A New Herbal Formula, KSG-002, Suppresses Breast Cancer Growth and Metastasis by Targeting NF-κB-Dependent TNFα Production in Macrophages

Tumor-associated macrophages (TAMs) in tumor microenvironment regulate cancer progression and metastases. In breast cancer, macrophage infiltration is correlated with a poor prognosis. While metastatic breast cancer is poor prognostic with a severe mortality, therapeutic options are still limited. In this study, we demonstrate that KSG-002, a new herbal composition of radices Astragalus membranaceus and Angelica gigas, suppresses breast cancer via inhibiting TAM recruitment. KSG-002, an extract of radices Astragalus membranaceus and Angelica gigas at 3 : 1 ratio, respectively, inhibited MDA-MB-231 xenograft tumor growth and pulmonary metastasis in nude mice, while KSG-001, another composition (1 : 1 ratio, w/w), enhanced tumor growth, angiogenesis, and pulmonary metastasis, in vivo. KSG-002 further decreased the infiltrated macrophage numbers in xenograft tumor cohorts. In Raw264.7 cells, KSG-002 but not KSG-001 inhibited cell proliferation and migration and reduced TNF-alpha (TNFα) production by inhibiting NF-κB pathway. Furthermore, a combinatorial treatment of KSG-002 with TNFα inhibited a proliferation and migration of both MDA-MB-231 and Raw264.7 cells. Taken together, we conclude that KSG-002 suppresses breast cancer growth and metastasis through targeting NF-κB-mediated TNFα production in macrophages.

Traditional herbal medicines have shown anti-cancer effects with no or less toxicity compared to other anti-cancer agents [8,[20][21][22]. Although anticancer effects of herbal extracts from radix Angelica gigas and/or radix Astragalus membranaceus have been revealed in different cancer cell types including hepatocellular carcinoma cells, cervical cancer, and prostate cancer cells [23][24][25][26][27][28][29][30], it is yet to be investigated whether those extracts affect highly metastatic cancers, especially breast cancer.
In this study, we developed new herbal formulae, KSG-001 and KSG-002, and identified that KSG-002 but not KSG-001 inhibits breast cancer growth, angiogenesis, and pulmonary metastasis by targeting TAMs. KSG-001 and KSG-002 were extracted from the herbal mixture of Astragalus membranaceus and Angelica gigas with 1 : 1 ratio (w/w) and 3 : 1 ratio (w/w), respectively. KSG-001 and KSG-002 showed no effect on a viability of different breast cancer cells and normal epithelial cells. However, KSG-002 but not KSG-001 effectively repressed TAM cell proliferation and migration with reductions of TNF production by inhibiting NF-B pathway. Thus, our present study suggests that KSG-002 may be useful for treating cancer growth and metastases.

Cell Viability, Migration, Invasion, and
Anchorage-Independent Assays. Cells were cultured in 96-well plates and treated with KSG-001 or KSG-002 for one to three days and subjected to the MTT assays. Assays were performed three times in sextuplicate. For the cell migration, cells cultured in 6-well plates were scratched and treated with extracts for 24 hours. Migrated cells were imaged using a camera connected to a microscope and counted. For the cell invasion, twochamber assays were performed. Cells were cultured on the upper chamber and treated with extracts for 24 hours. Cells were stained with hematoxylin and eosin to capture photo images and counted. For the anchorage-independent assays, cells were mixed with soft agars and cultured for 15 days. Extracts were added every three days. All experiments were performed three times in triplicate.

In Vivo Studies.
Animal experiments were approved by Kyung Hee University Institutional Animal Care of Use Committee (KHU-IACUC). Nude mice were subcutaneously injected with 1 × 10 6 MDA-MB-231 cells. When xenograft tumor volume reached 50 mm 3 , mice were randomly grouped ( = 5/group) and orally added with extracts (500 mg/kg/day). Xenograft tumor volume and mouse body weight were measured three times a week. At day 34 posttreatment, mice were sacrificed, and xenograft tumor bundles and other organs were fixed in 4% formaldehyde for further analyses. Tumor tissues were embedded in paraffin, dissected with 5 m, and stained with hematoxylin and eosin or appropriate antibodies. Anti-CD31 antibody (AbCam) was used for detecting angiogenic vessels. Anti-F4/80 antibody (AbD Serotec) for macrophages was kindly provided by Dr. Hyun-Soo Bae (Kyung Hee University, Seoul, Republic of Korea).

