Homoisocitrate dehydrogenase from Candida albicans: properties, inhibition, and targeting by an antifungal pro-drug


Correspondence: Iwona Gabriel, Department of Pharmaceutical Technology and Biochemistry, Gdańsk University of Technology, 11/12 Narutowicza Str., 80-233 Gdańsk, Poland. Tel.: +48 583486078; fax: +48 583471144; e-mail: iwogabri@pg.gda.pl


The LYS12 gene from Candida albicans, coding for homoisocitrate dehydrogenase was cloned and expressed as a His-tagged protein in Escherichia coli. The purified gene product catalyzes the Mg2+- and K+-dependent oxidative decarboxylation of homoisocitrate to α-ketoadipate. The recombinant enzyme demonstrates strict specificity for homoisocitrate. SDS-PAGE of CaHIcDH revealed its molecular mass of 42.6 ± 1 kDa, whereas in size-exclusion chromatography, the enzyme eluted in a single peak corresponding to a molecular mass of 158 ± 3 kDa. Native electrophoresis showed that CaHIcDH may exist as a monomer and as a tetramer and the latter form is favored by homoisocitrate binding. CaHIcDH is an hysteretic enzyme. The KM values of the purified His-tagged enzyme for NAD+ and homoisocitrate were 1.09 mM and 73.7 μM, respectively, and kcat was 0.38 s−1. Kinetic parameters determined for the wild-type CaHIcDH were very similar. The enzyme activity was inhibited by (2R,3S)-3-(p-carboxybenzyl)malate (CBMA), with IC50 = 3.78 mM. CBMA demonstrated some moderate antifungal activity in minimal media that could be enhanced upon conversion of the enzyme inhibitor into its trimethyl ester derivative (TMCBMA). TMCBMA is the first reported antifungal for which an enzyme of the AAP was identified as a molecular target.


l-Lysine is an essential amino acid for mammals including humans, whereas bacteria, plants, and fungi have developed pathways of lysine biosynthesis. There are two versions of this pathway: the diaminopimelic acid pathway, that is characteristic for bacteria, plants, and lower fungi, and the α-aminoadipate pathway (AAP) present in higher fungi, Ascomycetes and Basidiomycetes, including most of the human pathogenic yeast and filamentous fungi but also in some archaea and thermophilic bacteria. Enzymes that catalyze reactions of the first of the above-mentioned pathways are considered promising molecular targets for antibacterial chemotherapy (Hutton et al., 2003), whereas the second pathway could be a source of new targets for antifungal chemotherapy (Xu et al., 2006). Homocitrate synthase (HCS), homoaconitase (HA) and homoisocitrate dehydrogenase (HIcDH) catalyzing biosynthetic reactions present only in fungal cells and having no counterparts in mammalian cells, are the most obvious candidates for the molecular targets. It was shown that disruption of both genes encoding HCS in C. albicans leads to lysine auxotrophy (Kur et al., 2010) and diminished virulence in some systems of fungal cells lacking HCS of HA encoding genes was demonstrated (Liebmann et al., 2004; Schöbel et al., 2010).

Homoisocitrate dehydrogenase (EC catalyzes the fourth reaction of the AAP, i.e., the NAD+-dependent conversion of homoisocitrate to α-ketoadipate (α-Ka), as shown in Fig. 1. HIcDH is a member of the family of pyridine nucleotide-dependent β-hydroxyacid oxidative decarboxylases (Karsten & Cook, 2000). The reaction is metal ion-dependent and the enzyme selectively binds the Mg(II):homoisocitrate complex (MgHIc) (Lin et al., 2007).

Figure 1.

Reaction catalyzed by homoisocitrate dehydrogenase.

The genetics and biochemistry of homoisocitrate dehydrogenase have been studied to some extent for the enzyme from prokaryotic sources: Deinococcus radiodurans (DraHIcDH), Thermus thermophilus (TthHIcDH), Pyrococcus horikoshii (PhHIcDH), and its yeast versions from Saccharomyces cerevisiae (ScHIcDH) and Schizosaccharomyces pombe (SpHIcDH) (Miyazaki et al., 2003; Miyazaki, 2005ab; Yamamoto et al., 2007). Notably, there are no data concerning HIcDH from human pathogenic fungi. ScHIcDH has been the only homoisocitrate dehydrogenase overexpressed and purified as a His-tagged fusion protein. Addition of 10 His residues to the N-terminal end of that protein resulted in almost no effect on enzyme activity (Lin et al., 2007). A catalytic constant determined for the wild-type enzyme was 17 s−1 (Yamamoto et al., 2007) whereas kcat of the oligoHis-tagged HIcDH was 13 s−1 (Lin et al., 2007).

Substrate specificity of HIcDHs varies among the species studied. For example, SpHIcDH (Chen & Jeong, 2000) and Methanococcus jannaschii HIcDH (MjaHIcDH) (Howell et al., 2000) exclusively use homoisocitrate as a substrate and neither 3-isopropylmalate nor isocitrate serve as alternative substrates of these enzymes. On the other hand, homoisocitrate dehydrogenase from T. thermophilus exhibits a broader substrate spectrum and even prefers isocitrate to homoisocitrate by a factor of 20 (Miyazaki et al., 2003).

The potential of AAP enzymes as targets for antifungal chemotherapy has inspired the design and synthesis of structural analogs of AAP intermediates, including analogs of homoisocitrate, which have been tested for antifungal activity and/or inhibition of AAP enzymes. Particularly, thiahomoisocitrate was found to be an effective inhibitor of HIcDH from S. cerevisiae and D. radiodurans (Yamamoto et al., 2007; Yamamoto & Eguchi, 2008) but no data on antifungal effect of this compound are available. On the other hand, a variety of carboxyalkyl- and carboxyaryl-substituted d-malic acid derivatives and their methyl esters were synthesized as analogs of HIc and some of these compounds showed moderate inhibition of fungal growth, that could be partially restored by the presence of l-lysine in the growth medium (Palmer et al., 2004) but an actual target for these compounds has not been unequivocally identified.

