The Regulation of Carbon and Nutrient Assimilation in Diatoms is Significantly Different from Green Algae

Diatoms are the major players in the biochemical cycles of carbon, nitrogen, phosphorus, and silicon with a strong impact on global climate not only in the ocean but also in the freshwater environment. In the ocean, they tend to dominate the phytoplankton assemblage under nutrient-rich conditions; whereas, in freshwater ecosystems, high turbulence and the combination of low temperature/high nutrients favour diatom blooming. Their fascinating biodiversity with approximately 10,000 species has been recently taxonomically revised giving evidence that the diatoms form two clades: the Coscinodiscophytina and the Bacillariophytina, the latter containing two subclades Bacillariophyceae and Mediophyceae; all the more this tremendous species diversity evolved during the last 250 Ma and there is likely no phylogenetic group in which the evolutionary rate has occurred faster (Medlin and Kaszmarska 2004). Therefore, the diatoms can be considered as one of the most successful taxonomic groups with respect to evolution and ecology. However, the genetic and physiological basis why diatoms became so ecologically successful is still a matter of debate. The recently published genome of Thalassiosira pseudonana (Armbrust et al. 2004) and the EST sequence data available from Phaeodactylum tricornutum (Maheswari et al. 2005) have opened new perspectives in diatom research to understand in more detail their physiological peculiarities. The metabolism in oxygenic phototrophs depends on gene expression processes coordinated between the three different compartments: the nucleus, the plastid, and the mitochondrion. Diatoms originate from a secondary endosymbiotic event (see Fig. 1) as opposed to the primary endosymbiotic origin of the "Chloroplastida" (Adl et al. 2005), including among others, the charophytes, chlorophytes and prasinophytes. There is convincing evidence that the ancestral diatom derived from a putatively non-photosynthetic eukaryote which domesticated a eukaryotic cell phylogenetically close to a red alga: this "red line phylogenetics is clearly a work in progress" (Palmer 2003) because we neither know the nature of the secondary host nor whether all "chromophytes" originate from one single endocytobiotic event (Patron et al. 2004). Gene analyses based on whole genome sequencing of T. pseudonana revealed that the diatom genome contains about 800 proteins which align with proteins of animals but do not have homologues in the genome of Arabidopsis thaliana and the red alga Cyanidioschyzon merolae (Armbrust et al. 2004). This suggests that in diatoms the cytosolic metabolic pathways might be closely related to an animal-like rather than to a plant-like cell, which would likely affect the physiological regulation of photosynthesis. Considering the obligatory intracellular gene transfer in chromophytic phototrophs from the vanishing nucleus of the endosymbiont to the nucleus of the secondary host cell, various genes for proteins involved in photosynthesis had to be integrated into the new, more animal-like nucleus. Therefore, we would expect a different pattern in the nucleome-chondriome-plastome interaction with the consequence that metabolic pathways related to these compartments had been lost or that the respective genes had been transferred to the nucleus of the secondary endosymbiont. Therefore, the plastid had to develop a completely new interaction pattern with its genetic partners. To date, this regulatory network in diatoms is not yet understood. The stabilisation of a secondary endosymbiosis did not only require the development of protein targeting from the cytosol into the plastids crossing four membranes (Kilian and Kroth 2004), but also required a signal exchange in the reverse direction from the organelle to the nucleus. In the last few years it has been shown that in green plant cells photosynthesis-related nuclear genes are under chloroplast redox control by a new class of plastid signals (Goldschmidt-Clermont 1998). The sensor for this signal is the redox state of the plastoquinone (PQ) pool which depends on the relative activity of photosystems II and I (PS II and PS I). However, the redox state of the chloroplast does not only depend on the membrane-bound reactions, but also on the oxidative state of the stroma which is under the control of the antioxidative system via thioredoxin (TR), gluthathione, ascorbate, and their related metabolising enzymes. Although the biochemical nature of this retrograde signalling from the plastid to the nucleus is still not understood, the importance of this redox-controlled nuclear gene expression has recently been shown by Fey et al. (2005) demonstrating that this signal triggers the whole cellular network of gene expression. This