|Home | About | Journals | Submit | Contact Us | Français|
Peroxisomes are a diverse class of organelles involved in different physiological processes in eukaryotic cells. Although proteins imported into peroxisomes carry a peroxisomal targeting sequence at the C terminus (PTS1) or an alternative one close to the N terminus (PTS2), the protein content of peroxisomes varies drastically. Here we suggest a new class of peroxisomes involved in microtubule (MT) formation. Eukaryotic cells assemble MTs from distinct points in the cell. In the fungus Aspergillus nidulans, septum-associated microtubule-organizing centers (sMTOCs) are very active in addition to the spindle pole bodies (SPBs). Previously, we identified a novel MTOC-associated protein, ApsB (Schizosaccharomyces pombe mto1), whose absence affected MT formation from sMTOCs more than from SPBs, suggesting that the two protein complexes are organized differently. We show here that sMTOCs share at least two further components, gamma-tubulin and GcpC (S. pombe Alp6) with SPBs and found that ApsB interacts with gamma-tubulin. In addition, we discovered that ApsB interacts with the Woronin body protein HexA and is targeted to a subclass of peroxisomes via a PTS2 peroxisomal targeting sequence. The PTS2 motif was necessary for function but could be replaced with a PTS1 motif at the C terminus of ApsB. These results suggest a novel function for a subclass of peroxisomes in cytoskeletal organization.
Peroxisomes are ubiquitous organelles of eukaryotes which are surrounded by a single membrane (9, 30). They serve a variety of functions, depending on the species, the cell type, and the environmental or developmental conditions. In mammals, peroxisomes are involved in a number of catabolic and anabolic pathways, most importantly, peroxide metabolism, the β-oxidation of long-chain fatty acids, and the biosynthesis of ether phospholipids (17, 37). The vital importance of the organelle in humans is shown by the existence of a number of severe and often lethal disorders that occur when the biogenesis of the organelle is impaired (36). In plants, peroxisomes are involved in photorespiration and typically contain the glyoxylate cycle, as in protozoa and yeast (8).
Given the complexity of peroxisomal functions, it is obvious that a large number of proteins need to be targeted to these organelles. Peroxisomal membrane and matrix proteins are synthesized on free ribosomes in the cytosol and are imported posttranslationally into preexisting organelles (9). The apparatus of protein import is clearly distinct from the import machinery of other organelles because it translocates folded and even oligomeric proteins and there is evidence that they are descending from the endoplasmic reticulum (6). A large number of peroxisomal proteins employ a tripeptide sequence at the C terminus, PTS1 (S/A/C-K/R/H-L/M) (7). A second class of proteins uses a sequence close to the N terminus which is less conserved, consists of R/K-L/I/V-X5-H/Q-L/A, and is called PTS2 (33). In both cases, complex protein machineries are employed and some of the components appear to be used in PTS1- and PTS2-dependent protein translocation (9).
A very distinct class of peroxisomes is represented by the fungal Woronin body. This structure is named after a Russian mycologist who reported the characteristics of a distinct type of organelle in the fungus Ascobolus pulcherrimus (1, 39). Woronin bodies have been described in more than 50 species of ascomycota and deuteromycota but are missing in single-cell yeasts such as Saccharomyces cerevisiae and Schizosaccharomyces pombe. Thus, their function appears to be important for the filamentous life style. In Neurospora crassa, they appear as hexagonal bodies in the cell and upon cell damage plug the septal pores after a few seconds (15). This sealing mechanism is very important in syncytial organisms to prevent loss of the entire cytoplasm and hence death of the entire mycelium after one hypha is damaged (20). Their exact composition, however, remained obscure for many decades until G. Jedd and N.-H. Chua purified the organelle from N. crassa and identified the main constituent as a single protein named Hex1 (15, 42), because it forms hexagonal crystals. The existence of a PTS1 peroxisomal targeting sequence at the C terminus of the protein indicated that the Woronin bodies represent specialized peroxisomes. Hex1 displays some sequence similarity to eIF5, and it is thought that Hex1 derived from eIF5 during evolution by gene duplication and subsequent modification of its function (42). Another example of a peroxisome-associated function may be the Pro40 protein in Sordaria macrospora (5). This protein is implicated in the regulation of sexual development.
