|Home | About | Journals | Submit | Contact Us | Français|
Ergot alkaloids are indole-derived secondary metabolites synthesized by the phytopathogenic ascomycete Claviceps purpurea. In wild-type strains, they are exclusively produced in the sclerotium, a hibernation structure; for biotechnological applications, submerse production strains have been generated by mutagenesis. It was shown previously that the enzymes specific for alkaloid biosynthesis are encoded by a gene cluster of 68.5 kb. This ergot alkaloid cluster consists of 14 genes coregulated and expressed under alkaloid-producing conditions. Although the role of some of the cluster genes in alkaloid biosynthesis could be confirmed by a targeted knockout approach, further functional analyses are needed, especially concerning the early pathway-specific steps up to the production of clavine alkaloids. Therefore, the gene ccsA, originally named easE and preliminarily annotated as coding for a flavin adenine dinucleotide-containing oxidoreductase, was deleted in the C. purpurea strain P1, which is able to synthesize ergot alkaloids in axenic culture. Five independent knockout mutants were analyzed with regard to alkaloid-producing capability. Thin-layer chromatography (TLC), ultrapressure liquid chromatography (UPLC), and mass spectrometry (MS) analyses revealed accumulation of N-methyl-dimethylallyltryptophan (Me-DMAT) and traces of dimethylallyltryptophan (DMAT), the first pathway-specific intermediate. Since other alkaloid intermediates could not be detected, we conclude that deletion of ccsA led to a block in alkaloid biosynthesis beyond Me-DMAT formation. Complementation with a ccsA/gfp fusion construct restored alkaloid biosynthesis. These data indicate that ccsA encodes the chanoclavine I synthase or a component thereof catalyzing the conversion of N-methyl-dimethylallyltryptophan to chanoclavine I.
The ergot fungus Claviceps purpurea is a phytopathogenic ascomycete which infects the ears of several grasses, replacing the ovary and producing a hibernation structure, the so-called sclerotium, in which the ergot alkaloids are formed. These substances show a high level of structural homology to some neurotransmitters like serotonin and dopamine and can therefore bind to the same receptors in the central nervous system (CNS), which is the basis for the application of ergot alkaloids in a variety of clinical conditions (15).
The biochemistry of ergot alkaloid biosynthesis was first investigated by isolation of intermediates and postulation of a hypothetical pathway as well as enzymes needed for the successive biosynthetic steps of the production (Fig. (Fig.1).1). Most of the data were collected by pursuing the fate of radiolabeled precursors in feeding experiments (4). The first enzyme which could be assigned to alkaloid production was dimethylallyltryptophan synthetase (DMATS), which is the key enzyme of the pathway and is encoded by the gene dmaW (18). These analyses were performed with a Claviceps fusiformis strain, but a homolog of dmaW (AY259840) possessing a similar function could also be isolated in C. purpurea, as was confirmed by a knockout approach (N. Lorenz and P. Tudzynski, unpublished data). Using genome walking combined with cDNA screening, a 68.5-kb genomic region surrounding dmaW could be sequenced and revealed 14 open reading frames (ORFs) (putative genes) encoding, among others, nonribosomal peptide synthetases (NRPSs), a putative catalase, a CYP450-1 monooxygenase, a putative methyltransferase, and several oxidoreductases (6, 13, 19) (Fig. (Fig.2).2). Some of these genes were functionally and biochemically analyzed by a gene replacement approach which revealed their function within the pathway (2, 5, 7). However, there is still a deficit in functional analyses, especially with respect to the early steps within this pathway. The conversion from N-methyl-dimethylallyltryptophan (Me-DMAT) to agroclavine via chanoclavine I and chanoclavine I aldehyde includes successive oxidation and reduction steps mediated by a specific class of enzymes, the oxidoreductases (15) (Fig. (Fig.11).
These enzymes are involved in the biosynthesis of many fungal secondary metabolites. A prominent example is the family of the cytochrome P450 monooxygenases (named after the characteristic peak of 450 nm when complexed with carbon monoxide). Cytochrome P450 (CYP450) monooxygenases catalyze the transfer of one oxygen atom from molecular oxygen to various substrates, mostly accomplished by the involvement of NAD(P)H as an electron donor. The eas cluster of C. purpurea also includes a gene encoding a CYP450 monooxygenase: cloA is involved in the oxidation of elymoclavine, leading to the formation of paspalic acid (7).
