HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Shiau, A.-L. Articles by Wu, C.-L. Search for Related Content PubMed PubMed Citation Articles by Shiau, A.-L. Articles by Wu, C.-L.

Prothymosin Lacking the Nuclear Localization Signal as an Effective Gene Therapeutic Strategy in Collagen-Induced Arthritis¹

Ai-Li Shiau^*, Shih-Yao Chen, Meng-Ya Chang, Chih-Hau Su, Shih-Ye Chung, Yi-Te Yo^*, Chrong-Reen Wang^2, and Chao-Liang Wu^2,

^* Department of Microbiology and Immunology, Department of Biochemistry and Molecular Biology, and Section of Rheumatology and Immunology, Department of Internal Medicine, National Cheng Kung University Medical College, Tainan, Taiwan

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Prothymosin (ProT) is regulated by c-Myc, an oncoprotein overexpressed in synovium of rheumatoid arthritis, and is associated with cell proliferation. However, ProT also exerts immunomodulatory activities. The growth-promoting activity of ProT can be abolished by deleting its nuclear localization signal (NLS). In this study, we showed that AdProTNLS, an adenoviral vector encoding ProT lacking the NLS, did not enhance the proliferation of synovial fibroblasts. AdProTNLS treatment abolished the up-regulation of the MIP-1 promoter activity induced by TNF- in synovial fibroblasts. AdProTNLS suppressed macrophage chemotaxis and reduced macrophage infiltration into the ankle joints in rats with collagen-induced arthritis (CIA). Neutralization test confirmed the involvement of MIP-1 in macrophage chemotaxis. Administration of AdProTNLS reduced the severity of CIA in the clinical, radiographic, and histological aspects. The levels of TNF- (mean ± SEM, 1261.9 ± 107.9 vs 2880.1 ± 561.4 pg/mg total protein; p < 0.05), IL-1 (56.8 ± 8.0 vs 109.2 ± 4.9 pg/mg total protein; p < 0.01), and MIP-1 (41.7 ± 3.6 vs 55.2 ± 1.1 pg/mg total protein; p < 0.05) in the ankle joints were lower in the AdProTNLS-treated rats with CIA than those in their control counterparts. In the AdProTNLS-treated ankle joints, matrix metalloproteinase-9 expression was decreased by 40% and infiltrating macrophages reduced by 50%. Our results demonstrate that intra-articular delivery of AdProTNLS significantly ameliorated the clinical course of CIA in rats. This study is the first to suggest that ProT lacking the NLS may have therapeutic potential for the management of rheumatoid arthritis.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Rheumatoid arthritis (RA)³ is a chronic systemic autoimmune disease that is characterized by joint inflammation as well as progressive cartilage and bone erosion (1). Understanding the pathophysiology of RA has created a large number of possible molecular targets for developing therapeutic candidates. In RA, the normally thin synovial lining becomes dramatically thickened and hyperplastic due to persistent secretion of proinflammatory cytokines, such as IL-1 and TNF-, by macrophages and synoviocytes (2, 3). Recently, anti-TNF- Ab therapy has been used for treating RA. However, the expressions of physiological TNF- and immunoregulatory cytokines may be compromised by anti-TNF-, leading to side effects, such as microbial infections (4). Moreover, available treatments for RA fail to provide long-lasting control of the symptoms or disease progression (5). Gene therapeutic approaches have been explored for the treatment of RA. Genes of interest can be delivered locally and efficiently at the joint by intra-articular injection (6, 7).

Prothymosin (ProT) is an acidic, ubiquitous protein that contains 113 aa residues with thymosin l sequence at its N terminus (8). ProT is a nuclear protein known to play an essential role in the proliferation of mammalian cells (9). Moreover, ProT functions as a biological response modifier and has many immunomodulatory activities, which may be of great benefit in clinical applications (8). The mechanisms underlying the anticancer activity of ProT in experimental tumor models and in studies of melanoma and colorectal cancer patients include induction of tumoricidal peritoneal macrophages, enhancement of NK and lymphokine-activated killer cell activities, induction of IL-2 production, as well as induction of tumor-specific cytotoxic CD8⁺ and CD4⁺ Th cell activation when administered simultaneously with tumor cells (10, 11, 12).

ProT has been reported to exert in vitro immunomodulatory effects on autoimmune diseases (13, 14). It enhances the HLA-DR expression on monocytes from patients with multiple sclerosis, resulting in the restoration of the deficient autologous MLR (15). The depressed autologous and allogeneic MLR in patients with systemic lupus erythematosus can be enhanced in the presence of ProT (16). In addition, ProT is one of the downstream targets of c-Myc that directly regulates the expression of ProT gene by binding to the E-box of its promoter (17). In the rheumatic joint, pannus formation is attributable to the proliferation of synovial fibroblast-like cells and accumulation of infiltrating macrophages. Apart from an altered morphology, synovial fibroblasts exhibit elevated gene expression of proto-oncogenes, such as c-myc and c-ras (18, 19). It has been shown that Ras- and c-Myc-dependent signaling events cooperate to regulate the growth and invasiveness of synovial fibroblasts in RA (20). Previous clinical trials in autoimmune diseases have highlighted the benefits of using thymic peptides for treatment and to strengthen the effects of immunomodulators (14, 21). Thymic peptides, such as thymosin fraction 5, have been used for the treatment of RA. Thymosin fraction 5 reduces the overproduction of IL-2 by PBMC from RA patients (22). Previous studies suggest that regulatory T cells are active players in the regulation of inflammation in arthritis (23). Moreover, regulatory T cells derived from patients with active RA are functionally compromised (24). In this regard, thymosin fraction 5 has been shown to increase the suppressor cell activity of PBL from patients with RA (25).

