The use of excised roots as host partner in AM symbiosis was first proposed by Mosse and Hepper (1975). Isolated root can be propagated continuously in different solid and liquid media with high reproducibility. Clonal roots of some 15 p1ants have been established (Butcher, 1980), and this list has been enlarged during the last decades. Initiation of isolated roots requires pregermination of seeds previously surface sterilized with classical disinfectants (sodium hypochlorite, hydrogen peroxide), then thouroughly washed in sterile distilled water. Germination of seeds occurs after 48 h at 28°C in the dark on water agar or moistened filter papers. The tips (2 cm) of emerged can be transferred to a rich medium such as modified White medium (Bécard and Fortin, 1988) or Strullu and Romand medium (Strullu and Romand, 1987).  The pH of the medium is adjusted to 5.5 before autoclaving. Fast-growing roots are cloned by repeated subcultures.

Transformation of roots by the soil-borne microorganism Agrobacterium rhizogenes has provided a new way to obtain mass production of roots in a very short time. A 2-day old loopful of a bacterial suspension is used to inoculate sections of root organs. Genetically modified caroot (Daucus carrota L.) roots by A. rhizogenes show profuse roots two to four weeks later. Then, the tips are aseptically cultivated on rich medium (Bécard and Fortin, 1988). Several subcultures (3 to 4) are necessary in this medium enriched with antibiotics such as carbenicillin or ampicillin, to obtain free living roots without bacteria. A clonal culture derived from a single root is then established.

A. rhizogenes inserts in transformed roots copies of T-DNA (transfer DNA) which occurs in a large plasmid of A. rhizogenes (Chilton et al., 1982). Therefore, transformed roots  have a quick, vigorous and homogenous  growth in relative poor substrates without supplementation of hormonal substances. The negative geotropism of transformed roots facilitates contacts with hyphae of AM fungi (Bécard and Fortin, 1988; Mugnier and Mosse, 1987). Tepfer (1984) indicated they can survive for a long time without subculture, and Diop (1990) observed less contaminations in culture of transformed carrot roots due to their negative geotropism.

Pure culture of AM fungi is still a big challenge for microbiologists. The pre-symbiotique growth of AM is characterized by formation of a so-called running hyphae. After some weeks without additional host partner, growth of germinated AM propagules stops the hyphae septate and the cytoplasm is retracted. Several factors such as nutrition (Hepper, 1983), chemical treatments (Gianinazzi-Pearson et al., 1989; Bécard and Piché, 1989a), and genetical factors (Burggraaf and Beringer, 1989; Bécard and Pfeffer, 1993; Bianciotto and Bonfante, 1993) have been studied to explain the lack of pure growth of the extraradical phase of AM fungi. Low amounts of natural or synthesized flavonoids positively stimulated fungal growth (Morandi, 1989, Chabot et al, 1992b). Otherwise, certain amino acids (Hepper and Jacobsen, 1983), volatile compounds, particularly carbon dioxide and root exudations, promote hyphal elongation (Carr et al., 1985; Elias and Safir, 1987; Bécard and Piché, 1989b).

Hildebrant et al. (2002) mentioned the presence of slime-forming bacteria identified Paenibacillus validus on surface sterilized spores of Glomus intraradices. These bacteria stimulate the growth of G. intraradices up to spore formation in the absence of any plant tissue. Similar results on presymbiotic sporulation have been found with other Glomus sp (Diop, unpublished data). The authors explain this unusual sporulation by possible utilization by the fungi of a chemical component (or components) secreted by the bacteria. Recently Buee et al. (2000) isolated a semi-purified fraction called “Branched factor” from 8 mycotrophic plant species that are very active in stimulation of germination and nuclear division of Gigaspora sp.

Sheared-root of G. versiforme and G. intraradices develop on water agar medium extensive extraradical structures up to sporulation without supplemented living host partner during three months (Diop et al., 1994b). The richness of the intraradical forms of AM in lipid contents may support this growth irrespective of the decrease metabolic activity of the root explants by both surface sterilization and the shearing process. The results confirm that the quality rather than the quantity of root exudates is more involved in the stimulation of hyphal elongation (Carr et al., 1985; Elias and Safir, 1987).

The entire vegetative development of AM fungi in monoxenic cultures is followed by several works using either transformed or non transformed roots (Bécard and Piché, 1992; Fortin et al. 2002). Figure 2 shows morphological features of different AM fungi cultivated in vitro in association with isolated roots. A non exhaustive list of AM fungi cultivated monoxenically has been given (Fortin et al., 2002). The modified White medium (Minimal medium M) (Bécard and Fortin, 1988) and the modified Strullu and Romand medium (MSR medium) (Diop, 1995) are the most suitable medium to standardize the in vitro culture of AM fungi. The long term behaviour of G. margarita on Ri-TDNA transformed roots of carrot showed 80% of the fungal infections units were produced during the period of root aging (Bécard an Fortin, 1988; Diop et al, 1992).  

