AND OVERPRODUCING TOBACCO ANIONIC ... - PubAg - USDA

Journal of Chemical Ecology, Vol. 23, No. /0, /997

EXAMINATION OF DIFFERENT TOBACCO (Nicotiana spp.) TYPES UNDER- AND OVERPRODUCING TOBACCO ANIONIC PEROXIDASE FOR THEIR LEAF RESISTANCE TO He!icovelpa zeal

P. F. DOWD 2 ,* and L. M. LAGRIMINrJ 'Bioactive Agellls Research Unit, U.S. Department of Agriculture Agricultural Research Service, National Celller for Agricultural Utilization Research /8/5 N. University St., Peoria. fllinois 6/604 .JDepartll/elll of Horticulture and Crop Sciences Howlell Hall, 200/ Fyffe Court. 771e Ohio State University Columbus, Ohio 432/0

(Reeeived January 13. 1997: accepted June 12. 1997)

Abstract-First-instar larvae of the false tobacco budworm (com earwoml. Helicoverpa zeal that fed on either intact plants. leaf disks from undamaged plants. or leaf disks from insect-damaged plants of Nicotiano sylvestris and N. tabacIIIl/ Coker plants overproducing a tobacco anionic peroxidase generally caused significantly less damage than those caged with corresponding material from wild-type plants. In some cases mortality was significantly higher and weights significantly less for caterpillars feeding on leaf material from overproducing vs. wild-type plants. First-instar H. zea fed on the same type of leaf material from N. tabaclllll Xanthi underproducing tobacco anionic peroxidase generally caused significantly more damage than those fed leaf material from wild-type plants. However. first-instar H. zea fed on underexpressing leaf material from N. sylvestris did not cause significantly greater damage compared to wild-type material. In cases where peroxidase enzyme activity was determined. significantly higher mean peroxidase activity was seen in leaves of plant types that also had significantly less mean feeding mtings. This infomllltion suggests that peroxidase activity can contribute to leaf resistance to chewing insects. However. the context of the peroxidase (cooccurring substrJtes. additional inducible factors) can mediate the degree of influence seen by changes in individual peroxidase isozyme levels.

"To whom correspondence should be addressed. I Names are necessary to report factually on available data: however. the USDA neither guarantees nor warrants the standard of the product. and the use of the name by USDA implies no approval of the product to the exclusion of others that may also be suitable.

2357 009S-0J31'97ill}()O-2357/S 12.50/0

1997 Plenum Publishing CorporJtion

2358

DOWD AND L."GRIMINI

Key Words-Plant resistance. Helicol'elpa. false tobacco budwoml. Nicoliana. peroxidase. lignin. tmnsgenic plants. com earWOm1.

INTRODUCTION

Proteins are increasingly assigned a role in resistance to insects (Shah et aI., 1995). Most of these act as direct toxins, such as Bt crystal proteins and lectins (Gatehouse et aI., 1993). Others may act indirectly, such as protease inhibitors (Gatehouse et al., 1993, 1994). More recently, enzymes that interfere with insect processes, such as cholesterol regulation (Purcell et aI., 1993; Corbin et aI., 1994) or chitin fomlation (Cohen, 1993), have been described. Enzymes involved in host-plant responses to insects have also been investigated, and include polyphenol oxidases, peroxidases, and lipoxygenases (for review see Appel, 1993). The role of plant peroxidases in insect resistance is unclear. Material oxidized by peroxidases can be more toxic and/or less nutritious to insects (Felton et aI., 1989, 1992; Dowd, 1994; Dowd and Norton, 1995). However oxidation products that are thought to be produced by peroxidases can be less toxic than their precursors (Peng and Miles, 1988). Due to the complexity of interacting factors, predicting the role of peroxidase in a particular matrix is likely to be difficult (Appel, 1993; Miles and Oertli, 1993; Dowd and Lagrimini, 1997a). The production of plants (over- or under-) producing tobacco anionic peroxidase (Rothstein and Lagrimini, 1989; Lagrimini, 1992; Lagrimini et aI., 1990, 1992, 1993) allows for the more direct examination of the potential for peroxidase to promote resistance to insects. Earlier studies have demonstrated enhanced resistance of some excised plant tissues from tobacco or tomato (over)producing a tobacco anionic peroxidase to different insects compared to wild-type plants (Dowd and Lagrimini, 1997b). In some cases production of enhanced levels of foreign peroxidase in nonsource plant species has increased susceptibility to some insects (Dowd and Lagrimini, 1997c). Thus, a better understanding may be had if the level of a naturally occurring peroxidase isozyme were genetically altered in the source plant species. Earlier studies demonstrated resistance of excised mature stems of Nicotialla sylvestris overproducing tobacco anionic peroxidase to feeding by first instars of the false tobacco budwoml (com earwoml, Helicovelpa zea) compared to wild-type plants, but leaves of both plant types were equally susceptible to third instars (the only size tested) (Dowd and Lagrimini, 1997b). When seed became available for Nicotialla sylvestris overand underproducing tobacco anionic peroxidase, and for two varieties of N. tabacllm, one overproducing and the other underproducing tobacco anionic peroxidase (Lagrimini, 1992), it became possible to examine the activity in different, but closely related plants. Whole plant and leaf disk studies using caged first instars of the false tobacco budwoml (com earworm), HelicoVel]H/