No further monooxygenase-encoding genes seem to be present in the eas cluster, but several genes code for putative oxidoreductases (easA, easD, easE, easG, and easH). These oxidoreductases are most likely involved in the early steps within the pathway, but none of them has been functionally analyzed so far (15).
We initiated a functional analysis of the putative oxidoreductase-encoding gene ccsA (formerly easE) (Fig. (Fig.2).2). The coding region of ccsA (AJ011965; 1,503 bp) is composed of two exons interrupted by an intron of 52 bp, yielding a coding capacity of 483 amino acids (aa). The gene product shows highest similarity to putative oxidoreductases of other ergot alkaloid-producing fungi: EasE of C. fusiformis (e−160; ABV57823), EasE of Neotyphodium lolii (e−118; ABM91450) and CpoX1 of Aspergillus fumigatus (e−96; XM_751049). Analyses of the protein sequence using the program PROSITE revealed a flavin adenine dinucleotide (FAD)-binding domain (pfam01565) spanning the region from amino acids 14 to 161 and a berberine bridge enzyme domain (BBE domain; pfam08031) from amino acids 412 to 457. The role of CcsA in the alkaloid biosynthesis pathway was investigated by knockout of the corresponding gene, followed by functional and biochemical analyses of the deletion mutants.
The ku70-deficient strain derived from Claviceps purpurea strain P1 (ATCC 20102 [8, 19), which produces mainly ergotamine with small amounts of ergocryptine, was described earlier (5), as were the standard media and culture conditions (19). For alkaloid production, the fungus was cultivated in T25N medium with low (0.5 g/liter KH2PO4) and high (2.0 g/liter KH2PO4) levels of phosphates.
Standard cloning and DNA analysis techniques were performed according to the methods of Sambrook et al. (14). The Escherichia coli strains used for cloning by using plasmids pUC19 (Fermentas, St. Leon-Rot, Germany) and pCR2.1-TOPO (Invitrogen, Karlsruhe, Germany) and propagation of clones was TOP10F′ (Invitrogen, Karlsruhe, Germany). Extraction of genomic DNA, Southern and Northern blot analyses, and DNA sequencing were performed as described previously (11). For sequence comparisons and multiple sequence alignments, DNA STAR (Madison, Wisconsin) was used. For further analyses, the programs BLAST (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) and PROSITE (http://www.expasy.ch/prosite/) were utilized.
The replacement vector RV-ccsA is based on the vector pAN8.1_UM, including two multiple cloning sites in front of the gpdA promoter (Aspergillus nidulans) and downstream of the gpdA terminator (Aspergillus nidulans) of the phleomycin resistance cassette, an ampicillin resistance gene for selection in bacteria, and an origin of replication (Fig. (Fig.3).3). (For primers used for amplification of the flanks, see below.)
Transformation of protoplasts of C. purpurea was performed as described previously (2).
Total RNA extraction was done as described earlier (17) from 7-day-old mycelia using the RNAgents total RNA isolation system (Promega, Mannheim, Germany). Concentrations of purified RNA were determined using a BioPhotometer (Eppendorf, Hamburg, Germany), and RNA integrity was examined by electrophoresis in 1% formaldehyde agarose gels.
For PCR analysis, BioTherm (Genecraft, Lüdinghausen, Germany) polymerase was used according to the manufacturer's instructions.
For the construction of the replacement vector, the flanks were amplified by the primer pairs ccsA_VF_v2 (5′-CCA TTC AGG TAC CCG TCC AG-3′) and ccsA_VF_h (5′-AAG GAC AGA ATT CCT CCA CGC-3′) for the 5′ flank and ccsA_HF_v (5′-CGG AGC TAT CTA GAT AGC ATC-3′) and ccsA_HF_h (5′-TTG GAT GCC GCG GTG AGT GCA T-3′) for the 3′ flank. To check the transformants for homologous integration events, diagnostic PCR was performed using the primer pair d_ccsA_v (5′-GCA TCA AGA GCA CAA CAA CAC GAG-3′) and PAN3 (5′-GGT CAC CAG TCG CTG GCT TCC CG-3′) for the 5′ flank and the primer pair PAN2 (5′-CCG TAA CAC CCA ATA CGC CGG-3′) and d_ccsA_h (5′-TGC AGC ATA GAC CCC AGA CAG AC-3′) for the 3′ flank. Purification of the mutants was checked by the primer pair ccsA_easF_U (5′-ACC GTG GGT GCA GTA GGA GGC-3′) and ccsA_R (5′-GCT TCC GGC AAA TAC CTT CTG-3′).
