| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Eukaryotic Cell, June 2008, p. 1041-1052, Vol. 7, No. 6
1535-9778/08/$08.00+0 doi:10.1128/EC.00071-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Özgür Bayram,1
Silke Busch,1,
Michael Hoppert,3
Oliver Valerius,1
Stephan Seiler,1,2* and
Gerhard H. Braus1*
Molekulare Mikrobiologie und Genetik, Institut für Mikrobiologie und Genetik, Georg-August-Universität Göttingen, Grisebachstraße 8, D-37077 Göttingen, Germany,1 DFG Research Center for Molecular Physiology of the Brain (CMPB), D-37077 Göttingen, Germany;,2 Allgemeine Mikrobiologie, Institut für Mikrobiologie und Genetik, Georg-August-Universität Göttingen, Grisebachstraße 8, D-37077 Göttingen, Germany3
Received 8 March 2007/ Accepted 25 March 2008
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Important aspects of nuclear migration were elucidated by genetic analyses performed with the amenable fungal model organism Aspergillus nidulans. In a screen for nuclear migration mutant strains, the genes nudA, nudC, nudF, and nudG were discovered (60). The human homolog to nudF is Lis1, the haploinsufficiency of which causes autosomal-dominant lissencephaly type 1 (10). Mutations of Lis1 lead to retarded neuronal migration to the cerebral cortex, creating disorganized cortical layers. This defective brain development causes mental retardation, epilepsy, and early death. The characterization of LIS1-interacting proteins showed that LIS1 also plays a role in other diseases, like schizophrenia and neuronal degeneration (45). While during the nucleokinesis of neurons a single nucleus is transported through the cytoplasm, fungal growth is characterized by the distribution of multiple nuclei along filamentous hyphae. Mutations of nudF prevent the active transport of nuclei out of the spore into the growing hyphae, leading to the accumulation of nuclei near the spore remnant and only incidental nuclear motion, which leads to slow growth and excessive hyphal branching (61). In addition, nuclear migration is important for the development of asexual spores in A. nidulans, during which, after several rounds of nuclear division and the budding of specialized cells (sterigmata), uninucleate and green spores are finally released. Without NUDF, mainly anucleate sterigmata are formed, and conidia are generated only if a nucleus occasionally enters a bud (61). Thus, nuclear migration mutants of A. nidulans are easily recognized by their knobby, brownish colony appearance due to slow growth and the reduced production of colored spores. NUDF/LIS1 is a 50-kDa dimeric protein with an N-terminal 
-helical LisH dimerization motif, a coiled-coil helix, and a C-terminal WD40 domain, which is a β-propeller (29, 53).
The nudA and nudG nuclear migration mutations of A. nidulans that show the same phenotype as the nudF mutation affect genes encoding the heavy and light chains, respectively, of the microtubule (MT)-dependent motor dynein (60). NUDF is not a subunit of cytoplasmic dynein, but it regulates dynein motor function. Dynein heavy-chain mutations were isolated as extragenic suppressors of a temperature-sensitive nudF mutation (57) and subsequently were shown to affect an ATPase domain of dynein and its stem region (66). NUDF directly bound the first ATPase domain of the dynein heavy chain in a yeast two-hybrid analysis (48). Accumulated in vivo data suggest that NUDF is necessary for activating dynein's retrograde transport activity, and without NUDF, dynein accumulates at MT plus ends (63). Dynein heavy chain and NUDF were colocalized at MT plus ends and at spindle poles and influence the stability of MTs by their availability at plus ends (35). Mutations in the encoding genes reduce the frequency of MT catastrophe and rescue events, respectively, indicating that the intact dynein/NUDF complex promotes plus-end dynamics. This is further supported by the isolation of an -tubulin mutant, which suppresses the nudA and nudF defects (58). Furthermore, dynein and NUDF regulate the attachment of MT plus ends to the cell cortex; NUDF seems to interact with APSA (for anucleate primary sterigmata A), a cortical landmark protein, so that dividing nuclei attached to the MT minus ends can be brought to and fixed at specific cellular positions (56).
In addition, various other nud mutants have been described that regulate NUDF function. NUDE is a multicopy suppressor of the nudF7 mutation and has been shown to interact with NUDF's coiled-coil helix via its own N-terminal coiled-coil domain (13). Its C-terminal domain directs NUDE to MT plus ends, but NUDE also forms immobile specks along hyphae, suggesting additional cortical functions of NUDE (14). Whereas it has been shown that NUDF is recruited to MT plus ends by CLIPA and NUDE (14), it is unknown whether these proteins also are responsible for targeting NUDF to spindle poles. CLIPA belongs to the group of plus-end tracking proteins. The mammalian homolog CLIP-170 also regulates MT dynamics and mediates MT capture at cortical sites (20, 31). Furthermore, NUDF is stabilized by NUDC, and the level of NUDF is reduced in the nudC3 mutant, indicating that NUDC acts upstream of NUDF (61). In contrast to the other nud mutants, a nudC deletion in A. nidulans results in the loss of polar growth, abnormal cell walls, and the lysis of cells, which hints at additional essential functions (8).
Mammalian NUDC contains an extra N-terminal domain in addition to a conserved p23 domain. Murine NUDC interacts biochemically with LIS1 and copurifies with the dynein heavy and intermediate chains. Similarly, the colocalization of murine NUDC, LIS1, and dynein at MT-organizing centers near the nucleus indicate that NUDC exerts its function at least partly via the regulation of the dynein/LIS1 complex (38). In addition, murine NUDC was found at discrete foci at the cortical cytoskeleton. In different rat cell types, NUDC is localized at the region of the Golgi apparatus (39). Human NUDC (hNUDC) is essential for bipolar spindle formation, indicating a function in MT organization at spindle poles (64). hNUDC also is localized to the kinetochore and regulates MT attachment to chromosomes (41). During late mitosis, it is found at midzone MTs and the midbody in HeLa cells, which emphasizes its role for cytokinesis (1). Thus, NUDC plays a broad role during mitosis, which is in agreement with its high level of expression in proliferating cells. A mitotic function also can be attributed to LIS1. Together with dynein, it is localized at the cell cortex, the centrosome, and mitotic kinetochores, which also regulate spindle orientation, chromosome attachment, and the cortical tethering of astral MTs (16, 52). Unlike the case for Aspergillus, LIS1 overexpression in mammalian cells blocks mitotic progression (16), while the complete inhibition of mitotic entry by the reduction of LIS1 expression was discussed for nonneural mammalian cells (55).
