Dynamin-like protein-dependent formation of Woronin bodies in Saccharomyces cerevisiae upon heterologous expression of a single protein Christian Wu ¨ rtz, Wolfgang Schliebs, Ralf Erdmann and Hanspeter Rottensteiner Institut fu ¨ r Physiologische Chemie, Ruhr-Universita ¨ t Bochum, Germany The HEX1 protein of Neurospora crassa, identified by Jedd and Chua [1] and Tenney et al. [2], is the major component of a class of microbodies limited to euasco- mycetes and some deuteromycetes, the so-called Woro- nin body [3,4]. Because of the syncytial growth of filamentous fungi, wounding of hyphae can lead to a severe loss of cytoplasm and subcellular organelles, if the plasma membrane or a nearby septum is not rap- idly sealed. For this reason, the Woronin body is pres- ent in filamentous euascomycetes and plugs septal pores immediately after cells have been damaged [1,2]. In addition to septal pore sealing in cases of injury, Woronin bodies have also been described as being required for efficient pathogenesis, survival during nitrogen starvation [5] and conidiation [6] in various fungi. Although it is more than 140 years since the dis- covery of this very specialized organelle [4], our knowl- edge of the biogenesis of Woronin bodies remains incomplete. Electron microscopy studies provided the first evidence that Woronin bodies are derived from other microbodies [7]. These findings have been extended by reports showing that Woronin body formation is initiated in the vicinity of glyoxysomes and may proceed through fission from them [8], and by the demonstration that PEX14 is a key player in the biogenesis of both glyoxysomes and Woronin bodies [9]. Furthermore, the presence of a C-terminal canonical peroxisomal targeting signal type 1 (PTS1) is required for the proper topogenesis of HEX1 [9] and allows HEX1 to be imported into peroxisomes upon heterologous expression in yeast [1]. Peroxisomal Keywords filamentous fungi; Neurospora crassa; peroxisome; protein import; yeast Correspondence H. Rottensteiner, Institut fu ¨ r Physiologische Chemie, Abt. Systembiochemie, Ruhr- Universita ¨ t Bochum, D-44780 Bochum, Germany Fax: +49 234 321 4266 Tel: +49 234 322 7046 E-mail: hanspeter.rottensteiner@rub.de (Received 22 January 2008, revised 27 March 2008, accepted 2 April 2008) doi:10.1111/j.1742-4658.2008.06430.x Filamentous ascomycetes harbor Woronin bodies and glyoxysomes, two types of microbodies, within one cell at the same time. The dominant pro- tein of the Neurospora crassa Woronin body, HEX1, forms a hexagonal core crystal via oligomerization and evidence has accumulated that Woro- nin bodies bud off from glyoxysomes. We analyzed whether HEX1 is suffi- cient to induce Woronin body formation upon heterologous expression in Saccharomyces cerevisiae, an organism devoid of this specialized organelle. In wild-type strain BY4742, initial import of HEX1 into existing peroxi- somes enabled the formation of organelles with a hexagonal crystal. The observed structures mimicked the shape of genuine Woronin bodies, but exhibited a lower density and were significantly larger. Double-immuno- fluorescence analysis revealed that hexagonal HEX1 structures only occa- sionally co-localized with peroxisomal marker proteins, indicating that the Woronin-body-like structures are well separated from peroxisomes. In cells lacking Vps1p and Dnm1p, dynamin-like proteins required for the division of peroxisomes, the Woronin-body-like organelles remained attached to peroxisomes. The data indicate that Woronin bodies emerge after the formation of a HEX1 core crystal within peroxisomes followed by Vps1p- and Dnm1p-mediated fission. Abbreviations PMP, peroxisomal membrane protein; PNS, post-nuclear supernatant; PTS1, peroxisomal targeting signal type 1. 2932 FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS HEX1 provokes the formation of very small, mem- brane-bound protein granules that are hexagonal or spherical [1]. Although intriguing, this observation points to the existence of additional factors in filamen- tous fungi that contribute to the formation of mature and functional Woronin bodies. Although the transport of HEX1 to Woronin bodies via microbodies, i.e. peroxisomes, is well known, the budding of Woronin bodies from microbodies is still under investigation. Some players in peroxisomal fis- sion have been identified in recent years, including the dynamin-like proteins Vps1p [10] and Dnm1p [11]. The reduction in peroxisome numbers seen in vps1D and dnm1D single mutants is even more pronounced in the absence of both proteins. Most cells possess only a single peroxisome that is extended and exhibits a num- ber of constrictions [11]. The Pex11p family of proteins is also implicated in peroxisome fission, but is