Synthesis and Characterization of Urofuranoic Acids: In Vivo Metabolism of 2-(2-Carboxyethyl)-4-methyl-5-propylfuran-3-carboxylic Acid (CMPF) and Effects on in Vitro Insulin Secretion

Article abstract of DOI:10.1021/acs.jmedchem.6b01668

CMPF (2-(2-carboxyethyl)-4-methyl-5-propylfuran-3-carboxylic acid) is a metabolite that circulates at high concentrations in type 2 and gestational diabetes patients. Further, human clinical studies suggest it might have a causal role in these diseases. CMPF inhibits insulin secretion in mouse and human islets in vitro and in vivo in rodents. However, the metabolic fate of CMPF and the relationship of structure to effects on insulin secretion have not been significantly studied. The syntheses of CMPF and analogues are described. These include isotopically labeled molecules. Study of these materials in vivo has led to the first observation of a metabolite of CMPF. In addition, a wide range of CMPF analogues have been prepared and characterized in insulin secretion assays using both mouse and human islets. Several molecules that influence insulin secretion in vitro were identified. The molecules described should serve as interesting probes to further study the biology of CMPF.

Full text of DOI:10.1021/acs.jmedchem.6b01668

Subscriber access provided by AUSTRALIAN NATIONAL UNIV
Article
Synthesis and Characterization of Urofuranoic Acids: In Vivo
Metabolism of 2-(2-carboxyethyl)-4-methyl-5-propylfuran-3-
carboxylic acid (CMPF) and Effects on In Vitro Insulin Secretion.
Edith Nagy, Ying Liu, Kacey J. Prentice, Kyle W. Sloop, Phillip E. Sanders, Battsetseg
Batchuluun, Craig D. Hammond, Michael B. Wheeler, and Timothy B. Durham
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01668 • Publication Date (Web): 07 Feb 2017
Downloaded from http://pubs.acs.org on February 8, 2017
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and Characterization of Urofuranoic
Acids: In Vivo Metabolism of 2-(2-carboxyethyl)-
4-methyl-5-propylfuran-3-carboxylic 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
(CMPF) and Effects on In Vitro Insulin
Secretion.
1
2
2
1
1
Edith Nagy , Ying Liu , Kacey J. Prentice , Kyle W. Sloop , Phillip E. Sanders , Battsetseg
2
1
2
1
Batchuluun , Craig D. Hammond , Michael B. Wheeler , and Timothy B. Durham .*
1
Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN 46285, United States.
2
Department of Physiology, University of Toronto, Toronto, ON M5S 1A8, Canada.
KEYWORDS. CMPF, Type 2 Diabetes, Gestational Diabetes, Furanoic Acids, and Urofuranoic
Acids, Islets, Betaꢀcells, Insulin Secretion.
ABSTRACT. CMPF (2ꢀ(2ꢀcarboxyethyl)ꢀ4ꢀmethylꢀ5ꢀpropylfuranꢀ3ꢀcarboxylic acid) is a
metabolite which circulates at high concentrations in Type 2 and gestational diabetes patients.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Further, human clinical studies suggest it might have a causal role in these diseases. CMPF
inhibits insulin secretion in mouse and human islets in vitro and in vivo in rodents. However, the
metabolic fate of CMPF and the relationship of structure to effects on insulin secretion have not
been significantly studied. The synthesis of CMPF and analogues are described. These include
isotopically labeled molecules. Study of these materials in vivo has led to the first observation of
a metabolite of CMPF. In addition, a wide range of CMPF analogues have been prepared and
characterized in insulin secretion assays using both mouse and human islets. Several molecules
that influence insulin secretion in vitro were identified. The molecules described should serve as
interesting probes to further study the biology of CMPF.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
1
Diabetes is a significant health crisis. In 2012, the US Centers for Disease Control
reported 9.3% of the US population as being afflicted, with roughly 40% of those cases being
undiagnosed. Diabetes leads to a number of debilitating coꢀmorbidities that have significant
impact on the health care system. These morbidities include cardiovascular disease, stroke,
neuropathy, blindness, and amputation.
Diabetes presents in multiple forms. Type 1 diabetes mellitus (T1DM) is immune system
driven and typically occurs in the young. Type 2 diabetes mellitus (T2DM) can occur at any
stage of life. Gestational diabetes (GD) occurs during pregnancy. In the US, approximately 9.2%
2
of pregnant women experience GD. Typically, GD resolves after delivery of the baby, but in
some cases patients progress to T2DM. GD can cause a number of health problems for the
mother, including preeclampsia. In some GD cases, the fetus grows larger which leads to
ACS Paragon Plus Environment
2

