A CNS-Targeting Prodrug Strategy for Nuclear Receptor Modulators

Article abstract of DOI:10.1021/acs.jmedchem.0c00868

The blood-brain barrier is a major impediment for targeted central nervous system (CNS) therapeutics, especially with carboxylic acid-containing drugs. Nuclear receptor modulators, which often feature carboxylic acid motifs for target engagement, have emerged as a class of potentially powerful therapeutics for neurodegenerative CNS diseases. Herein is described a prodrug strategy that directs the biodistribution of parent drug nuclear receptor modulators into the CNS while masking them as functional receptor ligands in the periphery. This prodrug strategy targets a specific amidase, fatty acid amide hydrolase (FAAH), an enzyme with enriched expression in the CNS. Our results demonstrate that this prodrug strategy can be generalized to a variety of carboxylic acid-containing drug structures that satisfy the structural requirements of blood-brain barrier diffusion and FAAH substrate recognition.

Full text of DOI:10.1021/acs.jmedchem.0c00868

Subscriber access provided by CARLETON UNIVERSITY
Article
A CNS-targeting prodrug strategy for nuclear receptor modulators
Skylar Ferrara, and Thomas S Scanlan
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00868 • Publication Date (Web): 05 Aug 2020
Downloaded from pubs.acs.org on August 5, 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 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A CNS-targeting prodrug strategy for nuclear receptor
modulators
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
Skylar J. Ferrara and Thomas S. Scanlan*
Department of Chemical Physiology and Biochemistry, Oregon Health & Science University, 3181 SW Sam
Jackson Park Road, Portland, Oregon 97239, United States
KEYWORDS. prodrug, CNS, FAAH, nuclear receptors, blood-brain barrier
Abstract
The blood-brain barrier is a major impediment for targeted central nervous system (CNS) therapeutics,
especially with carboxylic acid-containing drugs. Nuclear receptors modulators, which often feature
carboxylic acid motifs for target engagement, have emerged as a class of potentially powerful therapeutics for
neurodegenerative CNS diseases. Herein is described a prodrug strategy which directs the biodistribution of
parent drug nuclear receptor modulators into the CNS while masking them as functional receptor ligands in
the periphery. This prodrug strategy targets a specific amidase, fatty-acid amide hydrolase (FAAH), an
enzyme with enriched expression in the CNS. Our results demonstrate that this prodrug strategy can be
generalized to a variety of carboxylic acid-containing drug structures that satisfy the structural requirements
of blood-brain barrier diffusion and FAAH substrate recognition.
Introduction
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 38
1
2
3
4
5
6
7
8
9
Nuclear receptors (NRs) are a validated class of drug targets and have recently emerged as
potentially powerful targets for diseases of the central nervous system (CNS).1-5 These ligand-activated
transcription factors are master regulators of vertebrate development as well as metabolism and homeostasis
in adults, and their synthetic modulators encompass approximately 13% of FDA approved drugs.6-8 NR
signaling has been shown to control lipid turnover and homeostasis, induce differentiation or phenotypic
changes of neurons and neuroglia, reduce amyloid burden, and suppress pro-inflammatory gene expression
among other beneficial effects depending upon the particular receptor being activated.1-2, 9 Many NRs are
expressed in CNS cell types linked to disease pathology enabling therapeutic intervention via modulation of
key signaling pathways a promising strategy with these highly druggable target receptors.
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
Neurodegenerative CNS diseases remain challenging to target therapeutically due in part to the
requirement that systemically dosed drugs pass through the blood-brain barrier (BBB), typically by passive
diffusion. Endogenous nuclear receptor ligands such as thyroid hormone access the CNS from systemic
circulation via active transport mechanisms. Many synthetic NR modulators, however, contain polar
functional groups like carboxylates required for high-affinity target engagement. Anionic functionality such as
carboxylates are a liability for passive diffusion across the BBB due to charge repulsion at the barrier surface
and the unfavorable electrostatic energetics of the cellular membrane potential. Isosteric replacement of
carboxylates can improve BBB penetration, but does not necessarily alter drug penetration into peripheral
tissues where undesired or adverse drug effects may occur. To circumvent excessive peripheral exposure, a
strategy favoring CNS distribution while minimizing peripheral drug exposure would be preferred.
A prodrug strategy that fulfills these criteria was recently developed in our lab for CNS-targeting of
the synthetic thyromimetic drug sobetirome.10-11 We have used this strategy to increase sobetirome
penetration into the CNS with mouse models of demyelination and lipid metabolism.12-13 This prodrug
strategy is underpinned by two pivotal processes and entails (1) converting the drug’s carboxylic acid
functional group into an N-methyl amide which imparts beneficial physicochemical properties for BBB
penetration via passive diffusion. As an amide, (2) the prodrug is a substrate for fatty acid amide hydrolase
(FAAH), an amidase with enriched expression in the CNS that cleaves the amide prodrug into the carboxylate-
containing parent drug. This results in enhanced CNS drug exposure and subsequent drug action (Figure 1A),
2
ACS Paragon Plus Environment

Page 3 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
while at the same time minimizing parent drug exposure and action in the periphery. In addition to our
development of a prodrug for sobetirome, other reports have successfully exploited FAAH for delivery of
bioactive substances and probes, however, none were deliberately used to increase the CNS penetration of a
drug.14-16
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
FAAH is a membrane bound serine hydrolase expressed in the CNS and select peripheral tissues and
functions to catabolize endogenous lipid endocannabinoids.17 Mice express only a single FAAH enzyme
whereas humans express two, FAAH1 and FAAH2, and there is considerable sequence overlap between
mouse FAAH and human FAAH1 with 84% shared amino acid identity.18 By targeting this specific enzyme, the
prodrug remains masked in peripheral circulation and devoid of receptor activity.11 The N-methyl amide of
sobetirome proved to be the optimal amide for delivering the most sobetirome to the CNS from a systemic
dose while at the same time minimizing the peripheral conversion of prodrug to parent drug. 11 However, an
open question remained as to whether this prodrug strategy could be successfully employed for other
carboxylate drugs that may have therapeutic applications in the CNS. Given the similarities between mouse
FAAH and human FAAH1, drugs adopting this prodrug strategy may prove useful as human therapeutics.
Herein we describe the results from testing the FAAH-targeted prodrug strategy on a series of different
known, and in some cases clinically relevant, carboxylic acid-containing nuclear receptor modulators. In
doing so, the substrate capacity of FAAH is examined with respect to potential small molecule N-methyl
amide substrates.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 38
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 1:FAAH-targeting CNS-selective prodrug strategy (A) and the series of nuclear receptor modulators
with nuclear receptor target in parentheses, and their corresponding prodrugs NRMA 1-10 (B).
4
ACS Paragon Plus Environment

