Immunological and structural comparisons of the pokeweed antiviral proteins from Phytolacca rigida and derived callus tissue


Material Information

Immunological and structural comparisons of the pokeweed antiviral proteins from Phytolacca rigida and derived callus tissue
Physical Description:
ix, 149 leaves : ill., photos. ; 29 cm.
Ervin, Sean Edward, 1953-
Publication Date:


bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1990.
Includes bibliographical references (leaves 139-148).
Statement of Responsibility:
by Sean Edward Ervin.
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 - 001581913
notis - AHK5831
oclc - 22973989
System ID:

Full Text







During the course of this research I have benefited greatly from discussion
and advice from my committee members E.M.Hoffmann, R. R.Schmidt, P. M.
McGuire and E. Hiebert, to whom I would extend my gratitude. I would like to
express my gratitude to D.M. Dusek and J.D. Rice, the two members of the
laboratory with whom I have been most intimately associated and from whom I
have benefited in conversation and expertise Of course, I must acknowledge
the debt I owe my wife, Krin, who put up with my late nights (and long days) in
the laboratory, my children for providing amusement and perspective, and my
mother, Pat, who helped out in so many ways. Lastly, my thanks go to Jim
Preston for encouraging me in this endeavor, and sharing his abiding love for
the mysteries of Nature, and his thorough 'hands-on' approach to research.


ACKNOWLEDGEMENTS. .............................................................. ii
LIST OF ABBREVIATIONS ............................................................ vi
ABSTRACT ..................................................................................... vii
I. INTRODUCTION .............................................................
Historical Perspective and Comparative Properties
of Ribosome-Inactivating Proteins ......................................... 1
Comparison of Pokeweed Antiviral Proteins and
other Ribosome-Inactivating Proteins ...................................... 6
Other Properties of Phytolacca s ........................................ 11
Ribosome-Inactivating Proteins Associated with
Plant Toxins .............................................................................. 13
Mechanism of Action .............................................................. 13
Biosynthesis of Lectins and RIPs .......................................... 18
Lectins as Analogs to the RIPs ............................................... 22
Objectives of the Present Research ..................................... 23

TO PAP-I, PAP-II AND PAP-S .................................................. 24
Introduction............................................................................ 24
Materials and Methods ......................................................... 27
Preparation of affinity columns ........................................ 27
Preparation of antibody ...................................................... 27
Fractionation of antibody on affinity columns ..................... 28
Preparation of plant extracts ..................................... ......... 31
Enzyme linked immunosorbent assay (ELISA) ................... 31
Electrophoresis and western blot analysis ......................... 32
Protein determination of plant crude extracts .................... 33
In vitro translation assay ........................................ .......... 33

R results ............................................................................................ 34
Specificities and cross-reactivities of antibodies .................. 34
Ability of monospecific antibody to detect the PAP
antigens in crude extracts from leaf or seed tissue ............... 44
Neutralization of translation inhibiting activity of PAP-I,
PAP-II and PAP-S by cross -reactive anti-PAP- .................. 47
Studies on the developmental staging of the PAP
proteins: PAP-I and PAP-II from Phytolacca da ............... 50
Discussion .................................................................................... 62
11. TISSUE CULTURE STUDIES .................................... ....... 67
Introduction ............................................................................. 67
Materials and Methods .............................................................. 68
Chemicals, reagents and media .................................... ..... 68
Preparation of monospecific antibodies .............................. 69
Preparation of antibody to tissue culture proteins ................. 69
Electrophoresis .......................................................................... 70
Western blot analysis (EITB) .................................... ........ 70
Lectin binding to western blots ................................... ....... 71
Enzyme linked immunosorbent assay ................................. 71
Carbohydrate ............................................................................. 72
Peptide sequencing ................................... ......................... 73
Protein synthesis inhibition ...................................... ........ .. 73
CNBr cleavage of proteins ........................................ ........... 73
Callus establishment ................................................................ 74
Fractionation of proteins .......................................... ........... 74
Immunocytochemical studies with callus tissue .................. 76
Results ..................................................................................... 77
Fractionation of PAP forms from extracts of callus
cultures .................................................... ............................. 77
Chemical properties ................................................................ 89
Amino terminal sequence data ..................................... ...... 95
Translation inhibitory activities ................................... ....... 95
Immunocytochemical localization of anti-PAP-I crm
and anti-PAP-Il crm in callus tissue ..................................... 99
SDS-PAGE of CNBr fragmented proteins ......................... 99
Discussion ................................................................................ 111

CULTURE PROTEINS ........................................................... 117
Introduction .................................................. ... 117
Materials and Methods ............................................... 117
Sequence characterization and synthetic peptide
preparation ................................................................. ........... 117
Preparation of peptide-BSA conjugate and antibody .......... 118
ELISA ...................................................................................... 119
E IT B .......................................................................................... 1 19
Results .................................................. 120
Characterization of antibody to peptide conjugate and
purification by affinity chromatography ............................... 120
EITB analysis ......................................... 125
Discussion ..................................................... 130
V. CONCLUSIONS......................................................... .. ... 134
REFERENCES .............................................................. .............. 139
BIOGRAPHICAL SKETCH.......................... ..1.49 149


absorbance at 280 nanometers

absorbance at 405 nanometers

bovine serum albumin


















cross-reactive material

deionized distilled water

western blot

enzyme linked immunosorbent assay


inhibitory concentration -50% value of control level







messenger RNA


pokeweed antiviral protein

phosphate buffered saline

phosphate buffered saline containing 0.3%

PBS:Tw with 0.1% BSA

ribosome-inactivating protein







ribosomal RNA

sodium dodecyl sulfate polyacrylamide gel

Tris buffered saline

vide supra

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



Sean Edward Ervin

May, 1990

Chairman: James F. Preston, III
Major Department: Microbiology and Cell Science
Phvtolacca rigida (common 'Southern' pokeweed) has been
shown to possess three proteins, PAP-I (spring leaf), PAP-II (summer
leaf) and PAP-S (seed), which are potent inactivators of eukaryotic
ribosomal function. These proteins are presumed to be separate
gene products and are developmentally regulated with respect to
their biosynthesis and tissue localization. They are all very basic
proteins with similar pl's, molecular masses, and with significant
amino acid homology (sufficient homology to predict common
evolutionary lineage). As inhibitors of in vitro translation they are
equivalent, acting as N-glycosidases removing the A4324 from the
28S rRNA on rat liver and rabbit reticulocyte ribosomes. These
proteins have collectively been termed ribosome-inactivating-proteins
and are widespread throughout the plant kingdom. This research was
undertaken to prepare antibody probes against the three PAP
proteins for immunological comparisons, to develop a tissue culture
system for the expression of the PAP-proteins and to utilize the
antibody preparations as probes for the PAP proteins in leaf or tissue

culture. Utilizing the antibody probes it was demonstrated that PAP-I
is immunologically cross-reactive with both PAP-II and PAP-S, but
that PAP-II and PAP-S show limited cross-reactivity. The tissue
culture system was found to express a protein, detected with
monospecific anti-PAP-1, which was the equivalent of the PAP-I
protein from spring leaf tissue. Monospecific anti-PAP-II identified
cross-reactive material which was of larger molecular weight,
glycosylated and possessed a unique amino terminal sequence
when compared to the leaf form. Monospecific antibodies to PAP-I
and to PAP-II were utilized in immunocytochemical studies to localize
the proteins in tissue culture. Antibody was developed against a
unique amino terminal sequence associated with the anti-PAP-ll
cross-reactive material and utilized as a probe of crude leaf extracts
for proteins which were immuno-reactive. In developmental studies
PAP-I and PAP-II were shown to be co-synthesized in leaf tissue.
The antibody probe against the unique amino terminal sequence
detected discrete molecular entities which may correspond to

processing precursors for the PAP-II protein.


Historical Perspective and Comparative Properties of Ribosome-Inactivating
Proteins which are toxic to cells (mammalian or otherwise) are found
widely distributed throughout Nature. Those found in sufficient quantity to
study are frequently obtained from plants. Two plants, Ricinus communis
and Abrus precatorius have a long history of medicinal use and it was early

recognized that they contained one or more toxic principles. The toxic
moieties were characterized in the late nineteenth century as proteins.

These two proteins, abrin (from Abrus precatorius) and ricin toxin (from

Ricinus communism) represent the first described toxic proteins from plants
and they have been extensively studied (Olsnes and Pihl, 1976). Lin et al.
(1970) demonstrated that both toxins have the same effect on mammalian
cells in culture, i.e., they are both comparable and potent inhibitors of protein

synthesis. Later investigations established that these proteins are

heterodimers, consisting of an A chain and a B chain. It was found that in

the presence of a reducing agent (e.g. mercaptoethanol) the two chains

could be separated and that the reduced toxin was no longer toxic to intact
cells but had the ability to inhibit in vitro cell-free protein synthesis. When

tested separately, it was demonstrated that only the toxin A chain could

inhibit protein synthesis in a cell-free system and that the B chain was

without effect. The B chain was shown to possess agglutinating

characteristics and served a binding function for the intact toxin. The B chain

acts as a lectin and either free or linked to the A chain by a disulfide bridge
preferentially binds to terminal galactose residues. This activity is
responsible for promoting the association of the holotoxin with cell surface
glycoproteins which may promote endocytosis. The A chain was shown to
possess a catalytic enzymatic activity which inactivated the 60S ribosomal
subunit of eukaryotic ribosomes. This activity is associated with a reduction
both in the binding of elongation factor EF-1 to the ribosomes and EF-2
dependent GTP-ase activity (Fernandez-Puentes and Vazquez, 1977; Obrig
et al., 1973). Because of the carbohydrate binding properties of the B chain
and the toxic effects of the A chain these proteins have been termed
cytotoxic lectins.
Since the original studies on abrin and ricin, a large number of proteins
from plants have been identified which share the ability to inhibit protein
synthesis in vitro and in vivo. Historically, these proteins were initially
termed ribosome-inhibiting proteins (Obrig et al., 1973), until it was
demonstrated that they irreversibly inhibited ribosome function. This
irreversible inhibition is effected through the N-glycosidase cleavage of a

specific adenine residue in 28S rRNA resulting in inactivation of the
ribosome (Endo et al., 1987). Because of their ability to inactivate the
eucaryotic ribosome these proteins have been more appropriately termed
'ribosome-inactivating proteins' or RIPs (Barbieri and Stirpe, 1982). Many

of these proteins are monomeric polypeptides which are functional analogs
of the ricin A chain. In order to distinguish the heterodimeric from the
monomeric proteins, the heterodimers are termed Type 2 RIPs, and the
monomers have been classified as Type 1 RIPs (Stirpe and Barbieri, 1986).
Interestingly, this class of proteins (Type 1 RIPs) is found within a large
number of different plant families and these proteins may be a ubiquitous

feature of all plants (Gasperi-Campani et al.,1985). However, even though
these proteins are found widely distributed throughout the plant kingdom,
there are significant differences in their absolute amounts between families
and orders. In particular, several families within the order Centrospermae
include species which show high levels of these proteins. In large scale
screening studies of crude extracts from plants representing many orders it
was found that proteins (crude extracts) from plants within the
Centrospermae may be 10 to 100-fold more potent than proteins from plants
outside the Centrospermae (Grasso and Shepherd, 1978). Within the
Centrospermae the highest concentrations are found in plants represented
by the families of Caryophyllaceae, and the Phytolaccaceae. Outside the
Centrospermae, two other orders have members which show high levels of
these proteins. In particular, the Cucurbitaceae (cucumber family in the
order Violales) and members of the order Euphorbiales include species
which are rich in these proteins. Ricinus communism (from which ricin toxin
has been isolated) is a member of the Euphorbiales. As more information
becomes available it appears that these proteins are more common to
members of the Caryophyllaceae, Phytolaccaceae and Cucurbitaceae
which contain greater amounts of these proteins than do members of the
Euphorbiaceae. Table 1.1 relates the taxonomic distributions and some
salient biophysical properties of RIPs in those plant species found to contain
significant amounts of monomeric ribosome-inactivating proteins. These

proteins are also represented in the Gramineae (grass family) of which those
from Triticum (wheat), Secale (rye) and Hordeum (barley) have been the
most well characterized (Coleman and Roberts, 1982; Asano et
al.,1984,1986 ). The proteins have been isolated from a variety of structures
including seeds, roots, leaves and latices. However, within an individual

Table 1.1 Properties of characteristic Type-1 ribosome-inactivating proteins


Biophysical Characteristics



Luffi ne


dianthin 30
dianthin 32






























6.3 >9.5

nr nr

1.7 8.6

0 9.4

1.3 >9.0

Table 1.1 (continued)







Biophysical Characteristics











peak 2
peak 3
peak 5









4.5 8.2

40.0 nri

nr nr

nr nr

nr nr

1.4 >9.5

a. Falasca et al., 1982; b. Stirpe et al., 1983; c. Stirpe et al., 1986; d. Ferreras et al., 1989;
e. Kishida et al., 1983; f. Maraganore et al., 1987; g. Casellas et al., 1988;
h. Asano et al., 1984; i. Coleman and Roberts, 1982; j. nr- no report

plant species they are usually found uniquely concentrated within a specific
tissue. As is apparent from Table 1.1, however, these proteins are most
frequently found associated with seed tissue (as is also the case for the Type
2 toxins, abrin and ricin) and/or leaf tissue.

Comparison of Pokeweed Antiviral Proteins and Other Ribosome-
Inactivating Proteins
The ribosome-inactivating proteins from Phytolacca species have been
well studied. In particular, the proteins from Phtolacca americana have
served as a source of study since the original observations of Duggar and
Armstrong (1925) suggested these proteins to possess an antiviral activity.
In this initial study the researchers demonstrated that the sap from P.

americana was able to inhibit viral transfer between tobacco and P.
americana. This ability to inhibit viral transfer was identified prior to the
characterization of the active principle as a protein. Several attempts made
to define the active principle in the plant extracts capable of causing viral
inhibition established the material as a protein (Kassanis and
Kleczkowski,1948) with a minimum molecular weight of 13,000 (Wyatt and
Shepherd, 1969). These observations resulted finally in the purification and
characterization of the active moiety in spring leaf tissue as a basic protein
with the catalytic property of inactivating eukaryotic ribosomes (Obrig et
al.,1973; Irvin, 1975). Subsequently, proteins with similar properties to
those ascribed to PAP-I (spring leaf tissue) were isolated from the summer
leaf tissue (PAP-II, Irvin et al.,1980) and the seeds (PAP-S, Barbieri et
al.,1982) of P. americana. There is reference to a PAP-R (from roots) but the
properties of that protein have not been reported (Cenini et al., 1988). The
PAP proteins isolated from these various tissues share several physical

Table 1.2 Properties of ribosome-inactivating proteins associated with
Phytolacca s.p. and callus tissue.







