CaGdt1 plays a compensatory role for the calcium pump CaPmr1 in the regulation of calcium signaling and cell wall integrity signaling in Candida albicans

Background Saccharomyces cerevisiae ScGdt1 and mammalian TMEM165 are two members of the UPF0016 membrane protein family that is likely to form a new group of Ca2+/H+ antiporter and/or a Mn2+ transporter in the Golgi apparatus. We have previously shown that Candida albicans CaGDT1 is a functional ortholog of ScGDT1 in the response of S. cerevisiae to calcium stress. However, how CaGdt1 together with the Golgi calcium pump CaPmr1 regulate calcium homeostasis and cell wall integrity in this fungal pathogen remains unknown. Methods Chemical sensitivity was tested by dilution assay. Cell survival was examined by measuring colony-forming units and staining with Annexin V-FITC and propidium iodide. Calcium signaling was examined by expression of downstream target gene CaUTR2, while cell wall integrity signaling was revealed by detection of phosphorylated Mkc1 and Cek1. Subcellular localization of CaGdt1 was examined through direct and indirect immunofluorescent approaches. Transcriptomic analysis was carried out with RNA sequencing. Results This study shows that Candida albicans CaGDT1 is also a functional ortholog of ScGDT1 in the response of S. cerevisiae to cell wall stress. CaGdt1 is localized in the Golgi apparatus but at distinct sites from CaPmr1 in C. albicans. Loss of CaGDT1 increases the sensitivity of cell lacking CaPMR1 to cell wall and ER stresses. Deletion of CaGDT1 and/or CaPMR1 increases calcium uptake and activates the calcium/calcineurin signaling. Transcriptomic profiling reveals that core functions shared by CaGdt1 and CaPmr1 are involved in the regulation of cellular transport of metal ions and amino acids. However, CaGdt1 has distinct functions from CaPmr1. Chitin synthase gene CHS2 is up regulated in all three mutants, while CHS3 is only up regulated in the pmr1/pmr1 and the gdt1/gdt1 pmr1/pmr1 mutants. Five genes (DIE2, STT3, OST3, PMT1 and PMT4) of glycosylation pathway and one gene (SWI4) of the cell wall integrity (CWI) pathway are upregulated due to deletion of CaGDT1 and/or CaPMR1. Consistently, deletion of either CaPMR1 or CaGDT1 activates the CaCek1-mediated CWI signaling in a cell wall stress-independent fashion. Calcineurin function is required for the integrity of the cell wall and vacuolar compartments of cells lacking both GDT1 and CaPMR1. Conclusions CaPmr1 is the major player in the regulation of calcium homeostasis and cell wall stress, while CaGdt1 plays a compensatory role for CaPmr1 in the Golgi compartment in C. albicans. Electronic supplementary material The online version of this article (10.1186/s12964-018-0246-x) contains supplementary material, which is available to authorized users.


Plain English summary
Congenital disorders of glycosylation (CDG) are inherited diseases with most of their genetic defects affecting the glycosylation process. The human ortholog for yeast ScGdt1, TMEM165, is one of these CDG-associated Golgi proteins and might function as a Ca 2+ /H + antiporter and/or a Mn 2+ transporter in the Golgi. This study provides evidence that the functional ortholog of ScGdt1, Candida albicans CaGdt1 plays a compensatory role for CaPmr1, the Ca 2+ /Mn 2+ ATPase that is required for Ca 2+ and Mn 2+ transport into the Golgi and involved in Ca 2+ dependent protein sorting processing. CaGdt1 and CaPmr1 work together at distinct sites of the Golgi apparatus to regulate the response of this human fungal pathogen to cell wall and ER stresses. Calcineurin function is required for the survival of C. albicans cells lacking both CaGdt1 and CaPmr1. These findings would contribute to our understanding of molecular mechanisms regulating TMEM165-associated CDG.

Background
Calcium ions regulate growth and programmed cell death as well as muscle contraction in the heart and taste in the mouth [1]. Calcium/calcineurin signaling pathway is highly conserved in eukaryotic cells. Functional counterparts of yeast calcium channels, pumps and exchangers exist and function in similar fashions in mammalian cells [2]. In Saccharomyces cerevisiae, calcium homeostasis is regulated through calcium transporters and sequestrates in the plasma and organelle membranes [2][3][4]. The plasma membrane calcium channel allows Ca 2+ influx in response to endoplasmic reticulum (ER) stress and pheromones, while the vacuolar Yvc1 releases Ca 2+ into the cytosol in response to hypotonic shock [4,5]. Excess calcium ions are transported into the vacuole for storage through the vacuolar calcium pump Pmc1 and the Ca 2+ /H + exchanger Vcx1; or into the ER/Golgi secretory pathway via the calcium pump Pmr1 and the Ca 2+ /H + exchanger Gdt1 [3,4,6]. Expression of PMR1 is positively controlled by the calcium/calcineurin signaling pathway and negatively controlled by the Rim101/Nrg1 pathway in S. cerevisiae [7]. Functional counterparts of Ca 2+ transporters and channels have been characterized in Candida albicans [8][9][10][11]. Rch1 is a novel negative regulator of calcium uptake in the plasma membrane of S. cerevisiae and C. albicans [12][13][14][15].
The cell wall of C. albicans is important in the interaction with its host during infection [16,22]. Deletion of CaPMR1 causes C. albicans cells to be hypersensitive to cell wall stress and to constitutively activate the Mkc1-mediated CWI signaling, which leads to a defect in glycosylation of cell wall proteins and thereby a weakened cell wall and reduced virulence [10]. C. albicans cells lacking MKC1 or CEK1 also show a reduced virulence in the mouse model of systemic infection [23,24]. Therefore, properly regulated CWI signaling is required for the virulence of C. albicans.
S. cerevisiae ScGdt1 and its human transmembrane protein 165 (TMEM165) belong to a well-conserved family of membrane proteins named UPF0016 (Uncharacterized Protein Family0016; Pfam PF01169), which exist in 919 bacterial species and 409 eukaryotic species [25]. Mutations of TMEM165 are linked to a subtype of inborn metabolic diseases affecting the glycosylation pathway, and TMEM165 is a functional homolog of the Golgi-localized ScGdt1 [7]. ScGdt1 may act as a Ca 2+ /H + antiporter and plays a major role in the calcium response induced by osmotic shock in the absence of ScPmr1 [26,27]. Here, we have characterized CaGdt1, the C. albicans homolog of S. cerevisiae ScGdt1, in the response to calcium and cell wall stress. CaGdt1 and CaPmr1 localize to distinct sites in the Golgi apartment. Transcriptomic profiling reveals overlapping and distinctive functions between CaGdt1 and CaPmr1. In addition, we demonstrate that CaGdt1 is involved in the Cek1-mediated, but not the Mkc1-mediated, cell signaling.

