Chemical synthesis and biological activity of novel brominated 7-deazaadenosine-3′,5′-cyclic monophosphate derivatives

Article abstract of DOI:10.1016/j.bmc.2019.03.024

Synthetic derivatives of cyclic adenosine monophosphate, such as halogenated or other more hydrophobic analogs, are widely used compounds, to investigate diverse signal transduction pathways of eukaryotic cells. This inspired us to develop cyclic nucleoti

Full text of DOI:10.1016/j.bmc.2019.03.024

Accepted Manuscript
Chemical synthesis and biological activity of novel brominated 7-deazaadeno-
sine-3’,5’-cyclic monophosphate derivatives
Marco Lelle, Maik Otte, Susanne Thon, Daniela Bertinetti, Friedrich W.
Herberg, Klaus Benndorf
PII:
S0968-0896(19)30217-2
https://doi.org/10.1016/j.bmc.2019.03.024
BMC 14816
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
7 February 2019
9 March 2019
11 March 2019
Please cite this article as: Lelle, M., Otte, M., Thon, S., Bertinetti, D., Herberg, F.W., Benndorf, K., Chemical
synthesis and biological activity of novel brominated 7-deazaadenosine-3’,5’-cyclic monophosphate derivatives,
Bioorganic & Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.bmc.2019.03.024
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Chemical synthesis and biological activity of novel brominated 7-deazaadenosine-
3’,5’-cyclic monophosphate derivatives
Marco Lellea, Maik Ottea, Susanne Thona, Daniela Bertinettib, Friedrich W. Herbergb, Klaus
Benndorfa*
a Institute of Physiology II, University Hospital Jena, Kollegiengasse 9, 07743 Jena, Germany
b
Department of Biochemistry, University of Kassel, Heinrich-Plett-Straße 40, 34132 Kassel,
Germany
*corresponding author
Keywords:
cAMP derivatives, CNG and HCN channels, protein kinase A, Epac1 and Epac2
Abstract:
Synthetic derivatives of cyclic adenosine monophosphate, such as halogenated or other
more hydrophobic analogs, are widely used compounds, to investigate diverse signal
transduction pathways of eukaryotic cells. This inspired us to develop cyclic nucleotides,
which exhibit chemical structures composed of brominated 7-deazaadenines and the
phosphorylated ribosugar. The synthesized 8-bromo- and 7-bromo-7-deazaadenosine-3’,5’-
cyclic monophosphates rank among the most potent activators of cyclic nucleotide-regulated
ion channels as well as cAMP-dependent protein kinase. Moreover, these substances bind
tightly to exchange proteins directly activated by cAMP.
1. Introduction:
Cyclic adenosine monophosphate (cAMP) is an important second messenger produced by
adenylyl cyclases and plays a fundamental role in many biological processes of eukaryotic
cells.1,2 Signal transduction via cAMP occurs intracellularly through its target molecules,
cyclic nucleotide-regulated ion channels, cAMP-dependent protein kinases (PKAs) and
exchange proteins directly activated by cAMP (Epacs), upon binding of first messengers,
such as hormones or neurotransmitters, to receptors in the cell membrane.3-7 These target
molecules share a highly conserved structural domain, the cyclic nucleotide-binding domain
(CNBD), which is involved in the interaction of the protein with cAMP.8 To study the structure
and function of these target proteins, various cyclic nucleotide derivatives with specific
properties, including high binding affinity, hydrolytic stability, the ability to permeate the
plasma membrane or a selectivity for the different CNBDs, are widely applied.9-11 The
brominated cAMP derivative 8-bromoadenosine-3’,5’-cyclic monophosphate (8Br-cAMP) and
7-deazaadenosine-3’,5’-cyclic monophosphate (7C-cAMP) display such characteristics
1

(Figure 1). 8Br-cAMP is a cell-permeable lipophilic analog that efficiently activates cAMP-
dependent protein kinases as well as ion channels in many biological systems.12-14 Moreover,
8Br-cAMP exhibits an increased stability against hydrolysis by different phosphodiesterases,
compared to the natural nucleotide.15 Substitution of the nitrogen atom with a hydrocarbon
motif in 7C-cAMP renders the molecule also more lipophilic and restricts hydrogen bonding.16
This substance has already shown superior binding to numerous target proteins of cAMP.
Mechanistically, this can be attributed to stronger interactions with the hydrophobic
environment of the binding pocket.17,18 However, a molecule containing both the bromination
and the deaza motif has never been investigated. Such a substance would be promising
because the properties of the precursor molecules could be potentially combined or even
enhanced. Furthermore, a higher target specificity of such a novel brominated 7-
deazaadenosine-3’,5’-cyclic monophosphate derivative could help to get deeper insights into
particular signaling pathways.
Figure 1: Chemical structures of adenosine-3’,5’-cyclic monophosphate (cAMP) and the corresponding
derivatives 8Br-cAMP as well as 7C-cAMP.
In the present work, we introduce the synthesis of 8-bromo-7-deazaadenosine-3’,5’-cyclic
monophosphate (8Br-7C-cAMP) and its 7-brominated analog (7Br-7C-cAMP). This
substitution pattern is typically incompatible with the natural nucleotide due to the presence
of the nitrogen atom in the purine ring system. To evaluate the potency of the unknown
brominated cyclic nucleotides, the effect of the compounds on various cyclic nucleotide-
regulated ion channels, cAMP-dependent protein kinase and exchange proteins directly
activated by cAMP is examined and compared to the parent nucleotides. In the course of this
study, activation of ion channels and PKA is investigated electrophysiologically and by a
spectrophotometric assay, respectively. To assess the binding strength to Epac proteins, a
competition assay is conducted, where a high-affinity fluorescent probe is displaced by the
applied nucleotide.
2. Results and discussion:
2.1 Nucleotide synthesis:
The brominated variants of 7C-cAMP were synthesized from their corresponding
halogenated nucleobases by glycosylation and subsequent phosphorylation (Scheme 1).
Such a synthesis route is more suitable than the direct halogenation of 7C-cAMP, since the
bromination of 7-deazaadenine with N-bromosuccinimide at room temperature occurs
2

exclusively at the 7-position, which has been conducted in our laboratory. As a consequence
the 8-bromo derivative would be difficult to access under mild reaction conditions.
Scheme 1: Synthesis of brominated 7-deazaadenosine-3’,5’-cyclic monophosphate derivatives.
Reagents and conditions: (i) 1. N,O-bis(trimethylsilyl)acetamide (1.2 equiv.), acetonitrile, argon, room
temperature, 15 min 2. 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribofuranose (1.2 equiv.), trimethylsilyl
trifluoromethanesulfonate (1.2 equiv.), acetonitrile, argon, 80 °C, 2 h; (ii) 7 M ammonia, methanol, 100
°C, 4 h; (iii) 1. phosphorus oxychloride (2 equiv.), trimethyl phosphate, argon, 0 °C, 1 h 2. 0.08 M
potassium hydroxide, water/acetonitrile (2:3), room temperature. Yields are given in parentheses.
The initial step was the glycosylation of the brominated nucleobases 1a and 1b. This step
was carried out via a one-pot procedure of the silyl-Hilbert-Johnson reaction under
Vorbrüggen conditions, which is typically employed for 7-halogenated 7-deazapurines.19,20
The utilization of 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribofuranose yielded almost entirely the
desired β-anomers of the protected nucleosides (2a, 2b). It is widely accepted that the
halogenation at the 7-position is a prerequisite for the efficient glycosylation, whereas non-
halogenated 7-deazapurine bases nearly resist the modification with the ribosugar under the
same reaction conditions.20,21 If the electron withdrawing substituent is solely at the 8-
position, has never been investigated so far. Our results demonstrate that 8-bromo-6-chloro-
7-deazapurine yields the protected nucleoside in an efficient way as well. Afterwards, the
nucleosides were deprotected with 7 M methanolic ammonia at 100 °C to obtain the
compounds 3a and 3b in high yield. Thereby, the 6-amino group was generated too.19,22 The
final step was the preparation of the cyclic nucleotides directly from their corresponding
nucleosides. This reaction was performed by a modified version of a procedure that is
established for several decades.23,24 To carry out the phosphorylation, the deprotected
nucleosides were reacted in trimethyl phosphate with a small excess of phosphorus
oxychloride under anhydrous conditions. Subsequently, the obtained reaction mixtures were
added to a large amount of aqueous potassium hydroxide in acetonitrile, to achieve
cyclization. Before the isolation of the brominated cyclic nucleotides by liquid
chromatography, low molecular weight impurities were removed through dialysis against
water. The sodium salts of the desired nucleotides were achieved via the addition of a 0.1 M
sodium hydrogen carbonate solution, to yield 7Br-7C-cAMP and 8Br-7C-cAMP. These newly
synthesized halogenated 7-deazaadenosine-3’,5’-cyclic monophosphate derivatives display
3

