Peculiar ultrastructural characteristics of fungal cells and of other elements apposed to and in vessel walls in plants of a susceptible carnation cultivar, infected with Fusarium oxysporum f.sp. dianthi race 2

B. Ouellette, Guillemond; P. Baayen, Robert; Rioux, Danny; Simard, Marie

doi:https://doi.org/10.7202/010905ar

Introduction

Of the material extrinsic to mature vessel elements in many plants affected by fungal wilt diseases, opaque matter (the coating) covering vessel walls is found in most of these diseases (Bishop and Cooper 1983, 1984; Pegg et al. 1976; Robb et al. 1975, 1979, 1987; Shi et al. 1992), and considered by many to be compounds of host origin and often as an expression of resistance. Investigations of wilt of carnation (Dianthus caryophyllus L.) caused by Fusarium oxysporum Schlechtend.:Fr. f.sp. dianthi (Prill. & Delacr.) Snyder & Hans. have revealed marked differences in tissue invasion and reactions between the resistant and the susceptible plants (Ouellette and Baayen 2000; Ouellette et al. 1999a, 2002, 2004a). The coating in both types of plants was associated with fine filamentous-like structures (Nicole et al. 1994; Ouellette et al. 1999a, 2002), related to vessel wall alterations, and did not appear per se to be a factor of resistance. The lack of or at least the less extensive wall modifications in the susceptible plants compared with the resistant plants, was congruent with profuse host wall colonization, generally by means of microhyphae (Ouellette et al. 1999a). This led to tissue collapse and leaf withering (Baayen and Elgersma 1985; Baayen et al. 1988).

In elm trees infected with Ophiostoma novo-ulmi Brasier, the coating is generally present at all stages of infection and associated with alterations of vessel walls (Ouellette et al. 2004b). Prominent networks (denoted alveolar) of layers, delimited by opaque bands, are also closely linked to the coating (Ibid). To our knowledge, the question concerning the origin and nature of coating in this and other wilt-affected plants has not yet been resolved adequately, despite the fact that some authors have stated that coatings may contain compounds, assigned a host origin, such as suberin, pectin, phenolics, and so on (Bishop and Cooper 1983, 1984; Gold and Robb 1995; Robb et al. 1978; Shi et al. 1992; Street et al. 1986). However, with gold-complexed probes, pectin, cellulose and chitin residues were found to be of exceptional occurrence in the coating in the diseases we have studied (Ouellette et al. 1999a, 2002, 2004a, 2004b; Rioux et al. 1998). Most histopathological studies of fungal wilt diseases concerned annual plants; generally, these provided little information about the advanced infection stages, during which coating material in particular could have become modified, and host walls, including those of vessels, increasingly altered. In this respect, the possible primary contribution of the pathogen in coating formation has been overlooked in the models proposed.

It is thus of paramount importance to discriminate between compounds of pathogen vs. host origin and to determine the various possible modes of pathogen development and to which extent its elements may withstand modifications in their walls without losing all their inherent metabolic activities. For example, O. novo-ulmi growing in its hosts as well as on various substrates, including growth through Millipore membranes (Ouellette et al. 1995), may produce various irregular growth forms and subsequently revert to so-called typical growth forms. Similar irregular growth forms were reported for Fusarium infection in carnation and tomato (Charest et al. 2004; Ouellette et al. 2001).

The present work concerns the occurrence and characteristics of various unreported elements on vessel walls in susceptible carnation infected with Fusarium, mostly in advanced stages of infection (early stages of infection have been described in the references referred to above). The association of these elements with vessel wall alterations and their links with more typical fungal forms have been noticed. Likewise, attention has been given to fungal wall modifications occurring in these cells. Many of the observations are based on cytochemical tests for cellulose and chitin. Other peculiar structures related to vessel wall alterations are reported in Ouellette et al. (2004c).

