Carborane-derived local anesthetics are isomer dependent

Article abstract of DOI:10.1002/cmdc.201402369

Clinically there is a need for local anesthetics with a greater specificity of action on target cells and longer duration. We have synthesized a series of local anesthetic derivatives we call boronicaines in which the aromatic phenyl ring of lidocaine was replaced with ortho-, meta-,C,C′-dimethyl meta- and para-carborane clusters. The boronicaine derivatives were tested for their analgesic activity and compared with lidocaine using standard procedures in mice following a plantar injection. The compounds differed in their analgesic activity in the following order: ortho-carborane = C,C′-dimethyl meta-carborane > para-carborane > lidocaine > meta-carborane derivative. Both ortho-boronicaine and C,C′-dimethyl meta-boronicaine had longer durations of analgesia than lidocaine. Differences in analgesic efficacies are rationalized by variations in chemical structure and protein binding characteristics.

Full text of DOI:10.1002/cmdc.201402369

DOI: 10.1002/cmdc.201402369
Communications
Carborane-Derived Local Anesthetics are Isomer
Dependent
George R. Kracke,[a, b] Monika R. VanGordon,[b] Yulia V. Sevryugina,[b] Peter J. Kueffer,[b]
Kuanysh Kabytaev,[b] Satish S. Jalisatgi,[b] and M. Frederick Hawthorne*[b]
Clinically there is a need for local anesthetics with a greater
specificity of action on target cells and longer duration. We
have synthesized a series of local anesthetic derivatives we call
boronicaines in which the aromatic phenyl ring of lidocaine
was replaced with ortho-, meta-, C,C’-dimethyl meta- and para-
carborane clusters. The boronicaine derivatives were tested for
their analgesic activity and compared with lidocaine using
standard procedures in mice following a plantar injection. The
compounds differed in their analgesic activity in the following
order: ortho-carborane = C,C’-dimethyl meta-carborane >
para-carborane > lidocaine > meta-carborane derivative. Both
ortho-boronicaine and C,C’-dimethyl meta-boronicaine had
longer durations of analgesia than lidocaine. Differences in an-
algesic efficacies are rationalized by variations in chemical
structure and protein binding characteristics.
Figure 1. Structures and electron-density distributions mapped with electro-
static potential of lidocaine (1), boronicaines 2–5, and an adamantyl deriva-
*
*
*
*
tive (6). Color code: B: ; CH: ; C: ; BH: .
recognized. For example, lidocaine, which contains a phenyl
ring, possesses anesthetic activity in man, whereas a lidocaine
analogue containing a cyclohexyl group does not.[3] In other
studies, compounds containing a phenyl ring were more than
an order of magnitude more potent at blocking sodium chan-
nels than comparable compounds lacking one.[4] A thiophene
ring can substitute for a phenyl ring in the local anesthetic, ar-
ticaine, which shows comparable anesthetic activity to lido-
caine.[5]
Since the discovery by Freud and Koller in the 1800s that the
natural plant product cocaine has analgesic properties, cocaine
and its synthetic congeners have defined the class of drugs
called local anesthetics. They decrease pain by blocking volt-
age-gated sodium channels in peripheral sensory neurons.
However, they can also affect these channels in other excitable
cells, such as cardiac cells and central nervous system (CNS)
neurons and thus are additionally used as antiarrhythmics, an-
ticonvulsants, and antidepressants. Presently, the goals of local
anesthetic drug discovery are to develop agents with greater
specificity of action on target cells, fewer unwanted side ef-
fects, and longer duration.[1,2]
Herein, we present a series of local anesthetic derivatives we
call boronicaines in which the aromatic phenyl ring of lido-
caine has been replaced with an isomeric, carborane cluster.
Carboranes are icosahedral, aromatic, dicarba-closo-dodecacar-
boranes with the formula C2B10H12. They are hydrophobic, neu-
tral, moderately polar (contingent upon the position of the
carbon atoms), and have a volume of 144 ꢀ3, somewhat larger
than the 102 ꢀ3 volume of a longitudinally rotated phenyl
ring.[6] Carboranes are easily derivatized and have been used as
a scaffold in drug design, such as, aspirin analogues,[7] estrogen
receptor modulators,[8,9] opioid receptor agonists,[10] transthyre-
tin stabilizers,[11] and antidepressants.[12]
Local anesthetics are typically small, amphiphilic molecules
consisting of a hydrophobic, aromatic moiety connected via an
amide or ester linkage to a terminal amine. Lidocaine (1;
Figure 1) belongs to the class of amide local anesthetics. It has
a dimethyl substituted phenyl ring attached to the amide side
chain. The dependence of anesthetic activity on the presence
of a hydrophobic aromatic moiety such as a phenyl ring is well
[a] Dr. G. R. Kracke
The boronicaines were synthesized by replacing the dime-
thylphenyl group of lidocaine (1), with ortho-, meta-, para-, and
C,C’-dimethyl meta- carborane clusters yielding ortho-boroni-
caine (2), meta-boronicaine (3), para-boronicaine (4), and C,C-
dimethyl meta-boronicaine (5), respectively (Figure 1 and Fig-
ure S4 in the Supporting Information). Additionally, we synthe-
sized adamantane lidocaine (6), in which the phenyl group of
lidocaine was replaced with a nonaromatic, hydrophobic, ada-
mantyl group (Figure 1). We found that two of the boroni-
caines have longer lasting analgesic activity than lidocaine. We
Dept. of Anesthesiology & Perioperative Medicine
School of Medicine, University of Missouri
One Hospital Drive, Columbia, MO 65212 (USA)
[b] Dr. G. R. Kracke, Dr. M. R. VanGordon, Dr. Y. V. Sevryugina, Dr. P. J. Kueffer,
Dr. K. Kabytaev, Dr. S. S. Jalisatgi, Dr. M. F. Hawthorne
International Institute of Nano & Molecular Medicine
University of Missouri
1514 Research Park Drive, Columbia, MO 65211-3450 (USA)
E-mail: hawthornem@missouri.edu
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402369.
