The well-dried DPS herbs were ground using a pestle and mortar until it was in a fine powder condition. After adding 100 ml 70% ethanol to 2 g fine powder, the mixture was shaken for 72 h on a rotary shaker to extract the majority of the compounds. The resulting extract was filtrated twice through a filter paper. The filtrated extract was concentrated and dried on an IKA RV10 BasicV rotary evaporator (IKA^® Korea. Ltd.), and the dried extract was dissolved in an ultrasonic bath with the gradual addition of a total of 7 ml pure methanol. The resulting extract was stored in a freezer at −30°C.

Compound fractions were separated by TLC with the solvent system ethyl acetate, formic acid, glacial acetic acid and distilled water in a ratio of 100:11:11:26 (v/v), and with rutin (0.05% in methanol) as a standard. DPSE (600 µl) was added to the TLC Sorbfil plates (20 µl per 1 cm extract and 3 µl per point standards; Sorbfil). Following separation, the TLC plates were examined and under ultraviolet (UV) light using the TLC-scanner. The developed spots were described by color and the R_f values are calculated based of the rutin standard with a R_f of 0.30. Under UV light, six fluorescent dark violet spots were detected with R_f values of 0.37, 0.50, 0.65, 0.75 and 0.90, which were designated as 1F, 2F, 3F, 4F, 5F and 6F, respectively. The developed spots were cut from the TLC plate in strips, crushed and placed into 15 ml Falcon tubes for elution in small volumes of pure 100% methanol in a Multi RS-60 Biosan multitator (Biosan) for 20 min. The eluates were transferred to 10 ml tubes and centrifuged at 3,500 × g for 15 min at room temperature and freeze-dried. All chemicals used were chromatography grade purity.

To analyze the components of the TLC fraction 6F of DPSE, the HPLC system (Agilent Technologies 1290 Series; Agilent Technologies, Inc.) was used. HPLC analysis was performed using a Poroshell column (100×3.0 mm, 2.7 µm). Mobile phase A was 0.1% formic acid in distilled water and mobile phase B was 0.1% formic acid in acetonitrile. The injection volume was 3 µl, the column temperature was 40°C, and the samples were eluted at 300 µl/min with gradient program (0–5 min, 2% B; 5–16 min, 40% B; 16–24 min, 95% B; 24–25 min, 2% B). The eluents separated from 6F were analyzed with Agilent Technologies 6530 Accurate-Mass (Q-TOF) (Agilent Technologies, Inc.). The optimal parameters were as follows: The positive ESI mode, gas temperature 300°C, sheath gas temperature 350°C, sheath gas flow rate 11 l/min, and fragmentor voltage 175 V.

Human DLBCL cell lines (DHL4, DHL6, Ly1, Ly8, Ly19 and HBL1) (13) and peripheral blood mononuclear cells (PBMCs) were maintained in RPMI-1640 medium (Hyclone; Cytiva) supplemented with 10% fetal bovine serum (FBS; Hyclone; Cytiva), 1% L-glutamine, 1% N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid buffer, and 1% penicillin/streptomycin at 37°C in a 5% CO₂ incubator. The DHL4, DHL6, Ly1, Ly8, Ly19 and HBL1 DLBCL cell lines were a generous gift from Dr Ricardo Aguiar (University of Texas Health Science Center at San Antonio, San Antonio, TX, USA). PBMCs were purchased from PDXen Biosystems Co. Normal cells (splenocytes and bone marrow cells) from wild-type C57BL/6 mice were cultured in DMEM (Hyclone; Cytiva) with 10% FBS and 1% penicillin/streptomycin in a 5% CO₂ incubator at 37°C.

