Computationally Driven Structure Optimization, Synthesis, and Biological Evaluation of Imidazole-Based Proprotein Convertase Subtilisin/Kexin 9 (PCSK9) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b00402

Proprotein convertase subtilisin/kexin 9 (PCSK9) is responsible for the degradation of the hepatic low-density lipoprotein receptor (LDLR), which in turn regulates the circulating low-density lipoprotein cholesterol (LDL-C) level. For this reason, the PCSK9 inhibition, by small molecules or peptides, is a validated therapeutic approach for fighting hypercholesterolemia and cardiovascular diseases. In this field, we have recently reported an imidazole-based peptidomimetic that has shown PCSK9 inhibitory activity in the micromolar range. Here, by applying advanced computational techniques, the binding mechanism of that imidazole peptidomimetic was predicted. Then, among a small set of poly-imidazole analogs, compounds showing the highest theoretical affinity were suitably synthesized, relying on a van Leusen reaction based multicomponent strategy. One compound (named RIm13) displayed a PCSK9 inhibitory activity 10-fold lower than the template compound, and, remarkably, at a concentration of 1 μM, it successfully prevented the LDLR degradation mediated by PCSK9 on HepG2 cells. As well as increasing the LDL uptake at the same concentration, RIm13 represents currently one of the most potent small molecules targeting the PCSK9/LDLR protein-protein interaction.

Full text of DOI:10.1021/acs.jmedchem.9b00402

Subscriber access provided by AUB Libraries
Article
Computationally Driven Structure Optimization, Synthesis,
and Biological Evaluation of Imidazole-based Proprotein
Convertase Subtilisin/Kexin-9 (PCSK9) Inhibitors
Carmen Lammi, Jacopo Sgrignani, Anna Arnoldi, Giordano
Lesma, Claudia Spatti, Alessandra Silvani, and Giovanni Grazioso
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00402 • Publication Date (Web): 18 Jun 2019
Downloaded from http://pubs.acs.org on June 18, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Computationally Driven Structure Optimization, Synthesis, and
Biological Evaluation of Imidazole-based Proprotein Convertase
Subtilisin/Kexin-9 (PCSK9) Inhibitors
9
Carmen Lammi,a Jacopo Sgrignani,b Anna Arnoldi,a Giordano Lesma,c Claudia Spatti,c Alessandra
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Silvani,*,c Giovanni Grazioso*,a
a Dipartimento di Scienze Farmaceutiche, Università degli Studi di Milano, Via L. Mangiagalli 25, 20133 Milan,
Italy.
b Istituto di Ricerca in Biomedicina (IRB), Università della Svizzera Italiana (USI), Via V. Vela 6, CH-6500,
Bellinzona, Switzerland.
c Dipartimento di Chimica, Università degli Studi di Milano, Via Golgi 19, 20133 Milan, Italy.
Abstract
Proprotein convertase subtilisin/kexin 9 (PCSK9) is responsible for the degradation of the hepatic low-
density lipoprotein receptor (LDLR), which in turn regulates the circulating low-density lipoprotein
cholesterol (LDL-C) level. For this reason, the PCSK9 inhibition, by small molecules or peptides, is a
validated therapeutic approach for fighting hypercholesterolemia and cardiovascular diseases. In this
field, we have recently reported on an imidazole-based peptidomimetic that has shown PCSK9 inhibitory
activity in the micromolar range. Here, by applying advanced computational techniques, the binding
mechanism of that imidazole peptidomimetic was predicted. Then, among a small set of poly-imidazole
analogs, compounds showing the highest theoretical affinity were suitably synthesized, relying on a van-
Leusen reaction-based multicomponent strategy. One compound (named RIm13) displayed a PCSK9
inhibitory activity 10-fold lower than the template compound and, remarkably, at the concentration of 1
M, it successfully prevented the LDLR degradation mediated by PCSK9 on HepG2 cells. As well as
increasing the LDL uptake at the same concentration, RIm13 represents currently one of the most potent
small molecules targeting the PCSK9/LDLR PPI.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 28
1
2
3
4
INTRODUCTION
Low-density lipoprotein receptors (LDLRs) are the major intracellular carriers of cholesterol in the
blood. The LDLR is particularly abundant in the liver, the primary organ responsible for removing excess
cholesterol from the body. Thus, the amount of LDLRs on the surface of liver cells determines how
quickly LDL-cholesterol (LDL-C) is removed from the bloodstream. The accumulation of LDL-C in the
plasma is known to cause the formation of the atherosclerotic plaque, a key risk factor of cardiovascular
disease (CVD).1, 2 The LDLR activity on the cell surface is mainly regulated by Proprotein Convertase
Subtilisin/Kexin 9 (PCSK9),3, 4 which plays an important role in regulating the degradation of hepatic
LDLR.5-7 For this reason, PCSK9 is a validated therapeutic target for hypercholesterolemia treatment
and CVD prevention. In this context, the main strategic approach consists of the use of specific
monoclonal antibodies (mAbs), such as evolocumab and alirocumab, which function by impairing the
PCSK9-LDLR protein-protein interaction (PPI). The results of the Fourier trial8 have shown that the
addition of evolocumab to statin therapy over several years has significantly reduced cardiovascular
morbidity and mortality in patients with evident atherosclerotic CVD. It is, however, important to
underline that mAbs certainly represent a successful therapeutic strategy for inhibiting PCSK9 but are
extremely expensive.
5
6
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Regardless, the PCSK9-LDLR PPI remains a validated target for new drug discovery, and many efforts
are currently underway to develop peptidomimetics or other small molecules capable of impairing this
PPI. Since the PCSK9/LDLR X-ray crystal structure9 clearly evidences that this PPI is mediated by a β-
sheet, we have recently synthesized an N-methyl tetra-imidazole derivative (1, MeIm, Figure 1), a
11
minimalist peptidomimetic capable of resembling a β-strand motif.10,
Interestingly, biological
investigations have suggested that MeIm prevents PCSK9 from binding to LDLR in a dose-response
manner, with an IC50 value of 11.2 µM. Moreover, MeIm improves the functional ability of human
hepatic HepG2 cells to uptake LDL from the extracellular environment, with an EC50 equal to 6.0 µM.
The literature has reported on the use of the Exploring Key Orientations (EKO) procedure, as well as one
round of combinatorial fragment-based virtual docking, to discover compounds capable of impairing the
PCSK9/LDLR PPI. Three hit compounds bound PCSK9 with IC50 values in the 20–40 μM range (2,
LDLL-1dnlr, Figure 1).12 Moreover, Portola Pharmaceuticals has patented some tetrahydroisoquinolines
(3, Figure 1)13 that increase the LDL uptake into liver cells and LDLR cell surface populations. Similarly,
by a virtual screening session, Min and coworkers have identified compound 4 (or CB_36, Figure 1),14
containing fragments structurally resembling the plasticizer 2,2-bis(4-hydroxyphenyl)propane (5, BPA).
However, these studies did not report any direct evidence that such small molecules bind PCSK9, so that
2
ACS Paragon Plus Environment

Page 3 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
they could instead act via another mechanism, such as PCSK9 modulators. Similarly, compound 6 (R-
IMPP or PF-00932239, Figure 1), as well as its analogs developed by Pfizer, act as antisecretagogues,
lowering PCSK9 secretion by liver cells. Unfortunately, however, they are endowed with severe
hematopoietic side effects.15, 16 As recently reviewed by Xu et al.,14 no substantial advancement in the
discovery of PCSK9 small-molecule inhibitors has been made so far. Therefore, the therapy by PCSK9
mAbs remains the most advanced treatment for hypercholesterolemia.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
H2NOC
O
O
CH3
N
CH3
N
H
N
N
iPr
N
N
N
O
H
N
N
N
O
N
N
N
CH3
CH3
HN
HOOC
O
NH
1, MeIm
O
O
H2N
NH
2, LDLL-1dlnr
H
N
O
NH
N
F
3
4
O
HO
OH
N
N
O
HN
5, BPA
6
Figure 1. Chemical structures of small molecules endowed with PCSK9 inhibitory activity.
Given this scenario, the present study presents a detailed investigation on some new small molecules
capable of inhibiting PCSK9. Specifically, the objective was to improve the biological activity of MeIm
(1) by substituting the imidazole N-methyl groups with new moieties, suitably designed by a target-based
computational approach. In order to avoid the expensive and time-consuming synthesis and biological
characterization of compounds clearly showing low computed affinity, only the compounds showing the
highest estimated affinity to PCSK9 were targeted for the experimental work. On the synthesized
compounds, biological studies ranging from the inhibition of the PCSK9-LDLR binding to the evaluation
of cell safety were performed. In addition, in order to evaluate the potential hypocholesterolemic activity,
the most promising compounds were investigated for assessing their ability to modulate the LDLR
expression on human hepatic HepG2 cells surface and to increase the capacity of the same cells to uptake
extracellular LDL.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 28
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
General overview of the design strategy. To improve the biological activity of MeIm by applying a
rational target-based computational approach, the first step involved the assessment of the MeIm binding
mode. Toward this aim, three replicas of “supervised Molecular Dynamics” (suMD) simulations were
performed.17-20 This algorithm simulates the ligand binding process in an accessible time scale,
monitoring the distance between the ligand atoms and the protein binding site during classical MD
simulations without the introduction of biases or external forces. Successively, a cluster analysis
algorithm was applied to the attained trajectory frames. Molecular Mechanics/Generalized-Born Surface
Area (MM-GBSA) permitted us to choose the suMD replica in which the ligand showed the highest
binding free energy and, reasonably, the most reliable binding mode. Starting from these data, the focused
substitution of the N-methyl groups exposed by the imidazole rings of MeIm was undertaken in the
rational design of a small set of new MeIm analogs (RIm compounds). Finally, MD simulations and
ligand binding free energy calculations, via the MM-GBSA approach, were used to select the compounds
eligible for synthesis and further biological evaluation (Figure 2).
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Workflow of the computational approach applied to select the best MeIm analogs for the synthesis
and the biological evaluation.
suMD simulations. The MeIm binding process was determined by performing suMD simulations.17-20
The PCSK9 area involved in the interaction with the EGF-A domain of LDLR was considered as the
target for ligand binding.21 Briefly, by a tabu-like algorithm, the suMD methodology exploits the distance
between the centers of mass of the ligand atoms and the protein binding site during reiterated short MD
simulations of 0.6 ns. MD simulations were performed until the ligand−target distance repeatedly reaches
4
ACS Paragon Plus Environment

