Inducible lectins from hemolymph of Anticarsia gemmatalis Hübner (Lepidoptera: Noctuidae)


Material Information

Inducible lectins from hemolymph of Anticarsia gemmatalis Hübner (Lepidoptera: Noctuidae)
Physical Description:
xi, 148 leaves : ill., photos ; 29 cm.
Heath, Martha A., 1945-
Publication Date:


Subjects / Keywords:
Velvet-bean caterpillar   ( lcsh )
Lectins   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1991.
Includes bibliographical references (leaves 134-147).
Statement of Responsibility:
by Martha A. Heath.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001747384
notis - AJG0207
oclc - 26371930
System ID:

Full Text

Anticarsia gemmatalis Hibner
(LEPIDOPTERA: Noctuidae)







First, I would like to thank my husband for encouraging

me to pursue this project which would not have been

undertaken without assurance of his confidence and support.

Next, I would like to thank the members of my graduate

committee for their continuous support. Each member helped

in many ways and if I did not take full advantage of the

opportunities afforded, it was to my detriment. I thank Dr.

Drion Boucias for serving as committee chairman, for

providing equipment and supplies, and allowing me to conduct

this research in his laboratory. Dr. Pendland served as my

supervisor during the sabbatical of Dr. Boucias. Dr. Nation

has always been available for consultation about the intent

of the Doctoral Dissertation. It was during a conversation

with Dr. Maruniak that I decided to pursue the proper

purification protocol. Dr. Greany has helped with

computerized photography and digitizing of gels. Dr.

Aldrich was available to help with EM experiments that I was

unfortunately unable to perform.

I would like to thank the Department of Entomology and

Nematology especially Dr. John Strayer and Dr. John Capinera

for the opportunity to work in this department. Ms. Myrna

Litchfield, Ms. Sheila Eldridge and Ms. Kris Faircloth have

provided wonderful support in their specialized areas.


ACKNOWLEDGEMENTS.................................... ii

LIST OF TABLES .................................... v

LIST OF FIGURES...................................... vii

ABBREVIATIONS ....................................... ix

ABSTRACT........ .. ............ .................... x


Lectins. .......... ... .... ... ... 1
Insect Lectins.......................... 11
Lectin Specificity and Function.......... 27
Research Aims......................... 29

Anticarsia aemmatalis........................ 32

Introduction...................... ..... 32
Materials and Methods.................... 34
Results..... ............ ....... .......... 35
Discussion.. ............o..... ............ 41

CHARACTERIZATION.......................... 43

Introduction.... .................... 43
Materials and Methods.................. 44
Results..... ... ..... ....... ... 49
Discussion............................... 61

4 SUGAR INHIBITION PROFILES................... 67

Introduction.......................... 67
Materials and Methods................... 68
Results............................... 71
Discussion.................... ........... 81


5 OPSONIC PROPERTIES OF Anticarsia gemmatalis
HEMOLYMPH....... ........ ... ..... ..... ....... 91

Introduction............................. 91
Materials and Methods.................. 93
Results..... ........ ......... ......... 96
Discussion............................... 105

6 SUMMARY AND CONCLUSIONS..................... 111



B MARTHA'S COOKBOOK............................. 127

REFERENCES CITED....................................... 134

BIOGRAPHICAL SKETCH............................. ........ 148


Table page

1-1. Properties of fucose lectins.................... .... 12

1-2. Hemagglutinin activity in insect hemolymph....... 16

1-3. Hemagglutinin activity in insect tissue other
than hemolymph ......................................... 20

1-4. Inducible hemagglutinins in insects .............. 22

2-1. Hemagglutinin activity in hemolymph of larval
Anticarsia gemmatalis challenged with various microbial
agents.................................................. 38

2-2. Mortality of insects challenged with various
microbial agents........................................ 39

2-3. Average titers of fungal challenged insects....... 40

2-4. Occurrence of hemagglutinin in life stages of
Anticarsia emmatalis.................................... 40

2-5. Occurrence of hemagglutinin in larvae of
A. gemmatalis after heat shock......................... 40

3-1. Inhibition by galactose and fucose of fractions of
fucose binding proteins eluting from Sephacryl S-300.... 54

3-2. Inhibition by galactose and fucose of fractions of
galactose binding proteins eluting from Sephacryl S-300. 54

3-3. Characteristics of fractions from A. gemmatalis
lectin purification.............................. ...... 57

3-4. Characteristics of reconstituted lyophilized
preparations of lectin from A gemmatalis................ 57

4-1. Titers of whole hemolymph at various intervals
following lectin induction and in the presence of various
buffer systems ....................................... 73

4-2. Hemagglutination inhibition of erythrocytes by
immune hemolymph of Anticarsia emmatalis.............. 74

4-3. Hemagglutination inhibition of trypsinized rabbit
erythrocytes exhibited by four units agglutinin using
PBS or BIS as buffer................................... 75

4-4. Hemagglutination inhibition of trypsinized rabbit
erythrocytes by 0.2 M sugar of twice purified lectins from
Anticarsia emmatalis ................................. 75

4-5. Sugar inhibition profiles of twice purified
galactose and fucose Anticarsia gemmatalis lectin using
HI-II ....... .................................. ......... 76

4-6. Block titration of serial twofold dilutions of whole
immune hemolymph and serial twofold dilutions of L-fucose
in BIS.................................... ..... ...... 77

4-7. Block titration of serial twofold dilutions of whole
immune hemolymph and serial twofold dilutions of
D-galactose in BIS....................................... 78

4-8. Hemagglutination titers of nontrypsinized human
erythrocytes ............................................ 79

4-9. Hemagglutination titers of trypsinized human
erythrocytes..... ..... ..... .... ......................... 79

5-1. Number of circulating hemocytes (x 106) at various
time intervals following treatment with opsonized and
nonopsonized fungi...................... .......... ...... 101

A-1. Activity recovered from BioGel 1.5A under various
buffer conditions and % of activity inhibited by galactose
and fucose ......................................... 126




Figure page

3-1. Absorbance profile (280 nm) of galactose binding
proteins from Sephacryl S-300........................... 52

3-2. Absorbance profile (280 nm) of fucose binding
proteins from Sephacryl S-300........................... 52

3-3. Hemagglutinin activity recovered from pooled
fractions eluted from Sephacryl S-300................... ... 53

3-4. SDS-PAGE of fucose binding proteins eluting in
various fractions from Sephacryl S-300................... 55

3-5. SDS-PAGE of galactose binding proteins eluting in
various fractions from Sephacryl S-300.................. 56

3-6. SDS-PAGE of fractions obtained during the
purification of fucose binding proteins by double
affinity chromatography.................. ............... 58

3-7. SDS-PAGE of fractions obtained during the
purification of galactose binding proteins by double
affinity chromatography.................................. 59

3-8. SDS-PAGE of 2X purified galactose and fucose lectins
purified by double affinity chromatography .............. 60

5-1. Clearance of opsonized and nonopsonized P. farinosus
from hemolymph of A. gemmatalis.......................... 99

5-2. Clearance of opsonized and nonopsonized H. rilevi
from hemolymph of A. Qemmatalis......................... 100

5-3. Individual hemagglutination profiles of normal
hemolymph samples from untreated 6th instar larval
Anticarsia gemmatalis .................................... 102

5-4. Hemagglutination profiles of hemolymph from A.
gemmatalis treated with opsonized and nonopsonized H.
rileyi........................................... .. .... 103


5-5. Hemagglutination profiles of hemolymph from A.
aemmatalis treated with opsonized and nonopsonized P.
farinosus............................................... 104

A-1. Elution patterns of hemagglutinin activity from
BioGel 1.5A under different buffer conditions and when
analyzing normal hemolymph.............................. 123

A-2. Protein elution profile from BioGel 1.5A with 0.2M
Gal-BIS as buffer....................................... 124

A-3. Protein elution profile from BioGel 1.5A with BIS
without Ca" as buffer...................... ............. 124

A-4. Protein elution profile from BioGel 1.5A with Tris
Buffer with 0.15M NaCl + EDTA as buffer.................. 124

A-5. Protein elution profile from BioGel 1.5A with BIS
as buffer ................................................ 125

A-6. Protein elution profile from BioGel 1.5A using
normal hemolymph and BIS as buffer......................125





a linkage
N-acetyl neuraminic acid (sialic acid)
P linkage
bovine submaxillary mucin
2-deoxy glucose
dextran sulfate
human ABO erythrocytes
human O erythrocytes
a-methyl galactose
a-methyl mannoside
neuraminidase treated
porcine submaxillary mucin
rabbit erythrocytes

not determined

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

Anticarsia gemmatalis Hibner
(LEPIDOPTERA: Noctuidae)


Martha A. Heath

May 1991

Chairman: D. G. Boucias
Major Department: Entomology and Nematology

The inducible hemagglutinins (lectins) of larval

Anticarsia gemmatalis were studied. Although agglutinins

were present in low levels in noninduced hemolymph,

injection of hyphal bodies from Nomuraea rileyi and

blastospores of other fungi induced appearance of high

titers of these molecules in the hemolymph. The induced

hemolymph was termed immune hemolymph. Injection of

bacteria and viruses did not induce similarly high titers.

The lectin system appears to consist of a galactose-

binding component and a fucose-binding component. Both

components were isolated by a novel sequential affinity

chromatography procedure. Immune hemolymph was diluted in

buffer containing fucose and applied to a galactose-agarose

column. Galactose-binding molecules were eluted with buffer

containing galactose. The fucose-binding proteins in the

first eluent were dialyzed to remove fucose, applied to a

fucose-agarose column and eluted with buffer containing

fucose. The galactose agglutinin and a partially purified

fucose agglutinin were obtained. When analyzed by SDS-PAGE,

the agglutinins of two sugar specificities were found to

have subunits of similar MW (around 45 kd). Native gel

electrophoresis and gel permeation chromatography

demonstrated that the native hemagglutinin had an MW >300


The hemagglutinin activity of the galactose lectin was

inhibited by Amolar quantities of D-galactose and galactose-

related sugars. The hemagglutinin activity of the fucose

lectin required the presence of Ca" and was inhibited by

mM quantities of L-fucose. Neither lectin showed

specificity for erythrocytes of the human ABO blood groups.

Opsonization of fungal blastospores possessing exposed

galactose residues with immune hemolymph accelerated

clearance of these particles from hemolymph. Opsonization

of fungal hyphal bodies lacking exposed galactose residues

with immune hemolymph, normal hemolymph did not effect

clearance of these particles from hemolymph. Clearance may

be a function of the galactose-binding hemagglutinin.



This chapter reviews highlights of the current

literature on lectins pertinent to this dissertation. It

addresses the evolution that has occurred in the concept of

a lectin over the last century and how these molecules are

being studied in the 1990s. This review places special

emphasis on animal lectins, insect lectins and fucose

lectins. Known and surmised functions of lectins are

discussed. The research presented in this dissertation is

introduced and objectives of the study are presented.


Lectin research began after Stillmark (1888) discovered

hemagglutinating activity in extracts from castor beans.

This proteinaceous carbohydrate-binding molecule was ricin.

Since the initial discovery, lectins have been found in

virtually every group of organisms investigated for their

occurrence. The term lectin (from the Latin "lectus," to

select) was coined by Boyd and Sharpleigh in 1954 to

emphasize their selective binding to complex

glycoconjugates. Because lectins are frequently detected by

their ability to agglutinate erythrocytes, until recently

they were narrowly defined based on specificity for

particular sugar moieties (frequently monosaccharides) and

their bi- or polyvalency which permitted cross-linking of

sugar residues. As cited in reviews by Gallagher (1984) and

Pusztai (1987), Makaela (1957) recognized four groups of

lectins based on configuration of the hydroxyl groups at the

C-3 and C-4 position on the monosaccharide. This scheme

ignored anomeric specificity at C-l, a property exhibited by

lectins with specificity for the human ABO blood groups.

Goldstein and Hayes (1978) attempted to improve this scheme

by listing sugar binding in order of specificity and

denoting any anomeric preference e.g. a-lactose > a-

galactose. Goldstein et al., 1980, proposed a definition

which required a lectin to have at least two sugar-binding

sites. Gallagher (1984) presented a classification scheme

based on whether 1) the lectin was inhibited by a

monosaccharide (Class I or simpler binding mode lectin) or

oligosaccharide (Class II or complex binding mode lectin);

2) the monosaccharide-inhibitable lectin showed requirement

for end-chain monosaccharides (obligate exolectin) or

recognized both end-chain and internal monosaccharides

(facultative exolectin); 3) the lectins which recognized

only specific sugar sequences oligosaccharidess) required

homotypic sugar sequences or heterotypic sugar sequences.

