The occurrence of killer activity in yeasts isolated from natural habitats *

Yeast’s ability to restrict the growth and kill other yeasts, fungi and bacteria has been known for over 50 years. Killer activity was detected in yeasts deposited in the world collections or isolated from natural habitats. In this study, isolates from the forest environment, leaves of fruit trees, flower petals, cereals and frozen fruit have been screened in terms of their killer activities. Killer activity was tested on strains belonging to six yeast species: Candida, Rhodotorula, Pichia, Pachysolen, Yarrowia, Trichosporon. The reference strains were Kluyveromyces lactis Y-6682 and Kluyveromyces marxinanus Y-8281, wellknown to be sensitive to yeast killer toxins. Among one hundred and two tested strains, 24 (23.5% of isolates) showed positive killer action, and 10 (9.8% of the isolates) a weak killer action against at least one sensitive reference strain. The highest killer activity was observed among isolates from forest soil and flowers.


INTRODUCTION
Many strains of yeasts secrete extracellular proteins or glycoproteins, known as killer toxins.They act by inhibiting the growth of other yeasts, fungi and even bacteria (Dabhole & Joishy, 2005;Bajaj et al., 2012).Strains capable of producing toxins may be simultaneously resistant to the killing effect.The various toxins differ in the mechanism of secretion, molecular size, optimum pH and temperature of their activity (Schmitt & Breinig, 2006).
Killer phenotype in yeast was first described in 1963 by Makower and Bevan (Makower & Bevan, 1963).Since then, killer phenomenon has been reported in almost 100 species belonging to more than 20 genera and their number is still increasing (Buzzini & Martini, 2001;Golubev, 2006).Over 11 different killer toxins are known, and they are produced by representatives of such species as Hanseniaspora, Pichia, Rhodotorula, Williopsis, Ustiliago etc. (Schmitt & Breinig, 2002;Santos et al., 2011).The killer phenomenon of yeast cells is widely distributed among strains isolated from the natural habitats: water, wine, soil, fruit, and among yeasts stored in the collections of pure cultures (Vadkertiová & Sláviková, 1995;Vadkertiová & Sláviková, 2007;de Lima et al., 2013;de Ullivarri et al., 2014).Potential use of killer yeasts and their toxins is intended for various industries.In the brewing industry, winemaking and in production of fermented vegetables, killer yeasts can be used as starter cultures to prevent infection and the development of spoilage strains that might negatively affect the sensory quality of the final products (Antonini et al., 2005;Waema et al., 2009).Killer yeasts have also been used in biological control of post-harvest diseases and have become an alternative to the use of chemical fungicides (Santos et al., 2004).Yeasts producing killer toxins may be used in medicine as novel tools against animal and human fungal infections (Magliani et al., 2004).A killer system may be also helpful in bio-typing industrially and clinically interesting yeast cultures (Ochigava et al. 2011).In addition, killer yeasts and their toxins have been used as model systems to understand the mechanisms of regulation in eukaryotic polypeptide processing and expression of eukaryotic viruses (Schmitt & Breinig, 2006).
In this study, yeast strains isolated from natural habitats (forest, leaves, fruit, cereals, flowers) were screened for their killer activity against yeast belonging to the species of Candida, Rhodotorula, Pichia Pachysolen, Yarrowia, and Trichosporon, in order to find out whether strains from these environments have similar or different spectrum of their killer activity.

MATERIALS AND METHODS
Yeast strains.One hundred and two yeast strains isolated from the forest soil, rotting trees, leaves of fruit trees and bushes, flower petals, cereals, and from frozen fruit were examined for their killer activity (Table 1).
Twelve yeast cultures belonging to six yeast species.Candida,Rhodotorula,Pichia Pachysolen,Yarrowia,Trichosporon (Table 2), maintained in the Culture Collection of the Department of Biotechnology, Human Nutrition and Science of Food Commodities University of Life Sciences in Lublin, were used as sensitive strains.The reference strains from NRRL (ARS Culture and Patent Culture Collections, US Department of Agriculture, Illinois) were Kluyveromyces lactis Y-6682 and Kluyveromyces marxianus Y-8281 well-known to be sensitive to yeast killer toxins (Vaughan-Martini et al., 1988).
Assay for killer phenotypes.The killer activity was investigated by a modified method described previously (Woods & Bevan, 1968;Vadkertiová & Sláviková, 1995;Santos et al., 2009).Sensitive strains were grown in YMB medium at 20ºC for 48 h.Each culture was mixed with YMA-MB containing 0.9% (w/v) agar, and the mixture (A 600 0.9-1.0)was poured as a lawn onto the surface of a Petri dish containing the assay medium.The plates were incubated for 2-3 h until the agar hardened.Then, wells (8 mm) were sterilely cut in the YMA-MB agar and the potential killer strains were seeded in the wells at 100μl of yeast inoculum per well.The plates were incubated at three different temperatures: 18, 22, 25ºC, for 7 days, and checked daily.A killer effect was recorded when the zone of inhibition around the tested isolates appeared on the plate.Killer activity was measured by subtracting diameter of the well from diameter of the inhibition zone.If the strain inoculated into the well was surrounded by bluish colored cells of a potentially sensitive strain, and a clear zone < 1mm or was only surrounded by a blue zone, the reaction was recorded as "w" (weak killer reaction).If the inoculated strain was surrounded by bluish colored cells and a clear zone ≥ 1mm, it was designated as "+", "++", "+++" (positive killer reaction).The experiments were performed in triplicate.