In this manuscript, we present results of our studies on identification of a gene encoding HIcDH in the human pathogenic yeast C. albicans, construction of the His-tagged version of the gene product, and character-ization of its properties, including inhibition by a homoisocitrate analog. Evidence is also provided for the antifungal efficiency of the HIcDH inhibitor studied that has been enhanced upon conversion of the enzyme inhibitor into its ester derivative.

Materials and methods

Bacterial strains and growth conditions

The E. coli TOP 10F' strain from Invitrogene was used in all the cloning procedures. The E. coli Rosetta (DE3) pLysS strain from Novagen was used for the overexpression of wild-type and oligoHis-tagged CaHIcDH. Escherichia coli strains were cultured at 37 °C on LB (Luria–Bertani) solid medium [1.0% (w/v) NaCl, 1.0% (w/v) tryptone, 0.5% yeast extract, and 1.5% (w/v) agar], and in LB liquid medium [1.0% (w/v) NaCl, 1.0% (w/v) tryptone, and 0.5% yeast extract] supplemented with 0.1 mg mL−1 ampicillin when necessary.

Fungal strains and culture condition

The reference strains used in the present study were: Candida albicans ATCC 10231, Candida dubliniensis CBS 7987, Candida glabrata DSM 11226, Candida krusei DSM 6128, Candida pseudotropicalis KKP 324, Candida tropicalis KKP 334, S. cerevisiae ATCC 9763, C. albicans SC5314. Fungal strains were grown at 30 °C in YPG (1% yeast extract, 1% peptone, 2% glucose) liquid medium.

Plasmids, enzymes, and other materials

The plasmids used were: pET15b+ (Novagen) and pUC19 (Fermentas, Lithuania). Restriction enzymes were purchased from Fermentas and New England Biolabs. Protein molecular mass markers were from Fermentas. DNA polymerase was purchased from DNA-Gdańsk. T4 DNA ligase was from Epicentre. d,l-3-Isopropylmalic acid was purchased from Wako Pure Chemicals, Osaka, Japan, and d,l-Isocitric acid from Sigma, St. Louis, MO. (2R,3S)-Homoisocitric acid was synthesized according to the recently published procedure (Ma & Palmer, 2000). (2R,3S)-3-(para-carboxybenzyl) malic acid (CBMA) and its trimethyl ester derivative (TMCBMA) were synthesized as described previously (Palmer et al., 2004). Purification of oligoHis-tagged proteins was performed on Ni2+-IDA (iminodiacetic acid) agarose (His Bind Resin, Novagen). Wild-type enzyme was purified by ion-exchange chromatography on a ResourceQ column. Monoclonal Anti-polyHistidine Peroxidase Conjugate Clone HIS-1 and 3,3′,5,5′-tetramethylbenzidine were from Sigma.

DNA manipulations

Candida albicans SC5314 genomic DNA was isolated according to the protocol of Bacterial & Yeast Genomic DNA Purification Kit (DNA-Gdańsk). Isolation of plasmid DNA was carried out according to the protocol of the Plasmid Mini kit (A&A Biotechnology). DNA fragments were isolated from agarose gels following the standard procedure of the DNA Gel-Out kit (A&A Biotechnology). DNA purification after enzyme treatment was performed according to the instructions in the DNA Clean-up kit (A&A Biotechnology). DNA digestion with restriction enzymes was carried out according to the enzyme supplier's instructions. DNA fragments were ligated and E. coli cells were transformed according to the standard methods (Sambrook et al., 1989).

Cloning of the CaLYS12 gene

The LYS12 gene was amplified from the C. albicans SC5314 genomic DNA by PCR. The primers used in the amplification were: Forward: DehhisN. 5′–TTTTACATGTTACATCATCATCATCATCATGCTGCTAGATCTTCAATTCGT–3′; Reverse: Deh2 5′– AAACTCGAGTTAGAATCTTCTAATGATATCATCTATAACCTCTTG–3′, for the LYS12HisN gene and Forward: Deh1: TTTTACATGTTAGCTGCTAGATCTTCAATTCGTCGT; Reverse: Deh2 for the wild-type LYS12 gene cloning. The artificial cloning sites (underlined) for PscI (DehhisN) and XhoI (Deh2) at the 5′-ends of the primers were introduced to facilitate the cloning procedure. The hexaHis-tag-encoding sequence introduced in the forward primer is bolded. The PCR products (1164 bp and 1146 bp, respectively) were purified from an agarose gel and cloned directionally into the SmaI site of the pUC19 vector, giving a recombinant cloning plasmid pUC19-LYS12HisN (3850 bp) and pUC19-LYS12 (3832 bp). The LYS12HisN and LYS12 genes, cut off from the pUC19- LYS12HisN and pUC19-LYS12 plasmids with PscI and XhoI, were cloned into the NcoI and XhoI sites of pET15b+, affording recombinant expression plasmids pET15b-LYS12HisN and pET15b-LYS12. The identity of the plasmids was confirmed by restriction analysis and DNA sequencing. The obtained constructs encoded the putative CaHIcDH containing an additional hexaHis-tag at the N-terminus (His6CaHIcDH) and the wild-type (CaHIcDH).