ChemMedChem 2015, 10, 62 – 67
62
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
also observed that the boronicaine isomers differ in their anal-
gesic efficacy, a feature rationalized by their structural chemis-
try and molecular docking characteristics.
group. The BꢀN bond length of 1.480(2) ꢀ was within the
range of distances reported for other carboranylamides.[13]
Crystallographic details are given in Tables S1 and S2 in the
Supporting Information.
The syntheses of boronicaine isomers 2–5 were accom-
plished using palladium-catalyzed coupling of boron-substitut-
ed mono-iodo-carboranes, 3-iodo-ortho-carborane, 9-iodo-
meta-carborane, 2-iodo-para-carborane, with 2-diethylaminoa-
cetamide to yield 2–5, respectively.[13] Scheme 1 illustrates the
synthesis of 2 from 3-iodo-ortho-carborane.
The analgesic properties of 2–6 were tested and compared
with 1 using the hot plate test as described in the Supporting
Information. We injected anesthetics in to the upper hindlimb
near the sciatic nerve as well as the hindlimb plantar
region.[19,20] Even though injections in the upper hindlimb gave
the desired results, we found that plantar injections were more
consistent (for details, see Supporting Information). Anesthetics
were dissolved at 10 mm in saline at pH 5.2 since 2–6 were not
soluble at 10 mm in a pH 7.4 solution. All carborane analogues
of lidocaine showed analgesic activity (Figure 2). The onset of
analgesia was rapid since the plantar injections deposit the an-
esthetics in close proximity to the sensory nerves in the foot
touching the hot plate. However, the compounds had quite
different analgesic characteristics. Figure 2a shows that both 2
and 5 had analgesic activity of much longer duration than 1,
while 3 had a shorter duration than 1. The kinetics of 4 looked
very similar to 1. As expected, analgesia was seen with neither
the saline control nor 6 (data not shown). Since the boroni-
caines were not soluble at 10 mm in a pH 7.4 buffer solution,
there was the possibility of them precipitating out in the
mouse interstitial fluid which has a pH of 7.4. Therefore, we in-
vestigated a unilamellar liposome formulation of compounds
1–4 at pH 7.4 for the upper hindlimb experiments. The overall
trend remained similar to nonliposomal experiments, but we
found that the kinetics of the analgesic effect were slower
than nonliposomal formulations (for details, see Supporting In-
formation). The transient nature of the analgesic responses to
1–5 might be explained by the complex pharmacokinetics of
local anesthetics.[21] For example, the short analgesic duration
of lidocaine is likely due to its vasodilator properties.[22] Boroni-
caines 2 and 5 might be lesser vasodilators than lidocaine and
thus have longer durations.
Scheme 1. Palladium-catalyzed coupling of boron-substituted, mono-iodo
carborane with 2-diethylaminoacetamide. Reagents and conditions: a) K3PO4,
*
DavePhos ligand, Pd2(dba)3, toluene, reflux, 3 h, 45%. Color code: CH: ; B:
*
; BH:
*
.
The positions of the positive charges on the tertiary amine
are indicated in Figure 1 by the dark blue regions in the elec-
tron density distributions of 1 and 2. The positive charge on
the tertiary amine is an important factor in the access and
binding of the local anesthetic to the target in the inner pore
of the sodium channel.[14] The natural bond orbital (NBO)
charge distributions of the protonated tertiary amine groups
of compounds 2–5 were not affected by the replacement of
the phenyl ring with the carborane cage (Table S3 in the Sup-
porting Information).
The molecular basis of the interaction of local anesthetics
with sodium channels involves interactions between the drugs
and specific aromatic amino acids lining the channel pore. Two
types of mechanisms have been proposed. One mechanism,
suggested by amino acid substitution experiments, involves
the protonated amine of the local anesthetic molecule forming
cation–p interactions with the electron-rich p face of the aro-
matic side chain of a phenylalanine residue in the sodium
channel pore.[15,16] The presence of carborane clusters should
not preclude cation–p interactions between the boronicaines
and their receptor sites. Another type of mechanism was sug-
gested in studies of voltage-gated sodium channels and model
channel peptides. This involved p–p stacking between local
anesthetic phenyl groups and amino acid phenyl groups in the
channel pore.[17,18] However, our experiments with boronicaines
suggest that p–p stacking is not a significant driving force for
sodium channel block since the replacement of the phenyl
ring of 1 with a three-dimensional, highly symmetric carborane
cage, incapable of comparable p interactions, resulted in mole-
cules with analgesic activity.
