Isolation and Functional Analysis of a Gene, tcsB, Encoding a Transmembrane Hybrid-Type Histidine Kinase from Aspergillus nidulans

Abstract

We cloned and characterized a novel Aspergillus nidulans histidine kinase gene, tcsB, encoding a membrane-type two-component signaling protein homologous to the yeast osmosensor synthetic lethal N-end rule protein 1 (SLN1), which transmits signals through the high-osmolarity glycerol response 1 (HOG1) mitogen-activated protein kinase (MAPK) cascade in yeast cells in response to environmental osmotic stimuli. From an A. nidulans cDNA library, we isolated a positive clone containing a 3,210-bp open reading frame that encoded a putative protein consisting of 1,070 amino acids. The predicted tcsB protein (TcsB) has two probable transmembrane regions in its N-terminal half and has a high degree of structural similarity to yeast Sln1p, a transmembrane hybrid-type histidine kinase. Overexpression of the tcsB cDNA suppressed the lethality of a temperature-sensitive osmosensing-defective sln1-ts yeast mutant. However, tcsB cDNAs in which the conserved phosphorylation site His⁵⁵² residue or the phosphorelay site Asp⁹⁸⁹ residue had been replaced failed to complement the sln1-ts mutant. In addition, introduction of the tcsB cDNA into an sln1Δ sho1Δ yeast double mutant, which lacked two osmosensors, suppressed lethality in high-salinity media and activated the HOG1 MAPK. These results imply that TcsB functions as an osmosensor histidine kinase. We constructed an A. nidulans strain lacking the tcsB gene (tcsBΔ) and examined its phenotype. However, unexpectedly, the tcsBΔ strain did not exhibit a detectable phenotype for either hyphal development or morphology on standard or stress media. Our results suggest that A. nidulans has more complex and robust osmoregulatory systems than the yeast SLN1-HOG1 MAPK cascade.

Living cells are equipped with mechanisms for sensing environmental stimuli, such as osmotic stress, oxidative stress, and hormones, which allow adaptation to the stimuli through a variety of cellular responses triggered by the sensing and subsequent signaling systems. Two-component signaling systems, which involve a phosphorelay from the histidine of the sensor kinase to the aspartic acid of the response regulator, are widespread in bacteria (23). Regulatory proteins similar to bacterial systems are also found in eukaryotes, including plants (ETR1 [6], CKI1 [16], ATHK1 [32], and others [13, 33]), slime molds (dokA [29], dhkA [35], and dhkB [40]), filamentous fungi (NIK1 [1], os-1 [28], tcsA [34], and FOS-1 [25]), and yeasts (SLN1 [22], mak1⁺ [phk3⁺], mak2⁺ [phk1⁺], and mak3⁺ [phk2⁺] [3, 4], COS1 [2], CaHK1 [5], CaNIK1, and CaSLN1 [19]). Some of the histidine kinases, including yeast SLN1, plant ATHK1, and fungal NIK1 and tcsA, are involved in osmoregulation. In Saccharomyces cerevisiae, Sln1p consists of an extracellular sensor, a kinase, and a response regulator domain in a single polypeptide and is thus a transmembrane hybrid-type histidine kinase (22). Under low-osmolarity conditions, a specific histidine residue within the histidine kinase domain is autophosphorylated. The phosphate moiety of the histidine kinase is transferred to an aspartic acid residue within the response regulator domain and then via a phosphorelay is transferred to the downstream proteins Ypd1p and Ssk1p, shutting off the high-osmolarity glycerol response 1 (HOG1) mitogen-activated protein kinase (MAPK) cascade (24). Histidine kinase activity and phosphorylation of Sln1p are essential for growth at low osmolarity. Increased osmolarity downregulates the histidine kinase activity of Sln1p, and this downregulation in turn activates the downstream HOG1 MAPK cascade. Activation of the HOG1 MAPK cascade results in the induction of genes responsible for osmotic adaptation, such as the genes for glycerol-3-phosphate dehydrogenase (GPD1) and glycerol-3-phosphatase (GPP2) (8, 11).

