Investigations into the decomposition of aminoacyl-substituted monosaccharide scaffolds from a drug discovery library

Article abstract of DOI:10.1039/c5ob00122f

This study investigated the unexpected decomposition and associated intermediates of compound 1, a specific member of a drug discovery library based on a monosaccharide scaffold. LC/MS and NMR spectroscopic analyses indicated that, under acidic conditions

Full text of DOI:10.1039/c5ob00122f

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Investigations into the decomposition of
aminoacyl-substituted monosaccharide scaffolds
from a drug discovery library†
Cite this: Org. Biomol. Chem., 2015,
13, 4070
Q. Q. He,a N. Wimmer,b G. Verquin,b W. Meutermansb and V. Ferro*a
This study investigated the unexpected decomposition and associated intermediates of compound 1, a
specific member of a drug discovery library based on a monosaccharide scaffold. LC/MS and NMR
spectroscopic analyses indicated that, under acidic conditions, 1 can be converted into the 4-amino-
galactoside 2, due to cleavage of the 4-aminobutanoyl side chain. The reaction occurs most likely
through an initial intramolecular amino–amide interaction, followed by an N- to O-acyl transfer of the
side chain from C-4 to the C-6 position to form an ester intermediate (5), detectable by NMR, and sub-
sequent hydrolysis. Similar decomposition reactions could be induced in selected compounds with
similar structures, containing a free hydroxyl group at C-6 and a 4-aminobutanoyl side chain at C-4 of an
aminogalactoside. Furthermore, three model compounds were synthesized without a C-6 hydroxyl group
and with different length aminoalkanoyl side chains at the C-4 position. The model compounds all
decomposed under acidic conditions, but at different rates and much slower when compared with com-
pound 1, suggesting that both the C-6 hydroxyl group and the length of the side chain have an influence
on stability.
Received 21st January 2015,
Accepted 19th February 2015
DOI: 10.1039/c5ob00122f
www.rsc.org/obc
rates and help to efficiently identify lead compounds for
potential therapeutic targets from the structurally related sets
Introduction
Monosaccharide scaffolds provide an excellent platform for of hits.1,2
drug discovery with the aim of accurately transposing known
In efforts to obtain systematic diverse libraries using mono-
peptide ligands onto a metabolically stable scaffold while saccharide scaffolds, a solid phase synthesis technology has
maintaining a good level of bioactivity.1–4 For example, in the been developed [Versatile Assembly on Sugar-like Templates
past two decades such scaffolds have been successfully used to (VAST™)] enabling parallel and efficient generation of
convert bioactive peptides into more drug-like molecules.2,5–7 structurally diverse libraries.3,9 Three orthogonally protected
A number of unique features of monosaccharide scaffolds con- monosaccharide scaffolds, based on D-galactopyranose, D-gluco-
tribute to the generation of diverse chemical libraries.4,8 For pyranose and D-allopyranose were identified.10,11 These
instance, with only one scaffold and three different substitu- three scaffolds were designed to include a resin attachment
ents, there are already 60 different ways to present these substi- point and utilised protecting groups that are stable to various
tuents to a binding site.4,8 This increases to almost 96 reaction conditions required to introduce the substituents.10
different presentations by changing just one of the substituted In addition, these scaffolds could be readily prepared in good
chiral centres.4,8 Therefore, compounds built on monosacchar- yield at large scale ranging from 100 g to >1 kg.10 These
ide-based templates can rapidly attain high molecular diversity scaffolds have been successfully used as the starting points in
within one class of compound, as well as leading to rigid pro- the preparation of systematic diverse libraries via solid phase
ducts with well-defined geometry.1,2 This could increase hit synthesis.10 This allows the main reactions for library prepa-
ration to be carried out on resin, simplifying the manipulation
of the library,4,9 and thus providing a platform for efficient
aSchool of Chemistry and Molecular Biosciences, the University of Queensland,
Brisbane, Qld 4072, Australia. E-mail: v.ferro@uq.edu.au; Fax: +617-33654299;
Tel: +617-3346 9598
synthesis of compounds with high molecular diversity for pre-
clinical assessment.
The galactoside derivative 1 (Fig. 1) was synthesized as part
of a diversity scanning library (DSL) production process.9–11
After approximately 2.5 years of storage at 2–8 °C (10 mM in
DMSO), this sample was identified as a hit against a prop-
bAlchemia Ltd, Eight Mile Plains, Qld 4113, Australia
†Electronic supplementary information (ESI) available: LC/MS data for com-
pound 1; 1H and 13C NMR data for compounds 1 and 2; copies of 1H and 13C
NMR spectra for compounds 1–5 and 11–22. See DOI: 10.1039/c5ob00122f
4070 | Org. Biomol. Chem., 2015, 13, 4070–4079
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 1 The structure of compound 1, decomposition product 2, deri-
vatives 3 and 4, and proposed intermediate 5.
rietary drug target following a high-throughput screening cam-
paign. However, LC/MS analysis showed that the sample had
decomposed to a compound with MW = 445 (Fig. S1,† tR
=
4.79 min, m/z 446.22 [M + H]+). Thus, the decomposition
product was suspected to be responsible for the observed
activity. Herein we report the results of investigations into the
structure, stability and decomposition of compound 1 and
some related compounds, the structural elucidation of the
decomposition product and a key intermediate, and propose a
mechanism for their formation.
Fig. 2 LC/MS analysis of 1 (1 mM) in H2O with 10% TFA, pH 1 and 80 °C:
(a) after 1 h; (b) after 13 h.
Results and discussion
Investigation of the structure of the decomposition product
with the same mass as the starting material was detected by
LC/MS (Fig. 2a, tR = 4.47 min, m/z 531 [M + H]+, ∼10%). After
13 hours, the starting material was totally decomposed to
another compound with mass 445 (Fig. 2b, tR = 4.80, m/z 446
[M + H]+). HPLC co-elution experiments showed that the
resynthesized decomposition product and the one observed
from the DSL were co-eluted when injected at similar concen-
trations (data not shown), hence, it was concluded that the
resynthesized decomposition product was likely to be the
same as the one observed in the original DSL sample. When 1
was treated with 10% TFA in DMSO, there was more intermedi-
ate formed; however, the formation of the final decomposition
product was much slower than when the reaction was per-
formed in water.
Due to the limited amount of the decomposition product avail-
able for further study, compound 1 was resynthesized9–11 at
larger scale (m = 510 mg) and was fully characterized by NMR
spectroscopy in both DMSO-d6 and MeOH-d4, and by LC/MS
1
(Fig. S2,† tR = 4.87 min, m/z 531.10 [M + H]+). All H and 13C
NMR chemical shifts were fully assigned using a combination
of COSY, HSQC and HMBC experiments (see ESI, Tables
S1–2†). The newly synthesized sample of 1 was inactive in the
bioassay. Tests were then conducted on compound 1 under
various conditions in order to identify the optimal conditions
to reproduce the originally observed decomposition. Interest-
ingly, when 1 was heated in DMSO solution (10 mM, to repro-
duce the storage conditions) containing 5% H2O in the
presence of TFA (10%, pH 1) at 50 °C, after 2 days, a more
polar intermediate with mass 530 was observed by LC/MS
(Fig. S3b,† tR = 4.38 min, m/z 531 [M + H]+, ∼5%). After 6 days,
a product with mass 445 had started to form slowly (data not
shown). Heating was continued for 13 days and then the
sample was left at room temperature. After 13 days heating,
the amount of the intermediate had increased significantly
(Fig. S3c,† tR = 4.37 min, m/z 531 [M + H]+, ∼30%). After 49
days, the more polar intermediate was no longer detectable
(data not shown). After 10 months, only a single product could
be detected by LC/MS (Fig. S3d,† tR = 5.11 min, m/z 446 [M +
H]+), which could possibly be the same as the decomposition
product initially observed from the DSL.
