Abstract--The fate of the tomato foliar phenolic, chlorogenic acid, in the digestive systems of Colorado potato beetle Leptinotarsa decemlineata. (Coleoptera: ...
Journal of Chemical Ecology, Vol. 18, No. 4, 1992
A V O I D A N C E OF A N T I N U T R I T I V E P L A N T D E F E N S E : R O L E OF M I D G U T pH IN C O L O R A D O P O T A T O BEETLE 1
G.W. FELTON,* J. WORKMAN, 2 and S.S. DUFFEY 2 Department of Entomology University of Arkansas Fayetteville, Arkansas 72701 (Received October 4, 1991; accepted November 25, 1991) Abstract--The fate of the tomato foliar phenolic, chlorogenic acid, in the digestive systems of Colorado potato beetle Leptinotarsa decemlineata (Coleoptera: Chrysomelidae) and Helicoverpa zea (Lepidoptera: Noctuidae) is compared. In larval H. zea and other lepidopteran species previously examined, approximately 35-50% of the ingested chlorogenic acid was oxidized in the digestive system by foliar phenolic oxidases (i.e., polyphenol oxidase and peroxidase) from the tomato plant. The oxidized form of chlorogenic acid, chlorogenoquinone, is a potent alkylator of dietary protein and can exert a strong antinutritive effect upon larvae through chemical degradation of essential amino acids. In contrast, in L. decemlineata less than 4% of the ingested dose of chlorogenic acid was bound to protein. In vitro experiments to determine the influence of pH on covalent binding of chlorogenic acid to protein showed that 30-45% less chlorogenic acid bound to protein at pHs representative of the beetle midgut (pH 5.5-6.5) than at a pH representing the lepidopteran midgut (pH 8.5). At an acidic pH, considerably more of the alkylatable functional groups of amino acids (--NH2, --SH) are in the nonreactive, protonated state. Hence, polyphenol oxidases are unlikely to have significant antinutritive effects against the Colorado potato beetle and may not be a useful biochemical source of resistance against this insect. The influence of feeding by larval Colorado potato beetle on foliar polyphenol oxidase activity in tomato foliage and its possible significance to interspecific competition is also considered. Key Words--Leptinotarsa decemlineata, Coleoptera, Chrysomelidae, Colorado potato beetle, Helicoverpa zea, Lepidoptera, Noctuidae, phenolics, poly* TO whom correspondence should be addressed. Approved by the Director of the Arkansas Agricultural Experiment Station. 2Department of Entomology, University of California, Davis, California 95616. 571 0098-0331/92/04004)571506.50/0 9 1992PlenumPublishingCorporation
572
FELTON ET AL. phenol oxidase, peroxidase, chlorogenic acid, Lycopersicon esculentum, tomato, host-plant resistance, midgut pH.
INTRODUCTION
Feeding by the Colorado potato beetle, Leptinotarsa decemlineata, is normally restricted to solanaceous plants including potato, tomato, and eggplant (Hsiao, 1974, 1978; Hare, 1990). Because of the considerable damage this insect may cause to these crops, substantial efforts have been made to develop host-plant resistance to the beetle; however, commercially acceptable varieties have not been released (Hare, 1990). Several phytochemicals have been specifically implicated in resistance to the Colorado potato beetle (CPB), including steroidal glycoalkaloids (Hare, 1983, 1987; but see Barbour and Kennedy, 1991), methyl ketones (Kennedy and Sorenson, 1985) and the sesquiterpene, zingiberine (Carter et al., 1989). Although CPB causes a massive induction of serine protease inhibitors in tomato leaves during feeding (Green and Ryan, 1972), the beetle is insensitive to their adverse effects (Wolfson and Murdock, 1987) because it relies not upon serine proteases such as trypsin and chymotrypsin, but instead upon the thiol proteases, cathepsins B and H, for protein digestion (Murdock et al., 1987; Thie and Houseman, 1990). Another important component of chemical defense in solanaceous plants is the enzymic oxidation of phenolics. In the wild potato species, Solanum berthaultii, type A glandular trichomes containing polyphenol oxidase are essential for resistance against CPB (Neal et al., 1989; Tingey, 1991). Moreover, type B trichomes containing sucrose esters increase