Dual Neutral Sphingomyelinase-2/Acetylcholinesterase Inhibitors for the Treatment of Alzheimer's Disease

Article abstract of DOI:10.1021/acschembio.0c00311

We report the discovery of a novel class of compounds that function as dual inhibitors of the enzymes neutral sphingomyelinase-2 (nSMase2) and acetylcholinesterase (AChE). Inhibition of these enzymes provides a unique strategy to suppress the propagation

Full text of DOI:10.1021/acschembio.0c00311

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Article
Dual Neutral Sphingomyelinase-2/Acetylcholinesterase
Inhibitors for the treatment of Alzheimer’s disease
Tina Bilousova, Bryan J. Simmons, Rachel R. Knapp, Chris J. Elias, Jesus Campagna,
Mikhail Melnik, Sujyoti Chandra, Samantha Focht, Chunni Zhu, Kanagasabai Vadivel, Barbara
Jagodzinska, Whitaker Cohn, Patricia Spilman, Karen H. Gylys, Neil K. Garg, and Varghese John
ACS Chem. Biol., Just Accepted Manuscript • Publication Date (Web): 30 Apr 2020
Downloaded from pubs.acs.org on May 1, 2020
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 34
ACS Chemical Biology
1
2
3
4
5
6
Dual Neutral Sphingomyelinase-2/Acetylcholinesterase Inhibitors for the
Treatment of Alzheimer’s disease
7
8
9
Tina Bilousova1,3, Bryan J. Simmons2, Rachel R. Knapp2, Chris J. Elias1, Jesus
Campagna1, Mikhail Melnik3, Sujyoti Chandra1, Samantha Focht1, Chunni Zhu1,
Kanagasabai Vadivel1, Barbara Jagodzinska1, Whitaker Cohn1, Patricia Spilman1,
Karen H. Gylys3, Neil K. Garg2*, Varghese John1*
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
1Drug Discovery Laboratory, Department of Neurology, Mary S. Easton Center for
Alzheimer’s Disease Research, University of California, Los Angeles, CA 90095, United
States
2Department of Chemistry and Biochemistry, University of California, Los Angeles, CA
90095, United States
3School of Nursing, University of California, Los Angeles, CA 90095, United States
*Corresponding authors
ABSTRACT
We report the discovery of a novel class of compounds that function as dual inhibitors of
the enzymes neutral sphingomyelinase-2 (nSMase2) and acetylcholinesterase (AChE).
Inhibition of these enzymes provides a unique strategy to suppress the propagation of
tau pathology in treatment of Alzheimer’s disease (AD). We describe the key SAR
elements that affect relative nSMase2 and/or AChE inhibitor effects and potency, in
addition to the identification of two analogs that suppress the release of tau-bearing
exosomes in vitro and in vivo. Identification of these novel dual nSMase2/AChE
inhibitors represents a new therapeutic approach to AD and has the potential to lead to
the development of truly disease-modifying therapeutics.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 34
1
2
3
4
INTRODUCTION
5
6
7
8
Alzheimer’s disease (AD) is the most prevalent age-related neurodegenerative
disorder and, currently, there are no effective disease-modifying therapies available for
the treatment of AD. The number of AD cases in the US is ~5.8 million patients and this
number is expected to rise to 50 million by 2050. The estimated global socioeconomic
costs of AD and related dementias are predicted to reach $2 trillion by the year 2030.1
AD brain tissue is characterized by the presence of senile plaques composed mainly
of aggregated amyloid-β peptide (Aβ), neurofibrillary tangles (NFTs) composed of
pathological forms of the microtubule-stabilizing protein tau, chronic neuroinflammation,
and loss of neurons.2 Clinically, it is thought that the underlying mechanisms of disease
are initiated as early as 20 years before the onset of signs and symptoms. During this
asymptomatic period, proteopathic proteins are believed to accumulate, leading to
structural alterations and the neuronal dysfunction and loss that leads frequently to Mild
Cognitive impairment (MCI). MCI then progresses to full-blown AD-related memory
deficits, decline of other cognitive skills, and in advanced AD, the inability to participate
in activities of daily living.3
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
While the exact mechanisms of disease progression have not been fully elucidated,
it is thought that increased Aβ production at the synapse and/or impaired clearance,
results in synaptic loss. Contemporaneously and in conjunction with Aβ accumulation,
there is hyperphosphorylation and oligomerization of tau that eventually leads to
neuronal toxicity, NFT formation, and neuronal cell death. Diseased neurons can
release these toxic phosphorylated forms of tau (p-tau) in proteopathic seeds, which
can then be taken up by surrounding or interconnected neurons, leading to templating
and propagation of the pathological aggregates in prion-like fashion. The propagation of
the disease follows a spatiotemporal pattern with Aβ plaques first appearing in the basal
forebrain, then the frontal, temporal and occipital lobes of the cortex. NFTs form in the
locus coeruleus and in the allocortex of the medial temporal lobe.4 Both Aβ and tau
pathologies spread through the brain during disease progression.2
Given the importance of tau, significant attention is now being paid to the
mechanisms of pathological tau spread in AD with the goal of identifying targets for
novel therapies to prevent disease progression.5 Historically, Aβ pathology has been
ACS Paragon Plus Environment