Methods
Strains and reagents C. albicans and S. cerevisiae strains, plasmids and primers are listed in Tables 1 and 2, and Additional file 1:  Table S1, respectively. All strains were routinely grown at 30°C in YPD medium or SD medium (0.67% yeast nitrogen base without amino acids, 2% glucose, and auxotrophic amino acids as needed). Chemical reagents were obtained from Sigma.

Construction of C. albicans mutants
We constructed the double-gene deletion mutant by replacing two CaGDT1 alleles in the pmr1/pmr1 mutant in the CAI4 background [10], with the SAT1 flipper cassette [28] and the hisG-URA3-hisG cassette [29], respectively, as described previously [30]. Genotypes of the resultant mutants were confirmed by PCR (data not shown).

Growth test
Strains were cultured overnight in liquid YPD or SD-URA medium at 30°C. Overnight cultures were serially diluted by 10-fold, and spotted onto appropriate plates. Phenotypes were recorded after plates were incubated for 2-3 days at 30°C. 45 Ca 2+ uptake assays Cellular accumulation of Ca 2+ was measured basically as described [32]. Cells were inoculated in YPD medium containing 5 μCi ml − 1 45 CaCl 2 with a specific activity of 9.6 × 10 6 cpm ml − 1 (PerkinElmer), and grown to log phase at 30°C in 96-well filtration plates. Cultures were harvested by filtration, washed four times with ice-cold buffer (5 mM Na-HEPES, pH 6.5, 10 mM CaCl 2 , 150 mM NaCl), and dried at room temperature. The

Assays of β-galactosidase activity
The CaUTR2p-lacZ reporter in the pGP8 plasmid [13] was integrated into the CaUTR2 locus in the wild type CAI4, the gdt1/gdt1 mutant, the pmr1/pmr1 mutant and the gdt1/gdt1pmr1/pmr1 mutant. Three independent transformants of each strain were assayed for β-galactosidase activity as described previously [13].

Analysis of cell survival and cell integrity
Test strains were grown to log phase before the culture was divided into two parts, which were treated with 25 μg ml − 1 CsA and ethanol (as control), respectively. Two parts were further incubated with shaking for 1 h, 2 h, 4 h and 8 h, respectively, before samples were taken for measuring colony-forming units (clonogenic survival) as well as co-staining with 20 μg ml − 1 Annexin V-FITC and 50 μg ml − 1 propidium iodide (CLONETECH Laboratories, USA). Vacuoles in these cells were labelled in vivo with N-(3-triethylammoniumpropyl)-4-(6 [4-(diethylamino) phenyl) hexatrienyl] pyridinium dibromide (FM4-64; Molecular Probes, USA) as described [33].

Protein extract and western blot analysis
Cells were grown to mid-exponential phase, and treated with 50 μg ml − 1 Calcofluor white for 2 h before they were collected for protein extraction and Western blot analysis as described [7]. The anti-phospho-p44/42 MAPK (Thr 202 /Tyr 204 ) antibody (Cell signaling, USA) was used to simultaneously detect phosphorylated forms of both CaMkc1 and CaCek1 [23,24]. Anti-HA antibody (Abmart, Shanghai) was used to detect CaMkc1-HA and CaCek1-HA proteins [7].

Co-localization of GDT1-GFP and PMR1-HA proteins
Cells of the WJCA111 strain was grown to log-phase in SD-URA-HIS medium before they were treated with 100 mM CaCl 2 or 1 mM EGTA for 2 and 6 h, respectively. Cells were treated with 3.7% formaldehyde for 10 min and incubated in fixation buffer (0.1 M KH2PO4,0.5 M MgCl2, 3.7% formaldehyde, pH 6.5) for 20 min. Cells were then resuspended in 1 ml buffer pFA-HA-URA3 HA-epitope integration vector with CaURA3 marker [34] pFA-HA-HIS1 HA-epitope integration vector with CdHIS1 marker [34] pSFS2 the SAT1 flipper cassette [28] pKC1 Two flanking regions of CaGDT1 at either side of the hisG-URA3-hisG cassette in p5921 This study pGP8 Plasmid containing the 1-kb promoter of UTR2 fused to the lacZ reporter [13] (0.1 M KH 2 P 4 , 1.2 M sorbitol, pH 6.5) supplemented with 4 μl of beta-mercaptoethanol and 4 μl of 5 mg ml − 1 Zymolyase 20 T (Seikagaku Biobusiness), and incubated at 37°C for 20 min. Spheroplasts were washed twice with 1 ml of PBS buffer, and suspended in 100 μl of PBS supplemented with 0.05% Tween 20 (PBST). BSA of 400 μg ml − 1 was added to the spheroplast mixture, and incubated for 20 min before mouse anti-HA antibodies was added at a dilution of 1:500. The mixture was incubated for 2 h, and washed twice with PBST before goat anti-mouse IgG conjugated to Alexa Fluor 555 (Invitrogen, USA) was added. The mixture was incubated in dark for 45 min before spheroplasts were collected, washed three times, and visualized under a Nikon 80i microscope equipped with DS-U2 CCD. The images were acquired at 1000 × and processed using the NIS -ELEMENTS F3.0 software.

