• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntphysLink to Publisher's site
Plant Physiol. May 1998; 117(1): 33–41.

Molecular Characterization of the Oxalate Oxidase Involved in the Response of Barley to the Powdery Mildew Fungus1


Previously we reported that oxalate oxidase activity increases in extracts of barley (Hordeum vulgare) leaves in response to the powdery mildew fungus (Blumeria [syn. Erysiphe] graminis f.sp. hordei) and proposed this as a source of H2O2 during plant-pathogen interactions. In this paper we show that the N terminus of the major pathogen-response oxalate oxidase has a high degree of sequence identity to previously characterized germin-like oxalate oxidases. Two cDNAs were isolated, pHvOxOa, which represents this major enzyme, and pHvOxOb', representing a closely related enzyme. Our data suggest the presence of only two oxalate oxidase genes in the barley genome, i.e. a gene encoding HvOxOa, which possibly exists in several copies, and a single-copy gene encoding HvOxOb. The use of 3′ end gene-specific probes has allowed us to demonstrate that the HvOxOa transcript accumulates to 6 times the level of the HvOxOb transcript in response to the powdery mildew fungus. The transcripts were detected in both compatible and incompatible interactions with a similar accumulation pattern. The oxalate oxidase is found exclusively in the leaf mesophyll, where it is cell wall located. A model for a signal transduction pathway in which oxalate oxidase plays a central role is proposed for the regulation of the hypersensitive response.

The generation of AOS has been recognized as a significant phenomenon during pathogen-plant interactions (Baker and Orlandi, 1995; Low and Merida, 1996). AOS have been suggested to be involved in plant defense responses in several ways: in cross-linking of lignin and proteins during cell wall modification; in signal transduction leading to gene regulation, hypersensitive cell death, and systemic acquired resistance; and as antimicrobial agents, which inhibit pathogen development directly.

We demonstrated recently that H2O2 accumulates at the sites of contact between epidermal cells undergoing an HR and the subjacent mesophyll cells (Thordal-Christensen et al., 1997). Several possible sources of defense-related AOS have been suggested, one of which is the H2O2-generating oxalate oxidase. Increased activity of this enzyme is found in extracts of barley (Hordeum vulgare) and wheat (Triticum aestivum) leaves following inoculation with the powdery mildew fungus (Blumeria [syn. Erysiphe] graminis; Dumas et al., 1995; Zhang et al., 1995; Hurkman and Tanaka, 1996b). Oxalate oxidases are known from a number of plant, fungal, and bacterial species (Pundir, 1991). Of these, only the wheat and barley enzymes have been classified as germin-like oxalate oxidases (Dumas et al., 1993; Lane et al., 1993).

The germin-like oxalate oxidases are homo-oligomeric, water-soluble, heat-stable, protease-resistant, SDS-tolerant glycoproteins originally known to be expressed in cell walls of cereal embryos at the onset of germination (Lane, 1994). In total, six potentially different oxalate oxidase and oxalate oxidase-like proteins have been characterized in barley, either based on amino acid or cDNA sequence: two salt-response proteins from roots (Hurkman et al., 1991), a root cDNA/protein (Lane et al., 1993), a root cDNA (Hurkman et al., 1994), a seedling protein (Dumas et al., 1993), and a leaf epidermal cDNA (Wei et al., 1998). Only the root protein of Lane et al. (1993) and the seedling protein of Dumas et al. (1993) have been demonstrated to possess oxalate oxidase activity.

The pathogen-response oxalate oxidase in barley expressed in response to the powdery mildew fungus (Dumas et al., 1995; Zhang et al., 1995) appears as two bands on SDS-PAGE as oligomers of approximately 95 and 100 kD. Both are serologically related to the germin-like oxalate oxidase of wheat (Zhang et al., 1995). In the present study, we present the N-terminal sequence from the 100-kD oligomer and its corresponding cDNA. A highly related cDNA is also presented, and these two clones represent the two oxalate oxidase genes apparently present in barley. Both genes express transcripts that accumulate following attack by the powdery mildew fungus. Pathogen-response oxalate oxidase is found to be strictly confined to the leaf mesophyll and may play a role in a signal transduction pathway for regulation of the HR.


Two near-isogenic lines of barley (Hordeum vulgare L. cv Pallas), P-01 and P-02, were grown in a growth chamber under 16 h of light (100 mE s−1 m−2) at 60% RH and 20°C and 8 h of dark at 80% RH and 5°C. Isolates C15 and A6 of the powdery mildew fungus (Blumeria [syn. Erysiphe] graminis f.sp. hordei; Bgh) were used for uniform, high-density (100–200 conidia/mm2) inoculation of the abaxial epidermis of first leaves, as described previously (Thordal-Christensen and Smedegaard-Petersen, 1988). P-01 and P-02 possess the Ml-a1 and Ml-a3 resistance genes to Bgh, respectively (Kølster et al., 1986). P-01 exhibits single-cell HR resistance to C15 and is susceptible to A6. P-02 is susceptible to C15 and exhibits a multicell HR resistance to A6. Irrespective of these outcomes of the different interactions, papilla formation always arrests a high percentage of the fungal conidia at the stage of penetration (Thordal-Christensen and Smedegaard-Petersen, 1988).

