One of the most striking features of eukaryotic cells is the organization of specific functions into organelles such as nuclei, mitochondria, chloroplasts, the endoplasmic reticulum, vacuoles, peroxisomes or the Golgi apparatus. These membrane-surrounded compartments are not synthesized de novo but are bequeathed to daughter cells during cell division. The successful transmittance of organelles to daughter cells requires the growth, division and separation of these compartments and involves a complex machinery consisting of cytoskeletal components, mechanochemical motor proteins and regulatory factors. Organelles such as nuclei, which are present in most cells in a single copy, must be precisely positioned prior to cytokinesis. In many eukaryotic cells the cleavage plane for cell division is defined by the location of the nucleus prior to mitosis. Nuclear positioning is thus absolutely crucial in the unequal cell divisions that occur during development and embryogenesis. Yeast and filamentous fungi are excellent organisms for the molecular analysis of nuclear migration because of their amenability to a broad variety of powerful analytical methods unavailable in higher eukaryotes. Filamentous fungi are especially attractive models because the longitudinally elongated cells grow by apical tip extension and the organelles are often required to migrate long distances. This review describes nuclear migration in filamentous fungi, the approaches used for and the results of its molecular analysis and the projection of the results to other organisms.
Nuclear migration requires cytoskeletal components, mechanochemical motor proteins and regulatory proteins which control nuclear movement both spatially and temporally. The phenomenon of nuclear migration was well documented in basidiomycetous fungi long before the advent of molecular techniques . Basidiomycetes grow as haploids until a haploid hypha encounters a hypha of a compatible mating partner [2–4]. These hyphae fuse and nuclei from one mating partner populate the entire mycelium of the second strain. This process involves very rapid nuclear movement [5, 6]. In contrast to the haploid fungus, the dikaryotic mycelium is able to produce fruiting bodies in which nuclear fusion and subsequent meiosis and spore production occurs. Another characteristic feature of dikaryotic vegetative hyphae is a complex form of cell division involving the formation of clamp connections. This mechanism ensures that every new dikaryotic cell contains one single haploid nucleus of different origin. The two features, clamp connections and the fruiting ability of dikaryotic mycelia, were used to follow nuclear migration with a simple assay (Fig. 1) . One haploid mycelium was grown on an agar plate until the colony had reached a certain diameter. Then a compatible strain was inoculated close to the mature colony. The hyphae fused and nuclei of the freshly inoculated strain migrated through the hyphae of the other strain. Hyphal pieces sampled from around the growing colony were analyzed for clamp formation upon septation and for the ability to form fruiting bodies. These studies allowed the detection of the dikaryotic stage which indicated the presence of nuclei of different origin. The distance of the identified dikaryotic hyphae from the point of original hyphal fusion and the time between inoculation and sampling allowed the calculation of a minimal speed for nuclear migration of 10–25 μm min−1 in Coprinus lagopus. These velocities were at least 10 times faster than hyphal tip extension and suggested active transport of nuclei. In Schizophyllum commune nuclear migration was followed microscopically and velocities of up to 43 μm min−1 were found, a value which was still quite low in comparison to 670 μm min−1 reported for C. congregatus[5, 7].
In addition to nuclear migration after hyphal fusion of two compatible basidiomycetous fungi, nuclear movement and positioning is also required during vegetative growth of filamentous fungi. Polarized growth is the major form of growth in filamentous fungi, with hyphae continuously extending at their tip and septa laid down at some distance behind the tip [8–11]. Most biosynthetic capacities of a hypha require the presence of a nucleus, which illustrates the importance of nuclear migration and the even distribution of nuclei along the cell axis. For a long time it was believed that nuclear movement was passive, driven by cytoplasmic streaming. However, the early studies of Buller revealed that nuclear migration was faster than tip elongation and thus probably an energy driven process. In addition, it was observed that the nucleus closest to the growing tip remained at a defined distance behind the apex, suggesting an active process also of nuclear positioning [1, 12, 13]. Another proof for active movement was the isolation of mutants in which the mutation of a single gene blocked nuclear movement . Nuclear migration was recently investigated in genetically engineered Aspergillus nidulans strains in which nuclei were visualized by the green fluorescent protein (GFP). Nuclei followed the growing hyphal tip at the same speed as the hyphae elongated (1 μm min−1), but much higher velocities have been observed for short periods (up to 40 μm min−1). In addition, individual nuclei within a given multinucleate hyphal compartment displayed a variety of different mobilities and even traveled in opposite directions .
The recent molecular analyses of nuclear migration in yeast and filamentous fungi have provided many insights into the complex machinery required for this important cell biological process. This review will discuss recent achievements in the investigation of nuclear movement in filamentous fungi and the impact of these findings on our knowledge of organelle movement in higher eukaryotes.
2How to study nuclear migration?
2.1In vivo observations of nuclear movement
While the indirect observation of nuclear migration through mycelia of basidiomycetous fungi allowed some conclusions about the behavior of these organelles, direct analyses are rather rare, because in many species nuclei in hyphae are difficult to observe using conventional phase or Nomarski interference contrast microscopy [5, 16–21]. The fungi most commonly used as model systems in the past years, such as Neurospora crassa and A. nidulans, are not very useful for direct observations. Most methods to visualize nuclei in hyphae include staining protocols that require fixation and death of the cell. In cases where cells are not fixed prior to staining, the DNA staining chemicals may interfere with the function of the DNA and thus cannot be used to follow nuclear migration over longer periods of time. Nevertheless, even in these species sometimes nuclei are visible for some time without staining and their behavior could be followed in some parts of the hyphae [22, 23].
Recently, the expression of the GFP opened the possibility to visualize living and functionally active nuclei in hyphae of A. nidulans using fluorescence microscopy [15, 24]. With this technique it was possible to follow nuclear movement over long distances and over long time periods (http://www.blacksci.co.uk/products/elecpro/anims/mole254.gif) (Fig. 2). It was found that nuclei migrate with varying velocities after mitosis until they reach a certain place. In the elongating hyphal tip compartment containing 5–6 nuclei, the movement decreased gradually from the septum to the tip. Thus, the basal nuclei arrested first, while the apical nuclei continued to move. This suggested that regulation of nuclear mobility within a given hyphal compartment operates at the level of the individual nucleus. This behavior resembled the gradual onset of mitoses in a cell compartment. However, mitoses start at the hyphal tip and proceed to the basal nuclei close to the septum .