Han and Prade described in silico reconstruction of the yeast HOG pathway in Aspergillus nidulans and cloned the hogA gene, which is the counterpart of S. cerevisiae HOG1, and they characterized its function in the hogA null mutant (9). However, membrane-localized osmosensor proteins that function upstream of the HOG pathway in filamentous fungi remain unidentified and uncharacterized. Involvement of other histidine kinases in two-component signaling systems has been reported for Neurospora crassa (NIK1 [1] and os-1 [28]), A. nidulans (tcsA [34]), and Aspergillus fumigatus (FOS-1 [25]), and the histidine kinases are thought to be involved in morphogenesis, conidium formation, or cell wall assembly. Three histidine kinases (Nik1, TcsA, and Fos-1), which lack plasma membrane-spanning regions, are thought to be localized in the cytoplasmic fraction (1, 25, 28, 34). If the HOG MAPK system (besides Nik1, TcsA, and Fos-1) is functional in filamentous fungi, osmoregulation in filamentous fungi seems to be controlled by a mechanism that is more complex than the mechanism observed in S. cerevisiae. It is important to understand the role of the fungal HOG MAPK system and the extent of its contribution to osmoregulation among the rest of the histidine kinases. In particular, the sensor molecule Sln1p, which is located at the entry of the HOG pathway, is essential in S. cerevisiae. Hence, for a comparative study with the yeast Sln1p-Hog1p system, we cloned the SLN1 homologue tcsB from A. nidulans, which encodes a novel two-component histidine kinase with membrane-spanning regions. To our knowledge, no cell surface osmosensor proteins, including transmembrane hybrid-type histidine kinases, have been isolated and characterized for filamentous fungi; this is the first report of a transmembrane hybrid-type histidine kinase found in filamentous fungi. We also obtained evidence, by using yeast mutants, that TcsB has a potential function as an osmosensor. Surprisingly, in vivo functional analysis of tcsB revealed that unlike S. cerevisiae sln1Δ, the A. nidulans tcsB null mutant (tcsBΔ) was viable under both regular and high osmotic conditions. We discuss the role of the two-component histidine kinase in osmoregulation in A. nidulans below.

MATERIALS AND METHODS

Strains, media, and growth conditions.

We used the A. nidulans biotin and arginine auxotroph FGSC A89 (biA1 argB2) for all genetic manipulations. This strain was grown in potato dextrose medium (Nissui, Tokyo, Japan) or Czapek Dox (CD) medium supplemented with 0.02 μg of biotin per ml and/or 200 μg of arginine per ml (20). A. nidulans FGSC A89 that had been transformed with the A. nidulans argB gene, called the wild-type strain here, was used as a control for phenotype analyses of the tcsB knockout derivatives. To test the responses of a tcsB disruption mutant (tcsBΔ) to various stress conditions, we grew the wild-type and tcsBΔ strains on CD agar containing one of the following compounds at 30 or 37°C for 3 days: NaCl (0.5, 1.0, or 1.5 M), KCl (0.5, 1.0, or 1.5 M), sorbitol (0.5, 1.0, or 1.5 M), sodium dodecyl sulfate (SDS) (0.001, 0.005, or 0.01%), H₂O₂ (0.001 or 0.005%), Congo red (Nacalai Tesque, Kyoto, Japan) (0.1, 2, or 20 μg/ml), calcofluor white (Nacalai Tesque) (0.1, 2, or 20 μg/ml), iprodione (Wako, Osaka, Japan) (0.1, 2, or 20 μg/ml), or fludioxonil (Wako) (0.1, 2, or 20 μg/ml). Hyphal growth rates were determined by dividing the radii (in millimeters) of 3-day-old colonies by the incubation time (in hours). Mycelia grown in CD liquid media containing osmotic solutes at 30 or 37°C for 3 days were also quantified by straining the mycelia through Advantec no. 2 filter paper and determining the wet weight. Conidiospores (1 × 10² conidiospores in 50 μl of CD medium) of one of the strains were deposited on microscope slides and incubated at 37°C for 16 h. Then the mycelial morphology was examined by using a Site Vision system attached to a DMRB microscope (Leica, Tokyo, Japan). We also examined the growth of both strains on wheat bran with a low moisture content. Conidiospores (2 × 10⁷ conidiospores in 2 ml of water) were inoculated into 100-ml flasks containing 2 g of sterilized wheat bran and incubated at 30°C for 3 days at a moisture level of 50%. We determined the effect of cell wall-degrading enzymes on protoplast formation for both strains by incubating hyphae (2 g [wet weight]) with 5 mg of Yatalase enzyme preparation (Takara, Tokyo, Japan) per ml in protoplast formation buffer (0.8 M NaCl, 10 mM NaH₂PO₄; pH 8.0) at 30°C for 3 h. The numbers of protoplasts were determined with a hemocytometer and an Olympus light microscope (magnification, ×400).

Isolation of A. nidulans genomic DNA and RNA.

