The PCI-37B metastatic SCCHN cell line, which strongly expresses CCR7, was a kind gift from the University of Pittsburgh (12). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen, Carlsbad, CA, USA), which contained 10% fetal bovine serum (Gibco-BRL Corporation, Grand Island, NY, USA), 100 U/ml penicillin and 100 U/ml streptomycin. When inhibitors were used, treatment at the dose used did not affect cell viability or CCR7 expression.

SCCHN tissue specimens were obtained from 75 patients by biopsy prior to chemotherapy or radiotherapy at the Department of Oral and Maxillofacial Surgery, School of Stomatology, China Medical University. The term ‘metastatic’ in the present study refers to patients with positive lymph nodes that were recognized at initial presentation or later based on histopathological diagnosis following neck dissection. The classification of SCCHN, including primary tumors (T), regional lymph nodes (N), distant metastasis (M) and stage grouping, was determined according to the regulations of the Union for International Cancer Control (UICC) for Head and Neck Cancer [tumor-node-metastasis (TNM) classification, 1997]. Ten samples of normal tissues adjacent to the benign tumor were chosen as controls. The study protocol was approved by the Medical Ethics Committee of the Affiliated Stomatological Hospital of China Medical University and performed according to the declaration of Helsinki. All the specimens were obtained with the consent of the patients prior to surgery and in accordance with the Health Insurance Portability. Written informed consent was obtained from all individuals.

The CCR7 chemokine ligand, CCL19 (MIP-3β) was purchased from R&D Systems (Minneapolis, MN, USA). Mouse anti-CCR7 antibody was purchased from BD Biosciences (San Jose, CA, USA). C3 exoenzyme (Rho inhibitor) was purchased from Cytoskeleton Inc. (Denver, CO, USA) and Y-27632 (ROCK inhibitor) was purchased from Sigma (Santa Clara, CA, USA). Tyrphostin A9 (Pyk2 inhibitor) was purchased from Calbiochem (San Diego, CA, USA). Anti-RhoA, anti-phospho-Pyk2 (Tyr402), anti-phospho-cofilin (Ser3) and anti-phospho-myosin light chain (Ser19) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Bio-Rad protein assay dye reagent was purchased from Bio-Rad Laboratories (Richmond, CA, USA). Enhanced chemiluminescence was purchased from Amersham Pharmacia Biotechnology (Piscataway, NJ, USA).

Immunohistochemical staining used conventional horseradish peroxidase immunohistochemical staining methods. Briefly, 5-μm sections of the specimens were deparaffinized and hydrated with 0.6% H₂O₂ in methanol to inhibit endogenous peroxidase. Antigen retrieval was performed and slides were incubated with normal blocking serum for 10 min. Sections were incubated with primary antibodies (1:100 dilution): CCR7-specific monoclonal antibody, and rabbit anti-RhoA polyclonal antibody overnight at 4°C. Immunodetection was performed using peroxidase-labeled secondary antibody (R&D Systems) and diaminobenzidine for visualization. The sections were counterstained with hematoxylin (Sigma). Negative controls included omission of the primary antibody. Cell morphology was analyzed by microscopy (Nikon Eclipse 80i; Tokyo, Japan) at ×100–400 magnification. According to the percentage of positive tumor cells, the cells were scored as negative (−), <10% or no staining; weak positive (+), 11–50%; positive (++), 51–75%; or strongly positive (+++), >75%.

Chemotaxis in response to chemokines was measured by counting cells migrating through a polycarbonate filter (8-μm pore size) in 24-well Transwell chambers in triplicate in DMEM with 0.5% (w/v) BSA (Invitrogen). Cell suspensions (2×10⁵ cells/200 μl) were placed in the top chamber of the filter. Aliquots of chemokines were added to wells. After 24 h, the cells in each lower well were counted under a light microscope in at least five different fields (original magnification, ×200). Means ± SD were recorded for each condition, and an index was calculated based on the control involving random migration.

Cell invasion was quantified in vitro using matrigel-coated semipermeable, modified inserts with a pore size of 8-μm. Analysis of the invasion assay was performed as described in the chemotaxis assay incubated with CCL19 for 36 h. An invasion index was normalized to non-specific cell invasion (from media-pulsed wells).

Migration of PCI-37B cells was measured using an in vitro monolayer wound assay (20). Cells grown in 24-well plates in DMEM to confluence were scraped with a pipette tip to create a cell-free area. Wounded monolayers were washed with serum-free media. The cells were pretreated with/without RhoA inhibitor C3 exoenzyme (50 ng/ml), Y-27632 ROCK inhibitor (10 M) and CCL19 (200 ng/ml) for 1 h. Wound closure was followed after 0 and 24 h. The wound-healing effect was calculated as the percentage of the remaining cell-free area compared to the initial wound area (arbitrarily set as 100%). Monolayer wound assays included serum-free DMEM controls. The wounds were observed by phase contrast microscopy (Nikon TE2000-S Eclipse) at ×200 magnification and images captured were documented with photography. Each experiment was performed in triplicate, analyzing four or five scratches per well. The relative cell-free area was calculated based on the control group.

