MPA (mycophenolic acid) is an immunosuppressive drug produced by several fungi in Penicillium subgenus Penicillium. This toxic metabolite is an inhibitor of IMPDH (IMP dehydrogenase). The MPA-biosynthetic cluster of Penicillium brevicompactum contains a gene encoding a B-type IMPDH, IMPDH-B, which confers MPA resistance. Surprisingly, all members of the subgenus Penicillium contain genes encoding IMPDHs of both the A and B types, regardless of their ability to produce MPA. Duplication of the IMPDH gene occurred before and independently of the acquisition of the MPAbiosynthetic cluster. Both P. brevicompactum IMPDHs are MPA-resistant, whereas the IMPDHs from a non-producer are MPA-sensitive. Resistance comes with a catalytic cost: whereas P. brevicompactum IMPDH-B is >1000-fold more resistant to MPA than a typical eukaryotic IMPDH, its kcat/Km value is 0.5% of ‘normal’. Curiously, IMPDH-B of Penicillium chrysogenum, which does not produce MPA, is also a very poor enzyme. The MPA-binding site is completely conserved among sensitive and resistant IMPDHs. Mutational analysis shows that the C-terminal segment is a major structural determinant of resistance. These observations suggest that the duplication of the IMPDH gene in the subgenus Penicillium was permissive for MPA production and that MPA production created a selective pressure on IMPDH evolution. Perhaps MPA production rescued IMPDH-B from deleterious genetic drift.

INTRODUCTION

Filamentous fungi produce a vast arsenal of toxic natural products that require the presence of corresponding defence mechanisms to avoid self-intoxication. The importance of these defence mechanisms is demonstrated by the presence of resistance genes within biosynthetic gene clusters, yet how production and resistance co-evolved is poorly understood. Insights into the inhibition of enzymes involved in self-resistance provide an intriguing strategy for the development of antifungal agents. Furthermore, the elucidation of the defence mechanisms is required for the design of heterologous cell factories producing bioactive compounds.

Self-resistance can involve expression of a target protein that is impervious to the toxic natural product, which suggests that resistance originates from a gene-duplication event. The biosynthetic cluster of the immunosuppressive drug MPA (mycophenolic acid) offers an intriguing example of this phenomenon. MPA is a well-characterized inhibitor of IMPDH (IMP dehydrogenase) [1], and the Penicillium brevicompactum MPA-biosynthetic cluster contains a second type of IMPDH gene that confers MPA resistance (IMPDH-B/mpaF) [2,3]. Curiously, both MPA producers and non-producers in Penicillium subgenus Penicillium contain two IMPDH genes encoding IMPDH-A and IMPDH-B [3]. Phylogenetic analysis suggests that the gene-duplication event occurred before or simultaneous with the radiation of Penicillium subgenus Penicillium [2,3]. How this gene-duplication event influenced the acquisition of MPA biosynthesis is not understood.

In the present study, we investigated the relationship between MPA production, MPA resistance and the properties of IMPDH-A and IMPDH-B. Whereas IMPDH-B from the producer organism P. brevicompactum is extraordinarily resistant to MPA, IMPDH-B from the non-producer Penicillium chrysogenum displays typical sensitivity. Both IMPDH-Bs are very poor enzymes, but P. chrysogenum IMPDH-B is almost non-functional. These observations suggest that acquisition of the MPA biosynthetic cluster may have rescued IMPDH-B from deleterious genetic drift.

EXPERIMENTAL

Fungal strains

P. brevicompactum IBT23078, P. chrysogenum IBT5857, Aspergillus nidulans IBT27263, and the 11 strains listed in Supplementary Table S1 (at http://www.BiochemJ.org/bj/441/bj4410219add.htm), were grown on CYA (Czapek yeast extract agar) at 25°C [4].

MPA treatment of fungi

Spores from P. brevicompactum IBT23078, P. chrysogenum IBT5857 and A. nidulans IBT27263 were harvested and suspended in sterile water. Then, 10 μl of serial 10-fold spore dilutions were spotted on to CYA plates with or without 200 μg/ml MPA. All plates contained 0.8% methanol. Stock solution (25 mg/ml MPA in methanol) and MPA plates were made briefly before the spores were spotted on to the plates in order to avoid MPA degradation. MPA was acquired from Sigma.

RNA purification and cDNA synthesis

Spores from P. brevicompactum IBT23078 were harvested and used to inoculate 200 ml of YES (yeast extract sucrose) medium in 300 ml shake flasks without baffles [4]. P. brevicompactum was grown at 25°C and 150 rev./min shaking. After 48 h, the mycelium was harvested and RNA was purified using the Fungal RNA purification Miniprep Kit (E.Z.N.A.) and cDNA was synthesized from the RNA using Finnzymes Phusion™ RT (reverse transcription)–PCR kit following the instructions of the two manufacturers.

Plasmid construction

Constructs for expression of His6-tagged IMPDHs in Escherichia coli were created by inserting the IMPDH coding sequences into pET28a that had been converted into a USER-compatible vector (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/441/bj4410219add.htm). pET28U was created by PCR-amplifying pET28 with the primers BGHA527/BGHA528 followed by DpnI treatment to remove the PCR template (see Supplementary Figure S1). The P. brevicompactum IMPDH-B (PbIMPDH-B) gene was amplified from cDNA from P. brevicompactum. Genes encoding A. nidulans IMPDH-A (AnImdA), P. brevicompactum IMPDH-A (PbIMPDH-A), P. chrysogenum IMPDH-A (PcIMPDH-A) and P. chrysogenum IMPDH-B (PcIMPDH-B) were obtained from gDNA (genomic DNA). In all cases, the three exons of each gene were individually PCR-amplified and purified and subsequently USER-fused using a method described previously [5,6].

