Protein oxidation is ubiquitous throughout the tree of life and is important for all organisms to respond to both biotic and abiotic stress conditions. This is especially true in photosynthetic organisms, because they must contend with a daily onslaught of variable light conditions. Under optimal photon flux densities, a plant or alga can use the majority of the light for photochemistry. However, photosynthetic organisms are often inundated with excess light that cannot all be used photochemically due to sink constraints. This excess absorbed excitation energy must be dissipated or else the oxidative stress, caused by the formation of reactive oxygen species (ROS), would cause detrimental damage. These organisms have evolved a number of different mechanisms to dissipate excess energy, collectively known as non-photochemical quenching (NPQ). However, even with these safeguards, excess energy remains and ROS are still formed. Over the last few decades, research has shown that ROS may be more than just damaging to the cell. In fact, ROS can function in a second level of defense through protein oxidation-induced regulation and signaling. To investigate the effects of protein oxidation in photosynthetic organisms, I completed a proteomic analysis to identify cysteine oxidation sites in both Nannochloropsis oceanica cells and Arabidopsis thaliana chloroplasts, and I characterized cysteine-modified mutants that affect target proteins identified from the proteomics.
In N. oceanica, I identified several hundred proteins containing the primary oxidized state of cysteine, cysteine sulfenic acid (Cys-SOH), from three light conditions: dark, low light (LL), and high light (HL). Additionally, the proteomic analysis showed an increase in the number of Cys-SOH modifications with increases in light intensity. From this screen, several targets were selected for reverse genetics. CRISPR/Cas9 ribonucleoprotein-mediated homologous recombination was used to knockout (KO) these target genes. Four mutants showed an NPQ phenotype, with lhcx1 standing out above the rest. The lhcx1 KO exhibited a complete loss of the rapidly reversible, feedback de-excitation component of NPQ (qE), demonstrating that it is central to early NPQ induction. I transformed the mutant with the wild type LHCX1 gene (lhcx1+WT) as well as two modified versions of the gene: a cysteine to alanine mutant (lhcx1+C162A) to mimic the reduced state of the cysteine and a cysteine to serine (lhcx1+C162S) to mimic the Cys-SOH state. These modifications were chosen to determine what role the cysteine oxidation plays in the function of this protein.
Analysis of the cysteine-modified transgenics showed that LHCX1 is regulated by the oxidation state of its lone cysteine residue. In the lhcx1+WT and lhcx1+C162A lines, there was a recovery of qE back to the wild type state. In contrast, this recovery was not complete in the lhcx1+C162S line, which showed only ~60% of wild type total NPQ, and there was an additional sustained quenching phenotype: up to 50% of the total NPQ was slowly reversible compared to only 20% in lhcx1+WT and lhcx1+C162A. This sustained quenching was reminiscent of the qZ type of NPQ, which is slow to relax and dependent on zeaxanthin (Zea) formation. I examined Zea levels to determine if this higher qZ quenching was due to overaccumulation of Zea and found that there were no major differences. To determine if this sustained quenching would occur in the wild type, the period of actinic light was increased to allow for ROS build up. After 20 min, the level of qZ in the lhcx1+WT line had increased to ~40%, while qZ in lhcx1+C162A, which cannot be oxidized, remained low at 20%. These results strongly suggest that the oxidation of C162 to the Cys-SOH state modulates the quenching dynamics from the qE state to a qZ state. Based on protein modeling, it is possible that this switch is caused by a structural change in LHCX1, which directly alters the Zea binding affinity and causes the relocation of Zea away from the qE site(s) on LHCX1 toward qZ sites most likely on VCP type proteins.
In A. thaliana, I focused on the chloroplast proteome, as we are most interested in light-induced oxidation reactions. Hundreds of proteins with the Cys-SOH modification were identified from the three light conditions (dark, LL, and HL). Similar to N. oceanica, there was an increase in the number of Cys-SOH modified proteins with increases in light intensity, with the HL samples having twice as many modified proteins as the dark samples. The HL sample has an enrichment of proteins, based on GO term analysis, that are involved in translation, transport, and phosphorylation, which might be linked to downstream responses to oxidative stress. Based on the proteomics results, T-DNA lines were acquired in photosynthesis-associated targets for further characterization of possible Cys-SOH regulation. Three lines (lhca6, prxq, and atr2) had NPQ phenotypes, and one line (cyp38) had a strong growth phenotype. lhca6 had the strongest NPQ enhancement in both LL-grown and HL-treated conditions, so it was transformed with the wild-type LHCA6 gene (lhca6+WT) as well as the two cysteine-modified versions (lhca6+C58A and lhca6+C58S). The NPQ phenotype of lhca6+WT and lhca6+C58A both returned to the wild-type level, but the lhca6+C58S stayed at the level of the lhca6 T-DNA line. This is consistent with the possibility of Cys-SOH oxidation causing inactivation of LHCA6. However, this T-DNA line was only a knockdown with highly variable levels of mRNA, ranging from 20-65% of WT, so for complete confidence in the phenotype analysis of a complete KO is required.
NPQ dissipates excitation energy. If not tightly regulated this process could dissipate usable energy when light levels return to the optimal range. The oxidative regulation of proteins associated with NPQ could represent a second level of control, thereby preventing sustained NPQ quenching during short bursts of HL but allowing for greater overall quenching during longer periods of excess light. Additionally, the activation or inactivation of proteins by Cys-SOH formation could play a major role in many signaling and direct stress responses, beyond the light-mediated stress response examined here.