Plant photosynthesis converts atmospheric CO₂ into sugars that sustain the global food chain. Under the present day high O₂/CO₂ ratio, photosynthetic efficiency is limited by Rubisco’s oxygenase activity, which generates 2-phosphoglycolate (2-PG), a potent inhibitor of the Calvin-Benson-Bassham (CBB) cycle enzymes sedoheptulose-1,7-bisphosphatase (SBPase) and triosephosphate isomerase (TPI) [1–2]. Photorespiration exclusively removes 2-PG to prevent metabolic inhibition, yet it is energetically costly and releases previously fixed CO₂ and NH₃⁺ [3]. Beyond its essential repair function, photorespiration supports de novo nitrogen and sulfur assimilation, amino acid biosynthesis, one-carbon metabolism, and redox balance, thereby contributing to acclimation under fluctuating conditions [4–6]. These Janus-faced roles, protective but metabolically expensive, make photorespiration a key target for improving plant productivity under current and future climates [7–8].

Leaf-mesophyll photorespiration is well-characterized, but its function in specific cell types remains largely unexplored. Although the pathway is highly compartmentalized, spanning chloroplasts, mitochondria, peroxisomes, cytosol, and vacuoles [6, 9–10], its interaction with other metabolic routes is not well resolved. Recent isotopically non-stationary metabolic flux analyses (INST-MFA) showed that a substantial fraction of carbon exits the canonical cycle, feeding other biosynthetic routes including C1-carbon metabolism, mainly as serine and glycine [11–13]. This implies that certain reactions exert disproportionate control over the photorespiratory flux and may regulate carbon utilization and partitioning in distinct compartments. Genetic studies have highlighted two key control points: mitochondrial glycine decarboxylase (GDC) and chloroplast-localized 2-PG phosphatase 1 (PGLP1). Both enzymes share a strong positive correlation with the photorespiratory flux, CBB cycle operation, starch biosynthesis, and plant growth, making them attractive targets for improved yield [1, 14–16].

Photorespiratory rates are determined by the CO₂/O₂ ratios in chloroplast, which largely depend on opening of stomata, formed by pairs of guard cells (GC) in the leaf epidermis. In addition to flux of CO₂ and O₂, stomatal opening regulates water vapor movements, balancing carbon gain with water conservation [17–19]. Stomatal size and density are inversely related and determine maximum stomatal conductance (g_smax); small, numerous stomata support fast stomatal kinetics [20] and higher gas exchange than few, large stomata. Across evolutionary timescales, high atmospheric CO₂ concentrations ([CO₂]) favored fewer, larger stomata, whereas low [CO₂] selected for smaller, denser stomata [21–23]. On daily timescales, stomatal size and density will not change, but GC respond dynamically to environmental cues, including light, internal [CO₂], temperature, humidity, and water availability. Light is a dominant driver, red light triggers GC osmoregulation and transmits mesophyll signals that align stomatal opening with photosynthetic demand [24], while blue light acts independently of photosynthesis, directly via phototropin kinases to induce rapid opening at low fluence rates, particularly at dawn and during transient sunlight fluctuations, maximizing carbon assimilation [25–27]. Stomatal closure is mainly achieved via the plant hormone abscisic acid (ABA), maintaining overall plant water status [17–18].

Intercellular CO₂ (C_i) is regarded as another important factor controlling stomata. For example, C_i elevation with decreasing photosynthesis, darkness or via raised external [CO₂] promotes stomatal closure, whereas light-dependent draw-down of C_i via photosynthetic CO₂ fixation maintains opening [28]. Rising global [CO₂] reduces stomatal aperture and density, lowering conductance and conserving water, yet potentially increasing leaf temperature under drought. At the molecular level, CO₂ responses require the protein kinase high leaf temperature 1 (HT1) and converge with abscisic acid (ABA) signaling, because elevated [CO₂] increases guard-cell ABA to induce closure. Thus, stomatal behavior integrates red/blue light signals, CO₂ feedback, and hormonal cues to balance carbon gain, water use, and thermal stability [29–31].