Materials and methods

Material

Rooted cuttings of susceptible carnation plants (cv. Early Sam) were greenhouse-grown for 3-4 wk before inoculation with an isolate of Fusarium oxysporum f.sp. dianthi race 2, as previously described (Baayen et al. 1996). Briefly, inoculation was performed by placing a droplet of a conidial suspension (10⁶to 10⁷ conidia mL^-1) in the axil of one of the lower leaves of the plants (which were then from 25-35 cm in height) and an incision was made across the droplet (Baayen and Schrama 1990). Inoculum was prepared from slant cultures of isolate WCS816. Four inoculated, two water-injected, and two intact plants were sampled (past the first cm above the incision), and sampled from 2 to 37 d after inoculation, with the last samples being taken when the plants showed external disease symptoms.

TEM observations

Samples (2 mm x 2 mm) were fixed in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide, dehydrated, and embedded in Epon 812, as described previously (Ouellette et al. 1999a). Ultrathin sections (straw color), cut with a Reichert Ultracut II were contrasted with uranyl acetate and lead citrate before examination at 80 kV with a Philips 300 electron microscope. At least two samples from each sampling date and ordinarily five sections from each sample were examined.

Cytochemical tests

Some of the sections were processed for cytochemical labelling (Ouellette et al. 1995). Cellulose was detected using an exoglucanase (EC 3.2.1.91), having affinity for ß-(1-4)-D-glucans; the enzyme directly conjugated to colloidal gold, as previously described (Ouellette et al. 1995), using a colloidal gold particle size of 15 nm (Frens 1973). Briefly, ultrathin sections were floated on a drop of phosphate buffer saline (PBS)-PEG (20 000, 0.02%) at pH 6.0 and incubated for 30 min on a drop of the enzyme complex. Specificity of the labelling was verified either by incubating sections over a drop of the exoglucanase-gold complex that was previously adsorbed with ß-(1-4)-D-glucans from barley (Sigma Chemical Co., St. Louis, MO, U.S.A.), or over the non-complexed enzyme followed by incubation with the complexed enzyme.

Chitin was detected using a gold-complexed wheat germ agglutinin (WGA) as described by Ouellette et al. (1999a). Sections on nickel grids were incubated on 25 µg of WGA per mL of PBS followed by exposure to an ovomucoid-gold complex (pH 4.95 in PBS-PEG). Control tests included omitting the lectin in the procedure or neutralizing it with N-acetylchitotriose. Both compounds were from Sigma Chemical Co.

Some illustrations are from samples tested with gold-complexed monoclonal antibodies to pectin, giving only sparse labelling. These microphotographs have thus been used here to illustrate other features.

Results

Irregularities in fungal walls

Fungal cells that closely contacted host walls or were included therein often had aberrant walls, as shown by marked differences in labelling for chitin. Whereas labelling was associated with lucent layers of septa or of walls facing vessel lumina, it was absent or linked to the outside opaque layer touching the host wall (Fig. 1a, b). In fact, some fungal cells were wholly or in part solely delimited by an opaque layer (Fig. 1c, d) and the host wall-fungal cell interface in other instances was marked by only a membrane-like structure (Fig. 1b). Marked differences in wall structure even occurred in the same cell (Fig. 1a, e); some walls also showed abrupt discontinuities in, or transfor-mation of, both their wall layers mostly again when these contacted host cell walls or content (Fig. 1c, f, g). In these cases, labelling for chitin was mainly associated with opaque matter, and walls of adjoining fungal cells also often appeared to be fused or separated by outside opaque layers (Figs. 1g, 2a). The plasma membrane in many of these cells was seemingly also indistinct in the parts showing altered walls but was distinct in the other parts delimited by more typical walls (Figs. 1a-c, g, 2a). Points of membrane discontinuities in these cases were marked by loops (Figs. 1g, 2a).

