Combination of l-Carnitine with Lipophilic Linkage-Donating Gemcitabine Derivatives as Intestinal Novel Organic Cation Transporter 2-Targeting Oral Prodrugs

Article abstract of DOI:10.1021/acs.jmedchem.7b00049

Novel organic cation transporter 2 (OCTN2, SLC22A5) is responsible for the uptake of carnitine through the intestine and, therefore, might be a promising molecular target for designing oral prodrugs. Poor permeability and rapid metabolism have greatly restricted the oral absorption of gemcitabine. We here describe the design of intestinal OCTN2-targeting prodrugs of gemcitabine by covalent coupling of l-carnitine to its N4-amino group via different lipophilic linkages. Because of the high OCTN2 affinity, the hexane diacid-linked prodrug demonstrated significantly improved stability (3-fold), cellular permeability (15-fold), and oral bioavailability (5-fold), while causing no toxicity as compared to gemcitabine. In addition, OCTN2-targeting prodrugs can simultaneously improve the permeability, solubility, and metabolic stability of gemcitabine. In summary, we present the first evidence that OCTN2 can act as a new molecular target for oral prodrug delivery and, importantly, the linkage carbon chain length is a key factor in modifying the affinity of the substrate for OCTN2.

Full text of DOI:10.1021/acs.jmedchem.7b00049

Subscriber access provided by University of Newcastle, Australia
Article
Combination of L-carnitine with lipophilic linkage-
donating gemcitabine derivatives as intestinal novel
organic cation transporter 2-targeting oral prodrugs
Gang Wang, Hong-xiang Chen, Dong-yang Zhao, Da-wei Ding, mengchi Sun, Long-fa Kou, Cong
Luo, Dong Zhang, Xiu-lin Yi, Jinhua Dong, Jian Wang, Xiaohong Liu, Zhonggui He, and Jin Sun
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00049 • Publication Date (Web): 24 Feb 2017
Downloaded from http://pubs.acs.org on February 25, 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 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Combination of Lꢀcarnitine with lipophilic linkageꢀdonating
gemcitabine derivatives as intestinal novel organic cation
transporter 2ꢀtargeting oral prodrugs
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
Gang Wanga, Hongxiang Chena, Dongyang Zhaoa, Dawei Dinga, Mengchi Suna,
Longfa Koua, Cong Luoa, Dong Zhanga, Xiulin Yic, Jinhua Donga, Jian Wangd,
Xiaohong Liua, Zhonggui Hea and Jin Sun*,a,b
a School of Pharmacy, Shenyang Pharmaceutical University, Wenhua Road,
Shenyang 110016, China
b Municipal Key Laboratory of Biopharmaceutics, School of Pharmacy, Shenyang
Pharmaceutical University, China
c State Key Laboratory of Drug Delivery Technology and Pharmacokinetics,
Tianjin Institute of Pharmaceutical Research, Tianjin, China
d Key Laboratory of StructureꢀBased Drug Design and Discovery, Shenyang
Pharmaceutical University, Ministry of Education, China
KEYWORDS: gemcitabine; new organic cation transporter 2 (OCTN2); prodrug;
bioavailability.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
5
ABSTRACT
6
7
8
9
The novel organic cation transporter 2 (OCTN2, SLC22A5) is responsible for the
uptake of carnitine through the intestine and, therefore, it may be a promising
molecular target for designing oral prodrugs. Poor permeability and rapid metabolism
have greatly restricted the oral absorption of gemcitabine. We here describe the design
of intestinal OCTN2ꢀtargeting prodrugs of gemcitabine by covalent coupling of
Lꢀcarnitine to its N4ꢀamino group via different lipophilic linkages. Due to high
OCTN2 affinity, the hexane diacidꢀlinked prodrug demonstrated significantly
improved stability (3ꢀfold), cellular permeability (15ꢀfold), and then increased oral
bioavailability (5ꢀfold), while causing no toxicity as compared to gemcitabine. In
addition, OCTN2ꢀtargeting prodrugs can simultaneously improve the permeability,
solubility and metabolic stability of gemcitabine. In summary, we present the first
evidence that OCTN2 can act as a new molecular target for oral prodrug delivery and,
importantly the linkage carbon chain length is a key factor in modifying the substrate
affinity toward OCTN2.
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
ACS Paragon Plus Environment

