Reconstructing fungal natural product biosynthetic pathways.

Large scale fungal genome sequencing has revealed a multitude of potential natural product biosynthetic pathways that remain uncharted. Here we describe some of the methods that have been used to explore them via heterologous gene expression. We focus on filamentous fungal hosts and discuss the technological challenges and successes behind the reconstruction of fungal natural product pathways. Optimised, efficient heterologous expression of reconstructed biosynthetic pathways promises progress in the discovery of novel compounds that could be utilised by the pharmaceutical and agrochemical industries.

Identification of genes encoding squalestatin S1 biosynthesis and in vitro production of new squalestatin analogues.

A gene cluster responsible for the biosynthesis of squalestatin S1 (SQS1, 1) was identified by full genome sequencing of two SQS1-producing ascomycetes: Phoma sp. C2932 and unidentified fungus MF5453. A transformation protocol was established and a subsequent knockout of one PKS gene from the cluster led to loss of SQS1 production and enhanced concentration of an SQS1 precursor. An acyltransferase gene from the cluster was expressed in E. coli and the expressed protein MfM4 shown to be responsible for loading acyl groups from CoA onto the squalestatin core as the final step of biosynthesis. MfM4 appears to have a broad substrate selectivity for its acyl CoA substrate, allowing the in vitro synthesis of novel squalestatins.

Design and utility of oligonucleotide gene probes for fungal polyketide synthases.

BACKGROUND: Recent advances in the molecular biology of polyketide biosynthesis have allowed the engineering of polyketide synthases and the biological ('combinatorial') synthesis of novel polyketides. Additional structural diversity in these compounds could be expected if more diverse polyketide synthases (PKS) could be utilised. Fungal polyketides are highly variable in structure, reflecting a potentially wide range of differences in the structure and function of fungal PKS complexes. Relatively few fungal synthases have been investigated, perhaps because of a lack of suitable genetic techniques available for the isolation and manipulation of gene clusters from diverse hosts. We set out to devise a general method for the detection of specific PKS genes from fungi. RESULTS: We examined sequence data from known fungal and bacterial polyketide synthases as well as sequence data from bacterial, fungal and vertebrate fatty acid synthases in order to determine regions of high sequence conservation. Using individual domains such as beta-ketoacylsynthases (KS), beta-ketoreductases (KR) and methyltransferases (MeT) we determined specific short (ca 7 amino acid) sequences showing high conservation for particular functional domains (e.g. fungal KR domains involved in producing partially reduced metabolites; fungal KS domains involved in the production of highly reduced metabolites etc.). Degenerate PCR primers were designed matching these regions of specific homology and the primers were used in PCR reactions with fungal genomic DNA from a number of known polyketide producing species. Products obtained from these reactions were sequenced and shown to be fragments from as-yet undiscovered PKS gene clusters. The fragments could be used in blotting experiments with either homologous or heterologous fungal genomic DNA. CONCLUSIONS: A number of sequences are presented which have high utility for the discovery of novel fungal PKS gene clusters. The sequences appear to be specific for particular types of fungal polyketide (i.e. non-reduced, partially reduced or highly reduced KS domains). We have also developed primers suitable for amplifying segments of fungal genes encoding polyketide C-methyltransferase domains. Genomic fragments amplified using these specific primer sequences can be used in blotting experiments and have high potential as aids for the eventual cloning of new fungal PKS gene clusters.

Synthesis and secretion of wheat alpha-amylase in Saccharomyces cerevisiae.

A wheat alpha-amylase cDNA clone has been fused to the phosphoglycerate kinase initiator methionine to enable synthesis in the yeast Saccharomyces cerevisiae of an alpha-amylase enzyme that is identical in size to the wild-type alpha-amylase. The alpha-amylase is synthesized with an N-terminal plant signal peptide which is recognized in the yeast host, leading to efficient processing and secretion into the medium. The secretion of alpha-amylase into the medium is quite efficient in rich medium, but barely detectable in a minimal medium.

Functional identification of a fatty acid delta5 desaturase gene from Caenorhabditis elegans.

We have identified a cDNA from the nematode worm Caenorhabditis elegans that encodes a fatty acid delta5 desaturase. Saccharomyces cerevisiae expressing the full-length cDNA was able to convert di-homo-gamma-linolenic acid to arachidonic acid, thus confirming delta5 desaturation. The 1341 bp delta5 desaturase sequence contained an N-terminal cytochrome b5 domain and was located within a kilobase of the C. elegans delta6 desaturase on chromosome IV. With an amino acid identity of 45% it is possible that one of these genes arose from the other by gene duplication. This is the first example of a delta5 desaturase gene isolated from an animal.