The expression constructs for chimaeric and mutated IMPDHs were created by fusing the two parts of the IMPDHs using the USER fusion method to introduce the mutation in the primer tail [6]. Chimaeric IMPDH of PbIMPDH-B (N-terminus) and PbIMPDH-A (C-terminus) was created by USER-fusing two P. brevicompactum IMPDH-B and IMPDH-A fragments, which were PCR-amplified with primer pairs BGHA529/BGHA667 and BGHA668/BGHA540 respectively, into pET28U as described above. Similarly, chimaeric IMPDH of PbIMPDH-A (N-terminus) and PbIMPDH-B (C-terminus) was created by amplifying PbIMPDH-A and PbIMPDH-B fragments with primer pairs BGHA539/BGHA669 and BGHA670/BGHA530 respectively, followed by USER-fusing into pET28U. Using gDNA from the proper source, the Y411F (Chinese-hamster ovary IMPDH2 numbering) mutation was introduced into A. nidulans IMPDH-A and into P. brevicompactum IMPDH-A and IMPDH-B by amplifying and fusing two IMPDH fragments with primer pairs BGHA545/BGHA359 and BGHA546/BGHA358; BGHA539/BGHA455 and BGHA540/BGHA454; and BGHA529/BGHA361 and BGHA530/BGHA360 respectively. For details of primers, see Supplementary Tables S2 and S3 (at http://www.BiochemJ.org/bj/441/bj4410219add.htm). The PfuX7 polymerase was used for PCR amplification in all cases [7].

Cladistic analysis

Alignment of DNA coding regions and protein were performed with ClustalW implemented in the CLC DNA Workbench 6 program (CLC bio) using the following parameters: gap open cost=6.0; gap extension cost=1.0; and end gap cost=free. A cladogram was constructed with the same software using the neighbour-joining method and 1000 bootstrap replicates [33]. The DNA sequence of IMPDH and β-tubulin from selected fungi were either generated by dPCR (degenerate PCR) (see Supplementary Table S1) or retrieved from the NCBI: A. nidulans β-tubulin (GenBank® accession number XM_653694) and IMPDH-A DNA sequence [GenBank® accession number BN001302 (ANIA_10476)]; Coccidioides immitis β-tubulin (GenBank® accession number XM_001243031) and IMPDH-A DNA sequence (GenBank® accession number XM_001245054); Penicillium bialowiezense IBT21578 β-tubulin (GenBank® accession number JF302653), IMPDH-A DNA sequence (GenBank® accession number JF302658) and IMPDH-B DNA sequence (GenBank® accession number JF302662); P. brevicompactum IBT23078 β-tubulin (GenBank® accession number JF302653) and IMPDH-A DNA sequence (GenBank® accession number JF302657); Penicillium carneum IBT3472 β-tubulin (GenBank® accession number JF302650), IMPDH-A DNA sequence (GenBank® accession number JF302656) and IMPDH-B DNA sequence (GenBank® accession number JF302660); P. chrysogenum IBT5857 β-tubulin (GenBank® accession number XM_002559715), IMPDH-A DNA sequence (GenBank® accession number XM_002562313) and IMPDH-B DNA sequence (GenBank® accession number XM_002559146); Penicillium paneum IBT21729 β-tubulin (GenBank® accession number JF302651), IMPDH-A DNA sequence (GenBank® accession number JF302654) and IMPDH-B DNA sequence (GenBank® accession number JF302661); and Penicillium roqueforti IBT16406 β-tubulin (GenBank® accession number JF302649), IMPDH-A DNA sequence (GenBank® accession number JF302655) and IMPDH-B DNA sequence (GenBank® accession number JF302659). The MPA gene cluster sequence from P. brevicompactum IBT23078, which contains the IMPDH-B gene sequence (mpaF) is available from GenBank® under accession number HQ731031. Full-length protein sequences were retrieved from the NCBI: Candida albicans: (GenBank® accession number EEQ46038), Chinese-hamster ovary IMPDH2 (GenBank® accession number P12269), E. coli (GenBank® accession number ACI80035), human IMPDH type 2 (GenBank® accession num-ber NP_000875), P. chrysogenum IMPDH-A gene (GenBank® accession number CAP94756), IMPDH-B (GenBank® accession number CAP91832), yeast IMD2 (GenBank® accession number P38697), IMD3 (GenBank® accession number P50095), and IMD4 (GenBank® accession number P50094).

dPCR

Primers and PCR conditions for amplifying part of the genes encoding IMPDH-A, IMPDH-B and β-tubulin were as described in [3]. Genomic DNA from 11 fungi from the Penicillium subgenus Penicillium subclade (see Supplementary Table S1) were extracted using the FastDNA® SPIN for Soil Kit (MP Biomedicals). PCR primer pairs BGHA236HC/BGHA246HC or BGHA531/BGHA532 were used to amplify DNA from the gene encoding IMPDH-A. The primer pair BGHA240HC/BGHA241HC was used to amplify DNA from the gene encoding IMPDH-B, and the primer pair BGHA343/BGHA344 was used to amplify β-tubulin.

Expression and purification of His6-tagged proteins

Plasmids were expressed in a ΔguaB derivative of E. coli BL21(DE3) [8]. Cells were grown in LB (Luria–Bertani) medium at 30°C until reaching a D600 of 1.0 for AnImdA and 1.5 for all other enzymes, and induced overnight at 16°C with 0.5 mM IPTG (isopropyl β-D-thiogalactopyranoside) for AnImdA and 0.1 mM IPTG for all other enzymes. Cells were harvested by centrifugation at 3000 g for 20 min and resuspended in buffer (pH 8.0) containing 20 mM sodium phosphate, 500 mM NaCl, 5 mM 2-mercaptoethanol, 5 mM imidazole and Complete™ protease inhibitor cocktail (Roche Diagnostics). Cell lysates were prepared by sonication followed by centrifugation at 40000 g for 30 min. All enzymes were purified using a HisTrap affinity column (GE Healthcare) on an ÄKTA Purifier (GE Healthcare) and dialysed into buffer containing 50 mM Tris/HCl, pH 8.0, 100 mM KCl, 1 mM DTT (dithiothreitol) and 10% glycerol. All enzymes were purified to >90% purity as determined by SDS/PAGE (12% gels), with the exception of PcIMPDH-B, which was only partially purified to ~45% purity due to poor expression and stability (see Supplementary Figure S2 at http://www.BiochemJ.org/bj/441/bj4410219add.htm).

