The synthesis of isopropylidene mRNA cap analogs modified with phosphorothioate moiety and their evaluation as promoters of mRNA translation

Article abstract of DOI:10.1016/j.bmcl.2013.05.001

Synthetic mRNA cap analogs are valuable tools in the preparation of modified mRNA transcripts with improved translational activity and increased cellular stability, and have recently attracted more attention because of their great potential in therapeutic applications. We have synthesized and tested isopropylidene dinucleotide cap analogs bearing a phosphorothioate group at the β position of the 5′,5′-triphosphate bridge (two diastereomers of 2′,3′-iPr-m⁷Gpp_SpG), as synthetically simpler alternatives to previously obtained phosphorothioate cap analogs. To evaluate the utility of the new compounds in biological systems we determined their affinity to translation initiation factor 4E (eIF4E), and tested their translational properties in rabbit reticulocyte lysates (RRL) and in human immature dendritic cells (hiDCs). In order to explain the properties of isopropylidene analogs we performed ¹H NMR conformational analysis and correlated the absolute configuration at the β-phosphorous atom with previously synthesized m⁷Gpp_SpG.

Full text of DOI:10.1016/j.bmcl.2013.05.001

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3753–3758
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The synthesis of isopropylidene mRNA cap analogs modified
with phosphorothioate moiety and their evaluation as promoters
of mRNA translation
a
a
b
a
a
Marcin Warminski , Joanna Kowalska , Janina Buck , Joanna Zuberek , Maciej Lukaszewicz ,
b
b,c
b,c
a
a,d,
⇑
Corina Nicola , Andreas N. Kuhn , Ugur Sahin , Edward Darzynkiewicz , Jacek Jemielity
a
Division of Biophysics, Institute of Experimental Physics, Faculty of Physics, University of Warsaw, Zwirki i Wigury 93, 02-089 Warsaw, Poland
Ribological GmbH, Hölderlinstr. 8, 55131 Mainz, Germany
Institute for Translational Oncology and Immunology (TRON), Langenbeck Str. 1, 55131 Mainz, Germany
Centre of New Technologies, University of Warsaw, Zwirki i Wigury 93, 02-089 Warsaw, Poland
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthetic mRNA cap analogs are valuable tools in the preparation of modified mRNA transcripts with
improved translational activity and increased cellular stability, and have recently attracted more attention
because of their great potential in therapeutic applications. We have synthesized and tested isopropyli-
Received 30 March 2013
Revised 29 April 2013
Accepted 1 May 2013
Available online 9 May 2013
0
0
dene dinucleotide cap analogs bearing a phosphorothioate group at the b position of the 5 ,5 -triphosphate
0
0
7
S
bridge (two diastereomers of 2 ,3 -iPr-m Gpp pG), as synthetically simpler alternatives to previously
obtained phosphorothioate cap analogs. To evaluate the utility of the new compounds in biological systems
we determined their affinity to translation initiation factor 4E (eIF4E), and tested their translational prop-
erties in rabbit reticulocyte lysates (RRL) and in human immature dendritic cells (hiDCs). In order to
Keywords:
mRNA cap
Transcription in vitro
Translation efficiency
Phosphorothioate
Nucleotide
1
explain the properties of isopropylidene analogs we performed H NMR conformational analysis and cor-
7
related the absolute configuration at the b-phosphorous atom with previously synthesized m Gpp
S
pG.
Ó 2013 Elsevier Ltd. All rights reserved.
0
The cap present on the 5 end of the messenger RNA (mRNA) is a
unique structure consisting of 7-methylguanosine connected via a
5
immunization through mRNA-encoded antigen delivery into den-
dritic cells. Synthetic antigen-encoding mRNAs bearing a phosp-
7
0
0
,5 -triphosphate bridge to the first nucleotide of the transcript.
horothioate-modified anti-reverse cap analog (b-S-ARCA D1, see
Fig. 1) have proved to be more stable and more efficiently translated
in immature dendritic cells, which has resulted in more pronounced
The cap fulfills a variety of functions during the synthesis and
expression of mRNA, including splicing, intracellular transport,
regulation of stability, initiation of translation and translational
repression via microRNA.1 Thus, synthetic analogs of the mRNA
cap have found a wide application in the study of cap-dependent
8
antigen expression and initiation of immune responses. The first
,2
promising results of such an approach, including embarking on
the first stage of clinical trials on mRNA vaccination against malig-
nant melanoma, have encouraged us to search for new cap analogs,
which would share the same translational characteristics as b-S-
ARCA D1, but might have other beneficial properties, for example,
be synthetically less challenging.
3
processes as well as in biotechnology. One particularly interesting
application is the use of in vitro transcribed mRNAs capped with
synthetic cap analogs for therapeutic purposes, such as gene ther-
apy, anticancer vaccination or stem cell reprogramming.
