Oxaphosphinanes: New therapeutic perspectives for glioblastoma

Article abstract of DOI:10.1021/jm201428a

This paper reports the design and the synthesis of a new family of compounds, the phostines, belonging to the [1,2]oxaphosphinane family. Twenty-six compounds have been screened for their antiproliferative activity against a large panel of NCI cancer cell lines. Because of its easy synthesis and low EC₅₀ value (500 nM against the C6 rat glioma cell line), compound 3.1a was selected for further biological study. Moreover, the specific biological effect of 3.1a on the glioblastoma phylogenetic cluster from the NCI is dependent on its stereochemistry. Within that cluster, 3.1a has a higher antiproliferative activity than Temozolomide and is more potent than paclitaxel for the SF295 and SNB75 cell lines. In constrast with paclitaxel and vincristine, 3.1a is devoid of astrocyte toxicity. The original activity spectrum of 3.1a on the NCI cancer cell line panel allows the development of this family for use in association with existing drugs, opening new therapeutic perspectives.

Full text of DOI:10.1021/jm201428a

Article
pubs.acs.org/jmc
Oxaphosphinanes: New Therapeutic Perspectives for Glioblastoma
Ludovic Clarion,†,‡,∥ Carine Jacquard,†,∥ Odile Sainte-Catherine,§ Severine Loiseau,‡ Damien Filippini,‡
́
Marie-Helene Hirlemann,‡ Jean-Noel Volle,‡ David Virieux,‡ Marc Lecouvey,§ Jean-Luc Pirat,*,‡
́
̀
̈
and Norbert Bakalara*,†
†INSERM U-1051, Institut des Neurosciences de Montpellier, 80 rue Augustin Fliche 34091 Montpellier, France
‡UMR 5253, ICG Montpellier, Equipe AM2N, ENSCM, 8, rue de l'Ecole Normale, 34296 Montpellier cedex 5, France
§UMR 7244 CSPBAT, Equipe CBS Universite
́
Paris 13, 74 rue Marcel Cachin, 93017 Bobigny cedex, France
S
* Supporting Information
ABSTRACT: This paper reports the design and the synthesis
of a new family of compounds, the phostines, belonging to the
[1,2]oxaphosphinane family. Twenty-six compounds have
been screened for their antiproliferative activity against a
large panel of NCI cancer cell lines. Because of its easy
synthesis and low EC50 value (500 nM against the C6 rat
glioma cell line), compound 3.1a was selected for further biological study. Moreover, the specific biological effect of 3.1a on the
glioblastoma phylogenetic cluster from the NCI is dependent on its stereochemistry. Within that cluster, 3.1a has a higher
antiproliferative activity than Temozolomide and is more potent than paclitaxel for the SF295 and SNB75 cell lines. In constrast
with paclitaxel and vincristine, 3.1a is devoid of astrocyte toxicity. The original activity spectrum of 3.1a on the NCI cancer cell
line panel allows the development of this family for use in association with existing drugs, opening new therapeutic perspectives.
cell line such as the C6 one and in a second step against the
human National Cancer Institute (NCI) cell line panel.
1. INTRODUCTION
Malignant gliomas are the most frequent primary brain tumors.
Glioblastomas (WHO grade IV) account for 60−70% of
primary malignant gliomas with poor prognosis. Moreover,
low-grade glioma will inescapably become malignant. The best
treatment, which consists of surgical resection followed by
chemotherapy with Temozolomide (TMZ) and radiotherapy,
gives a median survival of 15 months as compared to 55 days
with biopsy alone.1 We therefore realize the need of developing
new drugs with antiproliferative activity on glioblastoma cells.
So far, molecular agents currently in clinical trial for high-grade
glioma have emerged from existing molecules selected for other
types of cancer. Mainly, these molecules target intracellular
signaling pathways relative to either growth or angiogenesis and
have been efficient in preclinical models of gliomas.2 Never-
theless, systemic antiangiogenic treatment can increase
glioblastoma invasion in patients and induce side effects like
intracranial hemorrhage and ischemic stroke.3,4 Moreover,
inhibition of tyrosine kinase-coupled growth factor receptors
by small chemical compounds (Gefitinib) or monoclonal
antibodies (cetuximab) adds little in terms of overall survival
or progression free survival.5,6 Innovative aspects of glioblas-
toma treatment lie in combinations of molecules such as PCV
(procarbazine, lomustine, and vincristine)7 and in local
chemotherapy with biodegradable carmustine wafers (Gliadel).8
Improvements have been attained at the surgery level by the
use of magnetic resonance imaging (MRI)9 and fluorescent-
guided surgery10 to increase the extent of resection. To develop
a new antiglioma therapy, we decided to design and synthesize
a new family of compounds screened for their antiproliferative
activity, in a first intention on commonly reported glioblastoma
Phostines are a set of compounds that present an 1,2-
oxaphosphinane heterocyclic core and are considered as mimes
of glycosides. We thought that the phosphinolactone structure
possesses unique and unexploited features: besides purely
physical and geometrical considerations, the phosphinolactone
group can be regarded as a surrogate of lactol and could present
an opportunity for medicinal chemists to explore uncovered
regions of chemical space.11 In addition to its isosteric
relationship with the lactol group, phosphinolactone is an
outstanding hydrogen bond acceptor and metal-complexing
agent, which makes it an attractive structural motif for drug
discovery. Whatever the nature of the phosphoryl bond (P
O), such compounds exhibit high Lewis/Bronsted basicity.12 As
observed with glycopyranoside derivatives, phosphorus six-
membered rings show also stereoelectronic-dependent inter-
actions with phosphorus atoms that are closely related to the
anomeric effect.13 Another important feature of oxaphosphi-
nanes is that such phosphorus heterocycles are considerably
more resistant to the ring-opening/closure occurring between
the open-chain and the cyclic pyranose forms.14,15
Similar modifications of lactol groups have already been
exploited in C-glycosides and aza- or thio-sugars, leading to
compounds of therapeutic interests.16 However, these com-
pounds are different from the phostone17−19 family because of
the exocyclic P−C bond, which confers to the molecules a
higher stability. In this paper, we describe the synthesis20 of a
Received: November 2, 2011
Published: January 23, 2012
© 2012 American Chemical Society
2196
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
Scheme 1. Synthetic Route to hydroxy-oxaphosphinanes 3.1-3.7
Scheme 2. Synthetic Route to the α-D-Amino-glucose Analogues 6a and 7a
Scheme 3. Synthesis of the Deoxyphosphinosugar 9
new family of compounds, the phostines, obtained from the
reaction of H-phosphinates with protected furanosyl glycosides
in the presence of a base. The antiproliferative properties of the
phostines were screened on C6 glial cells.
mixture of stereoisomers. When the IC50 values were lower
than 15 μM, they were separated by preparative reverse phase
high-performance liquid chromatography (HPLC) and tested
separately.
The toxicity of the best compound, 3.1a, was compared with
its diastereoisomer 3.1b on cancer cell lines from different
tissues and species, C6 glial cells, and nonproliferative brain
cells. This study shows that 3.1a is an antiglioma compound
with low glial cell toxicity and leads us to think that it can be a
therapeutic compound in preclinical glioma models.
When the Ar group was a phenyl, the diastereomers 3.1a−d
were obtained in 28/26/19/27 proportion. The four stereo-
isomers were fully characterized, allowing the attribution of all
of the stereogenic centers. The first two diastereomers 3.1a and
3.1b are epimer at carbon 2, and they correspond, respectively,
to phosphino-α-D-glucose and phosphino-α-D-mannose.23 The
two other, 3.1c and 3.1d, can be perceived as the β-anomers,
but they were revealed to be less configurationally stable at the
phosphorus center due to the presence of the phenyl ring in
axial position. So, only the 3.1a and 3.1b isomers were finally
tested.
In a second set of reactions, different chemical modifications
were introduced on the carbon adjacent to the phosphorus
atom. Then, amino and urea derivatives 6a and 7a were
synthesized starting from the mannose-like stereoisomer, 3.1b.
The transformation of the alcohol function into a good leaving
group led to the mesylate 4.1b in 50% yield. Substitution by
trimethylsilylazide gave the azido derivative 5.1a with inversion
of configuration. As expected, the Staudinger reaction furnished
the glucoselike oxaphosphinane 6.1a. Finally, when 6.1a was
reacted with phenylisocyanate, the corresponding urea 7.1a was
isolated in 53% yield. Starting from the glucoselike stereo-
isomer 3.1a, a similar scheme allowed the formation of the
2. RESULTS
2.1. Oxaphosphinane Synthesis. Phostines were specif-
ically designed as a combination of glycopyranosides and C-
aryglycosides, and their synthesis was achieved by the reaction
of tris(O-benzyl protected)-D-arabinofuranose 1 with various
ethyl arylphosphinates 221 in the presence of catalytic amount
of potassium t-butoxide (Scheme 1).20,22 This tandem sequence
involved first the nucleophilic Pudovik addition of H-
phosphinate anions 2 to the open-chain form of arabinofur-
anose. The second step consisted of a subsequent prototropy
and a 6-exotet cyclization, affording the six-membered ring
phosphinolactones (21−67% yield, Scheme 1). This procedure
offered an easy tuning of the aryl substituent directly bound to
the phosphorus atom. Four enantiopure diastereomers 3.a−d
were formed, arising from the creation of two chiral centers
with moderate diastereoselectivities. The reaction appeared
general, and oxaphosphinanes 3.1−3.6 were initially tested as a
2197
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
Scheme 4. Synthesis of the α-D-Glucose Analogues
Scheme 5. Synthetic Routes of the α-D-Glucose Analogues
a
Table 1. Antiproliferative Values of the Phostines on C6
a
Values were calculated from at least three independent experiments, and for each set of experiments, each point was repeated six times. EC50
estimation was determined with Prism 4.0 software using the Hill equation.
2198
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
a
Table 2. Antiproliferative Activities of 3.1a and 3.1b on Different Cell Lines
tissues
brain
skin
liver
intestine
prostate
breast
colorectal
adeno-
carcinoma
epidermoid
carcinoma
hepato-
carcinoma
prostatic carcinoma
(brain metastasis)
adeno-carcinoma
(pleural metastasis)
cell types
glioma
melanoma
MDAMB-
435
cell lines
species
C6
rat
B16F10
mouse
0.5
A431
human
0.1
HuH7
human
0.1
CaCo2
human
0.5
DU145
human
5
MDAMB-231
human
human
0.001
b
3.1a LTC (μM) 0.001
0.1
10 μM
82
1
1
42
0
73
0
32
1
22
1
43
1
40
0
30
1
1
toxicity
c
(%)
d
1 μM
63
22
1
25
0
18
1
16
1
10
1
Nt
17
toxicity
(%)
e
f
g
EC50 (μM)
0.5 0.1 13.9 0.5
4.9 0.3 14.5 0.5
30
0.1
55
3
1
57
5
8
1
65
1
8
1
Nqe
5
3.1b LTC (μM)
0.1
31
0.1
43
25
Nt
5
10 μM
toxicity
(%)
1
0
33
1
15
14
18
1
a
Values were calculated from at least three independent experiments, and for each set of experiments, each point was repeated six times. Data are
b
results of MTT assay, 48 h after product addition. LTC is calculated after mean comparison between vehicle and product and ANOVA posthoc
c
d
Fisher test (p < 0.0001). Toxicity (10 μM) was calculated as the percentage of MTT absorbance loss as compared to the vehicle. Nt, no toxicity.
e
EC50 estimation was determined with Prism 4.0 software using Hill equation with four parameters and with the bottom fixed with Triton X-100 2%
f
g
and the top fixed with vehicle. EC50 values under 10 μM are in bold. Nqe, no quantified estimation.
phosphino-α-D-mannose analogues 4.1a, 5.1b, and 6.1b in,
respectively, 46, 86, and 54% yields (Scheme 2).
form from brain tissue. That hypothesis is reinforced by the low
sensitivity of DU-145 (65 8 μM) to 3.1a. As a matter of fact,
DU-145 is a metastatic brain tumor that started from a primary
prostate cancer (Figure 1 and Table 2). Furthermore, because
To estimate whether or not the presence of the free hydroxy
group affects the IC50, the corresponding deoxyphosphinolac-
tone 9 was also synthesized. This oxaphosphinane 9 was
directly accessible from the mixture of diastereomers 3.1a and
3.1b by the Barton−McCombie deoxygenation. The dithiocar-
bonate intermediates 8.1a and 8.1b were separated to further
test them (Scheme 3).