The primary antibodies against poly (ADP-ribose) polymerase 1 (PARP-1; cat. no. sc-7150, 1:1,000), pro-caspase 3 (cat. no. sc-7148, 1:1,000), Bcl-2 (cat. no. sc-7382, 1:2,000), Mcl-1 (cat. no. sc-819, 1:1,000), Bax (cat. no. sc-7480, 1:1,000), Bak (cat. no. sc-832, 1:1,000), Myc (cat. no. sc-40, 1:1,000), p53 (cat. no. sc-126, 1:1,000) and β-actin (cat. no. sc-47778, 1:1,000) were obtained from Santa Cruz Biotechnology, Inc. The antibody against Bcl-xL (cat. no. 14-6994-81, 1:1,000) was purchased from eBioscience. Secondary HRP-conjugated goat anti-rabbit (cat. no. A120-101p, 1:5,000) and secondary HRP-conjugated goat anti-mouse (cat. no. A90-116p, 1:5,000) IgG antibodies were obtained from Bethyl Laboratories.

The CellTiter 96 AQueous MTS assay (Promega Corporation) was used to determine the cytotoxicity of DPSE or isolated compounds (1~6F and 1~6T), according to the manufacturer's instructions. Briefly, DLBCL and normal cells (splenocytes and bone marrow cells from mice and PBMCs) were plated at a density of 3×10⁴ cells and 5×10⁵ cells/well, respectively, in 96-well cell culture microplates and treated with DPSE (0, 0.2, 0.4, 0.6, 0.8 and 1 mg/ml for 24 or 48 h) or TLC fractions of terpenoids or flavonoids (0, 100 and 200 µg/ml for 24 h) at 37°C, followed by the addition of 30 µl MTS reagent to the wells and incubation at 37°C for 2 h. Absorbance was measured at 450 nm on a GloMax™ Microplate multi-mode reader (Promega Corporation).

DHL4, Ly1, Ly8, HBL1 and normal murine cells were seeded at a density of 5×10⁵ cells/well in 12-well plates, treated with DPSE (0 and 1 mg/ml) for 24 h at 37°C, and stained using FITC Annexin V Apoptosis Detection kit I (BD Biosciences) according to the manufacturer's instructions. The apoptotic fraction was analyzed by a BD FACSVerse flow cytometer (BD Biosciences). Data acquisition and analysis were performed using BD FACSCanto II software (version 3.0; BD Biosciences).

To examine caspase 3/7 activity, cells were exposed to DPSE (0, 0.5 and 1 mg/ml for 24 h) at 37°C, and Caspase-Glo 3/7 Assay reagent (cat. no. G8090; Promega Corporation) was added and incubated for 1 h at room temperature. Luminescence was measured using GloMax™ Microplate multi-mode reader.

For the measurement of the mitochondrial membrane potential, cells were treated with DPSE (0 or 1 mg/ml for 24 h) at 37°C, followed by the incubation with JC-1 dye (Abnova) for 15 min at 37°C. The resulting fluorescence was then monitored by fluorescence microscopy or by using a BD FACSVerse flow cytometer (BD Biosciences). Data acquisition and analysis were performed using DP2-BSW software (version 2.1; Olympus Cooperation) and BD FACSCanto II software (version 3.0; BD Biosciences).

Following exposure to DPSE (0, 0.5 or 1 mg/ml for 24 h at 37°C), DHL4, Ly1, Ly8, and HBL1 cells were harvested and lysed in RIPA buffer (ELPIS Biotech. Inc.) with 1 mM Na-vanadate, 50 mM β-glycerophosphate disodium salt, β-mercaptoethanol (142 mM; Bioworld Technology, Inc.), ProteaseArrest™ (G-Biosciences) and EDTA (5 mM; G-Biosciences). Protein concentrations were measured using the BCA assay kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Samples were boiled at 100°C for 10 min in sample buffer, 25 µg proteins were loaded in the 10 or 12% polyacrylamide gels, and transferred onto Immobilon-P Transfer membranes, followed by blocking with 1% bovine serum albumin (BSA; MP Biomedicals) for 1 h at room temperature and probing with primary antibodies at 4°C overnight. After washing three times with TBS containing 0.1% Tween-20 (TBST) for 5 min each, the membranes were probed with secondary anti-mouse/rabbit antibodies for 1 h at room temperature. After washing three times with TBST for 10 min each, the membranes were exposed to a chemiluminescent substrate [EzWestLumi plus (ATTO Corporation)] and the protein bands were visualized using the Luminograph II (ATTO Corporation). Western blots were quantified using ImageJ 1.52a software (National Institutes of Health).