an unusual and non-natural scaffold, which is more hydrophobic than the one from its parent
nucleotide cAMP.
2.2 Activation of cyclic nucleotide-regulated ion channels:
Cyclic nucleotide-gated (CNG) and hyperpolarization-activated cyclic nucleotide-modulated
(HCN) channels belong to the same superfamily of tetrameric ion channels.3 Nevertheless,
their function is remarkably different, CNG channels elicit receptor potentials in olfaction and
vision, whereas HCN channels serve as electrical pacemaker in special neurons and
cardiomyocytes.25-28 However, both types of channels are activated by the binding of cyclic
nucleotides, such as cAMP, to the CNBD, although HCN channels require a sufficiently
hyperpolarizing membrane voltage as primary activating stimulus. Numerous cyclic
nucleotide derivatives with diverse substitutions in different positions of the purine ring have
been shown to generate an enhanced apparent affinity in CNG and HCN channels as
compared to the natural cyclic nucleotides, among them also the lipophilic analogs of cAMP,
7C-cAMP and 8Br-cAMP.17,29-32
Figure 2: Effect of the novel brominated deaza nucleotides on olfactory channels. (A) Voltage protocol
and CNGA2:CNGA4:CNGB1b current evoked by cAMP. The current amplitude was measured at the
end of the pulse to +10 mV (red arrow). (B) Concentration-activation relationships for heterotetrameric
CNGA2:CNGA4:CNGB1b channels with the five agonists specified in the inset. (C) Concentration-
activation relationships for homotetrameric CNGA2 channels with the same agonists. Data points were
fitted with Equation 1 (experimental section), yielding the EC50 and H values provided by Table 1.
In the present study we investigated the effect of the novel brominated deaza nucleotides on
three different types of ion channels by using the patch clamp technique and compared the
data with those obtained for the precursor nucleotides. In a first approach we tested the
effect of the substances on olfactory CNG channels of natural composition, which are
composed of two CNGA2 subunits, one CNGA4 subunit and one CNGB1b subunit.33 The
currents were measured in inside-out patches at +10 mV, following the voltage protocol
4

indicated in the inset of Figure 2 A. As cAMP, 8Br-cAMP and 7C-cAMP, the two new
derivatives 8Br-7C-cAMP and 7Br-7C-cAMP produced robust currents. To determine the
potency and efficiency of the compounds, full concentration-activation relationships were
determined by applying a series of concentrations for each substance (Figure 2 B). These
relationships were fitted by Equation 1 (experimental section), yielding the concentration of
half maximum activation, EC50, and the Hill coefficient, H (Table 1). The maximum currents
were normalized with respect to the current at saturating cAMP (500 µM). As expected,
compared to cAMP the potency was increased by 8Br-cAMP and further by 7C-cAMP.
Notably, the potency of both 7Br-7C-cAMP as well as 8Br-7C-cAMP was even higher than
that of 7C-cAMP and the potency of 7Br-7C-cAMP exceeded that of 8Br-7C-cAMP.
Moreover, all compounds generated at saturating concentrations maximum efficiency with
regard to cAMP. All Hill coefficients were approximately similar.
Table 1: EC50 of the different agonists and H values derived from the concentration-activation
relationships in CNG channels (n represents the number of measurements).
Heterotetrameric CNGA2:A4:B1b channels
Homotetrameric CNGA2 channels
Compound
EC50 (µM)
4.98 ± 0.15
1.60 ± 0.03
0.36 ± 0.02
0.63 ± 0.02
0.12 ± 0.002
H
n
7
7
4
5
7
EC50 (µM)
53.1 ± 0.6
15.4 ± 0.5
2.08 ± 0.08
6.40 ± 0.13
0.54 ± 0.01
H
n
4
7
6
6
7
cAMP
2.08 ± 0.12
2.10 ± 0.08
1.95 ± 0.17
1.83 ± 0.11
1.84 ± 0.05
2.12 ± 0.06
2.14 ± 0.14
2.27 ± 0.19
2.32 ± 0.10
2.17 ± 0.09
8Br-cAMP
7C-cAMP
8Br-7C-cAMP
7Br-7C-cAMP
Afterwards, we studied the effect of the five compounds on homotetrameric CNGA2
channels, which are known to be significantly less sensitive to cAMP as compared to
heterotetrameric channels.34 The results show that the sequence of potency among the
agonists was fully preserved. Noticeably, the gain in potency for the most effective agonist,
7Br-7C-cAMP, was even higher as compared to heterotetrameric channels (Figure 2 C and
Table 1). Again, all substances were fully efficient at saturating concentrations and the Hill
coefficients were approximately similar. This suggests that the binding sites of the CNGA4
and CNGB1b subunit in the heterotetrameric channels function in a similar manner than
those of the CNGA2 subunits. Together, these results demonstrate that the novel brominated
deaza nucleotides are highly potent and effective agonists in hetero- and homotetrameric
olfactory CNG channels.
5

Figure 3: Activation of HCN2 channels by the novel brominated deaza nucleotides. The novel
brominated deaza nucleotide agonists (C, D) activate the channels similarly as the known agonists (A,
B, E) by both accelerating the activation speed and enhancing the current amplitude at the end of the
hyperpolarizing pulses. (F) Representative concentration-activation relationship in HCN2 channels
with 7C-cAMP as agonist, the most potent binder in HCN2 channels.17 The amplitude of the current
was determined at -80 mV (red arrows in E) and related to the current with saturating cAMP (Imax). The
indicated parameters were obtained by fitting Equation 1 (experimental section).
We further tested whether or not our agonists activate also structurally related
homotetrameric HCN2 channels. These measurements were performed in inside-out patches
as well. The channels were activated from a holding potential of -30 mV by a hyperpolarizing
voltage pulse to -130 mV, followed by a short pulse to -80 mV (inset in Figure 3 A). The
recordings were performed under control conditions first and then in the presence of an
effective concentration of the respective agonist. As typical for cAMP, all substances had a
dual effect, they accelerated the current and increased the amplitude at the end of the
hyperpolarizing pulse (Figure 3 A-E). For the most effective known binder in isolated
monomeric CNBDs of HCN2 channels, 7C-cAMP, a full concentration-activation relationship
was determined, to point out that this and also the related compounds are efficient activators
in full channels (Figure 3 F).17 Half-maximum activation was determined to 15 nM. It should
be noted, however, that the degree of determinateness in these measurements was inferior
as compared to that in CNG channels because of a higher natural variability of HCN channel
currents. This finding is also supported by the bigger error bars, compared to those in the
diagrams of Figure 2 B and C. Therefore, the Hill coefficient in Figure 3 F should be
considered with caution. Despite these limitations, the results show that the novel brominated
deaza nucleotides are also highly effective substances to activate structurally related HCN2
channels.
6

