BRG1 and NRG1 form a novel feedback circuit regulating Candida albicans hypha formation and virulence

Abstract

In the opportunistic fungal pathogen Candida albicans both cellular morphology and the capacity to cause disease are regulated by the transcriptional repressor Nrg1p. One of the genes repressed by Nrg1p is BRG1, which encodes a putative GATA family transcription factor. Deletion of both copies of this gene prevents hypha formation. We discovered that BRG1 overexpression is sufficient to overcome Nrg1p-mediated repression and drive the morphogenetic shift from yeast to hyphae even in the absence of environmental stimuli. We further observed that expression of BRG1 influences the stability of the NRG1 transcript, thus controlling filamentation through a feedback loop. Analysis of this phenomenon revealed that BRG1 expression is required for the induction of an antisense NRG1 transcript. This is the first demonstration of a role for mRNA stability in regulating the key C. albicans virulence trait: the ability to form hyphae.

Figure 1. BRG1 is repressed by Nrg1p and its over-expression leads to increased invasion of solid medium

A. Quantitative real-time PCR analysis identified BRG1 as a target of Nrg1p-mediated repression. Expression of BRG1 is derepressed in the nrg1Δ strain (BCa23-3) compared to a wild-type strain (SC5314) in yeast conditions (YPD 28°C 6 h, left panel). Conversely, BRG1 expression is repressed when NRG1 is over-expressed in our tet-NRG1 strain (SSY50-B, No DOX) in hypha-inducing conditions (RPMI-1640 37°C 3 h, middle panel). Levels were normalized by ACT1 and expressed relative to the wild-type strain (left panel) or to the doxycycline-containing culture (right panel) therefore those results are 1.0. Error bars represent the standard error of the mean. B. Cells from an overnight culture were washed in sterile PBS, diluted 1:20 into fresh, pre-warmed YPD+10% FBS and incubated with shaking at 37°C for 3 h. The brg1Δ strain (ICY326) fails to form hyphae. C. A copy of BRG1 under the control of the tetO promoter was integrated at the RPS1 locus in a wild-type background (THE1-CIp10) and in our tet-NRG1 strain (SSY50-B) to construct strains ICY171 and ICY175, respectively. Cells were streaked on YPD plates with or without doxycycline (DOX) and the plates incubated at 30°C for 48 hours. The plates were photographed prior to (upper panels) and after (lower panels) surface cells were removed by gentle washing under running water. Cells which grew solely on the surface of the medium washed away while cells that had invaded the agar remained. D. Cells from an overnight culture were washed in sterile PBS, counted and spotted onto solid Lee medium (buffered to pH 4 or 7) in the presence or absence of doxycycline and incubated at 30°C or 37°C for 2 days. A wrinkled colony morphology indicates the presence of filamentous cells.

Figure 2. Over-expressing BRG1 promotes C. albicans hyphal growth and biofilm formation

A. The THE1-CIp10 control and tet-BRG1 (ICY171) strains were grown under yeast conditions (YPD at 28°C) with or without doxycycline. Over-expression of BRG1 stimulates the production of hyphae. Cell walls were visualized with calcofluor white and nuclei with DAPI. B. The tet-NRG1 (SSY50-B) and tet-NRG1+tet-BRG1 (ICY175) strains were grown in hypha inducing conditions (YPD at 37°C). BRG1 over-expression restores the ability of the tet-NRG1 strain to filament in the absence of DOX. Cell walls were visualized with calcofluor white and nuclei with DAPI. C. To examine biofilm formation, cells from the indicated strains were incubated in SD medium at 30°C for 24h in 96-well microtitre plates. The bottom of the wells was photographed after non-adherent cells were washed away. A dense mass of cells covering the bottom of the well is visible when biofilms formed.

