Di-aryl guanidinium derivatives: Towards improved α2-Adrenergic affinity and antagonist activity

Article abstract of DOI:10.1016/j.ejmech.2020.112947

Compounds with excellent receptor engagement displaying α₂-AR antagonist activity are useful not only for therapeutic purposes (e.g. antidepressants), but also to help in the crystallization of this particular GPCR. Therefore, based on our broad experience in the topic, we have prepared eighteen di-aryl (phenyl and/or pyridin-2-yl) mono- or di-substituted guanidines and 2-aminoimidazolines. The in vitro α₂-AR binding affinity experiments in human brain tissue showed the advantage of a 2-aminoimidazolinium cation, a di-arylmethylene core, a conformationally locked pyridin-2-yl-guanidine and a di-substituted guanidinium to achieve good α₂-AR engagement. After different in vitro [³⁵S]GTPγS binding experiments in human prefrontal cortex tissue, it was possible to identify that compounds 7a, 7b and 7c were α₂-AR partial agonist, whereas 8h was a potent α₂-AR antagonist. Docking and MD studies with a model of α_2A-AR and two crystal structures suggest that antagonism is achieved by compounds carrying a di-substituted guanidine which substituent occupy a pocket adjacent to TM5 without engaging S200^5.42 or S204^5.46, and a mono-substituted cationic group, which favorably interacts with E94^2.65.

Full text of DOI:10.1016/j.ejmech.2020.112947

European Journal of Medicinal Chemistry xxx (xxxx) xxx
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Di-aryl guanidinium derivatives: Towards improved a2-Adrenergic
affinity and antagonist activity
,
Michela McMullan a, Brendan Kelly b c, Helene B. Mihigo a, Aaron P. Keogh a,
,
Fernando Rodriguez a, Iria Brocos-Mosquera d e, Aintzane García-Bea d,
,
, *
Patricia Miranda-Azpiazu d, Luis F. Callado d e, Isabel Rozas a
a School of Chemistry, Trinity Biomedical Sciences Institute, Trinity College Dublin, 152-160, Pearse Street, Dublin 2, Ireland
b Department of Computer Science, Molecular and Cellular Physiology, Institute for Computational and Mathematical Engineering, Stanford University,
Stanford, CA, 94305, USA
c Department of Structural Biology, Institute for Computational and Mathematical Engineering, Stanford University, Stanford, CA, 94305, USA
d Department of Pharmacology, University of the Basque Country, UPV/EHU, Leioa, Bizkaia, Spain
e
ꢀ
ꢀ
Centro de Investigacion Biomedica en Red de Salud Mental (CIBERSAM), Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Compounds with excellent receptor engagement displaying a2-AR antagonist activity are useful not only
for therapeutic purposes (e.g. antidepressants), but also to help in the crystallization of this particular
GPCR. Therefore, based on our broad experience in the topic, we have prepared eighteen di-aryl (phenyl
and/or pyridin-2-yl) mono- or di-substituted guanidines and 2-aminoimidazolines. The in vitro a2-AR
binding affinity experiments in human brain tissue showed the advantage of a 2-aminoimidazolinium
cation, a di-arylmethylene core, a conformationally locked pyridin-2-yl-guanidine and a di-substituted
Received 3 July 2020
Received in revised form
18 September 2020
Accepted 13 October 2020
Available online xxx
guanidinium to achieve good a2-AR engagement. After different in vitro [35S]GTP
gS binding experi-
Keywords:
ments in human prefrontal cortex tissue, it was possible to identify that compounds 7a, 7b and 7c were
a2-AR partial agonist, whereas 8h was a potent a2-AR antagonist. Docking and MD studies with a model
of a2A-AR and two crystal structures suggest that antagonism is achieved by compounds carrying a di-
substituted guanidine which substituent occupy a pocket adjacent to TM5 without engaging S2005.42
Di-phenyl methane
Di-pyridin-2-yl methane
Benzylpyridin-2-yl
Guanidine
2-aminoimidazoline
Depression
a2-adrenoceptor ligands
a2-adrenoceptor antagonists
Human brain tissue in vitro experiments
Molecular modelling
or S2045.46, and a mono-substituted cationic group, which favorably interacts with E942.65
.
© 2020 Elsevier Masson SAS. All rights reserved.
1. Introduction
therapeutic approach to date [4], and it is now accepted that this
alteration in chemical concentration indirectly induces the gradual
strengthening of synapses (neuroplasticity) and synthesis of new
neurons (neurogenesis) over the course of treatment to elicit the
behavioural antidepressant response [5].
Noradrenergic transmission plays a central role in the mediation
of stress where alterations in its signalling as well as changes in
physical features of its plasticity are heavily implicated in stress-
related disorders [6]. Dysfunction of noradrenergic transmission
in these disorders has been hypothesised to be directly linked to the
a2-noradrenergic receptor (a2-AR), as post-mortem biochemical
assays as well as live patient models have illustrated an up-
regulation or increase in receptor density of the high-affinity
state of the a2-AR in depression [7,8]. This is further supported by
the effects of normal antidepressant treatments which have shown
Although in recent years there has been a huge advancement in
understanding neural regions and circuits implicated in stress re-
sponses of affective disorders such as depression [1e3], there is still
limited novel treatments to show for it. Pharmacological manipu-
lation of monoamine transmission remains the most successful
List of abbreviations: a2-ARs, a2-adrenoceptors; nBuLi, n-butyl lithium; Boc, tert-
butoxycarbonyl; Fmoc, Fluorenylmethyloxycarbonyl; PFC, prefrontal cortex; [3H]
RX821002, 2-methoxyidazoxan; Ki, inhibition constant; [35S]GTP
gS, radioligand
guanosine 50-O-[gamma-thio]triphosphate; GPCR, G-protein coupled receptor;
EC50, Half maximal effective concentration; Emax, maximum [35S]GTP
stimulation over basal; IMHB, intramolecular hydrogen bond.
* Corresponding author.
gS binding
E-mail address: rozasi@tcd.ie (I. Rozas).
https://doi.org/10.1016/j.ejmech.2020.112947
0223-5234/© 2020 Elsevier Masson SAS. All rights reserved.
Please cite this article as: M. McMullan, B. Kelly, H.B. Mihigo et al., Di-aryl guanidinium derivatives: Towards improved a2-Adrenergic affinity
and antagonist activity, European Journal of Medicinal Chemistry, https://doi.org/10.1016/j.ejmech.2020.112947

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
to desensitize or down-regulate the increase in a2-AR density [9].
Clinical a2-AR antagonists have shown to have faster behavioural
antidepressant responses than conventional antidepressants alone,
which are thought to be due to a combination of their effects on
noradrenaline transmission, as well as influencing key responses
involved in neuroplasticity and neurogenesis.
nM Ki values) antagonists of the a2-AR, we focused our attention
around the di-aryl core of 1. In a ligand-based drug design strategy,
we incorporated key elements from our previously developed
pharmacophore and functional modifications that favor a2-AR
antagonist activity such as the di-substitution of both guanidine
functionalities (compounds 6a-c, Fig. 2). Additionally, since we
previously established that the 2-aminoimidazoline derivatives
exhibit strong a2-AR binding, it was decided to combine these two
structural features to create asymmetric di-substituted di-aryl
guanidine/2-aminoimidazoline analogues (compounds 7a-f, Fig. 2).
Also, in order to probe if flexibility between the aryl rings is
important, compounds with a methylene linker as in ‘hit’ 1 (7a-c,
Fig. 2) and an ethylene linker (7d-f, Fig. 2) were investigated, since
good a2-AR antagonist activity has been obtained for previously
prepared ethylene linked di-phenyl molecules [11].
Our group has been interested in the development of aryl guani-
dines as antagonists of the a2-ARs for more than 15 years [10e15].
Thus, in
aminoimidazolinium compound 1 (Fig. 1) was identified as a ‘hit’
[11], because, even though it was an agonist, it showed excellent a2
a
previous study the di-phenylmethane bis-2-
-
AR binding affinity (Ki ¼ 1.58 nM). From these studies, shorter
monomeric analogues were prepared exhibiting reasonable affinities
(<300 nM Ki values), yet discrepancies between agonist and antago-
nist activity were seen amongst very structurally similar derivatives
(2 and 3, Fig. 1) [11e14].
Activity at the receptor is of paramount importance; therefore,
our group developed a 3D pharmacophore incorporating a wide
range of a2-AR antagonists, to underline the structural elements
needed for antagonistic activity [13]. This study indicated the re-
quirements of R2 substitution at the guanidinium moiety (4, Fig. 1)
resulting in ligands with exclusive antagonist activity. However, it
must be noted that this modification had a detrimental effect on the
a2-AR affinity. Later, we reported a series of pyridinyl guanidine and
2-aminoimidazoline derivatives (5, Fig. 1) as conformationally
controlled systems that showed markedly improved a2-AR affinity
and, again, consistent antagonist activity [14].
Now, we present the synthesis of a series of di-arylmethane
guanidines and 2-aminoimidazolines emanating from the prom-
ising pharmacological profiles of initial derivatives as well as the
pharmacological evaluation of their in vitro a2-AR binding affinity
and activity in human PFC brain tissue. The primary aim of this
work is to obtain compounds with binding affinities in the range of
compound 1 (excellent receptor engagement), while displaying
solely antagonist activity at the a2-AR (selective blockade).
Lastly, considering the encouraging results obtained in our pre-
vious pyridine mono-cation series (compounds 5, Fig. 1), this bio-
isosteric modificationwas also incorporated intothe core di-aromatic
framework, combining the guanidine and 2-aminoimidazoline moi-
eties on eithersideofdi-pyridin-2-yl or phenyl/pyridine-2-yl systems
(compounds 8a-i, Fig. 2).
2.1. Synthesis
Our usual synthetic pathway towards the guanidine-like de-
rivatives involves the mercury(II) chloride mediated guanidylation
of the corresponding di-aryl amines. In the case of compounds 6a-c,
the corresponding starting di-aniline is commercially available;
however, the proposed phenyl pyridinyl (8d-i) and di-pyridinyl
(8a-c) guanidine-like derivatives required the initial preparation
of the appropriate diamines (9 and 10, respectively) from available
2-aminopyridine starting materials (Scheme 1).
The synthesis of phenylpyridinyl diamine 9 involved the prepa-
ration of the di-benzyl protected 2-amino-5-bromopyridine (11),
which was then subjected to lithium halogen exchange using nBuLi,
and subsequently reacted with 4-nitrobenzaldehyde to give inter-
mediate 12. Dual reduction of the alcohol and nitro groups under
hydrogenation conditions afforded intermediate 13. It must be noted
that the benzyl groups of these 2-amino pyridine analogues were
found to be inert to cleavage under hydrogenation conditions up to
12 atm. De-benzylation was next achieved using concentrated
2. Results and discussion
To date no ligand of our in-house library has surpassed the a2
-
AR binding affinity of our initial ‘hit’ compound 1 (Ki ¼ 1.58 nM)
and considering that our objective is to prepare high affinity (low
Fig. 1. Structures of some a2-AR ligands previously described in our group highlighting structural features necessary for antagonist activity.
2