Page 5 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Results and Discussion
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
Prodrug strategy and structures
Ten small molecule carboxylate-containing nuclear receptor modulators of therapeutic relevance to
CNS disorders were chosen to test the generalizability of the FAAH-targeting prodrug strategy (Figure 1B).
The panel includes ligands for thyroid hormone receptor (TR), retinoid X receptor (RXR), peroxisome
proliferator-activated receptor (PPAR), liver X receptor (LXR), estrogen receptor (ER), and retinoic acid
receptor (RAR). Nuclear receptor ligands constitute a rather diverse set of molecules as a consequence of the
differences in ligand binding pockets which correspond to the adapted biology of each receptor.19-20
Considerable structural variation is represented in this set of prodrug molecules to test the substrate capacity
of FAAH with respect to small molecule drug-like structures. Each drug tested has sufficient drug-like
properties and has undergone appropriate pharmacological evaluation to substantiate potential therapeutic
utility in the CNS.21
The N-methyl amide motif featured in each prodrug was deliberately chosen because it was shown to be
optimal for sobetirome, and represents the smallest (MW and size) practical amide moiety that increases
lipophilicity while shrinking the molecule’s topological polar surface area.11 These attributes promote the
passive diffusion of the charge-neutral amide prodrug across the BBB. Installation of the N-methyl amide can
be performed in a modular fashion via several routes. In our hands, CDI-mediated amide coupling proved the
most suitable synthetic method given the range of solubility for the series. Activation of the parent drug
carboxylic acid with carbonyldiimidazole (CDI) followed by the addition of excess methylamine in THF
provides the N-methyl amide prodrugs in moderate-to-excellent yields. The advantage of this method relates
to the mild conditions for amide coupling of these rather expensive parent drug starting materials, which
typically generates product mixtures that are only amide product and starting material after work-up.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 38
1
2
3
4
5
6
7
8
9
1) CDI
dry THF, 50 C
O
O
2 h
Y
Y
OH
2) H2NMe 2M in THF
r.t.  50 C
16 h
HN
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
Scheme 1: General synthetic scheme for modular synthesis of N-methyl amide prodrugs.
Prodrugs alter biodistribution and operate via FAAH in vivo
To determine if the amide prodrugs favorably alter the biodistribution of parent drug into the CNS,
parent drug concentrations were measured in brain and sera of male, wild-type C57BL/6 mice over 6 hours
from the time of systemic administration (i.p., 9.14 mol/kg) to generate an area under the curve (AUC)
(Figure 2, Table 1). These AUC values represent brain and serum exposure of the parent drug, either from
administration of the parent drug itself or that which was liberated via cleavage of the prodrug over the 6-
hour window, and are used to calculate a brain-to-serum ratio (Kp), a metric of CNS penetration and tissue
distribution selectivity. Of the ten prodrugs tested, 6 prodrugs favorably altered the biodistribution toward
the CNS and away from peripheral circulation. Compared to the corresponding parent drug, brain exposure
increased 2-to-6-fold and brain-to-serum ratios were enhanced 2-to-94-fold (see Table 1). In particular, the
prodrug of thyroid hormone receptor antagonist NH-3, NRMA-2, displayed excellent selectivity with a 6.2-fold
increase in brain exposure and a 94-fold increase in Kp. In contrast, NRMA-1, the N-methyl amide of the other
TR targeted drug Triac did not substantially increase CNS Triac exposure or Kp compared to unmodified Triac
demonstrating SAR limits to the prodrug strategy with respect to TR ligands. The prodrugs of tazarotenic
acid (NRMA-10) and tesaglitazar (NRMA-7), targeting RAR and PPAR, respectively, displayed Kp
improvements of 17-fold and 15-fold, respectively. Additionally, NRMA-2 and NRMA-10 feature Kp values of
1.5 and 2.03, respectively, indicating preferred distribution to the CNS compared to the periphery. In general,
Tmax values associated with prodrugs appear to be either very similar to the Tmax of the parent drug or greater,
likely suggestive of a processing time corresponding to the FAAH hydrolysis process in the CNS. Moreover,
the LC-MS/MS experiment to detect bezafibrate parent drug levels derived from NRMA-8 was set up to
6
ACS Paragon Plus Environment

Page 7 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
include detection of 4-chlorobenzoate in negative ion mode, the cleavage product resulting from hydrolysis of
the bezafibrate core structure’s internal amide. While NRMA-8 successfully altered parent drug partitioning
to the CNS, no (<LoQ) 4-chlorobenzoate was detected at any timepoint, suggesting that internal amides are
9
Brain AUC0.5 – 6 h (ng/g*h) Serum AUC0.5 – 6 h (ng/g*h)
Compound (9.14 mol/kg, i.p.)
Kp
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
Triac
NRMA-1
NH-3
88.9
155.8
11.9
5108
4020
756.9
49.5
0.017
0.039
0.016
1.5
NRMA-2
GW3965
NRMA-3
Bexarotene
NRMA-4
GW7604
NRMA-5
73.8
1867
< LoQ
1910
< LoQ
228.2
18.03
8853
155.4
1834
4.933
169.3
17.77
0.21
−
1.04
−
1.35
1.01
precluded from hydrolysis during FAAH-mediated processing. These results demonstrate that the FAAH-
targeted prodrug strategy can be extended to other carboxylate-containing drug scaffolds and the inclusion of
internal amide functional groups are tolerated.
In contrast, the prodrugs for GW3965 (NRMA-3) and bexarotene (NRMA-4), targeting LXR and RXR,
respectively, failed to yield quantifiable amounts of parent drug in brain and generated low levels in serum.
This suggests that both prodrugs are not hydrolyzed to parent drug in the CNS or periphery, and instead are
eliminated as the intact amide or an uncharacterized amide metabolite. Interestingly, the prodrugs for
GW7604 (NRMA-5) and F3MethylAA (NRMA-9), targeting ER and LXR, respectively, produced paltry, yet
quantifiable concentrations of parent drug in brain and serum, however, the limited overall exposure of each
is likely insufficient for any useful drug delivery purpose.
Table 1:List of parent drug and prodrug AUCs
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 38
1
2
3
4
5
6
7
8
9
GW501516
NRMA-6
41.8
146.3
7.3
2330
2705
2107
374.1
420.7
470.6
2140
276.5
314.5
51.34
0.018
0.054
0.004
0.06
Tesaglitazar
NRMA-7
21.9
5.5
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
Bezafibrate
NRMA-8
0.013
0.06
26.3
67.35
12.64
38.5
104
F3MethylAA
NRMA-9
0.03
0.05
Tazarotenic acid
NRMA-10
0.12
2.03
8
ACS Paragon Plus Environment

