Nucleoside Diphosphate Prodrugs: Nonsymmetric DiPPro-Nucleotides

Article abstract of DOI:10.1021/acs.jmedchem.5b00737

Nonsymmetric DiPPro-nucleotides are described as nucleoside diphosphate (NDP) delivery systems. The concept is to attach different bis(acyloxybenzyl) moieties at the β-phosphate moiety of a NDP. DiPPro compounds bearing two alkanoylbenzyl residues and DiPPro compounds bearing an alkanoylbenzyl or a benzoylbenzyl group as bioreversible prodrug moieties were studied. Compounds bearing short chain alkanoyl esters led to a fast hydrolysis by chemical or enzymatic means. The ester group in the second prodrug group comprised a long lipophilic aliphatic or an aromatic residue. The lipophilicity of this group enabled the prodrug to penetrate the cell membrane. The introduction of two different groups allowed a controlled stepwise removal of the prodrug moieties to achieve a highly selective delivery of the NDP in CEM cell extracts. The compounds were highly active against HIV even in thymidine kinase-deficient CEM cells. Thus, the compounds, although charged at the α-phosphate group, were taken up by the cells and released NDPs.

Full text of DOI:10.1021/acs.jmedchem.5b00737

Page 1 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Nucleoside Diphosphate Prodrugs: Nonꢀsymmetric
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
DiPProꢀNucleotides
1
2
2
1
Lina Weinschenk , Dominique Schols , Jan Balzarini and Chris Meier *
1
Organic Chemistry, Department of Chemistry, Faculty of Sciences, University of Hamburg,
2
MartinꢀLutherꢀKingꢀPlatz 6, 20146 Hamburg, Germany and Katholieke Universiteit Leuven,
Rega Institute for Medical Research, Minderbroedersstraat 10, 3000 Leuven, Belgium.
*
Correspondence to: chris.meier@chemie.uniꢀhamburg.de
KEYWORDS antiviral agents – nucleoside analogs – pronucleotides – bioreversible protection –
drug delivery – antiꢀHIV – nucleoside diphosphate delivery
ABSTRACT Nonꢀsymmetric DiPProꢀnucleotides are described as nucleoside diphosphate
(
NDP) delivery systems. The concept is to attach different bis(acyloxybenzyl)ꢀmoieties at the βꢀ
phosphate moiety of a NDP. DiPProꢀcompounds bearing two alkanoylbenzyl residues and
DiPProꢀcompounds bearing an alkanoylbenzyl or a benzoylbenzyl group as bioreversible
prodrug moieties were studied. Compounds bearing short chain alkanoyl esters led to a fast
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 60
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
hydrolysis by chemical or enzymatic means. The ester group in the second prodrug group
comprised of a long lipophilic aliphatic or an aromatic residue. The lipophilicity of this group
enabled the prodrug to penetrate the cell membrane. The introduction of two different groups
allowed a controlled stepꢀwise removal of the prodrug moieties to achieve a highly selective
delivery of the NDP in CEM cell extracts. The compounds were highly active against HIV even
in thymidine kinaseꢀdeficient CEM cells. Thus, the compounds although charged at the αꢀ
phosphate group were upꢀtaken by the cells and released NDPs.
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
INTRODUCTION
Nucleoside analogues are extensively used as agents in antiviral and antitumor chemotherapy.
However, their biological activity is strongly dependent on their intracellular activation by host
cell kinases to give – via the monoꢀ and the diphosphate – the ultimately active triphosphate
derivative which inhibits the function of the viral DNA polymerase. However, in the case of
1,2
nucleoside analogues cellular kinases often catalyze these metabolic steps insufficiently. The
result might be a partial loss of antiviral activity or for some nucleoside analogues even a
complete failure to exhibit any antiviral activity. Moreover, inefficient phosphorylation led in
3
,4
some cases to adverse effects. The use of prodrug forms of phosphorylated nucleoside
analogues (pronucleotides) can circumvent rate limiting steps within the metabolic pathway by
5,6
bypassing at least one, if not several involved activating enzymes. This task has been
successfully achieved in the past for the intracellular delivery of monophosphates of nucleoside
7
ꢀ9
10,11
12
analogues using prodrug strategies such as the cycloSalꢀ,
Sateꢀ,
bisPOMꢀ and
13,14
phosphoramidateꢀnucleotide approaches.
Remarkably, in contrast to numerous examples of successful nucleoside monophosphate
prodrug approaches for the bypass of nucleoside kinases, the development of nucleoside
2
ACS Paragon Plus Environment

Page 3 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
diphosphate (NDP) prodrugs has been very rarely addressed. However, such a prodrug approach
might be desirable, for instance, in the case of 3’ꢀazidoꢀ3’ꢀdeoxythymidine (AZT), which is very
poorly phosphorylated from AZTMP to yield AZTDP by thymidylate kinase (TMPꢀK) in human
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
15,16
cells.
Hostetler et al. synthesized different nucleoside diphosphate diglycerides as potential NDP
1
7ꢀ20
prodrugs.
However, the hydrolysis of these compounds delivered the monophosphates
instead of the diphosphates as a result of the cleavage of the diphosphate moiety. A second
2
1ꢀ23
approach was reported by HuynhꢀDinh et al.
In their approach, the βꢀphosphate was acylated
with fatty acids forming to a mixed anhydride bond. In chemical hydrolysis studies the NDP was
formed. However, in cell extracts an undefined decomposition of the compounds was observed.
Since a few years, we turned our interest towards the bioreversible protection of nucleoside
diphosphates although, as compared to nucleoside monophosphate prodrugs, the design of such
compounds is more demanding due to a chemical reason. In NDPs, a complete lipophilic
modification of the charged and therefore highly polar diphosphate unit led to a considerable
decrease in chemical stability of this part of the molecule. As a result, the phosphate anhydride
bond of the nucleoside diphosphate was cleaved leading to two monophosphorylated fragments.
Earlier, we reported on the DiPProꢀapproach. In this approach, esterification of the βꢀ
phosphate of a NDP with two bis(acyloxybenzyl)ꢀmoieties neutralizes the charges at this
phosphate group. In previous studies this was achieved by masking the βꢀphosphate using two
2
4,25
26
identical bis(alkanoyloxybenzyl)ꢀ
or bis(benzoyloxybenzyl)ꢀmoieties
(symmetrical
1
2
modification, R = R ). The concept was based on an enzymeꢀdriven cleavage of these masking
units, which ensured that the release of NDPs preferably took place intracellularly due to the
higher concentrations of the esterases/lipases in the target cells. In contrast to the low stability in
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 60
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 presence of esterases, the chemical stability at physiological pH was still found to be high.
By choosing the appropriate substituent R in the ester moiety within the mask, it was possible to
2
5
“tune” both stability and lipophilicity of the prodrugs (Scheme 1).
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
Scheme 1. Hydrolysis of earlier reported symmetric bis(acyloxybenzyl)ꢀDiPProꢀcompounds 1.
Although we were able to prove that this approach works, an unwanted property of the
symmetric aliphatic or aromatic DiPProꢀcompounds was that also some nucleoside
monophosphate (NMP) was detected in addition to the wanted nucleoside diphosphate e.g. in
PBS solution, pH 7.3 or cell extract hydrolysis studies. Interestingly, it was shown that the
amount of NMP clearly correlated with the stability of the DiPProꢀcompounds and that the
stability of the compounds was dependent on the length of the alkyl residues R attached via an
ester linkage to the 4ꢀhydroxybenzyl group: the longer the alkyl group R, the higher the stability
of the DiPProꢀcompounds but the higher was also the amount of NMP formed in the chemical
hydrolyses and in the cell extract studies. As the reason for the formation of NMP a concurrence
4
ACS Paragon Plus Environment