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Fig. 2. Structures of the di-arylmethane bis-guanidine (compounds 6a-c) and guanidine/2-aminoimidazoline (compounds 7a-f and 8a-i) families synthesised and investigated as
potential a2-AR ligands.
Scheme 1. Synthesis of dI-aryl bis-amino precursors 9 and 10.
sulfuric acid affording diamine 9. Synthesis of di-pyridinyl diamine 10
involved a modification of literature procedures [16]. The adapted
pathway involved the di-benzylation of 2-amino pyridine affording
14, followed by electrophilic aromatic substitution with formalde-
hyde (37% solution inwater) employing acetic acid as an electrophilic
assistant. De-benzylation was subsequently achieved by stirring 15 in
concentrated sulfuric acid for 24 h affording 10 in good yield.
Next, reaction of these starting di-aryl amines with the appro-
priate mono/di-Boc protected thioureas (16a,b’,c,d, Schemes 2 and
3; preparation described in ESI) yielded the corresponding Boc-
protected bis-guanidylated (17a,b’,c, Scheme 2) or mono-
guanidylated (18a-l, Scheme 3) products, depending on the
equivalents of amine used.
The mono-guanidylated di-aryl amines (18a-l) were then sub-
jected to a second coupling reaction with N,N0-di(tert-butox-
ycarbonyl)imidazolidine-2-thione in the presence of mercury (II)
chloride (Scheme 3) to give the final Boc-protected imino-2-
imidazolidine intermediates (19a-l). The Boc-protected de-
rivatives were subsequently deprotected using either a 4 M solution
of HCl/dioxane or 1.25 M HCl/MeOH to yield the final guanidinium
hydrochloride salts (6a-c, 7a-f, and 8a-f) in excellent yield
(Schemes 2 and 3).
Preparation of the pyridin-2-ylguanidine di-aryl derivatives (8g-
i) is shown in Scheme 4. Ideally, the bis-Boc protected 2-
iminoimidazolidine system would be first introduced to the
phenyl ring of compound 9 prior to the guanidine, as the aniline
3

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Scheme 2. Synthesis of di-substituted bis-guanidinium di-phenyl derivatives 6a,b,c.
Scheme 3. Synthesis of guanidinium/2-aminoimidazolinium di-aryl derivatives 7a-f and 8a-f.
amino group is far more nucleophilic than the 2-aminopyridine
amine. Unfortunately, in this situation the 2-iminoimidazolidine
moiety is not stable under the guanidylation conditions and leads
to degradation products under mercury(II) chloride and other re-
agents that promote desulphurisation.
Hence, for analogues 8g-i, di-aryl amine 9 was initially Fmoc
protected affording amine 20, which was then subjected to stan-
dard guanidylation conditions with thioureas 16a,c-d to yield in-
termediates 21a-c (Scheme 4). After Fmoc deprotection using
excess diethylamine, derivatives 22a-c were guanidylated with
4

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Scheme 4. Synthesis of guanidinium/2-aminoimidazolinium di-aryl derivatives 8g-i
N,N0-di(tert-butoxycarbonyl)imidazolidine-2-thione affording in-
termediates 23a-c, which were subsequently Boc-deprotected to
the guanidinium salts (8g-i) using hydrochloric acid dioxane so-
lutions (Scheme 4).
Evaluation of the binding affinities of all di-cationic compounds
indicates that, in general, the presence of a 2-aminoimidazoline
functional group as one of the cationic components is required to
achieve good a2-AR engagement (Ki < 500 nM), as significantly
lower binding affinities were obtained for bis-guanidines 6a-c.
Moreover, the ethylene linked derivatives 7d-f exhibited poorer
affinities compared to the methylene linked analogues 7a-c; from
previous work we have found that alternate linkers such as NH, O,
CO or S also result in weaker binding [10], which highlights the
advantage of the CH2 linker for optimal a2-AR affinities. In partic-
ular, the di-phenyl N,N0-bis/mono-substituted guanidine and 2-
aminoimidazolidine derivatives 7a-c gave excellent a2-AR binding
affinities with the Ki values of 7b-c on par with that of ‘hit’ com-
pound 1. This illustrates that steric extension at the guanidinium
moiety can result in higher affinities than the mono-substituted
guanidines within the di-cationic series.
2.2. Pharmacology
2.2.1. Binding affinity for the a2-AR
The affinity toward a2-AR, in general, was measured in vitro
using human brain PFC tissue by competition assays against the
selective radioligand [3H]RX821002 (2-methoxyidazoxan), which
has a known high affinity for the a2-AR (Kd ¼ 0.89 nM) [17]. The PFC
region contains a2-ARs of different subtypes; however, the a2A
-
subtype is predominant (90e95%) with a reduced presence of the
a2C- subtype (5e10%). Even though there are cell-based assays that
allow to assess a2-AR subtype selectivity, some previous works (e.g.
reference [18]) have shown that selectivity and efficacy demon-
strated by different compounds in cell models do not always
correspond to those in native tissue itself. Thus, in our study we
In order to analyse the binding results obtained for the family of
compounds 8 (Table 1), we should consider first that the potential
formation of an IMHB implies a competition for the corresponding
cation (guanidinium/2-aminoimidazolinium) to interact inter- or
intra-molecularly and that could decrease the target engagement.
On the contrary, free rotation of the cation indicates that only inter-
have focused in the affinity of our compounds for the a2-ARs, in
general, more than in their selectivity for a specific subtype. In the
mentioned competition assay, the prepared membranes containing
the receptor proteins were incubated with the radioligand, at a
constant concentration of 2 nM. The results obtained (Table 1) are
expressed in terms of nM Ki values.
molecular interactions are possible and, hence,
a stronger
engagement can be achieved. Thus, there are three possible sce-
narios for the structures of compounds 8: (i) compounds 8a-c
5

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Table 1
Ki (nM) values for the a2-AR shown by all final salts prepared and ‘hit’ compound 1.
Compound
Structure
Ki (nM) SEM
1.58 0.12
1
6a
355 43
6b
6c
2485 372
742 59
7a
7b
12.30 1.11
1.77 0.21
7c
0.82 0.09
175 14
7d
7e
7f
418 50
1745 262
8a
8b
138 11
263 29
8c
89.9 13.5
8d
8e
674 81
433 113
6

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Table 1 (continued)
Compound
Structure
Ki (nM) SEM
120 14
8f
8g
8h
86.8 6.9
33.9 3.7
8i
67.6 8.1
where both the guanidinium and the 2-aminoimidazolinium sys-
tems can establish IMHBs with the corresponding adjacent pyri-
dine ring, (ii) compounds 8d-f where only the 2-
aminoimidazolinium system can establish IMHB with the adja-
cent pyridine while the guanidinium system is free to rotate and
interact with any residue in the binding site, and (iii) compounds
8g-i where only the guanidinium cation can establish IMHB with
the adjacent pyridine while the 2-aminoimidazolinium system is
free to rotate and interact with any residue in the binding site.
Looking at the Ki values obtained for this family of compounds, it
seems that the formation of an imidazolinium-pyridine IMHB plus
a guanidinium free to bind any residue in the binding site (com-
pounds 8d-f) is the most detrimental combination for target
engagement (Ki values between 120 and 674 nM). On the contrary,
the second best in terms of target engagement is the 8a-c set (both
guanidinium and imidazolinium forming IMHBs) with Ki values
between 89.9 and 138 nM. Finally, the best set in terms of binding is
that which compounds can form a guanidinium-pyridine IMHB
(8g-i) with Ki values between 33.9 and 86.8 nM. Therefore, it seems
that the most relevant feature for target engagement, in this
particular family of compounds, is the guanidinium-pyridine IMHB
formation. However, if we compare the results of the 8g-i set with
those of the 7a-c set (Ki between 0.8 and 12.3 nM) where both the
guanidinium and 2-aminoimidazolinium have free rotation and
there is not competition between inter- and intra-molecular in-
teractions, it could be concluded that, in general, the free rotation of
the 2-aminoimidazolinium system plays an important role in target
engagement.
antagonist functional activity. Interestingly, compounds 7a-c and
8h displayed agonist activity through the stimulation of GTP, but
only at high concentrations. This did not correspond with the
potent binding observed for these compounds, because a dose
corresponding to the binding affinity would be expected to exert
the corresponding activity.
Next, in order to confirm their a2-AR functional activity, all
compounds were subjected to another [35S]GTP
gS assay against the
known a2-AR agonist UK14304. Table 2 illustrates the antagonist
effect induced by known a2-AR antagonist RX821002 and com-
pounds 7a-c, 8c, 8f-i at a single concentration of 10ꢀ5 M on a2-AR
agonist UK14304’s stimulation of [35S]GTP
gS binding. All of these
compounds significantly induced an increase in the EC50 of
UK14304 by shifting the concentration-response curve to the right,
which indicates competitive antagonism of the a2-AR.
However, since derivatives 7a-c and 8h showed agonist prop-
erties in [35S]GTP
g
S binding experiments, but antagonized the a2
AR agonist UK14304 effect by increasing its EC50 (Table 2), a third
35S]GTP
S binding assay was performed in the presence or
absence of the selective a2-AR antagonist RX821002, to observe if
the ligand stimulation of [35S]GTP
S was occurring through a2-ARs.
As shown in Table 3, the stimulation of [35S]GTP
S binding induced
-
[
g
g
g
by compounds 7a, 7b and 7c was blocked by RX821002, which
suggests that these compounds are partial agonists of the a2-AR as
they stimulate [35S]GTP
gS binding by themselves but also antago-
nized the effect of a high-potency selective a2-AR agonist as
UK14304. Conversely, co-administration of RX821002 to compound
8h resulted in no change in its stimulation of [35S]GTP
g
S, clearly
indicating that compound 8h is acting as a potent a2-AR antagonist,
and that its stimulation of [35S]GTP
S is perhaps the result of acting
g
2.2.2. [35S]GTP
Compounds which displayed Ki affinity values < 125 nM (com-
pounds 7a-c, 8c and 8f-i) were subjected to in vitro [35S]GTP
gS binding functional assays
as an agonist at an alternative GPCR.
gS
binding experiments in human PFC tissue to determine their nature
as agonists or antagonists. As the a2-AR is a member of the GPCR
family, the direct evaluation of the degree of G-protein activation
upon ligand binding can be made by determining guanine nucleo-
Table 2
EC50 values (mM) obtained from the concentration-response curves for UK14304
(10ꢀ12 to 10ꢀ3 M, 10 concentrations) stimulation of [35S]GTP
absence or presence of the different compounds (10ꢀ5 M).
g
S binding in the
tide exchange using radiolabelled GTP
agonist, antagonist or inverse agonist activity. Activation of the a2-AR
leads to the dissociation of GDP-G
, followed by GDP-[35S]GTP
exchange, which can be used to measure the functional activity.
Here, the phosphodiester bond that links the -phosphate to the rest
g gS) to observe
S ([35S]GTP
Experiment
EC50 (mM) SEM
Emax (%) SEM
UK14304
1.77 0.16
87.2 7.5
20.2 0.4
11.6 3.0
30.0 1.5
10.2 0.3
4.14 0.33
39.9 3.2
52.3 1.9
30.8 2.8
153
98
4
a
gS
UK14304 þ RX821002
UK14304 þ 7a
UK14304þ 7b
UK14304þ 7c
UK14304 þ 8c
UK14304 þ 8f
UK14304 þ 8g
UK14304 þ 8h
UK14304 þ 8i
5
5
110
155
171
127
139
140
96
2
8
7
3
6
7
g
of the nucleotide cannot be hydrolysed to reform GDP, and hence
prevents the GTP binding protein from being inactivated, allowing
for facile scintillation counting of the radiolabelled analogue.
In this experiment compounds 8c, 8f, 8g and 8i did not stimu-
122
3
late
[ gS binding, which in principle is indicative of
35S]GTP
7

