Synthetic sulfoglycolipids targeting the serine–threonine protein kinase Akt

Article abstract of DOI:10.1016/j.bmc.2016.05.031

The serine–threonine protein kinase Akt, also known as protein kinase B, is a key component of the phosphoinositide 3-kinase (PI3K)–Akt–mTOR axis. Deregulated activation of this pathway is frequent in human tumors and Akt-dependent signaling appears to be critical in cell survival. PI3K activation generates 3-phosphorylated phosphatidylinositols that bind Akt pleckstrin homology (PH) domain. The blockage of Akt PH domain/phosphoinositides interaction represents a promising approach to interfere with the oncogenic potential of over-activated Akt. In the present study, phosphatidyl inositol mimics based on a β-glucoside scaffold have been synthesized as Akt inhibitors. The compounds possessed one or two lipophilic moieties of different length at the anomeric position of glucose, and an acidic or basic group at C-6. Docking studies, ELISA Akt inhibition assays, and cellular assays on different cell models highlighted 1-O-octadecanoyl-2-O-β-D-sulfoquinovopyranosyl-sn-glycerol as the best Akt inhibitor among the synthesized compounds, which could be considered as a lead for further optimization in the design of Akt inhibitors.

Full text of DOI:10.1016/j.bmc.2016.05.031

Accepted Manuscript
Synthetic sulfoglycolipids targeting the serine-threonine protein kinase Akt
Barbara Costa, Milind Dangate, Maria Vetro, Giulia Donvito, Luca Gabrielli,
Loredana Amigoni, Giuliana Cassinelli, Cinzia Lanzi, Michela Ceriani, Luca
De Gioia, Giulia Filippi, Laura Cipolla, Nadia Zaffaroni, Paola Perego, Diego
Colombo
PII:
S0968-0896(16)30355-8
DOI:
http://dx.doi.org/10.1016/j.bmc.2016.05.031
BMC 13015
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
14 January 2016
22 April 2016
17 May 2016
Please cite this article as: Costa, B., Dangate, M., Vetro, M., Donvito, G., Gabrielli, L., Amigoni, L., Cassinelli, G.,
Lanzi, C., Ceriani, M., Gioia, L.D., Filippi, G., Cipolla, L., Zaffaroni, N., Perego, P., Colombo, D., Synthetic
sulfoglycolipids targeting the serine-threonine protein kinase Akt, Bioorganic & Medicinal Chemistry (2016), doi:
http://dx.doi.org/10.1016/j.bmc.2016.05.031
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthetic sulfoglycolipids targeting the serine-threonine protein kinase Akt
Barbara Costa,[a] Milind Dangate,[b][†] Maria Vetro,[b][×] Giulia Donvito,[a] Luca Gabrielli,[a] Loredana
Amigoni,[a] Giuliana Cassinelli,[c] Cinzia Lanzi,[c] Michela Ceriani,[a] Luca De Gioia,[a] Giulia
Filippi,[a] Laura Cipolla,[a] Nadia Zaffaroni,[c] Paola Perego,[c]* and Diego Colombo[b]
[a]
Department of Biotechnology and Biosciences, University of Milano-Bicocca, P.za della
Scienza 2, 20126 Milan, Italy
[b] Department of Medical Biotechnology and Translational Medicine, University of Milano, Via
Saldini 50, 20133 Milan, Italy
[c] Molecular Pharmacology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Via Amadeo
42, 20133 Milan, Italy
[†]
Present address: Amrita Vishwa Vidyapeetham, School of Arts and Sciences, Department of
Chemistry, Amritapuri, Clappana P.O., Kollam – 690 525, Kerala, India
[×]
Present address: Center for Synthesis and Chemical Biology, University College Dublin,
Belfield, Dublin 4, Ireland
*Corresponding author: Paola Perego, paola.perego@istitutotumori.mi.it
ABBREVIATIONS
phosphoinositide 3-kinase, PI3K; PH, pleckstrin homology; 4IP, inositol-(1,3,4,5)-
tetrakisphosphate; PI3P, 3-phosphorylated phosphatidylinositol; PTC, papillary thyroid
carcinoma; SQAG, sulfoquinovosylacylglycerols.

Abstract
The serine-threonine protein kinase Akt, also known as protein kinase B, is a key component of
the phosphoinositide 3-kinase (PI3K)-Akt-mTOR axis. Deregulated activation of this pathway is
frequent in human tumors and Akt-dependent signaling appears to be critical in cell survival.
PI3K activation generates 3-phosphorylated phosphatidylinositols that bind Akt pleckstrin
homology (PH) domain. The blockage of Akt PH domain/phosphoinositides interaction
represents a promising approach to interfere with the oncogenic potential of over-activated Akt.
In the present study, phosphatydyl inositol mimics based on a -glucoside scaffold have been
synthesized as Akt inhibitors. The compounds possessed one or two lipophilic moieties of
different length at the anomeric position of glucose, and an acidic or basic group at C-6. Docking
studies, ELISA Akt inhibition assays, and cellular assays on different cell models highlighted 1-
O-octadecanoyl-2-O--D-sulfoquinovopyranosyl-sn-glycerol as the best Akt inhibitor among the
synthesized compounds, which could be considered as a lead for further optimization in the
design of Akt inhibitors.
KEYWORDS: sulfoquinovose, phosphatydyl inositol analogues, inhibitors, Akt, cancer
2

1. Introduction
Kinase inhibitors have attracted considerable attention over the past decade as anti-cancer
agents and several kinase targeting agents have reached the clinical setting [1-4]. Kinases,
through the phosphorylation of specific protein and non-protein targets, are enzymes involved in
the regulation of many crucial cellular processes such as proliferation, differentiation and
apoptosis [5-7]. Among them, the serine/threonine protein kinase Akt, also known as protein
kinase B, plays a key role as a component of the phosphoinositide 3-kinase (PI3K)-Akt-mTOR
axis, which is implicated in aberrant tumor cell signaling [8-9]. Deregulated activation of this
pathway is a common event in human tumors and Akt-dependent signaling appears to be critical
in cell survival [8]. Thus, inhibitors that target PI3K and its downstream effectors, including Akt,
are relevant for cancer therapy [10-14]. PI3K activation generates 3-phosphorylated
phosphatidylinositols (PI3Ps), that bind Akt pleckstrin homology (PH) domain [15-17]. The
interaction between the PH domain and phosphoinositide induces a conformational change of
Akt, allowing the kinase full activation by two phosphorylations (Thr308 and Ser473) [18].
Thus, the blockage of Akt PH domain/phosphoinositide interaction represents a promising
approach to interfere with the oncogenic potential of overactivated Akt.
Several inhibitors targeting the PH domain have been proposed in the last years, including
phosphoinositide analogues, alkylphospholipids and inositol phosphates [11]. Though synthetic
carbohydrate-based mimics of inositols have already been reported [19], a few examples on the
use of glucose as scaffold for the design of Akt inhibitors have been described [20, 21]. Here, we
present the synthesis and the biological activity of new sulphoglycolipids structurally related to
the natural sulfoquinovosylacylglycerols (SQAG, 1, Figure 1), a family of natural compounds
associated with photosynthetic organisms in which sulfoquinovose (6-deoxy-6-sulfoglucose) is
-linked to the sn-3 position of an acylglycerol [22, 23].
3

