Defense against Reactive Carbonyl Species Involves at Least Three Subcellular Compartments Where Individual Components of the System Respond to Cellular Sugar Status

Abstract

Methylglyoxal (MGO) and glyoxal (GO) are toxic reactive carbonyl species generated as by-products of glycolysis. The pre-emption pathway for detoxification of these products, the glyoxalase (GLX) system, involves two consecutive reactions catalyzed by GLXI and GLXII. In Arabidopsis thaliana, the GLX system is encoded by three homologs of GLXI and three homologs of GLXII, from which several predicted GLXI and GLXII isoforms can be derived through alternative splicing. We identified the physiologically relevant splice forms using sequencing data and demonstrated that the resulting isoforms have different subcellular localizations. All three GLXI homologs are functional in vivo, as they complemented a yeast GLXI loss-of-function mutant. Efficient MGO and GO detoxification can be controlled by a switch in metal cofactor usage. MGO formation is closely connected to the flux through glycolysis and through the Calvin Benson cycle; accordingly, expression analysis indicated that GLXI is transcriptionally regulated by endogenous sugar levels. Analyses of Arabidopsis loss-of-function lines revealed that the elimination of toxic reactive carbonyl species during germination and seedling establishment depends on the activity of the cytosolic GLXI;3 isoform. The Arabidopsis GLX system involves the cytosol, chloroplasts, and mitochondria, which harbor individual components that might be used at specific developmental stages and respond differentially to cellular sugar status.

Figure 1.

The Glyoxalase System Is the Main Pathway for MGO Detoxification in Eukaryotic Cells. (A) MGO reacts spontaneously with GSH to form a hemithioacetal (HTA), which is the substrate of GLXI, producing S-d-lactoylglutathione (SD-l-GSH). GLXII converts SD-l-GSH into d-lactate. (B) Simplified schematic overview of the phylogenetic relationships of all Arabidopsis GLXI proteins. GLXI;1 and GLXI;2 form a clade with the Ni²⁺-dependent GLXI proteins (E. coli), whereas GLXI;3 clusters together with the Zn²⁺-dependent GLXI proteins (Homo sapiens) (Kaur et al., 2013). (C) The Arabidopsis genome encodes five GLXII-like proteins; however, S-d-lactoylglutathione hydrolase activity has only been reported for GLXII;2, GLXII;4, and GLXII;5. Sequences were retrieved from arabidopsis.org, and MEGA7 was used to reconstruct the maximum likelihood tree. The maximum likelihood tree is drawn to scale, with branch lengths measured in the number of substitutions per site (scale bar). Numbers at each node indicate boot strap values.

Figure 2.

Arabidopsis Genes Encoding Enzymes of the GLX System. (A) Gene models of the predicted splice forms (SF; indicated by underscore) of GLXI and GLXII. Box, exon; black, coding sequence; gray, untranslated region. (B) Predicted subcellular localization of the proteins arising from the splice forms. aa, amino acids; C, chloroplast; M, mitochondrion; S, secretory pathway; N, nucleus. The AramLocCon score was taken from Schwacke et al. (2007) and describes a consensus subcellular localization prediction based on 20 individual prediction programs for C, M, S, or N. +3 correspond to 3 additional base pairs at the beginning of exon 2 in GLXII;4 and GLXII;5.

Figure 3.

Transcript Abundance of GLX Homologs and Their Respective Splice Forms in Different Arabidopsis Organs. (A) Amplification of predicted GLXI splice forms in Arabidopsis organs via RT-PCR. Specific primer pairs were used to amplify the GLXI splice forms from cDNA from leaves and roots of 3-week-old Arabidopsis plants grown on MS medium. M, molecular size standard. (B) FPKM values of GLXI splice forms. (C) FPKM values of GLXII splice forms. Raw data were taken from Liu et al. (2012) and mapped with high stringency to the TAIR 10 annotation (Berardini et al., 2015).

Figure 4.

Metal Activation Profile of Arabidopsis GLXI Isoforms. Absolute activities of GLXI;1 (A), GLXI;2 (B), and GLXI;3 (C) with MGO-GSH and GO-GSH (mmol µg⁻¹ min⁻¹) using different bivalent metal ion cofactors. e, scientific notation (times 10 raised to the power of exponent). Values are means ± se (n = 3) of independent enzyme preparations. Testing for significant differences was performed by two-way ANOVA for GLXI;1 (substrate, P = 0.021; ion, P < 0.0001; interaction, P < 0.0001), for GLXI;2 (substrate, P = 0.20; ion, P < 0.0001; interaction, P < 0.0001), and for GLXI;3 (substrate, P = 0.016; ion, P < 0.0001; interaction, P < 0.0001) (Supplemental File 2), with subsequent Dunnett’s multiple comparisons test to w/oI (without ion) control, a = P < 0.05.

Figure 5.

Functional Complementation of the S. cerevisiae Mutant glo1Δ. Growth of wild-type and glo1Δ strains transformed with pDR195 as empty vector control (EV) and glo1Δ transformed with pDR195:GLXI;1, pDR195:GLXI;2, or pDR195:GLXI;3 on 1 and 4 mM MGO or GO was monitored up to 48 h at 30°C. Cell numbers are indicated at the bottom.

Figure 6.