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Table 3
residues on TM5 are known to be important for endogenous ligand-
binding as well as many discovered adrenoceptor ligands, partic-
ularly S2005.42 and S2045.46 which likely form HBs with
noradrenaline [24]. Furthermore, in Pose B, water molecules fill this
particular pocket, which would be detrimental to the affinity of
molecules were this to be the correct pose. Thus, Pose A is predicted
to be the orientation in which compounds bind to the receptor
which fills the pocket at TM5. It also follows that molecules which
do not engage with key residues on TM5 S2005.42 or S2045.46 are
likely to act as antagonists which do not activate the receptor.
These simulations were used to give some insight into the re-
sults observed for our lead compounds 7a-c and 8h. Experimen-
tally, we have observed that the affinity of di-substituted
guanidiniums is higher than that of the mono-substituted ana-
logues and, as the simulation results suggest, this indicates that
they interact with the endogenous binding site adjacent to TM5.
The partial agonism of compounds 7a, 7b and 7c can be rationalized
by their proximity to the S200 and S204 in TM5. As for their pyri-
dine analogues, 8b-c and 8h-i, which show no partial agonism, an
IMHB between the pyridine N and guanidinium NH is maintained
throughout simulation and prevents the alkyl/furan-2-yl substitu-
ent from extending as close to TM5, seemingly conferring antago-
nist activity to these compounds (Fig. S3, ESI).
By December 2019, four crystal structures of the different a2-AR
subtypes have been subsequently deposited in the PDB: the crystal
structure of the human a2A-AR in complex with a partial agonist
(PDB: 6KUY [25]); that of the a2A-AR receptor of the Spodoptera
frugiperda in complex with an antagonist (PDB: 6KUX [26]), that of
the human a2C-AR with an antagonist (PDB: 6KUW [27]) and also
the structure of the Spodoptera frugiperda a2B-AR in complex with
GoA (PDB: 6K41 [28]). Accordingly, we have carried out computa-
tional studies to validate the results we obtained with the a2A-AR
homology model structure. Thus, we have docked compound 8h, as
a model of the compounds here studied, in the 6KUY, 6KUX and
6KUW crystal structures, which correspond to the two most
important adrenoceptor subtypes in the brain’s PFC (a2A- and a2C-).
When docking 8h into the a2A-AR structures (6KUX and 6KUY;
Figs. S4 and S5, respectively) we observed the same interactions
(mostly with E94 and D113) that we had seen in the docking of 8h
into the a2A-AR homology model developed by us. In the case of
docking 8h into the a2C-AR structure (6KUW, Fig. S6) some different
contacts are observed, but the interactions established are still
favorable for binding. Finally, we have superposed the ‘best pose’
obtained for 8h within our a2A-AR homology model and the ‘best
poses’ obtained with 6KUX and 6KUY and we observed a very good
alignment (RMSD ¼ 1.66 and 2.09, respectively; see Fig. 4).
EC50 values (m gS
M) obtained from the concentration-response curves for [35S]GTP
binding stimulation by compounds 7a, 7b and 7c (10ꢀ12 to 10ꢀ3 M, 10 concentra-
tions) alone or in the presence of RX821002 (10ꢀ5 M).
Experiment
EC50
18
(
m
M) SEM
Emax (%) SEM
126 3a
7a
4
7a þ RX821002
64020 44
122
4
7b
534 72
168 4a
7b þ RX821002
2356 107
130
4
7c
46
487 18
4
139 18a
110 11
7c þ RX821002
8h
194 21
252 27
149 7a
8h þ RX821002
154
9
a
Emax represents the maximum [35S]GTP
gS binding stimulation induced by each
compound over basal stimulation in the absence of the compound (100%).
2.3. Docking and molecular dynamics simulations
At the time of performing the present research, no crystal
structure of any of the a2-AR subtypes had been determined and
thus, a homology model of the a2A-AR (major subtype present in
the PFC) was constructed in the Prime program of the Schrodinger
suite [19], using a crystal structure of the b2-AR as template
(5JQH.pdb) [20], in order to carry out a docking study to understand
the di-cationic structural activity observed and structure-affinity
relationships of families 6, 7 and 8. Multiple conformers of com-
pounds 6a-c, 7a-f, and 8a-i were then docked into this model using
Glide [21]. Docking resulted in two dominant poses for every
compound, each with comparable scores. These poses are charac-
terized by two electrostatic interactions: (i) that between one of the
ligand’s cationic groups (guanidinium or 2-aminoimidazolinium)
and D1133.32, and (ii) another between the second ligand’s cationic
group and E942.65 (superscript numbers denote the Ballesteros-
Weinstein numbering scheme for Class A GPCRs [22]). The second
salt-bridge formed by these compounds with the receptor could be
driving their enhanced affinity relative to that of mono-cationic
derivatives previously prepared by us [10e15]. Further, position
2.65 is a glutamate in all
route to selectively targeting
position 2.65 is a glutamate in all dopamine receptors, but no in 5-
HT receptors [23].
a
-AR, but not in
b-AR, providing a potential
a
-AR. It should also be noted that
Molecular dynamics (MD) simulations were next carried out to
determine the stability of these poses and the likelihood for prev-
alence of one over the other that would facilitate structure-based
drug design (see examples in Fig. 3, Figs. S1 and S2, ESI). In gen-
eral, compounds with low a2-AR affinity (e.g. 6a, Ki ¼ 355 nM, in
Figs. S1 and S2, ESI), despite remaining stably bound to D113 in the
orthosteric binding pocket throughout simulation, show strong
fluctuations due to instability caused by solvent exposure, which
disrupts the ionic interaction with E94. However, high a2-AR af-
finity compounds (e.g. 7a, 7b and 7c, Ki ¼ 12.30, 1.77 and 0.82 nM,
respectively; in Figs. S1 and S2, ESI) remain in a stable pose
throughout simulation. Analysis of interactions with key Serine
residues on TM5 S2005.42 and S2045.46 could also shed light on
which molecules would have agonist or antagonist functional ac-
tivity. While both poses are relatively stable throughout simulation,
differences emerge. The pose in which the di-substituted arylgua-
nidinium forms a hydrogen bond (HB) with D1133.32 (Pose A),
places the substituent (alkyl, furan-2-yl) to fill the pocket formed
adjacent to TM5, while the other cationic group extends towards
the extracellular solvent to interact with E942.65 (Fig. 3a, showing
compound 8h, Ki ¼ 33.9 nM).
All the molecular modelling studies provide a possible rationale
for the pharmacological results observed for high affinity partial
agonists/antagonists, 7a-c and 8h. Compounds with di-substituted
guanidines where one of these substituents have been optimized to
take advantage of the pocket adjacent to TM5 without engaging
S2005.42 or S2045.46 are desirable, while di-substituted 2-
aminoimidazolines may also be considered. Moreover, the pres-
ence of an additional mono-substituted cationic group is favorable
to interact with E942.65, as there is little room for second sub-
stituents and they tend to extend into solvent.
3. Conclusions
Aiming to obtain compounds with excellent receptor engage-
ment, while displaying selective blockade at the a2-AR, we have
prepared a series of di-aryl (di-phenyl, phenyl/pyridin2-yl or di-
pyridin-2-yl) mono- or di-substituted guanidines and 2-
aminoimidazolines emanating from previous encouraging results.
Following different synthetic strategies three disubstituted bis-
When the cationic groups are in the opposite orientation (Pose
B), the pocket at TM5 is left vacant. This is likely unfavorable as
8

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
Fig. 3. Representative simulation snapshots of (a) Pose A demonstrated for compound 8h. The di-substituted guanidinium interacts with D113 and extends to fill the pocket formed
at TM5, the 2-aminoimidazolinium interacts with E94, (b) Pose B demonstrated for compound 8h. The 2-aminoimidazolinium interacts with D113, leaving the TM5 pocket vacant,
the di-substituted guanidinium interacts with E94, placing the furan-2-yl substituent towards solvent. (c) In pose A both salt bridges are stable and the binding pocket at TM5 is
filled by the ligand, whereas (d) in pose B the salt bridge to E94 is less stable due to solvent exposure of the furan-2-yl substituent and the binding pocket at TM5 is left vacant.
Fig. 4. Alignment of the docked ‘best-pose’ of compound 8h in the in-house a2A-AR homology model and that in the 6KUX (a2A-AR, RMSD ¼ 1.66) crystal structure (left) or that in
the 6KUY (a2A-AR, RMSD ¼ 2.09) crystal structure (right).
guanidines with a di-phenylmethane core were prepared as well as
fifteen hybrid guanidine/2-aminoimidazoline derivatives with di-
phenylmethane (three), di-phenylethane (three), di-prydin-2-
ylmethane (three), and benzylpyridin-2-yl (six) cores.
Additionally, the in vitro a2-AR binding affinity of all the com-
pounds prepared was studied in human PFC brain tissue. Despite
the a2A-AR being the predominant subtype in the human PFC, a
reduced presence of the a2C-AR is also acknowledged. Thus, we
cannot exclude a minimal incidence of this minority subtype in the
results shown. Thus, we found that the presence of a phenyl-2-
aminoimidazolinium cation, a CH2 linker between the aryl sys-
tems and the steric extension at the aryl-guanidinium contribute to
9