In this study, we intended to identify new binding partners of the fungal LIS1 homolog NUDF. We analyzed where NUDC is localized in A. nidulans and if there are physical interactions between fungal NUDC and NUDF. Furthermore, we asked whether unknown NUDF binding proteins can be identified in this filamentous fungus in order to gain further insights into the molecular function and regulation of this nuclear migration protein.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(59) and SURE (Stratagene, Amsterdam, The Netherlands) were employed for the preparation of plasmid DNA and were grown in Luria-Bertani (LB) medium (1% tryptophan, 0.5% yeast extract, 1% NaCl) in the presence of 100 µg/ml ampicillin. The bacterial strain KS272 carrying the pKOBEG plasmid was grown in LB medium in the presence of 25 µg/ml chloramphenicol (7). Minimal medium (MM; 1% glucose, 2 mM MgSO4, 70 mM NaNO3, 7 mM KCl, 11.2 mM KH2PO4, 0.1% trace element solution [28], pH 5.5, 2% agar) and YAG medium (2% glucose, 0.5% yeast extract, 2% agar) either with or without 0.6 mM KCl were used for the growth of Aspergillus strains and were supplemented with the appropriate amounts of pyridoxine-HCl (0.1%), uridine (5 mM), and uracil (5 mM) for agar plates. For two-hybrid experiments, Saccharomyces cerevisiae strain EGY48 (MAT trp1 his3 ura3 lexAops-LEU2) harboring pRB1840 (URA3 lexAop-lacZ; 2µ; Apr) (22) was grown on selective synthetic complete (SC) medium [0.15% yeast nitrogen base, 0.55% (NH4)2SO4, 0.1% (vol/vol) 200 mM myo-inositol, 0.2% amino acid mix-4, 2% glucose, 2% agar] lacking tryptophan, uracil, histidine, and, in the case of growth assays, leucine (SC-3 and SC-4). When needed, supplements were added according to Guthrie and Fink (24).
|
Plasmid and strain constructions for TAP.
A 
nudF strain was generated for the reintegration of a tandem affinity purification (TAP)-tagged nudF version for expression from the authentic promoter. For that purpose, a deletion cassette containing 250-bp 5'-untranslated region (5' UTR) and 3' UTR fragments of nudF was generated as described by Krappmann et al. (32) using the zeo-AfpyrG-zeo deletion cassette for targeted gene replacement. A 10-kb genomic BglII fragment containing nudF was isolated from an Aspergillus bacterial artificial chromosome library by using a 640-bp probe, which was amplified using the HexaLabel DNA labeling kit (Fermentas GmbH, St. Leon-Rot, Germany) and hybridized to the 5' UTR of nudF. The genomic fragment was subcloned into BamHI-digested pBluescript (pME2822). This plasmid and the nudF deletion cassette were transformed into E. coli KS272, and the plasmid resulting from homologous recombination (pME3231) was rescued from KS272. It contained the nudF::zeo::pyrG::zeo deletion cassette with a 6-kb 5' UTR and 3-kb 3' UTR and was linearized by NotI digestion prior to transformation into A. nidulans strain AGB152, yielding strain AGB257 (nudF::zeo::AfpyrG::zeo pyrG89 pyroA4). Transformants were selected on YAG medium containing pyridoxine-HCl and 0.6 M KCl and were tested for their lack of nuclear migration by the staining of nuclei, which was performed as described previously (60), and for small-colony growth by growth tests on MM pyridoxine plates and was compared to that of the parental strain. The pyrG marker of AGB257 was rescued by being plated on 5-fluoroorotic acid (5-FOA) medium according to Krappmann et al. (32), resulting in strain AGB294.
A TAP-tagged version of nudF was reintegrated into the nudF locus of AGB257. An MluI site was created 5' of the start codon of nudF in pME2822, which was employed for the integration of the N-terminal TAP tag with optimized codons and without an internal MluI site (ntap*; pME3232). This plasmid was digested with NotI and transformed into AGB257 with selection on MM containing sorbitol, pyridoxine-HCl, uracil, uridine, and 1 mg/ml 5-FOA, resulting in strain AGB249 (nTAP*::nudF pyrG89 pyroA4). Again, transformants were controlled by the microscopic analysis of nuclear migration and by growth tests, which proved the functionality of the tagged NUDF.
Plasmid and strain constructions for bnfA deletion.
The bnfA knockout strain was generated by applying the zeo-AfpyrG-zeo deletion cassette. A 998-bp 5' UTR and a 976-bp 3' UTR fragment were PCR amplified and cloned into the EcoRV and HpaI restriction sites of pME2409, respectively (OLKH241, 5'-ATC GTG ATT CTC TAT TGG ATC TCG-3'; OLKH242, 5'-GGA GCT GGC TGG AGA TCG-3'; OLKH243, 5'-GAC TCG CGA TAA GAG TAC AC-3'; and OLKH244, 5'-AAC CTG TTT CAT TGG ACC CGT AG-3'), resulting in plasmid pME3233. The linearized plasmid was transformed into the nkuA strain A1149, and transformants were selected on YAG medium containing pyridoxine-HCl and 0.6 M KCl. The pyrG marker was rescued by being plated on 5-FOA medium, resulting in strain AGB295. Since the backcrossing of these nkuA strains was not successful, the bnfA::zeo deletion also was generated in the AGB152 strain background (AGB296), and the last one was used for localization experiments.