Page 3 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
thought to be causative in AD,6-7 but multiple clinicopathological evaluations, as well as
recent in vivo imaging studies, suggest that the cognitive status of AD patients
correlates most closely with region-specific brain atrophy and distribution of the hyper-
phosphorylated and aggregated pathological forms of tau that lead to the formation of
NFTs.8-11 Longitudinal studies have confirmed that propagation of tau pathology
correlates significantly with cognitive decline.12-13 These data suggest that suppression
of propagation of tau pathology in AD may have a disease-modifying effect.
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
Prompted by the findings described above, we undertook a screening effort to
identify inhibitors of tau propagation. As will be discussed below, this led to the
discovery of dual inhibitors of two enzymes: neutral sphingomyelinase 2 (nSMase2)
and acetylcholine esterase (AChE), a key enzyme implicated in AD. In our in vitro
studies, the identified dual inhibitors prevented the spreading of tau in cell culture
systems using assays we have previously reported.14
NSMase2 is an enzyme responsible for hydrolysis of sphingomyelin to
ceramide/phosphatidylcholine and has been implicated in the spread of AD pathology.
Pharmacological inhibition or genetic depletion of nSMase2 has been shown to
suppress progression of both Aβ and tau pathology in animal models.15-17 nSMase2
activity plays an important role for normal brain function, but its activity increases with
age leading to dysregulation in sphingomyelin turnover.18-22 There is over-activation of
nSMase2 in AD, and brain ceramide levels have been found to be elevated in AD
patient cerebrospinal fluid (CSF), compared to age-matched control subjects.23 The
ceramide/sphingomyelin imbalance is greater in individuals that express apolipoprotein
E4 (ApoE4), the major genetic risk factor for sporadic, late onset AD. 24 NSMase2 is a
key enzyme involved in biogenesis of brain exosomes through the Endosomal Sorting
Complex Required for Transportation (ESCRT)-independent pathway.25
Brain
exosomes are a type of extracellular vesicle (EV), that are 40-150 nm in diameter and
are released by brain cells when multivesicular endosomes fuse with the plasma
membrane.25-26 They are involved in normal brain function, but a subset produced by
the ESCRT-independent pathway involving nSMase2 have been shown to carry
disease-propagating proteopathic seeds, such as tau oligomers, in AD.14-15, 17, 27 Tau
oligomers have been found to be associated with neuronal exosomes in both cell
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 34
1
2
3
4
5
6
culture medium and transgenic AD/tauopathy model brain tissue, as well as in AD
patient plasma and CSF.15, 28-33
Despite recent progress, the current armamentarium of nSMase2 inhibitors have
poor drug-like properties and oral brain permeability.34-35 Thus, our initial goal was to
identify nSMase2 inhibitors that overcome these limitations for the development of
preclinical candidates for AD. Using an nSMase2 inhibitor screening assay, we
identified a novel furoindoline compound ‘validated hit’. Further structural alterations of
this initial hit generated compounds that resulted in the identification of novel dual
inhibitor analogs that not only inhibit nSMase2 activity, but also inhibit
acetylcholinesterase (AChE) enzyme activity and suppress p-tau propagation.
AChE inhibitors (AChEIs) are currently one of only two classes of FDA-approved AD
therapeutics; they have demonstrated amelioration of symptoms in AD, being most
effective in mild and moderate AD.36 Inhibition of AChE leads to increased levels of
acetylcholine (ACh) at the synapse and in brain parenchyma, and provides support for
cholinergic synaptic plasticity even during progressive loss of cholinergic innervation
from the basal forebrain.37 However, AChEI’s treatment only provides short term
benefits in AD and does not block the progression of the disease.
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
The dual nSMase2/AChE inhibitors we describe herein represent a new therapeutic
paradigm enabling target engagement by a single agent of two enzymes in the brain
that play a role in spread of tau pathology. 38 This promising approach could lead to an
effective treatment for AD. These agents have the potential to be disease-modifying by
suppressing disease progression through exosome-mediated tau propagation, while
also providing symptomatic relief through support of ACh-mediated cognitive
enhancement. Interestingly, in mild to moderate AD, there is significantly decreased
cholinergic activity and high levels of p-tau in CSF-derived exosomes, thus treating
patients in these stages of the disease with dual nSMase2/AChE inhibitors could be
highly beneficial.33
We propose a mechanism of action for these dual inhibitors
involving nSMase2 mediated suppression of tau oligomer release in brain exosomes by
presynaptic neurons, increased ACh levels at the synapse through AChE inhibition,
along with the suppression of tau oligomer uptake through ACh receptors by
postsynaptic neurons. 38
ACS Paragon Plus Environment

Page 5 of 34
ACS Chemical Biology
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
Screening for and optimization of selective nSMase2 and dual nSMase2/AChE
inhibitors. We initiated the present study by screening a small molecule compound
library for effects on nSMase2 activity. The ~70 compound library were mostly fused
indolines we had prepared previously through interrupted Fischer indolization
methodology that have structural resemblance to phenserine and known AChE
inhibitors. Using an Amplex Red neutral sphingomyelinase enzyme activity assay
several hits were identified that inhibited >60% nSMase2 activity at a concentration of
50 µM, as shown in the scatterplot (Figure 1a). The known nSMase2 inhibitor
cambinol14 was used as a positive control for the screening assay. After retesting, one
hit (Figure 1a) was validated at 50 µM and selected for further hit-to-lead optimization.
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
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 34
1
2
3
4
5
6
7
8
9
Figure 1. Screening and identification of novel dual nSMase2/AChE inhibitors: a)
nSMase2 inhibitor screening using Amplex Red-coupled assay revealed several hits
that inhibited activity ≥60%; b) hit-to-lead optimization of the validated hit shows removal
of the nitrogen group from the furoindoline aryl ring (red arrow) and addition of nitrogen
to the carbamate phenyl ring at either the 3 or 4 positions (blue arrow) results in
enhanced potency for nSMase2 inhibition and varied AChE inhibition; c) Dose-response
analysis for compounds 1, 8, and 11 in the nSMase2 assay; and d) Dose-response
analysis for compounds 1, 8, and 11 in the AChE assay.
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
Optimization efforts led to the synthesis and evaluation of analogs. Our synthetic
approach to the validated hit and analogs will be described subsequently, but a
summary of our optimization effort leading to key dual inhibitor analogs 8 and 11 is
shown in Figure 1b. Given the structural similarity between the validated hit and known
AChE inhibitor phensvenine (1) we initially prepared this analog to check if it was also
an nSMase2 inhibitor. Dose-response analysis revealed that phensvenine (1) indeed
has nSMase2 inhibitory activity (Figure 1c) but is a more potent inhibitor of AChE with
an IC50 = 0.5 µM (Figure 1d). In contrast, the dual inhibitors 8 and 11 were more potent
nSMase2 inhibitors (IC50 = 0.5 µM) with varying AChE inhibitory activity (Figure 1c &
1d). Interestingly, replacement of the oxygen in the furoindoline ring of phensvenine (O
 N-CH3), as seen in phenserine a known potent AChE inhibitor, results in the loss of
any detectable nSMase2 inhibitory activity (IC50 >50 μM). Posiphen, the (+)-enantiomer
of phenserine, is reported to be a weak AChE inhibitor and did not show detectable
nSMase2 inhibitor activity (IC50 >50 μM).39 We therefore focused on the (–)-enantiomer
for the dual inhibitor optimization effort.
The synthesis of the validated hit and analogs were made possible by using the
interrupted Fischer indolization reaction and variants thereof.40 As an example, the
interrupted Fischer indolization route to (–)-phensvenine (1) is shown in Scheme 1.
Treatment of aryl hydrazine 17 and lactol 18 with acetic acid furnished furoindoline 19 in
67% yield. Subsequent N-methylation provided 23. At this stage, the enantiomers
could be resolved using chiral SFC. As depicted for the (–)-enantiomer, O-deprotection
was achieved using BBr3, thus furnishing (–)-26. Lastly, treatment with NaH and
phenylisocyanate furnished (–)-phensvenine (1). It should be noted that (–)-enantiomers
were specifically targeted given the known stereospecificity of phenserine for AChE
inhibition.39, 41
ACS Paragon Plus Environment