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
achieve good a2-AR engagement (Ki < 68 nM). Furthermore, even
though formation of an IMHB increases a2-AR affinity in pyridin-2-
yl-guanidine compounds, in general, the free rotation of a 2-
aminoimidazolinium system attached to the diaryl core plays an
even more important role in target engagement.
terminated by separating free ligand from bound ligand by rapid
filtration under vacuum (1450 Filter Mate Harvester, PerkinElmer)
through GF/C glass fiber filters. The filters were then rinsed three
times with 300
mL binding buffer, air-dried (120 min), and counted
for radioactivity by liquid scintillation spectrometry using
a
Compounds 7a-c, 8c and 8f-i, which displayed a2-AR Ki affinity
values < 125 nM, were assessed as agonists or antagonists in vitro
MicroBeta TriLux counter (PerkinElmer). Specific binding was
determined and plotted as a function of the compound concen-
tration. Nonspecific binding was determined in the presence of
idazoxan (10ꢀ3 M). Analysis of competition experiments to obtain
the inhibition constant (Ki) were performed by non-linear regres-
sion using the Graph Pad Prism 5 program. All experiments were
analysed assuming a one-site model of radioligand binding.
via [35S]GTP
g
S functional binding experiments in PFC tissue. Since
the results were not conclusive, all compounds were subjected to
an alternate [35S]GTP
S assay against UK14304 (known selective
g
a2-AR agonist) to elucidate their functional activity. Considering
that derivatives 7a-c and 8h showed agonist properties by them-
selves in [35S]GTP
g
S binding experiments, but increased the EC50 of
UK14304, a third [35S]GTP
S binding assay was performed in the
presence of RX821001 (selective a2-AR antagonist). This assay
indicated that the stimulation of [35S]GTP
S binding induced by 7a,
g
4.1.3. [35S]GTP
The incubation buffer for measuring [35S]GTP
membranes contained, in a total volume of 250
3 mM MgCl2, 100 mM NaCl, 50 mM GDP, 50 mM Tris-HCl at pH 7.4,
and 0.5 nM [35S]GTP
S. Protein aliquots were thawed and resus-
pended in the same buffer. The incubation was started by addition
gS binding assays
g
S binding to brain
L, 1 mM EGTA,
g
m
7b and 7c was blocked by RX821001, suggesting a2-AR partial
agonism, whereas lack of change in RX821001 stimulation of [35S]
g
GTP
antagonism.
Finally, computational studies (docking and MD in an in-house
gS when co-administered with 8h indicates potent a2-AR
of the membrane suspension (20 mg of membrane proteins per well)
to the previous mixture and was performed at 30 ꢁC for 120 min,
a2A-AR homology model) have thrown light to the pharmacolog-
ical results observed for high affinity partial agonists/antagonists,
7a-c and 8h. These studies suggest that antagonism is achieved by
diarylmethane derivatives with a di-substituted guanidine which
substituent occupy the pocket adjacent to TM5 without engaging
S2005.42 or S2045.46, and a mono-substituted cationic group, which
favorably interact with E942.65, as there is little room for second
substituents and they tend to extend into solvent. Comparative
docking studies with our own a2A-AR homology model and the
recently reported crystal structures of a2A-AR (complexed with a
partial agonist or an antagonist) and a2C-AR, showed similar results.
Summarizing, we not only have found a new antagonist with
excellent a2-AR receptor engagement, but also we have determined
the structural features that will result in optimal target engage-
ment. All this will allow us designing novel therapeutics to treat
depression (some of our previous compounds have already shown
to reach the brain of rats in in vivo microdialysis experiments and
exhibited anti-stress activities in animal tests [11,29]).
with shaking. In order to evaluate the influence of the compounds
on [35S]GTP S binding, ten concentrations (10ꢀ12 to 10ꢀ3 M) of the
g
different compounds were added to the assay. Incubations were
terminated by separating free ligand from bound ligand by rapid
filtration under vacuum (1450 Filter Mate Harvester, PerkinElmer)
through GF/C glass fiber filters. The filters were then rinsed three
times with 300
mL of ice-cold incubation buffer and air-dried
(90 min). The radioactivity trapped was determined by liquid
scintillation spectrometry (MicroBeta TriLux counter, PerkinElmer).
The [35S]GTP S bound was about 6e15% of the total [35S]GTP
g gS
added. Nonspecific binding of the radioligand was defined as the
remaining [35S]GTP
S binding in the presence of 10 M unlabelled
GTP S binding stimu-
S. Emax represents the maximum [35S]GTP
g
m
g
g
lation over basal.
4.2. Computational
4.2.1. Homology modelling and docking
A homology model of the a2-AR was constructed as described in
previous studies [15]. Briefly, we used the Modeller 9.12 software
[30] to construct homology models of the human a2A-AR using
sequences from www.uniprot.org (a2AAR, P08913), based on three
template structures of the homologous b2-adrenoceptor (PDB IDs:
b2-AR; 2R4S, 3D4S, and 3SN6), which has 38.9% homology with the
a2A-AR and 35.2% homology with the a2C-AR. Molecules were
prepared in the Maestro software and docked using Glide. Poses for
simulation were chosen based on those that incorporated ionic
interaction with D1133.32, which is known to be important to GPCR-
binding ligands, and a second ionic interaction between the ligand
4. Experimental
4.1. Pharmacology
4.1.1. Preparation of membranes
Neural membranes (P2 fractions) were prepared from the PFC of
human brains obtained at autopsy in the Basque Institute of Legal
Medicine, Bilbao, Spain. Post-mortem human brain samples of each
subject (~1 g) were homogenized using a Teflon-glass grinder (10
up-and-down strokes at 1500 rpm) in 10 vol of homogenization
buffer (5 mM Tris-HCl and 0.25 M sucrose, pH 7.4). The crude ho-
mogenate was centrifuged for 5 min at 1000 g (4 ꢁC), and the su-
pernatant was centrifuged again for 10 min at 40,000 g (4 ꢁC). The
resultant pellet was washed twice in 5 vol of resuspension buffer
(50 mM Tris-HCl, pH 7.5) and re-centrifuged in similar conditions.
Aliquots of 2 mg protein were stored at ꢀ70 ꢁC until assay. Protein
concentration was measured according to the Bradford method,
using bovine serum albumin as standard.
and E942.65
.
Additionally, the structures of the a2A-AR complexed with a
partial agonist (6KUY) or with an antagonist (6KUX) as well as that
of the a2C-AR complexed with an antagonist were retrieved from
the RCSB PDB to be used for docking studies.
4.2.2. System setup for molecular dynamics (MD) simulations
MD simulations of the a2-AR were based on complexes formed
from docking compounds 6a-6c, 7a-7f, 8a-8i to the constructed
homology model. Two poses (one with each cationic group in
contact with D1133.32/E942.65) were simulated for these 18 com-
pounds, resulting in a total of 36 simulations. Each simulation was
4.1.2. [3H]RX821002 binding assays
Specific [3H]RX821002 binding was measured in 0.25 mL ali-
quots (50 mM Tris-HCl, pH 7.5) of the human brain membranes,
which were incubated in 96-well plates with [3H]RX821002 (2 nM)
for 30 min at 25 ꢁC in the absence or presence of the competing
compounds (10ꢀ12 to 10ꢀ3 M, 10 concentrations). Incubations were
~250 ns in length, for a total of 9
atoms were added using Prime (Schrodinger Inc.), and protein
chain termini were capped with the neutral groups acetyl and
m
s. For each complex, hydrogen
€
10