(&i Q-3)


P. dodecandra

P. riaida







Peak 49
Peak 19
Peak 29












0 >8.1

0 >8.0
0 >8.0
0 >8.0

0 >8.0



a. Irvin et al., 1975
b. Irvin et al., 1980
c. Barbieri et al., 1982
d. Ready et al., 1984
e.Preston and Ervin, 1987
f. Barbieri et al., 1989
g. Ervin and Preston, 1988

and biological properties. They are all highly basic proteins with published
pl values of greater than 8.0, and similar molecular weights of about 30,000.
Most importantly, they are all equally effective as inhibitors of in vitro
translation, indicating that they share the same enzymatic activity. All
members of the Phytolaccaceae so far investigated (to this date restricted to
Phytolacca Sp.) possess a ribosome-inactivating protein analogous to
those described from L americana. Table 1.2 shows the biophysical
properties of those members of the Phytolaccaceae known to possess a RIP,
and for comparison the properties of the callus tissue proteins described in
the present research.
It is obvious from the data compiled in Tables 1.1 and 1.2 that the RIPs
comprise a large family of related proteins. While the overall similarities at
the biophysical level relate these proteins one to the other, the available
amino acid sequence data reveal differences which belie the overall
relatedness of these proteins. To demonstrate this point amino acid
sequence data for selected RIPs are presented in Table 1.3.
Inspection of the sequence data presented in Table 1.3 reveals little
homology in sequence between proteins from different taxonomic families,
while within families one sees good homology. This observed difference in
sequence between proteins which share extensive biophysical similarities
may reflect amino acid changes which do not affect the overall tertiary
structure of the proteins (Montecucchi et al., 1989).
In the case of the ribosome-inactivating proteins, the complete primary
structure (amino acid sequence) was first obtained for ricin A chain (RTA)
(Funatsu et al., 1979; Yoshitake et al., 1978) and the gene encoding a ricin
precursor was isolated, cloned and sequenced by Lamb et al. (1985).

Table 1.3 Amino terminal sequence comparisons between selected
RIPs, including PAP-like proteins from Phytolacca callus tissue.
Toxin Sequence

Common Name


PAP-lb (P.rigid.) data not available
PAP-Sb (P.rigid.) I NIIITFDA
PAP-Ilb (P.rigid.) N2- V F D VQ G A I
PAP-C0 (P.amer.) VN T I I Y NV
Pk4 Callusd(P. rigid.) Y 1I AIY I I: (incomplete sequence)


a. Montecucchi et al., 1989
b. unpublished data obtained from Protein Core Facility, Univ. FI
c. Barbieri et al. 1989
d. Ervin, 1989 (this work)
e. Asano et al., 1986
f. N/D signifies an ambiguous amino acid at that position, the preferred
amino acid is signified first.

Trichosanthin (from Trichosanthes kirilowii. Xuejun and Jiahuai, 1986) and
the barley protein inhibitor (BPI, Asano et al., 1986) from Hordeum vulgare
have recently been completely sequenced as pure proteins. Incomplete
amino terminal sequence data are available for a large number of RIPs,
(both Typel and Type 2) including PAP-I, PAP-II, PAP-S and dodecandrin
(from Phytolacca =a.) and the A chain from modeccin (Adenia digitata,
Ready et al., 1984). Sequence data for saporin 6 (SO-6, from Saponaria
officinalis) have been reported from DNA sequencing of a gene isolated
and cloned from leaf tissue which encodes a seed-like protein (Benatti et al.,
1989). Based on sequence data homology to RTA and known biophysical
and functional similarities, all of the ribosome-inactivating proteins are
hypothesized to have evolved from a common ancestral gene (Ready et al.,
1988). The availability of the gene sequence and the long history of study of
ricin culminated in the crystallographic imaging of that protein at 2.8 A
resolution (Montfort et al.,1987 ). Based on these imaging studies, as well
as site directed mutagenesis studies (Hovde et al., 1988) and amino
terminal sequence data obtained from a number of RIPs revealing five
highly conserved amino acids (Montecucchi et al., 1989), a picture of the
active site of these proteins has begun to emerge (Robertus, 1988).
In the early period of study of these proteins, prior to the detailed
crystallographic data now available for RTA, chemical modifications of the
PAP-I protein were performed to gain some understanding of the
requirements for the enzymatic activity of the protein. Such treatments as
titration of -SH groups with dithiothreitol and treatments with
diethylpyrocarbonate or phenyl glyoxal revealed some aspects of the amino
acid structures in the microenvironment of the active site (see Irvin, 1983 for
review). Such studies by chemical modification yield information which is

distinct from that obtained with proteolytic digests. For the single chain RIPs,
in particular those from Phytolacca. there is limited structural data based
on chemical or enzymatic proteolysis.
Trypsin has long been a favorite tool for structural analysis as its mode of
action is well defined and tryptic maps of peptides obtained from enzymatic
digests are frequently quite repeatable. Historically it had been observed
that the RIPs (both Type 1 and Type 2) are resistant to enzymatic proteolysis.
In particular, it was reported that intact ricin or abrin (and by extension, other
RIPs) are insensitive to treatment with proteolytic enzymes. However, the
isolated A or B chains are susceptible, with the A chain being more
sensitive (Olsnes and Pihl, 1982). Stirpe et al. (1983, 1986) were unable to
affect the enzymatic activity of bryodin or SO-6 by incubation with trypsin,
chymotrypsin or subtilisin at a 0.01 molar ratio of enzyme to RIP. The SO-6
seed protein from Saponaria. however, was cleaved with clostripain to give
small peptide fragments suitable for sequencing from which oligonucleotide
probes were synthesized (Benatti et al, 1989). From the preparation of tryptic
maps, it was demonstrated that ricin A chain and the related agglutinin A
chain were closely related (Olsnes and Pihl, 1976). From the analysis of

peptide maps prepared from tryptic digests of PAP -I and PAP-II, Irvin et al.
(1980) concluded that few, if any of the peptides could be considered to be
homologous between PAP-I and PAP- II.

Other Properties of Phytolacca spp.
In addition, the Phytolacca sp are a source for the pokeweed mitogens,
which are capable of differentially stimulating DNA synthesis of B and/or T
cells of murine and human origin. These mitogens have found extensive
use in immunological studies (Waxdal et al., 1976). The species native to
North Africa, dodecandra (endod) has been shown to possess a protein

which is nearly equivalent to the PAP-I from E americana (Ready et al.,
1984). This plant also has found utility as a specific molluscide in the control
of schistosomiasis, due to the saponins found in the stem and berries
(Kloos, 1979). The presence of these other metabolites appears to be
unrelated to the presence of the ribosome-inactivating proteins in these
plants. However, these unique chemical properties of Phytolacca prompted
the Korean researchers Misawa et al. (1975) and Woo and Kang (1976) to
develop a tissue culture system to study the production of these compounds.
Due to their catalytic activity and mode of action (v.s.), the toxic proteins
ricin and abrin (Lin et al., 1970) were first considered as potential anti-
tumour agents (Lord, 1987). The identification of analogous biophysical
properties and enzymatic activities of PAP-I and PAP-S with those of ricin A
chain led to the early evaluation of the PAP forms in preparing specific
immunotoxins. The non-specific binding of the Type 2 toxins (abrin, ricin,
modeccin) mediated by the B chain made it difficult to work effectively with
these toxins in vivo. However, the construction of antibody conjugates with
either free A chain from ricin or abrin, or with Type 1 toxins which possessed
only the A chain offered the potential of greater targeting specificity. The
efficacy of such toxin-antibody immuno-conjugates was demonstrated for
PAP-I by Masuho et al. (1982). Much research has been fostered in this

area and it continues to be an area of great activity (Frankel, 1988). In
addition to their application as cell-specific cytotoxic agents in immuno-
conjugates, many of the RIPs have been shown to possess an abortifacient
activity (Yeung et al., 1988). In at least one case (trichosanthin) extracts from
the plant bearing the RIP have been used medicinally as an abortifacient for
hundreds of years (Kuo-Fen, 1982). The activity as an abortifacient has
been shown to be related to the RIP present in the tuber (Maraganore et al.,

1987). The applications in medical sciences of the RIPs are all related to the
unique enzymatic activity of these proteins.
Ribosome-lnactivatina Proteins Associated with Plant Toxins
As mentioned previously, the ability to inactivate eukaryotic ribosomes
was originally found in a group of proteins termed cytotoxic lectins. Due to
the presence of a B chain, these proteins are frequently non-specific in their
cytotoxicity towards mammalian cells. These proteins are exemplified by
ricin which is isolated from the seeds of the castor bean plant (from Ricinus
communis). The most well studied of this class of heterodimeric toxins
includes abrin (from Abrus recatorius) and modeccin (from Adenia SP).
(see Olsnes and Pihl, 1982 for a review).
Ricin toxin is a heterodimeric glycoprotein with an apparent molecular
weight of 65 kD. The two polypeptide chains are bound covalently by a
single disulfide bond. The B chain is specific for terminal galactose residues
and is able to bind to cell surface glycoproteins by virtue of this property.
The bound toxin, A and B chains linked by a disulfide bond is internalized
and by an ill-defined process the undegraded A chain is delivered across
the vesicular membrane and into the cytosol. The A chain, which is
analogous to the monomeric toxins described earlier, is capable of
catalytically inactivating the 60S ribosomal subunit RNA.

Mechanism of Action
The mechanism of action of these toxins was the subject of much
research in the last two decades but was finally resolved by the work of
Endo et al. (1987). These researchers demonstrated that the ricin A chain
deadenylated a specific residue in the 28S ribosomal RNA from rat liver
reticulocytes. The mechanism of action was described as the hydrolysis of

the N-glycosidic bond between the adenine and ribose at residue A4324 in
the ribosomal 28S RNA (Endo and Tsurugi, 1987). The A chain then is an
N-glycosidase with a very specific substrate. It was soon demonstrated that
many of the proteins of this sort, capable of inhibiting in vitro translation,
shared this same mechanism of action (Endo et al., 1988; Stirpe et
al.,1988). Similar effects were seen with a recombinant ricin A chain on
yeast ribosomes (Bradley et al., 1987). It is interesting to note that the
bacterial toxin from Shigella dvsenteriae 1 (Shiga toxin, Hovde et al., 1988)
and that from enterohemmorrhagic Escherichia oli (Vero toxin), share this
same mechanism of inactivation of the 28S rRNA (Endo et al., 1988)
li is now known that the fungal toxin, alpha-sarcin from Aspergillus
giganteus is capable of the specific cleavage of the phosphodiester bond
between the guanosine residue at position 4325 and the adenosine residue
at position 4326 in the 28S rRNA from yeast or wheat germ (Endo and
Wool, 1982). Alpha-sarcin then is a phosphodiesterase whose site of
cleavage is one nucleotide removed from that of the RIPs. Alpha-sarcin
preferentially cleaves the large rRNA from ribosomes of rat, Xenopus, yeast,
Zea chloroplast, mouse mitochondria and human mitochondria and E. ci
(Wool, 1984). In contrast to this broad range of susceptible rRNA's known
for alpha-sarcin, the RIPs show a greater specificity. In general, mammalian
ribosomes are susceptible, as are those from yeast. Wheat germ is an
effective substrate for the PAPs but not apparently for ricin A chain (Preston
and Ervin, 1987 ). Ricin (and by analogy, other RIPs) does not affect
ribosomes from E. cli nor from rat liver mitochondria (Greco et al., 1974). In
addition, ribosomes from Sulfolobus (Archaebacterium) are insensitive to
the action of dianthin-32, gelonin, momordin, PAP-I and ricin A-chain
(Cammarano et al., 1985). Type 1 RIPs are differentially effective against the

isolated ribosomes from the protozoa Acanthamoeba., Tetrahymena,
Leishmania and Trvpanosoma, while Type 2 RIPs are virtually ineffective
(Cenini et al., 1988; Cenini et al., 1987).
While this differential susceptibility of mammalian, protozoan and
bacterial ribosomes might be expected because of the presence of unique
associated ribosomal proteins, different assay conditions or other unknown
factors, the situation with ribosomes from plants is even less clear. The
early observations of Owens et al. (1973) on pokeweed showed that PAP-I
inhibited in vitro translation by wheat germ and cowpea ribosomes but not
with pokeweed ribosomes. Coleman and Roberts (1982) showed that the
RIP from wheat germ does not affect wheat germ ribosomes. These studies
provided the basis for the general hypothesis that RIPs should act only on
the heterologous ribosomes (Stirpe and Barbieri, 1986). This problem was
addressed in a paper by Battelli et al. (1984) in which the response of
ribosomal preparations from a number of RIP producing plants was
evaluated against a number of RIPs, including the homologous RIP. Two
points emerged from this study. 1) The Type 2 toxins (heterodimers) as ricin,
abrin and modeccin are relatively ineffectual inhibitors throughout all
preparations assayed and 2) the Type 1 toxins assayed (MCI, dianthin,
gelonin, PAP-S) are variable in their ability to inhibit heterologous
ribosomes but where assayed show no effect on autologous ribosome
preparations. The available data are summarized in Table 1.4 which shows
the IC-50 (nM) on susceptible ribosomes (in vitro) and indicates (by an
asterisk) those which have been shown to possess an N-glycosidase

.1 6 ._ v ao qr
c o o o C o


o 0





x 4-

-- < 2 EC


c *c 4
0 -I = an o 0 .
" E3 oC *CI 0HH
: oC ( D < < < ca c
m O O a 2 C IQ CL )- F--


6o \i D



o c660

Q. -

'Z-j EO -
c; ;


oT c
6 C

- E


j 0


I *

c c


o 0

0 00
D 0 0




0 0

.0 o

-- oJ






.0 0 -0
'- i- r--
C) 0 CD
0 C) C>

m m m



co lo A



00 0
C") ,











0. -







01 0
CM 00


C -

C /


I -


5o 0
ca M a co

l.e,- 2 2
co Cu

O 0o

aDm a

8. o -
| ax
Co rF

xwo '" 010.2
D o.- 2
C ) C- a)
-. 1 0

0 Q 1e i
Sco 4m

cu~ -M.