Stable expression of maize auxin-binding protein in insect cell lines.

To achieve continuous expression of the major maize auxin-binding protein (ABP1) in insect cells, the ABP1 gene coding region was placed under control of a baculovirus immediate-early gene promoter and transfected into Spodoptera frugiperda Sf9 cells. The ABP1 gene was detected in twelve cell lines, one of which was selected for detailed analysis. Immunolocalisation demonstrated that ABP1 was targeted to and retained in the endoplasmic reticulum (ER), in accordance with its signal peptide and carboxy-terminal KDEL ER-retention signal. We discuss the advantages of stable-transformation over transient expression systems for characterising proteins targeted to the secretory system of insect cells.

Combined ribosomal DNA and morphological analysis of individual gyrodactylid monogeneans.

A method is presented for the isolation and analysis of hamuli, marginal hooks, and bars from individual gyrodactylid monogeneans using scanning electron microscopy (SEM), while simultaneously processing parasites for rDNA analysis using the polymerase chain reaction (PCR). The haptors of ethanol-fixed gyrodactylids were protease digested to liberate hooks for SEM, whereas DNA extracted from the bodies was used for PCR. The method resulted in hooks and hamuli being prepared from more than 90% of Gyrodactylus turnbulli individuals, a significant improvement on previously published digestion-based SEM techniques. PCR on the same parasites was less successful, but sequence data were obtained from 50% of individuals. Amplification of rDNA internal-transcribed spacer regions from individual worms used for SEM gave PCR products consistent with those predicted from our previous sequence analysis. This method allows the correlation of morphology and DNA sequence from the same individual and can be applied to ethanol-fixed material, such as field collected and museum specimens.

Anatomy of amplified mitochondrial DNA in "Ragged" mutants of Aspergillus amstelodami: Excision points within protein genes and a common 215 bp segment containing a possible origin of replication.

The extranuclearly-inherited ragged growth phenotype (Rgd) of Aspergillus amstelodami is always accompanied by excision and head-to-tail amplification of mtDNA sequences. In one mutant strain (Rgd1) the amplified mtDNA segment (rgd1 DNA, monomeric length 0.9 kb) maps downstream of the large subunit ribosomal RNA gene (Region 1), whereas in all other strains analyzed the amplified sequences (rdg3-7DNA) are located in Region 2 between genes coding for cytochrome b and ATPase subunit 6. The various region 2 sequences differ in lengths (1.5 to 2.7 kb) but have in common a 215 bp sequence mapping between an. unidentified protein gene (corresponding to URF4 of human mtDNA) and an arginine tRNA gene. This common sequence may contain an origin of replication, because a looped-out hairpin structure similar to that of yeast and human mitochondrial origin sequences can be formed. Furthermore, Region 2 DNA suppresses replication of Region 1 DNA, indicating that the former group of molecules contains the more efficient origin. The nucleotide sequence of the rgd6 repeat unit starts and ends within protein genes of mtDNA, and no homologies were found between heads and tails or their flanking sequences.

Characterization of a strawberry gene for auxin-binding protein, and its expression in insect cells.

A gene encoding an auxin-binding protein (ABP1) was isolated from strawberry by screening a genomic library with an ABP1 cDNA from maize. It resembles ABP1 genes from other sources both in structure (four introns) and in the high level of homology of the deduced amino acid sequence of the mature protein encoded in exons 2-5. Exon 1, encoding mainly the non-conserved signal peptide, was identified by a reverse transcriptase-polymerase chain reaction (RT-PCR) technique. Northern analysis indicated that ABP1 transcript levels were low during fruit development, but transcripts were detected by RT PCR at all stages of receptacle swelling (auxin-dependent) and ripening (inhibited by auxin), consistent with a role for ABP1 in auxin perception. Southern blot analysis indicated a small ABP1 gene family in octoploid cultivated strawberry, and four genes were identified by comparison of genomic and cDNA sequences. RT PCR was used to amplify the complete coding region for cloning as cDNA, and a recombinant baculovirus was constructed for the expression of strawberry ABP1 in insect cells. The coding region contains three consensus glycosylation sites, and multiple bands representing a range of glycoforms of the protein were detected on western blots of insect cell extracts. Only a single band was observed in extracts of tunicamycin-treated cells, and glycosylated protein yielded a unique N-terminal amino acid sequence, allowing determination of the signal peptide cleavage site.