Enzyme concentration determination

Enzyme concentration was determined using the Bio-Rad assay according to the manufacturer's instructions using IgG as a standard. The Bio-Rad assay overestimates the IMPDH concentration by a factor of 2.6 [9]. Concentration of active enzyme was determined by MPA titration using an equation for tight binding inhibition (eqn 1):

 
formula
(1)

Enzyme assays

Standard IMPDH assay buffer consisted of 50 mM Tris/HCl (pH 8.0), 100 mM KCl, 1 mM DTT, and various concentrations of IMP and NAD+ (concentrations used for each enzyme are listed in Supplementary Table S4 at http://www.BiochemJ.org/bj/441/bj4410219add.htm. Enzyme activity was measured by monitoring the production of NADH by changes in absorbance increase at 340 nm on a Cary Bio-100 UV–visible spectrophotometer at 25°C (ϵ340=6.2 mM−1·cm−1). MPA inhibition assays were performed at saturating IMP and half-saturating NAD+ concentrations to avoid complications arising from NAD+ substrate inhibition. IMP-independent reduction of NAD+ was observed in the partially purified preparation of PcIMPDH-B at ~25% of the rate in the presence of IMP. The rate of this background reaction was subtracted from the rate with IMP for each MPA concentration. Initial velocity data were obtained by fitting to either the Michaelis–Menten equation (eqn 2) or an uncompetitive substrate inhibition equation (eqn 3) using SigmaPlot (Systat Software).

 
formula
(2)
 
formula
(3)

Bacterial complementation assays

Cultures were grown overnight in LB medium with 25 μg/ml kanamycin. Then, 5 μl of 1:20 serial dilutions was plated on M9 minimal medium containing 0.5% casamino acids, 100 μg/ml L-tryptophan, 0.1% thiamine, 25 μg/ml kanamycin and 66 μM IPTG. Guanosine-supplemented plates contained 50 μg/ml guanosine. MPA plates contained 100 μM MPA. Plates were incubated overnight at 37°C and then grown at room temperature (25°C).

Sequences generated in the present study

See Supplementary Table S1 for GenBank® accession numbers.

RESULTS

The presence of IMPDH-B is not linked to MPA biosynthesis

The first gene encoding IMPDH-B was identified in P. brevicompactum where it is part of the MPA biosynthetic cluster and confers MPA resistance (P. brevicompactum IMPDH-B is also known as mpaF) [2]. Of the full genome sequences currently available for filamentous fungi, only P. chrysogenum (the sole representative of Penicillium subgenus Penicillium) contains an IMPDH-B gene. We previously analysed four other fungi in Penicillium subgenus Penicillium for the presence of IMPDH genes, and all contain both IMPDH-A and IMPDH-B [3]. We have now expanded this search to include 11 additional Penicillium strains of the Penicillium subgenus. dPCR analysis revealed that all 11 strains contain both genes encoding IMPDH-A and IMPDH-B (see Supplementary Table S1 and Supplementary Figure S3 at http://www.BiochemJ.org/bj/441/bj4410219add.htm), suggesting that the presence of IMPDH-B is widespread in species of Penicillium subgenus Penicillium, even though many of these strains do not produce MPA [10].

It is possible that these MPA non-producer strains may contain latent or cryptic MPA biosynthetic genes. A previously described example is that although P. chrysogenum does not produce viridicatumtoxin, griseofulvin and tryptoquialanine, it does contain sequence regions with high identity with these biosynthetic clusters, which suggests that these gene functions were lost during evolution [11,12]. Likewise, P. chrysogenum does not produce MPA. However, in this case, BLAST searches failed to identify any genes with significant sequence identity with the MPA biosynthesis gene cluster in P. chrysogenum. These searches did identify an IMPDH-B pseudogene in P. chrysogenum. This 260 bp gene fragment has highest identity with IMPDH-B from P. chrysogenum (77%) and 64% identity with IMPDH-B from P. brevicompactum.

We analysed the regions surrounding the MPA biosynthetic cluster in P. brevicompactum and used BLAST searches to identify similar genes in P. chrysogenum. In this way, adjacent genes that are highly similar and syntenic to genes found in P. chrysogenum were identified (Figure 1). Interestingly, the genes flanking one side of the cluster correspond to a locus on contig Pc0022, while the other side of the cluster corresponds to a locus located at another contig, Pc0021. Regions further away from the MPA cluster in P. brevicompactum contain additional homologues of P. chrysogenum genes, but these genes are located in many different contigs (see Supplementary Table S5 at http://www.BiochemJ.org/bj/441/bj4410219add.htm). P. chrysogenum contains small sequence regions with high identity with the flanking region of the MPA cluster. This identity is in the promoter regions and probably represents conserved regulatory regions. Despite a detailed search, we found no sequence similarity to the regions around the PcIMPDH-B and the pseudo IMPDH-B (Pc22g07570) (results not shown). Together, these findings suggest that the MPA cluster has never been present in P. chrysogenum and that a high degree of genome shuffling has occurred since the divergence of P. brevicompactum and P. chrysogenum.

Flanking regions of the MPA-biosynthesis gene cluster

Figure 1
Flanking regions of the MPA-biosynthesis gene cluster

P. brevicompactum and the corresponding syntenic loci on P. chrysogenum contigs Pc0022 and Pc0021 are shown. Regions marked with (*) indicate short conserved sequences (length <53 bp, identity >86%).

Figure 1
Flanking regions of the MPA-biosynthesis gene cluster

P. brevicompactum and the corresponding syntenic loci on P. chrysogenum contigs Pc0022 and Pc0021 are shown. Regions marked with (*) indicate short conserved sequences (length <53 bp, identity >86%).

PbIMPDH-B is MPA-resistant, but PcIMPDH-B is MPA-sensitive

The PbIMPDH-B gene confers MPA resistance on A. nidulans, suggesting that MPA resistance might be a common feature of all IMPDH-B proteins [3]. Consistent with this hypothesis, P. chrysogenum is less sensitive to MPA than A. nidulans, although significantly more sensitive than P. brevicompactum (Figure 2A). These observations suggest that P. chrysogenum IMPDH-B may be semi-resistant to MPA.

MPA resistance in fungi and bacteria expressing fungal IMPDHs

Figure 2
MPA resistance in fungi and bacteria expressing fungal IMPDHs

(A) Spot assay to determine sensitivity towards MPA. A 10-fold dilution series of spores from A. nidulans, P. chrysogenum and P. brevicompactum were grown on CYA plates containing no additives or 6.2 μM MPA. The spot with the highest number of spores in each row contains ~106 spores. (B) E. coli ΔguaB was cultured on minimal medium, with no additions or supplemented with 100 μM MPA or 50 μg/ml guanosine (Guo) and 100 μM MPA.

