Aglycone Ebselen and β- D -Xyloside Primed Glycosaminoglycans Co-contribute to Ebselen β- D -Xyloside-Induced Cytotoxicity

Article abstract of DOI:10.1021/acs.jmedchem.7b01835

Most β-d-xylosides with hydrophobic aglycones are nontoxic primers for glycosaminoglycan assembly in animal cells. However, when Ebselen was conjugated to d-xylose, d-glucose, d-galactose, and d-lactose (8A-D), only Ebselen β-d-xyloside (8A) showed significant cytotoxicity in human cancer cells. The following facts indicated that the aglycone Ebselen and β-d-xyloside primed glycosaminoglycans co-contributed to the observed cytotoxicity: 1. Ebselen induced S phase cell cycle arrest, whereas 8A induced G2/M cell cycle arrest; 2. 8A augmented early and late phase cancer cell apoptosis significantly compared to that of Ebselen and 8B-D; 3. Both 8A and phenyl-β-d-xyloside primed glycosaminoglycans with similar disaccharide compositions in CHO-pgsA745 cells; 4. Glycosaminoglycans could be detected inside of cells only when treated with 8A, indicating Ebselen contributed to the unique property of intracellular localization of the primed glycosaminoglycans. Thus, 8A represents a lead compound for the development of novel antitumor strategy by targeting glycosaminoglycans.

Full text of DOI:10.1021/acs.jmedchem.7b01835

Subscriber access provided by - Access paid by the | UCSB Libraries
The Aglycone Ebselen and #-D-Xyloside Primed Glycosaminoglycans
Co-contribute to Ebselen #-D-Xyloside-induced Cytotoxicity
Yang Tang, Siqi Zhang, Yajing Chang, Dacheng Fan, Ariane de Agostini, Lijuan Zhang, and Tao Jiang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01835 • Publication Date (Web): 27 Mar 2018
Downloaded from http://pubs.acs.org on March 28, 2018
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 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
The Aglycone Ebselen and βꢀDꢀXyloside Primed Glycosaminoglycans
Coꢀcontribute to Ebselen βꢀDꢀXylosideꢀinduced Cytotoxicity
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
Yang Tang†,‡,# Siqi Zhang†,# Yajing Chang†,‡ Dacheng Fan †, Ariane De Agostini§, Lijuan
Zhang‡,* and Tao Jiang†,•,*
†Key Laboratory of Marine Drugs, Ministry of Education, School of Medicine and Pharmacy,
Ocean University of China,
‡Medical Systems Biology Center for Complex Diseases, Affiliated Hospital of Qingdao
University, Qingdao, 266003, P. R. China
§Department of Gynecology and Obstetrics, Geneva University Hospitals and University of
Geneva, Geneva 14, Switzerland
•Laboratory for Marine Drugs and Bioproducts of Qingdao National of Laboratory for Marine
Science and Technology, Qingdao 266003, P.R. China
Abstract
Most of βꢀDꢀxylosides with hydrophobic aglycones are nonꢀtoxic primers for glycosaminoglycan
assembly in animal cells. However, when Ebselen was conjugated to Dꢀxylose, Dꢀglucose, Dꢀ
galactose, and Dꢀlactose (8AꢀD), only Ebselen βꢀDꢀxyloside (8A) showed significant
cytotoxicity in human cancer cells. The following facts indicated that the aglycone Ebselen and
βꢀDꢀxyloside primed glycosaminoglycans coꢀcontributed to the observed cytotoxicity: 1. Ebselen
induced S phase cell cycle arrest whereas 8A induced G2/M cell cycle arrest; 2. 8A augmented
early and late phase cancer cell apoptosis significantly compared to that of Ebselen and 8BꢀD ; 3.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 47
1
2
3
4
5
6
7
8
9
Both 8A and phenyl ꢀβꢀDꢀxyloside primed glycosaminoglycans with similar disaccharide
compositions in CHOꢀpgsA745 cells; 4. Glycosaminoglycans could be detected inside of cells
only when treated with 8A, indicating Ebselen contributed to the unique property of intracellular
localization of the primed glycosaminoglycans. Thus, 8A represents a lead compound for the
development of novel antitumor strategy by targeting glycosaminoglycans.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 47
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
9
Glycosaminoglycans (GAGs) are hexosamineꢀcontaining linear polysaccharides. Two major
types of GAGs are heparan sulfate (HS) and chondroitin sulfate (CS) consisting of repeating
disaccharide of hexosamine (GlcN/GlcNAc for HS and GalNAc for CS) and hexuronic acid
(GlcA and IdoA for both HS and CS). HS and CS share the same initiating tetrasaccharide
sequence (GlcAꢀGalꢀGalꢀXyl) where the xylose links to the core proteins of proteoglycans.
Biosynthesis of GAG chains can take place independently of proteoglycan core proteins by
priming on βꢀDꢀxylosides, where the hydrophobic aglycones of βꢀDꢀxylosides penetrate cell
membranes to initiate the GAG assemble on xylosides in animal cells.1 Because of the immense
structural diversity of GAG chains, GAGs play important roles in organizing the extracellular
matrix, in activating multiple signaling pathways by serving as scaffolds for the interactions
between growth factors/cytokines and their receptors in cytokinesis and in cell division.2ꢀ7
Particularly, it is wellꢀdocumented that GAG biosynthesis dysfunctions are common in many
pathological conditions, including tumor progression and inflammation.8ꢀ11 Considering that the
dysfunctions of GAG biosynthesis are common in many pathological conditions, the exploitation
of xylosides for drug development are reasonable and promising. Some xylosides have reported
antiꢀthrombosis12, cell morphology13ꢀ14, angiogenesis15 and antiꢀtumor effects16.
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
Most of βꢀDꢀxylosides with hydrophobic aglycones are not toxic to animal cells up to 100 ꢀM
concentrations.17ꢀ18 Mani et al reported that the xyloside 2ꢀ(6ꢀhydroxynaphthyl) ꢀDꢀxyloside
(XylNapOH), in contrast to 2ꢀnaphthyl ꢀDꢀxyloside (XylNap), specifically reduces tumor growth
both in vitro and in vivo,19ꢀ21 indicating the subtle differences in aglycone structures lead to
significant difference in tumor biology. The same research group recently reported that the CS
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 47
1
2
3
4
5
6
7
8
9
chains primed on both XylNapOH and XylNap from HCC70 cells have similar structure and
similar cytotoxicity, indicating the CS chains but not the difference in aglycones are responsible
for the observed cytotoxicity. In contrast, despite the HS chain primed on both XylNapOH and
XylNap from HCC70 cells have similar disaccharide compositions, the XylNapꢀprimed HS
chains inhibit the cytotoxic effect of the XylNapOHꢀ and XylNapꢀprimed CS chains on HCC70
cells, whereas the XylNapOHꢀprimed HS is inactive, which demonstrates that the difference in
the aglycones XylNapOH and XylNap are responsible for the observed biological activities22.
However, the underlying molecular mechanisms are largely unknown. Thus, our goal is to
develop novel βꢀDꢀxylosides by conjugating known cytotoxic compound to xylose in order to
understand the role of the aglycone and primed GAGs in contributing to βꢀDꢀxylosideꢀdependent
cytotoxicity in cancer cells.
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 current study, we chose cytotoxic Ebselen as an aglycone. Ebselen is a hydrophobic
selenoorganic compound potently inhibits lipid peroxidation through a glutathione peroxidaseꢀ
like action23ꢀ24 and exerts a wide spectrum of biological activities, ranging from antiꢀoxidant,
cytoprotective, neuroprotective, and antiꢀinflammatory. Moreover, Ebselen also inhibits other
enzymes such as cyclooxygenases, lipoxygenases, and indoleamine 2, 3ꢀdioxygenase, which play
a broad functional role in cancer signaling, as well as the regulation of the immune response.25ꢀ30
Furthermore, it has been reported that Ebselen sensitizes glioblastoma cells to Tumor Necrosis
Factor (TNFa)ꢀinduced apoptosis through NFꢀkB down regulation.31 In prior studies, we have
observed that the benzoisoselenazoloneꢀfragment containing compounds derived from Ebselen
have anticancer properties by inhibiting focal adhesion kinase (FAK), AKTꢀ1, and protein
kinase Cꢀa (PKCꢀa).32 In current study, we conjugated the known cytotoxic compound Ebselen
ACS Paragon Plus Environment

