Biocontrol of phytopathogenic fungi relevant for cacao- and olive-derived economy

DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contatar o autor, através do RepositóriUM da Universidade do Minho. Atribuição-NãoComercial-SemDerivações CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/ iii ACKNOWLEDGEMENTS Em primeiro lugar, quero agradecer à Professora Cândida Lucas, por me ter acolhido no seu grupo de investigação e por me ter dado a conhecer o tema que me apaixonou desde o primeiro minuto. Agradeço todo tempo, paciência e conselhos valiosos, principalmente nos momentos difíceis, que muito me ajudaram a conseguir alcançar este marco na minha vida pessoal e profissional. À Professora Fernanda Cássio, por toda a ajuda, disponibilidade e pelas palavras de incentivo ao longo destes quatro anos. A todos os colegas do Laboratório de Biodiversidade e do LBM II por toda a ajuda e boa disposição, em especial às minhas parceiras de bancada. À Giulia e à Joana, por me terem ajudado desde o primeiro minuto em tudo e me terem introduzido ao mundo da microbiologia, por toda a alegria e amizade. À Mariana e à Coralie, as minhas pupilas, por toda a ajuda em tantos momentos essenciais, principalmente nos mais complicados e por toda a boa disposição que tornaram o ambiente no laboratório tão bom. A todos os colegas do Programa Doutoral, em especial à Guida, Mário, Rosana e Rui por terem partilhado comigo o estatuto de cobaias e por todo o companheirismo que sempre demonstraram. A todas as pessoas que partilharam comigo a hora de almoço e as pausas para café nestes quatro anos, em especial à Catarina, Cláudia, Dário, Diana, Gabriel, Giulia, Joana Pereira, Joana Tulha, Maria, Mário e Rosana, por todos os momentos bons e por serem amigos que ficarão para a vida. A todas as pessoas do Departamento de Biologia e do IB-S que, de uma forma ou de outra, me ajudaram quando necessário. Um agradecimento especial à Lídia, Manuela e Sr. Luís por toda a disponibilidade. Aos meus pais e irmãos, por terem tornado possível este meu sonho e por me terem sempre apoiado incondicionalmente ao longo da minha vida. Muito obrigado por tudo! À Catarina, por todo o amor, carinho e paciência, principalmente por ter aturado estes últimos tempos stressantes da escrita da tese, eu sei que não foram fáceis. Obrigado por estares sempre lá para tudo! This work was supported by the strategic programme UID/BIA/04050/2013 (POCI-01-0145-FEDER-007569) funded by national funds through the FCT I.P. and by the ERDF through the COMPETE2020 - POCI, and the project EcoAgriFood (NORTE01-0145-FEDER-000009), supported by the NORTE 2020 under the PORTUGAL 2020 Partnership Agreement through the European Regional Development Fund (ERDF). PF is a student of the Doctoral Programme in Applied and Environmental Microbiology (DP_AEM) and FCT grantee PD/BD/113810/2015. We thank Mariana Amorim-Rodrigues (FCT grant PD/BD/145354/2019) for the precious contributions with her research to the Chapter 4. We thank the Brazilian companies, Cerlev, Lda., Ouro Preto, MG, and Fermentec, Lda. Soluções Tecnológicas e Industriais, Piracicaba, SP, as well as Professor João Paulo Sampaio from the Portuguese Yeast Culture Collection/UCIBIO, NOVA, Portugal for kindly supplying the yeast strains used in this work. We also thank Professor Pedro Talhinhas, from LEAF - Linking Landscape, Environment, Agriculture and Food, ISA, Universidade de Lisboa, Portugal, for kindly supplying two of the fungal strains used in this work. iv STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. v ABSTRACT Biocontrol of phytopathogenic fungi relevant for cacaoand olive-derived economy Plant diseases caused by fungal pathogens constitute an increasing threat to food production and security. This is the case of Witches’ Broom Disease (WBD) of cacao, caused by the basidiomycete Moniliophthora perniciosa , and of Olive Anthracnose (OA), caused by a consortium of Colletotrichum gloeosporioides and C. acutatum species complexes, which seriously impact the economy. Both share the lack of effective containment methods. The use of chemical fungicides is increasingly more precarious and restricted. More effective eco-friendly methods are required, such as the use of biocontrol agents. This work aimed to evaluate the ability of yeasts to antagonize the fungal causal agents of both diseases. A group of yeasts originating from Brazilian sugarcane-based fermentation industries were tested against M. perniciosa . Two isolates from cachaça production fermentation and one of the most used strains in bioethanol production, efficiently kill six strains of M. perniciosa in vitro . Antagonistic ability is maintained in non-optimal conditions. Microscopy analysis showed that fungal cells die upon contact with these yeasts, which physically attach and fuse to the mycelium and drain the cells, in what appears like a yeast predacious behaviour. Two further observations revealed the formation of connections between yeast and fungal cells, as well as fimbriae-like connections between yeast cells. These results agree in that, at a distance, only the proliferation of mycelia is diverted through the probable secretion of a non-volatile agar diffusible compound. Several proteins were identified specifically secreted by fungi and yeasts in single culture and co-culture. The methodology now developed was further used to study the antagonism of yeasts against the OA-causing Colletotrichum sp.. The same group of fermentative yeasts were tested along with others originating from the olive biome in Portuguese orchards. One of these last strains showed a promising antagonistic potential, by inhibiting the growth of the three fungal strains in all the conditions tested. Moreover, the possibility of using vinasse , a waste product from sugarcane bioethanol production process, to control the development of M. perniciosa was also evaluated. Immersing or spraying the mycelium with vinasse either kills the fungus or impedes its proliferation. This effect is not extensive the OA-causing fungi. The results support the exploration of microbial biodiversity of the infected plant-associated biome to generate greener and more sustainable alternatives to contain phytopathogens like those causing WBD and OA, contributing to alleviate the socio-economic impact of these diseases. Keywords: Antagonism; Biocontrol; Olive Anthracnose; Witches’ Broom Disease ; Yeasts vi RESUMO Biocontrolo de fungos fitopatogénicos importantes para a economia do cacau e da oliveira Fitopatologias causadas por fungos são uma ameaça crescente à produção e segurança alimentares. É o caso da doença da Vassoura-de-bruxa (VB) do cacau, causada pelo fungo Moniliophthora perniciosa , e da Antracnose da oliveira (AO), causada pelos complexos de espécies Colletotrichum gloeosporioides e C. acutatum , que se tornaram um risco para a economia, por não haver métodos eficazes de contenção. O uso de fungicidas químicos é cada vez mais ineficaz e mais restrito, pelo que é necessário encontrar alternativas mais amigas do ambiente, como por exemplo a utilização de biocidas. Este trabalho teve como objetivo avaliar a capacidade de leveduras para antagonizar os fungos causadores destas doenças. Um grupo de leveduras proveniente da fermentação de caldo-de-cana no Brasil foi testado contra M. perniciosa. Dois isolados do processo de produção de cachaça e uma das leveduras mais utilizadas na produção industrial de bioetanol matam eficientemente 6 estirpes de M. perniciosa , in vitro . O antagonismo mantém-se em condições adversas. Observações de microscopia confirmaram que o fungo morre em contacto com as leveduras que aderem e se fundem com o micélio, esvaziando-o, no que parece ser um comportamento predatório. Além disso, observou-se a existência de conexões entre células de levedura e do micélio, e entre células de levedura entre si. Isto concorda com o facto de, à distância, haver apenas redireccionamento do desenvolvimento do micélio, através da difusão de um composto não-volátil, tendo sido verificado que os fungos e as leveduras segregam proteínas diferentes em cultura simples ou em co-cultura. O método desenvolvido neste trabalho foi aplicado para verificar o antagonismo de leveduras contra os agentes causadores da AO. O mesmo conjunto de leveduras fermentativas foi testado, em confronto com estirpes provenientes do bioma de olivais portugueses. Uma destas últimas leveduras mostrou grande potencial antagonista contra 3 estirpes de Colletotrichum sp. em todas as condições testadas. Adicionalmente, foi ensaiada com sucesso a utilização da vinhaça, o resíduo denso e líquido que resulta da destilação do etanol após a fermentação do caldo-de-cana, no controlo do desenvolvimento do M. perniciosa . A imersão ou a pulverização do micélio com vinhaça mata ou impede drasticamente a proliferação do fungo. Este efeito não foi extensível aos fungos causadores da AO. Em suma, a exploração da biodiversidade microbiana do bioma associado às plantas infetadas pode permitir desenvolver metodologias alternativas, mais amigas do ambiente e mais sustentáveis, para a contenção da presente progressão dos agentes fitopatogénicos, como os que provocam a VB e a AO, contribuindo para aliviar o impacto socioeconómico destas doenças. Palavras-chave: Antagonismo; Antracnose da Oliveira; Biocontrolo; Leveduras; Vassoura-de-bruxa vii TABLE OF CONTENTS ACKNOWLEDGEMENTS ............................................................................................................................ .iii STATEMENT OF INTEGRITY ........................................................................................................................ iv ABSTRACT ............................................................................................................................................. v RESUMO .............................................................................................................................................. vi LIST OF FIGURES .................................................................................................................................... xi LIST OF TABLES .................................................................................................................................... xiv LIST OF COMMON ABBREVIATIONS ............................................................................................................ xvi LIST OF PUBLICATIONS .......................................................................................................................... viii CHAPTER 1 - General Introduction INTRODUCTION ...................................................................................................................................... 2 WITCHES’ BROOM DISEASE OF CACAO ........................................................................................................ 3 OLIVE ANTHRACNOSE .............................................................................................................................. 7 FUNGAL DISEASES MANAGEMENT ............................................................................................................ 12 YEASTS AS BIOCONTROL AGENTS .............................................................................................................. 16 Killer toxins .................................................................................................................................. 17 Lytic enzymes and mycoparasitism ............................................................................................... 20 Volatile compounds ...................................................................................................................... 20 Competition for nutrients and space ............................................................................................. 21 Quorum sensing ........................................................................................................................... 21 Environmental variables ................................................................................................................ 22 ADVANTAGES AND CHALLENGES OF MICROBIAL WARFARE ............................................................................... 23 THESIS OUTLINE ................................................................................................................................... 26 REFERENCES ....................................................................................................................................... 29 xiv LIST OF TABLES CHAPTER 1 - General Introduction Table 1. Success cases of using yeasts to antagonize the spoilage of fruits by filamentous fungi ...... 18 CHAPTER 2 - Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease Table 1. Strains of Moniliophthora perniciosa from CBS-KNAW (www.wi.knaw.nl), their primitive origin, and the assigned letter code used in this work .................................................................................. 56 Table 2. Yeast strains used in this work, their primitive origin, and their assigned code used ............ 57 Table 3. Results of the antagonism assays between M. perniciosa and yeasts in solid medium ........ 63 Table 4. Antagonism assays in liquid medium using the combinations between the fermentative yeast strains and M. perniciosa strains F and G.......................................................................................... 67 Table 5. Antagonism assays using selected combinations of yeast and fungal strains, challenged with environmental stressors. ................................................................................................................... 72 CHAPTER 4 - Wickerhamomyces anomalus from olive orchards microbiome efficiently antagonizes the Olive Anthracnose’s causal agents, Colletotrichum spp. Table 1. Phytopathogenic filamentous fungi and yeast strains and their origin ................................ 107 Table 2. Results of antagonism assays in solid medium, using the yeast strains from Table 1 against Colletotrichum gloeosporioides s.s., C. godetiae and C. nymphaeae ................................................ 109 Table 3. Antagonism between the OA causative agents Colletotrichum gloeosporioides s.s., C. godetiae and C. nymphaeae and the yeast strains in Table 1: results are percentage of inhibition in solid media ......... 110 xv Table 4. Antagonism between the phytopathogenic fungal strains Colletotrichum gloeosporioides s.s., C. godetiae and C. nymphaeae and the yeast strains in Table 1 in liquid media. .............................. 112 Table S1. Growth rates of the phytopathogenic fungal strains Colletotrichum gloeosporioides s.s., C. godetiae and C. nymphaeae at 25 and 30 °C, in two culture media and at different pH values. ....... 115 SUPPLEMENTARY MATERIAL Table 1. Proteins from yeasts/fungal strains co-cultures and control fungal single culture. SDS-PAGE bands were analyzed by PMF and fragmentation. ............................................................................ 149 xvi LIST OF COMMON ABBREVIATIONS µg Growth Rate (yeast) ANOVA Analysis of Variance AOX Alternative Oxidase ATP Adenosine Triphosphate COD Chemical Oxygen Demand cyt c Cytochrome c DNA Deoxyribonucleic Acid dsDNA Double-stranded Deoxyribonucleic Acid Eno2 Enolase 2 ER Endoplasmic Reticulum GAPDH Glyceraldehyde 3-phosphate dehydrogenase Gd Growth diameter Gr Growth Rate (fungi) GRAS Generally Recognized As Safe kDa kilodalton Kre1 Killer toxin Resistant 1 KT Killer Toxin MB Methylene Blue MDA Malondialdehyde ME Malt Extract MEA Malt Extract-Agar MOX Methanol Oxidase OA Olive Anthracnose OD Optical Density PBS Phosphate Buffered Saline PDA Potato-Dextrose-Agar PI Propidium Iodide PMF Peptide Mass Fingerprinting QSI Quorum Sensing Inhibitor QSM Quorum Sensing Molecule xvii RNA Ribonucleic Acid ROS Reactive Oxygen Species rpm Rotations per minute RPS Ribosomal Proteins of the Small subunit Scw4 Soluble Cell Wall 4 SDS Sodium Dodecyl Sulphate SDS-PAGE Sodium Dodecyl Sulphate - Polyacrylamide Gel Electrophoresis SEM Scanning Electron Microscopy SOD Superoxide Dismutase TCA Tricarboxylic Acid Tdh3 Triose-phosphate Dehydrogenase 3 TEM Transmission Electron Microscopy v/v volume per volume V-ATPase Vacuolar-type H+-ATPase VOC Volatile Compound w/v weight per volume WBD Witches’ Broom Disease Ynk1 Yeast Nucleoside diphosphate Kinase 1 YPD Yeast Extract-Peptone-Dextrose YPDA Yeast Extract-Peptone-Dextrose-Agar xviii LIST OF PUBLICATIONS The work performed during this PhD resulted in the following peer-reviewed publications: Ferraz, P., Cássio, F. and Lucas, C. (2019) Potential of Yeasts as Biocontrol Agents of the Phytopathogen Causing Cacao Witches’ Broom Disease : Is Microbial Warfare a Solution? Frontiers in Microbiology 10: 1766 Ferraz, P., Amorim-Rodrigues, M., Cássio, F. and Lucas, C. Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease (submitted) Ferraz, P., Cássio, F. and Lucas, C. Microscopic assessment of the antagonism effect of yeasts against Moniliophthora perniciosa , the fungal causal agent of Witches’ Broom Disease in cacao (manuscript in preparation) Amorim-Rodrigues, M., Ferraz, P., Cássio, F. and Lucas, C. Wickerhamomyces anomalus from olive orchards microbiome efficiently antagonizes the Olive Anthracnose’s causal agents, Colletotrichum spp. (manuscript in preparation) Ferraz, P., Amorim-Rodrigues, M., Cássio, F. and Lucas, C. Vinasse waste from sugarcane-based bioethanol production plants kills Moniliophthora perniciosa, the causative agent of cacao Witches’ Broom Disease . (submitted) CHAPTER 1 General Introduction Part of this chapter has been previously published in: Ferraz, P., Cássio, F. and Lucas, C. (2019) Potential of Yeasts as Biocontrol Agents of the Phytopathogen Causing Cacao Witches’ Broom Disease : Is Microbial Warfare a Solution? Frontiers in Microbiology 10: 1766 2 INTRODUCTION A great number of plant diseases are responsible for major crop losses with huge socio-economic impact, causing each year a worldwide estimated losses of 40 billion dollars (Syed Ab Rahman et al. , 2018). Particularly, the diseases caused by fungal pathogens are increasingly recognized as a global threat to food production and security. In fact, since 2000 the number of new fungal plant pathogen alerts has increased by more than 7-fold (Fisher et al. , 2012). Presently, fungal-generated diseases constitute 6467% of the total crop diseases reported globally (Fisher et al. , 2012; Fisher et al. , 2018), and account for 20% of the losses at the level of production and a further 10% at postharvest level (Fisher et al. , 2018). These authors estimate that fungal diseases are spreading northbound at a rate of almost 8 km/year. This could derive from increasingly common agricultural practices, such as the extensive monocultures and the use of a restricted number of plant cultivars, as well as the increased global international trade proportionating disease spreading over great distances (Fisher et al. , 2012). Climate change adds a burden to that equation, potentiating the development of microbes and vectors in unprecedented regions (Robert et al. , 2015; Fisher et al. , 2018). To prevent plant diseases and protect crops from pests and pathogens, widely spread methodologies mainly correspond to applying chemical fungicides. The continued use of chemical fungicides leads to the development of fungicide resistance in the fungal pathogen (Syed Ab Rahman et al. , 2018) and, in the absence of other control measures, to the re-emergence of virulence (Fisher et al. , 2018). Therefore, in spite that the use of pesticides brought clear improvements in crop quality and quantity during more than half a century, their progressive inefficacy to treat some of the most harmful plant diseases requires the utilization of higher dosages each year (Medeiros et al. , 2010; Syed Ab Rahman et al. , 2018). The use of fungicides heavily impacts on the microflora of agrarian ecosystems, destroying beneficial microbes, such as endophytic bacteria and fungi, as well as animals important for the quality of the soils (Syed Ab Rahman et al. , 2018). Ultimately, the systemic use of these drugs leads to the persistence of chemical residues in the environment, proportionating low dosage toxicity and contaminating species across trophic levels (Carvalho, 2006; Dukare et al. , 2018). Due to this scenario, new efficient and eco-friendly strategies to control fungal diseases are required, such as the use of biological control agents, particularly antagonistic yeasts. This possible solution offers some advantages comparing to the use of chemical fungicides, such as safer methods of application and the fact of being environmentally friendly (Dukare et al. , 2018). The introduction of the yeast-based biocontrol products does not harm the ecosystems due to the absence of toxic residues, and the low levels of toxicity 3 in combination with the high level of biodegradability make them suitable sustainable agriculture procedures and for human consumption (Ocampo-Suarez et al. , 2017). Besides the use in the food and agriculture, antagonistic yeast can also have an important role in medical applications, including its use as antimycotics for therapeutic treatment of human fungal infections, e.g., in the combat of pathogenic Candida infections (Hatoum et al. , 2012) WITCHES’ BROOM DISEASE OF CACAO One of the fungal diseases with recognized high negative socio-economic impact is the pathology of cacao plant and fruit known as Witches’ Broom Disease (WBD). This is caused by the basidiomycete fungus Moniliophthora perniciosa (Aime and Phillips-Mora, 2005) (formerly designated Crinipellis perniciosa ). The severity and extent of its manifestation is endangering the rapidly expanding and very qualitydemanding chocolate market. According to data from the International Cacao Organization (www.ICCO.org), more than 4 million tons of cacao beans are produced annually (Wickramasuriya and Dunwell, 2018). Cacao beans are the core raw material for the chocolate industry, although other cacaoderived products also have important world markets, such as cacao butter or liquor (Pohlan and Pérez, 2010; Wickramasuriya and Dunwell, 2018). The economic global market for chocolate reached US$ 110 billion in 2015, and the world demand is expected to grow exponentially in the next decade due to the globalization of consumption styles in expanding economies such as China and India (Squicciarini and Swinnen, 2016). Cacao is produced in countries located approximately in the same latitude interval of equatorial climate, forming the so-called Cacao Belt (Pohlan and Pérez, 2010). The biggest producers are therefore countries from Central and South America and Africa. The cacao plant is affected by several diseases, the more threatening of which is WBD (Purdy and Schmidt, 1996; Pereira, 1999; Griffith et al. , 2003; Aime and Phillips-Mora, 2005; Teixeira et al. , 2015) (Figure 1). It has severely affected South and Central America countries, where it has been responsible for major irreversible crop losses. The highest economic and social consequences of WBD are described to