
*Corresponding author. Tel.: +1-313-577-3761; fax: +1-313-577-
2033; e-mail: rag@wizard.pharm.wayne.edu
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00337-1

1764
T. J. Zahn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1763±1766
at the 3 position of FPP (Scheme 1). The synthetic route
begins, as described previously, with the conversion of
b-ketoester 2 to the vinyl tri¯ate 3 in a highly stereo-
selective fashion. We have recently established that the
cuprate reagent derived from [13C]-methylmagnesium
iodide and copper cyanide couples in an e€ective fash-
ion with tri¯ate 3, and this was employed as a key step
in the preparation of [5,15-bis-13C]-FPP.11a Since a wide
variety of alkyl, alkenyl, and aryl Grignard reagents are
commercially available, this appeared to be an attractive
procedure for the synthesis of a wide variety of 3-sub-
stituted FPP analogues.
allylmagnesium chloride yielded at best only a few per-
cent of the desired ester. Note also that the reaction
appears to proceed much better in pure ether than in
ether/THF solvent mixtures.10b Chloride, bromide and
iodide Grignard reagents all can be used e€ectively in
the coupling reaction. Reduction of the 3-substituted
esters, followed by the two-step phosphorylation proce-
dure of Poulter and co-workers,12 a€orded the desired
3-substituted FPP analogues 1d±l.
The newly synthesized FPP analogues were evaluated as
potential substrates or for inhibitors of mammalian
protein±farnesyl transferase. Kinetic values were deter-
mined using a modi®ed version of a continuous spec-
tro¯uorometric mFTase assay with the peptide
cosubstrate dansyl-GCVLS-OH.13,14 This assay was
employed because it can be readily used to determine if
new FPP analogues are substrates, or inhibitors. The
assay was used ®rst to determine an IC50 value for the
previously characterized analogue 1c, and this was in
reasonably good agreement with the previously deter-
mined value.
Preliminary studies indicated that the previously pre-
pared tert-butyl ester 4d could be synthesized in excel-
lent yield from the organocopper reagent prepared from
tert-butylmagnesium chloride and a slight excess of
copper cyanide. Therefore we have used the same pro-
tocol for the synthesis of the eight 3-substituted esters
4e±l (Scheme 1).11b The yields of the esters are modest in
many cases; however, note that these coupling reactions
(and the subsequent steps) are unoptimized and in most
cases the reported yield is from a single reaction
attempt. The majority of the Grignard reagents tried
produced the desired ester. However, the attempted
coupling of the cuprate reagent derived from 2-methyl-
(1-propenyl)magnesium bromide and tri¯ate 3 a€orded
only starting material. Moreover, several attempts with
The results observed with analogues 1e±1l con®rm that
subtle changes in functionality at the 3 position can lead
to dramatic changes in activity (Table 1 and Figure 2).
Removal of one methyl group from the potent tert-butyl
compound 1d leads to the equally potent isopropyl
analogue 1e. Interestingly, the similar isobutyl com-
pound 1f is ca. eight times less potent than 1e. Perhaps
the extra carbon in the isobutyl group of 1f interferes
with its tight binding to the active site of mFTase. To
expand on this hypothesis, we synthesized the neopentyl
analogue 1g. Consistent with this trend, 1g is a strik-
ingly poor inhibitor of mFTase. Note in particular the
several 100-fold di€erence in binding anity seen
between the tert-butyl analogue 1d and the neopentyl
analogue 1g. The sec-butyl analogue 1h was prepared to
generate an analogue intermediate in size between 1e
and 1f. It is evident that 1h retains high anity for the
Figure 1. Previously prepared 3-substituted FPP analogues. (a) Values
from ref 3. (b) Under the assay conditions used in this paper.
Scheme 1. Grignard-mediated synthesis of new FPP analogues.

T. J. Zahn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1763±1766
1765
Table 1. Relative abilities of compounds 1b±1l to act as inhibitors of
or substrates for mammalian protein±farnesyl transferasea
of a vinyl group and the tight ®tting isopropyl group
leads to a synergistic enhancement of binding to the
active site of mFTase. In fact, 1i is one of the most
potent FPP based inhibitors of mFTase developed.15 It
is quite surprising, however, that the isopropenyl ana-
logue 1i is not an alternative substrate for the enzyme
like the vinyl analogue 1b. The n-propyl FPP analogue
1l was synthesized to evaluate further the e€ect of
unsaturation at the 3-position, in comparison to the
allyl derivative 1c. In agreement with the increased
potency of 1i in comparison to 1e, the allyl derivative 1c
is signi®cantly more potent than 11. Note also the ben-
e®cial e€ects of branching on the 3-alkyl chain; compare
the n-propyl analogue 1l with the more potent 1e and
1h.
