FimH Antagonists: Phosphate Prodrugs Improve Oral Bioavailability

Article abstract of DOI:10.1021/acs.jmedchem.5b01923

The widespread occurrence of urinary tract infections has resulted in frequent antibiotic treatment, contributing to the emergence of antimicrobial resistance. Alternative approaches are therefore required. In the initial step of colonization, FimH, a lectin located at the tip of bacterial type 1 pili, interacts with mannosylated glycoproteins on the urothelial mucosa. This initial pathogen/host interaction is efficiently antagonized by biaryl α-d-mannopyranosides. However, their poor physicochemical properties, primarily resulting from low aqueous solubility, limit their suitability as oral treatment option. Herein, we report the syntheses and pharmacokinetic evaluation of phosphate prodrugs, which show an improved aqueous solubility of up to 140-fold. In a Caco-2 cell model, supersaturated solutions of the active principle were generated through hydrolysis of the phosphate esters by brush border-associated enzymes, leading to a high concentration gradient across the cell monolayer. As a result, the in vivo application of phosphate prodrugs led to a substantially increased C_max and prolonged availability of FimH antagonists in urine.

Full text of DOI:10.1021/acs.jmedchem.5b01923

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
FimH Antagonists - Phosphate Prodrugs Improve Oral Bioavailability
Simon Kleeb, Xiaohua Jiang, Priska Frei, Anja Sigl, Jacqueline
Bezencon, Karen Bamberger, Oliver Schwardt, and Beat Ernst
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01923 • Publication Date (Web): 09 Mar 2016
Downloaded from http://pubs.acs.org on March 9, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
FimH Antagonists - Phosphate Prodrugs Improve Oral Bioavailability
#
#
#
Simon Kleeb, Xiaohua Jiang, Priska Frei, Anja Sigl, Jacqueline Bezençon, Karen
*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bamberger, Oliver Schwardt, Beat Ernst
Institute of Molecular Pharmacy, Pharmacenter, University of Basel, Klingelbergstrasse 50,
4056 Basel (Switzerland)
*
To whom correspondence should be addressed: Prof. Dr. B. Ernst, Institute of Molecular
Pharmacy, Pharmacenter, University of Basel, Klingelbergstrasse 50, 4056 Basel,
Switzerland; Tel: +41 61 267 15 51, Fax: +41 61 267 15 52; Eꢀmail:
beat.ernst@unibas.ch
Keywords: aqueous solubility, FimH antagonists, oral bioavailability, phosphate prodrugs.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The widespread occurrence of urinary tract infections has resulted in frequent antibiotic
treatment, contributing to the emergence of antimicrobial resistance. Alternative approaches
are therefore required. In the initial step of colonization, FimH, a lectin located at the tip of
bacterial type 1 pili, interacts with mannosylated glycoproteins on the urothelial mucosa. This
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
initial pathogen/host interaction is efficiently antagonized by biaryl αꢀDꢀmannopyranosides.
However, their poor physicochemical properties, primarily resulting from low aqueous
solubility, limit their suitability as oral treatment option. Herein, we report the syntheses and
pharmacokinetic evaluation of phosphate prodrugs, which show an improved aqueous
solubility of up to 140ꢀfold. In a Cacoꢀ2 cell model, supersaturated solutions of the active
principle were generated through hydrolysis of the phosphate esters by brush borderꢀ
associated enzymes, leading to a high concentration gradient across the cell monolayer. As a
result, the in vivo application of phosphate prodrugs led to a substantially increased Cmax and
prolonged availability of FimH antagonists in urine.
2
ACS Paragon Plus Environment

Page 3 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Urinary tract infections (UTIs) are the most frequent cause of bacteriosis in humans, affecting
1
millions of people worldwide. They remain one of the most common indications for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
prescribing antibiotics to alleviate symptoms (dysuria, frequent and urgent urination,
2
bacteriuria, and pyuria) and to prevent complications (pyelonephritis and urosepsis).
However, the frequent and repeated use of antibiotics can induce antimicrobial resistance;
3
therefore, alternative prevention and treatment strategies are urgently needed.
Uropathogenic Escherichia coli (UPEC) strains are the causative agent of more than 70% of
4
,5
all UTI episodes. The initial step in pathogenesis involves bacterial adherence to the bladder
cell surface, preventing UPEC from being cleared by the bulk flow of urine and enabling the
6
bacteria to colonize the urothelial cells. Among the different adhesins expressed on the
bacterial surface, mannoseꢀbinding type 1 pili are the most prevalent. They consist of a helical
rod containing 500 to 3000 copies of the structural subunit FimA, as well as one copy of the
7
FimF, FimG, and FimH subunits. The FimH subunit is expressed at the distal tip of each
pilus and exhibits the carbohydrate recognition domain (CRD), which interacts with the
8
mannosylated glycoprotein uroplakin 1a on the mucosal surface of the bladder.
More than three decades ago, Sharon and coꢀworkers described various oligomannosides and
9,10
aryl αꢀ
D
ꢀmannosides as potential antagonists of FimHꢀmediated bacterial adhesion.
This
led to subsequent reports on several monovalent mannoseꢀbased FimH antagonists containing
11
12
13
various aglycones, such as nꢀalkyl,
phenyl,
dioxocyclobutenylaminophenyl,
12
14ꢀ19
20
21
22
umbelliferyl, biphenyl,
indol(in)ylphenyl, triazolyl or thiazolylamino. In addition,
23ꢀ29
different multivalent presentations of mannose derivatives have been explored.
Most
importantly, it was recently shown that these αꢀDꢀmannopyranosides did not elicit adverse
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
side effects caused by the nonꢀselective binding of FimH antagonists to various mammalian
3
0
mannose receptors.
In vivo studies in a mouse disease model confirmed the therapeutic potential of biaryl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
5ꢀ17,19
mannosides for an oral treatment of UTI.
However, since only low oral bioavailability
could be achieved, basic determinants such as aqueous solubility and membrane permeability
31
should be further optimized. One possible solution is offered by a phosphate prodrug
approach, which is applied either when the active principal exhibits high membrane
32
permeability but suffers from low aqueous solubility, or when the therapeutic dose exceeds
3
3
the maximum amount of drug that can be dissolved into intestinal fluids. In a phosphate
prodrug approach, the active principle is rapidly released by endogenous phosphatases, such
as alkaline phosphatase, an enzyme particularly abundant on the brush border of
34
enterocytes.
The application of this prodrug principle has led to various marketed drugs, such as
35
34,36
prednisolone phosphate (1),
the antiretroviral drug fosamprenavir (2)
or the
37
chemotherapeutic drug fludarabine phosphate.
-O O-
-O O-
P
O
P
O
O
O
O
H
O
H
HO
N
N
O
S
H
O
O
O
H
H
2
H N
O
Prednisolone phosphate (1)
Fosamprenavir (2)
Figure 1. Phosphate prodrugs of marketed drugs.
The goal of the present study was to optimize the physicochemical profile of the biaryl
mannosides 3ꢀ5 by phosphorylation, enhancing aqueous solubility and, consequently, the oral
availability.
4
ACS Paragon Plus Environment