For complementation of the deletion mutant, the gene ccsA was amplified by the primer pair ccsA_GFP_v (5′-CAC CAA TTC TAG AAG CAC CG-3′) and ccsA_GFP_h (5′-GGT ACC CAT TCC GAT AAG ACT GGA CGC-3′) and cloned into the pUC19 vector in front of the start codon of the gfp (green fluorescent protein) gene from Aequorea jellyfish (16), leading to a fusion protein where GFP is located at the C terminus of CcsA. The vector already contained a 1-kb fragment of the nitrate reductase gene (niaD) of C. purpurea (see Fig. Fig.9).9). The niaD gene could be used for targeted integration of the complementation construct because of the increased rate of homologous integration of the ku70-deficient strain of C. purpurea. Additionally, a hygromycin resistance cassette (consisting of the oliC promoter, the hygromycin resistance gene, and the trpC terminator of Aspergillus nidulans) was cloned via the XbaI restriction site into the vector, enabling direct selection of the transformants. Successful integration was tested by diagnostic PCR using the primer GFP_h (5′-TCG AAT TCT TAC TTG TAC AGC TCG TCC-3′).
For alkaloid extraction and determination, cultures were adjusted to pH 11 with concentrated aqueous ammonium hydroxide and extracted three times with chloroform, and after concentration, the resulting liquid was applied onto thin-layer chromatography (TLC) plates (silica gel 60; Merck, Darmstadt, Germany). Alkaloids were identified by comparison with the corresponding standards in chloroform-methanol-ammonium (80:20:0.2; vol/vol/vol). After separation, TLC plates were sprayed with Ehrlich's reagent for alkaloid visualization. Standards used were 1 mg/ml ergotamine (Sigma, Taufkirchen, Germany), 1 mg/ml ergocryptine (Sigma, Taufkirchen, Germany), 1 mg/ml ergocristine (Sigma, Taufkirchen, Germany), 1 mg/ml agroclavine (Sandoz AG, Basel, Switzerland), and 1 mg/ml d-lysergic acid (Sandoz AG, Basel, Switzerland).
High-pressure liquid chromatography (HPLC) separation was carried out on a LiChrospher 100 RP-18 (Merck/Hitachi, Darmstadt, Germany) column (250 × 4 mm inner diameter; 5-μm particle size), tempered at 25°C, and operated at a flow rate of 1 ml/min. Compounds were isocratically eluted within 40 min using a binary mobile phase containing 10 mM ammonium carbonate and acetonitrile (50:50, vol/vol). Ergot alkaloids (EA) were detected with an L-7400 UV detector operating at 230 nm. EA standards were used as described above.
An Acquity ultrapressure liquid chromatography (UPLC) system (Waters, Milford, Massachusetts), equipped with a 2996 polydiode array (PDA) detector operating at 225 nm and 310 nm, was used for EA analysis under UPLC conditions (12). Data were processed with Empower 2 software (Waters). Samples were analyzed on a Waters BEH C18 column (50 × 2.1 mm inner diameter; 1.7-μm particle size) under the following conditions: column temperature, 35°C; data sample rate, 20 data points (pts)/s; filter constant, 0.5; injection volume, 5 μl; analysis time, 12 min; flow rate, 0.4 ml/min. Mobile phases consisted of water (phase A) and acetonitrile (phase B), both containing 0.04% NH4OH. Gradient elution started at 5% acetonitrile (0 min), increasing linearly to 61% acetonitrile within 12 min. Each analysis was followed with a column-washing step (95% acetonitrile, 1 min) and equilibration step (1 min).
