Novel potent apoA-I peptide mimetics that stimulate cholesterol efflux and pre-β particle formation in vitro

Article abstract of DOI:10.1016/j.bmcl.2009.10.128

Reverse cholesterol transport (RCT) is believed to be the primary mechanism by which HDL and its major protein apoA-I protect against atherosclerosis. Starting from the inactive 22-amino acid peptide representing the consensus sequence of the class A amphipathic helical repeats of apoA-I, we designed novel peptides able to mobilize cholesterol from macrophages in vitro, and to stimulate the formation of 'nascent HDL' particles, with potency comparable to the entire apoA-I protein.

Full text of DOI:10.1016/j.bmcl.2009.10.128

Bioorganic & Medicinal Chemistry Letters 20 (2010) 236–239
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel potent apoA-I peptide mimetics that stimulate cholesterol efflux
and pre-b particle formation in vitro
b
a
b
a
Raffaele Ingenito a, , Charlotte Burton , Annunziata Langella , Xun Chen , Karolina Zytko ,
*
Antonello Pessi a, Jun Wang b, , Elisabetta Bianchi a
a IRBM P. Angeletti, Merck Research Laboratories Peptide Centre of Excellence, via Pontina km 30,600, 00040 Pomezia, Italy
b Merck Research Laboratories, Department of Cardiovascular Diseases, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Reverse cholesterol transport (RCT) is believed to be the primary mechanism by which HDL and its major
protein apoA-I protect against atherosclerosis. Starting from the inactive 22-amino acid peptide repre-
senting the consensus sequence of the class A amphipathic helical repeats of apoA-I, we designed novel
peptides able to mobilize cholesterol from macrophages in vitro, and to stimulate the formation of ‘nas-
cent HDL’ particles, with potency comparable to the entire apoA-I protein.
Received 10 September 2009
Revised 27 October 2009
Accepted 28 October 2009
Available online 30 October 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Peptide
Peptide mimetics
apoA-I
HDL (high-density lipoproteins)
Pre-b-HDL
Atherosclerosis
High-density lipoproteins (HDL) protect against atherosclerosis
by multiple mechanisms1 one of which is considered to be reverse
cholesterol transport (RCT), a process that involves the efflux of
cholesterol from arterial macrophages to HDL, ultimately resulting
in elimination of excess cholesterol to the liver for excretion in the
bile and feces.2 RCT is a regulated process promoted by specific
transporters in vivo,3 such as the ATP-binding cassette A1 (ABCA1),
that mediates cholesterol efflux from macrophages to the main
protein component of HDL, the apolipoprotein A-I (apoA-I), to gen-
erate nascent HDL particles.4 The anti-atherogenic effects of apoA-I
have been documented in animal models5 and in humans,6 and ef-
forts are being devoted to develop treatments based on direct infu-
sion of apoA-I, or pharmacological agents that can upregulate
apoA-I production.7 Due to difficulties in manufacturing apoA-I
at the scale required for therapeutic intervention, one alternative,
promising approach is to develop apoA-I peptide mimetics.8
apoA-I is composed of multiple repeats of a unique secondary
at the interface between the hydrophilic and the hydrophobic face.
Designed peptides with amphipathic helical character or peptides
based on more than one repeat of apoA-I have been shown to pro-
mote cholesterol efflux from cells and to have antioxidant and anti-
inflammatory effects.10 One of these peptides, D-4F, was shown to
have anti-atherogenic effect in animal models11 and has advanced
to human clinical trials.
Our approach to develop apoA-I peptide mimetics was to start
from the 22-mer ‘apoA-I consensus sequence’ (apoA-Icons, Fig. 1
and compound 1, Table 1) based on the tandem repeat sequences
of apoA-I.12 When tested in an in vitro cholesterol efflux assay,
apoA-Icons (1) showed no ability to promote cholesterol efflux from
macrophages (Table 1). Therefore we explored a series of substitu-
tions aimed at (i) stabilizing the class A a-helical motif, (ii) modu-
lating lipophilicity, and (iii) modulating the negative charge
density. The peptides were tested in an in vitro cholesterol efflux
assay and screened for cytotoxicity in macrophages (RAW cell)
and red blood cells.
