Hepatoselectivity of statins: Design and synthesis of 4-sulfamoyl pyrroles as HMG-CoA reductase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.11.124

4-Sulfamoyl pyrroles were designed as novel hepatoselective HMG-CoA reductase inhibitors (statins) to reduce myalgia, a statin-induced adverse effect. The compounds were prepared via a [3 + 2] cycloaddition of a Muenchnone with a sulfonamide-substituted alkyne. We identified compounds with greater selectivity for hepatocytes compared to L6-myocytes than rosuvastatin and atorvastatin. There was an inverse correlation of myocyte potencies and C log P values. A number of analogs were effective at reducing cholesterol in acute and chronic in vivo models but they lacked sufficient chronic in vivo activity to warrant further development.

Full text of DOI:10.1016/j.bmcl.2007.11.124

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1151–1156
Hepatoselectivity of statins: Design and synthesis of
4-sulfamoyl pyrroles as HMG-CoA reductase inhibitors
William K. C. Park,* Robert M. Kennedy, Scott D. Larsen, Steve Miller, Bruce D. Roth,
Yuntao Song, Bruce A. Steinbaugh, Kevin Sun, Bradley D. Tait, Mark C. Kowala,
Bharat K. Trivedi, Bruce Auerbach, Valerie Askew, Lisa Dillon, Jeffrey C. Hanselman,
Zhiwu Lin, Gina H. Lu, Andrew Robertson and Catherine Sekerke
Department of Chemistry, CVMED, Pfizer Global Research and Development, Michigan Laboratories, Ann Arbor, MI, USA
Received 27 September 2007; revised 28 November 2007; accepted 30 November 2007
Available online 5 December 2007
Abstract—4-Sulfamoyl pyrroles were designed as novel hepatoselective HMG-CoA reductase inhibitors (statins) to reduce myalgia,
a statin-induced adverse effect. The compounds were prepared via a [3 + 2] cycloaddition of a Munchnone with a sulfonamide-
¨
substituted alkyne. We identified compounds with greater selectivity for hepatocytes compared to L6-myocytes than rosuvastatin
and atorvastatin. There was an inverse correlation of myocyte potencies and ClogP values. A number of analogs were effective
at reducing cholesterol in acute and chronic in vivo models but they lacked sufficient chronic in vivo activity to warrant further
development.
Ó 2007 Elsevier Ltd. All rights reserved.
It is well documented that coronary heart disease is a lead-
ing cause of death in the US and is positively correlated
with a high plasma LDL-C level.1,2 3-Hydroxy-3-methyl-
glutaryl coenzyme A (HMG-CoA) reductase inhibitors
(statins) are the treatment of choice for lowering plasma
LDL-C. Since lovastatin was first introduced in the
1980s, other statins have made their way to the market
with improved efficacy: for example, pravastatin, simva-
statin, atorvastatin, and the latest addition, rosuvastatin.
(OATPs) since these uptake transporters are expressed
on hepatocytes but not on myocytes, and statins are
known to be viable substrates.17 By making statins more
hydrophilic, one could anticipate reducing non-selective
passive diffusion into all cells and increasing selectivity
for cells capable of internalizing statins through active
transport. However, at the design stage it is challenging
to determine the optimal hydrophilicity of novel statins.
Thus, we elected to use atorvastatin 1 as a starting point
and to evaluate the impact on tissue selectivity of alter-
ing its lipophilicity. In order to rapidly assess tissue
selectivity, we used hepatocytes and L6-myocytes as pri-
mary assays.
The most common adverse effect associated with the
statins is myalgia, manifested by muscle stiffness, muscle
weakness, fatigue, and cramps.3 It is thought that the
cause of this side effect is inhibition of myocytic
HMG-CoA reductase.4 It has also been suggested that
statins with greater hepatoselectivity may help reduce
the observed side effect as they are less available to mus-
cle tissues.5,6 One way to obtain hepatoselective statins
is to utilize organic anion transporting polypeptides
Statins have been successfully co-crystallized with
HMG-CoA reductase.18 As shown in Figure 1, salient
features of the binding of atorvastatin to the enzyme in-
clude (i) strong hydrogen bonding interactions provided
by the 3,5-dihydroxy heptanoic acid fragment of 1; (ii)
the isopropyl group fitting into a small lipophilic pocket;
and (iii) a hydrogen bonding interaction between the 4-
carboxamide oxygen and the hydroxyl group of Ser 565.
