Identification and Partial Characterization of the Pectin Methyltransferase “Homogalacturonan-Methyltransferase” from Membranes of Tobacco Cell Suspensions

Abstract

A membrane preparation from tobacco (Nicotiana tabacum L.) cells contains at least one enzyme that is capable of transferring the methyl group from S-adenosyl-methionine (SAM) to the C6 carboxyl of homogalacturonan present in the membranes. This enzyme is named homogalacturonan-methyltransferase (HGA-MT) to distinguish it from methyltransferases that catalyze methyletherification of the pectic polysaccharides rhamnogalacturonan I or rhamnogalacturonan II. A trichloroacetic acid precipitation assay was used to measure HGA-MT activity, because published procedures to recover pectic polysaccharides via ethanol or chloroform:methanol precipitation lead to high and variable background radioactivity in the product pellet. Attempts to reduce the incorporation of the ¹⁴C-methyl group from SAM into pectin by the addition of the alternative methyl donor 5-methyltetrahydrofolate were unsuccessful, supporting the role of SAM as the authentic methyl donor for HGA-MT. The pH optimum for HGA-MT in membranes was 7.8, the apparent Michaelis constant for SAM was 38 μm, and the maximum initial velocity was 0.81 pkat mg⁻¹ protein. At least 59% of the radiolabeled product was judged to be methylesterified homogalacturonan, based on the release of radioactivity from the product after a mild base treatment and via enzymatic hydrolysis by a purified pectin methylesterase. The released radioactivity eluted with a retention time identical to that of methanol upon fractionation over an organic acid column. Cleavage of the radiolabeled product by endopolygalacturonase into fragments that migrated as small oligomers of HGA during thin-layer chromatography, and the fact that HGA-MT activity in the membranes is stimulated by uridine 5′-diphosphate galacturonic acid, a substrate for HGA synthesis, confirms that the bulk of the product recovered from tobacco membranes incubated with SAM is methylesterified HGA.

The cell wall gives shape to cells and plays critical roles in plant development (Carpita and Gibeaut, 1993). Primary cell walls, those walls surrounding growing plant cells, are composed mainly of the polysaccharides cellulose, hemicellulose, and pectin. Three classes of pectin have been detected in plant cell walls: HGA, RG-I, and RG-II (Jarvis, 1984; O'Neill et al., 1990). HGA, the most abundant pectic polysaccharide, is a linear homopolymer of α-1,4-linked d-galacturonic acid that is partially derivatized by methylesterification at C-6, by acetylation at the C-2 or C-3 hydroxyl (De Vries et al., 1986; Ishii, 1995), and in some plants by xylosylation (De Vries et al., 1986; Schols et al., 1995).

The degree of methylesterification of HGA varies during cell culture, and it is believed that the amount and pattern of HGA methylation is important for wall function in growth and development (Jarvis et al., 1988; Schaumann et al., 1993). This belief is supported by the observation that HGA in the walls of young cells is highly methylesterified, whereas HGA in the walls of older cells has a lower degree of esterification (Schaumann et al., 1993). The differences in the degree of methylesterification of pectins are believed to be controlled by the activities of PMT in the Golgi apparatus (Vannier et al., 1992) and PME in the cell wall (Gaffe et al., 1992).

PMT activity has previously been identified in mung bean (Phaseolus aureus L.) seedlings (Kauss et al., 1969). The fact that the PMT from mung bean catalyzed the transfer of ¹⁴CH₃ from SAM to produce a product that released [¹⁴C]methanol after treatment with PME or base (Kauss et al., 1967) provided good evidence that the described enzyme methylated HGA. Furthermore, the rate of PMT activity in mung bean membranes increased in the presence of UDP-GalUA, suggesting that HGA synthesized in the membranes was the methyl acceptor (Kauss and Swanson, 1969). However, no sensitivity of the methylated product to cleavage by EPGase was reported and no subsequent work on the solubilization and purification of the mung bean PMT has been published. Putative PMT activity has also been detected in flax (Linum usitatissimum L.) hypocotyls (Vannier et al., 1992) and flax suspension-cultured cells (Schaumann et al., 1993). The PMT in flax was shown to synthesize a product from which [¹⁴C]methanol was released after treatment with a high concentration of base (1 m NaOH) (Vannier et al., 1992). However, further studies to prove that the enzyme transfers ¹⁴CH₃ from SAM specifically to HGA have not been reported. Recently, the PMT from flax was partially purified and characterized, but the specificity of this enzyme for the type of pectin substrate methylated was not ascertained (Bruyant-Vannier et al., 1996). The fact that the solubilized flax enzyme is activated by exogenous PGA supports its identity as a transferase that methylates homogalacturonan (Bruyant-Vannier et al., 1996).

Bourlard et al. (1995) observed that different types of pectins (e.g. rhamnogalacturonan versus HGA) stimulate the incorporation of methyl groups into pectin at different pH optima. Therefore, it is possible that unique PMTs exist that specifically methylate the different pectic polysaccharides HGA, RG-I, and RG-II to form methylesters or methylethers. Examples of methyl etherification include 2-O-methyl Xyl and 2-O-methyl Fuc in RG-II (Darvill et al., 1978; O'Neill et al., 1996) and 4-O-methyl GlcUA in a side branch of RG-I (An et al., 1995). Thus, unique methyltransferases must exist that incorporate methyl groups into these different pectic polysaccharides.

We report here the identification and characterization, in membranes from tobacco (Nicotiana tabacum L.) cell suspensions, of a PMT that methylates HGA, and we refer to this enzyme as HGA-MT. Tobacco cell suspensions were used, because the enzyme that synthesizes HGA, PGA-GalAT, has been identified and studied in these cells (Doong et al., 1995). A comparison of the characteristics of the HGA-MT from tobacco with the PMTs previously described from flax (Vannier et al., 1992) and mung bean (Kauss et al., 1969) is also presented.

