Introduction
The pathogenesis of rheumatoid arthritis (RA) is characterized by the infiltration of inflammatory cells into the pannus, followed by tissue destruction. The RA synovium contains elevated levels of cytokines and inflammatory cells such as lymphocytes and monocytes [
1,
2]. Chemokines and other inflammatory mediators drive the pathogenesis of RA and regulated production of proinflammatory cytokines is important for the orchestration of the inflammatory response [
3‐
5]. Current therapies are designed to block cytokines such as TNF-α or IL-6 [
6,
7]. However, despite the success of blocking these cytokines, not all RA patients respond adequately to anti-TNF-α or anti-IL-6 therapy.
Angiogenesis is a highly regulated process that results in the formation of new vessels. It is important in vasculoproliferative states such as wound repair and chronic inflammation, as seen in RA [
8,
9]. The angiogenic process is important in the progression of RA and may prove to be a promising therapeutic target [
10]. Cellular adhesion molecules expressed on endothelial cells (ECs) are involved in leukocyte extravasation into the synovium leading to perpetuation of RA synovial inflammation [
11].
Glycosylation is one of the most common posttranslational modification reactions, and many proteins in eukaryotes are glycosylated [
12]. Most of these are
N-linked and/or
O-linked glycan chains that are synthesized posttranslationally in the endoplasmic reticulum and the Golgi apparatus by various glycosyltransferases [
13]. Fucosylated glycans are synthesized by fucosyltransferases (futs). Thirteen fut genes have been identified in the human genome [
14]. Fucosyltransferase 1 (fut1) and fut2 are α(1,2)-fucosyltransferases responsible for synthesis of the H blood group antigen and related structures [
15,
16]. Fut1 is overexpressed in some cancers such as colon and pancreas [
17,
18]. These reports indicate that α(1,2)-linked fucose synthesized by futs are important for tumor growth. In regards to arthritis, mRNA levels of fut7 are upregulated in synovial fluid (SF) compared to peripheral blood T cells in patients with juvenile idiopathic arthritis [
19].
We have shown previously that the soluble form of E-selectin mediates angiogenesis via its endothelial ligand sialyl Lewis
x[
20]. We have also shown that Lewis
y-6/H/5-2 (Le
y/H), synthesized by fut1, and its glucose analog, 2-fucosyllactose (H-2 g) mediates angiogenesis and inflammatory cell adhesion [
21,
22]. However, a direct role for fut1 in RA has not been demonstrated. In this study, we found that α(1,2)-linked fucosylated proteins were expressed in RA synovium. Hence, we show that α(1,2)-linked fucosylated proteins are upregulated in RA synovial tissue (ST) and that fut1 in RA synovial fibroblasts is important in EC tube formation, leukocyte-synovial fibroblast adhesion, and synovial fibroblast proliferation, all critical aspects of inflammation in the RA joint synovium.
Methods
Patients
RA and osteoarthritis (OA) ST were obtained from patients undergoing arthroplasty or synovectomy. Normal (NL) ST samples were obtained from a National Disease Research Interchange and Cooperative Human Tissue Network. NL skin biopsies were obtained from the University of Michigan Tissue Procurement Service. For all human specimens used in this study, we obtained written informed consent with approval from the University of Michigan Institutional Review Board.
Homogenate preparation
ST homogenates were prepared in anti-protease buffer as we have done previously [
23]. Briefly, RA, OA and NL STs were homogenized in 5 ml of a 1% protease inhibitor cocktail (Pierce, Rockford, IL, USA) in PBS. Samples were centrifuged at 900 g for 10 minutes, and filtered through a 1.2-μm pore size Sterile Acrodisk, and frozen at -80°C until thawed for assay. The total protein concentration of each lysate was determined using a bicinchoninic acid assay (Pierce).
