Diatom Phytochromes Reveal the Existence of Far-Red-Light-Based Sensing in the Ocean

Abstract

The absorption of visible light in aquatic environments has led to the common assumption that aquatic organisms sense and adapt to penetrative blue/green light wavelengths but show little or no response to the more attenuated red/far-red wavelengths. Here, we show that two marine diatom species, Phaeodactylum tricornutum and Thalassiosira pseudonana, possess a bona fide red/far-red light sensing phytochrome (DPH) that uses biliverdin as a chromophore and displays accentuated red-shifted absorbance peaks compared with other characterized plant and algal phytochromes. Exposure to both red and far-red light causes changes in gene expression in P. tricornutum, and the responses to far-red light disappear in DPH knockout cells, demonstrating that P. tricornutum DPH mediates far-red light signaling. The identification of DPH genes in diverse diatom species widely distributed along the water column further emphasizes the ecological significance of far-red light sensing, raising questions about the sources of far-red light. Our analyses indicate that, although far-red wavelengths from sunlight are only detectable at the ocean surface, chlorophyll fluorescence and Raman scattering can generate red/far-red photons in deeper layers. This study opens up novel perspectives on phytochrome-mediated far-red light signaling in the ocean and on the light sensing and adaptive capabilities of marine phototrophs.

Figure 1.

Biochemical Analysis of Pt-DPH and Tp-DPH. (A) Schematic domain organization of Pt-DPH and Tp-DPH proteins. PAS, Period-Arnt-Sim domain; GAF, cyclic di-GMP phosophodiesterase/adenyl cyclase/Fhla domain; H, Histidine Kinase A Domain; KD, HATPase c domain; REC, Response receiver domain; C, cysteine residue putatively binding biliverdin IXα (BV). In Pt-DPH, the PAS domain was not predicted. (B) Absorption spectra of recombinant Pt-DPH-PSM, Pt-DPH-FL, and Tp-DPH-PSM purified from BV-producing E. coli after irradiation with LED emitting light at 757 and 765 nm (red line) or 639 and 690 nm (brown line) for Pt-DPH and Tp-DPH, respectively. In the inset, corresponding purified proteins visualized by Coomassie blue staining (CBB) or zinc-induced fluorescence (Zn). (C) and (D) Different absorption spectra of native (C) and acidic urea denatured (D) recombinant Pt-DPH-PSM purified from E. coli expressing BV (a), together with phytochromobilin (PØB) (b) or PEB (c) biosynthesis enzymes, and irradiated as above.

Figure 2.

Pt-DPH Expression in Different Light Conditions. (A) RT-qPCR analysis of Pt-DPH mRNAs in P. tricornutum cells grown in diurnal 12-h-light/12-h-dark cycles (LD) and in continuous dark (DD) for 2 d. RPS and TBP were used for normalization (n = 3; black bars = ±sd). (B) Immunoblot analysis of Pt-DPH-HA protein in LD and DD in a transgenic line expressing Pt-DPH_Pro:Pt-DPH-HA, using the antihemagglutinin antibody (α-HA) and the α-βCF1 antibody, as loading control. The sampling time points are indicated.

Figure 3.

R-FR-Light-Mediated Gene Expression Changes. (A) Experimental setup used for microarray analysis: P. tricornutum cells were collected at the end of the 12-h-dark period (T0) and following 30 min of irradiation with 660 nm (R) or 765 nm (FR) LED lights (5 μmol⋅m⁻²⋅s⁻¹) or an additional dark treatment (D₃₀). (B) Hierarchical clustering analysis of the 282 genes (Supplemental Data Set 3) obtained from the microarray analysis from R, FR, and D₃₀ conditions, with respect to the T0. The dendrogram on the left defined four major groups assigned to three classes of light regulated genes based on the expression profiles. (C) and (E) Wavelength dependence of Class 1 (HSF4.6a and DNAJ) (C) and Class 2 (ACOAT and GLNAIII) (E) gene expression induction performed by RT-qPCR on cells irradiated for 30 min with different LED lights (660, 688, 732, 740, or 765 nm, 5 μmol⋅m⁻²⋅s⁻¹). The graphs below the RT-qPCR data represent the absorption spectra of the pure Pt-DPH Pr (red line) and Pfr (brown line) forms, and the whole-cell absorption spectra of P. tricornutum (filled gray). (D) and (F) RT-qPCR analyses of the HSF4.6a and DNAJ genes (D) and ACOAT and GLNAIII genes (F), performed on wild-type cells irradiated as described in (A), in presence or absence of the photosystem II inhibitor DCMU. RPS was used as normalizer gene and values were relativized to T0 (n = 3; black bars = ±sd).

Figure 4.

Characterization of Pt-DPH Knockout Lines. (A) Scheme of the translation products of Pt-DPH in the wild-type and transformation control (Tc), and in two knockout lines (KO1 and 2). Sequences retrieved from KO1 and KO2 lines shown in red, for the different alleles, the mutations in the Pt-DPH locus compared with the wild type, and on the right the nucleotide insertions (+), deletions (−), and/or transversions. (B) and (C) RT-qPCR analysis of Class1 (HSF4.6a, DNAJ, HISK, and GAFHisK) (B) and Class 2 genes (ACOAT, GLNAIII, CHRT, and ZEP1) (C) on wild-type cells, Tc, KO1, and 2 lines, exposed for 30 min to 660 nm (R, red) or 765 nm (FR, brown) LED lights (5 µmol⋅m⁻²⋅s⁻¹) or maintained in darkness (D₃₀, black). Expression data were normalized using the T0 point (cf. Figure 3A) and the RPS gene as normalizer (n = 3; black bars = ±sd). (D) Autophosphorylation assay of recombinant holo-Pt-DPH-FL incubated with [ϒ-³²P]ATP under 639-nm (R) or 757-nm (FR) LED lights up to 3 min, revealed by SDS-PAGE and autoradiography.

Figure 5.

Exploration of DPH Abundance and Distribution and of the Light Sources Possibly Triggering DPH in the Marine Environment. (A) Phylogenetic analysis of PHY protein superfamily using the GAF-PHY region. The tree is midpoint-rooted and was obtained applying the Bayesian inference method. The posterior probability value support is shown only for the basal nodes (black dots). (B) Vertical distribution of T. rotula (yellow) and S. costatum (green) from the MAREDAT database. Each data point is the median of the abundance values (cells/dm³) for single depth (squares) or depth intervals centered on the median depth (vertical lines). The number of observations for each data point is provided. Dashed lines show interpolation of the data points. (C) Vertical profiles of the available scalar irradiance in the wavelength range of 610 to 800 nm. Open circles, irradiance derived from the attenuated sunlight E(sl) only; closed circles, total irradiance E(tot), deriving from sunlight, Raman scattering, and phytoplankton chlorophyll. Marine station codes are reported at the bottom. (D) Ratios between the relative molar absorptivities of Pt-DPH Pr and Pt-DPH Pfr at different depths in the same spectral range.