A dynamic continuum of pleiomorphic tubules and vacuoles in growing hyphae of a fungus

Fungal vacuoles, like higher plant vacuoles, play an important role in controlling the composition of the cytoplasm, storage and lytic activity (Boller and Wiemken, 1986), and are generally depicted in the literature as separate, more or less spherical, bodies enclosed by a single membrane. They are usually considered to act as a repository for material and an end-point for intracellular transport, in contrast to the endoplasmic reticulum (ER), other reticula and various types of vesicles, which play an intermediary role in transport. This rather static view of vacuoles is understandable, since most of our knowledge of the structure and functions of fungal vacuoles is based on electron microscopy of thin sections of chemically fixed material or biochemical analysis involving cell fractionation (Klionsky et al., 1990). The tendency of both these approaches to fragment organelle systems that form a reticulum and represent tubules or reticula as discrete spherical structures has been widely discussed (Mersey and McCully, 1978; Hopkins et al. 1990; Wilson et al., 1990; Tooze and Hollinshead, 1992).

Recent methods used to investigate organelles using fluorescent markers in living cells, freeze substitution, and electron-opaque markers of physiological significance avoid some of these problems. Data from these techniques, mostly from animal cells, have emphasised: firstly, the dynamic nature of organelle interactions and; secondly, the capacity of a number of organelles, such as the trans-Golgi network, endosomes and some lysosomes that were formerly thought to be discrete to form extensive tubular networks throughout the cell (see Tooze and Hollinshead, 1992). In fungi the vacuole system has been examined using fluorescent markers in living cells of only a few species, mostly yeasts (Makarow, 1985; Preston et al., 1987; Roos and Slavik 1987; Weisman et al., 1987; Banta et al., 1988; Weisman and Wickner, 1988; Basrai et al., 1990). These studies demonstrate that, at least in yeast, intervacuolar transport occurs at a particular stage in the life cycle and is a precisely controlled event (Weisman and Wickner, 1988), but the mechanism for transfer is far from clear. Fluorescent trails connecting vacuoles have been variously interpreted as chains of small vesicles moving along predetermined tracks in the cytoplasm (Weisman and Wickner, 1988), or tubular connections (Klionsky et al., 1990). In higher plants a dynamic association between tubular elements and vacuoles has been shown in developing Allium guard cells (Palevitz and O’Kane, 1981; Palevitz et al., 1981). It is also becoming clear that the “large central vacuole” of higher plant cells can be accompanied by other vacuole systems of different morphology and motility, some of which are tubular; these include the pleiomorphic canalicular system of tomato hairs (McCully and Canny, 1985) and the tubular system in cultured carrot cells (Hillmer et al., 1989). Although neglected, such motile vacuolar systems appear to be widespread amongst higher plant cells of different types (McCully and Canny, 1985). These observations contrast with a prevailing view of the static nature of vacuoles and the view that transfer of materials between cellular compartments, including vacuoles, is predominently or indeed almost exclusively, via vesicles.

We have loaded the vacuoles of young living hyphae of Pisolithus tinctorius, a mycorrhizal fungus, with 6-carboxyfluorescein (CF) and monitored the changes in form and interactions of the vacuoles using fluorescence microscopy. CF has been used as a symplastic tracer in higher plants (Oparka, 1991) and is known to be compartmented by vacuoles of Egeria densa (Goodwin et al., 1990). It is considered to be membrane impermeant in dissociated state, but can be loaded into cells as the non-fluorescent diacetate, which is cleaved by intracellular esterases to form the highly fluorescent CF (Goodall and Johnson, 1982). The images from fluorescence microscopy are compared with electron micrographs of freeze-substituted cells. Freeze substitution has well-known advantages over chemical fixation in capturing the morphology of dynamic vacuole and tubule systems in plants and fungi (McCully and Canny, 1985; Howard and O’Donnell, 1987). Together, these methods reveal a motile, pleiomorphic and interactive vacuole and tubule system, which has characteristic behaviour and transports the fluorochrome between vacuole clusters situated at intervals along the growing hyphal tips. The smaller tubules are morphologically indistinguishable from tubular smooth ER (see Hepler et al., 1990). However, their accumulation of fluorochrome, their patterns of motility and behaviour and their connection with vacuoles show similarities with the endosomal and lysosomal networks recently described in animal tissue culture cells.

