Flavin Mononucleotide-Based Fluorescent Protein as an Oxygen-Independent Reporter in Candida albicans and Saccharomyces cerevisiae

Abstract

Hypoxia is encountered frequently by pathogenic and apathogenic fungi. A codon-adapted gene encoding flavin mononucleotide-based fluorescent protein (CaFbFP) was expressed in Candida albicans and Saccharomyces cerevisiae. Both species produced CaFbFP and fluoresced even during hypoxia, suggesting that oxygen-independent CaFbFP is a useful, novel tool for monitoring hypoxic gene expression in fungi.

Many microbial pathogens attain their virulence by their ability to colonize hypoxic niches in the human host. During growth within organs, tissues, biofilms, cells, or phagocytotic vacuoles, microbial pathogens may encounter no or extremely low levels of oxygen. For example, hypoxia in the human fungal pathogen Candida albicans increases glycolytic activities to permit ATP production under hypoxic growth (8), generates morphological phenotypes not occurring during normoxia (2, 6, 10), and modifies in vivo virulence (4, 11). However, further details on hypoxic adaptation in fungi regarding metabolism, cellular differentiation, and virulence are still unknown (reviewed in reference 3). Green fluorescent protein (GFP) of the jellyfish Aequorea victoria and its derivatives have been valuable tools for monitoring multiple aspects of gene expression as well as biosynthesis, localization, and function of proteins (reviewed in reference 9). However, chromophore formation of GFP and related proteins requires molecular oxygen, and therefore, their fluorescence does not relate to protein levels under oxygen depletion. Drepper et al. (1) introduced oxygen-independent flavin mononucleotide (FMN)-based fluorescent proteins (FbFP) engineered from the blue-light photoreceptors of Bacillus subtilis and Pseudomonas putida. The light oxygen voltage domains of the photoreceptors were codon optimized for expression in Escherichia coli and in the facultative anaerobic bacterium Rhodobacter capsulatus; in vivo fluorescence studies revealed that in contrast to GFP, the FbFP proteins fluoresced both in the presence and in the absence of oxygen. Here, we introduce the use of FbFP proteins as a reporter protein for fungi, including the apathogenic yeast Saccharomyces cerevisiae and the pathogen C. albicans.

The FbFP gene was resynthesized (Geneart, Regensburg, Germany), with CUG codons omitted (7) and A/T residues preferred in accordance with the low G/C content of C. albicans genes (see Fig. S1 in the supplemental material). The resultant gene, CaFbFP (C. albicans-adapted FMN-based fluorescent protein), was inserted in fungal expression vectors. For C. albicans expression, the 427-bp BglII-BamHI CaFbFP fragment was inserted into the BamHI site downstream of the ACT1 promoter in pDS1044 (CaURA3 CaLEU2 CARS, kindly supplied by D. Sanglard) or downstream of the TDH3 promoter in pIE5 (unpublished results), generating vector pIE1.1 or pIE5.1, respectively (Fig. 1A). To express a tandem CaFbFP fusion, we first generated asymmetrical AvaI sites (CTCGGG) by PCR preceding the start and stop codon positions of CaFbFP (see Fig. S1 in the supplemental material; GenBank accession numbers FJ936118 and FJ943639) and then inserted the resultant AvaI fragment into CaFbFP containing a single AvaI site preceding the stop codon. The tandem construct was used to construct expression vectors pIE1.2 and pIE5.2, respectively (Fig. 1A). Plasmids were transformed into strain CAI4 (ura3Δ::imm434/ura3Δ::imm434), with selection for uridine prototrophy (8). For S. cerevisiae expression, the 427-bp BamHI-EcoRI CaFbFP fragment was inserted downstream of the GAL1 promoter in plasmid p426-GAL1 (URA3 2μm plasmid ori) (5). The resultant plasmid, pIE3, was transformed into S. cerevisiae strain THY.AP4 (MATa ura3 leu2 lexA::lacZ::trp1 lexA::HIS3 lexA::ADE2) (5a).

FIG. 1.

Expression of FMN-binding fluorescent protein (CaFbFP) in fungi. (A) The codon usage of EcFbFP (1) was adapted to C. albicans, and the resulting CaFbFP gene was placed under the transcriptional control of the GAL1 promoter for S. cerevisiae expression or under the control of the ACT1 or THD3 promoter for C. albicans expression. A tandem CaFbFP fusion was constructed using the indicated asymmetrical AvaI sites. (B) Detection of CaFbFP in transformants of C. albicans strain CAI4 carrying plasmids pIE1.1, pIE1.2, pIE5.1, and pIE5.2 and in S. cerevisiae strain THY.AP4 containing plasmid pIE3. Crude extracts (50 μg protein) were analyzed by immunoblotting using anti-EcFbFP antiserum; extracts of transformants carrying empty vectors pDS1044 and p426 were tested as controls (c). The migration of standard proteins was as indicated (kDa). (C) Detection of CaFbFP in crude extracts (500 μg protein) of S. cerevisiae strain THY.AP4 containing plasmid pIE3 by Coomassie staining of a two-dimensional polyacrylamide gel electrophoresis gel (left) and subsequent immunoblotting using anti-EcFbFP antiserum (right). Arrows indicate CaFbFP proteins.

