Fungal Genetics and Biology 43 (2006) 618–629

www.elsevier.com/locate/yfgbi

Transcriptional and bioinformatic analysis of the 56.8kb DNA region ampliWed in tandem repeats containing the penicillin gene

cluster in Penicillium chrysogenum

Francisco Fierro a, Carlos García-Estrada a,b, Nancy I. Castillo a, Raquel Rodríguez a,b, Tania Velasco-Conde a, Juan-Francisco Martín a,b,¤

a Instituto de Biotecnología de León (INBIOTEC) Science Park, Av. Real, 1, 24006 León, Spainb Área de Microbiología, Fac. CC. Biológicas y Ambientales, Universidad de León, Campus de Vegazana, s/n, 24071 León, Spain

Received 20 October 2005; accepted 28 March 2006Available online 18 May 2006

Abstract

High penicillin-producing strains of Penicillium chrysogenum contain 6–14 copies of the three clustered structural biosynthetic genes,pcbAB, pcbC, and penDE [Barredo, J.L., Díez, B., Alvarez, E., Martín, J.F., 1989. Large ampliWcation of a 35-kb DNA fragment carryingtwo penicillin biosynthetic genes in high penicillin producing strains of Penicillium chrysogenum. Curr. Genet. 16, 453–459; Smith, D.J.,Bull, J.H., Edwards, J., Turner, G., 1989. AmpliWcation of the isopenicillin N synthetase gene in a strain of Penicillium chrysogenum pro-ducing high levels of penicillin. Mol. Gen. Genet. 216, 492–497.] Barredo et al., 1989; Smith et al., 1989. The cluster is located in a 56.8 kbDNA region bounded by a conserved TGTAAA/T hexanucleotide that undergoes ampliWcation in tandem repeats [Fierro, F., Barredo,J.L., Díez, B., Gutiérrez, S., Fernández, F.J., Martín, J.F., 1995. The penicillin gene cluster is ampliWed in tandem repeats linked by con-served hexanucleotide sequences. Proc. Natl. Acad. Sci. USA 92, 6200–6204; Newbert, R.W., Barton, B., Greaves, P., Harper, J., Turner,G., 1997. Analysis of a commercially improved Penicillium chrysogenum strain series: involvement of recombinogenic regions in ampliW-cation and deletion of the penicillin biosynthesis gene cluster. J. Ind. Microbiol. Biotechnol. 19, 18–27.]. Transcriptional analysis of thisampliWed region (AR) revealed the presence of at least eight transcripts expressed in penicillin producing conditions. Three of them corre-spond to the known penicillin biosynthetic genes, pcbAB, pcbC, and penDE. To locate genes related to penicillin precursor formation, orpenicillin transport and regulation we have sequenced and analyzed the 56.8 kb ampliWed region of P. chrysogenum AS-P-78, Wnding atotal of 16 open reading frames. Two of these ORFs have orthologues of known function in the databases. Other ORFs showed similari-ties to speciWc domains occurring in diVerent proteins and superfamilies which allowed to infer their probable function. ORF11 encodes aD-amino acid oxidase that might be responsible for the conversion of D-amino acids in the tripeptide L-�-aminoadipyl-L-cysteinyl-D-valineor other �-lactam intermediates to deaminated by-products. ORF12 encodes a predicted protein with similarity to saccharopine dehydro-genases that seems to be related to biosynthesis of the penicillin precursor �-aminoadipic acid. A deletion mutant, P. chrysogenum npe10lacking the entire AR including ORF12, shows a partial requirement of L-lysine for growth. ORF13 encodes a putative protein containinga Zn(II)2-Cys6 fungal-type DNA-binding domain, probably a transcriptional regulator. Although some of the ORFs in the AR may playroles in increasing penicillin production, none of the 13 ORFs other than pcbAB, pcbC, and penDE seem to be strictly indispensable forpenicillin biosynthesis. The genes located in the P. chrysogenum AR have been compared with those found in the Aspergillus nidulans50 kb DNA region adjacent to the penicillin gene cluster, showing no conservation between these two fungi.© 2006 Elsevier Inc. All rights reserved.

Keywords: Transcriptional analysis; AmpliWed region; Penicillin biosynthesis; Gene clusters; Precursor biosynthesis; Regulatory genes

1087-1845/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.fgb.2006.03.001

* Corresponding author. Fax: +34 987 29 15 06.E-mail address: degjmm@unileon.es (J.-F. Martín).

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 619

1. Introduction

Clustering of genes involved in a biosynthetic pathway iscommon in bacteria but it is a rare phenomenon in eucary-otic cells. However, in Wlamentous fungi, genes involved inparticular biosynthetic or catabolic pathways are oftenclustered together (Keller and Hohn, 1997). This phenome-non is especially frequent in the so-called dispensable path-ways including: (1) pathways for the biosynthesis ofsecondary metabolites (Martín and Liras, 1989), like thegenes for biosynthesis of penicillin (Díez et al., 1990; Smithet al., 1990) or sterigmatocystin (Brown et al., 1996), and (2)catabolic pathways for the utilization of unusual nutrientslike proline (Hull et al., 1989). These clusters contain genesthat encode proteins with diVerent functions, mainly bio-synthetic enzymes, transport proteins (Martín et al., 2005),and regulatory proteins (Martín and Liras, 1989).

The genes involved in the biosynthesis of �-lactam anti-biotics are arranged in clusters in bacteria as well as in Wla-mentous fungi (Aharanowitz et al., 1992; Martín, 1998;Martín et al., 1997). In Acremonium chrysogenum, genesinvolved in the biosynthesis of cephalosporin C arearranged in two separate clusters located in diVerent chro-mosomes (Gutiérrez et al., 1999a; Skatrud and Queener,1989). The “early gene cluster” located in chromosome VII(4.6 Mb) contains the genes pcbAB and pcbC encoding theenzymes for the Wrst two steps of the pathway, the genescefD1 and cefD2 which are responsible for the epimeriza-tion of the isopenicillin N (Ullán et al., 2002a), and the cefTgene, encoding a transmembrane protein involved in theexport of the antibiotic (Ullán et al., 2002b). The “late genecluster” located in chromosome I (2.2 Mb) contains thegenes for the last steps of cephalosporin biosynthesis: cefEFand cefG (Gutiérrez et al., 1992).

In P. chrysogenum, the three penicillin biosyntheticgenes (pcbAB, pcbC, and penDE) are clustered in a 17 kbDNA region (Díez et al., 1990; Smith et al., 1990) locatedin chromosome I (10.4 Mb) (Fierro et al., 1993). The clus-ter is located in a DNA region that is ampliWed in tandemrepeats bounded by a conserved hexanucleotideTGTAAA/T in penicillin overproducing strains (Fierroet al., 1995; Newbert et al., 1997). The same region is fre-quently deleted, resulting in non-producing mutants likeP. chrysogenum npe10 (Fierro et al., 1996a), by a recom-bination mechanism not yet fully understood (Fierro andMartín, 1999). The ampliWable unit is either 106.5 or56.8 kb in length depending on the strains. Strains withhigher penicillin production capacity possess up to 14copies of the 56.8 kb ampliWed region (AR). The correla-tion of copy number with increased penicillin biosynthe-sis suggests that other genes related to penicillinbiosynthesis might be located within this region. A simi-lar organization of the pcbAB (acvA), pcbC (ipnA), andpenDE (aat) genes has been reported in Aspergillus nidu-lans (MacCabe et al., 1990) but the cluster is present in asingle copy, in the 3.0 Mb chromosome VI (Montenegroet al., 1992).