Extranuclear recombination in Aspergillus nidulans: closely-linked multiple chloramphenicol- and oligomycin-resistance loci.

A nuclear, chloramphenicol-sensitive mutant cas-1 has been isolated which is cross sensitive to a number of drugs, including oligomycin and cycloheximide. Approximately one-third of the chloramphenicol-resistant mutants isolated from mutagenized conidia of this strain were found to be extranuclear, and exhibited a variety of phenotypes. One of these mutants, designated (camB51), was slow growing on drug-free medium and recombined at low frequency with the previously described mutant (camA112) (Gunatilleke et al., 1975). The majority of extranuclear oligomycin-resistant mutants isolated from cas-1 were indistinguishable from (oliA1) (Rowlands and Turner, 1973). Two mutants, (oliB322) and (oliB332), with similar but not identical phenotypes to (oli A1), recombined with the latter at low frequency but not with each other, thus representing a new class of extranuclear mutants.

Split gene for mitochondrial 24S ribosomal RNA of Neurospora crassa.

The 60 kb circular mitochondrial genome of N. crassa has previously been shown to contain a single transcription unit for 17S and 24S rRNA mapping within the largest Eco RI fragment E1 (19.6 kb). This fragment was isolated from uncloned mitochondrial DNA and further analyzed by cleavage with restriction endonucleases Hind II, Hind III, Bam HI, Pvu II and BgI I, and by electron microscopy of rRNA/DNA hybrids. The resulting map shows a 2.3 kb intervening sequence interrupting the gene for 24S rRNA. The main part (2.7 kb) of this gene is separated from the 17S rRNA gene by a 5 kb segment which contains several transfer RNA genes. This segment is much longer than the putative 1 kb spacer sequence within the 32S precursor molecule for both rRNAs, suggesting a second splicing event in that region.

α-amylase genes of wheat are two multigene families which are differentially expressed.

The α-Amy1 and α-Amy2 genes of wheat produce distinct subsets of α-amylase isozymes which show different patterns of expression in wheat aleurone cells and in developing grain. In order to characterise the organisation and expression of these genes, clones of α-Amy1 and α-Amy2 cDNA have been isolated. The two types of cDNA clone were distinguished within a small library of α-amylase cDNA clones (Baulcombe and Buffard, Planta 157 493-501 [1983]) by restriction endonuclease mapping and by cross hybridisation. The identity of α-Amy1 or α-Amy2 type was assigned from the results of hybrid selected translation analysis in which small subfragments of the cDNA clones were used. These subfragments were derived from the 3' ends of the cDNA and did not cross hybridise between the different types of cDNA. Hybridisation of α-Amy1 and α-Amy2 cDNA probes to restriction enzyme digests of wheat nuclear DNA revealed that these are multigene families located on the group 6 (α-Amy1) and group 7 (α-Amy2) chromosomes. Studies on the levels of α-Amy1 and α-Amy2 mRNA in developing grain and in aleurone tissue indicated that the differences in isozyme expression are due to the patterns of mRNA accumulation. In aleurone tissue the α-Amy1 transcripts accumulate in parallel with other genes which are regulated by gibberellic acid, while the accumulation of α-Amy2 genes is sustained for 36 h longer.

Ketosynthase domain probes identify two subclasses of fungal polyketide synthase genes.

Analysis of fungal polyketide synthase gene sequences suggested that these might be divided into two subclasses, designated WA-type and MSAS-type. Two pairs of degenerate PCR primers (LC1 and LC2c, LC3 and LC5c) were designed for the amplification of ketosynthase domain fragments from fungal PKS genes in each of these subclasses. Both primer pairs were shown to amplify one or more PCR products from the genomes of a range of ascomycetous Deuteromycetes and Southern blot analysis confirmed that the products obtained with each pair of primers emanated from distinct genomic loci. PCR products obtained from Penicillium patulum and Aspergillus parasiticus with the LC1/2c primer pair and from Phoma sp. C2932 with both primer pairs were cloned and sequenced; the deduced protein sequences were highly homologous to the ketosynthase domains of other fungal PKS genes. Genes from which LC1/2c fragments were amplified (WA-type) were shown by a phylogenetic analysis to be closely related to fungal PKS genes involved in pigment and aflatoxin biosynthetic pathways, whereas the gene from which the LC3/5c fragment was amplified (MSAS-type) was shown to be closely related to genes encoding 6-methylsalicylic acid synthase (MSAS). The phylogenetic tree strongly supported the division of fungal PKS genes into two subclasses. The LC-series primers may be useful molecular tools to facilitate the cloning of novel fungal polyketide synthase genes.