Other modifications on the hydroxy group were introduced
by the formation of the corresponding ester and ether
derivatives. Starting from the diastereoisomer 3.1a, the desired
targets were prepared by Williamson or by acylation reactions
(Scheme 4).
Direct synthesis of amino-oxaphosphinanes can be per-
formed starting from the alkylamino-arabinofuranose. Thus, 15
and 15′ were obtained quantitatively by treatment of 2,3,5-tri-
O-benzyl-D-arabinofuranose with an excess of benzylamine or
allylamine.24,25
Then, the phostines 16 and 17 were synthesized according to
the conditions of reaction mentioned above. Only one
enantiopure stereoisomer 16.1a and 17.1a (among the four
diastereomers) was isolated after purification in 12 and 7%
yield, respectively (Scheme 5).
2.2. Antiproliferative and Stereoselective Activity of
3.1a on Brain Cancer Cell Lines. Phostines were screened on
the C6 rat glioblastoma cell line using a dimethylthiazolyl-2,5-
diphenyltetrazolium bromide (MTT) test. Compounds 3.1a
and 16.1a proved to be the most potent inhibitors of C6 cell
growth (Table 1). However, the following studies were carried
out with 3.1a, which is readily synthesized in yields that are
better than for 16.1a. Moreover, a promising diastereoselective
synthesis to increase the yield of 3.1a even further is already
under investigation.
Figure 1. Antiproliferative activities of 3.1a and 3.1b. Antiproliferative
activities were measured by MTT reduction, 48 h after drug addition
(3.1a and 3.1b) on different cell lines on rat C6 glioblastoma cells
(C6) and brain metastasis of prostatic cells (DU-145). OD 570 nm,
optical density measured at 570 nm; C, concentration.
3.1a always has a higher antiproliferative effect than 3.1b at 10
μM and lower LTC and EC50 values (Figure 1 and Table 2) on
skin cancer cell lines (MDA-MB-435, B16F10, and A431),
colorectal cancer cells (CaCo2) and breast cancer cells (MDA-
MB-231), a stereodependent mode of action of 3.1a is
suggested. Conversely, 3.1b has greater antiproliferative activity
on the hepatocarcinoma cell line (Huh7, Table 2).
In a next step, we investigated whether or not the brain
tumor specificity and stereospecificity of 3.1a hold also for
other human primary cell lines (ie. Leukemia, lung, ovarian,
etc.). For that purpose, a screen test on the NCI human cancer
cell lines has been performed.
In all cancer cell lines derived from other tissue than brain
(i.e., skin, liver, intestine, prostate, and breast) 3.1a at 1 and 10
μM is always less toxic and has a higher lower toxic
concentration (LTC) and half maximal effective concentration
(EC50) values (Table 2), suggesting specificity for cancers that
2.3. Higher Antiproliferative Activity of 3.1a Than
3.1b against a Large Panel of Human Cancer Cell Lines
and Brain Specificity. The NCI quantified the effects of 3.1a
and 3.1b at 10 μM on the growth of 57 cell lines (Figure 2A,B)
using the sulforhodamine test. The results obtained by this test
2199
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
Figure 2. Oxaphosphinanes effect on NCI60 GP and effects comparison with paclitaxel on the CNS cell lines. (A and B) Growth percentage of NCI-
60 cell lines after 48 h of incubation of 10 μM 3.1a (A) or 10 μM 3.1b (B) as quantified with sulforhodamine B assay. Gray circles represent NCI 60
CNS cell lines and cell lines used for the MTT test in the primary screening (Table 1). (C) Oxaphosphinanes, TMZ, and paclitaxel effects on brain
cell lines growth percentage. Comparison of CNS cell lines growth percentage with 10 μM 3.1a, 3.1b, TMZ, and paclitaxel. Black, 3.1a; white, 3.1b;
black and gray, paclitaxel; and white and gray, TMZ. Nsc, nonsmall cell; CNS, central nervous system; 1, SF-295; 2, SNB-75; 3, U-251; 4, SF-268; 5,
SNB-19; 6, SF-539; a, MDA-MB-435; b, DU-145; and c, MDA-MB-231.
on the MDA-MB-435, MDA-MB-231, and DU-145 cell lines
show similar levels of antiproliferative activity as those initially
observed with the MTT test (Table 2). This confirms on the
one hand that 3.1a has a greater antiproliferative activity than
3.1b (Figure 2A−C) and on the other hand its brain selectivity.
As a matter of fact, the median growth percent (GP) value of
3.1a is inferior to 50% only for the central nervous system
(CNS) cell line cluster (Figure 2A). Negative GP values,
indicating a cytotoxic effect, were found for two CNS cell lines,
SF-295 and SNB75 (Figure 2A). Interestingly, Hs-578T is the
only breast cancer cell line highly sensitive to 3.1a (Figure 2A).
Other cell lines are moderately sensitive to 3.1a (i.e., 0 < GP <
50). They include CNS cell lines (SF-268 and U251), nonsmall
lung cancer cells (HOP-62 and NCI-H522), ovarian cell line
(SK-OV-3), and one breast cancer cell line (MDA-MB-468).
compare the antiproliferative activities of 3.1a and 3.1b with
reference drugs used in the treatment of malignant glioma.
2.4. Comparison of the Antiproliferative Activities of
3.1a and 3.1b and Reference Drugs Used in the
Treatment of Malignant Glioma on the C6 Glioma
Model. To evaluate the position of the most active compound
of the new phostine family, the antiproliferative activity of 3.1a
was compared to 3.1b (Figure 3A) and to some therapeutic
agents currently prescribed or in clinical trial for glioblasto-
ma7,26,27 (Figure 3B). By measuring neutral red accumulation,
we confirmed the data obtained with MTT, which indicates
that 3.1a is more active on the C6 cell line than its
diastereoisomer 3.1b. We can see that the antiproliferative
activity of 3.1a is lower than the antimitotic drugs paclitaxel and
vincristine used at 0.1, 1, or 10 μM concentration. By adding
100 μM TMZ, a slight antiproliferative activity on C6 cells
appeared, but it still remained lower than for 10 μM 3.1a
(Figure 3B). In conclusion, although 3.1a is less active than the
alkylating agent carmustine, it shows a higher activity than
TMZ, the alkylat agent used in first line for primary and
recurrent glioblastoma.
The GP mean of 81%
18 of 3.1b indicates that it has a
generally weak antiproliferative activity. Because the median GP
values in different tissues are similar, we concluded that it is not
tissue specific. Nevertheless, two cell lines are moderately
sensitive to 3.1b with GP values comprised between 45 and
50%: a melanoma cell line (LOX-IMVI) and a renal cell line
(UO-31).
Next, the sensitivities of six CNS cell lines to paclitaxel,
TMZ, and 3.1a were compared. The antiproliferative activity of
3.1a used at 10 μM is higher than 10 μM paclitaxel on the two
cell lines, SF-295 and SNB75, and higher to 10 μM TMZ on
the six cell lines (Figure 2C). These last data prompted us to
To evaluate if the best antiproliferative phostine, 3.1a, is less
toxic for normal than clinical drugs and if the antiproliferative
activity occurs at lower concentrations than for toxicity, we next
measured its effect on primary culture of nonproliferative brain
cells.
2.5. Comparison of Toxicity between 3.1a and Clinical
Drugs on Nonproliferative Astrocytes in Primary
2200
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
and 3.1b, respectively. These values are much lower than the 10
and 25 μM measured using the LDH release test. Although the
LDH test is known as an insensitive assay, these differences
indicate that the MTT test measured inhibition of proliferation
rather than cell lysis.
2.7. Time−Course Effects of 3.1a on C6 Cells Viability.
To estimate the effect of 3.1a on cell density in time, we
measured impedance (cell index value) (Figure 6). Compound
3.1a cytotoxicity is confirmed at 25, 50, and 100 μM because
the cell index value measured at these concentrations was lower
than at t = 0. With 100 μM, the cell index value decreased as
soon as 1 h after drug addition. At 50 and 25 μM, a decrease of
the cell index value was only detected 2 and 3 h after 3.1a
addition. With 10 μM, the cell index value remained constant
throughout the experiment and below vehicle. This result
suggests that at 10 μM, 3.1a is cytostatic. The cell index value
still decreases with 5 μM 3.1a, while no decrease is observable
with 1 μM.
3. DISCUSSION AND CONCLUSION
From a chemical point of view, a two-step sequence starting
from aryl-H-phosphinates and substituted arabinofuranoses
afforded the synthesis of structurally diverse phostines. The
purification of the active compounds was achieved either by a
direct selective precipitation of phostines from the crude
mixture or by reverse-phase preparative HPLC. This method-
ology allowed an easy scale-up and offered the opportunities to
introduce stereochemical modifications leading to different
phostines. Qualification and/or quantification of the biological
effects perceived on the modified 2-oxo-2λ5-[1,2]-
oxaphosphinanes allowed the selection of five active
phosphino-heterocycles (3.1a, 3.7b, 11.1a, 16.1a, and 17.1a)
with IC50 values lower than 5 μM. Interestingly, the activity
profile was affected both by the nature of the substituent and by
their relative stereochemistry. The combination of these two
features has motivated the synthesis of the small focused library
presented in this paper. Refinement of our lead compounds
into a drug will require the extension of the chemical space.
Then, our work will be focused on the determination of a
quantitative structure−activity relationship to choose the best
drug candidate.
This study has evidenced that phostine drugs exhibit
stereospecific antiproliferative properties with respect to CNS
cell lines and, very importantly, that these drugs are devoid of
toxicity on normal astrocytes. Such an unusual brain tumor
specificity has been previously described for 2-methyl-9-
methoxyellipticinium.28 Also, when compared to paclitaxel,
3.1a has a different action on the CNS cell lines (Figure 3),
indicating different and potentially complementary modes of
action of these compounds.
In addition to CNS cell lines, 3.1a inhibits the growth of the
breast cancer HS578T and the nonsmall cell cancer lung HOP-
62 cell lines. Phylogenetic trees of the 57 NCI cell lines indicate
that these two cell lines cluster with the CNS mesenchymal cell
line cluster.29,30 Because microRNA expression has been shown
to be related to the chemosensitivity of tumor cells,31 we
looked into the microRNA expression of the NCI cancer cell
lines. By this analysis, HS578T also clusters with the CNS cell
lines cluster.32 Moreover, the HOP-62 also clusters with the
CNS cell lines (in dendrograms based on drug sensitivity).29
Although HS578T did not group with the CNS cluster on a
dendrogram-based cell line drug sensitivities,29 it appeared to
be sensitive to 3.1a. Therefore, we conclude that the drug
Figure 3. Comparison of the antiproliferative activities of phostines
with clinical drugs on C6 cells. Neutral red incorporation in C6 cells
was measured 48 h after product addition. The percentage of toxicity
was calculated in comparison to vehicle. nt, no toxicity. Results in bold
are above 50% of toxicity. OD, optical density.
Culture. Twenty-one day primary cell cultures of astrocytes
were not proliferative because they were not sensitive to
cytabarine (AraC) as measured by neutral red test (Figure 4A),
Hoechst nuclei staining (Figure 4B), and propidium iodide
incorporation (Figure 4C) and expressed markers of differ-
entiated cells like glial fibrillary acidic protein (GFAP) (data
not shown). At all concentrations, 3.1a has no toxic effect on
nonproliferative astrocytes, the major brain cell population. In
contrast, vincristine used at 1 and 10 μM induced a significant
decrease of viability (60% in Figure 4A), which was associated
with its cytotoxicity as shown by the incorporation of
propidium iodide (PI) in the majority of the remaining cells
(Figure 4C). However, at 0.1 μM, vincristine induced
cytotoxicity without a decrease of cell number (Figure 4A−
C). Paclitaxel used at 10 μM induced a weak decrease of
viability of 28% (Figure 4A) and a dose-dependent cytotoxicity
(Figure 4C) without affecting cell number (Figure 4B). Like
3.1a, TMZ and carmustine used at 0.1 and 100 μM neither
induced cytotoxicity nor decreased viability (Figure 4A−C).