Keywords: LDL-C; HMG-CoA reductase; Atorvastatin; Rosuvastatin;
Organic anion transporting polypeptides; Hydrophilicity; Lipophilic-
ity; ClogP; L6-myocytes; Hepatocytes.
Similarly, it is known that rosuvastatin 2, one of the
most efficacious statins, forms a strong hydrogen bond
with Ser 565 through a sulfonamide oxygen, and it
*
Corresponding author. Address: P&H Therapeutics Inc. 7 Coastwind
dr. Westerly, Rhode Island, 02891. Tel.: +1 734 649 2254.; e-mail:
wkcpark@gmail.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.11.124

1152
W. K. C. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1151–1156
Br
CO2H
O
OH
c)
a) and b)
HO
HO
O
O
F
F
3
4
O
O
F
N
O
O
O
O
O
O
d) and e)
O
HN
NH
N
F
F
O
COOMe
5
COOR
Figure 1. The structure of atorvastatin 1 and its key interactions with
the HMG CoA reductase (resolution of the X-ray co-crystal structure,
6 : R =CH3
7 : R= H
O
O
O
˚
2.22 A, RCSB PDB ID: 1 hwk).
TBIA = tert-Butylisopropylidene amine
H2N
O
Scheme 2. Reagents and conditions: (a) MeOH, cat. TsOH 100%; (b)
NBS/HBr in CCl4, 65 °C for 5 h, 85%; (c) TBIA 1.1 equiv, triethyl-
amine 1.5 equiv, acetonitrile, room temperature, 89%; (d) isobutyryl
chloride 1.1 equiv, triethylamine 1.25 equiv, DCM, 90%; (e) LiOH
1.5 equiv/MeOH, room temperature, 16 h, 99% (Ref. 21).
was of interest to us to see what impact this would have
on the activity and hepatoselectivity of pyrrole-based
statins.
CO2H
HO
(R)
provided a diastereomeric mixture of 4-fluorophenylgly-
cine derivative 5.11–13 Acylation of this amine with iso-
butyryl chloride proceeded smoothly to generate the
methyl ester 6, which was hydrolyzed with methanolic
HO
(R)
F
N
N
S
LiOH to give the Munchnone precursor 7.
¨
O
N
O
The synthesis of the alkynyl tertiary sulfonamides is
shown in Scheme 3.7 Reaction of methane sulfonyl chlo-
ride with the appropriate amine gave the sulfonamide 10
in high yields. To prepare alkynyl secondary sulfona-
mides, a 2,4,6-trimethoxy benzyl protecting group for
the NH was utilized, which was installed via reductive
amination of sulfonamides 10 (Scheme 3). Proton
abstraction from 10 with n-BuLi followed by reaction
with the appropriate ester gave good yields of the b-keto
sulfonamides 11. The alkynyl sulfonamides 12 were
formed in excellent yields by reaction of 11 with 2-
chloro-N-methyl pyridinium iodide under basic
conditions.7
Rosuvastatin 2
As a general chemistry strategy we retained the 5-mem-
ber heterocyclic core of atorvastatin as a key scaffold as
well as the essential interactions within the active site of
the enzyme, and elected to investigate replacement of the
carboxamide moiety with a sulfonamide.