Conidia from A. nidulans colonies on CD agar plates were harvested in 5 ml of 0.01% Tween 20 and used to inoculate 400 ml of YPD medium (1% yeast extract, 2% Polypeptone, 2% glucose). After incubation at 37°C for 2 days, mycelia were harvested by filtration and washed with 2 liters of water. Mycelia (5 g [wet weight]) were frozen in liquid nitrogen and ground to a fine powder in a mortar. The ground mycelia were transferred to a 50-ml tube, and 6 ml of RNA extraction solution (5 M guanidine isothiocyanate, 10 mM EDTA, 50 mM Tris-HCl; pH 7.5) and 1.2 ml of 2-mercaptoethanol were added and vigorously mixed with a vortex mixer. We added 30 ml of 4 M LiCl and incubated the mixture at 25°C for 20 min. The sample was homogenized by passing it through an injection needle (19 gauge), and debris was removed by centrifugation at 600 × g for 5 min at 4°C. The supernatant was then centrifuged at 10,000 × g for 90 min at 4°C. The resulting pellet was suspended in 20 ml of 3 M LiCl and homogenized by passage through an injection needle (21 gauge). After centrifugation at 10,000× g for 60 min at 4°C, the pellet was resuspended in 5 ml of TESDS (10 mM Tris-HCl [pH 7.5], 5 mM EDTA, 0.1% SDS) by passage through an injection needle (21 gauge). The total RNA in the sample was purified by two successive extractions with 5 ml of phenol-chloroform-isoamyl alcohol (25:24:1) and by ethanol precipitation. The resulting total RNA was resuspended in RNase-free water (3.2 μg/μl) and stored at −80°C until it was used for mRNA preparation.

A. nidulans genomic DNA was isolated from mycelia grown in YPD medium at 37°C for 2 days with a Qiagen DNeasy plant extraction kit (Qiagen, Tokyo, Japan) used according to the manufacturer's instructions (20).

Molecular cloning and sequencing of tcsB gene.

All basic molecular biology procedures were carried out as described by Sambrook and Russell (27). We used the BLAST network service (Blast2) to search the A. nidulans expressed sequence tag database (http://www.genome.ou.edu/fungal.html) for histidine kinase genes homologous to NIK1 of N. crassa or SLN1 of S. cerevisiae. We found three expressed sequence tags, g7e07a1.r1, c5e08a1.r1, and g2e07a1.r1, which contained conserved regions for the H-box, N-box, and response regulator domain, respectively. Fragments containing these sequences were amplified by PCR (95°C for 1 min, 55°C for 1 min, and 72°C for 1 min for 30 cycles) by using A. nidulans genomic DNA and the following three primer sets: H-box primers 5′-CGTCTGCGCGCTGCATCACC-3′ and 5′-GACCGTGTTGACGAGCGCCG-3′, N-box primers 5′-GTCATCCAGACACTCCTCCG-3′ and 5′-GCGGTTTTGCAGGTGCTGCG-3′, and response regulator domain primers 5′-CCCGGCTCATCTACTGATGC-3′ and 5′-CCATGGCTCGAATCTCCCGC-3′. After gel purification, the amplified fragments were subcloned into the pGEM-T Easy vector (Promega, Tokyo, Japan), and DNA sequences of the cloned fragments were determined by using the dye-primer cycle sequencing method and a DNA sequencer (model 373A; Applied Biosystems, San Jose, Calif.). The insert sizes of the cloned fragments were 314 bp for the H-box, 437 bp for the N-box, and 303 bp for the response regulator domain. The resulting plasmids were digested with EcoRI, and the fragments were used as probes for screening histidine kinase genes in an A. nidulans cDNA library, which was prepared as described below.

mRNA (8.4 μg) was purified from the total RNA (1 mg) described above with a μMACS mRNA isolation kit (Miltenyi Biotec, Auburn, Calif.) used according to the manufacturer's instructions. An A. nidulans cDNA library was prepared by using a ZAP-cDNA library kit (Stratagene, Tokyo, Japan) according to the manufacturer's instructions. To isolate full-length cDNAs of histidine kinase genes, we screened the cDNA library using plaque hybridization. Positive plaques were purified, and the cDNA inserts were subcloned into a pBluescript SK(−) phagemid (Stratagene, La Jolla, Calif.) by using an in vivo excision process according to the phagemid manufacturer's instructions. One positive clone, pBSK511, contained a 3,324-bp insert, and the cDNA sequence was determined on both strands with an ABI Prism BigDye terminator cycle sequencing Ready Reaction kit (PE Applied Biosystems, Chiba, Japan) by automated DNA sequencing with an ABI Prism 377 DNA sequencer (PE Applied Biosystems).

The 5′ flanking region of the tcsB gene was cloned by inverse PCR (27, 31). Because a PstI site was found 780 bp downstream of the predicted translation initiation site of tcsB, A. nidulans genomic DNA digested with PstI was separated on an agarose gel and then subjected to Southern analysis with a PstI fragment (728 bp) derived from pBSK511 as a probe, which gave a 1.8-kb positive band. DNA extracted from the portion of the gel in which the positive signal was observed was circularized by ligation and used as the template for PCR amplification (95°C for 1 min, 54.5°C for 1 min, and 72°C for 2 min for 30 cycles) with primers 5′-CCAGCAAGGGCCGTCAAGAG-3′ and 5′-CAGCTGCGGATATTCTGGGC-3′. A 1.2-kb fragment was amplified and subcloned into the cloning vector pGEM-T Easy, resulting in plasmid pGIPtcsB. Three independent clones were sequenced.