Mass spectra were measured on a matrix-assisted laser desorption-ionization reflectron time-of-flight (MALDI-TOF) mass spectrometer (BIFLEX; Bruker-Franzen, Bremen, Germany) equipped with a nitrogen laser (337 nm) and gridless delayed extraction ion source. The ion acceleration voltage was 19 kV, and the reflectron voltage was set to 20 kV. Spectra were calibrated externally using the monoisotopic [M+H]+ ion of matrix peaks 190.1 m/z and 379.1 m/z and peptide MRFA (Met-Arg-Phe-Ala) 524.3 m/z. Solutions (10 mg/ml) of α-cyano-4-hydroxy-cinnamic acid or 2,5-dihydrobenzoic acid in 50% acetonitrile with 0.3% acetic acid were used as MALDI matrices. Each collected UPLC fraction was dried on a vacuum concentrator, dissolved in water, and sonicated for 5 min prior to mass spectrometry analysis. A 1-μl sample of matrix solution was mixed with 1 μl of the aqueous solution of the sample, and 1 μl of this premix was loaded on the target and allowed to dry at ambient temperature. The MALDI-TOF and postsource decay (PSD) spectra were collected in reflectron mode.
For deletion of ccsA, a double-crossover approach was used, resulting in replacement of the ccsA locus by a phleomycin resistance cassette by homologous recombination (Fig. (Fig.4).4). The replacement construct was based on the phleomycin resistance cassette of vector pAN8.1_UM flanked by 1,149-bp and 1,030-bp 5′ and 3′ regions, respectively, of ccsA (Fig. (Fig.44).
As the recipient strain, a ku70-deficient derivative of the producer strain P1 was used, characterized by an enhanced efficiency for homologous recombination (5).
Successful integration was checked by diagnostic PCR using the primer pair d_ccsA_v and PAN3 for homologous integration of the 5′ flank (yielding a fragment of 1.8 kb) and primer pair PAN2 and d_ccsA_h for verification of the 3′ flank (yielding a band of 1.4 kb; Fig. Fig.4).4). Of 22 transformants checked, 13 showed homologous integration. Interestingly, 6 out of these 13 also missed the wild-type band when the primer pair ccsA_easF_U and ccsA_R was used, indicating that these transformants were already homokaryotic and contained only transformed nuclei.
Successful knockout was also confirmed by Southern analysis using SacI-digested DNA of wild-type strain P1 as well as five homokaryotic transformants (Fig. (Fig.5).5). The gene ccsA lacks any SacI restriction site, so in the wild type only one band of 4.8 kb could be observed. In contrast, the gene replacement construct includes two SacI restriction sites, resulting in the detection of a reduced band of 3.4 kb in the deletion mutants when hybridized with the 5′ flank and a band of 3.1 kb when the 3′ flank was used as a probe (Fig. (Fig.44).
Northern analyses confirmed the successful knockout of the gene: no ccsA signal could be detected in the knockout mutants, whereas the genes easD and easF were expressed as in the wild type, confirming that expression of the neighboring cluster genes was not affected by the gene replacement approach (Fig. (Fig.6).6). The integrity of the neighboring cluster genes was also verified by sequencing of genomic PCR fragments (data not shown).
The alkaloid-producing capabilities of five deletion mutants were analyzed by TLC (thin-layer chromatography), HPLC (high-pressure liquid chromatography), and UPLC (ultrapressure liquid chromatography ) with PDA detection (polydiode array), and MS (mass spectrometry), and the spectra were compared to the alkaloid spectrum of the wild type.
The first preliminary HPLC analyses showed that all knockout mutants had lost the ability to produce clavine alkaloids, and it was concluded, therefore, that the pathway was blocked at a step prior to the steps catalyzing ergoline ring formation. Moreover, TLC analyses confirmed not only the lack of clavines but also the lack of more complex alkaloids such as ergopeptines (data not shown).