The areas under the curve (AUC) values, which are measures
of analgesic efficacy, are shown in Figure 2b. The compounds
differed in their analgesic efficacy in the following order: 2=
5>4>1>3>6. Boronicaine derivatives 2 and 5 had signifi-
cantly greater analgesic efficacy than 1 and did not significant-
ly differ from each other. In the literature, differences in mea-
sured activities have been observed with other isomeric car-
borane-based compounds. For example, carborane-derived bis-
phenols had an estrogenic activity ranking of para>ortho>
meta.[23] In another example, ortho- but not meta-carborane de-
rivatives of indomethacin-inhibited cyclooxygenase.[24]
Studies of the binding of local anesthetics to mammalian
voltage-gated sodium channels have been hampered by the
absence of channel crystal structures. However, crystallization
studies of human serum albumin (HAS) complexed with lido-
caine (PDB code 3JQZ[25]) have been useful in understanding
local anesthetic–protein binding geometry since albumin and
the channel binding sites both have similar cation–p and hy-
drogen-bonding interactions with lidocaine.[25]
The crystal structure of 3 (CCDC 964297) along with the
atom numbering is shown in Figure S1 in the Supporting Infor-
mation. The carborane framework displayed no distortion
upon functionalization with the 2-diethylaminoacetamide
In order to investigate the differences in binding of the bor-
onicaine anesthetics, we docked lidocaine and compounds 2–
ChemMedChem 2015, 10, 62 – 67
www.chemmedchem.org
63
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
are shown in Figure S8 in the
Supporting Information. The
best docking score among car-
borane derivatives was obtained
for
2 with the TS value of
ꢀ3.3568 with a RMSD value of
1.8859 ꢀ, followed by 4 (TS=
ꢀ2.1059, RMSD=1.5153 ꢀ),
(TS= +0.3402, RMSD=0.7050 ꢀ)
and finally (TS= +3.3496,
5
3
RMSD=1.5744 ꢀ). The TS value
for the adamantyl derivative (6)
was ꢀ3.6990 (RMSD=0.7050 ꢀ).
The ranking for boronicaines
based on TS has a direct correla-
tion with the analgesic respons-
es we observed experimentally
in mice. The unfavorably high TS
value for compound 3 is due to
the nearly three times higher
clash score compared with the
corresponding value for 1.
Figure 3 shows lidocaine and
boronicaines 2–5 aligned inside
the active site of HSA. The fol-
lowing hydrogen bonds were
observed for all analyzed com-
pounds: Asp187ꢀCOOꢀ···HN+ꢀ
amine and Lys190ꢀHN+N···O=Cꢀ
NHꢀR (R=phenyl for 1, carbor-
ane for 2–5, adamantyl for 6).
The amide–amine sections of
the compounds resided in iden-
tical regions of the cavity, but
the spatial displacement of hy-
drophobic moiety and the posi-
tions of the mildly acidic CꢀH
groups[27] in the carborane cage
differed among the isomers.
Their position relative to the
boron-substituted amide linkage
would determine their exposure
to the amino acids on the sur-
rounding protein. As shown in
Figure 1, two or one of the CꢀH
protons in 2 and 4, respectively,
are shielded by the B-amido sub-
stituent, while both CꢀH protons
in 3 are positioned away from
the amide substituent and ex-
Figure 2. Time courses of the analgesic effects of saline-dissolved drugs 1–5 and saline control (ctrl). Analgesia
was measured in mice as paw withdrawal latency in seconds in the hot plate test after a single hindlimb plantar
injection of 50 mL of a 10 mm solution, equivalent to 0.5 mmols of drug. A) Time courses of analgesia are shown
from 5 min before to 45 min after injection. Data represent the mean ꢁSEM for 10–13 animals. b) Areas under
the curve (AUC) were calculated from the experiments summarized in panel a). Data are the mean, and error bars
represent ꢁSEM. An asterisk denotes a significant difference (p <0.05, ANOVA) between 2 and 1, and between 5
and 1. There is no significant difference between the results of compounds 2 and 5.
