Differential Sensitivity between Fks1p and Fks2p against a Novel β-1,3-Glucan Synthase Inhibitor, Aerothricin1*

Fks1p and Fks2p are catalytic subunits of β-1,3-glucan synthase, which synthesize β-1,3-glucan, a main component of the cell wall in Saccharomyces cerevisiae. Although Fks1p and Fks2p are highly homologous, sharing 88.1% identity, it has been shown that Fks2p is more sensitive than Fks1p to one of echinocandin derivatives, which inhibits β-1,3-glucan synthase activity. Here we show a similar differential sensitivity between Fks1p and Fks2p to a novel β-1,3-glucan synthase inhibitor, aerothricin1. To investigate the molecular mechanism of this differential sensitivity, we constructed a series of chimeric genes ofFKSs and examined their sensitivity to aerothricin1. As a result, it was shown that a region around the fourth extracellular domain of Fks2p, containing 10 different amino acid residues from those of Fks1p, provided Fks1p aerothricin1 sensitivity when the region was replaced with a corresponding region of Fks1p. In order to identify essential amino acid residues responsible for the sensitivity, each of the 10 non-conserved amino acids of Fks1p was substituted into the corresponding amino acid of Fks2p by site-directed mutagenesis. Surprisingly, only one amino acid substitution of Fks1p (K1336I) conferred Fks1p hypersensitivity to aerothricin1. On the other hand, reverse substitution of the corresponding amino acid of Fks2p (I1355K) resulted in loss of hypersensitivity to aerothricin1. These results suggest that the 1355th isoleucine of Fks2p plays a key role in aerothricin1 sensitivity.

Fks1p and Fks2p are catalytic subunits of ␤-1,3-glucan synthase, which synthesize ␤-1,3-glucan, a main component of the cell wall in Saccharomyces cerevisiae. Although Fks1p and Fks2p are highly homologous, sharing 88.1% identity, it has been shown that Fks2p is more sensitive than Fks1p to one of echinocandin derivatives, which inhibits ␤-1,3-glucan synthase activity. Here we show a similar differential sensitivity between Fks1p and Fks2p to a novel ␤-1,3-glucan synthase inhibitor, aerothricin1. To investigate the molecular mechanism of this differential sensitivity, we constructed a series of chimeric genes of FKSs and examined their sensitivity to aerothricin1. As a result, it was shown that a region around the fourth extracellular domain of Fks2p, containing 10 different amino acid residues from those of Fks1p, provided Fks1p aerothricin1 sensitivity when the region was replaced with a corresponding region of Fks1p. In order to identify essential amino acid residues responsible for the sensitivity, each of the 10 non-conserved amino acids of Fks1p was substituted into the corresponding amino acid of Fks2p by site-directed mutagenesis. Surprisingly, only one amino acid substitution of Fks1p (K1336I) conferred Fks1p hypersensitivity to aerothricin1. On the other hand, reverse substitution of the corresponding amino acid of Fks2p (I1355K) resulted in loss of hypersensitivity to aerothricin1. These results suggest that the 1355th isoleucine of Fks2p plays a key role in aerothricin1 sensitivity.
The fungal cell wall consists mainly of ␤-D-glucans, mannoproteins, a small amount of chitin, and several proteins, all of which are interconnected, providing cells their rigidity and protecting them from osmotic pressure (1,2). As the fungal cell wall is one of the essential architectures for fungal growth and, as mammalian cells do not have such architecture, enzymes that synthesize, assemble, retain, and remodel the fungal cell wall have been thought to be promising targets for antifungal agents (2)(3)(4).