Optimization of mRNAs intended for therapeutic applications
such as cancer vaccines include several structural RNA elements
one of which is the 5 -cap structure. On the one hand, the cap en-
ables efficient recognition by the translational machinery and
efficient translation and, on the other hand, protects mRNA from
premature degradation by 5 exonucleases. The great potential of
mRNAs containing a chemically modified cap structure has recently
been demonstrated by employing them as tools for anticancer
3
,4
0
Usually, to achieve 5 capping of synthetic mRNAs so-called
anti-reverse cap analogs (ARCA) are added to in vitro transcription
reactions. The use of ARCA, instead of standard cap analogs such as
0
5,6
7
m GpppG, ensures proper incorporation of the dinucleotide into
the mRNA transcript, that is, exclusively in the ‘forward’ and not
‘reverse’ orientation. This results in production of mRNAs with a
higher percentage of properly capped molecules and a higher
0
0
0
overall translational activity. In conventional ARCAs the 3 or 2 hy-
0
0
droxyl of 7-methylguanosine is replaced by a 3 O-methyl or 2 O-
methyl group in order to prevent the RNA polymerase from
initiating the transcription reaction from m Guo moiety.
⇑
Corresponding author. Tel.: +48 225540774; fax: +48 225540771.
E-mail address: jacekj@biogeo.uw.edu.pl (J. Jemielity).
7
9,10
0
960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.05.001

3
754
M. Warminski et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3753–3758
–
tried to verify whether the bulky isopropylidene substituent,
which significantly increases the rigidity of the m Guo ribose moi-
CH
N+
3
O
7
N
ety, influences the stability of the cap complex with eukaryotic
translation initiation factor 4E (eIF4E), and, if so, how this impacts
translation of such modified mRNAs. In order to do so, we deter-
mined the affinities of the new analogs for eIF4E, we introduced
them into mRNA by in vitro transcription and we evaluated the
translational properties of such mRNAs in vitro and in human
immature dendritic cells. Finally, we investigated whether the iso-
propylidene group could be effectively removed at the stage of the
transcript as a step towards transient protection against reverse
incorporation of the cap during in vitro transcription.
HO
OH
–
–
–
O
P
X
P
O
N
O
N
NH
2
O
O
O
P
O
O
H N
2
N
O
N
N
O
O
O
7
O
O
HN
1: iPr-m GpppG, X=O
7
2a: iPr-m Gpp pG (D1), X=S
S
7
–
2b: iPr-m Gpp pG (D2), X=S
CH3
N+
O
S
N
HO
OH
–
–
–
O
P
X
P
O
N
O
N
NH2
Our results suggest that the isopropylidene group, in a manner
as yet not fully explained, impairs the functionality of mRNA cap
O
O
O
O
P
O
O
H
2
N
N
O
N
N
O
O
O
0
resulting in a slightly lower translation yield compared to the 2 -
R1
R2
O
0
HN
or 3 -O methylated counterparts. However, the introduction of
7
1
2
m GpppG, X=O, R =R =H
the phosphorothioate moiety partially compensates for this loss.
Thus, the new analogs 2a and 2b, although constituting a synthet-
ically simpler alternative as mRNA capping reagents, are not ‘per-
fect’ substitutes for b-S-ARCA, since their translational
performance is notably impaired in the systems we investigated.
7,3'-O
1
2
m2
GpppG X=O, R =CH , R =H
3
m27,2'-OGpp pG (D1 and D2), X=S, R =H, R =CH
1
2
S
3
Figure 1. Structures of isopropylidene cap analogs bearing a phosphorothioate
group (2a and 2b) synthesized in this study and various reference cap analogs used
in comparative studies.
0
0
The synthetic route to 2 ,3 -O,O-isopropylidene cap analogs,
0
0
both the phosphate-modified in the 5 ,5 -triphosphate bridge (2a
and 2b) and the unmodified one (1), is depicted in Figure 2. Gener-
ally, the synthesis was accomplished employing P-imidazolide
nucleotide activation. In either case, the less modified nucleotide
was converted into a reactive imidazolide derivative (8 and 9)
However, several other ARCA-type modifications have been devel-
2 -fluoro, 2 -allyl, or LNA-
m Guo. Our particular attention has been drawn by the 2 ,3 -
O,O-isopropylidene ARCA (iPrm GpppG) reported by Kore et al.,
which was interesting due to the low cost of production of 2 ,3 -
0
0
9,10
0
11
0
12
oped, including 2 -/3 -deoxy-,
7
13
0
0
7
0
0
and coupled with appropriate 7-methyl-2 ,3 -O,O-isopropyliden-
2
enucleotide in DMF. The reaction was accelerated by excess ZnCl ,
0
0
O,O-isopropylidene guanosine compared to other ribose-modified
which improves the solubility of reactants, serves as a Lewis acid
catalyst activating imidazole as a leaving group, and acts as a tem-
plate coordinating both nucleotide subunits that form the pyro-
1
4
guanosine analogs. From the chemical point of view, the intro-
duction of an isopropylidene group instead of a methyl group is
0
9,20
significantly simpler, as it enables the laborious separation of 2 -
phosphate bond, thus accelerating a coupling reaction.