Taken together, these data indicate that 3.1a has a better
antiproliferative activity on glioma than TMZ and a lower
toxicity than paclitaxel and a comparable one to carmustine on
21 day astrocytes. With no toxicity at 100 μM (Figure 4) and
an antiproliferative activity starting at 1 nM (Table 1), 3.1a
shows a good in vitro “therapeutic index”.
2.6. Cytotoxic Effect of 3.1a and 3.1b on C6 Cells. To
evaluate if the antiproliferative effects of the two diaster-
eisomers could be due to their cytotoxicity, we measured the
release of lactate dehydrogenase (LDH) in the cell medium.
This test shows that 3.1a and 3.1b are toxic for the C6
glioblastoma cell line and that at 10 μM 3.1a is more toxic than
3.1b (Figure 5). It has to be noticed that the LTC values
measured with the MTT test were 0.001 and 0.1 μM for 3.1a
2201
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
Figure 4. Comparison of the toxicity of OXP430.1 with clinical drugs on nonproliferative astrocytes. Astrocyte viability was quantified 48 h after
drug addition on nonproliferative cells maintained in cell culture during 3 weeks after isolation from mouse pups cortical brain. (A and B) Graphics
with means
standard deviations (statview) show neutral red incorporation level (A) and Hoechst 33342 (5 μg/mL) cytofluor intensity
measurement. The mean fluorescence was then calculated from three different wells per condition. (C) Dual labeling for PI and Hoechst was
performed using the Cellomics array scan vti (Thermofisher). The object detection was fixed on the Hoechst 33342 channel, and the graph shows
the percentage of Hoechst 33342 cells stained with PI. The control fluorescence intensities were measured on Triton X-100 (2%)-treated cells. The
mean percentage of PI incorporation was calculated from three different wells per condition. OD, optical density; and FU, fluorescent unit. *,
significant activity (p < 0.05); ***, significant activity (p < 0.001); and $, significant activity (p < 0.0001).
sensitivity of the CNS cell lines, HOP-62 and HS-578T, can be
attributed to a common biochemical mechanism depending on
a set of genes commonly expressed by these cell lines. Running
the COMPARE program from the NCI evidenced no other
clinically used standard agents with a similar profile as 3.1a.
Taken together, these data suggest that 3.1a inhibits a
biochemical process that has not been targeted by standard
anticancer agents. Otherwise, the selective cytotoxicity to this
2202
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
astrocytes. The easy tuning of the substituent groups at the
phosphorus and carbon 2 atoms of phostines will allow the
development of this family for therapeutic use.
4. EXPERIMENTAL SECTION
4.1. Chemistry. Commercial reagents were purified by distillation
or sublimation, and 2,3,5-tri-O-benzyl-D-arabinofuranose was pur-
chased from Carbosynth and dried under vacuum before use. The
analytical method used to determine the purity of all of the
compounds described in this paper has been determined by
phosphorus and proton NMR confirming a purity ≥95%. Analytical
and preparative HPLC show a purity ≥99% for the compounds 3.1a
and 3.1b for which the biological activities have been described.
All reactions were carried out under dry nitrogen in flame-dried
glassware. Solvents were dried according to current method, distilled,
and stored under nitrogen atmosphere.
Figure 5. Cytotoxic activity of 3.1a and 3.1b on C6 cells. The graph
shows the mean
standard deviation (statview) optical density
measured at 490 nm 48 h after product addition on C6 glial cells. $,
significant activity (p < 0.0001); and ***, significant activity (p <
0.001).
Nuclear magnetic resonance (1H, 31P, 13C, and 19F NMR) spectra
were recorded at 400.13 MHz for proton, 161.97 MHz for
phosphorus, 100.6 MHz for carbon, and 376.50 for fluorine (Bruker
DRX 400). Chemical shifts are reported in ppm relative to the solvent
signals [1H: chloroform-d1 (7.26 ppm), acetone-d6 (2.05 ppm),
dimethylsulfoxide (DMSO)-d6 (2.50 ppm), D2O (4.79 ppm), or
methanol-d4 (3.31 ppm); 31P: H3PO4 85%; 13C: chloroform-d1 (77.16
ppm), acetone-d6 (29.84 ppm), DMSO-d6 (39.52 ppm), or methanol-
d4 (49.00 ppm); 19F: reported from CFCl3]. NMR coupling constants
are noted J and reported in Hertz (Hz). These abbreviations are used
for express the multiplicity: s (singulet), d (doublet), t (triplet), q
(quadruplet), qu (quintuplet), m (multiplet), and b (broad).
Low- and high-resolution mass spectra were acquired using a
WATERS Micromass Q-Tof spectrometer with as internal reference
H3PO4 (0.1% in water/acetonitrile, 1/1) in positive mode (ESI). Flash
chromatography was performed using a CombiFlash Companion/TS
with prepacked column (6−120 g scale) with particle size of 35−70
μm. Analytical HPLC (column, Waters SunFire C18, 5 μm, 4.6 mm ×
250 mm; eluent, acetonitrile/water (63:34); flow, 1 mL min−1).
Preparative HPLC (Waters SunFire Prep C18 OBD; column, Waters
SunFire C18, 50 mm × 250 mm; flow, 17 mL min−1).
4.2. General Procedure for Oxaphosphinane Synthesis:
Compounds 3. Ethyl arylphosphinate was prepared using the
method described in the literature by our group.21 Under dry
nitrogen, 2,3,5-tri-O-benzyl-D-arabinofuranose (1 equiv) was added
under nitrogen to a solution of ethyl arylphosphinate (1 equiv) in
THF (1.4 mmol/mL), and then, freshly sublimated potassium tert-
butoxide (0.2 equiv) was added to the solution. The reaction mixture
was stirred at room temperature for 15 h. After evaporation of the
solvent under vacuum, chloroform (3 mmol/mL) was added to the
crude oil. The organic solution was washed with a saturated aqueous
solution of ammonium chloride (3 × 1.5 mmol/mL). The organic
cell line cluster might be the consequence of a preferential
uptake in the sensitive lines. Thus, in such a way, stereo-
selectivity will play a central role in a metabolism where
enzymes are involved.
Finally, it is important to note that 3.1a is toxic for different
cell types identified as drug-resistant. Compound 3.1a kills one
multidrug-resistant CNS cell line (SF-295) and reduces the cell
growth of two other MDR cell types: the nonsmall cell lung
tumor cell line (HOP-62) and to a lesser extent the renal tumor
cell line (RXF-393).33 Moreover, a greater sensitivity to
carmustine was reported for SF-268 than for SF-295,34 and
we observe a greater toxicity of 3.1a on SF-295 than on SF-268.
This last result suggests that 3.1a could be used on strains
developing resistance against alkylating agents.
When taken into account, the results from the different
viability tests on C6 glioblastoma cells, the effects of 3.1a are
dual. At concentrations higher than or equal to 10 μM, 3.1a
induces necrosis as revealed by LDH release. At concentrations
below 5 μM, we observed stable cell index values, a decrease of
MTT reduction, a decrease of neutral red accumulation, and no
LDH release. These data suggest that at least two different
effects are mediated by the drugs. At high concentrations, 3.1a
is toxic, whereas at medium concentrations (between 1 and 10
μM), a cytostatic effect is observed.
In conclusion, this work characterized a new family of
compounds targeting CNS cancer without affecting normal
Figure 6. Time−course effect of 3.1a on C6 cell growth. Compound 3.1a was used at concentrations ranging between 100 nM and 100 μM, and
impedance was measured 25 h after plating (t = 0) for 44 h. The electrode impedance is displayed as a cell index (CI) value.
2203
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
layer was dried over sodium sulfate and filtered, and the solvent was
evaporated under vacuum.
4.2.4. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[3-
methylpheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.2a−d).
4.2.1. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-phenyl-
2-oxo-2λ5-[1,2]oxaphosphinane (3.1). From ethyl phenylphosphi-
nate (8.5 g, 50 mmol), a yellow oil residue containing a mixture of four
diastereomers was obtained (25.7 g, 28/26/19/27). The solid was
purified by normal phase chromatography on silica gel with a mixture
of chloroform/ethyl acetate (60/40) as the eluent to give a mixture of
four diastereomers 3.1a−d (18.3 g, 67%). After chromatography on
preparative reverse-phase HPLC with a Waters SunFire column C18,
50 mm × 250 mm, and with a mixture of acetonitrile/water (60:40,
flow of 17 mL min−1) as the eluent, 3.1a, 3.1b, 3.1c, and 3.1d were all
separated and recovered purely.
From ethyl 3-methylphenylphosphinate (3.00 g, 16.3 mmol),
2,3,5-tri-O-benzyl-D-arabinofuranose (6.85 g, 16.3 mmol), and
freshly sublimated potassium tert-butoxide (0.37 g, 3.3 mmol),
3.2 was isolated as a yellow oil containing a mixture of four
diastereomers (26/29/16/29, 2.54 g, 28%). 31P NMR (161.97
MHz, DMSO-d6): δ = 37.37, 35.68, 33.41, 32.35 ppm. HRMS
(ESI) m/z calcd for C33H36O6P (M + H)+, 559.2250; found,
559.2243.
4.2.2. (SPSSSR)-3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-
2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (3.1a).
4.2.5. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[3-
methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.3a−d).
1
31P NMR (161.97 MHz, CDCl3): δ = 36.1 ppm. H NMR
(400.13 MHz, CDCl3): δ = 3.69 (dd, 1H, J = 11.1, 2.3 Hz),
3.88 (ddd, 1H, J = 11.1, 2.7, 2.5 Hz), 3.90 (dd, 1H, J = 9.5, 9.3
Hz), 3.93 (dd, 1H, 3J = 9.8, 2.0 Hz), 4.10 (ddd, 1H, J = 9.8, 9.5,
2.6 Hz), 4.46 (dddd, 1H, J = 9.3, 3.1, 2.7, 2.3 Hz), 4.46 (d, 1H,
J = 12.1 Hz), 4.54 (d, 1H, J = 12.1 Hz), 4.59 (d, 1H, J = 10.8
Hz), 4.83 (d, 1H, J = 11.1 Hz), 4.84 (d, 1H, J = 10.8 Hz), 4.88
(d, 1H, J = 11.1 Hz), 7.38−7.43 (m, 15H), 7.50−7.55 (m, 3H),
7.75−7.80 (m, 2H) ppm. 13C NMR (100.6 MHz, CDCl3): δ =
68.64 (d, J = 9.2 Hz), 72.19 (d, J = 97.3 Hz), 73.45, 75.24 (d, J
= 5.2 Hz), 75.50, 76.25, 77.65, 84.34 (d, J = 6.9 Hz), 127.65,
127.76, 127.82, 127.84, 128.15, 128.45, 127.49 (d, J = 135.8
Hz), 128.71 (d, J = 13.4 Hz), 131.91 (d, J = 10.4 Hz), 133.30
(d, J = 2.8 Hz), 137.92, 137.98, 138.37 ppm. HRMS (ESI) m/z
calcd for C32H34O6P (M + H)+, 545.2093; found, 545.2089.
Analytical HPLC: column Waters SunFire C18, 5 μm, 4.6 mm
× 250 mm, acetonitrile/water (63/37); flow, 1 mL min−1;
retention time = 18.53 min.
From ethyl 3-methoxyphenylphosphinate (4.0 g, 20.0 mmol),
2,3,5-tri-O-benzyl-D-arabinofuranose (8.4 g, 20.0 mmol) and
freshly sublimated potassium tert-butoxide (0.45 g, 4.0 mmol),
3.3 was isolated as a yellow oil containing a mixture of four
diastereomers (23/26/31/20, 3.56 g, 31%). 31P NMR (161.97
MHz, DMSO-d6): δ 37.8, 36.6, 34.4, 34.2 ppm. HRMS (ESI)
m/z calcd for C33H36O7P (M + H)+, 575.2199; found,
575.2228.
