141 KB – 13 Pages

PAGE – 1 ============
Fungal Genetics and Biology 44 (2007) 139–151 www.elsevier.com/locate/yfgbi 1087-1845/$ – see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.fgb.2006.07.008 Suppression of the acuH13 and acuH31 nonsense mutations in the carnitine/acylcarnitine translocase ( acuH) gene of Aspergillus nidulans by the G265S substitution in the domain 2 of the release factor eRF1 Óscar Martínez a, Esther Marco b, Federico Gago b, Fernando Laborda a, J. Ramón De Lucas a,¤a Departamento de Microbiología y Parasitología, Facultad de Farmacia, Campus Universitario, Universidad de Alcalá, Carretera Madrid-Barcelona Km 33, Alcalá de Henares ES-28871, Madrid, Spain b Departamento de Farmacología, Facultad de Farmacia, Campus Universitario, Universidad de Alcalá, Carretera Madrid-Barcelona Km 33, Alcalá de Henares ES-28871, Madrid, Spain Received 5 April 2006, and in revised form 28 June 2006; accepted 24 July 2006 Available online 12 September 2006 Abstract A search for suppressors of the carnitine/acylcarnitine translocase (CACT) de Wciency in Aspergillus nidulans permitted the identi Wca- tion of the suaE7 mutation, mapping at a new translational suppressor ( suaE ) gene. The suaE gene is essential in A. nidulans and encodes the eukaryotic release factor 1 (eRF1). The suaE7 mutation suppresses two acuH alleles ( acuH13 and acuH31 ), both carrying nonsense mutations in the CACT encoding gene that involve the replacement of a CAG (Gln) codon with a premature TAG stop codon. In con- trast, the suaE7 gene does not suppress the acuH20 amber nonsense mutation involving a TGG !TAG change. The phenotype associ- ated to the suaE7 mutation strictly resembles that of mutants at the suaA and suaC genes, two translational suppressor genes previously identi Wed, suggesting that their gene products might functionally interact in translation termination. Sequencing of the suaE7 gene allowed the identi Wcation of a mutation in the domain 2 of the omnipotent class-1 eukaryotic release factor involving the Gly265Ser sub- stitution in the A. nidulans eRF1. This mutation creates a structural context unfavourable for normal eRF binding that allows the mis- reading of stop codons by natural suppressor tRNAs, such as the tRNAs Gln . Structural analysis using molecular modelling of A. nidulans eRF1 domain 2 bearing the G265S substitution and computer simulation results suggest that this mutation might impair the necess ary conformational changes in the eRF1 to optimally recognize the stop codon and simultaneously interact with the peptidyl transfer ase cen- tre of the 60S ribosomal subunit. 2006 Elsevier Inc. All rights reserved. Keywords:Aspergillus nidulans ; Carnitine/acylcarnitine translocase; suaE gene; eRF1; Release factors; Translation termination; Translational suppres- sors; Natural suppressor tRNAs; Nonsense suppressors; Stop codon translation 1. Introduction The carnitine/acylcarnitine translocase (CACT) medi- ates transport of acylcarnitines of di Verent length across the mitochondrial inner membrane for their Wnal oxidation in the mitochondrial matrix ( De Lucas etal., 1999; Palmi- eri, 2004 ). De Wciency of human CACT results in the most severe disorder of fatty acid -oxidation and is usually lethal within a few hours or days after birth ( Pande etal., 1993; Stanley etal., 1992 ). The fact that this de Wciency is associated with lethality throughout the metazoan lineage accounts for the lack of animal models for the disease in both vertebrates and invertebrates ( De Lucas etal., 1999 ).*Corresponding author. Fax: +34 918854621. E-mail address: joser.lucas@uah.es (J. Ramón De Lucas).