Plasmid and strain constructions for localization experiments. NUDC-green fluorescent protein (NUDC-GFP) was localized in an A. nidulans nudC3 strain after the transformation of pME2823 harboring a nudC::sgfp fusion into A779 (AGB241). pME2823 was created by cloning a KpnI fragment of nudC into pMCB32 (18) (OLKH115, 5'-AAG GTA CCA TGT CGG AAC AAG AAC CGT C-3'; and OLKH116, 5'-AAG GTA CCC CAA TCT TCG CAT TCG AAA AG-3').
For the colocalization of NUDF and NUDC, the encoding genes were fused to the N-terminal and C-terminal halves of enhanced YFP (eYFP), respectively, by fusion PCR and were cloned into the blunt PmeI and SwaI sites, respectively, of plasmid pME3160, which contains the bidirectional niiA/niaD promoter. For PCR, template plasmids pME3012 and pME3013 (4) and pME2822/pME2823 as well as the following primers, encoding an RSIAT linker for the nudF::neyfp fusion and an RPACKIPNDLKQKVMNH linker between NUDC and cEYFP, were used: OLKH262, 5'-ATG AGC CAA ATA TTG ACA GCT C-3'; OLKH263, 5'-CTC CTC GCC CTT GCT CAC CGT GGC GAT GGA GCG GCT GAA CAC CCG TAC AGA G-3'; OLKH264, 5'-CTC TGT ACG GGT GTT CAG CCG CTC CAT CGC CAC GGT GAG CAA GGG CGA GGA GCT G-3'; OLKH265, 5'-TTA GAT ATA GAC GTT GTG GCT GTT GTA G-3'; OLKH258, 5'-ATG TCG GAA CAA GAA CCG TC-3'; OLKH259, 5'-GTG GTT CAT GAC CTT CTG TTT CAG GTC GTT CGG GAT CTT GCA GGC CGG GCG ACC AAT CTT CGC ATT CGA AAA GTC-3'; OLKH260, 5'-CGC CCG GCC TGC AAG ATC CCG AAC GAC CTG AAA CAG AAG GTC ATG AAC CAC GCC GAC AAG CAG AAG AAC GGC ATC AAG-3'; and OLKH261, 5'-TTA CTT GTA CAG CTC GTC CAT GCC GAG AGT GAT CCC-3'. The resulting plasmid, pME3238, was transformed into A779 and XX20 to test for the complementation of their temperature-sensitive phenotypes at 42°C, yielding strains AGB302 and AGB334. AGB334 also was transformed with pME3173, expressing the nuclear marker pgpdA::mrfp::h2A (S. Krappmann, personal communication), resulting in strain AGB303. For colocalization with 
-tubulin, AGB334 was transformed with a PCR product comprising a mipA::mrfp::pgpdA::natR cassette (AGB335). For this fusion PCR, a template plasmid was generated: a (GA)5 linker, the mrfp gene, and the natR expression module were amplified with primers OLKH301 (5'-GGA ATT CGG AGC TGG TGC AGG CGC TGG AGC CGG TGC CGC CTC CTC CGA GGA CGT C-3'), OLKH302 (5'-ACC GGT CAC TGT ACA GAG CTG CGG CCG CTT AGG CGC CGG TGG AGT GG-3'), OLKH303 (5'-CCA CTC CAC CGG CGC CTA AGC GGC CGC AGC TCT GTA CAG TGA CCG GT-3'), and OLKH304 (5'-CGG GAT CCT CAG GGG CAG GGC ATG CTC-3') and were cloned into pBluescript via EcoRI/BamHI, yielding plasmid pME2984. This module (OLKH306, 5'-GGA GCT GGT GCA GGC GCT G-3'; and OLKH307, 5'-TCA GGG GCA GGG CAT GCT C-3') was fused to a 500-bp homology region that included a part of mipA (OLKH308, 5'-GCA TGA ACG ACC GAT TTC CG-3'; and OLKH310, 5'-TCC AGC GCC TGC ACC AGC TCC TAC TCC AAC TTC ATC CTT TCC-3') and a 500-bp region comprising the 3' UTR of mipA (OLKH311, 5'-GAG CAT GCC CTG CCC CTG AAT TCA GCC TAG ATA TAA TGT ACC-3'; and OLKH313, 5'-CAT TGA TAG GAA CAG GTG ACT TG-3'), which were amplified using genomic DNA by a final fusion PCR with OLKH309 (5'-GCA GAT GTC GTG GTC AAC C-3') and OLKH312 (5'-CCC AAT CCC CTG GTT GAG AC-3').
A gfp::bnfA fusion was expressed in the bnfA strain. For that purpose, an AscI/PacI fragment of the bnfA open reading frame (ORF) was cloned into pMCB17apx (OLKH246, 5'-AGG CGC GCC CGC CCC CGT CAA GTT CAC CAA G-3'; and OLKH247, 5'-CCT TAA TTA ATC ATG TTT CCT TTG CCT CAG-3'), and the resulting plasmid, pME3234, was transformed into AGB296 together with pME3173 for the expression of the nuclear marker pgpdA::mrfp::h2A, yielding strain AGB297. The same plasmids were transformed into strains XX20 and AGB294, resulting in strains AGB298 and AGB299.