In order to identify the structure of the decomposition
product, another sample of 1 was treated with aqueous 10%
TFA at 80 °C as described above. After 10 minutes, the more
polar intermediate was observed by LC/MS and after 20 hours,
the decomposition product had formed (tR = 4.91 min,
m/z 446.58 [M + H]+). After evaporation of the solvent, the crude
residue was purified by passage through a reverse phase SPE
C18 cartridge to afford the product in 94% yield. The resynthe-
sized decomposition product was fully characterized by NMR
spectroscopy in two different solvents, DMSO-d6 and MeOH-
1
d4. All H and 13C chemical shifts were fully assigned using a
combination of COSY, HSQC and HMBC experiments (see ESI,
Tables S3–4†). In comparison with the NMR spectra of 1, the
NMR spectra of the decomposition product revealed that the
peaks for the 4-aminobutanoyl side chain were no longer
present. In addition, the proton assigned to H-4 was shifted
In an attempt to more rapidly reproduce the decomposition
product, compound 1 (1 mM) was treated with 10% TFA in
H2O (pH 1) at 80 °C. After one hour, a more polar intermediate
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4070–4079 | 4071

View Article Online
Paper
Organic & Biomolecular Chemistry
upfield from 4.55 ppm to 3.58 ppm, consistent with the loss of 1 (6.5 mg) in a mixture of 10% TFA in anhydrous DMSO-d6
the amide group. Moreover, the NMR signals for the other (pH 1) was heated at 80 °C for 6 hours in an NMR tube. The
1
protons had similar chemical shifts in all respects to that of 1. reaction progress was monitored by H NMR and COSY spec-
The NMR data was thus consistent with 4-aminogalactoside 2 troscopy (see ESI†). The NMR spectra showed that the ratio
as the decomposition product. These NMR data were initially between 1 and the intermediate was approximately 3 : 2 after
surprising because compound 2 is a member of the DSL and 6 hours heating and no product 2 was formed in the absence
was inactive in the original high-throughput screening assay.
of water. The NMR data was analyzed and a possible structure
Further confirmation of the structure of the decomposition of the reaction intermediate was postulated from these data as
product was obtained by the preparation of two derivatives. the 6-O-ester 5 (Fig. 1).
1
Firstly, treatment with excess acetic anhydride and a catalytic
Compared with the H NMR spectrum of 1, in the reaction
amount of DMAP in pyridine at 50 °C for 1.5 hours followed by mixture it was noteworthy that H-4 of the intermediate was
purification of the reaction mixture by reverse phase SPE C18 shifted upfield from 4.56 ppm to 3.88 ppm, while H-5 and
cartridge gave the diacetate 3 (tR = 5.80 min, m/z 530.47 H-6a,b, respectively, were shifted downfield (3.88 to 4.27 ppm,
[M + H]+). A tetraacetate derivative was also detected by LC/MS and 3.31 to 4.13 ppm, respectively). These shifts are consistent
of the reaction mixture (tR = 6.44 min, m/z 614.51 [M + H]+, with migration of the 4-aminobutanoyl side chain from C-4 to
636.40 [M + Na]+), presumably due to acetylation of the guani- C-6. In addition, the signal for the amide proton attached to
dino group, however, this product was unstable and could not C-4 was shifted slightly downfield to 8.03 ppm and integrated
be isolated. Treatment of the decomposition product with di- for two protons, suggesting the formation of an amino group.
tert-butyldicarbonate and triethylamine in anhydrous DMF at Furthermore, no cross peak was observed in the HMBC spec-
room temperature for 65 minutes followed by purification by trum between this signal and the carbonyl group of the
reverse phase SPE C18 cartridge gave a single product, the car- 4-aminobutanoyl side chain. Moreover, a 13C–1H long range
1
bamate 4 (tR = 6.45 min, m/z 546.59 [M + H]+). The H NMR interaction between protons H-6a,b of the intermediate and
and COSY spectra of 3 and 4 showed the expected downfield the carbonyl group was detected by HMBC, confirming the
shifts for H-4 (4.61 and 4.18 ppm, respectively) and H-6 (for 3, presence of the 4-aminobutanoyl group at the C-6 position.
4.02 ppm). Finally, an authentic sample of 2 from the DSL Taken together, these NMR data were consistent with the pro-
(tR = 4.92 min, m/z 446.58 [M + H]+) was re-purified by reversed posed structure 5. To further support the presence of an ester
phase SPE C18 cartridge and analyzed by 1H NMR spectro- group at C-6, the reaction mixture containing 1 and the inter-
scopy, which showed that the two samples were identical. mediate (3 : 2) was treated with NaOMe (pH 11) at 50 °C for
The data confirm that the final decomposition product was 10 minutes. LC/MS analysis showed that 1 was stable to these
the 4-aminogalactoside 2, however, the nature of the initially conditions, while the intermediate was readily converted into
formed intermediate and the identity of the bioactive com- 4-aminogalactoside 2 presumably due to transesterification of
pound were unclear.
the aminobutanoyl side chain.
A proposed reaction mechanism for the formation of the
product 5 from 1 under acidic conditions is shown in Fig. 3.
Investigation of the structure of the reaction intermediate
In order to confirm that the observed reaction intermediate The reaction may proceed via an initial intramolecular amino–
could be converted into 2 under acidic conditions (pH 1), a amide interaction12–14 to give a cyclic hemi-orthoamide inter-
small amount of the intermediate (50 μg) was isolated by mediate. Such intermediates have been previously isolated and
HPLC from the reaction of 1 (1 mg) in 10% TFA in H2O after
10 minutes at 80 °C (tR = 3.97 min, m/z 531 [M + H]+). The iso-
lated intermediate was then treated with 10% TFA in H2O
(300 µL, pH 1) for two days at room temperature and the
reaction monitored by LC/MS. LC/MS analysis showed that the
isolated intermediate was slowly converted into 2 at room
temperature under these conditions (tR = 4.88 min, m/z 446
[M + H]+). The intermediate was then isolated on a larger scale
(2 mg) by HPLC from the reaction of 1 (20 mg) with 10% TFA
in H2O (pH 1) after 10 minutes at 80 °C (tR = 4.09, m/z 531
[M + H]+). The sample was dissolved in MeOH-d4 (pH 7) and
analyzed by 1H NMR and COSY spectroscopy. However, the
NMR spectra indicated that the intermediate had been con-
verted back into the starting material 1 under neutral con-
ditions. This was further confirmed by LC/MS analysis of the
NMR sample (tR = 4.97, m/z 531 [M + H]+).
Given that the reaction intermediate was unstable under
Fig. 3 Proposed reaction mechanism for the decomposition of 1. R1
neutral conditions and was difficult to isolate on a large scale
by HPLC, an NMR experiment was performed. A sample of CH3, R2 = COCH2Ar, R3 = CH2CH2NHC(vNH)NH2.
=
4072 | Org. Biomol. Chem., 2015, 13, 4070–4079
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
characterized by NMR spectroscopy15 but are known to be DSL, which showed that both compounds eluted at tR
=
unstable and rapidly decompose with migration of the acyl 6.87 minutes. No polar intermediate was observed during the
group from N to N′.16 Such an acyl migration pathway could reaction of 6, possibly because the rate of reaction was too fast
account for loss of the 4-aminobutanoyl side chain via for- and not detectable by LC/MS. However, for both compounds
mation of butyrolactam. However, to account for formation of 7 and 8 a more polar intermediate with the same mass as the
ester 5 we propose that the hemi-orthoamide undergoes de- starting material was observed after 20 minutes and one hour,
hydration followed by intramolecular nucleophilic attack from respectively (tR = 4.75, m/z = 559 [M + H]+; tR = 3.99, m/z = 511
the C-6 OH and proton transfer to give an intermediate which [M + Na]+), before loss of the aminobutanoyl side chain after
then breaks down with migration of the side chain to O-6. overnight reaction. On the other hand, compound 9 with an
Hydrolysis of this intermediate then generates the ester 5. The aminoacetyl side-chain was completely stable to the reaction
product 2 could then be formed by simple hydrolysis of the conditions. These results suggest that compounds with a
ester or by an intramolecular attack by the γ-amino group to 4-aminobutanoyl side chain undergo a similar decomposition
form butyrolactam and 2. We cannot rule out the possibility of reaction to 1 under acidic conditions, but with slight variations
direct N- to O-acyl transfer by attack of the C6–OH on the in the reaction rates, probably arising from the slight differ-
amide carbonyl to form a cyclic hemi-orthoamide which ences in their structures.
decomposes with transfer of the side chain to C6. Such N- to
O-acyl transfers have been reported for β,γ-bishydroxy amides
Model compound studies
under acidic conditions and have been postulated to proceed
Some simple model compounds (11–13) were next synthesized
via a five-membered ring hydroxyoxazolidine intermediate.17
in order to further investigate the effects of the length of the
Similarly, cyclosporins are known to undergo N- to O-acyl
side chain on stability, and to determine if the C-6 hydroxyl
transfers to form isocyclosporins.18 However, both these
group was critical to the degradation process. The synthesis of
11–13 required the preparation of the common intermediate
amine 19 from methyl 4-O-tosyl-α-D-xylopyranoside 14
(Scheme 1).
examples do not contain an additional amino group on the
acyl side chain while our additional studies of related com-
pounds with different length aminoacyl side chains (see
below) indicate that overall stability is dependent on the
Commercially available methyl α-D-xylopyranoside was
length of the side chain. Unlike the well-known O- to N-acyl
regioselectively tosylated via its stannylene acetal following the
and O- to S-acyl transfers,19–21 N- to O-acyl transfers as
procedure of Kosma22 to give a 3 : 1 mixture of 4-O- and 2-O-
described above are relatively uncommon and presumably the
D-galactoside scaffold positions the C-6 hydroxyl group in a
favourable orientation to facilitate this process.
tosylates from which 4-O-tosylate23 14 was readily isolated by
flash chromatography. Tosylate 14 was then benzoylated with
benzoyl chloride in pyridine at room temperature to give the
dibenzoate24,25 15 in 96% yield. The melting point was slightly
Quality control of similar DSL compounds
lower than previously reported, presumably due to the pres-
ence of trace impurities (visible in the 1H NMR spectrum).
