Thylakoid luminal θ-carbonic anhydrase critical for growth and photosynthesis in the marine diatom Phaeodactylum tricornutum

Abstract

The algal pyrenoid is a large plastid body, where the majority of the CO2-fixing enzyme, ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) resides, and it is proposed to be the hub of the algal CO2-concentrating mechanism (CCM) and CO2 fixation. The thylakoid membrane is often in close proximity to or penetrates the pyrenoid itself, implying there is a functional cooperation between the pyrenoid and thylakoid. Here, GFP tagging and immunolocalization analyses revealed that a previously unidentified protein, Pt43233, is targeted to the lumen of the pyrenoid-penetrating thylakoid in the marine diatom Phaeodactylum tricornutum The recombinant Pt43233 produced in Escherichia coli cells had both carbonic anhydrase (CA) and esterase activities. Furthermore, a Pt43233:GFP-fusion protein immunoprecipitated from P. tricornutum cells displayed a greater specific CA activity than detected for the purified recombinant protein. In an RNAi-generated Pt43233 knockdown mutant grown in atmospheric CO2 levels, photosynthetic dissolved inorganic carbon (DIC) affinity was decreased and growth was constantly retarded; in contrast, overexpression of Pt43233:GFP yielded a slightly greater photosynthetic DIC affinity. The discovery of a θ-type CA localized to the thylakoid lumen, with an essential role in photosynthetic efficiency and growth, strongly suggests the existence of a common role for the thylakoid-luminal CA with respect to the function of diverse algal pyrenoids.

Keywords: CGHR domain; CO2-concentrating mechanism; luminal carbonic anhydrase; marine diatom; pyrenoid.

Fig. 1.

Structural comparison of CGHR family members in the P. tricornutum genome and conserved regions in CGHR family proteins from other photoautotrophs. (A) Schematic drawing of P. tricornutum CGHR family genes on chromosome 1. Magenta arrows indicate the cDNA sequence regions for CGHR family genes, and the numbers above the arrows represent their corresponding lengths. The numbers shown below the arrows indicate the length of noncoding regions. (B) Comparison of the conserved regions in CGHR family proteins from bacteria, cyanobacteria, and eukaryotic algae. (C) Schematic drawings of notable domain structures of four P. tricornutum CGHR family members.

Fig. S1.

Phylogenetic tree of CGHR family proteins. Bootstrap values are shown at the node of the respective branches. The bar shown underneath the tree indicates the number of substitutions per site.

Fig. S2.

(A) N-terminal sequences of Pt43233 and Tp1093. (Upper) N-terminal sequence of a CGHR family protein, Pt43233, in P. tricornutum is shown. (Lower) N-terminal sequence of the T. pseudonana ortholog of Pt43233 is also shown. ER signals predicted by SignalP are underlined. Chloroplast transit sequences are boxed. Hydrophobic amino acid residues are represented by bold red font. The italicized bold K and R indicate positively charged Lys and Arg preceding the hydrophobic domain. Ala in the TTD AXA motif is represented by bold font. The TTD sequences deleted in the experiment presented in Fig. 2 I and J are indicated by horizontal parentheses with residue numbers in the N-terminal Pt43233 sequence. (B) Relief image conversion of a part of the photograph in Fig. 2E. An enhanced image of the membrane and pyrenoid structures displayed the chloroplast envelope area (dashed red line), the boundary between the stroma and the pyrenoid (dashed yellow line), and thylakoid membrane structures between the chloroplast envelope and the pyrenoid, or in the central part of the pyrenoid (green parentheses).

Fig. 2.

Subcellular localization and transcript abundance analysis of Pt43233. Superresolution microscopy analysis of Pt43233:GFP localization in a P. tricornutum Pt43233:GFP1 mutant: light image (A), chlorophyll autofluorescence (B, red), GFP fluorescence (C, green), and merged image of B and C (D). (Scale bars, 5 μm.) (E) Immunogold labeling TEM image of the Pt43233:GFP1 mutant. The Pt43233:GFP1 mutant was subjected to TEM following immunogold labeling with anti-GFP antibody. Representative gold particles are indicated by the white arrows in E–H. (F) Magnification of the box in E. (G) Immunogold labeling TEM image of the Pt43233:GFP2 mutant. (H) Magnification of the box in G. (I and J) Respective subcellular localization analysis of Pt43233Δ47–67:GFP and Pt43233Δ55–67:GFP by laser-scanning confocal microscopy. A light image (Left), chlorophyll autofluorescence (Left Center), GFP (Right Center), and merged image (Right) are shown. (Scale bars, 5 μm.) (K) Quantitative RT-PCR analysis of changes in transcript levels in response to changing CO₂ conditions. Pt43233 transcript levels in P. tricornutum cells cultured under 5% (vol/vol) CO₂ (HC), atmospheric air (LC), or very low CO₂ (VLC; <0.002%) with continuous illumination. The gapC2 gene was used as the internal standard. The error bars indicate SDs of three separate experiments. (L) Localization analysis of Pt43233:GFP in Pt43233:GFP2 mutant grown under 1% CO₂ (1%) and 0.04% CO₂ (Air) under light and dark conditions.

