Making thiamin work faster: Acid-promoted separation of carbon dioxide

Article abstract of DOI:10.1021/ja054165p

The conjugate of thiamin and benzoylformate, mandelylthiamin (MTh), undergoes decarboxylation about 10⁶ times slower than the analogous enzymic intermediate. It has now been discovered that the decarboxylation of MTh is accelerated by the acid component of pyridine and 4-picoline buffers. There is no role for a proton donor to stabilize the transition state for decarboxylation: catalysis must be achieved by the acid's trapping the product carbanion, preventing recarboxylation. This requires that diffusion of CO₂ is rate-determining, and that protonation of the carbanion allows this to occur. This interpretation correctly predicts that the same acid components will prevent a fragmentation reaction by protonating the intermediate, which fragments only as the conjugate base. Copyright

Full text of DOI:10.1021/ja054165p

Published on Web 08/13/2005
Making Thiamin Work Faster: Acid-Promoted Separation of Carbon Dioxide
Qingyan Hu and Ronald Kluger*
DaVenport Chemical Laboratories, Department of Chemistry, UniVersity of Toronto, Toronto, Canada M5S 3H6
Received June 23, 2005; E-mail: rkluger@chem.utoronto.ca
Almost 50 years ago, Breslow deduced that thiamin diphosphate
(ThDP)-dependent 2-ketoacid decarboxylases form C-2 thiazolium
ThDP conjugates with substrates to yield resonance-stabilized
carbanions upon loss of CO2.1,2 The route has been experimentally
confirmed,3 but the intermediates themselves are not sufficiently
reactive to account for the enzymic process.4-7
Mandelylthiamin (MTh) is the conjugate of thiamin and ben-
zoylformate,8 an analogue of the conjugate in benzoylformate
decarboxylase (BFD).9 Decarboxylation of MTh generates the C2R
conjugate base of 2-(1-hydroxybenzyl)thiamin (HBnTh) with a first-
order rate constant that is about 106 times smaller than kcat for BFD.8
Thus, delocalization of the carbanion does not account for the
catalytic power of the enzyme. Furthermore, while BFD efficiently
releases benzaldehyde and regenerates ThDP, the loss of CO2 from
Figure 1. MTh decarboxylation in acetate (O), bis-tris (b), phosphate (0),
pyridine (2), and 4-picoline ([) buffers, pH ) pKa of each buffer (acetate
MTh leads to a fragmentation reaction that destroys thiamin (in
competition with protonation to form HBnTh).10-15 The processes
are summarized in Scheme 1. Since the rate constant for nonen-
zymatic fragmentation is considerably larger than kcat for BFD,12
the enzyme’s mechanism for accelerating decarboxylation must
coincidentally suppress fragmentation.
at pH 4.6, bis-tris at pH 6.7, phosphate at pH 6.5, 4-picoline at pH 6.0, and
pyridine at pH 5.4). In the case of ethanol (]), 0.1 M phosphate at pH 6.5
was used as buffer, and ethanol was added as a solvent component. In the
case of ethyl pyridinium (9), 0.1 M phosphate at pH 6.5 was used as buffer,
and ethyl pyridinium chloride was added as a component for maintenance
of ionic strength.
Scheme 1. Fragmentation Competes with Protonation after Loss
of Carbon Dioxide
Figure 2. MTh decarboxylation rate constant in pyridine buffer versus
[pyridinium ion]/[total pyridine]. The total pyridine concentration varied
from 0.1 to 0.8 M (O for 0.1 M, b for 0.2 M, 0 for 0.4 M, 9 for 0.6 M,
and 4 for 0.8 M). [A] refers to the acid component, pyridinium ion, and
[B] refers to the base component, pyridine.
In the course of examining the reactivity patterns of MTh, we
discovered that its decarboxylation is accelerated by certain buffers,
although there is no apparent way that the C-C bond breaking
might be stabilized by an acid or base. The data in Figure 1 show
that the decarboxylation of MTh is accelerated by buffers containing
pyridine or 4-picoline. We used solutions at pH ) pKa (pH 4.6-
6.5), where the rate of decarboxylation of MTh is pH-independent
(kobs ) 3.2 × 10-4 s-1).