Page 9 of 38
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 2: Side-by-side 6 h AUCs for brain and serum for the series of parent drugs and their corresponding
prodrugs. Cohorts of male C57BL/6 mice (n=3) were administered (i.p.) 9.14 mol/kg parent drug or
equimolar prodrug and tissues were collected at 0.5 h, 2 h, and 6 h time points. Statistical significance was
determined by a 2-tailed, unpaired t test for comparisons between parent drug and prodrug at the same time
point denoted with asterisks (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). All graphs show mean ± SEM.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 38
1
2
3
4
5
6
7
8
9
To confirm FAAH as the critical hydrolase responsible for prodrug cleavage, FAAH inhibition
experiments were performed using the FAAH inhibitor PF-3845 that inhibits FAAH in both the CNS and
periphery from a systemic dose.22 Side-by-side comparisons of brain concentrations of nuclear receptor drug
were analyzed from mice receiving either equimolar prodrug or 1mg/kg PF-3845 followed by prodrug after 1
hour post-administration of the inhibitor (Figure 3). Mice receiving only prodrug displayed high
concentrations of parent drug in the brain, while those that received prodrug and FAAH inhibitor generally
showed diminished prodrug-to-parent drug conversion. Differences in parent drug levels were observed
between each pair except for those prodrugs which weren’t cleaved in vivo, suggesting the prodrugs are
hydrolyzed specifically by FAAH in the brain. However, NRMA-1, and to a much lesser extent NRMA-6,
produced significant quantities of parent drug in the brain in the presence of the FAAH inhibitor. This
indicates promiscuous cleavage of these prodrugs likely in peripheral circulation prior to distribution into the
CNS by hydrolases other than FAAH, which corroborates with the relatively high serum drug levels observed
in the AUCs following systemic administration (Table 1).
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: Comparison of cleaved parent drug levels in brain with or without the presence of global FAAH
inhibitor PF-3845 1 h post-dose. Cohorts of male C57BL/6 mice (n=3) were administered (i.p.) 9.14 mol/kg
prodrug or 1 mg/kg PF-3845 followed by 9.14 mol/kg prodrug after 1 h, then brain tissue was collected 1 h
after prodrug administration. Statistical significance was determined by a 2-tailed, unpaired t test for
comparisons between prodrug groups with and without receiving FAAH inhibitor and is denoted with
asterisks (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001). All graphs show mean ± SEM.
10
ACS Paragon Plus Environment

Page 11 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
FAAH hydrolysis and SAR
8
9
In an effort to extrapolate information regarding FAAH’s substrate capacity with respect to these
drug-like structures and how that relates to what makes a suitable prodrug, characterization of in vitro FAAH
hydrolysis rates were determined (Figure 4A). Within the series, FAAH cleavage rates using a mouse liver S9
fraction ranged from ~1 to >300 pmol*mg-1min-1. NRMA-6 and NRMA-1 were found to be cleaved the fastest,
which mirrors their rather facile peripheral cleavage following systemic administration previously observed
in the AUC study. Somewhat surprisingly, the prodrugs which altered the brain-to-serum ratio to the greatest
extent had among the slowest FAAH hydrolysis rates at only about 1-2 pmol*mg-1min-1. In fact, a negative
correlation was observed between FAAH hydrolysis rate and Kp, the change in brain-to-serum ratio from
parent drug to prodrug (Figure 4B). These data suggest that prodrugs which survive peripheral circulation,
likely due to their ability to act as poor substrates for FAAH expressed in the liver (or periphery in general) −
yet substrates nonetheless, allows for a greater fraction of prodrugs to segregate into the CNS for cleavage by
FAAH expressed in the CNS. In this way, the in vitro hydrolysis rate may serve as a potential indicator for the
CNS selectivity of a particular prodrug. Hydrolysis rates using a human brain S9 fraction as the source of
FAAH were also collected and recapitulate the results found in mouse liver S9 fractions, however greater
FAAH activity was generally observed in mouse liver compared to human brain (Figure 4C). Additionally,
there was an observed increase in NRMA-8 cleavage and a decrease in NRMA-9 cleavage in human brain S9
fraction compared to mouse liver S9 fraction. The augmented conversion of NRMA-8 in human S9 fraction
could suggest potential selectivity towards FAAH2. Nevertheless, the human brain S9 fraction activity
generally validates the translational potential of this prodrug strategy for use with human therapeutics.
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
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 38
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: In vitro FAAH hydrolysis rates for the series of prodrugs in mouse liver S9 fraction (A) and human
brain S9 fraction (C). Trending relationship between in vitro hydrolysis rates and CNS-selectivity from AUC
data (vide supra, Figure 2) for C57BL/6 mice (B).
12
ACS Paragon Plus Environment

Page 13 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure-activity relationships were analyzed to understand why certain prodrugs are not cleaved
efficiently by FAAH. Basic physicochemical properties like MW, lipophilicity, topological polar surface area,
number of rotatable bonds etc., failed to exhibit any sort of rational correlation with the prodrug CNS
distribution data (see boiled egg plot SI Figure S1). However, analysis of each prodrug’s molecular shape
elucidated a connection between structure and CNS distribution of the parent drug. Principal moments of
inertia (PMI) calculations on molecular mechanics optimized structures can afford a two-dimensional PMI
plot which triangulates the overall shape of a particular molecule (Figure 5).20, 23 Prodrugs that displayed
CNS selectivity within this series which were inhibited by FAAH inhibitor PF-3845, all reside close to the
vertex which represents linear, rod-like structures. Those prodrugs not cleaved appreciably by FAAH deviate
from linearity and spread closer toward the flat, disc-like molecular space, and to a lesser extent, the spherical
vertex. This PMI plot shows the preference for linear structures as acceptable substrates for FAAH,
disfavoring those with more exaggerated or branched three-dimensional structures (NRMA-3, 4, and 5). As
the endogenous substrates for FAAH are amides of fatty acids, it’s logical that prodrugs with structural
similarities to fatty acids would make admissible substrates within FAAH’s active site. This trend was
recognized prior to the completion of the series of prodrugs, and NRMA-9 and NRMA-10 were synthesized
and studied to validate this structural preference. Unlike the highly branched LXR agonist prodrug NRMA-3,
NRMA-9 is an LXR agonist prodrug that features a slimmer, more linear shape. Although NRMA-9 did not alter
the biodistribution of its corresponding parent drug in a usable manner, it did lead to small, but measurable
levels of cleaved parent drug in contrast to NRMA-3. Moreover, inclusion of NRMA-10 led to a bona fide CNS-
selective prodrug with a Kp of 2.03, thus validating molecular shape as the critical property which defines
substrate capacity for FAAH. For practical use as a predictor, molecules with PMI I1/I3 < 0.2 and PMI I2/I3 >
0.9 appear to make suitable substrates for FAAH, at least as defined within this series.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 38
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 5: Principal moments of inertia plot for the prodrug series including FAAH inhibitor PF-3845
demonstrating the molecular shape preference for acceptable FAAH substrates aggregating toward the linear,
rod-like shape vertex (top left).
Proof of concept: CNS vs peripheral effects
To validate that increased CNS exposure and decreased peripheral exposure lead to the appropriate
transcriptional responses, an experiment to quantify target gene expression from the brain and peripheral
tissues was performed. Cohorts of male C57BL/6 mice were dosed once daily for seven days with either
parent drug, equimolar prodrug, or vehicle to compare target gene expression levels between groups per
tissue. NRMA-7 and NRMA-10 were chosen to study due to their exceptional CNS distribution properties, with
15-fold and 17-fold increases in Kp relative to their corresponding parent drugs, respectively. NRMA-7 is the
14
ACS Paragon Plus Environment