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
Gemcitabine is a nucleoside analog that has been used as the firstꢀline
chemotherapy for pancreatic and nonꢀsmallꢀcell lung cancers1. Recently, a number of
gemcitabineꢀbased therapies in combination with cytotoxins or molecularly targeted
agents began to be evaluated in clinical trials for the treatment of many cancers2. The
antitumor activity of gemcitabine is elicited by inhibition of ribonucleotide reductase
and DNA synthesis3. Despite therapeutic effectiveness, the administration of modality
gemcitabine is currently limited to the intravenous route, which leads to poor
compliance and a number of safety issues4. In addition, rapid kidney excretion and
extensive deamination by cytidine deaminase (CD) in blood and liver result in its
short halfꢀlive and need for frequent administration. Consequently, there is an urgent
need to develop highꢀefficacy, lowꢀtoxicity oral gemcitabine products. However, so
far, there are no oral preparations commercially available for gemcitabine, due to its
poor membrane permeability and low metabolic stability.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Recently, a lipophilized prodrug strategy has been used to improve metabolic
stability and membrane permeability5. The valproate amide prodrug of gemcitabine
(Figure 1b) exhibited excellent stability in hepatic and intestinal homogenates and
reduced deaminated metabolites (2’, 2’ꢀdifluoroꢀ2’deoxyuridine, dFdU) by 50 %, in
comparison with gemcitabine. Moreover, the systemic exposure of gemcitabine (AUC)
increased 1.4ꢀfold after oral administration. The general advantages of this method
include altered transport across cell membranes via passive diffusion, and a lower
susceptibility to inactivation by CD. However, the poor aqueous solubility generally
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
7
8
9
limits its oral absorption and clinical application. Accordingly, carrier prodrug
strategy is more recommended due to the good aqueous solubility and improved
membrane permeability mediated by transporters. A series of amino acid ester
prodrugs of gemcitabine were synthesized to target intestinal peptide transporter 1
(PepT1)6, 7. The Lꢀvalyl (Figure 1c) and Dꢀphenylalanyl (Figure 1d) prodrugs
exhibited increased Cacoꢀ2 cells permeability by 4.0ꢀfold and 4.5ꢀfold, respectively.
However, no in vivo oral pharmacokinetic results for these prodrugs have been
reported, probably due to the fact that the rapid hydrolysis to the parent drug in the
gastrointestinal (GI) tract strongly limits in vivo efficacy. To obtain more stable
PepT1ꢀtargeting prodrugs of gemcitabine, a number of aminoacyl amide derivatives
of gemcitabine were synthesized and evaluated (Figure 1i)8. However, rapid
rearrangement to the corresponding carboxamide occurred in aqueous solution at near
neutral pH, resulting in the active parent drug being unable to be released from the
prodrugs (Figure 1j). Hence, it is essential to simultaneously optimize membrane
permeability, stability, and solubility for oral prodrugs of gemcitabine.
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 novel organic cation transporter 2 (OCTN2, SLC22A5) is an important
nutrient transporter to transport carnitine in small intestine, which serves as a shuttle
for the transport of longꢀchain fatty acids into mitochondria. OCTN2 is an attractive
target for the transport of carnitine analogues9. Two excellent examples of
OCTN2ꢀtargeting prodrugs are prednisolone10 and nipecotic11. Following conjugation
to Lꢀcarnitine, these prodrugs are recognized by OCTN2 and exhibit increased
delivery to human bronchial epithelial cells and brain following intraꢀtracheal
4
ACS Paragon Plus Environment

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
instillation and intraꢀperitoneal administration, respectively. More importantly,
Lꢀcarnitine prodrugs significantly increase efficacy in the treatment of asthma and
tonic convulsions. OCTN2 is highly expressed throughout small intestine segments,
and thus represents a good target for the design of oral prodrugs12. To the best of our
knowledge, no reports have been published in terms of OCTN2ꢀtargeting oral prodrug
strategy, involving gemcitabine.
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 present study mainly aimed to evaluate the usefulness of OCTN2ꢀtargeting
oral prodrug strategy, and investigate the effect of the fatty acid linkage chain length
(n=4ꢀ10) on the substrate affinity toward OCTN2. Four OCTN2ꢀtargeting oral
prodrugs of gemcitabine, differing in the carbon chain length of the fatty acid linkage,
and a nonꢀtargeting prodrug were designed and synthesized. Lꢀcarnitine was
conjugated with the N4ꢀamino group of gemcitabine since the amide bond is stable to
chemical and enzymatic hydrolysis. Also, importantly, GI tractꢀstable prodrugs will
maximize the utility of OCTN2 activity and reduce the cytotoxicity of gemcitabine
exposure to the enterocytes. As expected, gemcitabine OCTN2ꢀtargeting prodrugs
showed excellent stability in the GI tract and high permeability. Interestingly, the
hexane diacid linkage prodrug showed better OCTN2 affinity and oral bioavailability
than the other targeted prodrugs with different linkage.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
RESULTS AND DISCUSSION
6
7
Design and Synthesis OCTN2ꢀtargeting oral prodrugs
8
9
Firstly, four OCTN2ꢀtargeting oral prodrugs, Lꢀcarnitineꢀsuccinicꢀgemcitabine
(GSC), Lꢀcarnitineꢀhexylicꢀgemcitabine (GHC), Lꢀcarnitineꢀoctanedioicꢀgemcitabine
(GOC) and Lꢀcarnitineꢀdecanedioicꢀgemcitabine (GDC) were prepared using the
synthetic approach shown in Figure 2A. To evaluate the uptake mechanism of the
targeted prodrugs, palmitoylꢀLꢀcarnitine (PC), a carnitine analogue but not substrate
of OCTN2, was synthesized using the synthetic approach of the targeted prodrugs.
Additionally, one nonꢀtargeting lipophilic amide prodrug, succinicꢀgemcitabine (GS),
was synthesized as a negative control and was shown in Figure 2B. Lꢀcarnitine was
attached to the N4ꢀamino group of gemcitabine through fatty acid with different chain
length. The 3’ꢀhydroxyl group of Lꢀcarnitine was chosen as a conjugation site since it
is a nonꢀessential group for the carnitine interaction with OCTN213. All the prodrugs
were fully characterized by HPLC, MS and NMR. Detailed synthetic procedures,
yields, and HPLC purity could be found in the Experimental section and Supporting
information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Stability Studies
To target intestinal transporter, the prodrugs should be stable in the GI tract after
oral administration. The chemical stability was examined in different pH phosphate
buffers and the halfꢀlives (t1/2) were obtained from linear regression of
pseudoꢀfirstꢀorder kinetics. As reflected in Table1, GSC, GHC and GS underwent
slow hydrolysis at pH 1.2 and were more stable in pH 6.8ꢀ7.4 buffers. The enzymatic
6
ACS Paragon Plus Environment