the expression of resistance in the presence of type A trichomes (Neal et al., 1989; Pelletier and Smilowitz, 1990; Tingey, 1991). A significant amount of the total leaf polyphenol oxidase is associated with the chloroplast in potato and tomato. The majority of the work on phenolic oxidation in potato has focused on the entrapping properties of trichomes; little work has been done on the potential antinutritive properties of oxidized phenolics as a basis of resistance to CPB. Carter (1987) reported that resistance to CPB in tomato species was not associated with total foliar phenolic oxidation. Our paper compares the ability of an antinutritive plant defense (i.e., polyphenol oxidase) to operate against two insects, CPB and Helicoverpa (=Heliothis) zea, with distinctly different gut physiologies, the former having an acidic midgut and the latter an alkaline midgut. In the highly polyphagous lepidopteran species, H. zea and Spodoptera exigua, phenolic oxidation exerts a strong antinutritive effect against larvae (Felton et al., 1989a,b). When larvae disrupt plant cellular compartments during feeding, the enzyme polyphenol oxidase mixes with phenolic substrates such as chlorogenic acid, leading to the rapid formation of orthoquinones (Felton et al., 1989a). The quinones readily
BEETLE MIDGUT pH, AVOIDANCE OF PLANT DEFENSE
573
alkylate dietary protein, resulting in a substantial decline in the nutritive quality of the protein (Felton et al., 1989a,b; Duffey and Felton, 1989, 1991). The lepidopteran species tend to feed on fruiting structures, where polyphenol oxidase activity is markedly lower than in foliage, thereby avoiding a significant impact of phenolic oxidation upon food quality (Felton et al., 1989a). In contrast, CPB feeds primarily on foliage and hence is potentially exposed to high levels of phenolic oxidation products. We report here an assessment of the potential for polyphenol oxidase to operate as an antinutritive source of resistance to CPB. The role of midgut pH in mediating the impact of phenolic oxidation on food quality is also considered.
METHODS AND MATERIALS
lnsects and Plants. Eggs of CPB were obtained from a colony maintained on potato in the laboratory of Dr. George Kennedy, North Carolina State University. Emerging larvae were reared in environmental chambers on tomato, Lycopersicon esculentum (var. Castlemart) for one generation before experiments were begun. Eggs of H. zea were obtained from the Bioenvironmental Insect Control Laboratory, USDA (Stoneville, Mississippi). Plants used for feeding and induction assays were grown following Broadway et al. (1986). Chemicals and Isotopes. Tritiated chlorogenic acid was prepared by tritium exchange by Research Products International Corp. (Gif-sur-Yvette, France) and purified using thin-layer chromatography conforming to Isman and Duffey (1983). The specific activity of the purified [3H]chlorogenic acid was 153 mCi/ mmol. Enzyme Assays. Preparation of leaf tissue for enzyme analysis has been described previously (Felton and Duffey, 1990). Peroxidase and polyphenol oxidase activities were measured following Ryan et al. (1982) using guaiacol and chlorogenic acid as substrates, respectively. Leaf protein was measured using Coomassie blue G-250 dye following Jones et al. (1989). Metabolism of Chlorogenic Acid in Larval Digestive System. Two-microliter aliquots of [3H]chlorogenic acid containing ca. 205,000 dpm (ca. 215 ng) in 50 % aqueous methanol were applied to the surfaces of tomato leaflets weighing ca. 50 mg. After the aqueous methanol evaporated, fourth-instar CPB or fifth-instar H. zea larvae that had been starved for 12 hr were placed on individual leaflets containing the radioisotope. Larvae completely ingesting the leaflet within 6 hr were used in the assay. Feces were collected as described (Felton et al., 1989a) from nine larvae per species and pooled in groups of three larvae for each replicate. A total of three replicates per species was analyzed by twodimensional thin-layer chromatography (2 % formic acid followed by n-butanolethanol-2% NHaOH at 6 : 1 : 3 on cellulose) followed by liquid scintillation
574
FELTON ET AL.
