A multivalent approach towards linked dual-pharmacology prostaglandin F receptor agonist/carbonic anhydrase-II inhibitors for the treatment of glaucoma

Article abstract of DOI:10.1016/j.bmcl.2012.12.058

Lowering of intra-ocular pressure is the primary pharmacologic approach for the treatment of glaucoma and a number of distinct mechanisms of action have been clinically validated. Targeting of multiple mechanisms in combination therapies has proven effective both clinically and commercially although potential improvements with regards to efficacy, tolerability and dosing frequency remain. Application of Theravance's multivalent approach to drug discovery towards linked dual-pharmacology prostaglandin F receptor (FP) agonist/carbonic anhydrase (CA)-II inhibitor compounds is described. Compound 29 exhibits weak potency (pEC₅₀ = 5.7, IA >1.0) as an FP agonist with high binding affinity (pK_i = 8.1) to the CA-II enzyme, and has comparable corneal permeability to the CA-II inhibitor dorzolamide.

Full text of DOI:10.1016/j.bmcl.2012.12.058

Bioorganic & Medicinal Chemistry Letters 23 (2013) 939–943
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A multivalent approach towards linked dual-pharmacology
prostaglandin F receptor agonist/carbonic anhydrase-II inhibitors
for the treatment of glaucoma
⇑
Daniel D. Long , Bryan Frieman, Sharath S. Hegde, Craig M. Hill, Lan Jiang, Samuel Kintz,
Daniel G. Marquess, Hans Purkey, Jeng-Pyng Shaw, Tod Steinfeld, Michael S. Wilson, Kevin Wrench
Theravance, Inc., 901 Gateway Blvd, South San Francisco, CA 94080, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Lowering of intra-ocular pressure is the primary pharmacologic approach for the treatment of glaucoma
and a number of distinct mechanisms of action have been clinically validated. Targeting of multiple
mechanisms in combination therapies has proven effective both clinically and commercially although
potential improvements with regards to efficacy, tolerability and dosing frequency remain. Application
of Theravance’s multivalent approach to drug discovery towards linked dual-pharmacology prostaglan-
din F receptor (FP) agonist/carbonic anhydrase (CA)-II inhibitor compounds is described. Compound 29
exhibits weak potency (pEC50 = 5.7, IA >1.0) as an FP agonist with high binding affinity (pKi = 8.1) to
the CA-II enzyme, and has comparable corneal permeability to the CA-II inhibitor dorzolamide.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 20 November 2012
Revised 14 December 2012
Accepted 17 December 2012
Available online 22 December 2012
Keywords:
Prostaglandin F receptor (FP) agonist
Carbonic anhydrase (CA)-II inhibitor
Intra-ocular pressure (IOP)
Glaucoma
Multivalent
Dual-pharmacology
Glaucoma is an ocular disease in which apoptosis of retinal gan-
glion cells and cupping of the optic disc occurs leading to progres-
sive loss of visual field.1 It has been estimated that the disease will
affect more than 80 million individuals worldwide by 2020, with at
least 6–8 million patients becoming bilaterally blind.2 Glaucoma is
associated with increased aqueous humor fluid pressure in the
eye and an intra-ocular pressure (IOP) of >21 mmHg is a well-
established primary risk factor for progression of the disease.3
Whilst loss of vision as a result of glaucoma cannot be reversed
there are a number of therapeutic options for early intervention
for those at risk with elevated IOP. Surgical4 and laser5 procedures
primarily focus on facilitating the outflow of aqueous humor either
through disruption of the trabecular meshwork or creation of
filtering channels. Separate pharmacological approaches exist to
affect both aqueous humor drainage and formation.6,7
agonists of the prostaglandin F receptor (FP) resulting in increased
uveoscleral outflow.