0
14
O-Me and 3 -O-Me regioisomers, and also the use of dangerous
Although analog 1 has been synthesized previously, we explored
0
0
reactants such as diazomethane to be avoided – which both are
important aspects when large scale synthesis is considered. The
use of cap analogs that are accessible from inexpensive nucleosides
is a factor that could considerably lower the cost of the large-scale
production of mRNAs for therapeutic purposes.
an alternative route, in which the number of steps involving 2 ,3 -
O,O-isopropylideneguanosine moiety was reduced. The key inter-
mediate, mononucleotide 5, was synthesized from guanosine in
three steps, involving transacetalation with 2,2-dimethoxypropane
2
1
0
22
(DMP), subsequent 5 phosphorylation by POCl
3
,
and selective
7
0
0
0
Due to the susceptibility of mRNA to chemical and enzymatic
degradation, another important improvement of mRNA dedicated
for gene delivery methodologies was the augmentation of its stabil-
ity and translation yield by means of properly modified cap analogs.
One of the main pathways for degradation of functional mRNAs is
N -methylation of the resulting 2 ,3 -isopropylideneguanosine 5 -
monophosphate (4) using methyl iodide in DMSO. The resulting
0
0
0
2 ,3 -O,O-isopropylidene-7-methylguanosine
5 -monophosphate
(5) was then coupled with a 1.1-fold excess of GDP imidazolide
7
derivative (9) yielding iPr-m GpppG (1). The analogs modified at
0
0
0
0
the 5 ?3 exonucleolytic hydrolysis, which occurs after cap removal
the b position of 5 ,5 -triphosphate bridge with phosphorothioate
15
0
0
by a specific pyrophosphatase (e.g., Dcp2). We have previously
shown that blocking cap removal by chemical modification at the
moiety were synthesized by ZnCl
2
-mediated coupling of 2 ,3 -iso-
0
propylidene-7-methylguanosine 5 -O-(2-thiodiphosphate) 7 and
GMP imidazolide 8 (1.5-fold excess). As determined by RP HPLC
analysis, a roughly equimolar mixture of the two P-diastereoiso-
mers of 2 was observed due to the formation of a new stereogenic
center at the b-phosphorus atom. The diastereoisomers were sep-
arated using RP HPLC and designated D1 and D2 according to their
elution order (D1 = faster). Intermediate 7 was efficiently obtained
by converting 5 into its P-imidazolide and subsequent coupling to
b-phosphate group (e.g., by phosphorothioate as in b-S-ARCA, by
phosphoroselenoate or by boranophosphate)1
6,17
is beneficial for
mRNA stability and translation efficiency and that such modified
1
8
mRNAs make better platforms for protein biosynthesis in vivo.
In this study, in order to search for a synthetically simpler equiv-
0
0
alent of b-S-ARCA, we explored how a combination of the 2 ,3 -iso-
propylidene substituent at the ribose of 7-methylguanosine and the
phosphorothioate modification at the b-position of the triphosphate
bridge influences the biological properties of the cap itself and of
3ꢀ
23
thiophosphate (PSO
3
) triethylammonium salt. The structure
and homogeneity of newly synthesized cap analogs have been con-
1
8,19
1
31
capped mRNAs.
We have synthesized two diastereomers of
firmed by RP HPLC, HRMS, H and P NMR (for details see the Sup-
plementary data).
7
mRNA cap analog, iPrm Gpp
S
pG (D1 and D2; Fig. 1), bearing the
1
two aforementioned modifications, studied their physicochemical
and biological properties and compared them to several previously
developed compounds, including b-S-ARCA diastereomers
We analyzed the fine structures of the H NMR spectra of 1, 2a,
2b and two mononucleotides (5 and 7), in order to determine the
0
0
conformations of their 2 ,3 -O,O-isopropylideneribose rings. Scalar
0
0
7
,2 -O
7,2 -O
0
0
7
0
0
(
m
m
2
7
Gpp
S
pG D1 and D2), standard ARCAs (m
2
GpppG and
coupling constants between H1 –H2 , H2 –H3 and (if they could
0
,3 -O
7
0
0
2
GpppG), and iPr-m GpppG (1) which was obtained by a syn-
be observed) H3 –H4 protons of m iPrG moiety were measured
and hence the pseudorotation angle P and puckering amplitude
thetic route different from that reported earlier.
