3-[2-(Aminomethyl)-5-[(pyridin-4-yl)carbamoyl]phenyl] benzoates as soft ROCK inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.09.040

Clinical development of ROCK inhibitors has so far been limited by systemic or local ROCK-associated side effects. A soft drug approach, which involves predictable metabolic inactivation of an active compound to a nontoxic metabolite, could represent an a

Full text of DOI:10.1016/j.bmcl.2013.09.040

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6442–6446
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
3-[2-(Aminomethyl)-5-[(pyridin-4-yl)carbamoyl]phenyl] benzoates
as soft ROCK inhibitors
a
a
a
a
a
Sandro Boland a, , Olivier Defert , Jo Alen , Arnaud Bourin , Karolien Castermans , Nele Kindt ,
⇑
Nicki Boumans a, Laura Panitti a, Sarah Van de Velde b, Ingeborg Stalmans b, Dirk Leysen a
a Amakem N.V. Agoralaan A bis, Diepenbeek 3590, Belgium
b Laboratory of Ophthalmology, KU Leuven, Leuven 3000, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Clinical development of ROCK inhibitors has so far been limited by systemic or local ROCK-associated side
effects. A soft drug approach, which involves predictable metabolic inactivation of an active compound to
a nontoxic metabolite, could represent an attractive way to obtain ROCK inhibitors with improved
tolerability. We herein report the design and synthesis of a new series of soft ROCK inhibitors structurally
related to the ROCK inhibitor Y-27632. These inhibitors contain carboxylic ester moieties which allow
inactivation by esterases. While the parent esters display strong activity in enzymatic (ROCK2) and
cellular (MLC phosphorylation) assays, their corresponding carboxylic acid metabolites have negligible
functional activity. Compound 32 combined strong efficacy (ROCK2 IC50 = 2.5 nM) with rapid inactivation
in plasma (t1/2 <50). Compound 32 also demonstrated in vivo efficacy when evaluated as an IOP-lowering
agent in ocular normotensive New-Zealand White rabbits, without ocular side effects.
Received 12 August 2013
Revised 12 September 2013
Accepted 13 September 2013
Available online 21 September 2013
Keywords:
Kinase inhibitor
ROCK
Soft drug
SAR
Ó 2013 Elsevier Ltd. All rights reserved.
Esterase
Rho-associated, coiled coil containing protein kinase (ROCK)1 is
a serine/threonine kinase belonging to the AGC family.1 In humans,
ROCK consists of two isoforms, namely ROCK1 and ROCK2. These
isoforms share an overall homology of 65%, while being 95%
homologous in the kinase domain. ROCK plays an important role
in numerous cellular processes including smooth muscle cell con-
traction, cell proliferation, adhesion and migration.2 Consequently,
ROCK inhibitors are of potential interest for the treatment of multi-
ple indications including cardiovascular diseases,3 inflammatory
and autoimmune diseases4–6 (e.g., asthma, Chronic Obstructive
Pulmonary Disease, Inflammatory Bowel Disease), or eye diseases7
(e.g., glaucoma). The ROCK inhibitor fasudil has been approved in
Japan for the treatment of cerebral vasospasm. However, ROCK
inhibitors cause a pronounced decrease in blood pressure,7 thereby
limiting their usefulness for the systemic treatment of non-cardio-
vascular indications.7 As a consequence, recent development of
ROCK inhibitors has typically been restricted to topical applica-
tions, which results in lower systemic exposure. Several ROCK
inhibitors have been clinically tested as intraocular pressure
(IOP)-lowering agents for topical treatment of glaucoma.8 Those
include Y-39983 (Novartis, Basel, Switzerland), K-115 (Kowa Com-
pany Ltd, Nagoya, Japan), AR-12286 & AR-13324 (Aerie Pharma-
ceuticals, Inc., Bedminster, USA) and ATS907 (Altheos, Inc., South
San Francisco, USA). Even when used topically, ROCK inhibitors
are not devoid of side effects. In the context of ophthalmology, a
common adverse event resulting from topical administration of
ROCK inhibitors is mild to severe conjunctival hyperemia (redness)
due to mechanism-based smooth muscle cell relaxation in con-
junctival blood vessels.8–10 As a consequence, most available ROCK
inhibitors have a narrow therapeutic window.