In this study, we investigated the feasibility of adenovirus-mediated ProT gene therapy for the treatment of experimental arthritis. Because ProT is implicated as a nuclear protein associated with cell proliferation, and the basic cluster of amino acids, TKKQKT, at its C terminus is identified as the nuclear localization signal (NLS) (26), to circumvent the growth-promoting activity of ProT, we constructed an adenoviral vector encoding ProT lacking the NLS, designated AdProTNLS. We examined the prophylactic effect of AdProTNLS in the collagen-induced arthritis (CIA) model in rats. Our results suggest that intra-articular delivery of ProT gene lacking the NLS may be an effective treatment strategy for RA.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Generation of adenoviral vectors

Two adenoviral vectors, AdProT encoding wild-type ProT and AdProTNLS encoding NLS-deleted ProT (27), were generated, as previously described (28). AdLacZ, an adenoviral vector encoding -galactosidase, was also used as the control vector. Recombinant viruses were propagated and viral titers were measured by standard plaque assay using 293 cells.

Immunohistochemical staining of ProT in synovial tissues

Fresh synovial tissues were obtained from patients diagnosed with RA or osteoarthritis (OA) according to the diagnostic criteria set by the American College of Rheumatology. They were embedded in OCT compound, and cryostatic sections (5 µm) were prepared and incubated overnight at 4°C with mAb against ProT (ascites 1:100) (9, 29). Immunostaining was performed using a commercial streptavidin-biotin alkaline phosphatase complex Dako LSAB 2 kit (DakoCytomation), according to the manufacturer’s instructions, using aminoethyl carbazole staining and hematoxylin counterstaining.

Synovial fibroblast proliferation assay

Human synovial fibroblasts were prepared from synovial tissues of patients with RA by standard methods. They were cultured in RPMI 1640 medium supplemented with 2 mM L-glutamine, 20 mM HEPES, 10% FCS, and 50 µg/ml gentamicin for 48 h at 37°C in 5% CO₂. The adherent cells were cultured continuously until reaching confluence. After four or more passages, synovial fibroblasts became a homogeneous population and were used for further studies. The fibroblasts (10⁴/well) were seeded into 96-well plates and cultured overnight. Cells were then infected with various adenoviral vectors at 5, 10, or 100 multiplicities of infection (MOI) and incubated for 96 h. Cell proliferation was determined using the MTT assay. Metabolization of MTT was quantified by measuring the absorbance at 570 nm.

Induction and gene therapy of CIA in rats

To induce experimental arthritis, male Sprague-Dawley rats (6–7 wk of age) were immunized, by two dorsal intradermal injections (400 and 100 µg, respectively) 1 wk apart, with bovine type II collagen (Elastin Products) emulsified in IFA (Difco), as previously described (6). Because high doses of adenoviral vectors may cause proinflammatory effects and cell-mediated immune responses to viral Ags, which may have adverse effects on arthritis, we used a medium dose of 5 x 10⁷ PFU to compromise its potential adverse effects on arthritis. This dose has been used in our previous studies for gene therapy of CIA rats with adenoviral vectors encoding thrombospondin-1 or kallistatin gene (6, 7). Both ankles of the rats were then injected with AdProTNLS or AdLacZ four times at 5-day intervals starting on day 7. Clinical, radiographic, histopathological, and immunohistochemical assessments of arthritis were determined, as previously described (6, 7). Macrophages in the synovial tissue were detected by immunohistochemistry with mouse anti-rat ED-1 mAb (Chemicon International) (6), and quantified using the Image-Pro Plus software (Media Cybernetics). For single-treatment regimen, rats immunized with collagen were injected with a single dose of AdProTNLS or AdLacZ on day 7. Clinical scores and ankle circumferences of the rats were monitored. In a separate experiment, on day 16, some treated rats were killed, and their joints were homogenized and proteins extracted for chemotaxis assay.

Analysis of MIP-1 promoter activity in rat synovial fibroblasts

To construct pFRL2-MIP-1, a MIP-1 promoter reporter plasmid that carries both firefly luciferase gene driven by the MIP-1 promoter and Renilla luciferase gene controlled by the CMV promoter, the 0.5-kb MIP-1 promoter fragment, which was excised from the pGL3-basic-MIP-1 promoter reporter plasmid (30) by digestion with HindIII, filling in the cohesive end with Klenow enzyme, and digestion with KpnI, was subcloned into the KpnI/SmaI sites of pFRL2, a single dual-luciferase reporter vector (31). Synovial tissues were isolated on day 17 from rats with CIA, and primary culture of synovial fibroblasts was established. Cells from a fifth culture passage grown in 24-well plates were transfected with 1 µg of pFRL2-MIP-1 reporter plasmid using polyethylenimine (Sigma-Aldrich Diagnostics). After 24 h, cells were infected with AdProTNLS or AdLacZ at a MOI of 10 for 24 h. Cells were then treated with 10 ng/ml rat rTNF- (PeproTech) for 24 h, and their firefly and Renilla luciferase activities were measured with the Dual-Luc Reporter Assay System (Promega). The ratio of firefly luciferase activity to Renilla luciferase activity was expressed as relative light units.

Chemotaxis assay

Chemotaxis assay was preformed using a 48-well modified Boyden chamber (NeuroProbe) with 8-µm-pore polycarbonate membranes (Nucleopore). MC3T3-E1 osteoblastic cells cultured in DMEM supplemented with 10% FCS were infected with AdProTNLS or AdLacZ at a MOI of 10, and their conditioned medium was collected 48 h later. In addition, the supernatant from MC3T3-E1 cells stimulated with LPS (1 µg/ml) for 16 h served as the positive control for chemotaxis assay. Murine RAW 264.7 macrophage-like cells (5 x 10⁴) were added to the upper chamber, and the conditioned medium from MC3T3-E1 cells infected with the adenoviral vectors was added to the lower chamber of the Boyden chamber. After incubation for 5 h, cells attached to the lower filter surface of the membrane were fixed by methanol, stained with toluidine blue, and counted in a microscope. To determine whether MIP-1 was involved in microphage infiltration into the joint of rats with CIA, chemotaxis assay was performed, as described above, except that joint homogenate extracts were added to the lower chamber of the Boyden chamber. Joint homogenate extracts from AdLacZ-treated ankles were pretreated with 0.1 or 1 ng/ml anti-rat MIP-1 mAb (clone 175024; R&D Systems) for 30 min at 37°C. Joint homogenate extracts from AdLacZ-injected rats pretreated with or without anti-MIP-1 mAb, as well as those from rats injected with AdProTNLS or PBS, were added to the lower chamber of the Boyden chamber.

