Synthesis and investigation of polyhydroxylated pyrrolidine derivatives as novel chemotypes showing dual activity as glucosidase and aldose reductase inhibitors

Article abstract of DOI:10.1016/j.bioorg.2019.103298

Diabetes is a multi-factorial disorder that should be treated with multi-effective compounds. Here we describe the access to polyhydroxylated pyrrolidines, belonging to the D-gluco and D-galacto series, through aminocyclization reactions of two differenti

Full text of DOI:10.1016/j.bioorg.2019.103298

BioorganicChemistry92(2019)103298
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Synthesis and investigation of polyhydroxylated pyrrolidine derivatives as
novel chemotypes showing dual activity as glucosidase and aldose reductase
inhibitors
⁎
Lorenzo Guazzellia, Felicia D'Andreaa, , Stefania Sartinia, Francesco Giorgellia, Gianluca Confinia,
Luca Quattrinia, Ilaria Pianoa, Susanna Nencettia, Elisabetta Orlandinib,c, Claudia Garginia,
⁎
Concettina La Mottaa,
a Department of Pharmacy, University of Pisa, Via Bonanno 6/33, 56126 Pisa, Italy
b Department of Earth Sciences, University of Pisa, Via Santa Maria 53, 56126 Pisa, Italy
c Research Center “E. Piaggio”, University of Pisa, Pisa, 56122, Italy
A R T I C L E I N F O
A B S T R A C T
Keywords:
Diabetes is a multi-factorial disorder that should be treated with multi-effective compounds. Here we describe
the access to polyhydroxylated pyrrolidines, belonging to the D-gluco and D-galacto series, through aminocycli-
zation reactions of two differentially protected D-xylo-hexos-4-ulose derivatives. The prepared compounds
proved to inhibit both alpha-glucosidase, responsible for the emergence of hyperglycemic spikes, and aldose
reductase, accountable for the development of abnormalities in diabetic tissues. Accordingly, they show the dual
inhibitory profile deemed as ideal for diabetes treatment. Significantly, compound 17b reduced the process of
cell death and restored the physiological levels of oxidative stress when tested in the photoreceptor-like 661w
cell line, thus proving to be effective in an in vitro model of diabetic retinopathy.
Diabetes
Aldose reductase
Aldose reductase inhibitors
Glucosidase
Glucosidase inhibitors
Azafuranose
Pyrrolidine
Aminocyclization
1. Introduction
terminal renal diseases, which requires renal dialysis or even kidney
transplantation. Moreover, a great proportion of people with diabetes
Diabetes mellitus is a multi-factorial metabolic disorder character-
ized by high levels of blood glucose, resulting from defects in insulin
production, insulin action, or both. It is recognized as a public health
problem, as it affects a significant portion of the population worldwide
and is expanding to pandemic dimensions. Indeed, according to epi-
demiological studies, approximately 415 million of adults worldwide
were diagnosed with diabetes in 2015, and this number is expected to
rise to at least 640 million by 2040 [1].
show different forms of nervous system damage, including impaired
sensation of pain and physical disability. Severe forms of diabetic nerve
diseases are the major contributing cause of lower-extremity amputa-
tions, more than 60% of which occur among people affected by this
disease [2].
While research on the prevention of diabetes itself is promising,
efforts to control hyperglycemic spikes and prevent the resulting onset
of chronic functional alterations are still unfruitful. Therefore, the re-
search of novel and ground-breaking therapeutic solutions for diabetic
people is more relevant than ever.
Despite the administration of insulin and/or potent oral hypogly-
cemics, diabetes still remains a significant cause of morbidity and
mortality, especially due to the development of related disabling
complications. Patients with diabetes have a 2–4 times higher risk for
cardiovascular pathologies, including strokes, and show accelerated
formation of severe atherosclerotic lesions in peripheral, coronary, and
cerebral arteries. Diabetes is also the leading cause of visual impairment
and blindness, due to severe retinopathy and cataract, as well as
An attractive strategy to address diabetes is the development of a
multi-effective compound, able to combine a hypoglycaemic effect with
the ability to prevent long term diabetic complications.
The straightforward way to control sugar plasma concentration is to
take action on enzymes managing its availability, being alpha-glucosi-
dase (AG) the most relevant among others [3]. Up to 70% of the daily
Abbreviations: AG, alpha-glucosidase; ALR2, aldose reductase; Nrf2, nuclear factor erythroid 2-related factor 2; ROS, reactive oxygen species; Keap1, Kelch-like
ECH-associated protein 1; ARE, antioxidant response element; Sod1, Cu-Zn-superoxide dismutase
⁎ Corresponding authors at: Department of Pharmacy, University of Pisa, Via Bonanno 33, 56126 Pisa, Italy (F. D’Andrea). Department of Pharmacy, University of
Pisa, Via Bonanno 6, 56126 Pisa, Italy (C. La Motta).
E-mail addresses: felicia.dandrea@unipi.it (F. D'Andrea), concettina.lamotta@unipi.it (C. La Motta).
https://doi.org/10.1016/j.bioorg.2019.103298
Received 2 August 2019; Received in revised form 13 September 2019; Accepted 16 September 2019
Availableonline19September2019
0045-2068/©2019ElsevierInc.Allrightsreserved.

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
glucose load results from carbohydrate intake through the diet. Indeed,
sugars are digested on the brush border of the small intestine and AG
provides glucose for intestinal absorption through the regulation of the
speed and the extent of post-prandial hyperglycaemia. On the other
hand, the onset of pathological changes affecting the nervous, renal,
vascular and ocular systems of diabetic patients are generated by the
activation of aldose reductase (ALR2), which has been finally accepted
as the key and critical checkpoint of the main biochemical abnormal-
ities affecting diabetic tissues [4].
was carried in MeOH, a 3:2 mixture of 4-C-methoxy derivatives 4c,d
was obtained [17] (Scheme 1). By treating the mixture 4c,d with
CF3COOH in 2:1 MeCN-water, hydrolysis of all the acetal groups was
observed and D-xylo-hexos-4-ulose 6 was isolated [17]. However, by
employing milder reaction conditions (Et3N in 5:1 MeOH-H2O), it was
possible to selectively cleave the m-chlorobenzoate protecting group
(MCB) from 4a,b and unmask the carbonyl groups while leaving the
lateral chain isopropylidene moiety untouched. In doing so, the par-
tially protected 5,6-O-isopropylidene-D-xylo-hexos-4-ulose (5) was ob-
tained.
On these grounds, the ideal anti-diabetic agent should be able to
simultaneously block the catalytic activity of both AG and ALR2.
Broadly speaking, developing a dual-target agent is a challenging
task, as it is considerably hard to optimize multiple, selective and po-
tent activities within the same chemical structure. However, with re-
gard to compounds able to take action on glucose metabolism, a pri-
vileged class of derivatives do exist, which might come to the aid of
medicinal chemists for the design of novel effective agents.
1,4-Dicarbonyl hexose 6 exists predominantly as the two anomeric
α- and β- 4-keto pyranose forms 6α,β (Scheme 1), deriving from
hemiacetalyzation of the C-5 hydroxyl group onto the carbon of the
aldehyde group [17]. In the case of 5, a rather complex mixture of
tautomeric forms (as shown by NMR analysis) was instead observed as
the 5-OH is blocked into the isopropylidene acetal protecting group.
The two dicarbonyl compounds 5 and 6 were subjected without
further purification to the aminocyclization reaction. To a solution of
either 5 or 6 and the selected ammonium salt (1 eq) in MeOH,
NaBH3CN (2.2 eq) was added and the mixture was stirred at 60 °C for
24–48 h (Scheme 2). Overall, the yields of the aminocyclization reac-
tions were quite satisfactory, spanning from 37% to 50% (Table 1-SI,
Supporting Information), although somewhat lower than those usually
obtained in the
This is represented by the naturally-derived carbohydrate analogues
iminosugars, characterized by the presence of a nitrogen atom replacing
the endocyclic oxygen of carbohydrates. Thanks to their structural si-
milarity to natural sugars, iminosugars are able to modulate carbohy-
drate-processing enzymes and are therefore particularly suitable for the
development of the aforementioned bi-functional anti-diabetic agents.
Several iminosugars are already used as effective AG inhibitors,
including the oligosaccharide Acarbose [3] (1, Chart 1) and the pi-
peridine derivative Miglitol or Glyset [5] (2, Chart 1), both marketed
worldwide as the standard of care approach to control glucose plasma
concentration. Their ability of inhibit ALR2 has never been thoroughly
investigated but it is a fact that polyhydroxylated compounds, often of
natural origin, can be effective ALR2 inhibitors [6]. These findings
suggest that iminosugars might represent the reference chemotype for
the discovery of dual AG/ALR2 inhibitors.
frustrating were iDn-sxtyelaod-htehxeosd-5iffi-uclousletiessereinecsou[1n1te,1re8d,19w]h. ePnarttriyciunlgartloy
purify the pyrrolidine diastereoisomers via chromatography on silica
gel. Only in the case of the 11a,b mixture it was possible to separate the
D-gluco and D-galacto diastereoisomers. For 7a,b and 15a,b mixtures, a
partial separation was achieved and pure samples of only 7b and 15a
were obtained. Some of the pyrrolidine mixtures were also acetylated,
in order to facilitate NMR assignment (Scheme 2).
Conversely, really intriguing were the results in terms of the ste-
reoselectivity of the reaction. The configuration of the C-4 furanoside
epimers was deduced by NMR analysis. In the 13C spectra, signals of the
carbons adjacent to the nitrogen (C-1 and C-4), which are shielded and
easily recognizable by DEPT-135 experiments, were diagnostic for the
two epimers. In particular, in the
With this in mind, we designed a series of novel polyhydroxylated
pyrrolidine derivatives as possible bi-functional compounds, able to
both delay glucose absorption and prevent long term diabetic compli-
cations. Our strategy involved a combination of our synthetic expertise
in transforming common sugars into complex architectures, [7–9] in
particular into iminosugars derivatives via dicarbonyl compounds,
[10–12] with the long-lasting experience in the ALR2 inhibitors field
[13,14]. Results obtained are reported here below.
shifted by about 3–6 ppm towardsDh-gigluhceor dfieerlidvsattihvaens Cin-4thweas always
epimers (see experimental section). Very different ratios of Dt-hgealatwctoo
diastereoisomers were obtained by using diverse amine partners: from
15:85 to 70:30 D-gluco:D-galacto ratios (Table 1-SI, entries 3 and 6). The
results suggest that the stereo outcome of the reaction is particularly
sensitive to the steric hindrance of the amine; the D-galacto diastereoi-
somer is favored with NH3 and PhCH2NH2 (Table 1-SI, entries 1 and 3),
while the D-gluco diastereoisomer becomes the major reaction product
with Ph2CHNH2 and D/L-Phe-OMe (Table 1-SI, entries 4–6). This hy-
pothesis works for both dicarbonyl derivatives 5 and 6 tested in the
aminocyclization reaction.
2. Results and discussion
2.1. Chemistry
The retrosynthetic approach to prepare structurally-related families
of polyhydroxylated pyrrolidine derivatives is depicted in Fig. 1. The
double reductive amination reaction (aminocyclization) of 1,4-di-
carbonyl derivatives, which in turn can be obtained from a readily
available enol ether intermediate 3, represents the key step.
The aminocyclization of aldohexos-4-ulose derivatives is far less
studied than that of the more common aldohexos-5-uloses. For the
latter, in the case of compounds of the D-xylo series, a strong preference
for the “D-gluco” product is observed regardless of the structure of the
amine reacting partner [20]. Therefore, there is a huge difference in the
stereochemical outcome between the two dicarbonyl isomers of the
The known enol ether 3 [15,16] was subjected to the epoxidation
reaction with meta-chloroperoxybenzoic acid (MCPBA) in CH2Cl2 af-
fording the reported 7:3 mixture (81%) of C-4 epimeric 3-m-chlor-
obenzoates 4a,b [17] (Scheme 1). When the epoxidation reaction of 3
Chart 1. Oligosaccharide Acarbose and Miglitol or Glyset.
2

