Tubocapsanolide A Inhibits Transforming Growth Factor-β-activating Kinase 1 to Suppress NF-κB-induced CCR7*

Withanolides are C28 steroidal lactones isolated from plants that exhibit potent anti-cancer activity. The chemokine receptor CCR7 is important for lymphatic invasion of cancer cells and is overexpressed in metastatic breast cancer cells. A bioactive withanolide tubocapsanolide A (Tubo A) suppressed NF-κB-mediated CCR7 expression in breast cancer cells and attenuated their migration toward lymphatic endothelial cells. Chromatin immunoprecipitation assay confirmed that binding of NF-κBto the consensus site localized at the –398/–389 of human CCR7 promoter was repressed by Tubo A. Tubo A inhibited IκB kinase (IKK) and p38 kinase and downstream mitogen and stress-activated protein kinase 1 (MSK1) activity to reduce IκB degradation and to suppress NF-κB activation. Co-expression of IKK and MSK1 fully rescued Tubo A-induced inhibition. In addition, ectopic expression of transforming growth factor-β-activating kinase (TAK1), the common upstream kinase of IKK and MSK1, also completely reversed the inhibition by Tubo A. Most importantly, Tubo A reduced NF-κB activation, CCR7 expression, and lymph node metastasis of breast cancer in vivo. We conclude that Tubo A inhibits TAK1 to repress NF-κB-induced CCR7 expression in breast cancer cells and suggest that Tubo A may be useful for the prevention of lymphatic invasion of breast cancer cells.


Withanolides are C 28 steroidal lactones isolated from plants that exhibit potent anti-cancer activity. The chemokine receptor CCR7 is important for lymphatic invasion of cancer cells and is overexpressed in metastatic breast cancer cells. A bioactive withanolide tubocapsanolide A (Tubo A) suppressed NF-Bmediated CCR7 expression in breast cancer cells and attenuated their migration toward lymphatic endothelial cells. Chromatin immunoprecipitation assay confirmed that binding of NF-B to the consensus site localized at the ؊398/؊389 of human CCR7 promoter was repressed by Tubo A. Tubo A inhibited IB kinase (IKK) and p38 kinase and downstream mitogen and stress-activated protein kinase 1 (MSK1) activity to reduce IB degradation and to suppress NF-B activation. Co-expression of IKK and MSK1 fully rescued Tubo A-induced inhibition. In addition, ectopic expression of transforming growth factor-␤-activating kinase (TAK1), the common upstream kinase of IKK and MSK1, also completely reversed the inhibition by Tubo A. Most importantly, Tubo A reduced NF-B activation, CCR7 expression, and lymph node metastasis of breast cancer in vivo. We conclude that Tubo A inhibits TAK1 to repress NF-B-induced CCR7 expression in breast cancer cells and suggest that Tubo A may be useful for the prevention of lymphatic invasion of breast cancer cells.
Withanolides are steroidal lactones that were originally isolated from Withania somnifera, one of the most important herbs used as a traditional remedy for several illnesses in Asian countries (1). These compounds are biologically active and have been shown to inhibit the enzymatic activity of cyclooxygenase-2 and suppress inflammation (2). In addition, recent studies demonstrate that withanolides exhibit anti-cancer effect on human lung, colon, and breast cancer cells in vitro and exert immunopotentiating activity in vivo (3,4). Several potential mechanisms have been implicated in the inhibition of tumorigenesis by withanolides. First, withanolides induce growth arrest and apoptosis in cancer cells (4,5). Second, withanolides can inhibit angiogenesis by suppressing endothelial cell proliferation (6). Third, withanolides can reduce cancer cell invasion and metastasis (7). These data suggest that withanolides may be developed as a novel class of anti-cancer drugs.