ELISA

Synovial tissues taken from the ankle joints of rats with CIA on day 27 were snap frozen and homogenized, as described previously (32). The levels of MIP-1, IL-1, and TNF- in the ankle homogenates were quantified using ELISA kits (R&D Systems), according to the manufacturer’s instructions. Total protein concentrations were determined using a bicinchoninic acid protein assay kit (Pierce).

Gelatin zymography

The enzymatic activities of matrix metalloproteinase (MMP)-2 and MMP-9 were determined by gelatin zymography. The ankle homogenates were electrophoresed on 12% SDS-polyacrylamide gels containing 0.2% gelatin under nonreducing conditions. Following electrophoresis, the gels were washed with 1% Triton X-100 to remove SDS. After equilibration in enzyme substrate buffer (40 mM Tris-HCl (pH 7.5), 10 mM CaCl₂), the gels were incubated in the same buffer for 18 h at 37°C. They were then stained with Coomassie blue G-250 and destained in water. The clear bands with gelatinolytic activity on the zymogram were quantified by densitometric scanning to determine the amounts of MMP-2 and MMP-9 as integrated density value.

Statistical analysis

Data were expressed as mean ± SEM. Differences in ankle circumference and articular index score between the AdProTNLS- and AdLacZ-treated groups were compared by two-way ANOVA with Bonferroni adjustment for multiple comparisons. Differences in other data between the two groups were analyzed with Student’s t test. Values of p < 0.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Expression of ProT in the RA synovium

It has been documented that c-Myc is expressed in the RA synovium detected by immunohistochemical and molecular analysis (18, 33). Because ProT is a downstream target of the c-Myc signaling pathway (17), we examined its expression in the synovium of RA and OA tissues by immunohistochemical staining with mAb against ProT. The positively stained cells were found in the synovium of three different RA tissues (Fig. 1, D–H), whereas few positive cells were detected in three different OA tissues (Fig. 1, A–C). Of note, in two of the three RA samples, expression of ProT was detected mainly in the lining layer of the synovium (Fig. 1, D, E, G, and H), No false-positive cells were observed in the Ab control sample (Fig. 1I). Taken together, immunohistochemical analysis reveals expression of ProT appeared to be higher in RA synovium than OA synovium.

View larger version (132K): [in this window] [in a new window]   FIGURE 1. Expression of ProT in the synovia from patients with RA. Fresh frozen sections of human synovial tissues from three patients with OA (A–C) and three patients with RA (D–I) were immunostained with mouse mAb against ProT, followed by rabbit anti-mouse IgG (D–H). Control section (I) was stained, omitting the primary Ab to evaluate nonspecific staining. The sections were further stained with aminoethyl carbazole as a peroxidase substrate and counterstained with hematoxylin. Images D and G, as well as E and H, were from the same section with different magnifications. Images H and I were consecutive sections. Original magnifications of x200 in A–F and x100 in G–I.

Biological behaviors of AdProTNLS-treated synovial fibroblasts

Because ProT is known to play a role in cell proliferation (9), we first examined whether AdProTNLS that encoded ProT lacking the NLS enhanced the proliferation of synovial fibroblasts isolated from patients with RA. Whereas AdProT encoding wild-type ProT enhanced synovial fibroblast proliferation in a dose-dependent manner, AdProTNLS, like AdLacZ, had no effects on cell proliferation (Fig. 2A), confirming the abolishment of the growth-promoting potential of ProT lacking the NLS (27). The facts that macrophages play an important role in the elaboration of chemotactic cytokines in RA and MIP-1 plays a role in the selective recruitment of macrophages in synovial inflammation associated with RA prompted us to investigate whether AdProTNLS down-regulated the MIP-1 promoter activity. The MIP-1 promoter activity in the rat synovial fibroblasts following infection with AdProTNLS or AdLacZ was determined by the dual-luciferase reporter assay. In the absence of TNF-, the promoter activity of MIP-1 in the AdLacZ-treated cells was higher than that in their AdProTNLS-treated counterparts (Fig. 2B). The MIP-1 promoter activity was also up-regulated in response to TNF- stimulation in the synovial fibroblasts. Whereas AdLacZ treatment dramatically increased the MIP-1 promoter activity in the presence of TNF-, AdProTNLS treatment completely suppressed the transcriptional activity of the promoter induced by TNF- and adenoviral transduction. Suppression of the MIP-1 promoter activity by AdProTNLS was also detected in MC3T3-E1 osteoblastic cells (data not shown). Together, these results demonstrate that AdProTNLS suppressed the MIP-1 promoter activity in synovial fibroblasts either under basal conditions or in response to TNF- stimulation. These findings provided the impetus for using adenovirus-mediated ProTNLS gene transfer for the treatment of rats with CIA.

View larger version (22K): [in this window] [in a new window]   FIGURE 2. Biological behaviors of AdProTNLS-treated synovial fibroblasts and osteoblasts. A, Synovial fibroblasts from patients with RA were infected with various adenoviral vectors at 5, 10, or 100 MOI and incubated for 96 h. Cell proliferation was determined by the MTT assay (n = 4). B, The promoter activity of MIP-1 in synovial fibroblasts infected with adenoviral vectors. Rat synovial fibroblasts were transfected with pFRL-2-MIP-1 luciferase reporter plasmid, and infected with AdProTNLS or AdLacZ at a MOI of 10 at 24 h posttransfection. After 24 h, cells were treated with rat rTNF- (10 ng/ml) or left untreated, and cultured for an additional 24 h. Cell lysates were then harvested, and their firefly and Renilla luciferase activities were determined. The ratio of firefly luciferase activity to Renilla luciferase activity was expressed as relative light units (n = 3). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Effects of AdProTNLS on CIA in rats

