Ganoderma lucidum polysaccharides in human monocytic leukemia cells: from gene expression to network construction

The induction of TNF-α in F3-treated human monocytic cells THP-1

Upon the binding of TNF-α to TNFR1, monocytic cells are triggered to undergo apoptosis. This critical regulatory process is accomplished by activating the caspase cascade that results in the degradation of various important cellular proteins. Previous reports showed that lipopolysaccharide (LPS) could markedly stimulate the cytokine expression, especially TNF-α [37]. Compared with the TNF-α expression in LPS-induced THP-1 cells, we could estimate the optimal effect concentration of F3. We treated THP-1 cells with different F3 concentrations (1, 10, 50, 100, 200 μ g/mL) and LPS (1 μ g/mL) for 24 hours, and measured their TNF-α expressions. Figure 1 demonstrates that F3 was dose dependent in the activation of TNF-α expression. TNF-α expression stimulated by F3 at 100 μ g/mL and 200 μ g/mL was similar to that of LPS at 1 μ g/mL. From these TNF-α expression data, we calculated the EC₅₀ (50% effect concentration) of F3-induced effectiveness to be around 10 μ g/mL. If 70% of TNF-α is expressed in F3-induced cells compared to LPS-induced cells, then 30 μ g/mL of F3 is required to achieve the same effectiveness. For further experiments, we used 30 μ g/mL of F3 to treat THP-1 cells.

Apoptotic effect of F3 on THP-1 cells

After THP-1 cells were treated with F3 (30 μ g/mL) for 48 hours, we observed the change of cell morphology under phase-contrast microscope. Differences in cell morphology can be detected between un-treated and F3-treated THP-1 cells. In Figure 2, representative photos of DAPI-staining results are shown. Cell shrinkage, one of cell death characteristics, happened in THP-1 cells after 48 hours treatment with F3 (Figure 2A and 2B). During cell apoptosis, an early event is the nuclear chromatin condensation, leading to the degradation of genomic DNA. DAPI nuclear staining was performed to check the apoptotic changes shown by cell morphology (Figure 2C and 2D). The percentage of chromatin condensed cells in F3-treated culture saw a significant increase (Figure 2E). Shrunken nucleus and apoptotic bodies in DAPI staining were features in determining whether cells had undergone apoptosis. These results indicated that incubation of THP-1 cells with F3 for 48 hours would lead to cell aggregations and apoptosis.

Recent studies showed G. lucidum to induce apoptosis in many cancer cells such as lung cancer cells [38], leukemia cells [39, 40], murine skin carcinoma cells [41], colonic carcinoma cells [42], prostate cancer cells [6], and breast cancer cells [43]. In our study, we shared the same results with other reports, but more specifically in that F3 extracted from G. lucidum could in fact induce apoptosis of leukemia cells THP-1. In our effort to better understand the mechanism of this apoptotic effect, we used microarray, a high-throughput technique, to measure the dynamic gene expression in F3-treated THP-1 cells.

Microarray analysis of overall gene expression in F3-induced THP-1 cells

In order to identify patterns of gene expression associated with apoptotic effect in THP-1 cells induced by F3, we performed a transcriptomic analysis on the THP-1 control and F3-treated or LPS-treated THP-1 cells by oligonucleotide microarray. In Figure 3, we observed clear separation of control (0 and 6 hours) and F3-treated (6 and 24 hours) samples after performing principle component analysis on the gene expression profiles measured by microarray experiments. Figure 4 shows the flow chart for our microarray data analysis. In Figure 5, the intensities of gene expression of one experiment were plotted on the x axis and the intensities of the other experiment on the y axis. A single dot represents one gene as shown in the scatter plots. The upper three scatter plots showed no difference between the control experiments for 0 hour and 6 hours. These results showed the consistency of our duplicate microarray experiments.

Genes with 2-fold change in gene expressions between THP-1 control and F3- or LPS-treated THP-1 cells were selected for mapping significantly disturbed biological pathways. The pathway of apoptosis induction through the DR3 and DR4/5 death receptors was observed to be significant (p < 0.05) in F3-treated THP-1 cells.