Figure 2. Different patterns of development of A AM fungi produced in dual culture with isolated roots: Glomus intraradices (A), Glomus versiforme (B), Glomus aggregatum (C) and Gigaspora margarita (D).

Developing countries will benefit more AM symbiosis using the monoxenic system, which is cheaper than greenhouse culture (Diop et al., 1994a). We propose a methodology for recovering, for purifying, for cultivating tropical AM fungi  (Figure  3).  AM fungi can be isolated from surface soil layers to soil depths greater than 35 m (1, see Figure 3) (Diop, 1995). They are multiplied in association with a mycotrophic plant under greenhouse conditions (2). Isolated spores or vesicles can be monoxenically cultivated in presence of isolated roots (3,4) or on entire plant host (5). Intraradical forms of AM fungi (6, 7) are used as starter inocula to have monoxenic cultures (8) on M medium (Bécard and Fortin, 1988) or on MSR medium (Diop, 1995). Regular subcultures every 4 months with total transmission of AM symbiosis (9) allowed to obtain different fungal generations in continuous cultures. Probable loss of infectivity is avoided by introducing plant in the in vitro system (10). The use of Sunbags (11) in in vitro system is an other alternative to obtain large scale AM inoculum without contaminations. The axenic AM propagules are conserved at 4°C in the dark for several months or use for fundamental or inculation practices.

          Figure 3. Protocol of extraction, purification, cultivation and preservation of AM fungi.

The possibility of continuous culture and cryopreservation have resulted in an international collection of in vitro AM fungi (websites: http://www.mbla.uclac.be/ginco-bel and http://res2.agr.ca/ecorc/ginco.can/).

Although in vitro culture is an artificial system, it may be a valuable tool to study fundamental and practical aspect of AM symbiosis, complementing the experimental approaches. Development of extraradical mycelium under aseptic conditions is often accompanied by the production of so-called arbuscule-like structures (ALS) (Bago et al., 1998a) or branched absorbing structures (BAS) (Bago et al., 1998b). The authors suggest the possible implication of the ALS/BAS in the exchange sites between AM fungi and host plants.

Diop (1995) established a bank of germplasms of AM monoxenically cultivated in association with isolated tomato or transformed carrot roots. The produced propagules (spores, hyphae, infected roots) were able to germinate and to reinfect new plants efficiently. Encapsulation stabilize the biological properties of mycorrhizal roots and isolated vesicles or spores (Strullu et al., 1991; Declerk et al., 1996b). This immobilization also preserve the infectivity of AM propagules under in vitro or in vivo assays.

AM fungi can contribute to root disease suppression though mechanisms are not well understood (Linderman, 1994). The most obvious effect of AM fungi has been attributed to amelioration of nutrient uptake (P and others), resulting in more vigorously growing plants better able to ward off or tolerate root disease. St-Arnaud et al. (1995a) proposed a compartmentalized in vitro system to elucidate interactions between G. intraradices and the root pathogen Fusarium oxysporum f.sp chrysanthemi. Significant negative correlations were found between conidia production and G. intraradices hyphae or spore concentrations. McAllister et al. (1994) found no in vitro interactions between spores of G. mosseae and Trichoderma koningii or Fusarium solani. Life cycles of G. intraradices and the burrowing nematode, Radopholus similes, were achieved in monoxenic cultures (Elsen et al., 2001). The AM fungus increased protection of the root by reducing the nematode population by 50%.

The comparisons between in vivo and in vitro systems  for production of spores would be very much in favour for the axenic systems (Fortin, personal communication). However the effectiveness of these propagules under adverse conditions remains unclear. Some AM fungi lose their infectivity after several successive in vitro subcultures (Plenchette et al., 1996). Therefore it will be necessary to evaluate inoculum potential of different generations of AM propagules in continous monoxenic cultures. In order to preserve permanent fungal biodiversity, in vitro and in vivo collections must always be maintained.

The encapsulation of AM fungi produced monoxenically in alginate beads offers the possibility to diversify inoculation process. It will be useful to incorporate in beads containing AM fungi, some stimulatory compounds such as flavonoids (Bécard and Piché, 1989; Gianinazzi-Pearson et al., 1989) or synergic microorganisms (Hildebrant et al., 2002).

The development of arbuscule-like structures only in a few dual cultures (Karandashov et al., 2000) poses the questions whether they have taxonomic significance or they are produced in soils. The compartmentalized in vitro system (St-Arnaud et al., 1995b) may help to better understand the mechanisms involved in interactions between AM fungi and pathogenic/nonpathogenic rhizosphere microorganisms and also the metabolism of AM fungi.

Acknowledgements

The author wishes to thank the International Foundation for Science (Contract C/2896), the FNRAA project (AP28/SS310800), and the Inco-Enrich project (contract ICA4-CT-2000-30022) for financial support.