Page 5 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
to Dꢀxylose and three other monosaccharides to synthesize Ebselen βꢀDꢀglycosides (8A, 8B, 8C,
and 8D). We found only Ebselen βꢀDꢀxyloside (8A) showed significant cytotoxicity with IC50
ranged from 6.4 to 7.3 ꢀM in four human cancer cell lines tested. Most importantly, we
discovered for the first time that the aglycone and βꢀDꢀxyloside primed glycosaminoglycan
chains coꢀcontributed to the Ebselen βꢀDꢀxylosideꢀinduced cytotoxicity.
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
Results
Chemistry.
The general method for the synthesis of the four glycosyl conjugated Ebselen is shown in
Scheme 1. We used peracetylated xylose, glucose, galactose, and lactose as glycosyl donors and
coupled them to 2ꢀchloroselenobenzoyl chloride to generate the glycosyl conjugated Ebselen
compounds.
Cytotoxicity.
8A Inhibited the Growth of Human Cancer Cells.
The cytotoxicity of the Ebselen, 8A, 8B, 8C, and 8D was studied with a negative control
condition of 0.1% DMSO in four human cancer cell lines including two lung cancer cell lines
A549 and H1299 and two colon cancer cell lines HCT116 and HT29 and their IC50 values are
shown in Table 1. Compared to the cytotoxic effect of Ebselen, all four EbselenꢀβꢀDꢀglycosides
were more toxic to colon cancer cells. Most importantly, Ebselen βꢀDꢀxyloside (8A) had much
higher increased cytotoxicity towards all four human cancer cell lines tested compared to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 47
1
2
3
4
5
6
7
8
9
Ebselen and other EbselenꢀβꢀDꢀglycosides (8B, 8C, and 8D). Cancer cell proliferation was
significantly suppressed in the presence of 8A in a concentrationꢀdependent manner (Figure 1).
8A had the same IC50 values of 7.0 ꢁM in both lung cancer cell lines A549 and H1299. The IC50
values for 8A were 6.4 ꢁM and 7.3 ꢁM in two colon cancer cell lines HCT116 and HT29 cells,
respectively (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
Effects of 8A on Cell Cycle and Apoptosis
To test if the cytotoxicity of 8A was due to the presence of the cytotoxic Ebselen in the xyloside,
we compared the extent of apoptosis and cell cycle distribution by flow cytometry analysis
among 8A, Ebselen, 8B, 8C, and 8D.
To study how 8A affected the cell cycle, colon cancer cells HCT 116 were treated with 8A, 8B,
8C, 8D, or Ebselen, respectively, for 24 h (Figure S2A). Ebselen (40 ꢁM) caused cell cycle arrest
at S phase as expected. In contrast, 8A at 12.5 ꢁM concentration increased G2/M cell population
to 45.2%, 1.3ꢀfold increase of that of the control cells, and 8A at 15 ꢁM concentrations further
increased G2/M cell population to 51.9%. At 20 ꢁM concentration, 8B, 8C, and 8D had similar
effect on cell cycle as 8A at 12.5 ꢁM (Figure 2A).
For the apoptosis analysis, human colon cancer cells HCT 116 were stained with Annexin
V/propidium iodide and analyzed by flow cytometry. Viable cells (FITCꢀnegative) and early
apoptotic cells (FITCꢀpositive) were PIꢀnegative, whereas late apoptotic and necrotic cells were
ACS Paragon Plus Environment