Page 5 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reaction was identified in which water/hydroxide reacted at the phosphorus atom of the βꢀ
phosphate group. This led to the cleavage of the phosphate anhydride bond releasing the NMP.
This side reaction became more relevant with increasing stability of the ester linkage in the
prodrug moiety. On the other hand, we observed that after the removal of the first mask to give a
monoꢀmasked intermediate such as 2, no further increase of the amount of NMP was observed
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 1). The reason here was that after the first removal an additional charge appeared at the
βꢀphosphate and this obviously prevented a nucleophilic reaction at this moiety.
Taking all previous results together, these findings led to the development of a second
generation of DiPProꢀcompounds. Here, we disclose a study on series of non-symmetric DiPProꢀ
1
2
nucleotides 1 bearing two different masking units (R not equal to R ). The design was the
following: One masking group contained a short alkyl chain carboxylic acid ester which ꢀ due to
2
5
our previously reported data ꢀ will be rapidly cleaved by chemical or enzymatic means. The
second masking group contained a long alkyl residue carboxylic acid ester or a substituted
benzoic acid ester, which adds high lipophilicity to the molecule. It was expected that such a
construct would allow a rapid conversion of the DiPProꢀcompound 1 into the monoꢀmasked
intermediate 2 and thereby should avoid the side reaction to form the unwanted nucleoside
monophosphate. The second mask will be cleaved subsequently from the intermediate to form
the nucleoside diphosphate. It was expected that with these nonꢀsymmetric compounds a highly
selective conversion of the DiPProꢀcompounds into nucleoside diphosphates can be achieved.
RESULTS AND DISCUSSION
Synthesis
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 60
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 nonꢀsymmetric DiPProꢀNDPs 1, we first tried to use the protocol of the
2
4ꢀ26
symmetric counterparts.
According to that protocol two equivalents of an appropriate
4
ꢀacyloxybenzylalcohol were reacted with dichloroꢀN,Nꢀdiisopropylaminophosphoramidite to
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
2
give the corresponding phosphoramidite like 5 (R =R ) which was then coupled by an acidꢀ
mediated reaction with a nucleoside monophosphate followed by oxidation. For the new nonꢀ
symmetric compounds it was tried to introduce two different 4ꢀacyloxybenzyl masking units 3
and 4 by the same way. However, this approach failed.
Finally, in a new protocol 4ꢀacyloxybenzylalcohol 3 was reacted with phosphorus trichloride
followed by the addition of two equivalents of N,Nꢀdiisopropylamine (DIPA) which led to
bis(diisopropylamino)ꢀphosphoramidites 6 in yields up to 56% in a one-pot reaction. Then, one
of the diisopropylamino moieties was replaced by the second 4ꢀacyloxybenzylalcohol 4 after
activation with dicyanoimidazol (DCI) to yield the nonꢀsymmetric phosphoramidites 5 in very
high yields.
Next, a second acidꢀactivated coupling of these phosphoramidites 5 with the corresponding
nucleoside monophosphates was carried out. At the beginning, the conversions obtained in this
coupling reaction were surprisingly low, which was proven by NMRꢀ and HPLCꢀanalysis of the
crude mixtures. Changing the equivalents of phosphoramidite 5 and the DCIꢀactivator did not led
to an improvement. Finally, the problem was solved by the successive addition of small amounts
of the DCIꢀactivator solution. Starting with the addition of 0.5 equiv. DCIꢀactivator solution to a
solution of 1 equiv. of the nucleoside monophosphate and 1.5 equiv. of the phosphoramidite 5,
followed by the stepwise addition of 0.25 equiv. every 5 min up to a total of 1.25 or 1.75 equiv.
of DCI led to a quantitative conversion of the nucleotide. DiPProꢀnucleotides 7ꢀ10 were
successfully isolated by using automated flash RPꢀ18ꢀchromatography. Using the new protocol
6
ACS Paragon Plus Environment

Page 7 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
numerous combinations of DiPProꢀnucleotides were accessible in quantitative conversions, high
1
13
31
isolated yields and high purity, which was checked by means of Hꢀ, Cꢀ and PꢀNMR
spectroscopy, mass spectrometry and HPLC analysis. The route for the synthesis of nonꢀ
symmetric DiPProꢀNDPs 7ꢀ10 is summarized in Scheme 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
O
O
R1
5
O
4
O
R1
3
2
Cl
Cl
i)
Cl
P
N(iPr)2
up to 56%
1
O
P
3
6
N(iPr)2
OH
O
O
R1
R1
O
5
eO
f
R2
4
3
2
d
c
O
OH
O
1
b
a
4
iii)
0 - 85%
O
O
P
O-
N
P
O
O P
O
O
Onucleoside
ii)
up to 98%
5
O
R2
5'
2
a'
1'
R
b'
c'
f'
R1 and R2 = aliphatic or aromatic
2'
4'
e'
7-10
O
O
O
O
3
'
d'
phosphoamidites 5a-j
Scheme 2. Synthesis of nonꢀsymmetric DiPProꢀNDPs 7ꢀ10. i) a) 1 equiv. 3, 1 equiv. pyridine,
78 °C to rt, 24 h, THF. b) 6.1 equiv. N,Nꢀdiisopropylamine, ꢀ10 °C to rt, 24 to 48 h, THF. ii)
ꢀ
0
6
1
.67 equiv. 4, 0.67 equiv. DCIꢀactivator solution (0.25 molar in acetonitrile), 4 °C to rt, 30 to
0 min, acetonitrile. iii) a) 1 equiv. [N(nC H ) ]d4TMP or [N(nC H ) ]AZTMP, 1.5 equiv. 5,
4
9 4
4
9 4
.25 to 1.75 equiv. DCIꢀactivator solution (0.25 molar in acetonitrile), rt, 30 min, acetonitrile; b)
1.5 equiv. tertBuOOH solution (5.5 molar in nꢀdecane), rt, 15 min, acetonitrile. (indices shown
in compound 5,6 and 7ꢀ10 are used for NMR assignment in Experimental Section)
Structural formulae of DiPProꢀd4TDPs 7,8 and DiPProꢀAZTDPs 9,10 bearing different
2
2
aliphatic (R = C H ) and aromatic (R = XꢀPh) acyl moieties as the lipophilic masks in
n
2n+1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 60
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
combination with short aliphatic acyl moieties (R = CH , C H ) in the acyloxybenzylꢀmasking
3
4
9
group are summarized in Figure 1. Also three symmetric DiPProꢀAZTDPs 11aꢀc were prepared
by the earlier published method.
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
Figure 1. Nonꢀsymmetric DiPProꢀd4TDPs 7,8 and DiPProꢀATZDPs 9,10.
Hydrolysis Studies
All prepared compounds were studied with regard to their stabilities and their hydrolysis
products in different media.
Chemical stability in phosphate buffer, pH 7.3: We expected that the more labile mask would
be cleaved faster to form intermediate 2 bearing the long, lipophilic and hydrolytically more
stable mask. However, surprisingly in the chemical hydrolysis studies in most of the cases both
8
ACS Paragon Plus Environment

Page 9 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
intermediates were observed by means of HPLC analysis. In the monoꢀmasked intermediates 2,
the short alkanoyl ester moieties were more rapidly hydrolyzed as compared to the long alkyl
bearing moieties. Furthermore, the halfꢀlives determined in phosphate buffer, pH 7.3, of the nonꢀ
symmetric DiPProꢀcompounds were surprisingly higher as compared to those of the symmetric
DiPProꢀNDPs. This can be seen for example for compound 7b (CH /C H ꢀDiPProꢀd4TDP). In
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
3
9
19
this case the halfꢀlife of the cleavage of the first masking group bearing the acetate group was
0 h while the same masking group in the symmetrical diꢀCH ꢀDiPProꢀd4TDP was cleaved with
4
3
2
4
a halfꢀlife of 10 h only. Steric hindrance or aggregation due to the second, more lipophilic
moiety might be a reason for this increase in stability. The determined halfꢀlives of the formation
of the monoꢀmasked intermediate were between 15 h and 69 h as summarized in Table 1.
The chemical stabilities of DiPProꢀAZTDPs 9,10 were in the same range as those of the
DiPProꢀd4TDPs 7,8 having identical prodrug moieties. Only CH /C H ꢀDiPProꢀAZTDP 9c
3
11 23
showed a lower stability compared to the corresponding d4TDP derivative 7c for some unknown
reasons. Even in comparison with the DiPProꢀcompounds with shorter alkyl chains the stability
of CH /C H ꢀDiPProꢀAZTDP 9c was markedly low. Nevertheless, all DiPProꢀcompounds 7ꢀ
3
11 23
10 hydrolyzed in buffer to form almost exclusively the nucleoside diphosphate.
As reported before, some NMP was formed as well probably due to the unexpected high
hydrolytic stability compared to the symmetric DiPProꢀcompounds bearing short alkyl chains in
the masking group. It is important to mention, that the ratio of NDP to NMP in these chemically
24,25
driven hydrolyses was higher as described for the symmetric compound with longer chains.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 60
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
Table 1. Halfꢀlives of DiPProꢀd4TDPs 7,8 and DiPProꢀAZTDPs 9,10 in PBS and CEM/0 cell
extracts.
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
NDP = d4TDP 7
NDP = AZTDP 9
t1/2 [h]
t1/2 [h]
in CEM/0
0.04
t1/2 [h]
t1/2 [h]
in CEM/0
0.02
Compound
in PBS
36
in PBS
33
CH /C H ꢀDiPProꢀNDP 7a/9a
3
7
15
CH /C H ꢀDiPProꢀNDP 7b/9b
40
48
15
58
0.04
0.16
0.03
0.04
35
32
13
ꢀ
0.04
0.11
0.04
ꢀ
3
9
19
CH /C H ꢀDiPProꢀNDP 7c/9c
3
11 23
CH /CF ꢀPhꢀDiPProꢀNDP 7d/9d
3
3
CH /CH ꢀPhꢀDiPProꢀNDP 7e
3
3
NDP = d4TDP 8
NDP = AZTDP 10
C H /C H ꢀDiPProꢀNDP 8a/10a
36
40
52
15
50
0.71
1.05
1.91
0.88
0.96
46
50
63
19
ꢀ
0.83
1
4
9
7
15
C H /C H ꢀDiPProꢀNDP 8b/10b
4
9
9
19
C H /C H ꢀDiPProꢀNDP 8c/10c
1.65
0.83
ꢀ
4
9
11 23
C H /CF ꢀPhꢀDiPProꢀNDP 8d/10d
4
9
3
C H /CH ꢀPhꢀDiPProꢀNDP 8e
4
9
3
We also evaluated the chemical stability of DiPProꢀAZTDP 9b and 10b in Roswell Park
Memorial Institute (RPMI) medium in addition to PBS solution, pH 7.3 as described above. Both
compounds showed also under these conditions high stability (t1/2 = 19 h for 9b and t1/2 = 77 h
for 10b) and were found to be markedly more stable as compared to the CEM/0 cell extracts.
+
Incubations in cell extracts of wild type CD Tꢀlymphocyte cells (CEM/0 cell extracts): As in
4
the above described studies in PBS and RPMIꢀculture medium, the halfꢀlives of the DiPProꢀ
compounds were found to be independent on the nucleoside analogue attached to the molecule.
In contrast to the chemical hydrolysis studies, the prodrugs delivered the corresponding
10
ACS Paragon Plus Environment