Page 5 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
values within the interval of 2–5 Å. The suMD simulation runs can be examined by a graph in which the
suMD step (or the ligand interaction energy) and the distance between the ligand and the binding site are
plotted (Figure S1A, Supporting Information). As suggested by the suMD developers, the suMD
simulations were repeated three times.18 Then, cluster analysis was applied to extract the conformations
representative of each suMD replica to assess the convergence of the attained results. In detail, the MD
frames were divided into clusters by the complete average linkage algorithm,22 and the complex
conformations with the lowest root mean square deviation (RMSD) to the cluster centers were visually
inspected. By our simulations, each suMD replica showed a diverse MeIm binding mode. Thus, to
choose the most consistent binding pose, MM-GBSA calculations were performed on the most populated
clusters of conformations found in each replica. In fact, we argued that the MeIm preferred binding mode
could be found in the suMD replica in which the ligand showed the lowest binding free energy value and
in the highest populated cluster. This would also indicate that the ligand rested in an energy minimum of
the system free energy landscape. More specifically, in the complex conformations of replica 2, MeIm
was more tightly bound to PCSK9 (Figure S1B, Supporting Information). Nevertheless, to better sample
the conformational space of the PCSK9/MeIm complex, 1 s-long classical MD simulations were
performed on the PCSK9/MeIm complex retrieved from this replica. Interestingly, by monitoring the
PCSK9/MeIm contacts (Table S1, Supporting Information), these simulations excluded any ligand
unbinding events, confirming that the MeIm conformers fluctuated within a minimum of the potential
energy surface. However, the visual inspection of the MD trajectory and the RMSD plot (Figure 3A) led
us to suspect that MeIm could adopt different binding modes within PCSK9. Thus, to explore the
structural differences among them, hierarchical agglomerative cluster analysis was once more applied to
identify the most representative modes (Figures 3B and Figure S2, Supporting information). The attained
data evidenced that in the most populated cluster of conformations (59% of frames, Figure 3B), MeIm
was anchored on a PCSK9 β-strand shaped by residue C378-S381, while a cation-π stacking between the
phenyl ring of MeIm and the side chain of residue PCSK9-R194 may additionally stabilize the ligand on
the PCSK9 surface. Considering that the binding mode found in the first cluster was the most preferred
by MeIm, the design of new analogs was accomplished by considering the representative complex
conformation of this cluster as the design template.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 28
1
2
3
4
5
6
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. A) RMSD plot of MeIm heavy atoms with respect to their average structure. The PCSK9 backbone
atoms were previously aligned with respect to the initial minimized structure. B) Representative structural clusters
of MeIm as a result of clustering the MD trajectory frames. MeIm is represented as sticks, whereas the PCSK9
surface area is colored according to atomic partial charges: positive and negative areas are colored blue and red,
respectively. These figures were acquired by Pymol software (Schrödinger Inc., USA).
Design of novel MeIm analogs. Initially, to increase the synthetic feasibility and to obtain ligands
bearing a molecular weight lower than 500 Da (as stated by Lipinski rules), the new MeIm analogs
(named RIm compounds) contained only one phenyl ring bound to a chain of three suitably substituted
imidazole rings.
In principle, the moieties decorating the imidazole rings of the RIm compounds should show the
highest structural complementarity with the biological counterpart with which they directly interact. For
this reason, when examining the PCSK9 site involved in the LDLR recognition (Figure 4), we thought
that the PCSK9 inhibitors could exploit at least four regions: (i) the positively charged area close to K237
and R194; (ii) the hydrophobic pocket sized by I369, A239, F379, L158; (iii) the negatively charged area
close to D367; and (iv) the planar hydrophobic area enclosing the β-strand-shaped residues C378–S383
(Figure 4).
6
ACS Paragon Plus Environment

Page 7 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. PCSK9 surface exploited for drug design approach. The enzyme solvent-accessible surface is depicted
according to the partial charge of the residues, blue for positive and red for negative areas.
Taking into account these information, MeIm could interact with the positive charged area, the
hydrophobic pocket (though the methyl group could occupy it partially), and the β-strand/hydrophobic
area. At variance, the optimal inhibitor should putatively involve all the above-described areas.
Therefore, starting from the previously selected PCSK9/MeIm complex conformation, a small set of
compounds was rationally designed (see experimental section for details) by considering the following
assumptions:
– compounds RIm1 and RIm2, bearing a meta- or a para-methoxy group on the phenyl ring, were
designed to increase the electron density of the benzene ring, considering that it was in contact with the
PCSK9 positively charged residue R194.
– by compounds RIm3-RIm4, we explored substituents showing chemical properties different from
those of an aliphatic group of the first imidazole ring. For simplicity, we did not introduce any additional
modification of the methyl group since, in the most populated MeIm conformations cluster, R2 was
immersed in solvent molecules and did not show any contact with PCSK9.
– substituents in the second imidazole ring of compounds RIm5-RIm7 pointed to occupy the PCSK9
pocket shaped by PCSK9 residues I369, A239, L158, and F379.
– compounds RIm8-RIm12 were designed to promote the creation of a hydrogen bond/salt bridge
between the third imidazole ring of the inhibitors and the PCSK9 negatively charged area shaped by the
side chain of D367.
– compound RIm13 was designed to improve the basicity of the amine group of compound RIm12,
aiming to better stabilize the salt-bridge interaction.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 28
1
2
3
4
5
6
– compound RIm14 was finally inserted into this set of molecules as an attempt to assess the effects
triggered by R2 homologation on RIm13.
The PCSK9/RIm(n) complexes were created by using the editing tools available in the Maestro software
(Schrödinger Inc., USA). Geometry optimization, 100 ns-long MD simulations, and MM-GBSA
calculations performed on the attained complexes permitted the estimation of the binding free energy
values (ΔG*) of these compounds (Table 1), with the aim of sorting out which compound was the most
theoretically active for further experimental evaluation.
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Chemical structure of the compounds considered in this study and their calculated binding free energy values (ΔG*,
column 6). ΔΔG* values (column 7) represent the difference between the theoretical binding free energy value of MeIm and
those calculated for new derivatives.
R3
N
N
N
N
N
N
R2
R4
R1
ΔG*
Compound
R1
R2
R3
R4
(kcal/mol ± Std. Err. of Mean)
MeIm
RIm1
RIm2
RIm3
RIm4
RIm5
RIm6
RIm7
RIm8
RIm9
RIm10
RIm11
RIm12
RIm13
Rim14
P5
 ± 0.4
 ± 
 ± 0.2
 ± 0.2
 ± 0.3
 ± 0.3
 ± 0.4
 ± 0.3
 ± 0.5
 ± 0.2
 ± 0.4
 ± 0.3
 ± 0.3
 ± 0.3
 ± 0.5
 ± 0.51
-3-OCH3 -CH3
-4-OCH3 -CH3
-4-OCH3 -NH2
-4-OCH3 -COOH
-4-OCH3 -CH3
-4-OCH3 -CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH2Ph
-CH2Naph
-CH3
-CH3
-4-OCH3 -CH3 -CH2-4ClPh
-4-OCH3 -CH3
-4-OCH3 -CH3
-4-OCH3 -CH3
-4-OCH3 -CH3
-4-OCH3 -CH3
-4-OCH3 -CH3
-CH2Ph
-CH2Ph
-CH2Ph
-CH2Ph
-CH2Ph
-CH2Ph
-CH2-NH2
-(CH2)2-NH2
-(CH2)3-NH2
-(CH2)4-NH2
-(CH2)5-NH2
-(CH2)5-NH-CH3
-4-OCH3 -CH3 -CH2CH2Ph -(CH2)5-NH-CH3
1 This value was estimated performing MM-GBSA calculations on 500 ns-long MD simulations on the PCSK9/P5 complex
described on reference 21.
The results in Table 1 suggest that: 1) the p-methoxy was better than m-methoxy group as R1 substituent;
2) the amino- and carboxy- groups, as R2 substituent, reduced the calculated affinity of the resulting
compounds; 3) the benzyl group, as R3 substituent, was better than the other selected moieties in the
8
ACS Paragon Plus Environment

