SOME ASPECTS OF THE SEQUESTRATION OF CARDENOLIDES IN THE LARGE MILKWEED BUG, ONCOPELTUS FASCIATUS (DALLAS) (HEMIPTERA: LYGAEIDAE)
by LYNN MARIE VASINGTON MOORE B.Sc. (Magna Cum Laude) UNIVERSITY OF CONNECTICUT, 1975 M.Sc. UNIVERSITY OF MASSACHUSETTS, 1978
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF ZOOLOGY
We accept this thesis as conforming to the-f5e.auired standard
April, 1985 (c) Lynn Marie Vasington Moore, 1985 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.
Department of
The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3
OE-6 (3/81) ii
ABSTRACT
Specific aspects of the selective sequestration, excretion and tolerance of cardenolides in the large milkweed bug, Oncopeltus fasciatus have been studied using spectrophotometry assays, thin-layer chromatography, tracer studies, in vivo tolerance assays, and enzyme inhibition techniques. The cardenolide content of the dorsolateral space, gut, wings and fat body of Oncopeltus fasciatus was examined. The results indicate that the majority of cardenolides sequestered in the insect are concentrated in the dorsolateral space, which confirms the basic pattern of quantitative distribution of cardenolides in fJ. fasciatus determined in earlier work. Large amounts of cardenolides were not found in the gut, wings and fat body. The female fat body contained 4-5% of the total cardenolide content of the insect. The cardenolide content of male fat body, and gut and wings of both sexes was below the detection limit of the cardenolide assay. Thin-layer chromatography was used to determine the cardenolide array of various tissues and secretions of fJ. fasciatus reared on seeds of a single species of milkweed (A. speciosa) and adult extracts and dorsolateral space fluid of 0. fasciatus reared on seeds of two species of milkweed with different cardenolide arrays (A. speciosa and A. syriaca)•
The results indicate that cardenolides are not sequestered in the insect simply on the basis of polarity and that metabolism and differential excretion of cardenolides are involved in the sequestration of cardenolides
in fJ. fasciatus.. The similarities in the cardenolide profiles of
0. fasciatus reared on different food sources, and tissues of fj. fasciatus
reared on a single food source indicates that there is regulation of the cardenolide array in 0. fasciatus. An in vitro preparation of Malpighian i 1* i
tubules was used to investigate the excretion of the polar cardenolide,
ouabain, in 0. fasciatus. Both segments of the tubules were found to metabolize ouabain. The distal Segment (Segment II) secreted primary urine
and ouabain. Secretion of ouabain by Segment II was not observed to occur
against a concentration gradient and increased with increasing fluid
secretion. The proximal segment (Segment I) reabsorbed fluid and ouabain
but not metabolites of ouabain. Ouabain was reabsorbed against a strong
concentration gradient (23-fold), was independent of fluid reabsorption,
and increased with increasing fluid secretion by Segment II. In rapidly
secreting Malpighian tubules (a situation of high cardenolide secretion by
Segment II), the presence of Segment I reduced the excretion of ouabain by
84 - 93%, mainly by reducing ouabain concentration. It appears excretory
loss of cardenolides can be reduced in 0. fasciatus and thus may be a
factor in the sequestration of cardenolides in this insect. 0. fasciatus
tolerated 1954x and 7288x, respectively, the LD50 ouabain dose of
Schistocerca gregaria and Periplaneta americana when ouabain was injected
into the hemocoel of these insects. The maximum ouabain dose that could be
injected into 0_. fasciatus (200 nmoles) resulted in no mortality; this dose
is higher than the lethal ouabain doses recorded for vertebrates and
invertebrates. The ouabain concentration resulting in 50% inhibition
(I50) of Na,K-ATPase activity was determined in lyophilates of nervous
tissue of 0. fasciatus and brain and recta of j>. gregaria and were 2.0 x
4 6 6 lO" , 2.0 x 10" , and 1.0 x 10~ M, respectively. The I50
value for ouabain inhibition of Na,K-ATPase activity in the nervous tissue
of 0_. fasciatus is higher than the I50 values for nervous tissue in most other insects as well as many other invertebrate and vertebrate iv
tissues. Thus, the presence of ouabain resistant Na,K-ATPases appears to be a factor in the tolerance and sequestration of plant cardenolides in
0. fasciatus. V
Table of Contents Page Abstract . ii Table of Contents v List of Tables viii List of Figures x Acknowledgements xii CHAPTER 1: GENERAL INTRODUCTION 1 Toxic plant compounds in plants and insects 1 Cardenolide chemistry and toxicity 2 Cardenolides and the Asclepiadaceae (milkweeds) 5 Ouabain and digitoxin 6 Cardenolide sequestration in insects 10 Oncopeltus fasciatus 11 Cardenolide sequestration in Oncopeltus fasciatus 13 CHAPTER 2: SELECTIVE SEQUESTRATION OF MILKWEED (ASCLEPIAS sp.) CARDENOLIDES IN ONCOPELTUS FASCIATUS 17 Summary 17 Introduction 18 Material and Methods 20 1) Insects 20 2) Collection of samples 20 3) Extraction of cardenolides 21 4) Cardenolide determinations 22 a) Total cardenolide concentrations 22 b) Determinations of cardenolide profiles with thin-layer chromatography (TLC) 23 vi
Page Results 25 1) Total cardenolide content of the fat body, wings, gut and dorsolateral space fluid of fj. fasciatus reared on
A. speciosa 25
2) Cardenolide profiles of 0 fasciatus reared on A speciosa
and A. syriaca seeds 28 3) Cardenolide profiles of adult extracts, tissues and dorsolateral space fluid of 0. fasciatus reared on
A. speciosa seeds 38 4) Geographic differences in the cardenolide profile of A. syriaca seeds 41 Discussion 43 CHAPTER 3: EXCRETION OF OUABAIN BY MALPIGHIAN TUBULES OF ONCOPELTUS FASCIATUS 50 Summary 50 Introduction 51 Material and Methods 52 1) Insects 52 2) Salines 53 3) In vitro Malpighian tubule preparation 53 4) Chromatography 55 Results 59 1) Characteristics of ouabain transport by Segment II 59 a) Secretion of ouabain by Segment II 59 b) Metabolism of ouabain by Segment II 65
2) Ouabain Reabsorption in Segment 1 65 vii
Page
a) Direct analysis 65
b) Set droplets 72
3) Modification of Segment II secretion by Segment 1 76
Discussion 79
CHAPTER 4: OUABAIN RESISTANT NA.K ATPASES AND CARDENOLIDE
TOLERANCE IH THE LARGE MILKWEED BUG, ONCOPELTUS
FASCIATUS 84
Summary 84
Introduction 85
Material and Methods 86
1) Insects 86
2) Ouabain injections 87
3) Na++K+ -dependent ATPase activity and ouabain
sensitivity 88
Results 90
1) Sensitivity of insects to injections of ouabain into the
hemocoel 90
2) Activity and ouabain sensitivity of Na++K+ -dependent
ATPases in tissue lyophilates of 0. fasciatus and
S^. gregaria 92
Discussion 99
CHAPTER 5: GENERAL DISCUSSION 108
Literature cited 113 viii
List of Tables
Page
Table 2.1 Total cardenolide content of adult an 5th instar
Oncopeltus fasciatus samples 26
Table 2.2 Rn, values of cardenolides and non-cardenolides
detected in extracts of 0. fasciatus and its food
sources, A. speciosa or A. syriaca 32
Table 2.3 The variability in cardenolide profiles of
0_. fasciatus and A..speciosa..'. 40
Table 2.4 Distribution of cardenolides in Oncopeltus fasciatus
expressed as % of the adult total 44 •
Table 3.1 Chromatographic systems used for the separation of
cardenolides and their metabolites 57
Table 3.2 Ionic composition (mM) of whole-tubule secretions,
reabsorbed fluid and bathing saline 73
Table 4.1 Survival of _0. fasciatus injected with 10 - 200
nmoles ouabain 91
Table 4.2 Post-injection recovery times of 0_. fasciatus 93
Table 4.3 Ouabain tolerance of CJ. fasciatus, _S. gregaria and
P_. americana 96
Table 4.4 Total, Mg2+ and Na,K-ATPase activity of tissue
lyophilates of 0_. fasciatus and S^. gregaria 98
Table 4.5 Inhibition of Na,K-ATPase activity by ouabain in
nervous tissue of 0. fasciatus and brain and rectum
of S. gregaria 102 ix
Page
Table 4.6 Effects of ouabain on tissue processes and survival
in insect and millipede species 103
Table 4.7 Toxicity values for ouabain in invertebrates and
vertebrates 105 X
List of Figures
Page
Figure 1.1 General structure of cardenolides 4
Figure 1.2 General structures of the three groups of Asclepias
cardenolides 7
Figure 1.3 The chemical structures of ouabain and digitoxin.... 9
Figure 2.1 Cardenolide profiles of A. speciosa and A. syriaca
seeds and adult extracts and dorsolateral space
fluid of fJ. fasciatus reared on each seed 30
Figure 2.2 Cardenolide profiles of 0. fasciatus adults,
tissues and secretions and A. speciosa seeds 35
Figure 2.3 Cardenolide profiles of A. syriaca seeds from four
geographic locations 42
Figure 3.1 In vitro preparation of rj. fasciatus Malpighian
tubules 54
Figure 3.2 The effect of varying external ouabain concentration
on urine-to-plasma (U/P) ratios 60
Figure 3.3 Effect of varying external ouabain concentration on
fluid secretion rate 61
Figure 3.4 Effect of fluid secretion rate on the rate of
ouabain secretion by Segment II alone 63
Figure 3.5 Effect of fluid secretion rate on ouabain
concentration in fluid secreted by Segment II alone 64
Figure 3.6 Distribution of radioactivity in typical
chromatograms of experimental and control fluids 66 xi
Page
Figure 3.7 Amount of fluid, ouabain, metabolised ouabain, and
ouabain concentration in Segment II secretion,
reabsorption droplets and whole-tubule secretion... 68
Figure 3.8 Amount of fluid secreted by whole tubule in 2 h in
relation to ouabain reabsorption by Segment I
during same time period 71
Figure 3.9 Change in fluid volume, radioactive label, and
radioactive label concentration in four set
droplets and whole-tubule secretion over 2 h 74
Figure 3.10 Rate of ouabain secretion with time 77
Figure 3.11 Ouabain concentration in secreted fluid with time.. 78
Figure 4.1 Mortality (24 h) of S^. gregaria and P_. americana
injected with varying amounts of ouabain 94
Figure 4.2 Effect of varying ouabain concentrations on
Na,K-ATPase activity in lyophilate preparations of
fj. fasciatus nervous tissue and ^. gregaria brain
and rectum 100 xi i
Acknowledgements
I would like to thank my research supervisor, Dr. 6.G.E. Scudder, for his advice and support throughout my Ph.D. program. I would also like to thank the members of my research committee, Drs. J. Gosline, W. Milsom, J.
Phillips and G.H.N. Towers, for their constructive criticism and interest throughout my research program, and for reviewing this thesis. A special thanks to Joan Martin for sharing her scientific expertise, providing-lots of encouragement and masterfully slogging through the sludge that became my manuscripts and thesis!
The advice of Dr. J.N. Seiber on chromatography systems and the advice of Dr. J.H. Anstee on techniques for determining Na,K-ATPase activity in
insect tissues is gratefully acknowledged. Carl Edman cheerfully synthesized the 2,4,2',4'-tetranitrodiphenyl and Dr. M.B. Isman provided advice and encouragement.
To the 4th floor critters: Jill Lancaster, Jennifer Robinson, Edie
Bijdemast, Dick Cannings, Syd Cannings, John Ford, Deb Cavanaugh, Garry
Stenson and Rich Lechleitner, and Liz and Bob Hancock; thanks for loads of fun, friendship, advice and encouragement.
A special thanks to my husband, Richard, for his love, patience and fun.
This work was supported by operating Grant A0865 from the Natural
Sciences and Engineering Research Council of Canada to Dr. G.G.E. Scudder and by a University Graduate Summer Scholarship from the University of
British Columbia to L.V. Moore. 1 J
CHAPTER 1: GENERAL INTRODUCTION
The large milkweed bug, Oncopeltus fasciatus (Dallas) feeds exclusively on plants that contain cardenolides (Duffey and Scudder, 1972). Cardenolides are specific inhibitors of Na,K-ATPase activity, and are toxic to many vertebrates and invertebrates (Detweiter, 1967; Biekirch, 1977; Rafaeli-Bernstein and Mordue, 1978; Benson et al_., 1979; Stekhoven and Bonting, 1981; Anstee and Bowler, 1984) at concentrations as low as 10"8 M (Glynn, 1964). 0. fasciatus, however, stores large quantities of these compounds in its body (up to 707 jjg/insect; Isman, 1977), and seems to suffer no ill effects from them (Isman, 1977; Chaplin and Chaplin, 1981; Jones etjal_. 1983). It has been suggested that the majority of cardenolides in the insect are sequestered in the dorsolateral space, a highly vacuolated epithelial layer in the integument, and that these cardenolides function in chemical defense against potential predators of fj. fasciatus (Duffey and Scudder, 1974; Scudder and Meredith, 1982a).
The work described in this thesis was undertaken to investigate some of the possible processes involved in the sequestration and tolerance of potentially toxic cardenolides in £. fasciatus. Therefore, I have provided below a brief summary of various aspects of the cardenolide system pertinent to the chapters that follow.
Toxic plant compounds in plants and insects Toxic plant compounds defend plants from general attack by microorganisms, fungi, insects and other herbivores (for review see Whittaker and Feeny, 1971). However, despite their toxicity, plants rarely 2
escape insect herbivory. Many insects are resistant to the toxic effects of a specific group or array of plant chemicals (for review see Dowd et al., 1983), and like 0_. fasciatus feed specifically on plants containing these chemicals. In many of these insects, toxins from the host plants are sequestered in various tissues and secretory glands where they function in the insect's defense against its own predators (for review see Blum, 1981;
Blum, 1983). Insects that sequester plant toxins are often brightly colored and part of large mimicry complexes, presumably to warn potential predators of their distastefulness or toxicity (for review see Rothschild,
1972; Huheey, 1984). Cardenolides are one of the better studied groups of plant chemicals that are involved in defense systems of plants and insects, as well as in insect mimicry complexes.
Cardenolide chemistry and toxicity
Cardenolides are a class of cardiac glycosides, a group of compounds well known for their bitter, emetic, and toxic qualities (Hoch, 1961).
Cardenolides are toxic to many vertebrates and invertebrates (Hoch, 1961;
Treherne, 1966; Detweiler, 1967; Anstee and Bell, 1975; Biekirch, 1977;
Rafaeli-Bernstein and Mordue, 1979; Stekhoven and Bonting, 1981; Anstee and
Bowler, 1984), at concentrations as low as 10-8 M (Glynn, 1964). In vertebrates, the different toxicities of individual cardenolides have been related to differences in their polarities and/or partition coefficients
(Detweiler, 1967; Okita, 1967; Lullmann and Peters, 1973; Smith and
Haber, 1974). Cardenolides are classified as cardiac glycosides owing to their toxic effects upon vertebrate heart tissue and the presence of one to four sugars attached to the genin. Cardenolides are 23 carbon steroids and characterized by a 5 carbon ,8-unsaturated ^-lactone (butenolide) ring 3 attached at C-17, a cis juncture of rings C and D and a 14j£-uH group (Fig. 1.1, Seiber et _al_., 1983); all three characteristics are essential for the toxic activity of these compounds (Wright, 1960; Nover, 1972). Sugars are generally attached to the genin at C-3 through an hydroxy group, and are often unusual derivatives of deoxymethylpentose, normal hexoses such as glucose are rarely encountered (Paris, 1963; Seiber ^t , 1983).
Cardenolides exert their toxic effects by specifically inhibiting the enzyme Na++K+-ATPase (Na,K-ATPase; E.C. 3.6.1.3; Akera, 1977; Bodeman, 1981). This enzyme, also known as the sodium pump, is ubiquitous among animal species and all animal cells, and is essential in maintaining the electrochemical gradients for Na+ across the cell plasma membrane (Jtfrgensen, 1980; Stekhoven and Bonting, 1981). The enzyme is restricted to the basalateral plasma membrane of epithelia, with highest activities occurring in excitatory and secretory tissues (Stekhoven and Bonting,
1981) . Na,K-ATPases are involved in numerous vertebrate tissue functions, i.e. salt regulation in avian salt glands (Eastin et , 1982), bicarbonate and acid secretion in gastric mucosa (Helander and Durbin,
1982) , solute reabsorption (particularly Na+) and regulation of intracellular pH and Ca2+ levels in kidney tubules (Jtfrgensen, 1980) and proper functioning of nervous tissue (Stekhoven and Bonting, 1981). In invertebrates, Na,K-ATPases have been demonstrated to function in Malpighian tubule secretion (Pilcher, 1970; Atzbacher et jaK, 1974; Anstee and Bell, 1975), proper functioning of nervous tissue (Treherne, 1966; Farquharson, 1974), midgut ion fluxes and transepithelial potentials (O'Riordan, 1969; Prusch, 1978) and labial gland secretion (Kafatos, 1968).
Cardenolide induced mortality in vertebrates occurs when inhibition of 4
Fig. 1.1. General structure of cardenolides. 5
Na,K-ATPase activity in cardiac cells results in such large net losses of
K+ from the cytoplasm, the heart becomes arrhythmic and no longer functions danger, 1981). The actual cause of cardenolide induced death in invertebrates has not been determined, but may also be the result of cardiac arrest.
Cardenolides and the Asclepiadaceae (milkweeds)
Cardenolides have been isolated from twelve families of Angiosperms
(Singh and Rastogi, 1970) and are particularly abundant in the milkweed
(Asclepiadaceae) and dogbane (Apocynaceae) families (Paris, 1963). Most of the information available on the chemistry and distribution of plant cardenolides is from the Asclepiadaceae, the primary food plants of
0. fasciatus.
Nearly all the Asclepiadaceae examined to date contain cardenolides
(Seiber et jil_., 1983). The individual cardenolides in the plant (as many as 23; Brower et al_., 1984a) and their relative concentrations determine the plant's cardenolide array or profile. Comparative studies of the cardenolide content and profile of Asclepias species have revealed large species differences in terms of the quantity and array of cardenolides present; some differences within populations of the same species have also been noted (Seiber et ^1_., 1978; Brower et al., 1982; Brower et al., 1984a;
Moore and Scudder, 1985). Cardenolides are present in all parts of the plant (Roeske et _al_., 1976). There is a great deal of intraplant variation in terms of cardenolide content and array (Nelson et al_., 1981; Nishio et al., 1983), as well as seasonal variation in the gross cardenolide content of the various plant parts (Seiber et al_., 1983).
Plants synthesizing cardenolides appear to prevent autotoxicity by 6
physically isolating these toxins in leaf vacuoles (Loffelhardt et al.,
1979) or in the laticifer system where cardenolides are stored and transported in the latex in latices independent of the plant's vascular system (Nelson et a\_., 1981; Nishio et jH., 1983). Storage in the laticifer system is particularly characteristic of Asclepias species with high cardenolide content (Nelson et ^1_., 1981).
Three groups of cardenolides have been isolated in the Asclepiadaceae, simple cardenolide genins (i.e. uzarigenin and syriogenin) and their sugar derivatives, and two groups of 2,3-dihydroxy cardenolide derivatives with cyclic bridged sugars (Fig. 1.2, Seiber et _al_., 1983). The two latter groups are (1) the glycosides of calotropagenin (gomphoside, afroside, calactin, calotropin, asclepin, uscharidin, uscharin, voruscharin, and calotoxin) found in Calotropis sp. and A. curassavica, A. fruticosa,
A. vestita, A. cordifolia and A. californica, and (2) desglucosyrioside and its derivatives, with a 7,8 epoxy function (labriformidin, labriformin, syrioside, syriobioside) found in A. labriformis, A. eriocarpa, A., syriaca,
A. erosa and A. speciosa. Many of the asclepiad cardenolides are unidentified; with future investigation additional cardenolide groups may be discovered (Seiber et al_., 1983).
Ouabain and Digitoxin
Commercial preparations of host plant cardenolides of 0_. fasciatus are not available. Therefore, experiments investigating the sequestration of cardenolides in this insect have generally used two non-Asclepiad cardenolides, ouabain and digitoxin, owing to the contrasting polarities of these two compounds and to their availability in pure and radiolabeled preparations of known structure (Fig. 1.3). In addition, a number of the 7
Fig. 1.2. General structures of the three groups of Asclepias
cardenolides: A) simple cardenolide genins, B) glycosides of
calotropagenin, and C) desglucosyrioside and its derivatives. 8 9 10 metabolites of these compounds have been identified and are also available in pure preparations. Ouabain and digitoxin have a 5^ cis A/B steroid ring juncture rather than the 5 ot trans A/B steroid ring juncture characteristic of Asclepiad cardenolides. Cardenolide sequestration in insects Fifty insect species in seven orders are known to sequester cardenolides from their host plants (Rothschild and Reichstein, 1976). Like fJ. fasciatus, these insects are warningly colored and many are members of various mimicry complexes (Duffey and Scudder, 1972; Rothschild, 1972; Rothschild, 1976; Scudder and Duffey, 1972; Isman et al_., 1977). Various studies suggest that sequestered cardenolides are an effective part of the defense system of these insects (Rothschild, 1966; von Euw et al_., 1967; Rothschild and Kellet, 1972; Rothschild et ^1_., 1973; Rothschild et aK, 1978; Fink and Brower, 1981; Cohen and Brower, 1983; Malcolm and Rothschild, 1983). In most cases, insects appear to suffer no ill effects from sequestering large amounts of cardenolides (Erikson, 1973; Isman, 1977; Smith, 1978; Dixon et^l_., 1978; Chaplin and Chaplin, 1981; Jones ^t aK, 1983; Brower et al_., 1984a; but see Seiber et al_., 1980). A number of similarities in the overall process of cardenolide sequestration have been noted in the insects studied so far. In general, cardenolides are sequestered in insects in proportion to the cardenolide content of the food source (Roeske et al_., 1976; Isman, 1977; Cohen and Brower, 1983), although in the monarch cardenolide levels are relatively independent of the cardenolide content of the food source (Brower et al., 1982, 1984a,b). Comparisons of host plant and insect cardenolide profiles indicate that in general, not all of the cardenolides available in the host plant are found in the insect; this is a result, in part, of selective feeding on particular plant parts by the insect, as well as metabolism of 11
ingested cardenolides in the insect (Rothschild et al., 1970; von Euw et al_., 1971; Roeske ^t ail_., 1976; Duffey et al_., 1978; Seiber ^t al_., 1980; Brower et al., 1982; Scudder and Meredith, 1982b). Sequestration of cardenolides in insects also seems to be characterized by differential distribution of these compounds among the insects' tissues in terms of quantity, and in the monarch, also in terms of emetic potency (Duffey and Scudder, 1974; Brower and Glazier, 1975; Blum, 1983). However, despite the similarities in the overall process of cardenolide sequestration in insects stated above, the details of the different mechanisms involved appear to vary greatly between individual species, leading Blum (1983) to suggest that each species of insect has a unique process of sequestration, tolerance and defensive use of the toxins ingested with its food plant.