4.2.6. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[3-
fluoropheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.4a−d).
4.2.3. (SPRSSR)-3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-
2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (3.1b).
From ethyl 3-fluorophenylphosphinate (0.85 g, 4.5 mmol),
2,3,5-tri-O-benzyl-D-arabinofuranose (1.89 g, 4.5 mmol) and
freshly sublimated potassium tert-butoxide (0.10 g, 0.9 mmol),
3.4 was isolated as a yellow oil containing a mixture of four
diastereomers (27/29/18/26, 0.53 g, 21%). 31P NMR (161.97
MHz, DMSO-d6): δ 34.9 (d, J = 7.9 Hz), 34.1 (d, J = 7.9 Hz),
32.3 (d, J = 1.9 Hz), 29.8 (d, J = 5.9 Hz) ppm. HRMS (ESI) m/
z calcd for C32H33FO6P (M + H)+, 563.1967; found, 563.1964.
31P NMR (161.97 MHz, CDCl3): δ = 36.6 ppm. H NMR
1
(400.13 MHz, CDCl3): δ = 2.67 (s, 1H), 3.80 (dd, 1H, J = 11.3,
2.0 Hz), 4.00 (ddd, 1H, J = 11.3, 3.3, 3.2 Hz), 4.30 (dd, 1H, J =
9.5, 2.8 Hz), 4.36 (dd, 1H, J = 9.5, 2.8 Hz), 4.43 (dd, 1H, J =
3.5, 2.8 Hz), 4.56 (dddd, 1H, J = 10.0, 3.3, 3.2, 2.0 Hz), 4.59 (d,
1H, J = 12.1 Hz), 4.62−4.80 (m, 4H), 4.93 (d, 1H, J = 10.8
Hz), 7.24−7.38 (m, 15H), 7.48−7.54 (m, 2H), 7.62−7.68 (m,
1H,), 7.96−8.03 (m, 2H) ppm. 13C NMR (161.99 MHz,
CDCl3): δ = 66.81 (d, J = 105.1 Hz), 68.91 (d, J = 8.8 Hz),
72.48, 73.43, 73.74 (d, J = 2.2 Hz), 75.46 (d, J = 5.6 Hz), 75.64,
81.38 (d, J = 2.9 Hz), 126.53 (d, J = 138.8 Hz), 127.23, 127.68,
127.80, 127.84, 127.98, 128.20, 128.31 (d, J = 13.4 Hz), 128.42,
128.65, 132.99 (d, J = 9.9 Hz), 133.34 (d, J = 2.8 Hz), 137.32,
138.07, 138.09 ppm. HRMS (ESI) m/z calcd for C32H34O6P
(M + H)+, 545.2093; found, 545.2092. Analytical HPLC:
column Waters SunFire C18, 5 μm, 4.6 mm × 250 mm,
acetonitrile/water (63/37); flow, 1 mL min−1; retention time =
15.97 min.
4.2.7. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[3-
trifluoromethylpheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.5a−d).
From ethyl 3-trifluoromethylphenylphosphinate (1.07 g, 4.5
mmol), 2,3,5-tri-O-benzyl-D-arabinofuranose (1.89 g, 4.5
mmol) and freshly sublimated potassium tert-butoxide (0.10
g, 0.9 mmol), 3.5 was isolated as a yellow oil containing a
mixture of four diastereomers (26/32/19/23, 0.63 g, 23%). 31
P
NMR (161.97 MHz, DMSO-d6): δ 35.0, 34.3, 32.3, 30.1 ppm.
HRMS (ESI) m/z calcd for C33H33F3O6P (M + H)+, 613.1886;
found, 613.3000.
2204
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
4.2.8. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[4-
127.68, 127.78, 127.99, 128.17, 128.42, 128.64, 135.02, 135.12,
137.40, 138.12, 163.67 ppm. HRMS (ESI) m/z calcd for
C33H36O7P (M + H)+, 575.2199; found, 575.2191. Analytical
HPLC: column Waters SunFire C18, 5 μm, 4.6 mm × 250 mm,
acetonitrile/water (60:40); flow, 1 mL min−1; retention time =
20.12 min.
methylpheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.6a−d).
4.2.12. 3-O-Trifluoromethanesulfonyl-4,5-bis-benzyloxy-6-benzy-
loxymethyl-2-[4-methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane
(4a,b).
From ethyl 4-methylphenylphosphinate (3.68 g, 20.0 mmol),
2,3,5-tri-O-benzyl-D-arabinofuranose (8.40 g, 20.0 mmol) and
freshly sublimated potassium tert-butoxide (0.45 g, 4.0 mmol),
3.6 was isolated as a yellow oil containing a mixture of four
diastereomers (30/25/15/30, 3.83 g, 34%). 31P NMR (161.97
MHz, DMSO-d6): δ 37.1, 35.7, 33.3, 32.1 ppm. HRMS (ESI)
m/z calcd for C33H36O6P (M + H)+, 559.2250; found,
559.2245.
Under dry nitrogen, pyridine (0.63 g, 0.64 mL, 8.0 mmol) and
trifluoromethanesulfonic anhydride (1.13 g, 0.66 mL, 4.0
mmol) were added to a solution of a mixture of 3.1a/3.1b
(1.09 g, 2.0 mmol, 60/40) in dichloromethane (CH2Cl2) (20
mL) at 0 °C. The reaction mixture was stirred for 2 h, and then,
water (10 mL) was added. The aqueous layer was extracted
with ethyl acetate (3 × 10 mL). The combined organic layers
were dried over magnesium sulfate, filtered, and concentrated
under vacuum to give yellow oil. The oily residue was purified
by normal-phase chromatography on silica gel with a mixture of
n-heptane/ethyl acetate (60/20) as the eluent, and a mixture of
4.1a/4.1b (60/40) was obtained in 46% yield. 31P NMR
(161.97 MHz, DMSO-d6): δ = 24.29, 27.92 ppm.
4.2.9. 3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymethyl-2-[4-
methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.7a−d).
From ethyl 4-methoxylphenylphosphinate (4.42 g, 22.1 mmol),
2,3,5-tri-O-benzyl-D-arabinofuranose (9.27 g, 22.1 mmol) and
freshly sublimated potassium tert-butoxide (0.49 g, 4.4 mmol),
3.7 was isolated as a yellow oil containing a mixture of four
diastereomers (27/25/18/30, 6.32 g, 50%). 31P NMR (161.97
MHz, DMSO-d6): δ 37.3, 37.1, 35.2, 32.5 ppm. HRMS (ESI)
m/z calcd for C33H36O7P (M + H)+, 575.2199; found,
575.2206.
4.2.13. (SPSSSR)-3-O-Trifluoromethanesulfonyl-4,5-bis-benzyloxy-
6-benzyloxymethyl-2-[4-methoxypheny]l-2-oxo-2λ5-[1,2]-
oxaphosphinane (4.1a).
4.2.10. (SPSSSR)-3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymeth-
yl-2-[4-methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.7a).
31P NMR (161.97 MHz, CDCl3): δ 28.1 ppm. 19F NMR
(376.50 MHz, CDCl3): δ −75.00 ppm. 1H NMR (400.13 MHz,
CDCl3): δ 3.68 (dd, 1H, J = 11.2, 2.0 Hz), 3.88 (dt, 1H, J =
11.1, 2.5 Hz), 4.09 (t, 1H, J = 9.7 Hz), 4.36 (m, 1H), 4.40−4.60
(m, 4H), 4.80 (d, 2H, J = 10.6 Hz), 4.900 (d, 1H, J = 10.4 Hz),
4.98 (dd, 1H, J = 10.4, 1.0 Hz), 7.00−7.30 (m, 15H), 7.49 (m,
2H), 7.62 (m, 1H), 7.80 (m, 2H) ppm. 13C NMR (100.61
MHz, CDCl3): δ 67.85, 67.95, 73.51, 75.52, 75.58, 75.64, 78.49,
80.31, 80.35, 80.94, 81.90, 116.38, 119.56, 124.10, 125.54,
127.58, 127.67, 127.71, 127.80, 127.92, 127.93, 128.01, 128.10,
128.37, 128.52, 128.60, 128.62, 128.96, 129.10, 132.37, 132.48,
134.48, 134.51, 137.03, 137.32, 137.55. HRMS (ESI) m/z calcd
for C32H33O8PF3S (M + H)+, 677.1586; found, 677.1582.
4.2.14. (SPRSSR)-3-O-Trifluoromethanesulfonyl-4,5-bis-benzyl-
oxy-6-benzyloxymethyl-2-[4-methoxypheny]l-2-oxo-2λ5-[1,2]-
oxaphosphinane (4.1b).
31P NMR (161.97 MHz, CDCl3): δ 37.6 ppm. 1H NMR
(400.13 MHz, CDCl3): δ 3.68 (dd, 1H, J = 11.1, 2.1 Hz), 3.77
(s, 3H), 3.85−3.91 (m, 3H), 4.09 (dd, 1H, J = 25.5, 9.1 Hz),
4.44−4.60 (m, 4H), 4.81−4.89 (m, 3H), 6.88−7.72 (m, 19H)
ppm. 13C NMR (100.61 MHz, CDCl3): δ 55.43, 68.58, 72.57,
73.47, 75.09, 75.44, 76.15, 77.69, 113.66, 114.27, 114.41,
127.64, 127.75, 127.82, 127.85, 127.89, 128.06, 128.44, 128.46
ppm. HRMS (ESI) m/z calcd for C33H36O7P (M + H)+,
575.2199; found, 575.2193. Analytical HPLC: column Waters
SunFire C18, 5 μm 4.6 mm × 250 mm, acetonitrile/water
(60:40); flow, 1 mL min−1; retention time = 23.51 min.
4.2.11. (SPRSSR)-3-Hydroxy-4,5-bis-benzyloxy-6-benzyloxymeth-
yl-2-[4-methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (3.7b).
31P NMR (161.97 MHz, DMSO-d6): δ 33.4 ppm.
4.2.15. (SPSSSR)-3-Azido-4,5-bis-benzyloxy-6-benzyloxymethyl-2-
[4-methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (5.1a).
31P NMR (161.97 MHz, CDCl3): δ 36.9 ppm. 1H NMR
(400.13 MHz, CDCl3): δ 3.60−3.79 (m, 2H), 3.87 (s, 3H), 3.9
(d, 1H, J = 9.6 Hz), 4.20−4.35 (m, 4H), 4.40−4.90 (m, 6H),
4.83 (d, 1H, J = 10.6 Hz), 6.90 (dd, 2H, J = 8.6, 2.9 Hz), 7.00−
7.35 (m, 15H), 7.83 (dd, 2H, J = 11.5, 8.8 Hz) ppm. 13C NMR
(100.61 MHz, CDCl3): δ 55.39, 68.92, 69.00, 72.42, 73.42,
73.76, 75.33, 75.61, 76.75, 77.07, 77.39, 81.46, 113.87, 114.01,
2205
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
4.2.18. (SPRSSR)-3-Amino-4,5-bis-benzyloxy-6-benzyloxymethyl-
2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (6.1b). Starting from
5.1b.
Under dry nitrogen, trimethylsilyl azide (0.173 g, 0.197 mL, 1.5
mmol) and a solution of tetrabutylammonium fluoride (1.5
mmol) in THF (15 mL) were added to a solution of 4.1a/4.1b
(0.406 g, 0.6 mmol, 40/60) in THF (20 mL). The reaction was
stirred at 65 °C overnight, and then, the solvent was evaporated
under vacuum to give a yellow oil. The residue was purified by
chromatography on silica gel with as eluent a mixture of n-
heptane/ethyl acetate (60/40) to give 5.1b and 5.1a in an
overall yield of 62%.