Page 7 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Me
Me
MeO
Me
HO
•HCl
NaH, MeI
MeO
MeO
AcOH:H2O
O
O
+
NH2
O
N
H
DMF, 0 °C to 23 °C
5 h
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
N
H
H
N
60 °C, 5 h
H
Me
(±)-19
(±)-23
(67% yield)
(81% yield)
17
18
Me
Me
H
NaH
PhNCO
HO
N
O
1. Chiral SFC
O
O
O
H
H
N
N
2. BBr3, CH2Cl2
23 °C, 1 h
THF, 23 °C, 16 h
(30% yield)
Me
Me
(–)-26
(–)-1
Scheme 1. Synthesis of (─)-phensvenine (1)
The synthetic route was readily amenable to the synthesis of analogs, particularly
by exploiting intermediate (–)-26 as a means to access different carbamate substitution
patterns. Scheme 2 provides an example, in the context of the synthesis of analog (–)-8.
Alcohol (–)-26 readily underwent reaction with 29 to furnish carbonate (–)-30. Upon
treatment with 4-aminopyridine (32), (–)-8 was obtained in 61% yield. The syntheses of
other carbamate analogs are provided in the SI.
NH2
O
N
O
O
Me
Me
H
Me
N
O
HO
O
O
O
O
pyridine
32
N
O
O
N
N
O
+
N
O
O
O
H
H
N
N
H
N
CH3CN, 23 °C
(99% yield)
CH2Cl2, 23 °C
(61% yield)
O
O
O
Me
Me
Me
(–)-8
(–)-26
(–)-30
29
Scheme 2. Synthesis of furoindoline analog (─)-8
Scheme 3 shows the routes used to prepare two analogs bearing substitution on
the furoindoline ring. Interrupted Fischer indolization using hydrazine 20 and lactol 21
proceeded smoothly to give 22 as a mixture of diastereomers in racemic form. Upon
methylation, diastereomers 24 and 25 were accessed and could be separated by silica
gel chromatography. Each diastereomer was then elaborated through a sequence
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 34
1
2
3
4
5
6
involve separation by chiral SFC, demethylation, and carbamate formation. This
unoptimized synthetic sequence delivered (–)-9 and (–)-10, respectively for initial
biological 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
Me
Me
Me
Me
Me
Me
MeO
MeO
NaH, MeI
• HCl
NH2
MeO
MeO
AcOH:H2O (1:1)
Me
+
O
O
O
+
Me
N
O
DMF
0 °C to 23 °C, 5 h
N
H
H
N
H
N
H
60 °C, 4.5 h
(62% yield)
H
HO
Me
Me
20
21
(±)-22
Major
(±)-24
Minor
(±)-25
(67% yield)
Me
H
Me
Me
NaH
PhNCO
Me
N
O
1. Chiral SFC
HO
Major
(±)-24
O
O
O
H
N
2. BBr3, CH2Cl2
23 °C, 1 h
H
THF, 23 °C, 16 h
(16% yield)
N
Me
Me
(–)-9
(–)-27
Me
H
N
Me
NaH
PhNCO
Me
O
Me
1. Chiral SFC
HO
Minor
(±)-25
O
O
O
H
N
2. BBr3, CH2Cl2
23 °C, 1 h
THF, 23 °C, 16 h
(17% yield)
Me
H
N
Me
(–)-10
(–)-28
Scheme 3. Synthesis of furoindoline analogs (─)-9 and (─)-10
In total, sixteen analogs were prepared as part of the optimization efforts. The
structures of these analogs, dose-response analysis results, predicted brain
permeability data, and binding efficiency to human serum albumin (HSA) are shown in
Table 1. Our SAR analysis reveals structural elements in this series required for
enhanced nSMase2 and/or AChE inhibition. Substitutions in the carbamate phenyl ring
pointed to a critical role for positions 3 and 4 as key control elements for nSMase2
and/or AChE inhibition. Substitution in the 4-position leads to increased selectivity for
nSMase2 inhibition (such as for compounds 2, 4 and 8). In contrast, substitution in the
3-position leads to increased selectivity for AChE inhibition (such as 3 and 11).
Introduction of electron donating groups (3, 6, and 7) at position 3 increased potency of
AChE inhibition, while an electron withdrawing group (such as in 14) resulted in
decreased potency. Importantly, replacement of the phenyl ring with a pyridyl ring in the
carbamate moiety generally decreased potency of AChE inhibition and markedly
enhanced potency for nSMase2 inhibition (e.g. 8, 11, 12). Most of the analogs (except
4) showed high predicted brain permeability by in silico StarDrop analysis and in a
parallel artificial membrane permeability assay (PAMPA). A low degree of binding to
ACS Paragon Plus Environment

Page 9 of 34
ACS Chemical Biology
1
2
3
4
5
6
human serum albumin (HSA) measured for most of the compounds, especially 8, 11,
12, and 13.
7
8
9
Table 1. Structure and characteristics of carbamate furoindoline analogs.
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
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 34
1
2
3
4
5
6
7
8
9
a
a
b
IC50
MW
(g/mol)
IC50
AChE
(µM)
CNS perm
(PAMPA)
HSA
Compound
#
cLogP
Structure
nSMase2
(µM)
(% unbound)
Me
H
N
O
O
324.15
>10
0.7
1
0.5
3.6
0.1
>5
3.2
2
3.69
3.73
3.73
4.84
3.66
12
1
2
3
4
5
O
H
N
Me
Me
Me
Me
H
N
O
O
O
O
354.16
354.16
408.13
330.19
15
10
33
17
O
O
H
N
MeO
MeO
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
Me
H
N
O
2.3
0.7
2.8
H
N
Me
H
N
O
2.7
3.4
O
O
H
N
F3CO
Me
Me
H
N
O
0.3
O
H
N
Me
Me
H
Me
N
O
O
352.18
384.17
3.6
0.2
0.3
3.1
1.8
4.68
3.75
4
9
6
7
O
H
N
Me
Me
Me
H
N
O
MeO
O
2
O
O
H
N
Me
OMe
Me
H
N
O
325.14
352.18
0.5
>5
7
1.9
3
2.86
4.73
28
7
8
9
N
O
H
N
Me
Me
Me
H
Me
Me
N
O
O
H
2.5
O
O
N
Me
H
N
O
O
H
352.18
325.14
325.14
343.13
343.13
392.13
5
3.7
1.7
>5
>5
>5
>5
3.2
1.8
1.9
1.3
2
4.73
2.86
2.86
2.65
3.1
5
10
11
12
13
14
15
N
Me
Me
Me
Me
H
N
O
O
O
N
O
0.5
0.8
0.6
0.9
5
31
28
31
21
7
O
H
N
Me
H
N
N
O
O
O
H
N
Me
H
N
O
N
H
N
F
Me
Me
H
F
N
O
O
O
H
N
N
Me
Me
H
N
O
O
2.2
4.94
O
O
H
N
F3
C
Me
Me
N
Me
Me
O
337.42
>5
>5
NA
NA
NA
16
O
H
N
Me
a- Measured values, see details in SI; b-calculated value using StarDrop software
ACS Paragon Plus Environment