Rats that had been immunized on days 0 and 7 with collagen were administered with 5 x 10⁷ PFU of AdProTNLS or AdLacZ on days 7, 12, 17, and 22, and their ankle circumferences and articular index scores were monitored. The ankle circumferences on days 22 and 27 in the AdProTNLS-treated rats were significantly smaller than those in the AdLacZ-treated rats (Fig. 3A). Likewise, the articular index scores were significantly reduced in the AdProTNLS-treated rats compared with those in the control rats from day 15 onward (Fig. 3B). Representative photographs of the rats on day 27 reveal that adenovirus-mediated ProTNLS gene transfer resulted in a significant reduction of joint inflammation (decreased ankle width and swelling) (Fig. 3C). The radiograph of the AdLacZ-treated ankle revealed joint destruction with osteoporosis, erosion, soft tissue swelling, and joint space narrowing (Fig. 3D). Notably, the joint destruction was significantly reduced in the AdProTNLS-treated ankles (Fig. 3D). Radiographic scores were significantly lower in the AdProTNLS-treated joints than those in the AdLacZ-treated joints (1.10 ± 0.23 vs 2.40 ± 0.34; p < 0.01). We further confirmed the clinical and radiological findings with histological evaluation in the ankle joints. The AdLacZ-treated joints revealed synovial hyperplasia, bone erosion, and infiltration of inflammatory cells (Fig. 3E). In the AdProTNLS-treated ankles, the joint tissue showed significant reduction of synovial hyperplasia, bone erosion, and pannus formation (Fig. 3E). Notably, histological scores were significantly lower in the AdProTNLS-treated ankles than in their AdLacZ-treated counterparts (1.50 ± 0.20 vs 3.70 ± 0.68; p < 0.05). In the same CIA model, we also treated the rats before the onset of arthritis with a single dose of AdProTNLS or AdLacZ on day 7. The ankle circumferences (Fig. 3F) and articular index scores (Fig. 3G) were significantly reduced in rats treated with AdProTNLS compared with those treated with AdLacZ up to day 20. However, from day 24 onward, the clinical symptoms did not differ between the two groups.

View larger version (51K): [in this window] [in a new window]   FIGURE 3. Prophylactic effects of ProTNLS gene transfer on CIA in rats. Both ankles of rats were immunized intradermally with collagen on days 0 and 7, followed by intra-articular injection of AdProTNLS or AdLacZ (5 x 107 PFU) four times starting on day 7 at intervals of 5 days (A–E). The AdProTNLS-treated ankles (n = 20) showed a significant reduction in arthritis on hind limbs with regard to ankle circumference (A) on days 24 and 27 and articular index score (B) from day 15 onward, compared with their AdLacZ-treated counterparts (n = 20). C, Representative photographs of the rat ankles receiving AdProTNLS or AdLacZ on day 27. D, The radiograph of joint destruction shows osteoporosis, erosion, soft tissue swelling, and joint space narrowing in the AdLacZ-treated ankle, whereas joint destruction was significantly reduced in the AdProTNLS-treated ankle. E, The histological appearance of the synovial tissue from the AdProTNLS-treated ankle was characterized by reductions of synovial hyperplasia, bone erosion, and pannus formation, as compared with the AdLacZ-treated ankle (original magnification, x100). In a separate experiment, one injection of AdProTNLS or AdLacZ (5 x 107 PFU) was given on day 7 to rats that had been immunized with collagen on days 0 and 7. Ankle circumferences (F) and articular index scores (G) were significantly reduced in AdProTNLS-treated ankles (n = 12) from days 10 to 20, compared with their AdLacZ-treated counterparts (n = 12). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Effects of AdProTNLS on macrophage chemotaxis in vitro and infiltration into the ankle joints in rats with CIA

To establish whether AdProTNLS affected the migration of inflammatory cells, chemotaxis assay was conducted using RAW 264.7 cells as chemotactic cells. LPS, a potent chemoattractant serving as the positive control, enhanced macrophage chemotaxis 5.7-fold (Fig. 4A). Moreover, although the conditioned medium from AdLacZ-treated MC3T3-E1 cells triggered an increase in macrophage chemotaxis compared with the basal transmigration, the conditioned medium from AdProTNLS-treated cells significantly reduced the number of migrating macrophages. In the chemotaxis assay using joint homogenate extracts from rats with CIA as chemoattractants for RAW 264.7 cells, joint homogenate extracts from the AdLacZ-treated rats, but not from the AdProTNLS-treated rats, induced macrophage chemotaxis (Fig. 4B). Notably, neutralization of MIP-1 activity in the AdLacZ-treated joint homogenates by anti-MIP-1 mAb dose responsively abrogated macrophage chemotaxis induced by the joint homogenates from the AdLacZ-treated rats. In addition, we examined the effect of AdProTNLS on macrophage infiltration into the ankle joints in rats with CIA. Immunohistochemical staining shows a significant decrease in ED-1-positive cells that infiltrated to the joint tissues in the AdProTNLS-treated rats compared with those in the AdLacZ-treated rats (Fig. 4C). Quantitative analysis also confirmed significant lower numbers of infiltrating macrophages in the AdProTNLS-treated ankle joints compared with their AdLacZ-treated counterparts (Fig. 4D).

View larger version (30K): [in this window] [in a new window]   FIGURE 4. Effects of AdProTNLS on macrophage chemotaxis in vitro and infiltration into the ankle joints in rats with CIA. A, Chemotaxis assay reveals that RAW 264.7 cells exhibited less chemotactic activity in response to the conditioned medium from AdProTNLS-infected MC3T3-E1 cells than in response to the conditioned medium from their AdLacZ-infected counterparts. The conditioned medium from LPS-stimulated MC3T3-E1 cells served as the positive control (n = 8). B, The joint homogenates from AdProTNLS-treated rats with CIA significantly reduced macrophage chemotaxis compared with those from their AdLacZ-treated counterparts, as determined by chemotaxis assay. Neutralization of MIP-1 activity by anti-MIP-1 Ab (0.1 or 1 ng/ml) significantly abolished macrophage chemotaxis induced by the joint homogenates from the AdLacZ-treated rats (n = 3). C, Rats were immunized intradermally with collagen on days 0 and 7, followed by intra-articular injection of AdProTNLS or AdLacZ (5 x 107 PFU) four times at 5-day intervals starting on day 7. Representative tissue sections from day 27 ankle joints revealed that ED-1+ macrophages infiltrating into the ankle joints were decreased in rats treated with AdProTNLS compared with those treated with AdLacZ, as examined by immunohistochemical staining (original magnification, x200). D, Numbers of macrophages detected by immunohistochemistry and quantified by Image-Pro Plus software were significantly lower in the AdProTNLS-treated joints than in the AdLacZ-treated joints (n = 9). **, p < 0.01; ***, p < 0.001.

