α-Azido Esters in Depsipeptide Synthesis: C-O Bond Cleavage during Azido Group Reduction

Article abstract of DOI:10.1021/acs.joc.6b02309

α-Azido esters, when treated with dithiothreitol (DTT)/diisopropylethylamine (DIPEA), undergo both azido group reduction to give α-amino esters and C-O bond cleavage to give triazoles. The extent of triazole formation depends upon leaving group ability. Some C-O bond cleavage via triazole formation was also found to occur when a resin-bound peptide, which contained a terminal α-azido ester group, was treated with DTT/DIPEA. C-O bond cleavage also took place when this peptide was treated with PPh₃, PBu₃, or PMe₃; however, in these cases, C-O bond cleavage occurred via either triazole formation and/or hydrolysis of the ester bond in the iminophosphorane intermediate to give betaines. The mechanism that dominated for C-O bond cleavage depended upon the phosphine that was used for azido group reduction. C-O bond cleavage during reduction of the azido group in the peptide was minimized by performing the reduction with PBu₃ in the presence of a symmetric anhydride derived from an amino acid in dry THF followed by the addition of water. Surprisingly, these conditions provided the amine as the major product, while the expected amide was formed as a minor product. These conditions were employed in an improved synthesis of an analogue of the cyclic lipodepsipeptide antibiotic daptomycin.

Full text of DOI:10.1021/acs.joc.6b02309

Article
pubs.acs.org/joc
α‑Azido Esters in Depsipeptide Synthesis: C−O Bond Cleavage
during Azido Group Reduction
Chuda Raj Lohani, Jacob Soley, Braden Kralt, Michael Palmer, and Scott D. Taylor*
Department of Chemistry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada
*
S Supporting Information
ABSTRACT: α-Azido esters, when treated with dithiothreitol (DTT)/diisopropylethylamine (DIPEA), undergo both azido
group reduction to give α-amino esters and C−O bond cleavage to give triazoles. The extent of triazole formation depends upon
leaving group ability. Some C−O bond cleavage via triazole formation was also found to occur when a resin-bound peptide,
which contained a terminal α-azido ester group, was treated with DTT/DIPEA. C−O bond cleavage also took place when this
peptide was treated with PPh , PBu , or PMe ; however, in these cases, C−O bond cleavage occurred via either triazole
3
3
3
formation and/or hydrolysis of the ester bond in the iminophosphorane intermediate to give betaines. The mechanism that
dominated for C−O bond cleavage depended upon the phosphine that was used for azido group reduction. C−O bond cleavage
during reduction of the azido group in the peptide was minimized by performing the reduction with PBu in the presence of a
3
symmetric anhydride derived from an amino acid in dry THF followed by the addition of water. Surprisingly, these conditions
provided the amine as the major product, while the expected amide was formed as a minor product. These conditions were
employed in an improved synthesis of an analogue of the cyclic lipodepsipeptide antibiotic daptomycin.
INTRODUCTION
depsipeptide antibiotic daptomycin, as outlined in Scheme 1,
which shows the synthesis of the daptomycin analogue, Dap-
■
The synthesis of complex or challenging peptides often requires
the use of α-amino protecting groups beyond the traditional
Fmoc or Boc groups that are commonly employed in peptide
synthesis. One alternative α-amino protecting group with a
number of attractive features is the azido group. First, most
peptides bearing α-azido groups are stable to the basic
1
,2
E12-W13. Route A, from peptide 1, was developed first and
uses three azido acids to arrive at peptide 3. After each azido
acid was incorporated into the peptide, the azido group was
reduced to the amine using PMe in dioxane/H O. However, as
3
2
most α-azido acids are not commercially available, it was
desirable to reduce the number of azido acids used in the
synthesis. Hence, route B was developed which employs just
one azido acid that is incorporated into the peptide as an α-
azido ester (peptides 2 and 4). In route B, we chose to reduce
the azido group in peptide 4 with DTT/DIPEA, which are less
1
conditions that are used to remove Fmoc groups and to the
acidic the conditions that are used to remove acid-labile groups
such as Boc-protected amines and t-butyl esters. Hence, they
provide an additional orthogonal dimension which is useful for
the synthesis of complex targets. Second, they minimize side
reactions, such as diketopiperazine formation, that can occur
2
,3
expensive and have longer shelf lives than PMe . Route A
3
4
required the use of PMe because DTT/DIPEA were found to
3
using standard Fmoc chemistry. Third, they are useful for
1
5
,6
cause some removal of the Fmoc group.
hindered couplings. Fourth, in comparison to Fmoc or Boc
groups, the azido group is atom-economical, and α-azido acids
are readily prepared from α-amino acids using diazo transfer
reagents and often do not require chromatographic purifica-
During our studies on the utility of α-azido acids in peptide
synthesis, we recently discovered a previously unnoticed variant
of the Staudinger reaction on α-azido esters in which
phosphazides, generated from the reaction of trialkyl
phosphines (PMe , PBu , or POct ) with phenyl esters of α-
4
,7−10
tion.
Finally, the azido group is reduced to the
corresponding amino group usually in good to excellent yield
under very mild conditions using phosphines in aqueous
mixtures of organic solvents (Staudinger reaction) or with
3
3
3
azido acids in THF/H O, undergo an intramolecular cyclization
2
to provide 5-substituted 2H-1,2,3-triazol-4-ols 7 in good to
4
,5,11−13
dithiols in the presence of a mild organic base.
We have recently reported employing α-azido acids in two
Received: September 20, 2016
different solid phase syntheses of analogues of the cyclic
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 1. Two Different Approaches to the Synthesis of Daptomycin Analogues
14
excellent yields (Scheme 2). Triazole formation is most likely
performing the reduction with PBu in the presence of a
3
due to cyclization of the phosphazide 5 to give intermediate 6
symmetric anhydride derived from FmocGlu(OtBu)OH in dry
THF followed by the addition of water. Surprisingly, these
conditions provided amine 3 as the major product, while the
expected amide was formed as a minor product. These
conditions were employed in an improved synthesis of Dap-
E12-W13.