Pisolithus tinctorius (Pers.) Coker and Couch, strain DI-15, isolated by Grenville et al. (1986), was cultured on modified Melin-Norkrans (MMN) agar medium (Marx, 1969) at 22°C in the dark. Actively growing hyphae from cultures 1 to 3 weeks old were treated with 6-carboxyfluorescein diacetate (CFDA). A 20 μg/ml solution of CFDA (Molecular Probes, OR, USA) was made up by diluting a 1 μg/ml stock solution in acetone with reverse osmosis water (final pH 4.8). The solution was not buffered; both phosphate and zwitterionic buffers diminished cell viability, as indicated by cessation of cytoplasmic streaming. Solutions retained low fluorescence for at least 24 hours, indicating a low CFDA hydrolysis rate in the external medium. A large segment (about 4 mm × 4 mm) was cut from the growing edge of the fungal colony and floated upside-down in each solution to submerge the fungus at the agar-air interface. Optimal staining of the vacuole system was obtained with a 10-min pulse of CFDA followed by 30 min in medium without CFDA. Each mycelial segment was then placed agar-side-up on a slide, a coverslip was pressed gently on to it, and it was viewed by fluorescence microscopy. Observations were confined to terminal and penultimate cells of actively growing submerged hyphae. Any aerial hyphae and hyphae submerged more than 1 mm below the agar surface were disregarded. Cytoplasmic streaming continued and the vacuole system maintained its characteristic behaviour in the 6-carboxyfluorescein (CF)-loaded cells for a minimum of 1 h and up to 4 h during exposure to blue excitation.

Whole fungal colonies (4 replicates), cut from agar plates and floated in the CFDA solution as described above, were macerated with a mortar and pestle. Control colonies floated in distilled water, and agar soaked in either CFDA or water were similarly macerated. CF was obtained by hydrolysis of the diacetate solution. Samples of each were loaded on to thin-layer chromatography plates (Kieselgel 60; Merck, Darmstadt, GDR) and developed in n-butanol:acetic acid:pyridine:water, 15:3:10:12, by volume. The Rf values of fluorescent components in each solution were calculated following examination under UV light.

Fluorescence micrographs were taken on a Zeiss Axiophot microscope equipped with an HBO 50 mercury arc lamp, using filter combination BP450-490, FT 510 and LP 520. Photomicrographs were taken on Kodak Technical Pan film rated at 400 ISO and processed in either Technidol or HC 110 developer. To check whether CF-loading plus irradiation modified vacuole morphology or movement, CF-loaded and control (water-soaked) hyphae were compared using Nomarski differential interference contrast (DIC) optics.

The mycelium was allowed to grow over 5 mm discs of autoclaved Nuclepore brand ‘Membra-fil’ gridded membrane filter (8 mm pore size) placed on the agar surface just ahead of the growing front of the colony. The membrane filters were cut from the agar surface before the hyphal tips had reached the other side. Each filter was placed on a disc of aluminium foil attached to the end of a plunging rod, which was then slammed on to the polished surface of a liquid nitrogen-cooled copper block. The frozen hyphae were transferred to vials containing 4 ml of 2% osmium tetroxide in acetone with 3 Å molecular sieve, and substituted at –70°C for six days. Vials were returned to room temperature in stages (1 hour at -20°C, 1 hour at 4°C and 1 hour at room temperature), rinsed three times (3 × 10 min) in fresh dry acetone and infiltrated with epoxy resin (Spurr, 1969). Sections were collected on Formvar/carbon-coated copper slot grids and stained with methanolic uranyl acetate (10 min) followed by undiluted lead citrate for 20 min (Reynolds, 1963). Electron micrographs were taken with an Hitachi H-7000 transmission electron microscope at 100 kV.

Chromatograms of the CFDA-treated fungal macerate gave one yellow fluorescent spot (Rf 0.77) identical to the single spot obtained from a hydrolysed CFDA solution, indicating that the CFDA was pure and that the compartmented fluorochrome was the single compound, CF. Two pale blue fluorescent spots, with Rf values of 0.85 and 0.72, were present in both CFDA-treated and control fungal macerates, but the only fluorescence detected by microscopy of whole live cells under blue excitation was a faint orange autofluorescence in the hyphal walls. Freshly prepared CFDA solution also gave a faint spot of Rf 0.77, showing that the solution contained a small amount of hydrolysis product. The agar contained no fluorescent compounds and agar soaked in CFDA gave only the usual spot of Rf 0.77.

CF was rapidly accumulated by many of the hyphal tips. A low level of fluorescence was transiently present in the cytoplasm, and intense fluorescence rapidly appeared in an interconnected tubule and vacuolar network (Fig. 1A). This was only seen in growing hyphal tips (Fig. 1B). More mature cells of each hypha, usually including the penultimate cell, contained fewer, larger, more or less elliptical or rounded vacuoles. These were also interconnected and accumulated fluorochrome. There was no apparent effect of accumulated CF plus irradiation on either cytoplasmic streaming or the behaviour of the vacuole system in the short term, as demonstrated by comparison of CF-loaded and control hyphae under DIC optics. However, after about 3 h under blue excitation the tubules broadened and coalesced, and their movements slowed.