To address the production of possible intermediates in more detail, UPLC analyses were performed. In these analyses, the accumulation of a substance (after 4.32 min) was reproducibly detected showing the typical UV pattern of ergot alkaloids (maxima at wavelengths of 224 nm and 280 nm, respectively) (Fig. (Fig.7).7). Structure determination of the compound was performed by MS analysis. The acquired MALDI-TOF spectra revealed that the isolated intermediate had a main signal at m/z 287.2, and manual interpretation of its MS/MS fragmentation resulted in identification of N-methyl-dimethylallyltryptophan (Me-DMAT) as the main end product of the deletion mutants (Fig. (Fig.8).8). Dimethylallyltryptophan (DMAT) was also found in minor amounts in the UPLC fraction at the time point 4.58 min, with a MALDI-TOF signal at m/z 272.2. The identification of Me-DMAT as the main product in contrast to DMAT strengthens the assumption that the pathway is blocked at the biosynthetic step following DMAT methylation, which explains why Me-DMAT accumulates whereas DMAT is constantly converted to Me-DMAT and thus present in only minor amounts. The next pathway-specific intermediate not detected in the deletion mutants is chanoclavine I, supporting the presumption that CcsA is essential for the conversion from Me-DMAT to chanoclavine I. Therefore, the gene—originally called easE (15)—was named ccsA (for chanoclavine synthase).
To confirm that the failure of the mutants to produce clavine and more-complex alkaloids was due to the targeted knockout, the transformants were complemented with a gene construct designed to express a CcsA::GFP fusion protein. This construct could allow not only testing for restoration of alkaloid production but also analysis of the cellular localization of alkaloid biosynthesis, which is described elsewhere as compartmentalized (15).
The complementation vector cCcsA/GFP contained the full coding region of ccsA with its endogenous promoter (to avoid overexpression), the gfp gene fused in frame to the 3′ end of ccsA, and part of the niaD (nitrate reductase) locus for targeted integration in this nonessential gene (for details, see Materials and Methods) (Fig. (Fig.9).9). Integration of the complementation fragment in the niaD locus was confirmed by PCR (Fig. (Fig.9)9) as well as Southern analysis (data not shown), verifying also that—due to the integration outside the eas cluster—the gene replacement situation was still present.
The expression pattern, as well as the alkaloid spectrum of the complemented mutants, was analyzed by Northern blotting and HPLC. The expression of the ccsA/gfp fusion construct was confirmed by the presence of a larger ccsA homologous transcript (data not shown), and functionality of the resulting fusion protein was shown by restoration of ergotamine production in the complemented mutant (Fig. (Fig.10).10). The GFP signal obtained, however, was rather diffuse, not supporting the assumption that the biosynthesis is located in specialized cellular compartments (data not shown).
A knockout of the eas cluster gene ccsA resulted in accumulation of Me-DMAT and traces of DMAT, as indicated by UPLC and MS analyses, strongly suggesting that CcsA is involved in the “chanoclavine I synthase” activity converting Me-DMAT to chanoclavine I (Fig. (Fig.11).
Two successive oxidation steps are required for chanoclavine I formation from Me-DMAT (Fig. (Fig.11).11). The reaction most likely starts with the generation of the diene by desaturation of the C8-C9 bond and loss of a proton at C17 (4). This is followed by rotation around the C8-C9 bond and oxidation of the C7-C8, proposed to give an epoxide intermediate (4). Decarboxylation of this intermediate by a (spontaneous) SN2′ reaction could be coupled with bond formation between C5 and C10 and cleavage of the epoxide. In addition to binding the FAD cofactor, it is conceivable that the BBE domain is involved in the subsequent cyclization, either as a catalyst or by determining stereospecificity, as has been suggested for the BBE (3, 9). Following this, ring opening of the epoxide occurs by proton attack with shift of the double bond between C9 and C10 to the position between C8 and C9.
We showed here that CcsA is essential for this biosynthetic step, but it remains an open question whether the reaction is carried out by CcsA alone or whether CcsA belongs to a complex of enzymes catalyzing the two oxidation steps to form the diene and then the epoxide, as was postulated by Schardl et al. (15). Experiments involving the knockout of other putative oxidoreductase-encoding genes of the eas cluster, e.g., easA, easD, easG, or easH, could help in further elucidating this process.
We thank S. Pažoutová and M. Flieger, Prague, Czech Republic, for discussions and U. Keller, Berlin, Germany, for critical reading of the manuscript. The DFG (special focus program Evolution of Metabolic Diversity) provided financial support.
Published ahead of print on 29 January 2010.