counting for radioactivity associated with free chlorogenic acid, protein-bound chlorogenic acid, and polymeric chlorogenic acid (Felton et al., 1989a). Oxidative Enzymes in Larval Midgut. To determine the presence of phenolic oxidizing enzyme activity in the beetle digestive system, five newly molted fourth-instar larvae were placed on an undamaged six- to seven-leaf stage tomato plant. Larvae were removed after 24 hr and held on ice for 30 min. Midguts were removed, and each midgut, containing lumen contents, was mechanically homogenized in ice-cold 1.15 % KC1 for 30 sec. Samples were centrifuged at 10,000g for 20 min; the supernatant was used immediately as the enzyme source for peroxidase and polyphenol oxidase. One leaflet was excised with a razor blade from each leaf of the tomato plant and assayed for peroxidase and polyphenol oxidase activities as described above. The experiment was replicated five times. Protein concentration was quantified using Coomassie brilliant blue G-250 following Jones et al. (1989). Determination of pH in Larval Digestive System. Twenty fourth-instar CPB and 20 fifth-instar H. zea were placed individually in Petri dishes containing two tomato leaflets excised from six-leaf stage tomato plants. After 24 hr, regurgitate was individually collected from 10 larvae of each species for pH determination. Midgut luminal fluid was individually collected from the remaining 10 larvae per species following dissection of the digestive tract. The pH of the regurgitate and midgut fluid was determined with short-range pH paper. Influence of pH on Polyphenol Oxidase and on Covalent Binding of Chlorogenic Acid to Protein. To determine the influence of pH on polyphenol oxidase activity and the subsequent binding of chlorogenic acid to protein, tomato polyphenol oxidase was partially purified by ammonium sulfate fractionation to a specific activity of 2300 units/mg protein (Fetton and Duffey, 1991a). One unit of activity equals a change in OD470of 0.001/min with chlorogenic acid as a substrate. A 1-ml solution containing 20 mg bovine serum albumin and 3.0 /zmol chlorogenic acid was incubated with 30/~g polyphenol oxidase in 0.1 M sodium phosphate buffer at pH 5.5, 6.5, 7.5, or 8.5 for 60 min. Three replicates were performed at each pH. The polyphenol oxidase activity for each pH level was determined from 10-#1 aliquots performed in triplicate (Ryan et al., 1982). Control treatments for each pH were treated identically except chlorogenic acid was omitted. The solutions were then dialyzed (12,000 mol wt cutoff) at 1-4~ against 8 M urea for 24 hr to remove noncovalently bound chlorogenic acid (Barbeau and Kinsella, 1983), followed by dialysis against distilled H2O for 24 hr, and finally desalted in a PD-10 Sephadex G-25M column (Pharmacia LKB, Uppsala, Sweden) according to the manufacturer's suggestions. The amount of chlorogenic acid covalently bound to the protein was determined from a standard curve using a SLM-Aminco 3000 Array spectrophotometer following Broadway et al.
BEETLE MIDGUT pH, AVOIDANCE OF PLANT DEFENSE
575
(1986). The small amount of background absorbance in the respective control treatments was subtracted from each test protein. Effect of Colorado Potato Beetle Feeding on Leaf Phenolic Oxidases. To determine the effect of CPB larval feeding on polyphenol oxidase activity, tomato plants (var. Castlemart) were grown in the greenhouse in fine mesh cages to exclude unwanted insects. When plants reached the six- to seven-leaf stage, one fourth-instar larva was placed on each plant on the lowest true leaf. No effort was made to restrict larval feeding to a particular leaf position. Control plants were treated identically throughout with the exception of not being infested with beetle larvae. Seven plants per treatment were maintained in the cages during the experiment. After seven days, plants were transported to the laboratory where six leaflets (one terminal leaflet per leaf) per plant were removed for analyses. Leaflets included both damaged and undamaged samples. Leaflets for each replicate plant were pooled, homogenized, and assayed in triplicate for polyphenol oxidase and peroxidase as described above. RESULTS
Metabolism of Chlorogenic Acid in Larval Digestive System. The majority of excreted chlorogenic acid in the CPB was not covalently bound to protein but remained in unbound, monomeric form (Table 1). Less than 4% of excreted chlorogenic acid was covalently bound to protein and/or amino acids, but a significant amount (ca. 16%) of excreted chlorogenic acid was polymerized. In contrast, considerably more excreted chlorogenic acid was bound to protein and/ or amino acids in larval H. zea: nearly 35 % was covalently bound, and approximately 14% was polymerized. Polymerized chlorogenic acid remained at or TABLE 1. FATE OF EXCRETED [3H]-CHLOROGENIC ACID IN FECES OF COLORADO POTATO BEETLE Leptinotarsa decemlineata AND TOMATO FRUITWORM
Helicoverpa zea LARVAE Percent of total excreted radioactivity" Chemical
CPB
H. zea
Chlorogenic acid (unbound) Chlorogenic acid-protein conjugate Chlorogenic acid (polymer)
82.9 (+12.2) A 3.5 (+0.9) A 16.3 (_+4.4) A
34.1 (+7.5) B 34.8 (+7.6) B 13.7 (_+2.6) A
aMeans in rows followed by a different letter are statistically different at P < 0.05 by ANOVA and separation of means by 95 % confidence limits. Each value represents the mean of three replicates. Numbers in parentheses represent 95 % confidence limits.
576
FELTON ET AL.
near the origin in the TLC systems tested. Chlorogenic acid chemically oxidized with 0.1 N NaOH or with ferrous sulfate at pH 8.0 behaved similarly. Oxidative Enzymes in Larval Midgut. Substantial polyphenol oxidase activity was present in the midgut contents in CPB larvae feeding on tomato foliage (Table 2). Activity was approximately 36% of that found in tomato foliage used as the food source. It is unknown if the midgut contains endogenous polyphenol oxidase or is merely derived from ingested foliage; nevertheless, ingested polyphenol oxidase activity is reduced in the gut of CPB. Peroxidase activity in the midgut contents was nearly 12 • in excess of that present in ingested leaflets (Table 2). The enriched peroxidative activity of the midgut with respect to the food plant implies the presence of endogenous peroxidase. pH of Larval Digestive System. The pH of the CPB regurgitate ranged from 5.5 to 6.0, whereas the regurgitate from H. zea was 7.0-7.4. The beetle's midgut fluid was acidic at pH 6.0-6.5, but the caterpillar's midgut was alkaline, ranging from 8.0 to 8.7.
Influence of pH on Polyphenol Oxidase Activity and Covalent Binding of Chlorogenic Acid to Protein. Both polyphenol oxidase from tomato foliage and covalent binding of chlorogenic acid to protein were significantly affected by pH (Figure 1). Optimal oxidase activity occurred at pH 6.5 within the pH range of the beetle midgut lumen, but at pH 8.5, representative of the lepidopteran midgut, polyphenol oxidase activity was nearly 50% less. However, maximal covalent binding was observed at the alkaline pH 8.5. At pHs occurring in the digestive system of CPB (i.e., 5.5-6.5), ca. 25-40% less covalent binding occurred than at pH 8.5 (Figure 1). Effect of Colorado Potato Beetle Feeding on Leaf Phenolic Oxidases. Feeding by CPB for 48 hr on tomato foliage did not significantly (P > 0.05) affect foliar peroxidase activity (Figure 2). However, polyphenol oxidase activity in foliage increased by 92% when feeding damage by CPB occurred (Figure 2; P < 0.01).
TABLE2.
PHENOLIC OXIDIZING ENZYMES a IN TOMATO FOLIAGE AND MIDGUT OF COLORADO POTATO BEETLE
Source
Polyphenoloxidase
Peroxidase
Tomato foliage Beetle midgutb
3.940 (_0.122) 1.432 (+0.050)
7.481 (+0.078) 86.420 (+5.94)
aEnzyme activities expressedas change in OD470/min/mg protein. Numbers in parentheses represent 95% confidencelimits of the mean with five replicationsper tissue. Polyphenoloxidase measured with chlorogenicacid as a substrate and peroxidasewith guaiacol. bMidgut measurementsmade with midgut wall and lumen contents.
BEETLE MIDGUT pH, AVOIDANCE OF PLANT DEFENSE
577
A
A
100 /
90
"~. "-.
/
80 D.
Bound C H A
\'-,
/ /
\.
70
.> .4.,I
60
n-
50 40 30
//"
~
./
i
5.50
PPO
I
I
I
6.50
I
7.50
I
I
8.50
pH FIG. 1. Influence of pH on tomato foliar polyphenol oxidase (PPO) activity and on the covalent binding of chlorogenic acid (CHA) to protein. Each point represents the mean of three replicates. Error bars indicate 95 % confidence limits. Means on each line not followed by the same letter are significantly different at P < 0.05 by 95 % confidence limits. The absolute values for 100% relative polyphenol oxidase activity and 100% covalent binding are respectively, 2300 units/mg protein/rain and 144.4 nmol chlorogenic acid bound/mg protein. DISCUSSION
The fate of the foliar phenolic, chlorogenic acid, in the digestive system of the Colorado potato beetle was strikingly different from its fate in lepidopteran species (Felton et al., 1989a; Felton and Duffey, 1991a). In the Colorado potato beetle, the majority of ingested chlorogenic acid remained unoxidized in the beetle's digestive system (ca. 80%); however, a substantial amount (16%) was oxidatively polymerized. A small amount of chlorogenic acid ( < 4 % ) was covalently bound to amino acids and/or proteins in the beetle digestive system. In contrast, in the lepidopteran species H. zea, Spodoptera exigua, Manduca sexta, and Trichoplusia ni, approximately 35-50% of excreted chlorogenic acid was covalently bound to protein or amino acids (Felton et al., 1989a; Felton and Duffey, 1991a, unpublished data). The differences in phenolic-protein binding among these insects are largely explicable by differences in midgut pH. The midgut luminal pH in the lepidopteran species was moderately to highly alkaline (i.e., pH 8.0-11.0; Felton et al., 1989a; Felton and Duffey, 1991a; Dow, 1984), but the pH of the digestive fluids of the beetle was mildly acidic (i.e., pH 5.5-6.5; see also Wolfson and Murdock, 1987). In an acidic midgut, chlorogenic acid is less likely to be covalently bound to amino acids and proteins for two reasons. First, the nonen-
578
FELTON ET AL.