structural element defined as the class A amphipathic
In this helix, negatively charged amino acid residues are clustered
a
-helix.9
apoA-Icons has a negative charged residue at position 13 (E13
,
at the center of the polar face, while positively charged residues are
black arrow in Fig. 1) which occurs on the hydrophobic face of
the helix. To test the hypothesis that increasing the hydrophobic
character in the context of stabilizing the
a-helical conformation
might promote cholesterol efflux from macrophages, E13 was
* Corresponding author. Tel.: +39 06 91093582; fax: +39 06 91093654.
E-mail addresses: raffaele_ingenito@merck.com, raffaele.ingenito@gmail.com (R.
Ingenito).
Present address: Sundia MediTech Company, Zhangjiang Hightech Park, Shanghai
201203, China.
substituted with a-aminoisobutyric acid (Aib).
The analog with the substitution E13?Aib (peptide 2, Table 1)
was unable to stimulate cholesterol efflux in vitro. However the
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.128

R. Ingenito et al. / Bioorg. Med. Chem. Lett. 20 (2010) 236–239
237
analog with the double substitutions Aib13/Aib16 (E13?Aib and
A16?Aib (3) showed some activity with an EC50 = 23
M (Table 1).
1
6
9
13
22
P V L D E F R E K L N E E L E A1L7 K Q K L K
l
Surprisingly, the introduction of just one single bulky hydro-
phobic residue such as hexafluoroleucine (hF-Leu) in position 13
brought a large beneficial effect on potency (peptide 4, Table 1),
Q19
K
E12
E5
L
12
P1
AA
16
16
K
22
1
Q
E
E8
K9
V2
K
L
with EC50 = 2.3 lM.
K20
E15
18
15
A
17
apo AI
Consensus
E
To maintain the total negative charge density of the consensus
sequence and compensate the loss of the carboxylate at position
13, the substitution E13?Aib13 was combined with the introduc-
K22
L
D4
N11
E1133
F6
E
E
13
N
L9
c ,
-carboxyglutamic acid (Gla) either in place of E8 or E16
18
tion of
K
K
K
L17
17
E
R7
L
L10
10
R
both on the hydrophilic face of the helix.
14
L14 L3
3
F
Both analogs Gla8/Aib13 and Gla15/Aib13, peptides 5 and 6 (Table
E
D
L21
6
1), showed a gain in potency compared to the single-substituted
5
L
V
Aib13 analog (2), with EC50 = 44 and 41
lM, respectively. The effect
Helical wheel diagrams
P
on cholesterol efflux activity depended on both the Gla and Aib res-
idues, since the corresponding control peptides with Gla/Ala at the
same positions, peptides 7 and 8, were less potent, with EC50 = 142
1
Helnet/Snorkel diagrams
Figure 1. Helical wheel (left) and Helnet/Snorkel (right) diagrams of apoA-Icons. The
sequence of the consensus peptide is given above the helical wheel and helical net
diagrams. Color code: red, acidic; blue, basic; green, hydrophobic; bold black, Asn
and Gln. The negative charged E13 at position 13 (arrow) occurs on the hydrophobic
face of the helix.
and 234 lM, respectively (Table 1).
However, all peptides containing Gla (5–8) were cytotoxic, indi-
cating that their efflux activity could be due to disruption of the
membrane integrity.
To further explore the effect of altering hydrophobicity, we
introduced long-chain (C5–7) hydrocarbon amino acids containing
olefin-bearing tethers, in either R- or S-configuration (R5H S5H
,
Table 1
R6H, S7H, Fig. 2) at specific positions of the consensus sequence. In
addition to increasing the hydrophobic character of the peptide,
these residues offer the possibility of stabilizing its helical confor-
mation by formation of a cyclic, stapled structure between residues
In vitro cholesterol efflux from macrophages and cytotoxicity of compounds 1–8
Compd Sequence
CEa EC50
M)
Cytotoxicityb,c
b
(l
IC50
(l
M)
1
2
3
4
5
6
7
8
ApoA-Icons
PVLDEFREKLNEAibLEALKQKLK
na
6%@100
40
on the same face of the
a
-helix (positions i À i + 4 or i À i + 7).