Overexpression of the homoisocitrate dehydrogenase gene

Escherichia coli Rosetta (DE3)pLysS cells, transformed with the pET15b-LYS12HisN or pET15b-LYS12 expression plasmid, were grown overnight in the LB liquid medium containing ampicillin, at 37 °C. A sample of the culture (10 mL) was then transferred to a fresh LB broth (1 L) containing ampicillin and the cell suspension was grown at 37 °C. Expression was induced by the addition of 0.5-mM isopropyl-β-d-thiogalactopyranoside to the cultures grown to OD600 = 0.5–0.6 and incubation was continued for another 5 h. The cells were harvested by centrifugation at 5000 g for 10 min at 4 °C.

Purification of the oligoHis-tagged homoisocitrate dehydrogenase

The oligoHis-tagged CaHIcDH was purified by metal-affinity chromatography. The bacterial pellet was suspended in buffer A (20-mM Tris/HCl, pH 8, 500-mM NaCl, 5-mM imidazole, 0.1% Tween 20, and 0.5-mM PMSF) and the cells were disrupted by sonication (5 × 30 s bursts with 30 s intervals at a power setting of 30 W, using a Branson sonifier 250) on ice. The total lysate was centrifuged at 16 000 g for 20 min at 4 °C. The supernatant (crude extract) was applied to a Ni2+-IDA agarose column, bed volume 10 mL, pre-equilibrated with 4 vol. of buffer A. The column was then washed with 20 mL of the same buffer, followed by washing with 40 mL of buffer W (20-mM Tris/HCl, pH 8, 500-mM NaCl, 50-mM imidazole, 0.1% Tween 20, and 0.5-mM PMSF). The oligoHis-tagged protein was eluted by two 5-mL portions of elution buffer E (20-mM Tris/HCl, pH 8, 200-mM imidazole, 500-mM NaCl, 0.1%Tween 20 and 0.5-mM PMSF). For further assays, the eluates were concentrated by ultrafiltration using Vivaspin concentrators (10 kDa cut-off limit; Viva Science Ltd.) at 7000 g for 30 min at 4 °C and the buffer was exchanged for buffer B (50 mM HEPES-NaOH pH 7.8) using HiTrapTM Desalting Columns (LKB Biotech).

Purification of the wild-type homoisocitrate dehydrogenase

The wild-type CaHIcDH was purified by ion-exchange chromatography. The bacterial pellet was suspended in buffer F (20 mM HEPES-NaOH pH 7.8 and 0.5 mM PMSF) and the cells were disrupted by sonication (5 × 30 s bursts with 30 s intervals at a power setting of 30 W, using a Branson sonifier 250) on ice. The total lysate was centrifuged at 16 000 g for 20 min at 4 °C. The supernatant (crude extract) was applied to a ResourceQ column, bed volume 6 mL, pre-equilibrated with 4 vol. of buffer F. The column was then washed with 20 mL of the same buffer. The wild-type protein was eluted by the gradient of buffer G (buffer F containing 1 M NaCl). Fractions exhibiting HIcDH activity were pooled and concentrated by ultrafiltration.

Western blot analysis

Proteins were separated on a 12% polyacrylamide gel according to Laemmli (1970). Electrophoresis was followed by electroblotting to a nitrocellulose membrane with the use of a transfer buffer (25-mM Tris-HCl pH 8.3, 192-mM glycine, 20% methanol). Nitrocellulose membranes were incubated in 3% skimmed milk solution in washing buffer (10-mM Tris-HCl pH 8.0, 30-mM NaCl) for 1 h. After 3× wash with the washing buffer, the membranes were incubated for 1 h with Monoclonal Anti-polyHistidine Peroxidase Conjugate Clone HIS-1 1 : 2000 solution. His6HIcDH was detected by the addition of peroxidase substrate 3,3′,5,5′-tetramethylbenzidine (TMB) according to the manufacturer's instructions.

Determination of a molecular mass of the native protein

Size-exclusion chromatography was performed on Superdex 200 HR 10/30 (Pharmacia LKB). The protein was applied in the volume of 0.5 mL and then eluted at 0.5 mL min−1 by 25 mM potassium phosphate buffer (pH 7.0) containing 0.15 M NaCl. Protein elution was monitored at 280 nm. The molecular mass standards were: α-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), carbonic anhydrase (29 kDa) and cytochrom c (12.4 kDa).

Oligomeric structure was also analyzed by native PAGE electrophoresis using a NativePAGE™ Novex® 4-16% Bis-Tris Gels kit (Invitrogen). The experiments were run according to the manufacturer's procedure.

Determination of HIcDH activity and substrate specificities

Reaction catalyzed by CaHICDH was monitored by measuring the NADH formation at 340 nm with a Perkin Elmer Lambda 45 UV-vis spectrometer. The standard assay mixture contained 10-mM KCl, 5.0-mM MgCl2, 5-mM NAD+ and 0.1 mM of (2R,3S)-homoisocitrate, d,l-isocitrate, or d,l-3-isopropylmalate in 50-mM Hepes–NaOH (pH 7.8) buffer. The reaction mixtures containing all the required components except the enzyme were pre-incubated for about 3 min and the reaction was started by addition of the enzyme (50 μg) to the mixture. The formation of NADH was measured for 30 s at 20 °C. Kinetic measurements were performed in standard assay mixtures containing appropriate amount of enzyme and various amounts of NAD+ (0–5.0 mM) or HIc (0–0.2 mM). Data were analyzed by the Lineweaver–Burk double reciprocal plots. The data are presented as means of three independent experiments. One unit of the enzyme specific activity is defined as an amount of enzyme that converts 1.0 μmole of homoisocitrate to α-ketoadipate min−1 at pH 7.8.