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stability was studied in tissue homogenates and plasma. The t1/2 values of those
prodrugs in intestinal homogenate were large, indicating that the prodrugs could be
delivered across intestinal epithelial cells in their intact form, which is essential to
reduce the cytotoxicity to enterocytes for anticancer oral prodrugs. The t1/2 values of
two OCTN2ꢀtargeting prodrugs in plasma were roughly 4ꢀfold shorter compared with
that in hepatic homogenate (Figure 3), indicating that the primary bioactivation site
was the plasma and then the liver.
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
An ideal prodrug should exhibit excellent stability in the GI tract before
interaction with OCTN2 and then must be converted to the parent drug in the
systematic circulation. In general, the ester bond is not very stable to enzymatic
hydrolysis14. For example, 5’ꢀOꢀLꢀphenylalanylꢀdecitabine is rapidly hydrolyzed in
intestinal fluid, with a very short halfꢀlife (95 min)15. In this study, OCTN2ꢀtargeting
gemcitabine prodrugs showed excellent enzymatic stability, thereby reducing
intestinal degradation to gemcitabine and increasing the targeting delivery, very
consistent well with the previous study of the valproate amide prodrug of
gemcitabine4.
Contribution of Nucleoside Transporterꢀ1 (hENT1) to the Oral Absorption
To investigate the possible contribution of hENT1 to the oral absorption of the
targeted prodrugs (GSC and GHC), a competitive inhibitor dipyridamole was used to
study the cellular uptake of prodrugs in Cacoꢀ2 cells, which are known to express
hENT116. Figure S7 showed that 2.5 mM dipyridamole did not affect the cellular
uptake of GSC and GHC, confirming that these prodrugs entered cells in a nucleoside
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 48
1
2
3
4
transporterꢀindependent manner.
5
6
7
Cellular uptake mechanism of OCTN2ꢀtargeting oral prodrugs
8
9
The uptake of GSC and GHC by Cacoꢀ2 cells (naturally express OCTN2) was
reduced by 2ꢀ to 10ꢀfold where Na+ in medium was replaced with Nꢀmethylꢀ
Dꢀglucamine (Figure 4C), whereas no effect of Na+ on GS and gemcitabine was
observed (Figure 4D). The results suggested that the cellular uptake of the targeted
prodrugs was in a Na+ꢀdependent manner. Moreover, the uptake of GHC and GSC
was significantly higher at 37 °C than 4 °C, while GS and gemcitabine were not
affected, indicating the involvement of transporterꢀmediated transport (Figure 4E).
To confirm the contribution of OCTN2 on the transport of the targeted prodrugs,
OCTN2 gene transfected HEK293 (HEK293ꢀOCTN2) cells were used. As expected,
the uptake of GSC and GHC was, respectively, 3ꢀ to 8ꢀfold higher in
HEK293/OCTN2 than HEK293 cells, but there was no significant difference in
gemcitabine between the two cells (Figure 5), indicating that the targeted prodrugs
was specifically taken up by the cells via interaction with OCTN2. To further confirm
that the active transport was mediated by OCTN2, the competitive experiments
between the targeted prodrugs and Lꢀcarnitine were investigated. As shown in Figure
4F, the uptake of Lꢀcarnitine was decreased by 42% to 70% in the presence of 100 ꢁM
the targeted prodrugs. Moreover, the cellular uptake of the targeted prodrugs was
significantly reduced (by 2.2ꢀ3.7 fold) in the presence of 100 ꢁM Lꢀcarnitine, a
typical substrate of OCTN2 (Figure 4A), but there was no effect for GS and
gemcitabine (Figure 4B). However, the uptake amounts of the targeted prodrugs were
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
8
ACS Paragon Plus Environment

Page 9 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
not affected by the presence of palmitoylꢀLꢀcarnitine, a carnitine analogue but not
substrate of OCTN2 (Figure 4A). Therefore, these data suggested that the targeted
prodrugs were taken up into Cacoꢀ2 cells via OCTN2 protein.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In addition, GHC (6ꢀcarbon length linker) exhibited 1.8ꢀ to 2.5ꢀfold better uptake
in Cacoꢀ2 cells than GSC, GOC and GDC (4, 8 and 10ꢀcarbon length linker). Kinetic
analysis showed that Km value of GHC was roughly 2.2 ꢀ 3.9 folds lower than other
targeted prodrugs, but their Vmax values were similar (0.2013 nmol/mg/10 minꢀ0.4344
nmol/mg/10 min), suggesting the increased cellular uptake of GHC was due to an
increase in its binding affinity (Figure 6). These results indicated that the fatty acid
chain length has a determining effect on the affinity of the substrate toward OCTN2.
Computational docking studies
The crystal structure of OCTN2 protein has not been resolved, so we chose a
computational model of hOCTN2. Homology model of OCTN2 was constructed
similar to that described previously for OCT117. The sequence of human OCTN2 was
obtained from the Uniprot database (entry O76082; http://www.uniprot.org), and
GlpT (PDB entry: 1PW4) as template18. The carnitine (S467) binding site in the
computational model of hOCTN2 was used to investigate interactions between
transporter and prodrugs. From a thermodynamic view, a negative free energy (ꢂG<0)
indicates a favorable interaction system. The calculated binding energy values were
ꢀ3.37 KJ/mol for GHC and ꢀ2.23 KJ/mol for GSC, demonstrating that GHC had a
stronger interaction with OCTN2 than GSC. Moreover, due to long chain lipophilic
linkage, the oxygen atom on the pyridine ring of GHC formed Hꢀbond with Gly206
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
5
6
7
8
9
residue, which significantly contributes to stabilize the conformations of
proteinꢀprodrugs complex (Figure 7). The trend in the binding affinities was
consistent with the above cellular experiments.
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
Cacoꢀ2 Monolayer Permeability
The apparent transcellular permeability of gemcitabine and its prodrugs across
Cacoꢀ2 monolayer is shown in Figure 8. The permeability of the targeted prodrugs
was significantly higher in the AP→BL direction in comparison to the BL→AP
direction, whereas there was no significant difference for gemcitabine and GS,
indicating active transport of the prodrugs. All the targeted prodrugs increased
permeability of AP→BL direction by 5ꢀfold to 15ꢀfold, compared with gemcitabine,
and the permeability of GS was comparable with that of gemcitabine. Given the low
cellular uptake of GS in Figure 8D, poor permeability may be due to high retention in
the lipid bilayer of the plasma membrane. To confirm the contribution of OCTN2 to
the permeability of the prodrugs, a typical substrate Lꢀcarnitine was coꢀincubated with
the targeted prodrugs. The permeability of the targeted prodrugs across Cacoꢀ2
monolayer was reduced by 1.6ꢀ to 3.6ꢀfold in the presence of 100 ꢁM Lꢀcarnitine and
was not affect affected by the presence of 100 ꢁM palmitoylꢀLꢀcarnitine, indicating
that the improved permeability was mediated by OCTN2. Compared with
5’ꢀDꢀPhenylalanylꢀgemcitabine in a previous study6, OCTN2ꢀtargeting oral prodrugs
exhibited higher permeability (15ꢀfold VS 4.5ꢀfold), indicating that the transport
ability of OCTN2 was better than that of PepT1. More interestingly, GHC (6ꢀcarbon
length linker) exhibited a 1.7 fold increase in Papp value compared to GSC (4ꢀcarbon
10
ACS Paragon Plus Environment