SQAG, which inspired this work, have been already considered as promising compounds in
different therapeutic areas, including anti-cancer therapy. In fact, members of this class of
compounds have been described as inhibitors of HIV-reverse transcriptase or of the mammalian
DNA polymerase. Moreover, inhibitory activity on cancer cell proliferation as well as apoptosis
induction and in vivo antitumor and anti-angiogenic effects have been reported [24].
Nevertheless, to our knowledge their Akt inhibitory activity in relation to inhibition of cancer
cell proliferation has not been reported. Differently from SQAG, compounds 2a-c and 3a-c
(Figure 1) are based on a 2-O--D-glucopyranosyl-sn-glycerol scaffold, thus representing a novel
class of potential Akt inhibitors acting as phosphoinositide mimics targeting the Akt PH domain.
The structure of phosphatidylinositol-3-phosphate, a natural ligand of Akt PH domain, (PI3P,
Figure 1) as produced by PI3K, can resemble a suitably modified D-glucose scaffold, where
specific functional groups can be introduced.
Figure 1
Thus, we synthesized a series of D-glucopyranose compounds (Figure 1) endowed with the
following key features needed for the interaction with the PH domain: i) an anionic group
mimicking the phosphate group in position 3 of the inositol (position 6 of the sugar), that should
4

be able to form a salt bridge with a key arginine residue [20]; ii) a -configured substituent at the
anomeric carbon of D-glucopyranose (position 1 of inositol) possessing a hydrogen-donor
heteroatom and a lipophilic group mimicking the diacyl glycerol moiety of phosphoinositides.
The long lipophilic acyl chains of the natural phosphatidylinositol-3-phosphate substrate can be
substituted by shorter hydrophobic groups, since the long fatty acid chain is not needed for
enzyme recognition, but only for membrane anchoring [16]. In addition, the lack of the hydroxyl
group in position 2 of the inositol ring, substituted by the pyranosidic oxygen of the sugar, does
not alter enzyme recognition.
2. Results and Discussion
2.1 Chemistry. Synthesis of inhibitors 2-8
Recently, as part of an exploratory study evaluating the potential anti-tumor-promoting
activities of glycoglycerolipids, we synthesized [25, 26] a series of analogues of SQAG in which
sulfoquinovose is -linked to the 2 position of an acylglycerol (diacylglycerol 2a-c or
monoacylglycerol 3a-c). In the D-glucose-based skeleton, the presence of the stable (non-
hydrolysable) anionic sulfonate group at the 6 position and the -acylglycerol aglycone directly
linked to the sugar (lacking the phosphate bridge), convert such compounds into a sort of hybrid
between PI3P and the SQAG 1. For these reasons, we decided to evaluate the Akt inhibitory
activity of compounds 2a-c and 3a-c. In addition, to obtain further SAR data, we prepared and
tested as potential Akt inhibitors the fully deacylated sulfoquinovosyl-2-O--D-glycerol 4, the
glucoglycerolipid 5 lacking the characteristic sulfonate group of sulfoquinovosides, the methyl
and octyl--D-sulfoquinovosides 6 and 7 lacking the glycerol moiety and the derivative 8
5

carrying a cationic (instead of an anionic) amino group. 1,3-Di-O-acyl-2-O--D-
sulfoquinovopyranosyl-sn-glycerols 2a-c [25], 1,3-di-O-dodecanoyl-2-O--D-glucopyranosyl-sn-
glycerol 5 [25], 1-O-acyl-2-O--D-sulfoquinovopyranosyl-sn-glycerols 3a-c [26] and 6-amino-6-
deoxy-glucoside 8 [27] were prepared as previously reported. The fully deacylated 2-O--D-
sulfoquinovopyranosyl-sn-glycerol 4 was obtained by Zemplén reaction from a small amount of
compound 3a. Methyl -D-sulfoquinovopyranoside 6 [28]was prepared by converting the known
methyl 2,3,4-tri-O-acetyl-6-O-tosyl-β-D-glucopyranoside (9) [29] into the corresponding 6-
thioacetyl glucoside (10) by potassium thioacetate treatment in DMF, following Oxone®
oxidation to the corresponding sulfonate 11 and final acetyl groups removal by Zemplén reaction
(Scheme 1). A similar synthetic scheme was used to obtain the octyl sulfoquinovoside 7. To this
end octyl 2,3,4-tri-O-acetyl-6-O-tosyl-β-D-glucopyranoside 12, obtained one pot from
commercial octyl -D-glucopyranoside by tosyl chloride treatment in pyridine at low temperature
followed by direct acetylation with acetic anhydride, was converted into the corresponding
thioacetate 13, oxidized to the triacetylsulfonate 14 and deacetylated to yield the desired octyl -
D-sulfoquinovopyranoside 7 (Scheme 1).
Scheme 1 i:CH3COSK, DMF; ii: OXONE®, CH3COOH; iii: CH3ONa, CH3OH
6

2.2 Evaluation of Akt inhibition in cell-free assay (ELISA)
The synthesized derivatives 2-8, with the exception of compound 2a that was not soluble in
DMSO, were assayed in vitro for inhibitory activity against Akt, using a cell-free assay, at fixed
concentration of 100 µM (Figure 2). The data showed that, 2b and 3a were the most active
derivatives, possessing long or medium acyl chains and the anionic –SO3- group at carbon 6 of
the sugar moiety. Compounds 4 and 8, the first missing the acyl chains, the second with a
positively-chargeable amino group in place of the sulphonyl group at C-6 were among the least
active ones, in agreement with the docking studies (see below). These data suggest that a
lipophilic moiety mimicking the diacylglycerol group of PI3P is needed for a better interaction
with the kinase (polar glycerol OH groups in 4 are detrimental for activity), whereas cationic
groups could not be accepted by the enzymatic pocket, because of repulsive ionic interaction
between the ammonium group and the Arg side chain [20]. Furthermore, both the acyl and
anionic groups are needed for high activity, the lack of one of them lowering the activity below
13% inhibition (Figure 2). Compound 5, endowed with medium length acyl chains (dodecanoyl
groups), but possessing the –OH instead of the anionic group at C-6 displayed a tenfold lower
activity than the corresponding sulphonate 2b, thereby confirming that the anionic group is
relevant for the activity, as already observed for a glucuronosylglycerol series [30]. In addition,
the length of the acyl chain appears to influence the potency of the sulphonate series 2 and 3. In
fact, the longer the chain, the higher the activity, whereas the number of acyl chains seems to be
less relevant (Figure 2).
7