PAGE – 2 ============
140 Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 In the Wlamentous fungus Aspergillus nidulans , the CACT is encoded by the acuH gene. In contrast with ani- mals, the CACT de Wciency in A. nidulans is conditionally lethal because, although a functional CACT is required for growth on long-chain fatty acids and C 2 compounds such as acetate, the enzyme is not essential when alternative car- bon sources such as carbohydrates or amino acids are uti- lized ( De Lucas etal., 1999 ). This important metabolic feature, together with the amenability of A. nidulans for classical and molecular genetic studies, pinpoints this eukaryotic microorganism as an excellent model to investi- gate important molecular and biochemical aspects of this inborn error of metabolism. Accordingly, we previously developed a metabolic model that allows the functional characterization of human CACT mutations such that neu- tral, single residue substitution-causing polymorphisms can be discriminated from potentially pathogenic missense mutations ( Perez etal., 2003 ).In this work, we utilized the model Wlamentous fungus A. nidulans to identify mutations that suppress the de W-ciency of acuH mutants for growth on acetate and oleate media aiming to discover a putative therapeutic target whose modulation could help in preventing the lethality associated to the CACT de Wciency. Although no physiolog- ical suppressor of the CACT de Wciency was found, we iden- tiWed and further characterized a suppressor mutation mapping at a new translational suppressor gene in A. nidu- lans , the suaE gene, which suppresses two nonsense muta- tions in the acuH gene involving the substitution of a CAG (Gln) codon for a premature TAG amber stop codon. Cloning and sequencing of the suaE gene, as performed in this work, permitted the identi Wcation of the A. nidulans eRF1. The eukaryotic release factor 1 (eRF1) terminates protein synthesis by recognizing mRNA stop codons at the ribosomal A-site and interacting with the ribosomal pepti- dyl transferase centre where the peptidyl-tRNA ester bond is hydrolysed so as to release the free polypeptide ( Frolova etal., 1999; Ito etal., 2000; Inge-Vechtomov etal., 2003 ).The eRF1 is composed of three domains. The N-terminal domain (domain 1) is proposed to participate in stop codon recognition ( Song etal., 2000; Bertram etal., 2000; Chav- atte etal., 2001; Frolova etal., 2002 ). Domain 2 is responsi- ble for the peptidyl transferase hydrolytic activity ( Frolova etal., 1999; Seit-Nebi etal., 2001 ) and the C-terminal part of eRF1 (domain 3) interacts with eRF3 (eukaryotic release factor 3) ( Kisselev etal., 2003 ).The phenotype of suaE7 mutant strains clearly resem- bles the phenotype of mutants at the suaA and suaC genes, two translational suppressor genes previously identi Wed in A. nidulans (Roberts etal., 1979 ) and whose unknown gene products presumably interact with eRF1 in the termination of translation. By sequencing the suppressor suaE7 gene we identi Wed the G265S substitution in the eRF1 domain 2 that must result in limited competitiveness with natural suppressor tRNAs capable of illegitimately translating mRNA stop codons. Molecular modelling was used to build the three-dimensional structures of domain 2 of wild- type and mutant forms of A. nidulans eRF1. Our computer simulation results suggest that this missense mutation does not directly a Vect stop codon recognition or the presumed catalytic domain but likely impairs the necessary conforma- tional changes that must occur in the eRF1 to optimally recognize the stop codon on the mRNA and simulta- neously interact with the peptidyl transferase centre located at the 60S ribosomal subunit ( Klaholz etal., 2003; Nakam- ura and Ito, 2003 ).2. Materials and methods 2.1. Strains and plasmids The A. nidulans strains used in this work are listed in Table 1 . The chromosome-III-speci Wc cosmid library obtained from the Fungal Genetics Stock Center (Univer- sity of Missouri, Kansas, USA) was employed to clone the suaE gene of A. nidulans . The suaE gene was subcloned using the plasmid pUC19 and established molecular biol- ogy techniques ( Sambrook etal., 1989 ). Plasmid pAL3 containing the Neurospora crassa pyr-4 gene as a fungal selectable marker and the promoter of the A. nidulans alcA gene ( alcA P) (Waring etal., 1989 ) was also employed. Escherichia coli DH5 (BRL, USA) served as the recipient strain for plasmid transformation. Table 1 Aspergillus nidulans strains used in this study StrainGenotypeSource or reference R21pabaA1 yA2 