Mastering tricyclic ring systems for desirable functional cannabinoid activity

Article abstract of DOI:10.1016/j.ejmech.2013.09.038

There is growing interest in using cannabinoid receptor 2 (CB2) agonists for the treatment of neuropathic pain and other indications. In continuation of our ongoing program aiming for the development of new small molecule cannabinoid ligands, we have synthesized a novel series of carbazole and γ-carboline derivatives. The affinities of the newly synthesized compounds were determined by a competitive radioligand displacement assay for human CB2 cannabinoid receptor and rat CB1 cannabinoid receptor. Functional activity and selectivity at human CB1 and CB2 receptors were characterized using receptor internalization and [³⁵S]GTP-γ-S assays. The structure-activity relationship and optimization studies of the carbazole series have led to the discovery of a non-selective CB1 and CB2 agonist, compound 4. Our subsequent research efforts to increase CB2 selectivity of this lead compound have led to the discovery of CB2 selective compound 64, which robustly internalized CB2 receptors. Compound 64 had potent inhibitory effects on pain hypersensitivity in a rat model of neuropathic pain. Other potent and CB2 receptor-selective compounds, including compounds 63 and 68, and a selective CB1 agonist, compound 74 were also discovered. In addition, we identified the CB2 ligand 35 which failed to promote CB2 receptor internalization and inhibited compound CP55,940-induced CB2 internalization despite a high CB2 receptor affinity. The present study provides novel tricyclic series as a starting point for further investigations of CB2 pharmacology and pain treatment.

Full text of DOI:10.1016/j.ejmech.2013.09.038

Accepted Manuscript
Mastering tricyclic ring systems for desirable functional cannabinoid activity
Ravil R. Petrov, Lindsay Knight, Shao-Rui Chen, Jim Wager-Miller, Steven W.
McDaniel, Fanny Diaz, Francis Barth, Hui-Lin Pan, Ken Mackie, Claudio N.
Cavasotto, Philippe Diaz
PII:
S0223-5234(13)00619-3
10.1016/j.ejmech.2013.09.038
EJMECH 6441
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 3 July 2013
Revised Date: 16 September 2013
Accepted Date: 19 September 2013
Please cite this article as: R.R. Petrov, L. Knight, S.-R. Chen, J. Wager-Miller, S.W. McDaniel, F.
Diaz, F. Barth, H.-L. Pan, K. Mackie, C.N. Cavasotto, P. Diaz, Mastering tricyclic ring systems for
desirable functional cannabinoid activity, European Journal of Medicinal Chemistry (2013), doi: 10.1016/
j.ejmech.2013.09.038.
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.

ACCEPTED MANUSCRIPT
-
-
-
Discovery of a CB2 selective compound 64 based on a carbazole scaffold
3
5
64 internalized CB2 receptors and activity was confirmed using [ S]GTP-γ-S assays
64 had inhibitory effects on pain hypersensitivity in a rat model of neuropathic pain

ACCEPTED MANUSCRIPT
Mastering tricyclic ring systems for desirable functional cannabinoid activity
a
b
c
b
Ravil R. Petrov, Lindsay Knight, Shao-Rui Chen, Jim Wager-Miller, Steven W.
a
a
d
c
b
McDaniel, Fanny Diaz, Francis Barth, Hui-Lin Pan, Ken Mackie, Claudio N.
e
a
Cavasotto and Philippe Diaz *
a
Core Laboratory for Neuromolecular Production, Department of Biomedical and
Pharmaceutical Sciences, The University of Montana, Missoula, MT 59812, USA.
b
Department of Psychological & Brain Sciences, Gill Center for Biomedical Sciences, Indiana
University, Bloomington, IN 47405, USA.
c
Department of Anesthesiology and Perioperative Medicine, Unit 110, The University of Texas
MD Anderson Cancer Center, 1515 Holcombe Blvd., Houston, Texas. USA.
d
Exploratory Unit, Sanofi-aventis R&D, 371 rue du Professeur Joseph Blayac, Montpellier,
France.
e
Instituto de Investigación Biomedicina de Buenos Aires (IBioBA)-CONICET-Partner Institute
of the Max Planck Society, Godoy Cruz 2390, C1425FQD Buenos Aires, Argentina.
*
Corresponding Author. Telephone: 406-243-4362; fax: 406-243-5228; email address:
Philippe.diaz@umontana.edu.
1

ACCEPTED MANUSCRIPT
Abstract
There is growing interest in using cannabinoid receptor 2 (CB2) agonists for the treatment of
neuropathic pain and other indications. In continuation of our ongoing program aiming for the
development of new small molecule cannabinoid ligands, we have synthesized a novel series of
carbazole and γ-carboline derivatives. The affinities of the newly synthesized compounds were
determined by a competitive radioligand displacement assay for human CB2 cannabinoid
receptor and rat CB1 cannabinoid receptor. Functional activity and selectivity at human CB1 and
35
CB2 receptors were characterized using receptor internalization and [ S]GTP-γ-S assays. The
structure-activity relationship and optimization studies of the carbazole series have led to the
discovery of a non-selective CB1 and CB2 agonist, compound 4. Our subsequent research efforts
to increase CB2 selectivity of this lead compound have led to the discovery of CB2 selective
compound 64, which robustly internalized CB2 receptors. Compound 64 had potent inhibitory
effects on pain hypersensitivity in a rat model of neuropathic pain. Other potent and CB2
receptor–selective compounds, including compounds 63 and 68, and a selective CB1 agonist,
compound 74 were also discovered. In addition, we identified the CB2 ligand 35 which failed to
promote CB2 receptor internalization and inhibited compound CP55,940-induced CB2
internalization despite a high CB2 receptor affinity. The present study provides novel tricyclic
series as a starting point for further investigations of CB2 pharmacology and pain treatment.
Graphical Abstract
2

ACCEPTED MANUSCRIPT
KEYWORDS: Cannabinoid; CB2; GPCR; neuropathic pain; carbazole; CB1.
3

ACCEPTED MANUSCRIPT
1
. Introduction
Two cannabinoid receptors, CB1 and CB2, have been characterized and cloned[1, 2]. The
CB1 receptor is found predominantly in the brain[3]. Impairment of cognitive functions and
psychoactivity induced by cannabinoids ligands are mediated by CB1 receptors activation[4].
Selective activation of CB2 receptors has been proposed as a strategy to curtail the negative
central side effects seen with nonselective CB1/CB2 agonists. Several CB2-selective agonists
have been described previously[5-11]. Although multiple preclinical studies suggest that the CB2
receptor is a viable target to decrease both acute and neuropathic pain responses[12, 13],
synthetic CB2 agonists have not advanced through clinical trials. In part, this is due to a lack of
a thorough understanding of CB2-mediated analgesic mechanisms[14]. CB2 modulation is also
implicated in immunomodulation and neuroprotection but the functional profile of the CB2
ligands inducing these effects has not been clearly defined[15]. Activation of CB2 receptor
induces its coupling to the Gi/o class of G proteins. The dissociation of the α and βγ subunits
resulting from the CB2 activation can influence multiple effector systems including adenylyl
cyclase, p42/44 MAPK (ERK1/2) or ion channels[16]. Currently, the diversity of cannabinoid
agonists is very broad and continues to expand rapidly (figure 1). In our previous
communications, the SARs of potent and selective CB2- receptor agonists[17] and antagonists[6]
based on the isatin scaffold were described (MDA19 in Figure 1).
4

ACCEPTED MANUSCRIPT
Although the isatin core was identified as a good scaffold for designing CB2-selective
cannabinoid agonists, we hypothesized that the intramolecular hydrogen bond pattern in the
isatin scaffold can be replaced with a covalent C-C bond. Also considering the high
pharmacological potential of carbazole-based natural products[18], we focused our attention on
this scaffold for use of studying CB1/CB2 receptor pharmacology. Thus, we first identified
compound 4 as a potent non-selective CB1/CB2 agonist[19]. We recently reported the
cannabinoid and T-type calcium channel activities for three compounds from this novel series of
cannabinoid ligands[7]. In this article, we optimized the CB2 selectivity of this series. We also
explored the effect on CB2 activity of the introduction of quaternary ammonium moieties or
introduction of heteroatoms increasing the polar surface area to help future identification of
peripherally-restricted CB2 agonists in our tricyclic series. This research effort led to the
2
discovery of a selective CB2 agonists with a polar surface area > 70 Å and expected to have a
low blood-brain barrier (BBB) permeability. Functional activities of our most active compounds
3
5
were determined using [ S]GTP-γ-S assays and by characterizing their ability to internalize
human CB1 and CB2 receptors. The ability of these compounds to interact with the orthosteric
site on human CB2 receptor was assessed by determining their ability to block CP55940-induced
hCB2 internalization. Moreover, the binding mode prediction through ligand-steered modeling
highlighted a potential H-bond interaction between the alkylsulfonamide moiety in the N-1
position bored by compound 64 and N7.45 of the CB2 receptor. Similarly to what was
previously observed within the isatin series, the presence of the OMe fragment in the carbazole
scaffold turned out to be a key substituent responsible for the agonist-to-antagonist functionality
switch.
5