thought to act upstream of the dynamin-like proteins [12–17]. This study is concerned with the topogenesis of HEX1 using the heterologous Saccharomyces cerevisiae expression model system. We used diverse cell biologi- cal and biochemical approaches to scrutinize various HEX1-expressing strains for the appearance of Woro- nin-body-like structures. Furthermore, we determined whether the formation of Woronin bodies from peroxi- somes shares components of the peroxisomal fission machinery by examining the fate of HEX1 in a vps1Ddnm1D double-deletion strain. The results are dis- cussed in terms of a mechanism for the formation of Woronin bodies that is largely determined by the expression of HEX1. The system set-up is likely to be of use for studying peroxisome fission in a time- resolved manner. Results HEX1 is imported into organelles in S. cerevisiae in a PTS1-dependent manner To address whether Woronin body formation depends on factors specific for filamentous euascomycetes, we heterologously expressed Neurospora crassa HEX1 cDNA under the control of the constitutive PGK1 pro- moter in S. cerevisiae strain BY4742. Expression of HEX1 in BY4742, achieved by the strong constitutive PGK1 promoter, was verified by western blotting (Fig. 1A). In line with Jedd and Chua [1], the size of heterologously expressed HEX1 was the same as that of endogenous N. crassa HEX1, indicating that HEX1 is correctly synthesized in this yeast strain. Differential centrifugation confirmed the organellar localization of HEX1; most HEX1 was found in the organellar pellet fraction together with the peroxisomal marker proteins Pex13p and Cta1p, whereas only a small amount of HEX1 was located in the cytosolic supernatant (Fig. 1A). HEX1 was also expressed in a pex5D strain, in which PTS1 import is specifically com- promised because of the absence of the cognate signal receptor. In this strain, both HEX1 and the peroxi- somal matrix protein Cta1p were exclusively detected in the cytosolic fraction, whereas the PMP Pex13p was still located in the organellar pellet (Fig. 1B). There- fore, HEX1 is imported into organelles in a PTS1- dependent manner. Localization of HEX1 in density gradients The subcellular distribution of HEX1 was further ana- lyzed by sucrose density gradient centrifugation. Post- nuclear supernatant (PNS) obtained from the wild-type strain BY4742 was subjected to centrifugation at 100 000 g and the resulting pellet was loaded on top of a sucrose density gradient. As expected, the peroxi- somal membrane marker Pex13p peaked at a density of  1.20 gÆcm )3 and was clearly separated from the mitochondrial marker Aac2p (1.18 gÆcm )3 ). The perox- isomal matrix protein Fox3p exhibited a dual distribu- tion, with one peak corresponding to that of Pex13p and an additional peak at light fractions (Fig. 1C). The latter was likely to be caused by ruptured organ- elles that emerge upon resuspension of the pellet. This distribution pattern was not altered when a PNS from BY4742 expressing HEX1 was used. The majority of HEX1 showed a localization that differed from that of Fox3p (Fig. 1D), with one peak at a density of 1.23 gÆcm )3 and one peak at lighter fractions. The dis- tribution of HEX1 was also distinct from that of mitochondrial Aac2p, the Golgi ⁄ endosome marker Pep12p and the endoplasmic reticulum marker Kar2p. When compared with the distribution profile of HEX1 in a density gradient of a N. crassa wild-type strain (Fig. 1E), it became obvious that in yeast HEX1 did not sediment to the density of N. crassa Woronin bodies (1.28 gÆcm )3 ; fraction 5). Thus, in yeast, HEX1 appeared to form Woronin-body-like organelles, but with a lower density than genuine Woronin bodies. PTS1-dependent formation of giant Woronin bodies The subcellular localization of HEX1 was also analyzed by immunofluorescence microscopy. Decoration of the untransformed wild-type strain with anti-HEX1 serum only led to background staining (Fig. 2A). Analysis of the HEX1-expressing strain revealed some small C. Wu ¨ rtz et al. Heterologous Woronin body formation FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 2933 circular spots typical for peroxisomes. Strikingly, large hexagonal Woronin-body-like structures with a mean size of  1.5 · 1.5 lm were also detected upon HEX1 expression (Fig. 2B). Remarkably, typical Woronin bodies of N. crassa are smaller with an average size of  400–700 nm [18]. In a pex5D strain, staining with anti-HEX1 serum was diffuse (Fig. 2C), thereby verify- ing the PTS1-dependent import of HEX1 in yeast. Notably, formation of the hexagonal structures likewise depended on the peroxisomal import receptor Pex5p. To examine whether peroxisomal proteins co-localize with the hexagonal structures, double-immunofluores- cence