Lupeol Is One of Active Components in the Extract of Chrysanthemum indicum Linne That Inhibits LMP1-Induced NF-κB Activation

We have previously reported that seventy percent ethanol extract of Chrysanthemum indicum Linne (CIE) strongly reduces Epstein-Barr virus (EBV)-transformed lymphoblastoid cell line (LCL) survival by inhibiting virus-encoded latent infection membrane protein 1 (LMP1)-induced NF-κB activation. To identify an active compound(s) in CIE that inhibits LMP1-induced NF-κB activation, activity-guided fractionation was employed. The CH2Cl2 fraction of CIE strongly reduced LMP1-induced NF-κB activation and LCL viability with relatively low cytotoxic effects on primary human foreskin fibroblast (HFF), HeLa or Burkitt’s lymphoma (BL41) cells. Furthermore, lupeol, a pentacyclic triterpene, was identified in the CH2Cl2 fraction of CIE to attenuate LMP1-induced NF-κB activation and LCL viability. This study demonstrates that lupeol is one of active compounds in the CH2Cl2 fraction of CIE that inhibits LMP1-induced NF-κB activation and reduces NF-κB-dependent LCL viability.


Introduction
The NF-κB family of transcription factors plays an important role in tumoirgenesis, and aberrant NF-κB activation is a hallmark of many epithelial and lymphoid-derived cancers [1,2]. NF-κB promotes tumorigenesis by inducing expression of genes involved in cell proliferation, survival, tumor promotion, immortalization, angiogenesis and metastasis [1][2][3]. In addition, NF-κB is a critical regulator of inflammation and promotes inflammation-associated cancers [4].
A canonical NF-κB pathway involves the heterodimeric p65/p50 complexes that are sequestered in the cytoplasm by IκB proteins. In response to various stimuli including proinflammatory cytokines, TNF-α and IL-1, and lipopolysaccharide (LPS), IκB proteins are phosphorylated by IKKβ and degraded by the ubiquitin-proteasome pathway allowing nuclear translocation of the p65/p50 complexes. The non-canonical NF-κB pathway which is utilized by lymphotoxin β (LT-β), B cell-activating factor of the TNF family (BAFF) and CD40 involves NF-κB inducing kinase (NIK)-and IKKαmediated proteolytic processing of p100 into p52 and nuclear translocation of the RelB/p52 complexes [5,6].
We have previously reported that Chrysanthemum indicum Linne extract (CIE) inhibits LMP1-induced NF-κB activation and LCL viability without exhibiting any adverse effect on the viability of cells whose survival is independent of NF-κB activation [19]. Therefore, in this study, we have expanded our investigation to identify an active compound(s) in CIE that inhibits LMP1-induced NF-κB activation and LCL viability by using activity-guided fractionation.

The effect of the CH 2 Cl 2 fraction of CIE on IKK activation
Since LMP1 activates both non-canonical and canonical NF-κB pathways through CTAR1 and CTAR2, respectively, the effect of the CH 2 Cl 2 fraction of CIE on LMP1-induced IKKα or IKKβ activation was further investigated ( Figure 3A). BL41 cells or their LMP1 expressing counterparts were treated with either DMSO or the CH 2 Cl 2 fraction for 24hr, and IKKα or IKKβ activity was determined by Western blot analysis with anti-p100/p52 or anti-phospho-IκBα antibody, respectively. Since the promoters of IκBα and p100 contain κB-binding sites and are regulated by the canonical NF-κB pathway [20][21][22], the protein levels of p100 and IκBα were induced in BL41 cells expressing LMP1 as previously reported ( Figure 3A, compare lane 2 with lane 1) [6,9,23]. In DMSO treated cells, LMP1 induced p100 processing to p52 and IκBα phosphorylation ( Figure 3A, compare lane 2 with lane 1). On the other hand, the CH 2 Cl 2 fraction blocked LMP1-induced p100 processing and IκBα phosphorylation ( Figure 3A, compare lane 4 with lane 2). The protein levels of p100 and IκBα in cells treated with the CH 2 Cl 2 fraction were decreased due to inactivation of the canonical NF-κB pathway ( Figure 3A, compare lanes 3 and 4 with lanes 1 and 2). Furthermore, the CH 2 Cl 2 fraction significantly attenuated both LPS-and IL-1β-induced IKKβ activation ( Figure 3B and C, compare lanes 6 to 10 with lanes 1 to 5). Thus, these data suggest that the CH 2 Cl 2 fraction of CIE inhibits NF-κB activation possibly by interfering with IKK activation.

