Atf2 Transcription Factor Binds to the APP1 Promoter in Cryptococcus neoformans: Stimulatory Effect of Diacylglycerol

Abstract

The fungus Cryptococcus neoformans is an environmental human pathogen which enters the lung via the respiratory tract and produces a unique protein, called antiphagocytic protein 1 (App1), that protects it from phagocytosis by macrophages. In previous studies, we proposed genetic evidences that transcription of APP1 is controlled by the enzymatic reaction catalyzed by inositol phosphorylceramide synthase 1 (Ipc1) via the production of diacylglycerol through the activating transcription factor 2 (Atf2). We investigated here the mechanism by which Atf2 binds to the APP1 promoter in vitro and in vivo. To this end, we produced Atf2 recombinant proteins (rAtf2) and found that rAtf2 binds to ATF cis-acting element present in the APP1 promoter. Indeed, mutation of two key nucleotides in the ATF consensus sequence abolishes the binding of rAtf2 to the APP1 promoter. Next, we produced C. neoformans strains with a hemagglutinin-tagged ATF2 gene and showed that endogenous Atf2 binds to APP1 promoter in vivo. Finally, by a novel DNA protein-binding precipitation assay, we showed that treatment with 1,2-dioctanoylglycerol positively increases binding of Atf2-APP1 promoter in vivo. These studies provide new insights into the molecular mechanism by which Atf2 regulates APP1 transcription in vivo with important implications for a better understanding of how C. neoformans escapes the phagocytic process.

Cryptococcus neoformans is an environmental fungal pathogen that infects humans through inhalation. Once in the lung, phagocytosis by alveolar macrophages (AMs) represents the first line of defense against C. neoformans, and the killing of the organism is controlled by an efficient host-cellular response (1, 15-17). However, in conditions of cellular immune deficiency, C. neoformans can disseminate to other organs, especially the brain, where it causes a life-threatening meningoencephalitis (3). Thus, fungal factors that inhibit the phagocytosis by AMs may assume a critical role in the outcome of the infection (4).

In previous studies, we identified a novel cryptococcal gene encoding for an antiphagocytic protein 1 (App1), which specifically inhibits the phagocytosis of C. neoformans by AMs (11). APP1 gene transcription was found to be under the control of inositol phosphorylceramide synthase 1 (Ipc1), a key enzyme in the fungal sphingolipid pathway because it regulates the cellular level of phytoceramide, complex sphingolipids, and diacylglycerol (DAG) (7, 8). Using a reporter gene, we found that DAG positively regulates the activity of the APP1 promoter (13), suggesting that Ipc1 regulates APP1 through the formation of DAG. Further studies revealed the presence of two consensus sequences in the APP1 promoter, AP-2 and ATF cis-acting elements, and the involvement of the ATF2 as a positive regulator of APP1 transcription (13).

ATF consensus sequence belongs to the cyclic AMP response element (CRE) family, a palindromic octanucleotide (TGACGTCA) that has been identified in the transcriptional regulatory regions of a large number of eukaryotic genes. The transcription of many eukaryotic genes is regulated by the binding of sequence-specific transcription factors to modular cis-acting promoter and enhancer elements. Although a lot is known about the ATF/CREB proteins in eukaryotes (2, 5, 9, 14), not that much is known regarding ATF in yeast, and particularly in C. neoformans.

To study the mechanism by which Atf2 regulates transcription of APP1, we produced recombinant Atf2 protein (rAtf2) and by electrophoretic mobility shift assay (EMSA) we showed that Atf2 binds in vitro to ATF double-stranded oligonucleotides and to the APP1 promoter region. Mutation of two nucleotides in the ATF consensus sequence abolishes the binding of Atf2 to APP1 in vitro. We then tagged the endogenous ATF2 gene by using the HA epitope and, using two independent and complementary protein-DNA binding assays, we conclusively show that Atf2 binds to APP1 promoter in vivo and that DAG positively regulated this binding.

MATERIALS AND METHODS

Strain, growth media, and reagents.

