Interactions between the photosystem II subunit PsbS and xanthophylls studied in vivo and in vitro

Abstract

The photosystem II subunit PsbS is essential for excess energy dissipation (qE); however, both lutein and zeaxanthin are needed for its full activation. Based on previous work, two models can be proposed in which PsbS is either 1) the gene product where the quenching activity is located or 2) a proton-sensing trigger that activates the quencher molecules. The first hypothesis requires xanthophyll binding to two PsbS-binding sites, each activated by the protonation of a dicyclohexylcarbodiimide-binding lumen-exposed glutamic acid residue. To assess the existence and properties of these xanthophyll-binding sites, PsbS point mutants on each of the two Glu residues PsbS E122Q and PsbS E226Q were crossed with the npq1/npq4 and lut2/npq4 mutants lacking zeaxanthin and lutein, respectively. Double mutants E122Q/npq1 and E226Q/npq1 had no qE, whereas E122Q/lut2 and E226Q/lut2 showed a strong qE reduction with respect to both lut2 and single glutamate mutants. These findings exclude a specific interaction between lutein or zeaxanthin and a dicyclohexylcarbodiimide-binding site and suggest that the dependence of nonphotochemical quenching on xanthophyll composition is not due to pigment binding to PsbS. To verify, in vitro, the capacity of xanthophylls to bind PsbS, we have produced recombinant PsbS refolded with purified pigments and shown that Raman signals, previously attributed to PsbS-zeaxanthin interactions, are in fact due to xanthophyll aggregation. We conclude that the xanthophyll dependence of qE is not due to PsbS but to other pigment-binding proteins, probably of the Lhcb type.

FIGURE 1.

Single mutants phenotypes. NPQ activity of the single mutant genotypes used in this work. A, single Glu → Gln mutants have approximately 50% qE amplitude with respect to WT, whereas the double Glu → Gln mutant shows no qE. B, lut2 and npq1 mutants have reduced qE with respect to WT, and the npq1/lut2 double mutant has no qE. C, immunoblotting of thylakoid membranes with an anti-PsbS antibody showing that all genotypes have comparable amounts of PsbS on a chlorophyll content basis with respect to WT with the exception of the E122Q/E226Q double mutant, which has a 40% reduction.

FIGURE 2.

Double mutants (lut2 background) phenotypes. NPQ activity of the double mutants E122Q/lut2 and E226Q/lut2. A, E122Q/lut2 has reduced qE with respect to both lut2 and E122Q. B, E226Q/lut2 has reduced qE with respect to both lut2 and E226Q. C, immunoblotting of thylakoid membranes with anti-PsbS antibodies showing that lut2 double mutants have comparable amounts of PsbS with respect to WT.

FIGURE 3.

Double mutant (npq1 background) phenotypes. NPQ activity of the double mutant npq1 E122Q and genotypes. A, E122Q/npq1 has virtually no activity compared with npq1 and E122Q single mutants. B, E226Q/npq1 has the same NPQ behavior as npq4 (no activity). C, immunoblotting of thylakoid membranes with anti-PsbS antibodies showing that npq1 double mutants have comparable amounts of PsbS with respect to WT.

FIGURE 4.

Zeaxanthin pre-loading experiment. Effect of light pretreatment, inducing zeaxanthin synthesis, on the NPQ behavior of plants carrying single DCCD-binding residues on PsbS. Leaves were treated with 1260 μEm^-2 s^-1 white light for 8 min to induce zeaxanthin synthesis while NPQ was being measured, and the leaves were then dark-adapted for 15 min before a second measurement. NPQ was induced by 1260 μEm^-2 s^-1 actinic white light.

FIGURE 5.

Spectroscopic characterization of in vitro refolded samples. Spectral properties of fractions obtained from in vitro reconstitution of PsbS protein following the protocol for Lhc proteins (30). A, absorption spectra of the two gradient bands from samples reconstituted with total pigments and zeaxanthin at pH 7.8. B, absorption spectra of Lhcb1 samples reconstituted with total pigments. C, CD spectra of lower bands from gradients shown in A and B.

FIGURE 6.

Pictures of sucrose gradient bands. Fractions are obtained upon ultracentrifugation of samples reconstituted with zeaxanthin (zea) or lutein (lute), following the protocol described in Ref. .

FIGURE 7.

Spectral properties of sucrose gradient fractions from Fig. 6. A, absorption spectra of bands from zeaxanthin(Zea)-containing samples. B, absorption spectra of lutein (Lute)-containing samples. C, CD spectra of zeaxanthin samples (upper bands) with or without PsbS shown in Fig. 6. The two samples had the same optical activity upon normalization to absorption at 533 nm.

FIGURE 8.

Resonance Raman spectra. Resonance Raman spectra of the upper and lower gradient bands of zeaxanthin and lutein without PsbS for excitation at 514.5 nm (A). B shows a curve-fitting analysis of the C-H out-of-plane wagging vibrations of the spectra in the 900-1000 cm^-1 region: panel a, zeaxanthin, upper band; panel b, lutein, upper band; panel c, zeaxanthin, lower band; panel d, lutein, lower band. Experimental conditions: 2.2 cm^-1 spectral resolution and 30 milliwatts laser power at the sample; zeaxanthin, upper band, 600 s integration time; zeaxanthin, lower band, 3600 s integration time; lutein, upper band, 1200 s integration time; lutein, lower band, 2400-s integration time.

FIGURE 9.

Conformational analysis of recombinant PsbS refolded in vitro. A, CD spectrum of recombinant PsbS refolded without pigments. B, spectra of the same protein sample after and before addition of denaturing agent guanidine HCl; guanidine absorption below 210 nm disturbs optical measurements.