OUP user menu

The strange case of a biofilm-forming strain of Pichia fermentans, which controls Monilinia brown rot on apple but is pathogenic on peach fruit

Sara Giobbe, Salvatore Marceddu, Barbara Scherm, Giacomo Zara, Vittorio L. Mazzarello, Marilena Budroni, Quirico Migheli
DOI: http://dx.doi.org/10.1111/j.1567-1364.2007.00301.x 1389-1398 First published online: 1 December 2007


A biofilm-forming strain of Pichia fermentans proved to be most effective in controlling brown rot on apple fruit when coinoculated into artificial wounds with a phytopathogenic isolate of Monilinia fructicola. Culture filtrates and autoclaved cells had no significant influence on the disease. When sprayed onto the apple fruit surface, this yeast formed a thin biofilm but failed to colonize the underlying tissues. When inoculated into wounds artificially inflicted to peach fruit or when sprayed onto the surface of peach fruit, the same strain showed an unexpected pathogenic behaviour, causing rapid decay of fruit tissues even in the absence of M. fructicola. Both optical and scanning electron microscopy were used to evaluate the pattern of fruit tissue colonization by P. fermentans. While on apple surface and within the apple wound the antagonist retained its yeast-like shape, colonization of peach fruit tissue was always characterized by a transition from budding growth to pseudohyphal growth. These results suggest that pseudohyphal growth plays a major role in governing the potential pathogenicity of P. fermentans, further emphasizing the importance of a thorough risk assessment for the safe use of any novel biocontrol agent.

  • fruit decay
  • Candida lambica
  • biological control
  • risk assessment
  • biohazard
  • dimorphic transition


Among the fungal species causing brown rot of fruit, Monilinia fructicola is considered to be the most destructive pathogen on both pome and stone fruits, and is widely distributed in the United States, Canada, Australia, and New Zealand (Batra, 1991; Michailides et al., 2007). In Europe, this species is still limited to some areas and is included in the European and Mediterranean Plant Protection Organization (EPPO) A2 list of quarantine pests. Consequently, the possibility of further spread through infected fruit and plant material warrants defining effective control measures in a timely manner.

Monilinia fructicola usually appears with blossom blight, and develops into a twig blight and canker, thus providing inoculum for latent infection of green fruit. Further on, brown rot develops on fruit at harvest and during storage and transport to market (Hong et al., 1998). Fungicide resistance, lack of new effective fungicide chemistries, latent infections, and the capability of the pathogen to grow at low temperatures are among the factors that challenge the management of brown rot. Under such circumstances, biological control may represent an effective alternative to fungicides (Janisiewicz & Korsten, 2002; Passoth et al., 2006).

Despite the impact of Monilinia rot over a large scale, a relatively few reported cases of effective biocontrol agents against this disease are available in the literature, compared with other postharvest pathogens. Yeasts are particularly suitable as biocontrol agents against brown rot, showing effectiveness on several commodities (Janisiewicz & Bors, 1995; Chand-Goyal & Spotts, 1996; Wittig et al., 1997; Hong et al., 1998; Spotts et al., 1998; Karabulut et al., 2001; Spadaro et al., 2002; Tian et al., 2004; Chan & Tian, 2005; Yao & Tian, 2005; Restuccia et al., 2006).

With the intention of identifying new biocontrol agents against brown rot on apple, a collection of isolates representing 52 yeast species has been recently tested for their antagonistic efficacy against Monilinia spp. (S. Giobbe, unpublished).

In preliminary experiments, a strain of Pichia fermentans (coded DISAABA 726) proved most effective in controlling brown rot on apple fruit when inoculated into artificial wounds before introduction of a pathogenic isolate of M. fructicola. However, when tested on peach fruit, this strain showed a remarkable pathogenic behaviour, causing the rapid decay of fruit tissues even in the absence of the pathogen.

The aims of the present investigation were: to evaluate further the biocontrol potential of P. fermentans 726 against brown rot caused by M. fructicola on apple fruit, to report for the first time on the pathogenicity of this strain on peach fruit, and to conduct a primary investigation by optical and scanning electron microscopy (SEM) on the role of biofilm formation and pseudohyphal growth in the different behaviour of P. fermentans on apple and on peach fruit.