2.3 Activation of cAMP-dependent protein kinase:
Another important target protein of cAMP and its synthetic derivatives in eukaryotic cells is
the cAMP-dependent protein kinase (PKA), which regulates cell division, metabolism and
apoptosis.35 Two regulatory (R) and two catalytic (C) subunits form the inactive state of PKA,
the heterotetrameric holoenzyme. PKA holoenzymes occur in different isoforms named in
respect to their respective R-subunit. The human R-subunits are divided into four isoforms -
RIα, RIβ, RIIα and RIIβ - which differ in biochemical characteristics and cellular distribution.36
However, the α-isoforms are more abundant and are ubiquitously expressed. In the classical
concept of PKA activation, four molecules of cAMP bind cooperatively to the CNBDs of the
two R-subunits of the holoenzyme. This induces the release of the C-subunits into their
catalytically active form, which enables them to phosphorylate nearby substrates.37 Yet, this
concept has been modified recently, using high resolution imaging.38
Within the scope of this study, we tested the effect of the aforementioned synthetic cyclic
nucleotides on the activation of human PKA in vitro. For this purpose, the catalytic activity of
PKA was measured in dependence of the cyclic nucleotide concentration (Figure 4).
Sigmoidal concentration-response fitting of the resulting data sets allowed us to deduce the
apparent activation constants (Kact) for the respective cyclic nucleotides (Table 2). The Kact
corresponds to the cyclic nucleotide concentration that induces half-maximal activation of
PKA.
Figure 4: In vitro activation of PKA holoenzyme type Iα (A) and IIα (B) by cAMP and synthetic analogs.
Inactive PKA holoenzyme was supplemented with increasing concentrations of cyclic nucleotides. The
resulting substrate phosphorylation was followed via a photometric assay. Depicted are the data sets
of one protein preparation (with each cyclic nucleotide), which are representative for the independent
biological duplicates. Each data point depicts the mean  SEM of at least a technical triplicate. The
0 M data points indicate the PKA activity without nucleotide. Table 2 shows the fit-derived apparent
activation constants (Kact) of the cyclic nucleotides.
7

The Kact values for PKA were determined in the nanomolar range, in contrast to the
micromolar affinity of CNG channels and Epac proteins (see below). Both tested PKA
isoforms (Iα and IIα) have nearly identical activation constants for cAMP and 7C-cAMP
(Figure 4, Table 2), which is in contrast to CNG channels (Figure 2, Table 1), but in
agreement with earlier studies on PKA.17,39 The observation that 8Br-cAMP activates PKA at
a 2 to 4-fold lower cyclic nucleotide concentration than cAMP is also in line with published
data and reflects the same tendency as reported above for the CNG channel variants (Figure
2, Table 1).39 Combining the 8Br substitution with the 7C modification did not alter the
activation of PKA: the Kact for 8Br-7C-cAMP was nearly the same as for 8Br-cAMP,
irrespective of the PKA isoform. However, this finding is in line with the missing effect of the
sole 7C substitution on PKA activation and underlines a difference between the CNBDs of
PKA and CNG channels with regard to the effect of a deaza modification in position 7 of the
adenine moiety.
Interestingly, when the bromo substituent is located at the 7-position of the deaza analog
(7Br-7C-cAMP), the cyclic nucleotide has a 2 to 5-times higher activation potency on PKA
than cAMP. For PKA IIα this effect correlates with the observations for 8Br-cAMP and 8Br-
7C-cAMP. However, for PKA Iα the effect of the 7Br substituent appears to be even stronger
than 8Br. It is important to note that the Kact of 7Br-7C-cAMP (~30 nM) measured for PKA Iα
already approaches the lower detection limit of the assay, because titration effects may
occur in this assay with 20 nM PKA used.
Table 2: Apparent activation constants (Kact) for PKA holoenzyme type Iα and IIα. Depicted is the
mean of the activation constant ± SD of two independent biological experiments. For cAMP and 7C-
cAMP similar activation constants have been reported previously.17,40
PKA Iα
PKA IIα
Compound
Kact (µM)
Kact (µM)
cAMP
0.133 ± 0.006
0.056 ± 0.008
0.199 ± 0.045
0.043 ± 0.001
0.028 ± 0.002
0.244 ± 0.003
0.088 ± 0.001
0.238 ± 0.015
0.072 ± 0.017
0.097 ± 0.025
8Br-cAMP
7C-cAMP
8Br-7C-cAMP
7Br-7C-cAMP
2.4 Cyclic nucleotide binding of exchange proteins directly activated by cAMP:
In humans two isoforms of the exchange proteins directly activated by cAMP (Epac) protein
family exist, Epac1 and Epac2, encoded by independent genes, Rapgef3 and Rapgef4,
respectively.40 As the name implies, Epacs are activated by cAMP and act as nucleotide
exchange factor for the Rap subfamily of RAS-like small GTPases (Rap1 and Rap2).41
8

The isoforms display distinct subcellular and tissue-specific localization as well as different
physiological functions. Whereas Epac1 is ubiquitously expressed, Epac2 is enriched in the
nervous system and endocrine tissues. Dysregulated expression of Epac1 has an influence
on several human cancers.42,43 In contrast, Epac2 plays a major role in the potentiation of
glucose induced insulin secretion. Although Epac1 and Epac2 share high structural
homology, they differ in the number of CNBDs: Epac1 contains only one cAMP-binding
domain (cNBD-B), whereas Epac2 has two (cNBD-A and cNBD-B).44
In addition to the experiments with PKA, CNG and HCN channels, we investigated the
effects of the novel synthetic cyclic nucleotides on truncated constructs of the two human
Epac isoforms (Epac1 and Epac2). Both Epac constructs lack the N-terminal DEP
(dishevelled/Egl/pleckstrin) domain, which plays a role in targeting Epac to intracellular
membranes.45,46 Furthermore, the Epac2 construct is deprived of its first CNBD domain,
which has a low cAMP affinity and is not involved in the activation process.40,47 The apparent
affinities of the cyclic nucleotides were measured with a fluorescence polarization (FP)-based
competition assay, in which the FP signal of the fluorophore (8Fluo-cAMP) decreases in
dependence of the Epac binding of the competitor (i.e. the respective cyclic nucleotides).
Sigmoidal concentration-response fitting of the resulting data sets (Figure 5) yielded the
apparent affinities (EC50) of the cyclic nucleotides for the two isoforms of Epac (Table 3). It is
important to note that the EC50 value has a different meaning in this context than presented
for the CNG and HCN channels, since the FP competition experiment can only indirectly
detect the binding of unlabeled cyclic nucleotides. The fit-derived half-maximal concentration
is neither a dissociation nor an activation constant. Therefore, the resulting affinity measured
is indicated as an effective concentration (EC).
Figure 5: Affinity between Epac and cyclic nucleotides. Representative fluorescence polarization (FP)
competition experiments with Epac1 (A) and Epac2 (B) constructs. Increasing concentrations of the
cyclic nucleotides were incubated with 1 nM 8Fluo-cAMP and added to 200 nM of the respective Epac
protein. Normalized fluorescence polarization was plotted against the competing cyclic nucleotide
9

concentration. Each data point represents the mean  SD of triplicates. The most left data points
indicate FP signals for the protein incubated without competing nucleotide. The resulting apparent
EC50 values are provided by Table 3.
For both, Epac1 and Epac2, the apparent affinities of all tested cyclic nucleotides are in the
high nanomolar to low micromolar range (Table 3). With apparent cAMP affinities of 5.6 μM
for Epac1 and 3.9 μM for Epac2, our data are consistent with previous reports.40,47-50
All synthetic analogs in this study showed a higher binding affinity to the Epac isoforms, than
the physiologic activator cAMP, which corresponds, in principal, to the observations with
CNG channels (Table 1) and to some extent with PKA (Table 2). However, the potency of the
bromo substituent and the deaza modification on Epac did not always coincide with the
tendencies observed for the other two protein families.
Unlike PKA, both Epac isoforms seem to recognize the sole deaza modification.
Nevertheless, the affinities of 7C-cAMP were only increased by a factor of 2 compared to
cAMP. 8Br-cAMP, in contrast, even increased the affinity by a factor of 5 compared to
unmodified cAMP. This is in line with a published 8-fold increased affinity.48,50 This effect fits
to previous studies on the catalytic activation of Epac1, where the apparent Kact was 7-times
reduced by the 8Br-cAMP derivative.49 Notably, among the tested proteins, only CNG
channels seem to respond stronger to 7C-cAMP than to 8Br-cAMP. This further
substantiates a potential difference in the binding mode, or mechanism respectively, of cyclic
nucleotides to CNG channel CNBDs as hypothesized above.
Within the presented set of cyclic nucleotides, we identified the 8Br-7C-cAMP derivative as
the most affine binder for Epac where it developed an about 10-times higher binding affinity
compared to cAMP. Although this number reflects a similar increase in potency to that
observed for CNG channels (Figure 2, Table 1), it is yet remarkable that for both PKA and
CNG channels the effect of 7Br-7C-cAMP was even stronger than in the constellation when
the bromo substituent was located at the 8-position of the deaza analog. This was not the
case for the Epac constructs: here the apparent affinities of 7Br-7C-cAMP were found to be
lower than that of 8Br-7C-cAMP and even unchanged compared to 7C-cAMP.
Table 3: Apparent EC50 values for Epac1 and Epac2 isoforms derived from a fluorescence polarization
competition assay. At least two independent protein preparations were used for each assay and each
assay was performed in triplicate.
Epac1
Epac2
Compound
EC50 (µM)
5.62 ± 0.51
1.09 ± 0.01
2.75 ± 0.59
EC50 (µM)
3.87 ± 0.54
0.78 ± 0.11
2.06 ± 0.34
cAMP
8Br-cAMP
7C-cAMP
10