Wo 0

c-*Co 7

6 jCO E 2

U) C) CO
M *
0 M ~--






0 C
r "
(U --

While these data may be indicative of a general trend supportive of the
interpretation that homologous ribosomes are not affected by the
homologous RIP, they do not entirely delimit the boundaries of
susceptibility. At the least, it seems clear that mammalian ribosomes are
more sensitive to the action of these toxins than are plant ribosomes, with
the exception of those prepared from wheat germ extract. The basis for this
differential sensitivity is unknown. An individual toxin may show differential
activity towards ribosomes from different sources. Indeed, all the available
data indicate a specificity (mediated by unknown factors) in the interaction
between ribosomes and RIPs. Battelli et al. (1984) suggest that there are
perhaps several distinct populations of ribosomes each differentially
susceptible to the action of the autologous toxin. This would support the
interpretation that these proteins (RIPs) are involved in the developmental
staging and/or regulation of the cell life cycle. However, the cellular protein
synthetic machinery would necessarily have to be immune to the action of
the toxin unless the toxins were synthesized as an inactive precursor, or
sequestered co-translationally away from the cytosol. (v.s.)

Biosynthesis of Lectins and RIPs
Little is known about the biosynthesis of the Type 1 ribosome-
inactivating proteins. More is known about the biosynthesis and targeting of
the Type 2 toxin ricin. Ricin toxin is composed of two distinct N-
glycosylated subunits, an A chain (Mr 32,000) and a B chain (Mr 34,000)

linked by a single disulfide bond. The A and B chains of the toxic lectin are
synthesized in the form of a single precursor polypeptide (Butterworth and
Lord, 1983). During synthesis this precursor is co-translationally modified
on passage into the lumen of the endoplasmic reticulum. These

modifications include the proteolytic cleavage of an amino terminal leader
(signal) sequence and glycosylation. The modified precursors are
internalized in a second membrane bound compartment during transit to the
protein bodies. Within the protein bodies the precursors are proteolytically
cleaved (Harley and Lord, 1985) to release the A and B subunits. This
mechanism of post-translational precursor cleavage is a sequence common
to a number of lectins and plant storage proteins (Chrispeels, 1984;
Spencer, 1984).
The Type 1 RIPs are not encumbered by the presence of a B chain
during their processing. In addition, the majority of the mature proteins show
variable glycosylation. Glycosylation may be an important signal during
processing and targeting of proteins (Marshall, 1972). Such is the case for
the lectin Concanavalin A (Con A). Con A is first synthesized as an N-linked
glycoprotein (Marcus et al., 1984), but the mature protein is not glycosylated.
There are precedents for the removal of a glycosylated peptide to yield a
mature, non-glycosylated protein However, in contrast to the proteolytic

cleavage of the proricin in the protein body to give the two complete peptide
chains, Con A is synthesized from the proteolytic cleavage of a 15 amino
acid linker from residue 237 to residue 1. This processing step is followed
by a transpeptidation in which the cleavage of an additional nonapeptide
from the C terminus occurs simultaneously with the religation between
residues 118 and 119 (Sharon and Lis, 1986). This processing sequence
has so far been demonstrated only for Concanavalin A Three important
points derive from these observations. 1) The surface loop of 15 amino acids
negatively affects the carbohydrate-binding activity of the protein; once the
loop is cleaved all precursor forms bind carbohydrate while the form bearing

the loop cannot bind carbohydrate. The loop is glycosylated and this may

be important for correct folding. 2) The authors note that "Since the
ribosomal-binding proteins of the RER, the ribophorins, are Con-A binding
N-glycoproteins, it is improbable that newly synthesized Con-A would ever
be transported out of that compartment and to the vacuole if it was in active
form." (Bowles and Pappin,1988, p.63) 3) The data imply that there may be
only a short, precise time span during development when the function of
Con-A is required. Point 2 is potentially important to an understanding of the
biosynthesis/targetting of the ribosome-inactivating proteins as, in principle,
cytosolic or membrane bound ribosomal RNA should be susceptible to the
action of these toxins.
The biosynthetic schema presented above may be relevant to an
understanding of the processes operative for the monomeric RIPs. In the
two cases discussed briefly above, the final site of deposition of the proteins
is known to be in the protein bodies within the seed. While reports for the
localization of any of the monomeric RIPs are scanty, PAP-I (spring leaf
tissue) has been localized to the cell wall matrix of the E americana by
ferritin immunocytochemistry (Ready et al., 1986). The PAP-I protein could
be initially deposited during the formation of the phragmoplast which is the
biosynthetic origin of the middle lamella. Alternatively, for PAP-I to be
routed to the cell wall during development would require that the protein
translocate from the cytosol across the plasma membrane and be deposited

as part of the growing wall. This situation is reminiscent of that described for
the ricin toxin although there are some obvious differences. Transport
across the plasma membrane and deposition within the polysaccharide rich
matrix of the extracellular compartment (cell wall matrix) is not the same as
transport across the tonoplast of the vacuolar compartment from which the
protein bodies ultimately derive (Boiler and Wiemkin, 1986 ). Nonetheless,

one should expect the processes which drive the biosynthesis of these
molecules to share those features which are common to known eukaryotic
and prokaryotic targeting strategies, which include the presence of an
amino terminal leader sequence for transport across the membrane of the
endoplasmic reticulum (Saier et al. 1989; von Heijne, 1988). Finally, as a
correllary to any understanding of the targetting/localization of these proteins
is a demonstration of the function that these proteins serve for the plants
which synthesize them. Insight into this question has been dominated by the
ability of these proteins to serve as effective inhibitors of viral propagation by
causing cell death. As discussed earlier, the data are inconclusive as to the
ability of these proteins to inhibit their own ribosomes, a necessary
prerequisite to function as antiviral agents. The case for the heterodimeric
cytotoxic lectins (ricin, abrin, modeccin) as a defense mechanism against
mammalian seed predators is no more convincing but less ambiguous, as
there is no need for these proteins to be toxic to their own ribosome to be
effective. The extracellular localization described for PAP-I is supportive of a
role for that protein as an antiviral agent but is not conclusive evidence.
Without further information documenting the localization of the summer leaf
protein (PAP-II) and the seed protein (PAP-S) it is difficult to ascribe any
function to these proteins, defensive or otherwise.
The pokeweed antiviral proteins, PAP-I, PAP-II and PAP-S comprise an
assemblage of closely related proteins surmised to be transcribed from
distinct genes, whose transcription is apparently related to different
developmental stages of the plant. PAP-I is found in the early spring leaf
tissue and persists into the summer at which time PAP-II levels gradually
increase. PAP-S is never found in leaf tissue and neither PAP-I or PAP-II
has been reported from the seed tissue. The PAP proteins have also been

argued to be immunologically distinct (not cross-reactive with antibody to
PAP-I, but see Chapter 2). In addition, 2-D tryptic maps of the proteins
appear different (Irvin et al., 1980 ), and the amino terminal sequences are
unique (Bjorn et al., 1984). These points taken together argue for these
proteins being unique gene products. This pattern of strict
compartmentation, limited or inconclusive information on immunological
cross-reactivity and distinct biophysical parameters between proteins with
similar biological activity finds itself repeated in the glycoprotein lectins.

Lectins as Analogs to the RIPs

Lectins are abundant in seeds as seed storage proteins (Chrispeels,
1984; Pusztai et al., 1983 ), but it has been demonstrated that the vegetative
tissues of plants which possess lectins may contain glycoproteins with
similarities to the seed storage proteins. Some of these may be
immunologically related to the seed lectin from the same species, but others
may be distinct. Proteins similar to seed lectins have been demonstrated in
tissues from Dolichos Griffonia and Sophora (Family Leguminoseae)
(Etzler, 1985). Lectins whose levels of expression vary seasonally have
been demonstrated in the bark tissues of Sambucus and Robinia (Nsimba-
Lubaki and Peumans, 1986). Recently, a lectin related protein has been
described in leaf tissue which is immunologically related to the seed lectin
from soybean and accumulates with a lowered 'sink demand.' (Spilatro and
In addition, the expression of rice lectin mRNAs is developmentally and
spatially regulated, each lectin transcript exhibiting a distinct pattern of
temporal expression in the developing embryo. "Unlike animal systems, a
possible correlation between multiple mRNA transcripts and tissue-specific

or developmentally regulated genes has not been established in plants"
(Wilkins and Raikhel, 1989). These examples are potential homologs of the
pokeweed antiviral proteins and may serve to guide our thinking about the
expression and localization of these proteins.

Objectives of the Present Research
The research described here was undertaken to fulfill the following
objectives: 1) To clarify the immunological cross- reactivity of the proteins
from Phyolacca by the preparation of affinity purified antibodies against the
three major forms of the protein (PAP-1, PAP-II, and PAP-S) with the goal to
prepare specific antibody probes for these proteins. 2) To develop a tissue
culture system for the expression of these proteins as a tool to examine
biosynthetic processing and as a path to the establishment of a single cell
suspension culture for physiologic studies on antiviral activity of these
proteins and 3) To utilize the antibody probes prepared against these
proteins in ultra-structural studies and as tools to examine the biosynthesis
and structure of these proteins, from green leaf tissue and from the callus
tissue. The following chapters provide the details of these researches.



The pokeweed antiviral proteins are easily isolated from americana
(Irvin, 1975) or from P. rigid (Preston and Ervin, 1987) by extraction of the
homogenized leaf or seed tissues with buffer (sodium phosphate, 5mM, pH
6.2 ) or by the extraction of acetone powders (Preston, unpublished
observation ). When comparing their biophysical properties, these proteins
appear very similar. However, the two proteins from leaf tissue, PAP-I
(spring leaf) or PAP-II (summer leaf) and the seed protein, PAP-S are
sufficiently different in ionic charge to elute differentially from a CM-52 cation
exchange resin. This observation, coupled with their reported pl's, different
molecular weights as determined by SDS polyacrylamide gel
electrophoresis and the reported N-terminal sequences (Bjorn et al., 1984)
indicated that these proteins were unique and different gene products. The
amino terminal sequence data compiled by Houston et al. (1983)
demonstrated 17/28 residues from PAP-I and PAP-S to be homologous.
Subsequent data from Bjorn et al. (1984) demonstrated that PAP-II shared
10/29 residues with PAP-I and 11/27 with PAP-S. These data supported the
interpretation of these proteins as different gene products. Limited
serological data presented in several different reports strengthened the
belief that these proteins were unique and antigenically dissimilar. The

following statements are excerpted from the literature addressing the
relatedness of the PAP proteins and serve to illustrate the confusion
surrounding this point.

The major distinguishing feature of these three proteins is found in their
interaction with antibodies prepared against PAP ( PAP-I). Upon challenge
with anti-PAP antibodies, PAP-II fails to cross-react with the antibody as
analysed by Ouchterlony immunodiffusion and by the failure of the antibody
to neutralize the ability of PAP-II to inhibit cell-free protein synthesis (Irvin et
al.,1980). In contrast, PAP-S has a partial cross reaction with anti-PAP
antibodies as detected by immunodiffusion and furthermore, its protein
synthesis inhibiting ability is neutralized by a five-fold excess of antibody
over that required to neutralize the activity of PAP (Barbieri et al.,1982). from
J.D.Irvin 1983. p 374.

There is no cross reactivity among the three proteins and their respective
antibodies. from S. Ramakrishnan and L.Houston 1984. p 201.

The greater degree of homology between pokeweed antiviral protein and
pokeweed antiviral seed protein as compared to pokeweed antiviral protein
II is consistent with immunological cross reactivity between pokeweed
antiviral protein and pokeweed antiviral seed protein, but not pokeweed
antiviral protein II (Barbieri et al.1982). from M.J.Bjorn et al.,1984. p 161.

Ouchterlony immunodiffusion experiments show that dodecandrin cross-
reacts completely with antibody to pokeweed antiviral protein (PAP-I),
indicating that all antigenic sites are held in common Although clearly
not identical proteins, differing in 5/30 of the N-terminal amino acids. from
M.P.Ready et al.,1984. p 316 -317.

Using a radioimmunoassay which easily detects 100 pg of ricin A chain,
we found that no significant reaction occurred even at concentrations of
PAP(I) greater than lug /assay tube. On the other hand, 50% of the
radiolabelled ricin A chain was prevented from precipitating by a
concentration of 5.3 ng/mL (0.8 ng/assay tube). Furthermore, antibodies
directed against PAP do not inhibit the in vitro action of ricin A chain under
the same conditions in which ricin antibodies completely block polyuridylic
acid translation. Therefore, no common antigenic determinants exist
between PAP and ricin A chain. In addition, we confirmed previous
observations that antibodies generated against PAP and PAP-S do not
cross react, from L.L.Houston et al.,1983. p 9602.

The overall picture to emerge from these data is that PAP-I (from
P.americana) and dodecandrin (from P. dodecandra) are serologically

identical. PAP-S (from E. americana) shares antigenic determinants with
PAP-I but not PAP-II. PAP-II is serologically unique. None of the PAP
proteins cross-react with antibody to the ricin A chain.
It is difficult to reconcile these serological data with the biophysical and
chemical similarities of these proteins. Given the overall amino acid
homologies between the single chain ribosome-inactivating proteins and the
A-chain of the heterodimeric toxins (such as ricin ) (Houston et al., 1983;
Bjorn et al., 1984), their proposed common evolutionary origins (Ready et al.
1988 ) and their similar enzymatic activities (Endo et al.,1988 ), it seemed
reasonable that these proteins should share antigenic determinants in
common, and that these determinants might relate to the active site of these
proteins. We proposed to re-investigate the cross-reactivity of these
proteins utilizing polyclonal antibody prepared against each of the three
pokeweed antiviral proteins and to evaluate systematically their cross-
reactivity by ELISA and by Western blotting techniques. The antibody was
then affinity purified in order to prepare a panel of antibodies which showed
a differential pattern of recognition, from cross-reactive to monospecific. The
specificity of these antibodies when purified by immuno-affinity
chromatography is demonstrated on crude extracts from leaf and seed
tissue. The utility of these antibodies as probes for the expression of the PAP
proteins in developmental studies was evaluated in limited studies which
addressed the question of the developmental staging of the two leaf
proteins, PAP-I and PAP-II. There is no developmental study on the
synthesis of these proteins (and/or their subsequent decline) during the life
cycle of the plant. Such a study would be of utility to gain an understanding
of the significance of these proteins for the plant, as it might provide some
insight into the pressures mediating the expression of these proteins and

would yield information pertinent to the developmentally linked expression
of a set of closely related proteins. In the absence of a suitable nucleic acid
probe to examine the levels of messenger RNA of a specific PAP protein, we
have undertaken to address the question of developmental staging by
probing crude extracts from e rigid with monospecific antibodies directed
against either PAP-I or PAP-II.
In addition the ability of cross-reacting antibody to PAP-I to neutralize
effectively the in vitro inhibition of a cell free translation system by the PAP
proteins was evaluated.