We envisioned that a [3 + 2] cycloaddition of two com-
ponents, a Munchnone 13 and an alkyne substituted
with sulfonamide 12 (Scheme 1), might achieve efficient
construction of a penta-substituted 4-sulfamoyl pyr-
role.14 Formation of pyrroles from the reactions of al-
¨
kynes with Munchnones has been reported with
¨
sulfonyl alkynes.8,9
As shown in Scheme 4, heating of 7 with acetic anhy-
dride in toluene formed the Munchnone 13 in situ,
¨
which reacted with the alkynes 12 to afford the pyrroles
14 in good yields. Theoretically, two sulfamoyl pyrrole
regioisomers are possible during the cycloaddition. Only
the desired regioisomer, with the sulfonamide at the 4-
position, was formed.20 As evidenced in an X-ray crystal
structure of 16a (Fig. 2), the observed regioselectivity
can be rationalized by the preferred orientation of the
The synthesis of the Munchnone precursor 7 is shown in
¨
Scheme 2. 4-Fluorophenylacetic acid 3 was esterified to
produce the corresponding ester followed by bromina-
tion using NBS under acidic condition to afford the a-
bromo ester 4. A subsequent SN2 reaction with TBIA15
alkyne sulfonamide 12 in relation to the Munchnone 7
¨
CO2t-Bu
prior to the cycloaddition. Treatment of 14 with trifluo-
roacetic acid removed the trimethoxybenzyl group from
the sulfonamide nitrogen, cleaved the tert-butyl esters
and acetonide, affording the lactones 15. Finally, the lac-
tone was treated with one equivalent of 1 N NaOH and
lyophilized to yield the pyrrole sodium salts 16.
O
CO2t-Bu
O
O
O
N
F
N+
13
F
O
O
SO2NR1R2
SO2NR1R2
R3
As shown in Figure 2, an X-ray co-crystal structure of
16a and the enzyme confirmed the desired regioisomer.
It is noteworthy that the compound uses a water mole-
cule as a hydrogen bonding bridge to Ser 565. The
bridging water molecule appears to be a general feature
R3
12
14
Scheme 1. A [3 + 2] cycloaddition approach for preparing 4-sulfamoyl
pyrroles.

W. K. C. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1151–1156
1153
9
O
O
O
O
O
O
NR1R2
O
H
b)
S
S
S
NR1R2
Cl
NR1R2
a)
8
10
R3
11
b)
c)
OMe
O
O
O
O
O
O
S
d)
R3
S
S
N
H
NR1R2
N
R1
R1
10'
10"
12
OMe
MeO
Scheme 3. Reagents and conditions: (a) triethylamine/DCM, 0 °C, 4 h, 88%; (b) 1.1 equiv n-butyl lithium/THF, À78 °C, 2 h, followed by
p-substituted benzoic acid methyl ester, 78%; (c) 2-chloro-N-methylpyridinium iodide 1.5 equiv, triethylamine 2.5 equiv/DCM, room temperature,
16 h, 75%; (d) 2,4,6-trimethoxybenzaldehyde 1.1 equiv, NaBH4, THF at room temperature, 4 h, 89% (Ref. 21).
CO2-Na+
CO2 t-Bu
HO
O
O
HO
HO
O
O
N
F
F
F
b)
N
N
c)
a)
7 + 12
SO2NR1R2
SO2NR1R2
SO2NR1R2
R3
R3
R3
14
16
15
Scheme 4. Reagents and conditions: (a) acetic anhydride, toluene at 60 °C, 5 h, 60–80%; (b) 30 (v/v) % TFA/DCM, room temperature, 2 h,
quantitative; (c) 1 equiv of 1 N NaOH/THF, room temperature, 4 h, quantitative (Ref. 21).
CO2H
The ratio of Cmax/L6 value is more consistent with the
incidence of myalgia.
