Bioactivity-guided identification and cell signaling technology to delineate the immunomodulatory effects of Panax ginseng on human promonocytic U937 cells

Ginsenosides standards were purchased from Chromadex. HPLC analysis on the composition of ginsenosides in PGSE (2 mg in 1.5 ml of milli-Q water) was performed by using an Agilent 1200 liquid chromatography system that was equipped with a quaternary solvent delivery system, an autosampler and photodiode array detector. A reversed-phase column, Lichrospher C₁₈ (250 mm × 4.6 mm i.d., 5 μm), was used for all separations. The gradient program, modified from a previous report [18], consisted of (A) water and (B) acetonitrile at a flow of 1 mL/min, as follows: 0–6 min, 21–22% B; 6–7 min, 22–23% B; 7–25 min, 23–24% B; 25–30 min, 24–30% B; 30–40 min, 30–32% B; 40–45 min, 32–50% B; 45–60 min, 50–65% B; 60–61 min, 65–100% B; and 61–65 min, back to 21% B before the next injection. The injection volume was 15 μl and the UV detection wavelength was performed at 203 nm for all ginsenosides and PGSE.

The human promonocytic U937 cells [19] were obtained from American Type Culture Collection (ATCC accession no. CRL-1593.2™) and were cultured in RPMI 1640 medium (Invitrogen) supplemented with 10% foetal bovine serum (Invitrogen), penicillin (100 U/ml) and streptomycin (100 μg/ml) in a 5% CO₂ incubator at 37°C. Cells were incubated with TNF-α (20 units/ml) for 2 hours with or without the pre-treatment of PGSE for 24 hours and harvested for genechip analysis. The PGSE concentrations used in our report are based on previous studies of ginseng by other investigators [20, 21] and verified by our cytotoxicity tests. The effective doses of ginsenosides in other groups' in vitro studies ranged from 10 – 100 μM or 0.01 – 0.1 mg/ml. Similarly, the concentrations of individual ginsenosides in 3 mg PGSE used in our experiments ranged from 0.01 to 0.14 mg/ml (Table 1). Therefore, at these low concentrations, it is conceivable that the ginsenoside content of 3 mg/ml PGSE is achievable in vivo. In addition, we determined the cytotoxic effects of PGSE at 3 mg/ml by trypan blue exclusion assay. The viability of cells was over 90% after incubating U937 cells with the PGSE for 48 hours.

Cytotoxic effects of PGSE on U937 cells were examined by incubating 3 mg/ml of PGSE for 48 hours and the cell viability was determined by using trypan blue exclusion test. There is no significant sign of cytotoxicity found at 3 mg/ml of PGSE.

The amount of bacterial endotoxin in PGSE was measured by Pyrotell Limulus amebocyte lysate assay kit (Associates of Cape Cod) according to the manufacturer's protocol. Briefly, 0.2 ml of various concentrations of PGSE was added to a single test vial of Pyrotell. The reaction mixture was incubated at 37°C for 60 min and then inverted to observe the gel formation. Positive result is indicated by the formation of an intact gel which does not collapse upon inversion. The levels of endotoxin in PGSE at 10 mg/ml were lower than the detection limit of the test (<0.05 ng/ml) indicating that the biological effects of PGSE are not due to endotoxin contamination.

U937 cells (1 × 10⁶) were pretreated with or without 3 mg/ml PGSE for 24 hours followed by 20 units/ml TNF-α for 2 hours and Genechip analysis was followed by using Affymetrix's protocol. Briefly, total cellular RNA was extracted using TRIzol (Invitrogen) and further purified by RNeasy cleanup kit (Qiagen) according to the manufacturer's instructions. The RNA integrity was determined by the ratio of 28S/18S ribosomal RNA using Agilent 2100 Bioanalzyer. For genechip analysis, total RNA (1 μg) were reverse transcribed to the first-stranded cDNA by using oligo (dT) linked-T7 RNA polymerase promoter sequence and the double-stranded cDNA was synthesized by using RT Kit (Invitrogen). The biotin labelled-cRNA was generated by in vitro transcription kit (Invitrogen), purified by RNeasy mini columns (Qiagen), denatured and 15 μg cRNA was hybridized to Human Genome U133 Plus 2.0 arrays (Affymetrix). Then, the arrays were stained with a streptavidin-phycoerythrin conjugate and visualized with GeneArray scanner (Agilent). The genechip data were analyzed by using Agilent Genespring GX and Affymetrix GeneChip Operating Softwares (GCOS). The signal intensity of each gene was firstly normalized with the total intensity of all genes from the genechip, and then the normalized signal of each treatment was compared with the mock-treatment to determine the relative fold changes of gene expression. The threshold level for up- or down-regulation of gene expression was the level of changes ≥2-fold.