Page 11 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
length linker). But, GOC and GDC (8 and 10ꢀcarbon length linker) exhibited a 1.7ꢀ
and 3ꢀfold lower Papp value compared to GHC. Thus, it seems that the length of linker
play a crucial role in the performance of permeability. GHC exhibited a significantly
higher permeability than other targeted prodrugs, probably due to optimal chain
length of the linkage that facilitates an increased interaction of GHC with the binding
domain of OCTN2 transporter protein.
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
Correlation of Lipophilicity and Permeability
To explore the correlation between the membrane permeability and lipophilicity,
the octanolꢀwater distribution coefficients (log P) were determined at pH 7.4 and
37 °C using the nꢀoctanol/water shake flask method (Table S1). GS had the highest
log P value and gemcitabine had the lowest. However, the lipophilicity was not
correlated with the high Cacoꢀ2 membrane permeability of the prodrugs (Figure 9),
indicating that, rather than passive diffusion, another transcellular mechanism is
responsible for the high permeation rates of OCTN2ꢀtargeting prodrugs.
In situ singleꢀpass perfusion
The effective permeability (Peff) and the absorption rate (Ka) of gemcitabine and
the prodrugs were determined using the in situ singleꢀpass intestinal perfusion system
in rats19. The Ka and Peff values of GSC and GHC were generally higher than those of
gemcitabine in whole intestinal segments (Figure 10 and Table 1). The Peff and Ka
values of GS were significantly lower than those of GSC and GHC, suggesting that
the contribution of passive transport for GSC and GHC was low. The Peff values of
GSC and GHC were reduced 25ꢀ and 11ꢀfold in the presence of Lꢀcarnitine,
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 48
1
2
3
4
5
6
7
8
9
respectively (Figure S8), agreeing well with the results of Cacoꢀ2 cellular
permeability study. In addition, only trace amounts of gemcitabine were observed in
the perfusate, suggesting that OCTN2ꢀtargeting oral prodrugs are stable within the GI
tract.
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
Cytotoxicity
In vitro cytotoxicity of all the prodrugs and gemcitabine was evaluated on
BxPCꢀ3 cells and Cacoꢀ2 cells with the 3ꢀ[4, 5ꢀdimethylthiazolꢀ2ꢀyl]ꢀ2,
5ꢀdiphenyltetrazolium bromide (MTT) assay kit. The half inhibitory concentration
(IC50) values were calculated and listed in Table S2. As reflected in Figure S9A, the
OCTN2ꢀtargeting prodrugs caused significant cytotoxicity in BxPCꢀ3 cells in a
doseꢀdependent manner after 48 h incubation. Compare to gemcitabine, the targeted
prodrugs showed equivalent cytotoxicity, indicating that the amideꢀtype prodrugs can
rapidly release active gemcitabine molecules to produce cytotoxicity. To examine the
possible involvement of OCTN2, we next compared the in vitro cytotoxicity of GHC
between the presence and the absence of 100 ꢁM Lꢀcarnitine in Cacoꢀ2 cells. As
shown in Figure S9B, GHC was less effective in killing cancer cells in the presence of
Lꢀcarnitine and the IC50 value of GHC was significantly increased by 8 fold, but there
was no effect for gemcitabine. This finding showed that the cytotoxic of targeted
prodrug was dependent on the effective uptake into tumor cells via OCTN2. In
addition, the in vitro cytotoxicity results of GHC showed that much lower cytotoxicity
in Cacoꢀ2 cells than BxPCꢀ3 cells (IC50 40.9 VS 0.32 ꢁM), probably since Cacoꢀ2
cells were usually resistant to chemotherapy agents20 and gemcitabine had weaker
12
ACS Paragon Plus Environment

