Libraries characteristics

Cell viability was measured after exposure to OenB. A viability of 75% and 85% (data not shown) corresponding to the 90 and 180 min treatments, respectively, were chosen for RNA extraction and library construction. The libraries were constructed using the RDA technique, which is able to identify both up- and down-regulated differentially expressed genes between two cDNA populations [27]. The RDA experimental design included two conditions: (i) Paracoccidioides yeast cells exposed to McVeigh Morton (MMcM) medium with 0.32 M OenB for 90 and 180 min and (ii) Paracoccidioides yeast cells grown in MMcM medium for 90 and 180 min. For the up- regulated gene library construction, the first condition (i) was used as a tester population and the latter (ii) as the driver population. However, for the down-regulated library construction, the former (i) was used as the driver population and the latter (ii) as the tester population. Subtractions were performed by incubating the driver and tester populations. The selection of the cDNAs was achieved by constructing subtracted libraries in pGEM-T Easy, as described in the Methods section.

Bioinformatics results for libraries exposed to OenB

RDA analysis revealed that a large number of ESTs (1011) was differentially regulated in the four gene libraries constructed after OenB treatment. From the RDA library of yeast cells exposed to OenB for 90 min 463 up-regulated and 104 down- regulated genes were identified. After pipeline analysis, the 463 up-regulated sequences originated 36 contigs and 20 singlets; the 104 down-regulated sequences originated 12 contigs and 7 singlets. The same analysis was performed for the RDA library obtained from yeast exposed to OenB for 180 min; 301 up-regulated and 143 down-regulated genes were identified from this library. After pipeline analysis, the 301 up-regulated sequences formed 14 contigs and 8 singlets; the 143 down-regulated sequences formed 12 contigs and 6 singlets. The ESTs obtained were submitted to the National Center for Biotechnology Information (NCBI) under accession numbers: LIBEST_028147 Paracoccidioides oenoteinB 90up Library, LIBEST_028146 Paracoccidioides oenoteinB 90 down Library, LIBEST_028149 Paracoccidioides oenoteinB 180 up Library and LIBEST_028148 Paracoccidioides oenoteinB 180down Library.

Graphs were plotted to demonstrate the statistically enriched GO functions with up- or down-regulated genes after exposure to the compound (Additional file 1: Figure S1). The groups with the highest percentage of up-regulated genes were the unclassified proteins (30%), transcription (23%), functional unclassified proteins (22%), metabolism (10%), cell cycle and DNA processing (7%) and morphology (4%). The highest percentage of down-regulated genes was grouped in the signal transducer (27%), cellular transport, transport facilities and transport routes (22%), transcription (15%), cell rescue, defense and virulence (14%), metabolism (12%), unclassified proteins (4%), protein synthesis (4%) and morphology (2%) groups.

The distribution of differentially expressed genes in the different functional categories and biological processes was evaluated (Table 1; Figure 1). The analysis demonstrated that most of the genes that were down-regulated after 90 min participate in cellular processes related to transport, followed by signal transducers and those involved in transcription and metabolism. On the other hand, the up-regulated genes were mostly found to be functionally unclassified proteins, followed by unclassified proteins, and those involved in cell cycle and DNA processing and metabolism. After 180 min treatment, most of the down-regulated genes were also involved in signal transduction, metabolism and cellular transport. The up-regulated genes were unclassified proteins or those involved in transcription and morphology determination. The percentage of occurrence of each gene in the libraries was indicated in the Additional file 2: Table S1.