Figure 2
Moreover, the activity of 2c, 3c and 4 are comparable indicating that the short chains in 2c and
3c have poor influence. Finally the alkyl sulfoquinovosides 6 and 7 lacking the glycerol moiety
exhibited a very low activity. The Akt1/2 kinase allosteric inhibitor (1,3-dihydro-1-(1-((4-(6-
phenyl-1H-imidazo[4,5-g]quinoxalin-7-yl)phenyl)methyl)-4-piperidinyl)-2H-benzimidazol-2-
one trifluoroacetate salt hydrate, indicated as A6730 from Sigma), included in ELISA studies as
reference compound, showed 89% inhibition of Akt, demonstrating the strength of the test. For
the compound 3a a concentration-response analysis was performed (Figure 3). The compound
showed a concentration-dependent significant inhibitory effect of Akt1 activity (IC50 = 31.33 μM,
95% confidence interval 21.18 – 45.71, r2= 0.9484). It is worthy of note that, with the same
specific ELISA test [30], perifosine (a suggested Akt inhibitor targeting the PH domain) [31]
displayed a similar inhibition with an IC50 value of 43.35 µM, whereas the surfactant sodium
dodecyl sulphate (SDS) didn’t exhibit inhibition of Akt thus supporting a specific effect induced
by 3a.
8

Figure 3
To evaluate the selectivity of compound 3a against the Akt PH domain, we tested the effect on
the activity of another kinase containing the PH domain, i.e., BTK (Bruton's tyrosine kinase).
The results indicate that compound 3a as well as the standard Akt1/2 inhibitor A6730 did not
substantially affect the activity of such a kinase, suggesting that the effect on Akt is selective
(Figure 4). On the contrary, the prototypical ATP-competitive kinase inhibitor staurosporine
inhibited BTK activity (Figure 4) besides Akt activity (not shown).
Figure 4
2.3 Docking studies
9

In an attempt to qualitatively define the mechanism by which the compounds tested in vitro
could interact with Akt and inhibit its activity, docking simulations were carried out using the X-
ray crystal structure of the PH domain of human Akt (PDB entry: 1UNQ, 0.98 Å resolution)
[32]. To assess the docking procedure, we performed the first run of docking using 4IP (inositol-
(1,3,4,5)-tetrakisphosphate) as the ligand (in the crystal structure of Akt, the protein is co-
crystallized with this organic cofactor, PDB ID:1UNQ). Since all the studied compounds are
phosphatidylinositol (PIs) analogues, we used as docking space all the spheres generated by
DOCK6 program within 10Å of the PIs binding site. The anchor and growth procedure followed
by the algorithm of DOCK6 program allowed initially the docking of the sugar ring. Hereafter,
little segments of the acyl chains were added to the sugar ring, and minimized to obtain the lower
energy possible. Despite the high conformational variability of the molecules due to the long
acyl chains, the algorithm used in the docking procedure allowed us to obtain robust results and
identify some key features of the substituted sugar ring necessary for an efficient binding. The
resulting docking simulations poses are shown in Figure 5 and the relative Grid score docking
energies in Suppl Table 1.
10

Figure 5
The interpretation of these results well fit with the experimental data obtained from ELISA
tests. A structural difference between compounds 5 and 8 and the rest of the series is the lack of
the functional group SO3-. Since compounds 5 and 8, when compared with other molecules, are
found to bind in a different position, the sulfonate group seems to be fundamental for the correct
11

binding of the molecule, and therefore for its biological activity. In particular the orientation of
this group is in a zone of the protein surface rich in positively charged aminoacids (K14, R23,
R25, Figure 6). Only compound 4 showed a different orientation of the SO3- group, and this
observation could also explain the weak docking score of this compound with Akt. It must be
underlined that also the length of the acyl chains plays a fundamental role in easing the binding
of the compound. In fact, a longer chain appeared to provide a more stable interaction between
Akt and the inhibitor than a shorter one (e.g., docking energy was -73.78 for the most active 3a
vs -59.51 and -51.00 Kcal/mol for the less active 3c and 7, see Figure 2 and Suppl Table 1).
Figure 6
2.4 Cellular studies
ELISA assays and molecular modelling data highlighted compound 3a, as a promising
inhibitor of Akt in this series. To evaluate its potential as anticancer agent, the activity of
compound 3a was assayed in different cell lines in comparison with compounds 7 and 8 as
12

negative controls and the known Akt inhibitor A6730. Compound 2a could not be tested also in
cellular systems because of solubility limits. The used cell lines were endowed with an
hyperactivated Akt and included murine K-RAS-transformed fibroblasts (NIH3T3 K-RAS) and
human papillary thyroid carcinoma and ovarian carcinoma cell lines.
2.5 Akt phosphorylation inhibition studies on NIH 3T3 K-RAS cells
NIH3T3-K-RAS cells, expressing an activated form of the K-RAS gene and showing an
hyperactivation on the Akt signalling pathway have been used as cellular model to investigate
the effect of compounds on Akt phosphorylation under different conditions, i. e. serum-
starvation, serum-stimulation and pre-treatment with selected inhibitors (see materials and
methods). As reported in Figure 7, only compound 3a displayed an inhibitory activity on Akt,
both in pre-treated and in serum-stimulated cells. In agreement with docking studies and ELISA
assays, compound 8 did not affect P-Akt and compound 7 appeared to be inactive. The results of
the densitometric analysis are shown in Figure 7B, where the data are expressed as percentage of
the respective control after the normalization versus actin and the calculation of the P-Akt/total
Akt ratio. The data highlight the significant inhibition of the Akt activation induced by
compound 3a (about 40%). Neither compound 7 nor compound 8 exerted similar effect.
13

Figure 7
2.6 Papillary thyroid carcinoma and ovarian carcinoma cell growth inhibition
To further explore the features of compound 3a, we examined the growth inhibitory effect on
human tumor cell lines in which Akt signaling is known to play a role. Aberrant activation of the
PI3K/Akt pathway plays a fundamental role in the tumorigenesis and progression of thyroid
cancers [33]. The major oncogenic mutations found in papillary thyroid carcinomas (PTCs)
involving BRAF, RAS, PI3KCA, RET genes have been demonstrated to activate the PI3K/Akt
pathway. Thus, targeting this deregulated signaling cascade represents an attractive therapeutic
approach for treatment of PTC tumors. Moreover, it has been reported that thyroid cancer cells
characterized by PI3K/Akt signaling deregulation show a preferential sensitivity to inhibitors of
14