have occurred in Brazil. In the ten years after the onset of the disease in 1989, WBD reduced the cacao production in more than 70% (Pereira et al. , 1989; Santos Filho et al. , 1998; Trevizan and Marques, 2002; Meinhardt et al. , 2008; Pires et al. , 2009; Teixeira et al. , 2015), causing Brazil to shift from being the 2nd world producer to becoming a net importer of cacao beans (Bowers et al. , 2001; Marelli et al. , 2009; Teixeira et al. , 2015). During that 4 period, the most affected region of Bahia suffered losses around 90%, configuring a severe social crisis from losing more than 200,000 farm jobs (Trevizan and Marques, 2002; Teixeira et al. , 2015). Figure 1. M. perniciosa basidiocarps and the released spores on a dry broom of a cacao plant (A), a healthy cacao pod (B) and the damage caused by the pathogen in the infected cacao pods (B). Adapted from Meinhardt et al. , 2008 (A) and Bowers et al. , 2001 (C). The severity of WBD derives from several factors. Moniliophthora perniciosa does not form specialized infection structures such as appressoria like other fungal pathogens. Since it is a hemibiotrophic fungus, the full infectious cycle unfolds through two distinct phases: (i) biotrophic and (ii) saprotrophic. (i) The initial infection occurs in young meristematic tissues and susceptible actively growing tissues (e.g. buds, young leaves, flower cushions, young fruits). The fungus penetrates through the stomatal openings, the bases of damaged trichomes and the husk of young fruits (Aime and Phillips-Mora, 2005). After the initial infection, the fungus induces hypertrophy and hyperplasia, causing the loss of apical dominance. This corresponds to a disorganized proliferation of the infected vegetative meristems of axillary shoots that results in the formation of a broom-like structure of abnormal stems called a green broom (Aime and Phillips-Mora, 2005; Meinhardt et al. , 2008; Pires et al. , 2009). Shortly after the initial infection, the fungus starts growing intercellularly, forming a monokaryotic and parasitic mycelium without clamp connections, establishing a biotrophic relationship with the host that corresponds to its life cycle biotrophic phase. (ii) Usually 4 to 6 weeks after the development of the green brooms , a concerted series of infected plant cells death events occurs, and the infected tissues become necrotic forming a structure called dry A C B 5 broom . Necrotic or dead host cells are then colonized by the fungus (Evans, 1980; Meinhardt et al. , 2008), which at this point, suffers major morphological changes entering its saprotrophic phase (Lawrence et al. , 1991; Meinhardt et al. , 2008). The hyphae become dikaryotic, clamp connections are formed, and the fungus begins to grow intracellularly, as well as between cells. The exact mechanisms and signaling factors that trigger the switch from the biotrophic phase to a saprotrophic phase, controlling the developmental alterations, remain unknown (Meinhardt et al. , 2008). After the fungus proliferation and colonization of the dead host tissues, pink-colored basidiocarps (small mushrooms) are produced on any infected necrotic tissue. Upon alternate wet and dry periods, each basidiocarp can produce 2 to 3.5 million spores (basidiospores), this way completing the fungus life cycle (Rocha and Wheeler, 1985; Almeida et al. , 1997). The release of the spores occurs mainly at night and is related to a high level of humidity and favorable temperature (20-30 ºC). The spores are disseminated locally by water and over long distances by wind and can endure latent in the soil or inside pruned branches of the plants for long periods (Meinhardt et al. , 2008; Pohlan and Pérez, 2010). An overview of the disease progression in the plant and the parallel events of the fungus life cycle described above are summarized in Figure 2. The ability of Moniliophthora perniciosa to infect the plant in all stages of its life-cycle and the fact that virtually all the plant tissues can be infected, underlie this pest exceptional virulence. This, allied to the fungus high prevalence in the soil and plant dead material, explains why once a single plant develops symptoms the whole plantation can be compromised. The resilience of Moniliophthora perniciosa relies essentially on its capacity to colonize both alive and dead plant tissue, the biotrophic and necrotrophic life cycle phases above mentioned. The shift between the two phases involves a drastic morphological and lifestyle change. Alternative Oxidase (AOXp)- respiration was associated with this transition (Thomazella et al. , 2012). Possibly, this type of respiration allows the fungal cell to overcome the plant host defenses generated in the first stages of the WBD, as observed with better studied model fungus Ustilago maydis (Cárdenas-Monroy et al. , 2017). The plant defenses include the production of high amounts of NO, which affect the fungal mitochondria, namely inhibiting respiration complex IV, this way inducing the production of ROS (Thomazella et al. , 2012). AOXp is an alternative mitochondrial oxidase that constitutes alone a bypass to respiratory chain complexes III and IV, which function prevents collapse from drugs that target these complexes like cyanide or AntimycinA (Maxwell et al. , 1999; Ruy et al. , 2006; Vanlerberghe et al. , 2009). At the same time, AOXp-respiration contributes to cope with the electron flux overflow without phosphorylation, and therefore without producing ATP (Van Aken et al. , 2009) lowering the global energy yield of metabolism. AOX-encoding gene sequences are found in many organisms (including yeasts and fungi) (Elthon and McIntosh, 1987; 12 epidemic (Cacciola et al. , 2012). Moreover, the severity of the disease varies accordingly to phytopathogens that are infecting the olive orchards (Talhinhas et al. , 2018). C. acutatum s.s. and C. nymphaeae are found to be more virulent, therefore more frequently related with epidemic outbreaks (Schena et al. , 2014; Talhinhas et al. , 2015). Fruit ripeness is also a crucial factor, since the ripening rates are higher and faster in trees with fewer fruits, consequently increasing the host susceptibility. Since an epidemic outbreak of OA decreases the fruit load in the subsequent year, the probability of another epidemic event increases, contributing to the severity of this disease (Moral and Trapero, 2012). The necrotrophic stage of the pathogens often causes fruit rot, which in turn results in premature fruit mummification. The majority of the mummified drupes fall to the ground and are easily decomposed, while the few that remain on the tree are capable of releasing viable conidia at a constant rate during several months, probably serving as inoculum for spring infections (Moral and Trapero, 2012; Sergeeva, 2014; Talhinhas et al. , 2018). These latent spring infections can, in turn, serve as the main inoculum source for autumn infections (Moral et al. , 2009; Talhinhas et al. , 2011). It has also been reported that in controlled conditions, one infected fruit per tree can affect up to 100% of the fruits in a susceptible cultivar if favorable weather conditions persevere during autumn (Moral et al. , 2009). All these factors combined explain why once a single plant in an olive orchard starts to develop symptoms, the whole plantation can be endangered. FUNGAL DISEASES MANAGEMENT The chemical fungicides generally used are either copper-based compounds, such as cuprous oxide, or azole-containing molecules, particularly tebuconazole (Oliveira and Luz, 2005; Medeiros et al. , 2010). These are usually used to control the spread of other fungal plant diseases, such as grapevine downy mildew and olive peacock spot, but showed very low efficiency against M. perniciosa (Medeiros et al. , 2010) .as well as the above described C. gloeosporioides and C. acutatum species complexes (Cacciola et al. , 2012). Copper is per se a non-specific anti-microbial agent able to destroy naturally occurring microorganisms, including fungi, that is for decades applied as foliar sprays (Yang et al. , 2011; Husak, 2015). The lethal action of copper-based fungicides derives from their ability to free copper ions that are massively internalized by the fungal cells. Intracellularly, they bind various chemical groups (imidazoles, phosphates, sulfhydryls, hydroxyls) namely in proteins, causing their denaturation and loss of function (Husak, 2015; Mirković et al. , 2015). Ultimately, this leads to irreversible cell damage and membrane 13 leakage (Husak, 2015). Yet some fungi are resistant to copper ions. The mechanisms underlying this resistance are not well understood, although several studies have suggested that they might exert a combined action: the extracellular chelation and cell wall sequester of copper ions, and their decreased intake and intracellular complexing by metallothioneins or other proteins (Cervantes and GutierrezCorona, 1994). This last case includes the over-expression of the superoxide dismutase (SOD) that uses copper as inorganic co-factor (Naiki, 1980). SOD has been described to display the ability to buffer copper excess independently of its superoxide scavenging function (Culotta et al. , 1995). Additionally, the copperinduced accumulation of glycerol was also described to be involved in its extrusion (Gadd et al. , 1984). Neither of these mechanisms were ever described in association with Moniliophthora perniciosa , although increased levels of SOD would contribute the higher resistance to the above-mentioned plant-generated ROS. On the other hand, tebuconazole, as other azole-fungicides, acts on the synthesis of ergosterol, altering the structure and functionality of the fungal cell membrane (Price et al. , 2015) as well as vacuolar ion homeostasis through v-ATPase function (Zhang et al. , 2010b). In consequence of ergosterol synthesis disruption, mitochondrial function is also affected in its ability to form iron-sulfur clusters, which results in the deposit of insoluble iron inside mitochondria and concomitant radical formation and mitochondrial loss (Ward et al. , 2018). CytC harbors a heme group which availability for this enzyme proper assembly and function would be affected by iron homeostasis disruption, consequently affecting Complex IV function in respiration. This would justify why fungi that can respire through the AOXp could be resistant to azoles. Still, there is no reference to this possibility in the literature. Rather, in Candida species, the resistance to azole-fungicides implicates other types of mechanisms (Whaley et al. , 2017). Nevertheless, these are human commensals and pathogens, having therefore specificities that are not common with other yeasts or fungi. Fungicides fail to control the spread of the WBD but the mechanisms underlying the resistance of Moniliophthora perniciosa to these drugs are not studied. The use of fungicides is therefore not a routine practice in most cacao-producing countries also due to their high cost, and the risks associated with cacao chemical contamination which hinders commercialization. The reasons underlying this include the increasingly considered negative impact of fungicides on human health and the environment. Public concerns regarding the prevalence of agronomic pesticide residues in food, and their relation with the increasing advent of pesticide resistant pathogens, not only in plants but also in humans (Droby, 2006; Pal and McSpadden Gardener, 2006; Marelli et al. , 2009; Verweij et al. , 2009; Nunes, 2012) led to restrictions in Europe. Therefore, the most commonly used methods to control the WBD are exclusively 14 agronomic, through phytosanitary pruning, removing as much as possible the infected material, which is though often impossible, due to hidden fungal inoculum in the soil and cut branches and leaves 2 . Therefore, more effective and eco-friendly methods and strategies are needed to satisfy the consumer demands. In the case of OA management, it is mostly based on the use of chemical fungicides, which should be applied in a preventive manner, since the disease is nearly impossible to control once symptoms emerge (Sergeeva, 2011; Talhinhas et al. , 2011; Cacciola et al. , 2012; Moral et al. , 2012). Few fungicides have shown some efficacy in controlling the disease, including copper-based fungicides, dithiocarbamate, azoles (hexaconazole and tebuconazole) and strobilurins (azoxystrobin and trifloxystrobin) (Pennisi et al. , 1993; Sergeeva, 2011; Moral et al. , 2014). The effectiveness of the treatment with these drugs depends on several factors, such as the severity of the disease, the plant cultivar and environmental conditions, being therefore regionally variable. Furthermore, the timing, frequency and number of fungicide applications, as well as the preventive or curative nature of the treatment, are factors to take into account for the success of a particular chemical-based strategy (Cacciola et al. , 2012; Landum et al. , 2016). Regardless with the level of efficiency of chemical fungicides, other problems associated with their use also arise. The wash-off of the fungicides from olive crops due to rains leads to residue accumulation in soil and adjacent water areas, which could have negative and harmful effects on the surrounding ecosystem and even in human health (Komárek et al. , 2010; Lamichhane et al. , 2018). Additionally, since the active constituents of fungicides commercially exploited for disease management are chemically very similar and generally belong to the same molecular family, fungal phytopathogens easily adapt and acquire resistance (Ma and Michailides, 2005). Therefore, the overuse of these types of drugs eventually leads to their increasing inefficacy followed by loss of effect, leaving crops without possible treatment (Cacciola et al. , 2012). The general public awareness of food and environmental contamination caused by the misuse of chemical pesticides, as well as the risks associated to human health are pushing European policies to restrict the employment of fungicides in the control of phytopathogenic diseases. Consequently, OA management is nowadays increasingly based in agricultural strategies, which include the replacement or susceptible cultivars with resistant ones, early harvesting, strategic pruning and even orchard design (Talhinhas et al. , 2011; Cacciola et al. , 2012; Moral et al. , 2012; Leoni et al. , 2018). The efficiency of these strategies is though limited. The use of resistant cultivars should at term be the best strategy, although the complete replacement of the orchards takes a significant lag-time in production 2 Available at: https://www.icco.org/about-cocoa/pest-a-diseases.html [Accessed February 18, 2019] 15 and is altogether very expensive (Landum et al. , 2016). Additionally, in cases where ideal conditions for fungal growth are met, apparently the infection occurs anyway (Cacciola et al. , 2012). Therefore, new more effective and sustainable methods to control OA are essential. A concept that has gained considerable prominence in the agriculture sector in recent years is the use of nanoagroparticles, which are nanoparticles designed to mitigate agriculture-related problems, including plant pathologies (Parizi et al. , 2014; Parisi et al. , 2015). These nanoagroparticles include silver, copper, sulfur, zinc oxide and magnesium oxide nanoparticles (Baker et al. , 2017), and can act efficiently as fungicides, pesticides, herbicides and also insecticides. They can easily enter into the fungal cell wall. Once inside the cell, they act through different modes, which include (i) causing the disruption of metabolism, or of the cell membrane, with consequent loss of cellular content (Baker et al. , 2015; Baker et al. , 2017), (ii) promoting the release of toxic ions (Cd2+, Zn2+ and Ag+) that bind to sulfur-containing proteins, (iii) targeting the pathogen DNA, this way inducing cell death, (iv) interrupting electron transport, this way causing the collapse of membrane potential, (v) promoting the generation of ROS, or (vi) interfering with nutrient uptake (Alghuthaymi et al. , 2015). More than one of these mechanisms can occur simultaneously, conferring the ability of nanoparticles to be effective against different plant pathogens (Alghuthaymi et al. , 2015). Recently, nanoparticles were conjugated with some biomolecules (including biocide/killer toxins), forming bionano-hybrid agroparticles (Baker et al. , 2017). The free utilization of these promising phytopathology management tools still requires not only cytotoxicology studies to evaluate potential harm to human and animal health, but even more important, extensive ecotoxicology and biodegradability studies to evaluate their prevalence in the environment and food chains and their effect on the long run in the microflora of plant, soil and water. Presently few information on this regard is available (Alghuthaymi et al. , 2015). Nonetheless, the prospective of being able to use such a nanotool to effectively deliver a toxin and kill a phytopathogenic fungus is attractive, especially if carrying a bioderived killing agent. 16 YEASTS AS BIOCONTROL AGENTS One possible approach to fungal diseases in plants might be the use of biocides or biological control agents. In phytopathology, this term designates the use of introduced or resident living organisms to contain or suppress populations of pathogens (Pal and McSpadden Gardener, 2006). There are a few of these agents in the market, mostly used in the control of pests at small scale. They correspond to dry biomass of bacterial or filamentous fungal strains isolated from the endosphere or the rhizosphere of plants (O’Brien, 2017) that are re-hydrated and used as alive reproductive microorganisms. These include a taxonomically and biologically diverse group of endophytic fungi that are characterized by colonizing internally the plant host tissues without causing any external disease symptoms (Wilson, 1995; Rubini et al. , 2005). These endophytes can prevent pathogen infection and propagation directly by competition, mycoparasitism or antibiosis, or indirectly by inducing resistance responses in the plant (Bailey et al. , 2006). Despite this, the biocontrol agents that could be applied in phytopathology are not restricted to these two groups of organisms. Endophytic microorganisms also include Ascomycota and Basidiomycota yeasts, found in many species of trees from very diverse climates, but also in agricultural species (reviewed by Doty, 2013). Ascomycota yeasts reproduce exclusively by budding, as the most well-known yeast Saccharomyces cerevisiae . Basidiomycota grow dimorphically, shifting from a monokaryotic yeast-form to a dikaryotic filamentous form (Choudhary and Johri, 2009). This is the case of Rhodotorula and Cryptococcus sp. (Table 1). Generally, endophytic yeasts apparently thrive symbiotically or mutualistically, virtually colonizing diverse plant tissues (reviewed by Doty, 2013), in which they may cause structural changes (Luna, 2017). They consume sugars and assimilate amino acids generated by the plant and contribute to the plant wellbeing and stress response in many different ways, including the production of phytopheromones, catalase or siderophores (reviewed by Joubert and Doty, 2018). Importantly, endophytic like epiphytic yeasts can antagonize phytopathogenic filamentous fungi, either by occupying their niche or by antagonizing them in more complex ways. The antagonistic interaction of yeasts with particular phytopathogenic fungi has been described in the literature. For example, Suzzi et al. (1995) observed that natural wine yeast strains of Saccharomyces and Zygosaccharomyces inhibited in vitro the growth of 10 species of soil-borne fungal plant pathogens, namely Cladosporium variabile , Rhizoctonia fragariae , Phomopsis longicolla , Colletotrichum acutatum , Aspergillus niger , Sclerotinia sclerotiorum , Penicillium digitatum , Macrophomina phaseolina , Trichoderma viride and Botrytis squamosa . Also, Walker et al. (1995) reported that strains of 17 Saccharomyces cerevisiae and Pichia anomala (know designated Wickerhamomyces anomalus ) inhibited in vitro the growth of several wood decay basidiomycetes including Serpula lacrymans , Postia placenta , Lentinus lepideus and Ophiostoma ulmi and phytopathogenic fungi, such as Rhizoctonia solani , Fusarium equiseti , Botrytis fabae and Phytophthora infestans . Importantly, Rosa-Magri et al. (2011) described the antagonism effect of the yeast Torulaspora globosa against the phytopathogenic mold Colletotrichum graminicola , the causal agent of anthracnose disease in maize. All of these cases were reported as in vitro studies, none were performed in planta or in field . Otherwise, yeasts have often been proposed and used for the control of microbial contaminations at the postharvest phase (Table 1). The possibility of using yeasts as biocontrol agents of fungal or bacterial proliferation associated with food spoilage has been recognized since the early 1960s, when it was found that Saccharomyces cerevisiae strains secreted toxins that killed other yeast strains but are immune to their own toxin (Bevan and Makower, 1963). Killer toxins (KTs) can be encoded by cytoplasm-inherited double-stranded RNA viruses (Schmitt and Breinig, 2002) or linear dsDNA plasmids (Schaffrath and Meinhardt, 2005), but they can also be chromosomally encoded (Suzuki, 2005). The killer phenomenon is well characterized and studied in Saccharomyces cerevisiae. In this yeast species, KTs have been grouped into four types, K1, K2, K28, and Klus, based on their killing profiles and lack of cross-immunity (Schmitt and Breinig, 2006; RodríguezCousiño et al. , 2011). Each strain producing one specific toxin kills strains from the other groups but has self-protective immunity (Schmitt and Breinig, 2006). Killer toxins The modes of action of the Saccharomyces cerevisiae KTs are well known, with the exception of the recently found Klus toxin (Schmitt and Breinig, 2002). K1 and K2 toxins kill sensitive yeast cells in a receptor-mediated two-step process. The first step involves a fast, energy-independent binding to a primary toxin receptor (R1), consisting of β-1,6-D-glucan (Lukša et al. , 2015). Though a second energydependent step, the toxin is translocated from the cell wall to the plasma membrane, where it interacts with a secondary membrane receptor (R2), identified in the case of K1 toxin as Kre1p, an O -glycosylated protein of the yeast cell surface (Breinig et al. , 2002; 2004). After reaching the plasma membrane, K1 and K2 toxins disrupt its function by forming cation-selective channels, and promoting the release of ATP and other metabolites, thus causing a lethal effect on the target cell (Liu et al. , 2015). 