Materials and methods

Strains and culture conditions

Pichia fermentans Lodder [anamorph: Candida lambica (Lindner & Genoud) Uden & H.R. Buckley ex S.A. Mey. & Ahearn (1983)] DISAABA 726 (=DBVPG 3627) was isolated from wine must and is maintained in the culture collections of the Dipartimento di Scienze Ambientali Agrarie e Biotecnologie AgroAlimentari and of the Dipartimento di Protezione delle Piante, University of Sassari. The yeast strain is stored in YPD, consisting of 1% yeast extract, 2% bacto-peptone (Difco), 2% dextrose broth (Carlo Erba Reagenti), 2% bacto-agar (Fluka) at 4°C, and in 50% glycerol at −80°C.

Taxonomic identification was confirmed by sequencing the internal transcribed spacer (ITS1-5.8S-ITS4) region of the nuclear rRNA genes (Esteve-Zarzoso et al., 1999). The GenBank accession number of the ITS sequence of P. fermentans 726 is: DQ665310.

Cultures of P. fermentans 726 were routinely prepared as follows: a single colony grown for 48 h on YPD at 25°C was transferred into a 100 mL Erlenmeyer flask containing 20 mL of 1% yeast extract, 2% bacto-peptone, 1% dextrose broth (YEPD), and incubated for 24 h at 25°C on a rotary shaker (200 r.p.m.). This culture contained c. 9 × 108–2 × 109 cells mL−1 yeast-shaped cells, which were then recovered by centrifuging at 1500 g for 5 min, washed and resuspended in sterile Ringer solution, and brought to a final concentration of 1 × 108 cells mL−1 by direct counting with a haemocytometer.

To test the strain's ability to differentiate pseudohyphal development, P. fermentans was cultured on YPD and on nitrogen-limiting medium (SLAD, consisting of 0.17% yeast nitrogen base without amino acid or ammonium sulphate, 2% dextrose, 2% bacto-agar, 50 μM ammonium sulphate; Gimeno et al., 1992), by incubating Petri dishes for 3–5 days at 30°C.

An isolate (PVS-M01) of M. fructicola (G. Winter) Honey was obtained from decayed sweet cherry (Prunus avium L.) and kept in the collection of the Dipartimento di Protezione delle Piante, University of Sassari. This pathogenic strain is stored in potato dextrose agar (PDA, Merck) at 4°C and in 50% glycerol at −80°C, and routinely reisolated from artificially inoculated apples; a monosporic culture of the pathogen was used in biocontrol experiments.

A conidial suspension of the pathogen was prepared as follows: a mycelium fragment of M. fructicola was transferred to the centre of a Petri plate (90 mm diameter) containing a medium composed of 200 mL L−1 filtered Campbell's V8 Vegetable juice, 2.25 g L−1 CaCO3, 2% bacto-agar, amended with 100 μg mL−1 of each streptomycin sulphate and tetracycline hydrochloride (Sigma). Inoculated plates were kept under constant fluorescent light; after 2 weeks of incubation at 25°C, spores were collected by flooding the culture with 10 mL sterile Ringer solution, and by scraping the colony surface with a sterile scalpel. Spores were filtered through two layers of sterile cheese-cloth, counted with a haemocytometer, and brought to a final concentration of 1 × 106 conidia mL−1.

Biological control and fruit colonization experiments

Apple (Malus domestica Borkh., cvs Golden Delicious and Renetta) or peach [Prunus persica (L.) Batsch., cv Springcrest, and Prunus persica var. nucipersica (Suckow.) C.K. Schneid., cv Bigtop] fruits were disinfected in sodium hypochlorite (0.8% as chlorine) and rinsed under tap water, air dried, and punctured with a sterile micropipette tip at the equatorial region (3 mm depth, three wounds per fruit). Ten microlitres of a yeast cell suspension (1 × 108 yeast-shaped cells mL−1) were pipetted into each wound and the fruit were kept at 25°C. At different intervals (i.e. 2 h before, simultaneously, or 2 h after yeast inoculation), each wound was inoculated with M. fructicola by pipetting 10 μL of a conidial suspension of the pathogen (1 × 106 conidia mL−1). A healthy control was included by pipetting 20 μL of sterile Ringer solution into each wound.