In order to rule out overexpression effects, a bnfA::gfp fusion was expressed from the authentic bnfA promoter in strain AGB337. For this purpose, a nkuA strain was transformed with a PCR product comprising bnfA::gfp2-5::AfpyrG (AGB337). This module was generated by fusion PCR (OLKH342, 5'-GGT TTG TCC TCG TCT ATA CTG-3'; and OLKH344, 5'-CGT CTT TCG GCA TAT ACC AG-3') covering a 2.5-kb homology region that included the 5' UTR and the bnfA ORF (OLKH341, 5'-GAG CCT GTC GGC GTC ACG-3'; and OLKH317, 5'-TCC AGC GCC TGC ACC AGC TCC TGT TTC CTT TGC CTC AGT CTG-3'), a (GA)5 linker, the gfp2-5 gene, AfpyrG (OLKH306, 5'-GGA GCT GGT GCA GGC GCT G-3'; and OLKH314, 5'-CTG TCT GAG AGG AGG CAC TG-3'), and a 2.2-kb homology region with the 3' UTR of bnfA (OLKH343, 5'-CAG TGC CTC CTC TCA GAC AGC AGG GAG TCG ACT CGC GA-3'; and OLKH345, 5'-GAT TAC GCC TTC CCA GTC ATG-3'), for which genomic DNA and plasmid pFNO3 (62), respectively, served as templates. Furthermore, strain AGB337 harbored the mipA::mrfp::pgpdA::natR cassette, which previously was introduced into A1149 (AGB336).
To rule out overexpression effects when localizing the nudC-gfp fusion, we also transformed strain AGB336 with pME2985, which was generated by cloning a KpnI fragment with nudC into a pMCB17apx derivative (pME3322). In some transformants, this plasmid did not integrate ectopically but did so solely at the nudC locus, resulting in a gfp2-5-labeled nudC copy driven by the authentic promoter and an unlabeled copy driven by the alcA promoter from the plasmid. Therefore, this strain (AGB338) was grown in glucose medium for microscopic analysis so that no unlabeled NUDC was present in the cells.
For the localization of NUDF in the bnfA strain, nudF was cloned into pMCB17apx via AscI/PacI (OLKH272, 5'-AGG CGC GCC CAG CCA AAT ATT GAC AGC TCC-3'; and OLKH273, 5'-CCT TAA TTA ATT AGC TGA ACA CCC GTA CAG-3'), and a cfp-h2A plasmid (51a) was used to exchange h2A against a NotI/SmaI fragment with tubA (OLKH274, 5'-CTA AGA ATG CGG CCG CAG AGA AGT CAT TAG TTT GAA CG-3'; and OLKH275, 5'-GCC CCC GGG TTA GTA CTC AAC TTC CTC ACC-3'). The resulting plasmids, pME3235 and pME3236, were transformed into AGB296 and AG152 as a control, yielding strains AGB300 and AGB301.
All ORFs were amplified using Pfu (Stratagene, Amsterdam, The Netherlands), KOD Hifi DNA polymerases (Novagen/Merck Chemicals Ltd., Bad Soden am Taunus, Germany), or Phusion high-fidelity DNA polymerase (Finnzymes/New England Biolabs GmbH, Frankfurt am Main, Germany) and were sequenced on an ABI Prism 310 capillary sequencer (Applera Deutschland GmbH, Darmstadt, Germany). Fungal genomic DNA was prepared as described previously (30), and Southern analyses of all strains were carried out according to Southern (51).
TAP and protein identification by tandem mass spectrometry. TAP was performed by the modified method of Busch et al. (6). For the preparation of crude extracts, 45 g of dried vegetative mycelia of strains AGB152 and AGB249 were incubated with buffer B containing 300 mM NaCl (100 mM Tris-HCl, pH 7.2, 300 mM NaCl, 10% glycerol, 0.1% NP-40, 1 mM dithiothreitol [DTT], 14.3 mM β-mercaptoethanol) to disassociate NUDF from MTs. Crude extracts were dialyzed against 2 liters of buffer D (20 mM K-HEPES, pH 7.9, 50 mM KCl, 0.2 mM EDTA, pH 8.0, 0.5 mM DTT, 20% glycerol, 0.5 mM phenylmethylsulfonyl fluoride [PMSF], 2 mM benzamidine) for 3 h and then adjusted to 10 mM Tris-Cl, pH 8.0, 150 mM NaCl, and 0.1% NP-40. Immunoglobulin G (IgG) Sepharose 6 Fast Flow beads (200 µl; Amersham Biosciences/GE Healthcare, München, Germany) were washed with 5 ml of IPP150 (25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% NP-40) and incubated with 50 ml of crude extract on a rotating platform for 12 h. The suspension was poured into a Bio-Rad PolyPrep column (Bio-Rad Laboratories, GmbH, München, Germany), in which beads were washed twice with 10 ml of IPP300 (25 mM Tris-HCl, pH 8.0, 300 mM NaCl, 0.1% NP-40), with 10 ml of IPP150, and with 10 ml of tobacco etch virus (TEV) cleavage buffer (25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% NP-40, 0.5 mM EDTA, 1 mM DTT). The following protease inhibitors were added to buffer B, IPP300, and IPP150: 5 µl 1 M DTT, 50 µl 0.1 M PMSF, 40 µl 0.5 M benzamidine, 5 µl leupeptin (1 mg/ml), 10 µl pepstatin (1.4 mg/ml), 12 µl chymostatin (2 mg/ml), and 100 µl aprotinin (1.7 mg/ml) per 10 ml of buffer. In the sealed column, beads were incubated with 1 ml of TEV cleavage buffer containing 300 U TEV protease on a rotating platform for 4 h. Proteins cleaved from the beads were eluted into a Falcon tube, and elution was repeated with another milliliter of TEV cleavage buffer. After the addition of 6 ml calmodulin binding buffer (CBB; 25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM Mg acetate, 1 mM imidazole, 2 mM CaCl2, 10 mM β-mercaptoethanol) and 6 µl of 1 M CaCl2, the solution was incubated with 200 µl of calmodulin beads (Stratagene, Amsterdam, The Netherlands) on a rotating platform for 4 h. The calmodulin beads were equilibrated with 5 ml of CBB prior to incubation and were washed twice with 1 ml of CBB containing 0.1% NP-40 and with 1 ml of CBB containing 0.02% NP-40. Proteins were eluted with 2 ml of calmodulin elution buffer (25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.02% NP-40, 1 mM Mg acetate, 1 mM imidazole, 20 mM EGTA, 10 mM β-mercaptoethanol). The final eluate was precipitated by trichloroacetic acid (TCA), including the adjustment of aliquots to 25% TCA, incubation on ice for 30 min, centrifugation at 4°C in a Microfuge for 30 min, and two washes with cold acetone. After being dried in a speed vacuum, the pellet was resuspended in protein loading dye. After separation on a polyacrylamide gel (34), protein bands were silver stained (3) and cut out. Excised polyacrylamide gel pieces of protein bands were digested with trypsin according to Shevchenko et al. (49). Peptides then were separated by water-acetonitrile gradients on a Dionex-NAN75-15-03-C18-PM column on an ultimate-nano-HPLC system (Dionex, Idstein, Germany). Online ESI-MS/MS2 spectra were generated on an LCQ-DecaXPplus mass spectrometer (Thermo Finnigan/Thermo Scientific, Dreieich, Germany). MS2 analysis and protein diagnosis was done with the software SEQUEST/TURBOSEQUEST (Bioworks Browser 3.1; Thermo Finnigan) by using the Broad Institute public A. nidulans genome protein database (http://www.broad.mit.edu/annotation/genome/aspergillus_nidulans/Home.html).