300 250 > ~
200
]Control B
BeeUe Damage
(J
, N fU.I
150 100 50 0
Polyphenol Oxidase
Peroxidase
FIG. 2. Effect of foliar feeding by Colorado potato beetle larvae on activities of tomato polyphenol oxidase and peroxidase. Enzyme activity expressed in units of change in OD47o/min/gm foliage with guaiacol as substrate for peroxidase and chlorogenic acid as substrate for polyphenol oxidase. Controls represent undamaged plants. Error bars indicate 95 % confidence limits. Each treatment was replicated seven times. Activities of enzymes were determined after seven days feeding by one fourth-instar larva. Polyphenol oxidase activity was significantly higher (P < 0.01) in plants damaged by beetle larvae. No significant differences were observed in peroxidase activity.
zymic oxidation (i.e., autooxidation) of chlorogenic acid is minimized at acidic pH (Barbeau and Kinsella, 1983; CiUiers and Singleton, 1989). Second, the chemical reactivity of the nucleophilic functional groups of amino acids (--SH and - - N H or --NH2) is strongly influenced by pH. In alkaline conditions, many amino acid side chains are unprotonated and, consequently, more prone to nucleophilic attack by electrophiles (e.g., quinones, epoxides); on the other hand in acidic pH ( < 6.5), the side chains become protonated and hence less susceptible to alkylation (Pierpoint, 1983; Barbeau and Kinsella, 1983). For instance, the pKas of the reactive --SH, --NH, and --NH2 groups of cysteine, histidine, and lysine, are respectively 8.1, 8.8, and 9.0. Consequently, at acidic pH, few of these groups are unprotonated and the oxidative polymerization of phenolics is favored, thereby minimizing the antinutritive effects of phenolics on protein. The phenolic polymer of chlorogenic acid possesses negligible toxicity to lepidopteran species (Felton et al., 1989a); however, its toxicity to CPB has not been determined. The oxidative polymerization of phenolics by insect peroxidase and phenol oxidases has been hypothesized to be a detoxicative process in certain insects (e.g., aphids; Peng and Miles, 1988). It is conceivable that in insect species possessing an acidic midgut, peroxidase and polyphenol oxidase activities (Table
BEETLE M I D G U T pH, A V O I D A N C E O F P L A N T DEFENSE
579
2) contribute to detoxification of chlorogenic acid. However, it must also be noted that the midgut of the CPB presumably possesses a negative oxidation potential because of the necessity for thiol digestive proteases to be in a reducing environment (Murdock et al., 1987; Thie and Houseman, 1990). A reducing environment minimizes phenolic oxidation (Appel and Martin, 1990) and may partially account for our observations that substantially less total phenolic oxidation (i.e., polymer formation and protein-bound phenolic) occurs in CPB compared to the lepidopteran species, despite the fact that foliar polyphenol oxidases in tomato are highly active at the pH of the CPB midgut (Felton et al., 1989a). The relatively large differences in the pH of the digestive system between the leaf-feeding beetle, CPB (mildly acidic, ca. 5.5-6.5), and the leaf-feeding lepidopterans, Manduca sexta and Trichoplusia ni (highly alkaline, > 9.5) may represent two divergent physiological mechanisms for tolerating phenolics and their oxidation products. We have hypothesized that, although quinones are more reactive with amino acids in the highly alkaline midgut, the alkalinity would allow for greater solubilization of plant protein, thus compensating for any nutritional loss in essential amino acids due to conjugation of phenolics with amino acids (Felton and Duffey, 1991a). This appends the earlier interpretation of gut alkalinity as an adaptive mechanism to circumvent tannin toxicity through inhibition of hydrogen bond complexation (Berenbaum, 1980). Alternatively, an acidic midgut impedes the covalent conjugation of amino acids with the electrophilic quinones, thus preventing the loss of essential amino acids. Therefore, it follows that insects with mildly alkaline digestive systems (i.e., pH 89; H. zea and Spodoptera exigua) are predicted to be most susceptible to the antinutritive effects of phenolic oxidation. Hence, we predict that host-plant resistance based upon the antinutritive properties of polyphenol oxidase will not be particularly effective against insects possessing acidic midguts (e.g., many beetles and grasshoppers). The ability of an acidic midgut to inhibit phenolic-protein covalent interactions cannot fully account for the unexpectedly small amount of phenolicprotein conjugate observed in feces of the CPB (i.e.,