Scheme 2. Synthesis of Triazoles from Phenyl Esters of α-
Azido Acids Using Trialkyl Phosphines
RESULTS AND DISCUSSION
■
We first examined whether triazole formation occurs when α-
azido esters are treated with DTT/DIPEA by subjecting model
esters 10 and 11 to 4 equiv of DTT and 2 equiv of DIPEA in
DMF for 2 h and analyzing the crude reaction mixture by
HPLC (Scheme 3). Ethyl ester 10 gave a mixture of amine 12
Scheme 3. Products Formed upon Treatment of Esters 10
and 11 with DTT/DIPEA
(
triazole pathway) before it can lose nitrogen and give the
iminophosphorane 8 (Staudinger pathway). Hydrolysis of 6
followed by protonation and tautomerization gives the triazoles.
Ethyl esters gave an approximately 1:1 mixture of the triazole
and Staudinger product, suggesting that the rate-determining
step for triazole formation is loss of the leaving group.
Surprisingly, PPh gave only the Staudinger products and no
triazoles.
Here, we report that C−O bond cleavage via triazole
formation occurs when α-azido esters, including peptide 4, are
treated with DTT/DIPEA. We also report that C−O bond
cleavage occurs when peptide 4 is treated with phosphines
and triazole 13, and phenyl ester 11 gave, almost exclusively,
triazole 13 which was isolated in an 83% yield. This suggests
that the rate-determining step for triazole formation is loss of
the leaving group. These product distribution patterns are very
similar to those we found when these esters were subjected to
3
14
alkyl phosphines.
including PPh ; however, C−O bond cleavage can occur by two
The mechanism for the reduction of an azido group with a
dithiol has not been studied in detail. Cartwright et al. proposed
two possible mechanisms, one of which involves attack of the
thiolate on the terminal (γ) nitrogen of the azido group to give
3
different mechanisms, namely, via triazole formation or
hydrolysis of the ester bond in the iminophosphorane
intermediate. What mechanism dominates for C−O bond
cleavage depends upon the phosphine (PMe , PBu , or PPh )
11,15
intermediate 14 (Scheme 4).
Disulfide formation followed
3
3
3
used for azido group reduction. C−O bond cleavage during
reduction of the azido group in peptide 4 was minimized by
by loss of nitrogen from intermediate 15 gives the amine 12.
Triazole formation may occur by cyclization of intermediate 16
B
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 4. Possible Mechanism for the Formation of
Triazole 13 from α-Azido Esters Using Dithiols under Basic
Conditions
4, as our previous studies with α-azido phenyl and ethyl esters
revealed that this reagent gives the Staudinger products and no
14,17
triazole.
Thus, peptide 4 was subjected to 6 equiv of PPh3
in THF/H O for 15 h and then filtered (filtrate #1, Scheme 6).
2
The resin was then treated with TFA/thioanisole/TIS/H O
2
and filtered again (filtrate #2, Scheme 6). The filtrates from
both steps were subjected to HPLC, and the resulting products
were analyzed by MS. Surprisingly, filtrate #2 contained
peptides 19 and 20 in an approximately 1:1 ratio. However,
the filtrate from the first step did not contain triazole 18.
Instead, HRMS analysis indicated that filtrate #1 contained a
compound whose structure which we initially assigned to be
betaine 21 (Scheme 6). To confirm the structure of 21, we
prepared this compound by reacting PPh with N Trp(Boc)-
3
3
18
OH in dry ether. This resulted in the formation of a
precipitate which upon filtration yielded a white solid. When
subjected to HPLC, the precipitate eluted in a single peak that
had a retention time identical to that of compound 21. The
HRMS of this solid was also consistent with this compound
31
being betaine 21. However, the P NMR spectrum of this solid
in THF exhibited a peak at −39.7 ppm, indicating that, in this
solvent, it existed almost exclusively as the cyclic trigonal
bipyramidal phosphorane 24 (Scheme 7) and not the betaine,
to give triazolone 17 which, after disulfide formation,
tautomerizes to give triazole 13.
The above results with ethyl ester 10 prompted us to
examine whether triazole formation also occurs when peptide 4
is subjected to DTT/DIPEA. Thus, peptide 4 was subjected to
DTT/DIPEA in DMF for 2 h and then filtered (filtrate #1,
Scheme 5). The resin was then treated with TFA/thioanisole/
16
31
which would be expected to have a P NMR chemical shift
18−20
31
greater than 20 ppm.
In contrast, the P NMR spectrum
in THF/10% H O exhibited a large broad peak at 2 ppm,
2
suggesting that compound 24 is in equilibrium with betaine 21
in this solvent.
TIS/H O and filtered (filtrate #2, Scheme 5). The HPLC
2
chromatogram from the first filtrate exhibited one major peak,
which MS analysis revealed to be triazole 18. The HPLC
chromatogram from the second filtrate exhibited two major
peaks of almost equal intensity, which MS analysis revealed to
be peptide 19, resulting from C−O bond cleavage, and the
amine peptide 20. These findings indicate that some C−O
bond cleavage does indeed occur during reduction of this azido
group with DTT/DIPEA. This result most likely accounts for
the lower yield that we obtained for Dap-E12-W13 using route
B (2.3%, Scheme 1) in comparison to route A (3.7%).
The above studies with PPh
not cyclize but instead loses N
3
reveal that phosphazide 25 does
to give iminophosphorane 28
2
(Scheme 8). However, the C−O bond in iminophosphorane 28
is susceptible to hydrolysis, and ester hydrolysis competes
effectively with hydrolysis of the phosphorane moiety (Scheme
8).