A Western hybridization with a rabbit anti-calmodulin binding peptide antibody (Upstate/Millipore GmbH, Schwalbach, Germany) was performed with samples taken during the purification procedure, which were transferred to a nitrocellulose membrane by being electroblotted after sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and were detected by the enhanced chemiluminescence method (54).
Yeast two-hybrid analysis. Protein-protein interactions were tested by applying the interaction trap of Golemis et al. (22). In this system, plasmids pEG202 and pJG4-5 were used for the expression of ORFs encoding bait and prey proteins and for the activation of lacZ and leu2 reporter constructs to discriminate on the basis of color in the presence of 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) or to select for viability on medium lacking leucine. The complete ORF of nudF was PCR amplified from the sexual cDNA library pCNS4 (33), with specific primers containing flanking NcoI and XhoI sites, and was cloned into pEG202 (OLKH146, 5'-CAT CCA TGG ATG AG CCA AAT ATT GAC AGC TC-3'; and OLKH147, 5'-CCC CTC GAG TTA GCT GAA CAC CCG TAC AGA-3'; pME2938). nudF also was cloned into pJG4-5 as an XhoI fragment (OLKH226, 5'-CGC CTC GAG AGC CAA ATA TTG ACA GCT CC-3'; and OLKH147, 5'-CCC CTC GAG TTA GCT GAA CAC CCG TAC AGA-3'; pME3237). The nudC ORF was amplified using OLKH148 (5'-CG GAA TTC ATG TCG GAA CAA GAA CCG TC-3') and OLKH149 (5'-CG GAA TTC CTA ACC AAT CTT CGC ATT CG-3') and was cloned into pJG4-5 and pEG202, respectively, as EcoRI fragments (pME3245 and pME3246). The 1,065-bp cDNA encoding BNFA (for binding to NUDF) was PCR amplified and inserted as an EcoRI/XhoI fragment into pJG4-5 and pEG202 (OLKH194, 5'-CG GAA TTC ATG GCG ATG GCC CCC GTC-3'; and OLKH195, 5'-CCG CTC GAG TCA TGT TTC CTT TGCC TCA GTC-3'; pME3239 and pME3240). As a positive control, the first 585 bp encoding the coiled-coil region of NUDE were PCR amplified and cloned as an EcoRI/XhoI fragment into pJG4-5 (OLKH162, 5'-CG GAA TTC ATG CCT TCC GCC GAT GAG C-3'; and OLKH163, 5'-CCC CTC GAG GTT ATT GTT CCT GAG CCT CTC-3'; pME2939). As a negative control, vector pME2938 was used in combination with empty vector pJG4-5.
For the domain interaction analysis, the LisH motif/coiled-coil region of nudF (amino acids [aa] 1 to 100) was amplified with primers OLKM3 (5'-CCC CTC GAG ATG AGC CAA ATA TTG ACA GCT C-3') and OLKM15 (5'-CCC CTC GAG TTA CCA ATT GGT CGG ATC TTG G-3') and cloned into pJG4-5 and pEG202, respectively, via XhoI (pME3241 and pME3242). The region encoding the WD40 domain of nudF (aa 99 to 445) was amplified using primers OLKM5 (5'-CCC CTC GAG AAT TGG CTT CCT AAA CCG TC-3') and OLKH147 and was cloned as an XhoI fragment into both vectors (pME3243 and pME3244). The 5' part of nudC, encoding aa 1 to 28, was amplified with OLKH148 and OLKM12 (5'-CG GAA TTC TTA CTC AGC TTC TTC GGC GGC-3') and was cloned as an EcoRI fragment (pME3247 and pME3248). The region for the p23 domain (aa 29 to 116) was amplified using OLKM8 (5'-CG GAA TTC CAG GCG ACC CTC CCC TAC-3') and OLKM13 (5'-CG GAA TTC TTA GTT GAC TTT GTC AAG GTG GAT G-3') and was cloned as an EcoRI fragment (pME3249 and pME3250), and the C-terminal part (aa 117 to 199) was generated using OLKM10 (5'-CG GAA TTC CAG ATG GAG TGG TGG GC-3') and OLKM14 (5'-CG GAA TTC TTA ACC AAT CTT CGC ATT CGA AAA G-3') and also was cloned as an EcoRI fragment (pME3251 and pME3252).
All ORFs were amplified using Pfu or KOD Hifi DNA polymerases and were sequenced on an ABI Prism 310 capillary sequencer. The constructs were transformed in different combinations into strain EGY48(pRB1840).