Therefore, it was of great interest sequencing andanalyzing the transcriptional pattern of the 56.8 kbP. chrysogenum DNA region to elucidate if, as occurs inA. chrysogenum and in �-lactam producing bacteria, othergenes related to the biosynthesis, regulation or transport ofpenicillin are clustered together with the three structuralgenes. On the other hand, since the 56.8 kb AR is present ina high number of copies in penicillin overproducing strains(Fierro et al., 1995), all genes located in it have an increasedgene dosage, which may have side eVects also in cellularprocesses other than penicillin production.

2. Materials and methods

2.1. Fungal and bacterial strains

Penicillium chrysogenum Wis 54-1255 is a wild-type-derived strain containing a single copy of the AR. P. chrys-ogenum npe10 is a deletion derivative of Wis 54-1255 lack-ing the entire AR (Fierro et al., 1996a). Penicilliumchrysogenum AS-P-78, a high penicillin producing straincontaining 5–6 copies of the AR (Fierro et al., 1995), waskindly donated by Antibióticos S.A. (León, Spain). Esche-richia coli DH5� (Clontech, Palo Alto, California) was usedfor plasmid constructions and plasmid DNA ampliWcation.

2.2. Culture conditions

Penicillium chrysogenum strains were grown on Powermedium (Fierro et al., 1996a). Conidia from one Petri dishwere collected and inoculated into a Xask with 100 ml ofdeWned seed medium [glucose 20 g/l, lactose 15 g/l, ammo-nium acetate 1.15 g/l, CaCO3 5 g/l, NaNO3 1.27 g/l, KH2PO42 g/l, Na2SO4 5 g/l, Na2S2O3·5H2O 1 g/l, pH 5.8]. After 60 hof incubation in a rotary shaker at 25 °C, 7 ml of this seedculture were inoculated into 100 ml of complex penicillinproduction medium [corn steep solids 20 g/l, lactose 55 g/l,MgSO4·7H2O 3 g/l, CaCO3 10 g/l, KH2PO4 7 g/l, potasiumphenylacetate (40%) 6.5 ml/l, pH 6.8]. After growing for 48 hin the above-mentioned conditions the mycelium was col-lected by Wltration, washed with 0.9% NaCl and frozen inliquid nitrogen prior to RNA extraction.

2.3. RNA isolation

Total RNA from Penicillium chrysogenum AS-P-78 wasisolated from 48 h cultures using the phenol-chloroformextraction and lithium chloride precipitation method(Ausubel et al., 1987). mRNA was puriWed by Wltering100 �g of total RNA through an oligo(dT)-cellulose col-umn (Invitrogen) according to the manufacturers instruc-tions.

2.4. Northern blot and hybridization

Ten micrograms of total RNA or 0.4 �g of mRNA wereloaded into the wells of denaturing agarose gels and sub-

620 F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629

mitted to electrophoresis. Denaturing gels and Northernblotting onto nylon membranes were performed by stan-dard procedures (Sambrook et al., 1989). DNA probes werelabeled with (32P)-�-dCTP (Amersham) by nick translationaccording to the manufacturer’s speciWcations (PromegaCorporation). Hybridizations were carried out at 42 °C in a50% formamide standard buVer (Sambrook et al., 1989).After hybridization the nylon membranes were washedwith 2£ SSC, 0.1% SDS (42 °C, 20 min), 0.1£ SSC, 0.1%SDS (42 °C, 15 min), 0.1£ SSC, 0.1% SDS (65 °C, 10 min)and exposed in a phosphorimager (Instant Imager, Pack-ard) for 1–14 h and to an X-ray Wlm for up to three weeks.

2.5. DNA sequencing and bioinformatics analysis

Sequencing of DNA was performed by automatic sequenc-ing using the Abi Prism™ system (PE Applied Biosystems,Foster City, California). Searches for open reading frames ingenomic sequences were performed with the ORF Wnder pro-gram at the NCBI web site (http://www.ncbi.nlm.nih.gov/gorf/gorf.html), and with the Genebuilder system of the ITB(Institute of Biomedical Technologies, Milan, It.) (http://www.itba.mi.cnr.it/webgene/). Identity searches were per-formed with the WU-Blast2 tool at the EBI web site, on theSwiss-Prot database (http://www.ebi.ac.uk/blast2/), and theBlast tools available at the NCBI web site (http://www.ncbi.nlm.nih.gov/BLAST/). Alignments of nucleotideand amino acid sequences were carried out with the CLUS-TAL W facility at the Pole BioInformatique Lyonnais site(http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page Dnpsa_clustalw.html) and the MegAlign program of DNA-star™ (DNAstar, Inc). Searches for conserved motifs inproteins were carried out with the integrated InterProScanSequence Search tool at the EBI web site (http://www.ebi.ac.uk/InterProScan/). The Aspergillus nidulans data-base at the Broad Institute (http://www.broad.mit.edu/anno-tation/fungi/aspergillus/) was used to retrieve the DNAsequence of contig 1.45 and the annotated ORFs in a 50kbregion containing the cluster of penicillin biosynthetic genes.

3. Results

3.1. Transcriptional analysis of the 56.8 kb ampliWed region

To study how many transcripts from genes located in theAR were detectable in conditions of penicillin production, atranscriptional mapping was performed by Northern blotanalysis (Fig. 1). A total of 10 diVerent DNA probes cover-ing the entire 56.8 kb AR were used. Six of these probes(E–J) correspond to DNA fragments situated downstreamfrom the pcbAB gene and four (A–D) downstream from thepenDE gene. Total RNA from 48 h of growth in complexpenicillin production medium was used for the Northernblot analysis, as expression of the three penicillin biosyn-thetic genes is higher at this time (Gutiérrez et al., 1999b).

Northern blot analysis revealed the presence of at leastseven transcripts: Wve new mRNAs in addition to the ones

belonging to the pcbAB and penDE genes (Fig. 1). Two ofthe new transcripts correspond to genes downstream ofpcbAB and three to genes downstream of penDE. Thehybridization signals of the new genes present in the ARwere weaker than those belonging to the structural genespcbAB and penDE, indicating that these new genes are nothighly expressed at the time and conditions used. Hybrid-ization with probe A gave rise to two signals of 1.1 and0.7 kb, whereas probe B resulted in a faint 0.7 kb signal.Hybridization with probe C produced a 1.4 kb signal welldeWned in the mRNA lane and smeared in the total RNAlane. The 1.2 kb signal generated by probe D corresponds tothe penDE gene. Hybridization with probe E resulted in asmear along the gel corresponding to degradation of thelong 11.5 kb transcript of the pcbAB gene. Since this smearcould mask other transcripts present in the fragment cov-ered by probe E, new hybridizations using probes Ess, E1,E2, and E3 were performed (Fig. 1). Surprisingly, whileprobe E1, situated immediately downstream of pcbAB pro-duced no signal, probe E2, located further downstream,gave a hybridization signal identical to that with probe Ebut of lower intensity. The same signal was obtained withprobe Ess. The explanation for this result came after analy-sis of the nucleotide sequence of the fragment correspond-ing to probe E, where two imperfect DNA direct repeats(showing 75.4% of identity between themselves) werefound; one of the repeats was located just downstreamfrom the pcbAB gene (B2 in Fig. 1) and includes the last91 bp of the coding region of this gene, and the other (B1)was found 2.3 kb further downstream. Probes F, H, and Jgave rise to no signals in the Northern blot, indicating theabsence of genes expressed in the used conditions in theseregions. Hybridization with probe G resulted in one signalof about 0.9 kb and with probe I a signal of about 1.2 kbwas detected.

