α-alkyl substituted pirinixic acid derivatives as potent dual agonists of the peroxisome proliferator activated receptor alpha and gamma

Article abstract of DOI:10.1002/ardp.200700209

Peroxisome proliferator-activated receptors (PPAR) are nuclear receptors, playing a pivotal role in energy homeostasis. Activators of the PPARα subtype are in widespread use for the treatment of hyperlipidemia, while activators of the PPARγ subtype are in clinical use for the treatment of type-2 diabetes. Since both of these diseases are frequently associated, the combined treatment with one drug simultaneously activating PPARα and PPARγ seems worthwhile. Starting with pirinixic acid, which is a moderately active dual PPARα/γ agonist, we improved potency at the human PPARα and PPARγ by substituting the a-position with an aliphatic chain. The maximal effect was achieved at a chain length of four and six carbons, respectively, leading to an activity induction by a factor of 36 for PPARα and 18 for PPARγ, respectively.

Full text of DOI:10.1002/ardp.200700209

Arch. Pharm. Chem. Life Sci. 2008, 341, 191–195
O. Rau et al.
191
Short Communication
a-Alkyl Substituted Pirinixic Acid Derivatives as Potent Dual
Agonists of the Peroxisome Proliferator Activated Receptor
Alpha and Gamma
Oliver Rau*, 1, Yvonne Syha*, 1, Heiko Zettl1, Michael Kock2, Andreas Bock2,
Manfred Schubert-Zsilavecz1
1 Institute of Pharmaceutical Chemistry/ZAFES, Johann Wolfgang Goethe University Frankfurt, Frankfurt,
Germany
2 Phenion GmbH & Co. KG, Frankfurt, Germany
Peroxisome proliferator-activated receptors (PPAR) are nuclear receptors, playing a pivotal role
in energy homeostasis. Activators of the PPARa subtype are in widespread use for the treatment
of hyperlipidemia, while activators of the PPARc subtype are in clinical use for the treatment of
type-2 diabetes. Since both of these diseases are frequently associated, the combined treatment
with one drug simultaneously activating PPARa and PPARc seems worthwhile. Starting with pir-
inixic acid, which is a moderately active dual PPARa/c agonist, we improved potency at the
human PPARa and PPARc by substituting the a-position with an aliphatic chain. The maximal
effect was achieved at a chain length of four and six carbons, respectively, leading to an activity
induction by a factor of 36 for PPARa and 18 for PPARc, respectively.
Keywords: Peroxisome proliferator activated receptor / Pirinixic acid / PPARa/c-Dual agonist /
Received: October 19, 2007; accepted: November 19, 2007
DOI 10.1002/ardp.200700209
Supporting Information for this article is available from the author or on the WWW under http://www.wiley-vch.de/
contents/jc_2019/2008/200700209_s.pdf
associated, dual agonists of PPARa and PPARc evolved as
potential drugs for “killing two birds with one stone” in
the treatment of hyperlipidemic type-2 diabetics [5]. In
addition, both types of drugs showed, both in clinical
and experimental data, additional beneficial effects such
as inhibition of inflammatory processes, which are medi-
ated at least in part by PPAR activation [6–8].
Pirinixic acid (WY14,643) is a weak agonist of both
PPARa and PPARc (Fig. 1) [9–14]. Here, we describe the
optimisation of this lead structure leading to dual
PPARa/c agonists with low micromolar activities.
Introduction
Peroxisome proliferator activated receptors (PPAR) are
members of the nuclear receptor superfamily, playing a
pivotal role in metabolism and homeostasis of lipids and
glucose. PPAR activation leads to an increased transcrip-
tion of key enzymes of cellular lipid or glucose uptake
and metabolism [1].
Today, fibrate-type PPARa agonists and glitazone-type
PPARc agonists are in clinical use for the therapy of
hyperlipidemia and type-2 diabetes, respectively [2–4]. As
both hyperlipidemia and type-2 diabetes are frequently
Correspondence: Manfred Schubert-Zsilavecz, Johann Wolfgang
Goethe University Frankfurt, Institute of Pharmaceutical Chemistry /
ZAFES, Max-von-Laue-Str. 9, D-60438 Frankfurt, Germany.
