5′-Phosphorothiolate Dinucleotide Cap Analogues: Reagents for Messenger RNA Modification and Potent Small-Molecular Inhibitors of Decapping Enzymes

Article abstract of DOI:10.1021/jacs.8b02597

The 5′ cap consists of 7-methylguanosine (m⁷G) linked by a 5′-5′-triphosphate bridge to messenger RNA (mRNA) and acts as the master regulator of mRNA turnover and translation initiation in eukaryotes. Cap analogues that influence mRNA translation and turnover (either as small molecules or as part of an RNA transcript) are valuable tools for studying gene expression, which is often also of therapeutic relevance. Here, we synthesized a series of 15 dinucleotide cap (m⁷GpppG) analogues containing a 5′-phosphorothiolate (5′-PSL) moiety (i.e., an O-to-S substitution within the 5′-phosphoester) and studied their biological properties in the context of three major cap-binding proteins: translation initiation factor 4E (eIF4E) and two decapping enzymes, DcpS and Dcp2. While the 5′-PSL moiety was neutral or slightly stabilizing for cap interactions with eIF4E, it significantly influenced susceptibility to decapping. Replacing the γ-phosphoester with the 5′-PSL moiety (γ-PSL) prevented β-γ-pyrophosphate bond cleavage by DcpS and conferred strong inhibitory properties. Combining the γ-PSL moiety with α-PSL and β-phosphorothioate (PS) moiety afforded first cap-derived hDcpS inhibitor with low nanomolar potency. Susceptibility to Dcp2 and translational properties were studied after incorporation of the new analogues into mRNA transcripts by RNA polymerase. Transcripts containing the γ-PSL moiety were resistant to cleavage by Dcp2. Surprisingly, superior translational properties were observed for mRNAs containing the α-PSL moiety, which were Dcp2-susceptible. The overall protein expression measured in HeLa cells for this mRNA was comparable to mRNA capped with the translation augmenting β-PS analogue reported previously. Overall, our study highlights 5′-PSL as a synthetically accessible cap modification, which, depending on the substitution site, can either reduce susceptibility to decapping or confer superior translational properties on the mRNA. The 5′-PSL-analogues may find application as reagents for the preparation of efficiently expressed mRNA or for investigation of the role of decapping enzymes in mRNA processing or neuromuscular disorders associated with decapping.

Full text of DOI:10.1021/jacs.8b02597

Subscriber access provided by Kaohsiung Medical University
5#-Phosphorothiolate dinucleotide cap analogs: reagents for mRNA
modification and potent small molecular inhibitors of decapping enzymes
Blazej Andrzej Wojtczak, Pawel J. Sikorski, Kaja Fac-Dabrowska, Anna Nowicka, Marcin Warminski,
Dorota Kubacka, Elzbieta Nowak, Marcin Nowotny, Joanna Kowalska, and Jacek Jemielity
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 20 Apr 2018
Downloaded from http://pubs.acs.org on April 20, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
5′-Phosphorothiolate dinucleotide cap analogs: reagents for mRNA
modification and potent small molecular inhibitors of decapping en-
zymes
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
Blazej A. Wojtczak†,‡, Pawel J. Sikorski†,‡, Kaja Fac-Dabrowska†, Anna Nowicka¶, Marcin Warminski¶,
Dorota Kubacka¶, Elzbieta Nowak§, Marcin Nowotny§, Joanna Kowalska*,¶, and Jacek Jemielity*,†
†Centre of New Technologies, University of Warsaw, Banacha 2c St, 02-097 Warsaw, Poland
¶Division of Biophysics, Institute of Experimental Physics, Faculty of Physics, University of Warsaw, Pasteura 5 St, 02-093
Warsaw, Poland
§International Institute of Molecular and Cell Biology in Warsaw, 4 Ks. Trojdena St, 02-109 Warsaw, Poland
ABSTRACT: The 5′ cap consists of 7-methylguanosine (m7G) linked by a 5′-5′-triphosphate bridge to mRNA and acts as the master
regulator of mRNA turnover and translation initiation in eukaryotes. Cap analogs that influence mRNA translation and turnover
(either as small molecules or as part of an RNA transcript) are valuable tools for studying gene expression, which is often also of
therapeutic relevance. Here, we synthesized a series of 15 dinucleotide cap (m7GpppG) analogs containing a 5′-phosphorothiolate
(5′-PSL) moiety (i.e., an O-to-S substitution within the 5′-phosphoester) and studied their biological properties in the context of three
major cap-binding proteins – translation initiation factor 4E (eIF4E) and two decapping enzymes, DcpS and Dcp2. While the 5′-PSL
moiety was neutral or slightly stabilizing for cap interactions with eIF4E, it significantly influenced susceptibility to decapping.
Replacing the γ-phosphoester with the 5′-PSL moiety (γ-PSL) prevented β-γ-pyrophosphate bond cleavage by DcpS and conferred
strong inhibitory properties. Combining the γ-PSL moiety with α-PSL and β-phosphorothioate (PS) moiety afforded first cap-derived
hDcpS inhibitor with low nanomolar potency. Susceptibility to Dcp2 and translational properties were studied after incorporation of
the new analogs into mRNA transcripts by RNA polymerase. Transcripts containing the γ-PSL moiety were resistant to cleavage by
Dcp2. Surprisingly, superior translational properties were observed for mRNAs containing the α-PSL moiety, which were Dcp2-
susceptible. The overall protein expression measured in HeLa cells for this mRNA was comparable to mRNA capped with the trans-
lation augmenting β-PS analog reported previously. Overall, our study highlights 5′-PSL as a synthetically accessible cap modifica-
tion, which, depending on the substitution site, can either reduce susceptibility to decapping or confer superior translational properties
on the mRNA. The 5′-PSL-analogs may find application as reagents for the preparation of efficiently expressed mRNA or investiga-
tion of the role of decapping enzymes in mRNA processing or neuromuscular disorders associated with decapping.
INTRODUCTION
the messenger RNA (mRNA) 5′ cap structure, known as decap-
ping, is a key step in the regulation of mRNA stability and turn-
over.3, 4 The 5′ cap, composed of 7-methylguanosine (m7G)
linked by a 5′,5′-triphosphate bridge to the first transcribed nu-
cleotide, is the hallmark of the 5′ end of an mRNA molecule
and is recognized by several specialized proteins involved in
mRNA processing, transport, translation, and degradation (Fig-
ure 1).5-7 There are two major decapping enzymes engaged in
bulk mRNA degradation, DcpS and Dcp1-Dcp2. Interestingly,
while these two enzymes differ in substrate specificity, partici-
pate in different mRNA decay pathways, and cleave the mRNA
cap triphosphate bridge at different sites, they ultimately lead to
a similar outcome: products of decapping and degradation of
the mRNA (Figure 1A).
Phosphate and oligophosphate mono- and diesters are key struc-
tural elements of many biomolecules including nucleotides and
nucleic acids. The enzymes governing phosphoester and phos-
phoanhydride bond cleavage within DNA and RNA participate
in gene expression, cellular signaling, and metabolism.1, 2 Con-
sequently, phosphate ester and anhydride analogs have been in-
valuable tools in deciphering the mechanisms of biophosphate-
related enzymatic transformations.2 Unfortunately, many of the
analogs used to study these enzymes are complex, difficult to
synthesize, or have biocompatibility issues, thus restricting
their use and limiting our understanding of enzymatic function
during various processes, including RNA processing.
Phosphohydrolytic reactions play an important role in RNA bi-
ology. Indeed, cleavage of the phosphoanhydride bonds within
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Structure, function, and applications of the mRNA 5′ cap and its chemically modified analogs. (A) Schematic structure of mRNA
and its 5′ end. Sites for triphosphate bridge hydrolysis by major decapping enzymes are denoted with scissors. (B) The involvement of cap-
recognizing proteins, eIF4E, DcpS, and Dcp2, in mRNA translation and degradation. (C) Structure of β-S-ARCA (m27,2′-OGppSpG), a previ-
ously reported phosphorothioate cap analog existing as two P-diastereomers, both conferring superior translational properties to mRNA. (D)
Structure of the 5′-phosphorothiolate moiety, a non-diastereogenic modification incorporated into the mRNA cap structure in this study. (E)
The concept of immune system programming relying on transfection of immature dendritic cells (iDCs) with synthetic 5′-capped mRNAs
followed by DC maturation.
DcpS, a member of the histidine triad (HIT) superfamily of py-
rophosphatases, acts as a scavenger, degrading 7-methylgua-
nine nucleotides arising from 3′-to-5′ mRNA degradation by the
exosome, presumably to prevent their undesired accumulation
in the cell.8 DcpS cleaves the cap between the γ and β phos-
phates, releasing 7-methylguanosine monophosphate (m7GMP)
and a nucleoside or oligoribonucleotide 5′-diphosphate (Figure
1A,B). Moreover, DcpS has also been suggested to play a more
general role in gene expression, particularly in regulation of
mRNA splicing, transcript-specific degradation, and mi-
croRNA stability.9, 10 DcpS has also been identified as molecu-
to-S, γ-O-to-BH3, β-O-to-BH3).16-19 All these compounds inhib-
ited DcpS in vitro, although at concentrations at least 20-fold
higher than RG3039, therefore their therapeutic potential is ra-
ther limited.20
The other major decapping enzyme, Dcp2 belongs to the nucle-
oside diphosphate linked to another moiety X (Nudix) hydro-
lase superfamily and is a catalytic subunit of the heterodimeric
Dcp1-Dcp2 complex. Dcp2 cleaves 5′ capped mRNA between
the α and β phosphates, releasing 7-methylguanosine diphos-
phate (m7GDP) and a 5′-monophosphorylated mRNA that is ex-
posed to 5′-exonucleolytic degradation (Figure 1A,B).21-24
mRNA decapping by Dcp2 is the first irreversible step in 5′-to-
3′ mRNA decay, thereby contributing to the regulation of
mRNA fate. Consequently, Dcp2-resistant cap structures pro-
long mRNA half-life in vivo,25-27 and, if also characterized by
high affinity for translation initiation factor 4E (eIF4E), confer
superior translational properties to the mRNA.18, 27 One of the
Dcp2-resistant cap analogs with the most promising properties
is β-S-ARCA (Figure 1C). In β-S-ARCA, a single O-to-S sub-
stitution at the β-phosphate induces a longer cellular mRNA
half-life and higher translation efficiency.27 Such mRNAs with
enhanced properties can benefit therapeutic applications relying
lar target for therapy of spinal muscular atrophy (SMA),11
a
neuromuscular disorder caused by mutations in the survival mo-
tor neuron (SMN 1) gene.12 Notably, a linear correlation was
found between the effectiveness of quinazoline derivatives to
inhibit DcpS and their ability to improve motor function in a
mouse SMA model.13-15 One of the most potent DcpS inhibitors
– quinazoline derivative RG3039 – recently entered clinical tri-
als, unfortunately without success. Phosphate-modified cap an-
alogs are another class of DcpS-inhibitors with applications in
research and medicine. Among previously reported, DcpS-re-
sistant cap analogs were those carrying bridging β-γ modifica-
tions (O-to-CH2, O-to-NH) or non-bridging modifications (γ-O-
2
ACS Paragon Plus Environment

Page 3 of 26
Journal of the American Chemical Society
as well as their ability to inhibit decapping or enhance mRNA
on inducing protein expression in vivo, such as delivery of an-
1
2
3
4
5
6
7
tigens eliciting anti-cancer immune response (currently under
clinical trials; Figure 1E).28, 29 Beyond its role in bulk mRNA
degradation, Dcp2 activity has also been recently linked to a
reduced type I interferon antiviral immune response12 and found
to act as a quality control enzyme during small nuclear RNA
assembly.30 These findings open new directions for Dcp2-re-
lated research.
translational potential.