In addition to the Woronin body close to the septal pore (22), we had evidence in Aspergillus nidulans that microtubule (MT) polymerization is initiated at septa. Using an MT plus-end-associated protein, the kinesin motor KipA (kinesin-7), we showed that the cytoplasmic area close to septa acts as an active MT-organizing center (MTOC) (16). Furthermore, we identified a novel MTOC-associated protein, ApsB, and localized it to the spindle pole bodies (SPBs) and to septa (35). The presence of septal MTOCs is similar to that in S. pombe, but there is no evidence for such organelles in the S. cerevisiae-related filamentous fungus Ashbya gossypii (18, 19). Given that MTOCs are generally composed of a large protein complex with gamma-tubulin as one characteristic member, we anticipate that MT polymerization at septa also requires a protein complex (40). Here we show for the first time the presence of gamma-tubulin at septal MTOCs (sMTOCs) and that it physically interacts with ApsB. Surprisingly, ApsB is associated with a subclass of peroxisomes in the cytoplasm, and we propose that they are involved in septal MTOC formation.
The preparation of the supplemented minimal and complete media used for A. nidulans and the standard strain construction procedures used are described in reference 11. To isolate total DNA and RNA, corresponding strains were grown in liquid culture for 16 h. Mycelium was harvested and immediately processed for total DNA (see below). A list of the A. nidulans strains used in this study is given in Table 1. Standard laboratory Escherichia coli strains (XL-1 blue and Top10) were used. The plasmids used are listed in Table 2.
For live-cell imaging, cells were grown in glass-bottom dishes (FD35-100; World Precision Instruments, Berlin, Germany) in 2 ml of minimal medium (MM) containing either 2% glycerol or 2% glucose as a carbon source. Medium was supplemented with pyridoxine, p-aminobenzoic acid, biotin, arginine, uracil, or uridine, depending on the auxotrophy of the strains. Cells were incubated at room temperature for 1 to 2 days, and images were captured using an Axiophot microscope (Zeiss, Jena, Germany), a Plan-apochromatic 63× or 100× oil immersion objective lens, and an HBO50 Hg lamp. Alternatively, a Zeiss AxioImager Z1 with AxioVision software (V4.5) was used. Fluorescence was observed using standard Zeiss filter combinations no. 09 (fluorescein isothiocyanate, green fluorescent protein [GFP]) and no. 15 (monomeric red fluorescent protein 1 [mRFP1], DsRed). Laser images were obtained using the Zeiss Cell Observer SD, which combines the high-end Cell Observer microscopy platform and the CSU-X1 spinning-disc technology from Yokogawa for high-speed confocal microscopy. Images were collected and analyzed with a Hamamatsu Orca ER II camera system and the Wasabi software (version 1.2) or a Zeiss Axiocam and AxioVision software. Image and video processing was done with the Wasabi software from Hamamatsu, Adobe Photoshop, ImageJ (NIH, Bethesda, MD), and virtual dub (http://www.virtualdub.org).
Standard DNA transformation procedures were used for A. nidulans (41) and E. coli (27). For PCR experiments, standard protocols were applied using a Biometra Personal Cycler (Biometra, Göttingen, Germany) for the reaction cycles. DNA sequencing was done commercially (MWG Biotech, Ebersberg, Germany). Total DNA was extracted from A. nidulans in the following way. Spores were inoculated into liquid MM plus supplements and grown for 12 to 18 h at 30°C without shaking. Hyphal mats were harvested, dried with tissue paper, and ground in liquid nitrogen. The resulting powder was mixed with extraction buffer (50 mM EDTA, 0.2% sodium dodecyl sulfate [SDS]) and incubated for 30 min to 2 h at 68°C in a water bath. SDS and proteins were removed from the suspension by addition of potassium-acetate solution (8 M, pH 4.2) and centrifugation. Total DNA was precipitated from the supernatant with isopropanol, and the pellet was washed twice with 70% ethanol, air dried, resuspended in TE buffer with RNase A, and stored at 4°C. Southern hybridizations were performed according to the DIG Application Manual for Filter Hybridization (Roche Applied Science, Technical Resources, Roche Diagnostics GmbH, Mannheim, Germany).