E-mail: Schubert-Zsilavecz@pharmchem.uni-frankfurt.de
Fax: +49 69 798-29332
Synthesis
Compounds 5–17 were generally synthesized based on
the method described by d'Atri et al. in a four-step reac-
Abbreviations: peroxisome proliferator activated receptors (PPAR); thia-
zolidinedione (TZD)
* These authors contributed equally to this work.
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

192
O. Rau et al.
Arch. Pharm. Chem. Life Sci. 2008, 341, 191–195
1 – 4a (X = H), 1 – 4b (X = Me), 1 – 4c (X = Et), 1 – 4d (X = n-Pr), 1 – 4e (X = n-Bu), 1 –
4f (X = n-Hex), 1 – 4g (X = n-Oct), 1 – 4h (X = Ph), 1 – 4i (X = Naph); v (R = 2,3-dime-
thylphenyl), w (R = 4-bromo-2,3-dimethylphenyl), x (R = 4-bromo-3-methylphenyl), y
(R = 4-chloro-2,3-dimethylphenyl), z (4-indanyl); 4a – k (R = 2,3-dimethylphenyl); 4k
(X = di-Me), 4l – o (X = H); 4l (R = 4-bromo-2,3-dimethylphenyl); 4m (R = 4-bromo-3-
methylphenyl); 4n (R = 4-chloro-2,3-dimethylphenyl); 4o (4-indanyl); 5 (X = Me), 6 (X
= Et), 7 (X = n-Pr), 8 (X = n-Bu), 9 (X = n-Hex), 10 (X = n-Oct), 11 (X = Ph), 12 (X =
Naph), 13 (X = Y = Me), 5 – 13 (R = 2,3-dimethylphenyl), 14 (R = 4-bromo-2,3-dime-
thylphenyl), 15 (R = 4-bromo-3-methylphenyl), 16 (R = 4-chloro-2,3-dimethylphenyl),
17 (4-indanyl), 14-17 (X = H).
Scheme 1. Synthesis route of presented compounds.
methyl groups at the a-position of the chlorinated ethyl
ester 3a was carried out employing LiHMDS as non–
nucleophilic base, methyl iodide as alkylating agent in
absolute THF and HMPA as co-solvent forming 3k.
Results and discussion
Figure 1. Dose-response curves for WY14643 and selected
compounds A: PPARa; B: PPARc
Starting with pirinixic acid, we performed a target-spe-
cific optimisation of this lead structure (Table 1). Ali-
phatic a-substitution of pirinixic acid (compounds 5–10
and 13) led to both an increased activity at the human
PPARa and PPARc with a more than 30-fold enhancement
of PPARa activity and a nearly 20-fold increase in the
activity of PPARc. PPAR activity enhancement by a-substi-
tution was maximised at a chain length of four and six
carbon atoms. Further elongation of the chain by an
octyl-substituent did not further enhance PPAR agonism.
More bulky residues adjacent to the a-position like phe-
nyl or naphthyl represented by compounds 11 and 12
diminished PPAR activities.
tion as shown in Scheme 1 [15]. Reaction of the appropri-
ate bromo ethyl esters 1a–i with thiobarbituric acid (i)
led to compounds 2a–i (compounds 1a–h were commer-
cially available, 1i was synthesized by bromination of
ethyl-2-naphtylethanoate at its a-position using NBS (N-
bromosuccinimide) and dibenzylperoxide in carbon tet-
rachloride). Aromatic nucleophilic substitution with
POCl3 (ii) gave the chlorinated ester 3a–i, which were
subsequently linked to the aniline derivatives v–z using
sodium carbonate in refluxing ethanol, (iii) yielding com-
pounds 4a–o (anilines v, x and z were commerically
available, whereas compouds w and y were prepared
according to a procedure reported by Mitchell et al. [16]).
Compounds 4a–o were hydrolyzed (iii) to give the
desired carboxylic acids 5–17. The introduction of two
Modification of the substitution pattern of the aniline
(compounds 14–17) did not result in enhanced PPAR
activity. Interestingly, pirinixic acid has also a rather low
PPARd activity, which could still be observed in com-
pound 5 with a-positioned methyl group. However, rudi-
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

Arch. Pharm. Chem. Life Sci. 2008, 341, 191–195
a-Substituted Pirinixic Acid as Dual PPARa/c Agonist
193
Table 1. In-vitro human PPAR activities of pirinixic acid derivates, determined in a luciferase reporter gene assay in Cos7 cells
described earlier [18–22].
