Energy dissipation pathways in Photosystem 2 of the diatom, Phaeodactylum tricornutum, under high-light conditions

Abstract

To prevent photooxidative damage under supraoptimal light, photosynthetic organisms evolved mechanisms to thermally dissipate excess absorbed energy, known as non-photochemical quenching (NPQ). Here we quantify NPQ-induced alterations in light-harvesting processes and photochemical reactions in Photosystem 2 (PS2) in the pennate diatom Phaeodactylum tricornutum. Using a combination of picosecond lifetime analysis and variable fluorescence technique, we examined the dynamics of NPQ activation upon transition from dark to high light. Our analysis revealed that NPQ activation starts with a 2–3-fold increase in the rate constant of non-radiative charge recombination in the reaction center (RC); however, this increase is compensated with a proportional increase in the rate constant of back reactions. The resulting alterations in photochemical processes in PS2 RC do not contribute directly to quenching of antenna excitons by the RC, but favor non-radiative dissipation pathways within the RC, reducing the yields of spin conversion of the RC chlorophyll to the triplet state. The NPQ-induced changes in the RC are followed by a gradual ~ 2.5-fold increase in the yields of thermal dissipation in light-harvesting complexes. Our data suggest that thermal dissipation in light-harvesting complexes is the major sink for NPQ; RCs are not directly involved in the NPQ process, but could contribute to photoprotection via reduction in the probability of ³Chl formation.

Appendix

Exciton-radical pair equilibrium (ERPE) kinetic models describe energy migration in the PS2 antenna–RC complex, including charge separation processes, charge recombination, and electron transfer. Figure 9a shows an example of a ERPE kinetic model (Schatz et al. 1988; Ivanov et al. 2008b). Upon excitation of the PS2 antenna complex a rapid excitation equilibrium (<10 ps) (Holzwarth et al. 2006) is achieved between antenna and RC (exciton exchange is represented by trapping (k _t) and detrapping (k _−t) rate constants). Due to this rapid equilibrium, ERPE models assume that the excited states of antenna and RC are not separated and are, therefore, represented by only one compartment (i.e., k _t and k_−t are not determined within the framework of this model). The next step is the process of charge separation in the RC (with the rate constant of k ₁) that results in the formation of a radical pair (RP) P680⁺I⁻. There are 4 major pathways of energy conversion and recombination of the RP. First, RP could recombine to the antenna/RC excited state (with the rate constant k ₋₁). Alternatively, it may recombine (through non-radiative decay) to the ground state (k _2D), to the triplet excited state of P680 (k _2T, spin dephasing), or transfer electron to Q _a (k _2Q, charge stabilization). Charge stabilization takes place in open RCs, whereas the other two pathways dominate in closed RCs. Deactivation of the excited states in the antenna complex is represented by non-radiative (thermal dissipation, kDa) and radiative (fluorescence, k _F) processes.

Fig. 9

Exciton-radical pair equilibrium model (modified from Schatz et al. 1988) (a) and its simplified 2-compartment version used for fluorescence decay kinetics analysis (b). k _a—antenna deactivation rate constant, k ₁ and k ₋₁ are apparent rate constants of charge separation and charge recombination, k ₂—rate of non-photochemical losses, which includes: k _2D—rate constant of charge recombination to the ground state; k _T—rate constant of charge recombination to the triplet state; and k _2Q—rate constant of charge stabilization. See text for more details

From the two-compartment model presented on Fig. 9b one can derive up to four rate constants: antenna deactivation rate constant (that includes both heat dissipation and fluorescence), apparent rate constant of charge separation (k ₋₁) and charge recombination (k ₁), and rate constant of photochemical (k_Q, for open RC) or non-photochemical (k ₂ = k _T + k _D, for closed RC) losses in the RC. Analytically solution for the system of kinetic equations that describe this 2-compartment model predicts two exponential components for fluorescence decay kinetics (i.e., the kinetics is described by four parameters—two pre-exponential amplitudes and two lifetimes). Thus, in order to determine all four rates of the kinetic model from the experimental lifetime kinetics of chlorophyll a fluorescence decay, we should be able to allocate two lifetimes related to PS2 fluorescence at a given state of photosynthetic apparatus (e.g., open or closed).

In order to calculate rates of exciton and electron transfer for exciton-radical pair equilibrium model (Fig. 9c), we solved a system of differential equations using a Matlab(c) software.

$$ \left\{ {\begin{array}{l} {\frac{{\text{d}}X}{{\text{d}}t} = k_{\text{a}} \times Y} \\ {\frac{{\text{d}}Y}{{\text{d}}t} = - k_{\text{a}} \times Y - k_{1} \times Y + k_{ - 1} \times Z} \\ {\frac{{\text{d}}Z}{{{\text{d}}t}} = k_{ 1} \times Y - k_{ - 1} \times Z - k_{2} \times Z} \\ {\frac{{{\text{d}}W}}{{{\text{d}}t}} = k_{2} \times Z} \\ {X(t) + Y(t) + Z(t) + W(t) = 1} \\ {X(0) = Y(0) = Z(0) = 0} \\ {Y(0) = 1} \\ \end{array} } \right., $$

(10)

where X, Y, and Z are relative concentrations of ground state chlorophyll a (including P680), excited state of chlorophyll a (antenna and P680), and radical pair (P680⁺I⁻), respectively. In the case of open RC, W is the relative concentration of charge-stabilized state (although, strictly speaking, it also includes concentration of intermediates that lead to P680 triplet state formation), while for closed reaction centers W represents the total concentration of intermediates after RP non-radiative charge recombination (including triplet excited state of P680, Fig. 9a).

Below we present a solution of Eq. (10) for Y(t) [similar solutions could be obtained for X(t), Z(t), and W(t)]:

$$ \begin{aligned} Y\left( t \right) = A_{1} \times {\text{e}}^{{ - \frac{t}{{\tau_{1} }}}} + A_{2} \times {\text{e}}^{{ - \frac{t}{{\tau_{2} }}}} \hfill \\ A_{1,2} = \frac{1}{2}\left( {1 \mp \frac{{L_{1} }}{{\sqrt {L_{1}^{2} + 4k_{1} k_{ - 1} } }}} \right) \hfill \\ \frac{1}{{\tau_{1,2} }} = \frac{1}{2}\left( {L_{2} \mp \sqrt {L_{1}^{2} + 4k_{1} k_{ - 1} } } \right) \hfill \\ L_{1,2} = (k_{a} + k_{1} ) \mp \left( {k_{ - 1} + k_{2} } \right) \hfill \\ \end{aligned} $$

(11)

A ₁, A ₂, τ ₁, and τ ₂ depend on the four rate constants (k_a, k ₁, k ₋₁, and k ₂) of system (A ₁). Taking into account that A ₂ = 1 − A ₁, we have only three independent equations to connect experimentally determined amplitudes and lifetimes with four rates of exciton and electron transfer.

In order to obtain all four rates, we either have to fix one of the rates (or their combination) or determine them independently. For open reaction centers (F _o level), a common practice is to use the yield of charge stabilization (i.e., ${\text{W}}(t \to \infty )$) determined from picosecond absorption kinetics (Schatz et al. 1988) as it provides us with an additional equation that combines these 4 rate constants. We can determine the yield of charge stabilization from system Eq. (10):

$$ W(t \to \infty ) = W_{max} = \frac{{k_{1} k_{2} }}{{k_{1} k_{2} + k_{a} \left( {k_{ - 1} + k_{2} } \right)}} $$

(12)

To present it is well established that the yield of charge stabilization in the RC is ~90 % (Schatz et al. 1987; Caffarri et al. 2011). We followed the procedure described by (Lambrev et al. 2012) and used W _max = 87 %, which allowed us (using Eq. (11) and expression for W _max) to determine all four rates for open RC (F _o level, Table 3). In order to determine all four rates for closed reaction (F _M level), we used k _a values previously determined for open RC. A more complex approach was used to evaluate all four rates in the closed reaction centers with induced NPQ, because induction of NPQ is expected to change antenna dissipation rate constant (k _a). In order to resolve this problem, we performed measurements of variable fluorescence using FIRe technique along with the lifetime measurements. Analysis of the induction curves allows determination of PS2 functional absorption cross section (σ _PS2 (Gorbunov and Falkowski 2005)), which is proportional to the probability of charge stabilization in the RC. Thus, its relative changes upon NPQ activation would reflect changes in the quantum yield of charge stabilization in the RC of PS2. W _max for open RC (F _o level) also gives us the probability of charge stabilization. Thus, relative changes in σ _PS2 and W_max determined under quenched and non-quenched conditions would be equal.

$$ {\raise0.7ex\hbox{${\sigma_{\text{PS2}}^{\text{NPQ}} }$} \!\mathord{\left/ {\vphantom {{\sigma_{\text{PS2}}^{\text{NPQ}} } {\sigma_{\text{PS2}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${\sigma_{\text{PS2}} }$}} = {\raise0.7ex\hbox{${W_{ \hbox{max} }^{\text{NPQ}} \left( {F_{0} } \right)}$} \!\mathord{\left/ {\vphantom {{W_{ \hbox{max} }^{\text{NPQ}} \left( {F_{0} } \right)} {W_{ \hbox{max} }^{{}} \left( {F_{0} } \right)}}}\right.\kern-0pt} \!\lower0.7ex\hbox{${W_{ \hbox{max} }^{{}} \left( {F_{0} } \right)}$}} = {\raise0.7ex\hbox{${\left( {k_{1} k_{2Q} + k_{\text{a}} \times \left( {k_{ - 1} + k_{2Q} } \right)} \right)}$} \!\mathord{\left/ {\vphantom {{\left( {k_{1} k_{2Q} + k_{\text{a}} \times \left( {k_{ - 1} + k_{2Q} } \right)} \right)} {\left( {k_{1} k_{2Q} + k_{\text{a}} \times \left( {k_{ - 1} + k_{2Q} } \right)} \right)^{\text{NPQ}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${\left( {k_{1} k_{2Q} + k_{\text{a}} \times \left( {k_{ - 1} + k_{2Q} } \right)} \right)^{\text{NPQ}} }$}} $$

(13)

(here ${\text{k}}_{{2{\text{Q}}}}$ is the rate of charge stabilization).

If we assume that changes in the probability of charge stabilization (in open RC) are related to changes only in the antenna deactivation rate constant (k _a), while apparent rate constants of charge separation (k ₁) and charge recombination (k ₋₁), as well as charge stabilization (k ₂) remained constant upon NPQ activation, than from Eq. (13) we can derive the equation to determine antenna deactivation rates under NPQ conditions:

$$ k_{a}^{NPQ} = \frac{{\frac{{\sigma_{PS2}^{{}} }}{{\sigma_{PS2}^{NPQ} }} \times \left( {k_{1} k_{2Q} + k_{a} \times \left( {k_{ - 1} + k_{2Q} } \right)} \right) - k_{1} k_{2Q} }}{{k_{ - 1} + k_{2Q} }} $$

(14)

This assumption is based on the fact that open and closed RC have very different rates of exciton and electron transfer (k ₁, k ₋₁, k ₂) and, changes induced by actinic light would affect open reaction center rates to a lesser extent. Moreover: (1) we do not see significant changes in k ₁ during NPQ induction (Fig. 7b); (2) due to small values of k ₋₁ compared to other rate expressions (14) is not as “sensitive” to changes in this rate and even a two-fold change in k ₋₁ would not affect the value of k _a significantly (estimated changes are < 5 %); (3) there is a 8.1 10⁸ s⁻¹ increase in k ₂ in closed RC upon actinic light illumination related to non-photochemical losses in the RC (Table 3); if this is to be added to the rate of k ₂ for open RC (26.8 × 10⁸ s⁻¹, Table 3) and the resulting rate 34.9 × 10⁸ s⁻¹ is to be used in (A5), it would have only a minor (<2 %) effect on k _a as compared with using 26.8 × 10⁸ s⁻¹.

Using the expression for rate constant of antenna deactivation under NPQ conditions (${\text{k}}_{\text{a}}^{\text{NPQ}}$) and Eq. (11) equations, we can calculate other rate constants (k ₁, k ₋₁, k ₂) and yields of energy dissipation in the reaction center (W _max) and antennae (X _max = 1 − W _max) of PS2 upon NPQ activation ($F_{\text{M}}^{\text{NPQ}}$ level). Table 5 provides a brief summary of the above-described procedure.

Table 5 Algorithm of kinetic data analysis. Picosecond lifetime and variable fluorescence kinetics’ parameters, input parameters, were used to determine rates of exciton and electron transfer in PS2 antenna–RC complex (Fig. 9), output parameters, using Eqs. (11–14) for each of the state of photosynthetic apparatus: open (F ₀), closed (F _M), and quenched ($F_{\text{M}}^{\text{NPQ}}$)