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
difficulties during birth. In addition, high birth weight children have a higher risk of obesity,
which in turn raises the child’s risk of developing T2DM.
Recently, the occurrence of GD and T2DM was shown to correlate with plasma levels of
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 metabolite 2ꢀ(2ꢀcarboxyethyl)ꢀ4ꢀmethylꢀ5ꢀpropylfuranꢀ3ꢀcarboxylic acid (1, CMPF, Figure
3
1
). Independent studies also showed a strong correlation between elevated levels of CMPF and
beta cell dysfunction in women with GD, though others have found less or no correlation of
4
CMPF to impaired glucose tolerance in other human populations. CMPF was found at levels
3
around 150 ꢀM in plasma in both pregnant women with GD and T2DM patients. Additionally,
CMPF was found to dramatically increase in preꢀdiabetic patients compared to patients with
3
a
normal glucose tolerance (NGT). These observation have led to the proposal that CMPF may
have a causative role in GD and T2DM.
3
b
3a
3a
In the laboratory, treatment of lean, DIO, or ob/ob mice with CMPF leads to
impaired glucose clearance during intraperitoneal glucose tolerance test (ipGTT) experiments.
Mouse and human islets treated with CMPF show decreased glucose stimulated insulin secretion
3b
(
GSIS). Additionally, mice administered CMPF via intraperitoneal injection had higher
3
a
numbers of immature insulin granules compared to vehicle treated animals.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
OH
OH
O
O
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
OH
OH
O
O
2
O
OH
OH
O
O
O
3
O
OH
OH
O
O
OH OH
4
Figure 1. Urofuranoic Acids.
3b
CMPF enters cells via active transport by organic anion transporters.
When mouse
islets were treated with CMPF, reactive oxygen species (ROS) increased suggesting that CMPF
3
b
increases oxidative stress in the cell. In support of this hypothesis, cells treated with Nꢀacetyl
3
b
cysteine were protected from the effects of CMPF. Ex vivo study of islets from CMPF treated
mice shows they have reduced glycolysis and at the same time increased glucose uptake and
3a
glycated proteins. These ex vivo studies suggest that CMPF alters the islet’s energy production
pathways, diminishing cells’ ability to utilize glucose for energy and shifting energy generation
3a
to fatty acid oxidation.
ACS Paragon Plus Environment
4