We evaluated the new analogs as reagents for the preparation
of capped mRNAs dedicated to high-yield protein expression. We
Um were calculated, using the empirically generalized Karplus
equation introduced by Haasnoot et al. According to Röder
2
4

M. Warminski et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3753–3758
3755
HO OH
HO OH
–
–
–
–
O
O
P
O
O
P
O
O
O
P O
O
N
O
O P O–
O
O
N
H N
N
O
O
N
2
N
N
H
2
N
N
O
(
i)
HN
9
HN
N
8
6%
–
(ii)
CH
N
3
O
N
+
4
6%
N
–
HO OH
CH
N
O
–
O–
P
O
–
O
3
O
O
O
+
N
O
NH
2
N
O
P O
O
O P
O
N
NH
NH2
N
O
N
NH
–
H
2
N
N
O
–
O
O
O P
O
N
N
N
NH2
N
4
–
O
O
O
O
N
–
O
O
P O
O
HN
N
NH
O
O
2
N
HO
1
O
(
iii)
(iv)
5
O
O
O
3
O
O
8
6%
65%
(
i)
CH
–
–
3
O
CH
N
3
O
N
S
P
O
5
7%
N+
–
OH/ TEAH+
+
O
NH
N
–
–
O–
P O
O
O
N P
O
S
O P
N
N
N
O
N
6
NH
–
NH
2
2
O
O
O
(
ii)
O
–
CH
N
3
O
7
O
O
42%
O
O
(
ii)
+
N
3
5%
HO OH
–
S–
P
O
–
O
O
N
O
N
NH
2
O
P O
O
O P
O
O
O
N
H N
N
2
HO OH
HO OH
–
–
O
P
O
O
P
O
O
O
HN
–
N
O
O
O
N
2a and b
O
N
H N
N
O
N
N
H
2
N
N
O
2
O
HN
HN
8
N
(i)
90%
N
O
0
Figure 2. Synthesis of isopropylidene dinucleotide cap analogs 1 and 2. Reagents: (i) imidazole, 2,2 -DTDP, PPh
3
, TEA, DMF, rt; (ii) ZnCl
2
, DMF, rt; (iii) (1) POCl
3
, (CH
3
O)
3
PO,
0
°C; (2) H
2
O, TEAB (iv) MeI, DMSO, rt.
et al.,25 pseudorotation of 2 ,3 -O,O-isopropylideneguanosine is
0
0
at 337 nm (excitation at 280 nm).26 As the concentration of the cap
analog increases, the fluorescence emission intensity decreases due
to the quenching of intrinsic Trp fluorescence in the eIF4E-cap com-
plex. After reaching its plateau, the signal increases again on ac-
count of the natural fluorescence of the unbound cap analog. The
association constants KAS for 1, 2a and 2b derived from the titration
inhibited and its conformation is fixed. Thus, only two parameters
need to be derived (P and
Um) from measurements. The values de-
0
0
rived from spectra of the five compounds containing 2 ,3 -O,O-iso-
propylideneribose moiety (three cap analogs and two
mononucleotides) are consistent and close to each other. The
pseudorotation angle P varies from 298° to 306° (including 300°
for all three cap analogs), while puckering amplitudes
in the range of 53–60°. These values correspond to E (C1 -endo)
curves are presented in Table 1, together with the same data for two
0
7
,2 -O
7
U
m
remain
diastereoisomers of m
2
S
Gpp pG and unmodified m GpppG cap
1
0
0
0
as a reference. It appears that the presence of 2 ,3 -O,O-isopropyli-
0
conformation, which is different from the 3 C-endo conformation
usually adopted by the N -methylguanosine nucleotides in com-
dene moiety reduces the binding affinity for eIF4E by 3.5-fold com-
7
7
pared to the unmodified parent compound (m GpppG) or
2
6,27
7
plex with eIF4E.
The observed conformational constraints in
conventional ARCA. The decreased affinity of iPr-m GpppG may
1
and 2 may influence both the efficiency of stacking interactions
be at least partially explained as a result of the sterical hindrance
caused by the isopropylidene group. Although the crystal structures
of 7-methylguanosine with two Trp residues and the alignment
of the triphopshate linkage in eIF4E’s cap binding pocket.