Significant biological pathways related to F3-induced THP-1 cells

In the pathway of apoptosis induction through DR3 and DR4/5 death receptors, 9 genes were found to be up-regulated and 2 to be down-regulated among the 27 genes in F3-treated THP-1 cells after F3 treatment for 6 hours. The 9 up-regulated genes include tumor necrosis factor (ligand) superfamily, member 10 (TNFSF10 or TRAIL), tumor necrosis factor receptor superfamily, member 10b (TNFRSF10B or DR5), caspase 10, apoptosis-related cysteine peptidase (CASP10), BH3 interacting domain death agonist (BID), CASP8 and FADD-like apoptosis regulator (CFLAR), TNFRSF1A-associated via death domain (TRADD), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (NFKBIA), nuclear factor of kappa light polypeptide gene enhancer in B-cells 1 (NFKB1), and caspase 7, apoptosis-related cysteine peptidase (CASP7); all of these genes are involved in cell death. Two down-regulated genes were DNA fragmentation factor, 40kDa, beta polypeptide (caspase-activated DNase, DFFB) and caspase 6, apoptosis-related cysteine peptidase (CASP6). CASP6 cleavage by caspase-3 (CASP3), caspase-8 (CASP8) or -10 (CASP10) generates the two active subunits. In the microarray gene expression results, CASP3 and CASP8 showed no significant difference after F3 treatment. However, in our Q-PCR gene expression results, CASP8 showed a significant up-regulated expression. CASP3 and CASP7 were cleaved into their active forms after F3 treatment as shown in Figure 6. In summary, F3 may bind to death receptor 4/5, and activate downstream apoptosis-related cysteine peptidase such as CASP8, CASP3 and CASP7, leading to the apoptosis of THP-1 cells.

The NF-κ B signaling pathway includes 21 genes among which 6 genes were up-regulated after treatment with F3. These six genes were TRADD, myeloid differentiation primary response gene (88) (MYD88), interleukin 1, alpha (IL1A), interleukin 1, beta (IL1B), NFKBIA and NFKB1. All of these genes are involved and are crucial in cell death. IL1B at low concentration induced strong apoptotic responses as revealed by caspase-8 activation and DNA fragmentation [48]. MYD88 is an adapter protein involved in Toll-like receptor and IL-1 receptor signaling pathways in the innate immune response and acts via interleukin-1 receptor-associated kinase (IRAK1) and TNF receptor-associated factor 6 (TRAF6), leading to NF-κ B activation, cytokine secretion and inflammatory responses. MYD88 can also induce IL-8 transcription and may be involved in myeloid differentiation. TRADD was observed to be overexpressed over 25-fold after F3 treatment for 3 hours. TRADD is an adapter protein in the tumor necrosis factor receptor superfamily, in which member 1A (TNFRSF1A/TNFR1) specifically associates with the cytoplasmic domain of activated TNFRSF1A/TNFR1, mediating its interaction with Fas (TNFRSF6)-associated via the death domain (FADD) [31]. Overexpression of TRADD leads to two major TNF-induced responses, apoptosis and the activation of NF-κ B. NF-κ B is a transcription factor regulating the expressions of a large number of genes critical in apoptosis regulation. NF-κ B is strongly activated shortly after TNF engagement with TNFR1, generating a pro-survival signal that must be overcome in many cell types for TNF to induce apoptosis [31]. Many reports also showed that Apo2L/TRAIL could activate NF-κ B [29, 49]. In our results, activation of the NF-κ B signaling pathway caused further THP-1 cell death after F3 treatment.

The TNFR2 signaling pathway includes 17 genes among which 7 genes were differentially expressed in F3- or LPS-treated THP-1 cells. All of these seven genes were up-regulated after treatments with F3 or LPS, as shown in Table 2. TNF binds to two different receptors, TNF receptor 1 (TNFR1) and TNFR2 (TNFRSF1B). TNFR2 is produced by activating lymphocytes and can be cytotoxic to many types of tumors and cells. TNFR2 expression is to mediate the anti-tumor effect of TNF, and NO is necessary for this process, possibly by inhibiting angiogenesis in the tumor [50]. Like TNFR2, TNF receptor-associated factor 1 (TRAF1) was also up-regulated in F3- or LPS-treated THP-1 cells. TRAF1 is an adapter protein and signal transducer that links members of the TNFR family to different signaling pathways by association with the receptor cytoplasmic domain and kinases. TRAF1 is involved in apoptosis by mediating the activation of NF-κ B and c-Jun N-terminal kinase (JNK) [51]. TRAF family member-associated NF-κ B activator (TANK) acts as a regulator of TRAF function by maintaining TRAF in a latent state. Alpha-induced protein 3 (TNFAIP3), a tumor necrosis factor, was induced by TNF-α and identified as a regulatory component of a putative cytoplasmic signaling cascade that mediates NF-κ B activation in response to DNA damage [52].