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
Fig. 1. Retrosynthetic approach to the polyhydroxylated pyrrolidine.
same D-xylo stereochemical series, even when the same reaction con-
ditions are applied. It is commonly accepted [11] that the reaction
proceeds through two consecutive steps, the fast reductive amination of
the aldehydo function, followed by a second, slower intramolecular
condensation of the resulting amine with the keto group. In this way a
cyclic iminium ion is formed, on which the final hydride attack de-
termines the stereochemical outcome of the reaction [21].
the corresponding mixture of 21a and 21b was obtained and it was
possible to isolate both pyrrolidines derivatives as pure compounds by
chromatographic purification and further transform them into the
known peracetylated derivatives 10a [22] and 10b [23] (Scheme 4).
The most abundant galacto-diastereoisomer was also converted into the
known deprotected hydrochloride salt 22 [24] both from 21b and 10b
by treatment with the acid conditions (2 M HCl).
For D-xylo aldohexos-5-uloses, the two chair-like transition states
deriving from the two possible 6-membered half-chair iminium ions
have such a different energy that practically only one of them is ac-
cessible, hence the high if not complete stereoselectivity of the reaction.
In the cases under investigation here, the formation of 5-membered
envelope iminium ions (2E and E2, Fig. 2) can be envisaged. Based on
the results, it is likely that the E2 iminium ion gives rise to the most
stable transition state and thus preferentially affords the D-galacto dia-
stereoisomer as long as the amine partner has a reduced steric hin-
drance. If the steric hindrance of the amine comes into play, than the 2E
cyclic iminium ion becomes the preferential one for the hydride attack.
Studies on this mechanism are currently underway ad will be presented
in due course.
It is worth mentioning that NMR analysis of pyrrolidine amides
18a,b, 19a,b and 10a,b was complicated by the presence of CeN
amide bond rotational isomers (see Experimental Section). It is well
known that the rotation energy barrier could determine the presence of
rotational isomers for amides and these have been reported before for
piperidine [25,26] and pyrrolidine [27,28] amide derivatives. To the
best of our knowledge, this is the first time that this hindered rotation
has been observed for polyhydroxylated N-acyl-pyrrolidine derivatives
of the D-gluco and D-galacto series (azafuranose). Rotational isomers
were found for all the synthesized N-acyl-compounds (18a,b, 19a,b
and 10a,b) and their ratio depended on both the substitution pattern
and the configuration of the furanoside epimers. For example, in the
case of 18a and 18b, two rotational isomers (ratios 75:25 and 60:40
respectively) were observed by recording the NMR spectra at 20 °C. The
signals coalesced and sharpened when the temperature was increased to
80 °C (see experimental section).
In order to increase the number of structurally-related compounds
for biological evaluations, a few different protecting group manipula-
tions were carried out on those structures that were obtained as pure
compounds (11a, 11b and 15a) from the aminocyclization reaction
(Scheme 3). In particular, the benzyl group on the nitrogen was re-
moved by hydrogenolysis (11a → 17a, 11b → 17b) and the resulting
compounds were reacted with isobutyric anhydride (17a → 18a,
17b → 18b). Finally, the 5,6-O-isopropylidene protecting group was
removed from compounds 11b, 15a, 18a and 18b by acid hydrolysis to
afford a series of pyrrolydine derivatives characterized by an enhanced
hydrophilic character (11b → 20b, 15a → 16a, 18a → 19a, 18b →
19b).
2.2. Biological activity
2.2.1. Aldose reductase and alpha-glucosidase inhibitory assays
To get some insight into the bi-functional activity of the novel
pyrrolidine derivatives, compounds were tested as both ALR2 and AG
inhibitors. Results obtained are listed in Table 1.
Within the
subseries, the presence of the 5,6-O-iso-
propylidene groupD-gwluacsodetrimental for the activity, being the protected
compounds generally less effective than the hydrophilic counterparts.
Compare for example the isopropylidene derivative 18a, devoid of any
significant inhibitory efficacy, with the hydrophilic parent 19a, which
proved to halve ALR2 activity when tested at the same concentration.
Similar results were also achieved testing the D-gluco compounds
In the case of the mixture of 7a and 7b, for which it was only
possible to obtain a small sample of pure 7b by chromatographic pur-
ification on silica gel, an attempt to verify whether the removal of the
nitrogen benzyl group would favour the diastereoisomeric separation
was performed (Scheme 4). By treating 7a and 7b with H2 and Pd/C,
Scheme 1. Synthesis of 4-uloside derivatives 5 and 6. Reagents and conditions: (i) Et3N, 5:1 MeOH-H2O (quantitative yield); (ii) CF3COOH, 2:1 CH3CN-H2O
(quantitative yield).
3

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
Scheme 2. Reagents and conditions: (i) PhCH2NH2·HCl (for 7a,b, 50% yield), or NH3·AcOH·(for 10a,b, 55% yield), NaBH3CN, MeOH, 60 °C, 24–48 h, followed by 1:2
Ac2O-Py, rt, 12 h; (ii) Ph2CHNH2·AcOH (for 8a,b, 52% yield), PhCH2NH2·HCl (11a,b, 37% yield), D-Phe-OMe·HCl (for 12a,b, 42% yield) or L-Phe-OMe·HCl (for
15a,b, 44% yield), NaBH3CN, MeOH, 60 °C, 24–48 h; (iii) 1:2 Ac2O-Py, rt, 12 h (for 9a,b, 71% yield; for 13a,b, 75% yield); (iv) 80% aq AcOH, 40 °C, 15 min (for
14a,b, 46% yield).
against AG, thus rewarding the 1,4-immino-D-glucitol 19a as the best
dual inhibitor of the whole subseries.
is characterized by both the presence of ALR2, as evidenced in Fig. 3A,
and the high sensitivity to hyperglycemic conditions, thus turning out
to be an excellent in vitro model of diabetic retinopathy. [30] Tested at
the concentration of 100 µM, compound 17b proved to inhibit the ac-
tivity of the target enzyme by approximately 90% with respect to the
activity shown by the enzyme in the untreated hyperglycemic cells
(Fig. 3B).
As for the D-galacto subseries, an opposite inhibitory trend was ob-
served. Actually, in the case of ALR2, the best functional efficacy was
displayed by the 5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-ga-
lactitol 17b, and any attempt to modify its chemical structure, by re-
moving the protecting group as well as inserting suitable substituents
on both the hydroxy residues and the nitrogen atom, resulted in less
effective analogues. On the contrary, the aforementioned structural
modifications allowed to get promising AG inhibitors. Indeed, removal
of the isopropylidene bridge from 17b led to compound 22, which
displayed an IC50 value of 62.4 μM against the target enzyme. An even
more marked increase of efficacy was obtained by inserting a benzyl
Once verified the ability of compound 17b to inhibit the enzymatic
activity of ALR2, its efficacy in both increasing cell viability and re-
ducing the biological pathways that trigger hyperglycemia-induced cell
death was also evaluated. Fig. 4A shows how the treatment at two
concentrations (50 and 100 µM) of the test compound is able to in-
crease, although not significantly, the levels of cellular vitality. More-
over, Fig. 4B indicates the ability of 17b to reduce significantly the
apoptotic process, the major pathway of death of retinal neurons.
Finally, the ability of 17b to act also against oxidative stress trig-
gered by hyperglycemic conditions was evaluated. Fig. 5A shows the
immunocytochemistry for Nrf2 as a transcription factor for phase 2
antioxidant enzymes. Under physiological conditions, Nrf2 is found in
the cytosol complex with the inhibitory factor Keap1. Under high oxi-
dative stress, like in hyperglycemic conditions, ROS induce the dis-
ruption of the Nrf2-Keap1 complex and while Keap1 is degraded, Nrf2
translocates to the nucleus where it interacts with the specific DNA
sequence ARE, thus inducing the increase of the synthesis of antioxidant
enzymes such as Sod1. As can be seen from Fig. 5A, in the ctr-NG, the
localization of Nrf2 is cytosolic (green staining) and well separated
residue on the nitrogen atom of the de-protected 22, as in 20b (IC50
:
40.64 μM).
On the whole, D-galacto derivatives turned out to be more effective
than the D-gluco isomers against the two enzymes, thus proving to have
the molecular geometry that best fits the catalytic binding sites of both
ALR2 and AG at the same time. The observed inhibitory effects were
generally more marked toward AG rather than ALR2, nonetheless
suggestive for the intended dual functional profile.
2.2.2. In vitro assessment of ALR2 inhibition
Compound 17b, showing the best inhibitory profile against ALR2,
was tested on the murine retinoblastoma photoreceptor-like cell line
661w, [29] in order to evaluate its efficacy in vitro. The 661w cell line
Fig. 2. 2E and E2 cyclic iminium ions.
4

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
Scheme 3. Reagents and conditions: (i) 80% aq AcOH, 40 °C, 0.5–1 h (16a: 89%; 19a: 83%; 19b: 85%; 20b: 90%); (ii) H2, 10% Pd/C, absolute EtOH, 48 h (17a: 85%;
17b: 87%); (iii) isobutyric anhydride, Et3N, 3:1 MeOH-H2O, rt, 2 h (18a: 85%; 18b: 75%).
from the red nuclear staining, but in the ctr-HG the green cytosolic
staining disappears and a yellow-orange color becomes visible of the
nucleus, indicating that Nrf2 translocation occurred in the nucleus in
response to the increase in ROS. Treatment with 17b at a concentration
of 100 µM restores the localization of cytosolic Nrf2, and results
obtained are fully comparable to those achieved in ctr-NG. This result is
also confirmed by data obtained by biochemical evaluation of Sod1
levels. Indeed, Fig. 5B shows how the 100 µM dose of 17b is able to
reduce Sod1 levels, indicating a reduction of ROS levels.
Scheme 4. Reagents and conditions: (i) H2, 10% Pd/C, absolute EtOH, 2 h (21a: 30%; 21b: 52%); (ii) 1:2 Ac2O-Py, rt, 12 h (10a: 91%; 10b: 90%); (iii) 2 M HCl,
100 °C, 15 min (87%).
5

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
Table 1
ALR2, has been shown to have activity also on an in vitro model of
diabetic retinopathy. Indeed, once tested in photoreceptor-like 661w
hyperglycemic cell line, 17b reduced the processes of cell death and
restored the physiological levels of oxidative stress, probably reducing
the production of ROS mediated by ALR2.
Aldose Reductase (ALR2) and α-glucosidase (AG) inhibitory data of poly-
hydroxylated pyrrolidine derivatives.
% Inhibitiona
Entry
Compound
ALR2
AG
D-Gluco Derivatives
4. Experimental section
1
2
3
4
5
6
18a
19a
11a
17a
15a
16a
n.a.b
49.0%
n.a.b
n.a.b
27.2
15.8
33.3
54.5
21.0
34.0
35.6
n.a.b
4.1. Chemistry
4.1.1. General methods
Optical rotations were measured on a Perkin-Elmer 241 polarimeter
at 20
2 °C. Melting points were determined with a Kofler hot-stage
D-Galacto Derivatives
apparatus and are uncorrected. 1H NMR spectra were recorded in ap-
propriate solvents with a Bruker AC 200 (200.13 MHz) or Bruker
Avance II (250.13 MHz or 400 MHz). 13C NMR spectra were recorded
with the spectrometers operating at 50.33 or 62.9 MHz. The assign-
ments were made, when possible, with the aid of DEPT-135, HETCOR,
HSQC and COSY experiments and by comparison of values for known
compounds and applying the additivity rules [32]. In the case of mix-
tures, assignments were made by referring to the differences in the peak
intensities. The first order proton chemical shifts δ are referenced to
either residual CD3CN (δH 1.94, δC 1.28) or residual CD3OD (δH
3.31 ppm, δC 49.0 ppm) and J-values are given in Hz. All reactions were
followed by TLC on Kieselgel 60 F254 with detection by UV light and/or
with ethanolic 10% phosphomolybdic or sulfuric acid, and heating or
for exposure to iodine vapors. Kieselgel 60 (E. Merck, 70–230 and
230–400 mesh, respectively) was used for column and flash chroma-
tography. Some of flash chromatography were conducted by the auto-
mated system Isolera Four SVTM (Biotage®), equipped with UV detector
with variable wavelength (200–400 nm). All reactions involving air- or
moisture-sensitive reagents were performed under an argon atmosphere
using anhydrous solvents. Anhydrous methanol (MeOH) and pyridine
were purchased from Sigma-Aldrich. Other dried solvents were ob-
tained by distillation according to standard procedure [33] and stored
over 4 Å molecular sieves activated at least 12 h at 200 °C. All reagents
were purchased from Aldrich Chemical Co. and were used without
further purification. MgSO4 was used as the drying agent for solutions.
Elemental analysis was used to determine the purity of compounds.
Analytical results are within 0.40% of the theoretical values.
7
18b
19b
20b
17b
22
18.8%
32.2
32.6
57.1
35.2
29.6
42.8
98.7
29.4
31.4
62.1
93.2c
30.2
89.4d
n.a.b
33.3
n.t.e
8
9
10
11
12
7b
13
10b
Epalrestat
DNJ.HCl
72.3
a
Percentage of enzyme inhibition at 100 μM test compound, obtained as
mean of at least three determinations. Standard errors of the means (SEMs) are
≤10%.
b
c
d
e
Not active. No inhibition was observed at 100 μM test compound.
IC50: 40.6
IC50: 62.4
Not tested.
3.80 μM.
5.11 μM.
3. Conclusions
Polyhydroxylated pyrrolidine derivatives of the D-gluco and D-galacto
series have been prepared, exploiting the aminocyclization reaction of
two differentially protected D-xylo-hexos-4-ulose derivatives. The ste-
reoselection of the reaction was affected by the steric hindrance of the
amine partner employed. This stereochemical outcome strongly differs
from what has been reported before for D-xylo-hexos-5-ulose derivatives
and is probably related to the different size of the cyclic iminium ion
intermediate (6-membered versus 5-membered) as well as the stabilities
of the conformers involved in the stereodetermining step. These com-
pounds, belonging to both D-gluco and D-galacto stereochemical series,
were tested both as AG and as ALR2 inhibitors. Intriguing structure-
related activity trends were observed, which, far from being conclusive,
suggest that subtle interactions between the enzymes and the imino-
sugar are present and that the development of bi-functional molecules
of this kind deserves to be further investigated. Moreover, compound
17b, emerged as the one showing the best inhibitory activity against
The following standard procedure was used for acetylation: a solu-
tion of the compound (1.0 mmol) in a 2:1 (v/v) mixture (6 mL) of
pyridine and Ac2O was stirred at room temperature for 12–24 h, and
then repeatedly co-evaporated under diminished pressure with toluene,
and the residue was purified by flash chromatography on silica. “Solid
foam” refers to amorphous compounds, recovered pure by chromato-
graphy for which all attempts to crystallize failed. 3-deoxy-1,2:5,6-di-O-
isopropylidene-D-erythro-hex-3-enofuranose (3) [15,16], (4R)- and (4S)-
Fig. 3. (A) Representative western blot confirming
the presence of ALR2 in the experimental cell
model; (B) Results obtained in the presence of 17b.
Inhibition of ALR2 was assessed by using a specific
kit assay. Values are calculated as percentage of
maximum activity measured in hyperglycemic
control (ctr-HG), and are reported as mean
(n = 3, independent experiments).
SEM
6