The labelled vacuole and tubule system consisted of clusters of more or less spherical vacuoles located at intervals along the terminal cell with an interconnecting system of tubules (Fig. 1A,B). The variation in form in different hyphal tips at the growing edge of the same colony is shown in Fig. 1; the two extremes were either clusters of vacuoles with very few associated tubules (Fig. 1A,C), or a tubular reticulum with a few small vacuoles (Fig. 1D). Most tips contained a combination of both. Tubules were of various lengths and could be either single (Fig. 1C) or branched (Fig. 1D,E). Their diameter ranged from the limit of resolution to about 0.6 μm, and they were up to 60 μm long. They were invariably oriented approximately parallel or at a small angle to the long axis of the hypha (Fig. 1C,D,E,F) and, in some cells, they formed a continuous reticulum, which was interconnected for the full length of the cell.

The terminal cell could be subdivided into zones (apical, subapical, nuclear and basal) on the basis of the distribution and behaviour of the vacuole and tubule system. There was usually at least one vacuole cluster in each zone, with the exception of the zone containing the two nuclei, which was generally free of vacuoles, but often contained single tubules (Fig. 1C) or a tubular reticulum. Vacuoles were absent from the extreme tip of the cell and those in zones anterior to the nucleus were usually small (Fig. 1A,E), though not always (Fig. 1C). The basal zone, between the nuclei and septum, was the most obviously variable, being occupied either almost exclusively by tubules, or by several large vacuoles interconnected by bridges (Fig. 1F), or occasionally both.

During the observation period vacuole clusters tended to remain in the same location although they underwent saltatory movements, changed shape and showed transient fusions within the cluster (Fig. 2A-D). This contrasted with the movement of the tubular elements, which could rapidly extend and retract over long cellular distances. Individual tubules could extend from a vacuole in one cluster (Fig. 2E,F) into an adjacent zone and fuse with another vacuole or tubule. Tubule movement was intermittent, unpredictable and multidirectional. Some tubules underwent and completed their movements within a second or two, while others remained extended in a single plane for at least one minute, usually until the fluorochrome had faded (Fig. 2G,H). Although the direction and distance varied, it appeared that the tubules were always moving along specific pathways. The nuclear zone occupied a central position; tubules traversed it in either direction, or originated from it. They moved acropetally from vacuoles in the basal zone and transiently formed connections with subapical zone vacuoles at least 40 mm away, or they moved basipetally from the subapical zone into the nuclear and basal zones. Frequently a single tubule, originating from subapical zone vacuoles or the reticulum in the nuclear zone, projected into the apical zone to within a micrometre or two of the hyphal tip (Fig. 1E).

In addition to their extension and retraction the tubules also dilated and contracted by peristaltic movement and, in this way, transported globules of fluorescent material along the tubule (Fig. 2I-L). Fluorescent content from vacuoles of the subapical zone moved along the tubular element as dilations, giving the impression that the vacuole itself was moving along the tube. Extending tubules often had dilations at their tips and these were commonly observed to separate and produce a vacuole (Figs 1D, 2A-D). Dilations travelling along a tubule either produced a vacuole when they arrived at the tip, or their content was incorporated into an existing vacuole. A pulsating tubule often suddenly underwent transformation into a string of small vacuoles (about 1 μm diam.), connected by fine fluorescent bridges (Fig. 1E,F). These were commonly strung out and moved in unison along the same tracks, in continuity with the tubules, and they ultimately grouped into vacuole clusters (Fig. 2A-D).

Tubule movement was independent of cytoplasmic streaming. The cytoplasmic streams contained very small fluorescent particles (≤0.1 μm in diameter), presumed to be vesicles that accumulate fluorochrome, moving in them. There were usually several streams. The tubules moved in various directions relative to these cytoplasmic streams and at different rates from them. Fig. 2G-H shows changes in position and retraction of the tip of a tubule that branches from a reticulum that is connected to a vacuole. Fig. 3A-D shows the sequence of changes over a short period of time in a reticulum and its continued attachment to individual vacuoles throughout. In particular, one of the tubules retracts back to the branch point. A vacuole is attached to a tubule via a narrow bridge that contains fluorescent material. From fluorescence micrographs the vacuoles (dilated region) had a diameter of roughly 1.3 mm while the bridge (narrower region) varied from about 0.7 mm in diameter to the limit of resolution. Finally, the reticulum (Fig. 3E) shows changes characteristic of long irradiations and interpreted as damage, where the tubules thicken and form new fusions as their motility ceases.