Page 7 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
PIꢀpositive and FITCꢀpositive. As shown in Figure S2B, only viable cells were observed in the
untreated control cells (K3) whereas the 8A treatment increased the percentage of both early and
late apoptotic cells significantly. The data in Figure 2B showed that 8A at 7.5 ꢁM increased
early apoptotic cell population by 27.7ꢀfold compared to that of control cells in HCT 116 cells.
When the concentration of the 8A was increased to 12.5 ꢁM, the populations of early and late
apoptotic cell were further increased to 10.97% and 13.18%, respectively. In contrast, 40 ꢁM of
Ebselen, 8B, 8C, or 8D induced less early and late apoptotic cells compared to that of 8A at 7.5
ꢁM (Figure 2B).
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
Glycochemistry.
8A Primed Both CS and HS Chains.
8A was more toxic (Figure 1) with much lower IC50 values (Table 1) in four different human
cancer cell lines compared to that of Ebselen, 8B, 8C, or 8D. 8A also affected cell cycle and
induced apoptosis in a way different from that of Ebselen, 8B, 8C, or 8D (Figure 2).
Theoretically, 8A could prime for GAG chains whereas Ebselen, 8B, 8C, or 8D. 8A could not,
which suggested that the GAG priming ability of 8A might contribute to the observed
cytotoxicity. This was tested using a mutant Chinese hamster ovary (CHO) cell line pgsAꢀ745
that lack active xylosyltransferase.33 As a consequence, CHO pgsAꢀ745 cells cannot substitute
xylose on hydroxyl group of serine on proteoglycan core proteins, thus defective in GAG
biosynthesis. However, exogenous supply of βꢀxylosides in cell culture media could bypass the
xylosyltransferase deficiency and stimulate 3 to 5ꢀfold increase in GAG biosynthesis in CHO
pgsAꢀ745 cells compared to wildꢀtype CHO K1 cells. In our experiment, the CHO pgsAꢀ745
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 47
1
2
3
4
5
6
7
8
9
cells were incubated with 0.1% DMSO (negative control) or with 10 ꢀM of either 8A or phenylꢀ
βꢀDꢀxyloside (positive control) with a final 0.1% DMSO in cultural medium supplemented with
additional 100 mg/L sodium sulfate. After 48 h of incubation, the GAG chains were isolated by
anion exchange resins from cells and their culture media. After complete hydrolysis, the amount
of GlcN and GalN in each GAG sample was quantified. The results in Figure 3A showed that the
amount of GlcN and GalN detected in the GAGs extracted from the negative control cells
(Figure S3A, blue tracer) were barely detectable, which was consistent with the original report
about the CHO pgsAꢀ745 cells.33 The amount of GlcN and GalN detected in the GAGs extracted
from the 8Aꢀtreated CHO pgsAꢀ745 cells had 12.2ꢀfold increase in GlcN and 27.6ꢀfold increase
in GalN compared to that of the negative control, indicating 8A stimulated GAG biosynthesis in
CHO pgsAꢀ745 cells (Figure S3A, green tracer). However, the positive control, i.e. phenylꢀβꢀDꢀ
xylosideꢀtreated CHO pgsAꢀ745 cells, stimulated more GAG biosynthesis based on the amount
of GlcN and GalN detected (Figure S3A, red tracer). Since fetal bovine serum used in cell
culture media contains significant amount of GAGs,34 both GlcN and GalN were detected in the
GAGs extracted from the media of the negative control (Fig. S3B, blue tracer). 8A treatment did
not make big increase in the amount of both GlcN and GalN in GAGs extracted from the culture
media compared to the negative control (Fig. S3B, red tracer vs. blue tracer). However, there was
significant increase in the amount of GalN in GAGs extracted from the culture media of phenylꢀ
βꢀDꢀxylosideꢀtreated CHO pgsAꢀ745 cells. Based on the GlcN and GalN quantification data,
phenylꢀβꢀDꢀxyloside was more efficient at priming GAG chains and most of them were secreted
into the cell culture media compared to 8A. In contrast, only a limited fraction of the GAG
chains primed by 8A was secreted into the culture media.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 9 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The data shown in Figure 1 & 3 indicated that 8A was different from most of reported βꢀDꢀ
xylosides in three aspects: 1. potent cytotoxicity; 2. limited amount of primed GAGs in cell
culture media; and 3. significant amount of cellꢀassociated GAGs. In order to investigate whether
the cellꢀassociated GAG chains primed on 8A were located inside of the CHO pgsAꢀ745 cells,
we fixed the βꢀDꢀxylosideꢀtreated cells by 10% methanol to permeate the cell membrane and
make the staining of intracellular GAGs possible. The stained cells were then analyzed by flow
cytometry and confocal microscope independently. Figure 4A is the images under the confocal
microscope illustrating the treated CHO pgsAꢀ745 cells doubleꢀstained with antibodies against
chondroitinase ABC preꢀdigested monoꢀsulfated nonꢀreducing terminal of CS chains by monoꢀ
antibody 2B6 and nucleus by DAPI. Theoretically, fluorescence staining intensity on cells with
MAb 2B6 should be increased after chondroitinase ABC digestion. In the experiment, we
adjusted the exciting laser to an intensity under which the cells treated in control condition had
no visible stain and then used the identical exciting laser to stain the βꢀDꢀxylosideꢀtreated cells.
Under this condition, cells treated with phenylꢀβꢀDꢀxyloside exhibited few green fluorescence
around the nucleus (Figure 4A) when stained with MAb 2B6 (Figure 4A). In contrast, the green
fluorescence with direct MAb 2B6 stain in 8A treated cells was obvious and stronger than that in
the phenylꢀβꢀDꢀxyloside treated cells. After treatment with chondroitinase ABC, the green
fluorescence became stronger as expected, indicating specific CS staining.
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
Similar results were obtained when the antibody against heparin lyase digested HS chains (Mab
3G10) was used for immunostaining the xylosideꢀtreated and untreated CHO pgsAꢀ745 cells
(Figures 4B) where the fluorescence intensity was strongest in the cells treated with 8A after
heparin lyase digestion.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 47
1
2
3
4
5
6
7
8
9
To quantify the fluorescent intensity in the xylosideꢀtreated and ꢀuntreated CHO pgsAꢀ745 cells
after 2B6 and 3G10 plus Alexa 488 labeled secondary antibody staining, we performed flow
cytometry analysis on each stained sample (Figure 4C). 8A treated cells showed the strongest
fluorescent intensity followed by that of phenylꢀβꢀDꢀxylosideꢀtreated and 0.1% DMSO treated
cells (control). Overall results from Figure 4 indicated that 8Aꢀprimed GAG chains were more
abundant inside CHO pgsAꢀ745 cells. The primed GAG chains could be detected using
antibodies that recognize chondroitinase ABC or heparinase generated novel structural motifs
after disrupting the cell membrane with 4% formaldehyde.
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 further confirm the observation that GAGs were localized in cellular compartment of cells
after 8A treatment, we decided to use the cell surface FGF/FGFR/GAG ternary complex
formation assay we have used on CHO wildꢀtype and mutant cells previously35 to detect the 8A
primed GAGs inside of CHO pgsAꢀ745 cells. To this end, we chose FGF7 and FGFR2BꢀFc
fusion protein. We have previously shown that both HS and CS could facilitate FGF7 and
FGFR2Bꢀdependent signaling in BaF3 cells.34 The binding of FGFR2BꢀFc to CHO pgsAꢀ745
cells was monitored by confocal microscopy (Figure 5). The control CHO pgsAꢀ745 cells
showed no binding and the normal cellular morphology. The cells treated with phenylꢀβꢀDꢀ
xyloside also showed no binding. However, under the same detecting condition, we observed the
fluorescence signal within the cells treated with 8A. Interestingly, The dividing cells showed a
much stronger fluorescent signal, indicating the enrichment of the 8Aꢀprimed GAGs in these
cells (Figure 5), suggesting the intracellular GAGs might be responsible the observed
cytotoxicity.
ACS Paragon Plus Environment

Page 11 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
LCꢀMS Analysis of PMPꢀlabeled HS and CS Disaccharides
8
9
Quantitative disaccharide compositional analysis of the primed GAG chains is extremely
important in understanding the 8Aꢀinduced cytotoxicity. However, there is great need for a
sensitive, reliable and quantifiable method for the analysis of the GAGꢀderived disaccharides
from CHO pgsAꢀ745 cells that have grown in different culture conditions. Our laboratory has
established LCꢀMS analytic methods for the quantitative disaccharide compositional analysis of
GAGs from cells, tissue and blood that rely on application of stable isotope labeling.36 Briefly,
the 1ꢀphenylꢀ3ꢀmethylꢀ5ꢀpyrazolone (PMP) can be conveniently introduced into the reducing end
of GAGꢀderived disaccharides in alkaline condition. PMPꢀlabeling offers chromatographic
separation of GAGꢀderived disaccharides and also a strong UV absorbance at 245 nm.
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 confirmed the identity of each disaccharide qualitatively based on its retention time detected
by UV and molecular weights (Mw) detected by MS by injecting PMPꢀlabeled commercial
disaccharide standard one at a time. The Mw of PMPꢀlabeled disaccharides were calculated from
the sum of the disaccharide and PMP molecular weights, minus water and proton and are listed
in Table S1. The mass spectra of the 5 kinds of PMPꢀlabeled HS disaccharides showed that nonꢀ
sulfated disaccharide had only an [MꢀH]− molecularꢀion peak whereas monoꢀsulfated
disaccharides had both z1 [MꢀH]− and z2 [M−2H]2− molecularꢀion peaks. All 5 PMPꢀlabeled HS
disaccharides were completely separated using gradient ratio of 10 mM ammonium acetate and
acetonitrile as elution solution as described in the “Methods”.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 47
1
2
3
4
5
6
Pairꢀwised comparison of the quantitative changes in each CS and HS disaccharide in 8Aꢀ or
phenylꢀβꢀDꢀxylosideꢀtreated vs. 0.1% DMSOꢀtreated CHO pgsAꢀ745 Cells.
7
8
9
In order to quantify GAG disaccharides extracted from different sources, GAGs from the same
amount of cellular proteins were digested with either chondroitinase ABC to obtain CS
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
disaccharides or heparinase , ,
to obtain HS disaccharides. We then used the regular PMP
(5HꢀPMP) to label disaccharides derived from GAGs of 8Aꢀ or phenylꢀβꢀDꢀxylosideꢀtreated
CHO pgsAꢀ745 cells and deuterium substituted PMP (5DꢀPMP) to label the disaccharides
derived from GAGs of 0.1% DMSO treated control CHO pgsAꢀ745 cells. We then coꢀinjected
the disaccharides with different PMP tagging for the LCꢀMS analysis. This is an accurate
disaccharides quantitative method because the same disaccharide is eluted at the same position
after LC separation but with different molecular ions intensities detected by MS that can be
quantified pair wisely.36ꢀ37. All the disaccharides identified by the LCꢀMS analysis were
summarized in Table 2.
As shown in Figure 6A, the left peak (708.23) presented the nonꢀsulfated ꢁUAꢀGalNAc labeled
by 5HꢀPMP derived from cells treated with 8A (A), and the right peak (z1 718.29) which
presented the same nonꢀsulfated ꢁUAꢀGalNAc labeled by 5DꢀPMP from the control cells.
Similar increase was observed between phenylꢀβꢀDꢀxylosideꢀtreated cells and the control cells
(Figure S2A). These results confirmed that both 8A and phenylꢀβꢀDꢀxyloside could effectively
prime GAG chains. In contrast, the CHO pgsAꢀ745 cells were unable to synthesize significant
amount of GAGs in the absent of xylosides, which was consistent with published result.33 In
addition, we detected comparable amount of nonꢀsulfated ꢁUAꢀGalNAc disaccharides from the
cell cultural media of both 8Aꢀ or 0.1% DMSOꢀtreated cells, which was consistent with our
previous published results that fetal bovine serum contains significant amount of both HS and
ACS Paragon Plus Environment