RESULTS AND DISCUSSION
The optimal temperature for killer activity of the yeast isolates was 22ºC (Fig. 1), and it is similar to what was reported previously by others (Santos et al., 2009).At this temperature, among all tested strains, twenty four showed a positive killer reaction, and ten showed a weak killer reaction against reference strains and stains used as sensitive (Table 2).This corresponded to 23.5% and 9.8% of the isolates, respectively.Results of positive and weak killer activity at 22ºC are presented in Table 3.

Species
Strain No.

Yarrowia lipolytica Y1
Trichosporon cutaneum T7 The occurrence of killer activity in yeasts of P. tannophilus.Isolates 2, 17, 20b, derived from the forest soil, cherry tree, and flowers of vibrunum, respectively, were the only ones of all tested strains that acted lethally on Y. lipolytica.Yeasts derived from grains had a much more diverse activity against susceptible strains.Most of them worked by growth inhibition of C. fluviatilis, C. parapsilosis and Rh.pallida.Two strains derived from frozen strawberry fruits, no.39 and 39a, were characterized by identical killer profile against C14, C16, Rh1, Rh5, P5, P6 and T7.The broadest and identical killer spectrum was observed for two isolates (17, 20b) from flower petals, and they displayed such activity against eight of the twelve tested strains.Nine of isolated yeasts derived from forest soil, cherry and apple trees, colza, jasmine, frozen strawberry also showed a broad spectrum of activity, but only against seven strains.The strain number 31a, isolated from cereals, limited growth of only one strain, Rh. pallida, and furthermore it only demonstrated a weak killer action.The strongest sensitivity was displayed by the Candida freyschussi strain (C16).Precent, 31, of all isolates showed positive or weak action against C16, the largest zones of inhibition (≥ 10mm) were observed when isolates 2, 21a, 29b, 32a, 39 were used as killer strains.The highest number of wide and clear zones was observed among isolates from cereals against Rh1 (Fig. 2).None of the tested strains had the ability to kill the P. stipitis yeast (P4).
Killer potential of yeast isolates was also shown by Vadkertiová & Sláviková (2007).Yeast from water, soil and leaves were able to kill the yeast from the Candida genus: C. parapsilosis, C. albicans, C. krusei, C. tropicalis.Two strains from sediments of a fresh water lake were characterized by the highest activity, and they worked on all the examined test strains.However, the greatest number of killer strains came from leafy materials.On the other hand, yeast isolates from Amazon soil tended to kill yeasts belonging to Candida, Pichia, and Debaryomyces species from the same habitat, and did not have any activity against Rhodotorula or Trichosporon (Vital et al., 2002).Studies on the activity of strains isolated from natural environments were also carried out by Mushtaq et al. (2015).Strains isolated from dairy products, flowers' nectar, slime fluxes of trees and soil were tested on sensitive Pichia strains.It has been demonstrated that the tested isolates can act as killers, be neutral or sensitive, depending on the environment from which they originate.In another study, yeast isolates from flowers of a medicinal plant have been used.They gave positive reaction against Debaryomyces anomala (Dabhole & Joishy, 2005).Sensitivity of yeasts belonging to Cryptococcus, Candida, Debaryomyces, Kluyveromyces, Pichia and Saccharomyces species was investigated by Keszthelyi et al. (2008).Rhodotorula mucilaginosa, C. krusei, C. albicans, and P. membranifaciens strains were not sensitive to isolates of Filobasidium capsuligenum, ten strains derived from a wine cellar, dried fig, sake, cider, soil and grapefruit.F. capsuligenum proved to be active only against Cryptococcus neoformans and C. laurentii strains.Sensitivity of yeast strains belonging to 21 food spoilage species of 14 genera (Candida, Debaryomyces, Dekkera, Hanseniaspora, Issatchenkia, Kazachstania, Kluyveromyces, Pichia, Rhodotorula, Saccharomyces, Schizosaccharomyces, Torulaspora, Yarrowia and Zygosaccharomyces) was observed in different studies (Goretti et al. 2009).Searching for killer strains among those naturally occurring in different environments is very important, due to their future potential use in plant protection, medicine and industry.
In addition, knowledge about the susceptibility of a given strain can be helpful in choosing the best strain for biotyping pathogenic strains or for using as the most resistant starter culture.

Figure 1 .
Figure 1.Killer activity of tested isolates at different temperatures.

Figure 2 .
Figure 2. Killer activity of some strains isolated from cereals against Rhodotorula pallida.