Interconnections between tubules and vacuoles were confirmed at the ultrastructural level. Serial sections through vacuole clusters showed that each vacuole was connected via narrow membrane-enclosed bridges to usually more than one adjacent vacuole, indicating that the clusters are a complex fully interconnected system (Fig. 4). The interconnecting bridges were often wider, but could be as narrow as cisternae of adjacent rough ER. Vacuoles contained a dispersed electron-opaque material, were mostly 0.4 –0.8 μm in diameter (n=39) and ranged from more or less circular to ovate, pyriform, irregular or elongate in profile. Two adjacent vacuole clusters in the subapical zone are shown in Fig. 5A,B. An elongate vacuole has extended from one of the clusters and lies parallel to the long axis of the hypha (Fig. 5A). In another section through the same region (Fig. 6A,B) an elongate vacuole lies parallel to the hyphal long axis adjacent to both these vacuole clusters, without apparently connecting with them. This elongate vacuole shows an undulating profile and some constricted regions. In other cases vacuoles and constrictions alternate (Fig. 7B) to produce structures identical in appearance to the long strings of connected vacuoles seen by fluorescence microscopy (Fig. 1F). The dilations were 0.3 - 0.8 mm in diameter and the constrictions were 40 - 150 nm (n=4) (Fig. 7B). The narrowest constrictions were of similar diameter to the lumen (39 nm, n=4) of rough ER cisternae, which also lay more or less parallel to the hyphal long axis (Fig. 6). Rough ER was most obvious in interv acuolar regions (Figs 5, 6), where it was continuous over long distances with occasional breaks in individual sections, indicating that it occurred mostly as perforate sheets (Figs 5A, 6A). Smooth tubular cisternae were widespread. They occurred in peripheral and central parts of the terminal cells in areas both with and without vacuole clusters (Figs 5, 6, 7A), in cytoplasm around the nucleus (Fig. 7C), and associated with the dolipore septum at the base of the cell. Tubular cister-nae were often found in characteristic ring-like structures (Fig. 7A,C) and they frequently extended into the extreme hyphal tips, which did not contain vacuoles (Fig. 7A). The lumen was variable in diameter, from totally constricted at some points, to 50 - 73 nm (n=14) in dilated regions. Profiles of smooth tubular cisternae were often seen adjacent and parallel to elements of the cytoskeleton, oriented approximately parallel to the long axis of the hypha. Cytoskeletal elements were also located close and parallel to the elongate vacuoles (Figs 5, 6) and were identified as both microtubules (Fig. 7B) and microfilaments (Fig. 6A,B). Groups of microtubules (Fig. 7D,E) and bundles of microfilaments (Fig. 7E,F) were seen in cross-section in transverse sections of the hyphae. Golgi bodies were seen as clusters of vesicle-like profiles, short tubules with dilated ends, and occasionally a reticulum, with contents of moderate to high electron opacity (Figs 5, 6, 7A).

The system demonstrated here in the actively growing hyphal tips differs from any previously reported vacuole system in fungi in its motility, interconnectedness and pleiomorphism. It is clearly not induced by CF-loading and UV irradiation (cf. Lee and Chen, 1988), since it can also be shown in untreated, live cells by DIC microscopy and by electron microscopy after freeze-substitution. This constantly changing continuum of tubules and vacuoles must play an important role in intracellular transport in living hyphae. It provides an alternative transport pathway to either the cytoplasmic streams or endomembrane vesicles. Indeed, the tubules appear to be ferrying material specifically between the clusters of vacuoles stationed at intervals along the hyphal tip. Fluorescent globules are moved along the tubules and transferred between these vacuoles by peristalic movements and new vacuoles are produced by pinching off such globules from the tubule tip. Tubule movement is multidirectional and so the system can accommodate the bidirectional transport observed in hyphal tips. Transfer via tubules rather than vesicles allows larger volumes to be moved at one go and, if peristalsis is involved, without a concomitant transfer of membrane. It can be predicted that the ratio of membrane to fluid transfer along the pathway will be lower than for transfer of a similar volume via a collection of vesicles, where the ratio of membrane to internal content depends on the size of the vesicle. An involvement of the cytoskeleton in tubule movement is indicated by the constant close association of cytoskeletal elements.