The derived second-order rate constant for catalysis by pyridine is
0.001 M-1 s-1, and that for 4-picoline is 8 × 10-4 M-1 s-1. A
very small amount of complexation would give these apparent
values, despite a much larger actual value in the complex.
We tested whether the acceleration from pyridine and 4-picoline
arises from their effect on solvent polarity, as desolvation should
destabilize MTh.6,7 We used ethanol as an additive with a dielectric
constant similar to that of pyridine. The addition of ethanol (0.1-
0.8 M) to 0.1 M phosphate buffer does not change significantly
the observed rate coefficient for decarboxylation. N-ethyl pyridinium
chloride was used to assess the magnitude of the nonspecific
medium effect of pyridinium ions, and this is also minimal.
A plot of the observed rate constant for decarboxylation of MTh
as a function of the mole fraction of the acid component indicates
that this component is catalytic (Figure 2). We assume that curvature
is the result of saturation in formation of the pyridinium complex.
The thiazolium ring of MTh has no site where a proton donor
might reasonably associate to promote the loss of CO2. However,
if the back reaction is prevented by protonation prior to complete
separation of CO2, the overall reaction will be faster; diffusion of
CO2 is rate-limiting.16 (Guthrie’s analysis of decarboxylation by
the no barrier theory provides a very clear picture of the detailed
processes to consider, including changes in hybridization.17) Ad-
dition of CO2 to the HBnTh carbanion must compete with separation
of these species. This is reminiscent of the mechanism (in reverse)
proposed for biotin-enzyme, ATP-driven carboxylation reactions.18
Where “low entropy CO2” is generated in situ, an adjacent
nucleophile will react most efficiently. The mechanistic issues of
9
12242
J. AM. CHEM. SOC. 2005, 127, 12242-12243
10.1021/ja054165p CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
The general role of thiamin in promoting decarboxylation has
often been summarized by showing the thiazolium serving as a
“sink” for the electron density generated in conjunction with
formation of CO2 in a concerted transition state. However, our
results show that the transition state for breaking the C-C bond
need not involve such delocalization to be concerted with formation
of CO2.
The many aspects of ThDP as a catalyst should be emphasized.
Its ability to ionize to the C2 ylide, which can readily add to a
substrate’s carbonyl group, is central in positioning the substrate
for reaction, providing the benefits of forming a covalent intermedi-
ate.24 The adjacent positive charge of the thiazolium nitrogen
provides very significant electrostatic stabilization of anions and
of anionic transition states.25 After a bound substrate is converted
to a protein-associated ThDP conjugate in a decarboxylase,7 it can
be oriented to be destabilized by electrostatic stress26 and desol-
vation,6,7 while the bond breaking can be facilitated by protonation
of the product carbanion to facilitate the departure of CO2, a process
that our results suggest can be rate-determining. Finally, the need
for the quick arrival of a proton in the proper location can be
facilitated by the proton tunnel found in related enzymes.27
Figure 3. Increased yield of HBnTh by increasing acid concentration of
the buffer components. The data were collected at pH ) pKa (O bis-tris
pH 6.7, b pyridine pH 5.4, 0 (2,2,2)-trifluoroethylamine pH 5.7, 9
phosphate pH 6.5, and 4 acetate pH 4.6).
separation and recombination are also similar to those in stepwise
nucleophilic substitution, in which ion pair separation is kinetically
significant.19-21 Thus, a mechanism that accounts for our results
involves the acid catalyst as a spectator to the carbon-carbon bond-
breaking event22 (Scheme 2).
Scheme 2. A Brønsted Acid Intercepts the Carbanion in
Competition with Nascent CO2: Rehybridization17 and Partial
Separation are Schematically Incorporated into k2
Acknowledgment. We thank NSERC Canada for support, and
Peter Guthrie (University of Western Ontario) for critical insights.
References
(1) Breslow, R. Chem. Ind. (London) 1957, 893-894.
(2) Breslow, R. J. Am. Chem. Soc. 1957, 79, 1762-1763.
(3) Tittmann, K.; Golbik, R.; Uhlemann, K.; Khailova, L.; Schneider, G.;
Patel, M.; Jordan, F.; Chipman, D. M.; Duggleby, R. G.; Huebner, G.
Biochemistry 2003, 42, 7885-7891.
(4) Kluger, R. Chem. ReV. 1987, 87, 863-876.