Page 5 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Results and Discussion
Although exhibiting nanomolar affinity and high permeability, one drawback of FimH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
4,18,19,20
antagonists 3ꢀ5
is their low aqueous solubility, resulting in limited oral bioavailability.
They are therefore perfect candidates for a phosphate prodrug approach. Table 1 summarizes
14,18,19,20
their previously reported binding affinities (IC )
as well as their physicochemical
50
properties (solubility, lipophilicity, and permeability).
Table 1. Binding affinities and physicochemical properties of the biaryl αꢀ
Dꢀmannopyranosides 3ꢀ5. IC50 values,
1
8ꢀ20
aqueous solubility, log P, PAMPA and Cacoꢀ2 cell data were adopted from references,
the IC50 of antagonist
1
8
4
and Cacoꢀ2 Papp values of antagonist 5 were obtained according to the procedure described in.
OH
OH
O
[e]
ꢀ6
Caco-2 Papp [10 cm/s]
[
a]
[b]
[c]
Cpd
HO
HO
IC50
[
Solubility
log P
PAMPA
[d]
nM]
[µg/mL]
log P
e
[cm/s]
R
a→b
b→a
O
14,18
3
84.9
21
2.1
ꢀ4.7
10.0 ± 0.9
19.0 ± 1.2
Cl
O
19
4
10.1
20
192
24
2.1
1.9
ꢀ5.2
ꢀ5.5
2.2 ± 0.4
2.9 ± 0.6
22.1 ± 1.5
39.3 ± 5.8
CN
O
2
0
NO
2
5
N
3
8
[
a] IC50 values were determined in a cellꢀfree competitive binding assay; [b] Thermodynamic solubility for
39
compounds 5; kinetic solubility for compound 3 and 4; [c] Octanolꢀwater partition coefficients (log P) were
4
0
determined by a miniaturized shakeꢀflask procedure; [d] P = effective permeability: passive permeation
e
through an artificial membrane was determined by the parallel artificial membrane permeability assay
41,42
(PAMPA);
[e] Papp = apparent permeability: permeation through a Cacoꢀ2 cell monolayer was assessed in the
43,44
absorptive (a→b) and secretory (b→a) directions in triplicate.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
For identifying the optimal position of the phosphate promoiety on the mannoside core of
FimH antagonists 3ꢀ5, a series of phosphate esters was synthesized (Figure 2) and their
solubility determined. In the prodrugs 6aꢀd and 7aꢀd, the phosphate ester bond was directly
linked to the various hydroxyl groups of the mannose moiety. Alternatively, an acetal linker
was used in prodrugs 6e and 8 to increase the distance between the enzymatic cleavage site
and thereby enhance accessibility of the phosphate ester and subsequently the
dephosphorylation rate. When prodrugs 6e and 8 are dephosphorylated, the intermediate
hemiacetals is expected to collapse spontaneously releasing the active principle 3 or 5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
45
accompanied by formaldehyde.
O
O
P
ONH
ONH
4
OR6
OR
OR6
OR
O
OH
4
2
2
4
O
4
O
O
R O
R O
HO
HO
3
3
R O
R O
Cl
O
O
O
NO
2
N
CN
6
6
6
6
6
a: R2 = PO (NH ) ; R , R , R = H
3
4
6
7a: R2 = PO (NH ) ; R , R , R = H
3
4
6
8
3
4 2
3
4 2
b: R3 = PO (NH ) ; R , R , R = H
2
4
6
7b: R3 = PO (NH ) ; R , R , R = H
2
4
6
3
4 2
3 4 2
c: R4 = PO (NH ) ; R , R , R = H
2
3
6
7c: R4 = PO (NH ) ; R , R , R = H
2
3
6
3
4 2
3 4 2
d: R6 = PO (NH ) ; R , R , R = H
2
3
4
7d: R6 = PO (NH ) ; R , R , R = H
2
3
4
3
4 2
3 4 2
e: R2 = CH OPO (NH ) ; R , R , R = H
2
3
4
2
3
4 2
Figure 2. Phosphate monoesters 6a-e of biphenyl αꢀ
Dꢀmannopyranoside 3, 7aꢀd of substituted biphenyl αꢀ
Dꢀ
mannopyranoside 4, and 8 from indolinylphenyl αꢀ ꢀmannopyranoside 5.
D
Synthesis of Phosphates 6a-d. 2ꢀPhosphate 6a of biphenyl αꢀDꢀmannopyranoside (3) was
14,18
synthesized according to the procedure depicted in Scheme 1. Starting from 3,
a
benzylidene acetal (→ 9) was formed to protect the 4ꢀ and 6ꢀOH of the mannose moiety. The
3ꢀposition was subsequently protected by a regioselective dibutyltin oxideꢀmediated
benzylation (→ 10), and then phosphorylation using dibenzyl N,Nꢀdiisopropylꢀ
phosphoramidite in the presence of 1,2,4ꢀtriazole, followed by oxidation with tertꢀ
6
ACS Paragon Plus Environment

Page 7 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
butylhydroperoxide, afforded the protected intermediate 11. Global deprotection via catalytic
hydrogenation yielded 2ꢀphosphate 6a.
OH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
O
Ph
O
OH
O
HO
HO
O
RO
a)
O
O
3
9: R = H
b)
1
0: R = Bn
O
P
O
OBn
OBn
OH
P
ONH
ONH
4
Ph
O
O
O
O
HO
HO
O
4
O
BzO
c)
d)
O
O
11
6a
2 2
Scheme 1. a) PhCH(OMe) , pꢀTsOH, DMF, 50 °C, overnight (70%); b) i. Bu SnO, toluene, 135 °C, 3 h; ii.
BnBr, toluene, 115 °C, overnight (80%); c) dibenzyl N,Nꢀdiisopropylphosphoramidite, 1,2,4ꢀtriazole, MeCN, 0
°C to rt, overnight; then 70% aq. tertꢀBuOOH, rt, 1 h (62%); d) i. H (4 bar), Pd(OH) /C, EtOAc, cat. HOAc,
overnight; ii. 25% aq. NH /MeOH (4:1), rt, overnight (45%).
2
2
3
For the synthesis of 3ꢀphosphate 6b, 3ꢀ position of 3 was regioselectively benzylated (→ 12)
followed by perbenzoylation to give 13 (Scheme 2). Cleavage of the benzyl group by
hydrogenation (→ 14) and subsequent phosphorylation afforded an inseparable 3:2ꢀmixture of
the protected 3ꢀ and 2ꢀphosphates 15 & 16 due to partial migration of the 2ꢀbenzoyl moiety.
Upon deprotection via catalytic hydrogenation pure 3ꢀphosphate 6b could be isolated.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
OH
OH
OR
OR
OBz
OBz
O
O
RO
BnO
O
BzO
HO
HO
HO
a)
c)
d)
O
O
O
3
12: R = H
14
b)
13: R = Bz
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
OBz
OBz
OBz
OH
P
OBn
OBn
O
OH
BzO
BnO
O
BzO
BzO
O
e)
HO
P
O
+
O
H NO
O
4
P
BnO
H NO
O
O
4
O
O
O
15
16
6b
Scheme 2. a) i. Bu
pyr, rt, overnight (99%); c) H
N,Nꢀdiisopropylphosphoramidite (90%), 1,2,4ꢀtriazole, MeCN, 0 °C to rt, overnight; then 70% aq. tertꢀBuOOH,
rt, 1 h (80%, 3:2ꢀmixture of 2ꢀand 3ꢀphosphate derivatives); e) i. H (1 bar), Pd(OH) /C, EtOAc, 5 h; ii. 25% aq.
NH /MeOH (4:1), rt, overnight (7%).
2
SnO, MeOH, reflux, 5 h, ii. BnBr, toluene, 115 °C, overnight (36%); b) BzCl, cat. DMAP,
2
(4 bar), Pd(OH) /C, dioxane/EtOAc, cat. AcOH, rt, overnight (73%); d) dibenzyl
2
2
2
3
For the synthesis of 4ꢀphosphate 6c, 9 was benzoylated (→ 17). Reductive opening of the
benzylidene acetal with Me N·BH and AlCl3 afforded precursor 18 (Scheme 3).
3
3
Phosphorylation (→ 19) and subsequent deprotection yielded 4ꢀphosphate 6c.
OBn
Ph
O
O
HO
OH
O
Ph
O
O
BzO
OBz
O
OBz
O
HO
BzO
a)
b)
O
O
O
9
17
18
O
OBn
OBz
O
P
OH
OH
BnO
BnO
H NO
4
P
O
BzO
O
O
O
4
H NO
c)
d)
HO
O
O
1
9
6c
3 3 3 2
Scheme 3. a) BzCl, cat. DMAP, pyr, rt, overnight (60%); b) Me N·BH , AlCl , THF/H O, rt, 1 h (67%); c)
dibenzyl N,Nꢀdiisopropylphosphoramidite (90%), 1,2,4ꢀtriazole, MeCN, 0 °C to rt, overnight; then 70% aq. tertꢀ
8
ACS Paragon Plus Environment