7
7
of eIF4E-m GpppG and eIF4E-m GpppA complexes revealed that
0
0
Biochemical properties of novel synthesized cap analogs can be
predicted with a fair degree of accuracy by determination of the
binding affinity for eIF4E protein. This relatively simple and reliable
biophysical test is carried out by titration of eIF4E with a solution of
the cap analog and measuring the quenching of protein fluorescence
the 2 - and 3 -hydroxyls are not buried in the cap-binding cavity,
but rather exposed to the solution and do not contribute signifi-
2
6,27
cantly to complex stabilization,
the introduction of a relatively
bulky isopropylidene group, which rigidifies and changes confor-
mation of the 7-methylguanosine sugar ring, may still destabilize
Table 1
Equilibrium association constants (KAS) of cap analog – murine eIF4E complex, as determined by fluorescence quenching
Compounds
Abbreviation
K
AS
[l
M
ꢀ1]
Translational efficiency in vitro
7
m GpppG
9.4 ± 0.4
1.00 ± 0,02
1.32 ± 0.04
ND
1.91 ± 0.01
0.80 ± 0.02
1.33 ± 0.01
1.16 ± 0.00
0.15 ± 0.00
0
7
7
7
,3 -O
GpppG10
m
2
m
2
m
2
10.2 ± 0.3
43.1 ± 1.4
19.3 ± 2.2
2.9 ± 0.1
5.6 ± 0.1
9.2 ± 0.2
0
,2 -O
19
19
Gpp
Gpp
S
pG (D1)
pG (D2)
0
,2 -O
S
7
1
iPr-m GpppG
iPr-m Gpp
iPr-m Gpp
ApppG
7
2
2
a
b
S
pG (D1)
pG (D2)
7
S
2
6
0.0095 ± 0.0007
ND, not determined.

3
756
M. Warminski et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3753–3758
7
7
S S
Figure 3. Conversion of an unequimolar mixture of iPrm Gpp pG diastereomers into m Gpp pG diastereomers. The isopropylidene group removal was carried out in aqueous
7
7
buffer at pH 2 for 4.5 h and 37 °C. The reaction progress monitored by RP HPLC revealed that iPr-m Gpp
diastereomers have the same absolute configuration around the stereogenic b-phosphorus atom.
S S
pG D1 is converted into m Gpp pG D1, which indicates that the two
the complex. Additionally, the conformational constraints caused
by the isopropylidene group may also contribute to this effect.
The introduction of phosphorothioate modification at the b po-
catalyzed by SP6 polymerase. First, we synthesized short (5 nt)
capped transcripts and analyzed them by gel electrophoresis. We
found that the efficiency of incorporation of 2a and 2b into the
sition results in an increase in KAS value by ca. 2- and 3-fold for D1
transcripts was comparable to the reference compounds, that is,
0
7
7,3 -O
7
and D2 diastereoisomers of iPr-m Gpp
S
pG, respectively, which is in
m
2
GpppG (ARCA), m GpppG and 1 (data not shown). Next,
general agreement with the effect of phosphorothioate modifica-
tion on the cap-eIF4E complex reported earlier. As a result, the
D2 isomer (2b) binds to eIF4E with an association constant that
approximates that for m GpppG, whereas the two other analogs
display a lower affinity (Fig. 4). The fact that the KAS value for the
we transcribed full-length luciferase-encoding mRNAs capped
0
7
7,3 -O
with either 1, 2a, 2b, m GpppG or m
2
GpppG and determined
their translational efficiencies in rabbit reticulocyte lysate. In this
experiment, different concentrations of mRNA were added to the
lysate (which was optimized for cap-dependent translation) and
the luciferase activity was measured by luminometry 60 min after
the addition of mRNA. The results of a representative experiment
are depicted in Figure S1, and average translation efficiency values
from replicate experiments, calculated from the slopes of the lines
representing luciferase activity versus mRNA concentration, are
summarized in Table 1. The determined translation efficiencies
7
D2 isomer is higher than for D1 is in contrast with previous results
7
ꢀ1
7,2 -O
for diastereoisomers of m Gpp
S
pG (KAS = 45.0 ± 1.1
l
2
M
for D1
Gpp pG
S
0
ꢀ1
and KAS = 23.0 ± 0.4
l
M
for D2, respectively) or m
Table 2). One possible explanation for this fact is that the pres-
1
9
(
ence of an isopropylidene group, which rigidifies the 7-methylgu-
7
anosine ribose moiety, reverses the elution order of iPr-m Gpp
S
pG
7
7
diastereoisomers during RP HPLC (related to m Gpp
S
pG and
were normalized to the mRNA capped with m GpppG, for which
0
,2 -O
7
m
2
Gpp
To test this, we performed an isopropylidene group removal
reaction (Fig. 3) under acidic conditions to convert an iPr-
S
pG analogs).
translation efficiency was set to one. mRNA with a non-functional
cap analog (ApppG) was used to estimate the level of cap-indepen-
dent translation. Generally, the presence of an isopropylidene
group decreased the relative translation efficiency of mRNA in
7
7
m Gpp
S S
pG D1/D2 mixture to an m Gpp pG D1/D2 mixture. Sur-
0
prisingly, the reaction progress monitoring by RP HPLC revealed
RRL by around 40% compared to the corresponding 3 -O-Me cap
7
7
that iPr-m Gpp
S
pG D1 was converted into m Gpp
S
pG D1, not D2
analog. For example, the relative translation efficiency for iPr-
7
(
Fig. 3). This means that the reversed order in the binding affinities
m GpppG-mRNA was 0.80 ± 0.02, compared to 1.32 ± 0.04 for
0
7
7
7,3 -O
7
for iPr-m Gpp
S
pG D1 and D2 compared to m Gpp
S
pG D1 and D2 is
m
2
GpppG-mRNA. Both stereoisomers of iPr-m Gpp
S
pG were
7
not a result of opposite configurations at the b-phosphorous atoms.
translated more efficiently than iPr-m GpppG, which is generally
Hence, another possible explanation is that this observation may
consistent with the higher affinity of these two analogs for eIF4E.