Subcellular Localization of Arabidopsis GLXI Isoforms Using Tobacco Leaf Protoplasts. (A) The coding sequence of YFP was fused in-frame with the C terminus of the full-length coding sequence of GLXI;1_1, GLXI;2_1, GLXI;3_1, and GLXI;3_4 (GLXI:YFP fusions) and expressed in tobacco leaf protoplasts. GLXI proteins with high probability of mitochondrial localization (AramLocCon Score >3) were coexpressed with the mitochondrial marker protein mito:eq611. (B) YFP was fused in frame with the N terminus of the full coding sequence of GLXI;1_1, GLXI;2_1, and GLXI;3_4 (YFP:GLXI fusions) and coexpressed with a peroxisomal marker protein in tobacco leaf protoplasts. Images were taken after 3 d of infiltration. merge, protoplast overview; merge_zoom, 50 µm detail; YFP, YFP fluorescence; mitochondria or peroxisome, fluorescence of marker proteins for the organelles; Chl_auto, autofluorescence of chlorophyll; n.r., not required. Bar = 10 µm.

Figure 7.

Subcellular Localization of Arabidopsis GLXII Isoforms in Tobacco Leaf Protoplasts. The coding sequence of YFP was fused in-frame with the C terminus of the full-length coding sequence of GLXII;2, GLXII;4_1, GLXII;5_1, and GLXII;5_2 (GLXII:YFP fusions) and expressed in tobacco leaf protoplasts. GLXII proteins with high probability of mitochondrial localization were coexpressed with the mitochondrial marker protein mito:eq611. Images were taken after 3 d of infiltration. merge, protoplast overview; merge_zoom, 50 µm detail; YFP, YFP fluorescence; mitochondria, fluorescence of marker proteins for the organelles; Chl_auto, autofluorescence of chlorophyll; n.r., not required. Arrows indicate colocalization of YFP and marker protein fluorescence. Bar = 10 µm.

Figure 8.

Transcriptional Regulation of the GLXI Isoforms by Different Stresses. (A) Expression levels of GLXI;1, GLXI;2, and GLXI;3 relative to the untreated control (set to 0) in whole 2-week-old Arabidopsis seedlings grown in the presence of 0.5 mM MGO or 60 mM sucrose. (B) Expression levels of GLXI;1, GLXI;2, and GLXI;3 relative to the untreated control (set to 0) in 4-week-old Arabidopsis rosettes of wild-type and ADP-glucose pyrophosphorylase loss-of-function (adg1-1) plants after prolonged darkness (24 h dark). ACTIN2 was used as the endogenous reference. Values are means ± se (n = 3, biological replicates with three to four individual plants pooled, *P < 0.05, unpaired two-tailed Student’s t test). (C) Relative expression levels of GLXI;1, GLXI;2, and GLXI;3 after sucrose treatment of Arabidopsis seedlings (array data from Gonzali et al. [2006]). (D) Relative expression levels of GLXI;2 in Arabidopsis leaves after treatment with 100 mM sucrose for 16 h (*P < 0.001; array data from Müller et al. [2007]). (E) Relative expression levels of GLXI;1, GLXI;2, and GLXI;3 in Col-0 leaves after 48 h of high light (48 h HL) exposure (*P < 0.05, Bonferroni adjusted; array data from Schmitz et al. [2014]). n.d., no data; FC, fold change.

Figure 9.

Effect of Sucrose, MGO, and GO on the Germination and Seedling Establishment of Arabidopsis GLXI T-DNA Insertion Lines. (A) Germination (in percentage of total seeds sown) at 2 d after seed imbibition. The emergence of the radicle was used as an indicator of germination. Values are means ± se (n = 4) from independent experimental trials with 50 to 100 counted seeds for each genotype. Testing for significant differences was performed by two-way ANOVA on arcsine transformed data (treatment, P < 0.0001; genotype, P < 0.0001; interaction, P = 0.0005). (B) Root length after 12 d of germination, values are means ± se (n = 4) from independent experimental trials with 7 to 10 root measurements each. Testing for significant differences was performed by two-way ANOVA (treatment, P < 0.0001; genotype, P < 0.0001; interaction, P < 0.0001) (Supplemental File 2). Tukey’s multiple comparison test was used, where “a” indicates significant differences in treatments versus MS control and “b” indicates significant differences in genotypes versus Col-0 control. (C) Photographs of Col-0 and glxI;3 loss of function lines after 12 d of germination in MS medium and in the presence of sucrose, MGO, and GO. Bar = 1 cm.

Figure 10.

Proposed Production and Detoxification Pathways of MGO within a Plant Cell. The fluxes through glycolysis and the Calvin Benson cycle (CBC) determine the rate at which MGO molecules are formed in the cytosol and chloroplast, respectively. Our results indicate that in the cytosol, GLXI;2 and/or GLXI;3 convert MGO and GSH into S-d-lactoylglutathione. This can be further converted to d-lactate by cytosolic GLXII;2 or transported to the mitochondria, where it is converted to d-lactate by GLXII;4 and/or GLXII;5. In the chloroplasts, GLXI;1 and/or GLXI;3 convert MGO and GSH into S-d-lactoylglutathione, which is further converted to d-lactate by chloroplastic GLXII;4 and/or GLXII;5. In the mitochondria, d-lactate is finally converted by d-lactate dehydrogenase (D-LDH) to pyruvate, which is fed into the tricarboxylic acid cycle (TCA). TPI, triose phosphate isomerase; PGA, 3-phosphoglycerate; F-1,6-BP, fructose-1,6-bisphosphate. Dashed arrows indicate possible transport; dotted arrows indicate summarized reactions.