Page 9 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
BuOOH, 1 h (53%); d) H
56%).
2 2 3
, Pd(OH) /C, EtOH/EtOAc, 5 h, quant; e) 25% aq. NH /MeOH (4:1), rt, overnight
(
For the synthesis of 6ꢀphosphate 6d mannoside 3 was tritylated in the 6ꢀposition, followed by
perbenzoylation (→ 20) and removal of the trityl group (Scheme 4). Then, intermediate 21
was phosphorylated (→ 22) and final global deprotection gave 6ꢀphosphate 6d.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
OH
OTr
OBz
OH
OBz
HO
HO
O
BzO
BzO
O
BzO
BzO
O
a)
b)
O
O
O
3
20
21
O
O
O
O
BnO
BnO
H
H
4
NO
P
P
4
NO
OBz
O
OH
O
BzO
BzO
HO
HO
c)
d)
O
O
22
6d
Scheme 4. a) i. TrCl, cat. DMAP, pyr, 80 °C, overnight, ii. BzCl, 50 °C, overnight (88%); b) FeCl
rt, 5 h (62%); c) dibenzyl N,Nꢀdiisopropylphosphoramidite (90%), 1,2,4ꢀtriazole, MeCN, 0 °C to rt, overnight;
then 70% aq. tertꢀBuOOH, rt, 1 h (66%), d) i. H , Pd(OH) /C, EtOH/EtOAc, overnight, ii. 25% aq. NH /MeOH
4:1), rt, overnight (55%).
3 2
/H O, DCM,
2
2
3
(
Synthesis of Phosphates 7a-d. Due to the labile chloroꢀ and cyanoꢀsubstituents present in
19
biphenyl αꢀDꢀmannopyranoside 4, the 2ꢀphosphate 7a was obtained via a modified strategy
(Scheme 5), omitting a potentially intractable hydrogenation step. Therefore, after protecting
the 4ꢀ and 6ꢀOH of 4 with a benzylidene acetal (→ 23), the 3ꢀOH of the mannose moiety was
46
selectively benzoylated to afford 24. Phosphorylation of the 2ꢀOH group with bis[2ꢀ
trimethylsilyl)ethyl] N,Nꢀdiisopropylphosphoramidite in the presence of 1,2,4ꢀtriazole, and
(
subsequent oxidation with tertꢀbutylhydroperoxide, yielded intermediate 25. After cleavage of
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the (trimethylsilyl)ethyl esters with TFA (→ 26) and subsequent ester hydrolysis upon
treatment with NH
3
/MeOH, 2ꢀphosphate 7a was obtained.
SiMe
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
P
OH
OH
SiMe3
O
Ph
O
OH
O
Ph
O
O
BzO
O
O
O
O
HO
HO
O
RO
Cl
Cl
Cl
a)
c)
O
O
O
4
23: R = H
25
b)
24: R = Bz
CN
CN
CN
O
P
O
P
OH
OH
ONa
ONa
ONa
ONa
O
O
HO
BzO
O
HO
HO
O
Cl
Cl
d)
e)
O
O
2
6
7a
CN
CN
Scheme 5. a) PhCH(OMe)
trimethylsilyl)ethyl] N,Nꢀdiisopropylphosphoramidite, 1,2,4ꢀtriazole, MeCN, 0 °C to rt, 16 h; then 70% aq. tertꢀ
BuOOH, rt, 1 h (55%); d) TFA/DCM (1:4), rt, 2 h (61%); e) 25% aq. NH /MeOH (4:1), rt, 16 h (71%).
2
, pꢀTsOH, rt, 17 h (22%); b) BzCl, DCM/pyr, 0 °C to rt, 3 h (60%); c) bis[2ꢀ
(
3
Regioselective 3ꢀallylation of 4 (→ 27) followed by benzoylation of the 2ꢀ, 4ꢀ and 6ꢀOH gave
8 (Scheme 6). Subsequent cleavage of the allyl group with PdCl (→ 29) and
phosphorylation afforded intermediate 30 and final deprotection 3ꢀphosphate 7b.
2
2
1
0
ACS Paragon Plus Environment

Page 11 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
OH
OH
OR
OR
OBz
OBz
O
HO
HO
O
RO
AllO
O
BzO
HO
a)
c)
Cl
Cl
Cl
O
O
O
4
27: R = H
29
b)
28: R = Bz
CN
CN
CN
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OBz
OBz
OH
OH
BzO
O
HO
P
O
3
Me Si
d)
O
O
e)
NaO
NaO
O
Cl
Cl
P
O
O
O
O
O
30
7b
3
Me Si
CN
CN
Scheme 6. a) i. Bu
pyr, rt, overnight (87%); c) PdCl
phosphoramidite, 1,2,4ꢀtriazole, MeCN, 0 °C to rt, 15 h; then 70% aq. tertꢀBuOOH, rt, 2 h (32%); e) i.
TFA/DCM (1:4), rt, 1.5 h; ii. 25% aq. NH /MeOH (4:1), rt, overnight (45%).
2
SnO, toluene, 80 °C, 6 h; ii. AllBr, Bu
4
NI, toluene, 80 °C, 20 h (55%); b) BzCl, cat. DMAP,
2
, MeOH, 40 °C, 5 h (84%); d) i. bis[2ꢀ(trimethylsilyl)ethyl] N,Nꢀdiisopropylꢀ
3
4
ꢀPhosphate 7c was synthesized in three steps via regioselective dibutyltin oxideꢀmediated
47
acetylation of 4 in the 2ꢀ, 3ꢀ and 6ꢀposition (→ 31), phosphorylation of the 4ꢀOH (→ 32)
and subsequent cleavage of all protecting groups (Scheme 7).
OH
OH
OAc
OAc
HO
HO
O
HO
AcO
O
a)
b)
Cl
Cl
O
O
4
31
CN
CN
O
P
OAc
OAc
O
P
OH
OH
3
Me Si
O
NaO
NaO
O
AcO
O
O
HO
O
O
c)
Cl
Cl
3
Me Si
O
O
32
7c
CN
CN
Scheme 7. a) i. Bu SnO, MeOH, 70 °C, 2 h; ii. Ac O, MeCN, rt, 16 h (21%); b) i. bisꢀ[2ꢀ(trimethylsilyl)ethyl]
2
2
N,Nꢀdiisopropylphosphoramidite, 1,2,4ꢀtriazole, MeCN, 0 °C to rt, 15 h; then 70% aq. tertꢀBuOOH, rt, 2 h
62%); c) i. TFA/DCM (1:4), rt, 2 h; ii. 25% aq. NH /MeOH (4:1), rt, 2 h (98%).
(
3
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
For the synthesis of 6ꢀphosphate 7d a similar protection strategy as for 6d was used (Scheme
8
). The parent compound 4 was tritylated in the 6ꢀposition (→ 33), followed by
perbenzoylation (→ 34) and removal of the trityl group. Phosphorylation of intermediate 35
→ 36) and final global deprotection afforded 6ꢀphosphate 7d.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
OH
OH
OTr
OR
OH
OBz
HO
HO
O
RO
RO
O
BzO
BzO
O
a)
c)
Cl
Cl
Cl
O
O
O
4
35
3
3: R = H
b)
34: R = Bz
CN
CN
CN
Me
Si
3
Si
O
O
O
O
O
O
NaO
NaO
Me
3
P
P
OBz
O
OH
O
BzO
BzO
HO
HO
d)
e)
Cl
Cl
O
O
36
7d
CN
CN
Scheme 8. a) TrCl, cat. DMAP, pyr, 80 °C, 16 h (80%); b) BzCl, cat. DMAP, pyr, rt, 6 h (79%); c) FeCl
3
2
, H O,
rt, 5 h (85%); d) i. bisꢀ[2ꢀ(trimethylsilyl)ethyl] N,Nꢀdiisopropylphosphoramidite, 1,2,4ꢀtriazole, MeCN, 0 °C to
3
rt, 15 h; then 70% aq. tertꢀBuOOH, rt, 1.5h (59%); e) i. TFA/DCM (1:4), rt, 2 h; ii. 25% aq. NH /MeOH (4:1),
rt, overnight (68%).
Synthesis of Acetal-linked Phosphates 6e & 8. Synthesis of the acetalꢀlinked phosphate 6e of
biphenyl αꢀ ꢀmannopyranoside (3) was achieved by first introducing a 6ꢀOꢀ(thiomethyl)ꢀ
methyl group on intermediate 21 with DMSOꢀacetic anhydride under acidic conditions (→
7, Scheme 9). Subsequent treatment with phosphoric acid and Nꢀiodosuccinimide gave
phosphate 38. Finally, debenzoylation with NH /MeOH provided the target compound 6e.
D
3
3
1
2
ACS Paragon Plus Environment