The efficacy of living cells of P. fermentans 726, prepared as described previously, was also tested in comparison with autoclaved yeast cells (10 μL of a suspension adjusted to a final concentration of 1 × 108 conidia mL−1 pipetted into each wound), and with a culture filtrate (10 μL of culture filtrate pipetted into each wound), prepared by filtering the overnight culture broth (YEPD) through a 0.2 μm filter (Millipore). Living cells, autoclaved cells, and culture filtrates were pipetted into apple wounds 2 h before pathogen inoculation.

In a series of experiments, a suspension of P. fermentans 726 (1 × 108 cells mL−1) was sprayed with a hand-held nebulizer onto sound or artificially wounded apple and peach fruits to test the strain's ability to establish onto the fruit surface. Approximately 5 mL of suspension containing yeast-shaped cells were applied to each fruit, corresponding to 5 × 108 cells per fruit. Control treatments was sprayed with an equal volume of Ringer solution.

When dry, fruit were placed in plastic holders (60 cm × 40 cm × 15 cm, one layer), wrapped in transparent polyethylene bags to prevent evaporation, and stored at 25°C and 85±5% relative humidity. The incidence of brown rot around artificially inoculated wounds was determined by measuring the diameter of decayed tissue at 48 h intervals. Results are expressed as lesion diameter (in millimeter) after 7 days of incubation.

There were 10 fruits per treatment, the treatments were arranged in a completely randomized block design, and each experiment was repeated at least two times. The data were subjected to one-way anova, followed by multiple comparison by Dunnett's test, using minitab® for Windows release 12.1.

Population dynamics of P. fermentans 726 in apple and peach wound

Apple (cv Golden Delicious) or peach (cv Bigtop) fruit were selected for lack of defects and wounded at three locations in the equatorial region as described previously. Immediately after wounding, each wound was inoculated with 10 μL of a cell suspension (1 × 108 cells mL−1) of P. fermentans 726. Fifteen minutes later, fruit were placed in plastic holders and stored as described. On each sampling time, three fruit were selected and one tissue cylinder (10 mm diameter × 10 mm depth) was cut from each wound with a sterile cork borer. Three cylinders from each triplicate fruit were cut lengthwise with a sterile scalpel, placed in a sterile 50 mL polypropylene tube with 10 mL of Ringer's solution and vigorously shaken for 15 min. The resulting suspension was serially diluted in Ringer's solution and 100 μL aliquots were plated onto YPD. Plates were incubated at 25°C for 48 h and colony counts were expressed as the number of cells per wound at 0, 1, 2, 3, 4, 5, 6, and 7 days from the time of inoculation.

Biofilm formation in vitro by P. fermentans

Invasive growth assay

Pichia fermentans 726 was plated onto standard (2% agar) YPD plates, allowed to grow for 3 days at 30°C, and photographed. Sterile distilled water was used to wash noninvasive cells by rubbing the colony surface vigorously with a sterile gloved finger. Washed plates were incubated for 2 days at 30°C and then photographed a second time (Roberts & Fink, 1994).

Aqueous-hydrocarbon biphasic hydrophobicity assay

Pichia fermentans strain 726 was grown overnight in liquid synthetic complete medium (SC; Sherman et al., 1986) +2% (w/v) glucose, washed once with water, and resuspended in SC+0.1% glucose to an OD600 nm of 0.5. After a stationary incubation of 3 h at 25°C, the OD600 nm of the culture was measured. Then, 1.2 mL of the culture was added to each of three 13 mm × 100 mm borosilicate glass tubes and overlaid with 600 μL of n-octane. After vortexing the tubes for 3 min, the phases were allowed to separate and the OD600 nm of the aqueous layer was measured. The difference between the OD600 nm of the aqueous phase before and after addition of n-octane was used to determine the hydrophobicity (Zara et al., 2005).

Mat formation

Pichia fermentans 726 was evaluated for its ability to form an elaborate pattern of multicelluar growth (‘mat formation’), as described by Reynolds & Fink (2001): the isolate was inoculated onto the centre of YPD agar plates containing 0.3% agar and 2% glucose (soft agar). Triplicate plates were incubated at 25°C for 5 days in the darkness and photographed at different time points.