Page 15 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
prodrug of dual PPAR/ agonist tesaglitazar, and PPAR agonists have been shown to protect dopaminergic
neurons and reduce amyloid plaque burden.24-27 However, tesaglitazar failed in clinical trials due to renal and
cardiac safety issues, making tesaglitazar an ideal candidate for a drug that may benefit from this prodrug
strategy.28-29 Indeed, after seven days of dosing prodrug NRMA-7 was shown to significantly upregulate the
PPAR target gene Pgc1- (PPAR gamma coactivator 1 alpha) in the brain, while having no effect on
expression of Fasn (fatty acid synthase) in the kidney (Figure 6C+D). In contrast, administration of the
unmodified parent drug tesaglitazar showed no effect on Pgc1- in the brain, but downregulated kidney Fasn.
Tazarotenic acid, the parent drug of prodrug NRMA-10, is a pan-RAR agonist which positively regulates the
main alpha secretase Adam10 responsible for amyloid precursor protein processing.30 Seven days of
treatment with tazarotenic acid does not significantly upregulate target brain gene Adam10, however, the
corresponding prodrug NRMA-10 does increase brain expression of Adam10. Conversely, tazarotenic acid
treatment upregulates the liver target gene Rarres2, while NRMA-10 had no effect (Figure 6A+B).28 As
predicted, the CNS selectivity of these prodrugs affords greater exposure and subsequent target engagement
in the CNS while attenuating exposure and target engagement in the periphery. In this way, these prodrugs
offer not only greater on-target potency in the CNS, but also less potential on-target adverse effects in the
periphery resulting in a larger therapeutic index.
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: RT-qPCR comparisons between administered parent drug and prodrug for CNS and peripheral
tissue genes. Cohorts of male C57BL/6 mice (n=5) were administered (i.p.) 9.14 mol/kg parent drug,
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 38
1
2
3
4
5
6
7
8
equimolar prodrug, or vehicle for 7 days once daily by i.p. Statistical significance was determined by a 2-
tailed, unpaired t test for comparisons between groups with asterisks (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). All
graphs show mean ± SEM.
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
Conclusions
This study demonstrates that CNS-selective distribution of drugs can be achieved utilizing a FAAH-
targeted prodrug strategy. In order to successfully implement: 1) the parent drug must contain a carboxylic
acid functional group; 2) the parent drug must have sufficient drug-like properties and largely obey Lipinski’s
Rule of 5 to facilitate passive diffusion of the prodrug across the BBB; 3) conversion of the parent drug’s
carboxylic acid to an amide defines the prodrug; 4) the newly generated amide prodrug must have an overall
linear shape as defined by PMI calculations (PMI I1/I3 < 0.2 and PMI I2/I3 > 0.9). Additionally, we find that
prodrugs with slower in vitro FAAH-mediated hydrolysis rates (< 5 pmol*mg-1min-1 in this experiment in
mice) provided the highest degree of CNS selectivity, presumably by avoiding peripheral FAAH cleavage in
the liver and/or gut. Given that this prodrug strategy is generalizable to carboxylate-containing drugs other
than sobetirome, this blueprint should find utility within the fields of drug discovery and neuroscience and
facilitate the development of new CNS drugs and tool compounds.
Experimental Section
Parent drug abbreviations
All parent drugs were purchased from commercial sources with the exception of NH-3 which was synthesized
in our laboratory.32
Triac = [4-(4-hydroxy-3-iodophenoxy)-3,5-diiodophenyl]acetic acid
NH-3 = 2-(4-(4-hydroxy-3-isopropyl-5-((4-nitrophenyl)ethynyl)benzyl)-3,5-dimethylphenoxy)acetic acid
16
ACS Paragon Plus Environment

Page 17 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
GW3965 = 2-(3-(3-((2-chloro-3-(trifluoromethyl)benzyl)(2,2-diphenylethyl)amino)propoxy)phenyl)acetic
acid
7
8
9
Bexarotene = 4-(1-(3,5,5,8,8-pentamethyl-5,6,7,8-tetrahydronaphthalen-2-yl)vinyl)benzoic acid
GW7604 = (E)-3-(4-((E)-1-(4-hydroxyphenyl)-2-phenylbut-1-en-1-yl)phenyl)acrylic acid
GW501516 = 2-(2-methyl-4-(((4-methyl-2-(4-(trifluoromethyl)phenyl)thiazol-5-
yl)methyl)thio)phenoxy)acetic acid
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
Tesaglitazar = (S)-2-ethoxy-3-(4-(4-((methylsulfonyl)oxy)phenethoxy)phenyl)propanoic acid
Bezafibrate = 2-(4-(2-(4-chlorobenzamido)ethyl)phenoxy)-2-methylpropanoic acid
F3MethylAA = 2-(3-chloro-4-((3-((7-propyl-3-(trifluoromethyl)benzo[d]isoxazol-6-
yl)oxy)propyl)thio)phenyl)acetic acid
Tazarotenic acid = 6-((4,4-dimethylthiochroman-6-yl)ethynyl)nicotinic acid
Animal Studies
Experimental protocols were in compliance with the National Institutes of Health Guide for the Care and Use
of Laboratory Animals and approved by the Oregon Health & Science University Institutional Animal Care &
Use Committee. Wild-type C57BL/6 mice, aged 8−10 weeks, were purchased from Jackson Laboratory and all
mice were housed in climate-controlled rooms with a 12 h/12 h light−dark cycle with ad libitum access to
food and water. FAAH-KO mice on a C57/BL6 background were a kind gift of Prof. Benjamin Cravatt (Scripps
Research Institute).33 All injections were delivered intraperitoneally (i.p.) using DMSO as a vehicle except for
chronic 7 day once-daily injections which used a 1:1:8 Kolliphor:NMP:water vehicle. Injection volumes were
standardized to 150 μL/26 g mouse weight. FAAH inhibitor PF-3845 was purchased from Sigma-Aldrich and
was used as received.
Tissue Processing
Tissues were processed for LC-MS/MS analysis as previously described with slight modifications.10, 34
Standard curves were prepared in vehicle treated samples for each tissue to control for ion suppression in the
analysis. In each run, tissue from a vehicle treated mouse was brought through the stages of processing and
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 38
1
2
3
4
5
6
7
8
split to generate a standard curve using known concentrations of each parent drug. Deuterated d6-sobetirome
was used as an internal standard in each sample due to its robust and well-known behavior regarding
extraction recovery, ionization response, and chromatographic retention time.32
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
Serum. Whole blood was incubated for 15 min on ice and then spun at 7,800 rpm for 15 min at 4 °C. Serum
(100 μL) was transferred to a new tube and stored at −80 °C until further processed. Serum samples were
thawed on ice prior to treatment with 10 μL of internal standards (2.99 μM d6-sobetirome in DMSO) followed
by a crash with 125 L of EtOAc followed by 400 μL of MeCN. Following a vigorous vortexing for 20 s,
samples were centrifuged at 10,000 rpm for 15 min at 4 °C. The supernatant is dried under high vacuum, then
samples were treated with 400 μL 10% DMF in MeCN with vigorous vortexing. Samples were transferred to
eppendorf tubes, centrifuged at 10,000 rpm for 15 min at 4 °C, and the supernatant was transferred and
centrifuged again at 13,000 rpm for 15 min at 4 °C prior to analysis by LC-MS/MS.
Brain. Tissue weights were determined prior to homogenization. Whole brains were added to 1 mL of water
containing 10 L internal standard (2.99 M d6-sobetirome) in 2 mL tubes containing 3 GoldSpec 1/8 chrome
steel balls (Applied Industrial Technologies) and tissues were homogenized using a Bead Ruptor 24. The
homogenate was crashed and extracted with 1 mL of EtOAc and 4 mL of MeCN, then centrifuged at 10,000
rpm for 15 min at 4 °C. The supernatant was transferred to glass culture tubes and dried under high vacuum.
The sample residue was then treated with 400 μL of 10% DMF in MeCN and vigorously vortexed for 20 s. The
sample was transferred to an eppendorf tube and centrifuged at 10,000 rpm for 15 min at 4 °C. The resulting
supernatant was transferred again and was centrifuged at 13,000 rpm for 15 min at 4 °C and submitted for
LC-MS/MS analysis.
Liver and Kidney. Liver and kidney tissues were processed exactly as brain tissue above.
LC-MS/MS Analysis
Prodrugs and internal standard were analyzed using a QTRAP 4000 hybrid/triple quadrupole linear ion trap
mass spectrometer (Applied Biosystems, (Foster City, CA)) with electrospray ionization (ESI) in negative
mode. The mass spectrometer was interfaced to a Shimadzu (Columbia, MD) SIL-20AC XR auto-sampler
18
ACS Paragon Plus Environment