6 in the active site of the HSA using molecular docking soft-
ware FlexX[26] (for details, see Supporting Information). The top
ranking docking pose is the one with the lowest root mean
square deviation (RMSD) of coordinates against the reference
structure. The corresponding values of the FlexX total docking
scores (TS), which include contributions from the protein–
ligand interactions, clash penalty, and ligand rotation entropy
posed to the surrounding protein for hydrogen-bonding inter-
actions. The analgesic effects shown in Figure 2 led us to be-
lieve that hydrogen-bonding interactions of CꢀH protons with
the amino acid residues of the surrounding protein play an im-
portant role in the efficacy of these derivatives. To further
study the role of CꢀH protons of carborane derivatives in the
analgesic activity, we synthesized a C,C’-dimethyl meta-carbor-
ChemMedChem 2015, 10, 62 – 67
www.chemmedchem.org
64
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
Figure 3. Results of FlexX docking in HSA. The view of the human serum albumin (HSA) active site (PDB code 3JQZ[25]) with compounds 1–5 (balls and sticks).
Surface charge color code: negative, positive.
ane derivative (5), where hydrogens on the carbons were re-
placed with methyl groups. This modification resulted in a satis-
factory docking in the HAS crystal structure (Figure 3) and im-
proved analgesic activity over meta-boronicaine 3 (Figure 2).
The docking of 6 in HSA was successful (for details, see Sup-
porting Information); however, the experimental studies show
that the replacement of the lidocaine aromatic phenyl with
adamantyl ceases the analgesic activity. Thus we hypothesize
that there are three factors that contribute to the analgesic ac-
tivity of our carborane derivatives: 1) the presence of delocal-
ized electron density in the hydrophobic part of the molecule;
2) exposure of hydrophobic BꢀH vertices to the protein; 3) the
presence or absence of acidic CꢀH groups in the vicinity of the
substituent.
The relative potencies of 1 and 2 were established by meas-
uring the analgesic responses of animals injected with varying
*
*
Figure 4. Analgesia log concentration–response curves for 1 ( ) and 2 ( ).
Single hindlimb plantar injections were made with 50 mL of drug at the des-
ignated concentration. The equivalent molar amounts of compound injected
were 0.05, 0.15, 0.5, and 1.0 mmols, respectively. Data represent the mean
ꢁSEM for 6–13 animals. An asterisk denotes statistically different (p <0.05
ANOVA) mean values between the compounds.
concentrations of the compound. Data were fitted to the Hill
equation with a four-parameter (top, bottom, EC50 and Hill
slope), nonlinear regression curve. Figure 4 shows that the po-
tencies of 1 and 2, calculated from experiments similar to
those in Figure 2, were not significantly different. The EC50
value for 1 was 6.3 mm (range: 4.4–8.9 mm, with 95% confi-
dence interval) and for 2 was 5.3 mm (range: 3.0–9.5 mm, with
95% confidence interval). Hill coefficients (meanꢁ SEM) were
1.6ꢁ0.4 for 1 and 4.5ꢁ2.2 for 2. The analgesic efficacy of 2
was significantly greater than 1 at both 10 and 20 mm.
In summary, to the best of our knowledge, we have synthe-
sized the first carborane-based local anesthetics and studied
their structure–activity relationships. Analgesic studies in mice
showed that two of the carborane isomers, ortho- (2) and C,C’-
dimethyl meta- (5), had longer durations and greater efficacies
of analgesic effects than the parent lidocaine (1). The adaman-
tyl derivative (6) had minimal analgesic activity. The potencies
of 1 and 2 were not significantly different from each other, but
2 was more efficacious than 1. The analgesic activity differed
among the compounds in the following order: 2=5>4>1>
3>6. The variation in boronicaine analgesic activity can par-
tially be explained by molecular docking studies, which show
differences in the exposure of acidic CꢀH protons of the car-
borane cage to the surrounding proteins with more exposure
correlating with less analgesic activity.
Experimental Section
General: All coupling reactions were carried out under argon
using Schlenk techniques. Toluene was freshly distilled from CaH2
prior to use. Tetrahydrofuran (THF) was dried by passage through
a solvent purification system to give the solvent with a water con-
tent of less than 25 ppm as determined by a coulometric KF titra-
tor (Mettler, Toledo, USA). All other solvents were used as pur-
chased from commercial sources. 2-Dicyclohexylphosphino-2-(N,N-
dimethylamino) biphenyl (DavePhos), Et2NH, 2-bromoacetamide
and Pd2(dba)3 were used as purchased (Sigma–Aldrich, St. Louis,
MO, USA). ortho-, meta- and para-carboranes were obtained from
Katchem Ltd (Prague, Czech Republic). Starting materials 2-iodo-
para-carborane, 9-iodo-meta-carborane,[28] 3-iodo-ortho-carbor-
ane,[29] 1,7-dimethyl-meta-carborane,[30] and diethylaminoacetic acid
hydrochloride[31] were prepared using literature procedures. All
iodo-carboranes were azeotropically dried with benzene to remove
any trace of water. K3PO4 was ground to a fine powder and then
dried thoroughly by heating in vacuo at 1058C. All solids were
stored in an argon-filled glove box. Silica gel was used as pur-
chased (63–200 mm, Sorbent Technologies, Inc.). Analytical thin-
ChemMedChem 2015, 10, 62 – 67
www.chemmedchem.org
65
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
layer chromatography (TLC) for carborane identification was per-
formed by using pre-coated silica gel XHL plates (Sorbent Technol-
ogies, Inc.) and visualized by dipping into an acidified (HCl) solu-
tion of PdCl2 followed by heating.