In view of the potential of ROCK inhibitors for generating side
effects, a soft drug approach could represent an attractive way to-
wards novel ROCK inhibitors with improved tolerability. Soft
drugs, sometimes known as antedrugs, are biologically active com-
pounds that are designed to undergo metabolic inactivation by
controlled conversion of the parent molecule into a predictable,
nontoxic metabolite.11,12 Herein, we report the design and initial
evaluation of soft ROCK inhibitors.
Compound 113 (Fig. 1) was taken as a starting point for the de-
sign of soft ROCK inhibitors. 1 is structurally related to Y-27632, a
well-described ROCK inhibitor known for its good selectivity to-
wards ROCK.14 The crystal structure of a ROCK/Y-27632, revealed
the importance of the pyridine nitrogen and primary amine for
activity.15 Consequently, those pharmacophore points were not
altered. Modification of the 4-fluorophenyl moiety was instead
preferred. Introduction of ester side chains provided an inactiva-
tion mechanism for the resulting derivatives, by making them
potential substrates for esterases.
A number of ROCK inhibitors structurally derived from 1 were
synthesized. They were prepared in ca. 5–10 steps, starting from
⇑
Corresponding author.
E-mail address: sandro.boland@amakem.com (S. Boland).
the commercially available 3-bromo-4-methylbenzoic acid
2
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.09.040

S. Boland et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6442–6446
6443
Table 1
ROCK2 inhibition data for compounds 1, 10 and 11
Figure 1. Structure of compounds Y-27632 and 1.
R1
R2
ROCK2 IC50 (nM)
a
Compd
1
10
11
Me
H
H
4-F
4-CO2Me
3-CO2Me
1.2
102
2.9
(Fig. 2). Fisher esterification of 2, followed by radical bromination
of the benzylic group yielded the intermediate benzyl bromide 3.
This compound was transformed into the Boc-protected benzyl-
amine 4 through a sequence of three steps, comprising the substi-
tution of 3 by (Boc)2NH, selective removal of one Boc group by
trifluoroacetic acid (TFA), and finally hydrolysis of the methyl ester
under basic conditions. The resulting free acid 4 is reacted with
aminopyridines, using HATU/DMAP as coupling reagent to provide
intermediates 5–6. The key step to introduce the aromatic ring dis-
playing the carboxylic acid is a Pd-catalyzed Suzuki reaction, using
commercially available boronic acids. Further esterification of the
resulting carboxylic derivatives 7–9 with a selection of alcohols
can be achieved through various methods, including TBTU, Steglich
method or the use of Ghosez reagent, to activate the acid. Final
deprotection by TFA or HCl gas wraps up the synthesis of this ini-
tial series of compounds.
In a first step, we checked what the optimal position was for the
introduction of an ester side chain on the phenyl ring of 1. Replace-
ment of the 4-fluoro by a methyl ester yielded compound 10,
which had significantly reduced potency compared to 1. Introduc-
tion of the same methyl ester on position 3 was however tolerated,
resulting in the potent compound 11 (Table 1).
Molecular docking of 1 and 11 in ROCK explains the observed
difference of potency between compounds 10 and 11. The
proposed binding mode for compounds 1 and 11 (Fig. 3) is similar
to Y-27632, with the additional phenyl moiety positioned under
the P-loop (Gly-rich loop) of ROCK. Position 4 (4-F in 1) closely
a
Values are means of two experiments.
contacts residues from the P-loop, discouraging introduction of
bulky functional groups. In contrast, position 3 provides access to
a solvent-exposed cleft which can accommodate the ester group
with minimal perturbation of the binding mode.
In view of these initial results, priority was given to position 3
for further optimization of the series. On-target potency of the
candidate soft ROCK inhibitors was evaluated externally, with
Y-27632 as comparator. While Y-27632 appeared relatively potent
in this assay compared to literature data, its activity remained con-
sistent throughout independent experiments. Activity data for rep-
resentative compounds is provided in Table 2. Degradation of the
compounds in human plasma was used as a first estimate of their
soft drug characteristics.