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
Fig. 4. Effectiveness of compound 17b
in increasing cell viability. (A) The his-
togram shows that while the hypergly-
cemic condition impairs significantly
the cell viability, compared to the nor-
moglycemic control (ctr-NG), the treat-
ment with 17b at both 50 and 100 µM
concentration partially recovers the
level of cell viability; (B) The histo-
grams show results obtained for apop-
totic pathway inhibition. It is important
to note that the level of apoptotic pro-
cess is significantly reduced after the
treatment with 100 µM of 17b. The va-
lues are reported as mean
SEM
(n = 3, independent experiments).
3-O-m-chlorobenzoyl-4-C-hydroxy-1,2:5,6-di-O-isopropylidene-α-D-
xylo-hexofuranose (4a and 4b) [17] and D-xylo-hexos-4-ulose (6) [17]
were prepared according to the reported procedure.
analysis (1H and 13C) showed the presence of 1,4-dicarbonyl derivative
5 as a complex mixture of tautomeric forms that was submitted without
further purification to a double-reductive aminocyclisation.
4.1.2. 5,6-O-isopropylidene-D-xylo-hexos-4-ulose (5)
4.1.3. N-benzyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-glucitol
The known mixture of 3-O-m-chlorobenzoyl derivatives 4a,b [17],
(1.00 g, 2.42 mmol) was dissolved in 5:1 (v/v) MeOH-H2O (50 mL),
treated with Et3N (0.85 mL) and the solution was stirred at room
temperature until TLC analysis (EtOAc) revealed the complete dis-
appearance of the starting material (Rf 0.72) and the formation of a
more retained product. After 15 min the mixture was concentrated
under diminished pressure and repeatedly co-evaporated with toluene
(4 × 15 mL). The crude product (1.16 g) was a syrup and the NMR
(7a)
and
N-benzyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-
galactitol (7b)
A
solution of the crude
(6) [17] (1.10 g,
4.16 mmol, 1.0 eq) in dry MeOHD(-1x2ylmo-Le)sows-a4s-utrleoastement with a solution
of BnNH2 (0.51 mL, 4.58 mmol, 1.1 eq) and AcOH (0.38 mL) in dry
MeOH (23 mL) and stirred at room temperature. After 5 min NaBH3CN
(576 mg, 9.16 mmol, 2.2 eq) was added, the mixture was heated to
60 °C and stirred until the starting material was completely disappeared
Fig. 5. Antioxidant activity of 17b. (A) The
image obtained with the florescence micro-
scope shows the specific staining for Nrf20
(green) and the nuclear staining obtained
with ethidium bromide (red). The arrows
indicate the different localization of Nrf2,
cytosolic in normoglycemic conditions and
after treatment with 17b, while it is nuclear
(yellow-orange marking, co-localization) in
hyperglycemic conditions. (B) The histo-
gram shows the protein levels of the anti-
oxidant enzyme Sod1. To be noted that the
treatment with 17b is effective in restoring
protein levels in a way comparable to those
of ctr-NG. The values are reported as
mean
SEM (n = 3, independent experi-
ments). (C) Representative image of a wes-
tern blot experiment, the content of Sod1
was normalized for total protein content by
using the Stain Free Technology (BioRad)
[31].
7

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
(48 h, TLC, 1:1 EtOAc-MeOH). The solution was cooled to room tem-
perature, excess NaBH3CN was destroyed by adding methanol HCl
(1.3 N) until pH 1–2 and stirred at room temperature for 1 h. The so-
lution was neutralized by addition of solid NaHCO3, stirred for 30 min
and filtered over a short pad of Celite®, washed with MeOH, and the
combined organic phases were concentrated at diminished pressure.
The crude residue was dissolved in a mixture of pyridine and Ac2O (2:1
v/v, 90 mL) and stirred at room temperature until TLC analysis (EtOAc,
20 h) revealed the formation of a single spot (Rf 0.57). The solution was
co-evaporated with toluene (4 × 30 mL) under diminished pressure and
the crude residue was partitioned between CH2Cl2 (80 mL) and satd aq
NaHCO3 (25 mL) and the aqueous phase extracted with CH2Cl2
(4 × 40 mL). The organic layers were collected, dried (MgSO4) and
evaporated under diminished pressure to leave a crude residue (1.73 g)
constituted (NMR, CDCl3) by a mixture of 1-deoxyazafuranose deriva-
tives 7a and 7b in the ratio of 4:6, measured on the relative intensities
of two C-4 signals at δ 65.6 and 68.3 respectively. Purification of crude
product by flash chromatography on silica gel (first 7:3 hexane-
EtOAc + 0.1% Et3N and then 3:7 hexane-EtOAc + 0.1% Et3N) gave a
mixture of pure 7a and 7b (872 mg, 50%) as a clear syrup, Rf 0.57
(EtOAc). Compounds 7a and 7b were inseparable by TLC with several
elution systems and their structures were unequivocally inferred from
their NMR data. A second purification of a mixture of 7a and 7b by
flash-chromatography on silica gel (1:9 hexane-EtOAc) afforded a
sample of pure 7b, as a clear syrup, [α]D −31.8 (c 0.9 in CHCl3), 1H
NMR (250.13 MHz, CDCl3): δ 7.28–7.14 (m, 5H, Ar-H), 5.20 (ddd, 1H,
J5,6a 8.2 Hz, J4,5 6.2 Hz, J5,6b 2.5 Hz, H-5), 5.14 (m, 1H, H-3), 4.89 (m,
1H, H-2), 4.58 (dd, 1H, J6a,6b 12.1 Hz, H-6b), 4.11 (dd, 1H, H-6a), 4.03,
3.46 (AB system, 2H, JAB 13.5 Hz, CH2Ph), 2.96–2.89 (m, 2H, H-4, H-
1α), 2.65 (dd, 1H, J1β,1α 11.6 Hz, J1β,2 4.7 Hz, H-1β), 2.14–1.96 (m,
12H, 4 × MeCO); 13C NMR (62.9 MHz, CDCl3): δ 170.8–169.1 (4 × C]
O), 137.8 (AreC), 128.5–127.3 (AreCH), 77.2 (C-3), 76.0 (C-2), 70.9
(C-5), 68.3 (C-4), 63.1 (C-6), 59.1 (CH2Ph), 56.6 (C-1), 20.9–20.8
(4 × MeCO). Anal. Calcd for C21H27NO8: C, 59.85; H, 6.43; N, 3.32.
Found: C, 59.78; H, 6.39; N, 3.27.
Purification of crude product by flash chromatography on silica gel (9:1
EtOAc-MeOH + 0.1% Et3N) gave pure a mixture of 8a and 8b (237 mg,
52%) as a yellow foam, Rf 0.42 (10:3 EtOAc-MeOH). Compounds 8a
and 8b were inseparable by TLC with several elution systems and their
structures were unequivocally inferred from their NMR data. 1H NMR
(200.13 MHz, CD3CN-D2O) of 8a: δ 5.01 (s, 1H, CHPh2), 3.20 (dd, 1H,
J
1β,1α 10.8 Hz, J1β,2 4.8 Hz, H-1β), 3.06 (dd, 1H, J4,5 5.5 Hz, J3,4 3.2 Hz,
H-4), 2.33 (dd, 1H, J1α,2 4.1 Hz H-1α); of 8b: δ 4.94 (s, 1H, CHPh2),
2.90 (m, 1H, H-4), 2.87 (dd, 1H, J1β,1α 11.4 Hz, J1α,2 3.0 Hz H-1α), 2.71
(dd, 1H, J1β,2 5.1 Hz, H-1β); cluster of signals for 8a and 8b: δ
7.63–7.19 (m, 10H, Ar-H), 4.04 (m, 4H, 2 × H-2, 2 × H-3), 3.40 (m,
6H, 2 × H-5, 2 × H-6a, 2 × H-6b); 13C NMR (50.33 MHz, CD3CN-D2O)
of 8a: δ 79.4, 76.6 (C-2, C-3), 73.4 (C-5), 71.6 (CHPh2), 64.5 (C-6), 60.0
(C-4), 56.2 (C-1); of 8b: δ 80.1, 73.9 (C-2, C-3), 73.4 (CHPh2), 71.6 (C-
5), 64.6 (C-4), 64.5 (C-6), 57.7 (C-1); cluster of signals for 8a and 8b: δ
143.8, 142.8 (2 × AreC), 129.6–126.6 (AreCH). Anal. Calcd for
C
19H23NO4: C, 69.28; H, 7.04; N, 4.25. Found: C, 69.22; H, 6.99; N,
4.21.
4.1.5. N-benzhydryl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-
glucitol (9a) and N-benzhydryl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-
immino-D-galactitol (9b)
Routine acetylation of a mixture of 8a and 8b (237 mg, 0.72 mmol)
with 1:2 Ac2O-Py mixture (9.0 mL) gave after flash chromatography
(7:3 hexane-EtOAc + 0.1% Et3N) pure a 6:4 mixture of tetra-O-acetyl
azafuranose derivatives 9a and 9b (250 mg, 71%) as a syrup, Rf 0.36
(1:1 hexane- EtOAc). Compounds 9a and 9b were inseparable by TLC
with several elution systems and their structures were in equivocally
inferred from their NMR data. 1H NMR (200.13 MHz, CDCl3) of 9a: δ
5.30 (m, 2H, H-3, H-5), 5.23 (s 1H, CHPh2), 5.12 (m, 1H, H-2), 4.53
(dd, 1H, J6a,6b 12.0 Hz, J5,6b 3.5 Hz, H-6b), 4.08 (dd, 1H, J5,6a 7.7 Hz, H-
6a), 3.61 (dd, 1H, J4,5 7.0 Hz, J3,4 3.2 Hz, H-4), 3.34 (dd, 1H, J1β,1α
11.6 Hz, J1β,2 6.1 Hz, H-1β), 2.63 (dd, 1H, J1α,2 4.2 Hz H-1α), 2.13,
2.12, 1.97, 1.83 (4 s, each 3H, 4 × MeCO); of 9b: δ 5.12 (m, 3H, H-2, H-
3, H-5), 5.08 (s 1H, CHPh2), 4.62 (dd, 1H, J6a,6b 12.2 Hz, J5,6b 2.7 Hz, H-
6b), 4.18 (dd, 1H, J5,6a 8.2 Hz, H-6a), 3.20 (dd, 1H, J4,5 6.7 Hz, J3,4
2.0 Hz, H-4), 3.10 (dd, 1H, J1β,1α 12.0 Hz, J1α,2 2.3 Hz H-1α), 2.84 (dd,
1H, J1β,2 5.4 Hz, H-1β), 2.17, 2.13, 2.01, 1.94 (4 s, each 3H, 4 × MeCO);
cluster of signals for 9a and 9b: δ 7.38–7.26 (m, 10H, Ar-H); 13C NMR
(50.33 MHz, CDCl3) of 9a: δ 141.3, 139.2 (2 × AreC), 75.7, 75.6 (C-2,
C-3), 70.7 (CHPh2), 68.4 (C-5), 63.4 (C-6), 60.7 (C-4), 51.3 (C-1); of 9b:
δ 141.4, 139.0 (2 × AreC), 78.0, 75.9 (C-2, C-3), 71.0 (C-5), 70.5
(CHPh2), 66.2 (C-4), 63.4 (C-6), 53.3 (C-1); cluster of signals for 9a and
The NMR analysis of the mixture of 7a and 7b allowed assigning the
chemical shifts of the proton and carbon related to the azafuranosic
structure 7a. 1H NMR (200.13 MHz, CDCl3): δ 7.37–7.19 (m, 5H, Ar-H),
5.48 (dd, 1H, J3,4 6.5 Hz, J2,3 3.1 Hz, H-3), 5.22 (m, 1H, H-5), 5.00
(ddd, 1H, J1β,2 5.3 Hz, J1α,2 4.5 Hz, H-2), 4.26–4.20 (m, 2H, H-6a, H-
6b), 4.18, 3.54 (AB system, 2H, JAB 13.4 Hz, CH2Ph), 3.40 (dd, 1H, J4,5
5.1 Hz, H-4), 3.29 (dd, 1H, J1β,1α 11.5 Hz, H-1β),2.40 (dd, 1H, H-1α),
2.11–2.04 (m, 12H, 4 × MeCO); 13C NMR (50.33 MHz, CDCl3) δ
170.4–168.8 (C]O), 138.1 (AreC), 128.6–126.1 (AreCH), 76.6, 76.3
(C-2, C-3), 70.9 (C-5), 65.6 (C-4), 63.3 (C-6), 60.9 (CH2Ph), 55.9 (C-1),
20.6 (4 × MeCO). Anal. Calcd for C21H27NO8: C, 59.85; H, 6.43; N,
3.32. Found: C, 59.75; H, 6.35; N, 3.28.
9b:
δ
170.2–169.6 (C]O), 129.2–126.9 (AreCH), 21.0–20.3
(4 × MeCO). Calcd for C27H31NO8: C, 65.18; H, 6.28; N, 2.83. Found: C,
65.14; H, 6.25; N, 2.80.
4.1.6. N-acetyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-glucitol
(10a) and N-acetyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-
galactitol (10b)
4.1.4. N-benzhydryl-1,4-dideoxy-1,4-immino-D-glucitol (8a) and N-
benzhydryl-1,4-dideoxy-1,4-immino-D-galactitol (8b)
A solution of the crude D-xylo-esos-4-ulose (6) [17] (253 mg,
1.38 mmol, 1.0 eq) in dry MeOH (12 mL) was treatment with a solution
of Ph2CHNH2 (0.24 mL, 1.38 mmol, 1 eq) and AcOH (0.11 mL) in dry
MeOH (7 mL) and stirred at room temperature. After 5 min NaBH3CN
(173 mg, 2.76 mmol, 2 eq) was added, the mixture was heated to 60 °C
and stirred until TLC analysis (10:3 EtOAc-MeOH) showed the complete
disappearance of the starting material (5 h) and the formation of a
single spot (Rf 0.42). The solution was cooled to room temperature,
excess NaCNBH3 was destroyed by adding methanol HCl (1.3 N) until
pH 1–2 and stirred at room temperature for 1 h. The solution was
neutralized by addition of solid NaHCO3, stirred for 30 min and filtered
over a short pad of Celite®, washed with MeOH, and the combined
organic phases were concentrated at diminished pressure. The crude
residue (625 mg) constituted (NMR, CD3CN-D2O) by a mixture of aza-
furanose derivatives 8a and 8b in the ratio of 60:40, measured on the
relative intensities of two C-4 signals at δ 60.0 and 64.6 respectively.
A solution of the crude D-xylo-esos-4-ulose (6) [17] (560 mg,
1.65 mmol, 1.0 eq) in dry MeOH (23 mL) was treatment with a solution
of ammonium acetate (NH4+AcO-, 1.37 g, 17.8 mmol, 10.8 eq) in dry
MeOH (39 mL) and stirred at room temperature. After 5 min NaBH3CN
(228 mg, 3.62 mmol, 2.2 eq) was added, the mixture was heated to
60 °C and stirred until the starting material was completely disappeared
(90 h, TLC, 1:1 EtOAc-MeOH). The solution was cooled to room tem-
perature, excess NaCNBH3 was destroyed by adding methanol HCl
(1.3 N) until pH 1–2 and stirred at room temperature for 1 h. The so-
lution was neutralized by addition of solid NaHCO3, stirred for 30 min
and filtered over a short pad of Celite®, washed with MeOH, and the
combined organic phases were concentrated at diminished pressure.
Routine acetylation of crude residue with 1:2 Ac2O-Py (10.0 mL) gave
after work-up the reaction a crude syrup (714 mg) constituted (NMR)
by a mixture of azafuranose derivatives 10a and 10b in the ratio of
15:85, measured on the relative intensities of two C-4 signals at δ 56.1
8