Chart 1. Structures of the cap analogs synthesized in this
study
Despite the significant number of mRNA cap analogs devel-
oped to study cap-dependent processes, our understanding of
the multifaceted functions of DcpS and Dcp2 is still limited.
Furthermore, new decapping enzymes are also being continu-
ously discovered, creating the space for new molecules to be
designed to study them.4, 31 Of particular interest in this context,
are analogs that combine unique, beneficial biological proper-
ties with structural simplicity and synthetic availability, making
them effective as well as easy to manufacture.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In this study, we explored a new type of chemical modification
in the mRNA cap, namely a substitution of the 5′-O atom to an
S atom in the triphosphate chain, resulting in the formal replace-
ment of the 5′-phosphoester with a phosphorothiolate (5′-PSL)
moiety (Figure 1D, Chart 1). We synthesized and biochemically
characterized a set of 15 m7GpppG-derived, dinucleotide cap
analogs containing a 5′-PSL moiety (compounds 1–15, Chart
1). The analogs primarily differ in the position of the PSL moi-
ety (next to the m7G, next to guanosine (G), or at both sites), the
number of phosphate groups (di- or triphosphate), and the pres-
ence of additional phosphate modifications (none, phos-
phorothioate, or methylenebisphosphonate). Two compounds
were additionally modified with a 2′-O-methyl group within the
m7G, making them anti-reverse cap analogs (ARCAs) (i.e., an-
alogs that can be incorporated into mRNA by in vitro transcrip-
tion only in the forward orientation).32-34 After their synthesis,
we thoroughly investigated the properties of the new analogs in
the context of mRNA cap-related biological processes, includ-
ing eIF4E-mediated translation initiation and decapping by both
DcpS and Dcp1-Dcp2. To our knowledge, this is the first report
outlining the synthesis of multiple novel cap analogs using 5′-
PSL in addition to evaluating their structural and biochemical
properties, including their resistance to enzymatic degradation
RESULTS
Synthesis and physiological stability of 5′-PSL cap analogs.
Synthesis of P-S-C-5′ linkages in nucleotide and oligonucleo-
tide analogs is usually achieved by either SN2 S-alkylation of an
appropriate phosphorothioate-containing compound with a 5′-
iodonucleos(t)ide derivative or by taking advantage of phos-
phorus (III) chemistry.35-38 As the S-alkylation based ap-
proaches do not require nucleoside protection, these were our
first choice for the synthesis of the 5′-PSL cap analogs. We
tested two strategies, both based on a combination of S-alkyla-
tion and phosphorimidazolide chemistries (Schemes 1–3). The
first approach involves the multistep synthesis of a mononucle-
otide analog containing a terminal phosphorothioate (PS) moi-
ety, followed by S-alkylation with a 5′-deoxy-5′-iodonucleoside
in the final step yielding the target dinucleotide. The second ap-
proach, which was especially effective for the synthesis of the
heavily modified cap analogs (such as 9 or 14), involves the
convergent synthesis of two mononucleotide subunits, followed
by the transformation of one of them into an imidazolide deriv-
ative, and finally, coupling the subunits in a ZnCl2-mediated py-
rophosphate bond formation reaction.39
The 5′-deoxy-5′-iodoguanosine derivatives 18 and 19 were ob-
tained from unprotected nucleosides 16, 17 that were treated
with I2 and triphenylphosphine in DMF.40, 41 The resulting com-
pounds were then converted into the N7-methylated derivatives
20 and 21 using methyl iodide in DMSO (Scheme 1). These
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 26
reactions resulted in 77% and 44% isolated yields, respectively.
RP-HPLC. This purification resulted in isolated yields of 5–
51%.
1
2
3
4
5
6
7
8
Then, 5′-iodonucleosides 18 and 20 were reacted with trisodium
thiophosphate in a water-DMF solution to form the correspond-
ing 5′-phosphorothiolates 22, 25 (Scheme 1).42 After IE purifi-
cation, the yields of these compounds ranged from 53–65%.
Compounds 22 and 25 were then converted into the P-imidaz-
olides 23 and 26, respectively, using a dithiodipyridine/tri-
phenylphosphine oxidation-reduction procedure,43 followed by
coupling with triethylammonium phosphate or thiophosphate
(Scheme 1). These coupling reactions afforded compounds 24,
27 or 28 in good yields (70–80%).
Analogs 7–9 (Scheme 3), were obtained from imidazolides 23
and 26 via coupling with 5′-S-GDP (24), m7GpCH2p (34),16 or
m7,2′-OGDP (35)34 in the presence of excess ZnCl2. The isolated
yields after IE and RP-HPLC purification varied from 6 to 40%.
The synthesis of cap analogs 10–15 (Scheme 3), which bear two
or three modifications (α-PSL, γ-PSL, or β-PS),27 was more de-
manding and required the use of freshly prepared 5′-S-m7GDP-
βS (28) and m7GDP-βS (29), which due to their instability upon
storage had to be reacted immediately after purification. These
reactions involved coupling of 28 and 29 with 23 or 36, yielding
cap analogs 10–15, which contain both PS and PSL moieties
and are synthesized in the form of two P-diastereoisomers des-
ignated D1 and D2 according to their elution order during RP-
HPLC. (Scheme 3, Figure S2A). The crude products were puri-
fied by IE chromatography followed by semipreparative RP-
HPLC, resulting in product yields ranging from 28–56% (Fig-
ure S2B).
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
Cap analogs 1–6 were synthesized by reacting 18, 20, or 21 with
the specified nucleoside 5′-γ-thiotriphosphate18, 44, 45 (29–33) in
DMSO in the presence of DBU (Scheme 2). After depletion of
the starting material, which was determined by RP-HPLC anal-
ysis (Figure S1), the crude products 1–6 were purified by ion
exchange (IE) chromatography followed by semi-preparative
Scheme 1. Synthetic routes to guanine and N7-methylguanine mononucleotide analogs containing a 5′-PSL moiety
Conditions: i) I2, imidazole, PPh3, NMP; ii) Na3PSO3  12H2O, DMF; iii) imidazole, PPh3, TEA, DTDP, DMF; iv) H2PO4 ·Et3NH+,
−
−
DMF; v) H2PSO3 ·Et3NH+, DMF; vi) MeI, DMF.
Scheme 2. Synthesis of dinucleotides 1–6
Conditions: i) DBU, DMSO.
4
ACS Paragon Plus Environment

Page 5 of 26
1
Journal of the American Chemical Society
Scheme 3. Synthesis of dinucleotides 7–15
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conditions: i) ZnCl2, DMF.
Notably, nucleoside 5′-phosphorothiolates are susceptible to
hydrolytic cleavage under acidic conditions.46 Therefore, before
evaluating the biochemical properties of dinucleotides 1–15, we
tested their chemical stability in aqueous buffers pH 3–9. After
incubating for 24 h at 37 C, an RP-HPLC analysis did not re-
veal any traces of analog decomposition in the acidic and neu-
tral buffers (Figure S3). At pH 9, slow 7-methylguanine imid-
azole ring opening was observed – a reaction typical to all 7-
methylguanosine nucleotides 47, 48 (Figure S3, 4 open, 15 open).
Next, we evaluated the general enzymatic susceptibility of se-
lected cap analogs in fetal bovine serum (FBS). FBS contains
various nucleases capable of hydrolyzing pyrophosphate and
phosphodiester bonds and has been previously employed to
study the general stability of modified nucleotides and oligonu-
cleotides under physiological conditions.49 Cap analogs 2, 4, 5,
and 8–15 were incubated in 10% FBS at 37 °C and analyzed by
RP-HPLC at different time points (Figure S4A). The percentage
of cap analog remaining in the solution was plotted against time,
and an exponential decay model was fitted to the experimental
data to determine the half-life values (Figure S4B and Table
S1). The majority of the selected cap analogs had half-lives sim-
ilar to the reference analog, m7GpppG (t1/2 = 26.9 ± 1.3 min),
suggesting that the 5′-PSL modification has little influence on
analog stability in FBS (Table S1).
the mRNA 5′ cap and eIF4E is crucial for canonical initiation
of protein biosynthesis. 50, 51 In vitro transcribed mRNAs carry-
ing cap analogs with a lower affinity for eIF4E than the natural
cap show decreased translatability.25 Thus, high affinity for
eIF4E is a desirable feature of synthetic cap analogs designed
for mRNA engineering applications. To assess the influence of
the 5′-PSL moiety on the eIF4E-cap interaction, we determined
the binding affinities of analogs 1−15 for murine eIF4E using
time-synchronized fluorescence quenching titration (FQT, Fig-
ure 2A).52 The equilibrium association constants (KAS) calcu-
lated from replicate FQT experiments are summarized in Figure
2B and Table S2.
Our results indicate that the 5′-PSL modification generally does
not destabilize the complex with eIF4E, and in some cases, even
stabilizes the interaction (Figure 2B). For example, compound
4 carrying a γ-5′-PSL moiety (i.e., a 5′-O-to-S substitution next
to the m7G), had 1.6-fold higher binding affinity for eIF4E than
m7GpppG. Similarly, a corresponding 2′-O-methylated deriva-
tive of 4 (compound 6), had a KAS that was 1.7-fold higher com-
pared to that of the parent compound, m27,2′-OGpppG. Interest-
ingly, the presence of an α-PSL modification (i.e., a substitution
next to the G; compounds 2 and 7) or combination of the α-PSL
and γ-PSL moieties (compound 9) minimally influenced eIF4E
affinity. Analysis of the binding affinities for compounds 1, 3,
5, 8 and 10−15, which combine one or two 5′-PSL moieties with
The interaction of 5′-PSL cap analogs with eIF4E closely
mimics that with unmodified caps. The interaction between
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 26
other previously reported triphosphate modifications, also con-
firmed the general compatibility of the 5′-PSL moiety with cap-
moiety, but the overall alignment of the triphosphate bridge re-
mained unchanged.
1
eIF4E interaction.
2
3
4
5
6
7
8
9
Combination of the 5′-PSL moiety with the triphosphate bridge
shortened to diphosphate (compounds 1 and 3) or an additional
CH2 modification between the phosphates (compounds 5 and 8)
weakened the interaction with eIF4E compared to m7GpppG
(Table S2), as expected from previous studies.25, 52 However, the
beneficial effect of the γ-PSL moiety was still observed for
these analogs, as the γ-PSL-containing compounds 3 and 5 were
observed to have higher eIF4E affinity than the corresponding
compounds with α-PSL (compounds 1 and 8, respectively). In
contrast, the combination of the 5′-PSL-moiety with a β-phos-
phorothioate moiety (β-PS; compounds 10−15) yielded analogs
that had higher affinities for eIF4E than m7GpppG, with the D1
diastereomers having a higher affinity than the corresponding
D2 isomers (Table S2). These data are in good agreement with
previously reported properties of β-PS cap analogs.18 There was
no additional effect provided by α-PSL in combination with β-
PS, since analogs 10 and 11 had binding affinities very similar
to those for corresponding diastereomers of parent compounds,
m7GppSpG D1 or D2 (Table S2). Intriguingly, the combination
γ-PSL and β-PS (compounds 12 and 13) or the combination of
all three modifications (compounds 14 and 15) yielded analogs
with KAS values lower than those for m7GppSpG D1 or D2, re-
spectively, suggesting a slightly destabilizing effect of γ-PSL in
these analogs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Presence of the 5′-PSL moiety does not significantly alter
cap-binding by translation initiation factor eIF4E. (A) Representa-
tive time-synchronized fluorescence quenching titration (FQT)
curves of meIF4E with compounds 4, 15, and reference cap analog
(m7GpppG). (B) KAS values for cap-eIF4E complexes for analogs
1–15 determined by FQT (mean values from triplicates). (C) Struc-
ture of cap analog 9 in complex with meIF4E (colored sticks) over-
laid with the structure of m7GpppG–eIF4E (from PDB id: 1L8B)
(gray sticks, only the ligand shown for clarity).