For the Class I lectins, the definition included a


requirement that the lectin binding be inhibited by greater

than 50% in the presence of 10 mM or less of the inhibiting

monosaccharide. In a review of plant lectins, Etzler (1985)

expanded the definition to include monovalent carbohydrate

binding proteins.

In 1988, Barondes proposed to redefine lectins simply

as "carbohydrate binding proteins other than enzymes and

antibodies" (Barondes, 1988, p. 482) because of the growing

body of data on endogenous animal lectins which did not fit

the classical definition. This new definition encourages

investigators to look at lectins from the functional

viewpoint that lectins are recognition molecules and

Yoshizaki suggested additionally that the definition would

prove useful in studying the "molecular evolution of

carbohydrate-binding proteins" (Yoshizaki, 1990, p. 589).

As pointed out by Sharon and Lis "four different

monosaccharides can form 35,560 distinct tetrasaccharides,

whereas four different amino acids or nucleotides can form

only 24 tetrameric structures" (Sharon and Lis, 1989, p.

227). It has been hypothesized that "the specificity of

many natural polymers is written in terms of sugar residues

and not of amino acids or nucleotides" (Sharon, 1975, p.

26). Plant and animal lectins have been shown to possess

exquisite specificity in their ability to distinguish subtle

differences in carbohydrate structures and this suggests

that it is a lack of knowledge of carbohydrate chemistry

that is limiting our ability to understand the true

oligosaccharide specificity. Deciphering the information

carried by oligosaccharides may represent a new frontier in

biology. Feizi (1988) outlined strategies for decoding this

information using oligosaccharides as antigens, receptors

and probes.

Lectins are frequently named by several schemes. In

one common method, the first letters of the genus and

species are used and L (for lectin) or A (for agglutinin) is

added. Component lectins are designated by letters or

numbers. For example, the lectin from castor bean, Ricinus

communis, is termed RCA. It consists of two types of

molecules. The toxic A chain can be designated RCA-A and

the agglutinin, the B chain, RCA-B. The components could

also be termed RCL-I and RCL-II, RCL-A and RCL-B, or RCA-I

and RCA-II. The common name of a lectin is usually derived

from the species name and the suffix "-in" added e.g. ricin.

Frequently, investigators working on a particular lectin

have adopted their own conventional terminology which may

not be consistent with that adopted for another lectin.

Purified and partially purified lectin preparations are

widely available commercially. Two common sources are Sigma

Chemical Co. (St. Louis, MO) and EY Laboratories (San Mateo,

CA). A general source of information on lectins is the

Sigma catalog (Sigma Chemical Co., St. Louis, MO). Perusal

of the lectin section reveals most commercially available

lectins are of plant origin and reflects the relative

abundance of investigations on plant lectins compared to

those from other organisms. Several lectins from animal,

bacterial and fungal sources are also commercially

available. Also obvious are the broad sugar specificities

exhibited by lectins. Additionally, lectins have been

conjugated to agents like fluorescein isothiocyanate (FITC),

biotin, peroxidase and colloidal gold for use in assay

methods other than hemagglutination assays. Use of labelled

lectins has allowed characterization of exposed surface

carbohydrates of cells and microorganisms. Work conducted

in this laboratory using such probes has been useful in

delineating the surface carbohydrates of entomopathogenic

fungi and insect cells (Pendland and Boucias, 1984, 1986b,

in press).

Some of the well-studied plant lectins are interesting

because their biochemical diversity provides prototypes for

comparing and categorizing lectins from other species. Most

lectins are composed of multiple subunits and are often

glycosylated. The propensity of lectins to bind

carbohydrate moieties, including those possessed by

component subunits, makes it difficult to determine native

molecular weight. One of the classic lectins, concanavalin

A (Con A), is a product of jackbean (Canavalia ensiformis)

seeds. The lectin is glycosylated and consists of 4

subunits of MW 26,000 but undergoes varying degrees of

aggregation depending on pH (Kalb and Lustig, 1968). Con A

recognizes D-glucose and D-mannose residues of the a- anomer

and requires the divalent cations Ca+* and Mn* (Reeke et

al., 1974). Although not useful in blood typing, Con A

represents one of the few lectins which possesses mannose

specificity (its recognition capabilities for mannose are 10

times greater than for glucose). Other interesting lectins

are found in seeds of Bandeiraea simplicifolia. One

component lectin consists of 5 possible combinations

(isolectins) of subunits BS-I (A) and BS-I (B) and has an

aggregate MW of 114 kd. The BS-I (A) subunit recognizes a-

D-N-acetyl-galactosamine while the BS-I (B) subunit

recognizes a-D galactose. The possible compositions of

isolectins are A4:A3B1:A2B2:A1B3:B4 (Hayes and Goldstein,

1974; Murphy and Goldstein, 1977). The B. simplicifolia

lectin system also has another component, BS-II, which

recognizes a-D-N-acetyl glucosamine (GlcNAc) (Shankar Iyer

et Al., 1976). Wheat germ agglutinin (WGA) consists of two

subunits of 36 kd and exhibits specificity for N-acetyl-

glucosamine dimers and N-acetyl-neuraminic acid (Allen et

al., 1973; Nagata and Burger, 1974). An L-fucose-binding

lectin is produced by the asparagus pea, Lotus

tetragonobolus. It consists of 10 subunits all of which

exhibit affinity for a-L-fucose. They exist in the ratio of

4:2:4 and are of approximately the same MW (Yariv et Al.,

1967; Kalb, 1968). Another lectin system occurs in gorse,

Ulex europeus and is composed of two types of molecules:

UEA-I has specificity for a-L-fucose while UEA-II in

inhibited by N,N' acetylchitobiose and salicin (Matsumoto

and Osawa, 1969, 1970; Osawa and Matsumoto, 1972).

Plant lectins are ubiquitous and seeds of the families

of Leguminosae and Graminaceae are particularly rich

sources. The biological functions of plant lectins have

been reviewed by Pusztai (1987) but in general, the

functions of these molecules are obscure although they may

represent a significant percentage of the total protein of a

given source. Recently two papers have appeared in the

literature which may open up an entirely new area for the

study of plant and insect herbivore relationships. Murdoch

et al. (1990) studied the effect of plant lectins on the

cowpea weevil and found that five of the 17 lectins screened

delayed the development of these insects. Pratt et al.

(1990) found that addition to artificial diet of purified

lectin from the tepary bean (Phaseolus acutifolius) delayed

development in the bruchid beetle, Acanthoscelides obtectus.

Since most Phaseolus spp. are known to contain lectins,

these proteinaceous chemicals can now be argued to play a

role in plant defense against insect predation as suggested

by Janzen et al. (1976).

Lectins can also be useful as tools (Sharon, 1987), and

only a few actual applications and potential applications

will be discussed in this review. Lectins are routinely

used as diagnostic reagents for typing of the human ABO

blood system. Others, such as pokeweed mitogen and

phytohemagglutinin, function as mitogens for subsets of

lymphocytes and as such have been successfully used to

stimulate cell division and antibody production in normally

unreceptive B cells and for stimulating nonreceptive T

cells. These properties have proven extremely useful for

immunologists. The castor bean lectin, ricin, is being

researched for use in immunotherapy as an immunotoxin (see

Vitetta and Uhr, 1985, and Olsnes et al., 1989, for review).

Ricin is known to consist of a toxin component which can

recognize N-acetyl-D-galactosamine and P-D-galactose and an

agglutinin component which recognizes P-D-galactose. In

theory, an antibody is raised against a tumor specific

antigen and this antibody is completed to the ricin toxin.

After this complex is administered to a patient, the

antibody targets the tumor cells bearing the foreign

antigen. After binding, the antibody-toxin complex is

endocytosed by the cell which succumbs to the toxin.

In the animal kingdom, there has been an expanding body

of data on endogenous lectins but many of these molecules

failed to fit the classical definition of a lectin because

they lacked cross-linking capability (valency). As

mentioned above, Barondes (1988) proposed to redefine

"lectin" to include these molecules. As animal lectins have

become better studied, functional differences appeared and

two classes--membrane lectins and soluble lectins--could be

distinguished (Barondes, 1984; Leffler at al., 1989; Caron

et al., 1990). Membrane lectins require detergent

solubilization prior to purification and apparently function

to bind glycoconjugates to membranes. The soluble lectins

can move freely and interact with both soluble and membrane-

bound glycoconjugates. Recently, there has been growing

interest in a family of soluble galactoside-binding

vertebrate lectins. These can be subdivided into two

classes: S-type lectins and C-type lectins (Drickamer,

1988). C-type lectins require Ca*, possess cysteine

molecules as disulfides, show variable solubility, are

extracellular and have varying carbohydrate specificities.

In contrast, the S-type lectins do not require Ca** and

possess cysteine molecules as free thiols. They are soluble

in buffer and may exist intracellularly or extracellularly

and have preference for terminal P-galactosides. These

lectins will be discussed because they probably represent

the prototypes of the lectins occurring in the Anticarsia

cemmatalis system and in other insects.

As mentioned above, the C-type endogenous animal

lectins have a requirement for Ca". Membership in this

large family has been corroborated through amino acid

sequence data. Vertebrate members include the

asialoglycoprotein receptor, chicken hepatic lectin, mannose

binding protein, and lymphocyte Fc receptor. Two

invertebrate lectins, the Sarcophaga perearina lectin and

sea urchin lectin, have also been included in this group.

The common feature of such diverse molecules is a

carbohydrate recognition domain (CRD) with 18 conserved

amino acids within the 130 amino acid domain.

The S-type lectins are the second family of soluble

animal lectins. They show no requirement for divalent

cations but are usually assayed in the presence of thiols

since oxidation inhibits carbohydrate binding activity.

Another common feature of this group is molecule weight of

14-16 kd. Structural analysis of the carbohydrate domain

reveals similarity within the group and no similarity to CRD

of C-type lectins. Additionally, there are no cysteine

residues in the invariant region.

Some of the endogenous animal lectins have a more

complex nature than carbohydrate specificity. These have

been reviewed by Barondes (1988). At least two, discoidin

I, a lectin from the cellular slime mold Dictvostelium

discoideum, and the elastin receptor can mediate both

carbohydrate-protein and protein-protein interactions.

Discoidin I is a classical lectin that is developmentally

regulated. In the aggregating stage of this slime mold, the

lectin is exocytosed in multilaminar bodies which form the

matrix upon which the cells aggregate. The lectin also has

a protein binding site which recognizes the tripeptide Arg-

Gly-Asp and is essential for aggregate morphogenesis in this


organism. This tripeptide is identical to that required for

substratum adhesion in fibronectin and laminin. The elastin

receptor also shows bifunctionality since it can be affinity

purified on either an elastin or glycoconjugate affinity

matrix and eluted from these columns respectively by the

peptide Val-Gly-Val-Ala-Pro-Gly or lactose (Wrenn et al.

1988; Hinek et al., 1988). These investigators have also

suggested that the function of this receptor is to provide

for the proper alignment of the elastin microfibrils to

themselves and the extracellular matrix. One insect lectin

has been reported to be bifunctional. The inducible lectin

from Manduca sexta can function as a hemocyte coagulant

(Minnick et al., 1986).

Lectins which recognize L-fucose are unique in that

they recognize a biologically occurring sugar of the L

enantiomer. A fucose lectin appears to exist in A.

gemmatalis. For comparison with the A. gemmatalis fucose

binding proteins, properties of known fucose lectins are

summarized in Table 1-1. Although widely distributed in

nature, fucose lectins are considered uncommon (Gilboa-

Garber et al., 1988).

Insect Lectins

Hemagglutinins were first reported to occur in insects

by Bernheimer (1952). He detected this activity in larvae

and pupae in 10 of the 46 lepidopteran species examined.

t 3 t

OC l

541 -4 u




l- 4J








0 0

0 i

a u

r. 0

4 *a*


>im C

CO ** 0
** N ,
O ** O
(*4'm FI
g-4(v *-4


a o


u o

C a



x U



t<. IO

0 0
3O- ( *
A o N to

0 0 l
C 04 I V 0-
U0 to D V 4 C **
O V 0 4) V-l to 0 go"

V z 0 4I > t.
q0 00 40 V V0 3'-V 0-)
Wc *4 e 10 > 0
+ 3 0 0 i# 010 M k .0
P 0 0 W ** 44 M 0 *





0 *
N 10 #00
*^ *>**1* 1 ****

0 0)


M4 s

4.) H0

Go 0 1 -
(A *0 c=, I to rl
vC ,m M (A

o-4 0 4

** (a rl k

W FOAS 4-B 0

S0 t4 *41 k *

0 0M O 0 01
v) r I k N CCal >
% 0 O 4) tol i rd r)t
0 0 a (
4) 40 0- 1 0 W-

0 r'r.10 0 oo

r m qr Ln %0 R)- 0 0
a ^ o--> 3 i ( e -'-

None of the butterflies showed activity and the eggs and

adults of the moths were similarly negative when tested

against erythrocytes of various human blood groups (ABO,

Rh,M and N). Invertebrate lectins continued to receive

attention from numerous investigators (for reviews see

Ratcliffe and Rowley, 1983; Rowley et al., 1986). By 1986,

endogenous hemagglutinins had been found in six orders

including 15 species and inducible agglutinins were reported

from four species representing four orders (Ratcliffe and

Rowley et al., 1986). Recently (1986), in Hemocvtic and

Humoral Immunity in Arthropods, edited by A.P. Gupta, Hapner

and Stebbins summarized properties of purified lectins from

seven insects in three orders. The chapter by Rowley et

al., 1986, reviewed humoral recognition factors in insects.