Fungal transformants were tested for production of the CaFbFP protein and its fluorescence. S. cerevisiae THY.AP4 carrying pIE3 was grown in SGR medium (0.67% yeast nitrogen base, 2% galactose, and 1% raffinose), and crude cell extracts were prepared by shaking with glass beads. Following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (4 to 20% acrylamide), immunoblotting was carried out using anti-EcFbFP antiserum (1). Transformants were shown to contain an antibody-reactive protein of the expected size (15.7 kDa) (Fig. 1B). Interestingly, separation on two-dimensional gels split this protein species into two closely comigrating bands, which were both identified as CaFbFP by peptide mass fingerprinting (Fig. 1C). Possibly, only one of these species contains FMN or an alternative cofactor. Production of CaFbFP was also detected in C. albicans strain CAI4 transformants grown in SD medium (0.67% yeast nitrogen base, 2% glucose), revealing antibody-reactive bands of about 17 kDa (single) and 35 kDa (tandem) in pIE1.2 and pIE1.2 transformants, respectively (Fig. 1B). Notably, in C. albicans, the tandem CaFbFP-specific signal was significantly stronger than the hardly detectable signal produced by the monomeric protein, suggesting that the duplication had increased the stability of the fluorescence reporter.

We subsequently tested the fluorescence in cell extracts of fungal transformants. High fluorescence levels were obtained for the S. cerevisiae pIE3 transformant, and the fluorescence emission spectrum revealed a peak of around 495 nm, which is the wavelength for maximal emission of EcFbFP (1) (Fig. 2A). pIE3 transformant colonies and cells fluoresced (Fig. 2B and C). C. albicans transformant colonies carrying plasmid pIE1.2 or pIE5.2 but not plasmid pIE1.1 or pIE5.1 fluoresced also, suggesting that using higher levels of the tandem than of the single CaFbFP protein led to significant fluorescence. Interestingly, this correlation did not apply to S. cerevisiae, because in this species, the tandem CaFbFP protein neither generated higher protein levels nor improved fluorescence over that obtained with the single-copy version (data not shown). We then tested if CaFbFP fulfills its promise to show fluorescence under hypoxia, i.e., under conditions that prevent GFP fluorescence. For this purpose, we grew S. cerevisiae THY.AP4[pIE3] on SGR plates in a jar generating anaerobic conditions (Oxoid AnaeroGen), alongside a transformant (YKMI) expressing GFP from the GAL1 promoter (kindly provided by J. H. Hegemann); in parallel, plates were grown in air (normoxia). The GFP transformant colonies or grown streaks showed strong fluorescence under normoxia, and this fluorescence was more intense than that of the CaFbFP-producing transformant (Fig. 2B). In contrast, after growth under anoxia, only the CaFbFP colonies or streaks showed strong fluorescence, while the GFP-producing strains were not active. We wish to point out that hypoxically grown colonies producing GFP did not reveal fluorescence even after several hours of equilibration in air, demonstrating that the lack of oxygen during growth is not simply reversible. C. albicans transformants producing tandem versions of CaFbFP produced fluorescent colonies under normoxia but also following anoxic growth (Fig. 3A). In addition, anoxically grown pIE5.2 transformants expressing CaFbFP from the TDH3 promoter showed fluorescence of single cells (Fig. 3B), which was not observed for the pIE1.2 transformants (ACT1 promoter constructs).

FIG. 2.

CaFbFP fluorescence in yeast transformants. (A) Fluorescence emission spectra of crude extracts of S. cerevisiae transformants (0.5 mg protein/ml) carrying plasmid pIE3 (CaFbFP) or control vector p426-Gal1. Fluorescence was measured with a Perkin Elmer model LS50B luminescence spectrometer (excitation wavelength, 450 nm); fluorescence intensity is expressed in arbitrary units (A.U.). (B) Fluorescence of yeast transformants synthesizing GFP or CaFbFP following normoxic or anoxic growth. S. cerevisiae strain YKMI expressing a fusion of GFP to the GAL1 promoter or strain THY.AP4 carrying a GAL1p fusion to CaFbFP (pIE3) was grown on SGR agar (2% galactose-1% raffinose) at 30°C for 48 h under normoxia or for 72 h in a jar generating anoxic conditions (Oxoid AnaeroGen AN0025A). Grown colonies and streaks of transformants were photographed directly or examined for fluorescence by using a Zeiss Axioplan 2 microscope (filter set, Zeiss FS38; excitation BP, 470/40; emission, 525/50; magnification, 100-fold). (C) Fluorescence of CaFbFP in single cells of S. cerevisiae transformants carrying pIE3. Cells were examined by phase microscopy or by fluorescence microscopy using a Zeiss Axioplan 2 microscope equipped with an HBO100 lamp (filter set, Zeiss FS38; excitation BP, 470/40; emission, 525/50; magnification, 1,000-fold). Transformants carrying empty vector p426-Gal1 were used as controls.

FIG. 3.

(A) Fluorescence of C. albicans CAI4 transformants carrying plasmid pIE1.2 or pIE5.2 synthesizing tandem CaFbFP following growth on SD agar (2% glucose) at 30°C for 24 h under normoxia or for 72 h in a jar generating anoxic conditions (Oxoid AnaeroGen AN0025A). Grown colonies were photographed directly or examined for fluorescence by using a Zeiss Axioplan 2 microscope (filter set, Zeiss FS38; excitation BP, 470/40; emission, 525/50; magnification, 100-fold). (B) Fluorescence of tandem CaFbFP in single cells of anoxically grown C. albicans CAI4 transformants carrying pIE5.2.

We conclude that CaFbFP is a suitable fluorescent reporter for aerobically and hypoxically grown fungal cells. This proof of principle may be expanded in the future in multiple directions, including improved methods of CaFbFP fluorescence detection by novel CaFbFP variants.

Nucleotide sequence accession numbers.

The codon-optimized genes were submitted to GenBank under accession numbers FJ936118 and FJ943639.