Evaluation of the intra and interspecific variability in the genus Perkinsus. Proteomic analysis of the parasite and its interaction with the host

TESE DE DOUTORAMENTO EVALUATION OF THE INTRA AND INTERSPECIFIC VARIABILITY IN THE GENUS Perkinsus . PROTEOMIC ANALYSIS OF THE PARASITE AND ITS INTERACTION WITH THE HOST Asdo. .............................................................. SERGIO FERNÁNDEZ BOO INSTITUTO DE ACUICULTURA SANTIAGO DE COMPOSTELA 2015 CENTRO DE INVESTIGACIÓNS MARIÑAS (XUNTA DE GALICIA) UNIVERSIDAD DE SANTIAGO DE COMPOSTELA Evaluación de la variabilidad intra e interespecífica en el género Perkinsus. Análisis proteómico del parásito y de su interacción con el hospedador. Memoria presentada por SERGIO FERNÁNDEZ BOO para optar al título de doctor en Biología Codirectores Antonio Villalba García Asunción Cao Hermida Paulino Martínez Portela Santiago de Compostela, Enero 2015 D. Antonio Villaba García, Doctor en Biología e Investigador del Centro de Investigacións Mariñas de la Consellería de Medio Rural e do Mar (Xunta de Galicia). Dña María Asunción Cao Hermida, Doctora en Biología e Investigadora del Centro de Investigacións Mariñas de la Consellería de Medio Rural e do Mar (Xunta de Galicia). D. Paulino Martínez Portela, Catedrático del Departamento de Genética de la Universidad de Santiago de Compostela. INFORMAN: De que la memoria titulada “Evaluation of the intra and interspecific variability in the genus Perkinsus. Proteomic analysis of the parasite and its interaction with the host”, que presenta D. Sergio Fernández Boo para optar al grado de Doctor por la Universidad de Santiago de Compostela, ha sido realizada bajo su dirección y considerándola concluida, autorizan su presentación a fin de que pueda ser juzgada por el tribunal correspondiente. Y para que así conste, se firma el presente informe Fdo. Dr. Antonio Villalba García. Fdo. Dra. María Asunción Cao Hermida Fdo. Dr. Paulino Martínez Portela. Santiago de Compostela, a 30 de Enero de 2015 D. Carlos Pereira Dopazo, Doctor en Biología y Profesor Titular del Dpto. de Microbiología y Parasitología de la Universidad de Santiago de Compostela, INFORMA: De que la memoria titulada “Evaluation of the intra and interspecific variability in the genus Perkinsus. Proteomic analysis of the parasite and its interaction with the host”, que presenta D. Sergio Fernández Boo para optar al grado de Doctor por la Universidad de Santiago de Compostela, ha sido realizada bajo su tutela y considerándola concluida, autoriza su presentación a fin de que pueda ser juzgada por el tribunal correspondiente. Y para que así conste, firma el presente informe Fdo.: Dr. Carlos Pereira Dopazo Santiago de Compostela, a 30 de Enero de 2015. Durante el desarrollo del presente trabajo Sergio Fernández Boo ha disfrutado de una beca de formación de personal investigador de la Xunta de Galicia (DOGA nº11, 16 de Enero de 2008, 1058-1065). La presente tesis doctoral ha sido desarrollada en el Centro de Investigacións Mariñas (CIMA), perteneciente a la Xunta de Galicia, a través de los proyectos siguientes: Proyecto: “La perkinsosis en el litoral español: caracterización de variantes taxonómicas del parásito, de su ciclo de vida y de la respuesta inmunitaria del hospedador”. Duración: 01/10/2006-31/12/2009 (3 años). Financiado por el Ministerio de Ciencia e Innovación. Nº Ref: AGL2006-11809/ACU. Proyecto: “Análisis proteómico de la interacción del sistema inmune de la almeja con Perkinsus olseni”. Duración: 14/12/200913/12/2010 (1 año). Financiado por el Ministerio de Ciencia e Innovación. Nº Ref: AGL200913305C03-01. Proyecto: “Perkinsosis de la almeja: expresión proteínica en los hemocitos y plasma del hospedador durante su interacción”. Duración: 01/01/201231/12/2014 (2 años). Financiado por el Ministerio de Ciencia e Innovación. Nº Ref: AGL2011-30449-C02-01. Index Index I. INTRODUCTION ..................................................................................................... 21 I.1. PRODUCTION OF CLAMS ....................................................................................... 23 I.1.1. Pathological conditions of clams R. philippinarum and R. decussatus. .......... 25 I.2. IMMUNE SYSTEM OF R. philippinarum AND R. decussatus .................................. 26 I.2.1. Haemocytes .................................................................................................... 27 I.2.2. Functions of haemocytes ................................................................................ 28 I.3. PERKINSOSIS .......................................................................................................... 34 I.3.1. General overview ............................................................................................ 34 I.3.2. Life cycle .......................................................................................................... 34 I.3.3. Taxonomy and phylogeny ............................................................................... 38 I.3.4. Epidemiology .................................................................................................. 43 I.3.5. Diagnostic methods ........................................................................................ 44 I.3.6. In vitro culture of Perkinsus spp...................................................................... 46 I.3.7. Interaction Perkinsus – clam ........................................................................... 47 I.3.8. Fighting perkinsosis ......................................................................................... 52 I.4. MICROSATELLITES ................................................................................................. 54 I.5. PROTEOMICS ......................................................................................................... 55 I.5.1. Proteomic separation tools ............................................................................ 56 I.5.2. Mass spectrometry ......................................................................................... 62 I.5.3. Protein identification ...................................................................................... 67 I.5.4. Sequence database searching ........................................................................ 70 I.5.5. Proteomic applications for molluscs and parasites ........................................ 71 I.5.6. Functional genomics and proteomics ............................................................. 73 II. JUSTIFICATION AND OBJECTIVES .......................................................................... 75 III. EVALUATION OF THE GENETIC VARIABILITY OF Perkinsus olseni AMONG REGIONS OF THE SPANISH COAST USING MICROSATELLITE MARKERS ..................................... 79 III.1. ABSTRACT ............................................................................................................ 81 III.2. INTRODUCTION ................................................................................................... 83 III.3. MATERIALS AND METHODS ................................................................................ 84 Index 18 III.3.1. Parasite isolates and microsatellite markers ............................................ 84 III.3.2. Data analysis .............................................................................................. 85 III.4. RESULTS ............................................................................................................... 87 III.4.1. Genetic variability ...................................................................................... 87 III.4.2. HWE deviations and genotypic disequilibrium ......................................... 88 III.4.3. Distribution of MLGs within and among hosts .......................................... 88 III.4.4. Population structure and differenciation .................................................. 89 III.5. DISCUSSION ......................................................................................................... 93 IV. VARIABILITY OF THE CELL PROTEOME OF Perkinsus olseni AMONG REGIONS OF THE SPANISH COAST ................................................................................................. 97 IV.1. ABSTRACT ............................................................................................................ 99 IV.2. INTRODUCTION ................................................................................................. 101 IV.3. MATERIALS AND METHODS .............................................................................. 102 IV.3.1. Production of in vitro clonal cultures. ..................................................... 102 IV.3.2. Protein extraction .................................................................................... 102 IV.3.3. Two dimensional electrophoresis (2DE) and image analysis. ................. 103 IV.3.4. Protein identification and database searching ....................................... 104 IV.3.5. Statistical analysis .................................................................................... 105 IV.4. RESULTS ............................................................................................................. 106 IV.4.1. In vitro proliferation of Perkinsus olseni clones. ..................................... 106 IV.4.2. Protein expression patterns of Perkinsus olseni ..................................... 107 IV.4.3. Protein identification .............................................................................. 107 IV.5. DISCUSSION ....................................................................................................... 115 V. VARIABILITY OF PROTEIN EXPRESSION PROFILING IN THE EXTRACELLULAR PRODUCTS OF Perkinsus olseni AMONG REGIONS OF THE SPANISH COAST ............ 119 V.1. ABSTRACT ........................................................................................................... 121 V.2. INTRODUCTION .................................................................................................. 123 V.3. MATERIALS AND METHODS ............................................................................... 124 V.3.1. Production of in vitro clonal cultures ....................................................... 124 V.3.2. Protein extraction ..................................................................................... 125 V.3.3. Two dimensional electrophoresis and imaging analysis .......................... 125 V.3.4. Protein identification and database searching ........................................ 126 Index V.4. RESULTS .............................................................................................................. 127 V.4.1. Protein expression patterns of Perkinsus olseni extracellular products .. 127 V.4.2. Protein identification ............................................................................... 129 V.5. DISCUSSION ........................................................................................................ 135 VI. COMPARISON OF PROTEIN EXPRESSION PROFILES BETWEEN THREE Perkinsus spp.: P. olseni, P. marinus and P. chesapeaki .......................................................... 139 VI.1. ABSTRACT .......................................................................................................... 141 VI.2. INTRODUCTION ................................................................................................. 143 VI.3. MATERIALS AND METHODS .............................................................................. 145 VI.3.1. In vitro clonal cultures ............................................................................. 145 VI.3.2. Protein extraction .................................................................................... 145 VI.3.3. Two dimensional electrophoresis (2DE) ................................................. 145 VI.3.4. Protein visualisation and image analysis ................................................. 146 VI.3.5. Protein identification and database search ............................................ 146 VI.3.6. Statistical analysis .................................................................................... 147 VI.4. RESULTS ............................................................................................................. 147 VI.5. DISCUSSION ....................................................................................................... 156 VII. PROTEIN EXPRESSION PROFILING IN HAEMOCYTES OF THE MANILA CLAM Ruditapes philippinarum IN RESPONSE TO INFECTION WITH Perkinsus olseni ........ 161 VII.1. ABSTRACT ......................................................................................................... 163 VII.2. INTRODUCTION ................................................................................................ 165 VII.3. MATERIALS AND METHODS ............................................................................. 166 VII.3.1. Clams and exposure to Perkinsus olseni ................................................ 166 VII.3.2. Protein extraction ................................................................................... 167 VII.3.3. Two dimensional electrophoresis (2DE) and image analysis ................. 167 VII.3.4. Protein identification and database search ........................................... 168 VII.3.5. Statistics ................................................................................................. 169 VII.4. RESULTS ............................................................................................................ 169 VII.5. DISCUSSION ...................................................................................................... 174 VIII. PROTEIN EXPRESSION PROFILING IN PLASMA HAEMOLYMPH OF THE MANILA CLAM Ruditapes philippinarum IN RESPONSE Perkinsus olseni INFECTION ............. 179 VIII.1. ABSTRACT ........................................................................................................ 181 VIII.2. INTRODUCTION ............................................................................................... 183 Index 20 VIII.3. MATERIALS AND METHODS ............................................................................ 184 VIII.3.1. Clams and exposure to Perkinsus olseni ............................................... 184 VIII.3.2. Protein extraction.................................................................................. 185 VIII.3.3. Two dimensional electrophoresis (2DE) and image analysis ................ 185 VIII.3.4. Protein identification and database search .......................................... 185 VIII.3.5. Statistics ................................................................................................ 186 VIII.4. RESULTS ........................................................................................................... 186 VIII.5. DISCUSSION ..................................................................................................... 191 IX. GENERAL DISCUSSION ....................................................................................... 197 X. CONCLUSIONS .................................................................................................... 201 XI. REFERENCES ...................................................................................................... 205 XII. ANNEX I: SUPPLEMENTARY MATHERIAL AND METHODS ................................. 251 XIII. RESUMEN ........................................................................................................ 255 I. INTRODUCTION I. Introduction 23 I.1. PRODUCTION OF CLAMS Aquaculture of clams is an industry that is rising every year. A perspective of the world clam production in the last ten years shows an increment of the production of 74,5%. The clam species with the highest worldwide production is the Manila clam Ruditapes philippinarum. China is the main producer of this species, followed by Italy, South Korea, USA and Spain. The production of this species reached 3,785,311 tonnes in 2012. (FAO Fisheries and Aquaculture Statistics. http://www.fao.org/ fishery/ statistics/ en) (Fig. I.1). Fig. I.1. World clam production of the last 10 years. FAO Fisheries and Aquaculture Statistics. The production of venerid clams in Galicia, based both on fishery and aquaculture, is a very important socio-economic resource, directly employing 4000 people, most of them women. In some coastal populations, venerid clam industry is the main source of incomes. This activity generated about 67,6 m€ in the year 2013, with a production of 9,832 tonnes (www.pescadegalicia.com). The introduction of foreign species in marine ecosystems is an established method to increase productivity and generate incomes. The introduction of the Manila clam Ruditapes philippinarum in Europe during the 1970s and 80s is a good example (Gosling, 2002). Reproduction of Manila clam along the European coastline resulted in its rapid spread and naturalisation. R. philippinarum has proved to be well adapted and faster growing than the native European grooved carpet shell clam Ruditapes decussatus (Jensen et al., 2004; Hurtado et al., 2011). Consequently, R. philippinarum is the major contributor to clam landings in Europe. In Galicia, R. philippinarum culture rose the last years (Fig. I.2) being the clam species with the highest culture production in 2013 (1,350 tons), followed by Venerupis corrugata (273,5 tons) and R. decussatus (184,2 tons) (www.pescadegalicia.com). 0 500000 1000000 1500000 2000000 2500000 3000000 3500000 4000000 4500000 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Production in tonnes Worldwide production of R. philippinarum I. Introduction 24 Fig. I.2. Evolution of the annual aquaculture production of Ruditapes decussatus and Ruditapes philippinarum in Galicia for the period 2004-2013. www.pescadegalicia.com. Additionally, venerid clam production from fisheries Galicia is even higher (Fig. I.3). The fishery production of R. decussatus has remained constant in the last 13 years while the production of R. philippinarum shows an increasing trend since 2001. Nowadays, R. philippinarum is the venerid clam species with the highest production in Galicia. Fig. I.3. Evolution of annual fishery production (columns) of Ruditapes decussatus and Ruditapes philippinarum in Galicia for the period 2001-2013. The lines correspond to linear adjustment. www.pescadegalicia.com 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Production in tonne Aquaculture production of Ruditapes spp. in Galicia R. decussatus R. philippinarum 0 500 1000 1500 2000 2500 3000 Production in tonnes Fishery of Ruditapes spp. in Galicia R. decussatus R. philippinarum Lineal (R. decussatus) Lineal (R. philippinarum) I. Introduction 25 I.1.1. Pathological conditions of clams R. philippinarum and R. decussatus. The research in mollusc pathology progresses as mollusc aquaculture spreads. The main pathological problems of R. decussatus and R. philippinarum are summarised in Table I.1. Table I.1. Main pathogens of the clams Ruditapes decussatus and Ruditapes philipinarum Species Herpes-like virus Host type: R. philippinarum Mortality: Yes (larvae) Reference: Renault and Arzul, 2001. Species Picornavirus Host type: R. decussatus Mortality: Yes Reference: Novoa and Figueras, 2000. Species Rikettsia-like bacteria Host type: R. decussatus Mortality: Not reported Reference: Navas et al.,1992; Villalba et al., 1993; López et al., 1998. Species Vibrio tapetis Host type: R. philippinarum and R. decussatus Mortality: Yes Reference: Paillard and Maes, 1994; Borrego et al., 1996; Drummond et al., 2007. Species Minchinia tapetis Host type: R. decussatus and R. phillipinarum Mortality: Not reported Reference: Joly, 1982; Chagot et al., 1987; Villalba and Navas, 1988; López et al., 1998 Species Perkinsus olseni Host type: R. philippinarum and R. decussatus Mortality: Yes Reference: Azevedo, 1989; Santmarti et al., 1995, Villalba et al., 2005; Pretto et al., 2014 Spcies Cercaria tapidis Host type: R. philippinarum Mortality: Not reported Reference: Lee et al., 2001. Spcies Cercaria lata/Bacciger bacciger Host type: R. decussatus Mortality: ? Reference: Breber, 1985; Culurgioni et al., 2006; El-Wazzan and Radwan, 2013 I. Introduction 32 process could block trophozoite dissemination via the circulatory system (Montes et al., 1995a). Fig. I.4. Image of a histological section of a clam Ruditapes decussatus showing encapsulation of Perkinsus olseni (arrows) by layers of haemocytes. I.2.2.4. Phagocytosis The major cellular mechanism involved in the hemocyte mediated immune response of marine bivalves is the internalisation of foreign material. Recognition and binding of non-self material by receptors expressed at the surface of the haemocytes induce the mechanism of internalisation called phagocytosis (Fig. I.5) (Donaghy et al., 2009). The attraction of haemocytes to foreign particles is mediated by chemotaxis and the recognition of foreign particles is produced by interaction of haemocytes and PAMPs. In clams, engulfment of foreign particles by haemocytes occurs by invagination of the cell membrane followed by pseudopod formation and particle internalisation into an endocytic vacuole, also called the primary phagosome. Cytoplasmic lysosomal granules then migrate and fuse with the primary phagosome. Contents of granules with numerous hydrolases, including phosphatases, esterases, amidases, as well as carbohydrate hydrolases, and oxidative enzymes such as peroxidase and cytochrome c oxidase (López et al., 1997a; Cima et al., 2000; Donaghy et al., 2009), are subsequently discharged in the so-called secondary phagosome, accomplishing the enzymatic degradation of engulfed foreign material. After the intracellular destruction of a pathogen, granules of glycogen are released that could be utilised in haemocytes metabolic processes or liberated to plasma (Costa, 2008). I. Introduction 33 Fig. I.5. Immune response of the oyster Crassostrea virginica against Perkinsus marinus. Histological section of an oyster showing oyster haemocytes (arrows) that have phagocytosed P. marinus. Image taken from Villalba et al. (2011). I.2.2.5. Oxidative mechanisms A high proportion of reactive oxygen species (ROS) are generated during phagocytosis and this process is called “respiratory burst”. The respiratory burst is generally described as a cascade of enzymatic reactions that starts with the production of the superoxide anion (O2-) by NADPH oxydase complexes associated with the plasma membrane (Schott and Vasta, 2003; Donaghy et al., 2009). Superoxide is converted to hydrogen peroxide (H2O2) by superoxide dismutase (SOD), and in presence of chloride ion, H2O2 is the substrate for the production of hypochlorite (HOCl), in a reaction catalysed by myeloperoxidase (MPO) (Anderson et al., 1992; Greger et al., 1995). Cytotoxic effects of ROS include peroxidation of lipids, breakage of DNA strands, and inactivation of enzymes, especially those containing Fe-S centers (Schott and Vasta, 2003). ROS production is inhibited in C. virginica haemocytes when are exposed to live P. marinus trophozoites (La Peyre et al., 1995a; Volety and Chu, 1995; Anderson, 1999; Schott and Vasta, 2003). This inhibition allow to parasite to be alive inside oyster haemocytes. Haemocytes of R. philippinarum can also produce ROS upon their activation (Cima et al., 2000) but exposure to P. olseni seems to have no significant effect in the production of ROS by clams haemocytes (Hégaret et al., 2007; da Silva et al., 2008) as well as exposure to β-glucans (Costa et al., 2008). Among ROS produced in the respiratory burst, the superoxide anion combined with the nitric oxide (NO), by the enzyme NO synthase, can generate peroxinitrite (ONOO-), which has a great oxidant I. Introduction 34 power and cytotoxic activity (Rosen et al., 1995). NO is known to have a microbicidal activity against parasites, bacteria and viruses. NO constitutes an alternative method of killing invading pathogens (Tafalla et al., 2003). The ability of R. decussatus clams to produce NO in response to zymosan or bacterial lipopolysaccharide (LPS) was measured by Tafalla et al. (2003); their results showed an increase of NO production after exposure to zymosan, LPS and the pathogenic bacteria Vibrio tapetis. The production of NO was also independent from phagocytosis. Production of Nitric Oxide (NO) is increased in C. virginica haemocytes exposed to P. marinus; NO contributes in first term to elimination of P. marinus cells but the parasite overcome the NO damage in some term and the progression of the infection continues (Villamil et al., 2007). I.2.2.6. Apoptosis Apoptosis or Programmed Cell Death is a resource that the organisms can use to avoid proliferations of intracellular pathogens. With the death of the cells the pathogens are not liberated thus avoiding avoid their dispersion (Villalba et al., 2011). P. marinus inhibits the apoptosis of C. virginica haemocytes indicating the modulation of the host immune response (Sunila and LaBanca, 2003; Hughes et al., 2010). Instead of this, there were not significant changes in apoptosis rate in R. philippinarum haemocytes exposed to P. olseni cells (Hégaret et al., 