Since these reviews, lectins have been reported to occur in

other insects. A summary of the well-studied insect lectins

is presented in Table 1-2.

In most species, lectins have been detected in

hemolymph but are known to occur in other tissues. These

findings are summarized in Table 1-3. Most lectins have

been found to occur constitutively but several inducible

lectins are known. Known inducible insect lectins are

listed in Table 1-4. In several insect species, agglutinins

occur which recognize microorganisms. In Diptera, molecules

have been found which agglutinate Trvpanosoma, Leishmania,

and Crithidia (Ingram et al., 1983; Ingram et al., 1984;

Ibrahim et al., 1984) and in the lepidopteran Philosamia

ricini similar molecules are active against Bacillus

thurinaiensis (Bellah et al., 1988). In honeybees,

agglutinins are synthesized in response to a bacterial

pathogen (Bacillus larvae) (Gilliam and Jeter, 1970).

For the most part, the site of synthesis and function

of insect lectins are poorly understood. Although it seems

that insect hemagglutinins will exhibit the same broad range

of specificities found in lectins from species in other

kingdoms, a significant number of insect lectins appear to

be specific for galactosyl residues.

The best studied insect lectin occurs constitutively

in pupae of Sarcophaaa peregrina (Komano et al., 1980, 1983)

but can also be induced in larvae and adults (Kubo et Al.,

1984) by injury to the body wall. This activation has been

shown to be mediated by a humoral factor in vivo and in

cultured fat body (Shiraishi and Natori, 1988, 1989). It

has been suggested that this galactose-inhibitable lectin

functions during pupation in recognition of effete larval

tissue and as a wound response protein in larvae (Komano et

al., 1981). It has also assists in removal and lysis of
foreign tissue (Komano and Natori, 1985). This lectin is

synthesized by the fat body but can be detected on the

surface of hemocytes. Studies have been expanded to cloning

and sequencing of the lectin gene (Takahashi t al., 1985)

and cloning and in vitro transcription of the lectin gene


a U. I
4 0) .4 EU C

41 0 r- to
to 8 'A A A

i' i 0
00 0 G0 I-

544) 04) 4-r4 4
:3, O^ $4 <

S r4 -4 0
00 0

PC 00 t
1 on 4 II

t4) 0

-4 +41 i w I
U4 00 r.
> 'O CEl
> 40 00 C 0
H -# t t 0 in co
4 0 40 )
<> int- 0 An
c c

to :3 0 4 r-4
H Co 0
-A 00 0 0

F4 r-4 1 0
14 o It- p 04 COA 5 4 V UV V

4) U s ON 0a 0
0 0 0 0 0H
to ) t t t 0 g

*. *. r!

to z V
(0 r-4 4
0#-I -
0 0 to4
1 0

Q 2w
0d i0

F- i- 'S
0 Q CQ

F0 0*
) *r4
C -

0 1






V r-












M '-c






t r.


a ts
ca *0

cU r-

8s U


a >



(0 0

X; 0)


.9 0 ro

to -4 4,) 0 t
- 0 C

N to .4-
4000 M0

k0' j Dax
000l I 0

to o

~4 i
0 0r-4 0
0 5.44

0 I 4)' -o & C O 0
Iy 10 .%o i
S15 4) 034

S0 C
gnm^ 3U- X- 1 4O O''IM

S~~~~~~ a oioi



o N
r-I %.

0 0

C 04) 4 )
94 94 0 0
^' Ca 0a

0 V-4 0 V 0 1

4 I8 0) a

vW 0 A0 S.V 9 0 w m 9 0
V 0.) 0 0 0 00

00A 09$4W r-4 : 4 0 U 4 0

S4 00 c 1 I 2 4
CA14 Q Q00 r0 t g a

40 0 0 0

I C ) C C C
? O 0 U I
0 0 0*C 4 g
Ca 31 a 1o
4)1) ..4 4 3 0
0 C 0 .90
09 4H 0 4) "q
4 V A 4 g g0 00 FA

4ON OP O4 9 4

0 0 40# *0 3 4 414 .M

0 0V

0 0 0 0 0 0
C C d > C (C C (v o..

5 0

0 o I In .. M
4 6 0* 0 0
Cn Nn N ( %

> >rl %0r( N4 % r N r0

0 I f O 0S '0

94 H0-< 0 -* E

o n w I u0 *6E*

V +j n O O I t 0

04 toW to Ol 0 0 >I0
(i _. 0 9
0% 2 4 0 to 0 94 9 094 0 r4
4 0 t-


z z


.0 0
u<-d rlu




0 0


* --4
. t;








X ,



1 ,-

k t0 to tC
4 0 o

P4 r )

S 43 4 M
i< I a<<*> &< ^


04 I 00

0O O .4400 M o
4 0C00O


O m w Ag -4H14

S00 00 0
rC f H H 4 H V f N

10 000
04 O *H OV4 H
0 H| C 1 (l
C C 01 4 01

S04O 4 -4 rB 0 -
~~ 4))1N r4
FlA i wl l tol vl

V 64- 9 o5
.0C 00 4JW 00 r.
04 000 0 Q4 56

cn I c rn 0

*H ^^1 < 0 0
j30 B^A^ ^
CIA cxAA~ ii i)o





00 o
S.40 0N
4A *1 n

0- 0 U4 0
OZ r4o0
0 V 4S4 0 Z











p .p

V 4
5-I 54










0- 0
M( r-
eU *-





0 1

1 a




0 Of

* -
0 b


0 0



A 0





0 (




00 0
ca 3 k0

srl 4 (

rC4 r-4 0 0

c- c9 H'-* a0
%, 0 '40 4. 0 Ot


ea W
A 0 I-
ty >

0 4
C 0 0

a a
v0 0 v









-4 0

r-4 N

4) V

0 t
0 I0

0 O

0 4

0 0

H 0


0 *e
S% *0000
>IC ) 0
N 4 r-r 1O
N o i-l ,

cO c0 0
EU -S0% -UA
ft rl 0044

34 4J)4)4 C CO J
= z NH inU 04 z


V 4

# (



$14 4-)

4 >t
ot to
10 PI

u ^


C*^ a

.r4 r.
.0 to


0 (

4 4-a
A m

40 *
(0 0
NO 4
^*g &






& S

> r









M <~



cOU r-i0
0.0 04-%
o 0 -0
4 O *-I


04'a Z -H
B (0 x 0

4| (cUifl
WO '0h
a c ^ +>|

r4 4t N

(0 -r
to 0 t e
0m 04 M

(Q -^ S w 0 (

(Kobayashi et al., 1989). Other elegant studies have

demonstrated that a receptor for this lectin exists on the

surface of mouse macrophages (Ohkuma et al., 1988) and that

once stimulated, the macrophages are induced to produce a

protein toxic to murine sarcoma cells (Itoh et al., 1986).

A constitutive galactose/glucose lectin exists in adult

grasshoppers (Jurenka et al., 1982; Hapner, 1983; Stebbins

and Hapner, 1985) and is essentially the same in both

species studied (Melanoplus sanguinives and M.

differentialis). It is synthesized by the fat body (Stiles

et al., 1988) and can be localized on the surface of

hemocytes (Bradley et al., 1989). This lectin has been

reported to lack an opsonic function but the surface

carbohydrates of the test organism were not well-

characterized and further studies need to be conducted.

In the cockroach Leucophaea maderae (Amirante and

Mazzalai, 1978) and the giant silk moth Hyalophora cecroDia

(Yeaton, 1981), evidence for hemocytes as a synthetic source

of lectin exists. The lectins of these insects have not

been investigated for a function. The lectins from

Hyalophora cecropia are of interest because superficially

they appear similar to the group of isolectins found in the

plant, U. simplicifolia. However, the A chain is specific

for both galactose and N-acetylgalactosamine while the B

chain recognizes an unknown moiety which may not be a

carbohydrate. The MW is 160 kd with A subunits of 40 and 41

kd (N-acetyl-D-galactosamine and galactose) and B subunits

of 37 and 38 kd (specificity unknown).

In this laboratory, Pendland and Boucias have purified a

galactose lectin from Spodoptera exioua (Pendland and

Boucias, 1986a) and demonstrated that it can function as an

opsonin and enhance the phagocytosis of fungi which possess

exposed galactose residues on their cell wall surfaces

(Pendland et al., 1988). The site of synthesis of this

constitutive lectin has yet to be explored. An inducible

galactose lectin occurs in Anticarsia gemmatalis (Pendland

and Boucias, 1985). Attempts to purify this lectin

according to methods employed for the S. exigua lectin were

unsuccessful and crossabsorption studies using human type 0

and rabbit erythrocytes demonstrated the relative complexity

of this system. The purification and characterization of

components of this lectin system is the subject of this

dissertation and the research objective will be introduced

at the end of this review.

Additionally, in this laboratory, various economically

important noctuid larvae have been examined for the presence

of hemagglutinins. In the Plusiinae, Pendland and Boucias

(1985) failed to find lectin activity in hemolymph of larval

Trichoplusia ni and Heath (unpublished) failed to find

activity in Pseudoplusia includes. Challenge by

intrahemocoelic injection of hyphal bodies from Nomuraea

rilevi did not induce agglutinin production. In the

Agrotinae, both Heliothis zea and Heliothis virescens

possess endogenous lectins with galactose specificity

(Heath, unpublished). The naturally occurring hemagglutinin

titers of H. virescens are higher (>16,000) than those of H.

zea (range 16 >1024) when assayed against trypsinized

rabbit erythrocytes. The H. zea lectin is amenable to

purification by affinity chromatography using a galactose-

agarose resin and has essentially been purified. The MW is

in the 30 kd range. The twice purified H. zea lectin is

inhibited best by galactose and galactose derivatives but

can also be inhibited by glucose and glucose related sugars.

This activity is heat labile.

Since lectins and lectin-like molecules have been found

on cell surfaces (either sequestered or synthesized by the

cells), the lectin system found in the snail, Biomphalaria

glabrata, is of interest in that it may serve as a prototype

for the insect lectin function. Fryer et Al. (1989)

conducted experiments which can provide a model for

opsonization and phagocytosis experiments. Using yeast

(Saccharomvces cerevisiae) with known cell wall chemistry,

these investigators presented evidence that the circulating

hemolymph lectin and hemocyte recognition molecules exhibit

different specificities. They postulated that the opsonic

hemolymph factor (inhibitable by mannose) undergoes

conformational changes once bound to yeast mannans and these

changes subsequently allow binding to hemocyte surfaces

(inhibitable by P-1,3 glucan) and lead to enhanced

phagocytosis (Bayne and Fryer, 1989). In molluscs the

hemocytes are the putative site of lectin synthesis (Van der

Knaap and Loker, 1990).

The recognition of P-1,3 glucans has been linked to

nonself recognition in insects (Ashida et al., 1982) and

this recognition system resides in the hemolymph and exists

as a separate entity from the peptidoglycan recognition

system in B. mori (Yoshida et al., 1986). Ochiai and Ashida

(1988) have purified a 0-1,3 glucan recognition protein from

B. mori hemolymph. A 3-1,3 glucan binding protein has also

been isolated from Blaberus craniifer (S6derhall et al.,

1988). In the well-studied prophenoloxidase system of the

crayfish, Duvic and S6derhall (1990) purified a 3-1,3 glucan

binding protein from plasma. Recently, Matha et al., 1990a,

and Matha et al., 1990b, reported the existence of a P-1,3

glucan lectin in Galleria mellonella. It was isolated from

hemocyte lysate and localized in the plasmatocyte class of

hemocytes. Immunocytochemical evidence suggested that it

was synthesized in these cells also. As such, it may

represent the nonspecific recognition factor found in

insects. Additionally, Pendland et Al. (1988) detected

similar nonspecific binding of various microbial cells to

plasmatocytes in S. exiqua larvae.