Available compounds from the DSL (6–10, Fig. 4) with similar
structures to 1 were next investigated to determine if a similar
Interestingly, in the 1H NMR spectrum of 15 the signal for H-3
decomposition could occur for these samples. Compounds
6–9 were thus treated with 10% TFA in H2O (pH 1) at 80 °C
overnight and the reactions were analyzed by LC/MS. Com-
pounds 6–8, which contained a 4-aminobutanoyl side chain at
C-4, were all susceptible to decomposition and gave final
products corresponding to loss of the side chain. This was con-
firmed via a co-elution LC/MS experiment of the decompo-
sition product of 6 with compound 10, also available from the
at 5.88 ppm appeared as a second order multiplet (apparent
triplet of triplets) due to virtual coupling26 to H-1 (δ H-1–δ H-2
= 5.0 Hz; J1,2 = 3.5 Hz). The tosylate was then displaced with
inversion of configuration by heating with sodium azide in
DMF to give the azide24 16 in good yield (81%). Transesterifica-
tion of the benzoate groups gave the diol24 17 in 88% yield
which was then methylated with sodium hydride and methyl
iodide in anhydrous DMF to give 18 in 93% yield. Reduction of
the azido group was accomplished by hydrogenation in MeOH
Scheme 1 Reaction scheme for the synthesis of model compounds
11–13. Reagents and conditions: (a) BzCl, py, 0 °C → r.t., 18 h, 96%; (b)
NaN3, DMF, 90 °C, 48 h, 81%; (c) NaOMe, MeOH, 88%; (d) NaH, MeI,
DMF, 93%; (e) H2, 10% Pd/C, MeOH, 91%; (f) HBTU, DIPEA, DMF, Boc- or
Fmoc-protected amino acid, 13–32%; (g) 1 : 2 : 7 TFA–Et3SiH–CH2Cl2
(for Boc); and ethylenediamine resin–piperidine–DMF (for Fmoc),
91–95%.
Fig. 4 DSL compounds 6–10 with similar structures to
COCH2Ar, where Ar is 1-naphthyl for 6, 7 and 9, and 2-naphthyl for 8),
and model compounds 11–13.
1 (R =
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4070–4079 | 4073

View Article Online
Paper
Organic & Biomolecular Chemistry
with 10% Pd/C as catalyst to furnish the amine 19 in 91% 2. Selected compounds from the DSL with similar structures to
1
yield. The structure of 19 was supported by H NMR spectro- 1 are also capable of undergoing a similar decomposition reac-
scopy (H-4 was shifted significantly upfield to 3.30 ppm) and tion under acidic conditions with formation of an intermedi-
LRMS (m/z 192.0 [M + H]+), however, 19 was somewhat ate. Furthermore, three model compounds 11–13 with
unstable, and so was reacted immediately in the next step different length side chains at C-4 and without a C-6 hydroxyl
without further purification or characterization. Amine 19 was group were also decomposed with different reaction rates
coupled with (Boc-γ-amino)butyric acid, Fmoc-β-Ala-OH and under acidic conditions. It can be concluded that the stability
Boc-Gly-OH, respectively, using HBTU and Hünig’s base in 4-aminoacyl-D-galactoside derivatives is related to the length of
anhydrous DMF at room temperature to give the carbamates the side chain with a 4-aminobutanoyl side chain being the
20–22. The carbamates were obtained in unoptimised yields of most labile. In addition, the presence of a free hydroxyl group
13, 32 and 19%, respectively, following purification by flash at C-6 appears to accelerate the rate of decomposition and
chromatography, and were fully characterized. The target results in an unusual N- to O-acyl transfer to give an ester inter-
amines 11–13 were then obtained in excellent yield (91–95%) mediate. The nature of the bioactive species in the drug screen-
by removal of the Boc and Fmoc protecting groups via treat- ing assay is still unclear and remains of ongoing interest.
ment with 1 : 2 : 7 trifluoroacetic acid–triethylsilane–dry
dichloromethane (for Boc) and polymer supported ethylene-
diamine resin and piperidine in anhydrous DMF (for Fmoc),
respectively.
Experimental
General
Stability of model compounds under acidic conditions
Compound 1 and similar compounds from the DSL were
obtained from Alchemia Ltd. All reagents and solvents were
obtained from Sigma-Aldrich (Australia) and were used
without further purification, except MeOH, dichloromethane,
EtOAc, hexane and EtOH which were distilled prior to use. 1D
and 2D NMR spectra were recorded on a Bruker spectrometer
(1H NMR, COSY, HSQC and HMBC at 500 MHz, 13C NMR at
100 MHz). The residual solvent peaks served as internal stan-
dard (CDCl3: δ H 7.24 and δ C 77.0; DMSO-d6: δ H 2.49 and δ C
39.5; MeOD: δ H 4.78 and δ C 49.0). Melting points were deter-
mined on a DigiMelt MSRS apparatus. High resolution mass
spectra were recorded on a Bruker micrOTOFQ spectrometer in
positive ion ESI mode. Chromatographic separations were
carried out on silica gel (0.04–0.06 mm, 230–400 mesh, Merck)
or a reverse phase SPE C18 cartridge. Reactions were monitored
by reverse phase LC/MS or TLC, as appropriate. Reverse phase
LC/MS was used under the following conditions: column:
Zorbax SB-C18, flow rate: 2 mL min−1, time: 12 min, solvent A:
0.1% HCOOH in H2O, solvent B: 0.1% HCOOH in MeCN. TLCs
were visualized by staining/charring with ninhydrin (5% nin-
hydrin in ethanol), KMnO4 (1.5 g KMnO4, 10 g K2CO3 and
1.25 mL 10% NaOH in 200 mL water), or 5% sulfuric acid in
ethanol.
The model compounds 11–13 were treated with 10% TFA in
H2O (pH 1) at 80 °C overnight to investigate if their stability
was influenced by the length of the side chain at C-4 through
an intramolecular amino–amide interaction. Unfortunately,
the compounds were too polar for satisfactory analysis by
LC/MS, so reaction progress was analyzed by TLC (tert-butanol–
acetic acid–water, 2 : 1 : 1)27 and 1H NMR and COSY spec-
troscopy. No intermediates were observed during the reaction
via these techniques, only starting materials and products. In
addition, the reaction rates were much slower than for the DSL
compounds, with starting material still present after overnight
reaction. 1H NMR spectroscopic analysis after overnight reac-
tion confirmed the major decomposition product to be amine
19 in all cases but the reaction rate for each compound was
different. By comparison of the integral of the peaks for H-4
for the starting material and for the product amine 19 (at
3.64 ppm), it was possible to establish the relative stability of
the three model compounds. The relative ratio of starting
material to 19 increased with increasing chain length from
1 : 0.4 for 11, to 1 : 0.7 for 12 and 1 : 0.8 for 13. Whilst no inter-
mediates were detected, we hypothesize that compounds
12 and 13 are more labile because they can form cyclic hemi-
orthoamide intermediates15 via intramolecular amino–amide
interactions, whilst 11 cannot. The results also suggest that
decomposition is less facile without the presence of a free
hydroxyl group at C-6 to form an ester intermediate, but
further studies are needed to confirm this.
Re-synthesis of the decomposition product (2)
Compound 1 (20 mg, 40 μmol) was dissolved in a mixture of
10% TFA in H2O (36 mL, pH 1) and stirred at 80 °C for 20 h.
