Genomics and proteomics of immune modulatory effects of a butanol fraction of echinacea purpurea in human dendritic cells

Background Echinacea spp. extracts and the derived phytocompounds have been shown to induce specific immune cell activities and are popularly used as food supplements or nutraceuticals for immuno-modulatory functions. Dendritic cells (DCs), the most potent antigen presenting cells, play an important role in both innate and adaptive immunities. In this study, we investigated the specific and differential gene expression in human immature DCs (iDCs) in response to treatment with a butanol fraction containing defined bioactive phytocompounds extracted from stems and leaves of Echinacea purpurea, that we denoted [BF/S+L/Ep]. Results Affymetrix DNA microarray results showed significant up regulation of specific genes for cytokines (IL-8, IL-1β, and IL-18) and chemokines (CXCL 2, CCL 5, and CCL 2) within 4 h after [BF/S+L/Ep] treatment of iDCs. Bioinformatics analysis of genes expressed in [BF/S+L/Ep]-treated DCs revealed a key-signaling network involving a number of immune-modulatory molecules leading to the activation of a downstream molecule, adenylate cyclase 8. Proteomic analysis showed increased expression of antioxidant and cytoskeletal proteins after treatment with [BF/S+L/Ep] and cichoric acid. Conclusion This study provides information on candidate target molecules and molecular signaling mechanisms for future systematic research into the immune-modulatory activities of an important traditional medicinal herb and its derived phytocompounds.

or aerial parts) are among the top-selling medicinal or food supplement products in the United States and Europe [1][2][3]. Recent studies have shown that treatment with specific Echinacea extracts activates macrophages, natural killer cells, or other immune cells [4][5][6]. Echinacea extracts have also been reported to stimulate the secretion of cytokines such as tumor necrosis factor-alpha, interferon, interleukin-1 and interleukin-6 [7][8][9][10]. In vivo studies have shown that treatment with Echinacea extracts can increase the number of white blood cells in the circulatory system [11], enhance phagocytosis [12], and trigger the alternate complementary pathway [13]. Echinacea extracts have been marketed as possible immune stimulators or enhancers worldwide. These phyto-extracts have been actively evaluated in various clinical studies as candidate therapeutics or preventive remedies for upper respiratory tract infections, common cold, urogenital infection and wound healing [14][15][16][17][18][19]. However, the results from various studies on the efficacy of Echinacea extracts for prevention of experimental colds or common cold have been controversial [16][17][18][19][20]. The most recent study, as a meta-analysis for evaluating the effect of Echinacea, addressed again on the potential use and problems of Echinacea as remedy for common cold/flu [20,21].
Dendritic cells (DCs) are involved in a spectrum of immune cell functions, including antigen-presentation and phagocytic activity, and play important roles in both innate and adaptive immunities [22]. DCs can capture and transfer molecular or cellular information from the body's outside or interface environment to cells of different immune systems. These cells are not only critical for the induction of primary immune responses but are also involved in the regulation of T cell-mediated immunity [22]. Recently, a series of studies developed DC-based immunotherapy or vaccine protocols designed to elicit specific immunity against certain cancers [12,23,24]. Echinacea plant extracts have been shown to have immunemodulatory effects [3,[11][12][13], and we recently reported on the possible use of Echinacea purpurea (Ep) phytocompounds as immune-modifiers for human DC activity [25]. Echinacea extracts have significant and specific modulatory effects on human DCs, but these effects are plant tissue-specific, the bioactivity varying greatly between root and shoot plus leaf (S+L) tissues. In this follow-up study, we further investigated in detail the effect of a partially purified and chemically defined Ep phytocompound mixture on human DCs.
Recent gene expression profiling in DCs have shown that DCs can actively process environmental signals and activate different transcriptional programs in response to distinct stimuli [26]. In this study, we used functional genomics to analyze changes in gene expression in human immature DCs in response to treatment with the butanol-partitioned fraction (BF) of the S+L tissue extracts of E. purpurea [BF/S+L/Ep] and cichoric acid (a major component of this fraction) through Affymetrix gene chip microarray analyses. High-resolution 2-D gel electrophoresis, MALDI-TOF mass spectrometry (MS), tandem MS-MS analysis, and bio-informatics database systems were subsequently employed for proteomics studies in parallel with the genomics studies. Results from these analyses and cell-based bioactivity-guide assays suggest that groups of differentially expressed genes, specific functional genes, and the associated molecular signaling networks can be employed as potential targets for future systematic studies of the response of human DC systems, as a response to traditional herbal medicine formulations or their derived phytocompounds.