Co-localization of GDT1-GFP protein and Golgi-tracker red dye
For Golgi apparatus labeling, the pmr1/pmr1 GDT1 -GFP/gdt1 cells were grown to log-phase in SD-URA, harvested, washed twice with PBS, and suspended in 200 μl PBS supplemented with 4 μl beta-mercaptoethanol and 15 U lyticase enzyme (Sigma). Cells was incubated at 37°C for 40 min to partially digest the cell wall, and were washed once with PBS and suspended in 200 μl PBS before they were mixed with 2 μl Golgi-Tracker Red dye (33.3 mg ml − 1 ) (Beyotime Institute of Biochemistry, China) and incubated at 4°C for 30 min. Cells were then washed twice and incubated in 1 ml SD-URA at 30°C for 30 min, before they were visualized by the Nikon 80i fluorescent microscope.

RNA extraction and qRT-PCR assays
Total RNA was extracted as described previously (12). Contaminated genomic DNA in the reverse transcription products was examined by PCR of 35 cycles using a pair of primers CaACT1-g-UP and CaACT1-g/m-DOWN flanking the intron of the C. albicans ACT1 gene. Real-time qPCR was performed in the CFX96™ Real-Time System I (BIO-RAD) using SYBR Premix EX Taq™ II (TAKARA Biotechnology, China). All assays were performed in duplicate. For quantification, the abundance of each gene was determined relative to the transcript of CaACT1.

Transcript profiling and data analysis
Total RNA samples were extracted from the wild type CAI4, the gdt1/gdt1, the pmr1/pmr1, and the gdt1/gdt1 pmr1/pmr1 mutants grown to log phase in SD-URA at 30°C. RNA integrity was evaluated using an Agilent 2100 Bioanalyzer (Agilent Technologies, USA). RNA-seq libraries were constructed using Illumina's TruSeq RNA Sample Preparation Kit (Illumina Inc., USA). RNA sequencing, data analysis and sequence assembly were performed by Beijing BioMarker Technologies (Beijing, China). Preparation of the paired-end libraries and sequencing were performed following standard Illumina methods and protocols. The mRNA-seq library was sequenced on the Illumina Hiseq 2500 and Illumina Genome Analyzer system.
To obtain high-quality clean read data for de novo assembly, raw reads from mRNA sequencing were filtered by discarding the reads with adaptor contamination, masking low-quality reads with ambiguous 'N' bases and removing reads in which more than 10% bases had a Q-value < 20. Clean reads were assembled into full-length transcriptome from RNA-Seq data with the reference genome (http://www.candidagenome.org/).

Digital analysis of gene expression
Gene expression levels were measured as numbers of reads, and normalized with RPKM in RNA-Seq analyses as described [36]. Differentially expressed genes were identified in pair-wise comparison with the IDEG6 software [37], and all statistical test results were corrected for multiple testing with the Benjamini-Hochberg false discovery rate (FDR<0.01). Gene transcripts were considered to be significantly differentially expressed if the adjusted P value was <0.001 and there was at least a two-fold change (>1 or <-1 in log 2 ratio value) in PKM between two libraries.

Semi-quantitation of gene expression levels by RT-PCR
To confirm the RNA-sequencing data, we carried out semi-quantitative RT-PCR with reverse transcription products that were treated to remove their contaminated genomic DNA. RT-PCR was carried out as described previously [12]. PCR products were separated on agarose gel and stained with ethidium bromide. DNA contents in all lanes of the gel were quantified and analyzed using the Bio-Rad Gel DocTM XR+ System and the Image Lab software (Version 4.0.1). Relative levels of target gene transcripts in each strain were calculated using CaPGK1 as the internal control.

Statistical analysis
Significant differences were analysed by GraphPad Prism version 4.00. P values of < 0.05 were considered to be significant.