after which time it was filtered, and the filtrate was evaporated to
give a yellow crude solid. The crude product was purified by
column chromatography using silica gel ( MeOH/Et2NH/CH2Cl2,
0:0:100!5.7:0.3:94) to give a yellowish solid. Washing the solid
with Et2O gave the desired product as a white crystalline powder
1H, 13C and 11B NMR spectrometry was performed on Bruker Avance
400 and Avance 500 spectrometers. 11B NMR spectra were external-
ly referenced to BF3·Et2O; the peaks upfield of the reference were
designated as negative. 1H and 13C NMR chemical shifts (d) are
given in parts per million (ppm) relative to residual solvent signals
(CDCl3: d=7.24 ppm for 1H NMR and d=77.0 ppm for 13C NMR).
High-resolution mass spectra (HRMS) were obtained on an ABI
QSTAR and Mariner Biospectrometry Workstation (PerSeptive Bio-
systems). Melting points (mp) were obtained using an automated
melting point system OptiMelt (Stanford Research Systems). Meas-
urements were conducted in open-end Kimble Chase capillary
tubes from borosilicate glass (L=90 mm). The melting point range
is defined as the interval between the onset and clear points using
a 0.58Cminꢀ1 ramp rate, and values are uncorrected.
1
(432 mg, 86%): mp: 127.9–129.28C; H NMR (400 MHz, [D6]DMSO):
d=7.05 (s, 1H, CONH), 3.82 (s, 2H, CcarboraneꢀH), 2.80 (s, 2H, CH2),
2.42 (q, J=7.1 Hz, 4H, NEt2), 0.88 (t, J=7.2 Hz, 6H, NEt2), 3.2–
1.1 ppm (m, 9H, BH); 13C NMR (100.6 MHz, [D6]DMSO): d=173.6
(C=O), 58.0, 52.9 (CH2 +Ccarborane), 48.1, 12.2 ppm (NEt2); 11B NMR
(128.4 MHz, [D6]DMSO): d=ꢀ0.5 (s, 1B), ꢀ7.7 (d, J=139 Hz, 2B),
ꢀ11.6 (d, J=146 Hz, 1B), ꢀ14.2 (d, J=184 Hz, 2B), ꢀ15.8 (d, J=
167 Hz, 2B), ꢀ18.8 (d, J=173 Hz, 1B), ꢀ22.2 ppm (d, J=148 Hz,
1B); HRMS (ESI): m/z [M+H]+ calcd for C8H24N2OB10: 273.2970,
found: 273.2508.
2-(Diethylamino)-N-(1,12-dicarba-closo-dodecaboran-2-yl)aceta-
mide (4): To a dry 25 mL Schlenk flask in an argon-filled glove box
were added 2-iodo-p-carborane (1.0 g, 3.69 mmol), 2-diethylami-
noacetamide (630 mg, 4.85 mmol), K3PO4 (3.91 g, 18.45 mmol), Da-
vePhos ligand (71 mg, 0.18 mmol), Pd2(dba)3 (169 mg, 0.18 mmol).
To this mixture, dry THF (20 mL) was added by syringe at the
Schlenk line. The reaction mixture was heated at reflux for 2 days,
after which time it was filtered, and the red colored filtrate was
evaporated to give a reddish-brown residue. The crude product
was purified by column chromatography over silica gel gel (100%
CH2Cl2) to give the desired compound as a yellowish solid (940 mg,
93%). The product was further purified by recrystallization from
2-Diethylaminoacetamide: A Schlenk tube containing 2-bromoa-
cetamide (2.6 g, 18.9 mmol), Et2NH (1.38 g, 18.9 mmol), K2CO3
(2.8 g, 20.3 mmol), and KI (3.14 g, 18.9 mmol) was evacuated and
then refilled with argon. Dry CH3CN (50 mL) was added to the
tube. The mixture was stirred at reflux for 4 h. The solvent was
evaporated, the solid residue was redissolved with EtOAc (70 mL),
and the solution was filtered. The filtrate was washed with water,
dried with MgSO4, and then the solvent was evaporated. A white
solid product was obtained without any further purification steps
1
MeOH/ Et2O. H NMR (400 MHz, CDCl3): d=7.54 (br s, 1H, CONH),
1
(73%): H NMR (500 MHz, CDCl3): d=6.19 (br s, 2H, NH2), 3.01 (s,
3.71 (s, 1H, CcarboraneꢀH), 2.93 (s, 2H, CH2), 2.68 (s, 1H, CcarboraneꢀH),
2.55 (q, J=7.2 Hz, 4H, NEt2), 1.03 (t, J=7.2 Hz, 6H, NEt2), 3.1–
1.1 ppm (m, 9H, BH); 13C NMR (125.8 MHz, CDCl3): d=175.9 (C=O),
64.7, 60.5 (Ccarborane), 58.4 (CH2), 49.0, 12.5 ppm (NEt2); 11B NMR
(128.4 MHz, CDCl3): d=ꢀ5.1 (s, 1B), ꢀ13.9 (d, J=190 Hz, 2B), ꢀ15.5
(d, J=161 Hz, 4B), ꢀ16.5 (d, J=139 Hz, 2B), ꢀ19.9 ppm (d, J=
162 Hz, 1B); HRMS (ESI): m/z [M+H]+ calcd for C8H24N2OB10:
273.2970, found: 273.2803.