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
(for 10a) and 61.6 and 63.5 (for rotational isomers amide 10b). Pur-
ification of crude product by flash chromatography on silica gel (7:2:1
Et2O-EtOAc-iPrOH) gave pure a mixture of 10a and 10b (338 mg, 55%).
Compounds 10a and 10b were inseparable by TLC with several elution
systems and their NMR spectroscopic data of 10a and 10b was de-
scribed below.
formation of a major spot (Rf 0.50) visible under UV light. The solution
was cooled to room temperature and concentrated under diminished
pressure. The crude residue was partitioned between CH2Cl2 (30 mL)
and satd aq NaHCO3 (30 mL) and the aqueous phase was extracted with
CH2Cl2 (4 × 35 mL). The organic layers were collected, dried (MgSO4)
and evaporated under diminished pressure to leave a crude syrup
(772 mg) constituted (NMR) by a mixture of azafuranose derivatives
12a and 12b in the ratio of 6:4, measured on the relative intensities of
two C-4 signals at δ 64.5 and 68.4 respectively. Purification of crude
product by flash chromatography on silica gel (3:1 Et2O-EtOAc + Et3N)
gave pure a 7:3 mixture of 12a and 12b (375 mg, 42% calculated from
4a,b), impure for the excess of D-Phe-OMe. Compounds 12a and 12b
were inseparable by TLC with several elution systems and the NMR
analysis of the isolated fraction allowed assigning only the chemical
shifts of the carbon related to the azafuranosic structures. 13C NMR
(62.9 MHz, CDCl3) of 12a: δ 173.6 (C]O), 138.2 (AreC), 109.0
(CMe2), 79.3, 78.1 (C-3, C-2), 75.3 (C-5), 65.6 (C-6), 64.5 (C-4, CHN),
52.1 (C-1), 51.4 (OMe), 36.8 (CH2Ph), 26.2, 24.9 (CMe2); of 12b: δ
175.1 (C]O), 138.0 (AreC), 108.0 (CMe2), 79.2, 77.0 (C-3, C-2), 75.8
(C-5), 68.9 (C-4), 66.1 (C-6), 63.7 (CHN), 52.7 (C-1), 51.5 (OMe), 32.4
(CH2Ph), 26.1, 25.3 (CMe2); cluster of signals for 12a and 12b: δ
129.1–126.2 (AreCH).
4.1.7. N-benzyl-5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-glucitol
(11a)
and
N-benzyl-5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-
galactitol (11b)
The crude 5,6-O-isopropylidene-D-xylo-hexos-4-ulose (5) (1.16 g)
obtained from 4a,b (2.42 mmol) was dissolved in dry MeOH (50 mL),
and a solution of BnNH2·HCl (346 mg, 2.42 mmol, 1 eq) in dry MeOH
(20 mL) and NaBH3CN (303 mg, 4.84 mmol, 2.2 equiv.) were added.
The mixture was heated to 60 °C and stirred until TLC analysis (9:1
EtOAc-MeOH) showed the complete disappearance of the starting ma-
terial (40 h) and the formation of a two spot (Rf 0.52 and 0.38) visible
under UV light. The solution was cooled to room temperature and
concentrated under diminished pressure. The crude residue was parti-
tioned between CH2Cl2 (70 mL) and satd aq NaHCO3 (50 mL) and the
aqueous phase was extracted with CH2Cl2 (4 × 50 mL). The organic
layers were collected, dried (MgSO4) and evaporated under diminished
pressure to leave a crude syrup (680 mg) constituted (NMR) by a mix-
ture of azafuranose derivatives 11a and 11b in the ratio of 1:1, mea-
sured on the relative intensities of two C-4 signals at δ 66.7 and 73.7
respectively. Purification of crude product by flash chromatography on
silica gel (1:1 Et2O-EtOAc + 0.1% Et3N) gave pure 11a (90 mg, calcu-
lated from 4a,b 13%) and 11b (168 mg, 24% calculated from 4a,b).
N-benzyl-5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-glucitol
(11a) as a white solid, Rf 0.52 (9:1 EtOAc-MeOH); [α]D −23.9 (c 1.1 in
CHCl3), mp 83–85 °C (from EtOAc-hexane), 1H NMR (250.13 MHz,
CDCl3): δ 7.39–7.24 (m, 5H, Ar-H), 4.27 (m, 1H, H-5), 4.20 (dd, 1H, J3,4
5.9 Hz, J2,3 3.7 Hz, H-3), 4.15–4.08 (m, 2H, H-2, H-6b), 4.13, 3.58 (AB
system, 2H, JAB 13.2 Hz, CH2Ph), 3.96 (dd, 1H, J6a,6b 8.0 Hz, J5,6a
7.6 Hz,H-6a), 3.28 (dd, 1H, J1β,1α 10.5 Hz, J1β,2 5.7 Hz, H-1β), 3.22 (bt,
1H, J4,5 = J3,4 = 5.9 Hz, H-4), 3.50.3.30 (bs, 2H, 2 × OH), 2.35 (dd,
1H, J1α,2 5.5 Hz, H-1α),1.44, 1.37 (2 s, each 3H, CMe2); 13C NMR
(62.9 MHz, CDCl3): δ 138.0 (AreC), 128.9–127.3 (AreCH), 108.5
(CMe2), 78.9 (C-3), 76.1, 76.0 (C-2, C-5), 67.1 (CH2Ph), 66.5 (C-4),
60.4 (C-6), 58.3 (C-1), 26.4, 25.0 (CMe2). Anal. Calcd for C16H23NO4: C,
65.51; H, 7.90%; N, 4.77%; Found: C, 65.49; H, 7.85%; N, 4.74%.
N-benzyl-5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-galactitol
(11b) as a colourless syrup, Rf 0.38 (9:1 EtOAc-MeOH); [α]D +52.2 (c
1.2 in CHCl3), 1H NMR (200.13 MHz, CD3CN): δ 7.37–7.24 (m, 5H, Ar-
H), 4.29 (ddd, 1H, J4,5 6.6 Hz, J5,6a 7.2 Hz, J5,6b 6.4 Hz, H-5), 4.24, 3.42
(AB system, 2H, JAB 13.6 Hz, CH2Ph), 4.03 (dd, 1H, J6a,6b 8.4 Hz, H-6b),
3.93 (dd, 1H, H-6a), 3.82 (m, 1H, H-2), 3.76 (m, 1H, H-3), 3.39 (bs, 2H,
2 × OH), 2.61 (dd, 1H, J3,4 4.1 Hz, H-4), 2.54 (dd, 1H, J1α,1β 10.4 Hz,
J1β,2 4.8 Hz, H-1β), 2.70 (bd, 1H, H-1α),1.42, 1.33 (2 s, each 3H,
(CMe2); 13C NMR (50.33 MHz, CD3CN): δ 140.3 (AreC), 129.6–127.7
(AreCH), 109.6 (CMe2), 80.9, 78.8 (C-2, C-3), 77.4 (C-5), 73.7 (C-4),
67.0 (C-6), 60.3 (CH2Ph), 60.0 (C-1), 27.0, 25.6 (CMe2). Anal. Calcd for
C16H23NO4: C, 65.51; H, 7.90%; N, 4.77%; Found: C, 65.48; H, 7.86%;
N, 4.75%.
Routine acetylation of
a mixture of 12a and 12b (200 mg,
0.55 mmol) with 1:2 Ac2O-Py mixture (10.0 mL) gave after flash chro-
matography (6:4 hexane-EtOAc) pure a 1:1 mixture of di-O-acetyl
azafuranose derivatives 13a and 13b (185 mg, 75%) as a colourless
syrup, Rf 0.62 (6:4 hexane-EtOAc), Compounds 13a and 13b were in-
separable by TLC with several elution systems and their structures were
in equivocally inferred from their NMR data. 1H NMR (250.13 MHz,
CDCl3) of 13a: δ 5.24 (dd, 1H, J3,4 6.3 Hz, J2,3 3.7 Hz, H-3), 5.01 (m,
1H, H-2), 4.20–4.06 (m, 3H, H-5, H-6b, CHN), 3.90 (dd, 1H, J6a,6b
7.4 Hz, J5,6b 6.4 Hz, H-6b), 3.76 (dd, 1H, J4,5 3.3 Hz, H-4), 3.56 (s, 3H,
OMe), 3.45 3.34 (dd, 1H, J1β,1α 11.4 Hz, J1β,2 5.8 Hz, H-1β), 3.15 (m,
1H, H-1α), 1.36, 1.27 (2 s, each 3H, CMe2); of 13b: δ 4.95 (m, 1H, H-2),
4.88 (bd, 1H, J3,4 3.3 Hz, H-3), 4.04–4.00 (m, 2H, H-5, CHN), 3.76 (dd,
1H, J4,5 7.2 Hz, H-4), 3.68 (s, 3H, OMe), 3.63 (m, 2H, H-6a, H-6b), 3.45
(dd, 1H, J1β,1α 11.0 Hz, J1β,2 4.8 Hz, H-1β), 3.15 (dd, 1H, J1α,2 6.8 Hz,
H-1α,), 1.37, 1.31 (2 s, each 3H, CMe2), cluster of signals for 13a and
13b: δ 7.29–7.16 (m, 5H, Ar-H), 3.12–2.90 (m, 2H, CH2Ph), 2.06–2.04
(2 s, each 3H, 2 × MeCO); 13C NMR (62.9 MHz, CDCl3) of 13a: δ 172.2
(COOMe), 137.5 (AreC), 108.0 (CMe2), 77.0 (C-3), 75.8 (C-5), 75.1 (C-
2), 65.4 (C-6), 64.1 (CHN), 61.0 (C-4), 51.5 (OMe), 50.2 (C-1), 36.3
(CH2Ph), 26.1, 24.7 (CMe2); of 13b: δ 172.0 (COOMe), 138.0 (AreC),
109.2 (CMe2), 77.1 (C-3), 76.1 (C-5), 75.9 (C-2), 67.1 (C-4), 65.7 (C-6),
62.3 (CHN), 51.1 (OMe), 50.6 (C-1), 31.2 (CH2Ph), 26.3, 24.9 (CMe2);
cluster of signals for 13a and 13b: δ 169.9–169.3 (C]O), 129.1–126.3
(AreCH), 20.8 (4 × MeCO). Anal. Calcd for C23H31NO8: C, 61.46; H,
6.95%; N, 3.12%; Found: C, 61.43; H, 6.90%; N, 3.09%.
4.1.9. N-[(2′R)-1′-carbomethoxy-2′-benzylethyl]-2,3-di-O-acetyl-1,4-
dideoxy-1,4-immino-D-glucitol (14a) and N-[(2′R)-1′-carbomethoxy-2′-
benzylethyl]-2,3-di-O-acetyl-1,4-dideoxy-1,4-immino-D-galactitol (14b)
A solution of 1:1 mixture of di-O-acetyl azafuranose derivatives 13a
and 13b (81 mg, 0.18 mmol) in 80% aq AcOH (v/v,3.5 mL) was stirred
at 40 °C until the TLC analysis (2:8 hexane-EtOAc) revealed the com-
plete reaction of the starting material with formation of a major slower
moving products (Rf 0.35). The solution was then cooled to room
temperature and co-evaporated with toluene (6 × 5 mL) under dimin-
ished pressure. Purification of the crude residue by flash chromato-
graphy on silica gel (3:7 hexane-EtOAc) gave pure a mixture of aza-
furanose derivatives 14a and 14b (34 mg, 46%) in the ratio of 65:35,
measured on the relative intensities of two CH2Ph signals at δ 37.5 and
33.0 respectively. Compounds 14a and 14b as a colourless syrup, Rf
0.35 (2:8 hexane-EtOAc), were unseparable by TLC with several elution
systems and the NMR analysis of the isolated fraction allows to assign
only the chemical shifts of the carbon related to the azafuranosic
4.1.8. N-[(2′R)-1′-carbomethoxy-2′-benzylethyl]-2,3-di-O-acetyl-5,6-O-
isopropylidene-1,4-dideoxy-1,4-immino-D-glucitol (13a) and N-[(2′R)-1′-
carbomethoxy-2′-benzylethyl]-2,3-di-O-acetyl-5,6-O-isopropylidene-1,4-
dideoxy-1,4-immino-D-galactitol (13b)
The crude 5,6-O-isopropylidene-D-xylo-hexos-4-ulose (5) (534 mg,
2.45 mmol) obtained from 4a,b was dissolved in dry MeOH (30 mL),
and a solution of D-Phe-OMe·HCl (580.7 mg, 2.69 mmol, 1.1 eq) in dry
MeOH (6.0 mL) and a solution of NaBH3CN (338 mg, 5.38 mmol, 2.2
equiv.) in dry MeOH (6.0 mL) were added. The mixture was heated to
65 °C and stirred until TLC analysis (1:1 Et2O-EtOAc + Et3N) showed
the complete disappearance of the starting material (48 h) and the
9