ACCEPTED MANUSCRIPT
2
. Results and discussion
.1. Chemistry
2
The pathway of achieving the synthesis of the first series of compounds (4-11 and 14-23) is
outlined in Scheme 1. We found that desirable analogs 4-11 can be conveniently prepared from
commercially available carbazole via consecutive substitution with n-pentyl bromide under
alkaline conditions followed by electrophilic formylation using standard Vilsmeier–Haack
conditions, oxidation, and, finally, amidation under standard peptide coupling conditions. Later,
a more convenient method was developed for the synthesis of compound 4 that utilized a direct
Friedel-Crafts reaction using piperidinecarbonyl chloride. While our conditions for the Friedel-
Crafts reaction (compounds 4, 18, and 19) were not expected to be optimal, this strategy did
provide a rapid access to the desired compounds. Compound 19 was subjected to nucleophilic
attack by the corresponding amines to give analogs 20-22. Deprotection of the Boc protecting
group in compounds 12 and 13 furnished compounds 14–15. Analogs 16 and 17 were prepared
by coupling of corresponding bromomethylpyridyl derivatives and acid 3 in the presence of
TBAF under basic conditions [20]. The tertiary base 9 was converted to the quaternary form 23
by treatment with methyl iodide in diethyl ether at room temperature.
Next, we focused on modifications (Scheme 2) that led to a series of compounds in which the
original 3-carbonyl group in the carbazole framework was either derivatized or entirely
eliminated. Compounds 24-26, 28 and 30 were synthesized following protocols as outlined in
Scheme 2. Nitrile 24 was prepared by a one-pot solvent-free procedure from aldehyde 2 and
hydroxylamine. Thioamide 25 was prepared by heating amide 4 with Lawesson's reagent under
microwave conditions. Tertiary amine 26 was obtained upon treatment of 4 with LAH. The
6

ACCEPTED MANUSCRIPT
preparation of amines 28 and 30 was achieved by palladium-catalyzed coupling in analogy to
known literature methods (e.g.[21]).
The pyrido[3,4-b]indole-based analogs 33-35 were prepared as depicted in Scheme 3.
Commercially available ethyl 9H-pyrido[3,4-b]indole-3-carboxylate was alkylated with n-pentyl
bromide under microwave conditions. Hydrolysis of ester 31 gave the corresponding carboxylic
acid 31. Activation of acid 32, followed by coupling with the corresponding amines provided the
corresponding target carboxamides 33-35.
A series of N-substituted analogues were prepared according to Scheme 4. In an analogous
fashion as previously described for compound 4, N-alkylations of carbazole (compounds 36-39)
followed by Friedel–Crafts reaction using piperidinecarbonyl chloride onto the resultant
substrates led to analogs 40-42. N-alkyl derivatives 43-44 were prepared according to the
literature[22]. Intermediate 44 was prepared from carbazole and ethylene oxide under basic
conditions. Conversion of alkyl carbazoles 37, 44, and 45 to the corresponding aldehydes 46-49
was achieved using standard Vilsmeier–Haack conditions as described above. Vilsmeier–Haack
reaction allowed simultaneous installation of an aldehyde group into the aromatic core and a
halogen atom into the N-alkyl chain. Similarly, the obtained aldehydes were then oxidized to the
corresponding carboxylic acids 50-53. After activation, these acids were used for coupling with
the corresponding amines under standard peptide coupling conditions to furnish compounds 54-
5
5
7. The N-alkylated chloride derivatives 56 and 57 were converted to the corresponding iodides
8 and 59. Taking advantage of the ease of nucleophilic displacement of the iodide substituent,
compounds 60-64 were obtained by base-mediated aminolysis of compounds 58 and 59.
Although some of the targets from this series could still be prepared starting from carbazole, we
realized that it would more expedient to achieve the same goal by using Buchwald–Hartwig
7

ACCEPTED MANUSCRIPT
amination and subsequent palladium-mediated oxidative cyclization of the diarylamines [23].
Following this strategy, the diaryl amine 65 was formed and then oxidized with Pd(OAc) to give
2
the carbazole ester 66. Subsequent ester hydrolysis of 66 delivered the corresponding acid which
was amidated with piperidine to produce the generic compound 67. Alkylation of carbazole 67
afforded the target compounds 68-76. Deprotection of the Boc group within 72 afforded
compound 73 after aqueous work-up with sodium hydroxide.
Following a procedure analogous to that of Scheme 4 for compound 67, compounds 81, 84, and
8
5 were prepared as depicted in Scheme 5. Alternatively, further modification of the
carboxamide part of analog 55 was enabled starting from carboxylic acid 51, which was
subjected to amide coupling conditions to afford the desired amides 82 and 83.
2
.2.
Binding affinities for cannabinoid receptors and modeling: structure-activity
relationship
We previously identified a series of isatin acylhydrazone CB2 modulators[6, 17]. These
compounds showed potent CB2 agonist or antagonist activities. The lead compound among the
isatin series possessed potent antiallodynic effects in a rat model of neuropathic pain without
affecting the rat locomotor activity[13] at the therapeutic dose. We next turned our attention to
substituting the isatin acylhydrazone core with a carbazole scaffold, which is devoid of the
potential hydrazone bond isomerization and would have improved potency due to its rigid
structural template that may ensure the precise spatial orientation of important functional groups.
Starting with the lead compound 4 we carried out structural modifications in order to expand the
SARs and optimize binding affinity, selectivity, and functional activity associated with the CB2
8

ACCEPTED MANUSCRIPT
receptor. The CB1 and CB2 receptors binding affinities were determined by performing
3
[
H]CP55,940 radioligand competition binding experiments (Table 1-5). Previous studies have
shown that binding affinities of CB1 ligands were in the same range for human and rat CB1
receptors[12, 24].
The structural model of 4 complexed with the CB2 receptor is shown in Fig. 2. The ligand
occupies a cavity defined by helices 3, 5, 6 and 7, with the alkyl chain buried within the binding
site, and the piperidine moiety facing the extracellular side. In agreement with this predicted
binding mode, deletion of the chain born by the endocyclic nitrogen in 67 was detrimental to its
binding. Replacement of the piperidine ring by a nitrile moiety in 24 or a methyl ester moiety in
3
1 had a strong negative effect on CB2 affinity, probably due to the lack of ligand-receptor van
der Waals contacts. In order to further improve the physicochemical properties and metabolic
stability of compound 4, we decided to incorporate polar groups at position 4 of the piperidine
ring[26]. However, such a chemical modification did not produce significant improvement in
affinity towards the CB2 receptor for analogues 6 and 8. Alternatively, replacing the piperidine
ring by the more polar N-methylpiperazine ring (compound 9), decreased affinity at CB1 but
increased toward CB2, resulting in a 3-fold improvement of CB2 selectivity compared to the
piperidine analogue 4. It is notable that the removal of the methyl moiety (compound 15) from
the N-methylpiperazine analogue, dramatically decreased the CB2 binding affinity indicating
that a hydrophobic moiety in this part of the molecule may be essential for the CB2 affinity. In
contrast, the CB1 affinity in this case decreased by only 4-fold. Next, the quaternary ammonium
derivative 23 was prepared in order to probe electrostatic interactions with CB2 and investigate if
this strategy can be used to design CB2 agonists that do not cross the blood-brain barrier.
9

ACCEPTED MANUSCRIPT
Unfortunately, compound 23 turned out to be devoid of CB1 or CB2 affinity. This result lends
further support to our view that lipophilicity is a required feature on the carboxamide part of the
molecule. Since there was compelling evidence that a lipophilic moiety is required for a good
CB2 affinity, we decided to add a methylene moiety or a nitrogen spacer in between the carbonyl
and the lipophilic piperidine moiety to test whether the biological profile can be further
improved. However, as compared to the piperidine analogue 4, the N-aminopiperidine derivative
7
, the N-methylpiperidine derivative 20 and the methylpiperidine derivative 14 did not show
relevant CB1 or CB2 affinities, thus suggesting a scarce adaptability of the lipophilic pocket in
the CB2 binding site. One can notice that among common features present in these three
compounds is not only a longer substituent, but also a tertiary amine offering a site of N-cation
interactions. Therefore, it may be speculated that perhaps either the increased basicity of the
tertiary amine or the subsequent cation interactions are responsible for destabilizing the complex.
The thiomorpholine 1,1-dioxide analogue 21 exhibited moderate affinities for both CB1 and
CB2. Our attempts to increase the conformational flexibility in the carboxamide region by
opening the piperidine cycle, as exemplified by compound 5 slightly increased CB2 affinity and
the selectivity toward CB2. Replacement of the piperidine by the planar phenyl ring, as in
compound 18, resulted in a slight decrease of CB1 affinity, while the CB2 affinity was in the
range of the parent compound. These observations support the binding mode proposed for 4
(Figure 2). On the other hand, an increase of lipophilicity in compound 10, resulted in an
improvement in CB2 selectivity. In this case the adamantyl group is probably too bulky to fit
properly in the CB1 receptor-binding pocket. One of the unique benefits of “soft drugs” such as
esters for treating inflammatory disorders lies in the susceptibility of their ester functionality to
hydrolysis, which can limit their CNS penetration. Therefore, such molecules should lack the
1
0