Complementation analysis of tcsB and its derivatives in a temperature-sensitive sln1 mutant.

Yeast (S. cerevisiae) strain YHS-13 (MATa ura3 leu2 trp1 sln1-ts4) with a temperature-sensitive SLN1 allele (17) was used for complementation analysis. The expression plasmids used in this experiment were constructed with the expression vector pYES2 (Invitrogen, Tokyo, Japan), in which expression is under control of the galactose-inducible GAL1 promoter (15). A fragment containing the complete open reading frame (ORF) of the tcsB cDNA was excised from pBSK511 by digestion with SpeI and XhoI. dATP was added to the terminal end of the SpeI-XhoI fragment with Taq polymerase (Takara), and the fragment was subcloned into pGEM-T Easy, resulting in pGEMtcsB. pGEMtcsB was digested with NotI, and the NotI fragment containing tcsB cDNA was ligated into the NotI site of pYES2, resulting in the tcsB expression vector pYEStcsB.

To construct the catalytically inactive mutants pYEStcsB-H552Q and pYEStcsB-D989N, nucleotide substitutions for pYEStcsB were generated by using a QuikChange site-directed mutagenesis kit (Stratagene) (95°C for 0.5 min for one cycle; 95°C for 0.5 min, 54°C for 1 min, and 68°C for 18 min for 18 cycles; and 68°C for 10 min for one cycle) with primers 5′-CGCCAATATTTCTcagGAGCTCAAAAC-3′ and 5′-GTTTTGAGCTCctgAGAAATATTGGCG-3′ for H552Q (His [CAT] replaced by Gln [CAG]) and primers 5′-GATTTTTATGaatATTCAGATGCC-3′ and 5′-GGCATCTGAATattCATAAAAATC-3′ for D989N (Asp [GAT] replaced by Asn [AAT]) (lowercase letters indicate mutations). To make a transmembrane region deletion mutant, a HindIII site and a new translation initiation codon were generated just before the Ile⁴⁹³ site of the tcsB gene in pYEStcsB by using a QuikChange site-directed mutagenesis kit (95°C for 0.5 min for one cycle; 95°C for 0.5 min, 54°C for 1 min, and 68°C for 18 min for 18 cycles; and 68°C for 10 min for one cycle) with primers 5′-GGAGCGTAAGCTTatgATTACGGATGAAC-3′ and 5′-GTTCATCCGTAATcatAAGCTTACGCTGG-3′ (the HindIII sites are underlined, and the new start codons are indicated by lowercase letters), and the resultant plasmid was digested with HindIII to excise the transmembrane regions in the N-terminal half and religated, resulting in pYEStcsBΔTM (deletion of Arg²-Ile⁴⁹² in TcsB). Each mutation was confirmed by DNA sequencing.

pRS-SLN1 was used for constitutive expression of yeast SLN1 as the positive control (17, 32).

Plasmids containing tcsB cDNA, its derivatives, and yeast SLN1 were transformed into S. cerevisiae YHS-13 by using a lithium acetate method (14), and complementation analysis with the transformants was performed on YPD medium or YPG medium (in which the glucose in YPD medium was replaced with 2% galactose) at 37°C (restrictive conditions) or 24°C for 3 days.

Complementation analysis of tcsB and its derivatives in an sln1 sho1 deletion mutant (sln1Δ sho1Δ).

An S. cerevisiae sln1 sho1 double-deletion mutant (sln1Δ sho1Δ) was used for complementation analysis (32). This mutant harbors the expression plasmid for PTP2, which encodes tyrosine phosphatase under control of the GAL1 promoter, to prevent constitutive activation of Hog1p (17). The expression plasmids used in this experiment were constructed with the multicopy expression vector YEpGAP, in which expression is under the control of the constitutive GAP promoter (26). Plasmid YEpGAPtcsB was constructed by insertion of an SpeI-XhoI fragment containing the complete ORF from pBSK511 into the SpeI-XhoI sites of YEpGAP. To construct the catalytically inactive mutants YEpGAPtcsB-H552Q and YEpGAPtcsB-D989N, nucleotide substitutions for YEpGAPtcsB were generated by using a QuikChange site-directed mutagenesis kit under the conditions that were used to generate pYEStcsB-H552Q and pYEStcsB-D989N. sln1Δ sho1Δ was then transformed with YEpGAPtcsB, YEpGAPtcsB-H552Q, YEpGAPtcsB-D989N, or pRS-SLN1. Wild-type strain TM141 (MATα ura3 leu2 trp1 his3) and a hog1 deletion mutant (hog1Δ) were used as positive and negative controls, respectively. These yeast cells were cultured on YPD medium or YPD medium containing 0.9 M NaCl at 30°C for 3 days.

Construction of a tcsB::argB gene disruptant of A. nidulans.