MATERIALS AND METHODS

Chemicals

The chloride salt of SAM, the ammonium salt of UDP-GalUA, pectins of 31 to 93% degree of esterification, and 5-methyltetrahydrofolic acid were purchased from Sigma. Polygalacturonic acid was purchased from Sigma and ICN. Dextran standards were purchased from Pharmacia. Fluoral-P (4-amino-3-penten-2-one) was purchased from Acros (Pittsburgh, PA). [¹⁴C]Methyl SAM (specific activity, 55 mCi/mm) was purchased from American Radiolabeled Chemicals (St. Louis, MO) and Amersham.

Plant Material

Tobacco (Nicotiana tabacum L. cv Samsun) cell-suspension cultures were originally isolated from pith callus tissue (Eberhard et al., 1989). The cells were grown on Murashige and Skoog medium supplemented with 4.5 μm 2,4-D and 90 mm Suc, and subcultured every 12 d (Doong et al., 1995).

Preparation of Membranes from Tobacco Cell-Suspension Cultures

Membranes were prepared by a modification of the method of Villemez et al. (1966). Three- to 4-d-old tobacco cells (75 g) were collected by filtration and homogenized with a polytron in 100 mL of grinding buffer (50 mm Tris-HCl, pH 7.3, 0.4 m Suc, 1% [w/v] BSA, and 1 mm EDTA). The homogenate was strained through a nylon cloth (50-μm pore size) and the filtrate centrifuged at 3,500g for 15 min. The supernatant was centrifuged at 100,000g for 1 h to yield a membrane pellet and the pellet was resuspended in 5 mL of storage buffer (0.4 m Suc, 50 mm Tris-HCl, pH 6.8). HGA-MT activity was measured in membranes (4–6 mg mL⁻¹ protein) that were either assayed immediately after preparation or frozen and stored at −80°C until use. Protein content was determined using a Bradford assay (Bradford, 1976) with BSA as a standard.

HGA-MT Assay

The HGA-MT assay was a modification of that previously described by Kauss and Hassid (1967). The membranes (25 μL, 100–150 μg of protein) were incubated in 25 μL of reaction buffer (50 mm Tris-Mes, pH 8.5, 8 μm [¹⁴C]methyl SAM [0.01 μCi], and 12 μm SAM) at 30°C for times ranging from 5 min to 4 h. The reaction was stopped by the addition of 50 μL of 20% TCA to precipitate the methylated products. These mixtures were centrifuged for 5 min at 4000g. Unincorporated SAM was removed by washing the pellets twice with 200 μL of 2% TCA. The washed pellets were resuspended in 300 μL of water, and the radioactivity incorporated into the product was measured by liquid-scintillation counting using Scintiverse BD scintillation cocktail (Fisher Scientific).

Chemical Extraction of Radiolabeled Product

The pellets obtained after TCA precipitation were partially solubilized with boiling water, 0.5% boiling EDTA, 0.5% boiling ammonium oxalate, 0.5 m imidazole-HCl (pH approximately 6.0) at 25°C, or 0.1 n NaOH at 25°C. After one treatment with 0.5% boiling ammonium oxalate, the pellets were further treated with 0.5% boiling ammonium oxalate containing 1% Triton X-100. All of these treatments were performed for 1 to 2 h, except for the treatment with NaOH, which was performed for 4 to 12 h. After treatment the suspensions were centrifuged and the amount of radioactivity in the supernatant and pellet was measured. Uronic acids were measured by a meta-hydroxybiphenyl assay, as adapted from Blumenkrantz and Asboe-Hansen (1973).

PME Assay

PME activity was detected by the production of methanol (Wojciechowski and Fall, 1996). The assay was modified by incubating 25 μL of membranes, 610 μg of 93% esterified pectin, 90 μg of Fluoral-P, 4 units of alcohol oxidase (ICN), and 153 mm KH₂PO₄, pH 6.0, in the absence or presence of 0.01% Triton X-100. The change in optical density at 405 nm indicated the reaction of formaldehyde (produced by the oxidation of methanol by alcohol oxidase) with 4-amino-3-penten-2-one (Fluoral-P) to yield 3,5-diacetyl-1,4-dihydro-2,6-dimethylpyridine and was a measure of PME activity (Wojciechowski and Fall, 1996).

EPGase and PME Digestion of Product

Methylated product was demethylesterified and/or hydrolytically cleaved by complete digestion with a cloned Aspergillus aculeatus PME expressed in Aspergillus oryzae (Christgau et al., 1996) (gift of Hans Peter Heldt-Hansen, Novo Nordisk, Bagsvaerd, Denmark) and/or an EPGase from Aspergillus niger that were purified from the culture filtrates (gifts of Carl Bergmann, Complex Carbohydrate Research Center, Athens, GA). Methylated product was resuspended in 200 μL of 50 mm sodium acetate, pH 5.0, and treated with 1 to 4 units of EPGase and/or PME for 4 to 12 h at 30°C. The reaction products were analyzed by TLC or HPLC.

TLC of ¹⁴C-Labeled Products

TLC was performed using precoated TLC plates (Silica Gel 60 WF₂₅₄, EM Science, Gibbstown, NJ) run vertically in water:isopropanol:hexyltriethylammonium phosphate (2:1:0.02, v/v) (Q6 ion pair cocktail, Regis Chemical Co., Morton Grove, IL). Radiolabeled product was visualized by exposing the TLC plates to storage phosphor screens (Molecular Dynamics, Sunnyvale, CA). The exposed screens were autoradiographed using a phosphor imager (model 425F, Molecular Dynamics). Sugars were detected by spraying the plates with orcinol reagent (Sigma).