EC50 l SD values were calculated using SigmaPlot2001 based on the mean values of at least three determinations. For values given
as approx. or more than, it was not possible to calculate exact values since activities were too low and higher concentrations were
not tolerated by the Cos7 cells.
Values in brackets give the relative activation compared to the positive control. For compounds were there is given no EC50 value,
the given value represents the highest activation observed at 100 lM or concentration indicated. Positive controls were 100 lM
WY14643 for PPARa, pioglitazone for PPARc and L165041 for PPARd each at 1 lM; ia: inactive at 100 lM or concentration indicated,
nd: no data, a) at 30 lM, b) at 20 lM, c) at 10 lM.
mental PPARd activity got lost with longer a-substituents concluded from the PPARa agonistic fibrate type of drugs
(6–13). But in compounds (14–17), which are not substi- and some of their thio-analogues e.g., GW9578, the intro-
tuted in a-position, variations of the substitution pattern duction of two methyl groups in the a-position is suffi-
at the aniline residue gave rise to a slight increase in cient to induce PPARa agonism [9]. Here, we show that a-
PPARd activitiy.
dimethyl substitution present in compound 13 is effec-
PPAR agonists have an acidic head group in common, tive to induce PPARc agonism as well. However, with
usually a carboxyl group. Typical carboxyl-type dual respect to PPARc, compound 13 is not more active than
PPARa/c agonists are substituted in a-position. It is consis- the single methyl-substituted compound 5, but PPARa
tent with our results that the introduction of substitu- activity is induced by the second methyl group more
ents in a-position to the acidic head group seems to con- than threefold compared to 5; i.e., 13 is nearly equipo-
tribute to increased PPARa and PPARc acitivity. As can be tently activating PPARa compared to the most active
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

194
O. Rau et al.
Arch. Pharm. Chem. Life Sci. 2008, 341, 191–195
compounds 8 and 9 with single butyl or hexyl substitu- Chemistry
ents. Thus, suggesting an a-positioned di-methyl substitu-
Bromo-naphtalen-1-yl-acetic acid ethyl ester 1i
tion to be a contributor for PPARa preference.
Dibenzoyl peroxide (0.031 g, 0.13 mmol) was added to a stirred
mixture of ethyl-2-naphtylethanoate (5 g, 23.34 mmol) and NBS
(4.98 g, 28 mmol) in abs. carbon tetrachloride (70 mL). The reac-
tion mixture was heated at reflux for 3.5 h, filtered and concen-
trated in vacuo. Purification by distillation at 128–1308C and
0.002 mbar yielded compound 1i (80%).
Based on 14 crystal structures, Pirard introduced a che-
mometrical approach to formally classify the ligand-
binding pockets of the PPARs into a pocket accepting the
acidic head group, a pocket proximal to the carboxyl
group-accepting substituents, which are a- or b-posi-
tioned and a linker region, accepting the lipophilic back-
bone, which is commonly represented by a b- or c-posi-
tioned aromatic system [17]. While the proximal pocket
is suggested to exhibit only minor contribution to recep-
tor-subtype selectivity, a left and a right distal pocket
seem to be major determinants of subtype selectivity.
This hypothesis is consistent with our experimental
observations in which a-substitution alone did not dis-
criminate between PPARa and PPARc agonism.
It is a well established concept that substitution in a-
position to the acidic head group has an impact on PPAR
activitiy. Here, we show for the first time that a system-
atic variation of the chain length of an a-alkyl substitu-
tion leads to gradual differences in the PPARa and PPARc
activation profile with an optimal induction of both
PPARa and PPARc agonism at a chain length of a butyl
and a hexyl residue.