staining was carried out in strain BY4742 co-expressing the artificial peroxisomal marker protein GFP-SKL and HEX1. In a few cells, GFP-SKL showed a rim-like staining around the hexagonal HEX1 struc- tures (Fig. 2D, upper). In other cases, GFP-SKL was located in small dots in vicinity to the hexagonal struc- tures stained with anti-HEX1 serum. However, most hexagonal structures did not contain GFP-SKL, whereas small spots were often double-labeled for HEX1 and GFP-SKL (Fig. 2D, middle and lower). Ultrastructure of the hexagonal structures To corroborate the appearance of hexagonal HEX1 structures in BY4742, this strain was also analyzed by electron microscopy. The untransformed wild-type showed a normal distribution and morphology for all visible organelles (Fig. 3A). In the HEX1-expressing strain, however, large electron-opaque hexagonal (Fig. 3B) or rectangular (Fig. 3C) structures were visible. These clearly harbored a delimiting single A B C D E Fig. 1. Subcellular distribution of HEX1 upon heterologous expres- sion in S. cerevisiae. (A, B) Differential centrifugation. PNS of (A) BY4742 and BY4742 expressing HEX1, and of (B) the otherwise isogenic pex5D gene deletion strain, with or without expressing HEX1, were separated by centrifugation at 25 000 g for 20 min into a supernatant and an organellar pellet fraction. Equal amounts of each fraction were loaded onto an SDS gel and subjected to western blot analysis. Distribution of the peroxisomal matrix protein catalase (Cta1p), the PMP Pex13p and HEX1 was determined with appropriate antibodies. (C, D) Density gradient centrifugation. The 25 000 g organellar pellets of (C) BY4742 and (D) BY4742 express- ing HEX1 were loaded on top of a linear sucrose gradient (30–60% w ⁄ w) and subjected to centrifugation at 38 000 g for 2 h. Fractions (1 mL) were collected from the bottom (fraction 1) to the top (frac- tion 27) and assayed by western blot for the distribution of the per- oxisomal matrix protein Fox3p, PMP Pex13p and HEX1. Aac2p served as a marker for mitochondria, Kar2p for the ER and Pep12p for the Golgi ⁄ endosome compartment. Densities of the peak frac- tions of HEX1-containing organelles (1.23 gÆcm )3 ) and peroxisomes (1.20 gÆcm )3 ) are indicated. (E) Density of Woronin bodies in N. crassa. For comparison, a 25 000 g organellar pellet from a N. crassa wild-type strain was separated on a 30–60% w ⁄ w sucrose gradient and analyzed for the distribution of HEX1 (Woro- nin bodies), glyoxysomal ICL1 and mitochondrial TIM23. Densities of the peak fractions of Woronin bodies (1.28 gÆcm )3 ) and glyoxy- somes (1.20 gÆcm )3 ) are indicated. Heterologous Woronin body formation C. Wu ¨ rtz et al. 2934 FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS membrane and were therefore designated as giant Woronin bodies. The size of the hexagonal structures fitted with the measurements based on the immunoflu- orescence images. The rectangular structures were up to 1.5 lm on their short side and up to 7.4 lmon their long side, and may represent Woronin bodies in a different orientation with respect to the plane of the section. The short dimension fits with the measurement of the en face view of the hexagonal structures. In some rare cases, peroxisomes with areas of distinct electron density were denoted (Fig. 3D), probably representing HEX1-enriched regions that eventually bud off from the peroxisome. HEX1 assembles to a crystalline core in S. cerevisiae Because the observed Woronin-body-like hexagonal structures differed from N. crassa Woronin bodies in density (Fig. 1) and size (Figs 2 and 3), the question arose as to whether HEX1 is able to form dense crys- tals in S. cerevisiae. It has been shown previously that Woronin bodies sediment upon medium speed centrifugation even if the membrane is removed using detergent, but this requires proper formation of the HEX1 core crystal [19]. To this end, differential centri- fugation at various speeds was conducted, in the presence or absence of detergent. Examination of a wild-type PNS revealed increasing amounts of peroxi- somal marker proteins in the pellet fractions upon increasing centrifugation speed (Fig. 4). Disintegration of the organellar membranes by 0.5% Triton X-100 prevented the marker proteins from being sedimented except for trace amounts in the 15 000 g pellet. A simi- lar distribution of Cta1p and Pex13p was seen for the strain expressing HEX1. By contrast, HEX1 was pres- ent in the 1000 g pellet fraction and, more importantly, HEX1 was detected in the sediment even after A D B C Fig. 2. Pex5p-dependent appearance of giant Woronin bodies upon HEX1 expres- sion. Yeast strains BY4742 (A), BY4742 + HEX1 (B) and BY4742pex5D + HEX1 (C) were analyzed for the localization of HEX1 by