2007; da Silva et al., 2008) which is consistent with the extracellular proliferation strategy of P. olseni. I.3. PERKINSOSIS I.3.1. General overview Perkinsosis is the disease caused by the protozoan parasite Perkinsus spp. This disease has caused mass mortality of molluscs resulting in dramatic economic losses (Villalba et al., 2004). In 1950 John G. Mackin, H. Malcon Owen and Albert Collier discovered a parasite associated with high mortalities in the oyster Crassostrea virginica in the Gulf of Mexico (Mackin et al., 1950). In a first time this parasite was called Dermocystidium marinum and actually is called Perkinsus marinus. Since this first detection, several species of Perkinsus were found infecting a high variety of molluscs, bivalves and gastropods. This genus of parasites is present in North and South America, Europe, Africa, Asia and Australasia. I.3.2. Life cycle Parasites of the genus Perkinsus have three main cellular stages in the life cycle: trophozoite, hypnospore and zoospore. The trophozoite stage occurs in the host. The cell in this stage is spherical and has a large vacuole and peripheral prominent nucleus with a nucleolus; thus, the cell has “signet ring” appearance (Fig. I.6); in some species, there is a polymorphic I. Introduction 35 inclusion in the vacuole called vacuoplast (Villalba et al., 2004). The size of the trophozoite varies among species. The proliferation of the parasite through host tissues takes place by vegetative proliferation (palintomy) (Fig.I.7); trophozoites are divided by successive bi-partitioning (cycles of karyokinesis followed by cytokinesis) to yield up to 32 daughter cells that stay together inside the mother wall; eventually, the wall brakes and the daughter cells are liberated; the resulting immature throphozoites gradually enlarge and form the vacuole inside becoming mature trophozoites (Perkins, 1996; Villalba et al., 2011). Fig. I.6. Images of trophozoites of Perkinsus spp. A. Perkinsus chesapeaki in vitro-cultured mature thophozoites. The micrograph shows the typical “signet ring” appearance with the prominent nucleus (N) and the large vacuole (Va). B. Transmission electron microscopy micrograph of a mature throphozoite from a Perkinsus marinus culture. L: lipid dropet; N: nucleus; Nu. nucleolus; P: presuntive precursor material of vacuoplast in cisternae of the endoplasmatic reticulum; V: vacuoplast material inside the eccentric vacuole; Va: Vacuole; W: cell wall; white arrowhead: cluster of intranuclear virus-like particles. The image B is taken from Perkins (1996). I. Introduction 36 Fig.I.7. Micrographs showing progressive steps of vegetative multiplication of Perkinsus olseni. The trophozoite is divided in two and then four cells, which become united within the wall of the mother cell (schizont); eventually, that wall breaks and young trophozoites become free. The parasite has a dormant/resistant stage that is produced when the host dyes or is dying. Host death involves changes of environmental conditions in the host tissue, such as low oxygen, acidic pH and increased nutrient levels, change; when that occurs, thophozoites enlarge keeping the spherical shape and develope a thick wall; this stage is called hypnospore (Casas and La Peyre, 2013). If the infected host tissues are incubated in fluid thioglycollate medium (FTM), trophozoites also transfrom into hypnospores (Ray, 1952). Hypnospores can live for long periods waiting for suitable conditions to address the next stage of the life cycle, zoosporulation (Chu and Greene, 1989; Casas et al., 2002a). When hypnospores reach the water column and the environmental conditions (temperature and salinity) are suitable, the zoosporulation occurs through successive karyokinesis and cytokinesis to form hundreds of cells within the original thick cell wall; thus the hypnospore becomes a zoosporangium and the daughter cells become zoospores that are released through a discharge tube, which is formed very early in the zoosporulation process, even before the first cell division takes place (Fig. I.8). Zoospores are ellipsoidal and biflagellated (Fig. I.9) (Perkins and Menzel, 1967; Casas et al., 2002a; Villalba et al., 2004). I. Introduction 37 Fig. I.8. Progression of zoosporulation. A. Unicellular zoosporangia with the discharge tube (arrow) formed. B. Two-cells zoosporangium. C. Four-cells zoosporangium. D. Zoosporangium cell division between 4 to 8 cells. E. Eight-cells zoosporangium. F. Sixteen-cells zoosporangium. G. Hundreds of motile zoospores inside the zoosporangium. H. Release of the zoospores through the discharge tube. Images taken from Villalba et al. (2011). Fig. I.9. Transmission electronic microscopy micrograph of two Perkinsus olseni zoospores showing the ellipsoidal shape and the two flagella. Taken from Villalba et al., (2011). Transmission of Perkinsus sp. is direct mollusc to mollusc without intermediate host (Ray, 1954; Chu, 1996). After the death of the host, the zoospores released to the water can infect new molluscs (Chu, 1996; Ford et al., 2002) but new infections are possible without the death of the host. Trophozoites are released to the water through the faeces (Bushek et al., 2002; Park et al., 2010) and probably through the gonoducts (Moss et al., 2008), which can infect new hosts. The maximum transmission rate occurs when the mortality of the molluscs reaches a maximum, but the transmission is also I. Introduction 38 produced when there is not mortality, so the death of the host is not necessary for the dispersion of the disease (Ragone Calvo et al., 2003). The proximity among individuals increase the possibility of transmission (Andrews, 1965). A scheme of the life cycle of the parasites of the genus Perkinsus is shown in Fig. I.10. Fig. I.10. Life cycle of Perkinsus olseni parasitising clams. Image taken from Auzoux-Bordenave et al. (1995). I.3.3. Taxonomy and phylogeny The first parasite discovered of this group was described as a fungus, included in the genus Dermocystidium and called Dermocystidium marinum (Mackin et al., 1950). A new morphological study of the parasite concluded that it was a protozoan, and was included into the phylum Labyrinthomorpha as Labyrinthomyxa marina (Mackin and Ray, 1966). Later, according to the zoospore structure, the presence of the cortical alveoles, micropores and the conoid apical structure, the taxonomic position changed to the phylum Apicomplexa (Perkins, 1976). Then, based on this unique structure, Levine (1978) established the new genus Perkinsus, placing it within Apicomplexa. With the development of the molecular techniques, new taxonomic studies were done as molecular techniques were available; the analysis of the DNA sequence of the small subunit of the rRNA (SSU rRNA) showed that Perkinsus is closer to Dinoflagellates than to Apicomplexa (Goggin and Barker, 1993; Siddall et al., 1997). Studies based on the nuclear-encoded spliced leades (SL) RNA and mitochondrial genes, intron prevalence, and multi-protein phylogenies expanded support for the affiliation of the genus Perkinsus with an independent lineage (Perkinsozoa), I. Introduction 39 positioned between the phyla of Apicomplexa and Dinoflagellata (Zhang et al., 2011). The identification of plastid-associated major biosynthetic pathways in Perkinsus suggests the presence of a relic plastid. This plastic could suggest a common ancestor in Chromalveolata branch (nowadays Alveolata branch) as Fernandez-Robledo et al. (2011) hypothesised. Nowadays, genus Perkinsus is included into a supergroup defined as SAR (Fig. I.11), which name is derived from the acronym of the three groups united in this clade - Stramenopiles, Alveolata and Rhizaria -, and included into the Alveolata group, a group that includes Protalveolata, Dinoflagellata, Apicomplexa, Cilophora, Ellobiopsidae and Colponema (Saldarriaga et al., 2003; Cavalier-Smith and Chao, 2004; Adl et al., 2005; Adl et al., 2012). Fig. I.11. A. General view of the taxonomic position of the genus Perkinsus, according to the International Society of Protozoologists (Adl et al., 2012) where i.s. means “incertae sedis”. B. Phylogenetic relationships of the Perkinsus species based on ITS region of the rRNA gene, according to da Silva et al. (2014). I. Introduction 40 Since the discovery of Perkinsus marinus in 1950, more species have been described and 7 species are currently accepted within this genus. Morphological characters of Perkinsus do not allow discriminating species, thus taxonomy of the genus is mainly based on gene sequences, particularly those of the different regions of the rRNA gene but also actin type I and II (Dungan and Reece, 2006; Moss et al., 2008). The second discovered species of this genus was Perkinsus olseni. It was described in the abalone Haliotis rubra in Australia by Lester and Davis (1981). Later, a new species called Perkinsus atlanticus was described in carpet-shell clams Ruditapes decussatus from Portugal (Azevedo, 1989). However, due to the homology of the rRNA gene sequence, P. atlanticus was considered synonymous of P. olseni (Murrell et al., 2002). Blackbourn et al. (1998) described Perkinsus qugwadi as the agent of high mortality of the scallop Pactinopecten yessoensis in Canada. Two species of the genus were described in the period 2000 - 2001, in Chesapeake Bay, Perkinsus chesapeaki, infecting the soft-shell clam Mya arenaria (McLaughlin et al., 2000), and Perkinsus andrewsi in the Baltic clam Macoma balthica (Coss et al., 2001) but, again, P. andrewsi was considered synonymous of P. chesapeaki (Burreson et al., 2005). Casas et al. (2004) described Perkinsus mediterraneus infecting the European flat oyster Ostrea edulis in Mediterranean waters. Recently, two more species have been described in Asia, Perkinsus honshuensis infecting the Manila clam Ruditapes philippinarum in Japan (Dungan and Reece, 2006) and Perkinsus beihaiensis infecting the oysters Crassostrea hongkongensis and Crassostrea ariankensis in China (Moss et al., 2008). The current geographic distribution of Perkinsus spp. is shown in Fig. I.12. The information on valid species of Perkinsus spp. and their known hosts is summarised in Table I.2. Fig. I.12. Geographic distribution area of the Perkinsus spp. I. Introduction 41 Table I.2. Valid species of the genus Perkinsus, geographic distribution and their host Species Type host Other hosts Affected countries References Perkinsus marinus Crassostrea virginica Oysters: Crassostrea gigas, Crassostrea ariakensis, Crassostrea rhizophorae, Crassostrea corteziensis, Crassostrea gasar, Saccostrea palmula Clams: Mya arenaria, Macoma balthica, Macoma mitchelli, Mercenaria mercenaria USA, Mexico, Brazil Mackin et al., 1950; Burreson et al., 1994; Andrews, 1996; Calvo et al., 1999; Calvo et al., 2001; Coss et al., 2001; Cáceres-Martínez et al., 2008; Pecher et al., 2008; CáceresMartínez et al., 2012; da Silva et al., 2013; da Silva et al., 2014. Perkinsus olseni (= atlanticus) Haliotis rubra Clams: Ruditapes decussatus, Ruditapes philippinarum, Austrovenus stutchburyi, Venerupis corrugata,Venerupis aurea, Tapes rhomboides, Anadara trapecia, Pitar rostrata, Protothaca jedoensis, Paphia rhomboides; Paphia aurea, Paphia undulata, Cerastoderma glaucum Oysters: Crassostrea ariakensis, Crassostrea hongkongensis, Crassostrea rhizophorae, Crassostrea gasar, Pinctada máxima, Pinctada fucata Abalones: Haliotis laevigata, Haliotis scalaris, Haliotis cyclobates Australia, New Zealand, South Korea, Japan, China, Thailand, India, Portugal, Spain, Italy, Tunisia, Uruguay and Brazil Lester and Davis, 1981; Azevedo, 1989; Norton et al., 1993; Goggin and Lester, 1995; Hamaguchi et al., 1998; Fernández-Robledo et al., 2000; De la Herrán et al., 2000; Leethochavalit et al., 2004; Cremonte et al., 2005; Park et al., 2005; Park et al., 2006; Zhang et al., 2005; Abollo et al., 2006; Moss et al., 2007; Sanil et al., 2010; El Bour et al., 2012; Ramilo et al., 2012; da Silva et al., 2014; Pretto et al., 2014; Ramilo et al., 2015. Perkinsus qugwadi Patinopecten yessoensis Canada Bower et al., 1998; Itoh et al., 2013. Perkinsus chesapeaki (= andrewsi) Mya arenaria Clams: Macoma baltica, Macoma mitchelli, Mercenaria mercenaria, Tagelus plebeius, Cyrtopleura costata, Rangia cuneata, Mulinia lateralis; Ruditapes decussatus, Ruditapes philippinarum Cockles: Cerastoderma edule Oysters: Crassostrea virginica USA, Spain, France Kotob et al., 1999; Mc Lauglin et al., 2000; Coss et al., 2001; Dungan et al., 2002; Burreson et al., 2005; Reece et al., 2008; Arzul et al., 2012; Ramilo et al., 2012; Carrasco et al., 2014. Perkinsus mediterraneus Ostrea edulis Clams: Chamaelea gallina, Venus verrucosa, Arca noae. Pectinids: Chlamys varia Spain Casas et al., 2004; Moss et al., 2008; Ramilo et al., 2015. Perkinsus honshuensis Ruditapes philippinarum Japan Dungan and Reece, 2006. Perkinsus beihaiensis Crassostrea hongkongensis Oysters: Crassostrea ariakensis, Crassostrea madrasensis Clams: Anomalocardia brasiliana China, India, Brazil Moss et al., 2008; Sanil et al., 2012; Ferreira et al., in press I. Introduction 48 Fig. I.13. Schematic representation of P. olseni encapsulation by clam R. decussatus and R. philippinarum haemocytes. Image taken from Soudant et al. (2013). Although it is unclear how efficiently can be the cells of Perkinsus spp. degraded or eliminated, the encapsulation process could block trophozoite dissemination (Rodriguez and Navas, 1995; Montes et al., 1995a). López et al. (1997c) confirmed that R. decussatus haemocytes can phagocytose P. olseni trophozoites in in vitro assays, whereas P. olseni zoospores were not phagocytosed. The involvement of lectins in the recognition and opsonisation of P. olseni cells within the immune response of clams has been described in section 2.2. The process of engulfment of P. olseni trophozoites by clam haemocytes is shown in Fig. I.14. Some studies have been performed to improve the knowledge of the genes involved in the immune response against P. olseni in R. decussatus clams. Several genes related with immunity have been described by supression sustractive hybridisation (SSH) (Prado-Alvarez et al., 2009) and microarray at a transcriptomic level (Leite et al., 2013). I.3.7.2. Parasite virulence factors Perkinsus spp. serine proteases responsible for virulence and anti-oxidant activities allowing the parasite to elude host defense have been described and characterised. I. Introduction 49 Fig. I.14. Hypothetical process of how Manila clam Ruditapes philippinarum lectins (MCL) are involved in recognition and opsonisation of Perkinsus olseni trophozoites for subsequent phagocytosis and elimination by clam haemocytes. Image taken from Soudant et al. (2013). Lytic activity Extracellular products (ECPs) from in vitro cultured P. marinus were first studied by La Peyre et al., (1995a). In vitro experiments have shown that ECPs not only reduce several host haemocyte and humoral immune functions, including reactive oxygen species (ROS) and lysozyme production, cell mobility and haemagglutination (Garreis et al., 1996; Soudant et al., 2013), but also the bactericidal activity of haemocyte against Vibrio spp. bacteria (Tall et al., 1999). ECPs mainly consist of serine proteases, which have been proposed as possible virulence factors responsible for degradation of tissues of the infected oysters (La Peyre, 1996). In P. marinus the most abundant serine protease is a quimiotrypsin which was named as perkinsin (Faisal et al., 1999). The addition of eastern oyster homogenate to the culture medium specifically enhanced expression of serine proteases in P. marinus cultures (MacIntyre et al., 2003). Supplementation of P. marinus cultures with eastern oyster plasma or tissue homogenate also enhanced their infectivity (Earnhart et al., 2004). Serine protease activities such as trypsin and α-chymiotrypsin were found in ECPs of P. marinus but they were not in those of P. olseni, although other lytic activities as esterase, glucosidase and phosphatese were detected in ECPs of both species (Casas et al., 2002c). Several proteolitic bands with high differences among isolates from the same species were reported in ECPs of P. mediterraneus (Casas et al., 2008). Differences in proteolytic band profiles of culture supernadats between isolates of P. marinus had also been reported (La Peyre and Faisal, 1997b). A study that compared enzyme profiles in the supernatants of three Perkinsus spp. cultures (P. chesapeaki, P. marinus I. Introduction 50 and P. olseni) found hydrolytic activity in all the cultures but their relative concentration seemed to be unique to each species; P. chesapeaki supernatants showed proteolytic activity (Casas et al., 2009). No report on proteolytic activity in ECPs of P. olseni has been found in the revised literature. Antioxidant capacity of Perkinsus Studies in confrontation of C. virginica haemocytes with P. marinus trophozoites showed an inhibition of ROS production by C. virginica haemocytes, otherwise, P. marinus still live with high concentrations of superoxide anion (O2-) and hydrogen peroxide (Schott et al., 2003). These observations suggest that P. marinus can inhibit ROS products or scavenge its products. Indeed, three anti-oxidant enzymatic activities involved in counteracting the oxygen dependent anti-microbial system have recently been identified and characterised in P. marinus: acid phosphatases (inhibition of O2production), superoxide dismutases (neutralisation of O2-), and ascorbate peroxidases (H2O2 removal) (Volety and Chu, 1997; Wright et al., 2002, Schott et al., 2003; Schott and Vasta, 2003; Soudant et al., 2013) (Fig. I.15). P. marinus was found resistant to the nitric oxide (NO) produced by oyster haemocytes. The parasite resisted high concentration of NO, thus evading its damaging effects (Villamil et al., 2007). Parasite virulence genes With the availability of Perkinsus spp. cultures, several authors begun the search for genes related with virulence. One of the difficulties lies on the fact that the parasite quickly losses its virulence and pathogenicity when it is cultured under standard laboratory conditions (Volety and Chu, 1994; Bushek and Allen, 1996; Ford et al., 2002; Pales Espinosa et al., 2014). Wild-type P. marinus is significantly more virulent compared with cultured parasite cells (Ford et al., 2002). Despite loss of virulence in culture media, several genes related with virulence have been identified in the last years. Genes of two superoxide dismutases (PmSOD1 and PmSOD2) were reported in P. marinus (Schott and Vasta, 2003; Fernández-Robledo et al., 2008). The finding of these genes suggested the resistance of P. marinus to exogenous oxidative damage in host phagocytes. I. Introduction 51 Fig. I.15. Antioxidant capacities of Perkinsus marinus upon phagocytosis by eastern oyster Crassostrea virginica haemocytes (phagosome is delimited by a blue line and phagocytoszed Perkinsus cell is pink). Toxic and non-toxic molecules are noted in red and blue, respectively. Pro-oxidant activities of haemocytes are represented in orange, and the anti-oxidant activities and mechanisms of P. marinus are represented in purple. Hydrolytic enzymes are represented in green. Abreviation: NO, Nitric oxide; ONOO-, peroxynitrite; O2-, superoxideanion; HOCl, Hypochloride; iNOS, inducible nitric oxide synthase; SOD, superoxide dismutase; MPO, myeloperoxidase; AP, acid phosphatase; APX, ascorbate dependent peroxidase; Nramp, Natural Resistance-Associated Macrophage Protein. Image taken from Soudant et al. (2013). Using universal primers, Brown and Reece (2003) isolated and characterised serine protease genes from P. marinus, which are involved in parasite evasion of host defense mechanisms (Chaudhuri et al., 1989). Leite et al. (2008) found high expression of hypoxia-inducible factors (HIF), mainly promoted by HIF proxyl hydrolases (HPHs), when cultures of P. olseni were supplemented with haemolymph from susceptible species to P. olseni infection; however, no significant expression of HPHs was found in cultures supplemented with haemolymph of resistant species to P. olseni infection, thus showing a high relationship between HPH gene expression and parasite virulence. A series of genes over-expressed in P. olseni confronted with R. decussatus haemolymph were identified by Ascenso et al. (2007). A comparison of these genes among P. olseni exposed to haemolymph from species with different susceptibility to P. olseni infection were measured by macroarray hybridisation (Ascenso et al., 2009). I. Introduction 52 A study of the effect of pallial mucus on P. marinus gene expression was performed based on previous evidence of significant increase of the in vivo virulence of P. marinus exposed to oyster pallial mucus (Palles Espinosa et al., 2013). As a result, the exposure of P. marinus to mucus induced significant regulation of nearly 3,600 transcripts, many of which are considered as putative virulence factors (Pales Espinosa et al., 2014). Proliferation and virulence of Perkinsus spp. is modulated by intracellular Fe+2; peroxide antimalarial drugs as iron chelators can inhibit important metabolic pathways of the parasite. Gene expression of Fe+2 transport proteins as Nramp, described first in P. marinus by Lin et al. (2011), and calcium transport protein (ATP6/SERCA) are highly expressed after exposure to iron chelators showing a way of proliferation control (Araujo et al., 2013). Genome and transcriptome of P. marinus have been sequenced (Joseph et al., 2010); their analysis should provide a better knowledge of the parasite and future prospects in the study of virulence and epidemiology. I.3.8. Fighting perkinsosis The fight against perkinsosis is a very difficult battle. Several strategies have been developed in order to minimise the incidence of the parasite, but the eradication is almost impossible in open environment, especially when a disease has been established in an area for a long period (Villalba and Figueras, 2011). The inclusion of Perkinsus marinus and P. olseni in the list of notifiable diseases of the World Organization for Animal Health (O.I.E.) (http://www.oie.int/en/animal-health-in-theworld/oie-listed-diseases-2014/) involves restriction of movements of molluscs from affected areas and other measures that should contribute to avoid disease spreading, especially to non-affected areas. Regarding fighting measures in affected areas, significant advances have been achieved in the case of the infection of oysters C. virginica with P. marinus but almost no research has been focused on P. olseni. Changes in culture procedures and fishery management were addressed to minimise the effects of infection of oysters C. virginica with P. marinus in Chesapeake Bay (USA), taking advantage of key epidemiological information: (1) salinity influences infection, below a threshold P. marinus does not proliferate and, above it, high salinity favours infection (prevalence and infection intensity increase); (2) the influence of temperature determines a seasonal pattern of infection dynamics, with new infections concentrated in the warmest months and mortality peaks in warm months of the next year; (3) infected oysters are the source of new infections, especially when the oyster dies and zoospores become free in the water column (Andrews and Ray, 1988). According to the recommended protocol, oyster producers collected oyster seed from low salinity (non-affected) areas and moved them to high salinity areas for a faster growth; this transfer had to be done in early fall after the highest infective period I. Introduction 53 (summer) and the harvest should be done before the mortality peak (summer of the second year). All the transferred oysters had to be thoroughly harvested to avoid new infective foci for the next oyster introduction. This protocol worked well for a long period but a successive years of drought (late 1980s) provoked increased salinity in previously non-affected areas, which became colonised by P. marinus; since then this fighting procedure lost effectiveness (Villalba and Figueras, 2011). The temperature in the USA Gulf coast is warmer than in Chesapeake Bay, thus seasonality of P. marinus infection in the Gulf coast is not as marked as in Chesapeake Bay; management fighting measures are mostly based on salinity regimes (Ray, 1996; La Peyre et al., 2009). Additionally, the accumulation of historical data of infection intensity, mortality and environmental conditions from Perkinsus affected areas allowed the development of useful predictive models, which are helping for the management of the C. virginica culture and fishery (Hofmann et al., 1995; Powell et al., 1996, 1997; Soniat and Kortright, 1998; Ragone Calvo et al., 2000; Soniat et al., 2006). Selective breeding programs for oysters C. virginica (highly susceptible to P. marinus and Haplosporidium nelsoni infection) have been developed, with high success at mid-, long-term in some cases, enhancing disease tolerance/resistance and diminishing prevalence and infection intensity (Ford and Haskin, 1987; Gaffney and Bushek, 1996; Ragone Calvo et al., 2003; Abbe et al., 2010; Frank-Lawale et al., 2014). The criterion to select brood-stock for these programmes was very simple and intuitive: survivors with good growth rate under long-term pressure from