Other pharmacologic approaches target aqueous humor pro-
duction by the ciliary epithelium via distinct mechanisms such as
b1/2-adrenergic receptor antagonism [e.g., timolol (6)] and carbonic
anhydrase (CA)-II enzyme inhibition [e.g., dorzolamide (7), brinzo-
lamide (8) and acetazolamide (9)], (Fig. 1). Agonism of the a2-adre-
noceptor [e.g., brimonidine (10) has the dual outcome of both
increasing aqueous humor outflow and reducing aqueous humor
production, (Fig. 1). This dual effect is commonly replicated by
the use of combinations of the aforementioned agents as both
adjuvants or as fixed-dose combinations and can lead to improve-
ments in efficacy relative to monotherapy.9 In Europe, a number of
fixed-dose combinations containing the first-line prostaglandin
therapies as one component are available [e.g., latanoprost-timolol
maleate (XalacomÒ)] but none have been approved by the US
FDA.10 Advantages of the fixed-dose combinations relative to adju-
vant therapy include reduction of clinical side effects (largely due
to minimized exposure to eye-drop solution preservatives), and
convenience of dosing and patient compliance. In the US, dorzola-
mide hydrochloride-timolol maleate (CosoptÒ) and brimonidine
tartarate-timolol maleate (CombiganÒ) are the only fixed-dose
combinations currently available and whilst they exhibit improved
dosing convenience relative to simultaneous therapy with the indi-
vidual components, they remain twice-a-day products.11,12
A number of analogues of the prostaglandin PGF2 (1) [latano-
a
prost (XalatanÒ) (2), bimatoprost (LumiganÒ) (3), travoprost
(TravatanÒ) (4), tafluprost (TaflotanÒ) (5)] have been approved in
the US for the treatment of glaucoma and constitute a first line ap-
proach for the lowering of IOP, (Fig. 1).8 With the exception of
bimatoprost (3) (a prostamide) all the synthetic agents are ester
prodrugs which upon cleavage to the carboxylic acid act as
⇑
Corresponding author.
E-mail address: dlong@theravance.com (D.D. Long).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.12.058

940
D. D. Long et al. / Bioorg. Med. Chem. Lett. 23 (2013) 939–943
O
O
O
O
α-chain
OH
R
OiPr
OiPr
HO
HO
HO
HO
CF3
ω-chain
HO
HO
HO
HO
O
O
F
F
HO
HO
HO
PGF2α, 1
Latanoprost, 2 (R=OiPr)
Bimatoprost, 3 (R=NHEt)
Travoprost, 4
Tafluprost, 5
O
N
HN
O
N
N
N
N
O
S
O
O
S
N
N
NH2
S
S
NH2
N
H
N
H
N
X
H
S
N
O
S
O
Br
N
O
O
H
HO
Timolol, 6
Dorzolamide, 7 [X=CH-(S)-Me]
Acetazolamide, 9
Brimonidine, 10
Brinzolamide, 8 [X=N(CH2)3OMe]
Figure 1. Marketed agents for the treatment of glaucoma.
To date only one fixed-dose combination of an FP agonist and
Zn2+ ion it is evident that the remainder of the molecule can extend
out of the active site cleft via the -chain, and the -chain can be
accommodated; both chains forming potential close hydrophobic
interactions on the surface of the enzyme. We did not model the
interaction of this molecule and others at the FP receptor but it
CA-II inhibitor has been clinically investigated (travoprost–brinzo-
lamide) although no results have been reported and the most re-
a
x
cent Phase
3
study was terminated.13 The most significant
challenge to such an approach is in matching the pharmacokinetics
between the once-a-day FP agonist and the three-times-a-day CA-II
inhibitor for optimal dosing and efficacy. On the basis of supportive
clinical data14 highlighting the improved efficacy of a CA-II inhibi-
tor as an adjuvant to an FP agonist versus the FP agonist alone we
chose to investigate a multivalent, linked dual-pharmacology
approach15,16 whereby both activities would be incorporated into
a single compound. If suitable matched efficacy of each pharmaco-
logical component can be achieved with overall physicochemical
properties suitable for eye-drop formulation and convenient
topical ocular delivery then such an approach offers significant
patient benefit. An analogy can clearly be made with our prior
work on multivalent, linked dual-pharmacology compounds
combining muscarinic M3 acetylcholine receptor antagonism and
b2-adrenoceptor agonism (MABA) for the topical (lung selective)
treatment of COPD that has proven clinically successful in several
Phase 2 trials.16,17
The active site of the metallo-enzyme CA contains a key cata-
lytic Zn2+ ion and it is well established that aryl sulfonamide com-
pounds provide a common inhibitory binding motif.18 Our initial
strategy was to establish whether we could incorporate aryl sul-
fonamide fragments based on dorzolamide (7), brinzolamide (8),
acetazolamide (9) or simple phenyl sulfonamides onto the FP ago-
nist pharmacophore. We chose to retain the dihydroxylated cyclo-
was evident that loss of the terminal
detrimental to potency. As a result we subsequently focused on
keeping the FP agonist -chain intact and probed substitution of
a-chain carboxylic acid was
a
aryl sulfonamide fragments onto the
x-chain as the point of
attachment.