␤-1,3-Glucan is the most abundant component in the fungal cell wall (2) and is synthesized by ␤-1,3-glucan synthase (UDPglucose:1,3-␤-D-glucan 3-␤-D-glucosyltransferase; EC 2.4.1.34). In Saccharomyces cerevisiae, two kinds of catalytic subunits are encoded by FKS1/GSC1/CWH53/ETG1/CND1/PBR1/ YLR342W and FKS2/GSC2/G4074/YGRO32W (5)(6)(7)(8)(9). They are highly homologous at the amino acid sequence level, showing 88% identity. Although disruption of either gene alone does not express lethal phenotype, simultaneous disruption of both genes provokes synthetic lethality to the yeast cells (8,10). These suggest that Fks1p and Fks2p share the function, which is essential for growth. On the other hand, transcriptionally, it is known that their expression is differently controlled; the FKS1 expression is regulated in the cell cycle and predominates during growth on glucose, whereas FKS2 is expressed in the absence of glucose (10). In Candida albicans (11) and Cryptococcus neoformans (12), only one gene encoding the catalytic subunit has been isolated, and it is believed that the genes are essential for their growth because of the lack of success in the establishment of their null mutants. From other fungi, each single gene encoding the catalytic subunit of ␤-1,3-glucan synthase has been isolated, such as Aspergillus nidulans (13), Aspergillus fumigatus, 1 and Paracoccidioides brasiliensis (15). Catalytic subunits from these fungi share the same features, a size greater than 200 kDa and possession of putative 16 transmembrane domains. In addition, Rho1p, a small GTP-binding protein, is known as a regulatory subunit of the ␤-1,3-glucan synthase in S. cerevisiae (16 -18) and C. albicans (19).
Here we present a differential sensitivity against aerothri-cin1 between Fks1p and Fks2p of S. cerevisiae, similar to the characteristics observed with the echinocandin derivative, L-733,560 (6,10). Furthermore, we identify one determinant amino acid residue involved in this differential sensitivity by using a series of mutant catalytic subunits, Fks1p and Fks2p. Finally, we discuss a possible interaction between aerothricin1 and the catalytic subunit of ␤-1,3-glucan synthase. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18
Chimeric molecules are shown in Fig. 3. The first series of chimeric genes were constructed by general recombination techniques. Region A of both FKS1 and FKS2 and a remaining region containing the Cterminal of FKS1 were amplified by PCR. Sequences of primers were 5Ј-CCCCGAATTCCCATGAACACTGATCAACAACC-3Ј for the region A of FKS1, 5Ј-CCCCGAATTCTTATGTCCTACAACGATCCAAAC-3Ј for the region A of FKS2, 5Ј-TG(A/G)TTGTCGACCAATTGTTGATAGTT-3Ј for region A of both FKS1 and FKS2, and 5Ј-ATTGGTCGACAACCAA-CCTTTGGCTGCTTACAAG-3Ј and 5Ј-TTTTGGGCCCTTTCAGAATTA-CTGACACCGAAAGCTGCTCCG-3Ј for the remaining region containing C-terminal of FKS1. The amplified fragments were subcloned into pT7blue (Takara Ltd.) and subjected to sequencing for confirmation. A remaining region containing a C terminus of FKS2 was prepared as a SalI-ApaI fragment. The A regions and the remaining regions containing C terminus were ligated and inserted into the pRS414-pT as an EcoRI-ApaI fragment, generating FKS1-A2 and FKS2-A1 genes.
For For the third series of chimeric mutants, constructions were performed with the same method described above. In the first step, intermediate fragments corresponding to regions E-G of Fks1p and Fks2p were amplified by PCR with primers 5Ј-CGTAGACGTCCCAAGTTTA-GAGTTCAATTATC-3Ј and 5Ј-GAAAATAGACAATGTATAACGTCTC-ACCCAATC-3Ј for region E, 5Ј-GATTGGGTGAGACGTTATACATTGT-CTATTTTC-3Ј and 5Ј-GCCAATGTGCGACAGTACCGAACAGCAACA-TTAACATTG-3Ј for region F, and 5Ј-CAATGTTAATGTTGCTGTTCG-GTACTGTCGCACATTGGC-3Ј and 5Ј-GATTGGCGCCAAGGTACAAA-TGATGATACG-3Ј for region G. In the second step, the components were mixed and amplified, generating chimeric fragments with an AatII and a BbeI site at each end. Original AatII-BbeI fragments of FKS1 and FKS2 were replaced with the resulting chimeric fragments, FKS1-E2, -F2, and -G2, and FKS2-E1.
Chemicals and Other Techniques-Cycloheximide and 5Ј-fluoroorotic acid was purchased from Wako Pure Chemical Industries, Ltd. Aero-thricin1 was prepared as described previously (24). Restriction endonucleases and Taq polymerase were purchased from Takara Ltd. Yeast transformation, plasmid shuffling, PCR method, and gene manipulation were performed as described previously (28). DNA sequencing was done by using an automated DNA sequencer model 373A with a Dye Terminator Cycle Sequencing Core kit (Applied Biosystems). Sequence data were analyzed with GENETYX for windows version 4.0.1.0 (Software Development Co., Ltd.).