Aleurone nuclear proteins bind to similar elements in the promoter regions of two gibberellin-regulated alpha-amylase genes.

Binding of nuclear proteins from wild oat aleurone protoplasts to the promoter regions of two gibberellin-regulated wheat alpha-amylase genes (alpha-Amy1/18 and alpha-Amy2/54) has been studied by gel retardation and DNase 1 footprinting. Gel retardation studies using 300-430 bp fragments of the promoters showed similar binding characteristics with nuclear extracts from both gibberellin A1-treated and untreated protoplasts. DNase 1 footprints localised binding of nuclear proteins from gibberellin A1-treated aleurone protoplasts to regions in both promoters. Similar sequence elements in the promoter regions of both genes were protected from digestion although the location and number of footprints in each promoter region were different. Each footprint contained either a sequence similar to the cAMP and/or phorbol ester response elements, or a hyphenated palindrome sequence. The presence of cAMP and/or phorbol ester response element-like sequences in the footprints suggests that transcription factors of the bZIP type may be involved in the expression of alpha-amylase genes in aleurone cells. Footprints containing hyphenated palindrome sequences, found in the promoter regions of both genes, suggest the possible involvement of other classes of transcription factor. The conserved alpha-amylase promoter sequence TAA-CAGA was also shown to bind nuclear protein in the alpha-Amy2/54 promoter. These observations are discussed in relation to alpha-amylase gene expression in aleurone and to functional data concerning these genes.

Members of a new family of DNA-binding proteins bind to a conserved cis-element in the promoters of alpha-Amy2 genes.

The promoters of wheat, barley and wild oat alpha-Amy2 genes contain a number of conserved cis-acting elements that bind nuclear protein, we report here the isolation of two cDNAs encoding proteins (ABF1 and ABF2) that bind specifically to one of these elements, Box 2 (ATTGACTTGACCGTCATCGG). The two proteins are unrelated to each other except for a conserved region of 56-58 amino acids that consists of 25 highly conserved amino acids followed by a putative zinc finger motif, C-X4-5-C-X22-23-H-X1-H. ABF1 contains two such conserved regions, whereas ABF2 possesses only one but also contains a potential leucine zipper motif, suggesting that it could form homo- or heterodimers. ABF1 and ABF2 expressed in Escherichia coli bound specifically to Box 2 probes in gel retardation experiments; this binding was abolished by the transition-metal-chelating agent, 1,10-o-phenanthroline and by EDTA. We propose that ABF1 and ABF2 are representatives of two classes of a new family of plant sequence-specific DNA-binding proteins.

Physical map of Aspergillus nidulans mitochondrial genes coding for ribosomal RNA: an intervening sequence in the large rRNA cistron.

A detailed map of the 32 kb mitochondrial genome of Aspergillus nidulans has been obtained by locating the cleavage sites for restriction endonucleases Pst I, Bam H I, Hha I, Pvu II, Hpa II and Hae III relative to the previously determined sites for Eco R I, Hind II and Hind III. The genes for the small and large ribosomal subunit RNAs were mapped by gel transfer hybridization of in vitro labelled rRNA to restriction fragments of mitochondrial DNA and its cloned Eco R I fragment E3, and by electron microscopy of RNA/DNA hybrids. The gene for the large rRNA (2.9 kb) is interrupted by a 1.8 kb insert, and the main segment of this gene (2.4 kb) is separated from the small rRNA gene (1.4 kb) by a spacer sequence of 2.8 kb length. This rRNA gene organization is very similar to that of the two-times larger mitochondrial genome of Neurospora crassa, except that in A. nidulans the spacer and intervening sequences are considerably shorter.

Amplification of a mitochondrial DNA sequence in the cytoplasmically inherited 'ragged' mutant of Aspergillus amstelodami.

A comparison has been made between mtDNA of the cytoplasmically inherited 'ragged' mutant of Aspergillus amstelodami and that of the wild-type strain. Ragged mitochondria contain both the wild-type mitochondrial genome and several large DNA molecules which are not cleaved by the restriction endonucleases BamHI, HaeIII, HhaI, HindII, HindIII, PstI and MboI, but are converted by either EcoRI or HpaII into a single 820-840 base-pair fragment. Restriction analysis and molecular hybridization data indicate that this fragment contains sequences of wild-type mtDNA located within a 1200-base-pair segment of the 40,500-base-pair genome, for which a basic restriction map has been deduced. It is concluded that in the ragged mutant a small segment of wild-type mtDNA has been amplified as tandem repeats, which is reminiscent of the Rho- petite phenotype of yeast. The results are discussed in relation to the phenomenon of senescence in Podospora anserina.