Page 9 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
interaction with the PCSK9 hydrophobic cavity; 4) the alkyl-amino chains, as R4 substituents, containing
less than five methylene groups, were not capable to reach the negatively charged area on the PCSK9
surface; 4) the terminal -NH-Me group, as R4 substituent, led to a compound (RIm13) with a computed
affinity 3 kcal/mol lower than that of RIm12; 5) the homologation of the benzyl group on Rim13 did not
lead to the improvement of the computed ΔG* value, since the one of Rim14 was slightly lower than
that of RIm13. Consequently, it was possible thinking that the benzyl group was optimal for the
interaction with the hydrophobic cavity on the PCSK9 surface.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the light of these outcomes, RIm13 and RIm14 possessed the lowest estimated binding free energy
on PCSK9. For them, additional MD simulations were performed to gain accurate insights into their
binding mechanism. Toward this aim, MD simulations were run to 1 s, and cluster analysis was
subsequently performed. Then, MM-GBSA calculations were accomplished on the trajectory frames
found in the most populated clusters of complex conformations. By these, ΔG* values of 22.1±0.5 and
19.5±0.6 kcal/mol were attained for RIm13 and RIm14, respectively. The theoretical affinity of both
compounds remained higher than that of MeIm; moreover, RIm13 showed the lowest estimated ΔG*
value.
In the complex conformation representative of the most populated cluster (55% of frames), RIm13
(Figure 5A) was bound to PCSK9 by (1) creating a H-bond/salt bridge by the amino group with the
PCSK9 D367 residue, (2) projecting the benzyl ring in the hydrophobic pocket sized by A239, F379,
P156, and I369, and (3) generating a cation-π stacking by the third imidazole ring with the N-terminal
end of PCSK9 (S153). Remarkably, this putative binding mode showed high stability over MD
simulations, as demonstrated by monitoring the contacts between PCSK9 and the ligand atoms (see Table
S2, Supporting Information).
On the other hand, the PCSK9 residues involved in the interaction with RIm14 were essentially
similar to those previously described for RIm13, although the frequency of the contacts (Table S3,
Supporting information) decreased. Nevertheless, the theoretical binding mode of RIm14 (Figure 5B)
that resulted in the cluster of conformations populated by the majority of trajectory frames (43%) was
different from RIm13. The difference was found in the position of the phenyl ring, creating cation-π
stacking with the PCSK9 N-terminal end (S153), and in the formation of a salt bridge with D374.
Moreover, the phenethyl moiety in RIm14 seemed too large to be accepted in the hydrophobic pocket
sized by A239, F379, P156, and I369 (Figure 5B), since it fluctuated in the solvent environment from
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 28
1
2
3
4
5
6
the early stages of MD simulations. Additionally, the overall stability of RIm14 on the PCSK9 surface
was lower than that of RIm13, as expected by the RMSD plot (Figure S3, Supporting information).
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Hypothetical orientation of RIm13 (A, magenta sticks) and RIm14 (B, cyan sticks) on the PCSK9 surface. The
enzyme is represented as green sticks highlighting the secondary structure. The solvent-accessible surface of PCSK9 is colored
depending on the partial charge of the atoms: positive areas are depicted as blue, while red areas suggest the presence of
positively charged residues. H-bonds are represented as yellow dashed lines.
Based on these assumptions, compounds RIm13 and RIm14 were synthesized and experimentally
assayed to evaluate their biological activity. Together with these, a synthetic intermediate (compound 9,
see the next section) was also tested in order to evaluate the effects of the presence of the third imidazole
ring on the biological activity of compounds RIm13.
Synthesis of compounds RIm13 and RIm14. The targeted compounds RIm13 and RIm14 were
synthesized through a quasi-iterative process, pivoted on the alternation of van Leusen three-component
reactions (vL-3CRs) with formylation steps. The vL-3CR is able to generate trisubstituted imidazoles in
a single step, by a base-induced condensation between an aldehyde, a primary amine and tosylmethyl
isocyanide (TosMIC). This chemical path was conceived for ensuring the rapid construction of
functionalized triimidazole derivatives, exposing different N-substituents on imidazole rings. The proper
choice of amine components for the three sequential vL-3CRs allowed determining such N-substituents.
Formylation at C-2 of imidazole intermediates furnished step by step the required aldehyde components,
besides the starting p-anisaldehyde (Scheme 1).
10
ACS Paragon Plus Environment

Page 11 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
N
CHO
N
Me
Me
N
N
CHO
N
Me
R
a
b
c
N
N
MeO
H2N
MeO
MeO
7
8
MeO
9
(R = CH2Ph)
(R = CH2CH2Ph)
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Boc
10
N
Me
13
Boc
NHMe
NMe
5
N
5
CHO
N
N
N
N
N
N
N
N
N
Me
R
d
e
f
N
N
R
N
Me
Me
R
N
N
N
MeO
MeO
MeO
11
12
14
15
RIm13
RIm14
(R = CH2Ph)
(R = CH2CH2Ph)
(R = CH2Ph)
(R = CH2CH2Ph)
(R = CH2Ph)
(R = CH2CH2Ph)
Scheme 1. Reagents and conditions: a) MeNH2 40 wt% aq sol, DMF, rt, 2h; then TosMIC, K2CO3, 50 oC, 3h (82%). b) n-
BuLi, THF, - 78 oC, 2h; then DMF, rt, 2h (77%). c) Benzylamine (for 9) or phenethylamine (for 10), DMF, 70 oC, 2h; then
TosMIC, K2CO3, 70 oC, overnight (48% for 9, 45% for 10). d) n-BuLi, TMEDA, THF, - 78 oC, 2h; then DMF, rt, 2h (43%
for 11, 77% for 12). e) Amine 13, DMF, 70 oC, 2h; then TosMIC, K2CO3, 70 oC, overnight (45% for 17, 40% for 18). f) 4N
HCl in AcOEt, from 0 oC to rt, overnight; then NaHCO3/CH2Cl2 (quant. yield for both RIm13 and RIm14).
In detail, the first imidazole derivative 7 was obtained in good yield, starting from p-anisaldehyde,
methylamine and TosMIC. A precondensation time of 2 h ensured the in situ formation of the
intermediate imine from p-anisaldehyde and methylamine, after that isocyanide was added. Then,
compound 7 was treated with n-BuLi at low temperature and DMF as formylating agent, to give the
aldehyde derivative 8 still in good yield. The subsequent vL-3CR employed benzyl amine or,
alternatively, phenethylamine, besides compound 8 and TosMIC, and required a slightly higher
temperature, affording the diimidazole derivatives 9 and 10 in moderate yield, after careful purification
on neutral aluminum oxide. Formylation of compounds 9 and 10, to give 11 and 12 respectively, required
the addition of TMEDA in order to facilitate the solubilization of intermediate lithium anions and then
their reaction with DMF. The last vL-3CR needed the preliminary synthesis of amine 13, which was
carried out as reported in Scheme S1 (Supporting Information).
From reaction of aldehyde derivatives 11 and 12 with amine 13 under standard vL-3CR conditions,
the N-Boc protected triimidazole derivatives 14 and 15 were obtained. Finally, by acidic N-Boc
deprotection, the target compounds RIm13 and RIm14 were achieved quantitatively. Their structure, as
well as that of all intermediates, was confirmed through 1H NMR, 13C NMR and mass spectrometry.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 28
1
2
3
4
5
6
7
8
9
PCSK9-LDLR binding experiments. In order to evaluate the ability of compounds RIm13 and RIm14
to impair the PPI between PCSK9 and LDLR, in vitro binding experiments were performed. Both
compounds were tested in a range of concentrations from 0.01 to 500.0 µM (Figure 6A). Compounds
RIm13 and RIm14 showed dose-response behaviors with IC50 values of 1.38±0.1 and 3.06±0.3 µM,
respectively. Compound 9, which was tested at the concentrations of 10.0 and 100.0 μM, was instead
inactive, since, at the concentration of 100.0 μM, it impaired the PCSK9-LDLR binding by 18.2±8.4%,
but in a non-statistically significant way (Figure 6B).
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B
A
Figure 6. A) Inhibition of the PCSK9/LDLR PPI by compounds RIm13 (red curve), RIm14 (blue curve), MeIm
(orange curve), and peptide P5 (green curve). IC50 values were equal to 1.4, 3.1, 11.2, and 1.6 µM, respectively.
Data points represent averages ± s.d. of three independent experiments in duplicate. B) Compound 9 does not
impair the PCSK9/LDLR PPI at 10.0 and 100.0 μM. Data points represent averages ± s.d. of three independent
experiments in duplicate. ns: not significative.
Remarkably, these data confirmed the previous theoretical predictions in favor of improved affinities of
RIm compounds. In fact, RIm13 displayed an IC50 value 10-fold lower than that of MeIm (11.2 µM)10
and, accordingly, the RIm13-calculated ΔG* value was 5.4 kcal/mol lower than that of MeIm (Table 2).
Moreover, the lower experimental and estimated affinity shown by RIm14 could be due to the different
suspected binding mode and to the low stability of RIm14 in the PCSK9 binding site (RMSD plot, Figure
S3, Supporting information). Interestingly, binding experiments on compound 9 suggested that the
presence of the third imidazole ring in the structure of RIm13 is fundamental to the PCSK9 binding
affinity of the compound. MD simulations accomplished on the PCSK9/9 complex showed the
conformational instability of the ligand within the PCSK9 surface, this led rapidly to the ligand unbinding
after just 45 ns of MD simulations. Considering the trajectory frames in which the ligand interacted with
12
ACS Paragon Plus Environment