Effects of AdProTNLS on the expressions of MIP-1, IL-1, TNF-, MMP-2, and MMP-9 in the ankle joints of rats with CIA

We determined the levels of MIP-1, IL-1, and TNF- in the ankle joints of rats with CIA by ELISA. The levels of MIP-1 (Fig. 5A), IL-1 (Fig. 5B), and TNF- (Fig. 5C) were significantly lower in the AdProTNLS-treated ankles than those in the AdLacZ-treated rats. Moreover, the expressions of MMP-2 and MMP-9 were analyzed by gelatin zymography. The level of MMP-2 was unaffected by AdProTNLS treatment, whereas the production of MMP-9 was significantly decreased in the AdProTNLS-treated ankles compared with their AdLacZ-treated counterparts (Fig. 5D). Quantitative assessment performed by densitometric scanning shows that the level of MMP-9 in the AdProTNLS-treated ankles was reduced by 40% when compared with that in the AdLacZ-treated ankles (Fig. 5E).

View larger version (32K): [in this window] [in a new window]   FIGURE 5. Expressions of MIP-1, IL-1, TNF-, MMP-2, and MMP-9 in the ankle joints of rats with CIA that had been treated with AdProTNLS or AdLacZ. Rats were immunized intradermally with collagen on days 0 and 7, followed by intra-articular injection of AdProTNLS or AdLacZ (5 x 107 PFU) four times at 5-day intervals starting on day 7. Synovial tissues were taken from the ankle joints on day 27, and their homogenates were prepared. Levels of MIP-1 (A) (n = 3), IL-1 (B) (n = 4), and TNF- (C) (n = 9) in the homogenates were quantified by ELISA. D, A representative gelatin zymography showing MMP-9 and MMP-2 activities in the ankle joints of AdProTNLS- and AdLacZ-treated rats with CIA. E, The integrated density values of MMP-9 assessed by densitometric scanning of the zymography were shown (n = 4). *, p < 0.05; **, p < 0.01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

MIP-1, a member of the CC chemokine family, is an 8-kDa protein first isolated from LPS-stimulated RAW 264.7 cells (34). It is produced by monocytes and other types of leukocytes, endothelial cells, fibroblasts, and smooth muscle cells (35). MIP-1 is a monocyte chemoattractant and also has differential chemoattractant effects on other leukocytes. MIP-1 is expressed in the RA synovium in very early disease, accumulates over time, and correlates with increased macrophage infiltration in long-standing disease (36). Patients with RA have higher levels of MIP-1 in their synovial fluid than patients with OA (37). Furthermore, the chemotactic activity of RA synovial fluid for macrophages can be significantly inhibited upon incubation with anti-MIP-1 Ab (37). Therefore, MIP-1 plays an important role in the pathogenesis of RA by recruiting monocytes. In rodent models of CIA and adjuvant-induced arthritis, elevated expression of MIP-1 was observed in the ankle joint (38, 39). Passive immunization with anti-MIP-1 Ab inhibited the development of arthritis in anti-type II collagen Ab-induced arthritis and delayed the onset and reduced the severity of CIA (38, 40). Decreased levels of the MIP-1 mRNA transcript were reported in mice treated with thymic peptides (41). In this study, we showed that AdProTNLS down-regulated the MIP-1 promoter activity induced by TNF- in rat synovial fibroblasts, resulting in lower levels of MIP-1, IL-1, and TNF-, as well as decreased macrophage infiltration in the ankle joints of rats with CIA.

Adenoviral vectors have been used for gene delivery to osteoblastic cell lines, achieving 90% transfection efficiency with MOI of 500 (42). Osteoblasts can synthesize TNF-, IL-1, IL-6, and IL-8 in response to a variety of stimuli, such as LPS (43, 44, 45, 46). Furthermore, osteoblasts synthesize MIP-1 in response to TNF- and IL-1 stimulation (47). Because adenoviral vector activates the innate immune response, which involves the induction of a variety of proinflammatory cytokines or chemokines, including TNF-, IL-1, IL-6, MIP-1, MIP-1, MIP-2, and MCP-1 (48), it is conceivable that osteoblasts infected with AdLacZ or AdProTNLS would express some of these cytokines or chemokines, including MIP-1. In the present study, we show that the MIP-1 promoter activities of synovial fibroblasts and osteoblasts were suppressed by AdProTNLS. The conditioned medium from AdProTNLS-infected MC3T3-E1 osteoblastic cells may contain a lower level of MIP-1 compared with that from their AdLacZ-infected counterparts. As a result, in the chemotaxis assay RAW 264.7 cells exhibited less chemotactic activity in response to the conditioned medium from AdProTNLS-infected MC3T3-E1 cells than in response to the conditioned medium from AdLacZ-infected cells. Furthermore, by using the in vitro chemotaxis assay, we demonstrate that whereas joint homogenates from AdLacZ-treated rats stimulated macrophage chemotaxis, those from AdProTNLS-treated rats did not exert such effect. The chemotactic effect of the joint homogenates from AdLacZ-treated rats could be partly abolished by the addition of anti-MIP-1 Ab. Moreover, decreased macrophage infiltration into the joint tissues was also observed in rats treated with AdProTNLS compared with those treated with AdLacZ. Therefore, the prophylactic effect of AdProTNLS may have been mediated, in part, through the suppression of MIP-1 production during the clinical course of CIA.