Adherence to polystyrene surface

Pichia fermentans 726 was grown overnight at 30°C in SC+2% (w/v) glucose. Cells were then washed once in sterile water, resuspended in SC+0.1% (w/v) glucose, and brought to an OD600 nm of 1.0. Decuplicate 100 μL aliquots were then transferred into each well of a 96-well polystyrene plate and the cell suspension was incubated at 30°C for 3 h. An equal volume of 1% (w/v) crystal violet was then added. After 30 min, the wells were washed with sterile water. The adherence of cells was quantified by solubilizing the retained crystal violet with 100 μL of 10% SDS and an equal volume of sterile water after 30 min. Finally, 50 μL of the solution was transferred to a fresh polystyrene 96-well plate and the A570 nm and A590 nm were measured spectrophotometrically. As negative control, the nonadhering Saccharomyces cerevisiae strain S288c was used.


Tissue sections (c. 2 cm × 2 cm × 2 cm in size) to be observed by SEM were cut with a sterile surgical blade from inoculated fruit and fixed in 2.5% glutaraldehyde for 90 min, then washed three times for 5 min with Ringer solution, and postfixed in 1% OsO4 for 2 h. After three 5 min washes with distilled water, samples were acetone-dehydrated with two successive 10 min washes in 25%, 50%, 70%, 80%, 95%, and 100% acetone (v/v), respectively. Before OsO4 fixation, specimens including fruit wounds were cryo-fractured in liquid nitrogen to expose the internal cellular structures. After critical point drying using liquid CO2, the samples were sputter-coated with gold–palladium, examined and photographed using a Zeiss DMS 960a SEM. The interval included between fruit inoculation and sample processing is indicated in the figure legends.


Effect of P. fermentans 726 on brown rot of apple caused by M. fructicola

The biocontrol potential of P. fermentans 726 towards M. fructicola was evaluated on apple cv Golden Delicious. The results of two separate tests are shown in Table 1. In both experiments, the yeast was effective in reducing significantly brown rot on apple after a 1-week storage at 25°C, allowing a reduction of decay up to 100% of the inoculated control.

View this table:

Biological control activity of Pichia fermentans 726 against Monilinia fructicola decay on artificially inoculated apple fruit (cv Golden Delicious)

Lesion diameter (mm)
Application of the antagonistFirst experimentAdjusted P valueSecond experimentAdjusted P value
Control43.1 ± 3.041.7 ± 4.5
2 h before pathogen1.2 ± 4.0000
Coinoculated000.7 ± 2.30
2 h after pathogen4.8 ± 10.9000
  • * Results of two separate experiments are expressed as the extension of decay diameter (± SD) around the inoculated wound after 7 days of incubation at 25°C.

  • ** Values are significantly different from control by Dunnett test (P<0.01).

  • The yeast was inoculated at different intervals (2 h before, coinoculated, or 2 h after pathogen inoculation).

In order to obtain preliminary indication on the role of toxic metabolites in the biocontrol potential of P. fermentans 726, killed cells and cell-free culture filtrates of the antagonist were tested for their efficacy in controlling brown rot on apple cvs Golden Delicious and Renetta. While the living antagonist confirmed its ability to reduce the lesion diameter significantly when applied 2 h before pathogen inoculation, both autoclaved cells and culture filtrates had no biocontrol effect and tended to slightly increase the severity of disease, possibly due to the presence of nutrients at the wound site (Table 2).

View this table:

Biological control activity of living cells, autoclaved cells, and cell-free culture filtrates of Pichia fermentans 726 against Monilinia fructicola decay on artificially inoculated apple fruit (cv Golden Delicious and Renetta)

Lesion diameter (mm)
TreatmentGolden DeliciousAdjusted P valueRenettaAdjusted P value
Control36.7 ± 5.922.2 ± 5.0
Cell suspension005.1 ± 7.10.0003
Autoclaved cells38.6 ± 3.30.678520.6 ± 10.90.9575
Culture filtrate39.1 ± 5.90.515926.1 ± 6.10.6071
  • * Results of two separate experiments are expressed as the extension of decay diameter (± SD) around the inoculated wound after 7 days of incubation at 25°C.

  • ** Values are significantly different from control by Dunnett test (P<0.01).

Response of peach fruit to inoculation with P. fermentans 726

When inoculated into wounds artificially inflicted to peach fruit, alone or together with M. fructicola isolate PVS-M01, P. fermentans 726 caused extensive degradation of tissues around the lesion, appearing after 3–4 days of incubation at 25°C. One week after inoculation, discoloration of the epicarp was accompanied by tissue softening and browning, which extended until fruit stone, surrounding it completely (Fig. 1a and b).