Armitt etal. (1976) R153 wA3 ; pyroA4 Armitt etal. (1976) G191 pabaA1 pyrG89 ; uaY9 ; fwA1 De Lucas etal. (2001) WH8wA3 ; pyroA4 ; acuH8 De Lucas etal. (1997) WH13 wA3 ; pyroA4 ; acuH13 De Lucas etal. (1997) WH20 wA3 ; pyroA4 ; acuH20 De Lucas etal. (1997) WH24 wA3 ; pyroA4 ; acuH24 De Lucas etal. (1997) WH31 wA3 ; pyroA4 ; acuH31 De Lucas etal. (1997) YH8 pabaA1 yA2 ; acuH8 De Lucas etal. (1997) YH13 pabaA1 yA2 ; acuH13 De Lucas etal. (1997) YH20 pabaA1 yA2 ; acuH20 De Lucas etal. (1997) YH24 pabaA1 yA2 ; acuH24 De Lucas etal. (1997) YH31 pabaA1 yA2 ; acuH31 De Lucas etal. (1997) PPA14 biA1 ; acuH::pyr4 +; argB2 Perez etal. (2003) WSol7H wA3 suaE7 ; pyroA4 ; acuH13 This study YSol7H pabaA1 yA2 ; suaE7 ; acuH13 This study WSol7 wA3 suaE7 ; pyroA4 This study YSol7 pabaA1 yA2 ; suaE7 This study S7HG biA1 ; argB2 suaE7 ; pyroA4 ;acuH13 This study G351 pabaA1 ; alX4 suaA101 ; sB43 ; alcR125 ; fwA1 Glasgow stock G724 pabaA1 ; alX4 ; sB43 ; suaC109 ; fwA1 Glasgow stock H9pabaA1 ; alX4 suaB111 ; sB43 ; fwA1 Dr. H. Sealy-Lewis WG355 biA1; argB2 galA1 Perez etal., (2003 )MS2.7 biA1 ;wA3 ;galA1 ;pyroA4 ;facA303 ;sB3 ;nicB8 ;riboB2 Dr. M. Grindle

PAGE – 3 ============
Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 141 2.2. Aspergillus nidulans media and culture conditions Aspergillus nidulans was grown in Aspergillus complete medium (CM) and Aspergillus minimal medium (MM) (Armitt etal., 1976 ). The following carbon sources were added to the MM for phenotypic analysis of acuH mutants: 1% glucose (non-selective conditions), 0.1M acetate or 6 MM oleate in 1% Tergitol NP40 (Sigma) (selective condi- tions) (Valenciano etal., 1996 ). Growth tests of palB mutants were carried out in CM containing 0.2M PO 4HNa 2 (pH 8) (restrictive conditions) and CM containing 0.5M PO 4H2Na(pH 5) (non-selective conditions). A. nidulans cultures were normally incubated at 37°C. However, when growth tests of suaE7, suaA101, suaB111 and suaC109 strains were per- formed, the following temperatures were used: 23, 28, 37 and 43°C. The following nutritional requirements were added as necessary: 1 g p-aminobenzoic acid ml ¡1, 0.05 g pyrodoxine ml¡1, 0.02 g biotin ml ¡1, 200 g arginine ml ¡1, 50g methio- nine ml ¡1, 3g panthotenic acid ml ¡1, 2g D-pantolactone ml¡1, 2g nicotinic acid ml ¡1, 2.5 g ribo Xavin ml ¡1, 2.44mg uridine ml ¡1. A. nidulans was transformed by the method of Balance etal. (1983 ).2.3. Classical genetic techniques Standard techniques for genetic manipulation of A.nidulans (Pontecorvo etal., 1953; Clutterbuck, 1974 ) wereused to synthesize heterokaryons and diploids for comple- mentation tests, to generate ascospores (meiotic) progeny with various genotypes, and for chromosome assignment by means of the parasexual cycle. 2.4. Molecular genetic techniques Cosmid and plasmid DNA was isolated by using the Qiagen Plasmid Kit (Qiagen) following the manufacturer’s instructions. Genomic DNA was isolated from mycelium grown on 1% glucose MM employing the method described by De Lucas etal. (2001 ). Southern blotting was carried out by established procedures ( Sambrook etal., 1989 ). DNA sequencing was carried out by using the GeneAmp PCR9700 system on an ABI-PRISM3100 automatic sequencer (Applied Biosystems). The sequence of the A. nidulans suaE gene was deciphered by sequencing the plasmids pXB25 and pPS150 using appropriate oligonucleotides. Molecular char- acterization of the suaE7 allele was performed by PCR- ampli Wcation of genomic DNA from WSol7H strain in four overlapping regions followed by direct sequencing of the PCR-ampli Wed fragments. These PCRs were carried out using the Pfu DNA polymerase (Promega) and the following primers: ERFfw1: 5 -TCTCA AGAACCAGGTGGCCAAC-3 ,ERFfw2: 5 -TCAAGCC AATCAACACGTCC-3 , ERFf w3: 5- GTCGCCGGTA TAATTCTTGC-3 , ERFf w4: 5 -CCTC GAATGGCTTG CGGA-3 , ERFrev1: 5 -GA GGCCGCTG AACCGTCAATCT-3 , ERFrev2: 5 – CCTG TTCAGCTCA CTTCC-3 , ERFrev3: 5 -CCTTAACCCTC GCACACT-3 and ERFrev4: 5 -TCAAGGCTTCGGTGTGAA-3 .To analyse the essentiality of the suaE gene the pALERF1 vector, based on pAL3 ( Waring etal., 1989 ), was constructed. Primers LKPeRF1: 5 -CAGGTACC ATACAGCTC-3 (including a KpnI site) and RBAeRF1: 5 -GGATCC TCAACACCGTAACA-3 (including a Bam HI site) were used in a PCR to amplify a 1065-bp fragment from nucleotides ¡34 to +1032 of the A. nidulans suaE gene (Accession no. AF451327). The PCR-ampli Wed fragment lacking the last 135 codons of the suaE ORF was cloned into pGEMTeasy giving the vector pGMERF1. The Kpn I–BamHI insert of this vector was ligated into the expression vector pAL3, previously digested with the same enzymes, to give pALERF1. 