ELISA, Western Blot, and Luciferase
Assays. Cells were treated with KSGs for 24 hours, and then media were gathered for ELISA. Levels of TNF , IL-6, and IL-beta were analyzed with the appropriate ELISA kits (BD Biosciences, San Jose, CA, USA). For the western blot, 30 g of protein was loaded onto 6% to 12% SDS-PAGEs and transferred to PVDF membranes. Anti-p-NF-B, p-I B NF-B, p-IKK, iNOS, MMP-9, FAS, and COX-2 were purchased from Cell Signaling. Anti-Tubulin antibody was obtained from Sigma. Anti-I B antibody was purchased from Upstate Biotechnology. Anti-IKK antibody was obtained from Santa Cruz Biotechnology. For the luciferase assays, constructs for NF-B-luc (Stratagene) and TNF -luc were kindly provided from Dr. Mingyao Liu (Texas A&M University, Houston, TX, USA), and plasmids for IKK and p65 were received from Dr. Barat B. Aggarwal (University of Texas MD Anderson Cancer Center, Houston, TX, USA). Raw264.7 cells were transfected with the appropriate plasmids for 24 hours and then treated with extracts or not for another 6 hours. The luciferase activities were done by manufacturer's manuals (Promega).

2.5.
Statistics. Statistical significances were calculated by the Student's -test or one-way ANOVA, as appropriate. value less than 0.05 was considered statistically significant.
When tumor angiogenic vessels were examined with anti-CD31 antibody staining, KSG-002, compare to control or KSG-001, reduced angiogenic vessel numbers (Figure 1(c)).  As KSG-002 appeared to inhibit tumor progression and angiogenesis, we further examined whether KSG-002 inhibited tumor metastasis. Mice were sacrificed at day 34 posttreatment, and lungs from each group were dissected for investigating metastatic colony formation at the lungs. When metastatic foci at lung were counted, KSG-002 (avg. = 5.2), compared to control (avg. = 8.25), appeared to reduce metastatic colony numbers while it was not statistically significant, and KSG-001 (avg. = 10.2) increased those numbers (Figure 1(d)).  (Figure 2(a)).

KSG-002 Does Not Affect Tumor
We next examined breast cancer cell migration, invasion, and anchorage-independent growth. Both KSG-001 Evidence-Based Complementary and Alternative Medicine 5 and KSG-002 at 50 g/mL did not affect MDA-MB-231 cell migration, invasion, and anchorage-independent growth (Figures 2(b)-2(d)). Thus, KSG-001 or KSG-002 seems not to affect breast cancer cells directly.
Next, we examined whether KSG-002 affects chemotactic migration of macrophages. Raw264.7 cells on the upper chamber were treated with either KSG-001 or KSG-002, and MDA-MB-231 medium was added on the bottom chamber as a chemoattractant. KSG-002 at 500 g/mL inhibited Raw264.7 cell migration while KSG-001 at 500 g/mL did not affect it (Figure 3(d)), indicating that KSG-002 blocks macrophage infiltration in tumor bundle.