2H, CH2), 2.56 (q, J=7.2 Hz, 4H, NEt2), 1.04 ppm (6H, t, J=7.3 Hz,
NEt2); 13C NMR (125.8 MHz, CDCl3): d=175.6 (C=O), 57.4 (CH2),
48.6, 12.3 ppm (NEt2); HRMS (ESI): m/z [M+H]+ calcd for C6H15N2O:
131.1184, found: 131.0907.
2-(Diethylamino)-N-(1,2-dicarba-closo-dodecaboran-3-yl)aceta-
mide (2): The following were added to a dry 25 mL Schlenk flask in
an argon-filled glove box: 3-iodo-o-carborane (300 mg, 1.10 mmol),
2-diethylaminoacetamide (432 mg, 3.32 mmol), K3PO4 (1.13 g,
5.35 mmol), DavePhos ligand (42 mg, 0.1 mmol), Pd2(dba)3 (46 mg,
0.05 mmol) and a Teflon-coated magnetic stir bar. To this mixture,
dry toluene (4 mL) was added by syringe at the Schlenk line. The
reaction mixture was heated at reflux for 3 h, after which time it
was filtered. The filtrate was evaporated to leave a yellow oily resi-
due, which was purified by column chromatography using silica
gel (MeOH/Et2NH/CH2Cl2, 0:0:100! 1.43:0.075:98.5). Further purifi-
cation was carried out by treating a solution of the product with
active charcoal to give the desired compound as a white crystalline
solid (130 mg, 45%): 1H NMR (500 MHz, CDCl3): d=8.04 (s, 1H,
CONH), 4.52 (s, 2H, CcarboraneꢀH), 3.04 (s, 2H, CH2), 2.66 (q, J=6.7 Hz,
4H, NEt2), 1.12 (t, J=7.5 Hz, 6H, NEt2), 3.0–1.3 ppm (m, 9H, BH);
13C NMR (125.8 MHz, CDCl3): d=176.7 (C=O), 57.7, 54.6 (CH2 +
Ccarborane), 48.9, 12.2 ppm (NEt2); 11B NMR (160.5 MHz, CDCl3): d=
ꢀ4.8 (d, J=148 Hz, 2B), ꢀ6.7 (s, 1B), ꢀ11.1 (d, J=151 Hz, 1B),
ꢀ13.0 (d, J=167 Hz, 2B), ꢀ15.1 ppm (d, J=141 Hz, 4B); HRMS
(ESI): m/z [M+H]
1,7-Dimethyl-9-iodo-1,7-dicarba-closo-dodecaborane: To a solu-
tion of 1,7-dimethyl-1,7-dicarba-closo-dodecaborane (387 mg,
2.25 mmol) in CH2Cl2 (50 mL) were added I2 (600 mg, 2.36 mmol)
and AlCl3 (30 mg, 0.22 mmol). The reaction mixture was heated at
reflux for 4 h. The solvent was removed in vacuo, and the crude
product was purified by column chromatography using silica gel
(CH2Cl2/hexane, 1:10) to give the desired compound as a white
solid (470 mg, 70%): 1H NMR (400 MHz, CDCl3): d=3.8–1.9 (br m,
9H, BH), 1.71 ppm (s, 6H, CH3); 13C NMR (100.6 MHz, CDCl3): d=
72.5 (CH), 24.2 ppm (CH3); 11B NMR (128.4 MHz, CDCl3): d=ꢀ6.3 (d,
J=167 Hz, 2B), ꢀ8.5 (d, J=155 Hz, 1B), ꢀ9.0 (d, J=159 Hz, 2B),
ꢀ10.0 (d, J=165 Hz, 2B), ꢀ12.2 (d, J=181 Hz, 1B), ꢀ14.0 (d, J=
182 Hz, 1B), ꢀ23.5 ppm (s, BꢀI); HRMS (ESI): m/z [M]ꢀ calcd for
C4H15B10I: 298.1221, found: 298.1017.
2-(Diethylamino)-N-(1,7-dimethyl-1,7-dicarba-closo-dodecabor-
an-9-yl)acetamide (5): In a 10 mL round bottom flask, 2-diethyla-
minoacetamide (83 mg, 0.64 mmol) was added to a suspension of
NaH (60% dispersion in mineral oil, 26 mg, 0.64 mmol) in dioxane
(2 mL). The mixture was stirred at 1008C for 15 min, and
then 1,7-dimethyl-9-iodo-1,7-dicarba-closo-dodecaborane (94 mg,
0.32 mmol), Pd2(dba)3·CHCl3 (16 mg, 0.015 mmol), and 2,2’-bis(di-
phenylphosphino)-1,1’-binaphthyl (BINAP) (20 mg, 0.03 mmol) were
added to the reaction mixture. Stirring was continued at 1008C for
another 72 h. The solvent was removed in vacuo, and the crude
product was purified by silica gel chromatography (EtOAc/Hexane,
0:100 arrow 30:70) to give compound 5, an oil (34 mg, 35%):
calcd for C8H25N2OB10: 273.2970, found: 273.2799.