Page 13 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 9. a) DMSO, Ac
2 3 4
O/AcOH, rt, overnight (35%); b) H PO , NIS, THF, 0 °C to rt, 1 h (58%); c) 25% aq.
NH /MeOH (4:1), rt, overnight (50%).
3
The acetalꢀlinked phosphate 8 was prepared analogously from indolinylphenyl αꢀDꢀ
20
mannopyranoside 5 (Scheme 10). Selective TBSꢀprotection of the 6ꢀOH, followed by perꢀ
benzoylation (→ 39), and cleavage of the silyl ether yielded precursor 40. After the
introduction of a 6ꢀOꢀ(thiomethyl)methyl group with DMSOꢀacetic anhydride (→ 41),
treatment with phosphoric acid and Nꢀiodosuccinimide (→ 42) and debenzoylation, test
compound 8 was obtained.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
OH
OH
OTBS
OBz
HO
HO
O
BzO
BzO
O
a)
O
O
NO
2
NO2
5
39
N
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
S
OH
O
OBz
O
OBz
O
BzO
BzO
BzO
BzO
b)
c)
O
O
NO
2
NO
2
40
41
N
N
O
O
O
P
O
O
P
OH
ONH
ONH4
4
O
OH
O
OBz
O
OH
BzO
BzO
HO
HO
O
d)
e)
O
NO
2
NO
2
4
2
8
N
N
Scheme 10. a) i. TBSCl, cat. DMAP, pyr, rt, overnight, ii. BzCl, rt, 2 h, (quant.); b) 1 M H
73%); c) DMSO/Ac O/HOAc, rt, overnight (74%); d) H PO /NIS/THF, 0 °C to rt, 1 h (67%); e) 25% aq.
NH /MeOH/DCM, rt, overnight (41%).
2 4
SO /MeOH, rt, 1.5 h
(
2
3
4
3
Solubility. The thermodynamic solubility of the phosphate prodrugs was determined in
phosphate buffer (50 mM, pH 6.5). The expected improvement of aqueous solubility, which
exceeds those of the active principles 3ꢀ5 by several orders of magnitude (Table 2), could be
confirmed.
ALP-Mediated Hydrolysis. ALPꢀmediated hydrolysis of the various phosphate esters was
studied in Cacoꢀ2 cells, which express phosphatase on the apical brush border surface of the
48
confluent cell monolayer. The experimental halfꢀlife (t1/2) was calculated from the
concentration of remaining prodrug vs. incubation time (Table 2).
1
4
ACS Paragon Plus Environment

Page 15 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Aqueous solubility and Cacoꢀ2 phosphataseꢀmediated hydrolysis (t1/2) of prodrugs 6aꢀe, 7aꢀd, 8, and
their active principles 3ꢀ5, respectively.
Cpd
Solubility [µg/mL]
21
t
1/2 [min]
ꢀꢀꢀ
1
4,18
3
, active principle
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
a
>3000
>3000
2703
12
6b
13
6c
> 60
> 60
8.7
ꢀꢀꢀ
6d
>3000
>3000
192
6e
19
4
, active principle
7a
>3000
>3000
>3000
>3000
24
13
7b
12
7
c
43
7d
48
2
0
5
, active principle
ꢀꢀꢀ
8
>3000
11
a) 100
b) 100
6
a
6b
6c
7a
7b
7c
7d
10
10
1
6
d
6
e
1
8
0.1
0.1
0
20
40
60
0
20
40
60
Time [min]
Time [min]
Figure 3. Decomposition of phosphomonoester prodrugs 6a-d and 7a-d and phosphonooxymethyl ether
prodrugs 6e and 8 in the apical compartment of the Cacoꢀ2 cell assay: (a) 6aꢀe, 8; (b) 7a-d. Prodrugs dissolved in
Dulbecco’s Modified Eagle’s Medium (62.5 µM) were applied to the apical chamber and the concentrations of
unchanged prodrug were monitored by LCꢀMS.
Depending on the position of the promoiety, the prodrugs showed varying propensity to
dephosphorylation (Figure 3). The 2ꢀ and 3ꢀphosphate esters (6a, 6b, 7a and 7b) were rapidly
hydrolyzed (t1/2 < 15 min), whereas the 4ꢀ and 6ꢀphosphate esters (6c, 6d, 7c and 7d) showed
prolonged halfꢀlives (t1/2 > 40 min). Improved stability of the latter likely results from reduced
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
5,49,50
access to the ester bonds at C4 and C6 because of steric hindrance.
Therefore, with the
introduction of a linker (→ 6e, 8), accessibility can be improved and the susceptibility to
ALPꢀmediated cleavage was markedly increased (t = 8.7 min and 11 min, respectively).
1
/2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Owing to their high propensity to ALPꢀmediated hydrolysis, the phosphate esters 6a, 6b, 7a,
b and 8 were almost entirely converted to parent drug within 60 min (Figure 3). For
7
example, Figure 4 depicts the concentration of prodrug 7b 60 min after application to either
the apical or the basolateral side of the Cacoꢀ2 system. When applied to the apical chamber,
the prodrug was almost quantitatively hydrolyzed. However, when applied to the basal
chamber, the prodrug remained the prominent species detected, due to the lack of ALP on the
48
basal Cacoꢀ2 cell membrane. In addition, irrespective of dosing on the apical or basal side,
the prodrug could not be detected in the receiver compartment, which corroborates the poor
membrane permeability of the polar phosphate ester (Figure 4).
Figure 4. Conversion of prodrug 7b to the active principle 4 in a Cacoꢀ2 cell monolayer model after 60 min of
incubation. A prodrug solution (62.5 µM) was applied either into the apical or basal chamber. Columns represent
the percentage concentrations of prodrug and active principle 60 min after initiation of the experiment.
Concentration of prodrug 7b determined at time point t = 0 min is defined as 100%.
1
6
ACS Paragon Plus Environment

