Preservation of microorganisms is a main interest in two applied fields: (i) ex situ preservation of microbiodiversity in culture collections and (ii) food starter storage. The processes required in those domains implied specific biotechnologies in order to ensure optimal long-term viability, activity and genetic stability. The main preservation methods for microorganisms, also considered as stresses (Thammavongs et al. Reference Thammavongs, Denou, Missous, Guéguen and Panoff2008), are freezing (Yamasato et al. Reference Yamasato, Okuno and Ohtomo1973; Hubalek Reference Hubalek1996; Thammavongs et al. Reference Thammavongs, Poncet, Desmasures, Guéguen and Panoff2004), freeze-drying (Benedict et al. Reference Benedict, Sharpe, Corman, Meyers, Baer, Hall and Jackson1961; Morgan et al. Reference Morgan, Herman, White and Vesey2006) and spray drying (Corcoran et al. Reference Corcoran, Ross, Fitzgerald and Stanton2004).
Geotrichum candidum, the anamorphic form of Galactomyces candidus (de Hoog & Smith Reference de Hoog and Smith2004), is a filamentous yeast-like fungus with a thallic reproduction (Kirk et al. Reference Kirk, Cannon, David and Stalpers2001). In this microorganism, the asexual spores are called arthrospores and, according to Park & Robinson (Reference Park and Robinson1969), a hyphal tip cell shows three modes of behaviour: somatic growth by extension, cease extension-growth without morphological differentiation, or cease extension-growth and differentiation into arthrospores. This micro-eukaryote, widely used in the dairy industry as a ripening starter, requires a special interest owing to its role in the appearance and the improvement of the organoleptic characteristics of cheeses (Mogensen et al. Reference Mogensen, Salminen, O'Brien, Ouwehand, Holzapfel, Shortt, Fondén, Miller, Donohue, Playne, Crittenden, Bianchi Salvaldori and Zink2002; Wouters et al. Reference Wouters, Ayad, Hugenholtz and Smit2002; Boutrou & Guéguen Reference Boutrou and Guéguen2005), and in the inhibition of growth of pathogenic bacteria and undesirable moulds in the cheese ripening and malting processes (Dieuleveux et al. Reference Dieuleveux, Lemarinier and Guéguen1998; Foszczyńska et al. Reference Foszczyńska, Dziuba and Stempniewicz2004; Molimard et al. Reference Molimard, Buchet and Boivin2005). The G. candidum species present heterogeneity in genetic and phenotypic characteristics (Marcellino et al. Reference Marcellino, Beuvier, Grappin, Guéguen and Benson2001; Gente et al. Reference Gente, Desmasures, Panoff and Guéguen2002; Reference Gente, Sohier, Coton, Duhamel and Guéguen2006). According to literature, the intraspecies morphological diversity on solid medium shows two main morphotypes: “yeast” type strains considered as producing numerous arthrospores and “mould” type strains presumably characterized by preponderance of hyphae and few arthrospores. “Intermediate” type strains have been also described (Guéguen & Schmidt Reference Guéguen, Schmidt, Hermier, Lenoir and Weber1992; Boutrou & Guéguen Reference Boutrou and Guéguen2005).
Although different methods have been proposed for the improvement of fungi (including yeasts) preservation (Espinel-Ingroff et al. Reference Espinel-Ingroff, Montero and Marti-Mazuelos2004; Diogo et al. Reference Diogo, Sarpieri and Pires2005; Borman et al. Reference Borman, Szekely, Campbell and Johnson2006; Homolka et al. Reference Homolka, Lisa, Kubatova, Valqova, Janderova and Nerud2007), the few studies which have explored the preservation of G. candidum are focused on the physiological adaptation of this microorganism to freezing stress by homologous and heterologous mild stress pre-treatment (Thammavongs et al. Reference Thammavongs, Panoff and Guéguen2000; Dubernet et al. Reference Dubernet, Panoff, Thammavongs and Guéguen2002). Recent works have shown that the increase of the nucleation point by adding a biological ice nucleator (SNOMAX®), decrease the freezing lethality of G. candidum (Missous et al. Reference Missous, Thammavongs, Dieuleveux, Guéguen and Panoff2007).