DNase1 footprints suggest the involvement of at least three types of transcription factors in the regulation of alpha-Amy2/A by gibberellin.

To elucidate the mechanisms by which alpha-amylase genes are expressed in wild oat aleurone, two genes, alpha-Amy2/A and alpha-Amy2/D, were isolated. Both were shown to be positively regulated by gibberellin (GA) during germination and both contain the conserved cis-acting elements Box 2, GA-response element (TAACAGA) and TATCSATSS (where S is C or G). In addition, they possess a conserved initiator element (CATCA) that is present in both alpha-Amy2 and alpha-Amy1 genes, and also in a number of other plant TATA-containing and TATA-less promoters. DNase 1 footprint analysis showed the alpha-Amy2/A promoter to be a complex array of binding sites for a number of different classes of DNA-binding proteins. Our data suggest that the area around the initiator element (Inr) is bound by a large complex of general transcription factors, that the TATA box is bound by the TFIID complex, that Box 2 is bound by one or more WRKY proteins and that the GA-response element is bound by one or more MYBs. Two other elements containing the core sequence CCATGG/C are bound by nuclear protein and this sequence is the core of the Sph element. The regulation of alpha-Amy2 genes by GA therefore involves an interplay of at least three different types of transcription factor.

Auxin-binding protein--antibodies and genes.

Of several auxin-binding systems that have been characterised the auxin-binding protein (ABP) of maize coleoptile membranes is the best candidate for a true auxin receptor. ABP, which exists as a homodimer of 22 x 10(3) M(r) glycosylated subunits, has been purified, and monoclonal and polyclonal antibodies raised against it. Electrophysiological studies with antibodies indicated the presence of a functional population of auxin receptors on the exterior face of the plasmalemma; electrophysiological experiments with impermeant auxin analogues now reinforce this conclusion. An epitope mapping kit has been used to identify the major epitopes recognised by antibody preparations. Three major epitopes, bracketing the glycosylation site, have been identified in the polyclonal serum. They are also represented in antisera produced in other laboratories and are conserved in ABP prepared from other plants. One monoclonal antibody recognises an epitope close to the amino terminus of ABP and two others recognise the carboxy terminus. The latter antibodies have been used in a sandwich ELISA to demonstrate that auxin binding induces a conformational change in ABP. Maize ABP is encoded by a small gene family and cDNA and genomic clones have been isolated. With a single exception, predicted amino acid sequences indicate remarkably little heterogeneity. The exceptional cDNA sequence predicts 87% amino acid homology with the major class of proteins. Four introns are apparent in the sequence of a complete ABP gene; their sequences are very highly conserved in an incompletely-cloned second gene lacking the first exon. The major difference between the two genes lies in the length of the first intron, which has been estimated to exceed 5.2 kb in the incomplete gene. The site of initiation of transcription has not been unambiguously identified in the complete gene, and some evidence suggests that there may be an additional intron. Homology to maize ABP cDNA has been detected in the genomes of Arabidopsis, spinach and strawberry but not in that of tobacco. A sequence located within the 3'-half of the maize cDNA is highly repeated in the strawberry genome, from which clones with homology to both halves of the maize cDNA (i.e. putative ABP genes) have been isolated.

Isolation of a Delta5-fatty acid desaturase gene from Mortierella alpina.

Arachidonic acid (C20:4 Delta5,8,11,14) is a polyunsaturated fatty acid synthesized by the Delta5-fatty acid desaturation of di-homo-gamma-linolenic acid (C20:3 Delta8,11,14). In mammals, it is known to be a precursor of the prostaglandins and the leukotrienes but it is also accumulated by the filamentous fungus Mortierella alpina. We have isolated a cDNA encoding the Delta5-fatty acid desaturase from M. alpina via a polymerase chain reaction-based strategy using primers designed to the conserved histidine box regions of microsomal desaturases, and confirmed its function by expression in the yeast Saccharomyces cerevisiae. Analysis of the lipids from the transformed yeast demonstrated the accumulation of arachidonic acid. The M. alpina Delta5-desaturase is the first example of a cloned Delta5-desaturase, and differs from other fungal desaturases previously characterized by the presence of an N-terminal domain related to cytochrome b5.