October 4, 2005

Apple Tissue Culture Contamination By Rhodotorula Spp.: Identification and Prevention

By Nagy, J Kolozsvri; Sule, S; Sampaio, J P


Shoot cultures of apple cv. Pinova were contaminated with faint pink pigmented yeast. Yeast isolates were identified as Rhodotorula slooffiae with standard physiological methods and molecular analysis. Growth of isolated yeasts was tested against different fungicides. The following fungicides inhibited the growth of yeast isolates, and were not phytotoxic to apple shoots at concentrations lower than the minimal phytotoxic concentrations (MPC): ProClin 300, mancozeb, triforine, myclobutanil, thiabendazole, mancozeb + zoxamid, and silver nitrate. Some fungicides inhibited growth of yeasts, but were phytotoxic. These included miconazole, PPM(TM), copper sulfate, potassium sorbate, and cycloheximide. Benomyl was not phytotoxic, but was effective only at high doses. Decontamination of shoots was achieved using a combination of two treatments. Shoots were first soaked in half-strength Murashige and Skoog (MS) liquid medium containing silver nitrate (588 μM) and Silvet 77 (0.01%) for 1-2h, and then transferred to a solidified MS medium containing both mancozeb (15 mg l^sup -1^) and thiabendazole (40 mg l^sup -1^).

Key words: isothiazolones; mancozeb; micropropagation; silver nitrate; thiabendazole; yeast.


In vitro contamination of plant tissue cultures by microorganisms is one of the most serious problems in plant micropropagation (Cassells, 1997). Cultures can be infected with a wide range of microorganisms (filamentous fungi, yeasts, bacteria, viruses, and phytoplasmas) (Leifert and Cassells, 2001). Contaminants are either present in the mother explant or introduced during inadequate handling in the laboratory. Contaminating microorganisms may be readily observed immediately or can remain latent for long periods of time. This often makes it difficult to identify the source of contamination. During the establishment of a new culture, plant meristematic tissues are generally utilized as starting material. However, these tissues are frequently colonized by bacteria, filamentous fungi, or yeasts, which may not become visible during the initial stage of culture establishment. The presence of contaminant microorganisms in tissue culture is usually attributed to insufficient surface sterilization or poor sterile techniques (Leifert et al., 1994). Besides increased culture mortality, the presence of contaminants can result in variable growth, tissue necrosis, and reduced shoot proliferation and rooting. Microbial contamination is difficult to eradicate from established cultures. Contaminants of explants grown on rich tissue culture media containing high amounts of sugar multiply in high numbers and invade internal tissues. Sterilization of such highly contaminated tissues is much more difficult than establishment of a new culture. However, in certain circumstances, a valuable and contaminated culture must be saved and cleaned from contaminants.

During the course of micropropagation of apple cv. Pinova, which had been maintained through several passages, a pink pigmented bacterium-like contaminant was found. Preliminary observations, based on microscopic examinations, suggested the presence of yeasts. Using several antibiotic treatments and repetition of surface sterilization could not eliminate the contaminants. The present study was undertaken to identify the contaminant microorganism, and to circumvent the problem.


Materials. Chemicals used in tissue culture were from Sigma- Aldrich (Budapest). The fungicides mancozeb, myclobutanil, and mancozeb + zoxamid were from Dow AgroSciences Hungary (Budapest); triforine from American Cyanamid Company (Princeton, NJ, USA); benomyl from Chinoin (Budapest); ProClin 300 from Sigma-Aldrich (Budapest), and PPM(TM) Plant Cell Technology Inc. (Washington, DC, USA).