Results
CaGDT1 complements the function of ScGDT1 in cell wall stress As reported previously [30], expression of CaGDT1 in the S. cerevisiae vector, pHAC181, fully suppressed the sensitivity of the Scgdt1 single-gene deletion mutant to 0.6 M and 0.7 M CaCl 2 as well as the sensitivity of the Scgdt1/Scpmr1 double-gene deletion mutant to 0.4 M CaCl 2 (Additional file 1: Figure S1A). In addition, expression of CaGDT1 reverted the sensitivity level of the Scgdt1/Scpmr1 double-gene deletion mutant to that of the Scpmr1 single-gene deletion mutant under higher concentrations of CaCl 2 (0.6 M and 0.7 M) (Additional file 1: Figure S1A). Therefore, CaGDT1 is a functional homolog of ScGDT1 in calcium stress [30]. Although S. cerevisiae cells lacking ScGDT1 were not sensitive to 5 mM EGTA, 0.4 M CaCl 2 , 100 μg ml − 1 Congo red (CR) and 50 μg ml − 1 Calcofluor white (CFW) [7], we found that deletion of ScGDT1 increased the sensitivity of cells lacking ScPMR1 to all these reagents (Additional file 1: Figure S1A). Expression of CaGDT1 reverted the sensitivity level of the Scgdt1/Scpmr1 double-gene deletion mutant to that of the Scpmr1 single-gene deletion mutant under these cell-wall stress conditions (Additional file 1: Figure S1A). Taken together, these data demonstrate that CaGdt1 complements the function of ScGdt1 in the response of S. cerevisiae cells to cell-wall stress.
CaGDT1 shows genetic interactions with CaPMR1 in the response to cell wall and ER stress Since mutations of human TMEM165, the CaGdt1 homolog, affect the glycosylation pathway [25,38], we tested cells lacking CaGDT1 against cell wall and ER stresses. Surprisingly, C. albicans cells lacking CaGDT1 did not show significant alteration in their sensitivity to CR, CFW, caffeine and tunicamycin (Fig. 1a).
Unlike ScGDT1, C. albicans cells lacking CaGDT1 were not sensitive to 0.7 M CaCl 2 ( Fig. 1a; Additional file 1: Figure S1A; [25]). However, ScGDT1 genetically interacts with ScPMR1 in the calcium sensitivity of S. cerevisiae cells [25]. We next investigated possible interactions between CaGDT1 and CaPMR1 in cell wall stress. As reported previously [10], cells lacking CaPMR1 were hypersensitive to CR, CFW, caffeine and tunicamycin ( Fig. 1a and b). Further deletion of CaGDT1 increased the sensitivity of cells lacking CaPMR1 to CR, CFW, caffeine and tunicamycin, which could be reversed by the introduction of the CaGDT1 gene ( Fig. 1a and b). These results demonstrate that CaGDT1 has an additive effect on CaPMR1 in the response of C. albicans cells to cell wall and ER stresses.
Calcineurin function is required for C. albicans cells in response to cell wall stress Further deletion of CaGDT1 led cells lacking CaPMR1 hypersensitive to cyclosporine A (CsA) and FK506, specific inhibitors of calcineurin, in YPD medium with or without cell wall stress ( Fig. 1b; Additional file 1: Figure  S1B). This indicates that normal growth of C. albicans cells lacking both CaGDT1 and CaPMR1 requires the calcineurin function.
The homozygous mutant for CaGDT1 became hypersensitive to CR and CFW in the presence of 50 μg ml − 1 CsA, and its sensitive degree was comparable to that of C. albicans cells lacking both CaGDT1 and CaPMR1 on YPD plates containing these reagents without CsA (Compare Fig. 1b to a). Furthermore, addition of CsA in YPD plates made C. albicans cells more sensitive to CR and CFW independent of CaGDT1 and CaPMR1 (Compare Fig. 1b to a). This suggests that calcineurin function is additive to the overlapping function between CaGdt1 and CaPmr1 in the response of C. albicans cells to cell wall stress. Fig. 1 Roles of CaGDT1 and CaPMR1 in the sensitivity to cell wall-perturbing agents and calcium. Phenotypes of the Candida albicans single-gene gdt1/gdt1 and pmr1/pmr1 mutants as well as the double-gene gdt1/gdt1 pmr1/pmr1 mutant in the sensitivity to cell-wall perturbing agents in the absence (a) or presence of cyclosporine A (CsA) (b). Indicated C. albicans strains containing the pCR4 vector or the pCR4-CaGDT1 recombinant plasmid were grown overnight in SD-URA medium, serially diluted by 10 times and spotted onto plates. Plates were incubated at 30°C for 3 days before photos were taken. CR, Congo red. CFW, calcofluor white. TM, tunicamycin

