Ascospore development in Ceratocystis sensu lato ( Fungi ) : a review

Ceratocystis. Ceratocxstiopsis and Ophiostoma are important pathogens of trees and some agricultural crops and have recently been found on proteas and forest trees in South Africa. Taxonomic controversy exists regarding these genera and ultrastructural studies on the development of asci, uniquely shaped ascospores and centrum structure are inadequate. This rev iew summarises current knowledge of ascospore shape and development of the centrum obtained from light and electron microscope studies of Cemtocystis sensu lato. Important questions requiring further investigations are outlined. It is furthermore proposed that additional ultrastructural studies are required to clarify the current taxonomic disagreement in this group. Such studies could also identify relationships between these fungi and other Ascomycetes.

The generic concept in Ceratocxstis s.I. has been the basis of controversy and has changed regularly since the first eytological studies of these fungi in 1925.In a number of cases, the same fungi have been studied under different < names which could lead to confusion.In this review' we follow the taxonomic scheme of De Hoog & Scheffer (1984) where Ceratocxstis and Ophiostoma are treated as distinct genera.Currently used names of fungi treated in older literature under different taxonomic schemes are given in Table I with the appropriate synonymies.
The aim of this review, is to compile data from past eytological and recent ultrastructural studies, to summarize current knowledge of centrum structure and organization as well as ascospore shape, in Ceratocxstis s.I.The possible implications of these observations in the taxonomy of Ceratocxstis s.I. and their relationships with other Ascomycetes.comparing certain ultrastructural features, are discussed.