Apple tissue culture. Shoot cultures of apple cv. Pinova used in this study were initiated from axillary and actively growing buds. Individual buds were surface-sterilized in 70% ethanol for 1 min, and in 1.0% sodium hypochlorite containing a drop of Triton X surfactant for 20 min. After rinsing in sterile water, scales and outer leaves were removed. Buds were dipped in 70% ethanol for 5- 10s, followed by continuous agitation in 10% sodium hypochlorite for 5 min, and then rinsed twice with sterile distilled water. Meristem tips (1-3 mm) were aseptically excised, and transferred to Petri dishes, each containing 20 ml of the culture medium. After 4 d, contaminant-free meristematic shoot-tips were re-transferred to fresh medium. To confirm that the disinfestation process was successful, randomly selected explants were pressed onto plates with potato dextrose agar (PDA) medium, and plates were evaluated for growth of microorganisms after incubation at 28C for 7 d.

Cultures were kept in a culture room at 25C under 16-h cool white illumination (60 μmol m^sup -2^ s^sup -1^ ). Shoots (5-10 mm) were subcultured at monthly intervals. Shoot cultures were maintained on a proliferation medium consisting of MS salts (Murashige and Skoog, 1962), and B5 vitamins (Gamborg et al., 1968), supplemented with 4.4 μM 6-benzylarninopurine (BA), 0.5 μM indole-3-butyric acid (IBA), and 88 mM sucrose. Before autoclaving, the pH was adjusted to 5.6 with 1 N KOH, and the medium was solidified with 0.4% agar and 0.1% gelrite. Proliferating shoots (15- 20 mm) were used for phytotoxicity assays.

Yeast isolation. Yeasts were isolated from visibly-contaminated shoots. Isolations from contaminated shoots were done by collecting small stem fragments (1 mm3) from the basal ends of shoots, and triturating these tissues in a drop of sterile distilled water on a glass slide. All isolations were done by plating tissue triturates on PDA, pH 5.6, containing 100 mg 1^sup -1^ chloramphenicol and incubating aerobically at 25C for 6-7d.

Sixteen individual yeast colonies were randomly selected, and subcultured three times on PDA. Isolates were stored at - 70C in 15% glycerol. Working cultures were streaked onto PDA, incubated for 3- 4d at 25C, and then transferred to fresh PDA and incubated for 2 d.

Characterization of yeasts. Yeasts were identified by morphological, physiological, and biochemical characteristics according to standard methods employed in yeast taxonomy (Yarrow, 1998). The type strain of Rhodotorula glutinis PYCC 4177 (CBS 20) was included in all investigations as a reference.

Molecular sequence analysis of the D1/D2 domains of the 26S rDNA was conducted as follows. Total DNA was isolated using the protocol of Sampaio et al. (2001) and amplified using primers ITS5 (5' GGA ACT AAA ACT CGT AAC AAG G) and LR6 (5' CGC CAG TTC TGC TTA CC). Cycle sequencing of the 600-650 base pair region at the 5' end of the D1/D2 domains employed a forward primer NIJ (5' GCA TAT CAA TAA GCG GAG GAA AAG) and a reverse primer NL4 (5' GGT CCG TGT TTC AAG ACG G). Partial sequences were obtained using an Amersham Pharmacia ALF Express II automated sequencer. For preliminary identification, a BLAST search was conducted by comparing the obtained sequence with those deposited in GenBank (NCBI, http://www.ncbi.nlm.nih.gov). Sequence alignment was conducted using MegAlign (DNAStar). The 26S rDNA partial sequence has been deposited in the GenBank under the accession number AY621081.

Fungicide treatment. Fungicidal sensitivity was determined by growing yeast isolates on corn meal agar (Difco) and MS medium amended with different fungicides. The corn meal agar contained 2% glucose and 0.05% 2morpholinoethanesulfonic acid (MES). The pH of both media was adjusted to 5.6 before autoclaving. Stock solutions of most of the tested compounds were prepared in sterile water just before pouring them into plates. Thiabendazole and miconazole were dissolved in dimethyl sulfoxide (DMSO). and the final concentration of DMSO in the medium was

The minimal phytotoxic concentration (MPC) of the tested compounds was determined in MS medium using previously determined MIC values. Young (10-15mm long) shoots were incubated for l-2h in liquid cultures containing the fungicide treatment, and then transferred to a solidified MS medium containing the same concentration of fungicide. Phytotoxicity of the compound was determined visually by checking for necrosis, browning, chlorosis, and morphological changes. Growth of yeasts in tissue culture was visually observed for 14 d. Expiants with no visible contaminant growth were removed, placed into individual tubes containing potato broth (PB) with 2% glucose, and incubated for an additional 7 d to determine if undetected yeast was present. If the broth became turbid, the suspension was diluted and plated on PDA plates containing 100 mgl chloramphenicol.

