Sp3- dependent transcription in articular chondrocytes

Accepted Manuscript Fibronectin fragment inhibits xylosyltransferase-1 expression by regulating Sp1/Sp3dependent transcription in articular chondrocytes Hyun Sook Hwang, Mi Hyun Lee, Hyun Ah Kim PII:

S1063-4584(19)30031-7

DOI:

https://doi.org/10.1016/j.joca.2019.01.006

Reference:

YJOCA 4393

To appear in:

Osteoarthritis and Cartilage

Received Date: 26 May 2018 Revised Date:

9 January 2019

Accepted Date: 15 January 2019

Please cite this article as: Hwang HS, Lee MH, Kim HA, Fibronectin fragment inhibits xylosyltransferase-1 expression by regulating Sp1/Sp3-dependent transcription in articular chondrocytes, Osteoarthritis and Cartilage, https://doi.org/10.1016/j.joca.2019.01.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Fibronectin fragment inhibits xylosyltransferase-1 expression by regulating

2

Sp1/Sp3- dependent transcription in articular chondrocytes

3

Hyun Sook Hwang†‡, Mi Hyun Lee †‡, Hyun Ah Kim†‡*

5

†Division of rheumatology, Department of Internal Medicine, Hallym University Sacred Heart

6

Hospital, Kyunggi, 431-070, Korea; ‡Institute for Skeletal Aging, Hallym University, Chunchon

7

200-702, Korea

SC

RI PT

4

9

M AN U

8

Each author’s email address:

11

Hyun Sook Hwang: [email protected]

12

Mi Hyun Lee: [email protected]

13

Hyun Ah Kim: [email protected]

14

TE D

10

*Corresponding author:

16

Name: Hyun Ah Kim

17

Address: Division of rheumatology, Department of internal medicine, Hallym University

18

Sacred Heart Hospital, 896, Pyungchon, Anyang, Kyunggi, 431-070, Korea

19

Phone: 82-31-380-1826

20

Fax: 82-31-381-8812

21

E-mail: [email protected]

AC C

EP

15

1

ACCEPTED MANUSCRIPT ABSTRACT

2

Objective

3

We investigated the effects of 29-kDa amino-terminal fibronectin fragment (29-kDa FN-f) on

4

xylosyltransferase-1 (XT-1), an essential anabolic enzyme that catalyzes the initial and rate-

5

determining step in glycosaminoglycan chain synthesis, in human primary chondrocytes.

RI PT

1

6

Methods

8

Proteoglycan and XT-1 expression in cartilage tissue was analyzed using safranin O staining and

9

immunohistochemistry. The effects of 29-kDa FN-f on XT-1 expression and its relevant

10

signaling pathway were analyzed by quantitative real-time-PCR, immunoblotting, chromatin

11

immunoprecipitation, and immunoprecipitation assays. The receptors for 29-kDa FN-f were

12

investigated using small interference RNA and blocking antibodies.

M AN U

SC

7

13

Results

15

The expression of XT-1 was significantly lower in human osteoarthritis cartilage than in normal

16

cartilage. Intra-articular injection of 29-kDa FN-f reduced proteoglycan levels and XT-1

17

expression in murine cartilage. In addition, in 29-kDa FN-f-treated cells, XT-1 expression was

18

significantly suppressed at both the mRNA and protein levels, modulated by the transcription

19

factors specificity protein 1 (Sp1), Sp3, and activator protein 1 (AP-1). The 29-kDa FN-f

20

suppressed the binding of Sp1 to the promoter region of XT-1 and enhanced the binding of Sp3

21

and AP-1. Inhibition of mitogen-activated protein kinase and nuclear factor kappa B signaling

22

pathways restored the 29-kDa FN-f-inhibited XT-1 expression, along with the altered expression

23

of Sp1 and Sp3. Blockading toll-like receptor 2 (TLR-2) and integrin α5β1 via siRNA and

AC C

EP

TE D

14

2

ACCEPTED MANUSCRIPT 1

blocking antibodies revealed that the effects of 29-kDa FN-f on XT-1 expression were mediated

2

through the TLR-2 and integrin α5β1 signaling pathways.

3

Conclusion

5

These results demonstrate that 29-kDa FN-f negatively affects cartilage anabolism by regulating

6

glycosaminoglycan formation through XT-1.

RI PT

4

SC

7

Keywords

9

Xylosyltransferase-1, fibronectin fragments, osteoarthritis, cartilage matrix, specific protein 1,

10

M AN U

8

specific protein 3

11 12

TE D

15

EP

14

Running headline: XT-1 expression inhibited by 29-kDa FN-f

AC C

13

3

ACCEPTED MANUSCRIPT INTRODUCTION

2

Articular cartilage damage is a key event leading to joint deformity in various arthritides,

3

including rheumatoid arthritis (RA), osteoarthritis (OA), and septic arthritis1. The dynamic

4

equilibrium between the anabolic and catabolic activities of chondrocytes is of utmost

5

importance for the maintenance of the cartilage extracellular matrix (ECM). Although the

6

role of cartilage matrix catabolism in the pathogenesis of OA has been widely investigated,

7

the role of matrix anabolism in OA is not well understood.

SC

RI PT

1

Proteoglycans (PG) are major constituents of cartilage ECM, which, by covalently

9

attaching highly charged glycosaminoglycan (GAG) chain such as chondroitin sulfate (CS),

10

heparan sulfate (HS), and keratan sulfate, confer compressive integrity to cartilage2,3. The

11

biosynthesis of GAGs is accomplished through multiple steps, beginning with the formation

12

of a core protein linkage tetrasaccharide that functions as a primer for GAG chain elongation.

13

Xylosyltransferase-1 (XYLT1; XT-1; EC 2.4.2.26) catalyzes the first and rate-limiting step in

14

the biosynthesis of GAGs by transferring xylose to specific serine residues in PG core

15

proteins4.

TE D

M AN U

8

Previous reports have demonstrated that XT-1 functions as a key regulator of the cartilage

17

repair process5. XT-1 expression was lower in damaged regions of OA cartilage than in

18

unaffected regions, and modulation of XT-1 expression using RNA interference and

19

overexpression vectors revealed that XT-1 was closely associated with GAG synthesis6. In

20

vivo animal studies have shown that XT-1 expression was significantly reduced in rat

21

cartilage in antigen-induced arthritis, and XT-1 knockdown prevented cartilage repair by

22

decreasing PG synthesis5. In addition, XT-1 was identified as a regulator of chondrocyte

23

maturation and endochondral ossification in a recessive dwarf mouse mutant7. XT-1

24

expression was downregulated by the catabolic cytokine interleukin-1 beta (IL-1β) in human

AC C

EP

16

4

ACCEPTED MANUSCRIPT 1

primary chondrocytes and cartilage explants, and upregulated by the anabolic cytokine

2

transforming growth factor beta 1 (TGF-β1) in cartilage explants3,6. Chondrocytes express degradative enzymes and catabolic mediators in response to matrix

4

degradation products as well as proinflammatory cytokines and mechanical loading8.

5

Fibronectin fragments (FN-fs) are generated through the proteolysis of fibronectin (FN), a

6

component of articular cartilage matrix, and are detected at high levels in OA cartilage and

7

synovial fluids9,10. FN-fs, including 29-kDa, 50-kDa, and 140-kDa FN-fs, induce cartilage

8

catabolism by increasing the expression of matrix metalloproteinases (MMPs), and intra-

9

articular injection of 29-kDa FN-f into rabbit knee joints caused cartilage PG loss10-12.