An open question is whether nuclear division and nuclear migration are linked. In in vivo analyses it was observed that immediately after mitosis daughter nuclei separated from each other and migrated along the direction of hyphal growth . Both nuclei migrated tipwards but the leading nucleus moved faster and further than the second daughter nucleus. However, nuclear movement was also observed in growing tip cells independent of mitosis. Furthermore, in subterminal hyphal compartments, where branching occurs, it was found that nuclei also migrated independently of a prior mitotic event and that they were able to move in opposite directions within a hypha. Last but not least, nuclei which had arrived at a certain destination in a non-growing hyphal compartment were observed for several hours and it was found that nuclei moved saltatorily in different directions over short distances but returned usually to the starting point. This movement was also independent of mitosis . The results of these in vivo analyses of nuclear migration were consistent with the observation of N.R. Morris and some co-workers who dissected the two processes several years ago with the analysis of nuclear migration in mutants defective in mitosis. They found that nuclei in a nimA mutant, which is blocked in the G2 phase of the mitotic cycle at restrictive temperature, were still able to migrate . These results suggest that mitosis-dependent nuclear movement might contribute to nuclear distribution but that mitosis-independent migration also occurs.
The integrity of the cytoskeleton is important for nuclear migration which was first demonstrated using specific inhibitors such as benomyl, nocodazole or cytochalasin. Benomyl and nocodazole are known to destabilize microtubules whereas cytochalasin B leads to disassembly of actin filaments [26, 27]. It was found that microtubule drugs affect both mitosis and nuclear migration [16, 28, 29]. Interestingly, mitochondrial movement in A. nidulans was not affected by benomyl . Application of cytochalasin B on growing A. nidulans hyphae also affected nuclear distribution . Because the actin cytoskeleton serves several important functions in a cell it remains to be demonstrated that the observed nuclear positioning phenotype was primarily due to the disruption of the actin fibres. This illustrates the general problem of how specific certain drugs are when used to analyze cellular processes.
2.3The molecular approach
One of the most powerful tools to unravel complex cell biological processes at a molecular level in lower eukaryotes is the use of genetics. Fungi like Saccharomyces cerevisiae or A. nidulans are excellent candidates for this approach because they are haploid organisms and available as single cells or produce single cells (spores) during their life cycles [31–33]. In addition, sexual reproduction in these organisms allows classical genetic methods, mapping of genes, construction of double mutants and dominance assays. Molecular biological techniques are well established, the complete genomic sequence is available for S. cerevisiae (http://genome-www.stanford.edu) and the genomes of other filamentous fungi are well on their way to being sequenced . For the analysis of nuclear migration different strategies have been followed over the years, including the non-targeted, the targeted and the direct non-targeted genetic approaches and the biochemical approach which have led to the discovery of many interesting genes (Table 1).
Table 1. Cloned and characterized genes involved in nuclear migration in filamentous fungi
Nuclear migration is a very important process in filamentous fungi. Thus, mutagenesis of crucial components has severe effects on filamentous growth . However, temperature-sensitive (ts) strains can overcome this problem. These strains will grow as wild-type at the permissive temperature and exhibit a mutant phenotype only at the restrictive temperature. This scheme can be used to screen for mutants with a temperature-sensitive growth defect. The primary phenotype would be temperature sensitivity; the nuclear distribution phenotype must be detected in a second screen. This mutagenesis approach was successfully applied by N.R. Morris who screened approximately 200 000 A. nidulans mutants to find 1000 temperature-sensitive strains of which two displayed a nuclear distribution defect (nud). The primary phenotype (ts) was subsequently used for cloning the corresponding genes. A second approach was initiated by J. Clutterbuck, who isolated A. nidulans asexual developmental mutants with defined differentiation defects . Some of these mutants were primarily affected in nuclear migration . Both strategies are very laborious because large numbers of strains have to be analyzed in two subsequent screens.
2.3.2The targeted genetic approach
The non-targeted approaches identified a number of genes which were involved in nuclear migration. To study their molecular function in detail it is necessary to analyze their physical or functional interactions with other proteins in the cell. The identification of these proteins can be achieved by a targeted genetic approach such as a suppressor analysis [37–42]. An extragenic suppressor is a second gene whose mutation rescues the defect of the primary gene of interest. Both directly interacting proteins and non-interacting suppressors can be isolated with this method. In both cases, studying the second gene may help to understand the biochemical function of the original gene product.
2.3.3The direct non-targeted genetic approach
Analysis of nuclear migration using the targeted and non-targeted genetic approaches is dependent on secondary phenotypes, such as temperature sensitivity, developmental defects, morphological abnormalities or rescue of primary mutations. These limitations bias the number of genes which can be discovered. The technology of the GFP opened the fascinating possibility of searching directly for nuclear migration mutants . GFP expressed in nuclei of A. nidulans allowed the visualization of these organelles in living hyphae . After mutagenesis of spores and formation of short germ tubes nuclear distribution could be analyzed under the microscope and germlings with an aberrant nuclear distribution pattern could be separated using a micro-manipulator. This technique allows one to use nuclear distribution as the primary phenotype and opens the field for the discovery of new mutants. The advantage of this screen is that once an interesting nuclear migration mutant is discovered, the strain can be tested under various conditions to find a secondary phenotype, e.g. heat or cold sensitivity, sensitivity against microtubule drugs, sensitivity against detergents or other supplements. These secondary phenotypes then can be used for cloning of the corresponding genes.
2.3.4The biochemical approach
Although the genetic approach is very powerful for the identification of molecular components of a system, the resolution of the exact cellular function of these components requires biochemical methods. In addition, biochemical approaches can be used to identify novel proteins which interact with already characterized ones. In the laboratory of M. Schliwa, kinesin from N. crassa was studied in detail. The protein was purified from protein extracts using a microtubule gliding assay [29, 44–46]. Subsequently, the corresponding gene was isolated and mutagenized to determine its function in vivo. Deletion of the gene resulted in a defect of vesicle transport. In addition, an effect on nuclear distribution was observed. However, it is not clear yet whether the observed nuclear clustering is a primary or a secondary effect of the mutation.
3Molecular biology of nuclear migration in filamentous fungi
Eukaryotic cells are characterized by three major intracellular networks, the tubulin, the actin and the intermediate filament networks, which serve structural and several other functions. In filamentous fungi microtubules and actin have been intensively characterized, whereas intermediate filaments have not yet been detected in fungi (Fig. 3). Microtubules are hollow, 24 nm thick tubes composed of 13 protofilaments of αβ-tubulin dimers . Several isoforms are differently expressed and perform distinct functions in a cell. In A. nidulans two α- and two β-tubulin genes have been characterized [49, 50]. Microtubules require γ-tubulin, which was originally discovered in A. nidulans, for initiation of assembly [51–54]. Microtubules are well known for their function in mitosis where they are crucial for chromosome segregation [38, 55, 56]. In addition to intranuclear microtubules, cytoplasmic microtubules emanate from the spindle pole body into the cytoplasm [13, 57]. These microtubules, known as astral microtubules, are implicated in mitotic nuclear division . After completion of mitosis, cytoplasmic microtubules are found as long filaments oriented longitudinally within fungal hyphae . Microtubule filaments continually assemble (at the plus end) and disassemble (at the minus end); the ratio between assembly and disassembly determines whether a given filament grows or shrinks. This process is coupled to the cell cycle . Prior to mitosis cytoplasmic microtubules disassemble and spindle microtubules are formed and after mitosis cytoplasmic microtubules reassemble. One important question is from where the cytoplasmic microtubules emanate, from the spindle pole bodies of the nuclei as in S. cerevisiae or additionally from other microtubule organizing centers. In the rust fungus Uromyces phaseoli microtubules have been reported to initiate also at the hyphal tip . Further evidence for a microtubule-organizing center at the growing hyphal tip came from experiments with Allomyces macrogynus. In this fungus γ-tubulin, which is crucial for the initiation of microtubule assembly, was identified as a component of the Spitzenkörper in the hyphal tip . These results suggest that the polarity of the filaments in cells of these fungi is mixed, which has important consequences for the action of the different motor proteins. The polarity of microtubules in other fungi has not been studied.