Toll-like receptor pathway was observed to be significant (p < 0.05 and q < 0.05) in LPS-treated but not in F3-treated THP-1 cells (Table 1). This pathway includes 27 genes among which six and seven genes were differentially expressed in F3- and LPS-treated THP-1 cells, respectively. After F3 treatment for six hours, MYD88, eukaryotic translation initiation factor 2-alpha kinase 2 (EIF2AK2), NFKBLA, mitogen-activated protein kinase kinase 3 (MAP2K3), mitogen-activated protein kinase 8 (MAPK8) and v-Jun sarcoma virus 17 oncogene homolog (avian) (JUN) were up-regulated and CD14 antigen (CD14) was down-regulated. Interestingly, CD14 was up-regulated after treatments with LPS and F3 for 24 hours. CD14 cooperates with TLR4 to mediate the innate immune response to LPS and acts via MYD88 and TRAF6, leading to NF-κ B activation, cytokine secretion and the inflammatory response. Lymphocyte antigen 96 (LY96) was up-regulated after treatment with F3 or LPS for 24 hours. The downstream signaling pathway used by toll-like receptors are similar to that used by IL-1 receptors, which is activating the IL-1 receptor associated kinase (IRAK) through the MYD88 adaptor protein, and signaling through TRAF-6 and protein kinase cascades to activate NF-κ B and Jun [53]. In conclusion, the expression of these genes in the toll-like receptor pathway may implicate their associations with THP-1 apoptosis after F3 or LPS treatment.

Comparisons between oligonucelotide microarray and Q-PCR results

To further validate our findings from the microarray analysis, we selected a set of genes known for their involvement in apoptosis through death receptors, and carried out gene expression studies using Q-PCR. 18 gene expressions related to the death receptor pathway were examined between F3- or LPS-treated THP-1 cells for 3, 6, 12, and 24 hours. mRNAs were reverse-transcribed and amplified through Q-PCR using primers specific for each gene of interest; the housekeeping gene, GAPDH, was used as internal control. Each experiment was repeated three times. The results are shown in Figure 7 and 8. Significance analysis of Q-PCR measurements was performed by EDGE software package [54]. There were nine significantly differential gene expressions in F3-treated THP-1 cells, including TNFRSF10B, CASP7, CASP6, TRADD, CASP3, TNFSF12, baculoviral IAP repeat-containing 2 (BIRC2/c-IAP), conserved helix-loop-helix ubiquitous kinase (CHUK/IKKα), and NF-κ B (NFKB1); whereas in LPS-treated THP-1 cells these were TNFRSF12, CASP7, TNFSF12, CASP6, TRAF2, and NFKB1. Figure 9 shows that nearly all genes are consistent between the microarray and Q-PCR data.

We compared the 18 gene expressions in the time courses after F3 or LPS treatments, as shown in Figure 7 and 8. Statistically significant difference of time-course gene expression profiles between F3 and LPS treatment was observed for TRADD, CHUK, TRAF2, BIRC2, TNFSF10, FADD, CASP3, CASP10, CASP6, BCL2, and NFKB1, while other genes exhibited similar expression trends. These results suggest that THP-1 cells have similar but different gene expression response to F3 and LPS treatments.