Intramural cells were generally well confined by host wall structures and connected to small elements that were delimited by solely a membrane-like structure (Fig. 2b). Arrays of fine, parallel-oriented filamentous-like structures frequently occurred perpendicularly across walls of these fungal cells, reaching into host walls and in continuity with those surrounding the elements (Fig. 2c). A similar associa-tion of filamentous structures, often in large masses, and fungal cell walls also occurred in cells present in vessel elements or parenchyma cells (Fig. 3a-c) that likewise pervaded the host walls they contacted. Similar fine structures pervading vessel walls also occurred in the compact coating, appearing to be structurally similar to those other masses (Fig. 3d); paired opaque bands associated with these structures also occurred in this coating.

Elements apposed to vessel walls

Compact coating of varying thickness occurred in most vessel elements at all stages of infection, albeit predominantly in early infection (Figs. 3d, 4a). In samples collected 15 d after inoculation and later (however, sampling between d 8 and 14 was lacking), when host walls were profusely invaded and degraded by the fungus, other types of thin elements equally occurred for appreciable distances over vessel walls or extended into the vessel lumen, intermixed or not with other types of components (Fig. 4b-i). These elements were delimited by paired, opaque bands that were frequently of varying thickness and unequally spaced, the thinnest ones appearing as simple opaque lines (Fig. 4b, d-g). These bands, also associated with fine opaque material of filamentous appearance, were often contorted (Fig. 4b, f, i), with the band apposed to the vessel wall contouring pit borders (Fig. 4d, g). The elements had sparse content in the form of opaque or vesicular-like bodies and fine matter (Fig. 4b, g). Similar structures were connected to fungal cells (Fig. 4h) whereas the element delineating bands and fungal walls were of the same opacity and texture (Fig. 4d-f). For convenience, the peculiar elements just described will be referred to hereafter as IEs.

Labelling for cellulose and chitin

In labelling tests for cellulose, fungal walls, IEs, and the compact opaque coating were not labelled except when these elements impinged upon host walls (Fig. 4c, h; see also Ouellette et al. 2004a). The compact, opaque coating, even in intimate contact with labelled fungal cells, was also free of labelling for chitin, except over areas that contained lucent matter (not illustrated), or where it likewise possibly impinged upon the labelled vessel secondary walls.

The IEs, in comparison, generally labelled strongly and exclusively for chitin over their whole width (Fig. 5a, b), which was considered most meaningful when these IEs bordered unlabelled pit membranes or primary walls or the generally slightly labelled secondary walls and thickenings of small vessel elements (see below). Even when they were affixed to strongly labelled vessel secondary walls, the IEs were separated from them by their generally unlabelled delimiting bands (Fig. 5a). The opaque coating, not delimited by bands, did not label, unless it had encompassed labelled vessel wall material (Fig. 5b). Moreover, IEs were labelled only when they contained membranous or vesicular-like structures (Figs. 5a, b, 6a, b). Generally, this labelling was not clearly associated with a distinct lucent layer, as for typical fungal cells, but with more diffuse lucent or even opaque matter (Fig. 6b). Occasionally, labelling was restricted to a lucent band present across the IE (Fig. 6a). The apparently empty, tiny elements (as those illustrated in figure 4) were also labelled (Fig. 6b), but they were elsewhere connected with larger portions with visible content (Fig. 6a). Thin intramural microhyphae and larger but irregular cells in vessel lumina, which displayed only thin and apparently discontinuous lucent layers, labelled in the same manner as the IEs (Fig. 6b).

Peculiarities of labelling for chitin

Although few gold particles of the probe occurred in control tests (comparable, for example, to those present on walls of some small vessel elements, as in Fig. 7e), the number of particles in some vessel lumina was relatively high. This situation mainly occurred next to strongly labelled IEs that occurred on or in unlabelled primary walls and pit membranes (Fig. 7a-g). Single or superimposed IEs opened up at places on the vessel lumen (Fig. 7a, b; compare with Fig. 6a). Gold particles did not attach to free opaque matter in these cases, except when it was closely associated with IEs, membranous structures, or lucent bands (Fig. 7e, A inset). High numbers of particles were also noticeable in elements extending across vessel lumina, and, once more, the element-limiting, firm bands, and their extensions along or contouring vessel walls were free of labelling (Fig. 7d, f, g). Networks of meandering opaque bands, parts of or connected to the IE or connected to their limiting bands, were also free of labelling (Fig. 7d, g). The typical labelling of the lucent fungal wall layer in Fig. 7d, the absence of gold particles over the cell content, except over traces of material in the vacuoles (compare with Fig. 1a), and the buckling of the plasma membrane in this cell are to be noted.