HO
HO
(R)
(R)
Examples of the 4-sulfamoyl pyrrole analogs are shown
in an order of hepatoselectivity in Table 2. In general,
most of the analogs exhibited excellent potency
in vitro and selectivity toward hepatocyte over myocyte,
particularly, the N-cycloalkyl sulfonamides, that is, 16a,
16b, 16c, and 16d. The ring size seemed to have little ef-
fect on the selectivity until the azetidine sulfonamide,
16k. Interestingly, the rotation-restricted analog 16q
showed no markedly better in vitro potency than the
N-benzyl sulfonamide analog, 16h. In addition, it was
observed that the secondary sulfonamide 16h was more
hepatoselective than 16q or the N-methyl benzyl sulfon-
amide analogs, 16r and 16s. The meta-substituted phe-
nyl sulfonamides, that is, 16e, 16f, and 16g were more
hepatoselective than the para-substituted analogs, that
is, 16o and 16p. It was interesting to see that the phenyl
sulfonamide analogs 16t, 16u, and 16v showed the poor-
est hepatoselectivity as these analogs are structurally the
closest to atorvastatin 1. The observed low selectivity
was consistent with the higher ClogP values (5.95,
F
N
O
Ser 565
S
O
N
O
Figure 2. The X-ray crystal structure of 16a bound to HMG-CoA
reductase (PDB ID code 3BGL).
of binding between the sulfonamides and the enzyme,
which differs from both atorvastatin and rosuvastatin.
Table 1 contains selectivity data of some marketed stat-
ins. The highest hepatoselectivity is found in rosuvasta-
tin (926) and the lowest selectivity in cerivastatin (4.1).
The percent myalgia incidence correlates with the
hepatoselectivity with the exception of rosuvastatin.
Table 1. A correlation of myalgia frequency and the ratio of Cmax/ L6 myocyte
a
Statin
Hep (nM)
L6-Myo (nM)
Selectivity
Cmax
Cmax/L6
Myalgia incidence
Cerivastatin
Simvastatin
Rosuvastatin
Pravastatin
1.7
1.3
7
150
4.1
115
926
25
120
75
3.6
Withdrawn
18.2%b
0.80
0.30
0.01
0.27
17
250
7550
High dose limited
10.9%b
444
100
a Human Cmax at highest available dose (nM).
b Percent values are quoted from the PRIMO study results.10

1154
W. K. C. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1151–1156
Table 2. Biological activity of 3-sulfamoylpyrroles 16 in order of hepatoselectivity
HO
(R)
CO2-Na+
OH
R )
(
F
N
SO2NR1R2
R3
16
16
–NR1R2
R3
H
In vitroa IC50 (nM) Hepb IC50 (nM) L6 Myoc IC50 (nM) Selectivityd MAICSe (%) ClogPf
N
O
16a
4.5
2.9
2.6
2.0
0.80
0.67
0.5
10,000
7000
12,469
10,479
10,400
9167
À55.6
À56.0
À75.9
À61.3
4.00
5.24
4.69
5.24
N
16b
16c
16d
H
H
H
5200
N
2.4
22,000
H
N
16e
16f
F
F
18
5
36,000
10,000
7340
7299
ND
ND
4.96
3.51
H2NOC
H
N
CONH
2
1.0
9.4
1.4
H
N
OH
16g
16h
F
F
0.4
2500
2200
6250
5945
À44.2
À58.6
4.18
5.86
N
H
13
0.37
N
16i
16j
16k
H
H
H
2.7
6.5
1.8
3.0
10,500
1380
3559
3066
1954
ND
2.17
4.05
4.12
N
–NMe2
0.45
1.1
ND
N
2150
À34.8
16l
16m
–NMe2
–NHMe
F
F
6.1
0.75
0.22
1400
ND
1866
ND
À58.8
À33.0
À15.0
4.19
4.10
27
N
S
16n
16o
F
9.7
4.9
3.3
5700
1748
4.88
H
N
F
2.7
4380
1622
À15.0
3.99
SO2NH2
H
N
16p
16q
F
F
5.8
8.4
1.8
2.5
2340
2500
1300
984
À34.4
4.33
6.25
CONH2
N
ND
Rosuvastatin
16r
3.7
F
0.27
8.0
250
0.69
142
926
550
À82.4
797
1.90
ND
4.46
N
N
5.97
Atorvastatin
16s
3.8
H
0.99
8.5
144
140
À75.2
566
0.25
0.47
0.96
ND
5.82
5.95
5.76
H
N
16t
F
F
15
60
45
127
45
À51.8
ND
F
H
N
16u
2.0
(continued on next page)

W. K. C. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1151–1156
1155
Table 2 (continued)
16
–NR1R2
R3
H
In vitroa IC50 (nM)
9.0
Hepb IC50 (nM)
0.57
L6 Myoc IC50 (nM)
20
Selectivityd
34
MAICSe (%)
ClogPf
5.32
H
N
16v
À49.2
a Rat microsomal HMGCoA reductase inhibition.