Page 13 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
activity in colon cancer cells. These results also indicated that there was no toxicity of
GHC in intestinal epithelial cells.
7
8
9
Oral Pharmacokinetic Evaluation
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 plasma concentrationꢀtime profiles of gemcitabine, a nonꢀtargeting prodrug
(GS) and OCTN2ꢀtargeting prodrugs (GSC, GHC, GOC and GDC) were compared
after oral administration to rats (an equivalent gemcitabine dose of 50 mg/kg). The
plasma concentration of prodrugs and gemcitabine were simultaneously determined
using UPLCꢀMS/MS method. Figure 11 summarized the overall concentrationꢀtime
curves of all the drugs, and table 2 listed the pharmacokinetic parameters. The
detailed pharmacokinetic profiles of the prodrugs and gemcitabine are shown in
Figure S10ꢀS15. All the targeted prodrugs were rapidly absorbed after oral
administration and the absolute oral bioavailability ranged from 4 % to 16 %,
respectively. The bioavailability of the targeted prodrugs was about 1.2ꢀfold to
4.9ꢀfold greater than that of gemcitabine (3.2 %). Also, the halfꢀlives of the targeted
prodrugs were about 3ꢀfold greater than that of gemcitabine. The pharmacokinetic
results suggested that high membrane permeability together with enzymatic stability
markedly improved the oral bioavailability of gemcitabine. By contrast, GS showed
similar oral bioavailability to gemcitabine, indicating poor permeability and rapid
degradation by CD which probably limited its oral absorption.
To evaluate the role of OCTN2 in the oral absorption of GSC in vivo,
competitive inhibition studies were involving coꢀadministration with Lꢀcarnitine. The
absolute bioavailability of GSC decreased from 12% to 6% in the presence of
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
7
8
9
Lꢀcarnitine (P<0.05) (Table S3). These results further confirmed the improved oral
bioavailability was due to OCTN2ꢀmediated delivery. In addition, OCTN2 is
expressed in the brush border membranes of the proximal tubule cells and is
responsible for efficient reabsorption of filtered Lꢀcarnitine and acetylcarnitine (AC)
18. Thus, the targeted prodrugs may be reabsorbed into the system circulation
mediated by OCTN2, which probably contribute to the increased long residence time
of these targeted prodrugs. More importantly, GHC exhibited a significantly
prolonged plasma halfꢀlives and the highest oral bioavailability among the targeted
prodrugs, and the longer chains (8ꢀ10 carbons) did not help improve pharmacokinetic
behavior of gemcitabine through OCTN2. Therefore, the linker lengths are of a
crucial role in the performance of OCTN2ꢀtargeting oral prodrug strategy due to the
fatty acid length could affect the affinity of the substrate toward transporter, in good
agreement with the results of the cellular study.
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
To explore the bioactivation mechanism, systemic and portal vein plasma
concentrations of GSC and gemcitabine were determined. GSC was able to cross the
intestinal epithelium cells primarily in intact form, with a mean GSC concentration of
2456 ng/mL while the mean gemcitabine concentration was 199 ng/mL at 2h in portal
blood, suggesting that GSC remained stable after the intestinal absorption process.
Moreover, the mean concentration for gemcitabine in portal vein was only 10% that
of gemcitabine in the jugular vein at 2h, indicating that bioactivation mainly occurred
in the blood and liver. This result was different from our previous study of amino acid
esterꢀbased valylcytarabine and valyldidanosine21, 22, in which over 98% of prodrug
14
ACS Paragon Plus Environment

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
bioactivation occurred in the GI tract and intestinal cells. Also, our previous report
indicated that the relatively rapid degradation in the GI tract and intestinal cells was
unfavourable to improve the oral bioavailability and to reduce the intestinal toxicity of
transporterꢀtargeting prodrugs. By contrast, the introduction of an amide bond
contributed to the improvement in the stability of the prodrug, which changed the
bioactivation site in vivo. Therefore, the relatively good chemical and enzymatic
stability in the GI tract appears to help the oral bioavailability and to maintain the
minimal intestinal toxicity of OCTN2ꢀtargeting oral prodrugs.
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
Tissue distribution study
The tissue distribution was investigated in mice after a single oral administration
of 102 mg/kg of GHC (an equivalent gemcitabine dose of 50 mg/kg). The
concentrations of gemcitabine in vital organs were determined at 0.16 h, 0.5 h and 2 h
after administration (Figure 12). The tissue distribution results demonstrated that
GHC was mainly distributed in mice’s spleen, heart and lung at 0.5 h, but was rapidly
decreased after 2 h, indicating good safety of GHC in vital organs in the body.
In vivo safety studies
OCTN2 is widely expressed in all organs, including heart, liver, spleen, lung and
kidney, which may increase drug uptake and thus produce organ toxicity. In vivo
safety studies was carried out on heart, liver, spleen, lung and kidney tissue after
continuous administration of GHC for 14 days (50 mg/kg/day) using Hematoxylin
and eosin (H&E) histological analysis. Compared to PBS group, the microstructures
of organs from GHC treated mice were normal (Figure 13). No evidence of hydropic
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
7
8
9
degeneration, inflammatory infiltrates and fibrosis was observed in cardiac muscle,
hepatocytes, spleen, lung and kidney samples, indicating that there was no toxicity of
GHC in mice at short exposure duration, probably due to low tissue distribution and
rapid clearance in various tissues.
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
16
ACS Paragon Plus Environment

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
CONCLUSIONS
5
6
7
8
9
OCTN2ꢀtargeting oral prodrugs of gemcitabine simultaneously increased the
permeability and stability in comparison with PepT1ꢀtargeting prodrugs and
lipophilized prodrugs, resulting in higher oral bioavailability. No toxicity in mice was
observed after short exposure duration (14 days) following oral administration of
OCTN2ꢀtargeting gemcitabine prodrug. In particular, the hexane diacid linkage
exhibits a 2.2ꢀfold increased OCTN2 affinity and a 1.3ꢀfold increase in oral
bioavailability, compared with the succinic acid linkage. These findings support the
notion that in addition to PepT1, OCTN2 transporter can be used as a novel target to
deliver gemcitabine by the most practical oral route. Furthermore, the fatty acid
linkage chain length can modify the affinity of the substrate toward transporter. We
have shown that OCTN2ꢀtargeting oral prodrugs strategy may be feasible for the oral
deliver other antitumor agents.
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
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
Chemicals
Gemcitabine (98.8% purity) was purchased from Nanjing Chemlin Chemical
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
Industry Co., Ltd (Jiangsu, Nanjing, PR China); Lꢀcarnitine, isobutyl chlorocarbonate,
triethylamine, succinic anhydride, adipic anhydride, octandioic acid, decanedioic acid,
sulfoxide chloride, 1,4ꢀdioxane, DMF, benzyl bromide were supplied by Sigma
Aldrich and used without further purification; tetrahydrouridine (THU), a cytidine
deaminase inhibitor, was purchased from J﹠K Scientific (HPLC grade); Cacoꢀ2 cells
were obtained from ATCC and HEK293 cells were kindly provided by Dr XiuꢀLin YI
(Tian Jin Institute Research Pharmaceutical Corporation); Lꢀcarnitineꢀsuccinicꢀ
gemcitabine (GSC), Lꢀcarnitineꢀhexylicꢀgemcitabine (GHC), Lꢀcarnitineꢀoctanedioicꢀ
gemcitabine (GOC), Lꢀcarnitineꢀdecanedioicꢀgemcitabine (GDC), and succinicꢀ
gemcitabine (GS) were synthesized in Shenyang Pharmaceutical University
(Shenyang, China).
Synthesis of OCTN2ꢀtargeting prodrugs of gemcitabine
All compounds were analyzed by 1H NMR and MS. The purity of
OCTN2ꢀtargeting prodrugs was ≥95% as determined by HPLC.
Synthesis of GSC
A mixture of compound 2 (1.2 g, 0.003 mol), gemcitabine (0.78 g, 0.003 mol),
isobutyl chloroformate (0.6 g, 0.045 mol), triethylamine (TEA, 0.3 g, 0.003 mol) and
DMF (10mL) was stirred overnight at 60 °C, under an argon atmosphere. The solvent
was evaporated, and the residue was purified by column chromatography on a silica
18
ACS Paragon Plus Environment