Actin forms 10 nm thick filaments that also have a plus and a minus end like microtubules. Actin has been characterized in a variety of filamentous fungi . It forms patches at the cortex of the cells with a high concentration at the hyphal tip [64, 65]. From these patches, filaments extend into the cytoplasm. The function of these structures is mainly thought to be in vesicle movement. A high concentration of vesicles at the hyphal tip is visible in the microscope and was described as the Spitzenkörper [10, 66, 67]. In addition, in some species mitochondrial movement is dependent on actin filaments [27, 30].
Intermediate filaments have not yet been reported in filamentous fungi. However, an intermediate filament-like protein has been found in yeast [68, 69]. Mutagenesis of the corresponding gene, MDM1, resulted in an aberrant distribution of nuclei and mitochondria.
Different motor proteins have been described that convert chemical into mechanical energy [55, 70–73]. They are able to migrate along filamentous tracks and if the motor protein is coupled to a cargo, e.g. a vesicle or a nucleus, the organelles are transported correspondingly (Fig. 4). Two major classes of motor proteins have been described, microtubule-dependent and actin-dependent motors. The first class comprises two subclasses, one, the dynein motor proteins which migrate along microtubules to the minus end and the kinesins which migrate either to the plus or to the minus end [74–81]. Depending on the orientation of the microtubules in the cell or the attachment to certain motors, cargoes can be transported in different directions [82–84]. Both motor proteins, dynein and kinesin, have been described in A. nidulans and N. crassa[44, 85]. Mutations in dynein result in severe nuclear distribution defects. In dynein mutants of A. nidulans nuclei undergo mitosis but they rarely migrate out of a germinating conidiospore. This leads to limitation of hyphal growth and small colonies. In N. crassa the phenotype is more complex. Besides the nuclear migration phenotype, a hyphal morphology phenotype was observed: the hyphae are curled . Besides dynein, kinesin has been implicated in nuclear migration in N. crassa. Mutation of this motor protein resulted in a clustering of nuclei. In addition, hyphal growth was impaired and the number of vesicles transported to the hyphal tip seemed to be reduced [44, 79]. The future challenges will be to assign specific functions of motor proteins to different cellular properties. The complexity of this task is illustrated by the fact that in the S. cerevisiae genome at least six kinesins and one dynein have been identified. One important step to a better understanding of the different motor activities could be the application of the GFP technology which will allow one to study motor-dependent processes in living eukaryotic cells.
The second major class of motor proteins are the actin-dependent myosins, some of which are involved in intracellular membrane trafficking [71, 86]. In A. nidulans a gene encoding a class I myosin motor was identified using PCR based on the high sequence conservation of known myosins from different species . This enzyme was localized at the tip of growing hyphae where it is essential for polarized tip extension. Deletion of the gene severely impaired hyphal growth, probably because of impaired vesicle movement along actin fibers [27, 87]. Whether or not the actin-myosin system is also involved in the movement of other organelles is not known. However, there is evidence in some fungi that mitochondrial movement depends on the actin-myosin force-generating system rather than on microtubule dynein (see below).
3.2The nud genes
Mutant analysis is a powerful tool in filamentous fungi to yield insights into complex pathways like nuclear migration . The problem in using a genetic screen for nuclear migration mutants is that disturbing nuclear movement is not necessarily very obvious without staining of the nuclei. Therefore, N.R. Morris screened for temperature-sensitive mutants and from these isolated two strains which displayed a nuclear distribution phenotype, nudA and C (nud=nuclear distribution) . These two strains were the start of a very successful analysis of nuclear migration in A. nidulans which contributed significantly to our knowledge of organelle movement in other organisms. Neither mitosis nor movement of other organelles was affected in the A. nidulans nud mutants. Spores germinated in these strains and formed a short hypha at restrictive temperature; the nuclei divided but did not migrate from the spore into the hypha. However, after prolonged incubation of the germlings at restrictive temperature some nuclei moved into the hyphae to allow slow growth of the colonies (Fig. 5). The molecular cloning of some of these genes was possible through functional complementation of the ts phenotype.
3.2.1nudC regulates nudF
The first gene cloned and analyzed was nudC. In addition to the nuclear distribution phenotype, nudC strains displayed a developmental defect. Conidiophores produced metulae but only rarely phialides and spores. The corresponding gene encodes a 22-kDa protein and although homologs were found in other eukaryotes, including rat, mouse and Drosophila, a biochemical function could not yet been assigned to the protein. However, it was found that nudC regulates the expression of another important nuclear migration gene, nudF. Interestingly, deletion of nudC was lethal and additional functions could be assigned to the NudC protein . Conidia of the ΔnudC defective strain grew spherically and lysed. The cell wall composition and thickness in this strain was different from wild-type and actin polarization was affected. These results indicated that NudC serves some function in polarized growth and cell wall deposition. The mutant phenotype could be rescued by growth on osmotically buffered medium at low temperature. The interesting question of how polarized growth and nuclear migration might be connected cannot be answered at the moment but this example demonstrates nicely that several cell biological processes share common pathways and common biochemical components.
To gain insights into the molecular function of nudC in A. nidulans, a suppressor analysis was performed in which nine nudC-independent genes were discovered . They were named sncA–sncI (=suppressor of nudC). They all rescued the nudC defects, growth of the colony, nuclear distribution in the germling, conidiophore development and regulation of the expression of nudF, to some degree. They were either dominant or semi-dominant indicating that they were gain of function mutations. A more detailed picture of their molecular function and of their interplay with NudC or other nuclear migration proteins cannot be drawn at the moment. Only the sncB gene, characterized by the dominant mutation sncB69, was analyzed in detail . The genomic sequence of the locus revealed that the gene encodes a mutant tRNALeu in which the normal 5′-CAG-3′ anticodon of leucine is changed to the proline anticodon 5′-CGG-3′. This mutation in a wild-type strain did not lead to any detectable phenotype. Since the original mutation of the nudC3 allele, which was used for the suppressor analysis, is a missense mutation of amino acid 146 that causes replacement of leucine by proline, the effect of the suppressor mutation is probably a reversion of the defect in the NudC protein. Thus, the identified suppressor was not informative for the biochemical function of the NudC protein.