Lectin Specificity and Function

As lectins have become better studied, more

sophisticated methods are being used which argue for their

exquisite specificity. Many immunochemical techniques have

now been applied to the study of lectins and demonstrate

that lectins show some analogous functions with

immunoglobulins. These methods include use of equilibrium

dialysis and Scatchard plots to determine valency and

Michaelis-Menten experiments to determine binding kinetics.

In insects, the Allomvrina dichotoma lectin is being studied

by these methods (Yamashita et al., 1988). Recently, very

elegant work has been performed by Bhattacharyya et Al.

1989, 1990). These investigators found that purified lectin

(Con A or LTA), in the presence of homogenous cross-linked

glycopeptides, will precipitate and can be analyzed by

quantitative precipitin analysis (Kabat and Mayer, 1961).

Lectin, in the presence of heterologous complexes (two

different glycopeptides), fails to precipitate and exist as

soluble complexes. Inhibitory monosaccharides act as

haptens and also exist as soluble complexes. These results

can be analogous to the peptide-specific antibodies which,

when purified, can be analyzed by quantitative precipitation

techniques but exist physiologically as soluble complexes.

If inhibited by haptens, the immunoglobulins also exist as

soluble complexes. Further elegant work has been extended

to view these complexes with the aid of the electron

microscope (EM). Each lectin subunit generally has one

binding site and the valency of the molecule depends on the

number of subunits. A biantennary oligosaccharide has two

binding sites which may or may not be identical. These

antennae may act analogously to the arms of the antibody Fab

chains and have the ability to orient for proper cross-

linking. With identical antennae, ordered arrays of a

single type can be observed. With antennae of heterogenous

composition, a different ordered pattern can be observed.

These EM results corroborate data obtained from kinetic

studies demonstrating that binding kinetics are specific for

each glycopeptide inhibitor. It is also of interest that in

vertebrates, many anti-carbohydrate antibodies exist

constitutively. Frequently, the inhibitory carbohydrate

haptens have been identified as blood group substances which

an individual does not possess on his own tissue. These

antibodies are generally of the IgM subclass and are unique

in that they are not subject to class switching. Anti-

carbohydrate antibodies of the IgG class can be prepared by

injection of carbohydrates with MW above 10,000 daltons.

Although there has been no demonstration of sequence

homology between lectins and antibodies, work discussed

above suggests the presence of domain homology within the

large group of soluble vertebrate galactose lectins.

Since lectins show superficial functional similarity to

antibodies, they have been frequently considered to function

as nonself recognition molecules in insects (Yeaton, 1981).

Controversy continues over what can be considered an immune

response in invertebrates (Klein, 1989; Marchalonis and

Schluter, 1990). Insects are generally considered to lack

immunoglobulins, but there have been reports (Harrelson and

Goodman, 1988; Seeger et al., 1988; Bieber et al., 1989) of

insect proteins which possess adequate sequence homology

with immunoglobulins to place them in the immunoglobulin

superfamily (Williams and Barclay, 1988); these proteins

have been found to play roles in neural cell interaction.

Sun et al., 1990, reported that hemolin (previously

designated P4 by Boman, 1980) is an insect immune protein

which belongs to the immunoglobulin superfamily also. It is

induced by bacterial challenge in H. cecroDia. Since the

bacteria-induced P4 protein found in Manduca sexta shares

the common characteristics of MW, isoelectric point, amino

acid composition and N- terminal amino acid (Ladendorff and

Kanost, 1990) with the H. cecropia P4 protein, perhaps such

proteins are more widespread than previously thought.

Research Aims

Invertebrates are generally considered to lack specific

adaptable humoral immunity, I.g. immunoglobulins and their

cellular synthesizing machinery lymphocytess), but are

nonetheless known to possess an internal defense system

which may include components for carbohydrate recognition

(lectins), nonself recognition, constitutive and inducible

antibacterial proteins such as lysozymes, cecropins,

attacins, hemolin, and P-1,3 glucan inducible components

(prophenoloxidase system) (see Dunn, 1986 for review; Sun et

al., 1990).
In 1985, Pendland and Boucias reported the existence of

a galactose lectin in hemolymph of larval A. gemmatalis.

The overall goal of the proposed research was to further

understanding of the role lectins play in the insect humoral

immune defense system by studying the novel inducible lectin

of A. qemmatalis and exploring how it might function in the

immune defense system of this insect. To this end,

experiments were designed to effect maximum induction of the

lectin and to subsequently develop an appropriate

purification scheme for recovery of the lectin from the

immune hemolymph. Once purified lectin was obtained,

experiments would be set up to explore the sugar specificity

of the lectin. Additionally, experimentation would be

conducted to discover possible biological functions of the

lectin in the insect or, if no function could be

ascertained, then biological functions which could be ruled

out would be determined. As with many systems, the A.

gemmatalis lectin system could not be manipulated as easily

as first anticipated and the results reported in this

dissertation reflect where the system led the investigator

and what was learned about the system during the course of


this research. As suspected by Pendland and Boucias (1985),

the Anticarsia system was more complex than the system found

in Spodoytera exigua. The fundamental difference between

the anticipated and actual situation was demonstration of

the existence of a second lectin or lectin-like molecule

with specificity for L-fucose. The purification scheme

which effected separation of these molecules yielded

partially purified fucose lectin so both of these molecules

were studied in tandem. For studies on biological function,

whole immune hemolymph was used because it was felt that

preliminary work should consider how the intact lectin

system functions in the insect. Two general categories of

biological function studies were designed: experiments to

discover what agents caused induction of the lectin and

experiments to determine whether the galactose-binding

lectin could act as an opsonin.




The occurrence of a galactose/lactose hemagglutinin in

Sarc haqa pererina was reported by Natori st Al. in 1980.

This lectin occurred naturally in pupae but could be induced

in larvae by injury to the body wall. In 1985, Pendland and

Boucias published an account of an inducible galactose

lectin in larval Anticarsia gemmatalis. Titers could be

induced to higher levels by injection of hyphal bodies of

the entomogenous hyphomycete Nomuraea rilevi. Subsequently,

there have been only two additional reports of inducible

insect lectins. In Manduca sexta, one of the antibacterial

proteins, previously named M13 by Hurlbert et ai. (1985) was

found to have carbohydrate binding properties (Minnick e

al., 1986). This lectin has glucose specificity and bimodal

activity, functioning also as a hemocyte coagulant. Its

appearance in the hemolymph can be induced by hemocoelic

injection of Bacillus thurinaiensis (B&..) or oral

administration of a sublethal dose of the B.t. crystal toxin

(Hurlbert et Al., 1985; Rupp and Spence, 1985). Mori et al.

(1989) found a lectin in Bombyx mori which was induced by

cytoplasmic polyhedrosis virus administered per os. Since

diverse agents have been implicated in lectin induction,

experiments were designed to determine if agents other than

H. rileyi were able to induce lectin production in A.


Other studies were undertaken to look at constitutive

lectin titers in 5th instar, 6th instar, wandering, prepupal

and pupal A. cemmatalis. Natori et al. (1980) previously

theorized that the S. peregrina lectin functioned in pupae

in recognition of larval tissue. Bellah et al. (1989)

reported that an age dependent lectin occurred in

Philosamia ricini. This naturally occurring agglutinin was

heat labile in early instars and heat stable in older

larvae. It could also act as a bacterial agglutinin and

this antibacterial activity was resistant to heating at 70 C

in all larval stages studied suggesting a bimodal function

for the molecule. The occurrence of a hemagglutinin,

developmentally regulated by hemolymph ecdysteroids, has

been recently reported to occur in f. mori (Amanai et al.,

1990). The lectin exhibited specificity for glucuronic and

galacturonic acid. This activity did not increase in

response to injury to the body wall and reached maximum

titers prior to larval-larval ecdysis and pupation. These

investigators felt that the hemagglutinin played a role in

post-embryonic development but not in immune defense.


Additional experiments were conducted to determine if

lectin production could be induced by heat shock. Ezekowitz

and Stahl (1988), in a review of vertebrate mannose lectin-

like proteins, reported that these lectins were up-regulated

in response to heat shock and could function as opsonins of

mannose-bearing microorganisms. The mannose-binding

proteins (MBPs) function in cooperation with a cell-bound

mannose receptor which recognizes terminal mannose and

fucose residues. Interestingly, vertebrate MBPs show some

amino acid sequence homology with the 5. perearina lectin.

Materials and Methods

Insects maintenance. Colonies of A. aemmatalis were

maintained in culture at the USDA Insectary in Gainesville,

FL, and the insects were collected as eggs. They were

reared on artificial diet (Greene et al., 1976) and housed

in incubators at 26 C under a photoperiod of 14 hr light and

10 hr dark. For determination of instar, head capsule data

kindly supplied by Dr. G. Wheeler was used.

Experimental desicn--induction of hemagalutinin.

Insects were injected through a proleg with one of the

various agents. Cohorts of 20 insects from the same

treatment group were maintained with diet in paper cups

until they were bled 24 or 48 hr post injection (PI). Ten

individual insects were bled and an aliquot of 10 Al

hemolymph diluted with 70 Al buffered insect saline (BIS)

(Castro et al., 1987). The samples were analyzed for

hemagglutination activity. After bleeding, the cohorts of

20 insects were monitored for mortality.

Inducing agents. The fungi were chosen from cultures

maintained in the Insect Pathology Laboratory. The Candida

albicans strain, isolated from a human patient was provided

by Dr. D. Soll (Iowa State University, Iowa City, IA). The

Bacillus sphaericus strain 2362 was obtained from Mr. P.

Vilarinhos (Brasilia, Brazil). The Escherichia coli (DH-5a)

culture and Anticarsia gemmatalis nuclear polyhedrosis virus

(AgNPV) was supplied by Ms. A. Garcia-Canedo. The AgNPV was

from virus infected cell cultures and the amount of virus

calculated by the following formula: 0.1 x O.D.g2 (1 cm

light path) x dilution = mg/ml (Dr. J. E. Maruniak, personal


Heat shock study. Groups of six insects were

maintained at 30 or 37 C for 24 or 48 hr prior to bleeding.

Hemaaclutination assay (HA). Hemagglutination was

measured against trypsinized rabbit erythrocytes (RBC) and

the titer expressed as the reciprocal of the highest

dilution producing hemagglutination. A more detailed

protocol is presented in Chapter 4 and in Appendix B.


The responses of 6th instar larvae to challenge by

various pathogenic and nonpathogenic microbes are shown in

Table 2-1. Hemolymph samples from individual insects were

titered and a titer of 1024 was arbitrarily chosen as a high

titer since only 25% of untreated insects had this amount of

circulating hemagglutinin. Subsequent to injection of

pathogenic and nonpathogenic fungi, high titers of lectin

appeared in the hemolymph. In all treatments except one (Q.

farinosus 30,000 blastospores, 24 hr), the percentage of

insects responding peaked and decreased by 48 hr PI. This

treatment did show a slight decline from the 24 hr level

when average titer was considered. The decrease in titer

was most dramatic in 48 hr f. bassiana treated insects and

although not shown, the HA profiles from these insects

showed the reappearance of nonspecific ragged

hemagglutination. This phenomenon was absent in all other

insects challenged with fungi. The other agents--bacteria,

virus, injury, saline--failed to induce similarly high

titers of hemagglutinin. It is interesting that injection

of the dipteran pathogen D. sphaericus produced only 35%

mortality in A. aemmatalis.

Mortality data from these treatments are shown in Table

2-2. There was no attempt made to recover B. bassiana from

treated insects but this fungus sporulates readily and can

be isolated easily from cadavers. Results from one

additional experiment are included in this table. Candida

albicans (which is a vertebrate pathogen) was isolated from

only one of the challenged larvae, and was reinjected into a


cohort of insects to determine if passage through a nonhost

would enhance virulence for this nonhost. Passage did not

enhance the virulence of this fungus.

The average titer was determined for three of the

fungal agents and these results are shown in Table 2-3.

Titers are the average titer of ten insects expressed as the

log2 of the reciprocal of the dilution giving complete
agglutination. The most effective agent for agglutinin

induction was H. rilevi with a mean average titer of 212.8

for 30,000 hyphal bodies at 24 hr.

Data on the occurrence of constitutive levels of

hemagglutinin in various life stages are presented in Table

2-4. There is no evidence that the hemagglutinin is

activated upon pupation. In fact, there is a decreased

level of activity and the appearance of a prozone in

hemolymph samples from these insects. Although sham

injected insects showed an average titer of 26.8 compared to

saline injected controls with an average titer of 23, injury

does not seem to induce lectin production.