Lymph node invasion by cancer cells is an important step for tumor metastasis and is frequently correlated with early recurrence and poor prognosis. However, the underlying mechanism by which cancer cells metastasize into peripheral lymphatic capillaries is poorly defined. Recent studies indicate that the interaction between chemokine CCL21 and its cognate receptor CCR7 may play an important role in this process (8,9). The hypothesis suggests that lymphatic endothelial cells (LECs) 2 express and release chemotactic factors such as CCL19 and CCL21 to direct cancer cells with high expression of chemokine receptor CCR7 to grow and migrate toward lymphatic capillaries (10 -12). Indeed, many highly metastatic cancer cells express large amount of CCR7 receptor (13,14), and CCR7 expression is associated with lymph node metastasis in breast, lung, gastric, esophageal cancer, and melanoma (15)(16)(17)(18)(19). It is possible that inhibition of CCR7 expression or block of CCL21-CCR7 interaction may cause reduction of lymph node invasion and tumor metastasis. However, the control of CCR7 gene transcription in cancer cells is largely unclear, and no natural products have been shown to regulate CCR7 expression.
We have isolated a new bioactive withanolide tubocapsanolide A (Tubo A) from Tubocapsicum anomalum (20). Tubo A exhibited cytotoxic activity on various types of human cancer cells (20). In addition, our recent results demonstrated that Tubo A suppressed the transcription of Skp2 oncogene and up-regulated cyclin-dependent kinase inhibitory proteins p27 * This study was supported by the National Sun Yat-Sen University-Kaohsiung Medical University Joint Research Center and Center for Gene Regulation and Signal Transduction Research, National Cheng Kung University (to W.-C. H.). 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. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1-S6. 1 To whom correspondence should be addressed. and p21 to inhibit proliferation of human lung cancer cells (21). In this study, we address the following objectives: (a) the effect of Tubo A on the expression of CCR7 in breast cancer cells, (b) the effect of Tubo A on lymphatic migration of breast cancer cells, and (c) the underlying mechanism by which Tubo A regulates CCR7.

EXPERIMENTAL PROCEDURES
Plant Material-The initial collection of T. anomalum (Solanaceae) was made on July 2003 near NanTao County and identified by Dr. Hsin-Fu Yen (National Museum of Natural Science, Taichung, Taiwan). A larger amount of the same plant was collected at the Da-Han Mountain, Kaohsiung, in October 2004, and identified by Dr. Ming-Ho Yen (Graduate Institute of Natural Products, Kaohsiung Medical University, Taiwan). The samples were authenticated and deposited in the Graduate Institute of Natural Products, Kaohsiung Medical University, Taiwan. Extraction and isolation of Tubo A were performed as described previously (20).
In Vitro Invasion Assay-In vitro invasion assay was performed as described previously (22). 3000 cells in 100 l of medium with vehicle or Tubo A (0.5 M) were placed in the upper part of the Transwell unit and allowed to invade for 24 h. The lower part of the Transwell unit was filled with LECs. After incubation, invaded cells on the bottom surface of the membrane were fixed in formaldehyde, stained with Giemsa solution, and counted under a microscope. In some experiments, cells transfected with various expression vectors were collected by trypsinization and placed into the upper well in the absence or presence of Tubo A for invasion assay.
Immunoblotting-Extraction of cellular proteins and immunoblotting were performed as previously described (23).
Preparation of Cytoplasmic and Nuclear Fractions-Cells were plated in 75-cm 2 flasks and grown to 70 -80% confluence. Treated cells were washed with cold phosphate-buffered saline and lysed in hypotonic buffer (10 mM HEPES-KOH, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2 , 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride). Nuclei were spun down, and the supernatant was collected as the cytoplasmic fraction. Nuclei were further extracted in cold buffer (20 mM HEPES-KOH, pH 7.9, 25% glycerol, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride) on ice for 20 min. Cellular debris was clarified by centrifugation, and the supernatant was collected as nuclear fraction.
Flow Cytometry-For detection of CCR7 expression, cells were washed with ice-cold phosphate-buffered saline and incubated with biotinylated anti-CCR7 antibody at 4°C for 30 min. After washing, cells were incubated with fluorescein isothiocyanate-conjugated avidin and subjected to flow cytometric analysis as described previously (24).
Promoter Activity Assay-In brief, cells were plated onto 6-well plates at the density of 100,000 cells/well and grown overnight. Cells were transfected with 1 g of NF-B or AP-1 luciferase reporter plasmid. After transfection, cells were treated with vehicle (0.1% DMSO) or various concentrations of Tubo A in 10% FCS medium for 24 h. Promoter activity was determined and normalized for the concentration of cellular proteins.