C. neoformans var. grubii serotype A wild-type strain H99, the GAL7::IPC1 (12), IPC1/APP1::LUC, and the GAL7::IPC1/APP1::LUC strains (13), and four derivative strains in which ATF2 gene has been tagged with hemagglutinin (HA) epitope (IPC1/ATF2::HA, GAL7::IPC1/ATF2::HA, IPC1/APP1::LUC/ATF2::HA, and GAL7::IPC1/APP1::LUC/ATF2::HA) were used in the present study. The strains were routinely grown in yeast extract-peptone-dextrose (YPD) medium. Yeast nitrogen base (YNB) with 2% glucose was used in some experiments as indicated. For upregulation of Ipc1, the GAL7::IPC1/ATF2::HA or GAL7::IPC1/APP1::LUC/ATF2::HA strain was grown on 2% galactose as indicated. Hygromycin B (Calbiochem, San Diego, CA) at a concentration of 200 U/ml was added to YPD plates for selection of the IPC1/ATF2::HA, GAL7::IPC1/ATF2::HA, IPC1/APP1::LUC/ATF2::HA, and GAL7::IPC1/APP1::LUC/ATF2::HA strains.

Identification and cloning of C. neoformans ATF2 cDNA and expression of rAtf2.

To isolate C. neoformans ATF2 cDNA, reverse transcription-PCR using a GeneRace kit (Invitrogen) was performed. Total RNA was extracted from C. neoformans wild-type H99 strain using an RNeasy kit (Qiagen) and used for reverse transcription with an oligo(dT) primer [5′-GCT GTC AAC GAT ACG CTA CGT AAC GGC ATG ACA GTG(T)₂₄-3′] provided in the GeneRacer kit. For the identification of the 5′ of ATF2 gene, the cDNA obtained was subjected to PCR using the ATF2 specific primer Racer A (5′-TCG AAT CCG CCT TCA GCGTT-3′) and the GeneRacer 5′ primer (5′-CGA CTG GAG CAC GAG GAC ACT GA-3′). The resulting ∼900-bp fragment was then cloned into a PCR-TOPO cloning vector and sequenced. For the identification of the 3′ of ATF2 gene, the cDNA obtained was subjected to PCR using the ATF2 specific primer 3Race-2 (5′-GAA TTT CTT GGA GAG GAA TCG ACA AG-3′) and the GeneRacer 3′ primer (5′-GCT GTC AAC GAT ACG CTA CGT AACG-3′). The resulting ∼850-bp fragment was cloned into a PCR-TOPO cloning vector and sequenced. BLAST analysis of the two sequences with the C. neoformans genome databases revealed 100% homology with putative ATF2 gene. ATG start and stop codon were identified, and the full lengths of ATF2 cDNA were amplified by using the cDNA made above as a template and the primers BH-ATF (5′-CTA GGA TCC TAT GGC TGC AGT TGC ACA AGCA-3′) and KI-ATF (5′-CAT GGT ACC ATT CAT CGC AAC CTT CCC CCATA-3′), which contain BamHI and KpnI sites, respectively (underlined). The amplified ∼1.9-kb fragment was digested with BamHI and KpnI and ligated into a BamHI- and KpnI-restricted pRSETB vector (Invitrogen), yielding the pRSETB/cATF2 plasmid. This plasmid was sequenced to make sure that no mutations were found in the ATF2 cDNA gene.

For Atf2 protein expression, the pRSETB/cATF2 plasmid was transformed into BL21(DE3) competent cells (Stratagene) according to the chemical transformation procedure. Atf2 protein expression was achieved as follows. First, 2 ml of BL21 competent cells transformed with pRSETB/cATF2 were grown overnight in LB broth containing 50 μg of ampicillin/ml. The next morning, 1 ml of the overnight culture was used to inoculate a 200-ml culture of LB broth containing 50 μg of ampicillin/ml, which was grow in a shaker incubator at 250 rpm at 37°C until the optical density at 600 nm reached 0.6 (∼4 h). At this time, 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) was added, and the cells were further incubated for another 2 h and 30 min. The cells were harvested, washed with phosphate-buffered saline (PBS) containing 10 mM EDTA, washed again with PBS only, and then resuspended in a buffer containing 20 mM sodium phosphate, 0.5 M NaCl, and 20 mM imidazole (pH 7.4). Cells were than incubated with 1 mg of lysozyme (chicken egg white; Sigma)/ml for 30 min at 37°C; the samples were then sonicated for four cycles at 25% power for 8 s each time. Samples were then centrifuged at 4°C at 13,000 rpm, and the supernatant was loaded into a His-Trap FF affinity column (GE Healthcare) for purification. Eluted Atf2 was then concentrated and resuspended in a buffer containing 20 mM HEPES, 50 mM KCl, 25% glycerol, 0.2 mM EDTA, 1 mM dithiothreitol (DTT), and 1 mM phenylmethylsulfonyl fluoride, and it was stored at −80°C until ready to use. Western blot analysis of rAtf2 was performed using anti-Xpress antibody (Invitrogen).