Cell culture
Fresh STs were minced and digested in tissue enzyme digestion solution as described previously [
24]. The synovial fibroblasts were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% FBS. Cells were seeded in 6-well plates (BD Biosciences, Bedford, MA, USA) at a density of 1 × 10
5 cells per well, and were maintained in complete medium. After overnight serum starvation, cells were treated with 25 ng/ml TNF-α (R&D Systems, Minneapolis, MN, USA) for 24 hours. Cell-conditioned medium and cell lysates were collected. We used fibroblasts from three NL knees; three knees and one hip from OA patients; and six knees from RA patients.
Human dermal microvascular endothelial cells (HMVECs) were purified from digested skin biopsies using mouse anti-human CD31 MicroBeads (Miltenyi Biotec, Cambridge, MA, USA), according to the manufacturer’s protocol. Cells were cultured in EC basal medium (Lonza, Walkersville, MD, USA) with growth factors. In order to confirm EC purity, we used antibodies to EC markers von Willebrand factor and CD31 and immunohistochemistry. THP-1 cells (human acute monocytic leukemia cell line) were purchased from the American Type Culture Collection (Manassas, VA, USA). THP-1 cells were cultured in RPMI supplemented with 10% FBS.
Enzyme-linked lectin assay (ELLA)
For RA, OA and NL ST homogenate analysis, we measured the total protein content of each of our samples as done previously [
25]. We next used Ulex Europeaus Agglutinin 1 lectin (UEA-1) to measure α(1,2)-linked fucosylated proteins in the same samples. UEA-1 binds specifically to α(1,2)-fucose, the terminal sugar of blood group antigens H and Lewis
y. 2′fucosyllactose-bovine serum albumin (2′FL-BSA), 50 to 0.78 ng/ml (V-labs Inc, Covington, LA, USA) was used as a standard for measurement of total fucosyated proteins. Because 2′FL-BSA is linked to α(1,2)-fucose on BSA, and BSA is a well known standard for measuring proteins, we used 2′FL-BSA as a metric for quantifying total α(1,2)-linked fucosylated proteins in our samples. Results are presented as the ratio of the total fucosylated proteins in each of the ST homogenate samples using fucosylated BSA as standard (in ng of 2′FL-BSA), normalized to the total protein content of the same sample (in mg) as measured using nonfucosylated BSA as a standard (Pierce).
The ELLA was performed by adding samples and standards to 96-well plates with an overnight incubation at 4°C. The next day, plates were washed (PBS + 0.05% Tween), and blocked with Synblock (ImmunoChemistry Tech, Bloomington, MN, USA) for 2 hours. UEA-1, 2 μg/ml (Vector laboratories Inc, Burlingame, CA, USA) was added for 90 minutes followed by 10 μg/ml biotinylated goat anti-UEA-1 antibody (Vector laboratories Inc) for 60 minutes, then streptavidin-horseradish peroxidase (HRP) for 30 minutes. Tetramethylbenzine substrate (TMB) was used as a color development reagent and the plate was read at 450 nm following addition of 1 N H2SO4 on a BioTek Synergy plate reader (Winooski, VT, USA). Total fucosylated proteins from synovial fibroblast conditioned medium and cell lysates were measured by ELLA using 2′FL-BSA as a standard curve.
Immunofluorescence
Immunofluorescence staining on RA ST fibroblasts was performed as previously described [
25]. To determine if α(1,2)-linked proteins were expressed on RA ST synovial fibroblasts, mouse anti-human collagen-1 (Abcam, Cambridge, MA, USA) and goat anti-UEA-1 (Vector laboratories Inc.) were used. RA ST slides were fixed with cold acetone for 20 minutes. Then slides were blocked with 20% FBS and 5% donkey serum for 1 hour at 37°C, and incubated with UEA-1 (2 μg/ml, Vector laboratories Inc.) for 1 hour at 37°C. Goat anti-UEA-1 and rabbit anti-human collagen-1 were used as primary antibodies. Fluorescent conjugated donkey anti-goat (for UEA-1) and anti-mouse (for collagen-1) secondary antibodies were purchased from Life Technologies (Carlsbad, CA, USA). For nuclear staining, 4′, 6-diamidino-2-phenylindole (DAPI) was used. Images were taken at 100× magnification. Anti-collagen-1-positive fibroblasts were shown by fluorescent red staining, and fucosylation was shown by fluorescent green. Yellow cells were a result of merging the green and red fields.