10

Previously, we reported that the 29-kDa amino-terminal FN-f (29-kDa FN-f) stimulated

11

catabolic responses through the MyD88-dependent toll-like receptor 2 (TLR-2) signaling

12

pathway in primary human articular chondrocytes, suggesting that it may function as a

13

mediator of damage-associated molecular pattern (DAMP) and induce a vicious cycle

14

accelerating cartilage matrix damage1. It is unknown whether FN-fs also regulate anabolic

15

signaling in chondrocytes.

TE D

M AN U

SC

RI PT

3

In this study, we investigated the influence of 29-kDa FN-f on cartilage matrix anabolic

17

response by examining the transcriptional regulation of XT-1 and GAG synthesis in human

18

articular chondrocytes.

20

AC C

19

EP

16

21

MATERIALS AND METHODS

22

Materials, cartilage collection, primary chondrocyte culture, cartilage explants culture, and

23

surgically induced OA and intra-articular injection of 29-kDa FN-f in mice are described in

24

the supplementary information. 5

ACCEPTED MANUSCRIPT 1

Histology and immunohistochemistry (IHC) analysis

3

Safranin O staining and immunohistochemistry analysis of normal and OA human cartilage

4

and mouse joint cartilage tissues were performed as previously described13.

RI PT

2

5

Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR)

7

Total RNA was isolated from chondrocytes and cartilage tissues using TRIzol reagent. XT-1,

8

Sp1, and Sp3 gene expression were quantified as previously described13. The primer

9

sequences are shown in Table S1.

M AN U

SC

6

10

Transfection with siRNA and overexpression vectors

12

Transfections with siRNA and plasmids were performed using Lipofectamine 2000

13

(Invitrogen, Carlsbad, CA, USA) as previously described1. The siRNA sequences are shown

14

in Table S2.

TE D

11

Overexpression plasmids for Sp1 and Sp3 were prepared. Briefly, PCR products amplified

16

using Sp1- and Sp3-specific primers containing HindIII (AAGCTT) and BamHI (GGATCC)

17

restriction sites were ligated into a pGEM-T Easy vector and subcloned into a pcDNA3.1-

18

myc-His vector. Purified Sp1 and Sp3 overexpression vectors were transfected into

19

chondrocytes using Lipofectamine 2000. The expression of Sp1 and Sp3 was detected using

20

western blot analysis.

AC C

21

EP

15

22

Western blot analysis

23

Cell lysates were prepared with radioimmunoprecipitation assay (RIPA) lysis buffer and

24

immunoblotting was performed as described previously1. 6

ACCEPTED MANUSCRIPT 1

Chromatin immunoprecipitation (ChIP) and immunoprecipitation (IP) assays

3

ChIP was performed according to the manufacturer’s instructions (EpiQuik, Farmingdale, NY,

4

USA). Briefly, chondrocytes were cross-linked with 1% formaldehyde for 10 min and the

5

cross-linked protein–DNA complexes were extracted from cells. The extracted genomic DNA

6

was sheared into smaller pieces by sonication and immunoprecipitated with specific

7

antibodies against immunoglobulin G (IgG), Sp1, Sp3, and c-jun. Cross-linked DNA was

8

released from the antibody-target protein-DNA complex and purified using Fast-Spin

9

columns. The levels of DNA bound to each protein were determined using PCR and agarose

SC

M AN U

10

RI PT

2

gel electrophoresis.

For IP, cells were lysed with RIPA lysis buffer containing protease inhibitors. Cell lysates

12

were centrifuged at 13,000 × g for 10 min at 4°C and the supernatants were collected. The

13

supernatant samples containing equal amounts of proteins were pre-cleared with protein A

14

agarose bead slurry for 4 h at 4°C on a rotating shaker. The pre-cleared samples (1 µg) were

15

incubated with specific antibodies against c-jun, Sp1, or IgG in the presence of protein A

16

agarose beads at 4°C overnight with gentle rotation. Proteins were eluted by boiling in 2x

17

SDS sample buffer. A portion of the lysates was used as an input control. Eluted protein

18

samples were subjected to western blot analysis as described above.

EP

AC C

19

TE D

11

20

Direct enzyme-linked immunosorbent assay (ELISA) to detect the binding of 29-kDa

21

FN-f to TLR-2

22

Round-bottom microtiter plates were coated with 100 µL recombinant human TLR-2 (1

23

µg/mL) overnight at 4°C and blocked with 300 µL BSA (2% w/v) for 2 h at room

24

temperature. 29-kDa FN-f (1, 5, 10, and 30 µg/mL) or IgG was added to TLR-2-coated wells 7

ACCEPTED MANUSCRIPT for 2 h at room temperature. Anti-FN antibody (100 µL/well, 1:1000 dilution) was added to

2

each well for 1 h at room temperature, followed by incubation with HRP-conjugated anti-

3

mouse IgG. To detect HRP activity, the wells were incubated with 1-step Turbo TMB Direct

4

ELISA reagent (Thermo Fisher Scientific, Rockford, IL, USA) for 30 min, and then stop

5

solution was added to each well. The absorbance was measured at 450 nm.

RI PT

1

6

Neutralization assay using neutralizing antibodies against TLR-2 and integrin α5β β1

8

Chondrocytes were pre-incubated for 2 h with antibodies against TLR-2 or integrin α5β1

9

(Merck Millipore, Billerica, MA, USA) and treated with 29-kDa FN-f for 14 h. The

10

expression levels of XT-1, Sp1, and Sp3 in anti-TLR-2- and anti-integrin α5β1-treated cells

11

were measured by qRT-PCR.

12

M AN U

SC

7

Statistical analysis

14

For in vivo studies, 4-6 mice in each group were used. For in vitro studies, individual

15

experiments were performed using chondrocytes from more than three different donors with

16

duplicate or triplicate per donor. Multi-level mixed effects models were used for statistical

17

analyses by using Statistical Analysis System 9.4 (Statistical Analysis System, Cary, NC,

18

USA). Data are expressed as the mean ± standard deviation (SD). A value of P < 0.05 was

19

considered statistically significant.

EP

AC C

20

TE D

13

21 22

RESULTS

23

XT-1 expression is significantly downregulated in OA cartilage

8

ACCEPTED MANUSCRIPT To investigate whether XT-1 and ACAN expression changes in OA, we compared XT-1 and

2

ACAN expression in normal and OA cartilage. The mRNA expression of XT-1 and ACAN were

3

significantly downregulated in OA cartilage compared with normal cartilage (Fig. 1A). To

4

further investigate XT-1 protein levels in cartilage tissues, we analyzed XT-1 levels using IHC.

5

ECM, as assessed by safranin O staining, and XT-1 expression levels were lower in human OA

6

cartilage than in normal cartilage (Fig. 1B). In addition, XT-1 expression and ECM levels were

7

reduced in DMM-induced OA murine knee-joint cartilage relative to those of sham mice (Fig.

8

1C), suggesting that reduced XT-1 protein expression may be closely related to OA

9

pathogenesis.