Page 17 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Stability in biorelevant media. Since chemical stability at various pH conditions and stability
51,52
for degradation by digestive enzymes (simulated gastric and intestinal fluids,
for
composition see Table S1 in Supporting Information) turned out to be high (> 80% after 2 h;
Figure 5), the phosphate prodrugs 7a, 7c and 7d appear optimally suited to mitigate the
solubility problem.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Stability of the phosphate prodrugs 7a (black column), 7c (grey column), and 7d (white column) in
biorelevant media. Percentage of the remaining compound concentration relative to the initial concentration (t =
0 min) after 120 min of incubation in a) simulated gastric fluids (FaSSGF, fastedꢀstate simulated gastric fluid;
FeSSGF, fedꢀstate simulated gastric fluid; sGF, simulated gastric fluid; buffer sGF, prepared equally to sGF but
without pepsin) and b) simulated intestinal fluids (FaSSIF, fastedꢀstate simulated intestinal fluid; FeSSIF, fedꢀ
state simulated intestinal fluid; sIF, simulated intestinal fluid; buffer sIF, prepared equally to sIF but without
pancreatin) are shown. For composition see Table S1 in Supporting Information. H O was used as a reference
2
media. Data represent the mean (triplicates) with its corresponding standard deviation.
Oral Bioavailability. For oral administration of a phosphate prodrug several absorption rateꢀ
limiting factors need to be considered. In addition to unsuitably slow ALPꢀmediated
hydrolysis of the phosphate prodrug, low solubility, poor permeability, and efflux of the
active principle can also limit oral absorption.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In Vitro Pharmacokinetic Evaluation. To further examine the benefits associated with
improved solubility obtained upon phosphorylation of the active principle, the most labile
phosphate prodrugs 6e, 7b, and 8 (t1/2 ≤ 12 min, Table 2) were applied to the apical chambers
of the Cacoꢀ2 system, and the accumulation of active principles 3, 4, and 5 on the basal side
of the cell monolayer was monitored (Figures 6aꢀc). Applying either the active principle or
the respective prodrug at equal concentrations (62.5 µM) to the apical chamber resulted in
similar basolateral concentrations of the active principle, i.e. the hydrolysis is not the rate
limiting step. When higher apical doses of the phosphate prodrugs were applied, the
basolateral concentrations were markedly increased (Figure 6aꢀc).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Once a prodrug has been hydrolyzed in vivo, physicochemical properties of the active
principle (i.e. solubility and permeability) are becoming the rateꢀlimiting steps for absorption.
When after rapid hydrolysis (t1/2 < 15 min, Table 2) the active principle is precipitating due to
low solubility, the available amount for absorption is reduced. Although the active principles
3 and 5 show similar solubilities (Table 1), different basolateral concentrations were observed
(Figure 6a & 6c). In contrast, the basolateral concentrations of compound 3 were similar or
higher compared to those of compound 4 (Figure 6a & 6b), even though compound 4 has an
[
39]
8
ꢀfold higher solubility (Table 1). Based on permeability data derived from PAMPA, high
[42]
e
passive permeability is predicted for all three antagonists (log P < ꢀ5.7 cm/s, Table 1).
However, for the antagonists 4 and 5, efflux ratios (bꢀa/aꢀb) > 2 (Table 1) in biꢀdirectional
Cacoꢀ2 experiments were observed, indicating the involvement of an efflux transporter (e.g.
[44]
Pꢀgp). Due to the high apical concentrations reached from fast ALPꢀmediated hydrolysis of
prodrugs 7b and 8, saturation of the transporter mediated efflux occurred, leading to increased
basolateral concentrations of 4 and 5 (Figure 6b & 6c). In contrast, 3 exhibited an efflux ratio
<
2 (Table 1), i.e. the basolateral concentration rose proportionally to the amount of applied
[53]
prodrug and did not display saturation kinetics (Figure 6a).
1
8
ACS Paragon Plus Environment

Page 19 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Accumulation of active principle (a) 3, (b) 4, and (c) 5 in the basal receiver chamber of a Cacoꢀ2 cell
system 60 min after applying (a) active principle 3 or prodrug 6e, (b) active principle 4 or prodrug 7b, and (c)
active principle 5 or prodrug 8 into the apical chamber. The active principles were dosed at a concentration of
62.5 µM, which corresponds to approx. the aqueous solubilities of 3 and 5. The phosphate prodrugs were dosed
at four different concentrations (62.5 µM, 100 µM, 200 µM, and 400 µM).
In vivo pharmacokinetic studies. In the mouse PK model, an increased intestinal uptake was
anticipated for the prodrug 7c due to its increased solubility and slower hydrolysis rate. The
prodrugs 7b, 7c and 8 were administered per os at a dose of 10 mg/kg, but could not be
detected in plasma or urine samples. The oral bioavailability of active principle 4 upon
administering prodrug 7b and 7c and of active principle 5 upon administration of prodrug 8
are illustrated in Figure 7. Table 3 summarizes the pharmacokinetic parameters (Cmax, Tmax
,
and urine AUC0ꢀ24) of the different p.o. applications.
Table 3. Pharmacokinetic parameters (Cmax, Tmax, and urine AUC0ꢀ24) determined after a single p.o. application
a
of compounds 4, 7b, 7c, and 8, in female C3H/HeN mice.
7b
7c
8
Parameter
Dose
4
(
prodrug of 4)
(prodrug of 4)
(prodrug of 5)
1.25
7.7
10
10
10
(mg/kg) p.o.
C
max (µg/mL)
3.9 ± 1.9
23.6 ± 3.1
40.7 ± 11.4
57.0 ± 4.2
4.5 ± 1.6
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
T
max (h)
1.4 ± 0.6
1.5 ± 5.2
1.4 ± 0.2
0.9 ± 0.2
0.44 ± 0.2
3.0 ± 0.0
Urine AUC0ꢀ24
2
106.8 ± 18.0
103.9 ± 7.7
226.4* ± 42.8
24.6 ± 10.7
(µg x h/mL)
a
[54]
Values were calculated using PKsolver and are shown as mean with standard deviation.
concentration; Tmax, time were Cmax is reached; AUC, area under the curve. *Statistical differences at p < 0.001
TwoꢀWay ANOVA, Prism; GraphPad Software, Version 5.0f), compared to all other p.o. applications.
Cmax, maximal
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Given the almost identical propensity to ALPꢀmediated hydrolysis of 7b and 8, the different
urine profiles must be related to physicochemical properties of their active principles 4 and 5.
Indeed, antagonist 5, besides having an 8ꢀtimes lower solubility and therefore a higher
propensity to precipitate after release in the small intestines, also permeates biological
membranes less easily (shown in the PAMPA results, Table 1). In fact, urine AUC0ꢀ24 and
C
max of 5 were similar to the pharmacokinetic parameters of 4 when applied at a dose of 1.25
mg/kg (Table 3).
Figure 7. Urine concentrations of active principle 4 in C3H/HeN mice upon p.o. administration of 4 (1.25 mg/kg
and 7.7 mg/kg p.o.) or the related phosphate prodrugs 7b, 7c (10 mg/kg p.o.) and the urine concentration of the
active principle 5 upon p.o. administration of phosphate prodrug 8 (10 mg/kg p.o.). Shown are mean values with
standard deviations for groups of three (4, 1.25 mg/kg, 7b, 7c, and 8) or five (4, 7.7 mg/kg) mice.
2
0
ACS Paragon Plus Environment

Page 21 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Since prodrugs 7b and 8 are rapidly hydrolyzed in vitro, we also assessed the impact of
slower ALPꢀmediated bioactivation and availability of the related active principle in urine.
For this purpose, we administered prodrug 7c (in vitro t1/2 of 43 min, Table 2) at a dose of 10
mg/kg (Figure 7). After applying 7c, compound 4 reached Cmax (57 µg/mL, Table 3) after
only 26 min, as compared to 40.7 µg/mL at 54 min for 7b; it then remained at elevated levels
for the next 3 h. Within the observation period of 3ꢀ24 h postꢀadministration, urine levels of
the two prodrugs dropped steadily. However, at 6 h the concentration of active principle 4
originating from prodrug 7c was approx. three times higher than the one reached with 7b,
resulting in a 2ꢀfold increased urine AUC0ꢀ24 of 7c.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The lower concentration of active principle 4 upon p.o. application of 7b as compared to 7c
can be rationalized by several pharmacokinetic effects. First, the fast dephosphorylation of 7b
leads to a local accumulation of 4 at the brush border in the small intestine. Due to insufficient
solubility, the active principle 4 could suffer from precipitation. Because of the slower
cleavage of the promoiety in prodrug 7c, the solubility limit of 4 takes longer to reach,
enabling an improved absorption of the active principle. Moreover, although high
concentrations of 4 saturate Pꢀgp, its efflux activity still contributes to accumulated
concentrations of 4 and consequently to precipitation and intestinal elimination. A slower
prodrug hydrolysis apparently optimizes the timely interplay between parent drug uptake and
Pꢀgpꢀmediated efflux, leading to higher net absorption and urine concentrations.
Next, we addressed the question of whether increased solubility indeed leads to higher
availability of active principle in the urine. Therefore, urine concentration profiles upon p.o.
application of 1.25 mg/kg (based on maximal solubility) or 7.7 mg/kg of 4 (applied as a
suspension corresponding to 10 mg/kg of prodrug) were determined (Figure 7).
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Oral application of the suspension of 4 resulted in a urine AUC0ꢀ24 of 106.8 µg x h/mL (Cmax
of 23.6 µg/mL, Table 3) as compared to a urine AUC0ꢀ24 of 226.4 µg x h/mL (Cmax of 57
µg/mL, Table 3) achieved with the corresponding prodrug 7c, demonstrating the beneficial
effect of increased solubility on intestinal uptake. In addition, the application of 4 at two
doses (1.25 mg/kg and 7.7 mg/kg; 6.16 times greater) resulted in parallel curves with a similar
Tmax and an approximate difference in urine AUC0ꢀ24 of a factor of 5 and Cmax of a factor of 6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Table 3). This is a useful observation for future doseꢀfinding in patients.
Summary and Conclusions
For the successful oral application of a phosphate prodrug of an active principle that displays
moderate to high membrane permeability, but insufficient solubility, several prerequisites
need to be fulfilled. Obviously, the solubility of the prodrug should allow to dissolve the
required dose (e.g. in our case 10 mg/kg). Second, because phosphate prodrugs are too polar
to permeate the enterocyte layer, an efficient release of the active principle is required.
Finally, slow enzymatic hydrolysis is preferred to avoid precipitation of the poorly soluble
active principle.
We have demonstrated the advantages of the phosphate prodrug approach as applied to FimH
antagonists with an insufficient solubility but high passive permeability. An increase in
solubility of up to 140ꢀfold could be reached upon phosphorylation of the active principle.
Furthermore, either fast or slow hydrolysis was observed, depending on the position of the
phosphate promoiety on the mannose ring. When the phosphate ester bond was directly linked
at the C2ꢀ or C3ꢀposition of mannose (→ 6a, 6b, 7a, 7b), or when an acetal linker at C6ꢀ was
used (→ 6e & 8), enzymatic cleavage was fast (t1/2 < 15 min). In contrast, a phosphate at the
2
2
ACS Paragon Plus Environment