Epidermal and mesophyll (including the noninoculated epidermis) tissues were separated by making an incision near the leaf tip, and the abaxial, inoculated epidermis was then stripped off. For whole-leaf samples, approximately 6 cm of the central part of the leaf was collected.

Protein Purification, Preparation of Antibodies, and Amino Acid Sequencing

Oxalate oxidase was extracted and partially purified from infected leaves (P-01/A6, harvested 6 d after inoculation) according to the method of Zhang et al. (1996). Following preparative 6% SDS-PAGE (without prior boiling in reducing agent: oxalate oxidases are SDS tolerant), the gel strip containing the 100- and 95-kD oxalate oxidase proteins was used for immunizing a rabbit according to the method of Harlow and Lane (1988). Fifty micrograms of protein was injected seven times. In addition to recognizing the native 100- and 95-kD oxalate oxidase proteins, the obtained antiserum recognized numerous barley leaf proteins.

To immunopurify a specific oxalate oxidase antibody, the mature oxalate oxidase polypeptide with six His residues fused to the N terminus was expressed in Escherichia coli. This was accomplished by ligating a PCR-amplified fragment of pHvOxOa into the expression vector pQE-9 (Diagen, Hilden, Germany). Expression of the protein was induced, and the protein was purified subsequently on an Ni2+ column under denaturing conditions according to the recommendations of the manufacturer. This purified, denatured monomer was then dotted onto a nitrocellulose membrane. The membrane was dried and used to purify the oxalate oxidase-specific antibody. The membrane was blocked with 4% BSA and then washed with 0.1 m Gly-HCl, pH 2.8, for 5 min. After the membrane was incubated overnight with the antiserum and extensive washing with 1× TBS, purified antibody was eluted from the membrane with 0.1 m Gly-HCl, pH 2.8, for 2 min. The buffer with purified antibody was quickly transferred to 0.03 volume of 1 m Tris-base, thereby adjusting the pH to 7.0. The purified antibody specifically recognized oxalate oxidase. With approximately the same activity, it recognized two bands of the native (Fig. (Fig.7;7; see below) and one band of the denatured oxalate oxidase (data not shown). The native bands were confirmed by activity staining in an identical SDS gel (Fig. (Fig.7).7). Alternatively, the 100-kD oxalate oxidase was blotted onto a PVDF membrane and subjected to amino acid sequencing according to the method of Nielsen et al. (1993).

Figure 7
Tissue localization of the pathogen-response oxalate oxidase in barley leaves (P-01, 24 h after inoculation with Bgh [C15]) demonstrated on 12% SDS-PAGE gels using samples of the whole leaf (lanes 1), epidermal tissue (lanes ...

Plaque, Northern-, and Southern-Blot Hybridization

Plaque lifts from a λ-cDNA library were prepared on nitrocellulose (Millipore) essentially according to the method of Sambrook et al. (1989). Northern blots on Zetaprobe membranes (Bio-Rad) and genomic Southern blots on Hybond N+ membranes (Amersham) were prepared essentially according to the method of Sambrook et al. (1989). Even loading of northern blots was confirmed by methylene blue (0.02%) staining of rRNA on the blot prior to hybridization. 32P-labeled probes of insert DNA were prepared using the Megaprime DNA labeling system (Amersham). PCR fragments for gene-specific probes were synthesized using the KS primer (Stratagene) and 5′-CTTAACTTCCATGAGCCC-3′ on pHvOxOa, and the SK primer (Stratagene) and 5′-AATTCCTGGGAGCCTTC-3′ on pHvOxOb '. 32P-labeled gene-specific probes were subsequently prepared using the Megaprime DNA labeling system with the same two sets of primers replacing the random primers. Hybridizations of plaque lifts and northern blots were performed according to the method of Bryngelsson et al. (1994). Hybridizations of genomic Southern blots were performed in the aqueous solution described by Anderson and Young (1985). See figure legends for wash stringency conditions. Relative hybridization signals were determined by contour densitometry using ImageMaster1D software (Pharmacia).

RNA and DNA Extraction and Molecular Techniques

Total RNA was obtained from frozen leaf material essentially according to the method of Collinge et al. (1987). Barley genomic DNA was isolated according to the method of Ausubel et al. (1987). Selected cDNA clones of a λ-ZAPII library were converted to pBluescriptII SK(−) plasmid clones by the in vivo excision procedure recommended by the manufacturer (Stratagene).

DNA sequencing was performed using the Sequenase DNA sequencing kit (United States Biochemical) according to the manufacturer's recommendations, using [35S]-dATP (Amersham). The reactions were resolved on 8 m urea, 6% polyacrylamide buffer gradient (1–5× TBE) gels. The nucleotide sequences were determined in both orientations using the KS, SK, T3, and T7 primers on double-stranded plasmid DNA of the presented clones and derived subclones, all in the pBluescriptII SK(−) plasmid vector. Computer-assisted analysis of sequence data was performed using DNASIS software (version 5.02, Pharmacia) and the GCG (Genetics Computer Group, Madison, WI) package (Devereux et al., 1984). The sequence databases provided by EMBL and GenBank were searched for related sequences using the GCG package.