Recently, homologs of NudC were discovered in Drosophila and mammals indicating the high conservation of the nuclear migration machinery. The A. nidulans NudC protein is 68% identical to the homolog from rat and 52% identical to the Drosophila protein [91, 92]. The Drosophila gene encodes a 38.5-kDa protein and serves an essential function in the cell (Warrior, personal communication). The protein was localized in the cytoplasm but a co-localization to any other structure in the cytoplasm was not possible. The nudC defect in A. nidulans was restored with a chimeric protein comprised of the N-terminus of the A. nidulans and the C-terminal part of the Drosophila protein. Whether the full Drosophila protein can restore all functions in A. nidulans is not yet clear .
The rat nudC homolog has been identified in a screen for prolactin inducible genes in T cells . T cell activation, achieved by contact to antigen presenting cells or by application of prolactin, involves a massive reorganization and polarization of the cytoskeleton which leads to a reorientation of the centrosome microtubule-organizing center, its associated microtubules and the Golgi apparatus toward the site of cell contact. The reorientation of the microtubules optimizes the microtubule-dependent vesicle transport towards the target cell. One of the induced genes was found to share high sequence similarity with nudC from A. nidulans. Complementation experiments with the rat homolog in the fungus showed that the mammalian protein could fully substitute for all biochemical functions of NudC which demonstrates the high conservation of these proteins. It will be very important to analyze how nudC is implicated in vesicle transport and if nudC is also involved in nuclear migration in rat. It is unclear at the moment what the role of nuclear migration in antigen presenting cells could be and thus the findings in A. nidulans could lead to a better understanding of the biology of T cells in mammals.
Evidence for a regulatory role of NudC came from the analysis of a second nud gene, nudF. This gene was identified as a multi-copy suppressor of nudC during attempts to clone nudC. The isolated gene also complemented the nudF nuclear migration mutant, generated by UV mutagenesis, at a high frequency and it was shown that indeed the nudF gene was cloned. The relation between NudC and NudF was further analyzed and it was found that the NudC protein posttranscriptionally regulates the expression of NudF. In nudC3 mutant strains the amount of NudF protein was significantly decreased. This explains why extra copies of NudF could rescue the nudC defect. Although co-sedimentation experiments failed to demonstrate a direct protein-protein interaction, recent approaches in mammals provided evidence for such an interaction of the NudC and the NudF homologs . The 444-amino acid NudF protein, deduced from the A. nidulans DNA sequence, was found to contain six β-transducin-like repeats (WD-40) which are found in many heterotrimeric G-proteins. Because these proteins are known to be involved in signaling processes it is very likely that nudF is involved in the regulation of nuclear migration . In addition, a region with a very high probability for a coiled-coil formation was identified. The highest sequence similarity (42% identity) was found to the protein encoded by the human Miller-Dieker lissencephaly syndrome gene (LIS-1) [96, 97]. Patients with this syndrome suffer a profound misdevelopment of the cerebral cortex caused by a defect in neuronal cell migration. Since neuronal movement requires nuclear migration, it is very likely that the primary defect of the LIS-1 mutation is indeed an impairment of nuclear mobility. The Lis1 protein was independently cloned as a subunit of the platelet-activating factor acetylhydrolase . The catalytic function of the enzyme was due to other subunits, suggesting a regulatory role for Lis1. Lis1 and the two catalytic subunits of acetylhydrolase are expressed in brain regions undergoing active proliferation and differentiation. Because proteins containing WD-40 repeats are very diverse in function and because they often form protein complexes, Sapir et al. studied the interaction of Lis1 and other cellular proteins. They identified tubulin as an interacting protein and found that Lis1 is implicated in the regulation of microtubule stability which results in a net increase of the maximum length of the microtubules. This cytoskeletal interaction might explain the link to nuclear migration. Further evidence for an involvement of Lis1 in nuclear migration came from Yu-Lee and co-workers. They found in a two-hybrid assay that the mouse Lis1 directly interacts with mouse NudC . Although a Lis1 homolog was identified in S. cerevisiae and implicated in a nuclear migration pathway, it remains to be determined whether or not the nuclear migration phenotype of mutants in A. nidulans NudF or its yeast homolog is due to a stabilization of microtubule bundles . NudF was recently discovered in Drosophila where it is essential for viability (Warrior, personal communication).
3.2.2NudA and NudG: subunits of the motor
Two genes whose sequence immediately indicated their function were nudA and nudG. Both genes encode subunits of the motor protein dynein. The nudG gene encodes a 8-kDa small subunit (N.R. Morris, personal communication) whereas nudA encodes a 492-kDa protein with very high similarity to the heavy chain of cytoplasmic dynein (Fig. 4). Because dynein is a minus-end directed microtubule-dependent motor protein implicated in different cellular processes such as organelle motility, mitosis in mammalian cells, and nuclear migration in fungi, NudA and NudG are probably directly involved in nuclear translocation . The nudA gene was cloned by functional complementation of a temperature-sensitive A. nidulans nudA mutant. Deletion of the entire coding region resulted in a phenotype similar to the original mutant. Although nuclear distribution was severely affected, strains were still viable because some nuclei were transported into the hyphae . This could be due to a functional overlap with other motor proteins which could substitute for the dynein function. In immunofluorescence experiments dynein was localized at the hyphal tip. However, the enzyme was only detectable when overexpressed. Therefore, dynein also might be present somewhere else in the cell. Despite the important role of dynein in nuclear migration, the exact interaction of the enzyme with microtubules or the nucleus remains to be determined. Some functional interactions were revealed in suppressor analyses. Willins et al. described that a mutation in α-tubulin suppressed mutations in several nuclear migration genes. The mutated tubulin led to destabilized microtubules as compared to wild-type. Interestingly, the mutation suppressed not only the temperature-sensitive dynein mutation but also a dynein null mutant as well as mutations in nudC, nudF and nudG. From this, one can assume that NudA, NudC, NudF and NudG destabilize microtubules. Additional evidence for an interaction of NudF and dynein came from an analysis of suppressors of nudF function . Because a nudF nudA double mutant had the same phenotype as each of the single mutations, it was concluded that the two genes act in the same pathway with NudF being epistatic to NudA.
Because dynein is involved in different functions in eukaryotes, two strategies were followed to better understand the biochemical function of nudA. First, extragenic suppressors (snaA–E) were isolated depending upon their ability to rescue the nudA1 phenotype . Interestingly, some sna mutants displayed a developmental phenotype similar to the phenotype of apsA and apsB mutants (see below). Molecular cloning of the corresponding sna genes is necessary to understand their biochemical function during nuclear migration and in development.