Ester derivatives displaying C1–C6 linear alkyls were first syn-
thesized and evaluated. While such chains were initially tolerated,
increased length ultimately resulted in decreased activity, as illus-
trated by 16. A plausible explanation to this observation is that the
solvent-exposed cleft accommodating the ester chain has limited
affinity towards lipophilic groups. Flexible alkyl chains could also
lose conformational freedom upon binding, resulting in a substan-
tial loss of entropy and thereby in reduced potency. This hypothe-
sis was tested by replacing the n-hexyl moiety of 16 by a
Figure 2. Synthetic scheme for synthesis of compounds 10–35. (a) H2SO4, MeOH, 60 °C, 16 h; (b) NBS, AIBN, CCl4, reflux, 16 h; (c) Boc2NH, t-BuOK, DMF, rt, 16 h; (d) DCM/TFA
(50:1), 0 °C ? rt, 4 h; (e) NaOH, MeOH, 50 °C, 2 h; (f) HATU, DMAP, NEt3, DMA, 30 °C, 16 h; (g) Pd(dppf)Cl2, Na2CO3, H2O, DMF, 100 °C; 16 h; (h) ROH, TBTU, HOBT, DIEA, DMF,
rt, 16 h or ROH, DCC, DMAP, DCM, rt, 16 h or Me2C@C(Cl)NMe2, THF or DCM, rt, followed by ROH, 16 h; (i) DCM/TFA (7:1), 30 °C, 16 h or HCl(g) in DCM, 30 °C, 16 h.

6444
S. Boland et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6442–6446
lower functional efficacy observed for these compounds. Exposure
to alcohols resulting from hydrolysis of compounds 11–33 should
also be taken into account when designing/selecting soft ROCK
inhibitors. However, such exposure will highly depend on use, dose
and mode of administration of the compound. Such considerations
go beyond the scope of this exploratory work.
The soft drug concept involves conversion of the parent com-
pound to the inactive metabolite. This aspect was initially assessed
by evaluating the plasma stability of compounds 11–33. We ob-
served a clear influence of linear alkyl chain length on stability
(11–16). Derivatives displaying shorter alkyl chains were
hydrolyzed at a relatively slow, although measurable rate. Plasma
stability decreased with increasing chain length, to reach a mini-
mum with 13, which displayed a half life of 18 min. The stability
of the cycloalkylmethyl derivatives was, with respect to their linear
analogs, similar (25 vs 14) or lower (21 and 23 vs 16; 24 vs 15). The
cyclohexyl ester 19 displayed significantly increased plasma stabil-
ity compared to its n-hexyl analog 16. Replacement of cyclohexyl
moieties by phenyl had a variable effect. While a fast hydrolysis
was observed for the phenyl ester 20, the difference between the
benzyl derivative 26 and its cyclohexylmethyl analog 22 was not
striking. Replacement of alkyl chains by ether chains resulted in
compounds with relatively short plasma half-life. This was espe-
cially clear for compounds 31 (vs 19) and 32 (vs 23). In contrast,
the hydroxyethyl derivative 30 was more stable than the closely
related methoxyethyl, propyl or butyl analogs. Surprisingly, some
Figure 3. Molecular modeling of compounds 1 (cyan sticks) and 11 (orange sticks)
in ROCK1 (PDB 2ETR). Y-27632 is depicted as purple sticks for comparison.
3-fluoro-4-amino-pyridine derivatives were hydrolyzed at
a
slightly increased rate with respect to their 4-aminopyridine ana-
logs. Importantly, the carboxylic derivatives 34 and 35 were not
significantly degraded (<5%) over the course of the experiment,
indicating that compound degradation resulted, as intended, from
ester hydrolysis. This was further confirmed by simultaneously
monitoring disappearance of selected compounds and appearance
of the corresponding metabolite in plasma samples (data not
shown). One can therefore conclude that once in plasma, most
compounds will be rapidly converted into the functionally inactive
34 or 35.
The collected stability data clearly indicate that the degradation
of compounds 11–33 in plasma is due, as intended, to enzymatic
hydrolysis of the carboxylic ester moiety present in those com-
pounds. Firstly, the structurally related compounds 1, 34 and 35
were stable in plasma (<5% disappearance of parent compound)
over the course of the experiment. Secondly, the rapid hydrolysis
of compounds 11–33 in plasma was not reproduced in phosphate
buffered saline (PBS) pH 7.4, although some degree of hydrolysis
was noticed for some compounds such as 19 (data not shown).