Page 11 of 34
ACS Chemical Biology
1
2
3
4
5
6
A summary of the dual nSMase2/AChE inhibition SAR from our optimization
efforts is summarized in Figure 2. Key points are as follows: (1) replacement of nitrogen
in the furoindoline ring of the validated hit yields 1 (phensvenine), which is a dual
7
8
9
inhibitor showing weak nSMase2 inhibition (IC50 > 10 µM) but potent AChE (IC50=0.5
42
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
µM) inhibition;
(2) phenserine, a commercially available potent AChE inhibitor,
displays loss of nSMase2 inhibitory activity (IC50 (not depicted) >50 μM); (3) substitution
in the 3-position of carbamate ring favors enhanced AChE inhibition as seen in
compound 3 compared to compound 2; (4) the 4-pyridyl ring in the carbamate group
leads to 8, a dual inhibitor with ~14-fold selectivity for nSMase2 inhibition (IC50 = 0.5
μM) over AChE inhibition (IC50 = 7 µM); and (5) the 3-pyridyl carbamate compound 11
was a dual inhibitor with ~3-fold selectivity for nSMase2 (IC50=0.5 µM) and AChE
(IC50=1.7 µM) inhibition. The mode of inhibition by the dual inhibitors (shown below) of
both enzymes allow for comparison dual activity using IC50 values.39,41 Based on the
SAR, the two dual inhibitors, 8 and 11, with 10- and 3-fold selectivity for nSMase2
inhibition over AChE, respectively, were further evaluated in in vitro and in vivo assays
for exosomal tau release.
Me
Me
H
H
N
N
O
O
O
NMe
O
O
H
N
H
N
Me
Me
Phenserine
IC50 (nSMase2) > 50 µM
Phensvenine (1)
IC50 (nSMase2) > 10 µM
IC50 (AChE) = 0.3µM
IC50 (AChE) = 0.5 µM
Me
Me
H
H
N
O
N
O
MeO
O
O
O
O
H
H
N
N
MeO
Me
Me
2 (Dual nSMAse 2/AChEl)
3 (Dual nSMAse 2/AChEl)
IC50 (nSMase2) = 0.7 µM
IC50 (nSMase2) = 1.0 µM
IC50 (AChE) = 3.6 µM
IC50 (AChE) = 0.1 µM
Me
Me
H
H
N
O
N
O
N
O
O
N
O
O
H
H
N
N
Me
Me
8 (Dual nSMAse 2/AChEl)
11 (Dual nSMAse 2/AChEl)
IC50 (nSMase2) = 0.5 µM
IC50 (nSMase2) = 0.5 µM
IC50 (AChE) = 7 µM
IC50 (AChE) = 1.7 µM
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 34
1
2
3
4
5
Figure 2. Key structure-activity relationship (SAR) control elements for inhibition of
nSMase2 (blue highlight) and AChE (yellow or green highlight) activity are indicated.
6
7
8
9
Mechanism of AChE and nSMase2 inhibition by the novel furoindoline
compounds. To determine the type of nSMase2 inhibition by compounds 8 and 11,
kinetics assays were performed. As shown in Figures 3a and 3b, increasing
concentrations of compounds 8 and 11 resulted in decreasing Km (the Michaelis
constant) values as well as concomitant decreases in Vmax (the maximum rate)
indicative of a uncompetitive mechanism of inhibition of nSMase2.43 Thus, it can be
concluded that both compounds bind to the enzyme distal from the active site and can
inhibit enzyme-substrate cleavage.
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
Molecular docking analysis (see the SI for details) was performed using a
recently published crystal structure of the nSMase2 catalytic domain (pdb: 5UVG).44 We
found that both 8 and 11 could bind to nSMase2 at the distal DK-switch (Asp-Lys) site
away from the substrate sphingomyelin site, and thus in concordance with the kinetic
analysis, could be an uncompetitive inhibitor of the enzyme through modulation of the
DK-switch. This is similar to what we have previously published with the known
nSMase2 inhibitor cambinol, which was also shown by molecular docking and
simulation to bind the nSMase2 catalytic domain in the DK-switch region and prevent
enzyme activation by likely keeping the switch in the ‘off’ position.14 Molecular
Dynamics (MD) simulation was performed to determine the binding free energy of
compound 8 binding to nSMase2. Compound 8 stays at the DK-switch site of nSMase2
through the 50 ns simulation with an estimated binding energy of -14.3 kcal/mol. An
AMBER16 package was used to perform the MD simulation.
Kinetic studies of AChE inhibition by Phenserine was shown to be in a mixed
type manner (competitive/uncompetitive), we assumed that likewise the dual inhibitor
compounds 8 and 11 would bind to active site of AChE and show mixed inhibition. 41, 42
ACS Paragon Plus Environment

Page 13 of 34
ACS Chemical Biology
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. Mechanism of nSMase2 inhibition by compounds 8 and 11. Kinetics of
enzyme inhibition by compounds 8 (a) and 11 (b) are shown. The rate of the reaction is
plotted against substrate concentration at four different concentrations of the inhibitors;
corresponded values for Vmax and Km are presented in the tables below the graphs. c)
Modeling of 8 and 11 (green) binding to catalytic domain of nSMase2 predicts
compound binding preferably near DK-switch site than substrate binding site. d)
Molecular surface representation of the nSMase2 catalytic domain with 8 and 11
(green) bound near the DK-switch; the color representations are blue for positive
charge, red for negative charge and white for neutral charge. e) The nSMase2 residues
(yellow) within 5 Å radius surrounding inhibitors 8 and 11 (green); H-bonding between
the inhibitor and nSMase2 are shown in dashed lines (black).
In vitro inhibition of tau seed propagation by dual nSMase2/AChE inhibitors. We
previously developed a cell culture system based on a well-known tau RD biosensor cell
line (tau biosensors) for testing inhibitors of tau propagation.14 Using known nSMase2
inhibitors cambinol and GW4869, we had demonstrated the role of the nSMase2-
dependent pathway of EV biogenesis in tau transmission from donor to recipient cells in
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 34
1
2
3
4
5
6
this non-neuronal cell model using two different in vitro assays – the Donor plus
Recipient (D+R) assay and the EV-mediated transfer (EMT) assay.14
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. Dual nSMase2/AChE inhibitors 8 and 11 suppress tau propagation from
donor to recipient cells in vitro. The assay scheme for each assay is presented at the
top of the figure. a) Donor plus recipient (D+R) assay results are shown. Compounds 8
and 11 at a concentration of 20 µM or a corresponding volume of DMSO were added to
the D+R cultures for 48 hrs. Levels of FRET signal were analyzed in recipient cells
using flow cytometry. Combined data from three independent experiments are
presented. b) EV-mediated tau seed transfer (EMT) assay results are shown.
Compounds 8 and 11 at 20 µM concentration or DMSO were added to donor cell culture
medium and then donor cell-derived EVs were purified and transfected to recipient cells.
Levels of FRET signal were analyzed in recipient cells using flow cytometry. Four
technical replicates were used for each experimental condition. Combined data from
three independent experiments is presented. The histograms represent integrated
FRET density per each treatment group (mean ± SEM). c) Size distribution and
concentrations of the donor-derived EV samples were analyzed by Tunable Resistive
ACS Paragon Plus Environment