A 3′ flanking region fragment of tcsB was obtained from pBSK511K by digestion with KpnI and ApaI. To introduce a new KpnI site, pBSK511K was generated from pBSK511 by using a QuikChange site-directed mutagenesis kit (95°C for 0.5 min for one cycle; 95°C for 0.5 min, 50°C for 1 min, and 68°C for 12 min for 18 cycles; and 68°C for 10 min for one cycle) with primers 5′-AATCCCACAATTGGtaCCTTAAGTCCGTCG-3′ and 5′-CGACGGACTTAAGGtaCCAATTGTGGGATT-3′ (lowercase letters indicate mutations). A fragment of the 5′ flanking region of tcsB was obtained from pGIPtcsB by digestion with PstI and EcoRI. A fragment of the Aspergillus oryzae argB gene, which complements the argB2 mutation in A. nidulans, was obtained from pAORB (containing the argB gene; kindly provided by K. Gomi) by digestion with EcoRI and KpnI. Then the three fragments were simultaneously ligated into the PstI and ApaI sites of pSL1180 (Amersham Biosciences, Tokyo, Japan), resulting in pSLtcsB::argB. A. nidulans FGSC A89 (argB2) was transformed by the protoplast method (7) with the linear form of pSLtcsB::argB by ApaI digestion. Arginine prototrophs were obtained as candidates of A. nidulans tcsB::argB (tcsBΔ). To verify that modification of the tcsB locus occurred, five arginine prototrophs, which gave a 3.4-kb fragment for the targeted tcsB locus and a 1.8-kb fragment for the authentic tcsB locus, were analyzed by PCR by using their genomic DNAs as templates and primers 5′-ATATGAACGGGCGGCAATCGGTTTCCAATG-3′ and 5′-AAGAATCCCATTCAGCGGCGTTTTGAGCTC-3′. Two transformants in the five prototrophs indicated the PCR pattern for tcsBΔ. For further confirmation, genomic DNAs from the two candidates were digested with FspI and subjected to Southern blot analysis. When a probe derived from pBSK511 by digestion with NspV was used, genomic DNA from parental strain FGSC A89 produced a single 2.6-kb hybridized band originating from the authentic tcsB gene, and DNAs from the tcsBΔ candidates produced only the 4.1-kb band originating from the modified tcsB locus instead of the 2.6-kb band, as expected.

Nucleotide sequence accession number.

The nucleotide sequence of tcsB, previously called NHK1 by us, has been deposited in the DDBJ/EMBL/GenBank nucleotide database under accession number AB036054. Because another histidine kinase gene, tcsA, has been reported (34), we renamed NHK1 tcsB based on the nomenclature of A. nidulans genes.

RESULTS

tcsB encodes an A. nidulans hybrid-type histidine kinase.

From the A. nidulans cDNA library, we isolated a positive clone, pBSK511, with the longest insert (approximately 3.3 kb) after excision. The insert contains a 3,210-bp ORF encoding a single polypeptide consisting of 1,070 amino acid residues and having a predicted molecular mass of 117,573 Da. We also cloned its 5′ flanking region that is −1,130 bp upstream from the predicted translation initiation codon by inverse PCR (data not shown). Searches of current nucleotide and protein databases with the BLAST network service revealed that the sequence was unique. Thus, the gene was designated tcsB (two-component signaling protein B); this gene was previously deposited in databases as NHK1 (nidulans histidine kinase 1). The putative protein, TcsB, contains the following typical histidine kinase conserved sequences: an H-box with a conserved histidine (autophosphorylation site), an N-box (nucleotide binding), and ATP-binding or ATP recognition motifs (G1-, F-, and G2-boxes) (Fig. 1). In general, eukaryotic histidine kinases can be classified into the following two types on the basis of their localization (36): (i) transmembrane-type histidine kinases (Sln1p, CaSln1, DhkA, Athk1, Cki1, Etr1, and Ers1 [12]), which have two or three hydrophobic transmembrane domains in the N-terminal half (although Sln1p, CaSln1, DhkA, Athk1, and Cki1 contain a putative extracellular domain, Etr1 and Ers1 do not); and (ii) cytoplasm-type histidine kinases (Nik1 [Os1], CaNik1 [Cos1], CaHk1, TcsA, Fos-1, Mak1 [Phk3], Mak2 [Phk1], and Mak3 [Phk2]), which have no transmembrane domains (these kinases are thought to be localized in the cytoplasm). The cytoplasmic histidine kinases differ greatly in the structures of their N-terminal halves. Nik1 and CaNik1 contain six and five tandem repeats of 90 amino acids, respectively. TcsA, Fos-1, Mak1, Mak2, and Mak3 contain a PAS/PAC domain. The PAS domain is a signaling module found in a large number of proteins involved in sensing light, redox potential, oxygen, small ligands, and energy levels of cells (30, 39). CaHk1, Mak2, and Mak3 contain a serine/threonine kinase domain and a GAF domain. The functions of the domains are little understood, although involvement of the tandem repeat domain of Nik1 in osmoregulation has been genetically demonstrated (21). Hydrophobicity analysis with TMpred (10) revealed that TcsB has two putative transmembrane domains in the N-terminal half (data not shown). Therefore, we suggest that TcsB is a transmembrane-type histidine kinase. As expected, alignment of the amino acid sequences of TcsB and Sln1p showed a high degree of identity in the conserved motifs and a moderate degree of similarity in the transmembrane segments (Fig. 1).