Finally, the observation that increasing linear alkyl chain length
initially results in faster hydrolysis rates in plasma differs from
common rules governing the chemical stability of benzoic esters
wherein the derivatives displaying the shorter, less hindered alkyl
chains are more rapidly hydrolyzed;18 and are instead reminiscent
of substrate preferences observed with some hydrolytic enzymes,
including esterases.19
Among the various hydrolytic enzymes, carboxylic ester hydro-
lases (EC 3.1.1) probably play a major role in the metabolism of
compounds 11–33 in plasma. While being essentially devoid of
carboxylesterase activity, human plasma contains two main
esterases (butyrylcholinesterase and paraoxonase 1) and a pseudo-
esterase20 (serum albumin). Simultaneous involvement of these
esterases in drug hydrolysis has been demonstrated for com-
pounds as simple as aspirin.21 At this stage, the esterase(s) respon-
sible for hydrolysis of soft ROCK inhibitors has/have not been
formally attributed for each compound. Biochemical intervention
with specific inhibitors is a common way to identify the esterase(s)
involved in drug or prodrug metabolism,21 which could provide
further elucidation of the metabolism of compounds 11–33.
cyclohexyl, but compound 19 did not improve over 16. However,
introduction of a cyclopentylmethyl in 21 led to a 4-fold increase
in potency. This finding prompted the synthesis of additional
cycloalkylmethyl derivatives. Increasing the ring size yielded 22,
which displayed an activity comparable to Y-27632. On the other
hand, reducing ring size delivered 24 and 25, which had single digit
nanomolar potency. Replacement of the cyclohexylmethyl by a
benzyl did not restore activity (26). Introduction of heteroatoms
was generally well tolerated, with the linear ether derivatives 27
and 28 displaying an on-target activity similar or slightly better
than their isosteric alkyl analogs 14 and 15. As with previous deriv-
atives, extension of the ester chain with further lipophilic groups
resulted in decreased potency (27). However, introduction of a
hydroxyl group generated the potent compound 30. Similarly,
cyclic ethers turned out to be viable alternatives to cycloalkyls
(31–33). 3-Fluoro-4-aminopyridine derivatives had slightly re-
duced potency (2 to 3-fold) when compared to the corresponding
4-aminopyridine analogs (11 vs 17; 13 vs 18; 21 vs 23).
Myosin light chain (MLC) phosphorylation status is a direct
marker of the ROCK enzymatic activity in cells.16 Low ROCK activ-
ity results in decreased MLC phosphorylation. Evaluation of se-
lected compounds in
a cellular MLC phosphorylation assay
confirmed their functional efficacy. The most potent representa-
tives in the series (11, 13 and 32) displayed EC50 values in the
low nanomolar range, and were comparable to compound 1 and
clearly lower than with Y-27632.
An important aspect of the soft drug approach resides in the de-
creased activity of the pre-defined metabolite. Compounds 34 and
35, which represent the metabolites resulting from ester hydroly-
sis of 11–33, had lower on-target potency than most of their parent
compounds. This reduction in activity was further apparent in the
MLC phosphorylation assay, where 34 and 35 were almost devoid
of functional efficacy. Interestingly, compounds 34 and 35 both
displayed very low permeability (Papp <0.01 10ꢀ6 cm/s). This result
is consistent with literature data indicating lower membrane per-
meability for acidic or zwitterionic species compared to neutral or
basic species.17 As ROCK is an intracellular target, it is therefore
possible that reduced membrane permeability plays a role in the

S. Boland et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6442–6446
6445
Table 2
On-target potency, cellular efficacy and stability data for 11–35
a
b
Compd
X
R
ROCK2 IC50
(nM)
MLC-PP EC50
(nM)
t1/2 Plasma
(min)
Y-27632 NA NA
NA NA
54
200
6.1
8.0
>120d
>120d
91
41
18
22
58
50
65
19
>120
7
31
99
17
13
18
64
Figure 4. IOP-lowering effect of 32 (0.5% solution) following topical administration
to ocular normotensive NZW rabbits. ⁄P <0.05. ⁄⁄P <0.01.
1
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
1.2
2.9
<1.0c
1.4
5.5
4.9
22
7.8
5.1
20
H
H
H
H
H
H
F
F
H
F
H
H
F
F
F
H
H
H
F
F
F
F
F
H
F
Me
Et
nPr
nBu
nPen
nHex
Me
nPr
ND
9.1
ND
ND
ND
130
ND
ND
ND
70
UK). Higher hydrolytic activities were observed in cornea, conjunc-
tiva and sclera, indicating that 32 could actually be metabolized
and behave as a soft ROCK inhibitor in those tissues.