8Br-7C-cAMP
7Br-7C-cAMP
0.53 ± 0.07
2.54 ± 0.25
0.30 ± 0.07
2.36 ± 0.82
3. Conclusions:
In summary, we have presented the synthesis and functional characterization of two novel
cyclic nucleotides. Both substances were prepared from their corresponding nucleobases
under the same reaction conditions. Thus, we demonstrated that an electron withdrawing
bromo substituent in 8-position enables the efficient glycosylation by the established one-pot
procedure of the silyl-Hilbert-Johnson reaction under Vorbrüggen conditions as well. The
final synthesis step, the phosphorylation and subsequent cyclization, is always challenging,
also due to the reaction medium. The utilized trialkyl phosphates cause side reactions and
hamper the purification. Our approach employed for the first time dialysis prior to purification,
to remove undesired phosphates, which simplified the isolation of the products by liquid
chromatography.
Figure 6: Effect patterns of the investigated cyclic nucleotides in CNG channels, PKA and Epac
proteins expressed as EC50 and Kact values.
The brominated deaza nucleotides are highly efficient activators of cyclic nucleotide-
regulated ion channels, 7Br-7C-cAMP is 100-times more potent on homotetrameric CNGA2
channels than natural cAMP. This molecule can also activate PKA Iα in the low nanomolar
range. In contrast, Epac proteins respond to this substance like to the natural ligand and 7C-
cAMP, but are highly sensitive to 8Br-7C-cAMP. This finding is also supported by Figure 6,
which demonstrates that modifications in 7-position do not alter the affinity of the cyclic
nucleotide towards Epac1 and Epac2. A similar pattern is observed for PKA, with the
exception of 7Br-7C-cAMP and PKA Iα. Astonishingly, the opposite is the case for CNG
11

channels, where modifications in 7-position induce a bigger effect. Although there is no
perfect selectivity of the novel brominated deaza nucleotides for one of their target proteins,
there are markedly different patterns of the effects, which can help to learn more about the
corresponding CNBDs.
The herein described observation of different cyclic nucleotide recognition patterns by three
different protein families could be a valuable starting point for designing cyclic nucleotide
analogs, to specifically target only one protein family, while leaving other cAMP-dependent
proteins unaffected. Due to the presence of bromo substituents, further modification is
enabled. However, for this purpose it would be crucial to obtain structural models of the
proteins with the respective analogs bound, to map the key residues for the observed
differences in the interaction patterns.
4. Experimental section:
4.1 General information:
Chemicals, including solvents and reagents, were bought from commercial sources and
applied without further purification, unless otherwise stated. Cyclic nucleotides were
purchased from BIOLOG Life Science Institute (Bremen, Germany). Biotech cellulose ester
dialysis tubing (MWCO: 100-500 Da) was obtained from Repligen (Ravensburg, Germany).
Thin layer chromatography sheets (ALUGRAM SIL G/UV254) as well as silica gel for column
chromatography (0.04-0.063 mm) were ordered from Macherey-Nagel (Düren, Germany)
and used with appropriate solvent systems.
Reversed-phase high-performance liquid chromatography (RP-HPLC) was conducted on a
Agilent Technologies 1100 Series system (Waldbronn, Germany) with suitable solvent
delivery pumps (G1361A), a dual loop autosampler (G2258A) and a multi-wavelength
detector (G1365B). Analytical RP-HPLC was carried out on a AppliChrom (Oranienburg,
Germany) OTU LipoMare C18 column (250 x 4.6 mm) with 5 µm particle size as stationary
phase and a flow rate of 1 mL/min. Purification of the cyclic nucleotides was performed on a
AppliChrom OTU LipoMare C18 column (250 x 20 mm) at an appropriate flow rate of 15
mL/min and 5 µm silica as stationary phase. Product fractions were isolated with a fraction
collector (G1364C). The applied eluents were 25 mM (pH 7) triethylammonium acetate buffer
(A) as well as acetonitrile (B) and were utilized with a linear gradient from 0 min (100% A) to
25 min (75% A). All substances were simultaneously detected at 230 and 260 nm.
UV-Vis absorption spectra were measured on a Analytik Jena SPECORD 250 UV-Vis
spectrophotometer (Jena, Germany). Measurements were accomplished at ambient
temperature in quartz cuvettes utilizing aqueous solutions of the cyclic nucleotides.
1H, 13C and 31P NMR spectra were recorded at 300 K on a Bruker Avance I 300 MHz
spectrometer (Karlsruhe, Germany) or at 297 K on a Bruker Avance III 400 MHz
12

spectrometer. Chemical shifts are reported in parts per million, relative to the residual solvent
signals of DMSO-d6 and D2O or DSS-d6 and phosphoric acid.51 Coupling constants (J) are
given in hertz (Hz).
High-resolution electrospray ionization mass spectrometry measurements were carried out
on a Bruker Daltonics micrOTOF system (Bremen, Germany), equipped with an automatic
syringe pump for sample injection. The standard electrospray ion source was used to
generate ions and the instrument was calibrated in the m/z range 50 to 3000 utilizing an
internal calibration standard (Tunemix solution) from Agilent Technologies.
4.2 Chemical syntheses
4.2.1 Synthesis of compound 2a:
N,O-Bis(trimethylsilyl)acetamide (488.2 mg, 586.8 µL, 2.40 mmol) was added to a
suspension of 7-bromo-6-chloro-7-deazapurine (464.9 mg, 2.00 mmol) in 20 mL dry
acetonitrile. Afterwards, the solution was stirred for 15 min at room temperature in an argon
atmosphere, followed by the addition of 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribofuranose
(1.21 g, 2.40 mmol) and trimethylsilyl trifluoromethanesulfonate (533.4 mg, 434.4 µL, 2.40
mmol). The reaction mixture was heated for 2 h at 80 °C under argon with vigorous stirring.
After glycosylation, the solution was left to cool down and diluted with 400 mL ethyl acetate.
The organic phase was sequentially washed twice with a saturated sodium hydrogen
carbonate solution and brine. Subsequently, the organic layer was dried with magnesium
sulfate and the solvent was removed under vacuum, to afford the crude nucleoside.
Purification by column chromatography on silica gel with hexane/ethyl acetate (2:1) as eluent
1
yielded the product as colorless solid (864.0 mg, 1.28 mmol, 64%). H NMR (300 MHz,
DMSO-d6) δ 8.62 (1H, s), 8.32 (1H, s), 8.00-7.84 (6H, m), 7.69-7.61 (3H, m), 7.54-7.40 (6H,
m), 6.73 (1H, d, J = 5.0 Hz), 6.32-6.28 (1H, m), 6.16-6.11 (1H, m), 4.90-4.85 (1H, m), 4.84-
4.65 (2H, m); 13C NMR (75 MHz, DMSO-d6) δ 165.4, 164.7, 164.5, 151.4, 151.0, 150.3,
134.0, 133.9, 133.6, 129.4, 129.4, 129.3, 129.2, 129.2, 128.8, 128.6, 128.2, 114.9, 88.1,
86.6, 79.3, 73.5, 70.6, 63.4; HRMS (ESI): m/z calcd for C32H24BrClN3O7 [M + H]+ 676.0481,
found 676.0489
4.2.2 Synthesis of compound 2b:
This nucleoside derivative was prepared according to the synthesis of 2a with 8-bromo-6-
chloro-7-deazapurine (464.9 mg, 2.00 mmol), to afford the product as colorless solid (931.5
mg, 1.38 mmol, 69%). 1H NMR (300 MHz, DMSO-d6) δ 8.55 (1H, s), 7.97-7.84 (6H, m), 7.70-
7.60 (3H, m), 7.53-7.40 (6H, m), 7.10 (1H, s), 6.72-6.63 (1H, m), 6.50 (1H, d, J = 3.8 Hz),
6.45-6.35 (1H, m), 4.95-4.87 (1H, m), 4.84-4.55 (2H, m); 13C NMR (75 MHz, DMSO-d6) δ
165.3, 164.7,164.6, 150.8, 150.8, 150.0, 134.0, 133.9, 133.5, 129.4, 129.3, 129.2, 129.1,
13

