Molecular and Biochemical Characterization of Cytosolic Phosphoglucomutase in Maize

Abstract

Phosphoglucomutase (PGM) catalyzes the interconversion of glucose (Glc)-1- and Glc-6-phosphate in the synthesis and consumption of sucrose. We isolated two maize (Zea mays L.) cDNAs that encode PGM with 98.5% identity in their deduced amino acid sequence. Southern-blot analysis with genomic DNA from lines with different Pgm1 and Pgm2 genotypes suggested that the cDNAs encode the two known cytosolic PGM isozymes, PGM1 and PGM2. The cytosolic PGMs of maize are distinct from a plastidic PGM of spinach (Spinacia oleracea). The deduced amino acid sequences of the cytosolic PGMs contain the conserved phosphate-transfer catalytic center and the metal-ion-binding site of known prokaryotic and eukaryotic PGMs. PGM mRNA was detectable by RNA-blot analysis in all tissues and organs examined except silk. A reduction in PGM mRNA accumulation was detected in roots deprived of O₂ for 24 h, along with reduced synthesis of a PGM identified as a 67-kD phosphoprotein on two-dimensional gels. Therefore, PGM is not one of the so-called “anaerobic polypeptides.” Nevertheless, the specific activity of PGM was not significantly affected in roots deprived of O₂ for 24 h. We propose that PGM is a stable protein and that existing levels are sufficient to maintain the flux of Glc-1-phosphate into glycolysis under O₂ deprivation.

In cells of prokaryotic and eukaryotic organisms, PGM (EC 5.4.2.2), a phosphoenzyme, catalyzes an important trafficking point in carbohydrate metabolism. In one direction, Glc-1-P produced from Suc catabolism is converted to Glc-6-P, the first intermediate in glycolysis. In the other direction, conversion of Glc-6-P to Glc-1-P provides a substrate for synthesis of UDP-Glc, which is required for synthesis of a variety of cellular constituents, including cell wall polymers and glycoproteins. An obligatory step in the PGM reaction mechanism is the transfer of phosphate from the Ser of the phosphoenzyme to the substrate (Glc-6-P or Glc-1-P) to form the dephosphorylated enzyme and Glc-1,6-P₂, a catalytic intermediate and cofactor.

Two isoforms of PGM are present in plants, one located in the cytosol and the other in the chloroplast stroma (Mühlbach and Schnarrenberger, 1978). In maize (Zea mays L.) cytosolic isozymes of PGM are encoded by pgm1 and pgm2 (Goodman et al., 1980; Stuber and Goodman, 1983), which map to duplicated chromosomal regions on 1L and 5S, respectively (Helentjaris et al., 1988). The products of these loci are monomeric and are detected in cell extracts from various tissues, including roots, coleoptiles, leaves, scutella, and pollen (Stuber and Goodman, 1983). Despite the extensive use of cytosolic PGM isozymes as genetic markers, very little is known about pgm gene structure or expression in plants.

Gene expression in plants is regulated both developmentally and by the environment. Environmental stresses such as flooding (O₂ deprivation/hypoxia and anoxia) affect Glc-P metabolism in many agronomically important plants (for review, see Drew, 1997). During the first few hours of O₂ deprivation in maize, the synthesis of proteins made under aerobic conditions is rapidly suppressed, whereas the synthesis of a small group of proteins called ANPs is enhanced (Sachs et al., 1980). The increased synthesis of the ANPs is not only due to increased transcription of specific genes that encode them but also to selective mRNA translation (Bailey-Serres and Freeling, 1990; Russell and Sachs, 1992; Fennoy and Bailey-Serres, 1995; Sachs et al., 1996; Manjunath and Sachs, 1997; S. Fennoy, T. Nong, and J. Bailey-Serres, unpublished data).

Many of the proteins synthesized in O₂-deprived roots participate in Suc breakdown, Glc-P metabolism, or ethanolic fermentation (for review, see Sachs et al., 1996). The ANP Suc synthase catalyzes the formation of UDP-Glc, which is subsequently converted to Glc-1-P, the substrate of PGM. Enzymes involved in Glc-P metabolism are also up-regulated during the development of the kernel embryo and endosperm (Chourey, 1981). This up-regulation is due to transcriptional induction of these genes, as evidenced by transient expression analysis of alcohol dehydrogenase1 (adh1) promoter constructs (Klein et al., 1989) and increased transcription of shrunken1 (encoding Suc synthase) in nuclei isolated from developing kernels (Kodrzycki et al., 1989).

As a first step toward a molecular characterization of plant cytosolic PGM, we describe the isolation and characterization of full-length cDNAs that encode cytosolic PGMs of maize. Maize lines with distinct pgm1 and pgm2 genotypes were used to determine if the cDNAs encode the known cytosolic isozymes of PGM. Since enzymes involved in Glc-P metabolism are expressed during seed development and in response to O₂ deprivation, RNA analysis was performed to examine the developmental and environmental regulation of PGM mRNA accumulation. Molecular and biochemical techniques were used to confirm that PGM is a phosphoprotein in extracts from maize roots and to examine the synthesis of this phosphoprotein in O₂-deprived roots.