2-(4,6-Dichloro-pyrimidin-2-ylsulfanyl)-2-methyl-propionic
acid ethyl ester 3k
LiHMDS (1.0 M, 11.23 mL) and HMPA (1.96 mL) were added to a
solution of the ester 3a (3 g, 11.23 mmol) in abs. THF (100 mL) at
–788C. After 30 min at –788C, methyl iodide (1.59 g,
11.23 mmol) was introduced. The reaction was allowed to warm
to rt (room temperature) over 3.5 h, then cooled to –788C before
a second equiv of LiHMDS and HMPA was added. The reaction
was allowed to warm up to rt slowly and was stirred over night
for 18 h. The resulting solution was then partitioned between
EtOAc (100 mL) and water (100 mL). The organic phase was
washed with water (100 mL), saturated NaHCO3 solution
(100 mL) and once more with water (100 mL), dried over magne-
sium sulphate, filtered, and concentrated in vacuo. Purification
by silica gel chromatography with n-hexane/ethyl acetate (25 : 1)
gave 3k as pale yellow oil (52%).
2-[4-Chloro-6-(2,3-dimethyl-phenylamino)-pyrimidin-2-
ylsulfanyl]-hexanoic acid 8
We gratefully acknowledge Mona Tawab for carefully revi-
sing the manuscript.
The authors have declared no conflict of interest.
Mp. = 120–1218C. 1H-NMR (300.13 MHz, (CD3)2SO): d 0.80–0.83
(m, 3H, CH3 (n-Bu)), 1.21–1.38 (m, 4H, CH2 (n-Bu)), 1.68–1.78 (m,
2H, CH2 (n-Bu)), 2.05 (s, 3H, Ph-2-CH3), 2.26 (s, 3H, Ph-3-CH3), 4.19 (s
(br), 1H, S-CH), 6.21 (s, 1H, Pyr-5-H), 7.06–7.10 (m, 3H, Ph-4,5,6-H),
9.51 (s (br), 1H, NH), 12.82 (s (br), 1H, COOH). 13C-NMR
(50.32 MHz, (CD3)2SO): d 13.67 (Ph-2-CH3), 14.00 (CH3 (n-Bu)), 20.07
(Ph-3-CH3), 21.60 (CH2 (n-Bu)), 28.65 (CH2 (n-Bu)), 31.71 (CH2 (n-Bu)),
46.68 (S-CH), 98.94 (Pyr-5-C), 124.02 (Ph-6-C), 125.69 (Ph-4-C),
127.85 (Ph-5-C), 132.41 (Ph-2-C), 135.63 (Ph-3-C), 137.60 (Ph-1-C),
157.73 (Pyr-4-C), 162.17 (Pyr-2-C), 169.63 (Pyr-6-C), 172.47 (COOH).
MS (ESI–): m/e = 377.9 [M – 1]. Anal. (C18H22ClN3O2S) C, H, N: Anal.
Calcd. C 56.91, H 5.84, N 11.06; Found C 57.09, H 5.86; N 10.87.
Experimental
General
All commercially available chemicals and solvents are of reagent
grade and were used without further purification unless speci-
fied otherwise. Syntheses of compounds 1i and 3k were carried
out under argon atmosphere. Melting points were determined
on a Bꢀchi-Tottoli melting point apparatus (Bꢀchi Labortechnik,
2-[4-Chloro-6-(2,3-dimethyl-phenylamino)-pyrimidin-2-
ylsulfanyl]-octanoic acid 9
1
1
Flawil, Switzerland) and are uncorrected. H-NMR and 13C-NMR
Mp. = 958C. H-NMR (300.13 MHz, (CD3)2SO): d 0.83 (t, 3H, CH3
spectras were measured in CDCl3 or DMSO-d6 on a Bruker ARX
300 (1H-NMR) and AC 200 E (13C-NMR) spectrometer (Bruker,
Rheinstetten, Germany). Chemical shifts are reported in parts
per million (ppm) using tetramethylsilane (TMS) as internal
standard. Mass spectra were obtained on a Fisons Instruments
VG Platform II spectrometer measuring in the positive or nega-
tive ion mode (ESI–MS system). Column chromatography was
carried out on Merck silica gel 60 (Merck, Darmstadt, Germany).
Elemental analysis has been performed by the Microanalytical
Laboratory of the Institute of Organic Chemistry and Chemical
Biology, University of Frankfurt, on a Foss Heraeus CHN–O–
rapid elemental analyzer (Heraeus, Hanau, Germany).