The reaction was monitored by TLC (n-butanol–H2O–HOAc,
26/25/6)27 and LC/MS. After the reaction was complete, the
resulting mixture was diluted with H2O (50 mL), concentrated
and co-evaporated with toluene (5 mL) to give the crude
product. Purification by reverse phase SPE C18 cartridge (5%
Conclusions
In conclusion, the decomposition of compound 1 occurred i-PrOH–H2O + 0.1% HCOOH → 100% i-PrOH–H2O + 0.1%
during long term storage in DMSO (10 mM) to form the HCOOH) gave compound 2 (16 mg, 94%) as white solid. Rf =
4-aminogalactoside 2, most likely via a mechanism involving 0.61 (n-butanol–H2O–HOAc, 26/25/6); LC/MS: tR = 4.91 min,
migration of the side chain from C-4 to C-6 to form an ester m/z = 446.58 [M + H]+, 468.47 [M + Na]+; 1H and 13C NMR data
intermediate (5) followed by hydrolysis to furnish compound are given in Tables S3–4.†
4074 | Org. Biomol. Chem., 2015, 13, 4070–4079
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Acetylation of compound 2. To a solution of compound 2 vacuum. It was then subjected to analysis by 1H NMR and
(8 mg, 18 μmol) in pyridine (0.5 mL) was added Ac2O (47 µL, COSY spectroscopy. However, it was found that the product
0.50 mmol) and a catalytic amount of DMAP. The reaction had been converted back into 1 under neutral conditions at
mixture was sonicated at 50 °C under N2 for 1.5 h. The reaction room temperature, as judged by 1H NMR, COSY and LC/MS
was monitored by LC/MS. The reaction mixture was cooled which were identical in all respects to that of authentic 1.
(0 °C) and anhydrous MeOH (2 mL) was added to remove the
General experimental procedure for LC/MS experiments for
the quality control of similar DSL compounds
excess Ac2O. The solution was stirred at 0 °C for 1 h and was
then concentrated and co-evaporated with toluene (3 × 1 mL)
to give the crude product as a light yellow solid. Purification by DSL compounds 6–10 were first analysed by LC/MS. Samples
reverse phase SPE C18 cartridge (5% i-PrOH–H2O + 0.1% (0.15 mg) were dissolved in a mixture of 10% TFA in H2O
HCOOH → 100% i-PrOH–H2O + 0.1% HCOOH) gave the di- (300 μL, pH 1). The resulting solutions were stirred at 80 °C
acetate 3 (2 mg, 22%) as white solid. LC/MS: tR = 5.73 min, m/z overnight. The reaction progress was examined by LC/MS.
1
= 530.62 [M + H]+. H NMR (500 MHz, MeOH-d4): δ 7.96 (d, J =
6: tR = 6.58 min, m/z = 552.57 [M + H]+, 574.60 [M + Na]+; 7:
8.0 Hz, 1H, H-24), 7.81 (m, 1H, H-21), 7.74 (m, 1H, H-19), tR = 5.21 min, m/z = 559.68 [M + H]+; 8: tR = 4.50 min, m/z =
7.46–7.41 (m, 4H, H-23, H-22, H-18, H-17), 4.78 (1H, H-1, 489.65 [M + H]+, 511.61 [M + Na]+; 9: tR = 5.15 min, m/z =
overlap with MeOD peak), 4.61 (dd, J = 1.5, 4.0 Hz, 1H, H-4), 503.52 [M + H]+, 525.41 [M + Na]+; 10: tR = 6.55 min, m/z =
4.16 (dd, J = 4.0, 11.5 Hz, 1H, H-2), 4.06–3.94 (m, 5H, H-5, H-6, 467.48 [M + H]+. 489.51 [M + Na]+.
H-15), 3.61 (m, 1H, H-7a), 3.50 (dd, J = 4.5, 11.5 Hz, 1H, H-3),
3.44 (m, 1H, H-7b), 3.29 (m, 1H, H-8a), 3.25 (s, 3H, OCH3), Methyl α-D-xylopyranoside was tosylated following the method
3.18 (m, 1H, H-8b), 1.95 (s, 3H, CH3), 1.93 (s, 3H, CH3).
of Kosma22 to give a mixture of tosylates (1.0 g). Purification by
Methyl
4-O-p-toluenesulfonyl-α-D-xylopyranoside
(14).
Synthesis of carbamate 4. To a mixture of 2 (5 mg, 11 μmol) flash chromatography (EtOAc–toluene, 3 : 2) gave the 4-O-tosy-
and (Boc)2O (3.1 µL, 13 μmol, 1.2 eq.) in anhydrous DMF late 14 (745 mg, 75%) and the 2-O-tosylate (250 mg, 25%). 4-O-
(1 mL) was added Et3N to adjust the pH to 10. The mixture was Tosylate 14: Rf = 0.28 (EtOAc–toluene, 3 : 2), colourless needles,
sonicated at room temperature for 65 min, diluted with m.p. 50–55 °C (EtOAc–hexane; lit.28 59–60 °C (benzene); lit.29
toluene and evaporated under reduced pressure to remove 57–58 °C (benzene)); 1H NMR (500 MHz, CDCl3): δ 7.84–7.78
DMF. The crude product was purified by reverse phase SPE C18 (m, 2H, Ar), 7.35–7.32 (m, 2H, Ar), 4.67 (d, 1H, J1,2 = 4.0 Hz,
cartridge (30% i-PrOH–H2O + 0.1% HCOOH → 100% i-PrOH + H-1), 4.32 (ddd, 1H, J3,4 = 9.0 Hz, J4,5a = 5.8 Hz, J4,5b = 10.5 Hz,
0.1% HCOOH) to give the carbamate 4 (3 mg, 50%) as white H-4), 3.80 (dd, 1H, J2,3 = 9.2 Hz, H-3), 3.68 (dd, 1H, J5a,5b
=
solid. LC/MS: tR = 6.66 min, m/z = 546.59 [M + H]+. H NMR 11.2 Hz, H-5a), 3.58 (dd, 1H, H-5b), 3.46 (dd, 1H, H-2), 3.38 (s,
(500 MHz, MeOH-d4): δ 7.97 (d, J = 8.5 Hz, 1H, H-24), 7.81 (d, 3H, OCH3), 2.43 (s, 3H, ArCH3); 13C NMR (100 MHz, CDCl3):
J = 8.0 Hz, 1H, H-21), 7.74 (dd, J = 2.0, 7.5 Hz, 1H, H-19), δ 145.3, 133.0, 129.9, 128.0 (Ar), 98.8 (C-1), 77.9 (C-4), 72.3
7.48–7.36 (m, 4H, H-23, H-22, H-18, H-17), 4.74 (d, J = 4.0 Hz, (C-2), 72.0 (C-3), 58.8 (C-5), 55.6 (OCH3), 21.7 (ArCH3). 2-O-
1H, H-1), 4.18 (d, J = 3.0 Hz, 1H, H-4), 4.08 (dd, J = 4.0, Tosylate: Rf = 0.34 (EtOAc–toluene, 3 : 2), colourless needles
11.5 Hz, 1H, H-2), 3.97 (s, 2H, H-15), 3.82 (t, J = 5.0 Hz, 1H, m.p. 141–142 °C (EtOAc–hexane; lit.23 140–141 °C). 1H NMR
H-5), 3.61–3.57 (m, 1H, H-7a), 3.51–3.50 (m, 2H, H-6), 3.45 (dd, (500 MHz, CDCl3): δ 7.83–7.78 (m, 2H, Ar), 7.36–7.32 (m, 2H,
J = 4.0, 11.5 Hz, 1H, H-3), 3.40–3.36 (m, 1H, H-7b), 3.28–3.27 Ar), 4.62 (d, 1H, J1,2 = 3.5 Hz, H-1), 4.24 (dd, 1H, J2,3 = 9.5 Hz,
(m, 1H, H-8a), 3.25 (s, 3H, OCH3), 3.20–3.16 (m, 1H, H-8b), H-2), 3.87 (ddd, 1H, J3,OH = 3.5 Hz, J3,4 = 9.5 Hz, H-3),
1
1.37 (s, 9H, CH3).
3.66–3.61 (m, 2H, H-4, H-5a), 3.47 (dd, 1H, J4,5b = J5a,5b = 11.5
Purification of 2 from DSL. Resin-bound 2 was treated with Hz, H-5b), 3.26 (s, 3H, OCH3), 2.68 (d, 1H, 3-OH), 2.52 (d, 1H,
cleavage solution (TFA–TES–DCM, 1 : 2 : 7, 3 mL) under N2 for J4,OH = 3.0 Hz, 4-OH), 2.44 (s, 3H, ArCH3); 13C NMR (100 MHz,
3 h at room temperature. The cleavage solution was then CDCl3): δ 145.3, 133.9, 129.9, 128.0 (Ar), 97.4 (C-1), 79.5 (C-2),
drained under N2. The resin was then washed with MeCN (4 × 71.6 (C-3), 70.0 (C-4), 60.6 (C-5), 55.4 (OCH3), 21.7 (ArCH3).