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
CMPF is a urofuranoic acid. In humans, CMPF is produced by metabolism of Fꢀacids
6
(
Scheme 1). Fꢀacids are lipid metabolites produced in plants and believed to enter mammals
7
through the food chain. Notably, mammalian metabolism of Fꢀacids appears to differ across
species since rats and cows do not metabolize Fꢀacids through oxidation of the C3 methyl. Other
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
7
a, 8
urofuranoic acids have also been identified in plasma (ie. Figure 1, compounds 2 and 3).
CMPF has a very high binding affinity to plasma proteins. For this reason, CMPF has been
9
9a,
classified as a uremic toxin. Uremic toxins have been linked to chronic kidney disease (CKD).
9
b
CKD is a common coꢀmorbidity of diabetes.
O
OH
OH
OH
O
n
O
O
m
O
m
F-Acids
m = 2-4; n = 8-12
Urofuranoic acids
Scheme 1. Human Metabolism of Fꢀacids.
CMPF has been shown to have distinctly different reactivity to oxidants than the parent
8
, 10
Fꢀacids.
For instance, hypochlorite oxidation of CMPF produces dihydroxy furan 4 (Figure 1)
1
1
which, unlike the typical highly reactive oxidation products of Fꢀacids, can be isolated. CMPF
has also been shown to react with tertꢀbutyl peroxide to produce a radical species which can be
9
c
observed by electron paramagnetic resonance (EPR). It has been suggested that CMPF radicals
could promote reactions in human kidney cells leading to cell damage. Unfortunately, the
9
c
structure of these species was not able to be determined by EPR.
Given the data suggesting a potentially causative linkage of CMPF to diabetes, as well as
its longstanding connection to CKD, the details of the biology mediated by CMPF are of
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
significant interest. The exact mechanism(s) by which CMPF affects glucose metabolism is as of
yet unknown. Further, other structurally related urofuranoic acids have been identified, but their
potential linkage to diabetes or effects on insulin secretion have not been investigated. Our goal
was to use medicinal chemistry approaches to probe CMPF metabolism and explore the effect of
structural analogues on islets in order to identify potential tools to support future studies in
mechanistic biology. To that end, we report the synthesis of several CMPF analogues and have
profiled their effects on insulin secretion in mouse and human islets in vitro. These analogues
include a series of putative metabolites of CMPF along with other naturally occurring
urofuranoic acid derivatives. Finally, we also made isotopically labeled CMPF analogues to
explore the fate of CMPF in vivo. These studies have provided the first observation of a CMPF
metabolite in the plasma of a mammal.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chemistry
5
a, 12
CMPF has been synthesized previously.
Both approaches used a FriedelꢀCrafts
acylation to install the C5 propyl chain. To access CMPF and its analogues, we chose to exploit
13
5b
an alternate approach based on the work of Hanson et al which was used by Spiteller to
prepare the naturally occurring urofuranoic acid 2 (Figure 1). In this approach, a βꢀketoꢀester is
condensed with an acyloin under Lewis acid promoted conditions to give the tetraꢀsubstituted
furan. The requisite starting materials are easily prepared by a variety of means. For instance,
Meldrum’s acid 6 was acylated with commercially available succinyl chloride 5 and the resulting
product refluxed in methanol to provide βꢀketoꢀester 7 (Scheme 2). The necessary acyloins (11)
were obtained from commercial sources or were readily prepared by addition of lithiated ethyl
vinyl ether to the corresponding aldehyde followed by acid hydrolysis of the enol ether (scheme
2). Refluxing the acyloin (11a-h) with the appropriate βꢀketo ester (7, 12, or 13) in the presence
ACS Paragon Plus Environment
6

Page 7 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of zinc chloride in methanol provided the urofuranoic acid esters 14ꢀ23 in good yield (scheme 3).
Hydrolysis of both ester groups was then achieved using forcing conditions. Notably, using this
synthetic route we were able to easily make gram quantities of CMPF to support subsequent
chemistry and in vivo studies.
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 2.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
Compound
R
R
R
1
2
nPr
nPentyl
nPr
Me
Me
Me
Me
Me
Et
ꢀ(CH
ꢀ(CH
2
)
)
2
CO
CO
2
2
H
H
2
2
1
1
1
4
5
6
ꢀ(CH
ꢀ(CH
ꢀ(CH
ꢀ(CH
ꢀ(CH
ꢀ(CH
ꢀ(CH
2
2
2
2
2
2
2
)
)
)
)
)
)
)
2
2
2
2
2
2
2
CO
2
2
2
2
2
2
2
Me
Me
Me
Me
Me
Me
Me
nPentyl
Me
CO
CO
CO
CO
CO
CO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
Et
1
1
8
9
nPr
nPr
ꢀCH
(CH
)
CH
ꢀ
2
2
2
2
20
ꢀCH CH
2
2
OBn
Me
Me
Me
Me
Me
Et
2
2
1
2
nPr
nPr
nPr
Me
Et
ꢀ(CH
ꢀ(CH
nPr
2
)
3
OBn
)
OH
2
3
23
2
2
4
5
ꢀ(CH
ꢀ(CH
ꢀ(CH
2
2
2
)
)
)
2
2
2
CO
2
2
2
H
CO
CO
H
H
26
nPr
nPr
nPr
Me
Me
Me
2
2
7
8
ꢀ(CH
2
)
3
OH
CO
nPr
ꢀCH
CH
OBn
ꢀ(CH
)
H
2
2
2
2
2
29
nPr
Scheme 3.
Additional CMPF analogues were prepared using alternative synthetic approaches
(Schemes 4ꢀ8). Diketone analogue 31 was prepared from CMPF in three steps via the Weinreb
amide intermediate 30 (Scheme 4). CMPF was also decarboxylated using copper/quinolone to
give analogue 32 (Scheme 4).
ACS Paragon Plus Environment
8