128.8, 128.7, 128.6, 128.5, 128.3, 118.2, 117.6, 103.5, 88.3, 78.6, 72.8, 69.9, 62.4; HRMS
(ESI): m/z calcd for C32H23BrClN3NaO7 [M + Na]+ 698.0300, found 698.0309
4.2.3 Synthesis of compound 3a:
The protected nucleoside 2a (473.8 mg, 700 µmol) was suspended in 30 mL 7 M ammonia in
methanol and heated at 100 °C for 4 h in a sealed reaction vessel with stirring. After that, the
solution was cooled to ambient temperature and the solvent was evaporated under reduced
pressure. The obtained residue was purified by silica gel column chromatography using
dichloromethane/methanol (9:1) as eluent, to yield the product as colorless solid (174.0 mg,
1
504 µmol, 72%). H NMR (300 MHz, DMSO-d6) δ 8.10 (1H, s), 7.65 (1H, s), 6.82 (2H, br s),
6.04 (1H, d, J = 6.2 Hz), 5.35 (1H, d, J = 6.5 Hz), 5.25-5.07 (2H, m), 4.41-4.29 (1H, m), 4.12-
4.02 (1H, m), 3.93-3.83 (1H, m), 3.67-3.47 (2H, m); 13C NMR (75 MHz, DMSO-d6) δ 157.0,
152.4, 149.6, 121.8, 101.1, 86.8, 86.7, 85.2, 73.9, 70.5, 61.5; HRMS (ESI): m/z calcd for
C11H14BrN4O4 [M + H]+ 345.0193, found 345.0198
4.2.4 Synthesis of compound 3b:
This nucleoside derivative was prepared according to the synthesis of 3a with 2b (464.9 mg,
700 µmol), to afford the product as colorless solid (169.1 mg, 490 µmol, 70%). 1H NMR (300
MHz, DMSO-d6) δ 8.02 (1H, s), 7.28 (2H, br s), 6.80 (1H, s), 5.85 (1H, d, J = 6.8 Hz), 5.77-
5.69 (1H, m), 5.32 (1H, d, J = 6.5 Hz), 5.18-5.04 (2H, m), 4.21-4.13 (1H, m), 3.97-3.90 (1H,
m), 3.71-3.46 (2H, m); 13C NMR (75 MHz, DMSO-d6) δ 156.6, 151.5, 149.6, 109.1, 103.9,
102.9, 90.0, 86.1, 71.2, 70.9, 62.4; HRMS (ESI): m/z calcd for C11H14BrN4O4 [M + H]+
345.0193, found 345.0195
4.2.5 Synthesis of compound 7Br-7C-cAMP:
Compound 3a (86.3 mg, 250 µmol) was dried in a high vacuum prior to phosphorylation and
dissolved in 1.5 mL freshly distilled trimethyl phosphate by heating. Afterwards, freshly
distilled phosphorus oxychloride (76.7 mg, 46.5 µL, 500 µmol) was added and the reaction
mixture was stirred in an ice bath for 1 h under argon. Subsequently, the solution was poured
into 30 mL 0.08 M potassium hydroxide in water/acetonitrile (2:3) with stirring and neutralized
by the addition of 1 M hydrochloric acid. The volume was reduced and the residue was
diluted with 100 mL 0.1 M sodium hydrogen carbonate. Then, the solution was dialyzed
against water and freeze dried. The obtained solid was taken up in 25 mM triethylammonium
acetate buffer and the hydrolyzed acid chloride was isolated by RP-HPLC. The solvent of the
collected fractions was removed under reduced pressure and the solid was further
desiccated in a high vacuum. After that, the residue was dissolved in a small amount of 0.1
M sodium hydrogen carbonate, dialyzed against water and lyophilized, to afford the colorless
14

product as sodium salt (21.4 mg, 50 µmol, 20%). UV-Vis (H2O) λmax (ε) = 278 nm (7750 M-1
cm-1); 1H NMR (400 MHz, D2O) δ 8.01 (1H, s), 7.27 (1H, s), 6.20 (1H, s), 4.63-4.46 (3H, m),
4.34-4.25 (2H, m); 13C NMR (75 MHz, D2O) δ 156.7, 152.2, 148.1, 121.4, 101.6, 91.0, 89.1,
77.7 (d, J = 4.7 Hz), 72.5 (d, J = 7.9 Hz), 71.3 (d, J = 4.2 Hz), 67.4 (d, J = 6.7 Hz); 31P NMR
(162 MHz, D2O) δ -1.65; HRMS (ESI): m/z calcd for C11H11BrN4O6P [M - Na]- 404.9605,
found 404.9602
4.2.6 Synthesis of compound 8Br-7C-cAMP:
This nucleotide derivative was prepared according to the synthesis of 7Br-7C-cAMP with 3b
(86.3 mg, 250 µmol), to afford the product as colorless solid (22.5 mg, 53 µmol, 21%). UV-
1
Vis (H2O) λmax (ε) = 277 nm (14050 M-1 cm-1); H NMR (400 MHz, D2O) δ 8.09 (1H, s), 6.63
(1H, s), 6.17 (1H, d, J = 1.4 Hz), 5.34-5.25 (1H, m), 5.10-5.03 (1H, m), 4.54-4.42 (1H, m),
4.33-4.25 (1H, m), 4.24-4.15 (1H, m); 13C NMR (75 MHz, D2O) δ 155.9, 151.8, 150.0, 109.0,
104.0, 103.8, 92.9, 76.9 (d, J = 4.3 Hz), 71.5 (d, J = 4.4 Hz), 71.0 (d, J = 8.0 Hz), 67.2 (d, J =
7.0 Hz); 31P NMR (162 MHz, D2O) δ -1.45; HRMS (ESI): m/z calcd for C11H11BrN4O6P [M -
Na]- 404.9605, found 404.9596
4.3 Molecular biology and heterologous expression of CNG and HCN channels:
The subunits CNGA2 (accession No. AF126808), CNGA4 (accession No. U12623) and
CNGB1b (accession No. AF068572) of rat olfactory channels as well as of mouse HCN2
channels (NM008226) were subcloned prior to the T7 promoter of pGEMHEnew. The
respective cRNAs were produced by using the mMESSAGE mMACHINE T7 Kit (Ambion,
Austin, TX, USA).
Oocytes of Xenopus laevis were harvested either surgically under anesthesia (0.3% 3-
aminobenzoic acid ethyl ester) from female adults or purchased from Ecocyte® (Castrop-
Rauxel, Germany). The procedures had approval from the authorized animal ethical
committee of the Friedrich Schiller University Jena and were carried out in accordance with
the approved guidelines.
The oocytes were incubated for 105 min in Ca2+-free Barth’s medium containing collagenase
A (3 mg/ml; Roche, Grenzach-Wyhlen, Germany) and (in mM) 82.5 NaCl, 2 KCl, 1 MgCl 2,
and 5 Hepes, pH 7.4. Oocytes at stages IV and V were injected with 50-130 ng cRNA
encoding either CNGA2, CNGA2:CNGA4:CNGB1b (2:1:1 ratio) or HCN2 channels either by
hand or by means of an injection robot (RoboInject®). The injected oocytes were then
incubated at 18 °C for up to 6 days in Barth’s solution containing (in mM) 84 NaCl, 1 KCl, 2.4
NaHCO3, 0.82 MgSO4, 0.41 CaCl2, 0.33 Ca(NO3)2, 7.5 TRIS, cefuroxime (4.0 µg×ml-1),
penicillin/streptomycin (100 µg×ml-1), pH 7.4.
15