Dual immunofluorescence staining was done on RA ST sections embedded in optimal cutting temperature (OCT) medium and cryosectioned. The slides were fixed and blocked in 5% goat serum. Rabbit anti-human fut1 (Thermo Scientific, Waltham, MA), mouse anti-human CD68 macrophage marker (BD Biosciences, San Jose, CA), and mouse anti-human Cadherin-11 (fibroblast marker; R&D Systems) were used as primary antibodies at a concentration of 10 μg/ml and incubated for 1 hour at 37°C. The slides were washed with PBS and goat anti-rabbit fluorescein isothiocyanate (FITC)-conjugated and goat anti-mouse rhodamine-conjugated antibodies were used as secondary antibodies and were incubated for 1 hour at 37°C. The slides were again washed in PBS and 4′, 6-diamidino-2-phenylindole (DAPI) staining was used at 1:5,000 concentration. The slides were mounted and visualized under tetramethylrhodamine (TRITC), FITC, and DAPI wavelengths.
Immunohistologic analysis
We performed immunoperoxidase staining on cryosections from NL, OA, and RA ST as described previously [
25]. ST samples were blocked with 20% FBS and 5% goat serum in PBS, and incubated with mouse anti-human fut1 (10 μg/ml, Thermo Scientific) or purified nonspecific mouse IgG. ST samples were washed with PBS, and a 1:200 dilution of biotinylated goat anti-mouse antibody was added. After washing, antibody binding was detected using a Vectastain ABC Elite peroxidase system (Vector laboratories Inc.) and chromogen 3,3′-diaminobenzidine (DAB) (Kirkegaard & Perry Laboratories Inc., Gaithersburg, MD, USA). ST samples were counterstained with Harris hematoxylin. Staining was evaluated by a pathologist blinded to the experimental conditions. Each of the ST components was graded for immunostaining and scored 0% to 100%, in which 0% indicates no staining and 100% indicates that all cells were immunoreactive [
25].
Transfection of RA synovial fibroblasts with fut1 small interfering RNA (siRNA)
RA synovial fibroblasts were seeded in 6-well plates at a density of 1 × 105 cells per well. Cells were maintained in complete medium up to 70% confluency. siRNA (50 nM) against fut1 or control siRNA and transfection regent (Mirus, Madison, WI, USA) was mixed with TransIT-TKO transfection reagent according to manufacturer’s instructions and overlaid on the cells. Cells were incubated with the siRNA/TransIT-TKO for 24 hours at 37°C. Control and fut1 siRNA were purchased from Santa Cruz Biotechnology. To determine the transfection efficiency of cells, a fluorescein-conjugated nonsilencing siRNA (Santa Cruz Biotechnology) was transfected into cells with TransIT-TKO. Transfected cells were counted by fluorescence microscopy, and total cells were counted by bright field microscopy. The transfection efficiency was calculated as the percent of fluorescein-positive cells divided by the number of bright field cells. The percent knockdown of fut1 expression was confirmed using quantitative polymerase chain reaction (qPCR) and western blotting.
Cell lysis and western blotting
Cell lysis and western blotting were performed as previously described [
25]. RA synovial fibroblasts were transfected with fut1 siRNA or control siRNA. Cells were seeded in 6-well plates at a density of 1 × 10
5 cells per well. After overnight serum starvation, cells were stimulated with TNF-α (25 ng/ml). Membranes were probed with rabbit anti-human fut1 antibody (Epitomics Inc., Burlingame, CA, USA). The immunoblots were stripped and reprobed with rabbit anti-β-actin to verify equal loading. For cell signaling experiments, antibodies against phosphorylated and total JNK, NFκB, P38, and Erk1/2 (Cell Signaling Technology, Danvers, MA, USA) were used.