In other situations, numerous gold particles, bound to traces of dense material, occurred in the vicinity of opaque, paired or single bands and membranous structures which had seemingly pervaded unlabelled vessel walls (Fig. 8a, b). Sparse gold particles occurred elsewhere over the section, except in areas corresponding to fungal elements. The presence of these particles in vessel lumina could, at first sight, also have been associated with chipped portions or remnants of vessel secondary walls (Fig. 8c, d); however, in figure 8c, due to the occurrence of inter-mixed opaque bands and vesicular bodies and traces of ruptured, unlabelled middle lamella in conjunction with these, it was difficult to visualize what corresponded to the original cell walls. When vessel secondary walls were eroded, this alteration still appeared to be associated with opaque bands and more diffuse matter related to IEs (Fig. 8d; compare with the material confluent with cell F2 in Fig. 6b); it is to be noted here that the labelled band contacting the unlabelled pit membrane is texturally similar to a fungal cell wall (Fig. 8d). The only content in these cells was texturally similar to extracellular material, and in turn to coating that, in figure 8d, covers a pit membrane of increased opacity.

Links of fungal cells with IEs and the opaque matter

The IE-delimiting bands often appeared to be confluent with the outer fungal wall layer (Fig. 9a). In some cases, IEs’ bands seemed to encapsulate larger fungal elements (Fig. 9b, c; see also Fig. 6b), as indicated by the chitin-labelled content of IEs that were continuous with the fungal cell walls (Fig. 9b). Similarly, the IE-limiting band appeared to coalesce or to be continuous with material (analogous to coating) apposed to vessel walls of the same or adjoining vessel element (Fig. 9a, c). IEs’ bands also merged with fungal cell walls, making them appear as if they were double-layered at some locations (Fig. 9d, e). Typical fungal cells and IEs were also connected to one another by means of crowded arrays of fine filamentous-like structures, often grouped in large masses (Fig. 10a, b).

The content of many fungal cells was nearly homogeneous, not showing typical organelles and internal membranes, except for a few particles of ribosomal appearance bordering an area of more lucent filamentous structures (Figs. 9b, d, 10b). In contrast, the plasma membrane was regular, except for some small localized kinks (Fig. 9d, e). In some cases, similar ingrowths were the only visible membrane-like structures in the cell, whereas other juxtaposed cells, in the same hypha, had a distinct plasma membrane and a different type of content (Fig. 10a).

Figure 11a may be considered to summarize the main aspects of tissue invasion by the pathogen, characterized by tiny intramural hyphae similar to the IEs, by large masses of fine matter connected to these IEs and associated with larger fungal cells, and by similar tiny elements in vessel lumina. Connections of similar matter with fungal cells by means of filamentous structures and with coating material also occurred in vessel lumina (Fig. 11b).

Discussion

The crucial problem in this work was to discover the possible origin of the IEs and to obtain some insight into their nature. It must first be emphasized that using current terminology, based mainly on observations of fungal growth in vitro, was often problematic in attempting to describe these novel elements. For example, elements are described as filaments or layers that might have been part of a larger frondose body. Hence, these could cover a large surface or be part of ramifying elements and be difficult to visualize in their entirety, an undertaking, however, which could undoubtedly provide interesting information. One may raise questions as to whether the tiny elements could correspond to collapsed or degenerative hyphae. First, collapsed hyphae would not be expected to have one of their limiting structures closely contour vessel rims and pit borders, whereas others become deeply infolded or recurvate, or have bridged fungal cells (see Fig. 6b). Also, these IEs in vessel elements were quite similar to elements that were well confined within host walls. The same types of considerations would apply to the question of degenerative hyphae, and one would expect to find in this case various intermediate stages of fungal wall degradation; but this was not the case.