Performance of contaminated plantlets. Rooted plantlets contaminated with yeasts were gradually acclimatized and planted in soil in the greenhouse to determine if \they grew similar to control plants free of yeast infection.


Phenotypic investigation. From tissues with visible contamination, faint pink pigmented colonies appeared on PDA after 5- 7 d. To determine whether these colonies were of bacterial or yeast origin, different antibiotics were added to the medium. Surprisingly, neither cycloheximide nor chloramphenicol, each at 100 mgl^sup -1^, inhibited growth of isolated colonies. Further examination of these colonies under the microscope revealed that the contaminant organism was yeast. Cells were ovoid, and measured 3- 4x5-7 μm in size. Phenotypic selection, based on colony morphology and color, identified 16 isolates that were subsequently subjected to physiological evaluation. Using physiological tests, all of these isolates formed similar faint pink-colored colonies on PDA and corn meal agar, and had identical physiological and biochemical characters (Table 1). All isolates used glucose, D- xylose, sucrose, trehalose, glycerol, ethanol as carbon source, and urea as N source. None grew on L-rhamnose, melibiose, myo-inositol, Dl.-lactate, and starch. Based on a number of conventional criteria, these strains could be assigned to the basidiomycetous yeast genus Rhodotorula F. C. Harrison (Fell and Statzell-Tallman, 1998). Comparison of physiological and biochemical characters of apple isolates and Rhodotorula glutinis PYCC 4177 with available online database (CHS; http://www.cbs.knaw.nl/yeast/ (3w3m3i2fyv3oamehnroxv545)/ BioloMICS.aspx?Link = T&DB = O&Table = O&Descr = CBS%205 706&Fields = All&ExactMatch = T) for R. minuta and R. slooffiae, revealed several differences (Table 1). Apple isolates formed a distinct but homogeneous group, suggesting that these isolates originated from the same source, and spread from latent contaminated plantlets to new culture vessels. To determine the exact species, one of these apple isolates (Al) was subjected to molecular analysis and identified as R. slooffiae. By comparing D1/ D2 domains of the 26S rDNA, strain Al (GenBank accession number AY621081) had higher homology to R. slooffiae than to any other sequences of Rhodotorula spp., and only a single mismatch (in position 480) was found between this isolate and CBS 5706, the type strain of R. slooffiae (GenBank accession number AF189965).

Sensitivity of isolates to chemical compounds. Initial experiments showed that yeasts isolated from apple were not sensitive to ampicillin (100mgl^sup -1^), rifampicin (50mgl^sup - 1^), cefotaxime (100mgl ), carbenicillin (100mgl^sup -1^), gentamicin (15 mgl^sup -1^), chloramphenicol (100mgl^sup -1^), kanamycin (50mgl^sup -1^), streptomycin (100mgl^sup -1^), tetracycline (10mgl^sup -1^), timentin (100mgl^sup -1^), trimethoprim (30mgl^sup -1^), cycloserin (30mgl^sup -1^), potassium tellurite (100mgl^sup -1^), berberinc sulfate (500 mgl^sup -1^), and silver thiosulfate (20 mg l ).