This can be accomplished by ring-closing metathesis13 (RCM,
Fig. 2). For this purpose the fully elongated resin-bound peptide
was exposed to a ruthenium catalyst that promotes cross-linking
of the alkenyl chains through olefin metathesis,14 thereby forming
an all-hydrocarbon macrocyclic cross-link.
>160
>160
>160
75
75
40
PVLDEFREKLNEAibLEAibLKQKLK 23
PVLDEFREKLNEhFLLEALKQKLK 2.3
PVLDEFRGlaKLNEAibLEALKQKLK 44
PVLDEFREKLNEAibLGlaALKQKLK 41
PVLDEFRGlaKLNEALEALKQKLK
PVLDEFREKLNEALGlaALKQKLK
142
234
75
The hydrocarbon bridge was engineered on the hydrophobic
a
In vitro cholesterol efflux from RAW cells. Values are means of three experi-
face of the helix of the consensus sequence, and position 13 (E13
)
ments (na = not active).
was kept constant for replacement, based on previous results.
We initially formed an eight-member (i À i + 4) ring, by linking
amino acids at position 9–13 (K9/E13) (peptides 10 and 12, Table
b
RBC lysis assay.
No cytotoxicity up to 160
c
l
M in macrophage RAW cells.
1
6
9
P V L D E F R E K L N E1E3 L E A1L7 K Q K L 2K2
A
C
B
K
K
K
L
L
L
22
K
22
22
K
K
Q
Sn
Q
O
R
Q
K
K
K
L
H N
S
)
(
2
L
L
OH
18
A
18
E
18
17
A
A
E
17
17
E
L
n=1,2,3
L
L
S7H
S7H
E
E
E
E
13
N
13
13
N
L
9
N
L
L
9
9
K
K
K
E
E
E
9
R
Ruthenium-catalyzed olefin metathesis
9
9
R
R
Rn
O
R6H
R6H
F
E
E
E
H2N(R)
R
6
D
6
OH
6
D
D
L
L
L
V
V
V
P
n=1,2,3
P
P
1
1
1
R = H, Me
HELNET/SNORKEL analysis
6
13
Figure 2. apoA-Icons (A), with the olefinic amino acids R6H and S7H before (B), and after (C) RCM to form the hydrocarbon bridge (‘staple’). The sequence of apoA-Icons is
given above. Color code: red, acidic; blue, basic; green, hydrophobic; bold black, Asn and Gln.

238
R. Ingenito et al. / Bioorg. Med. Chem. Lett. 20 (2010) 236–239
2) or 13–17 (E13/L17) (peptide 14, Table 2) through RCM of C5-olefin
amino acids of both R- and S-configuration (R5H, S5H; Fig. 2). Both
the open and the stapled peptides were tested for their ability to
promote cholesterol efflux in vitro (Table 2). The stapled peptide
obtained with R5H between K9 and E13 (peptide 10, Table 2) showed
Circular dichroism spectroscopy studies revealed that the 22-
mer apoA-Icons peptide displays about 36% helicity in solution. Sub-
stitution of the long hydrocarbon chain amino acids R6H and S7H at
positions 6 and 13 of the sequence resulted in a large increase in
the helical content, which was essentially the same for the open
and stapled forms (62% and 65%, respectively, Fig. 3). The lack of
additional helical stabilization by stapling offers an explanation
for the finding that 16 is not more potent than 15. Overall optimal
cholesterol efflux activity depends on a balance of factors including
position, side-chain length, and stereochemistry of the olefinic
amino acids.
good cholesterol efflux capacity with an EC50 = 10 lM, while the
corresponding open precursor (peptide 9, Table 2) was also active,
albeit about fivefold less potent, hinting at the importance of struc-
tural stabilization in addition to increased hydrophobicity.
We found that the
D-configuration of the C5-olefinic amino acid
(R5H) was preferred over the
L-configuration (S5H), for both the
open and the stapled form. Furthermore, the position of the olefinic
amino acids was critical for activity, since substitution of R5H at
positions E13/L17 yielded inactive peptides, both in the open and
the stapled form (Table 2).