Our previous studies on ethyl ester 10 suggested that C−O
bond cleavage via triazole formation would occur if peptide 4
was subjected to PMe or PBu in THF/H O. However, our
3
3
2
results with peptide 4 and PPh indicate that C−O bond
3
In order to increase the yield of dap analogues via route B
Scheme 1), it was necessary to minimize C−O bond cleavage
during reduction of the azido group in peptide 4. Therefore, we
cleavage can also occur via hydrolysis of the ester bond in the
iminophosphorane intermediate. To determine if this is also the
case with PMe and PBu , peptide 4 was subjected to 3 equiv of
(
3
3
decided to use PPh in THF/H O to reduce the azido group in
PMe or PBu in THF/H O (1 h for PMe and 16 h for PBu ),
3
2
3
3
2
3
3
Scheme 5. Products Formed upon Treatment of Peptide 4 with DTT/DIPEA
C
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 6. Products Formed upon Treatment of Peptide 4 with Phosphines in THF/H O
2
Scheme 7. Products Formed upon Reaction of
N Trp(Boc)OH with PPh
reduction, and triazole and betaine products) are so different
between peptide 4 and ester 10 and peptide 31 when treated
3
3
with phosphines in THF/H O. The greater susceptibility of the
2
ester moiety in iminophosphoranes 28−30 than of the
analogous ester in the iminophosphorane derived from ethyl
ester 10 and peptide 31 may result from iminophosphoranes
28−30 adopting a conformation that allows a nearby proton to
act as a general acid and promote ester hydrolysis. It is also
possible that iminophosphoranes 28−30 may adopt a
conformation that results in the iminophosphorane abstracting
a proton from the amide nitrogen of the adjacent glycine, which
could promote ester cleavage through intramolecular attack of
the amide oxygen. Phosphazides 25 and 27 may also adopt a
conformation that prevents these phosphazides from cyclizing,
and hence triazoles are not formed when peptide 4 is treated
with PBu or PPh . Phosphazide 26 may adopt a similar
conformation; however, the smaller size of the trimethylphos-
phine moiety may still allow for cyclization to occur.
The ester moiety in phosphoranes 28−30 (Scheme 8)
appears to be fairly susceptible to hydrolysis in comparison to
the ester moiety in peptide 4, which does not undergo any
and the reaction was analyzed as described above for PPh3.
With PMe , the first filtrate contained triazole 18 and, to a
3
21
lesser extent, betaine 22 (Scheme 6). The second filtrate
contained peptide 19 as the major product, and only a relatively
3
3
small amount of peptide 20 was formed. Thus, with PMe , C−
3
O bond cleavage occurs by both triazole formation from
phosphazide 26 (Scheme 8) and hydrolysis of the ester in
2
2
iminophosphorane 29 (Scheme 8). With PBu , the filtrate
3
21
after the first step contained betaine 23 but no triazole 18. In
the second filtrate, peptide 19 was the major product and only a
relatively small amount of peptide 20 was formed. Therefore,
with PBu , loss of N from phosphazide 27 is faster than
detectable hydrolysis after 24 h in THF/H O, as determined by
2
HPLC. Scheme 10 shows two possible mechanisms for the
hydrolysis of these esters. The phosphorane nitrogen may
attack the carbonyl carbon followed by loss of the threonine-
derived peptides to give the three-membered ring intermediate
3
2
triazole formation, and hydrolysis of the ester moiety in the
iminophosphorane 30 is faster than phosphorane hydrolysis.
Interestingly, tripeptide 31 exhibited different reactivity
patterns compared to those of peptide 4 when subjected to
34, which then hydrolyzes to give the betaines (pathway a in
Scheme 10). Alternatively, direct attack of water on the
carbonyl carbon may occur, and the resulting tetrahedral
intermediate may be stabilized by the iminophosphorane group
possibly by forming a pentacoordinate intermediate such as 35
phosphines in THF/H O (Scheme 9). Subjecting 31 to 1.2
2
equiv of PMe or PBu in THF/H O for 1−3 h gave an
3
3
2
approximately equal mixture of triazole 18 and peptides 32 and
33 as determined by HPLC of the crude reaction mixture.
(
pathway b in Scheme 10). It is also possible that the
Subjecting 31 to 1.2 equiv of PPh in THF/H O for 20 h gave
3
2
iminophosphorane nitrogen or the γ-nitrogen of the initially
formed phosphazide acts as general base by deprotonation of a
water molecule and concomitant attack of the resulting
activated water molecule on the carbonyl group. Our data do
not allow us to distinguish between these two mechanisms or
rule out any others.
In order to avoid C−O bond cleavage, we attempted the
coupling of an activated form of FmocGlu(OtBu)OH with the
iminophosphorane derived from peptide 4. Lundquist and
Pelletier have reported the formation of peptide bonds in the
solid phase by reacting resin-bound peptides having an N-
only amine 32. After 20 h, this reaction was not complete, as
the iminophosphorane derived from peptide 31 and PPh was
3
evident in the HPLC chromatogram. Thus, peptide 31 exhibits
a product distribution fairly similar to that found with ester 10
14
with all of the phosphines. Betaines 21−23 were not formed
in any of the reactions, indicating that the ester moiety in the
iminophosphoranes derived from peptide 31 is not as
susceptible to hydrolysis as the analogous ester in the
iminophosphoranes (28−30 in Scheme 8) derived from
peptide 4.
It is not clear to us why the reactivity patterns (in terms of
the relative proportions of C−O bond cleavage, azide
terminal azido group with PMe in the presence of succinimidyl
3
D
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 8. Pathways for C−O Bond Cleavage When Peptide 4 Is Treated with Phosphines in THF/H O
2
Scheme 9. Products Formed upon Treatment of Tripeptide 31 with Phosphines in THF/H O
2
4
esters of Fmoc amino acids in dry dioxane. We attempted this
procedure on peptide 4 with FmocGlu(OtBu)OSu; however,
this resulted in the formation of mainly iminophosphorane 29.
Malkinson et al. conjugated an amino acid to a resin-bound
glycosyl azide by reacting the azide with the Fmoc amino acid
amount of peptide 19 could be reduced even further by
reacting peptide 4 with 5 equiv of the symmetric anhydride of
FmocGlu(OtBu)OH in the presence of 3 equiv of PBu for 5
3
min in dry THF, followed by the addition of about 5% water
and stirring for 18 h. However, very little of the desired product
36 was formed. Instead, the reduced peptide 20 was almost the
sole product, as determined by HPLC analysis of the crude
reaction mixture after cleavage from the support (Scheme 11).