indirect immunofluorescence, using anti-HEX1 serum in combination with Alexa Fluor 594-labeled anti-rabbit IgG. In wild-type cells, large hexagonal structures resembling Woronin bodies were detected upon expression of HEX1. (D) Double-immu- nofluorescence microscopy of BY4742 co-expressing HEX1 and the artificial peroxi- somal marker GFP-SKL. The three panels illustrate representative morphological differ- ences of HEX1-stained organelles. Detection was achieved with mouse monoclonal anti- bodies against GFP combined with rabbit anti-HEX1 serum. The secondary antibodies used were Alexa Fluor 488-labeled anti- mouse IgG and Alexa Fluor 594-labeled anti-rabbit IgG. Bar = 5 lm. C. Wu ¨ rtz et al. Heterologous Woronin body formation FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 2935 Triton X-100 treatment, whereas the majority of Pex13p and Cta1p were detected in the supernatant fractions. These data indicated that a typical HEX1 crystal core was formed in yeast which is likely to contain only minor inclusions of peroxisomal matrix proteins. Vps1p and Dnm1p: two dynamin-like proteins involved in fission of peroxisomes and Woronin bodies So far, we have been able to show that large Woronin- body-like structures are formed in S. cerevisiae upon heterologous expression of N. crassa HEX1. Because this is supposed to require budding from peroxisomes, we analyzed whether the typical peroxisomal fission machinery is also involved in the formation of Woronin bodies. Key players in peroxisomal fission are the dynamin-related proteins Vps1p and Dnm1p [10,11] whose concomitant absence typically results in the presence of just one giant peroxisome per cell [11]. Thus, if Vps1p and Dnm1p are also required for the formation of Woronin bodies, HEX1-containing microbodies should be caught in the act of separating from peroxisomes in this mutant strain. The effect of the vps1 dnm1 double-deletion on the subcellular distribution of HEX1 was first analyzed by differential centrifugation analysis. HEX1, as well as Pex13p and Cta1p, were detected in the pellet fraction (Fig. 5A), although trace amounts of Cta1p also A B C D Fig. 3. Ultrastructure of giant Woronin-body-like organelles and peroxisomes in BY4742. Cells were grown for 14 h on medium containing oleic acid as the sole carbon source and processed for electron microscopy. (A) Typical morphology of a wild-type cell. (B–D) BY4742 cells expressing HEX1. (B) A Woronin-body-like structure is viewed from the top, with the typical hexagonal shape of a N. crassa Woronin body. (C) A giant rectangular Woronin body is captured from the side. (D) A small Woronin body is still attached to a peroxisome (*), representing an intermediate of the fission pro- cess. N, nucleus; M, mitochondria; Ld, lipid droplets; V, vacuole; P, peroxisomes; Wb, Woronin-body-like structures. (A–C) Bar = 5 lm; (D) Bar = 2.5 lm. Fig. 4. Properties of the Woronin body core crystal. PNS prepared from BY4742- and BY4742-expressing HEX1 were subjected to centrifugation at 1000, 5000 and 15 000 g for 5 min, in the presence or absence of Tri- ton X-100. The resulting supernatant (S) and pellet (P) fractions were subjected to SDS- gel electrophoresis and analyzed by western blotting for the distribution of HEX1, the per- oxisomal marker proteins Pex13p and Cta1p. Disintegration of the membranes by Triton X-100 changed the distribution of Pex13p and Cta1p, but not that of HEX1. Heterologous Woronin body formation C. Wu ¨ rtz et al. 2936 FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS appeared in the supernatant fraction. Sucrose density gradient centrifugation showed that the distribution of peroxisomes was similar to that for the deletion and wild-type strains (Fig. 5B). Peroxisomal density was not affected upon expression of HEX1. The distribu- tion profile of HEX1 was similar to that in the wild- type strain, although with a shift of the peak towards that of peroxisomes (Figs 1D and 5C). To analyze whether fission of peroxisomes and Woro- nin-body-like structures still occurs in the vps1Ddnm1D mutant strain, electron microscopy studies were per- formed. The vps1Ddnm1Ddouble-deletion strain showed mitochondria with abnormal morphology and large interconnected peroxisomes (Fig. 6A). Upon expression of HEX1, smaller versions of rectangular or hexagonal Woronin-body-like structures were visible (Fig. 6B,C). Their mean size was  0.7 · 3.8 lm, which was about half the size of the Woronin-body-like structures in the wild-type. Interestingly, all discernible Woronin-body- like structures had small vesicles attached. Based on the gathered data, these vesicles were likely to represent peroxisomes. To support this hypothesis, immuno- fluorescence studies were performed. Extended single