this pathway [34]. On these bases, we tested the antiproliferative activity of sulfoglycolipids 3a,
including compounds 7 and 8 as negative controls, on three PTC cell lines characterized by
constitutive activation of the Akt pathway functionally related to RET/PTC-1 rearrangement
(TPC-1 cells), to PI3KCA gene E542K mutation (K1 cells), and to decreased expression of the
dual-specificity phosphatase PTEN (Nim-1 cells). A 72h-treatment with compound 3a induced a
concentration-dependent inhibition of PTC cell growth. The inhibitory effect (about 30%) at the
highest concentration tested (50 μM) was significant in TPC-1 and K1 cells (p<0.05 and p<0.007
versus untreated control, respectively) (Figure 8A). K1 cells treated with 50 μM 3a displayed
signs of cytotoxicity such as vacuolated and granulated cytoplasms (not shown). Compounds 7
and 8 did not induce a significative inhibition of PTC cell growth after 72h. The Akt1/2 kinase
allosteric inhibitor (A6730), included in this studies as reference compound, showed a
comparable antiproliferative activity on TPC-1 and K1 cells (IC50 = 11 ± 1.6 µM and IC50 = 10 ± 5
µM, respectively). Of note, Nim-1 cells were the most sensitive to this inhibitor (IC50 = 3 ± 1.6
µM) in keeping with data describing an increased sensitivity of PTEN-deficient cells to
PI3K/Akt inhibitors [35].
15

Figure 8
16

Next, the efficacy of the sulfoglycolipids compounds was assessed on the ovarian carcinoma
cell line IGROV-1 previously described as PI3K-addicted [36]. To facilitate drug uptake and
interaction with the target, cells were treated in the absence of serum. In this model, compound
3a effectively reduced cell growth in a concentration-dependent manner with an IC50 of 2 ± 1.4
µM, whereas compound 7 was substantially inactive, and compound 8 slightly affected cell
proliferation (Figure 8B). Thus, only, compound 3a exhibited a biological effect in the cellular
tests. The IGROV-1 cell line showed a high sensitivity to the reference compound (IC50 = 0.3 ±
0.15 µM) according with the relevance of the Akt pathway in mediating cell proliferation.
Biochemical analysis performed on whole IGROV-1 cell lysates confirmed the ability of 3a to
affect PI3K/Akt signaling after 24h of treatment. Indeed, a reduced serine phosphorylation of the
critical Akt downstream substrates GSK-3αand β[37], and BAD was observed in cells exposed
to a 3a concentration in the range of IC50 (3 µM) (Figure 8C). This finding supports the ability of
this sulfoglycolipid to affect target activation and confirm the compound ability to enter the cells.
Again, the standard Akt1/2 inhibitor was highly effective in inhibiting Akt downstream effects.
3. Conclusions
In the present work, we show that synthetic sulfo-glycoconjugates derived from a D-
glucosidic scaffold, designed to mimic phosphatidyl inositols are able to impair Akt kinase
activity. The synthesized compounds were tested both in ELISA assays with purified Akt and in
different cell lines characterized by Akt hyperactivation. Docking studies and ELISA assay
evidenced that the longer the acyl chain, the better is the activity, while an acidic anionic group
is fundamental to guarantee optimal electrostatic interactions within the binding pocket.
Compound 3a, endowed with these features, showed selectivity for Akt and biological activity in
17

the tested cellular systems. This compound is a promising candidate, useful as lead compound
for the design of more efficient inhibitors of Akt, taking into consideration key structural features
of the studied sulfoglycolipids and chemico-physical properties favouring solubility and cell
permeability.
4. Experimental
4.1 Chemistry
Methyl 2,3,4-tri-O-acetyl-6-O-tosyl-β-D-glucopyranoside (9) was synthesized according to
the literature procedure [29]. Octyl β-D-glucopyranoside was purchased from Aldrich. Optical
rotations were determined on a Perkin–Elmer 241 polarimeter at 20 °C, in a 1 dm cell. Melting
points were recorded on a Büchi 510 capillary melting point apparatus and were uncorrected. All
reagents and solvents used were reagent grade and were purified before use by standard methods.
Dry solvents and liquid reagents were distilled prior to use or dried on 4 Å molecular sieves.
Column chromatography was carried out on flash silica gel (Merck 230–400 mesh). TLC
analysis was carried out on silica gel plate (Merck 60F254) with detection by charring with 50%
sulfuric acid or anisaldehyde based reagent. Evaporation under reduced pressure was usually
performed below 40 °C otherwise specified. The structures of all the new synthesized
1
compounds were confirmed through full H and 13C NMR characterization and mass
1
spectroscopy. H NMR analysis were performed at 500 MHz with a Bruker FT-NMR
AVANCE™ DRX500 spectrometer using a 5 mm z-PFG (pulsed field gradient) broadband
13
reverse probe at 298 K unless otherwise stated, and C NMR spectra at 125.76 MHz were done
of all the new compounds. The signals were unambiguously assigned by 2D COSY and HSQC
18

experiments (standard Bruker pulse program). Chemical shifts are reported as δ(ppm) relative to
1
residual CHCl3 or CH3OD fixed at 7.24 and 3.30 ppm, respectively, for H NMR spectra and
13
relative to CDCl3 fixed at 77.0 ppm (central line) or CD3OD at 49.00 ppm (central line) for C
NMR spectra; scalar coupling constants are reported in hertz. Mass spectra were recorded in
negative or positive-ion electrospray (ESI) mode on a Thermo Quest Finnigan LCQ DECA™
ion trap mass spectrometer; the mass spectrometer was equipped with a Finningan ESI interface;
sample solutions were injected with a ionization spray voltage of 4.5 kV or 5.0 kV (positive and
negative-ion mode, respectively), a capillary voltage of 32 V or -15 V (positive and negative-ion
mode, respectively), and capillary temperature of 250 °C. Data were processed by Finnigan
Xcalibur software system. TLC, NMR and MS confirmed purity and identity of all synthesized
compounds. High resolution mass spectra (HRMS) of the new target compounds (4 and 6-8)
were also recorded at C.I.G.A. (Centro Interdipartimentale Grandi Attrezzature, University of
Milan) on a Fourier Transform Ion Ciclotron Resonance (FT-ICR) mass spectrometer APEX II
& Xmass software (Bruker Daltonics) – 4.7 T magnet (Magnex), in positive-ion electrospray
(ESI) mode. Samples were injected as <1 to 5 µg/mL methanol solutions and eluted with
methanol (120 µL/h).
4.2 Synthesis of 2-O-β-D-sulfoquinovopyranosyl-sn-glycerol potassium salt 4.
To a solution of 3a (0.041 g, 0.066 mmol) in methanol (0.5 mL) 0.1 mL of 0.7 M sodium
methoxide in methanol were added. The reaction was monitored by TLC (AcOEt:iPrOH:H2O
3:3:1 v/v) and after 4 h was neutralized with DOWEX-50 x 8H+. The resin was filtered and the
solvent dried under vacuum. Water (3 mL) was added and extracted with petroleum ether (1.5
mL x 2). The acidic water solution was then neutralized with 0.1 M KOH and washed again with
19