Oncopeltus fasciatus 0_. fasciatus is a member of the Lygaeidae (seed bugs), one of the largest families in the insect order, Hemiptera. The majority of Lygaeids feed upon mature seeds of plants. In common with other Hemipterans, the Lygaeids have piercing and sucking mouthparts and feed by liquefying their food with injections of saliva prior to ingestion. All Hemipterans undergo incomplete metamorphosis; there is little change during development except in size and reproductive maturity, and immatures and adults usually occupy similar habitats and utilize similar food resources. Hemipteran adults are active feeders and long-lived. In addition, both adults and immatures possess scent glands which may deter predators. 0. fasciatus is a relatively large Lygaeid (17-18 mm in length). It is widely distributed in the New World from Maine and Ontario, west to S. Dakota and California, and south throughout the Neotropics. The insect is less common in the northern 12 parts of its range, does not overwinter there, but migrates north each year from the south (Dingle, 1978). Six to seven additional Oncopeltus sp. occur in the southern and southeastern states (for further details of the Lygaeidae and Hemiptera see Slater, 1964 and Slater and Baranowski, 1978). The conspicuously colored orange and black 0. fasciatus belongs to the Lygaeinae, a class of Lygaeids characterized by their bright coloration. In general, Lygaeids are cryptically colored; it has been suggested that the brightly colored Lygaeinae are warningly colored to advertise their distastefulness or toxicity to predators (Scudder and Duffey, 1972). Brightly colored Lygaeinae are involved in mimicry complexes in Africa, North America and the Caribbean (Scudder and Duffey, 1972; Duffey and Scudder, 1972). The presence of cardenolides in the major food plants of the warningly colored Lygaeinae, as well as the presence of these plant-derived cardenolides in the Lygaeinae themselves, established a chemical basis for these mimicry complexes (Scudder and Duffey, 1972; Duffey and Scudder, 1972).
0. fasciatus feeds exclusively on plants containing cardenolides. The insect is found primarily on members of the Asclepiadaceae but also on members of the Apocynaceae. Mature seeds are an essential part of the diet of fj. fasciatus. Although nymphs feed on vegetative parts of the plant until mature seeds are available, fJ. fasciatus must feed on mature seeds to complete its development (Ralph, 1976; Blakley and Dingle, 1978). Seeds are a patchy food source, however, the high mobility and migratory behavior of fJ. fasciatus adults, as well as the oviposition behavior of the female, enable the insect to exploit this type of food resource (Ralph, 1977; Blakley, 1980; Dingle et al., 1980; Klausner et al., 1980; Chaplin and 13
Chaplin, 1981).
Cardenolide Sequestration in Oncopeltus fasciatus 0_. fasciatus sequesters cardenolides from the seeds and vegetative parts of its host plant throughout its life cycle (Duffey and Scudder, 1974; Feir and Suen, 1971; Moore and Scudder, 1985). Cardenolides are also found in the insect's warningly colored eggs (Duffey and Scudder, 1974). The wide range of cardenolide contents detected in field caught, as well as laboratory reared, insects (42 to 707 jiq cardenolide per insect) reflects both intra- and interspecies differences in the cardenolide content of the insects' host plants, and provides evidence that 0. fasciatus sequesters
y cardenolides in proportion to the cardenolide content of its food source (Duffey and Scudder, 1974; Isman, 1977; Isman et jil_., 1977; Duffey et al_., 1978). Quantitative regulation of cardenolide content in 0_. fasciatus has been indicated by the demonstration that a greater percentage of the seed cardenolide content is sequestered in insects reared on seeds of low cardenolide content than in insects reared on seeds of high cardenolide content (Vaughan, 1979). Evidence to date suggests that uptake of ingested cardenolides across the midgut is passive (Scudder and Meredith, 1982b). The presence of a dorsolateral space, a highly vacuolated epithelial layer in the integument, appears to be a specialization of the integument of 0_. fasciatus, and other members of the Lygaeinae, that is intimately involved with the sequestration of cardenolides in these insects (Scudder and Duffey, 1972; Scudder and Meredith, 1982a). The majority of cardenolides sequestered in 0_. fasciatus adults appear to be concentrated in the dorsolateral space where they are thought to function in the insect's defense against predators (Duffey and Scudder, 1974; Scudder and 14
Meredith, 1982a). When 0. fasciatus adults are roughly handled, cardenolide-rich dorsolateral space fluid is released in discrete droplets from cuticular openings along the thorax and abdomen where the fluid presumably repels predators by its emetic, bitter and/or toxic properties (Duffey and Scudder, 1974; Scudder and Meredith, 1982a). The dorsolateral space is present in the immatures, however, there are no cuticular openings, thus release of dorsolateral space fluid does not appear to be part of the defensive strategy of immatures (Duffey and Scudder, 1974). Investigations of the role of sequestered cardenolides in the defense of fJ. fasciatus against its potential predators is complicated by the presence of the insect's other defense mechanisms. Additional defenses of fJ. fasciatus include the release of volatile aldehydes and cardenolides from scent glands in the adult and immatures, the release of copious amounts of rectal fluid in immatures (Games and Staddon, 1973; Duffey and Scudder, 1974; Staddon and Daroogheh, 1981), and the presence of histamine or a histamine analogue in the rectal fluid and hemolymph of immatures and the hemolymph and dorsolateral space fluid of adults (Graham and Staddon, 1974). A further complication in assessing the role of sequestered cardenolides in the defense of fj. fasciatus is the variation in the cardenolide susceptibility of the insect's potential predators (Duffey, 1977). Therefore, the role of cardenolides in the insect's defense is at this date, ambiguous.
The actual sequestration system in the dorsolateral space is not understood, although interaction with an emulsion phase has been suggested (Duffey et al_., 1978). Ultrastructural and kinetic analysis indicate that cardenolide uptake into the dorsolateral space is passive, in proportion to the cardenolide content in the food source and the hemolymph, and 15 nonsaturable (Duffey et al_., 1978; Scudder and Meredith, 1982a, b). It has been suggested that the rapid uptake and accumulation of cardenolides in the dorsolateral space, which can occur when cardenolide levels in the hemolymph are as low as 3.5 x 10-7 M, may function in the insect's tolerance of these toxins by maintaining low levels of these compounds in the hemolymph (less than 6.5 x 10~6 M, Moore and Scudder, 1985). In addition, the very low, if any turnover of sequestered cardenolides in the dorsolateral space in the adult (Isman et^ aJL, 1977) may also be a factor in the insect's tolerance to these toxins.
In vivo and in vitro studies with digitoxin and ouabain, and in vivo studies with milkweed cardenolides, provide evidence of selective uptake of cardenolides by the gut and into the dorsolateral space, as well as evidence of differential metabolism of cardenolides at the level of the gut and elsewhere in the insect (Duffey and Scudder, 1974; Yoder et al_., 1976; Duffey et al_., 1978; Scudder and Meredith, 1982b). The chemical or structural basis for this selectivity is not known, however the role of polarity in the sequestration of cardenolides in 0. fasciatus has been suggested (Duffey, 1977; Duffey et al., 1978). Blum (1981, 1983) has stressed the need for detailed studies of the specifics of sequestration of plant toxins by insects. This thesis examines specific aspects of the sequestration and tolerance of cardenolides in 0_. fasciatus. Chapter 2 clarifies certain aspects of the differential distribution of natural cardenolides in 0. fasciatus and examines the cardenolide content of three organs, the gut, wings and fat body, that have not been investigated previously as possible sites of cardenolide accumulation. Chapter 2 also documents some of the capabilities of the sequestration process of cardenolides in 0. fasciatus 16 by determining and comparing the cardenolide array of various tissues and secretions of rj. fasciatus reared on seeds of a single species of milkweed, and of adult extracts and dorsolateral space fluid of insects reared on seeds of two species of milkweed with very different cardenolide profiles. Chapter 3 investigates the role of excretion in the sequestration of cardenolides in fJ. fasciatus by examining the function of the Malpighian tubules in the sequestration of the polar cardenolide, ouabain. Finally, since the relative sensitivity of 0. fasciatus to the toxic effect of cardenolides, as well as the possibility that the Na,K-ATPases in the insect are resistant to the inhibitory effects of cardenolides have never been determined, Chapter 4 compares the relative in vivo sensitivity of fJ. fasciatus to ouabain and the ouabain sensitivity of its Na,K-ATPases to that of two insects which do not normally encounter cardenolides in their diet. 17
CHAPTER 2: SELECTIVE SEQUESTRATION OF MILKWEED (ASCLEPIAS sp.) CARDENOLIDES IN ONCOPELTUS FASCIATUS
Summary The cardenolide content of the gut, wings and fat body of Oncopeltus fasciatus was examined. The female fat body contained 4-5% of the total cardenolide content of the insect. The cardenolide content of male fat body, and gut and wings of both sexes was below the detection limit of the cardenolide assay. Thin-layer chromatography was used to determine the cardenolide array of various tissues and secretions of 0. fasciatus reared on seeds of a single species of milkweed (A. speciosa) and adult extracts and dorsolateral space fluid of 0_. fasciatus reared on seeds of two species of milkweed with different cardenolide arrays (A. speciosa and A,, syriaca). The results indicate that cardenolides are not sequestered in the insect simply on the basis of polarity and that metabolism and differential excretion of cardenolides are involved in the sequestration of cardenolides in 0_. fasciatus. The similarities in the cardenolide profiles of _0. fasciatus reared on different food sources, and tissues of 0_. fasciatus reared on a single food source indicates that there is regulation of the cardenolide array in 0_. fasciatus. 18
INTRODUCTION
In this chapter, the cardenolide content of the gut, wings and fat body of fJ. fasciatus is determined to provide a more complete understanding of the quantitative distribution of cardenolides in this insect. In addition, the possibility of selective sequestration of host plant cardenolides in fJ. fasciatus in terms of the individual cardenolides sequestered and their relative concentrations is investigated. Finally, to document some of the capabilities of the sequestration of cardenolides in fJ. fasciatus, the cardenolide profile of adult extracts and dorsolateral space fluid of fj. fasciatus reared on seeds of two species of milkweed with very different cardenolide profiles, is determined.
The gut, wings and fat body have not been investigated previously as possible sites of cardenolide accumulation in fJ. fasciatus. It has been suggested that retention of toxic compounds in the gut lumen or tissue and in the fat body may aid in the tolerance of toxins in insects (Kilby, 1963; Brooks, 1976). Furthermore, the monarch butterfly (Danaus plexippus) which also sequesters cardenolides from its food plants, is known to sequester substantial amounts of plant cardenolides in both the gut lumen and the fat body (see Blum, 1981; 1983). Thus, it is of interest to see if cardenolides accumulate in the fat body and gut of 0. fasciatus. It is possible that the wings of fJ. fasciatus contain cardenolides since large concentrations of these compounds, thought to function in defense against predators, are sequestered in the wings of two lepidopterans, Danaus plexippus (Brower and Glazier, 1975; see Blum, 1981) and Cycnia inopinatus (Hy. Edwards) (see Blum, 1983) and in the elytra of the cerambycid beetle Tetraopes melanurus Schon. (Nishio et al_., 1983). 19
Therefore, in this chapter I also determine the cardenolide content of the wings of 0_. fasciatus. Selective sequestration in insects of host plant cardenolides, both in terms of the individual cardenolides sequestered and their relative concentrations, has been demonstrated in the aphid, Aphis nerii, the monarch butterfly, and two lygaeid bugs, Caenocoris nerii and Spilostethus pandurus (Rothschild et al., 1970; von Euw et al_., 1971; Roeske et al., 1976; Seiber etjH., 1980; Brower et al_., 1982). In addition, there is evidence that cardenolides of differing emetic potencies are selectively sequestered in various tissues of the monarch butterfly (Brower and Glazier, 1975). Therefore, in this chapter I investigate the possibility of selective sequestration of host plant cardenolides in 0_. fasciatus, as well as selective sequestration of cardenolides among the various tissues of the insect, by using thin-layer chromatography to determine the cardenolide array of various tissues and secretions of _0. fasciatus, as well as the cardenolide array of the insect's food source. In this chapter, I also determine the cardenolide array of adult extracts and dorsolateral space fluid of fJ. fasciatus reared on seeds of two species of milkweed with very different cardenolide profiles to document some of the capabilities of the sequestration process of cardenolides in 0. fasciatus. The cardenolide profiles are also used to determine the polarity distribution of cardenolides in 0. fasciatus and its tissues since the toxicity of cardenolides has been correlated with their polarity (Duffey, 1977; Smith and Haber, 1974; Detweiler, 1967) and it has been suggested that the sequestration of cardenolides in 0. fasciatus may be a function of their'polarity (Duffey, 1980).
This chapter provides evidence that cardenolides are differentially 20 distributed in fj. fasciatus in terms of quantity, but not polarity, and that cardenolides are not sequestered in the insect simply on the basis of polarity. This chapter also provides evidence that metabolism and differential excretion of cardenolides are part of the selective sequestration process of cardenolides in fJ. fasciatus. In addition, the similarities in the cardenolide profiles of fJ. fasciatus reared on different food sources, and between tissues of fj. fasciatus reared on a single food source indicates that there is regulation of the cardenolide array in fj. fasciatus.
MATERIALS AND METHODS 1) Insects Adult male and female and fifth instar larvae of Oncopeltus fasciatus taken from a laboratory culture maintained at 26° C under a 16:8 light:dark cycle were used in all experiments. The insects were reared either on commercial sunflower seeds (Heliothus annuus L.), or on milkweed seeds (Asclepias speciosa Torr. or A. syriaca L.). Insects reared on sunflower seeds provided controls since this food source does not contain cardenolides.
2) Collection of samples Gut (with contents) and fat body samples were dissected from insects in Berridge's dissecting saline initially (Berridge, 1966), and in later experiments in a saline based on the constituents of fj. fasciatus hemolymph (Meredith, jet jiK, 1984). Tissues were rinsed in three 0.5 ml aliquots of saline^nd placed directly in extracting solvent or blotted 1 second on each side and wet weight determined: no cardenolides were detected in the 21 saline rinses of fat body or gut tissue. No attempt was made to remove tracheae or remnants of the membrane that encloses the fat body. Insects used to assay gut (with contents) were anesthetized for 30 seconds with a low volume of CO2 or cooled at 4°C. The gut was ligatured with surgical thread (Ethicon, Inc.) at the Malpighian tubule-pylorus junctions, the anterior end of the 1st ventriculus and the posterior end of the hind gut, and removed by cutting distal to the ligatures: the ligatures ensured removal of the gut with its contents intact. Dissections were completed within 20 minutes. Dorsolateral space fluid and hemolymph were collected from the insect as described by Duffey and Scudder (1974). Urine/feces samples were collected by placing insects for 48-72 hours in small rearing dishes lined with Whatman Chromatography paper. An excess of seeds was provided and dechlorinated water was constantly available. After the collection period the filter paper, which had absorbed the urine/feces, was removed and extracted for cardenolides. Filter paper from a control rearing dish (N=l) set up identically to the rearing dishes for urine/feces samples but without insects was collected after 72 hours, extracted as for urine/feces samples and assayed for cardenolides: none was detected. Wings were removed from insects after immobilization at 4°C.
3) Extraction of Cardenolides With the exception of dorsolateral space fluid and hemolymph, cardenolides were extracted from all samples by either the insect or the seed method (Isman et al_., 1977). The seed method was modified by retaining the CHCI3 phase of the CHCl3:MeOH 10:1 extraction for cardenolide determination. Lipids in some samples interfered with both TLC 22 analysis and colorimetric determination of total cardenolide content. The seed method removed the interfering lipids and therefore, was used (i) to extract all samples for TLC analysis (except urine/feces), and (ii) with seed and fat body samples, used for total cardenolide determination. The insect method was used to extract whole insect, wing and gut samples for total cardenolide determinations and urine/feces samples for TLC analysis. All samples were extracted in a shaking incubator to facilitate diffusion of cardenolides and extraction liquors were concentrated to 10-20 jul by evaporation under N2 before application to TLC plates. Tests showed no difference in the cardenolide array in whole insect samples extracted by either method. For all samples, the petroleum ether discard of the seed method was assayed by TLC for cardenolides: none was detected. Even after seed method extraction, fat body extracts were difficult to apply to the TLC plate. Therefore, the extract was centrifuged (12000 g, 2 min) and the resulting two phases applied separately. Only one faint blue spot at the origin was detected in the lower liquid phase. Dorsolateral space fluid and hemolymph were collected from the insect, precipitated with acetone (5:1 acetone:hemolymph) and 95% EtOH (1 ^il dorsolateral space fluid/ml 95% EtOH), centrifuged (12000 g, 2 min) to remove proteins and the supernatant used for total cardenolide determination and TLC analysis of hemolymph. Resuspension of the pelleted hemolymph protein after 1 min of sonication in 95% EtOH yielded negligible cardenolide. Dorsolateral space fluid was applied directly after collection to TLC plates.
4) Cardenolide Determinations a) Total Cardenolide Concentrations Total cardenolide concentrations were determined by a spectrophoto- 23 metric assay using a Lambda 3 UV/VIS spectrophotometer (Perkin-Elmer) and the indicator 2,4,2',4'-tetranitrodiphenyl (TNDP) in the presence of base (NaOH) (Brower and Moffitt, 1974; Brower et al_., 1975). The TNDP reaction was run at room temperature. Sample absorbance at 626 nm was recorded 40 minutes after the colorimetric reaction was started, with 95% EtOH as the reference. Two controls were used to monitor any absorbance at 626 nm owing to substances in the individual sample extracts tested, as well as the reaction reagents TNDP and NaOH: Control 1 = the absorbance of the particular insect sample being tested, in the absence of TNDP; Control 2 = the absorbance of TNDP and NaOH, in the absence of insect sample. The absorbance of the two controls was subtracted from the absorbance of the experimental cuvette (insect sample, in the presence of TNDP and NaOH). Cardenolide concentrations were determined by comparison to a digitoxin standard, and expressed as molar or jjg equivalent amounts of digitoxin (jugn) to facilitate comparisons with other studies (Roeske et al_., 1976). Brower and Glazier (1975), using the TNDP reaction, found only minor differences in the extinction coefficients of nine cardenolides and digitoxin. Therefore, it was assumed that digitoxin and the individual cardenolides ingested by the insects have similar extinction coefficients.
b) Determination of Cardenolide Profiles with Thin-layer Chromatography (TLC) Thin-layer chromatography plates prepared with Silica Gel G (Redi/Plate, Fisher Sci. Co., gel 250 jm thick) were used. Plates were activated over concentrated H2SO4 24 h prior to and 12 h after applying samples (Duffey and Scudder, 1972). The relatively nonpolar cardenolide digitoxin was spotted as a standard on both sides of each sample. Plates 24 were developed in filter paper- lined, saturated chambers containing methylene chloride:methanol:formamide (105:15:1) (Isman, pers. communication) to a distance of 15-16 cm. TNDP followed by NaOH (Brower et al., 1982) were sprayed on the plates to detect cardenolides. To standardize results within and between TLC plates and obtain a measure of polarity of the cardenolides detected, TLC results were recorded as Relative-to-digitoxin (Rp,) values:
Rp spot
RD = — X Rp of digitoxin on each side of sample
distance spot moved from origin where Rp = distance solvent front moved from origin.
The Rp of digitoxin was 0.51 + 0.003 (n=18). The cardenolides of A. speciosa and A. syriaca seeds have not been identified. In addition, since Rp, values vary between different samples of a given tissue, and between TLC plates, we used the following characteristics to identify individual cardenolides: (1) color (red or blue, red indicating a non-cardenolide); (2) RQ value; (3) pattern (based on color, relative position to other cardenolides in sample and intensity). Intensity is an indication of concentration (Brower et al_., 1982). The detection limit of the TLC assay was 0.3-0.5yjg of digitoxin. For a further indication of polarity, the RQ values of the very polar cardenolide ouabain and the nonpolar cardenolide digitoxigenin were determined and found to be 0.081, (n=3) and 1.22 + 0.010 (n=9), respectively. Although samples from males and females were assayed separately, no sexual differences were found (except where noted in Fig. 25 2.1) and results from both sexes were pooled. The cardenolide profiles were compiled from TLC analysis of from 4 to 16 different extracts of each tissue or secretion. To ensure detection of cardenolides present in low concentrations, the amount of insect material used for the different extracts of the individual tissues was varied. The following control samples from insects reared on H. annuus were assayed for cardenolide profiles: adult whole insects (1 male (M), 1 female (F)), urine/feces (10F, 10M excreting for 72 hours), eggs (100 and 200), hemolymph (2 replicates each of 1/2 extract of 33M and 25F), and dorsolateral space fluid (3F and 3M). No cardenolides were detected in any of the samples.