The enhanced yellow fluorescent protein (eYFP)-tagged N-terminal half (YFPN) was amplified using primers 5′-CGGTACCATGGTGAGCAAGGGCGAGGAGCTG-3′ (fwd_Kpn_YFP-N) and 5′-CGGCGCGCCCGTGGCGATGGAGCGCATGATATAGACGTTGTGGCTGTTGTAG-3′. For the C-terminally eYFP-tagged (YFPC) half, primers 5′-CGGTACCATGGCCGACAAGCAGAAGAACGGCATCAAGG-3′ (fwd_Kpn_YFP-C) and 5′-CGGCGCGCCGTGGTTCATGACCTTCTGTTTCAGGTCGTTCGGGATCTTGCAGGCCGGGCGCTTGTACAGCTCGTCCATGCCGAGAGTGATCCC-3′ (rev_YFP-C_Li_Asc) were used. These primers introduced KpnI and AscI restriction sites (in italics) in addition to the protein linker sequences RSIAT (YFPN) and RPACKIPNDLKQKVMNH (YFPC) (underlined). eYFP was split at bp 460 to 462 by using the ATG codon as the start of the YFPC half. PCR fragments were subcloned into pCR2.1-Topo (Invitrogen, Karlsruhe, Germany), subsequently released with KpnI and AscI, and used to replace GFP2-5 of pMCB17apx-apsB (35), giving pDV7 (YFPN) and pDV8 (YFPC). Full-length apsB3.2 (3.2 kb) was taken from pDV21a and cloned into pDV7, giving pDV22b [alcA(p)::YFPN::apsB3.2], and full-length γtubulin1.8 was amplified using primers 5′-CGGCGCGCCCGGG ATGCCTAGGTATACCCTC-3′ (Gamma_tub_Asc_fwd) and 5′-CTTAATTAA TTATACTCCAACTTCATCCTTTCC-3′ (Gamma_tub_Pac_rev) (the AscI and PacI restriction sites are in italics, and the start and stop codons are underlined) and cloned into pDV8, giving pDV50 [alcA(p)::YFPC::γtubulin1.8]. For BiFC analysis, pDV22 and pDV50 were combined and transformed into GR5, giving SNZ11. The cloned gene length is indicated as exponents; e.g., the length of the cloned gene apsB3.2 is 3.2 kb. Similarly, hexA was amplified using primers 5′-CGGCGCGCCCGGGATGGGTTACTACGACGACG-3′ (hexA_Asc_fwd) and 5′-CTTAATTAATTATAGACGGGAAGAGTGGATGATC-3′ (hexA_ Pac_rev1; 680 bp to stop codon) or 5′-GTTAATTAACCTCAATCAAGTGCAAGGTTTCG-3′ [hexA_Pac_rev2; 1 kb, including the poly(A) site] and cloned into pDV8, giving pDV17 [alcA(p)::YFPC::hexA680], and into pDV7, giving pDV19a [alcA(p)::YFPN::hexA1.0]. For BiFC analysis, pDV17 and pDV22b were combined and transformed into GR5, giving SDV42, or pDV19a and pDV23a were combined, giving SDV43.
To prepare protein extracts, A. nidulans strains SNZ-SI 42 [alcA(p)::3×HA::apsB3.2], SNZ16 [alcA(p)::GFP::γtubulin1.8], and SNZ37 [alcA(p)::apsB3.2::3×HA alcA(p)::GFP::γtubulin1.8] were incubated in liquid MM for 24 h at 37°C. The medium was supplemented with 0.2% glucose and 2% threonine to induce the alcA promoter. The mycelium was harvested by filtration through Miracloth (Calbiochem, Heidelberg, Germany), dried between paper towels, and immediately ground in liquid nitrogen. Afterwards, the mycelial powder was resuspended in protein extraction buffer (20 mM Tris-HCl, pH 8, 150 mM NaCl, 0.01% Triton X-100) containing protease inhibitor (2 mM phenylmethylsulfonyl fluoride [PMSF]) and vortexed for 5 min at 4°C. Cell debris was pelleted by two centrifugations (Eppendorf centrifuge 5403; Eppendorf, Hamburg, Germany) at 13,000 rpm at 4°C for 10 min. A volume of 1 ml protein extract was adjusted to 300 mM NaCl and incubated with monoclonal antibody HA.11 (dilution, 1:200; clone 16B12; Hiss Diagnostics, Freiburg, Germany). After 1 h of incubation at 4°C, 50 μl protein-G-agarose (Roche, Mannheim, Germany) was added and the mixture was incubated for an additional 3 h. Agarose beads were pelleted by centrifugation in an Eppendorf centrifuge at 15,000 rpm at 4°C for 30 s and washed three times with 1 ml extraction buffer containing protease inhibitor (2 mM PMSF) with different NaCl molarities (150 mM NaCl, 500 mM NaCl, and no NaCl). After the denaturation of the samples, protein extracts and coimmunoprecipitated pellets were loaded onto an 8% SDS-polyacrylamide gel. For Western blotting, a polyclonal antibody raised against GFP (product G1544; dilution, 1:4,000; Sigma-Aldrich, Munich, Germany) with anti-rabbit IgG peroxidase conjugate secondary antibody (product A0545; dilution, 1:4,000; Sigma-Aldrich, Munich, Germany) in the case of gamma-tubulin and the anti-HA antibody (clone 16B12; dilution, 1:1,000) with anti-mouse IgG peroxidase conjugate secondary antibody (product A2304; dilution, 1:10,000; Sigma-Aldrich, Munich, Germany) in the case of ApsB were used. Nitrocellulose membranes used for blotting were from Schleicher & Schuell (Dassel, Germany).