Figure 3. BRG1 over-expression influences C. albicans virulence

A. BRG1 Over-expression attenuates virulence in a wild-type strain. Approximately 4.6×10⁵ (dark lines) or 2.4×10⁵ (light lines) yeast cells of C. albicans strain tet-BRG1 (ICY171) were injected into mice. Survival differences between Plus DOX and No DOX groups at both high and low doses were significantly different (p=0.0016 and p=0.0018, respectively). Examination of kidneys retrieved from mice that succumbed to the infection revealed extensive C. albicans growth (GMS staining, upper panels), and infiltration of numerous neutrophils (H&E staining, lower panels). Interestingly, examination of the kidney retrieved from a mouse that survived the infection (No DOX) revealed a subcapsular scar (H&E staining, far right panel) indicative of earlier tissue damage. B. For timed sacrifice experiments, approximately 1.7×10⁶ (6 h sacrifice) or 3.9×10⁵ (3 day sacrifice) yeast cells of the tet-BRG1 strain (ICY171) were injected into mice. The indicated tissues were homogenized and plated onto Sabouraud agar to determine the fungal burdens (group median indicated by a bar). There was a significant reduction (p=0.0212) in kidney fungal burdens upon BRG1 over-expression (6 h sacrifice). Examination of kidneys retrieved after 3 days from the BRG1 over-expression group (No DOX) revealed kidney necrosis and significant vascular thrombosis (H&E staining, upper panels) associated with invasive C. albicans growth (GMS staining, lower panel). C. BRG1 Over-expression results in attenuated virulence in the tet-NRG1 background. Approximately 6.0×10⁵ (dark lines) or 2.4×10⁵ (light lines) yeast cells of the tet-NRG1+tet-BRG1 strain (ICY175) were injected into mice. The dashed line represents the survival data produced by the parental tet-NRG1 strain (SSY50-B) when NRG1 levels are elevated (Saville et al., 2003). Survival differences between the Plus DOX and No DOX groups were not statistically significant at the high dose (p=0.1573) but were at the low dose (p=0.0031). Examination of kidneys retrieved from infected mice revealed extensive lesions containing hyphal cells irrespective of the presence or absence of doxycycline (H&E staining, upper panels; GMS staining, lower panels). This contrasts starkly with the tet-NRG1 strain that grows only as yeast cells in the kidney in the absence of DOX (Saville et al., 2003).

Figure 3. BRG1 over-expression influences C. albicans virulence

A. BRG1 Over-expression attenuates virulence in a wild-type strain. Approximately 4.6×10⁵ (dark lines) or 2.4×10⁵ (light lines) yeast cells of C. albicans strain tet-BRG1 (ICY171) were injected into mice. Survival differences between Plus DOX and No DOX groups at both high and low doses were significantly different (p=0.0016 and p=0.0018, respectively). Examination of kidneys retrieved from mice that succumbed to the infection revealed extensive C. albicans growth (GMS staining, upper panels), and infiltration of numerous neutrophils (H&E staining, lower panels). Interestingly, examination of the kidney retrieved from a mouse that survived the infection (No DOX) revealed a subcapsular scar (H&E staining, far right panel) indicative of earlier tissue damage. B. For timed sacrifice experiments, approximately 1.7×10⁶ (6 h sacrifice) or 3.9×10⁵ (3 day sacrifice) yeast cells of the tet-BRG1 strain (ICY171) were injected into mice. The indicated tissues were homogenized and plated onto Sabouraud agar to determine the fungal burdens (group median indicated by a bar). There was a significant reduction (p=0.0212) in kidney fungal burdens upon BRG1 over-expression (6 h sacrifice). Examination of kidneys retrieved after 3 days from the BRG1 over-expression group (No DOX) revealed kidney necrosis and significant vascular thrombosis (H&E staining, upper panels) associated with invasive C. albicans growth (GMS staining, lower panel). C. BRG1 Over-expression results in attenuated virulence in the tet-NRG1 background. Approximately 6.0×10⁵ (dark lines) or 2.4×10⁵ (light lines) yeast cells of the tet-NRG1+tet-BRG1 strain (ICY175) were injected into mice. The dashed line represents the survival data produced by the parental tet-NRG1 strain (SSY50-B) when NRG1 levels are elevated (Saville et al., 2003). Survival differences between the Plus DOX and No DOX groups were not statistically significant at the high dose (p=0.1573) but were at the low dose (p=0.0031). Examination of kidneys retrieved from infected mice revealed extensive lesions containing hyphal cells irrespective of the presence or absence of doxycycline (H&E staining, upper panels; GMS staining, lower panels). This contrasts starkly with the tet-NRG1 strain that grows only as yeast cells in the kidney in the absence of DOX (Saville et al., 2003).