Western hybridization experiments were performed for expression analysis using anti-hemagglutinin (Sigma-Aldrich Chemie GmbH, Munich, Germany) and anti-LexA antibodies (Santa Cruz Biotechnology, Inc., Heidelberg, Germany), respectively.
For interaction tests, overnight SC cultures were diluted to an optical density at 546 nm of 0.2 with SC medium, and 10 µl of the dilution was dropped onto SC-4 plates containing either 2% galactose-1% raffinose or 2% glucose (negative control) for growth tests. Dilutions at an optical density at 546 nm of 0.1 were spotted onto SC-3 plates containing 2% galactose-1% raffinose and 1.7 mM leucine for subsequent β-galactosidase filter assays. Plates were incubated at 30°C for 2 days. For the filter assay, a Whatman filter was placed onto the yeast plate, shock frozen in liquid nitrogen, and dried. The filter then was put onto another Whatman filter soaked with 3 ml of Z buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 0.1% X-Gal) and incubated at 30°C for 4 h.
Microscopy. Five hundred microliters of the appropriate medium was put onto coverslips, which were placed in petri dishes and inoculated with 4 x 104 spores. For growth at 42°C, coverslips were placed into 6-well plates with 5 ml of medium inoculated with 2 x 105 spores or into petri dishes with 20 ml of medium inoculated with 1 x 106 spores. After incubation at the respective temperatures, slides were mounted on glass slides with nail polish. For the staining of nuclei, 1 µl of 4',6'-diamidino-2-phenylindole (DAPI) was spotted onto the glass slide before placing the coverslip onto it. Cells were examined with a Zeiss 100 Axiovert microscope (Carl Zeiss MicroImaging GmbH, Jena, Germany), and photographs were taken using a Xillix Microimager digital camera (Xilix Technologies Corp., Richmond, Canada) and the Openlab software (Improvision, Coventry, United Kingdom).
For electron microscopy, embedding in Lowicryl K4M resin was performed as described previously (26, 46). Resin sections of about 80 nm in thickness were cut with glass knives. The sections were stained for 3 min with 3% [wt/vol] phosphotungstic acid solution, pH 7.0. Specimens were analyzed with a Philips EM 301 instrument (Philips, Hamburg, Germany) at calibrated magnifications and using IMAGO electron-sensitive films (Atomic Force F&E GmbH, Mannheim, Germany). Immunogold labeling was performed as described earlier (26).
In silico analyses. For manual annotations, we used the genome databases of CADRE (www.cadre-genomes.org.uk), TIGR (www.tigr.org/tdb/e2k1/afu1/), the Broad Institute (www.broad.mit.edu/annotation/fungi/fgi/), JGI (genome.jgi-psf.org), and the Center for Integrated Fungal Research (www.aspergillusflavus.org). Domain analyses were performed using MyHits (myhits.isb-sib.ch/cgi-bin/motif_scan) and the COILS server (www.ch.embnet.org/software/COILS_form.html). Protein alignments were carried out with NPSA (http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=npsa_clustalw.html).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
These sites were clearly identified as SPBs by colocalization with -tubulin (MIPA), which was fused to monomeric RFP (mRFP) in strain AGB335 (Fig. 2D). In this strain, a few prominent spots were observed among several fainter fluorescent spots in the cytoplasm, and the prominent spots clearly could be allocated to the mRFP-labeled SPBs. The SPBs often jerked in the cells but also were found immobilized near the cortex, in which case accurate superimposition with BiFC spots could be best achieved.
To investigate whether some of the dots observed for NUDC-GFP alone also colocalized with SPBs, we introduced a nudC::gfp fusion into a strain harboring the mipA::mrfp fusion (AGB338) (Fig. 1C). In addition, nudC was not overexpressed but was expressed from the authentic promoter, and it was fused to the more stable gfp2-5 version. In fact, NUDC-GFP spots could be colocalized with MIPA-mRFP signals even in mitotic nuclei (Fig. 1C, lower row).
BNFA is a novel protein that physically interacts with NUDF in A. nidulans. Since several aspects of NUDF localization and the regulation of its activity are still unclear, we searched for additional, unknown interaction partners of NUDF and performed TAP using an nTAP*-tagged NUDF. The use of this tag allowed two subsequent affinity purifications, the first on IgG Sepharose due to the presence of two protein A domains and a second on calmodulin resin in the presence of Ca2+ by virtue of a calmodulin binding epitope (CBP) directly N terminal to NUDF. Both tags were separated by a TEV cleavage site so that CBP-NUDF could be cleaved off the first matrix by incubation with TEV protease. Elution from the calmodulin beads was achieved through titrating out calcium by the addition of EGTA.
The fusion protein was expressed from the authentic nudF promoter by the reintegration of the nTAP*::nudF fusion (AGB249) into the nudF locus of a nudF strain (AGB257). The functionality of the fusion protein was confirmed by its complementation of the nuclear migration and growth defect of the nudF deletion (Fig. 3A). DAPI staining of nuclei showed that the even distribution of nuclei along hyphae was restored in the transformed strain AGB249, although a few nuclei accumulated in the remainder of the conidia. Nevertheless, the nearly full functionality of the fusion protein enabled this strain to grow like the parental strain AGB152 (wild-type nudF).