RESULTS 1) Total Cardenolide Content of the Fat Body, Wings, Gut and Dorsolateral Space Fluid of 0; fasciatus reared on A. speciosa The results indicate that large amounts of cardenolides do not accumulate in the fat body, wings or gut of adult 0_. fasciatus reared on A. speciosa (Table 2.1). The concentration of cardenolides in gut and wing extracts of both sexes and fat body samples of adult males was below the detection limit of the assay (6.5 x 10-6 M or 13.0 nmoles digitoxin equivalents). Fat bodies of adult females contained only 4-5% (mean 8 ^ugn) of the total cardenolide content detected in whole females. Similarly, the fat body of male and female fifth instar larvae does not appear to be a major site of cardenolide accumulation. Using 1/2 of the total cardenolide content of adults as a conservative estimate for the concentration of cardenolides in fifth instar larvae (Duffey and Scudder, 1974), the cardenolides sequestered in the fat body of such larvae account for less than 12% of the total cardenolides stored at this stage. 26
T«b1e 2.1. Total cardenolide content of adult and 5th instar Oncopeltus fasciatus samples. Cardenolide content Is measured 1n digitoxin equivalents tyjgpland 1s reported as mean *_ S.E. (range). BDL • below detec• tion limit of assay. All Insects were reared on Asdepias speciosa seeds collected 1n Pentlcton, B.C.
t In Total
each * of ^9D Per Insect per mg wet weight Sample Sample Samples or organ Insect or organ
Whole Insect (adult)
Female" 1 10 208.3 • 23.6 (100.8 - 336.1) 4.9 + 0.4 (2.9 7.5) Male' 1 10 208.8 • 21.1 (99.6 - 293.8) 5.8 • 0.6 (2.2 7.9) Female^ 1 8 152.5 • 9.6 (103.4 - 196.1) 3.4 • 0.3 (1.6 - 5.9)
Hale0 1 8 148.0 • 4.3 (66.8 - 191.6) 4.5 + 0.2 (1.3 - 7.8)
Fat Bodyc
Female adult 6-7 3 8.0 (4.0 - 10.6)0" 4.0 • 1.0 (2.0 - 5.3) Female 5th Instar 6-7 2 8.4 (8.0 - 9.0) Male adult 6 3 BDL1' BDL Male 5th Instar 7 2 5.3 (4.9 - 5.6)° 1.4 • 0.1 (1.3 1.5) Gut With Contents (adult)c Female 4 2 BDL BDL Male 4 2 BDL BDL Wings (set « fore • h1nd)b Female 1 set 4 BDL BDL Male 1 set 4 BDL BDL Dorsolateral Space Fluid (adult)1* Female 1 ^1 9 69.8 - 104.7« Male 1 ^Jl 11 88.0 - 132.0« * Reared on seeds collected September 1981 b Reared on seeds collected September 1982. c Reared on seeds collected September 1980. d Calculated by multiplying digitoxin equivalent per mg wet weight by mean fat body mass/insect. e Calculated by multiplying ^ug digitoxin equivalent per^jl dorsolateral space fluid by estimated volume of dorsolateral space fluid/insect (2-3^jl; Duffey and Scudder, 1974). Mean determined concentration of cardenolides in dorsolateral space fluid for adult females and males is 34.9 * 4.4 (range 24.2 - 67.0) and 44.0 + 3.9 (range 17.0 - 58.3)yigo/ul fluid, respectively. 27 In both sexes, cardenolide accumulation in the fat body appears to change during development from fifth instar larvae to adult (Table 2.1). The fat body of female fifth instar larvae and adults contained similar total amounts of cardenolides (mean 8.4, range 8.0 - 9.0 and mean 8.0, range 4.0 - 10.6, jigp/fat body, respectively). However, when expressed as mg per wet weight, on average 1/3 more cardenolides were detected in the adult (mean 4.0, range 2.0 - 5.3, vs mean 2.5, range 2.4 - 2.7,jugn/mg fat body). Measureable amounts of cardenolides were detected only in the fat body of male fifth instar larvae: none was detected in the adult. In both the fifth instar and adult, the male fat body contained lower concentrations of cardenolides than the female. The cardenolide content of the fat body of adult females may be affected by reproductive state. The cardenolide content of three samples of 6-7 fat bodies was determined; the average cardenolide content of the fat bodies in two of the samples was almost twice that found in the third sample (9.1 and 8.2 vs 4.8 jugn/fat body). In the third sample, one haM of the insects used had well developed ovarioles and eggs, whereas none of the individuals in the other two samples contained eggs in the ovarioles and were probably young females prior to egg development. The greatest concentration of cardenolides was found in the dorsolateral space fluid, with a mean of 34.9 +_ 4.4 (range 24.2 - 67.0) ^ign/^ul fluid in females and a mean of 44.0 + 3.9^jgp//Jl (range 17.0 - 58.3) fluid in males. The difference between the cardenolide content of dorsolateral space fluid in males and females is not significant (Scheffe's test for multiple comparisons). Using 2-3 JJI as an estimated total volume of dorsolateral space fluid in adult _0. fasciatus (Duffey and Scudder, 28 1974), the cardenolides in the dorsolateral space fluid account for 46-89% of the cardenolides in the insect. The sequestration sites of the cardenolides unaccounted for in this study of cardenolide distribution in tissues of fJ. fasciatus are unknown. It is possible that the cardenolides in the dorsolateral space fluid account for a greater percentage of the cardenolides in the insect than we estimated, since it is difficult to determine the total vacuolar volume of the inner epithelial layer of the epidermis. The total cardenolide content of adult 0. fasciatus was determined for insects reared on A. speciosa collected from Penticton, B.C. in September 1981 and September 1982. Less cardenolides were sequestered by insects reared on seeds collected in 1982 (mean 150 (range 67 - 196) jig^ vs. mean 208 (range 100 - 336) jug^/insect). The cardenolide content of males and females did not differ significantly on a per insect basis when reared on seeds collected either year or on a wet weight basis for insects reared on Sept. 1981 seeds (Student's t-test). However, the cardenolide content/mg wet weight of males was significantly greater than females for insects reared on September 1982 seeds (4.5 vs 3.4 ^ign,/mg wet weight, p<0.01, Student's t-test). 2) Cardenolide Profiles of 0. fasciatus Reared on A. speciosa and A. syriaca Seeds To investigate the potential capabilities and limitations of cardenolide sequestration in fj. fasciatus, the differences and similarities in the cardenolide array of dorsolateral space fluid and adult extracts from insects reared on the seeds of two different species of milkweed (A. speciosa collected in Penticton, B.C. and A. syriaca collected in Willimantic, Connecticut) were determined by thin-layer chromatography 29 (TLC) analysis. Since only minor differences were detected between males and females (6,(j), Fig. 2.1), data from both sexes were pooled. Table 2.2 lists the RQ values and identities of the cardenolides in each profile. The cardenolides of A. speciosa and A. syriaca seeds have not been identified, therefore, cardenolides were identified in this study by color, Rn. value and pattern (based on color, relative position to other cardenolides in sample and intensity). Cardenolides of a wide polarity range were available from A. speciosa seeds (RQ 0.14 - 1.31; A, Fig. 2.1). Insects reared on this food source, however, preferentially sequestered cardenolides of a more limited polarity range: very nonpolar (RQ> 1.20) and polar cardenolides (RQ<0.50) were absent or in low concentrations.^, C, Fig. 2.1, Table 2.2). A nonpolar cardenolide (RQ 1.32) was detected in male fJ. fasciatus and polar cardenolides were present in the fat body, but these were in very low concentrations (see Section 3). A more limited polarity range of cardenolides was present in A. syriaca seeds (RQ 0.38 - 1.24; D, Fig. 2.1). Very nonpolar cardenolides were not detected in insects reared on this food source either. However, polar cardenolides (RQ<0.50) were sequestered in greater concentrations in insects reared on A. syriaca than in those reared on A. speciosa (E, F, Fig. 2.1). r The cardenolide of highest concentration in insects reared on either seed was of intermediate polarity (RQ approx. 0.62,0, Fig. 2.1, Spot 19, Table 2.2). Another cardenolide of intermediate polarity (RQ approx. 0.57, Spot 16, Table 2.2;©, Fig. 2.1) and less polar cardenolides in the same polarity range as digitoxin (RQ 1.0,@, Fig. 2.1) were concentrated in insects reared on either species of milkweed seed. In addition, fairly nonpolar cardenolides (RQ 1.12 & 1.17, Spot 32 & 33, Table 2.2) were 30 FIGURE 2.1. Cardenolide profiles of A. speciosa and A. syriaca seeds and adult extracts and dorsolateral space (DLS) fluid of 0_. fasciatus reared on each seed. The relative concentrations of the cardenolides of the different samples are not comparable. Symbols: red (non-cardenolide);Cj , faint blue;Q » blue;^§), dark blue;^, very dark blue,^, dark blue or very dark blue depending on the extract; /\t seen in only one extract; "j-* seen in >50% of extracts;^, seen in all samples; (5^, seen in male only; ^ , seen in female only; J , light blue tailing. Sample sizes: A. speciosa seeds and samples of insects reared on A. speciosa as described in Figure 2.2. A. syriaca seeds, N=4; Adults, N=8 (4C)!,4c)); Dorsolateral space fluid, N=8 (40^4^), fluid from 1-3 adults per sample. v 31 A. speciosa reared A. syriaca reared 1.50 / \ o+ 1.25 digitoxigenin o o+ O* O o+ o o+ ®* digitoxin 1.00- >* O 0+ o+ 0+ o+ Q o* o+ o+ o o § 0.75 o* CD o o* •* •+ o 0.50 o o* o* o* o* 0.25 O 8: ouabain O 0 A B C D E F Standards A. speciosa Adult DLS A. syriaca Adult DLS Seeds Fluid Seeds Fluid TABLE 2.2. RQ values of cardenolides and non-cardenoUdes (NC)A detected In extracts of 0. fasciatus and Its food sources, A. speciosa or A. syriaca. (N) » number of samples. Rp values are reported as mean *_ S.E. Extracts from Insects Extracts from Insects reared on A. speciosa seeds reared on A. syriaca seeds A. speciosa 0. fasciatus Dorsolateral Fat body A. syriaca 0. fasciatus Dorsolateral TLC spot seeds adults space fluid Hemolymph Eggs (females) Urine/feces seeds adults space fluid number (9) (10) (15) (4) (7) (7) (16) (4) (7) (8) _ 1 0.00 • .000 — 2 - - - - 0.05 4 .003 - - - - 3 0.14 • .005 ------4 NC - - - - 0.14 4 .004 - - - - 5 0.19 • .002 - - - 0.19 4 .005 0.18 4 .003 - - - 6 0.22 .003 ------7 MC - - - - 0.23 4 .009 - - - - 8 0.26 4 .004 ------9 - - - - 0.31 4 .005 - - 0.27 4 .006 - 10 NC 0.32 • .008 ------11 0.35 + .002 - - - 0.36 4 .005 - - - - 12 0.40 4 .009 - - - 0.41 0.42 4_ .003 0.38 4 .007 0.34 4 .004 0.34 4 .005 13 NC 0.41 • .007 ------14 0.46 • .006 - - - 0.48 0.46 4 .005 0.46 - 15 0.52 4 .012 - - - - 0.52 4 .007 - - - 16 - 0.56 • .008 0.58 • .007 0.56 • .009 0.55 4 .008 0.56 4 .005 0.57 4 .007 0.50 4 .004 0.52 4 .004 17 0.59 .016 ------IB NC - - - - - 0.60 4 .010 - - - 19 - 0.64 4_ .007 0.66 • .008 0.66 + .009 0.63 4 .008 0.64 4 .003 0.61 4 .006 0.65 4_ .007 0.59 4 .005 0.62 4 .012 20 NC 0.63 • .011 ------21 ...... o.66 22 NX ------0.67 23 0.71 • .006 ... 0.71 ... 0.70 • .008 24 - 0.73 + .009 - - 0.76 + .009 - 0.75 • .015 - - 0.73 + .008 25 - 0.80 • .009 0.80 • .008 0.79 .007 0.81 0.78 - 0.78 • .003 0.79 • .004 0.79 +_ .007 26 ... 0.86 • .010 0.86 • .002 - 0.86 • .008 ... 27 - 0.89 + .009 0.90 • .006 - 0.88 - - 0.87 • .003 0.90 • .003 0.87 • .009 28 ... 0.94 • .010 0.94 ..... 29 - 0.96 • .005 0.97 • .003 - 0.96 + .006 0.96 0.95 + .009 - 0.97 • .009 0.97 •• .007 30 - 1.04 • .012 1.02^.005 1.00 006 1.02 • .010 1.02 1.01 +.018 - - 1.02 • .007 31 - - - 1.08 + .003 ..... 32 1.10 +.009 1.14 + .010 1.13 +.009 1.11 + .004 - - 1.12 +.004 - 1.10 • .007 1.12 • .008 33 1.19 - 1.20 +.009 - - - - 1.16 + .006 - 1.17 • .007 34 ------1.24 • .007 35 1.31 • .014 1.32 -- OO OJ Total number of cardenolides 13 9 8 7 11 13 12 8 11 10 a NoncardenoHdes (NC) are compounds that turned red when sprayed with colorimetric reagents, cardenolides color blue or purple. 34 concentrated in insects reared on A. syriaca. Cardenolides of these Rn. values were present in A. speciosa reared insects, but were in lower concentrations. In A. syriaca reared insects, four cardenolides in the Rn range of 1.0-1.2 were concentrated in the dorsolateral space fluid, but, only two diffuse spots were detected in this range in whole adult extracts. It is possible that all four cardenolides were present in the adult extracts, but could not be separated owing to interference from waxes, lipids or other compounds in the adult extracts. When reared on either A. syriaca or A. speciosa seeds, (D. fasciatus sequestered most of the individual cardenolides present in the seeds. Nine of the thirteen cardenolides detected in extracts of A. speciosa seeds and six of the eight cardenolides detected in extracts of A. syriaca seeds corresponded to cardenolides in insects reared on these seeds (Fig. 2.1 & 2.2, Table 2.2). However, twelve cardenolides detected in insects reared on A. speciosa did not correspond to A. speciosa cardenolides, ten of these occurring in Rn, ranges where no cardenolides were detected in the seeds (RQ 0.00-0.05 and 0.75-1.05, Fig. 2.1 and 2.2, Table 2.2). Similarly, eight of the cardenolides found in insects reared on A. syriaca did not correspond to A. syriaca seed cardenolides (Fig. 2.1, Table 2.2). Four of these insect cardenolides occurred in Rn. ranges where no cardenolides were detected in A. syriaca seeds (Rn. 0.27 and 0.97-1.12). The presence of cardenolides in 0. fasciatus that were not detected in its food source indicates metabolism of seed cardenolides in the seeds by salivary enzymes of 0. fasciatus, metabolism within the insect, or concentration in the insect of seed cardenolides that were in such low levels in the seeds they were not detected in TLC analysis. 35 FIGURE 2.2. Cardenolide profiles of 0. fasciatus adults, tissues and secretions and A. speciosa seeds. The relative concentrations of the cardenolides of the different samples are not comparable. Symbols as described in Figure 2.1; ^ , red tailing. Sample sizes: A. speciosa seeds, N=9, 0-5 replicates per sample; Adults, N=10 3o), 1-5 adults and 0-1 replicate per sample; Dorsolateral space (DLS) fluid, N=15 (76,8(j)), fluid from 1-3 individuals and no replicates per sample. hemolymph from 24-30 individuals and 3 replicates per sample; Eggs, N=7, 50-200 eggs and no replicates per sample; Fat body, N=7 (all females), 10-15 fat bodies and 0-1 replicate per sample; Urine/feces, N=16 (T3,9(J}), 180-1920 excrement hours (EH) = # of insects excreting x # hours insects excreted, 0-4 replicates per sample. 1.50 r OcV O digitoxigenin 1.25- O o+ o o+ o+ o+ o+ o+ ©• digitoxin O* 1.00- o O o+ o+ 8: o o+ Q o+ o DC o+ o+ o+ O o* o o 0.25 O 8: •+ ouabain O o+ 0 A. speciosa Adult DLS Hemo Eggs Fat Urine/ Standards Seeds Fluid lymph Body Feces 37 The cardenolide profiles of adult extracts and dorsolateral space fluid of insects reared on the two different milkweed seeds were quite similar. Nine of the ten cardenolides detected in adults and dorsolateral space fluid of insects reared on A. speciosa seeds corresponded closely to cardenolides found in the same extracts of A. syriaca reared insects (Table 2.2). Only four of the thirteen cardenolides present in adult extracts and dorsolateral space fluid of fj. fasciatus reared on A. syriaca were not found in the same extracts of A. speciosa reared insects. However, they were detected in low concentrations in extracts of other tissues of A. speciosa reared insects (Spots 9, 12, 14, 24; Table 2.2). In addition, the highly concentrated cardenolides in insects reared on A. speciosa (fe *®> Fig. 2.1; Spots 16, 19, 30, Table 2.2) corresponded to three of the most concentrated cardenolides in insects reared on A. syriaca. The similarities in the cardenolide profiles of fj. fasciatus reared on the two different milkweed seeds do not reflect similarities in the cardenolide array of the two seeds. Only two of the nine cardenolides found in common in adult extracts and dorsolateral space fluid of fj. fasciatus reared on either food source were found in both species of milkweed seeds. Furthermore, two of the highly concentrated cardenolides in A. syriaca reared insects were also the most concentrated cardenolides in their food source (Spots 16 & 19, Table 2.2), whereas these same cardenolides, although highly concentrated in A. speciosa reared insects, were not highly concentrated in A. speciosa seeds. 38 3) Cardenolide Profiles of Adult Extracts, Tissues and Dorsolateral Space Fluid of 0. fasciatus Reared on A. speciosa seeds To determine if different cardenolides were preferentially sequestered in different tissues of 0_. fasciatus, the cardenolide array of adult extracts, dorsolateral space fluid, urine/feces, fat body, hemolymph and egg extracts of insects reared on A. speciosa seeds was investigated by TLC analysis. The cardenolide profiles of all the extracts tested were similar in terms of the individual cardenolides sequestered and their relative concentrations (Fig. 2.2, Table 2.2). The cardenolide array of both sexes was analyzed for each extract, but, since only minor differences were detected between male and female samples ($, (j), Fig. 2.2) the data from both sexes were pooled. The cardenolide arrays of the adult, dorsolateral space fluid, hemolymph and eggs showed the greatest similarity: 7/8, 5/7 and 7/11 (87, 71 and 64%) of the cardenolides detected in the dorsolateral space fluid, hemolymph, and eggs, respectively, corresponded to the adult cardenolides. The most concentrated cardenolide in the insect (Rn 0.64) was also the most concentrated cardenolide in the dorsolateral space fluid, hemolymph and eggs. This cardenolide was also concentrated in the fat body and urine/feces, but both extracts contained two more polar cardenolides of similar concentration. The cardenolide arrays of the fat body and urine/feces were similar to the other insect tissues in the intermediate polarity range Rn 0.55-1.10, but differed owing to the presence of low concentrations of cardenolides in the polar range Rn<0.50 and the concentration of both intermediate and more polar cardenolides (A ®, Fig. 2.2). Large amounts of urine/feces, fat body and eggs were needed to obtain extracts with cardenolide 39 concentrations high enough to visualize in TLC analysis. As a consequence, the detection of cardenolides in low concentrations was enhanced in these tissues and may explain, in part, the greater number of cardenolides detected in these samples and the detection of cardenolides in the polar range RQ<0.50 in fat body and urine/feces samples. Large sample sizes could not be used for the other insect samples because of severe tailing of highly concentrated cardenolides. Cardenolides in the polar range RQ<0.50 may be present in low concentrations in the other samples since faint blue tailing was detected between the origin and RQ 0.55 in most samples (j , Fig. 2.2). Alternatively, this tailing may represent non-specific binding of cardenolides to various compounds such as proteins, waxes, pigments or lipids. Very nonpolar cardenolides were absent or in low concentrations in all extracts of the insect. Adult and dorsolateral space fluid extracts contained cardenolides of the lowest polarity (RQ 1.32 and 1.20, respectively), which correspond to the most nonpolar cardenolides detected in the seeds. However, the very nonpolar cardenolide (RQ 1.32) in the adult male was apparently in very low concentrations: it was only detected in a pooled sample of 5 males, never in extracts of single males. Low concentrations of very nonpolar cardenolides may also occur in the fat body since extracts produced faint blue tailing from RQ 1.04 to approximately 1.32 (j , Fig. 2.2). With the exception of urine/feces, the array of cardenolides detected in different extracts of the same insect tissue was quite constant (Table 2.3, + and * in Fig. 2.1 and 2.2). Much of the variability between extracts of the same tissue was the result of the difficulty in detecting cardenolides of low concentrations; owing to their weak intensity and/or 40 Table 2.3. The variability in cardenolide profiles of fj. fasciatus and A. speciosa. Variability is expressed as the percentage of individual cardenolides seen in 50% or more of the extracts of each sample. Sample % A. speciosa seeds 56 Urine/feces 25 Adults 70 Eggs 64 Hemolymph 100 Dorsolateral space fluid 100 Fat body 62 V 41 masking by tailing of cardenolides of higher concentrations they were difficult to detect. However, most of the cardenolides detected in urine/feces occurred in less than half of the extracts tested, indicating a real difference in cardenolide profiles between samples. In addition, the cardenolides of greatest concentration varied between samples of urine/feces Fig. 2.2), whereas, in different extracts of all other tissues the cardenolide of greatest concentration was constant (fe Fig. 2.1 and 2.2). 4) Geographic Differences in the Cardenolide Profile of A. syriaca seeds Previous studies of sequestration of Asclepiad cardenolides in CJ. fasciatus used A. syriaca seeds from Missouri and Ontario as the insects' food source (Feir and Suen, 1971; Duffey and Scudder, 1974). It is possible that differences in the results between these studies and the present study are a result of geographic variation in the cardenolide profile of seeds of A. syriaca. Geographical variation has not been established for the seeds of A. syriaca, therefore, I investigated this by determining the cardenolide array of A. syriaca seeds from the following four locations: Montebello, Quebec, (collected October, 1976); Ottawa, Ontario (collected September, 1976); Cleveland, Ohio (collected September, 1975); and Willimantic, Connecticut (collected September 1981). The cardenolide array of the seeds collected in Ontario, Ohio and Quebec showed only minor differences in the individual cardenolides present and their relative concentrations, but were dissimilar to the cardenolide profile of the Connecticut seeds (Fig. 2.3). 