In addition to photorespiration, the activity of Rubisco carboxylation and the flux through the CBB cycle also responds to fluctuating CO₂/O₂ ratios. It has been shown that SBPase activity might be a control point of the CBB cycle flux, because its overexpression in plants enhanced photosynthesis and yield [32–34]. Comparable effects have been observed in overexpressors of key photorespiratory enzymes such as GDC and PGLP1 [1, 14–16], whereas impairment of photorespiration diminished productivity [35–36]. Interestingly, the recently reported GC-specific manipulation of GDC suggested a functional link between mitochondrial photorespiratory metabolism and stomatal regulation [37]. If this finding is specific for mitochondria or GDC, releasing CO₂ during photorespiration thereby eventually affecting C_i, remains unknown. However, initial pharmacological studies on photorespiratory enzymes in epidermal peals indirectly supported a functional interaction between the activity of certain photorespiratory enzymes and stomatal movements [38]. Surprisingly, the role of PGLP1, the photorespiration-specific 2-PG degrading enzyme, in GC is still unclear. Gaining insights into its physiological significance is interesting because of the reduced GC chloroplast count and size, alongside with hinds for the sink-tissue-like characteristics of GC, and the ongoing debate to which extent GC rely on their own internal photosynthesis [39–41].

To address this question, we analyzed the impact of GC-specific manipulation of the central photorespiratory enzyme, chloroplastidal PGLP1. The transgenic lines with in- and de-creased GC-specific PGLP1 expression were assessed to study its role in GC metabolism and impact on stomata function, photosynthesis and biomass accumulation. By elucidating the role of photorespiration in these specialized cells, our work not only provides new insights into its significance for GC metabolism but also provided the foundation for new strategies to engineer crops with enhanced growth, water-use efficiency, and yield, key traits to meet the challenges of plant production under increasing climate change.

To test if growth changes are due to altered photosynthesis, we measured chlorophyll a fluorescence and gas exchange parameters of plants grown under photorespiratory conditions. Whilst PSI and PSII efficiencies and related parameters associated with photosynthetic light reactions did not significantly vary among the genotypes (Supp. Fig. S2; Supp. Table S2), light-dependent net CO₂ assimilation (A_N) and stomatal conductance (g_s) followed the growth pattern. Thus, A_N and g_s displayed a positive correlation with GC PGLP1 expression (Fig. 2A-C). Transpiration rates (E), maximum photosynthetic rates (A_max) and the slope of the light response curves (α_p) followed this tendency (Supp. Table S3). Intracellular CO₂ concentrations (C_i) were only significantly decreased in the antisense lines, whilst intrinsic water use efficiency (iWUE) showed only minor alterations (Supp. Table S4). To check if photosynthetic stimulations rely on altered photorespiratory 2-PG turnover in GC, we measured photosynthesis at three different O₂ concentrations (3, 21 and 40%) to suppress or stimulate 2-PG formation. At low photorespiratory flux requirements (3% O₂), no significant changes on A_N, g_s, and the CO₂ compensations points (Γ) were observed (Fig. 2D-F). However, at air O₂ levels (21%) A_N was increased in overexpressor (~ 19%) and decreased in antisense lines (~ 13%), whilst Γ displayed inverse tendencies (~ 10% lower in overexpressors and ~ 14% higher in antisense lines). Interestingly, g_s, a parameter directly related to stomatal opening, was positively correlated with GC PGLP1 protein amounts and was higher (~ 29%) or lower (~ 16%) in the overexpressor and antisense lines, respectively (Fig. 2E, Supp. Table S5). The described patterns were similar at 40% O₂, i.e. photorespiration-stimulating conditions, but with stronger specification. In the overexpression lines A_N and g_s were stimulated (~ 49% and 60%) and Γ decreased (~ 17%), whilst antisense repression caused a reduction in A_N and g_s (~ 33% and 27%) and a corresponding increase (~ 15%) in Γ (Fig. 2D-F). The calculated O₂ sensitivity revealed overexpression lines were less and antisense lines more sensitive to O₂ compared with the wildtype (Fig. 2D, inlet). Finally, the slope of the Γ-vs-O₂ concentration (γ), representing a measure of photorespiratory CO₂ release, revealed that GC PGLP1 overexpression caused a significant reduction, whilst antisense suppression an increase in photorespiratory CO₂ losses (Fig. 2F, inlet).