31P NMR (161.97 MHz, CDCl3): δ 39.5 ppm. 1H NMR
(400.13 MHz, CDCl3): δ 1.65 (s, 2H, NH2), 3.71 (dd, 1H, J =
11.1, 2.0 Hz), 3.77 (dd, 1H, J = 5.8, 3.3 Hz), 3.88 (ddd, 1H, J =
11.1, 3.0, 2.5 Hz), 4.25−4.31 (m, 2H, 4.41−4.65 (m, 6H), 4.84
(d, 1H, J = 10.84 Hz), 7.05−7.21 (m, 15H), 7.43−7.48 (m,
2H), 7.54−7.60 (m, 1H), 7.77−7.93 (m, 2H, J = 11.6, 7.4, 1.5
Hz) ppm. 13C NMR (100.6 MHz, CDCl3): δ 49.37, 50.34,
68.94, 69.04, 72.14, 73.29, 73.32, 73.40, 75.42, 75.60, 75.65,
80.99, 127.74, 127.87, 127.95, 128.39, 128.44, 128.54, 128.57,
132.69, 132.79, 133.16, 133.19, 137.75, 138.04, 138.14 ppm.
HRMS (ESI) m/z calcd for C32H35NO5P (M + H)+, 544.2253;
found, 544.2245.
1
31P NMR (161.97 MHz, CDCl3): δ 31.9 ppm. H NMR (400.13
MHz, CDCl3): δ 3.61 (dd, 1H, J = 10.6, 5.8 Hz), 3.70 (dd, 1H, J =
11.1, 2.0 Hz), 3.91 (dt, 1H, J = 11.1, 2.5 Hz), 4.00 (t, 1H, J = 9.7 Hz),
4.21 (m, 1H, J = 10.6, 9.3, 2.5 Hz), 4.40−4.70 (m, 4H), 4.80−4.90 (m,
3H), 7.10−7.35 (m, 15H), 7.42−7.52 (m, 2H), 7.55−7.63 (m, 1H),
7.78−7.86 (m, 1H) ppm. 13C NMR (100.61 MHz, CDCl3): δ 61.65,
62.60, 68.32, 68.41, 73.50, 75.27, 75.33, 75.53, 76.25, 78.41, 81.74,
81.77, 126.26, 127.63, 127.66, 127.67, 127.85, 127.94, 128.05, 128.11,
128.49, 128.52, 128.90, 129.04, 131.92, 132.02, 133.77, 133.80, 137.43,
137.67, 137.78 ppm. HRMS (ESI) m/z calcd for C32H33N3O5P (M +
H)+, 570.2158; found, 570.2151.
4.2.16. (SPRSSR)-3-Azido-4,5-bis-benzyloxy-6-benzyloxymethyl-2-
[4-methoxypheny]l-2-oxo-2λ5-[1,2]oxaphosphinane (5.1b).
4.2.19. (SPRSSR)-3-Phenylureido-4,5-bis-benzyloxy-6-benzyloxy-
methyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (7.1a).
31P NMR (161.97 MHz, CDCl3): δ 33.4 ppm. 1H NMR
(400.13 MHz, CDCl3): δ 3.77 (dd, 1H, J = 11.2, 1.9 Hz), 3.97
(dt, 1H, J = 11.2, 2.9, Hz), 4.26 (dd, 1H, J = 8.2, 3.2 Hz), 4.30
(d, 1H, J = 9.5 Hz), 4.50 (dd, 1H, J = 9.5, 3.2 Hz), 4.54 (dm,
1H, J = 9.9 Hz), 4.59 (d, 1H, J = 12.4 Hz), 4.64 (d, 1H, J = 10.6
Hz), 4.72 (d, 1H, J = 12.4 Hz), 4.79 (s, 2H), 4.94 (d, 1H, J =
10.9 Hz), 7.18−7.42 (m, 15 H), 7.52−7.60 (m, 2H), 7.68 (tq,
1H, J = 7.7, 1.3 Hz), 7.93−8.00 (m, 2H) ppm. 13C NMR (100.6
MHz, CDCl3): δ 59.04, 60.00, 68.67, 68.77, 68.72, 72.99, 73.41,
74.06, 74.08, 75.74, 76.01, 76.07, 80.81, 80.84, 127.69, 127.74,
127.84, 127.90, 127.93, 128.11, 128.42, 128.45, 128.54, 128.62,
128.68 ppm. HRMS (ESI) m/z calcd for C32H33N3O5P (M +
Under dry nitrogen, phenyl isocyanate (0.024 g, 0.20 mmol)
was added to a solution of 6.1a (0.100 g, 0.18 mmol) in
CH2Cl2 (3 mL). The reaction mixture was stirred at room
temperature overnight, and then, the solvent was evaporated
under vacuum. The residue was purified by chromatography on
silica gel with as eluent a mixture of n-heptane/ethyl acetate
(100/0 to 0/100) to give 7.1a (0.063 g, 53%).
1
31P NMR (161.97 MHz, CDCl3): δ 39.0 ppm. H NMR (400.13
MHz, CDCl3): δ 3.81 (dd, 1H, J = 11.1, 1.8 Hz), 3.94 (ddd, 1H, J =
10.9, 2.8, 2.5 Hz), 4.07 (ddd, 1H, J = 10.6, 9.4, 2.8 Hz), 4.18 (dd, 1H, J
= 9.3, 9.3 Hz), 4.53−4.83 (m, 8H), 6.87−8.49 (m, 25H) ppm. 13C
NMR (100.6 MHz, CDCl3): δ 50.73 (d, J = 97.3 Hz), 68.65 (d, J = 9.5
Hz), 73.36, 75.45, 76.27 (d, J = 5.9 Hz), 76.63, 78.30, 82.97 (d, J = 8.8
Hz), 118.25, 121.86, 125.47−128.73 (m, 15C), 127.10 (d, J = 23.6
Hz), 129.12 (d, J = 13.2 Hz), 131.74 (d, J = 11.0 Hz), 133.87, 137.59,
137.80, 137.87, 139.71, 155.22 ppm. HRMS (ESI) m/z calcd for
C39H40N2O6P (M + H)+, 663.2631; found, 663.2624.
H)+, 570.2167; found, 570.2158.
4.2.17. (SPSSSR)-3-Amino-4,5-bis-benzyloxy-6-benzyloxymethyl-
2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (6.1a).
4.2.20. (SPSSSR)-3-Dithiocarbonyl-S-methylester-4,5-bis-benzyl-
oxy-6-benzyloxymethyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane
(8.1a).
Under dry nitrogen, polymer-bound triphenylphosphine (3
mmol.g−1, 0.17 g, 0.510 mmol) and water (0.03 g, 0.03 mL,
1.67 mmol) were added to a solution of 5.1a (0.095 g, 0.167
mmol) in THF (2.5 mL). The reaction mixture was stirred at
60 °C for 3 h, cooled at room temperature, and filtered through
microporous filter. The solvent was evaporated under vacuum
to give 6.1a as colorless oil (0.074 g, 82%).
Under dry nitrogen, carbon disulfide (2.34 g, 1.90 mL, 30.7
mmol) and sodium hydride (60% in oil, 0.16 g, 4.0 mmol) were
added to a solution of a mixture 3.1a/3.1b (1.69 g, 3.1 mmol,
40/60) in DMF (15 mL). The reaction mixture was stirred at
room temperature for 30 min, and then, methyl iodide (4.36 g,
1.90 mL, 30.7 mmol) was added. After 1 h of stirring, water (30
mL) was added, and the aqueous layer was extracted with ethyl
acetate (3 × 30 mL). The combined organic layers were dried
over magnesium sulfate, filtered, and concentrated under
vacuum to give a white solid (1.70 g). The solid was purified
by normal-phase chromatography on silica gel with a mixture of
1
31P NMR (161.97 MHz, CDCl3): δ 38.7 ppm. H NMR (400.13
MHz, CDCl3) δ 1.46 (s, 2H), 2.97 (dd, 1H, J = 10.1, 9.8 Hz), 3.70
(dd, J = 11.8, 2.0 Hz), 3.82 (m, 1H, J = 10.1, 9.0, 2.5 Hz), 3.90 (dd,
1H, J = 9.6, 9.0 Hz), 3.92 (ddd, 1H, J = 11.8, 4.3, 2.5 Hz), 4.44−5.10
(m, 7H), 7.05−7.18 (m, 15H), 7.41−7.5 (m, 2H), 7.5−7.62 (m, 1H),
7.79−7.90 (m, 1H) ppm. 13C NMR (100.6 MHz, CDCl3): δ 53.82,
54.79, 68.61, 68.70, 73.49, 75.32, 75.36, 75.41, 76.35, 76.72, 78.82,
84.76, 84.82, 127.24, 127.66, 127.77, 127.82, 127.92, 127.99, 128.07,
128.45, 128.52, 128.59, 128.68, 128.73, 132.95, 133.05, 133.13, 133.11,
137.84, 137.93, 138.08 ppm. HRMS (ESI) m/z calcd for C32H35NO5P
(M + H)+, 544.2253; found, 544.2258.
2206
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
4.2.23. (SPSSSR)-3-Methoxy-4,5-bis-benzyloxy-6-benzyloxymeth-
n-heptane/ethyl acetate (70/30) as the eluent to give 8.1a and
8.1b (overall yield, 75%).
yl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (10.1a).
1
31P NMR (161.97 MHz, CDCl3): δ 30.1 ppm. H NMR (400.13
MHz, CDCl3): δ 2.39 (s, 3H), 3.69 (dd, 1H, J = 11.1, 1.5 Hz), 3.91 (d,
1H, J = 11.2 Hz), 4.08 (t, 1H, J = 9.7 Hz), 4.38−4.50 (m, 2H), 4.51−
4.65 (m, 3H), 4.7 (s, 2H), 4.80 (d, 1H, J = 10.9 Hz), 6.52 (d, 1H, J =
10.3 Hz), 7.18−7.25 (m, 15H), 7.35−7.43 (m, 2H), 7.45−7.58 (m,
1H), 7.72−7.83 (m, 2H). 13C NMR (100.6 MHz, CDCl3): δ 19.50,
68.31, 68.40, 73.50, 75.12, 75.17, 75.68, 76.28, 78.05, 78.20, 79.19,
81.92, 81.98, 126.30, 127.68, 127.78, 127.82, 127.87, 127.90, 128.08,
128.39, 128.48, 128.70, 128.83, 129.22, 130.80, 130.87, 132.05, 132.16,
133.68, 133.71, 137.66, 137.79, 193.66, 193.94, 194.70, 194.76, 215.47
ppm. HRMS (ESI) m/z calcd for C34H36O6PS2 (M + H)+, 635.1691;
found, 635.1682.
Under dry nitrogen, sodium hydride (50% in oil, 0.01 g, 0.021
mmol) was added to a solution of 3.1a (0.10 g, 0.18 mmol) in
dry THF (3 mL). The reaction mixture was stirred at room
temperature for 30 min, and then, methyl iodide (13 μL, 0.21
mmol) was added. The reaction mixture was stirred for one
night. After solvent evaporation under vacuum, ethyl acetate (3
mL) was poured into the crude oil. The combined organic
layers were dried over magnesium sulfate, filtered, and
concentrated under vacuum. The residue was purified by
chromatography on silica gel using a mixture of n-heptane/
ethyl acetate (100/0 to 0/100) as the eluent to give 10.1a as a
4.2.21. (SPRSSR)-3-Dithiocarbonyl-S-methylester-4,5-bis-benzyl-
oxy-6-benzyloxymethyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane
(8.1b).
white solid (0.015 g, 15%).