Page 13 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CS.38 However, the amount of nonꢀsulfated ꢁUAꢀGalNAc from the cell culture media of phenylꢀ
βꢀDꢀxylosideꢀtreated cells was 11 times of that from the control cells (Figure S2B). The amount
of nonꢀsulfated ꢁUAꢀGalNAc from the cell culture media of 8Aꢀtreated cells was also 2.24 times
of the control (Figure 6B). As shown in Figure 6C, the peak on the left (z1 788.18) represented
the 4ꢀOꢀsulfated ꢁUAꢀGalNAc (ꢁUAꢀGalNAc 4S) from the cells treated with 8A, and the peak
on the right (z1 798.25) represented ꢁUAꢀGalNAc4S from the control cells. Based on the MS
analysis, 8Aꢀtreatment increased the levels of monoꢀsulfated disaccharides, 4ꢀOꢀ, and 6ꢀOꢀ
sulfated ꢁUAꢀGalNAc similar to that of phenylꢀβꢀDꢀxyloside (Figure S2C). To be precise, we
detected 10.5 times of ꢁUAꢀGalNAc6S in phenylꢀβꢀDꢀxylosideꢀtreated cells and 3.4 times of
ꢁUAꢀGalNAc6S in 8Aꢀtreated cells compared to that of the control. Furthermore, due to limited
amount of 4ꢀOꢀsulfated ꢁUAꢀGalNAc was detected in the 0.1% DMSOꢀtreated control cells, the
amount of ꢁUAꢀGalNAc4S was up to 99. times and 34.5 times in phenylꢀβꢀDꢀxylosideꢀ and 8Aꢀ
treated cells, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We also digested each GAG sample with heparinase , , and analyzed the HS disaccharides
by the LCꢀMS. As shown in Figure 6D, the most abundant HS disaccharide, nonꢀsulfated ꢁUAꢀ
GlcNAc tagged by two 5HꢀPMPs (z1 708.23), were detected in both phenylꢀβꢀDꢀxylosideꢀ
(Figure S2D) or 8Aꢀtreated cells. In contrast, the peak on the right (z1 718.29) that theoretically
represented nonꢀsulfated ꢁUAꢀGlcNAc tagged by 2 molecules of 5DꢀPMPs (z1 718.23) from
cells treated with 0.1% DMSO was undetectable. These observations indicated that both 8A and
phenylꢀβꢀDꢀxyloside initiated HS synthesis in CHO pgsAꢀ745 cells.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 47
1
2
3
4
Discussion
5
6
7
8
9
In the current study, the xylosides were dissolved in DMSO and then diluted in cell culture
medium before adding to cells. A final concentration of 0.1% DMSO was used in all cell culture
experiments including the negative, positive, and 8Aꢀtreated cells.39 The reason behind this is
that DMSO at high concentration can lyse cells and therefore the xylosides dissolved in DMSO
should not be added directly to cells. DMSO is also known to cause differentiation of certain
cells, which may be associated with a change in GAG biosynthesis.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
It has been reported that fully acetylated xylosides are easier to cross the plasma membrane and
get into the endoplasmic reticulum/Golgi network for efficient GAG priming.40ꢀ41 Furthermore,
we have synthesized over 51 different xylosides with novel aglycones structures during the past
and found out that the acetylated compounds had almost identical activity as that deacetylated
compounds. Combined with the published knowledge and our previous research, we
intentionally left all EbselenꢀβꢀDꢀglycosides fully acetylated for our current study.
Xylosides are powerful tools in studying GAG biosynthesis. Historically, Dꢀxyloside was
identified as an initiator of glycosaminoglycans (GAGs) production by the discovery of its
ability to restore CS biosynthesis inhibited by puromycin in embryonic chick epiphyseal
cartilage.42 Subsequently, it was reported that βꢀDꢀxylosides, pꢀnitrophenylꢀβꢀDꢀxylose, and 4ꢀ
methylumbelliferylꢀβꢀDꢀxylose are more effective than Dꢀxylose in enhancing the production of
CS.43 By competing with proteoglycan core proteinꢀlinked xylose, βꢀDꢀxylosides induce the
secretion and biosynthesis of proteinꢀfree CS chains.44 Similar to the proteoglycan core proteins,
the structure of aglycones play an important role in regulating GAG biosynthesis in cells. Indeed,
the aglycone moiety of xyloside is largely responsible for the size, type, and fine structure of the
primed GAG chains. For initiating CS biosynthesis, phenyl OꢀβꢀDꢀxyloside and SꢀβꢀDꢀxyloside
ACS Paragon Plus Environment