Inhibition of calcineurin promotes cell death and disrupts cell wall and vacuolar integrity of cells lacking both CaGDT1 and CaPMR1
To further investigate the growth defect of C. albicans cells lacking CaGDT1 and CaPMR1 in the presence of CsA, we determined their clonogenic survival ability by measuring their colony-forming units. In response to 25 μg ml − 1 CsA in YPD medium, cells of the gdt1/gdt1 pmr1/pmr1 mutant lost their viability in both time-dependent and dose-dependent fashions ( Fig. 2a  and b). However, CsA treatment of the wild type did not affect its viability (Data not shown). This indicates that CsA promotes the death of cells lacking both CaGDT1 and CaPMR1.
Compared to untreated cells (Fig. 2c), most of cells treated with 25 μg ml − 1 CsA for 8 h were abnormal in their internal structures (Fig. 2d), and appeared to be flakes in culture solution (Fig. 2e). To examine their cell wall integrity, we co-stained these cells with Annexin V-FITC (green fluorescence) and propidium iodide (red fluorescence), which measure externalization of phosphatidylserine and loss of membrane integrity of yeast cells, two indicators of cell death [39,40]. Only a few of the wild-type and the mutant cells untreated with CsA were stained (Fig. 2f ). Comparing to the CsA-treated wild-type with only a few of stained cells, most of mutant cells treated with CsA were stained, by these two dyes (Fig. 2g), indicating susceptible cell walls of CsA-treated mutant cells for an easy entry of these dyes.
To understand the nature of the abnormal intracellular compartments in CsA-treated cells lacking both CaGDT1 and CaPMR1 (Fig. 2d), we examined the integrity of their vacuoles with the dye FM4-64 [33]. Untreated wild-type and mutant cells showed a uniform  (Fig. 2h). Wild-type cells treated with CsA showed a uniform staining pattern, while mutant cells treated with CsA exhibited various staining patterns (Fig. 2i), including cells with normal intracellular compartments and no intracellular staining, indicating their lost ability of transporting FM4-64 through endocytosis, and cells with abnormal intracellular compartments and strong whole-cell staining, indicating their disrupted vacuolar structures. Taken together, these results suggest that calcineurin function is essential for the cell survival as well as the cell wall and vacuolar integrity of cells lacking both CaGDT1 and CaPMR1.
Calcium uptake and activation of the calcium/calcineurin signaling in C. albicans cells lacking CaGDT1 and CaPMR1 Although deletion of CaGDT1 did not alter the sensitivity of C. albicans cells to calcium stress, we showed that CaGDT1 complemented the function of ScGDT1 in calcium stress ( [30]; Additional file 1: Figure S1A). Both ScGdt1 and its human ortholog TMEM165 are involved in the regulation of cytosolic Ca 2+ and pH homeostasis [7]. We next measured the uptake of 45 Ca 2+ in mutants for CaGDT1 and/or CaPMR1. As compared to the wild type, both the gdt1/gdt1 and the pmr1/pmr1 mutants exhibited similar high levels of 45 Ca 2+ uptake in YPD medium, while their double-gene mutant showed an even higher level of 45 Ca 2 uptake (Fig. 3a). Increased calcium uptake was also observed for cell lacking ScPMR1 [41]. These results suggest that CaGdt1 and CaPmr1 have additive functions in the regulation of calcium homeostasis. This is consistent with observations for ScGdt1 and ScPmr1 in S. cerevisiae [7,42]. Fig. 3 Function of CaGDT1 and CaPMR1 in calcium uptake, calcineurin signaling and gene expression. a Ca 2+ accumulation in C. albicans cells lacking CaGDT1 and/or CaPMR1. Cells of the wild type (CAI4), the gdt1/ gdt1 mutant, the pmr1/pmr1 and the gdt1/gdt1 pmr1/pmr1 mutant were grown to log-phase in YPD medium in the presence of 45 Ca 2+ , and OD 600nm values and intracellular 45 Ca 2+ levels of these strains were measured. b β-galactosidase activities of UTR2-lacZ reporter in the wild type and the three mutant cells in the absence or presence of 300 mM CaCl 2 . *indicates the statistically significant differences between each mutant and the wild-type (p < 0.05). # indicates the significant difference between the gdt1/gdt1 pmr1/pmr1 mutant and each of its two single-gene deletion mutants (< 0.05). c Shared differentially expressed genes (DEGs) between cells lacking the gdt1/gdt1, the pmr1/ pmr1 or the gdt1/gdt1 pmr1/pmr1 mutants. DEGs for each mutant are obtained from its comparison to the wild type, which include 249 (=116 + 29 + 76 + 28) genes for the gdt1/gdt1, 446 (=197 + 29 + 76 + 144) genes for the pmr1/pmr1 mutant, and 440 (=192 + 28 + 76 + 144) genes for the gdt1/gdt1 pmr1/pmr1 mutant, respectively. Total numbers of shared DEGs between two, or among three, mutants are indicated We next examined activation levels of the calcium/calcineurin signaling in these mutants with the lacZ reporter of CaUTR2, a downstream target of CaCrz1 [43]. There was no significant difference in the lacZ activity between the wild type and each of three mutants in YPD medium (Fig. 3b). In the presence of 0.2 M CaCl 2 , both the pmr1/pmr1 and the double-gene mutants showed higher lacZ activities than, but the gdt1/gdt1 mutant showed a similar lacZ activity to, the wild type (Fig. 3b). These CaUTR2p::lacZ activity results agree with our transcriptomic profiling data on the transcript levels of CaUTR2 (CAWG_02505) in these mutants and the wild type (SuppInfo 1 and 2 [GEO Accession number: GSE100737]). Taken together, these data indicate that CaPmr1 is the major controller of calcium homeostasis in C. albicans, which in consistent with observations on ScPmr1 in S. cerevisiae [7,42].
CaGdt1 plays a role in the CaCek1-mediated, but not CaMkc1-mediated, cell wall integrity signaling Deletion of CaPMR1 activates the Mkc1-mediated CWI signaling [10]. Since deletion of CaGDT1 increased the sensitivity of cells lacking CaPMR1 to cell wall stresses (Fig. 1a), we examined the effect of CaGDT1 deletion on CWI signaling. We chromosomally tagged the HA epitope to the C-terminus of the CaPKC1 allele and the CaMKC1 allele, respectively, in the wild type CAI4, the WJCAG22 (gdt1/gdt1), the NGY98 (pmr1/pmr1) and the WJCAG25 (gdt1/gdt1 pmr1/pmr1) strains (Additional file 1: Figure S2A, 2C and 2D). In YPD medium, phosphorylation levels of the wild type CaMkc1 (or the HA-tagged CaMkc1-HA fusion) were similar between the wild type and the gdt1/gdt1 strains in the absence or presence of 50 μg ml − 1 CFW (Fig. 4a). As reported previously [10], phosphorylation levels of CaMkc1 (or CaMkc1-HA) increased with similar degrees in both the pmr1/pmr1 and the gdt1/gdt1 pmr1/pmr1 mutants as compared to the wild type (Fig. 4a). Similar patterns for CaMkc1 phosphorylation levels were also observed in C-terminally HA tagged CaCEK1 strains (Fig. 4b). These results indicate that CaPmr1 does, but CaGdt1 does not, play a role in the CaMkc1-mediated CWI signaling -.
The CaCek1-mediated CWI signaling is also involved in cell wall construction [18]. Phosphorylated CaCek1 and CaCek1-HA) were not detected in both the wild type and the gdt1/gdt1 strains, while the gdt1/gdt1 pmr1/pmr1 mutant showed a higher level of phosphorylated CaCek1 and CaCek1-HA than the pmr1/pmr1 mutant in the absence or presence of CFW (Fig. 4b).
Similarly, in C-terminally HA tagged CaMKC1 strains, the gdt1/gdt1 pmr1/pmr1 mutant showed a higher level of phosphorylated CaCek1 than the pmr1/pmr1 mutant in the absence or presence of CFW, while phosphorylated CaCek1 was detected at very low levels in both the wild type and the gdt1/gdt1 strains, with a slightly higher level in the latter (Fig. 4a). Taken together, these data suggest that CaGdt1 and CaPmr1 are both involved in the activation of CaCek1-mediated CWI signaling independent of cell wall stress.