Page 19 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
followed by 2 LC-20AD XR LC pumps and column oven. The mass spectrometer was operated with the
following settings: source voltage -4500 kV, GS1 30, GS2 40, CUR 10, TEM 650, and CAD MEDIUM. The
scheduled MRM transitions monitored are listed in the Supporting Information (Table S2). Compounds were
infused individually, and instrument parameters optimized for each MRM transition. The column was a 3µm
Gemini-NX C18 50 x 2.1 mm, kept at 40°C. The autosampler was kept at 15°C. The gradient mobile phase was
delivered at a flow rate of 0.75 ml/min, and consisted of two solvents, A: 0.1% formic acid in water and B:
0.1% formic acid in acetonitrile. The gradient was as follows: initial concentration of B was 10%, held for 1
minute, followed by an increase to 95% B over 5 minutes, held for 2 minutes, returning to 10%B over 0.1
minutes, and held at 10% B for 3 minutes for a total run time of 11 minutes. Column eluant was directed to
the ionization source from 0.5 minutes to 8 minutes. Data were acquired using Analyst 1.6.2 software and
analyzed using Multiquant 3.0.2.
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
RT-qPCR
RNA was purified using the PureLink RNA Mini kit with TRIzol extraction (Life Technologies, for whole brain,
liver, and kidney). RNA was quantified using a Nanodrop, and cDNA was prepared with the QuantiTect
Reverse Transcription kit (Qiagen). Each qPCR experiment was performed on an Applied Biosciences 7500
Real-Time PCR system using and following the QuantiTect SYBR Green PCR kit protocols (Qiagen). Results
were analyzed using the ΔΔCt relative quantification.35 Experiments were performed using primers for the
housekeeping gene cyclophilin A (f-Ppia: 5′-AGGGTGGTGACTTTACACGC-3′; r-Ppia: 5′-
CTTGCCATCCAGCCATTCAG-3′), and target gene primers PPAR gamma coactivator 1- (f- Pgc-1a 5′-
TGAGAGGGCCAAGCAAAG-3′, r- Pgc-1a 5′-ATAAATCACACGGCGCTCTT-3′ ), A disintegrin and
metalloproteinase domain-containing protein 10 (f-Adam10 5'-GGGAAGAAATGCAAGCTGAA-3', r-Adam10
5'-CTGTACAGCAGGGTCCTTGAC-3'), Rarres2 (chemerin) (f-Rarres2 5'-TACAGGTGGCTCTGGAGGAGTTC-
3', r-Rarres2 5'-CTTCTCCCGTTTGGTTTGATTG-3'), and fatty acid synthase (f-Fasn 5’-
GCTGCTGTTGGAAGTCAGC-3’, r-Fasn 5’-AGTGTTCGTTCCTCGGAGTG-3’).36-40
In Vitro FAAH Hydrolysis Assays
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 38
1
2
3
4
5
6
7
8
9
S9 subcellular fractions derived from WT C57BL/6 and FAAH KO mouse livers were isolated as previously
described and diluted into TE buffer (50 mM Tris HCl, 1 mM EDTA, 3 mM MgCl2, pH 7.4) to 1 mg/ml.41 Human
brain S9 fractions were purchased commercially from Sekisui XenoTech (Kansas City, Kansas). Protein
concentrations were measured using a BCA protein assay (Pierce).
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
For observed rates, 49 L of S9 fraction at 1 mg/mL (WT mouse, FAAH KO mouse, human brain, or human
brain containing 10 M PF-3845) was prewarmed at 37 °C in a water bath for 5 min, followed by the addition
of 1 L of 5 mM prodrug (working concentration of 100 M) or DMSO vehicle and briefly vortexed. These
mixtures were incubated at 37 °C for 15 min then crashed with 250 L of 29.9 nM d6-sobetirome internal
standard in 10% DMF in MeCN, vortexed, and centrifuged at 10,000 rpm for 15 min at 4 °C. Samples were
prepared from this supernatant fresh immediately prior to submission for LC-MS/MS as described above. For
each prodrug, concentrations were determined from: [WT liver S9 sample (n=3) – average of FAAH KO liver
S9 samples (n>2)] or [human brain S9 sample (n=3) – average of human brain S9 samples containing 10 m
PF-3845 (n>2)]. In each run, a single S9 fraction containing 1 L DMSO vehicle instead of prodrug was split
and used to generate the standard curve.
PMI calculations and Plot
Principal moments of inertia (PMI) for each molecule were calculated using Chem3D (Version 15.1.0.144,
Perkin Elmer). Chemical structures drawn in ChemDraw (Perkin Elmer) were brought into Chem3D and an
MM2 energy minimization was performed. On this minimized structure, principal moments of inertia were
calculated to afford three values which correspond to the principal ‘vectors’ of the molecule I1, I2, and I3. The
two shorter vectors (I1 and I2) are each divided by the longest vector (I3) and plotted on a two-dimensional
graph (PMI plot). Vertices are represented by 1,3,5-hexatriyne (triacetylene), benzene, and adamantane to
describe the molecular shapes as linear and rod-like, flat and disc-like, and round/spherical, respectively. All
molecules calculated fall within this triangular plot which characterizes each molecule’s shape.
Statistical Analysis
20
ACS Paragon Plus Environment