|
-subunit (AN4218.1; three peptides with Xcorr values of 2.3, 3.0, and 3.4). The elongation factor also was identified in other control purifications, and since it was shown to bind to calmodulin (11), its copurification in this experiment probably was not due to a specific binding to NUDF. The lower band contained the gene product of ORF AN3213.1 (only one peptide, with an Xcorr value of 2.8), which we called BNFA. BLAST analyses showed that BNFA did not bear similarities to any known protein. Therefore, manual annotation regarding the AN3213 locus and its neighboring loci (AN3211, AN3212, and AN3214) was performed using the genomes of Aspergillus fumigatus, Aspergillus oryzae, Aspergillus terreus, Aspergillus niger, and Aspergillus flavus. A BLAST search of the A. niger genome identified locus 669910-670782 on scaffold 8 (protein identity 41676). The encoded protein bears 27% identity and 53% similarity to BNFA, indicating a divergent protein even in closely related species. The syntheny of this genomic region was confirmed by the identification of genes homologous to AN3212.3 and AN3211.3, respectively, on scaffold 8 (loci 672614 to 673576 and 674136 to 674961). In the other Aspergillus species, no bnfA homolog could be detected. In A. terreus, the genes homologous to AN3212.3 and AN3211.3 were located next to each other (ATEG_07921.1 and ATEG_07922.1). In A. fumigatus, four other genes were inserted between the genes homologous to AN3212.3 and AN3214.3 (Afu4g01060-30), and in A. oryzae and A. flavus the neighboring loci were scattered along the genome. Thus, not only is BNFA a divergent protein but also the whole genomic region is highly variable. Therefore, BNFA presumably exists but still is untraceable in other organisms.
Domain analyses of BNFA indicated the presence of proline-rich regions spanning aa 11 to 85 and 238 to 297 and a C-terminal coiled-coil region between aa 302 and 350 (Fig. 4). The similar hypothetical protein of A. niger also was predicted to form a C-terminal coiled-coil region (aa 250 to 287), but it showed gaps in the proline-rich regions of BNFA. The strongest similarities between BNFA and the hypothetical A. niger protein were found in the central part located between these regions.
|
BNFA is located at A. nidulans SPBs in a NUDF-dependent manner.
We deleted bnfA in order to investigate the cellular function of this novel protein. In the nkuA strain A1149 and in AGB152, the bnfA ORF was replaced by the zeo::pyrG::zeo blaster cassette with subsequent counterselection against pyrG. However, no nuclear migration phenotype was observed. DAPI staining of nuclei in germinating hyphae demonstrated an even distribution, and in growth assays a wild-type-like growth rate and asexual development were observed in both strain backgrounds (data not shown).
A gfp2-5::bnfA fusion was expressed in the bnfA strain AGB296 to gain insight into BNFA's cellular role through localization experiments. Again, a prominent nuclear dot pattern was observed in germinating hyphae characteristic of SPB proteins (Fig. 5A). In addition, a faint fluorescence was detected throughout the nuclei, which could indicate a role for BNFA in the nucleus, but it also might be caused by the rather high expression from the alcA promoter during growth on glycerol-containing medium. As a control, nuclei were counterstained by the expression of an mrfp::h2A fusion (Fig. 5A). In several cases, we observed two dots of GFP-BNFA in close proximity to each other, indicating that the respective nuclei were undergoing mitosis, during which the SPB was duplicated for spindle formation (Fig. 5A). For clarification, a bnfA::gfp2-5 fusion was expressed from the bnfA promoter in a strain with labeled SPBs (AGB337). Microscopy revealed the same fluorescent pattern within the cells, which was weaker due to the lower expression level, but it also corresponded to the -tubulin pattern. Again, an additional faint fluorescence was observed throughout nuclei, so that an overexpression effect could be ruled out.
|
BNFA might be an integral part of the SPB recruiting NUDF to the nucleus, or BNFA might be placed at the SPB by means of NUDF. Therefore, GFP-NUDF was localized in the bnfA strain and compared to the wild-type bnfA background (AGB300 and AGB301) to discriminate between these possibilities. GFP-NUDF was localized at MT plus ends during interphase and at the poles of mitotic nuclei, indicating that BNFA is not necessary for recruiting NUDF to SPBs (data not shown). The GFP-BNFA localization then was analyzed in the absence of NUDF. The gfp2-5::bnfA and mrfp::h2A fusions were expressed in a temperature-sensitive nudF6 strain (AGB298), and fluorescence microscopy was performed after growth in glycerol-containing medium at room temperature and 42°C, respectively. At low temperature, GFP-BNFA localized to SPBs as seen before (data not shown). At 42°C, the typical nuclear migration defect could be observed with nuclei clustering in the spore, indicating the inactivity of NUDF. In this case, the GFP-BNFA signals were only rarely detected in the vicinity of nuclei, but discrete dots were dispersed throughout the hyphae (Fig. 5B). The same cytoplasmic clustering arose in the nudF deletion strain (AGB299), in which definitively no NUDF was present (data not shown). This observation showed that NUDF is required for positioning BNFA at SPBs.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The dynein regulatory complex at SPBs. In Aspergillus, NUDF was previously found at the poles of spindles of various lengths, but it was not directly observed there during interphase (35). NUDA was localized there only late during mitosis, for which NUDF was required, among others. Thus, it is not unlikely that small amounts of NUDF also are present there during interphase but could not yet be detected. Using BiFC microscopy, we obtained data indicating that NUDF is present at SPBs at different stages of the cell cycle. In addition, by means of analyzing the presence or absence of the NUDF binding protein BNFA at SPBs in the presence or absence of NUDF, a subset of NUDF could be indirectly localized there during interphase. NUDF's homolog LIS1, together with dynein, was shown to be necessary for coupling the centrosome to the nucleus in mammalian neurons and Dictyostelium (44, 52). LIS1 also was found there throughout the cell cycle, and this localization was not MT dependent in Dictyostelium, in contrast to the case for neurons. Therefore, LIS1 was suggested to be an integral centrosome component. The SPB of filamentous fungi differs from the centrosome in that it is embedded in the nuclear envelope. However, this difference does not obviously influence NUDF/LIS1 localization at this MT-organizing center. NUDF was shown to be positioned at spindle poles in the absence of dynein or MTs (35), and it seems likely that NUDE and CLIPA also are the NUDF anchors at SPBs. Mouse NUDE binds to six centrosomal proteins and is important for MT organization (17), and human Nude1 was described to be necessary for centrosomal LIS1 localization and MT nucleation (23). These results indicate an important recruiting function of NUDE and argue against NUDF/LIS being an SPB/centrosome component.