Second, a synthetic lethal screen was used. All nud mutants are severely affected in hyphal extension at restrictive temperature, but the mutations are not lethal. Even strains in which the dynein gene is deleted, grow slowly and small colonies are formed. After mutagenesis of a strain in which NudA was downregulated due to the control of an inducible promoter, strains were isolated which did not grow under these conditions but which grew normally when NudA was present (sld=synthetic lethality without dynein). From nine genes identified, six are likely to play a role in mitosis . These results clearly demonstrate the biochemical link between different processes which depend on dynamic features of the cytoskeleton.
3.3Dissection of the dynein motor complex: the ropy genes
In N. crassa, a set of mutants was isolated with a characteristic hyphal growth phenotype with curled hyphae and a slower hyphal extension rate [104, 105]. The curled hyphal growth phenotype led to the name ropy (ro). Ropy mutants are also impaired for sexual development and nuclei are clustered in hyphae rather than evenly distributed (Fig. 6). Eight strains of the same class of mutants were isolated almost 20 years later as partial suppressors of the cot-1 mutation . The latter gene encodes a serine/threonine protein kinase and is involved in polarized hyphal growth . Several of the ropy genes have been cloned by complementation in the past few years.
ro-1 encodes the 495-kDa heavy chain of cytoplasmic dynein which is 72% identical to the A. nidulans cytoplasmic heavy chain of dynein . Mutation of dynein in N. crassa causes a nuclear distribution phenotype similar to that in the A. nidulans dynein mutants and also a hyphal morphology phenotype. In addition to dynein, in N. crassa the plus end directed microtubule-dependent motor protein kinesin was studied . Because mutation of this mechanochemical enzyme resulted in clustering of nuclei, the function could be required for nuclear positioning rather than for nuclear migration itself.
The second ropy gene, cloned independently in two laboratories, was ro-4 which encodes a 42-kDa actin-related protein [42, 107]. It is 62% identical with Arp1 or centractin, a major component of the Glued/dynactin complex which is required for the activation of cytoplasmic dynein-driven vesicle movement [108, 109]. This example demonstrates that ropy mutants are good candidates to dissect the biochemical composition and function of the dynein motor complex in eukaryotes.
The high potential of ropy mutants to analyze dynein function and the relatively simple mutant screen stimulated Plamann and co-workers to initiate a large screen for ropy strains. More than 1000 ropy mutants were isolated which defined 23 complementation groups only some of which have been analyzed in detail to date .
A detailed characterization was performed for the ropy genes ro-2 and ro-3[111, 112]. The ro-2 gene encodes a novel 80-kDa protein which has a homolog of unknown function in Caenorhabditis elegans. Several domains were found, which allowed speculation about a regulatory role of the protein. One motif was a domain with similarities to zinc-binding LIM or RING domains which are thought to mediate protein-protein interactions [113, 114]. In addition, several potential binding sites for Src homolog 3 (SH3) domains were detected in the polypeptide sequence. SH3 domains are often found in proteins involved in signal cascades . This suggests that Ro2 might interact with other regulatory proteins. However, these proteins and the targets of Ro2 remained to be determined. The final action of Ro2 could be the regulation of either the motor protein or the cytoskeletal organization. Indeed, some changes in the distribution of microtubules and actin were found in ro-2 mutant strains. It is not clear yet whether or not this is a primary effect of the mutation. Complete inactivation of the gene had an additional effect in comparison to the original mutant allele. Whereas ro-2B20 strains were slightly impaired in conidia production, the null mutants were aconidial. This interesting finding is consistent with the result that the expression of ro-2 is transcriptionally regulated during development suggesting an increased requirement for the gene product during spore formation.
ro-3 was also cloned by complementation. The deduced amino acid sequence displayed homology to the largest subunit of dynactin. Dynactin is a protein complex, comprised of nine different polypeptides, and is required for cytoplasmic dynein-driven vesicle transportation along microtubules (Fig. 4). Dynactin has been found in yeast, in vertebrates and in Drosophila. Whereas deletion in yeast did not affect viability, the protein was found to be essential in higher eukaryotes [116–120]. It was shown that dynactin associates to the actin-related Arp1 (=actin-related protein) filaments and physically interacts with the intermediate chain of dynein . Ro3 thus could be involved in the mediation of an interaction between cytoplasmic dynein and the nuclei or the cortical actin cytoskeleton. Indeed, there is genetic evidence for an interaction of Ro3 with dynein .
Recently, ro-10 and ro-11 were analyzed. ro-10 encodes another subunit of the dynactin complex and is thus also directly involved in the integrity of the motor protein complex (M. Plamann, personal communication). Additional experiments are required to resolve the exact biochemical functions of dynein, the motor associated proteins and their relation to nuclei and cytoskeleton.
3.4The aps genes: regulators of nuclear positioning
In a screen for developmental mutants J. Clutterbuck isolated two strains, apsA and apsB, in which asexual development was blocked at the metula stage [35, 36]. The developmental block was due to a failure of nuclei to enter metulae and ensure proper inheritance in the conidiospores. Because of the anucleate metulae, the gene was named aps (=anucleate primary sterigmata). In addition to the developmental phenotype, Clutterbuck described a misdistribution of nuclei in mature vegetative hyphae (Fig. 7). In a recent paper it has been shown that the clusters were not permanent but were rearranged after some time . The cause of clustering seemed to be that some nuclei suddenly started moving and were arrested at a different place. Although this might explain the nuclear distribution phenotype in hyphae the question of the link between nuclear positioning and development remains. It was hypothesized that the developmental defect, the anucleate metulae, could have two reasons. First, nuclei fail to migrate into the metulae and second, nuclei migrate into metulae as in wild-type but because of an increased motility leave these cells before a septum could be formed at the basis of the cell . Interestingly, a suppressor was isolated which rescued the asporogenous aps phenotype through an effect on septation [39, 123]. In summary, nuclear movement per se seemed not to be affected in aps mutants, but the correct positioning of nuclei was. This suggested that the corresponding genes probably serve a regulatory function. The question remained at which level the regulation might occur. One hypothesis was that when nuclei migrate through a fungal hypha, they would reach a certain place which would cause two events, the disconnection from the driving force, and in addition, a fixation of the nucleus. Although these events have not been studied well, the aps proteins could be involved in the regulation of the fixation of the nuclei because in hyphal tip compartments and in germinating hyphae only slight differences of nuclear distribution to wild-type were observed. This could be explained if nuclei migrate in the hyphal tip compartment, are released from the driving force, but the fixation is differently regulated than in wild-type allowing nuclei to start moving again in older hyphae, whereas wild-type nuclei stay at their position .