ACCEPTED MANUSCRIPT
undesired CNS-induced side effects associated with CB1 receptor activation. The pyridine ester
analogues 16 and 17 were synthesized to build aqueous solubility required for bioavailabilty via
the hydrolysable ester bonds. Surprisingly, both compounds exhibited higher affinity at the CB1
rather than at the CB2 receptor. This decrease in CB2 affinity is in good agreement with what is
reported above regarding the longer substituents. Further exploration of the SARs around the
carbazole series led us to examine the impact of the amide toward CB2 affinity. Replacement of
the carbonyl (compound 4) by a methylene in compound 26 resulted in a decrease of CB2
affinity while the CB1 affinity was in the same range, what could be explained in terms of the
lost hydrogen-bond with the OH group of S7.39 (cf. Figure 2). On the other hand, the
replacement of the carbonyl by the thiocarbonyl in 25 resulted in the CB2 affinity being restored.
Replacement of amide bonds by thio-amide bonds have been shown to destabilize hydrogen
bond by the higher steric demands imposed by the larger sulfur atom, which leads to non-optimal
angles required to form hydrogen bonding[27]. In addition, the sulfur atom has lower
electronegativity compared to oxygen atom. However, C=S bond in 25 may induce the required
conformation and with the correct position of the piperidine ring needed for high CB2
affinity[28]. Nevertheless, thioamide derivatives cannot be considered as viable alternatives to
design potent CB2 agonists since it was shown that thioamides behave as amide prodrugs in vivo.
As expected, the piperazine analogues 28 and 30 did not exhibit potent CB2 affinity even though
the CB1 affinity was slightly lower for 28 compared 9. Introduction of an extra nitrogen atom in
the carbazole scaffold was expected to impact both water solubility and positioning of the
lipophilic moiety attached to the carbonyl. As a result, CB2 affinity was slightly increased for the
carboline analogue 33 when compared to the respective carbazole analogue. For the former, the
CB1 affinity was in the same range as for its carbazole counterpart 4. Notably, the CB2 affinity
1
1

ACCEPTED MANUSCRIPT
was substantially higher for the N-aminopiperidine carboline derivative 34 as compared to the
carbazole derivative 7. This higher affinity could potentially result from a hydrogen bond
between the secondary amino group and the nitrogen on the carboline scaffold, thus providing
some critical assistance in accommodating the respective lipophilic moiety in the CB2 binding
site. Introduction of the neopentyl moiety in 35 resulted in increased affinity at both CB1 and
CB2 but the selectivity toward CB2 receptor was rather low (ca. 1.65).
In the course of our initial exploratory work on the structure–activity relationship for this novel
series of CB2 modulators, we decided to determine the optimal chain length attached to the
carbazole’s endocyclic nitrogen (compounds 40 to 54). Among the linear N-1 alkyl chains, a
pentyl chain seemed to be the most optimal for occupying the CB2 lipophilic cavity, because
systematically increasing and decreasing the length from n-pentyl in 4 negatively impacted the
respective CB2 affinities. Introduction of halogen atoms was shown to increase the affinity of
small molecules toward proteins even though halogen bond interactions are significantly weaker
than hydrogen bond interactions. Previous results from literature suggest that interactions
between C-Cl moieties and carbonyl groups or other H-bond acceptors increase affinity of
corresponding ligands to biological targets[29]. Introduction of a chlorine atom into the ethyl or
propyl chain of 56 and 57 did not increase CB2 affinity. The longer chloropropyl chain in 57
mimicked the pentyl chain in 4, while the use of a shorter one in 56 suffered a dramatic loss of
CB2 affinity. Hydrogen bond interactions were shown to provide specificity to the process of
receptor recognition although not always did they add much of free binding energy[30]. Since
we achieved only modest improvements in terms of CB2 selectivity, we next sought to expand
our library by targeting potential hydrogen bond interactions between the chain carried by the
1
2

ACCEPTED MANUSCRIPT
endocyclic nitrogen and CB2 receptor. First, we turned our attention on to alkyl ethers, despite
the fact that ethers are known to be weak hydrogen bond acceptors. However, an interesting
feature of ethers in this regard is that they exhibit a somewhat higher tolerance for angular
requirement between the hydrogen-donor and acceptor[30]. In order to explore the SARs within
this context, we synthesized acyclic (compound 68) and cyclic (compound 69) ethers. The cyclic
ethers are known to be stronger hydrogen bond acceptors than the acyclic ones[30]. Consistent
with our predictions (Fig. 2), 69 demonstrated higher CB2 affinity as compared to 4 although the
CB2 selectivity remained poor (Fig. 2). A replacement of the 4-methyltetrahydropyran fragment
by the 4-methylpyridine moiety as exemplified by compound 74 resulted in a 4-fold
improvement in CB2 selectivity, while the respective CB2 affinities were similar. The binding
mode of 74 is shown in Fig. 3, exhibiting a similar pose when compared to 4 with hydrogen
bonding occurring between the amide carbonyl of 74 and S7.39. In this case, the only difference
is that an additional hydrogen bond is present between the nitrogen of the 4-methylpypridine and
the side chain of W6.48. Consistent with this model, the 2- and 3-methylpyridine analogues
(compounds 75 and 76) exhibited low CB2 affinities. In contrast replacement of the pentyl chain
by the acyclic ether in compound 68 resulted in a significant increase of selectivity due to a
specific decrease CB1 affinity. Acyclic esters functionality such as in compound 70 resulted in a
dramatic improvement in selectivity due to significantly lower affinity of this compound at the
CB1 receptor. Interestingly, this effect was observed for both the ester analogue and the dimethyl
amine analogue 71. A significant loss of CB2 binding affinity in these cases may be attributed to
three obstacles such as higher degree of angular preferences, higher steric demands, or difference
in electronic properties. In comparison, the ethyl ester counterpart 55, which has the ester
functionality closer to the carbazole scaffold did not exhibit significant affinity at CB1, and its
1
3

ACCEPTED MANUSCRIPT
affinity at CB2 was not promising either. In part, this might be explained by poor ability of 55 to
promote hydrogen bonding with the corresponding receptors. When the nitrogen atom was free
of substituents as in 73 or carried a bulky Boc moiety as in 72, a significant loss of CB2 affinity
was observed. In the former case, the charged terminal amine group was likely to destabilize the
complex, which is consistent with our previous observations. Employment of alkyl chains
featuring the sulfonyl functional group is typical in cases when a further increase in the PSA
values of the corresponding analogs is desirable. The dual character of the weakly polar sulfonyl
group which allows it to form hydrogen bonds and at the same time participate in nonpolar
hydrophobic interactions through van der Waals contacts drawn our attention to the possibility of
applying this strategy to improve CB2 affinity[30, 31]. As expected, the methyl sulfonamide
derivative 64 demonstrated the most specific CB2 affinity.
The binding mode of compound 64 is shown in Figure 4; the pose is very similar to that of
compound 4 (Fig. 2). The first thing one can notice in this model is a weak hydrogen bond
between the oxygen of the sulfonyl group and the side chain of N7.45 in addition to a hydrogen
bond between the amide carbonyl of 64 and S7.39. Introducing a basic residue such as a
piperidine ring into the N-aliphatic side chain resulted in the inactive compounds (61 and 60). A
possible explanation for this might be that the charged amine in this case is buried within the
potential hydrophobic binding site. In contrast, the dioxothiomorpholine counterpart 62 regained
both good selectivity and CB2 affinity. A similar trend was observed for analogue 63 which
offered a 30-fold selectivity over CB1. For comparative purposes, we next decided to carry out
structure-activity relationship studies looking at the effects of ring substitution to determine if
they would parallel those in our previous isatin series. Indeed, we observed a similar result when
1
4