Page 13 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
PCSK9, a ΔG* value of 7 kcal/mol (-15.0 ± 0.3 kcal/mol) lower than that of Rim13 was calculated for
compound 9.
6
7
8
9
Characterization of the cholesterol-lowering effect of RIm13 on HepG2 cells. The 10-fold in vitro
binding affinity exhibited by RIm13 prompted us to investigate more deeply the molecular and
functional effects of PCSK9 inhibition on LDLR-pathway modulation. To achieve this goal, human
hepatic HepG2 cells were chosen as the cell system of interest. Initially, viability experiments (MTT)
were performed in order to exclude any potential cytotoxicity effect after the treatment of HepG2 cells
with increasing concentrations (0.1, 1.0, and 10.0 µM) of the compound RIm13. The attained results
suggested that RIm13 was safe for the HepG2 cells (Figure S4, Supporting Information), therefore these
concentrations were used for the further biological characterization of RIm13. Afterwards, the effects of
compound RIm13 on the modulation of the LDLR localized on the HepG2 cell surface were investigated
using an in cell western (ICW) assay.23 These experiments clearly indicated that the LDLR protein levels
decreased in the presence of PCSK9 alone by 30.5±4.4% versus the control cells, and that compound
RIm13 (at the fixed concentration of 1.0 μM) enhances the LDLR protein levels when co-incubated with
PCSK9. These experiments were performed using peptide P5 (positive control) and 9 (negative control)
as reference ligands at the fixed concentration of 1.0 and 10.0ꢀµM, respectively (Figure 7A). More in
details, HepG2 cells incubated with PCSK9 and RIm13 (1.0 µM) renewed the LDLR on cell membrane
up to 88.4 ± 3.7 %, similarly to the positive control peptide P5, which restored the receptor protein level
up to 93.9 ±4.8 %. On the contrary, the compound 9 (negative control) was not able to reestablish the
LDLR levels, suggesting that compound RIm13 is an effective PCSK9 inhibitor. Based on these results,
new ICW experiments were assessed testing RIm13 at 0.1, 1.0, and 10.0 µM. In particular, Figure 7B
clearly indicates that the LDLR levels dropped in the presence of PCSK9 alone by 47.2±12.2% versus
the control cells, and that increased the LDLR protein levels when co-incubated with PCSK9. In
particular, HepG2 cells incubated with PCSK9 and RIm13 (0.1, 1.0, and 10.0 µM) restored the LDLR
up to 77.7±15.7%, 101.2±21.8%, and 101.6±14.8%, respectively, indicating a concentration-response
dependence (Figure 7B).
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Moreover, functional experiments were carried out in order to evaluate the ability of the new
compound to modulate the capacity of HepG2 cell to uptake extracellular LDL. At the beginning, HepG2
cells were treated with PCSK9 alone or in the presence of the new and references compounds (9 and P5),
respectively. After 2 h of treatment with PCSK9 alone, a reduced ability of HepG2 cells to uptake
fluorescent LDL by 34.8 ± 7.2 % versus untreated cells was achieved (Figure 7C). Unlike to compound
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 28
1
2
3
4
5
6
7
8
9
9 (negative control), compound RIm13, tested at the fixed concentration of 1.0 μM restored the LDL-
uptake up to 103.7±17.3 % similarly to peptide P5 that, at the same concentration, increase the ability of
hepatic cells to absorb LDL up to 103.5±13.5% (Figure 7C). Moreover, as showed in the Figure 7D, the
treatment with PCSK9 alone reduced the capability of HepG2 cells to clear the fluorescent LDL by
67.0±4.5% versus untreated cells was observed (Figure 7D); however, this ability was improved by
RIm13 at all tested concentrations. In particular, RIm13, at the concentrations of 0.1, 1.0, and 10.0 µM,
improved the LDL uptake by 63.6±12.8%, 95.7±15.5%, and 101.0±24.1%, respectively.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. A) RIm13 at the fixed concentration of 1.0 μM, increases the LDLR protein level on cell membrane
similarly to peptide P5 (1.0 μM), whereas compound 9 was unable to modulate the active LDLR protein; B) RIm13
induces an increase of the LDLR protein level on the HepG2 cell surface. LDLR degradation mediated by PCSK9
(K9) is prevented by increasing concentrations of RIm13 (0.1, 1.0, and 10.0 µM). C) RIm13 (1.0 μM) restores the
ability of HepG2 to clear LDL from extracellular environment similar to the positive control, P5 (1.0 μM), on the
contrary, compound 9 was ineffective to modulate the functional ability of hepatic cells; D) The decreased ability
to uptake LDL by HepG2 cells mediated by K9 is prevented by increasing concentrations of RIm13 (0.1–10.0
µM). C represents the control without any treatment. Results are mean ± s.d. of three independent experiments.
(*) P < 0.05, (**) P < 0.001, and (****) P < 0.00001.
14
ACS Paragon Plus Environment

Page 15 of 28
Journal of Medicinal Chemistry
1
2
3
CONCLUSIONS
4
5
6
7
8
We applied an integrated computational approach, including suMD, MD simulations, cluster analysis,
and binding free energy estimation (MM-GBSA), to investigate the binding mechanism of the poly-
imidazole peptidomimetic MeIm to PCSK9. Starting from this information, a small set of compounds
(Table 3) were suitably designed by means of molecular modeling tools and a structure-based drug design
methodology. Then, MD simulations and MM-GBSA calculations were performed on this set of
compounds to select the most promising terms for further chemical synthesis and biological evaluation
of PCSK9. This computational procedure avoided the costly synthesis and biological characterization of
compounds with low computed affinity. Therefore, compounds RIm13 and Rim14 appeared to be
particularly interesting, since they displayed the highest estimated binding affinity to PCSK9. The
consequent synthesis, by van Leusen three-component reactions, and the experimental binding affinity
assays confirmed the theoretical predictions. In fact, RIm13 exhibited an improved inhibitory activity
on the PCSK9/LDLR PPI, displaying an IC50 value 10-fold lower than the template compound MeIm.
Remarkably, at the concentration of 1 M, RIm13 prevented the LDLR degradation mediated by PCSK9
on HepG2 cells (Figure 7B), as well as increased the LDL uptake at the same concentration (Figure 7C).
To the best of our knowledge, RIm13 represents currently one of the most potent small molecules
targeting the PCSK9/LDLR PPI. The attained results confirm that the computer-aided design of PCSK9
inhibitors can reliably guide the drug discovery process.
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General Methods. All commercial materials and solvents (> 95% purity grade) was used without further
purification. All reactions were carried out under a nitrogen atmosphere, unless otherwise noted. All
reactions were monitored by thin layer chromatography (TLC) on precoated silica gel 60 F254; spots
were visualized with UV light or by treatment with a 1% aqueous KMnO4 solution or 0.2% ninhydrin
solution in ethanol. Products were purified by flash chromatography (FC) on silica gel 60 (230–400
mesh) or aluminium oxide activated neutral 58 (150 mesh). NMR spectra were recorded on 300 or 400
13
MHz Bruker spectrometers, using tetramethylsilane (TMS) as the internal standard. For C NMR, the
APT pulse sequence was adopted. Chemical shifts are reported in parts per million relative to the residual
1
solvent. Multiplicities in H NMR are reported as follows: s = singlet, d = doublet, t = triplet, m =
multiplet, br s = broad singlet. The mass spectra were obtained in the ESI positive mode ((+)-HRESIMS),
from a Waters Micromass Q-Tof micro Mass spectrometer.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 28
1
2
3
4
5
6
7
8
9
The purity of tested compounds was confirmed to be > 95% by means of elemental analysis on a CHN
Perkin ‐ Elmer 2400 instrument. To exclude any spectroscopic interference by compounds with LDL
uptake cell-based test, fluorescence assays have been conducted. Moreover, the chemotypes found in our
compounds do not belong to Pan Assay Interference Compounds (PAINS), as confirmed by ZINC server
(http://zinc15.docking.org/patterns/home).
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
System setup, suMD, and classical MD simulations. After retrieving the X-ray coordinate from the
Protein Data Bank (PDB code 4NMX, resolution 1.85 Å),24 we built and optimized the PCSK9
computational model. The model was successfully utilized to perform several studies on natural and
synthetic peptides.21, 25, 26 Here, it was initially used to perform suMD simulations20 and to predict the
MeIm binding mechanism. The atomic partial charges of MeIm, as well as the charges of the new
compounds, were assigned by RESP calculations,27 as implemented in an antechamber. The applied MD
simulation protocol have been described by us.21, 26 FF4SB28 and GAFF29 force fields were used to
parameterize the complexes studied by the suMD algorithm and classical MD simulations. The default
parameters were set in the suMD configuration file (selection.dat), and the 0.91 version of this algorithm
was employed for these calculations.18 The ligand binding site on PCSK9 was the site shaped by residues
depicted by the presence of the EGF-A domain of LDLR in the X-ray structure (P155, R194, A239,
T377, F379, Q382).30 SuMD simulations were performed by the support ACEMD31 as an algorithm for
classical MD simulations. The 1 s-long production runs of MD simulations performed on PCSK9 in the
complex with MeIm, RIm13, and RIm14 were performed by pmemd.cuda modules of the Amber17
package.32 The MD trajectories were examined by visual inspection with VMD.33
Cluster analysis. Cluster analysis of the trajectory coordinates attained by suMD and classical MD
simulations on PCSK9/MeIm, PCSK9/RIm13, and PCSK9/RIm14 complexes were achieved via the
hierarchical agglomerative approach, with an average complete linkage algorithm. The Cpptraj module34
of Amber1732 was used toward this aim. The default parameters were used to process each trajectory file,
and the conformation with the lowest RMSD to the cluster center was examined and discussed in the
above sections. The RMSD of the ligand heavy atoms were used as a distance metric.
Ligand binding free energy estimation. The MM-GBSA approach was applied to estimate the ligand
binding free energy values. Once the geometrical stability of the ligand under investigation was acquired,
i.e., the ligand RMSD value remained stable over time, 100 snapshots were extracted from the attained
16
ACS Paragon Plus Environment