Page 21 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Statistical significance was determined using 2-tailed, unpaired t tests for comparisons between two specific
groups then plotted together as indicated. Replicates in each experiment were as stated in the specific figure
legend and in the corresponding methods. Experimental numbers for animal experiments were informed by
previous literature accounts in order to minimize total animal numbers as appropriate. Analysis was carried
out in GraphPad Prism 7 without further modifications. Significance level was set to *P ≤ 0.05, **P ≤ 0.01,
***P ≤ 0.001, ****P ≤ 0.0001. All graphs show mean ± SEM.
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
General Chemistry
1 H NMR were taken on a Bruker 400 MHz spectrometer. All spectra were calibrated to the chloroform CDCl3
NMR solvent reference peak. 1 H coupling constants (JHH, Hz) are reported as follows: chemical shift,
multiplicity (br = broad, s = singlet, d = doublet, sept = septet, m = multiplet, dd = doublet of doublets),
coupling constant, and integration. High resolution mass spectrometry (HRMS) with electrospray ionization
was performed by the Bioanalytical MS Facility at Portland State University. d6-sobetirome was synthesized
as previously described.29 All other reagents were purchased from Fisher, Sigma, or TCI and used as received.
Analytical HPLC analysis was performed on a Varian ProStar HPLC with a Grace Altima C18, 5 μm column (4.6
mm × 250 mm) with a gradient (solvent A = water + 0.1% TFA; solvent B = MeCN + 0.1% TFA) over 30 or 40
mins. with a flow rate of 1 mL/min. Purity analysis of final compounds was determined to be >95% by HPLC
(A235, 254, or 280 nm).
General synthetic procedure
35 mg (1 equivalent) of parent drug carboxylic acid was placed in a thick-walled tube containing
carbonyldiimidazole (CDI, 1.2 equivalents), a stir bar, and 3mL of dry THF. The tube was sealed, placed under
vacuum, then refilled with argon three times. While under argon, the mixture was heated for 2 h at 50° C then
cooled to room temperature. Three additional vacuum/argon refills were performed followed by the addition
of excess methylamine (~10 equivalents, 2M in THF) via syringe. The mixture was stirred at room
temperature for 2 h, then heated to 50° C for ~14 h. Upon cooling, the product mixture was diluted with
dichloromethane and washed with 1N HCl and brine. All organic layers were combined, dried with sodium
sulfate, and evaporated to dryness to give the crude product mixture typically containing only starting
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 38
1
2
3
4
5
6
material and product by TLC. The product was separated on silica with 1-10% MeOH in DCM usually as the
first band.
7
8
NRMA-1
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
2-(4-(4-hydroxy-3-iodophenoxy)-3,5-diiodophenyl)-N-methylacetamide
35 mg triac (0.056 mmol) yielded 30.45 mg (0.048 mmol, 87%) product as an off-white powder.
1H NMR (400 MHz, Chloroform-d) δ 7.78 (s, 2H), 7.10 (dd, J = 2.9, 1.1 Hz, 1H), 6.94 – 6.87 (m, 1H), 6.70 (s, 1H),
3.75 (s, 3H), 3.57 (s, 2H). HRMS (ESI) m/z [M+1]+ C15H13I3N1O3 + requires 635.8024, found 635.8049. HPLC
purity: 95.2%.
NRMA-2
2-(4-(4-hydroxy-3-isopropyl-5-((4-nitrophenyl)ethynyl)benzyl)-3,5-dimethylphenoxy)-N-
methylacetamide
35 mg NH-3 (0.074 mmol) yielded 18.7 mg (0.038 mmol, 52%) product as a yellow powder.
1H NMR (400 MHz, Chloroform-d) δ 8.24 (d, J = 8.9 Hz, 2H), 7.65 (d, J = 8.9 Hz, 2H), 7.04 (s, 1H), 6.74 (s, 1H),
6.67 (s, 2H), 5.77 (s, 1H), 4.52 (s, 2H), 3.93 (s, 2H), 3.29 (septet, 1H), 2.94 (d, J = 5.0 Hz, 3H), 2.26 (s, 6H), 1.26
(t, J = 6.1 Hz, 6H). HRMS (ESI) m/z [M+1]+ C29H31N2O5 + requires 487.2227, found 487.2225. HPLC purity:
95.1%.
NRMA-3
2-(3-(3-((2-chloro-3-(trifluoromethyl)benzyl)(2,2-diphenylethyl)amino)propoxy)phenyl)-N-
methylacetamide
35 mg GW3965 (0.06 mmol) yielded 13.6 mg (0.023 mmol, 38%) product as a colorless residue. Note: during
workup, the crude reaction mixture was washed with 0.5M NaOH(aq) and separated on column to give the
charge neutral species.
22
ACS Paragon Plus Environment

Page 23 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1H NMR (400 MHz, Chloroform-d) δ 7.50 (d, J = 7.1 Hz, 1H), 7.16-7.29 (m, 12H), 6.96 (t, 1H), 6.85 (d, J = 7.5 Hz,
1H), 6.74 – 6.64 (m, 2H), 5.42 (s, 1H), 4.16 (t, 1H), 3.81 (s, 2H), 3.73 (t, 2H), 3.57 (s, 2H), 3.17 (d, J = 7.7 Hz,
2H), 2.76 (s, 3H), 2.73 (t, 2H), 1.88 (pentet, 2H). HRMS (ESI) m/z [M+1]+ C34H35ClF3N2O2 + requires 595.2339,
found 595.2332. HPLC purity: 95.2%.
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
NRMA-4
N-methyl-4-(1-(3,5,5,8,8-pentamethyl-5,6,7,8-tetrahydronaphthalen-2-yl)vinyl)benzamide
35 mg bexarotene (0.1 mmol) yielded 29.3 mg (0.081 mmol) product as a white powder.
1H NMR (400 MHz, Chloroform-d) δ 7.67 (d, 2H), 7.33 (d, 2H), 7.12 (s, 1H), 7.09 (s, 1H), 6.09 (s, 1H), 5.78 (s,
1H), 5.29 (s, 1H), 3.01 (d, J = 4.7 Hz, 3H), 1.93 (s, 3H), 1.69 (s, 4H), 1.30 (s, 6H), 1.27 (s, 6H). HRMS (ESI) m/z
[M+1]+ C25H32N1O1 + requires 362.2478, found 362.2486. HPLC purity: 97.2%.
NRMA-5
(E)-3-(4-((E)-1-(4-hydroxyphenyl)-2-phenylbut-1-en-1-yl)phenyl)-N-methylacrylamide
35 mg GW7604 (0.094 mmol) yielded 8.3 mg (0.021 mmol, 23%) product as a yellowish residue.
1H NMR (400 MHz, Chloroform-d) δ 7.65 (d, J = 15.6 Hz, 1H), 7.44 (m, 1H), 7.24 (m, 2H), 7.20-7.07 (m, 3H),
6.87 (m, 2H), 6.73 (d, J = 8.1 Hz, 1H), 6.53 (d, J = 8.1 Hz, 1H), 6.40 (d, J = 15.5 Hz, 1H), 6.22 (d, J = 15.6 Hz, 1H),
5.69 (s, 1H), 5.59 (s, 1H), 2.96 (d, J = 19.8 Hz, 3H), 2.52 (q, 2H), 0.95 (t, 3H). HRMS (ESI) m/z [M+1]+
C26H26N1O2 + requires 384.1958, found 384.1958. HPLC purity: 97.4%.
NRMA-6
N-methyl-2-(2-methyl-4-(((4-methyl-2-(4-(trifluoromethyl)phenyl)thiazol-5-
yl)methyl)thio)phenoxy)acetamide
35 mg GW501516 (0.077 mmol) yielded 27.7 mg (0.059 mmol, 77%) product as an off-white powder.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 38
1
2
3
4
5
6
7
8
1H NMR (400 MHz, Chloroform-d) δ 7.98 (d, J = 8.1 Hz, 2H), 7.67 (d, J = 8.2 Hz, 2H), 7.22 (s, 1H), 7.19 (d, 1H),
6.70 (d, J = 9.1 Hz, 1H), 6.53 (s, 1H), 4.49 (s, 2H), 4.14 (s, 2H), 2.94 (d, J = 5.0 Hz, 3H), 2.24 (s, 3H), 2.22 (s, 3H).
HRMS (ESI) m/z [M+1]+ C22H22F3N2O2S2 + requires 467.1069, found 467.1067. HPLC purity: 97.8%.
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
NRMA-7
(S)-4-(2-(4-(2-ethoxy-3-(methylamino)-3-oxopropyl)phenoxy)ethyl)phenyl methanesulfonate
35 mg Tesaglitazar (0.086 mmol) yielded 15.6 mg (0.037 mmol, 43%) product as a yellow oil.
1H NMR (400 MHz, Chloroform-d) δ 7.36 (d, J = 8.7 Hz, 2H), 7.28 (d, J = 8.7 Hz, 2H), 7.14 (d, J = 8.7 Hz, 2H),
6.81 (d, J = 8.7 Hz, 2H), 6.48 (s, 1H), 4.16 (t, 2H), 3.91 (m, 1H), 3.44 (m, 1H), 3.13 (s, 3H), 3.11 (m, 2H), 2.86 (m,
1H), 2.79 (d, J = 5.0 Hz, 3H), 1.15 (t, 3H). HRMS (ESI) m/z [M+1]+ C21H28N1O6S1 + requires 422.1632, found
422.1638. HPLC purity: 96.7%.
NRMA-8
4-chloro-N-(4-((2-methyl-1-(methylamino)-1-oxopropan-2-yl)oxy)phenethyl)benzamide
35 mg Bezafibrate (0.097 mmol) yielded 29.8 mg (0.079 mmol, 82%) product as a white powder.
1H NMR (400 MHz, Chloroform-d) δ 7.63 (d, J = 8.6 Hz, 2H), 7.38 (d, J = 8.6 Hz, 2H), 7.12 (d, J = 8.5 Hz, 2H),
6.86 (d, J = 8.5 Hz, 2H), 6.76 (s, 1H), 6.21 (s, 1H), 3.67 (q, J = 6.1 Hz, 2H), 2.90 (m, 2H), 2.88 (s, 3H), 1.48 (s,
6H). HRMS (ESI) m/z [M+1]+ C20H24ClN2O3 + requires 375.1470, found 375.1472. HPLC purity: 97.4%.
NRMA-9
2-(3-chloro-4-((3-((7-propyl-3-(trifluoromethyl)benzo[d]isoxazol-6-yl)oxy)propyl)thio)phenyl)-N-
methylacetamide
35 mg F3MethylAA (0.072 mmol) yielded 11.2 mg (0.022 mmol, 31%) product as a white powder.
1H NMR (400 MHz, Chloroform-d) δ 7.58 (d, J = 8.8 Hz, 1H), 7.31 (m, J = 3.9 Hz, 2H), 7.15 (d, J = 9.9 Hz, 1H),
7.07 (d, J = 8.8 Hz, 1H), 5.52 (s, 1H), 4.26 (t, 2H), 3.50 (s, 2H), 3.20 (t, J = 7.1 Hz, 2H), 2.94 (t, 2H), 2.80 (d, J =
24
ACS Paragon Plus Environment