chloroform to remove the remaining traces of stearic acid. Water was dried by freeze-drying to
1
obtain 0.017 g of the potassium salt 4 (0.046 mmol, 71% yield, hygroscopic). H NMR (D2O): δ
=3.49 (dd, J5’,6’a= 9.7 Hz, J6’a,6’b= 14.8 Hz, 1H, H-6’a), 3.70 (dd, J3’,4’= 9.3 Hz, J4’,5’= 9.5 Hz, 1H, H-
4’), 3.75 (dd, J1’,2’= 8.0 Hz, J2’,3’= 9.3 Hz, 1H, H-2’), 3.82 (dd, J5’,6’b= 1.7 Hz, 1H, H-6’b), 3.93 (dd,
1H, H-3’), 4.10-4.21 (m, 4H, H-1a, H1b, H3a and H3b), 4.22 (ddd, 1H, H-5’), 4.32 (m, 1H, H-2),
5.03 (d, 1H, H-1’). 13C NMR (D2O): δ = 52.16 (C6’), 60.99 (C3 or C1), 61.14 (C1 or C3), 72.21
(C5’), 72.41 (C4’), 73.22 (C2’), 75.38 (C3’), 81.11 (C2), 101.95 (C1’). ESI-MS (CH3OH,
negative-ion mode): m/z = 317.1 [M-K]-. Calcd for C9H17O10SK, m/z 356,02 [M]. ESI-HRMS
(CH3OH, negative-ion mode): m/z = 317.05451 [M-K]- (error, 0.9 ppm). Calcd for C9H17O10S,
m/z = 317.05479 (monoisotopic mass).
4.3 Synthesis of the sulfonates 6 and 7
Methyl 2,3,4-tri-O-acetyl-6-deoxy-6-thioacetyl-β-D-glucopyranoside 10. The known tosyl 9 (0.10
g, 0.211 mmol) was dissolved in dry DMF (3 mL), and potassium thioacetate (0.08 g, 0.7 mmol)
was added. The mixture was stirred overnight under an atmosphere of Ar at room temperature.
After disappearing of the starting material as shown by TLC (petroleum ether:ethyl acetate 50:50
v/v) water (7 mL) was added and extracted with dichloromethane (20 mL x 3). The collected
organic layers were washed with water (10 mL x 2), dried with sodium sulphate and
concentrated under vacuum. DMF was co-evaporated with cyclohexane at reduced pressure at
45-50 °C. Subsequent flash chromatography (petroleum ether:ethyl acetate 60:40 v/v) of the
crude residue yielded the desired derivative 10 (0.075 g, 0.2 mmol, 94% yield). Mp: 91-92°C
1
20
(amorphous solid); [α]D = - 19.3 (CH2Cl2, c=1.1). H NMR (CDCl3): δ =1.93, 1.98 and 2.01 (3s,
9H, 3CH3CO), 2.28 (s, 3H, CH3COS), 3.00 (dd, 1H, J5,6a= 7.0 Hz, J6a,6b= 14.3 Hz, 1H, H-6a), 3.20
20

(dd, 1H, J5,6b= 3.0 Hz, 1H, H-6b), 3.43 (s, 3H, OCH3), 3.56 (ddd, J4,5= 9.5 Hz, 1H, H-5), 4.33 (d,
J1,2= 8.0 Hz, 1H, H-1), 4.88 (dd, J2,3= 9.5 Hz, 1H, H-2), 4.90 (dd, J3,4= 9.5 Hz, 1H, H-4), 5.11 (dd,
1H, H-3). 13C NMR (CDCl3): δ = 20.49 (CH3CO), 20.59 (2CH3CO), 30.01(C6), 30.29 (CH3COS),
56.76 (OCH3), 70.56 (C4), 71.14 (C2), 72.63 (C3), 72.78 (C5), 101.23 (C1), 169.28, 169.64 and
170.12 (3CO), 194.60 (SCO). ESI-MS (CH3OH, positive-ion mode): m/z = 401.0 [M+Na]+.
Calcd for C15H22O9S, m/z 378.1 [M].
Methyl 2,3,4-tri-O-acetyl-β-D-sulfoquinovopyranoside 11
Potassium monopersulfate triple salt (Oxone®, 0.182 g, 0.296 mmol) and potassium acetate
(0.281 g, 2.87 mmol) were added in that order to a solution of compound 10 (0.075 g, 0.198
mmol) in glacial acetic acid (2.5 mL). The suspension was stirred at room temperature overnight
after checking by TLC (petroleum ether:ethyl acetate 50:50 v/v and CH2Cl2:CH3OH 80:20 v/v)
and another amount of the oxidant was added (0.182 g, 0.296 mmol). The solvent was dried
under vacuum and the crude residue was submitted twice to flash chromatography purification
(CH2Cl2:CH3OH 90:10 to 80:20 v/v) yielding 0.073 g (0.19 mmol, 96% yield) of 11. Mp: 167-
1
20
173°C dec; [α]D = - 1.7 (CH3OH, c=1.0). H NMR (CD3OD): δ = 1.95, 2.00 and 2.03 (3s, 9H,
3CH3CO), 3.01 (dd, 1H, J5,6a= 5.2 Hz, J6a,6b= 15.0 Hz, 1H, H-6a), 3.03 (dd, 1H, J5,6b= 5.2 Hz, 1H,
H-6b), 3.50 (s, 3H, OCH3), 4.07 (ddd, J4,5= 9.5 Hz, 1H, H-5), 4.55 (d, J1,2= 8.0 Hz, 1H, H-1), 4.86
13
(dd, J3,4= 9.5 Hz, 1H, H-4), 4.87 (dd, J2,3= 9.5 Hz, 1H, H-2), 5.23 (dd, 1H, H-3). C NMR
(CD3OD): δ = 18.52, 18.59 and 18.67 (3CH3CO), 51.62 (C6), 55.48 (OCH3), 69.68 (C5), 70.77
(C4), 70.97 (C2), 72.40 (C3), 100.46 (C1), 169.25, 169.54 and 169.63 (3CO). ESI-MS (CH3OH,
negative-ion mode): m/z = 382.9 [M-1]-. Calcd for C13H20O11S, m/z 384.07 [M].
Methyl β-D-sulfoquinovopyranoside potassium salt 6
To a methanol solution (1.0 mL) of triacetate 11 (0.073 g, 0.19 mmol), 0.54 mL of 1M sodium
methoxide in methanol were added and the reaction was stirred at room temperature for 5 h
21