18 Table 1. Success cases of using yeasts to antagonize the spoilage of fruits by filamentous fungi. Yeast antagonist Host Fungal phytopathogen(s) References Preharvest application1 Candida (Pichia) guilliermondii Cherry tomato Fruit decay agents Zhao et al. , 2011 Candida sake Apple Penicillium expansum Teixidó et al. , 1999 Postharvest application Aureobasidium pullulans Pear Penicillium expansum Robiglio et al. , 2011 Apple Botrytis cinerea , Colletotrichum acutatum and Penicillium expansum Mari et al. , 2012 Candida (Pichia) guilliermondii Chilli Colletotrichum capsici Chanchaichaovivat et al. , 2007 Tomato Rhizopus nigricans Zhao et al. , 2008 Rhizopus stolonifer Celis et al. , 2014 Kiwifruit Botrytis cinerea Sui and Liu, 2014 Papaya Colletotrichum gloeosporioides Lima et al. , 2013 Candida oleophila Banana Colletotrichum musae , Fusarium moniliforme and Cephalosporium sp. Lassois et al. , 2008 Apple Penicillium expansum and Botrytis cinerea Liu et al. , 2012 Candida pelliculosa Tomato Botrytis cinerea Dal Bello et al. , 2008 Candida sake Apple Penicillium expansum Morales et al. , 2008 Cryptococcus infirmo-miniatus Sweet cherry Monilinia fructicola Spotts et al. , 2002 Cryptococcus laurentii Strawberry Botrytis cinerea Wei et al. , 2014 Sweet cherry Fruit decay agents Tian et al. , 2004 Debaryomyces hansenii Peach Rhizopus stolonifer Mandal et al. , 2007 Mandarin, orange Penicillium digitatum Taqarort et al. , 2008 Metschnikowia fructicola Apple Penicillium expansum Liu et al. , 2011 Grapefruit Penicillium digitatum Hershkovitz et al. , 2013 Meyerozyma caribbica Mango Colletotrichum gloeosporioides Bautista-Rosales et al. , 2013 Pichia membranefaciens Apple Monilinia fructicola, Penicillium expansum and Rhizopus stolonifer Chan and Tian, 2005 Rhodosporidium paludigenum Cherry tomato Botrytis cinerea Wang et al. , 2010 Rhodotorula mucilaginosa Pear Penicillium expansum Hu et al. , 2015 Rhodotorula rubra Tomato Botrytis cinerea Dal Bello et al. , 2008 Wickerhamomyces (Pichia) anomalus Banana Colletotrichum musae , Fusarium moniliforme and Cephalosporium sp. Lassois et al. , 2008 Orange Penicillium digitatum Aloui et al. 2015; Platania et al. , 2012 Papaya Colletotrichum gloeosporioides Lima et al. , 2013 Commercial yeast-biocontrol products2 Aureobasidium pullulans Pome Penicillium, Botrytis, Monilinia Boni Protect®, Bio-Ferm, AT Candida oleophila Pome Penicillium, Botrytis Nexy®, Lesaffre, BE Metschnikowia fructicola Pome, table grape, stone fruits, strawberry, sweet potato Penicillium, Botrytis, Rhizopus, Aspergillus Shemer®, Bayer/Koppert, NL 1 Preharvest application to prevent postharvest spoilage. 2 Wisniewski et al. , 2016. 19 The K28 KT mode of action is very different, since it enters a sensitive target yeast cell by endocytosis, in a cell wall receptor-mediated manner (Schmitt and Breinig, 2006). The cell wall receptor for K28 toxin has been identified as a mannoprotein with high molecular mass (Liu et al. , 2015). The K28 toxin is internalized through the secretory pathway (via Golgi and ER), and after entering the cytosol the β-subunit is ubiquitinated and degraded in the proteasome. The subsequently free small α-subunit has been suggested to enter the nucleus without the help of an active nuclear import machinery, therefore by a socalled passive diffusion (Schmitt and Breinig, 2006). Once inside the nucleus, the K28 toxin kills the host cell by irreversibly blocking the DNA synthesis. The target cells arrest in early S phase of the cell cycle, forming a medium-sized bud and a single, pre-replicated nucleus in the mother cell, eventually dying (Schmitt and Breinig, 2006). The killer phenomenon, despite being best characterized in Saccharomyces cerevisiae , is not confined to this yeast species, rather it is often found in other yeast species and genera (Magliani et al., 1997; Schmitt and Breinig, 2002). Some of these were described to damage the plasma membrane, very similarly to the Saccharomyces cerevisiae K1 toxin. This is the case of the KTs produced by Pichia kluyveri (Ahmed et al., 1999), Pichia membranifaciens (Santos and Marquina, 2004; Santos et al., 2009), Pichia farinosa (Suzuki et al., 2001) and Zygosaccharomyces bailii (Weiler and Schmitt, 2003). Other killer mechanisms include the damage of the cell wall upon the inhibition of the synthesis of β-glucans. Examples of this mode of action are the toxins produced by Hansenula mrakii (previously Williopsis mrakii ) (Marquina et al., 2002), Wickerhamomyces anomalus (formerly designated Pichia anomala or Hansenula anomala ) (Wang et al., 2007), Williopsis saturnus (Guyard et al., 2002; Peng et al., 2010) and Kluyveromyces phaffii (Comitini et al., 2009). Yet other yeast KTs act by blocking the cell cycle, namely the one produced by Kluyveromyces lactis (Klassen et al., 2004), and by triggering DNA damaging and the induction of apoptosis, which is the case of the toxins secreted by Pichia acaciae (Klassen and Meinhardt, 2005) and Wingea robertsiae (Klassen and Meinhardt, 2002). The blocking of calcium uptake was also described as killer mode of action, namely in the case of Ustilago maydis (Gage et al., 2001). Although there are plenty of reports in the literature regarding the interaction of nonSaccharomyces killer yeasts with a vast variety of sensitive targets, the actual mechanisms involved remain mostly unknown or superficially studied at the molecular level. 20 Lytic enzymes and mycoparasitism Several other mechanisms of yeast antagonism have been proposed that do not involve the secretion of a peptide/protein that may be classified as a KT. Other proteins that are secreted by the yeast and antagonize filamentous fungi are lytic enzymes that destroy the fungal cell wall (Spadaro and Gullino, 2004). This kind of antagonism is considered a form of mycoparasitism. An example is the manner in which Pichia guilliermondii antagonizes Botrytis cinerea (Wisniewski et al. , 1991). The authors observed that the fungus cell wall glucans and a yeast-secreted β-(1–3) glucanase form a lectin-like interaction resulting in a strong attachment of the antagonist to the fungal pathogen which culminates with the lysis of fungal cells. Besides the secretion of these antifungal compounds, other modes of action can be involved in the mycoparasitism behavior. Yeasts can attack the fungi pathogens by direct physical contact, reaching the fungal cells and killing them, although not necessarily through an invasion of the target cell (Mims et al. , 2007). Additionally, yeasts can act as predatory mycoparasites, physically penetrating their prey cell walls, through haustoria or penetration pegs (Junker et al. , 2019). Volatile compounds Yeasts that antagonize other yeasts can also produce several volatile compounds against filamentous fungi that inhibit the target growth (Mari et al. , 2016). These compounds include alkenes, alcohols, ketones, benzenoids, pyrazines, sulfides and terpenes (Schulz-Bohm et al. , 2017). These have in common the fact of having small molecular weight and physicochemical properties which facilitate evaporation and diffusion in soil and rhizosphere environments (Mari et al. , 2016; Schulz-Bohm et al. , 2017). This is the case of the yeast Aureobasidium pullulans that produces 2-methyl-1-butanol, 3-methyl1-butanol, 2-phenethyl alcohol and 2-methyl-1-propanol, with inhibitory effect against Botrytis cinerea , Colletotrichum acutatum , Penicillium expansum , Penicillium digitatum and Penicillium italicum (Di Francesco et al., 2015) . Interestingly, it appears that antagonizing yeasts may operate in different ways also because in some cases they strongly attach to the fungus hyphae. This was described in detail for the case of the yeasts Pichia membranefaciens and Cryptococcus albidus when challenged with three phytopathogenic fungi causing the postharvest deterioration of nectarines and apples ( Monilinia fructicola, Penicillium expansum and Rhizopus stolonifer ) (Chan and Tian, 2005). 21 Competition for nutrients and space Another effective mechanism of antagonism and possibly the most common is competition. Microbes compete for space, for oxygen and of course for nutrients, such as carbohydrates, vitamins, minerals and amino acids (Spadaro and Droby, 2016). Yeasts grow much faster than filamentous fungi, being thus able to quickly colonize the niches that fungi can occupy, such as plant wounds or tissue lesions, forming colonies or biofilms (Andrews et al. , 1994). Increasingly bigger yeast populations reduce the amount of nutrients available for fungi and make them difficult to access (Zhang et al. , 2010a). In the case of micronutrients, iron plays a crucial role in the growth, development and virulence of the fungal pathogens (Saravanakumar et al. , 2008). To compete with the pathogens for iron, yeasts secrete siderophores that deplete iron from the growth medium such as pulcherrimin, produced by Metschnikowia pulcherrima to compete with Botrytis cinerea , Alternaria alternata and Penicillium expansum (Saravanakumar et al. , 2008). Quorum sensing The production and secretion of quorum sensing molecules can be also involved in yeast antagonism. The concept of quorum sensing was introduced in 1994 (Fuqua et al. ), and consists in the cell-to-cell communication process through which each individual cells can adjust its phenotype in response to the presence of extracellular quorum sensing molecules (QSM) at a specific concentration (Wuster and Babu, 2007). When QSM reach those concentrations, they bind with a receptor, link with promotor sequences and activate the transcriptional regulators of specific genes (Mehmood et al. , 2019). Examples of these kind of molecules described in yeasts are 2-phenylethanol, tryptophol and tyrosol (Avbelj et al. , 2016). Fungi can also produce a particular type of secondary metabolites which can act as anti-microbial agents known as quorum sensing inhibitors (QSI). These microorganisms colonize several habitats interact with other organisms, such as other microbes, animals and plants. Due to the need to compete for nutrients and space, they deal with competitor organisms by producing secondary metabolites, enzymes and chemicals (Padder et al. , 2018). These compounds can act as QSI by degrading QSM, delaying their production or even blocking the receptors via homologs of QSM (Padder et al. , 2018; Mehmood et al. , 2019). An example of QSI is farnesol, a secondary metabolite secreted by many dimorphic yeasts (Wongsuk et al. , 2016; Mehmood et al. , 2019). 28 Chapter 3 presents a microscopy analysis of the interactions between the antagonistic yeasts and the phytopathogenic fungi in the antagonism assays performed in Chapter 2. Scanning Electron Microscopy technique was used. The yeast attachment to the pathogen hyphae, the occurrence of constriction of the fungal cells and an apparent draining of the fungal cellular content was observed. Moreover, yeast fusion with the hyphae and fimbriae-like connections between the yeast and the fungal cells and also between antagonistic yeasts are described. The possibility of a predacious-like behavior exerted by the antagonistic yeasts and its role in the antagonism exerted against the WBD fungal pathogens is discussed. Chapter 4 comprises the assessment of the capability of the group of yeast strains originating from the industrial fermentations of cachaça and from the industrial production of bioethanol used on Chapter 2 to antagonize Colletotrichum gloeosporioides and C. acutatum , the fungal causative agents of Olive Anthracnose. The ability of a group of yeasts from the olive biome in antagonizing the olive anthracnose fungal pathogens is also explored. A Wickerhamomyces anomalus , isolated from olive orchards in Portugal, is the most promising strain, being able to inhibit the growth of all Colletotrichum sp. tested. The validation of the application of the protocol developed and optimized in Chapter 2 for the evaluation of yeast antagonism is also presented. Chapter 5 explores the possibility of using vinasse , the main waste product from sugarcane bioethanol production process, to control the development of the Witches’ Broom Disease causal agents. The ability of vinasse per se to inhibit the growth and development and even kill M. perniciosa is reported. The potential of a solution based on the fertirrigation of cacao plantations with vinasse , which could contribute to contain the prevalence and spread of the disease is presented. Chapter 6 is a conclusion chapter which summarizes all the research performed in the scope of this thesis. The results and findings of the different experimental studies are here consolidated. The future perspectives in this area related with the research work performed is also proposed and discussed. Supplementary Material An exploratory assessment and identification of the protein content of the yeast vs fungus co-culture supernatants from the antagonism assays of Chapter 2 that could be involved in the antagonism effect against M. perniciosa is provided. 29 REFERENCES Ahmed, A., Sesti, F., Ilan, N., Shih, T. M., Sturley, S. L., and Goldstein, S. A. (1999). A molecular target for viral killer toxin: TOK1 potassium channels. Cell 99(3), 283-291. doi: 10.1016/S00928674(00)81659-1. Aime, M. C., and Phillips-Mora, W. (2005). The causal agents of witches’ broom and frosty pod rot of cacao (chocolate, Theobroma cacao ) form a new lineage of Marasmiaceae. Mycologia 97(5), 1012-1022. doi: 10.1080/15572536.2006.11832751. Alghuthaymi, M. A., Almoammar, H., Rai, M., Said-Galiev, E., and Abd-Elsalam, K. A. (2015). Myconanoparticles: synthesis and their role in phytopathogens management. Biotechnology & Biotechnological Equipment 29, 221-236. doi: 10.1080/13102818.2015.1008194. Alkan, N., Espeso, E. A., and Prusky, D. (2013). Virulence Regulation of Phytopathogenic Fungi by pH. Antioxidants & Redox Signaling 19(9), 1012-1025. doi: 10.1089/ars.2012.5062. Almeida, O. C., Chiacchio, F., and Rocha, H. M. (1997). Survival of Crinipellis perniciosa (Stahel) Singer on dried brooms of cacao ( Theobroma cacao L.) in the State of Bahia (in Portuguese). Agrotrópica (Brazil) 9(1), 23-28. Aloui, H., Licciardello, F., Khwaldia, K., Hamdi, M., and Restuccia, C. (2015). Physical properties and antifungal activity of bioactive films containing Wickerhamomyces anomalus killer yeast and their application for preservation of oranges and control of postharvest greenmold caused by Penicillium digitatum . International Journal of Food Microbiology 200, 22-30. doi: 10.1016/j.ijfoodmicro.2015.01.015. Andrade, M. J., Thorsen, L., Rodríguez, A., Córdoba, J. J., and Jespersen, L. (2014). Inhibition of ochratoxigenic moulds by Debaryomyces hansenii strains for biopreservation of dry-cured meat products. International Journal of Food Microbiology 170, 70-77. doi: 10.1016/j.ijfoodmicro.2013.11.004. Andrews, J. H., Harris, R. F., Spear, R. N., Lau, G. W., and Nordheim, E. V. (1994). Morphogenesis and adhesion of Aureobasidium pullulans . Canadian Journal of Microbiology 40, 6-17. doi: 10.1139/m94-002. 30 Arnold, A. E., Mejía, L. C., Kyllo, D., Rojas, E. I., Maynard, Z., Robbins, N. , et al. (2003). Fungal endophytes limit pathogen damage in a tropical tree. Proc. Natl. Acad. Sci. USA 100, 1564915654. doi: 10.1073pnas.2533483100. Avbelj, M., Zupan, J., and Raspor, P. (2016). Quorum-sensing in yeast and its potential in wine making. Applied Microbiology and Biotechnology 100, 7841-7852. doi: 10.1007/s00253-016-7758-3. Bailey, B. A., Bae, H., D., S. M., Roberts, D. P., Thomas, S. E., Crozier, J. , et al. (2006). Fungal and plant gene expression during the colonization of cacao seedlings by endophytic isolates of four Trichoderma species. Planta 224, 1449-1464. doi: 10.1007/s00425-006-0314-0. Baker, S., Kumar, M., Santosh, P., Rakshith, D., and Satish, S. (2015). Extracellular synthesis of silver nanoparticles by novel Pseudomonas veronii AS41G inhabiting Annona squamosa L. and their bactericidal activity. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 136, 1434-1440. doi: 10.1016/j.saa.2014.10.033. Baker, S., Volova, T., Prudnikova, S. V., Satish, S., and Prasad, M. N. N. (2017). Nanoagroparticles emerging trends and future prospect in modern agriculture system. Environmental Toxicology and Pharmacology 53, 10-17. doi: 10.1016/j.etap.2017.04.012. Ballet, N., Souche, J. L., and Vandekerckove, P. (2016). Efficacy of Candida oleophila , strain O, in preventing postharvest diseases of fruits. Acta Horticulturae 1144, 105-112. doi: 10.17660/ActaHortic.2016.1144.15. Baroncelli, R., Talhinhas, P., Pensec, F., Sukno, S. A., Le Floch, G., and Thon, M. R. (2017). The Colletotrichum acutatum Species Complex as a Model System to Study Evolution and Host Specialization in Plant Pathogens. Frontiers in Microbiology 8, 2001-2001. doi: 10.3389/fmicb.2017.02001. Bautista-Rosales, P. U., Calderon-Santoyo, M., Servín-Villegas, R., Ochoa-Álvarez, N. A., and RagazzoSánchez, J. A. (2013). Action mechanisms of the yeast Meyerozyma caribbica for the control of the phytopathogen Colletotrichum gloeosporioides in mangoes. Biological Control 65, 293-301. doi: 10.1016/j.biocontrol.2013.03.010. Bevan, E. A., and Makower, M. (1963). "The physiological basis of the killer-character in yeast," in Genetics Today, XIth International Congress of Genetics Vol. 1, ed. S. J. Geerts. (Oxford: Pergamon Press), 202-203. 31 Bonaterra, A., Badosa, E., Cabrefiga, J., Francés, J., and Montesinos, E. (2012). Prospects and limitations of microbial pesticides for control of bacterial and fungal pomefruit tree diseases. Trees 26, 215226. doi: 10.1007/s00468-011-0626-y. Bowers, J. H., Bailey, B. A., Hebbar, P. K., Sanogo, S., and Lumsden, R. D. (2001). The impact of plant diseases on world chocolate production. Plant Health Progress Online. doi: 10.1094/PHP-20010709-01-RV. Breinig, F., Schleinkofer, K., and Schmitt, M. J. (2004). Yeast Kre1p is GPI-anchored and involved in both cell wall assembly and architecture. Microbiology 150, 3209-3218. doi: 10.1099/mic.0.271750. Breinig, F., Tipper, D. J., and Schmitt, M. J. (2002). Kre1p, the plasma membrane receptor for the yeast K1 viral toxin. Cell 108, 395-405. doi: 10.1016/S0092-8674(02)00634-7. Cabral, A. S., de Carvalho, P. M. B., Pinotti, T., Hagler, A. N., Mendonça-Hagler, L. C. S., and Macrae, A. (2009). Killer yeasts inhibit the growth of the phytopathogen Moniliophthora perniciosa , the causal agent of Witches’ Broom disease. Brazilian Journal of Microbiology 40, 108-110. doi: 10.1590/S1517-83822009000100018 Cacciola, S. O., Faedda, R., Sinatra, F., Agosteo, G. E., Schena, L., Frisullo, S. , et al. (2012). Olive Anthracnose. Journal of Plant Pathology 94(1), 29-44. Campos-Martínez, A., Velázquez-del Valle, M. G., Flores-Moctezuma, H. E., Suárez-Rodríguez, R., RamírezTrujillo, J. A., and Hernández-Lauzardo, A. N. (2016). Antagonistic yeasts with potential to control Colletotrichum gloeosporioides (Penz.) Penz. & Sacc. and Colletotrichum acutatum J.H. Simmonds on avocado fruits. Crop Protection 89, 101-104. doi: 10.1016/j.cropro.2016.07.001. Cárdenas-Monroy, C. A., Pohlmann, T., Piñón-Zárate, G., Matus-Ortega, G., Guerra, G., Feldbrügge, M. , et al. (2017). The mitochondrial alternative oxidase Aox1 is needed to cope with respiratory stress but dispensable for pathogenic development in Ustilago maydis . PLoS One 12, e0173389. doi: 10.1371/journal.pone.0173389. Carvalho, F. P. (2006). Agriculture, pesticides, food security and food safety. Environmental Science & Policy 9, 685-692. doi: 10.1016/j.envsci.2006.08.002. Carvalho, M. T., Simões-Lopes, P., and Monteiro da Silva, M. J. (Year). "Influence of Different Olive Infection Rates of Colletotrichum acutatum on Some Important Olive Oil Chemical Parameters": International Society for Horticultural Science (ISHS), Leuven, Belgium), 555-558. 32 Castoria, R., Caputo, L., De Curtis, F., and De Cicco, V. (2003). Resistance of Postharvest Biocontrol Yeasts to Oxidative Stress: A Possible New Mechanism of Action. Phytopathology 93, 564-572. doi: 10.1094/PHYTO.2003.93.5.564. Celis, Z. C., Moreno, D. G., Sequeda-castañeda, L. G., García, A., Albarracín, D. M., Charry, B. , et al. (2014). Determining the effectiveness of Candida guilliermondii in the biological control of Rhizopus stolonifer in postharvest tomatoes. Universitas Scientiarum 19, 51-62. Cervantes, C., and Gutierrez-Corona, F. (1994). Copper resistance mechanisms in bacteria and fungi. FEMS Microbiology Reviews 14(2), 121-137. doi: 10.1111/j.1574-6976.1994.tb00083.x. Chan, Z., and Tian, S. P. (2005). Interaction of antagonistic yeasts against postharvest pathogens of apple fruit and possible mode of action. Postharvest Biology and Technology 36, 215-223. doi: 10.1016/j.postharvbio.2005.01.001. Chanchaichaovivat, A., Ruenwongsa, P., and Panijpan, B. (2007). Screening and identification of yeast strains from fruits and vegetables: Potential for biological control of postharvest chilli anthracnose ( Colletotrichum capsici ). Biological Control 42, 326-335. doi: 10.1016/j.biocontrol.2007.05.016. Chaudhuri, M., Ott, R. D., and Hill, G. C. (2006). Trypanosome alternative oxidase: from molecule to function. Trends in Parasitology 22, 484-491. doi: 10.1016/j.pt.2006.08.007. Choudhary, D. K., and Johri, B. N. (2009). "Basidiomycetous Yeasts: Current Status," in Yeast Biotechnology: Diversity and Applications, eds. T. Satyanarayana & G. Kunze. Springer), 19-46. Comitini, F., Mannazzu, I., and Ciani, M. (2009). Tetrapisispora phaffii killer toxin is a highly specific βglucanase that disrupts the integrity of the yeast cell wall. Microbial Cell Factories 8(55), 8. doi: 10.1186/1475-2859-8-55. Culotta, V. C., Joh, H. D., Lin, S. J., Slekar, K. H., and Strain, J. (1995). A Physiological Role for Saccharomyces cerevisiae Copper/Zinc Superoxide Dismutase in Copper Buffering. The Journal of Biological Chemistry 270(50), 29991–29997. doi: 10.1074/jbc.270.50.29991. Dal Bello, G., Mónaco, C., Rollan, M. C., Lampugnani, G., Arteta, N., Abramoff, C. , et al. (2008). Biocontrol of Postharvest Grey Mould on Tomato by Yeasts. Journal of Phytopathology 156, 257-263. doi: 10.1111/j.1439-0434.2007.01351.x. 33 Darsini, N. N., Sudana, I. M., Suprapta, I. D. N., and Nyana, D. N. (2017). Exploring Antagonistic Candidate Fungi for Controling Pathogenic Fungi ( Colletotrichum gloeosporioides ) Causing Anthracnose Disease in Kintamani Siam Orange Plants ( Citrus nobillis Lour Var. Hass). International Journal on Advanced Science, Engineering and Information Technology 7, 269-275. doi: 10.18517/ijaseit.7.1.1919. de Lima, J. R., Gonçalves, L. R., Brandão, L. R., Rosa, C. A., and Viana, F. M. (2013). Isolation, identification, and activity in vitro of killer yeasts against Colletotrichum gloeosporioides isolated from tropical fruits. Journal of Basic Microbiology 53(7), 590-599. de Oliveira, B. V., Teixeira, G. S., Reis, O., Barau, J. G., Teixeira, P. J., do Rio, M. C. , et al. (2012). A potential role for an extracellular methanol oxidase secreted by Moniliophthora perniciosa in Witches' broom disease in cacao. Fungal Genetics and Biology 49, 922-932. doi: 10.1016/j.fgb.2012.09.001. De Silva, D. D., Crous, P. W., Ades, P. K., Hyde, K. D., and Taylor, P. W. J. (2017). Life styles of Colletotrichum species and implications for plant biosecurity. Fungal Biology Reviews 31(3), 155168. doi: 10.1016/j.fbr.2017.05.001. Dean, R., Van Kan, J. A. L., Pretorius, Z. A., Hammond-Kosack, K. E., Di Pietro, A., Spanu, P. D. , et al. (2012). The Top 10 fungal pathogens in molecular plant pathology. Molecular Plant Pathology 13(4), 414-430. doi: 10.1111/j.1364-3703.2011.00783.x. Del Rio, D., Stewart, A. J., and Pellegrini, N. (2005). A review of recent studies on malondialdehyde as toxic molecule and biological marker of oxidative stress. Nutrition, Metabolism and Cardiovascular Diseases 15, 316-328. doi: 10.1016/j.numecd.2005.05.003. Di Francesco, A., Ugolini, L., Lazzeri, L., and Mari, M. (2015). Production of volatile organic compounds by Aureobasidium pullulans as a potential mechanism of action against postharvest fruit pathogens. Biological Control 81, 8-14. doi: 10.1016/j.biocontrol.2014.10.004. Doty, S. L. (2013). "Endophytic Yeasts: Biology and Applications," in Symbiotic Endophytes, Soil Biology, ed. R. Aroca. Springer, Berlin, Heidelberg), 335-343. Droby, S. (2006). Improving quality and safety of fresh fruit and vegetables after harvest by the use of biocontrol agents and natural materials. Acta Horticulturae 709, 45-51. doi: 10.17660/ActaHortic.2006.709.5. 