Page 15 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
are superior compared to Nꢀxyloside and homoꢀCꢀxyloside.45 For initiating HS biosynthesis, 3ꢀ
estradiolꢀβꢀDꢀxyloside and a series of synthetic and natural βꢀDꢀxylosides carrying two or more
fused aromatic rings as aglycones are potent inducers of both HS and CS biosynthesis in Chinese
hamster ovary cells.18, 46ꢀ47 However, most of xylosides reported so far have no cytotoxicity.
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 reported cytotoxic βꢀDꢀxyloside Xylꢀ2ꢀNapꢀ6ꢀOH19 has an IC50 value of 130±20 ꢀM vs. 7.0
± 0.37 ꢀM for 8A (Table 1) observed in our study, which indicated 8A was a unique cytotoxic
xyloside. We concluded that the aglycone Ebselen and βꢀDꢀxyloside primed GAGs coꢀ
contributed to the observed cytotoxicity based on the following facts: 1. instead of causing cell
cycle arrest at S phase by Ebselen, Ebselen βꢀDꢀxyloside induced G2/M cell cycle arrest (Figure
1); 2. Ebselen βꢀDꢀxyloside significantly increased early and late phase cancer cell apoptosis
compared to Ebselen and other Ebselen βꢀDꢀglycosides (Figure 2); 3. when treated CHOꢀ
pgsA745 cells with Ebselen βꢀDꢀxyloside along with phenyl ꢀβꢀDꢀxyloside, we found that both
xylosides not only primed GAGs based on monosaccharide quantification of media and cell
associated GAGs (Figure 3), but also shared the similar disaccharide compositions based on LCꢀ
MS analysis of the GAGs after chondroitinase ABC and heparinase I, II, III digestion (Figure 6);
4. GAGs could be detected inside of CHOꢀpgsA745 cells only in Ebselen βꢀDꢀxyloside treated
but not phenyl ꢀβꢀDꢀxyloside treated cells by fluorescenceꢀfacilitated confocal microscope
analysis (Figures 4&5), indicating Ebselen contributed to the unique property of intracellular
localization of the primed GAGs.
Conclusion
In current study, we conjugated the known cytotoxic compound Ebselen to Dꢀxylose and three
other monosaccharides to synthesize Ebselen βꢀDꢀglycosides. We found only Ebselen βꢀDꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 47
1
2
3
4
xyloside (8A) showed significant cytotoxicity. Most importantly, using the novel
5
6
immunostaining methods，we discovered for the first time that the aglycone and βꢀDꢀxyloside
7
8
9
primed glycosaminoglycan chains coꢀcontributed to the Ebselen βꢀDꢀxylosideꢀinduced
cytotoxicity. We also employed the stable isotope PMPꢀlabeling plus LCꢀMS analytical approach
to quantitatively compare the HS and CS disaccharide structures detected in the xylosideꢀtreated
and untreated CHO pgsAꢀ745 cells. The methods and conclusions drawn from current study
should be insightful for the development of novel antitumor strategy by targeting GAGs based on
novel cytotoxic βꢀDꢀxylosides.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Chemistry.
Commercial reagents and solvents were purchased from Sinopharm Chemical Reagent Company
(Shanghai, PRC) and directly used without further purification. All reactions were carried out in
glassware magnetically stirred and monitored by analytical thinꢀlayer chromatography (TLC)
using precoated silicaꢀgel 60 F254 plates (E. Merck). All newly synthesized compounds were
characterized with 1H NMR, 13C NMR, and high resolution mass spectrometry. 1H NMR and 13C
NMR spectra were obtained on a Propulse 500 MHz spectrometer (Agilent) with
tetramethylsilane (Me4Si) as the internal standard, and chemical shifts were recorded in δ values.
Mass spectra were recorded on a QꢀTOF Global mass spectrometer. Analysis of sample purity
was performed on a Waters 2998 Photodiode Array Deteter HPLC system with a YMC Pack
ODSꢀA (5 ꢁm; 10 × 250 mm + guard column). HPLC conditions were as follows: solvent A =
water, solvent B = methanol, and flow rate = 1.0mL/min. Compounds were eluted with a
ACS Paragon Plus Environment

Page 17 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
gradient of A:B = 10:90 at 0 min to 100:0 at 25 min . Purity was determined via integration of
UV spectra at 254 nm, and all tested compounds have a purity of >95% (Table S2).
7
8
9
1. Synthesis of 2,3,4ꢀtriꢀOꢀacetylꢀ1ꢀdeoxyꢀβꢀDꢀ xylosaminitol, 3A. The hydroxyl groups of
xylopyranose was protected by acetyl, and 2A was synthesized through bromination and
azidation. Then 2A was reducte by H2 with the catalysis of Pd/C to give 3A as white powder.
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 other glycosyl donors: 2,3,4,6ꢀtetraꢀOꢀacetylꢀ1ꢀdeoxyꢀβꢀDꢀglucamine (3B), 2,3,4,6ꢀtetraꢀOꢀ
acetylꢀ1ꢀdeoxyꢀβꢀDꢀgalactosamine (3C) and 2,3,6,2', 3', 4', 6'ꢀheptakisꢀOꢀacetylꢀ1ꢀdeoxyꢀβꢀDꢀ
lactosamine (3D) were also synthesized through acetylation, bromination, azidation and
reduction from starting monosaccharides. The yield of glycosyl donors from starting
monosaccharides were 27.0% (3A), 28.3% (3B), 23.7% (3C), 21.5% (3D).
2. Synthesis of Ebselen βꢀDꢀglycocosides (4ꢀ7): Synthesis of 2, 2’ꢀbisselanyldibenzoic acid
(6). Anthranilic acid (4) was generated from phthalimide; 1,3ꢀdihydroꢀ1,3ꢀdioxoisoindole
through the reaction in sodium hydroxide and bromine with a yield of 85.5%. Selenium powder
(1.6 g) was mixed with cetyltrimethyl ammonium bromide (CTAB) and 8% sodium hydroxide
(in water, 10 mL). NaBH4 solution (in water, 4%, 3 mL) that contained NaOH (0.08 g) was
added dropꢀwisely into the selenium mixture in nitrogen atmosphere in an ice bath. The reaction
was stirred at room temperature for 1 h and then at 90 for 0.5 h to from Na2Se2. Water (30 mL)
and hydrochloric acid (7 mL) was added into anthranilic acid (4, 20 mmol), into which NaNO2
(1.8 g in 5 mL water) was added in an ice bath. The mixture was stirred for 1.5 h to give the
desired diazonium salt. Freshly prepared Na2Se2 was mixed with NaOH (2.8 g in 3 mL water),
followed by adding diazonium salt solution dropꢀwisely under ice bath, after which the reaction
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 47
1
2
3
4
was stirred at 40 for 2 h. The mixture was then filtered and 6 was precipitated by adding excess
5
6
HCl (yield 93.2%).
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
3. Synthesis of 2ꢀchloroselenobenzoyl chloride (7). 2, 2’ꢀbisselanyldibenzoic acid (6) was
mixed with SOCl2 (30 mL) and heated to 80 for 4 h. The excessive SOCl2 was removed by
evaporation under vacuum, while the residue was extracted three times with anhydrous nꢀhexane.
The organic phases were combined, after which the solvent was evaporated under vacuum to
give 7 as yellow solid (10.1 mmol, yield 54.3%).
4. Synthesis of 2ꢀ(2,3,4ꢀtriꢀOꢀacetylꢀ1ꢀdeoxyꢀβꢀDꢀxylopyranosyl)ꢀbenzo[d][1,2]selenazolꢀ
3(2H)ꢀone (8A). 2,3,4ꢀtriꢀOꢀacetylꢀ1ꢀdeoxyꢀβꢀDꢀ xylosaminitol (3A, 4 mmol) was dissolved in
dry THF (40 mL) by stirring. 2ꢀchloroselenobenzoyl chloride (7, 4 mmol) and dry Et3N (8
mmol) were dissolved in dry THF (20 mL), respectively. Et3N solutions and 7 were
simultaneously added into 3A drop wisely cooled in an ice bath at the same rate, which took
approximately 30 min. The reaction was warmed to room temperature followed by overnight
stirring. The reaction mixture was filtered and then concentrated, followed by a silicon column to
give the final product 8A, a faint yellow powder (1.5 mmol, yield 36.7%, mp: 182ꢀ183 ).
1H NMR (500 MHz, DMSOꢀd6): δ 8.05ꢀ8.04 (d, J = 8.1 Hz, 1H, ArꢀH). 7.86ꢀ7,84 (d, J = 7.7 Hz,
1H, ArꢀH), 7.64ꢀ7.61 (t, J = 7.6 Hz, 1H, ArꢀH), 7.43ꢀ7.40 (t, J = 7.5 Hz, 1H, ArꢀH), 5.92ꢀ5.90 (d,
J = 9.0 Hz, 1H, C1ꢀH), 5.45 (t, J = 9.5 Hz, 1H, XylꢀH), 5.18 (t, J = 9.2 Hz, 1H, XylꢀH), 5.05 (td,
J = 10.2, 5.6 Hz, 1H, XylꢀH), 4.02 (dd, J = 11.1, 5.6 Hz, 1H, XylꢀH), 3.76 (t, J = 10.9 Hz, 1H,
XylꢀH), 2.02 (s, 3H, AcꢀH), 1.98 (s, 3H, AcꢀH), 1.81 (s, 3H, AcꢀH).
ACS Paragon Plus Environment