EMSA. (i) Preparation of the double-stranded DNA fragments.

ATF and AP-2 consensus sequences present in the APP1 promoter were used in a DNA-protein binding assay. Single-stranded palindromic oligonucleotides ATF-TF1 (5′-CTG TCA TGACGTCA AAA GTCG-3′) and ATF-TF2 (5′-CGA CTTT TGACGTCA TGA CAG-3′) containing an ATF cis-acting element (underlined) were obtained from IDT Technology. The oligonucleotides were resuspended in PBS, heated at 90°C, and slowly cooled to room temperature to allow alignment and the formation of double-stranded ATF DNA. Since the APP1 promoter contains another cis-acting element (AP-2), additional oligonucleotides—AP2-TF1 (5′-ACA TGT CCCCGCGGC TCG CCT-3′) and AP2-TF2 (5′-AGG CGA GCCGCGGGG ACA TGT-3) containing AP-2 (underlined)—were also ordered and aligned producing double-stranded AP-2. Single-stranded palindromic oligonucleotides ATF-deltaTF1 (5′-CTG TCA TGAAATCA AAA GTCG-3′) and ATF-deltaTF2 (5′-CGA CTTT TGATTTCA TGA CAG-3′) containing a mutated ATF cis-acting element (underlined) were also from IDT technology and used as a negative control for competition assays.

A large region of the APP1 promoter (256 bp) immediately upstream the ATG start site containing both ATF and AP-2 cis-acting elements was also produced by PCR using the primers APP1-1 (5′-CTT GGG ACA TTT GAA AAA ATC-3′) and APP1-2 (5′-AAC TCA ACG GAA GAA TGT AAT-3′). The resulting APP1-256 fragment was then cloned into a PCR-TOPO vector and sequenced. The resulting pTOPO-APP1-256 plasmid was subjected to PCR site-directed mutagenesis (Invitrogen) to mutate ATF, AP-2, or both cis-acting elements. The AP-2 cis-acting element (CCCCGCGGC) was mutated into CCCCGCAAC by using the primers AP-2 MUT1 (5′-ATG GGG ACA TGT CCC CGC AAC TCG CCT TGTC-3′) and AP-2 MUT2 (5′-GCG GGG ACA TGT CCC CAT AGT GGG GGTCG-3′) and the pTOPO-APP1-256 plasmid as a template. The resulting plasmid was named pTOPO-APP1-256/ΔAP-2. ATF cis-acting element (TGACGTCA) was mutated into TGAAATCA by using the primers ATF-MUT1 (5′-TGT AGT CAC CTG TCA TGA AAT CAA AAG TCGT-3′) and ATF-MUT2 (5′-TCA TGA CAG GTG ACT ACA TAG TAA TCGCT-3′) and pTOPO-APP1-256 plasmid as a template. The resulting plasmid was named pTOPO-APP1-256/ΔATF. The pTOPO-APP1-256/ΔAP-2 plasmid was also used as a template, with ATF-MUT1 and ATF-MUT2 generating pTOPO-APP1-256/ΔAP2-ΔATF in which both AP-2 and ATF consensus sequences have been mutated. All plasmids were sequenced to make sure that point mutations were inserted only in the desired locations. Plasmids were then digested with EcoRI, and the corresponding APP1p-WT, APP1p-ΔATF, APP1p-ΔAP2, and APP1p-ΔATF-ΔAP2 fragment was purified from the agarose gel for radioactive labeling.

(ii) Radioactive labeling of the double-stranded DNA fragments.

For radiolabeling, 1.87 pmol of double-stranded DNA fragments was labeled with 10 μCi of [α-³²P]ATP (Amersham) by using 10 U of T4 DNA kinase (Promega). For competition assay, cold wild-type or ATF mutated double-stranded DNA oligonucleotides were used at a 300× concentration (∼561 pmol). Recombinant Atf2 protein at the indicated concentrations was incubated in a binding buffer containing 2 μl of 10× buffer HDKE (50 mM KCl, 20 mM HEPES, 5% glycerol, 1 mM EDTA, 0.2 M DTT), 1 μl of a 1-μg/μl concentration of pdI.dc (Amersham), 1 μl of a 1-μg/μl concentration of pdN6 (GE Healthcare), 1 μl of a 10-mg/ml concentration of bovine serum albumin, 1 μl of 0.1 M DTT, and 1 μl of radiolabeled oligonucleotide probe (60,000 to 100,000 cpm). The reaction mixture was incubated for 20 min at room temperature and then stopped with 6 μl of Ficoll dye (25% bromophenol blue, 25% xylene, 15% Ficoll). For supershift experiments, 2 μl of a 1.3-μg/ml concentration of anti-His antibody was added to appropriate samples, followed by incubation for 30 min on ice before Ficoll termination. The reaction mixtures were separated on a 5% native polyacrylamide gel, and autoradiography was performed by exposure to Kodak film.

