in peak I has a Km of 217 micromolar for DIMBOA and its activity on DIBOA was too low .... A Waters3 model 244 high-performance liquid chromato- graph equipped .... and to decrease to 70% of optimum as the pH was adjusted to 8.0 and 9.0.
Plant Physiol. (1989) 90, 1071-1076
Received for publication November 8, 1988 and in revised form March 7, 1989
0032-0889/89/90/1071/06/$0.1 00/0
Hydroxamic Acid Glucosyltransferases from Maize
Seedlings' Bryan A. Bailey* and Russell L. Larson Department of Biochemistry (B.A.B.) and Department of Agronomy, and U.S. Department of Agriculture, Agricultural Research Service (R.L.L.), University of Missouri, Columbia, Missouri 65211 ABSTRACT
which differs from DIMBOA by a methoxyl group, is present at a lesser concentration (Fig. 1) (27). DIMBOA has been implicated in resistance to several diseases (3, 16) and insects (17) in maize although many of these relationships have not been substantiated. By far the most extensively characterized biological function proposed for DIMBOA in the developing maize plant is its resistance effect on the first brood of the European corn borer, Ostrinia nubilalis (Hbn.) (1 1-13, 21). The concentration of DIMBOA in young maize plants can be closely correlated with resistance to the European corn borer. Investigation of the pathway involved in hydroxamic acid production has been limited. The precursors of the hydroxamic acids, like many other secondary metabolites, are products of the shikimic acid pathway (20). Anthranilic acid and phosphoribosyl pyrophosphate have been shown to serve as the immediate precursors in hydroxamic acid synthesis in maize (23). Subsequent modifications of the hydroxamic acid structure include hydroxylation, methylation, and glucosylation. The present study concerns identification of two enzymes that are capable of catalyzing the addition of a glucose moiety at the 2-position of DIMBOA (Fig. 1). The extraction and characterization of these glucosyltransferases is seen as an initial step in the ultimate understanding of DIMBOA synthesis in maize.
Hydroxamic acids occur in several forms in maize (Zea mays L.) with 2,4-dihydroxy-7-methoxy-2H-1,4-benzoxazin-3(4H)-one (DIMBOA) being the predominant form and others including 2,4dihydroxy-1,4-benzoxazin-3-one (DIBOA) being found at lower concentrations. Two enzymes capable of glucosylating hydroxamic acids were identified in maize protein extracts and partially purified and characterized. The total enzyme activity per seedling increased during the first 4 days of germination and was concurrent with the accumulation of DIMBOA. Purification of the enzymes by ammonium sulfate precipitation followed by Sephadex G-200 and Q-Sepharose gel chromatography resulted in a 13-fold increase in specific activity. The enzymes are initially separated into two peaks (peak 1 and peak 2) of activity by 0-Sepharose gel chromatography. The peak 1 glucosyltransferase had 3.6% of the DIMBOA glucosylating activity when DIBOA was used as substrate, whereas this percentage increased to 57% for the peak 2 enzyme. The enzyme in peak 2 has a Km of 174 micromolar for DIMBOA and a Km of 638 micromolar for DIBOA; the enzyme in peak I has a Km of 217 micromolar for DIMBOA and its activity on DIBOA was too low to determine a Km. The identification of two glucosyltransferases capable of glucosylating hydroxamic acids in vitro serves as an initial step in the characterization of the enzymes involved in production of hydroxamic acids in maize.
MATERIALS AND METHODS
The presence of the cyclic hydroxamic acids in grain crops has been known since before 1957 (24). Most of the previous work on hydroxamic acids has centered on identification of the chemical structures involved ( 14, 20) and the relationship between the various compounds and biological phenomena such as herbicide tolerance (8) and insect and disease resist-
Chemicals The chemicals used were commercially available except DIMBOA, DIBOA, and their glucosides, which were extracted from seedlings of the maize inbred CI3 1A and Frederick rye, respectively.
ance (12, 16, 17).