Page 11 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
nucleoside diphosphate selectively within 2 to 7 hours only. In the case of the CH /R ꢀDiPProꢀ
3
NDPs 7 and 9 the hydrolysis led exclusively to the intermediates bearing the long lipophilic acyl
group with halfꢀlives of minutes (Table 1). In all cases the acetate ester was selectively cleaved
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
first. However, in comparison the cell extract hydrolysis of the C H /R ꢀDiPProꢀNDPs 8 and 10
4
9
released in addition to the intermediate bearing the long alkyl chain acyl group also the C H ꢀ
4
9
intermediate, although in a markedly smaller ratio. This is due to the higher stability of the
1
pentanoyl (R = C H ) ester moiety compared to the acetate group present in compounds 7 and 9.
4
9
But also in these cases the diphosphate was finally formed from both intermediates selectively.
The more lipophilic mask had no influence on the halfꢀlives of the cleavage of the short chain
1
1
acyl group (R = CH and R = C H , respectively, Table 1) except for the four DiPProꢀ
3
4
9
2
nucleotides with R = C H (7cꢀ10c) which showed different stabilities. Figure 2 summarizes
11
23
as an example the hydrolysis of CH /C H ꢀDiPProꢀd4TDP 7b in CEM/0 cell extracts followed
3
9
19
by means of RPꢀ18ꢀHPLCꢀchromatography. After 60 min virtually all of the prodrug 7b was
consumed. Only the C H ꢀintermediate 11b and d4TDP appeared as hydrolysis products at that
9
19
time point. Whether traces of d4TMP were also formed remained unclear because the detection
of d4TMP was not possible by RPꢀ18ꢀHPLCꢀchromatography due to remaining cell extract
components that eluted in the chromatogram with identical retention time. On the other hand, by
using a HILICꢀcolumn (HILIC: hydrophilic interaction liquid chromatography) in the
chromatography the separation of d4TMP from cell extract components was successful. With
this second analytical method it was proven that very small amounts of d4TMP were also present
in the CEM cell extract hydrolysis samples of the DiPProꢀcompounds.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 60
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
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
C
9
H
19-intermediate
2b
Figure 2. RPꢀ18ꢀHPLCꢀchromatograms of the hydrolysis of CH /C H ꢀDiPProꢀd4TDP 7b in
3
9
19
CEM/0 cell extracts.
In order to study if the detected d4TMP resulted either from the side reaction described above
or from a dephosphorylation of d4TDP to give d4TMP due to phosphatases present in the
extracts, both d4TDP and AZTDP were incubated as well in the CEM/0 cell extracts and aliquots
of the hydrolysis mixtures were analyzed after 2 h, 4 h, 7 h and 10 h of incubation. In both cases
d4TMP and AZTMP were detected. The halfꢀlife of dephosphorylation of d4TDP was 10 h while
that of AZTDP was 19 h in the experimental setꢀup. For this reason, we concluded that the trace
amounts of the nucleoside monophosphates found in the above described hydrolysis studies of
the DiPProꢀcompounds in CEM/0 cell extracts were most likely a result of a dephosphorylation
of the formed diphosphates from the prodrugs.
12
ACS Paragon Plus Environment

Page 13 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
In the case of the C H /R ꢀDiPProꢀnucleotides 8 and 10 the hydrolysis proceeded via both
4
9
possible intermediates and then led to the corresponding NDPs. As the halfꢀlives for the starting
1
1
DiPProꢀcompounds 1 bearing an acetyl (R = CH ) or a pentanoyl (R = C H ) residue, the halfꢀ
3
4
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
lives for the formed monoꢀmasked intermediates of type 2 were dependent on the acyl ester
group. The halfꢀlives of monoꢀmasked intermediates were determined by the decrease of the
corresponding peak areas in the HPLꢀchromatograms of these compounds in the hydrolysis
2
mixtures of the CH /R ꢀDiPProꢀnucleotides 7 and 9 in CEM/0 cell extracts after full
3
consumption of the starting DiPProꢀcompounds. The results are summarized in Table 2.
Table 2. Halfꢀlives of the intermediates 2 and 2’ in CEM/0 cell extracts.
2
NDP = d4TDP
2’ NDP = AZTDP
2
t
1/2 [min] in
CEM/0
99
t
1/2 [min] in
Intermediate (with R )
CEM/0
99
C H 2a/2’a
7
15
C H 2b/2’b
50
50
9
19
C H 2c/2’c
69
63
1
1
23
PhꢀCF 2d/2’d
690
230
ꢀ
3
PhꢀCH 2e
870
3
We assumed that the stability would increase with longer chain length. Interestingly,
intermediates 2a and 2’a were more stable than the ones with longer chain length. Therefore, it
seemed that the octanoyl group was a worse substrate for the ester cleaving enzymes.
In the case of the CF ꢀPhꢀintermediate 2d it is surprising that the attached nucleoside had
3
obviously an influence on the stability. The AZT intermediate 2’d was markedly more labile
than the d4T intermediate 2d for unknown reasons.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 60
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
Incubations with pig liver esterase: The nonꢀsymmetric aliphatic DiPProꢀAZTDPs 9aꢀc and
1
2
1
0aꢀc bearing a R = CH or C H in combination with long alkyl chains as the R residue were
3 4 9
incubated with pig liver esterase (PLE) to investigate the influence of the chain length on the
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
enzymatic cleavage by this esterase. Furthermore, CH /C H ꢀDiPProꢀd4TDP 7b and
3
9
19
C H /C H ꢀDiPProꢀd4TDP 8b were also included to study whether the attached nucleoside
4
9
9
19
analog has an effect on the hydrolysis pathway or on the compound stability.
The results of the hydrolysis experiments in the presence of PLE are summarized in Table 3.
Table 3. Halfꢀlives of the DiPProꢀcompounds 7ꢀ10 in the presence of PLE.
1
t1/2 [min]
of DiPProꢀNDP
0.14
t1/2 [min] of R ꢀ
intermediate
nm.*
Compound
CH /C H ꢀDiPProꢀd4TDP 7b
3
9
19
CH /C H ꢀDiPProꢀAZTDP 9a
0.26
0.21
0.45
0.19
0.27
0.27
1.2
nm.*
nm.*
nm.*
28
3
7
15
CH /C H ꢀDiPProꢀAZTDP 9b
3
9
19
CH /C H ꢀDiPProꢀAZTDP 9c
3
11 23
C H /C H ꢀDiPProꢀd4TDP 8b
4
9
9
19
C H /C H ꢀDiPProꢀAZTDP 10a
15
4
9
7
15
C H /C H ꢀDiPProꢀAZTDP 10b
19
4
9
9
19
C H /C H ꢀDiPProꢀAZTDP 10c
38
4
9
11 23
*nm. not measurable
All compounds were rapidly hydrolyzed and delivered the nucleoside diphosphates d4TDP or
AZTDP exclusively within a few minutes. The halfꢀlives determined for the CH /C H ꢀ and
3
7
15
CH /C H ꢀcompounds 9a,b as well as the C H /C H and C H /C H ꢀmasked pronucleotides
3
9
19
4
9
7
15
4
9
9
19
1
0a,b were found to be all in the same range (0.21 min to 0.27 min). The influence of the
nucleoside on the halfꢀlive was small although DiPProꢀd4TDPs 7b,8b proved to be somewhat
14
ACS Paragon Plus Environment