Page 23 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
C4ꢀ or C6ꢀposition (→ 6c, 6d, 7c & 7d) showed an enhanced enzymatic stability (t1/2 > 40
min).
Interestingly, even when the rates of hydrolysis were similar, e.g. for prodrugs 7b and 8,
different physicochemical properties (solubility and permeability) of their active principles
influenced the in vivo PK properties. For antagonist 5, which has an 8ꢀfold lower solubility as
compared to 4, the risk of precipitation from a supersaturated solution in the small intestines
needs to be taken into account.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Furthermore, a high concentration gradient across the Cacoꢀ2 cell monolayer, as was reached
with the more soluble phosphate prodrugs, promotes absorptive flux of the active principles 3,
4
, and 5, and apparently saturates the efflux carrier activity of 4 and 5. This observation was
confirmed in an in vivo PK study in mice, where urine AUC0ꢀ24 of the active principle 4 could
be doubled when prodrug 7c was applied instead of active principle 4. Moreover, in vivo
administration of slowly hydrolyzed prodrug 7c (t1/2 < 40 min) exhibited an increase in urine
AUC0ꢀ24 as compared to a phosphate prodrug with fast enzymatic cleavage (7b, t1/2 < 15 min).
Slower conversion to the active principle prolonged intestinal uptake and renal excretion by
improving the interplay between solubility, drug uptake, and saturation of Pꢀgp mediated
efflux.
Experimental Section
Synthesis
General Methods. NMR spectra were recorded on a Bruker Avance DMXꢀ500 (500.1 MHz)
1
13
spectrometer. Assignment of H and C NMR spectra was achieved using 2D methods
COSY, HSQC). Chemical shifts are expressed in ppm using residual CHCl or MeOH as
(
3
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
references. Optical rotations were measured with a PerkinElmer Polarimeter 341. Electrospray
ionization mass spectrometry (ESIꢀMS) data were obtained on a Waters Micromass ZQ
instrument. Microwaveꢀassisted reactions were carried out with a CEM Discover and
Explorer. Reactions were monitored by TLC using glass plates coated with silica gel 60 F254
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Merck) and visualized by UV light and/or by charring with a molybdate solution (0.02 M
solution of ammonium cerium sulfate dihydrate and ammonium molybdate tetrahydrate in
aqueous 10% H SO ). Medium pressure chromatography (MPLC) separations were carried
2
4
f
out on a CombiFlash Companion or R from Teledyne Isco equipped with RediSep normalꢀ
phase or RPꢀ18 reversedꢀphase flash columns. Commercially available reagents were
purchased from Fluka, Aldrich, or Alfa Aesar (Germany). Solvents were purchased from
Sigma–Aldrich (Buchs, Switzerland) or Acros Organics (Geel, Belgium) and were dried prior
to use where indicated. MeOH was dried by reflux with sodium methoxide and distilled and
stored under argon atmosphere. Dichlormethane (DCM) and acetonitrile (MeCN) were dried
2 3
by filtration over Al O (Fluka, type 5016 A basic) and stored over molecular sieves under
argon. Molecular sieves (4 Å) were activated in vacuo at 300 °C for 0.5 h before use.
Compound purity. Each test compound was purified by chromatography on silica
2
(DCM/MeOH) or reversedꢀphase chromatography (RPꢀ18, H O/MeOH) prior to HPLC,
HRMS, NMR and activity testing. The purity of all test compounds was determined by NMR
and HPLC [Method A: Beckman Coulter Gold, consisting of pump 126, DAD 168 (190ꢀ410
2
nm) and autoꢀsampler 508; column: Waters Atlantis T3, 3 ꢀm, 2.1 × 100 mm; A: H O + 0.1%
TFA; B: MeCN + 0.1% TFA; gradient: 5% B for 0.5 min, 5% B → 70% B over 19.5 min;
flow rate: 0.5 mL/min. Method B: Agilent 1100/1200 with UV detection (190ꢀ410 nm);
2
column: Waters Atlantis T3, 3 ꢀm, 2.1 × 100 mm; A: H O + 0.01% TFA; B: MeCN + 0.01%
TFA; gradient: 5% B for 1 min, 5% B → 95% B over 19 min; flow rate: 0.5 mL/min.
2
4
ACS Paragon Plus Environment