M AN U

SC

RI PT

1

10

29-kDa FN-f suppresses the expression of aggrecan (ACAN) and XT-1 in primary

12

chondrocytes and cartilage explant culture

13

We examined the basal expression of ACAN, a representative PG, and XT-1 in chondrocyte

14

monolayer cultures incubated in the presence or absence of FBS for 72 h. The mRNA

15

expression of XT-1 and ACAN oscillated during the 24 h of culture, peaking after 14 h of

16

culture (Fig. S1). Interestingly, at 72 h, XT-1 and ACAN expression tended to increase again

17

compared with baseline levels (Fig. S1). To determine whether 29-kDa FN-f modulates XT-1

18

and ACAN expression in primary chondrocytes, cells were stimulated with 29-kDa FN-f for 24

19

h. PCR analysis demonstrated that 29-kDa FN-f suppressed XT-1 expression at 6 h and at 14 h,

20

when XT-1 expression was maximally induced, but had no effect on XT-1 expression at 24 h

21

(Fig. 2A). Unlike its mRNA expression, XT-1 protein expression did not decrease at 24 h, and

22

treatment with 29-kDa FN-f suppressed the protein expression of XT-1 at 14 and 24 h (Fig. 2B).

23

In the case of ACAN, 29-kDa FN-f suppressed mRNA expression of ACAN throughout the 24-

24

h period (Fig. 2A). We examined whether ACAN has influence on XT-1 expression, and found

AC C

EP

TE D

11

9

ACCEPTED MANUSCRIPT out that regardless of the presence of 29-kDa FN-f, ACAN knockdown suppressed XT-1

2

expression (Fig. S2). Conversely, XT-1 knockdown led to ACAN suppression, suggesting that

3

ACAN and XT-1 reciprocally affect the expression of the other (Fig. S2). To further investigate

4

the effects of 29-kDa FN-f on the in vivo expression of XT-1 protein, 29-kDa FN-f (3 µM/10 µL)

5

was intra-articularly injected into the left knee joints of mice. The animals were sacrificed 7

6

days after injection. IHC analysis revealed that administration of 29-kDa FN-f decreased XT-1

7

expression in murine joint cartilage compared with PBS injection (Fig. 2C). Human cartilage

8

explants exposed to 29-kDa FN-f (300 nM) for 7 days showed ECM depletion and decreased

9

XT-1 expression (Fig. 2D) similar to the observations in 29-kDa FN-f-injected murine joint

10

cartilage (Fig. 2C). The results demonstrate that 29-kDa FN-f negatively regulated XT-1 and PG

11

expression in primary human chondrocytes and cartilage tissue.

M AN U

SC

RI PT

1

12

Transcription factors Sp1 and Sp3 transcriptionally regulate XT-1 expression

14

We analyzed the 5’-flanking region of the human XT-1 gene to identify transcription factors

15

that regulate XT-1 expression. Genome sequence analysis revealed three activator protein 1

16

(AP-1) and two Sp1 binding sites in the XT-1 promoter region3. In addition, Sp3 is a

17

transcription factor that competes for the same DNA region bound by Sp13. PCR analysis

18

demonstrated that Sp1 and Sp3 expression increased in the absence of 29-kDa FN-f after 14 h

19

of incubation, whereas 29-kDa FN-f treatment up-regulated Sp1 and Sp3 transcripts at 6 h,

20

followed by reduced Sp1 expression and increased Sp3 expression after 14 h of treatment

21

(Fig. 3A). Western blot results showed that 29-kDa FN-f treatment reduced Sp1 expression

22

and increased Sp3 expression at 14 h, and downregulated XT-1 expression (Fig. 3B). 29-kDa

23

FN-f exposure (at 6 h) did not affect XT-1 protein expression compared with control levels,

24

whereas its mRNA expression decreased (Fig. 2A and B). The results suggest that the

AC C

EP

TE D

13

10

ACCEPTED MANUSCRIPT 1

reduction in XT-1 expression was likely due to altered cellular levels of Sp1 and Sp3 in 29-

2

kDa FN-f-treated chondrocytes. Next, to clarify whether the alterations of Sp1 and Sp3 expression by 29-kDa FN-f affect XT-

4

1 expression, knockdown or overexpression of Sp1 and Sp3 was performed using Sp1 and Sp3

5

siRNA or Sp1 and Sp3 overexpression vectors. Whereas Sp1 knockdown further downregulated

6

XT-1, Sp1 overexpression slightly increased XT-1 expression in 29-kDa FN-f-treated

7

chondrocytes (Fig. 3C). On the other hand, Sp3 knockdown enhanced XT-1 expression, whereas

8

its overexpression downregulated XT-1 in 29-kDa FN-f-treated cells (Fig. 3D). Collectively,

9

these data demonstrate that 29-kDa FN-f regulates cellular levels of Sp1, a transcriptional

10

activator, and Sp3, a transcriptional repressor, which are involved in XT-1 expression,

11

suggesting that Sp1 and Sp3 may be transcriptional regulators of XT-1.

M AN U

SC

RI PT

3

12

AP-1 transcription factor modulates the expression of Sp1 and Sp3

14

The XT-1 promoter contains three sites that interact with the transcription factor AP-1 (c-

15

jun/c-fos)3. We examined whether 29-kDa FN-f affects AP-1 activity by examining the

16

phosphorylation of c-jun and c-fos. Western blot analysis of nuclear extract demonstrated that

17

29-kDa FN-f enhanced the nuclear levels of phospho-c-jun and phospho-c-fos (Fig. 3E),

18

indicating that 29-kDa FN-f activates the transcription factor AP-1. We also assessed whether

19

AP-1 is implicated in 29-kDa FN-regulated XT expression using 12-O-tetradecanoylphorbol-

20

13-acetate (TPA), an AP-1 activator, and Tanshinone IIA, an AP-1 inhibitor. AP-1 inhibition

21

by Tanshinone IIA recovered 29-kDa FN-f-suppressed XT-1 expression to the levels observed

22

in untreated cells, accompanied by increased Sp1 and decreased Sp3 expression (Fig. 3F). By

23

contrast, activation of AP-1 by TPA did not further decrease 29-kDa FN-f suppression of XT-

AC C

EP

TE D

13

11

ACCEPTED MANUSCRIPT 1

1. These results demonstrate that 29-kDa FN-f activates AP-1, modulating Sp1 and Sp3

2

expression and ultimately downregulating XT-1.

3

29-kDa FN-f modulates the binding of Sp1, Sp3, and AP-1 to the XT-1 promoter

5

We next investigated whether 29-kDa FN-f directly regulates the binding of Sp1, Sp3, and

6

AP-1 to the promoter region of XT-1 using a ChIP assay. The putative Sp1 and AP-1 binding

7

sites are shown in Figure 4A. Our ChIP data indicate that 29-kDa FN-f promoted the

8

recruitment of Sp3 and c-Jun, transcriptional repressors of the XT-1 gene, to the XT-1

9

promoter. By contrast, 29-kDa FN-f suppressed the binding of Sp1, a transcription activator,

10

to the XT-1 promoter (Fig. 4A). Therefore, the results indicate that 29-kDa FN-f modulated

11

the recruitment of transcriptional activators and repressors of the XT-1 gene to their

12

individual binding sites.

M AN U

SC

RI PT

4

c-Jun regulates a variety of genes through its associations with transcription factors such as

14

Sp1, ATF, and Smad14. To investigate whether the interaction between Sp1 and c-Jun is

15

mediated by 29-kDa FN-f, we performed IP using anti-c-Jun or -Sp1 antibodies. Binding

16

between c-Jun and Sp1 was observed in untreated control cells, whereas 29-kDa FN-f

17

treatment decreased the interaction between c-Jun and Sp1 (Fig. 4B), suggesting that 29-kDa

18

FN-f inhibits the formation of the c-Jun/Sp1 complex. Taken together, these results suggest

19

that 29-kDa FN-f regulates the binding of the transcription factors Sp1, Sp3, and AP-1 to the

20

XT-1 promoter, as well as the c-Jun/Sp1 complex.