Page 15 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
less stable. In contrast, both compounds 9c,10c (CH /C H ꢀ and C H /C H ꢀDiPProꢀAZTDP)
3
11 23
4
9
11 23
bearing the longest aliphatic chain showed higher stability against hydrolysis by PLE.
2
Remarkably, in the case of the hydrolysis of all CH /R ꢀDiPPro 9aꢀc,7b, exclusively the
3
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
hydrolysis intermediate bearing the long aliphatic chain was formed although in very small
amounts. Therefore, it was not possible to determine the halfꢀlives of these intermediates from
the hydrolysis mixtures. This means that both alkanoyl esters were cleaved in almost identical
rate although with a slight preference of the C ꢀalkanoyl group. As an example the HPLꢀ
1
chromatograms of the incubations of compound 9b with PLE are shown in Figure 3.
Figure 3. RPꢀ18ꢀHPLCꢀchromatograms of the PLEꢀmediated hydrolysis of CH /C H ꢀDiPProꢀ
3
9
19
AZTDP 9b.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 60
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
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
Figure 4. RPꢀ18ꢀHPLCꢀchromatograms of the PLEꢀmediated hydrolysis of C H /C H ꢀDiPProꢀ
4
9
9
19
AZTDP 10b.
2
In contrast, incubation of C H /R ꢀDiPProꢀNDPs 10aꢀc, 8b with PLE initially showed high
4
9
amounts of the C H ꢀintermediate (Figure 4, e.g. the 1 minꢀHPLC chromatogram). In addition, as
4
9
for the C H ꢀ and C H ꢀintermediate also the C H ꢀchain bearing intermediate 2’b was
7
15
11 23
9
19
detected in the HPLꢀchromatograms although to a minor extent. In contrast to the above study
with the CH ꢀbearing DiPProꢀcompounds, here obviously the longer lipophilic chain acyl groups
3
are cleaved faster than the C H ꢀalkanoyl ester. This can be concluded from the markedly
4
9
different amounts of the two possible intermediates in the HPLꢀchromatograms (Figure 4). From
both formed intermediates finally AZTDP was released with a halfꢀlife of about 19 min as the
sole hydrolysis product.
These observations contrasted also with the results obtained in CEM/0 cell extracts and point
to a different substrate specificity of the esterase from pig liver and the esterases present in
16
ACS Paragon Plus Environment

Page 17 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
human CEM/0 cells or other enzymes that play a significant role in the hydrolysis process of the
DiPProꢀcompounds described here.
Comparative hydrolysis studies: Finally, we compared the nonꢀsymmetric DiPProꢀnucleotides
described here with compounds 11 bearing identical masking groups. Therefore, nonꢀsymmetric
and in addition to the symmetric counterparts DiPProꢀAZTDP derivatives with R=C H (11a),
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
4
9
C H (11b) and C H (11c) were incubated in phosphate buffer at pH 7.3 and in CEM/0 cell
6
13
9
19
extracts. The hydrolysis studies in PBS led to significant differences in the ratios of formed NDP
and NMP. In all cases the symmetric compounds 11 formed markedly more NMP and less NDP
as compared to the studies of the nonꢀsymmetric DiPProꢀcompounds disclosed here. Even more
interesting was the comparison of the hydrolysis in CEM/0 cell extracts. Also under these
conditions the hydrolysis of the symmetric C H ꢀDiPProꢀAZTDP 11a and C H ꢀDiPProꢀ
4
9
6
9
AZTDP 11b resulted in high NDP/NMP ratios. In contrast, this ratio decreased significantly in
the case of the more lipophilic C H ꢀDiPProꢀAZTDP 11c which was correlated to the difference
9
19
in the rate of metabolism. As mentioned above, the short chain alkyl esters are rapidly cleaved by
the esterases present in the extracts and therefore the side reaction did not interfere as in the case
of the more stable long chain aliphatic esters. Figure 5 summarizes a comparison of the
hydrolysis of CH /C H ꢀDiPProꢀAZTDP 9b, C H ꢀDiPProꢀAZTDP 11c and the nucleoside
3
9
19
9
19
diphosphate AZTDP after 10 hours of incubation. The analysis of the hydrolysis aliquots was
performed by means of HILICꢀHPLꢀchromatography. The NDP/NMP ratio in the case of the
nonꢀsymmetric prodrug 9b was found to be 5:1, for the symmetric prodrug 11c the ratio was
only 1.5:1. Interestingly, starting from independently synthesized AZTDP, the ratio was found to
be 3:1 after 10 h of incubation in CEM/0 cell extracts.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 60
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
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
Figure 5. HPLꢀchromatograms (HILICꢀcolumn) of the incubations of CH /C H ꢀDiPProꢀ
3
9
19
AZTDP 9b, C H ꢀDiPProꢀAZTDP 11c and AZTDP in CEM/0 cell extracts after 10 hours.
9
19
*components from the cell extracts
The markedly higher ratio found in the case of the nonꢀsymmetric DiPProꢀprodrug 9b
compared to AZTDP implies that compound 9b is not susceptible to cleavage at the
pyrophosphate moiety. The formation of the NMP only took place by the enzymatic
dephosphorylation after AZTDP delivery and thus after cleavage of the masking units. As a
consequence, the NDP/NMP ratio in case of the prodrug 9b was even higher as compared to that
of AZTDP for which the dephosphorylation immediately starts at the beginning of the
incubation. On the other hand, due to the high stability of the symmetric DiPProꢀcompound 11c
against enzymatic cleavage of the prodrug moieties, the chemical cleavage of the diphosphate
unit in this compound occurred by the above described concurrence reaction which led to
AZTMP and consequently to a low NDP/NMP ratio (1.5:1).
Antiviral Activity
18
ACS Paragon Plus Environment