A full-length cDNA fragment of apsB was amplified with primers 5′-GGATCCGAATGACTCTAAAAGAGC-3′ and 5′-GTCGAC TCAAACTTCGATATCAAC-3′ and cloned into the BamHI-SalI restriction sites of pGBT9 (Clontech), giving pRS89, and into pGAD424 (Clontech), giving pRS88. A cDNA fragment from a cDNA library containing the full-length hexA gene was cloned into of the yeast GAL4-Matchmaker system (Clontech), giving pRS91. Transformation of yeast strains, selection for diploids, a histidine growth assay, and a β-galactosidase (β-Gal) assay were done as described in reference 2. The yeast strains used for transformation were AH109, Y187, and PJ69-4A (Clontech).
The peroxisomal target sequence of apsB was mutated using pDV21a as the template and the QuikChange XL site-directed mutagenesis kit from Stratagene. The last two amino acids were mutated using primer 5′-GCGATTTGGAGAAGCTACGTAAGACCAGCAGTCAGATAAGGAG-3′ and the corresponding antiparallel primer, giving pDV43. Successful mutagenesis was confirmed by commercial sequencing (MWG Biotech, Ebersberg, Germany).
pMCB17apx was used as the basic vector for the tagging of apsB or hexA with GFP, and pDM8 was used for tagging with mRFP1 (see reference 35). Full-length hexA was amplified from genomic DNA using primers 5′-CGGCGCGCCCGGGATGGGTTACTACGACGACG-3′ and 5′-CTTAATTAATTATAGACGGGAAGAGTGGATGATC-3′.
To obtain in vivo protein expression levels, we expressed the proteins under the control of the corresponding natural promoters. The apsB promoter (1.33 kb) was amplified from genomic DNA using primers 5′-GCCTAGGCAAGCCGCAACTCCC-3′ (apsB_nat(p)_AvrII_fwd) and 5′-CGGTACCGGATCTGCCACTGCG-3′ (apsB_nat(p)_KpnI_rev) (the AvrII and KpnI restriction sites are in italics), cloned instead of alcA(p) into pDV21, giving pNZ-SI37 [apsB(p)::GFP::apsB], and transformed into TN02A3, giving SNZ59. The gamma-tubulin promoter (1.16 kb) was amplified from genomic DNA using primers 5′-GGAATTCCATACCCAGCATAAATTCGG-3′ (Gamma_tub_nat(p)_EcoRI_fwd) and 5′-CCGTACGCTTTCTTGCTTGCCTTAAG-3′ (Gamma_tub_nat(p)_BsiwI_rev) (EcoRI and BsiwI restriction sites are in italics), cloned instead of alcA(p) into pNZ17, giving pNZ-SI36 (γtubulin(p)::GFP::γtubulin1.8), and transformed into TN02A3, giving SNZ61. AlpB AN4867 (S. pombe Alp6) was amplified via fusion PCR using primers 5′-GGGAGGACAAATACAAACTCG-3′ (Alp6_mitte_fwd) and 5′-ctccagcgcctgcaccagctccTTGCTCAGTCGAATCCTTCTTTTC-3′ (Alp6_linker_rev) to amplify the C-terminal fragment of AlpB without the stop codon and primers 5′-atcagtgcctcctctcagacagTAGCATACATGCAGTACATTTCTCG-3′ (Alp6_RB_link_fwd) (linkers in lower case letters) and 5′-ACCGTCATGGCAGAAACGAAG-3′ (Alp6_RB_rev) to amplify the right border of AlpB. The two PCR products were fused to a GFP-pyrG PCR cassette (kindly provided by S. Osmani, Ohio State University) to generate a 5.5 fusion PCR product using primers 5′-CCAGTCTCGAGACCTCAATTG-3′ (Alp6_Nprimer_fwd) and 5′-TTATCACCTGCTGGTTCTGAG-3′ (Alp6_Nprimer_rev). The fusion PCR product was transformed into A. nidulans strain SO451, giving SNZ-SH80 [alpB(p)::alpB::GFP].