Growth of all isolates in solidified media was inhibited (Table 2) by ProClin 300 (3 mgl^sup -1^), mancozeb (15mgl^sup -1^), mancozeb + zoxamid (50mgl^sup -1^), miconazole (20mgl^sup -1^), thiabendazole (40 mgl^sup -1^), silver nitrate (470 μM), potassium sorbate (3.3 μM), and PPM(TM) (2 ml l^sup -1^). Copper sulfate (1 mM). myclobutanil (250 mg l^sup -1^), triforine (200 mg l^sup -1^), cycloheximide (400 mg l^sup -1^), and benomyl (750 mg l^sup -1^) were less effective.

Phylotoxicity. The minimal phytotoxic concentrations (MPC) of different chemicals effective in yeast inhibition are summarized in Table 2. MIC values of ProClin 300, mancozeb, thiabendazole, silver nitrate, mycobutanil, triforin, and mancozeb + zoxamid were not phytotoxic. While effective concentrations of miconazole, potassium sorbate, and PPM(TM) were slightly phytotoxic, cycloheximide and copper sulfate were phytotoxic at MIC values. Benomyl was not phytotoxic at 750mgl^sup -1^, but its low dispersibility in the medium made it impractical for use.

Prevention and decontamination. Mancozeb (15mgl^sup -1^) has been used successfully in this study to prevent yeast contamination of apple tissue culture. To improve the elimination efficiency, we tested different combinations of fungicides (data not shown) and the following combination resulted in the highest efficacy. Thirty visibly contaminated shoots were selected and incubated for 1-2 h in halfstrength liquid MS salts (without any sugar) containing 588 μM silver nitrate and 0.01% Silvet 77 as a wetting agent for surface sterilization. Subsequently, treated shoots were inserted in semisolid complete MS medium containing mancozeb (15mgl ) and thiabendazole (40mgl ). After 14 d, none of the shoots showed contamination. Basal sections (2-3 mm) of these shoots were removed, homogenized, and plated on PDA without any fungicide. Three out of 30 shoots showed contamination, while the others were free from any contamination.





Growth of contaminated plants in the greenhouse. Contaminated and cleaned plants have grown normally, similar to control plants. No obvious differences were seen between contaminated and uncontaminated plants.


Although yeast contamination is a major problem in plant tissue culture (Leifert et al., 1990), so far, few attempts have been made to identify the species causing this contamination, and to develop means for preventing or treating contamination.

In this study, 16 isolates from apple shoot cultures were compared using physiological tests. All isolates demonstrated similar patterns of growth, indicating that contamination originated from the same source. Sequence analysis of the D1/D2 domains of the 26S rDNA revealed that the yeast contaminant belonged to the species R. slooffiae. This yeast was first described in 1962 in Hungary (Novk and Vros-Felkai, 1962); however, there were no reports on the occurrence of this yeast in plant tissue culture. Until recently, R. slooffiae has been considered a synonym of R. minuta (Fell and Statzell-Tallman, 1998). Thus, it is likely that these two species have long been present, but remained undetected.

Rhodotorula spp. are widely distributed in nature, and they are generally found on green leaves, fruits, and buds (Matteson Heidenreich et al., 1997; Buck, 2002), as well as in laboratory air. Previously, R. minuta, a species closely related to R. slooffiae, has been detected on apple trees (Chand-Goyal et al., 1996). This species has been often detected as an endophyte, growing vigorously within meristematic tissues, such as pine buds growing in vivo (Pirttila et al., 2003). The origin of tissue culture contamination in this study is most likely attributed to latent infected apple buds. It is conceivable that under natural conditions, these yeasts are under strict growth control of the host plant. However, once an in vitro culture is initiated from these buds, these yeast colonies begin to thrive and grow.

Various compounds have been described for their antimycotic activity against a broad range of fungi. Many of these compounds are potentially useful in yeast elimination from tissue culture. However, plants are more sensitive to the phytotoxie effects of chemicals in vitro, than in vivo. Therefore, prior to using any compound, the phytotoxie activity and yeast-killing activity of this compound should be tested. To eliminate contamination from apple shoot cultures, different antibiotics and fungicides have been evaluated. The best fungicides without phytotoxie activities included ProClin 300, mancozeb, thiabendazole, silver nitrate, and mancozeb + zoxamid. All other compounds are either phytotoxie, at fungicidal doses used, or ineffective against yeast isolates.