ChIP.

IPC1/APP1::LUC/ATF2::HA and IPC1/APP1::LUC strains were grown in YNB containing 2% glucose overnight in a 10-ml culture in a shaker incubator at 30°C. The day after, 100 μl was transferred into 50 ml of fresh YNB medium with 2% glucose until the optical density at 600 nm reached 1.5. Once the density was reached, formaldehyde was added at a 1% final concentration, and the culture was further incubated for 20 min at room temperature with occasional swirling. Glycine was then added at a final concentration of 125 mM, and the cells were further incubated for 5 min at room temperature. The cells were pelleted at 1,500 × g for 10 min and then washed first with cold Tris-buffered saline (TBS) containing 125 mM glycine and then with cold TBS. The cells were transferred into a 2.0-ml tube, pelleted, and resuspended in chromatin immunoprecipitation (ChIP) lysis buffer (50 mM HEPES [pH 7.5], 140 mM NaCl, 1% Triton X-100, 0.1% sodium deoxycholate, 1 mM EDTA). Protease inhibitors and glass beads were added, and the cells were broken in the homogenizer after four cycles of 40 s each with a 1-min pause between runs. The bottom of the tube was pierced with a hot 20-gauge 1.5 needle, and the tubes were then placed inside a 15-ml tube and centrifuged for 5 min at 1,500 × g. The cell pellet was emulsified with the supernatant and then sonicated five times with 1-min pauses in ice between runs at 35% power. The mixture was then centrifuged at 10,000 × g for 10 min at 4°C, the supernatant was transferred in another tube that was recentrifuged for 5 min at 10,000 × g at 4°C. The resulting supernatant was used for protein measurement according to the Bradford method. Before immunoprecipitation with anti-HA antibody, nonspecific interactions blocked with 1 mg of total protein were incubated with 25 μl of rec-protein G-Sepharose 4B conjugate (2.7 mg of protein per ml of beads; Zymed Laboratories) for 1 h with rotation in a cold room and then washed and incubated for 3 h with 15 μl of a 0.25-μg/μl concentration of rabbit anti-HA antibody (Zymed Laboratories). The mixture was then washed twice with 2 ml of ChIP lysis buffer containing protease inhibitors, twice with 1 ml of ChIP high-salt lysis buffer (50 mM HEPES [pH 7.5], 500 mM NaCl, 1% Triton X-100, 0.1% sodium deoxycholate, 1 mM EDTA) containing protease inhibitors, twice with 1 ml of ChIP wash buffer (10 mM Tris [pH 8.0], 250 mM LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, 1 mM EDTA), and finally twice with 1 ml of Tris-EDTA (pH 8.0). The DNA was eluted using ChIP elution buffer (50 mM Tris [pH 8.0], 1% sodium dodecyl sulfate, 10 mM EDTA), incubated at 65°C in a Thermomixer for 15 min, and centrifuged in a microcentrifuge to remove the beads. To reverse the cross-link, samples were incubated overnight at 65°C. The precipitated DNA was purified by using a Qiagen PCR purification kit, and amplification of the APP1 promoter was performed by using PCR and the primers APP1-1 (5′-CTT GGG ACA TTT GAA AAA ATC-3′) and APP1-2 (5′-AAC TCA ACG GAA GAA TGT AAT-3′). Primers CTRLR (5′-AGT GCT GGC GTT TCG AATT-3′) and CTRLF (5′-ACT CGT CTA TTG AGA ACT CT-3′) were used as a negative control. Briefly, these primers amplify the ATF2 gene. The PCR product was separated in a 0.7% agarose gel, visualized with ethidium bromide, and quantified by using densitometry.

Generation of C. neoformans ATF2::HA-tagged strains.

To tag ATF2 gene with HA, we used the following strategy. First, the pSK/5′UTR/HYG/3′UTR plasmid generated previously (13) was digested with BamHI and XbaI. This digestion generate two fragments: pSK/HYG/3′UTR and 5′UTR. The pSK/HYG/3′UTR was extracted from the gel and purified.