31P NMR (161.97 MHz, CDCl3): δ = 32.8 ppm. 1H NMR (400.13
MHz, CDCl3): δ 3,23 (s, 3H), 3.63 (d, 1H, J = 10.0 Hz), 3.67 (dd, 1H,
J = 11.1, 1.8 Hz), 3.88 (ddd, 1H, J = 11.1, 2.6, 2.4 Hz), 3.94 (dd, 1H, J
= 9.8, 9.6 Hz), 4.17 (ddd, 1H, J = 9.8, 9.6, 2.6 Hz), 4.43−4.62 (m,
3H), 4.78−4.89 (m, 2H), 7.12−7.24 (m, 15H), 7.43−7.57 (m, 3H),
7.81−7.85 (m, 2H) ppm. 13C NMR (100.6 MHz, CDCl3): δ 61.19,
68.59, 73.43, 74.85, 75.55, 76.43, 77.83, 81.30, 82.31, 83.46, 83.53,
127.59, 127.73, 127.77, 127.81, 127.99, 128.42, 128.44, 128.78,
128.91,131.78, 131.88, 133.25, 137.98, 138.32 ppm. HRMS (ESI)
m/z calcd for C33H36O6P (M + H)+, 559.2250; found, 559.2250.
4.2.24. (SPSSSR)-3-Benzyloxy-4,5-bis-benzyloxy-6-benzyloxy-
methyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (11.1a).
31P NMR (161.97 MHz, CDCl3): δ 32.1 ppm. 1H NMR
(400.13 MHz, CDCl3): δ 2.30 (s, 3H), 3.75 (dd, 1H, J = 11.1,
1.8 Hz), 3.95 (dt, 1H, J = 11.1, 2.8 Hz), 4.25 (t, 1H, J = 9.9
Hz), 4.45 (dd, 1H, J = 9.6, 2.5 Hz), 4.48 (d, 1H, J = 11.1 Hz),
4.50 (d, 1H, J = 11.9 Hz), 4.53 (d, 1H, J = 11.1 Hz), 4.57 (m,
1H), 4.77 (d, 1H, J = 11.1 Hz), 4.85 (d, 1H, J = 11.1 Hz), 6.88
(dd, 1H, J = 3.78, 2.78 Hz), 7.23−7.40 (m, 17H), 7.50 (m, 1H),
7.78 (m, 2H) ppm. 13C NMR (100.6 MHz, CDCl3): δ 18.02,
28.67, 67.85, 67.94, 71.11, 73.10, 73.25, 73.26, 74.06, 74.67,
75.05, 75.11, 78.67, 78.72, 123.82, 125.26, 126.53, 126.66,
126.76, 126.87, 126.92, 127.19, 127.35, 127.37, 127.39, 127.51,
131.72, 131.83, 132.59, 132.61, 136.18, 136.93, 137.10, 213.38
ppm. HRMS (ESI) m/z calcd for C34H36O6PS2 (M + H)+,
635.1691; found, 635.1698.
Under dry nitrogen, sodium hydride (50% oil, 0.01 g, 0.21
mmol) was added to a solution of 3.1a (0.10 g, 0.18 mmol) in
dry THF (3 mL). The reaction mixture was stirred at room
temperature for 30 min. Benzyl chloride (25 μL, 0.22 mmol)
was added, and the reaction mixture was stirred for one night.
After solvent evaporation under vacuum, ethyl acetate (3 mL)
was added to the crude oil. The combined organic layers were
dried over magnesium sulfate, filtered, and concentrated under
vacuum. The residue was purified by chromatography on silica
gel using as eluent a mixture of n-heptane/ethyl acetate (100/0
4.2.22. (SPSSR)-4,5-Bis-benzyloxy-6-benzyloxymethyl-2-phenyl-2-
oxo-2λ5-[1,2]oxaphosphinane (9).
Tributyltin hydride (0.23 g, 0.21 mL, 0.79 mmol) and AIBN
(0.002 g, 0.0122 mmol) were added to a solution of 8.1a/8.1b
(0.39 g, 0.61 mmol) in toluene (8 mL) at room temperature.
The reaction mixture was stirred at 115 °C for 3 h. The solvent
was evaporated affording a yellow oil. The residue was purified
by chromatography on silica gel with a mixture of n-heptane/
ethyl acetate (60:40) as the eluent to give 9 as a white solid
(0.094 g, 29%).
to 0/100) to give 11.1a as a white solid (0.045 g, 39%).
31P NMR (161.97 MHz, CDCl3): δ = 33.4 ppm. 1H NMR (400.13
MHz, CDCl3): δ 3.65 (dd, 1H, J = 11.2, 2.0 Hz), 3.85 (d, 1H, J = 10.2
Hz), 3.88 (ddd, 1H, J = 11.2, 2.6, 2.3 Hz), 3.92 (dd, 1H, J = 9.8, 9.6
Hz), 4.28 (ddd, 1H, J = 10.0, 9.8, 2.7 Hz), 4.39−4.61 (m, 6H), 4.84 (d,
1H, J = 10.7), 4.91 (d, 1H, J = 10.7), 6.80−6.90 (m, 2H), 6.95−7.05
(m, 2H), 7.05−7.15 (m, 3H), 7.15−7.25 (m, 13H), 7.35−7.45 (m,
2H), 7.05−7.15 (m, 3H), 7.50−7.55 (m, 1H), 7.65−7.75 (m, 2H)
ppm. 13C NMR (100.6 MHz, CDCl3): δ 68.52, 68.61, 73.44, 74.42,
74.46, 74.74, 75.54, 76.39, 77.43, 78.03, 78.43, 83.72, 83.79, 127.72,
127.86, 128.14, 128.17, 128.42, 128.44, 128.69, 128.82, 131.88, 131.98
ppm. HRMS (ESI) m/z calcd for C39H40O6P (M + H)+, 635.2563;
found, 635.2559.
1
31P NMR (161.97 MHz, CDCl3): δ 37.9 ppm. H NMR (400.13
MHz, CDCl3): δ 1.92 (m, 1H), 2.62 (dt, 1H, J = 13.6, 4.5 Hz), 3.68
(dd, 1H, J = 11.1, 2.0 Hz), 3.93 (m, 2H), 4.25 (m, 1H), 4.40−4.50 (m,
2H), 4.50−4.65 (m, 4H), 4.88 (d, 1H, J = 10.6 Hz), 7.25−7.30 (m,
15H), 7.30−7.47 (m, 2H), 7.49−7.58 (m, 1H), 7.70−7.78 (m, 2H)
ppm. 13C NMR (400.13 MHz, CDCl3): δ 32.47, 33.29, 68.96, 69.05,
72.14, 73.42, 75.44, 75.50, 75.53, 127.64, 127.69, 127.75, 127.80,
127.85, 127.90, 129.13, 130.51, 131.42, 131.53, 132.95, 132.98, 137.87,
138.09, 138.13 ppm. HRMS (ESI) m/z calcd for C32H34O5P (M +
H)+, 529.2144; found, 529.2136.
4.2.25. (SPSSSR)-3-Acetoxy-4,5-bis-benzyloxy-6-benzyloxymethyl-
2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (12.1a).
2207
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
Under dry nitrogen, acetic anhydride (0.062 g, 57 μL, 0.61
mmol), pyridine (30 μL, 0.41 mmol), and DMAP (0.005 g,
0.041 mmol) were added to a solution of 3.1a (0.225 g, 0.41
mmol) in dry CH2Cl2 (5 mL). The reaction mixture was stirred
for one night. After solvent evaporation under vacuum, ethyl
acetate (10 mL) was added to the crude oil. The combined
organic layers were dried over magnesium sulfate, filtered, and
concentrated under vacuum. The residue was purified by
normal-phase chromatography on silica gel using as eluent a
mixture of n-heptane/ethyl acetate (100/0 to 0/100) to give
12.1a (0.154 g, 64%).
acetate (100/0 to 0/100) as the eluent to give 14.1a (0.054 g,
31%).
1
31P NMR (161.97 MHz, CDCl3): δ 31.8 ppm. H NMR (400.13
MHz, CDCl3): δ 3.72 (d, 1H, J = 11.2 Hz), 3.92 (d, 1H, J = 11.0 Hz),
4.11 (t, 1H, J = 9.6 Hz), 4.41 (m,1H), 4.46−4.65 (m, 5H), 4.74 (d,
1H, J = 11.0 Hz), 4.84 (d, 1H, J = 11.0 Hz), 5.70 (d, 1H, J = 10.4 Hz),
6.90−7.60 (m, 21H), 7.70−7.85 (m, 2H), 7.94 (d, 1H, J = 7.2 Hz)
ppm. 13C NMR (100.6 MHz, CDCl3): δ 68.35, 68.44, 73.46, 75.29,
75.60, 76.33, 78.29, 81.60, 127.64, 127.83, 127.98, 128.27, 128.42,
128.48, 128.73, 128.87, 130.06, 131.90, 133.48, 137.43, 137.77, 137.85
ppm. MS (ESI) m/z for C39H38O7P (M + H)+, 649.3.
4.2.28. Compound 15.1.
1
31P NMR (161.97 MHz, CDCl3): δ 32.1 ppm. H NMR (400.13
MHz, CDCl3): δ 1.82 (s, 3H), 3.67 (dd, 1H, J = 11.2, 2.0 Hz), 3.90
(ddd, 1H, J = 11.2, 2.7, 2.4 Hz), 4.04 (t, 1H, J = 9.7 Hz), 4.23 (ddd,
1H, J = 10.4, 9.4, 2.9 Hz), 4.44−4.83 (m, 7H), 5.42 (dd, 1H, J = 10.4,
1.0 Hz), 7.12−7.25 (m, 15H), 7.40−7.50 (m, 2H), 7.50−7.60 (m,
1H), 7.74−7.80 (m, 2H) ppm. 13C NMR (100.6 MHz, CDCl3): δ
20.46, 68.31, 68.40, 70.16, 71.17, 73.46, 75.20, 75.25, 75.59, 76.21,
78.21, 81.77, 81.82, 127.60, 127.63, 127.87, 127.93, 128.49, 128.69,
131.82, 133.61, 133.64, 137.70, 137.83, 137.97, 169.54, 169.58 ppm.
HRMS (ESI) m/z calcd for C34H36O7P (M + H)+, 587.2199; found,
587.2186.
The synthesis was according to Cipolla et al.24
4.2.29. Compound 15.2.
4.2.26. (SPSSSR)-3-Trifluoroacetoxy-4,5-bis-benzyloxy-6-benzy-
The synthesis was according to Cipolla et al.24
loxymethyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (13.1a).
4.2.30. (SPSSSR)-3-Allylamino-4,5-bis-benzyloxy-6-benzyloxy-
methyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (16.1a).
Under dry nitrogen, trifluoroacetic anhydride (0.084 g, 56 μL,
0.40 mmol), pyridine (0.021 g, 21 μL, 0.27 mmol), and DMAP
(0.003 g, 0.027 mmol) were added to a solution of 3.1a (0.150
g, 0.27 mmol) in dry CH2Cl2 (3 mL). The reaction mixture was
stirred during night. Ethyl acetate (5 mL) was added to the
crude mixture, and the mixture was washed with water
saturated with NaCl. The organic layer was dried over
magnesium sulfate, filtered, and concentrated under vacuum.
The residue was purified by chromatography on silica gel with
as eluent a mixture of n-heptane/ethyl acetate (100/0 to 0/
Under dry nitrogen, compound 15.1 (10.0 g, 21.7 mmol) was
added to a solution of ethyl phenylphosphinate (3.7 g, 21.7
mmol) in THF (250 mL), and then, freshly sublimated
potassium tert-butoxide (0.7 g, 6.24 mmol) was poured to the
solution. The reaction mixture was stirred at room temperature
for one night. The crude mixture was washed with water
saturated with NaCl. The organic layer was dried over
magnesium sulfate, filtered, and concentrated under vacuum.
The residue was purified by chromatography on silica gel with
as eluent a mixture of n-heptane/ethyl acetate (100/0 to 0/
100) to give 16.1a (1.60 g, 12%).
100) to give 13.1a (0.069 g, 40%).