Materials and Methods

Preparation of affinity column
Affinity columns were prepared by the coupling of pokeweed antiviral
protein PAP-I, II or S to Affi-gel 10 (Bio-Rad). Protein was prepared as
described (Preston and Ervin, 1987). The purity of the pokeweed antiviral
proteins utilized for preparation of affinity columns and for antibody
production (see below) was routinely assessed by SDS-PAGE. By this
criterion, these proteins were greater than 90% pure (silver stained and

Coomassie stained gels detected only one major polypeptide). In addition,
amino terminal sequence data was obtained for PAP-II, and the presence of

a single amino terminal sequence is presumptive evidence for the purity of
this preparation. Such an analysis has not been done for the PAP-I protein.
The purified PAP proteins were individually coupled to Affi-gel 10 at a
ratio of 10 mg protein/5-10 mL of gel. The gel was washed with 0.1 M
MOPS, pH 7.0, and unreacted sites were blocked with 1.0 M glycine ethyl
ester for 4 h at room temperature with intermittent shaking. Gel prepared in

this way was then washed 5 times with PBS, 1 time with PBS containing 0.5

M NaCI, 1 time with 0.1 M Glycine-HCI (pH 2.3), and then extensively
washed with PBS to restore the pH. Gels were stored in PBS with 0.02%
NaN3 at 40 C.

Preparation of antibodies

Antibodies were prepared in goats by a conventional methodology.
Goats were injected with a solution containing 1.0 mg/mL of PAP-I, PAP-II or
PAP-S, prepared as described (Preston and Ervin, 1987) in complete
Freunds adjuvant. Antigen was readministered to goats at 10 days by a
similar regime except the protein was delivered in incomplete Freunds
adjuvant and the goats were then bled weekly. Antibody was isolated from
the goat serum by an adaptation of a procedure which utilizes n-octanoic
acid followed by ammonium sulfate fractionation (Steinbuch and Audran,
1969). Serum was titrated to pH 5.0 with 3.0 M acetic acid and then n-
octanoic acid was added dropwise with stirring, to a final ratio of 1 part n-
octanoic acid to 20 parts of serum. The solution was stirred at room
temperature for 30 min and then centrifuged at 16,000 x g for 30 min. The
supernatant solution was poured off through glass wool and an equal
volume of saturated ammonium sulfate was added. The solution was stirred
at room temperature for 30 min and then centrifuged for 30 min at 16,000 x
g. The pellet was washed with a small volume of 50% ammonium sulfate,
centrifuged and resuspended in a measured volume of PBS. The resulting
purified IgG preparation was dialysed against PBS and stored at 40C in
PBS containing 0.02% NaN3.

Fractionation of antibody on affinity columns
Antibody prepared as described was applied to the homologous affinity
column ( i.e., anti-PAP-I to PAP-I, etc.) and allowed to bind at 4C for 30 min.
The column was then washed extensively with PBS containing 0.1 M NaCI,
then PBS containing 0.5 M NaCI, then PBS. The column was eluted with 0.1
M glycine-HCI (pH 2.3) to elute specifically bound antibody and was then
washed extensively with PBS. One milliliter fractions were collected.
Fractions containing acid were neutralized by the addition of a small volume
(0.2 mL) of 1.0 M Tris (pH 10.95) and the absorbance of each fraction was
determined at 280 nm (A280) as a measure of protein. Protein (antibody)

containing fractions which washed through the column were pooled and re-
applied to the column in case the protein load had exceeded the binding
capacity of the column. Protein which was eluted from the affinity matrix at
pH 2.3 was considered to be specific antibody. These tubes were pooled
and dialysed against PBS overnight at 4C. This antibody was then applied
to a heterologous column and the procedure repeated. A simplified
nomenclature was adopted to designate the specific antibody pools
obtained from these trials. The antibody used in the treatment is specified
first, and its affinity for the column indicated is specified (as b+ for bind and
b- for not bind). For example, affinity purified antibody against PAP-I which

bound to the PAP-S column is designated as I b+, S b+. A flow sheet which
illustrates this process is presented in Figure 2.1. All pooled fractions from a
single affinity isolation were evaluated by ELISA for reactivity to the PAP
proteins (see below).

IgG (anti-PAP-1, prepared by octanoic acid and
ammonium sulfate fractionation)

PAP-I Affigel



PAP-II Affigel



PAP-S Affigel





Figure 2.1. Preparation of affinity purified antibodies.
Flow-sheet of the stepwise fractionation of goat antibody to PAP-I by
application to PAP-I, PAP-II and PAP-S affinity columns. Arrows are
representative of the affinity column indicated. Bound (b+) and unbound (b-
) fractions are indicated, preceded by the column by which they were
fractionated; I for PAP-I, II for PAP-II and S for PAP-S. The shorthand
notation is indicated beneath each fraction.



Preparation of plant crude extracts
Crude extracts were obtained from leaf tissue of locally (Gainesville, FL)
growing wild southern pokeweed (E rigida) Leaf tissue was collected from
a population of a minimum of three different plants. Collections were made
monthly, beginning in March and continuing through July (NS samples), or
were collected at four day intervals (BZR samples) beginning in April and
continuing through June. In addition, samples were collected from plants
which sprouted in late November of 1989 and showed young leaf tissue
which would be analogous to tissue from spring leaf, and from plants which
showed all developmental stages including seeds at that time. Ten grams of
tissue were collected and mixed with 40 mL of 5 mM Na-phosphate buffer,
pH 6.5. The tissue and buffer were blended in a Waring blender at high
speed for 60 s, and then filtered through four layers of cheesecloth. The
resulting bright green solution was heated with intermittent stirring to 680 C
and then cooled immediately on an ice bath to 4C. The resulting solution
was filtered through Whatman 2V paper in the cold. The straw colored

filtrate was then centrifuged at 16,000 x g for 30 min in a Sorvall RC-2B
centrifuge. The supernatant solution following centrifugation was saved and
assayed as described by ELISA and by SDS-PAGE and Western blot.

Enzyme linked immuno-sorbent assay (ELISA)

ELISA analysis of antibody was performed in 96 well Immulon
(Dynatech) polystyrene plates. The ELISA was a direct, non-competitive
ELISA in which purified antigens (i.e., PAP proteins) were plated individually

as 100 uL aliquots in amounts ranging from 1 to 1000 ngs per well in 0.5 M
sodium carbonate buffer ( pH 9.6). Antigen was allowed to bind at room

temperature for 1 h or overnight at 40C. The wells were washed with PBS
containing 0.3% Tween-20 (PBS:Tw) three times, 100 uL/well. Antibody
(100 uLwell in PBS:Tw as serial dilutions) was then added to determine
the relative titer. Antibody was bound at room temperature for 1-3 h or
overnight at 40C. The plates were incubated and washed as described.
Alkaline phosphatase linked secondary antibody (rabbit anti-goat) was
applied at 100 uLwell at the optimal dilution and incubated and washed as
described. Color was developed from the enzymatic hydrolysis of p-
nitrophenyl phosphate applied as a 1 mg/mL solution in carbonate buffer
containing 0.5 mM MgCl2, pH 9.1. Color development was monitored with

a 405 nm filter on a Bio-Rad EIA plate reader ( Model 2550).

Electrophoresis and Western blot analysis
Antigens (PAP proteins and standards) were electrophoresed at 18 mA
constant current in 12% SDS-polyacrylamide gels with 4% stacking gels.
The discontinuous buffer system of Laemmli (1970) was used. All buffers
and gel stocks were prepared as directed in the Hoefer Catalog ( Hoefer
Scientific, 1988) using analytical/biochemical grade reagents.
Electrophoresis was stopped when the tracking dye reached the bottom of
the gel, and the gels were placed in transfer buffer (0.19 M glycine, 0.025 M
Tris pH 8.3, 20% v/v methanol) for 15 min. Gels were electroblotted to
nitrocellulose (Schleicher and Schuell, NC-42) for 1 h at 1 amp. The
nitrocellulose was then removed from the blotting apparatus and placed into
blocking buffer ( PBS:Tw containing 5% (w/v) Carnation instant milk powder)
for 1 h. Primary antibody containing solution was then applied in the same
buffer for 1 h at room temperature with gentle shaking. The antibody
solution was removed and the blots were washed 3 x 5 min each in blocking

buffer and finally enzyme linked secondary antibody was applied in
blocking buffer at the appropriate dilution. After incubation for 1 h at room
temperature, the blots were washed as described and then washed for 2 x 5
min each in phosphatase substrate buffer ( 0.1 M Tris-HCI, 0.001M MgCI2,
pH 8.8). Finally, substrate was applied as a solution in phosphatase
substrate buffer containing 0.1 mg/mL nitro blue tetrazolium (NBT) and 0.05
mg/mL 5-bromo-4-chloro-3-indolyl phosphate (BCIP).
Gels which were to be stained with silver for protein visualization were
fixed overnight at room temperature in a solution containing ethanol, acetic
acid, water (4:1:5, v/v). The gels were swollen in deionized distilled water
(ddw), placed into 50% aqueous methanol and then swollen in ddw prior to
staining. Silver staining was performed according to the protocol of Wray et
al. (1981).

Protein determination of olant crude extracts
Protein content of crude extracts was determined utilizing the BCA
reagent (Pierce Chemical Co.). Bovine serum albumin (BSA) was utilized
as a standard.

In vitro translation assay
The assay for the inhibitory effect of the RIPs on in vitro translation
systems was performed utilizing a commercially obtained wheat germ
extract (Amersham Corp.). Brome mosaic virus (Promega) was the source
for mRNA for translation. Tritiated leucine (2,3,4,5 3H-L-Leucine, 115
Ci/mmole, ICN ) was added to the incubation mixture and the incorporation
of the radiolabel was determined by liquid scintillation counting. The
procedure was as follows: 10 uL of wheat germ extract were added to a 12 x

75 mm test tube and 10 uL of the sample to be assayed were added. When
assaying for the ability of antibody to neutralize the inhibitory effect of PAP
proteins on in vitro translation, antibody and PAP samples were
preincubated for 30 min, and then a 10 uL aliquot was added to the wheat
germ extract. The samples were then incubated at room temperature for 15
min and then 10 uL of a mixture containing 19 amino acids (minus leucine),
K+ and Mg2+, brome mosaic virus RNA (1 ug) and tritiated leucine (5 uCi)
were added. The mixture was incubated at room temperature for 60 min. A
100 uL aliquot of pancreatic ribonuclease (50 ug) was added and the
reaction mixture was incubated a further 15 min at 37C. The tubes were
placed in an ice bath at 40C and the protein was precipitated by the addition
of 1 mL of a cold 10% (w/v) solution of trichloroacetic acid and

pyrophosphate (TCAPP) in water. The acid precipitated protein was
collected on a glass fiber filter (Whatman, GF/C) by vacuum filtration. The
filter was washed with TCAPP (250C), then ethanol (95%):ether in the ratio
of 3:1(350-40oC) and finally with anhydrous ether. The filter was air dried for
5 min and then placed in an oven at 600C for 15 min. The filter was cooled
to room temperature and placed in a glass scintillation vial with 5 mL of
toluene based liquid scintillation cocktail and the radioactivity determined by
counting in a Beckman LS spectrophotometer.

Specificities and cross-reactivities of antibodies
Initial screening of goat antibody ( a protein fraction enriched for IgG from
goat serum prepared as described) against any of the three PAP proteins,
PAP-I, II, or S by ELISA is demonstrated in Figure 2.2 (a, b and c). From
these data it is clear that anti-PAP-I, anti-PAP-II or anti-PAP-S IgG

Figure 2.2. Detection by ELISA of cross reactivities between PAP-I, PAP-II
and PAP-S.
PAP-I ( -a-) PAP-11 ( ) and PAP-S ( .- )
were plated at amounts ranging between 3 and 100 nanograms per well in
carbonate buffer as described in the text.

a. ELISA detection of PAP antigens with goat anti-PAP-I IgG (1/1000
dilution) prepared from octanoic acid/ammonium sulfate fractionation of
serum as described in the text.

b.ELISA detection of PAP antigens with goat anti-PAP-Il IgG (1/2500
dilution) prepared from octanoic acid/ammonium sulfate fractionation of
serum as described in the text.

c.ELISA detection of PAP antigens with goat anti-PAP-S IgG (1/500 dilution)
prepared from octanoic acid/ammonium sulfate fractionation of serum as
described in the text.

0 20 40 60 80 100 120

nanograms antigen

nanograms antigen

*. 1 ........... ..
Ii --- .... .. ... ...

