IspH protein of the deoxyxylulose phosphate pathway: Mechanistic studies with C1-deuterium-labeled substrate and fluorinated analogue

Article abstract of DOI:10.1002/anie.200803452

(Chemical Equation Presented) The last step of the deoxylxylulose phosphate pathway is catalyzed by IspH to synthesize two precursors for isoprenoid biosynthesis. The lack of a primary kinetic isotope effect at C₁ and enzymatic evaluation with a fluorinated analogue (see scheme) suggest that the C₁-position is not involved in the IspH-catalyzed reaction. FldA = flavodoxin, Fpr = flavodoxin reductase, NADPH = nicotinamide adenine dinucleotide phosphate, PPi = P₂O₆^3-.

Full text of DOI:10.1002/anie.200803452

Communications
DOI: 10.1002/anie.200803452
Enzyme Catalysis
IspH Protein of the Deoxyxylulose Phosphate Pathway: Mechanistic
Studies with C1-Deuterium-Labeled Substrate and Fluorinated
Analogue**
Youli Xiao and Pinghua Liu*
Isoprenoids, such as steroids, terpenoids, carotenoids, and
ubiquinones, have important roles in physiological and
pathological processes in all organisms, including electron
transfer, photosynthesis, membrane stability, and cellular
signaling.[1,2] As one of the largest groups of natural products,
all isoprenoids are constructed from two precursors, isopen-
tenyl diphosphate (IPP, 2) and its isomer dimethylallyl
diphosphate (DMAPP, 3; Scheme 1). Two pathways for the
archaebacteria.[3] The unique distribution of these two
biosynthetic pathways among different kingdoms suggests
that mechanistic studies on the enzymes in the DXP pathway
may lead to the development of mechanism-based inhibitors
as antibiotics or herbicides.[4,5] In the MVA pathway, the end
product is 2, which is then isomerized to 3 by isopentenyl
pyrophosphate isomerase.[6–10] In contrast, the IspH enzyme in
the DXP pathway catalyzes the formation of both 2 and 3
(Scheme 1).[11–18] Several models have been proposed for the
IspH-catalyzed reductive dehydration of (E)-4-hydroxy-3-
methyl-2-butenyl diphosphate (HMBPP, 1) to form 2 and 3. In
the study reported herein, a substrate analogue and isotopi-
cally labeled substrate are utilized as probes to examine the
two models in Scheme 1.[19]
Recently, Rohdich et al. suggested that the IspH-cata-
lyzed transformation is a biological counterpart of the Birch
reduction of allylic alcohols with lithium in liquid ammonia
(model A in Scheme 1).[13] In model A, it is believed that the
conformational restriction at the enzyme-active site favors
the C4 hydroxy group instead of the C1-diphosphate as the
leaving group. In addition, the C4 hydroxy group is directly
ligated to the unique iron site of the iron–sulfur cluster to
facilitate the dehydration process. The iron–sulfur cluster is
involved in both the dehydration and reduction steps. In
model A, all hydrogen atoms in 1 are retained in 2 and 3,
which is consistent with results from feeding experiments
using isotopically labeled DXP pathway precursors.[3]
The steady-state kinetic analysis of an IspH substrate
analogue, (E)-4-hydroxy-3-methyl-2-butenyl diphsophonate
(4), has revealed that there is a significant decrease in kcat
(about 26-fold) relative to that of 1 (Scheme 1).[19] This
reduction in kcat could be because 4 is shorter than 1 by one
bridging oxygen atom, which results in nonoptimal electron
transfer or protonation. It might also be possible that the C1-
position of compound 1 is involved in the reaction, as
suggested in model B that follows an aconitase type of
dehydration (Scheme 1).[20] Because all hydrogen atoms are
retained during the reductive dehydration from 1 to 2 and 3,[3]
the same hydrogen atom removed during the deprotonation
step will have to be added back in the subsequent protonation
step in model B. Also, the proton abstracted should not
readily equilibrate with the solvent. The presence of this type
of “sticky proton” has been well documented in the case of
aconitase.[20]
Scheme 1. Two IspH mechanistic models and a substrate analogue.