ACCEPTED MANUSCRIPT
a methoxy group was introduced in position 6 of the carbazole scaffold resulting in compound 81
which displayed a higher CB2 affinity and was endowed with inverse agonist instead of agonist
functional activity.
In the carbazole series, we also briefly explored whether it is possible to impact the spatial
position of the N-alkyl chain by introducing substituents in position 4 of the piperidine ring. As
shown in Table 5, both the methylpiperazine analogue 82 and the dioxothiomorpholine analogue
8
3 turned out to be lacking CB2 affinity. This result may be explained by the fact that the
methylpiperazine and dioxothiomorpholine rings did not allow the chain containing the ester
moiety to be set further down into the lipophilic pocket in order to restore the CB2 affinity
within the same range of 70. As expected (compare with 72), the dioxothiomorpholine analogue
8
5 bearing a distal Boc moiety on the N-propyl chain showed no measurable activity against
both the CB1 and CB2 receptors.
2
.3.
Cannabinoid receptors functional activity
3
5
Functional activity for compound 64 was evaluated by using a [ S] guanosine-5'-triphosphate
(GTP)-γ-S assay in Chinese hamster ovarian cell membrane extracts expressing recombinant
3
5
hCB1 or hCB2 receptor. In this system, agonists stimulate [ S]GTP-γ-S binding, whereas
3
5
antagonists have no effect and inverse agonists decrease [ S]GTP-γ-S basal binding. Efficacies
Emax) for CB1 or CB2 were expressed as a percentage of the efficacy of compound CP55,940.
Compound 64 behaved as a selective CB2 agonist since it did not show any agonistic or
(
35
antagonistic activities at human CB1 receptors in the [ S] GTPγS assay at 1 µM whereas the
1
5

ACCEPTED MANUSCRIPT
EC value for the CB2 receptor was 7.3 nM, with an efficacy of almost 60%. Thus, compound
50
6
4 appears to be a selective CB2 agonist.
Receptor internalization and functional selectivity are important emerging pharmacological
concepts that increase the complexity of the biological profiling of G protein-coupled receptor
(GPCR) ligands. Recently, it has been shown that WIN55,212-2 and other aminoalkylindoles
failed to promote CB2 receptor internalization, whereas compound CP55,940 robustly
internalized CB2 receptors[14]. Despite these differences in term of CB2 induced-internalization,
both compounds activated CB2 receptors mediated ERK 1/2 phosphorylation and recruited β-
arrestin to the membrane. In contrast, whereas CP55,940 inhibited voltage-gated calcium
2
channels via CB2 receptor activation, WIN55,212-2 was ineffective on its own and antagonized
the effects of CP55,940. These differences in terms of functional activity and the
polypharmacology associated with cannabinoids ligands[7, 32] can possibly explain the
variability observed in preclinical models of neuropathic pain and the difficulties of translating
results from the preclinical models to human pain states. As a way for rapidly screening for
orthosteric interactions with CB1 and CB2 receptors, internalizations of these receptors induced
by our most active compounds was determined (Table 7). In our series of compounds both CB1
3
5
and CB2 receptor internalization correlated well with the GTPγ[ S] functional assays since 81, a
35
partial CB1 agonist and CB2 inverse agonist in the GTPγ[ S] functional assays only promoted
3
3% of CB1 internalization and failed to internalize CB2 at 10 µM. In contrast, 4 which was a
35
full CB1 agonist and a partial CB2 agonist in the GTPγ[ S] functional assay, promoted
internalization of both hCB1 and hCB2 to similar extent. Despite high CB2 affinities,
compounds 5, 9, 18 and 63 promoted only partial CB2 internalization. These data suggest that
1
6

ACCEPTED MANUSCRIPT
these compounds behave as partial agonist at CB2, at least with respect to the internalization.
Similarly, 6 and 35 behaved as CB1 partial agonists as they partially induced CB1 internalization
despite the high CB1 affinity. Interestingly, compound 26 failed to induce CB1 internalization
despite a CB1 affinity of around 100 nM. This result may be explained by the fact that the
carbonyl moiety between the carbazole scaffold and the piperidine is essential for CB1
internalization as exemplified by compound 4, which induced CB1 internalization. Low CB1
affinity is associated with low CB1 induced internalization as illustrated by compounds 70 or 20.
The lack of CB2 internalization observed with 35 could potentially result from the hydrogen
bond between the secondary amino group and the nitrogen on the carboline scaffold. This
hydrogen bond may compromise the conformational flexibility of 35 required by the CB2
receptor lipophilic pocket to accommodate the neopentyl moiety essential to induce
internalization. Compound 35 produced little CB1 internalization despite a good affinity,
behaving as a partial CB1 agonist.
To further explore the concentration dependence of this effect, we performed complementary
experiments with hCB2 cells. First, we applied a constant 10 nM compound CP55,940 with
cotreatments of 3 uM of selected compounds. As expected, the CB2 inverse agonist compound
8
1 reduced compound CP55,940-induced internalization in hCB2 cells (Fig. 5). Unexpectedly,
compound 18 and 35 reduced by more than 50% hCB2 internalization induced by CP55,940.
These compounds may behave as inverse agonists or despite their effect on internalization, they
may activate CB2 receptors without inducing receptor internalization as has recently been shown
for compound WIN55,212-2. A more detailed investigation into the CB2 functional activity
profiles of these compounds is required to confirm these hypotheses.
2
.4.
In vivo studies
1
7

ACCEPTED MANUSCRIPT
Peripheral nerve injury can cause clinically relevant chronic neuropathic pain. The L5 and L6
spinal nerve ligation produced long-lasting mechanical hypersensitivity (tactile allodynia) on the
ipsilateral hind paw in rats. Since compound 64 exhibited a favorable CB2 selectivity profile and
induced CB2 internalization, we were interested in characterizing its activity in the SNL in vivo
model of neuropathic pain. Compound 64 administered intraperitoneally (5 – 20 mg/kg)
significantly attenuated tactile allodynia in a dose-dependent manner. The higher doses (20
mg/kg and 10 mg/kg) produced a longer duration of the antiallodynic effect than that observed
with the 5 mg/kg of compound 64 (Figure 6).
3
. Conclusion
In this investigation, we presented a broad range of experimental data on the novel series of
carbazole-based cannabinoid ligands. Within this series, sulfonamide analogue 64 was identified
as a selective CB2 agonist. Our structure modeling and docking studies for compound 64 based
on the ligand-steered approach highlighted a potential H-bond interaction in a burrow-like site
between the alkylsulfonamide moiety at the N-1 position and N7.45 of the CB2 receptor. This
3
5
selective CB2 ligand exhibited functional agonist activity as assessed by using [ S]GTP-γ-S
assay and the CB2 internalization study. Similarly to our isatin series, the presence of a methoxy
group in position 7 of the carbazole scaffold in compound 81 turned out to be a key substituent
3
5
responsible for the agonist-to-antagonist functionality switch. Compound 81 exhibited [ S]GTP-
γ-S inverse agonist activity and inhibited compound CP55,940-induced CB2 internalization. On
the basis of our results, compound 35 has a high CB2 affinity, inhibits compound CP55,940-
mediated internalization, and merits further pharmacology characterization. Compound 64,
1
8

ACCEPTED MANUSCRIPT
which induced CB2 internalization, attenuated tactile allodynia in a dose-dependent manner in
the SNL in vivo model of neuropathic pain. In summary, we have identified a novel series of
tricyclic CB2 selective agonists with a well-defined CB2 functional activity that can be used as a
platform for the future development of specific CB2 agonists as treatments of pain. The present
study also provides an additional insight into the internalization of CB2 receptors induced by
CB2 agonist, which should further facilitate optimization of this novel class of tricyclic CB2
modulators for the treatment of pain.
4
. Experimental Section
.1. Synthesis
4
Unless otherwise stated, all reactions were carried out under a nitrogen or argon atmosphere,
using commercially available reagents and solvents. Anhydrous THF and Et O were obtained by
2
distillation from sodium and benzophenone followed by distillation from LAH. All other
solvents are reagent grade and were used without further purification. All procedures were
carried out at room temperature unless otherwise stated. Magnesium sulfate was used as the
drying agent. The crude products were purified by flash chromatography using prepacked
Biotage® cartridges on a Biotage® Isolera separation system. Analytical thin-layer
chromatography (TLC) was performed on precoated, aluminum-backed silica gel (200 µm thick,
Sorbent Technologies, UV254). Melting points were obtained on a Start SMP3 melting point
apparatus and are uncorrected. The microwave irradiation was effected using a Biotage®
Initiator microwave synthesizer. High Resolution mass spectrometry (HRMS) analyses were
performed on a Waters/Micromass LCT-TOF instrument. The HPLC systems used to analyze the
1
9