FIG. 1.

Comparison of the deduced amino acid sequences of A. nidulans TcsB and S. cerevisiae Sln1p. Identical amino acids in TcsB (DDBJ/EMBL/GenBank accession no. AB036054) and Sln1p (DDBJ/EMBL/GenBank accession no. U01835) are indicated by asterisks. Probable transmembrane domains (TM) and conserved motifs (H-, N-, G1-, F-, and G2-boxes) are underlined. The potential phosphorylation sites of TcsB, His⁵⁵² and Asp⁹⁸⁹, are indicated by boldface type.

TcsB acts as a histidine kinase in yeast.

Because TcsB is structurally related to the yeast osmosensor Sln1p, we investigated the in vivo function of TcsB by using yeast sln1 mutants. We introduced the tcsB cDNA into the temperature-sensitive sln1 mutant (sln1-ts) under control of the galactose-inducible GAL1 promoter to examine the catalytic activity of TcsB. The sln1-ts mutation of yeast is lethal because of constitutive activation of the HOG1 MAPK cascade at a restrictive temperature, 37°C (17). Overexpression of the tcsB cDNA suppressed the phenotype of the sln1-ts mutant in the presence of galactose (Fig. 2, spot TcsB). By contrast, cDNAs in which one of the putative phosphorylation sites had been altered (His⁵⁵² [TcsB-H552Q; His⁵⁵² to Gln] or Asp⁹⁸⁹ [TcsB-D989N; Asp⁹⁸⁹ to Asn]) did not complement the sln1-ts mutation (Fig. 2). These results suggest that TcsB acts as a histidine kinase in yeast. Similarly, mutant TcsBΔTM, whose transmembrane domains were deleted, did not complement the sln1-ts mutation. We concluded that TcsB contains the putative transmembrane and extracellular domains that are important for activation of TcsB.

FIG. 2.

Complementation of sln1-ts mutation by expression of the tcsB cDNA and its derivatives. sln1-ts mutants harboring the following plasmids were cultured on plates supplemented with 2% glucose or 2% galactose at 37 or 24°C for 3 days: pYES2 (Vector), pYEStcsB (tcsB wild type) (TcsB), pYEStcsB-H552Q (TcsB-H552Q), pYEStcsB-D989N (TcsB-D989N), pYEStcsBΔTM (transmembrane region deletion) (TcsBΔTM), and pRS-SLN1 (SLN1 wild type) (Sln1p).

TcsB confers high-osmolarity tolerance to the sln1Δ sho1Δ yeast double mutant.

S. cerevisiae contains the following two osmosensors (18, 36): synthetic high-osmolarity-sensitive protein 1 (Sho1p), a protein that contains transmembrane domains, and Sln1p. Mutants lacking both Sln1p and Sho1p (sln1Δ sho1Δ) are lethal under high-osmolarity conditions (32). If TcsB expressed in yeast is functional and thus becomes inactive in response to increases in osmolarity, the sln1Δ sho1Δ mutants should be able to grow by reactivation of the HOG1 MAPK cascade. To analyze the functionality of TcsB as an osmosensor in yeast, we introduced the tcsB cDNA into the sln1Δ sho1Δ mutant and then examined its viability on high-osmolarity medium. As shown in Fig. 3, the sln1Δ sho1Δ double mutant expressing either TcsB (sln1Δ sho1Δ TcsB) or Sln1p (sln1Δ sho1Δ Sln1p) grew as well as the wild-type strain on YPD medium containing 0.9 M NaCl, but the hog1 deletion mutant (hog1Δ) showed reduced growth under these conditions. However, introduction of TcsB-H552Q or TcsB-D989N did not confer high-osmolarity tolerance to the sln1Δ sho1Δ double mutant. This suggests that the catalytic activity of TcsB is required for growth under saline conditions. These results imply that the histidine kinase activity of TcsB was inactivated in response to increases in external osmolarity, and consequently the HOG1 MAPK cascade was activated.

FIG. 3.