Page 19 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
All DiPProꢀcompounds 7ꢀ11 were evaluated for their antiꢀHIV activity in HIVꢀ1ꢀ and HIVꢀ2ꢀ
infected wildꢀtype CEM/0 cell cultures and in HIVꢀ2ꢀinfected mutant thymidine kinaseꢀdeficient
ꢀ
CEM/TK cell cultures. Table 4 summarizes the antiviral and cytostatic data of the DiPProꢀ
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
compounds 7ꢀ10 and the parent nucleoside analogues d4T and AZT, respectively, as reference
compounds.
D4TDP derivatives 7,8 bearing aliphatic ester functions in the acyloxybenzyl units (aꢀc)
proved to be antivirally active against HIVꢀ1 and HIVꢀ2 at least in the same concentration range
as compared to the parent compound d4T in wildꢀtype CEM/0 cell cultures. Often the inhibitory
potency was even somewhat higher as compared to that of the parent nucleoside particularly for
compound 8c. The antiviral activity observed in the wildꢀtype CEM/0 cell cultures was
completely retained in the case of the lipophilic DiPProꢀcompounds 7b and 8b,c in mutant
ꢀ
thymidineꢀdeficient CEM cells (TK ). Moreover, the antiviral activity increased with the increase
of lipophilicity associated with the introduced acylꢀmoieties of the DiPProꢀprodrugs.
Disappointingly, DiPProꢀd4TDPs 7,8 bearing benzoyl ester functions in the masking units
(d,e) proved to be less potent as compared to the alkanoyl counterparts. They also were less
active as compared to d4T in the wildꢀtype CEM/0 cell cultures and lost more or less all antiviral
activity in the TKꢀdeficient CEM cell cultures (EC : ≥ 10 µM).
5
0
Alike the d4TDPꢀcontaining DiPProꢀcompounds, the AZTDP derivatives 9,10 were able to
inhibit the HIVꢀ1 and HIVꢀ2 induced cytopathogenic effect at very low concentrations in wildꢀ
type CEM/0 cell cultures.
Table 4: AntiꢀHIV and cytostatic activity of the DiPProꢀNDPs 7ꢀ11.
a)
b)
EC50 [ꢀM]
CC50 [ꢀM]
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 60
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
ꢀ
CEM/0
CEM/0
CEM/TK
Compound
HIVꢀ1 (HE)
HIVꢀ2 (ROD)
HIVꢀ2 (ROD)
CEM/0
CH /C H ꢀDiPProꢀd4TDP 7a
0.33 ± 0.11
0.33 ± 0.11
0.29 ± 0.16
0.72 ± 0.16
0.29 ± 0.057
0.33 ± 0.11
0.36 ± 0.046
0.10 ± 0.03
0.40 ± 0.00
0.3 ± 0.014
0.042 ± 0.008
0.016 ± 0.0
0.042 ± 0.008
0.012 ± 0.007
0.036 ± 0.0
0.034 ± 0.018
0.013 ± 0.005
0.036 ± 0.0
0.011 ± 0.0018
0.016 ± 0.0
0.026 ± 0.014
0.52 ± 0.32
0.044 ± 0.01
1.05 ± 0.30
1.36 ± 0.13
0.28 ± 0.04
1.38 ± 0.16
1.47 ± 0.29
1.47 ± 0.29
1.17 ± 0.40
0.28 ± 0.04
0.72 ± 0.16
1.45 ± 0.00
0.16 ± 0.08
0.14 ± 0.08
0.08 ± 0.00
0.06 ± 0.03
0.07 ± 0.01
0.29 ± 0.18
0.06 ± 0.02
0.06 ± 0.02
0.054 ± 0.01
0.058 ± 0.03
0.23 ± 0.15
2.23 ± 1.51
0.49 ± 0.57
6.31 ± 0.00
2.32 ± 1.6
1.77 ± 0.33
> 10
33 ± 27
15 ± 1
26 ± 0
51 ± 15
11 ± 6
66 ± 15
20 ± 1
30 ± 17
37 ± 31
14 ± 1
29 ± 3
20 ± 9
18 ± 3
100 ± 8
73 ± 12
41 ± 23
36 ± 28
58 ± 54
60 ± 7
47 ± 8
38 ± 10
79 ± 3
>250
3
7
15
CH /C H ꢀDiPProꢀd4TDP 7b
3
9
19
CH /C H ꢀDiPProꢀd4TDP 7c
3
11 23
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
CH /CF ꢀPhꢀDiPProꢀd4TDP 7d
3
3
CH /CH ꢀPhꢀDiPProꢀd4TDP 7e
10 ± 0.00
6.31 ± 0.00
0.83 ± 0.00
0.13 ± 0.07
>10
3
3
C H /C H ꢀDiPProꢀd4TDP 8a
4
9
7
15
C H /C H ꢀDiPProꢀd4TDP 8b
4
9
9
19
C H /C H ꢀDiPProꢀd4TDP 8c
4
9
11 23
C H /CF ꢀPhꢀDiPProꢀd4TDP 8d
4
9
3
C H /CH ꢀPhꢀDiPProꢀd4TDP 8e
10 ± 0.00
>10
4
9
3
CH /C H ꢀDiPProꢀAZTDP 9a
3
7
15
CH /C H ꢀDiPProꢀAZTDP 9b
>10
3
9
19
CH /C H ꢀDiPProꢀAZTDP 9c
10 ± 2.61
>10
3
11 23
CH /CF ꢀPhꢀDiPProꢀAZTDP 9d
3
3
C H /C H ꢀDiPProꢀAZTDP 10a
>10
4
9
7
15
C H /C H ꢀDiPProꢀAZTDP 10b
10 ± 0.00
8.16 ± 2.61
>10
4
9
9
19
C H /C H ꢀDiPProꢀAZTDP 10c
4
9
11 23
C H /CF ꢀPhꢀDiPProꢀAZTDP 10d
4
9
3
C H ꢀDiPProꢀAZTDP 11a
>10
4
9
C H ꢀDiPProꢀAZTDP 11b
>10
6
13
C H ꢀDiPProꢀAZTDP 11c
1.3 ± 0.0
150 ± 9
>250
9
19
d4T
AZT
[
a] Antiviral activity in CEM Tꢀlymphocyte cultures: 50% effective concentration; [b] Cytostatic activity: 50% cytostatic
concentration to inhibit CEM/0 cell proliferation by 50%.
ꢀ
In mutant thymidine kinaseꢀdeficient CEM/TK cell cultures only the lipophilic compounds 9c,
1
0b and 10c showed some antiviral activity (EC : ~ 10 µM). Interestingly, the symmetric
50
C H ꢀDiPProꢀcompound 11c was active at 1.3 ꢀM in the TKꢀdeficient CEM cell cultures. One
9
19
reason for this pronounced retention of antiviral activity of 11c might be that the lipophilicity of
this compound was sufficiently high to allow a passage through the cell membrane. In the
hydrolysis studies discussed above it was shown that all compounds bearing an acetyl ester as an
acyl moiety in the masking group were cleaved extremely fast. Thus, if these compounds are not
20
ACS Paragon Plus Environment

Page 21 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
lipophilic enough they will be cleaved in the cell culture medium before the uptake into the cells.
Then AZTDP will be formed extracellularly which will be enzymatically dephosphorylated to
give AZT. AZT is able to migrate through the membrane and this result in antiviral activity in
the wildꢀtype cell cultures. It has been reported that the phosphorylation of AZTMP to AZTDP
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
5,16
by dTMP kinase is rateꢀlimiting.
However, another reason for the failure of the DiPProꢀ
AZTDP compounds might be that even the delivered AZTDP is not efficiently metabolized into
the corresponding triphosphate. In such a case the delivery of AZTDP will not lead to increased
antiviral activity. It was already reported that in biochemical activity studies using nucleoside
diphosphate kinase (NDPꢀK), which is generally the enzyme that is involved in the
phosphorylation of nucleoside diphosphates into their triphosphate forms, the phosphorylation to
2
7
AZTTP was rather poor. So, maybe there is not only a bottleꢀneck in the metabolism of
28
AZTMP to AZTDP but also a bottleꢀneck in the conversion of AZTDP to form AZTTP which
becomes more prominent if the concentration of AZTDP is increased due to its delivery from a
prodrug. From our studies in CEM/0 cell extracts we clearly have proven that the exclusively
formed product from the prodrug was AZTDP.
SUMMARY AND CONCLUSION
Nonꢀsymmetric DiPProꢀNDPs 1 of the antiꢀHIV active nucleoside analogues d4T and AZT are
disclosed here bearing two different masking units attached to the βꢀphosphate group of the
corresponding nucleoside diphosphate. For the synthesis of the title compounds, a new protocol
for the synthesis of nonꢀsymmetric bis(acyloxybenzyl)ꢀphosphoramidites was developed, which
yielded the phosphoramidites in high chemical yields. These were coupled by weak acid
catalysis with nucleoside monophosphates followed by oxidation. Using this route, DiPProꢀ
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 60
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
d4TDP and DiPProꢀAZTDP compounds with different combinations of aliphatic and aromatic
acyl moieties were obtained in good yields. Hydrolysis studies in phosphate buffer, RPMIꢀ
culture medium, CEM/0 cell extracts and pig liver esterase confirmed the designed delivery
mechanism described above. Halfꢀlives in PBS and RPMIꢀculture medium were markedly higher
than those in CEM/0 cell extracts and incubations with PLE. Although in PBS still both the
nucleoside diphosphates and the monophosphates were formed, the exclusive products formed
by enzymatic hydrolysis of the prodrugs (cell extracts and PLE) were the nucleoside
diphosphates. As compared to the symmetric DiPProꢀNDPs with two identical masks reported
previously, the ratio of formed NDP to NMP was markedly higher in all media. Most of the nonꢀ
symmetric DiPProꢀcompounds were as active as or even more active than the corresponding
parent nucleosides in wildꢀtype CEM/0 cell cultures. Moreover, also high activities were
obtained in dependence on the lipophilicity of the DiPProꢀd4TDP prodrugs against HIVꢀ2 in
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
ꢀ
mutant CEM/TK cell cultures. This confirmed that these compounds were efficiently takenꢀup
by the cells and delivered intracellularly a phosphorylated form of d4T, most likely d4T
diphosphate. Surprisingly, the good antiviral activities of DiPProꢀAZTDPs in the wildꢀtype cell
cultures were only weakly retained in the mutant TKꢀdeficient cell cultures. This may point to a
possible further bottleꢀneck in the phosphorylation of the AZTDP to AZTTP. A clear conclusion
from the antiviral activity data was that DiPProꢀcompounds bearing short acetyl ester or benzoyl
ester groups in one of the prodrug moieties were not suitable for further development. The first
ester group is cleaved too fast and is not lipophilic enough to allow a successful uptake into the
target cells while the second type of ester moiety is also not lipophilic enough although we have
2
6
shown before that the stability of such compounds could be adjusted over a wide range.
22
ACS Paragon Plus Environment