U937 cells were treated as described in genechip analysis and the procedures of quantitative RT-PCR analysis were described in our previous studies [22‐24]. Briefly, DNase-treated RNA samples were reverse transcribed using TaqMan reverse transcription reagent kit (Applied Biosystems) and the levels of CXCL-10, IL-8 and TNFAIP3 mRNA as well as the reference gene 18S rRNA were assayed by the gene-specific TaqMan gene expression assays (Applied Biosystems). All samples and controls were run in triplicates on an ABI 7500 Real-time PCR system. The quantitative RT-PCR data was analyzed by the comparative cycle number threshold method and the fold inductions of samples were compared with the untreated samples.

U937 cells were pre-treated with or without PGSE (3 mg/ml) for 24 hours prior to TNF-α (20 units/ml) stimulation for 16 hours. After treatment, the levels of CXCL-10 and IL-8 in culture supernatant were measured by using the respective commercially available specific ELISA kits (R&D Systems).

U937 cells were pre-treated with or without PGSE (1 or 3 mg/ml) for 24 hours followed by TNF-α (20 units) stimulation for 2 hours. To prepare the whole cell lysate, cells were washed with PBS and lysed with ice-cold lysis buffer containing 1% Triton X-100, 25 mM HEPES, 5 mM EDTA, 100 mM NaCl, 0.1 mg/ml PMSF, 2 μg/ml aprotinin, 1 mM sodium orthovanadate, 2 μg/ml pepstatin, 2 μg/ml leupeptin, 50 mM sodium fluoride and 10 mM beta-glycerophosphate for 20 min on ice. The total protein was harvested by centrifugation at 13000 rpm for 10 min at 4°C. The supernatants were stored as aliquots at -70°C.

Protein concentration was determined by BCA protein assay reagent kit (Pierce) according to the supplier's procedures. Thirty micrograms of total protein lysate were separated by 10% SDS-PAGE, electroblotted onto nitrocellulose membranes (Schleicher & Schuell), and then probed with anti-phospho-ERK1/2 polyclonal antibodies or anti-phospho-p38 MAPK polyclonal antibodies (Cell signaling). Control blots were immunoblotted with anti-ERK1/2 or anti-p38 MAPK polyclonal antibodies for whole cell lysates. Immuoblots were then incubated with HRP-conjugated anti-rabbit antibodies (BD Bioscience). Finally, the blot was incubated with the Enhanced Chemiluminescence System (GE Healthcare) to detect the target proteins.

All data are presented as the mean ± standard deviation (SD) obtained from at least three separate experiments and statistically analyzed by two-tailed, paired t-test. The statistical significance was defined as *p < 0.05; ^†p < 0.01; ^ψp < 0.005.

To investigate the immunomodulatory activity of ginseng, U937 cells were treated with PGSE and followed by TNF-α stimulation. The gene expression profiles of total cellular RNA were examined by Affymetrix genechip analysis and the data were analyzed by using the Affymetrix GCOS and Genespring GX softwares as described in Methods. To increase the stringency of the analysis, we combined the gene lists from the two software analyses. Only the genes found in both gene lists were reported in this study. Cells with TNF-α or PGSE treatment only were included, and the fold induction of cytokines in cells with treatment was normalized with that of the untreated cells.

Following the sequential treatment of PGSE and TNF-α, we found that 102 upregulated genes and 64 downregulated genes were repeatedly shown in the gene list of two analyses (data not shown). To determine the effects of PGSE on TNF-α signalling pathways, the TNF-α-inducible cytokines and signalling proteins were grouped and summarized in Table 2. Our results showed that PGSE suppressed the transcription of TNF-α inducible genes including CXCL-10, NF-κB inhibitor alpha (IκB-α), G protein-coupled receptor 84, phosphodiesterase 4B, CXCL-11 and CCL-3 in U937 cells. In contrast, PGSE enhanced the transcription of IL-8 with TNF-α, but it did not affect the transcription of CXCL-2, CCL-2, IL-18 receptor, IL-1β and TNF-α-induced protein 3 (TNFIP3). The genechip results of CXCL-10 and IL-8 were validated by quantitative RT-PCR and ELISA. Consistently, PGSE showed inhibition on TNF-α-induced CXCL-10 expression (Figures 1A and 2A) but augmentation of TNF-α-induced IL-8 expression (Figures 1B and 2B). By contrast, there was no significant change of the transcription of TNFIP3 in TNF-α-treated U937 cells with PGSE treatment (Figure 1C).