Data from the heat shock experiment are shown in Table

2-5. The hemagglutinin does not seem to be a heat shock


Table 2-1 Hemagglutinin activity in hemolymph of larval
Anticarsia gemmatalis challenged with various microbial

Treatment Dose Time of % Response
Sampling >8 >1024

C. albicans

D. bassiana

E. farinosus

E. coli

B. sphaericus






1 x 106
1 x 106

1 x 106
1 x 106

0.53 mg/ml



24 hr
48 hr
24 hr
48 hr
9 da

24 hr
48 hr
24 hr
48 hr

24 hr
48 hr
24 hr
48 hr

24 hr
48 hr
24 hr
48 hr

24 hr
48 hr

24 hr
48 hr

24 hr
48 hr

24 hr
48 hr

24 hr
48 hr

24 hr






67 25


Table 2-2. Mortality of
microbial agents.

Treatment Dose



_Q. albicans
(from insect)

f. bassiana

. farinosus








E. coli 1 x 106

B.sphaericus 1 x 106




* 1/73 insects (1.4%); NE

insects challenged with various

%Mortality Time of
Death (da)

100 3
95 3

100 9

53 3
46 3

30 3
30 3

90 3
97.5 3

89 3
91 3

17.5 var.

35 var.

100 3-7

25 var.

10 var.

% Recovery
of Agent










determined; var. variable

)- not

Table 2-3. Average titers of fungal challenged insects.

Treatment Dose Time Average Titer*

H. rilevi 30,000 24 hr 12.6
30,000 48 hr 11.8
60,000 24 hr 11.3
60,000 48 hr 9.9

B. bassiana 30,000 24 hr 8.9
30,000 48 hr 8.5
60,000 24 hr 12.0
60,000 48 hr 3.9

P. farinosus 30,000 24 hr 9.7
30,000 48 hr 8.8
60,000 24 hr 11.9
60,000 48 hr 10.3

* average of titers from 10 insects expressed in log, of
reciprocal of dilution giving complete hemagglutination

Table 2-4. Occurrence of hemagglutinin in life stages of
Anticarsia aemmatalis.

Life stage # Insects % Response Prozone
>8 >1024

Pupae 10 30 0 yes
Prepupae 10 100 30 no
Wandering 10 100 40 no
6th Instar 12 67 25 yes
5th Instar 12 100 83 yes

Table 2-5. Occurrence of hemagglutinin in larvae of A.
gemmatalis after heat shock.

Temperature # Insects Length of % Response
Treatment >8 >1024

30 C 6 24 hr 33 0
6 48 hr 33 33
37 C 6 24 hr 33 0
6 48 hr 100 0


To date, there have been few reports of inducible

lectins in insects (see Table 1-4). Induction by fungi has

only been reported in larval A. gemmatalis (Pendland and

Boucias, 1985). Unlike S. perearina, the hemagglutinin does

not seem to exist constitutively in pupae but average titers

of 210'4 and 29.5 are found in wandering prepupae and

prepupae, respectively. Thus, if production is

developmentally regulated, the appearance of high lectin

titers in hemolymph may occur at a different life stage than

in the dipteran S. peregrina, and the A. gemmatalis lectin

may be more closely related to the hemagglutinin found in B.

mori than to the hemagglutinin found in S. perearina. The

function of the hemagglutinin in prepupae of A. aemmatalis

is unknown. Since both pathogenic and nonpathogenic fungi

can induce production of high titers of the lectin(s), the

induction may be a response to fungal mannans and

galactomannans which are recognized as nonself. The

prophenoloxidase system in insects is generally considered

to recognize another fungal wall component, P-1,3 glucan.

Pendland and Boucias (1985) found that laminarin (P-1,3

glucan) did not induce lectin production. The

prophenoloxidase system in A. gemmatalis is not as active as

the system found in Trichoplusia ni (Boucias and Pendland,

1987), and challenge with H. rileyi hyphal bodies renders it

even less active (personal observation). Once hemocytes are


removed from hemolymph, there is very little tendency for

the hemolymph to melanize, while the surface layers of the

pelleted hemocytes do eventually melanize. Furthermore,

larval A. gemmatalis, unlike certain other insects, do not

produce melanized nodules in response to injection of hyphal

bodies or mycelia of H. rilevi (unpublished observation).




The existence of a novel inducible galactose lectin in

larval Anticarsia gemmatalis was reported by Pendland and

Boucias (1985). Crossabsorption studies using trypsinized

type 0 human erythrocytes and trypsinized rabbit

erythrocytes provided evidence that the lectin was either

multispecific (a functionally heterogeneous molecule) or a

heteroagglutinin (structurally different agglutinin

molecules). In 1986, these investigators reported the

characterization of a galactose-specific hemagglutinin from

larval Spodoptera exioua and were able to purify this lectin

to homogeneity on an Affi-Gel ovalbumin column using an EDTA

elution protocol. Initial attempts to purify the A.

gemmatalis lectin by a similar method demonstrated the

relative complexity of the A. gemmatalis lectin system.

Many methods were attempted to effect purification and these

are discussed in Appendix A. Although purification was not

accomplished by these protocols, many of the methods

provided data on the lectins. This chapter chronicles the

development of the protocol currently used in this

laboratory for purification of the A. gemmatalis lectin(s).

Materials and Methods

Insect maintenance. Colonies of A. aemmatalis were

maintained in culture at the USDA Insectary in Gainesville,

FL. The insects were collected as eggs and reared on

artificial diet (Greene et al., 1976) under a photoperiod of

14 hr light and 10 hr dark at 26 C.

Fungal culture maintenance. Strains of Nomuraea rilevi

and other fungal cultures are stored at -70 C in the Insect

Pathology Laboratory and are maintained on Sabouraud Maltose

Yeast agar (SMY) or Sabouraud Dextrose Yeast broth (SDY).

The strain used in this study was the FL-78 strain which was

originally isolated from field collected A. gemmatalis

larvae (Boucias et al., 1982)

Lectin induction. For injection, fungal cells were

harvested as hyphal bodies (HB) by flooding the Petri plate

with sterile water. Using aseptic technique, the fungal

cells were washed several times in water and suspended in

sterile 0.85% NaC1. After an additional centrifugation, the

cells were resuspended in sterile saline with a vortex

mixer. The cells were counted with a hemacytometer and

diluted to the desired concentration. For preparation of

high lectin titer serum, late sixth instar A. gemmatalis

larvae were each inoculated with 30,000 washed HB in sterile


saline. Injections of 5-10 A1 were made into a proleg with

an ISCO injector (Instrumentation Specialties Co., Lincoln

NE) equipped with a tuberculin syringe fitted to a 30 gauge

needle (Thomas Scientific, Swedesboro, NJ). The insects

were bled at 24 hr postinjection (PI) by puncturing a

proleg. The hemolymph was collected on a sheet of parafilm

placed on an ice bath. Hemolymph was pooled in a prechilled

microcentrifuge tube containing a few crystals of

phenylthiourea (PTU) and centrifuged at 10,000 x g for 5

min. The cell free hemolymph was pooled and stored at -70 C

until processed. An average of 25 ul hemolymph could be

obtained from each larva.

Lectin purification. A novel sequential affinity

purification scheme was developed for purification of the A.

gemmatalis lectin(s). Both a galactose-agarose resin

(Pierce Chemical Co., Rockford, IL) and a fucose-agarose

resin (Sigma Chemical Co, St. Louis, MO) were employed.

Aliquots of 5 ml of pooled immune hemolymph were diluted 1:2

in buffered insect saline (BIS) (Castro et al., 1987) + 0.2

M L-fucose (fucose-BIS) and applied to a packed galactose-

agarose resin which had been equilibrated with fucose-BIS.

Approximately 15 ml of the eluent containing the fucose-

binding proteins (FBPs) were collected and dialyzed

extensively against BIS. The galactose-binding proteins

(GBPs) remaining on the column were washed with BIS + 0.5 M

NaCl (salt-BIS), fucose-BIS and the GBPs eluted with 0.4 M


galactose in BIS (galactose-BIS). Elution was monitored by

increase in absorbance at 280 nm and hemagglutination (HA)

assays. The GBPs were extensively dialyzed against BIS.

The resin was washed with salt-BIS and BIS and equilibrated

with fucose-BIS. The dialyzed GBPs were repurified by a

second passage through the galactose-agarose column

according to the protocol described above.

The dialyzed (FBPs) were applied to a packed fucose-

agarose resin equilibrated with BIS. The protein-rich first

eluent was reserved for other experimentation and the

adhering FBPs were washed with salt-BIS, BIS and finally

eluted with 0.4 M fucose in BIS. Elution was monitored by

increase in absorbance at 280 nm and HA assays. The FBPs

were extensively dialyzed against BIS and repurified

according to the fucose agarose protocol. The resin was

extensively washed with salt-BIS and BIS and equilibrated

prior to each use.

The twice purified GBPs and FBPs were dialyzed against

several changes of 1/10 BIS followed by dialysis against

1/50 BIS. The dialyzed fractions were then lyophilized and

reconstituted with approximately 1 ml sterile deionized


Lectin purification using affinity chromatographv

followed by gel permeation chromatographv. For the first

step, a modification of the above affinity chromatography

procedure was utilized. An aliquot of 2.5 ml of immune


hemolymph was diluted with an equal volume of galactose-BIS

and applied to the fucose-agarose column using galactose-BIS

as buffer. The FBPs were eluted with BIS+ 0.4M fucose. Two

column runs (5 ml immune hemolymph) were pooled, dialyzed

and lyophilized. The first eluents from the fucose-agarose

column were pooled, dialyzed to remove the galactose and

reapplied to a galactose-agarose column. The GBPs were

eluted with BIS + 0.4 M galactose, dialyzed and lyophilized.

Each lyophilized agglutinin was reconstituted with 300 gl of

deionized water. A 200 pl sample was diluted with 200 Al

BIS with 5% glycerol and applied to a 1 cm x 54 cm column of

Sephacryl S-300 gel filtration media (Pharmacia Fine

Chemicals, repackaged by Sigma, St. Louis, MO) in separate

experiments. Fractions of 2 ml were collected and pooled

based on HA activity and absorbance at 280 nm. For

convenience, the fractions were designated from strip chart

measurements as the distance, in cm, from the point of

sample application to the points at which fraction

collection was initiated and terminated. This distance

measurement, although arbitrary, allowed for comparison of

the fractions since the conditions for chromatography were

kept uniform. The chart speed was 2 mm/min and the flow

rate was 12 ml/hr. Molecular weight standards were also run

under the same conditions as the samples but were found not

to be particularly helpful in this study. Under these

conditions, thyroglobulin (MW 669 kd; 330 kd for half unit)

eluted at 4.4 cm, apoferritin (MW 440 kd; 220 kd for half

unit; 18.5 kd for subunit) eluted at 5.1 cm, P-amylase (MW

220 kd) eluted at 5.6 cm, alcohol dehydrogenase (MW 150 kd)

eluted at 6.1 cm, bovine serum albumin (MW 66.2 kd) at 6.4

cm and carbonic anhydrase (MW 31 kd) eluted at 7.3 cm. The

fractions were dialyzed, lyophilized and reconstituted.

Samples of the fractions were subjected to sodium dodecyl

sulfate polyacrylamide gel electrophoresis (SDS-PAGE)

under denaturing conditions and stained with Coomassie

Brilliant Blue. The remainder of each sample was tested for

protein, hemagglutination and hemagglutination inhibition.

Hemaqqlutination assay (HA). Hemagglutination was

measured against trypsinized rabbit erythrocytes (RBC) and

the titer expressed as the reciprocal of the highest

dilution producing hemagglutination. A more detailed

protocol is presented in Chapter 4 and in Appendix B.

Hemaqqlutination Inhibition assay (HI).

Hemagglutination inhibition was measured as the ability of

the sample, serially diluted in either 0.2 M fucose-BIS or

0.2 M galactose-BIS to inhibit HA of the rabbit RBC. A more

detailed protocol is presented in Chapter 4 and in Appendix


Sodium dodecvl Sulfate Dolvacrvlamide gel

electrophoresis (SDS-PAGE). Sodium dodecyl sulfate-

polyacrylamide gel electrophoresis was performed according

to the method of Laemmli (1970) and as outlined in the

Hoefer Protocols (Hoefer Scientific Instruments, San

Francisco, CA). Both standard gels (Protean Apparatus,

BioRad, Richmond, CA) and minigels (BioRad, Richmond, CA)

were used. Molecular weight standards (BioRad) were run on

each gel.

Native qel electrophoresis. A 4-20% gradient gel was

prepared according to a method supplied by Dr. P. Greany.

Molecular weight standards (Pharmacia Fine Chemicals,

Uppsala, Sweden) were run on each gel.