Figure 4. Analysis of BRG1 and NRG1 expression during hyphal induction

BRG1 and NRG1 transcription was examined during hyphal induction (YPD+10% FBS, 37°C) by northern blot analyses. RNA was isolated, transferred to a Nytran membrane and probed with DNA fragments specific to the BRG1 or NRG1 coding sequence. Ethidium bromide-stained rRNA bands are shown as a loading control. Numbers represent time post-induction. A. In a wild-type strain (SC5314), BRG1 transcript is undetectable in yeast cells (Time 0), is abundant 30 minutes after induction and subsequently declines. Expression of NRG1 is high in yeast cells (Time 0), falls during the first hour of hyphal induction and gradually returns to pre-induction levels. B. In the brg1Δ strain (ICY326), no BRG1 transcript is detectable and during hyphal induction, NRG1 levels remain high. C. In the tet-BRG1 strain (ICY171), when BRG1 expression is induced from a tet-regulated allele, BRG1 levels rise steadily over time, whilst NRG1 levels fall. When BRG1 is not over-expressed, through the addition of DOX to the medium, NRG1 levels remain stable (YPD 28°C). D. When BRG1 is over-expressed in yeast conditions (YPD 28°C) in the tet-BRG1 strains ICY171 and ICY175, NRG1 levels fall. Note that in ICY175, transcription from both the endogenous and tet-regulated (arrow) NRG1 alleles is affected. (-) No DOX, (+) Plus DOX. E. Quantitative real-time PCR analysis confirmed that NRG1 mRNA destabilization depends on BRG1. The tet-NRG1 (SSY50-B) and brg1Δ (ICY277) strains were grown overnight (YPD 28°C No DOX) then diluted intro fresh medium and grown under hypha inducing conditions (YPD+10% FBS 37°C Plus DOX). In the tet-NRG1 strain, NRG1 levels fall substantially by 30 minutes after induction whilst in the brg1Δ strain NRG1 levels remain stable. Levels were normalized by ACT1 and expressed relative to the respective time zero samples (therefore those results are 1.0). Error bars represent the standard error of the mean. F. During hyphal induction (YPD+10% FBS 37°C 2 h) NRG1 transcript is most abundant in the brg1Δ (ICY326) strain. Levels are returned to those observed in the BRG1/brg1Δ heterozygote (ICY318) by reintegrating a copy of BRG1 at its native locus (ICY330).

Figure 5. Examination of NRG1 transcript stability

A. One copy of NRG1 was replaced with lacZ under the control of the tetO promoter (ICY281). In this strain reporter activity, as indicated by the appearance of the blue precipitate, is modulated through the addition of doxycycline to the medium. In the tet-lacZ tet-BRG1 strain (ICY307) simultaneous over-expression of an ectopic copy of BRG1 under the control of the tetO promoter results in a wrinkled morphology as in the tet-BRG1 strain (ICY171), but does not affect expression of the lacZ reporter inserted at NRG1. B. lacZ expression was measured using quantitative real-time PCR. In agreement with the phenotypic results, over-expression of BRG1 (ICY307) reduced NRG1 expression but not lacZ expression. Levels were normalized by ACT1 and expressed relative to the tet-lacZ strain (ICY281) time zero samples (therefore those results are 1.0). Error bars represent the standard error of the mean. C. Constitutive expression of NRG1 from an ectopic ACT1 promoter-regulated allele (strain ICY293) is sufficient to block hyphal induction in YPD at 37°C. Addition of a tet-regulated BRG1 allele, to form strain ICY295, overcame this Nrg1p-mediated repression. D. Northern blot analysis. The abundance of both the endogenous and larger ectopic NRG1 (arrow) transcripts depends on BRG1 over-expression. Strains were grown in YPD at 37°C for 3 h. Ethidium bromide-stained rRNA bands are shown as a loading control. (-) No DOX. (+) Plus DOX