Plant were grown in air to stage 5.1 [43] to determine photosynthetic gas exchange as functions of light intensity and O₂ concentrations. Selected parameters of the light response curves are given as follows: (A) Net CO₂ uptake rates (A_N), (B) stomatal conductance (g_s), and (C) correlation plot of A_N versus g_s. Selected parameters of the CO₂ response curves are presented as follows: (D) A_N, including the oxygen inhibition of photosynthesis as inlet, (E) g_s, and (F) CO₂ compensation points (Γ), including the slope of the Γ-versus-O₂ concentration functions (γ) as inlet. Correlation plots of A_N, g_s and Γ at 21% O₂ (air) are displayed at the bottom of each figure. Given are means ± SD of 6 biological replicates. Values that do not share the same letter are significantly different from each other as determined by ANOVA. Further photosynthetic parameters and full numerical data, including statistical evaluation, are provided as Supp. Tables S2-S5.

As GC PGLP1 amounts correlated with g_s, we analysed stomata count and size of all genotypes grown in air and elevated CO₂ to compare the impact of photorespiratory and non-photorespiratory conditions. As displayed in Fig. 3, stomata size, index, and density (significant only in the overexpressors), corelated with GC PGLP1 expression, as these parameters were in- and decreased in overexpression and antisense lines, respectively (Fig. 3A, Supp Fig. S3). These changes were photorespiration-dependent as they were absent in high CO₂-grown plants (Supp. Fig. S3). Based on these findings, we hypothesized if altered GC PGLP1 expression and 2-PG amounts could serve as morphogenetic signal for stomatal development. To test this assumption, increasing 2-PG concentrations (0, 10, 50, 100 µM) were externally applied to Arabidopsis wildtype-plants during cultivation on agar plates. Interestingly, characteristic stomatal determinants showed a negative correlation with external 2-PG application, as we measured gradually decreased length, width and smaller stomatal area and index compared to the control plants, lacking 2-PG in the growth media. However, 2-PG treatment had only minor effects on stomata density (Fig. 3C).

Plant were grown on soil in (A) air (400 ppm CO₂) to stage 5.1 [43] to determine stomata parameters. Displayed are correlation plots of selected stomatal parameters and GC-specific PGLP1 protein expression in the transgenic lines and the wildtype. Displayed are means ± SD (Σ 120 stomata per genotype was analyzed, from 4 biological replicates and 30 stomata per leaf) The corresponding high CO₂ and all uncorrelated data are shown in Supp. Fig. S3. The full numerical data is also provided as Supp. Table S6). Values that do not share the same letter are significantly different from each other as determined by ANOVA. (B) Arabidopsis wildtype in air (400 ppm CO₂) on half strength MS media supplemented with different 2-PG concentrations (0, 10, 50 and 100 µM). After 2–3 weeks stomatal parameters were determined by microscopic analysis as means ± SD (Σ 120 stomata per genotype, from 4 biological replicates and 30 stomata per leaf). Numbers in brackets indicate the reduction (in %, compared to control) with increased 2-PG amounts: stomatal area (-3.2%, -11.4% and − 22.3%), stomatal length (-2.3%, -5.9% and − 9.5%), stomatal width (-2.7%, -3.0% and − 11.5%), stomatal density (no significant changes), and stomatal index (-10.7%, -20.4% and − 27.7%). Full numerical data is provided in Supp. Table S7). Values that do not share the same letter are significantly different from each other as determined by ANOVA.