(Hex), J = 6.3 Hz), 1.15–1.22 (m, 8H, CH2 (Hex)), 1.68–1.78 (m, 2H,
CH2 (Hex)), 2.05 (s, 3H, Ph-2-CH3), 2.26 (s, 3H, Ph-3-CH3), 4.21 (t (br),
1H, S-CH), 6.21 (s, 1H, Pyr-5-H), 7.07–7.11 (m, 3H, Ph-4,5,6-H), 9.51
(s (br), 1H, NH), 12.85 (s (br), 1H, COOH). 13C-NMR (50.32 MHz,
(CD3)2SO): d 13.79 (Ph-2-CH3), 13.94 (CH3 (Hex)), 20.00 (Ph-3-CH3),
21.85 (CH2 (Hex)), 26.34 (CH2 (Hex)), 28.03 (CH2 (Hex)), 30.89 (CH2
(Hex)), 31.91 (CH2 (Hex)), 46.89 (S-CH), 98.81 (Pyr-5-C), 123.94 (Ph-6-
C), 125.62 (Ph-4-C), 127.78 (Ph-5-C), 132.30 (Ph-2-C), 135.58 (Ph-3-C),
137.53 (Ph-1-C), 157.67 (Pyr-4-C), 162.11 (Pyr-2-C), 169.58 (Pyr-6-C),
172.42 (COOH). MS (ESI–): m/e = 406.1 [M – 1]. Anal. Calcd.
(C20H26ClN3O2S) C, H, N: Calc. C 58.88, H 6.42, N 10.30; Found C
58.63, H 6.32; N 10.24.
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

Arch. Pharm. Chem. Life Sci. 2008, 341, 191–195
a-Substituted Pirinixic Acid as Dual PPARa/c Agonist
195
2-[4-Chloro-6-(2,3-dimethyl-phenylamino)-pyrimidin-2-
ylsulfanyl]-2-methyl-propionic acid 13
References
Mp. = 162–1638C.1H-NMR (300.13 MHz, (CD3)2SO): d 1.55 (s, 6H, S-
C(CH3)2), 2.05 (s, 3H, Ph-2-CH3), 2.26 (s, 3H, Ph-3-CH3), 6.08 (s, 1H,
Pyr-5-H), 7.05–7.09 (m, 3H, Ph-4,5,6-H), 9.41 (s (br), 1H, NH), 12.54
(s (br), 1H, COOH). 13C-NMR (50.32 MHz, (CD3)2SO): d 13.95 (Ph-2-
CH3), 20.08 (Ph-3-CH3), 25.97 (S-(CH3)2), 51.01 (S-C(CH3)2), 98.50 (Pyr-
5-C), 124.21 (Ph-6-C), 125.84 (Ph-4-C), 127.95 (Ph-5-C), 132.56 (Ph-2-
C), 135.67 (Ph-3-C), 137.67 (Ph-1-C), 157.42 (Pyr-4-C), 162.33 (Pyr-2-
C), 170.32 (Pyr-6-C), 174.40 (COOH). MS (ESI–): m/e = 350.1 [M – 1].
Anal. Calcd. (C16H18ClN3O2S60.4 H2O [359.07]) C, H, N: Calc. C
53.52, H 5.28, N 11.70; Found C 53.75, H 4.76; N 11.04.
In-vitro transactivation assays, cell culture and transfection,
plasmids and propagation of the luciferase assay has been
described previously [18, 19]. In short: Cos7 cells were grown in
DMEM-supplemented FCS, sodium pyruvate, penicillin/strepto-
mycine and glutamine at 10% CO2 and 378C. Cells were seeded
in 96-well plates at a density of 30000 per well the day before
transfection. Cells were transfected with Lipofectamine2000
according to the manufacturer's protocol with pFR-Luc (Strata-
gene, La Jolla, CA, USA), pRL-SV40 (Promega-Deutschland, Man-
nheim, Germany) and the respective receptor expression plas-
mid pFA-CMV-hPAPRa-LBD, pFA-CMV-hPPARd-LBD or pFA-CMV-
hPAPRc-LBD. After transfection, medium was changed to DMEM
without phenolred and FCS, containing appropriate concentra-
tions of the test compounds. Each concentration was tested in
triplicate wells and each determination was repeated at least
three times.