3À
PPi=P2O6
.
biosynthesis of 2 and 3 have been discovered: the deoxy-
xylulose phosphate (DXP) pathway in green algae, the
chloroplasts of higher plants and most eubacteria, and the
mevalonic acid (MVA) pathway in animals, fungi, and
[*] Dr. Y. Xiao, Prof. Dr. P. Liu
Department of Chemistry, Boston University
Boston, MA 02215 (USA)
Fax: (+1)617-353-6466
À
To test the C H bond cleavage at the C1-position in
E-mail: pinghua@bu.edu
Scheme 1, the deuterated substrate, (E)-4-hydroxy-3-methyl-
2-[1,1-2H2]-butenyl diphosphate ([1,1-2H2]-1), was synthesized
and used to measure the primary kinetic isotope effect (KIE)
[**] This work is supported by the Boston University startup fund.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803452.
9722
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9722 –9725

Angewandte
Chemie
at the C1-position. To synthesize the desired [1,1-2H2]-1, we
modified a reported procedure with LiAlD4 as the reductant
to introduce the two deuterium atoms at the C1-position
(Scheme 2).[21] The deuterium-labeled allylic alcohol 6 was
at the C1-position, was synthesized and evaluated as a
substrate analogue. Because two fluorine atoms are used to
replace the two hydrogen atoms at the C1-position, compound
8 is an inhibitor and no catalytic turnover is expected if the
reaction follows model B. However, according to model A,
IspH will make use of compound 8 as a substrate.
The synthesis of compound 8 is outlined in Scheme 3.
Starting from commercially available methyl methacrylate,
(E)-3-bromo-2-methylacrylate (9) was readily synthesized
Scheme 2. Synthesis of [1,1-2H2]-1. a) LiAlD4, Et2O, reflux, 90%;
b) 1. p-TsCl, DMAP, CH2Cl2, RT, 65%; 2. MeOH, TsOH, RT, 82%;
c) TBAPP, MeCN, RT, 72%. THP=tetrahydropyran; TsCl=toluenesul-
fonyl chloride; DMAP=4-dimethylaminopyridine; TsOH=p-toluenesul-
fonic acid; TBAPP=tris(tetra-n-butylammonium)hydrogen pyrophos-
phate.
Scheme 3. Synthesis of 8. a) CuI, KI, DMF, 1408C, 6 h, 82%;
b) 1. DIBAL-H, CH2Cl2, À788C, 95%; 2. Ac2O, pyridine, DMAP, RT,
94%; c) (EtO)2P(O)CF2ZnBr, CuBr, THF, RT, 84%; d) 7n NH3 in
MeOH, RT, 86%; e) BzCl, Et3N, CH2Cl2, DMAP, RT, 98%; f) TMSBr,
CH3CN, RT, quantitative yield; g) 1. Bu3N, DMF, RT; 2. CDI, DMF, RT;
3. MeOH, DMF, RT; 4. Bu3NH3PO4, DMF, RT, 64%; h) 30% NH3·H2O,
308C, 88%. DIBAL-H=diisobutylaluminum hydride; Bz=benzoyl;
TMSBr=bromotrimethylsilane; CDI=1,1-carbonyldiimidazole.
converted into the allylic chloride 7 with p-TsCl and DMAP
in CH2Cl2 solvent. Removal of the THP group in compound 6
was achieved by using a catalytic amount of TsOH in MeOH
solvent.[22] To simplify the purification process, the diphos-
phate group was introduced as the last step by using TBAPP
in acetonitrile as the pyrophosphorylation reagent. The final
product, [1,1-2H2]-1, was purified by cation-exchange chro-
matography with the ammonium form of Dowex-50 WX8
resin followed by C18 reverse-phase chromatography using a
two-step solvent system, acetonitrile/10% ammonium hy-
droxide/H2O from 10:2.5:0.5 to 6:2.5:0.5, with an overall yield
of 34% (see the Supporting Information).
according to a reported procedure.[23] The synthesis of (E)-3-
iodo-2-methylallyl acetate (10) from 9 involved several steps.