Effect of metal ions and other reagents on enzyme activity

The effect of metal ions on His6CaHIcDH activity was determined by measuring the enzyme activity in standard assay mixtures containing MgCl2, MnCl2, and CaCl2 (0–15 mM in the presence of 1 mM KCl) or KCl and NaCl (0–10 mM in the presence of 5 mM MgCl2).

The effects of CBMA on the His6-HIcDH activity were determined by measuring the enzyme activity in standard assay mixtures containing various concentration of inhibitor (0–5 mM). Kinetic measurements were performed for fixed concentration of one of the substrates and different concentrations of another substrate in the presence of CBMA (0, 1, 2 or 4 mM).

Determination of antifungal activity of CBMA and TMCBMA

Minimum inhibitory concentration (MIC) values were determined by the serial twofold dilution microliter plate method under conditions described in the M27-A3 document (CLSI, 2008). Turbidity was measured using a microplate reader (Victor3V, Perkin Elmer, Shelton, CT).

Homology modeling

The 3-D structure of C. albicans homoisocitrate dehydrogenase was constructed by homology modeling using the modeller9v2 package (Sali & Blundell, 1993). The templates used were the structures of homoisocitrate dehydrogenase from Schizosaccharomyces pombe (PDB: 3ty4) and from Thermus thermophilus (PDB: 3asj), showing 64% and 46% sequence identity to the target respectively. The resulting model was energy-minimized with the GROMOS 43a2 force field, by means of the gromacs 4.55 package (Berendsen et al., 1995; Hess et al., 2008).

Docking calculations

Dockings of the substrate (homoisocitrate) and inhibitors to the C. albicans HIcDH model were performed with the autodock software version 4.2 (Morris et al., 2009). Two versions of the protein receptor were built. The first one was based on the structure of the modeled enzyme. The second one included the magnesium ion bound to the protein. As the magnesium was not present in any of the available homoisocitrate dehydrogenase structures, its position was deduced on the basis of the multiple sequence alignment (Fig. 2). According to the alignment with annotated sequences of S. cerevisiae and S. pombe proteins, the residues Asp272, Asp276, and Asp248′ were selected to form the putative metal ion-binding site. In our model of the C. albicans enzyme, these residues are located in the binding pockets close to each other and their spatial arrangement resembles the metal-binding sites found in other enzyme versions (Miyazaki et al., 2003). Electrostatic potential around the protein was calculated and spots preferential for placing positive ions were identified. In one of such spots found near the Asp272, Asp276, and Asp248′ residues, the magnesium ion was placed. The resulting receptor complex comprising the protein and metal ion was then again energy-minimized with the G43a2 force field.

Figure 2.

Multiple alignment of amino acid sequences of homoisocitrate dehydrogenase from Candida albicans (CaHIcDH), Saccharomyces cerevisiae (ScHIcDH), Schizosaccharomyces pombe (SpHIcDH) and Thermus thermophilus (ThHIcDH). Residues constituting the isocitrate/isopropylmalate dehydrogenase conserved site are framed; Lys and Tyr catalytic residues as well as the conserved Arg residues present in all β-decarboxylating dehydrogenases are in the bold frames; Asp residues constituting the magnesium binding site are on the red background; putative residues involved in binding of the 4-carboxy group of homoisocitrate/isocitrate (Arg85 in TtHIcDH, Lys149 in CaHIcDH, and corresponding Lys residues in ScHIcDH and SpHIcDH) are in the bold frames on the red background.

Initial geometries of ligands (homoisocitrate, CBMA and (2S, 3S) thiahomoisocitrate) were built by means of the Accelrys Discovery Studio. Final AutoDock topologies of ligands and receptors in the pdbqt format, including the atom types and partial charges appropriate for the program, were calculated and assigned with the AutoDock-accompanied scripts. The ‘grid approach’ of AutoDock was used to speed up calculations; the docking grid was centered in the middle of the binding site at the Arg122 guanidine moiety and its dimension was set to 75 × 75 × 75 points, with the default grid spacing of 0.375 Å. A semi-flexible methodology was used (the receptor fixed and the ligand flexible) and Lamarckian genetic algorithm (LGA) was chosen as a search engine. The initial size of the population was set to 150 random solutions and the number of generations was limited to 27 000. To ensure the convergence of the search procedure, the number of objective function evaluations was raised to 25 000 000. Docking procedure was repeated 50 times and the resulting ligand poses were clustered with the tolerance of 2.0 Å. The most abundant and the lowest energy clusters were selected for analysis.

Other methods

Protein concentration was determined by the Bradford method (Bradford, 1976). Discontinuous SDS/PAGE was performed by the method of Laemmli (Laemmli, 1970) using a 12% (w/v) separating gel and a 5% (w/v) stacking gel.

Results and discussion

It is known that in S. cerevisiae HIcDH is encoded by the LYS12 geneA blast search using the nucleotide sequence of LYS12 from the Saccharomyces Genome Database (www.yeastgenome.org) performed on the Candida Genome Database (www.candidagenome.org) revealed ORFs 19.2525 and 19.10060 named LYS12 on Ca21chrR of C. albicans, highly homologous to their S. cerevisiae counterpart. The CaLYS12 ORFs contain 1131 nucleotides and code for the protein of 376 aa with a theoretical molecular mass of 40 668.4 Da. The amino acid sequence of a putative CaLYS12 product shows 96% identity to that of HIcDH from Candida tropicalis, 69% identity to that of the S. cerevisiae enzyme, 64% to the S. pombe sequence, and 46% to that of T. thermophilus. Multiple sequence alignment (Fig. 2) and amino acid sequence analysis revealed the presence of a sequence known as a isocitrate/isopropylmalate dehydrogenase-conserved site: IPRO19818 (Sigrist et al., 2002) (268NLYGDILSDGAAALVGSLGV287) and the catalytic Lys-Tyr pair (Lys211 and Tyr147), corresponding to Lys206 and Tyr150, shown previously in ScHIcDH to catalyze the acid–base chemistry of the enzymatic reaction (Lin et al., 2009).