Measurement of Growth Inhibition-Growth inhibition was determined in a standard microdilution assay. Briefly, 10 4 cells were cultivated with 100 l of medium on 96-well microtiter plates at 30°C for 16 -24 h in the presence or absence of antifungal compounds. An A 595 of the exponentially growing cells in the 96-well plates was measured by EIA-reader (Bio-Rad). IC 50 values refer to the compound concentrations that gave 50% inhibition of cell growth compared with the control. A spotting assay was performed by spotting 10 4 cells onto YPD agar plates containing aerothricin1 at different concentrations from 0.003 to 1 g/ml. The minimum inhibitory concentration (MIC) 2 was determined after overnight incubation of the spotted plates at 30°C.

RESULTS
Differential Sensitivity to Aerothricin1 between S. cerevisiae fks1 and fks2 Null Mutants-It is known that S. cerevisiae fks1⌬ mutant is more sensitive to L-733,560, one of the echinocandin derivatives, than the wild type strain (6). This differential sensitivity is thought to be due to biochemical characteristics of Fks2p, which is more sensitive to this compound than Fks1p (10). These lines of evidence intrigued us to test the antifungal activity of aerothricin1 (Fig. 1), a novel ␤-1,3-glucan synthase inhibitor, against both fks1D and fks2D null mutants. As shown in Table I, the fks1⌬ null mutant appeared to be more sensitive to aerothricin1 than the fks2⌬ null mutant or the parental wild type strain A451.
In an attempt to address this differential sensitivity more precisely, each of the FKS1 and FKS2 expression plasmids was introduced into the fks1⌬ fks2⌬ double null mutant harboring the URA3-borne GALp-FKS1 plasmid, and this plasmid was then eliminated by 5Ј-fluoroorotic acid treatment. In the resultant cells, either the FKS1 or FKS2 gene can be expressed in the absence of endogenous Fks1p and Fks2p under the control of a constitutive GAP promoter (29). The introduction of either In the Type of sensitivity column, Fks1p means that its MIC value is equal to or more than 0.3 g/ml, and Fks2p means equal to or less than 0.03 g/ml. GAPp-driven FKS1 or GAPp-driven FKS2 suppressed the lethal phenotype of S. cerevisiae fks1⌬ fks2⌬ double null mutant, and the resulting mutant showed the same growth rate as the parental wild type strain, YPH499 (data not shown). At first, we confirmed the differential sensitivity between the two strains. As shown in Table I, the double null mutant expressing only Fks2p was more sensitive to aerothricin1 than that expressing only Fks1p. We also questioned whether they showed a differential sensitivity against another type of antifungal agent, cycloheximide. However, we observed no clear difference in their sensitivities against this agent (Table I). These results suggest that this differential sensitivity against aerothricin1 may simply rely on differences between Fks1p and Fks2p.
Identification of Regions Containing Determinant(s) for the Aerothricin1 Sensitivity of Fks2p-Based on the hypothesis that differences in the primary sequences of Fks1p and Fks2p may represent determinants for the aerothricin1 sensitivity, we first looked at the intracellular domain at the N terminus of Fks2p because this region is less homologous, even though Fks1p and Fks2p exhibit 88.1% identity throughout overall sequences. For this purpose, two kinds of chimeric genes, FKS1-A2 and FKS2-A1, were constructed by replacing the Nterminal region (Fig. 3, A and B) and introduced into the fks1⌬ fks2⌬ double null mutant under the control of the GAP promoter. For a rapid profiling of their sensitivities, we applied a spotting assay with plates containing different concentrations of aerothricin1. Although aerothricin1 sensitivities were determined as MICs in this assay (see "Experimental Procedures" and Fig. 2), we could see the same differential sensitivity as seen in the comparison of IC 50 values against the double null mutant cells expressing either Fks1p or Fks2p (Table I). Both chimeric proteins, Fks1-A2p and Fks2-A1p, appeared to suppress the synthetic lethal phenotype of fks1⌬ fks2⌬ double null mutant because no growth defects observed compared with the parent strain YPH499 (data not shown). As shown in Fig. 3B, the spotting assay revealed that Fks1-A2p failed to confer the mutant cells hypersensitive to aerothricin1. Surprisingly, the mutant cells expressing Fks2-A1p showed Fks2p-like sensitivity. These results indicate that the determinant(s) may exist in the C-terminal region of Fks2p, which is highly conserved between Fks1p and Fks2p, sharing 92.5% identity.