Discoloration of the epicarp (a) and tissue degradation (b) incited by Pichia fermentans 726 when applied by direct pipetting into wounds of nectarine peach fruit (cv Bigtop). Control fruit (left) showed no symptoms of decay. Inoculated fruit (right) was incubated for 7 days at 25°C and 85±5% relative humidity.

Sprayed peach fruit were covered by a dense and translucent yeast biofilm, and softening of underlying tissues occurred both on sound and wounded fruit 3–4 days after treatment (not shown).

The yeast was consistently reisolated in pure culture from decayed peach tissues and induced the same symptoms when inoculated on both sound and wounded peach fruit. Conclusive proof of the identity of P. fermentans 726 upon reisolation was achieved by sequencing the ITS1-5.8S-ITS4 region of the nuclear rRNA genes, thus fulfilling Koch's postulates.

Apple control fruit never showed symptoms of decay in the absence of M. fructicola, even after prolonged incubation at 25°C and 85±5% relative humidity.

Growth and colony morphology of P. fermentans 726 on different substrates

After 3 days of growth on YPD, P. fermentans 726 formed rough colonies, associated with the presence of hyphae and pseudohyphae (Fig. 2a and b). Invasive growth was evident after 3 days on YPD and cells adhering to the solid substrate were able to regenerate filamentous colonies (not shown). Unexpectedly, colonies developed on SLAD medium were composed merely by yeast-shaped cells (Fig. 2c and d).


Optical and scanning electron microscopy of the colony morphology of Pichia fermentans 726 developed after 3 days of growth at 30°C on YPD (a, b) or on SLAD (c, d) medium.

After 48 h of static growth in YNB amended with 4% ethanol, P. fermentans 726 was able to form a thick biofilm, which adhered to the glass tube surface (not shown).

When inoculated in the centre of soft agar plates, P. fermentans 726 formed a flat mat, growing in a radial form on the Petri plate, and ultimately covering the entire plate surface after 5 days.

When grown on SC medium+0.1% glucose, P. fermentans 726 was able to adhere to the polystyrene plates after repeated washes (OD 2.0±0.21 and 2.4±0.22 at 570 and 590 nm, respectively). On the contrary, the S. cerevisiae strain S288c was unable to adhere to polystyrene surface. This phenotype was associated with a high cell hydrophobicity. Accordingly, the OD of the aqueous layer after partition between water and n-octane was 80.03%±0.05 lower than the control value (mean of three replicates).

Growth of P. fermentans 726 on apple and peach tissue

When sprayed onto the apple fruit surface, P. fermentans formed a thin biofilm composed of yeast-shaped cells embedded within an abundant matrix network (Fig. 3a and b).


Morphology of Pichia fermentans 726 cells in biofilm formed 5–7 days after spraying onto the surface of sound apple (cv Golden Delicious; a, b) or nectarine (cv Bigtop; c, d) fruit, respectively.

During colonization of apple wounds, P. fermentans 726 formed a dense layer of cells embedded within an abundant matrix network and adhering to the inner wound surface, but retained its yeast-like growth (Fig. 4a).


Pattern of colonization of artificial wound on apple (cv Golden Delicious; a) and on nectarine peach fruit (cv Bigtop; b–d) by Pichia fermentans 726 inoculated by direct pipetting and incubated for 5 days at 25°C and 85±5% relative humidity. (a) Yeast-shaped cells developing in apple wound. (b–c) Yeast-shaped cells were still visible in the wound cavity, progressively converting to multicellular hyphae during multiplication and tissue invasion.

Conversely, pathogenic colonization of peach fruit was always associated with the extensive differentiation of hyphae, both at the surface (Fig. 3c and d) and at the wound level (Fig. 4b and d). On the peach surface, developing hyphae were embedded within an amorphous extracellular material. During the first phases of peach wound colonization, it was possible to observe the yeast-shaped cells of P. fermentans 726 in the wound cavity and their progressive conversion to multicellular hyphae during multiplication and tissue invasion (Fig. 4b and c).