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
methylamide. Titratable residues were left in their dominant pro-
tonation state at pH 7.0. All aspartate residues were deprotonated,
as is expected in the inactive state of GPCRs to which antagonists
bind. Both guanidinium/2-amino-imidazolinium groups were
positively charged. The prepared protein structures were aligned
on the transmembrane helices to the Orientation of Proteins in
Membranes (OPM) [31] structure of PDB 5JQH, and internal waters
added with Dowser [32]. The structures were then inserted into a
pre-equilibrated palmitoyl-oleoyl-phosphatidylcholine (POPC)
bilayer and solvated with 0.15 M NaCl in explicitly represented
water, then neutralized by removing sodium ions.
We used the CHARMM36 parameter set for protein molecules,
lipid molecules, and salt ions, and the CHARMM TIP3P model for
water; protein parameters incorporated CMAP terms [33]. Param-
eters for ligands were generated using the CHARMM General Force
Field (CGenFF [34]) with the ParamChem server (paramchem.org),
version 1.0.0. Hydrogen mass repartitioning was employed to
enable a timestep of 4.0 fs [35].
instrument with a WATERS 2690 autosampler and methanol/
acetonitrile as carrier solvent. Melting points were determined
using a Stuart Scientific Melting Point SMP1 apparatus and are
uncorrected. Infrared spectra were recorded on a PerkinElmer
Spectrum One FT-IR Spectrometer equipped with a Universal ATR
sampling accessory. HPLC purity analysis was carried out using a
Varian ProStar system equipped with a Varian Prostar 335 diode
array detector and a manual injector (20 mL). For purity assessment,
UV detection was performed at 245 nm and peak purity was
confirmed using a purity channel. The stationary phase consisted of
an ACE 5C18-AR column (150 ꢂ 4.6 mm), and the mobile phase
used the following gradient system, eluting at 1 mL/min: aqueous
formate buffer (30 mM, pH 3.0) for 10 min, linear ramp to 85%
methanol buffered with the same system over 25 min, hold at 85%
buffered methanol for 10 min. Minimum requirement for purity
was set at 95%.
4.3.1. General methods
Simulations were performed on GPUs using the CUDA version of
PMEMD (Particle Mesh Ewald Molecular Dynamics) in Amber15
[36]. Prepared systems were minimized, then equilibrated as fol-
lows: the system was heated using the Langevin thermostat from
0 to 100 K in the NVT ensemble over 12.5 ps with harmonic re-
straints of 10.0 kcal molꢀ1 Åꢀ2 on the non-hydrogen atoms of lipid,
protein and ligand, and initial velocities sampled from the Boltz-
mann distribution. The system was then heated to 310 K over 125
ps in the NPTensemble with semi-isotropic pressure coupling and a
pressure of 1 bar. Further equilibration was performed at 310 K with
harmonic restraints on the protein and ligand starting at
5.0 kcal molꢀ1 Åꢀ2 and reduced by 1.0 kcal molꢀ1 Åꢀ2 in a stepwise
fashion every 2 ns, for a total of 10 ns of additional restrained
equilibration. Simulations were conducted in the NPT ensemble at
310 K and 1 bar, using a Langevin thermostat and Monte Carlo
barostat. In each of these simulations, we performed 5 ns of un-
restrained equilibration followed by a production run of 250 ns.
Simulations used periodic boundary conditions, and a time step of
4.0 fs. Bond lengths to hydrogen atoms were constrained using
SHAKE. Non-bonded interactions were cut off at 9.0 Å, and long-
range electrostatic interactions were computed using the particle
4.3.1.1. Method A: generation of guanidine hydrochlorides from N-
(tert-butoxycarbonyl) protected guanidine and iminoimidazolidine
derivatives using hydrochloric acid in 1,4-dioxane. To the starting
Boc-protected guanidine was added 4 M HCl/dioxane (6.0 eq. per
Boc group) and a 1:1 solution of iPrOH:CH2Cl2, such as to maintain a
reaction concentration of 0.2 M. The mixture was stirred at 55 ꢁC for
3e4 h, after which all starting material was consumed as adjudged
by TLC analysis. Solvent and excess HCl were then removed under
vacuum and the crude salt was dissolved in a minimum volume of
H2O. It was washed with CH2Cl2 (2 ꢂ 5 mL) and then purified using
reverse phase chromatography (C-8 silica) using 100% H2O as mo-
bile phase.
4.3.2. Characterization of final salts
4.3.2.1. 4,40-Di[3-(furanyl-2-methyl)guanidino]diphenylmethane
dihydrochloride (6a). Following Method A, to a solution of 4 M HCl/
1,4-dioxane (12.0 eq., 600 mL, 2.28 mmol) was added compound 17a
(1.0 eq., 120 mg, 0.19 mmol) affording the title compound (89 mg,
91%) as yellow solid. Mp: 135e139 ꢁC. dH (400 MHz, D2O): 3.90 (s,
2H, CH2), 4.34 (s, 4H, CH2), 6.21 (app. t, 2H, J ¼ 4.0, Ar), 6.32 (app. t,
2H, J ¼ 4.0, Ar), 7.08 (d, 4H, J ¼ 8.2, Ar), 7.23 (d, 4H, J ¼ 8.2, Ar), 7.38
(app. t, 2H, J ¼ 4.0, Ar). dC (125 MHz, D2O): 38.1 (CH2), 48.7 (CH2),
107.1 (CH Ar), 110.3 (CH Ar), 126.3 (CH Ar), 130.1 (CH Ar), 132.1 (qC
Ar), 141.1 (qC Ar), 143.3 (CH Ar), 149.1 (qC Ar), 155.3 (qC CN). nmax
(ATR)/cmꢀ1: 3134 (NH), 2929, 1626, 1609, 1437, 1343, 1073, 1017,
975. HRMS: (m/z ESIþ): Found: 443.2190 (Mþ þ H, C25H27N6O2
Requires: 443.2193). Purity by HPLC: 98.4% (tR 23.63 min).
mesh Ewald (PME) method with an Ewald coefficient
b of
approximately 0.31 Å and B-spline interpolation of order 4. The FFT
grid size was chosen such that the width of a grid cell was
approximately 1 Å.
Trajectory snapshots were saved every 100 ps during production
simulations. Trajectory analysis was performed using VMD [37] and
CPPTRAJ [38], and visualization was performed using VMD. Two
metrics were used to determine ligand stability in their respective
poses: the distance between each cationic group of the ligand and
either D1133.32 or E942.65. Figures were rendered using PyMol [39].
4.3.2.2. 4,40-Di[3-(hydroxyethyl)guanidino]diphenylmethane dihy-
drochloride (6b). Following Method A, to a solution of 1.25 M HCl/
CH3OH (15.0 eq., 4.56 mL, 5.7 mmol) was added compound 17b (1.0
eq., 250 mg, 0.38 mmol) affording the title compound as a yellow
gum (148 mg, 88%). dH (400 MHz, D2O): 3.25 (t, 4H, J ¼ 8.0, CH2),
3.57 (t, 4H, J ¼ 8.0, CH2), 3.83 (s, 2H, CH2), 7.05 (d, 4H, J ¼ 8.0, Ar),
7.17 (d, 4H, J ¼ 8.0, Ar). dC (125 MHz, D2O): 40.1 (CH2), 42.6 (CH2),
59.8 (CH2), 125.9 (CH Ar), 130.1 (CH Ar), 132.2 (qC Ar), 140.9 (qC Ar),
155.7 (qC CN). nmax (ATR)/cmꢀ1: 3319 (NH), 3161 (NH), 2963, 1600,
1622, 1514, 1351, 1248, 1061, 1019, 920. HRMS: (m/z ESIþ): Found:
371.2190 (Mþ þ H, C19H27N6O2 Requires: 371.2199). Purity by HPLC:
95.5% (tR 19.95 min).
4.3. Synthesis
All commercial chemicals were obtained from Sigma-Aldrich or
Fluka and used without further purification. Deuterated solvents
for NMR use were purchased from Apollo. Dry solvents were pre-
pared using standard procedures, according to Vogel [40], with
distillation prior to use. Solvents for synthesis purposes were used
at GPR grade. Analytical TLC was performed using Merck Kieselgel
60 F254 silica gel plates or Polygram Alox N/UV254 aluminium
oxide plates. Visualization was by UV light (254 nm). NMR spectra
were recorded on Bruker DPXe400 Avance spectrometers, oper-
ating at 400.13 MHz and 600.1 MHz for 1H NMR; 100.6 MHz and
150.9 MHz for 13C NMR. Shifts are referenced to the internal solvent
signals. NMR data were processed using Bruker TOPSPIN software.
HRMS spectra were measured on a Micromass LCT electrospray TOF
4.3.2.3. 4,40-Di[3-(propy)guanidino]diphenylmethane
dihydro-
chloride (6c). Following Method A, to a solution of 4 M HCl/1,4-
dioxane (12.0 eq., 1.55 mL, 6.12 mmol) was added compound 17c
(1.0 eq., 290 mg, 0.51 mmol) affording the title compound as a
yellow solid (199 mg, 89%). Mp: 144e150 ꢁC. dH (400 MHz, D2O):
0.79 (t, 6H, J ¼ 8.0, CH3), 1.46 (m, 4H, CH2), 3.07 (t, 4H, J ¼ 8.0, CH2),
11