Page 15 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Pulse Sensing (TRPS). d) Donor-derived EVs were imaged using transmission electron
microscopy (TEM). e) Western blot (WB) representative images for exosomal markers
are shown. The same volume of EV fractions derived from a similar number of donor
cells or control tau biosensor cells treated with Lipofectamine 2000 (Control) were
loaded per well and probed against exosomal markers CD63, CD81, and Syntenin-1.
Densitometry analysis is shown below the WB image. Statistics were performed using
One-way ANOVA with post hoc Bonferroni and Holm multiple comparison test was used
for statistical analysis: * p<0.05, **<0.01.
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
The principles of the D+R and EMT assays are presented in schematic form in
Figures 4a and 4b, respectively. Our data demonstrates that treatment with 8 or 11 at
a concentration of 20 µM significantly suppresses tau seed transfer from donor to
recipient cells in the D+R and EMT assays (Figures 4a, 4b and Supplementary
Figure 10). Shuttling by tau-bearing EVs is not the only pathway of tau seed transfer
between cells in vivo or when donor and recipient cells are growing together in vitro, as
in the D+R assay. In contrast, the EMT assay lets us isolate the effect of the inhibitors
on EV-mediated tau seed transmission, which can explain the profound difference in the
magnitude of FRET fluorescence density by dual nSMase2/AChE inhibitor 11 between
the assays - 19.5% decrease from dimethyl sulfoxide (DMSO) treated cells in D+R
assay and 41.3% decrease in EMT assay.
We characterized EVs purified from the seeded donor cells growing in the
presence of our dual inhibitor compounds 8 and 11 or DMSO control. Successful
purification of EVs was confirmed by tunable resistive pulse sensing (TRPS) (Figures
4c), transmission electron microscopy (TEM) (Figures 4d), and western blotting
analysis with known exosomal markers (Figures 4e). Treatment with dual
nSMase2/AChE inhibitors 8 or 11 did not affect EV size distribution, but decreased the
concentrations of exosomal-type small EVs (Figures 4c). Levels of exosomal markers
CD63, CD81, and syntenin-1 were reduced in EVs purified from 8 and 11 treated cells
in comparison with the DMSO treated donors (Figures 4e). Relatively high suppression
of tau transfer by 11 compared to 8 in the EMT assay may be related to the greater
AChE inhibitory activity of 11 in conjunction with its nSMase2 inhibition and the role of
38
dual inhibitory activity in exosome-mediated transfer of tau seeds.
Interestingly, in
pilot testing using the (+) enantiomer of 8 and 11 in the EMT assay, we did not see any
suppression of EV-mediated tau seed transfer when compared to the (-) enantiomer.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 34
1
2
3
4
5
6
7
8
9
Cell viability and/or rate of proliferation may have an effect on tau seed transfer
from donor to recipient cells through different mechanisms. Thus, we evaluated effects
of tau seeding and treatment with nSMase2/AChE inhibitors on donor cell number and
viability. Twenty-four hour exposure to AD human brain synaptosomal extracts
decreased the rate of the donor cell survival in the next passage compared to cells
treated with lipofectamine 2000 (Supplementary Figure 11). We have not determined
the specific mechanisms of cell death in tau-seeded donor cultures. It is possible that a
subset of tau-bearing EVs affected by nSMase2 inhibitors are apoptotic exosome-like
vesicles (AEVs) that - in contrast to apoptotic bodies - represent a subtype of exosomes
originating from multivesicular endosomes (MVE) at the early apoptotic phase. AEV
biogenesis is controlled by the ESCRT-independent sphingosine1-phosphate
(S1P)/S1PRs signaling pathway, and can be partially inhibited by nSMase2 inhibitor
GW4869.45 Interestingly, AChE inhibitors are known to protect different cell types,
including HEK293T, from apoptosis,46 and thus dual inhibitor 11 with greater AChE
inhibition could potentially indirectly suppress AEV production. However, treatment of
donor cells with 11 for 48 hours didn’t affect donor cell numbers or survival compared to
DMSO or to compound 8 treated donor cells (Supplementary Figures 11a and 11b).
We also hypothesize that other factors may contribute to the greater effect of 11 on tau
seed transfer in the EMT assay. A recent report suggests that intracellular uptake of tau
can be mediated by the muscarinic acetylcholine receptors (mAChR) M1 and M3.38
Thus, accumulation of tau oligomers in the synapse may exacerbate the cholinergic
deficit in AD through suppression of ACh uptake via mAChR M1/M3 receptors on
postsynaptic terminals. Based on similar reasoning, inhibition of AChE could have a
direct effect on tau seed uptake through the increased levels of ACh in the synapse and
postsynaptic M1/M3 receptor occupancy.
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
Our preliminary experiments at 20 μM using rivastigmine, a potent AChE (but
with no nSMase2 activity) inhibitor, reveals that inhibition of AChE may partially
suppress EV-mediated transfer and uptake of tau seeds by recipient cells
(Supplementary Figures 12). We have previously published on cambinol, an nSMase2
inhibitor (IC50 = 7.7 μM but no AChE activity), that shows inhibition of tau seed transfer
in the EMT assay.14 The novel dual inhibitors 8 and 11 have both nSMase2/AChE
ACS Paragon Plus Environment