3.2. Sequence and bioinformatic analysis of the 56.8 kb ampliWed region

The sequence of the entire AR along with Xankingsequences was obtained. The sequence comprising exactly56,820pb between the conserved hexanucleotides situated atthe left and right ends of the AR (Fierro et al., 1995) has beenannotated and deposited in GenBank with the Accession No.DQ192518. Bioinformatic analysis predicted a total of 16ORFs, seven downstream of pcbAB and six downstream ofpenDE, plus the three structural genes pcbAB, pcbC, andpenDE (Fig. 1). Five of the new ORFs contained putativeintrons (Table 1). Searches in the databases gave high scorematches with hypothetical fungal proteins for several of theORFs (Table 1). Two of the ORFs had orthologues of knownfunction in the databases: ORF12 (a saccharopine dehydro-genase) and ORF14 (a glyoxalase family protein). OtherORFs showed conserved domains that allowed to infer theirpossible function; the protein encoded by ORF2 shows aputative N-acetylglucosaminyltransferase-IV conservedregion, ORF6 shows similarity to S-adenosylmethione-depen-

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 621

dent methyl transferases, ORF11 has a conserved DadAdomain (Glycine/D-amino acid oxidase), and ORF13 encodesa protein containing a putative Zn(II)2-Cys6 fungal-typeDNA-binding domain, which suggests that the encoded pro-tein might have a regulatory function by binding to DNA.

The hybridization signals obtained in the transcriptionalanalysis of the AR were correlated with the ORFs pre-dicted by bioinformatic analysis. The results, summarizedin Table 2, indicate that ORF3, ORF4, ORF6, ORF7,ORF11, and probably ORF13 were not signiWcantlyexpressed under the conditions used. The rest of ORFsseem to be transcribed, though clear evidence for expres-sion was only found for ORF5, ORF12, and ORF14 (inaddition to the pcbAB and penDE genes).

3.3. Two open reading frames are located in the 3.2 kb “shift fragment”

At the right of the AR there is a 3.2 kb region, named“shift fragment”, Xanked by the same hexanucleotidebounding the AR but in an inverted repeat arrangement,

which undergoes inversion seemingly by a similarsequence-speciWc mechanism as the ampliWcation, thusappearing in diVerent orientation depending on thestrains (Fierro et al., 1995). Two ORFs, ORF15, andORF16, are located within this shift fragment, whose ori-entation consequently changes from that in P. chrysoge-num AS-P-78 (Fig. 1) to the opposite one in strains like P.chrysogenum Wis 54-1255 or P. chrysogenum P2. ORF15contains ankyrin repeat domains (ANK) involved in pro-tein–protein interaction processes, and ORF16 is a puta-tive sterol desaturase family protein containing sixtransmembrane domains.

3.4. Direct repeats downstream of the pcbAB gene

As mentioned above, two imperfect direct repeats werefound downstream of the pcbAB gene (named B1 and B2,Fig. 1). They are 489 and 480 bp long, respectively, share75.4% of their nucleotide sequence, and Xank a DNAregion of 2353 bp which does not contain any long ORF.Interestingly, a 40 bp long poly-A stretch appears within

Fig. 1. Restriction map and transcriptional analysis of the 56.8 kb AR of P. chrysogenum, downstream of pcbAB (above), and downstream of penDE(below), showing the position of the ORFs (black arrows) found by computer analysis and the transcripts detected by Northern blot with probes A–J(indicated with thin double-head arrows below). The region between the pcbAB 3� end and the penDE gene is not included. White boxes inside ORFs indi-cate introns. EcoRI (E), BamHI (B), and SalI (S). Enzymes ClaI (C), HindIII (H), and SacI (Sa) were used to generate probes E3, E2, E1, C and B, and theyare shown only at these points. The DNA fragment corresponding to the AR is shown with a thick box, and Xanking sequences at the left and right with athin box. The “shift” fragment is highlighted with a darker grey box. Hexanucleotide sequences framed by boxes indicate the borders of the AR and the“shift” fragment. Grey arrows above the upper box correspond to the positions of direct imperfect DNA repeats found downstream of pcbAB: A1–A2 andB1–B2. Black boxes in the Xanking sequences at the right correspond to short DNA stretches showing similarity to a Mariner/Tc1-like transposase (t1)and to a retrotransposon gag/pol gene (t2) respectively. Numbers correspond to the base positions of the AR, starting from the TGTAAA (T D 1)sequence at the left end and ending at the TGTAAT (T D 56820) sequence at the right end. The size of the detected transcripts is indicated in kilobases atthe right of the Northern blot panels; m, polyadenylated mRNA; t, total RNA. The 11.5 kb pcbAB transcript is notoriously unstable (Díez et al., 1990) andgives a degradation smear in the hybridization with probes E and Ess (see text for explanation).

622 F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629

the inter-repeats region, just beside repeat B2. Databasesearches were performed looking for identities at DNA orprotein level with genes present usually in retrotransposons.The matches obtained showed low identity scores; a 174deduced amino acid sequence from the inter-repeats region

showed 29% identity (eD0.014) to the RNA polymerase IIlargest subunit of hamster, while a 46 deduced amino acidsequence in opposite orientation showed 43% identity(eD0.19) with hypothetical ribonuclease E and G fromThermobiWda fusca.

Table 1Description of the ORFs found in the P. chrysogenum AR

c, ORF situated on the complementary strand.a CDS size, number of deduced amino acids in brackets.b highest score match obtained with the BLAST search tool at the NCBI.c e-value, percentage of similarity in brackets.

Code Position Introns Sizea Matchb Identityc Other features

ORF1c 2847–1969 0 879 (292) No similarity — —ORF2c 4725–3499 0 1227 (408) Hypothetical protein AN5414_2

[Aspergillus nidulans FGSC A4]405 aa. (Accession No. EAA62574)

3e ¡ 163 (69%) Putative N-acetylglucosaminyltransferase-IV (GnT-IV) conserved region between aa 90 and 212. Putative integral membrane protein (A. fumigatus, Accession No. XP_749941) with 3 transmembrane spanning domains

ORF3 5170–6387 0 1218 (405) Hypothetical protein FG09369.1 [Gibberella zeae PH-1] 328 aa. (Accession No. EAA76688)

1e ¡ 3 (23%) —

ORF4 7551–9902 0 2352 (783) Predicted protein [Aspergillus nidulans FGSC A4] 741 aa. (Accession No. EAA62573)

0.0 (56%) Putative signal peptide (residues 1–25) and four transmembrane regions

ORF5 12451–13338 0 888 (295) Predicted protein [Aspergillus nidulans FGSC A4] 215 aa. (Accession No. EAA60057)