19F NMR (376.46 MHz, CDCl3): δ −74.4 ppm. 31P NMR (161.97
MHz, CDCl3): δ 28.7 ppm. 1H NMR (400.13 MHz, CDCl3): 3.81 (d,
1H, J = 11.2 Hz), 4.00 (d, 1H, J = 11.0 Hz), 4.18 (dd, 1H, J = 9.8, 9.6
Hz), 4.46 (dd, 1H, J = 9.4, 8.4 Hz), 4.50−5.00 (m, 5H), 5.48 (d, 1H, J
= 10.6 Hz), 7.15−7.45 (m, 15H), 7.50−7.63 (m, 2H), 7.65−7.75 (m,
1H), 7.80−7.95 (m, 2H) ppm. 13C NMR (100.6 MHz, CDCl3): δ
68.05, 68.13, 73.51, 74.11, 75.09, 75.42, 75.47, 75.67, 76.47, 78.17,
80.67, 127.66, 127.70, 127.74, 128.05, 128.53, 129.00, 129.14, 131.81,
131.91, 134.17, 137.16, 137.45, 137.64 ppm. MS (ESI) m/z for
C34H33F3O7P (M + H)+, 641.2.
1
31P NMR (161.97 MHz, CD3C(O)CD3): δ 35.5 ppm. H NMR
(400.13 MHz, CDCl3): 3.08−3.23 (m, 3H), 3.77 (dd, 1H, J = 11.0, 2.2
Hz), 3.98−4.15 (m, 3H), 4.45−5.05 (m, 9H), 5.45−5.65 (m, 1H),
7.20−7.40 (m, 15H), 7.46−7.56 (m, 1H), 7.58−7.66 (m, 1H), 7.85−
7.95 (m, 2H) ppm. Analytical HPLC [column, Waters SunFire C18, 5
μm, 4.6 mm × 250 mm; eluent, acetonitrile/water (80:20); flow, 1 mL
min−1; retention time = 13.66 min (99%)]. HRMS (ESI) m/z calcd
for C35H39NO5P (M + H)+, 584.2566; found, 584.2565.
4.2.27. (SPSSSR)-3-Benzoate-4,5-bis-benzyloxy-6-benzyloxymeth-
4.2.31. (SPSSSR)-3-Benzylamino-4,5-bis-benzyloxy-6-benzyloxy-
yl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (14.1a).
methyl-2-phenyl-2-oxo-2λ5-[1,2]oxaphosphinane (17.1a).
Under dry nitrogen, benzoic anhydride (0.090 g, 0.40 mmol),
pyridine (0.021 g, 21 μL, 0.27 mmol), and DMAP (0.003 mg,
0.027 mmol) were added to a solution of 3.1a (0.150 g, 0.27
mmol) in dry CH2Cl2 (3 mL). The reaction mixture was stirred
during night. Ethyl acetate (5 mL) was added to the crude
mixture and washed with water saturated with NaCl. The
organic layer was dried over magnesium sulfate, filtered, and
concentrated under vacuum. The residue was purified by
chromatography on silica gel with a mixture of n-heptane/ethyl
Under dry nitrogen, compound 15.2 (10.20 g, 19.6 mmol) was
added to a solution of ethyl phenylphosphinate (3.34 g, 19.6
mmol) in THF (250 mL), and then, freshly sublimated
potassium tert-butoxide (0.65 g, 5.8 mmol) was poured into the
solution. The reaction mixture was stirred at room temperature
for one night. The crude mixture was washed with a saturated
aqueous solution of NaCl. The organic layer was dried over
magnesium sulfate, filtered, and concentrated under vacuum.
2208
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
OXP (1 nM to 100 μM) were measured 25 h after plating (t = 0) for
44 h. The electrode impedance was displayed as a cell index (CI)
value. The mean CI value calculated from triplicate measurements was
normalized to vehicle-treated cells. The primary effects of OXP were
determined by a statistical analysis of the 0−5 h interval after drug
addition. Indeed, when plates were coated with an inhibitor of cell
adhesion like polyhydoxymethacrylate (data not shown), the CI value
remained constant over time. The primary effects of 3.1a were
determined by a statistical analysis on the 0−5 h interval after drug
addition.
4.6. Statistical Analyses. EC50 values were calculated using the
Hill equation of the dose−log response curves using Prism 5.0
software (Graphpad) using the vehicle as the top value and Triton X-
100 as the bottom value. For group comparison, one-way ANOVA
followed by Fisher's posthoc PLSD test was performed using the
Statview 5.0 software.
The residue was purified by chromatography on silica gel with
as eluent a mixture of n-heptane/ethyl acetate (100/0 to 0/
100) to give 17.1a (0.97 g, 7%).
31P NMR (161.97 MHz, CDCl3): δ = 36.8 ppm. 1H NMR (400.13
MHz, CD3C(O)CD3): 3.08 (dd, 1H, J = 10.1, 5.5 Hz), 3.64−3.70 (m,
3H), 3.844.10 (m, 3H), 4.30−5.00 (m, 7H), 6.61−6.90 (m, 2H),
6.90−7.05 (m, 3H), 7.05−7.15 (m, 2H), 7.15−7.30 (m, 13H), 7.30−
7.45 (m, 2H), 7.45−7.60 (m, 2H), 7.70−7.85 (m, 2H) ppm. 13C NMR
(100.6 MHz, CDCl3): δ 52.58, 52.65, 53.62, 58.63, 60.09, 68.69, 68.84,
71.76, 74.94, 75.02, 75.35, 76.47, 76.62, 77.12, 78.76, 83.96, 84.08,
126.88, 127.63, 127.71, 127.76, 127.84, 128.08, 128.13, 128.42, 128.47,
128.48, 128.59, 128.64, 128.80, 130.75, 131.87, 132.04, 132.98, 133.02
ppm. HRMS (ESI) m/z calcd for C39H41NO5P (M + H)+, 634.2722;
found, 634.2724.
4.3.1. Cell Cultures. Cell Lines. A431, MDA-MB-231,
DU145, MDA-MB-435, B16F10, CacO2, and HuH7 cells
were obtained from the American Type Culture Collection.
The C6 cells were a gift from M. Mersel (Institut des
neurosciences de Montpellier, INM). These cell lines were
cultured in Dulbecco's modified Eagle's medium (DMEM)
ASSOCIATED CONTENT
■
S
* Supporting Information
All spectra and analyses. This material is available free of charge
via the Internet at http://pubs.acs.org.
supplemented with 10% fetal calf serum (FCS), 2 mM L-
glutamine, 1 mM sodium pyruvate, and 50 U mL−1
streptomycin.
4.3.2. Astrocytes. Astrocytes were isolated from newborn mice (day
AUTHOR INFORMATION
2) cortex as previously described.35
■
Corresponding Author
4.4. Drug Addition. TMZ, carmustine, vincristine, paclitaxel,
cytosine arabinoside (AraC), MTT, neutral red, phosphate-buffered
saline (PBS), triton X 100, Hoechst 33342, and PI were from Sigma.
Phostine solutions were prepared in ethanol/DMSO (80/20 v/v). For
TMZ and paclitaxel, solutions at 100 mM were prepared in 100%
DMSO. Carmustine was at 100 mM in ethanol. Vincristine (100 mM)
was dissolved, and AraC (10 μM) was prepared in PBS.
*Tel: +33 467 147 243. Fax: +33 467 144 319. E-mail: jean-luc.
pirat@enscm.fr (J.-L.P.). Tel: +33 607 106 844. Fax: +33 499
636 020. E-mail: norbert.bakalara@enscm.fr (N.B.).
Author Contributions
∥L.C. and C.J. equally contributed to the paper.
4.5.1. Measurements of the Oxaphosphinane Effects on C6
and Astrocytes in Primary Culture. MTT Assay. The MTT
assay was carried out as previously described36 48 h after drug
addition.
Notes
The authors declare no competing financial interest.
4.5.2. Neutral Red Assay. The neutral red assay was carried out as
previously described36 24 and 48 h after drug addition.37
4.5.3. LDH Assay. LDH release was measured using the LDH kit
from Roche for 60 min and according to the manufacturer's
instructions.
4.5.4. NCI Assay. The NCI assays were carried out at 10 μM
performing a sulforhodamine test at 24 h.
ACKNOWLEDGMENTS
■
We thank Didier Petite, Clovis Rakotoarivelo, Marcel Mersel,
Omar Munoz, Marissa Teigell, and Alina Florentine Cristian for
expert technical support, Romain Filhol and Julien Cau for
Cellomic expertise, Nicolas Gauthier from Roche for
Xcelligence system expertise, and Jan de Weille for carefully
reading the manuscript and for English language. This research
was supported by l'Agence Nationale de la Recherche (ANR-
07-PCVI-0038-01), L'INCa (DM/FC/mc/25-03-09), l'Institut
4.5.5. Hoechst 33342 and Propidium Iodide Fluorescence
Measurements and Impedance Measurement. After 48 h of OXP
treatment, 5 μg/mL Hoechst 33342 and 10 μM PI were added in the
culture medium for 30 min at 37 °C. Cells were then washed in PBS
and fixed during 10 min with 4% paraformaldehyde. The mean
fluorescence at four positions in each well at 460 nm was measured
(Cytofluor multiwell plate reader 4000, Perseptive biosystems) with
excitation at 360 nm and a gain of 90 dB. The mean fluorescence of
three wells per condition was then calculated.
Dual labeling with PI and Hoechst 33342 was performed using the
Cellomics array scan vti (Thermofisher). Fluorescence was detected
with an ORCA camera (ORCA ER;0.63), a 20× objective, and a XF 93
filter, gain of 25 dB, and exposure times were 24 and of 280 ms for
Hoechst 33342 and PI, respectively. Images were analyzed using the
array scan vti 6.0 software. The object detection was fixed on the
Hoechst 33342 channel with a threshold intensity of 400. The object
area was fixed between 10 and 160 to separate nuclei. The object
counting validation was fixed at 2500/well counting less than 15 fields
containing a minimum of five objects. Control fluorescence intensities
for red and blue staining were measured on Triton X-100 (2%)-treated
cells. The mean percentage of PI incorporation was calculated from
three different wells per condition.
National de la Sante
and L'Ecole Nationale Super
(ENSCM).
́
et de la Recherche Med
́
icale (INSERM),
́
ieure de Chimie de Montpellier
DEDICATION
■
This paper is dedicated in honor of Prof. Francisco Palacios on
the occasion of his 60th birthday.
ABBREVIATIONS USED
■
NCI, National Cancer Institute; EC50, half maximal effective
concentration; PCV, procarbazine, lomustine, and vincristine;
MRI, magnetic resonance imaging; MTT, dimethylthiazolyl-
2,5-diphenyltetrazolium bromide; LTC, lower toxic concen-
tration; GP, growth percent; CNS, central nervous system;
AraC, cytabarine; GFAP, glial fibrillary acidic protein; PI,
propidium iodide; TMZ, Temozolomide; LDH, lactate
dehydrogenase; CH2Cl2, dichloromethane; HPLC, high-
performance liquid chromatography; INM, Institut des neuro-
sciences de Montpellier; DMEM, Dulbecco's modified Eagle's
medium; FCS, fetal calf serum; DMSO, dimethylsulfoxide
4.5.6. Impedance Measurement. The Xcelligence system, kindly
lent by Roche diagnostics (Meylan, France), measured the electrical
impedance across interdigitated microelectrodes integrated on the
bottom of tissue culture E-Plates (ACEA biosciences). Impedance was
recorded for 69 h using RTCA software every 10 min. The effects of
2209
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
(16) McDonald, F. E.; Zhu, H. Y. H.; Holmquist, C. R. Rhodium-
catalysed alkaline cyclotrimerization strategies for 6-arylglycoside
synthesis. J. Am. Chem. Soc. 1995, 117, 6605−6606.
(17) Harvey, T. C.; Simiand, C.; Weiler, L.; Withers, S. G. Synthesis
of Cyclic Phosphonate Analogs of Ribose and arabinose. J. Org. Chem.
1997, 62, 6722−6725.
(18) Hanessian, S.; Rogel, O. Synthesis of a phostone glycomimetic
of the endothelin converting enzyme inhibitor phosphoramidon.
Bioorg. Med. Chem. Lett. 1999, 9, 2441−2446.
(19) Stoianova, D. S.; Whitehead, A.; Hanson, P. R. P-Heterocyclic
Building Blocks: Application to the Stereoselective Synthesis of P-
Sugars. J. Org. Chem. 2005, 70, 5880−5889.