(TLC, CH2Cl2:CH3OH 80:20 v/v and AcOEt:iPrOH:H2O 3:3:1 v/v). After neutralizing with
DOWEX-50 x 8H+ the resin was filtered and the solvent dried under vacuum. Water (2 mL) was
added and the acidic solution was neutralized with 0.1 M KOH. To decolorize the yellowish
solution, it was diluted with 10 mL of methanol and water (1.5 mL) and refluxed with activated
charcoal. The solvent was reduced under vacuum, water was added (1 mL) and dried by freeze-
20
drying to obtain 6 (0.046 g, 0.15 mmol, 82% yield, hygroscopic) as potassium salt. [α]D = - 6.4
(D2O, c=0.9). 1H NMR (D2O): δ = 2.93 (dd, 1H, J5,6a= 9.7 Hz, J6a,6b= 14.8 Hz, 1H, H-6a), 3.13 (dd,
J3,4= 9.4 Hz, J4,5= 9.4 Hz, 1H, H-4), 3.15 (dd, J1,2= 8.0 Hz, J2,3= 9.3 Hz, 1H, H-2), 3.28 (dd, 1H,
J5,6b= 1.4 Hz, 1H, H-6b), 3.37 (dd, 1H, H-3), 3.42 (s, 3H, OCH3), 3.65 (ddd, 1H, H-5), 4.24 (d,
1H, H-1). 13C NMR (D2O): δ = 51.97 (C6), 57.10 (OCH3), 72.05 (C5), 72.32 (C4), 73.06 (C2),
75.54 (C3), 102.97 (C1). ESI-MS (CH3OH, negative-ion mode): m/z = 257.2 [M-K]-. Calcd for
C7H13O8SK, m/z 296.00 [M]. ESI-HRMS (CH3OH, negative-ion mode): m/z = 257.03332 [M-K]-
(error, 1.3 ppm). Calcd for C7H13O8S, m/z = 257.03366 (monoisotopic mass).
Octyl 2,3,4-tri-O-acetyl-6-O-tosyl-β-D-glucopyranoside 12
Tosyl chloride (0.76 g, 4.0 mmol) was added to a solution of octyl β-D-glucopyranoside (1g, 3.42
mmol) in pyridine (22 mL) at -5°C and left overnight at this temperature. After about 90%
conversion of the starting material as shown by TLC (ethyl acetate 100%), acetic anhydride (4.85
mL, 51.3 mmol) was added at 0°C under stirring. After 1 h the reaction mixture was left to
gradually reach the room temperature. After 4 h the reaction was completed (TLC, petroleum
ether:ethyl acetate 60:40 v/v), then poured into ice/water (100 mL) and extracted with
dichlorometane (30 mL x 5). The collected organic layers were then washed with 5M
hydrochloric acid (35 mL x 2), sodium bicarbonate (50 mL) and water (50 mL x 2), dried with
sodium sulphate and concentrated under reduced pressure. Crystallization of the crude material
(1.87 g) with a mixture of ethanol (18 mL) and dichloromethane (1 mL) yielded 1.332 g (2.33
22

1
20
mmol, 68% yield) of pure 12. Mp: 109-110°C; [α]D = -4.6 (CHCl3, c=1.0); H NMR (CDCl3): δ
= 0.86 (t, J= 6.7, 3H, CH3), 1.18-1.33 (m, 10H, 5 CH2), 1.50 (m, 2H, CH2), 1.96, 1.97 and 1.99
(3s, 9H, 3 CH3CO), 2.43 (s, 3H, CH3), 3.39 (m, 1H, OCHa), 3.71 (m, 1H, H-5), 3.75 (m, 1H,
OCHb), 4.04 (dd, J5,6a= 5.9 Hz, J6a,6b= 10.9 Hz, 1H, H-6a), 4.09 (dd, J5,6b= 3.0 Hz, 1H, H-6b), 4.41
(d, J1,2= 7.9 Hz, 1H, H-1), 4.86 (dd, J2,3= 9.5 Hz, 1H, H-2), 4.87 (dd, J3,4= 9.5 Hz, J4,5= 9.7 Hz, 1H,
13
H-4), 5.14 (dd, 1H, H-3). C NMR (CDCl3): δ = 14.03 (CH3), 20.50, 20.54 and 20.55 (3
CH3CO), 21.62 (CH3Ph), 22.61 (CH2), 25.80 (CH2), 29.22 (CH2), 29.24 (CH2), 29.35 (CH2), 31.77
(CH2), 67.91 (C6), 68.87 (C4), 70.15 (OCH2), 71.23 (C2), 71.58 (C5), 72.62 (C3), 100.60 (C1),
128.05 (2 CHPh), 129.86 (2 CHPh), 132.59 (CPh), 145.08 (CPh), 169.14, 169.46 and 170.16 (3
CO). ESI-MS (CH3OH, positive-ion mode): m/z = 595.0 [M+Na]+. Calcd for C27H40O11S, m/z
572.23 [M].
Octyl 2,3,4-tri-O-acetyl-6-deoxy-6-thioacetyl-β-D-glucopyranoside 13
Compound 12 (0.50 g, 0.87 mmol) was dissolved in dry DMF (4 mL), and potassium thioacetate
(0.360 g, 3.15 mmol) was added. The mixture was stirred under Ar at room temperature and the
progress of the reaction was monitored by TLC (petroleum ether:ethyl acetate 70:30 v/v). After 6
h the reaction was stopped and after water (9 mL) addition extracted with dichloromethane (30
mL x 3). The collected organic layers were washed with water (10 mL x 2), dried with sodium
sulphate and concentrated under vacuum. DMF was co-evaporated with cyclohexane at reduced
pressure at 45-50 °C and subsequent flash chromatography (petroleum ether:ethyl acetate 80:20
v/v) of the crude derivative 13 was obtained (0.396 g, 0.832 mmol, 95% yield). Mp: 58-59°C
1
20
(wax); [α]D = - 10.4 (CHCl3, c=1.0). H NMR (CDCl3): δ = 0.84 (t, J= 6.9, 3H, CH3), 1.17-1.33
(m, 10H, 5 CH2), 1.52 (m, 2H, CH2), 1.95, 1.98 and 2.04 (3s, 9H, 3 CH3CO), 2.30 (s, 3H,
CH3COS), 3.00 (dd, J5,6a= 7.1 Hz, J6a,6b= 14.2 Hz, 1H, H-6a), 3.22 (dd, J5,6b= 3.0 Hz, 1H, H-6b),
3.43 (m, 1H, OCHa), 3.56 (ddd, J4,5= 10.0 Hz, 1H, H-5), 3.79 (m, 1H, OCHb), 4.41 (d, J1,2= 8.0
23