It was found that reacting peptide 4 with 5 equiv of the
symmetric anhydride of FmocGlu(OtBu)OH in the presence of
23
in the presence of HOBt, DIC, and PBu in dry THF.
3
Applying these conditions to the coupling of FmocGlu(OtBu)-
OH to peptide 4 resulted in a complex mixture of products
which included the desired coupled product (36 in Scheme
1
1), as well as peptides 19 and 20. Surprisingly, when this
procedure was performed in the presence of 5% water, the
reduced peptide 20 was the major product and considerably
less C−O cleavage product (peptide 19) and coupled product
3 equiv of PBu in dry THF for 30 min or 5 h, followed by the
3
addition of water and stirring for 18 h, resulted in a more
complex mixture of products and an increase in the amount of
peptide 19.
36 were formed. Further experimentation revealed that the
E
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 10. Possible Mechanisms for the Hydrolysis of the
Ester Moiety in Iminophosphoranes 28−30
these compounds fragment in aqueous solution. Instead, in
α
β
aqueous solution, they readily undergo N −N fragmentation
25,26
to give the corresponding diazonium salts and amides.
We
are not aware of any detailed studies on the hydrolysis of N-
acyl-3-alkyltriazenes, although it appears that these species also
readily fragment in aqueous solution to give the respective
24,27
diazonium salts and amides.
Moreover, we were unable to
detect N-acyl triazene 43 (Scheme 14) when tripeptide 31 was
subjected to the conditions outlined in Schemes 11 and 12.
However, a small amount of hydrazone 38 was detected, which
suggests that a small amount of diazonium salt 45 was formed
in this reaction as hydrazone 38 was most likely produced by
28
PBu reduction of 45 (Scheme 14). It is possible that 45 was
3
produced by fragmentation of the N-acyl triazene 43 even
though we were unable to detect any FmocGlu(OtBu)NH2
(
44). It is also possible that 45 was produced by fragmentation
29
of the initially formed phosphazide.
Although the mechanism outlined in Scheme 13 has yet to be
supported by the detection of any of the proposed
intermediates, it is possible that the resin-bound peptide may
adopt a conformation that promotes the hydrolysis of 41
analogous to how a specific conformation in intermediate 30
may promote the hydrolysis of the ester group in the
phosphorane moiety to give betaine 23 (Schemes 7 and 8).
Having developed a procedure for reducing the azido group
in peptide 4, we applied it to the synthesis of Dap-E12-W13
Peptide 31 exhibited very different reactivity patterns
compared to those of peptide 4 when subjected to the
conditions outlined in Scheme 11. HPLC and MS analysis of
the reaction mixture revealed that the coupled product 37 was
by far the dominant peptide produced (Scheme 12). A small
amount of peptides 32 and 38 was also formed. We did not
detect any products resulting from C−O bond cleavage
(
betaine 23, triazole 18, or peptide 33). This was not entirely
unexpected since we had already found that the reactivity
patterns exhibited by resin-bound peptide 4 (Schemes 7 and 8)
can be very different from those exhibited by simpler substrates
such as ester 10 or peptide 31 in solution (Schemes 3 and 9).
The reason behind the preferential formation of peptide 20
over peptides 19 and 36 under the conditions described in
Scheme 11 is not entirely clear. Nevertheless, in Scheme 13, we
present a mechanism that might explain this phenomenon. In
(
Scheme 15). After the azido group in peptide 4 was reduced
using the procedure outlined in Scheme 11 to give mainly
peptide 3, standard Fmoc SPPS (DIC/HOBt) was employed to
incorporate Glu12 and D-Ser11, and then the alloc group was
1
removed using our standard conditions to give peptide 46. In
our previous syntheses of Dap-E12-W13, cyclization of peptide
46 to peptide 47 was achieved using PyBOP/HOAt/DIPEA in
DMF/DCM/NMP (3:2:2 (v/v)) containing 1% Triton X-100
this mechanism, peptide 4 reacts with PBu in dry THF to give
3
2
(
magic mixture) for 2 × 13 h. However, the cyclization
phosphazide 27. The latter can undergo loss of N to give
2
efficiency was only 40%. We have now found that the
cyclization can be achieved with 90% efficiency (by HPLC)
in just 2 × 1.5 h by using PyAOP/HOAt/2,4,6-collidine in
DMF containing 1% Triton X-100. The peptide was cleaved
from the resin, and side chain protecting groups were removed
using standard conditions. The overall yield of Dap-E12-W13
was 10.3%. The Dap-E12-W13 prepared in this manner
exhibited biological activity identical to that of Dap-E12-W13
prepared by our previous routes (MIC of 1 μg/mL against
iminophosphorane 30 (pathway A), which can produce C−O
bond cleavage products 19 and 23 or peptide 20. It is possible
that phosphazide 27 can also react with the excess anhydride to
give intermediate 39, which would be in equilibrium with
intermediate 40 (pathway B). Addition of water results in
24
protonation of 39 to give N-acyl triazenes 41 and 42. It is
possible that 42 hydrolyzes to give N and the reduced peptide
2
20, even though studies on N-acyl-1,3-dialkyltriazenes have
shown that hydrolysis is not the dominant pathway by which
Scheme 11. Products Formed upon Reaction of Peptide 4 with an Excess of the Symmetric Anhydride of FmocGlu(OtBu)OH in
the Presence of Excess PBu in Dry THF for 5 min Followed by the Addition of Water and Stirring for 18 h
3
F
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 12. Products Formed upon Reaction of Peptide 31 with an Excess of the Symmetric Anhydride of FmocGlu(OtBu)OH
in the Presence of Excess PBu in Dry THF for 5 min Followed by the Addition of Water and Stirring for 18 h
3
Scheme 13. Possible Mechanism for the Formation of Peptide 20
Scheme 14. Possible Route to Peptide 38 during the Reaction of Peptide 31 under the Conditions Outlined in Schemes 11 and
12
Bacillus subtilis 1046 and 3 μg/mL against B. subtilis PY79 in the
presence of 5 mM CaCl2).