PEROXIDASE IN TOBACCO LEAF RESISTANCE

2359

zea, were used to examine relative resistance of the different tobacco variants to this insect. METHODS AND MATERIALS

Insects. H. zea were reared on a pinto bean-based diet at 27 + 1°C. 40 ± 10% relative humidity, and a 14L: 100 photoperiod (Oowd, 1988). Newly hatched instars «24 hr old) were used in all assays. Plants. At least twice-selfed seed from single lines of the transgenic tobacco anionic peroxidase overproducer (Lagrimini et aI., 1990) and underproducer (Lagrimini, 1992) N. sylvestris, overproducer N. tabaclllll Coker (Lagrimini et aI., 1990), and underproducer N. tabacllm Xanthi (Lagrimini, 1992) was used. Seeds were planted in a previously described soil mix (Oowd and Lagrimini, 1997b). Plants were transplanted to 4-in.-diameter pots when two fully expanded true leaves were present (ca. 7-10 days). Plants were grown in an growth chamber (Environmental Growth Chambers G-IO) containing llO-W fluorescent lights (Phillips) and 40-W fan crystal light bulbs (General Electric). The tops of the pots were 50 cm from the lights of the growth chamber. Plants were grown at 27 ± 1°C, 40 ± 10% relative humidity, and a 14L: 100 photoperiod. After transplanting, whenever plants were watered, liquid fertilizer with chelated iron was used (Dowd and Lagrimini, 1997b). Intact Plant Bioassays. Cages were made to contain the small larvae on intact plants. Cages were made out of Solo clear plastic polyethylene cups with a 9-cm mouth diameter (10 oz.). The rolled lip of the cups was removed to facilitate pushing into the soil mix. For ventilation, ca. 2-cm squares were cut on opposite sides of the cup with the closest edge ca. 2 cm from the bottom of the cup. The holes were covered with organdy cloth secured with hot-melt glue. Plants were used for assays when they reached a size such that the largest leaf was ca. 3-5 cm. long (blade portion), which was approximately 20 days after germination. First-instars were added to the plant. the cup was inverted over the plant, and the top of the cup was pressed ca. I cm into the soil. Initially only two larvae per plant were used (with the first series of N. sylvestris). This number of larvae produced only limited damage, so 10 larvae per plant were used in subsequent assays. Plants remained in the growth chamber for one week and were then removed. The length of each leaf of a plant was measured, and the damage (in square millimeters) to each leaf was estimated. Initially the first four true leaves were not considered due to infrequent damage (with the first series of N. sylvestris) , but with the 10 larvae per plant used in subsequent assays, these leaves were often damaged and so were also included. Larvae present were removed, counted, and weighed. Plants were returned to the growth chamber for recovery so that individual leaves could be used in subsequent leaf

2360

DOWD AND LAGRlMINI

disk assays (see below). As many plants as possible that were of an equivalent size of the different types being examined at a particular time were used in assays, and the number ranged from 13 to 22 plants of a particular type. All plant types of a particular cultivar were examined at the same time. Leaf Disk Bioassays. For assays with disks from undamaged plants, leaf disks were removed from the newest leaf with a ca. 3 to 4-cm-Iong leaf blade. This size was reached approximately seven days after the leaf was initiated. In the first series of studies with N. sylvestris, both the younger leaves and leaves that were mature (second to third leaf from the bottom, approximately 14 days after the leaf initiation) were used. The maturity of the leaf influenced the amount of damage (see Results). Because age was initially determined to influence relative susceptibility of the leaf to caterpillar damage, it was important to select leaves of approximately the same maturity for assays. The younger leaves continued to be used due to their greater susceptibility to feeding. A ca. 2-cm disk was cut from the center of the leaf and used in bioassays. The disk was placed in a small Petri dish with a tight-fitting lid (Falcon 1006) along with 10 newly hatched H. zea larvae. The disk assay chambers were held under the same conditions used to rear the larvae except they were kept in the dark. Preliminary observations indicated cannibalism by the H. zea larvae was much lower if the leaves were kept in the dark. Assay chambers were examined for leaf feeding damage and caterpillar mortality after one and two days. Larvae were considered dead if they did not move when dishes were tapped. Mortalities were calculated as percentages of dead larvae of total larvae counted. Surviving larvae were weighed after two days using a Mettler AEI63 analytical balance with a IO-/lg accuracy. Damage to leaves was determined by two methods. When damage was generally less that II 10 of the entire leaf for a particular group, a rating of the number of equivalent 0.25-mm 2 holes (approximating the head capsule size of the first instars) was used; 75 of these units were approximately equivalent to 1110 of the leaf being consumed. When greater amounts of the leaf were consumed, a 1-10 scale was used. In cases where both degrees of damaged were seen in the same series, the conversion rate of 75 units = 1/ I0 of leaf consumed was used to convert values to those of the most prevalent rating unit. The number of leaves used depended on the number of plants available; typically 10-20 plants of a particular type were used. All assays with a particular plant cultivar were set up at the same time. Insect feeding has been shown to differentially induce defensive agents, including peroxidases (e.g., Stout et aI., 1994). Thus, it was of interest to test leaves from plants that had been damaged by caterpillars to see if the trend of resistance was the same as for leaves from plants that had not been damaged by the insects. The leaf used was the most terminal leaf from plants that had been damaged in the intact plant assays. After this leaf had expanded to usable