Hz, 1H, H-1), 4.90 (dd, J2,3= 9.6 Hz, 1H, H-2), 4.92 (dd, J3,4= 9.6 Hz, 1H, H-4), 5.13 (dd, 1H, H-
13
3). C NMR (CDCl3): δ = 14.01 (CH3), 20.55 (2 CH3CO), 20.63 (CH3CO), 22.59 (CH2), 25.77
(CH2), 29.20 (CH2), 29.38 (CH2), 30.22 (C6), 30.32 (CH3COS), 31.76 (CH2), 70.04 (OCH2),
70.85 (C4), 71.46 (C2), 72.81 (C3), 72.97 (C5), 100.62 (C1), 169.18, 169.67 and 170.20 (3 CO),
194.62 (SCO). ESI-MS (CH3OH, positive-ion mode): m/z = 499.0 [M+Na]+. Calcd for C22H36O9S,
m/z 476.21 [M].
Octyl 2,3,4-tri-O-acetyl-β-D-sulfoquinovopyranoside 14
Potassium monopersulfate triple salt (Oxone®, 2.0 g, 3.26 mmol) and potassium acetate (1.12 g,
11.7 mmol) were added in that order to a solution of compound 13 (0.389 g, 0.817 mmol) in
glacial acetic acid (15 mL). The suspension was stirred at room temperature and the reaction was
monitored by TLC (petroleum ether:ethyl acetate 70:30 v/v and CH2Cl2:CH3OH 90:10 v/v). After
disappearing of the starting material, the solvent was evaporated under vacuum, 5 M
hydrochloric acid (40 mL) was added and extracted with dichlorometane (20 mL x 5). The
collected organic layers were dried under vacuum and the crude residue purified by flash
chromatography (CH2Cl2:CH3OH 90:10 to 80:20 v/v) yielding 0.352g (0.73 mmol, 89% yield) of
1
20
compound 14 as a white foam. [α]D = - 16.8 (CH3OH, c=1.0). H NMR (CD3OD): δ = 0.89 (t, J=
6.9, 3H, CH3), 1.25-1.35 (m, 10H, 5 CH2), 1.54 (m, 2H, CH2), 1.95, 2.00 and 2.02 (3s, 9H, 3
CH3CO), 3.02 (dd, J5,6a= 5.0 Hz, J6a,6b= 14.8 Hz, 1H, H-6a), 3.03 (dd, J5,6b= 5.0 Hz, 1H, H-6b), 3.52
(m, 1H, OCHa), 3.92 (m, 1H, OCHb), 4.06 (ddd, J4,5= 10.0 Hz, 1H, H-5), 4.62 (d, J1,2= 8.0 Hz,
1H, H-1), 4.86 (dd, J3,4= 9.6 Hz, 1H, H-4), 4.87 (dd, J2,3= 9.6 Hz, 1H, H-2), 5.23 (dd, 1H, H-3).
13C NMR (CD3OD): δ = 12.40 (CH3), 18.53, 18.65 and 18.69 (3 CH3CO), 21.69, 25.03, 28.40,
28.43, 28.49 and 30.96 (6 CH2), 51.77 (C6), 68.98 (OCH2), 69.63 (C5), 70.85 (C4), 71.11 (C2),
72.39 (C3), 99.60 (C1), 169.13, 169.56 and 169.65 (3 CO). ESI-MS (CH3OH, negative-ion
mode): m/z = 481.1 [M-1]-. Calcd for C20H34O11S, m/z 482.18 [M].
24

Octyl β-D-sulfoquinovopyranoside potassium salt 7
To a methanol solution (2.0 mL) of triacetate 14 (0.352 g, 0.73 mmol), 2.6 mL of 1M sodium
methoxide in methanol were added and, following the procedure described for compound 6, the
reaction was stirred at room temperature overnight (TLC, CH2Cl2:CH3OH 80:20 v/v and
AcOEt:iPrOH:H2O 3:3:1 v/v). After neutralizing with DOWEX-50 x 8H+ the resin was filtered
and the solvent dried under vacuum. Water (1 mL) was added and the acidic solution was
neutralized with 0.4 M KOH. To decolorize the yellowish solution, it was diluted with 6 mL of
methanol and refluxed with activated charcoal. The solvent was reduced under vacuum, water
was added (1 mL) and dried by freeze-drying to obtain 7 (0.234 g, 0.59 mmol, 81% yield) as
1
20
potassium salt. Mp: 155-160°C dec; [α]D = - 3.9 (D2O, c=1.8). H NMR (D2O): δ = 0.78 (t, J=
6.9, 3H, CH3), 1.14-1.31 (m, 10H, 5 CH2), 1.54 (m, 2H, CH2), 2.97 (dd, J5,6a= 9.5 Hz, J6a,6b= 14.8
Hz, 1H, H-6a), 3.18 (dd, J3,4= 9.2 Hz, J4,5= 9.5 Hz, 1H, H-4), 3.19 (dd, J1,2= 8.0 Hz, J2,3= 9.2 Hz,
1H, H-2), 3.31 (dd, J5,6b< 1.0 Hz, 1H, H-6a), 3.41 (dd, 1H, H-3), 3.57 (m, 1H, OCHa), 3.68 (ddd,
13
1H, H-5), 3.83 (m, 1H, OCHb), 4.36 (d, 1H, H-1). C NMR (D2O): δ = 13.44 (CH3), 22.04,
25.12, 28.42, 28.49, 28.69 and 31.12 (6 CH2), 52.17 (C6), 70.63 (OCH2), 72.12 (C5), 72.45 (C4),
73.22 (C2), 75.74 (C3), 102.00 (C1). ESI-MS (CH3OH, negative-ion mode): m/z = 355.2 [M-K]-.
Calcd for C14H27O8SK, m/z 394.11 [M]. ESI-HRMS (CH3OH, negative-ion mode): m/z =
355.14257 [M-K]- (error, 1.8 ppm). Calcd for C14H27O8S, m/z = 355.14321 (monoisotopic mass).
4.4 Docking Studies
Docking studies were performed using the high resolution X-ray structure of PH domain of
Protein kinase B/Akt (pdb entry: 1UNQ [Milburn, C.C]). Ligands structures were built using the
chemistry package Molecular Operating Environment (MOE) [Chemical Computing Group Inc.
25

http://www.chemcomp.com/]. Receptors and ligand atom types and charges were calculated
using the AM1-BCC method of the antechamber package, available in the software UCSF
Chimera [38], in order to recognize atom and bond types, generating the residue topology file
and finding force field parameters. Docking simulations were performed using DOCK6 [39].
Starting from receptor surface, spheres with a maximum radius of 3 Å and a minimum radius of
1 Å were generated. The ensemble of spheres within 10Å of the bounded inositol-(1,3,4,5)-
tetrakisphosphate (4IP) molecule was assumed to represent the binding site core, and box in
which poses were generated was built extending the binding site core by additional 10Å along
the Cartesian axes. The grid was calculated including all receptor atoms and the distance
between grid point along each axes was of 0.3 Å. Non polar hydrogen atoms were modeled using
the all atom model: hydrogens attached. During dock calculations, ligands were kept flexible and
500 minimization cycle were performed both for the anchor (the largest rigid substructure of the
ligand) and the branches (other component of the ligand molecules). Very similar poses (RMSD
< 0,1) were clustered and the best 30 poses were ranked according to Grid Score (score is given
by approximate molecular mechanics interaction energies, consisting of van der Waals and
electrostatics components without solvation energy) [39]. When considering very flexible
molecules, such as 2a, several conformations (differing only for small differences in the
orientation of the alkyl chains) clustered within a narrow energy range. However, for the sake of
clarity, only the lowest energy poses are presented and discussed.
4.5 Biological Assays
4.5.1 Akt inhibition assays
26