34 Droby, S., Wisniewski, M., Macarisin, D., and Wilson, C. (2009). Twenty years of postharvest biocontrol research: Is it time for a new paradigm? Postharvest Biology and Technology 52. doi: 10.1016/j.postharvbio.2008.11.009. Duarte, H. S. S., Cabral, P. G. C., Pereira, O. L., Zambolim, L., Gonçalves, E. D., Vieira Neto, J. , et al. (2010). First report of anthracnose and fruit mummification of olive fruit ( Olea europaea ) caused by Colletotrichum acutatum in Brazil. Plant Pathology 59, 1170. doi: 10.1111/j.13653059.2010.02297.x. Dukare, A. S., Paul, S., Nambi, V. E., Gupta, R. K., Singh, R., Sharma, K. , et al. (2018). Exploitation of microbial antagonists for the control of postharvest diseases of fruits: a review. Critical Reviews in Food Science and Nutrition 16, 1-16. doi: 10.1080/10408398.2017.1417235. Elthon, T. E., and McIntosh, L. (1987). Identification of the alternative terminal oxidase of higher plant mitochondria. Proceedings of the National Academy of Sciences, USA 84, 8399-8403. doi: 10.1073/pnas.84.23.8399. Evans, H. C. (1980). Pleomorphism in Crinipellis perniciosa , causal agent of witches’ broom disease of cocoa. Transactions of the British Mycological Society 74, 515-523. doi: 10.1016/S00071536(80)80051-9. Fisher, M. C., Hawkins, N. J., Sanglard, D., and Gurr, S. J. (2018). Worldwide emergence of resistance to antifungal drugs challenges human health and food security. Science 360, 739-742. doi: 10.1126/science.aap7999. Fisher, M. C., Henk, D. A., Briggs, C. J., Brownstein, J. S., Madoff, L. C., McCraw, S. L. , et al. (2012). Emerging fungal threats to animal, plant and ecosystem health. Nature 484, 186-194. doi: 10.1038/nature10947. Fuqua, W. C., Winans, S. C., and Greenberg, E. P. (1994). Quorum sensing in bacteria: the LuxR-LuxI family of cell density-responsive transcriptional regulators. Journal of bacteriology 176(2), 269275. doi: 10.1128/jb.176.2.269-275.1994. Gadd, G. M., Chudek, J. A., Foster, R., and Reed, R. H. (1984). The Osmotic Responses of Penicillium ochro-chloron : Changes in Internal Solute Levels in Response to Copper and Salt Stres. Journal of General Microbiology 130, 1969-1975. doi: 10.1099/00221287-130-8-1969. 35 Gage, M. J., Bruenn, J., Fischer, M., Sanders, D., and Smith, T. J. (2001). KP4 fungal toxin inhibits growth in Ustilago maydis by blocking calcium uptake. Molecular Microbiology 41, 775-785. doi: 10.1046/j.1365-2958.2001.02554.x. Gouvinhas, I., Machado, N., Sobreira, C., Dominguez-Perles, R., Gomes, S., Rosa, E. , et al. (2017). Critical Review on the Significance of Olive Phytochemicals in Plant Physiology and Human Health. Molecules 22(11). doi: 10.3390/molecules22111986. Griffith, G. W., Nicholson, J., Nenninger, A., Birch, R. N., and Hedger, J. N. (2003). Witches’ brooms and frosty pods: two major pathogens of cacao. New Zealand Journal of Botany 41, 423-435. doi: 10.1080/0028825X.2003.9512860. Guyard, C., Dehecq, E., Tissier, J. P., Polonelli, L., Dei-Cas, E., Cailliez, J. C. , et al. (2002). Involvement of β-Glucans in the Wide-Spectrum Antimicrobial Activity of Williopsis saturnus var. mrakii MUCL 41968 Killer Toxin. Molecular Medicine 8, 686-694. doi: 10.1007/BF03402032. Hatoum, R., Labrie, S., and Fliss, I. (2012). Antimicrobial and probiotic properties of yeasts: From fundamental to novel applications. Frontiers in Microbiology 3, 421. doi: 10.3389/fmicb.2012.00421. Hershkovitz, V., Sela, N., Taha-Salaime, L., Liu, J., Rafael, G., Kessler, C. , et al. (2013). De-novo assembly and characterization of the transcriptome of Metschnikowia fructicola reveals differences in gene expression following interaction with Penicillium digitatum and grapefruit peel. BMC Genomics 14, 168. doi: 10.1186/1471-2164-14-168. Hu, H., Yan, F., Wilson, C., Shen, Q., and Zheng, X. (2015). The ability of a cold-adapted Rhodotorula mucilaginosa strain from Tibet to control blue mold in pear fruit. Antonie van Leeuwenhoek 108, 1391-1404. doi: 10.1007/s10482-015-0593-1. Husak, V. (2015). Copper and Copper-Containing Pesticides: Metabolism, Toxicity and Oxidative Stress. Journal of Vasyl Stefanyk Precarpathian National University 2(1), 38-50. doi: 10.15330/jpnu.2.1.38-50. Joseph-Horne, T., Hollomon, D. W., and Wood, P. M. (2001). Fungal respiration: a fusion of standard and alternative components. Biochimica et Biophysica Acta 1504, 179-195. doi: 10.1016/S00052728(00)00251-6. 36 Joubert, P. M., and Doty, S. L. (2018). "Endophytic Yeasts: Biology, Ecology and Applications," in Endophytes of Forest Trees. Forestry Sciences, eds. A. Pirttilä & A. Frank. Springer, Cham), 314. Kalogiannis, S., Tjamos, S. E., Stergiou, A., Antoniou, P. P., Ziogas, B. N., and Tjamos, E. C. (2006). Selection and evaluation of phyllosphere yeasts as biocontrol agents against grey mould of tomato. European Journal of Plant Pathology 116, 69-76. doi: 10.1007/s10658-006-9040-5. Klassen, R., and Meinhardt, F. (2002). Linear plasmids pWR1A and pWR1B of the yeast Wingea robertsiae are associated with a killer phenotype. Plasmid 48, 142-148. doi: 10.1016/S0147619X(02)00101-4. Klassen, R., and Meinhardt, F. (2005). Induction of DNA damage and apoptosis in Saccharomyces cerevisiae by a yeast killer toxin. Cellular Microbiology 7, 393-401. doi: 10.1111/j.14625822.2004.00469.x. Klassen, R., Teichert, S., and Meinhardt, F. (2004). Novel yeast killer toxins provoke S-phase arrest and DNA damage checkpoint activation. Molecular Microbiology 53, 263-273. doi: 10.1111/j.13652958.2004.04119.x. Komárek, M., Čadková, E., Chrastný, V., Bordas, F., and Bollinger, J.-C. (2010). Contamination of vineyard soils with fungicides: A review of environmental and toxicological aspects. Environment International 36(1), 138-151. doi: 10.1016/j.envint.2009.10.005. Lamichhane, J. R., Osdaghi, E., Behlau, F., Köhl, J., Jones, J. B., and Aubertot, J.-N. (2018). Thirteen decades of antimicrobial copper compounds applied in agriculture. A review. Agronomy for Sustainable Development 38(3), 28. doi: 10.1007/s13593-018-0503-9. Landum, M. C., Félix, M. d. R., Alho, J., Garcia, R., Cabrita, M. J., Rei, F. , et al. (2016). Antagonistic activity of fungi of Olea europaea L. against Colletotrichum acutatum . Microbiological Research 183, 100-108. doi: 10.1016/j.micres.2015.12.001. Lassois, L., de Bellaire, L., and Jijakli, M. H. (2008). Biological control of crown rot of bananas with Pichia anomala strain K and Candida oleophila strain O. Biological Control 45, 410-418. doi: 10.1016/j.biocontrol.2008.01.013. Lawrence, J. S., Campêlo, A., and Figueiredo, J. M. (1991). Diseases of cacao. li - Fungal diseases which occur on the leaves, branches and trunk (in Portuguese). Agrotrópica (Brazil) 3(1), 1-14. 37 Leoni, C., Bruzzone, J., Villamil, J. J., Martínez, C., Montelongo, M. J., Bentancur, O. , et al. (2018). Percentage of anthracnose ( Colletotrichum acutatum s.s.) acceptable in olives for the production of extra virgin olive oil. Crop Protection 108, 47-53. doi: 10.1016/j.cropro.2018.02.013. Lima, J. R., Gondim, D. M. F., Oliveira, J. T. A., Oliveira, F. S. A., Gonçalves, L. R. B., and Viana, F. M. P. (2013). Use of killer yeast in the management of postharvest papaya anthracnose. Postharvest Biology and Technology 83, 58-64. doi: 10.1016/j.postharvbio.2013.03.014. Liu, G. L., Chi, Z., Wang, G. Y., Wang, Z. P., Li, Y., and Chi, Z. M. (2015). Yeast killer toxins, molecular mechanisms of their action and their applications. Critical Reviews in Biotechnology 35(2), 222234. doi: 10.3109/07388551.2013.833582. Liu, J., Sui, Y., Wisniewski, M., Droby, S., and Liu, Y. (2013). Review: Utilization of antagonistic yeasts tomanage postharvest fungal diseases of fruit. International Journal of Food Microbiology 167, 153-160. doi: 10.1016/j.ijfoodmicro.2013.09.004. Liu, J., Wisniewski, M., Droby, S., Norelli, J., Hershkovitz, V., Tian, S. , et al. (2012). Increase in antioxidant gene transcripts, stress tolerance and biocontrol efficacy of Candida oleophila following sublethal oxidative stress exposure. FEMS Microbiology Ecology 80, 578-590. doi: 10.1111/j.15746941.2012.01324.x. Liu, J., Wisniewski, M., Droby, S., Tian, S., Hershkovitz, V., and Tworkoski, T. (2011). Effect of heat shock treatment on stress tolerance and biocontrol efficacy of Metschnikowia fructicola . FEMS Microbiology Ecology 76, 145-155. doi: 10.1111/j.1574-6941.2010.01037.x. Liu, S. Q., and Tsao, M. (2008). Biocontrol of dairy moulds by antagonistic dairy yeast Debaryomyces hansenii in yoghurt and cheese at elevated temperatures. Food Control 20, 852-855. doi: 10.1016/j.foodcont.2008.10.006. Long, C. A., Wu, Z., and Deng, B. X. (2005). Biological control of Penicillium italicum of Citrus and Botrytis cinerea of Grape by Strain 34-9 of Kloeckera apiculata . European Food Research and Technology 221, 197-201. doi: 10.1007/s00217-005-1199-z. Lopes, M. R., Klein, M. N., Ferraz, L. P., da Silva, A. C., and Kupper, K. C. (2015). Saccharomyces cerevisiae : A novel and efficient biological control agent for Colletotrichum acutatum during preharvest. Microbiological Research 175, 93-99. doi: 10.1016/j.micres.2015.04.003. 44 Ruy, F., Vercesi, A. E., and Kowaltowski, A. J. (2006). Inhibition of specific electron transport pathways leads to oxidative stress and decreased Candida albicans proliferation. Journal of Bioenergetics and Biomembranes 38, 129-135. doi: 10.1007/s10863-006-9012-7. Samuels, G. J., Pardo-Schultheiss, R., Hebbar, K. P., Lumsden, R. D., Bastos, C. N., Costa, J. C. , et al. (2000). Trichoderma stromaticum sp. nov., a parasite of the cacao witches broom pathogen. Mycological Research 104(6), 760-764. doi: 10.1017/S0953756299001938. Santos, A., and Marquina, D. (2004). Killer toxin of Pichia membranifaciens and its possible use as a biocontrol agent against grey mould disease of grapevine. Microbiology 150, 2527-2534. doi: 10.1099/mic.0.27071-0. Santos, A., San Mauro, M., Bravo, E., and Marquina, D. (2009). PMKT2, a new killer toxin from Pichia membranifaciens , and its promising biotechnological properties for control of the spoilage yeast Brettanomyces bruxellensis . Microbiology 155, 624-634. doi: 10.1099/mic.0.023663-0. Santos Filho, L. P., Freire, E. S., and Carzola, I. M. (1998). Estimation of cacao production losses caused by witches' broom ( Crinipellis perniciosa (Stahel) Singer) in Bahia (in Portuguese). Agrotrópica (Brazil) 10(3), 127-130. Saravanakumar, D., Ciavorella, A., Spadaro, D., Garibaldi, A., and Gullino, M. L. (2008). Metschnikowia pulcherrima strain MACH1 outcompetes Botrytis cinerea , Alternaria alternata and Penicillium expansum in apples through iron depletion. Postharvest Biology and Technology 49, 121-128. doi: 10.1016/j.postharvbio.2007.11.006. Scarpari, L. M., Meinhardt, L. W., Mazzafera, P., Pomella, A. W., Schiavinato, M. A., Cascardo, J. C. , et al. (2005). Biochemical changes during the development of witches' broom: the most important disease of cocoa in Brazil caused by Crinipellis perniciosa . Journal of Experimental Botany 56, 865-877. doi: 10.1093/jxb/eri079. Schaffrath, R., and Meinhardt, F. (2005). "Kluyveromyces lactis zymocin and related plasmid-encoded yeast killer toxins," in Microbial Protein Toxins. Topics in Current Genetics, vol 11, eds. M. J. Schmitt & R. Schaffrath. (Berlin, Heidelberg, Germany: Springer-Verlag), 133-155. Schena, L., Mosca, S., Cacciola, S. O., Faedda, R., Sanzani, S. M., Agosteo, G. E. , et al. (2014). Species of the Colletotrichum gloeosporioides and C. boninense complexes associated with olive anthracnose. Plant Pathology 63(2), 437-446. doi: 10.1111/ppa.12110. 45 Scherm, B., Ortu, G., Muzzu, A., Budroni, M., Arras, G., and Migheli, Q. (2003). Biocontrol activity of antagonistic yeasts against Penicillium expansum on apple. Journal of Plant Pathology 85, 205213. doi: 10.4454/jpp.v85i3.1032. Schmitt, M. J., and Breinig, F. (2002). The viral killer system in yeast: from molecular biology to application. FEMS Microbiology Reviews 26, 257-276. doi: 10.1111/j.15746976.2002.tb00614.x. Schmitt, M. J., and Breinig, F. (2006). Yeast viral killer toxins: lethality and self-protection. Nature Reviews Microbiology 4(3), 212-221. doi: 10.1038/nrmicro1347. Schulz-Bohm, K., Martín-Sánchez, L., and Garbeva, P. (2017). Microbial Volatiles: Small Molecules with an Important Role in Intraand Inter-Kingdom Interactions. Frontiers in Microbiology 8, 2484. doi: 10.3389/fmicb.2017.02484. Sergeeva, V. (2011). What is effective against anthracnose? Olive Press 22–24. The Olive Press 17. Sergeeva, V. (2014). The Role Of Epidemiology Data In Developing Integrated Management Of Anthracnose In Olives - A Review. Acta Horticulturae 1057, 163-168. doi: 10.17660/ActaHortic.2014.1057.19. Sharma, M., and Kulshrestha, S. (2015). Colletotrichum gloeosporioides : An Anthracnose Causing Pathogen of Fruits and Vegetables. Biosciences Biotechnology Research Asia 12, 1233-1246. doi: 10.13005/bbra/1776. Sharma, R. R., Singh, D., and Singh, R. (2009). Biological control of postharvest diseases of fruits and vegetables by microbial antagonists: A review. Biological Control 50, 205-221. doi: 10.1016/j.biocontrol.2009.05.001. Silveira, R. (2013). Produtores da Bahia serão os primeiros a usar biofungicida Tricovab nas lavouras de cacau. Available at: https://canalrural.uol.com.br/noticias/produtores-bahia-serao-primeirosusar-biofungicida-tricovab-nas-lavouras-cacau-29175/ [Accessed February 14, 2019]. Spadaro, D., and Droby, S. (2016). Development of biocontrol products for postharvest diseases of fruit: the importance of elucidating the mechanisms of action of yeast antagonists. Trends in Food Science & Technology 47, 39-49. doi: 10.1016/j.tifs.2015.11.003. 46 Spadaro, D., and Gullino, M. L. (2004). State of the art and future prospects of the biological control of postharvest fruit diseases. International Journal of Food Microbiology 91, 185-194. doi: 10.1016/S0168-1605(03)00380-5. Spotts, R. A., Cervantes, L. A., and Facteau, T. J. (2002). Integrated control of brown rot of sweet cherry fruit with a preharvest fungicide, a postharvest yeast, modified atmosphere packaging, and cold storage temperature. Postharvest Biology and Technology 24(3), 251-257. doi: 10.1016/S09255214(01)00155-7. Squicciarini, M. P., and Swinnen, J. (2016). The Economics of Chocolate. Oxford, UK: Oxford University Press. Stenmark, P., and Nordlund, P. (2003). A prokaryotic alternative oxidase present in the bacterium Novosphingobium aromaticivorans . FEBS Letters 552, 189-192. doi: 10.1016/S00145793(03)00920-7. Sui, Y., and Liu, J. (2014). Effect of glucose on thermotolerance and biocontrol efficacy of the antagonistic yeast Pichia guilliermondii . Biological Control 74, 59-64. doi: 10.1016/j.biocontrol.2014.04.003. Sundh, I., and Melin, P. (2011). Safety and regulation of yeasts used for biocontrol or biopreservation in the food or feed chain. Antonie van Leeuwenhoek 99(1), 113-119. doi: 10.1007/s10482-0109528-z. Suzuki, C. (2005). "Acidiphilic structure and killing mechanism of the Pichia farinosa killer toxin SMTK," in Microbial Protein Toxins. Topics in Current Genetics, vol 11, eds. M. J. Schmitt & R. Schaffrath. (Berlin, Heidelberg, Germany: Springer-Verlag), 189-214. Suzuki, C., Ando, Y., and Machida, S. (2001). Interaction of SMKT, a killer toxin produced by Pichia farinosa , with the yeast cell membranes. Yeast 18, 1471-1478. doi: 10.1002/yea.791. Suzzi, G., Romano, P., Ponti, I., and Montuschi, C. (1995). Natural wine yeasts as biocontrol agents. Journal of Applied Bacteriology 78, 304-308. doi: 10.1111/j.1365-2672.1995.tb05030.x. Syed Ab Rahman, S. F., Singh, E., Pieterse, C. M. J., and Schenk, P. M. (2018). Emerging microbial biocontrol strategies for plant pathogens. Plant Science 267, 102-111. doi: 10.1016/j.plantsci.2017.11.012. Talhinhas, P., Gonçalves, E., Sreenivasaprasad, S., and Oliveira, H. (2015). Virulence diversity of anthracnose pathogens ( Colletotrichum acutatum and C. gloeosporioides species complexes) on 47 eight olive cultivars commonly grown in Portugal. European Journal of Plant Pathology 142(1), 73-83. doi: 10.1007/s10658-014-0590-7. Talhinhas, P., Loureiro, A., and Oliveira, H. (2018). Olive anthracnose: a yield‐ and oil quality‐degrading disease caused by several species of Colletotrichum that differ in virulence, host preference and geographical distribution. Molecular Plant Pathology 19(8), 1797-1807. Talhinhas, P., Mota-Capitão, C., Martins, S., Ramos, A. P., Neves-Martins, J., Guerra-Guimarães, L. , et al. (2011). Epidemiology, histopathology and aetiology of olive anthracnose caused by Colletotrichum acutatum and C. gloeosporioides in Portugal. Plant Pathology 60(3), 483-495. Taqarort, N., Echairi, A., Chaussod, R., Nouaim, R., Boubaker, H., Benaoumar, A. A. , et al. (2008). Screening and identification of epiphytic yeasts with potential for biological control of green mold of citrus fruits. World Journal of Microbiology and Biotechnology 24, 3031-3038. doi: 10.1007/s11274-008-9849-5. Teixeira, P. J. P. L., Thomazella, D. P. d. T., and Pereira, G. A. G. (2015). Time for Chocolate: Current Understanding and New Perspectives on Cacao Witches’ Broom Disease Research. PLOS Pathogens 11(10), e1005130. doi: 10.1371/journal.ppat.1005130. Teixidó, N., Usall, J., and Viñas, I. (1999). Efficacy of preharvest and postharvest Candida sake biocontrol treatments to prevent blue mould on apples during cold storage. International Journal of Food Microbiology 50, 203-210. doi: 10.1016/S0168-1605(99)00105-1. Thomazella, D. P., Teixeira, P. J., Oliveira, H. C., Saviani, E. E., Rincones, J., Toni, I. M. , et al. (2012). The hemibiotrophic cacao pathogen Moniliophthora perniciosa depends on a mitochondrial alternative oxidase for biotrophic development. New Phytologist 194, 1025-1034. doi: 10.1111/j.1469-8137.2012.04119.x. Tian, S. P., Qin, G. Z., Xu, Y., and Wang, Y. S. (2004). Application of Antagonistic Yeasts Under Field Conditions and Their Biocontrol Ability Against Postharvest Diseases of Sweet Cherry. Acta Botanica Sinica 46(11), 1324-1330. Trevizan, S. D. P., and Marques, M. (2002). Social and economic impacts of the cacao crisis: study of a community case (in Portuguese). Agrotrópica (Brazil) 14, 127-136. Van Aken, O., Giraud, E., Clifton, R., and Whelan, J. (2009). Alternative oxidase: a target and regulator of stress responses. Physiologia Plantarum 137, 354-361. doi: 10.1111/j.13993054.2009.01240.x. 48 Vanlerberghe, G. C., Cvetkovska, M., and Wang, J. (2009). Is the maintenance of homeostatic mitochondrial signaling during stress a physiological role for alternative oxidase? Physiologia Plantarum 137, 392-406. doi: 10.1111/j.1399-3054.2009.01254.x. Veiga, A., Arrabaça, J. D., and Loureiro-Dias, M. C. (2003). Cyanide-resistant respiration, a very frequent metabolic pathway in yeasts. FEMS Yeast Research 3, 239-245. doi: 10.1016/S15671356(03)00036-9. Verweij, P. E., Snelders, E., Kema, G. H., Mellado, E., and Melchers, W. J. (2009). Azole resistance in Aspergillus fumigatus : a side-effect of environmental fungicide use? The Lancet Infectious Diseases 9, 789-795. doi: 10.1016/S1473-3099(09)70265-8. Walker, G. M., McLeod, A. H., and Hodgson, V. J. (1995). Interactions between killer yeasts and pathogenic fungi. FEMS Microbiology Letters 127, 213-222. doi: 10.1111/j.15746968.1995.tb07476.x. Wang, X., Chi, Z., Yue, L., and Li, J. (2007). Purification and characterization of killer toxin from a marine yeast Pichia anomala YF07b against the pathogenic yeast in crab. Current Microbiology 55, 396401. doi: 10.1007/s00284-007-9010-y. Wang, Y., Yu, T., Xia, J., Yu, D., Wang, J., and Zheng, X. (2010). Biocontrol of postharvest gray mold of cherry tomatoes with the marine yeast Rhodosporidium paludigenum . Biological Control 53, 178182. doi: 10.1016/j.biocontrol.2010.01.002. Ward, D. M., Chen, O. S., Li, L., Kaplan, J., Bhuiyan, S. A., Natarajan, S. K. , et al. (2018). Altered sterol metabolism in budding yeast affects mitochondrial iron-sulfur (Fe-S) cluster synthesis. Journal of Biological Chemistry 293, 10782-10795. doi: 10.1074/jbc.RA118.001781. Wei, Y., Mao, S., and Tu, K. (2014). Effect of preharvest spraying Cryptococcus laurentii on postharvest decay and quality of strawberry. Biological Control 73, 68-74. doi: 10.1016/j.biocontrol.2014.02.016. Weiler, F., and Schmitt, M. J. (2003). Zygocin, a secreted antifungal toxin of the yeast Zygosaccharomyces bailii , and its effect on sensitive fungal cells. FEMS Yeast Research 3, 69-76. doi: 10.1111/j.1567-1364.2003.tb00140.x. Whaley, S. G., Berkow, E. L., Rybak, J. M., Nishimoto, A. T., Barker, K. S., and Rogers, P. D. (2017). Azole Antifungal Resistance in Candida albicans and Emerging Nonalbicans Candida Species. Frontiers in Microbiology 7, 2173. doi: 10.3389/fmicb.2016.02173. 49 Wharton, P. S., and Diéguez-Uribeondo, J. (2004). The biology of Colletotrichum acutatum . Anales del Jardín Botánico de Madrid 61(1), 3-22. doi: 10.3989/ajbm.2004.v61.i1.61. Wickramasuriya, A. M., and Dunwell, J. M. (2018). Cacao biotechnology: current status and future prospects. Plant Biotechnology Journal 16, 4-17. doi: 10.1111/pbi.12848. Wilson, D. (1995). Endophyte: The Evolution of a Term, and Clarification of Its Use and Definition. Oikos 73(2), 274-276. doi: 10.2307/3545919. Wisniewski, M., Biles, C., Droby, S., McLaughlin, R., Wilson, C., and Chalutz, E. (1991). Mode of action of the postharvest biocontrol yeast, Pichia guilliermondii . I. Characterization of attachment to Botrytis cinerea . Physiological and Molecular Plant Pathology 39, 245-258. doi: 10.1016/08855765(91)90033-E. Wisniewski, M., Droby, S., Norelli, J., Liu, J., and Schena, L. (2016). Alternative management technologies for postharvest disease control: The journey from simplicity to complexity. Postharvest Biology and Technology 122, 3-10. doi: 10.1016/j.postharvbio.2016.05.012. Wongsuk, T., Pumeesat, P., and Luplertlop, N. (2016). Fungal quorum sensing molecules: Role in fungal morphogenesis and pathogenicity. Journal of Basic Microbiology 56(5), 440-447. doi: 10.1002/jobm.201500759. Wuster, A., and Babu, M. M. (2007). "Chemical molecules that regulate transcription and facilitate cellto-cell communication," in Wiley encyclopedia of chemical biology, ed. T. P. Begley. John Wiley & Sons, Chichester), 1-11. Yang, C., Hamel, C., Vujanovic, V., and Gan, Y. (2011). Fungicide: Modes of Action and Possible Impact on Nontarget Microorganisms. ISRN Ecology 2011. doi: 10.5402/2011/130289. Zhang, D., Spadaro, D., Garibaldi, A., and Gullino, M. L. (2010a). Selection and evaluation of new antagonists for their efficacy against postharvest brown rot of peaches. Postharvest Biology and Technology 55, 174-181. doi: 10.1016/j.postharvbio.2009.09.007. Zhang, V. Q., Gamarra, S., Garcia-Effron, G., Park, S., Perlin, D. S., and Rao, R. (2010b). Requirement for Ergosterol in V-ATPase Function Underlies Antifungal Activity of Azole Drugs. PLOS Pathogens 6(6), e1000939. doi: 10.1371/journal.ppat.1000939. Zhao, L., Zhang, H., Li, J., Cui, J., Zhang, X., and Ren, X. (2012). Enhancement of Biocontrol Efficacy of Pichia carribbica to Postharvest Diseases of Strawberries by Addition of Trehalose to the Growth 50 Medium. International Journal of Molecular Sciences 13, 3916-3932. doi: 10.3390/ijms13033916. Zhao, Y., Tu, K., Shao, X., Jing, W., and Su, Z. (2008). Effects of the yeast Pichia guilliermondii against Rhizopus nigricans on tomato fruit. Postharvest Biology and Technology 49, 113-120. doi: 10.1016/j.postharvbio.2008.01.001. Zhao, Y., Wang, R., Tu, K., and Liu, K. (2011). Efficacy of preharvest spraying with Pichia guilliermondii on postharvest decay and quality of cherry tomato fruit during storage. African Journal of Biotechnology 10(47), 9613-9622. doi: 10.5897/AJB11.870. CHAPTER 2 Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease The work presented in this chapter has been submitted for publication: Ferraz, P., Amorim-Rodrigues, M., Cássio, F. and Lucas, C. Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease 52 ABSTRACT BACKGROUND: Cacao Witches’ Broom Disease (WBD) is caused by Moniliophthora perniciosa, a fungus that infects all the plant tissues, at any stage of its life cycle. Chemical fungicides are ineffective against M. perniciosa. Presently, WBD is tentatively controlled using a biocide, Trichoderma stromaticum , expensively produced by demand, and therefore economically unsustainable. RESULTS: A group of wild yeasts from several species, isolates from Brazilian fermentation industries, were tested for their ability to antagonize M. perniciosa strains from South and Central America infected orchards. Tests were done in vitro, in solid and liquid media. Three yeasts were found that efficiently kill 6 strains of M. perniciosa . Death was confirmed by methylene blue and propidium iodide staining and requires the physical contact between the yeasts and the mycelium. Two yeast strains were chosen, Wickerhamomyces anomalus #1105 and Saccharomyces cerevisiae #1112 from spontaneous fermentations of the production of cachaça, the Brazilian spirit, and the S. cerevisiae PE2 from bioethanol industrial fermentations in Brazil. They are effective at three temperatures, under starvation, at different culture stages or growing old. CONCLUSIONS: These results are the initial step towards the formulation of a new eco-friendly and effective alternative for controlling WBD, based on the application of live yeast without the need for the purification of a specific antifungal peptide/compound. Moreover, results suggest that spontaneous fermentations may provide microbial biodiversity to be locally applied in the control of the disease, improving the sustainability of small cacao producers in remote areas. Keywords Cacao; Witches Broom Disease ; Moniliophthora perniciosa ; yeast; antagonism 53 INTRODUCTION The cacao plant ( Theobroma cacao L.) is one of the most valuable crops worldwide (Pohlan and Pérez, 2010; Teixeira et al. , 2015), responsible for important fractions of the economic revenue of the countries from Central and South America and Africa within the Cacao Belt. T. cacao is affected by several diseases, the most severe being the Witches’ Broom Disease (WBD) caused by the basidiomycete fungus Moniliophthora perniciosa (formerly Crinipellis perniciosa ) (Purdy and Schmidt, 1996; Aime and PhillipsMora, 2005; Teixeira et al. , 2015). This disease is responsible for major crop losses with large socialeconomic consequences, particularly in Brazil which cacao production decreased more than 70% in a period of 10 years after the onset of the disease (Trevizan and Marques, 2002; Meinhardt et al. , 2008; Teixeira et al. , 2015). The severity of WBD is closely related to the virulence of M. perniciosa , which comes from its ability to virtually infect all cacao plant tissues at all stages of the plant life cycle (Meinhardt et al. , 2008; Ferraz et al. , 2019). As a hemibiotrophic fungus, M. perniciosa has two distinct phases: a biotrophic and a saprotrophic (reviewed by Ferraz et al. , 2019). After the initial infection, the pathogen induces hypertrophy and hyperplasia, causing a disorganized proliferation of the infected vegetative meristems of axillary shoots, which results in the formation of green brooms, a structure composed of abnormal stems. Several weeks after the development of these structures, the infected plant tissues become necrotic due to a series of cell death events, forming a structure named dry broom (Meinhardt et al. , 2008). M. perniciosa then colonizes those necrotic plant cells and generates pink-coloured basidiocarps producing 2 to 3.5 million spores each (Almeida et al. , 1997). The spores are mainly released at night under optimal conditions of temperature and humidity, being disseminated by water and wind, and can endure and remain latent in the soil or inside pruned plant branches for long periods of time (Meinhardt et al. , 2008; Pohlan and Pérez, 2010). All these factors contribute to the exceptional virulence of M. perniciosa and explain why a whole plantation is compromised after the initial infection of one cacao plant. The conventional chemical fungicides used to control the spread of fungal plant diseases, such as copper or azole-based compounds, are ineffective against M. perniciosa (Medeiros et al. , 2010). In addition, the use of chemical fungicides has been restricted in most cacao producing countries due to their high cost, and to the risks associated with chemical contamination of the cacao fruits and chocolate (Marelli et al. , 2009; Verweij et al. , 2009; Nunes, 2012). Currently, WBD has only one management method implemented in Brazil, consisting in spraying the infected plants with Tricovab®, a live suspension of 60 RESULTS AND DISCUSSION Optimization of M. perniciosa cultivation conditions M. perniciosa strains used in this work (Table 1) were originally isolated from cacao plants and fruits infected with WBD in South American countries. This species of filamentous fungus is very aggressive and resilient but grows preferably within a relatively narrow range of temperatures, between 20-30 ºC (ICCO.org; Purdy and Schmidt, 1996). Considering that the majority of yeasts are best cultured at 30 ºC, this temperature was chosen to cultivate the fungal strains. Optimal pH, on the other hand, was determined by quantifying fungal growth rates in MEA, a well-known fungal growth media, adjusted to pH 4.0, 4.5, 5.0, 5.5 or 6.0. Results (Figure 1) showed that M. perniciosa strains vary considerably between each other. The fastest growing strain was F from Brazil with a specific growth rate of 2.7 mm.day-1 at pH 5, while the strain B from Ecuador was the slowest, growing at Gr=0.47 mm.day-1 at pH 4.5. No statistically significant differences were observed between the growth at pH 5 to 6. Growth rates were generally lowest at pH 4.0, with a latency phase lasting 1 day (strains C, D, F and I) or more (remaining strains). Similarly, a latency phase of 1 day was also observed for strains C, E, H and I at pH 4.5. Figure 1 . Variation of the growth rate of M. perniciosa strains in Table 1 determined in MEA at 30°C at different pH. Different letters represent significant differences across different pH values. § represents a latency phase prior to the initial mycelia growth (§ 1 day of latency; §§ >1 day of latency). Results are data from at least 3 independent replicates. 61 These assays were repeated in PDA and no statistically significant differences were observed between the growth rates in either media (not shown). Fungal strains were thus further cultivated on MEA or PDA at 30°C and pH 5.5 to avoid latency. From the same assays, it was also established that 10 days is enough to test fungal growth phenotypes. Optimization of yeast vs fungi antagonism assays The yeasts chosen for this work (Table 2) can be grouped in two sets. The first contains strains of Wickerhamomyces anomalus (previously Pichia anomala or Hansenula anomala ) obtained from a credited culture collection (PYCC - Portuguese Yeast Culture Collection), and the second contains isolates from fermentative processes in Brazil. These yeasts were chosen based on their dominant nature in microbial mixtures, particularly the wild yeasts isolated from the spontaneous fermentations used for the production of cachaça , the Brazilian spirit (da Conceição et al. , 2015), or used in a more controlled industrial process of sugarcane juice fermentation for bioethanol production (Lopes et al. , 2016). Moreover, W. anomalus strains were chosen considering their origin from diverse natural niches, possible sources for strong antagonism (Baysal and Silme, 2018), and the fact that this is a killer species active against a large variety of other yeasts (Abranches et al. , 1998; Fredlund et al. , 2002). Broad killers are known to be also effective against bacteria and filamentous fungi, including phytopathogens (Suzzi et al. , 1995; Walker et al. , 1995; Platania et al. , 2012; Lima et al. , 2013; Parafati et al. , 2015). W. anomalus strains in Table 2 have for long been kept in a culture collection, they do not originate directly from nature. Bearing in mind that they might have lost their killer ability, they were firstly assayed in that regard against a broad sensitive strain of Meyerozyma guilliermondii (previously Pichia guilliermondii ) (PYCC 2734) (Aguiar and Lucas, 2000; da Silva et al. , 2008) . W. anomalus strains were cultivated in the media and growth conditions established for M. perniciosa, i.e. on MEA with MB at pH 5.5 and 30 °C, in which conditions they displayed regular growth, and were therefore subsequently tested for their killer phenotype in these same conditions. All the strains of W. anomalus caused a blue inhibition halo on M. guilliermondii indicative of cell death (not shown), so none was excluded from the assessment of inhibitory effect over fungal growth. For this purpose, the M. perniciosa fastest growing strains F and G from Brazil and A from Ecuador were used. The development of mycelium was followed up to 10 days. All the W. anomalus strains affected the fungal growth although to a different extent. The mycelia developed freely in the opposite direction of the yeast strikeout, while the extent of its development varied 62 in the space between the yeast and the fungal plug. Based on this variation, three levels of response were identified, which were converted into an empirical scale of yeast/fungus interaction (Figure 2): level 0, corresponds to a high fungal resistance, in which the mycelium overgrows the yeast eventually filling the entire plate; level 1, corresponds to a weak inhibitory effect, wherein the mycelium fills the gap between the yeast and the plug and stops growing upon contact with the yeast culture; and level 2 corresponds to a clear inhibitory effect, with the formation of a blue inhibition halo facing the yeast. The results of each combination of yeast/fungal strain are presented in Table 3 according to this scale. As can be seen, M. perniciosa strains from Brazil were more resistant than the strain from Ecuador, which was inhibited by most of the W. anomalus strains. These results were obtained in MEA supplemented with MB. As controls, identical assays were made on PDA, with and without MB, and MEA without MB, plating all the strains of W. anomalus against the more sensitive strain of M. perniciosa , A from Ecuador. Results were identical to the ones in Table 3, showing that neither the medium nor the presence of MB affected the yeast-fungus antagonism (not shown). Assays were subsequently extended to the industrial yeasts in Table 2. Results (Table 3) showed a more irregular response from the fungus in the presence of these strains than with W. anomalus , which was weaker than predicted according to their origin (Abranches et al. , 1998; Alonsodel-Real et al. , 2019; Hatoum et al. , 2012; Mannazzu et al. , 2019). Moreover, the assays were also done using the fungal strains B and C, absent from Table 3. The growth of these fungi in the antagonism assays was extremely irregular generating unreliable results, reason why these fungi were excluded from further assays. In total, most fermentation yeasts strains were less able to antagonise M. perniciosa than W. anomalus. Noticeably, fungal strain F from Brazil stood out for displaying the highest resistance to all yeasts tested. Figure 2 . Empirical classification scale of antagonistic response in solid media. 0 represents the absence of inhibition of any kind, the fungus eventually growing on top of the yeast culture; 1 represents a weak inhibitory response in which case the fungus grows up to the limit of the yeast culture without overgrowing it, and 2 represents a clear antagonistic effect. 63 Table 3. Results of the antagonism assays between M. perniciosa and yeasts in solid medium. The fastest-growing fungal strains, A from Ecuador and F and G from Brazil were tested against all the yeasts in Table 2. Further assays were done using all the remaining fungal strains against the yeasts originating from fermentation processes. Results using M. perniciosa strains D and E from Ecuador and H from Venezuela are presented. The inhibitory effect was rated from 0 to 2, according to the empirical scale in Figure 2. Results presented were identical in three independent replicates. A table cell displaying 3 numbers (e.g. 1/1/2) represents a situation where each plate had a different result. M. perniciosa strains Yeasts Wickerhamomyces anomalus #2495 #2505 #3294 #4121 #4380 #4554 #5008 A 2 2 2 2 1 2 1 F 1 1 1 2 1 1 1 G 1 1 1 1 2 1 0 Fermentative yeast strains #1 #2 #3 #4 #5 #6 #7 CAT1 PE2 A 1 2 2 1 1 0 1 2 2 F 1 0 1 0 0 0 0 0 0 G 0 1 1/1/2 1 0 0 1 2 1 D 1 2 1 1 1 0 1 2 1 E 1 1/1/2 1 2 0 1 2 2 1/1/2 H 1 1 1 1 0 0 1 1 1 64 Is M. perniciosa growth inhibition caused by a volatile compound? Mycelia strong growth inhibition occurred at a certain distance between the yeast and the fungal cells ( e.g. Figure 2, antagonism level 2). This suggests that probably the yeast strains secreted some kind of compound that signals the fungal cells. Yeasts can secrete a soluble molecule which diffuses through the agar, eventually reaching the mycelium (Schmitt and Breinig, 2006; Lopes et al. , 2015; Liu et al. , 2018), or a volatile compound that may affect the fungus at some distance (Fialho et al. , 2016; Schulz-Bohm et al. , 2017). To evaluate whether the antagonistic response observed against M. perniciosa strains corresponded to a volatile compound secreted by the yeast, the above described antagonism assays were repeated using septate Petri dishes. These prevent the diffusion of molecules through the agar but allow the organisms in the two sides of the plate to share the atmosphere. M. perniciosa strain G was challenged with the presence of the yeast strains #3 and CAT1 (Table 3). The yeast strain #1 was used as negative control. Unlike before, no inhibitory response was observed for any of the combinations used ( e.g. Figure 3), indicating that the fungal growth inhibition probably involves the diffusion of a non-volatile compound through the agar. Nevertheless, the competition for nutrients and space cannot be discarded at this point. Figure 3. Example of an antagonism assay in divided Petri dishes (left picture), with the comparison with the previous tests in solid medium using the same yeast vs fungus combination (right picture). 65 Antagonism assays in liquid media Since the fungi have a slower growth rate in comparison with yeasts, it is necessary to incubate the plates for 10 days to score the antagonism effect. This is though unfavourable to the maintenance of a fully viable yeast culture. Hence, blue halo of cell death around the yeast biomass can appear, or in some cases the whole biomass can become blue. This could be a reason underlying the diverse and irregular response depicted in Table 3, since each yeast strain will stay alive and metabolically healthy for different periods of time. For this reason, the possibility of assaying antagonism more efficiently in liquid medium, which provides a direct contact between hyphae and yeasts, was considered. Moreover, the fermentation yeasts were chosen to proceed with the antagonism in liquid media, based on that (i) the set also includes one strain of W. anomalus isolated from spontaneous fermentations of cachaça production and this way the species is represented, (ii) these strains should be more resilient bearing in mind their putative future utilization in the field, and (iii) they originate from de same geographic region of the WBD, which in case of in field application, could avoid the potential imbalance of the cacao plantations ecosystem caused by the introduction of an alien microbe. Importantly, since these are strains used by the industry, their legalization, commercialization and acceptance should be facilitated. To optimize the conditions for the assays in liquid media, fungi and the industrial yeasts were firstly cultivated in ME or YPD, pH 5.5 and 30 °C. M. perniciosa produced abundant mycelia in the form one or several cotton ball-like large conglomerates growing in size over time (Figure 4A). Yeasts globally performed slightly better on YPD than on ME (Figure 4A), which is consistent with these yeast strains preference for glucose. Still, their growth rates allowed to consider the use of ME. Preliminary antagonism assays were performed choosing in Table 3 the stronger antagonizing yeasts against the faster growing fungal strains F and G: strains #1 and #3 vs F, using CAT1 as negative control, and #3 and CAT1 vs G, using #1 as negative control. Separate controls of fungi and yeasts growing alone were also performed. The fungi were inoculated through a mycelium agar plug placed in a glass tube inoculated with 108 cells/mL from an exponentially growing yeast culture. As expected, the yeasts grew faster than the fungi, filling the growth medium. After 10 days of co-incubation the medium was decanted to check for the presence or absence of mycelium. Interestingly, unlike in solid medium, yeasts inhibited fungal growth in all the combinations tested, including the negative controls (not shown). Based on these preliminary results, antagonisms assays in liquid medium were expanded to the full set of combinations between all the fermentation yeast strains and the fungal strains F and G. 66 67 Figure 4. (A) Fungal cultivation of M. perniciosa strain F in ME and YPD at 30 °C, with the development of abundant mycelia in the form one or several cotton ball-like large conglomerates (left panel) and the growth curves of the industrial yeast strains in ME and YPD at 30 °C, determined by OD at 600 nm (right panel). (B) Empirical classification scale of antagonistic response in liquid media. 0 represents the absence of inhibitory effect, the fungus grows three-dimensionally producing a large conglomerate of hyphae; 1 represents a limited inhibitory response in which case some small mycelium is still formed from the agar plug, and 2 represents a strong inhibitory effect in which case no mycelium is formed. The culture media was removed and substituted to fresh ME medium to allow a better visualization of the results, since the turbidity resulting from yeast growth made it difficult to visualize the degree of inhibition at the end of the assays. Results varied, allowing to establish a second antagonism empirical scale (Figure 4B), in which level 0 represents high fungal resistance, with the formation of the cotton ball-like conglomerates of mycelium identical to the fungal-alone control; level 1 corresponds to a weak inhibitory response, with the development of some mycelia around the fungal agar plug; and level 2 represents a strong inhibition with the total absence of fungal development. Results are scored in Table 4 according to this scale. A large increase in the inhibitory response compared to that obtained in solid medium was obtained. In fact, a strong inhibitory response was observed in 9 out of 18 combinations (50%), in opposition to the 2 out of 18 (11%) observed in solid media (Table 3). The yeast strains #1, #5, #6 and PE-2 elicited a strong inhibitory response, while CAT1 inhibitory effect was strong (level 2) or weak (level 1) in the case of strains G and F respectively. Table 4. Antagonism assays in liquid medium using the combinations between the fermentative yeast strains and M. perniciosa strains F and G. The presence of the inhibitory effect was evaluated on a scale from 0 to 2 according to Figure 2. Results were identical in three independent replicates. Yeasts M. perniciosa strains #1 #2 #3 #4 #5 #6 #7 CAT1 PE2 F 2 0 1 0 2 2 0 1 2 G 2 0 1 0 2 2 0 2 2 68 The difference in responses observed in solid and liquid media may derive from several factors. Liquid medium, in opposition to solid, propitiates the fast growth of the yeast population, faster than the fungal mycelium. The increasing yeast population possibly compete with the mycelium for nutrients. This was reported to occur between yeasts and phytopathogenic fungi, as one of the main modes of antagonism action (Andrews et al. , 1994; Saravanakumar et al. , 2008; Zhang et al. , 2010; Spadaro and Droby, 2016), showing that the inhibition of fungal growth does not always implicate death, being often reversible since the fungal cells can retain their viability (Spadaro and Droby, 2016). Additionally, liquid medium also allows the homogenous spread of the yeast cells and their contact with the hyphae. Finally, yeasts remain metabolically active for a longer period of time than in solid medium, and their permanent contact with the mycelia may trigger the production and secretion of peptides or other compounds with antifungal properties. This means that antagonism may correspond to fungal death or just an inhibition of its replication. Evaluation of fungal death by staining with Methylene Blue and Propidium Iodide To verify if the antagonism observed in liquid medium corresponds to actual fungal death, two well-known markers of cell death were used, Methylene Blue (MB) and Propidium Iodide (PI). MB is a cationic dye which can penetrate both live and dead cells, inside which it binds to negatively charged molecules such as nucleic acids. Living cells are able to reduce the dye and consequently remain colourless. PI can only enter the cells which membrane is disrupted; therefore, only necrotic cells are stained with red fluorescence. Cells stained with MB but not with PI lost viability but preserve the integrity of the plasma membrane, which means that they are possibly dying of apoptosis (Kwolek-Mirek and Zadrag-Tecza, 2014). Results are exemplified