2.5. Molecular modelling and computer simulations To build the initial model for domain 2 of A. nidulans eRF1, the automated comparative protein modelling server SWISS-MODEL ( Guex etal., 1999 ) was employed using as template the X-ray crystal structure of human eRF1 depos- ited in the Protein Data Bank ( http://www.rcsb.org/pdb/ )with identi Wcation code 1DT9 ( Song etal., 2000 ). For each mutated residue, a manual search for optimal side chain conformations was performed so as to select the rotamer producing the lowest non-bonded energy. A short optimisa- tion run restraining all non-H atoms to their initial coordi- nates allowed readjustment of covalent bonds and van der Waals contacts without changing the overall conformation of the protein. This was followed by immersion of the mole- cule in a truncated octahedron of »8200 TIP3P water mole- cules with sides extending 8Å away from any protein atom. Periodic boundary conditions were applied and electrostatic interactions were represented using the smooth particle mesh Ewald method with a grid spacing of »1Å. The cuto Vdistance for the non-bonded interactions was 9Å. Unre- strained molecular dynamics simulations at 300K and 1atm were then run for 10ns using the SANDER module in AMBER8 (URL: http://amber.scripps.edu/doc8/ ). All bonds involving hydrogens were restrained to their equilibrium distances using the SHAKE algorithm and an integration step of 2fs was used throughout. The simulation protocol was essentially as described elsewhere ( Marco etal., 2003 )and involves a series of progressive energy minimizations followed by slow heating and equilibration periods before data collection. System coordinates were saved every 2ps after the Wrst 500ps of equilibration for further analysis. An average structure of the wild-type eRF1 domain 2 was used to model the G265S substitution, and the mutant protein was simulated under the same conditions as before. 3. Results 3.1. Suppression analysis of the carnitine/acylcarnitine deWciency in A. nidulans In a previous work ( De Lucas etal., 1997 ), we isolated Wve di Verent mutants in the A. nidulans acuH gene. These

PAGE – 4 ============
142 Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 acuH mutants were shown to be CACT-de Wcient and thus unable to grow on acetate or oleate MM ( De Lucas etal., 1999 ) although they grow normally on non-gluconeogenic carbon sources such as carbohydrates. Preliminary charac- terization of the acuH mutant alleles allowed the identi Wca- tion of three nonsense mutations, a missense mutation and a 1 bp-deletion frameshift in the acuH coding region. Phe- notypic and biochemical analyses demonstrated that all these mutations belong to the loss-of-function class ( Perez etal., 2003 ).Results obtained in this work (see below) suggested the re-examination of all existing acuH mutations in order to con Wrm the codon change previously reported ( Perez etal., 2003 ). This study con Wrmed that the published nucleotide changes for all acuH mutations were correct ( Perez etal., 2003 ). Nevertheless, it was found that in the acuH nonsense mutations the amino acid previously reported ( Perez etal., 2003 ) as replaced by the TAG codon was the preceding to the correct one. Accordingly, in this work it was con Wrmed that all nonsense mutations in the acuH gene: acuH13 ,acuH20 and acuH31 (DacuH253 , Armitt etal., 1976 )belong to the amber (UAG) class. Also, it was unequivo- cally demonstrated that in two of these mutants a prema- ture TAG stop codon replaces a CAG (Gln) codon in the acuH gene, while in the acuH20 strain a TGG (Trp) codon is replaced by the TAG stop codon ( Table 2 ).Spores (5 £108 conidia per plate) from the non-leaky strain WH13 carrying the acuH13 mutation ( Tables 1 and 2) were inoculated on selective media (0.1M acetate MM or 6 mM oleate MM) and incubated for 10 days at 37°C. Seven revertants growing on acetate medium (Sac strains) and Wve strains growing on oleate medium (Sol strains) were isolated. These revertants were tested for their ability to use acetate and oleate as the sole carbon source. Only four revertants, 2 Sac and 2 Sol strains, were able to grow on both media indicating that they carry intragenic or extragenic mutations suppressing the e Vect of the acuH13 allele. Such strains were crossed with the wild-type strain A.nidulans R21 ( Table 1 ) and phenotypic analysis of the prog- eny con Wrmed that a strain, named WSol7H, carries an extragenic suppressor gene of the acuH13 mutation. This extragenic suppressor