Construction of cell death gene networks and possible pathways involved in F3-induced THP-1 cell death

Based on our time-course Q-PCR data, we constructed the cell death gene network in F3-treated THP-1 cells by our self-developed software tool, BSIP [55]. Assuming that gene regulatory network follows the S-system mathematical model [56], we used BSIP to estimate the modeling parameters, identify the optimized structure, compare the concordance, and infer a plausible regulatory network. Moreover, we established the interrelated apoptosis pathways initiated by F3 in THP-1 cells using our concurrent findings, as illustrated in Figure 10. Death receptors, which belong to tumor necrosis factor (TNF) gene superfamily, are cell surface receptors that transmit apoptosis signals and play a crucial role in apoptosis and cell survival. Our characterized death receptors are TNFR1 (tumor necrosis factor receptor-1), TNFR2 (tumor necrosis factor receptor-2), and DR4/5 (also called TRAIL-R1/2) [31]. TNF-α and TNFSF10, also called TRAIL, exhibit an increase in gene expression in F3-treated THP-1 cells. When TNF-α binds to TNFR1, adaptor protein TRADD recruitment follows, and then it interacts with another death domain-containing molecule FADD, leading to the subsequent cleavage of pro-caspase-8 [57]. The binding of TRAIL to DR4/5 also induces the recruitment of FADD and pro-caspase-8 of auto-proteolytic activation [37]. Interactions between pro-caspase-8 and FADD result in further activation of caspase-3 and caspase-7, and also the initiation of apoptosis [58]. In addition to apoptotic signals, recruitment of TRAF2 through TNFR2 binding TNF-α activates NF-κ B, thus producing anti-apoptotic signals [59]. TNFR1 also activates anti-apoptotic NF-κ B. NF-κ B-dependent activation of TRAF2, RIPK1, and BIRC2 gene expressions seemed to support this because TRAF2 and RIP interacting with TRADD prevented caspase-8 activation, while c-IAPs inhibited caspase-3 and caspase-7 activation [60]. When NF-κ B is activated, Iκ B is phosphorylated by protein kinase IKK and this phosphorylation serves as a signal for the ubiquitination and degradation of Iκ B. Free NF-κ B dimers are released and translocated to the nucleus, where they enhance the transcription of target genes. NF-κ B activation and Iκ B degradation prompt cell survival signals and mediate immune responses [61]. In contract, NF-κ B is a transcription factor that regulates expression of a large number of genes critical for the regulation of apoptosis, and many reports have showed that Apo2L/TRAIL can activate NF-κ B [29, 49]. All in all, NF-κ B activation can promote apoptosis or survival, depending on the cellular contents [49].

Materials

Crude Reishi extract (prepared via alkaline extraction (0.1N NaOH), neutralization and ethanol precipitation) was obtained from Pharmanex Co. (CA, USA). All chemicals and reagents were from Sigma-Aldrich Co., unless indicated.

Purification of Reishi extract

Crude Reishi powder (obtained from Pharmanex Co.) 6 g was dissolved in 120 mL of dd water, stirred at 4°C for 1 hour, and centrifuged (1000 rpm) for 1 hour to remove the insoluble. The resulting solution was concentrated at 40~ 50°C to give a small volume which then was lyophilized to generate 5 g (83%) powder of dark-brown color. This water soluble residue was stored at -20°C for further purification.

Isolation of the F3 fraction of Reishi polysaccharide [8]

F3 was isolated from the dark powder of water soluble residue of Reishi polysaccharide. The procedure during chromatography was maintained at 4°C in cool room. The 2.1 g sample was dissolved in a small volume of Tris buffer (pH7.0, 0.1N) containing 0.1% sodium azide, and purified by gel filtration chromatography using a Sephacryl S-500 column (95 × 2.6 cm) with 0.1N Tris buffer (pH 7.0) as the eluent. The flow rate was set at 0.6 mL/min, and 6.0 mL per tube was collected. After the chromatography, each fraction was subjected to phenol-H₂SO₄ method to detect the content of sugar in each tube. Five fractions were collected (fractions 1–5), F3 fraction was concentrated at 40~ 50°C to give a small volume in rotary vapor, which was dialyzed or purified through G-50 column to remove excessive salt and sodium azide and was then lyophilized to give 520 mg (25%) of F3.