90000
ApoA1cons PVLDE-F-REKLNE-E-LEALKQKLK
fH= 36.3%
15
16
PVLDE-R6H-REKLNE-S7H-LEALKQKLK
fH= 62.0%
fH= 65.0%
We then explored a larger ring size (i À i + 7) with C6–7 olefin
amino acids of both L- and D-configuration. We found that the cho-
PVLDE-R6H-REKLNE-S7H-LEALKQKLK
n=11
lesterol efflux capacity was increased, particularly with C6-olefin
(R6H) at F6 and C7-olefin (S7H) at E13 and vice versa (peptides 15–
17, Table 3).
50000
50000
Interestingly, the open sequences 15 and 17, where the posi-
tions of R6H and S7H are exchanged, showed the same activity as
the corresponding stapled peptide 16 (Table 3). We were unable
to produce the stapled form of peptide 17 in acceptable yields.
The potency shown by peptides 15–17 in the cholesterol efflux as-
say is the same as the peptide mimetic D-4F (18) and importantly,
only eightfold lower than the full-length apoA-I protein (19) (Table
3). Further studies are ongoing to clarify the preferred configura-
tion at each position.
0
190
200
220
240
250
-30000
-30000
Wavelength [nm]
Figure 3. Circular dichroism analysis of peptides 15 and 16 in phosphate buffer, in
comparison with apoA-Icons. The percent helicity was calculated according to Chen
et al.16
Table 2
In vitro cholesterol efflux from macrophages and cytotoxicity of compounds 9–14
Compd Sequencea
CEb EC50
M)
Cytotoxicityc,d
c
(
l
IC50
(
lM)
Cpd 15 (0.25mg/ml) incorporated into WT mouse HDL
1
9
ApoAIcons
na
46
10
7%@30
40
PVLDEFRER5HLNER5HLEALKQKLK
PVLDEFRER5H*LNER5H*LEALKQKLK
PVLDEFRES5HLNES5HLEALKQKLK
PVLDEFRESS5H*LNES5H*LEALKQKLK NA
PVLDEFREKLNER5HLEAR5HKQKLK
>160
>160
>160
>160
>160
>160
HDL
12
A
4
10
11
12
13
14
15
10
Total cholesterol
3
8
3%@30
6
4
2
0
2
1
0
PVLDEFREKLNER5H*LEAR5H*KQKLK 2%@30
a
VLDL
The asterisk indicates the presence of a side-chain to side-chain ring (stapling)
between the two residues.
Cholesterol efflux from RAW cells. Values are means of three experiments
(na = not active).
LDL
b
1
11
21
31
41
51
61
71
c
RBC lysis assay.
No cytotoxicity up to 160 lM in macrophage RAW cells.
Eluted Fractions
d
30
20
B
Table 3
In vitro cholesterol efflux from macrophages and cytotoxicity of compounds 1, 15–18,
and the apoA-I protein (19)
15
D-4F
Compd Sequencea
CEb EC50
M)
Cytotoxicityc,d
c
(
l
IC50
(lM)
10
0
1
15
16
17
18
19
ApoA-Icons
PVLDER6HREKLNES7HLEALKQKLK
PVLDER6H*REKLNES7H*LEALKQKLK 2.0 0.3
PVLDES7HREKLNER6HLEALKQKLK
dwfkafydkvaekfkeaf
h-ApoA-I
na
na
0.8 0.25 >160
>160
1.3 0.20 >160
2.0 0.25 75
ApoA-Icons
100 200 300 400 500 600
0
0.1 0.01 >160
[Peptide] (µg/ml)
a
The asterisk indicates the presence of a side-chain to side-chain ring (stapling)
between the two residues; lower case indicates ( )-amino acids.
D
Figure 4. (A) Incubation of peptide 15 with mouse plasma at 37 °C for 2 h in vitro
followed by size exclusion chromatography lipoprotein analysis; in red peptide
distribution among the lipoprotein fractions (HDL, LDL, and VLDL). (B) Titration
curves (ELISA with pre-b-Ab) for peptide 15, consensus sequence and D-4F
incubated with human plasma.
b
Cholesterol efflux from RAW cells. Values are means of three experiments
(na = not active).
c
RBC lysis assay.