CaGdt1 is localized in the Golgi apparatus but at distinct sites from CaPmr1
To determine the subcellular localization of CaGdt1, we chromosomally tagged the green florescent protein Fig. 4 Activation of the CaMkc1-and CaCek1-mediated MAP kinase signaling in cells lacking CaGDT1 and/or CaPMR1. Phosphorylation levels of CaMkc1, CaCek1 and their C-terminally HA-tagged versions were determined simultaneously by Western blotting using the Phospho-p44/42 MAPK Rabbit mAb in protein extracts prepared from log-phase growing cells of the wild-type (WT) CAI4, the WJCAG22 (gdt1/gdt1), the NGY98 (pmr1/pmr1) and the WJCAG25 (pmr1/pmr1 gdt1/gdt1) strains, in which either the CaMKC1 allele (Panel a) or the CaCEK1 allele (Panel b) was chromosomally HA-tagged, in the presence or absence of 50 μg ml − 1 Calcofluor white (CFW). The anti-HA antibody (Abmart, Shanghai, China) was used to reveal the total amount of CaMkc1-HA and CaCek1-HA proteins loaded. Each lane was loaded with 50 μg of total proteins (GFP) to the C-terminus of CaGdt1 in the heterozygous mutant for CaGDT1 generating WJCA102 (CAI4 pmr1pmr1 GDT1-GFP/gdt1) (Additional file 1: Figure  S2). Like the pmr1/pmr1 mutant, the WJCA102 mutant was not sensitive to CsA, while the pmr1/pmr1 gdt1/ gdt1 mutant was (Fig. 5a). This indicates that the CaGDT1-GFP allele is functional. In log-phase growing cells of WJCA102, CaGDT1-GFP was colocalized with the Golgi Tracker Red dye (Fig. 5b).
Similarly, we constructed the WJCA111 (RM1000 PMR1/PMR1::HA-HIS1 GDT1/GDT1::GFP-URA3) strain to examine the colocalization between CaGdt1 and CaPmr1 (Data not shown). Through indirect immunofluorescent approach, we found that the CaGDT1-GFP protein was not colocalized with the CaPMR1-HA in log-phase growing WJCA111 cells (Fig. 5c). It should be noted that there was a discrepancy in CaGDT1-GFP localization between Fig. 5b and c, which could be due to the difference in sample processing for florescent microscopy as described in the Methods section. In response to depletion of extracellular calcium (1 mM EGTA treatment) or high levels of extracellular calcium, the CaGDT1-GFP protein was also not colocalized with the CaPMR1-HA in WJCA111 cells (Additional file 1: Figure S3). Taken together, these results indicate that CaGdt1 is present in the Golgi apparatus but at distinct sites from CaPmr1 in C. albicans.

Transcriptomic profiling of cells lacking CaGDT1 and/or CaPMR1
To help us further understand functions of CaGdt1 and CaPmr1, we carried out transcript profiling for the wild type CAI4, the gdt1/gdt1, the pmr1/pmr1, and the gdt1/ gdt1 pmr1/pmr1 mutants, growing in log phase in YPD medium at 30°C (GEO accession number: GSE100737). Transcripts of 5901 genes at various expression levels were detected in each of these four strains (SuppInfo 1 [GEO Accession number: GSE100737]). As compared to the wild type, there are 249, 446 and 440 differentially expressed genes (DEGs) in the gdt1/gdt1, the pmr1/ pmr1 and the gdt1/gdt1 pmr1/pmr1 mutants, respectively ( Fig. 3c; SuppInfo 2). Deletion of CaGDT1 does not affect the transcript level of CaPMR1, and vice versa (SuppInfo 2).
There are 105 shared DEGs between the gdt1/gdt1 and the pmr1/pmr1 mutants, which accounts for 42% (105/ 249) of total DEGs due to deletion of CaGDT1 or 24% (105/446) of total DEGs due to deletion of CaPMR1 (Fig. 3c). This indicates that CaGdt1 and CaPmr1 have shared and distinct functions. Furthermore, there are 76 shared DEGs among the gdt1/gdt1, the pmr1/pmr1 and c Co-localization of CaGDT1-GFP and CaPMR1-HA in WJCA111 cells through indirect immunofluorescent approach. Images of differential interference contrast (DIC), GFP, red florescence derived from Golgi-Track Red dye (b) or from goat anti-mouse IgG conjugated to Alexa Fluor 555 (for CaPMR1-HA protein) (c) and their merged images are presented. Scale bar, 5 μm the gdt1/gdt1 pmr1/pmr1 mutants, with 23 of them being up-regulated and 53 of them being down-regulated ( Fig. 3c; Table 3). Using a semi-quantitative RT-PCR approach, we confirmed the transcript levels of selected 15 out of these 76 genes in the wild type, the gdt1/gdt1, the pmr1/pmr1 and the gdt1/gdt1 pmr1/pmr1 mutants (Additional file 1: Figure S4). The 23 up-regulated DEGs are involved in the cellular transport, including the transporters of long-chain fatty acids (PXA1 and PXA2), the calcium channel (CCH1) and the potassium transporter (TRK1) ( Table 3). Similarly, a big portion of the 53 down-regulated DEGs are also involved in the cellular transport, with different substrates including amino acids (MUP1, QDR1, ALP1, GNP1, TPO3 and TPO4) [44], lipids (PDR16), nucleosides (CNT) [45], calcium ions (RCH1) [13] and sodium ions (ENA2). In addition, 10 of these down-regulated DEGs are involved in the protein synthesis (Table 3). These data suggest that core functions shared by CaGdt1 and CaPmr1 are involved in the regulation of cellular transport of metal ions and amino acids.
CaGdt1 homologs in humans and S. cerevisiae have been shown to regulate glycosylation pathway [25,27]. In S. cerevisiae, there are two main types of glycosylation, the N-linked and the O-linked glycosylation (http:// www.kegg.jp/kegg/pathway.html). The N-linked glycosylation is the attachment of the oligosaccharide known as glycan to the amide nitrogen of an asparagine residue of a protein, There are 30 genes involved in the N-linked glycosylation in S. cerevisiae, with 20 of them having homologs in C. albicans (Additional file 1: Table S2). Transcript levels of CAWG_02847 (DIE2) and CAWG_03939 (STT3) are up-regulated only in the gdt1/gdt1 pmr1/pmr1 mutant, while that of CAWG_04868 (OST3) is up-regulated in both the pmr1/pmr1 mutant and the gdt1/gdt1 pmr1/pmr1 mutant, but not in the gdt1/gdt1 mutant (Additional file 1: Table S2). The O-linked glycosylation is the attachment of a sugar molecule (such as N-acetyl-galactosamine, N-acetylglucosamine and mannose) to an oxygen atom in serine or threonine residue of a protein. There are 13 genes involved in the O-linked glycosylation n S. cerevisiae, with 7 of them showing homologs in C. albicans (Additional file 1: Table S3). Expression of CAWG_05623 (PMT1) and CAWG_04353 (PMT4) is up-regulated only in the gdt1/gdt1 pmr1/pmr1 mutant (Additional file 1: Table S3). Furthermore, from the C. albicans genome database we identified 20 genes involved in the CWI pathway based on their homologs in S. cerevisiae (Additional file 1: Table S4) [46]. Only one gene CAWG_01201 (SWI4) was transcriptionally upregulated in each of three mutants, while expression of the gene CAWG_05506 (BEM2) was down regulated in the pmr1/pmr1 mutant, but not affected in both the gdt1/gdt1 mutant and the gdt1/gdt1 pmr1/pmr1 mutant (Additional file 1: Table S4).