THP-1 cell culture and induction of cell death

The monocyctic cell-line THP-1 cells were seeded at an initial concentration of 10⁵ cells/mL in RPMI 1640 Medium supplemented with 2 mg/mL sodium bicarbonate (Atlanta Biologicals, GA, USA), 4.5 mg/mL glucose, 2 mg/mL HEPES, 1% Antibiotic-antimycotic, 10% serum fetal bovine. Cells were cultured at 37°C in an incubator with controlled humidified atmosphere containing 5% CO₂. We detected apoptotic cells by 4,6-diamidino-2-phenylindole (DAPI; Sigma, St. Louis, MO, USA) staining. After washing with PBS and fixing with 4% paraformaldehyde for 15 min, the cells were stained with 2 mg/mL DAPI for 20 min. The stained cells were examined under a fluorescent microscope and cells were considered to undergo apoptosis based on the appearance of nuclear fragmentation.

Detection of TNF-α activity by colorimetric sandwich ELISA

The induction agent, lipopolysaccharide (LPS) or F3, was dissolved in RPMI 1640 Medium at a stock concentration of 4 mg/mL and 2 mg/mL, respectively; then stored at -20°C. THP-1 cells with 10⁵ cells/mL concentrations were seeded in 96-well microplates and incubated overnight. Then the cells (1.25 × 10⁴) were treated with F3 at the dosages indicated as 1 μ g/mL, 10 μ g/mL, 50 μ g/mL, 100 μ g/mL, 200 μ g/mL, and with LPS at the dose of 1 μ g/mL, respectively. The same volume of medium was applied as control. After 24 hours, the supernatants were collected at indicated time points by centrifugation at 1200 g for 5 min. In vitro TNF-α activity was determined using Human TNF-α Immunoassay Kit (Quantikine^®, RD systems) based on the manual provided. The procedure is briefly described as follows. Assay Diluent RD1F solution (50 μ L) and 200 μ L of sample supernatants, standard and control, were gently mixed and loaded into individual wells. After 2 hours of incubation at room temperature, each well was aspirated, washed three times with 400 μ L of Wash Buffer, loaded with 200 μ l of secondary antibody solution (TNF-α Conjugate solution) at room temperature for 1 hour. The same aspiration and wash procedures were performed; 200 μ L of substrate solution was then added to each well and incubated in darkness at room temperature for 20 min. The enzymatic reaction was finally terminated by the addition of 50 μ l of Stop Solution. The optical density was determined using a microplate ELISA reader (Molecular Devices Corporation, California, USA) set at 450 nm with the correction wavelength at 540 or 570 nm. The concentration of TNF-α was determined by plotting the sample reading against the standard curve. All the measurements were performed in triplicate.

Microarray gene expression analysis

THP-1 cells with 10⁷ cells/mL concentrations were seeded in 100 mm dish and incubated overnight. After that, cells were treated with F3 and LPS at a final concentration of 30 μ g/mL and 1 μ g/mL, respectively. After incubated for 6 and 24 hours, the cell pellets were collected by centrifugation at 250 g for 5 min, correspondingly. Controlled samples of un-induced cells were treated in the same way with the same amount of medium. Total cellular RNA was extracted from a minimum of 5 × 10⁶ cells with the use of TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA); RNA was additionally purified with phenol-chloroform-isoamylalcohol (25:24:1). Purity was confirmed by spectrophotometry (A₂₆₀/A₂₈₀ ratio) and capillary electrophoresis (Agilent 2100 Bioanalyzer, Agilent Inc, Foster City, CA). RNA processing and hybridization onto Affymetrix (Santa Clara, CA) HG-U133A GeneChip oligonucleotide microarray were performed according to the manufacturer's protocol. Initial data analysis was performed using Affymetrix Microarray Suite v5.0 software, setting the scaling of all probe sets to a constant value of 500 for each GeneChip. Additional data analysis was performed using GeneSpring v5.1 (Silicon Genetics Inc., Redwood City, California). Principle component analysis on gene expression profiles of control (0 and 6 hours) and F3-treated (6 and 24 hours) samples were performed using Cluster 3.0 [62]. Genes with 2-fold change in differential expression between THP-1 control (combining 0 and 6 hours in total of four arrays) and F3- or LPS-treated THP-1 cells were selected for further analysis. Based on fold-change expression, genes that were differentially expressed between the two repeated microarray experiments were subjected to a scatter-plot analysis.

Functional classification of differentially expressed genes

The differentially expressed genes were classified into groups based on their annotated functions using the BGSSJ software tool [63]. BGSSJ is an XML-based Java application that organizes lists of genes or proteins according to Gene Ontology database for biological interpretation [64], and organizes information based on molecular functions, biological processes, and cellular components for a number of different organisms. For our transcriptomic data, the UniGene IDs of the differentially expressed genes were used to input into BGSSJ with a database option of "NCBI + GO".