The influence of the 5′-PSL moiety on the cap-eIF4E interac-
tion was further studied by X-ray crystallography. The co-crys-
tal structure of analog 9 (which contains both α-PSL and γ-PSL
modifications and had an affinity for eIF4E comparable to
m7GpppG) was determined to 1.92 Å resolution (PDB id:
5OSX). Comparison of this structure with that of eIF4E-
m7GpppG complex (PDB id: 1L8B, Figure 2C),52 revealed very
similar protein and ligand conformations, and protein-ligand
contacts in both structures. This indicates that analog 9 closely
mimics m7GpppG in the eIF4E cap binding pocket, which is
also consistent with our FQT data (Table S2). Indeed, both lig-
ands were tightly bound by Trp56 and Trp102 (cation–π stack-
ing with m7G), Glu103 (hydrogen bonds to N1 and N2 of m7G),
Arg157 and Lys 162 (ionic contacts and hydrogen bonds with
the triphosphate bridge), as well as several ordered water mole-
cules buried in the binding pocket. The only differences result-
ing from the 5′-O-to-S substitutions included subtle changes in
the length of the hydrogen bonds involving the γ-phosphate
γ-PSL modification confers resistance to hydrolysis by
hDcpS. We next investigated the susceptibility of the PSL cap
analogs to hydrolysis by the human DcpS enzyme (Figure 3A,
Figure S5). The cap analogs (1–15) or reference analog
m7GpppG (20 μM) were incubated with 10 nM hDcpS enzyme
for 15, 30, or 60 min, followed by RP-HPLC analysis to esti-
mate the percentage of remaining substrate. The results are
summarized in Table 1, and representative RP-HPLC profiles
are shown in Figure S5A-D. In this assay, 48% of the m7GpppG
was hydrolyzed by hDcpS within 60 min. In contrast, cap ana-
logs 3–6, 8, 9, and 12–15 were completely resistant to hydroly-
sis by hDcpS under the same conditions, indicating that 5′-O-
to-S substitution next to the m7G effectively protects the cap
structure from degradation. These findings are similar to those
previously reported for O-to-CH2 substitution at the β-γ bridg-
ing position (exemplified here by compound 8) and several
other modifications within the γ- or β-phosphate.17, 19 In con-
trast, the PSL modification next to the G moiety (compounds 1,
2, 7, 10, or 11) minimally impacted cap susceptibility to hydrol-
ysis by hDcpS. The susceptibility of cap analogs 3–6, 8, 9, and
12–15 was also studied under a higher enzyme:ligand ratio (200
nM hDcpS and 10 μM cap analog), but this change did not af-
fect the results (Figure 3A, Figure S5E-F).
5′-PSL cap analogs resistant to hDcpS hydrolysis potently
inhibit its activity. The reduced hDcpS susceptibility observed
for some of the synthesized analogs prompted us to evaluate
6
ACS Paragon Plus Environment

Page 7 of 26
Journal of the American Chemical Society
d Values in brackets are from the HPLC assay: conditions as in
their inhibitory properties toward this enzyme. This was per-
b, except 60 µM m7GpppG was used as a substrate.
1
2
3
4
5
6
7
8
formed using a fluorescence-based high-throughput screening
(HTS) assay (Figure 3B, top panel; Figures S6 and S7).20 The
assay is based on the use of the unnatural hDcpS substrate 7-
methylguanosine 5′-fluoromonophosphate (m7GMPF). Upon
hydrolysis by hDcpS, m7GMPF releases fluoride ions, which
can then be quantified by reacting them with a fluorogenic
probe (bis-(tert-butyldimethylsilylfluorescein)), thereby gener-
ating a fluorescent signal that is directly proportional to F- con-
centration. In the presence of an hDcpS inhibitor, fluorescence
signal would be attenuated.
(monomer) and included 5′-PSL-containing cap analogs 1–15
and mononucleotides 25 and 27 along with known hDcpS in-
hibitors (RG3039,14 m7GDP,53 and m7GpSppG (D1/D2)18) as
references. The percentage of inhibition (%inhib) for each com-
pound (Figure S6) was calculated as the ratio of signal intensity
lost in the presence of inhibitor to the maximum response in the
absence of an inhibitor (Table 1). This screening revealed that
a group of cap analogs (compounds 4, 9 and 12–15) strongly
inhibit hDcpS (Figure S6). For these analogs, the IC50 parame-
ters were also determined using the same HTS assay (Figure
3C, Figure S7, and Table 1). Notably, compound 4 (i.e., the
m7GpppG analog containing a single γ-PSL) appears to be a po-
tent hDcpS inhibitor, with an IC50 of ~0.87 μM, which is com-
parable to the most potent cap-derived hDcpS inhibitor reported
in the literature (m7GppBH3pG D1, IC50 ~1 μM under the same
conditions).20 Structural alterations of analog 4, such as short-
ening the triphosphate to diphosphate (compound 3), introduc-
tion of an O-to-CH2 substitution at the α-β bridging position
(compound 5), or the addition of a 2′-O-methyl group to m7G
(compound 6), all decreased the inhibition potency (by 4-, 8.5-
, and 29-fold, respectively). Interestingly, combining the γ-PSL
and α-PSL moieties in one compound increased inhibition by 3-
fold (compound 9, IC50 0.285 μM). The inhibitory properties
were also enhanced when the γ-PSL and β-PS moieties were
combined, regardless of the configuration of the β-P-stereocen-
ter (compounds 12 and 13; 4.9- and 5.5-fold increased potency,
respectively). The combination of all three modifications (γ-
PSL, α-PSL, and β-PS) produced the strongest inhibitors among
the tested compounds (compounds 14 and 15; Table 1) with
compound 15 (IC50 43 nM) showing the highest level of inhibi-
tion, which was comparable to that of RG3039 in the same as-
say.
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
Our initial screening was carried out in the presence of 60 μM
substrate (m7GMPF), 5 μM inhibitor, and 50 nM hDcpS
Table 1. Susceptibility to hDcpS hydrolysis and hDcpS inhi-
bition potency of 5′-phosphorothiolate cap analogs
b
Compound
Susceptibility
to DcpS hy-
drolysisa
%
IC50 [μM]c
inhib
0.0478 ± 0.0106/
(0.0707 ± 0.0164)d
108  1
RG3039
NA
m7Gpp
resistant
hydrolyzed
hydrolyzed
resistant
5.2 ± 1.2
ND
69.5  1.4
37.4  9.5
38.8  6.8
75.0  4.8
m7GppSG (1)
m7GpppSG (2)
m7GSppG (3)
ND
2.93 ± 0.58
0.87 ± 0.17
6.2 ± 1.5
22.6 ± 6.3
72 ± 17
m7GSpppG (4)
m7GSppCH2pG (5)
m27,2′-OGSpppG (6)
m27,2′-OGpppSG (7)
m7GpCH2ppSG (8)
m7GSpppSG (9)
resistant
93.5  3.4
67.7  1.8
24.1  5.1
28.6  2.3
49.1  3.7
106  3
resistant
resistant
hydrolyzed
resistant
5.8 ± 1.3
0.285 ± 0.065
ND
To verify these results, analog 15 was further analyzed with two
other independent assays (Figure 3B, middle panel). An RP-
HPLC inhibition assay using m7GpppG as a substrate con-
firmed that the inhibition potency of 15 towards recombinant
hDcpS is at nanomolar level and similar to RG3039 (Figure
3D). Finally, a fluorescence-based HPLC assay using cytoplas-
mic HeLa cell extracts was applied to verify the inhibitory prop-
erties of 15 against endogenous hDcpS activity. In this experi-
ment, we monitored the degradation of fluorescein-labeled cap
analog (m7GpppA-FAM) in the extract (Figure 3B, bottom
panel). Notably, concentration-dependent inhibition of probe
cleavage was observed in the extracts in the presence of either
compound 15, RG3039, or m7GpCH2ppG – a previously re-
ported hDcpS-resistant cap analog with moderate inhibitory
properties16, 20 (Figure S8). As expected, both 15 and RG3039
were significantly more potent than m7GpCH2ppG (Figure 3E).
Interestingly, in contrast to the other in vitro assays, compound
15 was about 10-fold more potent than RG3039 in this cell ex-
tract assay (EC50 0.70 ± 0.09 nM and 6.82 ± 1.08 nM, respec-
tively).
resistant
m7GppSpSG D1 (10) hydrolyzed
m7GppSpSG D2 (11) hydrolyzed
25.4  12.9
33.4  4.6
107  1
ND
m7GSppSpG D1 (12)
m7GSppSpG D2 (13)
m7GSppSpSG D1 (14)
resistant
resistant
resistant
0.178 ± 0.042
0.159 ± 0.027
0.214 ± 0.040
107  1
107  2
0.0430 ± 0.0080/
109  1
m7GSppSpSG D2 (15)
resistant
(0.0447 ± 0.0089)d
m7GSp (25)
resistant
resistant
ND
13.3  12.5
84.8  3.9
m7GSpp (27)
1.52 ± 0.36
aDetermined by HPLC. Conditions: 20 µM cap analog, 10 nM
hDcpS, 50 mM Tris/HCl buffer containing 200 mM KCl, 0.5 mM
EDTA (pH 7.6), 20 C. Cap analogs assigned ‘resistant’ remained
unhydrolyzed after 60 min.
b From fluoride-release HTS assay: 60 µM m7GMPF (substrate),
5 µM inhibitor (1–15), 50 nM hDcpS, 30 C, 50 mM Tris/HCl, 200
mM KCl, 0.5 mM EDTA, 0.75 mg/ml BSA (pH 7.6). Values were
calculated as described in experimental; %inhib values above 100%
result from fluctuations in background fluorescence levels.
Finally, since analog 15 appears to interact with both DcpS and
eIF4E (Table S2, Table S3), we further investigated the selec-
tivity of this inhibitor. Cap analogs targeting eIF4E have been
shown to inhibit cap-dependent translation in cell-free sys-
tems.54 Therefore, we tested the effect of analog 15 on the trans-
lation of ARCA-capped mRNA encoding firefly luciferase in
c From fluoride-release HTS assay: conditions as in b, except a
10-point half-log dilution series of inhibitor starting from 158 or 50
µM was tested.
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 26
HeLa cell extracts at concentrations that are inhibitory to
domain relative to the C-terminal domain.55, 58, 59 In our hDcpS-
1
2
3
4
5
6
7
8
hDcpS. We found that translation was inhibited at compound
concentrations at least 100-fold higher than those required for
inhibition of hDcpS (Figure S9), indicating that the concentra-
tion of compound 15 required for hDcpS inhibition would not
disrupt normal mRNA translation.