Page 17 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
inhibitory activities in the same molecule and show suppression of tau transfer (Figures
4b). Thus, these dual nSMase2/AChE inhibitors 8 and 11 are agents that could be
developed to efficiently target both enzymes in the brain and suppress tau propagation.
Brain pharmacokinetics for lead compounds. Our goal was to identify brain
permeable dual nSMase2/AChE inhibitor analogs for further testing. We therefore
performed pharmacokinetic (PK) analysis on the leads 8 and 11 to determine brain
permeability using wild type mice. The compounds, whose brain permeability have not
been previously described, were subcutaneously (SQ) injected at a relatively high dose
of 20 mg per kg of body weight (mpk) to determine brain penetrance. Brain and plasma
samples were collected 1, 2, and 4 hours after dosing. Our PK analysis revealed that 8
and 11 reached peak brain levels (Cmax) around one hour after SQ dosing and brain
levels were detectable for both compounds 4 hours after injection (Figure 5a).
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. Pharmacokinetic analysis for lead compounds 8 and 11. a) Mice were
subcutaneously (SQ) injected with 20 mg/kg of compound 8 or 11; animals were
sacrificed 1, 2, and 4 hours after dosing (n=1 animal per time point). b) Mice were dosed
as in (a), but n = 6 per compound and sacrificed 1 hour after dosing. Compound levels
in brain tissue were analyzed using a LC-MS/MS method.
To carefully evaluate brain compound levels at the Cmax (1 hour) time point, 20
mpk SQ dosing of compounds 8 and 11 was performed again using 6 mice per group.
Average brain level of the compounds at the peak was equal to 61 ng/g (0.2 μM) and
262 ng/g (0.8 μM) for compounds 8 and 11, respectively (Figure 5b). This data
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 34
1
2
3
4
5
6
7
8
9
confirmed good brain permeability of the lead compounds as was predicted by in silico
and PAMPA analysis described earlier (Table 1). Compound 11 showed higher average
brain levels compared to 8, and the brain levels corresponded to ~ 2-fold IC50 for
nSMase2 and ~0.5-fold for AChE.
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
Inhibition of brain EV release by the dual nSMase2/AChE inhibitors in a rapid in
vivo assay. The chronic inflammation that is reported in AD and tauopathy models is
characterized by elevated levels of pro-inflammatory cytokines in brain parenchyma,
including interleukin 1β (IL1β), known to induce nSMase2 activity through the IL1-
Receptor 1 (IL1-R1).47 Neuroinflammation and upregulation of IL1 signaling is linked
with an early stage of tauopathy development; blocking of IL1 signaling in the 3xTg
mouse AD model attenuates tau pathology and rescues cognition.48-50 It was
demonstrated that striatal injection of IL1 to wildtype mice induced release of
astrocyte-derived EVs into the blood, resulting in peripheral acute cytokine responses 51
which can be suppressed by pre-treatment with nSMase2 inhibitors.34-35
In order to rapidly test our dual nSMase2 inhibitors in vivo, we used the Tau
P301S (PS19 line) tauopathy mouse model.48 For our in vivo assay there were 4
groups: group I (control) received SQ injection of vehicle (DMSO) and
intracerebroventricular (ICV) injection of another vehicle (0.0006% BSA in PBS, pH 7.4)
an hour after SQ treatment; group II (IL1) received SQ injection of vehicle and
unilateral ICV injection of 0.2 ng of IL1 an hour later; group III (8/IL1) – SQ treatment
with 20 mg/kg of 8 and ICV injection of 0.2 ng of IL1; group IV - SQ treatment with 20
mg/kg of 11 and ICV injection of 0.2 ng of IL1. The one-hour interval between
treatment with the inhibitors and IL1 ICV injection was chosen based on the brain PK
analysis presented above. All animals were sacrificed at 3 hours after compound or
vehicle treatment and 2 hours after ICV injection of IL1. Brain EVs were purified as
previously described.52
Size distribution and concentration of brain EVs were analyzed using the TRPS
method. There were no significant differences in EV size distribution between
experimental groups (Figure 6a). As previously reported,52 the collected fraction (F2)
consists mostly of small exosome-size EVs with a mode equal to 80±5 nm based on
TRPS analysis. A high abundance of exosome-sized EVs was confirmed by TEM
ACS Paragon Plus Environment

Page 19 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
analysis (Figure 6c). As expected we found that ICV injection of IL1 significantly
increased the concentration of small EVs (size 50-150 nm) purified from the brain, more
than 2 times that of the control (Figure 6b). Dual nSMase2/AChE inhibitors 11
suppressed IL1-induced exosomal release to the control level (Figure 6b), while the
less brain-permeable dual inhibitor 8 did not induce the same level of suppression.
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 6. Dual nSMase2/AChE inhibitor 11 diminished IL1β-induced brain EV release in
the rapid in vivo assay. Tau P301S (line PS19) mice were treated with compound 8 or
11 subcutaneously (SQ) at 20 mg/kg one hour before IL1β injection (unilateral ICV
injection of 0.2 ng). Two hours after IL1β injection, brain tissue was collected and used
for brain EV isolation. a) Size distribution and concentrations of the brain EV samples
were analyzed by Tunable Resistive Pulse Sensing (TRPS). b) Average concentrations
of 50-150 nm size EVs from each treatment condition were compared. c) A
representative transmission electron microscopy (TEM) image of the brain EV fraction is
shown. d) Representative images of western blot (WB) analysis of EV fractions from
individual animals is shown; membranes were probed against exosomal markers (CD63
and syntenin-1), tau protein, and cell-type specific markers (astrocytic glutamate-
aspartate transporter GLAST1, microglia marker CD11b, and neuronal isoform of
Bridging Integrator 1, BIN1). e) Densitometry analysis of the WB images is shown.
Histograms represent average relative signal intensity per each treatment group (mean
± SEM). Statistical analysis was performed using one-way ANOVA with post hoc
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 34
1
2
3
4
5
6
Bonferroni and Holm multiple comparison tests: #-P < 0.05 and ## - P < 0.01 compared
to control group, treated with vehicles for SQ and ICV injections, * - P < 0.05 and ** - P
< 0.01 compared to IL1β group.
7
8
9
Biochemical analysis of brain-derived EVs (Figures 6d & 6e) confirmed that
pretreatment with lead compound 11 led to a significant reduction of exosomal marker
CD63 in exosome-enriched F2 fractions compared to the group treated only with IL1.
In contrast to significant changes in common exosomal marker CD63, levels of
syntenin-1, a marker of a specific exosomal subpopulation generated through the
Syndecan-Syntenin-ALIX pathway 53 were not different between the groups (Figures 6d
& 6e). These results confirm that IL1β stimulation and nSMase2 inhibition have effects
on specific populations of exosomes.
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
Our data suggest that the nSMase2-dependent pathway of exosome biogenesis
is involved in tau-bearing exosome production in PS19 mice. Tau levels in the F2
fraction showed a strong trend of being elevated in animals treated with IL1β, with the
average tau level being around 6 times higher in the IL1β-treated group compared to
the control group (Figures 6d & 6e). Pretreatment with 11 significantly reduced IL1β-
induced tau release by exosomes. The lead compound 8 was less effective in this
study. The known variability of tau load between PS19 mice likely accounts for the lack
of statistical significance despite the high magnitude of tau changes.
Multiple brain cell types express IL1-R1, including subpopulations of neurons,
astrocytes, choroid plexus cells and ependymal cells;54 thus, the nSMase2-mediated
exosomal release by different types of brain cells can be affected differently in response
to acute increases in intracerebral IL1 concentration. We used a couple of cell-type
specific markers to assess the origin of the IL1β/nSMase2 sensitive exosomal
population. We found that levels of astrocytic glutamate-aspartate transporter (GLAST)
and microglial marker CD11b were significantly elevated in F2 fractions isolated from
IL1β-treated animals. GLAST is known to be sensitive to papain, the enzyme we used
for gentle brain tissue dissociation. Therefore, we used a 30 kDa fragment of GLAST
instead of full-length protein for the analysis.55 Pretreatment with the dual
nSMase2/AChE inhibitor 11 significantly reduced the level of astrocyte-derived
exosomes and showed the same trend for microglia-derived exosomes, but the
ACS Paragon Plus Environment