Analysis of Protein Stability-Protein stability of IB was measured by blocking protein synthesis with cycloheximide and harvesting the cells at various times. Cells were incubated with 10% FCS medium containing vehicle or various concentrations of Tubo A (0 or 0.5 M) for 24 h. Cells were treated with 10 g/ml cycloheximide, and cellular proteins were harvested at various times. Protein level of IB was determined by immunoblotting.
In Vivo Xenograft Study-We tested the anti-metastatic ability of Tubo A in nude mice and animal studies were approved by the Animal Care and Ethics Committee of the National Sun Yat-Sen University. Female BALB/cAnN-Foxn1 null mice (5 weeks old) were obtained from National Laboratory Animal Center (Taipei, Taiwan). MDA-MB-231 cells (1 ϫ 10 6 cells/ mice) were injected into left second thoracic mammary fat pad of nude mice. Tumor volumes were measured every 3 days from the second week after injection and were calculated using the formula, V ϭ (length) ϫ (width) 2 ϫ 0.5. After 4 weeks, tumors grew to ϳ0.18 -0.2 cm 3 . Mice were randomly divided into two groups (n ϭ 5 for each group) and subjected to treatment. Animals of the control group received intraperitoneal injection of 0.1% of DMSO every 3 days. On the other hand, animals of the treatment group received intraperitoneal injection of Tubo A (4 mg/kg) every 3 days. Animal were sacrificed after continuous treatment for 21 days, and the tumors were excised and subjected for RT-PCR analysis. Left axillary, brachial, and inguinal lymph nodes (defined as proximal lymph nodes because they located at the same side of tumor injection site) were excised and fixed in 4% paraformaldehyde. In addition, right axillary, brachial, and inguinal lymph nodes (defined as distal lymph nodes because they located at the counter side of tumor injection site) were also excised. All lymph nodes were embedded in paraffin using route procedures and subjected for immunohistochemical analysis.
Immunohistochemical Analysis-Lymph nodes were cut into 3-m sections, and paraffin was removed by xylene. Tissues were placed in citrate buffer and incubated at 85°C for 5 min for antigen retrieval. Tissues were blocked for 15 min with a 3% hydrogen peroxide solution to inhibit endogenous peroxidase activity and washed with phosphate-buffered saline. Tissues were probed with antibodies directed against lymphatic marker LYVE-1 for the confirmation of lymphatic tissues or epithelial marker EpCAM for the identification of lymphatic invasion of breast cancer cells. EpCAM (also referred as CD326) is a glycoprotein of ϳ40 kDa that was originally identified as a marker for carcinoma, attributable to its high expression on rapidly proliferating tumors of epithelial origin (25). MDA-MB-231 cells have been shown to expressed a high level of EpCAM (26), whereas lymph node tissues do not express this epithelial marker (27). Bound primary antibodies were detected with horseradish peroxidaseconjugated secondary antibody and then developed by diaminobenzidine substrate. Finally, sections were co-stained with hematoxylin. In addition, tissues were also subjected to hematoxylin & eosin stain for identify tumor metastasis. We also addressed the effect of Tubo A on NF-B activation in vivo. Tumor tissues were stained with anti-NF-B antibody, and the cell number with positive staining was counted for each tumor section. Percentages of positive staining cells with nuclear signal, which indicated nuclear translocation and activation of NF-B, were determined and compared between control and the Tubo A-treated group.
Statistical Analysis-Student's t test was used to evaluate the difference between various experimental groups. Differences were considered to be significant at p Ͻ 0.05.