Figure 2
MPA resistance in fungi and bacteria expressing fungal IMPDHs

(A) Spot assay to determine sensitivity towards MPA. A 10-fold dilution series of spores from A. nidulans, P. chrysogenum and P. brevicompactum were grown on CYA plates containing no additives or 6.2 μM MPA. The spot with the highest number of spores in each row contains ~106 spores. (B) E. coli ΔguaB was cultured on minimal medium, with no additions or supplemented with 100 μM MPA or 50 μg/ml guanosine (Guo) and 100 μM MPA.

To investigate this hypothesis further, IMPDHs from P. brevicompactum, P. chrysogenum and A. nidulans were heterologously expressed as N-terminal His6-tagged proteins in an E. coli ΔguaB strain that lacks endogenous IMPDH; these enzymes are denoted PbIMPDH-A, PbIMPDH-B, PcIMPDH-A, PcIMPDH-B and AnImdA (Figure 2B). All strains grow on minimal medium in the presence of guanosine (Figure 2B). Bacteria expressing AnImdA, PbIMPDH-A, PbIMPDH-B and PcIMPDH-A grow on minimal medium in the absence of guanosine, indicating that active IMPDHs were expressed in all four cases. Only bacteria expressing PbIMPDH-B were resistant to MPA (Figure 2B). PcIMPDH-B failed to grow on minimal medium in the absence of guanosine, suggesting that this protein was not able to complement the guaB deletion. This observation suggests that PcIMPDH-B may be non-functional.

All five fungal IMPDHs were isolated to investigate further the molecular basis of MPA resistance. With the exception of PcIMPDH-B, all enzymes were purified to >90% purity in one step using a nickel-affinity column (see Supplementary Figure S2). PcIMPDH-B was unstable and could only be purified to ~45% homogeneity. Western blot analysis using anti-(human IMPDH2) polyclonal antibodies showed that PcIMPDH-B had undergone proteolysis. This degradation could not be prevented by the inclusion of protease inhibitor cocktails during the purification. We performed MPA titrations to determine the concentration of active enzyme in all five protein samples. MPA traps an intermediate in the IMPDH reaction, so that only active enzyme will bind MPA [13]. All of the proteins were ~100% active, with the exception of PcIMPDH-B, which was 48% active, as expected. The oligomeric states of AnImdA, PbIMPDH-A and PbIMPDH-B were examined by gel-filtration chromatography. All three enzymes eluted at a molecular mass of approximately 220 kDa, consistent with being tetramers (see Supplementary Figure S3).

AnImdA, PcIMPDH-A and PcIMPDH-B are all sensitive to MPA, with values of IC50 ranging from 26 to 60 nM (Figure 3 and Table 1). Similar MPA inhibition has been reported for the mammalian and C. albicans enzymes [13,14]. These observations indicate that MPA resistance is not a general property of IMPDH-Bs.

MPA inhibition of fungal IMPDHs

Figure 3
MPA inhibition of fungal IMPDHs

●, AnImdA; ○, PbIMPDH-A; ▼, PbIMPDH-B; △, PcIMPDH-A; ■, PcIMPDH-B.

Figure 3
MPA inhibition of fungal IMPDHs

●, AnImdA; ○, PbIMPDH-A; ▼, PbIMPDH-B; △, PcIMPDH-A; ■, PcIMPDH-B.

Table 1
Characterization of IMPDHs from filamentous fungi

Data were analysed as described in the Experimental section. n.a., not applicable. PbIMPDH-B and PcIMPDH-B did not show NAD+ substrate inhibition up to 5 mM. The concentrations of IMP and NAD+ used for IC50 determination for each enzyme is included in Supplementary Table S4 at http://www.BiochemJ.org/bj/441/bj4410219add.htm.

Enzyme IC50 (MPA) (μM) kcat (s−1Km (IMP) (μM) kcat/Km (IMP) (M−1·s−1Km (NAD+) (μM) kcat/Km (NAD) (M−1·s−1Kii (NAD+) (mM) 
AnImdA 0.026±0.002 0.74±0.06 10±1 74000±8000 170±70 4400±100 1.5±0.6 
PbIMPDH-A 0.43±0.03 0.70±0.08 130±30 5400±1200 340±90 2100±900 4.7±1.1 
PbIMPDH-B 27±9 0.41±0.02 1400±200 300±80 790±140 500±100 n.a. 
PcIMPDH-A 0.036±0.004 0.84±0.03 40±6 21000±2000 290±60 2900±300 2.4±0.5 
PcIMPDH-B 0.06±0.01 0.0075±0.001 600±300 12±3 640±130 11±6 n.a. 
Enzyme IC50 (MPA) (μM) kcat (s−1Km (IMP) (μM) kcat/Km (IMP) (M−1·s−1Km (NAD+) (μM) kcat/Km (NAD) (M−1·s−1Kii (NAD+) (mM) 
AnImdA 0.026±0.002 0.74±0.06 10±1 74000±8000 170±70 4400±100 1.5±0.6 
PbIMPDH-A 0.43±0.03 0.70±0.08 130±30 5400±1200 340±90 2100±900 4.7±1.1 
PbIMPDH-B 27±9 0.41±0.02 1400±200 300±80 790±140 500±100 n.a. 
PcIMPDH-A 0.036±0.004 0.84±0.03 40±6 21000±2000 290±60 2900±300 2.4±0.5 
PcIMPDH-B 0.06±0.01 0.0075±0.001 600±300 12±3 640±130 11±6 n.a. 

MPA is a poor inhibitor of both PbIMPDH-A and PbIMPDH-B (Figure 3 and Table 1). PbIMPDH-A is 10-fold more resistant than IMPDHs from the other filamentous fungi whereas PbIMPDH-B is a remarkable ~1000-fold more resistant. Thus the ability of P. brevicompactum to proliferate while producing MPA can be explained by the intrinsic resistance of its target enzymes. These observations suggest that MPA production creates a selective pressure on IMPDH evolution. Curiously, similar IC50 values have been reported for Imd2 and Imd3 from Saccharomyces cerevisiae (IC50=70 μM and 500 nM for Imd2 and Imd3 respectively [15]). These enzymes are the only other examples of naturally occurring MPA-resistant IMPDHs from eukaryotes. Cladistic analysis indicates that the S. cerevisiae IMPDHs arose in a distinct gene-duplication event (Figure 4). Therefore P. brevicompactum IMPDH-B is unlikely to share a common ancestry with S. cerevisiae Imd2.