Page 25 of 38
Journal of Medicinal Chemistry
1
2
3
4
4.8 Hz, 3H), 2.24 (pentet, 2H), 1.72 (sextet, J = 7.5 Hz, 2H), 0.98 (t, 3H). HRMS (ESI) m/z [M+1]+
5
6
C23H25ClF3N2O3S1 + requires 501.1221, found 501.1218. HPLC purity: 97.7%.
7
8
NRMA-10
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
6-((4,4-dimethylthiochroman-6-yl)ethynyl)-N-methylnicotinamide
35 mg Tazarotenic acid (0.11 mmol) yielded 29.2 mg (0.087 mmol, 79%) product as a yellow powder.
1H NMR (400 MHz, Chloroform-d) δ 8.95 (s, 1H), 8.12 (dd, J = 8.1, 2.2 Hz, 1H), 7.64 – 7.54 (m, 2H), 7.31 – 7.21
(m, 1H), 7.08 (d, J = 8.1 Hz, 1H), 6.54 (s, 1H), 3.10 – 3.02 (m, 2H), 3.05 (s, 3H), 2.00 – 1.92 (m, 2H), 1.34 (s, 6H).
HRMS (ESI) m/z [M+1]+ C20H21N2O1S1 + requires 337.1369, found 337.1374. HPLC purity: 96.8%.
Associated Content
Author Information
Corresponding authors
*E-mail: scanlant@ohsu.edu
ORCHID
Skylar J. Ferrara https://orcid.org/0000-0001-7826-5621
Thomas S. Scanlan https://orcid.org/0000-0003-2612-5445
Funding
This work was supported by the National Institutes of Health (DK 052798, T.S.S., and T32DK007680, support
to S.J.F.).
Notes
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 38
1
2
3
4
5
6
7
8
The authors declare the following competing financial interest(s): S.J.F. and T.S.S. are inventors of a licensed
patent application claiming central nervous system-penetrating prodrugs of nuclear receptor modulators and
their uses. T.S.S. is a founder of Autobahn Therapeutics.
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
Acknowledgements
We thank Lisa Bleyle (OHSU Bioanalytical Shared Resource/ Pharmacokinetics Core) for technical assistance
with LC-MS/ MS, Prof. Benjamin Cravatt (Scripps Research Institute) for the FAAH-KO mice, and Dr. J.
Matthew Meinig for preparation and isolation of mouse liver S9 fractions.
Abbreviations
AUC, area under the curve, i.p. intraperitoneal; RT-qPCR, real-time quantitative polymerase chain reaction;
WT, wild type
Supporting Information
The Supporting Information contains tabulated physicochemical properties of the series, mass spectrometry
information, 1H NMR spectra (PDF) and molecular formula strings (CSV), and is available free of charge on the
ACS Publications website at DOI:
References
1.
Neurobiology of Disease 2014, 72, 104-116.
2. Moutinho, M.; Codocedo, J. F.; Puntambekar, S. S.; Landreth, G. E., Nuclear Receptors as
Skerrett, R.; Malm, T.; Landreth, G., Nuclear Receptors in Neurodegenerative Diseases.
Therapeutic Targets for Neurodegenerative Diseases: Lost in Translation. Annual Review of
Pharmacology and Toxicology 2019, 59, 237-261.
26
ACS Paragon Plus Environment