For in vitro toxicity testing, the splenocytes and bone marrow cells were isolated from wild-type male C57BL/6 mice (n=5; age, 8 weeks; weight, 18±2 g), following the lysis of red blood cells using ammonium chloride-based red cell lysis buffer. For in vivo toxicity testing, 10 athymic male nude mice (age, 8 weeks; weight, 20±4 g) were divided into two groups and injected intraperitoneally with vehicle (1% methanol; n=5) or DPSE (50 mg/kg; n=5) daily for 3 weeks, followed by histological analysis of the lung, heart, liver and kidney. C57BL/6 and athymic nude mice (BALB/c-nu) were purchased from Central Lab Animal Inc. and Japan SLC, Inc., respectively. For xenograft studies, 6 athymic nude mice (age, 4 weeks; weight, 17±1.6 g) were divided into two cohorts and inoculated in the flanks with 1×10⁷ Ly1 or HBL1 parental cells. The injection sites were inspected for possible tumor formation for 3 weeks. Mice were housed under a 12:12 h light/dark cycle at 26°C and provided food and water ad libitum. All mice used in this research were euthanized using a CO₂ chamber 3 weeks after the start of the indicated treatment. A 10-liter volume chamber was used and the displacement rate was fixed to 20% of the CO₂ chamber volume. Cervical dislocation was performed for secondary means to assure death after euthanasia. The animal protocol used in this study was reviewed and approved by the Pusan National University-Institutional Animal Care and Use Committee (approval no. PNU-2016-1158).

H&E staining was performed as previously described (14,15). After sacrificing the mice, the lung, heart, liver and one kidney per mouse were isolated, and the specimens were fixed with 10% neutral buffered formalin solution (Sigma-Aldrich; Merck KGaA) at 4°C overnight. Then, samples were embedded in paraffin and cut into 4-µm sections, followed by deparaffinization, hydration and staining with H&E (stained with hematoxylin for 15 and eosin for 5 min) at room temperature. H&E-stained sections were visualized using a light microscope (Olympus CX31 microscope, Olympus Corporation) at ×100 magnification. The representative images were captured using Motic Images Plus 2.0 software (Motic Co. Ltd.).

Generation of Ly1 control cells or Ly1 cells stably expressing c-Myc (Myc hereafter) using pCDH-CMV-MCS-EF1-copGFP lentiviral constructs (Systems Biosciences) was carried out as previously described (15), and then designated as Ly1 CDH and Ly1 CDH-Myc, respectively.

To measure the transcriptional level of Myc, total RNA was extracted from Ly1 CDH-control or CDH-Myc cells using TRIzol (Favorgen), and cDNA was synthesized using PrimeScript RT reagent kit with gDNA Eraser (Takara Bio, Inc.) according to the manufacturer's guidelines. qPCR was then performed using TOPreal qPCR PreMIX SYBR Green with low ROX (Enzynomics). The thermocycling conditions were as follows: 95°C for 10 min followed by 40 cycles of 95°C for 10 sec, 60°C for 15 sec and 72°C for 30 sed. The sequences for the TBP- and Myc-specific qPCR primers are as follows: Myc forward, 5′-CTCCTGGCAAAAGGTCAGAG-3′ and reverse, 5′-TCGGTTGTTGCTGATCTGTC-3′; and TBP forward, 5′-TATAATCCCAAGCGGTTTGCTGCG-3′ and reverse, 5′-AATTGTTGGTGGGTGAGCACAAGG-3′. Relative gene expression was analyzed using the 2^−∆∆Cq method (16).