Page 21 of 34
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
difference in CD11b levels between IL1β and 11/IL1β groups was not significant
(Figures 6d & 6e). This finding correlates with previously demonstrated IL1β-induced
nSMase2-mediated production of astrocyte-derived exosomes in wild type mice.34-35
Microglia play an important role in tau spread,15, 55 and inhibition of microglial nSMase2-
dependent exosome release suppresses tau propagation in mouse models.15 The low
levels of microglia response in our rapid in vivo assay may be attributed to saturation of
microglia responses in 5-6 month old PS19 mice. Microglia activation is already
detectable in 3 month old PS19 mice and precedes astrogliosis.48
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
Recently, Bridging Integrator 1 (BIN1), a known genetic risk factor for AD, was
connected to tau seed release through exosomes in human AD and male PS19 mice.56
We analyzed levels of BIN1 in our F2 samples. Neuronal BIN1 isoform 1, but not
microglia specific isoform 2, were highly enriched in the F2 fractions (Figure 6d). As in
the case of exosomal tau, we found a high magnitude increase in exosome-associated
BIN1 upon IL1β stimulation that was lower in the compound 11 treated group, but no
statistically significant changes were found due to the high variability of individual levels
of BIN1 within each group (Figure 6d). This data suggests that nSMase2 and BIN1
could be a part of the same exosomal pathway responsible for tau release and spread
in AD.
Overall, our rapid in vivo assay results demonstrate the effectiveness of novel
dual AChE/nSMase2 inhibitor 11 in suppression of IL1β-induced release of tau-bearing
exosomes in a tauopathy model.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 34
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 7. A putative mechanism for dual nSMase2/AChE inhibition and suppression of
EV/exosome-mediated propagation of tau pathology wherein nSMase2 inhibition
suppresses exosome biogenesis while AChE inhibition reduces exosome uptake and
cholinergic support.
Our discovery of a novel class of potent nSMase2/AChE dual inhibitors presents
an opportunity for further evaluation and development of these agents as a new
therapeutic approach for the treatment of Alzheimer’s disease. Our data supports the
ability of the dual inhibitors to suppress tau propagation in vitro and release of tau
carrying exosomes in vivo in an AD model. These dual nSMase2/AChE inhibitors would
enhance cholinergic synaptic plasticity, reduce neuroinflammation,37 and most
importantly suppress exosome-mediated tau propagation and tau uptake mediated
through the M1/M3 muscarinic ACh receptors.38, 57-59 This combination effect is unique,
has not been evaluated previously in the disease and clearly differentiate these agents
from currently available AChE inhibitors for the treatment of AD. In concert, these
mechanisms of action have the potential to not only address symptoms of AD by
enhancing cholinergic activity but also to suppress cell-to-cell tau propagation, (Figure
7), significantly altering an underlying cause of AD and thus be truly disease-modifying.
ACS Paragon Plus Environment

Page 23 of 34
ACS Chemical Biology
1
2
3
4
METHODS
5
6
7
8
Compound Synthesis. Unless stated otherwise, reactions were conducted in flame-
dried glassware under an atmosphere of N2 and commercially obtained reagents were
used as received. Sodium hydride, boron tribromide, boron trichloride, phenyl
isocyanate, N,N’-disuccinimidyl carbonate (29), cyclohexyl isocyanate, 3,5-
dimethylphenyl isocyanate, 3,5-dimethoxyphenyl isocyanate, 3-aminopyridine (33), 4-
aminopyridine (32), and 4-(trifluoromethyl)aniline (37) were obtained from Sigma-
Aldrich. Hydrazine (17), 4-methoxyphenyl isocyanate, 3- methoxyphenyl isocyanate,
and N-ethylmethylamine (37) were obtained from Oakwood Products, Inc. 4
(trifluoromethyl)phenyl isocyanate and 3-amino-5-fluoropyridine (35) were obtained from
Combi-Blocks. Methyl iodide was obtained from Alfa Aesar. Reaction temperatures
were controlled using an IKAmag temperature modulator, and unless stated otherwise,
reactions were performed at room temperature (approximately 23 °C). Thin-layer
chromatography (TLC) was conducted with EMD gel 60 F254 pre-coated plates (0.25
mm for analytical chromatography and 0.50 mm for preparative chromatography) and
visualized using a combination of UV, anisaldehyde, iodine, and potassium
permanganate staining techniques. Silicycle Siliaflash P60 (particle size 0.040–0.063
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
1
mm) was used for flash column chromatography. H NMR spectra were recorded on
Bruker spectrometers (500 MHz) and are reported relative to residual solvent signals.
1
Data for H NMR spectra are reported as follows: chemical shift (δ ppm), multiplicity,
coupling constant (Hz), integration. Data for 13C NMR are reported in terms of chemical
shift (125 MHz). 19F NMR spectra were recorded on Bruker spectrometers (at 376 MHz)
and reported in terms of chemical shifts (δ ppm). Data for IR spectra were recorded on
a Perkin-Elmer UATR Two FT-IR spectrometer and are reported in terms of
frequency absorption (cm-1). DART-MS spectra were collected on a Thermo Exactive
Plus MSD (Thermo Scientific) equipped with an ID-CUBE ion source and a Vapur
Interface (IonSense Inc.). Both the source and MSD were controlled by Excalibur
software v. 3.0. The analyte was spotted onto OpenSpot sampling cards (IonSense Inc.)
using volatile solvents (e.g. chloroform, dichloromethane). Ionization was accomplished
using UHP He (Airgas Inc.) plasma with no additional ionization agents. The mass
calibration was carried out using Pierce LTQ Velos ESI (+) and (–) Ion calibration
ACS Paragon Plus Environment

ACS Chemical Biology
Page 24 of 34
1
2
3
4
5
6
7
8
9
solutions (Thermo Fisher Scientific). Optical rotations were measured with a Rudolph
Autopol III Automatic Polarimeter. Any modification of the conditions shown in the
representative procedures are specified in the corresponding schemes. The detailed
methods for synthesis of the compounds are included in the Supporting Information.
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
Enzyme activity and kinetic analysis assay. To evaluate nSMase2 inhibitory activity
of tested compounds, cell lysates from HEK293T cells over-expressing human
nSMase2 were used as the source of the nSMase2 enzyme.34 The enzyme activity was
measured with or without inhibitor treatment using the Amplex Red Sphingomyelinase
activity assay.42 For acetylcholinesterase (AChE) assay we used human AChE
(Sigma). The enzyme activity was measured with or without inhibitor treatment using the
Amplex Red assay kit (Thermo Fisher A12217). The reaction was monitored for 60 min
and read at 530/590 nm. The detail methods for enzyme activity analysis are included
in the Supporting Information.
Modeling of compounds binding to SMase2. Molecular docking analysis of
compounds was performed using the Swiss Dock server on our Area51 Work Area 51
R4 linux workstation. Molecular Dynamics (MD) simulation was performed to determine
the binding free energy of compound 8 binding to nSMase2. An AMBER16 package
was used to perform the MD simulation. The Antechamber module in AMBER was used
to generate the parameters for compounds. The SwissDock web server was used to
predict compound binding to nSMase2. The detail methods for modeling analysis
included in the Supporting Information.
Parallel Artificial Membrane Permeability Assay (PAMPA). To predict the potential
for brain permeability, a Regis Technologies analytical column connected to an Agilent
HPLC system was used. We used the IAM.PC.DD column for determination of the
retention time of a compound to calculate the predicted CNS permeability. The detail
methods for modeling analysis included in the Supporting Information.
ACS Paragon Plus Environment