The corresponding aps genes were cloned by complementation of the developmental defect and subsequently sequenced. The apsA gene encodes a 183-kDa protein with a high probability of forming a coiled-coil structure . In the middle region of the protein three 34-amino acid repeats were found, but a function could not be assigned to them. Most indicative for the role of the protein was a PH domain found in the C-terminus of ApsA. Those domains have been described in many proteins with functions in signal transduction pathways [125–127]. For ApsA, it was found that the PH domain directs the protein to the cytoplasmic membrane although the protein itself is rather hydrophilic . Immunolocalization experiments detected the protein along the hyphae at the cortex. Removal of the PH domain resulted in a cytoplasmic aggregation of the polypeptide. A similar behavior has been recently described for some other PH domain containing proteins . At the membrane those proteins might interact with other partners of a signaling cascade . Comparison of ApsA with known protein sequences revealed homology to a yeast protein, Num1p (=nuclear migration). Mutagenesis of the num1 gene resulted in a high frequency of binucleate mother cells . The Num1 protein is 213 kDa, even larger than ApsA, contains 12 repeats and also a PH domain at the C-terminus. The protein was cell cycle-dependent, localized at the cortex of the mother and not at the daughter cell . Expression was dependent on cell cycle phase and a genetic interaction with microtubule mutations suggested some interaction with the cytoskeleton. Characterization of astral microtubules in the NUM1 mutant revealed, indeed, that these structures were elongated in comparison to the wild-type and thus a stabilization of the microtubules was postulated. The identification of a Ca2+-binding site suggested a connection to other cellular processes. Surprisingly, the gene was identified a second time in a screen for stationary phase mutants . There it could be demonstrated that most of the repeats in the protein were dispensable and not required for function.
Recently, a homolog of apsA was also cloned in Podospora anserina. Mutants in this gene produced anucleate microconidia. Thus the mutant phenotypes in the distantly related fungi, A. nidulans and P. anserina, are very similar. However, a detailed characterization of the Podospora homologue is not complete yet (F. Graia, personal communication).
A second gene, apsB, of A. nidulans involved in nuclear positioning was characterized recently . It encodes a 121-kDa protein with an extended region with a high probability for a coiled-coil formation. No motifs which could suggest a certain function were found in the sequence. The protein localized in the cytoplasm and a protein-protein interaction with ApsA has not been found. These results suggested that ApsA and ApsB are two components which are probably involved in the same signaling cascade without or with only a transient contact. The identification of interacting components of ApsA and of ApsB is crucial for a better understanding of the function of the two proteins. Interestingly, no homolog of ApsB was found in S. cerevisiae.
Recently, another mutant was described in which nuclear distribution was impaired in hyphae and during conidiogenesis . The corresponding gene was named anuA and localized on chromosome VII. A molecular analysis has not yet been performed.
4Nuclear migration in budding yeast
In single celled yeasts, nuclear distribution is tightly coupled to cell division [134, 135]. At the end of the G2 phase, the nucleus migrates from a position in the center to the budding neck of the emerging daughter cell. The nuclear envelope, with the spindle pole body ahead, then extends into the growing bud cell and mitosis occurs at the budding neck. At this stage it is very important that the mitotic spindle stays properly oriented until the end of nuclear division. Interestingly, a rapid oscillation of the nucleus was observed across the bud neck during this stage . After complete separation of the two daughter nuclei, each of them returns to the center of the respective cells and cytokinesis follows (Fig. 8). Disruption of the machinery responsible for DNA distribution results in chromosome segregation at inappropriate positions and thus in a high frequency of binucleate mother cells (Fig. 8). Nevertheless, vegetative growth is not severely affected in these strains. In comparison to filamentous fungi the distance nuclei have to migrate is short and might be covered by random movement of the organelle when the cytoskeletal organization is disturbed. Another explanation for the leaky phenotype of most nuclear migration mutations could be overlapping functions of different components which may substitute for each other when one is mutated .
A detailed overview of all genes and proteins described to be involved in nuclear movement to date merits a review of its own and exceeds the scope of this review because it would not necessarily contribute much to our understanding of the process of nuclear migration in filamentous fungi. In addition, some of the work has been recently reviewed [47, 73, 138]. Nuclear migration during budding has been reported to be dependent on microtubules and actin, and the motor proteins dynein, kinesin and myosin [57, 136, 139–147]. Many other proteins have been described which are involved in microtubule integrity or the composition of the motor proteins and thus also affect nuclear distribution when mutated [120, 131, 132, 148–155]. Not surprisingly, most of the genes involved in microtubule stability or motor function have pleiotropic effects when they are non-functional. Recently, the involvement of an intermediate filament-like protein (Mdm1) in nuclear migration was also reported. These structures and their associated proteins were not only necessary for nuclear but also for mitochondrial inheritance [68, 156–159].
The unknown arrangement and connection of all components does not permit a final picture of nuclear migration in yeast at the moment. However, recent experiments applying GFP technology have nicely shown that after positioning of the nucleus at the neck, astral microtubules emanate from the spindle pole body and reach the cortex of the bud cell with their distal ends. They then make contact to dynein which may be attached to the cortex of the cell through the protein Kar9. The dynein motor protein exerts the force on the astral microtubules and thus ‘pulls in’ the daughter nucleus [160, 161]. In this model the localization of dynein at the cortex is crucial. However, the protein was found at the spindle pole body and associated with the cytoplasmic microtubules, but only possibly at the cortex . Further experiments are required to elucidate the exact organization and action of the described components.
Besides nuclear migration during the vegetative cell cycle nuclei also move during mating prior to karyogamy. The process consists of two major steps, nuclear congression and nuclear fusion. These nuclear activities have been reviewed excellently by Rose and co-workers recently [138, 162]. Microtubules emanating from the spindle pole body of each of the two nuclei and overlapping in the center are of central importance for the movement . The kinesin-like motor protein Kar3 is involved in the cross-bridging of the microtubules and thus in the catalysis of the gliding. The protein was also found at the spindle pole body where it mediates depolymerization of the microtubule bundles. Both processes contribute to the translocation of the organelle [164–166]. The motor protein dynein is not required for karyogamy [139, 142].
5Models for nuclear migration in filamentous fungi
The phenomenon of nuclear migration in filamentous fungi has been observed for many decades and although many genes and proteins have been described that are involved in the process, it is too early to develop a consistent picture of the process. Although the importance of tracks and motor proteins is clearly known, it is still not clear which cytoskeletal components are primarily involved in the translocation. Microtubules are known to be involved in nuclear migration, however it is not clear yet whether actin and intermediate filament-like structures, if they exist in fungi, also play a role. The same holds true for the motor proteins. Dynein has been shown to be the major mechanochemical enzyme but there is also evidence for the involvement of kinesin and it cannot be excluded at this point that myosins also play a role in nuclear migration and positioning. As with the incomplete knowledge of all players, the cellular arrangement and the interaction of the known components has not been finally resolved. Therefore, four hypotheses will be discussed of how nuclei could be moved in the cytoplasm (Fig. 9). The four models take into account the arrangement of only three components, the microtubules, the nucleus and the motor protein dynein and they are largely dependent on the polarization of the microtubules and the localization of dynein . It is very likely that each of the four theories is an oversimplification of the situation in the cell and that several or all models work at the same time or during certain stages of nuclear migration.