To minimize regions containing the determinant(s), we performed the second round of chimeric gene analysis. As illustrated in Fig. 3A, the sequence encoding the C-terminal region of Fks2p was divided into three regions, named B, C, and D. Each of them was replaced with the corresponding region of FKS1 gene, resulting in chimeric genes FKS1-B2, FKS1-C2, and FKS1-D2. Mutant cells harboring each chimeric gene also grew normally (data not shown). By the spotting assay, it was shown that only Fks1-C2p conferred the mutant cells hypersensitive to aerothricin1 (Fig. 3B). We also tested an opposite substitution, Fks2-C1p, in which the region C of Fks2p was replaced with that of Fks1p (Fig. 3A). Interestingly, the replacement of region C in Fks2p resulted in a loss of hypersensitivity to aerothricin1 (Fig. 3B), suggesting that region C of Fks2p contains the determinant(s) of differential sensitivity to aerothricin1.
To localize a region containing the determinant(s) more precisely, we further divided a portion of Fks2p including region C into three parts (region E, F, G in Fig. 3A) and constructed chimeric genes by replacing the cognate region in FKS1 with that of FKS2 gene. Finally, it was found that an introduction of region E of Fks2p into Fks1p was enough to provide the aero-thricin1 sensitivity to the mutant cells (Fig. 3B, FKS1-E2). Conversely, Fks2p harboring a replacement of region E failed to confer the mutant cells sensitive to the inhibitor (Fig. 3B,  FKS2-E1).
The 1355th Ile Is Essential for the Aerothricin1 Sensitivity of Fks2p-From a series of analyses using chimeric proteins, it was suggested that determinant(s) could be located within a region shared by regions C and E (Fig. 4A). The shared region consists of 74 amino acids and contains 10 non-conservative amino acids between Fks1p and Fks2p (86.5% identical). Identification of these non-conservative amino acids prompted us to question which amino acid was essential for the aerothricin1 sensitivity of Fks2p. For this purpose, we mutagenized each of them in Fks1p with that of Fks2p by using site-directed mutagenesis. All 10 Fks1 mutant proteins were analyzed in the fks1⌬ fks2⌬ double null mutant cells with the spotting assay. Surprisingly, as summarized in Fig. 4B, only one mutant Fks1p (FKS1 K1336I ) conferred the cells sensitive to aerothricin1. We also examined the effects of substitution of the corresponding amino acid residue of Fks2p with that of Fks1p and found that this opposite substitution resulted in a complete loss of the aerothricin1 hypersensitivity (Fig. 4B, FKS2 I1355K ). The switching of the sensitivity due to these substitutions was further confirmed by determination of IC 50 values of aerothri-cin1 against the fks1⌬ fks2⌬ double null mutant cells expressing each mutant protein (Table II). Although IC 50 values of echinocandin B against these cells were also determined, no clear difference was observed in their sensitivities.
Next we investigated the effects of substitutions on biochemical properties of Fks1p and Fks2p. ␤-1,3-Glucan synthase complexes containing the mutant catalytic subunits were partially purified from the fks1⌬ fks2⌬ double null mutant cells After overnight incubation at 30°C, MICs of aerothricin1 were determined as minimum concentrations inhibiting the cell growth. In the Type of sensitivity column, Fks1p means that its MIC value is equal to or more than 0.3 g/ml, and Fks2p means equal to or less than 0.03 g/ml.
