3.1 The changes in metabolite content were similar between dark-exposed and illuminated guard cells
Overall, the metabolic changes were similar between dark-exposed and illuminated guard cells. Under both dark and light conditions, the accumulation of fumarate, pyruvate, aspartate and lactate and the degradation of sucrose, glucose, gamma-aminobutyric acid (GABA) and different amino acids was observed (Figure 1). However, also differential behaviour of specific metabolites was observed. After 60 min of illumination the content of both glutamate (Glu) and urea increased as compared to the time 0 of the experiment and fructose, sorbose and citrate levels decreased (Figure 1). Comparing dark and light treatments in each time point, decreased content of alanine, citrate, glucose, myo -inositol, sorbose and sucrose was observed in illuminated samples. Interestingly, the content of malate is lower at 10 and 60 min and higher at 20 min in the light, as compared to dark-exposed guard cells (Supplemental Figure S1). This analysis highlights that the dynamic of relative metabolite accumulation/degradation throughout time is slightly altered by light imposition. This idea is further supported by principal component analysis (PCA), in which no clear separation of dark and light treatments was observed (Figure 2).
3.2 Illumination triggers re-modelling of metabolic network topology in guard cells
We have recently shown that light-induced stomatal opening involves changes in the density and topology of the guard cell metabolic network (Freire et al. 2021). Here, no general pattern of increasing or decreasing network parameters such as clustering coefficient, centralization, density and heterogeneity was observed over time in metabolic networks created using metabolite content data (Table S1). However, the number of hub-like nodes, the preferential attachment and the appearance of hub-like nodes differ between dark and light samples. Whilst these parameters reduced to zero after 60 min in the dark, they increased in the light metabolic network over time (Table S1), in which malate, fumarate, GABA and glucose appear as important hubs of the metabolic network of illuminated guard cells (Figure 3). These results suggest that illumination alter the metabolic network structure of guard cells, with particularly strong impact on the pathways associated to the TCA cycle and sugar metabolism.
3.3 Guard cells have a highly active metabolism in the dark
Similar to the described for metabolite content data, dark-exposed and illuminated guard cells display similar trends in the relative13C-enrichment (R13C) data. Alanine, fumarate, glucose, glycine, sucrose and valine were labelled with statistical significance (P < 0.05) under both conditions in at least one time point, as compared to the time 0 of the experiment (Figure 4). However, comparing light and dark treatments in each time point, increased 13C-labelling in sucrose, fructose, sorbose, malate, pyruvate, succinate and glutamate was observed in illuminated guard cells, especially at 60 min of labelling (Supplemental Figure S2). In fact, PCA indicates that dark and light treatments differ mainly at 60 min, as evidenced by the separation by the first component (Figure 5). Analysis of the contribution of each metabolite for the PCA highlights that certain sugars (sucrose, fructose and glucose), organic acids (citrate and pyruvate), amino acids (Glu, Ile) and myo ­-­inositol are those that mostly contributed to the separation observed in PCA of 60 min (Supplemental Figure S3). These results indicate that several metabolic pathways are activated following PEPc-mediated CO2 assimilation in the dark, including gluconeogenesis and the TCA cycle. However, the combination of CO2 assimilation catalysed by both PEPc and RuBisCO leads to more dramatic changes in 13C-distribution over these and other metabolic pathways in illuminated guard cells.