Identification of an S. macrospora Atg12 homolog as an interaction partner of SmATG7 and SmATG3
A BLASTP search of predicted S. macrospora proteins  using S. cerevisiae ATG12 (P38316) as a query sequence identified a predicted protein of 215 amino acids encoded by the ORF SMAC_06998 as the top hit (3e-13). Compared with ATG12 proteins from other organisms, the predicted protein SMAC_06998 has an extended C-terminal region but no glycine residue at the C-terminus (Fig A in S1 File). Automatic annotation predicted two introns in SMAC_06998 (Fig A in S1 File); however RT-PCR amplification with primer pair 06998_f/06998_r and sequencing confirmed only splicing of the first intron, which leads to an earlier stop codon. The amplified cDNA encodes a protein of 159 amino acids with a conserved C-terminal glycine that shares a significant level of sequence similarity with ATG12 proteins of plants, animals that also have a conserved glycine at the C-terminus (Fig 1). The S. macrospora ATG12 homlog shares 81% and 50% identity to the previously described ATG12 proteins from Neurospora crassa and Magnaporthe oryzae [12,13], and only about 20% identity with homologs from S. cerevisiae, plants and animals (Fig 1). However, most of the residues involved in non-covalent interactions between ATG12 and ATG5, binding and conjugation of ATG3, or binding of ATG8 in the S. cerevisiae and the human ATG12 homologs are conserved in the S. macrospora ATG12 (Fig 1) [24,27,64–66]. Several residues within the BH3 domain of ATG12 homologs from mammals and non-mammalian vertebrates are also conserved. Similar to the S. cerevisiae and other fungal ATG12 homologs, a conserved aspartic acid residue (equivalent to D64 in human ATG12) that is required for binding of anti-apoptotic proteins is replaced by a serine residue at position 82 in the S. macrospora protein (Fig 1) .
Fig 1. Multiple sequence alignment of ATG12 orthologs from fungi, plants and animals.
ClustalX alignment was created using the following sequences: Smac [S. macrospora, Accession No. XP_003349162.1, excluding 56 C-terminal amino acids)], Ncra [Neurospora crassa, Q7S083.1], Pans [Podospora anserina, XP_001906089.1], Cglo [Chaetomium globosum, Q2GSG9.2], Mory [Magnaporthe oryzae, XP_368646.1], Anid [Aspergillus nidulans, Q5BCH0.2], Pchr [Penicillium chrysogenum, XP_002557636.1], Celg [Caenorhabditis elegans, CCD61524.1], Hsap [Homo sapiens, NP_004698.3], Dmel [Drosophila melanogaster, NP_648551.3], AthaB [Arabidopsis thaliana, Q9LVK3.1], AthaA [A. thaliana, Q8S924.1], Scer [S. cerevisiae, P38316]. Identical amino acids, which are conserved in all proteins, are shaded in black; residues conserved in at least 10 of 13 sequences are shaded in dark grey and residues conserved in at least eight sequences are shaded in light grey. The conserved C-terminal glycine residue for the covalent linkage to ATG5 and the conserved phenylalanine residue corresponding to Phe154 in the S. cerevisiae Atg12 is labelled in red , amino acids important for non-covalent interactions between ATG12 and ATG5 in S. cerevisiae according to Noda et al.  are marked by asterisks. Non-covalent contacts between ATG12 and ATG5 identified in the human homologs according to Otomo et al.  are marked by black squares. The red bar represents the turn—loop—alpha helix 2 segment (Asn105 –Phe123 of the human ATG12) which is associated with the interaction surface of ATG5 . White squares mark residues of the human ATG12, which are important for binding of ATG3 , #, indicates residues of the non-canonical AIM of ATG12 involved in interaction with ATG8 . The region of the BH3 domain identified in the human ATG12 homolog is indicated and the conserved aspartic acid residue is indicated in green. Amino-acid identity in % is given at the right margin.
In S. cerevisiae, direct protein-protein interactions between the UBL protein Atg12 and both E1/E2-like enzymes Atg7 and Atg3 have been reported [64,66,68]. To confirm that SmATG12 can interact with SmATG7 and SmATG3, yeast two-hybrid analysis was performed after cloning of Smatg12, Smatg7  and Smatg3 (S. macrospora ORF SMAC_05399) into prey and bait vectors pGBKT7 and pGADT7.
