A new widespread subclass of carbonic anhydrase in marine phytoplankton

Abstract

Most aquatic photoautotrophs depend on CO₂-concentrating mechanisms (CCMs) to maintain productivity at ambient concentrations of CO₂, and carbonic anhydrase (CA) plays a key role in these processes. Here we present different lines of evidence showing that the protein LCIP63, identified in the marine diatom Thalassiosira pseudonana, is a CA. However, sequence analysis showed that it has a low identity with any known CA and therefore belongs to a new subclass that we designate as iota-CA. Moreover, LCIP63 unusually prefers Mn²⁺ to Zn²⁺ as a cofactor, which is potentially of ecological relevance since Mn²⁺ is more abundant than Zn²⁺ in the ocean. LCIP63 is located in the chloroplast and only expressed at low concentrations of CO₂. When overexpressed using biolistic transformation, the rate of photosynthesis at limiting concentrations of dissolved inorganic carbon increased, confirming its role in the CCM. LCIP63 homologs are present in the five other sequenced diatoms and in other algae, bacteria, and archaea. Thus LCIP63 is phylogenetically widespread but overlooked. Analysis of the Tara Oceans database confirmed this and showed that LCIP63 is widely distributed in marine environments and is therefore likely to play an important role in global biogeochemical carbon cycling.

Fig. 1

CA activity of purified recombinant LCIP63. a CA activity with and without treatment with 5 mM EDTA or EDTA plus 10 mM of either Zn²⁺, Mn²⁺, Ca²⁺, Co²⁺, or Mg²⁺. n.s, not significant; ***p ≤ 0.001. Error bars show mean ± SD, n = 3–19. ‘a’ indicates values compared to the EDTA-treated sample. b CA activity in LCIP63 treated with 20 µM acetazolamide (AZA) versus control; *p ≤ 0.05. c Esterase activity of LCIP63 expressed as the production of p-nitrophenol (µmol l⁻¹). The arrow represents the time when the substrate p-nitrophenyl acetate was added to the reaction mixture. d Phylogenetic tree constructed from sequences of the different CA subclasses; numbers in parentheses are the mean percent identity with LCIP63. Pfal, Plasmodium falciparum; Prei, Plasmodium reichenowi; Plasmo, Plasmodium sp.; Pgab, Plasmodium gaboni; Pt, Phaeodactylum tricornutum; Fs, Fistulifera solaris; Fc, Fragilariopsis cylindrus; Tp, Thalassiosira pseudonana; Ehux, Emiliania huxleyi; Lpol, Lingulodinium polyedra; To, Thalassiosira oceanica; Esil, Ectocarpus siliculosus; Blasto, Blastocystis sp.; Pinf, Phytophtora infestans; Phal, Plasmopara halstedii; Sjap, Saccharina japonica; Tw, Thalassiosira weissflogii; Pm, Pseudo-nitzschia multiseries

Fig. 2

Transmission electron micrographs of immunogold labeled LCIP63 in T. pseudonana. a–e Cells cultured at 50 ppm CO₂. c Close-up of the rectangular box in (a). f, g Cells cultured at 20,000 ppm CO₂. Gold particles are indicated by yellow arrowheads. Ch, Chloroplast; Pyr, Pyrenoid. Bar = 500 nm

Fig. 3

LCIP63 overexpression and effect on photosynthesis in T. pseudonana. a Western blot against LCIP63 (upper) and SDS-PAGE to verify protein loading (lower) for control cells and three LCIP63-overexpressing clones. b Rate of net photosynthesis versus concentration of DIC. EV corresponds to the clone transformed with an empty vector; C1, C2, C3, to clones 1, 2, and 3 overexpressing LCIP63. Theoretical curves showing a close-up of fits obtained with experimental data (3 replicates). Fits were performed using Michaelis–Menten equation; dashed vertical lines represent the mean of K_0.5 [DIC] that are indicated in parentheses. All parameters and statistical analysis are given in Table 1

Fig. 4

Characterization of two oligomeric forms of LCIP63. a Elution profile of LCIP63 using size-exclusion chromatography. Arrows indicate elution volumes of proteins used for calibration: 0, Void volume/Blue dextran (2000 kDa); 1, ferritin (440 kDa); 2, catalase (232 kDa); 3, glyceraldehyde-3-phosphate dehydrogenase (150 kDa); 4, bovine serum albumin (68 kDa); 5 ovalbumin (45 kDa). b CA activity of the high molecular mass form (HMM, eluted at 56 ml) and the low molecular mass form (LMM, eluted at 64 ml). c SDS-PAGE of LCIP63 (2 µg): (1) ladder, (2) HMM fraction, (3) LMM fraction. d Schematic representation of the primary structure of LCIP63 from six diatom species. Yellow and green boxes represent ER transit and chloroplast targeting peptides, respectively, light blue boxes represent CaMKII-AD and dark blue boxes the ‘(H)HHSS’ motif. e, f CA activity of the high molecular mass (HMM) and low molecular mass (LMM) forms of the LCIP63 variant containing three (e) and two (f) domain repeats. **p ≤ 0.01; ***p ≤ 0.001. Error bars show mean ± SD, n = 3–6

Fig. 5

Alignment of the LCIP63 sequence from T. pseudonana and other organisms. a Alignment with the second LCIP63 domain was made with Mega6 software and analyzed in GeneDoc (http://www.psc.edu/biomed/genedoc). This domain has an identity of 60–68% with the other three LCIP63 domains and was the most similar to the sequences from the other species. The sequences with the highest score from diatoms (orange), bacteria (black), cyanobacteria (blue), green algae (green), and other Chromista algae (brown) are shown. Shading levels correspond to conserved amino acids: Black, 100% identity; dark gray, 80% identity; light gray, 60% identity. Complete alignment is given in Fig. S4. b Phylogenetic tree of proteins containing LCIP63-like domains. Bootstrap values are shown between nodes. The scale represents the number of substitutions per site. Colors are as in panel (a). More information about the sequences are in table S1

Fig. 6

Geographic and taxonomic distribution of LCIP63 homolog sequences found in the OGA webserver. a Geographic distribution and abundance of LCIP63 homologs in eukaryotes in surface water (SRF; upper panel) and deep chlorophyll maximum layer (DCM; lower panel). b Taxonomic distribution of all the hits found in the OGA webserver for eukaryotes. c Geographic distribution and abundance of LCIP63 homologs in prokaryotes and picoeukaryotes in SRF (upper panel), DCM (middle panel), and mesopelagic zone (MES; lower panel). d Taxonomic distribution of all the hits found for prokaryotes and picoeukaryotes. The size of the filled circles shown in (a, c) are proportional to the abundance of the hits in one location compared to the total number of hits at the different sampled depths, and colors represent the size fractionation range