Page 25 of 34
ACS Chemical Biology
1
2
3
4
5
6
Human Serum Albumin (HSA) Binding Assay. For HSA compound binding we use a
CHIRAL column immobilized with HSA on a Agilent HPLC system. The detail methods
for modeling analysis included in the 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
Preparation of brain-derived synaptosomal extracts. Cryopreserved brain tissue
was obtained from University of California Irvine was used for preparation of the
synaptosomal fractions (P2 fractions, ‘P2’). The detail methods for modeling analysis
included in the Supporting Information.
Tau propagation assays. We use the HEK293T Tau RD P301S FRET biosensor (tau
biosensor) for evaluation of the inhibitors in functional tau propagation “D+R” (Donors
plus Recipients) and EMT (EV-mediated transfer) assays. We have previously
published the detailed protocols and validation of the assays using known nSMase2
14
inhibitors GW4869 and cambinol. The detail methods for modeling analysis included
in the Supporting Information.
Rapid in vivo assay and brain EV purification. Using 5-6 month old male PS19 mice
expressing human tau with the P301S mutation under control of the murine prion
promoter intracerebroventricular (ICV) injections of IL-1β with or without pre-treatment
with dual nSMase2/AChE inhibitor followed isolation of EVs. The in vivo experiments
were performed under an approved IACUC protocol. The detail methods for modeling
analysis included in the Supporting Information.
Transmission electron microscopy (TEM). For quality control purposes, small
amounts of purified brain- or tissue culture- derived EVs were fixed on a copper mesh in
glutaraldehyde/paraformaldehyde solution, stained with 2% uranyl acetate solution and
imaged on a JEOL 100CX electron microscope at 29,000X magnification.
Tunable Resistive Pulse Sensing analysis. Size distribution and concentrations of EV
samples were analyzed by Tunable Resistive Pulse Sensing (TRPS) method using the
qNano Gold instrument (Izon Science). NP100 nanopore (particle size range: 50-330
ACS Paragon Plus Environment

ACS Chemical Biology
Page 26 of 34
1
2
3
4
5
6
nm) and CPC100 calibration particles were used for the analysis. Data analysis was
performed using qNano instrument software.
7
8
9
Immunoblot analysis of EV samples. Electrophoresis of proteins was performed
using 10–20% Tris-Glycine gels in non-reducing (only for tetraspanins) or reducing (with
addition of DTT) conditions; proteins were then transferred to PVDF membrane and
probed with primary antibodies (Supporting Table 2) followed by HRP conjugated
secondary antibodies. Chemiluminescent signals were generated with Super Signal
West Femto substrate (Thermo Scientific Pierce 34095) and detected using a
BioSpectrum 600 imaging system and quantified using VisionWorks Version 6.6A
software (UVP; Upland, CA). The detail methods for modeling analysis included in the
Supporting Information.
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. All data was expressed as the mean ± SEM. Significant
differences were determined by one-way ANOVA followed by Bonferroni and Holm
multiple
comparison
method
using
online
web
statistical
calculator
(http://astatsa.com/OneWay_Anova_with_TukeyHSD). Only a subset of pairs relative to
#
the DMSO group were simultaneously compared. Values of * or <0.05 and ** or
##<0.01 were considered statistically significant.
ASSOCIATED CONTENT
The Supporting Information is available free of charge on the ACS Publications website.
Detailed experimental procedures, compound characterization data submitted as a
(PDF).
ACS Paragon Plus Environment

Page 27 of 34
ACS Chemical Biology
1
2
3
4
AUTHOR INFORMATION
5
6
7
Corresponding Authors*
Neil Garg, Ph.D.
8
Professor & Chair
9
UCLA Chemistry & Biochemistry
Email: neilgarg@chem.ucla.edu
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
0000-0002-7793-2629
Varghese John, Ph.D.
Professor of Neurology
PI, Drug Discovery Lab
UCLA Neurology
Email: vjohn@mednet.ucla.edu
0000-0001-7675-2426
Authors
Tina Bilousova, Ph.D. - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Bryan J. Simmons Ph.D. - University of California Los Angeles,
Dept. of Chemistry and Biochemistry
607 Charles Young Drive East
Los Angeles, CA, USA 90095
Rachel R. Knapp - University of California Los Angeles,
Dept. of Chemistry and Biochemistry
607 Charles Young Drive East
Los Angeles, CA, USA 90095
Chris J. Elias - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Jesus Campagna - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Sujyoti Chandra, Ph.D. - University of California Los Angeles,
ACS Paragon Plus Environment

ACS Chemical Biology
Page 28 of 34
1
2
3
4
5
6
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
7
8
9
Samantha Focht - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
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
3138 Reed Building
Los Angeles, CA, USA 90095
Chunni Zhu, Ph.D. - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Kanagasabai Vadivel, Ph.D. - University of California Los Angeles,
Orthopaedic Surgery
615 Charles Young Drive East
Los Angeles, CA, USA 90095
Barbara Jagodzinska, Ph.D. - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Whitaker Cohn - University of California Los Angeles,
Department of Neurology
Drug Discovery Lab
3138 Reed Building
Los Angeles, CA, USA 90095
Karen H. Gylys, Ph.D.- University of California Los Angeles,
School of Nursing
Gylys Lab
Los Angeles, CA, USA 90095
Funding
The work was supported by Mary S. Easton Center for Alzheimer’s Disease Research
at UCLA to V.J.; National Institutes of Health (AG051386 to V.J.), and the National
Science Foundation (DGE-1144087 for B.J.S. and CHE-1464898 for N.K.G.) These
studies were supported by shared instrumentation grants from the NSF (CHE-1048804)
and the National Center for Research Resources (S10RR025631)
ACS Paragon Plus Environment

Page 29 of 34
ACS Chemical Biology
1
2
3
4
Notes
The authors declare no competing financial interests.
5
6
7
8
9
ACKNOWLEDGEMENTS
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
. The authors thank the funding sources that supported this work. We would like
to thank W. W. Poon from UC Irvine for human autopsy brain samples, C. Rojas from
John Hopkins School of Medicine for nSMase2-overexpressing HEK293 cells, and the
UCLA Janis V. Giorgi flow cytometry and Brain Research Institute Electron microscopy
core facilities.
ACS Paragon Plus Environment