The resulting bait and prey plasmids were transformed into yeast strains Y187 and AH109, respectively, which were then mated, or both plasmids were co-transformed into strain AH109. Transactivation of pBD-derivatives was tested by mating with the AH109 strain carrying the empty vector pGADT7 (data not shown). As a positive control and to confirm expression of the proteins encoded by the bait plasmids, strains were mated with yeast strain AH109 containing pAD-ranBPM. RanBPM directly interacts with the GAL4-binding domain, which provides a method for confirming that the gene cloned into the bait vector is expressed appropriately . The two-hybrid experiment clearly demonstrated an interaction between SmATG12 and both SmATG3 and SmATG7 (Fig 2A and 2B). SmATG12 and SmATG7 interacted only when SmATG7 was expressed as a GAL4-BD fusion protein (Fig 2A). Recently, Kaufmann et al.  demonstrated that the yeast ATG8-PE can interact directly with ATG12 via a non-canonical AIM in ATG12. However, interaction of SmATG12 and SmATG8 were not demonstrated in the yeast two-hybrid system (Fig 2C), although expression of SmATG8 and SmATG12 in yeast was verified by Western blot analysis (Fig B in S1 File).
Fig 2. Yeast-two hybrid interaction of SmATG12 with SmATG7, SmATG8 and SmATG3.
Full-length cDNAs of Smatg12, Smatg7 and Smatg3 were used to generate GAL4-DNA binding domain (BD) and activation domain (AD) plasmids. Smatg8 two-hybrid vectors were previously described in Voigt and Pöggeler . To select for the presence of both plasmids 20 μl of cells were spotted in serial delutions on SD medium lacking tryptophan and leucine (SD -Trp, -Leu) or to verify the interactions of the proteins on medium lacking additionally histidine or adenine (SD -Trp, -Leu, -His/-Ade). (A) SmATG12 and SmATG7 interacted only when SmATG7 was expressed as GAL4-BD fusion protein. (B) SmATG12 and SmATG3 interacted with each other as bait and prey proteins, respectively. (C) SmATG8 did not interact with SmATG12. Transformants carrying a bait plasmid and pAD-ranBPM  were used to confirm the appropriate expression of the bait proteins. Strains carrying empty plasmids pGADT7 and pGBKT7 served as negative control (A). Plates were incubated for 3–5 days at 30°C.
Furthermore, we tested the functional conservation of SmATG12 and its yeast counterpart by a yeast complementation assay with the Smatg12 cDNA expressed under the control of a MET25 promoter in an S. cerevisiae atg12Δ null mutant. Rescue of autophagy in the S. cerevisiae mutant was monitored using an aminopeptidase I (API) maturation assay based on the autophagy-dependent maturation of the precursor proaminopetidase I (prAPI) to the mature enzyme (mAPI) . In S. cerevisiae, API maturation relies on an intact UBL protein Atg12. However, the Smatg12 gene was unable to complement the yeast atg12Δ mutant (Fig C in S1 File). In S. cerevisiae, API is delivered by the cytoplasm-to-vacuole-targeting pathway to the vacuole. So far it is not clear if this pathway exits in S. macrospora. In addition, we therefore monitored autophagy in the complemented mutant by the GFP-Atg8 proteolysis assay. When GFP-Atg8 is delivered to the lumen of the vacuole, the Atg8 part of the fusion protein is sensitive to degradation, whereas the GFP moiety is relative resistant to hydrolysis. In a Western blot with an anti-GFP antibody the appearance of free GFP can be used to monitor delivery of autophagosomal membranes to the vacuole . The assay revealed that also no processing of GFP-Atg8 occurred in the atg12Δ yeast strain complemented with the Smatg12 gene (Fig D in S1 File).
Deletion of Smatg12 impaired vegetative growth and fruiting-body development
To analyze the effect of Smatg12 on autophagy, vegetative growth and sexual development of S. macrospora, we generated a ΔSmatg12 knockout mutant for phenotypic characterization (Fig E in S1 File). The EGFP-ATG8 proteolysis assay in the S. macrospora ΔSmatg12 mutant revealed that vacuolar degradation of the fusion protein is inhibited, whereas in the complemented ΔSmatg8 strain proper vacuolar proteolysis of the fusion protein can be observed (Fig 3). Although Smatg12 cannot complement the S. cerevisiae atg12Δ strain, this result suggests, that the isolated gene is an ortholog of the yeast ATG12.