2361

size (mean of ca. 5-8 cm of leaf blade length, depending on the cultivar) approximately seven days after initial damage, it was used in another series of leaf disk assays. Leaf tissue was removed, assays set up, and assays evaluated as described for the undamaged plant leaf disk assays. Peroxidase Assays. Leaf disks I cm in diameter were removed from the same leaves of wild-type and underproducing N. tabacum Xanthi and wild-type and overproducing N. tabacum Coker as those used in leaf disk feeding assays prior to the start of the assay. They were assayed for peroxidase activity using guaiacol as a substrate by previously published procedures (Lagrimini et aI., 1987). Statistical Analyses. Mortality data was analyzed using chi-square analysis and leaf feeding ratings, weights, and enzyme activity were analyzed using analysis of variance (SAS Institute, 1985). RESULTS

Differential Feeding Responses on Intact Plants. The total area of leaf consumed by caterpillars in intact plant assays was always less for plants overproducing peroxidase compared to wild-type plants and was significantly less for whole plants in some cases (Table 1). Intermediately aged leaves of the different plant types showed the greatest difference in feeding where obvious differences were present. For example, the sixth leaf from the bottom of the overproducing plants from the second series of N. sylvestris was damaged significantly less (P < 0.05) than the corresponding leaf from the wild-type plants. The fourth leaf from the bottom of the N. tabacum Coker overproducing plants was damaged significantly less (P < 0.05) than the corresponding leaf from the wild-type plants. More tissue was consumed by caterpillars caged with underproducing plants of N. tabacum Xanthi compared to wild-type plants, while the opposite was true for underproducing N. sylvestris (although differences for all leaves combined were not significantly different in any case). The sixth leaf from the bottom of the wild-type N. tabacum Xanthi was damaged significantly less (P < 0.05) compared to the corresponding leaf from underproducing plants. Because only small numbers of caterpillars were recovered from plants at the end of the seven-day assay period, weights are not reported. Differential Feeding Responses in Leaf Disk Assays. In leaf disk assays, leaf disks from plants overproducing tobacco anionic peroxidase were always damaged less than those from wild-type plants to a significant degree in all cases except for mature N. sylvestris leaf disks (Table 2). There were few differences in mortality or weights of larvae feeding on leaf disks from these respective plants, although in some cases mortality was significantly greater, and weights significantly less, for caterpillars fed on disks from overproducing compared to

2362

TABLE

DOWD AND LAGRIMINI

I.

EFFECTS OF FEEDING BY FIRST-INSTAR

H. zea

ON LEAVES OF INTACT

PLANTS OF WILD-TYPE AND TRANSGENIC TOBACCOS OVER- AND UNDERPRODUCING TOBACCO ANIONIC PEROXIDASE"

Plant type

N. wbac/l1/1 Xanthi Underproducer Wild type N. wbaclIIlI Coker Wild type Overproducer N. syil'cs/ris set I Underproducer Wild type Overproducer N. sY!"cs/ris set 2 Underproducer Wild type Overproducer

a

Leaf damage (mOle) of entire plant

123 82

± 20a ± Ila

250 178

± 15a

J:

"

_"'1_, :I:

54 38 123 151 88

:I:

:I:

-'-

±

56a

N

" 22

15 20

9a 13a 9a

15 22 13

lOa 13a 8b

20 19 19

Studies were initiated approximately 20 days after germination. N. sy!vcs/ris set I plants had two larvae per plant. while all other sets had 10 larvae per plant. Values are means ± standard errors. Values in columns of like studies followed by different letters are significantly different at P < 0.05 by analysis of variance.

wild-type plants. Leaf disks from underproducing plants were damaged to about the same degree as those from wild-type plants, and generally there were no differences in mortality and weights of larvae fed on these two different types of plant disks. Disks from old leaves of N. sylvestris were damaged much less that those from new leaves for each corresponding degree of expressing type of plant. Leaf disks from insect-damaged plants (Table 3) were generally damaged less compared to disks from leaves of plants that had not been fed upon, especially for N. tabacum varieties (Table 2), but the length of leaves used were in some cases dissimilar. However. the same trends that were noted in Tables 1 and 2 were also noted with the leaf disks from the insect-damaged plants. Leaves from overproducing plants were generally damaged to a significantly lesser extent than wild-type plant leaf disks. Damage to underproducing N. tabacum Xanthi was significantly greater than to wild-type plants, while no significant differences in the degree of damage were seen for the underproducing vs. wild-type

2363


TABLE

2.


