Convenient methods for the synthesis of P1-farnesyl-P 2-indicator diphosphates

Article abstract of DOI:10.1016/j.bmcl.2005.12.016

To access P¹-farnesyl-P²-indicator diphosphates, more efficient methods for the synthesis of farnesyl-phosphate and diphosphates were developed. The procedures reported here provide more flexible conditions than the conventional imidazolide and morpholidate coupling methods. Milder conditions for the synthesis of sensitive allylic diphosphates and greatly improved reaction efficiencies provide access to novel reagents for analysis of diphosphate-based enzymatic reactions.

Full text of DOI:10.1016/j.bmcl.2005.12.016

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1665–1667
1
Convenient methods for the synthesis of P -farnesyl-P -indicator
2
diphosphates
Birte K. Feld and Gregory A. Weiss*
Department of Chemistry, University of California, Irvine, 516 Rowland Hall, Irvine, CA 92697-2025, USA
Received 10 November 2005; accepted 5 December 2005
Available online 10 January 2006
1
2
Abstract—To access P -farnesyl-P -indicator diphosphates, more efficient methods for the synthesis of farnesyl-phosphate and
diphosphates were developed. The procedures reported here provide more flexible conditions than the conventional imidazolide
and morpholidate coupling methods. Milder conditions for the synthesis of sensitive allylic diphosphates and greatly improved reac-
tion efficiencies provide access to novel reagents for analysis of diphosphate-based enzymatic reactions.
Ó 2005 Elsevier Ltd. All rights reserved.
Plants, fungi, and other organisms synthesize terpenoid
natural products from allylic diphosphate starting mate-
1
rials. Sensitive assays for production of terpenoids
could link consumption of the allylic diphosphate start-
ing material with turnover of an indicator dye. To devel-
op novel substrates for monitoring such reactions, we
report efficient syntheses of farnesyl diphosphates at-
tached to indicator moieties as shown in Figure 1. The
improved syntheses described here are applicable to a
wide range of biosynthetic substrates.
Figure 1.
4
,5
functioning as a leaving group. In our hands, this
method, especially the second step, proved unsatisfacto-
ry due to instability of the resultant diphenyl diphos-
phate. Other coupling methods have also been
reported, but, in general, the procedures are unsatisfac-
tory, due to their slow reaction rates and poor yields
Allylic diphosphates are unstable, especially in the pres-
2
,3
ence of acids or strong nucleophiles, which compli-
cates their synthesis and manipulation. Furthermore,
such amphiphilic diphosphates often form emulsions,
preventing the removal of inorganic impurities by
extraction.
6
1
(
e.g., 2% overall yield for P -(tetra-O-acetyl-a-D-galac-
2
topyranosyl) P -farnesyl diphosphate from farnesol).
4
1
2
Typically, P -terpenoid-P -substituted diphosphates are
formed in the following three steps. First, an allylic
phosphate is synthesized via Cramer’s method (activa-
tion of the alcohol with trichloro-acetonitrile and reac-
tion with bis(triethylammonium) phosphate), which
Recently, Ryu and Scott reported an efficient method
for coupling commercially available diphosphates to al-
lyl chlorides or homoallyl tosylates. The yields achieved
7
through this method are very good, but require avail-
ability of the mono-substituted diphosphate. In most
cases, however, the diphosphate must be synthesized
first, which is often non-trivial. For example, the synthe-
sis of 5-bromo-4-chloro-3-indolyl diphosphate from the
commercially available monophosphate is quite time
consuming and provides only low yields of the diphos-
phate. Therefore, we focused on the development of a
different method for the synthesis of farnesyl diphos-
phate derivatives that would access a wide variety of
compounds from readily available starting materials.
3
usually results in a mixture of mono- and diphosphates.
Second, the monophosphate is converted to a P -allyl-
1
2
P -diphenyl diphosphate, which is finally coupled to a
second phosphate with the diphenyl phosphate moiety
Keywords: Disubstituted diphosphate; Farnesyl derivatives; Enzyme
pseudosubstrates; Indoxyl derivative.
*
Corresponding author. Tel.: + 011 949 824 5566; fax: +011 949 824
920; e-mail: gweiss@uci.edu
9
0
960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.016