To enable the
building block was required from which novel
x
-chain substitution strategy a functionalized
-chains could be
x
introduced. Scheme 2 details the synthesis of key aldehyde build-
ing blocks 17 and 18. Dess–Martin oxidation of the Corey lactone
14 and subsequent protection as the acetal afforded 15. The
a-chain was introduced via DIBALH reduction of the lactone and
Wittig reaction to give acid 16 with the Z-olefin configuration as
expected.20 The acid was then protected as either the isopropyl
or methyl ester, the hydroxy groups silyl-protected and the acetal
cleaved to provide the aldehydes 17 and 18.
Reductive amination of an amine functionalized acetazolamide
compound 1921 directly onto isopropyl ester 18, and subsequent
hydroxyl deprotection afforded the ester 20 which was hydrolyzed
to the acid analogue 21, Table 1, Scheme 3. Both the ester 20 and
acid 21 exhibited moderate binding to CA-II (pKi = 7.4 and 7.0,
respectively) albeit weaker than acetazolamide (9) (pKi = 8.1).
However, both compounds displayed no functional potency at
the FP receptor (pEC50 <5.3).
pentane core of PGF2 (1) and its FDA-approved derivatives 2–5,
Extending the aryl sulfonamide moiety along the FP pharmaco-
a
and explore both the
a- and
x
-chains as suitable points of attach-
phore x-chain by Wittig reaction of aldehyde 18, subsequent amide
ment to the CA ligand, (Fig. 1). Schemes 1–5 detail the preparation
of the compounds described in this Letter.
coupling of the resultant acid 22 with amines, followed by hydroxyl
deprotection, afforded the amide derivatives 23 and 24, Table 1,
Scheme 4. In the case of compound 23 it was hydrolyzed to the cor-
Simple amide derivatization of the FP agonist
a-chain carbox-
ylic acid by coupling of PGF2 (1) with amines afforded the aryl sul-
responding a-chain acids 25. Once again no functional potency at
a
fonamides 11–13, Table 1, Scheme 1. Interestingly, whilst the
functional potency at the FP receptor was ablated (pEC50 65.4) rel-
the FP receptor (pEC50 <5.3) was observed for any of the analogues
23–25. However, significant binding to CA-II (pKi = 8.9) was evident
for the acetazolamide derivative 24, demonstrating that the CA-II
enzyme is tolerant of the FP pharmacophore linked to an aryl sulfon-
ative to PGF2 (1) (pEC50 = 7.9) for all the compounds, significant
a
binding to CA-II was observed and in the case of the acetazolamide
derivative 13 the binding affinity to CA-II (pKi = 9.2) approached
that of brinzolamide (8) (pKi = 9.6) and was in fact greater than
acetazolamide (9) (pKi = 8.1) itself. This provided an interesting
first attempt and to rationalize the tolerance of the CA-II binding
site for the bulky FP pharmacophore the interaction of compound
13 at the active site of the CA-II enzyme was modeled, (Fig. 2).19
Fixing the key interaction of the sulfonamide with the catalytic
amide through either its
a- or x-chain.
An additional approach to derivatization of the
x
-chain in-
volved efficient Seyferth–Gilbert homologation22 of aldehyde 17
to the alkyne 26. ‘Click’ reaction23 of the alkyne with azide 27,24
hydroxyl deprotection to afford ester 28 and subsequent hydroly-
sis to acid 29 afforded two novel triazole compounds with interest-
ing activity, Table 1, Scheme 5. Both the ester 28 and acid 29

D. D. Long et al. / Bioorg. Med. Chem. Lett. 23 (2013) 939–943
941
R - Aryl Sulfonamide groups:
O
S
NH2
O
O
N
H
O
OH
R
HO
HO
(i)
11
O
N
HO
HO
O
S
N
HO
1
HO
11-13
N
N
O
O
N
H
S
O
S
NH2
S
13
O
S
O
NH2
12
Scheme 1. Reagents and conditions: (i) amine; EDCIÁHCl, 1-HOBt, 2,6-lutidine, DCM (11: 26%; 13: 27%) or DMF (12: 37%).