H. zea

ON LEAF DISKS FROM

UNDA!\IAGED PLANTS OF WILD-TYPE AND TRANSGENIC TOBACCOS OVER- AND UNDERPPRODUCING TOBACCO ANIONIC PEROXIDASE"

Mortality (%)

Damage mting Plant type

Day

N. ta!>aClIl1I Xanthi Underproducer 74 ± 16a Wild type 84 ± 12a N. ta!>aClIl1I Coker Wild type 2.4 ± O.4a Overproducer 1.2 ± O.lb N. sy/veslris set I (mature leaves) Underproducer 9 ± la Wild type 8 ± la Overproducer 7 ± la N. sy/veslris set I (young leaves) 46 ± 3a Underproducer Wild type 40 ± 4a Overproducer 24 ± 2b N. sy/veslris set 2 Underproducer 30 ± 3a Wild type 29 ± 3a 16 ± Ib Overproducer

Day 2

Day

Day 2

Weight (mg)

3.6 3.9

± 0.7a ± O.4a

5.6a I.3b

8.3a 9.0a

0.17 0.17

± O.Ola ± O.Ola

5.4

± 0.5a ± 0.5b

1.5 a 1O.5b

2.2a 6.4a

0.22 0.20

± O.Ola ± O.Ola

16 ± 2a 15 ± 2a 15 ± 2a

13.0a 10.5a 6.3a

11.8a 31.Ib 21.3ab

0.07 0.07 0.06

± O.Ola ± O.Ola

79 74 38

± 7a ± 6a ± 4b

10.2a 10.3a 15.2a

1O.6a 14.3a 28.3b

0.11 0.11 0.09

± O.Ola ± O.Ola ± O.Ola

154 149 68

± 37a ± 30a ± 16b

19.2a 21.3a 33.3b

27.9a 20.3a 23.0a

0.14 0.16 0.13

± O.Olab ± O.Ola ± O.Olb

' 1 -).,;;..

O.Ola

"Damage rating and weight values are means ± standard errors. Values in columns of like pammeters and studies followed by different letters are significantly different at P < 0.05 by analysis of variance (damage mtings and weights) or chi-square (mortality) analysis. Ratings for day 2 of N. tabaclIl1I Xanthi and both days of N. tabaclIl1I Coker are on a 0-10 scale. other mtings are in 0.25 mm'. Numbers of leaves for ecah treatment of all N. tabcalll1l studies are 20. For N. s.v/veslris. numbers of leaves used are 15 for underproducer. 22 for wild type. and 13 for overproducer for both leaf types of set I; and 10 for underproducer. 14 for wild type. and 16 for overproducer for set 2. Mature leaf disks used for N. sy/veslris were from leaves approximately 14 days after initiation; other leaf disks were from leaves approximately seven days after initiation.

N. sylvestris. As was seen for the other leaf disk assays, there were typically few differences in mortality or weights of caterpillars fed on the different leafdisk types, although those fed on overproducing N. sylvestris were significantly smaller than those fed on wild-type plants. Caterpillars fed on leaf disks from wild-type N. tabacum Xanthi were significantly smaller that those fed on leaf disks from corresponding underproducing leaves. OveralL the mean of leaf sizes from the different expressing plants used in all of these concurrent assays was not significantly different (P < 0.05). Peroxidase Activity of Leaf Tissues. Peroxidase activity was not signifi-

2364

DOWD AND L.".GRIMINI

TABLE

3.


H. zea

ON LEAF DISKS FROM

DAMAGED PLANTS OF WILD-TYPE AND TRANSGENIC TOBACCOS OVER- AND UNDERPPRODUCING TOBACCO ANIONIC PEROXIDASE

Mortality (%)

Damage rating Plant type

Day 2

Day

N. rabacwll Xanthi Underproducer Wild type N. rabaclIlll Coker Wild type Overproducer N. sylvesrris set 2 Underproducer Wild type Overproducer

u

Day I

Day 2

Weight (mg)

19 II

± 2a ± 2b

66 29

± 9a ± 4b

4.5a I.la

22.la 23.4a

0.07 0.04

± O.Ola ± O.Olb

17 9

± 2a ± Ib

47 23

± 5a

ND ND

44.2a 31.1a

0.10 0.11

± O.Ola ± O.Ola

21 17 16

± la

59 51 39

± 6a ± 6ab ± 4b

8.9a 9.8a 8.9a

19.0a 22.8a 19.6a

0.13 0.15 0.12

± O.Olab ± O.Ola ± O.Olb

±

lab Ib

4b

"Damage rating and weight values are means ± standard errors. Values in columns of like parameters and studies followed by different letters are significantly different at P < 0.05 by analysis of variance (damage ratings and weights) or chi-square (mortality) analysis. Leaf ratings are all 0.25 mm 2 • Numbers of leaves for each treatment are 17 for underproducer and 19 for wild type of N. rabaclIlll Xanthi. 11 for both leaf types of N. rabaclIlll Coker. and 20 for underproducer. 18 for wild type. and 19 for overproducer of N. sylvesrris. Leaf disks were from leaves approximately seven days after initiation. ND = not detemlined.