15 complex, the overall protein structure was similar to that pre-
viously described. However, significant differences were ob-
served in the conformation of several amino acid sidechains in
the enzyme active site in both the closed and open states (Figure
S10A). The closed state mimics the substrate-bound complex
and involves numerous contacts formed between the protein
and cap molecule, whereas in the open state, the ligand is less
Co-crystallization of hDcpS with analog 15 provides in-
sights into the molecular basis of enzymatic resistance and
inhibition. To elucidate the molecular basis of hDcpS re-
sistance and to better understand the role of the three O-to-S
substitutions in analog 15, we determined the X-ray crystal
structure of 15 in complex with hDcpS (residues 38–337) at
2.06 Å resolution (PDB id: 5OSY, Figure 3F). hDcpS is a ho-
modimeric enzyme with two active sites. Structural studies
have revealed that in the apo state, both sites are symmetric and
open.55 In the previously reported crystal structures of hDcpS in
complex with either m7GpppG (PDB id: 1ST0)56 or m7GDP
(PDB id: 1XMM),57 two distinct binding sites are observed and
are associated with the open and closed states.57 The confor-
mation change involves the movement of the entire N-terminal
57
tightly bound and mimics the product-bound complex.56,
Electron density maps (Fo-Fc maps) were clearly defined for in-
hibitor molecules bound to both active sites. In chain A (closed
state), we were able to model the whole molecule of analog 15,
while in chain B (open state), the electron density was only vis-
ible for the m7G portion (Figure S10B). After refinement, a dis-
continuity in the Fo-Fc map for the γ-phosphate of the ligand
bound in the closed state was observed, indicating an excess of
electron density in our model. This phenomenon is likely due to
a partial P–S bond break. Therefore, we also modeled two frag-
ments of the initial molecule: 5′-thio-7-methylguanosine and 5′-
thioguanosine-5′-(2-thiodiphosphate), assuming that the γ-
phosphate is disordered or absent in the cleaved product.
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
8
ACS Paragon Plus Environment

Page 9 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Compound 15 potently inhibits hDcpS. (A) Susceptibility of m7GpppG (upper) and analog 15 (lower) to hDcpS analyzed using
RP-HPLC. m7GpppG is completely hydrolyzed within 15 min, while compound 15 is resistant. Data for all compounds are shown in Table
1. (B) Enzymatic assays used for hDcpS inhibitor evaluation: HTS assay based on artificial substrate20 used for initial evaluation of com-
pounds 1–15; in vitro HPLC assay using unmodified m7GpppG as a substrate; and HeLa cell extract assay using fluorescein-labeled cap
analog as a substrate. (C-E) Dose response curves (mean values from triplicates) for selected cap analogs and RG3039 obtained in the three
assays. (F) X-Ray structure of cap analog 15 in complex with hDcpS (residues 38−337). Left: Overall view of hDcpS-compound 15 complex.
Color coding: teal, C-terminal domains; orange, N-terminal domains; purple, hinge regions; yellow sticks, histidine triads; gray sticks, over-
laid m7GpppG ligand from the previously reported complex with H277N-hDcpS (PDB id: 1ST0).56 Inset: Close-up view of the catalytic
center in the closed state with both ligands. Additional amino acid contacts observed only for cap analog 15 are marked in yellow, while the
direction of the nucleophilic attack by His277 is marked as a red dashed line. (G) Comparison of ligand conformations presented in two
orientations, showing different conformations and positions of triphosphate chains and guanosine moieties.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 26
These two alternate ligand set-ups (Figure S10B) were refined therefore, this time point was used for quantification (Table S2).
Transcripts capped with the m7GpppG, ARCA, and β-S ARCA
D2 reference compounds had decapping susceptibilities (DS, or
the fraction of decapped RNA after 15 min) of 0.69, 0.52, and
0.43 respectively. Further, RNA capped with compounds 2 and
4 had DSs of 0.68 and 0.18, respectively, indicating that the
presence of a 5′-O-to-S substitution next to the m7G (γ-PSL),
but not next to the G (α-PSL), notably decreased the suscepti-
bility of the RNA to SpDcp1-Dcp2. Consistent with this obser-
vation, the ARCA versions of caps 2 and 4 (i.e., compounds 7
and 6, ensuring their correct orientation) had analogous proper-
ties (DS = 0.52 and 0.07, respectively). Interestingly, the com-
bination of two 5′-PSL modifications in one compound ap-
peared to diminish the stabilizing effect compared to a single γ-
PSL moiety (compound 9; DS = 0.44). The combination of γ-
PSL with an O-to-CH2 substitution at the α-β-position resulted
in similar DS (compound 5, DS = 0.14) compared to com-
pounds containing only the γ-PSL. In contrast, combination of
α-PSL with an O-to-CH2 substitution at the β-γ-position, did not
confer resistance (compound 8, DS = 0.69). Surprisingly, the
combinations of PSL and β-PS modifications (as in compounds
10–15) increased resistance to decapping by SpDcp1-Dcp2 only
in the case of compounds 10 (α-PSL and β-PS combined, DS =
0.11) and 14 (α-PSL, γ-PSL, and β-PS combined, DS = 0.30).
1
2
3
4
5
6
7
8
to group occupancies of 0.59 and 0.41 for the cleaved and non-
cleaved molecules, respectively. The occurrence of partial 5′-
S–Pγ bond cleavage under crystallization conditions was con-
firmed by RP-HPLC (Figure S11). Despite this 5′-S–Pγ bond
breakage, the resulting fragments, with the exception of the γ-
phosphate moiety, remained very well ordered inside the active
site and retained the conformation of the initial, non-hydrolyzed
inhibitor. Surprisingly, cap analog 15 adopted a very compact
conformation (Figure 3G), with an S-shaped triphosphate chain
forming the core of the molecule. This chain appears to be sta-
bilized by intramolecular hydrogen bonds between the 2′- and
3′-hydroxyl groups of the ribose rings and the two nucleobases
located at the opposite sites. The arrangement of the m7G was
similar to that observed in the m7GpppG and m7GDP com-
plexes, with the m7G base stacked between Trp175 and Leu206
and the ribose in a 2′–endo puckering and C4′–C5′-sc (trans-
gauche) conformation. One notable difference was the slightly
elongated 5′-S–Pγ bond pointing in a different direction com-
pared to the 5′-O-Pγ bond in m7GpppG, placing the γ-phosphate
of 15 farther away from the HIT motif, particularly the His277
residue responsible for the nucleophilic attack during catalysis
(Figure 3F).56 The negatively charged β-PS sulfur atom was
sticking out toward a hinged α-helix and formed a strong ionic
contact with the positively charged amine group of Lys142 and
a hydrogen bond with a hydroxyl group of Ser272. One of the
oxygen atoms of the adjacent α-phosphate also interacted with
the hinge region residues (Lys142 and Tyr143), while the sec-
ond oxygen atom was hydrogen bonded to the phenolic group
of Tyr273, which plays a crucial role in closing the active site
upon cap binding and stabilizing the closed state.56
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
Next, we tested the influence of the γ-PSL modification on
RNA susceptibility to decapping by hDcp2. To this end, tran-
scripts capped with ARCA analogs 6 and 7 were incubated with
recombinant hDcp2 for up to 60 min and analyzed electropho-
retically (Figure 4). The hDcp2 decapping assay confirmed that
the 5′-PSL modification stabilized RNA against decapping, but
only if it was located at the γ-position (Figure 4 and Table 2).
The DSs for RNAs capped with compounds 6 and 7 were 0.18
± 0.05 and 0.60 ± 0.19, respectively, compared to 0.69 ± 0.03
and 0.55 ± 0.04 for the reference compounds m7GpppG and
ARCA. Interestingly, in contrast to the results from the
SpDcp1-Dcp2 assay, the susceptibility of transcripts capped
with compound 6 was comparable to that of RNA capped with
β-S ARCA D2 (DS = 0.14 ± 0.03).
RNAs capped with analogs carrying the γ-PSL moiety are
resistant to hydrolysis by hDcp2. In vitro resistance to the
mRNA decapping enzyme Dcp2 is an important predictor of
cap analog-mediated mRNA stabilization. Unlike DcpS, Dcp2
acts at the beginning of the mRNA decay pathway, cleaving the
cap between the α and β phosphates to release m7GDP and ex-
posing the 5′-phosphorylated RNA to further degradation. For
efficient catalysis, Dcp2 requires the presence of the RNA body
as well as divalent metal ions. The effect of 5′-PSL on the sus-
ceptibility of the cap to Schizosaccharomyces pombe Dcp1-
Dcp2 and human Dcp2 was tested on short (25 nt) RNAs carry-
ing 5′-PSL cap analogs.60
α-PSL enhances cap-dependent translation. After biophysi-
cally and biochemically characterizing all of the novel cap ana-
logs in the context of translation and decapping machinery, we
investigated the properties of the mRNA molecules capped with
these compounds.
SpDcp1-Dcp2 susceptibility was tested for RNAs capped with
analogs 2 and 4–15 (the diphosphate caps 1 and 3 were excluded
from the analysis due to their unfavorable interactions with
eIF4E). The caps were incorporated into the RNA by in vitro
transcription, followed by trimming of the 3′ ends of the tran-
scripts with DNAzyme 10-23 to reduce 3′-end heterogeneity
and facilitate analysis.61 The transcripts were incubated with the
SpDcp1-Dcp2 complex for various periods of time and resolved
in high resolution polyacrylamide gel to separate capped and
uncapped RNAs and estimate the progress of decapping over
time (Figure S12A). The initial fractions of capped RNA versus
total RNA (at time point 0) were used to estimate the capping
efficiencies (Table S2). For all tested analogs capping efficien-
cies were above 70% indicating efficient incorporation into
RNA by SP6 polymerase. To compare decapping susceptibili-
ties of different caps, the fraction of capped RNA present in
each sample was plotted as a function of time (Figure S12B).
The greatest differences were observed at the 15 min time point;
Figure 4. A 5′-PSL moiety next to the m7G (γ-PSL) protects RNA
from decapping by Dcp2. (A) Schematic representation of the
hDcp2 susceptibility assay. (B) Example results from the assay
10
ACS Paragon Plus Environment

Page 11 of 26
Journal of the American Chemical Society
(RNAs resolved by denaturing polyacrylamide gel electrophoresis
ARCA versions (compounds 4 and 2). Interestingly, these val-
ues were between those obtained for the reference ARCA-
capped mRNA (m27,3′-OGpppG; translation efficiency = 1.56 ±
0.14) and mRNAs capped with both β-S-ARCA diastereomers
(translation efficiencies = 2.81 ± 0.37 and 3.45 ± 0.42 for m7,2′-
OGppSpG D1 and D2, respectively), suggesting that the 5′-PSL
moiety in the ARCA-type compounds increases mRNA trans-
lation efficiency.
(PAGE) and stained with SYBR Gold). (C) Susceptibility of 25-nt
transcripts capped with various cap analogs to hDcp2. Fraction of
capped RNA remaining in the reaction mixture is plotted as a func-
tion of time. Data are reported as the average values from two bio-
logical replicates ± S.D. All data were normalized to time 0.
1
2
3
4
5
6
7
8
Table 2. Biochemical properties of compounds 6 and 7
Compound
Decapping Normalized relative mRNA
protein expressionb half-life (h)c
susceptibilitya
The translation properties of transcripts capped with ARCAs 6
and 7 were additionally assessed in HeLa cells. In vitro tran-
scribed mRNAs encoding firefly luciferase capped with com-
m7GpppG
1
9
0.69  0.03
0.55  0.04
1.35  0.34
1.28  0.29
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
7,3′-
m27,3′-OGpppG
pounds 6, 7, or with reference analogs m7GpppG, ARCA (m2
1.90  0.27
2.70  0.30
OGpppG), or β-S ARCA D2 (m27,2′-OGppSpG D2) were trans-
fected into HeLa cells. To account for variation in the transfec-
tion efficiency between samples, mRNA encoding Renilla lu-
ciferase capped with ARCA was used as an internal control for
each transfection.63 The firefly and Renilla luciferase activities
were measured using a dual reporter assay at different time
points post-transfection. Firefly luciferase activity normalized
to Renilla luciferase was plotted as a function of time (Figure
5A),
(ARCA)
m27,2′-OGppSpG D2
(β-S ARCA D2)
m27,2′-OGSpppG (6)
n.d.