The trans halogenation was achieved by Cu(I)-assisted
halogen exchange to produce
a more reactive vinyl
iodide.[24] After trans halogenation, the ester group was
reduced to produce an alcohol, which was then acetylated
to form 10. The coupling between the iodoalkene 10 and the
organometallic reagent (EtO)2P(O)CF2Cu·ZnBr2 was ach-
ieved in 84% yield by using a Shibuya–Yokomatsu coupling
to form the key intermediate, a difluorinated phosphonate,
(E)-4-(diethylphosphono)-4,4-difluoro-2-methylbut-2-enyl
acetate (11).[25] The acetate protecting group in 11 was
converted into a more stable benzoyl protecting group to
[1,1-2H2]-1 was characterized by steady-state kinetics.[19]
No primary KIE was observed as the same kinetic parameters
were obtained when either 1 or [1,1-2H2]-1 was used as a
substrate (Figure 2S in the Supporting Information). The
deuterium-labeled IPP and DMAPP produced from [1,1-
2H2]-1 were purified by HPLC and characterized by 1H NMR
spectroscopy and mass spectrometry. The 1H NMR spectrum
of isolated IPP is consistent with retention of the two
deuterium atoms at the C1-position during catalysis. The
1H NMR signal for the C2-position hydrogen atoms of IPP
produced from [1,1-2H2]-1 is a singlet (2.21 ppm), which
suggests the retention of the two C1-position deuterium atoms
in IPP and DMAPP (Figure 3S in the Supporting Informa-
tion). The results from high-resolution mass spectrometry
match those of deuterated IPP and DMAPP (Figure 3S in the
Supporting Information).
There are at least two different explanations for the lack
of primary KIE and deuterium washout when [1,1-2H2]-1 was
used as a substrate. All of these results are consistent with
model A (Scheme 1) because of the lack of involvement of
the C1-position. Alternatively, if the step involving the C1-
position in model B (Scheme 1) is not the rate-limiting step,
no primary KIE for [1,1-2H2]-1 will be observed either. To
differentiate between these two options, (E)-4-hydroxy-3-
methyl-2-butenyl-1,1-difluoro diphosphonate (8), fluorinated
generate
(E)-4-(diethylphosphono)-4,4-difluoro-2-methyl-
but-2-enyl benzoate (12) because removal of the ethyl
groups with TMSBr caused partial acetate ester hydrolysis
when 11 was used directly. Once (E)-4-(benzoyloxy)-1,1-
difluoro-3-methylbut-2-enylphosphonic acid (13) was pro-
duced by deprotection, the diphosphonate 8 was obtained
following a coupling procedure reported by us recently in the
synthesis of compound 4, with 30% total yield over nine steps
(see the Supporting Information).[19]
Once diphosphonate 8 was in hand, it was characterized
by steady-state kinetics by using our recently reported
nicotinamide adenine dinucleotide phosphate (NADPH)
consumption assay.[19] IspH can utilize 8 as a substrate.
However, compound 8 is an extremely poor substrate with a
kcat of 0.022 minÀ1, which is more than 500-fold lower than
that of the natural substrate 1.[19] To characterize products
from the turnover of 8, a single-turnover experiment was
Angew. Chem. Int. Ed. 2008, 47, 9722 –9725
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9723

Communications
carried out using IspH (300 mm) with NADPH, flavodoxin
(FldA), and flavodoxin reductase (Fpr) as the reducing
system. When the reaction was monitored by 19F NMR
spectroscopy (Figure 1), in addition to the signals from the
Scheme 4. IspH reaction using 8 as a substrate.
us when diphosphonate 4 was utilized as the substrate, which
produced only the IPP analogue (Figure 1S-b in the Support-
ing Information).[19]
Although compound 8 is a poor substrate, with an activity
about 20-fold lower than that of 4, the utilization of 8 by IspH
as a substrate is mechanistically important. The production of
14 from 8 by IspH suggests that C1-deprotonation/protona-
tion may not be part of the IspH-catalyzed reaction. Neither
the kinetic studies with isotopically labeled substrate nor the
results from use of the fluorinated substrate analogue support
the involvement of the C1-position in IspH-catalyzed reac-
tion.
A radical-mediated reaction finds a precedent in the
radical-mediated dehydration reactions of hydroxylacyl-coen-
zyme A (CoA) to enoyl-CoA.[26,27] In that case, because of the
extremely weak acidity of the b position, a complicated
radical mechanism is invoked. It is believed that the reduction
of a thioester carbonyl group by one electron leads to the
production of a radical anion, which is then used to mediate
the dehydration reaction to produce a ketyl radical. Recently,
the presence of an allylic ketyl radical has been directly
observed.[28] A similar model is proposed in model A
(Scheme 1) for the IspH-catalyzed reaction, in which the
reduction of an alkene instead of thioester is used to form a
radical anion. The radical anion produced is then used to
mediate a dehydration reaction to form an allylic radical
intermediate.