Sensitivity to high osmolarity of the sln1Δ sho1Δ double mutants transformed with tcsB cDNA. sln1Δ sho1Δ double mutants harboring the following plasmids were cultured on YPD or YPD medium containing 0.9 M NaCl at 30°C for 3 days: YEpGAP (Vector), pRS-SLN1 (SLN1 wild type) (Sln1p), YEpGAPtcsB (tcsB wild type) (TcsB), YEpGAPtcsB-H552Q (TcsB-H552Q), and YEpGAPtcsB-D989N (TcsB-D989N). A wild-type strain (TM141) and a hog1Δ strain were used as positive and negative controls, respectively.

tcsB is not essential in A. nidulans.

To investigate the in vivo function of tcsB in A. nidulans, we next constructed A. nidulans tcsBΔ strains in which part of the native tcsB gene was replaced by the A. oryzae argB selectable marker, which complements the argB2 mutation in A. nidulans. One-third of the tcsB coding region was deleted from the N terminus, and the deletion construct was confirmed by PCR and Southern hybridization analysis (Fig. 4). Reverse transcriptase PCR demonstrated that tcsB was not transcribed in the tcsBΔ strain (data not shown). Then we compared the phenotypes of the tcsBΔ strain under a variety of stress conditions with those of the wild-type strain. The stress conditions used included high osmolarity, the presence of SDS in the medium, oxidative stress, and the presence of fungicides in the medium. The tcsBΔ strain unexpectedly showed neither enhanced sensitivity nor resistance to any of the conditions examined, as monitored by the hyphal growth rates on agar plates (Table 1 and data not shown). The mycelial biomass of the tcsBΔ strain grown in liquid culture under high-osmolarity conditions was almost the same as that of the wild-type strain (data not shown). No difference in mycelial morphology of the wild-type and tcsBΔ strains in liquid culture was observed (data not shown). Furthermore, to investigate the involvement of tcsB in sensing environmental humidity, we compared the growth of the tcsBΔ strain on wheat bran with that of the wild-type strain. However, the tcsBΔ strain grew as well as the wild-type strain (data not shown), and no recognizable difference in phenotypes was observed for the two strains. A fos-1 null mutant of A. fumigatus is known to be more resistant to cell wall-degrading enzymes than the wild-type strain, suggesting that the cell wall compositions and/or structures of wild-type and fos-1 mutant strains are not identical (25). To determine whether TcsB is also involved in cell wall regulation, we tested the effect of cell wall-degrading enzymes on the wild-type and tcsBΔ hyphae, but the numbers of protoplasts generated from the two strains were almost the same (data not shown). These in vivo functional analyses of tcsB suggest that tcsB is not essential in A. nidulans, which is significantly different from the findings obtained for S. cerevisiae SLN1 and other fungal histidine kinases.

FIG. 4.

Generation of tcsB gene disruption. (A) Strategy for disrupting tcsB by replacing one-third of the tcsB coding region with the A. oryzae argB gene. The two arrows indicate the tcsB-specific primers. The gray box indicates the upstream region of tcsB. (B) Agarose gel electrophoresis of tcsB-specific PCR products after amplification of genomic DNA from the wild-type strain and arginine prototrophic transformants. The wild-type strain gave a 1.8-kb product (lane WT). Candidates of tcsB knockout strains gave only a 3.4-kb product (lanes 1 and 3). Transformants, which possessed ectopic integration, gave 1.8- and 3.4-kb products originating from the authentic and ectopic loci, respectively (lanes 2, 4, and 5). (C) Southern hybridization analysis of the candidates of knockout transformants. Each lane contained 20 μg of FspI-digested genomic DNA of the wild-type strain or a transformant (lanes as in panel B). Hybridization with the NspV fragment of pBSK511 showed the expected hybridization at 2.6 kb for the wild type (lane WT) and at 4.1 kb for the disruptants (lanes 1 and 3).

TABLE 1.

Growth rates of the wild-type and tcsBΔ strains in high-solute media^a

Solute	Concn (M)	Hyphal growth rate (mm/h)b
		30°C		37°C
		Wild type	tcsBΔ	Wild type	tcsBΔ
None		0.208 ± 0	0.202 ± 0.006	0.340 ± 0.011	0.338 ± 0.013
NaCl	0.5	0.194 ± 0.006	0.197 ± 0.006	0.340 ± 0.007	0.338 ± 0.011
	1.0	0.171 ± 0.006	0.175 ± 0.007	0.239 ± 0.006	0.243 ± 0.009
	1.5	0.139 ± 0.006	0.131 ± 0.009	0.168 ± 0.007	0.168 ± 0.007
Sorbitol	0.5	0.207 ± 0.007	0.204 ± 0.006	0.413 ± 0.015	0.403 ± 0.022
	1.0	0.199 ± 0.006	0.193 ± 0.010	0.356 ± 0.021	0.357 ± 0.016
	1.5	0.197 ± 0.006	0.197 ± 0.006	0.276 ± 0.010	0.264 ± 0.012

^aWild-type and tesBΔ conitiospores (10⁵ conidiospores) were inoculated onto CD agar plates with or without osmolants. The plates were incubated at 30 or 37°C for 72 h.