Protein assays. The protein content of various samples

was determined by the method of Bradford (1976) (Pierce

Coomassie Assay) or the BCA method (Pierce BCA Assay) using

a bovine serum albumin (BSA) standard. The reagents were

purchased from Pierce Chem. Co., Rockford IL. During

chromatography, protein was detected by absorbance at 280



The protocol used for purification and partial

purification of the lectin(s) from A. gemmatalis was

developed after extensive experimentation as discussed in

Appendix A. The initial successful experiment involved a

single passage through the fucose agarose column in the

presence of galactose and a single passage of the dialyzed

first eluent through the galactose resin. These

preparations were subjected to gel permeation on a Sephacryl

S-300 column. Figures 3-1 and 3-2 show the absorbance

profile of the FBPs and GBPs eluting from the Sephacryl S-

300 gel permeation column. Fractions were pooled based on

HA titer and absorbance data. Figure 3-3 depicts the

elution profile of the HA activity from each column. The

purpose of this graph is to illustrate the relative position

on the column where the activity eluted. As with the BioGel

1.5A column (Appendix A), molecular weight standards were

not particularly useful for determining the MW of

hemagglutinins because the HA activity did not elute in a

sharp peak so that a native MW could be determined. For the

fucose lectin, the peak activity eluted from 3.9 5.0 cm

and 5.0 cm 6.2 cm. At least some of the activity is in

the MW range of thyroglobulin (669 kd) and the thyroglobulin

half unit (330 kd). When the galactose lectin preparation

was analyzed, the majority of the protein eluted 4.2 7 cm

range but the majority of activity eluted from 8.5 9.7 cm

which was beyond the point of elution of carbonic anhydrase

(MW 31 kd). Because of the difficulty in representing

serial twofold dilutions in such a format, high activity is

not adequately represented on this graph.

Table 3-1 and Table 3-2 show the amount of this

activity which could be inhibited by fucose or galactose,

respectively. When the same fractions were electrophoresed

on an SDS-PAGE gel, the profiles in Figures 3-4 and 3-5 were

obtained. The results from these experiments demonstrated

that there existed a fraction in each preparation which

could be presumed, by a process of elimination, to be the

candidate lectin fraction (Fraction II of the FBPs and

Fraction IV of the GBPs). These putative lectin fractions

were of similar MW (about 45 kd) after SDS-PAGE. The fucose

lectin appeared to consist of isolectins and the diffuse

migration exhibited by the galactose lectin was suggestive

of a glycoprotein. Native gels were run on the same

preparations and the semipure fractions from SDS-PAGE showed

a band in the 400 kd range. Because of inefficiency in

purification and the inability to obtain discrete pure

fractions by gel permeation, it was decided to pursue a

double affinity purification according to the protocol

described above.

Initially the protocol utilized a fucose-agarose

purification followed by galactose-agarose purification

because of the cost of including L-fucose in the buffer.

When a galactose-agarose followed by fucose-agarose protocol

was adopted, much cleaner preparations were obtained and the

consistency in purification of the galactose lectin was


To determine the efficiency of the purification

protocol, three aliquots of 5 ml of immune hemolymph were

used for the analysis. These results are shown in Table 3-

3, Figure 3-6 and 3-7. The banding pattern of FBPs on the

SDS-PAGE gels were not in agreement with previous work.


Figure 3-1. Absorbance profile (280 nm) of galactose
binding proteins from Sephacryl S-300. Titer refers to
hemagglutination titer of rabbit erythrocytes.

c.m. 3.2 3.9 5.0 6.2 8.2 10.3 12.7
Fraction I II II IV V VI
Titer 8 16.384 8192 128 <8 <8

Figure 3-2. Absorbance profile (280 nm) of fucose binding
proteins from Sephacryl S-300. Titer refers to
hemagglutination titer of rabbit erythrocytes.

c.m. 3.6 4.2 7.1 8.5 9.7 10.6
Fraction I II III IV V
Titer 8 2048 2048 32.768 128

HA from
SghWcy S-300

HA fWm

Sgaccyl S-300

Mfraton 5 IV V
ODit m 1 2 3 4 5 6 7 8 9 10 11 12

Figure 3-3. Hemagglutinin activity recovered from pooled
fractions eluted from Sephacryl S-300.

Fhmati I

Table 3-1. Inhibition by galactose and fucose of fractions
of fucose binding proteins eluting from Sephacryl S-300.
Total units (U) were calculated as titer x volume.

Fr. ml Titer Total U HA # U inh. # U inh.
activity by Gal by Fuc

I 6 8 48 0 48
II 12 16,384 196,608 98,304 196,608
III 12 8,192 98,304 0 98,304
IV 20 128 2,560 0 2,560
V 22 <8 0 0 0
VI 24 <8 0 0 0

Table 3-2. Inhibition by galactose and fucose of fractions
of galactose binding proteins eluting from Sephacryl S-300.
Total units (U) were calculated as titer x volume.

Fr. ml Titer Total U HA # U inh. # U inh.
activity by Gal by Fuc

I 6 <8 0 0 0
II 30 2048 61,440 61,440 960
III 14 2048 28,672 28,672 224
IV 12 32,768 393,216 393,216 768
V 8 128 1,024 1.024 0

Figure 3-4. SDS-PAGE of fucose binding proteins eluting in
various fractions from Sephacryl S-300. P is the parental
affinity purified sample applied to the gel permeation
column. Fraction II contains the highest hemagglutination

Figure 3-5. SDS-PAGE of galactose binding proteins eluting
in various fractions from Sephacryl S-300. P is the
parental affinity purified sample applied to the gel
permeation column. Fraction IV contains the highest
hemagglutination titer.

97 kd
66 kd

4 -5 kd

32w kd

Aw11 kd

Table 3-3. Characteristics of fractions from A. gemmatalis
lectin purification. Specific activity was calculated as
titer/mg protein. Units were calculated as titer x volume.

Fraction Volume Protein Titer Specific #
mg/ml Activity units
whole 3 x 5 ml 35.05 8192 233.4 122880
FE (gal 3 x 30 ml 9.48 2048 216 184320
ix gal 45.5 ml 0.0042 128 30476 5824

2x gal 12 ml 0.0017 256 150588 3072
Ix fuc 18 ml 0.0419 2048 48878 36864
2x fuc 15 ml 0.0107 256 22756 3840
3x fuc 10.5 ml 0.0112 512 45714 5376

Table 3-4. Characteristics of reconstituted lyophilized
preparations of lectin from A gemmatalis. Units were
calculated as titer x volume.

Fraction Reconstituted Units

2X gal FE 2 ml 4096
2X gal salt cut 1 ml 128
2X gal lectin 0.5 ml 16384
3X fuc FE 0.5 ml 128
3X fuc lectin 0.5 ml 8192

Figure 3-6. SDS-PAGE of fractions obtained during the
purification of fucose binding proteins by double affinity
chromatography. P is the parental hemolymph; FE is the
first eluent; S is the NaCl cut; IX, 2X and 3X are the
purified fucose lectin fractions.

Figure 3-7. SDS-PAGE of fractions obtained during the
purification of galactose binding proteins by double
affinity chromatography. P is the parental hemolymph; FE is
the first eluent; S is the NaCI cut; 1X and 2x are the
purified galactose lectin fractions.

Figure 3-8. SDS-PAGE of 2X purified galactose and fucose
lectins purified by double affinity chromatography.

Figure 3-8 shows the appearance of purified fractions used

for antibody production and demonstrates inconsistency in

purification, especially for the FBPs. As determined from

assays of dialyzed fractions, the efficiency of purification

was 6% as determined by addition of units from 2X gal lectin

+ 3X fucose lectin divided by apparent units in whole

hemolymph. However, as shown in Table 3-4, when the

purified fractions were lyophilized and reconstituted, there

was an increase in the number of units recovered and the

efficiency was 20% if the total number of units in whole

hemolymph was used for the calculation. If the activity in

the first eluent was used for the calculation, there was 13%

recovery of activity. If all the activity recovered was

considered, 23.5% of the activity was recovered in some


This discussion will consider data presented in this

chapter and in Appendix A. The information obtained from

column chromatography (affinity chromatography and gel

permeation chromatography) yielded insights into the A.

gemmatalis lectin system. All of the gel permeation studies

using BioGel 1.5A failed to effect separation of a lectin

fraction and it was felt that the galactose-binding lectin

might be adhering to the galactose-based matrix (agarose)

thus causing elution over broad MW ranges. These

interactions were not overcome under any of the buffer

conditions tested (presence of galactose in the buffer, high

salt, EDTA, etc.) and the results from these experiments

could be analyzed only after data from the double affinity

purifications was obtained. The ability of the lectin to

migrate successfully in polyacrylamide gels suggested that

it did not stick to this resin. When the BioGel P 300 resin

was used, the activity eluted in the void volume as a

cohesive, impure peak. BioGel P has a nominal exclusion MW

of 300 kd. Preliminary results from native gel

electrophoresis had demonstrated that the native molecule of

both GBPs and FBPs had an apparent MW in the 360 kd and 720

kd range (data not shown). This would indicate that the

native molecule is at least an octamer and perhaps

aggregates into a 16-mer or simply a large aggregate with no

particular subunit composition. Finally, Sephacryl P-300

was chosen for use because it was polyacrylamide based and

could allow for separation of molecules over a broad MW

range (10 kd 1,500 kd). Evidence suggested that the

addition of galactose to the buffer might inhibit adherence

of GBPs to the matrix and to other hemolymph proteins. This

column provided critical information into the nature of the

A. aemmatalis lectin system but since hemagglutinin activity

did not elute in a narrow MW range on this gel and the

column did not effect efficient purification of either

lectin, a double affinity purification was the method of


As reported in Appendix A, small amounts of lectin can

also be purified from normal hemolymph and with a mannose-

agarose column. This supports the likelihood that one of

the lectin components, probably the fucose lectin is a

constitutive factor in the hemolymph. Dot blots (data not

shown) probed with peroxidase labelled lectins have provided

evidence that both 2X purified lectins bind to Con A and not

to peanut agglutinin or wheat germ agglutinin indicating

that the molecules are glycosylated and have glucose and/or

mannose residues. The fucose lectin, with its weak mannose-

binding capabilities may bind to the galactose lectin as

well as to the mannosylated arylphorin. If this occurs, it

would explain the difficulty in separating the component

lectins from each other and from arylphorin. Arylphorin has

been found to be the dominant serum protein in late instar

lepidopteran larvae (Telfer et al., 1983; Haunerland and

Bowers, 1986). Nonspecific aggregations of these molecules

would also explain the inconsistent results obtained for

purification protocols because seemingly minor

inconsistencies in purification procedures, e.g. length of

salt wash, could subsequently affect the purity of the

molecules which eluted in the presence of the inhibiting


The purification efficiency data are of interest in

that they corroborate other existing data which show that

the galactose lectin has high specific activity. The fucose

lectin does not have similarly high specific activity.

During the course of purification, there was an increase in

specific activity of 646 fold for the galactose lectin as

compared to 195 fold for the fucose lectin. The yield of

lectin is not good and most likely does not truly reflect

the amount of total hemagglutinin (especially FBPs) present

in the hemolymph. Some of the active molecules may elute in

the salt wash and, although not present in sufficient

quantity to give a detectable titer, are nonetheless

nonspecifically stripped away from the galactose lectin.

This could explain the dramatic loss in fucose activity

during the course of purification.

Results shown in Tables 3-2 and 3-3 point out the

fallacy of relying on HA titers for measuring activity.

After lyophilization and reconstitution, there was a

dramatic increase in activity most likely because the

molecules had undergone self aggregation. If 184,320 units

represent the total amount of fucose lectin, then 20% is

recovered with one passage through fucose agarose, but less

than 3% remains after 3X purification.

The amenability of the galactose lectin to

purification is much greater than the fucose lectin.

Although silver staining has not been performed, the

galactose lectin can apparently be purified to homogeneity

or near homogeneity while the fucose lectin shows the

presence of multiple bands even after 3X purification. As

shown in Figure 3-6, some of the fucose lectin may exist as

22.5 kd subunits. However, a previous experiment in which

2X purified lectin preparations had been boiled in lysis

buffer for up to 30 min failed to detect these low MW

subunits and they may represent degradation products of the


In the hemolymph, the galactose lectin may exist bound

to other hemolymph components which render it inactive (_.

g. (Fraction II of galactose lectin as shown in Figure 3-3).

This is evidenced in Figure 3-3 by comparing fractions II

and IV. Whether results from minigels and large gels are

completely compatible is also uncertain. Although fraction

II (Figure 3-3) contains more protein, fraction IV has 16

times greater activity and apparently the barely detectable

amount of lectin in fraction III represents the same amount

of activity as in fraction II. However, the total activity

of this preparation was 99.6% inhibitable by fucose in

addition to being totally inhibited by galactose.