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
O
O
OH
O
N
O
CH3
O
c
a, b
OH
N O
CH3
O
O
O
1
30
31
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
d
OH
O
3
2
Conditions: a: oxalyl chloride, CH Cl ; b: MeNOMe-HCl, Et N, ClCH CH Cl, 50% over two steps;
c: MeMgBr, THF, 0-45 °C, 50%; d: quinoline, copper, 180 °C, 3h, 45%.
2
2
3
2
2
Scheme 4.
Naturally occurring ketoꢀanalogue 3 was prepared according to the approach described
5
a
by Pfordt et al (Scheme 5). In our hands, FriedelꢀCrafts acylation of furan 33 proceeded best
when using the acid chloride rather than the anhydride. Hydrolysis of ester 34 provided
carboxylic acid 3.
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 5.
Furanoic acid 36 was prepared by exploiting the intrinsic chemoselectivty of
commercially available dibromide 35 (Scheme 6). The C5 propyl group was installed using a
Negishi reaction at room temperature. The resulting bromide was then reacted in a second
Negishi reaction at 100 °C. The ester product was then hydrolyzed to furnish acid 36.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 6.
Racemic tetrahydrofuran analogue 42 was prepared using a Lewis acid promoted [3+2]
addition reaction. Rhodium promoted cyclopropanantion of 1ꢀpentene with dimethylꢀ
14
diazomalnoate (37) gave diester 38. Tin (IV) chloride promoted [3+2] cycloaddition with
aldehyde 39 provided the tetrahydrofuran 40 as a 4:1 mixture of diastereomers. Decarboxylation
1
4
of diester 40 was accomplished using sodium cyanide under refluxing conditions to furnish
ester 41. Hydrolysis then gave tetrahdyrofuran diꢀacid 42.
ACS Paragon Plus Environment
10

Page 11 of 61
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
Scheme 7.
Oxidation products 43 and 4 were synthesized using sodium hypochlorite under buffered
1
1
conditions (Scheme 8). Notably, the oxidation products were stable to silica chromatography.
To guard against decomposition, the products were dissolved in DMSO and frozen until needed
for testing.
O
O
OR2 OR
OR
O
NaOCl
Acetone
pH = 7.4 Buffer
OR2
O
OH
O
O
HO
4
3: R = Me
1
4: R = Me
4
: R = H
1
: R = H
Scheme 8.
1
3
C labeled CMPF analogues were also prepared. We targeted three isotope labeling
patterns to allow suitable opportunity to observe CMPF metabolites in vivo. Analogue 46,
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
13
containing one C atom in the propyl chain was prepared from isotopically enriched propionic
acid (Scheme 9). FriedelꢀCrafts acylation of furan 33 followed by hydrolysis gave CMPF
analogue 45. Ketone reduction using triethylsilane promoted by BF ꢀOEt provided compound
3
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
6 in moderate yield.
O
a-c
OH
+
33
4
4
O
OH
d
13C
OH
O
O
O
45
O
OH
13C
OH
O
O
46
Conditions: a: oxalyl chloride, CH Cl ; b: AlCl ,
ClCH CH Cl, 42% over two steps; c: NaOH, MeOH,
2
2
3
2
2
MW 130 °C, 1 h, 90%; d: Et SiH, BF -OEt ,
3
3
2
CH Cl , reflux, 52%.
2
2
Scheme 9.
1
3
Double labeled analogue 50 was prepared from commercially available C succinic acid
4
7 (Scheme 10). Dehydrative cyclization using acetic anhydride followed by solvolysis gave the
succinate monoꢀester. Reaction with oxalyl chloride yielded the isotopically enriched acid
chloride which was further elaborated by the same approach used in Schemes 2 and 3 for the
synthesis of CMPF to provide analogue 50.
ACS Paragon Plus Environment
12

Page 13 of 61
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
Scheme 10.
13
The final isotopically labeled CMPF analogue 54 was prepared from C diethyl malonate
(51) (scheme 11). Conversion of the malonate to the mixed βꢀketoꢀester 52 was accomplished in
two steps. Ketoꢀester 52 was then subjected to our standard set of conditions to convert it to
labeled CMPF analogue 54.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 11.
Results and Discussion
Our investigations began with studies to explore the potential for metabolism of CMPF in
vivo. Plasma concentrations of CMPF in healthy humans range from 20ꢀ40 ꢀM. CMPF is
excreted as parent in urine in small amounts (~1ꢀ5 mg/day). No CMPF conjugates in urine
1
5
samples from healthy humans have been found. We wondered if CMPF might be metabolized
through other pathways. In particular, we were interested in seeing if CMPF was being processed
through lipid metabolism pathways or alternatively being oxidized in liver.
ACS Paragon Plus Environment
14