Tubo A Inhibited CCR7 Expression in Breast Cancer Cells and Attenuated
Their Migration toward LECs-The chemical structure of Tubo A is shown in supplemental Fig. S1. We found that CCR7 was highly expressed in metastatic MDA-MB-231 cells (Fig. 1A). Conversely, CCR7 expression was low in nonmetastatic MCF-7 cells. Flow cytometric analysis demonstrated a positive correlation between CCR7 mRNA and protein expression in these two breast cancer cell lines (Fig. 1A). Because MDA-MB-231 cells expressed high level of CCR7, we used this cell line to study the regulation of CCR7 and lymphatic invasion by Tubo A. Our previous studies have demonstrated that Tubo A at the concentration higher than 0.5 M showed significant cytotoxic activity on various human cancer cell lines (20,21). Because we addressed the anti-metastatic effect of Tubo A in this study, we consistently used low doses (0 -0.5 M) of Tubo A to treat cells to prevent the nonspecific effect caused by cell death. Tubo A at the concentrations of 0.125, 0.25, and 0.5 M inhibited CCR7 expression in a dose-and time-dependent manner (Fig. 1B). In addition, CCR7 mRNA level was also down-regulated by Tubo A in a dose-dependent manner (Fig. 1C). In addition to inhibit basal CCR7 expression in MDA-MB-231 cells, Tubo A also repressed tumor necrosis factor-␣induced CCR7 expression (supplemental Fig. S2). Therefore, Tubo A inhibited both basal and extracellular signal-stimulated CCR7 expression. We also compared the effect of another bioactive withanolide Withaferin A on CCR7 expression and found that these two withanolides exhibited similar potency on the inhibition of CCR7 (supplemental Fig. S3). We next tested whether CCR7 was required for the migration of breast cancer cells toward LECs. Primary cultured LECs used in these assays expressed high levels of CCL21 (supplemental Fig.  S4). LECs were placed on the lower part of a Transwell unit, and MDA-MB-231 cells were added on the upper part. As shown in Fig. 1D, Tubo A suppressed the invasion of MDA-MB-231 cells toward LECs in a dose-dependent manner. Our previous results also showed that knockdown of CCR7 expression in MDA-MB-231 cells reduced their migration toward LECs (28). Our data suggested that Tubo A inhibited CCR7 expression in breast cancer cells and reduced their migration toward LECs.
Tubo A Attenuated IB Degradation and NF-B (p65) Nuclear Translocation-Because CCR7 has been found to be a target gene for NF-B (29) and a bioactive withanolide withaferin A has been shown to inhibit NF-B-mediated gene transcription (30), we tested the effect of Tubo A on the IB degradation and NF-B activation. As shown in Fig. 2A, Tubo A up-regulated IB protein level in MDA-MB-231 cells. Pulsechase analysis indicated that protein stability of IB was increased by Tubo A in a dose-dependent manner and Tubo A at the concentration of 0.5 M effectively reduced IB degradation ( Fig. 2A). On the contrary, the protein level of NF-B was  not significantly changed. Increase of IB might sequester NF-B in cytoplasm to attenuate gene transcription. Therefore, we investigated the effect of Tubo A on nuclear translocation of NF-B. Fig. 2B showed that NF-B was mainly located in the nucleus of MDA-MB-231 cells possibly due to constitutive activation of some upstream signaling pathways, and this nuclear accumulation was attenuated by Tubo A. On the contrary, Tubo A treatment induced increase of cytoplasmic NF-B. To verify that NF-B in the nucleus was active, we used an antibody that specifically recognized the phosphorylated (at Ser-276) NF-B to perform immunoblotting and confirmed that phosphorylated NF-B was located in the nucleus of DMSO-treated cells while Tubo A treatment reduced nuclear accumulation (Fig. 2B). In agreement with these data, the NF-B-dependent reporter activity in MDA-MB-231 cells was also inhibited by Tubo A in a dose-dependent manner (Fig. 2C). This effect is NF-B-specific, because Tubo A did not significantly inhibit AP-1-mediated reporter activity under the same experimental condition. Analysis of the human CCR7 promoter sequence revealed two potential NF-B binding sites at Ϫ797/Ϫ787 and Ϫ398/Ϫ389 regions. Prediction of potential NF-B consensus sites in human CCR7 promoter had also been reported by Hopken et al. (29). We studied the in vivo binding of NF-B to CCR7 promoter by ChIP assays and found that binding of NF-B to the Ϫ797/Ϫ787 consensus sequence was very weak (data not shown). Conversely, NF-B constitutively bound to the Ϫ398/ Ϫ389 region of CCR7 promoter, and its binding was attenuated by Tubo A (Fig. 2D). Binding of NF-B to this consensus site was specific, because knockdown of NF-B by short hairpin RNA reduced the binding of NF-B to this site while a control short hairpin RNA targeting luciferase had no effect. Because no Myc binding site was found in the CCR7 promoter region amplified by PCR primers used in our study, addition of anti-Myc antibody did not get any nonspecific signal, which further verified the specificity of our ChIP assays (Fig. 2D). These data indicated that Tubo A repressed CCR7 by suppressing NF-B activation.