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
structures. 13C NMR (62.9 MHz, CD3CN) of 14a: δ 139.4 (AreC), 79.3,
78.9, 77.2 (C-3, C-5, C-2), 62.2 (C-6), 61.5 (C-4), 59.7 (CHN), 56.3 (C-
1), 52.1 (OMe), 37.5 (CH2Ph); of 14b: δ 139.9 (AreC), 79.0 (C-3), 76.3
(C-2), 70.8 (C-5), 67.0 (C-6), 64.9 (C-4), 63.9 (CHN), 52.1 (OMe), 50.6
(C-1), 33.0 (CH2Ph); cluster of signals for 14a and 14b: δ 171.1–171.0
(C]O), 130.9–127.3 (AreCH), 21.2–21.0 (MeCO). Anal. Calcd for
C20H27NO8: C, 58.67; H, 6.65%; N, 3.42%; Found: C, 58.64; H, 6.61%;
N, 3.39%.
J6a,6b 11.2 Hz, J5,6b 4.2 Hz, H-6b), 3.45 (dd, 1H, J5,6a 5.7 Hz, H-6a), 3.40
(m, 1H, H-1β), 3.22 (dd, 1H, J4,5 6.3 Hz, H-4), 2.90 (m, 1H, H-1α); 13C
NMR (62.9 MHz, CD3OD): δ 174.1 (C]O), 139.8 (AreC), 130.1–127.5
(AreCH), 79.1 (C-3), 76.4 (C-2), 73.0 (C-5), 64.8 (CHN), 64.3 (C-6),
64.0 (C-4), 52.4 (C-1), 51.7 (OMe), 38.0 (CH2Ph). Anal. Calcd for
C16H23NO6: C, 55.06; H, 7.13%; N, 4.31%; Found: C, 55.02; H, 7.10%;
N, 4.28%.
4.1.12. 5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-glucitol (17a)
To a solution of azafuranose derivative 11a (369 mg, 1.26 mmol) in
absolute EtOH (60 mL), 10% Pd on charcoal (230 mg) was added and
the mixture was stirred at room temperature under H2 (atmospheric
pressure) until TLC analysis (8:2 EtOAc-MeOH) showed the complete
disappearance of the starting material. After 48 h the suspension was
filtered through a small layer of Celite® and evaporated under dimin-
ished pressure. Purification of the crude residue by flash chromato-
graphy on silica gel (90:10:1 CHCl3-MeOH-H2O) gave pure of azafur-
anose derivatives 17a (218 mg, 85%) as white solid, Rf 0.38 (70:30:3
CHCl3-MeOH-H2O), mp 160–164 °C (dec.); [α]D −6.2 (c 1.0 in MeOH),
1H NMR (400 MHz, CD3CN-D2O): δ 4.15 (dt, 1H, J5,6a = J5,6b 6.2 Hz,
J4,5 86 Hz, H-5), 4.03 (dd, 1H, J6a,6b 8.4 Hz, H-6b), 3.98, 3.93 (2 m,
each 1H, H-2, H-3), 3.76 (dd, 1H, H-6b), 3.14–3.09 (m, 2H, H-1β, H-4),
2.62 (d, 1H, J1α,1β 11.9 Hz, J1α,2 1.5 Hz, H-1α), 1.35, 1.29 (2 s, each 3H,
CMe2); 13C NMR (100 MHz, CD3CN-D2O): δ 109.8 (CMe2), 78.0, 77.6
(C-2, C-3), 74.8 (C-5), 68.4 (C-6), 63.6 (C-4), 52.7 (C-1), 26.9, 25.6
(CMe2). Anal. Calcd for C9H17NO4: C, 53.19; H, 8.43%; N, 6.89%;
Found: C, 53.16; H, 8.39%; N, 6.86%.
4.1.10. N-[(2′S)-1′-carbomethoxy-2′-benzylethyl]-5,6-O-isopropylidene-
1,4-dideoxy-1,4-immino-D-glucitol (15a) and N-[(2′S)-1′-carbomethoxy-
2′-benzylethyl]-5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-galactitol
(15b)
The crude 5,6-O-isopropylidene-D-xylo-hexos-4-ulose (5) (470 mg)
obtained from 4a,b (2.15 mmol) was dissolved in dry MeOH (25 mL),
and a solution of
MeOH (7.0 mL) anLd-Pahes-oOluMtieo·HnColf (N5a1B1Hm3gC,N2.(32798mmmgo,l,41.7.14 emqm) oinl, d2r.2y
equiv.) in dry MeOH (7.0 mL) were added. The mixture was heated to
65 °C and stirred until TLC analysis (1:2 Et2O-EtOAc + Et3N) showed
the complete disappearance of the starting material (48 h) and the
formation of a major spot (Rf 0.51) visible under UV light. The solution
was cooled to room temperature and concentrated under diminished
pressure. The crude residue was partitioned between CH2Cl2 (30 mL)
and satd aq NaHCO3 (30 mL) and the aqueous phase was extracted with
CH2Cl2 (4 × 35 mL). The organic layers were collected, dried (MgSO4)
and evaporated under diminished pressure to leave a crude syrup
(689 mg) constituted (NMR) by a mixture of azafuranose derivatives
15a and 15b in the ratio of 7:3, measured on the relative intensities of
two C-4 signals at δ 63.4 and 69.6 respectively. Purification of crude
product by flash-chromatography on silica gel (3:1 Et2O-EtOAc + Et3N)
afforded a pure 15a (166 mg, 21%, calculated from 4a,b) and 35:65
mixture of 15a and 15b (182 mg), impure for the excess of L-Phe-OMe.
N-[(2′S)-1′-carbomethoxy-2′-benzylethyl]-5,6-O-isopropylidene-1,4-di-
deoxy-1,4-immino-D-glucitol (15a) as colourless syrup, Rf 0.52 (1:2
4.1.13. 5,6-O-isopropylidene-1,4-dideoxy-1,4-immino-D-galactitol (17b)
To a solution of azafuranose derivative 11b (508 mg, 1.73 mmol) in
absolute EtOH (60 mL), 10% Pd on charcoal (327 mg) was added and
the mixture was stirred at room temperature under H2 (atmospheric
pressure) until TLC analysis (8:2 EtOAc-MeOH) showed the complete
disappearance of the starting material. After 48 h, the suspension was
filtered through a small layer of Celite® and evaporated under dimin-
ished pressure. Purification of the crude residue by flash chromato-
graphy on silica gel (90:10:1 CHCl3-MeOH-H2O) gave pure 17b
(306 mg, 87%) as white solid, Rf 0.47 (70:30:3 CHCl3-MeOH-H2O), mp
129–130 °C (from EtOH), [α]D −13.9 (c 1.1 in MeOH), 1H NMR
(400 MHz, CD3CN): δ 4.14 (ddd, 1H, J4,5 5.4 Hz, J5,6a 7.00 Hz, J5,6b
6.4 Hz, H-5), 3.98 (dd, 1H, J6a,6b 8.2 Hz, H-6b), 3.85 (m, 1H, H-2), 3.76
(dd, 1H, H-6a), 3.64 (dd, 1H, J2,3 3.3 Hz, J3,4 5.2 Hz, H-3), 3.03 (dd, 1H,
J1α,1β 11.2 Hz, J1β,2 5.4 Hz, H-1β), 3.18 (dd, 1H, H-4), 2.73 (dd, 1H,
J1α,2 3.1 Hz, H-1α), 1.36, 1.30 (2 s, each 3H, CMe2); 13C NMR
(62.9 MHz, CD3CN,): δ 109.7 (CMe2), 77.7, 77.3 (C-2, C-3), 74.5 (C-5),
63.7 (C-4), 68.3 (C-6), 52.8 (C-1), 27.0, 25.6 (CMe2). Anal. Calcd for
C9H17NO4: C, 53.19; H, 8.43%; N, 6.89%; Found: C, 53.17; H, 8.40%; N,
6.85%.
Et2O-EtOAc + 0.1% di Et3N), 1H NMR (250.13 MHz, CDCl3):
δ
7.36–7.23 (m, 5H, Ar-H), 4.17–4.06 (m, 2H, H-3, H-3), 3.96–3.89 (m,
2H, H-5, CHN), 3.73 (m, 1H, H-6b), 3.71 (s, 3H, OMe), 3.58–3.42 (m,
3H, H-1β, H-4, H-6a), 3.15–2.85 (m, 5H, H-1α, CH2Ph, OH-2, OH-3),
1.33, 1.26 (2 s, each 3H, CMe2); 13C NMR (62.9 MHz, CDCl3): δ 175.2
(C]O), 138.1 (AreC), 129.0–126.5 (AreCH), 107.4 (CMe2), 78.4, 76.6
(C-3, C-2), 75.7 (C-5), 63.9 (CHN), 63.4 (C-4), 66.0 (C-6), 53.0 (C-1),
51.8 (OMe), 37.7 (CH2Ph), 26.2, 24.5 (CMe2). Anal. Calcd for
C
19H27NO6: C, 62.46; H, 7.45%; N, 3.83%; Found: C, 62.42; H, 7.41%;
N, 3.79%.
The NMR analysis of the 35:65 mixture of 15a and 15b allows to
assign only the chemical shifts of the carbon related to the azafuranosic
structure 15b. 13C NMR (62.9 MHz, CDCl3): δ 173.4 (C]O), 138.4
(AreC), 129.1–126.1 (AreCH), 109.1 (CMe2), 79.5, 78.2 (C-3, C-2),
75.6 (C-5), 64.0 (CHN), 69.4 (C-4), 66.4 (C-6), 54.2 (C-1), 51.4 (OMe),
32.5 (CH2Ph), 26.4, 25.5 (CMe2).
4.1.14. N-(2-methylpropanoyl)-5,6-O-isopropylidene-1,4-dideoxy-1,4-
immino-D-glucitol (18a)
4.1.11. N-[(2′S)-1′-carbomethoxy-2′-benzylethyl]-1,4-dideoxy-1,4-
immino-D-glucitol (16a)
To a solution of 17a (180 mg, 0.89 mmol) in 3:1 mixture of MeOH-
H2O (v/v, 2.5 mL), Et3N (0.16 mL, 1.14 mmol) and isobutyric anhydride
(0.30 mL, 1.80 mmol) were added and the solution was stirred at room
temperature until TLC analysis (8:2 EtOAc-MeOH) showed the com-
plete disappearance of the starting material. After 2 h, satd aq NaHCO3
(30 mL) was added and the mixture was stirred at room temperature for
15 min and the aqueous phase was extracted with EtOAc (4 × 15 mL).
The organic layers were collected, dried (MgSO4) and evaporated under
diminished pressure to afford a crude (syrup) which as purified by flash
chromatography on silica gel (EtOAc). Pure 18a (195 mg, 80%) was
obtained as white solid, Rf 0.35 (9:1 EtOAc-MeOH), mp = 115–118 °C
(chrom); [α]D −41.5 (c 1.0 in CHCl3). The NMR analysis (1H and 13C)
of the 18a was rather complex due to the presence of the two rotational
isomers of the amide bond in the ratio of 70:30, measured on the re-
lative integrals of two CHMe2 protons at δ 2.67 and 3.16 respectively.
A solution of azafuranose derivative 15a (48 mg, 0.131 mmol) in
80% aq AcOH (v/v, 2.6 mL) was stirred at 40 °C until the TLC analysis
(9:1 EtOAc-MeOH) revealed the complete reaction of the starting ma-
terial with formation of a major slower moving products (Rf 0.40). After
1 h the solution was then cooled to room temperature and co-evapo-
rated with toluene (6 × 5 mL) under diminished pressure. Purification
of the crude residue by flash chromatography on silica gel (95:15
EtOAc-MeOH) gave pure of azafuranose derivatives 16a (38 mg,
88.8%) as colourless syrup, Rf 0.30 (95:15 EtOAc-MeOH), [α]D −23.9
(c 1.1 in CHCl3), 1H NMR (250.13 MHz, CD3OD): δ 7.28 (m, 5H, Ar-H),
3.16, 2.96 (AB system, 2H, JA,B 13.6 Hz, CH2Ph), 4.11 (dd, 1H, J3,4
4.8 Hz, J2,3 6.8 Hz, H-3), 4.00 (m, 1H, H-2), 3.99 (dd, 1H, J2′,A 7.3 Hz,
J
2′,B 8.2 Hz, CHN), 3.78 (m, 1H, H-5), 3.67 (s, 3H, OMe), 3.55 (dd, 1H,
10