Discussion
Congenital disorders of glycosylation (CDG) are rare inherited diseases causing severe growth and psychomotor retardations in patients, with most of their genetic defects affecting the glycosylation process. The human ortholog for yeast ScGdt1, TMEM165, is one of these CDG-associated Golgi proteins and might function as a Ca2 + /H + antiporter and/or a Mn 2+ transporter in the Golgi [26,27,[46][47][48]. In line with these findings, we have provided evidence that the C. albicans functional ortholog of ScGdt1, CaGdt1, plays a compensatory role for CaPmr1, the high affinity Ca 2+ /Mn 2+ P-type ATPase that is required for Ca 2+ and Mn 2+ transport into the Golgi and involved in Ca 2+ dependent protein sorting processing, in the regulation of calcium homeostasis and cell wall stress. In addition to the role of CaPmr1 in activation of Mkc1-mediated CWI signaling [10], we show that deletion of CaPMR1 activates the Cek1-mediated CWI signaling (Fig. 4). This agrees with the function of Cek1 in cell wall construction [18]. In contrast, deletion of CaGDT1 alone does not affect the Mkc1-mediated CWI signaling, but activates the Cek1-mediated signaling. Although C. albicans cells lacking CaPMR1 are not sensitive to SDS [10], which targets both the cell wall and the plasma membrane, deletion of CaGDT1 renders C. albicans cells lacking CaPMR1 sensitive to SDS and more sensitive to hygromycin B (Additional file 1: Figure  S1D). We also demonstrate that CaGdt1 and CaPmr1 are present at distinct sites in the Golgi apparatus, which is not affected by the calcium depletion or high levels of calcium in the extracellular environment. CaGdt1 is a functional homolog of ScGdt1 in the response of S. cerevisiae cells to calcium and cell wall stresses (Additional file 1: Figure S1; 30). Therefore, our data supports the hypothesis that CaGdt1 together with human TMEM165 and ScGdt1 might form a new group of Golgi-localized Ca 2+ /H + exchangers [6].
As the central player in the calcium signaling, calcineurin has been shown to play a role in early synaptic dysfunction and neuronal death in mammalian cells [49]. In addition, ER stress can induce nonapoptotic cell death in yeasts, which can be blocked by the action of calcineurin [50]. Consistently, we show here that calcineurin function is required for the growth as well as the cell wall and vacuolar integrity of cells lacking both CaGDT1 and CaPMR1. Therefore, the calcium/calcineurin signaling, CaGdt1 and CaPmr1 work together to regulate the integrity of both the external cell wall and the intracellular vacuole in C. albicans.
Deletion of ScGDT1 increases the calcium sensitivity of S. cerevisiae cells lacking ScPMR1, while deletion of CaGDT1 does not affect the calcium sensitivity of C. albicans cells lacking CaPMR1 (Fig. 1a; Additional file 1: Figure S1A). It is interesting to note that although deletion of CaGDT1 increases the intracellular calcium concentration in C. albicans cells lacking CaPMR1 and elevates their activation levels of the calcium/calcineurin signaling, it does not affect their calcium sensitivity (Figs. 1 and 3a and b). Nevertheless, deletion of CaGDT1 increases the sensitivity of C. albicans cells lacking CaPMR1 to cell wall and ER stresses (Fig. 1). Therefore, the calcium homeostasis in the ER/Golgi apparatus is mainly involved in the regulation of the cell wall integrity. This is further supported by our recent observation on the deletion effect of CaRCH1, encoding a negative regulator of calcium uptake on the plasma membrane, on cells lacking CaPMR1 [51]. In addition, normal growth of C. albicans cells lacking both CaGDT1 and CaPMR1 is, but that of S. cerevisiae cells lacking both ScGDT1 and ScPMR1 is not, sensitive to the inhibition of calcineurin function ( Fig. 1b; Additional file 1: Figure  S1C). Furthermore, CaGdt1 and CaPmr1 localize at different sites of the Golgi compartment in C. albicans, irrespective of alterations in extracellular calcium levels ( Fig. 5; Additional file 1: Figure S3). In contrast, ScGdt1 and ScPmr1 are co-localized in the Golgi compartment of S. cerevisiae cells [6]. This difference in the relative localization between Gdt1 and Pmr1 might partially contribute to the differential response of cells lacking both GDT1 and PMR1 to the inhibition of calcineurin in these two microorganisms.