In spite of its metabolically labile nature, 32 had a pronounced
IOP-lowering effect when administered topically to ocular normo-
cHex
Phe
6.2
5.0
48
tensive NZW rabbits (40 ll of a 0.5% w/v solution in 1:1 PEG:water
CH2-cPen
CH2-cHex
CH2-cPen
CH2-cBu
CH2-cPr
Benzyl
(CH2)2OMe
(CH2)3OMe
(CH2)2OiPr
(CH2)2OH
Oxan-3-yl
vehicle with pH adjusted to 7.0). A significant IOP decrease vs. the
vehicle-treated contralateral eye was observed during the six
hours following administration (Fig. 4). The highest average IOP
reduction was achieved 3 h after instillation, and was 28.7% from
baseline (ꢀ3.85 mmHg). No obvious hyperemia was noticed over
the course of the experiment, suggesting that soft ROCK inhibitors
such as 32 could potentially separate the desired IOP-lowering
effect of ROCK inhibitors from this commonly observed side effect.
Further development and evaluation of this compound series
resulted in the discovery of AMA0076, Amakem’s clinical candidate
for the treatment of glaucoma. A detailed evaluation of AMA0076,
including dose ranging studies, rigorous scoring of hyperemia and
comparison to other IOP-lowering agents will be presented in an
upcoming dedicated paper.
In conclusion, we herein reported the first examples of soft
ROCK inhibitors. These compounds are structurally related to
Y-27632, but contain carboxylic ester functions allowing their
rapid hydrolysis by esterases, resulting in a functionally inactive
metabolite. A first in vivo proof of concept illustrating the potential
of soft ROCK inhibitors as IOP-lowering agents with reduced
hyperemic effects was also presented. Clinical development of
ROCK inhibitors for multiple indications, such as asthma, COPD,
IBD or even glaucoma, has so far been limited by ROCK-associated
side effects. In this context, we believe that soft ROCK inhibitors
displaying an improved therapeutic window might represent a no-
vel therapeutic approach towards many of these indications, espe-
cially those compatible with topical or local drug delivery.
ND
52
13
4.3
2.7
45
3.6
1.5
37
ND
ND
ND
170
28
ND
ND
ND
16
12
26
6
81
18
<5
7
1.4
8.7
CH2-oxolan-2-yl 2.5
CH2-oxan-2-yl
H
H
12
167
55
ND
>1000
>1000
>120d
>120d
a
b
c
Values are means of at least two experiments.
Values are means of three experiments.
ROCK2 concentration in the assay was 1 nM.
d
<5% Degradation over the course of the experiment.
Compound 32 was selected for in vivo evaluation since it
combined strong efficacy with rapid hydrolysis in plasma. Topical
administration to ocular normotensive New Zealand White
(NZW) rabbits and monitoring of the IOP was selected for this first
proof of concept. This animal model is easily implemented, corre-
sponds to a potential indication for ROCK inhibitors (IOP-lowering
agents for the treatment of glaucoma) and allows for an easy mon-
itoring of the expected side effect, conjunctival hyperemia.
The hydrolytic activity of rabbit eye tissues towards 32 was first
assessed by incubating the compound at a concentration of 20 lM,
in presence of tissue samples. A specific hydrolytic activity was
calculated for each tissue (Table 3) by normalizing the observed
degradation rate of 32 with respect to tissue weight. Low
hydrolytic activity (0.12 pmol minꢀ1 mgꢀ1) was found in aqueous
humor, corresponding to a half-life of 116 min in undiluted aque-
ous humor. This result was in line with the t1/2 value obtained
when incubating 32 with undiluted, pooled AH samples from
NZW rabbits (Sera Laboratories International, Haywards Heath,
Acknowledgments
The authors wish to thank Dr. Jean-Marie Stassen and Dr. Steve
Pakola for their contribution and advice during the review of this
manuscript. The authors also wish to thank Jelle Sterckx, Pieter
Schroeders and Jacques Geraets for their support during compound
evaluation.