TABLE

4. TOTAL PEROXIDASE ACTIVITY IN REPRESENTATIVE WILD-TYPE. OVER-. AND UNDERPRODUCING TOBACCO PLANTS"

Peroxidase activity Plant type

N. rabaclIlll Xanthi undamaged Underproducer Wild type N. rabaclIlll Xanthi damaged Underproducer Wild type N. rabaculll Coker undamaged Wild type Overproducer N. rabaclIlll Coker damaged Wild type Overproducer

(.6. AU/min/g fresh weight)

6.2 8.2

± 1.0a ± 1.0a

24.1 17.0

± 2.4a ± l.3b

18.7 79.6

± 6.la ± 15.5b

20.9 89.1

±

l.3a

± 28.3b

"Values are means ± standard errors. Values in columns of like leaf types followed by the same letter are not significantly different at P < 0.05 by analysis of variance.


2365

cantly different in underproducing vs. wild-type plants for undamaged leaves of N. tabacum Xanthi (Table 4). Peroxidase activity in underproducing leaves from damaged plants of N. tabacum Xanthi was unexpectedly significantly greater than that of the wild-type plants. In the cases of leaves from both damaged and undamaged plants of N. tabacum Coker, mean peroxidase activity in leaves from overproducing plants was approximately four to five times that of the wildtype plants. DISCUSSION

Peroxidase-Mediated Resistance in Undamaged Plants. Peroxidases can promote insect resistance in plants by converting substrates to highly reactive chemical species that subsequently undergo a variety of reactions that can adversely affect insects. These highly reactive compounds can reduce nutritional quality by binding to nutrients and making them less digestible, inhibit digestive enzymes, act as direct cytotoxins, and polymerize and/or crosslink to increase tissue toughness (e.g., Appel, 1993; Felton et a!., 1989, 1992; Dowd and Lagrimini, 1997a; Miles and Gertli, 1993). By examining the properties of the peroxidases in the different plants and differences in expression, substrate specificity, and substrate presence, it should be possible to gain some understanding of how changing peroxidase activity can alter resistance levels in transgenic plants compared to wild-type plants. Chiorogenic acid and rutin are the more common allelochemicals in tobacco and are suitable substrates for peroxidases (Dowd and Vega, 1996). For N. sylvestris (Snook et a!., 1986) and the different lines of N. tabacum (Schlotzhauer et a!., 1981; Snook et a!., 1986), the concentration of chlorogenic acid is about lO-fold that of rutin. Chlorogenic acid analogs are also common and are approximately equivalent in concentration (Snook et a!., 1986). Chlorogenic acid oxidized by peroxidase binds to proteins and reduces their nutritional quality to insects (Felton et a!., 1992). The products produced by peroxidase oxidation of chlorogenic acid and rutin are also more toxic to insects than the compounds themselves (Dowd and Vega, 1996). Thus, enhanced peroxidase activity in tobacco may result in more rapid generation of toxic metabolites, which additionally may bind to nutrients and reduce their digestibility. Reduced levels of peroxidase would tend to lower the production of these metabolites, making the tissues more acceptable to insects. Effects noted with leaf disks from undamaged plants were generally consistent with the idea of enhanced peroxidase activity being associated with enhanced insect resistance (overproducing plants), but reducing tobacco anionic peroxidase activity (underproducing plants) did not reduce insect resistance. The best explanation of these observations is that the total peroxidase activity is more important than the activity of individual peroxidase isozymes. Total peroxidase activity is enhanced by a significant level in the overproducing tobacco plants (present data, Dowd and