The effect of the CH 2 Cl 2 fraction of CIE on LCL viability
Since LMP1-induced NF-κB activation is essential for LCL survival, the effect of the CH 2 Cl 2 fraction of CIE on the viability of LCLs was investigated. LCLs were treated with 100µg/ml of CIE fractions, and the cell viability was measured using the CellTiter-Glo assay at 0, 3, 6, 9, 12 or 24hr after treatment ( Figure 4). Consistent with the NF-κB-dependent luciferase reporter data, the CH 2 Cl 2 fraction strongly reduced LCL viability ( Figure 4). After treatment with the CH 2 Cl 2 fraction, LCL viability was reduced by 65% at 9hr and by 94% at 24hr ( Figure 4). Interestingly, other fractions had almost no effect on LCL viability ( Figure 4).
To further assess the relative toxicity of the CH 2 Cl 2 fraction in other cell types whose survival is independent of NF-κB activation, LCLs, HFF, HeLa or BL41 cells were treated with 6.25, 12.5, 25, 50 or 100µg/ml of the CH 2 Cl 2 fraction, and the cell viability was measured by using the CellTiter-Glo assay at 24, 48 or 72hr after treatment ( Figure 5). The CH 2 Cl 2 fraction reduced LCL viability in a dose-and time-dependent manner, and the half maximal inhibitory concentration (IC 50 ) values for the cytotoxicity of the CH 2 Cl 2 fraction on LCLs at 24, 48 and 72hr were 97.3, 55.8 and 45.2mM, respectively ( Figure 5A and Table 1). Interestingly, the CH 2 Cl 2 fraction had very little cytotoxic effect on HFF or HeLa cells. Within the first 24hr after treatment, the CH 2 Cl 2 fraction had no adverse effect on the viability of HFF or HeLa cells ( Figure 5B and C). After 72hr treatment with the CH 2 Cl 2 fraction at 100µg/ml, HFF or HeLa cell viability was reduced by 91% and 80%, respectively ( Figure 5B and C). BL41 cells were slightly more sensitive to the CH 2 Cl 2 fraction than HFF or HeLa cells. After 24hr treatment with the CH 2 Cl 2 fraction at 100µg/ml, BL41 cell viability was reduced by 72% and 98% ( Figure 5D). The CH 2 Cl 2 fraction of CIE inhibits IKK activation. (A) Parental BL41 cells and their counterparts expressing LMP1 were treated with either DMSO or CH 2 Cl 2 fraction of CIE at 100µg/ml for 24hr, and equal amounts of cell extracts were subjected to Western blot analysis with anti-p100/p52, anti-phopho-IκBα, anti-IκBα, anti-LMP1 or antitubulin antibody. (B) Raw 264.7 cells were pre-treated with either DMSO or CH 2 Cl 2 fraction of CIE at 100µg/ml for 3hr and stimulated with LPS at 1µg/ml. At 0, 15, 30, 45 and 60 min after LPS treatment, equal amounts of cell extracts were subjected to Western blot analysis with anti-IκBα or anti-tubulin antibody. (C) HeLa cells were pre-treated with either DMSO or CH 2 Cl 2 fraction of CIE at 100µg/ml for 3hr and stimulated with IL-1β at 20ng/ml. At 0, 15, 30, 45 and 60 min after IL-1β treatment, equal amounts of cell extracts were subjected to Western blot analysis with anti-IκBα or anti-tubulin antibody. Nonetheless, the CH 2 Cl 2 fraction was evidently less cytotoxic to HFF, HeLa or BL41 cells with IC 50 values of 145.5, 109.7 and 91.4mM, respectively, at 72hr after treatment (Table 1).
Since NF-κB inhibition induces apoptosis in LCLs [9,10], the apoptotic effect of the CH 2 Cl 2 fraction of CIE in LCL was further assessed. LCLs were treated with 100µg/ml of either DMSO or the CH 2 Cl 2 fraction, and poly (ADP-ribose) polymerase (PARP) cleavage was determined by Western blot analysis at 0, 1, 3, 6, 9, 12 or 24hr after treatment ( Figure 6). At 6hr after treatment, the CH 2 Cl 2 fraction induced the PARP cleavage ( Figure 6, lane 11). At later time points, the PARP cleavage was further induced in cells treated with the CH 2 Cl 2 fraction ( Figure 6, compare lanes 11 to 14 with lanes 4 to 7). Taken together, the CH 2 Cl 2 fraction of CIE is more cytotoxic toward NF-κBdependent LCLs than NF-κB-independent HFF, HeLa or BL41 cells. The CH 2 Cl 2 fraction of CIE may reduce LCL viability by inducing apoptosis.