HPLC of ¹⁴C-Labeled Products

Radiolabeled methanol released from the ¹⁴C-labeled product by treatment with PME or NaOH was collected by distillation and separated over a Rezex ROA-organic acid column (Phenomenex, Torrance, CA) in water by HPLC using a chromatography system (flow rate, 0.6 mL min⁻¹; DX 500, Dionex, Sunnyvale, CA). Nonradiolabeled methanol was detected by pulsed amperometric detection with postcolumn addition of 400 mm NaOH (approximately 0.2 mL min⁻¹). Fractions containing radioactivity were measured by liquid-scintillation counting using Scintiverse BD scintillation cocktail. Samples were filtered through 5000 molecular weight pore-size microfilterfuge tubes before chromatography. The amperometric detector was operated with the following pulse sequences: E1 = 0.05 V (duration, 400 ms); E2 = 0.75 V (duration, 210 ms); and E3 = −0.15 V (duration, 390 ms). The sampling period was 200 ms and the response time was 1 s.

Size-Exclusion Chromatography of ¹⁴C-Labeled Products

Radioactive product solubilized by treatment of the pellet for 1 h with 0.5% ammonium oxalate (oxalate fraction) was dialyzed against water at 4°C for 24 h, and the resulting dialysate was lyophilized, resuspended in water (100 μL, 1600 cpm), spin filtered, and separated over a Superose 12 HR 10/30 fast-protein liquid chromatography size-exclusion column (Pharmacia) in 50 mm sodium acetate, pH 5.0, and 10 mm EDTA (flow rate, 0.48 mL min⁻¹). Fractions (0.48 mL) containing radioactivity were measured by liquid-scintillation counting using Scintiverse BD scintillation cocktail. The elution times of dextran molecular mass standards, pectin, and PGA were determined by pulsed amperometric detection.

RESULTS

Establishment of an Assay for HGA-MT

Precipitation of Methylated Product

A major obstacle to the development of an assay for the HGA-MT-catalyzed methylation of HGA by [¹⁴C]SAM was the nonspecific binding of SAM to the product and the resulting high background radioactivity. Vannier et al. (1992) previously reported that the precipitation of methylated pectins in 95% ethanol and subsequent washing of the product with 1 m NaCl in 60% ethanol was successful in decreasing [¹⁴C]SAM background. In our hands, however, these conditions were not sufficient to reduce background cpm, and the resulting pellets had very high backgrounds and variable cpm (Table I).

Table I.

Comparison of different procedures for the precipitation and washing of ¹⁴C products

Washing Procedure	Precipitation Method
	50% TCA	20% TCA	10% TCA	95% Ethanol	TCA-ethanol	Methanol:chloroform	4 n LiCl	8 m Guanidine
No wash	–	H, S	–	–	–	–	–	–
20% TCA	S	S	–	–	–	–	–	–
2% TCA, 1 time	–	H, S	–	–	–	–	–	–
2% TCA, 2 times	S	S	L	H, V	H, S	–	V, L	H, N
Methanol-chloroform	–	H, S	N	N	–	–	–	–
65% Ethanol	–	–	–	H, V	H, L	H, V	–	–
Ethanol-NaCl	–	–	–	H, V	–	–	–	–
Ethanol-TCA	–		–	H, N	H, L	–	–	–
2 n LiCl	–	–	–	–	–	–	H	–
4 n guanidine	–	–	–	–	–	–	–	H, N

Membranes were incubated with [¹⁴C]SAM, and the reaction was stopped by the addition of 50 μL of 10 to 50% TCA, 20% TCA-65% ethanol, 4 n LiCl or 8 m guanidine or 450 μL of 95% ethanol or methanol:chloroform (1:1, v/v). Each pellet was washed twice with 200 μL of 2 to 20% TCA, methanol:chloroform (1:1, v/v), 65% ethanol, 65% ethanol-1 m NaCl, 65% ethanol-2% TCA, 2 n LiCl, or 4 m guanidine. N, No activity; S, same value (1000–1500 cpm) as in the standard conditions; L, low incorporation (<100 cpm) compared with standard conditions; H, very high background; V, variable results from experiment to experiment. The boxed S indicates the selected standard method to analyze the HGA-MT activities.

We tested the efficacy of a number of different treatments to precipitate methylated products produced in tobacco membranes and to remove unincorporated [¹⁴C]SAM. The treatments included methanol:chloroform (1:1, v/v), as used for the precipitation of HGA (Doong et al., 1995); 4 n LiCl and 8 m guanidine, as used for precipitation of DNA or RNA (Ausubel et al., 1996); and TCA, as used for the precipitation of proteins (Ausubel et al., 1996) and of some polysaccharides (Kauss and Hassid, 1967). The use of 20 to 50% TCA for precipitation was the only system that resulted in the recovery of methylated compounds. The lowest background and the most reproducible cpm were achieved by washing the pellet twice with 2% TCA after precipitation. This assay method is most similar to that of Kauss and Hassid (1967). Table II shows the recovery of radioactivity in the pellet at each step of the HGA-MT assay. No incorporation of radioactivity into the pellet above background levels was obtained in control reactions containing the buffer and boiled or heat-treated membranes (Table II). Thus, the incorporation of radioactivity detected using the HGA-MT assay has the characteristics expected for an enzyme-catalyzed reaction.

Table II.

The recovery of radiolabeled product after each step of the HGA-MT assay

Treatment	Background	Assay
	cpm
20% TCA precipitation	700  ± 100	2200  ± 200
First wash with 2% TCA	200  ± 50	1600  ± 100
Second wash with 2% TCA	60  ± 30	1500  ± 40
Third wash with 2% TCA	60  ± 20	1500  ± 40
Boiled membranes (5 min)	120  ± 30	140  ± 25
Heated membranes (1 h at 60°C)	127  ± 20	138  ± 20

Membranes were incubated with [¹⁴C]SAM (18,000 cpm) for 30 min at 30°C. The reaction was stopped by the addition of 50 μL of 20% TCA and the pellet was washed twice with 2% TCA. The background cpm are those recovered when TCA was added to reaction buffer before the addition of enzymes (i.e. a time-0 control). The cpm recovered from control reactions containing membranes pretreated by boiling for 5 min or heating at 60°C for 1 h are also shown. The results are the average cpm ± sd from duplicate samples from two or three independent experiments.

Nature of the Methyl Donor

The methylation of HGA occurs by the transfer of a methyl group from a methyl donor to HGA. The methyl donor SAM has been reported to be the donor for HGA-MT (Kauss et al., 1967; Vannier et al., 1992). However, the possibility exists that other methyl donors such as 5-methyltetrahydrofolate may be the direct donor. To test this possibility the amount of the ¹⁴C-methyl group incorporated into the product in reactions containing 10 μm [¹⁴C]SAM in the presence and absence of 10 μm nonradiolabeled 5-methyltetrahydrofolate was determined. There was no decrease in the amount of radioactivity (98 ± 6% of controls, average ± se from duplicate samples from two experiments) incorporated into the product in the presence of 5-methyltetrahydrofolate, indicating that 5-methyltetrahydrofolate is not a methyl donor for HGA. Kauss and Hassid (1967) obtained comparable results using radiolabeled methyltetrahydrofolate. We conclude that SAM is the direct methyl donor for the HGA-MT studied here.

Characterization of HGA-MT

Initial attempts to identify HGA-MT activity using membranes prepared from tobacco by the method of Doong et al. (1995) did not generate detectable HGA-MT activity. Membranes containing detectable HGA-MT activity were obtained by the method of Vannier et al. (1992). It was determined that the homogenization and storage buffers used by Doong et al. (1995) contained inhibitors of membrane-bound HGA-MT (0.1% β-mercaptoethanol, 25% glycerol, and 25 mm KCl) that did not allow the detection of the HGA-MT activity.

Effect of Cations on HGA-MT Activity

The presence of 25 mm KCl in the homogenization buffer inhibited HGA-MT activity by 40%. An inhibition of mung bean PMT by 2.5 mm KCl has been reported by Kauss and Hassid (1967). The presence of 0.1 to 25 mm MgCl₂ did not affect the incorporation of methyl groups into product by microsomal membranes, as previously demonstrated for mung bean PMT by Kauss and Hassid (1967). There was also no effect of 0.1 to 50 mm MnCl₂ on HGA-MT activity.

Effect of Temperature on HGA-MT Activity

The temperature optimum for HGA-MT is between 25 and 40°C (Fig. 1A). This optimum is comparable to that of other plant membrane-bound enzymes (McNab et al., 1968; Misawa et al., 1996). The HGA-MT activity is reduced by more than 50% at temperatures above 55°C. A thermal-inactivation curve for HGA-MT at 60°C is shown in Figure 1B. HGA-MT undergoes an exponential decay of activity at 60°C, with a 50% reduction in activity after 5 min and a complete inactivation after 40 min.

Figure 1.

Effect of temperature on HGA-MT activity in membranes from tobacco cells. A, Temperature-optimum curve. Tobacco membranes were incubated with [¹⁴C]SAM for 5 min at the temperatures indicated and product was recovered by the HGA-MT assay. The data represent the average cpm ± sd of product recovered from duplicate samples from two independent experiments. B, Thermal-inactivation curve. Tobacco membranes were incubated at 60°C for the times indicated and then incubated with [¹⁴C]SAM for 5 min at 25°C. Product was recovered by the HGA-MT assay. The data represent the average cpm from duplicate samples from one experiment.

Effect of pH on HGA-MT Activity

The effect of reaction pH ranging from 5.0 to 9.3 on HGA-MT activity was determined (Fig. 2). A major peak of activity was obtained at pH 7.8 to 8.0. An apparent minor pH optimum (7.0–7.3), most apparent in the 5-min reaction, is the same as the reported pH optimum for flax (Bruyant-Vannier et al., 1996) and mung bean PMT (Kauss and Hassid, 1967).

Figure 2.

Effect of pH on HGA-MT activity in membranes from tobacco cells. Tobacco membranes were incubated with [¹⁴C]SAM for 5 (•) or 15 (▴) min in reaction buffer adjusted to the indicated pH by the appropriate mixture of Tris and Mes buffers. Product was recovered by the HGA-MT assay. The data represent the average cpm ± sd of product recovered from duplicate samples from three to seven independent experiments.

Effect of Triton X-100 on HGA-MT

Triton X-100 was used to permeabilize the membranes in an attempt to increase HGA-MT activity by increasing the amount of SAM accessible to the enzyme. The HGA-MT activity was not modified in the presence of 0.01% Triton X-100, a concentration commonly used to permeabilize membranes (Goubet et al., 1994). When a high concentration of 0.1% Triton X-100 was used, HGA-MT activity was inhibited by 30%.

Time Course of PME and HGA-MT Activities in Membranes

Kauss et al. (1969) and Bruyant-Vannier et al. (1995) have reported that PME activity is present in microsomal membranes isolated from mung bean and flax. We therefore tested whether any PME was present in the tobacco membranes. Tobacco cell suspensions were harvested 0, 2, 3, 4, 6, 9, 12, and 16 d after transfer of cells to fresh media, and membranes were isolated and analyzed for PME and HGA-MT activities (Table III). PME activity was detected in membranes prepared from 0- and 6- to 12-d-old cells. The PME may represent PME from the cell wall that bound to the membranes during tissue homogenization. Alternatively, PME may be located inside the Golgi apparatus or ER. To test whether a significant proportion of PME activity was located inside the membrane vesicles, membranes were permeabilized by treatment with 0.01% Triton X-100 and PME activity was measured. Permeabilization of the membranes did not result in an increase in PME activity. Thus, no direct evidence for the existence of PME within the membrane vesicles was obtained.