Page 17 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trajectory to ensure the lowest standard error in the free energy estimation and the lowest calculation
time. The MM-PBSA.py module35 was employed to perform these calculations, keeping all parameters at
the default values. In our calculations, the single trajectory approach of MM-GBSA was applied, and the
entropic contribution to ΔG was omitted. For this reason, our ligand binding free energy values were
termed <<ΔG*>> throughout the text.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Design of RIm compounds. New MeIm analogs (RIm compounds) were manually constructed by
adding or modifying the atoms comprising the ligand moieties in the PCSK9/MeIm complex
representing the most populated cluster of conformations. The editing tools available in the Maestro
software (Schrödinger, Inc.) were utilized toward this aim. The carboxylic or amino groups, if present in
the ligand structures, were considered in their ionized form, for consistency with the expected protonation
state at the physiological pH. Then, RIm compounds followed the computational protocols (MD
simulations, cluster analysis, and MM-GBSA calculations) as described above.
Synthetic procedures.
5-(4-Methoxyphenyl)-1-methyl-1H-imidazole (7)
p-Anisaldehyde (1 g, 7.3 mmol) was dissolved in DMF (8 mL). Methylamine (40 wt. % aqueous solution,
1.3 mL, 14.6 mmol) was added and the resulting mixture was kept under stirring for 2 h at room
temperature. Potassium carbonate (1.5 g, 10.9 mmol) and tosylmethyl isocyanide (1.7 g, 8.8 mmol) were
added sequentially, the reaction was stirred for additional 3 h at 50 °C and then left overnight at room
temperature under stirring. The resulting mixture was partitioned between AcOEt and water. The aqueous
phase was extracted with AcOEt (x3) and the organic phase was washed with brine (x5), dried over
Na2SO4 and concentrated under reduced pressure, to give a residue that was purified by trituration in
diethyl ether. The product was obtained as a yellowish solid (82% yield). 1H NMR (300 MHz, CDCl3) δ
7.48 (s, 1H), 7.29 (d, J = 8.7 Hz, 2H), 7.02 (s, 1H), 6.95 (d, J = 8.7 Hz, 2H), 3.83 (s, 3H), 3.61 (s, 3H);
13C NMR (100 MHz, CDCl3) δ 159.4 (Cq), 138.6, 133.2 (Cq), 129.9 (2C), 127.5, 122.2 (Cq), 114.2 (2C),
55.3, 32.3. HRMS (ESI) m/z: [M+H]+ Calcd for C11H13N2O 189.1022; Found 189.1026.
5-(4-Methoxyphenyl)-1-methyl-1H-imidazole-2-carbaldehyde (8)
To a solution of compound 7 (3.5 g, 18.6 mmol) in dry THF (37 mL) under nitrogen atmosphere cooled
to -78 °C, a solution of n-butyl lithium (1.6 M in hexane, 17.5 mL, 27.9 mmol) was added dropwise, and
the resulting mixture was kept under stirring for 2 h at the same temperature. Dimethylformamide (2.9
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 28
1
2
3
4
5
6
7
8
9
mL, 37.2 mmol) was added and the reaction was stirred for 2 h at room temperature. The resulting
mixture was quenched with distilled water and extracted with AcOEt (x2). The organic phase was washed
with brine (x4), dried over Na2SO4, and concentrated under reduced pressure, to give a residue that was
purified by FC (CH2Cl2/MeOH 99/1). The purified product was obtained as a yellowish solid (77%
yield). 1H NMR (300 MHz, CDCl3) δ 10.04 (s, 1H), 7.35 (d, J = 8.3 Hz, 2H), 7.25 (s, 1H), 7.05 (d, J =
8.3 Hz, 2H), 3.99 (s, 3H), 3.87 (s, 3H); 13C NMR (75 MHz, CD3OD) δ 163.2, 149.5, 134.5, 132.7 (2C),
129.6, 122.8, 116.1 (2C), 91.1, 56.3, 33.9. HRMS (ESI) m/z: [M+H]+ Calcd for C12H13N2O2 217.0972;
Found 217.0978.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3’-Benzyl-5-(4-methoxyphenyl)- 1-methyl-1H,3’H,2,4’-biimidazole (9)
Aldehyde 8 (300 mg, 1.4 mmol) was dissolved in DMF (1.4 mL) at 0 °C. Benzylamine (180 µL, 1.7
mmol) was added and the resulting mixture was kept under stirring for 2 h at 70 °C under nitrogen
atmosphere. Then, potassium carbonate (288 mg, 1.7 mmol) and tosylmethyl isocyanide (326 mg, 2.1
mmol) were sequentially added and the reaction was stirred for additional 2 h at 70 °C. After that, further
tosylmethyl isocyanide (326 mg, 2.1 mmol) was added. The reaction was left overnight at 70 °C under
stirring. The resulting mixture was partitioned between AcOEt and water. The aqueous phase was
extracted with AcOEt (x3) and the organic phase was washed with brine (x5), dried over Na2SO4, and
concentrated under reduced pressure, to give a residue that was purified by chromatography on neutral
aluminum oxide (AcOEt/ MeOH gradient from 97/3 to 92/8). The purified product was obtained as a
dark brown foamy solid (48% yield). 1H NMR (400 MHz, CDCl3) δ 7.63 (s, 1H), 7.25 – 7.13 (m, 6H),
7.10 (s, 1H), 6.96 – 6.88 (m, 4H), 5.34 (s, 2H), 3.77 (s, 3H), 3.15 (s, 3H); 13C NMR (101 MHz, CDCl3)
δ 160.2, 139.7, 139.0, 137.3, 135.2, 131.4, 130.7 (2C), 129.2 (2C), 128.5, 128.3 (2C), 127.8, 122.9, 122.8,
114.9 (2C), 56.0, 50.2, 32.9. HRMS (ESI) m/z: [M+H]+ Calcd for C21H21N4O 345.1710; Found 345.1716.
3'-Phenethyl-5-(4-methoxyphenyl)-1-methyl-1H,3'H-2,4'-biimidazole (10)
Aldehyde 8 (100 mg, 0.5 mmol) was dissolved in DMF (0.5 mL). Phenethylamine (70 µL, 0.6 mmol)
was added and the resulting mixture was kept under stirring for 2 h at 70 °C under nitrogen atmosphere.
Potassium carbonate (95 mg, 0.7 mmol) and tosylmethyl isocyanide (107 mg, 0.6 mmol) were
sequentially added and the reaction was stirred for additional 2 h at 70 °C before adding other tosylmethyl
isocyanide (107 mg, 0.6 mmol). Then the reaction was left overnight at 70 °C under stirring. The resulting
mixture was then partitioned between AcOEt and water. The aqueous phase was extracted with AcOEt
(x3) and the organic phase was washed with brine (x5), dried over Na2SO4 and concentrated under
18
ACS Paragon Plus Environment

Page 19 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reduced pressure, to give a residue that was purified by chromatography on neutral alumina oxide
(AcOEt/ MeOH gradient from 97/3 to 92/8). The purified product was obtained as a dark brown foamy
solid (45% yield). 1H NMR (400 MHz, CDCl3) δ 7.45 (s, 1H), 7.35 (d, J = 8.6 Hz, 2H), 7.22 – 7.07 (m,
5H), 6.99 (m, 4H), 4.63 (t, J = 6.9 Hz, 2H), 3.86 (s, 3H), 3.43 (s, 3H), 2.99 (t, J = 6.9 Hz, 2H); 13C NMR
(100 MHz, CDCl3) δ 160.3, 139.7, 138.4, 135.1, 131.6, 130.8 (2C), 130.6, 129.4 (2C), 129.1 (2C), 127.8,
127.2, 123.0 (2C), 115.0 (2C), 56.0, 47.7, 38.6, 33.6. HMRS (ESI) m/z: [M+H]+ Calcd for C22H23N4O
359.1866; Found 359.1861.
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3’-Benzyl-5-(4-methoxyphenyl)- 1-methyl-1H,3’H-[2,4’ biimidazole]-2’-carbaldehyde (11)
To a solution of compound 9 (530 mg, 1.5 mmol) and TMEDA (1.4 mL, 9 mmol) in dry THF (15 mL)
under nitrogen atmosphere cooled to - 78 °C, a solution of n-butyl lithium (1.6 M in hexane, 2.8 mL, 4.5
mmol) was added dropwise. The resulting mixture was kept under stirring for 2 h, allowing the
temperature to reach - 30 °C. The temperature was brought at - 78 °C again and dimethylformamide (1
mL, 12.8 mmol) was added. The reaction was stirred for additional 2 h at room temperature. The resulting
mixture was then quenched with distilled water and extracted with AcOEt (x2). The organic phase was
washed with brine (x4), dried over Na2SO4 and concentrated under reduced pressure. The product was
purified by FC (AcOEt/CH2Cl2 7/3). The purified product was obtained as a brown foamy solid (43%
yield). 1H NMR (300 MHz, CDCl3) δ 9.95 (s, 1H), 7.53 (s, 1H), 7.29 – 7.15 (m, 6H), 6.99 (d, J = 8.8 Hz,
2H), 6.92 – 6.85 (m, 2H), 6.06 (s, 2H), 3.85 (s, 3H), 3.06 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 182.6,
160.6, 144.8, 136.5, 135.8, 135.2, 131.3, 130.5, 130.4 (2C), 128.6 (2C), 128.0, 127.9, 127.5 (2C), 119.8,
114.6 (2C), 55.4, 48.6, 32.3. HRMS (ESI) m/z: [M+H]+ Calcd for C22H21N4O2 373.1659; Found
373.1651.
3’-Phenetyl-5-(4-methoxyphenyl)- 1-methyl-1H,3’H-[2,4’ biimidazole]-2’-carbaldehyde (12)
To a solution of compound 10 (530 mg, 1.5 mmol) and TMEDA (1.3 mL, 9 mmol) in dry THF (15 mL)
under nitrogen atmosphere cooled to - 78 °C, a solution of n-butyl lithium (1.6 M in hexane, 2.8 mL, 4.5
mmol) was added dropwise. The resulting mixture was kept under stirring for 2 h, allowing the
temperature to reach - 30 °C. The temperature was brought at - 78 °C again and dimethylformamide (1
mL, 12.8 mmol) was added. The reaction was stirred for additional 2 h at room temperature. The resulting
mixture was then quenched with distilled water and extracted with AcOEt (x2). The organic phase was
washed with brine (x4), dried over Na2SO4 and concentrated under reduced pressure. The product was
purified by flash chromatography (AcOEt/CH2Cl2 gradient from 6/4 to 9/1) to give a brownish foamy
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 28
1
2
3
1
solid (77% yield). H NMR (400 MHz, CDCl3) δ 9.92 (s, 1H), 7.37 (s, 1H), 7.34 (d, J = 8.7 Hz, 2H),
7.21 (s, 1H), 7.15 – 7.07 (m, 3H), 7.04 (d, J = 8.7 Hz, 2H), 6.95 (d, br, J = 7.8 Hz, 2H), 5.16 (t, J = 6.9
Hz, 2H), 3.90 (s, 3H), 3.25 (s, 3H), 3.08 (t, J = 6.9 Hz, 2H); 13C NMR (100 MHz, CDCl3) δ 182.6, 160.5,
144.9, 138.4, 137.7, 136.0, 132.6, 130.9 (2C), 129.6 (2C), 128.8 (2C), 128.5, 127.0, 122.4 (2C), 115.1
(2C), 56.1, 47.2, 37.8, 33.5. HMRS (ESI): m/z: [M+H]+ Calcd for C23H23N4O2 387.1816; Found
387.1821.
4
5
6
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tert-Butyl (5-(3'-benzyl-5-(4-methoxyphenyl)-1-methyl-1H,3'H,3''H-[2,4':2',4''-terimidazol]-3''-
yl)pentyl)(methyl)carbamate (14)
Aldehyde 11 (40 mg, 0.11 mmol) was dissolved in DMF (0.11 mL). Primary amine 13 (29 mg, 0.13
mmol) was added and the resulting mixture was kept under stirring for 2 h at 70 °C under nitrogen
atmosphere. Potassium carbonate (23 mg, 0.16 mmol) and tosylmethyl isocyanide (26 mg, 0.13 mmol)
were sequentially added and the reaction was stirred for additional 2 h at 70 °C. After that, further
tosylmethyl isocyanide (26 mg, 0.13 mmol) was added. The reaction was left overnight at 70 °C under
stirring. The resulting mixture was partitioned between AcOEt and water. The aqueous phase was
extracted with AcOEt (x3) and the organic phase was washed with brine (x5), dried over Na2SO4 and
concentrated under reduced pressure, to give a residue that was purified by chromatography on neutral
aluminum oxide (AcOEt/ MeOH gradient from 97/3 to 92/8). The purified product was obtained as a
brown foamy solid (45% yield). 1H NMR (400 MHz, CDCl3 , 4:1 rotameric mixture) δ 7.62 (s, br, 1H),
7.42 (s, 1H), 7.22 (d, J = 8.7 Hz, 2H), 7.20 (s, br, 1H), 7.20 – 7.16 (m, 3H), 7.14 (s, 1H), 6.98 (d, J = 8.7
Hz, 2H), 6.85 – 6.78 (m, 2H), 5.56 (s, 2H), 4.14 (t, br, J = 7.3 Hz, 1.6H), 3.94 (t, br, J = 7.3 Hz, 0.4H),
3.86 (s, 3H), 3.29 (s, 3H), 3.18 (m, br, 2H), 2.82 (s, br, 3H), 1.82 (quint, br, J = 7.4 Hz, 0.4H), 1.67
13
(quint, br, J = 7.4 Hz, 1.6H), 1.57 – 1.46 (m, 2H), 1.46 (s, 9H), 1.26 (quint, br, J = 7.3 Hz, 2H); C
NMR (101 MHz, CDCl3, 4:1 rotameric mixture) δ 160.3, 156.4, 140.8, 139.8, 139.0, 137.9, 135.3, 131.5,
131.4, 130.8 (2C), 129.2 (2C), 128.2, 128.1, 127.3 (2C), 124.2, 122.8, 122.1, 114.9 (2C), 79.9, 56.0, 49.2,
49.2 (br), 47.7 and 46.5 (1C), 34.8, 33.0, 31.5 and 31.3 (1C), 29.2 (3C), 28.0 and 27.8 (1C, br), 24.4.
HRMS (ESI) m/z: [M+H]+ Calcd for C35H44N7O3 610.3500, Found 610.3508.
tert-Butyl (5-(3'-phenetyl-5-(4-methoxyphenyl)-1-methyl-1H,3'H,3''H-[2,4':2',4''-terimidazol]-3''-
yl)pentyl)(methyl)carbamate (15)
Aldehyde 12 (90 mg, 0.25 mmol) was dissolved in DMF (0.25 mL). Primary amine 13 (65 mg, 0.3 mmol)
was added and the resulting mixture was kept under stirring for 2h at 70 °C under nitrogen atmosphere.
20
ACS Paragon Plus Environment