The effect of lupeol isolated from CIE on LCL viability
To assess the relative toxicity of lupeol in different cell types, LCLs, HFF, HeLa or BL41 cells were treated with either DMSO or lupeol at 3.125, 6.25, 12.5, 25 or 50µg/ml, and the cell viability was measured by using the CellTiter-Glo assay at 0, 3, 6, 9, 12, 24, 48 or 72hr after treatment ( Figure 9). Although lupeol reduced the viability of these cells in a dose-and timedependent manner at latter time points, it was more cytotoxic toward LCLs than HFF, HeLa or BL41 cells (Figure 9). Within the first 24hr after treatment at 50µg/ml, lupeol reduced LCL viability by 54% ( Figure 9A). The IC 50 values for the cytotoxicity of lupeol on LCLs at 24, 48 and 72hr were 109.9, 57.6 and 51.8mM, respectively (Table 2). On the other hand, lupeol had very little adverse effect on the viability of HFF or HeLa cells within the first 24hr after treatment ( Figure 9B and C). At 50µg/ml, lupeol reduced the viability of HFF cells by 48% and 93% at 48 and 72hr after treatment, respectively ( Figure 9B). At the same concentration, lupeol reduced the viability of HeLa cells by 78% and 96% at 48 and 72hr after treatment ( Figure  9C). BL41 cells were more sensitive to lupeol than HFF and HeLa cells. Within the first 24hr after treatment at 50µg/ml, lupeol reduced the viability of BL41 cells by 48% ( Figure 9D). At 48 and 72hr after treatment, 50µg/ml of lupeol further reduced the viability of BL41 cells by 61% and 92%, respectively ( Figure 9D). Nonetheless, lupeol was slightly more effective to reduce the viability of LCLs than BL41 cells. The IC 50 value for the cytotoxicity of lupeol on HFF, HeLa and BL41  (Table 2).
Furthermore, the apoptotic effect of lupeol in LCL was determined. LCLs were treated with 50µg/ml of lupeol, and PARP cleavage was determined by Western blot analysis at 0, 1, 3, 6, 9, 12 or 24hr after treatment (Figure 10). At 9hr after treatment, lupeol induced the PARP cleavage (Figure 9, lane 12). These data indicate that lupeol induces apoptosis in LCLs and is more cytotoxic to LCLs than HFF, HeLa or BL41 cells.
We have previously reported that CIE strongly reduces EBVtransformed LCL viability by inhibiting LMP1-induced NF-κB activation [19]. CIE had almost no adverse effects on the viability of cells in which NF-κB is not activated [19]. C. indicum has been used to treat inflammatory disease in traditional Korean and Chinese medicine [33][34][35][36][37][38][39][40]. CIE inhibits LPSinduced production of inflammatory cytokines possibly by down-regulating NF-κB and MAPK in RAW264.7 cells macrophages [41]. In addition, CIE inhibits LMP1-and IL-1βinduced NF-κB activation by blocking IKK activity [19]. How CIE inhibits IKK is unclear and is the subject of future studies. CIE may inhibit IKK activation directly by targeting IKK and/or indirectly by blocking the function of components upstream of IKK.
The CH 2 Cl 2 fraction of CIE was more effective to inhibit LMP1-induced NF-κB activation and selectively reduce LCL viability than lupeol. Lupeol may interact synergistically with unknown compounds in the CH 2 Cl 2 fraction of CIE to reduce LMP1-induced NF-κB activation and LCL viability. In addition to the CH 2 Cl 2 fraction of CIE, the EtOAc or n-BuOH fraction also reduced LMP1-induced NF-κB activation by 35% or 31%, respectively. Thus, these fractions may contain additional active compounds that contribute NF-κB inhibitory activity of CIE.

Fractionation and isolation of active compounds in CIE
The dried Chrysanthemum indicum Linne (1.2kg) was exhaustively extracted by 70% EtOH. The solvent was then evaporated under reduced pressure, at a temperature not exceeding 40°C, to yield 513 g of a semisolid dark yellow residue. The extract was re-suspended in distilled water and successively fractionated with n-Hexane, dichloromethane The CH 2 Cl 2 layer (3g) was chromatographed on a column of Silica (2.5 x 20 cm). Elution was carried out with ddH 2 O followed by 10% stepwise addition of methanol till 100% to give 9 sub-fractions. The 3rd sub-fraction was re-chromatographed on a column of sephadex LH-20. Elution was carried out with CH 2 Cl 2 :ddH 2 O:MeOH = 1:1:3, 1:1:2 and 1:1:1 and then an active compound was isolated by recycling HPLC with solvent, CH 3 Cl-MeOH ( Figure 6).

Western blot analysis
Cells were collected, fractionated, and transferred to nitrocellulose membranes as described previously [57]. Polyclonal rabbit antibody to p100/p52 was a kind gift from Dr. Ulrich Siebenlist (NIH). Antibodies to phospho-IκBα, IκBα and PARP were purchased from Cell Signaling Technology (Beverly, MA). An antibody to alpha-tubulin was purchased from Sigma Aldrich (St. Louis, MO). Enhanced chemiluminescence detection reagents (Pierce, Rockford, IL) and secondary peroxidase-labeled anti-mouse or anti-rabbit immunoglobulin G antibody (Amersham Biosciences, Piscataway, NJ.) were used according to the manufacturer's directions.

NF-κB luciferase reporter and cell viability assays
NF-κB luciferase reporter assay was performed as previously described [58]. Cell viability was determined using CellTiter-Glo luminescent cell viability assay (Promega, WI) according to the manufacturer's directions.