0
0
7
be a result of the different alignment of the 5 ,5 -triphosphate chain
in the cap-binding site of eIF4E, which is enforced by the confor-
Interestingly, iPr-m GpppG D1 was translated slightly more effi-
7
ciently than iPr-m GpppG D2 (relative tr. eff. values 1.33 and 1.16,
0
0
mation of 2 ,3 -isopropylidene 7-methylguanosine moiety, which
is different from the conformation of unmodified 7-methylguano-
sine (see the previous section).
respectively). This correlation was in contrast to the correlation
Next, we tested the new cap analogs 2a and 2b as substrates for
the preparation of mRNAs by in vitro transcription reaction
Table 2
0
Effect of differentially 5 -capped mRNAs on luciferase expression in hiDCs. Analysis of
the data shown in Figure 5 by mathematical modeling gives calculated values for
translational efficiency, the time point of maximal protein expression and overall
protein expression induced by the respective capped mRNA
0
5
-Cap structure on
Translational
efficiency
Time of max
(h)
Total protein
expression
*
⁄
mRNA
0
7
7
7
,3 -O
m
m
m
2
2
2
GpppG
1.00 ± 0.07
2.07 ± 0.21
1.85 ± 0.25
1.23 ± 0.09
9.25 ± 0.49
13.65 ± 0.07
8.75 ± 0.07
11.10 ± 0.14
1.00 ± 0.17
3.95 ± 0.38
1.65 ± 0.25
1.62 ± 0.13
0
,2 -O
Gpp
Gpp
S
pG (D1)
pG (D2)
0
,2 -O
S
7
iPr-m Gpp
S
pG (D1)
(
2a)
7
iPr-m Gpp
2b)
S
pG (D2)
1.08 ± 0.06
8.90 ± 0.28
0.97 ± 0.10
Figure 4. Comparison of equilibrium association constants (KAS) for complexes of
(
7
7
S
murine eIF4E with isopropylidene cap analogs 1 (iPr-m GpppG), 2a (iPr-m Gpp pG
*
7,3’-OGpppG–capped mRNA.
7
Values are given relative to results for m
2
S
D1) 2b (iPr-m Gpp pG D2) and their ribose-unmodified counterparts.

M. Warminski et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3753–3758
3757
0
observed previously for m 7
,2 -O
Gpp
pG D1 and D2 and other b-
2
S
0
0
7,2 -O
7
,2 -O
phosphate-modified cap analogs (m
2
GppSepG,
m
2
1
6–18
GppBH3pG).
In the case of all these analogs, when tested in rab-
bit reticulocyte lysates or mammalian epithelial cells (but not in
dendritic cells; see also the following paragraph), the D2 stereoiso-
mer was the one conferring higher translation efficiency to mRNA.
7
In the case of iPr-m Gpp
S
pG, the mRNAs capped with D1 diastereo-
isomer were translationally more active (see Table 1). A detailed
explanation of this finding is difficult. However, we assume it may
be a result of the different alignment of the phosphate chain in
the eIF4E cap-binding pocket, enforced by the conformational rigid-
ity of isopropylidene-7-methylguanosine.
Lastly, we also addressed the question whether the short RNA
transcripts may persist under conditions required for the isopro-
pylidene group removal from dinucleotide cap (Fig. 3). We have
demonstrated that 5 nt RNA transcripts capped with 1 or 2 do
not undergo degradation after incubation at pH 2, 37° for 4.5 h
(Fig. S2). This finding opens possibility for employing isopropyli-
dene group as a transient cap protection which can be removed
after in vitro transcription reaction. We are currently investigating
this issue more deeply.
0
Figure 5. Effect of differentially 5 -capped mRNAs on luciferase expression in
hiDCs. Luciferase activity was measured for 72 h following electroporation of mRNA
including the respective 5 -cap into hiDCs. Duplicate measurements were per-
0
In order to analyze the individual effects of the isopropylidene
phosphorothioate cap analogs on RNA stability and protein trans-
lation in antigen-presenting cells, mRNAs transcribed in vitro by
T7 RNA polymerase, coding for luciferase and including the respec-
tive cap analogs 2a or 2b, were electroporated into human imma-
ture dendritic cells (hiDCs). RNAs containing the cap dinucleotides
formed and the averaged bioluminescence signal is shown as a function of time.
substituent confers an additional stabilization effect to mRNA,
but we were unable to observe this in our experiments.