phosphines. The extent of triazole formation depends upon
leaving group ability. Some C−O bond cleavage and formation
of triazole 18 occurred when peptide 4 was treated with DTT/
DIPEA. In contrast to the simple α-azido ester 10, which gave
1
,2
SUMMARY AND CONCLUSIONS
■
only amine 12 when treated with PPh in THF/H O, peptide 4
We have shown that α-azido esters, when treated with DTT/
DIPEA, can undergo C−O bond cleavage to give triazoles,
analogous to when α-azido esters are treated with alkyl
3
2
gave a mixture of the desired amine product 20 as well as the
undesired C−O cleavage product 19 and betaine 21 when
G
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 15. Modified Route to Dap-E12-W13
1
treated with PPh in THF/H O. Peptides 19 and 20, betaines
shifts (δ) for H NMR spectra run in CD
3
OD are reported in parts per
3
2
1
3
million relative to residual solvent protons (δ 3.30). For C NMR
spectra run in CDCl , chemical shifts are reported in parts per million
relative to the solvent residual carbon (δ 77.0 for central peak). For
C NMR spectra run in CD OD, chemical shifts are reported in parts
per million relative to the CD OD residual carbon (δ 49.0 for the
central peak). All P NMR spectra are referenced to H PO (external
22 and 23, and/or triazole 18 were formed when peptide 4 was
treated with PMe and PBu . Overall, these results indicate that
3
3
3
one must exercise caution when azido esters are employed in
depsipeptide synthesis.
It was found that the azido group in peptide 4 could be
efficiently reduced to the desired amino peptide 3 by reacting
peptide 4 with an excess of the symmetric anhydride of
13
3
3
31
3
3
standard).
General Procedure for the Reduction of Esters 10 and 11
with DTT/DIPEA. To a solution of ester 10 or 11 in DMF (0.1 mmol
in 1 mL dry DMF) was added a solution of DTT (0.4 mmol) and
DIPEA (0.2 mmol) in dry DMF (0.5 mL). The mixture was stirred for
FmocGlu(OtBu)OH in the presence of excess PBu for 5 min
3
in dry THF, followed by the addition of about 5% water and
stirring for 18 h. By applying this new reduction procedure, as
well as an improved cyclization protocol, to the synthesis of the
daptomycin analogue, Dap-E12-W13, we achieved a much
higher yield in comparison to our previously reported
approaches.
2
h and then analyzed by analytical HPLC (linear gradient of 5%
ACN/95% H O (0.1% TFA) to 100% ACN over 45 min, λ = 230 nm,
2
see Figures S1 and S2 in the Supporting Information). Triazole 13,
from the reaction of phenyl ester 11, was purified by preparative
HPLC (linear gradient of 15% ACN/85% H O (0.1% TFA) to 25%
2
ACN/75% H O (0.1% TFA) over 60 min, t = 55 min). Yield = 14.5
mg (83%, white solid). The H NMR spectrum of compound 13
2
r
EXPERIMENTAL SECTION
1
■
General. 2-Chlorotrityl chloride resin, all Fmoc amino acids, and
coupling reagents used for peptide synthesis were purchased from
obtained in this manner was identical to that of compound 13
obtained previously by reduction of ester 11 using PMe in THF/
3
14
1
13
commercial sources. N Trp(Boc)OH was prepared according to a
literature procedure. ACS grade DMF, THF, piperidine, TFA, TIS,
thioanisole, and 2-methylpiperidine (2-MP) were purchased from
H
2
O: H NMR (CD
NMR (CD OD) δ 140.4, 129.4, 127.2, 30.0; HRMS (m/z) [M + H]
calcd for C 176.08184; found 176.08185.
3
OD) δ 7.17−7.31 (5H, m), 3.95 (2H, s); C
3
30
+
3
H10ON
9 3
commercial suppliers and used without further purification. CH Cl2
Reaction of Peptide 4 with DTT/DIPEA. To peptide 4 (0.025
mmol) was added a solution of DTT (2M) and DIPEA (1M) in dry
DMF (2 mL containing 0.616 g of DTT and 0.348 mL of DIPEA).
The mixture was stirred at rt for 2 h and filtered. The filtrate was
analyzed by analytical HPLC and HRMS (see Figure S5 in the
Supporting Information). The resin was washed with DMF (3×) and
2
(
DCM) was distilled from calcium hydride under nitrogen. Peptide 4
1
was prepared as previously described. Peptide syntheses and studies
with peptide 4 were performed manually using a rotary shaker for
31
agitation. Reversed-phase C18 columns (10 μM, 250 × 4.6 mm, or
10 μM, 250 × 20 mm) were used for analytical and semipreparative
HPLC at flow rates of 1.0 mL/min for analytical HPLC and 10 mL/
min for semipreparative HPLC. The HPLC was equipped with a UV
detector. All HPLC chromatograms (see the Supporting Information)
DCM (3×) and then subjected to TFA/thioanisole/TIS/H O
2
(88:2:5:5, 2 mL). The mixture was filtered, and the resin was rinsed
with an additional cleavage cocktail. The combined filtrates were
were obtained using a linear gradient of 10% ACN/90% H O (0.1%
concentrated to one-fourth the original volume using a stream of N
2
2
TFA) to 90% ACN/10% H O (0.1%TFA) over 50 min, with the
gas. The peptide was precipitated by adding cold ether (3 mL). The
precipitated peptide (white precipitate) was collected by centrifugation
and washed with cold ether (2×) and then analyzed by analytical
HPLC and HRMS (see Figure S6 in the Supporting Information).