Page 25 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
Detection: 254 nm] to be ≥ 95% (for H NMR spectra and HPLC traces see Supporting
Information).
General procedure for phosphorylation. To an iceꢀcooled solution (0 °C) of protected
mannoside (1 eq) and 1,2,4ꢀtriazole (4 eq) in dry MeCN was added dibenzyl N,Nꢀ
diisopropylphosphoramidite or bis[2ꢀ(trimethylsilyl)ethyl] N,Nꢀdiisopropylphosphoramidite
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
2 eq) and the mixture was stirred for 30 min at 0 °C and then overnight at rt. Then, 70% aq.
tertꢀbutylhydroperoxid (4 eq) was added and the solution was stirred for 1 h. The reaction was
quenched with 1 M aq. Na and 1 M aq. NaHCO and the mixture was extracted twice
with DCM. The combined organic layers were dried over Na SO , filtered and the solvents
2
S
2
O
3
3
2
4
removed in vacuo. The residue was purified by MPLC on silica gel (petroleum ether/EtOAc)
to yield the phosphorylated compounds.
Biphenyl 2-O-phosphoryl-α-
D-mannopyranoside diammonium salt (6a). Hydrogenolysis
of compound 11 (100 mg, 0.129 mmol) was conducted in a Parr shaker with 10% Pd(OH) /C
2
(
12 mg) and a catalytic amount of HOAc in EtOAc (6.0 mL) under hydrogen (4 bar) at rt
overnight. Then, the reaction suspension was filtered through celite and the filtrate was
concentrated in vacuo. The residue was stirred in 25% aq. NH (4 mL) and MeOH (1 mL)
3
overnight. Then, the solvents were removed under reduced pressure and the residue was
purified by MPLC on silica (DCM/MeOH/H O, 6:4:0.6) to give 6a (26.0 mg, 45%) as a white
2
2
D
0
1
solid. [α]
6), 3.87 (t, J = 10.0 Hz, 1H, Hꢀ4), 4.14 (ddd, J = 2.0, 3.0, 10.0 Hz, 1H, Hꢀ3), 4.62 (ddd, J =
.0, 3.0, 8.5 Hz, 1H, Hꢀ2), 5.89 (d, J = 1.5 Hz, 1H, Hꢀ1), 7.31 (m, 2H, ArꢀH), 7.44 (m, 1H,
2 2
+66.7 (c 0.12, H O); H NMR (D O, 500 MHz): δ = 3.77ꢀ3.82 (m, 3H, Hꢀ5, Hꢀ
2
1
3
ArꢀH), 7.55 (t, J = 7.5 Hz, 2H, ArꢀH), 7.71 (m, 4H, ArꢀH); C NMR (D
2
O, 126 MHz): δ =
6
0.5 (Cꢀ6), 66.6 (Cꢀ4), 70.1 (d, J = 5 Hz, Cꢀ3), 73.5 (Cꢀ5), 73.6 (d, J = 5 Hz, Cꢀ2), 97.0 (d, J =
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3 Hz, Cꢀ1), 117.2, 117.6, 126.7, 127.4, 128.0, 128.2, 129.1, 135.3, 140.0, 155.0 (12C, ArꢀC);
ꢀ
ESIꢀMS: m/z: Calcd for C H O P [Mꢀ2NH +H] : 411.08, found: 411.06.
18
20
9
4
Biphenyl 3-O-phosphoryl-α-D-mannopyranoside diammonium salt (6b). Hydrogenolysis
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of a 3:2ꢀmixture of 15 and 16 (130 mg, 0.144 mmol) in EtOAc (6 mL) was conducted in the
presence of 10% Pd(OH) /C (15 mg) with a hydrogen balloon at rt for 5 h. Then, the reaction
2
suspension was filtered through celite and the filtrate was concentrated in vacuo. The residue
was stirred in 25% aq. NH
removed under reduced pressure and the residue was purified by MPLC on silica gel
(DCM/MeOH/H O, 6:4:0.6) to provide 6b (4.5 mg, 7%, still containing some of the 2ꢀ
phosphate impurity) as white solid. [α]
3.75ꢀ3.83 (m, 3H, Hꢀ5, Hꢀ6), 3.92 (t, J = 9.5 Hz, 1H, Hꢀ4), 4.39 (m, 1H, Hꢀ2), 4.55 (dt, J =
3.0, 9.0 Hz, 1H, Hꢀ3), 5.72 (s, 1H, Hꢀ1), 7.32 (m, 2H, ArꢀH), 7.45 (m, 1H, ArꢀH), 7.54 (t, J =
3
(4 mL) and MeOH (1 mL) at rt overnight. The solvents were
2
20
1
D
2 2
+91.7 (c 0.12, D O); H NMR (D O, 500 MHz): δ
=
1
3
7
.5 Hz, 2H, ArꢀH), 7.71 (m, 4H, ArꢀH); C NMR (D
2
O, 126 MHz): δ = 61.4 (Cꢀ6), 66.9 (d, J
=
3 Hz, Cꢀ4), 70.0 (d, J = 3 Hz, Cꢀ2), 74.0 (Cꢀ5), 75.0 (d, J = 5 Hz, Cꢀ3), 98.5 (Cꢀ1), 118.1,
118.3, 127.31, 127.34, 128.0, 128.8, 129.8, 135.8, 140.6, 155.7 (12C, ArꢀC); ESIꢀMS: m/z:
ꢀ
Calcd for C18
H
20
O
9
P [Mꢀ2NH
4
+H] : 411.08, found: 411.12.
Biphenyl 4-O-phosphoryl-α-D-mannopyranoside diammonium salt (6c). Hydrogenolysis
of 19 (80 mg, 90 µmol) was done in EtOAc (4 mL) in the presence of 10% Pd(OH) /C (12 mg)
2
with hydrogen balloon at rt overnight. Then the reaction suspension was filtered through
celite, the filtrate was concentrated in vacuo to provide the debenzylated intermediate. The
3
crude intermediate (52 mg) was stirred in MeOH (1 mL) and 25% aq. NH (4 mL) overnight.
The solvent was removed under reduced pressure and the residue was purified by MPLC on
2
D
0
silica (DCM/MeOH/H
2
O, 4:1:0.1) to yield 6c (18 mg, 56%) as a white solid. [α]
+104.8 (c
1
0
.20, H O); H NMR (D O, 500 MHz):
δ
= 3.75ꢀ3.87 (m, 3H, Hꢀ5, Hꢀ6), 4.24 (m, 3H, Hꢀ2,
2
2
2
6
ACS Paragon Plus Environment

Page 27 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hꢀ3, Hꢀ4), 5.70 (s, 1H, Hꢀ1), 7.30 (d, J = 8.5 Hz, 2H, ArꢀH), 7.43 (t, J = 7.5 Hz, 1H, ArꢀH),
13
7
.54 (t, J = 8.0 Hz, 2H, ArꢀH), 7.70 (d, J = 8.0 Hz, 4H, ArꢀH); C NMR (D
2
O, 126 MHz): δ
=
60.6 (Cꢀ6), 69.6 (Cꢀ2), 69.7 (d, J = 5 Hz, Cꢀ4), 70.7 (Cꢀ3), 72.6 (d, J = 7 Hz, Cꢀ5), 97.9 (Cꢀ
1), 117.6, 126.7, 127.4, 128.2, 129.1, 135.3, 140.0, 155.1 (12C, ArꢀC); ESIꢀMS: m/z: Calcd
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
for C18
H
20
O
9
P [Mꢀ2NH
4
+H] : 411.08, found: 411.11.
Biphenyl 6-O-phosphoryl-α-
of 22 (64.0 mg, 70.0 µmol) was done in EtOAc/EtOH (5 mL, 3:2) in the presence of 10%
Pd(OH) /C (7.5 mg) with hydrogen balloon at rt overnight. The reaction suspension was
D-mannopyranoside diammonium salt (6d). Hydrogenolysis
2
filtered through celite and the filtrate was concentrated in vacuo. The crude intermediate was
stirred in MeOH (1 mL) and 25% aq. NH (4 mL) overnight. The solvents were removed
3
under reduced pressure and the residue was purified by MPLC on silica (DCM/MeOH/H
2
O,
2
0
1
4:1:0.1) to yield 6d (16.0 mg, 55%) as a white solid. [α]D +74.1 (c 0.16, H O); H NMR
2
(D
2
O, 500 MHz):
9.9 Hz, 1H, Hꢀ4), 4.08 (m, 1H, Hꢀ6b), 4.11 (dd, J = 3.5, 10.0 Hz, 1H, Hꢀ3), 4.21 (dd, J =
.8, 3.4 Hz, 1H, Hꢀ2), 5.69 (d, J = 1.4 Hz, 1H, Hꢀ1), 7.28 (m, 2H, ArꢀH), 7.49 (t, J = 7.4 Hz,
δ = 3.83 (m, 1H, Hꢀ5), 3.89 (ddd, J = 1.9, 4.9, 12.0 Hz, 1H, Hꢀ6a), 4.01 (t, J
=
1
13
1H, ArꢀH), 7.54 (t, J = 7.7 Hz, 2H, ArꢀH), 7.70 (d, J = 8.7 Hz, 4H, ArꢀH); C NMR (D
2
O,
1
9
26 MHz): δ = 62.6 (J = 4 Hz, Cꢀ6), 65.8 (Cꢀ4), 70.0, 70.1 (Cꢀ2, Cꢀ3), 72.9 (d, J = 8 Hz, Cꢀ5),
8.4 (Cꢀ1), 117.5, 127.4, 126.7, 128.3, 129.1, 135.3, 140.0, 155.1 (12C, ArꢀC); ESIꢀMS: m/z:
ꢀ
Calcd for C18
H
20
O
9
P [Mꢀ2NH
4
+H] : 411.08, found: 411.09.
Biphenyl 6-O-(phosphonooxymethyl)-α-D-mannopyranoside diammonium salt (6e).
Compound 38 (110 mg, 0.146 mmol) was stirred in MeOH (1 mL) and 25% aq. NH (4 mL)
3
overnight. The solvents were removed under reduced pressure and the residue was purified by
20
D
MPLC on RPꢀ18 (H
2
O/MeOH) to yield 6e (32.3 mg, 50%) as a white solid. [α]
+75.5 (c
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 86
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2 2
0.32, H O); H NMR (D O, 500 MHz): δ = 3.85 (dd, J = 2.0, 11.5 Hz, 1H, Hꢀ6a), 3.88 (m,
1
1
1
H, Hꢀ5), 3.93 (t, J = 9.5 Hz, 1H, Hꢀ4), 4.10 (dd, J = 3.5, 9.5 Hz, 1H, Hꢀ3), 3.97 (dd, J = 4.0,
1.5 Hz, 1H, Hꢀ6b), 4.22 (dd, J = 2.0, 3.5 Hz, 1H, Hꢀ2), 4.94 (A of ABX, J = 5.5, 11.0 Hz,
2 2
H, CH ), 5.04 (B of ABX, J = 5.5, 8.5 Hz, 1H, CH ), 5.69 (d, 1.5 Hz, 1H, Hꢀ1), 7.28 (m, 2H,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13
ArꢀH), 7.44 (m, 1H, ArꢀH), 7.54 (m, 2H, ArꢀH), 7.69ꢀ7.72 (m, 4H, ArꢀH); C NMR (D
2
O,
1
26 MHz): δ = 66.1 (Cꢀ4), 66.6 (Cꢀ6), 69.9 (Cꢀ2), 70.3 (Cꢀ3), 72.2 (Cꢀ5), 90.4 (d, J = 4 Hz,
CH
2
), 98.3 (Cꢀ1), 117.5, 126.7, 127.4, 128.3, 128.8, 129.1, 132.5, 135.4, 140.0, 155.0 (12C,
ꢀ
ArꢀC); ESIꢀMS: m/z: Calcd for C19
H
22
O
10P [Mꢀ2NH
4
+H] : 441.10, found: 440.92.
3'-Chloro-4'-(2-O-phosphoryl-α-
disodium salt (7a). Compound 26 (34.7 mg, 56.0 µmol) was dissolved in MeOH (0.25 mL)
and 25% aq. NH (1 mL). The mixture was stirred for 16 h at rt. The solvents were removed
in vacuo, the residue was dissolved in H O (0.5 mL) containing a drop of 1 M aq. NaOH and
purified by MPLC on RPꢀ18 (H O/MeOH) to yield pure 2ꢀphosphate 7a (20.4 mg, 71%) as
D-mannopyranosyloxy)-biphenyl-4-carbonitrile
3
2
2
2
D
0
1
the sodium salt. [α]
Hꢀ5, Hꢀ6), 3.92 (t, J = 9.8 Hz, 1H, Hꢀ4), 4.13 (dd, J = 3.0, 9.7 Hz, 1H, Hꢀ3), 4.66 (dt, J = 2.4,
.3 Hz, 1H, Hꢀ2), 5.97 (d, J = 1.2 Hz, 1H, Hꢀ1), 7.40 (d, J = 8.7 Hz, 1H, ArꢀH), 7.45ꢀ7.53 (m,
2 2
+17.9 (c 0.78, H O); H NMR (500 MHz, D O): δ = 3.70ꢀ3.82 (m, 3H,
8
1
3
4H, ArꢀH), 7.66 (d, J = 8.4 Hz, 2H, ArꢀH); C NMR (126 MHz, D
2
O):
δ = 62.0 (Cꢀ6), 68.6
(
Cꢀ4), 72.3 (d, J = 3 Hz, Cꢀ3), 73.8 (d, J = 4 Hz, Cꢀ2), 99.2 (d, J = 5 Hz, Cꢀ1), 111.0, 118.9,
1
21.2, 125.5, 128.2, 128.4, 130.1, 134.4, 135.4, 144.7, 152.5 (13C, 12 ArꢀC, CN); IR (KBr): ν
ꢀ1
ꢀ
9
= 3436 (vs, OH), 2230 (w, CN) cm ; ESIꢀMS: m/z: Calcd for C19H18ClNO P [Mꢀ2Na+H] :
470.04, found: 469.96.
3
'-Chloro-4'-(3-O-phosphoryl-α-D-mannopyranosyloxy)-biphenyl-4-carbonitrile
disodium salt (7b). A solution of 30 (89.1 mg, 90 µmol) in dry DCM (2 mL) was treated with
TFA (500 ꢁL) for 1.5 h at rt under argon. Then, the mixture was evaporated to dryness in
2
8
ACS Paragon Plus Environment