42 1.50 r 1.25 O* digitoxigenin O O* o* o* o* digitoxin 1.00 O O o* DC o* OA A O ©* •* ©• OA 0.50 0* o* ©« ©* o* ®* )* 0.25 - ouabain O Connecticut Ontario Ohio Quebec Standards FIGURE 2.3. Cardenolide profiles of A. syriaca seeds from four geographic locations. Symbols as described in Figure 2.1. Sample sizes: N=2 for Ontario, Ohio and Quebec seeds; N=4 for Connecticut seeds. 43 DISCUSSION This study was undertaken to clarify certain aspects of the differential distribution of natural cardenolides in 0. fasciatus, both in terms of quantity and polarity, and to document some of the capabilities of the sequestration process of cardenolides in 0_. fasciatus. My results show that large amounts of cardenolides do not accumulate in the fat body, gut and wings of 0. fasciatus; the greatest amount of cardenolides in the insect is in the dorsolateral space. The distribution of cardenolides in 0_. fasciatus is summarized in Table 2.4. Blum (1983) has suggested that each species of insect has a unique process of sequestration, tolerance and defensive use of the toxins ingested with its food plant. The differences in the distribution of cardenolides in 0_. fasciatus and two other insects, Danaus plexippus and Cycnia inopinatus, support this argument. Low amounts of cardenolides are sequestered in the wings, fat body, gut and hemolymph of 0_. fasciatus, whereas, substantial amounts of cardenolides are found in these tissues in the monarch (see Blum, 1981, 1983). C_. inopinatus differs from both the monarch and 0. fasciatus in the presence of large amounts of cardenolides in the hemolymph and wings, but negligible amounts of these compounds in the gut (see Blum, 1983). It appears that 0. fasciatus, unlike £. plexippus and C_. inopinatus, may not tolerate large quantities of cardenolide in the hemolymph. Thus, the rapid uptake and accumulation of cardenolides in the dorsolateral space may function in the insect's tolerance of sequestered cardenolides by maintaining low levels of cardenolides in the hemolymph (Duffey et aj_., 1978) and other tissues, i.e. fat body and wings (this study). Furthermore, the low cardenolide content of the wings in 0. fasciatus indicates that the anti-predator strategies 44 Table 2.4. Distribution of cardenolides in Oncopeltus fasciatus expressed as % of the adult total. BDL = below detection limit of assay (13 nmoles or 6.5 x IO-6 M digitoxin equivalents). Sample % of Adult Total Reference9 Dorsolateral space 60-95 1 Dorsolateral space fluid 46-89 2b Hemolymph BDL 1,2C Urine/feces BDL 1 Metathoracic gland BDL 1 Fat body (male) BDL 2 Fat body (female) 4-5 2 Gut with contents BDL 2 Wings BDL 2 a 1 = Duffey and Scudder, 1974 using Asclepias syriaca for the food source. 2 = This study, using A. speciosa for the food source. b Calculated using estimate of 2-3 jul for volume of dorsolateral space fluid in adult (Duffey and Scudder, 1974). c Unpublished results. 45 of fJ. fasciatus may differ from D_. plexippus and C. inopinatus. In the monarch, the greatest concentration of cardenolides is in the wings and this is thought to function as an anti-predator strategy, causing avian predators to reject the butterfly relatively unharmed and preventing attack of the more critical areas, the thorax and abdomen (Brower and Glazier, 1975). It is unlikely that the wings of fJ. fasciatus would function effectively in this manner. Owing to the small body size of fJ. fasciatus 17-18 mm) and the positioning of the wings at rest flat along the back, large predators probably snatch the entire insect upon attack. Thus, for fj. fasciatus, concentration of cardenolides in the dorsolateral space and release of dorsolateral space fluid along the thorax and abdomen is probably a more effective anti-predator strategy (Duffey and Scudder, 1974; Scudder and Meredith, 1982a). The results from the TLC analysis indicate that there are only minor differences in the cardenolides sequestered and concentrated in various tissues and secretions of CJ. fasciatus reared on seeds of a single species of milkweed. The presence in the fat body of low concentrations of polar cardenolides not detected in other tissues may indicate a difference in the cardenolide array sequestered in this tissue. Alternatively, this difference may be the result of enhanced detection of cardenolides in low concentrations in the fat body (see Results). Only minor differences were detected in the cardenolide profiles of adult extracts and dorsolateral space fluid of CJ. fasciatus reared on two food sources which appear to have different cardenolide arrays. This indicates that the cardenolide array of the insect can remain fairly constant despite differences in the cardenolide profiles of its food plants. These results differ from studies in which variation in 46 cardenolide profiles of monarchs has been related to interspecific differences in the cardenolide array of their food plants (Roeske et al., 1976; Brower et al., 1984a). However, two other studies of sequestration of cardenolides in 0. fasciatus reared on A. syriaca seeds (Feir and Suen, 1971; Duffey and Scudder, 1974) suggest that the cardenolide profile of 0_. fasciatus may exhibit differences related to intraspecific variation in the cardenolide profiles of its food plant. In my study seven cardenolides were detected in insects reared on A. syriaca seeds collected in Connecticut and most of the cardenolides in the seeds were sequestered in the insect. In contrast, Feir and Suen (1971) found that only four cardenolides were sequestered in insects reared on A. syriaca seeds collected in Missouri and only one of these corresponded to a seed cardenolide. I found only minor differences in the cardenolide profiles of adult extracts and dorsolateral space fluid of insects reared on A. syriaca seeds from Connecticut, and that the highly concentrated cardenolides in the insect covered a wide range of polarities. However, insects reared on A. syriaca seeds from Ontario exhibited a predominance of polar cardenolides and the cardenolides of the dorsolateral space fluid exhibited a much smaller polarity range than seen in the whole insect (Duffey and Scudder, 1974). The differences in the cardenolide profiles of 0_. fasciatus in these studies may be explained, in part, by the different extraction methods, solvent "systems and detection reagents used in the TLC analyses. In addition, the differences in the cardenolide arrays of the insect may reflect variation in the cardenolide profiles of the A. syriaca seeds the insects were reared on. Geographic variation in the cardenolide array of vegetative parts of A. syriaca has been well established (see Roeske et al_., 1976 for review). My results establishing a different cardenolide array in A. syriaca seeds collected in Connecticut than in 47 seeds of A. syriaca collected in Quebec, Ontario, and Ohio, and the differences in the numbers of cardenolides detected in the seeds of the different populations of A. syriaca used in the studies mentioned above, indicate that geographic variation also occurs in the seeds of A. syriaca. In vivo and in vitro evidence with two non-Asclepiad cardenolides, ouabain and digitoxin, indicates that differential excretion and metabolism of cardenolides in CJ. fasciatus as well as preferential uptake of individual cardenolides across the gut and into the dorsolateral space are involved in the selective sequestration of cardenolides in £. fasciatus (Scudder and Meredith, 1982b; Duffey et ^1_., 1978; Meredith et al_., 1984). However, ouabain and digitoxin do not occur in the host plants of fj. fasciatus. The present study provides evidence for differential excretion and metabolism in CJ. fasciatus of cardenolides present in its natural food plants. The presence of many cardenolides in CJ. fasciatus that were not detected in its food sources suggests metabolism of seed cardenolides or concentration in the insect of seed cardenolides present in such low concentrations they were not detected by TLC analysis. Metabolism of seed cardenolides may occur in the seed while it is being digested by the saliva of the insect, in the insect gut by bacteria or the gut milieu, and/or in other tissues within the insect. Two other insect species are known to metabolize cardenolides ingested with their food plant (Seiber et al., 1980; Brower et al., 1982; Levey, 1983): in one of these, the monarch, homogenates of both the gut and fat body metabolized the Asclepiad cardenolide, uscharidin (Marty and Krieger, 1984). Very polar and very nonpolar cardenolides were absent or in very low concentrations in fJ. fasciatus reared on A. speciosa or A. syriaca. The differential excretion of large amounts of intermediate and higher polarity 48 cardenolides relative to cardenolides of lower polarity in the urine/feces may explain, in part, the low levels of polar cardenolides in 0_. fasciatus when feeding upon seeds of A. speciosa. Rapid metabolism of nonpolar cardenolides in the insect may also explain the absence of these cardenolides in 0. fasciatus. Seiber et al_. (1980) have shown in the monarch that rapid metabolism of several less polar cardenolides to more polar metabolites results in the absence of or very low concentrations of these cardenolides in the larval tissue. The sequestration and concentration of cardenolides of a wide polarity range in _0. fasciatus reared on A. syriaca or A. speciosa suggests that cardenolides are not sequestered in the insect simply on the basis of polarity. The importance of physical-chemical characteristics other than polarity in cardenolide sequestration in Q_. fasciatus is also indicated by the sequestration and concentration of intermediate and more polar cardenolides in the fat body of 0_. fasciatus reared on A. speciosa. This was unexpected since the fat body often accumulates nonpolar compounds and toxins owing to their lipophilic nature (Kilby, 1963), and earlier observations indicated preferential sequestration of the nonpolar cardenolide digitoxin in the fat body of adult 0. fasciatus (Duffey et al., 1978). The importance of physical-chemical characteristics other than polarity in cardenolide sequestration has also been indicated in the monarch butterfly (Seiber ^t^., 1980; Brower et al., 1982) and Seiber et al., (1980) have suggested that physical-chemical characteristics that influence the chemical stability, solubility and ease of transport of individual cardenolides and their binding affinity to blood proteins and sequestration systems in the insect are involved. In summary, this study provides evidence that cardenolides are 49 differentially distributed in fJ. fasciatus in terms of total quantity. In contrast, only minor differences were detected in the cardenolide array of adults and five tissues and secretions of fJ. fasciatus. This study also provides evidence that cardenolides of a wide polarity range can be sequestered in fj. fasciatus, however, very nonpolar and very polar cardenolides are not sequestered or are present in extremely low concentrations. In addition, the results suggest that metabolism and differential excretion of cardenolides may be part of the selective sequestration process of cardenolides in fj. fasciatus. Finally, the constancy of the cardenolide profiles of fj. fasciatus reared on seeds of two species of milkweed with very different cardenolide arrays, and between tissues of fJ. fasciatus reared on a single species of milkweed seeds indicates that there is qualitative regulation of the cardenolide array in CJ. fasciatus, as in the monarch (Brower et al., 1982). (Material in this chapter is in press: Moore, L.V. and G.G.E. Scudder. 1985. Selective sequestration of milkweed (Asclepias sp.) cardenolides in Oncopeltus fasciatus (Dallas)(Hemiptera: Lygaeidae). J. Chem. Ecol.) 50 CHAPTER 3: EXCRETION OF OUABAIN BY MALPIGHIAN TUBULES OF ONCOPELTUS FASCIATUS Summary An in vitro preparation of Malpighian tubules was used to investigate the excretion of the polar cardenolide, ouabain, in 0. fasciatus. Both segments of the tubules were found to metabolize ouabain. The distal segment (Segment II) secreted primary urine and ouabain. Secretion of ouabain by Segment II was not observed to occur against a concentration gradient and increased with increasing fluid secretion. The proximal segment (Segment I) reabsorbed fluid and ouabain but not metabolites. Ouabain was reabsorbed against a strong concentration gradient (23-fold), was independent of fluid reabsorption, and increased with increasing fluid secretion by Segment II. In rapidly secreting Malpighian tubules (a situation of high cardenolide secretion by Segment II), the presence of Segment I reduced the excretion of ouabain by 84 - 93%, mainly by reducing ouabain concentration. It appears excretory loss of cardenolides can be reduced in 0. fasciatus and thus may be a factor in the sequestration of cardenolides in this insect. 51 INTRODUCTION In the previous chapter, selective sequestration of host plant cardenolides in CJ. fasciatus was demonstrated. In this chapter, the secretion, metabolism and reabsorption of the polar cardenolide, ouabain, is investigated under a variety of conditions to determine the possible role of the Malpighian tubules in the selective sequestration of cardenolides in 0. fasciatus. Since the Malpighian tubules are an important part of the excretory system in insects, they are a potential site of metabolism and loss of cardenolides that might otherwise be sequestered in CJ. fasciatus. The role of Malpighian tubules in insect excretory systems has been extensively reviewed by Phillips (1981) and the gross morphology and" physiology of the Malpighian tubules of fj. fasciatus have been well described by Meredith et j»l_. (1984). However, general aspects of insect Malpighian tubules, as well as specific aspects of the Malpighian tubules of CJ. fasciatus pertinent to the work described in the present chapter will be reviewed here. In insect excretory systems, the formation of primary urine from the hemolymph occurs in the Malpighian tubules. The primary urine contains many of the solutes present in the blood and is secreted into the hindgut where, in most insect, solutes and water can be conserved by reabsorption. In general, reabsorption of solutes and fluid does not occur in the tubules. The Malpighian tubules also function in the rapid clearance of toxins from the hemolymph by actively secreting these compounds. fj. fasciatus adults commonly have four similar Malpighian tubules, approximately 15 mm in length, that empty directly into the fourth 52 ventriculus (pylorus) of the gut at their proximal ends. Each tubule consists of at least two morphologically and physiologically distinct segments. The long (11 mm) distal Segment II functions in the secretion of primary urine, which appears to be secreted by the process common to many insects, active K+ secretion with Cl" as the accompanying ion (Phillips, 1982). The proximal Segment I functions in fluid reabsorption from the Malpighian tubule lumen to the hemolymph side of the tubule. In the following study, an in vitro preparation of Malpighian tubules is used to investigate the excretion of the polar cardenolide, ouabain, in 0. fasciatus. Preliminary experiments (Meredith et , 1984) indicated that ouabain is secreted by Segment II and reabsorbed by Segment I of the tubules. In this chapter, the secretion and metabolism of ouabain by Segment II and the reabsorption of ouabain and its metabolites in Segment I is examined under, a variety of conditions. In addition, the modification by Segment I of the primary secretion from Segment II is determined. Thin layer chromatography is used to detect the presence of ouabain and its metabolites in the secreted and reabsorbed fluids of the Malpighian tubules. Ouabain with [3H]ouabain is used owing to its relatively high solubility in saline which facilitates the detection of the cardenolide and its metabolites in the extremely small volumes (nl) of fluid secreted and reabsorbed by the Malpighian tubules. MATERIALS AND METHODS 1) Insects Adult male and female Oncopeltus fasciatus taken from a laboratory culture maintained at 26 °C under a 16:8 light:dark cycle were used in all 53 experiments. The insects were reared on milkweed seeds (Asclepias speciosa). 2) Salines The salines used in this study were based on the composition of fj. fasciatus hemolymph (unpublished observations; Florkin and Jeuniaux, 1974; Staddon and Everton, 1980) and contained the following: NaCl, 20 mM; KC1, 24 mM; MgCl2, 2 mM; CaCl2, 2 mM; glucose, 6.7 mM; NaH2P04, 2.5 mM; Naj^HPO^, 3.5 mM. In later experiments glutamine, 2.7 mM; alanine, 5 mM; and proline, 5 mM were included in salines. The pH equaled 6.9 in air and osmotic pressure was varied by adding sucrose. Tubules were dissected in saline with an osmotic pressure equal to that of CJ. fasciatus hemolymph (326 mOsm). Tubules were stimulated to secrete by including cyclic AMP in the bathing salines. Cyclic AMP (2.5 x 10"^ M) served to both increase (approximately 40 fold) and prolong fluid secretion. The excretion of ouabain was followed using [^ouabain (general label, specific activity 14.0 or 18.0 Ci/mM, 98.5% radiochemical purity) obtained from New England Nuclear. 3) In vitro Malpighian tubule preparation Malpighian tubules were prepared using an in vitro technique modified from Szibbo and Scudder (1979). Tubules were dissected using fine glass needles with ends that were sealed and slightly curved in a gas flame. The required tubule segments were severed, transferred to a drop of saline held in a depression in a watch glass lined with Sylgard (Dow Corning), and covered with liquid paraffin oil (Fig. 3.1). Tubule segments were tethered at their proximal end by a fine silk ligature, then drawn out into the oil 54 reabsorbed bathing fluid saline secreted fluid igature paraffi weights oil V Segment I Sylgard resin Segment H Figure 3.1. In vitro preparation of 0. fasciatus Malpighian tubules. 55 and anchored into position by small weights. Secretion was collected by nicking the tubules with a sharp micropipette immediately distal to the 1igature. In all cases dissection required 1.5 - 2 h and time 0 was considered to be the start of collection of secretion. Experiments were conducted at 21-23 °C and the bathing saline changed hourly. Under these conditions Malpighian tubules continued to secrete for at least 18 h, although most experiments were concluded by 8 h. Volumes and rates of fluid secretion and reabsorption were calculated according to the method of Maddrell (Maddrell, 1969), from measurements of the diameters of secreted drops. Radioactive samples were either counted directly in 10 ml of commercial scintillation fluid using a Beckman LS 9000 scintillation counter or chromatographed. Correction for varying counting efficiences was performed using the "H number method" (Anonymous, 1979). Ionic composition of whole tubule secretion, reabsorbed fluid and bathing saline was determined by electron microprobe analysis as described in Strange et al_. (1982). 