Populations of P. fermentans 726 in apple wound increased from 6.1 × 105 cells per wound at the inoculation time to a maximum of 1.0 × 107 cells per wound after 4 days, and then stabilizing at 4.6 × 106 cells per wound during the subsequent incubation period. On the contrary, populations dynamics in peach wound underwent a rapid increase from 4.7 × 105 cells per wound at the inoculation time to 9.5 × 106 cells per wound after 24 h, and then reached a concentration of 1.6 × 108 cells per wound after 7 days of incubation.


In this work, the dual nature of a strain of P. fermentans is reported on, which controls brown rot on apple fruit, but becomes a destructive pathogen when applied to peach fruit. On apple fruit, culture filtrates and autoclaved cells had no significant influence on M. fructicola, suggesting that biocontrol exerted by this strain should not depend upon the production of toxic metabolites. Several mechanisms have been proposed to play a role in the antagonistic potential of yeasts against brown rot caused by Monilinia species: competition for space and nutrients (Chand-Goyal & Spotts, 1996; Spadaro et al., 2002; Restuccia et al., 2006); survival ability on fruit surfaces and adaptability to postharvest storage conditions (Karabulut et al., 2001; Tian et al., 2004); production of extracellular hydrolases (Chan & Tian, 2005); and induction of resistance responses at the wound site (Yao & Tian, 2005).

Recently, the ability to form biofilms on the inner surface of wounds was indicated as a possible mechanism of biocontrol (Scherm et al., 2003; Ortu et al., 2005). Biological control experiments carried out with an S. cerevisiae flor strain able to form a biofilm in liquid medium have shown its effectiveness against Penicillium expansum, the cause of blue mold on stored apple fruit. The activity of this biofilm-forming strain was tightly correlated to the morphological phase during which the cells were collected: only yeast cells collected from the biofilm phase were effective in limiting pathogen growth, apparently being able to colonize the inner surface of artificial wounds with more efficiency (Ortu et al., 2005). Also, P. fermentans 726 actively colonized the wounds artificially inflicted to apple fruit, reaching population densities of 5 × 106–1.0 × 107 cells per wound after 7 days of incubation, while forming a dense layer of yeast-shaped cells embedded within an abundant extracellular matrix. The biocontrol efficiency observed against M. fructicola could therefore be due to the ability of yeast cells to remain adherent one to the other and to create a mechanical barrier interposing between the wound surface and the pathogen spore and germ tube, thus hampering infection by M. fructicola.

When P. fermentans 726 was applied to peach fruit (by direct pipetting into artificial wounds or by spraying onto the fruit surface), it acquired pathogenic features, inducing extensive tissue degradation and invading fruit tissues by reaching densities of >108 cells per wound after 7 days incubation.

Phytopathogenic yeast species are seldom reported in the literature. Two Pichia species (Pichia heedii and Pichia cactophila) have been associated with decaying tissue of cacti in North America (Phaff et al., 1978; Starmer et al., 1978). On Opuntia ficus-indica cladodes, a soft rot caused by Candida boidimi was described by Granata & Sidoti (2002), while Clavispora opuntiae was found in necrotic tissues of Opuntia sp. (Phaff et al., 1986). Swart & Swart (2002) isolated four species of yeasts from diseased cactus pear fruit: Hanseniaspora uvarum (the most frequent species), Pichia kluyveri, Pichia membranifaciens, and Candida sp.

Gognies (2002) reported on the potential virulence of S. cerevisiae, which caused growth retardation and plant death on grapevine. Pseudohyphal growth and (although not always) endopolygalacturonase activity were associated with pathogenicity (Gognies & Belarbi, 2002; Gognies et al., 2006).

In the case of P. fermentans 726, a biofilm was formed on apple and peach fruit surfaces by either yeast-shaped or filamentous cells, respectively. SEM observations revealed that both cell types were surrounded by an amorphous extracellular matrix. This is in accordance with several observations carried out on other yeast species, such as S. cerevisiae, Cryptococcus neoformans, Candida albicans, and Candida tropicalis, where the production of a matrix coat was compared with bacterial exopolysaccharides, emphasizing its role in biofilm maturation and multicellular community protection (Reynolds & Fink, 2001; Jefferson, 2004; Palková & Váchová, 2006). However, while for Cryptococcus neoformans and Candida spp. the production of an extracellular matrix represents an important pathogenicity factor (Mukherjee et al., 2005; Jain et al., 2006), in the case of P. fermentans 726 this feature may have an ambivalent role, by favouring both noninvasive surface colonization (and biocontrol), and, depending upon the substrate, pathogenic invasion of fruit tissue.