EP

AC C

21

TE D

13

22

MAPK and NF-κB signaling pathways are responsible for 29-kDa FN-f-modulated XT-1

23

expression

12

ACCEPTED MANUSCRIPT To determine whether the MAPK and NF-κB signaling pathways are involved in XT-1 and

2

ACAN expression, chondrocytes were exposed to MAPK and NF-κB inhibitors 2 h prior to

3

29-kDa FN-f treatment. Pretreatment with JNK inhibitor SP203580, p38 inhibitor SB600125,

4

ERK1/2 inhibitor PD98059, and IκBα inhibitor Bay11-7082 recovered XT-1 expression

5

suppressed by 29-kDa FN-f (Fig. 5A). In addition, ACAN downregulation by 29-kDa FN-f

6

was reversed by pretreatment with MAPK and NF-κB inhibitors (Fig. 5A). MAPK and NF-

7

κB inhibitors also restored the expression levels of Sp1 and Sp3 altered by 29-kDa FN-f to

8

those of untreated cells (Fig. 5B). These results demonstrate that the MAPK and NF-κB

9

signaling pathways play an important role in the regulation of XT-1 transcription factors Sp1

SC

and Sp3, as well as in XT-1 expression.

11

M AN U

10

RI PT

1

TLR-2 and integrin α5β β1 are receptors for 29-kDa FN-f in human chondrocytes

13

We previously reported that TLR-2 knockdown suppressed 29-kDa FN-f-mediated catabolic

14

factor expression and signaling pathways in human primary chondrocytes1. Previous reports

15

have shown that FN and 120-kDa FN-f interact with and stimulate members of the integrin

16

family 15-17. To verify the direct binding of 29-kDa FN-f to TLR-2, the interaction of TLR-2

17

with various concentrations of 29-kDa FN-f was assessed using a modified ELISA method.

18

TLR-2 was incubated with 29-kDa FN-f to induce complex formation, then 29-kDa FN-f in

19

the complex was probed using ELISA with an anti-FN antibody. Our ELISA data showed that

20

the interaction of 29-kDa FN-f with TLR-2 increased in a dose-dependent manner, suggesting

21

direct binding (Fig. 6A). A slightly increased interaction in the IgG-treated sample used as a

22

negative control was considered non-specific binding (Fig. 6A). However, we could not

23

confirm that 29-kDa FN-f directly binds integrin α5β1 with ELISA because purified integrin

24

α5β1 is unavailable.

AC C

EP

TE D

12

13

ACCEPTED MANUSCRIPT We further investigated whether TLR-2 and integrin α5β1 are implicated in 29-kDa FN-f-

2

mediated XT-1 and ACAN expression using siRNA knockdown and blocking antibodies.

3

First, human chondrocytes were transfected with siTLR-2, si-integrin α5β1, and control

4

siRNA, followed by incubation with 29-kDa FN-f. We found that TLR-2 and integrin α5β1

5

knockdown recovered XT-1 and ACAN expression suppressed by 29-kDa FN-f (Fig. 6B).

6

This suggests that both TLR-2 and integrin α5β1 are primarily, or at least partially,

7

responsible for 29-kDa FN-f-mediated pathways in the regulation of XT-1 and ACAN

8

expression. Next, TLR-2 and integrin α5β1 receptors on chondrocytes were blocked using

9

antibodies, followed by treatment with 29-kDa FN-f. Although XT-1 and ACAN expression

10

were suppressed in response to 29-kDa FN-f, their expression was partially recovered by

11

treatment with TLR-2 or integrin α5β1 antibodies (Fig. 6C). Furthermore, a combination of

12

two antibodies synergistically restored XT-1 and ACAN expression to the levels observed in

13

untreated cells (Fig. 6C). Blockading these receptors with siRNA and blocking antibodies

14

also reversed the effects of 29-kDa FN-f on Sp1 and Sp3 expression (Fig. 6B and C). Taken

15

together, these results demonstrate that both TLR-2 and integrin α5β1 are responsible for 29-

16

kDa FN-f modulation of the transcription factors Sp1 and Sp3 as well as XT-1 expression.

EP

TE D

M AN U

SC

RI PT

1

18

AC C

17

19

DISCUSSION

20

XT-1, a crucial enzyme involved in GAG synthesis, was downregulated in human OA cartilage

21

and surgically induced OA mouse cartilage compared with normal cartilage and sham surgery

22

mouse cartilage, respectively. In addition, 29-kDa FN-f suppressed the expression of both XT-1

23

and ACAN in human articular chondrocytes. This study focused on the mechanism by which

24

29-kDa FN-f regulates the expression of XT-1, the enzyme regulating the rate-limiting step of 14

ACCEPTED MANUSCRIPT GAG synthesis in cartilage matrix in primary human chondrocytes. 29-kDa FN-f downregulated

2

XT-1 expression via TLR-2 and integrin α5β1 receptor-mediated signaling pathways. The

3

transcription factors Sp1, Sp3, and AP-1 played particularly important roles in 29-kDa FN-f-

4

suppression of XT-1.

RI PT

1

In the pathogenesis of OA, a variety of proteins found in OA synovial fluids activate

6

inflammatory and catabolic signaling pathways and act as DAMPs18-21. For example,

7

S100A8/S100A9 in synovial fluid in OA and RA modulates the expression of catabolic factors,

8

including MMPs and IL-6, and anabolic factors, including ACAN and type II collagen22.

9

Exposure of human OA chondrocytes to high-mobility group box-1 (HMGB1) induces the

10

expression of nitric oxide (NO), IL-1β, TNF, and MMP-1323. We previously reported that 29-

11

kDa FN-f, which was found in the synovial fluid of OA joints, triggered TLR-2/MAPK/NF-kB

12

signaling pathways and was implicated in the induction of catabolic enzymes, such as MMP-1, -

13

3, and -131. A contradictory report found that low-concentration 29-kDa FN-f (1 nM) increased

14

the synthesis of PG and the release of anabolic factors, such as insulin-like growth factor-1 and

15

TGF-β1, indicating that it possesses both catabolic and anabolic activities24. In addition to

16

stimulating catabolic signaling, DAMPs, such as soluble biglycan, S100A8/S100A9, and ACAN

17

32-mer fragment, negatively regulated the expression of anabolic factors, including ACAN and

18

type II collagen in chondrocytes by modulating the TLR-2- or TLR-4-NF-κB pathways22,25,26.

19

TLR-4 activation by lipopolysaccharide (LPS) was found to decrease the expression of ACAN

20

and type II collagen, and blocking p38 inhibited the negative effects of LPS on matrix

21

biosynthesis27. These observations suggest that the DAMP/TLR pathway may be important in

22

both the catabolic and anabolic processes of ECM. Compared with the regulation of catabolic

23

responses by DAMPs in OA cartilage, relatively few studies have reported its role in anabolic

24

signaling pathways in the pathogenesis of OA.