2366

DOWD AND L.>\GRIMINI

Lagrimini, 1997b), but in the present study the underproducing leaves examined had approximately equal levels of total peroxidase activity compared to the wild types. Isoforms other than the tobacco anionic peroxidase apparently provide enough net peroxidase activity so that feeding resistance to insects is approximately equal initially in underproducing and wild-type plants. The fast (highly acidic) anionic peroxidases (equivalent to the tobacco anionic peroxidase studied in the present study) of different tobacco species can more rapidly oxidize the native substrates rutin and chlorogenic acid than slow anionic or cationic peroxidases (Sheen, 1974). However, the relative contribution of these peroxidase isozymes to total peroxidase activity varies in the different species (Sheen, 1974). For N. sylvestris, the rate of rutin oxidation by the fast anionic peroxidases is ca. 4 x, and chlorogenic acid, ca. 2 x, the combined rates for the other isozymes (Sheen, 1974). For N. tabaclIm, the contribution of the fast anionic peroxidases for rutin is ca. lOx the combined rate of the other peroxidases, and ca. 4 x the rate of the other peroxidases for chlorogenic acid (Sheen, 1974). In our studies, the degree of insect damage in the underproducer was little affected compared to the wild-type in N. sylvestris. The higher proportion of isozymes other than the fast forms in N. s.vlvestris compared to N. tabaclim may explain why the degree of insect damage of underproducing N. sylvestris was less different from the wild-type compared to corresponding plant types of N. tabaclim. This difference in species response of the underproducing vs. wild-type plants was much more dramatic in damaged plants of N. sylvestris and N. tabacllm compared to leaf disks from undamaged plants, suggesting other factors are interacting differentially with the changes in peroxidase activity in damaged vs. undamaged plants. Peroxidase-Mediated Resistance in Damaged Plants. Damaging plants can significantly change the resistance level to insects. Leaf disks of the same plant type from insect-damaged plants of the tobacco varieties tested generally received much less damage than those from the corresponding undamaged plants. Both insect and mechanical damage to tomato leaves can increase resistance through the induction of both chemical and biochemical factors (Stout et al., 1994, 1996: Stout and Duffey, 1996). Induction of trypsin inhibitor in tobacco and tomato by insects occurs to a similar degree (Jongsma et aI., 1994). Assuming additional factors are also induced by insect feeding on tobacco, this may explain the different effects seen with underproducing leaves from prefed upon vs. undamaged leaves of N. tabacliln Xanthi. Enhanced peroxidase activity has been reported in soybean (Glycine men) (Bi and Felton, 1995) and tomato (Lycopersicon esclilentllm) (Stout et aI., 1996) leaves in plants damaged by insects. However, in tomato, significantly enhanced peroxidase activity is only seen for the specific leaf that is damaged (Stout et al., 1996). Our results appear consistent with those with tomato in that leaves remote from the damaged leaf did not appear to have higher peroxidase activity compared to undamaged leaves in leaf disk assays.


2367

The results seen with the leaf disks from insect-damaged plants compared to the disks from the undamaged plants, as well as the whole plant assays, suggest additional factors are interacting with the peroxidases that are mediating insect resistance. While the overproducing plants demonstrate similar levels of enhanced enzyme activity and insect resistance relative to the wild-type plants, leaf disks from damaged underproducing plants generally were more susceptible to feeding compared to the wild-type plants (significantly so in some cases). Net feeding was always less than for leaf disks from the undamaged plants for the same plant types. When mechanically damaged or insect damaged, the fast anionic peroxidases are only slightly induced, while the slow anionic and cathodic forms are highly induced in the damaged leaf (Lagrimini and Rothstein, 1987; Dowd and Lagrimini, 1997b). Insect damage can greatly enhance nonanionic peroxidase activity in N. s.vlvestris, but in overproducing plants the anionic peroxidases still contribute most of the total peroxidase activity (Dowd and Lagrimini, 1997b). Wound-induced lignification occurs 24-48 hr sooner in the pith of plants overproducing the anionic peroxidase (Lagrimini, 1991). Properties of the overproducing tissue related to lignification include increased phenolic levels (Lagrimini, 1991) and smaller cells (Lagrimini, 1992). These may both increase the toughness of the tissue, which is a known deterrent to insect feeding (review Dowd and Lagrimini, 1997a). If a similar situation occurs in the leaves, then for the whole plant assays the advantage of the anionic peroxidase expression would be similar in N. sylvestris and N. tabacull1 if it is the more important defensive response, which was supported by data in the present study. By analogy, lignin production following damage in underproducing tissues should be significantly slower than in wild-type tissue with corresponding reductions in insect resistance: this was also seen with both plant species in the present study. Thus, the peroxidase reactions with rutin and chiorogenic acid may be more important in the short-tem1, immediate response (undamaged plants), while induction of lignification involving peroxidase would be more important in longtem1, induced resistance. This trend is somewhat similar to the conclusion reached for investigations of mechanisms contributing to varietal resistance of com. The allelochemical DIMBOA appeared to be more important in determining resistance to Ostrillia Ilubilalis in young leaves, while toughness of leaves (which can be mediated by peroxidase cross-linkings) was more important in determining resistance in older leaves (Bergvinson et aI., 1995). However, our results suggest relative peroxidase activity can mediate both allelochemical and structural resistance in undamaged and damaged leaves, with the allelochemical interactions more important in the undamaged leaves and structural factors more important in the damaged leaves. The age of the leaves examined appears to be important in making conclusions about peroxidase activity, because after leaves catch up in terms of maturity, there is no difference in resistance in uninduced tissues (mature leaf assays with N. sylvestris). This same eflect was noted with

2368

DOWD AND LAGRIMINI

wild-type and tobacco anionic peroxidase-producing leaves of sweetgum (Liqllidamber styracifllla) of differing maturity when fed to some species of caterpillars (Dowd and Lagrimini, 1997a,c). Multiple mechanisms in addition to peroxidase may be involved that also are mediated by plant age, and damage type, as discussed previously.