ProClin 300 is a preservative for the control of microorganisms in diagnostic reagents. It penetrates cell membranes and inhibits enzymatic activities in the Krebs cycle. At 3mgl^sup -1^, ProClin 300 was harmless to apple shoots, yet it killed all R. slooffiae isolates. PPM(TM) has been reported to prevent microbial contamination in tissue culture (Guri and Patel, 1998). To prevent R. slooffiae infection, PPM(TM) was used at 2 ml 1^sup -1^; however, this concentration was slightly phytotoxie to apple shoots. Miconazole, an ergosterol biosynthetic inhibitor, has been shown to be an effective antifungal agent in plant tissue cultures (Tynan et al., 1993) as it inhibited growth of different yeasts such as Candida spp. and Cryptococcus neoformans. However, miconazole was phytotoxie to apple shoots. Mancozeb, a protective contact fungicide effective against a wide range of foliage fungal diseases (Worthing, 1987) was highly active against R. slooffiae. At 15mgl , mancozeb prevented yeast growth on apple shoots. As the phytotoxie concentration of mancozeb was high (>100mgl ), this would allow its use for elimination of microorganisms from various plant tissue cultures.

Thiabendazole is mainly used against post-harvest diseases (Penicillium, Mucor, and Rhizopus) of oranges and other tropical fruits, while benomyl, another benzimidazole, is less effective as it only eliminates yeast infections at very high concentrations (750 mgl^sup -1^). This may be due to its low solubility and dispersibility. Potassium sorbate is used as a yeast inhibitor (Stratford, 1998), and it is only active at low pH. As the pH of tissue culture media is around 5.5, potassium sorbate is a good candidate against yeasts. However, in this study, potassium sorbate demonstrated inconsistent fungicidal effects. At 500 mgl , potassium sorbate was phytotoxic to apple shoots, and did not control yeast contamination. Therefore, additional experiments must be conducted to determine the appropriate usefulness of this compound in \tissue culture. Silver nitrate is reported to be beneficial for embryogenesis, shoot proliferation, and rooting of apple shoots (Ma et al., 1998). Its antimicrobial effect has long been known. Although it has been used to eliminate Agrobacterium turnefadens from tissue culture, its use as fungicide in plant tissue culture has not been reported until now; whereas, silver thiosulfate, a derivative of silver nitrate, is widely used in plant tissue culture as an inhibitor of ethylene biosynthesis. In this study, silver thiosulfate did not have any anti-yeast activity.

In conclusion, the endophytic colonization of apple shoots by yeasts made it difficult to eliminate yeasts by simple decontamination or fungicide treatment, but the severity of the problem could be reduced with the use of fungicides. Based on this study, several fungicides appeared to be effective against yeasts at concentrations that were not toxic to apple shoot culture. Further studies might reveal other plant protection compounds that could be useful against fungal infections in plant tissue culture. All apple shoot cultures used in this study now have been cleaned from yeasts by proper selection of expiants, surface sterilization with a silver nitrate solution, and growing shoots in a solidified MS medium containing both mancozeb and thiabendazole.

Both contaminated and uncontaminated plants grew well following transfer to the greenhouse, indicating that R. slooffiae is not harmful to apple plants grown in the field.


This work was partially supported by the Hungarian Scientific Research Fund OTKA T 042465. J. P. Sampaio was supported by grant POCTI/AGR/44 310/2002 (FCT, Portugal). The authors wish to thank Prof. L. Vajna for the help with microscopic examinations.


Buck, J. W. In vitro antagonism of Botrytis cinerea by phylloplane yeasts. Can. J. Bot. 80:885-891; 2002.

Cassells. A. C. Pathogen and microbial contamination management in micropropagation. Dordrecht: Kluwer Academic Publishers; 1997.