O
O
O
O
OH
OR
HO
TBDPSO
O
O
(i)
(ii)
(iii)
O
H
OH
O
O
O
HO
TBDPSO
O
O
O
O
O
14, Corey lactone
15
16
17: R=Me
18: R=iPr
Scheme 2. Reagents and conditions: (i) (a) Dess–Martin periodinane, NaHCO3, DCM, (b) 1,2-bis(trimethylsilyloxy)ethane, TMSOTf, DCM, 94% over 2 steps; (ii) (a) DIBALH,
toluene, DCM, 81%, (b) (4-carboxybutyl)triphenylphosphonium bromide, KOtBu, THF, 76%; (iii) For 17 (a) TMS-diazomethane, Et2O, toluene, MeOH, 94%, (b) TBDPS-Cl,
i
imidazole, DMF, 100%, (c) TMSOTf, 2,6-lutidine, DCM, 76%; For 18 (a) PrI, Cs2CO3, DMF, 88%, (b) TBDPS-Cl, imidazole, DMF, 81%, (c) TMSOTf, 2,6-lutidine, DCM, 77%.
O
O
O
OiPr
OiPr
OH
TBDPSO
HO
HO
H
NH
O
NH
O
H2N
(i)
O
TBDPSO
HO
(ii)
HO
N
N
N
N
O
S
O
S
N
S
O
N
HN
N
N
O
S
S
NH2
NH2
H
H
S
NH2
O
O
O
19
18
20
21
Scheme 3. Reagents and conditions: (i) (a) NaBH3CN, MeOH, (b) TBAF, THF, 30% over 2 steps; (ii) LiOH, MeOH, H2O, 80%.
R1 - Aryl Sulfonamide
groups:
O
O
O
O
S
OiPr
OiPr
OR2
TBDPSO
TBDPSO
HO
O
N
S
(i)
(ii)
R1
H
O
S
O
OH
TBDPSO
TBDPSO
HO
O
23 / 25
NH2
O
O
23-24: R2=iPr
25: R2=H
18
22
N
N
(iii)
O
S
N
H
S
NH2
O
24
Scheme 4. Reagents and conditions: (i) diethylphosphonoacetic acid, NaH, Et2O, 60%; for 23/25 (ii) (a) amine; EDCIÁHCl, 1-HOBt, 2,6-lutidine, DMF, (b) TBAF, THF, 47% over 2
steps; (iii) LiOH, tBuOH, H2O, 87%; for 24 (ii) (a) amine; EDCIÁHCl, 1-HOBt, 2,6-lutidine, DCM, (b) TBAF, THF, 10% over 2 steps.
showed good binding to CA-II (pKi = 8.1) and in the case of the acid
29 weak, full agonist potency at the FP receptor (pEC50 = 5.7,
IA >1.0) was observed.
With the first indication that a linked dual-pharmacology FP-
agonist/CA-II inhibitor might be possible through the identification
of triazole compound 29, the physicochemical properties of the

942
D. D. Long et al. / Bioorg. Med. Chem. Lett. 23 (2013) 939–943
O
O
O
OMe
OMe
OR1
TBDPSO
TBDPSO
HO
H
(i)
(ii)
O
S
O
TBDPSO
TBDPSO
HO
N
O
S NH2
O
O
N
NH2
H
N
N3
28: R1=Me
29: R1=H
27
17
26
(iii)
Scheme 5. Reagents and conditions: (i) dimethyl(1-diazo-2-oxopropyl)phosphonate, K2CO3, MeOH, 92%; (ii) (a) CuI, Et3N, 1,4-dioxane, 60 °C, (b) TBAF, THF, 100% over 2
steps; (iii) LiOH, tBuOH, H2O, 64%.