Figure 1. 19F NMR spectra of the substrate (8), the reaction mixture,
and the products. A) Pure 8. B) Control reaction without IspH, in
which compound 8 was not stable as indicated by the formation of
decomposition products. The reaction mixture contained 2.0 mm
NADPH, 60 mm flavodoxin (FldA), 26 mm flavodoxin reductase (Fpr),
and 0.3 mm 8 in 100 mm Tris-HCl, pH 8.0, with a total volume of
1.5 mL at 378C for 3 h. C) Single turnover of IspH reaction with
compound 8 (same conditions as (B) with the addition of 0.3 mm
IspH). D) Synthetic standard 15. E) Synthetic standard 14. By compar-
ing the 19F NMR results in (C–E), it is clear that 14 is the enzymatic
product of 8.
substrate (d = 105.51 ppm, Figure 1A) and the decomposition
products from 8 (d = À104.60, À117.38, À119.87 ppm, Fig-
ure 1B; the decomposition products have not been charac-
terized yet), there is clearly another double triplet (d =
À112.51 ppm), which is produced only when IspH is present
(Figure 1C). This product was purified by HPLC methods
In summary, our studies with a deuterium-labeled sub-
strate and fluorinated analogue suggest no significant influ-
ence of the C1-position in the IspH-catalyzed reaction unless a
different mechanism is employed when the substrate ana-
logue is utilized. Future work will be directed towards
characterization of the proposed allylic radical intermediate.
1
and analyzed by H NMR spectroscopy and mass spectrom-
Received: July 16, 2008
Published online: November 3, 2008
etry.
To guide the structural assignment of the products, two
potential turnover products, 1,1-difluoro-3-methyl-3-butenyl
diphosphonate (14, Figure 1E) and 1,1-difluoro-3-methyl-2-
butenyl diphosphonate (15, Figure 1D), were synthesized
Keywords: biosynthesis · enzymes · isotope effects · kinetics ·
.
reaction mechanisms
1
chemically (see the Supporting Information). The H NMR,
19F NMR, and mass spectrometry results of the isolated
product are all consistent with the production of 14 from 8 as
the IspH turnover product (Figure 1 and Figure 4S in the
Supporting Information). The production of 14 has also been
confirmed by high-resolution ESI mass spectrometry in the
negative mode, which gives a major signal at m/z 264.9851
(calculated [MÀH]À for 14 is m/z 264.9848; Figure 4S-b).
Compound 15 was not detected (Figure 1C and D, Figure 4S
in the Supporting Information). Instead of two products (2
and 3) from 1, there is only one detectable product (14)
produced from 8 (Scheme 4). Similar results were reported by
[1] J. C. Sacchettini, C. D. Poulter, Science 1997, 277, 1788 – 1789.
[2] Comprehensive Natural Products Chemistry: Isoprenoids Includ-
ing Carotenoids and Steroids (Ed.: D. E. Cane), Elsevier Science,
Oxford, 1999.
[3] W. Eisenreich, A. Bacher, D. Arigoni, F. Rohdich, Cell. Mol. Life
Sci. 2004, 61, 1401 – 1426.
[4] M. Rodriguez-Concepcion, Curr. Pharm. Des. 2004, 10, 2391 –
2400.
[5] M. Rohmer, C. Grosdemange-Billiard, M. Seemann, D. Tritsch,
Curr. Opin. Invest. Drugs 2004, 5, 154 – 162.
9724 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9722 –9725

Angewandte
Chemie
[6] B. W. Agranoff, H. Eggerer, U. Henning, F. Lynen, J. Biol. Chem.
1960, 235, 326 – 332.
[7] A. C. Ramos-Valdivia, R. van der Heijden, R. Verpoorte, Nat.
Prod. Rep. 1997, 14, 591 – 603.
[8] K. Kaneda, T. Kuzuyama, M. Takagi, Y. Hayakawa, H. Seto,
Proc. Natl. Acad. Sci. USA 2001, 98, 932 – 937.