Fig 3. EGFP-SmATG8 protein degradation in the ΔSmatg12 strain compared to the corresponding complemented ΔSmatg8::egfp-Smatg8ect strain.
Protein crude extracts of S. macrospora wt, ΔSmatg12::egfp-Smatg8ect and ΔSmatg8::egfp-Smatg8ect strains expressing EGFP or EGFP-SmATG8 were separated on a 12% SDS-PAGE gel. The Western blot hybridization using an anti-EGFP antibody verified the degradation of the EGFP-SmATG8 fusion protein in the complemented ΔSmatg8 strain by accumulation of free EGFP, whereas in the ΔSmatg12 mutant the EGFP-SmATG8 fusion protein accumulated.
In contrast to the wild-type (wt) strain and a complemented mutant strain (ΔSmatg12::egfp-Smatg12ect), the deletion mutant ΔSmatg12 did not form mature fruiting bodies when grown on solid SWG fructification medium or under histidine starvation conditions imposed using the drug 3-aminotriazole (3-AT) (Fig 4A). Prevention of autophagy is known to lead to the impairment of the foraging capability of filamentous ascomycetes [34,42], which describes the growth of filamentous fungi over a non-nutritious surface in order to reach nutrient-rich regions. The required nutrients are thought to be provided by autophagy taking place in the basal hyphae of mycelia [12,71]. To analyze the foraging capability of ΔSmatg12, wt and the complemented strain (ΔSmatg12::egfp-Smatg12ect), agar plugs were transferred into an empty cell-culture plate and incubated for 5 days. While the wt and the complemented strain underwent extensive mycelial growth, ΔSmatg12 was unable to grow over the inert plastic surface (Fig 4B). In addition, the deletion mutant displayed only a significant decrease in growth rate under histidine starvation conditions imposed by 3-AT (Fig 4C). In the complemented strain the growth defect was only partially complemented in comparison to the wt, probably due to an ectopic integration of the egfp-tagged Smatg12. Similarly, a slight growth defect of the complemented mutant could be observed under non-starvation conditions (SWG fructification medium), which might be also explained by ectopic integration of the egfp-tagged Smatg12 wt copy.
Fig 4. Phenotypic characterization of S. macrospora wt, Smatg12 deletion and complementation strain.
(A) Phenotype of wt, ΔSmatg12 and rescued strain ΔSmatg12::egfp-Smatg12ect grown on SWG or SWG medium supplemented with 2.5 mM 3-AT in petri dishes. Insets show a detailed view of perithecia or sterile mycelium. The images were taken seven days post inoculation. (B) Foraging capacity of indicated strains. Agar plugs of 0.5-cm diameter were transferred into empty cell-culture plates (6 well, 17.2 ml) and incubated for 5 d at 27°C in a damp chamber before photographed. Scale bars as depicted. (C) The growth velocity of wt, ΔSmatg12 and the complemented strain was analyzed by measuring the growth velocity in cm/day in 30-cm race tubes. Growth rates on SWG medium shown are averages from 7 independent measurements of three independent experiments (n = 21), standard deviations are indicated by error bars. Asterisks indicate significant differences according to Student´s t-test (p<0.0000001). (D) Microscopic investigation of sexual development of ΔSmatg12 compared to wt and the complemented strain. Strains were grown on SWG medium. Expression of the EGFP-SmATG12 fusion construct complemented the sterile phenotype of ΔSmatg12. The wt and the complemented strain ΔSmatg12::egfp-Smatg12ect form ascogonia at day 3, and protoperithecia at day 4 post inoculation. These develop to pigmented protoperithecia at day 5 and to mature perithecia at day 7. Sexual development of the mutant ΔSmatg12 is blocked at the stage of protoperithecia formation. ΔSmatg12 neither forms pigmented protoperithecia nor perithecia and ascospores. Scale bars as depicted.
Microscopic investigation of the fruiting-body development revealed that the mutant was arrested at the protoperithecia developmental stage and was unable to form mature fruiting bodies or ascospores (Fig 4A and 4D). Altogether, we analyzed five independent homokaryotic deletion mutants and two independent complemented mutants which displayed homogenous phenotypes. Thus, the observed phenotypes are associated with the deletion of the Smatg12 gene.