4) Chromatography Thin layer chromatography plates prepared with Silica Gel G (Redi/Plate, Fisher Sci. Co., gel 250 ^im thick) or cellulose (Eastman Kodak Chromatogram Sheet, adsorbent 160 ^im thick) were used. Silica Gel G plates were activated for 24 hours over concentrated H2SO4 in a closed chromatography tank prior to and 12 h after sample application. Samples 11 were applied 1 from the bottom of the plate. Fifty to seventy-fivejugm each of three cold standards, ouabain, ouabagenin and rhamnose (Sigma Chemical Co.), were spotted on top of the samples. Both ouabagenin and rhamnose are known metabolites of ouabain. Chromatography tanks were lined 56 with Whatman Chromatography Paper #1 (W&R Balston. Ltd.) and equilibrated with 300 mis of the solvent system. Plates were developed to a distance of 15 - 16 cm from the origin. To detect the cardenolides, ouabain and ouabagenin, plates were sprayed with 2.5 ml of 2,4,2',4'-tetranitrodiphenyl (TNDP, 0.5 gm/100 ml toluene) followed by NaOH (10% in MeOH) (Neher, 1969). Under these conditions, TNDP reacts with a lactone ring to form a blue color and is relatively specific for cardenolides (Nover, 1972). To detect rhamnose, plates were sprayed with a p-anisaldehyde-H2S04 reagent (Stahl and Kaltenbach, 1965). Plates could be sprayed for rhamnose then air dried and sprayed for cardenolide detection, but not vice versa. Samples were analyzed in three chromatographic systems (Table 3.1). Three additional ethyl acetate:MeOH systems were tried (EA:M 95:5, 90:10, and 50:50), however, they did not clearly separate the three standards. All calculations and statistical analysis were done with results from System I because it produced distinct spots with very little tailing. This system did not separate rhamnose and ouabain, but no peak of radioactivity indicating rhamnose was detected in any of the samples in System II and III. The plates were analyzed for radioactivity as follows. Chromatograms were divided into columns 1 cm wide extending from 1 cm below the origin to the solvent front and including the standard spots. These were divided into 0.5 cm sections, each of which was scraped into a scintillation vial containing 0.5 ml H2O. Samples were leached overnight and counted as described previously. Neither incubation in oil nor the mixing with cold Malpighian tubule secretion followed by overnight incubation in oil altered the chromatographic behavior of salines containing pure ouabain labelled with [3H]ouabain. Radioactivity in these controls and a nonincubated TABLE 3.1. Chromatographic systems used for the separation of cardiac glycosides and their metabolites. Solvent Retardation factor (Rf) System Stationary Phase Mobile Phase Front ouabain ouabagenin rhamnose Reference SIHca Gel G activated CHCl3:MeOH:H20 15-16 cm 0.33 0.42 0.33 Dutta, 1963 over H2S04 24 h prior (65:30:5) to and 12 h after applying samples (10) II Cellulose Impregnated n-Butyl alcohol 14 cm 0.45 0.64 0.34 modified from with water saturated saturated with von Schenker, 1954 with n-butyl alcohol water III Silica Gel G Ethyl acetate:MeOH 15 cm 0.39 0.13 0.70 modified from (75:25) developed 2 Brower, et al., 1982 times 58 saline control always chromatographed as a single peak coincident with cold ouabain and could be recovered with 89.9% efficiency in System I. The remaining label (10.1%) was evenly distributed throughout the chromatogram and judged to be a result of non-specific adsorption to the silica gel rather than to impurities. The amount of ouabain (nmoles) in experimental solutions (0e) was calculated as follows: _1 0e = [(Do* Dt ) + k] • De • Ds"l where D0 = disintegrations per minute (dpm) recovered from the ouabain standard spot in the chromatogram of the experimental solution D+; = dpm recovered between the origin and solvent front in the experimental solution chromatogram k = a correction for the fraction of ouabain which is non-specifically adsorbed to the TLC support (= 0.101 and is derived from (D+; - D0) • D-t~l for chromatograms of pure ouabain solutions De = dpm measured in the experimental solution Ds = dpm per nmole ouabain measured in standard ouabain solutions. From this value and the measured volume of experimental solutions, ouabain concentrations can be calculated. Metabolism (0Me), the amount of ouabain that must have been changed to give rise to the observed metabolism, was calculated as follows: 0Me = (De- Ds"l) - 0e where 0e is the amount of unchanged ouabain in the experimental 59 solution. All results are reported as mean +_ standard error of the mean. The variability among animals was found to be no greater than the variability among tubules from one animal, hence the number of tubules is reported. RESULTS 1) Characteristics of Ouabain Transport in Segment II a) Secretion of ouabain by Segment II The effect of external ouabain concentrations on the secretion of ouabain by Segment II was investigated. Urine/plasma (U/P) ratios of radioactive label and the secretion rates of fluid and radioactive label were determined in isolated Segment II's bathed in saline (321 mosmol) containing .003 - 5.0 mM ouabain (plus [3H]ouabain) + 2.5 x 10~4 M cAMP without amino acids. Fluid and radioactive label secretion rates were followed over 5 h and found to be relatively constant during the time period of the experiment, 2 - 5 h. Figures 3.2 & 3.3 summarize the results of these experiments. U/P ratios of less than 1 were always observed, even assuming no metabolism of ouabain, see Section II below (Fig 3.2,#). U/P ratios corrected for measured % metabolism are indicated (A). Increasing external ouabain concentrations did not consistently affect rates of fluid secretion although segments bathed in 1 mM ouabain secreted at reduced rates (Fig. 3.3). In preliminary experiments with unstimulated Segment II's, 1 mM ouabain did not affect rates of fluid secretion (data not shown). The effect of fluid secretion rates on the rate of ouabain secretion 60 i i i 1 1 + 1.0 0 -1.0 "2.0 "3.0 log ouabain concentration (mM) in bathing saline Figure 3.2. The effect of varying external ouabain concentration on urine-to-plasma (U/P) ratios, (%, if we assume no ouabain metabolism; A, corrected for measured ouabain metabolism) in Segment II. Each point represents mean + S.E. (where larger than symbol); number in parentheses indicates number of determinations from 4-8 tubules, h 2 and 3 are pooled. Figure 3.3. Effect of varying external ouabain concentration on fluid secretion rate, details as in Figure 3.2. 62 and its concentration in fluid secreted by isolated Segment U's was also determined. Fluid secretion rates were altered by varying the osmotic pressure of the saline (134, 223, 312, 401 mOsm). Bathing saline contained ImM ouabain (plus [3H]ouabain). The results of these experiments are shown in Figs. 3.4 & 3.5. The rate of ouabain secretion by Segment II was dependent on the rate of fluid secretion (Fig. 3.4). Corrections for metabolism (A) reduced these rates only slightly. The concentration of ouabain in secreted fluid (Fig. 3.5) approached 1 mM at the lowest rate of fluid secretion but decreased with higher fluid secretion rates. A permeability coefficient (b, cm-sec-1) was calculated using the data in Figs. 3.2 and 3.4 according to the equation b = a(U7P)(l-U/P)-1 (Ramsay, 1958) where U/P is the urine/plasma ratio of ouabain concentration and a is the rate at which water is actively pumped through unit area of wall into the lumen (^l«mrn-2.min-1, calculated by R = 2Tfrl_a where R is the rate at which urine issues from the tubule and 2irrL is the area of the tubule taken from gross morphological measurement, r = radius [1/2 the outer diameter of the whole tubule] and L = length). The permeability coefficient ranges from 0.25 - 2.7 x 10-6 cm-sec-1. Assuming constant metabolism, we found the permeability coefficient to be little changed (i.e., within experimental variability) by either changes in fluid secretion rate or bathing saline concentration of ouabain, suggesting ouabain transport is mainly by diffusion. / 63 .30 n (39) .25 cn o E c .20- C o a> u .15 ' to re ro .10- 3 O (18) "Ero .05' or 1 1 1 1 .2 .4 .6 .8 Fluid secretion rate (gl- h"1) Figure 3.4. Effect of fluid secretion rate on the rate of ouabain secretion by Segment II alone. # , label assumed to be 100% ouabain; A , corrected for measured metabolism of ouabain. Each point represents the mean + S.E. (where larger than symbol); number in parentheses indicates number of determinations from 4-8 tubules, h 3 - 5 are pooled. 64 Figure 3.5. Effect of fluid secretion rate on ouabain concentration in fluid secreted by Segment II alone. Horizontal arrow, ouabain concentration of bathing saline. Details as in Figure 3.4. 65 b) Metabolism of ouabain by Segment II When secretion collected from Segment II bathed in 326 mosmol saline containing 1 mM ouabain plus [3H]ouabain was chromatographed, 69.7 + 0.19 % of the radioactive label chromatographed as unchanged ouabain (74% in System II, 54% in System III, Fig. 3.6). Increasing the ouabain concentration of the bathing saline to 5 mM or increasing the rate of fluid secretion resulted in Segment II secreting a slightly greater percentage of unchanged ouabain (75% and 82%, respectively, in System I), although this difference was not significant (Student's t-test at the .05 level) 2) Ouabain Reabsorption in Segment I a)Direct analysis Whole tubules were placed in 326 mosmol bathing saline and the entire length of Segment I drawn out into paraffin oil. After the tubules equilibrated for 45 minutes the saline was replaced by 326 mosmol bathing saline with 1 mM ouabain plus [3H]ouabain. Tubules secreted fluid at the proximal end for at least 4 h and a reabsorption droplet often formed on the outside surface of Segment I over the first 3 h (Fig. 3.7). Tubules not secreting continuously throughout the experimental period were discarded. Reabsorption droplets were collected two hours after changing the bathing saline; whole tubule secretions were collected and analyzed every half hour during this period. Eight-six and one-half percent of the radioactive label in the reabsorbed fluid chromatographed as unchanged ouabain in System I (92% and 67% in System,II and III, respectively, Fig. 3.6). This percentage is not significantly different from the 1 mM ouabain plus [3H]ouabain saline controls (Student's t-test, p > 0.05), indicating that only ouabain was 66 Figure 3.6. Distribution of radioactivity in typical chromatograms of experimental and control fluids. Bathing saline in all cases was 326 mosmol containing 1 mM ouabain plus [3H]ouabain. Positions of origin (0 on x-axis), ouabain (clear oval), ouabagenin (cross-hatched oval), rhamnose (stippled oval) and solvent front (16, 14, 15) are indicated. Percentages indicate radioactivity co-chromatographing with ouabain standard. S, Segment II secretion; Rn, natural reabsorption droplets; R, set reabsorption droplets; W, whole-tubule secretion; C, bathing saline control. 67 SYSTEM I SYSTEM TI SYSTEM TJI CM FROM ORIGIN 68 Figure 3.7. Amount of fluid, ouabain, metabolized ouabain, and ouabain concentration in Segment II secretion, reabsorption droplets and whole-tubule secretion. Malpighian tubules were bathed in 326 mosmol bathing saline with 1 mM ouabain plus [3H] ouabain. Values are for a 2 h period (n = 8). Segment II secretion values are calculated from measured determinations of reabsorption droplets and whole- tubule secretion. Ouabain Metabol ized Ouabain pmo les ouabain concentration pmoles (%) m M 0 (0) 1.00 bathing saline segment II- 500 151.3 38.3 (20.2) 0.30 secretion segment II reabsorption J'\ 14± 3 98.8+ 23 0 (0) 6.91 droplet segment I — whole tubule ff\ secretion — 486 ± 36 29.2 ± 6 58. 1 (66.6) 0.06 ligature • 70 reabsorbed by Segment I. In contrast, only 23.3 + 0.54% of the label in whole tubule secretions chromatographed as ouabain in System I (35% and 33% in System II and III, respectively, Fig. 3.6). The amount of ouabain in whole tubule secretions increased to 67% when the tubules were bathed entirely in saline. The concentrations of unchanged and metabolized ouabain in whole tubule secretion and reabsorbed fluids were calculated using the equations given in Materials and Methods. To estimate Segment II secretion the following assumptions were made: a) the volume of Segment II secretion equals the volume of whole tubule secretion and reabsorption droplet b) the total radioactive label present in Segment II secretion equals the sum of that present in whole tubule and reabsorption droplets c) the ratio of unchanged and metabolized ouabain present in Segment II secretion equals that observed in chromatograms of secretions collected from Segment II alone. Fig. 3.7 summarizes these results. Concentrations of ouabain in the reabsorption droplets (6.9 mM) were substantially higher than those in whole tubule (.06 mM) or Segment II (.30 mM) secretion or in the bathing saline (1.0 mM). Sixty-five percent of the ouabain and 3% of the fluid secreted by Segment II was reabsorbed by Segment I. A single Malpighian tubule can remove 189.6 (151.3 + 38.3) pmoles of ouabain from the bathing saline in 2 h. Approximately 31% of this was metabolized: 20% by Segment II (38.3/189.6) and 10% by Segment I (58.1 -38.3/189.6), 52% (98.8/189.6) was reabsorbed by Segment I, while 15% (29.2/189.6) was finally excreted into the pylorus. The amount of ouabain reabsorbed may be positively correlated with whole tubule secretion rate (r= .95 excluding possible atypical tubules at fluid secretion rates of 600 nl, Fig. 3.8) but is 71 .3-, .21 100 200 300 400 500 600 700 Fluid secreted in two hours (nl) Figure 3.8. Amount of fluid secreted by whole tubule in 2 h in relation to ouabain reabsorption by Segment I during same time period. Values are for individual tubules. 72 independent of fluid reabsorption rate (data not shown). Ionic composition of whole tubule secretion, reabsorbed fluid and bathing saline was determined in these experiments by electron microprobe analysis. The results are shown in Table 3.2. The ionic composition of the three fluids differed substantially, indicating selective secretion and reabsorption of ions and/or differing permeabilities of the two segments. b) Set droplets To further investigate reabsorption along the length of Segment I and to provide controls for the direct analysis of reabsorption previously described, a second experiment was conducted. Tubules were set up as before and small (6 - 18 nl) droplets of saline without cAMP or ouabain were placed along the length of Segments I and II as diagrammed in Fig. 3.9. Droplets could be retrieved from tubules after 3 min with little change in volume. Although placed relatively close together they did not coalesce with each other or with the bathing saline or secreted fluid. Similar sized droplets incubated in paraffin oil did not change volume or ouabain concentration over the experimental period. These controls indicate that measured fluid volumes of set and reabsorbed droplets were not affected by our collection techniques, evaporation or coalescing of droplets. Droplets placed on the outside of Segment I immediately wet the surface of the tubule while those on Segment II did not. After 2 h, set droplets were removed from tubules and analysed. The results are presented in Figs. 3.6 and 3.9. The volume of droplets placed on the outside of Segment II decreased (indicated by a negative sign) over 2 h. Little radioactivity was detected in these droplets despite a chemical gradient favouring ouabain entry from the lumen. All droplets placed on the outside 73 Table 3.2. Ionic composition (mM) of whole-tubule secretions, reabsorbed fluid and bathing saline. Ion Reabsorbed fluid Whole tubule secretion Bathing saline i*l (n=3) (n - 3) (n - 1) Ca2+ 1.53 • 0.55 0.76 + 0.10 1.58 Na+ 52.71 6.65 8.50 • 0.55 27.66 65.35 + 7.88 170.37 4 3.86 23.06 ci- 119.83 + 11.68 126.53 4 0.47 43.33 2.84 + 0.77 3.28 4 0.17 1.93 Total S 2.03 • 0.58 1.34 4 0.09 0.16 Total P 10.91 • 1.43 37.28 4 2.91 5.96 Values are mean 4 SE. S, sulfur, P, phoshorus. 74 Figure 3.9. Change in fluid volume, radioactive label, and radioactive label concentration in four set droplets and whole-tubule secretion over 2 h. Negative values indicate a decrease in fluid volume. Malpighian tubules were bathed in 326 mosmol bathing saline containing 1 mM ouabain plus [3H]ouabain. Values are mean + S.E. (n= 8). Radioactive Radioactive label Fluid label concentration nl pmoles mM bathing saline -4.9± .6 0 0 segment II- "5.4 ± 1.1 0.3+0.2 0 +2.4+ 1.2 37.1 + 6.8 15.4 segment I +4.4+1.4 4.7±0.9 1.1 whole tubule secretion 110.8 ±27.4 13.412.5 0.1 I igature • 76 of Segment I increased in volume after 2 h. The concentration of radioactivity in this reabsorbed fluid was always higher than that in the bathing saline and chromatographic analysis (Fig. 3.6, System I and II) showed that all of the radioactive label was ouabain. In all cases it appeared most of the ouabain (approximately 88%) was reabsorbed from the lumen by the distal droplet thus reducing the amount of ouabain "seen" by the more proximally set droplet on Segment I. 3) Modification of Segment II secretion by Segment I In the previous experiments only the luminal side of Segment I was bathed in saline. Moreover reabsorption droplets covered only about 10% of the tubule surface. Both of these factors could result in low estimations of reabsorption. To overcome these problems we were able to analyze tubules sequentially since preliminary experiments had indicated that at fast secretion rates both whole tubules and Segment II's secreted fluid and ouabain at relatively constant rates between 2 - 5 h. Accordingly secretion was collected from whole tubules (less the 1.5 mm length of Segment I required to separate secretion) bathed entirely in 132 mosmol saline over the first 3 h. Then Segment I was withdrawn as described previously and secretion collected from Segment II alone. Fig. 3.10 & 3.11 present the results of these experiments. The presence of Segment I sharply reduces the rate of radioactive label excretion in whole Malpighian tubules by 85 - 94% (p < .001, analysis of variance, Scheffe's test, Fig. 3.10). Fig. 3.11 indicates this reduction is accompanied by a reduction of the ouabain concentration in the excreted fluid by about 71% (p < .001, analysis of variance, Scheffe's test). Chromatography of whole tubule and Segment II secretion produced under 77 ^Segment I,U ^Segment CD 1 2 3 4 5 Time Ch) Figure 3.10. Rate of ouabain secretion with time. At 3 h (stand dotted line) Segment I is removed. Traces are for individual tubules bathed in 132 mosmol saline containing 1 mM ouabain plus [3H]ouabain. 78 .Segment I,IL Segment H. > <- 0.2 H o CC c O C O 0.1 H o 03 .Q CD ZJ O T 1 2 3 4 5 Time Ch) Figure 3.11. Ouabain concentration in secreted fluid with time. At 3 h ( and dotted line) Segment I is removed. Details as in Figure 3.10. 79 these conditions of fast secretion rates indicated 58 and 82% respectively of the radioactive label was unchanged ouabain. Although the percentage of metabolized ouabain is greater in whole tubule secretion than in isolated Segment U's, this can be more than accounted for by reabsorption of unchanged ouabain, suggesting under these conditions little metabolism is occurring in Segment I. DISCUSSION This study demonstrates that in in vitro preparations the distal Segment II of the Malpighian tubules of 0. fasciatus secreted primary urine and ouabain into the Malpighian tubule lumen. The high K+/Na+ ratio and high CI" concentration of the fluid secreted by CJ. fasciatus Malpighian tubules suggest that the secretion of primary urine is by the process common to many insects - active K+ secretion with CI" as the accompanying anion (Phillips, 1982). The proximal Segment I functions in reabsorption as evidenced by the ability of this segment to reabsorb fluid and ouabain in vitro. The discovery in fJ. fasciatus of a Malpighian tubule segment that reabsorbs ouabain suggests that excretory loss of cardenolide from the insect could be minimized, possibly aiding in the insect's sequestration of large quantities of cardenolides. Secretion of ouabain by Segment II against a concentration gradient was never observed. Under all conditions tested, external ouabain concentrations of .03 - 5 mM and fluid secretion rates of .043 - .812 jjl/h, U/P ratios for ouabain were less than one. The secretion rate of ouabain increased with increasing fluid secretion. The permeability coefficient remained relatively constant throughout these experimental manipulations 80 (0.25 - 2.7 x 10"6 cm-sec-1) suggesting passive ouabain transport into the tubule lumen. Malpighian tubule segment II appears far more permeable than the midgut [Pouabain = 8,5 x 10_11cm'sec"l, (Scudder and Meredith, 1982b)] but caution must be used in such comparisons since gross measurement of tissue surface areas are used to estimate permeability coefficients and do not take into account true membrane area. Passive excretion of ouabain has been reported in Locusta migratoria and Zonocerus variegata fed on a cardenolide free diet (Rafaeli-Bernstein and Mordue, 1978). Indeed, the Malpighian tubules of a number of other insects are permeable to organic compounds, a process Maddrell and Gardiner (1974) suggest is an automatic way for insects to clear the hemolymph of toxins. Ouabain is a well known inhibitor of (Na+ + K+)-dependent ATPases (Stekhoven and Bonting, 1981). High external concentrations of ouabain (up to 5 mM) do not appear to inhibit fluid secretion rates in 0. fasciatus Malpighian tubules as one would expect if (Na+ + K+)-dependent ATPases are essential in the formation of primary urine in insects [reviewed by Phillips (1981)]. This may reflect a) suboptimal temperature or ion concentrations for ouabain inhibition of (Na+ + K+)-dependent ATPases [for discussion concerning conflicting reports of insect tissue sensitivity to cardenolides and experimental methodology see Anstee and Bowler (1979)]. b) the absence of (Na+ + K+)-dependent ATPases or the presence of (Na+ +K+)-dependent ATPases insensitive to ouabain (Jungreis and Vaughan, 1977; Vaughan and Jungreis, 1977; but see Anstee and Bowler, 1979). c) secretion of primary urine by ion pumps other than Na-K exchange [reviewed by Phillips (Phillips, 1981)]. 