expressing each of the mutants Fks1p and Fks2p, and then their sensitivities to aerothricin1 were examined. As shown in Table II, both mutated Fks1 K1336I p and Fks2 I1355K p revealed similar specific activities compared with their wild type proteins. Judging from IC 50 values, however, it was shown that Fks1 K1336I p was over 50-fold more sensitive to aerothricin1 than Fks1p. On the other hand, Fks2 I1355K p appeared to be more resistant to aerothricin1 than Fks2p, the same as Fks1p. These results imply that one isoleucine residue at the position 1355 of Fks2p is one of the determinants for aerothricin1 sensitivity. DISCUSSION Aerothricin1/RO0093655, identical to FR901469, is a potent and selective antifungal agent inhibiting the synthesis of ␤-1,3glucan, which is a main component of fungal cell wall. Although it has been shown that aerothricin1 inhibits in vitro ␤-1,3glucan synthesis of C. albicans and growth of various fungi, such as several Candida species and A. fumigatus (22)(23)(24)(25)(26)(27), the detailed molecular mechanisms of the inhibition are still unknown. In this report, we found that the fks1 null mutant was more sensitive to aerothricin1 than either the fks2 null mutant or the parental strain in S. cerevisiae. This observation is the first evidence suggesting that the catalytic subunit of ␤-1,3glucan synthase would be a molecular target of aerothricin1. In the course of our experiments shown here, we initially used a number of chimeric Fks proteins. Surprisingly, none of them resulted in impaired growth when expressed in the fks1⌬ fks2⌬ double null mutant of S. cerevisiae. These results indicate not only that these chimeric proteins are functional but also that Fks1p and Fks2p are highly structurally homologous.
Although we cannot exclude a possibility that other amino acid residues are involved in the aerothricin1 sensitivity of Fks2p, several lines of evidence presented here demonstrate that one amino acid residue, Ile-1355 of Fks2p, is one dominant determinant for its aerothricin1 sensitivity. Alternatively, Lys-1336 of Fks1p is the dominant one for the resistance to aero-thricin1. One possible explanation of these determinant residues in the interaction with aerothricin1 is that their charges may affect affinity between Fks proteins and aerothricin1; a positive charge of the 1336th lysine residue of Fks1p may interfere with the interaction of aerothricin1 with Fks1p molecules, because aerothricin1 has a positive-charged nitrogen at the ornithine moiety ( Fig. 1), which is essential for its inhibition activity (data not shown). Alternatively, the hydrophobicity of the 1355th isoleucine residue of Fks1p may be important for the aerothricin1 association.
Aerothricin1 exhibits growth inhibition effectively against at least C. albicans and A. fumigatus (24,25); the IC 50 values against C. albicans ATCC48130 and A. fumigatus CF1003 were 0.03 and 0.06 g/ml, respectively. As shown in Fig. 5, primary structures of the fourth extracellular domains, including the determinant residue, are conserved among these fungi. Interestingly, positions of the expected determinant are occupied with isoleucine or valine residues, which are non-charged and hydrophobic, supporting the importance of the hydrophobic residue of Fks2p for aerothricin1 interaction. It is interesting to question whether A. nidulans and P. brasiliensis are sensitive to this compound because the region including the determinant residue is also highly conserved and possesses the determinant isoleucine residue. C. neoformans is known to be less sensitive to aerothricin1 in growth inhibition assay (25) even though we can find an isoleucine residue in its Fks1p at the same position when aligned with Fks proteins from other sensitive fungi (Fig.  5). However, its sequence similarity is quite low against other Fks proteins, suggesting the region is structurally different from other Fks proteins.
Echinocandins share similar features with aerothricin1 in their chemical structure, such as cyclic macropeptides with a lipophilic side chain. In particular, S. cerevisiae Fks1p and Fks2p exhibit differential sensitivities against both types of inhibitors. Therefore, it is possible that they may share domains that interact with Fks proteins. However, it is unlikely that they share the same determinant(s) for their differential sensitivities because we failed to find any clear differences in the sensitivities of point-mutated Fks proteins (Table II). In addition, Fks1-A2p was more sensitive to echinocandin B than Fks1p, Fks2p, or Fks2-A1p (data not shown). These observations suggest that aerothricin1 and the echinocandins may interact differently with catalytic subunits via different determinant residue(s). By investigating the differential aerothricin1 sensitivity between Fks1p and Fks2p, we obtained a clue for understanding the mechanism of inhibition of ␤-1,3-glucan synthesis by aero-thricin1. For further analysis of the aerothricin1 inhibition, focus should be on the fourth extracellular domain. Monitoring a direct interaction between aerothricin1 and the fourth extracellular domain might be useful in understanding the actual  physical relationship between aerothricin1 and the catalytic subunits of ␤-1,3-glucan synthase. These results would be helpful for developing more potent derivatives from aerothricin1.