P. marinus and/or H. nelsoni were chosen (Abbe et al., 2010; Villalba and Figueras, 2011). Avalilability of molecular markers of tolerance/resistance to select brood-stock should allow more efficient selective breeding programmes, yielding higher survival in shorter term. Different studies of the C. virginica genome by amplified length polymorphism (Sokolova et al., 2006), quantitative trait locus (Yu and Guo, 2006) and microarray analysis (Wang et al., 2010) have been developed to search for differentially expressed genes involved in processes such as antimicrobial defence, pathogen recognition, antioxidation and apoptosis, which reveal resistance or tolerance to P. marinus infection (Wang et al., 2010). Serine protease inhibitor proteins were purified from plasma of C. virginica; these proteins inhibit perkinsin, the major extracellular protease of P. marinus, as well as other proteases (Xue et al., 2006, 2009). The overexpression of this gene as well as some mutations confers resistance against the parasite due to inhibition of the proliferation of the pathogen (La Peyre et al., 2010; Yu et al., 2011; He et al., 2012). Another tested strategy to minimise mortality was the use of triploid oysters, which have a faster growth and thus reach market size earlier than diploid ones. Nevertheless, triploid C. virginica oysters showed similar susceptibility to P. marinus as diploid ones (Barber and Mann, 1991; Meyers et al., 1991). I. Introduction 54 A more radical strategy to sustain a mollusc industry affected by a devastating disease is the substitution of the susceptible species for a resistant one (Villalba and Figueras, 2011). Two allochthonous species, Crassostrea gigas and Crassostrea ariakensis, have been tested in the Chesapeake Bay as a remedy against the devastating effects caused by P. marinus and H. nelsoni in the C. virginica industry (Mann et al., 1991; Luckenbach, 2008). However, the introduction of those exotic species into Chesapeake Bay was not advised due to the susceptibility of C. gigas to P. marinus (Calvo et al., 1999) and H. nelsoni (Burreson et al., 2000) and of C. ariakensis to P. marinus (Calvo et al., 2001; Paynter et al., 2008) and Bonamia sp. (Burreson et al., 2004). The use of therapeutic products to treat diseases of molluscs in the open marine environment is not advisable in open waters but could be successful in in-door culture facilities. Some compounds were tested in the infection of C. virginica with P. marinus (Calvo and Burreson, 1994; Faisal et al., 1999) and some of them, such as ciclohexamide and bacitracin, cause reduction of the infection level but do not eliminate the parasite completely. Other products inhibit or kill Perkinsus spp. in vitro (Gauthier and Vasta, 1994; Krantz, 1994; Dungan and Hamilton, 1995; Elandaloussi et al., 2003, 2005a,b; Lund et al., 2005; Panko et al., 2008; Alemán-Resto and FernándezRobledo, 2014). Some of these substances could be useful in close culture systems for a period of time to prevent infections (Villalba and Figueras, 2011). I.4. MICROSATELLITES Microsatellites, also known as simple sequence repeats (SSR) or short tandem repeats (STR), are non-coding repetitive DNA regions composed of small motifs of 1 to 6 nucleotides repeated in tandem, which occur in both eukaryotic and prokaryotic genomes (Field and Wills, 1998; Tóth et al., 2000). Broadly used as genetic markers, microsatellites have a particular attribute in that they suffer higher rates of mutation than the rest of the genome (Jarne and Lagoda, 1996). Microsatellites are classified according to the type of repeat sequence as perfect, imperfect, interrupted or composite. In a perfect microsatellite the repeat sequence is not interrupted by any base not belonging to their motif (e. g. TATATATATATATATA) while in an imperfect microsatellite there is a pair of bases between the repeat motifs that does not match the motif sequence (e. g. TATATATATACTATATA). In the case of an interrupted microsatellite there is a small sequence within the repeat sequence that does not match the motif sequence (e. g. TATATACGTGTATATATA) while in a composed microsatellite the sequence contains two adjacent distinctive sequence-repeats (e. g. TATATATAGTGTGTGTGT) (Oliveira et al., 2006). Microsatellites have taken advantage over other genetic markers such as amplified fragment length polymorphism (AFLP), randomly amplified polymorphic DNA I. Introduction 55 (RAPD) and restriction fragment length polymorphism (RFLP), because microsatellites have the highest rate of polymorphism, are codominants and have a mendelian segregation, the occurrence of a single genetic locus per microsatellite allows a clear observation and easy interpretation of bands in gels, and are selectively neutral (Golstein and Pollok, 1994; Vendramin et al., 1996). The use of microsatellites is widespread in science with a wide range of applications, such as construction of genetic maps of different types of organisms (Knapik et al., 1998; Cregan et al., 1999), association between the instability of the number of repeats and human genetic diseases (Stallings, 1994; Martin et al., 2009), and studies of population genetics and genotyping and paternity analysis (Borrel et al., 2002; Pardo et al., 2011; Thompson et al., 2011). Microsatellites can be used to estimate phylogenetic distance among strains or species set in different locations. Microsatellite markers were used to study populations of different parasites includes in Apicomplexa taxa, such as Plasmodium vivax (Gunawardena et al., 2010; Van den Eede et al., 2010), Trypanosoma cruzi (Llewellyn et al., 2009) and even for P. marinus (Thompson et al., 2011) Studies on genetics of Perkinsus marinus populations have been performed using RFLPs (Reece et al., 1997), variation at the ITS and NTS regions of the rRNA gene, and SOD1 and SOD2 genes (Thompson and Hare, 2005), and microsatellites (Thompson et al., 2008; Thompson et al., 2011, Thompson et al., 2014a). These studies showed no evidence for isolation by distance, and even asexual propagation seems the best way of reproduction especially when Perkinsus spp. recently arrived; movement of infected oysters may increase out crossing opportunities, potentially facilitating rapid evolution of the parasite (Thompson et al., 2014a). The development of 12 microsatellite markers in P. olseni genome (Pardo et al., 2011) provides useful tools for genetic population analysis in this species. I.5. PROTEOMICS The term “Proteome” was launched and defined by Wilkins et al. (1996) for the first time. The term was established to obtain an equivalent concept to “genome” and was defined as the full complement of proteins expressed by the genome of one organism, tissue or cell at a specific time. Imperceptibly, the proteome was transmuted into a new discipline, “proteomics”. This new discipline is defined as the high-throughput study of the proteome, including protein quantification, proteinprotein interaction, post-translational modifications and protein function (Diz et al., 2012). The proteome is very dynamic, their components vary among organisms, tissues, cells or organelles, and due to changes in their environment, stress, drugs administration, biochemical signals or their physiological or pathological state. All these factors increase the complexity of a proteome due to the activation or I. Introduction 56 suppression of gene expression, protein interactions, or post-translational modifications (Vázquez-Cobos, 2003). The main objective in proteomics is the identification of these differences and variations. It might be interesting to study the differences at a protein level of samples exposed to different conditions or in a comparison of samples from a healthy and unhealthy organism or tissue (ChicanoGálvez, 2010). The proteomic field is complementary to genomics in so much as it provides additional information about gene expression and its regulation in different tissues and/or cell types at different times. Nevertheless, proteomics offer information about expressed proteins and post-translational modifications that cannot be deduced from genomics. Studies on the transcriptome or the genome may be insufficient for understanding the phenotype because there is a lack of convergence between the proteome and the transcriptome (Diz et al., 2012). I.5.1. Proteomic separation tools In a global view, two main strategies are used to separate proteins: - Conventional: this strategy involves the separation of all the proteins occurring in a sample by two-dimensional electrophoresis (2-DE), which usually is followed by the identification of proteins with interest by mass spectrometry (MS). - Shot gun proteomics: It involves the sequencing of a complex mixture of peptides using liquid chromatography coupled with tandem mass spectrometry (LCMS/MS). The use of shot-gun proteomics allows for a greater number of proteins to be identified rapidly from a single sample, providing a more complete metabolic picture of cellular function and physiology (Schiffman et al., 2013) I.5.1.1. Conventional proteomics: two dimensional electrophoresis (2-DE) technique In 1975, Klose, O´Farrell and Scheele published various papers describing high resolution two-dimensional methods. Proteins can be separated in two steps in single spots in a gel according to their isoelectric point (pI) and molecular weight (Mw) (Fig. I.16). This new methods opened the door to the study of a new world of knowledge: the detailed workings of cellular machines (Anderson and Anderson, 1998). This technique can resolve proteins differing in a single charge and consequently, can be used in the analysis of in vivo modifications resulting in a change in charge. Proteins whose charge is changed by missense mutations can be identified (O´Farrel, 1975). Some limitations were found in 2-DE despites this technique has big advantages over other techniques, as the visualisation of all the proteins occurring in a sample and the possible identification of isoform proteins. However, proteins with extreme pI (below pH 3 and above pH 10) are very difficult to visualise, hydrophobic membrane proteins are largely absent from samples, and proteins present at less than 1,000 I. Introduction 57 copies per cell are likely undetectable (Görg et al., 2000; Wilkins and Appel, 2007). Another limitation of the technique is the loading capacity and the staining sensitivity of the 2-D gel process. Nevertheless, the combination of 2-DE and MS has become an important analytical technique for the characterisation of complex protein populations extracted from tissue, cell or subcellular fractions. The technique can separate and display more than 10,000 different proteins in a single experiment (Nordhoff et al., 2001). Fig. I.16. General scheme of a two-dimensional electrophoresis. Image taken from ChicanoGálvez (2010). I.5.1.1.1. First dimension: isoelectrofocusing (IEF) step A complex mixture of proteins can be separated by their isoelectric point (pI). The pI of a protein is the pH where its net charge is zero. The introduction of a protein mixture in a gel with pH gradient make that their N and C terminal extremes and their residues catch and release protons according to their pH. When an isoelectric field is applied, all molecules with positive net charge are attracted to the cathode and the molecules with negative net charge to the anode. When the proteins are close to their pI, they lose mobility until their charge is zero and they stop moving. It is in this stage when the proteins are “focused”. According to this property of the proteins, immobilised pH Gradient (IPG) strips were developed by Bjellqvist et al. (1982). This pH gradient was created incorporating covalently a gradient of acid and basic buffering groups into a polyacrylamide gel at the time it is cast. These buffering groups are a set of well-characterised molecules, each with a single acidic or basic buffering group Second Dimension (SDS-PAGE) First Dimension (IPG Strip) - I. Introduction 64 Fig. I.20. Scheme of Electrospray Ionization process. Image taken from www.magnet.fsu.edu I.5.2.2 Analysers Once the sample is ionised, it enters the mass spectrometer itself. The mass analysers explore different characteristics of the parent or fragment ions. In case of tandem experiments, the parent ion must be selected for further fragmentation, and the generated fragment ions will be detected. The combination of two or more analysers in the same mass spectrometer yields the high performance and resolution of the nowadays equipments. The main function of a mass analysers is to separate the ions according to their m/z ratio (March, 2009), basically by their behavior in electric or magnetic fields (El-Aneed et al., 2009). There are few types of analysers used in proteomic research: time-of-flight (TOF), quadrupole ion trap, orbitrap and Fourier Transform FT ion cyclotron resonance (ICR) analysers (Hernandez et al., 2007). I.5.2.2.1. Time of flight analyser TOF mass analysers measure ions that are accelerated in an electric field, then travel down a field-free vacuum tube towards an ion detector (Fig. I.21). All ions in the source are given the same amount of kinetic energy, but their velocity is a function of their mass and charge. The time needed to travel the distance between the source and the detector is therefore dependent on their m/z values. This can be calculated using the kinetic energy equation, given the tube length and the measured times of flight (Hernández et al., 2007). I. Introduction 65 Fig. I.21. General scheme of a MALDI TOF MS. Image taken from Hjernø and Højrup (2012). I.5.2.2.2. Quadrupole Quadrupole mass analysers (Fig. I.22) consist of four parallel and symmetrically arranged metallic roads. One couple of opposite rods have a positive electrical potential, while the other couple of opposite rods have a negative electrical potential. Ions oscillate while traversing the field along the central axis of the rods. Depending on the voltages applied, ions are either ejected from the quadrupole or sent to the detector, therefore the quadrupole is considered as mass filter. To obtain a complete spectrum, one couple continuously varies or scans the electromagnetic field in the quadrupole while the sample passes through the analyser (Hernández et al., 2007). I.5.2.2.3. Ion trap Ion trap mass analysers are devices that can store or trap charged molecules for long time. Ions are trapped by electric potentials produced by a ring-shaped electrode and two end-cap electrodes. This occurs in a space of 2-3 cm3 that is filled with an inert gas. Ions of different m/z values enter the trap at one of the end-cap electrodes and remain trapped, oscillating at frequencies that are related to their m/z values. The ions are then subjected to additional electric fields, which eject one ion species after another from the trap and they are detected, to produce a mass spectrum (Mann et al., 2001; Hernández et al., 2007). I. Introduction 66 Fig. I.22. Schematic representation of a quadrupole in a mass spectrometer. Image taken from www.files.chem.vt.edu. I.5.2.2.4. FT-ICR FT-ICR mass analysers allow ions to be accumulated and stored for periods as long as minutes. FT mass spectrometers consist of a cubic cell inside a strong magnetic field. Injected ions rotate around the magnetic field with a frequency typical for their m/z. By varying the electric fields, changes in the ion frequency of rotation can be measured and converted into m/z using a Fourier transformation (Hernández et al., 2007). I.5.2.3. Fragmentation There is no doubt that the fragmentation step of a precursor ion is a key point in proteomics analyses since it enables analyses at the tandem mass spectrometry (MS/MS) levels. In MS/MS analyses, the first analyser selects the ion(s) which proceeds to a subsequent section, where the excitation and dissociation steps will happen. Tandem mass spectrometry analyses are the result of two or more sequential separations of ions usually coupling two or more mass analysers (Glish et al., 2003; ElAneed et al., 2009). The most common fragmentation methods used in proteomics are: collision induced dissociation (CID) (Hadden et al., 1968), electron capture dissociation (ECD) (Zubarev et al., 1998) and electron transfer dissociation (ETD) (Syka et al., 2004). I. Introduction 67 I.5.2.4. Detectors The ions pass through the mass analyser and are then detected and transformed into an usable signal by a detector. Detectors are able to generate an electric current from the incident ions, which is proportional to their abundance. There are several types of detectors. The choice of detector depends on the design of the instrument and the analytical applications that will be performed. A variety of approaches are used to detect ions. However, detection of ions is always based on their charge, their mass or their velocity. Some detectors are based on the measurement of direct charge current that is produced when an ion hits a surface and is neutralised. Others are based on the kinetic energy transfer of incident ions by collision with a surface that in turn generates secondary electrons, which are further amplified to give an electronic current (de Hoffman and Stroobant, 2007). I.5.2.5. Computers A computer dedicated to mass spectrometry is usually capable of three basic operations: I) control of the mass spectrometer; II) acquisition and processing of data from the mass spectrometer; and III) interpretation of data. The computer can control the mass spectrometer by introducing the values and variations of different parameters. Because the computer processes digital data, whereas the mass spectrometer produces and receives analogic data, an interface is necessary to convert one type of data in another. A computer dedicated to mass spectrometry records the data given out by the mass spectrometer and converts them either into values of masses and peak intensities, or into total ionic current, temperatures, acceleration potential values, and so on. It is also capable of data processing. It allows calculation of average spectra and subtraction of one spectrum from another in order to eliminate the background noise or simply emphasise the differences between two spectra (de Hoffman and Stroobant, 2007). I.5.3. Protein identification There are two major methods that are widely used for protein identification by mass spectrometry (MS): Peptide mass fingerprinting and Tandem MS. I.5.3.1. Peptide mass fingerprinting Peptide mass fingerprinting (PMF) is an analytical technique for protein identification in which proteins are first digested using a site-specific proteolytic enzyme. The masses of the resulting peptides are then determined by MS. Since each protein has a different sequence, the masses obtained for each protein are an unique “fingerprint”. Protein identification is performed by comparing the experimentally determined peptide masses with theoretically determined peptide masses generated from protein sequences in databases by means of mass search programmes. Scoring I. Introduction 68 systems are used to rank proteins, whereby the high-ranking proteins from the database have the largest numbers of peptides in common with the protein that has been analysed. Scoring systems for PMF are critical, and should take into account many factors to produce a robust score. These factors include dissimilarities in the peptide masses due to calibration errors, expected peak intensities, noise, contaminant or missing peaks, presence of post-translational modifications… (Hernández et al., 2007). The PMF approach is rapid and efficient, but it has some limitations when: - Samples contain a mixture of proteins. The complexity of such spectra can result in false positives identifications. - MS spectra are searched against large sequence databases. As the specificity of the method is based on statistics, the larger database, the higher number of randomly matched peptide masses. - The proteins carry unexpected modifications reducing the number of matching peptide masses. - The proteins under analysis are very small or very large. Very small proteins produce very small number of peptides to be analysed. It is possible that these few peptides might not be present in a MS spectrum. In the case of large proteins, the number of theoretical peptides is so large that a portion of them are likely to randomly match nearly every spectrum. - The protein sequence under investigation is not represented in the protein sequence database (Hernández et al., 2007). I.5.3.2. Tandem mass spectrometry Protein identification from tandem mass spectra is one of the most versatile and widely used proteomics workflows, able to identify proteins, characterise posttranslational modifications, and provide semi-quantitative measurements of relative protein abundance (Edwars, 2011). MS/MS is widely used as an alternative to PMF. As a number of peptides are usually fragmented for each protein, the identification is more robust and less equivocal; it is more robust because there is no need to identify all peptides of a given protein to achieve confident identification; it is less equivocal because the identification of several peptides for a given protein confirms its presence in the sample (Hernández et al., 2007). Tandem mass spectra contain structural information related to the sequence of the peptide, rather than only its mass; these searches are generally more specific and discriminating (Mann et al., 2001; ChicanoGálvez, 2010). I. Introduction 69 I.5.3.2.1. The peptide fragment fingerprinting approach The principle of protein identification by peptide fragment fingerprinting (PFF) is similar to that of PMF. The aim is to correlate an experimental MS/MS spectrum with virtual MS/MS spectra constructed from the theoretical digestion of proteins to peptides and fragmentation of its peptides. A matching score is calculated that depends on the correlation between the experimental spectrum and the virtual spectrum of the peptide being compared (Hernández et al., 2007). The main advantages of this method are the higher discrimination capacity and the analysis of proteins mixtures allowing the use of massive sequencing techniques (Chicano-Gálvez, 2010). The computational analysis typically starts with the identification of the peptides that give rise to de acquired MS/MS spectra. In high-throughput studies, the most efficient peptide identification method is based on searching MS/MS spectra against protein sequence databases such as Sequest and Mascot (Nesvizhskii, 2007). I.5.3.2.2. De Novo Sequencing The growth of new sequenced species is simplifying the task of determining the primary structures of peptides and proteins, because open reading frames in the nucleotide sequence serve as templates for the construction of the corresponding proteins (Standing, 2003). The masses of the peptides produced by proteolytic digestion of an unknown protein can be compared with those predicted to arise from each protein in the database. This is often sufficient to identify a protein whose fulllength sequence is contained therein. But if the genome sequence of an organism is still unknown or the identification through PMF, or PFF does not work, the complete characterisation of the protein primary structure or de novo sequencing is required (Mann et al., 2001; Aebersold and Mann, 2003). De novo sequencing requires tandem mass spectrometry (MS/MS) in order to determine the order of the aminoacid sequences in a peptide. This includes accurate measurements of low mass ions, such as immonium ions. In this technique, a given parent (precursor) ion is selected in one mass spectrometer and then broken up, usually by collisions. The m/z values for the resulting daughter (product) ions are measured in a second mass spectrometer. The advantage of the de novo sequencing approach over the database search method is that it allows identification of peptides whose exact sequence is not present in the searched sequence database. However, de novo analysis is computationally intensive and requires high-quality MS/MS spectra. Peptide sequences extracted from MS/MS spectra using de novo algorithms need to be matched against the sequences of known proteins present in the sequence databases. In the case of organisms with nonsequenced or only partially sequenced genomes, the database search approach will fail to assign correct peptide sequences to many MS/MS spectra, thus using de novo sequencing tools becomes necessary (Liska and Shevchenko, 2003). I. Introduction 70 I.5.4. Sequence database searching For some organisms, multiple sequence databases are available (Apweiler et al., 2004). The most common used databases are: the National Center for Biotechnology Information (NCBI) Entrez Protein database, the NCBI Reference Sequence (RefSeq) database, that provides a non-redundant collection of sequences representing genomic data, transcripts and proteins (Pruitt et al., 2005), and The Universal Protein Source (UniProt) (consisting of Swiss-Prot, which is manually annotated and reviewed and its supplement TrEMBL, which is automatically annotated and is not reviewed) (Nesvizhskii, 2007). Sequence alignment has become the central tool for sequence comparison in molecular biology. In bioinformatics, a sequence alignment