0 20 40 60 80 100 120

nanograms antigen




< 0.4-


n n_

preparations show cross-reactivity with the non-homologous PAP proteins.
Anti-PAP-I (Figure 2.2 a) is cross-reactive with PAP-II and PAP-S. Anti-PAP-
II (Figure 2.2 b) cross-reacts with PAP-I and to a limited extent with PAP-S.
Anti-PAP-S (Figure 2.2 c) cross-reacts with PAP-I and very slightly with PAP-
II. When analysed by Western blotting techniques, the cross-reactivities of
these crude IgG preparations is also apparent (Figure 2.3 a, b and c).
Figures 2.3 a, b and c correspond to Western blots probed with anti- PAP-I,
anti-PAP-II and anti-PAP-S, respectively. In this experiment, 1.0, 0.50 and
0.25 ug of purified PAP-I (Lanes 2-4), PAP-II (Lanes 5-7) and PAP-S (Lanes
8-10) were electrophoresed and transferred to nitrocellulose as described.
The Western blots were then probed with antibody as indicated. It is clear
that anti-PAP-I recognizes all PAP proteins after blotting (panel a); anti-PAP-
II recognizes PAP-I and PAP-II but not PAP-S (panel b), and anti-PAP-S
recognizes PAP-S and PAP-I but not PAP-II (panel c). Pre-immune serum
obtained from the host animals prior to injection showed no reactivity.
The preparation of monospecific and cross-reactive antibodies from
each of the anti-PAP-IgG preparations proceeded by affinity chromatography
(see Figure 2.1). Figure 2.4 demonstrates the elution profiles obtained from
the affinity chromatography purification of anti-PAP-I by the stepwise
application of this antibody preparation to a PAP-I affinity matrix (Fig. 2.4 a ),
a PAP-II affinity matrix (Fig. 2.4 b) and a PAP-S ( Figure 2.4 c and d) affinity
matrix. The antibody populations obtained after each chromatographic step
were evaluated by ELISA for reactivity as a measure of the purification after
each step. The ELISA data which correspond to the chromatographic
separation at each step for this antibody are presented in Figure 2.5. This
analysis indicated a population of the antibody was monospecific for PAP-I,
and other populations which were cross-reactive. The affinity purification of

Figure 2.3. Detection of PAP proteins, PAP-I, PAP-II and PAP-S by
western blotting with goat IgG against PAP-I, PAP-II or PAP-S as a
measure of cross-reactivity. Antibodies were prepared as
described in the text from octanoic acid/ammonium sulfate
fractionation of goat serum. The lanes are the same between blots:
lane 1: Molecular weight standards; lanes 2-4: PAP-I at 1.0, 0.5
and 0.25 ug, respectively; lanes 5-7: PAP-II at 1.0, 0.5 and 0.25 ug,
respectively; and lanes 8-10:PAP-S at 1.0, 0.5 and 0.25 ug,

a. western blot probed with goat anti-PAP-I (1/1000 dilution).

b. western blot probed with goat anti-PAP-Il (1/1000 dilution).

c. western blot probed with goat anti-PAP-S (1/1000 dilution).


w aw __ -

50 -

50 -

50 -

30 -

1 2 3 4 5 6 7 8910


I 4"m

*&* ft.# a *

E 0,

L i) E.2~


-O. 0

r 0 03 0
< E iE 0' C 0
..L oo


_~ E, E
C.) Cu

0 IL ci
'= ~~ cu 20 E2
cc 7 (

r~tla V 0
Ca 7 E E f

E 0 3 Cu

C-4 .)E
UCL t5 .Ca, Uo -
Cts =

CL) Cu CO0C 0
<~l a- ctj 0 a

Co Cu x0 .! E >.
(D (D cz M: (D C/)
CU 4- L=

-~U0 0 O
W a.. ) CL C
a-- a
o ca PicL wo m

SuD c C a. .0
r- o 2u)! o.0-0
0 a- w LD 0)
o E >,uc..-2 t
0) "a C a) < CL
-0 0-0 a 6= CL -0 C
m a 0 CD tc ca ca
c )*4-U) -- 0
a) c Ca "Z6 = -a






L-- +
6o a-oE ^
-30? 3
So o


n z
+ -Q I,



0- .o co t

0 6

0O8 V o90 V

Figure 2.5. ELISA analysis of antibody pools obtained by affinity purification
of anti-PAP-I by chromatography on PAP-I Affigel, PAP-II Affigel and PAP-S

Affinity chromatography was performed as described in the text, utilizing
goat anti-PAP-I. The goat IgG was precipitated from serum utilizing n-
octanoic acid and ammonium sulfate Antibody was then dialyzed against
TBS and applied to the affinity colums as described in Figure 2.4. The
individual antibody pools obtained by this fractionation scheme were
analysed by ELISA for reactivity to the heterologous antigens as described.
The ELISA data correspond to the fractions described in Figure 2.4.
Antigens were uniformly plated at 50 ng per well, and antibody was diluted
as indicated in the figure. The symbols shown below correspond to the
ELISA response with the antigen plated as described.

--- PAP-I
---- PAP-II

a. anti PAP-1, b+ ( antibody dilutions as indicated)

b. anti-PAP-I, Ib+, Sb+ (antibody dilutions as indicated)

c. anti-PAP-1, lb+,Sb-, lSb+ (antibody dilutions as indicated)

d. anti-PAP-1, Ib+, Sb-, lib- (antibody dilutions as indicated)
d. anti-PAP-1, lb+, Sb-, lib- (antibody dilutions as indicated)

e. anti-PAP-I, lb+, Sb+, llb+ (antibody dilutions as indicated)

1 2 3 4
10 10 10

1/ Ab dilution









1/ Ab dilution

1/ Ab dilution

1 2 3
10 10 10
1/ Ab dilution






0.0 I .-
102 103
1/Ab dilution

10 3



uC 1



. i, .... ...,

the IgG preparations for PAP-II and PAP-S proteins proceeded in similar
fashion. The properties of the resultant antibodies from representative

separations which figured significantly in the present research are presented
in Table 2.1.

TABLE 2.1 Reactivity of affinity purified antibodies.

REACTIVITY (dil/ug)a

Antibody Preparation PAP-I PAP-II PAP-S

anti-PAP-I 50 6.6 2.5
anti-PAP-l,Sb+ 13.3 0 2.6
anti-PAP-l,Sb- 7.7 1.6 0
anti-PAP-l,Sb-,llb+ 5.3 2.1 0

anti-PAP-S 0.87 0 43.8
anti-PAP-S,lb+ 20 0 20
anti-PAP-S,lb- 0 0 9.0

anti-PAP-ll 0.50 6.8 0
anti-PAP-II,lb+ 0.50 0.50 0.50
anti-PAP-lI,lb-,Sb- 0 44 0

a. Calculated according to the dilution of antibody which gives an A 410 of 0.5 (on
an ELISA reader after 60 min when developed against 10 ng antigen per well
)divided by the amount (micrograms) of antibody used in the assay.

It is evident from this Table that an increase in specific reactivity of the

antibodies is obtained at each chromatographic step. Of interest are the

data pertaining to the preparation of the anti-PAP-Il antibody, and its

reactivity with PAP-S as antigen. Initially, antibody to PAP-II showed no

reactivity by ELISA with PAP-S. However, following passage over the PAP-I

affinity column, all antigens were equally reactive with this antibody. This

Figure 2.6. Detection of PAP proteins (PAP-I, II or S) by monospecific
antibody preparations against PAP-I, PAP-II and PAP-S analysed by
Western blot analysis. Crude extracts of spring leaf tissue, summer leaf
tissue or seed tissue of P. igida were prepared as described in the text and
electrophoresed on 12% polyacrylamide gels as described. Western blots
were prepared as described. The blots are formatted identically.
Lanes 1-3: 10 uL of a 0.1, 0.01 and 0.005 dilution of spring leaf crude extract
in sample buffer, respectively; Lanes 4-6: 25 uL of a 0.25, 0.1 and 0.01
dilution of summer leaf crude extract, respectively; Lanes 7-9: 7.5 uL of a 0,
0.25 and a 0.1 dilution of seed extract, respectively. Lane 10 contained the
purified antigen PAP-I, II and S at 0.5 ug each.

a. monospecific goat anti-PAP-I (1/1250 dilution)

b. monospecific gaot anti-PAP-Il (1/250 dilution)

c. monospecific goat anti-PAP-S (1/250 dilution)

1I I I I I I I I I
1 2 3 4 5 6 7 8 9 10


- 29.5

- 30.5

- 30.0

data suggests the specific enrichment of this antibody population for a small
number of antibodies which are reactive with PAP-S, which reactivity is
shared with PAP-I.

Ability of monospecific antibody to detect PAP antigens in crude extracts of
leaf or seed tissue.
The ability of the monospecific antibodies obtained by affinity
chromatography to detect the PAP antigens in crude extracts from leaf or
seed tissue is demonstrated in Figure 2.6 (a, b, c). Crude leaf (spring leaf
extract obtained in April, summer leaf extract obtained in July ) or seed
extracts (seed tissue obtained in October) were electrophoresed and
Western blots prepared as described. These results indicate that the
monospecific antibody pools obtained by fractionation on affinity columns
show similar specificity in Western blot analyses as demonstrated by ELISA
analyses (Figure 2.5 and Table 2.1). Monospecific anti-PAP-I reacted only
with a single band in crude extracts of spring leaf tissue (panel a, Figure 2.6,
lanes 1-3) or only with purified PAP-I (Lane 10). Monospecific anti-PAP-ll
reacted with some bands at the lowest dilution of crude extract from spring
leaf tissue (Lane 3, panel b, Figure 2.6) but not at the position expected for
PAP-II, but did react with PAP-II at the two lowest dilutions of summer leaf
crude extract ( Lanes 5 and 6), and with purified PAP-II (lane 10).
Monospecific anti-PAP-S reacted with PAP-S in crude seed extracts at all
dilutions (Lanes 7-9, panel c, Figure 2.6), and reacted weakly with PAP-I
(Lane 3) at the lowest dilution analyzed. These data are consistent with the
ELISA analyses and indicate that selected antibodies are specific probes of
PAP-related antigens in leaf tissue of P. rigida.


-00 -00 00000 000 000 0 00
.-o0 o-0 -o0 -C0 0 *oo 0

Figure 2.7. Neutralization of inhibitory activities of PAP-I, PAP-II
and PAP-S by goat anti-PAP-1 IgG. In vitro translation was
determined with wheat germ extract as described in the text. Two
concentrations, 10-11 and 10-10 of each form of PAP were
evaluated at three dilutions of antibody (1/100, 1/1000 and
1/10000). Antibody was preincubated with the PAP protein being
tested as described in the text. The figure shows for each
treatment in the lane marked '0' the inhibitory effect of the PAP
protein alone. The data are presented as percent of the untreated



Neutralization of translation inhibiting activity of PAP-I. PAP-II and PAP-S by
cross-reactive anti-PAP-1.
The ability of cross-reacting antibody (prepared against PAP-I) to
neutralize the inhibitory action in vitro of the PAP proteins is shown in Figure
2.7. This figure demonstrates the ability of anti-PAP-1 (cross-reactive with
PAP-I, II and S) at three dilutions to neutralize two different concentrations of
each of the PAP proteins. The amounts of the PAP proteins were chosen so
as to give approximately 50% inhibition and 90% inhibition in this assay.
This corresponds to a concentration of approximately 10-11 M and 10-10 M,
respectively. Antibodies were pre-incubated with the protein such that the
final relative dilution of the antibody in the reaction mixture was 0.01, 0.001
and 0.0001. The data show that anti-PAP-I cross-reactive antibody
effectively neutralizes the activity of the PAP proteins at low dilutions, which
effect is lost with increasing dilution or in the presence of greater amounts of
protein. In particular, anti-PAP-I at 0.01 and 0.001 dilution is capable of
totally abrogating the ability of PAP-I, PAP-II or PAP-S at 10-11 M to inhibit in
vitro translation. At a dilution of 0.0001, this antibody still relieves the
inhibition by PAP-I by about 35% above the control, but has little effect on

PAP-II or PAP-S. The ability of the cross-reactive antibody to PAP-I to
neutralize the PAP-proteins is much reduced at the 10-fold higher
concentration of protein. PAP-I at 10-10 M is still totally neutralized at the two
lowest dilutions, and there is still about 35% neutralization at the highest
dilution. The PAP-II and PAP-S proteins at 10-10 M are less susceptible to

neutralization by cross-reacting antibody. PAP-II is still totally neutralized at
0.01 dilution, but this neutralizing effect of antibody is totally eliminated at the
next two higher dilutions. In contrast, PAP-S is only about 40% neutralized

at 0.01 dilution of antibody, which effect is rapidly lost with increasing

Studies on the developmental staging of the PAP proteins: PAP-I and PAP-II
from Phytolacca rigida
Figures 2.8 a and 2.8 b demonstrate the appearance of the crude
extracts from NS samples and from BZR samples, respectively, when
visualized by silver staining of SDS-PAGE gels. NS samples were collected
on a monthly basis from March to August, while BZR samples represent 4
day sampling intervals from May to June. There is a gradual increase in the
appearance of the PAP-II protein over the time course of the samples. PAP-I
on the other hand remains relatively constant over this time period with a
slight decrease in staining density at the later time point (Lane 8, August 22).
In contrast, the BZR samples, collected more frequently, do not show this
same apparent increase in PAP-II (Figure 2.8 b, lanes 1-8).
The Western blot analysis of these gels is shown in Figure 2.9 and 2.10.
For NS samples (Figure 2.9 a and b), there is a marked increase in immuno-
reactive species detected by anti-PAP-ll (panel b) over the time course of

these samples while the level of anti-PAP-I reactive species (panel a)
remains fairly constant. The sampling on closer time intervals (BZR
samples, Figure 2.10 a and b) does not show this same clear result. Panel
a, which shows anti-PAP-I reactive species does not demonstrate a clear
increase in PAP-I, nor does panel b demonstrate a clear increase in PAP-II.
Of interest from both of these samples (NS and BZR) is the reactivity of
higher molecular weight components in crude extracts detected by anti-PAP-
II antibody (Figures 2.9 and 2.10, panel b ) and the reaction of this antibody
with a component in crude extracts of molecular weight about 22kD which

Figure 2.8. Silver stained polycrylamide gels of crude leaf extracts from P.
rigida samples collected at variuos time points during the growing season.
Leaf tissue was collected at various time intervals and extracted by
homogenization in buffer and heating as described in the text. Dates of
collection are indicated as month/day for each sampling Samples were
centrifuged and the crude fraction obtained was mixed with an equal volume
of sample buffer. The samples, 25 uL each, were electrophoresed in 12%
gels as described, and stained following the procedure of Wray et al. (1981).

a. NS samples corresponding to dates March through December. Lane 1:
PAP-I, 0.5 ug; Lane 2: PAP-II, 0.5 ug; Lane 3: 3/8; Lane 4: 4/10; Lane 5:
5/2; Lane 6: 6/4; Lane 7: 7/2; Lane 8: 8/22; Lane 9:12/5.

b. BZR samples corresponding to four day interval from 4/10 to 5/12. Lane
1: PAP-I 0.5 ug; Lane 2: PAP-II, 0.5 ug; Lane 3: 4/10; Lane 4: 4/14; Lane 5:
4/18; Lane 6: 4/22; Lane 7: 4/26; Lane 8: 5/2; Lane 9: 5/6; Lane 10: 5/10.