in Figure 5. All the yeast/fungal strain combinations evaluated that demonstrated a strong inhibitory response (level 2) presented an almost fully stained mycelium with both MB and PI, while in all the combinations with a weaker inhibitory response (level 1) the mycelium was only partially stained. These results showed that there are two mechanisms underlying M. perniciosa antagonism by yeasts. Yeasts kill the fungi, but in some cases, the relative amounts of yeast cells/mycelium do not appear sufficient to induce the death of the whole fungal culture in the time window of 10 days in which the assays take place. In this case, yeasts just delay fungal growth. Interestingly, it is possible to observe that isolated or clustered yeast cells appear to attach to the killed fungal hyphae, 69 suggesting that contact between the two organisms may be necessary to cause the effect and may even be involved in the mechanisms that cause the fungal cells death. Figure 5. Example of the yeast/fungal strains cultures staining with MB (upper panel) and PI (middle and lower panels, corresponding to the observed bright field and fluorescence of the same picture, respectively). Empirical classification scale based on the intensity of the staining: 0 represents an absence of staining; 1 corresponds to a weak staining; and 2 represents a strong staining of the mycelia. Scale bar: 20 μm. 0 1 2 MB PI 76 causal agent of Witches’ Broom disease. Brazilian Journal of Microbiology 40, 108-110. doi: 10.1590/S1517-83822009000100018 da Conceição, L. E., Saraiva, M. A., Diniz, R. H., Oliveira, J., Barbosa, G. D., Alvarez, F., et al. (2015). Biotechnological potential of yeast isolates from cachaça : the Brazilian spirit. Journal of Industrial Microbiology & Biotechnology 42(2), 237-246. doi: 10.1007/s10295-014-1528-y. da Silva, S., Calado, S., Lucas, C., and Aguiar, C. (2008). Unusual properties of the halotolerant yeast Candida nodaensis Killer toxin, CnKT. Microbiological Research 163(2), 243-251. doi: 10.1016/j.micres.2007.04.002. De Ingeniis, J., Raffaelli, N., Ciani, M., and Mannazzu, I. (2009). Pichia anomala DBVPG 3003 secretes a ubiquitin-like protein that has antimicrobial activity. Applied and Environmental Microbiology 75(4), 1129-1134. doi: 10.1128/aem.01837-08. Dukare, A. S., Paul, S., Nambi, V. E., Gupta, R. K., Singh, R., Sharma, K., et al. (2018). Exploitation of microbial antagonists for the control of postharvest diseases of fruits: a review. Critical Reviews in Food Science and Nutrition 16, 1-16. doi: 10.1080/10408398.2017.1417235. Ferraz, P., Cássio, F., and Lucas, C. (2019). Potential of Yeasts as Biocontrol Agents of the Phytopathogen Causing Cacao Witches’ Broom Disease: Is Microbial Warfare a Solution? Frontiers in Microbiology 10, 1766. doi: 10.3389/fmicb.2019.01766. Fialho, M. B., de Andrade, A., Bonatto, J. M., Salvato, F., Labate, C. A., and Pascholati, S. F. (2016). Proteomic response of the phytopathogen Phyllosticta citricarpa to antimicrobial volatile organic compounds from Saccharomyces cerevisiae . Microbiological Research 183, 1-7. doi: 10.1016/j.micres.2015.11.002. Fredlund, E., Druvefors, U., Boysen, M. E., Lingsten, K. J., and Schnurer, J. (2002). Physiological characteristics of the biocontrol yeast Pichia anomala J121. FEMS Yeast Research 2(3), 395402. doi: 10.1016/s1567-1356(02)00098-3. Hatoum, R., Labrie, S., and Fliss, I. (2012). Antimicrobial and probiotic properties of yeasts: From fundamental to novel applications. Frontiers in Microbiology 3, 421. doi: 10.3389/fmicb.2012.00421. ICCO.org | International Cocoa Organization Website: Pests and Diseases. [http://www.icco.org/aboutcocoa/pest-a-diseases.html (Accessed January 23, 2020)]. 77 Kwolek-Mirek, M., and Zadrag-Tecza, R. (2014). Comparison of methods used for assessing the viability and vitality of yeast cells. FEMS Yeast Research 14(7), 1068-1079. doi: 10.1111/15671364.12202. Lima, J. R., Gondim, D. M. F., Oliveira, J. T. A., Oliveira, F. S. A., Gonçalves, L. R. B., and Viana, F. M. P. (2013). Use of killer yeast in the management of postharvest papaya anthracnose. Postharvest Biology and Technology 83, 58-64. doi: 10.1016/j.postharvbio.2013.03.014. Liu, G. L., Chi, Z., Wang, G. Y., Wang, Z. P., Li, Y., and Chi, Z. M. (2015). Yeast killer toxins, molecular mechanisms of their action and their applications. Critical Reviews in Biotechnology 35(2), 222234. doi: 10.3109/07388551.2013.833582. Liu, J., Sui, Y., Wisniewski, M., Droby, S., and Liu, Y. (2013). Review: Utilization of antagonistic yeasts tomanage postharvest fungal diseases of fruit. International Journal of Food Microbiology 167, 153-160. doi: 10.1016/j.ijfoodmicro.2013.09.004. Liu, Z., Du, S., Ren, Y., and Liu, Y. (2018). Biocontrol ability of killer yeasts ( Saccharomyces cerevisiae ) isolated from wine against Colletotrichum gloeosporioides on grape. Journal of Basic Microbiology 58, 60-67. doi: 10.1002/jobm.201700264. Lopes, M. L., Paulillo, S. C., Godoy, A., Cherubin, R. A., Lorenzi, M. S., Giometti, F. H., et al. (2016). Ethanol production in Brazil: a bridge between science and industry. Brazilian Journal of Microbiology 47(1), 64-76. doi: 10.1016/j.bjm.2016.10.003. Lopes, M. R., Klein, M. N., Ferraz, L. P., da Silva, A. C., and Kupper, K. C. (2015). Saccharomyces cerevisiae : A novel and efficient biological control agent for Colletotrichum acutatum during preharvest. Microbiological Research 175, 93-99. doi: 10.1016/j.micres.2015.04.003. Magliani, W., Conti, S., Gerloni, M., Bertolotti, D., and Polonelli, L. (1997). Yeast killer systems. Clinical Microbiology Reviews 10(3), 369-400. doi: 10.1128/CMR.10.3.369. Mannazzu, I., Domizio, P., Carboni, G., Zara, S., Zara, G., Comitini, F., et al. (2019). Yeast killer toxins: from ecological significance to application. Critical Reviews in Biotechnology 39(5), 603-617. doi: 10.1080/07388551.2019.1601679. Marelli, J.-P., Maximova, S. N., Gramacho, K. P., Kang, S., and Guiltinan, M. J. (2009). Infection Biology of Moniliophthora perniciosa on Theobroma cacao and Alternate Solanaceous Hosts. Tropical Plant Biology 2, 149-160. doi: 10.1007/s12042-009-9038-1. 78 Marquina, D., Santos, A., and Peinado, J. M. (2002). Biology of killer yeasts. International Microbiology 5, 65-71. doi: 10.1007/s10123-002-0066-z. Medeiros, F. H. V., Pomella, A. W. V., de Souza, J. T., Niella, G. R., Valle, R., Bateman, R. P., et al. (2010). A novel, integrated method for management of witches' broom disease in Cacao in Bahia, Brazil. Crop Protection 29(7), 704-711. doi: 10.1016/j.cropro.2010.02.006. Mehmood, A., Liu, G., Wang, X., Meng, G., Wang, C., and Liu, Y. (2019). Fungal Quorum-Sensing Molecules and Inhibitors with Potential Antifungal Activity: A Review. Molecules 24, 1950. doi: 10.3390/molecules24101950. Meinhardt, L. W., Rincones, J., Bailey, B. A., Aime, M. C., Griffith, G. W., Zhang, D., et al. (2008). Moniliophthora perniciosa , the causal agent of witches’ broom disease of cacao: what’s new from this old foe? Molecular Plant Pathology 9(5), 577-588. doi: 10.1111/j.13643703.2008.00496.x. Nunes, C. A. (2012). Biological control of postharvest diseases of fruit. European Journal of Plant Pathology 133, 181-196. doi: 10.1007/s10658-011-9919-7. Oro, L., Feliziani, E., Ciani, M., Romanazzi, G., and Comitini, F. (2018). Volatile organic compounds from Wickerhamomyces anomalus , Metschnikowia pulcherrima and Saccharomyces cerevisiae inhibit growth of decay causing fungi and control postharvest diseases of strawberries. International Journal of Food Microbiology 265, 18-22. doi: 10.1016/j.ijfoodmicro.2017.10.027. Pal, K. K., and McSpadden Gardener, B. (2006). Biological Control of Plant Pathogens. The Plant Health Instructor . Online. doi: 10.1094/PHI-A-2006-1117-02. Parafati, L., Vitale, A., Restuccia, C., and Cirvilleri, G. (2015). Biocontrol ability and action mechanism of food-isolated yeast strains against Botrytis cinerea causing post-harvest bunch rot of table grape. Food Microbiology 47, 85-92. doi: 10.1016/j.fm.2014.11.013. Platania, C., Restuccia, C., Muccilli, S., and Cirvilleri, G. (2012). Efficacy of killer yeasts in the biological control of Penicillium digitatum on Tarocco orange fruits (Citrus sinensis). Food Microbiology 30, 219-225. doi: 10.1016/j.fm.2011.12.010. Pohlan, H. A. J., and Pérez, V. D. (2010). "Growth and Production of Cacao," in Soils, Plant Growth and Crop Production - Volume III - Encyclopedia of Life Support Systems (EOLSS) , ed. W. H. Verheye. (Oxford, UK: EOLSS Publishers). 79 Purdy, L. H., and Schmidt, R. A. (1996). STATUS OF CACAO WITCHES’ BROOM: Biology, Epidemiology, and Management. Annual Review of Phytopathology 34, 573-594. doi: 10.1146/annurev.phyto.34.1.573. Rodríguez-Cousiño, N., Maqueda, M., Ambrona, J., Zamora, E., Esteban, R., and Ramírez, M. (2011). A New Wine Saccharomyces cerevisiae Killer Toxin (Klus), Encoded by a Double-Stranded RNA Virus, with Broad Antifungal Activity Is Evolutionarily Related to a Chromosomal Host Gene. Applied and Environmental Microbiology 77(5), 1822-1832. doi: 10.1128/AEM.02501-10. Rosa-Magri, M. M., Tauk-Tornisielo, S. M., and Ceccato-Antonini, S. R. (2011). Bioprospection of Yeasts as Biocontrol Agents Against Phytopathogenic Molds. Brazilian Archives of Biology and Technology 54(1), 1-5. doi: 10.1590/S1516-89132011000100001´. Saravanakumar, D., Ciavorella, A., Spadaro, D., Garibaldi, A., and Gullino, M. L. (2008). Metschnikowia pulcherrima strain MACH1 outcompetes Botrytis cinerea , Alternaria alternata and Penicillium expansum in apples through iron depletion. Postharvest Biology and Technology 49, 121-128. doi: 10.1016/j.postharvbio.2007.11.006. Schmitt, M. J., and Breinig, F. (2006). Yeast viral killer toxins: lethality and self-protection. Nature Reviews Microbiology 4(3), 212-221. doi: 10.1038/nrmicro1347. Schulz-Bohm, K., Martín-Sánchez, L., and Garbeva, P. (2017). Microbial Volatiles: Small Molecules with an Important Role in Intraand Inter-Kingdom Interactions. Frontiers in Microbiology 8, 2484. doi: 10.3389/fmicb.2017.02484. Spadaro, D., and Droby, S. (2016). Development of biocontrol products for postharvest diseases of fruit: the importance of elucidating the mechanisms of action of yeast antagonists. Trends in Food Science & Technology 47, 39-49. doi: 10.1016/j.tifs.2015.11.003. Suzzi, G., Romano, P., Ponti, I., and Montuschi, C. (1995). Natural wine yeasts as biocontrol agents. Journal of Applied Bacteriology 78, 304-308. doi: 10.1111/j.1365-2672.1995.tb05030.x. Teixeira, P. J. P. L., Thomazella, D. P. d. T., and Pereira, G. A. G. (2015). Time for Chocolate: Current Understanding and New Perspectives on Cacao Witches’ Broom Disease Research. PLOS Pathogens 11(10), e1005130. doi: 10.1371/journal.ppat.1005130. Trevizan, S. D. P., and Marques, M. (2002). Social and economic impacts of the cacao crisis: study of a community case (in Portuguese). Agrotrópica (Brazil) 14, 127-136. 80 Verweij, P. E., Snelders, E., Kema, G. H., Mellado, E., and Melchers, W. J. (2009). Azole resistance in Aspergillus fumigatus: a side-effect of environmental fungicide use? The Lancet Infectious Diseases 9, 789-795. doi: 10.1016/S1473-3099(09)70265-8. Walker, G. M., McLeod, A. H., and Hodgson, V. J. (1995). Interactions between killer yeasts and pathogenic fungi. FEMS Microbiology Letters 127, 213-222. doi: 10.1111/j.15746968.1995.tb07476.x. Wang, X., Chi, Z., Yue, L., and Li, J. (2007). Purification and characterization of killer toxin from a marine yeast Pichia anomala YF07b against the pathogenic yeast in crab. Current Microbiology 55, 396401. doi: 10.1007/s00284-007-9010-y. Wuster, A., and Babu, M. M. (2007). "Chemical molecules that regulate transcription and facilitate cellto-cell communication," in Wiley Encyclopedia of Chemical Biology , ed. T. P. Begley. John Wiley & Sons, Chichester), 1-11. Zhang, D., Spadaro, D., Garibaldi, A., and Gullino, M. L. (2010). Selection and evaluation of new antagonists for their efficacy against postharvest brown rot of peaches. Postharvest Biology and Technology 55, 174-181. doi: 10.1016/j.postharvbio.2009.09.007. Zhang, D., Spadaro, D., Garibaldi, A., and Gullino, M. L. (2011). Potential biocontrol activity of a strain of Pichia guilliermondii against grey mold of apples and its possible modes of action. Biological Control 57(3), 193-201. doi: 10.1016/j.biocontrol.2011.02.011. CHAPTER 3 Microscopic assessment of the antagonism effect of yeasts against Moniliophthora perniciosa, the fungal causal agent of Witches’ Broom Disease in cacao The work presented in this chapter is in preparation for publication: Ferraz, P., Cássio, F. and Lucas, C. Microscopic assessment of the antagonism effect of yeasts against Moniliophthora perniciosa , the fungal causal agent of Witches’ Broom Disease in cacao 82 ABSTRACT Mycoparasitism of pathogenic fungi by yeasts could be associated with direct physical contact. Predacious yeasts can penetrate the fungal cell walls killing the fungal pathogen and/or feed on them, taking up nutrients from the fungal cells. Contact between yeasts and fungal phytopathogens was observed in antagonism yeast strains against Moniliophthora perniciosa , the fungal causal agent of the Witches’ Broom Disease of cacao (Ferraz et al ., submitted 2). Cell analysis by Scanning Electron Microscopy (SEM) revealed yeast binding and fusing with the hyphae of the phytopathogen, causing a constriction and deformation of its cells, and even a draining of the fungal cellular content. The formation of veil-like structures possibly associated with invasion of hyphae by yeasts was also observed. Moreover, SEM micrographs showed the occurrence of fimbriae-like connections between antagonistic yeast cells and of physical tube-like connection between yeast and fungal cells, which can be related with yeast cell-cell communication and to yeast/fungus recognition or even yeast predacious behavior, respectively. These results constitute the first description of a possible predacious-like behavior in nonSaccharomycopsis yeasts, which can be the basis of the antagonism exerted by yeasts against the fungal pathogens that cause WBD of cacao. Keywords SEM, Cacao; Witches Broom Disease ; Moniliophthora perniciosa ; yeast; antagonism; predation 83 INTRODUCTION Yeasts can antagonize other microorganisms such as other yeasts, molds, and bacteria (Hatoum et al. , 2012). The antagonistic effect of yeasts is based on several mechanisms of action, including the competition for nutrients and space (Parafati et al. , 2015) and/or the production and secretion of killer toxins (Schmitt and Breinig, 2006) and other antifungal compounds such as hydrolytic enzymes (Lopes et al. , 2015; Liu et al. , 2018), volatile compounds (Schulz-Bohm et al. , 2017) and quorum sensing molecules (Avbelj et al. , 2016). In all these mechanisms, yeasts are able to secrete compounds and exert their antagonistic effect at a distance. Despite this, some yeasts have the ability to attach to the hyphae of fungal pathogens and produce diverse extracellular cell wall lytic enzymes. This type of interaction is known as mycoparasitism (Dukare et al. , 2018). Some sequential events are required so that a microorganism can parasitize a fungus. First there must be close contact between the antagonist and the fungal pathogen cells followed by mutual recognition. Subsequently the antagonizing microorganism secretes lytic enzymes and ultimately, there is active growth of the antagonist into the fungal host cells through a kind of penetration peg (Spadaro and Gullino, 2004). The fungal cell wall is often targeted. The fungal cell wall is composed of polysaccharides, glucans and chitin, and glycoproteins, which together provide mechanical strength and structural integrity (Spadaro and Droby, 2016). The antagonist secretes hydrolytic enzymes, including chitinases, chitosanases, glucanases, cellulases or proteases which may act independently or in combination with each other or other enzymes (Spadaro and Droby, 2016). Ultimately, the destruction of cell wall leads to severe cytological damage, in most circumstances causing the lysis of the cell and subsequently death (Di Francesco et al. , 2015; Dukare et al. , 2018). Some examples of mycoparasitism and its association with the biocontrol of fungal phytopathogens have been described in the literature, particularly the mycoparasitism exerted by yeasts. This is the case of the antagonism of Botrytis cinerea by the yeast Meyerozyma guilliermondii (formerly known as Pichia guilliermondii ), in which the yeast cell attaches strongly to the host and secretes a β-(1–3) glucanase, which results in the lysis of the fungal cell (Wisniewski et al. , 1991). Is has also been reported that, this type of lytic enzymes also inhibits the germination of the pathogen spores and the elongation of its germtube (El-Tarabily and Sivasithamparam, 2006). Another example of mycoparasitism is the biocontrol of Penicillium expansum by the yeast Candida oleophila in harvested apples, associated with the secretion of an exo-β-1,3-glucanase, which results in the inhibition of conidial germination and mycelia growth (Tamayo-Urbina et al. , 2016). 84 The mycoparasites can be classified into different groups, according to their modes of action. The necrotrophic mycoparasites are very aggressive organisms with a broad range of preys, and can attack fungi at a distance by the secretion of toxins or lytic enzymes into the environment (Junker et al. , 2019). Contact necrotrophs attack the fungal cells by direct physical contact and can reach and kill the fungal pathogen using hyphae, although they do not necessarily invade the target cell (Mims et al. , 2007). On the other hand, invasive necrotrophs, also known as predatory mycoparasites, physically penetrate their prey fungal cells through haustoria or penetration pegs (Junker et al. , 2019). The predatory organism can kill the fungal prey, or simply take up nutrients from its cells, or both (Jeffries, 1995; Junker et al. , 2019). The majority of the well-studied mycoparasites are filamentous fungi, particularly species from the Trichoderma genus (Schmoll et al. , 2016). Although many yeast species are able to secrete antagonistic peptides such as killer toxins (Schmitt and Breinig, 2006), or lytic enzymes (Spadaro and Gullino, 2004), the only yeast species described as a necrotrophic mycoparasite belong to the Saccharomycopsis clade (Lachance and Pang, 1997). These yeasts are able to predate a wide range of other yeasts. All share a metabolic peculiarity, the inability to use sulfate as sole sulfur source. Despite not being well studied, their mode of predatory action involves the invasion of the prey cells, mediated by small haustoria-like penetration pegs, leading the subsequent death of the prey cells (Lachance et al. , 2012). S. schoenii is considered the most aggressive predacious yeast in this clade, and it was reported to attack and kill in vitro several clinical isolates of pathogenic Candida species, including the multi-drug resistant isolates of C. auris , thus having a great potential in medical biocontrol applications (Junker et al. , 2018). Additionally, some Saccharomycopsis species, have been described as successful potential biocontrol agents, namely S. schoenii , against plant pathogens developing in the surface of oranges (Pimenta et al. , 2008), and S. fibuligera against toxic molds on pork speck (Iacumin et al. , 2017). The most favorable environmental conditions that promotes predation varies significantly, depending on the predacious species. Some species are stimulated by the presence of rich nitrogenous nutrients and high concentrations of ammonium nitrogen, as well as by the presence of organic sulfur compounds, while other species may be inhibited under the same conditions (Lachance et al. , 2000). Moreover, the predacious yeasts are able to penetrate the cells of other described predacious species, and interestingly young cultures can penetrate older cultures of the same predacious strain (Lachance et al. , 2000). This is a very poorly studied phenomenon which has not been described for other yeast genera and species, and which predation modes of action have not been further studied. 85 In a recent study the causal agent of cacao Witches’ Broom Disease (WBD), the filamentous fungus Moniliophthora perniciosa , was shown to be killed by yeast isolates from spontaneous fermentations of cachaça production and strains used in bioethanol industrial processes (Ferraz et al. , submitted 2). Death of the fungal cell occurred after the co-incubation of both organisms and detected by staining with methylene blue and propidium iodide. The work hereby presented aimed at exploring microscopically the interaction between the two types of cells, unveiling that the yeasts physically attach to the hyphae, apparently fusing their walls surface, the two cells becoming as one, and that the hyphae are drained. Moreover, the study also showed that yeasts communicate physically between each other during the process. Results were obtained observing samples stained with methylene blue, as well as using Scanning Electron Microscopy (SEM). Further assessment by staining the cytoskeleton with a specific fluorescent dye, and the use of Transmission Electron Microscopy (TEM), will allow in the future the elucidation of the cytology of the actual merge between the two cells. 92 Figure 5. SEM micrographs of the yeast/fungal strains combinations #1105 vs CBS 441.80 and PE2 vs CBS 441.80, showing in close detail different morphological structures of the interaction: yeast pushing the hyphae and fusion between the cells of both organisms (white arrows); connections between the yeast and the fungal cells (blue arrows); and formation of a veil-like structure (red arrows). 93 1 µm 0.5 µm Figure 6. SEM micrographs of the connection between yeast cells. Left panels - Details of two yeast cells that apparently were connected and their small fimbriae-like structures on the cell surface. Right panels - Fimbriae-like structures in the surface of the yeast cells (blue arrows) and locations of previous connection structures (yellow arrows). 94 Fimbriae and pili have abundantly been described in bacteria (Carter et al. , 2016) as hair-like structures protruding from bacteria cell wall. They are involved in adherence to inert surfaces or living tissues, in cell-cell communication, as well as in the conjugational transfer of DNA between two cells (reviewed by Berne et al., 2018). In the case of yeasts, connection fibrils between two cells from the same species and strain have been reported in biofilms (Mamvura et al. , 2017) or in starved cell (Varon and Choder, 2000). To these fibrils have been attributed a role in the colony formation of some Candida species (Vargas et al. , 2004; Furlaneto et al. , 2012). Information is though very scarce, and there are no detailed studies on which yeasts are able or not to form them, on how or when they are formed, or on their structure and composition. 