Second, an ATF2-HA fragment was obtained by PCR using genomic DNA from C. neoformans wild-type H99 as a template and primers EXBA (5′-CAT TCTAGA ATG CTC ACT GGT CCC CAG GGCGC-3′) and TAGHA (5′-CAA GGATCC CTA AGC GTA GTC TGG GAC GTC GTA TGG GTA TCG CAA CCT TCC CCC ATA AGG CGC-3′), which contain BamHI and XbaI sites (underlined). The resulting fragment was digested with BamHI and XbaI and ligated with the pSK/HYG/3′UTR fragment, generating a pSK/ATF2-HA/HYG/3′UTR construct. This construct was first sequenced and then digested with XbaI and KpnI, yielding a linearized fragment ATF2-HA/HYG/3′UTR, which was biolistically delivered into IPC1/ATF2 (wild type, H99 strain), GAL7::IPC1, IPC1/APP1::LUC, and GAL7::IPC1/APP1::LUC strains to generate IPC1/ATF2::HA, GAL7::IPC1/ATF2::HA, IPC1/APP1::LUC/ATF2::HA, and GAL7::IPC1/APP1::LUC/ATF2::HA strains, respectively. Stable hygromycin-resistant transformants were selected and subjected to Southern analysis with appropriate probes to identify a double-crossover event at the ATF2 locus.

In vivo DNA protein-binding (DPB) precipitation assay.

This assay is a modified ChIP assay in which the protein of interest bound to the DNA can be visualized by Western blotting. The IPC1/APP1::LUC/ATF2-HA and GAL7::IPC1/APP1::LUC/ATF2::HA strains were grown for 16 h in a 15-ml culture of YNB containing 2% glucose at 30°C in a shaker incubator. Next, formaldehyde was added at 1% final concentration, and the cells were incubated for 20 min at room temperature with occasional swirling. Next, glycine was added at the final concentration of 125 mM, and the cells were further incubated for 5 min at room temperature. The cells were then pelleted at 3,000 rpm for 10 min and washed once with cold TBS containing 125 mM glycine and once with cold TBS. Lysates were prepared as described for the ChIP assay. A 0.66 volume of 100% ethanol was then added, and the mixture was incubated at −80°C for 2 h. The samples were then centrifuged at 13,000 rpm for 1 h at 4°C. Pellets were washed with 70% ethanol, allowed to dry, and resuspended in an appropriate volume of 8 M urea. The samples were then boiled for 20 min and cooled on ice. The DNA concentration was estimated by using a spectrophotometer, and approximately the same amount of DNA was loaded onto a 0.5% agarose gel. Stained DNA was then quantified by using a densitometer. For protein analysis, the same volume loaded onto an agarose gel was also subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were then transferred into a nitrocellulose membrane, and Western blot analysis was performed using rabbit anti-HA antibody and an enhanced chemiluminescence system.

Luciferase enzymatic activity.

Luciferase activity was determined according to a Promega protocol described for their luciferase reporter gene assay. Proteins were extracted according to the method of Luberto et al. (12); 20 μl of cell lysate was added to 100 μl of luciferase assay reagent (Promega), and the production of luciferase was immediately measured by using a Reporter Microplates luminometer (Turner Designs). The results were normalized per microgram of protein.

RESULTS AND DISCUSSION

C. neoformans Atf2 binds ATF consensus sequences in vitro.

In our previous studies, we showed that APP1 expression was negatively and positively regulated by AP-2 and ATF consensus sequences, respectively, located in the APP1 promoter. Here we sought to determine which transcription factor(s) is involved in binding these cis-acting elements and thus regulate APP1 transcription. To identify the potential transcription factor implicated in APP1 activation, the human AP-2 and ATF gene families were analyzed by using BLAST in the C. neoformans H99 Duke University Database. The only significant homology found was for the human ATF gene family. No homology was found for human AP-2. Thus, we decided to focus our investigation on the C. neoformans homolog of human ATF gene, which we named C. neoformans ATF2 (13). Using a 5′ and 3′ RACE (rapid amplification of cDNA ends) approach, we amplified and cloned the entire ATF2 cDNA fragment of ∼1.9 kb, which was used for Atf2 protein expression and purification. The yielded rAtf2 protein of about 75 kDa, produced in bacterial cells (Fig. 1), was used for further studies.

FIG. 1.

Production of Atf2 recombinant protein. (A) Coomassie blue stain of a polyacrylamide gel containing protein extracted from E. coli expressing rAtf2 or empty vector. The arrow indicates the putative rAtf2 protein. (B) Western blot analysis of a similar gel using anti-Xpress antibody showing rAtf2 protein.

