Synthesis and Catalytic Evaluation of Novel Mimics of Thiamine Pyrophosphate

Abstract

The triazolium mimic of thiamine was successfully prepared and isolated using a novel synthetic route, as this compound has not been previously reported in the literature to the best of our knowledge. The route employs inexpensive and commercially available starting material with all reactions carried out in water or ethanol. Thiamine was used to sacrificially to access the intricate 4-aminopyrimidinyl side substituent. In addition to the prime triazolium mimic of thiamine, three other analogues were also prepared starting with different esters and a lactone.
For these, the C3 carbon acid pKa values of the mimics were determined using a kinetic H/D exchange method. The pseudo first order rate constants kex (s−1) for exchange were estimated using 1H NMR spectroscopy. Second-order rate constants for deuteroxide ion catalysed exchange (kDO, M−1 s−1) were obtained from the slope of the plot of kex against deuteroxide concentration, which are referred to as kinetic acidities.
Experimental evidence was used to establish the absence of general base catalysis by analysis of H/D exchange in a range of formic acid buffer concentrations at the same buffer ratio. This informed the decision to remove the term for contributions to exchange from buffer catalysis and hence values for kDO may be calculated.
The triazolium based mimics 144, 183-185 show a fifteen-fold increase in acidity towards deuteroxide ion compared to the corresponding native thiamine. This is due to the presence of two extra ring nitrogen atoms which favours the stability of the carbene/ylide. The range of mimics do not show a significant variation in acidity (0.1 unit of pKa) mostly due to the 5-substituent being relatively far removed from the carbenic carbon atom. Aside from the 5-substituent on the triazolium ring, the mimics were structurally identical otherwise.
Interestingly, the simplest di-methyl triazolium salt 212 studied showed a much lower acidity to the triazolium mimics of thiamine, which suggest the N2-methylpyrimidinyl substituent to be electron-withdrawing, hence the observed lower pKa for all the direct triazolium mimics. The N-phenyl triazolium salt previously studied in the group showed similar kinetic acidities compared to the novel mimics which suggest similar effects between the N2-phenyl and the N2CH2-pyrimidinyl substituent. This is an important observation as it shows the triazolium mimics have closely similar acidity to one of the most widely used and versatile triazolium organocatalysts.
The reactions of the triazolium mimics with a range of typical aldehydes used in NHC-transformations were then explored (Scheme 7.2). Analysis of the 1H NMR spectra obtained from the catalytic evaluation of the triazolium mimics show the presence of the hydroxyaryl adduct in the reaction mixture in all cases. Values of k1, k−1 and K for reactions between novel triazolium mimics 144 and 183 were determined by analysis of changes in species concentration on approach to equilibrium. In addition, using global fitting software, values of forward, reverse and equilibrium constants were estimated from fitting reaction data. The 2-hydroxyethyl triazolium mimic 144 consistently gave larger forward rate constants across the board compared to thiamine 7. In addition, the 2-subtituent effect previously observed and reported by our group was found to hold true for the new mimics prepared in this work especially with ortho substrate 213.
In silico binding studies of TPP-dependent enzymes using crystal structures from the protein data bank have been reported using docking software. The results demonstrate the potential for the new diphosphate mimic to hydrogen bond using its N1 atom at the active site of TPP dependent enzymes. Triazolium and thiazolium organocatalysts show different chemoselectivities in a range of transformations and triazoliums demonstrate a broader reaction scope. It is hoped that the diverse chemistry of triazolium organocatalysts can be harnessed at TPP-dependent enzyme active sites.