0.14  0.03
0.18  0.05
0.60  0.19
2.07  0.19
2.75  0.29
1.54  0.40
1.34  0.23
m27,2′-OGpppSG (7)
a
Fraction of decapped RNA after a 15-min incubation with
hDcp2. Data shown are the average values from duplicates ± S.D.
b
Total protein expression in cell lysates of firefly luciferase
capped with given analog normalized to the expression of
m7GpppG-capped mRNA. The average values from three biologi-
cal replicates ± S.D are reported.
c
mRNA half-life of firefly luciferase transcript capped with
given analog in HeLa cells. The average values from three (two for
m7GpppG) biological replicates ± SEM are reported.
n.d. not determined
The influence of selected 5′-PSL cap analogs on mRNA expres-
sion was investigated in a cell-free translation system (rabbit
reticulocyte lysate, RRL) and in cultured HeLa cells. First, a set
of mRNAs encoding Renilla luciferase with modified 5′ ends
were synthesized in vitro with SP6 RNA polymerase and trans-
lated in RRL. The experiments were performed under condi-
tions optimized for cap-dependent translation as previously de-
scribed.62 The cap-dependent nature of the system was con-
firmed by the low expression of negative control mRNA carry-
ing an unmethylated (non-functional) 5′-cap analog (GpppG)
(Table S2). The results obtained for the mRNAs capped with
analogs 1–15 and functional reference cap structures
Figure 5. mRNAs capped with 5′-PSL cap analogs 6 and 7 are ef-
ficiently translated in living cells. (A) Representative time course
of firefly luciferase mRNA expression in HeLa cells for a single
biological replicate when the specified 5′-cap analogs were incor-
porated. (B) Total protein expressions in HeLa cells for various
capped mRNAs. Bars represent integers under the expression
curves shown in (A) normalized to m7GpppG. Data are reported as
the average values from three biological replicates ± S.D.
and total protein expression was calculated for each mRNA
based on the activity curve (Figure 5B and Table 2). Consistent
with the translation efficiencies obtained in our RRL experi-
ments, total protein expression was higher for mRNAs capped
with analog 7 compared to those capped with analog 6 or un-
modified ARCA (2.07 ± 0.19, 2.75 ± 0.29, and 1.90 ± 0.27, re-
spectively). Notably, transcripts capped with compound 7
yielded similar amounts of protein as mRNA capped with β-S
ARCA D2 (total protein expression = 2.70 ± 0.30) (Figure 5B
and Table 2). Interestingly, total protein expression for mRNAs
capped with analog 6 was comparable to values obtained for
ARCA capped mRNAs, despite that RNAs with compound 6 at
their 5’ ends were more stable against hDcp2 than RNAs
capped with ARCA or analog 7 in in vitro assay (Fig. 4). This
may suggest that resistance to hDcp2 hydrolysis may not trans-
late into increased stability in vivo. Therefore, stability of
mRNA capped with compound 6 and 7 in HeLa cells was in-
vestigated. To this end, HeLa cells were transfected with differ-
ently capped mRNAs, cells were lysed at various time points,
7,2′-
[m7GpppG, ARCA (m27,3ꞌ-OGpppG), and β-S ARCA D2 (m2
OGppSpG D2)] are summarized in Table S2, whereas Figure
S13 shows representative data from a single biological repli-
cate. mRNAs carrying the 5′-γ-PSL moiety (compound 4) or 5′-
α-PSL moiety (compound 2), or a combination of these two
modifications (compound 9) in non-ARCA analogs all had
translation efficiencies that were comparable to that of
m7GpppG-mRNA, indicating that these changes do not signifi-
cantly influence in vitro translation. Alternatively, compounds
containing the PSL modification and an additional meth-
ylenebisphosphonate moiety (α-β (compound 5) or β-γ (com-
pound 8)) had decreased translation efficiency compared to
m7GpppG-mRNA, whereas those with an additional β-PS moi-
ety (cap analogs 2, 4, and 9) produced mRNAs that were trans-
lated more efficiently. These results are consistent with previ-
ous reports for compounds carrying these modifications. As ex-
pected, the 5′-PSL-ARCAs (analogs 6 and 7) also augmented
translation efficiency (relative translation efficiencies = 1.73 ±
0.24 and 2.23 ± 0.31, respectively) compared to their non-
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 26
and levels of firefly, Renilla, and GAPDH mRNA were meas- interactions was also indirectly supported by the high transla-
1
2
3
4
ured by qRT-PCR. We found that the presence of cap analog 6,
7 or ARCA at 5’ end of firefly mRNA does not significantly
differentiate mRNA stability in HeLa cells (Table 2 and Fig.
S14).
tional activity observed for the mRNAs capped with 5′-PSL an-
alogs.
Conversely, the 5′-PSL moiety does significantly influence the
susceptibility of the cap analogs to decapping. Indeed, all of the
analogs containing the γ-PSL moiety were resistant to hDcpS,
whereas α-PSL had no influence. Moreover, these DcpS-re-
sistant compounds were also potent hDcpS inhibitors. Com-
pound 4, for example, had an IC50 < 1 μM, suggesting that it is
more potent than known hDcpS-resistant analogs containing γ-
PS, β-γ-imidodiphosphate, or β-γ-methylenebisphosphonate
moieties. Additional O-to-S substitutions further improved
these inhibitory properties, leading to the most potent inhibitor
– the triple-modified analog 15 (m7GSppSpSG D2) – with low
nanomolar potency. The hDcpS-15 complex crystal structure
suggests that resistance to decapping and increased affinity for
hDcpS likely arises from the accumulation of multiple steric
and electronic effects. The primary factor involved in resistance
is the altered γ-phosphate geometry caused by 5′-C-S-P bond
elongation, which pushes the γ-phosphorus about 4.1 Å away
from the site occupied by the unmodified cap, making it una-
vailable to the catalytic histidine triad. The additional α-PSL
moiety likely ensures higher conformational flexibility enabling
the ligand to adopt a more favorable conformation and form
new ionic contacts with Lys 142 and Tyr 143, which were not
observed in the previously reported structure of catalytically in-
active mutant in complex with m7GpppG.52 Finally, it is proba-
ble that the interaction with Lys 142 is further stabilized by fa-
vorable electronic effects introduced by the β-PS moiety. These
strong cap-enzyme contacts involving both the C- and N-hDcpS
terminal domains may stabilize the closed enzyme state, a fea-
ture required for potent inhibition.11 Notably, the potency of an-
alog 15 in two independent in vitro assays was comparable to
that of RG3039, a quinazoline-derived compound that alleviates
SMA symptoms in a mouse model, but failed in clinical trials.
Furthermore, in a HeLa cell lysate inhibition assay, 15 showed
10-fold greater potency, indicating its potentially higher selec-
tivity. This assay also demonstrated that compound 15 is stable
enough under physiological conditions to potently inhibit en-
dogenous hDcpS. Importantly, we established the concentra-
tions of compound 15 that provide complete hDcpS inhibition
without any effect on translation. Another advantage of com-
pound 15 is that unlike other small molecule inhibitors it can be
easily incorporated into the 5′ end of the RNA by in vitro tran-
scription. Thus, analog 15 may facilitate studies of hDcpS ac-
tivity, including the molecular mechanisms involved in long
and short capped RNA discrimination as well as definition of
the links between decapping, splicing modulation, and thera-
peutic relevance in SMA. Notably, other cellular and in vivo
studies could also be performed provided that the compound is
additionally modified or complexed with some carriers to en-
hance cell permeability.64-68 While additional work is necessary
to elucidate the full use of compound 15, the current study
firmly places it in the category of potent DcpS inhibitors.
5
6
7
8
DISCUSSION
5′-PSL is a sister modification to PS; however, in 5′-PSL, a sul-
fur occupies the bridging position within the modified phos-
phate moiety rather than the non-bridging position. Therefore,
unlike PS, 5′-PSL does not introduce a new P-stereogenic center
into the molecule, simplifying synthesis, purification, and bio-
logical evaluation of the resulting compounds. These character-
istics make them ideal for the synthesis of mRNA cap analogs
and the subsequent study of decapping enzymes, such as DcpS
and Dcp2. In the present study, we explored the 5′-PSL linkage
as an mRNA 5′ cap modification in 15 novel dinucleotide
m7GpppG-derived cap analogs. Notably, the present study is the
first to explore the synthetic accessibility and biological prop-
erties of these particular compounds.
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
The synthetic pathways used to synthesize the 5′-PSL cap ana-
logs, which were based on a combination of SN2 S-alkylation
for 5′-PSL insertion and P-imidazolide chemistry for pyrophos-
phate bond formation, turned out to be moderately efficient, but
thanks to the low number of synthetic steps appeared to be scal-
able and compatible with the practical applications of these
compounds. Additionally, the chemical stability of these ana-
logs in aqueous solutions of varying pH makes them compatible
with biological applications. Although nucleoside 5′-phos-
phorothiolates undergo acid-catalyzed P-S bond cleavage in
aqueous solutions,46 the 5′-PSL cap analogs synthesized here
were chemically stable at pH 3−9, indicating that dinucleotides
containing the 5′-PSL moiety are more stable than the corre-
sponding mononucleotides. The stability of the cap analogs in
blood serum (FBS) was also essentially unaffected by the 5′-
PSL moiety.
After evaluating the basic biocompatibility of the analogs, we
evaluated the biochemical properties of the compounds using a
set of assays involving three major cytoplasmic cap-recogniz-
ing proteins important for mRNA expression and turnover:
eIF4E, DcpS, and Dcp1-Dcp2 complex. Our results show that
the presence of the 5′-PSL modification had differential effects
on these proteins. For eIF4E, the 5′-PSL moiety did not affect
or even slightly stabilized the mRNA-enzyme interaction, indi-
cating that the longer bonds are generally well accommodated
in the cap-binding pocket. However, the single β-O-to-S substi-
tuted cap analogs (e.g. β-S-ARCA D1) had higher affinity than
corresponding analogs with multiple substitutions (e.g. analogs
10, 12, and 14), indicating that additional O-to-S substitutions
may disrupt the optimal conformation for ligands carrying β-PS
moiety. Dual 5′-PSL moieties in compound 9 did not impact
eIF4E binding affinity. This corresponded with the eIF4E co-
crystal structure, wherein triphosphate chain alignments in
compound 9 and parent cap analog (m7GpppG) were very sim-
ilar. This is likely because structural perturbations introduced
by longer C-S and S-P bonds compared to C-O and P-S bonds
are accommodated by C5′-S-P versus C5′-O-P angle differ-
ences. The effects of the 5′-PSL moiety in cap analog-eIF4E
Following incorporation into RNA, we also studied the suscep-
tibility of 5′-PSL cap analogs to Dcp2. All novel cap analogs
were efficiently incorporated into short RNAs during in vitro
transcription, consistent with previously reported findings that
nucleoside 5′-phosphorothiolates can serve as RNA polymerase
transcription initiators.40, 69 RNAs capped with analogs contain-
ing a γ-PSL moiety, but not an α-PSL moiety, had increased
12
ACS Paragon Plus Environment

Page 13 of 26
Journal of the American Chemical Society
insight into how it can be used to modulate the activity of dinu-
resistance to both SpDcp1-Dcp2 heterodimer or hDcp2 mono-
1
2
3
4
5
6
7
8
mer. Interestingly, this is the same position of modification that
also conferred resistance to DcpS. This is unexpected consider-
ing that the two enzymes belong to different protein families,
utilize different length RNA substrates, and cleave the cap at
different positions via different mechanisms. Although we can-
not yet explain Dcp2 resistance at the molecular level, it likely
results from shifting of the β-phosphate away from the catalytic
site in the Nudix domain caused by 5′-C-S-P bond elongation.