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
gel, eluting with methanol in dichloromethane (gradient 5ꢀ10 %). The pure
intermediate was dissolved in methanol, and then Pd/C (10 %) was added. The
mixture was stirred under H2 at room temperature for 15 minutes and filtered. The
solvent was evaporated, and product was collected in a vacuum. Yield: 67 % of a
white powder. C20H28F2N4O9, MS (ESI)：m/z = 507.2 [M+H+]. HPLC purity: tR = 7.9
min, 98.2 %. 1H NMR (400 MHz, DMSOꢀd6): δ 10.86 (d, J = 286.0 Hz, 2H), 8.31 (d,
J = 7.6 Hz, 1H), 7.20 (d, J = 7.5 Hz, 1H), 6.48 (s, 1H), 6.15 (t, J = 7.1 Hz, 1H), 5.86
(dd, J = 13.9, 4.8 Hz, 2H), 5.39 (s, 1H), 5.25 (s, 1H), 4.28 – 4.03 (m, 8H), 3.86 (d, J =
8.6 Hz, 1H), 3.76 (s, 1H), 3.61 (ddt, J = 16.3, 12.4, 7.8 Hz, 10H), 3.36 (d, J = 19.8 Hz,
10H), 3.08 (s, 9H), 2.75 (s, 2H), 2.56 (dd, J = 11.9, 4.5 Hz, 4H).
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
Synthesis of GHC
A mixture of compound 3(1.2 g, 0.003 mol), gemcitabine (0.78 g, 0.003 mol),
isobutyl chloroformate (0.6 g, 0.045 mol), triethylamine (TEA, 0.3 g, 0.003 mol) and
DMF (10 mL) was stirred overnight at 60 °C, under an argon atmosphere. The solvent
was evaporated, and the residue was purified by chromatography on a silicaꢀgel
column, eluting with methanol in dichloromethane (gradient 5ꢀ10 %). The pure
intermediates were dissolved in methanol, and then Pd/C (10 %) was added. The
mixture was stirred under H2 at room temperature for 15 minutes and filtered. The
solvent was evaporated, and product was collected in a vacuum. Yield: 35 % of a
white powder. C22H32F2N4O9, MS, (ESI)：m/z = 535.3 [M+H+]. HPLC purity: tR = 8.5
min, 97.4 %. 1H NMR (400 MHz, D2O) δ 8.24 (d, J = 7.6 Hz, 1H), 7.42 (d, J = 7.6 Hz,
1H), 6.28 (t, J = 7.3 Hz, 1H), 5.64 (d, J = 6.3 Hz, 1H), 4.39 (td, J = 12.0, 8.9 Hz, 1H),
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
5
6
4.18 – 4.10 (m, 1H), 4.04 (s, 1H), 3.94 – 3.82 (m, 2H), 3.65 (s, 1H), 3.36 (s, 1H), 3.22
(d, J = 15.9 Hz, 9H), 2.60 – 2.38 (m, 5H), 1.77 – 1.58 (m, 4H).
7
8
9
Synthesis of GOC
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
A mixture of compound 4(1.23 g, 0.003 mol), gemcitabine (0.78 g, 0.003 mol),
isobutyl chloroformate (0.6 g, 0.005 mol), triethylamine (TEA, 0.3 g, 0.003 mol) and
DMF (10mL) was stirred overnight at 60 °C, under an argon atmosphere. The solvent
was evaporated, and the residue was purified by chromatography on a silicaꢀgel
column, eluting with methanol in dichloromethane (gradient 5ꢀ10 %). The pure
intermediates were dissolved in methanol, and then Pd/C (10 %) was added. The
mixture was stirred under H2 at room temperature for 15 minutes and filtered. The
solvent was evaporated, and product was collected in a vacuum. Yield: 42 % of a
white powder. C24H36F2N4O9, MS (ESI)：m/z = 563.2 [M+H+]. HPLC purity: tR = 10.2
1
min, 98.2 %. H NMR (400 MHz, DMSOꢀd6) δ 11.02 (s, 1H), 8.29 (d, J = 7.5 Hz,
1H), 7.38 (q, J = 2.8, 2.3 Hz, 1H), 7.28 (d, J = 7.6 Hz, 1H), 6.16 (d, J = 7.5 Hz, 1H),
5.47 (d, J = 26.2 Hz, 1H), 4.20 (dt, J = 13.0, 6.4 Hz, 1H), 3.89 (d, J = 8.0 Hz, 1H),
3.80 (d, J = 12.8 Hz, 2H), 3.66 (dd, J = 12.8, 3.4 Hz, 2H), 3.14 (d, J = 19.3 Hz, 9H),
2.60 (s, 2H), 2.41 (t, J = 7.3 Hz, 2H), 2.31 (d, J = 8.1 Hz, 2H), 1.53 (s, 4H), 1.27 (s,
4H).
Synthesis of GDC
A mixture of compound 4(1.77 g, 0.003 mol), gemcitabine (0.78 g, 0.003 mol),
isobutyl chloroformate (0.6 g, 0.005 mol), triethylamine (TEA, 0.3 g, 0.003 mol) and
DMF (10 mL) was stirred overnight at 60 °C, under an argon atmosphere. The solvent
20
ACS Paragon Plus Environment