Page 21 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Potassium carbonate (52 mg, 0.37 mmol) and tosylmethyl isocyanide (59 mg, 0.3 mmol) were
sequentially added and the reaction was stirred for additional 2 hours at 70 °C. After that, further
tosylmethyl isocyanide (59 mg, 0.3 mmol) was added. The reaction was left overnight at 70 °C under
stirring. The resulting mixture was then partitioned between AcOEt and water. The aqueous phase was
extracted with AcOEt (x3) and the organic phase was washed with brine (x5), dried over Na2SO4 and
concentrated under reduced pressure, to give a residue that was purified by chromatography on neutral
alumina oxide (AcOEt/ MeOH gradient from 97/3 to 92/8). The purified product was obtained as a brown
foamy solid (40% yield). 1H NMR (400 MHz, CDCl3) δ 7.60 (s, 1H), 7.34 (d, J = 8.3 Hz, 2H),7.31 (s,
1H), 7.21 (s, 1H), 7.17 (s, 1H), 7.15 – 7.08 (m, 3H), 7.00 (d, J = 8.4 Hz, 2H), 6.89 (d, J = 7.7 Hz, 2H),
4.72 (t, br, J = 6.8 Hz, 2H), 4.08 (t, br, J = 7.0 Hz, 2H), 3.85 (s, 3H), 3.46 (s, 3H), 3.15 (m, br, 2H), 2.83
(t, br, J = 7.0 Hz, 2H), 2.78 (s, 3H), 1.70 (quint, br, J = 7.2 Hz, 2H), 1.47 (m, br, 2H), 1.42 (s, 9H), 1.24
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
(quint, br, J = 7.2 Hz, 2H); C NMR (100 MHz, CDCl3), δ 160.3, 156.4, 140.2, 139.7, 139.5, 138.0,
135.3, 131.1 (2C), 130.8 (2C), 129.4 (2C), 129.0 (2C), 127.9, 127.2, 124.0, 122.8, 122.2, 115.0 (2C),
79.9, 56.0, 49.2, 46.8, 46.6, 38.0, 34.8, 33.7, 31.3, 29.1 (3C), 27.9, 24.4. HMRS (ESI) m/z: [M+H]+ Calcd
for C36H46N7O3 624.3657; Found 624.3648.
1-(3'-Benzyl-5-(4-methoxyphenyl)-1-methyl-1H,3'H,3''H-[2,4':2',4''-terimidazol]-3''-yl)-N-
methylmethanamine (RIm13)
To a solution of compound 14 (80 mg) in AcOEt (0.5 mL) under nitrogen atmosphere, HCl 4 N in AcOEt
(0.5 mL) was added at 0 °C. The reaction was allowed to room temperature and left under stirring
overnight. The reaction mixture was filtered on a Hirsch funnel and the solid was partitioned between
NaHCO3 sat aq solution and CH2Cl2. The aqueous phase was extracted with CH2Cl2 (x3). The collected
organic phase was dried over Na2SO4 and concentrated under reduced pressure. The product was
obtained as a brown foamy solid (quantitative yield). 1H NMR (400 MHz, CDCl3) δ 7.62 (s, br, 1H), 7.42
(s, 1H), 7.31 – 7.16 (m, 6H), 7.14 (s, 1H), 6.99 (d, J = 8.6 Hz, 2H), 6.85 – 6.75 (m, 2H), 5.54 (s, 2H),
4.13 (t, br, J = 7.2 Hz, 2H), 3.86 (s, 3H), 3.85 (m, br, 1H), 3.27 (s, 3H), 2.55 (t, br, J = 6.8 Hz, 2H), 2.43
(s, br, 3H), 1.65 (quint, br, J = 7.2 Hz, 2H), 1.46 (quint, br, J = 7.1 Hz, 2H), 1.29 (quint, br, J = 7.2 Hz,
13
2H); C NMR (101 MHz, CDCl3) δ 159.7, 140.1, 139.2, 138.3, 137.3, 134.6, 130.8 (2C), 130.1 (2C),
128.6 (2C), 127.6, 127.5, 126.7 (2C), 123.4, 122.1, 121.5, 114.3 (2C), 55.4, 51.7, 48.5, 45.8, 36.4, 32.3,
30.8, 29.1, 24.3. HRMS (ESI) m/z: [M+H]+ Calcd for C30H36N7O 510.2976; Found 510.2988. Anal.
Calcd for C30H35N7O: C, 70.70; H, 6.92; N, 19.24; Found: C, 70.48; H, 7.01; N, 19.32.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 28
1
2
3
4
5
6
1-(3'-Phenetyl-5-(4-methoxyphenyl)-1-methyl-1H,3'H,3''H-[2,4':2',4''-terimidazol]-3''-yl)-N-
methylmethanamine (RIm14)
To a solution of compound 15 (80 mg) in AcOEt (0.5 mL) under nitrogen atmosphere, HCl 4 N in AcOEt
(0.5 mL) was added at 0 °C. The reaction was allowed to room temperature and left under stirring
overnight. The reaction mixture was filtered on a Hirsch funnel and the solid was partitioned between
NaHCO3 sat aq solution and CH2Cl2. The aqueous phase was extracted with CH2Cl2 (x3). The collected
organic phase was dried over Na2SO4 and concentrated under reduced pressure. The final product was
obtained as a dark brown foamy solid (quantitative yield). 1H NMR (400 MHz, CDCl3) δ 7.63 (s, 1H),
7.37 (d, J = 8.1 Hz, 2H),7.34 (s, 1H), 7.23 (s, 1H), 7.20 (s, 1H), 7.18 – 7.10 (m, 3H), 7.02 (d, J = 8.1 Hz,
2H), 6.93 (d, J = 8.1 Hz, 2H), 4.73 (t, br, J = 6.8 Hz, 2H), 4.11 (t, br, J = 7.0 Hz, 2H), 3.88 (s, 3H), 3.49
(s, 3H), 2.84 (t, br, J = 6.8 Hz, 2H), 2.53 (t, br, J = 7.0 Hz, 2H),2.40 (s, 3H), 1.71 (quint, br, J = 7.3 Hz,
2H), 1.46 (quint, br, J = 7.4 Hz, 2H), 1.38 (m, br, 1H), 1.31 (quint, br, J = 7.3 Hz, 2H); 13C NMR (100
MHz, CDCl3), δ 160.3, 140.2, 139.7, 139.5, 138.1, 135.3, 131.2, 131.0, 130.8 (2C), 129.4 (2C), 129.0
(2C), 128.0, 127.2, 124.0, 122.9, 122.3, 115.0 (2C), 56.1, 52.3, 46.9, 46.6, 38.1, 37.0, 33.7, 31.5, 29.8,
24.9. HMRS (ESI): m/z [M+H]+ Calcd for C31H38N7O 524.3132; Found 524.3138. Anal. Calcd for
C31H37N7O: C, 71.10; H, 7.12; N, 18.72; Found: C, 71.28; H, 7.19; N, 18.89.
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biological assay materials. The HepG2 cell line was bought from ATCC (HB-8065, ATCC from LGC
Standards, Milan, Italy). Dulbecco's modified Eagle's medium, (DMEM), 96-Well plates, L-glutamine,
fetal bovine serum (FBS), phosphate buffered saline (PBS), penicillin/streptomycin, and
chemiluminescent reagent were purchased from Euroclone (Milan, Italy). Janus green were bought from
Abcam (Cambridge, UK), while the antibodies against anti-rabbit Ig-HRP, was purchased from Santa
Cruz Biotechnology Inc. (Santa Cruz, CA, US). Antibody against LDLR and TMB substrate were
obtained from Pierce (Rockford, IL, US). LDL-DyLight™ 550 (Cayman Chemical Company, Ann
Arbor, MI, US). Synthetic peptides were synthesized by the company GenScript (Piscataway, NJ, USA)
at >95% purity.
PCSK9-LDLR binding Assay. Compound RIm13, RIm14, compound 9, and peptide P5 (0.01 μM –
100.0 μM) were tested using the in vitro PCSK9-LDLR binding assay (CycLex Co., Nagano, Japan),
following the manufacture instructions. The absorbance at 450 nm was measured using the Synergy H1
fluorescent plate reader (Biotek, Bad Friedrichshall, Germany). In particular, for the in vitro screening
of the synthetic PCSK9-LDLR inhibitors, at different concentrations, were added to the appropriate
22
ACS Paragon Plus Environment