First, we tested whether rAtf2 would bind to a short oligonucleotide of 20 bp containing ATF or AP-2 cis-acting elements by using an EMSA. As illustrated in Fig. 2A, partially purified extract from rAtf2-expressing Escherichia coli reacted with oligonucleotide containing ATF consensus sequence (lane 1), whereas partially purified extract from negative control (empty plasmid) did not react (lane 2), suggesting that rAtf2 is binding to the oligonucleotide. To confirm the specificity of the binding, a supershift assay was performed with anti-His antibody (Invitrogen), and competition assays were performed with cold wild-type ATF and ΔATF. As illustrated in Fig. 2B, a supershift was observed in the presence of anti-His antibody. A competition assay with cold ATF shows a significant decrease of binding, whereas no effect was observed when ΔATF was used (Fig. 2C). Thus, it can be concluded that Atf2 binds the ATF consensus sequence. In addition, partially purified extract from rAtf2-expressing E. coli did not react with oligonucleotide containing AP-2 consensus sequence (data not shown). To make sure that the His tag was not responsible for this interaction, a supershift was performed with rApp1::His (11) and the oligonucleotide containing ATF, and no binding or supershift was observed (data not shown).

FIG. 2.

EMSA results showing that rAtf2 binds to the ATF cis-acting element. (A) EMSA analysis of partially purified protein extract from E. coli expressing rAtf2 (lane 1) or partially purified protein extract from E. coli expressing empty vector (lane 2) with 20-bp oligonucleotides containing ATF. (B) Supershift assay with anti-His antibody. (C) Competition assay in which 300× cold ATF (∼561pmol) added to hot ATF (³²P-ATF) significantly decreases the binding of rAtf2 to ATF. Addition of 300× of cold ΔATF did not affect binding of rAtf2 to ATF. (D) EMSA analysis of rAtf2 using 256-bp APP1 promoter wild-type (APP1p-WT) and mutated forms, as indicated. (E) EMSA analysis of negative control proteins using 256-bp APP1 promoter wild-type (APP1p-WT) and mutated forms APP1p-ΔATF, APP1p-ΔAP2, and APP1p-ΔATF-ΔAP2, as indicated.

Next, we tested whether rAtf2 would bind to the APP1 promoter region (Fig. 2D). We used a fragment of 256 bp corresponding to the APP1 promoter (APP1p) region upstream from ATG start site of ATF2 gene. This region contains both ATF and AP-2 cis-acting elements. The 256-bp APP1p wild-type (APP1p-WT) fragment was produced by PCR as described in Materials and Methods. In addition, we created three mutated forms of the 256-bp APP1p fragments in which two nucleotides of ATF or AP-2 cis-acting elements were mutated (APP1p-ΔATF, APP1p-ΔAP-2, and APP1p-ΔATF-ΔAP-2) as described in Materials and Methods. As shown in Fig. 2D, rAtf2 bound to APP1p-WT (lane 1) and to APP1p-ΔAP-2 (lane 3) fragments, whereas it did not bind to APP1p-ΔATF (lane 2) or to APP1p-ΔATF-ΔAP-2 (lane 4) fragments. As a negative control, partially purified extract from E. coli expressing empty plasmid was tested using the same APP1p fragments (Fig. 2E), and no binding was observed. Taken together, these results show that Atf2 binds to ATF and not to AP-2 consensus sequences present in the APP1 promoter region.

APP1 promoter is the target of Atf2 in vivo.

Since Atf2 binds to the APP1 promoter in vitro, it became important to establish that this binding also occurs in vivo (leaving cells). Since, to our knowledge, antibody to C. neoformans Atf2 is not available, we epitope tagged the ATF2 gene with HA sequence. Because it was previously shown that expression of APP1 is regulated by Ipc1 activity (11), in addition to wild-type C. neoformans (IPC1/ATF2), the ATF2 gene was tagged also in the GAL7::IPC1, IPC1/APP1::LUC, and GAL7::IPC1/APP1::LUC strains to allow study of the effect of Ipc1 modulation on Atf2 expression and on the interaction of Atf2 to the APP1 promoter. The recipient strains were thus biolistically transformed with the pATF2::HA/HYG plasmid, as illustrated in Fig. 3A. Transformants in which plasmid integration was shown at the ATF2 locus were screened by Southern blot analysis (data not shown). One clone for each transformation showing the correct molecular mass shift at the Southern blot was selected, protein extracted, and assayed on a Western blot against anti-HA antibody. As shown in Fig. 3B, a single ∼77-kDa protein corresponding to the Atf2 molecular mass was detected. Thus, the corresponding strains were named IPC1/ATF2::HA, GAL7::IPC1/ATF2::HA, IPC1/APP1::LUC/ATF2::HA, and GAL7::IPC1/APP1::LUC/ATF2::HA, as indicated (Fig. 3B).