Effect of [BF/S+L/Ep] extract on expression of cell surface markers in human iDCs
Flow cytometry revealed that [BF/S+L/Ep] enhanced the expression of CD83, a key marker for DC maturation in iDCs, as compared with vehicle (0.1% DMSO) treatment ( Figure 1) and the percentage of CD83 + expressing cells increased, from 20% to 45% with 10, 50 and 100 μg/ml [BF/S+L/Ep]. However, treatment with the ethyl acetate (EA) fraction of the S+L extract at 1, 10, or 50 μg/ml significantly decreased the percentage of CD83 + expressing cells from 20% to 0% (Figure 1). The observed effects were not caused by cellular mechanisms related to cytotoxicity, since MTT assay results indicated that treatment with EA Specific bioactivities of three subfractions of stem and leaf (S+L) extracts of E. purpurea in human immature dendritic cells (iDC) Figure 1 Specific bioactivities of three subfractions of stem and leaf (S+L) extracts of E. purpurea in human immature dendritic cells (iDC). iDCs were treated for 24 h with the S+L tissue extracts and the derived ethyl acetate (EA), butanol (BuOH), or water fractions. Test cells were subsequently analyzed for cell-surface marker CD83 expression by flow cytometry.
The levels of cichoric acid (4) and rutin (7) Figure 2). The standardized [BF/S+L/Ep] fraction was then systematically used in the following bioactivity assays to evaluate its immune regulatory effects in human iDCs.

Affymetrix DNA microarray analysis of differential gene expression patterns in human iDCs in response to treatment with [BF/S+L/Ep] or cichoric acid
For DNA microarray analysis, iDCs differentiated from primary monocytes in vitro were incubated with [BF/S+L/ Ep] or cichoric acid for 4 or 24 h to characterize early-or late-responsive genes. A total of nine oligo-DNA chips were hybridized to determine the transcriptome profiles in human iDCs. The array data was normalized for identifying differentially expressed genes, clustering and annotation. We selected 382 genes for clustering by gap statistic analysis. Figure 3 shows a heat map of differential gene expression patterns and a plot of the resultant 6 possible gene clusters. The K-means clustering revealed that some genes exhibited a consistent response, for example with increased expression at both 4 h and 24 h after treatment with [BF/S+L/Ep]. Alternatively, some genes highly expressed at 4 h but not at 24 h after treatment, or at 24 h but not at 4 h post-treatment, and these may reflect other sets of specific and differential transient gene expression patterns.
Approximately 119 genes were up-regulated by ≥ 2-fold change in expression at 4 h post-treatment with [BF/S+L/ Ep], with 20 genes showing 11-to 31-fold change in expression (Table 1A). About 20 genes were down-regulated by > 2-fold change in expression at 4 h post-treatment (Table 1A)     up-or down-regulated genes, with 5.36-to 18.9-fold increase or 0.13-to 0.37-fold decrease in expression at 24 h were chosen for further evaluation (Table 2B). As shown in Table 3, among the groups of genes up-or down-regulated by cichoric acid versus [BF/S+L/Ep] treatment for 24 h, an overlapping expression pattern was observed for 31 genes.

Putative signaling networks/pathways involved in the modulatory effect of [BF/S+L/Ep] on iDCs
To identify the possible putative signal transduction pathways, we analyzed, for 4 hr and 24 hr treatments, the top 20 up-regulated genes in human iDCs in response to treatment with [BF/S+L/Ep] using TRANSPATH soft ware. Signal transduction pathways involving the MCP-1, IL-8, CCL5, JAK2 and TRIO genes with ≥ 5-fold change in expression on treatment (Table 1A) was identified. Signaling pathway networking and gene function analyses led to the hypothesis that treatment of iDC with [BF/S+L/Ep] may activate the cyclic AMP and PKC pathways leading to the regulation of a key down stream molecule, adenylate cyclase 8 (AC8), in a Ca 2+ dependent manner ( Figure 4).