The hydroxamic acids exist in the plant as glucosides that are readily hydrolyzed when the structural integrity of the
Seedling Culture
tissue is destroyed. The predominant hydroxamic acid in maize (Zea mays L.) is DIMBOA2 (25), whereas DIBOA,
Seed of the maize (Zea mays L.) inbred C13 1A was placed in aluminum pans lined with plastic wrap and one sheet of Whatman No. 1 filter paper. Thirty mL of 1.0 mM CaSO4 was added to each pan with vermiculite being used to absorb
Supported in part by funds to B. A. B. from the Food for the 21 st Century fellowship program of the University of Missouri, Columbia. Cooperative investigations, Agricultural Research Service, U.S. Department of Agriculture, and Missouri Agricultural Experiment Station, Columbia, Missouri 65211. Journal Series No. 10485. 2 Abbreviations: DIMBOA, 2,4-dihydroxy-7-methoxy-2H- 1,4-benzoxazin-3(4H)-one; DIBOA, 2,4-dihydroxy- 1,4-benzoxazin-3-one; E262, extinction coefficient at 262 nanometers; k', capacity factor; UDPG, uridine 5'-diphosphoglucose; NEM, N-ethylmaleimide.
R, +
OH
COH
UDPG
_
_
O>0 OGlu l l ~~+UDP Nw O
COH
Figure 1. Glucosylation of hydroxamic acids utilizing UDPG as a glucose doner. DIBOA (R=H); DIMBOA (R=O-CH3). 1071
1072
BAILEY AND LARSON
moisture. The pans were covered with aluminum foil and incubated at 30°C for up to 5 d.
excess
Purification of Standards DIBOA, DIMBOA, and their respective glucosides were purified using modifications of established methods. Aglycones were purified (7) by grinding seedling tissue in a mortar in distilled water and allowing the mixture to stand for 1 h after which the debris was removed by centrifugation at 26,000g for 20 min. The resulting supernatant was acidified to pH 2.0 with 6 N HCI and heated to 65°C, cooled on ice for 10 min, and centrifuged (26,000g, 20 min). This supernatant was extracted with 2.5 volumes of ethyl acetate, and the ethyl acetate was collected and evaporated under reduced pressure. The resulting brown residue was resuspended in 1 mL of acetone, to which hexane was added until it became cloudy. Amber colored crystals formed on standing overnight in the cold. At this point the isolation procedures diverged as follows: DIBOA crystals were quickly washed with a small volume of acetone, resuspended in acetone, and precipitated a second time with hexane. The resulting crystals were collected and resuspended in methanol to which activated charcoal was added to remove impurities. The charcoal was removed by centrifugation leaving the purified DIBOA in methanol which was stored at -20C. DIMBOA crystals were collected and washed with a small volume of a chloroform:methanol mixture (9:1 v/v), suspended in acetone, and precipitated again by hexane. The fine white crystals were resuspended in methanol and stored at -20°C. Glucosides were purified (1) by freezing the tissue with liquid nitrogen in a mortar and grinding it to a powder from which the glucosides were extracted with acetone. The debris was removed from the extract by centrifugation (26,000g, 20 min), and the supernatant evaporated leaving a thick yellow suspension in water. This suspension was spotted on freshly prepared silica gel G thin layer plates and developed in ethyl acetate:methyl ethyl ketone:formic acid:water (50:30:5:10, v/ v/v/v). A portion ofthe chromatogram was sprayed with 10% (w/v) ferric chloride in methanolic HCl (1%) after which DIBOA glucoside (RF: 0.25) and DIMBOA glucoside (RF: 0.30) were identified as intense blue bands. Silica gel at corresponding unstained portions of the plate was removed, and the glucosides were extracted with methanol and stored at -20°C.
Assay of Standards and Enzyme Products A Waters3 model 244 high-performance liquid chromatograph equipped with a C18 ,uBondapak column and a model 440 absorbance detector was used to quantitate standards and reaction products. Samples were applied to the HPLC in methanol or the upper phase (methanol:water) of the Folch extraction solvent mixture (5). Samples containing DIBOA and its glucoside were eluted from the column with methanol:acetic acid:water (15:10:75) and samples containing I Mention of trademark, proprietary product, or vendor does not constitute a guarantee or warranty by the U.S. Department of Agriculture or the University of Missouri and does not imply approval to the exclusion of other products or vendors that may also be suitable.