ACCEPTED MANUSCRIPT
target compounds were either: (i) Waters 2790 high-performance liquid chromatograph with an
autosampler connected to a Waters 2487 dual absorbance UV detector and Waters Micromass
LCT KC290 mass spectrometer or (ii) Waters Alliance e2695 high-performance liquid
chromatograph with an autosampler connected to a Waters 2998 photodiode array detector and
Waters Micromass LCT Premier mass spectrometer. NMR spectra were obtained either on
TM
13
Varian Inova-500 or Bruker Ascend -400 instruments. Chemical shifts for C NMR spectra
are recorded in parts per million using either the central peak of deuterated chloroform (77.23
ppm) or deuterated DMSO (39.51 ppm) as the internal standards. Characteristic splitting patterns
due to spin−spin coupling are expressed as follows: br = broad, s = singlet, d = doublet, t =
triplet, q = quartet, p = pentet, m = multiplet. All coupling constants are measured in hertz. All
synthesized piperidine-amides, except for thioamide analogs, displayed a characteristic
1
3
broadening effect in their C NMR spectra for α- and β-CH carbons of the piperidine ring,
2
reflecting a restricted rotation around the peptide bond. For this reason, the corresponding
chemical shifts values for these analogs were reported only when possible. 2-Hydroxypropyl-β-
cyclodextrin with an average degree of molar substitution of 4.4 was purchased from CTD
Holdings Inc (Alachua, FL, USA)
4
.1.1. 9-Pentyl-9H-carbazole (1). Method A. Under argon atmosphere, a solution of carbazole
(
2.5 g, 14.95 mmol), 1-bromopentane (2.225 mL, 17.94 mmol), and Cs CO (7.3 g, 22.41 mmol)
2
3
in DMF (10 mL) was subjected to microwave irradiation at 140°C for 1 h. The reaction mixture
was cooled, diluted with EtOAc (50 mL), and filtered. The organic solvents were evaporated in
vacuo. The resultant dark oil was distilled under reduced pressure (bp 125°C, 2 mmHg) to afford
2
0

ACCEPTED MANUSCRIPT
1
the title compound as light yellow oil (3.169 g, 89%). H NMR (500 MHz, CDCl ) δ 8.08 (dd, J
3
=
7.6, 0.8 Hz, 2H), 7.47 – 7.41 (m, 2H), 7.37 (d, J = 8.2 Hz, 2H), 7.21 (td, J = 7.5, 1.0 Hz, 2H),
13
4
.24 (t, J = 7.3 Hz, 2H), 1.89 – 1.77 (m, 2H), 1.40 – 1.26 (m, 4H), 0.85 (t, J = 7.1 Hz, 3H). C
NMR and DEPT (126 MHz, CDCl ) δ 140.64 (C), 125.75 (CH), 123.02 (C), 120.52 (CH),
3
1
18.88 (CH), 108.84 (CH), 43.23 (CH ), 29.62 (CH ), 28.87 (CH ), 22.69 (CH ), 14.15 (CH ).
2 2 2 2 3
Method B. A mixture of carbazole (10 g, 59.81 mmol), n-pentyl bromide (11 mL, 88.92 mmol),
and finely ground NaOH (4 g, 100 mmol) in dry acetone (100 mL) was refluxed for 16 h under
nitrogen. After all volatile components were removed by rotary evaporation in vacuo, the residue
was extracted with tert-butyl methyl ether (150 mL). The organic phase was washed with water,
brine, dried (MgSO ), filtered, and evaporated in vacuo. The obtained residue was crystallized
4
from ice-cold ethanol. Yield: 12.72 g (90%); mp 51°C. NMR spectra are identical with those by
Method A.
4
.1.2. 9-Pentyl-9H-carbazole-3-carbaldehyde (2). The title compound was prepared according
to a modified literature procedure[33]. POC1 (2.6 mL, 28.40 mmol) was added, over a period of
3
1
0 min, to an ice-cooled, stirred DMF (7.43 mL, 95.96 mmol) under nitrogen. The reddish
solution was allowed to stir at room temperature for 1 h. 9-Pentyl-9H-carbazole (1) (3.169 g,
3.35 mmol) was added over 10 min, and the obtained mixture was subjected to microwave
1
irradiation at 100°C for 1 h. The reaction mixture was cooled and then poured into crushed ice.
After warming to room temperature, the resultant product was extracted with EtOAc. The
organic phase was washed with water, brine, dried (MgSO ), filtered, and evaporated in vacuo.
4
The obtained residue was purified by column chromatography on silica gel using
heptanes/EtOAc in different proportions to afford the title compound as a white solid (3.49 g,
1
9
9%); mp 63-64°C. H NMR (500 MHz, CDCl ) δ 10.07 (s, 1H), 8.56 (s, 1H), 8.12 (d, J = 7.7
3
Hz, 1H), 7.98 (dd, J = 8.5, 1.1 Hz, 1H), 7.51 (t, J = 7.6 Hz, 1H), 7.43 (d, J = 8.4 Hz, 2H), 7.30 (t,
J = 7.4 Hz, 1H), 4.27 (t, J = 7.3 Hz, 2H), 1.86 (p, J = 7.2 Hz, 2H), 1.34 (d, J = 3.7 Hz, 4H), 0.87
2
1

ACCEPTED MANUSCRIPT
13
(
t, J = 6.9 Hz, 3H). C NMR and DEPT (126 MHz, CDCl ) δ 191.86 (C=O), 144.21 (C), 141.33
3
(C), 128.66 (C), 127.27 (CH), 126.85 (CH), 124.09 (CH), 123.21 (C), 123.15 (C), 120.87 (CH),
1
1
20.43 (CH), 109.55 (CH), 109.07 (CH), 43.54 (CH ), 29.50 (CH ), 28.78 (CH ), 22.59 (CH ),
2 2 2 2
4.08 (CH3).
4
.1.3. 9-Pentyl-9H-carbazole-3-carboxylic acid (3). To an ice-cold solution of 9-pentyl-3-
formylcarbazole (2) (2.96 g, 11.16 mmol) in water/acetone (100 mL, 1:1. v/v) was added
dropwise under stirring a solution of potassium permanganate (1.8 g, 11.39 mmol) in acetone (50
mL). The mixture was heated 3 h at reflux and then allowed to cool to room temperature. After
that the reaction mixture was quenched with ethanol (20 mL), and then stirred for 30 min at
®
reflux. After cooling to room temperature, the mixture was filtered through a pad of Celite and
concentrated in vacuo. The concentrated solution was diluted with water (100 mL), basified with
NaOH to pH ca. 10, and extracted with heptane/ether (4:1, v/v, 50 mL × 3) to remove the
unreacted starting material. The aqueous solution was cooled on an ice-water bath and then
acidified with ice-cold solution of sulfuric acid (20%) to pH ca. 2. The resultant bulky precipitate
was extracted with EtOAc (150 mL). The organic layer was washed with brine (30 mL), dried
over magnesium sulfate, filtered, and concentrated in vacuo. The precipitated product was
collected by filtration, washed with heptanes (20 mL), and dried overnight to produce the title
1
compound 3 (1.743 g, 55%) as a light yellow-to-greenish solid; mp 147°C. H NMR (500 MHz,
CDCl ) δ 12.66 (s, 1H), 8.91 (d, J = 1.5 Hz, 1H), 8.25 (dd, J = 8.6, 1.6 Hz, 1H), 8.15 (d, J = 7.7
3
Hz, 1H), 7.54 – 7.47 (m, 1H), 7.41 (d, J = 8.2 Hz, 1H), 7.39 (d, J = 8.7 Hz, 1H), 7.34 – 7.26 (m,
1
3
H), 4.27 (t, J = 7.3 Hz, 2H), 1.86 (p, J = 7.4 Hz, 2H), 1.40 – 1.29 (m, 4H), 0.87 (t, J = 7.0 Hz,
1
3
H). C NMR and DEPT (126 MHz, CDCl ) δ 173.58 (C=O), 143.80 (C), 141.22 (C), 128.04
3
(
(
(
CH), 126.60 (CH), 123.91 (CH), 123.15 (C), 122.85 (C), 120.92 (CH), 120.17 (CH), 119.84
C), 109.35 (CH), 108.43 (CH), 43.48 (CH ), 29.55 (CH ), 28.82 (CH ), 22.65 (CH ), 14.15
2
2
2
2
CH3).
.1.4. (9-Pentyl-9H-carbazol-3-yl)(piperidin-1-yl)methanone (4). Method A: amide coupling.
4
Carboxylic acid 3 (300 mg, 1.07 mmol), piperidine (215 mg, 2.53 mmol), DIPEA (363 µL, 2.14
mmol), and DMAP (156 mg, 1.28 mmol) were added to DCM (30 mL) under nitrogen. The
obtained solution was cooled down on an ice-water bath. EDC (350 mg, 1.83 mmol) was added
2
2