is a way of arranging the primary sequences of DNA, RNA or protein to identify regions of similarity that may be a consequence of functional, structural or evolutionary relationships between the sequences (Lesk, 2002). Aligned sequences of nucleotide or amino acid residues with identical or similar characters are aligned in successive columns. Very short of similar sequences can be aligned by hand. Most interesting problems require the alignment of lengthy, highly variable or extremely numerous sequences that cannot be aligned solely by human effort. Instead, human knowledge is primarily applied in constructing algorithms to produce high-quality sequence alignments, and occasionally, in adjusting the final results to reflect patterns that are difficult to represent algorithmically. A lot of computational algorithms, such as Basic Local Alignment Search Tool (BLAST) and Fast Alignment (FASTA) have been applied to sequence alignment problem, including slow but formally optimising methods like dynamic programming or probabilistic methods designed for large-scale database searches. Due to the high speed, sensitivity and the current availability of up-to-date sequence database, BLAST is the most widely used computer programme for database searches that has been developed; on the contrary, FASTA is much lower (Henikoff and Henikoff, 1992) but it is more suitable for nucleic acid sequence searches (Oladede et al., 2009). Data base searching is an essential element of large-scale proteomics. Eng et al. (1994) presented their search engine Sequest; the approach of this programme is based on cross-correlation between the fragment ion spectrum and the predicted list of m/z values of predicted ions for each potential peptide (Sadygov et al., 2004). Since then, several search engines have been developed based on Sequest software like the MS/MS-based search engine Mascot (www.matrixscience.com) which assigns a probability-based score and an expectation value to each search hit (Creasy and Cotrell, 2002). Three main types of sequence databases are suitable for searching mass spectrometric data. Non-redundant protein databases (nrdb) contain the known set of full-length protein sequences, extracted from the major sequence repositories and purged of duplicates. The nrdb maintained at the European Bioinformatics Institute I. Introduction 71 (EBI) allows consulting more than one hundred different databases and have several analysis tools of biological information as Blast and Fasta. The most popular protein databases are SwissProt and Uniprot Knowledgebase together with TrEMBL; they are accessible all together in the European Bioinformatics Institute website (www.ebi.ac.uk /Tools/sss/ncbiblast/) or in Uniprot (www.uniprot.org). The main server of SwissProt and TrEMBL is Expasy (www.expasy.ch), it is specialised in proteins and offers a high number of tools for sequence, structure and function analysis. Expressed sequence tag (EST) databases, such as dbEST at the National Center for Biotechnology Information (NCBI), contain millions of short one-pass sequences from random sequencing of cDNA libraries. These can be searched with appropriate software, usually by translating into the open reading frames. Genome databases can also be searched with MS data. The advantages of searching databases of completely sequenced genomes are that each peptide must be present by definition and that often the MS data can help to define the structure of the gene, such as start and stop and intron-exon structure (Mann et al., 2001). I.5.5. Proteomic applications for molluscs and parasites Proteomic analysis could provide an integrated “snapshot” of the functional proteins that can change in the course of all kinds of biological events, such as development, evolution and pathogenicity (Chen et al., 2011). This technology has been used in the study of molluscs for several applications, from ecological adaptation to search for biomarkers in contamination events or disease infection. López et al. (2001) detected differences in proteins expression between intertidal and cultured mussels, the intertidal mussels expressed more heat shock proteins than those cultured from rafts. Fuentes et al. (2002) found a lower viability of hybrid mussels between Mytilus edulis and M. galloprovincialis than “pure populations”; the lower viability was associated with higher parasitisation and lower expression level of stress proteins. Biomonitoring of aquatic environment and assessment of ecosystem health play essential roles in the development of effective strategies for the protection of the environment, human health and sustainable development. The growing interest in the application of proteomic technologies to solve toxicological issues and its relevance in the search for proteins involved in toxicological responses as biomarkes makes the proteomics a highly widespread tool (Monsinjon and Knigge, 2007; Lemos et al., 2010; Campos et al., 2012). Using heavy metals as indicator of pollution, changes in protein expression between clean and polluted areas were found. These changes disappeared when metal content was normalized, thus demonstrating that proteomics is an excellent tool for the search of biomarkers in invertebrates (Rodriguez-Ortega et al., 2003; Romero-Ruiz et al., 2006). Effects of Cadmium, an environmental stressor due to its toxicity, were measured by 2-DE in tissues of R. decussatus, the results showed a I. Introduction 72 diminution in protein expression of the affected tissues, especially in proteins involved in cytoskeletal maintenance (Chora et al., 2009). The level expression of heat shock proteins (hsp 60, 70 and 90 kDa) in response to the contaminant p,p´- dichlorodiphenyldichloroethylene (DDE) was studied in tissues of R. decussatus by immunobloting. The results showed a tissue-specific formation of reactive oxygen species in clams (Dowling et al., 2005). As well as in clams, toxicological effects were measured by 2-DE in abalones Haliotis diversicolor supertexta exposed to the endocrine disruptor bisphenol-A (Zhou et al., 2010), and in oysters Crassostrea angulata following the bioaccumulation of Hg in the food chain, suggesting that protemics would be further developed in application research of food safety (Zhang et al., 2013). The effects on marine invertebrates of toxic molecules produced by dinoflagellates and other microalgae can be evaluated by proteomic approaches. During seasonal harmful algal blooms many filter-feeding invertebrates can accumulate phycotoxins at extremely high levels representing a serious threat to human health. Proteomic analysis can be used for the detection and identification of biomarkers of biotoxin contamination and components that participate in the tissue response to the exogenous contaminant (Ronziti et al., 2008; Manfrin et al., 2012). As in human diseases, proteomics was used to detect mollusc proteins expressed in the host-parasite interaction, in order to identify new proteomic markers of disease resistance. That is the case of superoxide dismutase-like molecules from the oyster Saccostrea glomerata in the infection with the protozoan parasite Marteilia sydneyi, the causative agent of QX disease (Simonian et al., 2009a, b). Cao et al. (2009b) found differences in haemolymph protein expression between Ostrea edulis infected and non-infected with the protozoan parasite Bonamia ostreae; differences in the protein expression between Ostrea edulis (susceptible to infection with B. ostreae) and Crassostrea gigas (resistant to infection) were also recorded. Viral necrosis infection in Chlamys farrery was also studied by 2DE; the immune response of haemocytes against this infection was analysed and 48 proteins associated with immune response were reported (Chen et al., 2011). More recently, a similar approach developed by Castellanos-Martínez et al. (2014) addressed the immune response of the cephalopod Octopus vulgaris against the coccidian parasite Aggregata octopiana. Several proteins related to immune response were identified both in haemocytes and plasma. These kind of studies are particularly useful for a better comprehension of immune response of the host against parasites, especially in non-model species providing information for subsequent studies. Proteomics have been used to reveal protein expression of molluscs after challenge with bacteria as well. The scallop Chlamys farreri showed a differential expression of 27 proteins, most of them related with immune expression, after challenge with Vibrio harveyi (Huan et al., 2011). In mussels, the response after induction with Micrococcus luteus and Vibrio anguillarum I. Introduction 73 was measured in gills (Ji et al., 2013) and hepatopancreas (Wu et al., 2013), showing a differential protein expression after exposition with bacteria; Vibrio anguillarum provoked an increase of oxidative stress and disturbance in energy metabolism while M. luteus caused immune stress and disturbances in signaling pathways and protein synthesis. In the case of parasites theirselves, the study of protein expression of several parasites of the Apicomplexa group have been developed by 2DE. Proteomics analysis of the life cycle could help to settle basis for understanding intracellular and extracellular survival, invasion of host cell membranes, and evasion of host immune systems (Florens et al., 2002; Lasonder et al., 2002; Bautista et al., 2014). Proteomic maps of parasites such as Toxoplasma gondii (Cohen et al., 2002), Neospora caninum (Regidor-Cerrillo et al., 2012), Plasmodium falciparum (Lasonder et al., 2002; Vincensini et al., 2005) and Leishmania spp. (Drummelsmith et al., 2003; Walker et al., 2006; Cuervo et al., 2007; Vergnes et al., 2007) have been developed, some of them in different stages of the life cycle of the parasite, in order to get a snapshot of the parasite infection progress (Florens et al., 2002). Studies comparing species of the same genus or attenuated vs virulent strains allow the search for potential targets for drug design and provide a better knowledge of mechanisms associated with virulence. I.5.6. Functional genomics and proteomics The term proteome was first coined to describe the set of proteins encoded by the genome (Wilkins et al., 1996). The proteome is much more dynamic than the genome. Likewise, the term “proteome” led to a new research field called proteomics, defined as the high-throughput study of the proteome, including protein quantification, protein-protein interactions, post-translation modifications (PTMs) and protein function (Jensen, 2006; Schrimpf et al., 2009). The PMTs affect the structure, locations, function, and exchange, and affect activation and regulation functions, response to the environment, etc., and also are critical for control of the protein degradation processes (Gygi, 1999). Therefore, many proteins are present in multiple molecular forms. This essential information can only be determined by studying proteins and not genes. Besides, there is no direct correlation between the levels of RNA expression (transcription) and protein expression (Gygi, 1999). A short half-life messenger RNA (mRNA) can generate a long half-life protein, and vice versa. When the protein is present in the cell (and can be detected) the mRNA already cannot be detected. The expression of mRNA does not reveal the activity of the protein that it encodes or its possible combinations or interactions with other proteins that generate new functions, etc (Abbott, 1999). Post-translational modifications generate tremendous diversity, complexity and heterogeneity of gene product and their determination is one of the main challenges in proteomics research. Recent developments in mass spectrometry-based approaches for systematic, qualitative and Evaluation of the genetic variability of P erkinsus olseni 81 III.1. ABSTRACT Twelve microsatellite markers were used to characterize 130 clonal cultures of Perkinsus olseni derived from 30 clams from six different geographic locations. Only two loci were polymorphic in the four populations studied from Spanish coast (mean sample size = 31.2), and a third locus was variable in only two populations. In contrast, five parasites isolated from five clams from Japan and New Zealand showed variation at nine loci. Low genetic variation (2.08 ± 0.64 alleles per locus; mean genetic diversity: 0.101 ± 0.022), and very high FIS values (0.857 on average) were observed in Spanish populations. A total of 39 multilocus genotypes (MLGs) were identified in the whole sample (121 clonal isolates after excluding incomplete MLGs due to missing data). A three-level hierarchical analysis of molecular variance found significant levels of genetic variation within infrapopulations (all the parasites in a single host; ØIS = 0.679) and among infrapopulations within the component population (all the parasites among a host population; ØSC = 0.579). Differences among the component population from different geographic locations were not significant (ØCT = 0.057). These results suggest that an important fraction of FIS is explained by the Wahlund effect, but also strong inbreeding within infrapopulations. Another explanation for the high FIS within infrapopulations is the presence of haploid and diploid stages in the clam. Although fully aquatic system provides many opportunities for mixing of parasites from different clams, results are consistent with the consideration of all P. olseni in a clam as a cohesive genetic unit (i.e., deme). If the parasite was introduced into the Spanish coast with the importation of infected clams from Asia and Oceania, the low microsatellite polymorphism could be reflecting founder effects in the recent evolutionary history of P. olseni. The loss of alleles would be intensified in a scenario structured in numerous demes because of recurrent founder effects at microgeographic level. Evaluation of the genetic variability of P erkinsus olseni 83 III.2. INTRODUCTION The protozoan Perkinsus olseni is a parasite infecting a wide variety of bivalves around the world. It has been associated with extensive mortalities of commercially important species including oysters, clams, cockles and abalones (Villalba et al., 2004). The species was originally described in Australia (Lester and Davis, 1981) and is common in the Asian-Pacific region (Choi and Park, 2010). It has also been reported in Portugal (Azevedo, 1989), Spain (Robledo et al., 2000; Casas et al., 2002a), Italy (Abollo et al., 2006), France (Arzul et al., 2012), and Uruguay (Cremonte et al., 2005). Because of its significant impact on aquaculture, the infection by P. olseni is included in the OIE (Office International des Epizooties) list of diseases to be notified in order to prevent its spread. On the basis of its geographic distribution it has been hypothesized that the presence of the parasite in the Atlantic and the Mediterranean is due to imports of infected hosts from Asia and Oceania. Despite substantial economic losses caused by Perkinsus species, little is known about relevant population parameters such as genetic variability, patterns of dispersal, population genetic structure, and reproductive strategies. However, this knowledge is essential for the efficient management of the parasitic disease (Tibayrenc and Ayala, 2002). The trophozoite of P. olseni occurs in clams as a single cell which undergoes vegetative multiplication. The trophozoite adopts a multi-nucleated form and gives rise to new cells. Once the host dies, trophozoites in the anaerobic condition provided by the necrotic tissue enlarge and transform into hypnospores, a resistant dormant stage with a thick external wall. When the hypnospores are in contact with aerated seawater at appropriate temperature and salinity, they undergo zoosporulation and thus the occurrence of new successive cell bipartitions within the original hypnospore wall and the generation and releasing of multiple cells of motile bi-flagellated zoospores capable of infecting new hosts (Villalba et al., 2004; Choi and Park, 2010). It is unknown if the life-cycle includes meiosis, fusion of gametes and zygote formation. Because clonal amplification occurs both within and outside the host it is expected that clonal propagation shapes the population genetic structure. Therefore, populations of P. olseni may be characterized by different manifestations of clonality, such as linkage disequilibrium, fixed heterozygosity within loci, deviation from Hardy–Weinberg expectations towards a heterozygote excess, and the presence of overrepresented identical multilocus genotypes (Tibayrenc and Ayala, 2002; De Meeûs et al., 2006). The inference of clonal structure in organisms that may alternate clonal amplification and sexuality requires the definition of what constitutes a cohesive genetic unit (i.e., deme), because cryptic genetic structuring within the units defined as demes can both hide the expected excess of heterozygotes under clonality and overestimate the rate of sexual recombination (Halkett et al., 2005; Prugnolle and De Meeûs, 2010). For parasitic organisms, the deme could be represented by individual Chapter III 84 infrapopulations or the component population. The infrapopulation refers to all individuals of a parasite species in an individual host at a particular time and the component population refers to all infrapopulations at a given place and time (Bush et al., 1997). It is even possible that other different units constitute the deme, such as groups of infrapopulations (e.g., family groups of hosts, different host species…) and temporal groups distributed over several infrapopulations. Thus, the entity we call deme, an evolutionary unit characterized by certain persistence over time, may be very different depending on the life cycle of the parasite and the ecology of transmission (Criscione et al., 2005; Criscione and Blouin, 2006). Microsatellites are useful markers to investigate patterns of population genetic structure of parasitic protozoa. For instance, analyses of microsatellite variation revealed evidences of clonal structure in Sarcocystis (Asmundsson et al., 2006), Toxoplasma (Ajzenberg et al., 2004), Plasmodium (Razakandrainibe et al., 2005), and Trypanosoma (Simo et al., 2010). Microsatellite studies also revealed that Trypanosoma, Leishmania, Toxoplasma, Cryptosporidium, and Plasmodium, whose mode of reproduction has often been assumed to be predominantly clonal, actually display different population structures, ranging from clonality to panmixia (Anderson et al., 2000; Ajzenberg et al., 2004; Morrison et al., 2008; Koffi et al., 2009; Rougeron et al., 2009). Here, we carry out a study of genetic variation of P. olseni from the three main areas of clam production in Spain. We found very low microsatellite polymorphism, particularly in comparison with a sample from the Pacific area, which suggests strong genetic drift probably caused by recurrent founder effects. We also observed strong heterozygote deficits in the component population from different locations, indicating subdivision at some level below and suggesting that the infrapopulation constitutes the deme in P. olseni. III.3. MATERIALS AND METHODS III.3.1. Parasite isolates and microsatellite markers A total of 130 clonal cultures of P. olseni from six geographic regions were genotyped for 12 microsatellites. Five clonal cultures per individual host were analyzed from four locations through the Spanish coast: Ría de Arousa and Ría de Pontevedra in Galicia (NWSpain), Carreras River in Huelva (Andalucía, SW Spain), and Delta de l’Ebre in Catalonia (NE Spain) (Fig. III.1). Therefore, the three main areas of clam production in Spain (Galicia, western Andalucía and Catalonia) were included in the study. Five carpet shell clams (Ruditapes decussatus) from Arousa, four from Pontevedra, eight from Huelva, and eight Manila clams (Ruditapes philippinarum) from Catalonia were used to isolate P. olseni. The parasites were isolated from the gills and were allowed to proliferate in vitro as described by Casas et al. (2002b). The in vitro cultures (one per Evaluation of the genetic variability of P erkinsus olseni 85 host) were cloned by limiting dilution plating in 96-well culture plates, and five monoclonal derivatives of each isolate culture were expanded. Additionally, three P. olseni clonal cultures from three Manila clams collected in Japan and two more clonal cultures from two clams Austrovenus stutchburyi taken in New Zealand were used in the analyses. Parasites were previously identified as P. olseni by using a PCR-RFLP diagnostic assay of the rRNA internal transcribed spacer (ITS) region (Abollo et al., 2006). We studied the 12 microsatellite loci (PolUSC1-PolUSC12) previously described in P. olseni following the protocols for DNA extraction and microsatellite genotyping reported (Pardo et al., 2011). These markers confirmed that all Perkinsus isolates belong to a single species. The presence of Perkinsus mediterraneus, the other species reported from the Spanish coast (Casas et al., 2004), in the studied samples may be discarded because several markers suitable for distinguishing the two species were used (Pardo et al., 2011). Fig. III.1. Map showing the four Spanish locations where infected clams were collected: Ría de Arousa (A) and Ría de Pontevedra (P) in Galicia, Carreras River in Huelva, Andalucía (H) and Delta de l’Ebre in Catalonia (C). III.3.2. Data analysis Allele frequencies and estimates of genetic variation within populations (average number of alleles per locus, allelic richness, and heterozygosity) were calculated using FSTAT 2.9.3.2 (Goudet, 2001). After excluding clonal cultures with missing data (the number of isolates decreased from 130 to 121), genotypic diversity Chapter III 86 (G) was measured as the number of different multilocus genotypes (MLGs) over the total number of clonal cultures per geographic sample (n). All clonal cultures from one clam from Huelva showed incomplete MLGs due to missing data, so they were not used in the MLG analysis. Conformity to Hardy–Weinberg proportions was tested using exact test as implemented in GENEPOP 3.4 (Raymond and Rousset, 1995). Linkage disequilibrium between pairs of loci was tested with the randomization test based on the log-likelihood ratio G of genotypic frequencies from paired loci in contingency tables. Genotypes at two loci are associated at random a number of times and the loglikelihood G test statistic is recalculated on the randomized data set. However, tests were not possible in many cases due to the low microsatellite variation of P. olseni. We adjusted the P-values with the sequential Bonferroni correction. Linkage disequilibrium analyses were performed in FSTAT. Population structure was inferred by using the Weir and Cockerham’s (1984) unbiased estimators of Wright F-statistics (Wright, 1965) defined for three hierarchical levels (individuals within geographical populations, geographical populations within metapopulation, and metapopulation): FIT measures an overall inbreeding combining the homozygosity of individuals within populations relative to that measured between individuals (FIS) and the homozygosity caused by the Wahlund effect when several differentiated populations are treated as a single panmictic unit (FST). These statistics can be translated into biologically characteristics relevant to molecular epidemiology (De Meeûs et al., 2007). For instance, the three Fstatistics equals zero in a single panmictic population. Although the sample size per infrapopulation is small (only a maximum of five clonal cultures of P. olseni per clam), we calculated the mean FIS among infrapopulations because the number of polymorphic infrapopulations is relatively high. F-statistics were calculated with FSTAT and their significant deviation from zero was tested by randomizing alleles between individuals within subsamples for FIS and randomizing individuals among subsamples for FST. Randomizations (10,000) were done in FSTAT. The hierarchical distribution of genetic variation was also characterized using analysis of molecular variance (AMOVA). Three-level AMOVA was conducted in ARLEQUIN 3.0 (Excoffier et al., 2005) using an FST-like estimator under the infinite alleles model (IAM; Weir and Cockerham, 1984) and an RST-like estimator under the stepwise mutation model (SMM; Michalakis and Excoffier, 1996), since results may vary according to the assumptions of different mutation models (Balloux and Goudet, 2002; Balloux and Lugon-Moulin, 2002). The analysis was carried out to assess the amount of variance imputable to genetic differences between geographic populations (ØCT), among infrapopulations within geographic populations (ØSC), among individuals within infrapopulations (ØIS), and within individuals (ØIT). One thousand permutations were generated to assess whether levels of differentiation were significantly greater than zero (Excoffier et al., 2005). Genetic differentiation between geographical samples was also estimated with Dc genetic distance (CavalliSforza and Edwards, 1967), calculated using POPULATIONS 1.2.26 (Langella, 2002). A Neighbor-joining (NJ) tree of individuals (clonal isolates) Evaluation of the genetic variability of P erkinsus olseni 87 derived from shared allelic distance (Chakraborty