Page 29 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
vacuo to yield the benzoylated intermediate, which was dissolved in MeOH (1 mL) and 25%
aq. NH (4 mL). The mixture was stirred at rt overnight. The solvents were removed in vacuo,
3
2
the residue was dissolved in H O (1 mL) containing two drops of 1 M aq. NaOH and purified
2
D
0
by MPLC on RPꢀ18 (H O/MeOH, 95:5ꢀ4:1) to yield 7b (20.7 mg, 45%). [α] +42.5 (c 0.82,
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
H
2
O); H NMR (500 MHz, D
2
O):
δ
= 3.74 (dd, J = 5.8, 12.4 Hz, 1H, Hꢀ6a), 3.77ꢀ3.83 (m,
2
H, Hꢀ6b, Hꢀ5), 3.91 (t, J = 9.6 Hz, 1H, Hꢀ4), 4.43 (dd, J = 1.8, 3.4 Hz, 1H, Hꢀ2), 4.55 (dt, J
3.4, 8.8 Hz, 1H, Hꢀ3), 5.76 (d, J = 1.5 Hz, 1H, Hꢀ1), 7.42 (d, J = 8.7 Hz, 1H, ArꢀH), 7.58
dd, J = 2.3, 8.7 Hz, 1H, ArꢀH), 7.71 (d, J = 8.5 Hz, 2H, ArꢀH), 7.75 (d, J = 2.3 Hz, 1H, Arꢀ
=
(
13
H), 7.80 (d, J = 8.5 Hz, 2H, ArꢀH); C NMR (126 MHz, D
2
O): δ = 62.2 (Cꢀ6), 68.3 (d, J = 3
Hz, Cꢀ4), 70.9 (d, J = 4 Hz, Cꢀ2), 74.9 (d, J = 5 Hz, Cꢀ3), 75.3 (Cꢀ5), 99.8 (Cꢀ1), 111.1, 121.2,
1
25.7, 128.2, 128.6, 130.3, 134.4, 135.9, 144.9, 152.5 (13C, 12 ArꢀC, CN); IR (KBr):
ν
=
ꢀ
1
ꢀ
3436 (vs, OH), 2230 (w, CN) cm ; ESIꢀMS: m/z: Calcd for C19
H
18ClNO
9
P [Mꢀ2NH
4
+H] :
4
70.04, found: 470.12.
3
'-Chloro-4'-(2,3,4-O-benzoyl-4-O-phosphoryl-α-D-mannopyranosyloxy)-biphenyl-4-
carbonitrile disodium salt (7c). Prepared according to the procedure described for 7b.
Compound 32 (37.3 mg, 27 µmol) was subsequently treated with TFA (300 ꢁL) in DCM (1.2
3
mL) for 2 h, and then with MeOH (0.5 mL) and 25% aq. NH (2 mL) for 2 h. Purification by
2
D
0
MPLC on RPꢀ18 (H
2
O/MeOH, 95:5ꢀ4:1) yielded 7c (23.6 mg, 98%). [α]
+70.0 (c 0.81,
1
H
2
O); H NMR (500 MHz, D
2
O): = 3.64ꢀ3.71 (m, 2H, Hꢀ6a, Hꢀ5), 3.85 (dd, J = 4.1, 12.5
δ
Hz, 1H, Hꢀ6b), 4.17ꢀ4.28 (m, 3H, Hꢀ3, Hꢀ2, Hꢀ4), 5.56 (s, 1H, Hꢀ1), 7.11 (d, J = 1.9 Hz, 1H,
ArꢀH), 7.16 (d, J = 8.7 Hz, 1H, ArꢀH), 7.19 (m, 3H, ArꢀH), 7.41 (d, J = 8.3 Hz, 2H, ArꢀH);
1
3
C NMR (126 MHz, D
2
O): δ = 60.7 (Cꢀ6), 69.2 (d, J = 5 Hz, Cꢀ4), 69.6, 70.9 (Cꢀ2, Cꢀ3),
7
3.3 (d, J = 7 Hz, Cꢀ5), 98.6 (Cꢀ1), 109.5, 117.5, 119.5, 124.1, 126.6, 128.2, 132.8, 134.0,
2
9
ACS Paragon Plus Environment