gene, named suaE7 (see below), produces growth retardation and other pleotropic eVects at 37°C (see below) that allowed to distinguish clearly the four di Verent genotypes ( suaE7 ; acuH13 , suaE +;acuH13 , suaE7 ; acuH + and suaE +;acuH +) of the backcross progeny ( Fig.1 ), each representing 25% of the meiotic progeny analysed (200 colonies). In parallel, we utilized an A. nidulans acuH strain (PPA14) that had been constructed earlier by replacing the entire acuH ORF with the Neurospora crassa pyr-4 gene (Perez etal., 2003 ). Spores (5 £108 conidia per plate) from the PPA14 strain were inoculated on 0.1M acetate MM or 6mM oleate MM and incubated in the conditions described above. In addition, PPA14 spores were mutage- nized with uv-light (LD 90) and inoculated on the same media. No revertants able to grow on both acetate and ole- ate media were isolated suggesting the impossibility to obtain physiological suppressors of the CACT de Wciency in A. nidulans. 3.2. The WSol7H strain de Wnes a new translational suppressor (suaE) gene in A. nidulans According to the above statement, it was assumed that the extragenic suppressor represented by the WSol7H revertant belongs to the translational type. Nevertheless, appropriate genetic analyses were performed to con Wrm the mechanism of suppression. Strain WSol7H was crossed with complementary strains carrying the following acuH mutations: acuH , acuH8 , acuH13 , acuH20 , acuH24 and acuH31 (Table 1 ). The results obtained indicate that the identi Wed suppressor does not suppress the acuH8 (mis- sense) mutation, the acuH24 (frameshift) mutation or the acuH deletion. Among the acuH nonsense mutations, the acuH13 and acuH31 alleles were suppressed in a similar manner ( Fig.1 ) but not the acuH20 nonsense mutation (Table 2 ). This result demonstrates that the suppressor gene identi Wed in this work is allele speci Wc and belongs to the translational class. Interestingly, the acuH13 and acuH31 mutations involve a CAG !TAG change that causes the replacement of a glutamine codon with the amber stop codon, while the acuH20 allele consists of a TGG !TAG substitution that results in the replacement of a tryptophan codon with the same stop codon ( Table 2 ). These results show that the isolated suppressor is unable to suppress all amber nonsense mutations in the acuH gene, which sug- gests that its suppression capacity of these acuH mutations depends on the codon (amino acid) at which the nonsense mutation occurs. Table 2 Suppression analysis of Aspergillus nidulans acuH mutants by the translational suppressor suaE7 gene aPosition of nucleotide changes are from the A TG of the acuH genomic DNA sequence ( De Lucas etal., 1999 . EMBL Data Bank Accession No. AJ011563). StrainType of mutationNucleotide change aPredicted consequenceSuppression by suaE7 acuH8 Missense148C !TPro50LeuNo acuH13 Amber nonsense501C !TGln134StopYes acuH20 Amber nonsense861G !ATrp254StopNo acuH24 Frameshift430delTFrameshift after Asp109/Leu110StopNo acuH31 Amber nonsense76C !TGln26StopYes PPA14 acuH::pyr-4 +del(1-1590)No CACTNo

PAGE – 5 ============
Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 143 Since all the acuH nonsense mutants belong to the amber (UAG) class, to analyse whether the suaE7 allele is able to suppress all type of nonsense mutations, we used a diVerent gene, the palB gene. Three complete loss-of-func- tion palB mutants, each representing a di Verent type (amber, opal or ochre) of nonsense mutation (Herb N. Arst & Miguel A. Penalva personal communication to be detailed elsewhere) were utilized. The A. nidulans palB gene encodes a nuclear calpain-like protease ( Denison etal., 1995; Nozawa etal., 2003 ), which acts as a component of the alkaline ambient pH signal transduction pathway ( Arstand Penalva, 2003 ). Mutants at the palB locus are easily identi Wed by their inability to grow at alkaline pH (pH 8) while they grow at acidic pH (pH 5). Sexual crosses were established between strain WSol7 ( wA3 suaE7 ; pyroA4 ) and the three palB mutant strains. Phenotypic analyses of the meiotic progeny (over 200 colonies) from each cross indi- cated that the suaE7 gene weakly suppresses all palB muta- tions analysed. Therefore, the suaE7 mutation has to be considered as a suppressor of all type nonsense mutations. Further analysis of the phenotype of the WSol7H strain showed intriguing similarities with that caused by muta- tions in two translational suppressor genes previously iden- tiWed in A. nidulans , the suaA and suaC genes ( Roberts etal., 1979 ). The suaA and suaC genes, together