A PTS1 targeting sequence (SRL) was added to the C terminus of ApsB by amplifying the full-length mutated gene apsB3.2_PTS2mut in pDV43 using primers 5′-TTTGGGCGCGCCCGGCATGACTCTAAAAGAGCAAAGTAGTACG-3′ (apsB_ Asc_fwd) and 5′-CCTTAATTAATCAtagacgggaAACTTCGATATC-3′ (SRL_PTS1_PacI_rev) (PTS1 is in lowercase letters). The PCR product was cloned between the AscI and PacI restriction sites in the vector pMCB17apx and confirmed via sequencing, giving plasmid pNZ16, which was transformed into apsB10 mutant strain AJC1.7, generating strain SNZ34 [apsB10, alcA(p)::GFP::apsB_PTS2mut_SRL]. Ectopic integration of the construct and the presence of the mutated endogenous apsB locus were confirmed by PCR, Southern blotting, and sequencing of the PCR products. Likewise, transformation of the apsB construct was done with pNZ16 into apsB deletion strain SRS24, generating SNZ94 with the same rescue phenotype as in the case of AJC1.7.
Spores (103/ml) were inoculated with 0.5 ml MM on sterile coverslips for 12 to 24 h at room temperature (RT). Cells were fixed for 30 min with formaldehyde and digested for 1 h using digestion solution (Glucanex, β-d-glucanase, lyticase, and Driselase in Na-phosphate buffer with 50% egg white), washed with PBS, incubated in −20°C methanol for 10 min, and blocked with Tris-buffered saline-Tween 20 (TBST) plus 5% skim milk before incubation with the first monoclonal antibody (anti-gamma-tubulin T6657 at 1:500; Sigma-Aldrich) in TBST overnight at 4°C. Next, cells were washed and incubated with the Alexa Fluor546-labeled goat anti-mouse secondary antibody (A11003 at 1:200 in TBST; Molecular Probes) for 1 h at RT. Cells were washed and mounted on microscope slides (with VECTASHIELD mounting medium with DAPI [4′,6-diamidino-2-phenylindole]), sealed with nail polish, and stored at 4°C overnight in the dark before microscopy.
A. nidulans ApsB has been localized at SPBs and at septa, suggesting the presence of MTOCs at septa (Fig. 1) (35). MTOCs are large protein complexes which consist of several proteins, gamma-tubulin, and associated gamma-tubulin complex proteins, which are mostly conserved from yeast to humans (26). Unfortunately, the most important protein of MTOCs, gamma-tubulin, has not been identified at septa of A. nidulans before. In our own experiments, we were able to detect a very weak signal at septa when gamma-tubulin was expressed from its own promoter and fused to GFP. In S. pombe, it has also been reported that gamma-tubulin was present at nonnuclear MTOCs in very small amounts and thus was also not easy to detect (29). To further elucidate the composition of septal MTOCs, we searched the A. nidulans genome for a homologue of S. pombe Alp6 (S. cerevisiae Spc98, human Gcp3) and identified the open reading frame AN4867 (968 amino acids in length) with 35% identity to Alp6. In order to localize the corresponding protein, we constructed a C-terminal GFP fusion protein expressed from the native promoter and transformed it into A. nidulans (SO451). The protein localized to MTOCs at nuclei and at septa, indicating that the two MTOCs also share this protein (Fig. 1). During the course of our experiments, this gene was analyzed in the laboratory of B. Oakley and was named gcpC (40).
In order to demonstrate that ApsB and gamma-tubulin colocalize at MTOCs, we visualized gamma-tubulin in a GFP-ApsB-expressing strain (Fig. 2A). Next, we showed that ApsB not only colocalizes but also interacts with gamma-tubulin. To this end, we applied the BiFC assay system and fused full-length ApsB with the N-terminal part of YFP and full-length gamma-tubulin with the C-terminal part of YFP. Corresponding A. nidulans strains showed a YFP signal at nuclei and at septa (Fig. 2B). Interestingly, we also found a fluorescence signal at the tips of all actively growing hyphae (Fig. 2C). Previously, ApsB had already been found at the hyphal tip and growing MTs were also reported to originate from the hyphal tip in some cases (16, 35). Gamma-tubulin alone was not visible at the hyphal tip, probably due to the high cytoplasmic background. Some cytoplasmic spots were also observed, as shown before for ApsB alone (35). Control experiments with ApsB or gamma-tubulin alone did not result in any fluorescence.