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was evaporated, and the residue was purified by chromatography on a silicaꢀgel
column, eluting with methanol in dichloromethane (gradient 5ꢀ10 %). The pure
intermediates were dissolved in methanol, and then Pd/C (10 %) was added. The
mixture was stirred under H2 at room temperature for 15 minutes and filtered. The
solvent was evaporated, and product was collected in a vacuum. Yield: 47 % of a
white powder. C26H40F2N4O9, MS (ESI)：m/z = 591.3 [M+H+]. HPLC purity: tR = 11.3
min, 97.8 %. 1H NMR (400 MHz, DMSOꢀd6) δ 12.73 (s, 1H), 10.99 (s, 1H), 8.28 (d,
J = 7.6 Hz, 1H), 7.28 (d, J = 7.5 Hz, 1H), 6.40 (s, 1H), 6.16 (d, J = 7.5 Hz, 1H), 5.45
(s, 1H), 4.21 (d, J = 10.8 Hz, 1H), 3.91 – 3.88 (m, 1H), 3.82 (s, 2H), 3.68 (d, J = 3.5
Hz, 2H), 3.12 (s, 9H), 2.69 (d, J = 4.7 Hz, 2H), 2.40 (d, J = 7.3 Hz, 2H), 2.25 (t, J =
7.4 Hz, 2H), 1.53 (d, J = 6.5 Hz, 4H), 1.25 (s, 8H).
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
Synthesis of GS
A mixture of succinic anhydride (0.2 g, 0.002 mol), gemcitabine (0.78 g, 0.003
mol), triethylamine (TEA, 0.1 g, 0.001 mol) and DMF (10 mL) was stirred overnight
at 60 °C, under an argon atmosphere. The solvent was then evaporated, and the
residue was purified by preparative HPLC. Yield: 75 % of a white powder.
C13H17F2N3O5, MS (ESI)：m/z = 334.1 [M+H+]. HPLC purity: tR = 13.5 min, 97.5 %.
1H NMR (400 MHz, DMSOꢀd6): δ 10.99 (s, 2H), 8.26 (d, J = 7.6 Hz, 2H), 7.39 (d, J
= 24.9 Hz, 2H), 7.29 (d, J = 7.6 Hz, 2H), 6.15 (dd, J = 15.6, 8.0 Hz, 3H), 5.78 (d, J =
7.4 Hz, 1H), 4.25 ꢀ 4.10 (m, 3H), 3.88 (d, J = 8.5 Hz, 2H), 3.77 (dd, J = 18.8, 12.3 Hz,
4H), 3.68 ꢀ 3.59 (m, 3H), 2.38 (t, J = 7.3 Hz, 4H), 1.58 ꢀ 1.51 (m, 3H).
Stability
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
7
8
9
Plasma was obtained from rats and stored at ꢀ80 °C until required. Hepatic and
intestinal samples were washed with physiological saline, and 0.3 g tissue (weighed
accurately) was mixed with 1mL of physiological saline then homogenized for 1 min.
After centrifuging at 13,000 rpm at 4 °C for 5 min, the supernatant was stored at
ꢀ80 °C until required.
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 chemical stability of GS, GSC and GHC were tested in different pH
phosphate buffers (pH 1.2, 6.8, 7.4) at 37 °C for 24 h. Drug solutions were prepared
by dissolving the prodrugs in buffer to obtain a concentration of 5 ꢁM. In addition, the
enzymatic stability was determined in rat plasma, 20% rat hepatic homogenate and
20% rat intestinal homogenate at 37°C 24 h. The prodrugs were added to the plasma,
rat hepatic and intestinal homogenate to obtain a final concentration of 5 ꢁM. At the
appropriate intervals, samples were collected and then analyzed by HPLC. The
degradation halfꢀlives were calculated using the equation:
t1/2=0.693/k
Cacoꢀ2 Permeability
For transcellular transport studies, Cacoꢀ2 cells were grown on 12ꢀwell
polycarbonate filter inserts (0.4 ꢁm pore size, area 1.12 cm2, Corning, NY) at a
density of 1.0 × 105 cells/well and cultured in MEM for 21 days. The transepithelial
electrical resistance (TEER) over than 250 ꢃ·cm2 were used for the experiment. The
apparent permeability coefficients (Papp) were determined for the AL→BP and BP→
AL directions with and without the presence of 100 ꢁM Lꢀcarnitine. Drug solutions
were prepared in Han’s balanced salt solution (HBSS) to obtain a final concentration
22
ACS Paragon Plus Environment