The OenB treatment resulted in up-and down-regulated genes involved in different biological processes (Table 1; Figure 2). We analyzed the occurrence of the transcripts by determining the number of ESTs found in each transcript. The highest occurrence of up-regulated ESTs after 90 min of OenB treatment were as follows: DUF 1688 domain protein (154 ESTs), arginine N-methyltransferase- SKB1 (PRMT5) (52 ESTs), GATA-type sexual (NSDD) (22 ESTs), trehalose phosphatase (TPS1) (20 ESTs), histone deacetylase (RPD3) (12 ESTs), glutamine synthetase (GLN1) (11 ESTs), phosphatase regulatory subunit (GAC1) (10 ESTs) and protein with a PYP-like sensor domain (PAS) (10 ESTs). For down-regulated ESTs, the highest expression transcripts were the following: major facilitator superfamily transporter (MFS) (27 ESTs), guanine nucleotide binding protein alpha-1 subunit (GPA2) (23 ESTs) and C6 transcription factor (CTF1B) (15 ESTs).

Likewise, we analyzed the up-and down-regulated genes resulting from 180 min of OenB treatment and noted genes that might be involved in different biological process (Table 1; Figure 2). The highest transcription expression of transcripts after 180 min of treatment were the following for up-regulated ESTs: transcription factor fungi (65 ESTs), glutamine synthetase (GLN1) (1 EST) and arp 2/3 complex subunit Arc16 (12 ESTs). For down-regulated ESTs, the highest transcript expression of transcripts were the following: protein with PYP-like sensor domain (PAS) (44 ESTs), pathogenesis associated protein (CAP20) (21 ESTs), betaine aldehyde dehydrogenase (BADH) (13 ESTs) and RING finger protein (RNF) (15 ESTs). The most evident up- and down-regulated genes are shown in the Figure 2.

Quantitative real-time PCR (qRT-PCR)

RDA is a methodology that enriches for differentially expressed sequences within transcriptional profiles and specifically permits the detection of transcripts that are differentially represented even at low levels of expression, such as transcription factors and other regulatory genes. While representative in general terms, further confirmation of identified sequences is indicated, as in any high throughput approach, and qRT-PCR is a method of choice for this confirmation process.

The expression levels of the up- and down-regulated genes from the library of yeast treated for 90 min were confirmed as significantly higher for PAL1 and GLN1 and significantly lower for the MFS and ERG3 genes, compared to the controls, which was in accordance with the results from the RDA methodology. All four genes from the library of yeast treatment for 180 min also corroborated the results from the RDA methodology because the levels of expression were significantly higher for GLN1 and KRE6 and significantly lower for MFS and CAP20 genes (Figure 3A).

The FKS1 gene was analyzed because we found in our previous work that the gene was repressed in the presence of OenB [19]. Here, the decreased level of FKS1 after 90 and 180 min of treatment are in agreement with our previous data.

Previous studies using electron microscopy have demonstrated the formation of large vacuoles with almost no electron density, the presence of lysed cells, partial disruption of the cell wall and cytoplasmic organelle leakage out of OenoB-treated Paracoccidioides yeast cells [19]. Here, genes related to osmotic stress, such as RPD3, ATF1, HSF1 and GAC1 were up-regulated in Paracoccidioides yeast cells after OenB-treatment. In this way, the GAC1 gene, which responds to osmotic stress [28], was evaluated in the presence of KCl. It is noteworthy that KCl was used in this study as a stressor because it is an osmotic stress inducer and we found that it acted on the cell wall of Paracoccidioides in our previous study [29]. Our results indicated that the GAC1 gene responds to osmotic stress because its expression level was increased in the presence of OenB and KCl (Figure 3B).

Transcript levels in OenB-treated Paracoccidioidesyeast cells after internalization by macrophage cells

To investigate if the transcripts identified by RDA would also be up- or down- regulated in vivo, Paracoccidioides yeast cells were internalized by macrophage cells and the transcript levels were then quantified. J774A.1 mouse macrophage cells infected with Paracoccidioides were treated with OenB for 24 h, and qRT-PCR analysis was carried out for the genes KRE6, GLN1, FKS and ERG3. The expression levels of genes found to be up- and down-regulated in the libraries of yeast cells treated for 90 and 180 min were confirmed. Expression was significantly lower for ERG3 and significantly higher for KRE6 and GLN1 after 24 h of infection. The FKS1 gene, which was down- regulated in the presence of OenB in our previous work [19], was also dow- regulated during macrophage infection (Figure 3C). The results are in agreement with those found in the RDA and qRT-PCR assays when Paracoccidioides yeast cells were treated with OenB in vitro.