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To characterize the pharmacokinetics of CMPF, we conducted studies in mice (Figure 2).
Because CMPF is believed to be generated via metabolism of Fꢀacids in humans (vide supra), we
chose to use IV administration for PK studies rather than oral dosing. IV doses of CMPF
produced robust exposures with low volumes of distribution, consistent with a compound that is
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
highly bound to albumin (Table 1). The 3 mg/kg dose produced a C of ~130 ꢀM, similar to the
0
plasma levels observed in GD and T2DM patients. CMPF is cleared at a rate wellꢀbelow hepatic
blood flow. Based on these data, we opted to conduct our in vivo metabolism studies using 6
mg/kg doses since we anticipated this would provide a C of 200ꢀ250 ꢀM, replicating
0
concentrations we intended to use for in vitro work (vide infra).
Figure 2. Plasma Exposure of CMPF in Mice Following IV Administration.
Table 1. Pharmacokinetics of CMPF in Mice.
Dose (mg/kg)
AUC (nM*h)
89300
C0 (nM)
53300
T1/2 (h)
4.5
Cl (mL/min/kg)
0.66
Vd (L/kg)
0.2
1
3
210000
130000
3.27
0.85
0.2
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To explore the fate of CMPF in vivo, we chose to use isotopically labeled CMPF. We
decided to prepare three CMPF analogues with different label patterns so that we could have the
highest likelihood of not only detecting CMPF metabolites, but also determining where
modifications were being made. Because we wanted to prepare several analogues and our studies
were going to require samples to be shipped between several locations, use of nonꢀradioactive
labels was practical.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For these studies, CMPF was dosed at 6 mg/kg intraperitoneally for three days. On the
fourth and fifth days, labeled molecules 46, 50, and 54 were dosed as 1:1 mixtures of
labeled/unlabeled CMPF. On day five, the mice were euthanized 1 h postꢀdose and plasma,
adipose, muscle, and liver samples were collected. Samples were then analyzed by HPLCꢀMS.
Because 1:1 mixtures of labeled/unlabeled CMPF were dosed, any metabolites arising from
CMPF would produce distinctive isotope patterns in the MS allowing them to be readily
identified. It is important to note that CMPF has a specific fragmentation pattern in MS. The
most significant ions observed in negative ESMS of the labeled analogues 46, 50, and 54 are
shown in Table 2. These data show that CMPF undergoes preferential loss of the C2 carboxyl
during fragmentation.
Table 2. Significant Ions Observed in Negative ESMS.
Compound
Ions (m/z)
4
6
240, 196, 152
241, 197, 152
241, 196, 152
50
5
4
ACS Paragon Plus Environment
16

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In both plasma and tissue samples from treated mice, parent CMPF was the most
abundant material observed. However, in plasma samples of treated mice we were able to
identify low levels of a glucuronide conjugate of CMPF. MS/MS spectra for the glucuronide
product ions derived from the labeled materials are shown in Figure 3. Distinctive peaks at
416/417 m/z from the parent conjugate along with fragmentation peaks arising from CMPF
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
(240/241and 196/197 m/z) and glucuronic acid (193 and 175 m/z) support the assignment of the
material as the glucuronide conjugate. No other peaks were observed between 416/417 and
240/241 m/z. Based on the fragmentation pattern, it is not possible to assign which carboxyl
group is conjugated to glucuronic acid. To our knowledge, this is the first observation of a
metabolite derived from CMPF. Analysis of human samples to determine if the CMPFꢀ
glucuronide is formed are planned and will be reported as appropriate.
Interestingly, no metabolites arising from incorporation of CMPF into lipid biosynthetic
pathways (ie, conjugates to glycerol, monoꢀacylglycerol, or diacylglycerol species) were
observed in plasma, muscle, liver or adipose tissues. The study was conducted after several days
of dosing to ensure that any metabolic effects caused by refeeding would be observed. In our
analysis of the samples, we also did not detect any metabolites derived from oxidative
metabolism of CMPF in any compartment. While no other metabolic products were observed
besides the CMPFꢀglucuronide, we cannot exclude that such species are formed given the limit
of detection using nonꢀradiolabeled materials.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. MS/MS Spectra of Plasma Glucuronide Molecular Ion. No ions were observed in the
regions of the spectra between 250 and 400 m/z. A: Glucuronide derived from compound 50. B:
Glucuronide derived from compound 54. C: Glucuronide derived from compound 46.
Having explored the fate of CMPF in vivo, we turned our attention to studying the in
vitro pharmacology of CMPF. To investigate inhibition of insulin secretion from islets, we
prepared several analogues (vide supra). These included other known urofuranoic acids 2 and 3.
We also prepared compounds that were designed to allow deconstruction of CMPF (ie
ACS Paragon Plus Environment
18