with the suaB and suaD genes, were identi Wed in a search for transla- tional suppressors in A. nidulans (Roberts etal., 1979; Mar- tinelli, 1994 ). Mutations in the suaA and suaC genes, the Wrst class of translational suppressors, give rise to recessive suppression and cause pleiotropic alterations of morphol- ogy. It was proposed that such suppressors map either in a gene coding for a protein involved in translation termina- tion or for an essential ribosomal protein. In contrast, suaB and suaD mutant alleles, belonging to the second class and postulated to encode suppressor tRNAs, are semi-domi- nant and do not alter the phenotype in other ways ( Roberts etal., 1979; Martinelli, 1994 ).Genetic analysis testing for dominance/recessivity in dip- loids heterozygous for the suppressor suaE7 gene and homozygous for the acuH13 mutation showed that the iso- lated suppressor is recessive. In addition, growth tests of the WSol7H and WSol7 strains, both carrying the suaE7 allele (Table 1 ), were performed on CM and 1% glucose MM at diVerent temperatures: 23, 28, 37 and 43°C. Results obtained demonstrated similar phenotypic alterations to those described for the suaA and suaC mutants obtained in Fig.1. Growth tests on CM (A), 1% glucose (B), 0.1M acetate (C) or 6 mM oleate MM (D) of A. nidulans strains carrying the four genotypes ( suaE +;acuH +, suaE7 ; acuH +, suaE +; acuH13 and suaE7 ; acuH13 ) obtained from a sexual cross between A. nidulans R21 wild-type ( suaE +; acuH +) and the revert- ant WSol7H strain ( suaE7 ; acuH13 ). Cultures of MM plates were incubated for 3 days at 37°C, while the CM plate was incubated for 2 days. In the latter conditions, the growth retardation produced by the suaE7 gene is much clearer. suaE+; acuH+suaE7 ; acuH +suaE+; acuH13 suaE7 ; acuH13 CABD

PAGE – 6 ============
144 Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 the past ( Roberts etal., 1979; Martinelli, 1994 ). Such altera- tions include very sparse growth at 23 and 43°C, growth retardation at 28 and 37°C ( Fig.1 ) and increased sensitiv- ity to hygromycin. Representative results of growth tests of a suaE7 strain and available suaA , suaB and suaC mutant strains are shown in Table 3 . In addition, we quanti Wed the viability of suaE7 conidia. The 57% of the conidia formed by the suaE7 mutants germinated on CM, while the 81% of conidia produced by the R153 (wild-type) strain were via- ble. Also, we tested the ability of suaE7 strains to form sel- fed cleistothecia. The suaE7 mutants formed only a few cleistothecia which were surrounded by an unusually thick layer of Hülle cells. These cleistothecia contained only a few viable ascospores. These results demonstrate that, as reported for the majority of the suaA and suaC mutant alleles ( Bratt and Martinelli, 1988 ), the suaE7 mutation reduces conidial viability and fertility. Further genetic characterization of the identi Wed sup- pressor gene involved the use of parasexual genetics. Dip- loids YSol7::MS2.7 and YSol7H::MS2.7 ( Table 1 ) were obtained by using established procedures. Diploids were treated with benomyl (1mg/ml) as the haploidizing agent and phenotypic analyses of haploid segregants allowed unequivocally to locate the suppressor gene at chromosome III, as do the suaA and suaB genes. A sexual cross between strains WSol7H and G351, the latter carrying the suaA101 allele, was established. Phenotypic analysis of the meiotic progeny analysed (over 200 colonies) demonstrated no linkage between both suppressor mutations. Besides, a sec- ond cross was established between strains WSol7H and H9 which carries the suaB111 allele. Phenotypic analysis of the meiotic progeny (over 200 colonies) demonstrated that both suppressor mutations are unlinked. These results unequivocally indicate the identi Wcation of a new transla- tional suppressor gene in A. nidulans , which we have named suaE based on the nomenclature suggested by Roberts etal., 1979 ) for the translational suppressors in this model Wlamentous fungus. 