Cloning, expression, and purification of His6CaHIcDH in E. coli

The CaLYS12 gene was cloned into the expression vector pET15b+ in a system that enabled overexpression of a wild-type gene product and a recombinant protein containing the N-terminal oligoHis-tag sequence. The putative LYS12 gene fragment was amplified from C. albicans genomic DNA using the indicated primers (Dehhisn, Deh2 and Deh1, Deh2), cloned into the pUC19 vector with the use of the SmaI restriction enzyme and then subcloned into the pET15b+ expression plasmid. The resulting pET15b–LYS12HisN and pET15b–LYS12 plasmids were analyzed by restriction analysis and then sequenced. Both plasmids, in which the LYS12 gene was placed under control of the T7 promoter, were transformed into E. coli Rosetta (DE3)pLysS cells. Overexpression of the cloned gene induced upon IPTG addition resulted in high expression level of the wild-type and recombinant protein, constituting almost 30% of the soluble protein pool in cell-free extract, as revealed by the densitometric analysis. Purification of the recombinant His6CaHIcDH was achieved in a single step by metal-affinity chromatography. Unbound and weakly bound proteins were washed with 0–50 mM imidazole and His6CaHIcDH was eluted with 200 mM imidazole. Fractions from the Ni2+-IDA column were analyzed by SDS–PAGE (Fig. 3a) and Western blotting (Fig. 3b). SDS–PAGE analysis of the 200-mM imidazole eluate showed a single protein band corresponding to a protein with MW ≈43 kDa. The recombinant protein was purified to near homogeneity (> 95%, as revealed by a densitometric analysis) without considerable loss in protein yield. Western blotting analysis confirmed presence of an oligoHis-tagged fusion protein in the appropriate fraction. Purification of the wild-type CaHIcDH was also achieved in a single step by ion-exchange chromatography. The protein was purified with the efficiency sufficient for kinetic parameters comparison (purity > 80%, as revealed by the densitometric analysis).

Figure 3.

SDS-PAGE (a) and Western blotting (b) analysis of purification of N-terminally His6-tagged CaHIcDH overexpressed in Escherichia coli. Lane M – molecular-mass markers; lane 1 – total lysate of E. coli Rosetta (DE3)pLysS pET15b- LYS12HisN cells; lanes 2 and 3 – wash of the Ni(II)-IDA column with buffer W; lane 4 – 200 mM imidazole eluate from the Ni(II)-IDA column.

Characterization of His6CaHIcDH properties

The molecular mass of recombinant His6CaHICDH was determined under denaturing conditions by SDS-PAGE analysis and for the native protein, by size-exclusion chromatography. The SDS-PAGE analysis revealed the MW = 42.6 ± 1 kDa, slightly larger than the 41.491 Da value predicted from the amino acid sequence. In size- exclusion chromatography, the protein eluted in a single peak at an elution volume corresponding to the molecular mass of 158 ± 3 kDa. Comparison of the two values suggests a homotetrameric structure of the native His6CaHICDH. The same type of a quaternary structure was found for TthHICDH and DraHIcDH (Miyazaki et al., 2003; Miyazaki, 2005ab). On the other hand, for the already reported fungal HIcDHs, a dimeric or monomeric structure was suggested. As already reported, the SpHIcDH crystal structure indicates a homodimeric structure of the enzyme (Bulfer et al., 2012). Gel filtration analysis of HIcDH from S. lipolytica indicated that obtained results depend on enzyme concentration and the presence/absence of homoisocitrate. High protein concentration or the presence of the substrate stimulates formation of a dimeric form, whereas in the diluted solution and/or in the absence of the substrate, the enzyme exists as monomer (Gaillardin et al., 1982). A similar phenomenon was found in our studies for CaHIcDH. Electrophoresis under non-denaturizing conditions revealed that the wild-type CaHIcDH exists as monomer and homotetramer and that addition of HIc resulted in increasing intensity of bands corresponding to the homotetrameric structure (Fig. 4). A similar phenomenon was found for His6HIcDH.

Figure 4.

Native gel electrophoresis of CaHIcDH. Lane M – molecular mass markers; lane 1 – CaHIcDH pre-incubated with 0.1 mM HIc and 5 mM MgCl2; 2 – CaHIcDH pre-incubated with 0.05-mM HIc and 5-mM MgCl2; lane 3 – CaHIcDH alone.