CENTRUM DEVELOPMENT
Controversy exists as to whether species of Ceratocystis s.I. should be included in the Plectomycetes, characterized by closed ascocarps (cleistothecia) (Ainsworth et al. 1973;Malloch 1981), or Pyrenomycetes, with ostiolate ascocarps (perithecia) (Luttrell 1951).Luttrell (1951) includes members of Ceratocystis s.I., with their distinc tive ostiolate ascocarps, in the Pyrenomycetes.In contrast, Benny et al. (1980) considered the Ophiostomatales as an order of the Plectomycetes, with the Ophiostomataceae as the only family.An important feature of the Pyrenomycetes is the organization of the centrum.Luttrell (1951) proposed a developmental scheme, where centrum development is initiated by the formation of sterile and fertile cells.The fertile cells are initiated by curved hyphal branches and the sterile cells by additional branching hyphae.These cells, surrounded by other hyphae, give rise to the maturing perithecium.The central multinucleate fertile cell or ascogonium lies above the centre of the perithecium, rather than at its base, the asci therefore develop progressively and basipetally.The sterile cells are distributed between the fertile ascogonial cells, and they have the important function of providing space for the irregularly developing asci, by disintegrating during ascus development.
Asci in the Plectomycetes are evanescent and the ascospores fill the cleistothecial centrum at maturity with no special discharging mechanism.Luttrell (1951) emphasized that the irregular distribution of the asci within the centrum of Ceratocystis s.I., is reminiscent of the Plectomycetes.Redhead & Malloch (1977) concurred with this opinion and included Ophiostoma and Ceratocystis in this order.Furtherm ore, based on the presence of ascospores with galeate sheaths, these authors included the genera with numerous yeasts in the Endomycetaceae.Typically, Plectomycetes have closed ascocarps (Ainsworth et al. 1973), and thus Ophiostoma and Ceratocystis with beaked, ostiolate perithecia could equally be excluded from this group.Malloch (1981)  The development of the centrum of Thermoascus aurantiacus and Ceratocystis thermophile is typical of that of the Plectomycetes, with asci arising from croziers formed by ascogenous hyphae (Ellis 1981a(Ellis , 1981b)).Ultrastructural studies of these species, however, provide insufficient information to make logical comparisons with Ceratocystis s.I.The single-celled nature of yeasts apparently related to Ceratocystis s.I. precludes com parisons of centrum development.
In developmental studies of Ceratocystis s.I ., reference is commonly made to sterile cells, which include both cushion and pseudoparenchymatous cells (Table 2).However, their function has seldom been discussed.Pseudoparenchymatous cells are dispersed amongst the ascogenous hyphae and their arrangement appears to characterise centrum organization (Table 2).Although the ascogenous hyphae generally emerge towards the base of the perithecium, they may also extend to the base of the neck, with the asci developing towards the centre (Table 2).The ascogenous hyphae are usually separated from the perithecial wall by the cushion cells, with asci forming irregularly throughout the centrum.Bothalía 20,2 (1990) According to Luttrell (1951), the ascogonium is located above the centre of the perithecium (Figure 1A), developing basipetally (Figure IB).From available cytological studies we interpret centrum development in Ceratocystis s.I. as follows: ascus development is preceded by the development of ascogenous hyphae (= fertile cells) from the ascogonium.The ascogenous hyphae developing from the ascogonium, appear to form towards the base of the perithecium, with cushion cells (= sterile cells) surrounding them (Figure 1A).The ascogonium therefore develops basipetally (Figure IB) and asci are formed by acropetally developing ascogenous hyphae (Figure 1C).Ascospores then mature basipetally (Figure ID).
Gwynne -Vaughan & Broadhead (1936) stated that there is inadequate reference in most literature to the charac teristic shape of the ascospores in Ceratocystis s.I.The ascospore shape of each species previously studied cytologically, reproduced from descriptions or original illustrations, is compared in Table 2. Hutchinson (1950) emphasized that the number of perithecial wall layers and ascus shape are inadequate criteria for determination of natural relationships.More emphasis should therefore be placed on detailed cytological and ultrastructural studies of centrum organization, as well as ascospore shape and development, as potential criteria in the taxonomy of Ceratocystis s.I. Kimbrough 1987;Furtado 1971;Honegger 1985;Mainwaring 1967;Merkus 1973;Van Brummelen 1987).In general, previous ultrastructural studies of ascospore development have investigated wall formation (Lynn & Magee 1972;Moens 1971), the morphology of nuclear and membrane structure (Hashimoto et al. 1960;Oso 1969) and the role of lomasomes and plasmalemmasomes (M archant & Moore 1973).Ultrastructural studies of ascospore delimitation and development have also been conducted (Beckett 1981;Rosing 1985).A comprehensive review of this process has been provided elsewhere (Turian 1976).The latter review did not, however, cover Ceratocys tis s.I.
Development of ascospores appears to be similar in all Ascomycetes.The spore walls are formed between two delimiting membranes (Ellis 1981a;Stiers 1976).In a preliminary investigation of C. moniliformis (Figure 2A) and O. minus (Hedge.)H. & P. Syd. (Figure 2B) we have tentatively confirmed the presence of an electron transparent endospore and an electron dense epispore wall layer in ascospores of these species.Amongst fungi with evanescent asci, ultrastructural studies are available for Thermoascus aurantiacus Miehe (Ellis 1981a) and Chaetomium thermophile La Touche (Ellis 1981b) as well as Ceratocystis spp.However, in the former fungi, ascospores are elliptically shaped with a charac teristic germ pore and are therefore incomparable with Ceratocystis s.I.

CONCLUSION
With the exception of the hat-shaped ascospores in Ceratocystis fim briata, the characteristic and unusual ascospores in other species of Ceratocystis s.I. have not been illustrated ultrastructurally.It is proposed that additional ultrastructural studies are required to clarify the current taxonomic disagreement in this group.Further ultrastructural studies on centrum organization, ascus development and ascospore shape may provide new keys to relationships between genera of these fungi.A better understanding of Plectomycetes/Pyrenomycetes relation ships could result and might aid in interpretation of taxonom ic and evolutionary relationships in the Ascomycetes as a whole INTRODt CT1()\ Ceratocxstis EI1.& Halst.sensu lato includes the genera Ophiostoma H .& P. Sydow.Ceratocxstis sensu stricto and Ceratocxstiopsis Upadh.&.Kendr.(DeHoog &.Scheffer 1984: Upadhyay 1981;
suggested that the Plectomycete centrum has evolved towards progressive simplification.Sterile tissue, the ostiole, as well as forcible discharge of spores would therefore have been lost.This argument would justify inclusion of Ceratocystis s.I. in the Plectomycetes, despite their having ostiolate perithecia.Our knowledge of ascocarp development in these fungi is based on few examples with almost no ultrastructural information.Ultrastructural studies on the development of the centrum of Ceratocystis , Ophiostoma and related fungi could provide important clues clarifying their relationships.