To investigate Myc protein stability, Ly1 DLBCL cells were treated with cycloheximide (20 µg/ml; Sigma-Aldrich; Merck KGaA) in the presence or absence of DPSE (1 mg/ml for 0, 2, 4 or 8 h) (17). Western blotting was performed to analyze Myc protein levels.

Data are presented as the mean ± standard deviation. Statistically significant differences were identified by performing a Mann-Whitney U test or one-way ANOVA test with Tukey's post hoc test using Microsoft Office Excel 2010 (Microsoft Corporation) and GraphPad Prism version5 software (GraphPad Software, Inc.). All experiments were repeated at least three times independently to confirm reproducibility. P<0.05 was considered to indicate a statistically significant difference.

To investigate the effect of DPSE on the survival of DLBCL cells, one ABC (HBL1) and five GCB DLBCL (DHL4, DHL6, Ly1, Ly8 and Ly19) cell lines were exposed to increasing concentrations of DPSE for 24 and 48 h, followed by MTS assay. Cell viability was decreased dose-dependently as compared with the control (Fig. 1A). The IC₅₀ at 24 h ranged between 235.0 and 624.6 µg/ml, and HBL1 and DHL4 cells were the most and least sensitive to DPSE, respectively (Fig. 1B). Annexin V/PI double staining demonstrated that the apoptotic rate was significantly increased upon exposure to DPSE (Figs. 1C and S1), which was associated with an increase in cleaved PARP-1, the active form of caspase 3, and caspase 3/7 activities (Fig. 1D and E).

To corroborate the in vitro results, in vivo studies were performed using murine xenograft models. Two cell lines, Ly1 and HBL1, were selected due to their robust proliferation in culture, and 1×10⁷ Ly1 or HBL1 cells were injected subcutaneously into the flanks of athymic nude mice. However, tumors did not develop within 3 weeks (data not shown). Subsequently, the selectivity of DPSE toward cancer cells was investigated. Three different types of normal cells from humans and mice, human PBMCs, normal mouse splenocytes and bone marrow cells, were treated with DPSE, followed by the analysis of viability and apoptotic rates (Figs. 2A-C and S2). In contrast to DLBCL cells, the survival and apoptotic rates of normal cells were not affected, suggesting that the cytotoxic effect of DPSE is specific to cancer cells in vitro.

To further investigate the potential toxicity in normal cells, vehicle or DPSE was injected into athymic nude mice daily for 3 weeks. Tissues from the lung, heart, liver and kidney were then sectioned and H&E stained, followed by examination of their histopathology. Mice injected with either vehicle or DPSE did not show any obvious difference in terms of tissue architecture (Fig. 2D). Additionally, the measurement of body weights during drug administration suggests no systemic toxicity from DPSE treatment (Fig. 2E). These results collectively demonstrate that DPSE has minimal toxicity in normal cells and specifically affects cancer cells.

The increase in the cleaved/active form of caspase-3 and caspase-3/7 activities (Fig. 1D and E) indicates the involvement of these proteins in DPSE-induced cell killing. Therefore, the present study investigated the mechanism by which DPSE activates caspase-3/7 in DLBCL. It is well-established that an unbalanced ratio between pro- and anti-apoptotic Bcl-2 family members triggers disruption of mitochondrial membrane potential, which leads to cytochrome c release into the cytoplasm that culminates in the activation of caspases (18). Administration of DPSE markedly suppressed the levels of pro-survival proteins Mcl-1 and Bcl-xL (Fig. 3A), while it augmented the expression of pro-apoptotic Bax and Bak (Fig. 3B). This effect of DPSE was tightly associated with decreased mitochondrial membrane potential, which was confirmed by fluorescence microscopy and flow cytometry (Fig. 3C-E). Together, these data suggest that DPSE induces apoptosis in DLBCL by disrupting the mitochondrial membrane potential and increasing activities of caspases.