peroxisomes were observed in the double-deletion strain expressing GFP-SKL (Fig. 7A). Upon expression of HEX1, the synthetic peroxisomal marker GFP-SKL appeared in spots with tail-like extensions (Fig. 7B). The spots, but not the extensions, were also stained with A B C Fig. 5. Subcellular distribution of HEX1 in a BY4742vps1Ddnm1D mutant strain. (A) Differential centrifugation. PNS was subjected to centri- fugation at 25 000 g for 20 min and the resulting pellet and supernatant fractions were analyzed by immunoblotting for the presence of HEX1, the peroxisomal marker proteins Pex13p and Cta1p. The detection of HEX1 and Cta1p in the pellet fraction indicated that the vps1Ddnm1D mutant remained import competent. (B, C) Density gradient centrifugation. The 25 000 g organellar pellets of (B) BY4742vps1Ddnm1D and (C) BY4742vps1Ddnm1D expressing HEX1 were loaded on top of a linear sucrose gradient (30–60% w ⁄ w) and subjected to centrifugation for 2 h at 38 000 g. Fractions (1 mL) were collected from the bottom (fraction 1) to the top (fraction 27) and assayed by western blot for the distribution of HEX1, the peroxisomal marker proteins Fox3p and Pex13p. A B C Fig. 6. Ultrastructure of a BY4742vps1Ddnm1D mutant expressing HEX1. Cells were processed for electron microscopy after growth on medium containing oleic acid as the sole carbon source for 14 h. (A) Typical morphology of a BY4742vps1Ddnm1D cell with large, misshapen mitochondria. (B, C) Morphology upon expression of HEX1 in BY4742vps1Ddnm1D. Woronin-body-like structures were formed that are still attached to peroxisomes. (B) One rectangular and one nearly hexagonal Woronin body are visible. (C) A rectangular Woronin body with appended peroxisomal structures on both small sides. N, nucleus; M, mitochondria; Ld, lipid droplets; V, vacuole; P, peroxisomes; Wb, Woronin-body-like structures. Bar = 5 lm. C. Wu ¨ rtz et al. Heterologous Woronin body formation FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 2937 anti-HEX1 Ig, suggesting that the observed extensions represent tubular peroxisomes attached to Woronin- body-like objects. Thus, our study shows that HEX1- containing organelles can emerge from peroxisomes in yeast, in a process that depends on the typical peroxi- somal fission machinery. Discussion The Woronin bodies of filamentous ascomycetes are highly specialized organelles. They plug septal pores after hyphal damage and protect the mycel from exces- sive loss of cytoplasm. It is thought that the formation of the core crystal of Woronin bodies occurs by self- assembly of the HEX1 protein [1]. To analyze whether additional specific factors of filamentous fungi are needed for Woronin body formation we heterologously expressed N. crassa HEX1 in S. cerevisiae. This led to the Pex5p-dependent formation of giant Woronin- body-like structures containing a stable HEX1 crystal (Fig. 4). It is currently unclear why these impressing structures were not observed previously in a similarly designed experiment [1], but we theorize that Woronin- body-like organelles only emerge above a threshold expression level of HEX1. The giant organelles exhibited a lower density than genuine Woronin bodies from N. crassa. One reason for the lower density of these Woronin-body-like struc- tures might be the presence of trace amounts of peroxi- somal matrix proteins. For example, evidence for the appearance of some GFP-SKL or Cta1p was gathered in several of our localization experiments. The density of these structures could thereby be lowered without prohibiting formation of the HEX1 crystal. In Asper- gillus oryzae, formation of HEX1 multimers, and there- fore crystal formation, was shown to depend on HEX1 phosphorylation via protein kinase C [2,20]. It is also conceivable that the yeast kinase ortholog phosphory- lates HEX1 inefficiently and as a consequence crystal assembly is adversely affected. Without a doubt, bacte- rially expressed HEX1 can spontaneously assemble to crystals in vitro [1]. However, the in vitro crystals and Woronin-body-like organelles detected in yeast exceed the size of the Woronin bodies observed in living hyphae of N. crassa. HEX1 may be capable of forming large crystals without phosphorylation, but the crystal packing might be different when HEX1 is phosphory- lated. This would require solving the 3D structure of phosphorylated HEX1 and comparing it with the published structure of unphosphorylated HEX1 [19]. Separation of Woronin-body-like organelles from peroxisomes was demonstrated in immunofluorescence and electron microscopy images by the occasional occurrence of intermediate structures harboring zones of luminal material with distinct electron density, but not yet separated by a membrane. Similar structures have been