2-(Diethylamino)-N-(1,7-dicarba-closo-dodecaboran-9-yl)aceta-
mide (3): To a dry 25 mL Schlenk flask in an argon-filled glove box
were added 9-iodo-m-carborane (500 mg, 1.85 mmol), 2-diethyla-
minoacetamide (722 mg, 5.54 mmol), K3PO4 (1.96 g, 9.23 mmol),
DavePhos ligand (44 mg, 0.11 mmol), Pd2(dba)3 (42 mg, 0.05 mmol).
To this mixture, dry THF (20 mL) was added by syringe at the
Schlenk line. The reaction mixture was heated at reflux for 3 days,
ChemMedChem 2015, 10, 62 – 67
www.chemmedchem.org
66
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
1H NMR (400 MHz, CDCl3): d=2.93 (s, 2H, CH2), 2.54–2.49 (q, J=
7.3 Hz, 4H, NEt2), 1.67 (s, 6H, CH3), 0.99 (t, J=7.3 Hz, 6H, NEt2), 3.5–
1.5 ppm (m, 9H, BH); 13C NMR (100.6 MHz, CDCl3): d=175.2 (C=O),
67.8 (CH), 58.5 (CH2), 48.8 (CH2), 24.3 (CH3), 12.4 ppm (CH3);
11B NMR (128.4 MHz, CDCl3): d=ꢀ0.5 (s, 1B, BꢀN), ꢀ8.0 (d, J=
162 Hz, 2B), ꢀ10.9 (d, J=178 Hz, 3B), ꢀ12.4 (d, J=187 Hz, 2B),
ꢀ14.1 (d, J=186 Hz, 1B), ꢀ17.2 ppm (d, J=178 Hz, 1B); HRMS
(ESI): m/z [M+Na]+ calcd for C10H28B10N2ONa: 323.3103, found:
323.2695.
[3] N. Lçfgren, Studies on Local Anesthetics: Xylocaine, a new synthetic drug,
Ivar Haeggstroms, Stockholm, 1948, pp. 1–147.
[4] G. W. Zamponi, R. J. French, Biophys. J. 1994, 67, 1015–1027.
[5] M. Snoeck, Local Reg. Anesth. 2012, 5, 23–33.
[6] M. Scholz, E. Hey-Hawkins, Chem. Rev. 2011, 111, 7035–7062.
´
[7] M. Scholz, G. N. Kaluderovic, H. Kommera, R. Paschke, J. Will, W. S. Shel-
¯
drick, E. Hey-Hawkins, Eur. J. Med. Chem. 2011, 46, 1131–1139.
[8] Y. Endo, T. Iijima, Y. Yamakoshi, H. Fukasawa, C. Miyaura, M. Inada, A.
Kubo, A. Itai, Chem. Biol. 2001, 8, 341–355.
[9] M. L. Beer, J. Lemon, J. F. Valliant, J. Med. Chem. 2010, 53, 8012–8020.
[10] A. Eberle, O. Leukart, P. Schiller, J. L. Fauchꢂre, R. Schwyzer, FEBS Lett.
1977, 82, 325–328.
The general procedure for preparing hydrochloride salts of com-
pounds 2–5: The compound was dissolved in anhydrous Et2O (3–
6 mL), and anhydrous HCl gas was bubbled through the stirred so-
lution for 20–40 min at a medium rate. The product precipitated as
white crystals, which were isolated by filtration and then recrystal-
lized from MeOH/Et2O to give the hydrochloride salt as shiny white
crystals. The yield range was 50–70%.
[11] R. L. Julius, O. K. Farha, J. Chiang, L. J. Perry, M. F. Hawthorne, Proc. Natl.
Acad. Sci. USA 2007, 104, 4808–4813.
[12] S. M. Wilkinson, H. Gunosewoyo, M. L. Barron, A. Boucher, M. McDon-
nell, P. Turner, D. E. Morrison, M. R. Bennett, I. S. McGregor, L. M. Rendi-
na, M. Kassiou, ACS Chem. Neurosci. 2014, 5, 335–339.
[13] Y. Sevryugina, R. L. Julius, M. F. Hawthorne, Inorg. Chem. 2010, 49,
10627–10634.
2-(Diethylamino)-N-tricyclo[3.3.1.13,7]dec-1-ylacetamide (6): To
[14] B. Hille, J. Gen. Physiol. 1977, 69, 497–515.
[15] D. S. Ragsdale, J. C. McPhee, T. Scheuer, W. A. Catterall, Science 1994,
265, 1724–1728.