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
3.83 (s, 2H, CH2), 7.08 (d, 4H, J ¼ 8.1, Ar), 7.21 (d, 4H, J ¼ 8.1, Ar). dC
(125 MHz, D2O): 10.2 (CH3), 21.2 (CH2), 39.8 (CH2), 43.1 (CH2), 126.1
(CH Ar), 130.1 (CH Ar), 132.3 (qC Ar), 140.9 (qC Ar), 155.7 (qC CN).
nmax (ATR)/cmꢀ1: 3129 (NH), 2934, 1625, 1599, 1511, 1296, 1246,
1147, 1048, 805. HRMS: (m/z ESIþ): Found: 367.2605 (Mþ þ H,
4.3.2.8. {4-(2-Aminoimidazolino)-4’-[3-(furanyl-2-methyl)guani-
dino]}-1,2-diphenylethane dihydrochloride (7e). Following Method
A, to a solution of 4 M HCl/1,4-dioxane (18.0 eq., 540 mL, 2.16 mmol)
was added compound 19e (1.0 eq., 80 mg, 0.12 mmol) affording the
title compound as a yellow solid (57 mg, 82%). Mp: 237e240 ꢁC
decomposed. dH (400 MHz, D2O): 2.86 (br s, 4H, CH2), 3.59 (br s, 4H,
CH2), 4.34 (s, 2H, CH2), 6.28 (m, 1H, Ar), 6.32 (m, 1H, Ar), 7.04 (m,
4H, Ar), 7.39 (m, 4H, Ar), 7.39 (br s, 1H, Ar). dC (100 MHz, D2O): 35.8
(CH2), 38.0 (CH2), 42.6 (CH2), 108.1 (CH Ar), 110.9 (CH Ar), 124.3 (CH
Ar), 125.9 (CH Ar), 129.8 (CH Ar), 130.0 (CH Ar), 131.7 (qC Ar), 132.2
(qC Ar), 141.0 (qC Ar), 141.5 (qC Ar), 143.2 (CH Ar), 149.1 (qC Ar),
155.4 (qC CN), 158.7 (qC CN). nmax (ATR)/cmꢀ1: 3143 (NH), 2892,
1648, 1602, 1612, 1461, 1381, 1345, 1240, 1148, 1084, 1016, 825.
HRMS: (m/z ESIþ): Found: 403.2240 (Mþ þ H, C23H27N6O Requires:
403.2240). Purity by HPLC: 96.6% (tR 24.16 min).
C
21H31N6 Requires: 367.2605). Purity by HPLC: 98.3% (tR 24.60 min).
4.3.2.4. 4-Guanidino-4’-(2-aminoimidazoline)diphenylmethane
dihydrochloride (7a). Following Method A, to a solution of 4 M HCl/
1,4-dioxane (24.0 eq., 850 mL, 3.36 mmol) was added compound 19a
(1.0 eq., 100 mg, 0.14 mmol) affording the title compound as a
yellow solid (52 mg, 98%). Mp: 219e221 ꢁC decomposed. dH
(400 MHz, D2O): 3.59 (s, 4H, CH2), 3.88 (s, 2H, CH2), 7.08 (m, 4H, H-
2’, Ar), 7.22 (m, 4H, Ar). dC (100 MHz, D2O): 40.4 (CH2), 42.6 (CH2),
124.6 (CH Ar), 126.3 (CH Ar), 130.0 (CH Ar), 130.1 (CH Ar), 132.1 (qC
Ar), 133.5 (qC Ar), 140.6 (qC Ar), 144.6 (qC Ar), 156.3 (qC CN), 158.7
(qC CN). nmax (ATR)/cmꢀ1: 3384 (NH), 3270 (NH), 2993, 1602, 1518,
1406, 1232, 1147, 801. HRMS: (m/z ESIþ): Found: 309.1826 (Mþ þ H,
4.3.2.9. {4-[3-(Propyl)guanidino]-4’-(2-aminoimidazolino)}-1,2-
diphenylethane dihydrochloride (7f). Following Method A, to a so-
C
17H21N6 Requires: 309.1822). Purity by HPLC: 97.6% (tR 19.91 min).
lution of 4 M HCl/1,4-dioxane (18.0 eq., 890
mL, 3.6 mmol) was
added compound 19f (1.0 eq.,130 mg, 0.20 mmol) affording the title
compound as a yellow solid (80 mg, 91%). Mp: 271e275 ꢁC
decomposed. dH (400 MHz, D2O): 0.79 (t, 3H, J ¼ 7.4, CH3), 1.46 (m,
2H, CH2), 2.82 (br s, 4H, CH2), 3.07 (t, 2H, J ¼ 7.4, CH2), 3.59 (br s, 4H,
CH2), 7.02 (d, 4H, J ¼ 8.3, Ar), 7.11 (d, 4H, J ¼ 8.3, Ar). dC (100 MHz,
D2O): 10.2 (CH3), 21.4 (CH2), 35.8 (2 CH2), 42.6 (CH2), 43.0 (CH2),
124.1 (CH Ar), 125.8 (CH Ar), 129.8 (CH Ar), 129.9 (CH Ar), 132.0 (qC
Ar), 132.7 (qC Ar), 140.9 (qC Ar), 141.3 (qC Ar), 155.1 (qC CN), 158.6
(qC CN). nmax (ATR)/cmꢀ1: 3018 (NH), 2849, 1593, 1508, 1355, 1234,
1144, 1081, 1017, 826. HRMS: (m/z ESIþ): Found: 365.2305 (Mþ þ H,
4.3.2.5. 4-(2-Aminoimidazolino)-4’-[3-(furanyl-2-methyl)guanidino)
diphenylmethane dihydrochloride (7b). Following Method A, to a
solution of 4 M HCl/1,4-dioxane (18.0 eq., 1.30 mL, 5.4 mmol) was
added compound 19b (1.0 eq., 200 mg, 0.30 mmol) affording the
title compound as a white solid (111 mg, 80%). Mp: 201e206 ꢁC
decomposed. dH (400 MHz, D2O): 3.75 (s, 4H, CH2) 4.03 (s, 2H, CH2),
4.50 (s, 2H, CH2), 6.44 (d, 1H, J ¼ 4.0, Ar), 6.47 (t, 1H, J ¼ 4.0, Ar), 7.22
(d, 4H, J ¼ 8.3, Ar), 7.36 (d, 4H, J ¼ 8.3, Ar), 7.34 (br s, 1H, Ar). dC
(100 MHz, D2O): 38.0 (CH2), 40.0 (CH2), 42.6 (CH2), 108.7 (CH Ar),
110.7 (CH Ar), 124.8 (CH Ar), 126.4 (CH Ar), 130.2 (CH Ar), 130.5 (CH
Ar), 132.3 (qC Ar), 133.3 (qC Ar), 140.8 (qC Ar), 141.4 (qC Ar), 143.3
C
21H29N6 Requires: 365.2296). Purity by HPLC: 95.1% (tR 23.01 min).
(CH Ar), 149.2 (qC Ar), 155.5 (qC CN), 158.6 (q CN). nmax (ATR)/cmꢀ1
:
3140 (NH), 2981, 1647, 1570, 1387, 1244, 1096, 1022, 975. HRMS: (m/
z ESIþ): Found: 389.2084 (Mþ þ H, C22H25N6O Requires: 389.2094).
Purity by HPLC: 95.1% (tR 23.44 min).
4.3.2.10. 6-(2-Aminoimidazolino)-6’-(guanidino)dipyridinyl-3-
methane dihydrochloride (8a). Following Method A, to a solution of
4 M HCl/1,4-dioxane (24.0 eq., 840
mL, 3.36 mmol) was added
compound 19g (1.0 eq., 100 mg, 0.14 mmol) affording the title
compound as a yellow solid (49 mg, 92%). Mp: 149e151 ꢁC
decomposed. dH (400 MHz, D2O): 3.64 (s, 2H, CH2), 3.67 (s, 4H, CH2),
6.74 (d, 2H, J ¼ 8.1, H-30,H-3), 7.45 (dd, 2H, 3J ¼ 8.1, 4J ¼ 2.2, H-40, H-
4.3.2.6. 4-[3-(Propyl)guanidino]-4’-(2-aminoimidazolino)diphenyl-
methane dihydrochloride (7c). Following Method A, to a solution of
4 M HCl/1,4-dioxane (18.0 eq., 1.40 mL, 5.58 mmol) was added
compound 19c (1.0 eq., 200 mg, 0.31 mmol) affording the title
compound as a brown solid (126 mg, 96%). Mp: 221e225 ꢁC
decomposed. dH (400 MHz, D2O): 0.78 (t, 3H, J ¼ 8.3, CH3), 1.45 (m,
2H, CH2), 3.07 (t, 2H, J ¼ 8.3, CH2), 3.59 (s, 4H, CH2), 3.87 (s, 2H, CH2),
7.06 (d, 4H, J ¼ 8.1, Ar), 7.2 (d, 4H, J ¼ 8.1,Ar). dC (100 MHz, D2O): 10.3
(CH3), 21.4 (CH2), 40.0 (CH2), 42.6 (CH2), 43.0 (CH2), 124.6 (CH Ar),
126.2 (CH Ar), 130.0 (CH Ar), 130.1 (CH Ar), 132.3 (qC Ar), 133.0 (qC
Ar), 140.7 (qC Ar), 141.0 (qC Ar), 155.2 (qC CN), 158.7 (qC CN). nmax
(ATR)/cmꢀ1: 3159 (NH), 2968, 1717, 1650, 1592, 1485, 1468, 1284,
1240, 1088, 1018, 820, 758. HRMS: (m/z ESIþ): Found: 351.2292
(Mþ þ H, C20H27N6 Requires: 351.2292). Purity by HPLC: 99.7% (tR
22.49 min).
4
4), 7.99 (d, 2H, J ¼ 2.1, H-60, H-6). dC (100 MHz, D2O): 33.4 (CH2),
42.0 (CH2-Imid),112.6 (CH, Ar-30),113.3 (CH, Ar-3),131.9 (q, Ar-5, Ar-
50), 139.7 (CH, Ar-4, Ar-40), 146.0 (CH, Ar-60), 146.5 (CH, Ar-6), 148.8
(q, Ar-20 or Ar-2), 149.8 (q, Ar-2 or Ar-2’), 154.6 (q, C]N), 156.3 (q,
C]N). vmax (ATR)/cmꢀ1: 3185 (NH), 2980, 1624, 1600, 1481, 1355,
1242, 1027, 965. HRMS: (m/z ESIþ): Found: 311.1728 (Mþ þ H,
C15H19N8 Requires: 311.1733). Purity by HPLC: 98.4% (tR 1.75 min).
4.3.2.11. 6-(2-Aminoimidazolino)-6’-[2-(furanyl-2-methyl)guani-
dino]dipirydinyl-3-methane
Method A, to a solution of 4 M HCl/1,4-dioxane (18.0 eq., 400
dihydrochloride
(8b). Following
L,
m
1.62 mmol) was added compound 19h (1.0 eq., 64 mg, 0.09 mmol)
affording the title compound as a yellow solid (34 mg, 82%). Mp:
164e167 ꢁC decomposed. dH (400 MHz, D2O): 3.76 (s, 2H, CH2), 3.79
(s, 4H, CH2), 4.54 (s, 2H, CH2), 6.44 (br s, 2H, Ar), 6.91 (m, 2H, Ar),
7.51 (br s, 1H, Ar), 7.51 (dd, 1H, 3J ¼ 8.1, 4J ¼ 2.4, Ar), 7.61 (dd, 1H, 3J ¼
8.1, 4J ¼ 2.4, Ar), 7.96 (s, 1H, Ar), 8.11 (s, 1H, Ar). dC (100 MHz, D2O):
33.7 (CH2), 37.8 (CH2), 42.2 (CH2), 108.5 (CH Ar), 110.6 (CH Ar), 112.5
(CH Ar), 113.2 (CH Ar), 128.7 (qC Ar), 130.6 (qC Ar), 139.6 (CH Ar),
139.8 (CH Ar), 143.3 (CH Ar), 145.5 (CH Ar), 146.5 (CH Ar), 148.5 (qC
Ar), 149.5 (qC Ar), 149.5 (qC Ar), 153.7 (qC CN), 156.0 (qC CN). vmax
(ATR)/cmꢀ1: 3127 (NH), 2918, 2894, 1629, 1601, 1574, 1388, 1281,
1234, 1045, 1025, 828, 741. HRMS: (m/z ESIþ): Found: 391.1993
(Mþ þ H, C20H23N8O Requires: 391.1995). Purity by HPLC: 98.5% (tR
24.03 min).
4.3.2.7. [4-Guanidino-4’-(2-aminoimidazolino)]-1,2-diphenylethane
dihydrochloride (7d). Following Method A, to a solution of 4 M HCl/
1,4-dioxane (24.0 eq., 1.24 mL, 5.04 mmol) was added compound
19d (1.0 eq., 150 mg, 0.21 mmol) affording the title compound as a
yellow solid (71 mg, 86%). Mp: 210e215 ꢁC decomposed. dH
(400 MHz, D2O): 2.83 (s, 4H, CH2), 3.58 (s, 4H, CH2), 7.03 (m, 4H, Ar),
7.18 (m, 4H, Ar). dC (125 MHz, D2O): 35.8 (CH2), 35.8 (CH2), 42.6
(CH2), 124.1 (CH Ar), 125.7 (CH Ar), 129.6 (CH Ar), 129.9 (CH Ar),
131.7 (qC Ar), 132.7 (qC Ar), 140.5 (qC Ar), 141.5 (qC Ar), 156.2 (qC
CN), 158.6 (qC CN). nmax (ATR)/cmꢀ1: 3135 (NH), 2986, 1648, 1608,
1578, 1512, 149, 1255, 1019, 827. HRMS: (m/z ESIþ): Found: 323.1977
(Mþ þ H, C18H23N6 Requires: 323.1987). Purity by HPLC: 97.9% (tR
21.29 min).
12