Cell analysis by fluorescence microscopy

Calcofluor White (CFW) and Congo Red (CR) both stain the cell wall of fungi by binding to chitin chains [30]. CFW and CR dyes interact with various polysaccharides [31] but exhibit a particularly high affinity for chitin [32, 33].

To investigate the chitin level in the cell wall, Paracoccidioides Pb01 and Pb18 yeast cells were grown in liquid medium for 72 h with and without OenB, after which they were stained with Calcofluor White (CFW) and Congo Red (CR) and visualized by fluorescence microscopy. OenB-treated Paracoccidioides cells showed a stronger fluorescence when compared to the control, indicating increased chitin levels (Figure 4).

Carbohydrate content of the cell wall

The cell wall is divided in alkali-soluble (AS) and alkali-insoluble (AI) fractions according to its solubility in alkaline substances. In the Paracoccidioides Pb01 yeast phase, the glucan polymer consists mainly of 1,3-α-glucan (95%) which is present in the AS fraction, and a small amount of the 1,3-β-D-glucan (5%), which is present in the AI fraction [34]. GlcNAc residues are also found in the AI fraction. Figure 5A shows a decrease in total carbohydrate content in the AI fraction after OenB-treatment; no alteration in carbohydrate content was observed in the AS fraction. 1,3-β-D-glucan content was estimated after the digestion of the AI cell wall fraction with the 1,3-β-glucanase enzyme. As shown in Figure 5B, there was a decrease in 1,3-β-D-glucan content. In addition there was an increase in GlcNAc residue content in this fraction (Figure 5C), suggesting an increase in the chitin polymer concentration.

Plant material

E. uniflora leaves were collected in Anápolis, Goiás, Brazil and the sample of specimens were deposited at the Herbarium of the Universidade Federal de Goiás UFG, with code number UFG 25477. Access authorization and sample remittance of component of the genetic property is n°005/2007.

Extraction, isolation and identification of OenB from E. uniflora

The extraction, isolation and identification of OenB were realized as described previously [19]. Briefly, powdered leaves of E. uniflora (1.0 kg) were extracted with 50% aqueous acetone. The dry crude extract (174.27 g) was obtained by concentration in vacuum followed by freeze-drying and a part (150 g) was extracted with ethyl acetate. The aqueous layer was lyophilized to yield a 122.14 g fraction which was dissolved in methanol (MeOH) to separate the soluble (81.76 g) and the insoluble (39.22 g) methanolic fractions. The insoluble fraction (10 g) was submitted to separation and purification by Column Cromatography on Diaion HP-20, followed by CC on Sephadex LH-20, eluted with a water-MeOH gradient to obtain the IM-7 sub-fraction (791.5 mg). From this, OenB (62.4 mg) was isolated (purity of 96%) on a preparative HPLC system. The chromatographic system consisted of LC-10ADvp pumps (Shimadzu Corporation, Tokyo, Japan) connected to an SPD-10AVvp ultraviolet photodiode array detector. Chromatographic separations were performed using a LiChrospher RP-18 column (Shimadzu Corporation). The compounds were eluted using acetonitrile (A) 0.01 M H₃PO₄: 0.01 M KH₂PO₄ (B) (gradient 8% A in B to 18%) for 20 min at a flow rate of 1.0 mL/min.