AC C

EP

TE D

M AN U

SC

5

15

ACCEPTED MANUSCRIPT XTs, including XT-1 and XT-2, initiate GAG biosynthesis by transferring a sugar to an

2

acceptor protein or peptide28. They show different tissue expression patterns; for example, XT-1

3

expression is predominant in mouse testis, kidney, and brain, whereas XT-2 is dominant in the

4

liver, indicating the different physiological functions of the two enzymes29,30. A previous study

5

in dwarf mutant mice showed that XT-1, but not XT-2, was associated with skeletal

6

development, suggesting a lack of redundancy in XT expression in cartilage tissue31. Mutational

7

analysis of the XT gene in OA patients suggested that although single nucleotide

8

polymorphisms of XT-1 and XT-2 were not correlated with OA duration, joint involvement, XT

9

activity, and hip or knee phenotype, elevated serum levels of XT activity served as a marker for

10

the progression of articular cartilage damage32. Furthermore, XT activity was reported to

11

decrease depending on cartilage aging in rats33. Consistent with previous reports, our data

12

showed that XT-1 was significantly downregulated in human OA and DMM-induced mouse OA

13

cartilage relative to normal cartilage, and 29-kDa FN-f, a DAMP, significantly suppressed the

14

expression of XT-1 in monolayer chondrocyte culture, explant culture of human cartilage tissue,

15

and in vivo mouse joint cartilage, indicating that 29-kDa FN-f may be involved in the

16

pathogenesis of OA through the regulation of XT-1 expression. In a previous study, injection of

17

FN-fs into rabbit knee joints resulted in up to 70% depletion of proteoglycan within 7 days11,

18

while our study showed that such injection had little influence on ECM of mouse joint cartilage.

19

This discrepancy may result from the difference in experimental animal (species and age) and

20

the source of FN-f. For example, a mixture of FN-f (29- & 50-kDa FN-f) purified from rabbit

21

plasma was used in the previous study whereas we used 29-kDa FN-f isolated from human

22

plasma. XT-1 was found to be co-regulated with ACAN and XT-1 inhibition was shown before

23

loss of safranin O (Fig. S2 and Fig 2C). This suggests that XT-1 might serve as a surrogate

24

marker for ACAN or cartilage loss.

AC C

EP

TE D

M AN U

SC

RI PT

1

16

ACCEPTED MANUSCRIPT Sp1 and Sp3 are members of the specificity protein (Sp)/Krüppel-like transcription factor

2

family34. Sp1 has the capacity to form tetramers owing to its C-terminal D domain, which is

3

involved in synergistic transactivation by regulating chromatin looping between enhancer and

4

promoter regions of target genes35. By contrast, Sp3 binds as a monomer to Sp binding sites in

5

promoters without forming multimers, and shows no synergistic transactivation36. Sp1 and Sp3

6

participate in the expression of a variety of genes by competitively binding to the same region in

7

the promoters of target genes, and the Sp3/Sp1 ratio consequently influences promoter activity37.

8

In chondrocytes, IL-1β increases Sp3 protein expression and inhibits Sp1 protein biosynthesis

9

while decreasing the binding activity of both Sp1 and Sp3 to the COL2A1 promoter38. The

10

siRNA-mediated knockdown of Sp3 led to a highly significant reduction of XT-1 mRNA, and

11

the c-Jun/AP-1 transcription factor was essential for full XT-1 promoter activity in SW1353

12

cells39. On the other hand, Sp1 knockdown partially inhibited ADAMTS-4 induction by IL-1 in

13

chondrocytes39. Consistent with these reports, our results from Sp1 and Sp3 silencing and

14

overexpression vector transfections showed that 29-kDa FN-f altered the expression of Sp1 and

15

Sp3 and thus the Sp3/Sp1 ratio, leading to reduced XT-1 expression in human articular

16

chondrocytes. Sp1 and Sp3 interact directly and indirectly with transcription factors,

17

transcriptional regulators, and chromatin remodeling factors, and the activities of Sp1 and Sp3

18

are modulated by binding proteins as well37. The interactions of Sp1 and c-Jun are implicated in

19

the regulation of claudin-4, damage-induced neuronal endopeptidase, and integrin-linked

20

kinase40-42. Our IP analysis revealed that 29-kDa FN-f suppressed the association of Sp1 and c-

21

jun in the XT-1 promoter, reducing XT-1 expression.

AC C

EP

TE D

M AN U

SC

RI PT

1

22

We previously reported that chondrocytes expressed TLRs and that 29-kDa FN-f induced

23

catabolic responses by binding to TLR-2 and increasing TLR-2 expression8. Previous reports

24

demonstrated that integrin α5β1 also functioned as the receptor for FN and the FN-derived 17

ACCEPTED MANUSCRIPT proteolytic peptides 29-, 50-, and 140-kDa FN-fs16,43. Here, by using antibodies against TLR-2

2

and α5β1 integrin and siRNA tools, we demonstrated that both receptors were involved in the

3

regulation of XT-1 expression by 29-kDa FN-f, and blocking both using antibodies had

4

synergistic effects on reversing the effects of 29-kDa FN-f. The significant overlap of signaling

5

pathway of FN-f and IL-1β may be due to signaling through TLR-2 by FN-f which utilizes

6

MyD88-IRAK-MAPK-AP-1 pathway.

RI PT

1

In conclusion, our results demonstrate that 29-kDa FN-f suppresses anabolic responses by

8

activating TLR-2- and integrin α5β1-dependent signaling pathways in human articular

9

chondrocytes. Specifically, 29-kDa FN-f affects the pathogenesis of OA by reducing the

10

expression of cartilage matrix synthesis enzyme XT-1, suggesting that a relevant pathway may

11

be a useful therapeutic target in OA.

12

M AN U

SC

7

ACKNOWLEDGEMENTS

14

We would like to thank In Young Park for her assistance with histological processing.

15

TE D

13

Author contributions

17

HSH and HAK contributed to study conception and design. MHL contributed to acquisition of

18

data. HSH contributed to analysis and interpretation of data. HSH, MHL and HAK drafted the

19

article or revised it critically for important intellectual content. All authors read and approved

20

the final manuscript.

AC C

21

EP

16

22

Role of funding source

23

This study was supported by the Basic Science Research Program through the National

24

Research

Foundation

(NRF)

of

Korea 18

funded

by

the

Ministry

of

Education

ACCEPTED MANUSCRIPT 1

(2014R1A1A2059823 and 2017R1A2B200188) and in part by Hallym University Research

2

Fund.

3

5

Conflict of interests

RI PT

4

The authors declare no conflicts of interest.

AC C

EP

TE D

M AN U

SC

6

19

ACCEPTED MANUSCRIPT 1

References

2

1.

Hwang HS, Park SJ, Cheon EJ, Lee MH, Kim HA. Fibronectin fragment-induced expression of matrix metalloproteinases is mediated by MyD88-dependent TLR-2

4

signaling pathway in human chondrocytes. Arthritis Res Ther 2015;17(320.

5

2.

RI PT

3

Princivalle M, de Agostini A. Developmental roles of heparan sulfate proteoglycans: a comparative review in Drosophila, mouse and human. Int J Dev Biol 2002;46(3):267-

7

78.

8

3.

SC

6

Khair M, Bourhim M, Barre L, Li D, Netter P, Magdalou J, et al. Regulation of xylosyltransferase I gene expression by interleukin 1beta in human primary

10

chondrocyte cells: mechanism and impact on proteoglycan synthesis. J Biol Chem

11

2013;288(3):1774-84. 4.

13 14

Schwartz NB. Regulation of synthesis of chondroitin sulfate proteoglycan. Ups J Med Sci 1977;82(2):76-7.

5.

TE D

12

M AN U

9

Venkatesan N, Barre L, Magdalou J, Mainard D, Netter P, Fournel-Gigleux S, et al. Modulation of xylosyltransferase I expression provides a mechanism regulating

16

glycosaminoglycan chain synthesis during cartilage destruction and repair. FASEB J

17

2009;23(3):813-22. 6.

19

Xylosyltransferase-I regulates glycosaminoglycan synthesis during the pathogenic

20 21

Venkatesan N, Barre L, Bourhim M, Magdalou J, Mainard D, Netter P, et al.

AC C

18

EP

15

process of human osteoarthritis. PLoS One 2012;7(3):e34020.