and Jin, 1993) was built with the program POPULATIONS. The tree was visualized using TREEVIEW 1.4 (Page, 1996). III.4. RESULTS III.4.1. Genetic variability Low genetic variation was detected in P. olseni samples from the Spanish coast. The average number of alleles per locus was 2.08 ± 0.645, ranging from one to eight, and the mean genetic diversity in Spanish populations was Hs = 0.101 ± 0.022 (Table III.1).The number of polymorphic loci within populations (0.95 criterion) ranged from two (Pontevedra and Huelva) to three (Arousa and Catalonia), and the expected heterozygosity of polymorphic loci ranged from 0.025 to 0.711 (Table III.2). Samples from the Pacific Ocean (three clonal isolates from Japan and two from New Zealand) showed variation at nine loci. When the five parasites from the Asian-Pacific region were included in the analysis the average number of alleles per locus increased until 3.25 ± 0.733. This increase indicates high genetic variation and differentiation of these parasites with respect to those from the Spanish coast. After excluding individuals with missing data, a total of 39 distinct MLGs were identified in the whole data set (121 clonal cultures). Genotypic diversity within populations was low in Pontevedra and Huelva and moderate in Arousa and Catalonia (TableIII.1). The five MLGs from Japan and New Zealand were unique. Table III.1. Genetic variability in geographic populations of P. olseni. Sample size (N), sample size after excluding incomplete multilocus genotypes because missing data (n), Nei’s unbiased genetic diversity (Hs), average number of alleles (A), mean allelic richness per locus and population (AR), number of different complete multilocus genotypes (MLGs) and genotypic diversity (G). The number of individual hosts is shown in parentheses. N Hs A AR n MLGs G Arousa 25 (5) 0.128 1.58 1.56 25 (5) 13 0.52 Pontevedra 20 (4) 0.062 1.42 1.42 20 (4) 7 0.28 Huelva 40 (8) 0.067 1.42 1.36 32 (7) 7 0.22 Catalonia 40 (8) 0.149 1.83 1.69 39 (8) 16 0.41 Japan 3 (3) - 1.50 - 3 (3) 3 1.0 N. Zealand 2 (2) - 1.44 - 2 (2) 2 1.0 Chapter III 88 Table III.2. Genetic diversity at three polymorphic loci of 130 clonal isolates of P. olseni from Spain. Number of alleles (A), the total number of alleles per locus is shown in parentheses; expected heterozygosity (H) and deviation of the observed genotypic frequencies from the Hardy-Weinberg expectations (FIS). All FIS values except for PolUSC8 in Pontevedra are statistically significant (P < 0.001) Locus Arousa Pontevedra Huelva Catalonia Mean ± SE PolUSC7 A (8) 3 5 4 6 H 0.530 0.555 0.523 0.711 0.580 ± 0.044 FIS 1.000 0.820 0.940 0.856 0.904 ± 0.040 PolUSC8 A (5) 4 2 2 4 H 0.585 0.145 0.262 0.594 0.396 ± 0.114 FIS 0.863 0.655 1.000 0.832 0.837 ± 0.071 PolUSC4 A (3) 3 1 2 3 H 0.418 0.000 0.025 0.479 0.230 ± 0.109 FIS 0.801 - - 0.945 0.873 0.072 III.4.2. HWE deviations and genotypic disequilibrium There was a large deficit of heterozygotes within populations compared with genotypic frequencies predicted from Hardy–Weinberg expectations (Table III.2). The population FIS for all loci were: 0.893 (Arousa), 0.786 (Pontevedra), 0.931 (Huelva), and 0.872 (Catalonia). The overall mean value was 0.857 ± 0.043 (Table III.2). All FIS values except PolUSC8 in Pontevedra (P = 0.085) differed from the null hypothesis (FIS = 0; P < 0.001). The mean FIS value within infrapopulations was 0.502 ± 0.087. Given the low variation detected in the four Spanish populations, testing random association of genotypes at different loci was only possible for 10 pairs of loci. None of these pairs show significant linkage disequilibrium. III.4.3. Distribution of MLGs within and among hosts Only one MLG (10) was found in all four Spanish populations reaching a frequency of 0.269 in the whole Spanish sample. The remaining MLGs did not exceed the 0.1 frequency (Table III.3). In 79.2% (19 out of 24) of clams from Spain, more than Evaluation of the genetic variability of P erkinsus olseni 89 one parasite MLG was found among the four or five genotyped, and in 29.2% of them, three or more MLGs were found. These figures are high considering the low number of clonal isolates per clam (mean = 4.65) and the low genetic variability detected. The 66.7% (16 out of 24) of clams from Spain harbored at least one MLG not present in any other clam. The corresponding values within populations ranged from 28.6% (two out of seven) in Huelva to 100% (five out of five) in Arousa. The five clams from the Pacific area harbored unique MLGs. The 82% MLGs (32 out of 39) appeared in a single clam and the most frequent MLG was only present in 27.6% of clams (Table III.3). These results show a high level of genetic structure among infrapopulations. III.4.4. Population structure and differenciation Genetic structure among geographic populations from Spain was moderate but significant (FST = 0.137, P < 0.001). The large difference between the FST and the overall measure of deviation from panmixia (FIT = 0.897) is consistent with high levels of inbreeding (FIS = 0.881). All pairwise FST values except the one between Pontevedra and Huelva were significant (P < 0.008; Table III.4). Absolute genetic distances between geographical populations closely reflect FST values (Table III.4). No correlation was detected between genetic and geographical distances. In fact, the closest geographical populations (Pontevedra and Arousa) were genetically more related to Huelva and Catalonia, respectively (Fig. III.1; Table III.4). These results do not support isolation by distance model in Spanish coast. The Neighbor-joining tree based on shared allelic distances between MLGs shows that (1) the parasites from the Pacific area are highly differentiated, particularly those from New Zealand; and (2) differentiation between individuals from Spain is not associated with their geographic origin (Fig. III.2). The three-level hierarchical AMOVA showed significant levels of genetic differentiation between infrapopulations within geographic populations, between individuals within infrapopulations, and between individuals under IAM and SMM assumptions (Table III.5). After removal the variation attributable to different hierarchical levels including differences among individuals within infrapopulations, differentiation among geographic populations was not significant. The proportion of genetic variation was distributed similarly at all levels under both mutation models (although it was distributed more evenly across levels under the SMM). The largest proportion of variation was explained by differences among infrapopulations within geographic populations and to a minor extent between individuals within infrapopulations. Fixation indices confirmed that greater variation was observed among infrapopulations rather than between geographical populations (Table III.5). Chapter III 96 drift thus explaining the low number of alleles observed. Further studies aimed to identifying ploidy levels of P. olseni within clams and additional population genetic structure analysis involving larger sample sizes within infrapopulations are needed to estimate the rate of clonal reproduction and to determine whether cryptic structure is the main cause of the high homozygosity observed. IV. VARIABILITY OF THE CELL PROTEOME OF Perkinsus olseni AMONG REGIONS OF THE SPANISH COAST - The content of this chapter has been accepted for publication in: Fernandez-Boo, S., Villalba, A., Cao, A. In press. Variability of the cell proteome of Perkinsus olseni, protistan parasite of molluscs, among regions of the Spanish coast. Dis. Aquat. Organ. Variability of the cell proteome of Perkinsus olseni 99 IV.1. ABSTRACT The variability of the proteome of in vitro cultured Perkinsus olseni cells deriving from 4 regions of the Spanish coast was evaluated. The regions involved were: the Rías of Arousa and Pontevedra (Galicia, NW Spain), Carreras River in Huelva (Andalusia, SW Spain) and Delta de l’Ebre (Catalonia, NE Spain). P. olseni in vitro clonal cultures were produced starting from parasite isolates from four individual clams from each region. Those clonal cultures were used to extract cell proteins, which were separated by 2D electrophoresis. Qualitative comparison of P. olseni protein expression profiles among regions was performed with PD Quest software. Around 700 protein spots from parasites derived from each region were considered, from which 141 spots were shared by all the regions. Various spots were found to be exclusive of each region. Higher similarity was found among the proteome of P. olseni from the Atlantic regions than between the Mediterranean location and the Atlantic ones. A total of 54 spots were excised from the gels and sequenced. Nineteen proteins were annotated after searching in databases, 13 being shared by all the regions and 6 exclusive of one region. Most identified proteins were clustered into glycolysis, oxidation/reduction process, metabolism and response to stress. From proteins set analysed, no direct evidence of P. olseni variability associated with virulence was found although the differences in metabolic adaptation and stress response could be connected to pathogenicity. Variability of the cell proteome of Perkinsus olseni 101 IV.2. INTRODUCTION The genus Perkinsus includes protistan parasites infecting a wide range of molluscs over the world (Villalba et al., 2011). The species Perkinsus olseni was first described infecting abalones Haliotis ruber in Australia (Lester & Davis, 1981). Since then, P. olseni has been reported from gastropod and bivalve molluscs of Australasia, Europe, Asia and South America (Villalba et al., 2011). The name Perkinsus atlanticus, given to a parasite of carpet-shell clams Ruditapes decussatus from Portugal (Azevedo 1989), was considered synonym of P. olseni (Murrell et al., 2002). Likely, P. olseni was introduced into Europe by commercial trade of Manila clams Ruditapes philippinarum from Asia coasts (Hine, 2001, Elandaloussi et al., 2009a); it has spread through Italy (Da Ros and Canzonier, 1985, Abollo et al., 2006), Portugal (Azevedo, 1989), Spain (Casas et al., 2002a, Elandaloussi et al., 2009b) and France (Goggin, 1992; Arzul et al., 2012). Most studies on the genus Perkinsus have been devoted to P. marinus, which has been causing mass mortalities in Eastern oyster Crassostrea virginica populations of the Atlantic and Gulf coasts of the USA for more than 50 years (Ray, 1996; Andrews, 1996; Burreson and Ragone Calvo, 1996). The World Organisation for Animal Health (OIE) has included P. marinus and P. olseni in the list of notifiable diseases (http://www.oie.int/en/animal-health-in-the-world/oie-listed-diseases-2014/). Variability in virulence and other physiological aspects through populations of P. marinus has been demonstrated (La Peyre et al., 1995a; Bushek and Allen, 1996; Chu and Lund, 2006), which is consistent with the occurrence of genetic variability among geographic strains (Reece et al., 1997; Kotob et al., 1999; Reece et al., 2001; Thompson et al., 2011). The existence of races that vary in virulence or environmental tolerance has important management implications and spreading virulent races should be avoided. Spreading races with varying environmental tolerances can also be harmful by producing epidemic outbreaks in areas that are inhospitable to indigenous parasitic races (Bushek and Allen, 1996). Analysis of protein profiles is another way to search for variability and could be valuable for understanding biological processes including development, evolution and pathogenicity of these organisms. Proteomics involves the systematic analysis of gene expression at a protein level (Sperling, 2001; Lee et al., 2005). The proteomic techniques are powerful tools to compare different isolates of the same species (Shin et al., 2005; Regidor-Cerrillo et al., 2012) morphospecies (Chan et al., 2005), genetic variability (Mosquera et al., 2003), and differences in virulence among isolates or strains of the same species (Regidor-Cerrillo et al., 2012). A proteomic approach using two-dimensional electrophoresis (2-DE) coupled to Mass Spectrometry (MS) was performed to compare the protein expression profiles between P. olseni clonal cultures deriving from four regions of the Spanish coast, as a Chapter IV 102 way to search for variability, which could be the basis of geographical differences in virulence and environmental adaptation of P. olseni. IV.3. MATERIALS AND METHODS IV.3.1. Production of in vitro clonal cultures. Parasites Perkinsus olseni were isolated from 4 carpet shell clams R. decussatus from Ría de Arousa, 4 from Ría de Pontevedra in Galicia (NW Spain), 4 from Carreras River in Huelva (Andalucía, SW Spain) and from 4 Manila clams R. philippinarum from Delta de l´Ebre in Catalonia (NE Spain) (Fig.IV.1). The parasites were isolated from the gills and were allowed to proliferate in vitro as described by Casas et al. (2002b). The in vitro cultures (one per host) were cloned by limited dilution plating in 96-well culture plates (Casas and La Peyre, 2009). One monoclonal derivate of each isolate was in vitro expanded and used in the study. After 2 months, when the cultures were in exponential growth phase, parasites of each clonal culture were harvested; the viable cells were counted by staining with 50 mg/L neutral red and counting in a Malassez chamber. After centrifugation (800 x g, 10 min, 25 ºC), 5 x 106 cells were reseeded and the remaining cells were stored at -80 ºC for proteomic analysis. This process was repeated 5 more times up to collecting 150 x 10 cells of each clone. Species identification of the cells in each clonal culture was performed by PCR followed by restriction fragment length polymorphism assay as described by Abollo et al. (2006), thus confirming that the cells of every clon corresponded to P. olseni. IV.3.2. Protein extraction. A total of 150 x 106 frozen cells from each clonal culture were resuspended in 1 mL of lysis buffer (8M urea, 2M thiourea, 2% CHAPS, 1% DTT, 0.8% ampholites pH 310) during 2h 30 min for protein extraction. The protein concentration was determined by Lowry assay using the RC/DC Protein Assay Bio-Rad and measuring in a microplate lecture Expert 96 (Asys Hitech). Then 400 µg of protein were purified using the 2D CleanUp kit Bio-Rad and resuspended in 1 mL of rehydration solution (7M urea, 2M thiourea, 4% chaps, 0.3% DTT, 0.5% IPG buffer and bromophenol blue traces). Variability of the cell proteome of Perkinsus olseni 103 Fig. IV.1. Map showing the four Spanish locations where infected clams were collected: Ría de Arousa (A) and Ría de Pontevedra (P) in Galicia, Carreras River (C) in Huelva (Andalusia) and Delta de l’Ebre (D) in Catalonia. IV.3.3. Two dimensional electrophoresis (2DE) and image analysis. Preliminary trials had been performed to tune up the experimental conditions for separation of P. olseni proteins by 2DE. Each sample was processed in quadruplicate. For the first dimension, aliquots of 100 µg of protein were diluted to a final volume of 250 µL in rehydration solution. Samples were centrifuged for 1 min at maximum speed to remove bubbles and loaded onto the immobilised pH gradient (IPG)-strips (11cm, pH 5-8, Bio-Rad) by in-gel rehydration. Iso-Electro Focusing (IEF) was performed using a Protean®IEF System (Bio-Rad) at 20ºC, as follows: 50 V were applied during the rehydration step for 12 h, then the IEF proceeded through 5 steps, 150 V for 30 min, 300 V for 30 min, 1000 V for 60 min, 8000 V for 180 min and 8000 V until 35000 Vh. After IEF, the strips were incubated for 20 min in equilibration buffer (50 mM Tris–HCl pH 8.8, 6M urea, 30% glycerol, 2% SDS) with 10 mg/ml DTT to reduce the proteins and, in a second step, incubated for 20 min in this mix with iodoacetamide at 45 mg/ml to alkylate proteins. The second dimension involved separation in SDSPAGE, which was carried out at 15ºC through 12,5% polyacrylamide homemade gels, using a Protean®II xi cell (Bio-Rad) in two steps: 15 mA/gel for 15 min and 50 mA/gel until separation was finished (≈3:30 h). Molecular markers from 250 kDa to 10 kDa (BioRad) were run in the second dimension next to the sample for protein size determination. Four gels were produced from each clonal culture. The gels were silver stained with a protocol compatible with MS analysis: (1) incubation in 50% (v/v) methanol 5% (v/v) acetic acid for 1 hour or overnight; (2) incubation in 50% methanol Chapter IV 104 for 30 min; (3) wash with H2O milli-Q for 15 min (x 2); (4) incubation in 0.02% (w/v) sodium thiosulfate for 1 min; (5) wash with H2O milli-Q for 1 min (x2); (6) incubation in 0.1% (w/v) silver nitrate for 30 min; (7) wash with H2O milli-Q for 1 min (x3); (8) incubation in 0.04% (v/v) formaldehyde 2% (w/v) sodium nitrate until complete appearance of the spots; (9) incubation in 5% (v/v) acetic acid for 5 min to stop staining. Once stained, the gels were digitised with a GS-800 densitometer (Bio-Rad) and analysed with PD Quest V.7.4.0 software (Bio-Rad). Fig. IV.2 shows a scheme of the process of image analysis and comparison between geographic regions, which involved discarding the worst gel replicate from each clonal culture, thus using the best three gel replicates for the analysis. Highly reproducible replications were observed in gels of each clonal culture. The analysis of the gels was made region by region, producing a master gel of each region including just the spots shared by all the gels of the region; then the master gels of each region were compared to produce a final master gel in which spots shared by regions and spots exclusive of each region were discriminated. Percentages of similitude between gels were calculated as the number of common spots shared by two gels with regard to the total number of spots: PS = [C x 2 / (T1 + T2)] x 100 where PS is the percentage of similitude, C is the number of the common spots shared by the 2 regions, and T1 and T2 are the total number of spots in region 1 and 2, respectively. IV.3.4. Protein identification and database searching Nineteen spots shared by gels of every region and 35 spots that were exclusive of one region were selected for protein identification. The number of spots selected for identification was limited by funding availability, thus the selection criterion was to choose the most intense spots among the common and the region exclusive ones. The selected spots were excised manually within a laminar flow cabin with sterile scalpel blade. Protein identification by LC-MS/MS analysis and de novo sequencing were carried out in the LP-CSIC/UAB Proteomics laboratory (Barcelona, Spain), member of ProteoRed network. Details on the LC_MS/MS procedure are provided as “Supplementary Materials and Methods” (Annex I). All sequence tags obtained from de novo sequencing were manually confirmed and were submitted to a homology search using the pBLAST algorithm (NCBI, USA). Data were contrasted against nonredundant NCBI (National Center for Biotechnology Information, Maryland, USA) database using Alveolata as search category. Theoretical Mr and pI values were obtained with the compute pI/Mw tool at the ExPASy Proteomics Server (http://www.expasy.org/tools/pi_tool.html ). Variability of the cell proteome of Perkinsus olseni 105 Fig. IV.2. Scheme of gel analysis and comparison between regions. (A) Analysis of four Perkinsus olseni clonal cultures by triplicate from each region to obtain a master gel of each region including just the spots shared by all the gels. (B) Comparison among master gels of each region. (C) Final master gel discriminating the spots shared by regions and the spots exclusive of each region. IV.3.5. Statistical analysis In vitro proliferation of each P. olseni clon was measured as the number of cells counted in the clonal cultures after two months from starting culture with 5x106 cells. As explained above, 6 successive replicates of each clon were produced to harvest enough number of cells for proteomic analyses. Differences in in vitro proliferation were analysed by a two-factor nested ANOVA (geographical origin as main factor and clon nested under origin). The number of cells was root-squared transformed because of statistical requirements. As no significant effect due to clon was detected, the clones were grouped by origin and paired comparisons between origins were performed with Fisher’s test. Statistical analyses were performed using MINITAB 16 statistical software. The significance level was established at P ≤ 0.05. Chapter IV 112 Fig. IV.5. Fragments of master gels showing the spots exclusive of Perkinsus olseni from one region that were annotated. Left column corresponds to the areas of final master gel where the exclusive spots occur. The remaining columns correspond to the same areas in the master gel of each region. Each file corresponds to an exclusive spot labelled with the spot code (the same as in Table IV.3). EA: exclusive of Ría de Arousa; EC: exclusive of Carreras River, ED: exclusive of Delta de l’Ebre; EP: exclusive of Ría de Pontevedra. Variability of the cell proteome of Perkinsus olseni 113 Table IV.5. Peptides annotated after searching in Blast database against Alveolata group in non-redundant proteins database. Spot codes 1-13 correspond to spots shared by Perkinsus olseni from all the regions, “EA” corresponds to spots exclusively found in P. olseni from Ría de Arousa, “EC” to spots exclusive from Carreras River, “ED” to spots exclusive from Delta de l’Ebre and “EP” to spots exclusive from Ría de Pontevedra. pI theo and pI obs: the theoretical (Swiss Prot, www.expasy.ch/tools) value of isoelectric point and the one observed in gels, respectively. Mw theor and Mw obs: the theoretical molecular weight and the one observed in gels, respectively. Identified protein: protein name of identified peptides in the database. Organism: the species to which corresponds the protein recorded in the database. A.no.: accession number of proteins in the database. Biological function: function of the protein according to Gene Ontology Biological Process. Peptide sequence: peptide sequences with the highest individual ion scores indicating identity or extensive homology (p<0.05). Sco: Score of the match of peptides with proteins in NCBI database after BLAST searching; the higher the value of the score, the better the peptide match. E-value: expected value, which describes the number of hits expected to see by chance when searching a database of a particular size; the lower the E-value, the better the peptide match Spot code. pI obs/ pI theor Mw obs/ Mw theo (kDa) Identified protein; organism A. no. Biological function Peptide sequence Sco E value 1 6.25/6.00 62/56.55 Pyruvate kinase; Perkinsus marinus XP_002788066.1 Glycolysis VPSFQGTDHIIQSAINYGK 1080 0.0 2 6.15/8.35 54.5/17.08 Adenosylhomocysteinase; P. marinus XP_002781556.1 One carbon metabolic process NNAIVGNIGHFDNEIQmER 66.8 3,00 E-013 3 7.03/6.20 60/61.04 glucose-6-phosphate isomerase; P. marinus XP_002787676.1 Glycolysis FVAHIQQLDMESNGK 52.4 2,00 E-008 4 7.12/6.20 62/61.04 glucose-6-phosphate isomerase; P. marinus XP_002787676.1 Glycolysis FVAHIQQLDMESNGKR 55.8 2,00 E-009 5 7.12/6.04 54/47.33 Formate dehydrogenase ; P. marinus XP_002782414.1 Oxidoreductase DVEGMHLGTVA 32.5 0.057 6 6.50/5.23 49/48.63 Enolase 2; P. marinus XP_002785647.1 Glycolysis SGETEDTFIADIVVGLGTGQIK 70.2 2,00 E-014 7 6.70/5.23 50/48.63 Enolase 2; P. marinus XP_002785647.1 Glycolysis VNQIGSVTESIEANNK 52.4 1,00 E-010 8 6.50/5.82 48/44.77 Phosphoglycerate kinase 1; P. marinus XP_002778580.1 Glycolysis AGATSIIGGGDTAAmVEQQGK 65.5 2,00 E-012 9 6.05/5.58 32/25.05 Malate dehydrogenase; P. marinus XP_002784022.1 Carbohydrate metabolic process ImGLmSLDVTR 39.7 1.00 E-006 10 6.25/5.58 25/24.35 Conserved hypothetical protein; P. marinus XP_002776039.1 Unknown VEPATQP 21 0.76 11 7.05/5.86 22/21.55 Peroxiredoxin-2; P. marinus XP_002765329.1 Oxidation-Reduction process / Antioxidant activity VLDSIIETDEHGVVcPANWK 69.4 5,00 E-014 Chapter IV 114 Spot code. pI obs/ pI theor Mw obs/ Mw theo (kDa) Identified protein; organism A. no. Biological function Peptide sequence Sco E value 12 6.5/6.03 20.5/22.11 Iron-dependent superoxide dismutase; P. marinus XP_002788749.1 Superoxide metabolic process / Oxidation-Reduction process MSAETLQYHYG 37.1 1.00 E-002 13 5.58/6.28 17.5/17.00 Peroxiredoxin V; P. chesapeaki ABV22156.1 Oxidation-Reduction process / Antioxidant activity ALGVDFDVTPVLGNVR 52.4 2,00 E-008 EA1 5.65/5.39 81/81.53 Heat Shock Protein 90; P. marinus XP_002773238.1 Protein folding / Response to stress LVNSPAVVTSAISPQMR 55.8 2,00 E-009 EA2 6.50/5.23 47/42.8 Galactokinase; P. marinus XP_002788912.1 Galactose metabolic process SGQLVHLDcR 35.4 0.005 EA3 7.28/5.51 41.5/141.49 Conserved hypothetical Protein; P. marinus XP_002773742.1 Unknown SLLDQDPENAEIK 44.3 9.00 E-006 EP1 6.15/5.42 63.1/56.36 Pyruvate kinase; P. marinus XP_002788067.1 Glycolysis HVAVMLDTK 32.5 0.038 ED1 6.8/5.71 100/30.55 Malate dehydrogenase; P. marinus XP_002787252.1 Malate metabolic process LSLYVAcAGINPGR 46.9 1,00 E-006 EC1 6.85/5.79 36.5/36.38 Glyceraldehyde 3 phosphate dehydrogenase; P. marinus XP_002788755.1 Oxidation/Reduction process VIISAPPKDDTPMFVMGVNNK 72.3 7.00 E-015 Variability of the cell proteome of Perkinsus olseni 115 IV.5. DISCUSSION The results showed that there is a remarkable