This suggests that even subtle changes in the cap’s triphosphate
chain geometry may cause significant changes in the biochem-
ical properties of the resulting RNA molecules.
cleoside 5′,5′-oligophosphates in normal biological processes as
well as in disease.
ASSOCIATED CONTENT
Supporting Information. Supporting Info File 1: Supporting Ta-
bles S1–S3, Supporting Figures S1–S14, detailed experimental
procedures. Supporting Info File 2: 1H, 31P NMR, and HRMS spec-
tra for all novel compounds. This material is available free of
charge via the Internet at http://pubs.acs.org. The atomic coordi-
nates and structure factors files of eIF4E-9 and hDcpS-15 com-
plexes were deposited in Protein Data Bank under 5OSX and 5OSY
accession codes, respectively.
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
Finally, we inspected the translational properties of mRNAs
capped with 5′-PSL cap analogs in cell-free and cultured cell
models. Generally, the 5′-PSL modification did not interfere
with translation and in some cases increased the translational
potential of the mRNA, especially in the case of analogs carry-
ing the α-PSL moiety.70 However, a more detailed analysis of
this was not possible because most of the analogs could be in-
AUTHOR INFORMATION
Corresponding Author
*jkowalska@fuw.edu.pl
*j.jemielity@cent.uw.edu.pl
corporated into the mRNA in either the forward or reverse ori-
Author Contributions
70, 71
entations
and these two forms could not be differentiated
‡These authors contributed equally.
nor quantified. This issue was circumvented using analogs 6
and 7, which are ARCAs that contain the γ- and α-PSL moieties,
respectively. Their translation efficiencies in HeLa cells were
Funding Sources
This work was supported by the Foundation for Polish Science
(TEAM/2016-2/13), the National Centre of Research and Develop-
ment in Poland (LIDER/001/003/L-5/13/NCBR/2014), the Minis-
try of Science and Higher Education (Poland, DI2012 024842), and
the National Science Centre (ETIUDA 2017/24/T/NZ1/00345).
7,3′-
compared to those of known cap analogs m7GpppG, m2
7,2′-
OGpppG (a standard, commercially available ARCA) and m2
OGppSpG D2 (-S-ARCA D2, optimized for efficient transla-
tion in vitro and in vivo).70 Notably, the mRNAs capped with 7
were translated more efficiently than mRNAs capped with 6
(despite similar half-lives) and were comparable to -S-ARCA
D2-terminated mRNAs. Therefore, in our model, it appears that
the susceptibility of the 5′-PSL cap analogs to Dcp2 is not the
primary determinant of translational activity. Furthermore, the
high translational activity of compound 7 combined with its rel-
atively simple synthesis and lack of P-stereoisomerism makes
it an appealing alternative for the previously developed β-S-
ARCA.
ACKNOWLEDGMENT
We thank Mike Kiledjian (Rutgers University) for the hDcpS en-
coding plasmid, Christopher D. Lima (Sloan-Kettering Institute)
for the hDcp2 encoding plasmid, John D. Gross (University of Cal-
ifornia, San Francisco) for plasmids encoding the Dcp1-Dcp2 com-
plex from S. pombe, Stephen R. Ikeda (The National Institute on
Alcohol Abuse and Alcoholism) for providing the hRLuc-
pRNA2(A)128 plasmid, and Marek R. Baranowski (University of
Warsaw) for chemical reagents used in HTS assay. Diffraction data
have been collected on BL14.1 at the BESSY II electron storage
ring operated by the Helmholtz-Zentrum Berlin.72
CONCLUSIONS
In conclusion, we employed a wide range of biochemical assays
to evaluate a set of 15 chemically synthesized cap analogs con-
taining a 5ʹ-PSL moiety. We found that 5′-PSL stabilizes inter-
action of the cap structure with eIF4E, can enhance mRNA
translation, and in some cases, significantly reduces mRNA cap
susceptibility to decapping. These findings provide new struc-
tural insights into specific protein-mRNA cap interactions,
leading to the selection of novel cap analogs with superior bio-
logical properties. These superior analogs include: (i) com-
pound 6 (m27,2′-OGSpppG), which shows a notable decrease in
susceptibility to hDcp2, (ii) compound 7 (m27,2′-OGpppSG),
which enhances mRNA translational potential in a manner that
is comparable to β-S-ARCA D2 (m27,2′-OGppSpG D2), and (iii)
compound 15 (m7GSppSpSG D2), which inhibits hDcpS more
potently than clinically tested RG3039. Taken together, our
study demonstrates that the 5′-PSL moiety confers useful bio-
logical properties to mRNA caps, enabling modulation of ther-
apeutically relevant cap-dependent processes. While still some
additional work is necessary, this study highlights the 5′-PSL
moiety as a currently underexplored modification and provides
REFERENCES
1.
Bowler, M. W.; Cliff, M. J.; Waltho, J. P.; Blackburn, G. M.
Why did Nature select phosphate for its dominant roles in biology?
New J. Chem. 2010, 34 (5), 784-794.
2.
Korhonen, H. J.; Conway, L. P.; Hodgson, D. R. W. Phos-
phate analogues in the dissection of mechanism. Curr. Opin. Chem.
Biol. 2014, 21, 63-72.
3.
Coller, J.; Parker, R. Eukaryotic mRNA decapping. Annu.
Rev. Biochem. 2004, 73, 861-890.
4.
Grudzien-Nogalska, E.; Kiledjian, M. New insights into
decapping enzymes and selective mRNA decay. Wiley Interdiscip.
Rev.-RNA 2017, 8 (1), 1-11.
5.
Ramanathan, A.; Robb, G.; Chan, S. mRNA capping: bio-
logical functions and applications. Nucleic Acids Res. 2016, 44 (16),
7511-7526.
6.
ping: past and prospects. Adv. Virus Res. 2000, 55, 135-184.
7. Warminski, M.; Sikorski, P. J.; Kowalska, J.; Jemielity, J.
Furuichi, Y.; Shatkin, A. J. Viral and cellular mRNA cap-
Applications of Phosphate Modification and Labeling to Study
(m)RNA Caps. Top. Curr. Chem. 2017, 375:16 (1).
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 26
8.
Liu, H.; Rodgers, N. D.; Jiao, X.; Kiledjian, M. The scaven-
22.
van Dijk, E.; Cougot, N.; Meyer, S.; Babajko, S.; Wahle, E.;
ger mRNA decapping enzyme DcpS is a member of the HIT family of
pyrophosphatases. EMBO J. 2002, 21 (17), 4699-4708.
Seraphin, B. Human Dcp2: a catalytically active mRNA decapping en-
zyme located in specific cytoplasmic structures. Embo J. 2002, 21 (24),
6915-6924.
1
2
3
4
5
6
7
8
9.
binding protein-dependent processes? RNA Biol. 2008, 5 (4), 216-219.
10. Song, M.-G.; Bail, S.; Kiledjian, M. Multiple Nudix family
Bail, S.; Kiledjian, M. DcpS, a general modulator of cap-
23.
Wang, Z.; Jiao, X.; Carr-Schmid, A.; Kiledjian, M. The
hDcp2 protein is a mammalian mRNA decapping enzyme. Proc. Natl.
Acad. Sci. U. S. A. 2002, 99 (20), 12663-12668.
proteins possess mRNA decapping activity. RNA 2013, 19 (3), 390-
399.
24.
karyotic mRNA decapping. Curr. Opin. Struct. Biol. 2017, 47, 40-51.
25. Grudzien, E.; Kalek, M.; Jemielity, J.; Darzynkiewicz, E.;
Valkov, E.; Jonas, S.; Weichenrieder, O. Mille viae in eu-
11.
Singh, J.; Salcius, M.; Liu, S. W.; Staker, B. L.; Mishra, R.;
Thurmond, J.; Michaud, G.; Mattoon, D. R.; Printen, J.; Christensen,
J.; Bjornsson, J. M.; Pollok, B. A.; Kiledjian, M.; Stewart, L.; Jarecki,
J.; Gurney, M. E. DcpS as a therapeutic target for spinal muscular at-
rophy. ACS Chem. Biol. 2008, 3 (11), 711-722.
Rhoads, R. E. Differential inhibition of mRNA degradation pathways
by novel cap analogs. J. Biol. Chem. 2006, 281 (4), 1857-1867.
9
26.
Su, W.; Slepenkov, S.; Grudzien-Nogalska, E.; Kowalska,
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
12.
Li, Y.; Dai, J.; Song, M.; Fitzgerald-Bocarsly, P.; Kiledjian,
J.; Kulis, M.; Zuberek, J.; Lukaszewicz, M.; Darzynkiewicz, E.; Jemie-
lity, J.; Rhoads, R. E. Translation, stability, and resistance to decapping
of mRNAs containing caps substituted in the triphosphate chain with
BH3, Se, and NH. RNA 2011, 17 (5), 978-988.
M. Dcp2 Decapping Protein Modulates mRNA Stability of the Critical
Interferon Regulatory Factor (IRF) IRF-7. Mol. Cell. Biol. 2012, 32 (6),
1164-1172.
13.
Gogliotti, R. G.; Cardona, H.; Singh, J.; Bail, S.; Emery, C.;
27.
Grudzien-Nogalska, E.; Jemielity, J.; Kowalska, J.; Darzyn-
Kuntz, N.; Jorgensen, M.; Durens, M.; Xia, B.; Barlow, C.; Heier, C.
R.; Plasterer, H. L.; Jacques, V.; Kiledjian, M.; Jarecki, J.; Rusche, J.;
DiDonato, C. J. The DcpS inhibitor RG3039 improves survival, func-
tion and motor unit pathologies in two SMA mouse models. Hum. Mol.
Genet. 2013, 22 (20), 4084-4101.
kiewicz, E.; Rhoads, R. E. Phosphorothioate cap analogs stabilize
mRNA and increase translational efficiency in mammalian cells. RNA
2007, 13 (10), 1745-1755.
28.
Sahin, U.; Derhovanessian, E.; Miller, M.; Kloke, B.-P.; Si-
mon, P.; Löwer, M.; Bukur, V.; Tadmor, A. D.; Luxemburger, U.;
Schrörs, B.; Omokoko, T.; Vormehr, M.; Albrecht, C.; Paruzynski, A.;
Kuhn, A. N.; Buck, J.; Heesch, S.; Schreeb, K. H.; Müller, F.; Ortseifer,
I.; Vogler, I.; Godehardt, E.; Attig, S.; Rae, R.; Breitkreuz, A.; Tolliver,
C.; Suchan, M.; Martic, G.; Hohberger, A.; Sorn, P.; Diekmann, J.;
Ciesla, J.; Waksmann, O.; Brück, A.-K.; Witt, M.; Zillgen, M.; Roth-
ermel, A.; Kasemann, B.; Langer, D.; Bolte, S.; Diken, M.; Kreiter, S.;
Nemecek, R.; Gebhardt, C.; Grabbe, S.; Höller, C.; Utikal, J.; Huber,
C.; Loquai, C.; Türeci, Ö. Personalized RNA mutanome vaccines mo-
bilize poly-specific therapeutic immunity against cancer. Nature 2017,
547 (7662), 222-226.