0
0
0
7,2 -O
7
,3 -O
7,2 -O
m
2
GpppG, m
2
Gpp
S
pG (D1) and m
2
S
Gpp pG (D2),
In conclusion, we report two diastereomeric mRNA cap analogs
bearing a phosphorothioate group at the b-position of the triphos-
phate bridge and 2’,3’-O,O-isopropylidene at the ribose of 7-
methylguanosine, which we considered synthetically simpler
alternatives to previously developed phosphorothioate anti-
reverse cap analogs. The biological properties of these new com-
pounds indicate that the isopropylidene substituent compromises
the translational properties of the cap to some extent, but this ef-
fect may be partially compensated for by introduction of the
phosphorothioate moiety. The results reported here may be useful
for future developments of mRNA cap analogs with properties ben-
eficial for applications such as mRNA-based gene delivery or mRNA
vaccination.
which were previously analyzed in a comparable experimental
8
set-up, were used as references for evaluation of discrete cap-
dependent properties. As was the case for SP6 polymerase cata-
lyzed reactions, no significant differences between RNAs including
different cap analogs could be observed either for in vitro tran-
scription yield or for RNA integrity. Protein expression in hiDCs
post-electroporation was followed by measurement of the lucifer-
ase activity for a time period of 72 h (Fig. 5) and further analyzed
by mathematical modeling to yield values for translational effi-
ciency (given by the initial slope), the time point of maximal pro-
tein expression (indicative for RNA stability) and the total protein
expression (Table 2). The activity of the reporter gene product for
0
0
7,2 -O
7,3 -O
the reference RNAs capped with m
2
GpppG, m
pG (D2) in hiDCs are as previously seen.
An up-to-two-fold increase in translational efficiency can be ob-
2
S
Gpp pG
7
,2’-O
8
(
D1) and m
2
Gpp
S
Acknowledgments
0
,2 -O
7
served for m
2
Gpp
S
pG cap analogs compared to the assigned
GpppG, as well as an increase in total protein
expression. In line with former findings, the co-transcriptional
This study was supported by the Polish Ministry of Science and
Higher Education (N N204 089438, N N301 096339), the National
Science Centre, Poland (UMO-2012/05/E/ST5/03893), and by fund-
ing from the Innotop program of the State of Rhineland-Palatinate,
Germany. The authors are grateful to the Laboratory of Biological
NMR (Institute of Biochemistry and Biophysics of the Polish Acad-
emy of Sciences, IBB PAS) for access to their NMR apparatus, and
Jacek Oledzki from the Laboratory of Mass Spectrometry (IBB
PAS) for recording MS spectra.
0
,3 -O
7
reference cap m
2
0
,2 -O
7
insertion of the cap diastereomer m
2
S
Gpp pG (D1) results in a
larger quantity of reporter gene product and a higher stability of
the respective mRNA in hiDCs in comparison to the D2 diastereo-
mer (Fig. 5). For the isopropylidene phosphorothioate cap analogs,
such an overall behavior of the two respective diastereomers D1
and D2 can also be observed: the calculated ratio of values for
translational efficiency of D1 versus D2 is ꢁ1.1 in both cases and
total protein expression differs by
a
factor of ꢁ2.4 for
Supplementary data
0
,2 -O
7
m
2
Gpp
S
pG D1/D2 caps and ꢁ1.7 for isopropylidene phosp-
horothioate cap analogs with enhanced long-term stability of
mRNAs capped with the D1-diastereomer, respectively (Table 2).
Although the ratios of protein expression profiles from D1-to-D2
diastereomers of caps are comparable, indicating consistent effects
of phosphorothioate functionality, the absolute values of luciferase
activity in hiDCs for the investigated cap analogs differ. Luciferase
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.05.
0
01. These data include MOL files and InChiKeys of the most
important compounds described in this article.
References and notes
activity values of isopropylidene phosphorothioate capped mRNAs
0
7
,2 -O
are not as pronounced as for the m
2
Gpp
S
pG variants, again
1. Furuichi, Y.; Shatkin, A. J. Adv. Virus Res 2000, 55, 135.
2
.
Zdanowicz, A.; Thermann, R.; Kowalska, J.; Jemielity, J.; Duncan, K.; Preiss, T.;
Darzynkiewicz, E.; Hentze, M. W. Mol. Cell 2009, 35, 881.
Jemielity, J.; Kowalska, J.; Rydzik, A. M.; Darzynkiewicz, E. New J. Chem. 2010,
34, 829.
suggesting a role for constitutional or conformational differences
between the two cap di-nucleotide series. The previous study by
3.
1
4
0
0
Kore et al.,
suggested that the 2 ,3 -O,O-isopropylidene

3
758
M. Warminski et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3753–3758
4
.
.
.
.
.
.