Construction of proteome maps and identification of differentially expressed proteins in human iDCs
Treatment of iDCs with 75 μg/ml [BF/S+L/Ep] or 50 μg/ ml cichoric acid resulted in differential expression of 43 proteins (Figures 5 and Table 4) identified by MALDI-TOF-MS. These 43 proteins were further categorized by function with use of the Swiss-Prot database. Most of the proteins responding to [BF/S+L/Ep] (30.2%) were cell growth-or maintenance-related proteins ( Figure 6); these included structural and cytoskeleton proteins such as macrophage-capping protein, cofilin, profilin, F-actin capping protein β subunit, and laminin A/C (   tein 3, coronin-1C, beta adducin, vimentin, WD-repeat protein 1, T-complex protein 1 epsilon subunit, T-complex protein 1-beta subunit, cofilin, F-actin capping protein beta subunit 2, actin-related protein 2/3 complex subunit 2, profilin-1). These specific proteins may warrant further evaluation because of their apparent roles in modulation of DC maturation and function.

Confirmation of up-or down-regulation of specific proteins
Western blot analysis was conducted to confirm some of the up-regulated proteins observed from proteomic analysis. The expression of Mn-SOD (SOD2) was increased 2.65-to 1.99-fold with [BF/S+L/Ep] and 2.35-and 1.67fold with cichoric acid treatment at 12 and 24 h, respectively, as compared with vehicle controls, which is consistent with results from 2-D gel electrophoresis (Figure 7, Table 4). The levels of cofilin determined by 2-D gel electrophoresis ( Figure 8A Figure 8B).

Discussion
Results of some clinical studies are controversial about the effects of the traditional herbal medicine Echinacea extracts [14][15][16][17][18][19], even though recent studies seemed to provide some support for its possible beneficial effects for treatment of the common cold [20,21]. In addition, studies from animal experiments have shown that alkamides form Echinacea extracts may confer immune-modulatory activities [35,36]. Among these different observations, we believe it is important to evaluate systematically the specific and multiple effects of Echinacea phytocompounds on human immune cells, at the cellular and molecular  levels. Previously, we reported that un-fractionated, crude extracts of E. purpurea plant tissues could affect specific cell-surface markers of human DCs [25], which suggests the potential of using human DCs as an experimental system for in-depth study of the effect of Echinacea phytocompounds on important human immune cell systems.
In the current study, we evaluated and characterized the possible immune-modulatory effect(s) of phytocompounds from E. purpurea, as an organic solvent-fractionated phytocompound mixture or a single major component (cichoric acid), on human DCs. Using transcriptome and proteome experimental approaches, we analyzed the global and differential gene expression pat-terns at the mRNA and protein levels in human monocyte-derived iDCs treated with test phytocompounds.
Previous studies have demonstrated that DCs play important roles in initiating and regulating innate as well as adaptive immunities [22]. Under normal conditions, DCs are present in most tissues in a relatively immature state, but with inflammation, irritation or danger signals from foreign antigens or environment, iDCs undergo rapid changes and initiate a cascade of activities to defend the body system. The phenotypic and functional characteristics of DCs are intimately and dynamically linked to their stage of differentiation and maturation [37]. Human DCs  (Table 3), IFIT1 and the interferon alpha-inducible protein (Table 1B), were all down-regulated at 24 h post-treatment. Thus, several key chemokine genes are likely involved in the early response to treatment with specific Echinacea phytocompounds, and another set of interferon-inducible genes belong to the group of late-responsive genes. We also per-  formed gene-clustering analysis to classify the gene expression patterns in human iDCs. A total of 382 genes were grouped into six clustering sets on the basis of their expression in response to [BF/S+L/Ep] (Figure 3). In contrast, cichoric acid and vehicle treatment produced no clustering differences, which suggests the lack of a multidomain or substantial effect of cichoric acid on most expressed genes in iDCs.
We have identified a hypothetical signaling network involving [BF/S+L/Ep]-induced early response genes such as CCL2 (MCP-1), IL-8, CCL5, JAK2 and TRIO (Figure 4). This signaling network revealed possible activation of cAMP and PKC pathways leading to the regulation of AC8 through a Ca 2+ receptor, calmodulin. Previous studies have demonstrated that activation of cAMP induces DC maturation and migration [39]. PKC plays an important role in DC differentiation and antigen presenting function [40]. In addition, Ca 2+ dependent pathways have been shown to regulate DC maturation and migration [41]. Our results on increased expression of CD83 maturation marker coupled with candidate signal transduction pathways strongly suggest that [BF/S+L/Ep] may enhance the iDC maturation and function, which warrants further systematic investigation.
2-D gel electrophoresis revealed 100 of the 1,300 detectable protein spots (≈7.7%) affected in iDCs after 12-or 24h treatment with [BF/S+L/Ep]. We suggest that this 7.7% change in iDC protein levels may reflect a marked change in protein expression, since the other cell types we have studied in parallel (e.g., primary T-cells) showed only a 3% to 5% change (data not shown). We observed significant up-regulation of antioxidant defense enzymes such as Mn-SOD, catalase and peroxiredoxin 6 in iDCs in response to [BF/S+L/Ep] and cichoric acid. Endogenous antioxidants have been reported to play key role in dendritic cell survival, and ability to induce T cell activation and regulate the polarity of immune responses [42,43]. Several cytoskeletal and actin-binding proteins (Table 4) were significantly up-regulated in [BF/S+L/Ep] or cichoric acid treated iDCs. In DC, the changes in actin cytoskeleton components are essential for the formation of its characteristic dendrites and veils, as well as an immunological synapse necessary for antigen presentation [44]. In addition, actin cytoskeleton rearrangement is important for the motility and migration of cells and may influence the migration of DC to lymph nodes during their maturation [45]. Results from 2-D gel electrophoresis (Figures 7A and  8A) and Western blot analyses ( Figures 7B and 8B) confirmed in general the trend of changes in differential expression of Mn-SOD and cofilin. We therefore hypothesize that a specific group of phytocompounds in Echinacea may turn on the antioxidant defense system and regulate the rearrangement of cytoskeleton, which in turn may contribute to the positive enhancement of the immunemodulatory activities of DCs.
The present study has identified a pool of known or unknown genes associated with the differential expression of a spectrum of cytokines and chemokines and other immune cell activities in response to treatment with Echinacea compounds in DCs (Tables 1 to 4, and Figure 3). In addition, changes in level of molecules related to cell adhesion, immune-response and antigen presentation were observed in this human DC model, providing candidate target genes/proteins for future cross-talk studies of the biology of human DCs. These differentially expressed genes and proposed candidate signaling pathways/networks obtained via bioinformatics approaches have provided us with useful potential clues and molecular targets for future studies of molecular mechanisms underlying specific immune modulatory effect(s) of important medicinal herb Echinacea purpurea and derived phytocompounds in iDCs.