Page 23 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Nevertheless, this report on nonꢀsymmetric DiPProꢀcompounds proved the successful and
selective nucleoside diphosphate delivery as a further development of the DiPProꢀapproach.
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
EXPERIMENTAL SECTION
General. All experiments involving waterꢀsensitive compounds were conducted under
anhydride conditions and nitrogen atmosphere. All solvents were dried over an appropriate
drying agent. Triethylamine and acetonitrile were dried by heating under reflux over calcium
hydride for several days followed by distillation. Acetonitrile was stored over 3Å molecular
sieves. THF was dried by heating under reflux over potassium followed by distillation.
Petroleum ether 50ꢀ70 for chromatography was distilled before use. The evaporation of solvents
was carried out on a rotary evaporator under reduced pressure or using a highꢀvacuum pump.
Preparative radial chromatography was performed by chromatotron (Harrison Research, Model
7
924T) with glass plates coated with 1, 2, or 4 mm layers of VWR 60 PF254 silica gel containing
a fluorescent indicator (VWR no. 7749) were used. Analytical thinꢀlayer chromatography was
performed on preꢀcoated aluminum plates with a 0.2ꢀmm layer of silica gel containing a
fluorescent indicator (Macherey & Nagel Xtra Sil UV ); sugarꢀcontaining compounds were
2
54
visualized by wetting the plates with 10% sulphuric acid in water and heating with a fan.
1
Instrumentation. HꢀNMR Spectroscopy was carried out using a Bruker AMX 400 or AV
400 at 400 MHz or a Bruker AV 600 at 600 MHz with CDCl or MeODꢀd as internal standards.
3
4
13
CꢀNMR spectra were recorded on a Bruker AMX 400 or AV 400 at 101 MHz or on a Bruker
3
1
AV 600 at 150 MHz (CDCl or MeODꢀd as internal standard). PꢀNMR spectra were recorded
3
4
on a Bruker AMX 400 at 162 MHz or a Bruker AV 600 at 243 MHz (H PO as internal
3
4
19
1
31
13
standard). FꢀNMR spectra were recorded on a Bruker AV 600 MHz. All Hꢀ, Pꢀ and Cꢀ
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 60
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
13
31
NMR chemical shifts are quoted in parts per million (ppm). All Cꢀ and PꢀNMR spectra were
recorded in the protonꢀdecoupled mode. High resolution ESI mass spectra were recorded with an
Agilent 6224 ESIꢀTOF spectrometer in positive or negative modus. MALDI mass spectra were
recorded on an ultrafleXtreme MALDIꢀTOFꢀTOF mass spectrometer by Bruker Daltonik with
9ꢀAA as matrix. All HPLC measurements were carried out using a VWR Hitachi ‘Elite
LaChrom’ (pump Lꢀ2130, auto sampler Lꢀ2200, column oven Lꢀ2300, diode array detector Lꢀ
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
2
455). Analytical Rp18ꢀHPLC: Macherey & Nagel EC 125/3 NUCLEODUR 100ꢀ5 C18 EC. As
+
ꢀ
eluents CH CN/[(C H ) N] CH COO buffer (2 mM, pH 6.0) gradient was used. Method A:
3
4
9 4
3
0
ꢀ24 min CH CN (5ꢀ80%), 24ꢀ29 min CH CN (80%), flow rate 1 mL/min. Method B: 0ꢀ27 min
3 3
CH CN (5ꢀ90%), 27ꢀ30 min CH CN (90%), flow rate 1 mL/min, column temperature: 25 °C.
3
3
UV detection at 265 nm. The purities of the DiPPro nucleotides were evaluated by means of
HPLC and were in all cases ≥95%. Analytical HILICꢀHPLC: Macherey & Nagel EC 125/4
+
ꢀ
NUCLEODUR HILIC, 5ꢀm. As eluents CH CN/NH HCOO buffer (10 mM, pH 8.3) gradient
3
4
was used. Method C: 0ꢀ5 min CH CN (95%), 5ꢀ15 min CH CN (95ꢀ70%), 15ꢀ30 min CH CN
3
3
3
(
70%) flow rate 0.6 mL/min, column temperature: 25 °C. UV detection at 265 nm.
Synthesis.
General procedure A: Preparation of acyloxybenzyl bis(diisopropylamino)-
phosphoramidite 6: To a low concentrated solution of 1 equivalent phosphorus trichloride and
equivalent pyridine in THF 1 equivalent 4ꢀacyloxybenzyl alcohol, dissolved in THF, was
1
added dropwise over 1.5 h at ꢀ78 °C. After removing of the cold bath the reaction was stirred for
about 20 h at room temperature until the total reaction of 4ꢀacyloxybenzyl alcohol. Then
6
.1 equivalents N,Nꢀdiisopropylamine were added dropwise at ꢀ10 °C. After removing of the
24
ACS Paragon Plus Environment

Page 25 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cold bath the reaction was stirred for about one to two days at room temperature. Salts were
removed by filtration. The crude product was purified by radial chromatography at the
chromatotron with petroleum ether/triethylamine as eluent.
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
General procedure B: Preparation of non-symmetric bis(acyloxybenzyl)phosphor-
amidites 5: To a solution of 1.5 equivalents phosphordiamidite in acetonitrile 1 equivalent
4ꢀacyloxybenzyl alcohol and 1 equivalent of a DCIꢀactivator solution (0.25 molar in acetonitrile)
also dissolved in acetonitrile were added dropwise at 4 °C. After removing of the cold bath the
reaction mixture was stirred for 30 to 60 min. The reaction was stopped by removing the solvent
under reduced pressure. The residue was dissolved in petroleum ether/triethylamine (9:1) and
filtrated. The crude product was purified by radial chromatography at the chromatotron with
petroleum ether/triethylamine as eluent.
General procedure C: Preparation of DiPPro-nucleoside diphosphates: 1.5 equivalents
phosphoramidite 5 were added to a solution of 1 equivalent of the bis(tetraꢀnꢀbutylammonium)
nucleoside monophosphate in acetonitrile. The reaction was started by addition of
0
.5 equivalents 4,5ꢀdicyanoimidazole (0.25 M solution in acetonitrile) at room temperature.
Every 5 min again 0.25 equivalents 4,5ꢀdicyanoimidazole were added up to a total of 1.25 to
.75 equivalents. After stirring another 5 min after the last addition the mixture was oxidized by
1
addition of 1.5 equivalents tertbutylhydroperoxid (5.5 molar solution in nꢀdecane). The solvents
were removed under reduced pressure after 15 min. The compounds were isolated by automatic
flash RP18ꢀ chromatography (acetonitrile/water). After the tetraꢀnꢀbutylammonium ions were
+
changed with ammonia by elution over DOWEX 50WX8 (NH ), a second automatic flash
4
RP18ꢀ chromatography (acetonitrile/water) followed.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 60
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
24
25
25
Syntheses of 4ꢀ(hydroxymethyl)phenylacetate 3a, 4ꢀ(hydroxymethyl)phenylpentanoate 3b,
2
4
4
ꢀ(hydroxymethyl)phenyloctanoate 4a,
4ꢀ(hydroxymethyl)phenyldecanoate 4b,
2
5
4
ꢀ(hydroxylmethyl)phenyldodecanoate 4c,
4ꢀ(hydroxymethyl)phenylꢀ4’ꢀtrifluoromethylꢀ
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
26
26
benzoate 4d,
4ꢀ(hydroxymethyl)phenylꢀ4’ꢀmethylbenzoate 4e,
4ꢀ(hydroxymethyl)ꢀ
2
5
phenylheptanoate 4f have been described before. The analytical data were identical to those
reported before.
2
4,25
Syntheses of (N[n-Bu] ) -d4TMP and (N[n-Bu] ) -AZTMP have been described before.
4
2
4 2
The analytical data were identical to those reported before.
Acetoxybenzyl-bis(diisopropylamino)phosphoramidite 6a: General procedure B; 4ꢀ
Acyloxybenzylalcohol 3a (2.36 g, 14.4 mmol), dissolved in 15 mL THF and phosphorus
trichloride (1.26 mL, 14.4 mmol) and pyridine (1.17 mL, 14.4 mmol) in 35 mL THF. In the
following step DIPA (12.4 mL, 87.8 mmol) was added. The product (3.02 g, 7.62 mmol, 53%)
1
was obtained as a colorless waxꢀlike solid. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 7.39ꢀ7.37
3
(
m, 2 H, Hꢀ2); 7.05ꢀ7.04 (m, 2 H, Hꢀ3); 4.64ꢀ4.63 (m, 2 H, PhꢀCH ); 3.61ꢀ3.54 (m, 4 H, CNꢀH);
2
3
3
13
2
.29 (s, 3 H, Hꢀ6); 1.19 (d, J = 6.9 Hz, 12 H, isoPr); 1.19 (d, J = 7.0 Hz, 12 H, isoPr); Cꢀ
H,H H,H
3
NMR: (101 MHz, CDCl ): δ [ppm] = 169.6 (Cꢀ5); 149.6 (Cꢀ4); 138.3 (d, J = 10.6 Hz, Cꢀ1);
3
C,P
2
1
1
27.9 (Cꢀ2); 121.3 (Cꢀ3); 65.6 (d, J = 23.2 Hz, PhꢀCH ); 44.5 (d, J = 12.3 Hz, 2x NꢀCH);
C,P 2 C,N
2
2
31
2
4.7 (d, J = 8.0 Hz, isoPr); 24.0 (d, J = 5.6 Hz, isoPr); 21.2 (Cꢀ6). PꢀNMR (162 MHz,
C,N
C,N
~
ꢀ
CDCl , decoupled): δ/ppm = 123.4; IR
ν
[cm ]= 2970, 2923, 2901, 1766, 1506, 1454, 1302,
3
+
1
114, 1049, 950, 741. MS (MALDI) m/z = calcd for C H N O P: 589.331 [MꢀH +9ꢀ
3
4
46
3
6
ꢀ
aminoacridine (9ꢀAA) used as matrix] , found: 589.309.
Pentanoyloxybenzyl-bis(diisopropylamino)phosphoramidite 6b: General procedure B; 4ꢀ
Acyloxybenzylalcohol 3b (2.25 g, 10.8 mmol), dissolved in 15 mL THF and phosphorus
26
ACS Paragon Plus Environment