Another problem experienced during affinity

purification was that at the absorbance sensitivity used to

detect lectin, galactose also absorbs and frequently the

elution of the galactose lectin was barely detectable. In

addition, the fraction showed no HA activity while bound to

the galactose. Through experience, it was found that the

void volume of 5 ml of packed resin was about 2.5 ml and the

galactose lectin, which could barely be detected by

monitoring UV absorbance and was inactive because it was

bound to the galactose, would elute in approximately 8 ml.



Traditionally, lectins have been partially categorized

based on the monosaccharide or oligosaccharide which

provides most effective hemagglutination inhibition (see

Chapter 1). Previous work by Pendland and Boucias

demonstrated that binding of the Anticarsia gemmatalis

lectin to trypsinized rabbit RBC could be inhibited by

lactose, D-galactose and L-fucose. The agglutination of

human O erythrocytes, however, was inhibited solely by N-

acetyl neuraminic acid. The diluting buffer was phosphate

buffered saline (PBS).

During the course of affinity purification, it became

apparent from hemagglutination inhibition (HI) studies

(using HI method I) that L-fucose was a more potent

inhibitor of hemagglutination (HA) than D-galactose and

experiments were undertaken, initially using immune

hemolymph and later using purified and partially purified

lectin, not only to look at the inhibition profiles to

categorize the lectins but also to try to understand how the

lectins function in this insect. It was postulated that

inhibition studies might explain some of the difficulties

experienced during purification. Additionally, studies were

undertaken to ascertain if either lectin was blood group

specific and thus useful as a diagnostic reagent.

Materials and Methods

Erythrocytes. Rabbit RBC were obtained locally or from

Hazelton Research Products (Denver, PA). Human RBC were

obtained as outdated material from the Blood Bank at the JH

Miller Health Center in Gainesville, FL. Blood cells were

washed several times in PBS, pH 7.2 and usually trypsinized

prior to use according to the method of Novak et al. (1970).

For use, the cells were counted with a hemacytometer and

diluted to the appropriate concentration. For HA, a 2%

solution (3 x 108 cells) was generally used. For inhibition

studies, a 4% solution was usually used.

Hemaqqlutination Assay. For assay, serial twofold

dilutions of hemolymph, lectin or other test material were

made in V-bottom microtiter plates using either PBS or

buffered insect saline (BIS) (Castro et al., 1987) as

diluent. Frequently, it was desirable to conserve hemolymph

or purified lectin and 10 pl of hemolymph were added to 70

Al diluent giving a starting dilution of 1:8. Then 50 Al

were transferred to a well containing 50 Al diluent and

subsequently, serial twofold dilutions made. In the first

well, 30 Al of RBC were added. An equal volume (usually 50

~l) of erythrocytes was added to the following wells and

after 1 hr incubation at room temperature, the plates were

read. The plates were refrigerated and reread after several

hours or overnight. A positive reaction appeared as a

diffuse mat in the bottom of the well and a negative

reaction appeared as a red dot. Positive and negative

controls were included in each group of assays. The titer

was expressed as the reciprocal of the highest dilution

giving complete HA. One peculiarity of the test system was

the presence of incomplete hemagglutination. These results

were carefully and conservatively evaluated and will be

discussed later.

Hemaqqlutination Inhibition (HI) Assay I. Sugars were

usually obtained from Sigma Chemical Co. (St. Louis, MI) and

were of reagent grade. Test sugars were prepared as 200 mM

solutions in either PBS or BIS. Serial twofold dilutions of

hemolymph or lectin were prepared using the sugar solution

as diluent and after 1 hr at room temperature, an equal

volume of a 2% solution of test RBC were added. After an

additional incubation at room temperature for 1 hr, the

plates were read. Plates were reread after several hours or

overnight incubation at 4 C. Positive and negative controls

were included with each group of assays. The titer was

recorded as the reciprocal of the highest dilution giving

complete inhibition.

Hemagalutination Inhibition (HI) Assay II. For this

assay, the titer of the hemolymph, lectin or test substance

was determined and considered one unit j.g. with a titer of

1024, a 1:1024 dilution yields one unit. For HI, four units

were used e.g. a 1:256 dilution. Test monosaccharides were

usually prepared at a concentration of 800 mM in BIS or

occasionally in PBS. Other sugars such as disaccharides,

relatively insoluble sugars or expensive sugar derivatives

were prepared at a lower concentration. To the first well

of the microtiter plate were added 50 Al of test sugar. In

the second well 50 4l were added to 50 pl diluent and serial

twofold dilutions made. To each well were added 50 pl of

solution containing 4 units hemagglutinin. After an

incubation of 1 hr at room temperature, a solution of 4% RBC

was added. The HI assay was read after the 1 hr incubation

and after refrigeration for several hours or overnight.

Positive and negative controls were included with each assay

and consisted of hemagglutinin + RBC, inhibiting sugar +

RBC; diluent + RBC. In calculating the minimum inhibitory

concentration (MIC), the dilution of stock solution allowing

for dilution by hemagglutinin or diluent was recorded as the

MIC. Since four units of added lectin contained adequate

agglutinin, no dilution factor was considered. This was

also true for the indicator system the test RBC.


Optimization of HA test. These results are shown in

Table 4-1. The buffer giving consistently high titers was

BIS and this was generally used for the assays. One to two

per cent rabbit RBC were found to give optimum titers.

Hemaqglutination inhibition studies using A. gemmatalis

immune hemolvmph and variously treated ervthrocvtes.

Results are shown in Table 4-2 and illustrate the

inconsistencies obtained with different RBC. All hemolymph

samples were diluted in BIS. Trypsinized rabbit RBC showed

greater inhibition by L-fucose; trypsinized human 0 RBC

exhibited very high titers with no specificity;

nontrypsinized rabbit RBC showed a markedly decreased titer

and HA could be inhibited completely by galactose, L-fucose,

lactose and moderately by glucose and mannose;

nontrypsinized human 0 RBC were inhibited only by L-fucose.

Since the two dominant inhibitory monosaccharides were

D-galactose and L-fucose, HI Assay II was used to test the

ability of these sugars to inhibit HA of trypsinized rabbit

RBC using immune hemolymph and either PBS or BIS as diluent.

These results are shown in Table 4-3. The agglutinin best

inhibited by fucose seemed to require Ca" ions while the

agglutinin best inhibited by galactose did not seem to

require this divalent cation. These results suggested that

A. gemmatalis immune hemolymph contained two component

lectins which differ in sugar specificity.

Sugar inhibition profiles using HI-I. After twice

purified lectins were obtained, inhibition tests were

performed using 0.2 M galactose or fucose. Results are

shown in Table 4-4. The purified galactose lectin was

inhibited by both galactose and fucose while the partially

purified fucose lectin was inhibited completely by fucose

and slightly by galactose.

Sugar inhibition profiles using HI Assay II. These

results are presented in Table 4-5. The two lectins

exhibited different inhibition profiles. The twice purified

galactose lectin was inhibited by all of the galactose

derivatives and was inhibited best by the synthetic sugar m-

nitrophenyl galactopyranoside. The galactose lectin showed

no apparent anomeric specificity as indicated by equivalent

inhibition by a and P lactose. The twice purified fucose

lectin was inhibited best by L-fucose and the synthetic p-

nitrophenyl fucopyranoside. Since the m-nitrophenyl

galactopyranoside was relatively insoluble, the stock

solution was 50 mM.

Block titrations. To resolve some of the apparent

contradictions in data generated by the two HI assays, block

titrations were set up using whole immune hemolymph, BIS as

diluent, trypsinized RBC and either L-fucose or galactose as

inhibitory sugar. As shown in Table 4-6, in the presence of

L-fucose at concentrations of 400 mM, 200 mM and 100 mM,

there was complete inhibition regardless of hemolymph

Table 4-1. Titers of whole hemolymph
following lectin induction and in the
buffer systems.

at various intervals
presence of various

2%RBC 1%RBC 1%BSA+ 1%RBC 1%BSA+
6 hr PI 256 2048 128 256 64
12 hr PI 512 4096 512 128 64
18 hr PI 2048 4096 1024 128 128
24 hr PI 2048 4096 1024 1024 128
36 hr PI 1024 512 128 64 128
48 hr PI 1024 2048 256 64 128

Table 4-2. Hemagglutination inhibition of erythrocytes by
immune hemolymph of Anticarsia cemmatalis. Results are
given as the titer in the presence of the test sugar. The
numbers in parentheses are the number of wells reduction
represented by this titer.
Trypsinized Rabbit RBC Trypsinized Human O RBC -
Titer: 16,384 Titer: 65,536
+0.2 M galactose 2048 (3) +0.2 M galactose 4096(4)
+0.2 M glucose 4096 (2) +0.2 M glucose 8192(3)
+0.2 M mannose 4096 (2) +0.2 M mannose 8192(3)
+0.2 M lactose 1024 (4) +0.2 M lactose 8192(3)
+0.2 M L-fucose 64 (8) +0.2 M L-fucose 8192(3)

Nontrypsinized Nontrypsinized Human 0
rabbit RBC Titer: 512 RBC -Titer: 512
+0.2 M galactose <8 (>6) +0.2 M galactose 512 (0)
+0.2 M glucose 64 (3) +0.2 M glucose 512 (0)
+0.2 M mannose 128 (2) +0.2 M mannose 512 (0)
+0.2 M lactose 8 (6) +0.2 M lactose 512 (0)
+0.2 M L-fucose 8 (6) +0.2 M L-fucose 64 (3)
+0.4 M L-fucose <8 (>6) +0.4 M L-fucose 64 (3)

Table 4-3. Hemagglutination inhibition of trypsinized rabbit
erythrocytes exhibited by four units agglutinin using PBS or
BIS as buffer.

Inhibition HA titer (PBS) 512 HA titer (BIS) 4096
by (4 units in PBS = (4 units in BIS =
1:128 dilution) 1:1024 dilution)
fucose 50 mM fucose 6.25 mM fucose
(8.2 mg/ml) (1.02 mg/ml)
galactose 3.1 mM galactose 400 mM galactose
(0.56 mg/ml) (72.08 mg/ml)

Table 4-4. Hemagglutination inhibition of trypsinized
rabbit erythrocytes by 0.2 M sugar of twice purified lectins
from Anticarsia gemmatalis.
Inhibition 2X galactose lectin 2X fucose lectin
by Titer: 128,000 Titer:32,000
0.2M <16 4096
0.2M fucose 32 <16

Table 4-5. Sugar inhibition profiles of twice purified
galactose and fucose Anticarsia gemmatalis lectin using HI-

Inhibiting sugar concentration (mM) to inhibit
2x galactose 2x fucose
lectin lectin
D-galactose 0.39 >400
L-fucose 25.00 12.5
D-mannose >400 >100
D-glucose >400 >400
D-fucose 0.78 >400
trehalose >200 >400
a-lactose 0.39 >200

P-lactose 0.39 >200
p-nitrophenyl gal 0.78 100
m-nitrophenyl gal 0.10 >25
p-nitrophenyl fuc >200 25



I --H
o c

*4 P 0


:0 H

V *W
0 1

o O 'o
V 0

H0 *

*-I r 3
V0 0

M44 VA
H-'1 -I +

44.4 r-
94I 0

C t c
a M p a*t

oi -r i
0 +
o 3a toc

1- 0 -H
4- -0 '1

to 00

* rHl '0 -

CA i
0 c a)
E' 10VOX

M + + + + + + + + + I

" + + + + + + + I


S+ + + + + + + I I I I

n( + + + + + I I I I I

c + + + + + +| I I I

t+ + +

S + + + I I I I I

. + l l I I I I I

* I I +1 I I I I I I I I I

c a co w w I % I co Hr Hu H (N V o0 o% cn
S-i (- co H-

.C -"1

0 V V-


al 4


0 o
o m

4-) 4.1V
:3 H t

H 0

r-4 0)'

o4.) -r4

0 40o-.V

C0 to 9:
0-'4 0 0y
-raV) -H1
WQ.) :.0


J14 0+. a)-9
E) 44 %a)

04'4 V
a )Y

to ty t-'
I-rq to

IV 19 04)

EAI -V.40
.QV 00,
E4 0V..