Page 23 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
amount of His-tagged PCSK9 in the wells that had been coated with recombinant LDLR-AB domain,
followed by evaluation of inhibitory effect on PCSK9-LDLR interaction by measuring the amount of
His-tagged PCSK9 on the wells which is correlated to the absorbance signals at 450 nm, which were
measured using the Synergy H1 fluorescent plate reader (Biotek, Bad Friedrichshall, Germany).
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cell culture conditions and ICW assay. The HepG2 cell line was cultured in DMEM high glucose with
stable L-glutamine supplemented with 10% FBS, 100 U/mL penicillin, 100.0 µg/mL streptomycin and
incubated at 37 °C under 5% CO2 atmosphere. HepG2 cells were used for no more than 20 passages after
thawing, because the increase of the number of passages may change the cell characteristics and impair
assay results. A total of 3.0 x 104 HepG2 cells/well were seeded in 96-well plates, respectively. The
following day, cells were washed with PBS and then starved overnight (O/N) in DMEM without FBS.
HepG2 cells were treated with 4.0 μg/mL PCSK9 (K9) and 4.0 μg/mL K9 + RIm13 compound (0.1 –
10.0 µM), compound 9 (10.0 µM), peptide P5 (1.0 µM), and vehicle (H2O) for 2 h at 37 °C under 5%
CO2 atmosphere. Treated HepG2 cells were fixed in 4% paraformaldehyde for 20 min at room
temperature (RT). Cells were washed 5 times with 100.0 µL of PBS/well (each wash was for 5 min at
RT) and the endogenous peroxides activity was quenched adding 3% H2O2 in PBS for 20 min at RT.
Non-specific sites were blocked with 100.0 µL/well of 5% BSA in PBS for 1.5 h at RT. LDLR primary
antibody solution (1:3000 in 5% BSA in PBS, 25 µL/well) was incubated O/N at 4 °C. Subsequently, the
primary antibody solution was discarded and each sample was washed 5 times with 100.0 µL/well of
PBS (each wash was for 5 min at RT). Goat anti-rabbit Ig-HRP secondary antibody solution (1:6000 in
5% BSA in PBS, 50.0 µL/well) was added and incubated 1 h at RT. The secondary antibody solution
was washed 5 times with 100.0 µL/well of PBS (each wash for 5 min at RT). Freshly prepared TMB
substrate (100.0 µL/well) was added and the plate was incubated at RT until desired colour was
developed. The reaction was then stopped with 2 M H2SO4 and the absorbance at 450 nm was measured
using the Synergy H1 fluorescent plate reader from Biotek. Cells were stained by adding 1 x Janus green
stain, incubating for 5 min at RT. The dye was removed and the sample washed 5 times with water.
Afterward 0.1 mL 0.5 M HCl per well were added and incubated for 10 min. After 10 sec shaking, the
OD at 595 nm was measured using the Synergy H1 fluorescent plate reader from Biotek.
Fluorescent LDL uptake cell based assay. HepG2 cells (3.0 x 104/well) were seeded in black 96-well
plates and kept in complete growth medium for 2 d before treatment. The third day, they were treated
with 4.0 μg/mL K9 and 4.0 μg/mL K9 + RIm13 compound (0.1-10.0 µM), compound 9 (10.0 µM),
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 28
1
2
3
4
5
6
7
8
9
peptide P5 (1.0 µM), and, and vehicle (H2O) for 2 h with at 37 °C under 5% CO2 atmosphere. At the end
of the treatments, the culture medium was replaced with 50.0 μl/well LDL-DyLight™ 550 working
solution (Cayman Chemical Company, Ann Arbor, MI, US). The cells were additionally incubated for 2
h at 37 °C and then the culture medium was aspirated and replaced with PBS (100.0 μl/well). The degree
of LDL uptake was measured using the Synergy H1 fluorescent plate reader from Biotek (excitation and
emission wavelengths 540 and 570 nm, respectively).
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Statistical analysis of biological assays. Data are presented as mean ± s.d. using GraphPad Prism 6
(GraphPad, La Jolla, CA, USA). Statistical analyses were carried out by t student test and ANOVA. P-
values < 0.05 were considered to be significant.
ASSOCIATED CONTENT
Supporting Information
”Supporting Information.pdf” contains Figures S1-S3, Scheme S1, Tables S1-S3, and NMR spectra.
“Molecular Formula Strings” can be found in the file MFS.csv.
AUTHOR INFORMATION
Corresponding Author. Alessandra Silvani and Giovanni Grazioso equally contributed to this work.
Email: alessandra.silvani@unimi.it (A.S.), giovanni.grazioso@unimi.it (G.G.)
Author Contributions. The manuscript was written through contributions of all authors. All authors
have given approval to the final version of the manuscript.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENTS
GG thanks Prof. Stefano Moro and Dr. Mattia Sturlese for the availability of the suMD algorithm. We
acknowledge the CINECA and the Regione Lombardia award under the LISA initiative, for the
availability of high performance computing resources and support. GG gratefully acknowledge the
support of “NVIDIA Corporation” with the donation of the Titan Xp GPU: it was utilized for the majority
of the calculations here reported.
24
ACS Paragon Plus Environment