Because of the growth and photosynthetic responses of the transgenic lines, we quantified soluble sugars, starch, and 33 representatives of primary metabolism in leaves of all genotypes. Glucose and fructose levels were significantly higher in overexpression and lower in antisense lines. Sucrose was only higher in overexpressors (Fig. 4A-C), while transitory starch did not differ among genotypes (Fig. 4D). Further, leaf 2-PG and NAD⁺ amounts showed a negative, whilst 3-PGA a positive correlation with GC PGLP1 expression (Fig. 4E-F; Supp. Table S8). Among the other primary metabolites, we measured significant increases in glutamate, isoleucine, and isocitrate in the overexpression lines and significantly decreased arginine in the antisense lines. However, the calculation of total soluble sugars, amino and organic acid contents revealed all to increase and decrease in the GC PGLP1 overexpressors and antisense lines, respectively (Fig. 4G-H; Supp. Tab. S8).

Previous work suggested that GC starch and H₂O₂ amounts are involved in the energization and regulation of stomatal movements [44–46]. Hence, these parameters were measured under photorespiratory conditions, as no phenotypic or stomatal size variation were observed in high CO₂-grown plants. At one hand, we found a strong positive correlation between GC PGLP1 amounts and GC starch, which was significantly higher (~ 19–25%) in the overexpression and lower (~ 29–32%) in the antisense lines. On the other hand, H₂O₂ amounts were negative correlated with GC PGLP1 expression, being lower (~ 31–36%) in overexpression and higher (~ 27–29%) in antisense lines compared to the wildtype (Fig. 4I-J). Again, alterations in guard cell starch and H₂O₂ are photorespiration-dependent as both were statistically invariant among the analysed genotypes when grown under non-photorespiratory conditions, i.e. high CO₂ (Suppl. Table S8).

Plants were grown in air (400 ppm CO₂) to growth stage 5.1 [43]. Leaf-material was harvested at end of the day (11 h illumination) and analysed by (A-C) gas chromatography (GC), (D) spectrophotometrically, LC-MS/MS) liquid chromatography coupled to tandem mass spectrometry (E-H) and microscopic (I-J) analysis. Values are means ± SD (n > 6). Values that do not share the same letter are significantly different from each other as determined by ANOVA. Full numerical data set of all metabolites are provided in Supp. Table S8).

Given SBPase expression in leaves of various plants species positively correlates with growth and photosynthesis, and the fact that 2-PG is a potent inhibitor of SBPase activity [1, 32–34], we also manipulated the GC-specific SBPase protein expression. In clear contrast to PGLP1 manipulations, we did not observe any significant impact on the visual phenotype and quantitative growth parameters of overexpression (+ 11.3–15.7% GC SBPase protein expression) and antisense (-12.6-22.48% GC SBPase protein expression) lines under the same growth conditions (Supp. Fig. S4-S5). Furthermore, no significant change was seen on selected photosynthetic parameters, including A_N, g_s, and Γ (Supp. Fig. S4), measured as functions of varying light and CO₂.

Photorespiration is an unavoidable process in the primary C and N metabolism of plants, because it enables photosynthetic CO₂ assimilation by detoxifying the Rubisco oxygenation product 2-PG [5–6]. While the toxic effect of 2-PG on plant metabolism is well established [1–2, 47], it could also play a role as low CO₂-sensing molecule in oxygenic phototrophs as discussed to occur among cyanobacteria [48]. However, if 2-PG plays such a role among plants remained uncertain to date. To address the question if 2-PG is involved in CO₂-dependent stomata movements, we specifically manipulated the expression of the 2-PG-metabolizing enzyme PGLP1 in stomatal guard cells. This approach revealed a previously unrecognized function of PGLP1 or 2-PG in coordinating GC metabolism with stomatal behavior (Fig. 5).