seen in micrographs from the filamentous fungus Fusarium oxysporum, revealing the formation of an electron opaque matrix that buds off from microbodies [7]. These intermediate structures were predominant in a mutant yeast strain lacking the dynamin-like proteins Dnm1p and Vps1p, which are important players in the normal fission of peroxisomes [11]. Immunofluorescence microscopy studies further showed that co-localization of GFP-SKL and HEX1 occurs much more frequently in the vps1Ddnm1D Fig. 7. Subcellular localization of HEX1, the peroxisomal marker proteins GFP-SKL and Pex14p in BY4742vps1Ddnm1D cells. Indi- rect double immunofluorescence was used with cells co-expressing HEX1 and GFP-SKL or GFP-SKL alone. Detection was achieved with mouse mAbs against GFP in combina- tion with (A) rabbit anti-Pex14p and (B) rab- bit anti-HEX1 serum. The secondary antibodies used were Alexa Fluor 488- labeled anti-mouse IgG and Alexa Fluor 594- labeled anti-rabbit IgG, respectively. (A) Single large GFP-SKL-stained peroxisomes are visible. Upon heterologous expression of HEX1 (B), GFP-SKL-stained spots with tail- like extensions were discernible. The spots but not the extensions were also decorated by the anti-HEX1 serum. Bar = 5 lm. Heterologous Woronin body formation C. Wu ¨ rtz et al. 2938 FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS mutant than in the wild-type, in accordance with a budding and detachment of Woronin-body-like organ- elles from peroxisomes. This assumption gained further weight by the appearance of extensions from double- labeled spots that exclusively contained GFP-SKL. The methods applied did not, however, allow us to discern whether these tubular structures are indeed interconnected or represent clustered peroxisomes. In future studies, time-lapse microscopy will be needed to directly show the emergence of HEX1-containing organelles. Nonetheless, Vps1p and Dnm1p clearly con- trolled the budding of Woronin-body-like organelles from yeast peroxisomes and we thus suppose that both proteins are also key players in Woronin body forma- tion in filamentous fungi. A recent study by Liu et al. revealed that N. crassa vps1 and dnm1 single mutants were clearly compromised in peroxisomal fission, whereas Woronin body formation was still feasible, although the size and numbers of Woronin bodies were reduced [21]. It will be interesting to see whether the absence of both dynamin-like proteins prevents the fission of Woronin bodies in filamentous fungi. It is worth noting that in the vps1Ddnm1D strain, HEX1 was almost exclusively localized to the organel- lar pellet and exhibited a similar distribution in sucrose density gradients as in the wild-type. This also held largely true for Cta1p. Furthermore, the density of peroxisomes remained unchanged in the vps1Ddnm1D mutant, thereby indicating that an impaired fission process does not significantly interfere with peroxi- somal protein import. Expressing HEX1 in the well-characterized model organism S. cerevisiae could be a suitable tool to ana- lyze the mechanism of peroxisome division in more detail. Organellar budding can be easily followed by imaging techniques due to the different morphology of peroxisomes and Woronin bodies and, most impor- tantly, is arrested at intermediate states in vps1Ddnm1D deficient cells. The expression of dynamin-like proteins from inducible promoters in the double-deletion strains will allow studying the dynamic fission process in a time-dependent manner. Experimental procedures Strains and culture conditions For all plasmid amplifications and isolations Escherichia coli strain DH5a was used (Invitrogen, Carlsbad, CA, USA). The yeast wild-type strain BY4742 was used. The strain BY4742pex5D was obtained from the EUROSCARF strain collection (Frankfurt, Germany) and construction of the double-deletion strain BY4742vps1Ddnm1D was as described previously [11]. Transformation of yeast cells was performed as described previously [22]. Media for the culti- vation of yeast and bacterial strains were prepared as described elsewhere [23,24]. N. crassa strain FGSC#987 (St. Lawrence 74-OR23-1A, mat A) was obtained from the Fungal Genetics Stock Center (Kansas City, KS, USA). Plasmids and cloning procedures For heterologous expression in yeast, N. crassa HEX1 was amplified from a N. crassa cDNA library using PCR with primer pair RE951 (AAGAATTCATGGGCTACTACGA CGAC) ⁄ RE952 (AACTCGAGTTAGAGGCGGGAACC GTG) introducing recognition sites for EcoRI and XhoI (obtained from MWG-BIOTECH AG, Ebersberg, Germany). The resulting fragment was subcloned into pBluescript SK(+) (Stratagene, La Jolla, CA, USA) for sequencing purposes. The identity of the HEX1 fragment was verified by automated sequencing (MWG-BIOTECH). The insert was