Chand-Goyal, T.; Spotts, R. A. Enumeration of bacterial and yeast colonists of apple fruits and identification of epiphytic yeasts on pear fruits in the Pacific Northwest United States. Microbiol. Res. 151:427-432; 1996.

Fell, J. W.; Statzell-Tallman, A. Rhodotorula F. C. Harrison. In: Kurtzman, C. P.; Fell, J. W., eds. The yeasts, a taxonomic study. Amsterdam: Elsevier Science; 1998:77-100.

Gamborg, O. L.; Miller, R. A.; Ojima, K. Nutrient requirements of suspension cultures of soybean root cells. Exp. Cell Res. 50:151- 158; 1968.

Guri, A. Z.; Patel, K. N. Compositions and methods to prevent microbial contamination of plant tissue culture media. US Patent 5,750,402; 1998.

Leifert, C.; Cassells, A. C. Microbial hazards in plant tissue and cell cultures. In Vitro Cell. Dev. Biol. Plant 37:133-138; 2001.

Leifert, C.; Morris, C. E.; Waites, W. M. Ecology of microbial saprophytes and pathogens in tissue culture and field-grown plants: reasons for contamination problems in vitro. Crit. Rev. Plant Sci. 13:139-182; 1994.

Leifert, C.; Waites, W. M.; Nicholas, J. R.; Keetley, J. W. W. Yeast contaminants of micropropagated plant cultures. J. Appl. Bacteriol. 69:471-476; 1990.

Ma, J. H.; Yao, J. L.; Cohen, D.; Morris, B. Ethylene inhibitors enhance in vitro root formation from apple shoot cultures. Plant Cell Rep. 15:87-90; 1998.

Matteson Heidenreich, M. C.; Coral-Garcia, M. R.; Momol, E. A.; Burr, T. J. Russet of apple fruit caused by Aureobasidium pullulans and Rhodotorula glutinis. Plant Dis. 81:337-342; 1997.

Murashige, T.; Skoog, F. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15:473-497; 1962.

Novk, E. K.; Vrs-Felkai, G. Rhodoturala slooffii n. sp Acta Microbiol. Acad. Sci. Hung. 9:261-263; 1962.

Pirttil, A. M.; Pospiech, H.; Laukkanen, H.; Myllyla, R.; Hohtola, A. Two endophytic fungi in different tissues of Scots pine buds (Pinus sylvestris L). Microb. Ecol. 45:53-62; 2003.

Sampaio, J. P.; Gadanho, M.; Santos, S.; Duarte, F.; Pais, C.; Fonseca, A.; Fell, J. W. Polyphasic taxonomy of the genus Rhodosporidium: R. kratochvilovae and related anamorphic species. Int. J. Syst. Evol. Microbiol. 51:687-697; 2001.

Stratford, M. Evidence that sorbic acid does not inhibit yeast as a classic 'weak acid preservative.' Letters Appl. Microbiol. 27:203- 206; 1998.

Tynan, J. L.; Conner, A. J.; Macknight, R. C.; Poulter, R. T. M. Miconazole: an effective antifungal agent for plant tissue culture. Plant Cell Tiss. Organ Cult. 32:293-301; 1993.

Worthing, C. R., ed. The pesticide manual: a world compendium, 8th edn. Croydon: British Crop Protection Council; 1987.

Yarrow, D. Methods for the isolation, maintenance and identification of yeasts. In: Kurtzman, C. P.; Fell, J. W., eds. The yeasts, a taxonomic study. Amsterdam: Elsevier Science; 1998:77- 100.


Plant Protection Institute of the Hungarian Academy of Sciences, 1525 Budapest, P.O. Box 102, Hungary

2 CREM, SABT, Faculdade de Ciencias e Tecnologia, Universidade Nova de Lisboa, Lisbon, Portugal

(Received 13 July 2004; accepted 23 February 2005; editor S. S. Korban)

* Author to whom correspondence should be addressed: Email [email protected]

Copyright Society for In Vitro Biology Jul/Aug 2005