variability in the protein expression of in vitro cultured P. olseni cells depending on the geographic origin. Genetic and physiological variability depending on geographic origin had also been found in P. marinus (Reece et al., 1997; Kotob et al., 1999; Reece et al., 2001; Thompson et al., 2011). The variability between regions was higher than between P. olseni clones within each region; the percentages of similitude between clones from the same region was high (above 80% on average) in 3 locations and slightly lower in the other one. The number of spots shared by P. olseni from all the regions was higher than the numbers of spots exclusive of P. olseni from each region; this result contrasts with the comparison of the protein expression between 3 Perkinsus spp. (P. olseni, P. marinus and P. chesapeaki), in which the numbers of spots exclusive of each species was much higher than the number of spots shared by the 3 species (Chapter VI); lower variability between clones from the same species than between clones from different species was expected. Higher similarity was found among the proteome of P. olseni from the Atlantic regions than between the Mediterranean location and the Atlantic ones. Nevertheless, the highest similarity did not occur between the two closest locations (Arousa and Pontevedra) but between Carreras River and the other two Atlantic locations. This could be due to the fact that the movement of clams from the SW region of the Iberian Peninsula (Algarve and Huelva) to be immersed or depurated in the NW (Galicia) is much more frequent than movements of clams from NE (Catalonia) to the SW and NW or from the NW to the NE. Considering that the host source (R. decussatus) of the P. olseni isolates from the Atlantic locations was a different species from the host source (R: philippinarum) of the isolates from the Mediterranean location, to what extent the host difference influenced the variability remains unknown. Differences in P. olseni in vitro proliferation due to geographic origin were also detected, it was lower in clones from Ría de Arousa and Ría de Pontevedra and higher in Delta del’Ebre and Carreras river, which could be the result of local adaptation to environmental conditions; interestingly, water is warmer during summer in the two latter locations. Previous chapter using microsatellites to analyse genetic variability of P. olseni among the same locations as in this chapter showed no correlation between geographic and genetic distances. In the case of the spots shared by P. olseni from all the regions, a high proportion of the annotated proteins were involved in sugar associated metabolism, which could be due to the fact that the parasite cells were collected at the culture exponential growth phase. The occurrence of more than one spot in gels corresponding to glucose 6-P isomerase and enolase-2 suggests that several isoforms of those proteins occur in the parasite. In the case of glucose 6-P isomerase gene, different alleles were found in the oomycete Phytophthora infestans (Ospina-Giraldo Chapter IV 116 and Jones, 2003) as well as in Entamoeba histolytica (Razmjou et al., 2006). In the apicomplexan parasite Toxoplasma gondii, which develops metabolic and morphological changes during the infection, several stage-specific isoenzymes were found; these isoenzymes have different biochemical properties that maintain the major parasitic metabolism such as glycolysis in pace with the stage-specific requirements of carbohydrate or polysaccharide biosynthesis, including isoforms of glucose 6-P isomerase and enolase-2 (Tomavo, 2001). In P. olseni, the isoforms could have similar functions as in Toxoplasma gondii regulating the metabolism according to the progress of the infection. The strictly conserved plant enolase pentapeptide EWGWC insertion that had been found in P. marinus (Joseph et al., 2010) was also found in P. olseni enolase-2 in this study. This domain was found previously in other apicomplexan parasites as Plasmodium falciparum (Read et al., 1994) and Toxoplasma gondii (Tomavo, 2001). Glucose 6-P isomerase and pyruvate kinase were proposed as antiparasite drug targets in humans (Rigden et al., 1999; Tomavo, 2001; Muñoz and Ponce, 2003). A better understanding of these proteins in Perkinsus would help to find tools to fight the disease. Perkinsozoa group, where the genus Perkinsus is included, is considered the earliest group diverging from the lineage leading to dinoflagellates, branching close to the node shared by dinoflagellates and apicomplexans (Saldarriaga et al., 2003; Bachvaroff et al., 2011; Fernández Robledo et al., 2011). Members of both dinoflagellates and apicomplexans as well as P. olseni possess plastids and recent analysis of newly identified photosynthetic members of the apicomplexan lineage have shown that these plastids evolved from a single common secondary endosymbiosis with a red alga (Teles-Grilo et al., 2007; Moore et al., 2008; Janouskovec et al., 2010; Fernández Robledo et al., 2011). This suggests that non-photosynthetic relatives of both lineages, including Perkinsus, evolved from photosynthetic ancestors, raising the possibility that these lineages retain cryptic organelles (Keeling, 2010). This common lineage ancestor could explain the existence of proteins derived from a plant origin as these enolase-2 as well as superoxide dismutase and peroxiredoxin proteins that are discussed next. Enzymes linked to protection from host immune reaction were also well represented in the group of spots shared by all the regions that were annotated. A high constitutive expression of antioxidant proteins was detected. Proteins like peroxiredoxin (Prx) II and V and superoxide dismutase (SOD) were expressed in large quantities which suggests the importance of these proteins as defence mechanism. Their detoxifying role seems essential to infect the host and elude host defences; P. marinus inhibits the production of reactive oxygen intermediates when it is phagocytosed by host haemocytes, thus preventing oxidative damage (Robledo et al., 2008). In the case of the immune reaction of venerid clams against P. olseni, clam haemocytes encapsulate the parasite (Montes et al., 1995a; Villalba et al., 2004) and, according to our results, P. olseni also produces enzymes allowing to neutralise the reactive oxygen intermediates released from host haemocytes. Variability of the cell proteome of Perkinsus olseni 117 The S-adenosylhomocysteinase is a cytosolic enzyme that has important functions in the cytosolic metabolism linking processes of transmethylation, transsulfuration and purine metabolism (Kloor et al., 2000); the enzyme cleaves Sadenosylhomocysteine and provides homocysteine for the synthesis of cysteine and the regeneration of methionine (Takata and Fujioka, 1983). Methionine is also the source of cysteine, the limiting reagent for the synthesis of glutathione (GSH), one of the most important antioxidants in organisms (Company et al., 2011). The Sadenosylhomocysteinase was proposed as a target for chemotherapy in other protozoan parasites as Trichomonas vaginalis (Bagnara et al., 1996). In the case of spots exclusive of one region, the annotated proteins were classified into four categories according to Gene Onthology, metabolism (galactokinase in Ría de Arousa and malate dehydrogenase in Delta de l’Ebre), glycolysis (pyruvate kinase in Ría de Pontevedra), response to stress (heat shock protein 90 in Ría de Arousa), and oxidation/reduction (glyceraldehyde 3-P dehydogenase in Carreras River). All these proteins are involved in processes related to the Kreb´s cycle except the heat shock protein 90. Malate dehydrogenase, which was found in common proteins as well as Delta de l´Ebre specific protein,is involved in the metabolism of the intermediaries of the Kreb´scycle; the former catalyses the transformation of L-malate to pyruvate together with the coenzyme NADP+ (Sánchez et al., 1996). Pyruvate kinase catalyses the transformation of phosphoenolpyruvate to pyruvate, the previous step to the generation of Acetyl CoA. Two pyruvate kinases were detected in P. olseni from Ría de Pontevedra, the one shared by all the regions and one exclusive of this region. This specific protein from Ría de Pontevedra population seems to be the same proteins but with a posttranslational modification which rise its pI in 0.1 units. Galactokinase takes part in the Leloir route involving the transformation of galactose to glucose; it is an ATP-dependent enzyme catalysing the phosphorylation of galactose to form galactose1-phosphate (Chu et al., 2009). Glyceraldehyde 3-P dehydrogenase is a key enzyme in glycolysis and catalyses the oxidative phosphorilation of glyceraldehyde-3P into 1,3bisphosphoglycerate in the presence of NAD+ and inorganic phosphate (Daubenberger et al., 2000). Heat shock protein 90 had been studied in P. marinus due to its involvement in the infection progression; this protein is necessary for the adaptation of the parasite to the host environment. P. marinus, as many other organisms uses heat shock proteins as part of its adaptive survival repertoire (Tirard et al., 1995). Contrasting with the spots shared by P. olseni from all the regions, none of the annotated spots that were exclusively found in one region could be considered as directly linked to virulence, which would have supported the occurrence of virulence differences between P. olseni geographic strains; however, the detected differences (proteins involved in metabolism and stress response) could offer a better adaptation to environmental local conditions that would favour cell division / growth and thus possibly pathogenicity. Local adaptation of Perkinsus spp. to different ranges of salinity Chapter IV 118 and temperature can explain the adaptability to new environments, their proliferation and virulence (La Peyre et al., 2008; Thompson et al., 2014). Obviously, sequencing more spots would have allowed deeper analysis of P. olseni variability, but limited funding impeded going further. V. VARIABILITY OF PROTEIN EXPRESSION PROFILING IN THE EXTRACELLULAR PRODUCTS OF Perkinsus olseni AMONG REGIONS OF THE SPANISH COAST Variability of protein expression profiling in the extracellular products of P.olseni 121 V.1. ABSTRACT The variability of the protein expression profiling in the extracellular products (ECPs) of in vitro cultured Perkinsus olseni deriving from 4 regions of the Spanish coast was evaluated. The regions involved were the rías of Arousa and Pontevedra (Galicia, NW Spain), Carreras River (Andalusia, SW Spain) and Delta de l’Ebre (Catalonia, NE Spain). P. olseni in vitro clonal cultures were produced from parasite isolates from each of four clams from each region. Proteins released by the in vitro cultured parasites were isolated and separated by two dimensional electrophoresis (2DE). Qualitative comparison of protein expression profiles in the P. olseni ECPs among regions was performed with PD Quest software. Around 130 spots were counted in the gels from ECPs of P. olseni clones from each region, of which 23 spots were shared by all the regions and various spots were exclusive of one region. A total of 34 spots were excised from the gels and analysed for sequencing. The protein cathepsin B, involved in proteolysis, the signal recognition particle receptor subunit β, involved in protein transport through membranes, and a protein belonging to N-acetyl transferase superfamily, involved in biosynthesis, were identified in spots shared by P. olseni ECPs from all regions. Pepsin A precursor, involved in proteolysis; heat shock protein (HSP) 60; and phosphoserin aminotransferase, involved in biosynthesis, were exclusively identified in P. olseni ECPs from Ría de Arousa, while peroxiredoxin V, involved in oxidation-reduction, was exclusively identified in P. olseni ECPs from Ría de Pontevedra. Differences in released proteins suggest different virulence or resistance to host attack between parasites from different locations. Chapter V 128 Table V.1. Comparison of protein expression profiling between extracellular products of Perkinsus olseni clonal cultures from the same geographic origin. For each location, diagonal cells show the number of spots shared by the four gel replicates from each clonal culture. The number of spots shared between pairs of clonal cultures is shown in cells below the diagonal, and the percentage of similitude between pairs of clonal cultures is shown in cells above the diagonal. The mean percentage of similitude of the paired comparisons within each location is also shown. Location # Clon # Clon Mean percentage of similitude 1 2 3 4 Ría de Arousa 1 71 53.3 57.6 56.3 64.18 2 56 139 76.1 73.6 3 53 113 158 68.2 4 49 89 89 103 1 2 3 4 Ría de Pontevedra 1 135 90.8 87.6 77.4 82.61 2 119 127 85.1 77.4 3 124 117 148 77.3 4 95 91 99 108 1 2 3 4 Delta de l´Ebre 1 176 84.2 84.8 79.5 83.55 2 133 140 84.3 85.8 3 134 118 140 82.8 4 118 112 108 121 1 2 3 4 Carreras river 1 158 73.4 66.4 80.7 70.31 2 102 120 68.0 69.9 3 80 69 83 63.5 4 132 101 80 169 A master gel for P. olseni ECPs from Ría de Arousa was built including the 43 spots shared by every gel of that region, the master gel of Ría de Pontevedra included 86 spots, that of Carreras river 67, and the one of Delta de l´Ebre 144. The comparison between master gels from every region showed that 21 spots were shared by all regions; 5 spots (11.6%) of the master gel of Ría de Arousa were exclusive of that region (did not occur in the master gels of the other regions), 14 (16.3%) were exclusive of Ría de Pontevedra, 4 (6.0%) of Carreras river and 21 (14.6%) of Delta de l´Ebre. The percentages of similitude between regions (range: 43.6% – 72.5%; mean: 57.90%; Table V.2) were lower than those between clonal cultures within each region Variability of protein expression profiling in the extracellular products of P.olseni 129 (range: 64.18% - 83.55%; mean: 75.16%; Table V.1). Higher percentages of similitude between P. olseni from Delta de l´Ebre with Ría de Pontevedra and Carreras River than between parasites from Ría de Arousa with the other geographic origins were observed (Table V.2). V.4.2. Protein identification A total of 34 spots, including 11 spots shared by all populations and 23 spots exclusive from one population were excised for sequencing and identification. Ten out of 11 spots shared by P. olseni from all the regions that had been excised were successfully sequenced while 7 out of 23 excised spots exclusive from one region were sequenced (Table V.3). Some of the sequenced spots did not significantly match with any protein of the screened databases (Table V.3, Figs. V.4 and V.5). Table V.4 shows protein annotations. Table V.2. Number of spots shared between pairs of locations (above diagonal) and percentage of similitude between pairs of locations (below diagonal). Ría de Arousa Ría de Pontevedra Delta de l´Ebre Carreras River Ría de Arousa - 55.8 43.5 43.6 Ría de Pontevedra 36 - 70.5 61.4 Delta de l´Ebre 32 67 - 72.5 Carreras River 24 47 62 - Table V.3. Total number of spots and number of spots excised, sequenced and annotated, corresponding to the group of spots shared by extracellular products (ECPs) of Perkinsus olseni from all the regions and the groups of spots exclusive of (ECPs) of P. olseni from each region. Shared by all the regions Exclusive from Ría de Arousa Exclusive from Ría de Pontevedra Exclusive from Delta de l´Ebre Exclusive from Carreras river Total 23 5 14 21 4 Excised 11 4 7 10 2 Sequenced 10 4 2 1 0 Annotated 5 4 1 0 0 Chapter V 130 Fig. V.3. Digitised images of silver-stained gels produced by 2DE-PAGE of proteins occurring in extracellular products of Perkinsus olseni clonal cultures deriving from isolates from Ría de Arousa (A), Ría de Pontevedra (B), Delta de l’Ebre (C) and Carreras River (D). Variability of protein expression profiling in the extracellular products of P.olseni 131 Fig. V.4. Digitised image of a silver stained gel of proteins occurring in extracellular products of a Perkinsus olseni clonal culture deriving from Delta de l´Ebre. The spots that were annotated from those shared by parasites from all the regions are numbered in the image. Chapter V 132 Fig. V.5. Fragments of master gels showing the spots exclusive of extracellular products of Perkinsus olseni from one region that were annotated. Left column corresponds to the areas of final master gel where the exclusive spots occur. The remaining columns correspond to the same areas in the master gel of each region. Each file corresponds to an exclusive spot labelled with the spot code (the same as in Table V.4). ARía de Arousa; P - Ría de Pontevedra; C – Carreras River; D - Delta de l´Ebre. Variability of protein expression profiling in the extracellular products of P.olseni 133 Table V.4. Peptides annotated after searching in Blast database against Alveolata group in non-redundant proteins database. The C before numbers correspond to spots common to all Perkinsus olseni regions, “A” corresponds to exclusive spots of Ría deArousa region, “P” to exclusive spots exclusive of Ría de Pontevedra, “D” to exclusive spots exclusive of Delta de l´Ebre and finally “H” exclusive spots of Carreras river region. Protein identification: show the protein name of identified peptides in the database. A.no.: is the accession number of proteins to database. Biological function shows the function of the protein according to Gene Ontholgy Biological Process. Sco: Displays a score to BLAST results; the higher score the better the match to the searched sequence.E-value: Expected value describes the number of hits expected to see by chance when searching a database of a particular size. Mw: Molecular weight. pI: Isoelectric point. Spot code pI obs/pI theor Mw obs/Mw theo (kDa) Identified protein; organism A. no. Biological function Peptide sequence Sco E value C1 6.2/5.35 49/44.61 Cathepsin b, putative; Perkinsus marinus XP_002782575.1 Proteolysis VAVYSPEEEAQHXA 47.1 0.040 C2 6.6/- 50/- No identification - - QGPCKPVT, NXXDSFTCR - - C3 6.2/- 49/- No identification - - LAYDSTVR - - C4 6.55/6.54 40/30.5 Signal recognition particle receptor subunit beta, putative; Perkinsus marinus XP_002771937.1 Receptor EAAESLY 24.8 11 C5 6.55/- 40/- No identification - - VTVSADS, RGEAEKFL - - C6 6.5/- 28.2/- No identification - - AEAPESP, EGDMATLNY - - C7 5.8/8.69 17.84/15.78 Hypothetical protein Nacetyl transferase superfamily; Perkinsus marinus XP_002785377.1 N-Acetyltransferase activity VVAMYLVVNP, EAAESLY 34.1 0.055 C8 5.8/- 7.84/- No identification - - AGVQNDAG, VVTVSLP - - C9 5.4/6.54 14.74/30.46 Signal recognition particle receptor subunit beta, putative; Perkinsus marinus XP_002771937.1 Receptor EAAESLY 24.8 11 Chapter V 134 Spot code pI obs/pI theor Mw obs/Mw theo (kDa) Identified protein; organism A. no. Biological function Peptide sequence Sco E value C10 6.6/4.56 12.1/7.92 Hypothetical uncharacterised protein Perkinsus marinus XP_002788789.1 - ASDLAHLHYDG 38 5E-04 A1 5.52/5.19 68.75/59.45 Heat shock protein 60; Perkinsus marinus XP_002785716.1 Response to stress LNDALNA, EDFDPALLGSC 35.8 0.004 A2 6.65/5.89 42/41.88 Phosphoserine aminotransferase; Perkinsus marinus XP_002788117.1 L-Serine biosynthetic process LAEDSNGFYNA, DSNGFYNAPV 32.9 0.37 A3 7.38/7.51 40/38.8 Pepsin A precursor; Perkinsus marinus XP_002779904.1 Proteolysis LGTLQVGD 23.5 30 A4 5.57/8.69 18.15/15.78 Hypothetical uncharaterised protein Perkinsus marinus XP_002785377.1 - VVAMYLVV 27.4 0.8 P1 5.95/6.28 18.8/17.00 Peroxiredoxin V; Perkinsus chesapeaki ABV22156.1 Oxidationreduction process MXADXKADTA 27.4 2.1 P2 5.62/- 16.4/- No identification - - YDCGYTXDQGSGAP, AVCASDC - - D1 6.9/- 45.5/- No identification - - ADEGXV, GEAEKFX, KDYPY - - Variability of protein expression profiling in the extracellular products of P.olseni 135 V.5. DISCUSSION Results showed that P. olseni is able to release a high amount of proteins: up to 144 different spots were found in gels from ECPs of parasites from Delta de l’Ebre, which suggests a complex host-parasite biochemical interaction through the infection process. Furthermore, a remarkable variability in the protein profile of P. olseni ECPs depending on the geographic origin was found, because the percentage of similitude between clones within a region was higher than between clones of different regions. Interestingly, the variability of the protein profile of P. olseni ECPs is higher than that of cellular proteins recorded in chapter IV, because the percentage of similitude between clones within a region and between regions corresponding to ECPs was lower than those of cellular proteins, in spite of the number of cellular proteins was much higher than that of released proteins. Genetic and physiological variability depending on geographic origin had also been found in P. marinus (Reece et al., 1997, 2001; Thompson et al., 2011). The comparison of P. olseni ECPs between regions showed that the two closest regions were Carreras River and Delta de l´Ebre followed by Ría de Pontevedra and Delta de l´Ebre, while Ría de Arousa was the most distant region, which is difficult to explain and lacks geographic proximity support. These results are not consistent with those based on either P. olseni cellular proteins (Chapter IV) or microsatellites of P. olseni genomic DNA from the same locations (Chapter III), which showed more similitude between P. olseni clones from the Atlantic coast locations with regard to the clones from the Mediterranean coast. As discussed in chapter IV, it is difficult to estimate the influence of the host species source of P. olseni clones on the variability of the protein profiles of ECPs; in this case, the host species source seemed to play a less important role for differences between P. olseni clones of different regions. The comparison between ECPs was based on much fewer proteins than that between cellular proteins, which confers a more solid base to the latter to infer similitude between P. olseni clones from different geographic origin. Regarding protein annotation of sequences obtained from spots, the low success in the protein identification could be explained by the low quantity of ECP sequences of P. olseni or other species of the Alveolata group in the database, even after the complete genome of P. marinus has been sequenced (Joseph et al., 2010). Additionally, the low intensity (low protein concentration) of some spots made difficult obtaining a good spectrum by MS. Three annotated proteins corresponded to 4 spots shared by P. olseni ECPs from all the regions: One of these proteins, the signal recognition particle receptor subunit β (SRβ), corresponded to two different spots; why the same peptide sequence, annotated as SRβ, was identified in two spots with so different pI and molecular weight is puzzling. This receptor, only found in eukaryotes, is part of the complex signal recognition particle (SRP), a ribonucleoprotein complex that translocates proteins across the membrane to enter the secretory pathway (Egea Chapter V 136 et al., 2008; Akopian et al., 2013; Nyathi et al., 2013). This complex consists of three GTPases: the SRP54 subunit of the SRP and the αand βof the SR receptor (Helmers et al., 2003). The presence of SRβ in the ECP suggests that P. olseni could use the SRP complex as a transport channel to release proteins. Once the protein crosses into the complex, the SRβ could be released together with the protein. Nevertheless, the precise timing and coordination of events during the signal sequence transfer still remain poorly understood (Helmers et al., 2003; Nyathi et al., 2013). Other protein shared by P. olseni ECPs from all the regions was a cysteine peptidase, cathepsin B. Peptidases are essential for the establishment and survival of the parasites; they play an important role in the development and pathogenesis of several parasitic infections and have been proposed as targets in the structure-based strategy of drug design (Sakanari et al., 1997; Schaeffer et al., 2011). Cysteine peptidases from trophozoites of Entamoeba histolytica were found to have a cytopathic effect on mammalian cells (Keene et al., 1990); cysteine proteases of Plasmodium falciparum are involved in the degradation of host haemoglobin (Rosenthal, 1995); cathepsin B has been purified from intracellular and extracellular extracts from all stages of Trypanosoma cruzi (Nóbrega et al., 1998); cysteine peptidase involvement in host cell invasion by sporozoites of Eimeria tenella was deduced because cell invasion was inhibited by cysteine peptidase inhibitors (Schaeffer et al., 2011). In the case of the genus Perkinsus, other type of proteases, serine peptidases, have been found in P. marinus ECPs and been proposed as possible virulence factors responsible for tissue degradation in infected oysters (La Peyre, 1995b, 1996); the major extracellular protease (a N-glycosylated serine peptidase) produced by P. marinus in vitro was characterised and designed as perkinsin (Faisal et al., 1999). More recently, several protease sequences (cathepsin-like cysteine protease, subtilisin-like serine protease, rhomboid-like protease 1, cysteine protease, ATP-dependent protease, serine protease, metacaspase 1 precursor, and ubiquitin-specific proteases) were also identified in the P. marinus genome (Joseph et al., 