Page 19 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
13C NMR (126 MHz, DMSOꢀd6): δ 170.23, 170.16, 169.39, 167.69, 140.62, 132.92, 128.34,
127.92, 126.69, 126.54, 81.14, 73.07, 72.18, 68.99, 64.59, 21.15, 21.00, 20.80.
6
7
8
9
ESI MS m/z: 458.0339 [M + H]+, 480.0153 [M + Na]+.
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
Compound 8B, 8C and 8D were prepared according to the procedure described for the synthesis
of 8A by starting from 7 (4 mmol) and Et3N (8 mmol) added into corresponding glycosyl donors
3B, 3C or 3D (4 mmol) in dry THF (40 mL), The yield was 42.5% (8B), 39.3% (8C), 31.8%
(8D), respectively.
8B: 1H NMR (500 MHz, DMSOꢀd6): δ 8.05 (d, J = 8.1 Hz, 1H, ArꢀH), 7.85 (d, J = 7.3 Hz, 1H,
ArꢀH), 7.63 (t, J = 7.6 Hz, 1H, ArꢀH), 7.42 (t, J = 7.4 Hz, 1H, ArꢀH), 6.05 (d, J = 9.0 Hz, 1H,
C1ꢀH), 5.52 (t, J = 9.5 Hz, 1H, GlcꢀH), 5.17 (t, J = 9.3 Hz, 1H, GlcꢀH), 5.10 (t, J = 9.8 Hz, 1H,
GlcꢀH), 4.29 (m, 1H, GlcꢀH), 4.10 (dd, J = 6.9, 3.7 Hz, 2H, GlcꢀH), 2.02 (s, 6H, AcꢀH), 1.95 (s,
3H, AcꢀH), 1.82 (s, 3H, AcꢀH).
13C NMR (126 MHz, DMSOꢀd6): δ 170.51, 169.97, 169.88, 169.28, 167.48, 140.48, 132.78,
128.17, 127.81, 126.55, 126.41, 80.12, 73.25, 73.07, 71.97, 68.33, 62.35, 21.04, 20.93, 20.78,
20.65.
ESI MS m/z: 530.0548 [M + H]+, 552.0361 [M + Na]+.
8C: 1H NMR (500 MHz, DMSOꢀd6): δ 8.02 (d, J = 8.1 Hz, 1H, ArꢀH), 7.85 (d, J = 7.0 Hz, 1H,
ArꢀH), 7.64 (t, J = 7.6 Hz, 1H, ArꢀH), 7.43 (t, J = 7.8 Hz, 1H, ArꢀH), 6.00 (d, J = 8.8 Hz, 1H,
C1ꢀH), 5.44 (dd, J = 10.1, 3.4 Hz, 1H, GalꢀH), 5.35 (d, J = 2.9 Hz, 1H, GalꢀH), 5.31 – 5.26 (m,
1H, GalꢀH), 4.53 (t, J = 6.3 Hz, 1H, GalꢀH), 4.11 (dd, J = 11.4, 5.6 Hz, 1H, GalꢀH), 4.00 (dd, J =
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 47
1
2
3
4
5
6
11.4, 7.0 Hz, 1H, GalꢀH), 2.18 (s, 3H, AcꢀH), 1.99 (s, 3H, AcꢀH), 1.93 (s, 3H, AcꢀH), 1.84 (s,
3H, AcꢀH).
7
8
9
13C NMR (126 MHz, DMSOꢀd6): δ 170.39, 170.36, 169.93, 169.38, 167.33, 140.26, 132.81,
128.19, 127.83, 126.46, 126.45, 80.60, 72.50, 71.19, 69.68, 68.00, 62.07, 21.01, 20.93, 20.86,
20.75.
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
ESI MS m/z: 530.0551 [M + H]+, 552.0364 [M + Na]+.
8D: 1H NMR (CDCl3): δ 8.05ꢀ7.44 (m, 4H, ArꢀH), 5.91 (d, 1H, J = 8.7 Hz, C1ꢀH), 5.39 (m, 2H,
LacꢀH), 5.15 (m, 2H, LacꢀH), 4.99 (dd, 1H, J = 10.6 Hz, LacꢀH), 4.53 (d, 1H, J = 8.3 Hz , Lacꢀ
H), 4.51 (d, 1H, J = 12.4Hz, LacꢀH), 4.17ꢀ4,08 (m, 3H, LacꢀH), 3.91ꢀ3.88 (m, 3H, LacꢀH), 2.17ꢀ
1.92 (s, 21H, AcꢀH).
13C NMR (CDCl3): δ 170.4, 170.3, 170.2, 170.1, 169.7, 169.5, 169.0, 167.6, 138.6,132.9, 129.1,
126.4, 126.3, 124.2, 101.1, 80.6, 75.9, 75.4, 73.1, 71.7, 70.9, 70.8, 69.0,66.6, 61.8, 60.8, 20.9,
20.8, 20.7, 20.7, 20.6, 20.6, 20.5.
ESI MS m/z: 818.1397 [M + H]+.
Biology.
1. Cell Culture. The GAGꢀdeficient Chinese hamster ovary cells (CHO) mutants pgsAꢀ745
(xylosyltransferaseꢀdeficient) was a gift of Professor Jeffrey D. Esko.33 CHO pgsAꢀ745 cells
were grown in Fꢀ12 medium supplemented with 5% fetal bovine serum (Hyclone, Salt Lake
City, UT), 100 U/mL of penicillin, 0.1 mg/mL of streptomycin and sodium sulfate (100 mg/L) at
ACS Paragon Plus Environment