Since ginsenosides are major active ingredients in ginseng, we examined the composition of ginsenosides in PGSE by HPLC analysis and the results are shown in Figure 3. The calibration curves of the standard solutions containing 0.5–6.5 μg of each ginsenosides were plotted as the peak area versus the amount of selected ginsenosides. Individual ginsenosides from the PGSE were identified and quantified by retention time and peak areas, respectively, as compared to the commercially available pure standards. Nine ginsenosides including Rb₁, Rb₂, Rc, Rd, Re, Rf, Rg₁, Rg₃ and Rh₁ were identified in the PGSE. The amount, concentration and the percentage of each ginsenoside in 3 mg of PGSE are shown in Table 1.

To investigate whether the CXCL-10 suppressive effect by 3 mg of PGSE was due to a specific ginsenoside, U937 cells were treated with individual ginsenosides using the amount as listed in Table 1 for 24 hours and followed by TNF-α stimulation. The level of CXCL-10 transcription was measured by quantitative RT-PCR. With the exception of ginsenosides Rb₁ and Rb₂, our results showed that the CXCL-10 transcription were significantly inhibited by ginsenosides including Rd, Re, Rf, Rg₁ and Rg₃ (p < 0.01), as well as by Rc and Rh₁ (p < 0.05; Figure 4A). However, it is noted that the extent of the suppressive effect of individual ginsenosides on CXCL-10 transcription was still less than that of the PGSE mixture. As ginsenosides accounted for only 18.8% of PGSE by weight; and thus other constituents present in significant concentrations may modulate the activity of the ginsenosides.

We then investigated the combinatorial effect of the nine ginsenosides on TNF-α induced-CXCL-10 transcription. The nine ginsenosides were standardized to concentrations in the PGSE at 3 mg/ml according to Table 1. Moreover, we included a 10-fold dilution ginsenoside mixture to examine the dose-dependent effect on CXCL-10 suppression. Interestingly, the suppressive effect of the reconstituted mixture of ginsenosides at a dose equivalent to 3 mg/ml of PGSE on TNF-α induced-CXCL-10 transcription was comparable to the PGSE treatment (Figures 1A and 4B). Moreover, the suppressive effect of the mixture of ginsenosides occurred in a dose-dependent manner (Figure 4B). To examine the comparable inhibitory effects of PGSE and the mixture of ginsenosides, we measured the percentage change of TNF-α induced-CXCL-10 mRNA after the pretreatment of 3 mg/ml of PGSE, or the mixture of ginsenosides that were equivalent to their corresponding amounts in 3 mg/ml of PGSE. Our results showed that the mixture of ginsenosides gives comparable inhibition of CXCL-10 transcription to those with PGSE (p < 0.005, Figure 4C), but the percentage change of CXCL-10 mRNA between these two treatments was not statistically significance (p > 0.1). Hence, our results indicated that the suppressive effect of PGSE on TNF-α induced-CXCL-10 transcription can be due to the combinatorial effect of ginsenosides.

To investigate the underlying mechanisms of the suppressive effect of the PGSE on CXCL-10 induction, we measured the activities of MAP kinases, including ERK1/2 and p38MAPK, by Western analysis. Intense activation of phospho-ERK1/2 and phospho-p38MAPK was detected after TNF-α stimulation (lane 1, upper panel, Figure 5A and 5B). However, the level of ERK1/2 phosphorylation was decreased with PGSE pretreatment (lanes 2–3, upper panel, Figure 5A). In contrast, the PGSE did not show inhibitory effects on TNF-α activated phospho-p38MAPK activity (lanes 1–3, upper panel, Figure 5B). Interestingly, we found that PGSE inhibited the basal level of ERK1/2 phosphorylation at 1 or 3 mg/ml (lanes 2 and 3, Figure 5C). Equal loading amount of the proteins in the blot was shown by staining the immunoblot with anti-ERK1/2 antibodies (low panel, Figure 5C). In addition to the MAPK signalling pathways, we examined the effects of PGSE on the nuclear translocation of transcription factor NF-κB in the TNF-α treated cells by Western analysis. However, the PGSE did not inhibit the nuclear translocation of p50 and p65 subunits of NF-κB in the TNF-α treated-cells suggesting that the PGSE targets the ERK1/2 signalling pathways (data not shown).