4.4 Electrophysiology:
Macroscopic currents containing hundreds to several thousand of channels were recorded
from inside-out patches of the oocytes by the patch-clamp technique. The patch pipettes
were pulled from quartz tubing (P-2000, Sutter Instrument, Novato, USA) with an outer and
inner diameter of 1.0 and 0.7 mm (VITROCOM, New Jersey, USA). The corresponding
pipette resistance was 0.9-2.3 M. The bath and pipette solution contained (in mM): 150
KCl, 1 EGTA, 5 Hepes (pH 7.4 with KOH). Recording was performed by an Axopatch 200B
amplifier (Axon Instruments, Foster City, CA, USA). Electrophysiology was controlled by the
Patchmaster-software (HEKA Elektronik Dr. Schulze GmbH, Lambrecht, Germany). The
sampling rate was 5 kHz and the filter implemented in the amplifier (4-pole Bessel) was set
to 2 kHz. Measurements in HCN2 channels were begun 3.5 min after patch excision to
minimize channel run down during the measurement.52 The solutions with the different ligand
concentrations to be tested were applied via a multi-barrel device to the patches with a flow
rate of 0.8 to 1.2 mL/min. All experiments were carried out at room temperature.
4.5 Fitting steady-state concentration-activation relationships (CNG and HCN channels):
Concentration-activation relationships were fitted with the Igor software by the following
formula (Equation 1):
I/Imax = 1/(1+(EC50/[nucleotide])H)
(1)
where I is the actual current amplitude and Imax the maximum current amplitude at saturating
concentration specified for each cyclic nucleotide derivative separately. EC50 is the
concentration generating the half maximum current and H the Hill coefficient. Errors are
given as mean ± s.e.m.
4.6 Expression and purification of catalytic and regulatory subunits of PKA:
The human catalytic subunit (Cα1) and the two main human regulatory subunits (RIα, RIIα)
of PKA were recombinantly expressed in E. coli, purified via Sp-8-AEA-cAMPS agarose and
characterized as described previously.53 In vitro formation of PKA holoenzymes (i.e.
association of the heterotetramer R2C2) was performed as reported.40
4.7 PKA activity assay:
The cyclic nucleotide-dependent activity of PKA was monitored in vitro, using a coupled
spectrophotometric assay described by Cook et al..54 Each substrate phosphorylation event
catalyzed by an active PKA C-subunit produces one ADP molecule, which is regenerated to
ATP by a pyruvate kinase, in turn releasing pyruvate. Pyruvate is converted to lactic acid by
a lactate dehydrogenase under the consumption of NADH+H+. The oxidation of NADH+H+ is
stoichiometric with substrate phosphorylation by PKA and can be followed photometrically at
16

340 nm. The enzyme activity refers to the initial reaction rate (v0). The concentration of the
commonly used synthetic PKA substrate Kemptide (LRRASLG; GeneCust, Dudelange,
Luxembourg) was adopted from previous studies (260 µM).55 The cyclic nucleotide analogs
were applied in a concentration range of 100 pM to 1 mM. The concentration of PKA
holoenzyme was adapted to the specific activity of the catalytic subunit preparation (15 U/mg
active catalytic subunit), employing approximately 10 nM of holoenzyme in the assay. All
data points were obtained at least in triplicates and analyzed with GraphPad Prism 6.01
(GraphPad Software Inc., La Jolla, California). For each cyclic nucleotide, v0 was normalized
to the smallest (0%) and largest (100%) mean of the replicates. The apparent activation
constants (Kact) were determined by fitting the concentration-dependent normalized activity
with a sigmoid dose-response model (where top and bottom values were constrain to 100%
and 0%, respectively). In addition to the technical triplicate of each data point, the full data
set of each cyclic nucleotide analog was statistically verified by a biological duplicate, i.e.
with independent protein expressions, purifications and PKA holoenzyme formations.
4.8 Expression and purification of Epac constructs:
Expression plasmids with human His-tagged Epac1 157-881 (referred to as Epac1) were
transformed into E. coli TP2000 Δcya (gift from C. Kim, Baylor College of Medicine, Houston,
TX, USA) by electroporation. 20 mL seed cultures were grown overnight at 37 °C and
180 rpm. Main cultures were inoculated 1:500 with seed culture and incubated at room
temperature and 180 rpm until an OD600 of 0.6-0.8 was reached. Expression was induced by
adding 400 μM isopropyl β-D-1-thiogalactopyranoside and proteins were expressed
overnight at room temperature. Cells were harvested by centrifugation at 9,000 xg for 20 min
and cell pellets were stored at -20 °C until they were further processed.
Cell pellets were resuspended in buffer A (20 mM NaH2PO4, 300 mM NaCl2, 5 mM 2-
mercaptoethanol, pH8) containing cOmplete EDTA-free protease inhibitor cocktail (Roche)
and 1 mM PMSF. After homogenization, the suspension was lysed by 3 rounds of French
pressure cell (Thermo Electron Corp.) treatment, applying a pressure of ~1,000 bar. The cell
debris was spun down at 40,000 xg for at least 30 min at 4 °C. The supernatant was passed
through a 0.45 µm PVDF syringe-filter (Roth, Karlsruhe, Germany).
The purification was performed at 4°C with following FPLC assembly: ÄKTA purifier UPC 10,
fraction collector Frac-920, software Unicorn 5.31 (all GE Healthcare). The filtered
supernatant was loaded on a 1 ml Protino Ni-NTA column (Macherey-Nagel) via a superloop
(GE Healthcare). The purification was carried out at a flow rate of up to 1 mL/min and a
pressure limit of 0.7 MPa. After loading the lysate, the column was washed with 10 column
volumes of 20 mM imidazole in buffer A. Subsequently, His-tagged protein was eluted with
17

gradually increasing imidazole concentrations: from 40 to 250 mM imidazole, linearly
distributed over 20 column volumes. 1 mL fractions were collected.
Pooled elution fractions were concentrated and buffer exchanged via Amicon Ultra-15 30K
Centrifugal Filter Units (Merck Millipore), 2,500 xg, 50 min. Afterwards, 2 mL of the purified
His-tagged Epac1 157-881 were subjected to gel filtration on a Superdex 75 16/60 column
(GE Healthcare) in 20 mM MOPS, 150 mM NaCl, 5 mM 2-mercaptoethanol, pH 7 at a flow
rate of 2 mL/min and a pressure limit of 0.8 MPa. 2 mL fractions were collected. The column
was calibrated using the Gel Filtration Standard (Bio-Rad) according to the manufacturer's
protocol. SDS-PAGE was used to monitor protein expression and purification, in order to
ensure a homogeneity of ≥95%.
Mus musculus Epac2 ΔDEP (aa 280-998, referred to as Epac2)56 was kindly provided by H.
Rehmann (UMC Utrecht, Netherlands).
Epac concentrations were determined according to a calibration curve with Bovine Serum
Albumin (Pierce) using Bradford reagent (Bio-Rad).
4.9 Epac binding assay:
In order to measure the binding affinities of the synthetic cyclic nucleotides, a fluorescence
polarization-based competition assay, previously developed for PKA,57 was adapted to Epac.
The concentrations of both protein and 8Fluo-cAMP were fixed, while the concentration of
the competing cyclic nucleotide was varied over several orders of magnitude. When 8Fluo-
cAMP is bound to the protein, its rotational speed is slowed down in comparison to unbound
8Fluo-cAMP, which can be quantified according to the fluorescence polarization signal.
The FP measurements were performed in 150 mM NaCl, 20 mM MOPS with 0.005% (v/v)
CHAPS, pH 7 at room temperature. All data points were obtained in triplicates using a
CLARIOstar plate reader (BMG LABTECH). Data point acquisition took place with a rate of
10 s-1 at an excitation and emission wavelength of 482 nm and 520 nm, respectively, in black
384 well microtiter plates (BRANDplate, BRAND GMBH + CO KG).
Prior to determination of the binding affinities of the target analogs via a competition assay,
the interaction between a fluorescently labeled cAMP analog (8Fluo-cAMP, 8-[[2-
[(fluoresceinylthioureido)amino]ethyl]thio]-cAMP, BIOLOG Life Science Institute) and the
Epac isoforms was analyzed (supporting information). In this direct assay format, the
respective Epac isoform was varied (120 pM to 2 µM), whereas the 8Fluo-cAMP
concentration was held constant (1 nM). For the FP competition assay, increasing
concentrations of cAMP and target analog (ranging from 100 pM to 1 mM) were mixed with
1 nM 8Fluo-cAMP before addition of the respective Epac isoform. The Epac concentration
(200 nM) was adapted to the results of the direct interaction assay, in order to reach ~50% of
the maximum FP signal (~100 mPol). The FP measurement was started after 5 minutes of
18