The ApsB-gamma-tubulin interaction result was confirmed by coimmunoprecipitation using hemagglutinin (HA)-ApsB and GFP-gamma-tubulin tagged proteins. Gamma-tubulin was detected in the precipitate obtained with anti-HA antibodies (Fig. 2D).
To further analyze the role of ApsB, we employed a yeast two-hybrid analysis. The cDNA of apsB was cloned into pGBT9 (pRS89) and transformed into PJ69-4A. This strain was used as a recipient strain for a yeast two-hybrid gene bank kindly provided by S. Osmani (24). Besides ApsB itself, we identified five putative interacting clones (three unknown proteins, one Zn2+ finger protein, and one putative nucleoside transporter). The translation product of one of the clones displayed sequence identity to A. nidulans HexA, the homologue of the N. crassa Hex-1 protein (15). To prove the interaction between ApsB and HexA, we cloned the full-length hexA gene and tested it in the interaction screen. A yeast strain of mating type a (AH109) containing the GAL4 binding domain with apsB (Y2HapsB-BD) was crossed to a yeast strain of mating type α (Y187) containing the GAL4 activation domain with apsB (Y2HapsB-AD), hexA (Y2HhexA-AD), or just the empty vector (pGAD424) as a control. Diploids were identified by selective growth on YEPD agar medium lacking leucine and tryptophan (YEPD LT−) (Fig. 3A), as described by Cagney et al. (2). To look for positive protein-protein interactions, colonies were subsequently inoculated onto YEPD medium lacking leucine, tryptophan, and histidine (YEPD LTH−) (Fig. 3B). This medium allows growth only if the strains produce their own histidine due to a positive interaction of the respective proteins. Colonies with apsB/apsB and apsB/hexA grew well in comparison to the apsB/empty-vector combination. The weak background growth was reduced by the addition of 3 mM 3-amino-1,2,4-triazole (3-AT) (Fig. 3C). The positive interaction was confirmed with the β-Gal assay on membranes. ApsB interacted with itself, as well as with HexA, while in combination with the empty vector no reaction occurred (Fig. 3D). To exclude the possibility that neither HexA nor ApsB has an intrinsic affinity for other proteins and gives false positives in the yeast two-hybrid assay, we tested both of them and the empty vector against a library of Treponema pallidum with 73 different proteins (kindly provided by P. Uetz) (not shown). Growing on YPED LTH−, the empty vector produced three strains (4%) with a false-positive reaction, which were also seen for apsB and hexA. Despite these, no positive interaction with any of the remaining 70 T. pallidum proteins was found, indicating that neither ApsB nor HexA interacts randomly with given proteins and confirming that the HexA/ApsB interaction was specific. Sequence inspection of ApsB revealed a putative peroxisomal targeting sequence (PTS2), KIRDLEKQL, at amino acid positions 66 to 74. Likewise, proteins with sequence similarity to ApsB (29), such as mto1 (formerly known as mod20 or mbo1) and pcp1 (S. pombe), NCU02332.1 and NCU02411.1 (N. crassa), and AAH46878 (Drosophila melanogaster) and CDK5RAP2 (Homo sapiens) all have a possible PTS2, as identified with the software program psort (http://psort.hgc.jp/).
To obtain further proof of the peroxisomal localization of ApsB, we compared its localization with the localization of the peroxisomal enzyme AcuE (acetate-malate synthase). This protein was tagged with GFP (strain TALX207-10, kindly provided by M. Hynes and A. Andrianopoulos, Melbourne, Australia) and coexpressed in a strain with mRFP1-ApsB (strain SDV70b). Fourteen percent of the spots showed green and red fluorescence (Fig. 4A). In addition, we analyzed GFP-AcuE and mRFP1-tagged HexA (SDV78c), which confirmed the localization of mRFP1-HexA to peroxisomes (Fig. 4B). However, one important difference between ApsB and HexA was the frequency of colocalization with AcuE. While HexA and AcuE had about 95% hits, ApsB and AcuE showed only 14% colocalization, indicating that ApsB was transported only to a subclass of peroxisomes. In addition, we determined the frequency of mRFP1-ApsB and GFP-HexA colocalization to 10% (Fig. 4C). Similar results were obtained with S. macrospora, where Pro40 also colocalized only partially with HexA (5). An interaction between ApsB and HexA in A. nidulans was also shown in vivo using the BiFC assay system in strains coexpressing the N-terminal half of YFP (YN) fused to hexA and the C-terminal half of YFP (YC) fused to apsB or the other way around (Fig. 4D). ApsB-HexA colocalizing spots were found in the cytoplasm and at some septa (10%). To obtain a clearer picture of the ApsB and HexA structures at septa, we used deconvolution and laser-scanning spinning-disc microscopy. ApsB appeared normally as two spots in the center of the septal pore, whereas HexA localized normally on each side of the pore (Fig. 5). In three-dimensional (3D) reconstruction pictures, the spots appeared with a longer shape along the rim of the septum. Time course experiments revealed that ApsB colocalized with the constricting ring during septation (Fig. 6). These data show that at septa ApsB does not localize to peroxisomes or the Woronin body but rather the putative MTOC is embedded in the membrane of the septal pore.