[1] B. Desvergne, W. Wahli, Endocr. Rev. 1999, 20, 649–688.
[2] J. P. Despres, I. Lemieux, S. J. Robins, Drugs 2004, 64,
2177–2198.
[3] S. M. Grundy, J. I. Cleeman, C. N. Merz, H. B. Brewer Jr., et
al., J. Am. Coll. Cardiol. 2004, 44, 720–732.
[4] S. E. Inzucchi, Jama 2002, 287, 360–372.
[5] J. P. Berger, T. E. Akiyama, P. T. Meinke, Trends Pharmacol.
Sci. 2005, 26, 244–251.
[6] A. Castrillo, P. Tontonoz, Annu. Rev. Cell. Dev. Biol. 2004,
20, 455–480.
[7] R. Genolet, W. Wahli, L. Michalik, Curr. Drug Targets
Inflamm. Allergy 2004, 3, 361–375.
[8] M. Ricote, J. T. Huang, J. S. Welch, C. K. Glass, J. Leukoc.
Biol. 1999, 66, 733–739.
[9] T. M. Willson, P. J. Brown, D. D. Sternbach, B. R. Henke, J.
Med. Chem. 2000, 43, 527–550.
[10] T. Miyahara, L. Schrum, R. Rippe, S. Xiong, et al., J. Biol.
Chem. 2000, 275, 35715–35722.
[11] E. Raspe, L. Madsen, A. M. Lefebvre, I. Leitersdorf, et al., J.
Lipid Res. 1999, 40, 2099–2110.
[12] R. K. Vikramadithyan, J. Hiriyan, J. Suresh, C. Gershome,
et al., Obes. Res. 2003, 11, 292–303.
[13] P. Sauerberg, J. P. Mogensen, L. Jeppesen, L. A. Svensson,
et al., Bioorg. Med. Chem. Lett. 2005, 15, 1497–1500.
Biological assays
[14] Z. F. Cai, Q. Liu, P. P. Li, Z. R. Guo, Z. F. Shen, Acta Pharma-
col Sin. 2006, 27, 597–602.
Luciferase assay
Following overnight incubation with the test compounds, luci-
ferase assay was carried out using Dual-Glo luciferase assay sys-
tem (Promega) according to the manufacturer's protocol. Lumi-
nescence was measured with a GENios Pro luminometer (Tecan
Deutschland GmbH, Crailsheim, Germany).
[15] G. d'Atri, P. Gomarasca, G. Resnati, G. Tronconi, et al., J.
Med. Chem. 1984, 27, 1621–1629.
[16] R. L. Mitchell, Y.–H. Lai, R. V. Williams, J. Org. Chem.
1979, 44, 4733–4735.
[17] B. Pirard, J. Comput. Aided Mol. Des. 2003, 17, 785–796.
[18] O. Rau, M. Wurglics, A. Paulke, J. Zitzkowski, et al., Planta
Med. 2006, 72, 881–887.
Compounds
All compounds were dissolved in DMSO and diluted 1 : 1000
upon addition to the cells. Normalisation for transfection effi-
cacy and cell growth was done by renilla luciferase data. Activa-
tion factors were obtained by dividing by DMSO control. EC50
and standard deviation (SD) values were calculated with mean
values of at least three determinations by SigmaPlot 2001 (SPSS
Inc.) using four parameter logistic regression. All active com-
pounds were evaluated to act as full agonists by comparison of
the achieved maximum effect to that of reference compounds
WY14,643 and pioglitazone.
[19] S. Ulrich, S. M. Loitsch, O. Rau, A. von Knethen, et al.,
Cancer Res. 2006, 66, 7348–7354.
[20] O. Rau, M. Wurglics, T. Dingermann, M. Abdel–Tawab,
M. Schubert–Zsilavecz, Pharmazie 2006, 61, 952–956.
[21] S. Derksen, O. Rau, P. Schneider, M. Schubert–Zsilavecz,
G. Schneider, ChemMedChem 2006, 1, 1346–1350.
[22] L. Popescu, O. Rau, J. Bottcher, Y. Syha, M. Schubert–Zsi-
lavecz, Arch Pharm. Chem. Life Sci. 2007, 340, 367–371.
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com