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds 27, 29, 31, and 36). We also prepared a series of compounds to explore structural
modification at each of the four carbons of the furan.
To assess the importance of the furan moiety, we designed tetrahydrofuran 42 as a
saturated analogue of CMPF. Molecular modeling of tetrahydrofuran 42 shows that in the low
energy conformation all three ring substituents are in pseudoꢀequatorial orientations, similar to
CMPF (Figure 4).
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. Overlay of Furan 42 and CMPF. Overlay was generated using MOE. Molecules were
minimized in a neutral state in the gas phase using a AMBER10 forcefield. Cyan = CMPF.
Orange = Compound 42.
With a rich set of analogues in hand, we set out to explore the in vitro properties of our
molecules. Our initial profiling approach involved carrying out an assessment of intrinsic
cytotoxicity for all of the compounds (Table 3). Here, we chose to use the fold change in caspase
3 and caspase 7 as a measure of intrinsic toxicity. We used two cell lines for our assessment,
HEK293 cells and mouse primary hepatocytes. Compounds showing ≥ 2 fold change in caspase
3/7 were considered cytotoxic and were deprioritized for further evaluation. Notably, CMPF (1)
showed no change relative to DMSO in either cell line. None of the compounds showed
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cytotoxicity against HEK cells. In mouse hepatocytes, compounds 24, 36, and 42 showed
cytotoxicity. Thus we deprioritized these compounds for further followꢀup. The difference in
1
6
toxicity observed could be due to lower levels of expression of transporters in HEK cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3. Cytotoxicity Assay.
Fold Change in Caspase 3/7 Relative to DMSO
Mouse Primary
HEK Cells
Compound
Hepatocytes
Run 1
Run 2
Run 3
Run 1
Run 2
1
2
3
4
0.8
1.0
1.1
1.1
1.0
1.0
0.9
1.0
1.1
1.0
0.8
1.0
1.0
1.0
1.1
1.0
0.9
1.0
1.0
1.0
1.0
1.1
1.0
1.0
1.0
1.2
1.1
1.0
1.1
1.0
1.1
1.0
1.0
1.0
1.0
1.1
2.0
0.9
0.9
1.1
1.2
1.1
0.9
1.0
1.0
1.0
1.1
1.1
0.9
1.1
1.1
1.0
0.9
1.0
1.0
1.0
NT
1.5
0.9
0.9
NT
NT
NT
0.8
1.5
1.1
0.9
1.2
0.8
1.0
0.9
0.4
0.6
2.0
1.1
1.8
0.6
1.4
1.4
0.3
0.7
1.5
2.2
1.3
0.7
1.2
1.1
0.9
1.0
0.8
0.9
0.8
0.4
0.6
2.0
1.1
1.6
0.6
NT
1.4
0.3
0.7
NT
NT
NT
1
5
16
1
1
1
2
7
8
9
2
24
2
2
2
2
5
6
7
8
29
3
3
3
3
1
2
4
6
41
ACS Paragon Plus Environment
20