Table III.

PME and HGA-MT activity in membranes from tobacco cell suspensions as a function of cell age

	Day of Culture
	0	2	3	4	6	9	12	16
	pmol−1 s−1 g−1 cell
PMEa	23  ± 8	0	0	0	21  ± 8	21  ± 12	18  ± 13	0
PME + Tritona	24  ± 12	0	0	0	21  ± 9	22  ± 6	21  ± 7	7  ± 9
Fresh weight of cellsb	18  ± 6	23  ± 7	24	28  ± 9	62  ± 3	120  ± 14	224  ± 70	380  ± 28
HGA-MTc	0.029  ± 0.006	0.15  ± 0.02	0.18  ± 0.02	0.13  ± 0.01	0.09  ± 0.03	0.12  ± 0.07	0.10  ± 0.04	0.057  ± 0.006

^aPME activity was detected by the production of methanol. The activity was measured using pectin of 94% degree of esterification in the presence or absence of 0.01% Triton X-100. The data represent the average activity ± sd of duplicate samples from a minimum of three individual experiments. 

^bFresh weight (grams) data represent the average mass of cells from two individual experiments. Day 0 represents 12-d-old cells immediately after transfer to fresh medium. 

^cHGA-MT activity was detected as the incorporation of [¹⁴C]SAM into product in 15-min reactions. The products were precipitated by 20% TCA and washed with 2% TCA. The data represent the average activity ± sd of duplicate samples from two individual experiments except for d 3 data, which represent the average data ± sd of duplicate samples from one experiment. 

The highest HGA-MT activity was expressed at 2 to 4 and 9 to 12 d after the transfer of the cells (Table III). The activity observed during the lag phase has been described for flax cells (Schaumann et al., 1993); however, a second optimum later during culture has not been reported. Based on results from at least 20 experiments, the average HGA-MT activity in membranes from 2- to 4-d-old cells is 0.14 ± 0.02 pmol s⁻¹ g⁻¹ cell. The HGA-MT activity from 9- to 12-d-old cells is more variable (0.11 ± 0.07 pmol s⁻¹ g⁻¹ cell). This variability can be explained by the presence of PME in the membranes of these cells (Table III). An attempt was made to increase HGA-MT activity by permeabilizing membranes in the presence of the putative exogenous acceptor PGA. Incubation of membranes in the presence of 0.01% Triton X-100, with or without PGA, caused no increase in HGA-MT activity (data not shown). Higher concentrations of Triton X-100 (0.1%), with or without PGA, inhibited HGA-MT by 30% (data not shown). These results suggest that the endogenous acceptor may be present in excess and thus exogenous pectin does not increase HGA-MT activity. The standard HGA-MT assay is performed using 3- to 4-d-old cells in the absence of permeabilizing agents.

Reaction Kinetics

A time course of the incorporation of ¹⁴C into the product is shown in Figure 3. The rate of the reaction is linear during the first 10 to 15 min. The initial velocity is 0.17 pkat mg⁻¹ protein, which is significantly greater than the initial velocity of 0.005 pkat mg⁻¹ protein reported for PMT from flax (Bruyant-Vannier et al., 1996). The amount of HGA-MT activity at 60 min at 30°C was proportional to the amount of membrane protein assayed through 40 μg of protein per reaction. The apparent K_m of HGA-MT for SAM is 38 μm, and the V_max ± sd is 0.81 ± 0.05 pkat mg⁻¹ protein at pH 7.8 (Fig. 4). The V_max of 0.045 pkat mg⁻¹ protein reported for the PMT from mung bean (our calculation from data reported by Kauss et al. [1969]) is lower than the V_max for HGA-MT from tobacco. The reported apparent K_m of 30 μm SAM for PMT from flax cells (Bruyant-Vannier et al., 1996) and 60 μm SAM for PMT from mung bean (Kauss and Hassid, 1967) are similar to the value reported here for HGA-MT in tobacco.

Figure 3.

Time course of the incorporation of the ¹⁴C-methyl into precipitable products. Membranes were incubated with [¹⁴C]SAM (18,000 dpm) and product was recovered by the HGA-MT assay. The data represent the average cpm ± sd of product recovered from duplicate samples from three independent experiments.

Figure 4.

Hanes-Woolf plot of the kinetics of methyl incorporation into precipitable product in tobacco cell membranes. [SAM]/V₀ (initial velocity) is the concentration of SAM (μm) divided by V₀ (pmol methyl incorporated s⁻¹ mg⁻¹ protein). Membranes were incubated with [¹⁴C]SAM for 15 min and the product was recovered by the HGA-MT assay. The data represent the average of duplicate samples from two independent experiments.

Effects of Different Potential Oligosaccharide and Polysaccharide Substrates on HGA-MT Activity

To determine whether HGA-MT activity was dependent on a PGA-GalAT-catalyzed synthesis of homogalacturonan (i.e. stimulated by UDP-GalUA) or dependent on exogenous HGA or pectin substrates, we tested whether the addition of UDP-GalUA, OGA, or pectin could stimulate HGA-MT activity. The addition of OGA, PGA, or pectin with different degrees of esterification had no effect on the HGA-MT activity (data not shown). A similar lack of stimulation by PGA or pectin was also observed for PMT from mung bean (Kauss and Swanson, 1969) and flax (Bruyant-Vannier et al., 1996). The lack of stimulation of HGA-MT by the pectic oligosaccharides and polysaccharides was observed, regardless of whether the membranes were permeabilized by 0.01% Triton X-100 or not. UDP-GalUA, however, stimulated the incorporation of methyl groups into HGA (Fig. 5). The greatest amount of product was recovered in the presence of 20 to 50 μm UDP-GalUA. The activation of tobacco HGA-MT by UDP-GalUA, however, was only 20% compared with the reported 200% stimulation of the PMT from mung bean by UDP-GalUA (Kauss et al., 1969). The presence of 0.01% Triton X-100 to permeabilize the membranes did not affect the stimulatory effect of UDP-GalUA on HGA-MT activity (data not shown).

Figure 5.

Effect of UDP-GalUA on HGA-MT activity in tobacco cell membranes. Membranes were incubated with [¹⁴C]SAM for 15 min and the product was recovered by the HGA-MT assay. The data represent the average cpm ± sd of product recovered from duplicate samples from seven independent experiments.

EDTA, a chelator of calcium ions, solubilizes some pectins (Selvendran and O'Neill, 1995) and could potentially activate HGA-MT activity by allowing better access of HGA substrates to HGA-MT. The HGA-MT of tobacco was inhibited at EDTA concentrations greater than 2 mm, and no activation of HGA-MT was observed at any EDTA concentrations tested (0.1–10 mm). In contrast, Kauss and Hassid (1967) have shown that 2 mm EDTA activates the PMT activity from mung bean.

Characterization of the Product of HGA-MT

Radiolabeled Product Is Solubilized from the Pellet by Chemical Extractions and Enzymatic Hydrolysis

SAM donates methyl groups to DNA, RNA, protein, lipid, and carbohydrate acceptors (Chiang et al., 1996); therefore, it was necessary to determinate what fraction of the recovered product was methylated HGA. The total product was treated with boiling water, boiling EDTA, boiling ammonium oxalate, or imidazole (Table IV) in an attempt to solubilize pectin from the pellet (Mort et al., 1991; Schaumann et al., 1993; Schols et al., 1995; Selvendran and O'Neill, 1995). These treatments released at most 24% of the radioactive product. Similarly, only 24% of the total product was solubilized by treatment with EPGase, an enzyme specific for HGA (data not shown).

Table IV.

Chemical and enzymatic solubilization of total ¹⁴C-methylated products

Chemical Treatment	Radioactivity Released from the Product Pellet
	%
Water	12  ± 4
0.5% EDTA	8  ± 6
0.5% Ammonium oxalate	24  ± 7
0.5 m Imidazole	11  ± 7

After incubation of the membranes with [¹⁴C]SAM, the methylated products were precipitated by 20% TCA and the pellets washed twice with 2% TCA. The resulting pellets were treated with boiling water, 0.5% boiling EDTA, 0.5% boiling ammonium oxalate, or 0.5 m imidazole-HCl (pH 6.0) for 1 to 2 h. The radioactivity was measured in the supernatant and the pellet, and the percentage of radioactivity solubilized from the pellet in each fraction was calculated. The data represent the average percentage of radioactivity solubilized ± sd from duplicate samples from 3 to 12 independent experiments.

Because TCA precipitates polysaccharides, proteins, and lipids, we reasoned that a mixture of macromolecules was co-precipitated by the TCA, thus making it difficult to solubilize pectin from the complex pellet. The pellet was therefore first treated with ammonium oxalate to solubilize the easily accessible pectin, yielding a supernatant referred to as the oxalate fraction (Fig. 6). The remaining pellet was then treated with ammonium oxalate containing 1% Triton X-100 to yield a second supernatant called the oxalate-Triton fraction (Fig. 6). Triton X-100 was used to solubilize any protein-lipid-polysaccharide complex present in the pellet. The combined oxalate and oxalate-Triton fractions contained 82% of the radioactivity from the total pellet and 7.9% of the total proteins (Fig. 6).

Figure 6.

Flow diagram of how the radiolabeled methylated product was fractionated. The table inset gives the percentage of radioactivity in each fraction and the percentage of total membrane protein. ND, Not determined.

The oxalate and oxalate-Triton fractions were dialyzed and analyzed for uronic acid-positive material. Approximately 40 and 72 μg of uronic acid-positive material was recovered in the oxalate and oxalate-Triton fractions, respectively, from each initial product pellet.

Solubilization of Methylated Product by EPGase and PME

To confirm that the radiolabeled product was methylated HGA, we tested the sensitivity of the product in the oxalate and oxalate-Triton fractions to fragmentation by purified EPGase and PME. PME removes methyl groups specifically from the C-6 methylester of HGA and EPGase hydrolyzes the α-(1→4)-GalUA linkages in HGA that has at least four contiguous nonmethylesterified GalUAs (Chen and Mort, 1994; Benen et al., 1996).

A representative thin-layer chromatogram showing the separation of intact product and enzyme-treated product from the oxalate fraction is shown in Figure 7. Comparable results were obtained for product in the oxalate-Triton fraction (data not shown). The intact product in the oxalate fraction remained at the origin at a location comparable to PGA and pectin standards. Orcinol-positive material was also detected at the origin (orcinol reagent detects sugars). Digestion of the oxalate fraction with EPGase resulted in a loss of 52% of the radioactive product at the origin and the migration of the radioactivity to positions similar to di-, tri-, and heptagalacturonic acid. Digestion of the oxalate fraction with PME resulted in a loss of 80% of the radioactive product at the origin.

Figure 7.

Representative thin-layer chromatogram of intact, EPGase- and PME-treated product in the oxalate fraction. Membranes were incubated with [¹⁴C]SAM, and the pellet obtained after TCA precipitation and washing was treated for 1 h with 0.5% ammonium oxalate. After treatment, the solutions were centrifuged and the supernatant was dialyzed and lyophilized. Equal amounts of radioactive products were treated with 1 unit of EPGase or PME. The samples were separated by TLC. Radiolabeled product was visualized by exposing the TLC plate to storage phosphor screens for 3 to 4 weeks. Similar results were obtained in three individual TLCs from three separate experiments. The arrows represent the location of digalacturonic acid (DiGalA), trigalacturonic acid (TriGalA), and heptagalacturonic acid (HeptaGalA) standards. Commercially available PGA and pectin standards remained at the origin.