7.

Mis EK, Liem KF, Jr., Kong Y, Schwartz NB, Domowicz M, Weatherbee SD.

22

Forward genetics defines Xylt1 as a key, conserved regulator of early chondrocyte

23

maturation and skeletal length. Dev Biol 385(1):67-82.

24

8.

Kim HA, Cho ML, Choi HY, Yoon CS, Jhun JY, Oh HJ, et al. The catabolic pathway 20

ACCEPTED MANUSCRIPT 1

mediated by Toll-like receptors in human osteoarthritic chondrocytes. Arthritis Rheum

2

2006;54(7):2152-63.

3

9.

Xie DL, Hui F, Meyers R, Homandberg GA. Cartilage chondrolysis by fibronectin fragments is associated with release of several proteinases: stromelysin plays a major

5

role in chondrolysis. Arch Biochem Biophys 1994;311(2):205-12. 10.

7

fluid. J Rheumatol 1992;19(9):1448-52. 11.

9

Homandberg GA, Meyers R, Williams JM. Intraarticular injection of fibronectin fragments causes severe depletion of cartilage proteoglycans in vivo. J Rheumatol

10 11

SC

8

Xie DL, Meyers R, Homandberg GA. Fibronectin fragments in osteoarthritic synovial

1993;20(8):1378-82. 12.

M AN U

6

RI PT

4

Stanton H, Ung L, Fosang AJ. The 45 kDa collagen-binding fragment of fibronectin induces matrix metalloproteinase-13 synthesis by chondrocytes and aggrecan

13

degradation by aggrecanases. Biochem J 2002;364(Pt 1):181-90.

14

13.

TE D

12

Hwang HS, Park SJ, Lee MH, Kim HA. MicroRNA-365 regulates IL-1beta-induced catabolic factor expression by targeting HIF-2alpha in primary chondrocytes. Sci Rep

16

2017;7(1):17889. 14.

18

mediated dephosphorylation of c-Jun C terminus regulates phorbol ester-induced c-

19 20

Chen BK, Huang CC, Chang WC, Chen YJ, Kikkawa U, Nakahama K, et al. PP2B-

AC C

17

EP

15

Jun/Sp1 interaction in A431 cells. Mol Biol Cell 2007;18(3):1118-27.

15.

21

Forsyth CB, Pulai J, Loeser RF. Fibronectin fragments and blocking antibodies to alpha2beta1 and alpha5beta1 integrins stimulate mitogen-activated protein kinase

22

signaling and increase collagenase 3 (matrix metalloproteinase 13) production by

23

human articular chondrocytes. Arthritis Rheum 2002;46(9):2368-76.

24

16.

Homandberg GA, Hui F. Arg-Gly-Asp-Ser peptide analogs suppress cartilage 21

ACCEPTED MANUSCRIPT 1

chondrolytic activities of integrin-binding and nonbinding fibronectin fragments. Arch

2

Biochem Biophys 1994;310(1):40-8. 17.

4 5

Yasuda T. Cartilage destruction by matrix degradation products. Mod Rheumatol 2006;16(4):197-205.

18.

RI PT

3

Rosenberg JH, Rai V, Dilisio MF, Sekundiak TD, Agrawal DK. Increased expression of damage-associated molecular patterns (DAMPs) in osteoarthritis of human knee

7

joint compared to hip joint. Mol Cell Biochem 2017; 19.

9

Houard X, Goldring MB, Berenbaum F. Homeostatic mechanisms in articular cartilage and role of inflammation in osteoarthritis. Curr Rheumatol Rep

10

M AN U

8

SC

6

2013;15(11):375.

11

20.

Goldring MB. Articular cartilage degradation in osteoarthritis. HSS J 2012;8(1):7-9.

12

21.

Nefla M, Holzinger D, Berenbaum F, Jacques C. The danger from within: alarmins in

13

22.

TE D

14

arthritis. Nat Rev Rheumatol 2016;12(11):669-83.

Schelbergen RF, Blom AB, van den Bosch MH, Sloetjes A, Abdollahi-Roodsaz S, Schreurs BW, et al. Alarmins S100A8 and S100A9 elicit a catabolic effect in human

16

osteoarthritic chondrocytes that is dependent on Toll-like receptor 4. Arthritis Rheum

17

2012;64(5):1477-87. 23.

19

chondrocytes express the receptor for advanced glycation end products: Potential role

20 21

Loeser RF, Yammani RR, Carlson CS, Chen H, Cole A, Im HJ, et al. Articular

AC C

18

EP

15

in osteoarthritis. Arthritis Rheum 2005;52(8):2376-85.

24.

Homandberg GA, Hui F, Wen C, Purple C, Bewsey K, Koepp H, et al. Fibronectin-

22

fragment-induced cartilage chondrolysis is associated with release of catabolic

23

cytokines. Biochem J 1997;321 ( Pt 3)(751-7.

24

25.

Barreto G, Soininen A, Ylinen P, Sandelin J, Konttinen YT, Nordstrom DC, et al. 22

ACCEPTED MANUSCRIPT 1

Soluble biglycan: a potential mediator of cartilage degradation in osteoarthritis.

2

Arthritis Res Ther 2015;17(379.

3

26.

Lees S, Golub SB, Last K, Zeng W, Jackson DC, Sutton P, et al. Bioactivity in an Aggrecan 32-mer Fragment Is Mediated via Toll-like Receptor 2. Arthritis Rheumatol

5

2015;67(5):1240-9.

6

27.

RI PT

4

Bobacz K, Sunk IG, Hofstaetter JG, Amoyo L, Toma CD, Akira S, et al. Toll-like receptors and chondrocytes: the lipopolysaccharide-induced decrease in cartilage

8

matrix synthesis is dependent on the presence of toll-like receptor 4 and antagonized

9

by bone morphogenetic protein 7. Arthritis Rheum 2007;56(6):1880-93. 28.

11 12

M AN U

10

SC

7

Gotting C, Kuhn J, Kleesiek K. Human xylosyltransferases in health and disease. Cell Mol Life Sci 2007;64(12):1498-517.

29.

Cuellar K, Chuong H, Hubbell SM, Hinsdale ME. Biosynthesis of chondroitin and heparan sulfate in chinese hamster ovary cells depends on xylosyltransferase II. J Biol

14

Chem 2007;282(8):5195-200.

15

30.

TE D

13

Ponighaus C, Ambrosius M, Casanova JC, Prante C, Kuhn J, Esko JD, et al. Human xylosyltransferase II is involved in the biosynthesis of the uniform tetrasaccharide

17

linkage region in chondroitin sulfate and heparan sulfate proteoglycans. J Biol Chem

18

2007;282(8):5201-6. 31.

20

Mis EK, Liem KF, Jr., Kong Y, Schwartz NB, Domowicz M, Weatherbee SD. Forward genetics defines Xylt1 as a key, conserved regulator of early chondrocyte

21 22

AC C

19

EP

16

maturation and skeletal length. Dev Biol 2014;385(1):67-82.

32.

Schon S, Huep G, Prante C, Muller S, Christ R, Hagena FW, et al. Mutational and

23

functional analyses of xylosyltransferases and their implication in osteoarthritis.

24

Osteoarthritis Cartilage 2006;14(5):442-8. 23

ACCEPTED MANUSCRIPT 1

33.

Wolf B, Gressner AM, Nevo Z, Greiling H. Age-related decrease in the activity of

2

UDP-xylose:core protein xylosyltransferase in rat costal cartilage. Mech Ageing Dev

3

1982;19(2):181-90. 34.