3.3. The suaE7 mutation give rise to the G255S substitution in A. nidulans eRF1 Strain S7HG ( biA1 ; argB2 suaE7 ; pyroA4 ;acuH13 ) was derived from a cross between WSol7H and WG355 ( Table 1), and used as the recipient strain in transformation exper- iments to clone the suaE gene. To Wnd a positive selection method for cloning the gene, several growth conditions based on the phenotype associated to the suaE7 gene were tested. It was observed that strain S7HG is unable to grow in 1% glucose MM plus 200 M hygromycin at 42°C. Thus, these growth conditions were used to clone the suaE gene by transformation. DNA from 20-cosmid pools of the chromosome III-speci Wc library of A. nidulans (Brody etal., 1991 ) was mixed with equimolar amounts of the autono- mously replicating plasmid ARp1 ( Gems etal., 1991 ) and used in co-transformation experiments. Following the strat- egy described elsewhere ( De Lucas etal., 1999 ), a cosmid (L6H06) was identi Wed as capable of restoring all pleiotro- pic alterations of the S7HG strain associated to the suaE7 mutation. In addition, the hygromycin-resistant transfor- mants obtained were unable to grow on acetate and oleate media. All these results clearly suggested the identi Wcation of a positive cosmid that contains the suaE gene and not an unlinked suppressor (see below). DNA from this cosmid was digested with the following enzymes: Bam HI, Eco RI, Hin dIII, Pst I, XbaI and XhoI.Restriction fragments were used in combination with ARp1 DNA to co-transform the strain S7HG. Hygromycin-resis- tant transformants were obtained for all digestion mixtures except when Hin dIII restricts were used. The positive cos- mid was digested with Pst I and restriction fragments were used in transformation. Only transformation with a 15-kb PstI fragment gave hygromycin-resistant colonies. This fragment was cloned into pUC19 to give pPS150. This large vector, that transformed the S7HG strain to suaE + pheno- type at a frequency of 18 transformants per microgram of DNA, was digested with Xba I and restricts obtained were used in transformation. A positive 2.5-kb XbaI fragment was found. This genomic fragment was cloned into pUC19 to give the plasmid pXB25, which complements the suaE7 mutation at a similar transformation frequency than pPS150 does. The two strands of the 2.5-kb XbaI insert of plasmid pXB25 were sequenced. BLASTN and BLASTX searches unequivocally showed it to contain the ORF except for the last 54 codons of the gene encoding the eRF1 of A. nidulans (Han etal., 2005 ). The remaining 3 coding region of the A. nidulans suaE gene was identi Wed by partial sequencing of the vector pPS150. This vector contains the whole suaE gene surrounded by a large genomic region at its 5 and 3 ends. The physical map of the A. nidulans suaE Table 3 Growth tests and sensitivity to hygromycin of A. nidulans R153 (wild-type) strain and suaA , suaB, suaC and suaE mutants Growth on complete medium (CM) is expressed quantitatively from 0 (no growth) to 4 (maximal growth). Data were scored at the fo llowing incubation times: 5 days at 23°C, 4 days at 28°C and 3 days at 37°C and 43°C. Growth tests on 1% glucose MM were similar to those shown on CM.StrainGrowth on CMGrowth on CM+100 M hygromycin 23°C28°C37°C43°C23°C28°C37°C43°C R15344432231 suaA101 12320020 suaB111 44432220 suaC109 01430010 suaE7 12311120

PAGE – 8 ============
146 Ó. Martínez et al. / Fungal Genetics and Biology 44 (2007) 139–151 YEPD (repressing conditions) ( Romero etal., 2003 )showed that eRF1-depleted spores are unable to germinate (Fig.2 ). However, these transformants grew as the wild-type strain when incubated in 0.1M threonine MM (inducing con- ditions). Microscopic analysis showed that in repressing media these conidia swelled until they almost double their size. Nevertheless, these conidia did not show any other ger- mination feature. Consequently, germ tube emergence was never observed. Finally, these spores started to burst after 20h incubation at 37°C. These results demonstrate that the suaE gene is an essential gene in A. nidulans .Molecular characterization of the suaE7 mutation was per- formed by PCR ampli Wcation of the suaE7 allele from the revertant WSol7H strain as stated in Section 2. The 906G>A transition in the suaE ORF was identi Wed. This mutation results in the G265S substitution in the eRF1 of A. nidulans . In addition, we sequenced the suaE gene from several strains obtained in this work that carry the suaE7 allele or the wild- type suaE + gene ( Table 1 ). The results obtained unequivocally correlate the presence of the 906G>A transition at the suaE gene with the pleiotropic growth defects and the suppression of the acuH13 and acuH31 mutations ( Fig.1 ).It has been proposed that the omnipotent eRF1 and aRF1 (archaea class-1 release factor) have originated from an archaeal-like version in the common ancestor of eukay- otes and archaea ( Inagaki and Ford, 2000 ). In contrast, the two bacterial class-1 release factors (RF1 and RF2) do not posses any obvious sequence homology with eRF1 or aRF1 (Nakumura and Ito, 2003). An alignment of eRF1 sequences from