Ectopic Expression of IKK-␤ Only Partly Counteracted Tubo A-induced Down-regulation of CCR7-
Our aforementioned data suggested that Tubo A reduced IB degradation to attenuate NF-B-mediated transcription of CCR7. The main upstream kinase that phosphorylates IB and controls its protein stability is IKK, so we tested the effect of Tubo A on IKK activity. Phosphorylation on Ser-177/181 of IKK-␤ or Ser-176/180 on IKK-␣ is an indication for kinase activation. We used anti-phospho-IKK antibody to  investigate the activation of IKK, and our data indicated that IKK-␣ and -␤ activity were attenuated by Tubo A (Fig. 3A). Ectopic expression of IKK-␤ in MDA-MB-231 cells increased IKK-␤ activity and induced degradation of IB in these cells (Fig. 3B). In addition, overexpression of IKK-␤ increased the binding of NF-B to CCR7 promoter (Fig. 3C). However, reduction of NF-B promoter binding by Tubo A was only partly reversed. Moreover, Tubo A-induced down-regulation of CCR7 protein on the cell surface was only partly reversed by IKK-␤ (Fig. 3D). Similar results were observed when IKK-␣ was ectopically expressed in MDA-MB-231 cells (data not shown). Our data suggested that IKK was not the only one target for Tubo A to inhibit CCR7.
Tubo A Inhibited p38 and MSK1 Activation and Reduced NF-B Phosphorylation-Because enhancement of IB degradation by overexpression of IKK could not fully rescue Tubo A-induced inhibition of NF-B, we thought that Tubo A affected other signaling pathways that might directly regulate NF-B phosphorylation and activation. By using phospho-specific antibodies, we found that Tubo A suppressed the activation of p38, but not extracellular signal-regulated kinase (ERK) in MDA-MB-231 cells (Fig. 4A). Phosphorylation status of c-Jun N-terminal kinase (JNK) was not changed by Tubo A (data not shown). A main downstream effector kinase of p38, which can directly phosphorylate NF-B, is MSK1. We found that phosphorylation of MSK1 on Ser-581, a major phosphorylation site for p38, was also reduced by Tubo A (Fig. 4B). NF-B was phosphorylated by MSK1 on Ser-276 (31), and we found that Tubo A reduced Ser-276 phosphorylation of NF-B (Fig. 4B). So, p38 and MSK1 were involved in the inhibition of CCR7 by Tubo A. Ectopic expression of MSK1 partly reversed Tubo A-induced inhibition of NF-B binding to CCR7 promoter and CCR7 expression (Fig. 4,  C and D).

Co-activation of IKK-␤ and MSK1 Fully Rescued the Inhibition of NF-B Promoter Binding, CCR7 Expression, and Cell Migration by
Tubo A-We next tested whether combinational expression of IKK-␤ and MSK1 might fully counteract the inhibition of Tubo A. Our data showed that co-expression of IKK-␤ and MSK1 totally reversed Tubo A-induced inhibition of NF-B binding to CCR7 promoter and CCR7 expression (Fig. 5, A and B). In addition, the migration ability of MDA-MB-231 cells toward LECs was also fully rescued after co-expression of IKK-␤ and MSK1 (Fig. 5C). These results suggested that Tubo A simultaneously inhibited IKK and p38/MSK1 signaling pathways to suppress CCR7 expression.