EGFP-SmATG12 localizes to phagophore assembly sites
To investigate the in vivo localization of SmATG12, the deletion mutant ΔSmatg12 was transformed with plasmid pegfp-Smatg12. The ectopically integrated egfp-Smatg12 fusion construct, under the control of the endogenous promoter, was able to complement the sterile phenotype of ΔSmatg12 (Fig 4). Fluorescence microscopy of the functionally expressed EGFP-SmATG12 fusion protein in the proliferating mycelium (24 h after inoculation) revealed localization in the cytoplasm as small discrete dots or cup-shaped structures that were presumed to be phagophore assembly sites or growing phagophores (Fig 5A). In the ΔSmatg8 mutant, the EGFP-SmATG12 fluorescence signals appeared larger and more distinct than that of the complemented ΔSmatg12 mutant (Fig 5B). To investigate the role of SmATG12 in autophagy, we examined the autophagy-mediated engulfment of EGFP-labeled SmATG8 into the vacuole. Transformation of pRS-egfp-Smatg8  into ΔSmatg12 allowed us to follow the fluorescence signal of EGFP-labeled autophagosomes. In contrast to the complemented ΔSmatg8 mutant, in the ΔSmatg12 mutant EGFP-SmATG8 fluorescence was diffused throughout the cytoplasm, and large cytoplasmic aggregates were excluded from vacuoles, even in basal hyphae (Fig 5D). Vacuolar membranes were stained with the mebrane dys FM4-64. The localization did not change when we induced histidine starvation by 3-AT  or nitrogen limiting conditions. Fluorescence microscopy revealed the expected localization of EGFP-SmATG8 in vacuoles and in autophagosome-like structures (distinct spots in the cytoplasm) as well as in basal hyphal compartments in the lumen of the vacuoles in the complemented ΔSmatg8 (Fig 5E) . These results indicated that macroautophagy was disrupted in the ΔSmatg12 mutant.
Fig 5. Fluorescence microscopic localization of EGFP-SmATG12 and EGFP-SmATG8.
(A) Mutant ΔSmatg12 was transformed with plasmid pegfp-Smatg12. Expression of the EGFP-SmATG12 fusion construct complemented the sterile phenotype of ΔSmatg12. EGFP-SmATG12 localizes to the cytoplasm and at phagophore assembly sites indicated by small arrows and cup-shaped phagophores indicated by long arrows. Vacuolar membranes were stained using FM4-64. The merged picture shows a close up. (B) Localization of EGFP-SmATG12 in the ΔSmatg8 mutant . The fluorescence signals are larger and more distinct than signals in the complemented mutant ΔSmatg8::egfp-Smatg8ect (compare to E). (C) The fluorescence signal of free EGFP in the wt strain transformed with plasmid p1783-1 served as control. (D) When the mutant strain ΔSmatg12 expressed EGFP-SmATG8 (plasmid pRS-egfp-Smatg8 ), the fluorescence protein displays an equal diffused signal in the cytoplasm with large accumulating spots (arrow head) which are excluded from the vacuole. Vacuolar membranes were co-stained with FM4-64 and pictures were merged. (E) The previously constructed strain ΔSmatg8::egfp-Smatg8ect  was used to compare the localization of EGFP-SmATG8 in the complemented ΔSmatg8 mutant. Autophagosomes are indicated by arrows. DIC, differential interference contrast; EGFP, enhanced green fluorescence protein. Autophagy was induced by the addition of 2.5 mM 3-AT to the SWG medium or by nitrogen starvation conditions (SWG-N, without KNO3 and arginine). Scale bars as depicted.