Since the main challenges in fungal cryopreservation are preventing cell loss and physiological modifications, the aim of the present work was to investigate whether significant correlations exist between G. candidum morphotypes and freezing sensitivity, in relation with the arthrospore/hyphae involvement. In Reference Hashimoto and Blumenthal1978, Hashimoto & Blumenthal studied the arthrospores resistance to freezing in the dermatophyte Trichophyton mentagrophytes which is a thallospored fungus as G. candidum. These authors demonstrated that arthrospores are significantly more resistant to freezing than hyphae.
Materials and methods
Strains and culture medium
The strains retained for this study were chosen according to their thallus pattern on solid medium. Their belonging to Geotrichum candidum species as stated by de Hoog & Smith (Reference de Hoog and Smith2004) were confirmed by Gente et al. (Reference Gente, Sohier, Coton, Duhamel and Guéguen2006) with molecular studies. Experiments were performed using seven G. candidum strains (three strains of each main morphotype and one intermediate strain; Table 1, Fig. 1.A) from the laboratory collection (UCMA). Those fungal microorganisms are stored at −80°C as cell suspensions in 15% (v/v) glycerol.
(A) Thallus of G. candidum on malt extract agar at 25°C during 48 h. (B, C) Photomicrographs of G. candidum cells after 20 h and 48 h respectively, in malt extract broth at 25°C. “Yeast” strains: UCMA 454, UCMA 21 and UCMA 332 (1, 2, and 3, respectively). “Intermediate” strain: UCMA 91 (4). “Mould” strains: UCMA 96, UCMA 499 and UCMA 103 (5, 6 and 7, respectively). Scale bars: A=1 mm, B=100 μm.
G. candidum strains used in the present study
Growth experiments in solid or liquid media were carried out with MEA (Malt Extract Agar, AES, Bruz, France) or MEB (Malt Extract Broth, AES) media, respectively (Raper & Thom Reference Raper and Thom1949).
Growth conditions
Cells were spread on agar slope, incubated 48 h at 25°C, and suspended in 3ml 0·9% (w/v) NaCl in order to prepare the preinoculum for further experiments.
To obtain an initial OD600nm=0·05±0·005 (Spectronic 301, Bioblock Scientific, Illkirch, France), 1–3 ml preinoculum were added to 100 ml MEB medium in a 1L Erlenmeyer flask. The liquid culture was incubated at 25°C in the dark with orbital shaking at 150 rpm (AS850, LSL Biolafitte SA, St Germain en Laye, France).
Culture growth was measured simultaneously by spectrophotometry readings at 600 nm and by Thallus Forming Units (TFU) counting (25°C; 48 h). Measurements were performed hourly over at least 48 h for all strains. The TFU/OD ratios are shown in Fig. 3.
Freezing experiments
Cells from 48 h culture (stationary growth phase) were pelleted at 2400 g for 10 min (Eppendorf centrifuge 5810 R, Hamburg, Germany), washed and centrifuged twice (2400 g, 10 min) in 0·9% (w/v) NaCl. Cells were then resuspended in the NaCl solution and adjusted to an OD600nm=1 (±0·05). Suspensions aliquots (1 ml) were then frozen in 1·5 ml Eppendorf microtubes using a Peltier cooler/Heater (PCH-2, Grant-bio, Cambridgeshire, England) set at −10°C. After 24h at −10°C, cell suspensions were sampled, diluted in 0·9% (w/v) NaCl and spread on MEA medium for numeration. Results are the means of three experiments.