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
4
2
3
1.0
1.0
1.2
0.9
NT
2.0
0.3
NT
0.9
0.3
NT = Not Tested
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
We next profiled CMPF and a selection of structurally diverse analogues in in vitro
assays assessing their microsomal stability to and ability to inhibit cytochrome P450 enzymes
(
Table 4). CMPF itself did not show detectable levels of degradation in the presence of human or
mouse microsomes. Our other analogues had good to modest stability in microsomal
preparations, except for primary alcohol 22. CMPF had no detectable inhibition of CYP3A4,
CYP2D6, or CYP2C9 at 10 ꢀM. Like CMPF, most of the analogues we evaluated did not show
significant inhibition against the CYP enzymes.
Table 4. In vitro Metabolism Assessment of CMPF and Select Analogues.
Human
Micorosmal
Metabolism (%)
Mouse
Micorosmal
Metabolism (%)
Human CYP3A4
Inhibition (%)
Human CYP2D6
Inhibition (%)
Human CYP2C9
Inhibition (%)
Compound
CMPF (1)
ND
5
ND
7
0
3
0
3
0
9
2
3
ND
49
0
ND
60
16
76
100
21
7
3
11
28
12
20
0
10
27
3
1
1
1
2
5
6
9
2
59
22
48
0
16
100
0
2
0
25
0
3
17
0
2
2
6
7
3
0
0
10
21
ND
0
9
2
2
3
29
51
ND
20
0
13
21
3
5
18
0
3
3
1
2
0
0
9
36
20
0
0
0
ND = not determined.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Compounds were next screened in a mouse islet insulin secretion assay under high
glucose conditions to identify which analogues might have effects on islets (Figure 5). Because
3b
previous studies have shown CMPF impairs GSIS, our initial islet assays were performed in a
screening mode using only high glucose conditions (see experimental section for details). The
purpose of this assay was to help identify compounds with activity in islets that could be selected
for further study. Data are shown as fold change in insulin secretion relative to vehicle treatment
for an n=4 experiments. Consistent with previous results, CMPF treated islets showed
suppressed insulin secretion. The structurally similar metabolite 2 also showed suppression of
insulin secretion (Figure 5). Additionally, analogue 26, in which the C4 substituent has been
change to nPr, was also shown to suppress insulin secretion. The two esters 17 and 19 also
suppressed insulin secretion. Interestingly, the diacid analogue of 17, compound 25, did not
show a meaningful impact on insulin secretion. One possible explanation s is that compound 25
could be inactive in the islet assay due to an inability to penetrate islets. In partial support of this
hypothesis, we were unable to measure any passive permeability for compound 25 using a cell
permeability assay in MDCK cells (data not shown).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
22

Page 23 of 61
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 5. Insulin Secretion Screening Assay in Mouse Islets. Compounds were tested at 200 ꢀM.
The glucose concentration was 20 mM. Yꢀaxis shows fold change relative to vehicle control
(dashed line). Values are an average of n = 4 experiments. Error bars indicate standard error of
the mean. * = p ≤ 0.05 by the student’s tꢀtest.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Structures of Compounds Showing Effects in the Mouse Islet Screening Assay.
To our surprise, compounds 27, 43, and 4 all showed increased insulin secretion in mouse
islets under these assay conditions (Figures 5 and 6). Compounds 27 is a close analogue to
CMPF where the C2 carboxyl chain has been reduced to an alcohol. Compounds 43 and 4 are the
oxidized CMPF analogues.
Based on these results, we chose compounds 4 and 17, which had different responses in
the mouse islet screening assay, for more thorough evaluation. Accordingly, we conducted GSIS
studies under high and low glucose conditions and KCl stimulated insulin secretion assays in
mouse islets (Figure 7). We also measured the intracellular insulin content in islets treated with
these compounds. As shown in Figure 7, CMPF (1) displayed suppressed GSIS and KCl
ACS Paragon Plus Environment
24

