Genotype specific feedback responses towards a protistan grazer

The results within Publication IV of this thesis show that the interaction between a protistan grazer (Polykrikos kofoidii) and A. tamarense at the transcriptomic level is even more multifaceted than the previously discussed responses towards copepod grazers (Publication I & III). This study differs from the above investigated induced defense towards copepods in the sense that the associated A. tamarense strains are able to modify the interactions in a genotypic specific manner. Hence, the presence of the lytic A. tamarense strain A2 has an influence on the fitness of heterospecifics and conspecifics, which, in turn, feeds back to influence the phenotype expressed by this strain. This feedback mechanism was not given for the second investigated genotype of A. tamarense (A5) since this phenotype lacks the ability to produce lytic compounds. The lytic strain A2 therefore has the potential to alter its niche, a behavior that in turn influences the traits expressed by this genotype

8.2.1  Genotype specific gene expression patterns 

A highly intraspecific variation in gene expression in control conditions between the two strains of A. tamarense was observed, comprising 5% of all genes on the microarray. The variability in traits expressed by these two strains hence might go beyond the described changes in secondary chemistry (PSTs and lytic compound production). However, it cannot be excluded that the presence of the caged strain A5 caused those changes in the strain A2, at least to some extent.

However, the fold change values are unusually high (max. 900 FC) and such values were never observed in response to any treatment tested so far in Alexandrium tamarense (Publication I; III & IV), supporting the idea that intraspecific variation in gene expression generated those observed gene expression patterns. This conclusion is further supported by literature data, showing that i.e. intraspecific variation in gene expression in Arabidopsis spp. is a greater source of variance than treatment effects (Van Leeuwen et al. 2007, Kliebenstein 2008) The correlation of transcript copy number and genomic abundance of those genes in dinoflagellates (Bachvaroff & Place 2008) renders gene duplication events as a source of such observed transcriptional differences between strains. Such correlations between gene duplication and increased gene expression diversity within species have also been found for many other species (i.e. Arabidopsis thaliana, Drosophila melanogaster, Saccharomyces cerevisiae) (Gu et al. 2004, Landry et al. 2006, Kliebenstein 2008) and are considered to be important for shaping specific phenotypes as well as for species differentiation (Gu et al. 2004).

The observation of the occurrence of high intraspecific variation in gene expression can therefore be a valuable source for our understanding of the evolution and variance of specific traits not only in A. tamarense for several reasons:

I) In general, gene duplication events are a major source for the evolution of new traits and thereby facilitate adaptation to a dynamic environment.

II) Gene duplication and expression noise might decrease with increasing constraints acting on the respective genes’ function, which, in turn develops over time (Kliebenstein 2008). The highest rate of intraspecific gene expression variation can therefore be expected to occur in recent duplicated genes and those genes can indicate which

traits are currently under selection. The ability of dinoflagellates to reintegrate mRNA into the genome could furthermore significantly contribute to the expression and diversification of environmentally important traits.

III) Comparing intraspecific variations in gene expression patterns between different populations could enable tracing of traits responsible for local adaptations.

8.2.2  Genotype specific trait alterations inferred from functional genomics 

The results in Publication IV showed that a large range of the genome’s function of A. tamarense is involved in shaping the phenotype due to the presence of a protistan grazer. Based on functional annotations of Pfam families covering all genes on the microarray (Publication V), around 3% of those specific Pfam families were involved in the response to the respective biotic interaction. In addition to the integration of the response into different genotypes, the strain specific alteration of the co-occurring species interactions might have a major influence on the observed changes in the gene expression patterns (Publication IV). The production of allelochemical compounds in the strain A2 with a lytic effect on co-occurring heterospecifics (Tillmann & Hansen 2009), may indeed construct an environment different from the one that the strain A5 experiences under the otherwise comparable experimental conditions. The genotype specific modification of the environment might therefore be the cause of a comparable increased expenditure of cellular resources to build up endocytotic processes in the strain A2 than in the strain A5. As changes in the phenotype are aimed to increase fitness under changed conditions, it would seem mixotrophy increases fitness in A. tamarense even under optimized laboratory conditions. However, neither increased growth rates nor the presence of food vacuoles could be observed for the two A. tamarense strains (Tillmann & Hansen 2009), but are described for other A. tamarense strains (Jeong et al. 2005, 2010). Yet, the occurrence of food vacuoles might depend on the strains and culture conditions used. Batch cultures with ample nutrients might already thrive at maximum growth rates and benefits could therefore be hardly measureable. Other parameters such as nutrient stoichiometry or an increase in storage compounds might be better suited for identifying beneficial effects of mixotrophy. In addition, non-formation of food vacuoles large enough to be observed with light microscopy does not exclude the existence of endocytotic

processes per se. In any case, lytic compound production in A. tamarense seems not only to benefit the population by reducing grazing pressure from protistan grazers.

The recognition and translation of the cues emitted from the environment into different genotypes should have a common transcriptional catalyst that is specific to each interaction, as already discussed for the strain specific response towards copepod grazers. Hence, in the case of Polykrikos kofoidii grazing on conspecifics, the cues elicit transcriptional changes in both strains for genes that drive the conversion of diacylglycerol (DAG) into phosphatidic acid (PA), indicating that G-protein coupled receptors and DAG/PA signaling maybe the underlying intracellular response inducer (Publication IV). In addition, both strains respond with transcriptional changes that outnumber the observed gene expression changes in the copepod interactions (Publication I & III). Both strains further induce the expression of genes involved in DNA m⁶adenine methylation, implying that a high transcriptional activity is largely driven due to a reduction in DNA duplex stability.

Given that dinoflagellates lack or have a low abundance of histones (Lin et al. 2010, Lin 2011, Wisecaver & Hackett 2011) m⁶adenine methylation could be a convergent mechanism to histone acetylation that accelerates or even enables large transcriptional changes.

The differentially expressed genes induced by cues emanating from co-occurring species interactions further point towards changes in the outer glycosylated membrane surface. The outer membrane is the first contact area between an organism and its external environment, and therefore represents the bridge between the environment and associated intrinsic cellular processes and responses. Hence, the outer glycosylated membrane surface determines important traits such as self-self recognition, non-self discrimination as well as “cell taste”

(Gahmberg 1981, Wolfe 2000). Changes in its characteristics can subsequently influence further biotic interactions. Thus, the observed ranges of responses towards a protistan grazer provide strong implications that such feedback responses can shape community ecology processes in an unpredictable manner. In other words, if the presence of a protistan grazer modifies the niche of A.

tamarense (due to cell lysis) and expressed traits (setting-up of endocytosis and changes in membrane glycosylation), the outcome of interactions with a further species can differ from the interaction that would have occurred if the protistan grazer would not have been present. Consequences of species interaction therefore have the potential to scale up and might be one reason for the challenge of