b
Analogue (3-substituent)
IC50 (nM)
krel
Km (nM)
1b (vinyl)
1c (allyl)
1d (tert-butyl)
1e (isopropyl)
1f (isobutyl)
1g (neopentyl)
1h (sec-butyl)
1i (isopropenyl)
1j (4-¯uorophenyl)
1k (cyclopentyl)
1l (propyl)
173c
119 (189)c
31c
0.521c
0.053c
nd
0.013
0.000
0.042
0.000
0.000
0.002
0.072
0.078
156c
nd
nd
nd
nd
nd
nd
nd
nd
102
nd
35
290
6600
102
18
10,100
nd
453
aValues were determined using a ¯uorescence assay as previously
described in ref 3 (nd=not determined). Note that the concentrations
of all analogues were veri®ed by phosphate analysis.14 Recombinant
mFTase was expressed in baculovirus and puri®ed as previously
described.18
bRelative Vmax determined for each analogue at 10 mM, compared to
the Vmax determined for FPP (10 mM) under the same conditions.
cValues previously reported in ref 3.
To further probe the steric constraints of the active site
of FTase, additional bulky functional groups were
added to the 3-position of FPP. The 3-(4-¯uorophenyl)
analogue 1j was chosen for its hydrophobic phenyl
group, and the additional electron withdrawing char-
acteristics of the ¯uorine atom.16 Surprisingly, this
compound was the least potent FPP analogue examined
thus far, and in particular it should be noted that it
binds more than an order of magnitude more weakly to
mFTase than the corresponding unsubstituted 3-phenyl
FPP analogue (IC50=299 nM).3 To generate an analo-
gue intermediate in bulk between the potent isopropyl
analogue 1e and the ine€ective 3-(4-¯uorophenyl) ana-
logue 1j, we synthesized the cyclopentyl compound 1k.
One might assume, on the basis of steric considerations,
that 1k would exhibit a potency somewhere between 1e
and 1j. However, 1k was not an inhibitor at all, but was
instead a substrate. The Km for 1k is 102 nM, suggesting
that it binds tightly to mFTase, but its Vmax appears to
be relatively modest. Surprisingly, an IC50 value for this
active site of mFTase, with an IC50 value between those
of 1e and 1f. Note that all four of these analogues (1e-
1h) are very poor substrates for the enzyme.
Previous work in our laboratory has demonstrated that
the 3-vinyl FPP analogue 1b is an alternative substrate
for mFTase.3 It is possible that the additonal con-
jugated double bond in 1b makes it more reactive and
thus a better prenyl donor than the other FPP analo-
gues we have prepared. In view of this proposal, we
designed the isopropenyl analogue 1i, which has a com-
bination of the functionalities of both the potent iso-
propyl substituted inhibitor 1e and the alternative
substrate 1b. Apparently the combination of the pi bond
Figure 2. The 3-substituted FPP analogues synthesized in this study, with the IC50 values determined for them in parentheses.

1766
T. J. Zahn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1763±1766
analogue could not be determined, as no concentration
dependent decrease in mFTase activity was observed
with increasing concentrations of 1k. This ®nding is a
sobering reminder of the complicated kinetic mechan-
ism of mFTase,17 and its potentially complex mode of
interaction with a substrate analogue like 3-cyclopentyl
FPP in the presence of FPP and a peptide substrate.
8. Sum, F. W.; Weiler, L. Can. J. Chem. 1979, 57, 1431.
9. Mu, Y. Q.; Gibbs, R. A. Tetrahedron Lett. 1995, 36, 5669.
10. (a) Mu, Y. Q.; Gibbs, R. A.; Eubanks, L. M.; Poulter, C.
D. J. Org. Chem. 1996, 61, 8010. (b) A similar solvent e€ect
was also previously seen in the coupling of tert-butyl-
Cu(CN)Li with 3 (ref 10a).
11. (a) Zahn, T. J.; Ksebati, M. B.; Gibbs, R. A. Tetrahedron
Lett. 1998, 39, 3991. (b) Representative coupling procedure:
Ethyl 3-isopropenyl-7,11-dimethyldodeca-2(Z), 6(E), 10-trien-
oate (4i). In a ¯ame dried, argon-¯ushed ¯ask, CuCN (1.08
mmol, 97.0 mg) was dissolved in 1.04 mL anhydrous diethyl
ether. This solution was cooled to 78 ꢀC and isopropenyl
magnesium bromide (0.5 M in THF, 0.64 mmol, 1.28 mL) was
added dropwise. The mixture was warmed to 0 ꢀC and main-
tained at that temperature for 5 min, then recooled to 78 ꢀC.
A solution of tri¯ate 3 (0.43 mmol, 173 mg) in 1.0 mL ether
was added dropwise and the reaction was stirred for 2.5 h at
78 ꢀC. The reaction mixture was then warmed to 0 ꢀC and
quenched with 2 mL NH4Cl. The aqueous layer was extracted
with ether (3Â20 mL). The combined organic layers were dried
over MgSO4, ®ltered, and concentrated. Flash chromato-
graphy (98:2 hexane:ethyl acetate) gave 80 mg (64%) of the
ester 4i.