Macrophages play a pivotal role in RA because they are numerous in the inflamed synovial membrane and at the cartilage-pannus junction (49, 50). Macrophages possess widespread proinflammatory, destructive, and remodeling capabilities that contribute to the pathophysiology of RA. The abundance and activation of macrophages in the inflamed synovium and pannus significantly correlate with the severity of RA. In response to inflammatory stimuli, macrophages produce proinflammatory cytokines such as TNF- and IL-1, chemokines such as MIP-1, and MMP such as MMP-9, as well as NO and reactive oxygen species (49). Monocytes/macrophages in arthritic joints constitutively express MIP-1. Although resting synovial fibroblasts and articular chondrocytes release only low levels of MIP-1 under basal conditions, TNF- and IL-1, which are abundantly present in the RA joint, may further stimulate MIP-1 production by fibroblast-like synoviocytes (51). Notably, a positive feedback for macrophage recruitment exists, with the early production of MIP-1 and other chemokines by macrophages and synovial fibroblasts in the synovium tissue contributing to the recruitment of macrophages into the arthritic joint. Subsequently, production of these chemokines by infiltrating macrophages themselves amplifies this process. In this regard, although decreased macrophage infiltration into the ankle joint by ProTNLS gene transfer could be attributed to decreased MIP-1 production, lower levels of infiltrating macrophages may also result in lower MIP-1 production. Because the autoamplification loop of inflammation orchestrated by macrophages and proinflammatory cytokines and chemokines may be critical in the pathophysiologic development of RA, the strategy of suppressing MIP-1 expression may be useful for breaking off this autoamplification loop, and thus attenuating progression of the disease.

Regarding MMP, we show that adenovirus-mediated ProTNLS gene transfer significantly reduced MMP-9, but not MMP-2, production in the ankle joints of rats with CIA. This finding is in accordance with earlier work showing that rheumatoid synovial cells, which consist of variable proportions of fibroblasts, macrophages, and stellate cells, produce MMP-2, MMP-3, and MMP-9, and among which only MMP-9 production is mainly restricted to macrophages (52). Thus, down-regulation of the MMP-9 expression in the arthritic joints by AdProTNLS treatment may have been attributable to the reduced numbers of infiltrating macrophages in the ankle joints of rats with CIA.

In conclusion, this study is the first to suggest that ProT lacking the NLS may have therapeutic potential for the management of RA. Our results demonstrate that intra-articular delivery of adenoviral vectors encoding NLS-deleted ProT is an effective treatment strategy for CIA. Inhibition of macrophage recruitment into inflamed joints by suppressing MIP-1 production may play a crucial role in ameliorating CIA by ProTNLS gene therapy.

	Acknowledgments

 
We are indebted to P. Shore (School of Biological Sciences, University of Manchester, U.K.) and R. A. Hodin (Department of Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA) for providing pGL3-MIP-1 promoter reporter plasmid and pFRL2 plasmid, respectively.

	Disclosures

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
The authors have no financial conflict of interest.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work was supported by Grants NSC 92-2320-B-006-095, NSC 93-2320-B-006-080, and NSC 94-2311-B-006-006 (to C.-L.W.) from the National Science Council, Taiwan, and by the Foundation of Dr. Chen, Jieh-Chen Scholarship, Tainan, Taiwan.

² Address correspondence and reprint requests to Dr. Chao-Liang Wu, Department of Biochemistry and Molecular Biology, National Cheng Kung University Medical College, 1 Dashiue Road, Tainan 701, Taiwan; E-mail address: wumolbio{at}mail.ncku.edu.tw or Dr. Chrong-Reen Wang, Department of Internal Medicine, National Cheng Kung University Medical College, 1 Dashiue Road, Tainan 701, Taiwan; E-mail address: wangcr{at}mail.ncku.edu.tw

³ Abbreviations used in this paper: RA, rheumatoid arthritis; CIA, collagen-induced arthritis; MMP, matrix metalloproteinase; MOI, multiplicity of infection; NLS, nuclear localization signal; OA, osteoarthritis; ProT, prothymosin .

Received for publication March 14, 2006. Accepted for publication January 22, 2007.