incubation. For each cyclic nucleotide, the FP signal was normalized to the smallest (0%)
and largest (100%) mean of the replicates. Apparent EC50 values were calculated by fitting
with a nonlinear sigmoid dose-response model (where top and bottom values were
constrained to 100% and 0%, respectively) with the software GraphPad Prism 6.01. In
addition to the technical triplicate of each data point, a biological duplicate, i.e. with an
independent protein expression and purification, statistically verified the full data set of each
cyclic nucleotide analog.
Conflicts of interest:
There are no conflicts to declare.
Acknowledgements:
We thank K. Schoknecht, S. Bernhardt, A. Kolchmeier, U. Enke and C. Ranke from the
University Hospital Jena as well as M. Hansch and O. Bertinetti from the University of Kassel
for excellent technical assistance and L.L. Kailing (University of Kassel) for critically reading
and constructive discussions on the manuscript. Moreover, we would like to acknowledge the
NMR platform at Friedrich Schiller University Jena for support in NMR spectroscopy as well
as the mass spectrometry facility of Professor U.S. Schubert from the Friedrich Schiller
University Jena. This work was supported by the Deutsche Forschungsgemeinschaft (TRR
166 ReceptorLight, project A5 to K.B., and He 1818/10 to F.W.H.).
Supplementary data:
Supplementary data contain acquired NMR data, including the corresponding structure
assignments, UV-Vis absorption spectra, RP-HPLC chromatograms, high-resolution mass
spectra as well as additional information regarding the Epac binding assay. Supplementary
data associated with this article can be found, in the online version, at …
References:
1. Beavo JA, Brunton LL. Cyclic nucleotide research - still expanding after half a century. Nat
Rev Mol Cell Biol. 2002;3:710-718.
2. Bassler J, Schultz JE, Lupas AN. Adenylate cyclases: Receivers, transducers, and
generators of signals. Cell Signal. 2018;46:135-144.
3. Craven KB, Zagotta WN. CNG and HCN channels: Two peas, one pod. Annu Rev Physiol.
2006;68:375-401.
4. Kaupp UB, Seifert R. Cyclic nucleotide-gated ion channels. Physiol Rev. 2002;82:769-824.
5. Taylor SS, Zhang P, Steichen JM, Keshwani MM, Kornev AP. PKA: Lessons learned after
twenty years. BBA-Proteins Proteom. 2013;1834:1271-1278.
19

6. Bos JL. Epac: a new cAMP target and new avenues in cAMP research. Nat Rev Mol Cell
Biol. 2003;4:733-738.
7. Schmidt M, Dekker FJ, Maarsingh H. Exchange Protein Directly Activated by cAMP
(epac): A Multidomain cAMP Mediator in the Regulation of Diverse Biological Functions.
Pharmacol Rev. 2013;65:670-709.
8. Berman HM, Ten Eyck LF, Goodsell DS, Haste NM, Kornev A, Taylor SS. The cAMP
binding domain: An ancient signaling module. Proc Natl Acad Sci. 2005;102:45-50.
9. Poppe H, Rybalkin SD, Rehmann H, Hinds TR, Tang XB, Christensen AE, Schwede F,
Genieser H-G, Bos JL, Doskeland SO, Beavo JA, Butt E. Cyclic nucleotide analogs as
probes of signaling pathways. Nat Methods. 2008;5:277-278.
10. Schwede F, Maronde E, Genieser H-G, Jastorff B. Cyclic nucleotide analogs as
biochemical tools and prospective drugs. Pharmacol Ther. 2000;87:199-226.
11. Scott SP, Shea PW, Dryer SE. Mapping ligand interactions with the hyperpolarization
activated cyclic nucleotide modulated (HCN) ion channel binding domain using a soluble
Construct. Biochemistry. 2007;46:9417-9431.
12. Bartsch M, Zorn-Kruppa M, Kuhl N, Genieser H-G, Schwede F, Jastorff B. Bioactivatable,
membrane-permeant analogs of cyclic nucleotides as biological tools for growth control of C6
glioma cells. Biol Chem. 2003;384:1321-1326.
13. Blasic JR, Brown RL, Robinson PR. Phosphorylation of Mouse Melanopsin by Protein
Kinase A. PLoS ONE. 2012;7.
14. Huang H, Trussell LO. Presynaptic HCN Channels Regulate Vesicular Glutamate
Transport. Neuron. 2014;84:340-346.
15. Sandberg M, Butt E, Nolte C, Fischer L, Halbrugge M, Beltman J, Jahnsen T, Genieser
H-G, Jastorff B, Walter U. Characterization of Sp-5,6-Dichloro-1-Beta-D-Ribofuranosyl-
Benzimidazole-3',5'-Monophosphorothioate (Sp-5,6-Dcl-Cbimps) as a Potent and Specific
Activator of Cyclic-Amp-Dependent Protein-Kinase in Cell-Extracts and Intact-Cells. Biochem
J. 1991;279:521-527.
16. Werner K, Schwede F, Genieser H-G, Geiger J, Butt E. Quantification of cAMP and
cGMP analogs in intact cells: pitfalls in enzyme immunoassays for cyclic nucleotides.
Naunyn Schmiedebergs Arch Pharmacol. 2011;384:169-176.
17. Möller S, Alfieri A, Bertinetti D, Aquila M, Schwede F, Lolicato M, Rehmann H, Moroni A,
Herberg FW. Cyclic Nucleotide Mapping of Hyperpolarization-Activated Cyclic Nucleotide-
Gated (HCN) Channels. ACS Chem Biol. 2014;9:1128-1137.
18. Sartiani L, Mannaioni G, Masi A, Romanelli MN, Cerbai E. The Hyperpolarization-
Activated Cyclic Nucleotide-Gated Channels: from Biophysics to Pharmacology of a Unique
Family of Ion Channels. Pharmacol Rev. 2017;69:354-395.
20

19. Seela F, Peng XH. 7-functionalized 7-deazapurine ribonucleosides related to 2-
aminoadenosine, guanosine, and xanthosine: Glycosylation of pyrrolo[2,3-d]pyrimidines with
1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose. J Org Chem. 2006;71:81-90.
20. Seela F, Ming X. 7-Functionalized 7-deazapurine beta-D and beta-L-ribonucleosides
related to tubercidin and 7-deazainosine: glycosylation of pyrrolo[2,3-d] pyrimidines with 1-O-
acetyl-2,3,5-tri-O-benzoyl-beta-D or beta-L-ribofuranose. Tetrahedron. 2007;63:9850-9861.
21. Ingale SA, Leonard P, Seela F. Glycosylation of Pyrrolo[2,3-d]pyrimidines with 1-O-
Acetyl-2,3,5-tri-O-benzoyl-β-d-ribofuranose: Substituents and Protecting Groups Effecting the
Synthesis of 7-Deazapurine Ribonucleosides. J Org Chem. 2018;83:8589-8595.
22. De Ornellas S, Slattery JM, Edkins RM, Beeby A, Baumann CG, Fairlam IJS. Design and
synthesis of fluorescent 7-deazaadenosine nucleosides containing pi-extended
diarylacetylene motifs. Org Biomol Chem. 2015;13:68-72.
23. Genieser H-G, Butt E, Bottin U, Dostmann W, Jastorff B. Synthesis of the 3',5'-Cyclic
Phosphates from Unprotected Nucleosides. Synthesis. 1989;53-54.
24. Roy B, Depaix A, Perigaud C, Peyrottes S. Recent Trends in Nucleotide Synthesis.
Chem Rev. 2016;116:7854-7897.
25. Nache V, Wongsamitkul N, Kusch J, Zimmer T, Schwede F, Benndorf K. Deciphering the
function of the CNGB1b subunit in olfactory CNG channels. Sci Rep. 2016;6.
26. Shuart NG, Haitin Y, Camp SS, Black KD, Zagotta WN. Molecular mechanism for 3:1
subunit stoichiometry of rod cyclic nucleotide-gated ion channels. Nat Commun. 2011;2.
27. Peng CH, Rich ED, Varnum MD. Subunit configuration of heteromeric cone cyclic
nucleotide-gated channels. Neuron. 2004;42:401-410.
28. Thon S, Schmauder R, Benndorf K. Elementary Functional Properties of Single HCN2
Channels. Biophys J. 2013;105:1581-1589.
29. Chen L, Xu R, Sun FJ, Xue Y, Hao XM, Liu HX, Wang H, Chen XY, Liu ZR, Deng WS,
Han XH, Xie JX, Yung WH. Hyperpolarization-activated cyclic nucleotide-gated (HCN)
channels regulate firing of globus pallidus neurons in vivo. Mol Cell Neurosci. 2015;68:46-55.
30. V H, Frings S, Wiesner B, Helm S, Kaupp UB, Bendig J. [7-(dialkylamino)coumarin-4-
yl]methyl-caged compounds as ultrafast and effective long-wavelength phototriggers of 8-
bromo-substituted cyclic nucleotides. ChemBioChem. 2003;4:434-442.
31. Otte M, Schweinitz A, Bonus M, Enke U, Schumann C, Gohlke H, Benndorf K.
Hydrophobic alkyl chains substituted to the 8-position of cyclic nucleotides enhance
activation of CNG and HCN channels by an intricate enthalpy - entropy compensation. Sci
Rep. 2018;8.
32. Brown RL, Strassmaier T, Brady JD, Karpen JW. The pharmacology of cyclic nucleotide-
gated channels: Emerging from the darkness. Curr Pharm Des. 2006;12:3597-3613.
21