Conclusion
Since dendritic cells are well characterized and recognized as potent antigen presenting cells and have been shown to play critical roles in both innate and adaptive immunities, we have employed human DCs for functional genomics studies to evaluate possible immuno-modulatory effects   Based on these results, we conclude from this study that specific phytocompound mixtures present as major or active components in the test traditional herb may confer defined and significant immuno-modulatory effects on specific immune cell types. The candidate target molecules and molecular signaling mechanisms identified from this and previous study [25] provide us with useful information and hypothesis for future studies on characterization of human DCs in response to treatment with medicinal phytocompounds at the transcriptome and proteome levels.

Plant materials and crude extract preparations
E. purpurea plants at flowering stage were harvested from a reputable organic farm in Puli, Nantou County, Taiwan. Stem and leaf (S+L) tissues of fresh plants were extracted at room temperature by imbibition with 70% aqueous ethanol as previously reported [25]. The total 70% ethanolic extracts were concentrated (1 L) and successively partitioned with ethyl acetate (1 L × 3 times) and n-butanol (1 L × 3 times) to yield three sub-fractions, designated as the EA, BuOH and Water fractions. The percentage of yield for the three fractions was calculated as 10.3%, 7.23%, and 77.2%, respectively, of the 70% ethanolic extracts in dry weight.

Generation of human dendritic cells
Monocyte-derived dendritic cells were generated from peripheral blood mononuclear cells (PBMCs) isolated from healthy volunteers blood samples as described previously [25]. Briefly, CD14 + monocytes were isolated by passing the PBMCs through a magnetic cell separation system (Miltenyi Biotec, Germany). CD14 + cells at 95% purity were cultured for 7 days in AIM-V medium in the presence of GM-CSF (1000 U/ml) and IL-4 (500 U/ml) to obtain immature dendritic cells (iDCs).