ACCEPTED MANUSCRIPT
to the solution, and the reaction mixture was stirred for 16 h while warming at room temperature.
The solvent was removed in vacuo, and the obtained residue was extracted with EtOAc (100
mL). The organic layer was washed consecutively with 5% citric acid solution (50 mL × 3),
concentrated sodium bicarbonate (50 mL × 3), brine (50 mL), dried over magnesium sulfate,
filtered, and concentrated in vacuo. The residue was purified on silica gel using heptanes/EtOAc
1
in different proportions to afford the title compound as a light yellow gum (345 mg, 93%). H
NMR (500 MHz, CDCl ) δ 8.19 (d, J = 1.2 Hz, 1H), 8.08 (d, J = 7.7 Hz, 1H), 7.53 (dd, J = 8.4,
3
1
2
.6 Hz, 1H), 7.49 – 7.43 (m, 1H), 7.41 – 7.35 (m, 2H), 7.26 – 7.20 (m, 1H), 4.25 (t, J = 7.2 Hz,
H), 3.73 (br.m, 4H), 1.83 (p, J = 7.3 Hz, 2H), 1.72 – 1.65 (m, 3H), 1.61 (br.s, 3H), 1.36 – 1.29
1
3
(
m, 4H), 0.85 (t, J = 7.0 Hz, 3H). C NMR and DEPT (126 MHz, CDCl ) δ 171.53 (C=O),
3
1
40.99 (C), 140.90 (C), 126.74 (C), 126.13 (CH), 125.08 (CH), 122.74 (C), 122.43 (C), 120.54
(
(
CH), 119.88 (CH), 119.28 (CH), 109.01 (CH), 108.40 (CH), 43.20 (CH ), 29.41 (CH ), 28.71
2
2
+
CH ), 24.79 (CH ), 22.53 (CH ), 14.03 (CH ). ESI: m/z 349.5 (M + H) . HRMS calcd for
2
2
2
3
+
C H N O (M + H) 349.2280, found 349.2311.
23
29
2
4
.1.5. N,N-diethyl-9-pentyl-9H-carbazole-3-carboxamide (5). Using carboxylic acid 3 (112
mg, 0.40 mmol) and diethylamine (74 µL, 0.71 mmol) as starting compounds, the title compound
was prepared as a clear, colorless viscous oil according to the procedure described above for 4.
1
Yield: 44 mg (33%). H NMR (500 MHz, CDCl ) δ 8.16 (d, J = 1.5 Hz, 1H), 8.09 (d, J = 7.8 Hz,
3
1
H), 7.52 (dd, J = 8.4, 1.6 Hz, 1H), 7.50 – 7.46 (m, 1H), 7.42 (d, J = 8.2 Hz, 1H), 7.39 (d, J = 8.4
Hz, 1H), 7.26 – 7.22 (m, 1H), 4.30 (t, J = 7.2 Hz, 2H), 3.83 – 3.16 (br. m, 4H), 1.87 (p, J = 7.4
13
Hz, 2H), 1.35 (dq, J = 7.2, 3.6 Hz, 4H), 1.31 – 1.07 (m, 6H), 0.87 (t, J = 7.1 Hz, 3H). C NMR
and DEPT (126 MHz, CDCl ) δ 172.59 (C=O), 141.00 (C), 140.94 (C), 127.71 (C), 126.23 (CH),
3
1
1
24.61 (CH), 122.87 (C), 122.53 (C), 120.67 (CH), 119.36 (CH), 119.28 (CH), 109.11 (CH),
08.58 (CH), 43.36 (CH ), 29.57 (CH ), 28.85 (CH ), 22.67 (CH ), 14.16 (CH ). ESI: m/z 337.2
2
2
2
2
3
+
+
(
M + H) . HRMS calcd for C H N O (M + H) 337.2280, found 337.2309.
22 29 2
4
.1.6.
(9-Pentyl-9H-carbazol-3-yl)(1,1-dioxo-thiomorpholino)methanone
(6).
Using
carboxylic acid 3 (115 mg, 0.41 mmol) and 1,1-dioxo-thiomorpholine (80 mg, 0.59 mmol) as
starting compounds, the title compound was prepared as a clear, colorless gum according to the
procedure described above for 4. The obtained product solidified upon refrigeration to a white
2
3

ACCEPTED MANUSCRIPT
1
solid Yield: 83 mg (51%); mp 115°C. H NMR (500 MHz, CDCl ) δ 8.23 (d, J = 1.5 Hz, 1H),
3
8
.09 (d, J = 7.7 Hz, 1H), 7.55 (dd, J = 8.4, 1.7 Hz, 1H), 7.51 (ddd, J = 8.2, 7.2, 1.1 Hz, 1H), 7.43
(
(
(
(
(
3
t, J = 8.3 Hz, 2H), 7.30 – 7.24 (m, 1H), 4.30 (t, J = 7.3 Hz, 2H), 4.17 (s, 4H), 3.10 (s, 4H), 1.86
13
p, J = 7.4 Hz, 2H), 1.35 (dq, J = 7.2, 3.7 Hz, 4H), 0.91 – 0.83 (m, 3H). C and DEPT NMR
126 MHz, CDCl ) δ 172.50 (C=O), 141.69 (C), 141.08 (C), 126.72 (CH), 125.12 (CH), 123.87
3
C), 122.79 (C), 122.49 (C), 120.72 (CH), 120.47 (CH), 119.83 (CH), 109.31 (CH), 108.91
CH), 52.18 (CH ), 43.38 (CH ), 29.46 (CH ), 28.76 (CH ), 22.57 (CH ), 14.09 (CH ). ESI: m/z
2
2
2
2
2
3
+
+
99.1 (M + H) . HRMS calcd for C H N O S (M + H) 399.1742, found 399.1718.
2
2
27
2
3
4
.1.7. 9-Pentyl-N-(piperidin-1-yl)-9H-carbazole-3-carboxamide (7). Using carboxylic acid 3
(100 mg, 0.36 mmol) and 1-aminopiperidine (39 µL, 0.36 mmol) as starting compounds, the title
compound was prepared as a light yellow solid according to the procedure described above for 4
with the exception of not using 5% citric acid wash during the workup. Yield: 109 mg (84%); mp
1
1
1
58°C. H NMR (500 MHz, CDCl ) δ 8.51 (s, 1H), 8.11 (d, J = 7.7 Hz, 1H), 7.86 (d, J = 8.3 Hz,
3
H), 7.49 (ddd, J = 8.2, 7.1, 1.0 Hz, 1H), 7.41 (d, J = 8.2 Hz, 1H), 7.36 (d, J = 8.6 Hz, 1H), 7.26
(dd, J = 8.1, 6.8 Hz, 1H), 6.96 (s, 1H), 4.28 (t, J = 7.2 Hz, 2H), 2.93 (br. s, 4H), 1.85 (p, J = 7.5
1
3
Hz, 2H), 1.82 – 1.73 (m, 4H), 1.46 (br. s, 2H), 1.38 – 1.30 (m, 4H), 0.86 (t, J = 6.8 Hz, 3H). C
NMR and DEPT (126 MHz, CDCl ) δ 166.22 (C=O), 142.29 (C), 141.06 (C), 126.35 (CH),
3
1
1
2
24.90 (CH), 124.62 (C), 122.92 (C), 122.64 (C), 120.75 (CH), 120.02 (CH), 119.63 (CH),
09.16 (CH), 108.49 (CH), 57.45 (CH ), 43.31 (CH ), 29.47 (CH ), 28.75 (CH ), 25.57 (CH ),
2
2
2
2
2
+
3.50 (CH ), 22.58 (CH ), 14.09 (CH ). ESI: m/z 364.2 (M + H) . HRMS calcd for C H N O
2
2
3
23 30
3
+
(
M + H) 364.2389, found 364.2356.
4
.1.8. Morpholino(9-pentyl-9H-carbazol-3-yl)methanone (8). Using carboxylic acid 3 (100
mg, 0.36 mmol) and morpholine (62 µL, 0.71 mmol) as starting compounds, the title compound
was prepared as a white solid according to the procedure described above for 4. Yield: 109 mg
1
(
84%); mp 98°C. H NMR (500 MHz, CDCl ) δ 8.21 (d, J = 1.6 Hz, 1H), 8.09 (d, J = 7.7 Hz,
3
1
7
1
H), 7.54 (dd, J = 8.4, 1.7 Hz, 1H), 7.48 (ddd, J = 8.2, 7.1, 1.2 Hz, 1H), 7.40 (t, J = 8.5 Hz, 2H),
.25 (ddd, J = 7.9, 7.0, 0.9 Hz, 1H), 4.28 (t, J = 7.2 Hz, 2H), 3.73 (s, 8H), 1.90 – 1.79 (m, 2H),
1
3
.38 – 1.28 (m, 4H), 0.86 (t, J = 7.0 Hz, 3H). C NMR and DEPT (126 MHz, CDCl ) δ 171.72
3
(C=O), 141.26 (C), 140.98(C), 126.36 (CH), 125.45 (C), 125.29(CH), 122.66 (C), 122.57 (C),
2
4

