**3. Discussion**

The main function of the photorespiratory pathway is the efficient removal of critical intermediates, especially 2-PG, which severely inhibits carbon fixation and allocation [6,10]. Furthermore, it has been suggested that photorespiration plays a role in the abiotic stress response of plants [25–27]. Therefore, we hypothesized that an increased photorespiratory flux might facilitate plant abiotic stress tolerance. To test this hypothesis, we used transgenic lines overexpressing photorespiratory *PGLP*, previously shown to fix more carbon at high photorespiratory pressures and to display improved starch metabolism and stomatal movements [10].

Consistent with our hypothesis, we found improved photosynthetic performance after short-term and long-term abiotic stresses. In a first series of experiments, *PGLP* overexpression lines displayed higher CO2 assimilation accompanied with improvements in other gas exchange parameters after increasing the temperature to 30 ◦C during short-term high light treatment, anticipated to promote 2-PG production (Figure 2B). The maintenance of higher photosynthesis in *PGLP* overexpressors under short-term temperature stress is likely through diminished inhibition of the CB cycle by 2-PG, but we cannot rule out that increased photorespiration eventually helps to prevent the chloroplastidal electron transport chain from overreduction as suggested before [26,28,29]. Similar results were obtained if photosynthesis of wild-type and *PGLP* overexpressor plants was characterized under water-limiting conditions. Water shortage promotes stomatal closure [36] and eventually increases 2-PG levels due to a higher RuBP fraction being oxidized. Notably, both transgenic lines showed significant improvements in photosynthetic parameters after 13 days of water shortage (Figure 3). Interestingly, we did not only measure higher CO2 assimilation and lower CO2 compensation points, but also higher *gs*, indicating altered stomatal movements (Figure 3B). Given that photorespiration is also involved in proper guard cell metabolism [30,31], an optimized flux through the pathway in mesophyll cells could be beneficial for these specialized cells, too. Accordingly, it is likely to assume that the altered leaf-carbohydrate metabolism, in particular starch biosynthesis, can facilitate allocation of carbon from the mesophyll to the guard cells in order to enhance their energy supply. It should be noted that *PGLP* overexpression is driven through the *ST-LSI* promoter, therefore, changes in *PGLP* expression are not restricted to the mesophyll cells. Hence, changes in *gs* could also be directly caused by altered PGLP activity in guard cells.

Finally, long-term exposure at elevated temperatures was analyzed in more detail. This was done for three reasons: (i) elevated temperatures favor 2-PG production, (ii) higher PGLP activities were found to be beneficial for photosynthesis on a short-term (Figure 2), and (iii) future climate change scenarios predict an increase in temperature on a global scale [13,14]. Consequently, higher *PGLP* activity could eventually be a positive trait for plant engineering. In agreement with the short-term exposure to high light and elevated temperature (Figure 2) and growth under water-limiting conditions (Figure B), the overexpressor line O1 maintained higher photosynthesis after long-term exposure to elevated temperature (Figure 4). This change was accompanied by higher *gs* and increased transpiration, again suggesting changes in stomatal movements. Given the strong impact 2-PG has on CO2 fixation and carbohydrate utilization and allocation [6,10], we quantified amounts of starch and the soluble sugars sucrose, glucose and fructose (Figure 5). In line with previous findings [10], overexpressors of *PGLP* store somewhat more starch under standard conditions, without a significant impact on sucrose synthesis. Interestingly, these trends were kept after the shift to elevated temperature until day 7 at 30 ◦C. These data indicate faster 2-PG removal facilitates carbon allocation to starch biosynthesis also under the stress conditions. However, we cannot neglect the possibility that starch is degraded somewhat less fast in the transgenic lines. The assumption of faster starch synthesis is supported by the decreased amounts of soluble sugars in both lines during the first days at 30 ◦C. However, after 7 days at 30 ◦C, this trend became reversed, which might be a compensatory reaction of carbon metabolism under long-term temperature stress. The effects of *PGLP* overexpression seem to be mainly restricted to alterations in photosynthesis and carbohydrate metabolism, since amino and organic acid contents showed only minor changes.

Acclimation to elevated temperatures obviously activates the photorespiratory activity in wild-type plants, because a coordinated increase of mRNAs and proteins for many photorespiratory enzymes was observed (Figure 7), which is consistent with previous reports [25]. Given the lack of response on the expression of the photorespiratory enzymes in the transgenic lines, one might speculate upregulation of *PGLP*, and in turn lowering the steady-state content of 2-PG, already acts as signal to indicate sufficient acclimation of the photorespiratory flux. Additionally, this result once more demonstrates the importance of fast 2-PG removal via photorespiration, which might be sufficient to cope with the temperature stress. Furthermore, 2-PG could also play a regulatory role in the stress acclimation as shown for serine before [37]. Direct involvement of 2-PG as inducer for the transcription of genes involved in CO2 uptake was observed in cyanobacteria [38], but has not yet reported for plants. However, the *PGLP* knock out mutant showed strong and specific alterations in gene expression after a shift from high to low CO2 [31]. Therefore, it cannot be excluded completely that 2-PG mediated transcriptional reprogramming mechanisms exist in plants.