14.
Van Meerbeke, J. P.; Gibbs, R. M.; Plasterer, H. L.; Miao,
W.; Feng, Z.; Lin, M. Y.; Rucki, A. A.; Wee, C. D.; Xia, B.; Sharma,
S.; Jacques, V.; Li, D. K.; Pellizzoni, L.; Rusche, J. R.; Ko, C. P.;
Sumner, C. J. The DcpS inhibitor RG3039 improves motor function in
SMA mice. Hum. Mol. Genet. 2013, 22 (20), 4074-4083.
15.
Gopalsamy, A.; Narayanan, A.; Liu, S.; Parikh, M.; Kyne,
R.; Fadeyi, O.; Tones, M.; Cherry, J.; Nabhan, J.; LaRosa, G.; Petersen,
D.; Menard, C.; Foley, T.; Noell, S.; Ren, Y.; Loria, P.; Maglich-Good-
win, J.; Rong, H.; Jones, L. Design of Potent mRNA Decapping Scav-
enger Enzyme (DcpS) Inhibitors with Improved Physicochemical
Properties To Investigate the Mechanism of Therapeutic Benefit in Spi-
nal Muscular Atrophy (SMA). J. Med. Chem. 2017, 60 (7), 3094-3108.
29.
Kuhn, A. N.; Diken, M.; Kreiter, S.; Selmi, A.; Kowalska,
J.; Jemielity, J.; Darzynkiewicz, E.; Huber, C.; Türeci, O.; Sahin, U.
Phosphorothioate cap analogs increase stability and translational effi-
ciency of RNA vaccines in immature dendritic cells and induce supe-
rior immune responses in vivo. Gene Ther. 2010, 17 (8), 961-971.
16.
Kalek, M.; Jemielity, J.; Darzynkiewicz, Z. M.; Bojarska, E.;
Stepinski, J.; Stolarski, R.; Davis, R. E.; Darzynkiewicz, E. Enzymati-
cally stable 5′ mRNA cap analogs: Synthesis and binding studies with
human DcpS decapping enzyme. Bioorg. Med. Chem. 2006, 14 (9),
3223-3230.
30.
Shukla, S.; Parker, R. Quality control of assembly-defective
U1 snRNAs by decapping and 5'-to-3' exonucleolytic digestion. Proc.
Natl. Acad. Sci. U. S. A. 2014, 111 (32), E3277-E3286.
17.
Rydzik, A. M.; Kulis, M.; Lukaszewicz, M.; Kowalska, J.;
Zuberek, J.; Darzynkiewicz, Z. M.; Darzynkiewicz, E.; Jemielity, J.
Synthesis and properties of mRNA cap analogs containing imidodi-
phosphate moiety-fairly mimicking natural cap structure, yet resistant
to enzymatic hydrolysis. Bioorg. Med. Chem. 2012, 20 (5), 1699-1710.
31.
Taverniti, V.; Séraphin, B. Elimination of cap structures
generated by mRNA decay involves the new scavenger mRNA decap-
ping enzyme Aph1/FHIT together with DcpS. Nucleic Acids Res. 2015,
43 (1), 482-492.
18.
Kowalska, J.; Lewdorowicz, M.; Zuberek, J.; Grudzien-
32.
Jemielity, J.; Fowler, T.; Zuberek, J.; Stepinski, J.; Lewdor-
Nogalska, E.; Bojarska, E.; Stepinski, J.; Rhoads, R. E.; Darzynkie-
wicz, E.; Davis, R. E.; Jemielity, J. Synthesis and characterization of
mRNA cap analogs containing phosphorothioate substitutions that bind
tightly to eIF4E and are resistant to the decapping pyrophosphatase
DcpS. RNA 2008, 14 (6), 1119-1131.
owicz, M.; Niedzwiecka, A.; Stolarski, R.; Darzynkiewicz, E.; Rhoads,
R. E. Novel "anti-reverse" cap analogs with superior translational prop-
erties. RNA 2003, 9 (9), 1108-1122.
33.
Peng, Z.-H.; Sharma, V.; Singleton, S. F.; Gershon, P. D.
Synthesis and Application of a Chain-Terminating Dinucleotide
mRNA Cap Analog. Org. Lett. 2002, 4 (2), 161-164.
19.
Kowalska, J.; Wypijewska del Nogal, A.; Darzynkiewicz, Z.
M.; Buck, J.; Nicola, C.; Kuhn, A. N.; Lukaszewicz, M.; Zuberek, J.;
Strenkowska, M.; Ziemniak, M.; Maciejczyk, M.; Bojarska, E.;
Rhoads, R. E.; Darzynkiewicz, E.; Sahin, U.; Jemielity, J. Synthesis,
properties, and biological activity of boranophosphate analogs of the
mRNA cap: versatile tools for manipulation of therapeutically relevant
cap-dependent processes. Nucleic Acids Res. 2014, 42 (16), 10245-
10264.
34.
Stepinski, J.; Waddell, C.; Stolarski, R.; Darzynkiewicz, E.;
Rhoads, R. E. Synthesis and properties of mRNAs containing the novel
"anti-reverse" cap analogs 7-methyl(3′-O-methyl)GpppG and 7-me-
thyl(3'-deoxy)GpppG. RNA 2001, 7 (10), 1486-1495.
35.
Li, N. S.; Frederiksen, J. K.; Piccirilli, J. A. Synthesis, prop-
erties, and applications of oligonucleotides containing an RNA dinu-
cleotide phosphorothiolate linkage. Accounts Chem. Res. 2011, 44 (12),
1257-1269.
20.
Baranowski, M. R.; Nowicka, A.; Jemielity, J.; Kowalska, J.
A fluorescent HTS assay for phosphohydrolases based on nucleoside
5′-fluorophosphates: its application in screening for inhibitors of
mRNA decapping scavenger and PDE-I. Org. Biomol. Chem. 2016, 14
(20), 4595-4604.
36.
Eguaogie, O.; Cooke, L. A.; Martin, P. M.; Ravalico, F.;
Conway, L. P.; Hodgson, D. R.; Law, C. J.; Vyle, J. S. Synthesis of
novel pyrophosphorothiolate-linked dinucleoside cap analogues in a
ball mill. Org. Biomol. Chem. 2015, 14 (4), 1201-1205.
21.
She, M.; Decker, C. J.; Chen, N.; Tumati, S.; Parker, R.;
37.
Patel, B. K.; Eckstein, F. 5′-Deoxy-5′-thioribonucleoside-5′-
Song, H. Crystal structure and functional analysis of Dcp2p from
Schizosaccharomyces pombe. Nat. Struct. Mol. Biol. 2006, 13 (1), 63-
70.
triphosphates. Tetrahedron Lett. 1997, 38 (6), 1021-1024.
14
ACS Paragon Plus Environment

Page 15 of 26
Journal of the American Chemical Society
38.
Shah, R.; Maize, K. M.; Zhou, X.; Finzel, B. C.; Wagner, C.
54.
Cai, A.; Jankowska-Anyszka, M.; Centers, A.; Chlebicka,
R. Caught before Released: Structural Mapping of the Reaction Trajec-
tory for the Sofosbuvir Activating Enzyme, Human Histidine Triad Nu-
cleotide Binding Protein 1 (hHint1). Biochemistry 2017, 56 (28), 3559-
3570.
L.; Stepinski, J.; Stolarski, R.; Darzynkiewicz, E.; Rhoads, R. E. Quan-
titative assessment of mRNA cap analogues as inhibitors of in vitro
translation. Biochemistry 1999, 38 (26), 8538-8547.
1
2
3
4
55.
Liu, S. W.; Rajagopal, V.; Patel, S. S.; Kiledjian, M. Mech-
39.
Kadokura, M.; Wada, T.; Urashima, C.; Sekine, M. Efficient
anistic and kinetic analysis of the DcpS scavenger decapping enzyme.
synthesis of gamma-methyl-capped guanosine 5′-triphosphate as a 5′-
terminal unique structure of U6 RNA via a new triphosphate bond for-
mation involving activation of methyl phosphorimidazolidate using
ZnCl2 as a catalyst in DMF under anhydrous conditions. Tetrahedron
Lett. 1997, 38 (48), 8359-8362.
J. Biol. Chem. 2008, 283 (24), 16427-16436.
5
6
7
8
56.
Gu, M. G.; Fabrega, C.; Liu, S. W.; Liu, H. D.; Kiledjian,
M.; Lima, C. D. Insights into the structure, mechanism, and regulation
of scavenger mRNA decapping activity. Mol. Cell 2004, 14 (1), 67-80.
57.
Chen, N.; Walsh, M. A.; Liu, Y.; Parker, R.; Song, H. Crys-
tal Structures of Human DcpS in Ligand-free and m7GDP-bound forms
Suggest a Dynamic Mechanism for Scavenger mRNA Decapping. J.
Mol. Biol. 2005, 347 (4), 707-718.
9
40.
Kim, I.-H.; Shin, S.; Jeong, Y.-J.; Hah, S. S. Two-step syn-
thesis of 5′-deoxy-5′-thioguanosine-5′-monophosphorothioate and its
incorporation efficiency into 5′-terminus of RNA for preparation of
thiol-functionalized RNA. Tetrahedron Lett. 2010, 51 (26), 3446-3448.
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
58.
Neu, A.; Neu, U.; Fuchs, A. L.; Schlager, B.; Sprangers, R.
41.
logues: methods for the preparation of 2′,3′-secoguanosine, 5′-deoxy-
2′,3′-secoguanosine, and(R,S)-9-[1-(2-hydroxyethoxy)-2-hydroxy-
ethyl]guanine. Can. J. Chem. 1986, 64 (9), 1885-1889.
42. Zhang, B.; Cui, Z.; Sun, L. Synthesis of 5′-deoxy-5′-thiogua-
McGee, D. P. C.; Martin, J. C. Acyclic nucleoside ana-
An excess of catalytically required motions inhibits the scavenger
decapping enzyme. Nat. Chem. Biol. 2015, 11 (9), 697-704.
59.
Pentikainen, U.; Pentikainen, O. T.; Mulholland, A. J. Co-
operative symmetric to asymmetric conformational transition of the
apo-form of scavenger decapping enzyme revealed by simulations.
Proteins 2008, 70 (2), 498-508.
nosine- 5′-monophosphorothioate and its incorporation into RNA 5′-
termini. Org. Lett. 2001, 3 (2), 275-278.
60.
Steiger, M.; Carr-Schmid, A.; Schwartz, D. C.; Kiledjian,
43.
Mukaiyama, T.; Hashimoto, M. Phosphorylation by Oxida-
M.; Parker, R. Analysis of recombinant yeast decapping enzyme. RNA
2003, 9 (2), 231-238.
tion-Reduction Condensation. Preparation of Active Phosphorylating
Reagents. Bull. Chem. Soc. Jpn. 1971, 44 (8), 2284-2284.
61.
Coleman, T. M.; Wang, G.; Huang, F. Superior 5′ homoge-
44.
Strenkowska, M.; Wanat, P.; Ziemniak, M.; Jemielity, J.;
neity of RNA from ATP-initiated transcription under the T7 phi 2.5
promoter. Nucleic Acids Res. 2004, 32 (1), e14.
Kowalska, J. Preparation of synthetically challenging nucleotides using
cyanoethyl P-imidazolides and microwaves. Org. Lett. 2012, 14 (18),
4782-4785.