We and others have previously shown that apoptosis of DLBCL cells could be induced when Myc is downregulated by Myc inhibitors JQ1 and 10058-F4 (14,19). To investigate the potential involvement of Myc in DPSE-induced apoptosis, four different DLBCL cell lines were treated with DPSE for 24 h and it was identified that Myc levels were markedly reduced (Fig. 4A). Next, the present study aimed to determine whether DPSE affects Myc expression at the transcriptional or post-transcriptional level. The mRNA level of Myc was significantly downregulated upon treatment with DPSE (Fig. 5A), and Myc protein was less stable in the presence of DPSE when protein biosynthesis was inhibited by cycloheximide (Fig. 5B). These data clearly demonstrated that DPSE modulates Myc levels both transcriptionally and post-transcriptionally. To directly test whether Myc plays a critical role in DPSE-induced apoptosis, Ly1 cells ectopically expressing Myc were generated (Fig. S3). Ly1 parental cells and Ly1 cells transduced with a control vector exhibited a similar sensitivity to DPSE, while Ly1 cells transduced with a vector expressing Myc were significantly more resistant to DPSE treatment (Fig. 4B). Furthermore, Myc expression was inhibited by DPSE in Ly1 parental and Ly1 CDH control cells, but not in Ly1 CDH-Myc overexpressing cells (Fig. 4C). The same phenomenon was observed with JQ1, a well-characterized Myc inhibitor (20). In this previous study, JQ1 suppressed endogenous, but not exogenous, Myc expression, and ectopic Myc expression using the MSCV retroviral vector abolished the negative impact of JQ1 on cell proliferation. When treated with JQ1, the fractions of control and Myc-overexpressing cells in the S phase was decreased by 51 and 24%, respectively (20). Taken together, the results indicate that DPSE-triggered cell death is accomplished via, at least in part, a downregulation of Myc.

A previous study have shown that terpenoid and flavonoid compounds have anticancer activities (21). To gain insight into the nature of apoptosis-inducing compounds in DPSE, TLC fractionation was performed that resulted in total 12 fractions of terpenoids (1T-6T) and flavonoids (1F-6F). Ly1 and DHL4 DLBCL cells were exposed to each fraction for 24 h and cell viability was examined by the MTS assays. None of the terpenoid fractions elicited any discernible cytotoxic effect (Fig. S4). A significant dose-dependent decrease in cell viability was observed only with 6F, implying that most, if not all, of anti-lymphoma effect of DPSE stems from flavonoid compounds in 6F (Figs. 6A and S5).

The present study performed Annexin V/PI staining followed by FACS analysis to examine the apoptotic fraction after 6F treatment. Consistently, the apoptotic rate was significantly increased when Ly1 cells were exposed to 6F for 24 h (Figs. 6B and S6). To gain mechanistic insight into 6F-induced apoptosis, the levels of Myc and anti- and pro-apoptotic Bcl2 family proteins were examined after treatment with 6F. Myc and anti-apoptotic Bcl2 members, Mcl-1 and Bcl-xL, were significantly downregulated, while pro-apoptotic proteins Bax and Bak were significantly upregulated (Fig. 6C and D). These data suggest that 6F-induced apoptosis is tightly associated with down- and upregulation of anti- and pro-apoptotic factors, respectively. Next, HPLC/MS analysis was conducted to identify molecular formulas of flavonoids in 6F, and the most abundant compounds separated by the LC peaks are integrated (Fig. 6E and Table I). Expanded analyses of these integrated peaks from MS data are presented in Table SI.

Taken together, the data revealed the anticancer effect of DPSE in DLBCL and the underlying mechanism. DPSE-induced cell killing was shown to be mediated by disruption of the mitochondrial membrane potential via an imbalance between pro- and anti-apoptotic Bcl-2 proteins, in which p53 and Myc play a central role as upstream regulators of Bcl-2 family proteins (Fig. 7 and S7).