T. Sugimura et al. / Tetrahedron Letters 55 (2014) 6343–6346
6345
Scheme 3. Total synthesis of isodesmosine 1.
by sodium hydroxide (NaOH) and hydrogen peroxide (H2O2) then
afforded the anti-Markovnikov product 9 in 80% yield. Treatment
of the primary alcohol 9 with Dess–Martin periodinane (DMP) in
CH2Cl2 led to the desired compound 10 in 98% yield. It is notewor-
thy that the oxidation of the mono-Boc-protected version of 9 gave
the cyclized product.15 Thus, the aldehyde 10, which is a precursor
of the Chichibabin pyridine synthesis for the preparation of 1, was
prepared from 6 in an overall yield of 47% over four steps.
With the protected lysine 5 and aldehyde 10 in hand, we turned
our attention to the Pr(OTf)3-promoted Chichibabin pyridine syn-
thesis. Among the family of Ln(OTf)3 reagents,16,17 Pr(OTf)3 was
found to be the best reagent for the Chichibabin pyridine synthesis,
most likely because it possesses the appropriate level of Lewis
acidity for the Mannich and aldol condensation reactions that
occur.11 Given that the starting materials for the Chichibabin pyr-
idine synthesis were expected to be poorly soluble in H2O, the
effect of using MeOH as a co-solvent was investigated (Table 1).
Reactions between 1 equiv of 5 and 4 equiv of 10 were conducted
in the presence of 50 mol % Pr(OTf)3 at room temperature for 24 h.
When the reaction was performed in 100% H2O, it failed to afford
any of the desired products (Table 1, entry 1), likely due to the fact
that substrates 5 and 10 were insoluble in H2O. Different amounts
of MeOH were then added to the reaction (H2O/MeOH = 1:1, 1:2,
and 1:5 (v/v)) in an attempt to increase the solubility of the sub-
strates (Table 1, entries 2–4, respectively). However, only trace
amounts of the desired pyridinium compound 11 were observed
in these reactions, along with several byproducts, which had the
same m/z as 11 as determined by mass spectrometry (MS) analysis.
Pleasingly, when the reaction was run in a 2:1 (v/v) mixture of H2O
and MeOH, the isodesmosine-type Chichibabin product 11 was
obtained in 29% yield (Table 1, entry 5), along with several uniden-
tified byproducts, which were observed by thin layer chromatogra-
phy (TLC). Interestingly, and in contrast to the previous report,11
only a trace amount of the desmosine-type pyridinium compound
was identified in this particular case. Further increasing the ratio of
H2O to MeOH to 3:1 (v/v) resulted in a slight decrease in the yield
of 11 to 26% (Table 1, entry 6). Taken together, these results indi-
cated that the use of a 2:1 (v/v) mixture of H2O and MeOH (Table 1,
entry 5) was optimal for the Chichibabin pyridine synthesis in H2O/
MeOH and suggested that the solubility of starting materials 5 and
10 was a crucial factor for the success of this synthesis. These
conditions also promoted the selective synthesis of the isodesmo-
sine-type product.
completed in 27% yield (under the conditions of entry 5 in Table 1)
over five steps starting from commercially available 6-(benzyloxy-
carbony)-amino-2-(S)-4-[(tert-butoxycarbonyl)amino]-hexanoic
acid (3). The results obtained in the current study suggest that the
solubility of compounds 5 and 10 in the H2O/MeOH solvent system
was critical to the success of this Chichibabin pyridine synthesis.
Furthermore, this reaction system favored the selective formation
of the isodesmosine-type Chichibabin product, with only a trace
amount of the desmosine-type Chichibabin product being
detected. This selectivity implies that the formation of 1,3,4,5-tet-
rasubstituted pyridine was prevented in the presence of MeOH.
Further studies designed to elucidate the mechanisms involved
in the formation of the elastin crosslinkers isodesmosine 1 and des-
mosine 2 from lysine in nature are required in order to develop a
deeper understanding of these processes, and these studies are
currently underway in our laboratory. The preparation of isotopi-
cally labeled desmosines as internal standards for the analysis of
clinical COPD samples by the isotope-dilution LC–MS/MS9d is also
underway in our laboratory using the synthetic route described
above. This route is also being investigated for the preparation of
related crosslinking amino acids for the elucidation of the three-
dimensional crosslinking structure of elastin.18
Acknowledgments
We would like to thank Dr. Yong Y. Lin (Roosevelt Hospital Cen-
ter, New York) for many fruitful discussions. This work was sup-
ported by a Grant-in-Aid for Young Scientists (B) from the Japan
Society for the Promotion of Science (KAKENHI Grants 22710224
and 25750388).
Supplementary data
Supplementary data (experimental procedures and character-
ization data) associated with this article can be found, in the online
References and notes
The protecting groups in compound 11 were then removed as
follows. The three benzyl groups were removed by hydrogenation
over Pd/C and the seven tert-butoxycarbonyl groups and one tert-
butyl group were removed by trifluoroacetic acid (TFA) in a step-
wise manner. These deprotection steps proceeded quantitatively
to provide the desired pyridinium amino acid isodesmosine 1
(Scheme 3). Spectroscopic analysis of this material, including 1H
NMR, MS, and optical rotation, afforded data that were in good
agreement with those of natural 1.
In summary, we have developed a new method for the total
synthesis of the COPD biomarker isodesmosine 1 via the Pr(OTf)3-
promoted Chichibabin pyridine synthesis of the protected lysine 5
and aldehyde 10 in H2O and MeOH. This new synthesis of 1 was