RhoA-GTP was measured with recombinant purified glutathione-S-transferase rhotekin Rho-binding domain (GST-TRBD) bound to glutathione beads and purified as previously described (21). SCCHN cells were stimulated with CCL19 in serum-free medium lysed with ice-cold RIPA buffer and cleared by centrifugation. SCCHN cell lysates were incubated for 1 h at 4°C with GST or GST-TRBD coupled to glutathione-sepharose beads. Beads were then washed and boiled in SDS sample buffer and precipitates were analyzed with immunoblotting using an anti-RhoA antibody and then developed using chemiluminescence.

Cells were harvested in a lysis buffer (10 mM Tris-HCl, pH 7.6, 50 mM NaF, 1 mM NaV₃O₄, 1% Triton X-100 and 1X protease inhibitor of protein tyrosine phosphatases). Lysates were sonicated for 3 sec and centrifuged at 4°C at 12,000 rpm for 30 min. The supernatant was collected for protein quantification using the Bio-Rad protein assay dye reagent. Protein (50 mg) was size-fractionated through a 10% SDS-PAGE gel and transferred onto nitrocellulose filters which were blocked (1% non-fat dry milk, 0.1% Triton X-100, 150 mM NaCl, 50 mM Tris, pH 7.5) and incubated with primary antibody (1:1,000 dilution). Nitrocellulose filters were incubated with horseradish peroxidase-conjugated secondary antibodies and bands were visualized with enhanced chemiluminescence and quantified by scanning densitometry using ImageJ software.

Experiments were run in triplicate and repeated at least three times. Numerical data were expressed as means ± standard deviation (SD). Correlations were analyzed using the Spearman’s and χ² tests. Statistical differences between two groups were evaluated using an unpaired Student’s t-test. P<0.05 was considered to indicate a statistically significant result. Statistical analyses were performed with SPSS 13.0 software.

Immunohistochemistry was utilized to examine CCR7 and RhoA expression in SCCHN tumor tissues, metastatic and normal lymph nodes and oral mucosal tissues. CCR7 and RhoA were evident in the cell membrane and cytoplasm, and were mainly expressed in the stromal surroundings of tumors and metastatic lymph node cells. Stained cells were few or absent in normal lymph nodes and oral mucosal tissues (Fig. 1). CCR7 and RhoA expression was significantly correlated with cervical lymph-node metastasis and SCCHN clinical stages (P<0.05; Tables I and II). Spearman’s test indicated a significant positive correlation between the expression of CCR7 and RhoA (P<0.05; Table III).

Since CCR7 regulates actin organization (22) and the latter can be controlled by small GTPase Rho (16), we investigated whether CCR7 induces the activation of RhoA in an SCCHN cell line (PCI-37B). We measured activated RhoA (GTP-bound form) with Rho-binding domain (RBD) pull-down assays using the Rho effector protein, Rhotekin (23). Limited active bound RhoA was observed in non-stimulated cells, and this was increased after CCL19 stimulation. Although total RhoA remained unchanged, activated RhoA increased significantly 10 min after CCL19 treatment compared with the controls (P<0.05), whereas CCR7mAb, a C3-exo-enzyme, partly inhibited the CCL19-induced activation of RhoA compared with the CCL19-stimulation group (P<0.05; Fig. 2).

Using chemotaxis and matrigel invasion assay we analyzed the role of RhoA in PCI-37B migration and invasiveness in response to CCL19. The presence of CCL19 in the lower chamber stimulated a >1.5-fold increase in chemotaxis and invasion of PCI-37B across the filter membrane while C3 and Y-27632 blocked this activity (P<0.05; Figs. 3 and 4). In addition, we evaluated the migration ability of PCI-37B induced by CCR7 using a scrape wound-healing assay. The width of the wound, which was significantly reduced after 24 h in the presence of CCL19, and this was blocked by C3 and Y-27632 (P<0.05; Fig. 5). Data show that RhoA/ROCK has a role in the invasiveness of PCI-37B mediated by CCR7.

Western blotting indicated that CCL19-pretreated PCI-37B cells had an increased expression of p-Pyk2 (Fig. 6) and significantly decreased expression of p-cofilin (Fig. 7), which was reversed by CCR7mAb, suggesting CCR7-induced activation of Pyk2 and cofilin. As expected, CCL19-dependent Pyk2 phosphorylation and cofilin dephosphorylation was abrogated in the presence of RhoA, ROCK and Pyk2 inhibitors, suggesting that RhoA/ROCK may mediate the activation of Pyk2 and cofilin induced by CCL19. The results showed that MLC protein was not changed after CCL19 stimulation, indicating that CCR7 has no effect on MLC (Fig. 8).