95 CONCLUSIONS The present study allowed the visualization of (i) the antagonizing yeast cell binding and fusing to the dead hyphae of the phytopathogen, (ii) the constriction and deformation of the fungal cells caused by yeast attachment, (iii) the draining of the fungal cellular content in areas adjacent to the yeast attachment. Moreover, the analysis by SEM revealed (i) the formation of a veil-like structure that could be related with a possible invasion of hyphae by yeasts, (ii) the occurrence of a physical tube-like connection between yeast and fungal cells which can be related to yeast/fungus recognition or even associated with yeast predacious behavior, and (iii) the existence of fimbriae-like connections between antagonistic yeast cells, which can be associated with a mechanism of yeast cell-cell communication. Although neither of the antagonistic yeasts tested belong to the Saccharomycopsis clade, these characteristics resemble those of the predacious yeasts. This work presents the first description of a possible predacious-like behavior in nonSaccharomycopsis yeasts and might provide a basis to understand why the industrial yeasts are so resilient to harsh environments and prevail over other strains in their natural/industrial habitat. Nevertheless, a predacious behavior may not be independent of the secretion of some antifungal compounds by these yeasts, which may act synergistically, causing the antagonism effect and leading to the pathogen’s death. Future analysis by TEM and fluorescence microscopy can improve the understanding of this antagonism and clarify these yeasts mode of action. ACKNOWLEDGEMENTS This work was supported by the strategic programme UID/BIA/04050/2013 (POCI-01-0145-FEDER-007569) funded by national funds through the FCT I.P. and by the ERDF through the COMPETE2020 - Programa Operacional Competitividade e Internacionalizacão (POCI), and the project EcoAgriFood (NORTE-01-0145-FEDER-000009) from Norte Portugal Regional Operational Programme (NORTE 2020) under the PORTUGAL 2020 Partnership Agreement through the European Regional Development Fund (ERDF). PF is a PhD student of the Doctoral Programme in Applied and Environmental Microbiology (DP-AEM), (FCT grant PD/BD/113810/2015). We thank the Brazilian companies, Cerlev, Lda., Ouro Preto, MG, and Fermentec, Lda. Soluções Tecnológicas e Industriais, Piracicaba, SP for kindly supplying the yeast strains used in this work. 96 REFERENCES Avbelj, M., Zupan, J., and Raspor, P. (2016). Quorum-sensing in yeast and its potential in wine making. Applied Microbiology and Biotechnology 100, 7841-7852. doi: 10.1007/s00253-016-7758-3. Berne, C., Ellison, C. K., Ducret, A., and Brun, Y. V. (2018). Bacterial adhesion at the single-cell level. Nature Reviews Microbiology 16(10), 616-627. doi: 10.1038/s41579-018-0057-5. Carter, M. Q., Louie, J. W., Feng, D., Zhong, W., and Brandl, M. T. (2016). Curli fimbriae are conditionally required in Escherichia coli O157:H7 for initial attachment and biofilm formation. Food Microbiology 57, 81-89. doi: 10.1016/j.fm.2016.01.006. da Conceição, L. E., Saraiva, M. A., Diniz, R. H., Oliveira, J., Barbosa, G. D., Alvarez, F., et al. (2015). Biotechnological potential of yeast isolates from cachaça: the Brazilian spirit. Journal of Industrial Microbiology & Biotechnology 42(2), 237-246. doi: 10.1007/s10295-014-1528-y. Di Francesco, A., Ugolini, L., Lazzeri, L., and Mari, M. (2015). Production of volatile organic compounds by Aureobasidium pullulans as a potential mechanism of action against postharvest fruit pathogens. Biological Control 81, 8-14. doi: 10.1016/j.biocontrol.2014.10.004. Dukare, A. S., Paul, S., Nambi, V. E., Gupta, R. K., Singh, R., Sharma, K., et al. (2018). Exploitation of microbial antagonists for the control of postharvest diseases of fruits: a review. Critical Reviews in Food Science and Nutrition 16, 1-16. doi: 10.1080/10408398.2017.1417235. El-Tarabily, K. A., and Sivasithamparam, K. (2006). Potential of yeasts as biocontrol agents of soil-borne fungal plant pathogens and as plant growth promoters. Mycoscience 47(1), 25-35. doi: 10.1007/S10267-005-0268-2. Ferraz, P., Amorim-Rodrigues, M., Cássio, F., and Lucas, C. Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease. Submitted 2. Ferraz, P., Amorim-Rodrigues, M., Cássio, F. and Lucas, C. Vinasse waste from sugarcane-based bioethanol production plants kills Moliniophthora perniciosa , the causative agent of cacao Witches’ Broom Disease . Submitted 1. 97 Fullner, K. J., Lara, J. C., and Nester, E. W. (1996). Pilus Assembly by Agrobacterium T-DNA Transfer Genes. Science 273(5278), 1107-1109. doi: 10.1126/science.273.5278.1107. Furlaneto, M. C., Andrade, C. G. T. J., Furlaneto-Maia, L., de França, E. J. G., and Moralez, A. T. P. (2012). "Pathogenic Attributes of NonCandida albicans Candida Species Revealed by SEM," in Scanning Electron Microscopy , ed. V. Kazmiruk. IntechOpen). Hatoum, R., Labrie, S., and Fliss, I. (2012). Antimicrobial and probiotic properties of yeasts: From fundamental to novel applications. Frontiers in Microbiology 3, 421. doi: 10.3389/fmicb.2012.00421. Higuchi-Sanabria, R., Swayne, T. C., Boldogh, I. R., and Pon, L. A. (2016). "Imaging of the Actin Cytoskeleton and Mitochondria in Fixed Budding Yeast Cells," in Cytoskeleton Methods and Protocols: Methods and Protocols , ed. R. H. Gavin. (New York, NY: Springer New York), 63-81. Iacumin, L., Manzano, M., Andyanto, D., and Comi, G. (2017). Biocontrol of ochratoxigenic moulds ( Aspergillus ochraceus and Penicillium nordicum ) by Debaryomyces hansenii and Saccharomycopsis fibuligera during speck production. Food Microbiology 62, 188-195. doi: 10.1016/j.fm.2016.10.017. Jeffries, P. (1995). Biology and ecology of mycoparasitism. Canadian Journal of Botany 73(S1), 12841290. doi: 10.1139/b95-389. Junker, K., Bravo Ruiz, G., Lorenz, A., Walker, L., Gow, N. A. R., and Wendland, J. (2018). The mycoparasitic yeast Saccharomycopsis schoenii predates and kills multi-drug resistant Candida auris. Scientific Reports 8(1), 14959. doi: 10.1038/s41598-018-33199-z. Junker, K., Chailyan, A., Hesselbart, A., Forster, J., and Wendland, J. (2019). Multi-omics characterization of the necrotrophic mycoparasite Saccharomycopsis schoenii . PLOS Pathogens 15(5), e1007692. doi: 10.1371/journal.ppat.1007692. Lachance, M.-A., and Pang, W.-M. (1997). Predacious Yeasts. Yeast 13(3), 225-232. doi: 10.1002/(sici)1097-0061(19970315)13:3<225::aid-yea87>3.0.co;2-i. 98 Lachance, M.-A., Pupovac-Velikonja, A., Natarajan, S., and Schlag-Edler, B. (2000). Nutrition and phylogeny of predacious yeasts. Canadian Journal of Microbiology 46(6), 495-505. doi: 10.1139/w00-021. Lachance, M.-A., Rosa, C. A., Carvajal, E. J., Freitas, L. F. D., and Bowles, J. M. (2012). Saccharomycopsis fodiens sp. nov., a rare predacious yeast from three distant localities. International Journal of Systematic and Evolutionary Microbiology 62(11), 2793-2798. doi: 10.1099/ijs.0.043109-0. Lai, E.-M., and Kado, C. I. (1998). Processed VirB2 Is the Major Subunit of the Promiscuous Pilus of Agrobacterium tumefaciens . Journal of Bacteriology 180(10), 2711-2717. Liu, Z., Du, S., Ren, Y., and Liu, Y. (2018). Biocontrol ability of killer yeasts ( Saccharomyces cerevisiae ) isolated from wine against Colletotrichum gloeosporioides on grape. Journal of Basic Microbiology 58, 60-67. doi: 10.1002/jobm.201700264. Lopes, M. L., Paulillo, S. C., Godoy, A., Cherubin, R. A., Lorenzi, M. S., Giometti, F. H., et al. (2016). Ethanol production in Brazil: a bridge between science and industry. Brazilian Journal of Microbiology 47(1), 64-76. doi: 10.1016/j.bjm.2016.10.003. Lopes, M. R., Klein, M. N., Ferraz, L. P., da Silva, A. C., and Kupper, K. C. (2015). Saccharomyces cerevisiae : A novel and efficient biological control agent for Colletotrichum acutatum during preharvest. Microbiological Research 175, 93-99. doi: 10.1016/j.micres.2015.04.003. Mamvura, T. A., Paterson, A. E., and Fanucchi, D. (2017). The impact of pipe geometry variations on hygiene and success of orbital welding of brewing industry equipment. Journal of the Institute of Brewing 123(1), 81-97. doi: 10.1002/jib.398. Mims, C. W., Hanlin, R. T., and Richardson, E. A. (2007). Lightand electron-microscopic observations of Cladosporium sp. growing on basidia of Exobasidum camelliae var. gracilis . Canadian Journal of Botany 85(1), 76-82. doi: 10.1139/b06-153. Murtey, M. D., and Ramasamy , P. (2016). "Sample Preparations for Scanning Electron Microscopy – Life Sciences," in Modern Electron Microscopy in Physical and Life Sciences , ed. M. J. a. R. Kral. IntechOpen). 99 Parafati, L., Vitale, A., Restuccia, C., and Cirvilleri, G. (2015). Biocontrol ability and action mechanism of food-isolated yeast strains against Botrytis cinerea causing post-harvest bunch rot of table grape. Food Microbiology 47, 85-92. doi: 10.1016/j.fm.2014.11.013. Pimenta, R. S., Silva, F. L., Silva, J. F. M., Morais, P. B., Braga, D. T., Rosa, C. A., et al. (2008). Biological control of Penicillium italicum , P. digitatum and P. expansum by the predacious yeast Saccharomycopsis schoenii on oranges. Brazilian Journal of Microbiology 39, 85-90. doi: 10.1590/S1517-83822008000100020 Schmitt, M. J., and Breinig, F. (2006). Yeast viral killer toxins: lethality and self-protection. Nature Reviews Microbiology 4(3), 212-221. doi: 10.1038/nrmicro1347. Schmoll, M., Dattenböck, C., Carreras-Villaseñor, N., Mendoza-Mendoza, A., Tisch, D., Alemán, M. I., et al. (2016). The Genomes of Three Uneven Siblings: Footprints of the Lifestyles of Three Trichoderma Species. Microbiology and Molecular Biology Reviews 80(1), 205. doi: 10.1128/MMBR.00040-15. Schulz-Bohm, K., Martín-Sánchez, L., and Garbeva, P. (2017). Microbial Volatiles: Small Molecules with an Important Role in Intraand Inter-Kingdom Interactions. Frontiers in Microbiology 8, 2484. doi: 10.3389/fmicb.2017.02484. Spadaro, D., and Droby, S. (2016). Development of biocontrol products for postharvest diseases of fruit: the importance of elucidating the mechanisms of action of yeast antagonists. Trends in Food Science & Technology 47, 39-49. doi: 10.1016/j.tifs.2015.11.003. Spadaro, D., and Gullino, M. L. (2004). State of the art and future prospects of the biological control of postharvest fruit diseases. International Journal of Food Microbiology 91, 185-194. doi: 10.1016/S0168-1605(03)00380-5. Tamayo-Urbina, C., Guerrero-Prieto, V., Guigon-Lopez, C., Vargas-Albores, F., Berlanga-Reyes, D., AcostaMuniz, C., et al. (2016). Purification and Characterization of β-1,3-glucanase from Candida Oleophila for the Biocontrol of Penicillium expansum . Research & Reviews: Journal of Botanical Sciences 5(1), 38-45. 100 Vargas, K. G., Heying, J., Morris, R., and Srikantha, R. (2004). Switch Phenotypes of Candida albicans Differ in the Oral Colonization of Immunosuppressed Balb/c Mice. Oral Biosciences & Medicine 1, 267-275. Varon, M., and Choder, M. (2000). Organization and Cell-Cell Interaction in Starved Saccharomyces cerevisiae Colonies. Journal of Bacteriology 182(13), 3877-3880. doi: 10.1128/jb.182.13.38773880.2000. Wisniewski, M., Biles, C., Droby, S., McLaughlin, R., Wilson, C., and Chalutz, E. (1991). Mode of action of the postharvest biocontrol yeast, Pichia guilliermondii . I. Characterization of attachment to Botrytis cinerea . Physiological and Molecular Plant Pathology 39, 245-258. doi: 10.1016/08855765(91)90033-E. CHAPTER 4 Wickerhamomyces anomalus from olive orchards microbiome efficiently antagonizes the Olive Anthracnose’s causal agents, Colletotrichum spp. The work presented in this chapter is in preparation for publication: Amorim-Rodrigues, M., Ferraz, P., Cássio, F. and Lucas, C. Wickerhamomyces anomalus from olive orchards microbiome efficiently antagonizes the Olive Anthracnose’s causal agents, Colletotrichum spp. 108 RESULTS AND DISCUSSION Optimization of fungal cultivation conditions Yeast strains used in the present work (Table 1) can be divided in two groups. The first contains strains which were isolated from the olive biome of Portuguese orchards. The second group includes strains isolated from sugarcane-based spontaneous and industrial fermentation processes. The fungal strains that were used have different origins. The strain of C. gloeosporioides s.s. was originally isolated from infected orchards in Italy. The strains of C. godetiae and C. nymphaeae originate from infected olive trees in Portugal. Prior to the antagonism assays, the in vitro optimal culture conditions of fungal and yeasts strains were determined. Firstly, the fungal growth rates were assessed in MEA and PDA, at different pH (4.0, 4.5, 5.0, 5.5 and 6.0) and 30 °C. Only C. gloeosporioides s.s. was able to grow at this temperature. Therefore, assays were repeated at 25°C. In agreement with the literature (Wharton and DiéguezUribeondo, 2004; Talhinhas et al. , 2005), C. godetiae and C. nymphaeae (from the C. acutatum species complex) grew much slower than C. gloeosporioides s.s. (Table S1). Although this last actually grew better at 30 °C than at 25 °C, the difference was not significant (not shown). The selected temperature for further assaying all the fungi was therefore defined as 25 °C. Furthermore, differences in growth rates between assays in MEA or PDA (not shown) and at different pH (Table S1) were not significant. Based on these results, the optimal conditions were established for all the fungal strains as MEA or PDA, and pH 5.5 at 25 ºC. From the same assays, it was also shown that 4 to 8 days is enough to assess fungal growth. Antagonism in solid media Antagonism was tested using the yeasts and the phytopathogenic fungal strains in Table 1. Results were captured after 8 days of incubation (Table 2) and expressed according to the empirical scale of levels 0, 1 and 2 described in Ferraz et al. (submitted). A more detailed assessment of antagonism followed, determining the percentage of fungal growth inhibition (Table 3). As can be seen in these Tables, all the yeasts were able to inhibit the growth of at least one of the fungal strains, in most cases causing a low inhibitory response (level 1). Accordingly, some yeasts behaved generally as weak antagonists, like the strains #1 and PE2. In opposition, the yeast strain causing the strongest inhibitory response (level 2) was #18, which inhibited the growth of the three fungal strains in more than 70 % (Table 3). 109 Table 2. Results of antagonism assays in solid medium, using the yeast strains from Table 1 against Colletotrichum gloeosporioides s.s. (C1) , C. godetiae (C2) and C. nymphaeae (C3) . Assays were performed in MEA at 25 °C. The numbers displayed translate results according to the empirical scale described in Ferraz et al. (submitted). Single numbers represent the inhibitory response of 3 replicates. A table cell displaying 3 numbers (e.g. 1/0/0) represents a situation where each plate had a different result. Phytopathogens Yeasts #1 #2 #3 #4 #5 #6 #7 CAT1 PE2 #11 #12 C1 0 1/0/0 1/0/0 0 2/1/1 1 0 1/1/0 0 0 1/1/0 C2 1 1/1/0 1/1/0 1/1/0 1 2/1/0 1 1 1/0/0 1 1 C3 1 1/0/0 1/0/0 1/1/0 1/1/0 1/1/0 1 1/0/0 0 1/1/1 1/1/0 #13 #14 #15 #16 #17 #18 #19 #20 #21 #22 C1 1/0/0 1/1/0 1/0/0 1/0/0 1/1/0 2/2/1 1/1/0 1 1/1/0 1 C2 1 2/1/0 1 1/1/0 1 2 1 1 1 1 C3 1 1/0/0 1/1/0 1/1/0 1 2 1 1/1/0 1 1/1/0 The inhibition of the fungal growth was observed to occur at a certain distance as exemplified in Figure 1A, suggesting that yeasts possibly secrete a compound that affects the fungal multiplication. There are two types of possible compounds: (i) soluble, diffusing through the agar, such as killer toxins and lytic enzymes (Parafati et al. , 2015; Oro et al. , 2018); and (ii) volatile (Toffano et al. , 2017). To assess which might be the case, the antagonism on solid media was repeated in MEA using septate Petri dishes, equidistantly inoculated with the mycelial plug on one side of the septum, and the yeast streak on the other. This only allows the fungus and the yeast to share the atmosphere inside the dish. No inhibitory response was observed up to 2 weeks of incubation, as exemplified in Figure 1B. Moreover, the fungal strains, except C. godetiae , went over the septum , eventually growing on top of the yeast and covering the whole dish. This result strongly suggested that the inhibition of fungal growth by this yeast requires the diffusion of a molecule through the agar. Accordingly, Lima and collaborators (2013) found that when W. anomalus and M. guilliermondii cultures were killed by autoclaving and filtered, they lost the ability to inhibit the spore germination of C. gloeosporioides s.s., which concurs with the need for yeast to be metabolically active abovementioned. After 2 weeks of incubation, most of the yeast population is most probably dying or dead, consistently with the MB coloration of the yeast streak (not shown). 110 Table 3. Antagonism between the OA causative agents Colletotrichum gloeosporioides s.s. (C1) , C. godetiae (C2) and C. nymphaeae (C3) and the yeast strains in Table 1. Results are percentage of inhibition in solid media, according to Royse & Ries (1978). Values are average across replicates N≥3 and standard deviation. Shaded cells show highest inhibition of all fungi. Phytopathogens Yeasts #1 #2 #3 #4 #5 #6 #7 CAT1 PE2 #11 #12 C1 3.2 ± 5.5 17.5 ± 7.3 19.1 ± 4.8 25.4 ± 15.3 70.4 ± 1.4 46.0 ± 22.5 23.8 ± 4.8 57.1 ± 24.7 19.1 ± 4.8 0.0 ± 0.0 28.6 ± 31.2 C2 40.9 ± 7.9 22.7 ± 25.3 33.3 ± 26.6 25.8 ± 9.5 28.8 ± 9.5 21.2 ± 11.4 59.1 ± 4.6 10.6 ± 14.6 9.1 ± 4.6 43.2 ± 18.4 46.1 ± 17.9 C3 25.9 ± 3.6 33.3 ± 24.7 47.2 ± 17.5 54.7 ± 2.1 62.9 ± 7.6 51.2 ± 22.8 25.3 ± 17.1 37.9 ± 20.3 14.6 ± 12.4 30.2 ± 2.8 55.7 ± 3.1 #13 #14 #15 #16 #17 #18 #19 #20 #21 #22 C1 22.2 ± 22.5 31.8 ± 28.7 11.1 ± 7.3 14.3 ± 12.6 33.3 ± 35.9 70.8 ± 7.2 22.2 ± 5.5 28.6 ± 0.0 25.4 ± 24.0 28.6 ± 4.8 C2 38.1 ± 5.7 32.1 ± 27.8 33.3 ± 11.6 48.2 ± 5.6 55.2 ± 18.9 70.5 ± 0.8 41.9 ± 7.3 44.1 ± 10.6 52.2 ± 10.6 55.9 ± 16.0 C3 56.3 ± 5.5 65.9 ± 3.0 54.0 ± 9.7 43.6 ± 19.4 52.8 ± 14.1 74.5 ± 4.3 49.1 ± 4.3 61.3 ± 5.2 62.2 ± 7.9 65.4 ± 3.8 111 Figure 1. Yeast strain #18 against Colletotrichum gloeosporioides s.s. strain. Inhibitory response at a distance (A) and none in a septate petri dish (B). Antagonism in liquid media The antagonism assays were subsequently repeated in liquid medium. If a metabolite of some kind is produced and secreted by the yeasts that inhibits the fungal growth, its diffusion should be much more efficient in liquid medium and therefore might cause stronger inhibition. Moreover, the production of some antifungal metabolites is greater at the exponential growth (Lopes et al. , 2015), which is only possible in liquid cultures. Moreover, these also allow a greater nutrient availability promoted by the constant agitation, according to which yeasts should stay metabolically active for longer periods of time. Finally, the efficient antagonism in some cases can rely on physical contact between the inhibitor and the inhibited strains, as previously demonstrated with another phytopathogen (Ferraz et al. , submitted). no yeast #18 no yeast #18 A B 112 The three OA fungal strains were challenged with a metabolically active yeast inoculum and incubated for 4 days. Results of the antagonistic response using empirical scale with 0, 1, 2 levels of inhibition previously described (Ferraz et al. , submitted) are shown in Table 4, and represent the types of inhibition exemplified in Figure 2. Table 4. Antagonism between the phytopathogenic fungal strains Colletotrichum gloeosporioides s.s. (C1), C. godetiae (C2) and C. nymphaeae (C3) and the yeast strains in Table 1 in liquid media. The numbers displayed translate results according to a previously established empirical scale (Ferraz et al. , submitted; 0 = no inhibition; 1 = weak inhibition;2 = full inhibition) and represent at least three independent replicates with identical result. Phytopathogens Yeasts #1 #2 #3 #4 #5 #6 #7 CAT1 PE2 #11 #12 C1 2 0 0 0 2 2 2 2 2 0 1 C2 2 1 0 0 2 2 2 2 2 0 1 C3 2 0 0 0 2 2 2 2 2 0 1 #13 #14 #15 #16 #17 #18 #19 #20 #21 #22 C1 2 1 2 0 1 2 2 2 2 2 C2 2 0 2 1 2 2 2 2 2 2 C3 2 1 2 0 1 2 2 2 2 2 In general, as expected, the inhibitory response increased comparing to that obtained in solid media. In solid media, a strong inhibitory response (level 2 in Table 2; >70% in Table 3) was only observed in 6 out of 63 combinations, while in liquid medium, the fungi were strongly inhibited (level 2) in 39 out of 63 combinations. Results in Table 4 also show that 3 yeasts did not inhibit the growth of any of the fungal strains: #3, #4, and #11, and the #2 and #16 only weakly inhibited one of them. Based on these results, these yeasts were not further considered. 