33. Bonigk W, Bradley J, Muller F, Sesti F, I B, Ronnett GV, Kaupp UB, Frings S. The native
rat olfactory cyclic nucleotide-gated channel is composed of three distinct subunits. J
Neurosci. 1999;19:5332-5347.
34. Nache V, Zimmer T, Wongsamitkul N, Schmauder R, Kusch J, Reinhardt L, Bonigk W,
Seifert R, Biskup C, Schwede F, Benndorf K. Differential Regulation by Cyclic Nucleotides of
the CNGA4 and CNGB1b Subunits in Olfactory Cyclic Nucleotide-Gated Channels. Sci
Signal. 2012;5.
35. Skalhegg BS, Tasken K. Specificity in the cAMP/PKA signaling pathway. Differential
expression, regulation, and subcellular localization of subunits of PKA. Front Biosci.
2000;5:D678-D693.
36. Soberg K, Skalhegg BS. The Molecular Basis for Specificity at the Level of the Protein
Kinase a Catalytic Subunit. Front Endocrinol. 2018;9.
37. England JP, Hao YX, Bai LH, Glick V, Hodges HC, Taylor SS, Maillard RA. Switching of
the folding-energy landscape governs the allosteric activation of protein kinase A. Proc Natl
Acad Sci. 2018;115:E7478-E7485.
38. Smith FD, Esseltine JL, Nygren PJ, Veesler D, Byrne DP, Vonderach M, Strashnov I,
Eyers CE, Eyers PA, Langeberg LK, Scott JD. Local protein kinase A action proceeds
through intact holoenzymes. Science. 2017;356:1288-1293.
39. Ogreid D, Ekanger R, Suva RH, Miller JP, Sturm P, Corbin JD, Doskeland SO. Activation
of Protein-Kinase Isozymes by Cyclic-Nucleotide Analogs Used Singly or in Combination -
Principles for Optimizing the Isozyme Specificity of Analog Combinations. Eur J Biochem.
1985;150:219-227.
40. Schwede F, Bertinetti D, Langerijs CN, Hadders MA, Wienk H, Ellenbroek JH, de Koning
EJP, Bos JL, Herberg FW, Genieser H-G, Janssen RAJ, Rehmann H. Structure-Guided
Design of Selective Epac1 and Epac2 Agonists. PLoS Biol. 2015;13:e1002038.
41. Wang PY, Liu ZQ, Chen HY, Ye N, Cheng XD, Zhou J. Exchange proteins directly
activated by cAMP (EPACs): Emerging therapeutic targets. Bioorg Med Chem Lett.
2017;27:1633-1639.
42. Sun DP, Fang CL, Chen HK, Wen KS, Hseu YC, Hung ST, Uen YH, Lin KY. EPAC1
overexpression is a prognostic marker and its inhibition shows promising therapeutic
potential for gastric cancer. Oncol Rep. 2017;37:1953-1960.
43. Ichikawa H, Itsumi M, Kajioka S, Maki T, Lee K, Tomita M, Yamaoka S. Overexpression
of exchange protein directly activated by cAMP-1 (EPAC1) attenuates bladder cancer cell
migration. Biochem Biophys Res Commun. 2018;495:64-70.
44. Pereira L, Rehmann H, Lao DH, Erickson JR, Bossuyt J, Chen J, Bers DM. Novel Epac
fluorescent ligand reveals distinct Epac1 vs. Epac2 distribution and function in
cardiomyocytes. Proc Natl Acad Sci. 2015;112:3991-3996.
22

45. Ponsioen B, Gloerich M, Ritsma L, Rehmann H, Bos JL, Jalink K. Direct Spatial Control
of Epac1 by Cyclic AMP. Mol Cell Biol. 2009;29:2521-2531.
46. Li S, Tsalkova T, White MA, Mei FC, Liu T, Wang D, Woods VL, Cheng XD. Mechanism
of intracellular cAMP sensor Epac2 activation: cAMP-induced conformational changes
identified by amide hydrogen/deuterium exchange mass spectrometry (DXMS). J Biol Chem.
2011;286:17889-17897.
47. Rehmann H, Das J, Knipscheer P, Wittinghofer A, Bos JL. Structure of the cyclic-AMP-
responsive exchange factor Epac2 in its auto-inhibited state. Nature. 2006;439:625-628.
48. Christensen AE, Selheim F, de Rooij J, Dremier S, Schwede F, Dao KK, Martinez A,
Maenhaut C, Bos JL, Genieser H-G, Doskeland SO. cAMP analog mapping of Epac1 and
cAMP kinase - Discriminating analogs demonstrate that Epac and cAMP kinase act
synergistically to promote PC-12 cell neurite extension. J Biol Chem. 2003;278:35394-35402.
49. Rehmann H, Schwede F, Doskeland SO, Wittinghofer A, Bos JL. Ligand-mediated
activation of the cAMP-responsive guanine nucleotide exchange factor Epac. J Biol Chem.
2003;278:38548-38556.
50. Dao KK, Teigen K, Kopperud R, Hodneland E, Schwede F, Christensen AE, Martinez A,
Dskeland SO. Epac1 and cAMP-dependent protein kinase holoenzyme have similar cAMP
affinity, but their cAMP domains have distinct structural features and cyclic nucleotide
recognition. J Biol Chem. 2006;281:21500-21511.
51. Gottlieb HE, Kotlyar V, Nudelman A. NMR chemical shifts of common laboratory solvents
as trace impurities. J Org Chem. 1997;62:7512-7515.
52. DiFrancesco D. Characterization of Single Pacemaker Channels in Cardiac Sinoatrial
Node Cells. Nature. 1986;324:470-473.
53. Bertinetti D, Schweinsberg S, Hanke SE, Schwede F, Bertinetti O, Drewianka S,
Genieser H-G, Herberg FW. Chemical tools selectively target components of the PKA
system. BMC Chem Biol. 2009;9:3.
54. Cook PF, Neville ME, Vrana KE, Hartl FT, Roskoski R. Adenosine Cyclic 3',5'-
Monophosphate Dependent Protein-Kinase - Kinetic Mechanism for the Bovine Skeletal-
Muscle Catalytic Subunit. Biochemistry. 1982;21:5794-5799.
55. Chepurny OG, Bertinetti D, Diskar M, Leech CA, Afshari P, Tsalkova T, Cheng XD,
Schwede F, Genieser H-G, Herberg FW, Holz GG. Stimulation of Proglucagon Gene
Expression by Human GPR119 in Enteroendocrine L-cell Line GLUTag. Mol Endocrinol.
2013;27:1267-1282.
56. Rehmann H, Prakash B, Wolf E, Rueppel A, de Rooij J, Bos JL, Wittinghofer A. Structure
and regulation of the cAMP-binding domains of Epac2. Nat Struct Biol. 2003;10:26-32.
57. Moll D, Prinz A, Gesellchen F, Drewianka S, Zimmermann B, Herberg FW. Biomolecular
interaction analysis in functional proteomics. J Neural Transm. 2006;113:1015-1032.
23

24

25