MATERIALS AND METHODS

Reagent-grade chemicals were purchased from United States Biochemical, Sigma, or Boehringer Mannheim. [γ-³²P]ATP (3000 Ci mmol⁻¹) and [³⁵S]Met (1200 Ci mmol⁻¹) were purchased from DuPont-New England Nuclear and [α-³²P]dCTP was purchased from Andotek (Tustin, CA).

Plant Material

Genetic stocks were maize (Zea mays) inbred line B73 (Pgm1-9 and Pgm2-4; provided by Pioneer Hi-Bred International, Johnston, IA), Pgm1-null (Pgm1-null and Pgm2-3) and Pgm2-null (Pgm1-5 and Pgm2-null) bred from a line segregating Pgm1-null, Pgm1-5; and Pgm2-null, Pgm2-3 (a gift of Dr. Charles Stuber, North Carolina State University, Raleigh). Maize plants were grown in the field for collection of leaves of ear husks, immature ears, and silk at silk emergence, and pollen at anthesis. The embryo, aleurone, and endosperm were collected at 15, 20, and 25 DAP. “Endosperm” refers to all tissue within the pericarp and aleurone excluding the embryo. “Embryo” includes the scutellum and embryonic axis. “Aleurone” refers to the aleurone and attached pericarp. Maize seeds were germinated in the dark for 4 to 5 d at room temperature, as described by Fennoy and Bailey-Serres (1995). Primary roots (3–5 cm) and coleoptiles were cut directly into liquid N₂ and stored at −70°C until further use.

DNA Probes

We identified a rice EST cDNA (gift of Rice Genome Research Program NIAR/STAFF, Ibaraki, Japan; GenBank accession no. D24288) as a putative PGM clone on the basis of protein coding sequence similarity to human PGM (GenBank accession no. M83088). Rice and maize pgm cDNA inserts were obtained by digesting DNA clones with EcoRI and XhoI and electrophoresis on a low-melt agarose gel. cDNAs were labeled with [α-³²P]dCTP using a random-prime labeling kit (Promega). For adh1, a 893-bp PstI fragment was isolated by restriction digestion of the recombinant plasmid pZmL793 (Dennis et al., 1984). The 18S rDNA probe was a 210-bp fragment from a tomato clone DB292 (D. Bird, personal communication).

cDNA Library Screening and DNA Sequencing

A cDNA library constructed in the Uni-ZAP-XR vector from poly(A⁺) mRNA of 6-h O₂-deprived roots (3-d-old seedlings) of maize inbred B73 Ht was generously provided by Dr. M.M. Sachs, U.S. Department of Agriculture/Agricultural Research Station (Urbana, IL). Approximately 300,000 plaque-forming units were screened according to the manufacturer's protocol using a cDNA-synthesis kit (λ-ZAPII, Stratagene), with the rice EST as the probe at moderate stringency (hybridization in 7% SDS [w/v] and 50 mm NaPO₄, pH 7.0, at 55°C; washed with 0.1% SDS, 0.1 × SSC at 55°C). Both strands of cDNA were sequenced using the fmol DNA cycle-sequencing system (Promega) with commercially available and custom primers (Heligen Laboratory, Huntington Beach, CA). A portion of the sequencing was performed by the DNA Sequencing Core Laboratory (University of Florida, Gainsville). Nucleotide and amino acid sequence analyses were performed with the computer programs ClustalW (Thompson et al., 1994) and Alscript (Barton, 1993) from the Genetics Computer Group (Madison, WI).

Southern Analysis

Genomic DNA was isolated from 5-d-old maize roots using the mixed alkyltrimethylammonium bromide (M-7635, Sigma) extraction procedure (Boyce et al., 1989). Fifty micrograms of DNA was digested with HindIII (2.5 units μg⁻¹ DNA) overnight in a reaction volume of 200 μL. The DNA was concentrated by ethanol precipitation, fractionated on a 1% agarose gel, and transferred to a nylon membrane (MagnaGraph, MSI, Westbourgh, MA) (Sambrook et al., 1989). Prehybridization was carried out in 50% (v/v) formamide, 5× SSC, 20 mm NaPO₄, pH 6.8, 1% (w/v) SDS, Denhardt's solution (0.02% [w/v] PVP, 0.02% [w/v] Ficoll, and 0.02% [w/v] BSA), 5% dextran sulfate (w/v), and 100 μg mL⁻¹ of sheared and heat-denatured salmon-sperm DNA (Sambrook et al., 1989) for 4 h at 42°C. Hybridization to a [³²P]-labeled pgm2 cDNA probe was carried out with fresh prehybridization solution for 16 to 18 h at 42°C. Blots were washed and exposed to Hyperfilm (Amersham), as described by Manjunath and Sachs (1997).

Seed Germination, O₂-Deprivation Treatment, and Labeling in Vivo with [³⁵S]Met

O₂-deprivation treatment was by submergence of intact seedlings in induction buffer (0.5 mm Tris-HCl, pH 8.0, 7.5 μg mL⁻¹ chloramphenicol) sparged with argon in a closed Mason jar, as described by Fennoy and Bailey-Serres (1995). Dissolved O₂ concentration was measured with a Clark-type electrode (no. 97-08-00, Orion, Cambridge, MA) and meter (no. 611, Orion). For labeling proteins in vivo, the apical 1 cm of the primary root of intact seedlings was immersed in 150 μCi [³⁵S]Met mL⁻¹ induction buffer in a 1-mL plastic syringe with the plunger removed and the tip sealed with Parafilm. Seedlings were taken from open trays (aerobic) or after submergence in induction buffer sparged with argon for 22 h (22 h O₂ deprived). Labeling was carried out for the final 2 h of the treatment in a humidified Mason jar that was open and maintained in air (aerobic) or closed and sparged with argon (24 h O₂ deprived).

RNA Isolation and Northern Hybridization

Total RNA was extracted from various maize organs and tissues following the CsCl-gradient method described by Cone et al. (1986). RNA (20 μg) was electrophoresed on a 2.2 m formaldehyde/1.3% agarose gel and transferred onto a nylon membrane (MSI) by capillary action with 10× SSC. Hybridizations were performed as described for Southern blots. Blots were exposed to a phosphor imager screen and quantified using ImageQuant software (Molecular Dynamics, Sunnyvale, CA). To correct for differences in loading, 18S rDNA was used as a hybridization probe. When needed, blots were regenerated by soaking in 0.1× SSC, 0.5% SDS for 15 min at 90°C.

Root Protein Extraction and Phosphorylation in Vitro

All procedures were carried out at 4°C unless otherwise indicated. Roots were homogenized in 2 mL of extraction buffer (20 mm Tris-HCl, pH 8.0, 4 mm DTT, unless indicated otherwise in figure legends) per gram of fresh weight with a pestle (Kontes, Vineland, NJ) in a microcentrifuge tube. The extract was centrifuged for 5 min at 16,000g, and the crude supernatant fraction was collected for enzyme assays, for in vitro phosphorylation, or for estimation of soluble-protein concentration. Protein extracts (10–30 μg) were phosphorylated for 6 min at room temperature in a 30-μL reaction mixture containing root extract, 5 μCi [γ-³²P]ATP, and one or more of the following: 0.1 mm Glc-1,6-P₂, 5 mm MgCl₂, 2 mm Glc, and 2 units mL⁻¹ hexokinase (H-5375, Sigma), as indicated in the figure legends.

Assay of Enzyme-Specific Activity

Specific activity of enzymes was assayed using crude cell extracts in 20 mm Tris-HCl, pH 8.0, 4 mm DTT, and 5 mm MgCl₂. An assay coupled with Glc-6-PDH and the reduction of NADP⁺ was used to determine PGM activity. This assay was carried out in a 250-μL reaction mixture containing 20 mm Tris-HCl, pH 7.5, 10 mm MgCl₂, 1 mm Na₂-Glc-1-P, 0.25 mm NADP⁺, and 0.1 mm Glc-1,6-P₂. The reaction was initiated by the addition of 10 μL of extract (20–30 μg of protein) and 1 unit of Glc-6-PDH (G-5760, Sigma). The reduction of NADP⁺ was measured every 0.5 s for 1 min at 340 nm and at room temperature with a spectrophotometer (Lambda 3B, Perkin-Elmer Cetus). The specific activity of alcohol dehydrogenase in the direction of ethanol oxidation was determined as described by Kelley and Freeling (1984b). Specific activity was determined from reactions that were linear for at least 2 min. One unit of enzyme activity is defined as 1 μmol of NADP⁺ or NAD⁺ reduced min⁻¹. Specific activity is expressed in units per milligram of protein. Bradford reagent (United States Biochemical) was used to determine protein concentration using BSA as the standard.

Protein Gel Electrophoresis

Soluble root proteins (10–30 μg of extract per lane) were fractionated by native (nondenaturing)-PAGE at 4°C, as described by Bailey-Serres et al. (1992). Gels were stained for PGM activity immediately after electrophoresis by incubation in a fresh solution of 0.13 mm NADP⁺, 1 mm nitro blue tetrazolium, 0.4 mm phenazine methosulfate, 100 mm Tris, pH 7.5, 5 mm MgCl₂, 0.3 mm EDTA, 7 mm Na₂-Glc-1-P, and 0.4 unit mL⁻¹ Glc-6-PDH for 30 to 45 min at room temperature. Gel staining was stopped by immersion in 10% (v/v) glacial acetic acid for 5 min, and gels were stored in water prior to drying. Two-dimensional native/SDS-PAGE was carried out as described previously (Sachs et al., 1980). Total protein was visualized by staining with Coomassie blue R-250. Labeled proteins were visualized by exposure of dry gels to Hyperfilm at −70°C with an intensifying screen for [γ-³²P]ATP and at room temperature for [³⁵S]Met.

RESULTS

Maize Cytosolic PGM Is Encoded by Two Nearly Identical Genes

Screening of a maize root cDNA library with a rice EST cDNA that has deduced polypeptide similarity to mammalian PGMs resulted in the isolation of two nearly identical cDNA clones (GenBank accession nos. U89341 and U89342). Both of the putative PGM cDNAs encode an open reading frame of 1752 bp. The overall DNA sequence identity between the coding regions of the two cDNAs was 96.7%. DNA sequence identity was also high in the 5′ (87.0%) and 3′ (88.5%) UTRs, but sequence divergence increased in the 3′ UTRs distal to the stop codon.

Genomic DNA of lines with different wild-type and null alleles of pgm1 and pgm2 were digested with HindIII. Southern-blot hybridization with pgm cDNA probes revealed RFLPs that corresponded to functional and null alleles of each locus. Figure 1 presents a representative Southern blot from this analysis. Fragments of 4.7 and 2.4 kb were detected for Pgm1-5 and Pgm1-9 (labeled as 1-9/1-5) but absent for Pgm1-null, whereas fragments of 11.7, 8.3, and 4.4 kB were detected for Pgm1-null (labeled as 1-n) and absent for the functional alleles of pgm1. Fragments of 11.1 and 1.5 kb were detected for Pgm2-3 and Pgm2-4 (labeled as 2-3/2-4) but absent for Pgm2-null, whereas fragments of 10.4, 3.9, and 1.8 kB were detected for Pgm2-null (labeled 2-n) and absent for the functional alleles of pgm2. RFLPs were not identified that discriminated between the wild-type alleles Pgm1-5 and Pgm1-9 or Pgm2-3 and Pgm2-4. Due to the high level of sequence identity between the two PGM cDNAs, we were unable to generate gene-specific probes from the 3′ UTRs to confidently determine by Southern-blot hybridization which cDNA encodes PGM1 or PGM2. Nevertheless, the detection of RFLPs corresponding to the pgm1 and pgm2 genotype supports the conclusion that the cDNAs we characterized encode the cytosolic isozymes PGM1 and/or PGM2.

Figure 1.

Southern analysis of RFLPs identified by hybridization of a maize pgm cDNA probe to genomic DNA from maize lines with different pgm1 and pgm2 genotypes. Fifty micrograms of genomic DNA from B73 (Pgm1-9; Pgm2-4), 2-null (Pgm1-5; Pgm2-null), and 1-null (Pgm1-null; Pgm2-3) was digested with HindIII, fractionated by electrophoresis on a 1% agarose-TBE gel, transferred to a nylon membrane, probed with a full-length pgm2 cDNA under high-stringency conditions, and exposed to radiographic film. M, The position of migration of molecular mass markers (in kb). RFLPs characteristic of the pgm1 and pgm2 genotype are indicated to the right. Hybridizing fragments present in all genotypes are not labeled.

Both of the pgm cDNAs encoded a deduced protein of 583 amino acid residues with a calculated molecular mass of 63 kD. The deduced PGMs differed at nine residues, only two of which are nonconservative differences (Lys/Thr at residue 131 and Asp/Asn at residue 200; Fig. 2). The calculated pIs of the deduced PGMs were 5.47 and 5.48. It was not unexpected that the pIs of the cytosolic PGMs of B73 (Pgm1-9 and Pgm2-4) were nearly identical. These isozymes comigrate in a native-gel system (described below; see Fig. 5). In a specialized starch gel system the migration of PGM1-9 is only slightly more anodal (toward the positive pole) than that of PGM2-4 (Stuber and Goodman, 1983). On the basis of this information we tentatively identify the cDNA that encodes the PGM of lower pI (5.47) as pgm1 and the cDNA that encodes the PGM of higher pI (5.48) as pgm2.

Figure 2.

Alignment and comparison of the deduced amino acid sequence of PGMs. Z. mays (Zmpgm1 and Zmpgm2), Mesembryanthemum crystallinum (Mcpgm; GenBank accession no. U84888), Homo sapiens (Hspgm; M83088), Saccharomyces cerevisiae (Scpgm; P33401), A. tumefaciens (Atpgm; P39671), D. discoideum (Ddpgm; U61984), E. coli (Ecpgm1: M77127; Ecpgm2: U08369), and plastidic S. oleracea (Sopgm: X75898). The numbers above the alignment correspond to residues of the maize PGMs. Amino acid residues that are conserved in 7 of the 10 sequences are shaded. The predicted catalytic center and the metal-ion-binding site are boxed and labeled. The nine deduced amino acid sequence differences between the two maize cytosolic PGMs are boxed and indicated with an asterisk.

Figure 5.

In vitro phosphorylation of PGM in crude extracts of soluble proteins from aerobic roots. Root extracts were incubated with [γ-³²P]ATP for 6 min at room temperature and fractionated by native-PAGE. Lane 1, Activity gel stained in situ for PGM activity; lanes 2 to 4, proteins visualized by autoradiography. Lane 2, In vitro phosphorylation of crude root extract in extraction buffer. Lane 3, Same as lane 2, with the addition of 0.1 mm Glc-1,6-P₂. Lane 4, Same as lane 2, with the addition of 5 mm MgCl₂, 0.1 mm Glc-1,6-P₂, and 2 units mL⁻¹ hexokinase. The position of PGM in the activity-stained gel and the direction of electrophoresis are indicated.

Comparison of the deduced amino acid sequences of maize PGM1 and PGM2 to that of other prokaryotes and eukaryotes revealed long stretches of identical residues (Fig. 2). Of the sequences compared, the maize PGMs had the highest amino acid sequence identity with the deduced sequences of PGMs of the ice plant (81%) and humans (56%). Regions of high homology included the catalytic reaction center (Thr-Ala-Ser-His-Asp) at residues 122 to 126 in the multiple alignment, which contains the Ser of rat PGM that is phosphorylated during catalysis (Milstein and Sanger, 1961). The catalytic reaction center and metal-ion- binding loop of known animal and fungal PGMs are 100% conserved. On the basis of this identity, the phosphorylation site of PGM1 and PGM2 is most likely Ser-124. The hydrophobic-rich region at residues 300 to 304 in the multiple alignment shows very high identity to the metal-ion- binding loop Asp-Gly-Asp-Gly-Asp identified in rabbit muscle PGM (Dai et al., 1992). Many additional regions of unknown function are highly conserved between PGMs. However the PGMs of maize, ice plant, humans, yeast, Dictyostelium discoideum, and Agrobacterium tumefaciens had considerably lower homology to the PGMs of Escherichia coli and spinach plastids (see Discussion) (Fig. 2).

Analysis of PGM Transcript Accumulation during Development

PGM1 and PGM2 isozymes were detected at similar levels in all maize plant organs and tissues studied, including roots, coleoptiles, leaves of plants grown in light and dark, scutella, and pollen (Stuber and Goodman, 1983). To investigate the developmental expression of these genes at the mRNA level, northern analysis was performed with total RNA from roots, coleoptiles, leaves, pollen, silk, and developing seed by hybridization to the full-length pgm2 cDNA. This probe hybridizes to both pgm1 and pgm2; it was not possible to generate gene-specific probes due to the high sequence identity between the two cDNAs. The RNA blot shown in Figure 3 reveals that an approximately 2-kb PGM transcript accumulated to detectable levels in all samples except silk. Among the organs and tissues studied, PGM transcript accumulation was highest in roots and coleoptiles relative to 18S rRNA levels. PGM mRNA levels remained relatively unchanged throughout embryo development, decreased slightly during endosperm development, and decreased significantly during aleurone development.

Figure 3.

Developmental regulation of PGM mRNA accumulation. A, Total RNA was isolated from various tissues, and 20 μg was fractionated by electrophoresis and hybridized at high stringency to the coding region of a maize cytosolic pgm2 cDNA and 18S rRNA, as described in Methods. B, The hybridization signal in each lane was quantified with a phosphor imager, and the signal in each lane was normalized to the level of 18S rRNA in the same sample. RNA level in roots was given a value of 1.0, and fold increase/decrease over root levels is presented as the relative mRNA level.

PGM Transcript Accumulation Is Reduced during O₂ Deprivation

Increased glycolytic activity during O₂ deprivation of maize seedling roots correlates with increased transcription and translation of genes that encode enzymes involved in Suc breakdown, glycolysis, and ethanolic fermentation (Sachs et al., 1980; Rowland and Strommer, 1986; Russell and Sachs, 1989; Fennoy and Bailey-Serres, 1995; Manjunath and Sachs, 1997). Since PGM catalyzes the formation of Glc-6-P, the first glycolytic intermediate, we analyzed mRNA accumulation in O₂-deprived roots using the full-length pgm2 cDNA as a probe (Fig. 4, A and B). As mentioned earlier, this probe hybridizes to both pgm1 and pgm2. A 1.5-fold increase in PGM mRNA accumulation was observed in the first 2 h of O₂ deprivation, followed by a 4-fold reduction over 24 h relative to aerobic levels. By contrast, adh1 transcript accumulation increased over 40-fold relative to aerobic levels after 24 h of O₂ deprivation in the same samples.

Figure 4.

PGM mRNA accumulation in roots of seedlings deprived of O₂. A, Four- to five-day-old maize seedlings were deprived of O₂ for 0 to 24 h, total RNA was isolated, and PGM transcript levels were analyzed with use of the full-length pgm2 cDNA as described in Figure 3. The blot was sequentially stripped and re-probed with an adh1 cDNA and an 18S rRNA probe. Transcript levels were quantified with a phosphor imager. The corresponding aerobic values were given the value of 1.0, and the increase/decrease over aerobic control is expressed as the relative mRNA level.

Confirmation of Maize PGM as a Phosphoprotein

Cytosolic PGM of a number of organisms, as well as plastidic PGM, can be autophosphorylated in vitro (Ray and Peck, 1972; Salvucci et al., 1990). An in vitro phosphorylation reaction and native-PAGE system was developed to identify autophosphorylated PGM in crude extracts from maize roots. Incubation of the buffered extract with [γ-³²P]ATP resulted in the efficient phosphorylation of two proteins that were resolved by native PAGE. One of the phosphoproteins comigrated with cytosolic PGM activity in the native gel (Fig. 5, compare lanes 1 and 2). The reaction mixture was modified to improve the specificity of the phosphorylation assay. The addition of Glc-1,6-P₂, an activator of PGM activity (Ray and Peck, 1972), resulted in a marginal improvement in the specificity of phosphorylation of the polypeptide that comigrated with PGM activity (Fig. 5, lane 3). The addition of hexokinase and Mg²⁺ to stimulate conversion of Glc to Glc-6-[³²P] significantly increased the specificity of phosphorylation of the protein that comigrated with PGM activity (Fig. 5, lane 4) and reduced the labeling of the more slowly migrating phosphoprotein. These results confirm that cytosolic PGM of maize can be phosphorylated in vitro in crude extracts from roots.

Identification of a 67-kD Phosphoprotein as PGM in Aerobic and O₂-Deprived Roots

Sachs et al. (1980) described the synthesis of 10 major and 10 minor soluble ANPs in anaerobically stressed roots of maize seedlings. The same two-dimensional gel system was used to examine whether PGM is synthesized in roots under aerobic and O₂-deprivation conditions. Seedling roots were labeled in vivo with [³⁵S]Met to examine de novo protein synthesis. PGM was autophosphorylated by incubation of crude root extracts with [γ-³²P]ATP in the presence of MgCl₂, Glc-1,6-P₂, Glc, and hexokinase. Figure 6 shows a comparison of proteins of aerobic and 24-h O₂-deprived roots stained for PGM activity (Fig. 6A), labeled in vivo with [³⁵S]Met (Fig. 6B) and in vitro with [γ-³²P]ATP (Fig. 6, C and D).

Figure 6.

Two-dimensional gel analysis of soluble root protein from aerobic (control) and O₂-deprived seedling roots. Twenty-five micrograms of protein was fractionated on a native-polyacrylamide gel in the first dimension (A and C) and a 9% (w/v) polyacrylamide-SDS gel in the second dimension (B and D). A, Gel stained in situ for PGM activity. B, Proteins of seedling roots labeled for 2 h in vivo with [³⁵S]Met under aerobic conditions or after 22 h of O₂ deprivation. C and D, Proteins phosphorylated in vitro with [γ-³²P]ATP in the presence of 5 mm MgCl₂, 0.1 mm Glc-1,6-P₂, 2 mm Glc, and 2 units/mL hexokinase for 6 min at room temperature. Radiolabeled proteins were visualized by autoradiography. Arrows indicate the positions of PGM, the 67-kD polypeptide, and specific ANPs mentioned in Results.

The in vitro phosphorylation reactions (Fig. 6C) resulted in the efficient labeling of a protein that comigrated with PGM activity in the native gel (Fig. 6A). The phosphorylated PGM had an apparent molecular mass of 67 kD in both aerobic and O₂-deprived roots (Fig. 6D). The unidentified minor, slowly migrating phosphoprotein observed in Figure 5 (lanes 2 and 3) was detected after extended autoradiographic exposure and had an apparent molecular mass of 40 kD (data not shown). Twenty-four hours of O₂ deprivation resulted in a general change in the pattern of protein synthesis, as previously described (Sachs et al., 1980). A [³⁵S]Met-labeled protein that comigrated with PGM activity was observed in the samples from both aerobic and 24-h O₂-deprived roots (Fig. 6B, unlabeled arrow) and corresponded to the 67-kD phosphoprotein (Fig. 6D). There were ANPs of 64 or 65 kD, neither of which had the same electrophoretic mobility of PGM (Fig. 6B; ANP64 and ANP65). The synthesis of the 67-kD polypeptide was reduced over the time course of the O₂ deprivation relative to the synthesis of the Suc-synthase isozymes (Fig. 6B; ANP87). Since the synthesis of PGM in maize roots is reduced in response to O₂ deprivation, it is not an ANP.

PGM Specific Activity in Roots Is Not Significantly Affected by O₂ Deprivation

The specific activity of PGM in crude root extracts of soluble proteins from aerobic control and 24-h O₂-deprived seedlings was examined (Table I). PGM activity in root extracts was measured by quantitation of Glc-1-P conversion to Glc-6-P in a reaction coupled with exogenous Glc-6-PDH and NADP⁺. The assay conditions were optimized for pH and for concentration of Glc-1,6-P₂. The specific activity of PGM increased at least 25-fold by the addition of the cofactor Glc-1,6-P₂ to the reaction mixture (data not shown). An insignificant increase (approximately 30%) in PGM specific activity was observed over 24 h of O₂ deprivation compared with the aerobic control (Table I). A 7.2-fold increase in alcohol dehydrogenase specific activity was measured in the same root extracts.

Table I.

Dissolved O₂ concentration in induction buffer and PGM- and ADH-specific activities in control and O₂-deprived roots

O2 Deprivation	[O2] at 22°C	Specific Activity
		PGM	ADH
h	μM	units mg−1 protein
0 (aerobic)	272	5.00  ± 1.2a (1.0)a	0.20  ± 0.06a (1.0)
6	13	5.39  ± 1.7a (1.1)	0.45  ± 0.12a (2.3)
12	N.D.b	5.52  ± 1.7a (1.1)	0.81  ± 0.20b (4.1)
24	6	6.53  ± 1.6a (1.3)	1.43  ± 0.32c (7.2)

Dissolved O₂ concentration data are from one representative experiment. Specific activity data are means and sd from three experiments. Values followed by a different letter are significantly different (P ≤ 0.01) as determined by the Student's t test.

^aNumbers in parentheses are values relative to aerobic values. 

^bN.D., Not determined. 

DISCUSSION

Plant cells possess cytosolic and plastidic PGM activity. We identified cytosolic PGM as a phosphoprotein with an apparent molecular mass of 67 kD in roots of maize. We isolated two cDNAs of maize that encode PGMs of 98.5% deduced amino acid sequence identity and a calculated molecular mass of 63 kD. The pgm cDNAs most likely encode cytosolic PGM isozymes because: (a) Southern hybridization with a pgm cDNA probe identified RFLPs that correspond to pgm1 and pgm2 genotype; (b) the deduced amino acid sequences of maize PGM1 and PGM2 are 56% identical to PGM of humans (Whitehouse et al., 1992) and only 19% identical to the putative plastidic PGM of spinach (Penger et al., 1994); (c) pgm1 and pgm2 do not encode an N-terminal transit peptide typically found on chloroplast-targeted proteins, including the plastidic PGM of spinach (Heijne and Nishikawa, 1991; Penger et al., 1994); and (d) pgm transcripts and PGM1 and PGM2 isozymes accumulate in pollen (Stuber and Goodman, 1983), a cell type that in maize does not contain plastids.

RFLP analysis indicated that the two cDNAs encode pgm1 and pgm2. The two cDNAs were nearly identical, and gene-specific probes could not be generated to confirm which cDNA corresponded to pgm1 or pgm2. Therefore, we cannot exclude the possibility that PGM may be encoded by duplicate genes at either pgm1 or pgm2 and that we have determined only the gene sequence of one locus. Nonetheless, the Southern-blot data show that the cDNAs we isolated hybridize to both pgm1 and pgm2.

Maize cytosolic PGMs contain many regions that are conserved in human, yeast, D. discoideum, and A. tumefaciens PGMs (Fig. 2). One highly conserved region, Thr-Ala-Ser-His-Asp, was identified as the catalytic center of rat PGM (Milstein and Sanger, 1961) and is located at residues 122 to 126 of maize PGM. The hydroxyl O₂ of Ser-116 of human PGM (most likely Ser-124 of maize PGM) serves as the phosphate acceptor/donor in the catalytic process (Ray and Peck, 1972). A metal-ion-binding loop of PGM, Asp-Gly-Asp-Gly-Asp (Dai et al., 1992), located at residues 300 to 304 of maize PGM, is also highly conserved among these PGMs, except that the second Gly residue is replaced by Ala in plant PGMs. Figure 2 also shows the alignment of maize cytosolic PGMs with two PGMs of E. coli and a putative plastidic PGM of spinach.

The cDNA that encodes the plastidic PGM was isolated by complementation of a yeast hexokinase mutant and encodes a 57-kD protein with a transit peptide sequence (Penger et al., 1994). This PGM has limited sequence identity to the PGMs of maize (19%) and E. coli (19% identity to pgm2 [Lu and Klechner, 1994] and 27% identity to pgm1 [GenBank accession no. M77127]). Despite the low level of overall identity, the multiple alignment analysis shows conservation of the phosphorylated catalytic center and metal-ion-binding site in the spinach plastid and E. coli PGMs, suggesting a related enzymatic activity. Mutation of E. coli pgm2 led to partial blocking of Glc-1-P metabolism (Lu and Klechner, 1994), indicating the presence of an alternative Glc-1-P-utilization pathway. Subsequent analysis revealed that the algC gene encoding the phosphomannomutase of Pseudomonas aeruginosa can complement the pgm2 mutation in E. coli (Lu and Klechner, 1994). On the basis of these observations, pgm1 of E. coli may encode an alternative form of PGM or phosphomannomutase.

Our analysis confirms that cytosolic PGM of maize is distinct from the plastidic PGM reported for spinach. Whether the plastid PGM acts on Glc-P and/or other hexose-P sugars needs to be determined. Distinct structural and biochemical aspects of cytosolic and plastid PGMs are not unexpected, since the gene that encodes the plastid-targeted protein may have evolved from a divergent gene of an endosymbiont of cyanobacterial or proteobacterial origin.

PGM is involved in the trafficking of Glc-P in the synthesis and consumption of Suc. Because of its importance in housekeeping metabolic processes, it is not surprising that cytosolic PGM activity (Stuber and Goodman, 1983) and mRNA accumulation were detected in many plant organs and tissues (Fig. 3). Northern analysis revealed significant PGM mRNA accumulation in roots and coleoptiles. Transcript levels were lower per microgram of rRNA in mature green leaves, pollen, and immature ears. PGM transcripts were undetectable in silk, although a number of transcripts encoding translation factors and ribosomal proteins (S. Manjunath and J. Bailey-Serres, unpublished data; K. Szick and J. Bailey-Serres, unpublished data) have been detected in the same mRNA sample. The reason for the absence or low abundance of PGM mRNA in silk tissue is not known.

PGM mRNA accumulated throughout embryo development, decreased slightly during endosperm development, and decreased considerably during aleurone maturation. The early stages of maize kernel development (0–15 DAP) involve rapid cell division, DNA replication, and cell wall formation (Wilson, 1979). After about 15 DAP the kernel-fill period begins (Ingel et al., 1965), and the activities of many enzymes involved in starch and storage-protein synthesis increase rapidly (Prioul et al., 1990). Doehlert et al. (1994) established that shrunken1, α-zein, aldolase, waxy, shrunken2, and brittle2 transcripts peak at 15 or 30 DAP, then decrease to virtually undetectable levels by 55 DAP. We observed that pgm transcripts were high in embryo and endosperm from 15 to 25 DAP. Consistent with the expression of pgm transcripts in these tissues, Tsai et al. (1970) detected increased PGM activity throughout maize endosperm development.

O₂-deprived roots require accelerated Suc utilization to generate energy from substrate-level phosphorylation. This is accomplished by maintained or increased synthesis of enzymes involved in Suc metabolism, glycolysis, and ethanolic fermentation (for review, see Sachs et al., 1996; Drew, 1997). Therefore, we considered that PGM may also be synthesized in roots deprived of O₂. Northern analyses demonstrated that accumulation of PGM mRNA was transiently induced by 2 h of O₂ deprivation, then decreased over 24 h of flooding. Concomitant with the decrease in PGM mRNA levels, in vivo labeling of PGM with [³⁵S]Met was also lower in 24-h O₂-deprived roots than in aerobic roots. Since PGM synthesis is reduced during O₂ deprivation, we conclude that it is not an ANP. However, despite reduced de novo synthesis, the specific activity of PGM in crude root extracts was not significantly altered by 24 h of O₂ deprivation.

In comparison, Glc-P-isomerase (Kelley and Freeling, 1984a), Fru-1,6-P₂-aldolase (Kelley and Freeling, 1984b), and enolase (Lal et al., 1991) are synthesized at or above aerobic levels (i.e. are ANPs), but show no significant increase in specific activity and/or steady-state levels in O₂-deprived maize roots. However, cytosolic glyceraldehyde-3-P-dehydrogenase activity is maintained during O₂ deprivation due to the induced expression of two of four of the genes that encode this enzyme (Russell and Sachs, 1992; Manjunath and Sachs, 1997). Our results suggest that PGM may be a stable protein in flooded roots. It is possible that the synthesis of PGM is not up-regulated in response to O₂ deprivation because of sufficient enzyme activity to maintain the flux of Glc-6-P into glycolysis. This hypothesis could be tested by measurement of carbon flux through Glc-6-P into glycolysis in maize lines with varying levels of cytosolic PGM.

In summary, maize cytosolic PGM is encoded by two nearly identical genes, located on duplicated chromosomal regions of 1L and 5S (Stuber and Goodman, 1983; Helentjaris et al., 1988), which are transcribed in a number of tissues and organs and in response to O₂ deprivation. The failure to detect tissue-/organ-specific differences in the accumulation of PGM1 and PGM2 isozymes (Stuber and Goodman, 1983) indicates that pgm1 and pgm2 may be functionally redundant. High sequence identity throughout the pgm cDNAs made it impossible for us to determine if accumulation of mRNA encoding the two gene products was differentially regulated. Cytosolic PGM was identified as a 67-kD phosphoprotein that is synthesized at reduced levels in O₂-deprived roots. Nonetheless, PGM-specific activity was not affected by 24 h of O₂ deprivation. Our analysis indicates that the cytosolic PGMs of higher plants are distinct from the reported plastidic PGMs; however, the PGMs of the two subcellular compartments possess the phosphate-transfer catalytic center and metal-ion-binding site that are highly conserved in all known prokaryotic and eukaryotic PGMs and hexose mutases.

Abbreviations:

ANP

anaerobic polypeptide

DAP

days after pollination

EST

expressed sequence tag

Glc-1,6-P₂

Glc-1,6-bisphosphate

Glc-6-PDH

Glc-6-phosphate dehydrogenase

PGM

phosphoglucomutase

RFLP

restriction fragment-length polymorphism

UTR

untranslated region