In order to test the functionality of the PTS2 sequence in ApsB, we mutated the consensus sequence (Q73L and L74R) and fused the modified ApsB protein with GFP (pDV43). The construct (GFP-ApsB_PTS2mut) was first transformed into wild-type A. nidulans strain TN02A3 (resulting in SDV77) and then introduced into strains in which either hexA was deleted (SDV79) or apsB was mutated (SDV80) or both hexA and apsB were deleted (SDV95). In strain SDV77, the localization pattern of GFP-ApsB_PTS2mut at SPBs and at cytoplasmic spots looked like that of wild-type GFP-ApsB (Fig. 7A). However, it was surprising to find that the localization of GFP-ApsB_PTS2mut at septa (Fig. 7C) did not show the normal localization of GFP-ApsB (Fig. 7B) but resembled in 70% of the cases the pattern of GFP-HexA (Fig. 7D), whereas in 22% of the cases it was similar to the GFP-ApsB localization. These localization patterns could be achieved through a piggyback import mechanism of GFP-ApsB_PTS2mut along with HexA or ApsB, which were still present as fully functional proteins in SDV77. Only in a strain lacking both ApsB and HexA (SDV95), the specific localization of the mutated ApsB protein was lost at septa (Fig. 7E). The results obtained with the last strain clearly argue for a role for peroxisomes in the transport of ApsB to septal MTOCs. It remains to be elucidated how the assembly of the MTOC at septa occurs.
To test whether the altered localization pattern of GFP-ApsB_PTS2mut at septa prevents its biological function, we analyzed if the mutated ApsB protein is able to complement the oligosporogenic phenotype produced by an apsB mutation. Therefore, we transformed ApsB [alcA(p)::GFP::apsB] and the PTS2-mutated ApsB protein (GFP-ApsB_PTS2mut) into strain AJC1.5 (apsB6). Under repressing conditions (glucose), all three strains showed brown colonies, due to the reduced numbers of spores, which is typical for apsB mutant strains (4, 32). Under inducing conditions (glycerol or sorbitol), however, wild-type ApsB protein was able to complement the oligosporogenic phenotype (spores were produced), while PTS2-mutated ApsB did not complement it (Fig. 8A and B).
As we previously described, ApsB is important for the production of MTs at sMTOCs (35). Therefore, we wanted to know if the failure of PTS2-mutated ApsB to complement the oligosporogenic phenotype was due to an inability to restore the MTOC activity at septa. In a GFP-ApsB_PTS2mut strain with GFP-labeled MTs and an apsB6 background, the number of MTs was similar to the number of MTs in apsB mutant strains (data not shown). Therefore, we assume that PTS2 of ApsB is important for its function at septa.
Next we tested if the function of PTS2-mutated ApsB can be restored by adding a PTS1 targeting sequence (SRL) at the C terminus of ApsB. We transformed PTS2-mutated ApsB with the PTS1 signal fused to the C-terminal part of the protein (GFP-ApsB_PTS2mut_SRL) expressed from the alcA promoter into strain AJC1.7 (apsB10), resulting in SNZ34, and into apsB deletion strain (SRS24), resulting in SNZ94. The apsB10 mutation converts codon 83 into a stop codon, and thus the mutant lacks most of the 1,052-amino-acid-long ApsB protein (data not shown). The transformed plasmids were integrated ectopically. Transformants of both strains (SNZ34 and SNZ94) appeared with the brown apsB mutant-like phenotype under repressing conditions (glucose) and a wild-type-like, spore-producing phenotype under inducing conditions (sorbitol). These results suggest that the ApsB-PTS1 protein was able to complement the developmental phenotype (Fig. 8).