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
By recording the spectra at 80 °C (CD3CN-D2O), coalescence was ob-
served in the 13C spectrum, while 18a-major and 18a-minor are still
evident in the 1H spectrum. Selected 1H NMR (400 MHz, CD3CN-D2O,
20 °C) signals: 18a-major: δ 3.72 (dd, 1H, J1α,1β 11.2 Hz, J1β,2 4.3 Hz,
H-1β), 2.67 (m, 1H, (CHMe2), 1.35,1.28 (2 s, each 3H, CMe2), 1.01 (d,
3H, Jvic 6.5 Hz, CHMe2), 0.98 (d, 3H, Jvic 6.5 Hz, CHMe2); 18a-minor: δ
4.46 (m, 1H, H-4), 4.36 (m, 1H, H-5), 3.16 (m, 1H, CHMe2), 1.37, 1.31
(2 s, each 3H, CMe2), 1.04 (d, 3H, Jvic 5.8 Hz, CHMe2), 0.96 (d, 3H, Jvic
5.8 Hz, CHMe2); 13C NMR (100 MHz, CD3CN-D2O, 20 °C): 18a-major: δ
180.1 (C]O), 109.9 (CMe2), 77.5 (C-2), 77.3 (C-3), 74.8 (C-5), 70.0 (C-
6), 62.0 (C-4), 54.0 (C-1), 33.5 (CHMe2), 27.1, 26.6 (CMe2), 20.3, 19.3
(CHMe2); 18a-minor: δ 180.1 (C]O), 109.5(CMe2), 77.3 (C-2), 77.1
(C-3), 74.6 (C-5), 66.4 (C-6), 60.2 (C-4), 52.5 (C-1), 31.9 (CHMe2),
27.3, 56.0 (CMe2), 20.5, 19.6 (CHMe2); 13C NMR (50.33 MHz, CD3CN-
D2O, 80 °C): δ 179.8 (C]O), 109.8 (CMe2), 78.0 (C-2), 77.4 (C-3), 75.1
(C-5), 69.7 (C-6), 61.7 (C-4), 53.7 (C-1), 33.0 (CHMe2), 27.1, 26.1
(CMe2), 20.3, 19.2 (CHMe2). Anal. Calcd for C13H23NO5: C, 57.13; H,
8.48%; N, 5.18%; Found: C, 57.10; H, 8.45%; N, 5.14%.
diminished pressure. Purification of the crude residue (596 mg) by flash
chromatography on silica gel (EtOAc) gave pure of azafuranose deri-
vatives 18a (263 mg, 37% calculated from 11a + 11b) and 18b
(281 mg, 40% calculated from 11a + 11b) having NMR parameters
identical to those of the sample prepared above.
4.1.17. N-(2-methylpropanoyl)-1,4-dideoxy-1,4-immino-D-glucitol (19a)
A solution of azafuranose derivative 18a (55.2 mg, 0.20 mmol) in
80% aq AcOH (v/v, 3.0 mL) was stirred at 40 °C until the TLC analysis
(9:1 EtOAc-MeOH) revealed the complete reaction of the starting ma-
terial with formation of a major slower moving products (Rf 0.18). After
40 min the solution was then cooled to room temperature and co-eva-
porated with toluene (6 × 5 mL) under diminished pressure.
Purification of the crude residue by flash chromatography on silica gel
(9:1 EtOAc-MeOH) gave pure of azafuranose derivative 19a (38.7 mg,
83%) as white foam, Rf 0.18 (9:1 EtOAc-MeOH). The NMR analysis (1H
and 13C, CD3OD) of 19a, showed the presence of the two rotational
isomers of the amide bond 19a-major and 19a-minor in the ratio 9:1
calculated on the basis of the signals related to the C-6 carbons at δ 64.9
and 64.6 respectively. The 1H NMR analysis of the 19a allows to assign
only the chemical shifts of the proton related to the rotational isomer
19a-major. 1H NMR (250.13 MHz, CD3OD, at 20 °C) of 19a-major: δ
4.28–4.00 (m, 4H, H-2, H-3, H-4, H-5), 3.75 (dd, 1H, J1α,1β 11.3 Hz,
J1β,2 4.1 Hz, H-1β), 3.63 (dd, 1H, J6a,6b 11.3 Hz, J5,6b 8.5 Hz, H-6b),
3.56 (dd, 1H, J1α,2 2.0 Hz, H-1α), 3.50 (dd, 1H, J5,6a 6.5 Hz, H-6b), 2.46
(m, 1H, CHMe2), 1.13 (d, 3H, Jvic 6.8 Hz, CHMe2), 1.09 (d, 3H, Jvic
6.7 Hz, CHMe2); 13C NMR (62.9 MHz, CD3OD, at 20 °C) of 19a-major: δ
180.3 (C]O), 77.5, 75.6, 72.7 (C-2, C-3, C-5), 63.2 (C-4), 64.9 (C-6),
54.0 (C-1), 33.6 (CHMe2), 19.7, 18.9 (CHMe2); of 19a-minor: δ 180.3
(C]O), 79.4, 76.2 74.9 (C-2, C-3, C-5), 64.6 (C-6), 61.8 (C-4), 52.3 (C-
1), 33.4 (CHMe2), 20.1, 19.6 (CHMe2). Anal. Calcd for C10H19NO5: C,
51.49; H, 8.21%; N, 6.00%; Found: C, 51.46; H, 8.17%; N, 5.97%.
4.1.15. N-(2-methylpropanoyl)-5,6-O-isopropylidene-1,4-dideoxy-1,4-
immino-D-galactitol (18b)
To a solution of 17b (250 mg, 1.23 mmol) in 3:1 mixture of MeOH-
H2O (v/v, 3.3 mL), Et3N (0.22 mL, 1.60 mmol) and isobutyric anhydride
(0.41 mL, 2.46 mmol) were added and the solution was stirred at room
temperature until TLC analysis (8:2 EtOAc-MeOH) showed the com-
plete disappearance of the starting material. After 2 h, satd aq NaHCO3
(30 mL) was added and the mixture was stirred at room temperature for
15 min and the aqueous phase was extracted with EtOAc (4 × 15 mL).
The organic layers were collected, dried (MgSO4) and evaporated under
diminished pressure to afford a crude (syrup) which was purified by
flash chromatography on silica gel (EtOAc). Azafuranose derivatives
18b (252 mg, 75%) as obtained pure as white solid, Rf 0.46 (9:1 EtOAc-
MeOH), mp 106–108 °C (chrom); [α]D + 110.5 (c 1.0 in CHCl3). The 1H
NMR analysis of 18b showed the presence of the two rotational isomers
of the amide bond in the ratio of 60:40, measured on the relative in-
tensities of two C-1 signals at δ 53.5 and 55.1 respectively. By recording
the spectra at 80 °C (CD3CN-D2O) coalescence was observed in the 13C
spectrum. 13C NMR (100 MHz, CD3CN-D2O, 20 °C): 18b-major: δ 180.1
(C]O), 110.5 (CMe2), 78.9, 75.8, 74.8 (C-2, C-3, C-5), 69.6 (C-4), 67.5
(C-6), 53.5 (C-1), 32.3 (CHMe2), 26.4, 25.3 (CMe2), 20.0, 19.3
(CHMe2); 18b-minor: δ 179.6 (C]O), 109.9 (CMe2), 79.2, 77.3, 76.3
(C-2, (C-3, C-5), 67.1 (C-6), 64.5 (C-4), 55.1 (C-1), 32.8 (CHMe2), 26.3,
25.7 (CMe2), 20.0, 18.7 (CHMe2); 13C NMR (50.33 MHz, CD3CN-D2O,
80 °C): δ 179.9 (C]O), 110.5 (CMe2), 79.9, 77.2, 76.3 (C-2, C-3, C-5),
69.6 (C-6), 64.5 (C-4), 54.8 (C-1), 32.8 (CHMe2), 26.8, 25.9 (CMe2),
20.3, 19.3 (CHMe2). Anal. Calcd for C13H23NO5: C, 57.13; H, 8.48%; N,
5.18%; Found: C, 57.11; H, 8.44%; N, 5.15%.
4.1.18. N-(2-methylpropanoyl)-1,4-dideoxy-1,4-immino-D-galactitol
(19b)
A solution of azafuranose derivative 18b (58.9 mg, 0.22 mmol) in
80% aq AcOH (v/v, 3.2 mL) was stirred at 40 °C until the TLC analysis
(9:1 EtOAc-MeOH) revealed the complete reaction of the starting ma-
terial with formation of a major slower moving products (Rf 0.23). After
40 min the solution was then cooled to room temperature and co-eva-
porated with toluene (6 × 5 mL) under diminished pressure.
Purification of the crude residue by flash chromatography on silica gel
(9:1 EtOAc-MeOH) gave pure of azafuranose derivative 19b (43 mg,
85%) as white foam, Rf 0.23 (9:1 EtOAc-MeOH). The NMR analysis (1H
and 13C, CD3OD) of 19a, showed the presence of the two rotational
isomers of the amide bond 19b-major and 19b-minor in the ratio
75:25 calculated on the basis of the signals related to the C-1 carbons at
δ 55.8 and 54.1 respectively. 1H NMR (250.13 MHz, CD3OD, at 20 °C) of
19b-major: δ 4.17 (m, 1H, H-4), 4.14 (m, 1H, H-3), 4.10 (m, 1H, H-2),
4.02 (m, 1H, H-1β), 3.95 (m, 1H, H-5), 3.53–3.45 (m, 3H, H 1α, H-6a, H
6b), 2.82 (m, 1H, CHMe2), 1.12 (d, 3H, Jvic 6.8 Hz, CHMe2), 1.08 (d, 3H,
4.1.16. Preparation of azafuranose derivatives 18a and 18b from mixture
of 11a and 11b
To a solution of 2:3 mixture of azafuranose derivatives 11a and 11b
(757 mg, 2.58 mmol) in absolute EtOH (80 mL), 10% Pd on charcoal
(450 mg) was added and the mixture was stirred at room temperature
under H2 (atmospheric pressure) until TLC analysis (8:2 EtOAc-MeOH)
showed the complete disappearance of the starting material (48 h). The
suspension was filtered through a small layer of Celite®, evaporated
under diminished pressure gave a crude product (525 mg) constituted
(NMR) by a mixture of azafuranose derivatives 17a and 17b in the ratio
of 2:3. To a solution of crude mixture of 17a and 17b (530 mg) in 3:1
mixture of MeOH-H2O (v/v, 7.0 mL), Et3N (0.47 mL, 1.60 mmol) and
isobutyric anhydride (0.41 mL, 2.46 mmol) were added and the solution
was stirred at room temperature until TLC analysis (8:2 EtOAc-MeOH)
showed the complete disappearance of the starting material (2 h). Satd
aq NaHCO3 (15 mL) was added, stirred at room temperature (15 min),
the aqueous phase was extracted with CH2Cl2 (4 × 15 mL) and the
organic layers were collected, dried (MgSO4) and evaporated under
J
vic 6.7 Hz, CHMe2); of 19b-minor: δ 4.05 (m, 1H, H-1β), 4.10 (m, 1H,
H-2), 4.05 (m, 1H, H-3), 3.95 (m, 2H, H-4, H-5), 3.65 (dd, 1H, J6a,6b
12.1 Hz, J5,6b 3.5 Hz, H-6b), 3.58 (dd, 1H, J5,6a 4.4 Hz, H-6a), 3.22 (bd,
1H, J1α,1β 12.2 Hz, H-1a), 3.17 (m, 1H, CHMe2), 1.10 (d, 3H, Jvic 6.5 Hz,
CHMe2), 1.04 (d, 3H, Jvic 7.0 Hz, CHMe2); 13C NMR (62.9 MHz, CD3OD,
at 20 °C) of 19b-major: δ 181.2 (C]O), 76.5 (C-2), 75.6 (C-3), 74.7 (C-
5), 67.7 (C-4), 65.0 (C 6), 55.8 (C 1), 33.4 (CHMe2), 20.0, 18.7
(CHMe2); of 19b-minor: δ 180.2 (C]O), 79.8 (C-3), 75.8 (C-2), 71.5
(C-5), 68.8 (C-4), 64.6 (C 6), 54.1 (C 1), 32.8 (CHMe2), 19.3, 19.9
(CHMe2). Anal. Calcd for C10H19NO5: C, 51.49; H, 8.21%; N, 6.00%;
Found: C, 51.47; H, 8.18%; N, 5.96%.
4.1.19. N-benzyl-1,4-dideoxy-1,4-immino-D-galactitol (20b)
A solution of azafuranose derivative 11b (58.9 mg, 0.20 mmol) in
80% aq AcOH (v/v, 3.2 mL) was stirred at 40 °C until the TLC analysis
11