2
detector set to 280 nm unless stated otherwise. High-resolution
+
positive ion electrospray (ESI HRMS) mass spectra were obtained
using an orbitrap mass spectrometer. 1:1 MeOH/H O + 0.1% formic
2
acid was used as a solvent. The biological activity of Dap-E12-W13 was
Reaction of Peptide 4 with Phosphines in THF/H O. To
2
2
determined as previously described.
peptide 4 (0.025 mmol) was added a solution of PPh (6 equiv) or
3
1
H NMR and J-modulated spin−echo (J-MOD) 13C NMR spectra
PMe or PBu (3 equiv) in 2 mL of THF/H O (8:2 v/v). The mixture
3
3
2
were recorded on a 300 MHz NMR instrument. Chemical shifts (δ)
for H NMR spectra run in CDCl are reported in parts per million
was agitated for 1 h for Me P and 13 h for Bu P and Ph P and then
3 3 3
1
filtered and washed with DMF and DCM. The filtrate was analyzed by
3
relative to the internal standard tetramethylsilane (TMS). Chemical
analytical HPLC and HRMS (see Figures S7, S16, and S18 in the
H
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Supporting Information). The resin was treated with TFA/
Me P and Bu P reactions and 20 h for Ph P reaction and then
3
3
3
thioanisole/TIS/H O (88:2:5:5, 2 mL). The mixture was filtered,
and the resin was rinsed with an additional cleavage cocktail. The
analyzed by analytical HPLC (linear gradient of 5% ACN/95% H O
(0.1% TFA) to 100% ACN over 60 min, λ = 230 nm; see Figures
2
2
combined filtrates were concentrated to one-fourth the original
S23−S25 in the Supporting Information) and LRMS.
volume using a stream of N gas. The peptide was precipitated by
Reaction of Peptide 31 with Excess PBu in the Presence of
2
3
adding cold ether (3 mL). The precipitated peptide (white precipitate)
was collected by centrifugation and washed with cold ether (2×). The
peptide was analyzed by analytical HPLC and HRMS (see Figures S8,
S17, and S19 in the Supporting Information).
an Excess Amount of the Symmetric Anhydride of FmocGlu-
(
OtBu)OH and Water. A solution of the symmetric anhydride of
FmocGlu(OtBu)OH in dry THF (2 mL, 5 equiv) was added to
peptide 31 (0.023 g, 0.030 mmol) followed by 3 equiv of Bu P and
3
stirred for 5 min. H O (0.1 mL) was added, and the mixture was
Reaction of N Trp(Boc)OH with PPh . To a solution of
2
3
3
stirred for 18 h. An aliquot was withdrawn from the reaction mixture,
N Trp(Boc)OH (0.200 g, 0.61 mmol) in dry ether (4 mL) was
3
diluted in CH CN/H O (1:1), and then analyzed by analytical HPLC
added a solution of PPh (0.159g, 0.61 mmol) in dry ether (2 mL).
3
2
3
(
λ = 230 nm; see Figure S26 in the Supporting Information).
Synthesis of Dap-E12-W13 Starting from Peptide 4. To
The mixture was stirred at rt for 2 h and filtered. The filter cake was
washed with ether and dried under high vacuum, yielding a white solid
peptide 4 (0.025 mmol) was added a solution of 5 equiv of the
symmetric anhydride of FmocGlu(OtBu)OH in THF (0.5 mL)
(
0.102 g, 30%). The analytical HPLC chromatogram of this precipitate
shows one major peak with t = 36.6 min as well as well as a few very
r
followed by Bu P (3 equiv). The mixture was agitated for 5 min, and
minor peaks due to PPh and POPh (see Figure S9 in the Supporting
3
3
3
1
H O (0.1 mL) was added and the mixture agitated for 18 h. The
2
Information): H NMR (CDCl ) δ 8.10 (1H, d, J = 8.4 Hz), 7.60 (1H,
3
mixture was filtered and washed with DMF (3×) and DCM (3×).
FmocGlu(OtBu)OH and Fmoc-D-Ser(OtBu)OH were incorporated
using standard Fmoc SPPS techniques employing DIC/HOBt as
coupling agents (3.5 h) and 20% 2-methylpiperidine/DMF (3 × 10
min) for Fmoc removal. The alloc group was removed using DMBA
(dimethylbarbituric acid, 10 equiv) in the presence of a catalytic
amount of Pd(PPh ) in DCM/DMF (3:1 v/v) for 1 h. After filtration,
the resin was washed with DCM (3×) and a 1.0% solution of sodium
diethyldithiocarbamate trihydrate in DMF (3×) to remove excess Pd
catalyst and then DCM (3×) and DMF (3×). On-resin cyclization was
performed using PyAOP/HOAt/2,4,6-collidine (5/5/10 equiv) in the
presence of 1% Triton X-100 in DMF (2 mL) (2 × 1.5 h). Crude Dap-
E12-W13 was obtained by treating peptide 47 with a solution of TFA/
d, J = 7.4 Hz), 7.47 (1H, s), 7.19−7.40 (19H, m), 4.17 (1H, ddd, J =
1
1
1
0.6, 9.0, 3.2 Hz), 3.15 (1H, dd, J = 14.8, 3.2 Hz), 3.14 (1H, dd, J =
4.8, 9.0 Hz), 1.63 (9H, s); 13C NMR (CDCl ) δ 172.3, 149.4, 135.4,
3
34.4, 133.6 (d, J = 19.5 Hz), 131.4 (d, J = 10.8 Hz), 131, 130.4, 128.3
(
3
5
d, J = 14.1 Hz), 124.5, 124.0, 122.5, 119.2, 116.8, 115.2, 83.5, 56.2,
31
0.0, 28.1; P NMR (CDCl ) δ −10.9, −2.8 (PPh ), 31.7 (POPh ),
3
3
3
1.2; 3 P NMR (THF) δ −39.4, −0.4 (PPh ), 23.8 (POPh );
1
3
1
P
3 4
3
3
NMR (THF/10% H O) δ 2.05, 27.5 (POPh ); HRMS (m/z) [M +
H] calcd for C H O N P 565.22507; found 565.22446. The HPLC
2
3
+
34
34
4
2
chromatogram of a mixture of this precipitate with compound 21/24
isolated from filtrate #1 in the reaction of peptide 4 with PPh in
3
THF/H O shows a single peak, suggesting that these two compounds
2
are identical (see Figure S10 in the Supporting Information).
thioanisole/TIS/H O (88:2:5:5, 2 mL) for 2 h. The mixture was
Reaction of Peptide 4 with Excess PBu in the Presence of
2
3
the Symmetric Anhydride of FmocGlu(OtBu)OH and Water.
The symmetric anhydride of FmocGlu(OtBu)OH was prepared by
stirring FmocGlu(OtBu)OH (0.25 mmol) and DIC (0.125 mmol) in
filtered, and the resin was rinsed with an additional cleavage cocktail.