5.1The microtubule gliding model
In this model, microtubules within the cytoplasm have different polarizations because of different microtubule-organizing centers. The leading nucleus is attached to the growing tip by an additional microtubule bundle. Antiparallel filaments overlap and are connected through dynein motor proteins and action of the motor results in the gliding of two adjacent filaments. The movement of the filaments would lead to a translocation of the organelles [42, 168]. The number of dynein molecules involved in force generation would depend on the length of the overlap of the filaments and thus nuclei would not move any more once the same number of dyneins would be located on each side of the nucleus. This model would nicely account for the even spacing of nuclei in hyphae. A similar model of gliding microtubules was suggested for nuclear migration during karyogamy in yeast. After cell fusion nuclei are pulled together by the action of the minus end-directed motor protein Kar3 [138, 162, 164]. The microtubule gliding model for nuclear migration in filamentous fungi is also supported by the observation in Basidiobolus that loss of microtubules induced by a UV microbeam in the cytoplasm behind a nucleus resulted in a forward movement and loss of microtubules ahead of the nucleus in a backward movement of the organelle . However, there is no evidence that cytoplasmic microtubule bundles emanate in opposite directions from one nucleus and only some experiments suggested that additional microtubule bundles originate at the hyphal tip [61, 62]. Furthermore, the essential motor protein was found at the hyphal tip in A. nidulans and N. crassa rather than distributed along microtubules as would be postulated for this model . Nevertheless, it could well be that the dynein identified in these experiments, in which the enzyme was overexpressed, only represents one subclass of molecules which are not crucial for nuclear movement. However, the dynein localization should be accounted for, which leads to the second possible model for nuclear migration.
5.2The ‘pulling in’ model
In this scenario, dynein is located at the hyphal tip and is connected to the nuclei through long cytoplasmic microtubules. This arrangement is comparable to the situation in S. cerevisiae during mitosis, where astral microtubules reach the cortex of the bud and are involved in the pulling of the daughter nucleus into the bud cell. This ‘pulling in’ model requires that, after mitosis, cytoplasmic microtubules originate from the spindle pole bodies of the nuclei. The plus end of the growing filament would eventually reach the hyphal tip and get into contact with the motor protein. If the motor protein is connected to the cortex and starts moving towards the minus end of the filament, it results in pulling on the filament attached to the nucleus. However, most microtubules in A. nidulans do not reach the hyphal tip and it is difficult to imagine that all nuclei of a hyphal tip compartment are individually regulated with the same motor proteins at the apex . Recent studies have revealed that nuclear migration is linked to mitosis and this could be explained in model three.
5.3The mitotic movement model
As with nuclear division in S. cerevisiae, it has been observed in A. nidulans that daughter nuclei migrate apart from each other at the end of mitosis . This movement could be driven by astral microtubules emanating from the spindle pole bodies and reaching the cortex of the cell where they interact with the dynein positioned at certain places along the hypha. The attachment of the motor to the cortex could be mediated by actin. This model is supported by recent findings which link mitosis with nuclear migration . Further evidence comes from the comparison of the num1 yeast nuclear migration mutant with the apsA nuclear migration mutant of A. nidulans. In S. cerevisiae in the absence of Num1, astral microtubules are elongated and stabilized; thus the mutation prevents nuclear separation during mitosis, resulting in binucleate mother cells . This suggests that ApsA might serve similar functions in A. nidulans. Indeed the protein was localized at the cortex of the cell along the hyphae . Despite the attractive implications of this model for nuclear migration in filamentous fungi it cannot account for the long distance movement of interphase nuclei as it occurs in basidiomycetes or as it was observed in hyphae of A. nidulans[5, 15]. This movement might be explained by the fourth model.
5.4The migrating nucleus model
In this model the dynein motor protein is linked to the nuclear envelope and attached to cytoplasmic microtubules. Depending on the polarization of the filaments, a nucleus could move forward or backward. Although there is no experimental evidence for this arrangement of the components at the moment, it seems likely that it at least partly accounts for the long distance movement of nuclei in hyphae.
6Nuclear migration is linked to other cellular processes
6.1Nuclear migration and development
Most filamentous fungi are capable of forming differentiated structures in which spores are generated [32, 33, 169, 170]. As an example conidiophore development of A. nidulans will be considered [171–173]. After a period of vegetative growth some hyphae differentiate into a thick-walled foot cell which produces an aerial hypha, the stalk. After growing about 70 μm into the air, the stalk swells terminally to form a vesicle from which a layer of about 70 single cells bud off. These primary sterigmata, or metulae, give rise to a second layer of cells, the secondary sterigmata or phialides which are the sporogenous cells . Each metula produces two, sometimes three phialides which then continuously generate single-cell haploid spores, the conidia. The morphological changes are dependent on regulatory genes, which encode transcriptional activators or repressors and on structural genes which encode, for example, enzymes involved in the synthesis of the green spore pigment [175–183]. In addition to those development-specific genes, the concerted action of a set of genes is required which also serve important functions during vegetative growth [167, 184–187]. One example is the regulation of nuclear positioning. During conidiophore development a switch from multinucleate hyphal compartments in vegetative hyphae to cells with a single nucleus such as metulae, phialides and conidia occurs [124, 188]. This suggests that nuclear distribution is crucial for completion of development. If one of those cells does not receive a nucleus, development stops at that stage. In the two mutants, apsA and apsB, metulae remain anucleate and conidiophore development thus is blocked . An effect of nuclear migration mutations on developmental processes was also seen with several nud and ropy mutations in A. nidulans and N. crassa, respectively [111, 167].
In basidiomycetous fungi, nuclear migration is similarly important for spore formation when postmeiotic nuclei migrate from the basidium into the basidiospore. In Lentinus edodes it was found that the spindle pole body was located on the leading side of the migrating nucleus and microtubules extended from there through the sterigmata into the basidiospore . This suggested the involvement of these cytoskeletal components and a ‘pulling in’ mechanism for the nuclei.
6.2Nuclear migration, dikaryon formation, mating and nuclear fusion
After mating of two compatible basidiomycete strains, rapid nuclear migration is induced which leads to dikaryotization of the mycelia . The onset of this process is regulated by the B mating-type genes, which encode seven-transmembrane receptors and pheromones [2, 4, 190]. The signal cascade to the target genes which induce nuclear migration is under investigation and not clear at the moment.
In addition to the long distance movement of nuclei observed in basidiomycetes, nuclear distribution is also important for the maintenance of the dikaryotic stage during cytokinesis . Both nuclei of different mating type divide in one hyphal compartment synchronously, one in the center of the cytoplasm and one into the clamp connection [13, 192]. This rather complicated mode of nuclear division suggests a highly regulated system to ensure proper nuclear orientation before mitosis. Recently, it was found that the position of two paired nuclei within a cell is defined and changes after each division . Another example of the connection between mating and nuclear migration or nuclear positioning was reported during fruiting of the ascomycetous fungus P. anserina. Nuclei of different mating types remain separate until they fuse in the hook cells of the fruitbody. Exact nuclear positioning and recognition of nuclear identity is crucial for establishing the dikaryotic stage and subsequent nuclear fusion [194, 195]. The same probably holds true for other ascomycetous fungi such as A. nidulans. In the yeast S. cerevisiae, many mutants have been described that are blocked in nuclear fusion which results in sterility .