On the contrary, the pathogenicity of P. fermentans 726 on peach and nectarine fruit was always associated with pseudohyphal and filamentous growth. Many fungal pathogens are capable of reversible transition from yeast-shaped cells to a filamentous form depending on environmental stimuli. In the case of Candida albicans, dimorphic transition was shown to correlate with infectivity and tissue invasion (Sudbery et al., 2004). Mechanical penetration into solid surfaces can be more easily achieved by hyphae rather than by yeast-shaped cells, through the pressure generated by the hyphal tip on the underlying tissue (Gow et al., 2002). Moreover, the hyphal tip is the site of secretion of enzymes, which can degrade cellular components, facilitating yeast infiltration into solid substrates and tissues (Palková & Váchová, 2006). It has not yet been determined whether endopolygalacturonase or other extracellular enzymatic activities are implicated in tissue degradation of peach fruit caused by P. fermentans 726. Further investigations will be focused on the substrate factors regulating both dimorphic transition from yeast-shaped cells to filamentous growth and enzyme secretion under different conditions.

In vitro biofilm formation assays clearly indicate that P. fermentans 726 is able to undergo this phenotype in all the tests performed on both solid and liquid surfaces. Moreover, it has elevated hydrophobicity, capacity to adhere to plastic surfaces, and the ability to develop an evident mat on soft agar.

Biofilm formation is dependent on cells' ability to adhere to different surfaces. Yeast adhesion, which is conferred by specialized cell-surface proteins called ‘adhesins’ or ‘flocculins’, can be divided into two main groups: lectin-like and sugar-insensitive adhesion (Verstrepen & Klis, 2006). Adhesins of the latter group, e.g. S. cerevisiae Flo11p, increase cell surface hydrophobicity, thereby promoting hydrophobic interaction between cells and certain abiotic surfaces (polystyrene) and are involved in biofilm formation on liquid surfaces (Reynolds & Fink, 2001; Zara et al., 2005). Hydrophobicity and crystal violet assays performed with P. fermentans strain DISAABA 726 lead to the hypothesise that a sugar-insensitive adhesion mechanism is activated by this strain.

Differences observed in P. fermentans strain 726 morphology between apple and peach fruit could represent different stages of biofilm development. Indeed, studies carried out on Candida albicans show that the initial stages of biofilm formation consist of yeast-shaped cells, while in a mature biofilm hyphal elements are predominant (Malandra et al., 2003). It has also been observed that both yeast-shaped cells and hyphae are able to form a biofilm, indicating that this type of cellular aggregation is not morphology-specific (Chandra et al., 2001).

To the best of the authors' knowledge, this is the first report of a yeast that may switch from a safe and promising biological control agent to a destructive pathogen depending upon the host tissue. While this ‘Jekyll-and-Hyde’ behaviour may be quite disappointing from a phytosanitary point of view, it represents an intriguing model for studying the forces acting on the thin diaphragm that separates antagonism from pathogenicity.

Four potential adverse effects are commonly identified as safety issues associated with the use of biocontrol agents. These are: displacement of nontarget organisms, allergenicity to humans and other animals, toxicity and pathogenicity, and genetic stability (Cook et al., 1996). Except for allergenicity, all factors that enhance inoculant survival and efficacy also increase its putative adverse effect. Therefore, a minimal potential biohazard is inherent to any application of biocontrol agents (Migheli, 2001). Based on the pathogenic attitude of P. fermentans 726 on peach fruit, it is proposed that the capability to differentiate hyphae and pseudohyphae under particular growth conditions should be considered as a new potential biohazard factor for biocontrol yeasts. On the other hand, in addition to their potential pathogenicity on certain fruit commodities, isolates of P. fermentans have been recently implicated in bloodstream infections (Pfaller & Diekema, 2004), and as a cause of polyarthritis in a patient suffering from alcoholism (Trowbridge et al., 1999), further emphasizing the need for a thorough risk assessment before allowing any deliberate release of biocontrol agents.


This work was carried out within the frame of the Research Program MiPAAF – CIPE ‘FRU.MED.’– Project ‘DAFME’, publication no. 4. The authors wish to thank Dr Virgilio Balmas for help in optical microscopy observations.


View Abstract