CONCLUSIONS

Leaf material from two species of tobacco overproducing tobacco anionic peroxidase was often damaged significantly less than leaf material from wildtype plants by H. zea larvae. Leaf material underproducing tobacco anionic peroxidase was either damaged at the same rate or significantly more (depending on the plant species and leaf source) compared to wild-type leaf material. These results suggest expression of this particular peroxidase can mediate the degree of resistance of tobacco to feeding by chewing insects. This is likely to be one of several factors involved in resistance to insects, but nevertheless can be an important one. Because peroxidases are widely distributed in plants fed on by insect herbivores, peroxidase activity of appropriate isozymes is also likely to be important in determining relative insect resistance in other plants. However, the context of the peroxidase is important in determining its relative importance, due to the complexity of substrates and reactions involved. AckllowledgmeJlts-We thank C. M. Anderson and B. D. C. Hansen for technical assistance. T. C. Nelsen (USDA. Agricultural Research Service. Midwest Area Biometrician) for suggestions on statistical analysis. and 1. Estruch. G. W. Felton. R. A. Norton. L. Privalle. 1. L. Richard. and anonymous reviewers for comments on prior versions of this manuscript.

REFERENCES ApPEL. H. M. 1993. Phenolics in ecological intemctions: The importance of oxidation. J. Chon. Ecol.19:1521-1553. BERGVINSON. D. J•• LARSEN. 1. S.. and ARNASON. 1. T. 1995. Effect of light on changes in maize resistance against the European com borer. Ostrinia Ilubilalis (Hubner). Can. Emolllol. 127: 111-122. Bl. 1. L.. and FELTON. G. W. 1995. Foliar oxidative stress and insect herbivory: Primary compounds. secondary metabolites. and reactive oxygen species as components of induced resistance.1. Chelll. Ecol. 21:1511-1530. COHEN. E. 1993. Chitin synthesis and degmdation as targets for pesticide action. Arch. Illsect Biochelll. Physiol. 22:245-261. CORBIN. D. R.. GREENPLATE. 1. T .. WONG. E. Y.. and PURCELL. J. P. 1994. Cloning of an insecticidal cholesterol oxidase gene and its expression in bacteria and protoplasts. Appl. Enviroll. Microbiol. 60:4239-4244. DOWD. P. F. 1988. Toxicological and biochemical interactions of the fungal metabolites fusaric


2369

acid and kojic acid with xenobiolics in Heliothis cea and Spodoptera frugiperda. Pestic. Biochem. Physiol. 32:123-134. DowD. P. F. 1994. Enhanced maize (Zea mays L.) pericarp browning: Associations with insect resistance and involvement of oxidizing enzymes. J. Chem. Ecol. 20:2497-2523. DowD. P. F.. and L."-GRIMINI. L. 1'v1. 1997a. The role of peroxidase in host insect defenses. pp. 195-223. in N. Carozzi and M. Koziel (eds.). Transgenic Plants for Control of Insect Pests. Taylor and Francis. London. DowD. P. F.. and LAGRIMINI. L. M. 1997b. Examination of transgenic tobacco and tomato overexpressing tobacco anionic peroxidase for resistance to insects. Submitted. DowD. P. F.. and LAGRIMINI. L. M. 1997c. Examination of transgenic sweet gum overexpressing tobacco anionic peroxidase for leaf resistance to insects. Submitted. DowD. P. F.. and NORTON. R. A. 1995. Browning-associated mechanisms of resistance to insects in com callus tissue. J. Chem. Ecol. 21 :583-600. DOWD. P. F.. and VEGA. F. E. 1996. Enzymatic oxidation products of allelochemical as a potential direct resistance mechanism against insects: Effects on the com leafhopper Dalbulus maidis. Nm. Toxins 4:85-91. FELTON. G. W.. DONATO. K. DELVECCHIO. R. J.. and DUFFEY. S. S. 1989. Activation of plant foliar oxidases by insect feeding reduces nutritive quality of foliage for noctuid herbivores. 1. Chem. Ecol. 15:2667-2694. FELTON. G. W.. DONATO. K. K.. BROADWAY. R. M.. and DUFFEY. S. S. 1992. Impact of oxidized plant phenolics on the nutritional quality of dietary protein to a noctuid herbivore. Spodoptera exigua. J. Insect Physiol. 38:277-285. GATEHOUSE. A. M. R.. SHI. Y .. POWELL. K. S.. BROUGH. C.. HILDER. V. A.• HAMILTON. W. D. 0 .. NEWELL. C. A.. MERRYWEATHER. A.. BOULTER. D.. and GATEHOUSE. J. A. 1993. Approaches to insect resistance using transgenic plants. Phil. Trans. R. Soc. London B. 342:279-286. GATEHOUSE. A. M. R.. HILDER. V. A.. POWELL. K. S.. WANG. M.. DAVISON. G. M.. GATEHOUSE. L. M.. DOWN. R. E.. EDMONDS. H. S.. BOULTER. D.. and NEWELL. C. A. 1994. Insect-resistant transgenic plants: Choosing the gene to do the "job." Biochem. Soc. Trans. 22:944-949. JONGSMA. M. A.. BAKKER. P. L.. VISSER. B.. and STEIKMA. W. 1. 1994. Trypsin inhibitor acitvity in mature tobacco and tomato plants is mainly induced locally in response to insect attack. wounding and virus infection. Plallls 195:29-35. L\GRIMINI. L. M. 1991. Wound-induced deposition of polyphenols in transgenic plants overexpressing peroxidase. Plalll Physiol. 96:577-583. LAGRIMINI. L. M. 1992. Plant peroxidases: Under and over-expression in transgenic plants and physiological consequences. pp. 59-69. in C. Penel. T. Gaspar. and H. Greppin (eds.). Plant Peroxidases 1980-1990. University of Geneva. Geneva. L\GRIMINI. L. M.. and ROTHSTEIN. S. 1987. Tissue specificity of tobacco peroxidase isozymes and their induction by wounding and tobacco mosaic virus infection. Plalll Physiol. 84:438-442. LAGRIMINI. L. M.. BURKHART. W.. MOYER. M.. and ROTHSTEIN. S. 1987. Molecular cloning of complementary DNA encoding the lignin-fomling peroxidase from tobacco: Molecular analysis and tissue-specific expression. Pmc. Natl. Acad. Sci. U.S.A. 84:7542-7546. LAGRIMINI. L. M.. BRADFORD. S.. and ROTHSTEIN. S. 1990. Peroxidase-induced wilting in transgenic tobacco plants. Plalll Cell 2:7-18. LAGRIMINI. L. M.. VAUGHN. 1.. FINER. J .. KLOTZ. K.. and RUBAIHAYO. P. 1992. Expression of a tobacco peroxidase gene in transfoffiled tomato plants. 1. Am. Hortic. Soc. 117:1012-1016. LAGRIMINI. L. M.. VAUGHN. J.. ERB. W. A.. and MILLER. S. A. 1993. Peroxidase overproduction in tomato: Wound-induced polyphenol deposition and disease resistance. HortScience 28:218-221.