6.3Nuclear migration and pathogenesis
Many filamentous fungi are important pathogens which attack living plants [197–201]. After spore germination on the plant surface, the first crucial step for a successful infection is the entrance into the plant. Many fungi achieve this with a specialized structure called the appressorium. As an example the infection of rice with the rice blast fungus Magnaporthe grisea will be discussed . First, the conidium germinates and forms a short germ tube, which terminally swells and develops into the appressorium. These morphological changes are accompanied by a mitotic division of the nucleus. One of the two daughter nuclei migrates into the incipient appressorium, while the other returns to the cell of the conidium from which the mother nucleus originated . After nuclear migration a septum separates the appressorium from the germ tube and the appressorium develops a very thick cell wall which is required for resistance to the high turgor pressure build-up inside the cell . This pressure is necessary for the emergence of a penetration peg which ruptures the cuticle and enters the epidermis of the rice leaf. Growth of the fungus inside the leaf then causes the disease symptoms, chlorotic lesions which result in cell death. Although it is not known how strains would develop if nuclear migration were impaired one can speculate that disturbances of precise nuclear distribution in the early stages of development could be as severe as to prevent an infection of the plant.
7Nuclear migration in other eukaryotes
Nuclear migration and precise positioning is crucial for the establishment of polarized cells which are found in many developmental processes. During D. melanogaster embryogenesis, nuclei divide several times in the egg before they align at the cortex of the embryo. Here, cellularization takes place and the syncytial stage is replaced by uninucleate cells . The alignment of nuclei at the cortex occurs in two steps, an axial distribution followed by cortical movement. The movements are dependent on the microtubule and the actin cytoskeleton, respectively [205–207]. For the cortical, microtubule-dependent movement a microtubule gliding model was described . Nuclear positioning is also important during development of other eukaryotes such as C. elegans or mouse [148, 208–210]. Especially neuronal migration, as it is required during brain development, is dependent on nuclear movement [211, 212]. The recent findings that the mouse lis1 gene product, responsible for brain misdevelopment when mutated, interacts with the homolog of the A. nidulans nuclear migration protein NudC, illustrates the importance of the nucleus-translocating machinery for this differentiation .
In several algal species nuclear movement has been reported to be of great importance. In the green alga Spirogyra crassa the position of the nucleus defines the cleavage plane of the cell. After mitosis the daughter nuclei migrate back to a central position in each cell . The positioning of the nucleus within the cell depends on microtubules and actin with distinct functions for both cytoskeletal components . In higher plants, nuclear movement is especially obvious during pollen tube formation and growth in angiosperms . These examples demonstrate that the understanding of intracellular nuclear transport is not only a fundamental cell biological problem, but a key to development.
In contrast to the movement of nuclei to establish polarized cells, the function of nuclear movement in other cells is less obvious. In the alga Pleurenterium tumidum, which consists of two half-cells, the nucleus leaves its central position at the end of cell development and moves towards the cortical cytoplasm of the isthmus area. There, it performs a circular motion along the cell wall ring of the isthmus independently from other cell organelles and cytoplasmic streaming. The velocity of this movement has been determined to be 1.8–4.8 μm min−1. The nucleus turns at least 12 and sometimes up to 70 times and may change its direction of motion several times. To enter the next mitosis the nucleus migrates centripetally towards the cell center along a track of microtubules. However, there is also evidence that movement of the nucleus in this system requires actin . The challenge will be to determine the biological meaning of this interesting movement and to identify the regulating circuits which link nuclear migration to the development of these algae.
8Movement of other organelles
Besides nuclei, other eukaryotic organelles have to be distributed during cell division to ensure proper inheritance. One very important compartment for every cell is the mitochondrion . In fungi they form an interconnected tubular system and thus inheritance combines mitochondrial movement and mitochondrial morphology . Although the mode of inheritance has been predicted theoretically, the knowledge of the molecular biology of the distribution is very poor . In yeast, several genes were discovered whose function is important for transmittance of mitochondria [68, 156, 158, 159, 221]. It appears that the movement is dependent on microtubules, actin and intermediate filaments and the involvement of the different cytoskeletal systems is apparently different in different fungal species. The molecular data in this field are quite limited and a consistent picture cannot be drawn at the moment.
The knowledge of movement and the dynamics of other organellar systems, such as the Golgi apparatus, the endoplasmic reticulum, vacuoles etc., in eukaryotic cells is even more limited. However, it seems likely that processes similar to those involved in nuclear migration will be detected. Dynein, for example, has been shown to be important for the translocation of the Golgi apparatus as well as for the formation of the endoplasmic reticulum networks [222, 223]. The Golgi apparatus becomes fragmented during mitosis and distributed employing astral and mitotic spindles . After mitosis the fragments reassemble to the intact organelle. For vacuole inheritance the actin system seems to be more important and the force generating enzyme is a myosin .
9Conclusions and perspectives
The analysis of nuclear movement in eukaryotic cells revealed insights into an interesting machinery driving organelles. The small number of molecular components discovered to date allows us to imagine that a concerted action of many different proteins is necessary for the translocation process and that this process is tightly linked to several other cellular phenomena. The identification of nudC and nudF in Drosophila, rat and human show nicely that most of the results from investigations in lower eukaryotes can be transferred to higher eukaryotes and provide first insights into the functions of the corresponding genes in those organisms. However, for a full understanding of nuclear migration, research in many different organisms is required. Studies in yeast and filamentous fungi certainly will unravel many basic features but the fine tuning of the movements and the coordination with other processes will be often species specific. One of the most fascinating areas will be the investigation of the regulation of nuclear movement and the coupling to other cellular features, such as mitochondrial movement, vacuole inheritance but also nuclear fusion, development, or pathogenicity. The recent use of many different strategies for the investigation of organelle movement and the rapid increase of the number of identified components indicate a very fruitful and fascinating field of research.
I would like to thank N.R. Morris (Piscataway, NJ, USA), M. Plamann (Kansas City, MO, USA), R. Warrior (Los Angeles, CA, USA), R. Debuchy (Paris, France) and F. Graia (Paris, France) for providing results before they were published and M. Plamann and M.D. Rose (Princeton, NJ, USA) for original pictures used in Figures 6 and 8. I am grateful to the students of my laboratory, L.S. Robertson (Cambridge, MA, USA) and E. Kothe (Jena, Germany) for critically reviewing the manuscript. The research of my laboratory was supported by the DFG, the Philipps-University and the Max-Planck-Institute of Marburg.