Macro and microscopic approaches
Thallus of G. candidum strains were macroscopically observed after 48 h incubation on solid media (MEA) and after 72 h incubation in liquid media (MEB) in test tubes under orbital shaking (120 rpm) at 25°C. The microscopic observations were performed using a microscope (Leika, DMLB, Germany) coupled with a video camera (Linkam Scientific Instruments Ltd., UK) after 20, 48 and 72 h growth in MEB at 25°C.
Results and Discussion
Macroscopic and microscopic observations
Fig. 1 shows thallus morphological characteristics (A) of the strains and corresponding microscopic observations (C). As described in the literature (Guéguen & Jacquet Reference Guéguen and Jacquet1982; de Hoog et al. Reference de Hoog, Smith and Guého1986; Boutrou & Guéguen Reference Boutrou and Guéguen2005), the strains used in this study exhibit two thallus patterns, and thus are regarded as two different morphotypes: (i) “yeast” type (UCMA 454, 21 and 332), characterized by cream coloured colonies with a greasy aspect (Fig. 1.A1, 1.A2 and 1.A3) and (ii) “mould” type (UCMA 96, 499 and 103), showing white felting colonies forming filaments (Fig. 1.A5, 1.A6 and 1.A7). Strain UCMA 91 which shows characteristics between the types described above is considered as “Intermediate” (Fig. 1.A4).
In the micro-fungus G. candidum, hyphae are composed of vegetative cells which can differentiate into arthrospores (de Hoog & Smith Reference de Hoog and Smith2004). According to Guéguen & Schmidt (Reference Guéguen, Schmidt, Hermier, Lenoir and Weber1992) the intra species diversity which is currently associated with the variability of the macroscopic morphotypes, is related to this differentiation phenomenon. The expectation was to find a relationship between thallus patterns and microscopic observations. The three “yeast” strains show an important predominance of arthrospores (Fig. 1.C1, 1.C2 and 1.C3) whereas the “mould” strains UCMA 499 and 103 exhibit an abundance of hyphae (Fig. 1.C6, 1.C7). The “intermediate” strain UCMA 91 shows microscopic pattern similar to those of the “yeast” strains (Fig. 1.C4). Interestingly, the “mould” strain UCMA 96 appears with relatively equal parts of arthrospores and hyphae (Fig. 1.C5). According to these results, the relationship between macroscopic aspect and arthrospore/hyphae ratio, which has been proposed by Guéguen & Jacquet (Reference Guéguen and Jacquet1982), remains questionable. Two hypotheses might be proposed: (i) there is not a systematic correlation between the macroscopic morphotypes of G. candidum and the corresponding arthrospores/hyphae ratio; (ii) the macro and microscopic aspects are partially related to the culture media (solid or liquid).
Aspects of G. candidum strains in liquid cultures are clearly distinct (Fig. 2). The “mould” strains cultures are divided in two parts: the main one as a thick crust at the surface of the broth, and the other one, slight and cloudy at the bottom (Fig. 2.C). In parallel, the “yeast” strains cultures present an important cloudy bottom as single part (Fig. 2.A). Interestingly, the “mould” strain UCMA 96 shows an important cloudy bottom (data not shown).
Different aspects of cultures on Malt Extract Broth at 25°C during 72 h of three G. candidum strains [“yeast” strain UCMA 21, “intermediate” strain UCMA 91 and “mould” strain UCMA 499 (A, B and C, respectively)] and corresponding microscopic observations (1, 2 and 3).
As shown through Fig. 2 corresponding to the “yeast” strain UCMA 21 (Fig. 2.A) and the “mould” strain UCMA 499 (Fig. 2.C), microscopic observations revealed that cloudy bottoms of all studied strains cultures correspond to arthrospores (Fig. 2.1 & 2.3) whereas the surface of the mould strains cultures correspond to hyphae (Fig. 2.2).