representative eukaryotes including the fungi A. nidulans and Saccharomyces cerevisiae , the ciliate Tetrahymena thermophila and the mammal Homo sapiens ,together with the aRF1 se quences from the archaeas Met- hanosarcina mazei and Pyrococcus abysii , is shown in Fig.3 . It can be seen that a signi Wcant level of sequence identity exists among domains 1 (residues 1–141 using the numbering of human eRF1) of eRF1s and aRF1s. The domain 1 characteristic motif NIKS involved in recogni- tion of all stop codons is present in both aRF1s and eRF1s except for that from Tetrahymena thermophila , a ciliate with a nonstandard genetic code that utilizes only UGA as stop ( Lozupone etal., 2001; Kisselev etal., 2003 ). Also, a signi Wcant level of sequence identity is found among domains 2 (residues 145–275 of human eRF1) of eRF1s and aRF1s. As can be seen, the characteristic GGQ tripep- tide in domain 2 that is required for peptidyl-tRNA hydro- lysis is also conserved in both eukaryotic and archaeal RFs. In contrast, a very low level of sequence identity exists among domains 3 (residues 276–437 of human eRF1) of both types of class-1 RFs. Interestingly, the gly- cine 265 residue of domain 2 of the A. nidulans eRF1 (G253 in the human eRF1) is present in all of these sequences, which are representative of both Eukarya and Archaea. The conservation of this glycine in all eukaryotic and archaeal sequences analysed (not shown) clearly sug- gests a relevant role of this amino acid in eRF1 structure and/or function. 3.4. Structural consequences of the G265S substitution in domain 2 of A. nidulans eRF1 The overall architecture of domain 2 of A. nidulans eRF1 is very similar to that of human eRF1, as expected from the high degree of sequence identity ( Figs. 3 and 4 ). Likewise, Gly265, which is positionally equivalent to Gly263 in human eRF1, is located close to the N-terminus of the kinked -helix that connects domains 2 and 3. Our modelling and sim- ulation results for domain 2 of both wild-type and G265S mutant fungal eRF1 suggest that no gross conformational changes take place in this domain as a consequence of this mutation ( Fig.4 ). The evolution of the root-mean-square deviation from the initial structure ( Fig.5 ) shows that most of the Xuctuation is due to the high mobility of the loop embedding the GGQ motif, as expected, whereas the -helix remains very stable in both cases. Nonetheless, we note that the Xexibility of this helix, which has been suggested to par- ticipate in the hinge motions involving both protein domains, is likely to be a Vected by the good hydrogen bonds that the hydroxyl group in the side chain of the mutated Ser265 can establish with the backbone of adjacent protein residues that make up the preceding loop. 4. Discussion A search for suppressors of the CACT de Wciency in the model fungus A. nidulans has suggested the impossibility of isolating physiological suppressor genes for this genetic and metabolic disease, the most severe disorder of fatty acid -oxidation in humans. Nevertheless, this suppressor analysis has allowed the identi Wcation and further genetic and struc- tural characterization of a novel mutation ( suaE7 ) in the omnipotent class-1 release factor of eukaryotes (eRF1) that suppresses two of the three existing amber nonsense muta- tions in the A. nidulans CACT encoding ( acuH ) gene. Translation termination in eukaryotes is controlled by the release factor complex, a heterodimer formed by associ- ation of eRF1 and eRF3. This complex recognizes a stop codon located in the ribosomal A-site and triggers the release of the nascent peptide (see Inge-Vechtomov etal., 2003 , for a review). Genetic and biochemical studies in S.cerevisiae, in which the essential SUP45 and SUP35 genes encode, respectively, the eRF1 and eRF3 proteins, have demonstrated that eRF1 acts as the key factor in the termi- nation of translation while the eRF3 has a stimulating role (Bertram etal., 2001; Inge-Vechtomov etal., 2003 ).Resolution of the three-dimensional structure of human eRF1 by X-ray crystallography has revealed the molecular mechanism of eRF1 activity ( Song etal., 2000 ). The overall shape and dimensions of eRF1 resemble those of a tRNA molecule with domains 1, 2, and 3 of eRF1 corresponding to the anticodon loop, aminoacyl acceptor stem and T stem of a tRNA molecule, respectively ( Fig.3 ). Domain 1, con- taining the highly conserved motif NIKS, is proposed to participate in stop codon recognition ( Song etal., 2000; Bertram etal., 2000; Chavatte etal., 2001; Frolova etal.,

141 KB – 13 Pages