_immm i, ilii, ,.i -a = W

1 2 3 4 5 6 7 8 9 10




Figure 2.9. Western blot analysis of the proteins detected in crude
leaf extracts of rigid prepared from leaf tissue collected at
various time intervals during the growing season. Leaf tissue was
collected from locally growing specimens of P. rigida and
processed by heat precipitation following blending in sodium
phosphate buffer as described in the text. The blots are formatted
identically. Lane 1: 3/8; Lane 2: 4/2; Lane 3: 5/2; Lane 4: 5/17;
Lane 5: 6/4; Lane 6: 7/2; Lane 7: 7/21; Lane 8: 8/22; Lane 9-10:
PAP-I (panel a) or PAP-II (panel b) at 0.5 and 1.0 ug, respectively.
Samples were prepared for electrophoresis by boiling an equal
volume of leaf extract with sample buffer for three minutes. 25 uL
were electrophoresed on 12% polyacrylamide gels and Western
blots were prepared as described in the text.

a. Western blot treated with monospecific goat anti-PAP-I (1/250

b. Western blot treated with monospecific goat anti-PAP-II (1/250

1 2 3 4 56 7 8 9 10


Figure 2.10. Western blot analysis of the proteins detected in
crude leaf extracts of P rigida prepared from leaf tissue collected
at various time intervals during the growing season. Leaf tissue
was collected from locally growing specimens of P. rigida and
processed by heat precipitation following blending in sodium
phosphate buffer as described in the text. The blots are formatted
identically. Dates of collection are indicated as month/day. Lane
1: molecular weight standards; Lane 2: PAP-I, 1.0 ug; Lane 3:
PAP-II, 1.0 ug; Lane 4: blank; Lane 5: 3/24; Lane 6: 4/12; Lane 7:
4/24; Lane 8: 5/6; Lane 9: 5/22; Lane 10: 6/18. Samples were
prepared for electrophoresis by boiling an equal volume of leaf
extract with sample buffer for three minutes. 25 uL were
electrophoresed on 12% polyacrylamide gels and Western blots
were prepared as described in the text.

a. Western blot treated with monospecific goat anti-PAP-I (1/250

b. Western blot treated with monospecific goat anti-PAP-lI (1/250

123 4 5 6 7 8 9 10




migrates faster than PAP-I or PAP-II. These are not detected by anti-PAP-1
(Figures 2.9 and 2.10, panel a)
Finally, Figure 2.11 demonstrates the results of probing young leaf tissue
from plants which emerged in late October with anti-PAP-I or anti-PAP-Il.
Also included in this Figure are samples from the NS or BZR series
(presented above) to serve as controls. For example, PAP-I is present in all
leaf extracts from NS or BZR samples (Figures 2.9 and 2.10 a), while PAP-II
is absent from the earliest time points in the NS series ( lane 1, Figure 2.9,
panel b.) Two time points were chosen from these samples as 'controls'.
Figure 2.11, corresponds to a Western blot probed with monospecific anti-
PAP-I (panel a) and monospecific anti-PAP-ll (panel b). It is clear from this
figure that PAP-I and PAP-II occur simultaneously in this young leaf tissue
(see lanes 5 and 6, Figure 2.11 a and b ), whereas in control lanes ( lanes 1
and 3) PAP-II is not detectable in the earliest spring leaf tissue.
ELISA analyses corresponding to these data are presented in Figure
2.12. ELISA analysis was performed as described on a 100 uL aliquot of a
0.02 dilution of crude extract, and the data were corrected for protein
concentration (O.D.405/ug protein assayed). Inspection of the data reveals
variability in the ELISA reaction between time points within a collection.
However, the overall trends inferred from the Western blots presented
previously (Figures 2.9 and 2.12) are supported by these data.

As discussed previously, the cross-reactivity of the PAP antigens towards
immune sera against one of the PAP forms (PAP-I) was not well defined.
Utilizing the double diffusion technique in limited studies, it was determined
that the PAP antigens were not cross-reactive ( Irvin, 1975). The
experiments described here were designed to address the question of the

Figure 2.11. Western blot analysis of the proteins detected in
crude leaf extracts of P. riida prepared from leaf tissue collected
at various time intervals during the growing season. In particular,
samples were collected from plants which showed young(newly
emergent) leaf tissue in late October, 1989. Tissue was collected
from locally growing specimens of rigida and processed by heat
precipitation following blending in sodium phosphate buffer as
described in the text. The blots are formatted identically.
Lane 1: 3/8/87 (NS); Lane 2:3/24/89 (BZR), ; Lane 3: 8/22/87
(NS); Lane 4: 5/14/89 (BZR); Lane 5: young leaf tissue 10/27/89;
Lane 6: young leaf tissue, second sample 10/27/89; Lane 7: leaf
tissue from plant showing all developmental stages on 10/27/89;
Lane 8: leaf tissue from plant showing dry berries only 10/27/89;
Lanes 9-10: PAP-I (panel a) or PAP-II (panel b) at 1 or 0.5 ug
respectively. Samples were prepared by boiling an equal volume
of leaf extract with sample buffer for three minutes. 25 uL were
electrophoresed on 12% polyacrylamide gels and Western blots
were prepared as described in the text.

a. Western blot treated with monospecific goat anti-PAP-I (1/250

b. Western blot treated with monospecific goat anti-PAP-ll (1/250


t~m. l 7- ^C



12345 678910




1 2 3 4

6 7 8910




3/8 4/10 5/2 5/17 6/4 7/2 8/22
date of sample collection



< 0.1.


3/8 4/10 5/2 5/17 6/4 7/2 8/22
date of sample collection

Figure 2.12. ELISA reactivity of crude extracts of leaf tissue
with monospecific anti-PAP-I and monospecific anti-PAP-Il.

ELISA analysis of crude extracts from leaf tissue was
performed as described in the text. 100 uL aliquots of 0.02
dilutions of crude extract obtained from heat precipitation of
ground leaf tissue was plated in carbonate buffer, pH 9.6.
Antibody was supplied as 1: 100 fold dilutions in PBS:Tw,
and secondary antibody was alkaline phosphatase
conjugated. NS samples (a and b) and BZR samples (c
and d) correspond to the same fractions assayed by
western blotting (Figures 2.9 and 2.10) and in the SDS-
PAGE of Figure 2.8 a and b, respectively.
a. NS samples probed with anti-PAP-I (1:100 dilution)
b. NS samples probed with anti-PAP-II (1:100 dilution)
c. BZR samples probed with anti-PAP-I (1:100 dilution)
d. BZR samples probed with anti-PAP-II (1:100 dilution)





0.i. I I
3/17 3/24 4/1 4/12 4/194/24 5/6 5/22 6/1 6/18
date of sample collection

2 d

< 0

3/24 4/12 4/24 5/22 6/18
3/17 4/1 4/19 5/6 6/1
date of sample collection

Figure 2.12 (continued)

cross-reactivity of the PAP antigens and to prepare monospecific antibodies
as probes for the expression of a given PAP antigen in leaf, seed or callus
tissue. In addition, it was envisioned that these probes would serve as
useful tools for isolating PAP related proteins ( perhaps of developmental
significance) and would be of utility in immunocytochemical studies. It is

clear from the data presented above that both cross-reactive and
monospecific antibodies may be isolated from immune goat sera. While the
argument may be made that these animals are omnivores and may have
been exposed to pokeweed as a forage food, thus provoking the cross-
reactivity observed, two observations mitigate against this possibility. 1) Pre-
immune sera show no reactivity at the dilutions where the immune sera are
reactive, and 2) PAP-S as antigen shows very limited cross-reactivity with
PAP-II, and PAP-II as antigen shows no reactivity with PAP-S. This result
would not be expected if the goats were feeding freely on pokeweed and

supports previous serological observations obtained with studies utilizing
rabbit antibody. This same pattern of reactivity has been observed in our
laboratory utilizing antibody prepared in mouse (Preston and Ervin, 1987),
chickens and rabbits ( Preston, unpublished observations). The cross-
reactivity observed may depend upon the sensitivity of the assay system
utilized in the present study. Typically, 10 ng are easily detected by the
current ELISA system. If one measures reactivity by ELISA and then
standardizes the observed response (color development) by factoring in the
amount of antibody protein in the reaction, one can arrive at a measure of
cross-reactivity. Such an analysis was performed by Preston and Ervin
(1987). They were able to demonstrate that anti-PAP-I had a reactivity of 1
with PAP-I, 0.46 with PAP-II and 0.11 with PAP-S. Ricin A chain was of
limited reactivity (0.04). The data presented in Table 2.1 are in agreement

with those data reported previously. In addition, the ability of cross-reacting
antibodies to neutralize the enzymatic activity of the PAP proteins (Figure
2.7) argues for the presence of shared epitopes. Those epitopes which are
shared may be near the active site, although an influence on activity due to
conformational changes in the PAP protein on binding antibody cannot be
ruled out. Antibody alone at the lowest dilution used here had no effect on in
vitro translation when supplied to the reaction mixture.
From the data presented here, it is clear that the PAP proteins may be
effectively immobilized on Affi-gel. This result would be anticipated due to
the highly basic pl values reported for these proteins. In addition, coupling

via iso-amide linkage does not affect the ability of the PAP proteins to bind
antibody. This is crucial for the successful utility of these columns as
reagents for purifying monospecific antibody.
The major points to emerge from these results may be summarized as
1). Purified IgG or crude antisera show cross-reactivity. PAP-I is cross-
reactive with all PAP species. PAP-II and PAP-S are negligibly cross-
reactive, but each is cross- reactive with anti-PAP-l.
2). The highly basic proteins PAP-I, PAP-II and PAP-S are easily coupled
to NHS-activated Sepharose and retain their antigenicity following covalent
linkage to Sepharose.
3). Antibody prepared by affinity chromatography showed the expected
specificity and both cross-reactive and monospecific antibody preparations

could be obtained by affinity chromatography.
4). Monospecific antibody prepared by immunoaffinity chromatography
recognizes the native antigen from crude extracts on Western blots. By this

criterion PAP-S is not found in crude leaf extracts and PAP-I or PAP-II is not
found in crude seed extracts.

5). Cross-reactive antibody ( made against PAP-I ) is able to neutralize
the inhibitory activity of PAP-I, PAP-II or PAP- S in a wheat germ translation
PAP-I and PAP-II are closely related proteins which occur in leaf tissue of

americana and E digida. PAP-I is associated with and isolated from
'spring' leaf tissue, and PAP-II is isolated from 'summer' leaf tissue. The
present report deals with the immuno-detection of these proteins in crude
extracts from leaf tissue utilizing monospecific antibody probes to PAP-I and
PAP-II. From the data presented in Figure 2.9, it is clear that PAP-I (Figure
2.9, panel a) is the dominant protein in leaf tissue collected in March-April.
PAP-II increases over the course of the sampling period (Figure 2.9, panel
b), until PAP-I and PAP-II occur simultaneously in the late summer leaf tissue
(Figures 2.9 and 2.10, panel b). The appearance of PAP-II in late leaf tissue
has been related to an undefined (transcriptional or translational) "switch"
(Lord and Roberts, 1987) in the expression of these two forms. It is clear
that PAP-II increases, but not at the expense of PAP-I. PAP-I remains fairly
constant in its expression ( at least as detected by immunoblot) over the time
course of the sampling period. This methodology cannot distinguish
whether the increase in PAP-II is related to an increased transcriptional rate
or an increase in translational activity. Neither can one eliminate the
possibility of regulated protein degradation as a mechanism for controlling
the relative amounts of these proteins in leaf tissue.
As a percent of total protein, the ratio of PAP-I to PAP-II does decrease,
further confirming the apparent increase in PAP-II. In addition, the

appearance of higher molecular weight species uniquely detected by

monospecific anti-PAP-II is of interest. These components appear to
increase in amount with respect to the sampling period (see Figures 2.9 and
2.10, panels b). Previous work in this laboratory ( Ervin and Preston, 1988,
and see Chapter 3) has identified the presence of putative intermediates of
PAP-II in tissue cultures from e rigida and the presence of these might not
be unexpected in crude extracts from leaf tissue. Boness and Mabry (Dept.
of Botany, Univ. Texas, Austin personal communication) have obtained
information on the appearance of PAP-like proteins of altered
electrophoretic mobility from suspension cultures of E dodecandra. P
dodecandra is reported to produce a protein (dodecandrin) which is
homologous to PAP-I from americana ( Ready et al., 1984), and by
logical extension one that should be homologous to PAP-I from E rigida as
well. dodecandra is not reported to possess a homolog of the PAP-II
from these two other Phvtolaccca species.
The idea has been promulgated in the literature surrounding these
proteins that their association with 'spring' leaf tissue or 'summer' leaf tissue
reflects a temporal sequence of biosynthetic events. However, by analysis
of young leaf tissue from plants which emerged late in the year, we were
able to show that PAP-I is a constitutive part of all leaf tissue, young or old
(Figures 2.9, 2.10 and 2.11 ). PAP-II levels increase over the course of the
growing season, but in young leaf tissue from plants newly emerged late in
the growing season (i.e., November), the levels of PAP-I and PAP-II as
detected by immunoblotting/ELISA are about the same. (see Figures 2.11
and 2.12). The ELISA data in Figure 2.12 have implications for a sampling
strategy designed to elucidate developmental expression of these proteins.
When sampled on a monthly basis (Figure 2.12 a and b), trends in the
expression of PAP-I (Figure 2.12 a) and PAP-II (Figure 2.12 b) are apparent.

However, when trying to define more narrowly the window of expression by

sampling on more closely (four day) spaced intervals for PAP-II no such

trends are apparent. This observation suggests that the expression of PAP-II

in 'summer' leaf tissue is a gradually cumulative effect, and not an all-or-

none event.

The antibodies prepared in this study have been shown to be useful

immunologic probes for the presence of PAP related antigens in tissue

extracts and are thus candidates for extended studies utilizing immunogold
techniques or in polysome isolations.