Meinders M, Kulu DI, van de Werken HJ, Hoogenboezem M, Janssen H, Brouwer

RI PT

4 5

RW, et al. Sp1/Sp3 transcription factors regulate hallmarks of megakaryocyte

6

maturation and platelet formation and function. Blood 2015;125(12):1957-67. 35.

Deshane J, Kim J, Bolisetty S, Hock TD, Hill-Kapturczak N, Agarwal A. Sp1

SC

7

regulates chromatin looping between an intronic enhancer and distal promoter of the

9

human heme oxygenase-1 gene in renal cells. J Biol Chem 2010;285(22):16476-86.

10

36.

M AN U

8

Yu B, Datta PK, Bagchi S. Stability of the Sp3-DNA complex is promoter-specific:

11

Sp3 efficiently competes with Sp1 for binding to promoters containing multiple Sp-

12

sites. Nucleic Acids Res 2003;31(18):5368-76. 37.

14 15

Li L, Davie JR. The role of Sp1 and Sp3 in normal and cancer cell biology. Ann Anat

TE D

13

2010;192(5):275-83. 38.

Chadjichristos C, Ghayor C, Kypriotou M, Martin G, Renard E, Ala-Kokko L, et al. Sp1 and Sp3 transcription factors mediate interleukin-1 beta down-regulation of

17

human type II collagen gene expression in articular chondrocytes. J Biol Chem

18

2003;278(41):39762-72. 39.

20

Sylvester J, Ahmad R, Zafarullah M. Role of Sp1 transcription factor in Interleukin-1induced

21 22

AC C

19

EP

16

ADAMTS-4

(aggrecanase-1)

gene

expression

in

human

articular

chondrocytes. Rheumatol Int 2013;33(2):517-22.

40.

Ikari A, Atomi K, Yamazaki Y, Sakai H, Hayashi H, Yamaguchi M, et al.

23

Hyperosmolarity-induced up-regulation of claudin-4 mediated by NADPH oxidase-

24

dependent H2O2 production and Sp1/c-Jun cooperation. Biochim Biophys Acta 24

ACCEPTED MANUSCRIPT 1 2

2013;1833(12):2617-27. 41.

Kiryu-Seo S, Kato R, Ogawa T, Nakagomi S, Nagata K, Kiyama H. Neuronal injuryinducible gene is synergistically regulated by ATF3, c-Jun, and STAT3 through the

4

interaction with Sp1 in damaged neurons. J Biol Chem 2008;283(11):6988-96.

5

42.

RI PT

3

Tang Q, Zhao S, Wu J, Zheng F, Yang L, Hu J, et al. Inhibition of integrin-linked kinase expression by emodin through crosstalk of AMPKalpha and ERK1/2 signaling

7

and reciprocal interplay of Sp1 and c-Jun. Cell Signal 2015;27(7):1469-77.

9

43.

Johansson S, Svineng G, Wennerberg K, Armulik A, Lohikangas L. Fibronectinintegrin interactions. Front Biosci 1997;2(d126-46.

M AN U

8

SC

6

10

AC C

EP

TE D

11

25

ACCEPTED MANUSCRIPT 1

FIGURE LEGENDS

2

Figure 1. Reduced xylosyltransferase-1 (XT-1) expression in human osteoarthritis (OA)

4

articular cartilage and destabilization of medial meniscus (DMM)-induced OA murine

5

joint cartilage. (A) The relative expression of XT-1 and ACAN in normal and OA cartilage was

6

determined using Quanti Fast SYBR Green-based real-time PCR (RT-PCR). The expression

7

ratios of XT-1 relative to glyceraldehyde 3-phosphate dehydrogenase (GAPDH), the internal

8

control, are shown. Data represent the mean ± standard deviation (SD) for duplicate data from

9

10 separate donors for XT-1 and from 8 normal and 17 OA donors for ACAN. P value was

10

presented between two groups. Dot indicates the representative value obtained from each donor.

11

Expression of XT-1 in normal and OA cartilage (B) and in joint cartilage of sham and DMM

12

mice (C) was determined by immunohistochemistry (IHC) using anti-XT-1 antibodies. Data

13

represent the results from three normal and OA cartilage samples (B) and four sham and six

14

DMM 10-week-old mice (C). Scale bars = 100 µm for 100× and 200× magnifications and 50 µm

15

for ×400 magnification.

SC

M AN U

TE D

EP

16

RI PT

3

Figure 2. The 29-kDa fibronectin fragment (FN-f) suppresses XT-1 expression in primary

18

chondrocytes and joint cartilage. (A) mRNA expression of XT-1 and aggrecan (ACAN)

19

decreased following treatment with 29-kDa FN-f. Chondrocytes were stimulated with 29-kDa

20

FN-f (150 and 300 nM) for 6, 14, and 24 h. The mRNA levels of XT-1 and ACAN were

21

measured using RT-PCR. Data represent the mean ± standard deviation (SD) of duplicate data

22

from more than three different donors. P value was presented between two groups. Dot indicates

23

the representative value obtained from each donor. (B) XT-1 protein expression was

24

significantly suppressed following 24 h incubation with 29-kDa FN-f. Chondrocytes were

AC C

17

26

ACCEPTED MANUSCRIPT treated with 29-kDa FN-f (300 nM) for 6, 14, and 24 h. Protein levels of XT-1 were measured

2

using western blot analysis. The blot shown is representative of blots from six independent

3

donors. (C, D) IHC and safranin O staining in (C) murine joint cartilage injected with 29-kDa

4

FN-f (3 µM/10 µL; n = 6) or PBS (10 µL; n = 6) and (D) explant culture of human cartilage in

5

the presence of 29-kDa FN-f (300 nM) for 7 days. XT-1 expression was determined by IHC

6

using anti-XT-1 antibody and cartilage degeneration was examined by safranin O staining. i.a.,

7

intra-articular. Scale bars = 100 µm for 100× and 200× magnifications and 50 µm for ×400

8

magnification.

SC

RI PT

1

M AN U

9

Figure 3. Specificity protein 1 (Sp1), Sp3, and activator protein 1 (AP-1) are

11

transcriptional regulators of XT-1. (A, B) 29-kDa FN-f regulates the expression of Sp1 and

12

Sp3 at both the (A) mRNA and (B) protein levels. Chondrocytes were treated with 29-kDa FN-f

13

(300 nM) for 6, 14, and 24 h. mRNA and protein levels of Sp1 and Sp3 were measured using

14

RT-PCR and western blot analysis, respectively. RT-PCR data is expressed as the mean ±

15

standard deviation (SD) of duplicate data from three independent experiments. P value was

16

presented between two groups. Dot indicates the representative value obtained from each donor.

17

(C, D) Knockdown and overexpression of Sp1 and Sp3 affect XT-1 expression in primary

18

human chondrocytes. (C) Chondrocytes were transfected with control siRNA (Sc), small

19

interfering Sp1 RNA (siSp1), and Sp1 overexpression (O.E.) vector. (D) Chondrocytes were

20

transfected with small interfering Sp3 RNA (siSp3), empty vector (E), and Sp3 O.E. vector.