Plant Physiol. Vol. 90,1989
DIMBOA and its glucoside were eluted from the column with methanol:acetic acid:water (20:10:70). Column eluants were monitored at 280 nm. Aglycone concentrations were determined using established extinction coefficients of E254 = 8500 for DIBOA and E262 = 10,000 for DIMBOA (9, 26). DIBOA glucoside concentrations were determined using an extinction coefficientof E255 = 8500 (9). DIMBOA glucoside was quantitated by converting it to the aglycone with fl-glucosidase and quantitating the aglycone (10) as described above. Enzyme Extraction and Purification
CI3 1 A seedling tissue (50-60 g) was homogenized 60 s in a precooled Waring blender with the addition of added Polyclar AT (0.3 g/g tissue) and 200 mL of extraction buffer (50 mM Hepes, 5 mM DTE [pH 7.5]). The debris was removed by filtering through Pellon followed by centrifugation (26,000g, 10 min). A portion of the crude fraction was retained for assay and the remainder precipitated with ammonium sulfate (30-60% saturation). The precipitate was collected by centrifugation (26,000g, 10 min) and resuspended in 3 mL extraction buffer. The ammonium sulfate fraction was initially passed over a Sephadex G-200 superfine column (1.5 x 20 cm) which had been equilibrated in 20 mm Hepes, (pH 7.5) at 8°C. The protein was eluted with the equilibration buffer at a flow rate of 15 mL/h and fractions of 2.3 mL were collected and assayed for activity as described below. Active fractions from the Sephadex G-200 were pooled and loaded onto a Q-Sepharose ion exchange column (1 x 10 cm) which had been equilibrated and washed in 20 mM Hepes buffer (pH 7.5). The protein was eluted from the column with a 0.10 to 0.25 M KCI gradient at a flow rate of 40 mL/h and fractions of 2.4 mL were collected and assayed for activity as described below. Mol Wt Estimation
The G-200 column previously described was standardized for mol wt determination using the marker proteins gamma globulin (160,000), albumin (67,000), and myoglobin (17,800). The ammonium sulfate fraction was applied to the column and total glucosyltransferase activity was determined for each fraction (1 mL). Enzyme Assay Unless otherwise specified, the assay mixture included 20 tsg of protein, 1.06 mm UDPG, 372 ,M DIMBOA, and Hepes buffer (50 mM, 5 mM DTE [pH 8.2]) in a volume of 200 ,uL. Samples were incubated at 37°C for 5 min, at which time reactions were terminated by the addition of 0.8 mL of chloroform:methanol (2:1, v/v) + 1 % HC1. This Folch partitionment (5) extracted almost 100% of the glucoside into the upper phase as determined by HPLC analysis. The ammonium sulfate fraction was used to determine optima for pH, temperature, and incubation time for the reaction. The pH optimum for enzyme activity was determined using a mixture of Hepes, Ches, Mes, Bicine, each at
HYDROXAMIC ACID GLUCOSYLATION IN MAIZE
50 mm plus 5 mM DTE. The pH was varied by 0.5 units from 6.5 to 10.0 with additional samples at pH 8.25 and 8.75. The temperature optimum was determined by assaying enzyme activity over a range of 25 to 55C. Assay mixtures were preincubated for 5 min before initiation of the reaction by the addition of UDPG. A suitable time of incubation was determined by varying incubation time from 0.5 to 10 min. The inhibitory effect of NEM (1 mM) and EDTA (5 mM) on the DIMBOA glucosyltransferase was assayed using the ammonium sulfate fraction. N-Eyhylmaleimide was added to the assay mixture with and without added DTE and EDTA without added Ca2". The effects of the reducing agents DTE, ascorbate, glutathione, and 2-mercaptoethanol and the divalent cations Ca2", Cu2+, Fe2+, Mg2+, and Zn2+ on glucosyltransferase activity were determined using the Sephadex G-200 fraction. Enzyme assays were of 1 min duration and included 10 ug of protein and 5 mm reducing agent and added cation. The hydroxamic acids DIMBOA and DIBOA, the flavonoid quercetin, and the coumarin derivative esculetin were considered as possible substrates for glucosylation using the Q-Sepharose fractions. DIMBOA, DIBOA, quercetin, and esculetin were included in the assays at concentrations of 1.08, 0.97, 0.83, and 1.0 mM, respectively. Isoquercetin, the glucoside of quercetin, was identified and quantitated as previously described (6), and esculin, the glucoside of esculetin, was identified and quantitated using the same system as previously described for DIBOA. Michaelis-Menten constants (Km) were determined by the intercept replot method (18) where UDPG was used as the glucose doner with either DIBOA or DIMBOA using the Q-Sepharose fractions. The substrate levels for UDPG, DIMBOA and DIBOA ranged from 0.067 to 1.0 mm. The initial intercepts were plotted against the reciprocal of the substrate concentrations to determine the respective Km values.
Developmental Curve Glucosyltransferase activity and DIMBOA concentrations were determined for the inbreds CI31A, Mol7, and OH43 using dry seed and 1 to 5 d old seedlings. A seedling consisted of the developing roots and shoots minus the scutellum. Individual seedlings were ground in a mortar in Hepes buffer (50 mM, 5 mM DTE [pH 7.5]) and filtered and centrifuged to yield a supernatant that contained the enzyme and the DIMBOA glucosides. One portion of the supernatant was permitted to stand for 15 min to allow hydrolysis of the DIMBOA glucoside by in situ f3-glucosidases. This latter portion of the supernatant was then acidified to pH 2.0 with HCI and extracted with 5 mL of ethyl acetate. The ethyl acetate was evaporated in a boiling water bath under nitrogen and the residue suspended in methanol and the DIMBOA concentration determined by HPLC. Enzyme activity is expressed as nmol of product formed per min and specific activity as nmol of product formed per min per mg of protein, or per seedling in the developmental experiments. Protein concentration was determined by the Bradford method (2) using bovine serum albumin (fraction V) as a standard.
1 073
RESULTS Identification of Standards
The identity of DIBOA and DIMBOA was verified by several criteria including reaction with FeCl3 to produce a blue complex and the UV spectra observed. The spectra for DIBOA, Xmax(H20) 253 and 280 nm, and for DIMBOA, Xmax(H20) 262 nm and a shoulder at 285 nm, are comparable to published data (9, 22). When subjected to HPLC, one major peak was observed for DIBOA yielding a k' of 1.33 and one for DIMBOA with a k' of 2.01. Heating DIMBOA at 100°C for 15 min resulted in the production of a compound which eluted at the same time as commercially available MBOA which is the breakdown product of DIMBOA. When the major peaks were collected and exposed to direct probe mass spectrometry analysis, the ionization patterns were as follows: DIBOA mle 181 (36%), 135 (57%), 108 (42%), and 79 (100%), and DIMBOA mle 211 (10%), 195 (42%), 193 (9%), 165 (100%), and 150 (40%), which correspond to the expected respective structures. The glucosides were identified based on their reaction to form a blue complex with FeCl3 on TLC, their UV spectra which are similar to the aglycones, and our ability to react the compounds with ,B-glucosidase to release the aglycones. The glucosides of DIBOA and DIMBOA eluted from the HPLC column just prior to their aglycones having capacity factors of 1.0 and 1.7, respectively. Enzyme Extraction and Purification Column chromatography resulted in a 13-fold purification of the enzyme (Table I). The Sephadex G-200 column maintained activity while increasing the specific activity five-fold, whereas the Q-Sepharose column doubled the specific activity but only half of the applied activity was recovered. The activity eluted from the Q-Sepharose column in two peaks (peak 1 at 0.17 M and peak 2 at 0.22 M KCI) that were of near equal total activity (Fig. 2). Mol Wt Estimation
The coefficient of determination for the mol wt standard In repeated determinations the glucosyltrans-
curve was 0.996.
Table I. Purification of Uridine Diphosphoglucose:DIMBOA Glucosyltransferase from Maize Seedlings
Purification Crude Ammonium sulfate
Activt
Specific Activity Enrichment Yield
units"
units/mg
X-fold
9660 8460 7020
96.0 132.0 510.0
1.0
protein
1.4 5.3
%
100 88 73
Sephadex G-200 Sepharose Q 11.4 18 1028.0 Peak 1 (0.17M KCI) 1740 14.1 17 1362.0 Peak 2 (0.22 M KCI) 1680 12.7 35 1218.0 3420 Peak 1 + peak 2 a A unit of activity is equal to the transformation of 1 nmol of substrate per minute.
Plant Physiol. Vol. 90, 1989
BAILEY AND LARSON
1 074
0.3 2 2.E
5 10 15 20 25 30 35 4045 0.2o0 -
E 20
C:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. 3
0
6
8
10 12 14 16 18 20
Fraction number
0
0
Figure 2. 0-Sepharose chromatography. KCI gradient (A); Elution profile (B), (O-O), glucosyltransferase activity; (.), protein concentration.
5 10 15 20 25 30 35 40 45 Elution volume (ml)
ferase activity eluted as a single peak from Sephadex G-200 with an estimated mol wt of 50,000 (Fig. 3).