8e ¡ 14 (29%) Putative signal peptide residues 1–25

ORF6c 15676–13796 0 1881 (626) Conserved hypothetical protein [Aspergillus fumigatus Af293] 649 aa. (Accession No. EAL86016)

1e ¡ 144 (51%) SigniWcant identity (1e¡ 7) with an Arabidopsis thaliana putative S-adenosylmethionine-dependent methyltransferase

ORF7c 18194–16301 3 1599 (532) Hypothetical protein [Neurospora crassa] 1619 aa. (Accession No. XP_325620)

5e ¡ 10 (40%) Putative IMP dehydrogenase/GMP reductase domain, residues 84–486

ORF8c 34654–23318 0 11337 (3778) P. chrysogenum ACV synthetase (Accession No. AAA63415)

0.0 (100%) —

ORF9 35670–36665 0 996 (331) P. chrysogenum IPN synthase(Accession No. CAA35480)

0.0 (100%) —

0RF10 38163–39437 3 1074 (357) P. chrysogenum acyltransferase (Accession No. AAA33692)

0.0 (100%) —

ORF11c 42020–40456 1 1260 (419) Conserved hypothetical protein [Aspergillus fumigatus Af293] 515 aa. (Accession No. EAL89177)

2e ¡ 34 (29%) Putative conserved DadA domain (Glycine/D-amino acid oxidases (deaminating)) related to amino acid transport and metabolism, residues 116–391. FAD/NADP(H) binding domain superfamily

ORF12 42560–43915 0 1356 (451) Saccharopine dehydrogenase and related proteins [Anabaena variabilis ATCC 29413] 384 aa. (Accession No. ZP_00203444)

9e ¡ 17 (26%) Conserved Saccharopine dehydrogenase and Lys9 domains

ORF13 46934–47398 0 465 (154) Hypothetical protein UM04688_1 [Ustilago maydis 521] 698 aa. (Accession No. XP_760835)

3e ¡ 09 (32%) Similar to AlcR transcriptional activator DNA-binding domain from A. nidulans (65 aa., P21228). Zn(II)2Cys6 fungal-type DNA-binding domain, aa 26–75

ORF14c 52819–52260 1 477 (158) Glyoxalase family protein [Aspergillus fumigatus Af293] 128 aa. (Accession No. EAL91136)

2e ¡ 31 (52%) Belongs to the glyoxalase/bleomycin resistance protein superfamily

ORF15 54319–55297 2 834 (277) Conserved hypothetical protein [Aspergillus fumigatus Af293] 436 aa. (Accession No. EAL90417)

4e ¡ 20 (51%) Contains ANK domains (ankyrin repeats domains) for protein–protein interaction. Residues 11–67

ORF16 55514–56662 2 987 (328) Sterol desaturase family protein,putative [Aspergillus fumigatus Af293] 296 aa. (Accession No. EAL84630)

2e ¡ 143 (68%) Contains six transmembrane spanning domains

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 623

Other repeated sequence was found further downstreamof pcbAB, close to the left end of the AR. In this case, twodirect repeats (A1 and A2, Fig. 1) of 334 and 339 bp (shar-ing 91% of their sequences) Xank a DNA region of 4.6 kb.Two ORFs are located within this region, ORF1 andORF2, the latter encoding an N-acetylglucosaminyltrans-ferase related protein, but no matches to retransposon ele-ments were found.

3.5. ORF12 encodes a saccharopine dehydrogenase that contributes to lysine biosynthesis but is not essential for growth

The protein encoded by ORF12 showed high similarityin its amino terminal half to saccharopine dehydrogen-ases from diVerent bacteria, being the best match with asaccharopine dehydrogenase of Anabaena variabilis

(amino acids 11–280, 26% identity; 9e¡ 17) (Fig. 2). Thecarboxyl terminal region of the ORF12-encoded proteinis less conserved. Clustal alignments with diVerent saccha-ropine dehydrogenases from fungi and bacteria were per-formed. ORF12 showed overall identity to bacterialsaccharopine dehydrogenases, like those from Anabaenavariabilis (19.9% identity) or Gloeobacter violaceus(19.3%). It showed lower identity to the lysA-encoded pro-tein (saccharopine dehydrogenase) from A. nidulans(15.2%) and its orthologues in other fungi, like the lys1from Neurospora crassa (12.8%) or the lys1 from Saccha-romyces cerevisiae (12.4%) (Fig. 2).

The identity of ORF12 was also low with thepreviously cloned (Naranjo et al., 2001) saccharopinereductase (Lys7) of P. chrysogenum (15.6%), althoughgood (end to end) conservation (37.7% positives) wasobserved, including the carboxyl terminal region

Table 2Correlation between the ORFs present in the AR and the Northern blot signals obtained with the probes covering this region

Probe Position ORFs included (positions) Hybridization signals (kb) Comments

A Beyond 56820–52500 16 (55514–56662)15 (54319–55297)14 (52819–52260)

1.1 and 0.7 ORF16 (987 bp without introns) and ORF15 (834 bp) cannot correspond due to their larger size to the 0.7 kb transcript. Therefore the 1.1 kb transcript is the result of ORF15 and/or ORF16 expression. The 0.7 kb transcript corresponds to ORF14, whose size (477 bp) is in agreement with the length of the transcript.

B 52499–47254 14 (52819–52260)13 (46934–47398)

0.7 The 0.7 kb transcript is the same as the obtained with probe A, and corresponds to ORF14. Since there are no other signals with this probe, ORF13 (465 bp) seems not to be transcribed, although overlapping of weak transcription signals from both ORFs, similar in size, cannot be ruled out.

C 47253–42688 13 (46934–47398)12 (42560–43915)

1.4 The 1.4 kb transcript corresponds to ORF12 (1356 bp), since no signal of this size was observed with the previous probe B, which included part of the ORF13.

D 42687–38163 12 (42560–43915)11 (42020–40456)penDE gene

1.2 The intensity of the 1.2 kb transcript is the result of the high penDE gene expression. No transcripts bigger than 1.2 kb are present, and therefore ORF11 (1260 bp) seems not to be transcribed.

E 24190–18156 pcbAB gene7 (18194–16301)

11.5 This strong signal corresponds to the transcript of the pcbAB gene (see text). A very small fragment of ORF7 (38 bp) is included in this probe, which is too short to give rise to a detectable signal.

F 18155–14348 7 (18194–16301)6 (15676–13796)

none Neither ORF7 nor ORF6 are being transcribed under penicillin production conditions.

G 14804–9827 6 (15676–13796)5 (12451–13338)4 (7551–9902)

0.9 Due to the lack of signal for ORF6 given by probe F, and the large size of ORF4 (2352 bp), it can be concluded that the 0.9 kb transcript is the result of ORF5 (888 bp) expression.

H 9826–5758 4 (7551–9902)3 (5170–6387)

none Neither ORF4 nor ORF3 are being transcribed under penicillin production conditions.

I 5757–451 3 (5170–6387)2 (4725–3499)1 (2847–1969)

1.2 Since ORF3 is not transcribed, as observed with probe H, the 1.2 kb transcript has to be the result of the expression of ORF1 (879 bp) or ORF2 (1227 bp).

J 895–beyond 1 none none No ORFs included in this region.