KSG-002 Reduces TNF Production of TAMs by Inhibiting
NF-B Phosphorylation. TNF released from macrophages activates NF-B-mediated signaling pathway in cancer, resulting in cancer progression and metastasis [16]. To examine whether KSGs affect macrophagic TNF , Raw264.7 cells were cultured with conditional medium from MDA-MB-231 cells, and media from Raw264.7 were subjected to ELISA assays. KSG-002, but not KSG-001, reduced TNF level produced by Raw264.7 cells by approximately 86.5% (Figure 4(a)).
As TNF expression is regulated by NF-B signaling pathway, we further examined NF-B pathway molecules. Raw264.7 cells were cultured in conditioned medium from MDA-MB-231 cells and treated with 500 g/mL of either KSG-001 or KSG-002 for 24 hours. Both KSG-001 and KSG-002 reduced phosphorylation of IKK, I B, and NF-B (Figure 4(b)). Furthermore, both extracts inhibited expression levels of NF-B-dependent genes such as iNOS, COX2, Fas, and MMP-9 (Figure 4(c)). Thus, we further examined whether KSGs affect NF-B transcriptional activity. In the luciferase assays, while KSG-002 inhibited NF-B transcriptional activity in cells overexpressed with p65 and NF-B-luc, KSG-001 did not affect it (Figure 4(d), top). To confirm this inhibitory effect of KSG-002 on NF-B transcriptional activity, we performed the luciferase assays with TNF promoter construct (TNF -luc). In Raw264.7 cells transfected with TNF -luc and p65, KSG-002 but not KSG-001 decreased NF-B-dependent TNF promoter activity (Figure 4(d), bottom). Thus, our data indicate that KSG-002 selectively inhibits NF-B-dependent TNF production.
Hence, we further examined whether KSG-002 affects macrophagic TNF -induced breast cancer cell migration. MDA-MB-231 cells were treated with TNF alone or cotreated with TNF and KSGs for 24 hours. KSG-002 at 500 g/mL inhibited TNF -induced MDA-MB-231 cell migration by approximately 43%, while KSG-001 did not affect it (Figure 4(e)). We further examined whether KSG-002 affects the viability of either tumor cells or macrophages in TNF -confluent tumor microenvironment. Whereas TNF alone did not affect cell viability of either Raw264.7 or MDA-MB-231 cells (Figure 4(f)), KSG-002 plus TNF reduced the viabilities of both cell types by approximately 60% (Raw264.7) and 50% (MDA-MB-231), respectively (Figure 4(g)). Those data indicate that KSG-002 inhibits both cell proliferation and migration by enhancing cellular sensitivity to TNF .
As mentioned, infiltrating TAM numbers is associated with a poor prognosis [31][32][33]. TAMs release different cytokines and regulate tumor cell growth, migration, and invasion [4,6]. TNF produced by TAMs is known to promote cancer growth [12,15]. A number of F4/80-positive TAMs were reduced in tumor burdens of KSG-002-treated groups in our  Migrating cell numbers were counted, and data were presented as means ± standard deviations. Experiments were performed three times. * < 0.05, TNF alone versus TNF plus KSG-002. (f and g) Cells indicated were treated with TNF alone (f) or together with KSGs (g) for 48 hours and subjected to MTT assays. Right triangles indicate different concentrations of TNF (10 −3 , 10 −2 , 10 −1 , 1, 10, and 10 2 ng/mL). For the combination, 10 2 ng/mL of TNF was treated with different concentrations of KSGs as indicated. Experiments were performed three times in triplicate. Representative data show mean ± standard deviation. * < 0.05. in vivo study. Furthermore, KSG-002 reduced TNF production from TAMs and repressed cell growth and chemotactic migration of both TAMs and MDA-MB-231 cells. Consistently, we found a reduction of TNF level in the blood of KSG-002-treated mice, while levels of IL-1 and IL-6 were not altered (data not shown). TNF production involves NF-B transcriptional activation via phosphorylation of I B and NF-B [16]. Thus, an inhibition of I B/NF-B phosphorylation by KSG-002 may result in a reduction of TNF production from TAMs, suppressing breast cancer growth and metastasis ( Figure 5). We found that KSG-002 selectively repressed NF-B-dependent regulation of TNF expression, while both KSG-001 and KSG-002 affected expression levels of NF-B-dependent genes such as iNOS, COX-2, MMP-9, and FAS. In addition, we found that KSG-002 inhibited JAK1/2 phosphorylation but not phosphorylation of STAT3 (data not shown). Thus, it remains to examine crosstalks between JAK/STAT pathway and NF-B pathway in KSG-002 effect on  tumor microenvironment. In addition, while an extract from Angelica gigas showed cytotoxicity in rat intestinal epithelial (RIE) cells, KSG-002, a mixture of extracts from Astragalus membranaceus and Angelica gigas with 3 : 1 ratio, did not affect RIE cell viability (data not shown). In addition, KSG-001 rather increased tumor size. Thus, KSG-002 seems to have both the efficacy and safety, when two components are extracted with the appropriate ratio (3 : 1 in our study). As KSG-002 is herbal extracts containing multiple components, it is possible that KSG-002 may target other stromal cells such as vessel cells, fibroblasts, and immune cells. In our histochemistry data, CD-31-positive vessel numbers were reduced by KSG-002. While TAMs regulate tumor angiogenesis [40,41], it is also plausible that KSG-002 may directly target endothelial cells. Thus, it still remains to decipher multitarget effects of KSG-002 in tumor microenvironment. In addition, it is also worth to investigate effects of KSG-002 on different cancer types as well as different breast cancer subtypes. In this study, we focused on roles of KSG-002 in breast cancer cells, especially highly metastatic breast cancer cells, and found that the combination of KSG-002 with TNF reduced MDA-MB-231 cell viability, while KSG-002 or TNF alone did not affect it. Thus, further researches for KSG-002 effects on hormone-positive or HER2-positive breast cancer progression will provide evidence on anticancer effects of KSG-002 in different types of breast cancer cells, in depth and breadth. Furthermore, KSG-002 can be applied in different cancer cell types that are highly metastatic, as TNF may be plentiful in metastatic tumor cohorts. When Raw264.7 cells were treated with KSG-002, cell viability was reduced. Likewise, KSG-002 decreased viabilities of Raw264.7 and MDA-MB-231 cells when cotreated with TNF . However, KSG-002 did not affect the viability of Raw264.7 cells cultured in tumor medium. Thus, certain factors released from tumor cells may negatively affect KSG-002 effect, although it remains yet to know factors or mechanisms by which KSG-002 effect is abolished when macrophages and tumor cells are cocultured.
In this study, we demonstrate that KSG-002, a novel herbal formula, suppresses highly metastatic breast cancer cells by inhibiting NF-B-mediated TNF production from TAMs. Therapeutic agents for highly metastatic TNBC are yet to be developed, and regimens used for other cancer types are not working well. Traditional herbal medicines have been used for anticancer therapy, since those compared to anticancer drugs used in present have no or less side effect. While we need more efforts to determine KSG-002 functions in cancer disease, our present study for KSG-002 effect on breast cancer provides knowledge for therapeutic options.

Conclusion
In this study, we demonstrate that a new herbal formula, KSG-002, suppresses breast cancer growth and metastasis by blocking NF-B-dependent TNF production from macrophages. Although we still make efforts to clearly understand its inhibitory roles in different cancer types and biological mechanisms in cellular and physiological levels, our present data suggest that KSG-002 may be useful for treating highly metastatic breast cancer.