No cytotoxicity up to 160 lM in macrophage RAW cells.
d

R. Ingenito et al. / Bioorg. Med. Chem. Lett. 20 (2010) 236–239
239
Having identified peptide 15 as the most potent analog, we fur-
References and notes
ther explored its mechanism of action. Pre-b-HDL particles (‘nas-
cent HDL’) have been shown to play a key role in RCT5, and D-4F
in particular has been shown to cause an increase in pre-b-HDL.5
When peptides 1 and 15 were incubated in plasma for 2 h at
37 °C, only peptide 15 was selectively incorporated into the HDL
lipoprotein fraction (as monitored by mass-spectrometry of each
fraction after size exclusion chromatography lipoprotein analysis,
Fig. 4).
Moreover, similarly to D-4F, peptide 15 induced an increase in
the level of pre-b-HDL particles as revealed by size exclusion chro-
matography analysis followed by ELISA with an apoA-I pre-b-spe-
cific antibody15 (Fig. 4).
In conclusion we have shown here that by using as a starting
point the consensus sequence of apoA-I, it is possible to design
apoA-I peptide mimetics able to mobilize cholesterol from macro-
phages in an ABCA1-dependent pathway, and to promote the for-
mation of ‘nascent HDL’ particles, with potency superior to other
peptide mimetics in development and, importantly, comparable
to the apoA-I protein, which is known to be efficacious in humans.
1. Linsel-Nitschke, P.; Tall, A. R. Nat. Rev. Drug Disc. 2005, 4, 193.
2. Rader, D. J. J. Clin. Invest. 2006, 116, 3090.
3. Wang, X.; Collins, H. L.; Ranalletta, M.; Fuki, I. V.; Billheimer, J. T.; Rothblat, G.
H.; Tall, A. R.; Rader, D. J. J. Clin. Invest. 2007, 117, 2216.
4. Wang, N.; Silver, D. L.; Costet, P.; Tall, A. R. J. Biol. Chem. 2000, 275, 33053.
5. (a) Rubin, E. M.; Krauss, R. M.; Spangler, E. A.; Verstuyft, J. G.; Clift, S. M. Nature
1991, 353, 265; (b) Plump, A. S.; Scott, C. J.; Breslow, J. L. Proc. Natl. Acad. Sci.
U.S.A. 1994, 91, 9607.
6. (a) Nissen, S. E.; Tsunoda, T.; Tuzcu, E. M.; Schoenhagen, P.; Cooper, C. J.; Yasin,
M.; Eaton, G. M.; Lauer, M. A.; Sheldon, W. S.; Grines, C. L.; Halpern, S.; Crowe,
T.; Blankenship, J. C.; Kerensky, R. JAMA 2003, 290, 2292; (b) Tardif, J. C.;
Heinonen, T.; Noble, S. Curr. Atheroscler. Rep. 2009, 11, 58.
7. (a) Plump, A. S.; Scott, C. J.; Breslow, J. L. Proc. Natl. Acad. Sci. U.S.A. 1994, 91,
9607; (b) Berthou, L.; Staels, B.; Saldicco, I.; Berthelot, K.; J Casey, J.; Fruchart, J.
C.; Denefle, P.; Branellec, D. Arterioscler. Thromb. Vasc. Biol. 1994, 14, 1657; (c)
Pastore, L.; Belalcazar, L. M.; Oka, K.; Cela, R.; Lee, B.; Chan, L.; Beaudet, A. L.
Gene 2004, 327, 153.
8. (a) Navab, M.; Anantharamaiah, G. M.; Reddy, S. T.; Van Lenten, B. J.; Datta, G.;
Garber, D.; Fogelman, A. M. Curr. Opin. Lipidol. 2004, 15, 645; (b) Shah, P. K.;
Chyu, K. Y. Trends Cardiovasc. Med. 2005, 15, 291.