Co-culturing HMVECs and RA synovial fibroblasts in the EC tube formation assay
In order to confirm the effect of fut1 on HMVEC tube formation, a facet of angiogenesis, we co-cultured HMVECs and RA synovial fibroblasts using a Costar transwell system (Corning Inc., Lowell, MA, USA). RA synovial fibroblasts were first transfected with either control or fut1 siRNA as described above and were plated in the top inserts of the transwell system. HMVECs were grown on the bottom wells. HMVECs and RA synovial fibroblasts were co-cultured with serum-free endothelial basal medium (EBM) for 24 hours. HMVECs were collected from the co-culture plates, and subsequently plated on Matrigel (BD Biosciences) with the co-cultured conditioned media for 6 hours at 37°C. Tubes formed by HMVECs were counted by a blinded observer [
26].
RNA extraction and qPCR of RA synovial fibroblasts
RNA extraction and qPCR were performed as previously described [
27]. Fut1, monocyte chemoattractant protein 1 (MCP-1)/CCL2, epithelial-derived neutrophil-activating peptide 78 (ENA-78)/CXCL5, vascular endothelial growth factor (VEGF) and β-actin primer pairs were purchased from Integrated DNA Technologies (Coralville, IA, USA). The following primers were used; fut1 forward 5′-GTGCCCGTATCCAGAGTGAT-3′; reverse 5′-AGGACCCAGGGGAGAGTAAA-3′; MCP-1/CCL2 forward 5′-TCCAGCATGAAAGTCTCTGC-3′; reverse 5′-TGGAATCCTGAACCCACTTC-3′; ENA-78/CXCL5 forward 5′-GAGAGCTGCGTTGCGTTTG-3′; reverse 5′-TTTCCTTGTTTCCACCGTCCA-3′; VEGF forward 5′-ATGAACTTTCTGCTGTCTTGGGT-3′; reverse 5′-TGGCCTTGGTGAGGTTTGATCC-3′; β-actin forward 5′-GCTAGGCAGCTCGTAGCTCT-3′; reverse 5′-GCCATGTACGTTGCTATCCA-3′. All samples were run in duplicate and analyzed using Applied Biosystems software (Life Technologies).
ELISA for MCP-1/CCL2, ENA-78/CXCL5, and VEGF
ELISA was performed in a manner described previously [
28]. Fut1 siRNA, control siRNA or nontreated RA synovial fibroblasts were stimulated with TNF-α (25 ng/ml) for 24 hours, and cell supernatants were collected. Levels of MCP-1/CCL2, ENA-78/CXCL5, and VEGF were measured.
In vitro cell adhesion assay
Adhesion of THP-1 cells to nontreated, control siRNA or fut1 siRNA treated RA synovial fibroblasts grown to confluence in 96-well plates was examined [
25]. RA synovial fibroblasts were serum-starved overnight. The next day, cells were treated with TNF-α (25 ng/ml) for 24 hours. THP-1 cells were collected and labeled with 5 μM Calcein AM fluorescent dye (Life Technologies) for 30 minutes. After washing twice, 1 × 10
5 THP-1 cells were added to each well and incubated for 30 minutes at room temperature. Nonadherent cells were washed off and fluorescence was measured using a Synergy HT fluorescence plate reader (BioTek Instruments, Winooski, VT).
Cell surface ELISA for adhesion molecule expression
Nontreated, control siRNA-transfected, or fut1 siRNA-transfected RA synovial fibroblasts (1 × 10
5/well) were seeded in 96-well plates. Confluent RA synovial fibroblasts were serum-starved overnight prior to stimulation with TNF-α (25 ng/ml) for 24 hours. Cells were fixed with 3.7% formalin in PBS, and cell surface ELISA was performed as previously described [
29]. Mouse anti-human antibodies specific for intercellular adhesion molecule 1 (ICAM-1), 10 μg/ml, (R&D Systems) or vascular cell adhesion molecule 1 (VCAM-1) were used, and the plates were read with an ELISA reader at 450 nm.