4 mL), a solution of MeCN with 10% H2O (5 mL), and co-evap-
Methyl 2,3-di-O-benzoyl-4-O-p-toluenesulfonyl-α-D-xylopyra-
orated with toluene (2 × 2 mL) to afford the crude cleavage noside (15). Benzoyl chloride (0.61 mL, 5.28 mmol) was added
product. It was then purified by reversed phase SPE C18 dropwise to a solution of the 4-O-tosylate 14 (700 mg,
cartridge (5% i-PrOH–H2O + 0.1% HCOOH → 100% i-PrOH + 2.20 mmol) in anhydrous pyridine (8 mL) at 0 °C under N2. The
0.1% HCOOH) to give the product 2 (2 mg) as white solid. colourless solution turned yellow and a colourless solid precipi-
LC/MS: tR = 4.76, m/z = 446.65 [M + H]+. The 1H NMR and tated. The ice bath was removed and the reaction mixture was
COSY spectra of the product were identical in all respects to stirred at room temperature for 18 h. The resulting mixture was
that prepared by decomposition of compound 1.
dissolved in CHCl3 (50 mL) and washed with H2O (50 mL). The
Attempted isolation of the reaction intermediate. Com- aqueous layer was extracted with CHCl3 (40 mL), the organic
pound 1 (20 mg, 40 μmol) was dissolved in a mixture of 10% layers were combined, washed with aq 0.1 M HCl (50 mL) and
TFA in H2O (36 mL, pH1) and stirred at 80 °C for 10 min. The sat. aq NaHCO3 (50 mL), and dried (MgSO4). The solution was
reaction was stopped by cooling the reaction mixture (dry ice– concentrated and the solid residue was crystallized from
acetone bath). The solution was then purified by HPLC to give n-hexane–EtOAc to give the dibenzoate 15 as colourless
the intermediate as white solid (2 mg) after drying under high needles (1.11 g, 96%), Rf = 0.60 (toluene–EtOAc, 4 : 1), m.p.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4070–4079 | 4075

View Article Online
Paper
Organic & Biomolecular Chemistry
177.5–178.6 °C (lit.24 180–183 °C; lit.25 185–186 °C); The (s, 1H, OH); 13C NMR (100 MHz, CDCl3): δ 99.6 (C-1), 70.7
observed melting point was low due to the presence of traces (C-3), 70.0 (C-2), 61.1 (C-4), 60.4 (C-5), 56.8 (OCH3). LRMS: m/z
of impurities (∼3%), as seen in the 1H NMR spectrum. 1H = 212.0 [M + Na]+; HRMS: m/z calcd for C6H11N3O4Na [M +
NMR (500 MHz, CDCl3): δ 7.88–7.85 (m, 2H, Ar), 7.68–7.65 (m, Na]+: 212.0642; found: 212.0647.
2H, Ar), 7.62–7.57 (m, 2H, Ar), 7.49–7.44 (m, 2H, Ar), 7.33–7.27
Methyl 4-azido-4-deoxy-2,3-di-O-methyl-β-L-arabinopyrano-
(m, 4H, Ar), 6.93–6.90 (m, 2 H, Ar), 5.88 (app. tt, 1H, H-3), 5.03 side (18). Diol 17 (280 mg, 1.48 mmol) was dissolved in anhy-
(dd, 1H, J1,2 = 3.5 Hz, J2,3 = 9 Hz, H-2), 5.03–5.01 (m, 1H, H-1), drous DMF (15 mL), and NaH (8.4 eq., 0.30 g, 12.4 mmol,
4.66 (ddd, 1H, J3,4 = 9.4 Hz, J4,5a = 6.0 Hz, J4,5b 10.7 Hz, H-4), prewashed with hexanes) was added protionwise to the solu-
4.04 (dd, 1H, J5a,5b = 11.2 Hz, H-5a), 3.89 (dd, 1H, H-5b), 3.39 tion at room temperature. The resulting solution was stirred at
(s, 3H, OCH3), 2.15 (s, 3H, ArCH3); 13C NMR (100 MHz, room temperature under N2 for 50 min. MeI (4.8 eq., 0.44 mL,
CDCl3): δ 165.7, 164.9 (CvO), 133.4, 133.0, 132.8, 129.9, 129.7, 7.10 mmol) was then added dropwise and the mixture stirred
129.0, 128.8, 128.4, 128.1, 127.6 (C–Ar), 96.8 (C-1), 75.5 (C-4), for 2.5 h under N2 at r.t. with protection from light. MeOH was
71.7 (C-2), 69.2 (C-3), 59.2 (C-5), 55.6 (OCH3), 21.6 (ArCH3).
added dropwise to decompose excess NaH. The resulting
Methyl 4-azido-2,3-di-O-benzoyl-4-deoxy-β-L-arabinopyrano- mixture was diluted with EtOAc (50 mL) and washed with H2O
side (16). Dibenzoate 15 (1.09 g, 2.07 mmol) was dissolved in (3 × 100 mL). The combined organic layers were washed with
anhydrous DMF (8 mL) and NaN3 (5 eq., 10.4 mmol, 673 mg) 1 M HCl (100 mL), brine (100 mL), dried (MgSO4) and concen-
was added. The reaction mixture was then stirred at 90 °C for trated to give the dimethyl ether 18 as a slightly yellow solid
48 h under N2. The resulting mixture was dissolved in CHCl3 (300 mg, 93%); Rf = 0.84 (CHCl3–MeOH, 9 : 1); m.p. 230–234 °C
(100 mL) and washed with H2O (3 × 100 mL). The combined (dec.; EtOAc–hexane); [α]2D0 +5.6 (c 0.8, CHCl3); 1H NMR
organic layers were washed with sat. aq. NaHCO3 (100 mL) and (500 MHz, CDCl3): δ 4.82 (d, 1H, J1,2 = 3.5 Hz, H-1), 3.99–3.97
brine (100 mL), dried (MgSO4) and concentrated. The residual (m, 1H, H-4), 3.76 (dd, 1H, J4,5a = 1.8 Hz, J5a,5b = 12.5 Hz,
oil was dissolved in EtOAc (3 mL) and n-hexane was added H-5a), 3.69 (dd, 1H, J2,3 = 9.6 Hz, J3,4 = 3.8 Hz, H-3), 3.63 (dd,
until a colourless solid precipitated. The crystals were collected 1H, J4,5b = 2.2 Hz, H-5b), 3.56 (dd, 1H, H-2), 3.51 (s, 3H,
and the filtrate was evaporated to dryness and to afford an 2-OCH3), 3.50 (s, 3H, 3-OCH3), 3.40 (s, 3H, 1-OCH3); 13C NMR
additional product as a yellow solid (less pure product). The (100 MHz, CDCl3): δ 98.4 (C-1), 78.8 (C-3), 77.8 (C-2), 60.2
total yield of the product was 663 mg (81%); Rf = 0.65 (C-5), 59.3 (2-OCH3), 59.0 (C-4), 58.2 (3-OCH3), 55.6 (1-OCH3);
(toluene–EtOAc, 5 : 1). β-L-arabinoside 16: colourless needles, LRMS: m/z
m.p. 102–104 °C (EtOAc–hexane; lit.24 104–105.5 °C); C8H15N3O4Na [M + Na]+: 240.0955; found: 240.0944.
[α]2D0 +84.5 (c 0.49, CHCl3); FTIR: 2112 (N3), 1723 (CvO). 1H
Methyl 4-deoxy-4-(2′-tert-butoxycarbonylamino)acetamido-
= 240.0 [M +
Na]+; HRMS: m/z calcd for
NMR (500 MHz, CDCl3): δ 8.02–7.95 (m, 4H, Ph), 7.53–7.48 (m, 2,3-di-O-methyl-β-L-arabinopyranoside (20). The dimethyl
2H, Ph), 7.39–7.34 (m, 4H, Ph), 5.84 (dd, 1H, J2,3 = 10.5 Hz, J3,4 ether 18 (50 mg, 0.23 mmol) was dissolved in anhydrous
= 4.0 Hz, H-3), 5.56 (dd, 1H, J1,2 = 3.5 Hz, H-2), 5.12 (d, 1H, MeOH (1 mL), and Pd/C (30 mg) was added. The mixture was
H-1), 4.26–4.24 (m, 1H, H-4), 4.05 (dd, 1H, J4,5a = 1.0 Hz, J5a,5b stirred at room temperature under an atmosphere of H2 for
= 12.5 Hz, H-5a), 3.77 (dd, J4,5b = 2.0 Hz, 1H, H-5b), 3.41 (s, 3H, 19.5 h. The catalyst was then removed by filtration (celite), and
OCH3); 13C NMR (100 MHz, CDCl3): δ 165.9, 165.8 (CvO), the filtrate was concentrated to give the amine 19 (40 mg,
133.5, 133.3, 129.9, 129.8, 129.2, 128.8, 128.5, 128.4 (Ph), 97.9 91%) as light yellow solid; LRMS: m/z = 192.0 [M + H]+; 1H
(C-1), 69.6 (C-3), 69.1 (C-2), 60.4 (C-4), 60.1 (C-5), 55.7 (OCH3).
NMR (500 MHz, CDCl3): δ 4.81 (d, 1H, J1,2 = 3.3 Hz, H-1), 3.80
Methyl 4-azido-4-deoxy-β-L-arabinopyranoside (17). The (dd, 1H, J4,5a = 2.2 Hz, J5a,5b = 12.0 Hz, H-5a), 3.56 (dd, 1H, J4,5b
azide 16 (663 mg, 1.67 mmol) was suspended in anhydrous = 2.3 Hz, H-5b), 3.48 (s, 3H, 2-OCH3), 3.47 (m, 1H, H-2), 3.44
MeOH (12 mL), and a 1 M solution of NaOMe (3 mL) was (s, 3H, 3-OCH3), 3.44 (m, 1H, H-3), 3.40 (s, 3H, 1-OCH3), 3.30
added. The reaction mixture was then stirred at room tempera- (m, 1H, H-4). A mixture of Boc–Gly–OH (40 mg, 0.21 mmol),
ture for 1 h (pH 10). The resulting mixture was neutralized by HBTU (0.42 mmol, 159 mg) and DIPEA (0.21 mmol, 37 µL) was
addition of strongly acidic ion exchange resin (Amberlite stirred at 0 °C for 5 minutes in anhydrous DMF (1 mL, dried
IR-120, H+ form, pre-washed with anhydrous MeOH, 3 × over 3 Å molecular sieves). The resulting solution was then
20 mL) in small portions to pH 7, and straight away filtered added to a solution of amine 19 (40 mg, 0.21 mmol) and
through a Büchner-funnel. The resin was washed with anhy- DIPEA (0.21 mmol, 37 µL) in anhydrous DMF (1 mL). The reac-
drous MeOH (5 × 20 mL), and the combined filtrate and wash- tion mixture was stirred at 0 °C for 1 hour, and stirred at room
ings were concentrated in vacuo to give the crude product as a temperature for 24 hours. The resulting mixture was then
yellow solid. Flash chromatography (CHCl3–MeOH, 4 : 1 → 1 : 1 diluted with EtOAc (50 mL), and the mixture was washed with
→ 0 : 10) gave the diol 17 (280 mg, 88%) as colourless needles, 1 M HCl (2 × 30 mL), saturated NaHCO3 (2 × 30 mL) and satu-
m.p. 86–88 °C (CHCl3–hexane; lit.24 86–88 °C; lit.30 81–83 °C); rated NaCl (2 × 30 mL). The organic layer was dried (MgSO4)
[α]2D0 +193 (c 0.53, CHCl3); Rf = 0.68 (CHCl3–MeOH, 1 : 1). 1H and concentrated to afford a yellow solid. The crude product
NMR (500 MHz, CDCl3): δ 4.77 (d, 1H, J1,2 = 4.0 Hz, H-1), was then purified by flash column chromatography (CHCl3–
3.95–3.90 (m, 2H, H-3, H-4), 3.81 (dd, 1H, J4,5a = 1.0 Hz, J5a,5b
=
MeOH, 9 : 1) to give the carbamate 20 (15 mg, 19%) as colour-
12.5 Hz, H-5a), 3.78 (dd, 1H, J2,3 = 9.0 Hz, H-2), 3.68 (dd, 1H, less oil; Rf = 0.48 (CHCl3–MeOH, 9 : 1). 1H NMR (500 MHz,
J4,5b = 2.5 Hz, H-5b), 3.41 (s, 3H, OCH3), 2.47 (s, 1H, OH), 2.02 CDCl3): δ 6.41 (d, 1H, J = 6.3 Hz, NHCvO), 5.12 (br s, 1H,
4076 | Org. Biomol. Chem., 2015, 13, 4070–4079
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
NHCO2), 4.79 (d, 1H, J1,2 = 3.5 Hz, H-1), 4.40 (br s, 1H, H-4), 30 mL) and saturated NaCl (2 × 30 mL). The organic layer was
3.80–3.76 (m, 3H, CH2, H-5a), 3.66 (dd, 1H, J4,5b = 2.1 Hz, J5a,5b dried (MgSO4) and concentrated to give a yellow solid. The
= 12.2 Hz, H-5b), 3.62 (dd, J3,4 = 4.6 Hz, J2,3 = 9.4 Hz, 1H, H-3), crude product was then purified by flash chromatography
3.49 (s, 3H, 2-OCH3), 3.41 (s, 3H, 1-OCH3), 3.39 (s, 3H, 3- (CHCl3–MeOH, 9 : 1) to give the carbamate 22 (11 mg, 13%) as
OCH3), 3.26 (dd, 1H, H-2), 1.44 (s, 9H, Me3); 13C NMR colourless oil; Rf = 0.52 (CHCl3–MeOH, 9 : 1). 1H NMR
(100 MHz, CDCl3): δ 169.8 (NHCvO), 156.1 (NHCO2), 98.1 (500 MHz, CDCl3): δ 6.69 (br s, 1H, NHCvO), 4.82 (d, 1H, J1,2
=
(C-1), 78.4 (Me3C), 77.2 (C-2), 76.8 (C-3), 60.5 (C-5), 59.2 3.6 Hz, H-1), 4.74 (br s, 1H, NHCO2), 4.48–4.45 (m, 1H, H-4),
(2-OCH3), 57.2 (3-OCH3), 55.6 (1-OCH3), 46.8 (C-4), 44.6 (CH2), 3.78 (dd, 1H, J4,5a = 2.2 Hz, J5a,5b = 12.1 Hz, H-5a), 3.65–3.61
28.3 (CH3). LRMS: m/z = 349.1 [M + H]+, 371.0 [M + Na]+, 387.0 (m, 2H, H-5b, H-3), 3.50 (s, 3H, 2-OCH3), 3.49–3.45 (m, 1H,
[M + K]+; HRMS: m/z calcd for C15H28N2O7Na [M + Na]+: H-2), 3.41 (s, 3H, 1-OCH3), 3.38 (s, 3H, 3-OCH3), 3.26–3.14 (m,
371.1789; found: 371.1793.
Methyl
2H, CH2NHCO2), 2.26 (t, 2H, J = 7.0 Hz, NHCOCH2), 1.79
4-deoxy-4[3′-(fluoren-9-yl)methoxycarbonylamino]- (quintet, 2H, J = 6.9 Hz, CH2CH2NHCO2), 1.43 (s, 9H, Me3); 13
C
propanamido-2,3-di-O-methyl-β-L-arabinopyranoside
(21). A NMR (100 MHz, CDCl3): δ 172.9 (NHCvO), 156.7 (NHCO2),
mixture of Fmoc–β-Ala–OH (65 mg, 0.21 mmol), HBTU 98.1 (C-1), 79.4 (Me3C), 77.2 (C-2), 76.9 (C-3), 61.0 (C-5), 59.0
(0.42 mmol, 159 mg) and DIPEA (0.21 mmol, 37 µL) was (2–OCH3), 57.0 (1–OCH3), 55.5 (3–OCH3), 46.7 (C-4), 39.3
stirred at 0 °C for 5 min in anhydrous DMF (1 mL, dried over (CH2NHCO2),
33.7
(NHCOCH2),
28.4
(CH3),
27.0
3 Å molecular sieves). The resulting solution was then added (CH2CH2NHCO2). LRMS: m/z = 399.1 [M + Na]+, 775.2 [2M +
to a solution of amine 19 (40 mg, 0.21 mmol) and DIPEA Na]+; HRMS: m/z calcd for C17H32N2O7Na [M + Na]+: 399.2101;
(0.21 mmol, 37 µL) in anhydrous DMF (1 mL) at 0 °C. The reac- found: 399.2115.