62.
Rydzik, A. M.; Lukaszewicz, M.; Zuberek, J.; Kowalska, J.;
Darzynkiewicz, Z. M.; Darzynkiewicz, E.; Jemielity, J. Synthetic dinu-
cleotide mRNA cap analogs with tetraphosphate 5′,5′ bridge containing
methylenebis(phosphonate) modification. Org. Biomol. Chem. 2009, 7
(22), 4763-4776.
45.
Kowalska, J.; Lewdorowicz, M.; Darzynkiewicz, E.; Jemie-
lity, J. A simple and rapid synthesis of nucleotide analogues containing
a phosphorothioate moiety at the terminal position of the phosphate
chain. Tetrahedron Lett. 2007, 48 (31), 5475-5479.
63.
Warminski, M.; Sikorski, P. J.; Warminska, Z.; Lukasze-
46.
Brear, P.; Freeman, G. R.; Shankey, M. C.; Trmcic, M.;
wicz, M.; Kropiwnicka, A.; Zuberek, J.; Darzynkiewicz, E.; Kowalska,
J.; Jemielity, J. Amino-Functionalized 5′ Cap Analogs as Tools for
Site-Specific Sequence-Independent Labeling of mRNA. Bioconjugate
Chem. 2017, 28 (7), 1978–1992.
Hodgson, D. R. W. Aqueous methods for the preparation of 5′-substi-
tuted guanosine derivatives. Chem. Commun. 2009, (33), 4980-4981.
47.
Darzynkiewicz, E.; Stepinski, J.; Tahara, S. M.; Stolarski,
R.; Ekiel, I.; Haber, D.; Neuvonen, K.; Lehikoinen, P.; Labadi, I.; Lö-
nnberg, H. Synthesis, Conformation and Hydrolytic Stability of
p1,p3−Dinucleoside Triphosphates Related to mRNA 5′-cap, and Com-
parative Kinetic Studies on their Nucleoside and Nucleoside Mono-
phosphate Analogs. Nucleosides Nucleotides 1990, 9 (4), 599-618.
64.
Zochowska, M.; Piguet, A. C.; Jemielity, J.; Kowalska, J.;
Szolajska, E.; Dufour, J. F.; Chroboczek, J. Virus-like particle-medi-
ated intracellular delivery of mRNA cap analog with in vivo activity
against hepatocellular carcinoma. Nanomedicine 2015, 11 (1), 67-76.
65.
Neamtu, I.; Rusu, A. G.; Diaconu, A.; Nita, L. E.; Chiriac,
48.
Mikkola, S.; Salomaki, S.; Zhang, Z.; Maki, E.; Lonnberg,
A. P. Basic concepts and recent advances in nanogels as carriers for
medical applications. Drug Deliv. 2017, 24 (1), 539-557.
H. Preparation and properties of mRNA 5′-cap structure. Curr. Org.
Chem. 2005, 9 (10), 999-1022.
66.
Gerson, T.; Makarov, E.; Senanayake, T. H.; Gorantla, S.;
49.
Conway, J. W.; McLaughlin, C. K.; Castor, K. J.; Sleiman,
Poluektova, L. Y.; Vinogradov, S. V. Nano-NRTIs demonstrate low
neurotoxicity and high antiviral activity against HIV infection in the
brain. Nanomed.-Nanotechnol. Biol. Med. 2014, 10 (1), 177-185.
H. DNA nanostructure serum stability: greater than the sum of its parts.
Chem. Commun. 2013, 49 (12), 1172-1174.
50.
Jackson, R.; Hellen, C.; Pestova, T. The mechanism of eu-
67.
– An unsolvable task? Antivir. Chem. Chemother. 2017, 25 (3), 69-82.
68. Gollnest, T.; de Oliveira, T. D.; Schols, D.; Balzarini, J.;
Chris, M. Nucleoside diphosphate and triphosphate prodrugs
karyotic translation initiation and principles of its regulation. Nat. Rev.
Mol. Cell Biol. 2010, 11 (2), 113-127.
51.
Culjkovic, B.; Topisirovic, I.; Borden, K. Controlling gene
Meier, C. Lipophilic prodrugs of nucleoside triphosphates as biochem-
ical probes and potential antivirals. Nat. Commun. 2015, 6:8716.
expression through RNA regulons - The role of the eukaryotic transla-
tion initiation factor eIF4E. Cell Cycle 2007, 6 (1), 65-69.
69.
Zhang, L.; Sun, L.; Cui, Z.; Gottlieb, R. L.; Zhang, B. 5′-
52.
Niedzwiecka, A.; Marcotrigiano, J.; Stepinski, J.; Jankow-
Sulfhydryl-Modified RNA:ꢀ Initiator Synthesis, in Vitro Transcription,
and Enzymatic Incorporation. Bioconjugate Chem. 2001, 12 (6), 939-
948.
ska-Anyszka, M.; Wyslouch-Cieszynska, A.; Dadlez, M.; Gingras, A.-
C.; Mak, P.; Darzynkiewicz, E.; Sonenberg, N.; Burley, S. K.; Stolar-
ski, R. Biophysical Studies of eIF4E Cap-binding Protein: Recognition
of mRNA 5′ Cap Structure and Synthetic Fragments of eIF4G and 4E-
BP1 Proteins. J. Mol. Biol. 2002, 319 (3), 615-635.
70.
Grudzien-Nogalska, E.; Stepinski, J.; Jemielity, J.; Zuberek,
J.; Stolarski, R.; Rhoads, R. E.; Darzynkiewicz, E., Synthesis of anti-
reverse cap analogs (ARCAs) and their applications in mRNA transla-
tion and stability. In Translation Initiation: Cell Biology, High-
Throughput Methods, and Chemical-Based Approaches, Lorsch, J., Ed.
2007; Vol. 431, pp 203-227.
53.
Wypijewska, A.; Bojarska, E.; Lukaszewicz, M.; Stepinski,
J.; Jemielity, J.; Davis, R. E.; Darzynkiewicz, E. 7-methylguanosine
diphosphate (m7GDP) is not hydrolyzed but strongly bound by Decap-
ping Scavenger (DcpS) enzymes and potently inhibits their activity. Bi-
ochemistry 2012, 51 (40), 8003-8013.
71.
Pasquinelli, A. E.; Dahlberg, J. E.; Lund, E. Reverse 5′ caps
in RNAs made in vitro by phage RNA polymerases. RNA 1995, 1 (9),
957-967.
72. Mueller, U.; Förster, R.; Hellmig, M.; Huschmann, F.U.; Kastner
A.; Malecki P.; Pühringer S.; Röwer M.; Sparta K.; Steffien M.; Ühlein
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 26
M.; Wilk P.; Weiss M.S. The macromolecular crystallography beam-
lines at BESSY II of the Helmholtz-Zentrum Berlin: Current status and
perspectives, Eur. Phys. J. Plus 2015, 130 (141).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
ACS Paragon Plus Environment

Page 17 of 26
Journal of the American Chemical Society
SYNOPSIS TOC (Word Style “SN_Synopsis_TOC”).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chart 1. Structures of the cap analogs synthesized in this study.
115x168mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 19 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Structure, function, and applications of the mRNA 5′ cap and its chemically modified analogs. (A)
Schematic structure of mRNA and its 5′ end. Sites for triphosphate bridge hydrolysis by major decapping
enzymes are denoted with scissors. (B) The involvement of capꢀrecognizing proteins, eIF4E, DcpS, and
Dcp2, in mRNA translation and degradation. (C) Structure of βꢀSꢀARCA (m27,2′ꢀOGppSpG), a previꢀously
reported phosphorothioate cap analog existing as two Pꢀdiastereomers, both conferring superior
translational properties to mRNA. (D) Structure of the 5′ꢀphosphorothiolate moiety, a nonꢀdiastereogenic
modification incorporated into the mRNA cap structure in this study. (E) The concept of immune system
programming relying on transfection of immature dendritic cells (iDCs) with synthetic 5′ꢀcapped mRNAs
followed by DC maturation.
206x131mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Presence of the 5′ꢀPSL moiety does not significantly alter capꢀbinding by translation initiation
factor eIF4E. (A) Representaꢀtive timeꢀsynchronized fluorescence quenching titration (FQT) curves of
meIF4E with compounds 4, 15, and reference cap analog (m7GpppG). (B) KAS values for capꢀeIF4E
complexes for analogs 1–15 determined by FQT (mean values from triplicates). (C) Strucꢀture of cap analog
9 in complex with meIF4E (colored sticks) overꢀlaid with the structure of m7GpppG–eIF4E (from PDB id:
1L8B) (gray sticks, only the ligand shown for clarity).
207x187mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 21 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Compound 15 potently inhibits hDcpS. (A) Susceptibility of m7GpppG (upper) and analog 15
(lower) to hDcpS analyzed using RPꢀHPLC. m7GpppG is completely hydrolyzed within 15 min, while
compound 15 is resistant. Data for all compounds are shown in Table 1. (B) Enzymatic assays used for
hDcpS inhibitor evaluation: HTS assay based on artificial substrate20 used for initial evaluation of
compounds 1–15; in vitro HPLC assay using unmodified m7GpppG as a substrate; and HeLa cell extract
assay using fluoresceinꢀlabeled cap analog as a substrate. (CꢀE) Dose response curves (mean values from
triplicates) for selected cap analogs and RG3039 obtained in the three assays. (F) XꢀRay structure of cap
analog 15 in complex with hDcpS (residues 38−337). Left: Overall view of hDcpSꢀcompound 15 complex.
Color coding: teal, Cꢀterminal domains; orange, Nꢀterminal domains; purple, hinge regions; yellow sticks,
histidine triads; gray sticks, overlaid m7GpppG ligand from the previously reported complex with H277Nꢀ
hDcpS (PDB id: 1ST0).56 Inset: Closeꢀup view of the catalytic center in the closed state with both ligands.
Additional amino acid contacts observed only for cap analog 15 are marked in yellow, while the direction of
the nucleophilic attack by His277 is marked as a red dashed line. (G) Comparison of ligand conformations
presented in two orientations, showing different conformations and positions of triphosphate chains and
guanosine moieties.
190x194mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. A 5′ꢀPSL moiety next to the m7G (γꢀPSL) protects RNA from decapping by Dcp2. (A) Schematic
representation of the hDcp2 susceptibility assay. (B) Example results from the assay (RNAs resolved by
denaturing polyacrylamide gel electrophoresis (PAGE) and stained with SYBR Gold). (C) Susceptibility of 25ꢀ
nt transcripts capped with various cap analogs to hDcp2. Fraction of capped RNA remaining in the reaction
mixture is plotted as a funcꢀtion of time. Data are reported as the average values from two biological
replicates ± S.D. All data were normalized to time 0.
189x117mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 23 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. mRNAs capped with 5′ꢀPSL cap analogs 6 and 7 are efficiently translated in living cells. (A)
Representative time course of firefly luciferase mRNA expression in HeLa cells for a single biological replicate
when the specified 5′ꢀcap analogs were incorpoꢀrated. (B) Total protein expressions in HeLa cells for various
capped mRNAs. Bars represent integers under the expression curves shown in (A) normalized to m7GpppG.
Data are reported as the average values from three biological replicates ± S.D.
94x48mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthetic routes to guanine and N7ꢀmethylguanine mononucleotide analogs containing a 5′ꢀPSL
moiety
97x31mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 25 of 26
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of dinucleotides 1–6
99x43mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Synthesis of dinucleotides 7–15
161x109mm (300 x 300 DPI)
ACS Paragon Plus Environment