Page 21 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
37°C under an atmosphere of 5% CO2 in air and 100% relative humidity. Cells were passaged
with trypsin every 3– 4 days and revived periodically from frozen stocks.
7
8
9
Human lung cancer cell lines A549 and H1299 were obtained from the Chinese Academy of
Sciences (Shanghai, China). They were maintained in RPMI 1640 media supplemented with 5%
FBS, 100 U/mL of penicillin, and 0.1 mg/mL of streptomycin at 37°C with 5% CO2. Human
colon cancer cell lines HCT 116 and HT 29 were obtained from the Chinese Academy of
Sciences (Shanghai, China). They were maintained in 5A media supplemented with 5% FBS,
100 U/mL of penicillin, and 0.1 mg/mL of streptomycin at 37°C with 5% CO2.48 Cells were
passaged with trypsin every 3– 4 days and revived periodically from frozen stocks.
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. Cell Proliferation Assay. A549, H1299, HCT116 or HT29 (2000 cells/well) cells were
seeded in 96ꢀwells plates. After 24 h incubation, the cells were treated with 200 ꢁL complete
media containing serial Ebselen, 8AꢀD dilutions or 0.1% DMSO as negative control. After 48 h
treatments, each well was added 20 ꢁL of resazurin (2 mg/mL). The fluorescent signal was
monitored using 544 nm excitation and 595 nm emission wavelengths by Spectramax M5 plate
reader (Molecular Devices).49 The assay was performed in triplicate for each 8A treatment along
with control. The same experiment was repeated twice.
Cell Cycle Analysis. HCT116 cells (6×104 cells/well, 2 mL/well) were seeded in 6ꢀwell plates.
After 24 h, cells were treated with different concentrations of Ebselen, 8AꢀD or 0.1% DMSO in
2 mL of complete media. After another 24 h, media were collected and combined with adherent
cells that were detached by brief trypsinization. Cell pellets were resuspended in 70% ethanol at
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 47
1
2
3
4
5
6
7
8
4°C overnight. After centrifugation, the supernatant was removed and cells were incubated with
0.5 mL Propidum iodide (PI)/RNase Staining Buffer (BD Biosciences) for 15 min at room
temperature. Cell cycle was analyzed using the Beckman cell analyzer FC500ꢀmpl.50
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
3. Detection of Apoptosis. Apoptotic cells were quantified by Annexin V/propidium iodide (PI)
double staining assay kit (BD Biosciences, San Diego CA, USA). Briefly, HCT116 cells were
treated with or without Ebselen and 8AꢀD in cell culture media containing 0.1% DMSO and
harvested as described for the cell cycle analysis. After washing with binding buffer, cells were
suspended in 100 ꢁL binding buffer containing Annexin V and PI, and incubated for 15 min at
room temperature. The apoptotic cells were identified using the Beckman cell analyzer FC500ꢀ
mpl.
4. Confocal Microscopic Analysis of cellular GAGs by Immunostaining. For confocal
imaging, CHO pgsAꢀ745 cells were seeded on glass coverslips in complete cell culture media
overnight before treated with phenylꢀβꢀDꢀxyloside or 8A at the concentration of 10 ꢁM with
0.1% DMSO in cell culture media for 24 h at 37 °C. After that, the media were removed, and the
cells were fixed with 4% formaldehyde in complete media for 15 min. The fixed cells were
washed twice and incubated in PBS (8 g/L NaCl, 0.2 g/L KCl, 1.44 g/L Na2HPO4, 0.24 g/L
KH2PO4), 1 mU chondroitinase ABC (BIOKANGTAI, PRC) or 1 mU heparinase , ,
(BIOKANGTAI, PRC) respectively for 20 min at 37°C. After washing with PBS, the cells were
incubated in 0.5% primary monoclonal antibody 2B6 (AMSBIO, UK) for CS or monoclonal
antibody 3G10 (AMSBIO, UK) for HS overnight at 4 followed by incubated with 0.1% Alexa
ACS Paragon Plus Environment

Page 23 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Fluor 488 labeled secondary antibodies (Beyotime, PRC). The nucleuses were stained with DAPI
for 15 min. Doubleꢀlabeled cells were preserved in PBS at room temperature and observed and
photographed at 459 nm and 488 nm excitation and emission wavelengths by Laser Scanning
Confocal Microscope (Zeiss LSM 710, GER).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5. Detection of GAGs with Immunostaining by Flow Cytometry. CHO pgsAꢀ745 cells were
seeded on glass coverslips in complete cell culture media overnight before treated with phenylꢀβꢀ
Dꢀxyloside or 8A at the concentration of 10 ꢁM with 0.1% DMSO in cell culture media for 24 h
at 37 °C. After that, the media were removed, and the cells were harvested by brief trypsinization
and fixed with 4% formaldehyde in complete media for 15 min. The fixed cells were washed
twice and incubated in PBS, 1mU chondroitinase ABC (BIOKANGTAI, PRC) or 1mU
heparinase , ,
(BIOKANGTAI, PRC), respectively, for 20 min at 37°C. After washing
with PBS, the cells were incubated in 0.5% primary monoclonal antibody 2B6 (AMSBIO, UK)
for CS or monoclonal antibody 3G10 (AMSBIO, UK) for HS overnight at 4 followed by
incubated with 0.1% Alexa Fluor 488 labeled secondary antibodies (Beyotime, PRC). After 15
min of incubation, the cells were washed twice with 1 mL of PBS and reꢀsuspended in 100 ꢁL of
PBS. Flow cytometry was performed with Beckman cell analyzer FC500ꢀmpl.
6. Detection of FGF7/GAGs/FGFR2B Ternary Complex Formation by Confocal Analysis.
CHO pgsAꢀ745 cells were seeded on glass coverslips in complete cell culture media overnight
before being treated with phenylꢀβꢀDꢀxyloside or 8A at the concentration of 10 ꢁM with 0.1%
DMSO in cell culture media for 24 h at 37 °C. After that, the media were removed, and the cells
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 47
1
2
3
4
5
6
7
8
9
were fixed with 4% formaldehyde in complete media for 15 min. The fixed cells were washed
twice and incubated in PBS, or sequentially in FGFR2β/Fc, FGF7 and Protein AꢀAlexa Fluor
647. FGFR2BꢀFc is a soluble protein containing extracellular domain of FGFR2B and human Fc
domain of IgG1 (Pro100ꢀLys333) that can be recognized by florescenceꢀlabeled protein A or
florescenceꢀlabeled antiꢀhuman IgG antibody. The fluorescenceꢀtagged FGFR2β/Fc was
observed and photographed at 665 nm excitation wavelength by Laser Scanning Confocal
Microscope (Zeiss LSM 710, GER).
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
Glycochemistry.
1. Purification of GAGs. CHO pgsAꢀ745 cells were seeded on 15 cm diameter tissue culture
dishes, after 24 h, cells were treated with phenylꢀβꢀDꢀxyloside or 8A at the concentration of 10
ꢁM in cell culture media containing 0.1% DMSO for 48 h at 37 . Cells were lysed in 100 mM
sodium hydroxide for 8 h at 4℃. After neutralization, the amount of protein was determined by
BCA Protein Assay Kit,51 and each sample was digested with 0.83 g/L engineering protease
(ChongQingTianShi, PRC) in pH 6.0 for another 8 h. GAGs were isolated by anion exchange
resin and precipitated in 80% alcohol.
2. Monoꢀ and Disaccharide Analysis. For monosaccharide analysis, each purified GAG sample
was hydrolyzed in 500 ꢁL 6M hydrochloric acid sealed in an ampoule at 100°C for 3 h to
generate respective monosaccharides. The hydrolysates were tagged with 1ꢀphenylꢀ3ꢀmethylꢀ5ꢀ
pyrazolone (PMP) (Sigma) for 1 h at 70°C. The labeled samples are analyzed by HPLC (Agilent)
with ZORBAX SBꢀC18 column. Labeled monosaccharide samples were eluted by using 83% 0.1
ACS Paragon Plus Environment