FIG. 3.

Tagging ATF2 gene in C. neoformans with HA epitope tag. (A) Diagram illustrating the plasmid DNA, the recipient strains used for biolistic transformation, and the resulting tagged strains. (B) Western blot analysis of one transformant for each transformation, selected upon screening by Southern analysis, using anti-HA antibody showing that ATF2 gene has been successfully tagged. A strain in which the plasmid DNA was not transformed (IPC1/APP1::LUC) was used as a negative control.

Next, we analyzed the recruitment of Atf2::HA on the APP1 promoter, using the strains carrying the tagged protein (IPC1/ATF2::HA), by ChIP. A pilot experiment was performed to determine at which PCR cycle the APP1 promoter would be identified (data not shown). As a negative control for immunoprecipitation, the IPC1 strain carrying the untagged ATF2 was used. As a negative amplicon control for the PCR, CTRL primers (Neg amplicon) amplifying a different gene were used. Chromatin was immunoprecipitated using anti-HA antibody, and the precipitated DNA was subjected to PCR using two primers targeting the APP1 promoter region, yielding a fragment of 256 bp, or the CTRL primer amplifying a region of ATF2 gene. After 26 cycles we were able to clearly detect the APP1 promoter in the sample in which the DNA was extracted and precipitated from IPC1/ATF2::HA (Fig. 4A, 26 cycles, lane 1) and not in the sample in which negative control amplicon was used (Fig. 4A, 26 cycles, lane 2). Importantly, no amplification using either APP1 or CTRL primers was observed when DNA extracted from IPC1/ATF2 strain was used because ATF2 gene is untagged in this strain (Fig. 4A, 26 cycles, lane 3 and 4). A graphical representation and semiquantitation of the DNA intensity obtained from the ChIP assay is illustrated in Fig. 4B. These results clearly show that Atf2 binds to APP1 promoter in vivo.

FIG. 4.

ChIP assay. (A) Electrophoretic gel stained with ethidium bromide showing the DNA fragment of ∼250 bp corresponding to the APP1 promoter (PCR 26 cycles, lane 1) after the immunoprecipitation (After IP) of DNA using anti-HA antibody. The use of a negative amplicon with control primers directed against a different gene did not show any DNA amplification (PCR 26 cycles, lane 2). Also, the APP1 promoter was not amplified when DNA extracted from the IPC1/ATF2 strain, in which ATF2 gene is untagged, was used for ChIP assay (Fig. 4A, 26 cycles, lanes 3 and 4). Importantly, APP1 promoter was amplified using DNA from IPC1/ATF2::HA (ATF2::HA) or IPC1/ATF2 (ATF2) before immunoprecipitation (Before IP), as indicated. (B) Semiquantitative analysis of the DNA intensity illustrated in panel A.

Atf2 binding to APP1 promoter is stimulated by DAG and Ipc1.

In our previous studies, we showed that DAG stimulates APP1 transcription in a dose- and time-dependent manner (13). We also showed that the regulation of APP1 transcription by DAG may be under the control of Atf2 because when we deleted ATF2 gene, activation of APP1 by DAG was lost. Thus, we wondered whether upon treatment with DAG the binding of Atf2 with the APP1 promoter would increase. To address this question, we treated the IPC1/APP1::LUC/ATF2::HA strain with 20 μM DAG and at 30, 60, and 90 min we measured the amount of Atf2 bound to the DNA by using an in vivo DPB precipitation assay. The amount of Atf2 protein visualized by Western blotting (Fig. 5B) was normalized with the amount of DNA that was precipitated (Fig. 5A). As a negative control, a sample with no DAG treatment was used. As illustrated in Fig. 5C, DAG treatment significantly increased the binding of Atf2 to DNA in a time-dependent manner. After 90 min of DAG exposure, the amount of Atf2 bound to the DNA is decreasing and will reach the base level after 3 h of treatment (data not shown). Activation of luciferase activity upon DAG treatment was observed in the IPC1/APP1::LUC/ATF2::HA strain (Fig. 5D). These results suggest that, upon DAG treatment, Atf2 binds to the APP1 promoter, which will then drive the transcription of APP1.