Biological pathway annotations

The differentially expressed genes were annotated to specific biological pathways. For each UniGene ID, we retrieved its biological pathways from either BioCarta [44] or KEGG [45] through the existing NCI CGAP gene information database [46]. Biological pathways were mapped and sorted on the order of matching significance using the ArrayXPath web tool [47].

Statistical analysis for identification of significantly disturbed pathways

The disturbed pathways with statistical significance were identified based on Fisher's exact test and false discovery rate. The statistical analysis was performed using ArrayXPath.

Primer design

Quantitative real-time PCR

Gene expression of the genes in death receptor pathway was determined by RT-PCR. Total RNA were isolated from cell lines using TRIzol reagent according to the manufacturer's protocol. First-strand cDNA synthesis was performed on 5 μ g of total RNA in a volume of 20 μ l with 1 μ l ThermoScript™ Reverse Transcriptase (Invitrogen) and 1 μ l oligo(dT). Extracted first-strand cDNAs were analyzed using a BioRad iCycler iQ Real-Time Detection System with SYBR Green dye (Molecular Probes, Eugene, OR). SYBR Green yields a strong fluorescent signal on binding double-stranded DNA enabling the quantification of gene expression by measurement of the intensity of the fluorescent light. SYBR Green will bind to any PCR product amplified by primers and is therefore, non-specific and less accurate than designing specialized probes. Targeted products were between 100–200 bp. For each experimental setup, a standard curve was prepared consisting of 5 dilutions of PCR products. Software produced for the BioRad iCycler will aid in evaluating collected data. mRNA expression of these genes were normalized to RNA content for each sample by using GADPH gene products as internal controls. The relative expression was calculated as the ratio of expression from each F3-treated THP-1 cells compared with the control THP-1 cells.

Significance analysis of Q-PCR experiments

Significance analysis of Q-PCR experiments was performed by the EDGE software tool [54, 66]. EDGE is based on Optimal Discovery Procedure, a new method that finds the optimal rule for calling differentially expressed genes, and Time Course Methodology.

Western blotting

F3-treated and untreated THP-1 cells were washed with PBS twice. Cell pellets (1 × 10⁷ cells) were solubilized in lysis buffer containing 7 M urea, 4% CHAPS, 2M thiourea, and 0.002% bromophenol blue. Lysates were centrifuged at 13,200 × g for 30 mins. Proteins were loaded into 10% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA) at 150 V for 1.5 hours. After blocking in 5% nonfat milk in PBST containing 0.05% Tween 20 (Sigma) at room temperature overnight with gentle rocking, membranes were probed with antibodies. Primary antibodies involved in this study include CASP3 (IMGENEX, San Diego, CA) and CASP7 (Upstate, Lake Placid, NY) diluted in 1:200 and 1:300, respectively. Membranes were incubated with corresponding primary antibody and then incubated with secondary antibodies (biotinylated anti-mouse IgG-HRP, 1:2000 dilution, Abcam, Cambridge, UK). After incubation with secondary antibodies, immunoblots were visualized with the ECL detection kit (Amersham Biosciences) and exposed to X-ray film. β-actin was used as an internal loading control.

Network construction

Based on the time-course Q-PCR data, we reconstructed a plausible gene regulatory network in F3-treated THP-1 cells using the BSIP web tool [55]. We developed BSIP to solve the reverse engineering problems for gene regulatory network or other biological networks. BSIP is a PHP-based web server for identification of biological networks using experimental time-series data obtained from microarray, Q-PCR, proteomics, or metabolomics measurements. Based on the S-system modeling formalism, ${\dot{X}}_i={\alpha }_i{\displaystyle \prod _j{X}_j^{g_{ij}}-}{\beta }_i{\displaystyle \prod _j{X}_j^{h_{ij}}}$, where X _i represents the gene expression level and α _j , β _j , g _ij , and h _ij represent the regulatory parameters, an evolutionary optimization method with data collocation [56] was used to estimate the modeling parameters with the measured time-series data of gene expression and determine the regulatory network structure.