Table 3
Supplementary data
Specific hydrolytic activity of rabbit eye tissues towards 32
Tissue
Specific hydrolytic activity (pmol minꢀ1 mgꢀ1
)
Supplementary data (experimental protocols regarding synthe-
sis of compounds 11–35, full characterization of compound 32 and
of its metabolite 35 and experimental protocols regarding in vitro
and in vivo evaluation of compounds 11–35) associated with this
article can be found, in the online version, at http://dx.doi.org/
10.1016/j.bmcl.2013.09.040.
Aqueous humor
Vitreous humor
Conjunctiva
Sclera
0.12
0.22
1.52
1.00
2.71
Cornea

6446
S. Boland et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6442–6446
11. Bodor, N.; Buchwald, P. In Burger’s Medicinal Chemistry, Drug Discovery and
References and notes
Development, 6th ed.; John Wiley & Sons Inc.: New York, 2003; pp 533–608.
12. Erhardt, P. W.; Reese, M. D. In Prodrugs and Targeted Delivery: Towards Better
ADME Properties; Rautio, J., Ed.; Wiley-VCH Verlag GmbH & Co.: Weinheim,
2011; pp 385–413.
13. Van Rompaey, P.; Arzel, P.; Defert, O.; Leysen, D. WO/2007/042321, 2007.
14. Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P. Biochem. J. 2000, 351, 95.
15. Jacobs, M.; Hayakawa, K.; Swenson, L.; Bellon, S.; Fleming, M.; Taslimi, P.;
Doran, J. J. Biol. Chem. 2006, 281, 260.
16. Schröter, T.; Griffin, E.; Weiser, A.; Feng, Y.; LoGrasso, P. Biochem. Biophys. Res.
Commun. 2008, 374, 356.
17. Gleeson, M. P. J. Med. Chem. 2008, 51, 817.
18. Hancock, C. K.; Falls, C. P. J. Am. Chem. Soc. 1961, 83, 4214.
19. Khersonsky, O.; Tawfik, D. S. Biochemistry 2005, 44, 6371.
20. Li, B.; Sedlacek, M.; Manoharan, I.; Boopathy, R.; Duysen, E. G.; Masson, P.;
Lockridge, O. Biochem. Pharmacol. 2005, 70, 1673.
1. Matsui, T.; Amano, M.; Yamamoto, T.; Chihara, K.; Nakafuku, M.; Ito, M.;
Nakano, T.; Okawa, K.; Iwamatsu, A.; Kaibuchi, K. EMBO J. 1996, 15, 2208.
2. Riento, K.; Ridley, A. J. Nat. Rev. Mol. Cell Biol. 2003, 4, 446.
3. Liao, J. K.; Seto, M.; Noma, K. J. Cardiovasc. Pharmacol. 2007, 50, 17.
4. LoGrasso, P. V.; Feng, Y. Curr. Top. Med. Chem. 2009, 9, 704.
5. Doe, C.; Bentley, R.; Behm, D. J.; Lafferty, R.; Stavenger, R.; Jung, D.; Bamford,
M.; Panchal, T.; Grygielko, E.; Wright, L. L.; Smith, G. K.; Chen, Z.; Webb, C.;
Khandekar, S.; Yi, T. J. Pharmacol. Exp. Ther. 2007, 320, 89.
6. Fernandes, L. B.; Henry, P. J.; Goldie, R. G. Ther. Adv. Respir. Dis. 2007, 1, 25.
7. Kast, R.; Schirok, H.; Figueroa-Pérez, S.; Mittendorf, J.; Gnoth, M. J.; Apeler, H.;
Lenz, J.; Franz, J. K.; Knorr, A.; Hütter, J.; Lobell, M.; Zimmermann, K.; Münter,
K.; Augstein, K. H.; Ehmke, H.; Stasch, J. P. Br. J. Pharmacol. 2007, 152, 1070.
8. Mandell, K. J.; Kudelka, M. R.; Wirostko, B. Expert Rev. Ophthalmol. 2011, 6, 611.
9. Tokushige, H.; Inatani, M.; Nemoto, S., et al Invest. Ophthalmol. Vis. Sci. 2007, 48,
3216.
21. Bahar, F. G.; Ohura, K.; Ogihara, T.; Imai, T. J. Pharm. Sci. 2012, 101, 3979.
10. Tanihara, H.; Inatani, M.; Honjo, M.; Tokushige, H.; Azuma, J.; Araie, M. Arch.
Ophthalmol. 2008, 126, 309.