Figure 6. RT-PCR and quantitative real-time PCR analysis of antisense NRG1 production

A. RNA was isolated from the wild-type (SC5314) and the brg1Δ (ICY326) strains grown under hypha-inducing conditions (YPD+10% FBS, 37°C) for 30 minutes. RNA was treated with DNaseI and used as a template for reverse transcription with either strand-specific primers or oligo_dT, as indicated above each lane. Different primer combinations (indicated to the right of each panel) were used in PCR reactions with these single stranded cDNA templates. Aliquots of each PCR reaction were electrophoresed on an agarose gel and visualized with ethidium bromide. The housekeeping gene EFB1 serves as a control for the presence of genomic DNA or total cDNA. A no reverse transcriptase sample serves as a control for the RT-PCR. The diagram at the bottom indicates the relative positions of the primers used in this experiment (not to scale). F=NRG1_FOR, S2=NRG1_S2, S=NRG1-S S4=NRG_S4, A=NRG1-A, A2=NRG1_A2, A3=NRG1_A3, A4=NRG1_A4, R=NRG1_REV, dT=oligo_dT, G=gDNA, N=no reverse transcriptase control. B. The wild-type (dark bars) and the brg1Δ (light bars) strains were grown under hypha-inducing conditions (YPD+10% FBS, 37°C). Samples were removed prior to, as well as 20 and 30 minutes after induction and cDNA synthesized with oligo_dT for ACT1 and BRG1 real-time PCR reactions, and NRG1_S2 and NRG1_REV for the antisense and sense reactions, respectively. Levels were normalized by ACT1 and expressed relative to the yeast control culture (wild-type time 0 sample, therefore its results are 1.0). Error bars represent the standard error of the mean. BRG1 expression is rapidly upregulated in the early stages of hyphal induction (upper panel). The NRG1 sense transcript falls in the wild-type, but remains constant in the brg1Δ strain. Conversely, the antisense NRG1 transcript rises concurrently with BRG1 in the wild-type (SC5314) but remains similar to time 0 levels in the brg1Δ strain (ICY326). C. BRG1 expression does more than stimulate antisense NRG1 production. In yeast conditions, the nrg1Δ strain (BCa23-3) is constitutively pseudohyphal. Strains over-expressing BRG1 (ICY171 and ICY175), however, form hyphae under the same conditions. Cell walls were visualized with calcofluor white and nuclei with DAPI.

Figure 7. A model for Brg1p function in C. albicans hyphal induction

BRG1 is an important component of hyphal induction in C. albicans. A. Under yeast conditions, BRG1 is repressed in an Nrg1p-mediated fashion. During hyphal induction, BRG1 expression is rapidly and substantially increased. This leads to Brg1p-dependent induction of an antisense NRG1 transcript (wavy line) and subsequent destruction of the NRG1 sense transcript (thin arrow) by an, as yet unknown, argonaute-independent mechanism (dashed lines). B. Brg1p acts in a feedback mechanism with NRG1 to regulate hypha-specific gene (HSGs) expression. BRG1 expression is upregulated by appropriate environmental stimuli, resulting in decreased NRG1 transcript levels and stimulation of hypha-specific gene expression. As BRG1 levels fall after the initial phase of induction, NRG1 levels are able to rise and the system is reset.