Homology modeling of CaHIcDH

Looking for the molecular basis for the formation of a tetrameric structure by native CaHIcDH, homology modeling of this protein was performed. The amino acid sequences of S. pombe and C. albicans HIcDHs show over 60% identity, so that the PDB: 3ty4 structure of the former was used as a starting template for comparative modeling of the C. albicans protein. As the native S. pombe HIcDH is a homodimer, the resulting model was also dimeric. Assuming a possibility of tetrameric structure of CaHIcDH, known previously for the T. thermophilus enzyme and suggested by the results of our MW determination, the PDB: 3asj structure of the thermophilic archeon was used as a new template in the remodeling of the dimeric model obtained in the previous stage, giving finally the model shown in Fig. 5. Based on the multiple sequence alignment of S. cerevisiae, S. pombe, C. albicans, and T. thermophilus enzymes (Fig. 2), the most divergent regions of the sequences of eukaryotic versions of the enzyme are located in the loop areas 153–158, 193–202 and 228–233 (C. albicans numbering). The second and the third of these regions form solvent-exposed loops, distant from the binding site, but the first one, namely 153–158 according to the alignment with the sequence of T. thermophilus HIcDH, may be involved in the stabilization of the enzyme quaternary structure. The homologous loop of the prokaryotic protein (131–136, T. thermophilus numbering) forms the interface between subunits of the tetramer (Fig. 5). In all known eukaryotic versions of HIcDH, this loop is of a similar length and is 5–6 residues longer than the corresponding loop in the T. thermophilus protein. However, as mentioned earlier, even among eukaryotic HIcDHs, this is a very diver-gent region. It is relatively rich in charged residues (155KKED158) in the sequence of C. albicans HIcDH, has two such residues in the sequence of S. cerevisiae HIcDH (159DK160) and none in S. pombe sequence. The charged residues present in the C. albicans HIcDH 153–158 loop may participate in intersubunit interaction. These are not shown in our model resulting from the homology modeling based on the T. thermophilus template but may be present in the real protein, where the intermolecular distances might be shorter. Moreover, Tyr154 found in the C. albicans sequence, corresponds to Tyr132 of T. thermophilus but is replaced by Val140 in the sequence of S. pombe (Fig. 2). The actual quaternary structure of SpHIcDH is dimeric, whereas both the former proteins are tetramers. It is well known that the side chain of tyrosine is often involved in the quaternary structure stabilization of multimeric proteins, due to its ability to participate in polar, hydrophobic, and cation-π interactions (Brinda & Saraswathi, 2005). In the tetramer of T. thermophilus HIcDH, the hydroxyl moiety of Tyr132 forms a hydrogen bond with Asp139′ whereas its phenyl ring interacts with the hydrophobic patch formed by side chains of Val141′ and Val126′ (Nango et al., 2011). Similar interactions are preserved in our model of the C. albicans enzyme. The hydroxyl moiety of Tyr154 is involved in a strong hydrogen bond with Glu164′, whereas its phenyl ring is facing the guanidine moiety of Arg168′ and thus can be involved in the favorable cation-π type of interactions (Fig. 5). Interestingly, an analogous tyrosine residue can be identified in the sequence of S. cerevisiae HIcDH (Tyr157, Fig. 2) suggesting a possible tetrameric quaternary structure for this enzyme as well. On the other hand, the sequence corresponding to the CaHIcDH 153–158 loop in ScHIcDH contains fewer-charged residues than that of the former.

Figure 5.

Model of the tetrameric CaHIcDH (top) and close-up to the interface between subunits forming tetrameric structures of the CaHIcDH model (bottom left) and the TtHIcDH crystal structure (bottom right). Charged amino acids (Arg, Glu, Asp) proposed to be participating in the interactions are drawn as thin sticks. In the TtHIcDH model, Tyr132 and other interacting residues (Asp139′ and Val141′) are drawn as thick sticks. Respective Tyr154, Glu164′ and Arg168′ residues are shown in the CaHIcDH model. For the sake of clarity, only one set (out of four, one for each subunit of the tetramer) is shown.

Activity and substrate specificity of the enzyme

Enzymatic activity of purified His6CaHICDH was investigated using d,l-isocitrate (Ic), (2R,3S) homoisocitrate (HIc), and d,l-3-Isopropylmalic acid (3-IPM) as alternative substrates, and NAD+ as the oxidative substrate. The measured specific activity with HIc was 1.5 U mg−1, whereas with Ic and 3-IPM, the activity was not measurable (< 0.005). This finding suggests a strict substrate specificity of His6CaHIcDH for HIc, a feature shared by this enzyme with other HIcDHs from fungal sources known so far, i.e., ScHIcDH (Miyazaki et al., 2003), or SpHIcDH (Chen & Jeong, 2000). On the other hand, proteins form archebacteria demonstrating HIcDH activity seem to be the multifunctional proteins. Methanococcus jannaschii HIcDH (MjaHIcDH) not only catalyzes NAD+-dependent oxidation of Hic, but also that of (homo)nisocitrate (n = 2, 3) (Howell et al., 2000). The enzyme from T. thermophilus prefers isocitrate to homoisocitrate by a factor of 20 (Miyazaki et al., 2003), whereas for the HIcDH from Deinococcus radiodurans, homoisocitrate is preferred to isocitrate by 1.5-fold. DrHIcDH also catalyzes the oxidation of 3-isopropylmalate, but with the rate as low as 0.1% of that of the reaction with HIc (Miyazaki, 2005ab). It was shown that the possibility of using Ic as a substrate by T. thermophilus HIcDH was strongly dependent on presence of Arg85 and replacing this residue with Val prevented isocitrate binding and enhanced activity with HIc as a substrate (Miyazaki et al., 2003). According to the multiple sequence alignment, Arg85 of TtHIcDH corresponds to Val109 in the CaHIcDH sequence (Fig. 2), therefore, the strict substrate specificity of the latter enzyme version is not surprising.