Page 27 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trichloride (0.94 mL, 10.8 mmol) and pyridine (0.88 mL, 10.8 mmol) in 35 mL THF. In the
following step (9.24 mL, 65.8 mmol) was added. The product (2.65 g, 6.04 mmol, 56%) was
1
obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 7.38ꢀ7.36 (m, 2 H, Hꢀ2);
3
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
7
.04ꢀ7.02 (m, 2 H, Hꢀ3); 4.64ꢀ4.62 (m, 2 H, PhꢀCH ); 3.62ꢀ3.52 (m, 4 H, CNꢀH); 2.29 (t,
2
3
JHH = 7.5 Hz, 2 H, Hꢀ6); 1.78ꢀ1.70 (m, 2H, Hꢀ7); 1.50ꢀ1.40 (m, 2H, Hꢀ8); 1.18 (d,
3
3
3
JH,H = 6.8 Hz, 12 H, isoPr); 1.18 (d, J = 6.9 Hz, 12 H, isoPr); 0.97 (t, J = 7.4 Hz, 2 H, Hꢀ
H,H
H,H
1
3
3
9). CꢀNMR: (101 MHz, CDCl ): δ [ppm] = 172.5 (Cꢀ5); 149.7 (Cꢀ4); 138.2 (d, J = 10.7 Hz,
3 C,P
2
1
Cꢀ1); 127.9 (Cꢀ2); 121.3 (Cꢀ3); 65.7 (d, JC,P = 23.0 Hz, PhꢀCH ); 44.6 (d, J = 12.4 Hz, 2x Nꢀ
2
C,N
2
2
CH); 34.3 (Cꢀ6); 27.2 (Cꢀ7); 24.8 (d, JC,N = 7.9 Hz, isoPr); 24.0 (d, J = 5.7 Hz, isoPr); 22.4
C,N
3
1
~
ꢀ
(Cꢀ8); 13.9 (Cꢀ9). PꢀNMR (162 MHz, CDCl , decoupled): δ/ppm = 123.5. IR
ν
[cm ]= 2964,
3
2930, 2870, 1760, 1507, 1390, 1195, 1182, 1144, 1016, 951, 779. MS (MALDI) m/z = calcd for
C H N O P: 631.386 [M+ 9ꢀaminoacridine (9ꢀAA) used as matrix], found: 631.368.
37
52
3
6
CH /C H -phosphoramidite 5a: General procedure C; Acetoxybenzylꢀbis(diisopropylꢀ
3
7
15
amino)phosphoramidite 6a (1.24 g, 3.12 mmol) dissolved in 15 mL acetonitrile and
4ꢀ(hydroxymethyl)phenyloctanoate 4a (520 mg, 2.08 mmol) and DCIꢀactivator solution
(8.32 mL, 2.08 mmol) dissolved in 5 mL acetonitrile. The product (932 mg, 1.71 mmol, 82%)
1
was obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 7.36ꢀ7.34 (m 4H, Hꢀ2,
3
Hꢀ2‘); 7.05ꢀ7.03 (m, 4H, Hꢀ3, Hꢀ3‘); 4.77ꢀ4.73 (m, 2H, PhꢀCH ); 4.70ꢀ4.66 (m, 2H, PhꢀCH );
2
2
3.72ꢀ3.66 (m, 2H, 2x NCꢀH); 2.56ꢀ2.53 (m, 2H, Hꢀ6‘); 2.29 (s, 3H, ꢀCH ); 1.78ꢀ1.73 (m, 2H, Hꢀ
3
3
3
7‘); 1.44ꢀ1.29 (m, 12H, Hꢀ8‘ – Hꢀ11‘); 1.20 (d, J = 6.8 Hz, 12H, isoPr); 0.90 (d, J = 6.8
H,H
H,H
1
3
Hz, 3H, Hꢀ12‘). CꢀNMR: (101 MHz, CDCl ): δ [ppm] = 172.5 (Cꢀ5‘); 169.6 (Cꢀ5); 150.1 (Cꢀ
3
3
3
4‘); 150.0 (Cꢀ4); 137.3 (d, J = 7.6 Hz, Cꢀ1‘); 137.1 (d, J = 7.5 Hz, Cꢀ1); 128.1, 128.1 (Cꢀ2,
C,P C,P
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 60
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
2
2
Cꢀ2‘); 121.5, 121.5 (Cꢀ3, Cꢀ3‘); 65.0 (d, JC,P = 4.7 Hz, PhꢀCH ); 64.9 (d, J = 4.8 Hz, Phꢀ
2
C,P
1
CH ); 43.3 (d, J = 12.5 Hz, NꢀCH); 34.6 (Cꢀ6‘); 31.8, 29.2, 29.6, 22.7, (Cꢀ8‘ – Cꢀ11‘); 25.1
2
C,N
3
1
(
Cꢀ7‘); 24.8, 24.8 (2x isoPr); 21.3 (Cꢀ6); 14.2 (Cꢀ12‘). PꢀNMR (162 MHz, CDCl , decoupled):
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
ꢀ
~
δ/ppm = 148.0. IR
ν [cm ]= 2965, 2929, 2859, 1760, 1608, 1507, 1366, 1192, 1163, 1005, 973,
+
+ +
756, 504. HRMS (ESI ) m/z = calcd for C H NO P: 546.2979 [M+H ] , found: 546.2951.
3
0
45
6
CH /C H -phosphoramidite 5b: General procedure C; Acetoxybenzylꢀbis(diisopropylꢀ
3
9
19
amino)phosphoramidite 6a (520 mg, 1.31 mmol) dissolved in 10 mL acetonitrile and
ꢀ(hydroxymethyl)phenyldecanoate 4b (244 mg, 0.88 mmol) and DCIꢀactivator solution
4
(
3.52 mL, 0.88 mmol) dissolved in 6 mL acetonitrile. The product (437 mg, 0.76 mmol, 87%)
1
was obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 7.36ꢀ7.34 (m 4H, Hꢀ2,
3
Hꢀ2‘); 7.05ꢀ7.02 (m, 4H, Hꢀ3, Hꢀ3‘); 4.77ꢀ4.64 (m, 4H, 2x PhꢀCH ); 3.73ꢀ3.64 (m, 2H, 2x NCꢀ
2
H); 2.56ꢀ2.52 (m, 2H, Hꢀ6‘); 2.29 (s, 3H, ꢀCH ); 1.79ꢀ1.71 (m, 2H, Hꢀ7‘); 1.44ꢀ1.27 (m, 12H, Hꢀ
3
3
3
13
8
‘ – Hꢀ13‘); 1.20 (d, J = 6.8 Hz, 12H, isoPr); 0.89 (d, J = 6.8 Hz, 3H, Hꢀ14‘). CꢀNMR:
H,H H,H
(
101 MHz, CDCl ): δ [ppm] = 172.5 (Cꢀ5‘); 169.6 (Cꢀ5); 150.2 (Cꢀ4‘); 150.0 (Cꢀ4); 137.2 (d,
3
3
2
JC,P = 7.4 Hz, Cꢀ1, Cꢀ1‘); 128.2, 128.1 (Cꢀ2, Cꢀ2‘); 121.5, 121.5 (Cꢀ3, Cꢀ3‘); 65.1 (d, J = 3.2
C,P
2
1
Hz, PhꢀCH ); 64.9 (d, J = 3.4 Hz, PhꢀCH ); 43.3 (d, J = 12.4 Hz, NꢀCH); 34.6 (Cꢀ6‘);
2
C,P
2
C,N
3
2.0, 29.6, 29.4, 29.3, 22.8, (Cꢀ8‘ – Cꢀ13‘); 25.1 (Cꢀ7‘); 24.8, 24.8 (2x isoPr); 21.3 (Cꢀ6); 14.2
3
1
~
ꢀ
(Cꢀ14‘). PꢀNMR (162 MHz, CDCl , decoupled): δ/ppm = 148.0. IR
ν [cm ]= 2964, 2926,
3
+
2
855, 1759, 1507, 1192, 1005, 972, 755, 503. HRMS (ESI ) m/z = calcd for C H NO P:
32
49
6
+
+
5
74.3292 [M+H ] , found: 574.3196.
28
ACS Paragon Plus Environment