tion mixture was stirred at 0 °C for 1 h, and then at room
Methyl 4-(2′-amino)acetamido-4-deoxy-2,3-di-O-methyl-β-L-
temperature for 22 h. The resulting mixture was then diluted arabinopyranoside (11). The carbamate 20 (15 mg,
with EtOAc (50 mL), and the mixture was washed with 1 M HCl 0.043 mmol) was treated with cleavage solution (TFA–TES–dry
(2 × 30 mL), saturated NaHCO3 (2 × 30 mL) and saturated NaCl DCM, 1 : 2 : 7, 0.5 mL) for 4 h at room temperature. The clea-
(2 × 30 mL). The organic layer was dried (MgSO4) and concen- vage solution was then dried under N2, and co-evaporated with
trated to give a yellow solid. The crude product was then puri- a mixture of toluene–i-PrOH (10 : 1) (5 × 5 mL) to afford the
fied by flash column chromatography (CHCl3–MeOH, 9 : 1) to crude product. The crude amine was then purified using a
give the carbamate 21 (35 mg, 32%) as colourless oil; Rf = 0.52 reverse phase SPE C18 cartridge (5 g; 10% ACN in H2O, 0.1%
(CHCl3–MeOH, 9 : 1). 1H NMR (500 MHz, CDCl3): δ 7.74 (d, HCOOH → 100% ACN, 0.1% HCOOH) to give the amine 11 as
2H, J = 7.6 Hz, ArH), 7.56 (d, 2H, J = 7.5 Hz, ArH), 7.37 (d, 2H, a colourless oil (10 mg, 93%). LCMS: tR = 1.10 min, m/z =
J = 7.4 Hz, ArH), 7.29 (d, 2H, J = 7.4 Hz, ArH), 5.93 (d, 1H, J = 271.39 [M + Na]+; LRMS: m/z = 249.1 [M + H]+, 271.0 [M + Na]+;
7.4 Hz, NHCvO), 5.54 (br s, 1H, NHCO2), 4.78 (d, 1H, J1,2
=
1H NMR (500 MHz, MeOD): δ 4.78 (1H, H-1, overlapped with
2.8 Hz, H-1), 4.45–4.44 (m, 1H, H-4), 4.39–4.30 (m, 2H, MeOD solvent peak), 4.43–4.41 (m, 1H, H-4), 3.76 (dd, 1H, J4,5a
NHCO2CH2), 4.18 (t, 1H, J = 6.95 Hz, CH–Fmoc), 3.79 (dd, 1H, = 2.0 Hz, J5a,5b = 12.1 Hz, H-5a), 3.65 (s, 2H, CH2), 3.55 (dd,
J4,5a = 2.3 Hz, J5a,5b = 12.2 Hz, H-5a), 3.63 (dd, 1H, J4,5b
=
1H, J2,3 = 9.6 Hz, J3,4 = 4.5 Hz, H-3), 3.46 (dd, 1H, J4,5b = 2.7 Hz,
2.5 Hz, H-5b), 3.61 (dd, 1H, J2,3 = 9.5 Hz, J3,4 = 4.7 Hz, H-3), H-5b), 3.40 (s, 3H, 3-OCH3), 3.37 (dd, 1H, J1,2 = 3.6 Hz, H-2),
3.52–3.49 (m, 2H, CH2NHCO2), 3.45 (s, 3H, 2-OCH3), 3.40 (s, 3.32 (s, 6H, 3-OCH3, 1-OCH3); 13C NMR (100 MHz, CDCl3):
3H, 1-OCH3), 3.39 (s, 3H, 3-OCH3), 3.26 (dd, 1H, H-2), 2.46 (t, δ 167.4 (CvO), 99.5 (C-1), 78.5 (C-2), 78.4 (C-3), 62.0 (C-5), 59.1
J = 5.5 Hz, 2H, NHCOCH2); 13C NMR (100 MHz, CDCl3): (2-OCH3), 57.8 (3-OCH3), 55.8 (1-OCH3), 48.4 (C-4), 41.5 (CH2).
δ 171.6 (NHCvO), 156.5 (NHCO2), 143.9, 141.3, 127.7, 127.0,
Methyl 4-(3′-amino)propanamido-4-deoxy-2,3-di-O-methyl-
125.1, 120.0 (Ar), 98.1 (C-1), 77.2 (C-2), 76.9 (C-3), 66.8 β-L-arabinopyranoside (12). The carbamate 21 (17.5 mg,
(NHCO2CH2), 60.6 (C-5), 59.1 (2-OCH3), 57.4 (3-OCH3), 55.6 (1- 0.036 mmol) was dissolved in anhydrous DMF (2 mL), and
OCH3), 47.2 (CH–Fmoc), 46.8 (C-4), 37.3 (CH2NHCO2), 36.4 PL-EDA resin (10 eq., 0.36 mmol, 72 mg) was added. The
(NHCOCH2). LRMS: m/z = 485.1 [M + H]+, 507.1 [M + Na]+, resulting mixture was stirred at 50 °C under N2 for 10 minutes;
523.1 [M + K]+, 544.1 [M + ACN + NH4]+; HRMS: m/z calcd for it was then covered with aluminium foil and shaken for
C26H32N2O7Na [M + Na]+: 507.2101; found: 507.2105.
further 4 h at room temperature. Piperidine (50 µL) was added
Methyl 4-deoxy-4-(4′-tert-butoxycarbonylamino)butanamido- and the resulting mixture was shaken at 4 °C overnight (∼17 h)
2,3-di-O-methyl-β-L-arabinopyranoside (22). A mixture of (Boc- in a cold room. The resin was then filtered off and washed
γ-amino)butyric acid (43 mg, 0.21 mmol), HBTU (0.42 mmol, with MeOH (5 × 20 mL) and dried. It was co-evaporated with
159 mg) and DIPEA (0.21 mmol, 37 µL) was stirred at 0 °C for toluene (5 × 20 mL) at room temperature to afford a white
5 minutes in anhydrous DMF (1 mL, dried over 3 Å molecular solid which was purified on a reverse phase SPE C18 cartridge
sieves). The resulting solution was then added to a solution of (5 g; 10% i-PrOH in H2O, 0.1% HCOOH → 100% i-PrOH, 0.1%
amine 19 (40 mg, 0.21 mmol) and DIPEA (0.21 mmol, 37 µL) HCOOH) to give the amine 12 as a colourless oil (9 mg, 95%);
in anhydrous DMF (1 mL). The reaction mixture was stirred at LC/MS: tR = 1.20 min, m/z = 285.33 [M + Na]+; LRMS: m/z =
1
0 °C for 1 h, then at room temperature for 5 h. The resulting 285.2 [M + Na]+, 263.2 [M + H]+. H NMR (500 MHz, MeOD):
mixture was then diluted with EtOAc (50 mL), and the mixture δ 4.82 (1H, H-1, overlap with MeOD peak), 4.40 (br s, 1H, H-4),
was washed with 1 M HCl (2 × 30 mL), saturated NaHCO3 (2 × 3.74 (dd, 1H, J4,5a = 1.5 Hz, J5a,5b = 11.5 Hz, H-5a), 3.53 (dd,
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4070–4079 | 4077

View Article Online
Paper
Organic & Biomolecular Chemistry
1H, J2,3 = 9.5 Hz, J3,4 = 4.5 Hz, H-3), 3.46–3.41 (m, 2H, H-5b,
H-2), 3.40 (s, 3H, 2-OCH3), 3.32 (s, 3H, 1-OCH3), 3.32 (s, 3H,
3-OCH3), 3.12–3.05 (m, 2H, NHCOCH2), 2.62–2.56 (m, 2H,
CH2NH2); 13C NMR (100 MHz, CDCl3): δ 168.4 (CvO), 98.1
(C-1), 77.1 (C-2), 77.0 (C-3), 60.7 (C-5), 57.6 (2-OCH3), 56.4
((3-OCH3), 54.4 (1-OCH3), 46.7 (C-4), 35.7 (NHCOCH2), 31.2
(CH2NH2).
Methyl 4-(4′-amino)butanamido-4-deoxy-2,3-di-O-methyl-β-L-
arabinopyranoside (13). The carbamate 22 (12 mg,
0.032 mmol) was treated with cleavage solution (TFA–TES–dry
DCM, 1 : 2 : 7, 0.5 mL) for 4 h at room temperature. The clea-
vage solution was then evaporated under N2, and co-evapo-
rated with a mixture of toluene–i-PrOH (10 : 1) (5 × 5 mL) to
afford the crude product. The crude amine was then purified
on a reverse phase SPE C18 cartridge (5 g; 10% MeCN in H2O,
0.1% HCOOH → 100% MeCN, 0.1% HCOOH) to give the
2 K. C. Nicolaou, J. M. Salvino, K. Raynor, S. Pietranico,
T. Reisine, R. M. Freidinger and R. Hirschmann, in Pep-
tides, Chemistry, Structure and Biology: Proceedings of the
11th American Peptide Symposium, ed J. E. Rivier and
G. R. Marshall, ESCOM, Leiden, 1990, pp. 881–884.
3 G. Abbenante, B. Becker, S. b. Blanc, C. Clark, G. Condie,
G. Fraser, M. Grathwohl, J. Halliday, S. Henderson, A. Lam,
L. Liu, M. Mann, C. Muldoon, A. Pearson, R. Premraj,
T. Ramsdale, T. Rossetti, K. Schafer, G. L. Thanh,
G. Tometzki, F. Vari, G. r. Verquin, J. Waanders, M. West,
N. Wimmer, A. Yau, J. Zuegg and W. Meutermans, J. Med.
Chem., 2010, 53, 5576–5586.