The pokeweed antiviral proteins, PAP-I, PAP-II and PAP-S are ribosome-
inactivating-proteins (RIPs) which have been isolated from the spring leaves,
summer leaves and seeds respectively, of P. americana (Irvin, 1983; Stirpe
and Barbieri,1986). These three forms of the PAP protein are all basic
proteins with pl values of 8.1, 8.5, and 8.3 and molecular weight values of
29.5 kD, 30.0 kD and 31.0 kD for the PAP-I, PAP-S and PAP-II proteins,
respectively. When isolated from green plant tissue, no carbohydrate is
detectable on any of the mature PAP forms. Similar, if not identical forms
have been isolated and purified from the southern pokeweed, P. rigid
(Preston and Ervin, 1987).
Each PAP protein has nearly the same inhibitory activity toward in vitro
translation systems (Irvin, 1983; Preston and Ervin,1987). The pokeweed
antiviral proteins are typical of other RIPs in that their inhibitory activity is the
result of an N-glycosidase, depurinating a specific site in the 28S rRNA
(Endo et al.,1988; Stirpe et al., 1988). The occurrence of three forms of this
protein, apparently independently transcribed from different genes (Ready et
al., 1984), has provoked my investigations into the developmental
biosynthesis of these proteins. I have established callus cultures from the
spring leaves of P. riida and demonstrate the ability of this callus tissue to
synthesize two forms of the PAP protein, PAP-I and PAP-II, which are
detected by affinity purified antibodies prepared against the proteins

isolated from green leaf tissue. In addition we present evidence for
glycosylated forms of the proteins reactive with anti-PAP-Il.
Immunocytochemical studies were employed to localize proteins bearing
PAP-I or PAP-II epitopes in the callus tissue. The localization of the anti-
PAP-I cross reactive material and anti-PAP-II cross reactive material (crm)
from the callus tissue was examined.
Structural comparisons of the PAP-like proteins from callus tissue with
native leaf proteins were performed by cyanogen bromide mediated
cleavage to establish the identity of the tissue culture proteins with the leaf

Materials and Methods
Chemicals. reagents and media
All chemicals for the preparation of media were ACS certified or
biochemical grade. CM-52 cation exchange resin was from Whatman, the
ion retardation resin AG 11X8 was from BioRad and Sephadex G-75 gel
filtration resin was from Pharmacia. All commercial antibodies and reagents
for ELISA and EITB were from Sigma. PAP-I, PAP-II and PAP-S standard
proteins were isolated and purified from locally growing ri. gida using
minor modifications of published procedures (Irvin, 1975; Barbieri et
al.,1982; Preston and Ervin, 1987). All proteins were prepared by cation
exchange chromatography on CM-52 resin followed by gel filtration on
Sephadex G-75 in 0.1M NH4HCO3. Each protein after purification gave a
single band on SDS-PAGE when detected by Coomassie blue with
apparent molecular weights of 29.0, 30.0, and 31.0 kD for PAP-I, PAP-S and
PAP-II, respectively.

Preparation of monospecific antibodies
Monospecific and cross-reactive antibodies directed against PAP-I,
PAP-II and PAP-S were generated in goats and purified by affinity
chromatography as described previously (Chapter 2).
Preparation of antibody to tissue culture proteins
Proteins were purified from callus tissue as described below. Proteins
bearing PAP-II epitopes were isolated as 44 kD and 34 kD fractions. These
were prepared as 0.1 mg/mL solutions in PBS and injected into chickens.
Antigen solutions were injected subcutaneously at 0 and at 14 days into the
wing (0.7 mL) and into the foot pad (0.3 mL). Antibodies against both tissue
culture proteins were obtained from immunized chicken egg yolks by an
adaptation of the procedure initially described by Jensenius et al. (1981).
Briefly, yolks were isolated from the surrounding fluid and washed gently
with 50 mL of PBS. The washed yolk was disrupted at room temperature by
mixing (magnetic stirrer) in five times the volume of ddw. The pH of the
resulting solution was adjusted by the addition of 1N NaOH to 7.0 after
stirring for 15 min. This solution was frozen and then thawed to room

temperature. After thawing the solution was centrifuged for 30 min at 10,000
rpm in a GSA type rotor in a 150 ml polycarbonate tube. The resulting
supernatant fluid was assayed directly, following filtration through three
layers of cellulose tissue ( KIMWIPE ), by ELISA. Screening of egg yolks by
this rapid freeze-thaw method allowed the identification of eggs which
showed high titers of these antibodies. These eggs were pooled and
processed by the method of Poison et al. (1985) which involves precipitation
of the immunoglobulin fraction from egg yolk with polyethylene glycol.

Analysis of specific fractions by SDS-PAGE was performed in a Hoefer
vertical slab gel apparatus with 4% stacking gel and 12% running
(separating) gel as described by Laemmli (1970). Electrophoresis was
performed at a constant 150 V for 1 h and the voltage was then increased to
250 V for 2.5 h. Gels were removed and stained with 0.1% Coomassie blue
for 1 h and destined in ethanol (95 %): acetic acid: water (60:10:30) until
the background was clear. Alternatively gels were stained with ammoniacal
silver according to the method of Wray et al. (1981). Gels which were to be
utilized in western blotting experiments were placed into transfer buffer
(0.192 M glycine, 0.025 M Tris, pH 8.3, 20 % v/v methanol) for 30 min prior to
Gels which were utilized for the analysis of low molecular weight
peptides generated from the cyanogen bromide hydrolysis of the PAP and
PAP-like proteins from callus were prepared according to the methodology
of Giulian and Graham (1985) and visualized with the silver staining
methodology described by Giulian et al. (1983). As it was difficult to obtain
uniform results with the electrophoretic transfer from these gels, gels to be
utilized in the western blot analyses of these low molecular weight peptides
were prepared according to the methodology described by Christy et al.
(1989). This method involves the addition of 0.1 M NaOAc to the anode
buffer during electrophoresis and is otherwise identical to SDS-PAGE as
described by Laemmli (1970).
Western blot analysis (EITB)
Gels which had been equilibrated in transfer buffer were assembled in
cassettes and then transferred for 1 h at 1 amp (constant current) in a Hoefer
transblot apparatus. Nitrocellulose membrane was from Schleicher and

Schuell (NC-42). After transfer, membranes were incubated in PBS:Tw:BSA
(0.1 %) for 1 h, then in antibody containing solution at the appropriate
dilution for 1 h. Membranes were washed 4 x 15 min in PBS:Tw:BSA and
then incubated in rabbit anti-goat IgG conjugated to peroxidase. Color was
developed from the oxidation of benzamidine-HCI. Alternatively, the blots
were developed by exposure to secondary antibody conjugated to alkaline
phosphatase, in which case the color was developed from NTB
(nitrotetrazolium blue) and BCIP (5-bromo-4-chloro-3-indolyl-phosphate).
Lectin binding to Western blots
The binding of the lectin Concanavalin A (ConA) to proteins following
blotting to nitrocellulose was performed according to the following
procedure. Proteins on SDS-PAGE gels were electrophoretically
transferred to nitrocellulose membranes as described. After transfer, the
blots were incubated in PBS containing gelatin (3 %), 1 mM Ca2+, 1 mM
Mg2+, 0.1 mM Mn2+ (PBSG+) for 1 h. The blots were then placed in a
solution containing ConA (10 ug/mL) in PBSG+ or with ConA plus 0.1 M
alpha methyl-D-mannoside for 1 h at room temperature. Following exposure

to the ConA, the blots were washed 3 x15 min in PBSG+, and then exposed
to goat anti-ConA at the optimal dilution. The blots were incubated for 1 h at
room temperature, washed as described and incubated with rabbit anti-goat
IgG conjugate to alkaline phosphatase at the optimal dilution for 1 h. After
washing, the blot was developed with tetrazolium blue with BCIP as the

substrate. Ovalbumin and ricin A chain (ICN) were included on the gels as
molecular weight markers for glycoproteins.
Enzyme-linked-immuno-sorbent assay (ELISA)
ELISA analysis was performed on dilutions of samples plated as 100 uL

aliquots on Immulon polystyrene plates. The ELISA was a direct, non-

competitive ELISA which allowed for the rapid screening of chromatographic
fractions. (see Chapter 2)

Carbohydrate was determined by the phenol/sulfuric acid method of
Dubois et al. (1956) with the following modifications. Five hundred uL of
sample were mixed with 25 uL of 80% (v/v) phenol in water in a 13 x 100
mm test tube and 1.25 mL of concentrated sulfuric acid was rapidly added .
Glucose was used as standard and all data are presented as gram
equivalents of anhydrohexose. Descending paper chromatography of acid
hydrolysates was utilized to determine the major carbohydrates present.
Hydrolysis of isolated glycoprotein was performed in 3.0 N HCI for 3 h at
1100 C. The hydrolysate was lyophilized repeatedly from distilled water and
then resolubilized in a minimal volume of distilled water. The material was
passed through an ion retardation column and the neutral sugar fraction was
collected, pooled and concentrated by rotoevaporation under vacuum at 400
C. Aliquots were spotted on Whatman No.1 chromatography paper and

chromatographed in pyridine: butanol: water (6: 1: 3) for 24 h. Carbohydrate
was detected by staining the chromatogram with ammoniacal silver nitrate
(Trevelyan et al., 1950).
Glucose was quantified by a coupled enzyme assay utilizing hexokinase
and glucose-6-phosphate dehydrogenase to catalyze the conversion of
glucose to glucose-6-phosphate to 6-phosphogluconate. The reduction of
NADP to NADPH was monitored by measuring the absorbance at 340 nm.
With NADP in excess, the reaction was followed to completion on a

recording spectrophotometer. The stoichiometric conversion of glucose to 6-
phosphogluconate and NADP to NADPH was quantified using a molar

absorptivity of 6.22 mM-lcm-1 for NADPH (CRC Handbook of Biochemistry,

Peptide sequencing
Amino terminal sequence of peptides was performed by automated
Edman degradation utilizing a sequencer (Applied Biosystems) in the
Protein Core Facility Department of Biochemistry and Molecular Biology,
University of Florida. Protein prepared as described below was dialyzed
extensively against distilled water to remove trace amounts of ammonium
bicarbonate and lyophilized. Alternatively, protein was blotted directly to
PVDF membrane (Millipore) following SDS-PAGE and sequenced directly
from the blot (Matsudaira,1987).
Protein synthesis inhibition
The in vitro translation system derived from wheat germ extract
(Amersham) was assayed by the incorporation of 3H-leucine into acid
insoluble product as described previously. The ability of isolated proteins to
inhibit this activity was assayed by preincubating the extract with the protein
sample of interest, and then assaying for activity as described. ( see Chapter
CNBr cleavage of proteins
PAP proteins from leaf tissue and callus were prepared as previously
described. Each protein (1 mg) was dialysed against ddw overnight to
remove salts and then lyophilized. The lyophilized powders were
solubilized in a 0.50 mL volume of 70% (v/v) solution of formic acid. A
measured volume of CNBr in 70% (v/v) formic acid was added so as to yield
a 100-fold excess of CNBr over methionine. The reaction mixture (final
volume < 1.0 mL) was placed in an hydrolysis tube, flushed with nitrogen
and allowed to stand overnight at room temperature. The reaction was

stopped by the addition of ten volumes of ice cold ddw and lyophilized. The
lyophilized product was resolubilized in ddw and lyophilized. This process
was repeated three times.
Callus establishment.
Fresh leaves were collected from locally growing specimens of P. riida
Leaf tissue was washed well in sterile phosphate buffered saline (PBS),
rinsed three times in sterile water and briefly immersed in a 3% solution of
hypochlorite. After several rinses in sterile water, leaf segments were
placed onto a modified Murashige-Skoog (M-S) medium (pH 5.7) which
contained in addition to the M-S salts (Smith, 1986), 2 g/L casein
hydrolysate, 30 g/L sucrose, 0.1 g/L myo-inositol, 2,4-dichloro phenoxy
acetic acid (4.5 uM) kinetin (0.9 uM) and the vitamins nicotinic acid (50
ug/L), thiamine-HCI (40 ug/L) and pyridoxine (50 ug/L). Leaf explants were
maintained in the dark at room temperature. White callus produced from the
edges of the cut tissue was excised and related onto fresh medium. A firm,
viable callus was established after two passages and has since been
maintained on this medium, transferred every fourteen days. Several of the
calli obtained produced a red-pigmented tissue. These were maintained as
a separate line from the white tissue. The red pigmented callus was stable
over the duration of these experiments. Callus cultures of both lines are
shown in Figure 3.1.
Fractionation of proteins.
Callus tissue harvested for these experiments was stored at 40C for four
days. Alternatively, tissue was stored at -200C immediately after harvesting

a b

Figure 3.1. Appearance of two individual lines of callus tissue established
from leaf tissue explants of Eigida. Callus tissue was derived from
wounded leaf tissue explanted to a modified M-S media as described in the
text. Callus was maintained in the dark at room temperature and transferred
aseptically at 14 day intervals.

a. Appearance of a white line of callus tissue from leaf explants of
P. riaida.

b. Appearance of a red line of callus from leaf explants of P.

or used fresh. Callus was ground in a mortar to a slurry and then blended in
a Waring blender for 15 s. The resulting slurry was then subjected to
breakage in a French pressure cell at 500-750 psi and the cell debris
removed by centrifugation at 15,000 rpm in an SS-34 rotor for 30 min. The
supernatant fraction was filtered through Whatman (No. 3 MM) filter paper
and applied to a CM-52 column (15 x 2.7 cm) equilibrated in 5 mM sodium
phosphate buffer, pH 6.7 at room temperature. The column was then eluted
at room temperature with a 450 mL linear gradient from 0 to 0.3 M NaCI in 5
mM sodium phosphate buffer. After elution, the individual fractions were
assayed for absorbance at 280 and 310 nm and reactivity to antibody by
ELISA. Protein containing fractions which were established to be reactive
with either anti-PAP-I or anti- PAP-II were further examined by SDS-PAGE
and EITB analysis.
Fractions identified by ELISA, SDS-PAGE and EITB analysis to contain
either anti -PAP-I or anti-PAP-Il reactive material were pooled. This resulted
in two pools from the CM-52 chromatography, designated Pool 1 (PAP-I like
or anti-PAP-I crm) (0.1 to 0.15 M NaCI) and Pool 2 (PAP-II like or anti-PAP-ll

crm) (0.15 to 0.18 M NaCI). Each pool was individually concentrated at
room temperature to 7.5 mL in a 50 mL Amicon stirred flow cell on a YM-10
membrane at 25 psi. The retentate from this concentration step was then
applied to an SG-75 column which had been previously equilibrated in 0.1
M ammonium bicarbonate buffer (pH 8.9). The column was 140x 2.5 cm and
was eluted at a flow rate of 0.6 mL/min. Fractions obtained from this
chromatographic step were assayed as previously described for absorbance
and reactivity to antibody by ELISA, SDS-PAGE and EITB.