Photorespiration is a highly compartmentalized process involving reactions in the chloroplast, peroxisome, mitochondrion, cytosol, and vacuole. While its roles and physiological implications are well established at the whole-leaf level, particularly in the mesophyll, its function in specific tissues and cell types remains largely enigmatic. Guard cell-specific manipulation of photorespiratory 2-PG removal, achieved through upregulation or antisense repression of PGLP1, reveals photorespiration as a key component of guard cell metabolism as well as stomatal behavior and development. Mechanistically, we propose that changes in CO₂ availability are sensed via the capacity of the photorespiratory flux, with the amount of the photorespiration-specific entrance metabolite 2-PG in chloroplasts serving as a key determinant. On the one hand, high chloroplast PGLP1 activity maintains low 2-PG levels and alleviates negative impacts on CBBC performance, starch biosynthesis, and ROS accumulation, particularly H₂O₂, thereby supporting the carbohydrate and energy status required to drive stomatal movements. On the other hand, reduced PGLP1 activity leads to 2-PG accumulation in chloroplasts, which slows the CBBC, decreases carbohydrate availability and metabolism, and promotes H₂O₂ accumulation. Consequently, stomata tend to remain more closed to prevent further damage to guard cells and the whole leaf resulting from Rubisco oxygenation and impaired photorespiration. Abbreviations: 2-PG − 2-phosphoglycolate; 3-PGA − 3-phosphoglycerate; 3-HP – 3-hydroxypyruvate; CBBC – Calvin-Benson-Bassham cycle; CAT – catalase; ETR – electron transport chain; GDC – glycine decarboxylase; g_s – stomatal conductance; Hex-P – hexose phosphates; OAA – oxaloacetate; PEP – phosphoenolpyruvate; PGLP – 2-PG phosphatase; Pyr – pyruvate; ROS – reactive oxygen species; SHMT1 – serine-hydroxymethyltransferase 1; TCA – tricarboxylic acid cycle

Our results provide a consistent picture: enhanced expression of PGLP1 in GC resulted in improved photosynthesis, higher stomatal conductance and enhanced growth, whereas antisense repression had opposite effects compared to wildtype (Fig. 1–2). Importantly, growth stimulation depended on Rubisco-mediated 2-PG formation and its subsequent photorespiratory metabolization, as the transgenic lines were indistinguishable from the wildtype under photorespiration-suppressing conditions (Fig. 1; Supp. Table S1). Similarly, the differences in the photosynthetic parameters became larger when gas exchange measurements were done under high O₂ conditions, stimulating photorespiration, while at lowered O₂ levels they were virtually absent (Fig. 2). These findings suggest that GC metabolism is naturally constrained by PGLP1 activity under ambient, i.e., by the capacity of photorespiratory flux, and that increasing this capacity can benefit stomatal function through enhanced movement dynamics and energization (Fig. 5). This is particularly noteworthy given the ongoing debate regarding the extent to which GC perform photosynthesis and, perhaps, rely on active photorespiratory metabolism. Furthermore, GC PGLP1 limitation could be explained by specific features of these specialized cells, as they contain significantly fewer and smaller chloroplasts ([49–51] and, thus, have generally naturally lower abundances of photorespiratory proteins. Until recently, the overall significance of photorespiratory metabolism in GC remained unclear due to the absence of guard cell-specific transgenic approaches. The first GC-specific manipulation of the mitochondrial photorespiratory enzyme glycine decarboxylase (GDC) provided evidence that these specialized cells are indeed capable of, and to some extent dependent on, active mitochondrial photorespiration [37]. This conclusion is consistent with two recent proteomic studies on mitochondria isolated from GC and their specific ATP metabolism [52–53] and earlier omics studies, showing the transcription and translation of the full photorespiratory core cycle in GC [42, 54].

