Detection of Ca²⁺-Dependent Transglutaminase Activity in Root and Leaf Tissue of Monocotyledonous and Dicotyledonous Plants

Abstract

Protein extracted from root and leaf tissue of the dicotyledonous plants pea (Pisum sativum) and broad bean (Vicia faba) and the monocotyledonous plants wheat (Triticum aestivum) and barley (Hordeum vulgare) were shown to catalyze the incorporation of biotin-labeled cadaverine into microtiter-plate-bound N′,N′-dimethylcasein and the cross-linking of biotin-labeled casein to microtiter-plate-bound casein in a Ca²⁺-dependent manner. The cross-linking of biotinylated casein and the incorporation of biotin-labeled cadaverine into N′,N′-dimethylcasein were time-dependent reactions with a pH optimum of 7.9. Transglutaminase activity was shown to increase over a 2-week growth period in both the roots and leaves of pea. The product of transglutaminase's protein-cross-linking activity, ε-(γ-glutamyl)-lysine isodipeptide, was detected in root and shoot protein from pea, broad bean, wheat, and barley by cation-exchange chromatography. The presence of the isodipeptide was confirmed by reversed-phase chromatography. Hydrolysis of the isodipeptide after cation-exchange chromatography confirmed the presence of glutamate and lysine.

Transglutaminases are Ca²⁺-dependent enzymes that catalyze an acyl-transfer reaction between primary amino groups and protein-bound Gln residues. The γ-carboxamide group of protein-bound Gln is the exclusive acyl-donor substrate of transglutaminases, but a variety of primary amino groups may act as amine donors. These include the ε-amino group of protein-bound Lys residues, which result in the formation of inter- or intramolecular protein cross-links via ε-(γ-glutamyl)-Lys isodipeptide bonds. The covalent cross-links of ε-(γ-glutamyl)-Lys are stable and resist chemical, enzymic, and mechanical disruption (Folk and Finlayson, 1977). Alternatively, the primary amino groups of polyamines may be incorporated into Gln, resulting in covalent posttranslational modification of proteins through the formation of N-(γ-glutamyl)-polyamine bonds. The occurrence of both types of reaction products has been widely reported in animal tissues, and the function of several mammalian transglutaminases has been studied extensively (Folk, 1980; Griffin and Smethurst, 1994).

Factor XIII is the heterotetrameric form of transglutaminase present in human plasma. Proteolytic cleavage by thrombin to the dimeric form activates the enzyme, which then stabilizes the fibrin clot during the final stage of the blood-clotting sequence (Sobel and Gawinowicz, 1996). Keratinocyte transglutaminase present in skin is involved in the terminal differentiation of keratinocytes via cross-linking of a number of structural proteins (Thacher and Rice, 1985). Prostate transglutaminase is responsible for the clotting of rodent seminal plasma (Folk, 1980). Tissue transglutaminase is the most widespread member of the transglutaminase family. Proposed roles for the enzyme include an involvement in programmed cell death (apoptosis) (Fesus et al., 1987; Knight et al., 1991), in exocytosis (Bungay et al., 1986), in stabilization of the extracellular matrix (Aeschlimann et al., 1995), and in transmembrane signal mediation, acting as a G-protein (Gα_h) (Im et al., 1997).

Transglutaminase activity has been reported in higher plants, although no clear role for the enzyme in plant tissue has been defined. Incorporation of radiolabeled polyamines into the animal transglutaminase substrate N′,N′-dimethylcasein by crude cell extracts derived from pea (Pisum sativum) and Jerusalem artichoke has been demonstrated (Icekson and Apelbaum, 1987; Serafini-Fracassini et al., 1988). Radiolabeled polyamines covalently linked to endogenous proteins were detected by autoradiography in extracts prepared from explants of Jerusalem artichoke tubers (Dinella et al., 1992; Grandi et al., 1992). Chloroplasts prepared from Jerusalem artichoke leaf tissue were shown to contain N-(γ-glutamyl)-putrescine, N¹,N⁴-bis-(γ-glutamyl)-putrescine, and N¹,N⁸-bis-(γ-glutamyl)-spermidine.

The isolation of these polyamine conjugates provides unequivocal proof of a catalytically active transglutaminase present in Jerusalem artichoke chloroplasts (Del Duca et al., 1995). The large subunit of Rubisco has been demonstrated as a substrate in Jerusalem artichoke and alfalfa, suggesting a possible role for transglutaminase in photosynthesis (Margosiak et al., 1990; Del Duca et al., 1994). Actin and tubulin have been identified as substrates in apple pollen (Del Duca et al., 1997). More recently, a Ca²⁺-independent enzyme with transglutaminase-like activity has been purified from leaves of soybean using a [¹⁴C]putrescine-incorporation assay (Kang and Cho, 1996).

Transglutaminases in mammalian systems are Ca²⁺-dependent enzymes (Folk, 1980). However, in the intracellular environment this Ca²⁺ activation is regulated by the binding of GTP and ATP (Smethurst and Griffin, 1996). Recent findings suggest that Ca²⁺ ions stimulate the activity of plant transglutaminase but are not an absolute requirement (for review, see Serafini-Fracassini et al., 1995). Other investigators have shown Ca²⁺ ions to have an inhibitory effect at concentrations greater than 2 mm (Aribaud et al., 1995; Kang and Cho, 1996). The assays used to demonstrate Ca²⁺-independent amine incorporation involve the incorporation of radiolabeled polyamines into protein substrates such as N′,N′-dimethylcasein and Rubisco. Transglutaminase assays of this type may not be appropriate for screening crude plant extracts because of the possibility of interference by enzymes such as diamine oxidases. Diamine oxidases are able to incorporate [¹⁴C]putrescine into N′,N′-dimethylcasein in a Ca²⁺-independent reaction via Schiff base formation (Siepaio and Meunier, 1995).

During this study the root and leaf tissue of wheat (Triticum aestivum), barley (Hordeum vulgare), pea, and broad bean (Vicia faba) were screened for transglutaminase activity using the conventional [¹⁴C]putrescine-incorporation assay (Lorand et al., 1972) and two microtiter-plate-based assays (Slaughter et al., 1992; Lilley et al., 1997). In agreement with the findings of other studies, no absolute Ca²⁺-ion dependence was demonstrated using the [¹⁴C]putrescine-incorporation assay. However, Ca²⁺-ion-dependent transglutaminase activity was observed in all tissues using a biotin-labeled casein-cross-linking assay (Lilley et al., 1997) and a biotin-labeled polyamine-incorporation assay (Slaughter et al., 1992). Polyamine incorporation was found to be activated by nanomolar concentrations of Ca²⁺, whereas millimolar concentrations were required to activate protein cross-linking. Transglutaminase activity was shown to increase over a 2-week growth period in both the root and leaves of pea.

The definitive evidence for the presence of transglutaminase within a biological system is the presence of the ε-(γ-glutamyl)-Lys isodipeptide (Folk and Finlayson, 1977). We present the first evidence to our knowledge of ε-(γ-glutamyl)-Lys isodipeptide in plant tissue and show that the isodipeptide is present in the root and shoot tissue of the dicotyledonous plants pea and broad bean and the monocotyledonous plants wheat and barley.

MATERIALS AND METHODS

Treatment of Plant Material

Seeds of broad bean (Vicia faba var Aquadulce), pea (Pisum sativum var Feltham First), barley (Hordeum vulgare var Pipkin), and wheat (Triticum aestivum var Apollo) were soaked overnight in running water and then germinated in damp vermiculite in a greenhouse at 20°C. PPFD of 300 to 400 μmol m⁻² s⁻¹ was provided by natural daylight supplemented with high-pressure sodium lamps and a 16-h photoperiod. Root and leaf tissue were harvested after 14 d unless stated otherwise. Fifty grams of tissue was homogenized in a Waring blender in ice-cold 50 mm Tris-HCl, pH 7.4, containing 250 mm Suc, 10 mm EDTA, 10 mm 2-mercaptoethanol, 5 μm leupeptin, 1 μm pepstatin, 1 mm PMSF, and 5% (w/v) polyvinylpolypyrrolidone. The homogenate was filtered through two layers of muslin, and the pH was readjusted to 7.4 using solid Tris. The extract was centrifuged at 13,000g for 20 min at 4°C. The supernatant was clarified further by centrifugation at 80,000g for 45 min at 4°C to sediment the membrane fraction. The supernatant protein was precipitated by the addition of solid (NH₄)₂SO₄ to 90% saturation at 4°C. Precipitated protein was collected by centrifugation at 13,000g for 20 min at 4°C, redissolved in 50 mm Tris-HCl, pH 7.4, containing 1 mm 2-mercaptoethanol, and dialyzed against 2.5 L of the same buffer at 4°C. Aliquots of dialyzed protein were stored at −20°C.

Proteolytic Digestion of the Plant Proteins

Twenty-five milligrams of plant proteins extracted from leaf or root tissue was precipitated from solution by the addition of TCA to a final concentration of 10% (w/v). The precipitate was collected by centrifugation in a microcentrifuge for 5 min at 4°C (approximately 10,000g) and then washed twice in 10% (w/v) TCA, twice in diethyl ether:ethanol (1:1, v/v), and twice in diethyl ether. The pellet was dried and resuspended in 1.0 mL of 0.1 m (NH₄)₂CO₃, pH 8.0 (a crystal of thymol was included to inhibit bacterial growth). The digestion of plant proteins was carried out by the sequential addition of proteolytic enzymes (subtilisin, pronase, prolidase, Leu aminopeptidase, and carboxypeptidase Y) according to the method of Griffin et al. (1982). The incubation time with each of the proteolytic enzymes was 24 h at 37°C. The digests were mixed with 3.6 mL of chloroform:methanol:HCl (200:100:2, v/v) and centrifuged (model GPKR, Beckman) at 2,500g (approximately 3,000 rpm) for 5 min. The aqueous phase was separated from the organic phase and both were dried using a centrifugal evaporator (model RC 10.22, Jouan, Winchester, VA). The aqueous phase containing amino acids and the isodipeptide was resuspended in 1.0 mL of distilled water. The material in the dried organic layer was assayed for protein to determine the percentage of hydrolysis.

Purification of the Isodipeptide Using Anion-Exchange Chromatography

Three-hundred microliters of the protein digests (pH adjusted to 12.6) was applied to a Dowex (2 × 8–200; Cl form) anion-exchange column (1.5 × 1.1 cm) equilibrated with ultrapure water (Milli-Q system, Millipore) (>18 MΩ), pH 12.6, and washed with 25 mm NH₄HCO₃, pH 7.6, at a flow rate of 4.0 mL min⁻¹. After 70.0 mL had passed through the column to elute the majority of the Leu, the column was washed with 10.0 mL of ultrapure water, pH 12.6. The isodipeptide was eluted with 0.1 n HCl. Five-milliliter fractions were collected and their pH monitored. When the pH decreased to less than 7.0, the next 20.0 mL of eluate was pooled, the pH was adjusted to 7.0, and the eluate was freeze-dried. After freeze-drying the samples were redissolved in 0.3 mL of ultrapure water and stored at −20°C. After anion-exchange chromatography approximately 70% of the isodipeptide was recovered and 95% of the Leu was removed.

ε-(γ-Glutamyl)-Lys Isodipeptide Analysis

A sample of either crude plant protein digest or anion-exchange-purified material was applied to an amino acid analyzer (Alpha Plus 4151, LKB, Bromma, Sweden) with a 5 × 250 mm ion-exchange column (Ultrapac 8, LKB; particle size 8 ± 0.5 μm; lithium form). The amino acids and isodipeptide were eluted with lithium-citrate buffers and detected using postcolumn orthophthalaldehyde-2-mercaptoethanol derivatization as described previously (Griffin et al., 1988) but with the following modifications: the column temperature was 21°C and the pH of the elution buffer (c) was adjusted to 3.15. A 2.0-mL postcolumn reaction loop was incorporated between the mixing loop and the fluorescence detector (model LS1, Perkin-Elmer; 360 nm excitation, 450 nm emission). The data were recorded on a computer (model SL1, Viglen, UK) using a series interface (Analytical 900, Nelson, Cupertino, CA) and chromatography software (model 2600 V5, Nelson).

Hydrolysis of Ion-Exchange-Purified Sample

Fifty microliters of anion-exchange-purified wheat root sample was applied to the amino acid analyzer and eluted (profile described above) without postcolumn derivatization. A 5.0-mL fraction was collected at the elution point of the isodipeptide, the pH was adjusted to 7.0, and the sample was dried on a centrifugal evaporator (Jouan) and redissolved in 0.3 mL of ultrapure water, pH 8.0. One-hundred microliters of sample in 6.0 n HCl was sealed in a glass tube under N₂ and incubated overnight at 120°C. The samples were dried using the centrifugal evaporator and redissolved in 0.1 mL of ultrapure water. The pH was adjusted to 8.0 with 8.4 n KOH and the volume was adjusted to 0.2 mL. Pre- and posthydrolysis reversed-phase analyses were conducted with the appropriate standard additions.

HPLC Analysis

An autosampler (model 507, Beckman) mixed 25 μL of sample with an equal volume of orthophthalaldehyde-2-mercaptoethanol derivatizing reagent (Sigma) and injected a fixed 50 μL onto a reversed-phase column (4.6 × 15.0 cm, particle size 5.0 μm; Ultrasphere ODS, Beckman) using an HPLC system (Beckman Gold) and a fluorescence detector (340 nm excitation, 450 nm emission; model 167, Beckman). The isodipeptide was eluted at a flow rate of 2.0 mL min⁻¹ with a gradient of 60 mm CH₃CO₂K, pH 5.9, and methanol. The presence of the isodipeptide was confirmed by coelution with an authentic standard.

Protein Assay

The protein content of crude plant extracts was determined using a modified bicinchoninic acid method (Brown et al., 1989). BSA was used as the standard protein.

¹⁴C-Labeled Putrescine-Incorporation Assay

The method used was a modification of that described by Lorand et al. (1972). The assay was carried out at 37°C in 100 μL of 77.5 mm Tris-HCl containing 5 mm CaCl₂, 10 mm DTT, 5 mg mL⁻¹ N′,N′-dimethylcasein, 1.2 mm [1,4-¹⁴C]putrescine (specific activity 3.97 μCi μmol⁻¹), and 40 to 770 μg of plant protein. The reaction pH was 7.8 at 37°C. Putrescine incorporation was terminated after 60 min by pipetting 10-μL aliquots of the reaction mixture onto 1-cm² Whatman no. 1 filter paper squares presoaked in 1% (w/v) methylamine and 100 mm EDTA. Protein was precipitated by washing the filter papers once in ice-cold 10% (w/v) TCA, three times in ice-cold 5% (w/v) TCA, once in acetone:ethanol (1:1, v/v), and once in acetone. The filter papers were then dried and placed into 2.0 mL of liquid scintillant and counted for 5 min in a liquid scintillation counter (model 300C, Packard Instruments, Meriden, CT). One unit of transglutaminase activity was defined as 1 nmol of putrescine incorporated into N′,N′-dimethylcasein per hour.

Biotin-Labeled Cadaverine-Incorporation Assay

The assay was carried out according to the method of Slaughter et al. (1992) with the following modifications. Microtiter-plate wells were blocked with 3% (w/v) BSA in 0.1 m Tris-HCl, pH 8.5. The incubation time for the transglutaminase reaction was 60 min and the reaction pH was 7.9. Biotin cadaverine was replaced with 5-({[N-(biotinoyl)amino}hexanoyl]amino)pentylamine (biotin-X-cadaverine). Streptavidin-alkaline phosphatase (0.25 mg mL⁻¹ [1:150]) was replaced with extravidin peroxidase (2.0 mg mL⁻¹ [1:5000]). Phosphatase substrate was replaced with 100 mm NaC₂H₃O₂, pH 6.0, containing 0.310 mm 3,3′,5,5′-tetramethyl benzidine and 0.004% (v/v) H₂O₂. Color development was terminated by the addition of 50 μL per well of 5.0 m H₂SO₄. The A₄₅₀ was read using a multiscan ELISA spectrophotometer (Titertek, Flow Laboratories, McLean, VA). One unit of transglutaminase activity was defined as a change in A₄₅₀ of 1.0 per hour.

Biotin-Labeled Casein-Cross-Linking Assay

The ε-(γ-glutamyl)-Lys-formation assay was carried out according to the method of Lilley et al. (1997). One unit of transglutaminase activity was defined as a change in A₄₅₀ of 1.0 per hour.

RESULTS AND DISCUSSION

There is a mounting body of evidence to support the presence of transglutaminase in plant tissue. However, this is the first report to show a Ca²⁺-dependent transglutaminase and the presence of ε-(γ-glutamyl)-Lys isodipeptide bonds in plant tissue, the soundest evidence for transglutaminase activity.

The ability of crude plant extracts to incorporate radiolabeled polyamines into proteins such as N′,N′-dimethylcasein has been demonstrated (Icekson et al., 1987; Serafini-Fracassini et al., 1988; Margosiak et al., 1990; Aribaud et al., 1995). Table I shows that the [¹⁴C]putrescine-incorporation assay detected transglutaminase activity in only three of the eight tissues screened. The Ca²⁺-chelating agents EDTA and EGTA at 5 mm were unable to effect more than 35% inhibition of [¹⁴C]putrescine-incorporation activity of extracts. Similar results have been shown by other investigators using comparable assays, leading to the proposal that plant transglutaminase has no absolute Ca²⁺-ion requirement (Icekson et al., 1987; Serafini-Fracassini et al., 1988, 1995). Recent research (Siepaio and Meunier, 1995) indicated the presence of a contaminating diamine oxidase in crude plant extracts that is able to incorporate [¹⁴C]putrescine into N′,N′-dimethylcasein in a Ca²⁺-independent manner, masking the transglutaminase activity.

Table I.

Specific activity of transglutaminase in four plant species screened using three assay systems

Tissue Type	Polyamine-Incorporation Assays		Casein-Cross-Linking Assay
	[14C]Putrescine assay	Biotin-cadaverine assay
	units mg−1
Pea root	1.84  ± 0.95	1.21  ± 0.10	1.02  ± 0.08
Pea leaf	2.57  ± 0.31	0.29  ± 0.02	0.26  ± 0.02
Broad bean root	N.D.a	0.20  ± 0.02	0.42  ± 0.01
Broad bean leaf	0.11  ± 0.01	0.07  ± 0.003	0.16  ± 0.03
Wheat root	N.D.	2.74  ± 0.15	2.18  ± 0.23
Wheat leaf	N.D.	0.18  ± 0.002	0.16  ± 0.01
Barley root	N.D.	1.78  ± 0.11	0.71  ± 0.12
Barley leaf	N.D.	0.29  ± 0.01	0.32  ± 0.02

Crude extracts were incubated at 37°C for 60 min (n = 4). Extract boiled for 20 min was used as a negative control. Values are means ± se.

^aN.D., Not detected. 

In contrast, Table I shows that soluble transglutaminase activity was detected in all of the extracts screened using both the biotin-cadaverine-incorporation assay (Slaughter et al., 1992) and the casein-cross-linking assay (Lilley et al., 1997). Contaminating activities do not appear to interfere with the biotin-cadaverine-incorporation or casein-cross-linking assays, since chelation of Ca²⁺ by 1 mm EDTA and 1 mm EGTA resulted in more than 80% inhibition of pea root and leaf transglutaminase activity and 100% inhibition of all of the other extracts screened (data not shown). These data suggest that both plate assays are more suitable for the study of transglutaminase from crude plant cell extracts than the conventional [¹⁴C]putrescine-incorporation assay and confirm the Ca²⁺ dependency of the transglutaminase found in the plant tissue studied.

Table I shows that 14-d-old root tissue exhibited a higher specific activity in all species than leaf tissue of the same age. Using both assays the greatest specific activity was observed in wheat root. In pea root extract both the biotin-cadaverine-incorporation and casein-cross-linking reactions were time dependent and linear up to 60 min (Fig. 1); therefore, this time period was used for all subsequent experiments.

Figure 1.

Time-dependence curves for soluble pea root transglutaminase extracted from 14-d-old tissue. •, Biotinylated casein cross-linking; ○, biotin-cadaverine incorporation. Data points represent means ± se of four replicates. CaCl₂ (5 mm) was replaced by 250 μm EDTA in the negative control reaction buffer.

Resting levels of plant cytosolic Ca²⁺ have been found to be in the nanomolar range, with brief increases to micromolar levels in response to the appropriate stimulation. Ca²⁺ levels in the extracellular matrix and in Ca²⁺ stores have been detected in the millimolar range (Bush, 1995). Figure 2a shows that activation of the biotin-cadaverine-incorporation activity of soluble pea root transglutaminase by Ca²⁺ was observed at levels of 20 nm with an apparent K_m for Ca²⁺ of 50 nm. Maximum activity was achieved at 94 nm free Ca²⁺, suggesting that soluble pea root transglutaminase is able to incorporate polyamines into proteins at resting levels of cytosolic Ca²⁺. Figure 2b shows that activation of the protein-cross-linking function of soluble pea root transglutaminase occurs at 250 μm free Ca²⁺ and peaks at 3 mm, with an apparent K_m for Ca²⁺ of 2 mm. This observation may also suggest that to carry out the protein-cross-linking reaction, soluble pea root transglutaminase must be in a high-Ca²⁺ environment, such as the extracellular matrix, or in the intracellular environment when Ca²⁺ stores are released through cellular damage.

Figure 2.

Ca²⁺-activation curve for biotin-cadaverine incorporation (a) and casein cross-linking (b) by soluble pea root transglutaminase extracted from 14-d-old tissue. •, Biotinylated casein cross-linking; ○, biotin-cadaverine incorporation. Zero free Ca²⁺ was achieved by the addition of 1 mm EGTA to the reaction buffer. The amount of CaCl₂ required to give the desired free Ca²⁺-ion concentrations at pH 7.9 and 37°C was then calculated using a computer program (Fuhr et al., 1993). Data points represent means ± se of four replicates.

Figure 3 shows that the optimum pH for both casein-cross-linking and biotin-cadaverine-incorporation activity of soluble pea root transglutaminase was 7.9. The profiles of both pH plots are similar, suggesting that both assays are measuring the same enzymic activity. Other investigators have demonstrated pH optima between 7.9 and 8.4 for transglutaminase-like activities in different tissues of Jerusalem artichoke (Falcone et al., 1993).

Figure 3.

pH-dependence curve for soluble pea root transglutaminase extracted from 14-d-old tissue. •, Biotinylated casein cross-linking; ○, biotin-cadaverine incorporation. The pH values were measured at 37°C after the addition of pea root extract. Biotin-cadaverine incorporation was measured in the presence of 100 μm CaCl_2. Casein cross-linking was measured in the presence of 5 mm CaCl_2. Data points represent means ± se of four replicates. EDTA (250 μm) was used as a negative control.

Mammalian transglutaminases have a Cys residue at the active site and are irreversibly inhibited by reagents such as iodoacetamide (Smethurst and Griffin, 1996). In this study the biotin-cadaverine-incorporation activities of soluble pea root and leaf extracts were inhibited by 32% and 24%, respectively, by 10 mm iodoacetamide, suggesting that the active site of plant transglutaminase may be similar but not identical to the active site of mammalian tissue transglutaminase, since a greater inhibition may have been expected at this concentration of iodoacetamide. Furthermore, the activity of mammalian tissue transglutaminase is regulated by GTP at low concentrations of Ca²⁺ (Smethurst and Griffin, 1996). The biotin-cadaverine-incorporation activity of pea root transglutaminase was not inhibited by 1 mm GTP at 1 μm free Ca²⁺, indicating that in this respect plant transglutaminase may be different from the mammalian tissue transglutaminase (data not shown).

A relationship exists between transglutaminase activity and the age of pea root tissue. Figure 4a shows that transglutaminase activity increased during the first 18 d of growth, using both the biotin-cadaverine-incorporation and the casein-cross-linking assays. This was followed by a decrease in activity in 22- to 32-d-old root tissue, indicating that transglutaminase may be involved in early root growth and development. This suggestion is supported by a similar observed increase and decrease in transglutaminase activity in developing roots of Chrysanthemum morifolium (Aribaud et al., 1995). In pea leaf tissue (Fig. 4b) both casein-cross-linking activity and biotin-cadaverine-incorporation activity increased to a peak at d 25 and 15, respectively. Activity did not decline rapidly, as in pea root, but remained at a level above that detected at d 8. The profiles for the casein-cross-linking activity and biotin-cadaverine-incorporation activity in root and shoot tissue during the 30-d period assayed were similar, indicating, like the pH profile (Fig. 3), that one enzyme is capable of the two activities. Transglutaminase has been detected in both developing and mature leaf tissue, and this observation may support other reports proposing roles for transglutaminase in photosynthesis (Margosiak et al., 1990; Del Duca et al., 1994, 1995).

Figure 4.

Variation in soluble transglutaminase activity with age of developing pea root (a) and shoot (b) tissue. •, Biotinylated casein cross-linking; ○, biotin-cadaverine incorporation. EDTA (250 μm) replaced 5 mm CaCl₂ in the negative controls. Data points represent means ± se of four replicates.

Despite reports of the presence of transglutaminase in plants (Serafini-Fracassini et al., 1995; Kang and Cho, 1996; Del Duca et al., 1997), there have been no previous reports of ε-(γ-glutamyl)-Lys in plant tissue, which is the definitive proof of the presence of the enzyme (Folk and Finlayson, 1977). This may indicate that the isodipeptide is not present, suggesting that the reports of transglutaminase activity in plants are indeed reports of other types of enzymes such as diamine oxidases (Siepaio and Meunier, 1995). Alternatively, the levels of the cross-link in plant tissue may be low, making detection difficult.

This work shows that the low levels of ε-(γ-glutamyl)-Lys isodipeptide in plant tissue can be detected if an additional purification step is undertaken before cation-exchange chromatography. In addition, application of a concentrated crude digest of plant protein to the amino acid analyzer results in the Leu and the Tyr peaks masking the ε-(γ-glutamyl)-Lys peak. To overcome this problem, the resolution between Tyr and Leu was maximized by modification of the pH of the elution buffer and the temperature of the column. Figure 5a shows that levels of ε-(γ-glutamyl)-Lys isodipeptide in plant tissue were low. Figure 5b shows that detectable levels of the isodipeptide could be observed using the amino acid analyzer after the sample had been purified using anion-exchange chromatography.

Figure 5.

Amino acid profile from digested wheat root protein using cation-exchange chromatography. Protein extracted from wheat root tissue was digested with proteolytic enzymes. a, Ten microliters of the digest was applied to an amino acid analyzer. b, Fifty microliters of anion-exchange-purified material was applied to an amino acid analyzer. The amino acids and isodipeptide (solid lines) were eluted by increasing the salt concentration, and the position of the isodipeptide was confirmed by a run containing 1.0 nmol of authentic standard plus the wheat root samples (dashed lines).

To confirm the presence of ε-(γ-glutamyl)-Lys, the peak from the amino acid analyzer was collected and subjected to reversed-phase chromatography. Figure 6a shows the profile of the isodipeptide peak analyzed by reversed-phase chromatography (37 pmol). The presence of the isodipeptide was confirmed by coelution with authentic standard. In addition, the underivatized sample from the amino acid analyzer was hydrolyzed overnight in 6.0 n HCl at 120°C. Figure 6b shows the changes that occurred in the wheat root sample hydrolyzed by HCl; the isodipeptide peak (18.5 pmol applied to the column) decreased and this was mirrored by increases in glutamate (17 pmol) and Lys (16.9 pmol). To confirm the presence of glutamate and Lys, Figure 6b shows the hydrolyzed wheat root sample with additions of authentic glutamate, Lys, and ε-(γ-glutamyl)-Lys isodipeptide.

Figure 6.

Reversed-phase chromatography profile from digested wheat root protein after anion-exchange and cation-exchange chromatography. Protein extracted from wheat root tissue was digested with proteolytic enzymes. The digest was purified using anion-exchange resin and applied to an amino acid analyzer. The underivatized ε-(γ-glutamyl)-Lys isodipeptide peak was collected, concentrated, and hydrolyzed in 6.0 n HCl. Twenty-five microliters of isodipeptide before (a) and after (b) hydrolysis was applied to a reversed-phase column. The isodipeptide (solid lines) was eluted by increasing the methanol concentration (dotted/dashed lines), and the position of the isodipeptide (dashed lines) was confirmed by a run containing 100 pmol of authentic isodipeptide plus the wheat root isodipeptide sample or 50 pmol of authentic isodipeptide, 25 pmol of Lys, and 25 pmol of glutamate plus the hydrolyzed wheat root isodipeptide sample.

Table II shows the ε-(γ-glutamyl)-Lys isodipeptide present in 14-d-old leaf and root protein of the dicotyledonous plants pea and broad bean and the monocotyledonous plants wheat and barley, with more cross-linking in the proteins extracted from root tissue. The isodipeptide content follows the pattern of extractable transglutaminase shown in Table I. The level of ε-(γ-glutamyl)-Lys isodipeptide present in plant protein is approximately 3% of that found in clotted fibrin (Griffin and Wilson, 1984). In pea root protein the levels of ε-(γ-glutamyl)-Lys isodipeptide during a 32-d period varied between 0.15 and 0.55 nmol mg⁻¹ (data not presented), without any pattern being apparent.

Table II.

ɛ-(γ-Glutamyl)-Lys isodipeptide content of digested plant proteins analyzed using cation-exchange chromatography (n = 3)

Sample	ɛ-(γ-Glutamyl)-Lys
	nmol mg−1 hydrolyzed protein
Pea leaf	0.44  ± 0.01
Pea root	1.03  ± 0.03
Broad bean leaf	0.42  ± 0.01
Broad bean root	0.45  ± 0.02
Wheat leaf	0.52  ± 0.05
Wheat root	1.30  ± 0.07
Barley leaf	0.33  ± 0.06
Barley root	0.54  ± 0.03

Values are means ± se.

This is the first report, to our knowledge, to show the presence of ε-(γ-glutamyl)-Lys isodipeptide and a Ca²⁺-dependent transglutaminase in plant tissue. The protein-cross-linking function of the plant transglutaminase is activated by millimolar concentrations of Ca²⁺, suggesting a possible extracellular protein-cross-linking role or a role in cell death similar to that of the mammalian tissue transglutaminase when Ca²⁺ levels are likely to increase within the cell (Fesus et al., 1987). In contrast, the polyamine-incorporation function is activated by nanomolar concentrations of Ca²⁺, suggesting that this is an intracellular function of plant transglutaminase. Furthermore, polyamine incorporation is not inhibited by 1 mm GTP at low free-Ca²⁺ concentrations, indicating that in plants the enzyme may not be coupled to transmembrane signal transduction and that substrate availability may regulate intracellular plant transglutaminase activity.

The activity of plant transglutaminase is altered during the growth of pea root and leaf tissue, suggesting an involvement in developmental processes, possibly including cell wall development, root-tip development, organelle development, or cellular differentiation. In mammalian species the presence of ε-(γ-glutamyl)-Lys cross-links in proteins gives rise to protein stability (Folk, 1980; Smethurst and Griffin, 1994), so it is not unreasonable to assume that similar cross-links fulfill the same role in plant tissue. Further work will be required to characterize the enzyme and to identify cellular protein substrates to help establish a role for transglutaminase in plant tissue.