Page 27 of 38
Journal of Medicinal Chemistry
1
2
3
4
3.
Kabat, G. C.; Etgen, A. M.; Rohan, T. E., Do Steroid Hormones Play a Role in the Etiology of
5
6
7
8
Glioma? Cancer Epidemiology Biomarkers &amp; Prevention 2010, 19, 2421-2427.
4. Littleton-Kearney, M. T.; Ostrowski, N. L.; Cox, D. A.; Rossberg, M. I.; Hurn, P. D., Selective
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
Estrogen Receptor Modulators: Tissue Actions and Potential for CNS Protection. CNS Drug Reviews 2002,
8, 309-330.
5.
Kim, C.-H.; Han, B.-S.; Moon, J.; Kim, D.-J.; Shin, J.; Rajan, S.; Nguyen, Q. T.; Sohn, M.; Kim, W.-G.;
Han, M.; Jeong, I.; Kim, K.-S.; Lee, E.-H.; Tu, Y.; Naffin-Olivos, J. L.; Park, C.-H.; Ringe, D.; Yoon, H. S.;
Petsko, G. A.; Kim, K.-S., Nuclear Receptor Nurr1 Agonists Enhance Its Dual Functions and Improve
Behavioral Deficits in an Animal Model of Parkinson’s Disease. Proceedings of the National Academy of
Sciences 2015, 112, 8756-8761.
6.
Reviews Drug Discovery 2006, 5, 993-996.
7. Moore, J. T.; Collins, J. L.; Pearce, K. H., The Nuclear Receptor Superfamily and Drug Discovery.
ChemMedChem 2006, 1, 504-523.
8. Ana Maria, A.; Oscar Andréas, M.-R.; Haider, N. B., Role of Nuclear Receptors in Central Nervous
Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L., How Many Drug Targets Are There? Nature
System Development and Associated Diseases. Journal of Experimental Neuroscience 2015, 9s2,
JEN.S25480.
9.
Alzheimer’s Disease. Journal of Lipid Research 2017, 58, 1937-1949.
10. Meinig, J. M.; Ferrara, S. J.; Banerji, T.; Banerji, T.; Sanford-Crane, H. S.; Bourdette, D.; Scanlan, T.
Moutinho, M.; Landreth, G. E., Therapeutic Potential of Nuclear Receptor Agonists in
S., Targeting Fatty-Acid Amide Hydrolase with Prodrugs for CNS-Selective Therapy. ACS Chemical
Neuroscience 2017, 8, 2468-2476.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 38
1
2
3
4
11.
Meinig, J. M.; Ferrara, S. J.; Banerji, T.; Banerji, T.; Sanford-Crane, H. S.; Bourdette, D.; Scanlan, T.
5
6
S., Structure–Activity Relationships of Central Nervous System Penetration by Fatty Acid Amide
7
8
Hydrolase (FAAH)-Targeted Thyromimetic Prodrugs. ACS Medicinal Chemistry Letters 2019, 10, 111-116.
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
12.
Hartley, M. D.; Banerji, T.; Tagge, I. J.; Kirkemo, L. L.; Chaudhary, P.; Calkins, E.; Galipeau, D.;
Shokat, M. D.; DeBell, M. J.; Van Leuven, S.; Miller, H.; Marracci, G.; Pocius, E.; Banerji, T.; Ferrara, S. J.;
Meinig, J. M.; Emery, B.; Bourdette, D.; Scanlan, T. S., Myelin Repair Stimulated by CNS-selective Thyroid
Hormone Action. JCI Insight 2019, 4, e126329.
13.
Hartley, M. D.; Shokat, M. D.; DeBell, M. J.; Banerji, T.; Kirkemo, L. L.; Scanlan, T. S.,
Pharmacological Complementation Remedies an Inborn Error of Lipid Metabolism. Cell Chemical Biology
2020, 27, 551-559.
14.
Vu, C.; Bemis, J.K.; Benson, E.; Bista, P.; Carney, D.; Fahrner, R.; Lee, D.; Liu, F.; Lonkar, P.; Milne,
J.C.; Nichols, A.J.; Picarella, D.; Shoelson, A.; Smith, J.; Ting, A.; Wensley, A.; Yeager,M.; Zimmer, M.;
Jirousek, M.R., Synthesis and Characterization of Fatty Acid Conjugates of Niacin and Salicylic Acid.
Journal of Medicinal Chemistry 2016, 59, 1217-1231.
15.
Vu, C.B.; Bridges, R.J.; Pena-Rasgado, C.; Lacerda, A.E.; Bordwell, C.; Sewell, A.; Nichols, A.J.;
Chandran, S.; Lonkar, P.; Picarella, D.; Ting, A.; Wensley, A.; Yeager, M.; Liu, F., Fatty acid cysteamine
conjugates as Novel and Potent Autophagy Activators That Enhance the Correction of Misfolded
F508del-Cystic Fibrosis Transmembrane Conductance Regulator (CFTR). Journal of Medicinal
Chemistry 2017, 60, 458-473.
16.
Mofford, D.M.; Adams, S.T.; Reddy, G.S.K.K.; Reddy, G.R.; Miller, S.C., Luciferin Amides Enable in
Vivo Bioluminescence Detection of Endogenous Fatty Acid Amide Hydrolase Activity. Journal of the
American Chemical Society 2015, 137, 8684-8687.
17.
McKinney, M. K.; Cravatt, B. F., Structure and Function of Fatty Acid Amide Hydrolase. Annual
Review of Biochemistry 2005, 74, 411-432.
28
ACS Paragon Plus Environment

Page 29 of 38
Journal of Medicinal Chemistry
1
2
3
4
18.
Giang, D. K.; Cravatt, B. F., Molecular Characterization of Human and Mouse Fatty
5
6
7
8
AcidꢀAmideꢀHydrolases. Proceedings of the National Academy of Sciences 1997, 94, 2238-2242.
19. Li, Y.; Lambert, M. H.; Xu, H. E., Activation of Nuclear Receptors: A Perspective from Structural
Genomics. Structure 2003, 11, 741-746.
20. Aldeghi, M.; Malhotra, S.; Selwood, D. L.; Chan, A. W. E., Two- and Three-dimensional Rings in
Drugs. Chemical Biology & Drug Design 2014, 83, 450-461.
21. Lipinski, C. A., Drug-like Properties and the Causes of Poor Solubility and Poor Permeability.
Journal of Pharmacological and Toxicological Methods 2000, 44, 235-249.
22. Ahn, K.; Johnson, D. S.; Mileni, M.; Beidler, D.; Long, J. Z.; McKinney, M. K.; Weerapana, E.;
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
Sadagopan, N.; Liimatta, M.; Smith, S. E.; Lazerwith, S.; Stiff, C.; Kamtekar, S.; Bhattacharya, K.; Zhang, Y.;
Swaney, S.; Van Becelaere, K.; Stevens, R. C.; Cravatt, B. F., Discovery and Characterization of a Highly
Selective FAAH Inhibitor that Reduces Inflammatory Pain. Chemistry & Biology 2009, 16, 411-420.
23.
Molecules. Future Medicinal Chemistry 2016, 8, 1753-1767.
24. Corona, J. C.; Duchen, M. R., PPARγ as a Therapeutic Target to Rescue Mitochondrial Function in
Neurological Disease. Free Radical Biology and Medicine 2016, 100, 153-163.
25. Jiang, Q.; Heneka, M.; Landreth, G. E., The Role of Peroxisome Proliferator-Activated Receptor-γ
PPARγ) in Alzheimer’s Disease. CNS Drugs 2008, 22, 1-14.
26. Moosecker, S.; Gomes, P.; Dioli, C.; Yu, S.; Sotiropoulos, I.; Almeida, O. F. X., Activated PPARγ
Meyers, J.; Carter, M.; Mok, N. Y.; Brown, N., On the Origins of Three-dimensionality in Drug-like
Abrogates Misprocessing of Amyloid Precursor Protein, Tau Missorting and Synaptotoxicity. Frontiers in
Cellular Neuroscience 2019, 13, 239.
27.
Villapol, S., Roles of Peroxisome Proliferator-Activated Receptor Gamma on Brain and Peripheral
Inflammation. Cellular and Molecular Neurobiology 2018, 38, 121-132.
29
ACS Paragon Plus Environment