The combined filtrates were concentrated to one-fourth the original
volume using a stream of N gas. The peptide was precipitated by
2
adding cold ether (3 mL). The precipitated peptide was collected by
centrifugation and washed with cold ether (2×). Pure Dap-E12-W13
was obtained using semipreparative HPLC employing a linear gradient
32
dry THF (2 mL) for 1 h at rt. The mixture was filtered. The filtrate
was added to peptide 4 (0.025 mmol), followed by Bu P (3 equiv),
3
and agitated for 5 min. H O (0.1 mL) was added, and the mixture was
2
agitated for 18 h. The resin was washed with DMF (3×) and DCM
of 35:65 CH CN/H O (0.1% TFA) to 43:57 CH CN/H O (0.1%
3 2 3 2
(
3×) and then treated with TFA/thioanisole/TIS/H O (88:2:5:5, 2
TFA) over 60 min. Fractions containing Dap-E12-W13 were collected,
concentrated by high vacuum, and lyophilized to give pure Dap-E12-
W13 as a white powder (4.3 mg, 10.7% yield based on resin loading):
2
mL). The mixture was filtered, and the resin was rinsed with an
additional cleavage cocktail. The combined filtrates were concentrated
+
+
to one-fourth the original volume using a stream of N gas. The
HRMS (ESI ) calcd for C72H N O25 1602.70708 [M + H] ; found
100 17
2
peptide was precipitated by adding cold ether (3 mL). The
precipitated peptide (white precipitate) was collected by centrifugation
and washed with cold ether (2×). The peptide was analyzed by
analytical HPLC (see Figure S20 in the Supporting Information).
1602.70525. See Figures S27 and S28 in the Supporting Information
for the analytical HPLC chromatogram and mass spectral data of
purified Dap-E12-W13.
Peptide 31. To a solution of N Trp(Boc)OH (0.660 g, 2.00
3
ASSOCIATED CONTENT
■
mmol) in dichloromethane (5 mL) was added diisopropylcarbodii-
mide (DIC, 0.154 mL, 1.00 mmol). The mixture was stirred for 1 h.
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.6b02309.
33
ZThrGlyOBn (0.4 g, 1.0 mmol) and DMAP (0.012 g, 0.1 mmol)
were added, and the mixture was stirred for 24 h. The mixture was
filtered, and the filtrate wsa dissolved in Et O and washed with 0.25 M
2
1
13
31
NaOH, 0.1 M HCl, and saturated brine. The organic layer was dried
Analytical HPLC chromatograms, H, C, and P NMR
spectra, and HRMS data (PDF)
(
Na SO ) and concentrated, giving a yellow oil. The oil was subjected
2 4
to flash chromatography (30% EtOAc/70% hexane) which provided
1
peptide 31 as a white foam (0.606 g, 85% yield): H NMR (CDCl ) δ
3
8
6
5
.13 (1H, d, J = 7.9 Hz), 7.51−7.55 (2H, m), 6.87−7.32 (12H, m),
.89 (1H, t, J = 5.3 Hz), 5.68 (1H, d, J = 8.5 Hz), 4.48−5.51 (1H, m),
.13 (2H, s), 5.11 (2H, s), 4.48 (1H, d, J = 5.3 Hz), 4.23 (1H, t, J = 6.4
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: s5taylor@uwaterloo.ca.
Hz), 3.99 (2H, br s), 3.23 (1H, dd, J = 5.3, 14.8 Hz), 3.09 (1H, dd, J =
13
8
.5, 14.8 Hz), 1.65 (9H, s), 1.23 (3H, d, J = 6.4 Hz); C NMR
Notes
(
CDCl ) δ 169.2, 168.8, 168.7, 156.3, 149.4, 135.8, 135.3, 134.9, 129.9,
3
The authors declare no competing financial interest.
1
28.5, 128.2, 128.0, 124.6, 124.3, 122.5, 118.6, 115.3, 114.8, 83.7, 71.9,
+
6
7.4, 67.2, 61.7, 57.7, 41.3, 28.1, 27.1, 15.9; HRMS (m/z) [M + H]
calcd for C H O N 713.29295; found 713.29287.
Reaction of Peptide 31 with Phosphines in THF/H O. To a
solution of peptide 31 (0.02 g, 0.028 mmol) in THF (0.124 mL) and
37
41
9
6
ACKNOWLEDGMENTS
■
2
This research was supported by a Discovery Grant from the
Natural Sciences and Engineering Research Council (NSERC)
of Canada to S.D.T. (Grant No. 155283-2012).
H O (0.010 mL) was added a 1 M solution of PMe , PBu , or PPh in
2
3
3
3
THF (0.056 mL, 1.2 equiv). The mixture was stirred for 1 h for the
I
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
that this stepwise approach to the reduction of the azido group in
peptide 4 would still result in the formation of peptide 19 and betaines
REFERENCES
■
(
1) An exception to this are peptides bearing an N-terminal α-
2
(
5
(
1−23.
23) Malkinson, J. P.; Falconer, R. A.; Toth, I. J. Org. Chem. 2000, 65,
249−5252.
24) Strazewski and co-workers have reported the formation of N-
azidoaspartate residue which can undergo elimination of azide ion
when treated with reagents that are commonly used for removing the
Fmoc group during SPPS. See: Lohani, C. R.; Rasera, B.; Scott, B.;
Palmer, M.; Taylor, S. D. J. Org. Chem. 2016, 81, 2624−2628.
2) Lohani, C. R.; Taylor, R.; Palmer, M.; Taylor, S. D. Org. Lett.
015, 17, 748−751.
3) Lam, H. Y.; Zhang, Y.; Liu, H.; Xu, J.; Wong, C. T. T.; Xu, C.; Li,
X. J. Am. Chem. Soc. 2013, 135, 6272−6279.
4) Lundquist, J. T.; Pelletier, J. C. Org. Lett. 2001, 3, 781−783.
5) Meldal, M.; Juliano, M. A.; Jansson, A. M. Tetrahedron Lett. 1997,
8, 2531−2534.