cloned as an EcoRI–XhoI fragment into appropriately prepared pYPGE15 [25], designed for constit- utive expression in S. cerevisiae. Antibodies Antibodies against GFP (BD Biosciences, Franklin Lakes, NJ, USA), Pep12p (Molecular Probes, Eugene, OR, USA), Kar2p [26], TIM23 [27], ICL1 [28], Cta1p [29], HEX1 [30], Fox3p [31], Pex14p [31], Aac2p [32], and Pex13p [33] have been described previously. SDS ⁄ PAGE and immuno- blotting were performed according to standard protocols [23]. Horseradish peroxidase-coupled anti-rabbit and anti- mouse IgG, in combination with the ECLÔ system (GE Healthcare, Munich, Germany), was used to detect immuno- reactive complexes. Subcellular fractionation Preparation of PNS from S. cerevisiae cells and differential centrifugation at 25 000 g were conducted as described pre- viously [24]. Density gradient centrifugation was carried out as described previously [34] with the modification that instead of a PNS, a 100 000 g organellar pellet was loaded on top of the gradients, and a 30–60% (w ⁄ w) sucrose den- sity gradient with a 2 mL 65% sucrose cushion was used. The preparation of a N. crassa PNS and the separation of a 25 000 g organellar pellet by density gradient centri- fugation has been described elsewhere [9]. Electron microscopy The ultrastructure of yeast cells was studied with oleate- induced cells that had been fixed with 1.5% KMnO 4 and processed as described previously [24]. C. Wu ¨ rtz et al. Heterologous Woronin body formation FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 2939 Immunofluorescence microscopy All light microscopy studies were performed with an Axio- plan microscope and axiovision 4.6 software (Zeiss, Jena, Germany) as described previously [35]. Antibodies and used dilutions were as follows: N. crassa anti-HEX1 serum (1 : 100), S. cerevisiae anti-Pex14p serum (1 : 500) and anti- GFP IgG (1 : 100). The secondary antibodies applied were obtained from Molecular Probes (Alexa Fluor 594 goat anti-rabbit IgG and Alexa Fluor 488 goat anti-mouse IgG). Acknowledgements We thank F. Nargang for the N. crassa cDNA library and M. Bu ¨ rger for technical assistance. This work was supported by grants from the Deutsche Forschungs- gemeinschaft (project B10 of SFB480) and by the Fonds der Chemischen Industrie. References 1 Jedd G & Chua N-H (2000) A new self-assembled per- oxisomal vesicle required for efficient resealing of the plasma membrane. Nat Cell Biol 2, 226–231. 2 Tenney K, Hunt I, Sweigard J, Pounder JI, McClain C, Bowman EJ & Bowman BJ (2000) hex-1, a gene unique to filamentous fungi, encodes the major protein of the Woronin body and functions as a plug for septal pores. Fungal Genet Biol 31, 205–217. 3 Buller AHR (1933) Researches in Fungi. Hafner, New York, NY. 4 Woronin MS (1864) Zur Entwicklungsgeschichte des Ascobolus pulcherrimus Cr. und einiger Pezizen. Abh Senkenb Naturforsch Ges 5, 333–344. 5 Soundararajan S, Jedd G, Li X, Ramos-Pamplona M, Chua NH & Naqvi NI (2004) Woronin body function in Magnaporthe grisea is essential for efficient pathogen- esis and for survival during nitrogen starvation stress. Plant Cell 16, 1564–1574. 6 Simon UK, Bauer R, Rioux D, Simard M & Oberwin- kler F (2005) The vegetative life-cycle of the clover pathogen Cymadothea trifolii as revealed by transmis- sion electron microscopy. Mycol Res 109, 764–778. 7 Wergin WP (1973) Development of Woronin bodies from microbodies in Fusarium oxysporum f. sp. lycoper- sici. Protoplasma 76, 249–260. 8 Tey WK, North AJ, Reyes JL, Lu YF & Jedd G (2005) Polarized gene expression determines Woronin body formation at the leading edge of the fungal colony. Mol Biol Cell 16, 2651–2659. 9 Managadze D, Wu ¨ rtz C, Sichting M, Niehaus G, Veenhuis M & Rottensteiner H (2007) The peroxin PEX14 of Neurospora crassa is essential for the biogene- sis of both glyoxysomes and Woronin bodies. Traffic 8, 687–701. 10 Hoepfner D, van den Berg M, Philippsen P, Tabak HF & Hettema EH (2001) A role for Vps1p, actin, and the Myo2p motor in peroxisome abundance and inheritance in Saccharomyces cerevisiae. J Cell Biol 155, 979–990. 11 Kuravi K, Nagotu S, Krikken AM, Sjollema K, Deckers M, Erdmann R, Veenhuis M & van der Klei IJ (2006) Dynamin-related proteins Vps1p and Dnm1p control peroxisome abundance in Saccharomyces cerevi- siae. J Cell Sci 119, 3994–4001. 12 Erdmann R & Blobel G (1995) Giant peroxisomes in oleic acid-induced Saccharomyces cerevisiae lacking the peroxisomal membrane protein Pmp27p. J Cell Biol 128, 509–523. 13 Marshall PA, Krimkevich YL, Lark RH, Dyer JM, Veenhuis M & Goodman JM (1995) Pmp27 promotes peroxisomal proliferation. J Cell Biol 129, 345–355. 14 Schrader M, Reuber BE, Morrell JC, Jimenez-Sanchez G, Obie C, Stroh TA, Valle D, Schroer TA & Gould SJ (1998) Expression of PEX11beta mediates peroxisome proliferation in the absence of extracellular stimuli. J Biol Chem 273, 29607–29614. 