4 G. Thanh Le, G. Abbenante, B. Becker, M. Grathwohl,
J. Halliday, G. Tometzki, J. Zuegg and W. Meutermans,
Drug Discovery Today, 2003, 8, 701–709.
5 R. Hirschmann, K. C. Nicolaou, S. Pietranico, E. M. Leahy,
J. Salvino, B. Arison, M. A. Cichy, P. G. Spoors,
W. C. Shakespeare, P. A. Sprengeler, P. Hamley, A. B.
Smith III, T. Reisine, K. Raynor, L. Maechler, C. Donaldson,
W. Vale, R. M. Freidinger, M. R. Cascieri and C. D. Strader,
J. Am. Chem. Soc., 1993, 115, 12550–12568.
6 R. Hirschmann, K. C. Nicolaou, S. Pietranico, J. Salvino,
E. M. Leahy, P. A. Sprengeler, G. Furst, C. D. Strader,
A. B. Smith, M. A. Cascieri, M. R. Candelore, C. Donaldson,
W. Vale and L. Maechler, J. Am. Chem. Soc., 1992, 114,
9217–9218.
7 R. Hirschmann, W. Yao, M. A. Cascieri, C. D. Strader,
L. Maechler, M. A. Cichy-Knight, J. Hynes, R. D. van Rijn,
P. A. Sprengeler and A. B. Smith, J. Med. Chem., 1996, 39,
2441–2448.
amine 13 as a colourless oil (8 mg, 91%). LC/MS: tR
=
1.24 min, m/z = 299.76 [M + Na]+; LRMS: m/z = 277.1 [M + H]+;
1H NMR (500 MHz, MeOD): δ 4.78 (1H, H-1, overlapped with
MeOD solvent peak), 4.40–4.38 (m, 1H, H-4), 3.74 (dd, 1H, J4,5a
= 2.5 Hz, J5a,5b = 12.0 Hz, H-5a), 3.51 (dd, 1H, J2,3 = 9.5 Hz, J3,4
= 4.5 Hz, H-3), 3.44–3.41 (m, 2H, H-5b, H-2), 3.39 (s, 3H,
2-OCH3), 3.31 (s, 3H, 3-OCH3), 3.30 (s, 3H, 1-OCH3), 2.92–2.88
(m, 2H, NHCOCH2), 2.37–2.34 (m, 2H, CH2NH2), 1.87–1.80 (m,
2H, CH2CH2NH2); 13C NMR (100 MHz, CDCl3): δ 173.3 (CvO),
98.1 (C-1), 77.0 (C-2), 76.9 (C-3), 60.7 (C-5), 57.6 (2-OCH3), 56.3
(3-OCH3), 54.4 (1-OCH3), 38.9 (NHCOCH2), 32.2 (CH2NH2),
23.1 (CH2CH2NH2).
General experimental procedure for the stability of three
model compounds under acidic conditions
8 W. Meutermans, G. Le Thanh and B. Becker, ChemMed-
Chem, 2006, 1, 1164–1194.
A model compound (2 mg) was dissolved in a mixture of 10%
TFA in H2O (3.6 mL, pH 1). The resulting solution was stirred
at 80 °C for 24 h. The reaction progress was analysed by TLC
(tert-butanol–acetic acid–water, 2 : 1 : 1) and 1H NMR and
9 G. Le Thanh, G. Abbenante, G. Adamson, B. Becker,
C. Clark, G. Condie, T. Falzun, M. Grathwohl, P. Gupta,
M. Hanson, N. Huynh, P. Katavic, K. Kuipers, A. Lam,
L. Liu, M. Mann, J. Mason, D. McKeveney, C. Muldoon,
A. Pearson, P. Rajaratnam, S. Ryan, G. Tometzki,
G. Verquin, J. Waanders, M. West, N. Wilcox, N. Wimmer,
A. Yau, J. Zuegg and W. Meutermans, J. Org. Chem., 2010,
75, 197–203.
10 P. Rajaratnam, P. Gupta, P. Katavic, K. Kuipers, N. Huyh,
S. Ryan, T. Falzun, G. B. Tometzki, L. Bornaghi, G. Le
Thanh, G. Abbenante, L. Liu, W. Meutermans, N. Wimmer
and M. L. West, Aust. J. Chem., 2010, 63, 693–699.
11 M. L. West, P. Andrews, T. E. Ramsdale, W. Meutermans,
G. Le Thanh, C. Clark, G. Abbenante and L. Liu, Derivatives
of monosaccharides for drug discovery, PCT Int. Appl
WO2004/014929 A1, 2004.
COSY spectroscopy. The compounds were too polar (tR
=
1.20 min, LC/MS), and hence, the reaction progress could not
be analysed by LC/MS. The resulting solution was concentrated
under reduced pressure, and dried under high vacuum over-
night. The reaction mixture was then examined by 1H NMR
and COSY spectroscopy.
Acknowledgements
We wish to thank Dr Tri Le for assistance with NMR
spectroscopy, and Assoc. Prof. Craig Williams and Prof. Mary
Garson for useful discussions. This work was supported by the
University of Queensland and Alchemia Ltd.
12 J. A. Davies, C. H. Hassall and I. H. Rogers, J. Chem. Soc. C,
1969, 1358–1363.
13 W. J. le Quesne and G. T. Young, J. Chem. Soc., 1952, 594–
597.
14 K. Poduska, G. S. Katrukha, A. B. Silaev and J. Rudinger,
Collect. Czech. Chem. Commun., 1965, 30, 2410–2433.
Notes and references
1 R. Hirschmann, L. Ducry and A. B. Smith III, J. Org. Chem., 15 A. M. Reverdito, M. Garcia, A. Salerno, M. Blanco and
2000, 65, 8307–8316.
I. A. Perillo, Spectrosc. Lett., 2001, 34, 335–343.
4078 | Org. Biomol. Chem., 2015, 13, 4070–4079
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
16 P. C. de Visser, C. Govaerts, P. A. V. van Hooft, 23 Y. Tsuda, M. Nishimura, T. Kobayashi, Y. Sato and
H. S. Overkleeft, A. Van Schepdael and J. Hoogmartens,
K. Kanemitsu, Chem. Pharm. Bull., 1991, 39, 2883–
J. Chromatogr. A, 2004, 1058, 183–189.
2887.
17 T. R. Hurley, C. E. Colson, G. Hicks and M. J. Ryan, J. Med. 24 D. W. C. Jones, R. J. Nash, E. A. Bell and J. M. Williams,
Chem., 1993, 36, 1496–1498. Tetrahedron Lett., 1985, 26, 3125–3126.
18 L. Arnoldi, A. Manzocchi, F. Magni, M. DelPuppo and 25 Y. Kondo, Carbohydr. Res., 1982, 107, 303–311.
M. G. Kienle, Bioorg. Chem., 1997, 25, 110–116.
19 J. Dheur, N. Ollivier and O. Melnyk, Org. Lett., 2011, 13,
1560–1563.
20 N. Ollivier, J. Dheur, R. Mhidia, A. Blanpain and
O. Melnyk, Org. Lett., 2010, 12, 5238–5241.
26 P. J. Stevenson, Org. Biomol. Chem., 2011, 9, 2078–
2084.
27 R. A. B. Bannard, A. A. Casselman, W. F. Cockburn and
G. M. Brown, Can. J. Chem., 1958, 36, 1541–1549.
28 Y. Kondo, Carbohydr. Res., 1982, 110, 339–344.
21 W. Hou, X. Zhang, F. Li and C.-F. Liu, Org. Lett., 2010, 13, 29 J. G. Buchanan and R. Fletcher, J. Chem. Soc. C, 1966,
386–389. 1926–1931.
22 B. Müller, M. Blaukopf, A. Hofinger, A. Zamyatina, 30 A. J. Dick and J. K. N. Jones, Can. J. Chem., 1965, 43, 977–
H. Brade and P. Kosma, Synthesis, 2010, 3143–3151.
982.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4070–4079 | 4079