[16] C. A. Ahern, A. L. Eastwood, D. A. Dougherty, R. Horn, Circ. Res. 2008,
102, 86–94.
[17] H. L. Li, A. Galue, L. Meadows, D. S. Ragsdale, Mol. Pharmacol. 1999, 55,
134–141.
[18] Y. Kuroda, K. Miyamoto, K. Tanaka, Y. Maeda, J. Ishikawa, R. Hinata, A.
Otaka, N. Fujii, T. Nakagawa, Chem. Pharm. Bull. 2000, 48, 1293–1298.
[19] A. M. Binshtok, P. Gerner, S. B. Oh, M. Puopolo, S. Suzuki, D. P. Roberson,
T. Herbert, C. F. Wang, D. Kim, G. Chung, A. A. Mitani, G. K. Wang, B. P.
Bean, C. J. Woolf, Anesthesiology 2009, 111, 127–137.
[20] K. Mitchell, E. E. Lebovitz, J. M. Keller, A. J. Mannes, M. I. Nemenov, M. J.
Iadarola, Pain 2014, 155, 733–745.
[21] S. Leeson, G. Strichartz, Anesth. Analg. 2013, 116, 694–702.
[22] R. A. Johns, Anesthesiology 1989, 70, 805–811.
[23] T. Ogawa, K. Ohta, T. Iijima, T. Suzuki, S. Ohta, Y. Endo, Bioorg. Med.
Chem. 2009, 17, 1109–1117.
a
solution of diethylaminoacetic acid hydrochloride (168 mg,
1 mmol) and 1-aminoadamantane (150 mg, 1 mmol) in CH2Cl2
(10 mL) were added at RT 1-ethyl-3-(3-dimethylaminopropyl)carbo-
diimide hydrochloride (EDCI) (233 mg, 1.5 mmol), hydroxybenzo-
triazole (HOBT) (150 mg, 1.1 mmol) and N,N’-diisopropylethylamine
(DIPEA) (390 mg, 3 mmol). The reaction mixture was stirred at RT
for 24 h. The solvent was removed in vacuo, and the crude product
was purified by silica gel column chromatography (EtOAc, 100%)
to give compound 6 as a colorless oil (83 mg, 29%): 1H NMR
(400 MHz, CDCl3): d=7.17 (br s, 1H, NH), 2.86 (s, 2H, CH2), 2.53–
2.47 (q, J=7.0 Hz, 4H, NEt2), 2.07–2.02 (m, 3H, CH), 1.99–1.95 (m,
6H, CH2), 1.69–1.62 (m, 6H, CH2), 0.99 ppm (t, J=7.0 Hz, 6H, NEt2);
13C NMR (125.8 MHz, CDCl3): d=171.0 (C=O), 58.5 (CH2), 51.0 (Cꢀ
N), 48.8 (CH2), 41.7 (CH2), 36.4 (CH2), 29.4 (CH), 12.6 ppm (CH3);
HRMS (ESI): m/z [M+Na]+ calcd for C16H28N2ONa: 287.2099, found:
287.2122.
[24] M. Scholz, A. L. Blobaum, L. J. Marnett, E. Hey-Hawkins, Bioorg. Med.
Chem. 2011, 19, 3242–3248.
[25] K. L. Hein, U. Kragh-Hansen, J. P. Mortha, M. D. Jeppesen, D. Otzen, J. V.
Møller, P. Nissen, J. Struct. Biol. 2010, 171, 353–360.
[26] M. Rarey, B. Kramer, T. Lengauer, G. Klebe, J. Mol. Biol. 1996, 261, 470–
489.
[27] K. Hermansson, M. Wojcik, S. Sjoberg, Inorg. Chem. 1999, 38, 6039–
6048.
[28] W. Jiang, C. B. Knobler, C. E. Curtis, M. D. Mortimer, M. F. Hawthorne,
Inorg. Chem. 1995, 34, 3491–3498.
Acknowledgements
The authors thank Zachary H. Houston and Chimwemwe Manda
(laboratory assistance), and Seth Schuster (animal experiments).
Financial support to G.R.K. from the University of Missouri Intel-
lectual Property Fast Track Funding Program is acknowledged.
[29] J. Li, C. F. Logan, M. Jones, Jr., Inorg. Chem. 1991, 30, 4866–4868.
[30] M. F. Hawthorne, D. C. Young, P. M. Garrett, D. A. Owen, S. G. Schwerin,
F. N. Tebbe, P. A. Wegner, J. Am. Chem. Soc. 1968, 90, 862–868.
[31] D. A. Peak, T. T. Watkins, J. Chem. Soc. 1950, 445–452.
Keywords: carboranes · lidocaine · local anesthetic agents ·
pain · sensory blockade
[1] B. T. Priest, G. J. Kaczorowski, Proc. Natl. Acad. Sci. USA 2007, 104, 8205–
8206.
Received: August 29, 2014
[2] J. C. Rowlingson, Anesth. Analg. 2013, 117, 1045–1047.
Published online on November 24, 2014
ChemMedChem 2015, 10, 62 – 67
www.chemmedchem.org
67
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim