Identification and characterization of a plastid-localized Arabidopsis glyoxylate reductase isoform: comparison with a cytosolic isoform and implications for cellular redox homeostasis and aldehyde detoxification

Abstract

Enzymes that reduce the aldehyde chemical grouping (i.e. H-C=O) to its corresponding alcohol could be crucial in maintaining plant health. Recently, recombinant expression of a cytosolic enzyme from Arabidopsis thaliana (L.) Heynh (designated as glyoxylate reductase 1 or AtGR1) revealed that it effectively catalyses the in vitro reduction of both glyoxylate and succinic semialdehyde (SSA). In this paper, web-based bioinformatics tools revealed a second putative GR cDNA (GenBank Accession No. AAP42747; designated herein as AtGR2) that is 57% identical on an amino acid basis to GR1. Sequence encoding a putative targeting signal (N-terminal 43 amino acids) was deleted from the full-length GR2 cDNA and the resulting truncated gene was co-expressed with the molecular chaperones GroES/EL in Escherichia coli, enabling production and purification of soluble recombinant protein. Kinetic analysis revealed that recombinant GR2 catalysed the conversion of glyoxylate to glycolate (K(m) glyoxylate=34 microM), and SSA to gamma-hydroxybutyrate (K(m) SSA=8.96 mM) via an essentially irreversible, NADPH-based mechanism. GR2 had a 350-fold higher preference for glyoxylate than SSA, based on the performance constants (k(cat)/K(m)). Fluorescence microscopic analysis of tobacco (Nicotiana tabacum L.) suspension cells transiently transformed with GR1 linked to the green fluorescent protein (GFP) revealed that GR1 was localized to the cytosol, whereas GR2-GFP was localized to plastids via targeting information contained within its N-terminal 45 amino acids. The identification and characterization of distinct plastidial and cytosolic glyoxylate reductase isoforms is discussed with respect to aldehyde detoxification and the plant stress response.

Fig. 1.

ClustalW comparison of the predicted amino acid sequences for the full-length GR1 and GR2. Numbers indicate amino acid number. The first 51 amino acids in GR2 represent its putative targeting sequence, and the arrows indicate peptide cleavage sites as predicted by three web-based algorithms (ChloroP, TargetP, PSORT) for predicting subcellular localization. Identical and similar amino acids are shown by black and grey shading, respectively.

Fig. 2.

Expression and purification of recombinant GR2 extracted from E. coli cells co-expressing the GroES/GroEL chaperone complex. The top row illustrates SDS-PAGE analysis of protein lysates from cells transformed with the truncated gene insert (A) or the empty insert (B), respectively, whereas the bottom row illustrates SDS (C) and immunoblot (D) analysis of protein lysates from cells transformed with the truncated gene insert. The gels were stained with Coomassie Blue, whereas the immunoblot was probed with an anti-His antibody. The EL subunit of the chaperone complex (1) and the recombinant protein (2) are indicated.

Fig. 3.

Dependence of GR2 activity on pH. Activity was determined using saturating glyoxylate and NADPH as substrates, and 2-morpholino-ethanesulphonic acid (pH 5.5–6.8), HEPES (pH 6.8–8.2), N-tris(hydroxymethyl)methyl-4-aminobutanesulphonic acid (pH 8.2–9.6), and 3-(cyclohexylamino)-1-propanesulphonic acid (pH 9.7) as buffers. Data represent the mean ±SE. of triplicate measurements from a typical enzyme preparation.

Fig. 4.

Localization of GR1-GFP and GR2-GFP in transformed BY-2 cells. Images represent the localization of transiently-expressed or endogenous proteins in (co-) transformed (via biolistic bombardment) cells including (A) GR1-GFP and (B) RFP, (C) GR2-GFP and (D) NAGK, (F) GR2 1–45-GFP and (G) NAGK, and (I) GR2 Δ2–45-GFP and (J) NAGK. (E) and (H) represent the corresponding merged images of the same cells shown in (C) and (D), and (F) and (G), respectively. Bar=10 μm.

Fig. 5.

Localization of GR2-GFP in differentially-permeabilized BY-2 cells. Images shown represent individual GR2-GFP-transformed cells that were fixed and permeabilized with either triton X-100 (top rows) or digitonin (bottom rows) and then immunostained with either (A) anti-NAGK and anti-tubulin or (B) anti-GFP and anti-tubulin antibodies. Note that in the GR2-GFP-transformed cells that were permeabilized with digitonin, neither endogenous NAGK, nor GR2-GFP, was immunostained. Bar=10 μm.