Page 25 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
M phosphate buffer (13.6 g KH2PO4, 1.8 g NaOH, PH 6.7) and 17% acetonitrile. The UV signal
of PMP labeled monosaccharide were detected at 245 nm. For disaccharide analysis, each
purified GAGs sample was digested by 1mU chondroitinase ABC (BIOKANGTAI, PRC) or
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
1mU heparinase , ,
(BIOKANGTAI, PRC) in a final volume of 50 ꢁL in 5 mM NH4Ac pH
7.0 at 37°C for 24 h. The digested GAGs were labeled by DꢀPMP in ammonium hydroxide, in
the other hand, the phenylꢀβꢀDꢀxyloside or 8A treated ones were labeled by PMP. In order to
make quantitative analysis, a portion of GAGs corresponding to 1 mg cellular total proteins were
measured out from every PMP labeled xylosideꢀtreated samples and mixed with D5ꢀPMP labeled
control samples corresponding to the same amount of cellular proteins. The mixed differentially
labeled GAGs were made up to constant volume and monitored with LCꢀMS (Thermo) equipped
with Agilent 1200 infinity capillary liquid chromatography system with a ZORBAX Stable Bond
chromatogram column (5 ꢁm, 0.5 x 250 mm). Solvent A was 10 mM ammonium acetate (pH
5.0); Solvent B was 100% acetonitrile. For HS analysis, the gradient used was 17ꢀ22% B in 60
min and 23% B for another 15 min and equilibrated with 17% B. For CS analysis, the gradient
used was 13% B in 30 min, 20% B for 10 min and equilibrated with 13% B. The flow rate was
10 ꢁL/min. The UV absorbance of PMP labeled products was monitored at 245 nm by using an
online UV detector and the total ion chromatogram (TIC) was collected using a Thermo LTQꢀ
XL mass spectrometer operating in negative ion mode by scanning the m/z ranged from 300–
2000. The data were analyzed in Xcalibur software (Thermo).
Author Information
Corresponding Author
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 47
1
2
3
4
5
6
*Eꢀmail: T.J. (email: jiangtao@ouc.edu.cn; Tel: +86 532 82033054) or to L.Z. (email:
zhanglj@qduhospital.cn; Tel: +86 532 82031615).
7
8
9
Author Contributions
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 manuscript was written through contributions of all authors. All authors have given
approval to the final version of the manuscript.
#Y.T. and S.Z. contributed equally to this work.
Notes
The authors declare no competing financial interest.
Acknowledgements
We thank Professor Jeffrey D. Esko of UCSD for providing us with phenyl ꢀβꢀDꢀxyloside and
CHO pgsAꢀ745 cells. This work was supported by the National Natural Science Foundation of
China (Grants 21171154 and 81672585), Key Technology Fund of Shandong Province (Grant
2016ZDJS07A07), the Taishan Scholar Fellowship, and the “Double First Class fund” of
Shandong Province in China to L.Z.
ACS Paragon Plus Environment

Page 27 of 47
Journal of Medicinal Chemistry
1
2
3
4
Abbreviations and Acronyms
5
6
7
8
9
GAG
HS
Glycosaminoglycan
Heparan Sulfate
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
CS
Chondroitin Sulfate
Glucosamine
GlcN
GlcNAc
GalNAc
GlcA
IdoA
Gal
N-Acetylglucosamine
N-Acetylgalactosamine
Glucuronic acid
Iduronic acid
Galactose
Xyl
Xylose
DUA
FGF
4,5-unsaturated hexuronic acid
Fibroblast Growth Factor
Fibroblast Growth Factor Receptor
1-phenyl-3-methyl-5-pyrazolone
1-(phenyl-2, 3, 4, 5, 6-deuterated)-3-methyl-5-pyrazolone
Dimethyl sulfoxide
FGFR
PMP
5D-PMP
DMSO
DAPI
FITC
PI
4',6-diamidino-2-phenylindole
Fluorescein Isothiocyanate
Propidium Iodide
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 47
1
2
3
4
5
6
7
8
9
REFERENCE
1. Fritz, T. A.; Esko, J. D. Xyloside priming of glycosaminoglycan biosynthesis and
inhibition of proteoglycan assembly. Proteoglycan protocols 2001, 317ꢀ323.
2. Lindahl, U. Heparan sulfateꢀprotein interactionsꢀa concept for drug design? Thrombosis
and haemostasis 2007, 98 (1), 109ꢀ115.
3. Gallagher, J. Fell Muir Lecture: Heparan sulphate and the art of cell regulation: a
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
polymer chain conducts the protein orchestra. International journal of experimental pathology
2015, 96 (4), 203ꢀ231.
4.
Mizumoto, S.; Yamada, S.; Sugahara, K. Molecular interactions between chondroitin–
dermatan sulfate and growth factors/receptors/matrix proteins. Current opinion in structural
biology 2015, 34, 35ꢀ42.
5.
Izumikawa, T.; Kanagawa, N.; Watamoto, Y.; Okada, M.; Saeki, M.; Sakano, M.;
Sugahara, K.; Sugihara, K.; Asano, M.; Kitagawa, H. Impairment of embryonic cell division and
glycosaminoglycan biosynthesis in glucuronyltransferaseꢀIꢀdeficient mice. Journal of biological
chemistry 2010, 285 (16), 12190ꢀ12196.
6.
Mizuguchi, S.; Uyama, T.; Kitagawa, H.; Nomura, K. H.; Dejima, K.; GengyoꢀAndo, K.;
Mitani, S.; Sugahara, K.; Nomura, K. Chondroitin proteoglycans are involved in cell division of
Caenorhabditis elegans. Nature 2003, 423 (6938), 443ꢀ448.
7.
in molecular biology and translational science 2010, 93, 1ꢀ17.
8. Taylor, K. R.; Gallo, R. L. Glycosaminoglycans and their proteoglycans: hostꢀassociated
Zhang, L. Glycosaminoglycan (GAG) biosynthesis and GAGꢀbinding proteins. Progress
molecular patterns for initiation and modulation of inflammation. The FASEB journal 2006, 20
(1), 9ꢀ22.
ACS Paragon Plus Environment

Page 29 of 47
Journal of Medicinal Chemistry
1
2
3
4
9.
Vallen, M. J.; van der Steen, S. C.; van Tilborg, A. A.; Massuger, L. F.; van Kuppevelt,
5
6
T. H. Sulfated sugars in the extracellular matrix orchestrate ovarian cancer development:‘when
7
8
sweet turns sour’. Gynecologic oncology 2014, 135 (2), 371ꢀ381.
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
10.
research reviews 2017, 37 (5), 1051ꢀ1094.
11. Bishop, J. R.; Schuksz, M.; Esko, J. D. Heparan sulphate proteoglycans fineꢀtune
mammalian physiology. Nature 2007, 446 (7139), 1030ꢀ1037.
12. Toomey, J.; Abboud, M.; Valocik, R.; Koster, P.; Burns Kurtis, C.; Pillarisetti, K.;
Ghiselli, G. Drug Mediated Regulation of Glycosaminoglycan Biosynthesis. Medicinal
Danoff, T.; Erhardt, J. A comparison of the β D xyloside, odiparcil, to warfarin in a rat model
of venous thrombosis. Journal of thrombosis and haemostasis 2006, 4 (9), 1989ꢀ1996.
13.
Farach, M.; Tang, J.; Decker, G.; Carson, D. Differential effects of pꢀnitrophenylꢀDꢀ
xylosides on mouse blastocysts and uterine epithelial cells. Biology of reproduction 1988, 39 (2),
443ꢀ455.
14.
Margolis, R.; Goossen, B.; Tekotte, H.; Hilgenberg, L.; Margolis, R. Effects of betaꢀ
xylosides on proteoglycan biosynthesis and morphology of PC12 pheochromocytoma cells and
primary cultures of rat cerebellum. Journal of cell science 1991, 99 (2), 237ꢀ246.
15.
Raman, K.; Ninomiya, M.; Nguyen, T. K. N.; Tsuzuki, Y.; Koketsu, M.; Kuberan, B.
Novel glycosaminoglycan biosynthetic inhibitors affect tumorꢀassociated angiogenesis.
Biochemical and biophysical research communications 2011, 404 (1), 86ꢀ89.
16.
Belting, M.; Borsig, L.; Fuster, M. M.; Brown, J. R.; Persson, L.; Fransson, L.ꢀÅ.; Esko,
J. D. Tumor attenuation by combined heparan sulfate and polyamine depletion. Proceedings of
the national academy of sciences 2002, 99 (1), 371ꢀ376.
ACS Paragon Plus Environment