IMPDH phylogeny

Figure 4
IMPDH phylogeny

(A) Rooted cladogram based on full-length amino acid sequences. (B and C) Rooted cladograms based on (B) IMPDH cDNA sequences (372–387 bp) and (C) β-tubulin cDNA sequences (950 bp) from eight species from Penicillium subgenus Penicillium and two filamentous fungi outside. P., Penicillium and A., Aspergillus. Bootstrap values (expressed as percentages of 1000 replications) are shown at the branch points, and the scale represents changes per unit. In (B), MPA production is indicated by + or −. The subclades with Penicillium subgenus Penicillium genes are marked in light grey (IMPDH-A) and dark grey (IMPDH-B). E. coli has been used as an outgroup in (A), and C. immitis as an outgroup in (B) and (C).

Figure 4
IMPDH phylogeny

(A) Rooted cladogram based on full-length amino acid sequences. (B and C) Rooted cladograms based on (B) IMPDH cDNA sequences (372–387 bp) and (C) β-tubulin cDNA sequences (950 bp) from eight species from Penicillium subgenus Penicillium and two filamentous fungi outside. P., Penicillium and A., Aspergillus. Bootstrap values (expressed as percentages of 1000 replications) are shown at the branch points, and the scale represents changes per unit. In (B), MPA production is indicated by + or −. The subclades with Penicillium subgenus Penicillium genes are marked in light grey (IMPDH-A) and dark grey (IMPDH-B). E. coli has been used as an outgroup in (A), and C. immitis as an outgroup in (B) and (C).

IMPDH-Bs are very poor IMPDHs

The reactions of the fungal IMPDHs were characterized to determine how MPA resistance affects enzymatic function. The values of kcat are similar for AnImdA, PcIMPDH-A, PbIMPDH-A and PbIMPDH-B, and generally typical for eukaryotic IMPDHs [13] (Table 1). Typical values of Km (NAD+) and Km (IMP) are also observed for AnImdA and PcIMPDH-A; the values of Km for PbIMPDH-A are high, but not unprecedented. In contrast, the value of Km (IMP) for PbIMPDH-B (1.4 mM) is ~100-fold higher than the highest value reported for any other IMPDH. Like PbIMPDH-B, PcIMPDH-B exhibits an abnormally high value of Km (IMP). The value of kcat of PcIMPDH-B is more than a factor of 100 less than those of the other fungal IMPDHs. The corresponding values of kcat/Km are the lowest ever observed for an IMPDH by factors of 1000 (IMP) and 200 (NAD+) (see [13] for compilation of values). Therefore both IMPDH-Bs are inferior enzymes.

Identification of a major determinant of MPA resistance

MPA inhibits IMPDHs by an unusual mechanism [16]. The IMPDH reaction proceeds via the covalent intermediate E-XMP*, which forms when the active-site cysteine residue attacks the C2 position of IMP, transferring a hydride to NAD+ [13]. The resulting NADH dissociates, allowing MPA to bind to E-XMP*, preventing hydrolysis to form XMP. The crystal structure of the E-XMP*–MPA complex has been solved for Chinese-hamster ovary IMPDH type 2 [17], revealing that MPA stacks against the purine ring of E-XMP* in a similar manner to the nicotinamide of NAD+. Surprisingly, all of the residues within 4 Å (1 Å=0.1 nm) of MPA are completely conserved in the fungal IMPDHs (Figure 5). Likewise, the IMP-binding site is also highly conserved. Only residue 411 is variable (Chinese-hamster ovary IMPDH2 numbering); the IMPDH-A enzymes contain a tyrosine residue at this site, whereas IMPDH-B enzymes have phenylalanine residues (Figure 5). Substitution of phenylalanine for Tyr411 did not increase the MPA resistance of PbIMPDH-A, nor did the substitution of tyrosine for Phe411 decrease the MPA resistance of PbIMPDH-B (Table 2). Therefore the structural determinants of MPA resistance must reside outside the active site.

Table 2
MPA inhibition of fungal IMPDH variants

Phe411 is replaced with tyrosine in IMPDH-B and Tyr411 is replaced by phenylalanine in IMPDH-A*. nd, no data.

 IC50 (μM) 
Enzyme Wild-type 411* C-terminal swap C-terminal swap +411* 
PbIMPDH-A 0.43±0.03 0.16±0.02 3±0.2 4.1±0.3 
PbIMPDH-B 27±9 25±2 4.1±0.3 3.1±0.3 
AnImdA 0.026±0.002 0.026±0.002 nd nd 
 IC50 (μM) 
Enzyme Wild-type 411* C-terminal swap C-terminal swap +411* 
PbIMPDH-A 0.43±0.03 0.16±0.02 3±0.2 4.1±0.3 
PbIMPDH-B 27±9 25±2 4.1±0.3 3.1±0.3 
AnImdA 0.026±0.002 0.026±0.002 nd nd 

The MPA-binding site is conserved among drug-sensitive and drug-resistant eukaryotic IMPDHs

Figure 5
The MPA-binding site is conserved among drug-sensitive and drug-resistant eukaryotic IMPDHs

(A) Amino acid polymorphism in IMPDH of A. nidulans (An), P. brevicompactum (Pb) and P. chrysogenum (Pc). Shown are residues within 4 Å of E-XMP* (XMP*) and MPA. Residues in red are polymorphic between IMPDH-A and IMPDH-B. (B) An alignment of IMPDH-A and IMPDH-B from P. brevicompactum 23078 and P. chrysogenum, IMPDH-A from A. nidulans and IMPDH2 from Chinese-hamster ovary (CHO) was mapped on to the X-ray crystal structure of the MPA complex of Chinese-hamster ovary IMPDH (PDB code 1JR1) [17]. Key: 100% conserved, dark magenta; 58% conserved, tan; 17% conserved, dark cyan. MPA, green ball and stick; IMP intermediate, grey ball and stick; K+, gold; residues within 4 Å of MPA and IMP are shown in stick form. (C) Alignment of the C-terminal region. The region aligned represents the fragments that were swapped in the chimaeric PbIMPDH-A and PbIMPDH-B.