ACCEPTED MANUSCRIPT
1
2
20.63 (CH), 120.29 (CH), 119.50 (CH), 109.14 (CH), 108.59 (CH), 67.11 (CH ), 43.28 (CH ),
2
2
+
9.45 (CH ), 28.74 (CH ), 22.57 (CH ), 14.08 (CH ). ESI: m/z 351.2 (M + H) . HRMS calcd for
2
2
2
3
+
C H N O (M + H) 351.2073, found 351.2071.
22
27
2
2
4
.1.9. (4-Methylpiperazin-1-yl)(9-pentyl-9H-carbazol-3-yl)methanone (9). Using carboxylic
acid 3 (110 mg, 0.39 mmol) and 1-methylpiperazine (71 mg, 0.71 mmol) as starting compounds,
the title compound was prepared as a clear, orange viscous oil according to the procedure
described above for 4 with the exception of not using 5% citric acid during the workup. Yield:
1
1
7
34 mg (94%). H NMR (500 MHz, CDCl ) δ 8.20 (d, J = 1.3 Hz, 1H), 8.09 (d, J = 7.7 Hz, 1H),
3
.55 (dd, J = 8.4, 1.6 Hz, 1H), 7.48 (ddd, J = 8.2, 7.1, 1.1 Hz, 1H), 7.41 (d, J = 8.3 Hz, 1H), 7.39
(d, J = 8.5 Hz, 1H), 7.27 – 7.22 (m, 1H), 4.28 (t, J = 7.2 Hz, 2H), 3.78 (br. s, 4H), 2.46 (br. s,
13
4
H), 2.33 (s, 3H), 1.85 (p, J = 7.3 Hz, 2H), 1.37 – 1.32 (m, 4H), 0.86 (t, J = 7.0 Hz, 3H). C
NMR and DEPT (126 MHz, CDCl ) δ 171.63 (C=O), 141.19 (C), 140.97 (C), 126.30 (CH),
3
1
1
1
3
25.97 (C), 125.31 (CH), 122.72 (C), 122.51 (C), 120.64 (CH), 120.21 (CH), 119.44 (CH),
09.11 (CH), 108.52 (CH), 46.21 (CH ), 43.29 (CH ), 29.47 (CH ), 28.76 (CH ), 22.58 (CH ),
3
2
2
2
2
+
+
4.09 (CH ). ESI: m/z 364.2 (M + H) . HRMS calcd for C H N O (M + H) 364.2389, found
3
23 30
3
64.2356.
4
.1.10. N-(1-adamantyl)-9-pentyl-9H-carbazole-3-carboxamide (10). Using carboxylic acid 3
(112 mg, 0.40 mmol) and 1-adamantylamine (74 µL, 0.71 mmol) as starting compounds, the title
compound was prepared as a light yellow glass according to the procedure described above for
1
compound 4. Yield: 91 mg (62%). H NMR (500 MHz, CDCl ) δ 8.48 (d, J = 1.7 Hz, 1H), 8.11
3
(
d, J = 7.7 Hz, 1H), 7.85 (dd, J = 8.6, 1.8 Hz, 1H), 7.48 (ddd, J = 8.2, 7.1, 1.2 Hz, 1H), 7.40 (d, J
8.2 Hz, 1H), 7.35 (d, J = 8.6 Hz, 1H), 7.29 – 7.22 (m, 1H), 5.97 (s, 1H), 4.26 (t, J = 7.2 Hz,
H), 2.20 (d, J = 2.6 Hz, 6H), 2.15 (br. s, 3H), 1.88 – 1.80 (m, 2H), 1.79 – 1.69 (m, 6H), 1.37 –
=
2
1
1
1
3
.28 (m, 4H), 0.85 (t, J = 7.0 Hz, 3H). C NMR and DEPT (126 MHz, CDCl ) δ 167.51 (C=O),
3
42.14 (C), 141.13 (C), 126.79 (C), 126.31 (CH), 124.69 (CH), 123.03 (C), 122.62 (C), 120.70
(
CH), 119.59 (CH), 119.56 (CH), 109.19 (CH), 108.41 (CH), 52.36 (C), 43.36 (CH ), 42.01
2
(
CH ), 36.64 (CH ), 29.74 (CH), 29.51 (CH ), 28.80 (CH ), 22.62 (CH ), 14.13 (CH ). ESI: m/z
2
2
2
2
2
3
+
+
4
15.2 (M + H) . HRMS calcd for C H N O (M + H) 415.2749, found 415.2762.
28 35 2
2
5

ACCEPTED MANUSCRIPT
4
.1.11. N-(4-chlorophenethyl)-9-pentyl-9H-carbazole-3-carboxamide (11). Using carboxylic
acid 3 (105 mg, 0.37 mmol) and 2-(4-chlorophenyl)ethanamine (90 mg, 0.58 mmol) as starting
compounds, the title compound was prepared as a white solid according to the procedure
1
described above for 4. Yield: 38 mg (23%); mp 153°C. H NMR (500 MHz, CDCl ) δ 8.57 –
3
8
7
.43 (m, 1H), 8.07 (d, J = 7.6 Hz, 1H), 7.81 (ddd, J = 8.6, 1.7, 0.8 Hz, 1H), 7.49 (ddd, J = 8.2,
.1, 1.1 Hz, 1H), 7.41 (d, J = 8.2 Hz, 1H), 7.35 (d, J = 8.6 Hz, 1H), 7.30 – 7.22 (m, 3H), 7.17 (d,
J = 8.4 Hz, 2H), 4.26 (t, J = 7.3 Hz, 2H), 3.76 – 3.70 (m, 2H), 2.94 (t, J = 7.0 Hz, 2H), 1.84 (p, J
1
3
=
7.4 Hz, 2H), 1.37 – 1.31 (m, 4H), 0.86 (t, J = 7.0 Hz, 3H). C NMR and DEPT (126 MHz,
CDCl ) δ 168.39 (C=O), 168.32 (C=O), 142.37 (C), 141.17 (C), 137.85 (C), 132.46 (C), 130.42
3
(CH), 128.92 (CH), 126.49 (CH), 125.15 (C), 125.12 (C), 124.65 (CH), 122.96 (C), 122.77 (C),
1
4
20.74 (CH), 119.90 (CH), 119.76 (CH), 109.25 (CH), 108.58 (CH), 43.42 (CH ), 41.38 (CH ),
2 2
1.26 (CH ), 35.47 (CH ), 35.46 (CH ), 29.54 (CH ), 28.82 (CH ), 22.63 (CH ), 14.14 (CH ).
2
2
2
2
2
2
3
+
+
ESI: m/z 419.4 (M + H) . HRMS calcd for C H N OCl (M + H) 419.1890, found 419.1870.
2
6
28
2
4
.1.12. Tert-butyl 4-((9-pentyl-9H-carbazole-3-carboxamido)methyl)piperidine-1-carbo-
xylate (12). Carboxylic acid 3 (700 mg, 2.49 mmol), HOBt (404 mg, 2.99 mmol), DIPEA (847
µL, 4.98 mmol), DMAP (365 mg, 2.99 mmol), and EDC (573 mg, 2,99 mmol) were added upon
stirring to DCM (50 mL) under nitrogen. The obtained solution was cooled down on an ice-water
bath. Tert-butyl 4-(aminomethyl)piperidine-1-carboxylate (749 mg, 2.99 mmol) was added in
one portion, and the resulting reaction mixture was then allowed to warm to room temperature
and stirred for 16 h. The solvent was removed in vacuo, and the obtained residue was extracted
with EtOAc (100 mL). The organic layer was washed consecutively with 5% citric acid solution
(50 mL × 3), concentrated sodium bicarbonate (50 mL × 3), and brine (50 mL). The organic
layer was then separated, dried over magnesium sulfate, filtered, and concentrated in vacuo. The
crude product was purified by flash chromatography eluting with heptane/EtOAc (gradient
1
elution) to give 1.096 g (92%) of 14 as a yellow glass. H NMR (400 MHz, CDCl ) δ 8.56 (d, J =
3
1
.1 Hz, 1H), 8.08 (d, J = 7.7 Hz, 1H), 7.90 (dd, J = 8.6, 1.4 Hz, 1H), 7.47 (ddd, J = 8.3, 7.1, 1.2
Hz, 1H), 7.40 (d, J = 8.2 Hz, 1H), 7.36 (d, J = 8.6 Hz, 1H), 7.24 (t, J = 7.5 Hz, 1H), 6.64 (br s,
1
1
H), 4.26 (t, J = 7.2 Hz, 2H), 4.07 (br m, 2H), 3.39 (br s, 2H), 2.68 (t, J = 12.1 Hz, 2H), 1.91 –
.79 (m, 3H), 1.75 (br d, J = 13.2 Hz, 2H), 1.45 (s, 9H), 1.37 – 1.29 (m, J = 8.9, 5.2 Hz, 4H),
2
6