Flow cytometry
At 24 h after treatment with E. purpurea crude extract, the S+L extract, or the three subfractions, iDCs were harvested and analyzed by immunofluorescence staining. Monoclonal antibodies labeled with fluorescent dyes i.e., HLADR-FITC, CD14-PE, CD32-PE, and CD83-FITC from Immunotech (Fullerton, CA) and CD86-FITC from Pharmingen (Fullerton, CA) were used. After incubation with specific antibodies at 4°C for 30 min, cells were washed twice with PBS and fixed with 1% paraformaldehyde, and subjected to analysis using a Coulter EPICS XL flow cytometer (Beckman/Coulter, Durham, NC).

Cell viability assay
Human iDCs (4 × 10 5 cells/ml) were incubated with 0.1% DMSO (vehicle control), test plant extract or derived phytochemicals in basal medium in 96-well plates for 24 h in a 5% CO 2 incubator. All treatments were performed in triplicate. Viability of iDCs was determined by MTT assay, and percentage cell viability was calculated as previously described [25]. DNA microarray analysis of differential gene expression patterns Total RNA was isolated by use of TRIzol ® Reagent (Invitrogen) according to the manufacturer's instructions to generate cRNA targets. A total of 7 μg of RNA from each sample was used to synthesize the first-strand cDNA with T 7 -Oligo (dT) primer and T7 RNA polymerase by in vitro transcription reaction. The biotinylated cRNA products were then cleaned according to the Affymetrix protocol. Aliquots of 15 μg of total cRNA per sample were then hybridized to the Affymetrix gene chip HG-U133A containing approximately 22,283 probe sets following the manufacturer's instructions. Images were obtained by use of standard Affymetrix scanners. The cell intensities and detection from the nine hybridized oligonucleotide microarrays were derived by use of the Affymetrix Microarray Suite 5 (MAS 5.0) with default settings. Normalization was by dChip with the PM-MM model-based approach to obtain the expression indices for each probe set [46]. The pool of the replicate arrays, CM-00h-1 and CM-00h-2, was used as a baseline for normalization and was applied to the calculation of log2 ratios, M, for the test arrays. Comparative expression analysis of the two arrays by the MAS 5.0 system reported a level of "no change in gene expression pattern" of 98.0% (with a correlation of 0.9916). This high reproducibility of results indicates very high reliability and very low variation of our experimental data obtained from the in-house Affymetrix system. The gap statistic and K means analysis were used to distinguish gene clusters as described previously [47,48]. The putative signal transduction pathways were analyzed using TRANSPATH Professional 7.1 (BIO-BASE Biological Databases, Germany). The microarray data obtained from this study have been deposited to the Gene Expression Omnibus database at NCBI (GEO; http://www.ncbi.nlm.nih.gov/projects/geo/) under the accession number GSE12259.
(Bio-Rad). Protein spots on gels were automatically detected with use of the PDQuest system. The 2-D gel images of iDCs proteins obtained with IPG strips (pH 5-8) were matched separately, with each match set containing 9 images, with 3 images for control replicates. The image with the highest number of spots was selected as the master gel. Automatically detected images of protein spots in test gels were then manually edited to include the low intensity spots and corrected for spot artifacts. For the match set containing images from pI 5-8, the spot volume (intensity integrated over the spot area) was normalized to the total volume of spots in the gel. Data were then exported to Microsoft Excel.

Protein identification by MALDI-TOF-MS
Protein extraction was performed as previously described [50]. Briefly, each gel spot was cut into small pieces with a scalpel, washed with 700 μl of double distilled H 2 O, and subjected to reduction reaction by DTT. Alkylation reaction then carried out by adding 55 mM iodoacetamide for 1 h at room temperature in the dark. Gel spots were washed with ammonium carbonate then acetonitrile. Gel pieces were dried and then immersed in trypsin solution. The in-gel digestion was performed at 37°C overnight. Resulting tryptic peptides were extracted twice from the test gel into 70% acetonitrile/5% HCOOH by sonication then centrifugation. The combined supernatant was dried under a Savant Speed Vac, and 6 μl of 1% HCOOH was added to each test sample. Protein identification by MALDI-Q-TOF-MS [51] was performed at the Proteomics Core Facility of the Institute of Biological Chemistry,