Table 1
mFP IAc
Caco-2 Kp (1 Â 10À6 cm/s)d
Rabbit corneal Kp (1 Â 10À6 cm/s)e
a
b
Compound
Scheme
hCA-II pKi
mFP pEC50
PGF2 : 1
Latanoprost acid
Dorzolamide: 7
Brinzolamide: 8
—
—
—
—
—
1
1
1
3
3
—
—
7.9
8.2
—
—
1.0
1.0
—
—
—
—
—
0.9
—
—
3.3
3.5
5.2
—
2.5
—
9.7
1.2
2.6
—
3.3
—
a
9.7
9.9
8.1
7.4
8.3
9.2
7.4
7.0
7.3
8.9
7.3
8.1
8.1
Acetazolamide: 9
—
11
12
13
20
21
23
24
25
28
29
<5.3
<5.3
5.4
<5.3
<5.3
<5.3
<5.3
<5.3
<5.3
5.7
—
—
1.6
2.1
3.0
—
—
—
6.4
3.1
0.94
—
—
—
4
4
4
5
—
—
—
—
3.2
3.3
2.1
1.7
5
>1.0
a
b
c
Human carbonic anhydrase II (hCA-II) binding affinities were determined using a dansylamide (DNSA) competition assay using human recombinant CA-II.
Mouse FP receptor (mFP) potencies were determined using a intracellular calcium mobilization assay (FLIPR) using 3T3 fibroblast cells endogenously expressing mFP.
Maximum compound-evoked response (minus basal) expressed as a percentage of the maximum response evoked by prostaglandin PGF2
Intestinal permeabilities were measured using a Caco-2 transport assay.
.
a
d
e
Corneal permeabilities were measured using rabbit corneal tissues mounted on vertical Ussing chambers.
were screened in the Caco-2 cell line (intestinal permeability mea-
sure) and a rabbit corneal preparation and compared with the
known ocular agents, Table 1. Whilst it is clear that both 29 and
28 are poorly permeable in both the intestinal (Kp = 3.3 and 3.2,
respectively) and corneal assays (Kp = 1.7 and 2.1, respectively)
and that there is little effect of the prodrug strategy, these com-
pounds are comparable to the CA-II inhibitors dorzolamide (7)
(Kp = 5.2 and 2.6, intestinal and corneal permeability, respectively)
and acetazolamide (9) (Kp = 2.5 and 3.3, intestinal and corneal per-
meability, respectively). The poor corneal permeability of dorzola-
mide (7) compared to latanoprost (2) correlates clinically with a
higher administered dose (600 lg/drop vs 1.5 lg/drop) to achieve
similar aqueous humor concentrations of active drug.25
In conclusion, evaluation of a multivalent approach towards
single molecule, linked dual-pharmacology FP-agonist/CA-II inhib-
itors reveals that potent CA-II inhibitors can be obtained through
the linkage of appropriate aryl sulfonamide moieties to either the
Figure 2. Docking model depicting compound 13 binding to the active site of the
CA-II enzyme.
a
- or
functionality is likely necessary for potent FP agonism. A number
of -chain linked aryl sulfonamides were prepared and the triazole
x-chain of the FP agonist pharmacophore. The a-chain acid
x
compound 29 indicates that weakly potent FP agonism can be
achieved in tandem with potent CA-II inhibition. The physico-
chemical properties of this compound 29 result in poor permeation
across an ocular membrane. Subsequent studies are focused on
exploring alternate linkages and points of attachment to the
compound were considered. Since the majority of the ocular agents
discussed in this Letter are administered topically to the eye it is
necessary that they have good solubility (important for eye drop
formulation and access to both the aqueous humor and relevant
tissue) and permeation properties (corneal and conjuctival path-
ways are common). This is clearly evidenced by the requirement
of a permeable ester prodrug of the FP-receptor agonists, for exam-
ple, latanoprost (2) (Kp >40, corneal permeability). To assess the
permeability of triazole 29 and its methyl-ester prodrug 28 both
a- or x-chain with a view to optimizing dual-pharmacology
potency as well as physicochemical properties to deliver appropri-
ate preclinical pharmacokinetics and pharmacodynamics in animal
models of IOP.

D. D. Long et al. / Bioorg. Med. Chem. Lett. 23 (2013) 939–943
943
M. Arch. Ophthalmol. 2010, 128, 825; (c) Lorenz, K.; Rosbach, K.; Matt, A.;
Pfeiffer, N. Clin. Ophthalmol. 2011, 5, 1745; (d) Pfeiffer, N. Graefes Arch. Clin. Exp.