Paracoccidioidesculture and determination of cell viability

Paracoccidioides (ATCC MYA 826) has been studied extensively in our laboratory [75–78]. Paracoccidioides Pb01 yeast cells were sub-cultured every seven days on solid Fava-Netto’s medium (1.0% w/v peptone, 0.5% w/v yeast extract, 0.3% w/v protease peptone, 0.5% beef extract, 0.5% w/v NaCl, 4% w/v glucose and 1.4% w/v agar, pH 7.2) [79] at 36°C and used throughout this study. For viability experiments, yeast cells were grown in the presence or absence of OenB for specific time intervals and were kept in MMcM chemically defined medium [80] at 36°C. The cellular viability was determined by the trypan blue method [81]. The quantity of OenB used was 500 μg/mL (0.32 M), which corresponds to minimal inhibitory concentration (MIC) [19]. In brief, cells were incubated with the dye solution (0.1% trypan blue stain) for 5 min at room temperature, and viability was assessed by counting viable and unviable cells in a Neubauer chamber.

J774 A.1 mouse macrophage cell culture and infection with Paracoccidioides

The J774 A.1 mouse macrophage cells were purchased from a cell bank in Rio de Janeiro (Rio de Janeiro, Brazil) and were maintained in cell culture bottles in an atmosphere of 5% carbon dioxide at 37°C in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (Vitrocell/Embriolife, Campinas, SP, Brazil), 10% amino acid solution (Sigma Biochemical, St. Louis, MO, USA) and 0.2% of gentamicin solution (Sigma Biochemical).

The macrophages were quantified by counting them in a Neubauer chamber, after which they were plated at a density of 10⁶ cells per well on glass cover slips in 24-well culture plates. The macrophages were then infected with 2.5 × 10⁶Paracoccidioides Pb01 yeast cells, also determined by counting Neubauer chamber. The cells were co-cultured for 24 h in the presence of 0.32 M OenB at 36°C and 5% CO₂ to allow the internalization of the fungus. After this period, the supernatants were then aspirated, and the layer was observed microscopically, to visualize the fungal cells that had been internalized by the macrophages. The number of viable fungi co-cultivated with the macrophages was determined by quantifying the number of colony forming units (CFUs). Then, the monolayer was gently washed with 1× PBS to remove any non-adherent/internalized yeast cells, and the total RNA was isolated.

RNA isolation and cDNA synthesis

Paracoccidioides Pb01 yeast cells were cultured in MMcM medium only or in the same medium with 0.32 M OenB for 90 and 180 min. For RNA isolation, cells were collected by centrifugation and the RNAs of driver and tester cultures were extracted with Trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA quality was assessed by determining the A_260nm/A_280nm ratio and by the visualization of RNAs by 1.5% agarose gel electrophoresis. The RNAs were used to construct subtracted cDNA libraries. The first-strand cDNAs were synthesized from 1 μg of total RNAs using reverse transcriptase (RT Superscript III, Invitrogen) and were then used as a template from which to synthesize the second-strand using a SMART PCR cDNA synthesis kit (Clontech Laboratories, Palo Alto, CA, USA).

Bioinformatics analysis

The reading quality was checked by the Phred program module [84], and were chose to analyze the sequences with at least 50 nucleotides and a quality grater or equal to 20. Subsequently, vector sequences were trimmed using the Crossmatch program [85]. Readings that had passed the quality check were next submitted to a CAP3 program [86] to obtain the final set of contigs and singlet sequences. All these tools were integrated in a specific pipeline [87]. The sequences obtained were dynamically translated and compared against the GenBank [88] non-redundant (nr) database from NCBI using the Blast X program [89] and the nucleotide database from the Paracoccidioides structural genome [90]. The database sequence matches were considered significant at e-values ≤10^-5.

Sequence analysis and the annotation pipeline were set up using the Blast2GO program [91] that joints in one GO application based on a similarity search with statistical analysis and highlighted visualization on a directed acyclic graph [92]. The Blast2GO annotation algorithm already took multiple parameters into account such as sequence similarity, blast HSP (highest scoring pair) length and e-values, the GO hierarchical structure and GO term evidence codes [92, 93]. The sequences were grouped into functional categories according to the classification of the MIPS functional catalog [94].