Ectopic Expression of TAK1 Also Fully Rescued the Inhibitory Effect of Tubo A-A common upstream kinase of IKK and p38 is TAK1. Therefore, we tested whether overexpression of TAK1 activity might reverse the inhibition of Tubo A. Our data showed that activation of TAK1 completely reversed the inhibition of Tubo A on the binding of NF-B to CCR7 promoter (Fig. 6A). Tubo A-induced inhibition of CCR7 mRNA and protein expression was also reversed (Fig. 6B). In addition, TAK1 activation potently up-regulated NF-B reporter activity and counteracted the inhibitory effect of Tubo A (Fig. 6C). Similarly, TAK1 activation significantly increased cell invasion and reversed Tubo A-induced inhibition (Fig. 6D).
Tubo A Effectively Reduced Lymph Node Metastasis of Breast Cancer in Vivo-We next used a xenograft animal model to address the anti-metastatic activity of Tubo A in vivo. MDA-MB-231 cells were injected into left second thoracic mammary fat pad of nude mice to induce tumor growth and lymph node metastasis. To mimic real drug treatment, intraperitoneal injection (but not direct intratumor injection) was used as the drug administration route. After a 21-day treatment, mice were sacrificed and tumors were removed. Our data indicated that intraperitoneal injection of Tubo A (4 mg/kg) did not inhibit tumor growth at the injection site (Fig. 7A). In addition, no significant alterations of the weights of body, liver, lung, heart, kidney, and pancreas were found in Tubo A-treated group (data not shown). However, an 80% of reduction of CCR7 expression was detected in the tumors of Tubo A-treated group (Fig. 7B). We investigated whether reduction of CCR7 in the tumors was due to attenuation of NF-B activation. Our data indeed demonstrated that nuclear translocation of NF-B was inhibited by 57% by Tubo A in vivo (Fig. 7C). Left and right axillary, brachial, and inguinal lymph nodes of each mouse were excised and were confirmed to be lymphatic tissues by using the specific marker LYVE-1 (data not shown). Lymphatic metastasis of breast cancer was studied by using EpCAM as a marker. Because MDA-MB-231 cells expressed high level of EpCAM (26) and lymph node tissues did not express this epithelial marker (27), we thought it is a suitable marker for detection of lymph node metastasis of MDA-MB-231 cells. Similar application had also been used by others for the detection of lymph node metastasis of cancer cells (26,27,32,33). Lymphatic metastasis of MDA-MB-231 was detected in 100% of left lymph nodes (n ϭ 15, three lymph nodes of left side for each mouse) of the DMSO-treated group (Fig. 7D).
Hematoxylin & eosin stain also confirmed the tumor nest in the lymph node (supplemental Fig. S5). Conversely, only 53% of lymph nodes exhibited metastasis in Tubo A-treated group. Moreover, breast cancer cells (EpCAM-positive staining cells) detected in lymph nodes were much less in Tubo A-treated group (compared the staining signals of the two experimental groups, Fig. 7D). We defined left axillary, brachial, and inguinal lymph nodes as proximal lymph nodes, because they were located at the same side of tumor injection site. Right axillary, brachial, and inguinal lymph nodes (refereed as distal lymph nodes because they were located on the counter side of the tumor injection site) were also studied. Our data showed that 83% of right lymph nodes of the DMSO-treated group had tumor metastasis. On the contrary, only 40% of right lymph nodes of the Tubo A-treated group exhibited tumor metastasis. Collectively, we conclude that Tubo A effectively inhibited lymph node metastasis of breast cancer in vivo.

DISCUSSION
Recent studies indicate that withanolides may repress NF-B-mediated gene transcription via repression of IKK activation (30,34). Consistent with these studies, we found that withanolides (Tubo A in this study) inhibit IKK activation by suppressing the phosphorylation of Ser-177/181 on IKK-␤ or Ser-176/180 on IKK-␣, which are required for their activation. Although reduction of IKK activity and IB␣ degradation are commonly observed in withanolide-treated cells, in vitro kinase assay indicated that IKKs are not direct targets for withanolides (30,34). For the first time, we demonstrate that withanolides may inhibit the p38/MSK1 signaling pathway. A previous study reported that treatment of withaferin A reduced the phosphorylation of NF-B (p65) (30). The authors showed that tumor necrosis factor-induced phosphorylation of NF-B on Ser-536 was attenuated by withanolides. In this study, we demonstrated that the Tubo A inhibited p38 activation in MDA-MB-231 cells. We propose a link between p38 activation and NF-B phosphorylation and find a possible candidate, MSK1, because MSK1 can be activated by p38 and may directly phosphorylate the Ser-276 of NF-B (31,35). Our data indeed demonstrated that inhibition of p38 led to a reduction of phosphorylation of MSK1 on Ser-581, a specific phosphorylation site for p38. In addition, the phosphorylation of Ser-276 of NF-B was also attenuated. The functional importance of the p38/MSK1 signaling in the inhibition of CCR7 by Tubo A was further strengthened by the results that ectopic expression of MSK1 might reverse the inhibitory effect of Tubo A (Fig. 4D).