624 F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629

(Fig. 2C), indicating that both proteins are related,although they are clearly diVerent.

It is important to note that the primary metabolism sac-charopine dehydrogenase gene of P. chrysogenum (homolo-gous to lysA of A. nidulans) has not been cloned yet. Thepossibility that the ORF12 encodes this enzyme was testedby comparing the growth on minimal and L-lysine-supple-mented medium of P. chrysogenum npe10, which lacks theAR, and its parental strain Wis 54-1255. As shown in Fig. 3,

the strain npe10 (lacking ORF12) grows more slowly inCzapek than the parental strain, although it does notstrictly require lysine for growth. Normal growth ofP. chrysogenum npe10 was restored by supplementing themedium with L-lysine. These results indicate that the sac-charopine dehydrogenase encoded by ORF12 is a “second-ary metabolism type” enzyme contributing to lysinebiosynthesis in P. chrysogenum, but diVerent from the pri-mary (LysA) saccharopine dehydrogenase (see Section 4).

Fig. 2. Phylogenetic tree and amino acid sequence alignment of the ORF12 encoded protein with saccharopine dehidrogenases and saccharopine reduc-tases from diVerent organisms. (A) ORF12 shows higher identity with saccharopine dehydrogenases from diVerent cyanobacteria, like Anabaena variabilis,Gloeobacter violaceus, and Synechocystis sp. (B) Sequence alignment of ORF12 deduced amino acid sequence with saccharopine dehydrogenases fromAnabaena variabilis (YP_324887.1) and Gloeobacter violaceus (Q7NJW0); the amino terminal half of the sequence shows good conservation with the bac-terial proteins. (C) Sequence alignment of the ORF12 amino acid sequence with the P. chrysogenum Lys7 saccharopine reductase, showing low identity(15.6%) but good end to end conservation (37.7% positives).

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 625

3.6. Sequences Xanking the 56.8 kb AR: possible degenerated transposable elements

No ORFs were found in the adjacent DNA regionsXanking left and right the borders of AR (outside of theTGTAAA/T hexamer in each end). However, shortstretches in the right Xanking region (t1 and t2 in Fig. 1)showed similarity to transposases from class II (Tc1/mari-ner-like) transposons of Aspergillus (54% in 33 amino acids,4e¡03) and to retrotransposon gag-pol polyproteins ofplants (57% in 40 amino acids, 3e¡ 04), respectively,although they did not form long ORFs. These sequencesmight belong to old degenerated transposable elements.

3.7. The gene organization in the P. chrysogenum AR is not conserved in A. nidulans

No reports about tandem repeat ampliWcation of thepenicillin biosynthetic genes in A. nidulans have been made.

To know if some of the genes found in P. chrysogenumadjacent to the three structural genes or within the AR werealso present in similar arrangement in A. nidulans, we per-formed database searches and analyzed the ORFs anno-tated in the A. nidulans genome database (Broad Institute)in a DNA region of 50 kb containing the three penicillinstructural genes, which corresponds to the left part of con-tig 1.45. Nine ORFs other than the structural genes arepresent in this region, Wve downstream of pcbAB and fourdownstream of penDE (Fig. 4). Positions of ORFs andintrons were inferred from the annotated ORFs in contig1.45 and further informatic analysis. Some annotatedORFs were modiWed on checking that other conWgurationsWt better the consensus intron/exon junction sequences andbased on alignments resulting from searches in databases.Thus, the large ORF AN2620.2 was splitted in two:AN2620_1.2 and AN2620_2.2, showing the Wrst one iden-tity to a cellobiose dehydrogenase of similar size. A modi-Wed intron/exon structure of ORF AN2624.2 was found. In

Fig. 2 (continued)

Fig. 3. Growth of strains Wis54-1255 (prototroph), npe10 (lacking the 56.8 kb AR including ORF12), L2 (lysine auxotroph, defective in homoaconitase),and SR1¡ (lysine auxotroph, disrupted in the lys7 gene encoding saccharopine reductase) on Czapeck minimal medium (with NaNO3 as nitrogen source)and Czapeck-lysine (with lysine as sole nitrogen source). Note that P. chrysogenum npe10 grows less than its parental strain Wis54-1255 on Czapeck andrecovers full growth on Czapeck-lysine. Strain SR1¡ has a very weak growth on Czapeck-lysine, as saccharopine reductase is involved in lysine catabolism,while strain L2 has a good growth on this lysine supplemented medium.

626 F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629

most cases the annonated ORFs were maintained, althoughin cases like AN2617.2 and AN2618.2, alternative in frameATG codons were found which would eliminate the Wrstintron and the Wrst very short exon in both putative genes.Three of the ORFs in the 50 kb region showed no similari-ties to genes present in databases, while the other sixshowed diVerent degrees of identity to proteins with diVer-ent functions (Fig. 4C). The ORFs present in the P. chrys-ogenum AR and A. nidulans 50 kb region did not show anypattern of conservation other than that of the three struc-tural genes, which are very similar (MacCabe et al., 1990;Montenegro et al., 1990).

Of particular interest is the presence adjacent to the A.nidulans penDE gene of two putative transporter proteinsbelonging to the major facilitator superfamily (MFS).

Apart from the ORFs, it is interesting to note the pres-ence of the MATE-4 element 14.5 kb downstream from theaatA (DpenDE) gene. MATE-4 is a member of the MATE(mobile Aspergillus transformation enhancer) family ofgenetic elements believed to have or have had transpositionability in the past (Aleksenko and Clutterbuck, 1997).Inserted within MATE-4 lies a 1363-bp fragment, in whichthe Wrst exon of ORF AN2627.2, encoding a transposase, islocated (Fig. 4). The remaining exons described in the fea-ture map of contig 1.45 from the A. nidulans genomic data-

base lie within the sequences belonging to MATE-4, andtherefore must not correspond to coding regions ofAN2627.2. Clutterbuck (2004) has identiWed this 1363-bpinsert as a Mariner/Tc1-like transposable element. NeitherMATE nor Mariner/Tc1-like elements have been foundwithin the P. chrysogenum AR, but a short sequence situ-ated in the right Xanking region adjacent to the AR (t1 inFig. 1) shows similarity to Mariner/Tc1 class transposases(see above). A screening performed with a MATE probe ona P. chrysogenum genomic library and on total DNA bySouthern blot failed to identify sequences hybridizing withthis element (Fierro et al., 1996b), which indicates theabsence of MATE-like elements in P. chrysogenum.

4. Discussion

The 56.8 kb ampliWed region present in high-producingindustrial strains of Penicillium chrysogenum does not con-tain any gene essential for growth since strains lacking thiswhole region, like P. chrysogenum npe10 (Cantoral et al.,1993; Fierro et al., 1996a) are viable and phenotypicallysimilar to the parental strains except for the lack of penicil-lin production ability. However, some of the genes in theAR may provide precursors or confer advantages forincreased penicillin production.

Fig. 4. Map of the 50 kb DNA region from the A. nidulans contig 1.45 which contains the penicillin biosynthetic structural genes, downstream the pcbABgene (A) and downstream the penDE gene. (B) Relative positions and sizes of the annotated ORFs are indicated with black arrows; introns appear aswhite boxes. Positions of ORFs and introns were inferred from the annotated ORFs in contig 1.45 and further analysis in search for start and end codonsand introns (see text). Only the Wrst annotated exon of ORF AN2627.2 is shown. At the right end of the analyzed 50 kb region, the MATE-4 element (Clut-terbuck, 2004) is indicated with a grey bar, see text. (C) Conserved domains, highest score matches and putative function of the diVerent ORFs.