81 One might expect to find ouabain insensitive Malpighian tubules in CJ. fasciatus, however, since the insect ingests and assimilates cardenolides from its natural food sources (Duffey and Scudder, 1972). The ability of both Malpighian tubule segments to metabolize ouabain is interesting since such metabolism was not found in mammals (Dutta, et al, 1963; Kolenda, et &]_, 1971) nor in fJ. fasciatus gut, hemolymph or dorsolateral space fluid, although a nonpolar cardenolide, digitoxin, was metabolized (Scudder and Meredith, 1982b). Further study is necessary to elucidate the function of cardenolide metabolism in CJ. fasciatus. When Segment I was directly analyzed in slowly secreting tubules, it reabsorbed 52% of the ouabain excreted by Segment II. This reabsorption occurred against a strong concentration gradient (23 fold) and was specific to ouabain, i.e. no labelled metabolites, although present in the lumen, were reabsorbed. Reabsorption of ouabain was independent of fluid reabsorption^but increased with increasing fluid secretion by Segment II. In rapidly secreting whole Malpighian tubules, the presence of Segment I reduced the excretion of ouabain by 84 - 93%, mainly by reducing ouabain concentration. These results suggest that ouabain reabsorption is an active process and remains unsaturated under the limited conditions used in this study. Moreover, the results suggest that in in vivo situations requiring fast excretion rates (a situation of high cardenolide secretion by Segment II) CJ. fasciatus is able to recover by reabsorption in Segment I most of the cardenolide that otherwise would be lost. The high concentration of Na+ in the reabsorption fluid indicates that ouabain may be linked to Na+ transport as seen in the reabsorption of glucose and trehalose in the Malpighian tubules of L_. migratoria (Ramsay, 1958). Rafaeli-Bernstein and Mordue (1978) report that excretion of ouabain 82 (i.e., radioactive label) by Malpighian tubules of 2. variegata is increased, with U/P ratios about 3, when the insect's diet contains cardenolides. The authors suggest they have induced an active ouabain pump. No evidence was found in the present study for active excretion in (). fasciatus Malpighian tubules. The active pump we have described in Segment I differs from that proposed for 2. variegatus in both its orientation, resulting in active reabsorption, and also in the higher concentration gradient achieved. It would be interesting to determine whether this pump is inducible as in 1. variegatus. Reabsorption of an organic molecule such as ouabain appears to be rare in Malpighian tubules of insects studied so far. Only a few other organics, glucose and trehalose have been shown to be reabsorbed by Malpighian tubules (Knowles, 1975; Rafaeli-Bernstein and Mordue, 1979); Maddrell and Phillips (1976) state that amino acids are actively reabsorbed in the lower tubule of Rhodnius. Other than these examples, reabsorption by Malpighian tubules appears to be limited to ion recycling [K+ (Irvinev, 1969; Maddrell and Phillips, 1976) or Na+ in the case of some blood feeders (Gee, 1976)] fluid reabsorption (Irvine, 1969) and possibly bicarbonate reabsorption (Wigglesworth, 1931). Reabsorption by Segment I may aid in fluid reabsorption in 0. fasciatus. Although under conditions of direct analysis fluid reabsorption in 0. fasciatus is only 3% of excretion, it may be significantly higher in vivo. Sequential analysis suggests the presence of Segment I reduces secretion rates 1.1 - 5.1 times, although this could be partly a result of restricted fluid flow. Segment I could also play a role in ionic regulation since the Na+/K+ ratio in the reabsorbate (0.61) more closely resembles that in the hemolymph (0.96) than that in whole tubule secretion (0.05). 83 This study demonstrates that in in vitro preparations the distal segment of the Malpighian tubules of fJ. fasciatus secretes the polar cardenolide ouabain whereas the proximal Segment I reabsorbs ouabain. Both segments were found to metabolize this cardenolide. Each of these processes occurs simultaneously and provides potential points of control for cardenolide accumulation in vivo. fJ. fasciatus may be forced to lose cardenolides by excretion since Malpighian tubules are necessarily permeable structures. The discovery of a specialized Malpighian tubule segment that can actively reabsorb a cardenolide, however, suggests this excretory loss could be minimized. Further investigation is necessary to establish the role in vivo of the Malpighian tubules in the overall process of sequestration of cardenolides in fJ. fasciatus. (Material in this chapter was done in collaboration with J. Meredith and is in Meredith, J., Moore, L. and G.G.E. Scudder. 1984. Excretion of ouabain by Malpighian tubules of Oncopeltus fasciatus. Am. J. Physiol. 246 (Regulatory Integrative Comp. Physiol. 15): R705-715). 84 CHAPTER 4: OUABAIN RESISTANT NA.K-ATPASES AND CARDENOLIDE TOLERANCE IN THE LARGE MILKWEED BUG, ONCOPELTUS FASCIATUS Summary Oncopeltus fasciatus tolerated 1954x and 7288x, respectively, the LD50 ouabain dose of Schistocerca gregaria and Periplaneta americana when ouabain was injected into the hemocoel of these insects. The maximum ouabain dose that could be injected into Ch fasciatus (200 nmoles) resulted in no mortality; this dose is higher than the lethal ouabain doses recorded for vertebrates and invertebrates. The ouabain concentration resulting in 50% inhibition (I50) of Na,K-ATPase activity was determined in lyophilates of nervous tissue of fJ. fasciatus and brain and recta of ^. gregaria and were 2.0 x 10~4, 2.0 x 10~6, and 1.0 x 10"6 M, respectively. The I5Q value for ouabain inhibition of Na,K-ATPase activity in the nervous tissue of CJ. fasciatus is higher than the I50 values for nervous tissue in most other insects as well as many other invertebrate and vertebrate tissues. Thus, the presence of ouabain resistant Na,K-ATPases appears to be a factor in the tolerance and sequestration of plant cardenolides in 0. fasciatus. 85 INTRODUCTION Cardenolides are toxic to both vertebrates and invertebrates (Hoch, 1961; Treherne, 1966; Detweiler, 1967; Anstee and Bell, 1975; Biekirch, 1977; Rafaeli-Bernstein and Mordue, 1978; Benson et aK, 1979; Stekhoven and Bonting, 1981; Anstee and Bowler, 1984), at concentrations as low as 10-8 M (Glynn, 1964). Cardenolides exert their toxic effects by specifically inhibiting the enzyme Na++K+-ATPase (Na,K-ATPase; E.C. 3.6.1.3; Akera, 1977; Bodeman, 1981), which is essential in maintaining the electrochemical gradients for Na+ across the cell (Jtfrgensen, 1980; Stekhoven and Bonting, 1981). In invertebrates, the cardenolide, ouabain, has been demonstrated to interfere with Malpighian tubule secretion (Pilcher, 1970; Atzbacher et aK, 1974; Anstee and Bell, 1975), proper functioning of nervous tissue (Treherne, 1966; Farquharson, 1974) midgut ion fluxes and transepithelial potentials (O'Riordan, 1969; Prusch, 1978) and labial gland secretion (Kafatos, 1968). However, despite the potential toxicity of cardenolides, large amounts of these compounds are ingested by, and sequestered in, (). fasciatus from its host plants with no apparent ill effects (Isman, 1977; Chaplin and Chaplin, 1981; Jones et_aK, 1983). Among vertebrates, species differences in cardenolide sensitivity are known to parallel differences in the cardenolide sensitivity of their Na,K-ATPases (Akera, 1977; Schwalb et al_., 1982). However, neither the relative sensitivity of 0. fasciatus to the toxic effects of cardenolides or the possibility that the presence of cardenolide resistant Na,K-ATPases in 0_. fasciatus is a factor in the apparent insensitivity of this insect to cardenolide intoxication has been determined. Therefore, in this chapter, the in vivo resistance of 0. fasciatus to ouabain is compared to that of two insects, namely Schistocerca gregaria and Periplaneta americana, which 86 do not normally encounter cardenolides in their diet. In addition, the ouabain sensitivity of Na,K-ATPases is determined in tissue lyophilates of CJ. fasciatus and j>. gregaria. The results of this chapter provide evidence that CJ. fasciatus is more resistant to cardenolides than S. gregaria and IP. americana, as well as many other invertebrates and vertebrates. Furthermore, the relative insensitivity of Na,K-ATPases in the nervous tissue of CJ. fasciatus indicates that the presence of cardenolide resistant Na,K-ATPases may be another factor in the ability of this insect to ingest and sequester cardenolides. MATERIALS AND METHODS 1) Insects Adult Oncopeltus fasciatus (Dallas) were obtained from laboratory colonies maintained at 26°C with 65 % relative humidity (RH) and a 16 h light: 8 h dark (L:D) light regime. Insects were reared on one of two diets: 1) milkweed seeds (Asclepias speciosa Torr. collected Sept. 1982 in Penticton, B.C.), which contain cardenolides and 2) sunflower seeds (Heliothus annuus L.), which are cardenolide-free. Sunflower-reared bugs were obtained from a colony raised on sunflower seeds for two years; no cardenolides were detected in these insects (Moore and Scudder, 1985). Schistocerca gregaria L. adults were obtained from a laboratory colony reared on a cardenolide-free diet (bran, milk powder, alfalfa and lettuce) and were maintained at 29°C, 55 % RH, 12:12 L:D. Periplaneta americana L. adults were supplied by Carolina Biological Supply Company (Burlington, N. C.) and were maintained at room temperature on a cardenolide-free diet 87 (potato, lettuce, bran, milk powder, alfalfa and rat chow). 2) Ouabain Injections Female 0_. fasciatus tolerated larger injection volumes than males, therefore adult females of"all three insect species, (). fasciatus and P_. americana of random ages, and j>. gregaria 16-18 days after ecdysis, were used. Ouabain solutions (ouabain octahydrate, Sigma Chem. Co.) in 0.9% NaCl were injected into the hemocoel by fine glass pipettes (Fisher alkali-free coagulation capillary tubes). 0_. fasciatus were injected between abdominal sterna 5 and 6, ^. gregaria between abdominal sterna 1 and 2, and £. americana between abdominal terga 9 and 10. These injection sites minimized backwelling of hemolymph and injected solutions, and hemorrhaging in P_. americana. Oneyul volumes were injected into J>. gregaria and P_. americana; one to five jul volumes were injected into (). fasciatus. To facilitate injections into 0. fasciatus, hemolymph volume was reduced by maintaining insects with food, but no water, for 3 h for injections of 1-2 ^ul and 13-20 h for injections of 3-5 jul. Insects were immobilized by cooling at -20°C (5, 8, and 9 min for 0. fasciatus, P_. americana and Si. gregaria, respectively), injected, and then placed in cages under 100 watt bulbs, with food and water. Recovery time was defined as walking to the top of the cage. After recovery, insects were returned to their usual laboratory maintenance regime (stated above) and survival was monitored for 8 days; only those insects that were upright and moved when disturbed were considered survivors. Control insects were injected with 1-5 ^1 volumes of saline (0.9% NaCl) and treated in the same manner as ouabain-injected insects. 88 3) Na+ + K+-dependent ATPase Activity and Ouabain Sensitivity Tissue lyophilates of adult male CJ. fasciatus (random ages) and S_. gregaria (1-4 weeks after ecdysis) reared on cardenolide-free diets were prepared according to the methods of Peacock (1979, 1981a, 1981b) and Tolman and Steele (1976); all solutions and procedures were at 4°C. Tissues were dissected and rinsed in dH20 (gut tissues were slit longitudinally and contents removed prior to rinsing), pooled by tissue type in 0.5 ml dH20 until all tissues were dissected, homogenized with a teflon pestle for 3 min at 3000 rpm, frozen immediately in a MeOH/dry ice bath, lyophilized overnight and used immediately or stored at -80°C for a maximum of 18 days. Lyophilates were reconstituted by homogenization with a teflon pestle at 0°C for 1 min at 1500 rpm in buffer (30 mM imidazole, 250 mM mannitol, 5 mM EDTA, pH 7.2; modified from Anstee and Bowler 1984). Incubation and assay conditions for Na,K-ATPase activity were modified from the methods of Anstee and Bell (1975), Jtfrgensen (1974), and Peacock (1979). Five reaction media were used: 1) 4 mM MgCl2, 100 mM NaCl, 20 mM KC1 2) 4 mM MgCl2 3) 4 mM MgCl2, 100 mM NaCl 4) 4 mM MgCl2, 20 mM KCL 5) 4 mM MgCl2, 100 mM NaCl, 20 mM KC1 with varying concentrations of 8 ouabain (10~ to 1.2 x 10-3 M)# All media were buffered with 50 mM imidazole, pH 7.4, and contained 3 mM ATP (Tris salt,' Sigma Chem. Co.) and 0.1 mM EDTA. Anion concentration was kept constant by the addition of choline CI" (Hanrahan and Phillips, 1983). Total ATPase activity was determined in medium 1 and ouabain 89 inhibition of Na,K-ATPase activity was determined in medium 5. Four different determinations were averaged to estimate Mg2+-ATPase activity (media 2-5; Anstee and Bowler, 1984); the common estimate of Mg2+-ATPase activity (medium 5 with 10"3 M ouabain, Anstee and Bowler, 1984) was only used with tissues of gregaria since the presence of ouabain-insensitive Na,K-ATPases was suspected in 0. fasciatus. Na,K-ATPase activity was determined by total ATPase activity (medium 1) minus Mg2+-ATPase activity (average activity in media 2-4 or 2-5, Anstee and Bowler, 1984). Reaction media (1 ml) were preincubated at reaction temperature (30°C) for 10 min in 1.5 ml Eppendorf tubes, the reaction started by the addition of 10-15 ^il of lyophilate suspension (20 to 75 ^ug protein) and terminated after 20 min by the addition of 0.1 ml of 50% trichloroacetic acid. Under these assay conditions, the reaction was linear with respect to both time and enzyme concentration. Tubes were placed on ice, centrifuged at 4°C at 12,000 g for 5 min to remove any precipitated protein and kept on ice until use. Enzyme activity was measured by a sensitive microdetermination of inorganic phosphate (Pi) release (lower limit 0.15 nmoles Pi; Chen et al., 1956), using a Lambda 3 UV/VIS spectrophotometer (Perkin-Elmer). Contaminating Pi present in the lyophilate suspension or released by ATP degradation was determined by two control tubes treated in the same manner as experimental tubes: 1) reaction medium 1 with lyophilate suspension added, in the absence of ATP, and 2) reaction medium 1 in the absence of lyophilate suspension. Pi released in experimental tubes was corrected by subtracting the Pi present in both control tubes. Protein was determined by a modified Lowry assay (Schacterle and Pollack, 1973) with bovine plasma 90 gamma globulin standard (BioRad). RESULTS 1) Sensitivity of Insects to Injections of Ouabain into the Hemocoel 0. fasciatus (N=16) suffered no mortality 48 h after injection with the maximum dose of ouabain (200 nmoles, Table 4.1): 200 nmoles was the maximum dose of ouabain that could be injected into the hemocoel of 0. fasciatus owing to the limited solubility of ouabain in the saline carrier (1 gm/75 ml H2O) and to the limited volume that could be injected into this small insect (5 jul = 45% of the total hemolymph volume). Previous exposure to cardenolides or the presence of large amounts of sequestered plant cardenolides (150 jxq per insect; Moore and Scudder, 1985) did not affect the tolerance of £. fasciatus to ouabain since all insects injected with 200 nmoles of ouabain survived (48 h) whether they were reared on a diet containing cardenolides or were obtained from a colony that had not been exposed to cardenolides for 2 years (Table 4.1). Similarly, longer term survival (8 days) was unaffected by the presence or absence of cardenolides in the diet or by the large amounts of ouabain injected in this study. Although slight mortality at 8 days occurred in one group of insects injected with 200 nmoles of ouabain (11%, N=9), similar mortality was seen in a saline-injected control group of insects reared on the same food source (10%, N=10). The possibility of sublethal toxic effects of ouabain was investigated in 0. fasciatus by measuring the time taken for recovery from injections of increasing doses of ouabain. Although recovery times from injections were variable, no significant differences were found among recovery times from injections of 10 - 200 nmoles of ouabain, nor did recovery times of ouabain 91 Table 4.1. Survival of fJ. fasciatus injected with 10 - 200 nmoles ouabain.3 Ouabain Presence of Number of % Survival Injection cardenol ides insects (nmoles) in diet injected 48 h 8 days 10 4 100 100 20 4 100 100 40 5 100 100 200 9 100 89 200 7 100 100 Saline control'5 10 100 90 Saline controlb 8 100 100 a Adult females of random ages. b Maximum injection volume (5 jx\) 92 injected insects differ significantly from recovery times of control individuals reared on the same food source and injected with the same volume of saline (Student's t-test, p> 0.05; Table 2). In contrast to the insensitivity of 0. fasciatus to injections of as much as 200 nmoles of ouabain, mortality occurred in groups of S_. gregaria and P_. americana injected with as little as 3.0 and 0.5 nmoles of ouabain, respectively (Fig. 4.1). The ouabain dose causing 50% mortality (LDgg) in S_. gregaria and P_. americana was calculated by probit analysis (Fig. 4.1) and is reported in Table 4.3. 0. fasciatus is clearly insensitive to ouabain, and tolerated 33x and 222x the LD^Q ouabain dose of ^. gregaria and P_. americana (6.1 and 0.9 nmoles ouabain per insect, respectively). Correcting for weight and hemolymph volume differences between the three insect species, 0. fasciatus tolerated 1954x and 7288x the LD50 ouabain dose of J5. gregaria and P_. americana, respectively. 2) Activity and Ouabain Sensitivity of Na"1" + K+-dependent ATPases in Tissue Lyophilates of 0. fasciatus and S. gregaria Na,K-ATPase activity was determined in lyophilates of the following four tissues in £. fasciatus, the number in parentheses indicating the number of tissues pooled per lyophilate: 1) 1st ventriculus of the gut (30), 2) 2nd and 3rd ventriculi of the gut (30), 3) nervous tissue (head capsule contents plus prothoracic and central ganglion) (15), and 4) dorsal blood vessel (15). Na,K-ATPase activity was also determined in lyophilates of two tissues of the ouabain-sensitive j>. gregaria: 1) brain (4) and 2) rectum (2). 93 Table 4.2. Post-injection recovery times of fj. fasciatus. Ouabain Injection Presence of Number of Recovery Injection volume cardenolides insects time (nmoles) (jul) in diet injected X+SE (min) 10 1 - 4 1.5+0.29 20 1 - 4 1.7+0.86 40 2 5 5.4 + 2.96 200 5 - 9 3.2 + 1.33 200 5 +7 9.8+5.68 Saline injection 1-5 4.6 +_ 4.04 2 - 5 1.8+0.13 5 - 7 1.9+0.47 " 5 + 7 1.5 + 0.24 94 Figure 4.1. Mortality (24 h) of S_. gregaria and P_. americana injected with varying amounts of ouabain. Each data point represents the % mortality, expressed in probits, seen in a group of 5-6 insects. 00 / O A 9 000) i O P americana As gregaria 7 (977) O O A/ oi 5(50.0) H O o/o A A 3(2.3) H 10 4- -1D -05 6 ' 05 ' 10 logl0 nmoles ouabain injected 96 Table 4.3. Ouabain tolerance of fJ. fasciatus, j>. gregaria and P. americana3 Initial ouabain nmoles (^ig) nmoles (jig) ouabain concentration in ouabain injected/mg wet hemolymph after Insect injected/insect weight insect injection (M)b 0. fasciatus 200.0 (145.7) 4.3 (3.2) 12.5 x 10"3 100% survival S. gregaria 6.1 (4.4) 2.2 x IO-3 (1.6 x 10'' 3) 3.0 x IO"5 LD50 P. americana 0.9 (0.6) 5.9 x IO-4 (4.3 x 10"• 4) 5.4 x IO-6 LD50 3 Adult females - CJ. fasciatus and P_. americana random ages, S_. gregaria 16-18 days after ecdysis. Survival after 24 h (48 h survival is the same). b Calculated using the following hemolymph volume estimates: fj. fasciatus and !>. gregaria 11 and 200 jjl, respectively (Meredith, unpublished results), P_. americana 165 jul (Guthrie and Tindall, 1968), and assuming complete mixing of injected ouabain with hemolymph. 97 The results are shown in Table 4.4. No ATPase activity of any kind was detected in a lyophilate of 15 dorsal blood vessels (data not shown) and Na,K-ATPase activity in the 1st ventriculus and 2nd plus 3rd ventriculi lyophilates was low, only 19.3 and 15.8 nm Pi/mg protein/min, respectively. Much higher Na,K-ATPase activity was detected in lyophilates of the nervous tissue of CJ. fasciatus (148.6 nm Pi/mg protein/min) and the brain and rectum of ^. gregaria (258.5 and 225.2 nm Pi/mg protein/min, respectively). Preincubation with the solubilizing agent sodium deoxycholate (NaDOC) or NaDOC (0.1%) and 2mM sodium iodide (Nal), increases Na,K-ATPase activity in many microsomal Na,K-ATPase preparations (Nakao et al_., 1965; Peacock et al., 1972; J0rgensen, 1974; Anstee and Bowler, 1984). Therefore, preincubation with these solubilizing agents was used in an attempt to increase Na,K-ATPase activity in the two gut lyophilates of CJ. fasciatus. Gut lyophilates were preincubated in reconstitution buffer (see Material and Methods) containing various concentrations of NaDOC or NaDOC (0.1%) plus 2mM Nal for 30 min at 4°C. Preincubation with NaDOC (0.1%) plus 2 mM Nal did not effect Na,K- or Mg2+-ATPase activity in either gut lyophilate of fJ. fasciatus (N=l, data not shown). Similarly, no effect was seen in the Na,K- and Mg2+-ATPase activity of 1st ventriculus lyophilates preincubated with concentrations of NaDOC ranging from .04 to .35% (5 concentrations, N=l-2, data not shown) or in a 2nd plus 3rd ventriculi lyophilate preincubated with concentrations of NaDOC ranging from .18 to .52% (4 concentrations, N=l, data not shown). To investigate the possible role of ouabain resistant Na,K-ATPases in the ouabain resistance of 0. fasciatus, ouabain inhibition curves Table 4.4. Total, rig* and Na,K-ATPase activity of tissue lyophilates of 0. fasciatus and ^. gregaria. X ATPase activity (nmPj/mg protein/minute) + S.E. (N» number of lyophilates) 0. fasciatus 0. fasciatus 0. fasciatus j>. gregaria S. gregaria 1st ventrlculus 2nd A 3rd ventrlculus nervous tissue brain rectum Ionic Media3 (N • 3) (M - 3) (N - 6) (N • 5) (N - 5) 1. Total (Mg.Na.K) 111.2 + 5.1 83.7 + 3.6 203.0 + 18.4 362.0 + 21.2 336.3 • 32.0 2. Mg 90.1 • 2.3 65.5 • 2.9 56.9 + 5.3 106.6 + 7.5 122.7 + 11.2 3. Mg.Na 95.6 • 7.0 72.5 + 8.1 53.4 + 6.4 109.8 • 8.7 115.7 • 8.5 4. Mg.K 89.8 • 3.1 65.6 + 2.3 52.8 + 2.4 95.2 • 7.3 99.2 • 6.9 5. Mg.Na,K plus 1.2 x 10"3 ouabain b b b 102.1 + 10.8 106.4 • 11.8 X Mg2+-ATPase activity0 91.9 + 2.2 67.9 + 2.5 54.4 • 2.6 103.4 + 3.8 111.0 + 4.6 Na.K-ATPase act1v1tyd 19.3 15.8 148.6 258.5 225.2 for details of Ionic composition see Materials and Methods Medium 5 was not used to determine Mg2+-ATPase activity 1n 0. fasciatus because the presence of ouabain Insensitive Na,K-ATPases was suspected. Average of the Mg2+-ATPase activity determined In media 2-4 for 0. fasciatus tissues and media 2-5 for S. gregaria tissues. Na.K-ATPase activity was determined by total ATPase activity (medium 1) minus x Mg2+-ATPase activity. In all cases, total ATPase activity was significantly greater than Mg2+-ATPase activity (p < 0.05, Student's t-test). 