^bThe growth rates were calculated by dividing the radius of each colony by the culture time (72 h). Hyphal growth rates were determined from 10 individual data points; the values are means ± standard deviations.

DISCUSSION

We isolated the tcsB gene, which encodes a transmembrane hybrid-type histidine kinase, from A. nidulans and suspected that it is a homologue of S. cerevisiae Sln1p on the basis of the deduced amino acid sequences (Fig. 1). This was clearly confirmed by in vivo functional complementation experiments performed with both S. cerevisiae sln1-ts and sln1Δ sho1Δ double mutants (Fig. 2 and 3). The importance of the histidine-aspartate phosphorelay was demonstrated with H552Q and D989N mutants (Fig. 2 and 3). Transmembrane segments were also required for functionality of TcsB in the yeast system.

In S. cerevisiae, deletion of SLN1 is lethal, which is attributable to constitutive activation of the HOG1 MAPK cascade by the deletion (17), because Sln1p is predicted to act as a negative regulator of the HOG1 MAPK cascade. The dimorphic pathogenic fungus Candida albicans possesses two histidine kinases: CaSln1, which is an Sln1p homologue, and CaNik1. A casln1 null mutant of C. albicans showed several phenotypes, including growth retardation and morphological change, in the presence of 1.5 M NaCl (19). Nik1 is a cytoplasmic histidine kinase involved in osmoregulation in N. crassa, and a nik1 null mutant is known to be hypersensitive to NaCl (>4%) (21). tcsA was the first two-component signaling histidine kinase gene discovered in A. nidulans, and TcsA should be one of the cytoplasmic histidine kinases containing the PAS domain (34). The tcsAΔ disruption mutant produced a detectable defect in either sporulation or morphology on standard growth media, and the tcsAΔ phenotype was suppressed by growth on 1 M sorbitol, suggesting that tcsA is involved in both osmoregulation and morphogenesis (34).

Since the histidine kinase mutants of other fungi showed recognizable phenotypes, we constructed a tcsBΔ strain of A. nidulans to investigate the in vivo function of tcsB through analysis of its phenotype. However, surprisingly, the tcsBΔ strain did not show any recognizable morphological change on standard or stress media (data not shown). Furthermore, the growth rate of the tcsBΔ strain on CD agar plates (Table 1) or in CD liquid medium (data not shown) with or without osmotic stress (NaCl or sorbitol) was almost the same as that of the wild-type strain. The normal growth phenotype of the A. nidulans tcsBΔ strain suggests that some other systems suppress the tcsBΔ mutation.

The presence of a HogA MAPK cascade in A. nidulans, which would be the counterpart of the S. cerevisiae HOG1 MAPK cascade, was predicted by the in silico method (9), and transcriptional responses of several genes to osmotic stress through the HogA MAPK cascade were reported by Han and Prade (9). Although some histidine kinase genes have not yet been assigned in A. nidulans, this fungus seems to have another histidine kinase, AnNIK1 (partially sequenced), which is another homologue of N. crassa NIK1 (2) along with TcsA and TcsB. Recently, the N. crassa os-2 gene was identified as a homologue of S. cerevisiae HOG1, and both os-2 and nik1 mutants were sensitive to high osmolarity and resistant to phenylpyrrole fungicides (38). These findings suggest that Nik1 might regulate the N. crassa MAPK cascade, consisting of os-2 MAPK, in response to osmotic stress. This might imply that two histidine kinases, TcsB and AnNik1, regulate the same HogA MAPK cascade in A. nidulans.

Since A. nidulans possesses at least two other histidine kinases, TcsA and AnNik1, besides TcsB, the normal phenotype of the tcsBΔ mutant even in the presence of osmotic stress might be attributable to suppression of the tcsBΔ mutation by TcsA and/or AnNik1, which prevents constitutive activation of the HogA MAPK cascade. Multiple histidine kinases (including TcsA, TcsB, and AnNik1) might organize a more complex and robust osmoregulatory system in A. nidulans than the yeast Sln1p-Hog1p system. It would be interesting to study whether the three histidine kinases act cooperatively or independently in osmoregulation (and/or morphogenesis) in A. nidulans and to identify the cellular components, either downstream or upstream, in the phosphorelay system. In C. albicans, double deletion of two histidine kinase genes, CaSLN1 and CaNIK1, is thought to be lethal, and single disruptions of the two genes produced distinguishable phenotypes (37). Multiple disruptions of the histidine kinase genes might also be lethal or produce distinguishable phenotypes in A. nidulans. Construction of double mutants such as tcsAΔ tcsBΔ, tcsAΔ AnNIK1Δ, and AnNIK1Δ tcsBΔ will be important for understanding the organization of the multiple kinases.