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
(8:2 EtOAc-MeOH) revealed the complete reaction of the starting ma-
terial with formation of a major slower moving products (Rf 0.40). After
40 min the solution was then cooled to room temperature and co-eva-
porated with toluene (6 × 5 mL) under diminished pressure.
Purification of the crude residue by flash chromatography on silica gel
(8:2 EtOAc-MeOH) gave pure of azafuranose derivative 20b (47 mg,
90%) as white solid, Rf 0.40 (8:2 EtOAc-MeOH); mp 133–135 °C (from
EtOH); [α]D −25.5 (c 1.0 in CHCl3). 1H NMR (200.13 MHz, CD3OD): δ
7.39–7.20 (m, 5H, Ar-H), 4.20, 3.52 (AB system, 2H, JA,B = 13.6 Hz,
CH2Ph), 4.09 (m, 1H, H-3), 3.95–3.84 (m, 2H, H-2, H-5), 3.72 (dd, 1H,
5.40 (m, 2H, H-3, H-5), 5.21 (m, 1H, H-2), 4.61 (dd, 1H, J4,5 7.1 Hz, J3,4
5.9 Hz, H-4), 4.18 (dd, 1H, J6a,6b 11.9 Hz, J5,6b 3.8 Hz, H-6b), 4.09 (dd,
1H, J5,6a 8.1 Hz, H-6a), 3.70 (dd, 1H, J1β,1α 11.8 Hz, J1β,2 6.0 Hz, H-1β),
3.39 (dd, 1H, J1α,2 4.0 Hz, H-1α), 2.04, 2.02, 2.00, 1.97, 1.96 (5 s, each
3H, 5 × MeCO); 13C NMR (50.33 MHz, CDCl3 at 20 °C) of 10a-major: δ
74.3, 73.3 69.8 (C-2, C-3, C-5), 63.5 (C-6), 56.1 (C-4), 50.6 (C-1), of
10a-minor: δ 74.6, 73.38, 70.9 (C-2, C-3, C-5), 62.3 (C-6), 58.2 (C-4),
48.2 (C-1); cluster of signals for 10a-minor and 10a-major
170.6–169.4 (C]O), 22.2 (MeCON), 20.9–20.6 (MeCOO).
J
6a,6b 11.1 Hz, J5,6b 5.6 Hz, H-6b), 3.68 (dd, 1H, J5,6a 6.2 Hz, H-6a), 2.91
4.1.22. N-acetyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-
galactitol (10b)
(dd, 1H, J4,5 4.6 Hz, J3,4 2.7 Hz, H-4), 2.86 (m, 1H, H-1β), 2.71 (dd, 1H,
J1α,1β 10.7 Hz, J1α,2 4.4 Hz, H-1α); 13C NMR (62.9 MHz, CD3OD,): δ
138.6 (AreC), 128.4, 127.9, 126.7 (AreCH), 79.2 (C-3), 75.7 (C-2),
73.3 (C-4), 71.2 (C-5), 63.7 (C-6), 60.7 (CH2Ph), 58.9 (C-1). Anal. Calcd
for C13H19NO4: C, 61.64; H, 7.56%; N, 5.53%; Found: C, 61.61; H,
7.52%; N, 5.50%.
Routine acetylation of azafuranose derivative 21b (140 mg,
0.41 mmol) with 1:2 Ac2O-Py mixture (12.0 mL) gave after flash chro-
matography (7:2:1 Et2O-EtOAc- iPrOH + 0.1% Et3N) pure 10b
(139 mg, 90%) as a clear syrup, Rf 0.32 (EtOAc), [α]D + 77.5 (c 1.07 in
CHCl3). The NMR analysis carried out in various solvents (CDCl3 or
CD3OD or C6D6) showed a mixture of the two rotational isomers of the
amide bond 10b-minor and 10b-major in the ratio of 4:6, measured in
CDCl3 on the relative intensities of the H-1 signals at δ 3.28 and 3.51
respectively. 1H NMR (400 MHz, CDCl3) of 10b-major: δ 5.43 (bt, 1H,
4.1.20. 2,3,5,6-Tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-glucitol (21a)
and 2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-galactitol (21b)
To a solution of mixture of azafuranose derivatives 7a and 7b
(402 mg, 0.96 mmol) in absolute EtOH (30 mL), 10% Pd on charcoal
(180 mg) was added and the mixture was stirred at room temperature
under H2 (atmospheric pressure) until the starting compound was
completely reacted (TLC, EtOAc, 2 h). The suspension was diluted with
MeOH (20 mL), filtered over a pad of Celite®, washed with MeOH, and
the combined organic phases were concentrated at diminished pressure.
Purification of crude syrup (332 mg) by flash chromatography on silica
gel (7:2:1 Et2O-EtOAc- iPrOH + 0.1% Et3N) gave pure 21a (99 mg,
30%) and 21b (171 mg, 52%).
J
J
4,5 = J5,6a 7.3 Hz, J5,6b 3.2 Hz, H-5), 4.40 (d, 1H, H-4), 4.35 (dd, 1H,
6a,6b 12.3 Hz, H-6b), 4.07 (dd, 1H, H-6a), 4.01 (dd, 1H, J1β,1α 12.1 Hz,
J1β,2 6.7 Hz, H-1β); 3.43 (dd, 1H, J1α,2 3.4 Hz, H-1α); of 10b-minor: δ
5.31 (ddd, 1H, J4,5 8.3 Hz, J5,6a 5.9 Hz, J5,6b 4.4 Hz, H-5), 4.34 (dd, 1H,
J6a,6b 12.5 Hz, H-6b), 4.20 (dd, 1H, J1β,1α 14.0 Hz, J1β,2 7.0 Hz, H-1β),
4.05 (m, 1H, H-6a), 3.94 (d, 1H, H-4), 3.23 (dd, 1H, J1α,2 2.0 Hz, H-1α);
cluster of signals for 10b-minor and 10b-major: δ 5.11–5.02 (m, 1H,
H-2, H-3), 2.05–2.1.97 (m, 5 × 15H, MeCO); 13C NMR (100 MHz,
CDCl3) of 10b-major: δ 77.3, 75.7 (C-2, C-3), 70.3 (C-5), 63.5 (C-6),
61.6 (C-4), 52.4 (C-1), 22.3 (MeCON); of 10b-minor: δ 77.9, 74.3 (C-2,
C-3), 68.6 (C-5), 63.5 (C-4), 62.9 (C-6), 50.6 (C-1), 22.0 (MeCON);
cluster of signals for 10b-minor and 10b-major: δ170.4–169.5 (C]O),
20.8–20.5 (MeCOO). Anal. Calcd for C16H23NO9: C, 51.47; H, 6.21%; N,
3.75%; Found: C, 51.45; H, 6.17%; N, 3.73%.
2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-glucitol (21a) as a
clear syrup, Rf 0.32 (7:2:1 Et2O-EtOAc- iPrOH + 0.1% Et3N), [α]D
−18.3 (c 1.0 in CHCl3), 1H NMR (200.13 MHz, C6D6): δ 5.60 (bd, 1H,
J
3,4 4.2 Hz, H-3), 5.24 (ddd, 1H, J4,5 7.1 Hz, J5,6a 5.4 Hz, J5,6b 2.5 Hz, H-
5), 4.98 (ddd, 1H, J1β,2 5.2 Hz, J1α,2 2.3 Hz, H-2), 4.61 (dd, 1H, J6a,6b
12.2 Hz, H-6b), 4.12 (dd, 1H, H-6a), 3.42 (dd, 1H, H-4), 3.17 (dd, 1H,
J1β,1α 12.9 Hz, H-1β), 2.66 (dd, 1H, H-1α),1.79, 1.75, 1.67, 1.53 (4 s,
each 3H, 4 × MeCO); 13C NMR (50.33 MHz, C6D6): δ 170.2, 169.7,
169.1, 169.0 (4 × C]O), 78.5, 75.9 (C-2, C-3), 69.8 (C-5), 64.4 (C-6),
60.3 (C-4), 51.8 (C-1), 20.6, 20.4, 20.3, 20.2 (4 × MeCO). Anal. Calcd
for C14H21NO8: C, 50.75; H, 6.39%; N, 4.23%; Found: C, 50.72; H,
6.36%; N, 4.20%.
4.1.23. 1,4-Dideoxy-1,4-immino-D-galactitol hydrochloride (22)
A solution of 10b (56 mg, 0.15 mmol) or 21b (52 mg, 0.15 mmol) in
aq HCl (2 M, 5 mL) was stirred at 100 °C until the TLC analysis (EtOAc)
revealed the complete disappearance of the starting material (Rf 0.32).
After 15 min the mixture was repeatedly co-evaporated with toluene
(4 × 20 mL) under diminished pressure. The trituration of the crude
product with Et2O afforded pure the 1,4-dideoxy-1,4-immino-D-ga-
2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-galactitol (21b) as
colourless syrup, Rf 0.24 (7:2:1 Et2O-EtOAc- iPrOH + 0.1% Et3N),
[α]D + 44.3 (c 1.0 in CHCl3); 1H NMR (200.13 MHz, C6D6): δ 5.41 (m,
1H, H-3), 5.02 (m, 1H, H-5), 4.54 (dd, 1H, J6a,6b 11.6 Hz, J5,6b 4.2 Hz,
H-6b), 4.48 (m, 1H, H-2), 4.37 (dd, 1H, J5,6a 5.0 Hz, H-6a), 4.27 (m, 1H,
H-4), 3.30 (dd, 1H, J1α,2 2.3 Hz H-1α), 3.06 (dd, 1H, J1β,1α 12.1 Hz,
lactitol hydrochloride (22) (22 mg, 87%), as
a white solid; mp
98–101 °C; [α]D −24.1 (c 0.8 in MeOH). Lit. [24] [α]D −25.3 (c 1.0 in
MeOH), mp 102 °C (from CHCl3-MeOH). The NMR spectroscopic data
was completely agreed with those reported in literature [24].
J1β,2 5.8 Hz, H-1β),1.80, 1.71, 1.58, 1.54 (4 s, each 3H,4 × MeCO); 13
C
4.2. Biological experimental
NMR (50.33 MHz, C6D6): δ 172.6, 170.6, 169.4, 169.3 (4 × C]O),
76.5, 75.4 (C-2, C-3), 71.7 (C-5), 67.1 (C-6), 66.7 (C-4), 52.3 (C-1),
21.9, 20.5, 20.2, 20.1 (4 × MeCO). Anal. Calcd for C14H21NO8: C,
50.75; H, 6.39%; N, 4.23%; Found: C, 50.74; H, 6.37%; N, 4.21%.
Human recombinant aldose reductase from Escherichia coli, alpha-
glucosidase from Saccharomices cerevisiae, NADPH, D,L-glyceraldehyde,
p-nitrophenyl-α-D-glucopyranoside and phosphate buffer were pur-
chased from Sigma Aldrich. Epalrestat was obtained from Haorui
Pharma-Chem Inc., NJ, USA; 1-deoxynojirimycin chlorohydrate
(DNJ·HCl) was purchased from Carbosynth, UK.
4.1.21. N-Acetyl-2,3,5,6-tetra-O-acetyl-1,4-dideoxy-1,4-immino-D-glucitol
(10a)
Routine acetylation of azafuranose derivative 21a (80 mg,
0.23 mmol) with 1:2 Ac2O-Py mixture (9.0 mL) gave after flash chro-
4.2.1. ALR2 enzymatic assay
i
matography (7:2:1 Et2O-EtOAc- PrOH + 0.1% Et3N) pure 10a (79 mg,
The enzyme activity was determined spectrophotometrically at
340 nm by monitoring the change in absorbance of the NADPH cofactor,
according to a previously reported procedure [34]. The reaction mixture
contained 100 µL sodium phosphate buffer pH = 6.2 (0.1 M), 50 µL
NADPH (0.15 mM), 50 µL ALR2 (0.50 µM) and 50 µL D,L-glyceraldehyde
(10 mM) in a 96-multiwell plate. The reaction was monitored for 5 min at
30 °C, and the variation of pyridine coenzyme concentration versus time
was analysed through the WorkOut program from PerkinElmer.
91%) as a clear syrup, Rf 0.36 (EtOAc), [α]D + 52.6 (c 0.8 in CHCl3);
Lit. [22] + 53 (CHCl3). The NMR analysis (CDCl3) of 10a, showed the
presence of the two rotational isomers of the amide bond in the ratio
9:1 calculated on the basis of the signals related to the C-1 carbons at δ
50.6 and 48.2 respectively. The 1H NMR analysis of the 10a allows to
assign only the chemical shifts of the proton related to the rotational
isomer 10a-major. 1H NMR (400 MHz, CDCl3 at 20 °C) of 10a-major: δ
12