Protein Electrophoresis, Immunoblotting, and Activity Assay

Water extracts of frozen tissue powders (2 g fresh tissue/mL) were loaded onto SDS-PAGE in a loading buffer lacking reducing agent and without boiling (Sambrook et al., 1989). Proteins were blotted onto nitrocellulose (Bio-Rad) in a semidry blotter according to the method of Harlow and Lane (1988). Subsequent immunodetection of proteins on the blot was performed according to standard procedures. Alkaline phosphatase-conjugated goat anti-rabbit antibody (Sigma) was used as a secondary antibody. Oxalate oxidase active proteins were revealed by the in-gel assay described by Zhang et al. (1996). Rainbow Mr standards (Amersham) were used.

In Situ Detection of Oxalate Oxidase Activity

Approximately 3 × 5-mm2 leaf specimens of 8-d-old P-02 plants, harvested 24 h after inoculation with C15, were incubated at room temperature in an oxalate oxidase activity developer solution (40 mm succinic acid/NaOH, pH 3.5, 2 mm oxalic acid, 0.5 mg/mL 4-chloro-4-naphthol, and 3.5 mm EDTA) adapted from that of Dumas et al. (1995) for in situ activity detection. Very clear (positive) dark-blue staining appeared in the specimens after 4 h of incubation. Specimens incubated in developer solution lacking oxalic acid were used as activity-staining controls. The stained specimens were fixed in 4% paraformaldehyde in PBS (130 mm NaCl, 7 mm Na2HPO4, and 3 mm NaH2PO4, pH 7.0). After being washed in PBS, the specimens were infiltrated in a series of gelatin solutions (5–20%) in PBS at 40°C and embedded in 20% gelatin. The blocks were frozen to −20°C and stabilized with ice, and 30-μm sections were made by cryostat-sectioning in a rotary retracting microtome (model 5030, Bright, Huntingdon, UK). Sections were examined by light microscopy and photographed.


N-Terminal Amino Acid Sequence

We (Zhang et al., 1995) and Dumas et al. (1995) have showed previously that there is a significant elevation of the oxalate oxidase activity in extracts of barley leaves in response to inoculation with the powdery mildew fungus. The oligomer of this pathogen-response oxalate oxidase appears as activity bands and as immunobands of 95 and 100 kD, respectively, on SDS-PAGE (Zhang et al., 1995; Fig. Fig.7).7). The major 100-kD isoform, which will be designated here as HvOxOa for H. vulgare oxalate oxidase a, was purified and denatured to its monomeric form for N-terminal sequencing. A sequence of 30(28) amino acids was obtained, demonstrating a 93 to 100% identity with published barley oxalate oxidases (Fig. (Fig.1). 1).

Figure 1
N-terminal amino acid sequence of the monomer of the 100-kD oligomeric oxalate oxidase HvOxOa. Alignment to the deduced amino acid sequences of the cDNA clones pHvOxO-Lane (Lane et al., 1993) and pHvOxOb (Hurkman et al., 1994) and to the N-terminal ...

Isolation and Characterization of Oxalate Oxidase cDNA Clones

Two barley root oxalate oxidase cDNA clones have been reported (Lane et al., 1993; Hurkman et al., 1994), and a comparison between our N-terminal amino acid sequence and the polypeptides encoded by these two cDNAs suggests that the pathogen-response oxalate oxidase transcript will be nearly identical to these (Fig. (Fig.1).1). Hybridization of the cDNA pHvOxO-Lane of Lane et al. (1993) to blots of RNA extracted from barley leaves following inoculation with the powdery mildew fungus indicated expression of a pathogen-response oxalate oxidase transcript (Fig. (Fig.2). 2). Therefore, a cDNA library was screened to clone and characterize this pathogen-response oxalate oxidase transcript. A cDNA library (no. 2) of Thordal-Christensen et al. (1992), prepared from barley (P-01) leaf poly(A+) RNA extracted 6 h after inoculation of 10-d-old plants (Thordal-Christensen et al., 1992), was screened using pHvOxO-Lane. The northern blot data in Figure Figure22 suggest that pathogen-response oxalate oxidase cDNAs should be present in this library. One positive clone was identified among approximately 6000 plaques. Partial sequence analysis suggested that this cDNA represents a full-length oxalate oxidase transcript. This clone was used as a probe to screen another 6500 plaques, and two additional (a full-length and a partial) oxalate oxidase cDNA clones were isolated.

Figure 2
Accumulation of oxalate oxidase transcripts in barley leaves inoculated with the powdery mildew fungus. Total RNA was extracted from P-01 and P-02 barley leaves following inoculation with isolate C15 of Bgh (+) and from noninoculated control ...