1666
B. K. Feld, G. A. Weiss / Bioorg. Med. Chem. Lett. 16 (2006) 1665–1667
0
The reactivity of a phosphate or diphosphate anion is
strongly influenced by the associated counter ion. Solubil-
ity, stability, and nucleophilicity of the respective phos-
phate can be tuned by the choice of counter ion.
Davisson et al. have demonstrated that the tetra-butylam-
monium counter ion greatly improves the yield of terpe-
noid diphosphate formation from allylic chlorides or
Table 1. Coupling of phosphates with 1,1 -carbonyldiimidazole
0
a
8
R
R
Time
87 h
Yield
bromides. Therefore, we also used a tetra-butylammoni-
um salt for the formation of farnesyl monophosphate.
b
c,d
c,d
Me (4)
C
15
H
25 (3)
25 (3)
16% (8)
16% (9)
Accordingly, 85% phosphoric acid was titrated with 40%
tetra-butylammonium hydroxide to pH 9.8. Lyophiliza-
tion yielded tacky, highly hygroscopic bis(tetra-n-butyl-
ammonium) hydrogen phosphate (2). This salt was
reacted with freshly prepared farnesyl bromide in
e
(
5)
C
15
H
87 h
f
c,d
g
(6)
C
15
H25 (3)
145 h 14% (10)
9
DMF for 18 h. The crude product was submitted to a
column of Dowex 50WX8 (NH -form). Inorganic
+
4
phosphate was removed through precipitation with
organic solvents and the resulting product was extracted
with ether to remove traces of unreacted farnesyl bro-
mide (1). The farnesyl monophosphate (3) obtained by
this method was used for subsequent reactions without
further purification (Scheme 1).
b
h
(
7)
n.r.
c,d
c,d
b
C
C
15
H
H
25 (3)
25 (3)
Me (4)
87 h
12% (8)
f
15
(6)
145 h 41% (10)
Two different methods for coupling farnesyl phosphate
to a series of different phosphates were investigated.
c,d
b
(7) 87 h
0
C
15
H
25 (3)
3% (11)
Cramer, Schaller, and Staab introduced 1,1 -carbonyldi-
imidazole (CDI) as a useful coupling reagent for phos-
1
0
phates,
but Ryu and Scott reported very slow
a
b
c
32ꢀ
Based upon the initial monophosphate (ROPO ).
Free acid.
Bisammonium salt.
coupling of geranyl phosphate with the imidazolides of
adenosine and guanosine monophosphates, including
7
d
e
reaction times greater than a month. By optimizing
15
C H25 = farnesyl.
Bis(cyclo-hexylammonium) salt.
Free acid or p-toluidine salt.
the reaction conditions, we were able to decrease the
reaction time to 3–6 days and obtain good yields for
the resulting diphosphates.
f
g
h
After removal of methylcarbamate moiety.
2ꢀ
Initial monophosphate (ROPO
CDI.
3
) decomposed upon reaction with
Phosphates 4–6 were converted to bis(triethylammoni-
um) salts and were then treated with five equivalents
of CDI. Imidazolide formation was complete after two
example, reaction of 7 with CDI led to the formation
of three different products as determined by P NMR.
3
1
31
hours, as shown by an upfield shift of the P NMR sig-
nals. Excess CDI was quenched with a stoichiometric
amount of methanol. Subsequently, farnesyl phosphate
Most likely, this mixture resulted from hydrolysis of
the ester by imidazole formed during the reaction. Sim-
ilarly, in the reaction with the indigo dye precursor, 5-
bromo-4-chloro-3-indolyl phosphate 6, the indolyl
nitrogen reacted with CDI, and the methyl carbamate
formed. This was removed by stirring the crude product
in aqueous methanol with K CO at room temperature
2 3
for 1–3 days, though more efficient reaction conditions
could be envisioned.
3
1
3
was added and the reaction was monitored by
P
NMR. After a few hours, two new signals were detected
that marked formation of the diphosphate. After three
to six days, only traces of starting material remained
and no further product formation could be detected.
The crude compounds were submitted to preparative
LC–MS under basic conditions (Table 1). This proce-
dure has worked well with a wide range of starting mate-
rials, ill suited to previously reported conditions.
To improve the synthesis of diphosphates 10 and 11, the
reverse reaction order was explored. Farnesyl phosphate
3 was first converted to the imidazolide and then reacted
with phosphates 4, 6, and 7, respectively. Through this
route, diphosphate 10 was obtained in good yield. The
ester-based phosphate 11 could also be synthesized by
this strategy, though yields remained low.
The presence of CDI and imidazole in the reaction mix-
ture compromises the robustness of the reaction. For
Due to the difficulties encountered in the synthesis of
diphosphates 10 and 11, other coupling methods were
evaluated. The tetrazole-catalyzed coupling of phos-
1
1,12
pho-morpholidates to other phosphates, for exam-
ple, was attempted. Unfortunately, the Moffatt and
Scheme 1.