Kinetic properties and effect of metal ions

The determined KM values of the purified His6CaHIcDH for NAD+ and homoisocitrate were 1.09 ± 0.08 mM and 73.7 ± 0.5 μM, respectively, whereas kcat was 0.38 s−1. Catalytic efficiency of this recombinant enzyme expressed as kcat/KM was 0.35 × 103 M−1 s−1 and 5.16 × 103 M−1 s−1, in respect to NAD+ and Hic respectively. For the wild-type enzyme, kcat was 0.40 s−1, kcat/KM (NAD+) = 0.44 × 103 M−1 s−1 and kcat/KM (HIc) = 8.75 × 103 M−1 s−1. These values could be compared to those determined for the oligoHis-tagged and wild-type ScHIcDH: KM(NAD) 0.45 mM, KM(MgHIc) 4.2 μM, kcat = 13 s−1 (Lin et al., 2008). This comparison indicates that CaHIcDH has much lower catalytic efficiency than its baker's yeast counterpart. Apparent identity of kinetic constants of the wild-type and oligoHis-tagged CaHIcDH prompted us to use the latter in further studies, as it could be easily purified to near homogeneity. Additional kinetic and oligomeric structure analysis of His6CaHIcDH suggest that homoisocitrate dehydrogenase from C. albicans might be like other dehydrogenases, i.e., a hysteretic enzyme. Direct recording of the increase in absorbance at 340 nm at low enzyme or HIc concentration revealed a lag in the reduction of NAD+. This effect was substantially reduced by pre-incubating the enzyme with HIc and Mg2+ and starting the reaction with NAD+. HIc or Mg2+ ions alone had no effect on the lag. Slow response of the enzyme, in terms of kinetic characteristics, with a rapid change of substrate concentration is probably related to the conformational change as the rate-limiting step in the enzymatic reaction. The same phenomenon was reported previously for other hysteretic enzymes (Frieden, 1970, 1979), including HIcDH of Saccharomycopsis lypolytica (Gaillardin et al., 1982).

Data accumulated so far for HIcDHs from different sources indicated that the enzyme is activated by K+, Mn2+,or Mg2+ ions (Gaillardin et al., 1982; Miyazaki, 2005ab; Yamamoto et al., 2007). In our hands, His6CaHIcDH was slightly active in the absence of K+ (0.4 U mg−1) but enzyme activity progressively increased upon addition of K+ ions, to reach 1.5 U mg−1 in the presence of 10-mM KCl, whereas addition of Na+ had no effect on enzyme activity (Fig. 6a). These properties are similar to those reported for the enzyme from S. cerevisiae (Lin et al., 2008) or S. lipolytica (Gaillardin et al., 1982). On the other hand, HIcDH from D. radiodurans was completely inactive in the absence of monovalent cations (Miyazaki, 2005ab).

Figure 6.

Effect of mono and divalent cations on His6CaHIcDH activity. (a) Effect of K+ (▲) and Na+ (■) in the presence of 5-mM Mg2+; (b) Effect of Mg2+ (●), Mn2+ (○) and Ca2+ (×) in the presence of 1 mM K+.

His6CaHIcDH was completely inactive in absence of Mg2+ or Mn2+ ions but the activity was restored upon addition of one of these cations, as illustrated in Fig. 6b. The rate of the reaction catalyzed by His6CaHIcDH was the highest in the presence of 5 mM Mg2+. Concentrations higher than 5 mM resulted in a slight inhibitory effect. In the presence of Mn2+, the rate of reaction peaked at 1 mM and dropped as the cation concentration increased. Addition of Ca2+ ions had no effect on enzyme activity.

Requirement of divalent cations, Mg2+ or Mn2+, is well known for HIcDHs from different sources. ScHIcDH, TthHIcDH, and HIcDH from S. lipolytica (SlHIcDH) require Mg2+ for their activity (Broquist, 1971; Gaillardin et al., 1982). Some enzyme versions, namely DraHIcDH, ScHIcDH, SlHIcDH, and HIcDH from Pyrococcus horikoshii, are active in the presence of both Mg2+ and Mn2+ (Rowley & Tucci, 1970; Gaillardin et al., 1982; Miyazaki, 2005ab). For DraHIcDH, the rate of reaction was the highest in the presence of 2 mM Mg2+. Concentrations higher than 2 mM did not increase relative activity, yet no inhibition was observed with Mg2+ concentrations up to 10 mM. In the presence of Mn2+, the rate of the reactions peaked at 0.2 mM and dropped as the cation concentrations rose above 1 mM (Miyazaki, 2005ab). This property is similar to that obtained for His6CaHIcDH.

Inhibition of CaHIcDH and antifungal effect of the malic acid analogue

Several specific inhibitors of HIcDHs from different sources have been reported. In the present studies, we analyzed the influence of (2R,3S)-3-(p-carboxybenzyl)malic acid (CBMA, Fig. 7) on His6CaHIcDH activity. It was previously found that the trimethyl ester of CBMA (TMCBMA) inhibited growth of Aspergillus nidulans in lysine-free minimal medium (Palmer et al., 2004). In the present study, we found that CBMA inhibited His6CaHIcDH activity with IC50 = 3.78 mM. Kinetic analysis of the enzyme inhibition revealed that this was competitive with respect to HIc, with Ki = 2.91 mM and noncompetitive with respect to NAD+ (Fig. 8). Comparison of the Ki/Km ratio = 39 for the His6CaHIcDH inhibition by CBMA with those noted previously for other HIcDH inhibitors tested against ScHIcDH indicates that CBMA may be regarded as a moderate inhibitor, similar to 3-hydroxypropylidene malate and 3-carboxypropylidene malate (Ki/Km = 28 and 4, respectively) (Yamamoto et al., 2007). Thiahomocitrate, the strongest HIcDH inhibitor known so far had Ki = 97 nM and Ki/Km = 5.4 × 10−3 (Yamamoto & Eguchi, 2008).

Figure 7.

Structure of CBMA.

Figure 8.

Kinetic analysis of inhibition of His6CaHIcDH by CBMA fitted to Lineweaver-Burk equation to determine the kinetic parameters for (a) HIc (b) NAD+. Concentration of CBMA: (■) 0 mM, (●) 1 mM, (▲) 2 mM, (▼) 4 mM.