Alicia Knudson and I just returned from the 1st International Symposium on Fungal Multicellular Development, which was held on Sunday, April 3, in Paris. Twenty-one participants discussed recent advances in fungal developmental biology and evolution, drawing on molecular biology, anatomy, cultural studies, comparative genomics, transcriptomics, and phylogenetics. The workshop was organized by László Nagy of the Biological Research Centre, Szeged, Hungary (with nominal assistance from me), as a satellite meeting of the 13th European Congress of Fungal Genetics. Several talks illustrated the tremendous progress that hars been made in the well-established ascomycete model systems Sordaria macrospora and Neurospora crassa and the basidiomycetes Coprinopsis cinerea and Schizophyllum commune. There were also presentations on emerging basidiomycete systems, including Agrocybe aegerita and Lentinus tigrinus and taxa that cannot (yet) be induced to produce fruiting bodies in culture, such as the ectomycorrhizal gasteromycete Pisolithusmicrocarpus and the enigmatic Neolecta irregularis, the latter representing the only genus in Taphrinomycotina that produces a multicellular fruiting body (although Taphrina produces a hymenium…). The workshop was held in a school belonging to the École Militaire. Our group photo was taken on the playground, which featured murals of Noah’s ark and Little Red Riding Hood, along with a very threatening wolf and some anthropomorphized mushrooms.
L>R: David Hibbett, Maira d. F. Pereira, Robert Herzog, Annegret Kohler, Florian Hennicke, Minou Nowrousian, Jeff Townsend, Stefanie Pöggeler, Jason Stajich, Robin A. Ohm, Cissé Ousmane, Ursula Kües, Weeradej Khonsuntia, Elke-Martina Jung, Kathryn Ford, Torda Varga, Krisztina Krizsán, Alicia Knudson, Éva Almási, László G. Nagy; not present: Pascale Marie-Aimée Dozolme.
9:00 – 9:10 Welcome & Opening
9:10 – 9:40 Stefanie Pöggeler.Functions of autophagy-related genes in fruiting-body development of Sordaria macrospora
9:40 – 10:10 R. Herzog, I. M. Solovyeva, D. K. Gupta, R. Sharma, M. Rühl, M. Thines, Florian Hennicke. Making the poplar mushroom popular again – Agrocybe aegerita as a model basidiomycete
10:10 – 10:40 Jeffrey Townsend. Developmental origins of fungal fruiting body phenotypes in the evolution of the transcriptome
10:40 – 11:00 Coffee break
11:00 – 11:30 Ursula Kües, S. Subba, Y. Yu, M. Sen, W. Khonsuntia, W. Singhadaung, K. Lange, O. Voigt, K. Lakkireddi. Regulation of fruiting body development in Coprinopsis cinerea
11:30 – 12:00 Cissé Ousmane, A. Nguyen, D. Hewitt, M. Nowrousian, G. Jedd, J. Stajich. Two origins of complex multicellular organization in the Ascomycota
12:00 – 12:30. Alicia Knudson, S. Kovaka, J. Gibbons, L.G. Nagy, D.S. Hibbett. Fruiting body plasticity in Lentinus tigrinus and light signaling pathways in Agaricomycetes
12:30 – 14:00. Lunch
14:00 – 14:30 Elke-Martina Jung, K. Katrin, K. Nicole, E. Kothe. Aspects in intracellular regulation and microbial interactions of Schizophyllum commune
14:30 – 15:00. S. Traeger, D. Schumacher, S. Gesing, F. Altegoer, Minou Nowrousian. Comparative genomics and transcriptomics to study fruiting body development in ascomycetes
15:00 – 15:30 Annegret Kohler, M. d. F. Pereira, A. N. da Rocha Campos, T. C. Anastácio, M. D. Costa, F. Martin. Insights into the basidiocarp development of the ectomycorrhizal fungus Pisolithus microcarpus
15:30 – 16:00 Coffee break
16:00 – 16:30 Pascale Marie-Aimée Dozolme, B. Batailler, S.M. Moukha. Leveraging Idiophasic metabolism shift localization at mycelial and hyphal levels: case of mycotoxin biosynthesis and secretion
16:30 – 17:00 I. Marian, J. Pelkmans, M. Schuller, H. M. Brewer, A. Carver, A. Copeland, J. Grimwood, H. Lee, E. Lindquist, A. Lipzen, J. Martin, H. Park, S. O. Purvine, W. Schackwitz, M. Tegelaar, A. Tritt, S. Baker, I.-G. Choi, L. G. Lugones, H. A. B. Wösten, I. V. Grigoriev, Robin A. Ohm.Leveraging the diversity in the hypervariable species Schizophyllum commune to understand mushroom development and lignocellulose degradation
17:00 – 17:30 László G. Nagy. An overview of the evolution of complex multicellularity in Fungi: synthesys and future challenges.
17:00 – 17:50 Discussion and closing remarks