The macroscopic differences between “yeast” and “mould” strains, in liquid media, reflect the predominant morphological form (arthrospore or hyphae) in each cell suspension, since arthrospores constitute dense and cloudy bottoms unlike the hyphae which compose solid crust at the surfaces, probably owing to low density. Densities of the different fractions have not been measured. These observations which permit determination of the predominant morphological form, support the second hypothesis that macro and microscopic aspects of G. candidum are partially related to solid or liquid culture media.
Growth behaviour
G. candidum growth behaviour was studied by following the evolution of the TFU×ml −1/OD 600nm ratio as a function of time (h). This ratio involves two factors that characterize the microbial growth: (i) The Optical Density (OD) which is representative of the biomass evolution, and, (ii) the number of TFU which is partially (temporally and quantitatively) related to the sporulation ability. The growth curves of the “yeast” strains (Fig. 3.a, b & c) are characterized by a triphasic structure: initially, the ratio TFU×ml −1/OD 600nm decrease regularly owing to the increase of the biomass. The second phase corresponds to the increase of the ratio consequently to the TFU/ml increase. The last phase shows a decrease (Fig. 3.a & b) or a relative stabilization (Fig. 3.c) of the TFU×ml −1/OD 600nm ratio, according to the strains. Except for UCMA 96 (Fig. 3.e), the growth curves of the “mould” strains are characterized by an irregular monophasic structure (Fig. 3.f & g).
Growth behaviour of G. candidum strains expressed by the TFU×ml −1/OD 600nm ratio as a function of time in hours (t/h), cultured in liquid MEB medium. Data for each strain correspond to 3 replicates (square, circle and triangle) of 2 complementary experiments (open and closed symbols). “Yeast” strains: UCMA 454, 21 and 332 (a, b and c, respectively). “Intermediate” strain: UCMA 91 (d). “Mould” strains: UCMA 96, 499 and 103 (e, f and g, respectively).
In the triphasic curves of the “yeast” strains, the initial decrease of the TFU×ml −1/OD 600nm ratio corresponds to the germination of the arthrospores (initial inoculum) followed by the filaments elongation which leads to the biomass increase (Guéguen & Jacquet Reference Guéguen and Jacquet1982). At the second phase, the ratio TFU×ml −1/OD 600nm increase corresponds to the release of the sporulation by cellular differentiation of the vegetative cells to arthrospores (Fig. 1.B1, 1.B2 and 1.B3) which initiate the TFU/ml increase. Thus, the decrease/stabilization of the ratio TFU×ml −1/OD 600nm at the third phase corresponds to the germination of the arthrospores generated during the previous phase.
In the monophasic curves of “mould” strains UCMA 499 and UCMA 103, the low amplitude of the TFU×ml −1/OD 600nm ratio could be related to an undefined sporulation phase (Fig. 1.B6 & 1.B7). Since sporulation of the hyphae occurs concomitantly with elongation, the TFU×ml −1/OD 600nm ratio remains relatively constant.
The strain UCMA 91, with an “intermediate” colonial pattern (Fig. 1.A4), shows growth curve (Fig. 3.d) similar to that of the “yeast” strains which is in accordance with its microscopic aspect (Fig. 1.B4).
Interestingly, the strain UCMA 96 which has a macroscopic aspect (Fig. 1.A5) in accord with the “mould” morphotype unlike its microscopic characteristics (Fig. 1.B5), shows a growth curve similar to that of the “yeast” strains (Fig. 3.e).
Freezing sensitivity
The freezing sensitivity of the different G. candidum strains studied was evaluated by cell numeration after exposure at −10°C for 24 h of cell suspensions initially adjusted to an OD600nm=1±0·05. TFU/ml values obtained in these control conditions (Table 2) are significantly different. The “yeast” strains values are higher than 3·1×106 TFU/ml, while those corresponding to the “mould” strains are less than 2·2×106 TFU/ml. Interestingly, since the suspensions were adjusted as described above, these results show that production of arthrospores is more important in the yeast strains compared with the mould strains.