Immunocytochemical studies with callus tissue
Callus tissue of E rigida grown as described was fixed in 3%
glutaraldehyde, rinsed and dehydrated to 70% EtOH in ddw. This fixed
tissue was embedded in LRW plastic, and the plastic hardened at 600C for
24 h. Thin sections (100 A) were obtained and placed onto nickel grids
coated with Formvar film. The grids with sections were placed on a solution
of sodium phosphate buffer (pH 7.2) containing 0.1 % Tween-20 and 1%
ovalbumin (PB:Tw:OA) for 15 min. The grids were then floated on a solution
containing anti-PAP-I (1:5000) or anti-PAP-II (1:5000) for 45 min, and then
rinsed by floating on PB:Tw:OA 3x15 min each. The grids were then floated
on a solution containing rabbit anti-goat IgG for 45 min and then washed as
above. The grids were then floated on a solution of colloidal gold coupled to
Protein-A for 45 min, washed as described, and then blotted dry. The
sections were visualized (without post-staining) for the distribution of gold
label in an HU-11 E transmission electron microscope.

Fractionation of PAP forms from extracts of callus cultures
Figure 3.2 shows the typical elution profile for callus culture extract from
CM-52 cation exchange resin. The monospecific and cross-reactive
antibody preparations allowed me to screen each fraction for material
reactive with antibodies directed against PAP-I, PAP-II or PAP-S. Anti-PAP-
S did not detect any antigens from callus extracts. The reactivity (as
measured by ELISA) with anti-PAP-I (a polyclonal antibody which showed
cross reactivity with PAP-I, II and S) or monospecific anti-PAP-Il IgG of
individual fractions from the CM-52 elution is also demonstrated in Figure
3.2. The cross-reactive anti-PAP-I provided a convenient probe for detecting



co .

aC, as



ri 0

4-" O 0

0 Cc c c


.1 r

0 E


CM C) mCQQ 7

o IO o OIN

0 N 0 -
d d c1-- I--






o o
OJ~ 08z r

Figure 3.3. EITB analysis of fractions from CM-52 chromatography
of pokeweed callus extract. Individual tubes from the CM-52
fractionation of callus extract were selected from peak fractions ( see
Figure 3.2). 200 uL of each sample were boiled for three minutes in
sample buffer and applied to a 12% polyacrylamide gel. Western
blots were prepared as described in the text.

a. Lane 1: Crude extract of callus tissue after homogenization,
breakage and centrifugation, 25 uL; Lanes 2-5: selected fractions
(tube numbers 45, 50, 55 and 60) from Pool 1; Lanes 6-7: selected
fractions (tube numbers 64 and 66) from Pool 2; Lane 8: PAP-I
standard, 4 ug; Lane 9: PAP-II standard, 4 ug The western blot was
probed with anti-PAP-I antibody at a dilution of 1:1000.

b. same as for 2 a except the antibody used in the western blot
was anti-PAP-Il at a dilution of 1:1500. Lanes 1-4 of panel b
correspond to Lanes 6-9 of panel a.

1 2 3 4


44.0 -

30.5 -



1 2 3 4 5 6 7 8 9


34.0 *e

1 2 3 4 5 6 7 8 9 10

Figure 3.4. Western blot analysis of fractions from CM-52
chromatography of tissue culture extract demonstrating the reactivity
of two protein bands with monospecific anti -PAP-II.

The western blot was prepared from 12% polyacrylamide gels as
described. The separation performed here was identical in all
conditions to that described previously in Figure 3.3. Fractions for
SDS-PAGE were chosen from Pool 1 ( containing anti-PAP-I crm) and
from Pool 2 (containing anti-PAP-Il crm). 200 uL of sample was
electrophoresed unless otherwise indicated. Monospecific antibody
to PAP-II was utilized at a dilution of 1:1500.

Lane 1: crude extract of callus tissue after homogenization,
breakage and centrifugation, 25 uL; Lanes 2-5: selected fractions
from Pool 1 ( see chromatogram of Figure 3.2); Lanes 6-9: selected
fractions from Pool 2 ( see chromatogram of Figure 3.2); Lane 10:
PAP-I and PAP-II standards, 4 ug each.

any PAP-related proteins present in the tissue culture extract. The presence
or absence of protein species bearing PAP-II antigens or PAP-S antigens
was then further examined by utilizing monospecific anti-PAP-II or
monospecific anti-PAP-S. Anti-PAP-S detected no antigens in these
analyses. Anti-PAP-I detected a heterogeneous peak of cross- reactive
material eluting with 0.1 M NaCI, and two later peaks at 0.12 to 0.18 M
NaCI. Analysis of individual fractions from the CM-52 chromatography by
ELISA with anti-PAP-Il specific antibody revealed that this antibody reacted
with an early eluting peak at 0.12 to 0.15 M NaCI, corresponding to a similar
pattern of reactivity seen with the anti-PAP-I antibody. In contrast to the peak
detected by the anti-PAP-I antibody at tube number 60, anti-PAP-il showed a
diminished reactivity in this region of the profile. The anti-PAP-ll antibody
reacted with two later eluting components corresponding to a salt

concentration of 0.18 M to 0.20 M (tube numbers 68-80).
Figure 3.3 demonstrates the reactivity of specific fractions from the CM-
52 chromatography when analyzed by EITB. Figure 3.3 a shows the
material reactive with cross-reactive anti-PAP-I antibody and Figure 3.2 b

shows the material reactive with monospecific anti-PAP-Il antibody. The

specificity of this antibody for the PAP-II protein is evident from this blot.

Lane 3, panel b, shows no reactivity of this antibody with PAP-I compared to
the analogous lane 8 in panel a.
This analysis indicates that the peak eluting with 0.1 M NaCI, reactive
with anti-PAP-I but not anti-PAP-ll, (lanes 2-4, Fig.3.2 a) is composed of

several discrete components of high molecular weight. A component with a

molecular mass of 29.5 kD (lane 4-5, Fig. 3.3 a) eluted with 0.14 M NaCI and
was nearly identical in size to native PAP-I isolated from spring leaf tissue.

The material eluting later in the gradient at 0.18 M NaCI was comprised of a

single band, evident in this analysis which migrated with an apparent mass
of 44 kD. Subsequent analyses demonstrated that two proteins were
contained in Pool 2, one with an apparent mass of 44 kD and the other with
a mass of 34 kD. This is demonstrated in the Western blot depicted in Figure
3.4. Figure 3.4 is analogous to Figure 3.3 in that it depicts discrete fractions
from a CM-52 fractionation of tissue culture extract subjected to SDS-PAGE
and western blotting. The appearance of two diffuse protein species present
in Pool 2 from CM-52 chromatography when detected by monospecific anti-
PAP-II antibody is shown in Figure 3.4. These two bands were uniquely
detected by monospecific anti-PAP-II (Fig. 3. 3b, lanes 1,2; Figure 3.4, lanes
8-9) while no band was detected at the position expected for native PAP-II
(lane 4, Figure 3.3 b) isolated from summer leaf tissue. Note that the diffuse
nature of this material after SDS-PAGE and Western blot analysis makes it
difficult to assign a discrete molecular weight. This higher apparent
molecular weight for the anti-PAP-Il crms and their diffuse appearance on
SDS gels when compared to PAP-II from summer leaf tissue suggested that
they might represent a form of PAP-II modified by glycosylation and/or
possessing additional peptide sequence. Carbohydrate was present in the
fractions comprising this peak (see below).
The fractions from the CM-52 chromatography containing anti-PAP-I
reactive material were pooled and designated, Pool 1. Those fractions
reactive with anti-PAP-Il were pooled and designated, Pool 2. These pools
were independently concentrated to 7.5 mL and applied to an SG-75
column and eluted with 0.1 M NH4CO3. The results from the gel filtration of
Pool 1 and Pool 2 are given in Figure 3.5 a and Figure 3.5 b, respectively.
As is evident from Figure 3.5 a, gel chromatography on SG-75 resolved Pool
1 into 5 peaks identified by A 280g

Figure 3.5. SG-75 gel sieving chromatography elution profiles for
Pool 1 ( containing anti-PAP-I cross-reactive proteins) and Pool 2
(containing anti-PAP-ll cross-reactive proteins) obtained by CM-52
chromatography of callus tissue extract as described. The column
(140 x 2.5 cm) was developed with 0.1 M NH4C03 (pH 8.6) at a flow
rate of 0.6 mL/min. 6.0 mL fractions were collected and assayed for
A280 (protein) and A405 (ELISA) as described in the text.

--0-0-0- (Protein, A 280)
-o-o- ( ELISA reactivity (A405); anti-PAP-I, 1:1000)
( ELISA reactivity (A405) ; anti-PAP-II, 1:1500)

a. SG-75 gel sieving chromatography elution profile of Pool 2

b. SG-75 gel sieving chromatography elution profile of Pool 1



50 70









Figure 3.6. EITB analysis of fractions from SG-75 chromatography of
Pool 1 and Pool 2 from CM-52 chromatography of callus extract.

a. Lanes 1-3: tubes 51-53 (peak 1) from SG-75 gel
chromatography of Pool 2 (anti-PAP-ll crm, see Figure 3.2); Lanes 4-
6: tubes 62-64 (peak 2) from SG-75 gel chromatography of Pool 2;
Lanes 7-9: tubes 76-78 (peak 4, anti-PAP-I crm) from SG-75 gel
chromatography of Pool 1; Lane 10: PAP-I and PAP-II standards of 4
ug each. The blots were prepared as described. Anti-PAP-I antibody
cross-reactive with PAP-II and PAP-S was utilized at a dilution of
1:1000 to screen extracts.

b. Same as for the above except that the blot was developed with
anti-PAP-ll antibody (monospecific) at a dilution of 1:1000. Blots were
prepared as described previously.



44.0 -




1 2 3 4 5 6

7 8 9 10

I -

" 3Q4B6 1.

1 2 3 4 5 6 7 8 9 10

Peak 4 corresponds to PAP-I with an apparent molecular weight of 29.5 kD
(based on elution position from SG-75) while peaks 1-3 are higher
molecular weight species which are reactive with this anti-PAP-I antibody.
Pool 2 (Figure 3.5 b) containing anti-PAP-Il reactive material, was resolved
into three distinct peaks on SG-75, two of which were reactive with anti-PAP-
II antibody.
Figure 3.6 shows the reactivity of specific fractions from SG-75 of Pool 1
and Pool 2 when analyzed by EITB. Figure 3.6a demonstrates the reactivity
to the anti-PAP-I antibody and Figure 3.6b shows the reactivity to the anti-
PAP-II antibody.
Chemical properties

Both peak fractions 1 and 2 from Pool 2 and 1-4 from Pool 1 were
assayed for total carbohydrate. Table 3.1 shows the relative amounts of
protein and carbohydrate for the various fractions. The carbohydrate
associated with the two anti-PAP-Il cross reactive proteins was further
characterized by paper chromatography. These two compounds were
individually hydrolysed in mineral acid for one h, and the hydrolysate was

chromatographed as described. The mobilities of the released

monosaccharides were compared to the Rfs of pure standards of glucose
and mannose, which indicated the presence of glucose and mannose as
the only saccharides (data not shown). On the basis of relative spot intensity

glucose accounted for greater than 90% of the total carbohydrate. The
glucose was further quantified by an enzyme coupled analysis as described.
By this analysis, glucose accounted for only 15-25% of the total CH20

detected by the phenol-sulfuric acid assay which raised the question of the
accuracy of the colorimetric assay when applied to these glycoproteins.

To identify specific carbohydrate containing proteins, the reactivity of these
fractions to the lectin ConA was examined on Western blots. Individual
fractions from both CM-52 and SG-75 chromatography were shown to be
reactive with ConA, following EITB. Figure 3.7 shows the results of one
such analysis. Ovalbumin and RTA (lanes 1 and 2) are strongly reactive
while PAP-I and PAP-II (lane 3) are not reactive, indicating that the
procedure does correctly detect glycoproteins after electrophoretic transfer.
Lanes 4-6 represent fractions from CM-52 ion exchange chromatography
which correspond to fractions reactive with anti-PAP-I (see Figure 3.3 a). By
this analysis one may infer the presence of glycosylated, ConA reactive
species of relatively high molecular weight (greater than 30 kD with a
prominent band at 50 kD), bearing anti-PAP-I epitopes. Lanes 7-10
correspond to CM-52 fractions reactive with anti-PAP-ll, while lane 11 is a
concentrated pool of anti-PAP-ll crm with a mobility corresponding to a
polypeptide with a molecular weight of 34 kD. The anti-PAP-ll crm of 34 kD
(lanes 7, 8 ) is clearly reactive with ConA, while the material with a greater
molecular weight of 44 kD (lanes 9,10) is detected as a blanched (negative
staining) component apparently giving an anomalous reaction with ConA
and/or anti-ConA-alkaline phosphatase. Such a negative staining around a
reactive species by a lectin blot analysis has been observed by Faye and
Chrispeels (1985) and is ascribed to an overabundance of the transferred
protein. Lane 12 is a concentrated Pool I (containing anti-PAP-I crm from
CM-52 fractionation ) prior to SG-75 gel sieving chromatography while lanes
13 and 15 are purified anti-PAP-I crm of molecular weight 29.5 kD after SG-
75 gel sieving. These lanes correspond to peak 4 Figure 3.4 a, which is
strongly reactive with anti-PAP-I. This PAP-I like material is clearly non-
reactive with ConA except for high molecular weight components which may

Table 3.1. Carbohydrate and protein
callus tissue.

composition of PAP proteins from

Starting Material Protein Glucose CH20/Protein
(mg) equivalents (gm/gm)

Pool 1 CM-52 3.38 3.04 0.99
Fraction SG-75 0.39 0.69 1.7
Fraction2 SG-75 0.39 0.42 1.1
Fraction3 SG-75 0.51 0.45 0.88
Fraction4 SG-75 1.62 0.15 0.09
Pool 2 CM-52 8.00 11.34 1.4
Fraction SG-75 2.07 3.09 1.49
Fraction2 SG-75 2.25 0.81 0.36