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
4.3.2.12. 6-(2-Aminoimidazolino)-6’-[2-(propyl)guanidino]dipir-
ydinyl-3-methane dihydrochloride (8c). Following Method A, to a
4.3.2.16. 2-Guanidino-5-[(2-aminoimidazolino)benzyl]pyridine
dihydrochloride (8g). Following Method A, to a solution of 4 M HCl/
solution of 4 M HCl/1,4-dioxane (18.0 eq., 378
mL, 1.52 mmol) was
1,4-dioxane (24.0 eq., 480 mL,1.92 mmol) was added compound 23a
added compound 19i (1.0 eq., 55 mg, 0.084 mmol) affording the
title compound as a yellow solid (30 mg, 85%). Mp: 148e151 ꢁC
decomposed. dH (400 MHz, D2O): 0.81 (t, 3H, J ¼ 7.4, CH3), 1.50 (m,
2H, CH2), 3.11 (t, 2H, J ¼ 7.4, CH2), 3.62 (s, 2H, CH2), 3.66 (s, 4H, CH2),
6.75 (d, 2H, J ¼ 8.4, Ar), 7.45 (m, 2H, Ar), 7.94 (d, 1H, 4J ¼ 2.2, Ar),
8.00 (d, 1H, 4J ¼ 2.2, Ar). dC (100 MHz, D2O): 10.3 (CH3), 21.1 (CH2),
30.1 (CH2), 33.7 (CH2), 42.7 (CH2), 111.0 (CH Ar), 112.6 (CH Ar), 130.2
(q Ar), 132.0 (qC Ar), 139.7 (CH Ar), 139.8 (CH Ar), 145.7 (CH Ar),
148.8 (CH Ar), 148.8 (qC Ar), 151.6 (qC Ar), 156.0 (qC CN), 156.5 (qC
CN). vmax (ATR)/cmꢀ1: 3280 (NH), 2918, 1629, 1601, 1483, 1348,
1281, 1243, 1044, 1027, 1025, 926, 829, 741. HRMS: (m/z ESIþ):
Found: 353.2206 (Mþ þ H, C18H25N8 Requires: 353.2202). Purity by
HPLC: 98.5% (tR 23.93 min).
(1.0 eq., 53 mg, 0.08 mmol) affording the title compound as an off-
white solid (25 mg, 82%). Mp: 155e159 ꢁC decomposed. dH
(400 MHz, D2O): 3.58 (s, 4H, CH2), 3.78 (s, 2H, CH2), 6.78 (d, 1H, J ¼
8.4, Ar), 7.07 (d, 2H, J ¼ 8.3, Ar), 7.18 (d, 2H, J ¼ 8.3, Ar), 7.50 (dd, 1H,
3J ¼ 8.4, 4J ¼ 2.2, Ar), 8.03 (d, 1H, 4J ¼ 2.2, Ar). dC (100 MHz, D2O):
36.8 (CH2), 42.6 (CH2), 113.5 (CH Ar), 124.5 (CH Ar), 130.0 (CH Ar),
132.9 (qC Ar), 133.4 (qC Ar), 139.5 (qC Ar), 140.0 (CH Ar), 146.0 (CH
Ar), 149.2 (qC Ar), 158.0 (qC CN). vmax (ATR)/cmꢀ1: 3338 (NH), 2934,
1652, 1606, 1492, 1391, 1240,1068,1025, 845, 757. HRMS (m/z ESIþ):
Found: 310.1783 (Mþ þ H, C16H20N7 Requires: 310.1775). Purity by
HPLC: 98.8% (tR 21.47 min).
4. 3. 2.17. 2-[2-(Furanyl-2-methyl)guanidino]-5-[4-(2-
aminoimidazolino)benzyl]pyridine
Following Method A, to a solution of 4 M HCl/1,4-dioxane (18.0 eq.,
315 L, 1.26 mmol) was added compound 23b (1.0 eq., 47 mg,
dihydrochloride
(8h).
4.3.2.13. 2-(2-Aminoimidazolino)-5-(4-guanidino)benzyl]pyridine
dihydrochloride (8d). Following Method A, to a solution of 4 M HCl/
m
1,4-dioxane (24.0 eq., 660
m
L, 2.64 mmol) was added compound 19j
0.07 mmol) affording the title the compound as an off-white solid
(29 mg, 90%). Mp: 134e137 ꢁC, decomposed. dH (400 MHz, D2O): 3.75
(s, 4H, CH2), 3.91 (s, 2H, CH2), 4.55 (s, 2H, CH2), 6.47 (br s, 2H, Ar), 6.90
(d,1H, J¼ 8.5,Ar),7.20(d, 2H, J¼ 8.1, Ar),7.29(d,2H, J¼8.1, Ar), 7.53(br
s,1H, Ar), 7.61 (dd,1H, 3J ¼ 8.4, 4J ¼ 1.6, Ar), 8.07 (d,1H, 4J ¼ 1.6, Ar). dC
(100 MHz, D2O): 36.8(CH2), 37.8(CH2), 42.6(CH2),108.4(CH Ar),110.6
(CH Ar),113.2 (CH Ar),124.3 (CH Ar),129.9 (CH Ar),132.7 (qC Ar),133.3
(qC Ar),139.6 (qC Ar),139.9 (CH Ar),143.3 (CH Ar),145.7 (CH Ar),148.7
(1.0 eq., 80 mg, 0.11 mmol) affording the title compound as a white
solid (41 mg, 98%). Mp: 167e170 ꢁC decomposed. dH (400 MHz,
D2O): 3.67 (s, 4H, CH2), 3.84 (s, 2H, CH2), 6.79 (d,1H, J ¼ 8.4, Ar), 7.10
(d, 2H, J ¼ 8.3, Ar), 7.19 (d, 2H, J ¼ 8.3, Ar), 7.52 (dd, 1H, 3J ¼ 8.4, 4J ¼
2.2, Ar), 8.07 (d, 1H, 4J ¼ 2.2, Ar). dC (100 MHz, D2O): 36.4 (CH2), 42.2
(CH2),112.7 (CH Ar),126.1 (CH Ar),130.1 (CH Ar),132.3 (qC Ar),132.8
(qC Ar), 139.8 (CH Ar), 140.4 (qC Ar), 146.6 (CH Ar), 148.7 (qC Ar),
155.7 (qC CN), 156.2 (qC CN). nmax (ATR)/cmꢀ1: 3136 (NH), 2923,
2853, 1575, 1629, 1603, 1485, 1391, 1242, 1144, 1079, 1023, 926.
HRMS: (m/z ESIþ): Found: 310.1781 (Mþ þ H, C16H20N7 Requires:
310.1780). Purity by HPLC: 96.1% (tR 21.05 min).
(qC Ar), 149.6 (qC Ar), 153.9 (qC CN), 158.4 (qC CN). vmax (ATR)/cmꢀ1
3337 (NH), 2945, 2834, 1655, 1449, 1019, 1113, 1045, 956, 729. HRMS
(m/z ESIþ): Found: 390.2051 (Mþ
H, C21H24N7O Requires:
:
þ
390.2037). Purity by HPLC: 98.6% (tR 24.91 min).
4.3.2.14. 2-[2-Aminoimidazolino]-5-{4-[3-(furanyl-2-methyl)guani-
4.3.2.18. 2-[2-(Propyl)guanidino)]-5-[4-(2-aminoimidazolino)
dino]benzyl}pyridine dihydrochloride (8e). Following Method A, to
benzyl]pyridine dihydrochloride (8i). Following Method A, to a so-
a solution of 4 M HCl/1,4-dioxane (18.0 eq., 320
mL, 1.26 mmol) was
lution of 4 M HCl/1,4-dioxane (18.0 eq., 315 mL, 1.26 mmol) was
added compound 19k (1.0 eq., 44 mg, 0.07 mmol) affording the title
compound as a yellow solid (26 mg, 80%). Mp: 122e130 ꢁC
decomposed. dH (400 MHz, D2O): 3.64 (s, 4H, CH2), 3.79 (s, 2H, CH2),
4.30 (s, 2H, CH2), 6.24 (d, 1H, J ¼ 2.8, Ar), 6.28 (m, 1H, Ar), 6.77 (d,
1H, J ¼ 8.4, Ar), 7.06 (d, 2H, J ¼ 9.0, Ar), 7.17 (d, 2H, J ¼ 9.0, Ar), 7.34
(br s, 1H, Ar), 7.50 (dd, 1H, 3J ¼ 8.4, 4J ¼ 2.3, Ar), 8.05 (d, 1H, 4J ¼ 2.3,
Ar). dC (100 MHz, D2O): 36.7 (CH2), 42.6 (CH2), 47.6 (CH2), 108.1 (CH
Ar), 110.5 (CH Ar), 112.6 (CH Ar), 126.1 (CH Ar), 130.1 (CH Ar), 132.2
(qC Ar), 132.8 (qC Ar), 139.8 (CH Ar), 140.3 (qC Ar), 143.1 (CH Ar),
146.5 (CH Ar), 148.1 (qC Ar), 149.0 (qC Ar), 155.0 (qC CN), 156.3 (qC
CN). nmax (ATR)/cmꢀ1: 3199 (NH), 2923, 29854, 1636, 1599, 1511,
1242, 1147, 1076, 1016, 925, 742, 678. HRMS: (m/z ESIþ): Found:
390.2038 (Mþ þ H, C21H24N7O Requires: 390.2042). Purity by HPLC:
96.4% (tR 23.95 min).
added compound 23c (1.0 eq., 47 mg, 0.07 mmol) affording the title
compound as a yellow solid (27 mg, 98%). Mp: 164e168 ꢁC,
decomposed. dH (400 MHz, D2O): 0.78 (t, 3H, J ¼ 7.4, CH3), 1.47 (m,
2H, CH2), 3.08 (m, 2H, CH2), 3.67 (m, 4H, CH2), 3.87 (s, 2H, CH2), 6.80
(d, 1H, J ¼ 8.4, Ar), 7.08 (d, 2H, J ¼ 8.4, Ar), 7.18 (d, 2H, J ¼ 8.4, Ar),
7.53 (dd, 1H, 3J ¼ 8.4, 4J ¼ 2.2, Ar), 8.03 (d, 1H, 4J ¼ 2.2, Ar). dC
(100 MHz, D2O): 9.9 (CH3), 21.3 (CH2), 36.8 (CH2), 42.6 (CH2), 42.7
(CH2), 112.9 (CH Ar), 124.3 (CH Ar), 127.5 (qC Ar), 129.8 (CH Ar),
133.4 (qC Ar), 139.5 (CH Ar), 139.7 (qC Ar), 145.5 (CH Ar), 149.5 (qC
Ar), 153.3 (qC CN), 158.6 (qC CN). vmax (ATR)/cmꢀ1: 3350 (NH),
2955, 2921,1650,1606, 1509,1492,1391,1278,1034, 837, 736. HRMS
(m/z ESIþ): Found: 352.2243 (Mþ
þ
H, C19H26N7 Requires:
352.2244). Purity by HPLC: 98.8% (tR 23.23 min).
Declaration of competing interest
4.3.2.15. 2-(2-Aminoimidazolino)-5-[(4-(2-propylguanidino)benzyl]
pyridine dihydrochloride (8f). Following Method A, to a solution of
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
4 M HCl/1,4-dioxane (18.0 eq., 360
mL, 1.44 mmol) was added
compound 19l (1.0 eq., 53 mg, 0.08 mmol) affording the title
compound as a light yellow solid (30 mg, 87%). Mp: 175e179 ꢁC
decomposed. dH (400 MHz, D2O): 0.78 (t, 3H, J ¼ 7.4, CH3), 1.47 (m,
2H, CH2), 3.08 (m, 2H, CH2), 3.67 (br s, 4H, CH2), 3.87 (s, 2H, CH2),
6.80 (d, 1H, J ¼ 8.4, Ar), 7.08 (d, 2H, J ¼ 8.4, Ar), 7.18 (d, 2H, J ¼ 8.4,
Ar), 7.53 (dd, 1H, 3J ¼ 8.4, 4J ¼ 2.2, Ar), 8.03 (d, 1H, 4J ¼ 2.2, Ar). dC
(100 MHz, D2O): 9.9 (CH3), 21.2 (CH2), 36.8 (CH2), 42.0 (CH2), 43.0
(CH2), 112.0 (CH Ar), 126.2 (qC Ar),130.1 (CH Ar), 132.5 (qC Ar),133.1
(qC Ar), 139.1 (CH Ar), 146.6 (CH Ar), 148.7 (qC Ar), 155.7 (qC CN).
nmax (ATR)/cmꢀ1: 3324 (NH), 2963, 2833, 1655, 1572, 1395, 1202,
1164, 1097, 1020, 774. HRMS: (m/z ESIþ): Found: 352.2254 (Mþ þ H,
Acknowledgments
Thanks are given to the School of Chemistry at Trinity College
Dublin (M.McM.) and to the Irish Research Council (A.K. -GOIPG/
2014/457- and H.B.M. -GOIPG/2017/834-) for postgraduate support.
This study was also supported by the Ministerio de Economía y
Competitividad of Spain (SAF2013-48586-R) and the Basque Gov-
ernment (IT1211-19). The authors would like to thank the staff
members of the Basque Institute of Legal Medicine for their coop-
eration in the study.
C
19H26N7 Requires: 352.2250). Purity by HPLC: 97.0% (tR 22.64 min).
13