In this paper, we show that ApsB interacts with gamma-tubulin at SPBs, at septa, at the tips of growing hyphae, and in spot-like structures in the cytoplasm. This is the first evidence for the presence of gamma-tubulin at septa and in the hyphal tip region. We had evidence before that MTOCs exist at septa, but the nature of these MTOCs remained elusive (35). Our new results show that at least two other proteins associated with nuclear MTOCs exist in septal MTOCs, GcpC and the crucial protein gamma-tubulin. These findings are in agreement with the recent localization of GcpC (40). However, it is still unclear if sMTOCs share more or all proteins with nuclear MTOCs or whether specific proteins exist only at one or the other place. The biggest unsolved question is still the anchorage of sMTOCs. Structurally, the nuclear MTOC of S. cerevisiae has been studied the best and recently similar results were obtained with A. gossypii (14, 18, 19). It is likely that the situation is similar in A. nidulans and that nuclear MTOCs are embedded in the nuclear envelope. However, structural information about sMTOCs is still missing. Our fluorescence microscopy studies indicate that the MTOC appears as two dots inside the septal rim. The structure is clearly different from that of Woronin bodies at septa. Sometimes the two ApsB dots appeared to be connected through a third small dot. This has been described before in S. pombe for the equatorial MTOCs (eMTOCs), which are also characterized by the ApsB-homologous protein mto1 (formerly named mod20 or mbo1) (10, 29). In this yeast species, MTs are generated from nuclear MTOCs, eMTOCs, and interphase MTOCs (10, 28). The importance of non-nucleus-associated MTOCs was nicely demonstrated in enucleate cells (3).
We also identified the ApsB-gamma-tubulin interaction in the tips of growing hyphae. This is also the first evidence for gamma-tubulin in the hyphal tip. In comparison, in the chytridiomycete Allomyces macrogynus, gamma-tubulin has been identified as a component of the Spitzenkörper (21). Further evidence that gamma-tubulin may be functional in the hyphal tip comes from our observation that some MTs emanate from the hyphal tip and grow into the cytoplasm (16). We speculated at the time that either MTs which did not stop growth after reaching the hyphal tip or MT fragments close to the hyphal tip could be the origin of polymerization. However, our new results point to the possibility that MTOCs exist in the apical region of the hypha.
Several lines of evidence show that the spot-like appearance of ApsB and the ApsB-gamma-tubulin interaction are due to peroxisomal localization: colocalization with AcuE and HexA and the drastic reduction of the number of cytoplasmic spots in a pexC mutant. One very strong argument is the importance of the PTS2 sequence and the rescue of the PTS2 mutation by the addition of a PTS1 sequence to the C terminus. The nonfunctionality of ApsB with a mutated PTS2 sequence could still be explained by the fact that this region appears to be evolutionarily conserved from yeast to humans (29), but the rescue of the mutation by the addition of the PTS1 sequence speaks clearly against this possibility. We envisage three possible explanations for the role of the peroxisomal localization. (i) Peroxisomes serve as hosts for sMTOCs. (ii) Peroxisomes catalyze a reaction that is required for MTOC function at the septum and is ApsB dependent. (iii) Peroxisomes serve as transport vehicles for sMTOC-associated proteins. Our results point to a transport function for peroxisomes. In agreement with such a role is the observation of fast-moving mto1 (ApsB) spots in S. pombe (29). These structures could represent peroxisomes. However, many open questions remain to be solved, e.g., how the proteins are further recruited from the peroxisomes to the sMTOCs. Against all three possibilities speaks the observation that the septal localization of ApsB and sMTOC function in mutants with defects in PTS1 or PTS2 peroxisomal protein import or in pexC mutants lacking peroxisomes appeared similar to the situation in the wild type (results not shown). However, it has to be considered that the pexC mutant strain displays pleiotropic phenotypes and that the possibility of a piggyback import mechanism might mask the possible effects of PTS1 or PTS2 defects (12).
From our results we conclude that ApsB defines a new class of peroxisomes that is—besides the Woronin bodies—the second example of peroxisomes as organelles with a function beyond metabolic pathways (30).
We are grateful to M. Hynes (University of Melbourne, Melbourne, Australia) for sending us peroxisomal marker proteins and pex mutant strains and to G. Jedd for helpful discussions and the hexA deletion strains. We thank R. Suelmann, Björn Titz, and Sabrina Hettinger for initial help with this project.
The work was funded by the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie, and the special program Lebensmittel und Gesundheit of the Landesstiftung Baden Württemberg. N.Z. was partly supported by a fellowship from the Syrian ministry.
Published ahead of print on 26 March 2010.