113 Figure 2. Liquid media antagonism assay against Colletotrichum gloeosporioides s.s. strain. Examples of the type of inhibitory response. Exploratory assays of confirmation of the death of the fungal cells were conducted. The fungal inoculum plug, after the co-incubation in liquid medium with the strongest antagonizing yeast strains, was stained with Methylene Blue (MB) and Propidium Iodide (PI) (not shown). The first dye binds to negatively charged molecules when it is not reduced, thus staining dead cells, and the second only enters the cell when the plasma membrane is disrupted (Kwolek-Mirek and Zadrag-Tecza, 2014). In some combinations both mycelium and spores were well stained, while other combinations showed no or weak staining. These preliminary results suggest that some yeast strains could be killing the fungi, but in the combinations that were not stained, yeast strains are probably just inhibiting fungal proliferation, allowing these to retain viability. This exploratory microscopic inspection of all combinations did not reveal yeasts attaching to the hyphae as previously observed with another phytopathogen (Ferraz et al., submitted). CONTROL NO inhibition WEAK inhibition STRONG inhibition 1 2 3 114 CONCLUSIONS All taken, results suggested that fungal growth inhibition might occur through the secretion of a diffusible compound by antagonizing yeasts. For this to happen, yeasts have to be metabolically active. Whether this compound causes the death of the fungi, or simply inhibits their replication, remains to be confirmed. Moreover, results also showed that yeast strain #18, a W. anomalus isolated from olive orchards in Portugal, has a strong ability to antagonize the three Colletotrichum sp. strains used in this work. These results agree with other studies proposing the potential of W. anomalus (formerly known as Pichia anomala or Hansenula anomala ) as biocontrol agent (Friel et al. , 2007; Oro et al. , 2014; Parafati et al. , 2015). This species possesses other desirable characteristics, like the ability to grow under high osmotic pressure, at low pH, and at a broad range of temperatures (Fredlund et al. , 2002). The fact that the best yeast candidate originates from the olive trees biome suggests that, better than using well-known yeasts for long captive of laboratory or industrial processes and accordingly domesticated, the exploitation of an appropriate niche biodiversity yields better strains, able to efficiently colonise the olive tree orchards. The niches where the natural wild yeast and the phytopathogens co-exist are most probably ideal. The results in this work suggested as much and can be considered a steppingstone for the formulation of a new eco-friendly and effective alternative for controlling OA. Further studies are required in order to fully understand the mechanisms by which these yeasts are able to antagonize the phytopathogenic fungi. ACKNOWLEDGEMENTS This work was supported by the strategic programme UID/BIA/04050/2013 (POCI-01-0145-FEDER-007569) funded by national funds through the FCT I.P. and by the ERDF through the COMPETE2020 - Programa Operacional Competitividade e Internacionalização (POCI), and the project EcoAgriFood (NORTE-01-0145-FEDER-000009), supported by the Norte Portugal Regional Operational Programme (NORTE 2020) under the PORTUGAL 2020 Partnership Agreement through the European Regional Development Fund (ERDF). PF and MA-R are PhD students of the Doctoral Programmes in Applied and Environmental Microbiology (DP_AEM), (FCT grant PD/BD/113810/2015) and in Molecular and Environmental Biology (PDBMA), (FCT grant PD/BD/145354/2019), respectively. We thank the Brazilian companies, Cerlev, Lda., Ouro Preto, MG, and Fermentec, Lda. Soluções Tecnológicas e Industriais, Piracicaba, SP, as well as Professor João Paulo Sampaio from the Portuguese Yeast Culture Collection/UCIBIO, NOVA, Portugal for kindly supplying the yeast strains used in this work. We also thank Professor Pedro Talhinhas, from LEAF - Linking Landscape, Environment, Agriculture and Food, ISA, Universidade de Lisboa, Portugal, for kindly supplying two of the fungal strains used in this work. 115 SUPPLEMENTARY MATERIAL Table S1. Growth rates of the phytopathogenic fungal strains Colletotrichum gloeosporioides s.s. , C. godetiae and C. nymphaeae at 25 and 30 °C, in two culture media and at different pH values. Growth rate (mm.day -1) pH 4.0 pH 4.5 pH 5.0 pH 5.5 pH 6.0 30 ºC C. gloeosporioides PDA 3.90 4.43 5.31 5.13 5.86 MEA 4.10 5.89 6.14 5.76 6.14 25 ºC C. gloeosporioides PDA 3.17 4.84 5.00 5.38 5.08 MEA 4.06 4.80 5.56 5.89 5.76 C. godetiae PDA 1.88 2.77 1.77 2.68 2.56 MEA 1.77 2.21 2.76 2.25 2.46 C. nymphaeae PDA 2.17 2.47 2.71 3.15 2.54 MEA 2.41 2.39 3.18 2.90 2.64 116 REFERENCES Almeida, M. J. V. (1899). La gaffa des olives en Portugal. Bulletin de la Societe Mycologique de France 15, 90-94. Cacciola, S. O., Faedda, R., Sinatra, F., Agosteo, G. E., Schena, L., Frisullo, S., et al. (2012). Olive Anthracnose . Journal of Plant Pathology 94(1), 29-44. de Lima, J. R., Gonçalves, L. R. B., Brandão, L. R., Rosa, C. A., and Viana, F. M. P. (2013). Isolation, identification, and activity in vitro of killer yeasts against Colletotrichum gloeosporioides isolated from tropical fruits. Journal of Basic Microbiology 53(7), 590-599. doi: 10.1002/jobm.201200049. EUR-Lex. (2018). Access to European Union Law. [https://eur-lex.europa.eu/legalcontent (Accessed January 20, 2020)] Ferraz, P., Amorim-Rodrigues, M., Cássio, F., and Lucas, C. Saccharomyces cerevisiae and Wickerhamomyces anomalus are able to kill Moniliophthora perniciosa , the causal agent of cacao Witches’ Broom Disease . Submitted. Ferraz, P., Cássio, F., and Lucas, C. (2019). Potential of Yeasts as Biocontrol Agents of the Phytopathogen Causing Cacao Witches’ Broom Disease : Is Microbial Warfare a Solution? Frontiers in Microbiology 10, 1766. doi: 10.3389/fmicb.2019.01766. Fredlund, E., Druvefors, U., Boysen, M. E., Lingsten, K. J., and Schnurer, J. (2002). Physiological characteristics of the biocontrol yeast Pichia anomala J121. FEMS Yeast Research 2(3), 395402. doi: 10.1016/s1567-1356(02)00098-3. Friel, D., Pessoa, N. M. G., Vandenbol, M., and Jijakli, M. H. (2007). Separate and Combined Disruptions of Two Exo-β-1,3-Glucanase Genes Decrease the Efficiency of Pichia anomala (Strain K) Biocontrol Against Botrytis cinerea on Apple. Molecular Plant-Microbe Interactions 20(4), 371379. doi: 10.1094/mpmi-20-4-0371. Hagens, S., and Loessner, M. J. (2010). Bacteriophage for Biocontrol of Foodborne Pathogens: Calculations and Considerations. Current Pharmaceutical Biotechnology 11, 58. doi: 10.2174/138920110790725429. 117 Hatoum, R., Labrie, S., and Fliss, I. (2012). Antimicrobial and probiotic properties of yeasts: From fundamental to novel applications. Frontiers in Microbiology 3, 421. doi: 10.3389/fmicb.2012.00421. Hernández, A., Martín, A., Córdoba, M. G., Benito, M. J., Aranda, E., and Pérez-Nevado, F. (2008). Determination of killer activity in yeasts isolated from the elaboration of seasoned green table olives. International Journal of Food Microbiology 121(2), 178-188. doi: 10.1016/j.ijfoodmicro.2007.11.044. Hothorn, T., Bretz, F., and Westfall, P. (2008). Simultaneous Inference in General Parametric Models. Biometrical Journal 50(3), 346-363. doi: 10.1002/bimj.200810425. Kwolek-Mirek, M., and Zadrag-Tecza, R. (2014). Comparison of methods used for assessing the viability and vitality of yeast cells. FEMS Yeast Research 14(7), 1068-1079. doi: 10.1111/15671364.12202. Landum, M. C., Félix, M. d. R., Alho, J., Garcia, R., Cabrita, M. J., Rei, F., et al. (2016). Antagonistic activity of fungi of Olea europaea L. against Colletotrichum acutatum . Microbiological Research 183, 100-108. doi: 10.1016/j.micres.2015.12.001. Leoni, C., Bruzzone, J., Villamil, J. J., Martínez, C., Montelongo, M. J., Bentancur, O., et al. (2018). Percentage of anthracnose ( Colletotrichum acutatum s.s.) acceptable in olives for the production of extra virgin olive oil. Crop Protection 108, 47-53. doi: 10.1016/j.cropro.2018.02.013. Lima, J. R., Gondim, D. M. F., Oliveira, J. T. A., Oliveira, F. S. A., Gonçalves, L. R. B., and Viana, F. M. P. (2013). Use of killer yeast in the management of postharvest papaya anthracnose. Postharvest Biology and Technology 83, 58-64. doi: 10.1016/j.postharvbio.2013.03.014. Liu, G. L., Chi, Z., Wang, G. Y., Wang, Z. P., Li, Y., and Chi, Z. M. (2015). Yeast killer toxins, molecular mechanisms of their action and their applications. Critical Reviews in Biotechnology 35(2), 222234. doi: 10.3109/07388551.2013.833582. Lopes, M. R., Klein, M. N., Ferraz, L. P., da Silva, A. C., and Kupper, K. C. (2015). Saccharomyces cerevisiae : A novel and efficient biological control agent for Colletotrichum acutatum during preharvest. Microbiological Research 175, 93-99. doi: 10.1016/j.micres.2015.04.003. 124 Brazil produces bioethanol mostly from yeast fermentation of sugarcane juice. The process involves the physical extraction of the juice from the cane which is then fermented by yeasts. After each fermentation cycle, the must is centrifuged, and the yeast biomass separated and recycled. The subsequent distillation process yields a dark liquid waste called vinasse (Carrilho et al., 2016). Brazilian industry generates an average 12 L of vinasse per litre of ethanol produced (Lopes et al., 2016), meaning that each year the country generates around 300-400 billion litres of vinasse (Carrilho et al., 2016). This residue has been used for more than 50 years mostly for fertirrigation (fertilization + irrigation) of the sugarcane fields (Carrilho et al., 2016). This appears to be a virtuous cycle, since vinasse is rich in potassium (Naspolini et al. , 2017), avoiding the need for chemical fertilization (Prado et al., 2016), which in the long run is beneficial for soil quality and sugarcane production yield (Naspolini et al ., 2017). But in reality, it configures a serious environmental problem since the amounts of vinasse produced are far bigger than the fields can absorb, leading to their saturation. Additionally, the vinasse’s low pH (3.5 – 5.0) and the corrosive nature (Fuess et al. , 2017) are promoting undesirable changes of soil composition and physicochemical properties (Christofoletti et al., 2013), and serious contamination of ground water (Botelho et al., 2012). These problems are exacerbated by illicit discharges and inappropriate storage. The most studied possibilities of vinasse utilization relate with its reuse in the energy equation of the ethanol plant (Naspolini et al. , 2017; Rodrigues Reis and Hu, 2017). Nevertheless, vinasse is rich in minerals and organic matter (Goldemberg et al., 2008; Nitayarvardhana et al., 2013; Carrilho et al., 2016), around 100 – 130 g/L (COD) (Goldemberg et al., 2008), and includes sugars, organic acids, ethanol and glycerol (Nitayarvardhana et al., 2013; Carrilho et al., 2016), which suggest vinasse should be fit for microbial growth. Studies proposing this possibility include the cultivation of economically promising filamentous fungi such as the edible Rhizopus oligosporus (Nitayarvardhana et al., 2013), or the biotechnology relevant Aspergilus oryzae (Money, 2016; Gmoser et al., 2018) and Neurospora intermedia (Nair and Taherzadeh , 2016). In the present work, we explored the possibility of using vinasse for controlling the proliferation of M. perniciosa . Three different vinasse s were tested against two strains of the fungus originating from Brazil via a credited international culture collection. The fungi did not grow on vinasse . Instead, both died upon immersing or spraying with two of them, although at varying time/dosage, and were inhibited from proliferating by the third. One of the three tested vinasses inhibited the growth of the fungi. This killing effect of vinasse was not shared by another genetically distant phytopathogenic fungus suggesting specificity. Results suggest that using vinasse for fertirrigation of cacao plantations could, in time, clean the resident fungus inoculum. If not eradicating it, it could at least reduce the disease to manageable levels. 125 MATERIALS AND METHODS Moniliophthora perniciosa strains CBS 441.80 and 442.80, and Colletotrichum gloeosporioides CBS 100471 were purchased from CBS-KNAW Collections, The Netherlands. They were maintained at 4 °C on MEA (20 g/L malt extract w/ 20 g/L agar), and cultivated in the same medium at 30 °C. Vinasse was obtained from Fermentec Lda (https://www.fermentec.com.br/), and originated from three different bioethanol producing plants in the State of São Paulo, Brazil: Usina Alta Mogiana (http://www.altamogiana.com.br/), Usina Batatais (http://www.usinabatatais.com.br/) and Usina da Pedra (https://www.pedraagroindustrial.com.br/). Vinasse was stored at 4 °C. Prior to utilization it was centrifuged for 30 min at 13.000 x g and 4°C to eliminate particles in suspension, and autoclaved (121 ºC, 1 atm, 20 min). Growth assays were performed inoculating 20 mL of vinasse in a glass tube (Ø 3 cm; 13 cm height) with a one-week ME-grown fungus agar plug of approximately 0.8x0.8 cm. Tubes were incubated at 30 ºC and 200 rpm orbital shaking. Cultures identically inoculated and incubated in liquid ME medium were used as control. Growth was assessed after 10 days, decanting the culture supernatant and checking for mycelium development. The putative viability of any remaining fungal cells was assessed incubating this agar plug in MEA at 30°C for 1 week. These assays were repeated in the same manner, using vinasse supplemented with 2% of malt extract as a supplementary nutrient source. Death along incubation time in vinasse was assessed using a fully-grown fungal culture previously grown in liquid ME medium in a glass tube (until the formation of a globular mycelia), which was then transferred to a glass tube containing 20 mL of vinasse , under the same conditions described above. Each sample/incubation time corresponding to an independent tube identically inoculated from fully-grown fungal cultures. The incubation period varied between 1 and 10 days (T1 to T10), and at increasing time points, the mycelia was taken from the vinasse , gently washed with sterile water and its viability assayed in MEA at 30°C for 1 week. The death-inducing effect without immersing the cells in vinasse was performed by spraying vinasse on a MEA plate containing fully grown mycelia. The assay was performed during 10 days, applying a single spray at T0, or repeating it once each 24h. The plates were then incubated at 30°C for 3 days. All assays were performed at least in three independent replicates (n ≥ 3). 126 RESULTS AND DISCUSSION The vinasse used in this work originated from three bioethanol plants from the State of São Paulo in Brazil, all of which identically producing bioethanol from a sugarcane juice fermentation process. According to the literature, the composition and physical properties of vinasses do not differ significantly (e.g. Goldemberg et al., 2008; Nitayarvardhana et al., 2013). The three vinasses used in this work have ±6 % solids in suspension (mostly ashes), pH ≈ 4.5, 0.115 ± 0.060 % (w/v) sugars and 0.9 % (w/v) organic acids (information kindly supplied by Fermentec, Lda (https://www.fermentec.com.br/), values that fall within published data. The pH of vinasse is adequate to support the growth of M. perniciosa which optimal growth occurs at pH between 4.5 and 5.5 (our unpublished results), and the presence of carbon source (totalling ±0.5% (w/v)) should allow some fungal multiplication. Therefore, vinasse was centrifuged, to remove the solids in suspension, and the remaining liquid fraction was sterilized and used to inoculate the two Moniliophthora perniciosa strains. After 10 days (Figure 1A) there was no visible growth of either fungus. Figure 1. Two M. perniciosa strains originating from contaminated cacao fruits and trees in Brazil were grown in unsupplemented vinasse originating from three Brazilian bioethanol plants. (A) Results were scored after a 10 days incubation at 30°C, before and after decanting the vinasse showing the inoculum agar plug. Control growth in identical period and temperature in MEA showing the fungal biomass ball. (B) The plugs from (A) were inoculated in MEA solid medium. Only the plugs originating from the incubation in vinasse from Pedra showed mycelia growth. 127 Microscopic inspection confirmed the absence of hyphae associated with the agar plugs used as inocula. To check whether there were still residual viable mycelia, the fungal plugs were taken from the vinasse (Figure 1A), re-inoculated in solid MEA medium and incubated for a further 10 days at 30°C. As can be seen in Figure 1B, the plugs from the vinasses of Alta Mogiana and Batatais did not develop any growth, suggesting the fungus was killed. The fungal plug taken from the vinasse from Pedra developed mycelia, showing that it did not exert a killing effect, but only a strong inhibition of fungal growth. Considering that the carbon source present in vinasse might not be sufficient to allow the fungal to develop generously, and/or that the death of mycelia could occur at such different speed that the 10 days contact with the vinasse would not be enough to detect death of the mycelia, two different assays were performed. The first consisted in increasing the amount of carbon source to support survival and growth in vinasse by adding malt extract 2% (w/v) and repeating the incubation in the same conditions used in the previous assay. Results were identical to the previous ones (not shown), indicating that the growth inhibition observed when inoculating the fungal plugs in vinasse is not related with nutrient deprivation. The second approach aimed to establish how fast M. perniciosa was dying on vinasse . The assay consisted in having ME-grown mycelia and replacing ME with vinasse (Figure 2A). At increasing time points, the mycelia were taken from the vinasse and their viability was assayed in solid MEA medium as before. This was done using the vinasse originating from Alta Mogiana plant, for having proved able to kill the two fungal strains. Results showed that the complete death of the mycelia from the two fungal strains indeed occurred when the fungi were immersed in vinasse , but at different times. The strain CBS 441.80 lost full viability between day 6 and day 7 (Figure 2B), while the strain CBS 442.80 lost full viability faster, between day 2 and day 3 (Figure 2B). Vinasse has thus a different time/dosage effect on the two strains of M. perniciosa , which must be considered for the putative application in the field. In order to check whether the death-inducing effect could be obtained without the need to immerse the mycelium in vinasse , this was used to spray MEA fully grown mycelia. The spray was applied once every 24h for 10 days and the evolution of the mycelium was photographed. (Figure 3). This time the three vinasses were used. Mycelium progressively shrunk, becoming darker coloured. Vinasses from the different plants had different time/dosage effect. The vinasse from Batatais killed the two strains faster than the one from Alta Mogiana, and Pedra’s presented the weaker inhibitory effect. Moreover, it was also possible to verify that each vinasse had a different time-effect on each of the M. perniciosa strains, being CBS 442.80 more sensitive since the mycelium disappeared between days 5 and 10 (except when using the Pedra plant vinasse ). 128 Figure 2. (A) Scheme of the procedure used to assay the effect of vinasse on the viability of the M. perniciosa along time. (B) Representative results of the turning point in the loss of viability which occurs at different time points for each fungal strain. The controls show the growth of the mycelium ball fragmented, which occurs whenever the inoculum is large. 129 Figure 3. Chosen examples of the effect of spraying vinasse on the top of fully-grown mycelia on MEA once every 24h. Arrows indicate shrunk residual mycelium after 7 applications of vinasse . These results are consistent with the ones obtained by immersing mycelium in vinasse . Results further indicate that the spraying has to be repeated over time in order to observe the desired effect, corroborating the hypothesis that the continuity of the treatment could contribute to clean the field of hidden mycelium over time. Finally, as an attempt to understand whether the ability of vinasse to kill filamentous fungi was specific for M. perniciosa and/or the correspondent taxa, the assays of fungal growth in vinasse for 10 days were repeated using another phytopathogenic fungus, Colletotrichum gloeosporioides (Weir et al., 2012). This is one of the causative agents of olive anthracnose, a disease causing serious economic losses in the Mediterranean region (Talhinhas et al., 2018) and it was chosen because it is a very resilient fungal 130 species. Results (Figure 4) showed that C. gloeosporioides was able to grow without the need for nutrient supplementation, fully occupying the volume of vinasse with mycelium which thus became altogether a hard-solid mass. These results indicate that the ability of vinasse to kill M. perniciosa does not correspond to a generalized antifungal activity, and therefore depends on some specific trait of M. perniciosa biology. Vinasse chemical composition is very complex (Rodrigues Reis and Hu, 2017). Future approaches might allow the identification of the compounds or chemical groups in vinasse that are active against M. perniciosa , eventually enabling the creation of an engineered simpler and more efficient solution which considers the environmental effects in the long run. Moreover, the evaluation of the potential spectrum of vinasse application as a specific fungicide will have to be addressed experimentally. In particular, there is one possibility that stands out, which is that vinasse might affect the proliferation of M. roreri (Barbosa et al., 2018) , another phytopathogenic filamentous fungus genetically very close to M. perniciosa, also causing a severe and economically threatening disease in cacao fruits known as moniliasis or frosty pod rot (Bailey et al., 2018). Figure 4. C. gloeosporioides grown in unsupplemented vinasse originating from three different bioethanol plants from Brazil. Inoculum was done using a fully-grown mycelia agar plug as used for M. perniciosa . Results were scored after a 10 days incubation at 30°C (1), and after decanting the remaining vinasse showing the fungal mycelium that was formed (2). Control growth was performed in MEA in identical period and temperature. C. gloeosporioides did not form a mycelium ball, instead it spread eventually filling the full liquid volume available. 131 CONCLUSIONS Vinasse is the main waste product from sugarcane bioethanol production process. It has a very high fertilization ability, but if applied in excess it becomes a critical pollution problem for soils and groundwater. Many solutions for vinasse disposal have been suggested in literature, mainly biotechnological applications, including the growth of some economically interesting microorganisms or aerobic/anaerobic digestion. Additionally, many suggestions in the literature regard solutions that promote the reduction of the amount of vinasse produced per litre of ethanol, its concentration, clearing or neutralization. For now, the most cost-effective application for vinasse keeps being fertirrigation, in time improving soil quality and crop productivity, provided it is done carefully and responsibly, guaranteeing the control over all the environmental implications. Results clearly show the potential of vinasse to control the filamentous fungus M. perniciosa , cacao Witches’ Broom Disease causative agent. Considering the huge amounts of vinasse produced in Brazil every year, and the fact that this country was the most affected by the disease, the utilization of vinasse for its containment is an expressive possibility. There may be logistic and cost problems hindering the utilization of vinasse in the short term deriving from the distances separating the bioethanol plants from the cacao fields, associated with this waste physicochemical properties, in particular its pH and corrosive nature. Nevertheless, these problems should in time be technically overcome in view of the advantage of contributing to solve two serious and urgent problems with a single procedure: introducing a new alternative for the disposal of high amounts of vinasse , reducing the serious danger to the environment and groundwater of unlawful discharges and precarious storage, and the containment of cacao Witches’ Broom Disease . Together, the procedure should impact positively in the overall economical equation and contribute significantly to the socio-economic recovery of the regions of Brazil that most suffered with the crisis caused by cacao production fall-out in the last decades. 132 ACKNOWLEDGEMENTS Funding: This work was supported by the strategic programme UID/BIA/04050/2013 (POCI-01-0145-FEDER007569) funded by national funds through the FCT I.P. and by the ERDF through the COMPETE2020 - Programa Operacional Competitividade e Internacionalização (POCI), and the project EcoAgriFood (NORTE-01-0145-FEDER000009) from Norte Portugal Regional Operational Programme (NORTE 2020) under the PORTUGAL 2020 Partnership Agreement through the European Regional Development Fund (ERDF). PF and MA-R are PhD students of the Doctoral Programmes in Applied and Environmental Microbiology (DP_AEM), (FCT grant PD/BD/113810/2015) and in Molecular and Environmental Biology (PDBMA), (FCT grant PD/BD/145354/2019), respectively. The images of Witches’ Broom Disease and cacao tree and fruit used in the Graphical Abstract were obtained at https://www.forestryimages.org/: credit by Scott Bauer, USDA Agricultural Research Service, Bugwood.org, licensed under a Creative Commons Attribution 3.0 License. 133 SUPPLEMENTARY MATERIAL Figure S1. Graphical abstract.