15 Smith JJ, Marelli M, Christmas RH, Vizeacoumar FJ, Dilworth DJ, Ideker T, Galitski T, Dimitrov K, Rachu- binski RA & Aitchison JD (2002) Transcriptome profil- ing to identify genes involved in peroxisome assembly and function. J Cell Biol 158, 259–271. 16 Rottensteiner H, Stein K, Sonnenhol E & Erdmann R (2003) Conserved function of Pex11p and the novel Pex25p and Pex27p in peroxisome biogenesis. Mol Biol Cell 14, 4316–4328. 17 Tam YY, Torres-Guzman JC, Vizeacoumar FJ, Smith JJ, Marelli M, Aitchison JD & Rachubinski RA (2003) Pex11-related proteins in peroxisome dynamics: a role for the novel peroxin Pex27p in controlling peroxisome size and number in Saccharomyces cerevisiae. Mol Biol Cell 14, 4089–4102. 18 Trinci A & Collinge AJ (1974) Occlusion of the septal pores of damaged hyphae of Neurospora crassa by hexagonal crystals. Protoplasma 80, 57–67. 19 Yuan P, Jedd G, Kumaran D, Swaminathan S, Shio H, Hewitt D, Chua N-H & Swaminathan K (2003) A HEX-1 crystal lattice required for Woronin body func- tion in Neurospora crassa. Nat Struct Mol Biol 10, 264– 270. 20 Juvvadi P, Maruyama J & Kitamoto K (2007) Phos- phorylation of the Aspergillus oryzae Woronin body protein, AoHex1, by protein kinase C: evidences for its role in the multimerization and proper localization of the Woronin body protein. Biochem J 405, 533– 540. 21 Liu F, Ng SK, Lu Y, Low W, Lai J & Jedd G (2008) Making two organelles from one: Woronin body bio- genesis by peroxisomal protein sorting. J Cell Biol 180, 325–339. Heterologous Woronin body formation C. Wu ¨ rtz et al. 2940 FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 22 Schiestl R & Gietz R (1989) High efficiency transforma- tion of intact yeast cells using single stranded nucleic acids as a carrier. Curr Genet 16, 339–346. 23 Sambrook J, Fritsch E & Maniatis T (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 24 Erdmann R, Veenhuis M, Mertens D & Kunau W (1989) Isolation of peroxisome-deficient mutants of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 86, 5419–5423. 25 Brunelli J & Pall M (1993) A series of yeast shuttle vectors for expression of cDNAs and other DNA sequences. Yeast 9, 1299–1308. 26 Rose MD, Misra LM & Vogel JP (1989) KAR2, a kary- ogamy gene, is the yeast homolog of the mammalian BiP ⁄ GRP78 gene. Cell 57, 1211–1221. 27 Mokranjac D, Paschen S, Kozany C, Prokisch H, Hop- pins S, Nargang F, Neupert W & Hell K (2003) Tim50, a novel component of the TIM23 preprotein translocase of mitochondria. EMBO J 22, 816–825. 28 Maeting I, Schmidt G, Sahm H, Revuelta J, Stierhof Y & Stahmann K (1999) Isocitrate lyase of Ashbya gos- sypii – transcriptional regulation and peroxisomal local- isation. FEBS Lett 444, 15–21. 29 Gurvitz A, Rottensteiner H, Kilpelainen S, Hartig A, Hiltunen J, Binder M, Dawes I & Hamilton B (1997) The Saccharomyces cerevisiae peroxisomal 2,4–dienoyl- CoA reductase is encoded by the oleate-inducible gene SPS19. J Biol Chem 272, 22140–22147. 30 Schliebs W, Wu ¨ rtz C, Kunau W, Veenhuis M & Rottensteiner H (2006) A eukaryote without catalase- containing microbodies: Neurospora crassa exhibits a unique cellular distribution of its four catalases. Eukary- otic Cell 5, 1490–1502. 31 Erdmann R & Kunau W (1994) Purification and immu- nolocalization of the peroxisomal 3–oxoacyl-CoA thio- lase from Saccharomyces cerevisiae. Yeast 10, 1173–1182. 32 Palmieri L, Rottensteiner H, Girzalsky W, Scarcia P, Palmieri F & Erdmann R (2001) Identification and functional reconstitution of the yeast peroxisomal ade- nine nucleotide transporter. EMBO J 20, 5049–5059. 33 Erdmann R & Blobel G (1996) Identification of Pex13p a peroxisomal membrane receptor for the PTS1 recogni- tion factor. J Cell Biol 135, 111–121. 34 Scha ¨ fer A, Kerssen D, Veenhuis M, Kunau WH & Schliebs W (2004) Functional similarity between the peroxisomal PTS2 receptor binding protein Pex18p and the N–terminal half of the PTS1 receptor Pex5p. Mol Cell Biol 24, 8895–8906. 35 Girzalsky W, Rehling P, Stein K, Kipper J, Blank L, Kunau W & Erdmann R (1999) Involvement of Pex13p in Pex14p localization and peroxisomal targeting signal 2-dependent protein import into peroxisomes. J Cell Biol 144, 1151–1162. C. Wu ¨ rtz et al. Heterologous Woronin body formation FEBS Journal 275 (2008) 2932–2941 ª 2008 The Authors Journal compilation ª 2008 FEBS 2941 . Dynamin-like protein- dependent formation of Woronin bodies in Saccharomyces cerevisiae upon heterologous expression of a single protein Christian Wu ¨ rtz,. HEX1 expression (Fig. 2B). Remarkably, typical Woronin bodies of N. crassa are smaller with an average size of  400–700 nm [18]. In a pex5D strain, staining