In any event, these IEs and associated elements and structures need to be considered in a new perspective. If they were of pathogen origin, it means that they could maintain a potential for infection, as shown here and in other works for more typical larger fungal cells with defective walls, particularly in microhyphae spanning, non septate, long distances in host cell walls (Charest et al. 2004; Ouellette et al. 2001, 2002, 2004a, for example). Alterations or absence of fungal walls in hyphae penetrating host walls have now also been well illustrated in other systems using immunocytochemical means (Charest et al. 2004; O’Connell and Ride 1990; Tenberge et al. 1996). Moreover, some fungi may grow as protoplasts in culture as well as in nature (Tyrrell and Macleod 1972). Granted that the general or localized absence of typical walls would not be fatal to fungal development in vivo, then the ability of the fungus to develop into innumerable forms and sizes, as presently perceivable, can be more easily understood.

Some pivotal points of convergence for the pathogen origin of IEs and some problems inherent to this situation have been noted thus: as structural analogies between the outer opaque layer of typical fungal cells and the IE-delimiting bands and networks of associated networks; as an association of these elements or even of their individual limiting bands with host wall penetration and alteration; and the inferred analogy in structure and content of the IEs and associated matter lining vessel walls with the invasive microhyphae (see also Ouellette and Baayen 2000; Ouellette et al. 1999a, 2001). Similar microhyphae, which were demonstrated in sterilized wood sections and in elm trees (Casagrande and Ouellette 1971; Ouellette et al. 2004d), have also been well characterized by Hale and Eaton (1985a, 1985b). Hence, these types of elements are not to be ranked as inert components. However, IEs occurring in vessel elements would be less confined by host walls than generally are the elements first described as microhyphae. In this perspective, the following is theoretically perceivable: if the content of a fungal cell of about 3 µm in diam became flattened out possibly into a 0.1 µm thick plate-like IE (as some of these were estimated to be), the host wall surface then covered by the same fungal mass would be many times greater; similarly, if this mass divided into tiny filaments, the possible points of attack would be that much greater. Also, this type of development, in the absence of rigid limiting structures, could translate into a greater efficacy of fungal components apt to be translocated more rapidly towards the pathogen invasion front. Previous observations on the intense physical interaction of microhyphae with middle lamellae and primary walls of vessels in carnation (Ouellette et al. 1999a) indicated that surface interaction may be critical to infection by the pathogen. Indeed, close contact of fungal elements, particularly with thin or defective walls, might promote the efficacy of pectic and other cell wall degrading enzymes (Baayen et al. 1997; Charest et al. 2004; Tenberge et al. 1996). The possible role of these enzymes, as discussed by the last authors, may not necessarily be properly evaluated according to their amounts obtainable from cultures of the pathogen (Cooper et al. 1981, as an example).

Striking transitions were observed between IEs and typical fungal cells. Clearly, in assuming that IEs were of fungal origin, they must initially have formed from typical fungal cells, as an extension of microhyphal-like elements or otherwise. Connections between these elements and fungal cells were observed. However, most IEs extended freely for long distances along or even inside vessel walls (Fig. 5a, for example; see also Ouellette and Baayen 2000), or across vessel lumina (Fig. 4b, i, for example), without displaying evident links with typical fungal cells. Moreover, as shown by the types of connections between IEs and hyphae, it appeared that some larger hyphal cells might even have formed as enlargements from the former (Fig. 9b-d). Fungal cells and IEs were also interconnected by means of microfilamentous-like structures, whereas arrays of similar structures present in host walls were traceable to fungal cells, IEs, the coating and similar opaque matter. These observations add to previous ones from other studies (Ouellette et al. 1995, 1999a, 2002, 2004a, 2004d) showing the possible general occurrence of these structures in fungal infection and the need to unravel their exact nature.