Graphs were plotted to demonstrate the statistically enriched GO functions with up- or down-regulated genes, respectively, for the times of exposure to the compound (Additional file 1: Figure S1). The percentage of occurrence of each gene in relation to the total number of genes from the libraries was calculated and shown Additional file 2: Table S1.

Quantitative real-time PCR

For the qRT-PCR analysis of a subset of genes indicated as differentially expressed, gene-specific primers were designed using the software Primer Express (Applied Biosystems, Foster City, CA, USA). The predicted product lengths varied between 100 bp and 200 bp. The reference α-tubulin gene was used for normalization. Total RNA was extracted from control and OenB-treated yeast using Trizol reagent. The total RNA was also extracted from control and OenB-treated cells after infection with Paracoccidioides.

First strand cDNA was produced using a SuperScript III (Invitrogen) and oligo (dT)₁₅ primer. All qRT-PCR assays were run using a SYBR Green (Applied Biosystems) protocol in a StepOnePlus^TM Real-Time PCR system (Applied Biosystems). The amplification protocol was 40 cycles of 95°C for 15 s; 60°C for 1 min. The qRT-PCR assays were performed in triplicate.

The SYBR Green PCR master mix (Applied Biosystems) was used as the reaction mixture, and 10 pmol of each specific primer and 40 ng of template cDNA were added to a final volume of 25 μL. A melting curve analysis was performed to confirm a single PCR product. The data were normalized against α-tubulin in each set of qRT-PCR experiments. The relative expression levels of the genes of interest were calculated using the standard curve method for relative quantification [95]. The Student’s t test was used for statistical comparisons and P values ≤ 0.05 were considered statistically significant. The specific sense and antisense primers are listed in Table 2.

Dosage of cell wall polymers

To measure the amount of total carbohydrates, Paracoccidioides Pb01 yeast cells were grown in liquid Fava-Netto’s medium at 36°C for 72 h with or without OenB. The cell wall polymers were obtained as previously described [29]. In brief, yeast cells were collected and washed with acetone, ethyl alcohol and ether. Then, the cells were suspended in 50 mM TrisHCl pH 7.5 and disrupted with glass beads. Cell walls were separated from the cytosolic fraction by centrifugation. The alkali-soluble (AS) fraction was extracted with NaOH and sodium borohydride. Total carbohydrates were determined by the phenol sulfuric acid procedure. The amount of 1,3-β-D-glucan in the AI fraction was estimated by measuring the release of reducing sugars.

The amount of N-acetylglucosamine was determined as previously described [29]. Briefly, the Paracoccidioides Pb01 yeast cells enriched cell wall fraction was subsequently hydrolyzed in HCl and neutralized with NaOH. The sample was added to solution A and followed by solution B. Absorbance at 520 nm was measured and compared to absorbance values from a standard curve of glucosamine taken through the same reactions.

Fungus cell wall integrity analysis

CFW and CR (Sigma Biochemical) were utilized to stain the Paracoccidioides Pb01 and Pb18 cell wall, in order to show the effect of OenB by means of the fluorescence of the stain. Paracoccidioides was grown in liquid Fava-Netto’s medium at 36°C for 72 h with or without 0.32 M of OenB. Briefly, the cells were fixed in 100% methanol at -80°C for 20 min and then at -20°C for 20 min, and they were then washed and centrifuged. The cells were collected, stained with 100 μg/mL CFW and CR in PBS for 15 min and washed with PBS. The specimens were analyzed under a fluorescence microscope (Zeiss Axiocam MRc – Scope A1).

RDA libraries data accession

The RDA data from this study is available on the EST database (dbEST) at http://www.ncbi.nlm.nih.gov/nucest/?term= under series LIBEST_028147 Paracoccidioides oenoteinB 90 up Library, LIBEST_028146, Paracoccidioides oenoteinB 90 down Library, LIBEST_028149 Paracoccidioides oenoteinB 180 up Library and LIBEST_028148 Paracoccidioides oenoteinB 180 down Library.