The stimulation of growth, photosynthesis and stomatal conductance is consistent with earlier studies showing that GC-specific modifications of different processes can positively influence g_s and A_N [37, 55]. Interestingly, the photorespiration-specific results presented here, in conjunction with our earlier report, show that reprogrammed photorespiration fluxes in different subcellular organelles have a clear and consistent impact on overall plant performance under ambient laboratory conditions. However, it remains unresolved whether higher photosynthetic rates arise directly from changes in g_s or whether PGLP1, and thereby photorespiratory flux, modifications in GC signals increased CO₂ demand, which in turn prompts stomata to open more widely to support mesophyll photosynthesis and facilitate a higher energy status of the cells. Notably, the latter interpretation aligns with reports of photorespiratory optimizations, mainly PGLP1 and GDC overexpression, at the whole-leaf level, which also resulted in higher stomatal conductance [1, 14–16]. Given the previously used ST-LSI promoter is not fully mesophyll-specific and also drives expression in GC, it could well be that the observed responses are also caused by expression changes of both enzymes in GC. Taken together, these observations support the hypothesis that photorespiratory flux capacity, mediated through reinforcement or alleviation of negative feedback on carbon utilization, could serve as a key determinant for sensing and translating changes in external (C_a) and internal (C_i) CO₂ availability. More specifically, and supported by our findings, we suggest that chloroplastidal 2-PG could mechanistically serve as signaling metabolite translating altered photorespiratory fluxes in response to changes in CO₂ availability. The ultimate readout of such a mechanism could be shifts in the availability of photosynthates and other biomolecules at the whole-leaf level. Indeed, the metabolite profiles of the transgenic models support this hypothesis, given GC PGLP1 protein expression positively correlated with soluble sugars, as well as the total amino acid and organic acid contents (Fig. 4; Supp. Table S8). It should also be noted that changes in the GC photorespiratory flux, i.e. the chloroplastidal 2-PG amount, seem to be causative for optimized photosynthesis and growth, rather than alleviated negative feedback inhibition of the central CBB cycle enzyme SBPase as GC overexpression of the latter did not result in similar physiological responses (Supp. Figure 3–4).

GC starch availability and metabolism were reported to be a key determinate of GC energization and their rapid movements to acclimate to environmental fluctuations ([45, 56–57]. Although transitory starch stocks underwent no significant changes on the whole leaf basis among our transgenic plants, GC starch accumulation correlated with GC PGLP expression in the transgenic lines (Fig. 1, 4; Supp. Table S6). Hence, starch availability and turnover seem to be, at least to some extent, controlled by GC photorespiration. Similar alterations were found before, i.e., lowered 2-PG levels due to PGLP1 overexpression stimulated starch synthesis and elevated 2-PG levels due to PGLP1 antisense repressed starch accumulation on whole leaf basis [1]. However, if the different starch amounts in GC are a direct effect of altered GC photorespiration or its reprogrammed mesophyll metabolism and carbon import thereof has to be analyzed at higher resolution in the future. Nevertheless, increased GC starch seems to be a general response of GC overexpression of photorespiratory enzymes as similar observations were made on corresponding GDC manipulations [37].

In addition to starch, the GC-localized amounts of H₂O₂, another central player in stomatal regulation [44, 46, 58–59], were inversely correlated with GC PGLP1 expression in air (Fig. 4J). The exact origin of the altered H₂O₂ in our lines remains an open question. Photorespiratory H₂O₂ production seems unlikely, as flux scaling would predict opposite trends between overexpression and antisense lines, and the lack of H₂O₂ variations in high CO₂ (Supp. Table S8). Alternative sources, such as imbalances in mitochondrial or chloroplastidial electron transport and thereby produced H₂O₂, also lack support from our fluorescence and rETR(i) data. By contrast, NADPH oxidase activity emerges as a plausible candidate, potentially explaining the observed discrepancy between H₂O₂ accumulation and stomatal aperture. An additional possibility is that changes in ROS detoxification capacity contribute to the altered H₂O₂ profiles. Hence, in addition to the observed metabolic alterations, our findings highlight a potential role of H₂O₂ in linking GC PGLP1 activity and its potential role in CO₂ sensing to stomatal function. Typically, low concentrations of H₂O₂ promote stomatal opening via nuclear localization of KIN10 and subsequent induction of BAM1 and AMY3, driving starch degradation [46, 58]. At higher concentrations, however, H₂O₂ triggers stomatal closure through activation of Ca²⁺ channels and the anion channel SLAC1, largely mediated by NADPH oxidases (RBOHs) [60–61].