The strong labelling of IEs for chitin can be considered meaningful in cases where they contacted unlabelled vessel walls, another feature substantiating that IEs were not primarily of host origin. When labelled IEs were in contact with labelled secondary vessel walls, they either generally contained membranous structures or were separated from them by an unlabelled opaque band, except where they impinged upon these vessel walls, as did the compact coating. The strong labelling of vessel walls for chitin is in agreement with reports that chitin analogues occur in vessel walls in other plants (Benhamou and Asselin 1989). In regular fungal cells, the probe specifically attached to the inner, lucent wall layers that are known as the chitin backbone of the fungal wall (Ruiz-Herrera 1992; Wessels 1994).

The abundance of gold particles of the chitin probe in vessel lumina, compared with the few elsewhere in void spaces and controls, could have been related to substrates released from the vessel wall. In this instance, the presence of the cellulose probe particles should have been alike in similarly colonized vessel lumina (as verified in some contiguous sections), but those were found to be almost nil therein. In comparison, breakdown products of host walls surrounding intramural fungal cells were strongly labelled (see also Ouellette and Baayen 2000; Ouellette et al. 2004a). As numerous gold particles occurred next to strongly chitin-labelled IEs apposed to unlabelled host walls for an appreciable distance (e.g. young vessels with only sparse thickenings), chitin-labelled compounds in vessel lumina could then have mainly come from IEs. The fact that chitin or chitosan moieties can be released from fungal cells has previously been demonstrated in some Fusarium-infected plants (Hadwiger et al. 1981). In staghorn sumac, vessel secondary walls did not label with the WGA probe and gold particles were present in vessel lumina only again in proximity of labelled fungal cells (unpublished). In similar and simultaneous tests with samples of the resistant carnation plants (Ouellette et al. 1999a, 2002), gold particles were few over vessel lumina and the uncommon coating except where it also encroached on vessel walls (see also Fig. 12C in Nicole et al. 1994); moreover, IEs were lacking in the resistant cv. At any rate, when vessel secondary walls were eroded, this action was associated with IEs and other matter (see Ouellette et al. 2004c), a feature which would be most striking in figure 8c, if the largest labelled mass in vessel V1 was really part of the vessel wall. In any event, the features illustrated in this figure and others indicate that host wall alterations occurred in a non-classical manner.

If the occurrence of gold particles in IEs have bound to possibly dispersed chitin moieties, the occurrence of this substrate therein would still need to be explained. As one possiblility, their labelling patterns could primarily have been related to autolysis of chitin or to its alteration as an effect of extraneous enzymes, with these being apt to permeate the often thick, element-delimiting opaque bands. This possibility can at least be questioned. If, in opposition, the profuse chitin occurrence in IEs corresponded to native chitin, then its particular location would seem to deviate from the main model proposed for chitin synthesis and deposition, as established for fungi growing in culture (Ruiz-Herrera 1992; Sietsma et al. 1995; Stewart and Rogers 1983; Wessels 1994). However, exceptions to this have been noted that were attributed to stress or other unfavorable factors (see Gooday and Schofield 1995). Hence, it would seem that an interaction of several chitin synthase genes, some existing in “latent” forms with boosted activities in some instances, might be involved and active at the plasmalemma level. Examples are the high chitin synthase activities in protoplasts of Aspergillus sp. (Archer 1977; Hardy and Gooday 1983) and the chitin-rich septa in “morphologically aberrant hyphae in Neurospora due to a disruption of a chitin synthase gene” (Yarden and Yanofsky 1991, citation on p. 115 in Gooday and Schofield 1995). In this perspective, it is noteworthy that irregularities in chitin labelling of typical fungal cells in the present and other cases (unpublished observations) were associated with apparent anomalies of the plasmalemma. In any event, labelling for chitin in IEs or at their margins was noticeable only when they contained cytoplasmic-like structures or were confluent with portions containing these structures.