b Cholesterol-synthesis inhibition in hepatocyte.
c Cholesterol-synthesis inhibition in L6 myocyte.
d Ratio of hepatocyte IC50/L6 IC50
e Acute % cholesterol-synthesis inhibition at 1 mpk in mouse; for in vitro assay procedures, see Ref. 16.
.
f ClogP is a calculated hydrophilicity of a compound. It is defined as the logarithm of its partition coefficient between n-octanol and water
log(Coctanol/Cwater). The ClogP values were calculated based on the carboxylic acid form instead of the sodium salt for a comparison purpose to
other statins.
12000
10000
8000
were less efficacious in reducing LDL-C (16a À11%,
16c À38%, 16l À30%) than rosuvastatin (À62%).
In conclusion, we have designed, prepared, and
tested 4-sulfamoyl pyrrole analogs as novel inhibitors
of HMG-CoA reductase. Some of the most potent
analogs showed excellent selectivity (superior to rosu-
vastatin) toward hepatocytes over myocytes. It was
observed that the L6-myocytic potency is inversely
correlated with the ClogP of the analog. Although
some analogs effectively inhibited cholesterol synthe-
sis acutely in a mouse model and lowered LDL-C
chronically, none of the compounds had efficacy
greater than or equal to that of atorvastatin or
rosuvastatin.
6000
4000
2000
0
5.5
2
2.5
3
4
4.5
5
6
3.5
ClogP
Figure 3. A plot of L6-myocyte potency versus ClogP of 4-sulfamoyl
pyrrole analogs. The red filled triangles represent the analogs of
R3 = F, and the blue triangles represent the analogs R3 = H. The
commercially available statins are also shown: gray (rosuvastatin),
green (simvastatin), yellow (pravastatin), and transparent
(atorvastatin).
Acknowledgments
W.K.C. Park is grateful to Drs. Ronald W. Sarver and
Barry C. Finzel at Pfizer for their help in deposition of
X-ray cocrystal structure of PF-00335759 and HMG-
CoA reductase to PDB.
5.76, and 5.32, respectively) compared to that of atorva-
statin (4.46) as shown in Table 2.
A plot of L6 myocyte IC50 vs ClogP is presented in Fig-
ure 3. In general, the L6-myocyte potency showed an in-
verse correlation to the ClogP values. Interestingly, the
majority of the 4-sulfamoyl pyrrole analogs possess
higher myocyte IC50s than rosuvastatin, despite having
higher ClogPs and similar hepatocyte IC50s. This sug-
gests that the observed inverse correlation of L6 IC50
and ClogP likely holds only within a specific structural
class. Based on the data presented in Table 1 for mar-
keted statins, we would expect these new sulfamoylpyr-
roles to have reduced potential for myalgia relative to
current therapies.
References and notes
1. Pedersen, T. R.; Kjekshus, J.; Pyo¨ra¨la¨, K.; Olsson, A. G.;
Cook, T. J.; Musliner, T. A.; Tobert, J. A.; Haghfelt, T.
Am. J. Cardiol. 1998, 81, 333.
2. Tobert, J. A. Nat. Rev. Drug Discov. 2003, 2, 517.
3. Abstract of papers, 5th Annual Conference On Arterio-
sclerosis, Thrombosis and Vascular Biology, May 6–8,
2004; P474.
4. Flint, O. P.; Masters, B. A.; Gregg, R. E.; Durham, S. K.
Toxicol. Appl. Pharmacol. 1997, 145, 91.