Figure 5
The MPA-binding site is conserved among drug-sensitive and drug-resistant eukaryotic IMPDHs

(A) Amino acid polymorphism in IMPDH of A. nidulans (An), P. brevicompactum (Pb) and P. chrysogenum (Pc). Shown are residues within 4 Å of E-XMP* (XMP*) and MPA. Residues in red are polymorphic between IMPDH-A and IMPDH-B. (B) An alignment of IMPDH-A and IMPDH-B from P. brevicompactum 23078 and P. chrysogenum, IMPDH-A from A. nidulans and IMPDH2 from Chinese-hamster ovary (CHO) was mapped on to the X-ray crystal structure of the MPA complex of Chinese-hamster ovary IMPDH (PDB code 1JR1) [17]. Key: 100% conserved, dark magenta; 58% conserved, tan; 17% conserved, dark cyan. MPA, green ball and stick; IMP intermediate, grey ball and stick; K+, gold; residues within 4 Å of MPA and IMP are shown in stick form. (C) Alignment of the C-terminal region. The region aligned represents the fragments that were swapped in the chimaeric PbIMPDH-A and PbIMPDH-B.

Inspection of the structure suggested that the C-terminal segment (residues 498–527) was another likely candidate for a structural determinant of MPA resistance (Figure 5). This region includes part of the site that binds a monovalent cation activator, as well as segments that interact with the active site residues. Deletion of this segment inactivates the enzyme [18]. Swapping the 30-residue C-terminal segments revealed that this region is responsible for ~7 of the 60-fold difference in MPA resistance between PbIMPDH-A and PbIMPDH-B (see Table 2). Additional structural determinants of MPA resistance remain to be identified.

DISCUSSION

Phylogenetic analysis indicates that the IMPDH gene was duplicated before the radiation of Penicillium subgenus Penicillium (Figure 4 and see Supplementary Figure S4 at http://www.BiochemJ.org/bj/441/bj4410219add.htm). No remnants of the MPA biosynthetic cluster are present in P. chrysogenum, which suggests that the ability to produce MPA arrived after duplication of the IMPDH genes. Consistent with this hypothesis, both IMPDHs from the MPA-producer P. brevicompactum are more resistant to MPA than the IMPDHs from A. nidulans and P. chrysogenum. Thus both P. brevicompactum IMPDH-A and IMPDH-B have evolved in response to the presence of MPA. PbIMPDH-B is 1000-fold more resistant than typical eukaryotic IMPDHs. The structural basis of these functional differences is not readily discernible. Whereas prokaryotic IMPDHs contain substitutions in their active sites that can account for MPA resistance, no such substitutions are present in the P. brevicompactum IMPDHs. The substitution of serine or threonine for Ala249 is associated with MPA resistance in C. albicans and S. cerevisiae [14,15]. This substitution is found in PbIMPDH-A, but not in the more resistant PbIMPDH-B. Substitutions at positions 277 and 462 have also resulted in modest MPA resistance [19], and substitutions at position 351 are proposed to cause resistance [20], but again no correlation is observed in the fungal IMPDHs. We identified the C-terminal segment as a new structural determinant of MPA sensitivity, but this segment accounts for only 7 of the 60-fold difference in resistance between PbIMPDH-B and PbIMPDH-B. Importantly, the extraordinary resistance of PbIMPDH-B comes with a catalytic cost as this enzyme has lost much of the catalytic power of a typical IMPDH, with values of Km (IMP) that are ~100-fold greater and kcat/Km (IMP) that are 100–1000-fold lower than ‘normal’. This observation suggests that PbIMPDH-B would be a very poor catalyst at normal cellular IMP concentrations of 20–50 μM [21]. However, IMP concentrations increase 10–35-fold in the presence of MPA [1,22,23]. Thus P. brevicompactum may contain millimolar concentrations of IMP during MPA biosynthesis, so that the turnover of PbIMPDH-B would be comparable with a ‘normal’ fungal IMPDH.

Curiously, MPA production does not follow the phylogenetic relationship established by analysis of the IMPDH and β-tubulin genes (Figure 4). P. chrysogenum does not contain MPA-biosynthetic genes, although MPA producers are found on the neighbouring branches of the phylogenetic tree. As noted above, P. chrysogenum is more resistant to MPA than is A. nidulans (Figure 2A). This resistance could be the result of a gene-dosage effect as MPA resistance is associated with amplification of IMPDH genes in several organisms, consistent with this hypothesis [20,2427]. We propose that the presence of two IMPDH genes in Penicillium subgenus Penicillium was permissive for MPA production. The failure to observe MPA production in other Penicillium species may reflect inappropriate culture conditions rather than absence of the biosynthetic cluster. If so, it is possible that MPA biosynthesis has been obtained in several independent events. It will be interesting to map the evolution of MPA biosynthesis and resistance as more genomic sequences of filamentous fungi become available.

Gene duplication is generally believed to be a driving force in evolution, allowing one copy to diverge constraint-free, while the other copy maintains essential functions. The ‘escape from adaptive conflict’ subfunctionalization model is particularly attractive for the evolution of new enzyme activities [28,29]. In this model, duplication relieves the constraint of maintaining the original function and allows the emergence of enzymes with new catalytic properties. The blemish in this appealing hypothesis has long been recognized: most mutations are deleterious, so that the extra copy is quickly lost or converted into a pseudogene. Although several elegant and rigorous investigations of the in vitro evolution of new enzyme function demonstrate the feasibility of subfunctionalization (reviewed in [29,30]), genetic drift is widely recognized as the dominant mechanism in the wider evolution field. Several observations suggest that the emergence of MPA-resistant PbIMPDH-B could be an example of neofunctionalization by a classic Ohno-type mechanism: a duplication followed by (largely deleterious) drift until a new environment, MPA production, provides a selection.

PcIMPDH-B displays the effects of deleterious genetic drift: this protein is labile to proteolysis, and the value of kcat ~1% of that of typical eukaryotic IMPDH. Microarray data indicate that P. chrysogenum expresses both IMPDH-A and IMPDH-B under most conditions, so at present there is no evidence for specialized functions for these genes [31,32]. We cannot rule out the possibility that PcIMPDH-B plays another role within the cell, perhaps by interacting with another protein, or even as a heterotetramer with PcIMPDH-A. Nevertheless, our data suggest that the enzymatic properties of both IMPDH-Bs have declined over the course of evolution. In the case of P. brevicompactum, gaining the ability to produce MPA may have rescued IMPDH-B from non-functionalization.