Contrasting with the presence of opaque coating and other associated matter which were present at all stages of infection in both the susceptible carnation plants, IEs were not detected in early infection in the susceptible plants nor in the resistant plants at all stages of infection. However, some components similar to those of the IEs also occurred in the coating, including homogeneous opaque matter often of filamentous appearance, and paired bands. Small filament-like structures delimited by similar bands also occur in vessel lumina in elms and non-hosts infected with Ophiostoma novo-ulmi (Ouellette et al. 2004b), and in staghorn sumac infected with Fusarium oxysporum f.sp. callistephi (unpublished observa-tions). In support to this aspect, the following cases may be referred to: a pure, cellulose synthesizing enzyme, similarly fixed for TEM observations, was illustrated as having a filamentous appearance (Colvin 1980), and movement proteins located on tubules have been observed to extend freely outside virus–infected host cells (Ritzenhaler et al. 1995). Likewise, numerous fungi are known to produce fimbriae (Poon and Day 1975), now shown to contain RNA and collagen as some main components (Celerin et al. 1994, 1996), that extend freely into the surrounding medium. Present observations, therefore, might be of germane importance not only in the study of plant diseases but also of other systems.

If the opaque coatings were somehow related to IEs and microhyphae, one may assume that host defence reactions might contribute to hinder further action or development of coatings. Hence, the fungitoxic and wall-enforcing compounds formed in resistant cv. Novada (Baayen et al. 1996; Ouellette et al. 2002) might also be active in fixing the coating material. In this scheme, a pathogen origin of the coating would not necessarily preclude the subsequent or simultaneous occurrence of other components of host origin. It seems appropriate to re-emphasize that the coating and IEs did not label for pectin and cellulose in the infected plants studied (Ouellette et al. 1999a, 2002, 2004b, 2004d, references cited therein, and unpublished observations).

In summary, reliable evidence has been presented that the IE elements were related to host tissue and cell alterations, despite their not displaying a so-called normal organization, and hence were most probably of pathogen origin. Fine matter, often of a filamentous appearance, associated with IEs, fungal cells and coating also seemed to be involved in host attack. This matter was not only connected to fungal cells (as also shown in Ouellette et al. 2004c) but was similar to the observed, nearly homogeneous and dense content of some fungal cells, including a large seemingly unbound portion corresponding to the nucleus region, conditions which can be considered to be atypical of eukaryotes. Similar observations have been made concerning O. novo-ulmi growing in elm trees as well as on other substrates (Chamberland and Ouellette 1977; Ouellette et al. 1995, 1999b) and Fusarium in sumac (not illustrated). Evidence for the production and extrusion of constituents from fungal cells, of this and other pathogens that are apparently able to invade in an unrestricted manner host tissue is increasing (see discussions in Ouellette and Baayen 2000; Ouelletteet al. 1999a, 1999b, 2000, 2002, 2004c, 2004d). Material in the form of tiny filaments has also been observed to be associated with fungal cells in staghorn sumac and eggplant infected respectively with F. oxysporum and Verticillium dahliae Kleb., and likewise with cell disturbances (unpublished observations). However, it is not so much the extrusion of such matter that may cause concern, but its spread and activity once outside the cell.

Finally, granted that IEs are active elements, and hence most probably of pathogen origin, they could play an important role in final intensive tissue invasion, as in other wilt diseases (Bishop and Cooper 1983). Alternatively, IEs could be involved in disease recurrence in perennial plants, as for Dutch elm disease (Ouellette and Rioux 1993), even though only a limited number of typical fungal cells might be directly regenerated at one point or another from these elements, or from their fusions. From these few cells, other cells might proliferate to invade more fully the moribund tissues and then to produce the next source of inoculum.

Abstract

Résumé

Introduction