Grudzien-Nogalska, E.; Stepinski, J.; Jemielity, J.; Zuberek, J.; Stolarski, R.;
Rhoads, R. E.; Darzynkiewicz, E. Methods Enzymol. 2007, 431, 203.
Kreiter, S.; Diken, M.; Selmi, A.; Tureci, O.; Sahin, U. Curr. Opin. Immunol. 2011,
17. Su, W.; Slepenkov, S.; Grudzien-Nogalska, E.; Kowalska, J.; Kulis, M.; Zuberek,
J.; Lukaszewicz, M.; Darzynkiewicz, E.; Jemielity, J.; Rhoads, R. E. RNA 2011, 17,
978.
18. Grudzien-Nogalska, E.; Jemielity, J.; Kowalska, J.; Darzynkiewicz, E.; Rhoads, R.
E. RNA 2007, 13, 1745.
19. Kowalska, J.; Lewdorowicz, M.; Zuberek, J.; Grudzien-Nogalska, E.; Bojarska, E.;
Stepinski, J.; Rhoads, R. E.; Darzynkiewicz, E.; Davis, R. E.; Jemielity, J. RNA
2008, 14, 1119.
20. Kadokura, M.; Wada, T.; Urashima, C.; Sekine, M. Tetrahedron Lett. 1997, 38,
8359.
5
6
7
8
9
2
3, 399.
Kuhn, A. N.; Diken, M.; Kreiter, S.; Vallazza, B.; Tureci, O.; Sahin, U. RNA Biol.
011, 8, 35.
2
Kuhn, A. N.; Beissert, T.; Simon, P.; Vallazza, B.; Buck, J.; Davies, B. P.; Tureci, O.;
Sahin, U. Curr. Gene Ther. 2012, 12, 347.
Kuhn, A. N.; Diken, M.; Kreiter, S.; Selmi, A.; Kowalska, J.; Jemielity, J.;
Darzynkiewicz, E.; Huber, C.; Tureci, O.; Sahin, U. Gene Ther. 2010, 17, 961.
Stepinski, J.; Waddell, C.; Stolarski, R.; Darzynkiewicz, E.; Rhoads, R. E. RNA
21. Hampton, A. J. Am. Chem. Soc. 1961, 83, 3640.
2
001, 7, 1486.
22. Yoshikawa, M.; Kato, T.; Takenishi, T. Tetrahedron Lett. 1967, 5065.
23. Kowalska, J.; Lewdorowicz, M.; Darzynkiewicz, E.; Jemielity, J. Tetrahedron Lett.
2007, 48, 5475.
1
1
0. Jemielity, J.; Fowler, T.; Zuberek, J.; Stepinski, J.; Lewdorowicz, M.;
Niedzwiecka, A.; Stolarski, R.; Darzynkiewicz, E.; Rhoads, R. E. RNA 2003, 9,
1
108.
24. Haasnoot, C. A. G.; Deleeuw, F. A. A. M.; Deleeuw, H. P. M.; Altona, C. Org. Magn.
Reson. 1981, 15, 43.
1. Kore, A. R.; Shanmugasundaram, M.; Charles, I.; Cheng, A. M.; Barta, T. J. Bioorg.
Med. Chem. Lett. 2007, 17, 5295.
2. Kore, A. R.; Charles, I. Bioorg. Med. Chem. 2010, 18, 8061.
3. Kore, A. R.; Hodeib, M.; Hu, Z. T. Nucleos. Nucleot. Nucl. 2008, 27, 1.
4. Kore, A. R.; Shanmugasundaram, M.; Vlassov, A. V. Bioorg. Med. Chem. Lett.
25. Roder, O.; Ludemann, H. D.; Vongoldammer, E. Eur. J. Biochem. 1975, 53, 517.
26. Niedzwiecka, A.; Marcotrigiano, J.; Stepinski, J.; Jankowska-Anyszka, M.;
Wyslouch-Cieszynska, A.; Dadlez, M.; Gingras, A. C.; Mak, P.; Darzynkiewicz,
E.; Sonenberg, N.; Burley, S. K.; Stolarski, R. J. Mol. Biol. 2002, 319, 615.
27. Tomoo, K.; Shen, X.; Okabe, K.; Nozoe, Y.; Fukuhara, S.; Morino, S.; Sasaki, M.;
Taniguchi, T.; Miyagawa, H.; Kitamura, K.; Miura, K.; Ishida, T. J. Mol. Biol. 2003,
328, 365.
1
1
1
2
008, 18, 4828.
5. van Dijk, E.; Cougot, N.; Meyer, S.; Babajko, S.; Wahle, E.; Seraphin, B. EMBO J.
002, 21, 6915.
1
1
2
6. Kowalska, J.; Lukaszewicz, M.; Zuberek, J.; Darzynkiewicz, E.; Jemielity, J.
Chembiochem 2009, 10, 2469.