Abbreviations

     
  • CYA

    Czapek yeast extract agar

  •  
  • dPCR

    degenerate PCR

  •  
  • DTT

    dithiothreitol

  •  
  • gDNA

    genomic DNA

  •  
  • IMPDH

    IMP dehydrogenase

  •  
  • AnImdA

    IMPDH-A from Aspergillus nidulans

  •  
  • PbIMPDH-A

    IMPDH-A from Penicillium brevicompactum

  •  
  • PbIMPDH-B

    IMPDH-B from P. brevicompactum

  •  
  • PcIMPDH-A

    IMPDH-A from Penicillium chrysogenum

  •  
  • PcIMPDH-B

    IMPDH-B from P. chrysogenum

  •  
  • IPTG

    isopropyl β-D-thiogalactopyranoside

  •  
  • LB

    Luria–Bertani

  •  
  • MPA

    mycophenolic acid

AUTHOR CONTRIBUTION

Bjarne Hansen, Hans Genee, Xin Sun and Lizbeth Hedstrom wrote the paper. All authors contributed to the editing of the paper before submission. Bjarne Hansen, Hans Genee, Christian Kaas and Jakob Nielsen built constructs and performed degenerate PCR and spot assays. Bjarne Hansen and Hans Genee performed the bioinformatics analysis. Xin Sun purified the proteins, performed the biochemical analysis and carried out the complementation experiment. Uffe Mortensen, Jens Frisvad and Lizbeth Hedstrom contributed to experimental design. Lizbeth Hedstrom supervised the project.

We thank Martin Engelhard Kornholt for valuable technical assistance in the laboratory.

FUNDING

This work was supported by The Danish Council for Independent Research Technology and Production Sciences [grant numbers 09-064967 and 09-064240 (to B.G.H. and U.H.M.)] and by the National Institutes of Health [grant number GM054403 (to L.H.)].

References

References
1
Bentley
R.
Mycophenolic acid: a one hundred year odyssey from antibiotic to immunosuppressant
Chem. Rev.
2000
, vol. 
100
 (pg. 
3801
-
3826
)
2
Regueira
T. B.
Kildegaard
K. R.
Hansen
B. G.
Mortensen
U. H.
Hertweck
C.
Nielsen
J.
Molecular basis for mycophenolic acid biosynthesis in Penicillium brevicompactum
Appl. Environ. Microbiol.
2011
, vol. 
77
 (pg. 
3035
-
3043
)
3
Hansen
B. G.
Genee
H. J.
Kaas
C. S.
Nielsen
J. B.
Mortensen
U. H.
Frisvad
J. C.
Patil
K. R.
A new clade of IMP dehydrogenases with a role in self-resistance in mycophenolic acid producing fungi
BMC Microbiol.
2011
, vol. 
11
 pg. 
202
 
4
Frisvad
J. C.
Samson
R. A.
Polyphasic taxonomy of Penicillium subgenus Penicillium: a guide to identification of food and air-borne terverticillate Penicillia and their mycotoxins
Stud. Mycol.
2004
, vol. 
49
 (pg. 
1
-
174
)
5
Geu-Flores
F.
Nour-Eldin
H. H.
Nielsen
M. T.
Halkier
B. A.
USER fusion: a rapid and efficient method for simultaneous fusion and cloning of multiple PCR products
Nucleic Acids Res.
2007
, vol. 
35
 pg. 
e55
 
6
Hansen
B. G.
Salomonsen
B.
Nielsen
M. T.
Nielsen
J. B.
Hansen
N. B.
Nielsen
K. F.
Regueira
T. B.
Nielsen
J.
Patil
K. R.
Mortensen
U. H.
Versatile enzyme expression and characterization system for Aspergillus nidulans, with the Penicillium brevicompactum polyketide synthase gene from the mycophenolic acid gene cluster as a test case
Appl. Environ. Microbiol.
2011
, vol. 
77
 (pg. 
3044
-
3051
)
7
Norholm
M. H.
A mutant Pfu DNA polymerase designed for advanced uracil-excision DNA engineering
BMC Biotechnol.
2010
, vol. 
10
 pg. 
21
 