retrograde plastid to nucleus signalling must be basically different in green algae and diatoms because of two reasons: first, the cooperating nuclei are of different phylogenetic origin, and second, the redox control in green algae and diatoms is found to be basically different as outlined below (see sections on "Regulation of photosynthetic electron flow and the fate of photosynthetic electrons" and "Regulation of the enzyme activity in the Calvin cycle in diatoms"). However, we do not know what the consequences of the secondary endocytobiosis on the metabolic regulations were, nor whether the physiological peculiarities of diatoms observed in the past reflect this different host-endosymbiont relationship. In this review, we focus on cell functions which clearly separate the diatoms physiologically from algae of the chloroplastidic lineage. •The thylakoid membranes in diatoms are arranged in groups of three and are not differentiated into grana/stroma lamellae (Berkaloff et al. 1990; Owens 1988) having consequences not only for the PSII/PSI ratio and the regulation of the wavelength-dependent energy distribution between both PSs via so-called state 1-state 2 transitions, but also for the redox signalling (Owens 1986).•The light-harvesting proteins of diatoms are not differentiated into minor and major complexes and the protein domains which trigger membrane stacking in Chloroplastida by modulation of phosphorylation do not exist (Larkum and Vesk 2003; Westermann and Rhiel 2005).•In Chloroplastida, photoprotection is supported by cycling of xanthophylls bound to the minor antenna proteins in the presence of the PsbS protein. In diatoms, xanthophylls contribute even more to the photoprotection; however, in the genome of T. pseudonana, neither a psbS gene nor genes encoding for minor light-harvesting units could be identified.•Diatoms lack the α-carotene pathway with the consequence that photoprotective and light-harvesting pigments are synthesised on the same metabolic branch asking for sophisticated mechanisms of regulation (Lohr and Wilhelm 1999).•In contrast to Chloroplastida, metabolic activity in long dark periods leads to an enhanced reduction state of the PQ pool which is accompanied by the build-up of a proton gradient strong enough to drive the de-epoxidation reaction of xanthophyll cycle (XC) pigments (Jakob et al. 1999). Here again, the impact of unusual redox control of gene expression is obvious.•In chloroplasts of the Chloroplastida, photosynthetic CO₂ fixation is regulated by light via the redox state of thiol groups. There is now evidence that in diatoms, the Calvin cycle is controlled by a different mechanism and that the oxidative pentose phosphate cycle (OPP) is not present (Michels et al. 2005).•It has been observed that inhibitory effects of a high concentration of oxygen/low concentration of CO₂ on photosynthetic activity are less pronounced in diatoms than in green algae (Beardall 1989; McMinn et al. 2005). This has been discussed in the light of an active transport and accumulation of inorganic carbon within the cell (Beardall 1989) and the C4 fixation metabolism in diatoms (Reinfelder et al. 2004). However, some crucial plastid-localised enzymes being prerequisites in this metabolic pathway could not be identified in the diatom genome.•The metabolic carbon flow in diatom photorespiration is not yet elucidated. However, there is clear evidence that in diatoms, the photorespiratory carbon flux is not coupled to the nitrogen flux as in the typical carbon-oxidation cycle in green algae. Surprisingly, the genome analysis of T. pseudonana revealed that in contrast to green phototrophs, the diatoms possess a complete set of genes for a functional urea cycle (Armbrust et al. 2004). In this context, it might be noteworthy that the genes encoding the enzymes for ß-oxidation and catalase were found in the genome (Armbrust et al. 2004); however, it is not yet clear whether this pathway is located in the peroxisome.•Diatoms synthesise lipids and the ß-1,3 glucan chrysolaminarin as energy storage products. In contrast to starch, this polymer does not form crystals. In green algae, the carbon storage product starch induces cell sinking; whereas in diatoms, the storage product lipid can increase the cell′s buoyancy.•Several in situ iron fertilisation experiments have demonstrated that diatoms can survive even in areas that are severely iron limited (e.g. high nutrient low carbon [HNLC] regions), and have the potential to form large blooms upon relief of the iron limitation (Boyd et al. 2000). This response can in part be explained by the plasticity of the Fe:C ratio in diatoms, possibly achieved by changes in the stoichiometry of photosynthetic proteins and the replacement of ferredoxin by flavodoxin (LaRoche et al. 1996; Strzepek and Harrison 2004).