2010) and would be expected to degrade host protein substrates to acquire nutrients necessary for normal cell function and proliferation (Soudant et al., 2013). No protease activity had been found in previous studies on P. olseni ECPs although other hydrolytic enzyme activities were reported (Casas et al., 2002b, 2008, 2009). Hosts counteract the proteases released by parasites by producing protease inhibitors. Serine protease inhibitors that may inhibit P. marinus proliferation were found in C. virginica oysters (Faisal et al., 1998; Oliver et al., 2000; Xue et al., 2006, 2009; La Peyre et al., 2010). The occurrence of cysteine peptidase in P. olseni ECPs raise the question of the production of protease inhibitors in the Manila clam R. philippinarum; 23 contigs with homology to serine, cystein, Kunitz-type and Kazal-type protease inhibitors and metalloprotease inhibitors were identified in Manila clams (Moreira et al., 2012). Other identified protein shared by P. olseni ECPs from all the regions was a hypothetical protein of the N-acetyl transferase superfamily; there are many different proteins within this superfamily, these enzymes Variability of protein expression profiling in the extracellular products of P.olseni 137 use acetyl coenzyme A to transfer an acetyl group to a substrate, a reaction implicated in various functions of cellular processes. Three annotated proteins -pepsin A precursor, phosphoserine aminotransferase, and heat shock protein (HSP) 60corresponded to spots exclusively found in the ECPs of the four P. olseni clones from Ría de Arousa. Pepsin A precursor is an aspartic protease occurring in many organisms. The catalytic site is formed by two aspartate residues, one of which have to be protonated and the other deprotonated for the activation of the protein (Campos and Sancho, 2003). This is the second protease identified in P. olseni ECPs in this study. Considering the important role of proteases released by parasites in their pathogenicity, the fact that the spot corresponding to pepsin A precursor was not found in clones from the other regions could support higher virulence of P. olseni clones from Ría de Arousa. Phosphoserine aminotransferase (PSAT) is involved in serine biosynthesis; L-serine serves as a building block for protein synthesis and also plays an important role in various metabolic pathways for the generation of essential compounds, including glycine, L-cysteine, Lmethionine, phosphatidyl-L-serine, sphingolipids, taurine, porphyrins, purines, thymidine, and neuromodulators D-serine that are essential for the growth of the organisms (Ali and Nozaki, 2006). HSP 60 is included in a highly conserved family of proteins present in all organisms; they were originally identified as cellular proteins produced in response to stress or elevated temperature, but most HSPs were found to be constitutively expressed in cells and they are essential for cellular growth under normal conditions (Syin and Goldman, 1996). Most of these proteins play essential roles in protein biosynthesis and are involved in the transport, translocation and folding of proteins. Extracellular HSPs are important mediators of intercellular signaling and transport, once released may then bind to the surfaces of adjacent cells and initiate signal transduction cascades as well as the transport of cargo molecules (Calderwood et al., 2007). The role of extracellular HSP 60 is not well understood, induction of immunosupression was suggested (Flohe et al., 2007) and they have been shown to induce inflammation (Tian et al., 2013) or even autoimmune aggression (Cappello et al., 2009); tumour cells have been shown to release HSP 60 (Merendino et al., 2010). Thus, pathogenic effect of HSP60 released from P. olseni could be suspected; if so, it would further support higher virulence of P. olseni clones from Ría de Arousa. Peroxiredoxin V annotation corresponded to one spot occurrying exclusively in the four P. olseni clones from Ría de Pontevedra. Peroxiredoxins belong to the peroxidase family. In some parasites they may participate in the defence against the host attacks involving reactive oxygen intermediates (ROIs), because peroxiredoxins prevent the accumulation of H2O2, and even exert potent immune-modulatory effects (Kawazu et al., 2008; Ishii et al., 2012; Robinson et al., 2013). Peroxiredoxin V was reported within in vitro cultured P. chesapeaki and P. olseni cells, the latter also deriving from Ría de Pontevedra (Chapter VI), and the gene coding this protein was Chapter VI 144 European and South American countries (Casas et al., 2002a; Cremonte et al., 2005; Choi and Park, 2010; Dungan et al., 2007; Goggin and Lester, 1995; Hamaguchi et al., 1998; Leethochavalit et al., 2004; Park et al., 2006; Sanil et al., 2010; Wu et al., 2011). P. olseni has been blamed for mass mortalities of R. philippinarum in Korea (Park and Choi, 2001) and China (Liang et al., 2001) and has been associated with mortality of R. decussatus in Portugal (Azevedo, 1989) and Spain (Villalba et al., 2005). The life cycle of Perkinsus spp. involves that a cell type, usually called trophozoite, proliferates throughout host tissues undergoing successive bipartitioning to yield daughter cells that stay together inside a wall; daughter cells become independent and mature into trophozoites. When host infected tissues are incubated in fluid thioglycollate medium (which mimics natural tissue rotting after host death), the trophozoites transform into a resistant stage, called hypnospore. When hypnospores are transferred into seawater, zoosporulation begins and progresses with successive karyokinesis and cytokinesis, leading to formation and release of hundreds motile biflagellated zoospores. Zoosporulation is rarely seen in P. marinus when placed in sea water (da Silva et al., 2013; La Peyre, pers. comm.). Trophozoites, hypnospores and zoospores are able to initiate infection in a healthy host (Villalba et al., 2004). Procedures for in vitro cultivation of Perkinsus spp. have been established (La Peyre, 1996). Various approaches have been used in biological, ecological and evolutionary studies to characterise and understand the phenotypic variability at the biochemical level, such as genomics, transcriptomics and proteomics (Biron et al., 2006; Derome and Bernatchez, 2006; Roberge et al., 2008). All of them have limitations and numerous studies have shown examples of lack of correlation between mRNA and protein abundance (Griffin et al., 2002; Gygi et al., 1999; Ideker et al., 2001; Lippolis and Reinhardt, 2010). This lack of correspondence has been used as a justification for the application of proteomics in studying expression differences between species or strains (Brobey et al., 2006; Chan et al., 2005; Enard et al., 2002; Wang et al., 2008), with the aim of the identification of putative virulence factors as it was shown in several proteomic studies on parasites of the Alveolata group (Briolant et al., 2010; Cuervo et al., 2007; Drummelsmith et al., 2003). Information derived from these techniques are complementary, thus the combination of all, including metabolomics, are necessary for an in-depth understanding of parasite biology, accelerate the search for new proteins and therefore facilitate a comprehensive view of the cellular process including gene expression, mRNA content, protein expression and cellular response. The purpose of our study was to establish 2-DE as a tool to study protein expression patterns of the 3 Perkinsus spp. that have been blamed for host mortality, P. marinus, P. olseni, and P. chesapeaki, with the ultimate goal of employing proteomic technologies to identify common and species-exclusive proteins. If differences were found, they could contribute to a better understanding of Perkinsus spp. infection mechanisms, and differences of virulence and host preferences among species. Various proteins related with detoxification, electron transport, synthesis and signal Comparison of protein expression profiles between three Perkinsus spp. 145 transduction have been identified. The functions of identified proteins have been assigned according to Gene Ontology and the potential pathways of these proteins are discussed in light of bioinformatics analyses. VI.3. MATERIALS AND METHODS VI.3.1. In vitro clonal cultures Isolates from P. olseni (Pag6-07-P2-A1), P. marinus (GTLA-34) and P. chesapeaki (PRA-65) were used to develop in vitro clonal cultures. P. olseni culture was established from gills of clams R. decussatus from the Ría de Pontevedra (Galicia, NW Spain), following the method of La Peyre et al. (1993); P. marinus culture was provided by Jerome La Peyre (Department of Veterinary Science, Louisiana State University Agricultural Center), which had been established from heart from an infected Crassostrea virginica from Lower Barataria Bay (Louisiana, USA); P. chesapeaki culture was provided by Chris Dungan (Maryland Department of natural Resources, Cooperative Oxford Laboratory), which had been established from hypnospores of Mya arenaria from Chesapeake Bay (Maryland, USA). All cultures were maintained at 25ºC and were subcultured every 2-3 months in the protein-free culture medium JLODRP-2A (Casas et al., 2002b). In vitro clonal cultures of P. olseni, P. marinus and P. chesapeaki were established as described Casas and La Peyre (2009). Parasites in exponent phase of growth were collected by centrifugation at 1000g for 10 min at 25°C. Parasite density was estimated by counting with a haemocytometer. VI.3.2. Protein extraction Proteins were extracted by suspending 150x106 trophozoites from in vitro clonal cultures (one culture of each parasite species) in lysis buffer (8M urea, 2M thiourea, 2% CHAPS, 1% DTT, 0.8% ampholites (pH 3-10) and a 1/100 dilution of a protease inhibitor (Sigma Protease Inhibitor P2714). Proteins were solubilised for 2 h 30 min at 4°C with vigorous shaking. The lysate was centrifuged at 16000g for 30 min at 4ºC. The supernatant was purified using the 2-D Clean Up Kit (GE Healthcare) and finally, the pellet was resuspended in rehydration solution (8M urea, 2% CHAPS, 0.5% IPG buffer, 0.8% ampholites (pH 3-10), 40mM DTT and bromophenol blue traces). Protein concentration was measured using the DC protein assay kit (Bio-Rad), according to Lowry et al. (1951). VI.3.3. Two dimensional electrophoresis (2DE) For the first dimension separation, aliquots of 150 µg of protein samples were diluted to a final volume of 350 µl in rehydration solution and were incubated for 30 min at room temperature (23-26°C). Samples were centrifuged 1 min at maximum Chapter VI 146 speed to remove bubbles and loaded onto the immobilized pH gradient (IPG)-strips (18 cm, pH 4-7 linear, GE Healthcare) by in-gel rehydration. After 6-h passive and 6-h active (50V) rehydration, Iso-Electro Focusing (IEF) was performed (20ºC, 50 µA/strip) in a Protean®IEF System (Bio Rad) using six steps: 500 V, 90 min; 1000 V, 90 min; 2000 V, 90 min; 4000 V, 90 min; 8000 V, 120 min, and 8000 V, until 60 000 Vh (4 h) according to Romero-Ruiz et al. (2006). After IEF, proteins were reduced (10 mg/ml DTT, 20 min) and alkylated (45 mg/ml iodoacetamide, 20 min) in equilibration buffer (6M urea, 50mM Tris pH 8.8, 2% SDS, 30% glycerol) before separation in the second dimension. The proteins on the equilibrated IPG strips were separated across 12.5% SDS-PAGE gels using a vertical system (PROTEAN Plus Dodeca Cell, BioRad) and standard Tris/glycine/SDS buffer. Gels were run at 2.5 W/gel for 15 min followed by 12.5 W/gel for 5h 30 min. Molecular weight markers (BioRad) were run in the second dimension next to the problem sample for protein size determination. VI.3.4. Protein visualisation and image analysis Four replicates gels from the same protein extraction of each species were prepared. Protein spots in the gels were visualized by Sypro ruby (BioRad) staining following the manufacturer's specifications. Images of the four replicate gels from each in vitro clonal culture, (4 replicates x 3 species = 12 gels) were digitised with a FXImager scanner (BioRad). Automatic gel analysis using the ProteomweaverTM software 4.0 (Bio-Rad) was performed to properly assign the spots, correcting possible slight better match spots correcting possible slight gel deformations. All the images were highly reproducible in terms of number of spots, position and intensity. Spot patterns were analysed and compared between parasites in search for differences in protein expression profiles. The gels were calibrated using a select set of reliable identification landmarks distributed throughout the entire gel to determine experimental isoelectric point (pI) and molecular weight (Mw) coordinates for each single spot. All spot volumes were normalized to get normalized spot intensities. The theoretical Mw and pI values were obtained with the compute pI/Mw tool at the ExPASy Proteomics Server (http://www.expasy.org/tools/pi_tool.html ). VI.3.5. Protein identification and database search Fourteen spots that occur in the gels of the three Perkinsus spp. at the same position (common spots), 14 spots that were exclusive of P. marinus, 14 spots exclusive of P. olseni and 14 spots exclusive of P. chesapeaki were selected for protein identification. The number of spots selected for identification was limited by funding availability, thus the selection criteria was to choose the most intense spots among the common and the exclusive ones. Candidate protein spots were excised from the gels with a robot station (Genomic Solutions Propic®) at the Proteomics Facilities of Comparison of protein expression profiles between three Perkinsus spp. 147 University of Córdoba (SCAI, University of Córdoba, a member of ProteoRed network). In the case of the 14 spots common to the three Perkinsus spp., one spot from a gel of each species was excised (14 x 3 species = 42), to assure that the proteins were really the same in the three species. A total of 84 (42 exclusive spots + 42 common spots) excised spots were sent to Proteomics Laboratory of Autonomous University of Barcelona (LP-CSIC/UAB) and digested with trypsin as described by Romero-Ruiz et al. (2006). Extracted peptides were analysed by matrix-assisted laser desorption/ ionisation/time-of-flight mass spectrometry (MALDI-TOF MS) using a Voyager DE-PRO instrument (Applied Biosystems, Foster City, CA, USA). An ion-trap mass spectrometer Finningan LCQ IT (ThermoQuest, Finnigan MAT, San José, CA, USA) was used to verify proper digestion and sequencing by nanoelectrospray ion trap MS/MS (nESI-IT MS/MS). Peaks software (Thermo Electron Corporation, San Jose, CA USA) was used for de novo peptide sequencing. SEQUEST software was used for automated searching protein sequences against Uniprot - SwissProt database, in first time sequences were searched against non-redundant protein database of all organisms and then against Alveolate taxa database. All sequence tags obtained were manually confirmed and were submitted to a homology search using the pBLAST algorithm software (NCBI, USA). Data were contrasted against Alveolata taxa included in NCBI (National Center for Biotechnology Information, Maryland, USA) database. Correct identifications were considered valid with a score (ALC%) ≥ 65 (PEAKS) and a Sf value ≥ 0.90 (SEQUEST). VI.3.6. Statistical analysis Protein expression similarity between Perkinsus spp. was evaluated by considering only the spots occurring in at least one species, with a low variation coefficient among the four gels of the same species, to assure homogeneous intensity, an average linkage cluster analyses of the 12 gels (4 gels x 3 Perkinsus spp.) was performed based on the intensity of the spots to produce a dendrogram with correlation coefficient distance. Additionally, differences between the three Perkinsus spp. in the intensity of the 14 common spots that had been excised for protein identification were tested through one way ANOVA, followed by Tukey´s multiple comparison tests. MINITAB 15 software was used for the statistical analysis. VI.4. RESULTS Fig. VI.1 shows representative 2-DE gels of P. olseni, P. marinus and P. chesapeaki. An average of 1894 spots (range: 1518 – 2093) were detected in the gels of P. olseni, 1660 (range: 1394 – 2101) in those of P. marinus and 1999 (range: 1881 – 2125) in those of P. chesapeaki. The comparison between the three Perkinsus spp. showed 213 spots shared by the three species; P. chesapeaki and P. marinus shared Chapter VI 148 310 spots, P. chesapeaki and P. olseni shared 315 spots and P. marinus and P. olseni shared 242 spots. A number of spots were exclusive of each Perkinsus species: 1161 spots were exclusive of P. chesapeaki, 1124 of P. olseni and 895 of P. marinus (Fig.VI.2). Cluster analysis grouped the four gels of each Perkinsus sp.; furthermore, P. marinus and P. olseni gels were grouped in a cluster different from P. chesapeaki, although the distance between the P. chesapeaki cluster and the P. marinus + P. olseni cluster was short (Fig. VI.3). Significant differences between the three Perkinsus spp. in the intensity of some of the 14 common spots that were excised for identification were found (Fig. VI.4). Table VI.1 summarises the data of the identified spots. Forty-two of the 84 excised spots were successfully sequenced and 28 were identified, which represents 67% of successfully sequenced spots. Regarding common spots, C3 and C7 were identified as receptor for activated C kinase and pseudouridine synthase, respectively, by the same peptide sequences in all three species; spot C1 had peptides identified pairwise among the three species, corresponding to vacuolar ATPase b subunit; spot C13 was identified as phosphate acetyltransferase in P. chesapeaki and P. marinus; and spot C14 was identified as vacuolar ATPase subunit g in P. olseni and P. marinus. The spot C2 was identified as malate dehydrogenase only in P. olseni. Five more common spots were sequenced but they could not be annotated; their sequences showed some differences between the 3 Perkinsus spp. except for one of them, which showed the same sequence in the three Perkinsus spp. With regard to species-exclusive spots, six spots were identified in P. olseni: O1 as an uncharacterized protein of P. marinus; O3 as formate dehydrogenase; O6 as a proteasome subunit; O7 as triose phosphate isomerase and O14 as peroxiredoxin V. Two specific spots were successfully identified in P. marinus: M1 as phosphoglycerate kinase and M4 as 40S ribosomal protein S3. Seven specific spots were identified in P. chesapeaki: three of them corresponding to spots Ch1, 6 and 9 were annotated as uncharacterized proteins of P. marinus; Ch7 as glutathione S-transferase; Ch8 as peroxiredoxin II; and Ch11 and Ch12 as peroxiredoxin V. According to Gene Onthology (GO) annotation of biological process, the identified proteins of three Perkinsus species could be classified in 7 categories (Fig. VI.5). The majority of the identified proteins were clustered into electron transport; other identified proteins were involved in antioxidant functions, protein synthesis, carbohydrate metabolism and signal transduction; a minority were involved in metabolic processes and proteolysis. Comparison of protein expression profiles between three Perkinsus spp. 149 Fig. VI.1: Digitised images of Sypro ruby-stained gels produced by 2D-SDS-PAGE of Perkinsus olseni (A), Perkinsus marinus (B), and Perkinsus chesapeaki (C). The spots that were excised for identification are pointed out; the spots C1-14 were common to the three Perkinsus spp., O114 were exclusive of P. olseni, M1-14 were exclusive of P. marinus, and Ch1-14 were exclusive of P. chesapeaki. (A) (B) (C) Chapter VI 150 Fig. VI.2. Venn diagram showing the results of the protein analysis. Fig. VI.3. Dendrogram of combined expression data corresponding to 4 gels of each Perkinsus spp. (P. olseni, P marinus, and P. chesapeaki) considering only the spots occurring in at least one species with a low variation coefficient among the four gels of the same species to assure homogeneous intensity. The 4 gel replicates of each species are technical replicates of the same sample. Comparison of protein expression profiles between three Perkinsus spp. 151 Fig. VI.4. Mean intensity corresponding to the 14 spots occurring in the gels of the three Perkinsus spp. that were excised for identification. Perkinsus spp. are distinguished with different bar patterns. Different letters in the patterns of each spot indicate significant differences between Perkinsus spp. Spot 1: vacuolar ATPase subunit b; spot 2: malate dehydrogenase; spot 3: receptor for activated C kinase, spot 7: pseudouridine synthase; spot 13: phosphate acetyl transferase; spot 14: vacuolar ATPase subunit g; the remaining spots were sequenced but did not match with proteins in databases. Chapter VI 152 Fig. VI.5. Pie diagrams of each species showing the percentage of each protein group and the identified proteins in the study Comparison of protein expression profiles between three Perkinsus spp. 153 Table VI.1. Peptides annotated after searching in Blast database against Alveolata group in non-redundant proteins database. The initial letter “C” in the code corresponds to spots common to the three Perkinsus spp., “Ch” corresponds to spots exclusively found in P. chesapeaki, “M” to spots exclusive of P. marinus and “O” to spots exclusive of P. olseni. Mw: Molecular weight. pI: Isoelectric point. Sf: The Sf score for each peptide is calculated by a neural network algorithm that incorporates the Xcorr, DeltaCn, Sp, RSp, peptide mass, charge state, and the number of matched peptides for the search; the higher the value of the Sf score, the better the peptide match. E-value: Expectation value. The number of different alignents with scores equivalent to or better than S that are expected to occur in a database search by chance. The lower the E value, the more significant is the score. Function was stated according to Gene Onthology. Spot No. Mw Theor./ Mw Exp. pI Theor./ pI Exp. PEAKS sequence SEQUEST sequence Sf Identified protein; organism (automated search) UNIPROT AC Blast (manual search) E-value Function C1 o 55.6 (94.0) 5.1 (5.2) K.EM*IQTGISAIDTM*NSVVR.G 0.98 Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 Vacuolar ATPase beta subunit 1,0E-10 Electron Transport K.LPLFSAAGLPHNEIAAQVCR.Q 0.92 Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 3,0E-15 R.NDFEENGSM*ENVVLFM*NLANDP TIER.I 0.97 Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 3,0E-20 FTAAVNAAAVTR Tipc, putative; Toxoplasma gondii B9QHT5 5,1E-03 MDLTAAEYFAYER Putative vacuolar H + ATPase subunit B; Toxoplasma gondii Q86N78 2,0E-04 C1 m K.QVYPPINVLPSLSR.L 0.93 Chromosome undetermined scaffold_14; Paramecium tetraurelia A0C0D1 2,0E-09 FMAAALNAAAVTR Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 8,0E-04 K.EM*IQTGISAIDTM*NSVVR.G 0.96 Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 1,0E-10 GDAAVNAAAVTR Putative uncharacterized protein; Toxoplasma gondii Q1JSG6 5,1E-03 FDAAALNAAAVTR Putative uncharacterized protein; Toxoplasma gondii B9Q850 9,0E-03 C1 ch M(CamC)PETGLSALDTFN SVVR Vacuolar ATPase beta subunit; Amphidinium carterae B4ZFY9 2,0E-04 QVYPPLNVLPAESR Chromosome undetermined scaffold_42; Paramecium tetraurelia Q3SEB3 4,0E-03 FDAAALNAAAVTR Putative uncharacterized protein; Toxoplasma gondii B9Q850 9,0E-03 XIII. Resumen 257 El cultivo de almejas es una actividad que crece cada año a nivel mundial. La especie de almeja de mayor producción en el mundo es la almeja japonesa Ruditapes philippinarum. El país con mayor producción de esta especie es China, seguido por Italia, Corea del Sur, EE. UU. y España. La almeja japonesa se introdujo en Europa durante las décadas de los 70 y 80 del pasado siglo y se expandió rápidamente, naturalizándose en varios países europeos. R. philippinarum está bien adaptada y crece más rápido que la especie nativa Ruditapes decussatus, por lo que se ha convertido en la almeja de mayor producción en Europa. La producción de almejas venéridas en Galicia, considerando marisqueo y cultivo, es un recurso socio-económico muy importante. También en Galicia el cultivo de almeja japonesa se ha incrementado en los últimos años, convirtiéndose en la especie de almeja cultivada con producción más alta. La infección por Perkinsus olseni es una de las enfermedades más serias que afectan a las almejas. Este parásito está muy ampliamente distribuido por el mundo, afectando a una larga lista de moluscos en los cinco continentes. La infección por P. olseni se ha asociado a episodios de mortandad de R. decussatus y R. philippinarum en áreas del Sur de Europa así como de R. philippinarum en varios países asiáticos. Dos especies del género Perkinsus, P. olseni y P. marinus, están incluidas en la lista de enfermedades de declaración obligatoria de la Organización Mundial de Sanidad Animal, lo que indica el interés internacional por frenar su expansión. El conocimiento sobre P. olseni es escaso si se compara con el de P. marinus; ésta provoca mortandades masivas de ostras Crassostrea virginica en EE. UU. La transcendencia económica de la perkinsosis de las almejas justifica la investigación encaminada a conocerla mejor, para encontrar vías por las que minimizar sus efectos. Con esta intención se desarrolló este estudio, para conocer (1) la variabilidad de P. olseni en la costa española, con énfasis en la variabilidad de su virulencia, y (2) la modulación de la expresión proteínica de la almeja japonesa debida a la infección por P. olseni, con énfasis en la búsqueda de marcadores proteínicos de resistencia a esta infección. El estudio de la variabilidad de P. olseni se abordó analizando la estructura genética poblacional así como comparando el proteoma de clones de P. olseni derivados de varias regiones repartidas por el litoral español. Además se comparó el proteoma de P. olseni con los de otras dos especies de Perkinsus spp., P. marinus y P. chesapeaki, para así ampliar la perspectiva de la variabilidad del parásito. El análisis de la modulación de la expresión proteínica de la almeja japonesa debida a la infección por P. olseni se enfocó en los hemocitos y la hemolinfa, con objeto de concentrar la atención en la modulación de la respuesta inmunitaria de la almeja por el parásito; se consideró el efecto en la almeja de la exposición al parásito a corto plazo así como tras un periodo muy prolongado. XIII. Resumen 258 Los objetivos concretos de este estudio fueron los siguientes: 1. Evaluación de la variabilidad genética de P. olseni en el litoral español usando marcadores microsatélite. 