8
MacPherson
I. S.
Kirubakaran
S.
Gorla
S. K.
Riera
T. V.
D'Aquino
J. A.
Zhang
M.
Cuny
G. D.
Hedstrom
L.
The structural basis of Cryptosporidium-specific IMP dehydrogenase inhibitor selectivity
J. Am. Chem. Soc.
2010
, vol. 
132
 (pg. 
1230
-
1231
)
9
Wang
W.
Papov
V. V.
Minakawa
N.
Matsuda
A.
Biemann
K.
Hedstrom
L.
Inactivation of inosine 5′-monophosphate dehydrogenase by the antiviral agent 5-ethynyl-1-β-D-ribofuranosylimidazole-4-carboxamide 5′-monophosphate
Biochemistry
1996
, vol. 
35
 (pg. 
95
-
101
)
10
Frisvad
J. C.
Smedsgaard
J.
Larsen
T. O.
Samson
R. A.
Mycotoxins, drugs and other extrolites produced by species in Penicillium subgenus Penicillium
Stud. Mycol.
2004
, vol. 
49
 (pg. 
201
-
241
)
11
Chooi
Y. H.
Cacho
R.
Tang
Y.
Identification of the viridicatumtoxin and griseofulvin gene clusters from Penicillium aethiopicum
Chem. Biol.
2010
, vol. 
17
 (pg. 
483
-
494
)
12
Gao
X.
Chooi
Y. H.
Ames
B. D.
Wang
P.
Walsh
C. T.
Tang
Y.
Fungal indole alkaloid biosynthesis: genetic and biochemical investigation of the tryptoquialanine pathway in Penicillium aethiopicum
J. Am. Chem. Soc.
2011
, vol. 
133
 (pg. 
2729
-
2741
)
13
Hedstrom
L.
IMP dehydrogenase: structure, mechanism and inhibition
Chem. Rev.
2009
, vol. 
109
 (pg. 
2903
-
2928
)
14
Kohler
G. A.
Gong
X.
Bentink
S.
Theiss
S.
Pagani
G. M.
Agabian
N.
Hedstrom
L.
The functional basis of mycophenolic acid resistance in Candida albicans IMP dehydrogenase
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
11295
-
11302
)
15
Jenks
M. H.
Reines
D.
Dissection of the molecular basis of mycophenolate resistance in Saccharomyces cerevisiae
Yeast
2005
, vol. 
22
 (pg. 
1181
-
1190
)
16
Link
J. O.
Straub
K.
Trapping of an IMP dehydrogenase-substrate covalent interemediate by mycophenolic acid
J. Am. Chem. Soc.
1996
, vol. 
118
 (pg. 
2091
-
2092
)
17
Sintchak
M. D.
Fleming
M. A.
Futer
O.
Raybuck
S. A.
Chambers
S. P.
Caron
P. R.
Murcko
M.
Wilson
K. P.
Structure and mechanism of inosine monophosphate dehydrogenase in complex with the immunosuppressant mycophenolic acid
Cell
1996
, vol. 
85
 (pg. 
921
-
930
)
18
Nimmesgern
E.
Black
J.
Futer
O.
Fulghum
J. R.
Chambers
S. P.
Brummel
C. L.
Raybuck
S. A.
Sintchak
M. D.
Biochemical analysis of the modular enzyme inosine monophosphate dehydrogenase
Protein Expression Purif.
1999
, vol. 
17
 (pg. 
282
-
289
)
19
Farazi
T.
Leichman
J.
Harris
T.
Cahoon
M.
Hedstrom
L.
Isolation and characterization of mycophenolic acid resistant mutants of inosine 5′-monophosphate dehydrogenase
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
961
-
965
)
20
Lightfoot
T.
Snyder
F. F.
Gene amplification and dual point mutations of mouse IMP dehydrogenase associated with cellular resistance to mycophenolic acid
Biochim. Biophys. Acta
1994
, vol. 
1217
 (pg. 
156
-
162
)
21
Hartman
N.
Ahluwalia
G.
Cooney
D.
Mitsuya
H.
Kageyama
S.
Fridland
A.
Broder
S.
Johns
D.
Inhibitors of IMP dehydrogenase stimulate the phosphorylation of the anti-human immunodeficiency virus nucleosides 2′,3′-dideoxyadenosine and 2′,3′-dideoxyinosine
Mol. Pharm.
1991
, vol. 
40
 (pg. 
118
-
124
)
22
Balzarini
J.
Karlsson
A.
Wang
L.
Bohman
C.
Horská
K.
Votruba
I.
Fridland
A.
Van Aerschot
A.
Hedewijn
P.
De Clerq
E.
EICAR (5-ethynyl-1-β-D-ribofuranosylimidazole-4-carboxamide): a novel potent inhibitor of inosinate dehydrogenase activity and guanylate biosynthesis
J. Biol. Chem.
1993
, vol. 
33
 (pg. 
24591
-
24598
)
23
Wu
J. C.
Mycophenolate mofetil: molecular mechanisms of action
Perspect. Drug Discovery Des.
1994
, vol. 
2
 (pg. 
185
-
204
)
24
Collart
F. R.
Huberman
E.
Amplification of the IMP dehydrogenase gene in Chinese hamster cells reistant to mycophenolic acid
Mol. Cell. Biol.
1987
, vol. 
7
 (pg. 
3328
-
3331
)
25
Wilson
K.
Collart
F.
Huberman
E.
Stringer
J.
Ullman
B.
Amplification and molecular cloning of the IMP dehydrogenase gene of Leishmania donovani
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
1665
-
1671
)
26
Wilson
K.
Berens
R. L.
Sifri
C. D.
Ullman
B.
Amplification of the inosinate dehydrogenase gene in Trypanosoma brucei gambienese due to an increase in chromosome copy number
J. Biol. Chem.
1994
, vol. 
269
 (pg. 
28979
-
28987
)
27
Kohler
G. A.
White
T. C.
Agabian
N.
Overexpression of a cloned IMP dehydrogenase gene of Candida albicans confers resistance to the specific inhibitor mycophenolic acid
J. Bacteriol.
1997
, vol. 
179
 (pg. 
2331
-
2338
)
28
O'Brien
P. J.
Herschlag
D.
Catalytic promiscuity and the evolution of new enzymatic activities
Chem. Biol.
1999
, vol. 
6
 (pg. 
R91
-
R105
)
29
Khersonsky
O.
Tawfik
D. S.
Enzyme promiscuity: a mechanistic and evolutionary perspective
Annu. Rev. Biochem.
2010
, vol. 
79
 (pg. 
471
-
505
)
30
Tracewell
C. A.
Arnold
F. H.
Directed enzyme evolution: climbing fitness peaks one amino acid at a time
Curr. Opin. Chem. Biol.
2009
, vol. 
13
 (pg. 
3
-
9
)
31
van den Berg
M. A.
Albang
R.
Albermann
K.
Badger
J. H.
Daran
J. M.
Driessen
A. J.
Garcia-Estrada
C.
Fedorova
N. D.
Harris
D. M.
Heijne
W. H.
, et al. 
Genome sequencing and analysis of the filamentous fungus Penicillium chrysogenum
Nat. Biotechnol.
2008
, vol. 
26
 (pg. 
1161
-
1168
)
32
Hoff
B.
Kamerewerd
J.
Sigl
C.
Zadra
I.
Kück
U.
Homologous recombination in the antibiotic producer Penicillium chrysogenum: strain ΔPcku70 shows up-regulation of genes from the HOG pathway
Appl. Microbiol. Biotechnol.
2010
, vol. 
85
 (pg. 
1081
-
1094
)
33
Saitou
N.
Nei
M.
The neighbour-joining method: a new method for reconstructing phylogenetic trees
Mol. Biol. Evol.
1987
, vol. 
4
 (pg. 
406
-
425
)

Author notes

1

Present address: Novozymes A/S, Krogshoejvej 36, 2880 Bagsvaerd, Denmark.

2

Present address: Department of Chemistry, Boston University, 590 Commonwealth Avenue, Boston, MA 02215-252, U.S.A.

Supplementary data