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 25 ꢁM. For AL→BP transcellular transport, HBSS buffer containing the drugs was
added to the apical side and free HBSS buffer to the basolateral side. For BP→AL
transcellular transport, the drug solution was added to the basolateral side and free
HBSS buffer to the apical side. The samples were collected from the apical side or
basolateral side at 15, 30, 45, 60, 90, and 120 min at 37 °C. Each experiment was
repeated three times. Drug concentrations in both sides were determined by
UPLCꢀMS/MS and Papp was calculated using the formula:
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
Papp = dCr/dt × Vr × 1/A × 1/C0
where dCr/dt is the steadyꢀstate flux across the monolayer (ꢁmol/s), Vr is the receiver
volume, A is the surface area of the monolayer (1.13 cm2), and C0 is the initial
concentration (ꢁM) in the test solution. The concentrations of gemcitabine and its
prodrugs were analyzed by UPLC−MS/MS.
Pharmacokinetic Evaluation
All surgical and experimental procedures were performed in accordance with
institutional guidelines and approved by the Shenyang Pharmaceutical University
Animal Care and Use Committee.
Six groups of Wistar rats (six animals each group) were orally administrated of
gemcitabine, GSC, GHC, GOC, GDC and GS at a gemcitabine dose of 50 mg/kg. In
addition, GSC was orally coꢀadministered with 40 mM Lꢀcarnitine to the rats. Then,
serial blood samples were collected at 5, 10, 15, 30 min and 1, 2, 4, 6, 8, 12, 24h. To
calculate the absolute bioavailability, gemcitabine was also given intravenously to rats
at a dose of 7.5 mg/kg. Serial blood samples were collected at 15, 30 min and 2, 4, 8,
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3
4
5
6
7
8
9
12, 24 h and then transferred to THUꢀpretreated heparinized tubes (10 mg/mL) after
centrifuging at 3000 × g for 15 min. The supernatant of samples were stored at ꢀ80 °C
until analysis. The absolute bioavailability of gemcitabine was calculated using the
following equation:
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
Fabs =AUCpo × Div/AUCiv/Dpo × 100%
where Fabs is the absolute bioavailability, AUCpo is the area under the plasma
concentrationꢀtime curves of oral administration, Div is the intravenous administration
dose, AUCiv is the area under the plasma concentrationꢀtime curves of intravenous
administration, and Dpo is the dose orally administered.
Tissue distribution study
Nine mice were randomly divided into three groups (n = 3) and orally
administrated of GHC at the dose of 102 mg/kg. The tissue samples, including heart,
liver, spleen, lung, kidney and brain were collected at 0.16, 0.5 and 2 h postꢀdosing.
Tissue samples were washed with physiological saline, and 0.3 g tissue (weighed
accurately) was mixed with 1mL of physiological saline then homogenized for 1 min.
After centrifuging at 13,000 rpm at 4 °C for 5 min, the supernatant was stored at
ꢀ80 °C until analysis.
In vivo toxicity study
Two groups of Kunming mice (four animals each group) were orally
administrated of GHC or physiological saline (control) at the dose of 102 mg/kg once
daily for 14 days. Body weight of each mouse was recorded every other day for 14
days. For histology analysis, mice were sacrificed and the heart, liver, spleen, lung,
24
ACS Paragon Plus Environment

Page 25 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and kidney tissue samples were removed, and fixed in 10% buffered formalin
immediately, followed by embedding in paraffin, sectioning, and hematoxylin and
eosin staining. The histological sections were imaged using an optical microscope.
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
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
5
6
ASSOCIATED CONTENT
Supporting Information
7
8
9
Additional supporting documents and experimental results are available,
including measurement of the WaterꢀOctanol Partition Coefficient (Log P), cell
culture, Cellular uptake mechanism study, Lꢀcarnitine Uptake Inhibition, Contribution
of Nucleoside transporter to the oral absorption, Cytotoxicity, In situ singleꢀpass
perfusion, the systemic and portal vein pharmacokinetic evaluation, development of
analytical method, statistical analysis and Molecular formula strings and some data
(CSV). This material is available free of charge via the Internet at http://pubs.acs.org.
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
26
ACS Paragon Plus Environment

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
AUTHOR INFORMATION
5
6
7
8
9
∗ Corresponding author at: Eꢀmail addresses: sunjin66@21cn.com (J. Sun). Phone:
+86 24 23986325; fax: +86 24 23986325.
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
Notes
The authors declare no competing financial interest
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
ABBREVIATIONS USED
5
6
7
8
9
The organic cation transporter II, OCTN2; cytidine deaminase, CD;
2’,2’ꢀdifluoroꢀ2’deoxyuridine, dFdU; gastrointestinal, GI; Lꢀcarnitineꢀsuccinicꢀ
gemcitabine, GSC; Lꢀcarnitineꢀhexylicꢀgemcitabine, GHC; Lꢀcarnitineꢀoctanedioicꢀ
gemcitabine, GOC; Lꢀcarnitineꢀdecanedioicꢀgemcitabine, GDC; palmitoylꢀLꢀ
carnitine, PC; succinicꢀgemcitabine, GS; Nucleoside Transporterꢀ1, hENT1; 3ꢀ[4,
5ꢀdimethylthiazolꢀ2ꢀyl]ꢀ2, 5ꢀdiphenyltetrazolium bromide, MTT; halfꢀmaximum
inhibitory concentration, IC50; Lꢀcarnitine and acetylcarnitine, AC; Hematoxylin and
eosin , H&E; tetrahydrouridine, THU; transepithelial electrical resistance, TEER;
apparent permeability coefficients, Papp; Han’s balanced salt solution, HBSS;
calculated, calcd; electrospray ionization, ESI; dimethylformamide, DMF; Michaelis
constant, Km; highꢀperformance liquid chromatography, HPLC; hertz, Hz; intravenous,
iv; coupling constant (in NMR spectrometry), J; nuclear magnetic resonance, NMR;
minute(s), min; milliliter, mL; millimolar, mM; mass spectrometry, Ms; Protein Data
Bank, PDB; halfꢀtime, t1/2.
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
28
ACS Paragon Plus Environment

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
ACKNOWLEDGMENTS
5
6
7
8
9
We are grateful for financial support from the National Natural Science
Foundation of China (No. 81473164) and Technology bureau in Shenyang (No.
ZCJJ2013402). We are also grateful for financial support from the Project for New
Century Excellent Talents of Ministry of Education (No.NCETꢀ12ꢀ1015) and Specific
Science Foundation of Shenyang Pharmaceutical University (No. ZCJJ2014409).
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
29
ACS Paragon Plus Environment