M. McMullan, B. Kelly, H.B. Mihigo et al.
European Journal of Medicinal Chemistry xxx (xxxx) xxx
receptors, British J. Pharmacol. 169 (2013) 887e899.
[19] Schrodinger Release 2016-1: Prime, Schrodinger, LLC, New York, NY, 2016.
[20] D.P. Staus, R.T. Strachan, A. Manglik, B. Pani, A.W. Kahsai, T.H. Kim,
L.M. Wingler, S. Ahn, A. Chatterjee, A. Masoudi, A.C. Kruse, E. Pardon,
J. Steyaert, W.I. Weis, R.S. Prosser, B.K. Kobilka, T. Costa, R.J. Lefkowitz, Nature
535 (2016) 448e452.
[21] R.A. Friesner, R.B. Murphy, M.P. Repasky, L.L. Frye, J.R. Greenwood,
T.A. Halgren, P.C. Sanschagrin, D.T. Mainz, Extra precision Glide: docking and
scoring incorporating a model of hydrophobic enclosure for protein-ligand
complexes, J. Med. Chem. 49 (2006) 6177e6196.
[22] J.A. Ballesteros, H. Weinstein, Integrated methods for the construction of
three-dimensional models and computational probing of structure-function
Appendix A. Supplementary data
€
€
Supporting Information is available for the calculated total en-
ergies, synthetic procedures and spectroscopic data for all in-
termediates and NMR spectra and HPLC chromatograms of all the
final hydrochloride salts.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2020.112947.
References
relations in
366e428.
ꢀ
G protein-coupled receptors, Methods Neurosci. 25 (1995)
[23] G. Pandy-Szekeres, C. Munk, T.M. Tsonkov, S. Mordalski, K. Harpsøe,
A.S. Hauser, A.J. Bojarski, D.E. Gloriam, GPCRdb in 2018: adding GPCR struc-
ture models and ligands, Nucleic Acids Res. 46 (2017) D440eD446.
[1] E.J. Nestler, The biological basis of depression: insights from animal models,
in: J.J. Strain, M. Blumenfield (Eds.), Depression as a Systemic Illness, Oxford
University Press, New York, 2018, pp. 1e13.
[2] F. Cathomas, J.W. Murrough, E.J. Nestler, M.-H. Han, S.J. Russo, Neurobiology of
resilience: interface between mind and body, Biol. Psychiatr. 86 (2019)
410e420.
[3] C.H. Stanton, A.J. Holmes, S.W.C. Chang, J. Joormann, From stress to anhedonia:
molecular processes through functional circuits, Trends Neurosci. 42 (2019)
23e42.
[24] M. Masureel, Y. Zou, L.P. Picard, E. van der Westhuizen, J.P. Mahoney,
J.P.G.L.M. Rodrigues, T.J. Mildorf, R.O. Dror, D.E. Shaw, M. Bouvier, E. Pardon,
J. Steyaert, R.K. Sunahara, W.I. Weis, C. Zhang, B.K. Kobilka, Structural insights
into binding specificity, efficacy and bias of a
Biol. 14 (2018) 1059e1066.
b2AR partial agonist, Nat. Chem.
[25] L. Qu, Q.T. Zhou, Y. Xu, Y. Guo, X.Y. Chen, D. Yao, G.W. Han, Z.-J. Liu,
R.C. Stevens, G.S. Zhong, D. Wu, S.W. Zhao, Structural basis of the diversity of
adrenergic receptors, Cell Rep. 29 (2019), 2929-2935 and 2936-2943.
[26] Qu, L.; Zhou, Q. T.; Wu, D.; Zhao, S.W. Crystal structures of the alpha2A
adrenergic receptor in complex with an antagonist RSC. Released in PDB the
4th of December 2019. To Be Published. DOI: 10.2210/pdb6KUX/pdb.
[27] Chen, X. Y.; Wu, D.; Wu, L. J.; Han, G. W.; Guo, Y.; Zhong, G. S. Crystal structure
of human alpha2C adrenergic G protein-coupled receptor. Released in PDB the
4th of December 2019. To be published. DOI: 10.2210/pdb6KUW/pdb.
[28] D. Yuan, Z. Liu, J. Kaindl, S. Maeda, J. Zhao, X. Sun, J. Xu, P. Gmeiner,
H.W. Wang, B.K. Kobilka, Activation of the alpha2B-adrenoceptor by the
sedative sympatholytic dexmedetomidine, Nat. Chem. Biol. 16 (2020)
507e512.
ꢁ
[4] G. Di Giovanni, D. Svob Strac, M. Sole, M. Unzeta, K.F. Tipton, D. Mück-Seler,
I. Bolea, L. Della Corte, M. Nikolac Perkovic, N. Pivac, I.J. Smolders, A. Stasiak,
W.A. Fogel, P. De Deurwaerdere, Monoaminergic and histaminergic strategies
ꢂ
and treatments in brain diseases, Front. Neurosci. 10 (2016) 541.
[5] S.R. Wainwright, L.A. Galea, The neural plasticity theory of depression:
assessing the roles of adult neurogenesis and PSA-NCAM within the hippo-
campus, Neural Plast. 2013 (2013), 805497.
[6] S.A. Flavin, D.G. Winder, Noradrenergic control of the bed nucleus of the stria
terminalis in stress and reward, Neuropharmacology 70 (2013) 324e330.
[7] C. Cottingham, Q. Wang, a2 Adrenergic receptor dysregulation in depressive
disorders: implications for the neurobiology of depression and antidepressant
therapy, Neurosci. Biobehav. Rev. 36 (2012) 2214e2225.
[8] V. Maletic, A. Eramo, K. Gwin, S.J. Offord, R.A. Duffy, The role of norepineph-
[29] C. Muguruza, F. Rodriguez, I. Rozas, J.J. Meana, L. Uriguen, L.F. Callado, Anti-
depressant-like properties of three new alpha2-adrenoceptor antagonists,
Neuropharmacology 65 (2013) 13e19.
rine and its a-adrenergic receptors in the pathophysiology and treatment of
major depressive disorder and Schizophrenia: a systematic review, Front.
Psychiatr. 8 (2017) 42.
ꢁ
[30] A. Sali, T.L. Blundell, Comparative protein modelling by satisfaction of spatial
restraints, J. Mol. Biol. 234 (1993) 779e815.
ꢀ
[9] J. García-Sevilla, P. Ventayol, V. Perez, G. Rubovszky, D. Puigdemont, M. Ferrer-
[31] M.A. Lomize, I.D. Pogozheva, H. Joo, H.I. Mosberg, A.L. Lomize, OPM database
and PPM web server: resources for positioning of proteins in membranes,
Nucleic Acids Res. 40 (2012) D370eD376.
ꢀ
ꢀ
Alcon, A. Andreoli, J. Guimon, E. Alvarez, Regulation of platelet a2A-adreno-
ceptors, Gi proteins and receptor kinases in major depression: effects of
Mirtazapine treatment, Neuropsychopharmacol 29 (2004) 580e588.
[10] F. Rodriguez, I. Rozas, J.E. Ortega, J.J. Meana, L.F. Callado, Guanidine and 2-
aminoimidazoline aromatic derivatives as alpha2-adrenoceptor antagonists,
1: towards new antidepressants with heteroatomic linkers, J. Med. Chem. 50
(2007) 4516e4527.
[11] F. Rodriguez, I. Rozas, J.E. Ortega, A.M. Erdozain, J.J. Meana, L.F. Callado,
Guanidine and 2-aminoimidazoline aromatic derivatives as alpha2-
adrenoceptor antagonists, 2: exploring alkyl linkers for new antidepres-
sants, J. Med. Chem. 51 (2008) 3304e3312.
[32] J. Gumbart, L.G. Trabuco, E. Schreiner, E. Villa, K. Schulten, Regulation of the
protein-conducting channel by
1453e1464.
a bound ribosome, Structure 17 (2009)
[33] J. Huang, A.D. MacKerell Jr., CHARMM36 all-atom additive protein force field:
validation based on comparison to NMR data, J. Comput. Chem. 34 (2013)
2135e2145.
[34] K. Vanommeslaeghe, E. Hatcher, C. Acharya, S. Kundu, S. Zhong, J. Shim,
E. Darian, O. Guvench, P. Lopes, I. Vorobyov, A.D. MacKerell Jr., CHARMM
general force field: a force field for drug-like molecules compatible with the
CHARMM all-atom additive biological force fields, J. Comput. Chem. 31 (2010)
671e690.
[12] F. Rodriguez, I. Rozas, J.E. Ortega, A.M. Erdozain, J.J. Meana, L.F. Callado,
Guanidine and 2-aminoimidazoline aromatic derivatives as a2-adrenoceptor
ligands: searching for Structure-Activity relationship, J. Med. Chem. 52
(2009) 601e609.
[35] C.W. Hopkins, S. Le Grand, R.C. Walker, A.E. Roitberg, Long-time-step molec-
ular dynamics through hydrogen mass repartitioning, J. Chem. Theor. Comput.
11 (2015) 1864e1874.
[13] D.H. O’Donovan, C. Muguruza, L.F. Callado, I. Rozas, Guanidine-based
a2-
adrenoceptor ligands: towards selective antagonist activity, Eur. J. Med.
Chem. 82 (2014) 242e254.
[14] B. Kelly, M. McMullan, C. Muguruza, J.E. Ortega, J.J. Meana, L.F. Callado,
[36] D.A. Case, J.T. Berryman, R.M. Betz, D.S. Cerutti, T.E. Cheatham III, T.A. Darden,
R.E. Duke, T.J. Giese, H. Gohlke, A.W. Goetz, N. Homeyer, S. Izadi, P. Janowski,
J. Kaus, A. Kovalenko, T.S. Lee, S. LeGrand, P. Li, T. Luchko, R. Luo, B. Madej,
K.M. Merz, G. Monard, P. Needham, H. Nguyen, H.T. Nguyen, I. Omelyan,
A. Onufriev, D.R. Roe, A. Roitberg, R. Salomon-Ferrer, C.L. Simmerling,
W. Smith, J. Swails, R.C. Walker, J. Wang, R.M. Wolf, X. Wu, D.M. York,
P.A. Kollman, AMBER 2015, University of California, San Francisco, 2015.
[37] W. Humphrey, A. Dalke, K. Schulten, VMD - visual molecular dynamics, J. Mol.
Graph. 14 (1996) 33e38.
[38] D.R. Roe, T.E. Cheatham III, PTRAJ and CPPTRAJ: software for processing and
analysis of molecular dynamics trajectory data, J. Chem. Theor. Comput. 9
(2013) 3084e3095.
[39] The PyMOL Molecular Graphics System, Version 2.0 Schrodinger, LLC.
[40] A.I. Vogel, A.R. Tatchell, B.S. Furnis, A.J. Hannaford, Vogel’s Textbook of Prac-
tical Organic Chemistry, fifth ed., 1996.
I. Rozas, a2-Adrenoceptor antagonists: synthesis, pharmacological evaluation
and molecular modelling investigation of pyridinyl guanidine/2-
aminoimidazoline and their derivatives, J. Med. Chem. 58 (2015) 963e977.
[15] M. McMullan, A. Garcia-Bea, P. Miranda-Azpiazu, L.F. Callado, I. Rozas,
Substituted conformationally restricted guanidine derivatives: probing the
a
2-adrenoceptors’ binding pocket, Eur. J. Med. Chem. 123 (2016) 48e57.
[16] M.P. Groziak, L.B. Townsend, A new and efficient synthesis of guanosine,
J. Org. Chem. 51 (1986) 1277e1282.
[17] L.F. Callado, J.J. Meana, Low-affinity conditions for agonists increase the
binding of the antagonist [3H]RX821002 to the a2B/C-adrenoceptor subtypes
in human brain and rat kidney, Eur. J. Pharmacol. 332 (1997) 109e112.
[18] P. Marini, M.-G. Cascio, A. King, R.G. Pertwee, R.A. Ross, Characterization of
cannabinoid receptor ligands in tissues natively expressing cannabinoid CB2
€
14