(20) Pirat, J.-L.; Virieux, D.; Volle, J.-N.; Clarion, L.; Bakalara, N.;
Mersel, M.; Monbrun, J.; Cristau, H.-J. New phosphorus containing
heterocyclic compounds, sugar analogs, and compositions having
anticancer activity containing the same. PCT Int. Appl. WO 2009/
004096 A1, 2009.
REFERENCES
■
(1) Bauchet, L.; Mathieu-Daude, H.; Fabbro-Peray, P.; Rigau, V.;
Fabbro, M.; Chinot, O.; Pallusseau, L.; Carnin, C.; Laine, K.; Schlama,
A.; Thiebaut, A.; Patru, M. C.; Bauchet, F.; Lionnet, M.; Wager, M.;
Faillot, T.; Taillandier, L.; Figarella-Branger, D.; Capelle, L.; Loiseau,
H.; Frappaz, D.; Campello, C.; Kerr, C.; Duffau, H.; Reme-Saumon,
M.; Tretarre, B.; Daures, J. P.; Henin, D.; Labrousse, F.; Menei, P.;
Honnorat, J. Oncological patterns of care and outcome for 952
patients with newly diagnosed glioblastoma in 2004. Neuro-Oncology
2010, 12, 725−735.
(2) Van Meir, E. G.; Hadjipanayis, C. G.; Norden, A. D.; Shu, H. K.;
Wen, P. Y.; Olson, J. J. Exciting new advances in neuro-oncology: The
avenue to a cure for malignant glioma. CA Cancer J. Clin. 2010, 60,
166−193.
(3) Gerstner, E. R.; Chen, P. J.; Wen, P. Y.; Jain, R. K.; Batchelor, T.
T.; Sorensen, G. Infiltrative patterns of glioblastoma spread detected
via diffusion MRI after treatment with cediranib. Neuro-Oncology 2010,
12, 466−472.
(4) Fraum, T. J.; Kreisl, T. N.; Sul, J.; Fine, H. A.; Iwamoto, F. M.
Ischemic stroke and intracranial hemorrhage in glioma patients on
antiangiogenic therapy. J. Neurooncol. 2011, 105, 287−289.
(5) Uhm, J. H.; Ballman, K. V.; Wu, W.; Giannini, C.; Krauss, J. C.;
Buckner, J. C.; James, C. D.; Scheithauer, B. W.; Behrens, R. J.; Flynn,
P. J.; Schaefer, P. L.; Dakhill, S. R.; Jaeckle, K. A. Phase II Evaluation of
Gefitinib in Patients With Newly Diagnosed Grade 4 Astrocytoma:
Mayo/North Central Cancer Treatment Group Study N0074. Int. J.
Radiat. Oncol. Biol. Phys. 2011, 80, 347−353.
(6) Neyns, B.; Sadones, J.; Joosens, E.; Bouttens, F.; Verbeke, L.;
Baurain, J. F.; D'Hondt, L.; Strauven, T.; Chaskis, C.; In't Veld, P.;
Michotte, A.; De Greve, J. Stratified phase II trial of cetuximab in
patients with recurrent high-grade glioma. Ann. Oncol. 2009, 20,
1596−1603.
(7) Brada, M.; Stenning, S.; Gabe, R.; Thompson, L. C.; Levy, D.;
Rampling, R.; Erridge, S.; Saran, F.; Gattamaneni, R.; Hopkins, K.;
Beall, S.; Collins, V. P.; Lee, S. M. Temozolomide versus procarbazine,
lomustine, and vincristine in recurrent high-grade glioma. J. Clin.
Oncol. 2010, 28, 4601−4608.
(8) Westphal, M.; Hilt, D. C.; Bortey, E.; Delavault, P.; Olivares, R.;
Warnke, P. C.; Whittle, I. R.; Jaaskelainen, J.; Ram, Z. A phase 3 trial of
local chemotherapy with biodegradable carmustine (BCNU) wafers
(Gliadel wafers) in patients with primary malignant glioma. Neuro-
Oncology 2003, 5, 79−88.
(9) Sabbah, P.; Foehrenbach, H.; Dutertre, G.; Nioche, C.;
DeDreuille, O.; Bellegou, N.; Mangin, J. F.; Leveque, C.; Faillot, T.;
Gaillard, J. F.; Desgeorges, M.; Cordoliani, Y. S. Multimodal anatomic,
functional, and metabolic brain imaging for tumor resection. Clin.
Imaging 2002, 26, 6−12.
(21) Volle, J.-N.; Filippini, D.; Midrier, C.; Virieux, D.; Pirat, J.-L.;
Sobecki, M.; Drag, M. Revisited Synthesis of Aryl-H-phosphinates.
Synthesis 2011, 2490−2494.
(22) Cristau, H.-J.; Monbrun, J.; Schleiss, J.; Virieux, D.; Pirat, J.-L.
First synthesis of P-aryl-phosphinosugars, organophosphorus ana-
logues of C-arylglycosides. Tetrahedron Lett. 2005, 46, 3741−3744.
(23) Filippini, D.; Loiseau, S.; Bakalara, N.; Dziuganowska, Z. A.; Van
der Lee, A.; Volle, J.-N.; Virieux, D.; Pirat, J. L. Dramatic effect of
modified boranes in diastereoselective reduction of chiral cyclic α-
ketophosphinates. RSC Adv. 2012, 2, 816−818.
(24) Cipolla, L.; Lay, L.; Nicotra, F.; Pangrazio, C.; Panza, L.
Synthesis of Azasugars by Grignard Reaction on Glycosylamines.
Tetrahedron 1995, 51, 4679−4690.
(25) Behr, J. M.; Guillerm, G. Synthesis of 6-deoxy-homoDMDP and
its C(5)-epimer: Absolute stereochemistry of natural products from
Hyacinthus orientalis. Tetrahedron: Asymmetry 2002, 13, 111−113.
(26) Stupp, R.; Mason, W. P.; Van den Bent, M. J.; Weller, M.;
Fisher, B.; Taphoorn, M. J.; Belanger, K.; Brandes, A. A.; Marosi, C.;
Bogdahn, U.; Curschmann, J.; Janzer, R. C.; Ludwin, S. K.; Gorlia, T.;
Allgeier, A.; Lacombe, D.; Cairncross, J. G.; Eisenhauer, E.;
Mirimanoff, R. O. Radiotherapy plus concomitant and adjuvant
Temozolomide for glioblastoma. N. Engl. J. Med. 2005, 352, 987−996.
(27) Koziara, J. M.; Lockman, P. R.; Allen, D. D.; Mumper, R. J.
Paclitaxel nanoparticles for the potential treatment of brain tumors. J.
Controlled Release 2004, 99, 259−269.
(28) Acton, E. M.; Narayanan, V. L.; Risbood, P. A.; Shoemaker, R.
H.; Vistica, D. T.; Boyd, M. R. Anticancer specificity of some
ellipticinium salts against human brain tumors in vitro. J. Med. Chem.
1994, 37, 2185−2189.
(29) Scherf, U.; Ross, D. T.; Waltham, M.; Smith, L. H.; Lee, J. K.;
Tanabe, L.; Kohn, K. W.; Reinhold, W. C.; Myers, T. G.; Andrews, D.
T.; Scudiero, D. A.; Eisen, M. B.; Sausville, E. A.; Pommier, Y.;
Botstein, D.; Brown, P. O.; Weinstein, J. N. A gene expression
database for the molecular pharmacology of cancer. Nat. Genet. 2000,
24, 236−244.
(30) Ross, D. T.; Scherf, U.; Eisen, M. B.; Perou, C. M.; Rees, C.;
Spellman, P.; Iyer, V.; Jeffrey, S. S.; Van de Rijn, M.; Waltham, M.;
Pergamenschikov, A.; Lee, J. C.; Lashkari, D.; Shalon, D.; Myers, T. G.;
Weinstein, J. N.; Botstein, D.; Brown, P. O. Systematic variation in
gene expression patterns in human cancer cell lines. Nat. Genet. 2000,
24, 227−235.
(31) Blower, P. E.; Chung, J. H.; Verducci, J. S.; Lin, S.; Park, J. K.;
Dai, Z.; Liu, C. G.; Schmittgen, T. D.; Reinhold, W. C.; Croce, C. M.;
Weinstein, J. N.; Sadee, W. MicroRNAs modulate the chemosensitivity
of tumor cells. Mol. Cancer Ther. 2008, 7, 1−9.
(32) Blower, P. E.; Verducci, J. S.; Lin, S.; Zhou, J.; Chung, J. H.; Dai,
Z.; Liu, C. G.; Reinhold, W.; Lorenzi, P. L.; Kaldjian, E. P.; Croce, C.
M.; Weinstein, J. N.; Sadee, W. MicroRNA expression profiles for the
NCI-60 cancer cell panel. Mol. Cancer Ther. 2007, 6, 1483−1491.
(33) Lee, J. S.; Paull, K.; Alvarez, M.; Hose, C.; Monks, A.; Grever,
M.; Fojo, A. T.; Bates, S. E. Rhodamine efflux patterns predict P-
(10) Stummer, W.; Pichlmeier, U.; Meinel, T.; Wiestler, O. D.;
Zanella, F.; Reulen, H. J. Fluorescence-guided surgery with 5-
aminolevulinic acid for resection of malignant glioma: a randomised
controlled multicentre phase III trial. Lancet Oncol. 2006, 7, 392−401.
(11) Volle, J.-N.; Filippini, D.; Krawczyk, B.; Kaloyanov, N.; Van der
Lee, A.; Maurice, T.; Pirat, J.-L.; Virieux, D. Drug discovery:
Phosphinolactone, in vivo bioisostere of the lactol group. Org. Biomol.
Chem. 2010, 8, 1438−1444.
(12) Chemical properties and reactions of phosphine chalcogenides.
In The Chemistry of Organophosphorus Compounds; Hartley, F. R., Ed.;
John Wiley & Sons: Chichester, 1992; Vol. 2, pp 287−408.
́
(13) Hernandez, J.; Ramos, R.; Sastre, N.; Meza, R.; Hommer, H.;
Salas, M.; Gordillo, B. Conformational analysis of six-membered ring
dioxaphosphinanes. Part 1: Anancomeric thiophosphates. Tetrahedron
2004, 60, 10927−10941.
(14) Wadsworth, S.; Emmons, W. D. Bicyclic phosphites. J. Am.
Chem. Soc. 1962, 84, 610−617.
(15) Volle, J.-N.; Virieux, D.; Starck, M.; Monbrun, J.; Clarion, L.;
Pirat, J.-L. Chiral phosphinyl analogues of 2-C-arylmorpholinols: 2-
Aryl-3,5-diphenyl-[1,4,2]-oxazaphosphinanes. Tetrahedron: Asymmetry
2006, 17, 1402−1408.
2210
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211

Journal of Medicinal Chemistry
Article
glycoprotein substrates in the National Cancer Institute drug screen.
Mol. Pharmacol. 1994, 46, 627−638.
(34) Rutka, J. T.; Giblin, J. R.; Dougherty, D. Y.; Liu, H. C.;
McCulloch, J. R.; Bell, C. W.; Stern, R. S.; Wilson, C. B.; Rosenblum,
M. L. Establishment and characterization of five cell lines derived from
human malignant gliomas. Acta Neuropathol. 1987, 75, 92−103.
(35) Audouy, S.; Mallet, J.; Privat, A.; Gimenez y Ribotta, M.
Adenovirus-mediated suicide gene therapy in an in vitro model of
reactive gliosis. Glia 1999, 25, 293−303.
(36) Mosmann, T. Rapid colorimetric assay for cellular growth and
survival: Application to proliferation and cytotoxicity assays. J.
Immunol. Methods 1983, 65, 55−63.
(37) Repetto, G.; del Peso, A.; Zurita, J. L. Neutral red uptake assay
for the estimation of cell viability/cytotoxicity. Nat. Protoc. 2008, 3,
1125−1131.
2211
dx.doi.org/10.1021/jm201428a | J. Med. Chem. 2012, 55, 2196−2211