Complete sequencing demonstrated that the two full-length clones exhibit a high nucleotide identity. They were subsequently named pHvOxOa and pHvOxOb' (see also below). The partial cDNA clone is identical to the 3′ end of pHvOxOa. pHvOxOa is 970 bp and pHvOxOb' is 954 bp in length. Comparison with the reported barley oxalate oxidase cDNA sequences showed that pHvOxOa has a nucleotide identity of 98% with pHvOxO-Lane (Fig. (Fig.3). 3). However, because the cDNA of Lane et al. (1993) covers only the sequence encoding the mature protein (the region between positions 164 and 766 of pHvOxOa), the sequence analysis cannot determine unambiguously that these two cDNAs represent copies of the same gene rather than individual members of the gene family (see also below). pHvOxOb' is identical to the region between positions 28 and 981 of the cDNA of Hurkman et al. (1994; Fig. Fig.3). 3). We will therefore refer to this latter cDNA in the following description, and we propose to name it pHvOxOb. pHvOxOa and pHvOxOb both contain open reading frames of 672 bp, encoding polypeptides of 224 amino acids (Fig. (Fig.3).3). The overall identity between pHvOxOa and pHvOxOb is 84%, the main differences occurring outside the reading frame. In the coding regions, the two transcripts exhibit a nucleotide identity of 90%. In the 5′ UTR, the nucleotide identity is 71%, whereas in the 3′ UTR, it is 70% (Fig. (Fig.3).3). We therefore consider these two clones to represent individual members of the oxalate oxidase gene family (see also below).

Figure 3
Nucleotide sequence of pHvOxOa. Alignment to the oxalate oxidase encoding cDNA clones pHvOxO-Lane (Lane et al., 1993) and pHvOxOb (Hurkman et al., 1994) are shown. Dashes indicate introduced gaps; dots indicate nucleotides also found in pHvOxOa. ...

Amino Acid Sequence Comparison of Barley Oxalate Oxidases

The N-terminal sequence of 30(28) amino acids of HvOxOa purified from the inoculated barley leaves matched perfectly the pHvOxOa-encoded polypeptide from its amino acid no. 24 (Fig. (Fig.4).4). This suggests that the immature protein has a 23-amino acid leader sequence, which causes the mature protein to be synthesized into the ER, and to be potentially exported to the apoplast. This 23-amino acid sequence apparently has the required lipophilic nature. Furthermore, a cleavage site at this position was found in all N-terminal sequence analyses of related barley proteins (Fig. (Fig.4).4). After cleavage, the molecular masses of HvOxOa and HvOxOb, the deduced polypeptides of pHvOxOa and pHvOxOb, were 21.3 and 21.1 kD, respectively. Alignment of the polypeptide encoded by pHvOxOa, the 30(28) amino acids of HvOxOa, and the 52(51) amino acids of the three sequences of Dumas et al. (1993) showed 100% identity. We therefore infer these to be products of the same gene. Within the 52(51) amino acids of the sequences of Dumas et al. (1993), one (the N-terminal) differed in relation to the deduced polypeptide of pHvOxO-Lane and three (including the N-terminal) differed in relation to HvOxOb. This further supports our assumption that the polypeptide of Dumas et al. (1993) and the pHvOxOa-encoded polypeptide are identical. Only 3 of 201 amino acids differed between the polypeptides encoded by pHvOxOa and pHvOxO-Lane, which possibly represent gene copies (Fig. (Fig.4).4). The mature polypeptides, HvOxOa and HvOxOb, have an amino acid identity of 95%.

Figure 4
Amino acid sequence encoded by the pHvOxOa cDNA. Alignment to the deduced amino acid sequences of the cDNA clones pHvOxO-Lane (Lane et al., 1993) and pHvOxOb (Hurkman et al., 1994) and to the amino acid sequences of HvOxO-Dumas (Dumas et al., 1993 ...

Which Oxalate Oxidase Transcript Is Pathogen Responsive?

When either pHvOxOa or pHvOxOb' was used to probe northern blots of RNA samples from inoculated barley leaves, the same accumulation pattern shown in Figure Figure22 was obtained (data not shown). We investigated whether this was the result of cross-hybridization between the highly similar (84%) nucleotide sequences or the presence of two pathogen-response transcripts by using gene-specific probes generated from the relatively divergent 3′ UTRs of pHvOxOa and pHvOxOb' (underlined in Fig. Fig.3).3). These two 3′ probes were hybridized to a set of identical northern blots with the RNA samples also used in Figure Figure22 (Fig. (Fig.5). 5). After parallel hybridization with equally labeled probes, a parallel high-stringency wash (0.1× SSC, 68°C), and parallel exposure and film development, indistinguishable transcript accumulation patterns were resolved. However, the signal obtained with pHvOxOa 3′ probe was generally 6-fold stronger as determined by densitometry. The specificity of the two probes appears sufficient for distinguishing the individual transcripts as evaluated by the almost complete lack of cross-hybridization to the other cDNA clone present on the northern blot (Fig. (Fig.5).5). Note also that hybridizations using the same 3′ probes to genomic Southern blots were gene specific after only a medium-stringency wash (see below). These results strongly suggest that the HvOxOa transcript is pathogen responsive, whereas some doubt remains in relation to the HvOxOb transcript. However, the weak but clear signal with the pHvOxOb' 3′ probe, our isolation of the pHvOxOb' cDNA from this particular library, and the general lack of oxalate oxidase transcript signal in the noninoculated control samples suggest that the HvOxOb transcript is pathogen responsive as well, albeit at a very low level.

Figure 5
HvOxOa and HvOxOb transcripts accumulate in barley leaves inoculated with the powdery mildew fungus. Total RNA samples of Figure Figure22 were applied. The transcripts were detected on northern blots by parallel hybridization with the gene-specific ...

Detailed Accumulation Patterns of the Oxalate Oxidase Transcripts in Different Barley-Powdery Mildew Interactions

The accumulation patterns of the pathogen-response oxalate oxidase transcripts during plant-pathogen interactions were studied in a detailed time-course experiment of barley exhibiting different interaction phenotypes to the powdery mildew fungus (Fig. (Fig.6).6). Barley line P-01 exhibited a single cell HR to isolate C15. P-02 exhibited a multicell HR to isolate A6, whereas it was susceptible to isolate C15. All interactions were characterized by a high rate of papilla formation, which arrests approximately 90 to 95% of all penetration attempts. Northern hybridization was performed using the full-length pHvOxOa. This probe will hybridize to both the HvOxOa and the HvOxOb transcripts; however, the patterns of accumulation appear the same, the only difference being in amount (see above). The hybridization showed accumulation of oxalate oxidase transcripts in the leaves in both compatible and incompatible interactions (Fig. (Fig.6).6). A very small amount of oxalate oxidase transcript appeared in noninoculated control leaves. Transcript accumulation was obvious from 6 h after inoculation in all interactions. The accumulation patterns differed slightly between compatible and incompatible interactions. In the P01/C15 and P02/A6 (both incompatible) interactions, one peak of transcript accumulation was detected at 15 and 24 h after inoculation, respectively. In the P02/C15 (compatible) interaction, there were two peaks. One appeared at 15 h and the other at 96 h. The accumulation profiles agree well with those shown in Figures Figures22 and and5. 5.

Figure 6
Detailed expression pattern of oxalate oxidase transcripts demonstrated on northern blots of total RNA extracted from P-01 and P-02 barley leaves following inoculation with isolates A6 and C15 of Bgh. The transcripts were detected by hybridization ...

Oxalate Oxidase Localization in Leaf Tissue

Thordal-Christensen et al. (1997) detected H2O2 generation in the inoculated barley leaves undergoing HR by the use of a 3,3′-diaminobenzidine-uptake method. H2O2 appeared in the mesophyll tissue under epidermal cells undergoing HR. The staining commenced in a few of the mesophyll cells attached to the epidermal HR cell and then extended to all mesophyll cells under the HR cell. If oxalate oxidase is responsible for the H2O2 generation under epidermal HR cells, the enzyme activity should be detectable in mesophyll tissue of inoculated barley leaves. Protein extracts from different parts of inoculated leaves were studied with an in-gel activity assay and an oxalate oxidase antibody. Figure Figure77 shows that the pathogen-response oxalate oxidase occurred specifically in the mesophyll tissue, not in epidermal tissue. Note that the amount of protein loaded onto each lane represents the same amount of fresh weight tissue and that the epidermal tissue constitutes only approximately one-tenth of the whole leaf. Therefore, the epidermal sample represents 10 times as much leaf area. Mesophyll location has previously been determined at the transcript level (Gregersen et al., 1997).

Cross-sections of leaves stained for oxalate oxidase activity further confirm the mesophyll location of the pathogen-responsive oxalate oxidase. Leaf segments collected 24 h after inoculation gave a only a very faint response in the chloroplasts when oxalic acid was omitted from a developer solution adapted from Dumas et al. (1995; Fig. Fig.8A).8A). On the other hand, a very strong dark-blue staining reaction was observed after incubation in the developer solution containing oxalic acid (Fig. (Fig.8B).8B). That this H2O2-requiring staining reaction is dependent on oxalic acid strongly indicates specificity for oxalate oxidase. The faint response in the chloroplasts is likely to be due to an independent generation of AOS in these organelles. The oxalic acid-dependent staining reaction occurred throughout the entire mesophyll, where it was located at the margin of the cells (Fig. (Fig.8B).8B). A very strong response occurred in the vascular tissue, especially in the phloem tissue and various types of parenchyma (Fig. (Fig.8C).8C). The resolution of the light micrograph was not sufficient to determine whether the epidermal cell walls facing the mesophyll were stained; however, the extract of epidermal strips, which contains this particular cell wall, did not contain oxalate oxidase (Fig. (Fig.7). 7).

Figure 8
Cross-sections of oxalate oxidase activity-stained barley leaves (P-02, 24 h after inoculation with Bgh [C15]). The staining reaction was performed without (A) and with (B and C) oxalate. Images are of epidermis (top)/mesophyll section ...

Genomic Organization of Oxalate Oxidase Genes

Information concerning the barley oxalate oxidase genes was provided by a set of identical genomic Southern blots following hybridization with pHvOxOa and pHvOxOb' (Fig. (Fig.9,9, A and B). Nearly identical and rather simple band patterns were obtained with these two probes following a low- to medium-stringency wash. The specific identity of the bands was easily unravelled by hybridization using the 3′ UTRs (Fig. (Fig.9,9, C and D). This identification is partially confirmed in the relative intensity of the bands obtained after hybridization using the full-length cDNAs. However, pHvOxOa-specific bands are relatively strong in the pHvOxOb' hybridization, which is compatible with the interpretation that the gene for HvOxOa exists in more than one copy. It appears that there are no introns in the open reading frames of the genes for HvOxOa and HvOxOb. This is clear from the result in the HindIII lane, in which a single prominent band of approximately 0.8 kb was found for both of the full-length cDNAs. This band reflected an internal, approximately 0.8-kb HindIII fragment in both cDNAs (Fig. (Fig.3).3). Therefore, the approximately 0.8-kb HindIII band in Figure Figure9,9, A and B, must be at least a double band.