Ophthalmol. 2011, 249, 1065.
Acknowledgment
The authors would like to thank Dr. Jeff Finer from Theravance,
Inc. for his assistance in reviewing this article.
15. Long, D. D.; Aggen, J. B.; Christensen, B. G.; Judice, J. K.; Hegde, S. S.; Kaniga, K.;
Krause, K. M.; Linsell, M. S.; Moran, E. J.; Pace, J. L. J. Antibiot. 2008, 61,
595.
16. Hughes, A. D.; Chin, K. H.; Dunham, S. L.; Jasper, J. R.; King, K. E.; Lee, T.-W.;
Mammen, M.; Martin, J.; Steinfeld, T. Bioorg. Med. Chem. Lett. 2011, 21, 1354.
17. Hughes, A. D.; McNamara, A.; Steinfeld, T. Prog. Med. Chem. 2012, 51, 71.
18. Krishnamurthy, V. M.; Kaufman, G. K.; Urbach, A. R.; Gitlin, I.; Gudiksen, K. L.;
Weibel, D. B.; Whitesides, G. M. Chem. Rev. 2008, 108, 946.
19. Software used for modelling work: LigPrep (3D ligand co-ordinate generation),
Maestro (Protein preparation) and Glide (Docking), all from Schrodinger.
20. Corey, E. J.; Weinshenker, N. M.; Schaaf, T. K.; Huber, W. J. Am. Chem. Soc. 1969,
91, 5675.
21. Amine prepared via amide coupling of acetazolamide. HCl with N-Boc-glycine
and subsequent Boc removal (i) N-Boc-glycine, HATU, DIPEA, DMF, 64%; (ii)
TFA, DCM, 92%.
22. (a) Seyferth, D.; Marmor, R. S.; Hilbert, P. J. Org. Chem. 1971, 36, 1379; (b)
Gilbert, J. C.; Weerasooriya, U. J. Org. Chem. 1982, 47, 1837.
23. Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40,
2004.
References and notes
1. Zhang, K.; Zhang, L.; Weinreb, R. N. Nat. Rev. Drug Disc. 2012, 11, 541.
2. Quigley, H. A. Lancet 2011, 377, 1367.
3. Musch, D. C.; Gillespie, B. W.; Niziol, L. M.; Lichter, P. R.; Varma, R.
Ophthalmology 2011, 118, 1766.
4. Morgan, W. H.; Yu, D.-Y. Clin. Exp. Ophthalmol. 2012, 40, 388.
5. Meyer, J. J.; Lawrence, S. D. Curr. Opin. Ophthalmol. 2012, 23, 111.
6. Lee, A. J.; Goldberg, I. Expert Opin. Emerg. Drugs 2011, 16, 137.
7. Carta, F.; Supuran, C. T.; Scozzafava, A. Expert Opin. Ther. Patents 2012, 22, 79.
8. Stewart, W. C.; Stewart, J. A.; Nelson, L. A.; Kruft, B. Eur. J. Ophthalmol. 2008, 18,
199.
9. Aptel, F.; Chiquet, C.; Romanet, J.-P. Drugs 2012, 72, 1355.
10. Aptel, F.; Denis, P. Curr. Med. Res. Opin. 2011, 27, 1949.
11. Khouri, A. S.; Realini, T.; Fechtner, R. D. Drugs Aging 2007, 24, 1007.
12. Feldman, R. M.; Bell, N. P.; Nagi, K. S. Exp. Rev. Ophthalmol. 2011, 6, 151.
13. ClinicalTrials.gov, NCT00767494.
24. Azide prepared via one-pot displacement of diazo analogue derived from 4-
aminobenzene sulfonamide (i) NaN3, tBuNO2, tBuOH, H2O, 66%.
25. http://labeling.pfizer.com/ShowLabeling.aspx?id=613 and http://www.merck.
com/product/usa/pi_circulars/t/trusopt/trusopt_pi.pdf.
14. (a) Lesk, M. R.; Koulis, T.; Sampalis, F.; Sampalis, J. S.; Bastien, N. R. Ann. Pharm.
2008, 42, 498; (b) Tanna, A. P.; Rademaker, A. W.; Stewart, W. C.; Feldman, R.