Because both IKK and p38 kinase activities are attenuated by Tubo A, we hypothesize that Tubo A may target a common upstream kinase of these two kinases. Several kinases, including NF-B-inducing kinase, NF-B-activating kinase, mitogen-activated protein kinase kinase 1 (MEKK1), MEKK3, and TAK1 have been shown to phosphorylate the IKK complex and induce NF-B activation when overexpressed in cells (36 -39). Interestingly, TAK1 can also activate the main upstream activating kinases for p38, MKK3, and MKK6 (40). TAK1, originally identified as a transforming growth factor-␤-activating kinase, plays critical roles in the regulation of diverse cellular processes (41). Activation of TAK1 required its binding proteins (TAB1-3) and is ubiquitin-dependent (42). So, we address the role of TAK1 and show that activation of TAK1 fully reverses the inhibitory effect of Tubo A on CCR7 expression. We conclude that TAK1 is a molecular target for Tubo A (supplemental Fig. S6).
The interaction between chemokines and their cognate receptors is important for tumor metastasis. For example, the chemokine receptor CXCR4 and its ligand stroma-derived factor-1 have been shown to be critical for the dissemination of cancer cells (8,9). CCR7 is important for the adhesion and chemotaxis of leukocyte and dendritic cell to lymph nodes. Because LECs express high level of CCR7 ligands CCL19 and CCL21, up-regulation of CCR7 in cancer cells may promote the migration of cancer cells toward LECs and enhance lymph node invasion. Indeed, many metastatic cancer cells expressed large amount of CCR7 (13,14), and CCR7 expression is associated with lymph node metastasis in many cancers (15)(16)(17)(18)(19). Thus, inhibition of CCR7 expression or block of CCL21-CCR7 interaction may cause reduction of lymph node invasion and tumor metastasis. Indeed, we have shown that knockdown of CCR7 by small interference RNA in metastatic MDA-MB-231 cells reduced their migration toward lymphatic cells (28). However, treatment of human diseases by small interference RNA gene targeting is still at a very premature stage, and local application of small interference RNA for therapy of age-related macular degeneration has just undergone phase I clinical trail (43). Therefore, development of drugs for the inhibition of CCR7 or CCR7-CCL21 interaction is urgently needed. However, very little chemical compounds or natural products have been shown to suppress CCR7 expression. One potential compound is triptolide. This compound is a major active component isolated from the Chinese herb Tripterygium wilfordii Hook F, which has been used in traditional Chinese medicine for the treatment of autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematous (44,45). A very recent study demonstrated that triptolide might repress CCR7 expression and lymphatic migration of dendritic cells (46).
In this study, we provide the first evidence that Tubo A, a bioactive withanolide, inhibits CCR7 expression and lymphatic invasion of breast cancer in vitro and in vivo. Lack of inhibition of primary tumor growth by Tubo A is not unexpected, because the dose used for treatment in this study was very low (4 mg/kg). Previous studies showed that another bioactive withanolide Withaferin A needed almost a 10-fold dose (ϳ30 -50 mg/kg) to inhibit the growth and angiogenesis of melanoma in vivo (7,47). However, Tubo A at this low concentration still exhibited potent inhibitory effect on lymph node metastasis of breast cancer in vivo. These data are encouraging, because lymph node metastasis is a very poor prognostic factor for breast cancer and no effective drugs have been reported to reduce the lymph node metastasis at present. Taken together, we conclude that Tubo A is an effective compound for the inhibition of CCR7 and may be useful for the prevention lymphatic invasion of breast cancer.