The average amount of radioactivity released by enzymatic treatment of product in the oxalate and oxalate-Triton fractions from at least three independent experiments is shown in Table V. EPGase treatment of the oxalate and oxalate-Triton fractions resulted in the fragmentation of 60 and 35% of the product, respectively, into oligosaccharides that migrated away from the origin during TLC. Thus, at least 35% of the original product pellet contained radioactivity in HGA. This value, however, is a minimum estimate, because any EPGase products that remained at the origin would not have been included. Considerably more radioactivity was released from the oxalate fraction (82%) and the oxalate-Triton fraction (67%) by PME. Because PME is specific for the C-6 methylester of HGA, these results show that at least 59% of the radioactivity incorporated into the intact product was HGA. Similar results were obtained by treatment of the corresponding fractions with 0.1 n NaOH for 10 to 12 h at 25°C, confirming that the methyl groups were present in a methylester linkage. The radioactivity released by PME or NaOH (data not shown) was not detected by TLC, because the methanol evaporated when the samples were lyophilized or at the time of their application to TLC. In contrast, when the product treated by PME or NaOH was directly separated or first distilled and then separated by HPLC using a Rezex ROA-organic acid column, a single peak of radioactivity with a retention time identical to that of methanol was detected (data not shown). These results confirm that at least 59% of the radioactive product produced in tobacco membranes was methylated HGA.

Table V.

Sensitivity of oxalate and oxalate-Triton fractions of methylated products to EPGase, PME, and base

Treatment	Radioactivity Releaseda		Total Product Releasedb
	Oxalate fraction	Oxalate-Triton fraction
	%
EPGase	60  ± 15	35  ± 21	35
PME	82  ± 8	67  ± 9	59
PME + EPGase	84  ± 6	69  ± 11	60
0.1 n NaOH	88  ± 5	66  ± 8	59

^aMembranes were incubated with [¹⁴C]SAM. The pellet obtained after TCA precipitation and washing was treated one or two times for 1 h with 0.5% boiling ammonium oxalate and 1 h with 0.5% boiling ammonium oxalate with 1% Triton X-100. The resulting fractions were treated with EPGase for 16 h at 25°C, PME for 16 h at 25°C, or NaOH for 8 h at 25°C. The radioactivity released from polymeric product (located at origin) after the different treatments was calculated from the counts detected at the origin in a TLC plate before and after treatment. Data represent the average ± sd from duplicate samples from at least three independent experiments. 

^bThe percentage of radioactivity released from total product by each treatment was estimated as follows: (% radioactivity released in the oxalate fraction × 0.24 [see Fig. 5]) + (% radioactivity released in the oxalate-Triton fraction × 0.58 [see Fig. 5]). 

Size-Exclusion Chromatography of Radiolabeled Product

Size-exclusion chromatography of product solubilized from the intact pellet by ammonium oxalate resulted in several peaks of radioactivity that eluted from the column from 16 to 40 min (Fig. 8). The size of the products, compared with dextran standards, ranged from approximately 200 to 1.7 kD. The broad, overlapping peaks eluting from 16 to 28 min (40–200 kD) represent 39% of the radioactivity and elute with retention times similar to those of pectin standards of either 90 or 30% esterification. The two small peaks eluting from 29 to 33 min (6–40 kD) represent 10% of the radioactivity and elute with retention times comparable to those of commercially available pectin and PGA. The large peak from 34 to 40 min represents 51% of the radioactivity and elutes similarly to commercially available PGA. We do not know why some PGA elutes relatively late during size-exclusion chromatography, although an interaction between PGA and the column matrix is a possibility. The anomalous behavior of GalUA-containing polymers during chromatography has previously been noted (Mort et al., 1991). It is likely that the methylated products are greater in size than a decagalacturonide, because the radiolabeled product in the oxalate fraction remained at the origin during TLC under conditions in which homogalacturonans of degrees of polymerization less than 11 migrate away from the origin (see Fig. 7). We interpret the complexity of the size-exclusion chromatography profile to indicate that the methylated products are heterogeneous in size and/or in their degree of methylesterification.

Figure 8.

Size-exclusion chromatography of products solubilized from intact pellet by ammonium oxalate (oxalate fraction). Membranes were incubated with [¹⁴C]SAM (70,000 dpm) for 60 min, and the product was recovered by the HGA-MT assay. Products were separated by size-exclusion chromatography over a Superose 12 HR 10/30 column and fractions (0.48 mL) were collected for 50 min. The elution times of dextran molecular mass standards (480, 70, 40, 10, and 6 kD) are indicated by arrows. The resolution limit for the dextran was calculated as 1.7 kD. The range of elution times for 30 and 90% esterified pectins (- - - -) and for two commercially available sources of PGA (——) are indicated.

DISCUSSION

A PMT in tobacco cell membranes that methylates HGA (HGA-MT) has been identified. The product synthesized by membranes incubated with [¹⁴C]SAM was shown to contain at least 59% HGA based on the hydrolysis of the ¹⁴C-methyl group by PME. The HGA-MT has a pH optimum of 7.8, a K_m for SAM of 38 μm, and a V_max of 0.81 pkat mg⁻¹ protein. The characteristics of the tobacco HGA-MT are compared with previously described PMTs from mung bean (Kauss et al., 1969) and flax (Vannier et al., 1992) in Table VI. The previously described PMTs have a similar K_m for SAM but a different pH optimum than the HGA-MT from tobacco.

Table VI.

Comparison of the kinetics and product of the PMTs studied from mung bean (Kauss et al., 1969) and flax (Bruyant-Vannier et al., 1996) with the HGA-MT from tobacco

	Plant
	Flax	Mung bean	Tobacco
pH optimum	6.8	7.1	7.8
Km for SAM (μm)	30	59	38
Vmax for SAM (pkat mg−1 protein)	n.d.a	0.045b	0.81
PME-sensitive product	n.d.	+c	+c
EPGase-sensitive product	n.d.	n.d.	+d

^an.d., Not determined. 

^bThe V_max is our calculation using data from Kauss et al. (1969). 

^c+, Radioactive methanol is released from product by treatment with PME. 

^d+, Product is fragmented by treatment with EPGase. 

The identification of HGA-MT was complicated by the nonspecific binding of SAM to compounds in the product pellet and by the presence of PME in the microsomal membranes. The use of a modification of the TCA-precipitation method of Kauss and Hassid (1967) eliminated most of the nonspecific binding of SAM. Bruyant-Vannier et al. (1996) assayed the PMT from flax membranes by ethanol precipitation and washing of the methylated products. This technique did not work with tobacco membranes. It is possible that differences between the flax and tobacco PMTs or differences in the methylated pectins produced may account for the requirement of different assays for flax versus tobacco PMTs. The use of 2- to 4-d-old tobacco cells, which express no PME activity, overcame the problems associated with the demethylesterification of the methylated HGA product by PME in older cells. The PME may have been introduced as a contaminant from the cell wall during membrane preparation. Alternatively, it is possible that PME resides inside the membrane vesicles, although the lack of latency of the PME activity provides no evidence to support this. Previously, Bruyant-Vannier et al. (1995) reported that PME was associated with flax membranes, and these authors suggested that PME is naturally present inside membranes. The presence of PME inside the membranes would not be totally unexpected, because proteins to be transported to the wall are processed through the ER and the Golgi apparatus.

Maximum HGA-MT activity was detected during the lag phase of growth, when the fresh or dry cell weight does not change but cell division has probably started. A similar result has been obtained in flax cell suspensions (Schaumann et al., 1993). The methylated pectins are synthesized and integrated in the cell wall during cell elongation and division (Schaumann et al., 1993). The variability of HGA-MT activity during the exponential phase of cell growth can be explained by the presence of PME activity in the membranes.

The pH optimum for HGA-MT was 7.8, with a minor optimum at pH 7.3. Although we have only described the HGA-MT activity at pH 7.8 in this paper, similar analyses were performed for HGA-MT activity at pH 7.3 (F. Goubet and D. Mohnen, unpublished results) in an effort to determine whether the minor pH optimum at 7.3 represented an HGA-MT isoenzyme. No evidence was obtained to support the hypothesis that the HGA-MT activity at these two pH values originated from unique enzymes.

To study the relationship between HGA-MT and PGA-GalAT, the effect of the nucleotide sugar UDP-GalUA (i.e. the substrate for PGA-GalAT) and HGA (the putative substrate for HGA-MT) on HGA-MT activity was tested. The addition of UDP-GalUA to membranes stimulates HGA-MT by 20%. The activation of HGA-MT by UDP-GalUA has previously been described (Kauss and Swanson, 1969). In contrast, OGA, PGA, and pectin of different degrees of esterification have no effect on HGA-MT activity. Kauss and Swanson (1969) and Bruyant-Vannier et al. (1996) have also shown no stimulation of membrane-bound PMT by exogenous pectins. Four factors can explain the absence of a stimulatory effect of HGA, PGA, or pectin on HGA-MT activity. First, the endogenous pectin substrates may be present in excess. Second, the exogenous pectin may be either larger or smaller than the size required for recognition by HGA-MT. Third, the PGA and the HGA may not penetrate the membranes. Attempts to overcome the latter potential limitation by permeabilization of the membranes using 0.01 or 0.1% Triton X-100 did not yield any stimulation of HGA-MT activity by exogenous PGA or pectin. Fourth, the HGA-MT may require that HGA is being actively synthesized. Evidence to support the fourth possibility is the stimulation of HGA-MT by the addition of UDP-GalUA, a substrate for HGA synthesis by PGA-GalAT.

It is not known whether one enzyme both synthesizes and methylates HGA or whether PGA-GalAT and HGA-MT exist as separate enzymes. The fact that UDP-GalUA increases the amount of [¹⁴C]SAM incorporated into HGA suggests that GalUA from UDP-GalUA is first incorporated into HGA and that the HGA is subsequently methylesterified. The inability of exogenous SAM to stimulate the in vitro incorporation of UDP-GalUA into HGA (Kauss and Hassid, 1967; Doong et al., 1995) suggests that methylesterification is not directly linked to the synthesis of HGA. To determine if one or more enzymes is required for HGA synthesis and methylation, it will be necessary to purify the enzymes.

At least 59% of the radiolabeled product recovered by TCA precipitation is methylated HGA. The remainder of the radioactivity is insoluble material present in the pellet after TCA treatment (18%) and radioactive compounds solubilized by oxalate or oxalate-Triton X-100 but not fragmented by EPGase, PME, or NaOH (23%). The radioactive product that was not hydrolyzed by either NaOH or PME may contain methyl groups incorporated into a nonmethylester linkage such as a methylether linkage (Vannier et al., 1992).

In conclusion, HGA-MT has been detected in tobacco cells and a procedure to study the activity of this enzyme has been described. The identification of HGA-MT will facilitate its solubilization and purification to better understand how pectin is synthesized and to study the role of HGA-MT in plant growth and development.

Abbreviations:

EPGase

endopolygalacturonase

HGA

homogalacturonan

HGA-MT

homogalacturonan-methyltransferase

OGA

oligogalacturonides of degree of polymerization of 7 to 23

PGA

polygalacturonate

PGA-GalAT

polygalacturonate-4-α-galacturonosyltransferase

PME

pectin methylesterase

PMT

pectin methyltransferase

RG-I and RG-II

rhamnogalacturonan I and II, respectively

SAM

S-adenosyl-l-Met