21

After 48 h, chondrocytes were stimulated with 29-kDa FN-f (300 nM) for 24 h. (B–D) The

22

protein levels of XT-1, Sp1, and Sp3 were measured using western blot analysis. β-actin served

23

as the loading control. The blots shown are representative of three or more independent

24

experiments using chondrocytes from three different donors. E, empty vector. (E) Treatment

AC C

EP

TE D

10

27

ACCEPTED MANUSCRIPT with 29-kDa FN-f activates transcription factor AP-1. Chondrocytes were treated with 29-kDa

2

FN-f (300 nM) for 24 h. The levels of c-jun/phospho-c-jun (Ser63) and c-fos/phospho-c-fos

3

(Ser32) in the cytosol and nuclear fractions of 29-kDa FN-f-treated chondrocytes were measured

4

by western blot analysis. β-actin and lamin B served as loading controls for the cytoplasmic and

5

nuclear fractions, respectively. Cyt, cytosolic; Nuc, nuclear. (F) AP-1 inhibition recovers XT-1

6

expression suppressed by 29-kDa FN-f treatment. Chondrocytes were exposed to Tanshinone

7

IIA, an AP-1 inhibitor, for 6 h, and to 12-O-tetradecanoylphorbol acetate (TPA), an AP-1

8

activator, for 24 h, followed by treatment with 29-kDa FN-f for 24 and 48 h. The levels of XT-1,

9

Sp1, and Sp3 were determined using western blot analysis. The blots shown are representative

SC

M AN U

10

RI PT

1

of three or more independent experiments using chondrocytes from three different donors.

11

Figure 4. 29-kDa FN-f alters the binding of transcription factors Sp1, Sp3, and AP-1 to the

13

XT-1 promoter. (A) Chromatin immunoprecipitation (ChIP) assay for the binding of Sp1, Sp3,

14

and AP-1 to the XT-1 promoter region in 29-kDa FN-f-treated chondrocytes. ChIP assays were

15

performed using 29-kDa FN-f-treated chondrocytes. Chromatin was sheared by sonication and

16

protein-DNA complexes were immunoprecipitated with antibodies against Sp1, Sp3, and c-jun.

17

DNA retrieved from the precipitates was amplified with primers for specific sequences

18

encompassing the Sp1 (#1), Sp3 (#1), and AP-1 (#2) binding regions in the XT-1 promoter. The

19

PCR products were resolved on agarose gels and stained with GelRed. IgG was used as a

20

negative control. Input represents 5% of the total cross-linked chromatin before

21

immunoprecipitation. #1, Sp1/Sp3-40 and Sp1/Sp3-125; #2, AP1-590 and AP1-640 regions of

22

the XT-1 promoter. Data are representative of three independent experiments using

23

chondrocytes from three different donors. (B) 29-kDa FN-f inhibited the association between

24

Sp1 and c-jun. 29-kDa FN-f-stimulated chondrocytes were lysed with RIPA buffer and the

AC C

EP

TE D

12

28

ACCEPTED MANUSCRIPT nuclear extracts were immunoprecipitated with antibodies against Sp1 or c-jun. The proteins

2

were analyzed by western blot analysis with antibodies against Sp1, c-jun, and β-actin. Input,

3

nuclear extracts. Data are representative of three independent experiments using chondrocytes

4

from three different donors.

RI PT

1

5

Figure 5. 29-kDa FN-f downregulates XT-1 expression by activating nuclear factor

7

kappa B (NF-κB) and mitogen-activated protein kinase (MAPK) signaling pathways.

8

The NF-κB and MAPK signaling pathways are implicated in the expression of (A) XT-1 and

9

ACAN and (B) Sp1 and Sp3. Chondrocytes were pretreated for 2 h with various inhibitors

10

and stimulated with 29-kDa FN-f for 14 h. mRNA levels of XT-1 were measured using

11

Quanti Fast SYBR Green-based RT-PCR. SP (SP6001250), JNK inhibitor; SB (SB203580),

12

p38 inhibitor; PD (PD98059), MEK1/2 inhibitor; Bay (Bay 11-7082), NF-κB inhibitor. Data

13

are expressed as the mean ± standard deviation (SD) of duplicate data from three independent

14

donors. P value was presented between two groups (vs. 29-kDa FN-f-treated cells). Dot

15

indicates the representative value obtained from each donor.

M AN U

TE D

EP

16

SC

6

Figure 6. Toll-like receptor 2 (TLR-2) and integrin α5β β1 function as receptors for 29-

18

kDa FN-f in human primary chondrocytes. (A) 29-kDa FN-f significantly binds to TLR-2.

19

29-kDa FN-f (1, 5, 10, and 30 µg/mL) was added to TLR-2-coated microplates for 2 h, and

20

antibodies specific to FN or IgG were added to each well for 1 h. Data are expressed as the

21

mean ± standard deviation (SD) of triplicate data from three independent donors. P value was

22

presented between two groups (vs. untreated controls). (B) Effect of TLR-2 or integrin α5β1

23

knockdown on 29-kDa FN-f-suppressed XT-1, ACAN, Sp1, and Sp3 expression. 29-kDa

24

FN-f-inhibited XT-1 expression was recovered by knockdown of endogenous TLR-2 and

AC C

17

29

ACCEPTED MANUSCRIPT integrin expression. Chondrocytes were transfected with siTLR-2, si-integrin α5β1, and

2

control siRNA. After 48 h, cells were stimulated with 29-kDa FN-f for 14 h. XT-1, ACAN,

3

Sp1, and Sp3 expression were measured using Quanti Fast SYBR Green-based RT-PCR.

4

Data are expressed as the mean ± standard deviation (SD) of duplicate data from more than

5

three independent donors. P value was presented between two groups (vs. scRNA- and 29-

6

kDa FN-f-treated cells). (C) Effects of TLR-2 and integrin α5β1 antibodies on 29-kDa FN-f-

7

suppressed XT-1 expression in primary chondrocytes. To neutralize TLR-2 and integrin

8

α5β1, chondrocytes were incubated with antibodies specific to TLR-2, integrin α5β1, and

9

IgG for 24 h followed by treatment with 29-kDa FN-f for 14 h. XT-1, ACAN, Sp1, and Sp3

10

expression were measured by Quanti Fast SYBR Green-based RT-PCR. Data are expressed

11

as the mean ± standard deviation (SD) of duplicate data from three independent donors. P

12

value was presented between two groups (vs. 29-kDa FN-f-treated cells). Dot indicates the

13

representative value obtained from each donor.

TE D

M AN U

SC

RI PT

1

14

Figure S1. mRNA expression of XT-1 and ACAN according to culture period. XT-1 and

16

ACAN expression in human OA chondrocytes. Chondrocytes were cultured in the presence

17

or absence of FBS for 72 h. mRNA levels of XT-1 and ACAN were measured using RT-PCR.

18

P value was presented between two groups. Dot indicates the representative value obtained

19

from each donor.

AC C

20

EP

15

21

Figure S2. The effect of cellular level of ACAN on XT-1 core protein expression. (A)

22

The expression of ACAN was suppressed by XT-1 knockdown. (B) The expression of XT-1

23

was suppressed by ACAN knockdown. Chondrocytes were transfected with control siRNA

24

(Sc), (A) small interfering XT-1 RNA (si-XT-1), and (B) si-ACAN. After 48 h, chondrocytes 30

ACCEPTED MANUSCRIPT 1

were stimulated with 29-kDa FN-f (300 nM) for 24 h. mRNA levels of ACAN and XT-1

2

were measured using RT-PCR. P value was presented between two groups. Dot indicates the

3

representative value obtained from each donor.

RI PT

4

AC C

EP

TE D

M AN U

SC

5

31

ACCEPTED MANUSCRIPT 1

The English in this document has been checked by at least two professional editors, both

2

native speakers of English. For a certificate, please see:

3 4

http://www.textcheck.com/certificate/5Qes2h

AC C

EP

TE D

M AN U

SC

RI PT

5

32

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT