Unravelling the different components of nonphotochemical quenching using a novel analytical pipeline

Introduction

Photosynthesis is arguably one of the most important biological processes in Nature (Blankenship, 2008; Johnson, 2016). In this process, the incoming solar radiation is captured by photosynthetic organisms and converted to chemical energy, and so underpins most of the food chains on Earth (Nelson & Yocum, 2006; Blankenship, 2008; Johnson, 2016). However, if oxygenic photosynthetic organisms are exposed to light intensities at which the rate of photon absorption exceeds the rate of photochemical quenching of excitations and thus photosynthetic metabolism, the excess excitations present within the photosystems can cause photodamage. Specifically, high light results in a large portion of the reaction centres (RCs) within the photosystems being in the closed state. With increasing irradiance, it is more likely that an excitation encounters a closed RC leading to triplet states being formed by back reactions within the reaction centre, which in turn can produce singlet oxygen (Vass, 2011; Telfer, 2014). In order to minimise such photodamage, oxygenic photosynthetic organisms activate photoprotective mechanisms, designed to safely remove excess excitations from the photosystems via nonphotochemical quenching (NPQ) (Horton et al., 1996; Demmig-Adams & Adams, 1996b; De Bianchi et al., 2010; Ruban et al., 2012). In photosystem II (PSII), NPQ is mediated by several different molecular mechanisms that act together to quench excitations (Demmig-Adams & Adams, 1996a,b; D'Haese et al., 2004; Li et al., 2004, 2009; Johnson et al., 2009; Jahns & Holzwarth, 2012; Ruban et al., 2012; Sylak-Glassman et al., 2014; Goldschmidt-Clermont & Bassi, 2015; Armbruster et al., 2016; Ruban, 2016, 2019; Farooq et al., 2018; Townsend et al., 2018; Van Amerongen & Chmeliov, 2020; Ruban & Wilson, 2021; Long et al., 2022). The fastest of these mechanisms are triggered by the accumulation of protons in the lumenal space, and these processes have been studied for several decades (Demmig-Adams & Adams, 1996a,b; D'Haese et al., 2004; Li et al., 2004, 2009; Johnson et al., 2009; Jahns & Holzwarth, 2012; Ruban et al., 2012; Sylak-Glassman et al., 2014; Goldschmidt-Clermont & Bassi, 2015; Armbruster et al., 2016; Ruban, 2016, 2019; Townsend et al., 2018; Ruban & Wilson, 2021; Long et al., 2022). These processes are associated with the protonation of the PsbS protein and the activation of the violaxanthin de-epoxidase enzyme (VDE) leading to the accumulation of zeaxanthin via antheraxanthin. Whilst these processes are known to be important for NPQ, many studies of the kinetics of their induction and relaxation reveal a significant latency in the overall NPQ response (Kromdijk et al., 2016; Ruban, 2017; Wang et al., 2020). This latency is thought to make photosynthesis significantly less efficient upon a decrease in the actinic light intensity, whilst upon a sudden increase in light intensity, this latency could temporarily leave the photosynthetic apparatus under protected. Accordingly, these processes have been widely studied with a view to improving the overall efficiency of photosynthesis and photoprotection (Kromdijk et al., 2016; De Souza et al., 2022).

Typically, the molecular processes involved in the NPQ response are probed in vivo utilising either changes in the steady-state fluorescence and absorption spectra of photosynthetic organisms following exposure to, or changes in, actinic light. Several studies have correlated particular changes in the absorption spectra with the activation of VDE leading to the accumulation of zeaxanthin (Bilger & Björkman, 1990; Li et al., 2000; Johnson et al., 2009) and the formation of a quenching species thought to be associated with the protonation of PsbS (Johnson & Ruban, 2010). In addition to these changes, it has also been noted that the size of the trans-thylakoid voltage can be probed by monitoring changes in the absorption spectra due to electrochromic shifts (ECS) originating from the electric field generated by the trans-thylakoid voltage (Bailleul et al., 2010), and by monitoring changes in the far-red part of the spectra the oxidation state of the PSI RC can be measured (Harbinson & Woodward, 1987; Klughammer & Schreiber, 1998). Such studies have been used to explore different aspects of NPQ and have yielded some insights into the induction of the underlying processes. However, these measurements are broadly speaking difficult to perform and analyse due to the complexity created by the high degree of convolution of a large number of different chromophore containing molecular species with nontrivial absorption spectra.

Alternatively, photosynthesis can also be monitored in vivo by changes in the Chl fluorescence yield (Harbinson, 2018). Whilst there are some experimental indications that there may also be a small degree of variable PSI fluorescence (Lazár, 2013; Schreiber & Klughammer, 2021; Schreiber, 2023), it is generally accepted to be minor in comparison with variable fluorescence from PSII. Therefore, changes in Chl fluorescence yield are normally only useful for examining PSII (Harbinson, 2018), even though fluorescence-derived PSII parameters are often in error due to the presence of PSI fluorescence. Commonly, fluorescence yield is measured using pulse–amplitude–modulated (PAM) fluorometry (Maxwell & Johnson, 2000; Baker, 2008; Harbinson, 2018). Whilst using only Chl fluorescence decreases the overall number of distinct processes that can be monitored independently, the variables that can be extracted by applying the saturating pulse method (often used in PAM measurements) and utilising the Butler model (Butler, 1978) describe important photosynthetic parameters such as the proxies for the quantum yield of PSII ($$_PSII), the fraction of open PSII RCs (qP) and NPQ (Maxwell & Johnson, 2000). Due to its sensitivity, this technique has been widely applied to study photosynthetic phenomena such as NPQ. In the case of NPQ, this parameter can be calculated directly from the PAM data using the Stern–Volmer equation ($$, where $$ is the maximum fluorescence yield with closed RCs in the dark adapted state and $$ is the maximum fluorescence yield with closed RCs in the presence of NPQ) (Bilger & Björkman, 1990; Maxwell & Johnson, 2000; Harbinson, 2018). Over the past few decades, applying this technique to both wild-type (wt) and mutant plants has illustrated that there are many processes contributing to the overall NPQ developed by vascular plants in response to the trans-thylakoid membrane proton potential difference established during illumination and the associated decrease in lumen pH that ensues. These mechanisms include but are not limited to the protonation of the PsbS protein and possibly the antenna complexes (Li et al., 2004; Johnson & Ruban, 2010; Mou et al., 2013; Dong et al., 2015; Ruan et al., 2023), the activation of the xanthophyll cycle (Bilger & Björkman, 1990; Jahns & Holzwarth, 2012), and photoinhibition (qI) (Taylor et al., 2019; Vetoshkina et al., 2023). As multiple processes play a role on similar timescales, it has thus far been difficult, or even impossible, to unambiguously separate the contribution of the different mechanisms in time. One possible approach to disentangle these processes is the recently developed frequency domain analysis of Chl fluorescence yield. However, due to the complex nature of the data, these approaches are not yet capable to quantitatively discriminate between these different processes (Nedbal & Lazár, 2021; Niu et al., 2023, 2024).

Here, we develop a novel multivariate analysis pipeline to deconvolute a data set of NPQ induction curves, obtained via PAM fluorometry (using red actinic light, Supporting Information Fig. S1b, to minimise chloroplast movement (Baránková et al., 2016)), to reveal the induction profiles of different NPQ components. Focussing on data sets for wt and npq1 (zeaxanthin-lacking) A. thaliana, this pipeline reveals four distinct components ($$, $$, $$ and $$). Comparison with chemical treatments and previous studies allows the molecular processes associated with these components to be assigned. This allows the contributions of each of the different mechanisms underlying NPQ to be unambiguously identified directly from Chl fluorescence yield measurements for the first time.