Although this system has not previously been demonstrated in living hyphae, circumstantial evidence indicates that it may be widespread in fungi. The fluorescent threads or “trails” observed connecting vacuoles in yeast (Weisman and Wickner, 1988) bear a strong resemblance to the tubular elements seen in Pisolithus, and the view that these are “tracks” of vesicles, manifest as fluorescent threads, may need re-interpretation. Failure to appreciate the tubular connections from electron microscopy probably arises from the use of thin sections of chemically fixed material. However, there is some evidence for the existence of tubular systems in fungi from electron microscopy. A tubular vacuole system with only a few small associated spherical vacuoles is shown in freeze-substituted hyphal tips of the basidiomycete, Sclerotium rolfsii (Robertson and Fuller, 1988). Tubular vesicular complexes (TVC 1) are described by Knauf et al. (1989) in the cytoplasm of Uromyces appen - diculatus, another basidiomycete, and consist of cisternae with an irregular luminar width of 30 - 70 nm. They were considered to be a specialised portion of rough ER, but most are smooth cisternae associated with larger structures that are circular in profile (see, for example, their figure 5). A system is reported in Erisyphe graminis (Dahmen and Hobot, 1986). Tubules in Pisolithus hyphae, identical in appearance and behaviour to those that accumulate CF, also accumulate the fluorochrome DiOC6(3), whilst the vacuoles do not. DiOC6(3) was widely considered to be a specific marker for the ER (Terasaki et al., 1984; Quader and Schnepf, 1986), but recently it has been reported to be non-specific and to stain other intracellular membranes, including Golgi and endosomes (Terasaki and Reese, 1992). Furthermore, there are other vacuolar systems with tubular elements and properties very similar to those in Pisolithus that are not considered to be synonymous with smooth ER; for example, the tubular vacuolar network that develops in Allium guard cells during differentiation (Palevitz and O’Kane, 1981; Palevitz et al., 1981). This tubular network initially arises from large spherical vacuoles, exhibits complex movements and shape changes, and is pleiomorphic. The elements are of similar dimensions to those of Pisolithus, and they enlarge to form dilations. The reticulate form is transient and as differentiation proceeds the network is transformed into larger cisternae, which eventually coalesce into a single large vacuole. Tubular networks sometimes continuous with vacuoles have also been found in cultured plant and animal cells (see for example, Hillmer et al., 1989; Cole et al., 1990; Tooze and Hollinshead, 1992). Such networks would be almost impossible to distinguish from smooth ER cisternae in electron micrographs on morphology alone. However, several tubule systems that form a reticulum in animal cells have been identified as physiologically distinct from the ER compartment by use of specific markers. Among these are the trans-Golgi reticulum, tubular endosomes and tubular lysosomes (Swanson et al., 1987; Hopkins et al., 1990; Luo and Robinson, 1992; and see Tooze and Hollinshead, 1992) The Golgi tubules can be recognised by their electron-opaque content in freeze-substituted Pisolithus hyphae, while the tubule system is electron lucent or contains electron-opaque dispersed material. It bears a strong resemblance in appearance and behaviour under the fluorescence microscope to the tubular endosomes described by Hopkins et al. (1990), and to the same system viewed in the electron microscope (Tooze and Hollinshead, 1992). Its characterisation in fungal hyphae will depend upon the use of specific electron-opaque markers, as with animal cells. Vacuoles in fungi are considered analogous to animal cell lysosomes because of their low pH and hydrolytic enzyme content (Klionsky et al., 1990) and there are many reports of the apparent endocytosis of fluorescent probes in fungi (for example, Makarow, 1985; Preston et a1., 1987; Basrai et al., 1990). The tips of growing hyphae are a very likely site for endocytosis, since fungi are heterotrophic and the highest rates of uptake occur at their tips. However, endocytosis remains unproven in fungi, because the cell wall precludes the use of large molecules as tracers and the mechanism of uptake of the fluorescent probes used is controversial (Oparka, 1991; Wright et al., 1992). It is probable that movement of the tubules is under control of the cytoskeleton. We have noted the close association between both the tubules and elongated vacuoles, and microtubules and microfilaments. The functional relationship remains to be determined, but the mechanism of transport by peristalsis may be widespread in other reticular systems in plant and animal cells, as an alternative to transfer via vesicles.

The work was supported by an Australian Research Council grant awarded to A.E.A.; D.A.O. was in receipt of an Australian Postgraduate Research Award. The authors are grateful to Suzanne Bullock for printing the plates, Bill Allaway for comments on the manuscript, Lydia Kupsky for photographic assistance, Renate Sandeman for photographic advice, and Carl Zeiss Pty Ltd. for the loan of an Axiophot microscope.