Page 25 of 28
Journal of Medicinal Chemistry
1
2
3
ABBREVIATIONS
4
5
6
7
8
CVD, cardio-vascular disease; HRP, horseradish peroxidase; mAbs, monoclonal antibodies; MD,
molecular dynamics; MM-GBSA, Molecular Mechanics-Generalized Born Surface Area; suMD,
supervised Molecular Dynamics; LDL-C, low-density lipoprotein-cholesterol; LDLR, low-density
lipoprotein receptors; O/N, overnight; PCSK9, Proprotein Convertase Subtilisin/Kexin-9; PPI, protein-
protein interaction; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; PBS,
phosphate buffered saline; RT, room temperature; TMB, 3,3',5,5'-tetrametilbenzidina.
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
1.
Austin, M. A.; Hutter, C. M.; Zimmern, R. L.; Humphries, S. E. Familial Hypercholesterolemia and
Coronary Heart Disease: a Huge Association Review. Am. J. Epidemiol. 2004, 160, 421-429.
2. Civeira, F. Guidelines for the Diagnosis and Management of Heterozygous Familial
Hypercholesterolemia. Atherosclerosis 2004, 173, 55-68.
3.
Seidah, N. G. PCSK9 as a Therapeutic Target of Dyslipidemia. Expert Opin. Ther. Targets 2009, 13, 19-
28
4.
Zaid, A.; Roubtsova, A.; Essalmani, R.; Marcinkiewicz, J.; Chamberland, A.; Hamelin, J.; Tremblay, M.;
Jacques, H.; Jin, W.; Davignon, J.; Seidah, N. G.; Prat, A. Proprotein Convertase Subtilisin/kexin Type 9 (PCSK9):
Hepatocyte-specific Low-density Lipoprotein Receptor Degradation and Critical Role in Mouse Liver
Regeneration. Hepatology 2008, 48, 646-654.
5.
Zhang, D. W.; Lagace, T. A.; Garuti, R.; Zhao, Z.; McDonald, M.; Horton, J. D.; Cohen, J. C.; Hobbs, H.
H. Binding of Proprotein Convertase Subtilisin/kexin Type 9 to Epidermal Growth Factor-like repeat a of Low
Density Lipoprotein Receptor Decreases Receptor Recycling and Increases Degradation. J. Biol. Chem. 2007, 282,
18602-18612.
6.
Grefhorst, A.; McNutt, M. C.; Lagace, T. A.; Horton, J. D. Plasma PCSK9 Preferentially Reduces Liver
LDL Receptors in Mice. J. Lipid Res. 2008, 49, 1303-1311.
7. Lagace, T. A.; Curtis, D. E.; Garuti, R.; McNutt, M. C.; Park, S. W.; Prather, H. B. Secreted PCSK9
Decreases the Number of LDL Receptors in Hepatocytes and in Livers of Parabiotic Mice. J. Clin. Invest. 2006,
116, 2995-3005.
8.
Giugliano, R. P.; Pedersen, T. R.; Park, J. G.; De Ferrari, G. M.; Gaciong, Z. A.; Ceska, R.; Toth, K.;
Gouni-Berthold, I.; Lopez-Miranda, J.; Schiele, F.; Mach, F.; Ott, B. R.; Kanevsky, E.; Pineda, A. L.; Somaratne,
R.; Wasserman, S. M.; Keech, A. C.; Sever, P. S.; Sabatine, M. S. Clinical Efficacy and Safety of Achieving very
Low LDL-cholesterol Concentrations with the PCSK9 Inhibitor Evolocumab: a Prespecified Secondary Analysis
of the FOURIER Trial. Lancet 2017, 390, 1962-1971.
9.
Bottomley, M. J.; Cirillo, A.; Orsatti, L.; Ruggeri, L.; Fisher, T. S.; Santoro, J. C.; Cummings, R. T.;
Cubbon, R. M.; Lo Surdo, P.; Calzetta, A.; Noto, A.; Baysarowich, J.; Mattu, M.; Talamo, F.; De Francesco, R.;
Sparrow, C. P.; Sitlani, A.; Carfi, A. Structural and Biochemical Characterization of the Wild Type PCSK9-
EGF(AB) Complex and Natural Familial Hypercholesterolemia Mutants. J. Biol. Chem. 2009, 284, 1313-1323.
10.
Stucchi, M.; Grazioso, G.; Lammi, C.; Manara, S.; Zanoni, C.; Arnoldi, A.; Lesma, G.; Silvani, A.
Disrupting the PCSK9/LDLR Protein-Protein Interaction by an Imidazole-based Minimalist Peptidomimetic. Org.
Biomol. Chem. 2016, 14, 9736-9740.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 28
1
2
3
4
5
6
7
8
11.
Stucchi, M.; Cairati, S.; Cetin-Atalay, R.; Christodoulou, M. S.; Grazioso, G.; Pescitelli, G.; Silvani, A.;
Yildirim, D. C.; Lesma, G. Application of the Ugi Reaction with Multiple Amino Acid-derived Components:
Synthesis and Conformational Evaluation of Piperazine-based Minimalist Peptidomimetics. Org. Biomol. Chem.
2015, 13, 4993-5005.
12.
Taechalertpaisarn, J.; Zhao, B.; Liang, X.; Burgess, K. Small Molecule Inhibitors of the PCSK9.LDLR
Interaction. J. Am. Chem. Soc. 2018, 140, 3242-3249.
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13. Pettersen, D.; Fjellström, O. Small Molecule Modulators of PCSK9 – A Literature and Patent Overview.
Bioorg. Med. Chem. Lett. 2018, 28, 1155-1160.
14. Xu, S.; Luo, S.; Zhu, Z.; Xu, J. Small Molecules as Inhibitors of PCSK9: Current Status and Future
Challenges. Eur. J. Med. Chem. 2019, 162, 212-233.
15.
Disney, M. D. Inhibiting Translation One Protein at a Time. Trends Biochem. Sci. 2017, 42, 412-413.
16.
Londregan, A. T.; Aspnes, G.; Limberakis, C.; Loria, P. M.; McClure, K. F.; Petersen, D. N.; Raymer, B.;
Ruggeri, R. B.; Wei, L.; Xiao, J.; Piotrowski, D. W. Discovery of N-(piperidin-3-yl)-N-(pyridin-2-
yl)piperidine/piperazine-1-carboxamides as Small Molecule Inhibitors of PCSK9. Bioorg. Med. Chem. Lett. 2018,
28, 3685-3688.
17.
Salmaso, V.; Sturlese, M.; Cuzzolin, A.; Moro, S. Exploring Protein-Peptide Recognition Pathways Using
a Supervised Molecular Dynamics Approach. Structure 2017, 25, 655-662.e2.
18. Cuzzolin, A.; Sturlese, M.; Deganutti, G.; Salmaso, V.; Sabbadin, D.; Ciancetta, A.; Moro, S. Deciphering
the Complexity of Ligand–Protein Recognition Pathways Using Supervised Molecular Dynamics (SuMD)
Simulations. J. Chem. Inf. Mod. 2016, 56, 687-705.
19.
Ciancetta, A.; Sabbadin, D.; Federico, S.; Spalluto, G.; Moro, S. Advances in Computational Techniques
to Study GPCR-Ligand Recognition. Trends Pharmacol. Sci. 2015, 36, 878-890.
20. Sabbadin, D.; Moro, S. Supervised Molecular Dynamics (SuMD) as a Helpful Tool to Depict GPCR–
Ligand Recognition Pathway in a Nanosecond Time Scale. J. Chem. Inf. Mod. 2014, 54, 372-376.
21. Lammi, C.; Zanoni, C.; Aiello, G.; Arnoldi, A.; Grazioso, G. Lupin Peptides Modulate the Protein-Protein
Interaction of PCSK9 with the Low Density Lipoprotein Receptor in HepG2 Cells. Sci. Rep. 2016, 6, 29931.
22. Shao, J.; Tanner, S. W.; Thompson, N.; Cheatham, T. E. Clustering Molecular Dynamics Trajectories: 1.
Characterizing the Performance of Different Clustering Algorithms. J. Chem. Theory Comput. 2007, 3, 2312-2334.
23. Lammi, C.; Zanoni, C.; Arnoldi, A. A simple and High-throughput in-cell Western Assay Using HepG2
Cell Line for Investigating the Potential Hypocholesterolemic Effects of Food Components and Nutraceutics. Food
Chem. 2015, 169, 59-64.
24.
Zhang, Y.; Eigenbrot, C.; Zhou, L.; Shia, S.; Li, W.; Quan, C.; Tom, J.; Moran, P.; Di Lello, P.; Skelton,
N. J.; Kong-Beltran, M.; Peterson, A.; Kirchhofer, D. Identification of a Small Peptide that Inhibits PCSK9 Protein
Binding to the Low Density Lipoprotein Receptor. J. Biol. Chem. 2014, 289, 942-955.
25.
Grazioso, G.; Bollati, C.; Sgrignani, J.; Arnoldi, A.; Lammi, C. The First Food-Derived Peptide Inhibitor
of the Protein-Protein Interaction between Gain-Of-Function PCSK9D374Y and the LDL Receptor. J. Agric. Food
Chem. 2018, 66, 10552-10557.
26.
Lammi, C.; Sgrignani, J.; Roda, G.; Arnoldi, A.; Grazioso, G. Inhibition of PCSK9D374Y/LDLR Protein–
Protein Interaction by Computationally Designed T9 Lupin Peptide. ACS Med. Chem. Lett. 2019, 10, 425–430.
27. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Kollman, P. A. Application of RESP charges to Calculate
Conformational Energies, Hydrogen Bond Energies, and Free Energies of Solvation. J. Am. Chem. Soc. 1993, 115,
9620-9631.
26
ACS Paragon Plus Environment

Page 27 of 28
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
28.
Maier, J. A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; Hauser, K. E.; Simmerling, C. ff14SB:
Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB. J. Chem. Theory Comput.
2015, 11, 3696-3713.
29.
Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development and Testing of a General
Amber Force Field. J. Comput. Chem. 2004, 25, 1157-1174.
9
30. McNutt, M. C.; Kwon, H. J.; Chen, C.; Chen, J. R.; Horton, J. D.; Lagace, T. A. Antagonism of Secreted
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PCSK9 Increases Low Density Lipoprotein Receptor Expression in HepG2 Cells. J. Biol. Chem. 2009, 284, 10561-
10570.
31.
Harvey, M. J.; Giupponi, G.; Fabritiis, G. D. ACEMD: Accelerating Biomolecular Dynamics in the
Microsecond Time Scale. J. Chem. Theory Comput. 2009, 5, 1632-1639.
32. Case, D. A.; Cerutti, D. S.; T.E. Cheatham, I.; Darden, T. A.; Duke, R. E.; Giese, T. J.; Gohlke, H.; Goetz,
A. W.; Greene, D.; Homeyer, N.; Izadi, S.; Kovalenko, A.; Lee, T. S.; LeGrand, S.; P. Li, C. L.; Liu, J.; Luchko,
T.; Luo, R.; Mermelstein, D.; Merz, K. M.; Monard, G.; Nguyen, H.; Omelyan, I.; Onufriev, A.; Pan, F.; Qi, R.;
Roe, D. R.; Roitberg, A.; Sagui, C.; Simmerling, C. L.; Botello-Smith, W. M.; Swails, J.; Walker, R. C.; Wang,
J.; Wolf, R. M.; Wu, X.; Xiao, L.; York, D. M.; Kollman, P. A. AMBER 2017; University of California: San
Francisco, 2017.
33.
38.
Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual Molecular Dynamics. J. Mol. Graph. 1996, 14, 33-
34.
Roe, D. R.; Cheatham, T. E., 3rd. PTRAJ and CPPTRAJ: Software for Processing and Analysis of
Molecular Dynamics Trajectory Data. J. Chem. Theory Comput. 2013, 9, 3084-3095.
35. Miller, B. R., 3rd; McGee, T. D., Jr.; Swails, J. M.; Homeyer, N.; Gohlke, H.; Roitberg, A. E.
MMPBSA.py: An Efficient Program for End-State Free Energy Calculations. J. Chem. Theory Comput. 2012, 8,
3314-3321.
36.
Gardner, R. A.; Ghobrial, G.; Naser, S. A.; Phanstiel, O. t. Synthesis and Biological Evaluation of New
Acinetoferrin Homologues for Use as Iron Transport Probes in Mycobacteria. J. Med. Chem. 2004, 47, 4933-4940.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 28
1
2
3
Table of Contents Graphic
4
5
6
7
8
9
10
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28
ACS Paragon Plus Environment