Figure 9
Southern blot of barley (P-01) genomic DNA hybridized with full-length cDNA clones (A and B) and with gene-specific 3′ UTRs (C and D). Low-/medium-stringency washes with 2× SSC at 68°C (A and B), and medium-stringency washes ...

The parts of the genes located outside the HindIII fragments were necessarily accounted for by the very faint bands that were dissimilar for the pHvOxOa and pHvOxOb' probes, i.e. approximately 1.4 and 2.2 kb, respectively, for pHvOxOa and approximately 6.0 kb for pHvOxOb' (only one such band could be detected with this probe). A potential third, less-related gene was represented by the approximately 1.6-kb HindIII fragment in Figure Figure9,9, A and B. This potential third gene was possibly also represented by the unidentified bands in the remaining restriction enzyme lanes (Fig. (Fig.9,9, A and B).


In previous reports of elevated levels of oxalate oxidase activity in barley-powdery mildew interactions (Dumas et al., 1995; Zhang et al., 1995), as well as of oxalate oxidase transcript accumulation induced by B. graminis in wheat (Hurkman and Tanaka, 1996b), no attempts were made to identify the specific pathogen response isoenzyme/transcript species. Here we report a barley cDNA (pHvOxOa) representing a pathogen-response oxalate oxidase transcript. We found that the cDNA represents a pathogen response transcript based on N-terminal amino acid sequencing of the pathogen-response oxalate oxidase, based on northern-blot hybridization with 3′ UTRs and on the fact that pHvOxOa was isolated from a library prepared from plant material following B. graminis inoculation. The closely related HvOxOb transcript is also suggested to be pathogen responsive, albeit at a very low level. The HvOxOb enzyme has not been identified in extracts following inoculation.

Our data suggest the presence of two germin-like oxalate oxidases in barley. In the present study we identified cDNAs for both pHvOxOa and pHvOxOb'. Based on a coding sequence nucleotide identity of 98%, we suggest that a partial cDNA identified by Lane at al. (1993) and pHvOxOa represents copies of the same genes. A cDNA identified by Hurkman et al. (1994) is 100% identical to but longer than pHvOxOb', and therefore we propose to designate that clone pHvOxOb. The overall nucleotide identity between pHvOxOa and pHvOxOb is 84%. The presence of two germin-like oxalate oxidases in barley is also supported by the results of genomic Southern analysis presented here. pHvOxOa and pHvOxOb' clearly hybridize to different restriction fragments, and since no other dominant bands can be detected, we conclude that no other gene members of this close relationship exist in barley. However, more distantly related members of the germin-like oxalate oxidase family do exist.

Hurkman et al. (1991) presented 13(12) amino acid N-terminal sequences of two salt-response root proteins that show 77 and 85% identity to the oxalate oxidases described here (Fig. (Fig.4).4). Whether these root proteins have oxalate oxidase activity is not known. In addition, a multicopy gene that encodes an oxalate oxidase-like protein, HvOxOLP, with only 46% amino acid identity to HvOxOa and HvOxOb, has been identified by Wei et al. (1998). This protein, which seemingly exhibits no oxalate oxidase activity, is also pathogen responsive. HvOxOLP is exclusively expressed in the epidermis, in contrast to HvOxOa. In summary, the germin-like oxalate oxidase family appears to consist of several closely and more distantly related members, and the matter is further complicated by the fact that certain members are encoded by genes of a few to many copies. No amino acid sequence was obtained from the 95-kD pathogen-response oxalate oxidase. However, we speculate that this protein may be HvOxOb. Alternatively, it may represent a posttranslational modification of the 100-kD HvOxOa. This is based on the fact that expression of a single wheat oxalate oxidase gene in tobacco gave rise to a double protein band (Berna and Bernier, 1997).

Oxalate oxidase transcript (this study) as well as enzyme (Zhang et al., 1995) accumulates in both compatible and incompatible interactions. The transcript is detected 6 h after inoculation, whereas the enzyme is detected 15 to 24 h after inoculation. However, while the transcript level declines (particularly rapidly in the incompatible interactions), the enzyme level remains high, which presumably reflects the unusual stability of this enzyme. In this context, it is of interest that the relative level of the oxalate oxidase transcript is very low compared with other pathogen response transcripts. For instance, its level is only approximately 5% of the chitinase transcript (Gregersen et al., 1997).

Oxalate oxidase accumulates in barley seedlings upon onset of germination (Dumas et al., 1993, 1995). The transcript accumulates to high levels in barley roots upon germination and after treatment with NaCl, ABA, methyl salicylate, ABA, and IAA, as demonstrated by hybridization using what we now call pHvOxOb, followed by a high-stringency wash (Hurkman and Tanaka, 1996a). To understand the gene regulation, it is of interest to know the extent to which those signals are due to the HvOxOa and the HvOxOb transcripts.

We have localized the pathogen-response oxalate oxidase as being evenly distributed throughout the mesophyll tissue. In situ activity staining, which suggests that the enzyme is located at the margin of the mesophyll cell, the fact that the immature protein has a leader sequence, and the failure of attempts to wash the enzyme out of the leaf intercellular space (data not shown) suggest that this water-soluble protein is trapped in the mesophyll cell walls. Such a location suggests the enzyme to be a possible source for the H2O2 observed in the zones of attachment between an epidermal cell undergoing HR and subjacent mesophyll cells (Thordal-Christensen et al., 1997).

A potential role of this H2O2 in signaling for HR will require a specific posttranslation activation of oxalate oxidase in the incompatible interaction. Oxalate oxidase has maximum activity at pH 3.2 (Sugiura et al., 1979) and will be inactive at the intercellular pH of 5.5 to 6.0 in a nonstressed leaf. We therefore hypothesize that signaling for HR involves acidification of the apoplast under epidermal cells determined to undergo HR. An increase in the plasma membrane H+-ATPase of such epidermal cells will potentially lead to an activation of the pathogen-response oxalate oxidase in the mesophyll cell wall, partly because of a more suitable pH but also because the substrate, oxalate, otherwise will be inaccessible in salts of divalent cations. Pathogen-elicitor-stimulated plasma membrane H+-ATPase activity has been reported in barley (Wevelsiep et al., 1993; Knogge, 1996) and tomato (Vera-Estrella et al., 1994; Xing et al., 1996). In the tomato system, stimulation was found only in incompatible interactions. As indicated above, a decrease in extracellular pH will not only solubilize oxalate; it will also increase the free Ca2+ concentration (Trewavas and Gilroy, 1991). H2O2 and Ca2+ have been suggested to be involved in signaling for HR (Levine et al., 1996; Price et al., 1996). We are currently attempting to resolve such an array of HR-signaling events in which oxalate oxidase potentially plays a central role.


We are indebted to Professor Jim Dunwell (University of Reading, UK; formerly of Zeneca Seeds, UK) for providing a barley oxalate oxidase cDNA clone.


active oxygen species
hypersensitive response
untranslated region


1This study was supported by the Danish Agricultural and Veterinary Research Council, by the Daloon Foundation, Denmark (F.Z.), and by the Carlsberg Foundation (H.T.-C.).

The nucleotide sequence of pHvOxOa reported in this paper appears in the EMBL nucleotide database under the accession number Y14203.