	References

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 

1. Firestein, G. S.. 2003. Evolving concepts of rheumatoid arthritis. Nature 423: 356-361. [Medline]
2. Ritchlin, C., E. Dwyer, R. Bucala, R. Winchester. 1994. Sustained and distinctive patterns of gene activation in synovial fibroblasts and whole synovial tissue obtained from inflammatory synovitis. Scand. J. Immunol. 40: 292-298. [Medline]
3. Bucala, R., C. Ritchlin, R. Winchester, A. Cerami. 1991. Constitutive production of inflammatory and mitogenic cytokines by rheumatoid synovial fibroblasts. J. Exp. Med. 173: 569-574. [Abstract/Free Full Text]
4. Zagury, D., R. C. Gallo. 2004. Anti-cytokine Ab immune therapy: present status and perspectives. Drug Discov. Today 9: 72-81. [Medline]
5. Bendele, A., T. McAbee, G. Sennello, J. Frazier, E. Chlipala, D. McCabe. 1999. Efficacy of sustained blood levels of interleukin-1 receptor antagonist in animal models of arthritis: comparison of efficacy in animal models with human clinical data. Arthritis Rheum. 42: 498-506. [Medline]
6. Jou, I. M., A. L. Shiau, S. Y. Chen, C. R. Wang, D. B. Shieh, C. S. Tsai, C. L. Wu. 2005. Thrombospondin 1 as an effective gene therapeutic strategy in collagen-induced arthritis. Arthritis Rheum. 52: 339-344. [Medline]
7. Wang, C. R., S. Y. Chen, C. L. Wu, M. F. Liu, Y. T. Jin, L. Chao, J. Chao. 2005. Prophylactic adenovirus-mediated human kallistatin gene therapy suppresses rat arthritis by inhibiting angiogenesis and inflammation. Arthritis Rheum. 52: 1319-1324. [Medline]
8. Low, T. L., G. B. Thurman, M. McAdoo, J. McClure, J. L. Rossio, P. H. Naylor, A. L. Goldstein. 1979. The chemistry and biology of thymosin. I. Isolation, characterization, and biological activities of thymosin 1 and polypeptide 1 from calf thymus. J. Biol. Chem. 254: 981-986. [Abstract/Free Full Text]
9. Wu, C. L., A. L. Shiau, C. S. Lin. 1997. Prothymosin promotes cell proliferation in NIH3T3 cells. Life Sci. 61: 2091-2101. [Medline]
10. Baxevanis, C. N., A. D. Gritzapis, G. V. Dedoussis, N. G. Papadopoulos, O. Tsolas, M. Papamichail. 1994. Induction of lymphokine-activated killer activity in mice by prothymosin . Cancer Immunol. Immunother. 38: 281-286. [Medline]
11. Baxevanis, C. N., A. D. Gritzapis, G. Spanakos, O. E. Tsitsilonis, M. Papamichail. 1995. Induction of tumor-specific T lymphocyte responses in vivo by prothymosin . Cancer Immunol. Immunother. 40: 410-418. [Medline]
12. Eckert, K., E. Grunberg, P. Immenschuh, F. Garbin, E. D. Kreuser, H. R. Maurer. 1997. Interleukin-2-activated killer cell activity in colorectal tumor patients: evaluation of in vitro effects by prothymosin 1. J. Cancer Res. Clin. Oncol. 123: 420-428. [Medline]
13. Pineiro, A., O. J. Cordero, M. Nogueira. 2000. Fifteen years of prothymosin : contradictory past and new horizons. Peptides 21: 1433-1446. [Medline]
14. Cordero, O. J., H. R. Maurer, M. Nogueira. 1997. Novel approaches to immunotherapy using thymic peptides. Immunol. Today 18: 10-13. [Medline]
15. Baxevanis, C. N., C. Sfagos, E. Anastasopoulos, G. J. Reclos, M. Papamichail. 1990. Prothymosin- enhances HLA-DR antigen expression on monocytes from patients with multiple sclerosis. J. Neuroimmunol. 27: 141-147. [Medline]
16. Baxevanis, C. N., G. J. Reclos, M. Papamichail, G. C. Tsokos. 1987. Prothymosin restores the depressed autologous and allogeneic mixed lymphocyte responses in patients with systemic lupus erythematosus. Immunopharmacol. Immunotoxicol. 9: 429-440. [Medline]
17. Gaubatz, S., A. Meichle, M. Eilers. 1994. An E-box element localized in the first intron mediates regulation of the prothymosin gene by c-myc. Mol. Cell. Biol. 14: 3853-3862. [Abstract/Free Full Text]
18. Qu, Z., C. H. Garcia, L. M. O’Rourke, S. R. Planck, M. Kohli, J. T. Rosenbaum. 1994. Local proliferation of fibroblast-like synoviocytes contributes to synovial hyperplasia: results of proliferating cell nuclear antigen/cyclin, c-myc, and nucleolar organizer region staining. Arthritis Rheum. 37: 212-220. [Medline]
19. Volin, M. V., A. E. Koch. 2000. Cell cycle implications in the pathogenesis of rheumatoid arthritis. Front. Biosci. 5: D594-D601. [Medline]
20. Pap, T., M. Nawrath, J. Heinrich, M. Bosse, A. Baier, K. M. Hummel, P. Petrow, S. Kuchen, B. A. Michel, R. E. Gay, et al 2004. Cooperation of Ras- and c-Myc-dependent pathways in regulating the growth and invasiveness of synovial fibroblasts in rheumatoid arthritis. Arthritis Rheum. 50: 2794-2802. [Medline]
21. Bodey, B., B. Bodey, Jr, S. E. Siegel, H. E. Kaiser. 2000. Review of thymic hormones in cancer diagnosis and treatment. Int. J. Immunopharmacol. 22: 261-273. [Medline]
22. Wolf, R. E., V. C. Hall. 1987. Interleukin-2 production and effect of thymosin fraction 5 on interleukin-2 production in rheumatoid arthritis. J. Rheumatol. 14: 898-901. [Medline]
23. Cao, D., V. Malmstrom, C. Baecher-Allan, D. Hafler, L. Klareskog, C. Trollmo. 2003. Isolation and functional characterization of regulatory CD25^brightCD4⁺ T cells from the target organ of patients with rheumatoid arthritis. Eur. J. Immunol. 33: 215-223. [Medline]
24. Ehrenstein, M. R., J. G. Evans, A. Singh, S. Moore, G. Warnes, D. A. Isenberg, C. Mauri. 2004. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNF therapy. J. Exp. Med. 200: 277-285. [Abstract/Free Full Text]
25. Zatz, M. M., J. Oliver, A. L. Goldstein, C. Novak, R. P. Jacobs. 1984. Suppressor cell responses in patients with rheumatoid arthritis: the effect of thymosin. Thymus 6: 205-218. [Medline]
26. Manrow, R. E., A. R. Sburlati, J. A. Hanover, S. L. Berger. 1991. Nuclear targeting of prothymosin . J. Biol. Chem. 266: 3916-3924. [Abstract/Free Full Text]
27. Shiau, A. L., P. R. Lin, M. Y. Chang, C. L. Wu. 2001. Retrovirus-mediated transfer of prothymosin gene inhibits tumor growth and prolongs survival in murine bladder cancer. Gene Ther. 8: 1609-1617. [Medline]