As discussed above, stomatal conductance could be directly or indirectly related to the reprogrammed GC metabolism via GC-specific PGLP1 manipulation. However, in contrast to the manipulation of GC-specific photorespiration due to GDC expression changes ([37], GC-specific PGLP1 manipulations also affected stomatal morphology. Specifically, GC PGLP1 abundance positively correlated with stomatal size and, to some extent, density (Fig. 3; Supp. Fig. S3 and Table S6). These morphological adaptations can certainly contribute to altered stomatal conductance, as maximal conductance (g_smax) largely depends on stomatal size and density [62]. Thus, it seems reasonable to assume that enhanced PGLP1 activity, and thereby more efficient degradation of GC 2-PG, leading to lower steady-state GC 2-PG levels, could underlie the observed changes in stomatal size. Given direct quantification of GC 2-PG remains technically challenging, we tested whether exogenous 2-PG influences stomatal traits in Arabidopsis wild type. Indeed, increasing external 2-PG supply gradually reduced stomatal dimensions (Fig. 4C; Supp. Table S7), supporting the hypothesis that optimal GC PGLP1 activity, through 2-PG detoxification, is fundamental for maintaining proper GC and stomatal morphology. This observation also provides evidence that not changed amounts of GC PGLP1, but directly its substrate 2-PG, serves as signaling molecule. This finding, thus, could be taken as direct hint for its role in the signaling of different CO₂ levels not only via impacting on GC movements, but also GC development in plants. This statement is in line with evolutionary observations that low CO2 (high 2-PG) selected for smaller, whilst high CO2 (low 2-PG) for larger stomata [21–23].

Overall, our findings strongly support the view that GC photorespiration, including GC-specific 2-PG degradation, is a fundamental component of stomatal metabolism and behavior. This metabolic framework may serve as the basis for coordinating environmental variations that strongly influence photorespiratory fluxes with GC behavior and mesophyll metabolism. By modulating 2-PG detoxification and ROS homeostasis within GC, PGLP1 influences both stomatal size and conductance, thereby regulating CO₂ availability for photosynthesis. Together, these results highlight GC photorespiration as an underappreciated target for enhancing crop productivity, particularly under conditions where photorespiration is active.

Guard cell-specific transgenic Arabidopsis lines were generated to achieve overexpression or antisense-mediated reduction of PGLP1 and SBPase expression. The binary plant transformation vector pG0229:AtGC1:35STer, containing the guard cell-specific GC1 promoter [37, 42], served as the expression backbone. The full coding sequence (CDS) of Solanum lycopersicum PGLP1 (SlPGLP1; 1119 bp) was synthesized de novo (BaseClear, Leiden, The Netherlands). The CDS of Arabidopsis SBPase (AtSBPase; 1182 bp) was PCR-amplified from Col.0 cDNA using primers P967 and P968 (sequences listed in Supp. Table S9) with a proof-reading DNA polymerase and cloned into pJET2.1 (ThermoFisher Scientific, Schwerte, Germany) for sequence verification and amplification. The coding fragments were excised from their entry vectors using BamHI (SlPGLP1) and XmaI (AtSBPase), respectively, and ligated into pG0229:AtGC1:35STer in sense and antisense orientations to create overexpression constructs pG0229:AtGC1:SlPGLP1_sense:35STer and pG0229:AtGC1:AtSBPase_sense:35STer and antisense constructs pG0229:AtGC1:SlPGLP1_antisense:35STer and pG0229:AtGC1:AtSBPase_antisense:35STer (see Supp. Fig. S1 and S3). All final constructs were verified by sequencing (Microsynth, Göttingen, Germany). Subsequently, the constructs were introduced into Agrobacterium tumefaciens GV3101 + pSOUP, the drug resistant colonies verified via standard PCR procedures, and used for Arabidopsis floral dip transformation [63]. T1 seeds were surface sterilized and selected on half-strength MS media supplemented with 20 µg mL^− 1 phosphinothricin (BASTA). Resistant seedlings were transplanted to soil, PCR-verified for the presence of the transgene, and propagated to homozygous T3 or T4 lines, used for all physiological experiments. For comprehensive characterization, two independent SlPGLP1 and three independent AtSBPase overexpression and antisense lines were used.