  • Anderson MLM, Young BD. Quantitative filter hybridisation. In: Hames BD, Higgins SJ, editors. Nucleic Acid Hybridisation—A Practical Approach. Oxford, UK: IRL Press; 1985. pp. 73–111.
  • Ausubel FM, Brent R, Kingston RE, Moore DD, Smith JA, Seidman JG, Struhl K (1987) Current Protocols in Molecular Biology 1987–1988. J Wiley & Sons, New York
  • Baker JC, Orlandi EW. Active oxygen in plant pathogenesis. Annu Rev Phytopathol. 1995;33:299–321. [PubMed]
  • Berna A, Bernier F. Regulated expression of a wheat germin gene in tobacco: oxalate oxidase activity and apoplastic localization of the heterologous protein. Plant Mol Biol. 1997;33:417–429. [PubMed]
  • Bryngelsson T, Sommer-Knudsen J, Gregersen PL, Collinge DB, Ek B, Thordal-Christensen H. Purification, characterization and molecular cloning of basic PR-1-type pathogenesis-related proteins from barley. Mol Plant-Microbe Interact. 1994;7:267–275. [PubMed]
  • Collinge DB, Milligan DE, Dow JM, Scofield G, Daniels MJ. Gene expession in Brassica campestris showing a hypersensitive response to the incompatible pathogen Xanthomonas campestris pv. vitians. Plant Mol Biol. 1987;8:405–414. [PubMed]
  • Devereux J, Haeberli P, Smithies O. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 1984;12:387–395. [PMC free article] [PubMed]
  • Dumas B, Freyssinet G, Pallett KE. Tissue-specific expression of germin-like oxalate oxidase during development and fungal infection of barley seedlings. Plant Physiol. 1995;107:1091–1096. [PMC free article] [PubMed]
  • Dumas B, Sailland A, Cheviet J-P, Freyssinet G, Pallett K. Identification of barley oxalate oxidase as a germin-like protein. CR Acad Sci Paris. 1993;316:793–798. [PubMed]
  • Gregersen PL, Thordal-Christensen H, Förster H, Collinge DB. Differential gene transcript accumulation in barley leaf epidermis and mesophyll in response to attack by Blumeria graminis f.sp. hordei. Physiol Mol Plant Pathol. 1997;51:85–97.
  • Harlow E, Lane D. Antibodies: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1988.
  • Hurkman WJ, Lane BG, Tanaka CK. Nucleotide sequence of a transcript encoding a germin-like protein that is present in salt-stressed barley (Hordeum vulgare L.) roots. Plant Physiol. 1994;104:803–804. [PMC free article] [PubMed]
  • Hurkman WJ, Tanaka CK. Effect of salt stress on germin expression in barley roots. Plant Physiol. 1996a;110:971–977. [PMC free article] [PubMed]
  • Hurkman WJ, Tanaka CK. Germin gene expression is induced in wheat leaves by powdery mildew infection. Plant Physiol. 1996b;111:735–739. [PMC free article] [PubMed]
  • Hurkman WJ, Tao HP, Tanaka CK. Germin-like polypeptides increase in barley roots during salt stress. Plant Physiol. 1991;97:366–374. [PMC free article] [PubMed]
  • Knogge W. Fungal infection of plants. Plant Cell. 1996;8:1711–1722. [PMC free article] [PubMed]
  • Kølster P, Munk L, Stølen O, Løhde J. Near-isogenic barley lines with genes for resistance to powdery mildew. Crop Sci. 1986;26:903–907.
  • Lane BG. Oxalate, germin, and the extracellular matrix of higher plants. FASEB J. 1994;8:294–301. [PubMed]
  • Lane BG, Dunwell JM, Ray JA, Schmitt MR, Cuming AC. Germin, a protein marker of early plant development, is an oxalate oxidase. J Biol Chem. 1993;268:12239–12242. [PubMed]
  • Levine A, Pennell RI, Alvarez ME, Palmer R, Lamb C. Calcium-mediated apoptosis in a plant hypersensitive disease resistance response. Curr Biol. 1996;6:427–437. [PubMed]
  • Low PS, Merida JR. The oxidative burst in plant defense: function and signal transduction. Physiol Plant. 1996;96:533–542.
  • Nielsen KK, Mikkelsen JD, Kragh KM, Bojsen K. An acidic class III chitinase in sugar beet: induction by Cercospora beticola, characterization and expression in transgenic tobacco plants. Mol Plant-Microbe Interact. 1993;6:495–506. [PubMed]
  • Price A, Knight M, Knight H, Cuin T, Tomos D, Ashenden T. Cytosolic calcium and oxidative plant stress. Biochem Soc Trans. 1996;24:497–483. [PubMed]
  • Pundir CS. Purification and properties of oxalate oxidase from Sorghum leaves. Phytochemistry. 1991;30:1065–1067.
  • Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989.
  • Sugiura M, Yamamura H, Hirano K, Sasaki M, Morikawa M, Tsuboi M. Purification and properties of oxalate oxidase from barley seedlings. Chem Pharm Bull. 1979;27:2003–2007.
  • Thordal-Christensen H, Brandt J, Cho BH, Rasmussen SK, Gregersen PL, Smedegaard-Petersen V, Collinge DB. cDNA cloning and characterization of two barley peroxidase transcripts induced differentially by the powdery mildew fungus Erysiphe graminis. Physiol Mol Plant Pathol. 1992;40:395–409.
  • Thordal-Christensen H, Smedegaard-Petersen V. Comparison of resistance-inducing abilities of virulent and avirulent races of Erysiphe graminis f.sp. hordei and a race of Erysiphe graminis f.sp. tritici in barley. Plant Pathol. 1988;37:20–27.
  • Thordal-Christensen H, Zhang Z, Wei Y, Collinge DB. Subcellular localization of H2O2 in plants. H2O2 accumulation in papillae and hypersensitive response during the barley-powdery mildew interaction. Plant J. 1997;11:1187–1194.
  • Trewavas A, Gilroy S. Signal transduction in plant cells. Trends Genet. 1991;7:356–361. [PubMed]
  • Vera-Estrella R, Barkla BJ, Higgins VJ, Blumwald E. Plant defense response to fungal pathogens, activation of host-plasma membrane H+-ATPase by elicitor-induced enzyme dephosphorylation. Plant Physiol. 1994;104:209–215. [PMC free article] [PubMed]
  • Wei Y, Zhang Z, Andersen CH, Schmelzer E, Gregersen PL, Collinge DB, Smedegaard-Petersen V, Thordal-Christensen H. An epidermis/papilla-specific oxalate oxidase-like protein in the defence response of barley attacked by the powdery mildew fungus. Plant Mol Biol. 1998;36:101–112. [PubMed]
  • Wevelsiep L, Rüpping E, Knogge W. Stimulation of barley plasmalemma H+-ATPase by phytotoxic peptides from the fungal pathogen Rhynchosporium secalis. Plant Physiol. 1993;101:297–301. [PMC free article] [PubMed]
  • Xing T, Higgins VJ, Blumwald E. Regulation of plant defense response to fungal pathogens: two types of protein kinases in the reversible phosphorylation of the host plasma membrane H+-ATPase. Plant Cell. 1996;8:555–564. [PMC free article] [PubMed]
  • Zhang Z, Collinge DB, Thordal-Christensen H. Germin-like oxalate oxidase, a H2O2-producing enzyme, accumulates in barley attacked by the powdery mildew fungus. Plant J. 1995;8:139–145.
  • Zhang Z, Yang J, Collinge DB, Thordal-Christensen H. Ethanol increases sensitivity of oxalate oxidase assay and facilitates direct activity staining in SDS gels. Plant Mol Biol Rep. 1996;14:266–272.

Articles from Plant Physiology are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...