To further investigate the mechanism by which both CaGdt1 and CaPmr1 regulate the survival of C. albicans cells, we have examined the transcriptomic profiles of C. albicans cells lacking CaGDT1 and/or CaPMR1. Cells lacking CaGDT1 or CaPMR1 share a significant part of DEGs (Fig. 3c), suggesting that CaGdt1 and CaPmr1 have both shared and distinct functions in C. albicans cells. This is consistent with their distinct localizations in the Golgi compartments of C. albicans cells. We have shown that 192 DEGs are exclusively present in the gdt1/gdt1 pmr1/pmr1 mutant (Fig. 3c). Functions of these genes should be responsible for the unique phenotypes of C. albicans cells lacking both CaGDT1 and CaPMR1. Some of these genes might be regulated by the calcium signaling, so inhibition of their transcription by CsA might be responsible for the failure of C. albicans cells lacking both CaGDT1 and CaPMR1 to survive in the presence of CsA. As compared to the wild type cells, cells lacking CaGDT1, CaPMR1 or both share 76 DEGs, and most of them involved in the protein synthesis and the transport of amino acids, lipids and nucleosides are down-regulated (Table 3) This might explain why C. albicans cells lacking CaGDT1 and/or CaPMR1 grow slower than the wild type (Additional file 1: Figure S5).
During N-linked glycosylation, an oligosaccharide chain is assembled on the carrier molecule dolichyl pyrophosphate in the following order: 2 molecules of N-acetylglucosamine (GlcNAc), 9 molecules of mannose, and 3 molecules of glucose. DIE2 encodes an alpha-1,2 glucosyltransferase that catalyzes the addition of the third glucose moiety [52]. This 14-residue oligosaccharide core is then transferred by the oligosaccharyl transferase complex (OST complex), consisting of nine ER integral membrane protein subunits Ost1-6, Stt3, Swp1 and Wbp1, to an asparagine residue on a nascent prolylpeptide in the ER. As this protein progresses through the Golgi apparatus, its oligosaccharide core is modified by trimming or extension. As a result, a diverse array of N-glycosylated proteins is generated [52,53]. STT3 is an essential gene, and its conditional mutants are defective in cell wall biosynthesis [54]. On the other hand, protein O-mannosylation (POM) is an evolutionarily conserved, essential posttranslational modification that impacts a variety of cellular processes in both fungi and mammals, which is initiated at the ER by a family of dolichyl phosphate mannose-dependent protein O-mannosyltransferases (PMTs). PMTs transfer mannose residues from dolichyl phosphate-D-mannose to protein seryl/threonyl residues. The POM is essential for yeast cell wall integrity [55,56]. In S. cerevisiae, the PMT family is highly redundant (Pmt1-Pmt6) and is classified into three subfamilies, PMT1, PMT2, and PMT4, which mannosylate distinct target proteins [56]. In the present study, we have shown that lack of CaGDT1 and CaPMR1 upregulates the expression of three N-linked glycosylation genes CaDIE2, CaOST3 and CaSTT3 as well as two major O-linked glycosylation genes CaPMT1 and CaPMT4 (Additional file 1:  Tables S2 and S4). Since the activation level of the calcium/ calcineurin signaling is significantly elevated due to deletion of CaGDT1 and CaPMR1, our current results indicate that expression of these six glycosylation genes might be under control of the transcription factor CaCrz1. In this way, the defect in cell wall integrity due to deletion of CaGDT1 and CaPMR1 could be compensated by enhanced glycosylation process, specifically the increased biosynthesis of the cell wall beta1,6-glucan [54].
Chitin is an essential component of the fungal cell wall, and C. albicans has four chitin synthases, two class I enzymes encoded by CaCHS2 and CaCHS8, one class II enzyme encoded by CaCHS1 and one Class IV enzyme encoded by CaCHS3. All four genes are stimulated by CFW treatment. In addition, only expression of CaCHS2 and CaCHS8 is mediated by CaMkc1-mediated CWI signaling [57]. Our transcript profiling data indicates that expression of both CHS1 (CAWG_05611) and CHS8 (CAWG_02407) was not affected due to deletion of CaGDT1 and/or CaPMR1. CHS2 (CAWG_02178) was up regulated in all three mutants, while CHS3 (CAWG_00138) was only up regulated in the pmr1/ pmr1 and the gdt1/gdt1 pmr1/pmr1 mutants (SuppInfo 2 [GEO Accession number: GSE100737]). In addition, we have observed that expression of CaSWI4 is upregulated due to deletion of CaGDT1 and/or CaPMR1, while expression of BEM2 is down regulated due to the deletion of CaPMR1 (Additional file 1: Table S4). CaSWI4 encodes one of terminal transcription factors of the CWI pathway and its mutants are sensitive to the cell wall-interfering agent caffeine [58]. Bem2 is a negative regulator of the central CWI regulator Rho1 in S. cerevisiae, albeit not characterized in C. albicans [46]. These results agree with our observations on the activation of Mkc1-and Cek1-mediated CWI signaling in these mutants (Fig. 4) as well as the previous study [57]. Since Ca 2+ -induced upregulation of CHS promoters is independent of the CWI signaling [57], inhibition of expression of both CaCHS2 and CaCHS3 by CsA might contribute to defects in the integrity of the external cell wall and intracellular vacuoles of C. albicans cells lacking both CaGDT1 and CaPMR1, leading to their death.