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 627

The most likely candidates to be ancillary proteins forpenicillin biosynthesis are ORF11, ORF12, and ORF13.They are contiguous and situated downstream of thepenDE gene, thus forming a six-gene cluster together withthe three structural penicillin biosynthetic genes. An inter-esting gene present downstream of the penDE is ORF11,encoding a hypothetical protein with a D-amino acid oxi-dase (deaminating) domain. D-amino oxidases of yeasts andfungi (e.g., Trigonopsis variabilis) are known to convert theD-�-aminoadipic chain of cephalosporin C or penicillin Nto �-ketoglutarate (Sentheshanmuganathan and Nicherson,1962; Pollegioni et al., 2004). �-Lactam by-products con-taining �-ketoglutarate instead of �-aminoadipate havebeen reported in �-lactam producer strains (Kitano et al.,1976a,b), e.g., the formation of 6-(5-hydroxy-n-valeram-ido)-penicillanic acid, a product of the deaminatingdecarboxylation of penicillin N (presumably formed by aD-amino acid oxidase), along with penicillin N was reportedin the cephalosporin producer Paecilomyces carneus(Kitano et al., 1976a). Similarly, in addition to isopenicillinN and penicillin N, Cephalosporium chrysogenum andmutant strains of Cephalosporium polyaleurum and C. acre-monium produce a cephalosporin compound, 7�-(4-carb-oxybutanamido)-3-cephem-4-carboxylic acid, having an�-ketoglutarate side chain (Kitano et al., 1976b). Thiscompound is probably produced by an oxidase-mediateddeamination of a penicillin or cephalosporin intermediate.The possible involvement of the ORF11-encoded protein indeamination or isomerization of �-aminoadipic is a subjectof great interest.

The Wnding that ORF12 encodes a saccharopine dehy-drogenase-like protein is important because of its possibleinvolvement in the biosynthesis of lysine and the penicillinprecursor �-aminoadipic acid. High penicillin-producingstrains possess a very active �-aminoadipic pathway forsynthesis of the large amount of this intermediate requiredfor penicillin biosynthesis. In yeasts and Wlamentous fungithere are two separate enzymes, saccharopine reductaseand saccharopine dehydrogenase involved in the conver-sion of �-aminoadipic semialdehyde into lysine, whereas inhigher eucaryotic organisms, such as Arabidopsis thaliana,Mus musculus, and Homo sapiens, there is a bifunctionalsaccharopine reductase/dehydrogenase enzyme. In animals,this enzyme is involved in lysine catabolism, since lysine isan essential amino acid and animals lack the ability to syn-thesize it.

In P. chrysogenum we have reported that lysine is catab-olized to �-aminoadipic acid by two pathways (Esmahanet al., 1994; Martín de Valmaseda et al., 2005); one of theminvolves a lysine/ornithine omega-aminotransferaseencoded by the oat1 gene (Naranjo et al., 2005) and theother is a reversal of the lysine biosynthetic pathway cataly-sed by saccharopine dehydrogenase/saccharopine reduc-tase.

Since P. chrysogenum npe10 lacking ORF12 growspoorly on Czapek minimal medium when compared withthe parental Wis 54-1255 strain, it seems clear that the sac-

charopine dehydrogenase encoded by ORF12 contributesto lysine biosynthesis, but it is not strictly essential forgrowth, i.e., there is a diVerent isoenzyme (LysA) contribut-ing to primary metabolism (lysine biosynthesis). The sac-charopine dehydrogenase encoded by ORF12 is similar tothat of cyanobacteria and this similarity may indicate thatit has been transferred horizontally from those bacteria toP. chrysogenum. Duplication of enzymes for the biosynthe-sis of precursors of secondary metabolites structurally anal-ogous to primary metabolites is frequent in gene clustersfor secondary metabolites (Martín and Liras, 1989).

Particularly interesting is the presence in the AR ofORF13, encoding a hypothetical protein with a conservedZn(II)2-Cys6 fungal-type DNA-binding domain, similar tothat of the A. nidulans transcriptional activator AlcR. Sofar, several transcription factors have been reported to con-trol expression of the penicillin genes and bind to regula-tory boxes in their promoters, both in A. nidulans and in P.chrysogenum, including PacC, AreA, and CreA (Brakhage,1998; Martín, 2000; Fierro et al., 2002); however, no path-way-speciWc transcriptional regulator for penicillin biosyn-thesis has yet been found. This contrasts with the presenceof speciWc regulators in other secondary metabolite path-ways as is the case of aXatoxins (Yu et al., 1996) or tricho-thecene (Proctor et al., 1996), where speciWc regulators havebeen reported. It is likely that the protein encoded byORF13 regulates transcription of the penicillin genes. Thisis being studied in more detail.

One of the signiWcant Wndings in this work was the lackof conservation of the genes adjacent to the structural peni-cillin biosynthetic genes in P. chrysogenum and A. nidulans.Even though other genes adjacent to the structural onesmay have a role in penicillin production, only the threestructural genes have been conserved clustered in bothfungi during evolution. The considerably higher penicillinproduction in P. chrysogenum than in A. nidulans is proba-bly related to the presence of speciWc genes enhancing peni-cillin biosynthesis in the former strain.

Another interesting Wnding was that while two putativeMFS transporter genes are located downstream of thepenDE gene in A. nidulans, no apparent transporter genesare present in the P. chrysogenum AR. Moreover, the geneatrD, reported to have a role in penicillin production in A.nidulans (Andrade et al., 2000) is not located in the proxim-ity of the penicillin gene cluster in this fungus. The mecha-nisms by which penicillin and �-lactams in general aresecreted from the cell in fungi remains poorly understood(Martín et al., 2005).

The presence of repeated DNA sequences as those founddownstream of the pcbAB gene (A1–A2 and B1–B2) is fre-quent in eucaryotic genomes. In most cases these non-tan-dem repeats originate from transposable elements.Retrotransposons carry long terminal direct repeats (LTR)and genes gag and pol, the latter containing retrotranspos-ase and RNase H activities; they also contain often apurine-rich sequence just before the right end LTR(Kempken and Kück, 1998). These features are present in

628 F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629

the repeats B1–B2, where a polyA sequence with someinterspersed thymines is present just before B2. To checkthe possibility that A1–A2 and B1–B2 may representdegenerated retrotransposons, thorough database searcheswere performed. No identities to known transposable ele-ments or genes present habitually in them were detected.Two weak matches in the B1–B2 inter-repeats region werefound with a RNA polymerase II largest subunit fromhamster, a gene not reported to be present in retrotranspo-sons, and other with a ribonuclease E and G in the oppositeorientation. The low signiWcance of these matches, and theabsence of other features typical of LTR transposable ele-ments, like short inverted repeats at the ends of each LTR,did not provide evidence to support the presence of adegenerated retrotransposon. The polyA sequence can alsobe interpreted as the terminal polyA of a LINE-like ele-ment (i.e., a retrotrasposon devoid of LTRs) but no goodmatches were found to support this possibility.