Fig. S3.

Localization of CGHR family proteins in P. tricornutum. (A) SIM images of P. tricornutum mutants expressing Pt43234:GFP, Pt43232:GFP, or Pt32401:GFP. The light image (Left), merged image (Center), and magnified merged image (Right) are shown. (Scale bars, 5 μm.) The GFP signal was likely located at the surface or outside the chloroplast in all images. (B) Immunogold labeling TEM images of P. tricornutum cells expressing Pt43234:GFP, Pt43233:GFP, Pt43232:GFP, or Pt32401:GFP. Only Pt43233:GFP was targeted to the pyrenoid-penetrating thylakoid; however, in contrast, Pt43234:GFP was targeted to the mitochondria; Pt43232:GFP was targeted at the space between the chloroplast and the vacuole; and Pt32401 was targeted at the space between the mitochondria and the chloroplast. Arrows are used to represent gold particles. Ch, chloroplast; Mt, mitochondria; N, nucleus; pTk, pyrenoid-penetratng thylakoid; Py, pyrenoid; Tk, thylakoid; V, vacuole.

Fig. S4.

Transcript levels quantified by qRT-PCR under changing CO₂ conditions. Transcript levels of CGHR genes were measured in P. tricornutum cells grown under high-CO₂ (open bar), atmospheric air (gray bar), or very-low-CO₂ (closed bar) conditions with continuous illumination. Transcript levels were normalized to the transcript levels of the gapC2 gene. Data represent the mean ± SD of three separate experiments. The transcript level of Pt43232 under the high-CO₂ condition was 3.1-fold higher than for air-grown cells. Respective Pt43234 and Pt32401 transcript levels were 5.4-fold and 120-fold higher, under the air-grown condition compared with the transcript levels of 1% CO₂-grown cells. The transcript level of Pt32401 was more significantly responsive to low CO₂ than the other CGHR factors, and the accumulation level of Pt32401 under the air-grown condition was higher than the other CGHR factors. Under the very-low-CO₂ condition, the Pt43234 and Pt32401 transcript levels were reduced significantly.

Fig. 3.

CA activity of Pt43233. (A) CA activity of the purified recombinant Pt43233 in the absence (gray) and presence (green) of Zn. (B) CO₂ hydration (Hyd) and HCO₃⁻ dehydration (Dhyd) activities of cell lysates of WT (open bar) and Pt43233:GFP1 mutant (closed bars) grown under 1% CO₂. The result of the t test is indicated (*P < 0.05). (C) Western blot analysis of Pt43233:GFP following immunoprecipitation with anti-GFP antibody. E, eluted fraction; IP, input; N, nonbinding fraction; ni, nonimmune. The protein band corresponding to Pt43233:GFP is represented by the closed arrowhead, and the IgG protein band is represented by the open arrowhead. (D) CA activity of Pt43233:GFP immunoprecipitated with anti-GFP antibody. As a control, WT and Pt43233:GFP1 (G1) cell lysates were immunoprecipitated with anti-GFP IgG and nonimmune rabbit IgG, respectively. The error bar indicates the SD of three separate experiments.

Fig. S5.

Purification of recombinant Pt43233 protein and determination of esterase activity. (A) Purification of recombinant Pt43233 from E. coli. The N-terminal His₆-tagged Pt43233 protein was purified by an IMAC column, followed by His₆-tag removal by thrombin cleavage. The fractions were separated by SDS/PAGE and visualized by Coomassie Brilliant Blue (CBB) staining. (B) Time course of esterase activity detected in the absence (open symbols) and presence (closed symbols) of Pt43233. The circles, squares, and triangles represent data from three independent measurements; the arrow indicates the addition point of the substrate, p-nitrophenyl (PNP) acetate.

Fig. S6.

Confirmation of overexpression and silencing of Pt43233. (A) Western blot analysis of Pt43233:GFP-fusion protein in air-grown WT cells, Pt43233:GFP1 transformant (#1) cells, and Pt43233:GFP2 transformant (#2) cells probed with an anti-GFP rabbit antibody. Immunoblots indicate an 85-kDa immunoreactive band, the estimated size of Pt43233:GFP, in lysates of Pt43233:GFP cells (red arrowhead), but not in WT lysates. (B) Semi–qRT-PCR analysis of Pt43233 transcript level in Pt43233-i1 cells grown under 5% (vol/vol) CO₂ and air. The gapC2 gene was used as the internal standard.

Fig. 4.

Effect of Pt43233 overexpression and down-regulation on photosynthetic parameters and growth of P. tricornutum. (A) Growth curves of WT, Pt43233:GFP1 (G1), and Pt43233-i1 (I1) cells. Cells were cultured under 5% (vol/vol) CO₂ (5%) or 0.04% CO₂ (Air). Data represent mean ± SD of three separate experiments. (B) Kinetic plots of photosynthetic rate in WT, G1, and I1 cells cultivated under 5% (vol/vol) CO₂ (5%; Left) and 0.04% CO₂ (Air; Right). (Insets) Plots at low DIC concentrations. In all plots, open circles represent WT cells, closed gray circles represent G1 cells, and closed black circles represent I1 cells.