2. Evaluación de la variabilidad de los perfiles proteínicos celulares y extracelulares de P. olseni en el litoral español. 3. Comparación de los perfiles proteínicos celulares de tres especies del género Perkinsus: P. olseni, P. marinus y P. chesapeaki. 4. Identificación de proteínas de hemocitos y plasma de la almeja R. philippinarum cuya expresión está modulada por la infección por P. olseni. 5. Identificación de proteínas marcadoras de resistencia a la infección por P. olseni en la almeja R. philippinarum. Capítulo 1. Evaluación de la variabilidad genética de Perkinsus olseni entre regiones del litoral español mediante el uso de marcadores microsatélite . Hay escasa información sobre una serie de parámetros poblacionales de P. olseni relevantes, como variabilidad genética, patrones de dispersión, estructura genética poblacional y estrategias reproductivas. Se sabe que el parásito se propaga por amplificación clonal, tanto dentro como fuera del hospedador, y esto debería determinar la estructura genética poblacional, aunque la existencia de reproducción sexual ha sido postulada previamente sin que todavía se haya determinado en qué punto del ciclo de vida tiene lugar. Conocer bien todos estos aspectos es muy importante para la gestión de cualquier enfermedad parasitaria. Para conseguir información en este contexto, se desarrolló un estudio de la variabilidad genética de P. olseni considerando las tres regiones de mayor producción de almejas en España, usando marcadores microsatélite. En concreto se analizaron muestras de dos rías gallegas (Arousa y Pontevedra), Andalucía occidental (río Carreras, Huelva) y el Delta del Ebro. Además, como referencia se utilizaron cultivos clonales de P. olseni derivados de parásitos aislados de moluscos de Japón y Nueva Zelanda. Para garantizar la disponibilidad de un número suficiente de células idénticas del parásito en cada réplica de cada región, se utilizaron cultivos clonales producidos in vitro. Un total de 130 cultivos clonales de P. olseni, provenientes de las seis regiones mencionadas, fueron genotipados con 12 microsatélites diseñados expresamente para el estudio. El análisis de la variación de los 12 microsatélites de las diferentes regiones de la Península Ibérica mostró un déficit de heterozigotos muy marcado con respecto a XIII. Resumen 259 las expectativas relativas al equilibrio Hardy-Weinberg. Tanto el polimorfismo como el número de alelos fueron bajos, observándose una estructura genética moderada entre poblaciones y una elevada estructura genética entre parásitos de almejas diferentes. A pesar del marcado déficit de heterozigotos, su existencia sugiere que el parásito tiene una fase diploide en su ciclo de vida, confirmando la reproducción sexual. Este déficit de heterozigotos es explicado por el efecto Wahlund (la reducción de la heterozigosis en una población es causada por la existencia de subpoblaciones). Ello sugiere que las poblaciones del parásito están formadas por infrapoblaciones constituidas por los parásitos presentes en cada uno de los hospedadores infectados. Las distancias genéticas entre clones de diferentes almejas del litoral español fue relativamente alta, independientemente de la región de origen, y las infrapoblaciones mostraron diferenciación genotípica significativa. Estos resultados sugieren endogamia dentro de las infrapoblaciones, con una frecuente recombinación sexual y no una estructura poblacional únicamente clonal. Las muestras procedentes del área Asia-Pacífico eran muy diferentes de las del litoral español y mostraron mayor variabilidad. Este hallazgo es coherente con la hipótesis de que P. olseni fue introducido en la Península Ibérica debido a importaciones de almeja japonesa infectada con el parásito, con el consiguiente efecto fundador que explicaría una pérdida severa de variación genética en las poblaciones españolas del parásito. Por otro lado, se observó una falta de correlación entre distancia génica y geográfica en el litoral español, lo que sugiere que la dispersión del parásito debida a las corrientes oceánicas tiene una menor importancia que la dispersión ocasionada por el movimiento de lotes comerciales de almejas. Capítulo 2. Variabilidad del proteoma celular de Perkinsus olseni entre regiones de la costa española. En este capítulo se abordó la comparación del proteoma celular de P. olseni de las regiones del litoral español incluidas en el capítulo 1, con el fin de evaluar la variabilidad del parásito en el litoral español con metodología diferente. Para ello se aislaron parásitos P. olseni de cuatro almejas R. decussatus de la ría de Arousa, cuatro de la ría de Pontevedra, cuatro del río Carreras y cuatro almejas R. philippinarum del Delta del Ebro. Con los parásitos aislados de cada almeja se iniciaron cultivos in vitro (uno por almeja), que a su vez se clonaron. Un clon de P. olseni derivado de cada una de las almejas, es decir cuatro clones de cada origen geográfico (16 clones en total) se cultivaron in vitro para obtener un número de células idénticas suficientes de cada clon para análisis proteómico. Se aislaron las proteínas celulares de muestras de cada clon y se separaron mediante electroforesis bidimensional, obteniéndose marcas proteínicas (spots) en geles de poliacrilamida. Los geles se tiñeron con plata para visualizar las marcas proteínicas y se digitalizaron. Se produjeron cuatro geles de cada clon, XIII. Resumen 260 descartándose el peor de ellos, de forma que se utilizaron tres geles (réplicas analíticas) de cada clon. Los geles digitalizados se analizaron con el programa informático PD Quest V.7.4.0. Se obtuvieron más de 600 marcas por gel, de las que 141 eran comunes a todos los clones de las 4 regiones, mientras que 39 marcas eran exclusivas de la ría de Arousa, 57 exclusivas de la ría de Pontevedra, 37 del río Carreras y 47 del Delta del Ebro. La comparación entre regiones por pares mostró porcentajes de similitud entre 55 y 72%; Las regiones atlánticas mostraron más similitud entre ellas que la estimada entre el Delta del Ebro y las regiones atlánticas, aunque el porcentaje de similitud entre las regiones atlánticas no se correlacionaba con la distancia geográfica. Debido a una insuficiencia presupuestaria, sólo una parte de las marcas proteínicas pudieron procesarse para su secuenciación por espectrometría de masas y posterior identificación en bases de datos. En concreto 19 marcas comunes a todos los clones de las 4 regiones y 35 marcas exclusivas de alguna de las regiones (7 de la ría de Arousa, 11 de la ría de Pontevedra, 8 del río Carreras y 9 del Delta del Ebro) se procesaron. Del total de las 54 marcas procesadas, se pudieron secuenciar 22 e identificar 19 proteínas (13 proteínas comunes a todas las regiones y 6 específicas de clones de una región). La mayor parte de las proteínas identificadas correspondieron a proteínas implicadas en procesos metabólicos, pero entre ellas cabe destacar la enolasa-2. Esta proteína presenta un pentapéptido estrictamente conservado en plantas, ya encontrado previamente en P. marinus, lo que es coherente con la filogenia del parásito, con un ancestro común con dinoflagelados. Se constató también una elevada expresión constitutiva de proteínas antioxidandes como la superóxido dismutasa y las peroxiredoxinas II y V, que podrían estar implicadas en la neutralización de especies reactivas tóxicas de oxígeno liberadas por los hemocitos del hospedador como parte de la respuesta inmunitaria de éste. En el caso de las proteínas exclusivas de cada población, se identificaron la galactokinasa y una proteína de choque térmico 90 (HSP 90) en clones de la Ría de Arousa, la piruvato kinasa en la Ría de Pontevedra, la malato deshidrogenasa en el Delta del Ebro y la glyceraldehido 3-fosfato deshidrogenasa en el río Carreras. Estas proteínas del grupo de las exclusivas de una zona están todas implicadas en procesos del ciclo de Krebs excepto la HSP90; es reseñable que ninguna de las proteínas exclusivas identificadas estaba directamente implicada en procesos de virulencia o toxicidad, por lo que no se detectaron diferencias asociadas a virulencia entre regiones. Capítulo 3. Variabilidad de los perfiles de expresión proteínica en los productos extracelulares entre regiones de la costa española. En este capítulo se abordó la comparación de los perfiles proteínicos de los productos extracelulares de P. olseni de las regiones del litoral español mencionadas XIII. Resumen 261 en los capítulos previos. Se había constatado en varias especies del género Perkinsus, la liberación de enzimas hidrolíticas que degradan los tejidos del hospedador, por lo que los productos extracelulares del parásito incluyen factores de virulencia. El estudio incluido en este capítulo pretendió evaluar la variabilidad del perfil proteínico de los productos extracelulares del parásito en el litoral español, con énfasis en las diferencias asociadas a la virulencia. Como en el capítulo anterior, se utilizaron cuatro clones de cada origen geográfico (16 clones en total). Cada clon derivaba de un hospedador diferente. Los clones se cultivaron in vitro y las proteínas liberadas por el parásito al medio de cultivo se aislaron y se separaron mediante electroforesis bidimensional en geles de poliacrilamida. El tratamiento de los geles fue similar al descrito en el capítulo previo. Se visualizaron entre 118 y 144 marcas en los geles, de las que 23 eran comunes a todos los clones de las 4 regiones, mientras que 5 eran específicas de la ría de Arousa, 14 de la ría de Pontevedra, 21 del Delta del Ebro y 4 del río Carreras. La comparación entre regiones por pares mostró porcentajes de similitud entre 44 y 72%, siendo la ría de Arousa la región más dispar. La insuficiencia presupuestaria sólo permitió procesar 34 marcas para su secuenciación por espectrometría de masas y posterior identificación en bases de datos, 11 comunes a los clones de todas las regiones y 23 específicas de una región (4 de la ría de Arousa, 7 de la ría de Pontevedra, 10 del Delta del Ebro y 2 del río Carreras). De las 34 marcas procesadas, se consiguió secuenciar 17 e identificar 10 proteínas (5 proteínas comunes a clones de todas las regiones y 6 específicas de clones de una región) Entre las proteínas comunes a clones de todas las regiones, una de las identificadas fue la subunidad β del receptor de reconocimiento de señales, que forma parte de un complejo implicado en el transporte de proteínas a través de las membranas celulares. .También se identificó la cathepsina β, una peptidasa cisteínica, siendo ésta la primera vez que se identifica una proteasa cisteínica en los productos extracelulares de P. olseni. Se ha constatado el papel relevante de este grupo de peptidasas en otros protistas parásitos, contribuyendo a la invasión y degradación de los tejidos del hospedador. En cuanto a las proteínas exclusivas de clones de una región, tres de ellas correspondían a los clones de la ría de Arousa, pepsina A, que permite al parásito la obtención de nutrientes y energía mediante la degradación de tejidos del hospedador a la par que facilita el progreso de la infección, fosfoserina aminotransferasa, que está implicada en la síntesis de L-serina, componente esencial de la síntesis proteica, y la proteína de choque térmico 60 (HSP60), relacionada con procesos de estrés celular, con papel esencial en la síntesis, transporte y plegado de las proteínas. Además, se identificó una peroxiredoxina correspondiente a una marca exclusiva de la ría de Pontevedra; la función de las peroxiredoxinas en la neutralización de especies reactivas tóxicas de oxígeno liberadas por los hemocitos del hospedador se ha mencionado en el capítulo previo. En suma, la variabilidad entre clones de P. olseni XIII. Resumen 262 de las cuatro regiones del litoral español en las proteínas extracelulares fue mayor que en las celulares; algunas de las diferencias detectadas entre regiones estaban asociadas a virulencia. Las diferencias en virulencia entre clones de regiones diferentes del litoral español deberían estudiarse más a fondo, dada su transcendencia para la gestión de la perkinsosis. Capítulo 4. Comparación de los perfiles de expresión proteínica entre tres especies de Perkinsus, P. olseni, P. marinus y P. chesapeaki. El análisis de la variabilidad del agente etiológico de la perkinsosis se completó a través de la comparación del proteoma celular de tres especies del género Perkinsus. Se partió de un cultivo clonal de cada especie, P. marinus, P. olseni y P. chesapeaki. Tras aislar las proteínas celulares, éstas se separaron mediante electroforesis bidimensional en geles de poliacrilamida (cuatro réplicas analíticas por especie). Los geles se tiñeron con Sypro Ruby, se digitalizaron y se compararon con el programa informático Proteomweaver 4.0. Más de 1600 marcas proteínicas se observaron en los geles de cada especie. Se constató que 213 marcas eran comunes a las tres especies, 310 eran compartidas por P. marinus y P. chesapeaki, 315 por P. chesapeaki y P. olseni, y 242 por P. marinus y P. olseni. Se encontraron marcas específicas de cada especie: 1161 de P. chesapeaki, 1124 de P. olseni y 825 de P. marinus. Un análisis de similitud mostró mayor proximidad entre las especies P. marinus y P. olseni, siendo P. chesapeaki la más alejada de las tres, lo que es coherente con resultados previos basados en la comparación de las secuencias del gen ribosómico y de la actina. Se seleccionaron 14 marcas comunes a las tres especies y 14 marcas específicas de cada especie para su secuenciación por espectrometría de masas y su posterior identificación en bases de datos. Las 14 marcas comunes se separaron por triplicado (una marca de un gel de cada especie). De las 84 marcas procesadas, se logró la secuencia de 42 y 28 secuencias se identificaron con proteínas de las bases de datos del NCBI y SwissProt. Únicamente 6 proteínas comunes pudieron ser identificadas, las subunidades β y g de la ATPasa vacuolar, la malato deshidrogenasa, el receptor de la quinasa C activada, la pseudouridina sintasa, y la fosfato acetiltransferasa. Con respecto a las proteínas específicas, en P. olseni se identificó la formato deshidrogenasa, una subunidad del proteasoma, la triosafosfato isomerasa y la peroxiredoxina V; en P. marinus la fosfogliceto kinasa y la proteína 40S ribosomal S3 y en P. chesapeaki se identificó la glutatión S-transferasa, la peroxiredoxina II y la peroxiredoxina V (ésta en dos marcas). Las peroxiredoxinas y la glutation S-transferasa están implicadas en funciones de detoxificación y antioxidación, pudiendo actuar para neutralizar las especies reactivas tóxicas de oxígeno que genera el hospedador para luchar contra el avance de la infección. También el receptor de la quinasa C activada tiene un papel relevante en la regulación de Ca+2 en el hospedador; permitiría al XIII. Resumen 263 parásito modular procesos como la apoptosis, proliferación celular y disponibilidad de nutrientes. Las restantes proteínas identificadas regulan procesos vitales en el organismo como son el metabolismo, modificaciones post-transcripcionales, diferenciación celular y regulación del pH intracelular del parásito. Capítulo 5. Expresión proteínica de los hemocitos de la almeja japonesa Ruditapes philippinarum en respuesta a la infección por Perkinsus olseni. La hemolinfa es la sangre de los moluscos, en ella se distinguen dos componentes, las células (hemocitos) y el plasma. Los hemocitos están implicados en la reacción inmunitaria de la almeja a través de la fagocitosis o encapsulación de patógenos y su posterior destrucción y liberación de moléculas implicadas en la respuesta inmune. El plasma también contiene factores implicados en la respuesta inmunitaria, como lectinas, péptidos antimicrobianos, lisozimas o inhibidores de proteasas liberadas por parásitos, entre otros. Para conocer la respuesta inmunitaria de la almeja japonesa R. philippinarum ante la infección por P. olseni, se analizó la modulación del perfil de expresión proteínica de los hemocitos y del plasma de la almeja cuando ésta se enfrenta al parásito. En este capítulo se recoge el estudio correspondiente a los hemocitos y en el siguiente capítulo el correspondiente al plasma. Se plantearon dos situaciones experimentales, para conocer los efectos del enfrentamiento a corto plazo así como tras un enfrentamiento prolongado. En el análisis de los efectos a corto plazo se utilizaron almejas recogidas en un banco natural de Camariñas (A Coruña), donde nunca se había detectado la presencia de P. olseni. De ellas, 40 almejas se expusieron en vasos individuales a 106 zooesporas de P. olseni en agua de mar durante 24 horas, tras lo que se mantuvieron en un tanque durante 7 días; Otras 40 almejas se sometieron a las mismas condiciones pero sin añadir zooesporas a los vasos, para utilizarlas como control del experimento. Se extrajo hemolinfa de cada almeja, mezclándose la hemolinfa por grupos de 20 almejas, para producir así dos réplicas biológicas de almejas enfrentadas al parásito y otras dos de almejas no enfrentadas. Las muestras de hemolinfa se procesaron para separar los hemocitos del plasma y analizar la expresión proteínica, comparando la de los hemocitos de almejas enfrentadas al parásito con la de los hemocitos de almejas no enfrentadas. En el análisis de la exposición prolongada al parásito, se recogieron almejas de tamaño comercial (al menos dos años de edad) en un banco natural afectado por P. olseni, con valores altos de intensidad de infección, localizado en Vilalonga (ría de Arousa). Se trataba por tanto de almejas expuestas de manera natural al parásito durante un tiempo prolongado. Se extrajo la hemolinfa de 200 almejas y se diagnosticó la intensidad de la perkinsosis de cada una de ellas mediante el método de incubación de branquia en caldo de tioglicolato; se seleccionó la hemolinfa de 42 almejas que XIII. Resumen 264 presentaban una infección avanzada (entre 3 y 5 en la escala de Mackin) así como la hemolinfa de otras 42 almejas en que no se detectó infección. La hemolinfa de las almejas seleccionadas se mezcló por grupos de 14 almejas, para producir así tres réplicas biológicas de almejas con infección avanzada y otras tres de almejas no infectadas. Las muestras de hemolinfa se procesaron para separar los hemocitos del plasma y analizar la expresión proteínica, comparando la de los hemocitos de almejas con infección avanzada con la de las almejas no infectadas. En ambos planteamientos experimentales, las proteínas hemocitarias se aislaron y se separaron mediante electroforesis bidimensional en geles de poliacrilamida. Los geles se tiñeron con plata, se digitalizaron y se compararon (almejas expuestas frente a no expuestas en la exposición en laboratorio; almejas con infección avanzada frente a almejas no infectadas en la exposición natural prolongada) usando el programa informático PD Quest 8.4.0. Los geles mostraron un promedio de 570 marcas proteínicas en ambos planteamientos experimentales. No se detectaron diferencias significativas en el número de marcas por gel entre tratamientos. En el caso de la exposición a P. olseni en laboratorio, se han observaron tres marcas exclusivas en muestras de almejas expuestas y cinco exclusivas en las almejas no expuestas; en el caso de exposición prolongada en el medio natural, se detectaron dos marcas exclusivas en muestras de almejas infectadas y una en las no infectadas. Todas esas marcas se seleccionaron para su secuenciación por espectrometría de masas y posterior identificación en bases de datos; de ellas sólo siete pudieron ser identificadas. Parece que la exposición a P. olseni en el laboratorio indujo la expresión de la proteína de choque térmico 70 (HSP 70) 12B y la integrina-α PS3 en los hemocitos de las almejas, pues ambas proteínas se identificaron en marcas exclusivas de almejas expuestas. La proteína HSP 70 12B está implicada en la respuesta inmune de los moluscos bivalvos; se ha constatado su sobreexpresión en varias especies de moluscos expuestas a patógenos. La integrina-α PS3 es una glicoproteína implicada en el proceso de encapsulación de patógenos por hemocitos, así como en la fagocitosis de células apoptóticas y de bacterias, en varios invertebrados. Por el contrario, la exposición al parásito en el laboratorio pudo haber inhibido la expresión de citocromo C oxidasa I, actina y ankyrina-3, puesto que se identificaron en marcas exclusivas de hemocitos de almejas no expuestas. La disminución de la expresión de actina parece una respuesta común de los invertebrados frente a la infección. Actina y Ankyrina están implicadas en procesos de movilidad celular, regulación del citoesqueleto e incluso de apoptosis, mientras que la citocromo C oxidasa I está implicada en la obtención de energía de la cadena de transferencia de electrones en la mitocondria. En el caso de la exposición prolongada al parásito en el medio natural se identificó el regulador posttranscripcional ATRX, correspondiente a una marca XIII. Resumen 265 exclusiva de hemocitos de almejas no infectadas. Esta proteína está implicada en varios procesos celulares como, metilación del DNA, transcripción, ciclo celular y apoptosis. Su función en procesos de respuesta inmune no está bien dilucidada. La Rho GTPasa 6 se identificó en una marca exclusiva de hemocitos de almejas no infectadas, por tanto podría tratarse de una proteína cuya expresión se inhibe por la infección o bien de una proteína cuya presencia impide que la almeja se infecte por P. olseni o que la infección no progrese. Si éste fuese el caso, esta proteína podría constituir un marcador molecular de resistencia a la perkinsosis. Su función está relacionada con la organización de actomiosina, adhesión, y proliferación; aunque también forma parte de rutas de señalización usadas por receptores de antígenos que regulan la respuesta inmune. La enorme importancia de contar con un marcador molecular de resistencia a la perkinsosis para implementar programas de selección genética para producir estirpes de almeja resistentes a la perkinsosis, aconseja analizar a fondo si realmente la Rho GTPasa 6 es un marcador de resistencia al parásito. Capítulo 6. Estudio de la expresión proteínica del plasma de almejas Ruditapes philippinarum en respuesta a la infección por Perkinsus olseni. El planteamiento experimental es el expuesto en el capítulo previo. En este caso, las muestras de plasma se procesaron para aislar las proteínas y separarlas mediante electroforesis bidimensional en geles de poliacrilamida. Los geles se tiñeron con plata, se digitalizaron y se compararon. No se detectaron diferencias significativas en el número de marcas proteínicas entre muestras de plasma de almejas expuestas y no expuestas al parásito en el laboratorio ni entre almejas con infección avanzada y almejas no infectadas tras exposición prolongada en el medio natural. Sin embargo, el número de marcas en los geles del plasma de las almejas del experimento de exposición prolongada (174 marcas por gel de media) fue significativamente menor que el de las almejas del experimento de exposición en laboratorio (389 marcas de media), lo que podría deberse a las diferencias en las condiciones ambientales, con gran influencia en el plasma. Sólo se encontraron marcas exclusivas de tratamiento en los geles del experimento reexposición en laboratorio, con 15 marcas exclusivas del plasma de las almejas no expuestas y 10 marcas exclusivas de las almejas expuestas. Todas esas marcas se procesaron para su secuenciación y posterior identificación en bases de datos; de ellas únicamente se identificaron 14. Parece que la exposición de las almejas a P. olseni en laboratorio indujo en el plasma la expresión de la proteína X1 interactuante con Pin2, la demetilasa 3B especifica de lisina, la proteína de unión de calcio sarcoplasmático y de una lisozima, puesto que dichas proteínas se identificaron en marcas exclusivas de plasma de almejas expuestas. La proteína X1 interactuante con Pin2 está implicada en la regulación de la telomerasa. La demetilasa 3B especifica de lisina provoca cambios en