FIG. 5.

In vivo DPB precipitation assay upon DAG treatment. (A) DNA analysis extracted from IPC1/APP1::LUC/ATF2::HA strain untreated (0) treated with 20 μM DiC8 for 30, 60, or 90 min and precipitated with 100% ethanol. (B) Western blot analysis of the precipitated DNA with anti-HA antibody. (C) Semiquantitative analysis of the amount of Atf2 protein upon treatment with DAG detected in B normalized with 1 μg of precipitated DNA. (D) Luciferase activity of IPC1/APP1::LUC/ATF2::HA strain grown in glucose and treated with 20 μM DiC8 (DAG) for 30, 60, or 90 min.

In fungal cells, including C. neoformans, DAG is one of the products of the Ipc1 reaction (7, 8, 12). Thus, we wondered whether the binding of Atf2 to the DNA would increase upon the upregulation of Ipc1. To answer this question, the GAL7::IPC1/APP1::LUC/ATF2::HA strain was grown on glucose for 16 h, switched to galactose, and after 30, 60, 90, and 180 min we measured the amount of Atf2 bound to the DNA. At 180 and 360 min of growth in galactose, we also measured the luciferase activity. As illustrated in Fig. 6A, upregulation of Ipc1 increased Atf2 binding to the DNA at 30 and 60 min, and after 90 min of Ipc1 stimulation the binding returned to its basic level. As expected, luciferase activity increased at 180 and 360 min of Ipc1 stimulation. Interestingly, the delay between protein-DNA (Atf2-APP1p) binding and protein activity (luciferase) is confirmed in this experiment, supporting the hypothesis that APP1 transcription is regulated by a tight cascade of events under the control of Ipc1-DAG.

FIG. 6.

DPB precipitation assay upon Ipc1 upregulation. (A) Semiquantitative analysis of the amount of Atf2 protein obtained from the GAL7::IPC1/APP1::LUC/ATF2::HA strain grown in glucose (0) and switched to galactose medium for 30, 60, 90, or 180 min. (B) Luciferase activity of GAL7::IPC1/APP1::LUC/ATF2::HA strain grown in glucose (0) and switched to galactose medium for 180 or 360 min.

The mechanism by which DAG stimulates the binding of Atf2 to the APP1 promoter is still under investigation. One possibility is that DAG might stimulate the phosphorylation of Atf2 by Pkc1. This hypothesis is supported by our previous studies in which we showed that DAG activates Pkc1 in C. neoformans (7, 8) and that treatment with calphostin C, but not genestein, abolishes the activation of APP1 transcription (measured as luciferase activity) stimulated by DAG (13). This hypothesis is also supported by mammalian studies showing that CREB/ATF transcription factors are common substrates for protein kinases such as protein kinase C (10, 18). On the other hand, our in vitro studies with rAtf2, which is not phosphorylated, suggest that phosphorylation is not required for the binding of Atf2 to the APP1 promoter. Thus, whether Atf2 phosphorylation is present in vivo and whether it would regulate the interaction between Atf2 and the APP1 promoter region awaits further study.

In conclusion, we identified here the mechanism by which Atf2 activates APP1 transcription and validate that this regulation is under the control of DAG and Ipc1. The signaling events downstream Ipc1 are particularly important because this enzyme regulates key aspects of fungal pathogenesis, such as phagocytosis (11) and cell growth in acidic environments (12). Phagocytosis is under the regulation of Ipc1-DAG-Atf2-App1 pathway (11, 13), whereas growth at acidic pH is controlled by the production of complex sphingolipids initiated by the Ipc1 reaction (J. Garcia et al., unpublished data). Of interest, IPC1 mRNA has been found to be significantly increased in intracellular C. neoformans (within macrophages) compared to extracellular C. neoformans (6). Intracellularly, C. neoformans is mostly found within the phagolysosome of macrophages, which is characteristically an acidic environment. Thus, at the fungal-macrophage interface, Ipc1-DAG functions at two levels: it limits the internalization of fungal cells by macrophages, and it allows fungal survival once within. Considering that C. neoformans replicates faster intracellularly compared to extracellularly, these functions will have a significant biological effect on the overall fungal fitness and pathogenic trait of C. neoformans. Thus, studies addressing how these molecules (e.g., Ipc1, DAG, App1, and Atf2) would regulate these biological and pathological processes would produce significant insights into a better understanding of the molecular mechanisms of pathogenicity, with important implications for the development of new therapeutic strategies.