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stimulated insulin secretion but did not change the intracellular insulin content. This is
consistent with previous observations from the Wheeler lab which have shown that CMPF
treatment of mice increases the amount of immature insulin granules in islets as well as the
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
3a
amount of proꢀinsulin but had no effect on total insulin content.
In contrast to what was observed in the screening assay where islets were exposed only
to high glucose media (Figure 5), in these assays compound 4 treatment had no significant effect
on GSIS, KCl stimulated insulin secretion, or insulin content. Accordingly, we do not know if
compounds 27 and 43 will increase insulin secretion in the standard GSIS assay.
Compound 17 suppressed GSIS and lowered insulin content in this assay. The
observation that compound 17 lowered total insulin content but that CMPF did not suggests the
two compounds may act via different mechanisms to reduce GSIS. While the current body of
work cannot provide a definitive explanation for this difference, we hypothesize that some of the
mechanistic differences may arise from differences in compound reactivity in cells. Like
9
c
CMPF, compound 17 could induce formation of ROS but might do so either to a larger extent
or within different microenvironments/compartments of the cell, including the endoplasmic
reticulum (ER). Indeed, oxidative and ER stress are have a wide range of effects on protein
17
synthesis and trafficking which are linked to a host of human diseases, including diabetes.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
B
C
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
*
*
Figure 7. Effects of Compounds 1, 4, and 17 in Mouse Islets. All compounds were tested at 200
ꢀM. Error bars indicate standard error of the mean. A: Glucose Stimulated Insulin Secretion
assay. Values are the average of a n = 4ꢀ5 experiments. * = p <0.05 by the student’s tꢀtest; B:
ACS Paragon Plus Environment
26

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Potassium Chloride Stimulated Insulin Secretion. The KCl concentration for this assay was 30
mM. Values are the average of n = 3ꢀ4 experiments; C: Intracellular insulin content in mouse
islets with a screen assay at 20 mM glucose. Values are the average of n = 3ꢀ4 experiments. * = p
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
<
0.05 by the student’s tꢀtest.
To assess the relevance of our observations in mouse islets with human biology, we
tested the compounds from the mouse islet screen in a similar assay using human islets under
high glucose conditions. Islets for these studies were obtained from two donors. Because islets
are obtained from human donors postꢀmortem, the preparations often have greater heterogeneity
in overall health and function than those obtained from healthy lab mice (ie. Figure 5). This is
exemplified by the data generated with compound 16 where donor islets gave antipodal
responses. Thus, in evaluating the limited data set here, one must use caution in drawing
conclusions. Generally, compounds showed a higher propensity to suppress secretion in human
islets than in mouse islets. As observed in mouse islets, CMPF, compounds 2 and 19 showed
suppression of insulin secretion in islets from both donors. Notably, compounds 43, 4, and 27
failed to show increased insulin secretion in human islets, in contrast to the effects observed in
mouse islets. Interestingly, compound 15 enhanced insulin secretion in human islets but had no
effect on insulin secretion in mouse islets. The apparent discrepancy in experiments using mouse
versus human islets with compounds such as 43 and 15 may be due to species differences in islet
biology. It is well established that human and rodent islets can have different biological
18
responses.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 61
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Insulin Secretion in Human Islets. Compounds were tested at 200 ꢀM. The glucose
concentration was 20 mM. Y axis shows fold change relative to vehicle control (dashed line).
Conclusion
Based on the data we have gathered, CMPF does undergo phase 2 metabolism in vivo.
The observation of a glucuronide in the plasma of mice suggests additional analysis of human
samples looking for this and other potential metabolites is warranted.
ACS Paragon Plus Environment
28

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The inhibitory effect of CMPF on insulin secretion in mouse islets is a function of
molecular structure (Figure 5). Reduction of the C2 propyl chain carboxylate (ie compounds
1
→27→29) significantly changes performance in the mouse islet insulin secretion screening
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
assay. The C3 carboxyl is also important for insulin secretion activity as shown by compound 32.
The inactivity of diketone 31 in the mouse or human islets also suggests that the presence of the
negative charge in these regions is important. Small increases in size of the C4 and C5
substituents did not appear to meaningfully impact insulin secretion (ie. compound 2). However,
it is interesting that incorporation of a carbonyl at C5 (compound 3) resulted in a loss of insulin
secretion activity in the mouse islet screening assay.
Under standard assay conditions, compound 17 showed suppressed GSIS and insulin
content (Figure 7). Compound 4, which showed increased insulin secretion in the screening
assay (figure 5), did not show any statistically significant effects under these conditions (Figure
7). A possible explanation for this difference is the fact that islets under screening conditions
were exposed only to high glucose media.
The apparent connection between CMPF and GD as well as T2DM suggest that this area
of human biology is worthy of additional study. Understanding the mechanism by which CMPF
may facilitate the onset of GD and T2DM may help to identify alternative therapeutic approaches
for patients. The studies described here suggest that the in vivo metabolism of CMPF in humans
should be investigated further, as metabolites with potential biological activity might be
produced.
Experimental Methods
ACS Paragon Plus Environment
29