5. Tirona, R. G.; Leake, B. F.; Merino, G.; Kim, R. B.
J. Biol. Chem. 2001, 276, 35669.
6. Masters, B. A.; Palmoski, M. J.; Flint, O. P.; Gregg, R. E.;
Wangiverson, D.; Durham, S. K. Toxicol. Appl. Pharma-
col. 1995, 131, 163.
7. Leclercq, M.; Brienne, M. J. Tetrahedron Lett. 1990, 31,
3875.
8. Padwa, A.; Burgess, E. M.; Gingrich, H. L.; Roush, D. M.
J. Org. Chem. 1982, 47, 786.
9. Croce, P. D.; Fariboldi, P.; La Rosa, C. J. Heterocycl.
Chem. 1987, 24, 1793.
In general, cycloalkyl sulfonamide analogs (e.g., 16a,
16b, 16c, and 16d) showed high in vivo acute choles-
terol-synthesis inhibition (MAICS). Incidentally, these
analogs were also highly hepatoselective (ca. > 10,000).
Several analogs exceeded 50% reduction in MAICS as
seen in column (e) in Table 2.
The best analogs were run in the chronic in vivo hamster
efficacy model.19 Unfortunately, the best compounds
10. Buckert, E.; Hayem, G.; Dejager, S.; Yau, C.; Begaud, B.
Cardiovasc. Drugs Ther. 2005, 19, 403.

1156
W. K. C. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1151–1156
11. Sliskovic, D. R.; Roth, B. D.; Wilson, M. W.; Hoefle, M.
L.; Newton, R. S. J. Med. Chem. 1990, 33, 31.
12. Miyachi, N.; Yanagawa, Y.; Iwasaki, H.; Ohara, Y.;
Hiyama, T. Tetrahedron Lett. 1993, 34, 8267.
13. Bertolini, G.; Casagrande, C.; Norcini, G.; Santangelo, F.
Synth. Commun. 1994, 24, 1833.
14. A similar approach was described in the following
Hutchings, R; Park, W. K. C.; Askew, V.; Dillon, L.,
Hanselman, J. C.; Lin, Z.; Lu, G. H.; Robertson, A;
Sekerke, C.; Harris, M. S.; Pavlovsky, A.; Bainbridge, G.;
Caspers N. Bioorg. Med. Chem. Lett., submitted for
publication.
17. Yoshihisa, S.; Masaru, H.; Hitoshi, S.; Yuichi, S.
J. Pharmacol. Exp. Ther. 2004, 311, 228.
´
literature reference Castro, J.; Coteron, J. M.; Fraile, M.
T.; Carcıa-Ochoa, S.; de la Heras, F. G.; Martın-Cuesta,
18. Tabernero, L.; Rowell, V. W.; Stauffacher, C. V. J. Biol.
Chem. 2003, 278, 19933.
´
´
A. Tetrahedron Lett. 2002, 43, 1851.
19. For description of the chronic hamster model used in the
study Spady, D. K.; Turley, S. D.; Dietschy, J. M. J. Lipid
Res. 1985, 26, 465.
20. A typical example of the regioselectivity is shown in the X-
ray co-crystal structure of 16a and the HMG CoA
reductase in Figure 2.
21. Experimental and corresponding data are placed in
supporting information section as well as in US Patent
No. 7250444 B2.
15. Brower, P. L.; Butler, D. E.; Deering, C. F.; Le, T. V.;
Millar, A.; Nanninga, T. N.; Roth, B. D. Tetrahedron
Lett. 1992, 33.
16. (a) Bratton, L. D.; Auerbach, B.; Choi, C.; Dillon, L.;
Hanselman, J. C.; Larsen, S. D.; Lu, G.; Olsen, K.;
Pfefferkorn, J. A.; Robertson, A.; Sekerke, C.; Trivedi, B.
K.; Unangst, P. C. Bioorg. Med. Chem. 2007, 15, 5576; (b)
Pfefferkorn, J. A.; Choi, C; Larsen, S. D.; Auerbach, B.;