B. G. McCarthy et al. / Tetrahedron Letters 57 (2016) 502–504
503
this transformation.14,25 Furthermore, before the amine on lysine’s
side chain can be converted into an alcohol, the amine on the back-
bone must be selectively protected, which is a synthetic challenge
in itself.
Despite the variety of methods that have been described for
synthesizing hydroxynorleucine, this compound remains an
expensive building block—for example, Sigma Aldrich sells Boc-
protected hydroxynorleucine for $500 per gram36—which suggests
that there remains a need for new synthetic strategies that can
more efficiently access this highly useful core structure. Recently,
we have developed a method that utilizes N-nitrosodichloroac-
etamides to convert primary amines into alcohols,37 and herein
we show how this method can be applied to transform L-lysine into
6-hydroxynorleucine using a convenient, 1-purification sequence.
Results and discussion
The first challenge of starting a synthesis with lysine (2) is to
efficiently differentiate the two amino groups. After esterifying
lysine’s carboxylic acid group, we found that treating diamine 3
with methyl dichloroacetate allows for preferential functionaliza-
tion of the unhindered amine on the side chain versus the elec-
tron-deficient, branched amine on the backbone (Scheme 1). As a
result, pure dichloroacetamide 4 can be easily isolated on a multi-
gram scale with only a single aqueous workup required as purifica-
tion. Some bis(amide) 5 is produced, but this byproduct can be
washed or filtered away while product 4 remains dissolved in an
acidic aqueous phase. None of the isomer of 4 (in which the back-
bone is functionalized, but the side chain remains unreacted) was
observed, presumably due to a secondary kinetic resolution effect
that would readily transform this compound into bis(amide) 5.
The second challenge of converting lysine into hydroxynor-
leucine is to convert the side chain’s nitrogen-based functional
group into an oxygen-based one. Our recent work37 has shown that
N-nitrosodicholoroacetamides can be converted into dichloroac-
etate esters through a thermal rearrangement under mild condi-
tions (40 °C). The mechanism of this transformation (Scheme 2A)
begins with migration of the acetate group in 6 onto the nitroso
group’s oxygen atom to form diazo acetate 7, which then disasso-
ciates into diazoalkene 8 and carboxylic acid 9 and finally recom-
bines into ester 10 with release of nitrogen gas.
Scheme 2. Amide-to-ester conversion. (A) General mechanism; (B) conversion of
lysine’s side chain. Conditions: (a) DCM (0.1 M), 4-dimethylaminopyridine
(1 equiv), ethyl trifluoroacetate (2 equiv), 30 °C bath, 23 h, 99% yield; (b) 1:1
dichloroacetic acid/trifluoroacetic anhydride (0.3 M), ice bath, sodium nitrite
(2 equiv), 2 h, sodium nitrite (additional 2 equiv), additional 4 h, 79% mass recovery,
use directly for rearrangement; (c) toluene (0.1 M), 40 °C bath, 24 h, 42% yield.
Here we show that the rearrangement of N-nitrosodichloroac-
etamides can be used to install the oxygen-based functionality
required for hydroxynorleucine’s side chain (Scheme 2B). Starting
with amine 4, the backbone amino group can be simply protected
Scheme 3. Endgame. Conditions: (a) MeOH (0.2 M), triethylamine (2 equiv), 2 h,
95% yield; (b) 1:1 THF/water (0.2 M), NaOH (3 equiv), 23 h, use directly for next
step; (c) add sodium bicarbonate (5 equiv), acetone, Fmoc-OSu (1.05 equiv), 24 h,
98% yield; (d) add sodium bicarbonate (4 equiv), di-tert-butyl dicarbonate
(1.5 equiv), THF, sodium carbonate (excess), 22 h, 98% yield.
as a trifluoroacetamide. Although dichloroacetamides and trifluo-
roacetamides are structurally similar, the electron-deficiency of
the trifluoroacetamide in molecule 11 causes that group to be
much less reactive in the subsequent nitrosylation reaction.
Acidified sodium nitrite in the presence of an anhydride cosolvent
enables the nitrosylation of amide 11. A mixture of dichloroacetic
acid and trifluoroacetic anhydride enables full nitrosylation of the
side chain’s dichloroacetamide group (as measured by 1H NMR)
with only minimal (approximately 10%) reaction of the backbone’s
trifluoroacetamide group. Crude N-nitrosoamide 12 is directly dis-
solved in toluene and warmed to 40 °C for 24 h to complete the
rearrangement. Dichloroacetate ester 13 is isolated by silica-gel
chromatography, which is the only chromatographic purification
required in the entire synthetic sequence.
Scheme 1. Differentiation of lysine’s amino groups. Conditions: (a) MeOH (0.9 M),
SOCl2 (1.3 equiv), reflux 17 h, quantitative mass recovery; (b) MeOH (0.3 M),
diisopropylethylamine (3 equiv), ice bath, methyl dichloroacetate (1.5 equiv), to
ambient temperature, 24 h, 51% yield (2 steps).
The final steps to complete the synthesis are efficient and
straightforward (Scheme 3). First, mild conditions—2 equiv of