2370

DOWD AND L."'GRIMINI

MILES. P. W.. and OERTLI. 1. 1. 1993. The significance of aniioxidants in aphid-plant interactions: The redox hypothesis. Emolllol. Erp. Appl. 67:275-283. PENG. Z.. and tvllLES. P. W. 1988. Acceptability of catechin and its oxidative condensation products to the rose aphid. MacrosiphulIl rosae. EnlOlIlol. Erp. Appl. 47:255-265. PURCELL. 1. P.. GREENPLATE. J. T .. JENNINGS. M. G.. RYERSE. J. S.. PERSHING. J. C.. SIMONS. S. R.. PRINSEN. M. 1.. CORBIN. D. R.. TRAN. M.. and SAMMONS. R. D. 1993. Cholesterol oxidase: A potent insecticidal protein active against boll weevil larvae. Biochelll. Biophys. Res. COIllIllIllI.

196:1406-1413.

ROTHSTEIN. S. 1.. and LAGRIMINI. L. M. 1989. Silencing gene expression in plants. Oxf

SIIIT.

PlmI{ lvlol. Cell. Bioi. 6:221-246.

SAS INSTITUTE. Inc. 1985. SAS User's Guide. SAS Institute. Inc .. Cary. North Carolina. SHAH. D. M.. ROMMENS. C. M. T.. and BEACHY. R. N. 1995. Resistance to diseases and insects in transgenic plants: Progress and applications to agriculture. Trends Biolech. 13:362-368. SHEEN. S. J. 1974. Polyphenol oxidation by leafperoxidases in Nicoliana. Bol. Gaz. 135:155-161. SCHLOTZHAUER. W. S., MARTIN. R. M.. SNOOK. M. E.. and WILLIAMSON, R. E. 1981. Pyrolytic studies on the contribution of tobacco leaf constituents to the fonnation of smoke catechols. Tob. Sci. 30:372-374.

SNOOK. M. E., MASON. P. F., and SISSON. Y. A. 1986. Polyphenols in the Nicoliana species. Tob. Sci. 30:43-49. STOUT. M. 1.. and DUFFEY. S. S. 1996. Characterization of induced resistance in tomato plants. Emolllol. Erp. Appl. 79:273-283.

STOUT. M. 1.. WORKMAN. 1.. and DUFFEY. S. S. 1994. Differential induction of tomato foliar proteins by arthropod herbivores. J. Chelll. Ecol. 20:2575-2594. STOUT. M. 1.. WORKMAN, K. Y., and DUFFEY. S. S. 1996. Identity, spatial distribution. and variability of induced chemical responses in tomato plants. Emolllol. Erp. Appl. 79:255-271.

Supplied by U.S. Dept. of National Center for Utilization Research, Peorh:1,