A novel downstream regulatory element cooperates with the silencing machinery to repress EPA1 expression in Candida glabrata

Abstract

Candida glabrata, an opportunistic fungal pathogen, adheres to mammalian epithelial cells; adherence is mediated primarily by the Epa1 adhesin. EPA1 is a member of a large gene family of ≈ 23 paralogues, which encode putative adhesins. In this study, we address how EPA1 transcription is regulated. Our data show that EPA1 expression is subject to two distinct negative regulatory mechanisms. EPA1 transcription is repressed by subtelomeric silencing: the Sir complex (Sir2-Sir4), Rap1, Rif1, yKu70, and yKu80 are required for full repression. Activation of EPA1 occurs immediately after dilution of stationary phase (SP) cells into fresh media; however, transcription is rapidly repressed again, limiting expression to lag phase, just as the cells exit stationary phase. This repression following lag phase requires a cis-acting regulatory negative element (NE) located in the EPA1 3'-intergenic region and is independent of telomere proximity. Bioinformatic analysis shows that there are 10 copies of the NE-like sequence in the C. glabrata genome associated with other EPA genes as well as non-EPA genes.

Figure 1

Regulation of the expression of EPA1. (A) Schematic representation of the EPA1 genomic locus. (B) EPA1 transcript levels measured by S1 nuclease protection. BG14 (WT) cells were grown for 48 hr and 30° in YPD media. Cells were diluted into fresh media and samples were taken at different time points (see Materials and Methods). Time 0 is undiluted stationary phase (SP) cells. (C) Adherence of C. glabrata cells to HeLa cells. C. glabrata wild-type strain BG14 (WT) and strain BG64 (epa1Δ) were grown for 48 hr at 30° in YPD media. Cells were diluted into fresh media and samples were taken at different time points. Cells were adjusted to OD₆₀₀ of 1.0 in HBSS supplemented with 5 mM CaCl₂. Cell suspensions were diluted serially in sterile water and appropriate dilutions were made and plated on YPD plates to determine input colony forming units (CFU) (see Materials and Methods). Each experiment was made in triplicate. (D) EPA1 promoter activity measured by FACS. Strains BG198 (P_EPA1::GFP) and BG201 (P_EPA1::GFP::pYIplac211) were grown for 48 hr at 30° in YPD media. Cells were diluted into fresh media and samples were taken every 2 hr. Yeast cells were washed and resuspended in 1 ml PBS and fluorescence was assessed by FACS analysis using a BD FACSCalibur flow cytometer (see Materials and Methods). EPA1 promoter GFP fusion is at the chromosomal EPA1 locus. GFP was used as reporter of the activity of the EPA1 promoter. (E) EPA1 promoter activity measured by FACS as in D, but cells were grown in SC −Ura media and all constructs are borne in plasmids. Strain BG14 (WT) carrying plasmids pAP353 (promoterless control, GFP::3′UTR_HIS3), pAP354 (P_EPA1::GFP::3′UTR_HIS3), pAP385 (P_EPA1::GFP::3′UTR_EPA1NE_(3.1Kb)), pSP38 (P_EPA1::GFP::3′UTR_HIS3:: NE_(200bp)), and pSP105 (NE_200bp::P_EPA1::GFP::3′UTR_EPA1).

Figure 2

Mapping of the negative element (NE). The activity of the NE was assayed by measuring EPA1 promoter activity by FACS analysis of the GFP reporter fused to the EPA1 promoter. BG14 (WT) strain carrying a collection of plasmids containing serially deleted fragments (5′ to 3′ and 3′ to 5′) of the intergenic region between EPA1 and EPA2, were grown in SC −Ura media and assayed as described in Figure 1D legend. Fold of repression is the maximal expression of EPA1 at 2 hr divided by the expression at 8 hr. Line 1 is pSP21, line 2 is pSP26, line 3 is pSP20, line 4 is pSP7, line 5 is pSP19, line 6 is pAP407, line 7 is pSP31, line 8 is pSP32, line 9 is pSP33, line 10 is pSP34, line 11 is pSP35, and line 12 is pAP385 (Table 2). Experiments were done in triplicate and SDs are shown.

Figure 3

NE effect on adherence. C. glabrata wild-type strain BG14 (WT) (hatched bar) and strain BG646 containing a Tn7 insertion between EPA1 and the NE (open bar) were grown for 48 hr at 30° in YPD media. Cells were diluted to OD₆₀₀ of 1.0 in HBSS supplemented with 5 mM CaCl₂. The adherence assays were done as described in the Figure 1C legend. Adherent cells were recovered from the epithelial cells and were plated for viable accounts. See Materials and Methods.

Figure 4

Silencing effect on the expression of EPA1. Schematic representation of the reporter strains. (A) EPA1 was replaced by the URA3 gene and the NE was replaced by the bacterial cat gene and recombined in the chromosome. URA3 reports the activity of the EPA1 promoter. (B) The parental strains NE⁺ and neΔ and the strains carrying null mutations in SIR2–SIR4 (sir2Δ–sir4Δ), (C) HDF1 (yKu70), HDF2 (yKu80), RIF1, and rap1-21 (hdf1Δ, hdf2Δ, and rap1-21), and (D) HST1 and HST2 (hst1Δ and hst2Δ) were grown for 48 hr in YPD. Strains were diluted to OD_600nm 0.5 with distilled water and 10-fold serial dilutions were spotted onto YPD, SC −Ura and SC +5-FOA plates. Plates were incubated at 30°. Ura⁺ cells die on SC +5-FOA plates. Only cells with the URA3 gene transcriptionally repressed can grow on SC +5-FOA. See Materials and Methods.

Figure 5

Blast analysis of the NE in the C. glabrata genome. The last 60 bp of the NE is associated with other EPAs and non-EPA genes. EPA1, EPA2, EPA6, EPA7, AWP2, and CAGLOH10626 (15) encode cell wall proteins, all are subtelomeric and the NE associated is localized at their 3′ ends, except for AWP2 and EPA22, which is localized between these two divergently transcribed genes. The NE associated to non-EPA genes is located at the 5′ regions (promoters) of these divergently expressed genes. HXT6 (high-affinity glucose transporter of the major facilitator superfamily), MDH2 (cytoplasmic malate dehydrogenase), YPS1 (GPI-anchored aspartyl protease), PSY2 (putative subunit of an evolutionarily conserved protein phosphatase complex containing the catalytic subunit Pph3p and the regulatory subunit Psy4p), PFY1 (profilin, binds actin involved in cytoskeleton organization), ACS1 (acetyl-coA synthetase isoform), FCY2 (purine-cytosine permease), and CAGLOA2255, CAGL0M07747, and CAGL0M07766 are of unknown function (Saccharomyces Genome Database http://www.yeastgenome.org/ and C. glabrata Genome Database http://www.genolevures.org/cagl.html). Arrows indicate direction of transcription; numbers show the distance in kilobases between the negative element (NE) and the genes or the telomere (Tel); and Chr-(letter) denotes chromosome notation.

Figure 6

Model of EPA1 regulation. In stationary phase (SP), EPA1 is not expressed and cells are nonadherent. Upon dilution into fresh media, (1) a log-phase specific transcriptional activator (LP-Ac) induces expression of EPA1 and cells become adherent. This transcriptional activation is counteracted in log phase (LP) by the concerted action of silencing and the NE (2). The Sir complex silences the expression of EPA1 and yKu70/yKu80 repress EPA1 expression through the NE. These two regulatory mechanisms assure that EPA1 is not expressed even in the presence of the LP-Ac, which is active throughout LP.