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
4.2.2. ALR2 enzymatic inhibition
culture plates and allowed to stabilize overnight in a 5% CO2 incubator
at 37 °C. The cells were then cultured in 24.5 mM (NG) or 55 mM (HG)
glucose-containing medium in the absence or presence of 17b (10-6 M)
for 24 h. After the induction of HG condition, adherent and non-ad-
herent cells were collected by centrifugation and incubated with
Guava® ViaCount® Reagent or Muse™ Annexin V & Dead Cell Kit
(Millipore), the reagent distinguishes viable and non-viable cells based
on differential permeabilities of two DNA-binding dyes in the Guava
ViaCount Reagent. The nuclear dye stains only nucleated cells, while
the viability dye brightly stains dying cells. This proprietary combina-
tion of dyes enables the Guava ViaCount Assay to distinguish viable,
apoptotic, and dead cells. Debris is excluded from results based on
negative staining with the nuclear dye. The data collected were ana-
lyzing with Muse 1.5 Software (Millipore).
Inhibition studies were performed by adding the test compounds
(25 μL, 100 μM) to the above mentioned reaction mixture. All the
compounds were initially solubilized in DMSO, and then diluted with
water, before the addition to the assay test. DMSO concentration in the
test solutions never exceed 4%, and proved to have no inhibitory effects
on the target enzyme. In order to avoid absorption due to the tested
molecules, a blank containing all the components, with the exception of
the substrate, was analysed. Assays were conducted in triplicate.
Epalrestat was used as the reference compound.
4.2.3. AG enzymatic assay
The enzyme activity was determined spectrophotometrically at
405 nm following a literature procedure with some modification [35],
by monitoring the change in absorbance of p-nitrophenyl α-D-gluco-
pyranoside, which is converted into sodium p-nitrophenoxyde by re-
action with the target enzyme. The reaction mixture contained 100 µL
sodium phosphate buffer pH = 6.8 (0.1 M), 25 µL enzyme (0.2 U/mL)
and 25 µL of p-nitrophenyl α-D-glucopyranoside (0.375 mM), in a 96-
multiwell plate. All the reagents were incubated at 37 °C for 10 min.
The reaction was then stopped with 100 µL of saturated Na2CO3. Ab-
sorbance of the test solutions were analysed with the program Wallac
1420 of PerkinElmer.
4.2.8. Immunocytochemistry
The cells were directly washed in PBS (1X; KCl 2.67 mM, KH2PO4
1.47 mM, NaCl 139.9 mM, Na2HPO4, 8.1 mM, pH 7.4) and fixed in PAF
for 1 min. Once the cells were fixed, they were washed in PBS twice for
10 min and incubated in with PBS containing 0.03% Triton and 5% of
bovine serum albumin (1 h at room temperature), and incubated with
the primary antibody anti-Nrf2 (1:50; Sigma-Aldrich) overnight at 4 °C.
After incubation with the primary antibody, the samples were washed
three times in PBS and subsequently incubated with the secondary goat
anti-rabbit immunoglobulin G (IgG; 1:1000; Vector Laboratories) 488-
Alexa fluor conjugated for 2 h at room temperature. After washing in
PBS, the samples were incubated with Ethidium Bromide (1:5000,
Sigma-Aldrich) for 5 min. Finally, after washing twice in PBS, the
chambers were detached from the slide and mounted with Vectashield
mounting medium (Vector Laboratories). This immunolabeling pro-
tocol was modified respect to Wang et al., 2019 [36]. The slides were
examined with a fluorescent microscope (Nikon E-Ri). Images were
processed with ImageJ software.
4.2.4. AG enzymatic inhibition
Inhibition studies were performed by adding the test compounds
(25 µL, 100 μM) to the above mentioned reaction mixture. Compounds
found to be active were tested at additional concentrations between 10-
and 10-7 M. All the compounds were initially solubilized in DMSO,
5
and then diluted with water, before the addition to the assay test.
DMSO concentration in the test solutions never exceed 4%, and proved
to have no inhibitory effects on the target enzyme. In order to avoid
absorption due to the tested molecules, a blank containing all the
components, with the exception of the substrate, was analysed. Assays
were conducted in triplicate. Positive control was represented by 1-
deoxynojirimycin chlorohydrate (DNJ·HCl). IC50 data were obtained by
linear regression analysis of the log dose-response curve.
4.2.9. Western blot
For western blot analysis, the proteins were resolved using dena-
turing sodium dodecyl sulfate-poly-acrylamide gel electrophoresis
(SDS-PAGE) and transferred to PVDF membranes (Bio-Rad, Madrid,
Spain). Levels of Sod1 proteins were normalized by the levels of their
corresponding total proteins by using the Stain Free Technology
(BioRad) [31]. Antibody anti-Sod1 was purchased from Santa Cruz
Biotechnology (Palo Alto, CA) and secondary antibody anti-mouse HRP
conjugated was purchased from (Sigma-Aldrich). The chemilumines-
cence was analyzed using the Chemidoc Quantified (Bio-Rad) and the
images obtained analyzed with ImageLAb Software.
4.2.5. 661w cell culture
The 661 W cell line, derived from immortalized cone photoreceptors
(provided by Muayyad Al-Ubaidi, University of Oklahoma), was
maintained in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen,
ThermoFisher Scientific, Waltham, MA) containing 4.5 g/l (24.5 mM)
glucose and supplemented with 10% (v/v) heat-inactivated FBS, 1%
Lglutammine and 1% pen/strep. For the experiments, confluent cells
were maintained in 24.5 mM glucose (normoglycemic condition) or
cultured in 55 mM glucose (hyperglycemic condition) during 24 h in
the presence or absence of 17b (10-6 M).
Acknowledgements
The authors would like to thank Professor Giorgio Catelani, who
inspired the preparation of polyhydroxylated pyrrolidine, and the
postgraduate fellow Martina Landi (University of Pisa), who con-
tributed to the synthesis of compounds 5, 8a,b, 9a,b and 10a,b.
4.2.6. In vitro ALR2 enzymatic inhibition
Cells were seeded in 24-well tissue culture plates and allowed to
stabilize overnight in a 5% CO2 incubator at 37 °C. The cells were then
cultured in 24.5 mM (NG) or 55 mM (HG) glucose-containing medium
in the absence or presence of 17b (10-6 M) for 24 h; after that the cells
were mechanically lysate and the supernatant were collected by cen-
trifugation and incubate with Aldo-Keto Reductase (AKR) Activity
Assay Kit (AbCam) in a 96-well for 2 h, and absorbance was measured
with a spectrophotometer (NanoQuant) at 450 nm. The kit works by
measuring the absorbance of the NADH cofactor, which is consumed
during the glucose conversion to sorbitol mediated by the ALR2 en-
zyme. Therefore, a reduction in the absorbance value corresponds to an
increased activity of the enzyme itself.
Funding
This work was supported in part by the University of Pisa (grant
number PRA_2017_51).
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bioorg.2019.103298.
4.2.7. Analysis of cellular viability and detection of apoptosis
References
Quantification of cellular viability and apoptosis were performed
using flow cytometry methods. Cells were seeded in 24-well tissue
[1] W.R. Rowley, C. Bezold, Y. Arikan, E. Byrne, S. Krohe, Popul Health Manage. 20
13

L. Guazzelli, et al.
BioorganicChemistry92(2019)103298
(2017) 6–12.
[19] D. Cuffaro, M. Landi, F. D'Andrea, L. Guazzelli, Carbohydr. Res. 482 (2019) 107744.
[2] M. Browlee, Diabetes 54 (2005) 1615–1625.
[20] R.V. Stevens, Acc. Chem. Res. 17 (1984) 289–296.
[3] M. Hanefeld, F. Schaper, The Role of Alpha-Glucosidase Inhibitors (Acarbose), in:
C.E. Mogensen (Ed.), Pharmacotherapy of Diabetes: New Developments, Springer,
Boston, MA, 2007.
[21] E.W. Baxter, A.B. Reitz, J. Org. Chem. 59 (1994) 3175–3185.
[22] Y. Kuszmann, L. Kiss, Carbohydr. Res. 153 (1986) 45–53.
[23] C. Liu, L. Hou, A. Meng, G. Han, W. Zhang, S. Jiang, Carbohydr. Res. 429 (2016)
135–142.
[4] W.H. Tang, K.A. Martin, J. Hwa, Front Pharmacol. 3 (2012) 87.
[5] L.K. Campbell, D.E. Baker, R.K. Campbell, Ann. Pharmacother. 34 (2000)
1291–1301.
[24] R.E. Lee, M.D. Smith, J.R. Nash, C.R. Griffiths, M. McNeil, K.R. Grewal, W. Yan,
S.G. Besra, J.P. Brennan, W.J.G. Fleet, Tetrahedron Lett. 38 (1997) 6733–6736.
[25] T.R. Hoye, M.K. Renner, J. Org. Chem. 61 (1996) 2056–2064.
[26] G. Han, M.G. LaPorte, M.C. McIntosh, S.M. Weinreb, J. Org. Chem. 61 (1996)
9483–9493.
[6] L. Quattrini, C. La Motta, Expert Opin. Ther Pat. 29 (2019) 199–213.
[7] R. Bianchini, M. Rolla, J. Isaad, G. Catelani, L. Guazzelli, M. Corsi, M. Bonanni,
Carbohydr. Res. 356 (2012) 104–109.
[8] L. Guazzelli, R. Ulc, L. Rydner, S. Oscarson, Org. Biomol. Chem. 13 (2015)
6598–6610.
[27] T. Iwasaki, H. Horikawa, K. Matsumoto, M. Miyoshi, J. Org. Chem. 44 (1979)
1552–1554.
[9] L. Guazzelli, O. McCabe, S. Oscarson, Carbohydr. Res. 433 (2016) 5–13.
[10] L. Guazzelli, G. Catelani, F. D'Andrea, T. Gragnani, A. Griselli, Eur. J. Org. Chem. 29
(2014) 6527–6537.
[28] T. González, O. Abad, M.C. Santano, C. Minguillón, Synthesis 8 (2004) 1171–1182.
[29] E. Tan, X.-Q. Ding, A. Saadi, N. Agarwal, M.I. Naash, M.R. Al-Ubaidi, Invest.
Ophthalmol. Vis. Sci. 45 (2004) 764–768.
[11] V. Pistarà, A. Corsaro, A. Rescifina, G. Catelani, F. D'Andrea, L. Guazzelli, J. Org.
Chem. 78 (2013) 9444–9449.
[30] A.I. Arroba, A. Mazzeo, D. Cazzoni, E. Beltramo, C. Hernández, M. Porta, R. Simó,
Á.M. Valverde, Mol. Vision 22 (2016) 1522–1531.
[12] L. Guazzelli, G. Catelani, F. D'Andrea, Carbohydr. Res. 345 (2010) 369–376.
[13] F. Balestri, L. Quattrini, V. Coviello, S. Sartini, F. Da Settimo, M. Cappiello,
R. Moschini, A. Del Corso, U. Mura, C. La Motta, Cell Chem Biol. 25 (2018)
1414–1418.
[31] A. Gürtler, N. Kunz, M. Gomolka, S. Hornhardt, A.A. Friedl, K. McDonald,
J.E. Kohn, A. Posch, Anal. Biochem. 433 (2013) 105–111.
[32] K. Bock, C. Pedersen, Adv. Carbohydr. Chem. Biochem. 41 (1983) 27–66.
[33] D.D. Perrin, W.L.F. Armarego, D.R. Perrin, Purification of Laboratory Chemicals,
2nd ed., Pergamon Press, Oxford, 1980.
[14] S. Cosconati, L. Marinelli, C. La Motta, S. Sartini, F. Da Settimo, A.J. Olson,
E. Novellino, J. Med. Chem. 52 (2009) 5578–5581.
[34] A. Ramunno, S. Cosconati, S. Sartini, V. Maglio, S. Angiuoli, V. La Pietra, S. Di
Maro, M. Giustiniano, C. La Motta, F. Da Settimo, L. Marinelli, E. Novellino, Eur. J.
Med. Chem. 51 (2012) 216–226.
[15] D. Horton, J.P. Roski, P. Norris, J. Org. Chem. 61 (1996) 3783–3793.
[16] A. El Nemr, T. Tsuchiya, Carbohydr. Res. 303 (1997) 267–281.
[17] E. Attolino, G. Catelani, F. D’Andrea, M. Landi, Carbohydr. Res. 341 (2006)
2498–2506.
[35] Y. Tao, Y. Zhang, Y. Cehng, Y. Wang, Biomed. Chromatogr. 27 (2013) 148–155.
[36] J. Wang, J. Zhao, X. Cui, B.A. Mysona, S. Navneet, A. Saul, M. Ahuj, N. Lambert,
I.G. Gazaryan, B. Thomas, E.K. Bollinger, S.B. Smitha, Free Radical Biol. Med. 134
(2019) 604–616.
[18] P.L. Barili, G. Berti, G. Catelani, F. D’Andrea, F. De Rensis, L. Puccioni, Tetrahedron
53 (1997) 3407–3416.
14