Cell proliferation assay
Control or fut1 siRNA-transfected RA synovial fibroblasts were seeded in 96-well plates at 5 × 104 cells/ml. Cells were serum-starved overnight then treated with 10 μg/ml lipopolysaccharide (LPS) from Escherichium coli 0111 (Sigma-Aldrich) for 4 and 24 hours. Each treatment group experiment was performed in four replicate wells. DNA was measured using a CyQuant cell proliferation assay kit (Life Technologies) following the manufacturer’s instructions. For the assay, cells were lysed and total cellular nucleic acid was measured using fluorescence at 520 nm emission after excitation at 480 nm.
Statistical analysis
All data were analyzed using parametric tests, namely the Student’s t-test assuming equal variances. Data are reported as the mean ± standard error of the mean (SEM). P-values less than 0.05 were considered statistically significant. All error bars represent the SEM and n represents the number of independent experiments. Of note, all data represented in the manuscript that were significant using a parametric test, were similarly significant using the one-tailed Mann Whitney test.
Discussion
Posttranslational modifications, such as glycosylation, citrullination or NH
2-terminal truncation of natural cytokines, change their biological activity [
30,
31]. Nabeshima
et al. reported that the cytokine glycosylation on receptor binding changed biological activity [
32]. These reports indicate that glycosylated cytokines may contribute to disease pathogenesis. Over half of known proteins are modified by covalently bound glycans, which are important for physiological processes including protein folding, degradation, signaling, and immune function [
33]. The complexity of the glycoproteome is thought to be several orders of magnitude greater than the proteonome [
33]. Human ABO blood group antigens and Lewis systems are oligosaccharides synthesized by sequential actions of futs and these antigens are important in blood typing [
34]. The α(1,2)-fucosyltransferases fut1 and fut2 are the enzymes responsible for catalyzing an α(1,2)-linkage of fucose to terminal beta galactosidase [
35]. H and Lewis antigens are expressed most abundantly in endodermal epithelial cells, where the majority of human cancers arise [
36].
We hypothesized that fut1 in RA is overexpressed, and mediates angiogenesis, cell adhesion, and fibroblast proliferation. Indeed, we found that α(1,2)-linked fucosylated proteins were overexpressed in RA. Fucosylation is one of the most common modifications involving oligosaccharides on glycoproteins, and their structures are involved in a variety of biological processes in eukaryotic organisms, angiogenesis, fertilization, cell adhesion, inflammation, and tumor metastasis [
37]. We and others have previously reported that sialyl Lewis
x, synthesized by α(1,3)-fucosyltransferases, is involved in angiogenesis [
21]. In addition, we reported that soluble H and Lewis
y antigens, both synthesized by fut1, are potent mediators of cell adhesion, angiogenesis, and monocyte recruitment [
22,
38,
39]. Our study clearly demonstrates that α(1,2)-linked fucosylated proteins are more highly expressed on RA synovial fibroblasts than on NL synovial fibroblasts. Przybysz
et al. showed that the expression of α(1,6)-linked fucose in synovial fibronectins was related to RA disease activity [
40]. Kratz
et al. showed that the proportions of fucosyl determinants of intact synovial IgA and IgG were lower in the early RA group compared to the advanced RA group [
41], suggesting that fucosylated antibodies may be important in chronic RA pathogenesis. These findings suggest that α(1,2)-linked fucosylation has an important role in RA.
We next focused on fut1 expression and function in RA tissues. Fut1 is overexpressed in some cancers such as colon and pancreas. Fut1 mRNA in cancer tissues was elevated compared to normal tissues [
17,
18]. Thus far, there have been no reports of fut1 in RA. We examined a potential relationship between lining cells and lining thickness score, however there was not a correlation between them. Nonetheless, we and others have shown that angiogenesis is important in the growth and proliferation of the RA ST pannus, and in the ingress of leukocytes, and that cytokines play a key role in this process [
9,
20,
42]. We also found that HMVECs from a co-culture system with fut1 siRNA-treated RA synovial fibroblasts had decreased HMVEC tube formation compared with HMVECs from a similar co-culture system with control siRNA-treated or nontreated RA synovial fibroblasts. This is in agreement with Mathieu
et al. who showed that fut1-deficient hepatocarcinoma cells had reduced angiogenic responses [
43].