99 (10-8 to 1.2 x 10-3 M) of Na,K-ATPase activity were determined for the following lyophilates: 1) nervous tissue of rj. fasciatus, 2) brain of S_. gregaria, and 3) rectum of j>. gregaria (Fig. 4.2). As a result of their low Na,K-ATPase activity, ouabain inhibition curves could not be constructed for dorsal blood vessel or gut lyophilates of 0. fasciatus. PI50 and I50 values (the -login molar concentration and the molar concentration, respectively, of ouabain that inhibits 50% of the Na,K-ATPase activity) were calculated by linear regression of ouabain concentration vs. % inhibition of Na,K-ATPase activity expressed as probits. The results are reported in Table 4.5. The ouabain concentration resulting in 50% inhibition of the Na,K-ATPase activity in. nervous tissue lyophilates of 0. fasciatus is 2.0 x 10-4 M. This concentration is lOOx greater than the ouabain concentration resulting in 50% inhibition of the Na,K-ATPase activity of lyophilates of _S. gregaria brain (2.0 x 10-6 M), and 200x greater than the ouabain concentration resulting in 50% inhibition of Na,K-ATPase activity in S. gregaria rectum (1.0 x 10-6 M). DISCUSSION It is clear that £. fasciatus is very resistant to the toxic effects of ouabain. High concentrations of ouabain (_> 10-3 M) do not affect survival or Malpighian tubule fluid secretion in this insect (this chapter, Chapter 3), whereas lower ouabain concentrations (10~4 to 10-6 M) effect various tissue processes and survival in many other insects and a millipede (Table 4.6). Tissue processes in a number of insect species have been reported to be unaffected by high concentrations 100 Figure 4.2. Effect of varying ouabain concentrations on Na,K-ATPase activity in lyophilate preparations of fj. fasciatus nervous tissue and _S. gregaria brain and rectum (means +_ S.E. where larger than symbol; n = number of lyophilates). 101 XX) O. fasciatus nervous tissue piso"37 n«6 >» < ~'°8io ouabain concentration CM) 102 Table 4.5. Inhibition of Na,K-ATPase activity by ouabain in nervous tissue of 0. fasciatus and brain and rectum of S. gregaria. a b Species Tissue pi™ Icn (M) CL fasciatus nervous tissue 3.7 2.0 x 10" _S. gregaria brain 5.7 2.0 x 10" S. gregaria rectum 6.0 1.0 x 10" 9 P^O = (_^°9l0 °^ tne m°lar concentration, resulting in 50% inhibition. b I50 = the molar concentration resulting in 50% inhibition. 103 Table 4.6. Effects of ouabain on tissue processes and survival 1n Insect and millipede species. Ouabain concentration Species Tissue process (M); effect Reference Carausius morosus Malpighian tubule 2.5 x IO*4; inhibition Pilcher fluid secretion (1970) Na efflux from IO-5; Inhibition Treherne nerve cord (1966) Glomeris marginata Na efflux from 5 x IO"6 - IO-3; Farquharson nerve cord inhibition (1974) Drosophila hydei Malpighian tubule 1 ^jl 3 x IO-4 Atzbacher e1 dye excretion Injection; Inhibition al_. (1974) Locusta migratoria Malpighian tubule IO"6 - IO'3; Anstee and fluid secretion Inhibition Bell (1975) Peripi aneta Midgut ion fluxes IO"5 - 10"3; reduces O'Riordan americana and transepi- potential; IO-4 (1969) thelial potential reduces 1on fluxes 5.4 x IO-6 this study Na efflux from IO"5; Inhibition Treherne nerve cord (1966) -5 Schistocerca LD50 3.0 x IO this study gregaria Sarcophaga bullata Midgut transepi- IO-5; potential Prusch thelial potential to zero (1978) Antherea pernyi Labial gland 5 x IO"4; 50% Kafatos secretion inhibition (1968) 104 of ouabain (10~3 M), however, these results are controversial (for review, see Anstee and Bowler, 1979). 0. fasciatus is also insensitive to much larger ouabain doses than those resulting in death in at least nine vertebrate species, including rats which are considered resistant to cardenolide intoxication (Table 4.7; Stekhoven and Bonting, 1981). Although lethal injection doses of ouabain in insects and vertebrates may not be directly comparable owing to the differences between open and closed circulatory systems, the LD50 doses for j>. gregaria and £. americana are within the range of lethal doses recorded for vertebrates, indicating that lethal doses in these taxa may be similar. If this is true, the very large difference between the ouabain sensitivity of 0. fasciatus and that recorded for J>. gregaria, _P. americana, and the vertebrates species listed in Table 4.7 strongly suggests that 0. fasciatus is more resistant to ouabain than most vertebrates, as well as many invertebrates. Among vertebrates, species differences in cardenolide sensitivity are known to parallel differences in the cardenolide sensitivity of their Na,K-ATPases (Akera, 1977; Schwalb et al_., 1982). The PI50 value (3.7) for ouabain inhibition of Na.K-ATPase activity in 0. fasciatus nervous tissue is much lower than the PI50 values for nervous tissue in most other insects (PI50 5-6; Anstee and Bowler, 1984) as well as many other invertebrate and vertebrate tissues (pi50 5 - 8, this study; Anstee and Bowler, 1984; Phil ippot jit , 1972; Bonting, 1966; Bakkeeren and Bonting, 1968; Ridderstrap and Bonting, 1969; Pfeiler and Kirschner, 1972; Schwalb^t^K, 1982). Thus, as in many vertebrates, the cardenolide resistance seen in /J. fasciatus may be related to Na,K-ATPase cardenolide resistance. Ouabain resistant Na,K-ATPases have been isolated from the nervous 105 Table 4-7. Toxicity values for ouabain In invertebrates and vertebrates. Type of Animal Injection Dose (mg/kg)a References 0. fasciatus hemocoel 100% survival 3200.0 this study H S. gregaria hemocoel LD 50 1.63 M hemocoel LD L0 0.80 M P. americana hemocoel LD50 0.43 •t LDL0 0.24 Cat Intravenous MLD 0.11-0.15 Duffey (1977) Dog Intravenous LDL0 0.054 Lewis « Tatken (1979) H Dog subcutaneous LDL0 150. H Rat Intravenous LD50 14. tl Rat intraperitoneal LD50 47. H Monkey Intravenous LDL0 0.106 M Rabbit subcutaneous LDL0 0.120 Rabbit intravenous LD 0.1-0.2 Hoch (1961) Guinea pig Intravenous LDL0 0.288 Lewis & Tatken (1979) M Pigeon Intravenous LDL0 0.285 H Mouse intraperitoneal LD50 12. Frog subcutaneous LD50 0.417 Hoch (1961) A LDJQ: dose resulting in 50% mortality. LDJ_Q: lowest dose given over any period of time in one or more portions known to cause death. MLD: mean lethal dose. LD: lethal dose. 106 tissue of one other insect, the monarch butterfly (PI50 <3; Vaughan and Jungreis, 1977) and in a mutant cell line of the mosquito Aedes al bop ictus (Mento et al_., 1979). It is interesting that the monarch specializes on food sources containing cardenolides, and like fJ. fasciatus, sequesters large amounts of these compounds (Brower and Glazier, 1975). In vertebrates, ouabain resistant Na,K-ATPases have been isolated from rats and ouabain resistant mutant cell lines of several other species (Bakkeren and Bonting, 1968; Schwalb et al_., 1982; Robbins and Baker, 1977; Baker et al., 1974; Landolph et al_., 1980; Mankovitz et al_., 1974). The ouabain resistance of these vertebrate Na,K-ATPases and the mutant mosquito cell line appear to be the result of structural changes in the enzymes resulting in unstable enzyme-ouabain complexes, a decrease in the number of ouabain binding sites on ouabain-resistant cells, and/or a changed response to bound ouabain. It is unlikely that the low Na,K-ATPase activity of fJ. fasciatus midgut tissue is the result of the predominance of K+-ATPases in the functioning of this tissue as has been suggested for the midgut epithelial tissues of several lepidopteran species in which Na,K-ATPase activity could not be demonstrated (Jungreis and Vaughan, 1977; Harvey ^t^., 1983). Harvey jst al_. (1983) suggest that K+-ATPases function in ion transport in the midgut of these insects since the gut lumen and hemolymph have high K+/Na+ ratios and, in one species, midgut Mg2+-ATPase activity was stimulated by K+. However, in fJ. fasciatus the Na/K ratio in the hemolymph is >_ one (Mullen, 1957; Staddon and Everton, 1980; Meredith, unpublished results), and midgut Mg2+-ATPase activity is not stimulated by K+ (this study). It is possible that in lyophilate preparations of CJ. fasciatus midgut tissue Na,K-ATPase activity is latent; 107 indeed, for some insect gut tissues an elaborate purifying procedure, requiring larger tissue volumes than available for the present study, is necessary to obtain preparations with high Na.K-ATPase activity (Anstee, pers. comm.). In summary, this chapter provides evidence that 0. fasciatus is resistant to the toxic effects of cardenolides. This tolerance does not seem to depend upon previous exposure to cardenolides, nor does it appear to be affected by the usual presence of large amounts of sequestered plant cardenolides in the adult insect. The ouabain resistance of Na,K-ATPases in the nervous tissue of 0. fasciatus indicates that this tissue is well protected from cardenolide toxicity. If the ouabain resistance of nervous tissue Na,K-ATPases is indicative of ouabain resistance of Na,K-ATPases in the rest of the insect then other tissues in fJ. fasciatus will also be protected. Thus, the presence of Na,K-ATPases resistant to cardenolide intoxication appears to be one more factor in the ability of this insect to tolerate and sequester cardenolides from its food plants. 108 CHAPTER 5: GENERAL DISCUSSION The work described in the previous three chapters examined specific aspects of the sequestration and tolerance of cardenolides in £. fasciatus. Chapter 2, Selective sequestration of milkweed cardenolides in 0. fasciatus, is concerned with several aspects of the quantitative distribution of cardenolides in 0. fasciatus and is the first investigation of the selective sequestration of individual milkweed cardenolides in the insect. Chapter 3, Excretion of ouabain by the Malpighian tubules of 0. fasciatus, is the first investigation of the role of the excretory system in cardenolide sequestration in insects. In Chapter 4, Ouabain resistant Na,K-ATPases and cardenolide tolerance in (). fasciatus, the relative insensitivity of the insect and its Na,K-ATPases to cardenolides was established for the first time. My investigation of the selective sequestration of milkweed cardenolides in £. fasciatus (Chapter 2) confirms the basic pattern of quantitative distribution of cardenolides in _0. fasciatus determined by Duffey and Scudder (1974). Duffey and Scudder found that the majority of cardenolides sequestered in 0. fasciatus are concentrated in the dorsolateral space while low levels are maintained in the hemolymph. The demonstration in Chapter 2 of the same pattern of cardenolide distribution in £. fasciatus reared on a different species of milkweed than that used in the Duffey and Scudder study, suggests that the sequestration of cardenolides primarily in the dorsolateral space and the maintenance of low cardenolide levels in the hemolymph is characteristic of this insect. The work in Chapter 2 also indicates that minimal amounts of cardenolides are sequstered in the gut, wings and fat body of £. fasciatus. This not only 109 provides a more complete understanding of the quantitative distribution of cardenolides in (h fasciatus, it also provides evidence that the sequestration of cardenolides in £. fasciatus differs markedly from that of the two other insects whose cardenolide distribution is known (C_. inopinatus and D. plexippus; Blum, 1981, 1983). This lends further credence to Blum's hypothesis that each insect has a unique way of handling the toxins it encounters in its host plant (Blum, 1983). Brower et al. (1982) suggest that earlier arguments advocating the prominent role of polarity in the sequestration of cardenolides in 0. fasciatus are premature. My results in Chapter 2 support this suggestion. The cardenolide profiles of 0. fasciatus, its tissues and secretions clearly indicate that cardenolides of a wide polarity range are sequestered and concentrated in the insect. This provides further evidence that a number of physical-chemical characteristics of cardenolides must be involved in their sequestration in insects as suggested by Seiber et al. (1980), and is similar to the situation described in the monarch butterfly (Brower et al_., 1982). Brower et al_. (1982, 1984a, 1984b) have also suggested that it might be possible to identify the species of milkweed an insect has been reared on by the insect's cardenolide profile, and give evidence for this in monarch butterflies feeding on three species of Asclepias. The similarity of the cardenolide profiles of 0. fasciatus reared on seeds of two species of Asclepias (Chapter 2), however, indicate that identification of an insect's food plant by the insect's cardenolide profile is either not possible with all species of Asclepias, or may not be possible with CL fasciatus. The presence of cardenolides in rj. fasciatus that were not detected in its food source, and the high concentration of intermediate and higher polarity cardenolides in the urine/feces of fj. fasciatus (Chapter 2) is the 110 first evidence suggesting metabolism and differential excretion of host plant cardenolides in fj. fasciatus. Metabolism of host plant cardenolides has been well documented in the monarch butterfly (Seiber et al_., 1980; Brower et al., 1982; Marty and Kreiger, 1984). Chapter 2 also provided the first evidence of the role of excretion in the sequestration of cardenolides in insects, and initiated the investigation of the excretion of ouabain by the Malpighian tubules of CJ. fasciatus in Chapter 3. In Chapter 3 the secretion, metabolism and reabsorption of ouabain was investigated under a variety of conditions to determine the possible role of the Malpighian tubules in the sequestration of cardenolides in CJ. fasciatus. The results suggest that secretion, metabolism and reabsorption of cardenolides by the Malpighian tubules are all potential points of control for cardenolide accumulation in CJ. fasciatus. The passive secretion and metabolism of ouabain by the Malpighian tubules of CJ. fasciatus indicates that the tubules are a potential site of cardenolide loss in the insect; £. fasciatus may be forced to lose cardenolides by Malpighian tubule secretion since tubule permeability to organic compounds appears to be an automatic way for insects to clear the hemolymph of toxins (Maddrell and Gardiner, 1974). However, the active reabsorption of ouabain in the proximal segment of the tubules suggests that this excretory loss could be minimized. Indeed, in rapidly secreting Malpighian tubules, a situation of high cardenolide secretion by the tubules, reabsorption recovered 84-93% of the ouabain that otherwise would be excreted. Reabsorption of an organic molecule such as ouabain appears to be rare in Malpighian tubules of insects studied so far; only two other organics, glucose and trehalose, have been shown to be reabsorbed by Malpighian tubules (Knowles, 1975; Rafaeli-Bernstein and Mordue, 1979). It may be Ill that the ability of the Malpighian tubules of 0. fasciatus to reabsorb cardenolides that would otherwise be excreted is a unique specialization, and of major importance in the sequestration of cardenolides in (). fasciatus. Chapter 4 documents two additional specializations in fJ. fasciatus that may function in the ability of this insect to tolerate and sequester cardenolides from its food plants; in vivo tolerance of the insect to large amounts of cardenolides injected into the hemolymph and the presence of cardenolide resistant Na,K-ATPases in the insect's nervous tissue. The results in Chapter 4 provide evidence that 0. fasciatus is resistant to doses of injected ouabain much larger than the lethal doses recorded for both vertebrates and invertebrates. Furthermore, insects reared on cardenolide-free diets as well as those reared on their normal cardenolide-rich diet were both extremely resistant to the toxic effects of injected ouabain. Thus, the in vivo cardenolide tolerance of 0. fasciatus does not seem to depend upon previous exposure to cardenolides, nor does it appear to be affected by the usual presence of large amounts of sequestered cardenolides in the insect. This apparent independence of the cardenolide tolerance of the insect from the cardenolide content of its food plants, as well as the cardenolides present in its tissues, may be a factor in the insect's ability to feed on food plants that vary greatly in their cardenolide content (Isman, 1977; Vaughan, 1979; Chaplin and Chaplin, 1981; Jones et al_., 1983). Among vertebrates, species differences in cardenolide sensitivity are known to parallel differences in the cardenolide sensitivity of their Na,K-ATPases (Akera, 1977; Schwal b et al_., 1982). The results in Chapter 4 suggest that similarly, the in vivo sensitivity of 5>. gregaria and 112 fJ. fasciatus to cardenolides also parallels the sensitivities of their Na,K-ATPases to cardenolide inhibition. This provides the first evidence that cardenolide sensitivity in invertebrates may be related to the cardenolide sensitivity of their Na.K-ATPases. Furthermore, the isolation of ouabain resistant Na,K-ATPases from the nervous tissue of 0. fasciatus indicates that ouabain resistant Na,K-ATPases are another factor in the tolerance and sequestration of plant cardenolides in this insect. Chapter 4 is the second documentation of ouabain resistant Na,K-ATPases in an insect that ingests and sequesters cardenolides from its host plant; ouabain resistant Na,K-ATPases have also been isolated from the nervous tissue of the monarch butterfly (Vaughan and Jungreis, 1977). It therefore seems possible that cardenolide resistant Na,K-ATPases may be characteristic of insects that ingest and sequester large amounts of cardenolides from their host plants. In summary, the work described in this thesis has clarified several aspects of the quantitative distribution of cardenolides in fj. fasciatus, as well as provided evidence that cardenolides are not sequestered in the insect simply on the basis of polarity. In addition, this work suggests that metabolism and differential excretion of host plant cardenolides may be involved in the selective sequestration of these compounds in fj. fasciatus. Finally, this thesis documents three specializations that may factor in the tolerance and sequestration of host plant cardenolides in 0. fasciatus - active reabsorption of cardenolides by the Malpighian tubules, high in vivo tolerance to cardenolides and the presence of cardenolide resistant Na,K-ATPases. 113 LITERATURE CITED Akera, T. 1977. Membrane adenosinetriphosphatase: a digitalis receptor? Science 198: 569-574. Anonymous. 1979. LS 8000 Series Liquid Scintillation System Manual, California: Beckman Instruments, Inc., 169pp. Anstee, J.H. and Bell, D.M. 1975. Relationship of Na+-K+-activated ATPase to fluid production by Malpighian tubules of Locusta migratoria. J. Insect Physiol. 21: 1779-1784. Anstee, J.H. and Bowler, K. 1979. Ouabain-sensitivity of insect epithelial tissues. Comp. Biochem. Physiol. A 62: 763-769. Anstee, J.H. and Bowler, K. 1984. Techniques for studying Na+,K+-ATPase, pp. 198-220, in T.J. Bradley and T.A. Miller (eds.). Measurement of Ion Transport and Metabolic Rate in Insects. Springer-Verlag, New York. Atzbacher, U., Hevert, F., Weber-Von Grotthuss, E. and Wessing, A. 1974. The influence of ouabain on the elimination of injected and orally applied dyes in Drosophila hydei. J. Insect Physiol. 20: 1989-1997. Baker, R.M., Brunette, D.M., Mankovitz, R., Thompson, L.H., Whitmore, G.F., Siminovitch, L. and Till, J.E. 1974. Ouabain-resistant mutants of mouse and hamster cells in culture. Cell 1: 9-21. Bakkeren, J.A.J.M. and Bonting, S.L. 1968. Studies on (Na++K+)- activated ATPase. XX. Properties of (Na++K+)-activated ATPase in rat liver. Biochim. Biophys. Acta 150: 460-466. Beikirch, H. 1977. Toxicity of ouabain on Drosophila melanogaster. Experentia 33: 494-495. Benson, J.M., Seiber, J.N., Bagley, C.V., Keller, R.F., Johnson, A.E., and Young, S. 1979. Effects on sheep of the milkweeds Asclepias eriocarpa and A. labriformis and of cardiac glycoside-containing derivative material. Toxicon 17: 155-165. Berridge, M.J. 1966. Metabolic pathways of isolated Malpighian tubules of the blowfly functioning in an artificial medium. J. Insect Physiol. 12: 1523-1538. Blakley, N. 1980. Divergence in seed resource use among Neotropical milkweed bugs (Oncopeltus). Oikos 35: 8-15. Blakley, N.R. and Dingle, H. 1978. Competition: butterflies eliminate milkweed bugs from a Caribbean island. Oecologia (Berl.) 37: 133-136. 114 Blum, M.S. 1981- Chemical Defenses of Arthropods. Academic Press, New York, 562 pp. Blum, M.S. 1983. Detoxication, deactivation, and utilization of plant compounds of insects, Chapter 15, pp. 265-275, in P.A. Hedin (ed.). Plant Resistance to Insects. American Chemical-S~oc. Symposium Series No. 208. Bodemann, H.H. 1981. The current concept for the cardiac glycoside receptor. Clin. Cardiol. 4: 223-228. Bonting, S.L. 1966. Studies on sodium-potassium-activated adenosinetriphosphatase. XV. The rectal gland of the elasmobranchs. Comp. Biochem. Physiol. 