6) Tornøe, C. W.; Davis, P.; Porreca, F.; Meldal, M. J. Pept. Sci.
000, 6, 594−602.
7) Goddard-Borger, E. D.; Stick, R. V. Org. Lett. 2007, 9, 3797−
800.
8) Yan, R. B.; Yang, F.; Wu, Y.; Zhang, L. H.; Ye, X. S. Tetrahedron
Lett. 2005, 46, 8993−8995.
9) Katritzky, A. R.; El Khatib, M.; Bolshakov, O.; Khelashvili, L.;
Steel, P. J. J. Org. Chem. 2010, 75, 6532−6539.
10) Nyffeler, P. T.; Liang, C. H.; Koeller, K. M.; Wong, C. H. J. Am.
Chem. Soc. 2002, 124, 10773−10778.
11) Cartwright, I. L.; Hutchinson, D. W.; Armstrong, V. W. Nucleic
Acids Res. 1976, 3, 2331−2339.
12) Staros, J. V.; Bayley, H.; Standring, D. N.; Knowles, J. R.
Biochem. Biophys. Res. Commun. 1978, 80, 568−572.
13) Bayley, H.; Standring, D. N.; Knowles, J. R. Tetrahedron Lett.
acyl triazenes from the reaction of an alkyl azide with PBu , Fmoc, or
3
(
2
(
Boc amino acids, HOBt, and DIC at 0 °C. See: Charafeddine, A.;
Dayoub, W.; Chapuis, H.; Strazewski, P. Chem. - Eur. J. 2007, 13,
5
(
566−5584.
25) Smith, R. H.; Wladkowski, J. B.; Herling, J. A.; Pfaltzgraff, T. D.;
Pruski, B.; Klose, J.; Michejdal, C. J. J. Org. Chem. 1992, 57, 6448−
454.
26) Smith, R. H.; Wladkowski, J. B.; Herling, J. A.; Pfaltzgraff, T. D.;
Pruski, B.; Klose, J.; Michejda, C. J. J. Org. Chem. 1992, 57, 654−661.
27) Wakimoto, T.; Miyata, K.; Ohuchi, H.; Asakawa, T.; Nukaya, H.;
Suwa, Y.; Kan, T. Org. Lett. 2011, 13, 2789−2791.
28) The reduction of diazonium salts to hydrazones with
(
(
3
(
2
(
3
(
6
(
(
(
phosphines has been well-documented. For example, see: Yasui, E.;
Wada, M.; Nagumo, S.; Takamura, N. Tetrahedron 2013, 69, 4325−
(
4
330.
(
29) For an example where such a fragmentation may occur, see ref
4.
30) Valverde, I. E.; Bauman, A.; Kluba, C. A.; Vomstein, S.; Walter,
M. A.; Mindt, T. L. Angew. Chem., Int. Ed. 2013, 52, 8957−8960.
31) Wavreille, A. S.; Garaud, M.; Zhang, Y.; Pei, D. Methods 2007,
2, 207−219.
32) Mikolajczyk, M.; Kielbasinski, P. Tetrahedron 1981, 37, 233−
84.
33) Seebach, D.; Sommerfeld, T. L.; Jiang, Q.; Venanzi, L. M. Helv.
Chim. Acta 1994, 77, 1313−1330.
(
1
(
(
(
4
(
2
(
(
1
(
(
978, 19, 3633−3634.
(
14) Taylor, S. D.; Lohani, C. R. Org. Lett. 2016, 18, 4412−4415.
15) The other mechanism proposed by Cartwright et al. involved
attack at the α-nitrogen.
(
16) There are many possible variations on the mechanism for
triazole formation shown in Scheme 4, including one that involves
attack at the α-nitrogen.
(
17) We had previously avoided using PPh for reducing the azido
3
groups as we found that the reaction proceeded considerably slower in
comparison to PMe3.
(
18) Gololobov and co-workers reported that the reaction of
triphenylphosphine with azidoacetic acid in dry THF yielded a
precipitate that exhibited a 31P NMR chemical shift of 25.2 ppm in
+
−
CDCl , which they attributed to the betaine, Ph P -NHCH COO .
3
3
2
See: Gololobov, Y. G.; Gusar, N. I.; Chaus, M. P. Tetrahedron 1985,
4
1, 793−799. Earlier, these workers had reported that the reaction of
2-azidobutanoic acid or N ValOH with PPh in THF or benzene also
3
3
gave the corresponding betaines after recrystallization from acetoni-
trile; however, the 31P NMR chemical shift of these compound in
DMF was −28.4 and −37 ppm, respectively, indicating that these
compounds exist as the corresponding cyclic trigonal bipyramidal
phosphoranes in DMF. See: Chaus, M. P.; Gusar, N. I.; Gololobov, Y.
G. Zh. Obshch. Khim. 1982, 52, 24−30.
(
19) Garcia, J.; Urpi, F.; Vilarrasa, J. Tetrahedron Lett. 1984, 25,
841−4844.
20) Zaloom et al. reported that the reaction of ethyl
diphenylphosphinite with N ValOH gave the corresponding cyclic
4
(
3
3
1
trigonal bipyramidal phosphorane which exhibited a P NMR
chemical shift of −35.2 ppm in an undisclosed solvent. See: Zaloom,
J.; Calandra, M.; Roberts, D. C. J. Org. Chem. 1985, 50, 2601−2603.
(
21) We have chosen to show compounds 22 and 23 as betaines for
simplicity sake, even though it is possible that they could be in
equilibrium with their corresponding cyclic trigonal bipyramidal
phosphoranes.
(
22) We previously reported that when α-azido ester 10 was
subjected to PMe in dry THF for 30 min followed by reaction with
3
water for 2 h, only the reduced amine 12 was formed and no C−O
bond cleavage occurred (see ref 14). This suggested another potential
way to avoid C−O bond cleavage when reducing the azido group in
peptide 4. However, the fact that the ester bonds in the
iminophosphoranes 28−30 are susceptible to hydrolysis indicates
J
DOI: 10.1021/acs.joc.6b02309
J. Org. Chem. XXXX, XXX, XXX−XXX