In conclusion, we have extended the utility and applic-
ability of the vinyl tri¯ate route for the synthesis of 3-
substituted FPP analogues. In particular, we have
expanded our synthetic capability through the use of a
copper-cyanide mediated coupling of a vinyl tri¯ate
with alkyl, alkenyl and aryl magnesium halides. Fur-
thermore, the preliminary biological assessment of these
FPP mimetics has demonstrated that subtle changes in
functionality at the three position can lead to large and
surprising di€erences in activity. In addition, we report
one of the most potent FPP based inhibitors designed
thus far, the isopropenyl derivative 1i. It is evident that
mimics of the natural isoprenoid substrate of mFTase
present unique and useful targets for the design of
potent inhibitors of the enzyme.
12. Davisson, V. J.; Woodside, A. B.; Neal, T. R.; Stremler,
K. E.; Muehlbacher, M.; Poulter, C. D. J. Org. Chem. 1986,
51, 4768.
13. Pompliano, D. L.; Gomez, R. P.; Anthony, N. J. J. Am.
Chem. Soc. 1992, 114, 7945.
14. Cassidy, P. B.; Dolence, J. M.; Poulter, C. D. Methods
Enzymol. 1995, 250, 30.
Acknowledgements
This work was supported by the NIH (CA 78819 to
RAG and GM 46372 to Professor Patrick J. Casey).
This paper was taken from the dissertation submitted by
TJZ to WSU in partial ful®llment of the requirements
for the Ph.D. degree. RAG was supported in part by the
American Cancer Society Junior Faculty Research
Award (JFRA-609), and TJZ was supported in part by
a WSU GRA award. We thank Professor Mark Dis-
tefano and his laboratory for assistance with the
mFTase variant of the ¯uorimetric assay.
15. Some representative potent FPP-derived mFTase inhibi-
tors: (a) a-Hydroxyfarnesylphosphonate (Ki=5 nM): Pom-
pliano, D. L.; Rands, E.; Schaber, M. D.; Mosser, S. D.;
Anthony, N. J.; Gibbs, J. B. Biochemistry 1992, 31, 3800. (b)
Two potent phosphonate derivatives of farnesol and farnesyl
amine (IC50=75 and 83 nm): Patel, D. V.; Schmidt, R. J.;
Biller, S. J.; Gordon, E. M.; Robinson, S. S.; Manne, V. J.
Med. Chem. 1995, 38, 2906. (c) A farnesyl phosphonate deri-
vative of phenylalanine (IC50=80 nm): Lamothe, M.; Perrin,
D.; Blotieres, D.; Leborgne, M.; Gras, S.; Bonnet, D.; Hill, B.;
Halazy, S. Bioorg. Med. Chem. Lett. 1996, 6, 1291. (d) ref 16.
(e) Very potent (low nanomolar IC50), structurally diverse
(non-isoprenoid) FPP-competitive mFTase inhibitors have
also been described; for example, see: Aoyama, T.; Satoh, T.;
Yonemoto, M.; Shibata, J.; Nonoshita, K.; Arai, S.; Kawa-
kami, K.; Iwasawa, Y.; Saroh, H.; Tamaka, K.; Monden, Y.;
Koder, T.; Arakawa, H.; Suzuki-Takahashi, I.; Kamei, T.;
Tomimofo, K. J. Med Chem. 1998, 41, 143.
References and Notes
1. Cox, A. D.; Der, C. J. Biochem. Biophys. Acta 1997, 1333, F51.
2. (a) Oli€, A. Biochem. Biophys. Acta 1999, 1423, C19. (b)
Leonard, D. M. J. Med. Chem. 1997, 40, 2971.
3. Gibbs, B. S.; Zahn, T. J.; Mu, Y. Q.; Sebolt-Leopold, J.;
Gibbs, R. A. J. Med. Chem. 1999, 42, 3800.
4. Klinge, S.; Demuth, M. Synlett. 1993, 783.
16. -Tri¯uoromethyl FPP is the most potent FPP-based inhi-
bitor of yeast FTase (Ki=11 nm): Dolence, J. M.; Poulter, C.
D. Proc. Natl. Acad. Sci. USA 1995, 92, 5008.
5. Altman, L. J.; Ash, L.; Marson, S. Synthesis, 1974, 129.
6. Bouzbouz, S., Kirschleger, B. Synlett, 1994, 763.
7. (a) Gibbs, R. A.; Krishnan, U.; Dolence, J. M.; Poulter, C.
D. J. Org. Chem 1995, 60, 7821. (b) Shao, Y.; Eummer, J. T.;
Gibbs, R. A. Org Lett. 1999, 1, 627.
17. (a) Tschantz, W. R.; Fur®ne, E. S.; Casey, P. J. J. Biol.
Chem. 1997, 272, 9989. (b) Huang, C.-C.; Hightower, K. E.;
Fierke, C. A. Biochemistry 2000, 39, 2593.
18. Moomaw, J. F.; Zhang, F. L.; Casey, P. J. Methods
Enzymol. 1995, 250, 12.