17: 953-966. Brooks, G.T. 1976. Penetration and distribution of insecticides, pp. 3-58, in C.F. Wilkinson (ed.). Insecticide Biochemistry and PhysioTogy. Plenum Press, New York. Brower, L.P., Edmunds, M. and Moffitt, CM. 1975. Cardenolide content and palatability of a population of Danaus chrysippus butterflies from West Africa. J. Entomol. Ser. A. Physiol. Behav. 49: 183-196. Brower, L.P. and Glazier, S.C. 1975. Localization of heart poisons in the Monarch butterfly. Science (Wash., D.C.) 188: 19-25. Brower, L.P. and Moffitt, CM. 1974. Palatability dynamics of cardenolides in the Monarch butterfly. Nature (Lond.) 249: 280-283. Brower, L.P., Seiber, J.N., Nelson, CJ., Lynch, S.P., Hoggard, M.P., and J.A. Cohen. 1984b. Plant-determined variation in cardenolide content and thin-layer chromatography profiles of monarch butterflies, Danaus plexippus reared on milkweed plants in California: 3. Asclepias californica. J. Chem. Ecol. 10: 1823-1857. Brower, L.P., Seiber, J.N., Nelson, C.J., Lynch, S.P. and Holland, M.M. 1984a. Plant-determined variation in the cardenolide content, thin-layer chromatography profiles, and emetic potency of monarch butterflies, Danaus plexippus L. reared on milkweed plants in California: 2. Asclepias speciosa. J. Chem. Ecol. 10: 601-639. Brower, L.P., Seiber, J.N., Nelson, C.J., Lynch, S.P. and Tuskes, P.M. 1982. Plant-determined variation in the cardenolide content, thin-layer chromatography profiles, and emetic potency of Monarch butterflies Danaus plexippus reared on the milkweed, Asclepias eriocarpa in California. J. Chem. Ecol. 8: 579-633. Chaplin, S.J. and Chaplin, S.B. 1981. Growth dynamics of a specialized milkweed seed feeder (Oncopeltus fasciatus) oh seeds of familiar and unfamiliar milkweeds (AsclepiasTspp.). ElTtomol. Exp. Appl. 29: 345-356. 115 Chen, P.S. Jr., Toribara, T.Y. and Warner, H. 1956. Microdetermination of phosphorus. Anal. Chem. 28: 1756-1758. Cohen, J.A. and Brower, L.P. 1983. Cardenolide sequestration by the dogbane tiger moth (Cycnia tenera; Arctiidae). J. Chem. Ecol. 9: 521-532. Detweiler, D.K. 1967. Comparative pharmacology of cardiac glycosides. Federation Proc. 26: 1119-1124. Dingle, H. 1978. Migration and diapause in tropical, temperate, and island milkweed bugs, pp. 254-276, in H. Dingle (ed.). Evolution of Insect Migration and Diapause. Springer, N. Y. Dingle, H., Blakley, N.R. and Miller, E.R. 1980. Variation in body size and flight performance in milkweed bugs (Oncopeltus). Evolution 34: 371-385. Dixon, C.A., Erickson, J.M., Kellett, D.N. and Rothschild, M. 1978. Some adaptations between Danaus plexippus and its food plant, with notes on Danaus chrysippus and Euploea core (Insecta: Lepidoptera). J. Zool., Lond. 185: 437-467. Dowd, P.F., Smith, CM. and Sparks, T.C. 1983. Detoxification of plant toxins by insects. Insect Biochem. 13: 453-468. Duffey, S.S. 1977. Arthropod allomones: chemical effronteries and antagonists. Proc. XV Int. Congr. Entomol., Wash., D.C. (1976), XV: 323-394. Duffey, S.S. 1980. Sequestration of plant natural products by insects. Ann. Rev. Entomol. 25: 447-477. Duffey, S.S., Blum, M.S., Isman, M.B. and Sucdder, G.G.E. 1978. Cardiac glycosides: a physical system for their sequestration by the milkweed bug. J. Insect Physiol. 24: 639-645. Duffey, S.S. and Scudder, G.G.E. 1972. Cardiac glycosides in North American Asclepiadaceae, a basis for unpalatability in brightly coloured Hemiptera and Coleoptera. J. Insect Phys. 18: 63-78. Duffey, S.S. and Scudder, G.G.E. 1974. Cardiac glycosides in Oncopeltus fasciatus (Dallas) (Hemiptera: Lygaeidae). I. The uptake and distribution of natural cardenolides in the body. Can. J. Zool. 52: 283-290. Dutta, S., Marks. B.H., and Smith, CR. 1963. Distribution and excretion of ouabain-3H and dihydro-ouabain-^H in rats and sheep. J. Pharmacol. Exp. Ther. 142: 223-230. Eastin, W.C Jr., Fleming, W.J. and Murray, H.C. 1982. Organophosphate 116 inhibition of avian salt gland Na,K-ATPase activity. Comp. Biochem. Physiol. 73C: 101-107. Erickson, J.M. 1973. The utilization of various Asclepias species by larvae of the monarch butterfly, Danaus plexippus. Psyche 80: 230-244. Farquharson, P.A. 1974. A study of Malpighian tubules of the pill millipede, Glomen's marginata (Villers)- II. The effect of variations in osmotic pressure and sodium and potassium concentrations on fluid production. J. Exp. Biol. 60: 29-39. Feir, D. and Suen, J. 1971. Cardenolides in the milkweed plant and feeding by the milkweed bug. Ann. Entomol. Soc. Am. 64: 1173-1174. Fink, L.S. and Brower, L.P. 1981. Birds can overcome the cardenol ide defence of monarch butterflies in Mexico. Nature 291: 67-70. Florkin, M. and Jeuniaux, C. 1974. Haemolymph: composition, pp. 255-301, vn M. Rockstein (ed.). The Physiology of Insecta. New York: Academic Press, vol. V, 2nd edition. Games, D.E. and Staddon, B.W. 1973. Composition of scents from the larva of the milkweed bug Oncopeltus fasciatus. J. Insect Physiol. 19: 1527-1532. Gee, J.L. 1976. Active transport of sodium by the Malpighian tubules of the tsetse fly Glossina moristans. J. Exp. Biol. 64: 357-368. Glynn, I.M. 1964. The action of cardiac glycosides on ion movements. Pharm. Rev. 16: 381-407. Graham, J.D.P. and Staddon, B.W. 1974. Pharmacological observations on body fluids from the milkweed bug, Oncopeltus fasciatus (Dallas) (Heteroptera: Lygaeidae). J. Ent. (A) 48: 177-183. Guthrie, D.M. and Tindall, A.R. 1968. The biology of the cockroach. Arnold, Ltd., London. Hanrahan, J. and Phillips, J.E. 1983. Mechanism and control of salt absorption in locust rectum. Am. J. Physiol. 244 (Regulatory Integrative Comp. Physiol. 13): R131-R142. Harvey, W.R., Cioffi, M. and Wolfersberger, M.G. 1983. Chemiosmotic potassium ion pump of insect epithelia. Am. J. Physiol. 244 (Regulatory Integrative Comp. Physiol. 13): R163-175. Helander, H.F. and Durbin, R.P. 1982. Localization of ouabain-binding sites in frog gastric mucosa. Am. J. Physiol. 243 (Gastrointest. Liver Physiol. 6): G297-G303. 117 Hoch, J.H. 1961. A Survey of the Cardiac Glycosides and Genins. Columbia: University of South Carolina Press, 1961, 93pp. Huheey, J.E. 1984. Warning coloration and mimicry, pp. 257-297, in W.J. Bell and R.T. Carde (eds.). Chemical Ecology of Insects. Chapman and Hall, Ltd. Irvine, H.B. 1969. Sodium and potassium secretion by isolated insect Malpighian tubules. Am. J. Physiol. 217: 1520-1527. Isman, M.B. 1977. Dietary influence of cardenolides on larval growth and development of the milkweed bug Oncopeltus fasciatus. J. Insect Physiol. 23: 1183-1187. Isman, M.B., Duffey, S.S. and Scudder, G.G.E. 1977. Variation in cardenolide content of the lygaeid bugs, Oncopeltus fasciatus and Lygaeus kalmii kalmii and of their milkweed hosts (Asclepias spp.) in central California. J. Chem. Ecol. 3: 613-624. Jones, C.G., Hoggard, M.P. and Blum, M.S. 1983. Is sequestration structure-specific in the milkweed bug, Oncopeltus fasciatus? Comp. Biochem. Physiol. C 76: 283-284. Jtfrgensen, P.L. 1974. Isolation of Na.K-ATPase. Methods Enzymol. 32B: 277-290. J0rgensen, P.L. 1980. Sodium and potassium ion pump in kidney tubules. Physiol. Rev. 60: 864-917. Jungreis, A.M. and Vaughan, G.L. 1977. Insensitivity of lepidopteran tissues to ouabain: absence of ouabain binding and Na+-K+ ATPases in larval and adult midgut. J. Insect Physiol. 23: 503-509. Kafatos, F.C. 1968. The labial gland: a salt secreting organ of saturniid moths. J. Exp. Biol. 48: 435-453. Kilby, B.A. 1963. The biochemistry of the insect fat body, pp. 111-174, in J.W.L. Beament, J.E. Treherne and V.B. Wigglesworth (eds.). AcTvances in Physiology. Vol. 7. Academic Press, London. Klausner, E., Miller, E.R., and Dingle, H. 1980. Nerium oleander as an alternative host plant for south Florida milkweed bugs, Oncopeltus fasciatus. Ecol. Entomol. 5: 137-142. Knowles, G. 1975. The reduced glucose permeability of the isolated Malpighian tubules of the blowfly Calliphora vomitoria. J. Exp. Biol. 62: 327-340. Kolenda, K-D., Lullmann, H. and Peters, T. 1971. Metabolism of cardiac glycosides studied in the isolated perfused guinea-pig liver. Br. J. Pharmacol. Chemother. 41: 661-673. 118 Landolph, J.R., Bhatt, R.S., Telfer, N. and Heidelberger, C. 1980. Comparison of adriamycin- and ouabain-induced cytotoxicity and inhibition of ^rubidium transport in wild-type and ouabain- resistant C3H/10T1/2 mouse fibroblasts. Cancer Res. 40: 4581-4588. Langer, G.A. 1981. Mechanism of action of the cardiac glycosides on the heart. Biochem. Pharmac. 30: 3261-3264. Lewis, R.J. Sr. and Tatken, R.L. (eds). 1979. Registry of Toxic Effects of Chemical Substances. U.S. Dept. of Health and Human Services (NIOSH) Publication No. 80-111. Levey, B. 1983. Plant allelochemicals and the evolution of host-plant relationships in the genus Spilostethus. Ph.D. thesis. University of Witwatersrand, Johannesburg, 205pp. Loffelhardt, W., Kopp, B. and Kubelka, W. 1979. Intracellular distribution of cardiac glycosides in leaves of Convallaria majalis. Phytochem. 18: 1289-1291. Lullmann, H. and Peters, T. 1973. Enteric absorption rate of different glycosides and the dependency upon intestinal blood flow, pp. 210-212, jijiO. Storstein (ed.). Digitalis. Gyldendal Norsk Forlag, Oslo. Maddrell, S.H.P. 1969. Secretion by the Malpighian tubules of Rhodnius. The movements of ions and water. J. Exp. Biol. 51: 71-97. Maddrell, S.H.P. and Gardiner, B.O.C. 1974. The passive permeability of insect Malpighian tubules to organic solutes. J. Exp. Biol. 60: 641-652. Maddrell, S.H.P. and Phillips, J.E. 1976. Regulation of absorption in insect excretory systems, pp. 179-185, vn P. Spencer (ed.). Perspectives in Experimental Biology. Davies, Oxford: Pergamon Press, Vol. I, Zoology. Malcolm, S. and Rothschild, M. 1983. A danaid mullerian mimic, Euploea core amymone (Cramer) lacking cardenolides in the pupal and adult stages. Biol. J. Linnean Soc. 19: 27-33. Mankovitz, R., Buchwald, M. and Baker, R.M. 1974. Isolation of ouabain-resistant human diploid fibroblasts. Cell 3: 221-226. Marty, M.A. and Kreiger, R.I. 1984. Metabolism of uscharidin, a milkweed cardenolide, by tissue homogenates of monarch butterfly larvae, Danaus plexippus L. J. Chem. Ecol. 10: 945-956. Mento, S.J., Malinoski, F. and Stollar, V. 1979. Phenotypic characterization of ouabain-resistant Aedes albopictus cells. J. Cell 119 Physiol. 101: 515-522. Meredith, J., Moore, L. and Scudder, G.G.E. 1984. The excretion of ouabain by the Malpighian tubules of 0. fasciatus. Am. J. Physiol. 246 (Regulatory Integrative Comp. Physiol. 15): R705-R715. Moore, L.V. and Scudder, G.G.E. 1985. Selective sequestration of milkweed (Asclepias sp.) cardenolides in Oncopeltus fasciatus (Dallas) (Hemiptera: Lygaeidae). J. Chem. Ecol. Tn press. Mullen, J.A. 1957. Sodium, potassium and calcium ions in the haemolymph of Oncopeltus fasciatus (Dallas). Nature 180: 813-814. Nakao, T., Tashima, Y., Nagano, K. and Nakao, M. 1965. Highly specific sodium-potassium-activated adenosine triphosphatase from various tissues of rabbit. Biochem. Biophys. Res. Commun. 19: 755-758. Neher, R. 1969. TLC of steroids and related compounds, pp. 311-362, vn E. Stahl (ed.). Thin-Layer Chromatography, A Laboratory Handbook, New York: Springer-Verlag. Nelson, C.J., Seiber, J.N. and Brower, L.P. 1981. Seasonal and intraplant variation of cardenolide content in the California milkweed, Asclepias eriocarpa, and implications for plant defense. J. Chem. Ecol.7: 981-1010. Nishio, S., Blum, M.S., and Takahashi, S. 1983. Intraplant distribution of cardenolides in Asclepias humistrata (Asclepiadaceae), with additional notes on their fates in Tetraopes melanurus (Coleoptera: Cerambycidae) and Rhyssomatus 1ineaticol1 is (Coleoptera: Curculionidae). Mem. Coll. Agric., Kyoto Univ. 122: 3-52. Nover, L. 1972. Cardiac glycosides and their genins, pp. 350-392, _in K. Macek (ed.). Pharmaceutical Applications of Thin-layer and Paper Chromatography. Elsevier Pub. Co., Amsterdam. Okita, G. T. 1967. Species difference in duration of action of cardiac glycosides. Federation Proc. 26: 1125-1130. O'Riordan, A.M. 1969. Electrolyte movement in the isolated midgut of the cockroach (Periplaneta americana). J. Exp. Biol. 51: 699-714. Paris, R. 1963. The distribution of plant glycosides, pp. 337-358, in T. Swain (ed.). Chemical Plant Toxonomy. Academic Press, London. Peacock, A.J. 1979. A comparison of two methods for the preparation of Mg2+-dependent, (Na+-K+) stimulated ATPase from the locust rectum. Insect Biochem. 9: 481-484. Peacock, A.J. 1981a. Further studies of the properties of locust rectal 120 Na++K+-ATPase, with particular reference to the ouabain sensitivity of the enzyme. Comp. Biochem. Physiol. C 68: 29-34. Peacock, A.J. 1981b. Distribution of (Na++K+)-ATPase activity in the mid- and hindguts of adult Glossina morsitans and Sarcophaga nodosa and the hindgut of Bombyx mori larvae. Comp. Biochem. Physiol. A 69: 133-136. Peacock, A.J., Bowler, K. and Anstee, J.H. 1972. Demonstration of a Na++K+-Mg2+-dependent ATPase in a preparation from hindgut and Malpighian tubules of two species of insect. Experentia 28: 901-902. Pfeiler, E. and Kirschner, L.B. 1972. Studies on gill ATPase of rainbow trout (Salmo gairdneri). Biochim. Biophys. Acta 282: 301-310. Philippot, J., Thuet, M. and Thuet, P. 1972. Properties of the (Na+-K+)-ATPase from pleopods of Sphaeroma serratum (Fabricius). Comp. Biochem. Physiol. B 41: 231-243. Phillips, J.E. 1981. Comparative physiology of insect renal function. Am. J. Physiol. 241: R241-R257. Phillips, J.E. 1982. Hormonal control of renal functions in insects. Federation Proc. 41: 2348-2354. Pilcher, D.E.M. 1970. The influence of diuretic hormone on the process of urine secretion by the Malpighian tubules of the stick insect Carausius morosus. J. Exp. Biol. 53: 465-484. Prusch, R.D. 1978. Active Na+ uptake in the isolated midgut of larval Sarcophaga bullata. J. Insect Physiol. 24: 81-85. Rafaeli-Bernstein, A. and Mordue, W. 1978. The transport of the cardiac glycoside ouabain by the Malpighian tubules of Zonocerus variegatus. Physiol. Entomol. 3: 59-63. Rafaeli-Bernstein, A. and Mordue, W. 1979. The effects of phlorizin, phloretin and ouabain on the reabsorption of glucose by the Malpighian tubules of Locusta migratoria migratorioides. J. Insect Physiol. 25: 241-247. Ralph, CP. 1976. Natural food requirements of the large milkweed bug, Oncopeltus fasciatus (Hemiptera: Lygaeidae), and their relation to greganousness and host plant morphology. Oecologia (Berl.) 26: 157-175. Ralph, CP. 1977. Effect of host plant density on populations of a specialized, seed-sucking bug, Oncopeltus fasciatus. Ecology 58: 799-809. 121 Ramsay, J.A. 1958. Excretion of the stick insect, Dixippus morosus (Orthoptera, Phasmidae): amino acids, sugars and urea. J. Exp. Biol. 35: 871-891. Ridderstap, A.S. and Bonting, S.L. 1969. Na-K-activated adenosine triphosphatase and pancreatic secretion in the dog. Am. J. Physiol. 216:.547-553. Robbins, A.R. and Baker, R.M. 1977. (Na.K)ATPase activity in membrane preparations of ouabain-resistant HeLa cells. Biochemistry 16: 5163-5168. Roeske, CM., Seiber, J.N., Brower, L.P. and Moffitt, CM. 1976. Milkweed cardenolides and their compartive processing by monarch butterflies (Danaus plexippis L.). Recent Adv. Phytochem. 10: 93-167. Rothschild, M. 1966. Experiments with captive predators and the poisonous grasshopper Poekilocerus bufonius (Klug). Proc. R. Entomol. Soc. Lond. (C) 31T3T: Rothschild, M. 1972. Secondary plant substances and warning colouration in insects, pp. 59-83, vn Insect/Plant Relationships. H. F. van Emden (ed.). Roy. Ent. Soc. Lond., Symposium 6. Rothschild, M. and Kellett, D.N. 1972. Reactions of various predators to insects storing heart poisons (cardiac glycosides) in their tissues. J. Entomol. 46: 103-110. Rothschild, M. and Reichstein, T. 1976. Some problems associated with the storage of cardiac glycosides by insects, pp. 507-550, _in Luckner, M. and Mothes, K. (eds.). Secondary Metabolism and Coevolution: Cellular, intercellular, and interorganismic aspects. Nova Acta Leopoldina Supplementum Numrner 7. Rothschild, M., Marsh, N. and Gardiner, B. 1978. Cardioactive substances in the monarch butterfly and Euploea core reared on leaf-free artificial diet. Nature 275: 649-650. Rothschild, M., von Euw, J. and Reichstein, T. 1970. Cardiac glycosides in the oleander aphid, Aphis nerii. J. Insect Physiol. 16: 1141-1145. Rothschild, M., von Euw, J. and Reichstein, T. 1973. Cardiac glycosides (heart poisons) in the polka-dot moth Syntomeida epilais Walk. (Ctenuchidae: Lep.) with some observations on the toxic qualities of Amata (=Syntomis) phegea (L.). Proc. R. Soc. Lond. B. 183: 227-247. Schacterle, G.R. and Pollack, R.L. 1973. A simplified method for the quantitative assay of protein biologic material. Anal. Biochem. 51: 654-655. 122 Schwalb, H., Dickstein, Y. and Heller, M. 1982. Interactions of cardiac glycosides with cardiac cells. III. Alterations in the sensitivity of (Na++K+)-ATPase to inhibition by ouabain in rat hearts. Biochim. Biophys. Acta 689: 241-248. Scudder, G.G.E. and Duffey, S.S. 1972. Cardiac glycosides in the Lygaeinae (Hemiptera: Lygaeidae). Can. J. Zool. 50: 35-42. Scudder, G.G.E. and Meredith, J. 1982a. Morphological basis of cardiac glycoside sequestration by Oncopeltus fasciatus (Dallas) (Hemiptera: Lygaeidae). Zoomorphology (Berl.) 99: 87-101. Scudder, G.G.E. and Meredith, J. 1982b. The permeability of the midgut of three insects to cardiac glycosides. J. Insect Physiol. 28: 689-694. Seiber, J.N., Tuskes, P.M., Brower, L.P., and Nelson, C.J. 1980. Pharmacodynamics of some individual milkweed cardenolides fed to larvae of the monarch butterfly (Danaus plexippus L.). J. Chem. Ecol. 6: 321-339. ~ Seiber, J.N., Lee, S.M. and Benson, J.M. 1983. Cardiac glycosides (cardenolides) in species of Asclepias (Asclepiadaceae), pp. 43-83, in R.F. Keeler and A.T. Tu (eds.TI Plant and Fungal Toxins. Handbook "bT Natural Toxins Vol. 1. Marcel Dekker, Inc., New York. Seiber, J.N., Roeske, C.N. and Benson, J.M. 1978. Three new cardenolides from the milkweeds Asclepias eriocarpa and A. labriformis. Phytochem. 17: 967-970. Singh, B. and Rastogi, R.P. 1970. Cardenolides - glycosides and genins. Phytochem. 9: 315-331. Slater, J.A. 1964. A Catalogue of the Lygaeidae of the World. Univ. of Connecticut, Storrs., 1668pp. Slater, J.A. and Baranowski, R.M. 1978. How to Know the True Bugs (Hemiptera-Heteroptera). Wm.C. Brown Co. Pub., Dubuque, Iowa, 256 pp. Smith, D.A.S. 1978. The effect of cardiac glycoside storage on growth rate and adult size in the butterfly Danaus chrysippus (L.). Specialia 15: 845-846. Smith, T.W. and Haber, E. 1974. Digitalis. Little, Brown and Co., Boston, 110 pp. Staddon, B.W. and Daroogheh, H. 1981. Allometry of Cs and CQ alk-2-enals and alka-2,4-dienals in the metathoracic scent gland of Oncopeltus fasciatus. Comp. Biochem. Physiol. 68B: 593-598. Staddon, B.W. and Everton, I.J. 1980. Haemolymph of the milkweed bug Oncopeltus fasciatus (Heteroptera: Lygaeidae). Inorganic constituents 123 and amino acids. Comp. Biochem. Physiol. 65A: 371-374. Stahl, E. and Kaltenbach, U. 1965. Sugars and derivatives, pp. 461-469, in E. Stahl (ed.). Thin-layer Chromatography: A Laboratory Handbook. TIew York: Springer-Verlag. Stekhoven, F.S. and Bonting, S.L. 1981. Transport adenosine triphosphatases: properties and functions. Physiol. Rev. 61: 1-76. Strange, K., Phillips, J.E. and Quamme, G.A. 1982. Active HC03-secretion in the rectal salt gland of a mosquito larva inhabiting NaHC03-Co3 lakes. J. Exp. Biol. 101: 171-186. Szibbo, CM. and Scudder, G.G.E. 1979. Secretory activity of the segmented Malpighian tubules of Cenocorixa bifida (Hung.) (Hemiptera, Corixidae). J. Insect Physiol. 25: 931-937. Tolman, J.H. and Steele, J.E. 1976. A ouabain-sensitive, (Na++K+)- activated ATPase in the rectal epithelium of the American cockroach, Periplaneta americana. Insect Biochem. 6: 513-517. Treherne, J.E. 1966. The effect of ouabain on the efflux of sodium ions in the nerve cord of two insect species (Periplaneta americana and Carausius morosus). J. Exp. Biol. 44: 355-362. Vaughan, F.A. 1979. Effect of gross cardiac glycoside content of seeds of common milkweed, Asclepias syriaca, on cardiac glycoside uptake by the milkweed bug Oncopeltus fasciatus. J. Chem. Ecol. 5: 89-100. Vaughan, G.L. and Jungreis, A.M. 1977. Insensitivity of lepidopteran tissues to ouabain: physiological mechanisms for protection from cardiac glycosides. J. Insect Physiol. 23: 585-589. von Euw, J., Fishelson, L., Parsons, J.A., Reichstein, T. and Rothschild, M. 1967. Cardenolides (heart poisons) in a grasshopper feeding on milkweeds. Nature 214: 35-39. von Euw, J., Reichstein, T. and Rothschild, M. 1971. Heart poisons (cardiac glycosides) in the Lygaeid bugs Caenocoris nerii and Spilostethus pandurus. Insect Biochem. 1: 373-384. von Schenker, E., Hunger, A., and Reichstein, T. 1954. Zur papierchromatographie von stark polaren herzwirksamen glykosiden und aglykonen. Helv. chim. Acta 37: 680-685. Whittaker, R.H. and Feeny, P.P. 1971. Allelochemics: chemical interactions between species. Science 171: 757-770. Wigglesworth, V.B. 1931. The physiology of excretion in a blood-sucking insect, Rhodnius prolixus (Hemiptera, Reduviidae). III. The mechanism 124 of uric acid excretion. J. Exp. Biol. 8: 443-451. Wright, S.E. 1960. The Metabolism of Cardiac Glycosides. CC. Thomas, Springfield, Illinois: 10-18. Yoder, C. A., Leonard, D. E. and Lerner, J. 1976. Intestinal uptake of ouabain and digitoxin in the milkweed bug, Oncopeltus fasciatus. Experentia 32: 1549-1550.