B. K. Feld, G. A. Weiss / Bioorg. Med. Chem. Lett. 16 (2006) 1665–1667
1667
Khorana procedure for morpholidate synthesis caused
decomposition of 6, and no product could be detected.
However, a simple change of the solvent system from
H O/t-butanol to H O/i-butanol led to quantitative for-
In summary, we present a significantly faster method for
the synthesis of farnesyl monophosphate, which can eas-
ily be extended to the synthesis of other allyl monophos-
phates. We also evaluated and optimized two different
monophosphate coupling methods for the formation
2
2
mation of the desired morpholidate. This new solvent
system led to better yields and cleaner products in the
morpholidate formation for phosphates 4–6 (Table 2).
This improvement is likely due to the fact that i-butanol
and water are only slightly miscible. Therefore, the
phosphate likely remains mainly in the aqueous phase
and the morpholine in the organic phase. This separa-
tion diminishes hydrolysis of the phosphate by the
strongly basic morpholine. Morpholidate formation of
phosphates 3 and 7 could not be achieved using either
method. Therefore, diphosphate 11 remains inaccessible
through the morpholidate method.
1
2
of P -farnesyl-P -substituted diphosphates. Though the
use of morpholidate intermediate compounds can result
in better yields for diphosphates with imidazole- or
CDI-sensitive functionalities, it is less convenient than
the CDI-catalyzed coupling reactions, requiring two
separate reaction steps in place of one. Furthermore,
as demonstrated here, a broader variety of diphosphates
are accessible via CDI coupling, including compounds
inaccessible via the morpholidate route, such as diphos-
phate 11. The reactions presented in this publication
demonstrate that the CDI-based reaction conditions
can be used in the presence of esters, Michael acceptors,
and unprotected amines.
The synthesis of diphosphate 10 demonstrates the use-
fulness of the tetrazole-catalyzed coupling reaction for
allylic diphosphates. After two days reaction time, the
solvent was evaporated, and diphosphate 10 was isolat-
ed via preparative LC–MS in 55% yield (Scheme 2).
Supporting information
Experimental procedures, NMR, and mass spectral data
for compounds 1–3 and 8–14 are available.
Table 2. Synthesis of phosphomorpholidates by improved conditions
Acknowledgments
a
a
We gratefully acknowledge support by the ACS Petro-
leum Research Fund and the Arnold and Mabel Beck-
man Foundation. We also thank Brett Howard for
preliminary experiments and Dr. John Greaves for the
acquisition of high resolution mass spectral data.
R
Yield
O/t-BuOH)
Yield
(H O/i-BuOH)
(
H
2
2
b
6)
c
(
n.r.
99% (12)
84% (13)
e
d
n.a.
Me (4)
References and notes
e
5)
(
39% (14)
100% (14)
1. Cane, D. E. Chem. Rev. 1990, 90, 1089.
2
. Brody, E. P.; Gutsche, C. D. Tetrahedron 1977, 33,
23.
3. Cramer, F.; Rittersdorf, W. Tetrahedron 1967, 23, 3015.
7
f
c
c
C
15
H25 (3)
n.r.
n.r.
4
. Warren, C. D.; Jeanloz, R. W. Biochemistry 1972, 11,
2565.
e
c
c
(
7)
n.r.
n.r.
5
. Kanegasaki, S.; Wright, A.; Warren, C. D. Eur. J.
Biochem. 1983, 133, 77.
a
b
c
d
e
f
0
-Morpholine-N,N -dicyclohexylcarboxamidinium salt of morpholidate.
6. Danilov, L. L.; Maltsev, S. D.; Shibaev, V. N.; Kochet-
kov, N. K. Carbohydr. Res. 1981, 88, 203.
7. Ryu, Y.; Scott, A. I. Org. Lett. 2003, 5, 4713.
8. Davisson, V. J.; Woodside, A. B.; Neal, T. R.; Stremler,
K. E.; Muehlbacher, M.; Poulter, C. D. J. Org. Chem.
1986, 51, 4768.
4
p-Toluidine salt.
No reaction.
Not available.
Free acid.
Ammonium salt.
9
. Davisson et al. report the use of acetonitrile. In the
formation of farnesyl monophosphate, though, we noted
formation of a small amount of bisfarnesyl monophos-
phate. This side reaction was minimized with DMF as the
solvent.
1
1
0. (a) Cramer, F.; Schaller, H.; Staab, H. A. Chem. Ber. 1961,
4, 1612; (b) Schaller, H.; Staab, H. A.; Cramer, F. Chem.
Ber. 1961, 94, 1621.
9
1. Moffatt, J. G.; Khorana, H. G. J. Am. Chem. Soc. 1961,
83, 649.
12. Wittmann, V.; Wong, C. H. J. Org. Chem. 1997, 62, 2144.
Scheme 2.