After defining the activity of fut1 using HMVEC tube formation assays with RA synovial fibroblasts, we assessed the expression of pro-angiogenic mediators from fut1 siRNA-transfected RA synovial fibroblasts. We found that MCP-1/CCL2, ENA-78/CXCL5, and VEGF mRNA in TNF-α-stimulated fut1 siRNA-transfected RA synovial fibroblasts were decreased compared to control siRNA-transfected RA synovial fibroblasts. We also found that secretion of MCP-1/CCL2, ENA-78/CXCL5, and VEGF in TNF-α-stimulated fut1 siRNA-transfected RA synovial fibroblasts was decreased compared to control siRNA-transfected RA synovial fibroblasts. These findings suggest that fut1 expressed in synovial fibroblasts is important in RA angiogenesis by contributing to the production of pro-angiogenic mediators.
We next examined the role of fut1 in leukocyte adhesion. Leukocyte retention in the synovium is an active process mediated in part by cellular adhesion molecules [
44]. We found that adhesion of myeloid THP-1 cells to fut1 siRNA-transfected RA synovial fibroblasts was significantly decreased compared with control or nontreated RA synovial fibroblasts. These findings are consistent with Palumberi
et al. who showed that adhesion of human epidermoid carcinoma cells to fut1 and fut2 siRNA-transfected ECs was decreased compared with control siRNA-transfected ECs [
45]. On the other hand, Kwiatkowski
et al. reported that EC surface expression of terminally sialylated structures by high-level fut1 activity reduced monocyte adherence and activation [
46]. However, this group did not examine which adhesion molecules were differentially expressed during fut1 inhibition in ECs. In addition, the former group used bovine post-capillary venular ECs that bind to epidermoid carcinoma cells, while the other group used porcine EC monolayers to examine monocyte adhesion. Perhaps the different type of ECs, the use of epidermoid carcinoma cells, along with the different cell isolation methods could account for the differences in cellular adhesion. It could also be that overexpression of futs may have limits, and that highly elevated levels of fut activity may actually inhibit monocyte and EC interactions, at least in
in vitro systems. Nonetheless, we found that cell-surface adhesion molecules such as ICAM-1 or VCAM-1 on fut1 siRNA-transfected RA synovial fibroblasts were decreased compared to control fibroblasts. These findings indicate that fut1 in RA synovial fibroblasts is not only important for cell adhesion, but indicate that these interactions may also lead to activation of inflammatory cells and perpetuation of inflammation in RA synovium.
RA synovial fibroblasts proliferate and invade cartilage [
47]. We found that fut1 siRNA inhibits cell proliferation of LPS-stimulated RA synovial fibroblasts. Interestingly, Palumberi
et al. also reported that fut1 and fut2 siRNA-treated human epidermoid carcinoma cells have reduced cell proliferation when transfected with fut1 and fut2 siRNA [
45]. In agreement with Palumberi, we found that fut1 facilitates fibroblast proliferation, indicating that α(1,2)-linked fucosylation by fut1 may contribute to fibroblast overgrowth in the RA pannus.
Finally, we found that fut1 siRNA inhibited phosphorylated JNK signaling in RA synovial fibroblasts. On the other hand, fut1 siRNA did not inhibit phosphorylation of NFκB, P38, and Erk1/2 signaling in RA synovial fibroblasts. Overall, our results demonstrate that JNK plays key roles in mediating angiogenesis, cell adhesion and RA synovial fibroblast proliferation through fut1.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Conception and design by AEK and TI. Acquisition of data by TI, JHR, PT, CMH, (GE) and GKH. Analysis and interpretation of data by TI, JHR, MAA, PLC, (GE) and AEK. Drafting of manuscript by TI, JHR, MAA, PLC, and AEK. All authors read and approved the final manuscript.