(5) Alvarez, F. J.; Ermer, J.; Huebner, G.; Schellenberger, A.; Schowen, R.
L. J. Am. Chem. Soc. 1991, 113, 8402-8409.
(6) Crosby, J.; Stone, R.; Lienhard, G. E. J. Am. Chem. Soc. 1970, 92, 2891-
2900.
(7) Zhang, S.; Liu, M.; Yan, Y.; Zhang, Z.; Jordan, F. J. Biol. Chem. 2004,
279, 54312-54318.
(8) Hu, Q.; Kluger, R. J. Am. Chem. Soc. 2002, 124, 14858-14859.
(9) Polovnikova, E. S.; McLeish, M. J.; Sergienko, E. A.; Burgner, J. T.;
Anderson, N. L.; Bera, A. K.; Jordan, F.; Kenyon, G. L.; Hasson, M. S.
Biochemistry 2003, 42, 1820-1830.
(10) Hu, Q.; Kluger, R. J. Am. Chem. Soc. 2004, 126, 68-69.
(11) Kluger, R.; Lam, J. F.; Pezacki, J. P.; Yang, C.-M. J. Am. Chem. Soc.
1995, 117, 11383-11389.
This mechanism predicts that the catalytic Brønsted acids will
be uniquely effective in blocking fragmentation.10 They can transfer
a proton prior to dissociation of CO2 and also after diffusion has
occurred. The yields of fragmented products versus formation of
HBnTh are summarized in Figure 3. As predicted, the conjugate
acid of pyridine is especially effective at blocking fragmentation,
requiring a considerably lower concentration to have a maximal
effect.
(12) Kluger, R.; Moore, I. F. J. Am. Chem. Soc. 2000, 122, 6145-6150.
(13) Moore, I. F.; Kluger, R. Org. Lett. 2000, 2, 2035-2036.
(14) Moore, I. F.; Kluger, R. J. Am. Chem. Soc. 2002, 124, 1669-1673.
(15) Oka, Y.; Yoshioka, K.; Hirano, H. Chem. Pharm. Bull. 1967, 15, 119-
122.
(16) Longmuir, I. S.; Forster, R. E.; Woo, C. Y. Nature 1966, 209, 393-394.
(17) Guthrie, J. P. Bioorg. Chem. 2002, 30, 32-52.
(18) Sauers, C. K.; Jencks, W. P.; Groh, S. J. Am. Chem. Soc. 1975, 97, 5546-
5553.
On the basis of these results, we expect that enzymes can promote
decarboxylation of a ThDP conjugate through early protonation of
the carbanion, preventing recombination with nascent CO2. Gao’s
calculations of OMP decarboxylase conclude that separation can
be the major rate-liming process.23 Since recombination should have
a lower barrier than diffusion, this could account for the additional
acceleration in BFD compared to MTh and BFD’s avoidance of
fragmentation. A conformation of the conjugate base that would
lead to overlap with the thiazolium (along with rehybridization17)
would be less efficient as it would increase the barrier to protonation
on an enzyme.
(19) De la Mare, P. B. D.; Vernon, C. A. J. Chem. Soc. 1954, 2505-2510.
(20) Winstein, S.; Schreiber, K. C. J. Am. Chem. Soc. 1952, 74, 2165-2170.
(21) Goering, H. L.; Thies, R. W. J. Am. Chem. Soc. 1968, 90, 2967-2968.
(22) Venkatasubban, K. S.; Schowen, R. L. J. Org. Chem. 1984, 49, 653-
655.
(23) Gao, J.; Byun, K. L.; Kluger, R. Top. Curr. Chem. 2004, 238, 113-136.
(24) Zhang, X. Y.; Houk, K. N. Acc. Chem. Res. 2005, 38, 379-385.
(25) Halikdes, C. J.; Frey, P. A.; Tobin, J. B. J. Am. Chem. Soc. 1993, 115,
3332-3333.
(26) Wu, N.; Mo, Y.; Gao, J.; Pai, E. F. Proc. Natl. Acad. Sci. U.S.A. 2000,
97, 2017-2022.
(27) Frank, R. A. W.; Titman, C. M.; Ventkatesh, P. J.; Luisi, B. F.; Perham,
R. N. Science 2004, 306, 872-876.
JA054165P
9
J. AM. CHEM. SOC. VOL. 127, NO. 35, 2005 12243