In our experiments, BNFA was seen only at SPBs and not on MTs or at their plus ends, where there is an abundant supply of NUDF. Therefore, we conclude that BNFA is recruited directly to SPBs by NUDF independently of dynein or MTs. Other anchors for BNFA seem unlikely, since the depletion of NUDF was sufficient to displace BNFA from SPBs. The expression of bnfA-gfp and mipA::mrfp in a nudF6 or nudF strain could further confirm the NUDF-dependent localization of BNFA to SPBs. Our observation also makes it unlikely that BNFA is an integral component of the SPB. At present it remains unknown if the NUDF complex is bound to the inner plate or outer plate of SPBs, but BNFA localization within the nucleus indicates binding to both sides of SPBs.
Furthermore, the data presented here indicate that fungal NUDC also is part of the dynein/NUDF complex at SPBs, which suggests that, in A. nidulans, NUDC also is involved in MT organization for nuclear migration and spindle formation. In our BiFC study, the NUDF-NUDC interaction was observed in the cytoplasm in addition to SPBs. This localization could be a functional association, or it could be due to the overexpression of the fusion proteins. Overexpression studies of NUDA and NUDE fusions, respectively, showed that in addition to the authentic comets at MT plus ends, immobile cytoplasmic specks also appeared, which were assumed to be artifacts (12). Therefore, we conclude that the SPB localization of NUDC, like that of NUDF, is authentic and due to a specific function at this location, whereas the cytoplasmic localization is an artifact; however, at the moment we cannot rule out that it also is significant.
At present, is remains unclear how NUDC is attached to the SPB or the cortex. NUDF might be at least partly responsible for NUDC localization, but the opposite could be true as well. The colocalization of hNUDC with MTs could indicate direct MT binding (42), but it also is possible that NUDC binds to integral SPB components, other dynein complex subunits like NUDE, or additional regulatory proteins. Therefore, a screening for proteins binding to NUDC or to BNFA will be a target of our future research to confirm these localization studies and to clarify the molecular function of the dynein regulatory complex at this MT-organizing center.
Different subsets of NUDF and NUDC. The colocalization of NUDF and BNFA at SPBs, but not MT plus ends, suggests the existence of different subsets of NUDF. Similarly, there might be a difference between the NUDC population at the cortex and the NUDC associated with NUDF at SPBs. At SPBs, important regulatory mechanisms perform functions. Mitotic kinases and phosphatases like NIMA, PLKA, and BIMG are located there in A. nidulans and control mitotic events through phosphorylation/dephosphorylation cascades (2, 9, 19). LIS1 is a phosphoprotein (47), and NUDF was shown to be modified (our unpublished data). This raises the possibility that NUDF is phosphorylated at SPBs and therefore associates with specific proteins, like BNFA, at this site but not at others. The described phosphorylation of hNUDC by PLK1 (65) suggests that fungal NUDC also is phosphorylated. hNUDC was shown to bind PLK1, and upon its phosphorylation by PLK1 it associates with kinetochores (41). A similar mechanism also may apply to NUDC in Aspergillus. It will be very interesting to investigate if the phosphorylation of these nuclear migration proteins is a means to control the associations with specific factors, which in turn account for the different functions of dynein during the cell cycle.
The dynein/LIS1 complex contains more proteins than previously known. The determination of protein structures and multiple interaction studies lead to a model for LIS1 interaction with Ndel1 and dynein proposed by Tarricone et al., in which LIS1 dimerizes via its LisH domain and interacts with a Nudel dimer through the coiled-coil helix (53). In addition, the LisH domains of LIS1 contact the dynein heavy-chain stems, while the WD40 domains reside near the first AAA domain of the heavy-chain heads. In this study, we showed that fungal NUDC can associate with NUDF/LIS1, too, and that it can bind to NUDF's WD40 domain, which is in agreement with the described structural model, because the WD40 domains seem accessible to additional binding partners. How BNFA is incorporated in this complex is unknown, but it might bind NUDF's coiled-coil helix via its own C-terminal helix or might also contact the WD40 domain. In our two-hybrid analyses, BNFA did not bind to NUDC, which argues against the latter possibility, but BNFA is able to dimerize and therefore might also form heterodimers with other coiled-coil proteins (our unpublished results).
Here, we showed that there is at least one additional protein that is specifically recruited by NUDF to the dynein complex at SPBs. The function of BNFA in the dynein/NUDF complex, the existence of homologous genes, and which other proteins might assume its role in other organisms remains to be elucidated. The continuous improvement of genome annotations might complete the search for yet-undiscovered proteins in other organisms. On the other hand, it cannot be excluded that BNFA remains unique to fungi. For another formerly unique SPB-associated protein, SNAD, which affects septation in A. nidulans (36), analogous proteins now can be found in other fungi. Similarly, the SPB protein APSB seems to be required for MT organization only in fungi (56). Thus, given the slight difference between the SPB of fungi and the centrosome of higher eukaryotes, there could be a difference in dynein function or MT organization at these sites that involves BNFA. It is an interesting goal for future research to decipher BNFA's molecular function and identify analogous proteins in higher eukaryotes.
|
ACKNOWLEDGMENTS |
|---|
nkuA strain, and Reinhard Fischer for pMCB17apx and the EYFP plasmids. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) through the DFG Research Center for Molecular Physiology of the Brain (CMPB) and by grants from the Volkswagen-Stiftung and the Fonds der Chemischen Industrie.
|
FOOTNOTES |
|---|
Published ahead of print on 4 April 2008.
Present address: Abteilung Gastroenterologie, Hepatologie und Endokrinologie, Medizinische Hochschule Hannover, Carl-Neuberg-Straβe 1, D-30625 Hannover, Germany.
Present address: Zentrum für Arzneimittelforschung, Entwicklung und Sicherheit (ZAFES), Institut für Klinische Pharmakologie, Klinikum der Johann Wolfgang Goethe-Universität Frankfurt, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