Page 29 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CH /C H -phosphoramidite 5c: General procedure C; Acetoxybenzylꢀbis(diisopropylꢀ
3
11 23
amino)phosphoramidite 6a (620 mg, 1.57 mmol) dissolved in 12 mL acetonitrile and
ꢀ(hydroxymethyl)phenylꢀdodecanoate 4c (319 mg, 1.04 mmol) and DCIꢀactivator solution
4
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
(
4.16 mL, 1.04 mmol) in 5 mL acetonitrile. The product (453 mg, 0.75 mmol, 72%) was
1
obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 7.36ꢀ7.33 (m 4H, Hꢀ2, Hꢀ
3
2
‘); 7.05ꢀ7.02 (m, 4H, Hꢀ3, Hꢀ3‘); 4.78ꢀ4.64 (m, 4H, 2x PhꢀCH ); 3.73ꢀ3.64 (m, 2H, 2x NCꢀH),
2
2
.56ꢀ2.52 (m, 2H, Hꢀ6‘); 2.29 (s, 3H, ꢀCH ); 1.78ꢀ1.71 (m, 2H, Hꢀ7‘); 1.43ꢀ1.26 (m, 16H, Hꢀ8‘ –
3
3
3
13
Hꢀ15‘); 1.20 (d, J = 6.8 Hz, 12H, isoPr); 0.89 (d, J = 6.8 Hz, 3H, Hꢀ16‘). CꢀNMR: (101
H,H
H,H
MHz, CDCl ): δ [ppm] = 172.5 (Cꢀ5‘); 169.5 (Cꢀ5); 150.2 (Cꢀ4‘); 150.0 (Cꢀ4); 137.2 (Cꢀ1, Cꢀ
3
2
2
1
‘); 128.2 (Cꢀ2, Cꢀ2‘); 121.5 (Cꢀ3, Cꢀ3‘); 65.0 (d, J = 18.4 Hz, PhꢀCH ); 65.0 (d, J = 18.9
C,P 2 C,P
1
Hz, PhꢀCH ); 43.3 (d, J = 12.2 Hz, NꢀCH); 34.6 (Cꢀ6‘); 29.8, 29.7, 29.6, 29.5, 29.4, 29.4,
2
C,N
31
2
9.3 (Cꢀ8‘ – Cꢀ15‘); 25.1 (Cꢀ7‘); 24.8, 24.8 (2x isoPr); 21.3 (Cꢀ6); 14.4 (Cꢀ16‘). PꢀNMR (162
~
ꢀ
MHz, CDCl , decoupled): δ/ppm = 148.0. IR
ν
[cm ]= 2964, 2925, 2854, 1760, 1608, 1507,
3
+
1
365, 1192, 1163, 1005, 973, 756, 503. HRMS (ESI ) m/z = calcd for C H NO P: 602.3605
3
4
53
6
+
+
[
M+H ] , found: 602.4100.
CH /CF -Ph-phosphoramidite 5d: General procedure C; Acetoxybenzylꢀbis(diisopropylꢀ
3
3
amino)phosphoramidite 6a (544 mg, 1.37 mmol) dissolved in 12 mL acetonitrile and 4ꢀ
(hydroxymethyl)phenylꢀ4’ꢀtrifluoromethylbenzoate 4d (271 g, 0.915 mmol) and DCIꢀactivator
solution (3.66 mL, 0.915 mmol) in 5 mL acetonitrile. The product (360 mg, 0.61 mmol, 67%)
1
was obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 8.33ꢀ8.32 (m 2H, Hꢀ7’);
3
7
.79ꢀ7.78 (m, 2H, Hꢀ8’); 7.43ꢀ7.41 (m, 2H, Hꢀ2’); 7.37ꢀ7.36 (m, 2H, Hꢀ2); 7.19ꢀ7.18 (m, 2H, Hꢀ
3
’); 7.06ꢀ7.05 (m, 2H, Hꢀ3); 4.81ꢀ 4.67 (m, 4H, 2x PhꢀCH ); 3.74ꢀ 3.68 (m, 4H, CNꢀH); 2.29 (s,
2
29
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 30 of 60
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
3
13
3
H, CH ); 1.22 (d, J = 6.7 Hz, 6H, isoPr); 1.22 (d, J = 6.5 Hz, 6H, isoPr). CꢀNMR (101
3
H,H
H,H
3
MHz, CDCl ): δ [ppm] = 169.7 (Cꢀ5); 164.2 (Cꢀ5‘); 150.0 (Cꢀ4); 149.9 (Cꢀ4); 137.7 (d, J
=
3
C,P
3
3
7
.5 Hz, Cꢀ1‘); 137.2 (d, JC,P = 7.8 Hz, Cꢀ1); 135.2 (d, J = 33 Hz, Cꢀ9‘), 133.0 (Cꢀ6‘); 130.7
C,F
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
3
(
Cꢀ8‘); 128.3 (Cꢀ2‘); 128.3 (Cꢀ2); 125.8 (d, JC,F = 3.6 Hz, Cꢀ7‘); 125.7 (CF ); 121.5, 121.5 (Cꢀ3,
3
2
2
1
Cꢀ3‘); 65.0 (d, J = 18.5 Hz, PhꢀCH ); 65.0 (d, J = 18.6 Hz, PhꢀCH ); 43.2 (d, J = 12.3
C,P
2
C,P
2
C,N
2
19
Hz, NꢀCH); 24.8, 24.8 (d, J = 7.3 Hz, 2x isoPr); 21.3 (Cꢀ6). FꢀNMR (565 MHz, CDCl ):
C,N
3
31
~
ꢀ
δ/ppm = 67.1. PꢀNMR (162 MHz, CDCl , decoupled): δ/ppm = 148.1. IR
ν
[cm ]= 2968,
3
2932, 2869, 1762, 1741, 1608, 1506, 1324, 1264, 1193, 1129, 1075, 1013, 770, 552. HRMS
+
+ +
(
ESI ) m/z = calcd for C H F NO P: 592.2071 [M+H ] , found: 592.2209.
30 34 3 6
CH /CH -Ph-phosphoramidite 5e: General procedure C; Acetoxybenzylꢀbis(diisopropylꢀ
3
3
amino)phosphoramidite 6a (505 mg, 1.27 mmol) dissolved in 10 mL acetonitrile and 4ꢀ
(hydroxymethyl)phenylꢀ4’ꢀmethylbenzoate 4e (205 mg, 0.85 mmol) and DCIꢀactivator solution
(
3.40 mL, 0.85 mmol) in 5 mL acetonitrile. The product (323 mg, 0.60 mmol, 71%) was
1
obtained as a colorless oil. HꢀNMR: (400 MHz, CDCl ): δ [ppm] = 8.11ꢀ8.09 (m 2H, Hꢀ7’/Hꢀ
3
8‘); 7.42ꢀ7.36 (m, 4H, Hꢀ2, Hꢀ2‘); 7.32ꢀ7.30 (m, 2H, Hꢀ7’/Hꢀ8‘); 7.19ꢀ7.17 (m, 2H, Hꢀ3’); 7.07ꢀ
7
.05 (m, 2H, Hꢀ3); 4.81ꢀ 4.67 (m, 4H, 2x PhꢀCH ); 3.76ꢀ3.67 (m, 4H, CNꢀH); 2.45 (s, 3H, ꢀPhꢀ
2
3
3
CH ); 2.29 (s, 3H, ꢀCH ); 1.22 (d, J = 6.8 Hz, 6H, isoPr); 1.22 (d, J = 6.8 Hz, 6H, isoPr).
3
3
H,H
H,H
1
3
CꢀNMR (101 MHz, CDCl ): δ [ppm] = 169.6 (Cꢀ5); 165.4 (Cꢀ5‘); 150.3 (Cꢀ4); 150.0 (Cꢀ4‘);
3
2
2
1
7
1
44.5 (Cꢀ9‘); 137.3 (d, J = 8.0 Hz, Cꢀ1/Cꢀ1‘); 137.2 (d, J = 7.8 Hz, Cꢀ1/Cꢀ1‘); 130.3 (Cꢀ
C,P C,P
2
‘); 129.4 (Cꢀ8‘); 128.2, 128.2 (Cꢀ2/Cꢀ2‘); 127.0 (Cꢀ6‘); 121.7, 121.5 (Cꢀ3, Cꢀ3‘); 65.0 (d, J
=
C,P
2
1
8.4 Hz, PhꢀCH ); 65.1 (d, J = 18.3 Hz, PhꢀCH ); 43.3 (d, J = 12.4 Hz, NꢀCH); 24.8, 24.8
2
C,P
2
C,N
2
31
(
d, J = 7.2 Hz, 2x isoPr); 21.9 (PhꢀCH ); 21.3 (Cꢀ6). PꢀNMR (162 MHz, CDCl3,
C,N
3
30
ACS Paragon Plus Environment