Antifungal in vitro activity of CBMA and its trimethyl ester was determined against some human pathogenic fungi from the Candida genus and S. cerevisiae and results of this experiment are shown in Table 1. The ester derivative inhibited growth of all microorganisms tested except Candida glabrata in RPMI medium recommended by CLSI for broth-dilution tests of antifungal activity, with MIC values in the 0.5–2 mg mL−1 range. CBMA was inactive for most strains, although low activity was observed against C. dubliniensis and C. lusitaniae. A very similar pattern of activity was found in the minimal YNB medium. Growth inhibitory effect of TMCBMA in YNB was reversed when the medium was supplemented with 5 mM l-lysine, thus confirming that growth inhibition was due to the inhibition of AAP. It should be noted that the MIC of TMCBMA = 1 mg mL−1 corresponds to 3.3 mM, which when compared to the Ki value of CBMA makes it likely that HIcDH is the intracellular target for CBMA. Lower antifungal activity of CBMA in comparison with its trimethyl ester is probably a consequence of a less efficient uptake of the former. The fact that the antifungal activities of TMCBMA in RPMI and YNB are very similar is a bit surprising, as the former contains 0.27 mM l-lysine. This level of exogenous l-lys seems, therefore, insufficient to protect fungal cells against consequences of the HIcDH inhibition. This observation is promising for HIcDH as a potential target for antifungals, as the l-lys content in RPMI mimics the natural concentration level of this amino acid in the human blood serum. In contrast to this finding, it was previously demonstrated that the C. albicans auxotrophic mutants lacking functional genes encoding HCS (the enzyme catalyzing the first step of AAP) could grow in the RPMI 1640 medium (Kur et al., 2010) and HCS-deficient C. albicans and A. fumigatus mutants were found virulent in the disseminated fungemia murine models (Kur et al., 2010; Schöbel et al., 2010).

Table 1. Antifungal activity of CBMA and its methyl ester derivative
Minimum inhibitory concentration (MIC) (mg mL−1)
MediumRPMIYNBYNB + 5 mM l-Lys
C. krusei > 42> 42> 4> 4
C. albicans > 42212> 4
C. glabrata > 4> 4> 4> 4> 4> 4
C. tropicalis > 41> 41> 42
C. pseudotropicalis > 41> 42> 4> 4
C. dubliniensis 222122
C. lusitaniae 21> 42> 4> 4
S. cerevisiae > 40.5> 42> 4> 4

Molecular modeling of CaHIcDH: inhibitor interaction

Interactions of HIcDH with its substrate and CBMA inhibitor were studied by docking simulations. The putative geometry of the complex of (2S,3S)-thiahomoisocitrate with CaHIcDH was deduced by superimposing the 3asj structure of the TtHIcDH:inhibitor complex and our model of the C. albicans enzyme. This model was further improved by introduction of the magnesium ion located at the most probable site at the active center. The docking simulations were performed to this modeled receptor. The most preferred ligand positions (90% of results) obtained in this way closely resembled the one found in the crystal structure (Fig. 9), additionally shedding light on the important role of magnesium ion in the process of substrate or substrate analog binding at the active site of HIcDH.

Figure 9.

Docking of HIc and CBMA in complex with Mg(II) into the active site of the CaHIcDH model. Left – the most abundant pose of homoisocitrate; Right – the most abundant pose of CBMA docked to the same receptor. Magnesium ion is shown as a pink sphere. Transparent ball and stick representations show the position of superimposed Arg85 and Glu127 residues found in the binding site of the TtHIcDH.

Homoisocitrate binds with its 1-hydroxyl and 1-carboxylate groups chelating the magnesium ion, with the former group pointing toward the hydroxyl of Tyr147 and the 2-carboxylate group bound to the guanidine moiety of Arg140. These interactions, as well as all residues involved, closely resemble the pattern of interactions characteristic for the β-decarboxylating dehydrogenases and malate-binding enzymes (Miyazaki et al., 2003). Interestingly, the remaining part of the substrate chain with its 4-carboxylate group extends toward the ε-amino group of the Lys149 side chain and is locked in this position by strong electrostatic interactions between these two oppositely charged moieties. The CBMA molecule is significantly larger than that of homoisocitrate. The inhibitor cannot easily fit between Lys149 and magnesium in the same conformation as the substrate HIc does, thus its malate part binds in a slightly different mode. In the most abundant docking position, the 1-hydroxy and 2-carboxylate groups bind the magnesium ion whereas the 1-carboxylate of the inhibitor is bound by Arg140 and Arg112 guanidine moieties. With such adaptation of the malate group position (with respect to its orientation in the bound substrate), the benzoate carboxyl group can be still extended toward the amino group of Lys149 and is locked in this conformation by electrostatic interactions with these oppositely charged groups.

In conclusion, the LYS12 gene of C. albicans was unequivocally identified as coding for homoisocitrate dehydrogenase. The gene was cloned, sequenced, and overexpressed in E. coli as N-his-tagged fusion protein. His6CaHIcDH is the first HIcDH from human pathogenic fungi that was isolated and characterized. The enzyme was purified by metal affinity chromatography that yielded an active and near homogenous homoisocitrate dehydrogenase. Similar to Tth-HICDH and Dra-HICDH, purified, recombinant His6CaHIcDH was tetramer, required K+ and Mg2+ or Mn2+ for activity, and used NAD+ as a coenzyme. Our results indicate that (2R,3S)-3-(p-carboxybenzyl)malic acid was a competitive inhibitor with respect to HIc and noncompetitive with respect to NAD+. The inhibitory effect of CBMA on the enzyme and antifungal growth inhibitory activity of the CBMA trimethyl ester in lysine-free YNB minimal medium, partially reversed when lysine was added, suggest that CaHIcDH is the actual molecular target for CBMA in C. albicans cells.


Financial support of these studies by the Foundation for Polish Science is gratefully acknowledged.