ACCEPTED MANUSCRIPT
13
1
.20 (qd, J = 12.5, 4.3 Hz, 2H), 0.86 (t, J = 6.8 Hz, 3H). C and DEPT NMR (101 MHz, CDCl )
3
δ 168.49 (C=O), 154.76 (C), 142.05 (C), 140.86 (C), 126.12 (CH), 125.05 (C), 124.86 (CH),
1
7
22.76 (C), 122.43 (C), 120.43 (CH), 119.80 (CH), 119.38 (CH), 108.96 (CH), 108.19 (CH),
9.24 (C), 45.45 (CH ), 43.00 (CH ), 36.48 (CH), 29.91 (CH ), 29.20 (CH ), 28.50 (CH ), 28.41
2
2
2
2
2
+
(
CH ), 22.33 (CH ), 13.87 (CH ). ESI: m/z 478.2 (M + H) . HRMS calcd for C H N O (M +
3 2 3 29 40 3 3
+
H) 478.3070, found 478.3007.
4
.1.13. Tert-butyl 4-(9-pentyl-9H-carbazole-3-carbonyl)piperazine-1-carboxylate (13). Using
carboxylic acid 3 (700 mg, 2.49 mmol) and tert-butyl piperazine-1-carboxylate (556 mg, 2.99
mmol) as starting compounds, the title compound was prepared as a light yellow glass according
1
to the procedure described above for compound 14. Yield: 1.03 g (92%). H NMR (400 MHz,
CDCl ) δ 8.20 (d, J = 1.2 Hz, 1H), 8.09 (d, J = 7.7 Hz, 1H), 7.54 (dd, J = 8.4, 1.5 Hz, 1H), 7.48
3
(
(
(
J = 8.3, 7.0, 1.2 Hz, ddd, 1H), 7.41 (t, J = 8.1 Hz, 2H), 7.26 (ddd, J = 7.0, 5.9, 1.1 Hz, 1H), 4.30
t, J = 7.2 Hz, 2H), 3.68 (br s, 4H), 3.50 (br s, 4H), 1.93 – 1.82 (m, 2H), 1.48 (s, 9H), 1.39 – 1.31
1
3
m, 4H), 0.87 (t, J = 7.0 Hz, 3H). C and DEPT NMR (101 MHz, CDCl ) δ 171.98 (C=O),
3
1
54.83 (C), 141.38 (C), 141.09 (C), 126.45 (CH), 125.71 (C), 125.32 (CH), 122.77 (C), 122.69
(C), 120.70 (CH), 120.35 (CH), 119.59 (CH), 109.20 (CH), 108.66 (CH), 80.42 (C), 43.39
(
CH ), 29.53 (CH ), 28.81 (CH ), 28.56 (CH ), 22.62 (CH ), 14.11 (CH ). ESI: m/z 450.1 (M +
2
2
2
3
2
3
+
+
H) . HRMS calcd for C H N O (M + H) 450.2757, found 450.2777.
2
7
36
3
3
4
.1.14. 9-Pentyl-N-(piperidin-4-ylmethyl)-9H-carbazole-3-carboxamide (14). Dry hydrogen
chloride gas was passed through a solution containing 760 mg (1.59 mmol) of 12 dissolved in 30
mL of EtOAc for about ten minutes. The solution was stirred overnight, and then concentrated in
vacuo. The residue was dissolved in EtOAc (100 mL) and washed with 1N NaOH. The organic
layer was washed with brine (30 mL), dried (MgSO ), and filtered. The solvent was evaporated
4
under reduced pressure, and the residue self-crystallized. The white crystalline solid was
collected by filtration, washed with heptanes (20 mL) and dried in vacuo. Yield (as a free base):
1
5
7
27 mg (88%); mp 163°C. H NMR (400 MHz, CDCl ) δ 8.55 (d, J = 1.4 Hz, 1H), 8.09 (d, J =
3
.7 Hz, 1H), 7.90 (dd, J = 8.6, 1.7 Hz, 1H), 7.48 (ddd, J = 8.3, 7.1, 1.2 Hz, 1H), 7.40 (d, J = 8.2
Hz, 1H), 7.36 (d, J = 8.6 Hz, 1H), 7.24 (ddd, J = 8.0, 7.0, 1.0 Hz, 1H), 6.57 (t, J = 5.7 Hz, 1H),
.26 (t, J = 7.2 Hz, 2H), 3.39 (t, J = 6.2 Hz, 2H), 3.09 (d, J = 12.1 Hz, 2H), 2.60 (td, J = 12.2, 2.5
4
2
7

ACCEPTED MANUSCRIPT
Hz, 2H), 1.98 (s, 1H), 1.91 – 1.80 (m, 2H), 1.80 – 1.71 (m, 3H), 1.40 – 1.29 (m, 4H), 1.23 (qd, J
13
=
12.8, 3.7 Hz, 2H), 0.86 (t, J = 6.9 Hz, 3H). C and DEPT NMR (101 MHz, CDCl ) δ 168.51
3
(C=O), 142.32 (C), 141.15 (C), 126.40 (CH), 125.41 (C), 124.79 (CH), 122.99 (C), 122.73 (C),
1
4
20.73 (CH), 119.82 (CH), 119.67 (CH), 109.21 (CH), 108.50 (CH), 46.47 (CH ), 46.20 (CH ),
2 2
3.37 (CH ), 36.91 (CH), 31.37 (CH ), 29.49 (CH ), 28.77 (CH ), 22.59 (CH ), 14.09 (CH ).
2
2
2
2
2
3
+
+
ESI: m/z 378.2 (M + H) . HRMS calcd for C H N O (M + H) 378.2545, found 378.2583.
2
4
32
3
4
1
.1.15. (9-Pentyl-9H-carbazol-3-yl)(piperazin-1-yl)methanone (15). Starting with compound
3 (973 mg, 2.16 mmol), the title compound was prepared as a light yellow glass according to
1
the procedure described above for compound 14. Yield (as a free base): 732 mg (97%). H NMR
(
400 MHz, CDCl ) δ 8.23 (d, J = 1.1 Hz, 1H), 8.13 (d, J = 7.7 Hz, 1H), 7.57 (dd, J = 8.4, 1.5 Hz,
3
1
1
2
1
1
4
H), 7.52 (ddd, J = 8.3, 7.0, 1.2 Hz, 1H), 7.44 (t, J = 8.6 Hz, 2H), 7.29 (ddd, J = 7.9, 7.0, 1.0 Hz,
H), 4.33 (t, J = 7.2 Hz, 2H), 3.72 (br s, 4H), 2.95 (br s, 4H), 2.02 (s, 1H), 1.90 (p, J = 7.3 Hz,
1
3
H), 1.45 – 1.31 (m, 4H), 0.90 (t, J = 7.0 Hz, 3H). C and DEPT NMR (101 MHz, CDCl ) δ
3
71.74 (C=O), 141.20 (C), 141.04 (C), 126.32 (CH), 126.18 (C), 125.31 (CH), 122.80 (C),
22.60 (C), 120.67 (CH), 120.21 (CH), 119.47 (CH), 109.13 (CH), 108.55 (CH), 46.47 (CH2),
+
3.34 (CH ), 29.50 (CH ), 28.78 (CH ), 22.60 (CH ), 14.09 (CH ). ESI: m/z 350.1 (M + H) .
2
2
2
2
3
+
HRMS calcd for C H N O (M + H) 350.2232, found 350.2235.
2
2
28
3
4
.1.16. Pyridin-4-yl-methyl 9-pentyl-9H-carbazole-3-carboxylate (16). To a stirred mixture of
carboxylic acid 3 (200 mg, 0.71 mmol), 4-(bromomethyl)pyridine (147 mg, 0.85 mmol),
triethylamine (178 µL, 1.27 mmol), and sodium carbonate (1 g, 9.43 mmol) in THF (3 mL) in a
2
0 mL microwave vessel was added 1 M TBAF in THF (853 µL, 0.853 mmol). The mixture was
stirred at room temperature for 16 h under a nitrogen atmosphere, and then subjected to
microwave irradiation at 60°C for 1 h. After cooling to room temperature, the reaction mixture
was diluted with DCM (30 mL) and transferred to a round-bottomed flask. The volatiles were
removed in vacuo, and the obtained residue was dissolved in DCM (150 mL). The organic layer
was washed with 0.05 M NaOH (50 mL × 2), dried over magnesium sulfate, filtered, and
concentrated in vacuo. The obtained residue was purified by flash chromatography eluting with
heptane/EtOAc (gradient elution) on a Biotage® KP-NH cartridge yielding compound 12 as a
1
yellowish oil that solidified on standing to a white solid. Yield: 203 mg, (77%); mp 104°C. H
2
8