[17] T. Grawert, J. Kaiser, F. Zepeck, R. Laupitz, S. Hecht, S.
Amslinger, N. Schramek, E. Schleicher, S. Weber, M. Haslbeck,
J. Buchner, C. Rieder, D. Arigoni, A. Bacher, W. Eisenreich, F.
Rohdich, J. Am. Chem. Soc. 2004, 126, 12847 – 12855.
[18] K. J. Puan, H. Wang, T. Dairi, T. Kuzuyama, C. T. Morita, FEBS
Lett. 2005, 579, 3802 – 3806.
[9] R. Laupitz, S. Hecht, S. Amslinger, F. Zepeck, J. Kaiser, G.
Richter, N. Schramek, S. Steinbacher, R. Huber, D. Arigoni, A.
Bacher, W. Eisenreich, F. Rohdich, Eur. J. Biochem. 2004, 271,
2658 – 2669.
[19] Y. Xiao, Z. K. Zhao, P. Liu, J. Am. Chem. Soc. 2008, 130, 2164 –
2165.
[20] H. Beinert, M. C. Kennedy, C. D. Stout, Chem. Rev. 1996, 96,
2335 – 2374.
[10] T. Kuzuyama, H. Seto, Nat. Prod. Rep. 2003, 20, 171 – 183.
[11] F. Rohdich, A. Bacher, W. Eisenreich, Bioorg. Chem. 2004, 32,
292 – 308.
[12] F. Rohdich, S. Hecht, K. Gartner, P. Adam, C. Krieger, S.
Amslinger, D. Arigoni, A. Bacher, W. Eisenreich, Proc. Natl.
Acad. Sci. USA 2002, 99, 1158 – 1163.
[13] F. Rohdich, F. Zepeck, P. Adam, S. Hecht, J. Kaiser, R. Laupitz,
T. Grawert, S. Amslinger, W. Eisenreich, A. Bacher, D. Arigoni,
Proc. Natl. Acad. Sci. USA 2003, 100, 1586 – 1591.
[14] P. Adam, S. Hecht, W. Eisenreich, J. Kaiser, T. Grawert, D.
Arigoni, A. Bacher, F. Rohdich, Proc. Natl. Acad. Sci. USA 2002,
99, 12108 – 12113.
[15] F. Rohdich, S. Hecht, A. Bacher, W. Eisenreich, Pure Appl.
Chem. 2003, 75, 393 – 405.
[16] M. Wolff, M. Seemann, B. Tse Sum Bui, Y. Frapart, D. Tritsch, A.
Garcia Estrabot, M. Rodriguez-Concepcion, A. Boronat, A.
Marquet, M. Rohmer, FEBS Lett. 2003, 541, 115 – 120.
[21] S. Amslinger, K. Kis, S. Hecht, P. Adam, F. Rohdich, D. Arigoni,
A. Bacher, W. Eisenreich, J. Org. Chem. 2002, 67, 4590 – 4594.
[22] S. Hecht, S. Amslinger, J. Jauch, K. Kis, V. Trentinaglia, P. Adam,
W. Eisenreich, A. Bachera, F. Rohdicha, Tetrahedron Lett. 2002,
43, 8929 – 8933.
[23] S. Werle, T. Fey, J. M. Neudorfl, H. G. Schmalz, Org. Lett. 2007,
9, 3555 – 3558.
[24] H. V. Thulasiram, R. M. Phan, S. B. Rivera, C. D. Poulter, J. Org.
Chem. 2006, 71, 1739 – 1741.
[25] T. Yokomatsu, K. Suemune, T. Murano, S. Shibuya, J. Org.
Chem. 1996, 61, 7207 – 7211.
[26] J. Kim, M. Hetzel, C. D. Boiangiu, W. Buckel, FEMS Microbiol.
Rev. 2004, 28, 455 – 468.
[27] W. Buckel, B. T. Golding, Annu. Rev. Microbiol. 2006, 60, 27 – 49.
[28] J. Kim, D. J. Darley, W. Buckel, A. J. Pierik, Nature 2008, 452,
239 – 242.
Angew. Chem. Int. Ed. 2008, 47, 9722 –9725
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9725