Cell concentrations of G. candidum suspensions before [Control (TFU/ml); O.D.600nm=1] and after frozen [Survival (TFU/ml and %)]
The freezing survival evaluated by numeration showed a global decrease of the TFU/ml values (Table 2).
The relative freezing sensitivity was evaluated by the survival rate values (%). Interestingly the classification based on these values does not match with the previous ones. Depending on the strains, the survival rate is comprised between 13·7 and 75·3%. Except for the “mould” strain UCMA 96, strains could be still gathered into the groups described previously, with values higher than 51% for the “yeast” strains, and lower than 19% for the “mould” strains.
The research described in this article was developed on the hypothesis that the different macroscopic morphotypes observed within the species G. candidum are correlated with the freezing sensitivity. Our results support clearly this hypothesis on liquid medium and partially on solid medium. The yeast strains are significantly more resistant to freezing stress (−10°C) than the mould strains UCMA 103 and UCMA 499. It has been reported that in the dermatophyte fungus Trichophyton mentagrophytes, the arthrospores are more resistant to freezing than the hyphae (Hashimoto & Blumenthal Reference Hashimoto and Blumenthal1978). This is in accord with our microscopic observations which show an arthrospores predominance in yeast strains suspensions of G. candidum. Interestingly, freezing sensitivity of the “mould” strain UCMA96 is relatively low. Nevertheless, this is in accord with the microscopic aspect of this strain which show numerous arthrospores compared with the mould strains UCMA 103 and UCMA 499. The arthrospores predominance and the growth behaviour of the intermediate strain UCMA 91 are in accord with the high resistance to freezing stress and thus corroborate the link between the arthrospores proportions of the strains and the freezing sensitivity.
Depending on culture conditions and environmental factors, two types of arthrospores, cylindrical and ellipsoidal, have been described in G. candidum (Carmichael Reference Carmichael1957; Butler Reference Butler1960; Trinci & Collinge Reference Trinci and Collinge1974; Kier et al. Reference Kier, Allermann, Floto, Olsen and Sortkjaer1976; Quinn et al. Reference Quinn, Patton and Marchant1981; de Hoog et al. Reference de Hoog, Smith and Guého1986). In our study, only ellipsoidal arthrospores have been observed. These arthrospores contain more carbohydrates and fewer proteins than the corresponding mycelium (Allermann et al. Reference Allermann, Floto, Olsen, Sortkjær and Kier1978). In addition, arthrospores walls are two fold thicker than the hyphae walls (unpublished data). This is in accord with the results obtained by Steele & Fraser (Reference Steele and Fraser1973). Thus, the freezing sensitivity of G. candidum seems to be related to the sporulation level.
In conclusion, freezing resistance must be related to specific metabolic characteristics in arthrospores that are insignificant or missing in hyphae. Induction of arthrospores differentiation (Hoff et al. Reference Hoff, Schmitt and Kück2005) could be considered as a tool to optimise freezing preservation of fungal starters used in dairy industry such as G. candidum.
Microscopic aspects, growth behaviour and freezing resistance of the studied strains are closely correlated and aid discussion of the validity of the macroscopic discrimination of the “yeast”, “mould” and “intermediate” morphotypes usually described in G. candidum: The three morphotypes could be regarded as integral parts of a single continuum, or else, only both extreme morphotypes (yeast and mould) would have a real and determined physiological existence. Under these conditions, the definition of morphotype in G. candidum remains questionable and has to consider the macroscopic aspect on liquid medium.
G. candidum is a relevant food micro-eukaryote model to study the strain specific variability of the freezing stress response related to the differentiation phenomenon and to improve the cryopreservation process. In this context, studies of the proteomic expression of G. candidum submitted to cold stress using 2D-DIGE technology are in progress.
This project was supported by the “Ministère de l'Enseignement Supérieur et de la Recherche”, the “Conseil Régional de Basse-Normandie” and the ”Conseil Général du Calvados”.