Figure 3. Mol wt determination by Sephadex G-200 filtration. A, Mol wt (MW) estimation; a, -y-globulin (160,000); b, albumin (67,000); c, glucosyltransferase (estimated MW 50,000); d, myoglobin (17,800). ), proB, Elution profile; (@-@), glucosyltransferase activity; ( tein concentration.
Optimum Assay Conditions The pH optimum for the transferase was found to be 8.5 and to decrease to 70% of optimum as the pH was adjusted to 8.0 and 9.0. Activity was 10% of optimum at pH 6.5 and 30% of optimum at pH 10.0. The temperature optimum for enzyme activity was 45°C and decreased to 88% of optimum for 37 and 55°C and 60% of optimum at 25C. Activity responded linearly with time for 5 min (activity = 2.0 nmol/ min) and showed only a slight decrease at 10 min (activity = 1.7 nmol/min). Enzyme extracted without reducing agent was totally inactive. The inclusion of NEM (1 mM) in the assays minus DTE resulted in a 96% decrease in activity (Table II), but when DTE was included in the assay, the inhibitory effect of NEM was totally alleviated. Addition of the cation scavenger, EDTA resulted in a decrease of 16% in activity. The addition of Mg2" or Ca2" to the assay enhanced enzyme activity approximately 30% (Table II). The optimum concentration for Ca2" was determined to be 7.5 mm. The remaining metal cations reduced activity with Fe2" and Cu2" showing almost total inhibition of the activity. All the reducing agents that were assayed enhanced activity with DTE providing the greatest benefit (96%) and ascorbate the least (42%). Enzyme eluting from the Q-Sepharose column in peak 1 had activities of 8.08, 0.29, 0.21, and 0 nmol/min for the substrates DIMBOA, DIBOA, quercetin, and esculetin, re-
Table II. Influence of Inhibitors, Cations, and Reducing Agents on Uridine Diphosphoglucose:DIMBOA Glucosyltransferase from Maize Seedlings Addition
EDTA(-CaCI2) NEM (+DTE) NEM (-DTE)
CaCl2 CuCI2 FeSO4 MgCI2
ZnCl2 Ascorbate DTE Glutathione ,B-Mercaptoethanol
Concentration mM
Activity % control
5.0 1.0 1.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0
83 102 4
131 3 5 130 45 142 196 192 179
spectively. Enzyme eluting in peak 2 had activities of 6.41 and 3.67 nmol/min on DIMBOA and DIBOA, respectively, and had no detectable activity on quercetin and esculetin. Enzyme Kinetics The Km values for DIMBOA were similar for the enzymes
HYDROXAMIC ACID GLUCOSYLATION IN MAIZE
from the two Q-Sepharose peaks. The Km values for DIMBOA were 217 and 174 ,M for peak 1 and peak 2, respectively, and the Km values for UDPG were 286 glM for peak 1 and 200 ,M for peak 2. The peak 2 enzyme had a Km of 638 gM for DIBOA and 280 uM for UDPG when DIBOA was the hydroxamic acid substrate. The activity of peak 1 on DIBOA was too low to determine a valid Km value. Developmental Curve
The enzyme activity and DIMBOA accumulation patterns were similar in all three inbreds evaluated; such results for the inbred OH43 are presented in Figure 4. Total activity per seedling increased steadily reaching a high of 154 nmol/min at 4 d. DIMBOA accumulated steadily from 2 through 5 d, whereas seedling fresh weight increased at a progressively higher rate beginning with imbibition and continuing through the 5 d study.
DISCUSSION The presence of DIMBOA and its contribution to resistance to the corn borer in the young maize plant is a long-established fact (13). The existence of DIMBOA in the glucosidic form was established in 1959 (25). However, little information about the biosynthesis of these compounds has been established since that time. This report is concerned with the
extraction and partial purification and characterization of enzymes that catalyze the addition ofglucose at the 2-position of DIMBOA. Two distinct enzymes are identified based on their separation by Q-Sepharose chromatography and substrate specificity differences. Our data confirm the presence of DIMBOA and its glucoside in the maize plant and relate the accumulation of these compounds to the development of the young maize plant. Assay of the enzymes was complicated by the instability of DIMBOA at the optimum pH and
~~~~x
t /
Z150
XII10S 102z
wB1001.
I'
(/)
/ /106 Q. 2z /
>-100- E E
w ~~~< ~~~0 1U) ~~