28. Hitt, M., A. J. Bett, L. Prevec, F. L. Graham. 1998. Construction and propagation of human adenovirus vectors. J. E. Celis, Jr, ed. Cell Biology: A Laboratory Handbook 479-490. Academic Press, San Diego.
29. Li, K. J., A. L. Shiau, Y. Y. Chiou, Y. T. Yo, C. L. Wu. 2005. Transgenic overexpression of prothymosin induces development of polycystic kidney disease. Kidney Int. 67: 1710-1722. [Medline]
30. Bristow, C. A., P. Shore. 2003. Transcriptional regulation of the human MIP-1 promoter by RUNX1 and MOZ. Nucleic Acids Res. 31: 2735-2744. [Abstract/Free Full Text]
31. Malo, M. S., M. Abedrapo, A. Chen, M. Mozumder, P. Pushpakaran, F. Alkhoury, W. Zhang, E. Fleming, R. A. Hodin. 2003. Improved eukaryotic promoter-detection vector carrying two luciferase reporter genes. BioTechniques 35: 1150-1152, 1154. [Medline]
32. Woods, J. M., K. J. Katschke, M. V. Volin, J. H. Ruth, D. C. Woodruff, M. A. Amin, M. A. Connors, H. Kurata, K. Arai, G. K. Haines, et al 2001. IL-4 adenoviral gene therapy reduces inflammation, proinflammatory cytokines, vascularization, and bony destruction in rat adjuvant-induced arthritis. J. Immunol. 166: 1214-1222. [Abstract/Free Full Text]
33. Asahara, H., T. Hasunuma, T. Kobata, H. Inoue, U. Muller-Ladner, S. Gay, T. Sumida, K. Nishioka. 1997. In situ expression of protooncogenes and Fas/Fas ligand in rheumatoid arthritis synovium. J. Rheumatol. 24: 430-435. [Medline]
34. Wolpe, S. D., G. Davatelis, B. Sherry, B. Beutler, D. G. Hesse, H. T. Nguyen, L. L. Moldawer, C. F. Nathan, S. F. Lowry, A. Cerami. 1988. Macrophages secrete a novel heparin-binding protein with inflammatory and neutrophil chemokinetic properties. J. Exp. Med. 167: 570-581. [Abstract/Free Full Text]
35. Menten, P., A. Wuyts, J. Van Damme. 2002. Macrophage inflammatory protein-1. Cytokine Growth Factor Rev. 13: 455-481. [Medline]
36. Katrib, A., P. P. Tak, J. V. Bertouch, C. Cuello, H. P. McNeil, T. J. M. Smeets, M. C. Kraan, P. P. Youssef. 2001. Expression of chemokines and matrix metalloproteinases in early rheumatoid arthritis. Rheumatology 40: 988-994. [Abstract/Free Full Text]
37. Koch, A. E., S. L. Kunkel, L. A. Harlow, D. D. Mazarakis, G. K. Haines, M. D. Burdick, R. M. Pope, R. M. Strieter. 1994. Macrophage inflammatory protein-1: a novel chemotactic cytokine for macrophages in rheumatoid arthritis. J. Clin. Invest. 93: 921-928. [Medline]
38. Kasama, T., R. M. Strieter, N. W. Lukacs, P. M. Lincoln, M. D. Burdick, S. L. Kunkel. 1995. Interleukin-10 expression and chemokine regulation during the evolution of murine type II collagen-induced arthritis. J. Clin. Invest. 95: 2868-2876. [Medline]
39. Shahrara, S., M. A. Amin, J. M. Woods, G. K. Haines, A. E. Koch. 2003. Chemokine receptor expression and in vivo signaling pathways in the joints of rats with adjuvant-induced arthritis. Arthritis Rheum. 48: 3568-3583. [Medline]
40. Kagari, T., H. Doi, T. Shimozato. 2002. The importance of IL-1 and TNF-, and the noninvolvement of IL-6, in the development of monoclonal antibody-induced arthritis. J. Immunol. 169: 1459-1466. [Abstract/Free Full Text]
41. Sosne, G., C. C. Chan, K. Thai, M. Kennedy, E. A. Szliter, L. D. Hazlett, H. K. Kleinman. 2001. Thymosin 4 promotes corneal wound healing and modulates inflammatory mediators in vivo. Exp. Eye Res. 72: 605-608. [Medline]
42. Hiltunen, M. O., M. Ruuskanen, J. Huuskonen, A. J. Mahonen, M. Ahonen, J. Rutanen, V. M. Kosma, A. Mahonen, H. Kroger, S. Yla-Herttuala. 2003. Adenovirus-mediated VEGF-A gene transfer induces bone formation in vivo. FASEB J. 17: 1147-1149. [Abstract/Free Full Text]
43. Gowen, M., K. Chapman, A. Littlewood, D. Hughes, D. Evans, G. Russell. 1990. Production of tumor necrosis factor by human osteoblasts is modulated by other cytokines, but not by osteotropic hormones. Endocrinology 126: 1250-1255. [Abstract]
44. Swolin-Eide, D., C. Ohlsson. 1998. Effects of cortisol on the expression of interleukin-6 and interleukin-1 in human osteoblast-like cells. J. Endocrinol. 156: 107-114. [Abstract]
45. Chae, H. J., S. C. Kim, S. W. Chae, N. H. An, H. H. Kim, Z. H. Lee, H. R. Kim. 2001. Blockade of the p38 mitogen-activated protein kinase pathway inhibits inducible nitric oxide synthase and interleukin-6 expression in MC3T3E-1 osteoblasts. Pharmacol. Res. 43: 275-283. [Medline]
46. Dovio, A., M. L. Sartori, R. G. Masera, L. Peretti, L. Perotti, A. Angeli. 2004. Effects of physiological concentrations of steroid hormones and interleukin-11 on basal and stimulated production of interleukin-8 by human osteoblast-like cells with different functional profiles. Clin. Exp. Rheumatol. 22: 79-84. [Medline]
47. Taichman, R. S., M. J. Reilly, L. S. Matthews. 2000. Human osteoblast-like cells and osteosarcoma cell lines synthesize macrophage inhibitory protein 1 in response to interleukin 1 and tumor necrosis factor stimulation in vitro. Br. J. Haematol. 108: 275-283. [Medline]
48. Muruve, D. A.. 2004. The innate immune response to adenovirus vectors. Hum. Gene Ther. 15: 1157-1166. [Medline]
49. Kinne, R. W., R. Brauer, B. Stuhlmuller, E. Palombo-Kinne, G. R. Burmester. 2000. Macrophages in rheumatoid arthritis. Arthritis Res. 2: 189-202. [Medline]
50. Kawanaka, N., M. Yamamura, T. Aita, Y. Morita, A. Okamoto, M. Kawashima, M. Iwahashi, A. Ueno, Y. Ohmoto, H. Makino. 2002. CD14⁺, CD16⁺ blood monocytes and joint inflammation in rheumatoid arthritis. Arthritis Rheum. 46: 2578-2586. [Medline]
51. Szekanecz, Z., G. Szucs, S. Szanto, A. E. Koch. 2006. Chemokines in rheumatic diseases. Curr. Drug Targets 7: 91-102. [Medline]
52. Tetlow, L. C., M. Lees, Y. Ogata, H. Nagase, D. E. Woolley. 1993. Differential expression of gelatinase B (MMP-9) and stromelysin-1 (MMP-3) by rheumatoid synovial cells in vitro and in vivo. Rheumatol. Int. 13: 53-59. [Medline]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Shiau, A.-L. Articles by Wu, C.-L. Search for Related Content PubMed PubMed Citation Articles by Shiau, A.-L. Articles by Wu, C.-L.