Upper ocean oxygenation, evolution of RuBisCO and the Phanerozoic succession of phytoplankton

Abstract

Evidence is compiled to demonstrate a redox scale within Earth's photosynthesisers that correlates the specificity of their RuBisCO with organismal metabolic tolerance to anoxia, and ecological selection by dissolved O₂/CO₂ and nutrients. The Form 1B RuBisCO found in the chlorophyte green algae, has a poor selectivity between the two dissolved substrates, O₂ and CO₂, at the active site. This enzyme appears adapted to lower O₂/CO₂ ratios, or more "anoxic" conditions and therefore requires additional energetic or nutrient investment in a carbon concentrating mechanism (CCM) to boost the intracellular CO₂/O₂ ratio and maintain competitive carboxylation rates under increasingly high O₂/CO₂ conditions in the environment. By contrast the coccolithophores and diatoms evolved containing the more selective Rhodophyte Form 1D RuBisCO, better adapted to a higher O₂/CO₂ ratio, or more oxic conditions. This Form 1D RuBisCO requires lesser energetic or nutrient investment in a CCM to attain high carboxylation rates under environmentally high O₂/CO₂ ratios. Such a physiological relationship may underpin the succession of phytoplankton in the Phanerozoic oceans: the coccolithophores and diatoms took over the oceanic realm from the incumbent cyanobacteria and green algae when the upper ocean became persistently oxygenated, alkaline and more oligotrophic. The facultatively anaerobic green algae, able to tolerate the anoxic conditions of the water column and a periodically inundated soil, were better poised to adapt to the fluctuating anoxia associated with periods of submergence and emergence and transition onto the land. The induction of a CCM may exert a natural limit to the improvement of RuBisCO efficiency over Earth history. Rubisco specificity appears to adapt on the timescale of ∼100 Myrs. So persistent elevation of CO₂/O₂ ratios in the intracellular environment around the enzyme, may induce a relaxation in RuBisCO selectivity for CO₂ relative to O₂. The most efficient RuBisCO for net carboxylation is likely to be found in CCM-lacking algae that have been exposed to hyperoxic conditions for at least 100 Myrs, such as intertidal brown seaweeds.

Graphical abstract

Fig. 1

A comparison of algal biomarker records from rock and oil samples with the I/Ca record of upper ocean oxygenation. The C28/C29-sterane ratio of 500 rock samples are plotted, averaged in 50 Myr steps (rhomboids), and in 25 Myr steps (filled circles) compared with 400 oil samples (squares) analysed by Grantham and Wakefield [8] and presented in Schwark and Empt, [2]. Candlestick plot showing ranges of I/Ca values for Paleozoic (dark blue) and Meso-Cenozoic (light blue) from Lu et al., [7]. Boxes mark the 25th and 75th percentiles of values at each locality, and the whiskers show the maximum and minimum. Also shown in grey bars are the inferred abundances of different algal groups throughout the Phanerozoic. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.).

Fig. 2

a) The sensitivity of the equilibrium dissolved ratio of O₂/CO₂ (CO₂ [22]) (O₂ [23]) concentrations to temperature and salinity (S; 0, open circles, and 35 ppt, closed circles) for the modern (with an atmosphere of 400 ppmV) compared to that at the LGM with invariant O₂ but a CO₂ atmosphere of 180 ppmV. This environmental O₂/CO₂ provides a calibration for the redox gradient to RuBisCO of different algal groups and their ecology showing the relative substrate affinities K_o/K_c of RuBisCO as a first order determinant of RuBisCO specificity. Species abbreviation labels Rhodophyta: Gsu Galdieria sulfiraria, Gmo Griffithsia monilis, Pcr Porphyridium cruentum, Haptophytes: Plu Pavlovale lutheri, Pca Pleurochrysis carterae, Iga Isochrysis galbana, Heterokonts: Phaeodactylum tricornutum, Cyl cylindrotheca spp, Cfu Cylindrus fusiformis, Cmu Chaetoceros muellerae, Chrysophyte: Olu Olisthodiscus luterae, Fra Fragilariopsis sp, Cca Chaetoceros calcitrans, Tps Thalassiosira pseudonana, Sco: Skeletonema costatum, Green algae: Cre Chlamydomonas reinhardtii, Sob Scenedesmus obliquus, Egr Euglena gracilis, Cyanobacteria: Syn synechococcus, Pro Prochloroccus Plants: C3: Triticum aestivium C4 Zea Mays, Anaerobes: Tde Thiobacillus denitrans, Rsp Rhodobacter sphaeroides, Rru Rhodospirillum rubrum, Mbe Methanococcoides burtonii, Tko Thermococcus kodakarensis. Also labeled are mean ecologies of different groups of algae ranging from obligate anaerobes, through facultative anaerobes to obligate aerobes and hyperoxic tolerant. b) The number of anaerobic metabolic pathways in the genomes (PFL, PFL-AE, PFO/PNO, HYDA, HydEF, HydG, ADHE, ACK, PTA, ASCT, ADP-ACS) of the labeled organisms where RuBisCO specificity has also been determined taken from Atteia et al., [24]. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.).

Fig. 3

Phylogenetic Tree for RbcL within Haptophyta showing branches under positive selection (magenta) and those with no positive selection (black) adapted from Young et al., [48]. Black numbers above branch is statistical significance (p-value) of positive selection after a likelihood ratio test comparing nested models and using Bonferroni correction. Grey bars denote 95% confidence intervals for date estimates and grey numbers are posterior probability values. Yellow diamonds with corresponding letter are fossil calibration dates. Also indicated is the presence/absence of some components of a CCM with a blue tick for the presence of a δCA (methodology from Heureux et al. [20]), and a red bar for the absence or presence (blue) of a pyrenoid (Eason Hubbard unpublished; P. lutheri (now Diacronema lutheri) [50,51] Pavlova pinguis [52,53], Pavlova salina (now Rebecca salina) [51,53], Exanthemachrysis gayraliae [51,54], Phaeocystis globosa [55], Phaeocystis pouchetii [56], Hyalolithus neolepis (now Prymnesium neolepis) [57], Prymnesium parvum [58], Prymnesium patelliferum (now Prymnesium parvum haploid stage) [59], Imantonia rotunda (now Dicrateria rotunda) [60], Gephyrocapsa oceanica [61], Helicosphaera carteri (Eason Hubbard unpubl), Emiliania huxleyi [62,63], Isochrysis galbana [64], Chrysotila lamellosa [65], Pleurochrysis placolithoides [66], Pleurochrysis carterae [67], Pleurochrysis elongate [68], Coccolithus pelagicus [69], Calcidiscus leptoporus (Eason Hubbard unpubl), Umbilicosphaera sibogae var foliosa [70], Calyptrosphaera sphaeroidea (now Holococcolithophora sphaeroidea) [71], Cruciplacolithus neohelis [72]. Also indicated is the RuBisCO specificity in numerical values where data is available. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

a) Comparison between RuBisCO specificity (dark green bars) and the Redfield ratio (C:N measured in cells under exponential growth, light green bars [73]). These are data measured on a wide variety of species under the exact same conditions, important for internal consistency given the propensity for plasticity in the Redfield ratio. Here the Redfield ratio is interpreted to reflect nutrient efficiency of C fixation and C:N is higher when carbon fixation is more nutrient efficient and requires less proteins of a CCM. B) The relationship between higher RuBisCO specificity and larger carbon isotopic fractionation of the RuBisCO enzyme in vitro taken from Tcherkez et al. [74], and updated with data for S. costatum from Boller et al., [75]. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.).