The signiWcant degree of divergence of the A1–A2 repeats(91% identity) and especially the B1–B2 repeats (75.4% iden-tity), may be due to accumulation of mutations, in which casethe repeats must have appeared at an early time in the evolu-tion of the fungus, or they may have been produced by spe-ciWc mechanisms which, similarly to the RIP mutationoperating in Neurospora crassa (Selker and Garret, 1988) andAspergillus nidulans (Nielsen et al., 2001; Clutterbuck, 2004),mutate repeated sequences in the genome.

Note added in proof

While this article was being evaluated we learnt that verysimilar results have been obtained by Marco van der Berget al. at DSM (Delft, Holland).

Acknowledgments

This work was supported by a Grant of the CICYT(BIO2000-1726-C02-02) and the European Union (Euro-fung II). N.I. Castillo was supported by a fellowship of theAECI. Ministry of Foreign AVairs, Madrid, Spain. Wethank R.L. Bovenberg (DSM, The Netherlands) for sup-port and B. Martín, J. Merino, A. Casenave and M. Álvarezfor excellent technical assistance.

References

Aharanowitz, Y., Cohen, G., Martín, J.F., 1992. Penicillin and cephalospo-rin biosynthetic genes: structure, organization, regulation, and evolu-tion. Annu. Rev. Microbiol. 46, 461–495.

Aleksenko, A., Clutterbuck, A.J., 1997. Autonomous plasmid replicationin Aspergillus nidulans: AMA1 and MATE elements. Fungal Genet.Biol. 21, 373–387.

Andrade, A.C., Van Nistelrooy, J.G.M., Peery, R.B., Skatrud, P.L., DeWaard, M.A., 2000. The role of ABC transporters from Aspergillusnidulans in protection against cytotoxic agents and in antibiotic pro-duction. Mol. Gen. Genet. 263, 966–977.

Ausubel, F.M., Brent, R., Knigston, R.E., Moore, D., Smith, J.A., Seidman,J.G., Struhl, K. (Eds.), 1987. Current Protocols in Molecular BiologyWiley, New York.

Barredo, J.L., Díez, B., Alvarez, E., Martín, J.F., 1989. Large ampliWcationof a 35-kb DNA fragment carrying two penicillin biosynthetic genes inhigh penicillin producing strains of Penicillium chrysogenum. Curr.Genet. 16, 453–459.

Brakhage, A., 1998. Molecular regulation of �-Lactam biosynthesis in Wla-mentous fungi. Microbiol. Mol. Biol. Rev. 62, 547–585.

Brown, D.W., Yu, J.-H., Kelkar, H.S., Fernandes, M., Nesbitt, T.C., Keller,N.P., Adams, T.H., Leonard, T.J., 1996. Twenty Wve co-regulated tran-scripts deWne a sterigmatocystin gene cluster in Aspergillus nidulans.Proc. Natl Acad. Sci. USA 93, 1418–1422.

Cantoral, J.M., Gutiérrez, S., Fierro, F., Gil-Espinosa, S., van Liemptg, H.,Martín, J.F., 1993. Biochemical characterization and molecular genet-ics of nine mutants of Penicillium chrysogenum impaired in penicillinbiosynthesis. J. Biol. Chem. 268, 737–744.

Clutterbuck, A.J., 2004. MATE transposable elements in Aspergillus nidu-lans: evidence of repeat-induced point mutation. Fungal Genet. Biol.41, 308–316.

Díez, B., Gutiérrez, S., Barredo, J.L., van Solingen, P., van der Voort,L.H.M., Martín, J.F., 1990. The cluster of penicillin biosynthetic genes.J. Biol. Chem. 265, 16358–16365.

Esmahan, C., Álvarez, E., Montenegro, E., Martín, J.F., 1994. Catabolismof lysine in Penicillium chrysogenum leads to formation of 2-aminoa-dipic acid, a precursor of penicillin biosynthesis. Appl. Environ. Micro-biol. 60, 1705–1710.

Fierro, F., Gutiérrez, S., Díez, B., Martín, J.F., 1993. Resolution of fourlarge chromosomes in penicillin producing Wlamentous fungi: the peni-cillin gene cluster is located on chromosome II (9.6 Mb) in Penicilliumnotatum and chromosome I (10.4 Mb) in Penicillium chrysogenum. Mol.Gen. Genet. 241, 573–578.

Fierro, F., Barredo, J.L., Díez, B., Gutiérrez, S., Fernández, F.J., Martín,J.F., 1995. The penicillin gene cluster is ampliWed in tandem repeatslinked by conserved hexanucleotide sequences. Proc. Natl. Acad. Sci.USA 92, 6200–6204.

Fierro, F., Montenegro, E., Gutiérrez, S., Martín, J.F., 1996a. Mutantsblocked in penicillin biosynthesis show a deletion of the entire penicil-lin gene cluster at a speciWc site within a conserved hexanucleotidesequence. Appl. Microbiol. Biotechnol. 44, 597–604.

Fierro, F., Kosalkova, K., Gutiérrez, S., Martín, J.F., 1996b. Autono-mously replicating plasmids carrying the AMA1 region in Penicilliumchrysogenum. Curr. Genet. 29, 482–489.

Fierro, F., Martín, J.F., 1999. Molecular mechanisms of chromosomalrearrangement in fungi. CRC Crit. Rev. Microbiol. 25, 1–17.

Fierro, F., Martín, J.F., Kosalková, K., 2002. Sulfur-containing �-lactamantibiotics: enzymes, genes and regulation of the biosynthesis. In: Fier-roy, F., Martín, J.F. (Eds.), Microbial Secondary Metabolites: Biosyn-thesis, Genetics and Regulation. Research Signpost, Trivandrum,India, pp. 179–210.

Gutiérrez, S., Velasco, J., Fernández, F.J., Martín, J.F., 1992. The cefG geneof Cephalosporium acremonium is linked to the cefEF gene and encodesa deacetylcephalosporin C-acetyltransferase closely related to homo-serine O-acetyltransferase. J. Bacteriol. 174, 3056–3064.

Gutiérrez, S., Fierro, F., Casqueiro, J., Martín, J.F., 1999a. Gene organiza-tion and plasticity of the �-lactam genes in diVerent Wlamentous fungi.Antonie van Leeuwenhoek 75, 81–94.

Gutiérrez, S., Marcos, A.T., Casqueiro, J., Kosalkova, K., Fernández, F.J.,Velasco, J., Martín, J.F., 1999b. Transcription of the pcbAB, pcbC andpenDE genes of Penicillium chrysogenum AS-P-78 is repressed by glu-cose and the repression is not reversed by alkaline pHs. Microbiology145, 317–324.

Hull, E.P., Green, P.M., Arst Jr., H.N., Scazzochio, C., 1989. Cloning andphysical characterization of the L-proline catabolism gene cluster ofAspergillus nidulans. Mol. Microbiol. 3, 553–559.

Keller, N.P., Hohn, T.M., 1997. Metabolic pathway gene clusters in Wla-mentous fungi. Fungal Genet. Biol. 21, 17–29.

Kempken, F., Kück, U., 1998. Transposons in Wlamentous fungi—factsand perspectives. BioEssays 20, 652–659.