Table 4-8. Hemagglutination titers of nontrypsinized human

Normal HL Immune HL 2X galactose 2X fucose
lectin lectin
Human A 1 64 512 <8 2048
Human A 2 128 512 <8 1024
Human B 64 512 <8 1024
Human 0 128 512 <8 1024

Table 4-9. Hemagglutination titers of

trypsinized human

Normal HL Immune HL 2X gal 2X fuc
Human A 1 2048 32,768 64 32,768
Human A 2 1024 131,072 64 16,384
Human B 1024 4096 64 32,768
Human 0 1024 32,768 64 4096

concentration. Starting at 0.05 mM, the inhibition was

proportional to the dilution. As shown in Table 4-7, when

galactose was used as the inhibitory sugar, there was no

complete inhibition regardless of hemolymph concentration

even at the 400 mM level. In the presence of high hemolymph

concentrations (1:8; 1:16; 1:32), there was complete

inhibition in the presence of sugar concentrations as low as

3.125 mM.

ABO blood group specificity studies. The results from

these assays are shown in Tables 4-8 and 4-9. Normal

hemolymph showed limited ability to agglutinate

nontrypsinized human RBC and although trypsin treatment

increased the HA titer substantially, the erythrocytes were

agglutinated regardless of ABO blood type. Immune hemolymph

showed an increased titer compared to normal hemolymph and

trypsin treatment dramatically increased HA titer.

Typsinized type A 2 cells were agglutinated readily but

other blood types showed high titers. The purified

galactose lectin showed no hemagglutinin activity towards

nontrypsinized human RBC of known ABO specificity and very

slight hemagglutinin activity towards trypsinized human ABO

blood groups. The twice purified fucose lectin agglutinated

human erythrocytes of all types and trypsin treatment

increased HA titer.


This series of experiments provided information on

several aspects of the A. gemmatalis lectin system. Initial

findings demonstrated that the presence of the divalent

cation calcium improved hemagglutination titers (Table 4-1).

When twice purified fucose lectin was obtained and tested,

it was found that the absence of Ca" ions resulted in

complete loss of hemagglutinating activity (data not shown).

The same was not true for the twice purified galactose

lectin. This could explain why Pendland and Boucias (1986)

found no inhibition of human O cells by L-fucose. It is not

known why the concentration of Ca++ in whole hemolymph might

not be adequate for hemagglutination. Unfortunately, the

ion composition of A. gemmatalis hemolymph has yet to be

researched. Cohen and Patana (1982) and Bindokas and Adams

(1987) have determined various ion concentrations in

insects. In fifth instar larvae of the noctuid Spodoptera

exicua, the calcium concentration was found to be 8 mM which

should be adequate for HA since the calcium concentration in

BIS is only 1 mM. Cohen and Patana (1982) also analyzed

starved larvae and were able to show that the calcium

concentration decreased to 2 mM in these insects. Fungi

like Nomuraea rileyi are thought to kill their host through

nutrient depletion (McCoy et al., 1989) and perhaps this

occurs in Anticarsia and the concomitant calcium depletion

renders the endogenous fucose lectin inactive.

The data in Table 4-2 shows the effects of trypsin

treatment on rabbit and human type 0 cells. For both

trypsinized rabbit RBC and nontrypsinized human 0 cells,

fucose was the dominant inhibitory sugar. Nontrypsinized

rabbit RBC showed a great decrease in agglutinability but

were inhibited equally well by galactose, fucose and

lactose. Conversely, trypsinized human O cells exhibited a

greatly elevated titer but were inhibited quite

nonspecifically by all sugars tested. The reasons for this

are unclear as the purified galactose lectin shows limited

ability to agglutinate trypsinized Human 0 cells (Tables 4-8

and 4-9).

As shown in Table 4-3, using HI-II, immune hemolymph in

the presence of BIS was best inhibited by fucose. In the

presence of PBS, and at a much higher concentration,

galactose was the best inhibitor. We have prepared

monoclonal antibodies against the galactose lectin but

initial studies using Enzyme Linked Immunosorbent Assay

(ELISA) (Engvall and Perlmann, 1971; Van Weemen and Schuurs,

1971) have been unable to distinguish between the galactose-

inhibitable and fucose-inhibitable components suggesting

that they possess common epitopes. Thus, the relative

concentrations of the two component lectins in the whole

hemolymph have yet to be determined and future work should

be directed towards quantitatively partitioning the system.

With suitable inhibition protocols and ELISA, these


experiments should be feasible. In general, the assay using

PBS probably reflected the galactose lectin only while the

BIS assay reflected activity of both lectins. These data

can be used to interpret some results shown in Table 4-1.

If titer in PBS can be considered an index of galactose

lectin activity, the galactose lectin reaches peak levels at

24 hr post injection.

It would also be interesting to test normal hemolymph

for the presence of the component lectins by ELISA because

purification of small quantities of lectin from hemolymph of

nonchallenged insects has been effected (see Appendix A) and

normal hemolymph has the ability to agglutinate human

erythrocytes. As alluded to in the materials and methods

section, the HA results from normal serum were extremely

difficult to interpret because of the presence of heavy

agglutination which either formed a prozone and appeared

negative or moderate agglutination that appeared as a

reaction. This occurred using both PBS and BIS and the

addition of 1 mM Mg* to BIS also did not eliminate the

problem (data not shown).

Although partial purification of the component lectins

has been effected, knowledge of the native molecular

structure is lacking. The native lectin appears to be an

octamer, but the subunit composition is unknown. Each

component of the system retains activity after purification

but the biological function of the molecule may require

interaction of the two components. The twice purified

fucose lectin most likely contains some contaminating

galactose lectin and perhaps the galactose lectin contains

residual fucose lectin. It would be helpful if there were

known prototype lectins similar to those in the Anticarsia

gemmatalis lectin system. The isolectins from Bandeiraea

simplicifolia have subunits of similar MW and specificity

galactosee and N-acetyl galactosamine). Superficially, the

Hvalochora cecropia lectin system appears similar to the f.

simDlicifolia lectins but the A subunits exhibit specificity

for galactose and N-acetyl galactosamine while the

specificity of the B subunits is unknown and may not even be

directed against a carbohydrate (Castro et al., 1987).

Although a few lectins have been reported to show

crossreactivity with galactose and fucose, this phenomenon

does not appear to be common but could be explained in our

system if there existed a single lectin with two distinct

binding sites which recognized an oligosaccharide with both

galactose and fucose moieties. Lectins with two distinct

binding sites per subunit have not been described. The

sugar inhibition studies demonstrated that the galactose

lectin was 32 times more sensitive to inhibition by

galactose and galactose-related sugars than was the fucose

lectin to L-fucose. This was also shown by differences in

specific activity as discussed in Chapter 3. The twice

purified lectin was inhibited by m-nitrophenyl galactose >

D-galactose = a-lactose = P-lactose > D-fucose>> L-fucose.

By the criterion of Gallagher (1984), the fucose lectin

would be a borderline candidate for classification as a

monosaccharide inhibitable lectin and perhaps it recognizes

oligosaccharides. Inhibition studies using oligosaccharides

have yet to be conducted. The ability of L-fucose to

inhibit the galactose lectin can be explained in several

ways. The most obvious is that there is indeed cross

reactivity and although less efficient than the sugars

closely related to D-galactose, it can recognize the L-

configuration and bind. Similar results were obtained when

constant concentration of sugar (0.2 M) was used to inhibit

twice purified lectin (see Table 4-3). Results from the

block titration (Table 3-9), however, suggest that such a

potent preparation of lectin titerr 128,000) should require

higher concentrations of sugar to achieve this level of

inhibition. Sigma Chemical Co. purchases their L-fucose

from another source and does not analyze for optical purity

so this sugar could contain from 1-3% D-fucose (personal

conversation with representative from Sigma). This small

percentage of D-fucose could account for results obtained in

Table 4-5 but not necessarily those in Table 4-4. It is

also interesting that there is no apparent anomeric

specificity exhibited by the galactose lectin as both a- and

P-lactose were equally effective in inhibiting agglutination

by the galactose lectin. Evidence shown in Table 4-4

suggested that in the presence of high concentrations of

many sugars, which may simulate locally high sugar

concentrations on microbes or other self glycoproteins or

glycolipids, binding may occur to sugars otherwise nominally

not crossreactive, e.g. mannose.

The discrepancies in results between HI-I and HI-II

were analyzed by block titrations (Tables 4-6 and 4-7) and

show some interesting properties of the lectins. In the

presence of high concentrations of sugar, fucose was the

better inhibitor and as concentration of lectin and sugar

decreased the two components continued to interact and

hemagglutination was inhibited. The galactose lectin, on

the other hand, was sensitive to low concentrations of

galactose but required high concentrations of hemolymph for

inhibition. Again, this may be due to gross differences in

the concentrations of the two component lectins in immune

hemolymph--the fucose lectin being present at much higher

concentrations. At high concentrations of lectin, higher

concentrations of inhibitor are required. As the

concentration of lectin decreases, the concentration of

sugar required for inhibition decreases proportionally.

Conversely, the galactose lectin occurs in low

concentrations but can be inhibited by minute quantities of

sugar. These titrations also point out problems in

interpreting results from HI tests. Using HI method I at a


concentration of 200 mM, fucose appeared to be the dominant

inhibitory sugar compared to 200 mM galactose. Using HI

method II (4 units of hemagglutinin e. g., 512 dilution),

fucose also appeared to be the dominant inhibitory sugar.

It would be instructive to conduct these studies using

equivalent concentrations of the purified lectin components

but sufficient quantities of purified lectin are unavailable

at this time. It would also be interesting to analyze the

lectins by quantitative precipitin and equilibrium dialysis

methods to see if these results reflect differences in the

number of binding sites, binding affinity or binding


How these lectins function in the insect is unknown.

Perhaps the lectins act in concert and the endogenouss"

fucose lectin detects high concentrations of sugar, e.g.

locally high concentrations on surface of fungal cells or

nonself glycoproteins and glycolipids. After induction, the

galactose binding lectin is produced and the system is more

sensitive to minimal concentrations of galactose residues.

This might be analogous to the class switching observed in

vertebrate immune systems where the large IgM molecules are

replaced by smaller, more efficient IgG molecules which

continue to undergo affinity maturation.

A protein of MW 66-70 kd commonly co-purified with the

agglutinins (see Chapter 3). The major hemolymph protein in

late instar larvae of other lepidopterans is arylphorin


(Telfer et al., 1983; Haunerland and Bowers, 1986; Karpells

et al., 1990). In the insects studied (Heliothis zea,

Manduca sexta, Hyalophora cecropia, and Lymantria dispar)

the arylphorins are antigenically similar and glycosylated.

Both the H.zea and H. cecropia proteins contain glucosamine

and mannose in a 1:5 molar ratio. In H. zea, there is 2.5%

carbohydrate present. One of the lectins (most likely the

fucose lectin) in the Anticarsia system could conceivably

interact with the mannose residues on this protein as small

quantities of lectin have been partially purified using a

mannose-agarose affinity column (data not shown). This

remains to be tested in the near future.

Since fucose lectins with blood group specificity are

relatively rare, studies were undertaken to determine if

either of the lectins might be useful as a diagnostic

reagent. As shown in Table 4-8 and 4-9, the galactose

lectin showed little ability to agglutinate any of the blood

types (even the nominally galactose specific Type B) and the

nominal fucose lectin readily agglutinated all blood types.

This difference may also reflect the ability of the fucose

lectin to recognize a species distinct nonterminal

oligosaccharide sequence found on all blood types and the

inability of the galactose lectin to recognize terminal

monosaccharides on any of the human blood types. The lack

of apparent anomeric specificity exhibited by the galactose

lectin may also explain its inability to discriminate a-


lactose from a-lactose. Perhaps the cells were fragile due

to age and trypsin treatment rendered them very susceptible

to agglutination in the presence of the fucose lectin. It

would be valuable to repeat these tests with fresh human

cells to clearly rule out whether either lectin is blood

group specific.

It is interesting that many insect hemolymph lectins

show galactose specificity while the target erythrocytes

used are variable. In the human ABO group, galactose is

considered the inhibitory monosaccharide for blood group B

but, as demonstrated in these experiments, the Anticarsia

galactose lectin does not recognize it. Unfortunately, the

surface carbohydrate composition of rabbit RBC has not been

well-characterized and the target mono- or oligosaccharide

recognized by the Anticarsia lectins remains unknown.

Komano et al.,1980, used sheep RBC as their target cells and

these cells possess a dominant antigen (the

lipopolysaccharide Forssman antigen) which the rabbit RBC do

not possess. Castro et al., 1987, were unable to inhibit

agglutination of rabbit erythrocytes with galactose using a

galactose-specific lectin and perhaps the RBC from the

rabbit were anomalous. The effect of the trypsin treatment

is also unknown but may clear away some peptides which might

interfere with effective cross-linking of lectin molecules.

However, results in Table 4-2 showed that agglutination of

nontrypsinized rabbit RBC could be inhibited by galactose,