9. (a) Segrest, J. P. FEBS Lett. 1974, 38, 247; (b) Segrest, J. P.; Jones, M. K.; De Loof,
H.; Brouillette, C. G.; Venkatachalapathi, Y. V.; Anantharamaiah, G. M. J. Lipid
Res. 1992, 33, 141; (c) Segrest, J. P.; Li, L.; Anantharamaiah, G. M.; Harvey, S. C.;
Liadaki, K. N.; Zannis, V. Curr. Opin. Lipidol. 2000, 11, 105; (d) Shah, P. K.; Kaul,
S.; Nilsson, J.; Cercek, B. Circulation 2001, 104, 2498.
Key to our success is the use of
a-alkenyl amino acids at suit-
10. (a) Datta, G.; Chaddha, M.; Hama, S.; Navab, M.; Fogelman, A. M.; Garber, D. W.;
Mishra, V. K.; Epand, R. M.; Epand, R. F.; Lund-Katz, S.; Phillips, M. C.; Segrest, J.
P.; Anantharamaiah, G. M. J. Lipid Res. 2001, 42, 1096; (b) Remaley, A. T.;
Thomas, F.; Stonik, J. A.; Demosky, S. J.; Bark, S. E.; Neufeld, E. B.; Bocharov, A.
V.; Vishnyakova, T. G.; Patterson, A. P.; Eggerman, T. L.; Santamarina-Fojo, S.;
Brewer, H. B. J. Lipid Res. 2003, 44, 828; (c) Navab, M.; Anantharamaiah, G. M.;
Srinivasa, T. R.; Hama, S.; Hough, G.; Grijalva, V. R.; Wagner, A. C.; Frank, J. S.;
Datta, G.; Garber, D.; Fogelman, A. M. Circulation 2004, 3215; (d) Natarajan, P.;
Forte, T. M.; Chu, B.; Phillips, M. C.; Oram, J. F.; Bielicki, J. K. J. Biol. Chem. 2004,
279, 24044; (e) Sethi, A. A.; Stonik, J. A.; Thomas, F.; Demosky, S. J.; Amar, M.;
Neufeld, E.; Brewer, H. B.; Davidson, W. S.; D’Souza, W.; Sviridov, D.; Remaley,
A. T. J. Biol. Chem. 2008, 283, 32273.
able positions of the consensus sequence.
Although more studies are required to elucidate the mechanism
of action of the apoA-I mimetic peptides described here, and to
determine whether the observed in vitro activity correlates with
an effect on atherosclerotic plaques in vivo, these results are
encouraging, and represent a step towards the development of
peptide-based therapies aimed at reducing cardiovascular risk in
humans.
Acknowledgments
11. (a) Navab, M.; Anantharamaiah, G. M.; Hama, S.; Garber, D. W.; Chaddha, M.;
Hough, G.; Lallone, R.; Fogelman, A. M. Circulation 2002, 105, 290; (b) Li, X.;
Chyu, K. Y.; Faria Neto, J. R.; Yano, J.; Nathwani, N.; Ferreira, C.; Dimayuga, P. C.;
Cercek, B.; Kaul, S.; Shah, P. K. Circulation 2004, 110, 1701.
12. Anantharamaiah, G. M.; Venkatachalapathi, Y. V.; Brouillette, C. G.; Segrest, J. P.
Arteriosclerosis 1990, 10, 95.
13. (a) Blackwell, H. E.; Grubbs, R. H. Angew. Chem., Int. Ed. 1998, 37, 3281; (b)
Schafmeister, C. E.; Po, J.; Verdine, G. L. J. Am. Chem. Soc. 2000, 122, 5891.
14. Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446.
15. Troutt, J. S.; Alborn, W. E.; Mosior, M. K.; Dai, J.; Murphy, A. T.; Beyer, T. P.;
Zhang, Y.; Cao, G.; Konrad, R. J. J. Lipid Res. 2008, 49, 581.
We thank Steven Harper for the helpful discussion and support,
Silvia Pesci for help with the NMR experiments and Manuela Emili
for the artwork.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.10.128.
16. Chen, Y. H.; Yang, J. T.; Chau, K. H. Biochemistry 1974, 13, 3350–3359.