Kitano, K., Fujisawa, Y., Katamoto, K., Nara, K., Nakao, Y., 1976a.Occurrence of 7b—(4-carboxybutanamido)-cephalosporin compounds

F. Fierro et al. / Fungal Genetics and Biology 43 (2006) 618–629 629

in the culture broth of some strain of the genus Cephalosporium. J. Fer-ment. Technol. 54, 712–719.

Kitano, K., Kintaka, K., Suzuki, S., Katamoto, K., Nara, K., Nakao, Y.,1976b. A novel penicillin produced by strains of the genus Paecillomy-ces. J. Ferment. Technol. 54, 705–711.

MacCabe, A.P., Riach, M.B.R., Unkles, S.E., Kinghorn, J.R., 1990. TheAspergillus nidulans npeA locus consists of three contiguous genesrequired for penicillin biosynthesis. EMBO J. 9, 279–287.

Martín de Valmaseda, E.M., Campoy, S., Naranjo, L., Casqueiro, J., Mar-tín, J.F., 2005. Lysine is catabolized to 2-aminoadipic acid in Penicil-lium chrysogenum by an �-aminotransferase and to saccharopine by alysine 2-ketoglutarate reductase. Characterization of the omega-ami-notransferase. Mol. Genet. Genomics 274, 272–282.

Martín, J.F., Casqueiro, J., Liras, P., 2005. Secretion systems for secondarymetabolites: how producer cells send out messages of intercellularcommunication. Curr. Opin. Microbiol. 8, 282–293.

Martín, J.F., Liras, P., 1989. Organization and expression of genes involvedin the biosynthesis of antibiotics and other secondary metabolites.Annu. Rev. Microbiol. 43, 173–206.

Martín, J.F., 2000. Molecular control of expression of penicillin biosynthe-sis genes in fungi: regulatory proteins interact with a bidirectional pro-moter region. J. Bacteriol. 182, 2355–2362.

Martín, J.F., 1998. New aspects of genes and enzymes for �-lactam antibi-otic biosynthesis. Appl. Microbiol. Biotechnol. 50, 1–15.

Martín, J.F., Gutiérrez, S., Demain, A.L., 1997. �-Lactams. In: Anke, T.(Ed.), Fungal Biotechnology Chapman & Hall GmbH, Weinheim Ger-many, pp. 91–127.

Montenegro, E., Barredo, J.L., Gutierrez, S., Diez, B., Alvarez, E., Martin,J.F., 1990. Cloning, characterization of the acyl-CoA:6-aminopenicill-anic acid acyltransferase gene of Aspergillus nidulans and linkage to theisopenicillin N synthase gene. Mol. Gen. Genet. 221, 322–330.

Montenegro, E., Fierro, F., Fernández, F.J., Gutiérrez, S., Martín, J.F.,1992. Resolution of chromosomes III and VI of Aspergillus nidulans bypulsed-Weld gel electrophoresis shows that the penicillin biosyntheticpathway genes pcbAB, pcbC, and penDE are clustered on chromosomeVI (3.0 megabases). J. Bacteriol. 174, 7063–7067.

Naranjo, L., Lamas-Maceiras, M., Ullán, R.V., Campoy, S., Teijeira, F.,Casqueiro, J., Martín, J.F., 2005. Characterization of the oat1 gene ofPenicillium chrysogenum encoding an omega-aminotransferase: induc-tion by L-lysine, L-ornithine and L-arginine and repression by ammo-nium. Mol. Genet. Genomics 274, 283–294.

Naranjo, L., Martin de Valmaseda, E., Bañuelos, O., López, P., Riaño, J.,Casqueiro, J., Martín, J.F., 2001. Conversion of pipecolic acid intolysine in Penicillium chrysogenum requires pipecolate oxidase and sac-charopine reductase: characterization of the lys7 gene encoding sac-charopine reductase. J. Bacteriol. 183, 7165–7172.

Newbert, R.W., Barton, B., Greaves, P., Harper, J., Turner, G., 1997. Anal-ysis of a commercially improved Penicillium chrysogenum strain series:

involvement of recombinogenic regions in ampliWcation and deletionof the penicillin biosynthesis gene cluster. J. Ind. Microbiol. Biotechnol.19, 18–27.

Nielsen, M.L., Hermansen, T.D., Aleksenko, A., 2001. A family of DNArepeats in Aspergillus nidulans has assimilated degenerated retrotrans-posons. Mol. Genet. Genomics 265, 883–887.

Pollegioni, L., Caldinelli, L., Molla, G., Sacchi, S., Pilone, M.S., 2004. Cata-lytic properties of D-amino acid oxidase in cephalosporin C bioconver-sion: a comparison between proteins from diVerent sources.Biotechnol. Prog. 20, 467–473.

Proctor, R.H., Hohn, T.M., McCormick, S.P., Desjardins, A.E., 1996. Tri 6encodes and unusual zinc Wnger protein involved in regulation oftrichothecene biosynthesis in Fusarium sporotrichioides. Appl. Environ.Microbiol. 61, 1923–1930.

Sambrook, J., Fritsh, E.F., Maniatis, T., 1989. Molecular Cloning: A Labo-ratory Manual. Cold Spring Harbor Laboratory Press, Cold SpringHarbor, New York.

Selker, E.U., Garret, P.W., 1988. DNA sequence duplications triggergene inactivation in Neurospora crassa. Proc. Natl. Acad. Sci. USA85, 6870–6874.

Sentheshanmuganathan, S., Nicherson, W.J., 1962. Transaminase and D-amino acid oxidase of Trigonopsis variabilis. J. Gen. Microbiol. 27,465–471.

Skatrud, P.L., Queener, S.W., 1989. An electrophoretic molecular karyo-type for an industrial strain of Cephalosporium acremonium. Gene 78,331–338.

Smith, D.J., Burnham, M.K.R., Bull, J.H., Hodgson, J.E., Ward, J.M.,Browne, P., Brown, J., Barton, B., Earl, A.J., Turner, G., 1990. �-lactamantibiotic biosynthetic genes have been conserved in clusters in pro-karyotes and eukaryotes. EMBO J. 9, 741–747.

Smith, D.J., Bull, J.H., Edwards, J., Turner, G., 1989. AmpliWcation ofthe isopenicillin N synthetase gene in a strain of Penicillium chrysog-enum producing high levels of penicillin. Mol. Gen. Genet. 216,492–497.

Ullán, R.V., Casqueiro, J., Bañuelos, O., Fernández, F.J., Gutiérrez, S.,Martín, J.F., 2002a. A novel epimerization system in fungal second-ary metabolism involved in the conversion of isopenicillin N intopenicillin N in Acremonium chrysogenum. J. Biol. Chem. 277, 46216–46225.

Ullán, R.V., Liu, G., Casqueiro, J., Gutiérrez, S., Bañuelos, O., Martín,J.F., 2002b. The cefT gene of Acremonium chrysogenum C10encodes a putative multidrug eZux pump protein that signiWcantlyincreases cephalosporin C production. Mol. Genet. Genomics 267,673–683.

Yu, J.H., Butchko, R.A.E., Fernandes, M., Keller, N.P., Leonard, T.J.,Adams, T.H., 1996. Conservation of structure and function of the aXa-toxin regulatory gene aXR from Aspergillus nidulans and AspergillusXavus. Curr. Genet. 29, 549–555.