ELISA test. The Akt kinase activity assay was performed by employing a commercial ELISA
kit designed specifically to screen inhibitors or activators of Akt. The inhibitory activity of
compounds was tested employing the CycLex AKT/PKB kinase Assay/Inhibitor Screening Kit
(CycLex, Eppendorf, Milano, Italy). Plates were pre-coated with “AKTide-2T” which can be
efficiently phosphorylated by Akt1. The detector antibody specifically detects the
phosphorylated “AKTide-2T”. Particularly, to perform the test, the samples were dissolved in
DMSO (except for 4 which was dissolved in distilled water), and diluted in Kinase Buffer to a
final concentration of 100 µM. For compound 3a, a concentration-response analysis was also
performed employed a concentration range of 1-250 µM. Compounds were added together with
constitutive active form of human Akt1 (25 m units/well), and allowed to phosphorylate the
bound substrate following the addiction of Mg2+ and ATP. The amount of phosphorylated
substrate was measured by binding it with horseradish peroxidase conjugate of an anti-phpspho-
AKTide-2T monoclonal antibody, which then catalyzes the conversion of the chromogenic
substrate tetra-methylbenzidine from the colourless reduced form to the yellow oxidized product,
after the addition of the stopping reagent. The absorbance of the resulting solution is determined
spectrophotometrically at max= 450 nm, and it is related to Akt1 activity in the tested solution.
The test was performed three times, each time in triplicate. As reference compound, the Akt1/2
kinase
inhibitor
(1,3-Dihydro-1-(1-((4-(6-phenyl-1H-imidazo[4,5-g]quinoxalin-7-
yl)phenyl)methyl)-4-piperidinyl)-2H-benzimidazol-2-one trifluoroacetate salt hydrate, indicated
as A6730 from Sigma, St. Louis, MO) was also used. Furthermore Staurosporine (Sigma-
Aldrich, Milano, Italy) at the final concentration of 1µM was employed as “inhibitor control” as
indicated in the assay protocol to verify the function of the kit Statistical analyses were
performed using Kruskal-Wallis non parametric ANOVA. A value of p<0.05 was considered
statistically significant. The IC50 value is calculated through the GraphPad software.
27

BTK assay. The BTK kinase activity assay was performed by employing a commercial ADP-
Glo™ Kinase Assay (Promega), a luminescent assay that measures ADP formed from a kinase
reaction. The compounds were dissolved in DMSO and used at a final concentration of 100 µM.
Compound 3a was added together with BTK, and allowed to phosphorylate the substrate
following the addition of ATP. The kinase reaction was performed according to the
manufacturer’s instructions. The test was performed twice, each time in triplicate. As reference
compound, the Akt1/2 kinase inhibitor A6730 was used. Staurosporine was employed as positive
control as indicated in the assay protocol.
Western blot analysis. NIH 3T3 cells transformed by an activated form of the K-ras oncogene
(NIH 3T3 K-ras) [40] were grown at 37 °C in Dulbecco’s modified Eagle’s medium (DMEM)
(Euroclone) supplemented with 10% heat-inactivated NBCS (Gibco-BRL). 3 x 105 cell/mL cells
were plated in 12 well plates and starved for 16h in DMEM supplemented with 0.5% NCS. Cells
were pre-treated with the different compounds at 100 µM or with vehicle (DMSO) for 1 h and
then incubated in DMEM supplemented with 10% serum (NCS) for additional 30 min. Cell were
then harvested and lysed in Lysis Buffer (50 mM Tris-HCl, pH 7.4, 5% glycerol, 200 mM NaCl,
1% IGEPAL (Sigma), 1 µM aprotinin, 1 µM leupeptin and 1 mM PMSF). Equal amounts of total
protein extracts (15 µg) were loaded onto a 8% SDS-polyacrylamide gel and blotted to
nitrocellulose membranes. Western blots were probed with anti-phospho-Akt (Ser 473) (Cell
Signaling) antibodies; subsequently, filters were stripped and re-probed with anti-Akt (Cell
Signaling) antibodies. Anti-actin antibody produced in rabbit (Sigma-Aldrich) was used as
loading control. Secondary antibodies conjugated to horseradish peroxidase (HRP) (GE
Healthcare) were used. All signals were detected after the HRP was activated by enhanced
chemiluminescence.
28

4.5.2 Cellular Studies
Compound solution preparation. Compound 3a and 7 were easily dissolved in 100% DMSO at
10 mM, whereas compound 8 was dissolved in 100% DMSO at the same concentration after
prolonged stirring. A6730 (Sigma-Aldrich, St. Louis, MO) used as reference compounds was
dissolved like the other compounds.
Cell lines, cell culture and cell growth assay. The human papillary thyroid carcinoma (PTC) cell
lines TPC-1 and Nim-1 were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Lonza,
Verviers, Belgium) supplemented with 10% fetal bovine serum whereas the K1 cell line was
cultured in DMEM: Ham’s F12 (Lonza): MCDB 105 (Sigma) (2: 1: 1) supplemented with 20%
serum. The ovarian carcinoma cell line IGROV-1 [41] was routinely maintained in RPMI
o
medium (Lonza) containing 10% serum. All cell lines were incubated at 37 C in 5% CO2
atmosphere. TPC-1 cells harbor the RET/PTC-1 rearrangement, Nim-1 cells the BRAFV600E
mutation, and K1 cells both BRAF V600E and PI3K E542K mutation [42]. Down-regulation of
PTEN has been described in Nim-1 cells [35]. IGROV-1 cells harbor hetT319F deletion and
frameshift in PTEN, hetR38C mutation in the p85 binding domain of p110 α, and an additional
hetX1069W mutation that extends the C-terminus of p110 αby four amino acids [36]. For cell
growth assay, cells were plated in 12 well-plates at 5000 cells/cm2 (Nim-1), 3000 cells/cm2
(TPC-1), or 4000 cells/cm2 (K1), in complete medium. One day after seeding, cells were exposed
to solvent (DMSO, 0.5% final concentration) or to increasing drug concentrations for 72h.
IGROV-1 cells were plated in complete medium (at 5000 cells/cm2). The day after, cells were
serum starved and treated with the drugs for 24h. The supernatant was then removed and
complete medium was added for further 48 h. Next, cells were trypsinized and counted by a
Coulter Counter (Coulter Electronics, Luton, UK). Each experiment was performed in
29